diff --git a/sft_full/hyperrouter/added_tokens.json b/sft_full/hyperrouter/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d3d3a1b74d87e381e471f7b33784015d2dc0ea --- /dev/null +++ b/sft_full/hyperrouter/added_tokens.json @@ -0,0 +1,13 @@ +{ + "<|assistant|>": 32001, + "<|endoftext|>": 32000, + "<|end|>": 32007, + "<|placeholder1|>": 32002, + "<|placeholder2|>": 32003, + "<|placeholder3|>": 32004, + "<|placeholder4|>": 32005, + "<|placeholder5|>": 32008, + "<|placeholder6|>": 32009, + "<|system|>": 32006, + "<|user|>": 32010 +} diff --git a/sft_full/hyperrouter/config.json b/sft_full/hyperrouter/config.json new file mode 100644 index 0000000000000000000000000000000000000000..46dae14670cc46d1f0d913a9496d7b3ef53c103c --- /dev/null +++ b/sft_full/hyperrouter/config.json @@ -0,0 +1,170 @@ +{ + "_name_or_path": "/cm/archive/namnv78/checkpoints/phi35-siglip224/pft", + "architectures": [ + "LlavaPhiForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "auto_map": { + "AutoConfig": "configuration_phi3.Phi3Config", + "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM" + }, + "balance_loss_coef": 0.1, + "bos_token_id": 1, + "clip_smoe": true, + "dropout": false, + "embd_pdrop": 0.0, + "eos_token_id": 32000, + "freeze_mm_mlp_adapter": false, + "hidden_act": "silu", + "hidden_size": 3072, + "image_aspect_ratio": "pad", + "initializer_range": 0.02, + "intermediate_size": 8192, + "local_rank": 0, + "max_position_embeddings": 131072, + "mlp_smoe": true, + "mm_hidden_size": 1152, + "mm_patch_merge_type": "flat", + "mm_projector_lr": null, + "mm_projector_type": "moe", + "mm_use_im_patch_token": false, + "mm_use_im_start_end": false, + "mm_vision_select_feature": "patch", + "mm_vision_select_layer": -2, + "mm_vision_tower": "google/siglip-so400m-patch14-224", + "model_type": "llava_phi", + "moe_name": "hyperrouter", + "num_attention_heads": 32, + "num_experts": 4, + "num_hidden_layers": 32, + "num_key_value_heads": 32, + "num_layers": 3, + "num_selected": 2, + "original_max_position_embeddings": 4096, + "pad_token_id": 32000, + "resid_pdrop": 0.0, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "long_factor": [ + 1.0800000429153442, + 1.1100000143051147, + 1.1399999856948853, + 1.340000033378601, + 1.5899999141693115, + 1.600000023841858, + 1.6200000047683716, + 2.620000123977661, + 3.2300000190734863, + 3.2300000190734863, + 4.789999961853027, + 7.400000095367432, + 7.700000286102295, + 9.09000015258789, + 12.199999809265137, + 17.670000076293945, + 24.46000099182129, + 28.57000160217285, + 30.420001983642578, + 30.840002059936523, + 32.590003967285156, + 32.93000411987305, + 42.320003509521484, + 44.96000289916992, + 50.340003967285156, + 50.45000457763672, + 57.55000305175781, + 57.93000411987305, + 58.21000289916992, + 60.1400032043457, + 62.61000442504883, + 62.62000274658203, + 62.71000289916992, + 63.1400032043457, + 63.1400032043457, + 63.77000427246094, + 63.93000411987305, + 63.96000289916992, + 63.970001220703125, + 64.02999877929688, + 64.06999969482422, + 64.08000183105469, + 64.12000274658203, + 64.41000366210938, + 64.4800033569336, + 64.51000213623047, + 64.52999877929688, + 64.83999633789062 + ], + "short_factor": [ + 1.0, + 1.0199999809265137, + 1.0299999713897705, + 1.0299999713897705, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0699999332427979, + 1.0999999046325684, + 1.1099998950958252, + 1.1599998474121094, + 1.1599998474121094, + 1.1699998378753662, + 1.2899998426437378, + 1.339999794960022, + 1.679999828338623, + 1.7899998426437378, + 1.8199998140335083, + 1.8499997854232788, + 1.8799997568130493, + 1.9099997282028198, + 1.9399996995925903, + 1.9899996519088745, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0799996852874756, + 2.0899996757507324, + 2.189999580383301, + 2.2199995517730713, + 2.5899994373321533, + 2.729999542236328, + 2.749999523162842, + 2.8399994373321533 + ], + "type": "longrope" + }, + "rope_theta": 10000.0, + "router_z_loss_coef": 0.01, + "scales": [ + 1, + 3 + ], + "sliding_window": 262144, + "tie_word_embeddings": false, + "tokenizer_model_max_length": 2048, + "tokenizer_padding_side": "right", + "topk_max": 2, + "topk_min": 1, + "torch_dtype": "bfloat16", + "training": true, + "transformers_version": "4.43.0", + "tune_mm_mlp_adapter": false, + "use_cache": true, + "use_mm_proj": true, + "vocab_size": 32064 +} diff --git a/sft_full/hyperrouter/generation_config.json b/sft_full/hyperrouter/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..dad5c4578f0dc5969b38755d095fc30c368bb54a --- /dev/null +++ b/sft_full/hyperrouter/generation_config.json @@ -0,0 +1,12 @@ +{ + "_from_model_config": true, + "bos_token_id": 1, + "do_sample": true, + "eos_token_id": [ + 32007, + 32001, + 32000 + ], + "pad_token_id": 32000, + "transformers_version": "4.43.0" +} diff --git a/sft_full/hyperrouter/model-00001-of-00003.safetensors b/sft_full/hyperrouter/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3a64420456184c6745f020f1ec3ae7dee7d05354 --- /dev/null +++ b/sft_full/hyperrouter/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77070ba911bb125ab3cd8424cd59c16fe002d181caf04bf90cfe15a51301c6b4 +size 4972489328 diff --git a/sft_full/hyperrouter/model-00002-of-00003.safetensors b/sft_full/hyperrouter/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ace25bce91022edc2e1c10e7cbc8acfc33013c4f --- /dev/null +++ b/sft_full/hyperrouter/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c92aba44ae28d81e099be78e65ce61b5095566a4646d0882415ff4cc63270f08 +size 4995022432 diff --git a/sft_full/hyperrouter/model-00003-of-00003.safetensors b/sft_full/hyperrouter/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7c7c12f9e0d22de2dddb385d5973f82fcf0b1064 --- /dev/null +++ b/sft_full/hyperrouter/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fdcfe6f0f1bf56818412c4eec4443cbe87476db39f0f4782d0f29247dc905d28 +size 342468696 diff --git a/sft_full/hyperrouter/model.safetensors.index.json b/sft_full/hyperrouter/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..db030b0a892103170e46b0e3031402e452e46911 --- /dev/null +++ b/sft_full/hyperrouter/model.safetensors.index.json @@ -0,0 +1,1117 @@ +{ + "metadata": { + "total_size": 10309820576 + }, + "weight_map": { + "lm_head.weight": "model-00003-of-00003.safetensors", + "model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.hypernet.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.hypernet.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.hypernet.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.hypernet.2.weight": "model-00003-of-00003.safetensors", + "model.norm.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.position_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.hyper_embedding": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.hypernet.0.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.hypernet.0.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.hypernet.2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.hypernet.2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors" + } +} diff --git a/sft_full/hyperrouter/special_tokens_map.json b/sft_full/hyperrouter/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..3e4d5a5bc1cb51753cc9ae0305ece0da60052b10 --- /dev/null +++ b/sft_full/hyperrouter/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/sft_full/hyperrouter/tokenizer.model b/sft_full/hyperrouter/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/sft_full/hyperrouter/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/sft_full/hyperrouter/tokenizer_config.json b/sft_full/hyperrouter/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d579bb0b91b24b214ea3c2e487e27a65017cdc4a --- /dev/null +++ b/sft_full/hyperrouter/tokenizer_config.json @@ -0,0 +1,132 @@ +{ + "add_bos_token": false, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": false + }, + "32000": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "32001": { + "content": "<|assistant|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32002": { + "content": "<|placeholder1|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32003": { + "content": "<|placeholder2|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32004": { + "content": "<|placeholder3|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32005": { + "content": "<|placeholder4|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32006": { + "content": "<|system|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32007": { + "content": "<|end|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32008": { + "content": "<|placeholder5|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32009": { + "content": "<|placeholder6|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32010": { + "content": "<|user|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "chat_template": "{% for message in messages %}{% if message['role'] == 'system' and message['content'] %}{{'<|system|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'user' %}{{'<|user|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>\n' + message['content'] + '<|end|>\n'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}", + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "legacy": false, + "model_max_length": 2048, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +} diff --git a/sft_full/hyperrouter/trainer_state.json b/sft_full/hyperrouter/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..8fbaf3f8ace19ecd751d2addbd53c78c9826f923 --- /dev/null +++ b/sft_full/hyperrouter/trainer_state.json @@ -0,0 +1,99835 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999549082382648, + "eval_steps": 500, + "global_step": 11088, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 9.018352347026198e-05, + "flos": 20416263415680.0, + "grad_norm": 57.00024842325251, + "learning_rate": 0.0, + "loss": 2.0008, + "num_input_tokens_seen": 26480, + "step": 1 + }, + { + "epoch": 0.00018036704694052397, + "flos": 22494810728160.0, + "grad_norm": 54.76518143543624, + "learning_rate": 4.773623799730706e-07, + "loss": 1.9301, + "num_input_tokens_seen": 52215, + "step": 2 + }, + { + "epoch": 0.0002705505704107859, + "flos": 23986593935040.0, + "grad_norm": 34.817962710987985, + "learning_rate": 7.566014715123208e-07, + "loss": 1.6523, + "num_input_tokens_seen": 81590, + "step": 3 + }, + { + "epoch": 0.00036073409388104793, + "flos": 22241235660000.0, + "grad_norm": 55.03509622419361, + "learning_rate": 9.547247599461412e-07, + "loss": 1.9717, + "num_input_tokens_seen": 108995, + "step": 4 + }, + { + "epoch": 0.0004509176173513099, + "flos": 26064784673760.0, + "grad_norm": 40.07419571381783, + "learning_rate": 1.108401121501769e-06, + "loss": 1.5705, + "num_input_tokens_seen": 140930, + "step": 5 + }, + { + "epoch": 0.0005411011408215718, + "flos": 15355606232160.0, + "grad_norm": 59.82754961854425, + "learning_rate": 1.2339638514853914e-06, + "loss": 1.7504, + "num_input_tokens_seen": 168255, + "step": 6 + }, + { + "epoch": 0.0006312846642918339, + "flos": 24900518654880.0, + "grad_norm": 35.35920411193792, + "learning_rate": 1.3401256270225321e-06, + "loss": 1.4796, + "num_input_tokens_seen": 200695, + "step": 7 + }, + { + "epoch": 0.0007214681877620959, + "flos": 15538138421760.0, + "grad_norm": 27.048388037185436, + "learning_rate": 1.4320871399192119e-06, + "loss": 1.4154, + "num_input_tokens_seen": 226405, + "step": 8 + }, + { + "epoch": 0.0008116517112323579, + "flos": 21547055793120.0, + "grad_norm": 16.10454924385207, + "learning_rate": 1.5132029430246416e-06, + "loss": 1.2858, + "num_input_tokens_seen": 253615, + "step": 9 + }, + { + "epoch": 0.0009018352347026198, + "flos": 20742984914880.0, + "grad_norm": 16.46113030855282, + "learning_rate": 1.5857635014748399e-06, + "loss": 1.3977, + "num_input_tokens_seen": 279335, + "step": 10 + }, + { + "epoch": 0.0009920187581728818, + "flos": 11639569365120.0, + "grad_norm": 17.9625583761882, + "learning_rate": 1.6514025108267924e-06, + "loss": 1.4724, + "num_input_tokens_seen": 304145, + "step": 11 + }, + { + "epoch": 0.0010822022816431437, + "flos": 22423693510080.0, + "grad_norm": 14.815109087529212, + "learning_rate": 1.711326231458462e-06, + "loss": 1.353, + "num_input_tokens_seen": 333970, + "step": 12 + }, + { + "epoch": 0.0011723858051134058, + "flos": 16557568121280.0, + "grad_norm": 7.849771931244792, + "learning_rate": 1.7664507107987104e-06, + "loss": 1.1956, + "num_input_tokens_seen": 356980, + "step": 13 + }, + { + "epoch": 0.0012625693285836677, + "flos": 24209498217600.0, + "grad_norm": 5.595695455655739, + "learning_rate": 1.8174880069956024e-06, + "loss": 1.1579, + "num_input_tokens_seen": 386575, + "step": 14 + }, + { + "epoch": 0.0013527528520539298, + "flos": 22747493722080.0, + "grad_norm": 4.876320880060846, + "learning_rate": 1.8650025930140899e-06, + "loss": 1.1153, + "num_input_tokens_seen": 414945, + "step": 15 + }, + { + "epoch": 0.0014429363755241917, + "flos": 15574607689920.0, + "grad_norm": 5.614686454245882, + "learning_rate": 1.9094495198922823e-06, + "loss": 1.2292, + "num_input_tokens_seen": 441905, + "step": 16 + }, + { + "epoch": 0.0015331198989944536, + "flos": 24208903501440.0, + "grad_norm": 4.60576920503978, + "learning_rate": 1.9512009899507514e-06, + "loss": 1.0754, + "num_input_tokens_seen": 470450, + "step": 17 + }, + { + "epoch": 0.0016233034224647158, + "flos": 22606411548480.0, + "grad_norm": 4.33890202534264, + "learning_rate": 1.990565322997712e-06, + "loss": 1.1294, + "num_input_tokens_seen": 499605, + "step": 18 + }, + { + "epoch": 0.0017134869459349777, + "flos": 37577848144800.0, + "grad_norm": 3.492866057624997, + "learning_rate": 2.027800787770518e-06, + "loss": 1.0404, + "num_input_tokens_seen": 532040, + "step": 19 + }, + { + "epoch": 0.0018036704694052396, + "flos": 24391138332960.0, + "grad_norm": 4.874433044330603, + "learning_rate": 2.06312588144791e-06, + "loss": 1.1335, + "num_input_tokens_seen": 560030, + "step": 20 + }, + { + "epoch": 0.0018938539928755017, + "flos": 24860109392160.0, + "grad_norm": 3.326765328921013, + "learning_rate": 2.0967270985348526e-06, + "loss": 1.0392, + "num_input_tokens_seen": 586835, + "step": 21 + }, + { + "epoch": 0.0019840375163457636, + "flos": 28436513706240.0, + "grad_norm": 5.329131179623811, + "learning_rate": 2.128764890799863e-06, + "loss": 0.9558, + "num_input_tokens_seen": 617945, + "step": 22 + }, + { + "epoch": 0.0020742210398160257, + "flos": 36487189717920.0, + "grad_norm": 3.3794795790234655, + "learning_rate": 2.1593783012990145e-06, + "loss": 0.9578, + "num_input_tokens_seen": 649070, + "step": 23 + }, + { + "epoch": 0.0021644045632862874, + "flos": 21622113005760.0, + "grad_norm": 2.782376264921038, + "learning_rate": 2.188688611431533e-06, + "loss": 1.1077, + "num_input_tokens_seen": 675560, + "step": 24 + }, + { + "epoch": 0.0022545880867565495, + "flos": 23735026033920.0, + "grad_norm": 3.2216965604665764, + "learning_rate": 2.216802243003538e-06, + "loss": 1.0862, + "num_input_tokens_seen": 703870, + "step": 25 + }, + { + "epoch": 0.0023447716102268116, + "flos": 17869644040320.0, + "grad_norm": 4.520480703263119, + "learning_rate": 2.243813090771781e-06, + "loss": 0.9281, + "num_input_tokens_seen": 733050, + "step": 26 + }, + { + "epoch": 0.0024349551336970737, + "flos": 21476087254080.0, + "grad_norm": 4.98197740630697, + "learning_rate": 2.269804414536962e-06, + "loss": 1.0944, + "num_input_tokens_seen": 758950, + "step": 27 + }, + { + "epoch": 0.0025251386571673354, + "flos": 31203977903520.0, + "grad_norm": 5.810245856173359, + "learning_rate": 2.2948503869686733e-06, + "loss": 1.0051, + "num_input_tokens_seen": 791220, + "step": 28 + }, + { + "epoch": 0.0026153221806375975, + "flos": 63704303184000.0, + "grad_norm": 2.5243527668755172, + "learning_rate": 2.3190173696980436e-06, + "loss": 0.6974, + "num_input_tokens_seen": 884040, + "step": 29 + }, + { + "epoch": 0.0027055057041078597, + "flos": 23587216133760.0, + "grad_norm": 2.4685266128103778, + "learning_rate": 2.3423649729871604e-06, + "loss": 1.0075, + "num_input_tokens_seen": 910420, + "step": 30 + }, + { + "epoch": 0.0027956892275781214, + "flos": 18817361805600.0, + "grad_norm": 2.143913249481254, + "learning_rate": 2.364946941580084e-06, + "loss": 1.0044, + "num_input_tokens_seen": 937625, + "step": 31 + }, + { + "epoch": 0.0028858727510483835, + "flos": 14983330919520.0, + "grad_norm": 2.6855840728957925, + "learning_rate": 2.3868118998653532e-06, + "loss": 1.0676, + "num_input_tokens_seen": 962050, + "step": 32 + }, + { + "epoch": 0.0029760562745186456, + "flos": 70229663131680.0, + "grad_norm": 2.2868821106930737, + "learning_rate": 2.408003982339113e-06, + "loss": 0.6433, + "num_input_tokens_seen": 1059370, + "step": 33 + }, + { + "epoch": 0.0030662397979889073, + "flos": 24536606538240.0, + "grad_norm": 2.1916661507880275, + "learning_rate": 2.4285633699238223e-06, + "loss": 0.9158, + "num_input_tokens_seen": 1088155, + "step": 34 + }, + { + "epoch": 0.0031564233214591694, + "flos": 48399556650720.0, + "grad_norm": 4.269600493348615, + "learning_rate": 2.4485267485243007e-06, + "loss": 0.9468, + "num_input_tokens_seen": 1120235, + "step": 35 + }, + { + "epoch": 0.0032466068449294315, + "flos": 22492060165920.0, + "grad_norm": 1.8521889701156484, + "learning_rate": 2.467927702970783e-06, + "loss": 0.9835, + "num_input_tokens_seen": 1148730, + "step": 36 + }, + { + "epoch": 0.003336790368399693, + "flos": 25296365328960.0, + "grad_norm": 2.160964014603693, + "learning_rate": 2.4867970569753584e-06, + "loss": 1.0025, + "num_input_tokens_seen": 1176690, + "step": 37 + }, + { + "epoch": 0.0034269738918699553, + "flos": 18118758737280.0, + "grad_norm": 3.2039539288031387, + "learning_rate": 2.5051631677435883e-06, + "loss": 0.8433, + "num_input_tokens_seen": 1202550, + "step": 38 + }, + { + "epoch": 0.0035171574153402174, + "flos": 39290156769600.0, + "grad_norm": 2.8976393266411478, + "learning_rate": 2.523052182311031e-06, + "loss": 0.8118, + "num_input_tokens_seen": 1237825, + "step": 39 + }, + { + "epoch": 0.003607340938810479, + "flos": 23950236176160.0, + "grad_norm": 2.955931606446148, + "learning_rate": 2.540488261420981e-06, + "loss": 1.0162, + "num_input_tokens_seen": 1267345, + "step": 40 + }, + { + "epoch": 0.0036975244622807412, + "flos": 24828249174240.0, + "grad_norm": 2.220288777342096, + "learning_rate": 2.557493775753984e-06, + "loss": 0.8151, + "num_input_tokens_seen": 1298280, + "step": 41 + }, + { + "epoch": 0.0037877079857510034, + "flos": 25155134476320.0, + "grad_norm": 1.83756745402769, + "learning_rate": 2.5740894785079235e-06, + "loss": 0.9309, + "num_input_tokens_seen": 1327675, + "step": 42 + }, + { + "epoch": 0.0038778915092212655, + "flos": 24499802742240.0, + "grad_norm": 2.498780239088275, + "learning_rate": 2.5902946576685834e-06, + "loss": 0.9341, + "num_input_tokens_seen": 1356790, + "step": 43 + }, + { + "epoch": 0.003968075032691527, + "flos": 25156286738880.0, + "grad_norm": 2.4684913596345304, + "learning_rate": 2.606127270772933e-06, + "loss": 0.8622, + "num_input_tokens_seen": 1384125, + "step": 44 + }, + { + "epoch": 0.004058258556161789, + "flos": 28253200951680.0, + "grad_norm": 2.7506361138956374, + "learning_rate": 2.62160406452641e-06, + "loss": 0.8348, + "num_input_tokens_seen": 1414960, + "step": 45 + }, + { + "epoch": 0.004148442079632051, + "flos": 19145213521440.0, + "grad_norm": 2.10890961210875, + "learning_rate": 2.636740681272085e-06, + "loss": 0.9975, + "num_input_tokens_seen": 1442965, + "step": 46 + }, + { + "epoch": 0.004238625603102313, + "flos": 22312464387360.0, + "grad_norm": 2.124738452747591, + "learning_rate": 2.651551754008722e-06, + "loss": 1.0333, + "num_input_tokens_seen": 1471250, + "step": 47 + }, + { + "epoch": 0.004328809126572575, + "flos": 19253469063360.0, + "grad_norm": 2.093272254299793, + "learning_rate": 2.6660509914046035e-06, + "loss": 0.9696, + "num_input_tokens_seen": 1498710, + "step": 48 + }, + { + "epoch": 0.004418992650042837, + "flos": 32801303259840.0, + "grad_norm": 3.2826028124127444, + "learning_rate": 2.6802512540450642e-06, + "loss": 0.9285, + "num_input_tokens_seen": 1531230, + "step": 49 + }, + { + "epoch": 0.004509176173513099, + "flos": 22678681029120.0, + "grad_norm": 2.130306462372658, + "learning_rate": 2.694164622976609e-06, + "loss": 0.9148, + "num_input_tokens_seen": 1558370, + "step": 50 + }, + { + "epoch": 0.0045993596969833616, + "flos": 19181608450080.0, + "grad_norm": 1.830559633598101, + "learning_rate": 2.707802461463072e-06, + "loss": 0.8516, + "num_input_tokens_seen": 1586025, + "step": 51 + }, + { + "epoch": 0.004689543220453623, + "flos": 22349639880960.0, + "grad_norm": 1.9170224716429534, + "learning_rate": 2.7211754707448516e-06, + "loss": 0.8946, + "num_input_tokens_seen": 1612485, + "step": 52 + }, + { + "epoch": 0.004779726743923885, + "flos": 20633019564000.0, + "grad_norm": 1.9349332851175605, + "learning_rate": 2.734293740486721e-06, + "loss": 0.8748, + "num_input_tokens_seen": 1641400, + "step": 53 + }, + { + "epoch": 0.0048699102673941475, + "flos": 24498092933280.0, + "grad_norm": 2.210801838193833, + "learning_rate": 2.747166794510033e-06, + "loss": 0.9196, + "num_input_tokens_seen": 1671910, + "step": 54 + }, + { + "epoch": 0.004960093790864409, + "flos": 28323909302400.0, + "grad_norm": 1.9792441168906416, + "learning_rate": 2.759803632328562e-06, + "loss": 0.8877, + "num_input_tokens_seen": 1701190, + "step": 55 + }, + { + "epoch": 0.005050277314334671, + "flos": 27633223392960.0, + "grad_norm": 2.236156473693491, + "learning_rate": 2.772212766941744e-06, + "loss": 1.0072, + "num_input_tokens_seen": 1729240, + "step": 56 + }, + { + "epoch": 0.005140460837804933, + "flos": 23262077810400.0, + "grad_norm": 2.6314084981629327, + "learning_rate": 2.7844022592828385e-06, + "loss": 0.9949, + "num_input_tokens_seen": 1756890, + "step": 57 + }, + { + "epoch": 0.005230644361275195, + "flos": 26172780027360.0, + "grad_norm": 2.030030926435944, + "learning_rate": 2.7963797496711145e-06, + "loss": 0.8476, + "num_input_tokens_seen": 1785850, + "step": 58 + }, + { + "epoch": 0.005320827884745457, + "flos": 22896121356960.0, + "grad_norm": 2.077043983548429, + "learning_rate": 2.80815248657541e-06, + "loss": 0.9385, + "num_input_tokens_seen": 1813755, + "step": 59 + }, + { + "epoch": 0.005411011408215719, + "flos": 66614856721920.0, + "grad_norm": 1.3836573374528762, + "learning_rate": 2.819727352960231e-06, + "loss": 0.6167, + "num_input_tokens_seen": 1915265, + "step": 60 + }, + { + "epoch": 0.005501194931685981, + "flos": 21583450721760.0, + "grad_norm": 2.2149348164562093, + "learning_rate": 2.8311108904541717e-06, + "loss": 0.8974, + "num_input_tokens_seen": 1942385, + "step": 61 + }, + { + "epoch": 0.005591378455156243, + "flos": 28034236663680.0, + "grad_norm": 1.8451566987316133, + "learning_rate": 2.842309321553155e-06, + "loss": 0.8448, + "num_input_tokens_seen": 1973990, + "step": 62 + }, + { + "epoch": 0.005681561978626505, + "flos": 25262237755680.0, + "grad_norm": 1.6805959684418699, + "learning_rate": 2.8533285700471737e-06, + "loss": 0.831, + "num_input_tokens_seen": 2004565, + "step": 63 + }, + { + "epoch": 0.005771745502096767, + "flos": 25077029343360.0, + "grad_norm": 2.3634519405042367, + "learning_rate": 2.8641742798384237e-06, + "loss": 0.9689, + "num_input_tokens_seen": 2032840, + "step": 64 + }, + { + "epoch": 0.005861929025567029, + "flos": 23261297245440.0, + "grad_norm": 1.9346551005208115, + "learning_rate": 2.874851832300479e-06, + "loss": 0.925, + "num_input_tokens_seen": 2060190, + "step": 65 + }, + { + "epoch": 0.005952112549037291, + "flos": 16704002740320.0, + "grad_norm": 2.1514117374387585, + "learning_rate": 2.8853663623121834e-06, + "loss": 0.9263, + "num_input_tokens_seen": 2084785, + "step": 66 + }, + { + "epoch": 0.006042296072507553, + "flos": 18999262109280.0, + "grad_norm": 3.6662288028638814, + "learning_rate": 2.895722773085839e-06, + "loss": 0.9538, + "num_input_tokens_seen": 2112460, + "step": 67 + }, + { + "epoch": 0.0061324795959778146, + "flos": 28763436178080.0, + "grad_norm": 3.776492570285668, + "learning_rate": 2.905925749896893e-06, + "loss": 0.9218, + "num_input_tokens_seen": 2142930, + "step": 68 + }, + { + "epoch": 0.006222663119448077, + "flos": 30365890961280.0, + "grad_norm": 1.717567661401882, + "learning_rate": 2.915979772811335e-06, + "loss": 0.9116, + "num_input_tokens_seen": 2172770, + "step": 69 + }, + { + "epoch": 0.006312846642918339, + "flos": 25849240003680.0, + "grad_norm": 2.5433373718762944, + "learning_rate": 2.925889128497372e-06, + "loss": 0.8707, + "num_input_tokens_seen": 2202520, + "step": 70 + }, + { + "epoch": 0.0064030301663886005, + "flos": 21695014372320.0, + "grad_norm": 2.1292479074684425, + "learning_rate": 2.9356579211992906e-06, + "loss": 0.9701, + "num_input_tokens_seen": 2230980, + "step": 71 + }, + { + "epoch": 0.006493213689858863, + "flos": 31529413584960.0, + "grad_norm": 2.2386965274867867, + "learning_rate": 2.9452900829438533e-06, + "loss": 0.8412, + "num_input_tokens_seen": 2262440, + "step": 72 + }, + { + "epoch": 0.006583397213329125, + "flos": 27487457829600.0, + "grad_norm": 2.4839060598722367, + "learning_rate": 2.954789383042727e-06, + "loss": 0.8741, + "num_input_tokens_seen": 2288125, + "step": 73 + }, + { + "epoch": 0.006673580736799386, + "flos": 72446334160800.0, + "grad_norm": 1.2170255149229159, + "learning_rate": 2.9641594369484293e-06, + "loss": 0.6854, + "num_input_tokens_seen": 2386595, + "step": 74 + }, + { + "epoch": 0.006763764260269649, + "flos": 22674815374080.0, + "grad_norm": 3.1184061292765213, + "learning_rate": 2.9734037145158586e-06, + "loss": 0.899, + "num_input_tokens_seen": 2414850, + "step": 75 + }, + { + "epoch": 0.006853947783739911, + "flos": 19247744920320.0, + "grad_norm": 2.2209214880685337, + "learning_rate": 2.982525547716659e-06, + "loss": 0.898, + "num_input_tokens_seen": 2441125, + "step": 76 + }, + { + "epoch": 0.006944131307210172, + "flos": 15131252328960.0, + "grad_norm": 2.664736486293126, + "learning_rate": 2.9915281378493246e-06, + "loss": 0.8682, + "num_input_tokens_seen": 2463890, + "step": 77 + }, + { + "epoch": 0.007034314830680435, + "flos": 31528707359520.0, + "grad_norm": 2.1190262020880013, + "learning_rate": 3.000414562284102e-06, + "loss": 0.8927, + "num_input_tokens_seen": 2496130, + "step": 78 + }, + { + "epoch": 0.0071244983541506966, + "flos": 24681442857600.0, + "grad_norm": 3.350384414024863, + "learning_rate": 3.009187780778246e-06, + "loss": 0.8862, + "num_input_tokens_seen": 2526185, + "step": 79 + }, + { + "epoch": 0.007214681877620958, + "flos": 18852790320480.0, + "grad_norm": 2.5517017520391883, + "learning_rate": 3.017850641394051e-06, + "loss": 0.8788, + "num_input_tokens_seen": 2551390, + "step": 80 + }, + { + "epoch": 0.007304865401091221, + "flos": 45485286136800.0, + "grad_norm": 1.9257490608047638, + "learning_rate": 3.0264058860492832e-06, + "loss": 0.7478, + "num_input_tokens_seen": 2583315, + "step": 81 + }, + { + "epoch": 0.0073950489245614825, + "flos": 24317902438560.0, + "grad_norm": 2.166011332171411, + "learning_rate": 3.0348561557270548e-06, + "loss": 0.9362, + "num_input_tokens_seen": 2612770, + "step": 82 + }, + { + "epoch": 0.007485232448031745, + "flos": 64804254050880.0, + "grad_norm": 0.8748958048771477, + "learning_rate": 3.043203995369939e-06, + "loss": 0.6459, + "num_input_tokens_seen": 2701920, + "step": 83 + }, + { + "epoch": 0.007575415971502007, + "flos": 22202833564320.0, + "grad_norm": 3.5726065246204683, + "learning_rate": 3.051451858480994e-06, + "loss": 0.9798, + "num_input_tokens_seen": 2730395, + "step": 84 + }, + { + "epoch": 0.007665599494972268, + "flos": 32038682397600.0, + "grad_norm": 2.618383606837419, + "learning_rate": 3.05960211145252e-06, + "loss": 0.809, + "num_input_tokens_seen": 2761515, + "step": 85 + }, + { + "epoch": 0.007755783018442531, + "flos": 31602352121280.0, + "grad_norm": 1.9907251713100373, + "learning_rate": 3.0676570376416543e-06, + "loss": 0.8102, + "num_input_tokens_seen": 2791690, + "step": 86 + }, + { + "epoch": 0.007845966541912792, + "flos": 20710083943680.0, + "grad_norm": 3.090316254070269, + "learning_rate": 3.0756188412103647e-06, + "loss": 0.9078, + "num_input_tokens_seen": 2818570, + "step": 87 + }, + { + "epoch": 0.007936150065383054, + "flos": 26722941309600.0, + "grad_norm": 1.9036793494937867, + "learning_rate": 3.083489650746004e-06, + "loss": 0.8856, + "num_input_tokens_seen": 2846040, + "step": 88 + }, + { + "epoch": 0.008026333588853317, + "flos": 21039422449920.0, + "grad_norm": 1.8977251011312204, + "learning_rate": 3.0912715226772975e-06, + "loss": 0.9147, + "num_input_tokens_seen": 2873275, + "step": 89 + }, + { + "epoch": 0.008116517112323578, + "flos": 23407546015680.0, + "grad_norm": 1.618538039186167, + "learning_rate": 3.098966444499481e-06, + "loss": 0.8267, + "num_input_tokens_seen": 2905390, + "step": 90 + }, + { + "epoch": 0.00820670063579384, + "flos": 21586647321120.0, + "grad_norm": 2.2874551278804685, + "learning_rate": 3.1065763378212426e-06, + "loss": 0.8945, + "num_input_tokens_seen": 2934495, + "step": 91 + }, + { + "epoch": 0.008296884159264103, + "flos": 23586993115200.0, + "grad_norm": 1.6115771291977532, + "learning_rate": 3.1141030612451554e-06, + "loss": 0.8953, + "num_input_tokens_seen": 2965185, + "step": 92 + }, + { + "epoch": 0.008387067682734364, + "flos": 62278767956640.0, + "grad_norm": 0.8732800950926431, + "learning_rate": 3.1215484130924052e-06, + "loss": 0.6769, + "num_input_tokens_seen": 3051530, + "step": 93 + }, + { + "epoch": 0.008477251206204626, + "flos": 21841077293760.0, + "grad_norm": 1.9394212856541229, + "learning_rate": 3.128914133981793e-06, + "loss": 0.8988, + "num_input_tokens_seen": 3079185, + "step": 94 + }, + { + "epoch": 0.008567434729674889, + "flos": 26904953122560.0, + "grad_norm": 2.7741177436576607, + "learning_rate": 3.136201909272287e-06, + "loss": 0.8931, + "num_input_tokens_seen": 3109260, + "step": 95 + }, + { + "epoch": 0.00865761825314515, + "flos": 25991920476960.0, + "grad_norm": 2.1501188600198677, + "learning_rate": 3.1434133713776735e-06, + "loss": 0.8554, + "num_input_tokens_seen": 3138075, + "step": 96 + }, + { + "epoch": 0.008747801776615412, + "flos": 19983263293920.0, + "grad_norm": 1.8875340187089205, + "learning_rate": 3.15055010196128e-06, + "loss": 0.8819, + "num_input_tokens_seen": 3165025, + "step": 97 + }, + { + "epoch": 0.008837985300085675, + "flos": 41112951121920.0, + "grad_norm": 2.568253547605851, + "learning_rate": 3.157613634018135e-06, + "loss": 0.8753, + "num_input_tokens_seen": 3196850, + "step": 98 + }, + { + "epoch": 0.008928168823555935, + "flos": 23007127461120.0, + "grad_norm": 1.9552268963438804, + "learning_rate": 3.1646054538514336e-06, + "loss": 0.9013, + "num_input_tokens_seen": 3226975, + "step": 99 + }, + { + "epoch": 0.009018352347026198, + "flos": 28868234932320.0, + "grad_norm": 2.3908802968258365, + "learning_rate": 3.1715270029496797e-06, + "loss": 0.9338, + "num_input_tokens_seen": 3255725, + "step": 100 + }, + { + "epoch": 0.00910853587049646, + "flos": 28616072315040.0, + "grad_norm": 1.6604318906053428, + "learning_rate": 3.1783796797704243e-06, + "loss": 0.9132, + "num_input_tokens_seen": 3287410, + "step": 101 + }, + { + "epoch": 0.009198719393966723, + "flos": 19725896910240.0, + "grad_norm": 2.0865551894801007, + "learning_rate": 3.185164841436142e-06, + "loss": 0.8399, + "num_input_tokens_seen": 3314510, + "step": 102 + }, + { + "epoch": 0.009288902917436984, + "flos": 18708400038240.0, + "grad_norm": 2.653160093185487, + "learning_rate": 3.1918838053473723e-06, + "loss": 0.8796, + "num_input_tokens_seen": 3341505, + "step": 103 + }, + { + "epoch": 0.009379086440907246, + "flos": 34191038274720.0, + "grad_norm": 1.6678987796588922, + "learning_rate": 3.198537850717922e-06, + "loss": 0.785, + "num_input_tokens_seen": 3372185, + "step": 104 + }, + { + "epoch": 0.009469269964377509, + "flos": 24974200586400.0, + "grad_norm": 2.0397665043511597, + "learning_rate": 3.205128220036622e-06, + "loss": 0.8388, + "num_input_tokens_seen": 3401880, + "step": 105 + }, + { + "epoch": 0.00955945348784777, + "flos": 21622261684800.0, + "grad_norm": 1.9944135010295607, + "learning_rate": 3.2116561204597917e-06, + "loss": 0.8378, + "num_input_tokens_seen": 3431870, + "step": 106 + }, + { + "epoch": 0.009649637011318032, + "flos": 61086358695840.0, + "grad_norm": 0.8630207756255796, + "learning_rate": 3.218122725138335e-06, + "loss": 0.6675, + "num_input_tokens_seen": 3522085, + "step": 107 + }, + { + "epoch": 0.009739820534788295, + "flos": 24317716589760.0, + "grad_norm": 9.193565390656943, + "learning_rate": 3.224529174483104e-06, + "loss": 0.8717, + "num_input_tokens_seen": 3550490, + "step": 108 + }, + { + "epoch": 0.009830004058258556, + "flos": 29527655340000.0, + "grad_norm": 2.447845219861327, + "learning_rate": 3.2308765773719435e-06, + "loss": 0.8152, + "num_input_tokens_seen": 3580390, + "step": 109 + }, + { + "epoch": 0.009920187581728818, + "flos": 18488878203840.0, + "grad_norm": 1.919473279388928, + "learning_rate": 3.2371660123016323e-06, + "loss": 0.8659, + "num_input_tokens_seen": 3607695, + "step": 110 + }, + { + "epoch": 0.010010371105199081, + "flos": 29630781455040.0, + "grad_norm": 1.9000270138878073, + "learning_rate": 3.2433985284876787e-06, + "loss": 0.7929, + "num_input_tokens_seen": 3638605, + "step": 111 + }, + { + "epoch": 0.010100554628669342, + "flos": 20419779419040.0, + "grad_norm": 1.527120198233759, + "learning_rate": 3.2495751469148143e-06, + "loss": 0.8636, + "num_input_tokens_seen": 3666420, + "step": 112 + }, + { + "epoch": 0.010190738152139604, + "flos": 18124297031520.0, + "grad_norm": 5.1452259935694125, + "learning_rate": 3.2556968613407816e-06, + "loss": 0.8374, + "num_input_tokens_seen": 3693230, + "step": 113 + }, + { + "epoch": 0.010280921675609867, + "flos": 18627693022080.0, + "grad_norm": 1.8847568333870408, + "learning_rate": 3.2617646392559094e-06, + "loss": 0.9047, + "num_input_tokens_seen": 3719965, + "step": 114 + }, + { + "epoch": 0.010371105199080128, + "flos": 19618830800640.0, + "grad_norm": 1.9727746304787384, + "learning_rate": 3.2677794228007836e-06, + "loss": 0.9175, + "num_input_tokens_seen": 3745190, + "step": 115 + }, + { + "epoch": 0.01046128872255039, + "flos": 36122459866560.0, + "grad_norm": 2.0330298035665737, + "learning_rate": 3.273742129644185e-06, + "loss": 0.8506, + "num_input_tokens_seen": 3775615, + "step": 116 + }, + { + "epoch": 0.010551472246020653, + "flos": 15319397152320.0, + "grad_norm": 2.1659498006933435, + "learning_rate": 3.279653653823352e-06, + "loss": 0.8709, + "num_input_tokens_seen": 3802000, + "step": 117 + }, + { + "epoch": 0.010641655769490914, + "flos": 21109796272800.0, + "grad_norm": 2.4132804715003093, + "learning_rate": 3.285514866548481e-06, + "loss": 0.865, + "num_input_tokens_seen": 3830415, + "step": 118 + }, + { + "epoch": 0.010731839292961176, + "flos": 64429679946720.0, + "grad_norm": 2.010715034231121, + "learning_rate": 3.2913266169732838e-06, + "loss": 0.9277, + "num_input_tokens_seen": 3865140, + "step": 119 + }, + { + "epoch": 0.010822022816431439, + "flos": 23443792265280.0, + "grad_norm": 2.759129023577912, + "learning_rate": 3.2970897329333017e-06, + "loss": 0.8463, + "num_input_tokens_seen": 3890865, + "step": 120 + }, + { + "epoch": 0.0109122063399017, + "flos": 24354260197440.0, + "grad_norm": 1.9159953420578437, + "learning_rate": 3.302805021653585e-06, + "loss": 0.8425, + "num_input_tokens_seen": 3919770, + "step": 121 + }, + { + "epoch": 0.011002389863371962, + "flos": 19501989044160.0, + "grad_norm": 2.0504124482371204, + "learning_rate": 3.3084732704272426e-06, + "loss": 0.8548, + "num_input_tokens_seen": 3947505, + "step": 122 + }, + { + "epoch": 0.011092573386842225, + "flos": 24637874165280.0, + "grad_norm": 2.807387558544354, + "learning_rate": 3.314095247266304e-06, + "loss": 0.8565, + "num_input_tokens_seen": 3973630, + "step": 123 + }, + { + "epoch": 0.011182756910312485, + "flos": 24278794117440.0, + "grad_norm": 5.2435946672277876, + "learning_rate": 3.3196717015262255e-06, + "loss": 0.8211, + "num_input_tokens_seen": 4000955, + "step": 124 + }, + { + "epoch": 0.011272940433782748, + "flos": 24241061077440.0, + "grad_norm": 2.6527130813654005, + "learning_rate": 3.325203364505307e-06, + "loss": 0.8469, + "num_input_tokens_seen": 4030425, + "step": 125 + }, + { + "epoch": 0.01136312395725301, + "flos": 17389893750720.0, + "grad_norm": 2.944089114343484, + "learning_rate": 3.3306909500202442e-06, + "loss": 0.8951, + "num_input_tokens_seen": 4055555, + "step": 126 + }, + { + "epoch": 0.011453307480723271, + "flos": 23734579996800.0, + "grad_norm": 3.109355677252304, + "learning_rate": 3.3361351549589145e-06, + "loss": 0.9231, + "num_input_tokens_seen": 4083935, + "step": 127 + }, + { + "epoch": 0.011543491004193534, + "flos": 27267155430240.0, + "grad_norm": 2.14906906555158, + "learning_rate": 3.341536659811494e-06, + "loss": 0.7971, + "num_input_tokens_seen": 4113720, + "step": 128 + }, + { + "epoch": 0.011633674527663796, + "flos": 15319583001120.0, + "grad_norm": 2.3532032866202615, + "learning_rate": 3.346896129180904e-06, + "loss": 0.8822, + "num_input_tokens_seen": 4139575, + "step": 129 + }, + { + "epoch": 0.011723858051134057, + "flos": 27230277294720.0, + "grad_norm": 1.5864040646838096, + "learning_rate": 3.35221421227355e-06, + "loss": 0.8655, + "num_input_tokens_seen": 4170730, + "step": 130 + }, + { + "epoch": 0.01181404157460432, + "flos": 19035211000800.0, + "grad_norm": 5.206887657434952, + "learning_rate": 3.357491543371255e-06, + "loss": 0.8867, + "num_input_tokens_seen": 4197565, + "step": 131 + }, + { + "epoch": 0.011904225098074582, + "flos": 20889568212960.0, + "grad_norm": 1.6384883227661229, + "learning_rate": 3.3627287422852543e-06, + "loss": 0.815, + "num_input_tokens_seen": 4226190, + "step": 132 + }, + { + "epoch": 0.011994408621544843, + "flos": 23404423755840.0, + "grad_norm": 1.7008885389630057, + "learning_rate": 3.3679264147930497e-06, + "loss": 0.8891, + "num_input_tokens_seen": 4255825, + "step": 133 + }, + { + "epoch": 0.012084592145015106, + "flos": 26722160744640.0, + "grad_norm": 3.200313022324056, + "learning_rate": 3.37308515305891e-06, + "loss": 0.8084, + "num_input_tokens_seen": 4287045, + "step": 134 + }, + { + "epoch": 0.012174775668485368, + "flos": 24785981423520.0, + "grad_norm": 1.834870113380589, + "learning_rate": 3.3782055360387313e-06, + "loss": 0.9264, + "num_input_tokens_seen": 4316580, + "step": 135 + }, + { + "epoch": 0.012264959191955629, + "flos": 32989039215840.0, + "grad_norm": 1.89301086454503, + "learning_rate": 3.3832881298699633e-06, + "loss": 0.8214, + "num_input_tokens_seen": 4346600, + "step": 136 + }, + { + "epoch": 0.012355142715425892, + "flos": 19582435872000.0, + "grad_norm": 2.4065919112996617, + "learning_rate": 3.388333488247249e-06, + "loss": 0.8339, + "num_input_tokens_seen": 4375235, + "step": 137 + }, + { + "epoch": 0.012445326238896154, + "flos": 32369321845440.0, + "grad_norm": 1.7134564812384774, + "learning_rate": 3.393342152784406e-06, + "loss": 0.9117, + "num_input_tokens_seen": 4407085, + "step": 138 + }, + { + "epoch": 0.012535509762366415, + "flos": 19655077050240.0, + "grad_norm": 2.0349183044156223, + "learning_rate": 3.3983146533633376e-06, + "loss": 0.8682, + "num_input_tokens_seen": 4433680, + "step": 139 + }, + { + "epoch": 0.012625693285836678, + "flos": 24646088682240.0, + "grad_norm": 2.1374532772557093, + "learning_rate": 3.403251508470442e-06, + "loss": 0.9234, + "num_input_tokens_seen": 4462520, + "step": 140 + }, + { + "epoch": 0.01271587680930694, + "flos": 19436447290080.0, + "grad_norm": 2.033440333297442, + "learning_rate": 3.408153225521043e-06, + "loss": 0.8247, + "num_input_tokens_seen": 4490805, + "step": 141 + }, + { + "epoch": 0.012806060332777201, + "flos": 29163557374560.0, + "grad_norm": 2.2396188304364992, + "learning_rate": 3.413020301172361e-06, + "loss": 0.7541, + "num_input_tokens_seen": 4520415, + "step": 142 + }, + { + "epoch": 0.012896243856247463, + "flos": 21295153364160.0, + "grad_norm": 1.7193660672342985, + "learning_rate": 3.4178532216255024e-06, + "loss": 0.845, + "num_input_tokens_seen": 4548350, + "step": 143 + }, + { + "epoch": 0.012986427379717726, + "flos": 30584297702880.0, + "grad_norm": 4.213414640205477, + "learning_rate": 3.422652462916924e-06, + "loss": 0.9013, + "num_input_tokens_seen": 4578295, + "step": 144 + }, + { + "epoch": 0.013076610903187987, + "flos": 29746582458240.0, + "grad_norm": 1.6384155654966908, + "learning_rate": 3.4274184911998124e-06, + "loss": 0.7635, + "num_input_tokens_seen": 4608405, + "step": 145 + }, + { + "epoch": 0.01316679442665825, + "flos": 19545632076000.0, + "grad_norm": 2.173951178921118, + "learning_rate": 3.4321517630157976e-06, + "loss": 0.9525, + "num_input_tokens_seen": 4635210, + "step": 146 + }, + { + "epoch": 0.013256977950128512, + "flos": 18780334991040.0, + "grad_norm": 3.872806297290507, + "learning_rate": 3.4368527255573845e-06, + "loss": 0.9191, + "num_input_tokens_seen": 4662245, + "step": 147 + }, + { + "epoch": 0.013347161473598773, + "flos": 34228027919520.0, + "grad_norm": 1.9002852376353925, + "learning_rate": 3.4415218169214994e-06, + "loss": 0.8318, + "num_input_tokens_seen": 4694005, + "step": 148 + }, + { + "epoch": 0.013437344997069035, + "flos": 34991429346720.0, + "grad_norm": 1.6108908319974775, + "learning_rate": 3.4461594663544882e-06, + "loss": 0.8371, + "num_input_tokens_seen": 4725580, + "step": 149 + }, + { + "epoch": 0.013527528520539298, + "flos": 33821141826720.0, + "grad_norm": 2.4140942025186245, + "learning_rate": 3.450766094488929e-06, + "loss": 0.8029, + "num_input_tokens_seen": 4755245, + "step": 150 + }, + { + "epoch": 0.013617712044009559, + "flos": 21804496516320.0, + "grad_norm": 1.8533688518267568, + "learning_rate": 3.4553421135725735e-06, + "loss": 0.9033, + "num_input_tokens_seen": 4784355, + "step": 151 + }, + { + "epoch": 0.013707895567479821, + "flos": 27779323484160.0, + "grad_norm": 2.257846079645944, + "learning_rate": 3.45988792768973e-06, + "loss": 0.8719, + "num_input_tokens_seen": 4815590, + "step": 152 + }, + { + "epoch": 0.013798079090950084, + "flos": 21585755246880.0, + "grad_norm": 1.7928255975626175, + "learning_rate": 3.464403932975393e-06, + "loss": 0.9205, + "num_input_tokens_seen": 4843955, + "step": 153 + }, + { + "epoch": 0.013888262614420345, + "flos": 25771580907840.0, + "grad_norm": 1.7537544711566464, + "learning_rate": 3.468890517822395e-06, + "loss": 0.9115, + "num_input_tokens_seen": 4874525, + "step": 154 + }, + { + "epoch": 0.013978446137890607, + "flos": 20419816588800.0, + "grad_norm": 2.2743633259690905, + "learning_rate": 3.473348063081853e-06, + "loss": 0.8482, + "num_input_tokens_seen": 4901030, + "step": 155 + }, + { + "epoch": 0.01406862966136087, + "flos": 22242016224960.0, + "grad_norm": 1.8347056050317745, + "learning_rate": 3.4777769422571727e-06, + "loss": 0.8107, + "num_input_tokens_seen": 4930025, + "step": 156 + }, + { + "epoch": 0.01415881318483113, + "flos": 26649296547840.0, + "grad_norm": 1.8486346180821875, + "learning_rate": 3.4821775216918497e-06, + "loss": 0.8522, + "num_input_tokens_seen": 4963575, + "step": 157 + }, + { + "epoch": 0.014248996708301393, + "flos": 24937136602080.0, + "grad_norm": 1.7933598509438693, + "learning_rate": 3.4865501607513164e-06, + "loss": 0.8031, + "num_input_tokens_seen": 4996240, + "step": 158 + }, + { + "epoch": 0.014339180231771656, + "flos": 30766086497280.0, + "grad_norm": 3.604709929754867, + "learning_rate": 3.4908952119990423e-06, + "loss": 0.9249, + "num_input_tokens_seen": 5024610, + "step": 159 + }, + { + "epoch": 0.014429363755241916, + "flos": 55666822437120.0, + "grad_norm": 0.9140730963198742, + "learning_rate": 3.495213021367122e-06, + "loss": 0.6271, + "num_input_tokens_seen": 5117515, + "step": 160 + }, + { + "epoch": 0.014519547278712179, + "flos": 30219753700320.0, + "grad_norm": 2.712795269624348, + "learning_rate": 3.4995039283215464e-06, + "loss": 0.8763, + "num_input_tokens_seen": 5143085, + "step": 161 + }, + { + "epoch": 0.014609730802182442, + "flos": 23771866999680.0, + "grad_norm": 2.180965472795058, + "learning_rate": 3.5037682660223533e-06, + "loss": 0.8699, + "num_input_tokens_seen": 5171385, + "step": 162 + }, + { + "epoch": 0.014699914325652702, + "flos": 20055755793120.0, + "grad_norm": 2.13299012055861, + "learning_rate": 3.508006361478857e-06, + "loss": 0.8845, + "num_input_tokens_seen": 5199690, + "step": 163 + }, + { + "epoch": 0.014790097849122965, + "flos": 18270285613440.0, + "grad_norm": 2.101397993604937, + "learning_rate": 3.5122185357001253e-06, + "loss": 0.8368, + "num_input_tokens_seen": 5227625, + "step": 164 + }, + { + "epoch": 0.014880281372593228, + "flos": 23004488408160.0, + "grad_norm": 2.6103207821499406, + "learning_rate": 3.5164051038408817e-06, + "loss": 0.9091, + "num_input_tokens_seen": 5254215, + "step": 165 + }, + { + "epoch": 0.01497046489606349, + "flos": 24718767030240.0, + "grad_norm": 1.8755976042483455, + "learning_rate": 3.5205663753430093e-06, + "loss": 0.8946, + "num_input_tokens_seen": 5283430, + "step": 166 + }, + { + "epoch": 0.015060648419533751, + "flos": 25300602681600.0, + "grad_norm": 1.741686693407969, + "learning_rate": 3.5247026540727915e-06, + "loss": 0.9496, + "num_input_tokens_seen": 5313155, + "step": 167 + }, + { + "epoch": 0.015150831943004013, + "flos": 27267415618560.0, + "grad_norm": 2.5163919926322085, + "learning_rate": 3.5288142384540645e-06, + "loss": 0.7994, + "num_input_tokens_seen": 5345475, + "step": 168 + }, + { + "epoch": 0.015241015466474276, + "flos": 26504311549440.0, + "grad_norm": 2.2273125706861667, + "learning_rate": 3.532901421597421e-06, + "loss": 0.8852, + "num_input_tokens_seen": 5376045, + "step": 169 + }, + { + "epoch": 0.015331198989944537, + "flos": 29594498035680.0, + "grad_norm": 2.9439747653688233, + "learning_rate": 3.5369644914255915e-06, + "loss": 0.7978, + "num_input_tokens_seen": 5400890, + "step": 170 + }, + { + "epoch": 0.0154213825134148, + "flos": 25993890474240.0, + "grad_norm": 2.221235790132099, + "learning_rate": 3.5410037307951596e-06, + "loss": 0.8508, + "num_input_tokens_seen": 5431340, + "step": 171 + }, + { + "epoch": 0.015511566036885062, + "flos": 28907045895360.0, + "grad_norm": 1.9215228370621673, + "learning_rate": 3.545019417614725e-06, + "loss": 0.899, + "num_input_tokens_seen": 5464405, + "step": 172 + }, + { + "epoch": 0.015601749560355323, + "flos": 20929865966400.0, + "grad_norm": 2.030276255860237, + "learning_rate": 3.5490118249596387e-06, + "loss": 0.8953, + "num_input_tokens_seen": 5492020, + "step": 173 + }, + { + "epoch": 0.015691933083825584, + "flos": 23513645711520.0, + "grad_norm": 2.35781384285395, + "learning_rate": 3.5529812211834352e-06, + "loss": 0.904, + "num_input_tokens_seen": 5522460, + "step": 174 + }, + { + "epoch": 0.015782116607295848, + "flos": 21620143008480.0, + "grad_norm": 1.9019169573460302, + "learning_rate": 3.5569278700260707e-06, + "loss": 0.8563, + "num_input_tokens_seen": 5551315, + "step": 175 + }, + { + "epoch": 0.01587230013076611, + "flos": 22387447260480.0, + "grad_norm": 2.0622751517989286, + "learning_rate": 3.5608520307190746e-06, + "loss": 0.8384, + "num_input_tokens_seen": 5580890, + "step": 176 + }, + { + "epoch": 0.01596248365423637, + "flos": 33751920266400.0, + "grad_norm": 1.6895564935185357, + "learning_rate": 3.564753958087731e-06, + "loss": 0.7929, + "num_input_tokens_seen": 5612435, + "step": 177 + }, + { + "epoch": 0.016052667177706634, + "flos": 19728312944640.0, + "grad_norm": 1.8725643232474276, + "learning_rate": 3.5686339026503684e-06, + "loss": 0.8645, + "num_input_tokens_seen": 5642500, + "step": 178 + }, + { + "epoch": 0.016142850701176895, + "flos": 26685617136960.0, + "grad_norm": 6.801543778600326, + "learning_rate": 3.5724921107148806e-06, + "loss": 0.9393, + "num_input_tokens_seen": 5672890, + "step": 179 + }, + { + "epoch": 0.016233034224647155, + "flos": 21221285583840.0, + "grad_norm": 2.0876911235734825, + "learning_rate": 3.576328824472552e-06, + "loss": 0.9137, + "num_input_tokens_seen": 5701850, + "step": 180 + }, + { + "epoch": 0.01632321774811742, + "flos": 25075728401760.0, + "grad_norm": 5.447218880054452, + "learning_rate": 3.5801442820892838e-06, + "loss": 0.9471, + "num_input_tokens_seen": 5730545, + "step": 181 + }, + { + "epoch": 0.01641340127158768, + "flos": 26504088530880.0, + "grad_norm": 1.4646115538626225, + "learning_rate": 3.583938717794313e-06, + "loss": 0.8785, + "num_input_tokens_seen": 5761535, + "step": 182 + }, + { + "epoch": 0.01650358479505794, + "flos": 26610188226720.0, + "grad_norm": 2.1552181262725973, + "learning_rate": 3.5877123619664928e-06, + "loss": 0.8982, + "num_input_tokens_seen": 5791780, + "step": 183 + }, + { + "epoch": 0.016593768318528206, + "flos": 27701701558080.0, + "grad_norm": 2.268264811104713, + "learning_rate": 3.5914654412182268e-06, + "loss": 0.8737, + "num_input_tokens_seen": 5821560, + "step": 184 + }, + { + "epoch": 0.016683951841998466, + "flos": 23042481636480.0, + "grad_norm": 2.486006770676395, + "learning_rate": 3.595198178477127e-06, + "loss": 0.8931, + "num_input_tokens_seen": 5851315, + "step": 185 + }, + { + "epoch": 0.016774135365468727, + "flos": 45958717567200.0, + "grad_norm": 4.59175201355159, + "learning_rate": 3.5989107930654757e-06, + "loss": 0.7914, + "num_input_tokens_seen": 5885980, + "step": 186 + }, + { + "epoch": 0.01686431888893899, + "flos": 23002369731840.0, + "grad_norm": 1.906621427777382, + "learning_rate": 3.6026035007775437e-06, + "loss": 0.9214, + "num_input_tokens_seen": 5912870, + "step": 187 + }, + { + "epoch": 0.016954502412409252, + "flos": 17359631832480.0, + "grad_norm": 2.198903794713853, + "learning_rate": 3.6062765139548636e-06, + "loss": 0.826, + "num_input_tokens_seen": 5939170, + "step": 188 + }, + { + "epoch": 0.017044685935879513, + "flos": 29889820477920.0, + "grad_norm": 9.224987131332064, + "learning_rate": 3.6099300415594945e-06, + "loss": 0.7223, + "num_input_tokens_seen": 5971245, + "step": 189 + }, + { + "epoch": 0.017134869459349777, + "flos": 30437156858400.0, + "grad_norm": 1.8110270598995015, + "learning_rate": 3.6135642892453575e-06, + "loss": 0.7385, + "num_input_tokens_seen": 6004155, + "step": 190 + }, + { + "epoch": 0.01722505298282004, + "flos": 25557485858400.0, + "grad_norm": 1.5372220571239295, + "learning_rate": 3.6171794594277004e-06, + "loss": 0.8714, + "num_input_tokens_seen": 6035355, + "step": 191 + }, + { + "epoch": 0.0173152365062903, + "flos": 21039682638240.0, + "grad_norm": 4.11226027697315, + "learning_rate": 3.620775751350745e-06, + "loss": 0.8068, + "num_input_tokens_seen": 6063520, + "step": 192 + }, + { + "epoch": 0.017405420029760563, + "flos": 17430265843680.0, + "grad_norm": 4.162017262736004, + "learning_rate": 3.6243533611535794e-06, + "loss": 0.8954, + "num_input_tokens_seen": 6085410, + "step": 193 + }, + { + "epoch": 0.017495603553230824, + "flos": 25628565906720.0, + "grad_norm": 1.9594690668380716, + "learning_rate": 3.627912481934351e-06, + "loss": 0.848, + "num_input_tokens_seen": 6115620, + "step": 194 + }, + { + "epoch": 0.017585787076701085, + "flos": 22569310394400.0, + "grad_norm": 2.0892465468986727, + "learning_rate": 3.6314533038128e-06, + "loss": 0.831, + "num_input_tokens_seen": 6141860, + "step": 195 + }, + { + "epoch": 0.01767597060017135, + "flos": 65342706858720.0, + "grad_norm": 0.8074622599519082, + "learning_rate": 3.6349760139912048e-06, + "loss": 0.612, + "num_input_tokens_seen": 6225910, + "step": 196 + }, + { + "epoch": 0.01776615412364161, + "flos": 25738902955200.0, + "grad_norm": 1.833591731963898, + "learning_rate": 3.638480796813769e-06, + "loss": 0.9177, + "num_input_tokens_seen": 6251540, + "step": 197 + }, + { + "epoch": 0.01785633764711187, + "flos": 24864049386720.0, + "grad_norm": 3.5871003037937954, + "learning_rate": 3.641967833824504e-06, + "loss": 0.8522, + "num_input_tokens_seen": 6283125, + "step": 198 + }, + { + "epoch": 0.017946521170582135, + "flos": 22312538726880.0, + "grad_norm": 3.639026339962641, + "learning_rate": 3.645437303823663e-06, + "loss": 0.848, + "num_input_tokens_seen": 6312710, + "step": 199 + }, + { + "epoch": 0.018036704694052396, + "flos": 36923780182560.0, + "grad_norm": 2.5436691983673225, + "learning_rate": 3.64888938292275e-06, + "loss": 0.8351, + "num_input_tokens_seen": 6343685, + "step": 200 + }, + { + "epoch": 0.01812688821752266, + "flos": 20383087132320.0, + "grad_norm": 2.20228452410027, + "learning_rate": 3.6523242445981603e-06, + "loss": 0.8593, + "num_input_tokens_seen": 6372390, + "step": 201 + }, + { + "epoch": 0.01821707174099292, + "flos": 21877509392160.0, + "grad_norm": 1.7694094449661861, + "learning_rate": 3.655742059743495e-06, + "loss": 0.8217, + "num_input_tokens_seen": 6403210, + "step": 202 + }, + { + "epoch": 0.018307255264463182, + "flos": 20303495208960.0, + "grad_norm": 2.562111615777435, + "learning_rate": 3.659142996720576e-06, + "loss": 0.9125, + "num_input_tokens_seen": 6428180, + "step": 203 + }, + { + "epoch": 0.018397438787933446, + "flos": 20565433473120.0, + "grad_norm": 16.191238030640022, + "learning_rate": 3.6625272214092135e-06, + "loss": 0.8223, + "num_input_tokens_seen": 6455440, + "step": 204 + }, + { + "epoch": 0.018487622311403707, + "flos": 69057702972480.0, + "grad_norm": 0.6901464461324848, + "learning_rate": 3.6658948972557535e-06, + "loss": 0.5874, + "num_input_tokens_seen": 6559170, + "step": 205 + }, + { + "epoch": 0.018577805834873968, + "flos": 22642063081920.0, + "grad_norm": 1.8043749650110203, + "learning_rate": 3.6692461853204432e-06, + "loss": 0.7736, + "num_input_tokens_seen": 6587570, + "step": 206 + }, + { + "epoch": 0.018667989358344232, + "flos": 20929791626880.0, + "grad_norm": 1.989508888755896, + "learning_rate": 3.672581244323656e-06, + "loss": 0.7932, + "num_input_tokens_seen": 6616235, + "step": 207 + }, + { + "epoch": 0.018758172881814493, + "flos": 33426261566400.0, + "grad_norm": 1.9046878980553252, + "learning_rate": 3.6759002306909926e-06, + "loss": 0.7274, + "num_input_tokens_seen": 6650150, + "step": 208 + }, + { + "epoch": 0.018848356405284754, + "flos": 19836791505120.0, + "grad_norm": 1.7841109339303154, + "learning_rate": 3.67920329859731e-06, + "loss": 0.8628, + "num_input_tokens_seen": 6677980, + "step": 209 + }, + { + "epoch": 0.018938539928755018, + "flos": 20784509270400.0, + "grad_norm": 3.079956594136537, + "learning_rate": 3.6824906000096923e-06, + "loss": 0.8522, + "num_input_tokens_seen": 6706700, + "step": 210 + }, + { + "epoch": 0.01902872345222528, + "flos": 28247365299360.0, + "grad_norm": 5.074474348835768, + "learning_rate": 3.6857622847294067e-06, + "loss": 0.8585, + "num_input_tokens_seen": 6734750, + "step": 211 + }, + { + "epoch": 0.01911890697569554, + "flos": 17315356914720.0, + "grad_norm": 3.2053753647018777, + "learning_rate": 3.6890185004328626e-06, + "loss": 0.8478, + "num_input_tokens_seen": 6760115, + "step": 212 + }, + { + "epoch": 0.019209090499165804, + "flos": 38162062660800.0, + "grad_norm": 2.0227029417092295, + "learning_rate": 3.6922593927116113e-06, + "loss": 0.7912, + "num_input_tokens_seen": 6791665, + "step": 213 + }, + { + "epoch": 0.019299274022636065, + "flos": 21184370278560.0, + "grad_norm": 2.617883481532759, + "learning_rate": 3.695485105111406e-06, + "loss": 0.8372, + "num_input_tokens_seen": 6816890, + "step": 214 + }, + { + "epoch": 0.019389457546106326, + "flos": 22132496911200.0, + "grad_norm": 1.7343066777699303, + "learning_rate": 3.698695779170352e-06, + "loss": 0.8194, + "num_input_tokens_seen": 6844875, + "step": 215 + }, + { + "epoch": 0.01947964106957659, + "flos": 20820941368800.0, + "grad_norm": 1.999391224715717, + "learning_rate": 3.7018915544561744e-06, + "loss": 0.8391, + "num_input_tokens_seen": 6873510, + "step": 216 + }, + { + "epoch": 0.01956982459304685, + "flos": 20857187618400.0, + "grad_norm": 10.134203836628892, + "learning_rate": 3.7050725686026164e-06, + "loss": 0.7976, + "num_input_tokens_seen": 6900735, + "step": 217 + }, + { + "epoch": 0.01966000811651711, + "flos": 23256911213760.0, + "grad_norm": 2.2682201386778273, + "learning_rate": 3.708238957345014e-06, + "loss": 0.91, + "num_input_tokens_seen": 6925130, + "step": 218 + }, + { + "epoch": 0.019750191639987376, + "flos": 23516544952800.0, + "grad_norm": 1.8013248574706684, + "learning_rate": 3.7113908545550482e-06, + "loss": 0.8882, + "num_input_tokens_seen": 6953215, + "step": 219 + }, + { + "epoch": 0.019840375163457637, + "flos": 22019111942400.0, + "grad_norm": 1.6637470501559022, + "learning_rate": 3.7145283922747028e-06, + "loss": 0.8423, + "num_input_tokens_seen": 6981205, + "step": 220 + }, + { + "epoch": 0.019930558686927898, + "flos": 21286455640320.0, + "grad_norm": 1.606210727434194, + "learning_rate": 3.7176517007494612e-06, + "loss": 0.9359, + "num_input_tokens_seen": 7010190, + "step": 221 + }, + { + "epoch": 0.020020742210398162, + "flos": 36121679301600.0, + "grad_norm": 1.987828231195153, + "learning_rate": 3.7207609084607496e-06, + "loss": 0.7871, + "num_input_tokens_seen": 7044110, + "step": 222 + }, + { + "epoch": 0.020110925733868423, + "flos": 22816417924320.0, + "grad_norm": 2.648233681547019, + "learning_rate": 3.723856142157645e-06, + "loss": 0.8958, + "num_input_tokens_seen": 7070420, + "step": 223 + }, + { + "epoch": 0.020201109257338683, + "flos": 27084883428960.0, + "grad_norm": 1.61571446656186, + "learning_rate": 3.726937526887885e-06, + "loss": 0.8547, + "num_input_tokens_seen": 7100240, + "step": 224 + }, + { + "epoch": 0.020291292780808948, + "flos": 28909127401920.0, + "grad_norm": 3.149932884833434, + "learning_rate": 3.7300051860281798e-06, + "loss": 0.8078, + "num_input_tokens_seen": 7130385, + "step": 225 + }, + { + "epoch": 0.02038147630427921, + "flos": 16411839727680.0, + "grad_norm": 2.5254692661346474, + "learning_rate": 3.733059241313852e-06, + "loss": 0.9273, + "num_input_tokens_seen": 7152920, + "step": 226 + }, + { + "epoch": 0.02047165982774947, + "flos": 26577212916000.0, + "grad_norm": 1.9321812257968711, + "learning_rate": 3.736099812867827e-06, + "loss": 0.8359, + "num_input_tokens_seen": 7182140, + "step": 227 + }, + { + "epoch": 0.020561843351219734, + "flos": 20857596485760.0, + "grad_norm": 2.233252716981778, + "learning_rate": 3.73912701922898e-06, + "loss": 0.9051, + "num_input_tokens_seen": 7211450, + "step": 228 + }, + { + "epoch": 0.020652026874689994, + "flos": 65837999841600.0, + "grad_norm": 0.878677236602209, + "learning_rate": 3.742140977379868e-06, + "loss": 0.6176, + "num_input_tokens_seen": 7299215, + "step": 229 + }, + { + "epoch": 0.020742210398160255, + "flos": 21112100797920.0, + "grad_norm": 2.632496800724125, + "learning_rate": 3.745141802773854e-06, + "loss": 0.8779, + "num_input_tokens_seen": 7326385, + "step": 230 + }, + { + "epoch": 0.02083239392163052, + "flos": 24353851330080.0, + "grad_norm": 2.3925921755008357, + "learning_rate": 3.748129609361645e-06, + "loss": 0.8379, + "num_input_tokens_seen": 7353150, + "step": 231 + }, + { + "epoch": 0.02092257744510078, + "flos": 29824836270240.0, + "grad_norm": 1.7210877905425153, + "learning_rate": 3.7511045096172555e-06, + "loss": 0.8216, + "num_input_tokens_seen": 7386195, + "step": 232 + }, + { + "epoch": 0.02101276096857104, + "flos": 19800693934560.0, + "grad_norm": 2.1314582297885294, + "learning_rate": 3.7540666145634137e-06, + "loss": 0.8385, + "num_input_tokens_seen": 7414470, + "step": 233 + }, + { + "epoch": 0.021102944492041305, + "flos": 21586201284000.0, + "grad_norm": 1.696132198057856, + "learning_rate": 3.7570160337964225e-06, + "loss": 0.8215, + "num_input_tokens_seen": 7443970, + "step": 234 + }, + { + "epoch": 0.021193128015511566, + "flos": 21253034292480.0, + "grad_norm": 2.982908198008954, + "learning_rate": 3.7599528755104913e-06, + "loss": 0.8769, + "num_input_tokens_seen": 7471075, + "step": 235 + }, + { + "epoch": 0.021283311538981827, + "flos": 23146351146720.0, + "grad_norm": 1.836263684454838, + "learning_rate": 3.7628772465215515e-06, + "loss": 0.8432, + "num_input_tokens_seen": 7499375, + "step": 236 + }, + { + "epoch": 0.02137349506245209, + "flos": 28939129131840.0, + "grad_norm": 3.8432023813569045, + "learning_rate": 3.7657892522905666e-06, + "loss": 0.8378, + "num_input_tokens_seen": 7528425, + "step": 237 + }, + { + "epoch": 0.021463678585922352, + "flos": 37579743802560.0, + "grad_norm": 5.455281790247938, + "learning_rate": 3.7686889969463542e-06, + "loss": 0.7592, + "num_input_tokens_seen": 7558670, + "step": 238 + }, + { + "epoch": 0.021553862109392613, + "flos": 22860618502560.0, + "grad_norm": 2.107107908037579, + "learning_rate": 3.771576583307928e-06, + "loss": 0.8177, + "num_input_tokens_seen": 7585190, + "step": 239 + }, + { + "epoch": 0.021644045632862877, + "flos": 23115308663520.0, + "grad_norm": 2.26964039630377, + "learning_rate": 3.7744521129063722e-06, + "loss": 0.8778, + "num_input_tokens_seen": 7612515, + "step": 240 + }, + { + "epoch": 0.021734229156333138, + "flos": 19726937663520.0, + "grad_norm": 1.8173092802116093, + "learning_rate": 3.7773156860062653e-06, + "loss": 0.8682, + "num_input_tokens_seen": 7641000, + "step": 241 + }, + { + "epoch": 0.0218244126798034, + "flos": 25519269611520.0, + "grad_norm": 2.2485852562109945, + "learning_rate": 3.7801674016266554e-06, + "loss": 0.9411, + "num_input_tokens_seen": 7667140, + "step": 242 + }, + { + "epoch": 0.021914596203273663, + "flos": 69281127631680.0, + "grad_norm": 0.6194952060754128, + "learning_rate": 3.7830073575616035e-06, + "loss": 0.563, + "num_input_tokens_seen": 7767010, + "step": 243 + }, + { + "epoch": 0.022004779726743924, + "flos": 20378701100640.0, + "grad_norm": 2.0583892083660165, + "learning_rate": 3.785835650400313e-06, + "loss": 0.8827, + "num_input_tokens_seen": 7794790, + "step": 244 + }, + { + "epoch": 0.022094963250214185, + "flos": 32114520175200.0, + "grad_norm": 2.2163708248003364, + "learning_rate": 3.7886523755468334e-06, + "loss": 0.751, + "num_input_tokens_seen": 7823310, + "step": 245 + }, + { + "epoch": 0.02218514677368445, + "flos": 26431112824800.0, + "grad_norm": 1.804629106113587, + "learning_rate": 3.7914576272393746e-06, + "loss": 0.8358, + "num_input_tokens_seen": 7853825, + "step": 246 + }, + { + "epoch": 0.02227533029715471, + "flos": 22568715678240.0, + "grad_norm": 3.768930211352993, + "learning_rate": 3.7942514985692284e-06, + "loss": 0.9506, + "num_input_tokens_seen": 7879165, + "step": 247 + }, + { + "epoch": 0.02236551382062497, + "flos": 21076560773760.0, + "grad_norm": 2.4852748177001733, + "learning_rate": 3.797034081499296e-06, + "loss": 0.8347, + "num_input_tokens_seen": 7907175, + "step": 248 + }, + { + "epoch": 0.022455697344095235, + "flos": 18080877018240.0, + "grad_norm": 2.0085113140506414, + "learning_rate": 3.7998054668822595e-06, + "loss": 0.8227, + "num_input_tokens_seen": 7935370, + "step": 249 + }, + { + "epoch": 0.022545880867565496, + "flos": 24935947169760.0, + "grad_norm": 2.2053081976962248, + "learning_rate": 3.8025657444783776e-06, + "loss": 0.8198, + "num_input_tokens_seen": 7964420, + "step": 250 + }, + { + "epoch": 0.022636064391035757, + "flos": 24275114311200.0, + "grad_norm": 2.394180346816224, + "learning_rate": 3.80531500297293e-06, + "loss": 0.7552, + "num_input_tokens_seen": 7991905, + "step": 251 + }, + { + "epoch": 0.02272624791450602, + "flos": 14549044980000.0, + "grad_norm": 3.1318161443091403, + "learning_rate": 3.8080533299933147e-06, + "loss": 0.8124, + "num_input_tokens_seen": 8017095, + "step": 252 + }, + { + "epoch": 0.022816431437976282, + "flos": 28762952971200.0, + "grad_norm": 2.612097775830152, + "learning_rate": 3.8107808121258067e-06, + "loss": 0.808, + "num_input_tokens_seen": 8047070, + "step": 253 + }, + { + "epoch": 0.022906614961446543, + "flos": 21585457888800.0, + "grad_norm": 2.2667908924318994, + "learning_rate": 3.813497534931985e-06, + "loss": 0.91, + "num_input_tokens_seen": 8076250, + "step": 254 + }, + { + "epoch": 0.022996798484916807, + "flos": 24245001072000.0, + "grad_norm": 2.367477651005186, + "learning_rate": 3.816203582964841e-06, + "loss": 0.8724, + "num_input_tokens_seen": 8103985, + "step": 255 + }, + { + "epoch": 0.023086982008387068, + "flos": 21039719808000.0, + "grad_norm": 1.8080245042887335, + "learning_rate": 3.818899039784565e-06, + "loss": 0.8986, + "num_input_tokens_seen": 8130890, + "step": 256 + }, + { + "epoch": 0.02317716553185733, + "flos": 21947808875520.0, + "grad_norm": 2.0095007066575947, + "learning_rate": 3.821583987974031e-06, + "loss": 0.9026, + "num_input_tokens_seen": 8158980, + "step": 257 + }, + { + "epoch": 0.023267349055327593, + "flos": 26176311154560.0, + "grad_norm": 2.890607844121387, + "learning_rate": 3.8242585091539755e-06, + "loss": 0.7827, + "num_input_tokens_seen": 8188740, + "step": 258 + }, + { + "epoch": 0.023357532578797854, + "flos": 22387224241920.0, + "grad_norm": 2.7243666236505386, + "learning_rate": 3.8269226839978895e-06, + "loss": 0.9287, + "num_input_tokens_seen": 8216920, + "step": 259 + }, + { + "epoch": 0.023447716102268115, + "flos": 15681190592640.0, + "grad_norm": 2.8791250407130695, + "learning_rate": 3.82957659224662e-06, + "loss": 0.8021, + "num_input_tokens_seen": 8242085, + "step": 260 + }, + { + "epoch": 0.02353789962573838, + "flos": 34040440642560.0, + "grad_norm": 2.9812175565029904, + "learning_rate": 3.8322203127226855e-06, + "loss": 0.7322, + "num_input_tokens_seen": 8272885, + "step": 261 + }, + { + "epoch": 0.02362808314920864, + "flos": 18598025820000.0, + "grad_norm": 1.7952523243831562, + "learning_rate": 3.834853923344326e-06, + "loss": 0.9535, + "num_input_tokens_seen": 8300230, + "step": 262 + }, + { + "epoch": 0.0237182666726789, + "flos": 30365890961280.0, + "grad_norm": 2.1692057062075754, + "learning_rate": 3.837477501139285e-06, + "loss": 0.8467, + "num_input_tokens_seen": 8330580, + "step": 263 + }, + { + "epoch": 0.023808450196149165, + "flos": 27525042190560.0, + "grad_norm": 2.0377223262131565, + "learning_rate": 3.840091122258324e-06, + "loss": 0.8585, + "num_input_tokens_seen": 8360325, + "step": 264 + }, + { + "epoch": 0.023898633719619426, + "flos": 35280321420480.0, + "grad_norm": 1.840233698702878, + "learning_rate": 3.84269486198849e-06, + "loss": 0.767, + "num_input_tokens_seen": 8393910, + "step": 265 + }, + { + "epoch": 0.023988817243089686, + "flos": 22204506203520.0, + "grad_norm": 2.370090731240115, + "learning_rate": 3.845288794766121e-06, + "loss": 0.8639, + "num_input_tokens_seen": 8421665, + "step": 266 + }, + { + "epoch": 0.02407900076655995, + "flos": 17432979236160.0, + "grad_norm": 1.913177223228062, + "learning_rate": 3.847872994189619e-06, + "loss": 0.8415, + "num_input_tokens_seen": 8447780, + "step": 267 + }, + { + "epoch": 0.02416918429003021, + "flos": 24681740215680.0, + "grad_norm": 3.883740688345962, + "learning_rate": 3.8504475330319805e-06, + "loss": 0.9056, + "num_input_tokens_seen": 8476950, + "step": 268 + }, + { + "epoch": 0.024259367813500472, + "flos": 20201186828640.0, + "grad_norm": 1.9575666049426939, + "learning_rate": 3.853012483253093e-06, + "loss": 0.826, + "num_input_tokens_seen": 8504470, + "step": 269 + }, + { + "epoch": 0.024349551336970737, + "flos": 31676851787520.0, + "grad_norm": 2.0102205197274627, + "learning_rate": 3.855567916011802e-06, + "loss": 0.8541, + "num_input_tokens_seen": 8535815, + "step": 270 + }, + { + "epoch": 0.024439734860440997, + "flos": 19107889348800.0, + "grad_norm": 2.056890430838823, + "learning_rate": 3.858113901677755e-06, + "loss": 0.8528, + "num_input_tokens_seen": 8563690, + "step": 271 + }, + { + "epoch": 0.024529918383911258, + "flos": 24131318745120.0, + "grad_norm": 2.2269162616445533, + "learning_rate": 3.860650509843034e-06, + "loss": 0.8071, + "num_input_tokens_seen": 8594790, + "step": 272 + }, + { + "epoch": 0.024620101907381522, + "flos": 21185336692320.0, + "grad_norm": 2.080194461058651, + "learning_rate": 3.863177809333563e-06, + "loss": 0.8303, + "num_input_tokens_seen": 8623485, + "step": 273 + }, + { + "epoch": 0.024710285430851783, + "flos": 20784026063520.0, + "grad_norm": 1.9570416052386341, + "learning_rate": 3.86569586822032e-06, + "loss": 0.8208, + "num_input_tokens_seen": 8652075, + "step": 274 + }, + { + "epoch": 0.024800468954322044, + "flos": 29710224699360.0, + "grad_norm": 2.2312636687557292, + "learning_rate": 3.868204753830331e-06, + "loss": 0.8108, + "num_input_tokens_seen": 8681245, + "step": 275 + }, + { + "epoch": 0.02489065247779231, + "flos": 32110245652800.0, + "grad_norm": 2.229176803518284, + "learning_rate": 3.870704532757476e-06, + "loss": 0.7635, + "num_input_tokens_seen": 8712735, + "step": 276 + }, + { + "epoch": 0.02498083600126257, + "flos": 24719138727840.0, + "grad_norm": 1.8908812924530847, + "learning_rate": 3.8731952708730974e-06, + "loss": 0.8737, + "num_input_tokens_seen": 8741885, + "step": 277 + }, + { + "epoch": 0.02507101952473283, + "flos": 23437659254880.0, + "grad_norm": 2.1883671812803676, + "learning_rate": 3.8756770333364085e-06, + "loss": 0.8693, + "num_input_tokens_seen": 8767765, + "step": 278 + }, + { + "epoch": 0.025161203048203094, + "flos": 25955711397120.0, + "grad_norm": 3.75070552546874, + "learning_rate": 3.878149884604725e-06, + "loss": 0.7674, + "num_input_tokens_seen": 8797395, + "step": 279 + }, + { + "epoch": 0.025251386571673355, + "flos": 21221099735040.0, + "grad_norm": 2.363844118791325, + "learning_rate": 3.8806138884435125e-06, + "loss": 0.8194, + "num_input_tokens_seen": 8824825, + "step": 280 + }, + { + "epoch": 0.025341570095143616, + "flos": 24714492507840.0, + "grad_norm": 1.9126386679449001, + "learning_rate": 3.883069107936248e-06, + "loss": 0.7746, + "num_input_tokens_seen": 8856240, + "step": 281 + }, + { + "epoch": 0.02543175361861388, + "flos": 23698705444800.0, + "grad_norm": 2.3004738775609543, + "learning_rate": 3.885515605494114e-06, + "loss": 0.9022, + "num_input_tokens_seen": 8881215, + "step": 282 + }, + { + "epoch": 0.02552193714208414, + "flos": 23439666421920.0, + "grad_norm": 2.192839400360342, + "learning_rate": 3.8879534428655145e-06, + "loss": 0.9004, + "num_input_tokens_seen": 8908075, + "step": 283 + }, + { + "epoch": 0.025612120665554402, + "flos": 21511478599200.0, + "grad_norm": 1.8136968935888833, + "learning_rate": 3.890382681145432e-06, + "loss": 0.8822, + "num_input_tokens_seen": 8938480, + "step": 284 + }, + { + "epoch": 0.025702304189024666, + "flos": 29852830833120.0, + "grad_norm": 2.05636099194057, + "learning_rate": 3.892803380784608e-06, + "loss": 0.8763, + "num_input_tokens_seen": 8969985, + "step": 285 + }, + { + "epoch": 0.025792487712494927, + "flos": 53529823404480.0, + "grad_norm": 0.9740359161538354, + "learning_rate": 3.8952156015985725e-06, + "loss": 0.6414, + "num_input_tokens_seen": 9057285, + "step": 286 + }, + { + "epoch": 0.025882671235965188, + "flos": 21803678781600.0, + "grad_norm": 2.3734173771982707, + "learning_rate": 3.897619402776516e-06, + "loss": 0.844, + "num_input_tokens_seen": 9086520, + "step": 287 + }, + { + "epoch": 0.025972854759435452, + "flos": 19356966876000.0, + "grad_norm": 2.126785986352267, + "learning_rate": 3.900014842889995e-06, + "loss": 0.9181, + "num_input_tokens_seen": 9112620, + "step": 288 + }, + { + "epoch": 0.026063038282905713, + "flos": 30329793390720.0, + "grad_norm": 2.0534386846631536, + "learning_rate": 3.902401979901503e-06, + "loss": 0.7632, + "num_input_tokens_seen": 9144740, + "step": 289 + }, + { + "epoch": 0.026153221806375974, + "flos": 19363099886400.0, + "grad_norm": 5.898054946721081, + "learning_rate": 3.904780871172884e-06, + "loss": 0.8866, + "num_input_tokens_seen": 9172585, + "step": 290 + }, + { + "epoch": 0.026243405329846238, + "flos": 23071033745760.0, + "grad_norm": 3.0675356200496577, + "learning_rate": 3.907151573473601e-06, + "loss": 0.7185, + "num_input_tokens_seen": 9199655, + "step": 291 + }, + { + "epoch": 0.0263335888533165, + "flos": 24246264843840.0, + "grad_norm": 1.6077713615411307, + "learning_rate": 3.909514142988868e-06, + "loss": 0.8503, + "num_input_tokens_seen": 9232010, + "step": 292 + }, + { + "epoch": 0.02642377237678676, + "flos": 30617384522880.0, + "grad_norm": 2.209071365923621, + "learning_rate": 3.911868635327639e-06, + "loss": 0.8084, + "num_input_tokens_seen": 9262825, + "step": 293 + }, + { + "epoch": 0.026513955900257024, + "flos": 29562489138720.0, + "grad_norm": 1.4789743253634353, + "learning_rate": 3.914215105530455e-06, + "loss": 0.8182, + "num_input_tokens_seen": 9294745, + "step": 294 + }, + { + "epoch": 0.026604139423727285, + "flos": 25337592326400.0, + "grad_norm": 1.7327303950376143, + "learning_rate": 3.916553608077179e-06, + "loss": 0.8326, + "num_input_tokens_seen": 9325680, + "step": 295 + }, + { + "epoch": 0.026694322947197546, + "flos": 29964320144160.0, + "grad_norm": 1.9830907435543692, + "learning_rate": 3.91888419689457e-06, + "loss": 0.8284, + "num_input_tokens_seen": 9357525, + "step": 296 + }, + { + "epoch": 0.02678450647066781, + "flos": 22239340002240.0, + "grad_norm": 3.073695835103232, + "learning_rate": 3.921206925363754e-06, + "loss": 0.7786, + "num_input_tokens_seen": 9387045, + "step": 297 + }, + { + "epoch": 0.02687468999413807, + "flos": 15866213156160.0, + "grad_norm": 2.612375236342927, + "learning_rate": 3.923521846327559e-06, + "loss": 0.7545, + "num_input_tokens_seen": 9414210, + "step": 298 + }, + { + "epoch": 0.02696487351760833, + "flos": 13825569608640.0, + "grad_norm": 2.569050116537577, + "learning_rate": 3.925829012097725e-06, + "loss": 0.7959, + "num_input_tokens_seen": 9440680, + "step": 299 + }, + { + "epoch": 0.027055057041078596, + "flos": 25954559134560.0, + "grad_norm": 2.4982755435556583, + "learning_rate": 3.928128474462e-06, + "loss": 0.7585, + "num_input_tokens_seen": 9467725, + "step": 300 + }, + { + "epoch": 0.027145240564548857, + "flos": 24354037178880.0, + "grad_norm": 2.770711929606323, + "learning_rate": 3.930420284691115e-06, + "loss": 0.8462, + "num_input_tokens_seen": 9493175, + "step": 301 + }, + { + "epoch": 0.027235424088019117, + "flos": 18485681604480.0, + "grad_norm": 2.272465239071187, + "learning_rate": 3.932704493545644e-06, + "loss": 0.8715, + "num_input_tokens_seen": 9520760, + "step": 302 + }, + { + "epoch": 0.02732560761148938, + "flos": 24860555429280.0, + "grad_norm": 2.2129621637266954, + "learning_rate": 3.934981151282745e-06, + "loss": 0.9255, + "num_input_tokens_seen": 9547230, + "step": 303 + }, + { + "epoch": 0.027415791134959643, + "flos": 25848793966560.0, + "grad_norm": 1.8985724381578752, + "learning_rate": 3.9372503076628006e-06, + "loss": 0.8341, + "num_input_tokens_seen": 9576465, + "step": 304 + }, + { + "epoch": 0.027505974658429903, + "flos": 21148904593920.0, + "grad_norm": 1.7949266568557087, + "learning_rate": 3.939512011955941e-06, + "loss": 0.8578, + "num_input_tokens_seen": 9605080, + "step": 305 + }, + { + "epoch": 0.027596158181900168, + "flos": 14882026122720.0, + "grad_norm": 5.410046605951233, + "learning_rate": 3.941766312948463e-06, + "loss": 0.8341, + "num_input_tokens_seen": 9632400, + "step": 306 + }, + { + "epoch": 0.02768634170537043, + "flos": 17686256946240.0, + "grad_norm": 1.8840323647515385, + "learning_rate": 3.944013258949147e-06, + "loss": 0.8083, + "num_input_tokens_seen": 9660480, + "step": 307 + }, + { + "epoch": 0.02777652522884069, + "flos": 22824743950560.0, + "grad_norm": 2.0784272856024733, + "learning_rate": 3.946252897795465e-06, + "loss": 0.6582, + "num_input_tokens_seen": 9687235, + "step": 308 + }, + { + "epoch": 0.027866708752310954, + "flos": 15464939697120.0, + "grad_norm": 2.9189654056371825, + "learning_rate": 3.9484852768596935e-06, + "loss": 0.8472, + "num_input_tokens_seen": 9713295, + "step": 309 + }, + { + "epoch": 0.027956892275781214, + "flos": 23005566331200.0, + "grad_norm": 2.117278375118142, + "learning_rate": 3.950710443054923e-06, + "loss": 0.7386, + "num_input_tokens_seen": 9743365, + "step": 310 + }, + { + "epoch": 0.028047075799251475, + "flos": 33208338031680.0, + "grad_norm": 1.9730788580810192, + "learning_rate": 3.952928442840981e-06, + "loss": 0.7461, + "num_input_tokens_seen": 9771565, + "step": 311 + }, + { + "epoch": 0.02813725932272174, + "flos": 20703170368320.0, + "grad_norm": 1.8452763462643316, + "learning_rate": 3.955139322230243e-06, + "loss": 0.8566, + "num_input_tokens_seen": 9799445, + "step": 312 + }, + { + "epoch": 0.028227442846192, + "flos": 32588806510080.0, + "grad_norm": 2.5057544284885327, + "learning_rate": 3.957343126793365e-06, + "loss": 0.765, + "num_input_tokens_seen": 9832070, + "step": 313 + }, + { + "epoch": 0.02831762636966226, + "flos": 34300223060640.0, + "grad_norm": 1.8404356051089465, + "learning_rate": 3.959539901664921e-06, + "loss": 0.8152, + "num_input_tokens_seen": 9862895, + "step": 314 + }, + { + "epoch": 0.028407809893132525, + "flos": 19472619200160.0, + "grad_norm": 3.35537524198507, + "learning_rate": 3.9617296915489425e-06, + "loss": 0.9288, + "num_input_tokens_seen": 9886995, + "step": 315 + }, + { + "epoch": 0.028497993416602786, + "flos": 33455854428960.0, + "grad_norm": 4.523756052241371, + "learning_rate": 3.963912540724387e-06, + "loss": 0.775, + "num_input_tokens_seen": 9914340, + "step": 316 + }, + { + "epoch": 0.028588176940073047, + "flos": 22314248535840.0, + "grad_norm": 2.340356124285931, + "learning_rate": 3.966088493050501e-06, + "loss": 0.8338, + "num_input_tokens_seen": 9942000, + "step": 317 + }, + { + "epoch": 0.02867836046354331, + "flos": 24062989259040.0, + "grad_norm": 2.0107087497978595, + "learning_rate": 3.968257591972113e-06, + "loss": 0.8161, + "num_input_tokens_seen": 9970080, + "step": 318 + }, + { + "epoch": 0.028768543987013572, + "flos": 26249286860640.0, + "grad_norm": 2.087699231902369, + "learning_rate": 3.970419880524835e-06, + "loss": 0.912, + "num_input_tokens_seen": 10001425, + "step": 319 + }, + { + "epoch": 0.028858727510483833, + "flos": 25520570553120.0, + "grad_norm": 2.226687243798634, + "learning_rate": 3.972575401340192e-06, + "loss": 0.7954, + "num_input_tokens_seen": 10030445, + "step": 320 + }, + { + "epoch": 0.028948911033954097, + "flos": 23954845226400.0, + "grad_norm": 2.1820507295902005, + "learning_rate": 3.974724196650656e-06, + "loss": 0.8704, + "num_input_tokens_seen": 10060870, + "step": 321 + }, + { + "epoch": 0.029039094557424358, + "flos": 26649296547840.0, + "grad_norm": 3.152502573543167, + "learning_rate": 3.976866308294617e-06, + "loss": 0.8196, + "num_input_tokens_seen": 10089795, + "step": 322 + }, + { + "epoch": 0.02912927808089462, + "flos": 33206145015840.0, + "grad_norm": 4.300794998133112, + "learning_rate": 3.979001777721269e-06, + "loss": 0.7875, + "num_input_tokens_seen": 10124185, + "step": 323 + }, + { + "epoch": 0.029219461604364883, + "flos": 21695720597760.0, + "grad_norm": 2.709936655296617, + "learning_rate": 3.981130645995424e-06, + "loss": 0.7954, + "num_input_tokens_seen": 10153590, + "step": 324 + }, + { + "epoch": 0.029309645127835144, + "flos": 22569979450080.0, + "grad_norm": 2.013480176412692, + "learning_rate": 3.983252953802248e-06, + "loss": 0.8999, + "num_input_tokens_seen": 10183050, + "step": 325 + }, + { + "epoch": 0.029399828651305405, + "flos": 68919222682080.0, + "grad_norm": 0.7203496050431835, + "learning_rate": 3.9853687414519285e-06, + "loss": 0.6339, + "num_input_tokens_seen": 10285075, + "step": 326 + }, + { + "epoch": 0.02949001217477567, + "flos": 21144481392480.0, + "grad_norm": 2.0434896414101984, + "learning_rate": 3.987478048884265e-06, + "loss": 0.8717, + "num_input_tokens_seen": 10310665, + "step": 327 + }, + { + "epoch": 0.02958019569824593, + "flos": 26212631743680.0, + "grad_norm": 1.9421343598956295, + "learning_rate": 3.989580915673196e-06, + "loss": 0.802, + "num_input_tokens_seen": 10340905, + "step": 328 + }, + { + "epoch": 0.02967037922171619, + "flos": 24135890625600.0, + "grad_norm": 1.937562572330232, + "learning_rate": 3.991677381031255e-06, + "loss": 0.8029, + "num_input_tokens_seen": 10368060, + "step": 329 + }, + { + "epoch": 0.029760562745186455, + "flos": 21076151906400.0, + "grad_norm": 1.8536871457906818, + "learning_rate": 3.993767483813953e-06, + "loss": 0.8974, + "num_input_tokens_seen": 10397615, + "step": 330 + }, + { + "epoch": 0.029850746268656716, + "flos": 27376934932320.0, + "grad_norm": 2.039819920202332, + "learning_rate": 3.995851262524104e-06, + "loss": 0.741, + "num_input_tokens_seen": 10427625, + "step": 331 + }, + { + "epoch": 0.02994092979212698, + "flos": 26977482791520.0, + "grad_norm": 1.9822803766273553, + "learning_rate": 3.997928755316079e-06, + "loss": 0.8206, + "num_input_tokens_seen": 10457155, + "step": 332 + }, + { + "epoch": 0.03003111331559724, + "flos": 64155984571200.0, + "grad_norm": 0.8252025355638019, + "learning_rate": 4e-06, + "loss": 0.6197, + "num_input_tokens_seen": 10544555, + "step": 333 + }, + { + "epoch": 0.030121296839067502, + "flos": 26213598157440.0, + "grad_norm": 3.470577961173401, + "learning_rate": 3.999999914674486e-06, + "loss": 0.7899, + "num_input_tokens_seen": 10573780, + "step": 334 + }, + { + "epoch": 0.030211480362537766, + "flos": 32368058073600.0, + "grad_norm": 1.9985208405989234, + "learning_rate": 3.999999658697952e-06, + "loss": 0.8054, + "num_input_tokens_seen": 10605560, + "step": 335 + }, + { + "epoch": 0.030301663886008027, + "flos": 65416500299520.0, + "grad_norm": 0.6758771618822085, + "learning_rate": 3.9999992320704185e-06, + "loss": 0.5577, + "num_input_tokens_seen": 10693430, + "step": 336 + }, + { + "epoch": 0.030391847409478288, + "flos": 24718655520960.0, + "grad_norm": 2.6720779074201326, + "learning_rate": 3.999998634791922e-06, + "loss": 0.8409, + "num_input_tokens_seen": 10718025, + "step": 337 + }, + { + "epoch": 0.030482030932948552, + "flos": 33754670828640.0, + "grad_norm": 2.7957995174388715, + "learning_rate": 3.999997866862515e-06, + "loss": 0.8064, + "num_input_tokens_seen": 10748735, + "step": 338 + }, + { + "epoch": 0.030572214456418813, + "flos": 19616414766240.0, + "grad_norm": 1.998132131478497, + "learning_rate": 3.999996928282262e-06, + "loss": 0.9095, + "num_input_tokens_seen": 10776155, + "step": 339 + }, + { + "epoch": 0.030662397979889074, + "flos": 16084322539680.0, + "grad_norm": 2.3870830351534758, + "learning_rate": 3.999995819051244e-06, + "loss": 0.8191, + "num_input_tokens_seen": 10801910, + "step": 340 + }, + { + "epoch": 0.030752581503359338, + "flos": 29557768579200.0, + "grad_norm": 1.7349367348957068, + "learning_rate": 3.9999945391695536e-06, + "loss": 0.7827, + "num_input_tokens_seen": 10830420, + "step": 341 + }, + { + "epoch": 0.0308427650268296, + "flos": 23074713552000.0, + "grad_norm": 3.7318267427645533, + "learning_rate": 3.999993088637302e-06, + "loss": 0.9072, + "num_input_tokens_seen": 10860090, + "step": 342 + }, + { + "epoch": 0.03093294855029986, + "flos": 25807715648160.0, + "grad_norm": 2.018336047552139, + "learning_rate": 3.999991467454612e-06, + "loss": 0.8867, + "num_input_tokens_seen": 10888715, + "step": 343 + }, + { + "epoch": 0.031023132073770124, + "flos": 30182726885760.0, + "grad_norm": 1.538285964795855, + "learning_rate": 3.999989675621622e-06, + "loss": 0.7365, + "num_input_tokens_seen": 10922060, + "step": 344 + }, + { + "epoch": 0.031113315597240385, + "flos": 27342621510240.0, + "grad_norm": 1.7207967189314175, + "learning_rate": 3.999987713138485e-06, + "loss": 0.8265, + "num_input_tokens_seen": 10952525, + "step": 345 + }, + { + "epoch": 0.031203499120710645, + "flos": 71623152292320.0, + "grad_norm": 0.788415208970884, + "learning_rate": 3.999985580005369e-06, + "loss": 0.666, + "num_input_tokens_seen": 11051690, + "step": 346 + }, + { + "epoch": 0.031293682644180906, + "flos": 21257234475360.0, + "grad_norm": 2.1021194570844512, + "learning_rate": 3.999983276222455e-06, + "loss": 0.8934, + "num_input_tokens_seen": 11077485, + "step": 347 + }, + { + "epoch": 0.03138386616765117, + "flos": 21804608025600.0, + "grad_norm": 1.9100899714105242, + "learning_rate": 3.999980801789941e-06, + "loss": 0.8321, + "num_input_tokens_seen": 11107585, + "step": 348 + }, + { + "epoch": 0.031474049691121435, + "flos": 30948135480000.0, + "grad_norm": 1.8550812969232857, + "learning_rate": 3.999978156708036e-06, + "loss": 0.8263, + "num_input_tokens_seen": 11138590, + "step": 349 + }, + { + "epoch": 0.031564233214591696, + "flos": 18925505838240.0, + "grad_norm": 2.123710128652395, + "learning_rate": 3.9999753409769675e-06, + "loss": 0.9303, + "num_input_tokens_seen": 11162785, + "step": 350 + }, + { + "epoch": 0.031654416738061956, + "flos": 66610842387840.0, + "grad_norm": 0.7179795987676258, + "learning_rate": 3.999972354596975e-06, + "loss": 0.6088, + "num_input_tokens_seen": 11252365, + "step": 351 + }, + { + "epoch": 0.03174460026153222, + "flos": 25083831409440.0, + "grad_norm": 2.1526203289358983, + "learning_rate": 3.999969197568314e-06, + "loss": 0.773, + "num_input_tokens_seen": 11280120, + "step": 352 + }, + { + "epoch": 0.03183478378500248, + "flos": 17209926274560.0, + "grad_norm": 2.434739824216788, + "learning_rate": 3.999965869891253e-06, + "loss": 0.8726, + "num_input_tokens_seen": 11305605, + "step": 353 + }, + { + "epoch": 0.03192496730847274, + "flos": 22712065207200.0, + "grad_norm": 2.0777755099242827, + "learning_rate": 3.999962371566075e-06, + "loss": 0.8496, + "num_input_tokens_seen": 11334870, + "step": 354 + }, + { + "epoch": 0.03201515083194301, + "flos": 20966112216000.0, + "grad_norm": 2.0328001709548777, + "learning_rate": 3.999958702593082e-06, + "loss": 0.8908, + "num_input_tokens_seen": 11362935, + "step": 355 + }, + { + "epoch": 0.03210533435541327, + "flos": 27525636906720.0, + "grad_norm": 1.820950637751659, + "learning_rate": 3.999954862972583e-06, + "loss": 0.8473, + "num_input_tokens_seen": 11394775, + "step": 356 + }, + { + "epoch": 0.03219551787888353, + "flos": 24100685129280.0, + "grad_norm": 1.8005149131311957, + "learning_rate": 3.999950852704908e-06, + "loss": 0.7695, + "num_input_tokens_seen": 11424210, + "step": 357 + }, + { + "epoch": 0.03228570140235379, + "flos": 19982817256800.0, + "grad_norm": 1.5785104262038168, + "learning_rate": 3.9999466717903995e-06, + "loss": 0.8811, + "num_input_tokens_seen": 11450810, + "step": 358 + }, + { + "epoch": 0.03237588492582405, + "flos": 57576536889120.0, + "grad_norm": 0.7129580531173719, + "learning_rate": 3.999942320229413e-06, + "loss": 0.594, + "num_input_tokens_seen": 11542735, + "step": 359 + }, + { + "epoch": 0.03246606844929431, + "flos": 17140890563040.0, + "grad_norm": 2.3964779083488037, + "learning_rate": 3.99993779802232e-06, + "loss": 0.9526, + "num_input_tokens_seen": 11569130, + "step": 360 + }, + { + "epoch": 0.03255625197276458, + "flos": 24641888499360.0, + "grad_norm": 1.9390903836688447, + "learning_rate": 3.999933105169506e-06, + "loss": 0.8128, + "num_input_tokens_seen": 11595935, + "step": 361 + }, + { + "epoch": 0.03264643549623484, + "flos": 20020438787520.0, + "grad_norm": 1.6018745676576625, + "learning_rate": 3.999928241671373e-06, + "loss": 0.8216, + "num_input_tokens_seen": 11626830, + "step": 362 + }, + { + "epoch": 0.0327366190197051, + "flos": 50767568707200.0, + "grad_norm": 2.4771690227637047, + "learning_rate": 3.999923207528334e-06, + "loss": 0.7173, + "num_input_tokens_seen": 11658320, + "step": 363 + }, + { + "epoch": 0.03282680254317536, + "flos": 21694940032800.0, + "grad_norm": 1.5110708814919378, + "learning_rate": 3.9999180027408196e-06, + "loss": 0.8187, + "num_input_tokens_seen": 11688520, + "step": 364 + }, + { + "epoch": 0.03291698606664562, + "flos": 28719309939360.0, + "grad_norm": 2.0930819228424817, + "learning_rate": 3.9999126273092735e-06, + "loss": 0.8002, + "num_input_tokens_seen": 11713595, + "step": 365 + }, + { + "epoch": 0.03300716959011588, + "flos": 25990805384160.0, + "grad_norm": 5.2071228248230295, + "learning_rate": 3.999907081234156e-06, + "loss": 0.9126, + "num_input_tokens_seen": 11742855, + "step": 366 + }, + { + "epoch": 0.03309735311358615, + "flos": 22568975866560.0, + "grad_norm": 2.4107196464680807, + "learning_rate": 3.999901364515938e-06, + "loss": 0.7908, + "num_input_tokens_seen": 11770265, + "step": 367 + }, + { + "epoch": 0.03318753663705641, + "flos": 21617243767200.0, + "grad_norm": 2.1171958331968153, + "learning_rate": 3.999895477155108e-06, + "loss": 0.8709, + "num_input_tokens_seen": 11797585, + "step": 368 + }, + { + "epoch": 0.03327772016052667, + "flos": 57111840352320.0, + "grad_norm": 0.7218602100525112, + "learning_rate": 3.999889419152169e-06, + "loss": 0.5753, + "num_input_tokens_seen": 11890860, + "step": 369 + }, + { + "epoch": 0.03336790368399693, + "flos": 16813001677440.0, + "grad_norm": 1.8524861024034103, + "learning_rate": 3.999883190507638e-06, + "loss": 0.8192, + "num_input_tokens_seen": 11919400, + "step": 370 + }, + { + "epoch": 0.033458087207467194, + "flos": 21658879632000.0, + "grad_norm": 2.169901877790957, + "learning_rate": 3.999876791222044e-06, + "loss": 0.8047, + "num_input_tokens_seen": 11948130, + "step": 371 + }, + { + "epoch": 0.033548270730937454, + "flos": 21731855338080.0, + "grad_norm": 2.144203702999903, + "learning_rate": 3.999870221295936e-06, + "loss": 0.7738, + "num_input_tokens_seen": 11976850, + "step": 372 + }, + { + "epoch": 0.03363845425440772, + "flos": 34263456434400.0, + "grad_norm": 1.8254511262551822, + "learning_rate": 3.999863480729875e-06, + "loss": 0.787, + "num_input_tokens_seen": 12009675, + "step": 373 + }, + { + "epoch": 0.03372863777787798, + "flos": 27630063963360.0, + "grad_norm": 3.2919265708845487, + "learning_rate": 3.999856569524433e-06, + "loss": 0.7589, + "num_input_tokens_seen": 12038820, + "step": 374 + }, + { + "epoch": 0.033818821301348244, + "flos": 21073438513920.0, + "grad_norm": 1.7463905713702752, + "learning_rate": 3.999849487680202e-06, + "loss": 0.7604, + "num_input_tokens_seen": 12067815, + "step": 375 + }, + { + "epoch": 0.033909004824818505, + "flos": 39620498859360.0, + "grad_norm": 1.6053789768040532, + "learning_rate": 3.999842235197786e-06, + "loss": 0.7538, + "num_input_tokens_seen": 12100470, + "step": 376 + }, + { + "epoch": 0.033999188348288766, + "flos": 28399040854560.0, + "grad_norm": 2.4347875125610536, + "learning_rate": 3.999834812077803e-06, + "loss": 0.7529, + "num_input_tokens_seen": 12130575, + "step": 377 + }, + { + "epoch": 0.034089371871759026, + "flos": 25557337179360.0, + "grad_norm": 2.113565647922343, + "learning_rate": 3.999827218320886e-06, + "loss": 0.8727, + "num_input_tokens_seen": 12160305, + "step": 378 + }, + { + "epoch": 0.034179555395229294, + "flos": 19360795361280.0, + "grad_norm": 2.32607860362178, + "learning_rate": 3.999819453927685e-06, + "loss": 0.7336, + "num_input_tokens_seen": 12188015, + "step": 379 + }, + { + "epoch": 0.034269738918699555, + "flos": 18998890411680.0, + "grad_norm": 1.9398712905975162, + "learning_rate": 3.999811518898861e-06, + "loss": 0.8195, + "num_input_tokens_seen": 12214610, + "step": 380 + }, + { + "epoch": 0.034359922442169816, + "flos": 16229716405440.0, + "grad_norm": 2.45263074038379, + "learning_rate": 3.999803413235092e-06, + "loss": 0.834, + "num_input_tokens_seen": 12242205, + "step": 381 + }, + { + "epoch": 0.03445010596564008, + "flos": 21185150843520.0, + "grad_norm": 1.99027762024441, + "learning_rate": 3.999795136937068e-06, + "loss": 0.9289, + "num_input_tokens_seen": 12270525, + "step": 382 + }, + { + "epoch": 0.03454028948911034, + "flos": 21694679844480.0, + "grad_norm": 1.999522450177246, + "learning_rate": 3.999786690005496e-06, + "loss": 0.8077, + "num_input_tokens_seen": 12298770, + "step": 383 + }, + { + "epoch": 0.0346304730125806, + "flos": 24317567910720.0, + "grad_norm": 2.2362134079474143, + "learning_rate": 3.999778072441098e-06, + "loss": 0.8026, + "num_input_tokens_seen": 12329845, + "step": 384 + }, + { + "epoch": 0.034720656536050866, + "flos": 20054900888640.0, + "grad_norm": 2.693393012729577, + "learning_rate": 3.999769284244608e-06, + "loss": 0.8488, + "num_input_tokens_seen": 12356150, + "step": 385 + }, + { + "epoch": 0.03481084005952113, + "flos": 28976787832320.0, + "grad_norm": 1.5027439927263628, + "learning_rate": 3.999760325416775e-06, + "loss": 0.836, + "num_input_tokens_seen": 12387385, + "step": 386 + }, + { + "epoch": 0.03490102358299139, + "flos": 22460162778240.0, + "grad_norm": 1.770667109448213, + "learning_rate": 3.999751195958366e-06, + "loss": 0.9061, + "num_input_tokens_seen": 12416085, + "step": 387 + }, + { + "epoch": 0.03499120710646165, + "flos": 59128912538400.0, + "grad_norm": 0.792734459151157, + "learning_rate": 3.999741895870157e-06, + "loss": 0.5985, + "num_input_tokens_seen": 12507880, + "step": 388 + }, + { + "epoch": 0.03508139062993191, + "flos": 21216788042880.0, + "grad_norm": 4.057710142023415, + "learning_rate": 3.999732425152944e-06, + "loss": 0.8639, + "num_input_tokens_seen": 12533705, + "step": 389 + }, + { + "epoch": 0.03517157415340217, + "flos": 23406914129760.0, + "grad_norm": 1.6440745827354675, + "learning_rate": 3.999722783807533e-06, + "loss": 0.8791, + "num_input_tokens_seen": 12564120, + "step": 390 + }, + { + "epoch": 0.03526175767687244, + "flos": 33025657163040.0, + "grad_norm": 2.655119442165785, + "learning_rate": 3.999712971834748e-06, + "loss": 0.7522, + "num_input_tokens_seen": 12593590, + "step": 391 + }, + { + "epoch": 0.0353519412003427, + "flos": 17578373101920.0, + "grad_norm": 1.8292318400573409, + "learning_rate": 3.999702989235427e-06, + "loss": 0.8361, + "num_input_tokens_seen": 12621770, + "step": 392 + }, + { + "epoch": 0.03544212472381296, + "flos": 22093946136480.0, + "grad_norm": 2.9838943493832435, + "learning_rate": 3.999692836010419e-06, + "loss": 0.9127, + "num_input_tokens_seen": 12648465, + "step": 393 + }, + { + "epoch": 0.03553230824728322, + "flos": 17577406688160.0, + "grad_norm": 2.351039181900796, + "learning_rate": 3.999682512160593e-06, + "loss": 0.9621, + "num_input_tokens_seen": 12672705, + "step": 394 + }, + { + "epoch": 0.03562249177075348, + "flos": 59439591825120.0, + "grad_norm": 0.7735432253930699, + "learning_rate": 3.99967201768683e-06, + "loss": 0.5766, + "num_input_tokens_seen": 12757070, + "step": 395 + }, + { + "epoch": 0.03571267529422374, + "flos": 21294298459680.0, + "grad_norm": 3.3547015662002786, + "learning_rate": 3.999661352590023e-06, + "loss": 0.8091, + "num_input_tokens_seen": 12785635, + "step": 396 + }, + { + "epoch": 0.03580285881769401, + "flos": 25407111244800.0, + "grad_norm": 3.514502587754212, + "learning_rate": 3.999650516871083e-06, + "loss": 0.8261, + "num_input_tokens_seen": 12815180, + "step": 397 + }, + { + "epoch": 0.03589304234116427, + "flos": 22532543768160.0, + "grad_norm": 1.9090574883966422, + "learning_rate": 3.9996395105309365e-06, + "loss": 0.8487, + "num_input_tokens_seen": 12843385, + "step": 398 + }, + { + "epoch": 0.03598322586463453, + "flos": 28468411093920.0, + "grad_norm": 1.665958023417146, + "learning_rate": 3.99962833357052e-06, + "loss": 0.8488, + "num_input_tokens_seen": 12873435, + "step": 399 + }, + { + "epoch": 0.03607340938810479, + "flos": 26905250480640.0, + "grad_norm": 2.2534806562075502, + "learning_rate": 3.999616985990789e-06, + "loss": 0.8048, + "num_input_tokens_seen": 12903180, + "step": 400 + }, + { + "epoch": 0.03616359291157505, + "flos": 24099421357440.0, + "grad_norm": 2.011855630014064, + "learning_rate": 3.9996054677927104e-06, + "loss": 0.7611, + "num_input_tokens_seen": 12933930, + "step": 401 + }, + { + "epoch": 0.03625377643504532, + "flos": 26210178539520.0, + "grad_norm": 2.786760593075569, + "learning_rate": 3.9995937789772675e-06, + "loss": 0.7645, + "num_input_tokens_seen": 12962935, + "step": 402 + }, + { + "epoch": 0.03634395995851558, + "flos": 72012196900320.0, + "grad_norm": 0.726054475435064, + "learning_rate": 3.999581919545458e-06, + "loss": 0.602, + "num_input_tokens_seen": 13055140, + "step": 403 + }, + { + "epoch": 0.03643414348198584, + "flos": 18197086888800.0, + "grad_norm": 3.023555309750527, + "learning_rate": 3.9995698894982935e-06, + "loss": 0.8476, + "num_input_tokens_seen": 13078670, + "step": 404 + }, + { + "epoch": 0.0365243270054561, + "flos": 23734059620160.0, + "grad_norm": 1.6815251433698617, + "learning_rate": 3.9995576888368e-06, + "loss": 0.8738, + "num_input_tokens_seen": 13109050, + "step": 405 + }, + { + "epoch": 0.036614510528926364, + "flos": 21694717014240.0, + "grad_norm": 2.0490315266348524, + "learning_rate": 3.9995453175620194e-06, + "loss": 0.8119, + "num_input_tokens_seen": 13136915, + "step": 406 + }, + { + "epoch": 0.036704694052396625, + "flos": 64390831813920.0, + "grad_norm": 3.310617883871567, + "learning_rate": 3.999532775675007e-06, + "loss": 0.6794, + "num_input_tokens_seen": 13164455, + "step": 407 + }, + { + "epoch": 0.03679487757586689, + "flos": 35903607087840.0, + "grad_norm": 1.9575332332825202, + "learning_rate": 3.9995200631768326e-06, + "loss": 0.7956, + "num_input_tokens_seen": 13195475, + "step": 408 + }, + { + "epoch": 0.03688506109933715, + "flos": 16922595330720.0, + "grad_norm": 2.4724462338610107, + "learning_rate": 3.9995071800685815e-06, + "loss": 0.9498, + "num_input_tokens_seen": 13220215, + "step": 409 + }, + { + "epoch": 0.036975244622807414, + "flos": 24937433960160.0, + "grad_norm": 2.632725745336702, + "learning_rate": 3.999494126351352e-06, + "loss": 0.7943, + "num_input_tokens_seen": 13246735, + "step": 410 + }, + { + "epoch": 0.037065428146277675, + "flos": 23807964570240.0, + "grad_norm": 3.3851792159315006, + "learning_rate": 3.99948090202626e-06, + "loss": 0.829, + "num_input_tokens_seen": 13278145, + "step": 411 + }, + { + "epoch": 0.037155611669747936, + "flos": 36776007452160.0, + "grad_norm": 1.7333985750440086, + "learning_rate": 3.999467507094431e-06, + "loss": 0.8263, + "num_input_tokens_seen": 13310430, + "step": 412 + }, + { + "epoch": 0.0372457951932182, + "flos": 32841601013280.0, + "grad_norm": 1.8716034961388144, + "learning_rate": 3.999453941557011e-06, + "loss": 0.7279, + "num_input_tokens_seen": 13342040, + "step": 413 + }, + { + "epoch": 0.037335978716688464, + "flos": 60179496230400.0, + "grad_norm": 0.7299851229209908, + "learning_rate": 3.999440205415154e-06, + "loss": 0.5552, + "num_input_tokens_seen": 13432250, + "step": 414 + }, + { + "epoch": 0.037426162240158725, + "flos": 18051507174240.0, + "grad_norm": 2.4360804214293976, + "learning_rate": 3.999426298670035e-06, + "loss": 0.8706, + "num_input_tokens_seen": 13458855, + "step": 415 + }, + { + "epoch": 0.037516345763628986, + "flos": 22271497578240.0, + "grad_norm": 1.9300256965586835, + "learning_rate": 3.9994122213228385e-06, + "loss": 0.8661, + "num_input_tokens_seen": 13486285, + "step": 416 + }, + { + "epoch": 0.03760652928709925, + "flos": 24604006780320.0, + "grad_norm": 1.9917303420971657, + "learning_rate": 3.9993979733747675e-06, + "loss": 0.8527, + "num_input_tokens_seen": 13515230, + "step": 417 + }, + { + "epoch": 0.03769671281056951, + "flos": 26212483064640.0, + "grad_norm": 2.2215309445400098, + "learning_rate": 3.999383554827037e-06, + "loss": 0.8266, + "num_input_tokens_seen": 13543590, + "step": 418 + }, + { + "epoch": 0.03778689633403977, + "flos": 23917075016640.0, + "grad_norm": 1.9638398717254328, + "learning_rate": 3.999368965680876e-06, + "loss": 0.7834, + "num_input_tokens_seen": 13572800, + "step": 419 + }, + { + "epoch": 0.037877079857510036, + "flos": 22678755368640.0, + "grad_norm": 2.1140663254076624, + "learning_rate": 3.999354205937531e-06, + "loss": 0.8675, + "num_input_tokens_seen": 13602120, + "step": 420 + }, + { + "epoch": 0.0379672633809803, + "flos": 21986805687360.0, + "grad_norm": 1.984358900110008, + "learning_rate": 3.999339275598261e-06, + "loss": 0.8683, + "num_input_tokens_seen": 13628700, + "step": 421 + }, + { + "epoch": 0.03805744690445056, + "flos": 32913982003200.0, + "grad_norm": 2.8668661743054824, + "learning_rate": 3.99932417466434e-06, + "loss": 0.8048, + "num_input_tokens_seen": 13661060, + "step": 422 + }, + { + "epoch": 0.03814763042792082, + "flos": 21288722995680.0, + "grad_norm": 3.2177304625354823, + "learning_rate": 3.999308903137056e-06, + "loss": 0.7354, + "num_input_tokens_seen": 13688725, + "step": 423 + }, + { + "epoch": 0.03823781395139108, + "flos": 23079768639360.0, + "grad_norm": 2.290654901202669, + "learning_rate": 3.999293461017711e-06, + "loss": 0.8736, + "num_input_tokens_seen": 13718415, + "step": 424 + }, + { + "epoch": 0.03832799747486134, + "flos": 19249789257120.0, + "grad_norm": 1.9150503917866692, + "learning_rate": 3.9992778483076255e-06, + "loss": 0.8123, + "num_input_tokens_seen": 13744975, + "step": 425 + }, + { + "epoch": 0.03841818099833161, + "flos": 19867202102400.0, + "grad_norm": 1.8733764613041428, + "learning_rate": 3.99926206500813e-06, + "loss": 0.8798, + "num_input_tokens_seen": 13772530, + "step": 426 + }, + { + "epoch": 0.03850836452180187, + "flos": 16557493781760.0, + "grad_norm": 2.164931250310965, + "learning_rate": 3.999246111120571e-06, + "loss": 0.7509, + "num_input_tokens_seen": 13798285, + "step": 427 + }, + { + "epoch": 0.03859854804527213, + "flos": 25770094117440.0, + "grad_norm": 1.5918898853673924, + "learning_rate": 3.999229986646311e-06, + "loss": 0.8956, + "num_input_tokens_seen": 13826625, + "step": 428 + }, + { + "epoch": 0.03868873156874239, + "flos": 29309583126240.0, + "grad_norm": 4.19330888636932, + "learning_rate": 3.999213691586723e-06, + "loss": 0.7926, + "num_input_tokens_seen": 13858335, + "step": 429 + }, + { + "epoch": 0.03877891509221265, + "flos": 31494616956000.0, + "grad_norm": 1.7634225797993768, + "learning_rate": 3.9991972259432e-06, + "loss": 0.8658, + "num_input_tokens_seen": 13888455, + "step": 430 + }, + { + "epoch": 0.03886909861568291, + "flos": 24458798763360.0, + "grad_norm": 1.5614415207513985, + "learning_rate": 3.999180589717147e-06, + "loss": 0.9067, + "num_input_tokens_seen": 13918300, + "step": 431 + }, + { + "epoch": 0.03895928213915318, + "flos": 19400052361440.0, + "grad_norm": 2.8620447921286765, + "learning_rate": 3.999163782909983e-06, + "loss": 0.7365, + "num_input_tokens_seen": 13945050, + "step": 432 + }, + { + "epoch": 0.03904946566262344, + "flos": 27596716955040.0, + "grad_norm": 1.7731045313753953, + "learning_rate": 3.99914680552314e-06, + "loss": 0.8573, + "num_input_tokens_seen": 13974120, + "step": 433 + }, + { + "epoch": 0.0391396491860937, + "flos": 27816052940640.0, + "grad_norm": 2.172912130476062, + "learning_rate": 3.999129657558069e-06, + "loss": 0.7524, + "num_input_tokens_seen": 14005540, + "step": 434 + }, + { + "epoch": 0.03922983270956396, + "flos": 64380864584640.0, + "grad_norm": 0.7215456289612991, + "learning_rate": 3.999112339016234e-06, + "loss": 0.5683, + "num_input_tokens_seen": 14088300, + "step": 435 + }, + { + "epoch": 0.03932001623303422, + "flos": 20200926640320.0, + "grad_norm": 2.031814848784808, + "learning_rate": 3.999094849899109e-06, + "loss": 0.8357, + "num_input_tokens_seen": 14116265, + "step": 436 + }, + { + "epoch": 0.039410199756504484, + "flos": 20784211912320.0, + "grad_norm": 1.9750025326676366, + "learning_rate": 3.99907719020819e-06, + "loss": 0.7423, + "num_input_tokens_seen": 14144870, + "step": 437 + }, + { + "epoch": 0.03950038327997475, + "flos": 19290718896480.0, + "grad_norm": 2.4890127020174972, + "learning_rate": 3.999059359944982e-06, + "loss": 0.8511, + "num_input_tokens_seen": 14173715, + "step": 438 + }, + { + "epoch": 0.03959056680344501, + "flos": 35464972286400.0, + "grad_norm": 1.6192051192485415, + "learning_rate": 3.999041359111007e-06, + "loss": 0.8775, + "num_input_tokens_seen": 14205700, + "step": 439 + }, + { + "epoch": 0.03968075032691527, + "flos": 23771532471840.0, + "grad_norm": 2.6832826036985065, + "learning_rate": 3.999023187707801e-06, + "loss": 0.8954, + "num_input_tokens_seen": 14232815, + "step": 440 + }, + { + "epoch": 0.039770933850385534, + "flos": 21293815252800.0, + "grad_norm": 1.711819147241302, + "learning_rate": 3.999004845736913e-06, + "loss": 0.8864, + "num_input_tokens_seen": 14264895, + "step": 441 + }, + { + "epoch": 0.039861117373855795, + "flos": 25885226064960.0, + "grad_norm": 2.296494017683128, + "learning_rate": 3.9989863331999096e-06, + "loss": 0.8979, + "num_input_tokens_seen": 14295700, + "step": 442 + }, + { + "epoch": 0.039951300897326056, + "flos": 21476050084320.0, + "grad_norm": 2.4577111787761496, + "learning_rate": 3.99896765009837e-06, + "loss": 0.7556, + "num_input_tokens_seen": 14324090, + "step": 443 + }, + { + "epoch": 0.040041484420796324, + "flos": 23951053910880.0, + "grad_norm": 1.7063015353625772, + "learning_rate": 3.998948796433888e-06, + "loss": 0.7913, + "num_input_tokens_seen": 14354845, + "step": 444 + }, + { + "epoch": 0.040131667944266584, + "flos": 23843616103680.0, + "grad_norm": 4.255044523873574, + "learning_rate": 3.998929772208073e-06, + "loss": 0.8646, + "num_input_tokens_seen": 14384780, + "step": 445 + }, + { + "epoch": 0.040221851467736845, + "flos": 70347173943840.0, + "grad_norm": 0.6805027643367614, + "learning_rate": 3.998910577422547e-06, + "loss": 0.5952, + "num_input_tokens_seen": 14483575, + "step": 446 + }, + { + "epoch": 0.040312034991207106, + "flos": 25516853577120.0, + "grad_norm": 1.653001153990521, + "learning_rate": 3.99889121207895e-06, + "loss": 0.8638, + "num_input_tokens_seen": 14513215, + "step": 447 + }, + { + "epoch": 0.04040221851467737, + "flos": 17286172919520.0, + "grad_norm": 1.7836166854900934, + "learning_rate": 3.9988716761789324e-06, + "loss": 0.8485, + "num_input_tokens_seen": 14541015, + "step": 448 + }, + { + "epoch": 0.04049240203814763, + "flos": 21112026458400.0, + "grad_norm": 2.2272432056374325, + "learning_rate": 3.998851969724161e-06, + "loss": 0.7624, + "num_input_tokens_seen": 14568470, + "step": 449 + }, + { + "epoch": 0.040582585561617895, + "flos": 20492346257760.0, + "grad_norm": 2.245456983447726, + "learning_rate": 3.998832092716319e-06, + "loss": 0.8731, + "num_input_tokens_seen": 14593630, + "step": 450 + }, + { + "epoch": 0.040672769085088156, + "flos": 26539925913120.0, + "grad_norm": 1.6287822721948026, + "learning_rate": 3.998812045157102e-06, + "loss": 0.802, + "num_input_tokens_seen": 14627175, + "step": 451 + }, + { + "epoch": 0.04076295260855842, + "flos": 19435964083200.0, + "grad_norm": 2.15720324246817, + "learning_rate": 3.998791827048219e-06, + "loss": 0.873, + "num_input_tokens_seen": 14652825, + "step": 452 + }, + { + "epoch": 0.04085313613202868, + "flos": 28321344588960.0, + "grad_norm": 2.279692130969797, + "learning_rate": 3.998771438391396e-06, + "loss": 0.7655, + "num_input_tokens_seen": 14682550, + "step": 453 + }, + { + "epoch": 0.04094331965549894, + "flos": 65365311806400.0, + "grad_norm": 0.6604202916219326, + "learning_rate": 3.9987508791883725e-06, + "loss": 0.5738, + "num_input_tokens_seen": 14775700, + "step": 454 + }, + { + "epoch": 0.0410335031789692, + "flos": 16557828309600.0, + "grad_norm": 3.198801090239434, + "learning_rate": 3.998730149440904e-06, + "loss": 0.9436, + "num_input_tokens_seen": 14800335, + "step": 455 + }, + { + "epoch": 0.04112368670243947, + "flos": 24386938150080.0, + "grad_norm": 2.31800126071267, + "learning_rate": 3.998709249150758e-06, + "loss": 0.8739, + "num_input_tokens_seen": 14827205, + "step": 456 + }, + { + "epoch": 0.04121387022590973, + "flos": 23218137420480.0, + "grad_norm": 3.01170608379308, + "learning_rate": 3.998688178319717e-06, + "loss": 0.8486, + "num_input_tokens_seen": 14855235, + "step": 457 + }, + { + "epoch": 0.04130405374937999, + "flos": 20857968183360.0, + "grad_norm": 1.783888351966769, + "learning_rate": 3.9986669369495805e-06, + "loss": 0.8091, + "num_input_tokens_seen": 14881760, + "step": 458 + }, + { + "epoch": 0.04139423727285025, + "flos": 22095507266400.0, + "grad_norm": 2.3715296035196087, + "learning_rate": 3.998645525042161e-06, + "loss": 0.8788, + "num_input_tokens_seen": 14909805, + "step": 459 + }, + { + "epoch": 0.04148442079632051, + "flos": 13128416160960.0, + "grad_norm": 6.614443159245135, + "learning_rate": 3.998623942599284e-06, + "loss": 0.8937, + "num_input_tokens_seen": 14935840, + "step": 460 + }, + { + "epoch": 0.04157460431979077, + "flos": 25188853182240.0, + "grad_norm": 1.989781621136046, + "learning_rate": 3.998602189622793e-06, + "loss": 0.8184, + "num_input_tokens_seen": 14964945, + "step": 461 + }, + { + "epoch": 0.04166478784326104, + "flos": 18525533320800.0, + "grad_norm": 1.8938307892842914, + "learning_rate": 3.998580266114542e-06, + "loss": 0.8609, + "num_input_tokens_seen": 14992295, + "step": 462 + }, + { + "epoch": 0.0417549713667313, + "flos": 25994225002080.0, + "grad_norm": 1.703717959712149, + "learning_rate": 3.998558172076404e-06, + "loss": 0.7283, + "num_input_tokens_seen": 15021380, + "step": 463 + }, + { + "epoch": 0.04184515489020156, + "flos": 24639658313760.0, + "grad_norm": 1.8644099305858055, + "learning_rate": 3.998535907510262e-06, + "loss": 0.7549, + "num_input_tokens_seen": 15047520, + "step": 464 + }, + { + "epoch": 0.04193533841367182, + "flos": 25812138849600.0, + "grad_norm": 2.2854228068025555, + "learning_rate": 3.998513472418016e-06, + "loss": 0.8359, + "num_input_tokens_seen": 15077640, + "step": 465 + }, + { + "epoch": 0.04202552193714208, + "flos": 23079471281280.0, + "grad_norm": 2.1808235735365717, + "learning_rate": 3.998490866801582e-06, + "loss": 0.8586, + "num_input_tokens_seen": 15103940, + "step": 466 + }, + { + "epoch": 0.04211570546061234, + "flos": 18227869183680.0, + "grad_norm": 1.7720572860577026, + "learning_rate": 3.998468090662886e-06, + "loss": 0.8139, + "num_input_tokens_seen": 15131270, + "step": 467 + }, + { + "epoch": 0.04220588898408261, + "flos": 26139879056160.0, + "grad_norm": 1.7930704058915279, + "learning_rate": 3.998445144003874e-06, + "loss": 0.7204, + "num_input_tokens_seen": 15160850, + "step": 468 + }, + { + "epoch": 0.04229607250755287, + "flos": 21804645195360.0, + "grad_norm": 2.265201818472664, + "learning_rate": 3.998422026826504e-06, + "loss": 0.849, + "num_input_tokens_seen": 15189305, + "step": 469 + }, + { + "epoch": 0.04238625603102313, + "flos": 19253654912160.0, + "grad_norm": 2.6816241770196143, + "learning_rate": 3.998398739132746e-06, + "loss": 0.8538, + "num_input_tokens_seen": 15216675, + "step": 470 + }, + { + "epoch": 0.04247643955449339, + "flos": 25299859286400.0, + "grad_norm": 1.9391307735796102, + "learning_rate": 3.99837528092459e-06, + "loss": 0.7811, + "num_input_tokens_seen": 15244490, + "step": 471 + }, + { + "epoch": 0.042566623077963654, + "flos": 23257394420640.0, + "grad_norm": 2.3744726097405207, + "learning_rate": 3.998351652204034e-06, + "loss": 0.8726, + "num_input_tokens_seen": 15272295, + "step": 472 + }, + { + "epoch": 0.042656806601433915, + "flos": 29746582458240.0, + "grad_norm": 1.9926227168986694, + "learning_rate": 3.998327852973098e-06, + "loss": 0.867, + "num_input_tokens_seen": 15304985, + "step": 473 + }, + { + "epoch": 0.04274699012490418, + "flos": 14299261227360.0, + "grad_norm": 2.5783564608160723, + "learning_rate": 3.99830388323381e-06, + "loss": 0.7734, + "num_input_tokens_seen": 15330850, + "step": 474 + }, + { + "epoch": 0.042837173648374444, + "flos": 22202090169120.0, + "grad_norm": 1.866587969562931, + "learning_rate": 3.998279742988216e-06, + "loss": 0.8554, + "num_input_tokens_seen": 15359080, + "step": 475 + }, + { + "epoch": 0.042927357171844704, + "flos": 19655783275680.0, + "grad_norm": 1.638442849851347, + "learning_rate": 3.998255432238377e-06, + "loss": 0.7923, + "num_input_tokens_seen": 15389510, + "step": 476 + }, + { + "epoch": 0.043017540695314965, + "flos": 22751285037600.0, + "grad_norm": 4.17776587689736, + "learning_rate": 3.9982309509863656e-06, + "loss": 0.8984, + "num_input_tokens_seen": 15417930, + "step": 477 + }, + { + "epoch": 0.043107724218785226, + "flos": 26394940914720.0, + "grad_norm": 3.164716789507094, + "learning_rate": 3.998206299234272e-06, + "loss": 0.7719, + "num_input_tokens_seen": 15447615, + "step": 478 + }, + { + "epoch": 0.04319790774225549, + "flos": 25265099827200.0, + "grad_norm": 2.322170245984706, + "learning_rate": 3.998181476984198e-06, + "loss": 0.841, + "num_input_tokens_seen": 15477915, + "step": 479 + }, + { + "epoch": 0.043288091265725755, + "flos": 21294112610880.0, + "grad_norm": 2.011371115860162, + "learning_rate": 3.998156484238263e-06, + "loss": 0.7503, + "num_input_tokens_seen": 15506375, + "step": 480 + }, + { + "epoch": 0.043378274789196015, + "flos": 25045169125440.0, + "grad_norm": 2.385690920415955, + "learning_rate": 3.998131320998599e-06, + "loss": 0.7611, + "num_input_tokens_seen": 15534660, + "step": 481 + }, + { + "epoch": 0.043468458312666276, + "flos": 27925386405600.0, + "grad_norm": 2.4892592491843146, + "learning_rate": 3.998105987267353e-06, + "loss": 0.8712, + "num_input_tokens_seen": 15563860, + "step": 482 + }, + { + "epoch": 0.04355864183613654, + "flos": 25009926459360.0, + "grad_norm": 1.7553372057533845, + "learning_rate": 3.998080483046687e-06, + "loss": 0.7504, + "num_input_tokens_seen": 15593625, + "step": 483 + }, + { + "epoch": 0.0436488253596068, + "flos": 38925798615840.0, + "grad_norm": 1.7755549892975806, + "learning_rate": 3.998054808338776e-06, + "loss": 0.7739, + "num_input_tokens_seen": 15626815, + "step": 484 + }, + { + "epoch": 0.04373900888307706, + "flos": 22459493722560.0, + "grad_norm": 2.102157338417814, + "learning_rate": 3.998028963145812e-06, + "loss": 0.7849, + "num_input_tokens_seen": 15654870, + "step": 485 + }, + { + "epoch": 0.043829192406547327, + "flos": 20529410242080.0, + "grad_norm": 1.9944381865483218, + "learning_rate": 3.99800294747e-06, + "loss": 0.837, + "num_input_tokens_seen": 15682775, + "step": 486 + }, + { + "epoch": 0.04391937593001759, + "flos": 28289744559360.0, + "grad_norm": 1.7139315948991103, + "learning_rate": 3.99797676131356e-06, + "loss": 0.8592, + "num_input_tokens_seen": 15713090, + "step": 487 + }, + { + "epoch": 0.04400955945348785, + "flos": 25046172708960.0, + "grad_norm": 3.047981100559463, + "learning_rate": 3.997950404678726e-06, + "loss": 0.6813, + "num_input_tokens_seen": 15738190, + "step": 488 + }, + { + "epoch": 0.04409974297695811, + "flos": 18561259193760.0, + "grad_norm": 2.4225707008587345, + "learning_rate": 3.997923877567746e-06, + "loss": 0.9383, + "num_input_tokens_seen": 15765720, + "step": 489 + }, + { + "epoch": 0.04418992650042837, + "flos": 35172809273760.0, + "grad_norm": 1.6000627147781255, + "learning_rate": 3.9978971799828855e-06, + "loss": 0.653, + "num_input_tokens_seen": 15798950, + "step": 490 + }, + { + "epoch": 0.04428011002389863, + "flos": 23225422693440.0, + "grad_norm": 5.9463576251641355, + "learning_rate": 3.997870311926421e-06, + "loss": 0.8143, + "num_input_tokens_seen": 15827285, + "step": 491 + }, + { + "epoch": 0.0443702935473689, + "flos": 25448152393440.0, + "grad_norm": 3.8271207776813694, + "learning_rate": 3.997843273400645e-06, + "loss": 0.7809, + "num_input_tokens_seen": 15855320, + "step": 492 + }, + { + "epoch": 0.04446047707083916, + "flos": 28835631319200.0, + "grad_norm": 2.0840514153491627, + "learning_rate": 3.997816064407865e-06, + "loss": 0.7999, + "num_input_tokens_seen": 15882870, + "step": 493 + }, + { + "epoch": 0.04455066059430942, + "flos": 42169258956960.0, + "grad_norm": 1.6271111639412925, + "learning_rate": 3.997788684950402e-06, + "loss": 0.8434, + "num_input_tokens_seen": 15915905, + "step": 494 + }, + { + "epoch": 0.04464084411777968, + "flos": 19909767211200.0, + "grad_norm": 1.9417271799745779, + "learning_rate": 3.997761135030593e-06, + "loss": 0.8669, + "num_input_tokens_seen": 15943965, + "step": 495 + }, + { + "epoch": 0.04473102764124994, + "flos": 23331299370720.0, + "grad_norm": 1.8343139612287653, + "learning_rate": 3.997733414650789e-06, + "loss": 0.7238, + "num_input_tokens_seen": 15972710, + "step": 496 + }, + { + "epoch": 0.0448212111647202, + "flos": 28070594422560.0, + "grad_norm": 2.907186541860079, + "learning_rate": 3.9977055238133554e-06, + "loss": 0.7827, + "num_input_tokens_seen": 16000410, + "step": 497 + }, + { + "epoch": 0.04491139468819047, + "flos": 26431261503840.0, + "grad_norm": 4.136940061360457, + "learning_rate": 3.99767746252067e-06, + "loss": 0.7028, + "num_input_tokens_seen": 16031125, + "step": 498 + }, + { + "epoch": 0.04500157821166073, + "flos": 26139581698080.0, + "grad_norm": 1.9556654620472746, + "learning_rate": 3.997649230775129e-06, + "loss": 0.8262, + "num_input_tokens_seen": 16059995, + "step": 499 + }, + { + "epoch": 0.04509176173513099, + "flos": 24719919292800.0, + "grad_norm": 2.2343294365778847, + "learning_rate": 3.9976208285791395e-06, + "loss": 0.7614, + "num_input_tokens_seen": 16089250, + "step": 500 + }, + { + "epoch": 0.04518194525860125, + "flos": 70225277100000.0, + "grad_norm": 0.7747091762869349, + "learning_rate": 3.997592255935127e-06, + "loss": 0.6706, + "num_input_tokens_seen": 16178085, + "step": 501 + }, + { + "epoch": 0.045272128782071513, + "flos": 25552542280320.0, + "grad_norm": 1.982557290248797, + "learning_rate": 3.997563512845529e-06, + "loss": 0.8487, + "num_input_tokens_seen": 16209510, + "step": 502 + }, + { + "epoch": 0.045362312305541774, + "flos": 22315177779840.0, + "grad_norm": 1.8946492895733915, + "learning_rate": 3.9975345993127975e-06, + "loss": 0.7804, + "num_input_tokens_seen": 16237985, + "step": 503 + }, + { + "epoch": 0.04545249582901204, + "flos": 20529410242080.0, + "grad_norm": 2.24987888010181, + "learning_rate": 3.9975055153393985e-06, + "loss": 0.835, + "num_input_tokens_seen": 16266455, + "step": 504 + }, + { + "epoch": 0.0455426793524823, + "flos": 23731717925280.0, + "grad_norm": 1.84351075770046, + "learning_rate": 3.997476260927816e-06, + "loss": 0.842, + "num_input_tokens_seen": 16293850, + "step": 505 + }, + { + "epoch": 0.045632862875952564, + "flos": 39616447355520.0, + "grad_norm": 1.8385902881046712, + "learning_rate": 3.997446836080545e-06, + "loss": 0.7973, + "num_input_tokens_seen": 16325260, + "step": 506 + }, + { + "epoch": 0.045723046399422825, + "flos": 24317716589760.0, + "grad_norm": 2.381265907534834, + "learning_rate": 3.997417240800095e-06, + "loss": 0.8745, + "num_input_tokens_seen": 16352195, + "step": 507 + }, + { + "epoch": 0.045813229922893085, + "flos": 23001923694720.0, + "grad_norm": 3.579384527951417, + "learning_rate": 3.997387475088994e-06, + "loss": 0.9224, + "num_input_tokens_seen": 16379090, + "step": 508 + }, + { + "epoch": 0.045903413446363346, + "flos": 18124408540800.0, + "grad_norm": 2.5983316791065065, + "learning_rate": 3.99735753894978e-06, + "loss": 0.9206, + "num_input_tokens_seen": 16405720, + "step": 509 + }, + { + "epoch": 0.045993596969833614, + "flos": 26030657100480.0, + "grad_norm": 1.9886189531827463, + "learning_rate": 3.997327432385006e-06, + "loss": 0.8231, + "num_input_tokens_seen": 16436105, + "step": 510 + }, + { + "epoch": 0.046083780493303875, + "flos": 19727383700640.0, + "grad_norm": 1.9542723680660343, + "learning_rate": 3.997297155397244e-06, + "loss": 0.7535, + "num_input_tokens_seen": 16463890, + "step": 511 + }, + { + "epoch": 0.046173964016774136, + "flos": 25083013674720.0, + "grad_norm": 1.8373392800348842, + "learning_rate": 3.997266707989074e-06, + "loss": 0.743, + "num_input_tokens_seen": 16492680, + "step": 512 + }, + { + "epoch": 0.046264147540244396, + "flos": 18998853241920.0, + "grad_norm": 2.4312260408373194, + "learning_rate": 3.997236090163097e-06, + "loss": 0.7451, + "num_input_tokens_seen": 16520870, + "step": 513 + }, + { + "epoch": 0.04635433106371466, + "flos": 20602423117920.0, + "grad_norm": 2.633357322760654, + "learning_rate": 3.9972053019219235e-06, + "loss": 0.9553, + "num_input_tokens_seen": 16549885, + "step": 514 + }, + { + "epoch": 0.04644451458718492, + "flos": 18342852452160.0, + "grad_norm": 2.7009105983543282, + "learning_rate": 3.997174343268181e-06, + "loss": 0.8386, + "num_input_tokens_seen": 16575935, + "step": 515 + }, + { + "epoch": 0.046534698110655186, + "flos": 33571172225280.0, + "grad_norm": 3.1599191923311882, + "learning_rate": 3.9971432142045115e-06, + "loss": 0.8306, + "num_input_tokens_seen": 16607545, + "step": 516 + }, + { + "epoch": 0.04662488163412545, + "flos": 68849480745120.0, + "grad_norm": 0.7989540280686571, + "learning_rate": 3.99711191473357e-06, + "loss": 0.6619, + "num_input_tokens_seen": 16691105, + "step": 517 + }, + { + "epoch": 0.04671506515759571, + "flos": 26573384430720.0, + "grad_norm": 2.0380792695371874, + "learning_rate": 3.99708044485803e-06, + "loss": 0.8386, + "num_input_tokens_seen": 16721380, + "step": 518 + }, + { + "epoch": 0.04680524868106597, + "flos": 22237221325920.0, + "grad_norm": 2.1331283838450354, + "learning_rate": 3.997048804580574e-06, + "loss": 0.8292, + "num_input_tokens_seen": 16749795, + "step": 519 + }, + { + "epoch": 0.04689543220453623, + "flos": 38957993361600.0, + "grad_norm": 2.187312417843124, + "learning_rate": 3.997016993903901e-06, + "loss": 0.7719, + "num_input_tokens_seen": 16780335, + "step": 520 + }, + { + "epoch": 0.04698561572800649, + "flos": 22929542704800.0, + "grad_norm": 2.073992068445679, + "learning_rate": 3.996985012830728e-06, + "loss": 0.9012, + "num_input_tokens_seen": 16808390, + "step": 521 + }, + { + "epoch": 0.04707579925147676, + "flos": 37688185193280.0, + "grad_norm": 1.605576857293893, + "learning_rate": 3.996952861363782e-06, + "loss": 0.7671, + "num_input_tokens_seen": 16843245, + "step": 522 + }, + { + "epoch": 0.04716598277494702, + "flos": 23953953152160.0, + "grad_norm": 1.9795300066534003, + "learning_rate": 3.9969205395058064e-06, + "loss": 0.827, + "num_input_tokens_seen": 16873525, + "step": 523 + }, + { + "epoch": 0.04725616629841728, + "flos": 22423879358880.0, + "grad_norm": 2.627313865652736, + "learning_rate": 3.99688804725956e-06, + "loss": 0.8192, + "num_input_tokens_seen": 16902335, + "step": 524 + }, + { + "epoch": 0.04734634982188754, + "flos": 57547990513440.0, + "grad_norm": 1.3947278052542031, + "learning_rate": 3.996855384627815e-06, + "loss": 0.7766, + "num_input_tokens_seen": 16941655, + "step": 525 + }, + { + "epoch": 0.0474365333453578, + "flos": 16623444403200.0, + "grad_norm": 2.2233056592958564, + "learning_rate": 3.996822551613357e-06, + "loss": 0.8581, + "num_input_tokens_seen": 16966820, + "step": 526 + }, + { + "epoch": 0.04752671686882806, + "flos": 34985630864160.0, + "grad_norm": 2.1838588383087356, + "learning_rate": 3.996789548218989e-06, + "loss": 0.7541, + "num_input_tokens_seen": 16997650, + "step": 527 + }, + { + "epoch": 0.04761690039229833, + "flos": 26686509211200.0, + "grad_norm": 1.7718391105340698, + "learning_rate": 3.996756374447526e-06, + "loss": 0.7095, + "num_input_tokens_seen": 17029415, + "step": 528 + }, + { + "epoch": 0.04770708391576859, + "flos": 33461206874400.0, + "grad_norm": 2.516809693960593, + "learning_rate": 3.9967230303018005e-06, + "loss": 0.8214, + "num_input_tokens_seen": 17056565, + "step": 529 + }, + { + "epoch": 0.04779726743923885, + "flos": 23328771827040.0, + "grad_norm": 2.0506348766587834, + "learning_rate": 3.996689515784655e-06, + "loss": 0.805, + "num_input_tokens_seen": 17083450, + "step": 530 + }, + { + "epoch": 0.04788745096270911, + "flos": 23400446591520.0, + "grad_norm": 2.7577838842187914, + "learning_rate": 3.996655830898951e-06, + "loss": 0.7949, + "num_input_tokens_seen": 17111230, + "step": 531 + }, + { + "epoch": 0.04797763448617937, + "flos": 16481358646080.0, + "grad_norm": 2.9191078505787598, + "learning_rate": 3.996621975647562e-06, + "loss": 0.8218, + "num_input_tokens_seen": 17137020, + "step": 532 + }, + { + "epoch": 0.04806781800964964, + "flos": 39400382308800.0, + "grad_norm": 2.100467893155766, + "learning_rate": 3.996587950033377e-06, + "loss": 0.7552, + "num_input_tokens_seen": 17171870, + "step": 533 + }, + { + "epoch": 0.0481580015331199, + "flos": 19035136661280.0, + "grad_norm": 2.7081358226257897, + "learning_rate": 3.996553754059299e-06, + "loss": 0.7464, + "num_input_tokens_seen": 17199960, + "step": 534 + }, + { + "epoch": 0.04824818505659016, + "flos": 21328240184160.0, + "grad_norm": 2.9546196601458883, + "learning_rate": 3.996519387728245e-06, + "loss": 0.8214, + "num_input_tokens_seen": 17229140, + "step": 535 + }, + { + "epoch": 0.04833836858006042, + "flos": 26758630012800.0, + "grad_norm": 1.3859565466912758, + "learning_rate": 3.9964848510431495e-06, + "loss": 0.8042, + "num_input_tokens_seen": 17259620, + "step": 536 + }, + { + "epoch": 0.048428552103530684, + "flos": 31308070432320.0, + "grad_norm": 1.982175494137675, + "learning_rate": 3.996450144006957e-06, + "loss": 0.8786, + "num_input_tokens_seen": 17289170, + "step": 537 + }, + { + "epoch": 0.048518735627000945, + "flos": 23987225820960.0, + "grad_norm": 2.128178631391025, + "learning_rate": 3.99641526662263e-06, + "loss": 0.7991, + "num_input_tokens_seen": 17319385, + "step": 538 + }, + { + "epoch": 0.04860891915047121, + "flos": 20163267939840.0, + "grad_norm": 1.9388118732547666, + "learning_rate": 3.996380218893145e-06, + "loss": 0.9033, + "num_input_tokens_seen": 17348555, + "step": 539 + }, + { + "epoch": 0.04869910267394147, + "flos": 28107100860480.0, + "grad_norm": 2.2093422462377217, + "learning_rate": 3.996345000821491e-06, + "loss": 0.8454, + "num_input_tokens_seen": 17378850, + "step": 540 + }, + { + "epoch": 0.048789286197411734, + "flos": 26431596031680.0, + "grad_norm": 2.055515021133563, + "learning_rate": 3.996309612410674e-06, + "loss": 0.9291, + "num_input_tokens_seen": 17408205, + "step": 541 + }, + { + "epoch": 0.048879469720881995, + "flos": 25156732776000.0, + "grad_norm": 1.9349810931048441, + "learning_rate": 3.996274053663713e-06, + "loss": 0.7561, + "num_input_tokens_seen": 17436690, + "step": 542 + }, + { + "epoch": 0.048969653244352256, + "flos": 69534591190560.0, + "grad_norm": 0.7895070325140813, + "learning_rate": 3.996238324583643e-06, + "loss": 0.618, + "num_input_tokens_seen": 17527420, + "step": 543 + }, + { + "epoch": 0.049059836767822516, + "flos": 21038753394240.0, + "grad_norm": 2.4513809905196604, + "learning_rate": 3.996202425173512e-06, + "loss": 0.8856, + "num_input_tokens_seen": 17553550, + "step": 544 + }, + { + "epoch": 0.049150020291292784, + "flos": 22095804624480.0, + "grad_norm": 2.0674652674201144, + "learning_rate": 3.996166355436383e-06, + "loss": 0.7768, + "num_input_tokens_seen": 17582665, + "step": 545 + }, + { + "epoch": 0.049240203814763045, + "flos": 21512519352480.0, + "grad_norm": 1.8750185011904599, + "learning_rate": 3.996130115375333e-06, + "loss": 0.8301, + "num_input_tokens_seen": 17611975, + "step": 546 + }, + { + "epoch": 0.049330387338233306, + "flos": 70656663798240.0, + "grad_norm": 0.8792182877076518, + "learning_rate": 3.996093704993456e-06, + "loss": 0.6335, + "num_input_tokens_seen": 17698215, + "step": 547 + }, + { + "epoch": 0.04942057086170357, + "flos": 19035545528640.0, + "grad_norm": 2.960934440188475, + "learning_rate": 3.996057124293857e-06, + "loss": 0.9031, + "num_input_tokens_seen": 17723965, + "step": 548 + }, + { + "epoch": 0.04951075438517383, + "flos": 26905361989920.0, + "grad_norm": 1.3661720646637594, + "learning_rate": 3.996020373279659e-06, + "loss": 0.8713, + "num_input_tokens_seen": 17755430, + "step": 549 + }, + { + "epoch": 0.04960093790864409, + "flos": 16848950568960.0, + "grad_norm": 3.685465364510014, + "learning_rate": 3.995983451953996e-06, + "loss": 0.8303, + "num_input_tokens_seen": 17782940, + "step": 550 + }, + { + "epoch": 0.049691121432114356, + "flos": 24972193419360.0, + "grad_norm": 1.7847763091260356, + "learning_rate": 3.99594636032002e-06, + "loss": 0.8424, + "num_input_tokens_seen": 17808990, + "step": 551 + }, + { + "epoch": 0.04978130495558462, + "flos": 29124374713920.0, + "grad_norm": 3.834275516876539, + "learning_rate": 3.995909098380894e-06, + "loss": 0.7069, + "num_input_tokens_seen": 17839470, + "step": 552 + }, + { + "epoch": 0.04987148847905488, + "flos": 29018944073760.0, + "grad_norm": 1.536908031107113, + "learning_rate": 3.995871666139799e-06, + "loss": 0.7056, + "num_input_tokens_seen": 17872720, + "step": 553 + }, + { + "epoch": 0.04996167200252514, + "flos": 23298732927360.0, + "grad_norm": 2.008812918588595, + "learning_rate": 3.995834063599928e-06, + "loss": 0.7988, + "num_input_tokens_seen": 17901855, + "step": 554 + }, + { + "epoch": 0.0500518555259954, + "flos": 25406553698400.0, + "grad_norm": 2.2844965577561944, + "learning_rate": 3.99579629076449e-06, + "loss": 0.8498, + "num_input_tokens_seen": 17931865, + "step": 555 + }, + { + "epoch": 0.05014203904946566, + "flos": 25776041279040.0, + "grad_norm": 2.0755695894099526, + "learning_rate": 3.9957583476367084e-06, + "loss": 0.7943, + "num_input_tokens_seen": 17960645, + "step": 556 + }, + { + "epoch": 0.05023222257293593, + "flos": 21658136236800.0, + "grad_norm": 2.5433604324050365, + "learning_rate": 3.995720234219819e-06, + "loss": 0.8184, + "num_input_tokens_seen": 17989625, + "step": 557 + }, + { + "epoch": 0.05032240609640619, + "flos": 18270397122720.0, + "grad_norm": 2.1070780363111226, + "learning_rate": 3.995681950517075e-06, + "loss": 0.7268, + "num_input_tokens_seen": 18018595, + "step": 558 + }, + { + "epoch": 0.05041258961987645, + "flos": 20602200099360.0, + "grad_norm": 2.241684169299162, + "learning_rate": 3.995643496531743e-06, + "loss": 0.8367, + "num_input_tokens_seen": 18045025, + "step": 559 + }, + { + "epoch": 0.05050277314334671, + "flos": 18926100554400.0, + "grad_norm": 2.3464603265984567, + "learning_rate": 3.9956048722671044e-06, + "loss": 0.8306, + "num_input_tokens_seen": 18070625, + "step": 560 + }, + { + "epoch": 0.05059295666681697, + "flos": 26358806174400.0, + "grad_norm": 1.7672412192951767, + "learning_rate": 3.995566077726454e-06, + "loss": 0.7993, + "num_input_tokens_seen": 18101340, + "step": 561 + }, + { + "epoch": 0.05068314019028723, + "flos": 22715187467040.0, + "grad_norm": 2.3031928954345795, + "learning_rate": 3.995527112913103e-06, + "loss": 0.7816, + "num_input_tokens_seen": 18130560, + "step": 562 + }, + { + "epoch": 0.0507733237137575, + "flos": 23917223695680.0, + "grad_norm": 2.5878310171496, + "learning_rate": 3.995487977830375e-06, + "loss": 0.7713, + "num_input_tokens_seen": 18160205, + "step": 563 + }, + { + "epoch": 0.05086350723722776, + "flos": 20232043463040.0, + "grad_norm": 2.476648082517961, + "learning_rate": 3.9954486724816105e-06, + "loss": 0.9139, + "num_input_tokens_seen": 18185020, + "step": 564 + }, + { + "epoch": 0.05095369076069802, + "flos": 21111506081760.0, + "grad_norm": 2.157704070375769, + "learning_rate": 3.995409196870161e-06, + "loss": 0.8596, + "num_input_tokens_seen": 18212095, + "step": 565 + }, + { + "epoch": 0.05104387428416828, + "flos": 64237886753280.0, + "grad_norm": 0.7778164753672437, + "learning_rate": 3.995369550999398e-06, + "loss": 0.5782, + "num_input_tokens_seen": 18296240, + "step": 566 + }, + { + "epoch": 0.05113405780763854, + "flos": 16776978446400.0, + "grad_norm": 3.626067826416779, + "learning_rate": 3.995329734872702e-06, + "loss": 0.8439, + "num_input_tokens_seen": 18322715, + "step": 567 + }, + { + "epoch": 0.051224241331108804, + "flos": 20523611759520.0, + "grad_norm": 1.8584949750978583, + "learning_rate": 3.9952897484934706e-06, + "loss": 0.9338, + "num_input_tokens_seen": 18350315, + "step": 568 + }, + { + "epoch": 0.05131442485457907, + "flos": 18962606992320.0, + "grad_norm": 2.118743887473322, + "learning_rate": 3.995249591865115e-06, + "loss": 0.8233, + "num_input_tokens_seen": 18378870, + "step": 569 + }, + { + "epoch": 0.05140460837804933, + "flos": 26905436329440.0, + "grad_norm": 1.785632047121898, + "learning_rate": 3.995209264991063e-06, + "loss": 0.7795, + "num_input_tokens_seen": 18408970, + "step": 570 + }, + { + "epoch": 0.05149479190151959, + "flos": 21583785249600.0, + "grad_norm": 2.5506844144377534, + "learning_rate": 3.995168767874756e-06, + "loss": 0.8674, + "num_input_tokens_seen": 18437680, + "step": 571 + }, + { + "epoch": 0.051584975424989854, + "flos": 43518547539360.0, + "grad_norm": 1.76689622065428, + "learning_rate": 3.995128100519648e-06, + "loss": 0.8025, + "num_input_tokens_seen": 18472300, + "step": 572 + }, + { + "epoch": 0.051675158948460115, + "flos": 25258372100640.0, + "grad_norm": 2.2002291213281246, + "learning_rate": 3.995087262929209e-06, + "loss": 0.8339, + "num_input_tokens_seen": 18500300, + "step": 573 + }, + { + "epoch": 0.051765342471930376, + "flos": 22825747534080.0, + "grad_norm": 2.426529183929723, + "learning_rate": 3.995046255106925e-06, + "loss": 0.8285, + "num_input_tokens_seen": 18530095, + "step": 574 + }, + { + "epoch": 0.05185552599540064, + "flos": 20344870885440.0, + "grad_norm": 2.1782045412796114, + "learning_rate": 3.995005077056293e-06, + "loss": 0.8844, + "num_input_tokens_seen": 18558290, + "step": 575 + }, + { + "epoch": 0.051945709518870904, + "flos": 18416497213920.0, + "grad_norm": 2.141133825710301, + "learning_rate": 3.9949637287808284e-06, + "loss": 0.7096, + "num_input_tokens_seen": 18586870, + "step": 576 + }, + { + "epoch": 0.052035893042341165, + "flos": 25265545864320.0, + "grad_norm": 1.9189504686436771, + "learning_rate": 3.994922210284057e-06, + "loss": 0.7695, + "num_input_tokens_seen": 18616450, + "step": 577 + }, + { + "epoch": 0.052126076565811426, + "flos": 17609675773440.0, + "grad_norm": 2.251472521115055, + "learning_rate": 3.994880521569524e-06, + "loss": 0.8527, + "num_input_tokens_seen": 18642510, + "step": 578 + }, + { + "epoch": 0.05221626008928169, + "flos": 25229411124000.0, + "grad_norm": 21.78760933774738, + "learning_rate": 3.994838662640785e-06, + "loss": 0.8298, + "num_input_tokens_seen": 18672235, + "step": 579 + }, + { + "epoch": 0.05230644361275195, + "flos": 17323534261920.0, + "grad_norm": 2.7353508652043024, + "learning_rate": 3.9947966335014116e-06, + "loss": 0.8564, + "num_input_tokens_seen": 18698720, + "step": 580 + }, + { + "epoch": 0.052396627136222215, + "flos": 13970740455840.0, + "grad_norm": 2.6888628320858894, + "learning_rate": 3.99475443415499e-06, + "loss": 0.7958, + "num_input_tokens_seen": 18724255, + "step": 581 + }, + { + "epoch": 0.052486810659692476, + "flos": 21002507144640.0, + "grad_norm": 2.1700406458339745, + "learning_rate": 3.994712064605121e-06, + "loss": 0.814, + "num_input_tokens_seen": 18752580, + "step": 582 + }, + { + "epoch": 0.05257699418316274, + "flos": 19472433351360.0, + "grad_norm": 2.134501029855422, + "learning_rate": 3.99466952485542e-06, + "loss": 0.9693, + "num_input_tokens_seen": 18778215, + "step": 583 + }, + { + "epoch": 0.052667177706633, + "flos": 23589223300800.0, + "grad_norm": 2.2352164224727207, + "learning_rate": 3.994626814909518e-06, + "loss": 0.8328, + "num_input_tokens_seen": 18806600, + "step": 584 + }, + { + "epoch": 0.05275736123010326, + "flos": 20201521356480.0, + "grad_norm": 2.8589188675738297, + "learning_rate": 3.994583934771056e-06, + "loss": 0.7992, + "num_input_tokens_seen": 18832980, + "step": 585 + }, + { + "epoch": 0.05284754475357352, + "flos": 23729562079200.0, + "grad_norm": 2.0170168543703864, + "learning_rate": 3.9945408844436955e-06, + "loss": 0.8793, + "num_input_tokens_seen": 18862840, + "step": 586 + }, + { + "epoch": 0.05293772827704379, + "flos": 24464076869280.0, + "grad_norm": 2.256141862893215, + "learning_rate": 3.994497663931109e-06, + "loss": 0.8453, + "num_input_tokens_seen": 18892360, + "step": 587 + }, + { + "epoch": 0.05302791180051405, + "flos": 22452691656480.0, + "grad_norm": 1.9657459045972956, + "learning_rate": 3.994454273236984e-06, + "loss": 0.8394, + "num_input_tokens_seen": 18919000, + "step": 588 + }, + { + "epoch": 0.05311809532398431, + "flos": 23443792265280.0, + "grad_norm": 1.674629438034046, + "learning_rate": 3.994410712365023e-06, + "loss": 0.8244, + "num_input_tokens_seen": 18948680, + "step": 589 + }, + { + "epoch": 0.05320827884745457, + "flos": 19508530921920.0, + "grad_norm": 2.392727394897192, + "learning_rate": 3.994366981318943e-06, + "loss": 0.9249, + "num_input_tokens_seen": 18974560, + "step": 590 + }, + { + "epoch": 0.05329846237092483, + "flos": 21796467848160.0, + "grad_norm": 2.3341078721576705, + "learning_rate": 3.9943230801024765e-06, + "loss": 0.8299, + "num_input_tokens_seen": 19002050, + "step": 591 + }, + { + "epoch": 0.05338864589439509, + "flos": 19756270337760.0, + "grad_norm": 2.1716792615328635, + "learning_rate": 3.9942790087193666e-06, + "loss": 0.8652, + "num_input_tokens_seen": 19025650, + "step": 592 + }, + { + "epoch": 0.05347882941786536, + "flos": 18124371371040.0, + "grad_norm": 2.0398845550184292, + "learning_rate": 3.994234767173376e-06, + "loss": 0.8125, + "num_input_tokens_seen": 19053555, + "step": 593 + }, + { + "epoch": 0.05356901294133562, + "flos": 19545074529600.0, + "grad_norm": 2.021438376967609, + "learning_rate": 3.994190355468279e-06, + "loss": 0.7105, + "num_input_tokens_seen": 19082490, + "step": 594 + }, + { + "epoch": 0.05365919646480588, + "flos": 33425295152640.0, + "grad_norm": 1.8277217061499915, + "learning_rate": 3.994145773607865e-06, + "loss": 0.821, + "num_input_tokens_seen": 19114300, + "step": 595 + }, + { + "epoch": 0.05374937998827614, + "flos": 21583041854400.0, + "grad_norm": 1.7956079410920394, + "learning_rate": 3.994101021595938e-06, + "loss": 0.7914, + "num_input_tokens_seen": 19143660, + "step": 596 + }, + { + "epoch": 0.0538395635117464, + "flos": 20529930618720.0, + "grad_norm": 2.368043851652585, + "learning_rate": 3.9940560994363165e-06, + "loss": 0.8384, + "num_input_tokens_seen": 19171945, + "step": 597 + }, + { + "epoch": 0.05392974703521666, + "flos": 23407322997120.0, + "grad_norm": 2.2885053041116206, + "learning_rate": 3.994011007132833e-06, + "loss": 0.6978, + "num_input_tokens_seen": 19200735, + "step": 598 + }, + { + "epoch": 0.05401993055868693, + "flos": 62850270414720.0, + "grad_norm": 0.7054222725294166, + "learning_rate": 3.993965744689337e-06, + "loss": 0.6482, + "num_input_tokens_seen": 19277815, + "step": 599 + }, + { + "epoch": 0.05411011408215719, + "flos": 39292424124960.0, + "grad_norm": 2.0145294997535275, + "learning_rate": 3.993920312109687e-06, + "loss": 0.8042, + "num_input_tokens_seen": 19307315, + "step": 600 + }, + { + "epoch": 0.05420029760562745, + "flos": 25917569489760.0, + "grad_norm": 1.9076729763290377, + "learning_rate": 3.993874709397764e-06, + "loss": 0.8165, + "num_input_tokens_seen": 19334050, + "step": 601 + }, + { + "epoch": 0.05429048112909771, + "flos": 23079396941760.0, + "grad_norm": 2.1596937460485437, + "learning_rate": 3.993828936557454e-06, + "loss": 0.8063, + "num_input_tokens_seen": 19362170, + "step": 602 + }, + { + "epoch": 0.054380664652567974, + "flos": 36886753368000.0, + "grad_norm": 2.135059556331399, + "learning_rate": 3.993782993592667e-06, + "loss": 0.7827, + "num_input_tokens_seen": 19393805, + "step": 603 + }, + { + "epoch": 0.054470848176038235, + "flos": 25119780300960.0, + "grad_norm": 1.989318996531614, + "learning_rate": 3.993736880507321e-06, + "loss": 0.8616, + "num_input_tokens_seen": 19423325, + "step": 604 + }, + { + "epoch": 0.0545610316995085, + "flos": 33712997794080.0, + "grad_norm": 2.4352898970722117, + "learning_rate": 3.99369059730535e-06, + "loss": 0.766, + "num_input_tokens_seen": 19454750, + "step": 605 + }, + { + "epoch": 0.05465121522297876, + "flos": 24680810971680.0, + "grad_norm": 1.509603119033559, + "learning_rate": 3.993644143990706e-06, + "loss": 0.8243, + "num_input_tokens_seen": 19486300, + "step": 606 + }, + { + "epoch": 0.054741398746449024, + "flos": 17429262260160.0, + "grad_norm": 1.9683374245084557, + "learning_rate": 3.99359752056735e-06, + "loss": 0.742, + "num_input_tokens_seen": 19514880, + "step": 607 + }, + { + "epoch": 0.054831582269919285, + "flos": 25775929769760.0, + "grad_norm": 2.0567793840076187, + "learning_rate": 3.993550727039261e-06, + "loss": 0.6757, + "num_input_tokens_seen": 19542830, + "step": 608 + }, + { + "epoch": 0.054921765793389546, + "flos": 22496371858080.0, + "grad_norm": 2.2501044778938573, + "learning_rate": 3.993503763410431e-06, + "loss": 0.8405, + "num_input_tokens_seen": 19572060, + "step": 609 + }, + { + "epoch": 0.05501194931685981, + "flos": 29236756099200.0, + "grad_norm": 1.9398283965125007, + "learning_rate": 3.9934566296848686e-06, + "loss": 0.7378, + "num_input_tokens_seen": 19600805, + "step": 610 + }, + { + "epoch": 0.055102132840330074, + "flos": 21216602194080.0, + "grad_norm": 5.100078559719215, + "learning_rate": 3.993409325866595e-06, + "loss": 0.8989, + "num_input_tokens_seen": 19625255, + "step": 611 + }, + { + "epoch": 0.055192316363800335, + "flos": 60503110593600.0, + "grad_norm": 0.6288683388223979, + "learning_rate": 3.993361851959645e-06, + "loss": 0.5966, + "num_input_tokens_seen": 19727255, + "step": 612 + }, + { + "epoch": 0.055282499887270596, + "flos": 24427310243040.0, + "grad_norm": 2.6400764392132503, + "learning_rate": 3.993314207968071e-06, + "loss": 0.8654, + "num_input_tokens_seen": 19759040, + "step": 613 + }, + { + "epoch": 0.05537268341074086, + "flos": 31714027281120.0, + "grad_norm": 1.8575803285996322, + "learning_rate": 3.993266393895938e-06, + "loss": 0.8079, + "num_input_tokens_seen": 19789525, + "step": 614 + }, + { + "epoch": 0.05546286693421112, + "flos": 17503464568320.0, + "grad_norm": 1.8914618413445652, + "learning_rate": 3.993218409747326e-06, + "loss": 0.826, + "num_input_tokens_seen": 19816050, + "step": 615 + }, + { + "epoch": 0.05555305045768138, + "flos": 26831382700320.0, + "grad_norm": 1.9805532584133205, + "learning_rate": 3.993170255526328e-06, + "loss": 0.8125, + "num_input_tokens_seen": 19845155, + "step": 616 + }, + { + "epoch": 0.055643233981151646, + "flos": 22566076625280.0, + "grad_norm": 3.316244779684054, + "learning_rate": 3.993121931237054e-06, + "loss": 0.8473, + "num_input_tokens_seen": 19873765, + "step": 617 + }, + { + "epoch": 0.05573341750462191, + "flos": 21841300312320.0, + "grad_norm": 1.8144068717193997, + "learning_rate": 3.993073436883627e-06, + "loss": 0.8383, + "num_input_tokens_seen": 19902620, + "step": 618 + }, + { + "epoch": 0.05582360102809217, + "flos": 23549334414720.0, + "grad_norm": 2.1564867193784694, + "learning_rate": 3.993024772470184e-06, + "loss": 0.7995, + "num_input_tokens_seen": 19931765, + "step": 619 + }, + { + "epoch": 0.05591378455156243, + "flos": 19181013733920.0, + "grad_norm": 2.0946059114438143, + "learning_rate": 3.992975938000878e-06, + "loss": 0.804, + "num_input_tokens_seen": 19959440, + "step": 620 + }, + { + "epoch": 0.05600396807503269, + "flos": 66215367411360.0, + "grad_norm": 0.6545705194593023, + "learning_rate": 3.992926933479876e-06, + "loss": 0.586, + "num_input_tokens_seen": 20054830, + "step": 621 + }, + { + "epoch": 0.05609415159850295, + "flos": 25666670644320.0, + "grad_norm": 1.9176649797033092, + "learning_rate": 3.9928777589113595e-06, + "loss": 0.8456, + "num_input_tokens_seen": 20082440, + "step": 622 + }, + { + "epoch": 0.05618433512197322, + "flos": 35830073835360.0, + "grad_norm": 2.4262289963942676, + "learning_rate": 3.992828414299524e-06, + "loss": 0.7815, + "num_input_tokens_seen": 20116225, + "step": 623 + }, + { + "epoch": 0.05627451864544348, + "flos": 23043150692160.0, + "grad_norm": 1.8564684293011016, + "learning_rate": 3.992778899648579e-06, + "loss": 0.8188, + "num_input_tokens_seen": 20147370, + "step": 624 + }, + { + "epoch": 0.05636470216891374, + "flos": 20456397366240.0, + "grad_norm": 2.0449939266944743, + "learning_rate": 3.992729214962751e-06, + "loss": 0.8388, + "num_input_tokens_seen": 20174535, + "step": 625 + }, + { + "epoch": 0.056454885692384, + "flos": 25812399037920.0, + "grad_norm": 1.9622505928512028, + "learning_rate": 3.992679360246279e-06, + "loss": 0.857, + "num_input_tokens_seen": 20201480, + "step": 626 + }, + { + "epoch": 0.05654506921585426, + "flos": 23553311579040.0, + "grad_norm": 4.515712644789039, + "learning_rate": 3.992629335503416e-06, + "loss": 0.8506, + "num_input_tokens_seen": 20232140, + "step": 627 + }, + { + "epoch": 0.05663525273932452, + "flos": 36666190780320.0, + "grad_norm": 1.9408208014199784, + "learning_rate": 3.9925791407384304e-06, + "loss": 0.8021, + "num_input_tokens_seen": 20261360, + "step": 628 + }, + { + "epoch": 0.05672543626279479, + "flos": 20711496394560.0, + "grad_norm": 4.641648304946897, + "learning_rate": 3.992528775955606e-06, + "loss": 0.8601, + "num_input_tokens_seen": 20287825, + "step": 629 + }, + { + "epoch": 0.05681561978626505, + "flos": 19727866907520.0, + "grad_norm": 2.618656850537447, + "learning_rate": 3.992478241159239e-06, + "loss": 0.7891, + "num_input_tokens_seen": 20317400, + "step": 630 + }, + { + "epoch": 0.05690580330973531, + "flos": 20893768395840.0, + "grad_norm": 2.214761937697555, + "learning_rate": 3.992427536353643e-06, + "loss": 0.8415, + "num_input_tokens_seen": 20346965, + "step": 631 + }, + { + "epoch": 0.05699598683320557, + "flos": 25593286070880.0, + "grad_norm": 2.796421035494757, + "learning_rate": 3.992376661543143e-06, + "loss": 0.8058, + "num_input_tokens_seen": 20373590, + "step": 632 + }, + { + "epoch": 0.05708617035667583, + "flos": 17686926001920.0, + "grad_norm": 2.2952751764284307, + "learning_rate": 3.992325616732081e-06, + "loss": 0.793, + "num_input_tokens_seen": 20398630, + "step": 633 + }, + { + "epoch": 0.057176353880146094, + "flos": 25516444709760.0, + "grad_norm": 3.4166825146811535, + "learning_rate": 3.992274401924811e-06, + "loss": 0.9222, + "num_input_tokens_seen": 20426720, + "step": 634 + }, + { + "epoch": 0.05726653740361636, + "flos": 21403185887520.0, + "grad_norm": 1.772917616679934, + "learning_rate": 3.992223017125704e-06, + "loss": 0.7406, + "num_input_tokens_seen": 20454095, + "step": 635 + }, + { + "epoch": 0.05735672092708662, + "flos": 19108112367360.0, + "grad_norm": 1.9534549892132662, + "learning_rate": 3.992171462339145e-06, + "loss": 0.848, + "num_input_tokens_seen": 20480590, + "step": 636 + }, + { + "epoch": 0.057446904450556883, + "flos": 21216007477920.0, + "grad_norm": 2.1201956799006396, + "learning_rate": 3.992119737569532e-06, + "loss": 0.8866, + "num_input_tokens_seen": 20508885, + "step": 637 + }, + { + "epoch": 0.057537087974027144, + "flos": 21258312398400.0, + "grad_norm": 2.1916737656665704, + "learning_rate": 3.992067842821277e-06, + "loss": 0.8164, + "num_input_tokens_seen": 20536730, + "step": 638 + }, + { + "epoch": 0.057627271497497405, + "flos": 19909544192640.0, + "grad_norm": 3.3438178795315108, + "learning_rate": 3.99201577809881e-06, + "loss": 0.9493, + "num_input_tokens_seen": 20562815, + "step": 639 + }, + { + "epoch": 0.057717455020967666, + "flos": 18890002983840.0, + "grad_norm": 2.450964200770434, + "learning_rate": 3.991963543406574e-06, + "loss": 0.8386, + "num_input_tokens_seen": 20589290, + "step": 640 + }, + { + "epoch": 0.057807638544437934, + "flos": 21983534748480.0, + "grad_norm": 2.1097471685931857, + "learning_rate": 3.991911138749024e-06, + "loss": 0.8156, + "num_input_tokens_seen": 20618200, + "step": 641 + }, + { + "epoch": 0.057897822067908195, + "flos": 19179378264480.0, + "grad_norm": 2.265219185099528, + "learning_rate": 3.991858564130633e-06, + "loss": 0.8431, + "num_input_tokens_seen": 20646845, + "step": 642 + }, + { + "epoch": 0.057988005591378455, + "flos": 17170706444160.0, + "grad_norm": 2.4122073257882852, + "learning_rate": 3.991805819555885e-06, + "loss": 0.7976, + "num_input_tokens_seen": 20670870, + "step": 643 + }, + { + "epoch": 0.058078189114848716, + "flos": 21364635112800.0, + "grad_norm": 1.9918386714939207, + "learning_rate": 3.991752905029283e-06, + "loss": 0.8216, + "num_input_tokens_seen": 20700685, + "step": 644 + }, + { + "epoch": 0.05816837263831898, + "flos": 24099235508640.0, + "grad_norm": 2.0900196607477195, + "learning_rate": 3.991699820555341e-06, + "loss": 0.6953, + "num_input_tokens_seen": 20732560, + "step": 645 + }, + { + "epoch": 0.05825855616178924, + "flos": 16340127793440.0, + "grad_norm": 2.1916063186788355, + "learning_rate": 3.991646566138588e-06, + "loss": 0.8214, + "num_input_tokens_seen": 20760085, + "step": 646 + }, + { + "epoch": 0.058348739685259506, + "flos": 23224493449440.0, + "grad_norm": 1.7934400151061127, + "learning_rate": 3.991593141783567e-06, + "loss": 0.7415, + "num_input_tokens_seen": 20790995, + "step": 647 + }, + { + "epoch": 0.058438923208729766, + "flos": 19254398307360.0, + "grad_norm": 1.8206369659313464, + "learning_rate": 3.991539547494839e-06, + "loss": 0.8442, + "num_input_tokens_seen": 20818825, + "step": 648 + }, + { + "epoch": 0.05852910673220003, + "flos": 23116052058720.0, + "grad_norm": 2.130872465909891, + "learning_rate": 3.991485783276974e-06, + "loss": 0.8147, + "num_input_tokens_seen": 20847585, + "step": 649 + }, + { + "epoch": 0.05861929025567029, + "flos": 31130184462720.0, + "grad_norm": 2.9598331858050666, + "learning_rate": 3.991431849134563e-06, + "loss": 0.8731, + "num_input_tokens_seen": 20879715, + "step": 650 + }, + { + "epoch": 0.05870947377914055, + "flos": 28289298522240.0, + "grad_norm": 2.72403072652418, + "learning_rate": 3.991377745072205e-06, + "loss": 0.713, + "num_input_tokens_seen": 20909100, + "step": 651 + }, + { + "epoch": 0.05879965730261081, + "flos": 19582064174400.0, + "grad_norm": 2.418453432970395, + "learning_rate": 3.991323471094517e-06, + "loss": 0.8471, + "num_input_tokens_seen": 20937330, + "step": 652 + }, + { + "epoch": 0.05888984082608108, + "flos": 29888705385120.0, + "grad_norm": 1.9041441870878206, + "learning_rate": 3.991269027206131e-06, + "loss": 0.7918, + "num_input_tokens_seen": 20965385, + "step": 653 + }, + { + "epoch": 0.05898002434955134, + "flos": 63859844394240.0, + "grad_norm": 0.7935056700837415, + "learning_rate": 3.9912144134116916e-06, + "loss": 0.6377, + "num_input_tokens_seen": 21062515, + "step": 654 + }, + { + "epoch": 0.0590702078730216, + "flos": 23079842978880.0, + "grad_norm": 2.2320897079679622, + "learning_rate": 3.99115962971586e-06, + "loss": 0.7426, + "num_input_tokens_seen": 21090305, + "step": 655 + }, + { + "epoch": 0.05916039139649186, + "flos": 25805076595200.0, + "grad_norm": 2.2703809608381644, + "learning_rate": 3.991104676123308e-06, + "loss": 0.8526, + "num_input_tokens_seen": 21115670, + "step": 656 + }, + { + "epoch": 0.05925057491996212, + "flos": 70386951320640.0, + "grad_norm": 0.6586139200726209, + "learning_rate": 3.991049552638727e-06, + "loss": 0.5725, + "num_input_tokens_seen": 21212170, + "step": 657 + }, + { + "epoch": 0.05934075844343238, + "flos": 17432124331680.0, + "grad_norm": 2.1939318241943266, + "learning_rate": 3.99099425926682e-06, + "loss": 0.7815, + "num_input_tokens_seen": 21237675, + "step": 658 + }, + { + "epoch": 0.05943094196690265, + "flos": 26722941309600.0, + "grad_norm": 2.2795238496269494, + "learning_rate": 3.990938796012304e-06, + "loss": 0.787, + "num_input_tokens_seen": 21266395, + "step": 659 + }, + { + "epoch": 0.05952112549037291, + "flos": 14403279416640.0, + "grad_norm": 2.0054509267992087, + "learning_rate": 3.990883162879912e-06, + "loss": 0.9717, + "num_input_tokens_seen": 21290875, + "step": 660 + }, + { + "epoch": 0.05961130901384317, + "flos": 25120263507840.0, + "grad_norm": 2.6634931655145877, + "learning_rate": 3.990827359874391e-06, + "loss": 0.8836, + "num_input_tokens_seen": 21318575, + "step": 661 + }, + { + "epoch": 0.05970149253731343, + "flos": 23803838726880.0, + "grad_norm": 1.51025674784529, + "learning_rate": 3.990771387000503e-06, + "loss": 0.8201, + "num_input_tokens_seen": 21349215, + "step": 662 + }, + { + "epoch": 0.05979167606078369, + "flos": 28289038333920.0, + "grad_norm": 2.9662018129571748, + "learning_rate": 3.990715244263023e-06, + "loss": 0.849, + "num_input_tokens_seen": 21380170, + "step": 663 + }, + { + "epoch": 0.05988185958425396, + "flos": 71870668689600.0, + "grad_norm": 0.7100865647776274, + "learning_rate": 3.990658931666741e-06, + "loss": 0.6051, + "num_input_tokens_seen": 21473305, + "step": 664 + }, + { + "epoch": 0.05997204310772422, + "flos": 23880977446080.0, + "grad_norm": 1.8618913820998777, + "learning_rate": 3.990602449216463e-06, + "loss": 0.7371, + "num_input_tokens_seen": 21501045, + "step": 665 + }, + { + "epoch": 0.06006222663119448, + "flos": 29813350814400.0, + "grad_norm": 2.2833912872236435, + "learning_rate": 3.990545796917008e-06, + "loss": 0.7226, + "num_input_tokens_seen": 21528980, + "step": 666 + }, + { + "epoch": 0.06015241015466474, + "flos": 19138299946080.0, + "grad_norm": 1.8006149513099294, + "learning_rate": 3.99048897477321e-06, + "loss": 0.8635, + "num_input_tokens_seen": 21556390, + "step": 667 + }, + { + "epoch": 0.060242593678135004, + "flos": 20925145406880.0, + "grad_norm": 2.081459296072949, + "learning_rate": 3.990431982789917e-06, + "loss": 0.8378, + "num_input_tokens_seen": 21583160, + "step": 668 + }, + { + "epoch": 0.060332777201605264, + "flos": 24427235903520.0, + "grad_norm": 1.8677862635826998, + "learning_rate": 3.990374820971992e-06, + "loss": 0.753, + "num_input_tokens_seen": 21610510, + "step": 669 + }, + { + "epoch": 0.06042296072507553, + "flos": 49529620756800.0, + "grad_norm": 2.280014957914309, + "learning_rate": 3.990317489324312e-06, + "loss": 0.704, + "num_input_tokens_seen": 21643140, + "step": 670 + }, + { + "epoch": 0.06051314424854579, + "flos": 45266693546400.0, + "grad_norm": 2.530246017466489, + "learning_rate": 3.99025998785177e-06, + "loss": 0.8007, + "num_input_tokens_seen": 21674615, + "step": 671 + }, + { + "epoch": 0.060603327772016054, + "flos": 19910436266880.0, + "grad_norm": 1.7677985122448259, + "learning_rate": 3.990202316559271e-06, + "loss": 0.7587, + "num_input_tokens_seen": 21701840, + "step": 672 + }, + { + "epoch": 0.060693511295486315, + "flos": 25520496213600.0, + "grad_norm": 2.2189996068540676, + "learning_rate": 3.990144475451738e-06, + "loss": 0.8667, + "num_input_tokens_seen": 21730880, + "step": 673 + }, + { + "epoch": 0.060783694818956575, + "flos": 21472816315200.0, + "grad_norm": 1.663289925054071, + "learning_rate": 3.9900864645341036e-06, + "loss": 0.8169, + "num_input_tokens_seen": 21759050, + "step": 674 + }, + { + "epoch": 0.060873878342426836, + "flos": 22205472617280.0, + "grad_norm": 1.7287341653204849, + "learning_rate": 3.990028283811319e-06, + "loss": 0.8233, + "num_input_tokens_seen": 21786480, + "step": 675 + }, + { + "epoch": 0.060964061865897104, + "flos": 26975996001120.0, + "grad_norm": 1.7104433643830015, + "learning_rate": 3.989969933288348e-06, + "loss": 0.8129, + "num_input_tokens_seen": 21817470, + "step": 676 + }, + { + "epoch": 0.061054245389367365, + "flos": 61331756416800.0, + "grad_norm": 0.6583589318485473, + "learning_rate": 3.98991141297017e-06, + "loss": 0.5502, + "num_input_tokens_seen": 21906585, + "step": 677 + }, + { + "epoch": 0.061144428912837626, + "flos": 23917186525920.0, + "grad_norm": 2.396016343355561, + "learning_rate": 3.989852722861778e-06, + "loss": 0.7651, + "num_input_tokens_seen": 21936300, + "step": 678 + }, + { + "epoch": 0.061234612436307886, + "flos": 49768136338560.0, + "grad_norm": 0.7164350214747846, + "learning_rate": 3.98979386296818e-06, + "loss": 0.578, + "num_input_tokens_seen": 22018330, + "step": 679 + }, + { + "epoch": 0.06132479595977815, + "flos": 40561600407360.0, + "grad_norm": 1.8661101996252487, + "learning_rate": 3.989734833294398e-06, + "loss": 0.7964, + "num_input_tokens_seen": 22050905, + "step": 680 + }, + { + "epoch": 0.06141497948324841, + "flos": 20018840487840.0, + "grad_norm": 1.8126150292036223, + "learning_rate": 3.989675633845469e-06, + "loss": 0.8832, + "num_input_tokens_seen": 22077845, + "step": 681 + }, + { + "epoch": 0.061505163006718676, + "flos": 16704188589120.0, + "grad_norm": 4.136264332999801, + "learning_rate": 3.989616264626443e-06, + "loss": 0.7959, + "num_input_tokens_seen": 22105475, + "step": 682 + }, + { + "epoch": 0.06159534653018894, + "flos": 24935686981440.0, + "grad_norm": 1.9920464559254907, + "learning_rate": 3.989556725642388e-06, + "loss": 0.8522, + "num_input_tokens_seen": 22137655, + "step": 683 + }, + { + "epoch": 0.0616855300536592, + "flos": 25185024696960.0, + "grad_norm": 2.672108868631039, + "learning_rate": 3.989497016898382e-06, + "loss": 0.8697, + "num_input_tokens_seen": 22167180, + "step": 684 + }, + { + "epoch": 0.06177571357712946, + "flos": 27374964935040.0, + "grad_norm": 1.7349025164654595, + "learning_rate": 3.98943713839952e-06, + "loss": 0.8118, + "num_input_tokens_seen": 22198775, + "step": 685 + }, + { + "epoch": 0.06186589710059972, + "flos": 71348613479520.0, + "grad_norm": 0.6731368347374361, + "learning_rate": 3.9893770901509125e-06, + "loss": 0.6063, + "num_input_tokens_seen": 22300785, + "step": 686 + }, + { + "epoch": 0.06195608062406998, + "flos": 26654165786400.0, + "grad_norm": 1.8004992878707753, + "learning_rate": 3.989316872157682e-06, + "loss": 0.8085, + "num_input_tokens_seen": 22333970, + "step": 687 + }, + { + "epoch": 0.06204626414754025, + "flos": 15865841458560.0, + "grad_norm": 1.8845608140419638, + "learning_rate": 3.989256484424968e-06, + "loss": 0.8934, + "num_input_tokens_seen": 22359480, + "step": 688 + }, + { + "epoch": 0.06213644767101051, + "flos": 23115345833280.0, + "grad_norm": 2.7315559023856104, + "learning_rate": 3.98919592695792e-06, + "loss": 0.7821, + "num_input_tokens_seen": 22387800, + "step": 689 + }, + { + "epoch": 0.06222663119448077, + "flos": 24464076869280.0, + "grad_norm": 1.9309336233401089, + "learning_rate": 3.9891351997617096e-06, + "loss": 0.8452, + "num_input_tokens_seen": 22415570, + "step": 690 + }, + { + "epoch": 0.06231681471795103, + "flos": 30908395272960.0, + "grad_norm": 2.4301080370461534, + "learning_rate": 3.989074302841514e-06, + "loss": 0.757, + "num_input_tokens_seen": 22447750, + "step": 691 + }, + { + "epoch": 0.06240699824142129, + "flos": 17319408418560.0, + "grad_norm": 2.1903826657068404, + "learning_rate": 3.989013236202533e-06, + "loss": 0.9272, + "num_input_tokens_seen": 22473460, + "step": 692 + }, + { + "epoch": 0.06249718176489155, + "flos": 24317456401440.0, + "grad_norm": 1.5309549942776843, + "learning_rate": 3.988951999849974e-06, + "loss": 0.8541, + "num_input_tokens_seen": 22502300, + "step": 693 + }, + { + "epoch": 0.06258736528836181, + "flos": 26905064631840.0, + "grad_norm": 2.2408583685782872, + "learning_rate": 3.988890593789064e-06, + "loss": 0.7328, + "num_input_tokens_seen": 22529620, + "step": 694 + }, + { + "epoch": 0.06267754881183207, + "flos": 24390692295840.0, + "grad_norm": 2.1510548687422744, + "learning_rate": 3.9888290180250415e-06, + "loss": 0.8059, + "num_input_tokens_seen": 22561640, + "step": 695 + }, + { + "epoch": 0.06276773233530233, + "flos": 30293732989920.0, + "grad_norm": 3.8411819439068866, + "learning_rate": 3.988767272563161e-06, + "loss": 0.7833, + "num_input_tokens_seen": 22591225, + "step": 696 + }, + { + "epoch": 0.06285791585877261, + "flos": 17650010696640.0, + "grad_norm": 2.0516952948144302, + "learning_rate": 3.988705357408691e-06, + "loss": 0.7592, + "num_input_tokens_seen": 22614965, + "step": 697 + }, + { + "epoch": 0.06294809938224287, + "flos": 21876542978400.0, + "grad_norm": 2.5976850694477456, + "learning_rate": 3.9886432725669146e-06, + "loss": 0.7787, + "num_input_tokens_seen": 22644270, + "step": 698 + }, + { + "epoch": 0.06303828290571313, + "flos": 21368203409760.0, + "grad_norm": 2.0659437617488488, + "learning_rate": 3.988581018043128e-06, + "loss": 0.7753, + "num_input_tokens_seen": 22670560, + "step": 699 + }, + { + "epoch": 0.06312846642918339, + "flos": 27196075381920.0, + "grad_norm": 2.4720287521134607, + "learning_rate": 3.988518593842645e-06, + "loss": 0.8243, + "num_input_tokens_seen": 22699805, + "step": 700 + }, + { + "epoch": 0.06321864995265365, + "flos": 19909692871680.0, + "grad_norm": 1.8545457331041817, + "learning_rate": 3.9884559999707906e-06, + "loss": 0.8129, + "num_input_tokens_seen": 22725890, + "step": 701 + }, + { + "epoch": 0.06330883347612391, + "flos": 22889430800160.0, + "grad_norm": 4.299205688922555, + "learning_rate": 3.988393236432906e-06, + "loss": 0.8291, + "num_input_tokens_seen": 22755285, + "step": 702 + }, + { + "epoch": 0.06339901699959417, + "flos": 23406876960000.0, + "grad_norm": 2.0030999002059477, + "learning_rate": 3.988330303234347e-06, + "loss": 0.8911, + "num_input_tokens_seen": 22782965, + "step": 703 + }, + { + "epoch": 0.06348920052306443, + "flos": 39728865910560.0, + "grad_norm": 1.7094089874291964, + "learning_rate": 3.988267200380483e-06, + "loss": 0.7687, + "num_input_tokens_seen": 22818335, + "step": 704 + }, + { + "epoch": 0.0635793840465347, + "flos": 24354817743840.0, + "grad_norm": 1.8579525034185094, + "learning_rate": 3.988203927876698e-06, + "loss": 0.8647, + "num_input_tokens_seen": 22848610, + "step": 705 + }, + { + "epoch": 0.06366956757000496, + "flos": 28653768185280.0, + "grad_norm": 1.9472489121924157, + "learning_rate": 3.988140485728391e-06, + "loss": 0.7063, + "num_input_tokens_seen": 22877915, + "step": 706 + }, + { + "epoch": 0.06375975109347522, + "flos": 18598397517600.0, + "grad_norm": 1.8632216214089616, + "learning_rate": 3.988076873940975e-06, + "loss": 0.7662, + "num_input_tokens_seen": 22904915, + "step": 707 + }, + { + "epoch": 0.06384993461694548, + "flos": 31604247779040.0, + "grad_norm": 2.155208059476525, + "learning_rate": 3.9880130925198786e-06, + "loss": 0.8121, + "num_input_tokens_seen": 22934890, + "step": 708 + }, + { + "epoch": 0.06394011814041575, + "flos": 20128471310880.0, + "grad_norm": 2.9232383444072587, + "learning_rate": 3.987949141470543e-06, + "loss": 0.8267, + "num_input_tokens_seen": 22961320, + "step": 709 + }, + { + "epoch": 0.06403030166388601, + "flos": 25627190625600.0, + "grad_norm": 1.8715805025838397, + "learning_rate": 3.987885020798425e-06, + "loss": 0.8166, + "num_input_tokens_seen": 22990670, + "step": 710 + }, + { + "epoch": 0.06412048518735627, + "flos": 27225296546880.0, + "grad_norm": 2.6582338890520956, + "learning_rate": 3.987820730508996e-06, + "loss": 0.7847, + "num_input_tokens_seen": 23017890, + "step": 711 + }, + { + "epoch": 0.06421066871082654, + "flos": 21694865693280.0, + "grad_norm": 1.803478612224605, + "learning_rate": 3.987756270607742e-06, + "loss": 0.9102, + "num_input_tokens_seen": 23042690, + "step": 712 + }, + { + "epoch": 0.0643008522342968, + "flos": 71756205797760.0, + "grad_norm": 0.6637011274028455, + "learning_rate": 3.987691641100162e-06, + "loss": 0.5649, + "num_input_tokens_seen": 23137420, + "step": 713 + }, + { + "epoch": 0.06439103575776706, + "flos": 20893694056320.0, + "grad_norm": 1.8944575659197105, + "learning_rate": 3.987626841991771e-06, + "loss": 0.9015, + "num_input_tokens_seen": 23165910, + "step": 714 + }, + { + "epoch": 0.06448121928123732, + "flos": 20966558253120.0, + "grad_norm": 3.125926681449495, + "learning_rate": 3.987561873288099e-06, + "loss": 0.8498, + "num_input_tokens_seen": 23192440, + "step": 715 + }, + { + "epoch": 0.06457140280470758, + "flos": 31130742009120.0, + "grad_norm": 2.066183557751762, + "learning_rate": 3.987496734994688e-06, + "loss": 0.7638, + "num_input_tokens_seen": 23222060, + "step": 716 + }, + { + "epoch": 0.06466158632817784, + "flos": 24172545742560.0, + "grad_norm": 5.06942031134634, + "learning_rate": 3.987431427117097e-06, + "loss": 0.847, + "num_input_tokens_seen": 23246460, + "step": 717 + }, + { + "epoch": 0.0647517698516481, + "flos": 22277965116480.0, + "grad_norm": 2.1358196591601732, + "learning_rate": 3.9873659496608985e-06, + "loss": 0.8274, + "num_input_tokens_seen": 23273705, + "step": 718 + }, + { + "epoch": 0.06484195337511836, + "flos": 36850953155520.0, + "grad_norm": 3.2239932714160155, + "learning_rate": 3.987300302631678e-06, + "loss": 0.7707, + "num_input_tokens_seen": 23305395, + "step": 719 + }, + { + "epoch": 0.06493213689858862, + "flos": 20492346257760.0, + "grad_norm": 2.0384459751016712, + "learning_rate": 3.987234486035039e-06, + "loss": 0.8533, + "num_input_tokens_seen": 23334705, + "step": 720 + }, + { + "epoch": 0.0650223204220589, + "flos": 20814845528160.0, + "grad_norm": 2.4411649294796005, + "learning_rate": 3.987168499876595e-06, + "loss": 0.8664, + "num_input_tokens_seen": 23360195, + "step": 721 + }, + { + "epoch": 0.06511250394552916, + "flos": 15683680966560.0, + "grad_norm": 1.9994929982334875, + "learning_rate": 3.987102344161978e-06, + "loss": 0.8279, + "num_input_tokens_seen": 23387990, + "step": 722 + }, + { + "epoch": 0.06520268746899942, + "flos": 16047667422720.0, + "grad_norm": 2.186984301133536, + "learning_rate": 3.987036018896832e-06, + "loss": 0.8531, + "num_input_tokens_seen": 23412510, + "step": 723 + }, + { + "epoch": 0.06529287099246968, + "flos": 22314545893920.0, + "grad_norm": 1.8381529731269257, + "learning_rate": 3.986969524086817e-06, + "loss": 0.8776, + "num_input_tokens_seen": 23440810, + "step": 724 + }, + { + "epoch": 0.06538305451593994, + "flos": 22124208054720.0, + "grad_norm": 2.4339404402618285, + "learning_rate": 3.986902859737605e-06, + "loss": 0.7954, + "num_input_tokens_seen": 23467540, + "step": 725 + }, + { + "epoch": 0.0654732380394102, + "flos": 37652719508640.0, + "grad_norm": 2.3633975670303484, + "learning_rate": 3.986836025854886e-06, + "loss": 0.7502, + "num_input_tokens_seen": 23500970, + "step": 726 + }, + { + "epoch": 0.06556342156288046, + "flos": 26249361200160.0, + "grad_norm": 2.1753697442785467, + "learning_rate": 3.986769022444362e-06, + "loss": 0.7687, + "num_input_tokens_seen": 23530465, + "step": 727 + }, + { + "epoch": 0.06565360508635072, + "flos": 21068123238240.0, + "grad_norm": 7.700531869329328, + "learning_rate": 3.986701849511751e-06, + "loss": 0.8106, + "num_input_tokens_seen": 23558330, + "step": 728 + }, + { + "epoch": 0.06574378860982098, + "flos": 19109041611360.0, + "grad_norm": 2.3688642650411538, + "learning_rate": 3.986634507062782e-06, + "loss": 0.7952, + "num_input_tokens_seen": 23585035, + "step": 729 + }, + { + "epoch": 0.06583397213329124, + "flos": 19800136388160.0, + "grad_norm": 3.3981091570910227, + "learning_rate": 3.986566995103204e-06, + "loss": 0.8327, + "num_input_tokens_seen": 23612175, + "step": 730 + }, + { + "epoch": 0.0659241556567615, + "flos": 27338532836640.0, + "grad_norm": 2.5607654935037933, + "learning_rate": 3.986499313638776e-06, + "loss": 0.7473, + "num_input_tokens_seen": 23641580, + "step": 731 + }, + { + "epoch": 0.06601433918023177, + "flos": 24425860622400.0, + "grad_norm": 2.1522162481021674, + "learning_rate": 3.986431462675272e-06, + "loss": 0.7415, + "num_input_tokens_seen": 23671700, + "step": 732 + }, + { + "epoch": 0.06610452270370204, + "flos": 22857384733440.0, + "grad_norm": 1.526645432467539, + "learning_rate": 3.9863634422184835e-06, + "loss": 0.8332, + "num_input_tokens_seen": 23699905, + "step": 733 + }, + { + "epoch": 0.0661947062271723, + "flos": 19181385431520.0, + "grad_norm": 2.4845399621861493, + "learning_rate": 3.986295252274213e-06, + "loss": 0.8486, + "num_input_tokens_seen": 23727200, + "step": 734 + }, + { + "epoch": 0.06628488975064256, + "flos": 27814975017600.0, + "grad_norm": 1.7732619144045352, + "learning_rate": 3.9862268928482796e-06, + "loss": 0.8634, + "num_input_tokens_seen": 23759215, + "step": 735 + }, + { + "epoch": 0.06637507327411282, + "flos": 28725703138080.0, + "grad_norm": 3.7816884440813134, + "learning_rate": 3.986158363946515e-06, + "loss": 0.7258, + "num_input_tokens_seen": 23789460, + "step": 736 + }, + { + "epoch": 0.06646525679758308, + "flos": 22824520932000.0, + "grad_norm": 1.8279148209215208, + "learning_rate": 3.9860896655747685e-06, + "loss": 0.8157, + "num_input_tokens_seen": 23818195, + "step": 737 + }, + { + "epoch": 0.06655544032105334, + "flos": 27050384158080.0, + "grad_norm": 3.1672704188027194, + "learning_rate": 3.9860207977388994e-06, + "loss": 0.8907, + "num_input_tokens_seen": 23847240, + "step": 738 + }, + { + "epoch": 0.0666456238445236, + "flos": 22751879753760.0, + "grad_norm": 2.8202349792139976, + "learning_rate": 3.9859517604447854e-06, + "loss": 0.7433, + "num_input_tokens_seen": 23876450, + "step": 739 + }, + { + "epoch": 0.06673580736799387, + "flos": 16699988406240.0, + "grad_norm": 1.5282982382840764, + "learning_rate": 3.985882553698317e-06, + "loss": 0.7837, + "num_input_tokens_seen": 23903725, + "step": 740 + }, + { + "epoch": 0.06682599089146413, + "flos": 19144767484320.0, + "grad_norm": 2.3404922223596527, + "learning_rate": 3.985813177505399e-06, + "loss": 0.8755, + "num_input_tokens_seen": 23930865, + "step": 741 + }, + { + "epoch": 0.06691617441493439, + "flos": 29055115983840.0, + "grad_norm": 1.7099702095203002, + "learning_rate": 3.985743631871951e-06, + "loss": 0.7651, + "num_input_tokens_seen": 23959335, + "step": 742 + }, + { + "epoch": 0.06700635793840465, + "flos": 25884817197600.0, + "grad_norm": 2.264987513863545, + "learning_rate": 3.985673916803907e-06, + "loss": 0.812, + "num_input_tokens_seen": 23986685, + "step": 743 + }, + { + "epoch": 0.06709654146187491, + "flos": 22889839667520.0, + "grad_norm": 2.379785534948957, + "learning_rate": 3.985604032307215e-06, + "loss": 0.838, + "num_input_tokens_seen": 24008255, + "step": 744 + }, + { + "epoch": 0.06718672498534518, + "flos": 25265174166720.0, + "grad_norm": 2.6610166975814207, + "learning_rate": 3.985533978387839e-06, + "loss": 0.8218, + "num_input_tokens_seen": 24037790, + "step": 745 + }, + { + "epoch": 0.06727690850881544, + "flos": 15392818895520.0, + "grad_norm": 1.9198803979656327, + "learning_rate": 3.985463755051756e-06, + "loss": 0.855, + "num_input_tokens_seen": 24063690, + "step": 746 + }, + { + "epoch": 0.0673670920322857, + "flos": 26577547443840.0, + "grad_norm": 2.2860143370892145, + "learning_rate": 3.9853933623049575e-06, + "loss": 0.8122, + "num_input_tokens_seen": 24093415, + "step": 747 + }, + { + "epoch": 0.06745727555575597, + "flos": 26540371950240.0, + "grad_norm": 1.7968845533034088, + "learning_rate": 3.98532280015345e-06, + "loss": 0.8094, + "num_input_tokens_seen": 24122685, + "step": 748 + }, + { + "epoch": 0.06754745907922623, + "flos": 43007643257280.0, + "grad_norm": 2.5292532309518787, + "learning_rate": 3.985252068603254e-06, + "loss": 0.7349, + "num_input_tokens_seen": 24154835, + "step": 749 + }, + { + "epoch": 0.06763764260269649, + "flos": 19177482606720.0, + "grad_norm": 11.204046363132678, + "learning_rate": 3.985181167660406e-06, + "loss": 0.8452, + "num_input_tokens_seen": 24182745, + "step": 750 + }, + { + "epoch": 0.06772782612616675, + "flos": 20856407053440.0, + "grad_norm": 3.755769995361356, + "learning_rate": 3.985110097330953e-06, + "loss": 0.8388, + "num_input_tokens_seen": 24211680, + "step": 751 + }, + { + "epoch": 0.06781800964963701, + "flos": 25264430771520.0, + "grad_norm": 1.6147265298983051, + "learning_rate": 3.985038857620962e-06, + "loss": 0.806, + "num_input_tokens_seen": 24244075, + "step": 752 + }, + { + "epoch": 0.06790819317310727, + "flos": 70900834917120.0, + "grad_norm": 0.6786357151857163, + "learning_rate": 3.9849674485365094e-06, + "loss": 0.5889, + "num_input_tokens_seen": 24334200, + "step": 753 + }, + { + "epoch": 0.06799837669657753, + "flos": 14335618986240.0, + "grad_norm": 2.076265119466557, + "learning_rate": 3.98489587008369e-06, + "loss": 0.7395, + "num_input_tokens_seen": 24359125, + "step": 754 + }, + { + "epoch": 0.06808856022004779, + "flos": 23807927400480.0, + "grad_norm": 2.477358620898829, + "learning_rate": 3.98482412226861e-06, + "loss": 0.6573, + "num_input_tokens_seen": 24386210, + "step": 755 + }, + { + "epoch": 0.06817874374351805, + "flos": 16373437632000.0, + "grad_norm": 1.9671823830269095, + "learning_rate": 3.984752205097391e-06, + "loss": 0.8517, + "num_input_tokens_seen": 24410160, + "step": 756 + }, + { + "epoch": 0.06826892726698833, + "flos": 18051953211360.0, + "grad_norm": 1.9798197689294468, + "learning_rate": 3.984680118576171e-06, + "loss": 0.8223, + "num_input_tokens_seen": 24437620, + "step": 757 + }, + { + "epoch": 0.06835911079045859, + "flos": 23916591809760.0, + "grad_norm": 1.8006023061554295, + "learning_rate": 3.984607862711099e-06, + "loss": 0.7695, + "num_input_tokens_seen": 24467580, + "step": 758 + }, + { + "epoch": 0.06844929431392885, + "flos": 21075185492640.0, + "grad_norm": 3.673515050954582, + "learning_rate": 3.984535437508341e-06, + "loss": 0.8507, + "num_input_tokens_seen": 24494785, + "step": 759 + }, + { + "epoch": 0.06853947783739911, + "flos": 69445186450560.0, + "grad_norm": 0.6598908381786133, + "learning_rate": 3.984462842974078e-06, + "loss": 0.5937, + "num_input_tokens_seen": 24581350, + "step": 760 + }, + { + "epoch": 0.06862966136086937, + "flos": 21875242036800.0, + "grad_norm": 3.3740959369472043, + "learning_rate": 3.984390079114502e-06, + "loss": 0.8252, + "num_input_tokens_seen": 24609370, + "step": 761 + }, + { + "epoch": 0.06871984488433963, + "flos": 14954444282400.0, + "grad_norm": 1.9724768226528568, + "learning_rate": 3.984317145935824e-06, + "loss": 0.931, + "num_input_tokens_seen": 24636840, + "step": 762 + }, + { + "epoch": 0.06881002840780989, + "flos": 26139098491200.0, + "grad_norm": 1.9687181500477393, + "learning_rate": 3.984244043444264e-06, + "loss": 0.7867, + "num_input_tokens_seen": 24666720, + "step": 763 + }, + { + "epoch": 0.06890021193128015, + "flos": 24282027886560.0, + "grad_norm": 1.7677185340037043, + "learning_rate": 3.984170771646062e-06, + "loss": 0.7648, + "num_input_tokens_seen": 24695255, + "step": 764 + }, + { + "epoch": 0.06899039545475041, + "flos": 18926360742720.0, + "grad_norm": 2.1018557432564693, + "learning_rate": 3.9840973305474695e-06, + "loss": 0.8621, + "num_input_tokens_seen": 24722870, + "step": 765 + }, + { + "epoch": 0.06908057897822067, + "flos": 20201335507680.0, + "grad_norm": 1.7500059945140076, + "learning_rate": 3.984023720154752e-06, + "loss": 0.8656, + "num_input_tokens_seen": 24749620, + "step": 766 + }, + { + "epoch": 0.06917076250169094, + "flos": 21622447533600.0, + "grad_norm": 1.7114801112542535, + "learning_rate": 3.9839499404741915e-06, + "loss": 0.8257, + "num_input_tokens_seen": 24779265, + "step": 767 + }, + { + "epoch": 0.0692609460251612, + "flos": 65853908498880.0, + "grad_norm": 0.740461933517555, + "learning_rate": 3.983875991512082e-06, + "loss": 0.6089, + "num_input_tokens_seen": 24870130, + "step": 768 + }, + { + "epoch": 0.06935112954863147, + "flos": 23001143129760.0, + "grad_norm": 2.514115822690666, + "learning_rate": 3.9838018732747345e-06, + "loss": 0.8278, + "num_input_tokens_seen": 24897810, + "step": 769 + }, + { + "epoch": 0.06944131307210173, + "flos": 24136299492960.0, + "grad_norm": 1.846010535742223, + "learning_rate": 3.9837275857684716e-06, + "loss": 0.8045, + "num_input_tokens_seen": 24926915, + "step": 770 + }, + { + "epoch": 0.06953149659557199, + "flos": 18670741337760.0, + "grad_norm": 4.195377740050813, + "learning_rate": 3.983653128999634e-06, + "loss": 0.9482, + "num_input_tokens_seen": 24950860, + "step": 771 + }, + { + "epoch": 0.06962168011904225, + "flos": 26539963082880.0, + "grad_norm": 2.0597877007872567, + "learning_rate": 3.983578502974574e-06, + "loss": 0.845, + "num_input_tokens_seen": 24979985, + "step": 772 + }, + { + "epoch": 0.06971186364251251, + "flos": 19217520171840.0, + "grad_norm": 1.9241646254528526, + "learning_rate": 3.983503707699658e-06, + "loss": 0.8022, + "num_input_tokens_seen": 25005175, + "step": 773 + }, + { + "epoch": 0.06980204716598278, + "flos": 21039422449920.0, + "grad_norm": 1.6260132931521336, + "learning_rate": 3.983428743181268e-06, + "loss": 0.844, + "num_input_tokens_seen": 25032040, + "step": 774 + }, + { + "epoch": 0.06989223068945304, + "flos": 20958603924480.0, + "grad_norm": 1.8389551564070292, + "learning_rate": 3.983353609425802e-06, + "loss": 0.8796, + "num_input_tokens_seen": 25056770, + "step": 775 + }, + { + "epoch": 0.0699824142129233, + "flos": 25846972648320.0, + "grad_norm": 1.5709124884996133, + "learning_rate": 3.983278306439671e-06, + "loss": 0.809, + "num_input_tokens_seen": 25088485, + "step": 776 + }, + { + "epoch": 0.07007259773639356, + "flos": 28361679512160.0, + "grad_norm": 1.525914999573122, + "learning_rate": 3.983202834229297e-06, + "loss": 0.8472, + "num_input_tokens_seen": 25117175, + "step": 777 + }, + { + "epoch": 0.07016278125986382, + "flos": 49673044625280.0, + "grad_norm": 1.9790534806930709, + "learning_rate": 3.983127192801123e-06, + "loss": 0.8922, + "num_input_tokens_seen": 25151220, + "step": 778 + }, + { + "epoch": 0.07025296478333408, + "flos": 26314382577600.0, + "grad_norm": 2.3595989196826603, + "learning_rate": 3.983051382161602e-06, + "loss": 0.8484, + "num_input_tokens_seen": 25180425, + "step": 779 + }, + { + "epoch": 0.07034314830680434, + "flos": 21257643342720.0, + "grad_norm": 1.7043194563193964, + "learning_rate": 3.982975402317203e-06, + "loss": 0.837, + "num_input_tokens_seen": 25211490, + "step": 780 + }, + { + "epoch": 0.07043333183027461, + "flos": 20164420202400.0, + "grad_norm": 2.181424341093293, + "learning_rate": 3.982899253274409e-06, + "loss": 0.8346, + "num_input_tokens_seen": 25238970, + "step": 781 + }, + { + "epoch": 0.07052351535374488, + "flos": 19906979479200.0, + "grad_norm": 2.379698272299978, + "learning_rate": 3.982822935039717e-06, + "loss": 0.7996, + "num_input_tokens_seen": 25265625, + "step": 782 + }, + { + "epoch": 0.07061369887721514, + "flos": 19654631013120.0, + "grad_norm": 2.0974429002521418, + "learning_rate": 3.982746447619638e-06, + "loss": 0.7483, + "num_input_tokens_seen": 25293240, + "step": 783 + }, + { + "epoch": 0.0707038824006854, + "flos": 72866012384640.0, + "grad_norm": 0.7085152327031379, + "learning_rate": 3.9826697910207e-06, + "loss": 0.5827, + "num_input_tokens_seen": 25381345, + "step": 784 + }, + { + "epoch": 0.07079406592415566, + "flos": 28179965057280.0, + "grad_norm": 1.791827667750143, + "learning_rate": 3.982592965249442e-06, + "loss": 0.795, + "num_input_tokens_seen": 25409230, + "step": 785 + }, + { + "epoch": 0.07088424944762592, + "flos": 27664749083040.0, + "grad_norm": 2.3480636573233644, + "learning_rate": 3.982515970312422e-06, + "loss": 0.8573, + "num_input_tokens_seen": 25435575, + "step": 786 + }, + { + "epoch": 0.07097443297109618, + "flos": 20273642158080.0, + "grad_norm": 2.5101220260077435, + "learning_rate": 3.982438806216207e-06, + "loss": 0.7852, + "num_input_tokens_seen": 25460700, + "step": 787 + }, + { + "epoch": 0.07106461649456644, + "flos": 39838645412640.0, + "grad_norm": 2.041937310581958, + "learning_rate": 3.982361472967382e-06, + "loss": 0.7992, + "num_input_tokens_seen": 25492560, + "step": 788 + }, + { + "epoch": 0.0711548000180367, + "flos": 23225013826080.0, + "grad_norm": 1.9177134046706805, + "learning_rate": 3.982283970572546e-06, + "loss": 0.8858, + "num_input_tokens_seen": 25518260, + "step": 789 + }, + { + "epoch": 0.07124498354150696, + "flos": 12951385095840.0, + "grad_norm": 2.527436008567731, + "learning_rate": 3.982206299038311e-06, + "loss": 0.7895, + "num_input_tokens_seen": 25544100, + "step": 790 + }, + { + "epoch": 0.07133516706497722, + "flos": 17064904106400.0, + "grad_norm": 1.835321525968535, + "learning_rate": 3.9821284583713054e-06, + "loss": 0.888, + "num_input_tokens_seen": 25570860, + "step": 791 + }, + { + "epoch": 0.07142535058844748, + "flos": 17980166937600.0, + "grad_norm": 2.7546667902054933, + "learning_rate": 3.98205044857817e-06, + "loss": 0.7955, + "num_input_tokens_seen": 25597555, + "step": 792 + }, + { + "epoch": 0.07151553411191776, + "flos": 69745601149920.0, + "grad_norm": 0.7417931964897562, + "learning_rate": 3.981972269665561e-06, + "loss": 0.6117, + "num_input_tokens_seen": 25686020, + "step": 793 + }, + { + "epoch": 0.07160571763538802, + "flos": 26723164328160.0, + "grad_norm": 1.8184253914196056, + "learning_rate": 3.98189392164015e-06, + "loss": 0.8065, + "num_input_tokens_seen": 25718720, + "step": 794 + }, + { + "epoch": 0.07169590115885828, + "flos": 27265259772480.0, + "grad_norm": 1.85028946648132, + "learning_rate": 3.981815404508621e-06, + "loss": 0.854, + "num_input_tokens_seen": 25744515, + "step": 795 + }, + { + "epoch": 0.07178608468232854, + "flos": 21476310272640.0, + "grad_norm": 2.2086293603909213, + "learning_rate": 3.981736718277674e-06, + "loss": 0.8194, + "num_input_tokens_seen": 25773010, + "step": 796 + }, + { + "epoch": 0.0718762682057988, + "flos": 15683457948000.0, + "grad_norm": 2.160046096945659, + "learning_rate": 3.9816578629540235e-06, + "loss": 0.8478, + "num_input_tokens_seen": 25798695, + "step": 797 + }, + { + "epoch": 0.07196645172926906, + "flos": 28939352150400.0, + "grad_norm": 2.605844542069865, + "learning_rate": 3.981578838544398e-06, + "loss": 0.8231, + "num_input_tokens_seen": 25826695, + "step": 798 + }, + { + "epoch": 0.07205663525273932, + "flos": 66627308591520.0, + "grad_norm": 0.6683741895058267, + "learning_rate": 3.981499645055539e-06, + "loss": 0.566, + "num_input_tokens_seen": 25927685, + "step": 799 + }, + { + "epoch": 0.07214681877620958, + "flos": 60721368656160.0, + "grad_norm": 0.6514686379645989, + "learning_rate": 3.981420282494204e-06, + "loss": 0.6001, + "num_input_tokens_seen": 26024045, + "step": 800 + }, + { + "epoch": 0.07223700229967984, + "flos": 70613095105920.0, + "grad_norm": 0.7441136312720067, + "learning_rate": 3.981340750867166e-06, + "loss": 0.5993, + "num_input_tokens_seen": 26109055, + "step": 801 + }, + { + "epoch": 0.0723271858231501, + "flos": 23878635751200.0, + "grad_norm": 1.6217453715536356, + "learning_rate": 3.981261050181209e-06, + "loss": 0.7995, + "num_input_tokens_seen": 26138125, + "step": 802 + }, + { + "epoch": 0.07241736934662037, + "flos": 25301977962720.0, + "grad_norm": 2.0498223669735434, + "learning_rate": 3.9811811804431355e-06, + "loss": 0.7667, + "num_input_tokens_seen": 26166785, + "step": 803 + }, + { + "epoch": 0.07250755287009064, + "flos": 24281396000640.0, + "grad_norm": 2.0129495059491624, + "learning_rate": 3.981101141659759e-06, + "loss": 0.8529, + "num_input_tokens_seen": 26195365, + "step": 804 + }, + { + "epoch": 0.0725977363935609, + "flos": 19069896120480.0, + "grad_norm": 1.8044621050230885, + "learning_rate": 3.98102093383791e-06, + "loss": 0.8408, + "num_input_tokens_seen": 26223310, + "step": 805 + }, + { + "epoch": 0.07268791991703116, + "flos": 27956280209760.0, + "grad_norm": 5.54675481961576, + "learning_rate": 3.9809405569844315e-06, + "loss": 0.7335, + "num_input_tokens_seen": 26251745, + "step": 806 + }, + { + "epoch": 0.07277810344050142, + "flos": 20489335507200.0, + "grad_norm": 1.682567097543211, + "learning_rate": 3.980860011106182e-06, + "loss": 0.7746, + "num_input_tokens_seen": 26279365, + "step": 807 + }, + { + "epoch": 0.07286828696397168, + "flos": 23297617834560.0, + "grad_norm": 1.756388787049668, + "learning_rate": 3.980779296210033e-06, + "loss": 0.7503, + "num_input_tokens_seen": 26310470, + "step": 808 + }, + { + "epoch": 0.07295847048744195, + "flos": 19764224666400.0, + "grad_norm": 2.8354308461309574, + "learning_rate": 3.980698412302874e-06, + "loss": 0.8912, + "num_input_tokens_seen": 26337455, + "step": 809 + }, + { + "epoch": 0.0730486540109122, + "flos": 21585383549280.0, + "grad_norm": 1.9911978264227805, + "learning_rate": 3.980617359391604e-06, + "loss": 0.8262, + "num_input_tokens_seen": 26364510, + "step": 810 + }, + { + "epoch": 0.07313883753438247, + "flos": 30873747323040.0, + "grad_norm": 1.9883403656870613, + "learning_rate": 3.98053613748314e-06, + "loss": 0.7691, + "num_input_tokens_seen": 26395895, + "step": 811 + }, + { + "epoch": 0.07322902105785273, + "flos": 43882310976960.0, + "grad_norm": 1.5394512789215968, + "learning_rate": 3.980454746584413e-06, + "loss": 0.6743, + "num_input_tokens_seen": 26426045, + "step": 812 + }, + { + "epoch": 0.07331920458132299, + "flos": 20274534232320.0, + "grad_norm": 2.971641818584707, + "learning_rate": 3.9803731867023665e-06, + "loss": 0.8981, + "num_input_tokens_seen": 26452705, + "step": 813 + }, + { + "epoch": 0.07340938810479325, + "flos": 22601914007520.0, + "grad_norm": 2.103363966721277, + "learning_rate": 3.9802914578439596e-06, + "loss": 0.7979, + "num_input_tokens_seen": 26481080, + "step": 814 + }, + { + "epoch": 0.07349957162826351, + "flos": 24537052575360.0, + "grad_norm": 1.8534308555842822, + "learning_rate": 3.980209560016167e-06, + "loss": 0.9172, + "num_input_tokens_seen": 26509240, + "step": 815 + }, + { + "epoch": 0.07358975515173379, + "flos": 27158565360480.0, + "grad_norm": 1.9296369661826405, + "learning_rate": 3.980127493225975e-06, + "loss": 0.904, + "num_input_tokens_seen": 26539635, + "step": 816 + }, + { + "epoch": 0.07367993867520405, + "flos": 23735806598880.0, + "grad_norm": 2.8457384807665527, + "learning_rate": 3.980045257480387e-06, + "loss": 0.8436, + "num_input_tokens_seen": 26568420, + "step": 817 + }, + { + "epoch": 0.0737701221986743, + "flos": 24026817348960.0, + "grad_norm": 2.542446569220783, + "learning_rate": 3.9799628527864205e-06, + "loss": 0.7707, + "num_input_tokens_seen": 26595545, + "step": 818 + }, + { + "epoch": 0.07386030572214457, + "flos": 23917929921120.0, + "grad_norm": 1.990036584910802, + "learning_rate": 3.979880279151106e-06, + "loss": 0.7968, + "num_input_tokens_seen": 26624950, + "step": 819 + }, + { + "epoch": 0.07395048924561483, + "flos": 26030136723840.0, + "grad_norm": 1.92123414806808, + "learning_rate": 3.979797536581489e-06, + "loss": 0.7548, + "num_input_tokens_seen": 26655695, + "step": 820 + }, + { + "epoch": 0.07404067276908509, + "flos": 26905250480640.0, + "grad_norm": 2.721056431205836, + "learning_rate": 3.97971462508463e-06, + "loss": 0.6517, + "num_input_tokens_seen": 26682665, + "step": 821 + }, + { + "epoch": 0.07413085629255535, + "flos": 27924828859200.0, + "grad_norm": 2.4042053453142773, + "learning_rate": 3.979631544667603e-06, + "loss": 0.7959, + "num_input_tokens_seen": 26713390, + "step": 822 + }, + { + "epoch": 0.07422103981602561, + "flos": 38230801014240.0, + "grad_norm": 2.6770994628737, + "learning_rate": 3.979548295337496e-06, + "loss": 0.8239, + "num_input_tokens_seen": 26740670, + "step": 823 + }, + { + "epoch": 0.07431122333949587, + "flos": 26208654579360.0, + "grad_norm": 1.806425834465787, + "learning_rate": 3.9794648771014146e-06, + "loss": 0.7777, + "num_input_tokens_seen": 26772125, + "step": 824 + }, + { + "epoch": 0.07440140686296613, + "flos": 22350903652800.0, + "grad_norm": 1.9473580019971943, + "learning_rate": 3.9793812899664745e-06, + "loss": 0.8144, + "num_input_tokens_seen": 26799605, + "step": 825 + }, + { + "epoch": 0.0744915903864364, + "flos": 29161736056320.0, + "grad_norm": 2.946695261193683, + "learning_rate": 3.979297533939809e-06, + "loss": 0.7694, + "num_input_tokens_seen": 26827190, + "step": 826 + }, + { + "epoch": 0.07458177390990665, + "flos": 23480521721760.0, + "grad_norm": 2.181609588101744, + "learning_rate": 3.979213609028564e-06, + "loss": 0.788, + "num_input_tokens_seen": 26855695, + "step": 827 + }, + { + "epoch": 0.07467195743337693, + "flos": 20857336297440.0, + "grad_norm": 1.8694883718646744, + "learning_rate": 3.979129515239901e-06, + "loss": 0.8258, + "num_input_tokens_seen": 26884355, + "step": 828 + }, + { + "epoch": 0.07476214095684719, + "flos": 26867517440640.0, + "grad_norm": 1.8729428132141233, + "learning_rate": 3.979045252580994e-06, + "loss": 0.8165, + "num_input_tokens_seen": 26915595, + "step": 829 + }, + { + "epoch": 0.07485232448031745, + "flos": 37615767033600.0, + "grad_norm": 2.2098663215425804, + "learning_rate": 3.978960821059034e-06, + "loss": 0.7551, + "num_input_tokens_seen": 26946505, + "step": 830 + }, + { + "epoch": 0.07494250800378771, + "flos": 23695731864000.0, + "grad_norm": 2.0371921189676097, + "learning_rate": 3.978876220681225e-06, + "loss": 0.8504, + "num_input_tokens_seen": 26973305, + "step": 831 + }, + { + "epoch": 0.07503269152725797, + "flos": 21549100129920.0, + "grad_norm": 1.928913470873135, + "learning_rate": 3.978791451454786e-06, + "loss": 0.7399, + "num_input_tokens_seen": 27001795, + "step": 832 + }, + { + "epoch": 0.07512287505072823, + "flos": 24934163021280.0, + "grad_norm": 1.8941111305286393, + "learning_rate": 3.978706513386949e-06, + "loss": 0.771, + "num_input_tokens_seen": 27031210, + "step": 833 + }, + { + "epoch": 0.0752130585741985, + "flos": 54843020149920.0, + "grad_norm": 2.7602820213636465, + "learning_rate": 3.978621406484962e-06, + "loss": 0.8213, + "num_input_tokens_seen": 27065670, + "step": 834 + }, + { + "epoch": 0.07530324209766875, + "flos": 28143681637920.0, + "grad_norm": 3.3147366727841145, + "learning_rate": 3.978536130756086e-06, + "loss": 0.8009, + "num_input_tokens_seen": 27096090, + "step": 835 + }, + { + "epoch": 0.07539342562113902, + "flos": 22931698550880.0, + "grad_norm": 2.0545988462635187, + "learning_rate": 3.978450686207599e-06, + "loss": 0.8743, + "num_input_tokens_seen": 27125370, + "step": 836 + }, + { + "epoch": 0.07548360914460928, + "flos": 17104569973920.0, + "grad_norm": 2.1742913505777612, + "learning_rate": 3.978365072846789e-06, + "loss": 0.791, + "num_input_tokens_seen": 27151560, + "step": 837 + }, + { + "epoch": 0.07557379266807954, + "flos": 25702247838240.0, + "grad_norm": 9.479424531105144, + "learning_rate": 3.9782792906809625e-06, + "loss": 0.7594, + "num_input_tokens_seen": 27180425, + "step": 838 + }, + { + "epoch": 0.0756639761915498, + "flos": 20675250144960.0, + "grad_norm": 2.3906603043885184, + "learning_rate": 3.97819333971744e-06, + "loss": 0.7943, + "num_input_tokens_seen": 27207455, + "step": 839 + }, + { + "epoch": 0.07575415971502007, + "flos": 23147540579040.0, + "grad_norm": 3.5926348739079006, + "learning_rate": 3.978107219963553e-06, + "loss": 0.8211, + "num_input_tokens_seen": 27235885, + "step": 840 + }, + { + "epoch": 0.07584434323849033, + "flos": 19210755275520.0, + "grad_norm": 2.551057809321985, + "learning_rate": 3.978020931426651e-06, + "loss": 0.718, + "num_input_tokens_seen": 27263610, + "step": 841 + }, + { + "epoch": 0.0759345267619606, + "flos": 28145168428320.0, + "grad_norm": 2.1969911937997257, + "learning_rate": 3.977934474114096e-06, + "loss": 0.7589, + "num_input_tokens_seen": 27294335, + "step": 842 + }, + { + "epoch": 0.07602471028543085, + "flos": 70408552684800.0, + "grad_norm": 0.6404249762908303, + "learning_rate": 3.977847848033267e-06, + "loss": 0.5779, + "num_input_tokens_seen": 27390535, + "step": 843 + }, + { + "epoch": 0.07611489380890112, + "flos": 27669692661120.0, + "grad_norm": 1.7542349118740048, + "learning_rate": 3.977761053191553e-06, + "loss": 0.8523, + "num_input_tokens_seen": 27417140, + "step": 844 + }, + { + "epoch": 0.07620507733237138, + "flos": 25518154518720.0, + "grad_norm": 2.8495926415290045, + "learning_rate": 3.977674089596361e-06, + "loss": 0.7894, + "num_input_tokens_seen": 27448470, + "step": 845 + }, + { + "epoch": 0.07629526085584164, + "flos": 52979773631520.0, + "grad_norm": 0.6889488602263723, + "learning_rate": 3.97758695725511e-06, + "loss": 0.5748, + "num_input_tokens_seen": 27533360, + "step": 846 + }, + { + "epoch": 0.0763854443793119, + "flos": 31822877539200.0, + "grad_norm": 2.631057018171106, + "learning_rate": 3.977499656175236e-06, + "loss": 0.8038, + "num_input_tokens_seen": 27566515, + "step": 847 + }, + { + "epoch": 0.07647562790278216, + "flos": 25478042614080.0, + "grad_norm": 2.7337026586911852, + "learning_rate": 3.977412186364187e-06, + "loss": 0.717, + "num_input_tokens_seen": 27596320, + "step": 848 + }, + { + "epoch": 0.07656581142625242, + "flos": 19764113157120.0, + "grad_norm": 2.0678200743595454, + "learning_rate": 3.977324547829428e-06, + "loss": 0.8328, + "num_input_tokens_seen": 27623210, + "step": 849 + }, + { + "epoch": 0.07665599494972268, + "flos": 16988954819520.0, + "grad_norm": 2.28700950658613, + "learning_rate": 3.977236740578435e-06, + "loss": 0.8326, + "num_input_tokens_seen": 27648160, + "step": 850 + }, + { + "epoch": 0.07674617847319294, + "flos": 23553460258080.0, + "grad_norm": 2.959456181444925, + "learning_rate": 3.9771487646187015e-06, + "loss": 0.7957, + "num_input_tokens_seen": 27677070, + "step": 851 + }, + { + "epoch": 0.07683636199666322, + "flos": 17246135354400.0, + "grad_norm": 4.0982408424749375, + "learning_rate": 3.9770606199577325e-06, + "loss": 0.8456, + "num_input_tokens_seen": 27702665, + "step": 852 + }, + { + "epoch": 0.07692654552013348, + "flos": 19545706415520.0, + "grad_norm": 2.8382022946435344, + "learning_rate": 3.9769723066030505e-06, + "loss": 0.7394, + "num_input_tokens_seen": 27730085, + "step": 853 + }, + { + "epoch": 0.07701672904360374, + "flos": 20122115281920.0, + "grad_norm": 3.2459840170701253, + "learning_rate": 3.976883824562191e-06, + "loss": 0.8679, + "num_input_tokens_seen": 27755805, + "step": 854 + }, + { + "epoch": 0.077106912567074, + "flos": 19581729646560.0, + "grad_norm": 2.025307163442801, + "learning_rate": 3.976795173842703e-06, + "loss": 0.807, + "num_input_tokens_seen": 27782290, + "step": 855 + }, + { + "epoch": 0.07719709609054426, + "flos": 32401107723840.0, + "grad_norm": 4.0048055392523825, + "learning_rate": 3.97670635445215e-06, + "loss": 0.7852, + "num_input_tokens_seen": 27810420, + "step": 856 + }, + { + "epoch": 0.07728727961401452, + "flos": 23297766513600.0, + "grad_norm": 2.079061770293525, + "learning_rate": 3.976617366398112e-06, + "loss": 0.7781, + "num_input_tokens_seen": 27837010, + "step": 857 + }, + { + "epoch": 0.07737746313748478, + "flos": 24463444983360.0, + "grad_norm": 1.418401814894529, + "learning_rate": 3.976528209688181e-06, + "loss": 0.8371, + "num_input_tokens_seen": 27866500, + "step": 858 + }, + { + "epoch": 0.07746764666095504, + "flos": 24024512823840.0, + "grad_norm": 1.737964661303684, + "learning_rate": 3.976438884329965e-06, + "loss": 0.8125, + "num_input_tokens_seen": 27898645, + "step": 859 + }, + { + "epoch": 0.0775578301844253, + "flos": 25443394664160.0, + "grad_norm": 1.82928990475363, + "learning_rate": 3.976349390331085e-06, + "loss": 0.9374, + "num_input_tokens_seen": 27928350, + "step": 860 + }, + { + "epoch": 0.07764801370789556, + "flos": 23913432380160.0, + "grad_norm": 2.528453082285461, + "learning_rate": 3.976259727699178e-06, + "loss": 0.7743, + "num_input_tokens_seen": 27955385, + "step": 861 + }, + { + "epoch": 0.07773819723136582, + "flos": 18338169062400.0, + "grad_norm": 2.4403667708976586, + "learning_rate": 3.976169896441895e-06, + "loss": 0.7625, + "num_input_tokens_seen": 27980565, + "step": 862 + }, + { + "epoch": 0.07782838075483609, + "flos": 23881497822720.0, + "grad_norm": 2.3610849815368913, + "learning_rate": 3.976079896566898e-06, + "loss": 0.8401, + "num_input_tokens_seen": 28009140, + "step": 863 + }, + { + "epoch": 0.07791856427830636, + "flos": 29706805081440.0, + "grad_norm": 2.0940087002024224, + "learning_rate": 3.97598972808187e-06, + "loss": 0.8733, + "num_input_tokens_seen": 28038040, + "step": 864 + }, + { + "epoch": 0.07800874780177662, + "flos": 22788497700960.0, + "grad_norm": 2.127885318204249, + "learning_rate": 3.975899390994501e-06, + "loss": 0.808, + "num_input_tokens_seen": 28067820, + "step": 865 + }, + { + "epoch": 0.07809893132524688, + "flos": 22349231013600.0, + "grad_norm": 1.6350397862251127, + "learning_rate": 3.975808885312502e-06, + "loss": 0.8713, + "num_input_tokens_seen": 28097390, + "step": 866 + }, + { + "epoch": 0.07818911484871714, + "flos": 22132534080960.0, + "grad_norm": 1.755623908797971, + "learning_rate": 3.975718211043594e-06, + "loss": 0.8553, + "num_input_tokens_seen": 28126580, + "step": 867 + }, + { + "epoch": 0.0782792983721874, + "flos": 22095321417600.0, + "grad_norm": 2.5973598684010315, + "learning_rate": 3.975627368195515e-06, + "loss": 0.7237, + "num_input_tokens_seen": 28153765, + "step": 868 + }, + { + "epoch": 0.07836948189565766, + "flos": 27635565087840.0, + "grad_norm": 1.6951607491216132, + "learning_rate": 3.975536356776015e-06, + "loss": 0.7772, + "num_input_tokens_seen": 28186050, + "step": 869 + }, + { + "epoch": 0.07845966541912792, + "flos": 66662997294720.0, + "grad_norm": 0.7400085185179259, + "learning_rate": 3.975445176792861e-06, + "loss": 0.6516, + "num_input_tokens_seen": 28282075, + "step": 870 + }, + { + "epoch": 0.07854984894259819, + "flos": 20963733351360.0, + "grad_norm": 2.0388874548832803, + "learning_rate": 3.975353828253831e-06, + "loss": 0.815, + "num_input_tokens_seen": 28312830, + "step": 871 + }, + { + "epoch": 0.07864003246606845, + "flos": 24791148020160.0, + "grad_norm": 4.035827167345328, + "learning_rate": 3.97526231116672e-06, + "loss": 0.8093, + "num_input_tokens_seen": 28342420, + "step": 872 + }, + { + "epoch": 0.07873021598953871, + "flos": 26321630680800.0, + "grad_norm": 3.554651281151334, + "learning_rate": 3.975170625539338e-06, + "loss": 0.7496, + "num_input_tokens_seen": 28371795, + "step": 873 + }, + { + "epoch": 0.07882039951300897, + "flos": 24536086161600.0, + "grad_norm": 1.6224494137394339, + "learning_rate": 3.975078771379507e-06, + "loss": 0.8062, + "num_input_tokens_seen": 28404155, + "step": 874 + }, + { + "epoch": 0.07891058303647923, + "flos": 21294335629440.0, + "grad_norm": 3.093818124348594, + "learning_rate": 3.974986748695064e-06, + "loss": 0.7186, + "num_input_tokens_seen": 28429285, + "step": 875 + }, + { + "epoch": 0.0790007665599495, + "flos": 61866492248640.0, + "grad_norm": 0.7207659355436961, + "learning_rate": 3.974894557493862e-06, + "loss": 0.5894, + "num_input_tokens_seen": 28510930, + "step": 876 + }, + { + "epoch": 0.07909095008341976, + "flos": 18598546196640.0, + "grad_norm": 2.012769465443201, + "learning_rate": 3.974802197783768e-06, + "loss": 0.8524, + "num_input_tokens_seen": 28538705, + "step": 877 + }, + { + "epoch": 0.07918113360689003, + "flos": 13607051357760.0, + "grad_norm": 2.15857626334586, + "learning_rate": 3.974709669572661e-06, + "loss": 0.8665, + "num_input_tokens_seen": 28564580, + "step": 878 + }, + { + "epoch": 0.07927131713036029, + "flos": 17687409208800.0, + "grad_norm": 1.7951466084487906, + "learning_rate": 3.974616972868436e-06, + "loss": 0.799, + "num_input_tokens_seen": 28592085, + "step": 879 + }, + { + "epoch": 0.07936150065383055, + "flos": 23405538848640.0, + "grad_norm": 1.7979503565078077, + "learning_rate": 3.974524107679003e-06, + "loss": 0.8137, + "num_input_tokens_seen": 28618260, + "step": 880 + }, + { + "epoch": 0.07945168417730081, + "flos": 31093938213120.0, + "grad_norm": 1.8351025569851185, + "learning_rate": 3.974431074012286e-06, + "loss": 0.7984, + "num_input_tokens_seen": 28650475, + "step": 881 + }, + { + "epoch": 0.07954186770077107, + "flos": 24172359893760.0, + "grad_norm": 2.7073299799095762, + "learning_rate": 3.974337871876223e-06, + "loss": 0.8322, + "num_input_tokens_seen": 28676020, + "step": 882 + }, + { + "epoch": 0.07963205122424133, + "flos": 29778256827360.0, + "grad_norm": 1.5512948903144426, + "learning_rate": 3.974244501278766e-06, + "loss": 0.7748, + "num_input_tokens_seen": 28705870, + "step": 883 + }, + { + "epoch": 0.07972223474771159, + "flos": 31058398188960.0, + "grad_norm": 1.719952110910203, + "learning_rate": 3.974150962227883e-06, + "loss": 0.7575, + "num_input_tokens_seen": 28740180, + "step": 884 + }, + { + "epoch": 0.07981241827118185, + "flos": 19837014523680.0, + "grad_norm": 2.0535579230770717, + "learning_rate": 3.9740572547315535e-06, + "loss": 0.7342, + "num_input_tokens_seen": 28768565, + "step": 885 + }, + { + "epoch": 0.07990260179465211, + "flos": 22860321144480.0, + "grad_norm": 2.458758839997156, + "learning_rate": 3.973963378797775e-06, + "loss": 0.8462, + "num_input_tokens_seen": 28797515, + "step": 886 + }, + { + "epoch": 0.07999278531812237, + "flos": 24243551451360.0, + "grad_norm": 1.775859573590467, + "learning_rate": 3.973869334434556e-06, + "loss": 0.8221, + "num_input_tokens_seen": 28828660, + "step": 887 + }, + { + "epoch": 0.08008296884159265, + "flos": 47853112344480.0, + "grad_norm": 1.8202007734204222, + "learning_rate": 3.973775121649922e-06, + "loss": 0.8226, + "num_input_tokens_seen": 28862345, + "step": 888 + }, + { + "epoch": 0.08017315236506291, + "flos": 23336726155680.0, + "grad_norm": 1.7856026723585332, + "learning_rate": 3.973680740451911e-06, + "loss": 0.7212, + "num_input_tokens_seen": 28894665, + "step": 889 + }, + { + "epoch": 0.08026333588853317, + "flos": 19655077050240.0, + "grad_norm": 2.08504624203419, + "learning_rate": 3.9735861908485776e-06, + "loss": 0.7691, + "num_input_tokens_seen": 28922125, + "step": 890 + }, + { + "epoch": 0.08035351941200343, + "flos": 38413184524800.0, + "grad_norm": 3.135304209233298, + "learning_rate": 3.973491472847987e-06, + "loss": 0.8286, + "num_input_tokens_seen": 28952670, + "step": 891 + }, + { + "epoch": 0.08044370293547369, + "flos": 34152859197600.0, + "grad_norm": 1.6937555995721914, + "learning_rate": 3.973396586458222e-06, + "loss": 0.738, + "num_input_tokens_seen": 28985555, + "step": 892 + }, + { + "epoch": 0.08053388645894395, + "flos": 24680253425280.0, + "grad_norm": 4.32823649238731, + "learning_rate": 3.97330153168738e-06, + "loss": 0.7929, + "num_input_tokens_seen": 29016095, + "step": 893 + }, + { + "epoch": 0.08062406998241421, + "flos": 25808198855040.0, + "grad_norm": 1.7307702484290888, + "learning_rate": 3.973206308543571e-06, + "loss": 0.7519, + "num_input_tokens_seen": 29048180, + "step": 894 + }, + { + "epoch": 0.08071425350588447, + "flos": 22460497306080.0, + "grad_norm": 2.145541927464039, + "learning_rate": 3.973110917034918e-06, + "loss": 0.7792, + "num_input_tokens_seen": 29076770, + "step": 895 + }, + { + "epoch": 0.08080443702935473, + "flos": 29455125671040.0, + "grad_norm": 1.7170659224397251, + "learning_rate": 3.973015357169563e-06, + "loss": 0.7156, + "num_input_tokens_seen": 29105410, + "step": 896 + }, + { + "epoch": 0.080894620552825, + "flos": 40602604386240.0, + "grad_norm": 2.5607225720077125, + "learning_rate": 3.972919628955659e-06, + "loss": 0.6998, + "num_input_tokens_seen": 29135255, + "step": 897 + }, + { + "epoch": 0.08098480407629526, + "flos": 19172092991520.0, + "grad_norm": 2.4355246535987587, + "learning_rate": 3.972823732401373e-06, + "loss": 0.7948, + "num_input_tokens_seen": 29158075, + "step": 898 + }, + { + "epoch": 0.08107498759976552, + "flos": 26431112824800.0, + "grad_norm": 1.7873505643255667, + "learning_rate": 3.972727667514888e-06, + "loss": 0.7994, + "num_input_tokens_seen": 29185440, + "step": 899 + }, + { + "epoch": 0.08116517112323579, + "flos": 22314917591520.0, + "grad_norm": 1.6284353915859549, + "learning_rate": 3.972631434304402e-06, + "loss": 0.7667, + "num_input_tokens_seen": 29216030, + "step": 900 + }, + { + "epoch": 0.08125535464670605, + "flos": 18597914310720.0, + "grad_norm": 2.1150796116844934, + "learning_rate": 3.972535032778124e-06, + "loss": 0.9202, + "num_input_tokens_seen": 29241495, + "step": 901 + }, + { + "epoch": 0.08134553817017631, + "flos": 27632888865120.0, + "grad_norm": 2.251274807383989, + "learning_rate": 3.97243846294428e-06, + "loss": 0.8914, + "num_input_tokens_seen": 29271195, + "step": 902 + }, + { + "epoch": 0.08143572169364657, + "flos": 20346840882720.0, + "grad_norm": 1.6176341448997686, + "learning_rate": 3.972341724811111e-06, + "loss": 0.9421, + "num_input_tokens_seen": 29299090, + "step": 903 + }, + { + "epoch": 0.08152590521711683, + "flos": 22932925152960.0, + "grad_norm": 2.0009858562650202, + "learning_rate": 3.972244818386872e-06, + "loss": 0.7491, + "num_input_tokens_seen": 29327530, + "step": 904 + }, + { + "epoch": 0.0816160887405871, + "flos": 22605407964960.0, + "grad_norm": 1.782753360178617, + "learning_rate": 3.972147743679828e-06, + "loss": 0.7455, + "num_input_tokens_seen": 29354465, + "step": 905 + }, + { + "epoch": 0.08170627226405736, + "flos": 21221545772160.0, + "grad_norm": 2.1277871646568913, + "learning_rate": 3.972050500698265e-06, + "loss": 0.7689, + "num_input_tokens_seen": 29384960, + "step": 906 + }, + { + "epoch": 0.08179645578752762, + "flos": 25447669186560.0, + "grad_norm": 1.8350451309516598, + "learning_rate": 3.971953089450481e-06, + "loss": 0.7965, + "num_input_tokens_seen": 29412875, + "step": 907 + }, + { + "epoch": 0.08188663931099788, + "flos": 21221545772160.0, + "grad_norm": 1.8200103116005346, + "learning_rate": 3.971855509944784e-06, + "loss": 0.876, + "num_input_tokens_seen": 29440935, + "step": 908 + }, + { + "epoch": 0.08197682283446814, + "flos": 27592070735040.0, + "grad_norm": 1.662810235010306, + "learning_rate": 3.971757762189504e-06, + "loss": 0.791, + "num_input_tokens_seen": 29472000, + "step": 909 + }, + { + "epoch": 0.0820670063579384, + "flos": 27232693329120.0, + "grad_norm": 2.495047814287545, + "learning_rate": 3.9716598461929785e-06, + "loss": 0.7925, + "num_input_tokens_seen": 29501575, + "step": 910 + }, + { + "epoch": 0.08215718988140866, + "flos": 23844210819840.0, + "grad_norm": 1.7272191934868149, + "learning_rate": 3.971561761963563e-06, + "loss": 0.7786, + "num_input_tokens_seen": 29529775, + "step": 911 + }, + { + "epoch": 0.08224737340487893, + "flos": 15610147714080.0, + "grad_norm": 2.057406638751551, + "learning_rate": 3.971463509509628e-06, + "loss": 0.8818, + "num_input_tokens_seen": 29557115, + "step": 912 + }, + { + "epoch": 0.0823375569283492, + "flos": 16557939818880.0, + "grad_norm": 2.8614366113712872, + "learning_rate": 3.9713650888395555e-06, + "loss": 0.7934, + "num_input_tokens_seen": 29581215, + "step": 913 + }, + { + "epoch": 0.08242774045181946, + "flos": 19508828280000.0, + "grad_norm": 1.9445990054448061, + "learning_rate": 3.9712664999617425e-06, + "loss": 0.7925, + "num_input_tokens_seen": 29605060, + "step": 914 + }, + { + "epoch": 0.08251792397528972, + "flos": 24464411397120.0, + "grad_norm": 2.0632417674333143, + "learning_rate": 3.971167742884603e-06, + "loss": 0.7766, + "num_input_tokens_seen": 29634025, + "step": 915 + }, + { + "epoch": 0.08260810749875998, + "flos": 19212948291360.0, + "grad_norm": 1.9772674750813373, + "learning_rate": 3.971068817616564e-06, + "loss": 0.8532, + "num_input_tokens_seen": 29661645, + "step": 916 + }, + { + "epoch": 0.08269829102223024, + "flos": 28289558710560.0, + "grad_norm": 2.0481083772116864, + "learning_rate": 3.970969724166064e-06, + "loss": 0.8271, + "num_input_tokens_seen": 29693300, + "step": 917 + }, + { + "epoch": 0.0827884745457005, + "flos": 69651841814400.0, + "grad_norm": 0.8297232553712384, + "learning_rate": 3.970870462541559e-06, + "loss": 0.5777, + "num_input_tokens_seen": 29785140, + "step": 918 + }, + { + "epoch": 0.08287865806917076, + "flos": 24718246653600.0, + "grad_norm": 1.5500944908627432, + "learning_rate": 3.97077103275152e-06, + "loss": 0.821, + "num_input_tokens_seen": 29816060, + "step": 919 + }, + { + "epoch": 0.08296884159264102, + "flos": 31891392874080.0, + "grad_norm": 2.222156971397337, + "learning_rate": 3.970671434804428e-06, + "loss": 0.788, + "num_input_tokens_seen": 29848965, + "step": 920 + }, + { + "epoch": 0.08305902511611128, + "flos": 24898251299520.0, + "grad_norm": 1.8294752954033264, + "learning_rate": 3.970571668708784e-06, + "loss": 0.7985, + "num_input_tokens_seen": 29879670, + "step": 921 + }, + { + "epoch": 0.08314920863958154, + "flos": 67144606072320.0, + "grad_norm": 0.6616010036444748, + "learning_rate": 3.9704717344731e-06, + "loss": 0.5871, + "num_input_tokens_seen": 29980395, + "step": 922 + }, + { + "epoch": 0.0832393921630518, + "flos": 66261463647360.0, + "grad_norm": 0.718787358183867, + "learning_rate": 3.9703716321059026e-06, + "loss": 0.603, + "num_input_tokens_seen": 30081175, + "step": 923 + }, + { + "epoch": 0.08332957568652208, + "flos": 26066457312960.0, + "grad_norm": 2.3480193861724814, + "learning_rate": 3.9702713616157325e-06, + "loss": 0.891, + "num_input_tokens_seen": 30110635, + "step": 924 + }, + { + "epoch": 0.08341975920999234, + "flos": 24063026428800.0, + "grad_norm": 2.269593229777068, + "learning_rate": 3.9701709230111455e-06, + "loss": 0.8155, + "num_input_tokens_seen": 30137155, + "step": 925 + }, + { + "epoch": 0.0835099427334626, + "flos": 17687818076160.0, + "grad_norm": 2.5700549892375997, + "learning_rate": 3.970070316300713e-06, + "loss": 0.7696, + "num_input_tokens_seen": 30164485, + "step": 926 + }, + { + "epoch": 0.08360012625693286, + "flos": 35576164239360.0, + "grad_norm": 2.1045901813394408, + "learning_rate": 3.969969541493017e-06, + "loss": 0.8216, + "num_input_tokens_seen": 30194195, + "step": 927 + }, + { + "epoch": 0.08369030978040312, + "flos": 20050737875520.0, + "grad_norm": 2.0605723302633834, + "learning_rate": 3.969868598596658e-06, + "loss": 0.8541, + "num_input_tokens_seen": 30222150, + "step": 928 + }, + { + "epoch": 0.08378049330387338, + "flos": 34081556130720.0, + "grad_norm": 1.500275824936222, + "learning_rate": 3.969767487620249e-06, + "loss": 0.7429, + "num_input_tokens_seen": 30254750, + "step": 929 + }, + { + "epoch": 0.08387067682734364, + "flos": 23257505929920.0, + "grad_norm": 1.9687469119093155, + "learning_rate": 3.969666208572416e-06, + "loss": 0.8504, + "num_input_tokens_seen": 30284795, + "step": 930 + }, + { + "epoch": 0.0839608603508139, + "flos": 23224976656320.0, + "grad_norm": 1.957541576250899, + "learning_rate": 3.969564761461802e-06, + "loss": 0.8678, + "num_input_tokens_seen": 30310190, + "step": 931 + }, + { + "epoch": 0.08405104387428416, + "flos": 26030582760960.0, + "grad_norm": 1.6644169246422738, + "learning_rate": 3.969463146297062e-06, + "loss": 0.7815, + "num_input_tokens_seen": 30342865, + "step": 932 + }, + { + "epoch": 0.08414122739775443, + "flos": 23915216528640.0, + "grad_norm": 4.297376568737624, + "learning_rate": 3.969361363086867e-06, + "loss": 0.8263, + "num_input_tokens_seen": 30366530, + "step": 933 + }, + { + "epoch": 0.08423141092122469, + "flos": 31855890019680.0, + "grad_norm": 3.053481075954685, + "learning_rate": 3.9692594118399014e-06, + "loss": 0.7701, + "num_input_tokens_seen": 30397145, + "step": 934 + }, + { + "epoch": 0.08432159444469496, + "flos": 22715038788000.0, + "grad_norm": 2.238093343306863, + "learning_rate": 3.969157292564865e-06, + "loss": 0.8378, + "num_input_tokens_seen": 30424185, + "step": 935 + }, + { + "epoch": 0.08441177796816522, + "flos": 33059896245600.0, + "grad_norm": 1.762896564861614, + "learning_rate": 3.96905500527047e-06, + "loss": 0.7911, + "num_input_tokens_seen": 30453595, + "step": 936 + }, + { + "epoch": 0.08450196149163548, + "flos": 25229150935680.0, + "grad_norm": 1.9748865191535807, + "learning_rate": 3.968952549965445e-06, + "loss": 0.7897, + "num_input_tokens_seen": 30483890, + "step": 937 + }, + { + "epoch": 0.08459214501510574, + "flos": 20566102528800.0, + "grad_norm": 1.7107471377582792, + "learning_rate": 3.968849926658532e-06, + "loss": 0.8352, + "num_input_tokens_seen": 30511970, + "step": 938 + }, + { + "epoch": 0.084682328538576, + "flos": 23185942674720.0, + "grad_norm": 2.4441740170600346, + "learning_rate": 3.9687471353584866e-06, + "loss": 0.7179, + "num_input_tokens_seen": 30541910, + "step": 939 + }, + { + "epoch": 0.08477251206204627, + "flos": 21549806355360.0, + "grad_norm": 2.420708652334288, + "learning_rate": 3.9686441760740795e-06, + "loss": 0.8203, + "num_input_tokens_seen": 30571305, + "step": 940 + }, + { + "epoch": 0.08486269558551653, + "flos": 31676219901600.0, + "grad_norm": 2.1435737070269534, + "learning_rate": 3.968541048814098e-06, + "loss": 0.7459, + "num_input_tokens_seen": 30604665, + "step": 941 + }, + { + "epoch": 0.08495287910898679, + "flos": 22933148171520.0, + "grad_norm": 2.7751805805183354, + "learning_rate": 3.968437753587339e-06, + "loss": 0.7328, + "num_input_tokens_seen": 30630315, + "step": 942 + }, + { + "epoch": 0.08504306263245705, + "flos": 19653961957440.0, + "grad_norm": 1.590137876254722, + "learning_rate": 3.968334290402616e-06, + "loss": 0.85, + "num_input_tokens_seen": 30658965, + "step": 943 + }, + { + "epoch": 0.08513324615592731, + "flos": 33788203685760.0, + "grad_norm": 1.7279051656396227, + "learning_rate": 3.968230659268759e-06, + "loss": 0.7023, + "num_input_tokens_seen": 30692780, + "step": 944 + }, + { + "epoch": 0.08522342967939757, + "flos": 20237990624640.0, + "grad_norm": 2.0413350854580017, + "learning_rate": 3.968126860194609e-06, + "loss": 0.8769, + "num_input_tokens_seen": 30720260, + "step": 945 + }, + { + "epoch": 0.08531361320286783, + "flos": 27046890200640.0, + "grad_norm": 1.794794819084258, + "learning_rate": 3.968022893189025e-06, + "loss": 0.8591, + "num_input_tokens_seen": 30749690, + "step": 946 + }, + { + "epoch": 0.0854037967263381, + "flos": 20857447806720.0, + "grad_norm": 1.680511201670801, + "learning_rate": 3.967918758260874e-06, + "loss": 0.8396, + "num_input_tokens_seen": 30778415, + "step": 947 + }, + { + "epoch": 0.08549398024980837, + "flos": 17285987070720.0, + "grad_norm": 2.2047760100140548, + "learning_rate": 3.967814455419044e-06, + "loss": 0.8514, + "num_input_tokens_seen": 30804665, + "step": 948 + }, + { + "epoch": 0.08558416377327863, + "flos": 24055741155840.0, + "grad_norm": 2.0376613670058465, + "learning_rate": 3.967709984672434e-06, + "loss": 0.8491, + "num_input_tokens_seen": 30832875, + "step": 949 + }, + { + "epoch": 0.08567434729674889, + "flos": 26722792630560.0, + "grad_norm": 3.0658941339379506, + "learning_rate": 3.967605346029959e-06, + "loss": 0.8368, + "num_input_tokens_seen": 30862650, + "step": 950 + }, + { + "epoch": 0.08576453082021915, + "flos": 21877174864320.0, + "grad_norm": 1.6468727676405426, + "learning_rate": 3.9675005395005466e-06, + "loss": 0.8855, + "num_input_tokens_seen": 30891145, + "step": 951 + }, + { + "epoch": 0.08585471434368941, + "flos": 23115159984480.0, + "grad_norm": 1.9170709750230315, + "learning_rate": 3.967395565093139e-06, + "loss": 0.8285, + "num_input_tokens_seen": 30921200, + "step": 952 + }, + { + "epoch": 0.08594489786715967, + "flos": 22934709301440.0, + "grad_norm": 1.6337768043473195, + "learning_rate": 3.967290422816693e-06, + "loss": 0.8144, + "num_input_tokens_seen": 30949205, + "step": 953 + }, + { + "epoch": 0.08603508139062993, + "flos": 21840371068320.0, + "grad_norm": 2.315642618583341, + "learning_rate": 3.967185112680183e-06, + "loss": 0.7716, + "num_input_tokens_seen": 30976660, + "step": 954 + }, + { + "epoch": 0.08612526491410019, + "flos": 21949518684480.0, + "grad_norm": 2.1240506014786, + "learning_rate": 3.96707963469259e-06, + "loss": 0.8843, + "num_input_tokens_seen": 31002525, + "step": 955 + }, + { + "epoch": 0.08621544843757045, + "flos": 23699114312160.0, + "grad_norm": 1.8161040680260807, + "learning_rate": 3.966973988862917e-06, + "loss": 0.8001, + "num_input_tokens_seen": 31031415, + "step": 956 + }, + { + "epoch": 0.08630563196104071, + "flos": 25079928584640.0, + "grad_norm": 1.5636800506095991, + "learning_rate": 3.966868175200178e-06, + "loss": 0.862, + "num_input_tokens_seen": 31062195, + "step": 957 + }, + { + "epoch": 0.08639581548451097, + "flos": 25809611305920.0, + "grad_norm": 2.143371157077604, + "learning_rate": 3.9667621937134e-06, + "loss": 0.8684, + "num_input_tokens_seen": 31091295, + "step": 958 + }, + { + "epoch": 0.08648599900798125, + "flos": 33935121511680.0, + "grad_norm": 3.6948907784926788, + "learning_rate": 3.966656044411627e-06, + "loss": 0.7202, + "num_input_tokens_seen": 31122600, + "step": 959 + }, + { + "epoch": 0.08657618253145151, + "flos": 19248897182880.0, + "grad_norm": 1.5848735378609748, + "learning_rate": 3.966549727303918e-06, + "loss": 0.7805, + "num_input_tokens_seen": 31149590, + "step": 960 + }, + { + "epoch": 0.08666636605492177, + "flos": 23297989532160.0, + "grad_norm": 2.728778364983836, + "learning_rate": 3.966443242399341e-06, + "loss": 0.8414, + "num_input_tokens_seen": 31178055, + "step": 961 + }, + { + "epoch": 0.08675654957839203, + "flos": 61322575486080.0, + "grad_norm": 0.717374937542389, + "learning_rate": 3.966336589706985e-06, + "loss": 0.5985, + "num_input_tokens_seen": 31271395, + "step": 962 + }, + { + "epoch": 0.08684673310186229, + "flos": 26939043526080.0, + "grad_norm": 2.5027386923261985, + "learning_rate": 3.966229769235948e-06, + "loss": 0.8246, + "num_input_tokens_seen": 31299855, + "step": 963 + }, + { + "epoch": 0.08693691662533255, + "flos": 17177248321920.0, + "grad_norm": 2.0614161099785275, + "learning_rate": 3.966122780995345e-06, + "loss": 0.8887, + "num_input_tokens_seen": 31325450, + "step": 964 + }, + { + "epoch": 0.08702710014880281, + "flos": 19144098428640.0, + "grad_norm": 2.2229535753736744, + "learning_rate": 3.966015624994306e-06, + "loss": 0.8412, + "num_input_tokens_seen": 31352000, + "step": 965 + }, + { + "epoch": 0.08711728367227307, + "flos": 21148124028960.0, + "grad_norm": 2.0767623908491926, + "learning_rate": 3.9659083012419735e-06, + "loss": 0.7837, + "num_input_tokens_seen": 31379710, + "step": 966 + }, + { + "epoch": 0.08720746719574334, + "flos": 42679717201920.0, + "grad_norm": 2.5716774513511296, + "learning_rate": 3.965800809747505e-06, + "loss": 0.7422, + "num_input_tokens_seen": 31410870, + "step": 967 + }, + { + "epoch": 0.0872976507192136, + "flos": 15282890714400.0, + "grad_norm": 1.712343935572707, + "learning_rate": 3.965693150520071e-06, + "loss": 0.779, + "num_input_tokens_seen": 31437990, + "step": 968 + }, + { + "epoch": 0.08738783424268386, + "flos": 14044273708320.0, + "grad_norm": 3.536080895995925, + "learning_rate": 3.96558532356886e-06, + "loss": 0.7475, + "num_input_tokens_seen": 31463975, + "step": 969 + }, + { + "epoch": 0.08747801776615412, + "flos": 17760422084640.0, + "grad_norm": 5.890506166428189, + "learning_rate": 3.9654773289030704e-06, + "loss": 0.8568, + "num_input_tokens_seen": 31488785, + "step": 970 + }, + { + "epoch": 0.08756820128962439, + "flos": 26613124637760.0, + "grad_norm": 2.0117580965508965, + "learning_rate": 3.9653691665319176e-06, + "loss": 0.8244, + "num_input_tokens_seen": 31518530, + "step": 971 + }, + { + "epoch": 0.08765838481309465, + "flos": 39473841221760.0, + "grad_norm": 1.9908009809140945, + "learning_rate": 3.96526083646463e-06, + "loss": 0.739, + "num_input_tokens_seen": 31551320, + "step": 972 + }, + { + "epoch": 0.08774856833656491, + "flos": 34444055796480.0, + "grad_norm": 2.4269113657091097, + "learning_rate": 3.9651523387104526e-06, + "loss": 0.7625, + "num_input_tokens_seen": 31584265, + "step": 973 + }, + { + "epoch": 0.08783875186003517, + "flos": 19910287587840.0, + "grad_norm": 2.2250946024810347, + "learning_rate": 3.965043673278641e-06, + "loss": 0.8237, + "num_input_tokens_seen": 31613170, + "step": 974 + }, + { + "epoch": 0.08792893538350544, + "flos": 26904432745920.0, + "grad_norm": 3.416777554705223, + "learning_rate": 3.964934840178469e-06, + "loss": 0.8166, + "num_input_tokens_seen": 31643465, + "step": 975 + }, + { + "epoch": 0.0880191189069757, + "flos": 21689736266400.0, + "grad_norm": 1.784693618360481, + "learning_rate": 3.964825839419221e-06, + "loss": 0.8606, + "num_input_tokens_seen": 31670050, + "step": 976 + }, + { + "epoch": 0.08810930243044596, + "flos": 21148495726560.0, + "grad_norm": 2.2179746750674867, + "learning_rate": 3.964716671010199e-06, + "loss": 0.7944, + "num_input_tokens_seen": 31696740, + "step": 977 + }, + { + "epoch": 0.08819948595391622, + "flos": 27779583672480.0, + "grad_norm": 1.5910451038232394, + "learning_rate": 3.9646073349607165e-06, + "loss": 0.7759, + "num_input_tokens_seen": 31729405, + "step": 978 + }, + { + "epoch": 0.08828966947738648, + "flos": 16302246074400.0, + "grad_norm": 2.1875080653450647, + "learning_rate": 3.964497831280105e-06, + "loss": 0.9417, + "num_input_tokens_seen": 31756330, + "step": 979 + }, + { + "epoch": 0.08837985300085674, + "flos": 18525421811520.0, + "grad_norm": 3.2764875455258986, + "learning_rate": 3.964388159977705e-06, + "loss": 0.8134, + "num_input_tokens_seen": 31780990, + "step": 980 + }, + { + "epoch": 0.088470036524327, + "flos": 29451631713600.0, + "grad_norm": 1.4916533467783177, + "learning_rate": 3.964278321062876e-06, + "loss": 0.7154, + "num_input_tokens_seen": 31812755, + "step": 981 + }, + { + "epoch": 0.08856022004779726, + "flos": 24604898854560.0, + "grad_norm": 1.7068179627997786, + "learning_rate": 3.96416831454499e-06, + "loss": 0.8303, + "num_input_tokens_seen": 31839880, + "step": 982 + }, + { + "epoch": 0.08865040357126754, + "flos": 68337164012160.0, + "grad_norm": 0.6991876350171508, + "learning_rate": 3.964058140433434e-06, + "loss": 0.6297, + "num_input_tokens_seen": 31932035, + "step": 983 + }, + { + "epoch": 0.0887405870947378, + "flos": 20237433078240.0, + "grad_norm": 1.7839410097094148, + "learning_rate": 3.963947798737606e-06, + "loss": 0.8727, + "num_input_tokens_seen": 31958775, + "step": 984 + }, + { + "epoch": 0.08883077061820806, + "flos": 13969365174720.0, + "grad_norm": 1.8431336617355192, + "learning_rate": 3.963837289466923e-06, + "loss": 0.83, + "num_input_tokens_seen": 31980445, + "step": 985 + }, + { + "epoch": 0.08892095414167832, + "flos": 29345308999200.0, + "grad_norm": 2.0477719270996664, + "learning_rate": 3.9637266126308145e-06, + "loss": 0.784, + "num_input_tokens_seen": 32012105, + "step": 986 + }, + { + "epoch": 0.08901113766514858, + "flos": 61511203516320.0, + "grad_norm": 0.7208950128424861, + "learning_rate": 3.963615768238724e-06, + "loss": 0.6184, + "num_input_tokens_seen": 32104065, + "step": 987 + }, + { + "epoch": 0.08910132118861884, + "flos": 67234728504960.0, + "grad_norm": 1.6911580738423142, + "learning_rate": 3.963504756300107e-06, + "loss": 0.6979, + "num_input_tokens_seen": 32138965, + "step": 988 + }, + { + "epoch": 0.0891915047120891, + "flos": 26645802590400.0, + "grad_norm": 1.6698277140165332, + "learning_rate": 3.96339357682444e-06, + "loss": 0.8497, + "num_input_tokens_seen": 32168365, + "step": 989 + }, + { + "epoch": 0.08928168823555936, + "flos": 33021903017280.0, + "grad_norm": 1.40589226305058, + "learning_rate": 3.963282229821206e-06, + "loss": 0.8272, + "num_input_tokens_seen": 32201585, + "step": 990 + }, + { + "epoch": 0.08937187175902962, + "flos": 18890226002400.0, + "grad_norm": 1.689455735750331, + "learning_rate": 3.963170715299906e-06, + "loss": 0.8234, + "num_input_tokens_seen": 32229700, + "step": 991 + }, + { + "epoch": 0.08946205528249988, + "flos": 26467507753440.0, + "grad_norm": 1.6099747216271951, + "learning_rate": 3.963059033270056e-06, + "loss": 0.7347, + "num_input_tokens_seen": 32259460, + "step": 992 + }, + { + "epoch": 0.08955223880597014, + "flos": 24312884520960.0, + "grad_norm": 1.722437353780498, + "learning_rate": 3.9629471837411855e-06, + "loss": 0.7591, + "num_input_tokens_seen": 32288865, + "step": 993 + }, + { + "epoch": 0.0896424223294404, + "flos": 24791965754880.0, + "grad_norm": 1.6984364780679135, + "learning_rate": 3.962835166722838e-06, + "loss": 0.7558, + "num_input_tokens_seen": 32318975, + "step": 994 + }, + { + "epoch": 0.08973260585291068, + "flos": 22714927278720.0, + "grad_norm": 1.8169685395854778, + "learning_rate": 3.96272298222457e-06, + "loss": 0.8193, + "num_input_tokens_seen": 32346540, + "step": 995 + }, + { + "epoch": 0.08982278937638094, + "flos": 22237109816640.0, + "grad_norm": 1.8239783183741654, + "learning_rate": 3.962610630255956e-06, + "loss": 0.9378, + "num_input_tokens_seen": 32374140, + "step": 996 + }, + { + "epoch": 0.0899129728998512, + "flos": 54662495127360.0, + "grad_norm": 2.070910673277125, + "learning_rate": 3.96249811082658e-06, + "loss": 0.7789, + "num_input_tokens_seen": 32408295, + "step": 997 + }, + { + "epoch": 0.09000315642332146, + "flos": 50555623770240.0, + "grad_norm": 0.8063006850878286, + "learning_rate": 3.962385423946046e-06, + "loss": 0.6734, + "num_input_tokens_seen": 32492565, + "step": 998 + }, + { + "epoch": 0.09009333994679172, + "flos": 29927739366720.0, + "grad_norm": 1.4734049950227228, + "learning_rate": 3.962272569623966e-06, + "loss": 0.7846, + "num_input_tokens_seen": 32522910, + "step": 999 + }, + { + "epoch": 0.09018352347026198, + "flos": 20196800796960.0, + "grad_norm": 1.806524753463259, + "learning_rate": 3.9621595478699704e-06, + "loss": 0.7023, + "num_input_tokens_seen": 32548530, + "step": 1000 + }, + { + "epoch": 0.09027370699373224, + "flos": 15902831103360.0, + "grad_norm": 1.7090777039248328, + "learning_rate": 3.962046358693703e-06, + "loss": 0.8051, + "num_input_tokens_seen": 32576780, + "step": 1001 + }, + { + "epoch": 0.0903638905172025, + "flos": 15464976866880.0, + "grad_norm": 1.8998943751238375, + "learning_rate": 3.961933002104822e-06, + "loss": 0.8911, + "num_input_tokens_seen": 32602345, + "step": 1002 + }, + { + "epoch": 0.09045407404067277, + "flos": 24791333868960.0, + "grad_norm": 2.396981956353546, + "learning_rate": 3.961819478112999e-06, + "loss": 0.7361, + "num_input_tokens_seen": 32630660, + "step": 1003 + }, + { + "epoch": 0.09054425756414303, + "flos": 22203651299040.0, + "grad_norm": 1.9466961455508194, + "learning_rate": 3.961705786727921e-06, + "loss": 0.8332, + "num_input_tokens_seen": 32659095, + "step": 1004 + }, + { + "epoch": 0.09063444108761329, + "flos": 29528324395680.0, + "grad_norm": 2.090451275231894, + "learning_rate": 3.961591927959288e-06, + "loss": 0.734, + "num_input_tokens_seen": 32689065, + "step": 1005 + }, + { + "epoch": 0.09072462461108355, + "flos": 68435904095520.0, + "grad_norm": 0.7148235297339802, + "learning_rate": 3.961477901816816e-06, + "loss": 0.5937, + "num_input_tokens_seen": 32782835, + "step": 1006 + }, + { + "epoch": 0.09081480813455382, + "flos": 16740137480640.0, + "grad_norm": 1.9895372302187853, + "learning_rate": 3.961363708310233e-06, + "loss": 0.874, + "num_input_tokens_seen": 32808945, + "step": 1007 + }, + { + "epoch": 0.09090499165802408, + "flos": 20929940305920.0, + "grad_norm": 2.0379331537875918, + "learning_rate": 3.961249347449286e-06, + "loss": 0.8369, + "num_input_tokens_seen": 32837745, + "step": 1008 + }, + { + "epoch": 0.09099517518149434, + "flos": 20998418471040.0, + "grad_norm": 1.6141587796442796, + "learning_rate": 3.961134819243728e-06, + "loss": 0.7937, + "num_input_tokens_seen": 32867425, + "step": 1009 + }, + { + "epoch": 0.0910853587049646, + "flos": 21768101587680.0, + "grad_norm": 2.221937656726313, + "learning_rate": 3.961020123703335e-06, + "loss": 0.8174, + "num_input_tokens_seen": 32897060, + "step": 1010 + }, + { + "epoch": 0.09117554222843487, + "flos": 21403966452480.0, + "grad_norm": 2.0163321031005523, + "learning_rate": 3.960905260837892e-06, + "loss": 0.8209, + "num_input_tokens_seen": 32925585, + "step": 1011 + }, + { + "epoch": 0.09126572575190513, + "flos": 25483655247840.0, + "grad_norm": 1.8251770900349378, + "learning_rate": 3.960790230657199e-06, + "loss": 0.78, + "num_input_tokens_seen": 32954945, + "step": 1012 + }, + { + "epoch": 0.09135590927537539, + "flos": 40672011795360.0, + "grad_norm": 1.4227139020840514, + "learning_rate": 3.960675033171072e-06, + "loss": 0.7566, + "num_input_tokens_seen": 32987665, + "step": 1013 + }, + { + "epoch": 0.09144609279884565, + "flos": 31381417836000.0, + "grad_norm": 1.5469972966718573, + "learning_rate": 3.960559668389341e-06, + "loss": 0.7535, + "num_input_tokens_seen": 33018775, + "step": 1014 + }, + { + "epoch": 0.09153627632231591, + "flos": 20200926640320.0, + "grad_norm": 1.7757106110754333, + "learning_rate": 3.960444136321847e-06, + "loss": 0.8117, + "num_input_tokens_seen": 33044725, + "step": 1015 + }, + { + "epoch": 0.09162645984578617, + "flos": 22714815769440.0, + "grad_norm": 2.0211031494393916, + "learning_rate": 3.960328436978451e-06, + "loss": 0.8955, + "num_input_tokens_seen": 33072980, + "step": 1016 + }, + { + "epoch": 0.09171664336925643, + "flos": 62461709013600.0, + "grad_norm": 0.655175623863157, + "learning_rate": 3.960212570369024e-06, + "loss": 0.5943, + "num_input_tokens_seen": 33160125, + "step": 1017 + }, + { + "epoch": 0.09180682689272669, + "flos": 21731446470720.0, + "grad_norm": 1.7464560318508893, + "learning_rate": 3.9600965365034515e-06, + "loss": 0.7479, + "num_input_tokens_seen": 33189685, + "step": 1018 + }, + { + "epoch": 0.09189701041619697, + "flos": 20673614675520.0, + "grad_norm": 1.9332047504657934, + "learning_rate": 3.959980335391634e-06, + "loss": 0.8605, + "num_input_tokens_seen": 33218220, + "step": 1019 + }, + { + "epoch": 0.09198719393966723, + "flos": 20561753666880.0, + "grad_norm": 1.7399904912929953, + "learning_rate": 3.959863967043487e-06, + "loss": 0.7481, + "num_input_tokens_seen": 33246165, + "step": 1020 + }, + { + "epoch": 0.09207737746313749, + "flos": 22163093357280.0, + "grad_norm": 1.8166287849862988, + "learning_rate": 3.9597474314689405e-06, + "loss": 0.8371, + "num_input_tokens_seen": 33274855, + "step": 1021 + }, + { + "epoch": 0.09216756098660775, + "flos": 20925554274240.0, + "grad_norm": 1.9180326431726544, + "learning_rate": 3.959630728677937e-06, + "loss": 0.8284, + "num_input_tokens_seen": 33302430, + "step": 1022 + }, + { + "epoch": 0.09225774451007801, + "flos": 19581246439680.0, + "grad_norm": 1.6507557843072451, + "learning_rate": 3.959513858680434e-06, + "loss": 0.7857, + "num_input_tokens_seen": 33329455, + "step": 1023 + }, + { + "epoch": 0.09234792803354827, + "flos": 18846731649600.0, + "grad_norm": 1.9879376727260114, + "learning_rate": 3.959396821486405e-06, + "loss": 0.8988, + "num_input_tokens_seen": 33353565, + "step": 1024 + }, + { + "epoch": 0.09243811155701853, + "flos": 66527830846560.0, + "grad_norm": 0.8170644124038668, + "learning_rate": 3.959279617105835e-06, + "loss": 0.6243, + "num_input_tokens_seen": 33429600, + "step": 1025 + }, + { + "epoch": 0.09252829508048879, + "flos": 32369321845440.0, + "grad_norm": 1.9051549298821469, + "learning_rate": 3.9591622455487235e-06, + "loss": 0.8174, + "num_input_tokens_seen": 33458965, + "step": 1026 + }, + { + "epoch": 0.09261847860395905, + "flos": 18120877413600.0, + "grad_norm": 1.7494604777679588, + "learning_rate": 3.959044706825087e-06, + "loss": 0.8826, + "num_input_tokens_seen": 33485085, + "step": 1027 + }, + { + "epoch": 0.09270866212742931, + "flos": 29414604899040.0, + "grad_norm": 3.2836719112455683, + "learning_rate": 3.958927000944954e-06, + "loss": 0.7158, + "num_input_tokens_seen": 33517580, + "step": 1028 + }, + { + "epoch": 0.09279884565089958, + "flos": 27014509606080.0, + "grad_norm": 2.034276497561301, + "learning_rate": 3.958809127918368e-06, + "loss": 0.8949, + "num_input_tokens_seen": 33547820, + "step": 1029 + }, + { + "epoch": 0.09288902917436984, + "flos": 21149313461280.0, + "grad_norm": 1.5828315876810577, + "learning_rate": 3.958691087755387e-06, + "loss": 0.7456, + "num_input_tokens_seen": 33577265, + "step": 1030 + }, + { + "epoch": 0.09297921269784011, + "flos": 27415225518720.0, + "grad_norm": 1.9416649260397163, + "learning_rate": 3.958572880466081e-06, + "loss": 0.6838, + "num_input_tokens_seen": 33606580, + "step": 1031 + }, + { + "epoch": 0.09306939622131037, + "flos": 23334272951520.0, + "grad_norm": 1.6738357454747403, + "learning_rate": 3.9584545060605385e-06, + "loss": 0.7942, + "num_input_tokens_seen": 33636595, + "step": 1032 + }, + { + "epoch": 0.09315957974478063, + "flos": 22204915070880.0, + "grad_norm": 1.6373621147284203, + "learning_rate": 3.958335964548859e-06, + "loss": 0.774, + "num_input_tokens_seen": 33663610, + "step": 1033 + }, + { + "epoch": 0.0932497632682509, + "flos": 20234087799840.0, + "grad_norm": 1.9228314136633204, + "learning_rate": 3.958217255941156e-06, + "loss": 0.858, + "num_input_tokens_seen": 33691620, + "step": 1034 + }, + { + "epoch": 0.09333994679172115, + "flos": 37179325248000.0, + "grad_norm": 1.7053371376912798, + "learning_rate": 3.95809838024756e-06, + "loss": 0.7595, + "num_input_tokens_seen": 33725895, + "step": 1035 + }, + { + "epoch": 0.09343013031519141, + "flos": 21731855338080.0, + "grad_norm": 1.7527225977733851, + "learning_rate": 3.957979337478212e-06, + "loss": 0.7755, + "num_input_tokens_seen": 33753810, + "step": 1036 + }, + { + "epoch": 0.09352031383866168, + "flos": 16849508115360.0, + "grad_norm": 3.0437325729717193, + "learning_rate": 3.957860127643272e-06, + "loss": 0.827, + "num_input_tokens_seen": 33781825, + "step": 1037 + }, + { + "epoch": 0.09361049736213194, + "flos": 21293815252800.0, + "grad_norm": 2.0054345205645516, + "learning_rate": 3.95774075075291e-06, + "loss": 0.8127, + "num_input_tokens_seen": 33810785, + "step": 1038 + }, + { + "epoch": 0.0937006808856022, + "flos": 20820681180480.0, + "grad_norm": 1.9765890971569264, + "learning_rate": 3.957621206817312e-06, + "loss": 0.8661, + "num_input_tokens_seen": 33837630, + "step": 1039 + }, + { + "epoch": 0.09379086440907246, + "flos": 23773205111040.0, + "grad_norm": 2.724527381018707, + "learning_rate": 3.957501495846679e-06, + "loss": 0.7922, + "num_input_tokens_seen": 33867465, + "step": 1040 + }, + { + "epoch": 0.09388104793254272, + "flos": 68154631822560.0, + "grad_norm": 0.6870542077954159, + "learning_rate": 3.957381617851225e-06, + "loss": 0.5991, + "num_input_tokens_seen": 33951650, + "step": 1041 + }, + { + "epoch": 0.09397123145601298, + "flos": 42169742163840.0, + "grad_norm": 1.5771722467001212, + "learning_rate": 3.9572615728411776e-06, + "loss": 0.7318, + "num_input_tokens_seen": 33984345, + "step": 1042 + }, + { + "epoch": 0.09406141497948325, + "flos": 21840817105440.0, + "grad_norm": 7.576439738916402, + "learning_rate": 3.957141360826781e-06, + "loss": 0.6945, + "num_input_tokens_seen": 34009450, + "step": 1043 + }, + { + "epoch": 0.09415159850295352, + "flos": 21767655550560.0, + "grad_norm": 3.163328636683149, + "learning_rate": 3.957020981818292e-06, + "loss": 0.7916, + "num_input_tokens_seen": 34036540, + "step": 1044 + }, + { + "epoch": 0.09424178202642378, + "flos": 29782233991680.0, + "grad_norm": 2.6409308803747575, + "learning_rate": 3.956900435825982e-06, + "loss": 0.7993, + "num_input_tokens_seen": 34069285, + "step": 1045 + }, + { + "epoch": 0.09433196554989404, + "flos": 23114862626400.0, + "grad_norm": 1.5165866278743896, + "learning_rate": 3.9567797228601364e-06, + "loss": 0.7599, + "num_input_tokens_seen": 34099545, + "step": 1046 + }, + { + "epoch": 0.0944221490733643, + "flos": 23371336935840.0, + "grad_norm": 1.7179088268090346, + "learning_rate": 3.956658842931055e-06, + "loss": 0.8816, + "num_input_tokens_seen": 34128905, + "step": 1047 + }, + { + "epoch": 0.09451233259683456, + "flos": 23443197549120.0, + "grad_norm": 1.5434252099990053, + "learning_rate": 3.956537796049052e-06, + "loss": 0.8235, + "num_input_tokens_seen": 34158920, + "step": 1048 + }, + { + "epoch": 0.09460251612030482, + "flos": 68588434555200.0, + "grad_norm": 0.7183814140287046, + "learning_rate": 3.956416582224457e-06, + "loss": 0.6188, + "num_input_tokens_seen": 34259480, + "step": 1049 + }, + { + "epoch": 0.09469269964377508, + "flos": 20529447411840.0, + "grad_norm": 2.2048335383202198, + "learning_rate": 3.956295201467611e-06, + "loss": 0.8454, + "num_input_tokens_seen": 34285505, + "step": 1050 + }, + { + "epoch": 0.09478288316724534, + "flos": 21766205929920.0, + "grad_norm": 1.5770358113327136, + "learning_rate": 3.956173653788872e-06, + "loss": 0.8386, + "num_input_tokens_seen": 34315150, + "step": 1051 + }, + { + "epoch": 0.0948730666907156, + "flos": 29381443739520.0, + "grad_norm": 2.2559159176828887, + "learning_rate": 3.95605193919861e-06, + "loss": 0.8453, + "num_input_tokens_seen": 34341465, + "step": 1052 + }, + { + "epoch": 0.09496325021418586, + "flos": 26832274774560.0, + "grad_norm": 1.7050968619209628, + "learning_rate": 3.955930057707211e-06, + "loss": 0.7962, + "num_input_tokens_seen": 34369440, + "step": 1053 + }, + { + "epoch": 0.09505343373765612, + "flos": 22673625941760.0, + "grad_norm": 2.0405819664896803, + "learning_rate": 3.955808009325075e-06, + "loss": 0.8209, + "num_input_tokens_seen": 34398550, + "step": 1054 + }, + { + "epoch": 0.0951436172611264, + "flos": 39000595640160.0, + "grad_norm": 2.499743448445147, + "learning_rate": 3.955685794062615e-06, + "loss": 0.7806, + "num_input_tokens_seen": 34429920, + "step": 1055 + }, + { + "epoch": 0.09523380078459666, + "flos": 21585346379520.0, + "grad_norm": 2.210864062488129, + "learning_rate": 3.95556341193026e-06, + "loss": 0.9079, + "num_input_tokens_seen": 34458980, + "step": 1056 + }, + { + "epoch": 0.09532398430806692, + "flos": 26502007024320.0, + "grad_norm": 1.8692834175758364, + "learning_rate": 3.955440862938452e-06, + "loss": 0.9041, + "num_input_tokens_seen": 34487540, + "step": 1057 + }, + { + "epoch": 0.09541416783153718, + "flos": 22605556644000.0, + "grad_norm": 2.0522663660929203, + "learning_rate": 3.955318147097647e-06, + "loss": 0.8077, + "num_input_tokens_seen": 34512720, + "step": 1058 + }, + { + "epoch": 0.09550435135500744, + "flos": 20193195330240.0, + "grad_norm": 1.888577448033752, + "learning_rate": 3.955195264418316e-06, + "loss": 0.7729, + "num_input_tokens_seen": 34538510, + "step": 1059 + }, + { + "epoch": 0.0955945348784777, + "flos": 24353962839360.0, + "grad_norm": 1.5654016636215888, + "learning_rate": 3.955072214910944e-06, + "loss": 0.6684, + "num_input_tokens_seen": 34567395, + "step": 1060 + }, + { + "epoch": 0.09568471840194796, + "flos": 25229150935680.0, + "grad_norm": 2.451756478312839, + "learning_rate": 3.954948998586032e-06, + "loss": 0.7321, + "num_input_tokens_seen": 34595090, + "step": 1061 + }, + { + "epoch": 0.09577490192541822, + "flos": 32040838243680.0, + "grad_norm": 1.895626671674314, + "learning_rate": 3.954825615454089e-06, + "loss": 0.8159, + "num_input_tokens_seen": 34626550, + "step": 1062 + }, + { + "epoch": 0.09586508544888848, + "flos": 21432630071040.0, + "grad_norm": 1.7714925847814862, + "learning_rate": 3.954702065525649e-06, + "loss": 0.8038, + "num_input_tokens_seen": 34653420, + "step": 1063 + }, + { + "epoch": 0.09595526897235875, + "flos": 18489249901440.0, + "grad_norm": 2.061899849243839, + "learning_rate": 3.954578348811248e-06, + "loss": 0.835, + "num_input_tokens_seen": 34679070, + "step": 1064 + }, + { + "epoch": 0.096045452495829, + "flos": 32879519902080.0, + "grad_norm": 1.8257104300857587, + "learning_rate": 3.954454465321447e-06, + "loss": 0.7834, + "num_input_tokens_seen": 34707800, + "step": 1065 + }, + { + "epoch": 0.09613563601929928, + "flos": 15719667027840.0, + "grad_norm": 2.660832659391451, + "learning_rate": 3.954330415066813e-06, + "loss": 0.8906, + "num_input_tokens_seen": 34733180, + "step": 1066 + }, + { + "epoch": 0.09622581954276954, + "flos": 20196317590080.0, + "grad_norm": 2.322643509158473, + "learning_rate": 3.954206198057932e-06, + "loss": 0.7945, + "num_input_tokens_seen": 34762870, + "step": 1067 + }, + { + "epoch": 0.0963160030662398, + "flos": 28689791416320.0, + "grad_norm": 1.659487457200406, + "learning_rate": 3.954081814305403e-06, + "loss": 0.8664, + "num_input_tokens_seen": 34792750, + "step": 1068 + }, + { + "epoch": 0.09640618658971006, + "flos": 23768001344640.0, + "grad_norm": 1.7116777046885268, + "learning_rate": 3.953957263819839e-06, + "loss": 0.7465, + "num_input_tokens_seen": 34823260, + "step": 1069 + }, + { + "epoch": 0.09649637011318032, + "flos": 28908755704320.0, + "grad_norm": 1.6746594780344466, + "learning_rate": 3.953832546611867e-06, + "loss": 0.7436, + "num_input_tokens_seen": 34853175, + "step": 1070 + }, + { + "epoch": 0.09658655363665059, + "flos": 23808447777120.0, + "grad_norm": 2.3535542211889693, + "learning_rate": 3.953707662692129e-06, + "loss": 0.8358, + "num_input_tokens_seen": 34883135, + "step": 1071 + }, + { + "epoch": 0.09667673716012085, + "flos": 25630089866880.0, + "grad_norm": 2.385418176997566, + "learning_rate": 3.95358261207128e-06, + "loss": 0.739, + "num_input_tokens_seen": 34913695, + "step": 1072 + }, + { + "epoch": 0.0967669206835911, + "flos": 21727878173760.0, + "grad_norm": 1.647665412483044, + "learning_rate": 3.953457394759992e-06, + "loss": 0.8008, + "num_input_tokens_seen": 34943570, + "step": 1073 + }, + { + "epoch": 0.09685710420706137, + "flos": 24900778843200.0, + "grad_norm": 1.6714307961431656, + "learning_rate": 3.953332010768947e-06, + "loss": 0.8227, + "num_input_tokens_seen": 34970055, + "step": 1074 + }, + { + "epoch": 0.09694728773053163, + "flos": 25229002256640.0, + "grad_norm": 2.6235432897143838, + "learning_rate": 3.9532064601088436e-06, + "loss": 0.7732, + "num_input_tokens_seen": 34995740, + "step": 1075 + }, + { + "epoch": 0.09703747125400189, + "flos": 18196938209760.0, + "grad_norm": 3.7796347454752954, + "learning_rate": 3.953080742790396e-06, + "loss": 0.9563, + "num_input_tokens_seen": 35020250, + "step": 1076 + }, + { + "epoch": 0.09712765477747215, + "flos": 66301835740320.0, + "grad_norm": 0.7631252508139608, + "learning_rate": 3.95295485882433e-06, + "loss": 0.6606, + "num_input_tokens_seen": 35115530, + "step": 1077 + }, + { + "epoch": 0.09721783830094242, + "flos": 70827450343680.0, + "grad_norm": 0.6775273603284836, + "learning_rate": 3.952828808221387e-06, + "loss": 0.6279, + "num_input_tokens_seen": 35209785, + "step": 1078 + }, + { + "epoch": 0.09730802182441269, + "flos": 33389160412320.0, + "grad_norm": 4.907389125883224, + "learning_rate": 3.9527025909923225e-06, + "loss": 0.7874, + "num_input_tokens_seen": 35240030, + "step": 1079 + }, + { + "epoch": 0.09739820534788295, + "flos": 23916926337600.0, + "grad_norm": 1.6099551738398439, + "learning_rate": 3.952576207147906e-06, + "loss": 0.9007, + "num_input_tokens_seen": 35267820, + "step": 1080 + }, + { + "epoch": 0.09748838887135321, + "flos": 26103521297280.0, + "grad_norm": 1.8801847616674783, + "learning_rate": 3.95244965669892e-06, + "loss": 0.8467, + "num_input_tokens_seen": 35299050, + "step": 1081 + }, + { + "epoch": 0.09757857239482347, + "flos": 40822051881120.0, + "grad_norm": 2.763490877189231, + "learning_rate": 3.952322939656165e-06, + "loss": 0.7637, + "num_input_tokens_seen": 35331340, + "step": 1082 + }, + { + "epoch": 0.09766875591829373, + "flos": 20781461350080.0, + "grad_norm": 1.8719412331812153, + "learning_rate": 3.952196056030451e-06, + "loss": 0.8048, + "num_input_tokens_seen": 35361095, + "step": 1083 + }, + { + "epoch": 0.09775893944176399, + "flos": 22607117773920.0, + "grad_norm": 2.488034030963684, + "learning_rate": 3.952069005832605e-06, + "loss": 0.7749, + "num_input_tokens_seen": 35391975, + "step": 1084 + }, + { + "epoch": 0.09784912296523425, + "flos": 24901336389600.0, + "grad_norm": 1.9333283829899692, + "learning_rate": 3.951941789073468e-06, + "loss": 0.8319, + "num_input_tokens_seen": 35421600, + "step": 1085 + }, + { + "epoch": 0.09793930648870451, + "flos": 22168259953920.0, + "grad_norm": 1.7913136541073307, + "learning_rate": 3.9518144057638955e-06, + "loss": 0.8338, + "num_input_tokens_seen": 35449130, + "step": 1086 + }, + { + "epoch": 0.09802949001217477, + "flos": 22095618775680.0, + "grad_norm": 1.8324394503344021, + "learning_rate": 3.951686855914755e-06, + "loss": 0.8014, + "num_input_tokens_seen": 35476465, + "step": 1087 + }, + { + "epoch": 0.09811967353564503, + "flos": 30803224821120.0, + "grad_norm": 2.25980381780873, + "learning_rate": 3.9515591395369305e-06, + "loss": 0.8328, + "num_input_tokens_seen": 35505745, + "step": 1088 + }, + { + "epoch": 0.0982098570591153, + "flos": 24900592994400.0, + "grad_norm": 1.5219385842703284, + "learning_rate": 3.95143125664132e-06, + "loss": 0.7939, + "num_input_tokens_seen": 35535255, + "step": 1089 + }, + { + "epoch": 0.09830004058258557, + "flos": 26430964145760.0, + "grad_norm": 1.975275638515803, + "learning_rate": 3.951303207238833e-06, + "loss": 0.8326, + "num_input_tokens_seen": 35563650, + "step": 1090 + }, + { + "epoch": 0.09839022410605583, + "flos": 21838735598880.0, + "grad_norm": 1.939849843673568, + "learning_rate": 3.951174991340399e-06, + "loss": 0.7815, + "num_input_tokens_seen": 35592225, + "step": 1091 + }, + { + "epoch": 0.09848040762952609, + "flos": 24609545074560.0, + "grad_norm": 1.8081227424573205, + "learning_rate": 3.9510466089569546e-06, + "loss": 0.7683, + "num_input_tokens_seen": 35623955, + "step": 1092 + }, + { + "epoch": 0.09857059115299635, + "flos": 21471478203840.0, + "grad_norm": 1.618556967754744, + "learning_rate": 3.950918060099456e-06, + "loss": 0.7713, + "num_input_tokens_seen": 35652505, + "step": 1093 + }, + { + "epoch": 0.09866077467646661, + "flos": 18197755944480.0, + "grad_norm": 1.6865670486227704, + "learning_rate": 3.950789344778871e-06, + "loss": 0.8575, + "num_input_tokens_seen": 35680665, + "step": 1094 + }, + { + "epoch": 0.09875095819993687, + "flos": 24347904168480.0, + "grad_norm": 1.6289958974928274, + "learning_rate": 3.950660463006184e-06, + "loss": 0.7485, + "num_input_tokens_seen": 35710405, + "step": 1095 + }, + { + "epoch": 0.09884114172340713, + "flos": 29234600253120.0, + "grad_norm": 1.4939117056814128, + "learning_rate": 3.950531414792389e-06, + "loss": 0.8741, + "num_input_tokens_seen": 35739640, + "step": 1096 + }, + { + "epoch": 0.0989313252468774, + "flos": 22751619565440.0, + "grad_norm": 1.639008661336863, + "learning_rate": 3.950402200148498e-06, + "loss": 0.8122, + "num_input_tokens_seen": 35769110, + "step": 1097 + }, + { + "epoch": 0.09902150877034765, + "flos": 25624514402880.0, + "grad_norm": 4.600139715311984, + "learning_rate": 3.950272819085538e-06, + "loss": 0.7833, + "num_input_tokens_seen": 35798030, + "step": 1098 + }, + { + "epoch": 0.09911169229381792, + "flos": 15355420383360.0, + "grad_norm": 3.7405414823738905, + "learning_rate": 3.9501432716145474e-06, + "loss": 0.8923, + "num_input_tokens_seen": 35825300, + "step": 1099 + }, + { + "epoch": 0.09920187581728818, + "flos": 20091816193920.0, + "grad_norm": 1.8976657381915947, + "learning_rate": 3.950013557746579e-06, + "loss": 0.7023, + "num_input_tokens_seen": 35853555, + "step": 1100 + }, + { + "epoch": 0.09929205934075844, + "flos": 22824483762240.0, + "grad_norm": 1.6314927019241872, + "learning_rate": 3.949883677492703e-06, + "loss": 0.8425, + "num_input_tokens_seen": 35883485, + "step": 1101 + }, + { + "epoch": 0.09938224286422871, + "flos": 22933928736480.0, + "grad_norm": 2.4406216512698595, + "learning_rate": 3.9497536308639994e-06, + "loss": 0.825, + "num_input_tokens_seen": 35912730, + "step": 1102 + }, + { + "epoch": 0.09947242638769897, + "flos": 23294607084000.0, + "grad_norm": 2.353285351643218, + "learning_rate": 3.949623417871565e-06, + "loss": 0.8566, + "num_input_tokens_seen": 35941465, + "step": 1103 + }, + { + "epoch": 0.09956260991116923, + "flos": 21618210180960.0, + "grad_norm": 1.6100532806062635, + "learning_rate": 3.949493038526511e-06, + "loss": 0.7874, + "num_input_tokens_seen": 35969995, + "step": 1104 + }, + { + "epoch": 0.0996527934346395, + "flos": 21622521873120.0, + "grad_norm": 1.7310832635345148, + "learning_rate": 3.949362492839961e-06, + "loss": 0.7583, + "num_input_tokens_seen": 35997810, + "step": 1105 + }, + { + "epoch": 0.09974297695810976, + "flos": 24280578265920.0, + "grad_norm": 1.824915554757995, + "learning_rate": 3.949231780823054e-06, + "loss": 0.7663, + "num_input_tokens_seen": 36027465, + "step": 1106 + }, + { + "epoch": 0.09983316048158002, + "flos": 23698705444800.0, + "grad_norm": 1.8351556289951123, + "learning_rate": 3.949100902486945e-06, + "loss": 0.753, + "num_input_tokens_seen": 36055650, + "step": 1107 + }, + { + "epoch": 0.09992334400505028, + "flos": 23917595393280.0, + "grad_norm": 1.8488668141527176, + "learning_rate": 3.948969857842799e-06, + "loss": 0.7744, + "num_input_tokens_seen": 36084280, + "step": 1108 + }, + { + "epoch": 0.10001352752852054, + "flos": 25483989775680.0, + "grad_norm": 1.6190882767519656, + "learning_rate": 3.948838646901798e-06, + "loss": 0.7951, + "num_input_tokens_seen": 36113175, + "step": 1109 + }, + { + "epoch": 0.1001037110519908, + "flos": 21905764143360.0, + "grad_norm": 2.057720657882815, + "learning_rate": 3.948707269675138e-06, + "loss": 0.7935, + "num_input_tokens_seen": 36140155, + "step": 1110 + }, + { + "epoch": 0.10019389457546106, + "flos": 23444349811680.0, + "grad_norm": 1.8959918182977162, + "learning_rate": 3.948575726174028e-06, + "loss": 0.7933, + "num_input_tokens_seen": 36168130, + "step": 1111 + }, + { + "epoch": 0.10028407809893132, + "flos": 18815503317600.0, + "grad_norm": 2.0280699265442315, + "learning_rate": 3.9484440164096935e-06, + "loss": 0.8811, + "num_input_tokens_seen": 36194215, + "step": 1112 + }, + { + "epoch": 0.10037426162240158, + "flos": 20201186828640.0, + "grad_norm": 2.0862207113304856, + "learning_rate": 3.948312140393372e-06, + "loss": 0.827, + "num_input_tokens_seen": 36221370, + "step": 1113 + }, + { + "epoch": 0.10046444514587186, + "flos": 16630915524960.0, + "grad_norm": 2.154159527686247, + "learning_rate": 3.948180098136316e-06, + "loss": 0.7762, + "num_input_tokens_seen": 36245940, + "step": 1114 + }, + { + "epoch": 0.10055462866934212, + "flos": 28617744954240.0, + "grad_norm": 8.60109726758993, + "learning_rate": 3.948047889649791e-06, + "loss": 0.7735, + "num_input_tokens_seen": 36276015, + "step": 1115 + }, + { + "epoch": 0.10064481219281238, + "flos": 22532469428640.0, + "grad_norm": 1.9969348665674767, + "learning_rate": 3.947915514945079e-06, + "loss": 0.7875, + "num_input_tokens_seen": 36304965, + "step": 1116 + }, + { + "epoch": 0.10073499571628264, + "flos": 24206487467040.0, + "grad_norm": 1.677080268936043, + "learning_rate": 3.947782974033474e-06, + "loss": 0.8682, + "num_input_tokens_seen": 36335460, + "step": 1117 + }, + { + "epoch": 0.1008251792397529, + "flos": 23735174712960.0, + "grad_norm": 2.0151753478510672, + "learning_rate": 3.9476502669262866e-06, + "loss": 0.7666, + "num_input_tokens_seen": 36364535, + "step": 1118 + }, + { + "epoch": 0.10091536276322316, + "flos": 17833249111680.0, + "grad_norm": 1.706936342582102, + "learning_rate": 3.947517393634839e-06, + "loss": 0.7779, + "num_input_tokens_seen": 36391570, + "step": 1119 + }, + { + "epoch": 0.10100554628669342, + "flos": 25921249296000.0, + "grad_norm": 1.7928353719198216, + "learning_rate": 3.947384354170469e-06, + "loss": 0.8424, + "num_input_tokens_seen": 36420985, + "step": 1120 + }, + { + "epoch": 0.10109572981016368, + "flos": 20705214705120.0, + "grad_norm": 2.059699729590267, + "learning_rate": 3.947251148544528e-06, + "loss": 0.844, + "num_input_tokens_seen": 36450750, + "step": 1121 + }, + { + "epoch": 0.10118591333363394, + "flos": 15937627732320.0, + "grad_norm": 2.294352489206173, + "learning_rate": 3.947117776768382e-06, + "loss": 0.8327, + "num_input_tokens_seen": 36474995, + "step": 1122 + }, + { + "epoch": 0.1012760968571042, + "flos": 23737144710240.0, + "grad_norm": 1.4728865304298246, + "learning_rate": 3.9469842388534105e-06, + "loss": 0.7869, + "num_input_tokens_seen": 36508735, + "step": 1123 + }, + { + "epoch": 0.10136628038057446, + "flos": 25447334658720.0, + "grad_norm": 2.548484808355822, + "learning_rate": 3.946850534811009e-06, + "loss": 0.8177, + "num_input_tokens_seen": 36537690, + "step": 1124 + }, + { + "epoch": 0.10145646390404472, + "flos": 29856027432480.0, + "grad_norm": 1.779014456117165, + "learning_rate": 3.946716664652585e-06, + "loss": 0.7919, + "num_input_tokens_seen": 36568675, + "step": 1125 + }, + { + "epoch": 0.101546647427515, + "flos": 14407814127360.0, + "grad_norm": 2.905221244279074, + "learning_rate": 3.94658262838956e-06, + "loss": 0.7161, + "num_input_tokens_seen": 36593260, + "step": 1126 + }, + { + "epoch": 0.10163683095098526, + "flos": 25448226732960.0, + "grad_norm": 1.728715684595681, + "learning_rate": 3.946448426033373e-06, + "loss": 0.7251, + "num_input_tokens_seen": 36622405, + "step": 1127 + }, + { + "epoch": 0.10172701447445552, + "flos": 25040671584480.0, + "grad_norm": 1.7992148382332367, + "learning_rate": 3.946314057595473e-06, + "loss": 0.843, + "num_input_tokens_seen": 36649955, + "step": 1128 + }, + { + "epoch": 0.10181719799792578, + "flos": 30001607147040.0, + "grad_norm": 1.9435906200750208, + "learning_rate": 3.946179523087326e-06, + "loss": 0.8669, + "num_input_tokens_seen": 36679080, + "step": 1129 + }, + { + "epoch": 0.10190738152139604, + "flos": 28140596547840.0, + "grad_norm": 1.5875316556642913, + "learning_rate": 3.9460448225204104e-06, + "loss": 0.72, + "num_input_tokens_seen": 36709150, + "step": 1130 + }, + { + "epoch": 0.1019975650448663, + "flos": 22023497974080.0, + "grad_norm": 2.022088031659844, + "learning_rate": 3.945909955906221e-06, + "loss": 0.7962, + "num_input_tokens_seen": 36738045, + "step": 1131 + }, + { + "epoch": 0.10208774856833656, + "flos": 29782977386880.0, + "grad_norm": 2.040465358313883, + "learning_rate": 3.945774923256264e-06, + "loss": 0.8017, + "num_input_tokens_seen": 36767910, + "step": 1132 + }, + { + "epoch": 0.10217793209180683, + "flos": 15606021870720.0, + "grad_norm": 2.9397100697499834, + "learning_rate": 3.945639724582062e-06, + "loss": 0.8736, + "num_input_tokens_seen": 36792320, + "step": 1133 + }, + { + "epoch": 0.10226811561527709, + "flos": 26249286860640.0, + "grad_norm": 2.070798651278821, + "learning_rate": 3.94550435989515e-06, + "loss": 0.7708, + "num_input_tokens_seen": 36819935, + "step": 1134 + }, + { + "epoch": 0.10235829913874735, + "flos": 25739572010880.0, + "grad_norm": 1.9691006181735806, + "learning_rate": 3.945368829207079e-06, + "loss": 0.8236, + "num_input_tokens_seen": 36851120, + "step": 1135 + }, + { + "epoch": 0.10244848266221761, + "flos": 24751221964320.0, + "grad_norm": 2.6591007866783425, + "learning_rate": 3.945233132529414e-06, + "loss": 0.7611, + "num_input_tokens_seen": 36879395, + "step": 1136 + }, + { + "epoch": 0.10253866618568787, + "flos": 25260081909600.0, + "grad_norm": 1.704314839917385, + "learning_rate": 3.9450972698737304e-06, + "loss": 0.8017, + "num_input_tokens_seen": 36907670, + "step": 1137 + }, + { + "epoch": 0.10262884970915814, + "flos": 26249435539680.0, + "grad_norm": 1.975344195733967, + "learning_rate": 3.944961241251623e-06, + "loss": 0.7876, + "num_input_tokens_seen": 36936540, + "step": 1138 + }, + { + "epoch": 0.1027190332326284, + "flos": 30001123940160.0, + "grad_norm": 1.4546043495154104, + "learning_rate": 3.9448250466746985e-06, + "loss": 0.771, + "num_input_tokens_seen": 36968260, + "step": 1139 + }, + { + "epoch": 0.10280921675609866, + "flos": 27852410699520.0, + "grad_norm": 2.335733666345827, + "learning_rate": 3.944688686154578e-06, + "loss": 0.8327, + "num_input_tokens_seen": 36998375, + "step": 1140 + }, + { + "epoch": 0.10289940027956893, + "flos": 27043693601280.0, + "grad_norm": 1.7488490134315184, + "learning_rate": 3.944552159702894e-06, + "loss": 0.763, + "num_input_tokens_seen": 37025685, + "step": 1141 + }, + { + "epoch": 0.10298958380303919, + "flos": 23115717530880.0, + "grad_norm": 1.668552769924751, + "learning_rate": 3.944415467331299e-06, + "loss": 0.793, + "num_input_tokens_seen": 37053080, + "step": 1142 + }, + { + "epoch": 0.10307976732650945, + "flos": 23630004261120.0, + "grad_norm": 1.3309338207601553, + "learning_rate": 3.944278609051455e-06, + "loss": 0.7575, + "num_input_tokens_seen": 37085560, + "step": 1143 + }, + { + "epoch": 0.10316995084997971, + "flos": 25227069429120.0, + "grad_norm": 1.6562959916548872, + "learning_rate": 3.944141584875039e-06, + "loss": 0.7426, + "num_input_tokens_seen": 37115570, + "step": 1144 + }, + { + "epoch": 0.10326013437344997, + "flos": 21112769853600.0, + "grad_norm": 1.5911836343906574, + "learning_rate": 3.944004394813743e-06, + "loss": 0.8059, + "num_input_tokens_seen": 37143530, + "step": 1145 + }, + { + "epoch": 0.10335031789692023, + "flos": 17687483548320.0, + "grad_norm": 1.7037777830840675, + "learning_rate": 3.943867038879273e-06, + "loss": 0.8281, + "num_input_tokens_seen": 37171695, + "step": 1146 + }, + { + "epoch": 0.10344050142039049, + "flos": 41658763542240.0, + "grad_norm": 1.749582308921626, + "learning_rate": 3.943729517083349e-06, + "loss": 0.8386, + "num_input_tokens_seen": 37201090, + "step": 1147 + }, + { + "epoch": 0.10353068494386075, + "flos": 24390878144640.0, + "grad_norm": 1.7796969419367894, + "learning_rate": 3.943591829437705e-06, + "loss": 0.8333, + "num_input_tokens_seen": 37230455, + "step": 1148 + }, + { + "epoch": 0.10362086846733101, + "flos": 11967904287840.0, + "grad_norm": 1.795184990958847, + "learning_rate": 3.9434539759540895e-06, + "loss": 0.7485, + "num_input_tokens_seen": 37256395, + "step": 1149 + }, + { + "epoch": 0.10371105199080129, + "flos": 29089912612800.0, + "grad_norm": 1.605437192583517, + "learning_rate": 3.943315956644264e-06, + "loss": 0.7414, + "num_input_tokens_seen": 37285590, + "step": 1150 + }, + { + "epoch": 0.10380123551427155, + "flos": 17468816618400.0, + "grad_norm": 2.2068109723107203, + "learning_rate": 3.943177771520006e-06, + "loss": 0.8717, + "num_input_tokens_seen": 37313935, + "step": 1151 + }, + { + "epoch": 0.10389141903774181, + "flos": 24391063993440.0, + "grad_norm": 1.9090242678706875, + "learning_rate": 3.9430394205931065e-06, + "loss": 0.8467, + "num_input_tokens_seen": 37339075, + "step": 1152 + }, + { + "epoch": 0.10398160256121207, + "flos": 45189703506240.0, + "grad_norm": 1.692869051614899, + "learning_rate": 3.942900903875369e-06, + "loss": 0.6768, + "num_input_tokens_seen": 37373410, + "step": 1153 + }, + { + "epoch": 0.10407178608468233, + "flos": 25046878934400.0, + "grad_norm": 1.6429157068938431, + "learning_rate": 3.942762221378614e-06, + "loss": 0.8065, + "num_input_tokens_seen": 37403820, + "step": 1154 + }, + { + "epoch": 0.10416196960815259, + "flos": 25669161018240.0, + "grad_norm": 1.5117222985514311, + "learning_rate": 3.942623373114673e-06, + "loss": 0.8369, + "num_input_tokens_seen": 37437965, + "step": 1155 + }, + { + "epoch": 0.10425215313162285, + "flos": 20784843798240.0, + "grad_norm": 2.662684646436222, + "learning_rate": 3.942484359095396e-06, + "loss": 0.7905, + "num_input_tokens_seen": 37465450, + "step": 1156 + }, + { + "epoch": 0.10434233665509311, + "flos": 20929419929280.0, + "grad_norm": 1.718466163086814, + "learning_rate": 3.942345179332642e-06, + "loss": 0.7382, + "num_input_tokens_seen": 37494625, + "step": 1157 + }, + { + "epoch": 0.10443252017856337, + "flos": 26795024941440.0, + "grad_norm": 1.4553178218574687, + "learning_rate": 3.942205833838287e-06, + "loss": 0.7531, + "num_input_tokens_seen": 37526875, + "step": 1158 + }, + { + "epoch": 0.10452270370203363, + "flos": 20748337360320.0, + "grad_norm": 1.9142802834687944, + "learning_rate": 3.9420663226242204e-06, + "loss": 0.8954, + "num_input_tokens_seen": 37554690, + "step": 1159 + }, + { + "epoch": 0.1046128872255039, + "flos": 23006532744960.0, + "grad_norm": 2.3398094111041923, + "learning_rate": 3.941926645702348e-06, + "loss": 0.8514, + "num_input_tokens_seen": 37583090, + "step": 1160 + }, + { + "epoch": 0.10470307074897416, + "flos": 19217594511360.0, + "grad_norm": 5.466827403395913, + "learning_rate": 3.941786803084586e-06, + "loss": 0.7907, + "num_input_tokens_seen": 37610635, + "step": 1161 + }, + { + "epoch": 0.10479325427244443, + "flos": 24682260592320.0, + "grad_norm": 1.853779105783027, + "learning_rate": 3.941646794782867e-06, + "loss": 0.7518, + "num_input_tokens_seen": 37638865, + "step": 1162 + }, + { + "epoch": 0.10488343779591469, + "flos": 18999336448800.0, + "grad_norm": 2.543975865039662, + "learning_rate": 3.941506620809137e-06, + "loss": 0.8173, + "num_input_tokens_seen": 37664580, + "step": 1163 + }, + { + "epoch": 0.10497362131938495, + "flos": 27596977143360.0, + "grad_norm": 2.2755553353743316, + "learning_rate": 3.941366281175357e-06, + "loss": 0.7941, + "num_input_tokens_seen": 37692470, + "step": 1164 + }, + { + "epoch": 0.10506380484285521, + "flos": 23407843373760.0, + "grad_norm": 1.374019880364754, + "learning_rate": 3.941225775893502e-06, + "loss": 0.7399, + "num_input_tokens_seen": 37723460, + "step": 1165 + }, + { + "epoch": 0.10515398836632547, + "flos": 31967751028320.0, + "grad_norm": 1.6667583281424974, + "learning_rate": 3.941085104975559e-06, + "loss": 0.7792, + "num_input_tokens_seen": 37754475, + "step": 1166 + }, + { + "epoch": 0.10524417188979573, + "flos": 59940385932480.0, + "grad_norm": 0.7532567997121308, + "learning_rate": 3.9409442684335325e-06, + "loss": 0.6227, + "num_input_tokens_seen": 37842490, + "step": 1167 + }, + { + "epoch": 0.105334355413266, + "flos": 53210712315840.0, + "grad_norm": 2.8564040795635113, + "learning_rate": 3.940803266279438e-06, + "loss": 0.794, + "num_input_tokens_seen": 37877185, + "step": 1168 + }, + { + "epoch": 0.10542453893673626, + "flos": 23444015283840.0, + "grad_norm": 1.764875539870767, + "learning_rate": 3.9406620985253076e-06, + "loss": 0.7424, + "num_input_tokens_seen": 37903955, + "step": 1169 + }, + { + "epoch": 0.10551472246020652, + "flos": 21876282790080.0, + "grad_norm": 1.9174174275828701, + "learning_rate": 3.940520765183187e-06, + "loss": 0.7754, + "num_input_tokens_seen": 37933355, + "step": 1170 + }, + { + "epoch": 0.10560490598367678, + "flos": 22678606689600.0, + "grad_norm": 2.0014913482288432, + "learning_rate": 3.940379266265134e-06, + "loss": 0.8391, + "num_input_tokens_seen": 37960870, + "step": 1171 + }, + { + "epoch": 0.10569508950714704, + "flos": 31126764844800.0, + "grad_norm": 3.255368713497049, + "learning_rate": 3.940237601783223e-06, + "loss": 0.769, + "num_input_tokens_seen": 37989010, + "step": 1172 + }, + { + "epoch": 0.1057852730306173, + "flos": 27778877447040.0, + "grad_norm": 1.6687153701903734, + "learning_rate": 3.940095771749542e-06, + "loss": 0.8038, + "num_input_tokens_seen": 38018105, + "step": 1173 + }, + { + "epoch": 0.10587545655408757, + "flos": 22642843646880.0, + "grad_norm": 2.1431727040772874, + "learning_rate": 3.939953776176192e-06, + "loss": 0.8612, + "num_input_tokens_seen": 38046555, + "step": 1174 + }, + { + "epoch": 0.10596564007755783, + "flos": 40057944228480.0, + "grad_norm": 1.3666075430075837, + "learning_rate": 3.939811615075288e-06, + "loss": 0.8291, + "num_input_tokens_seen": 38081210, + "step": 1175 + }, + { + "epoch": 0.1060558236010281, + "flos": 21178869154080.0, + "grad_norm": 2.4536661899612655, + "learning_rate": 3.9396692884589616e-06, + "loss": 0.8689, + "num_input_tokens_seen": 38108575, + "step": 1176 + }, + { + "epoch": 0.10614600712449836, + "flos": 25519827157920.0, + "grad_norm": 2.3028743529868345, + "learning_rate": 3.9395267963393565e-06, + "loss": 0.7871, + "num_input_tokens_seen": 38137205, + "step": 1177 + }, + { + "epoch": 0.10623619064796862, + "flos": 19612103074080.0, + "grad_norm": 1.889808541545929, + "learning_rate": 3.939384138728631e-06, + "loss": 0.8412, + "num_input_tokens_seen": 38163340, + "step": 1178 + }, + { + "epoch": 0.10632637417143888, + "flos": 31783992236640.0, + "grad_norm": 1.7987245411824124, + "learning_rate": 3.939241315638956e-06, + "loss": 0.7901, + "num_input_tokens_seen": 38192810, + "step": 1179 + }, + { + "epoch": 0.10641655769490914, + "flos": 31456623727680.0, + "grad_norm": 2.1544184991945348, + "learning_rate": 3.93909832708252e-06, + "loss": 0.7877, + "num_input_tokens_seen": 38221975, + "step": 1180 + }, + { + "epoch": 0.1065067412183794, + "flos": 24825089744640.0, + "grad_norm": 1.5513028821761161, + "learning_rate": 3.938955173071523e-06, + "loss": 0.8096, + "num_input_tokens_seen": 38249905, + "step": 1181 + }, + { + "epoch": 0.10659692474184966, + "flos": 68405642177280.0, + "grad_norm": 0.6261193741880557, + "learning_rate": 3.938811853618179e-06, + "loss": 0.6089, + "num_input_tokens_seen": 38349945, + "step": 1182 + }, + { + "epoch": 0.10668710826531992, + "flos": 23951834475840.0, + "grad_norm": 4.11843488353192, + "learning_rate": 3.938668368734717e-06, + "loss": 0.8395, + "num_input_tokens_seen": 38376455, + "step": 1183 + }, + { + "epoch": 0.10677729178879018, + "flos": 23258732532000.0, + "grad_norm": 2.8964616564980585, + "learning_rate": 3.93852471843338e-06, + "loss": 0.6797, + "num_input_tokens_seen": 38406510, + "step": 1184 + }, + { + "epoch": 0.10686747531226044, + "flos": 20565544982400.0, + "grad_norm": 1.794744273169021, + "learning_rate": 3.9383809027264254e-06, + "loss": 0.7853, + "num_input_tokens_seen": 38433050, + "step": 1185 + }, + { + "epoch": 0.10695765883573072, + "flos": 23006978782080.0, + "grad_norm": 2.1754705731193096, + "learning_rate": 3.938236921626124e-06, + "loss": 0.8435, + "num_input_tokens_seen": 38462230, + "step": 1186 + }, + { + "epoch": 0.10704784235920098, + "flos": 20233455913920.0, + "grad_norm": 2.4903793232937748, + "learning_rate": 3.938092775144761e-06, + "loss": 0.8007, + "num_input_tokens_seen": 38488745, + "step": 1187 + }, + { + "epoch": 0.10713802588267124, + "flos": 24967621538880.0, + "grad_norm": 2.14187073483695, + "learning_rate": 3.9379484632946355e-06, + "loss": 0.7255, + "num_input_tokens_seen": 38517485, + "step": 1188 + }, + { + "epoch": 0.1072282094061415, + "flos": 24026594330400.0, + "grad_norm": 2.0839777394897876, + "learning_rate": 3.937803986088062e-06, + "loss": 0.8489, + "num_input_tokens_seen": 38545845, + "step": 1189 + }, + { + "epoch": 0.10731839292961176, + "flos": 20274051025440.0, + "grad_norm": 2.784554479360936, + "learning_rate": 3.937659343537367e-06, + "loss": 0.7589, + "num_input_tokens_seen": 38574335, + "step": 1190 + }, + { + "epoch": 0.10740857645308202, + "flos": 21549062960160.0, + "grad_norm": 1.795758041196013, + "learning_rate": 3.937514535654893e-06, + "loss": 0.8745, + "num_input_tokens_seen": 38602780, + "step": 1191 + }, + { + "epoch": 0.10749875997655228, + "flos": 22969282911840.0, + "grad_norm": 2.564876106051146, + "learning_rate": 3.937369562452996e-06, + "loss": 0.8074, + "num_input_tokens_seen": 38629650, + "step": 1192 + }, + { + "epoch": 0.10758894350002254, + "flos": 21841560500640.0, + "grad_norm": 1.8170231253622493, + "learning_rate": 3.937224423944044e-06, + "loss": 0.8576, + "num_input_tokens_seen": 38656985, + "step": 1193 + }, + { + "epoch": 0.1076791270234928, + "flos": 27919550753280.0, + "grad_norm": 1.7784468026086309, + "learning_rate": 3.937079120140423e-06, + "loss": 0.7341, + "num_input_tokens_seen": 38686445, + "step": 1194 + }, + { + "epoch": 0.10776931054696307, + "flos": 21585755246880.0, + "grad_norm": 1.633163122468688, + "learning_rate": 3.936933651054531e-06, + "loss": 0.8415, + "num_input_tokens_seen": 38715725, + "step": 1195 + }, + { + "epoch": 0.10785949407043333, + "flos": 23843393085120.0, + "grad_norm": 2.214675708565502, + "learning_rate": 3.936788016698779e-06, + "loss": 0.9186, + "num_input_tokens_seen": 38745120, + "step": 1196 + }, + { + "epoch": 0.1079496775939036, + "flos": 21584937512160.0, + "grad_norm": 1.609257443773359, + "learning_rate": 3.936642217085594e-06, + "loss": 0.8325, + "num_input_tokens_seen": 38775235, + "step": 1197 + }, + { + "epoch": 0.10803986111737386, + "flos": 21285452056800.0, + "grad_norm": 1.9583597451992372, + "learning_rate": 3.936496252227417e-06, + "loss": 0.7125, + "num_input_tokens_seen": 38801845, + "step": 1198 + }, + { + "epoch": 0.10813004464084412, + "flos": 20013450872640.0, + "grad_norm": 4.497784174984523, + "learning_rate": 3.936350122136703e-06, + "loss": 0.7831, + "num_input_tokens_seen": 38829085, + "step": 1199 + }, + { + "epoch": 0.10822022816431438, + "flos": 66403958271840.0, + "grad_norm": 0.6856375305574693, + "learning_rate": 3.936203826825919e-06, + "loss": 0.6296, + "num_input_tokens_seen": 38924855, + "step": 1200 + }, + { + "epoch": 0.10831041168778464, + "flos": 21073512853440.0, + "grad_norm": 1.745855191381485, + "learning_rate": 3.9360573663075475e-06, + "loss": 0.7623, + "num_input_tokens_seen": 38954710, + "step": 1201 + }, + { + "epoch": 0.1084005952112549, + "flos": 21075966057600.0, + "grad_norm": 2.1557257194580575, + "learning_rate": 3.935910740594087e-06, + "loss": 0.8255, + "num_input_tokens_seen": 38981855, + "step": 1202 + }, + { + "epoch": 0.10849077873472517, + "flos": 27775420659360.0, + "grad_norm": 2.5220327335081754, + "learning_rate": 3.935763949698047e-06, + "loss": 0.8111, + "num_input_tokens_seen": 39011690, + "step": 1203 + }, + { + "epoch": 0.10858096225819543, + "flos": 20524541003520.0, + "grad_norm": 1.774302872242882, + "learning_rate": 3.935616993631954e-06, + "loss": 0.8627, + "num_input_tokens_seen": 39038165, + "step": 1204 + }, + { + "epoch": 0.10867114578166569, + "flos": 28287700222560.0, + "grad_norm": 2.288877928912312, + "learning_rate": 3.935469872408345e-06, + "loss": 0.7954, + "num_input_tokens_seen": 39067405, + "step": 1205 + }, + { + "epoch": 0.10876132930513595, + "flos": 21403520415360.0, + "grad_norm": 1.898546820110704, + "learning_rate": 3.935322586039776e-06, + "loss": 0.8254, + "num_input_tokens_seen": 39095850, + "step": 1206 + }, + { + "epoch": 0.10885151282860621, + "flos": 35357199951360.0, + "grad_norm": 1.6783054077239514, + "learning_rate": 3.935175134538811e-06, + "loss": 0.7375, + "num_input_tokens_seen": 39130685, + "step": 1207 + }, + { + "epoch": 0.10894169635207647, + "flos": 14189630404320.0, + "grad_norm": 3.5924416192075483, + "learning_rate": 3.935027517918034e-06, + "loss": 0.7446, + "num_input_tokens_seen": 39153435, + "step": 1208 + }, + { + "epoch": 0.10903187987554674, + "flos": 17900723693280.0, + "grad_norm": 2.453175538143829, + "learning_rate": 3.93487973619004e-06, + "loss": 0.8341, + "num_input_tokens_seen": 39179770, + "step": 1209 + }, + { + "epoch": 0.109122063399017, + "flos": 13861444160640.0, + "grad_norm": 2.2605846243014063, + "learning_rate": 3.934731789367438e-06, + "loss": 0.8106, + "num_input_tokens_seen": 39206280, + "step": 1210 + }, + { + "epoch": 0.10921224692248727, + "flos": 63469393068960.0, + "grad_norm": 0.7607109641252708, + "learning_rate": 3.9345836774628505e-06, + "loss": 0.6084, + "num_input_tokens_seen": 39282975, + "step": 1211 + }, + { + "epoch": 0.10930243044595753, + "flos": 21585978265440.0, + "grad_norm": 1.8282359814987317, + "learning_rate": 3.934435400488917e-06, + "loss": 0.7939, + "num_input_tokens_seen": 39311545, + "step": 1212 + }, + { + "epoch": 0.10939261396942779, + "flos": 15355346043840.0, + "grad_norm": 2.036736927017205, + "learning_rate": 3.934286958458289e-06, + "loss": 0.8387, + "num_input_tokens_seen": 39337400, + "step": 1213 + }, + { + "epoch": 0.10948279749289805, + "flos": 21039979996320.0, + "grad_norm": 1.636098011088895, + "learning_rate": 3.934138351383632e-06, + "loss": 0.8643, + "num_input_tokens_seen": 39366110, + "step": 1214 + }, + { + "epoch": 0.10957298101636831, + "flos": 39364990963680.0, + "grad_norm": 1.6067145124938043, + "learning_rate": 3.933989579277626e-06, + "loss": 0.7352, + "num_input_tokens_seen": 39399135, + "step": 1215 + }, + { + "epoch": 0.10966316453983857, + "flos": 22313728159200.0, + "grad_norm": 6.26646512837007, + "learning_rate": 3.933840642152966e-06, + "loss": 0.7943, + "num_input_tokens_seen": 39430820, + "step": 1216 + }, + { + "epoch": 0.10975334806330883, + "flos": 20857001769600.0, + "grad_norm": 2.254542639421568, + "learning_rate": 3.933691540022359e-06, + "loss": 0.8282, + "num_input_tokens_seen": 39456750, + "step": 1217 + }, + { + "epoch": 0.10984353158677909, + "flos": 22751359377120.0, + "grad_norm": 1.8479738506886345, + "learning_rate": 3.933542272898527e-06, + "loss": 0.7707, + "num_input_tokens_seen": 39484670, + "step": 1218 + }, + { + "epoch": 0.10993371511024935, + "flos": 65117572390560.0, + "grad_norm": 0.7584193444642928, + "learning_rate": 3.933392840794207e-06, + "loss": 0.5656, + "num_input_tokens_seen": 39572555, + "step": 1219 + }, + { + "epoch": 0.11002389863371961, + "flos": 13314851175360.0, + "grad_norm": 2.305652170825168, + "learning_rate": 3.93324324372215e-06, + "loss": 0.9066, + "num_input_tokens_seen": 39597525, + "step": 1220 + }, + { + "epoch": 0.11011408215718989, + "flos": 37761272408640.0, + "grad_norm": 1.888808962433688, + "learning_rate": 3.9330934816951185e-06, + "loss": 0.8265, + "num_input_tokens_seen": 39628215, + "step": 1221 + }, + { + "epoch": 0.11020426568066015, + "flos": 20158956247680.0, + "grad_norm": 2.0556044472501727, + "learning_rate": 3.932943554725893e-06, + "loss": 0.8301, + "num_input_tokens_seen": 39655915, + "step": 1222 + }, + { + "epoch": 0.11029444920413041, + "flos": 57507798535680.0, + "grad_norm": 0.7611089759952951, + "learning_rate": 3.932793462827265e-06, + "loss": 0.6396, + "num_input_tokens_seen": 39746620, + "step": 1223 + }, + { + "epoch": 0.11038463272760067, + "flos": 25154205232320.0, + "grad_norm": 1.5816507517687584, + "learning_rate": 3.932643206012041e-06, + "loss": 0.8611, + "num_input_tokens_seen": 39776965, + "step": 1224 + }, + { + "epoch": 0.11047481625107093, + "flos": 20820755520000.0, + "grad_norm": 2.12638764794223, + "learning_rate": 3.932492784293043e-06, + "loss": 0.7321, + "num_input_tokens_seen": 39804025, + "step": 1225 + }, + { + "epoch": 0.11056499977454119, + "flos": 26467619262720.0, + "grad_norm": 1.9344425072545555, + "learning_rate": 3.932342197683104e-06, + "loss": 0.792, + "num_input_tokens_seen": 39834255, + "step": 1226 + }, + { + "epoch": 0.11065518329801145, + "flos": 27487903866720.0, + "grad_norm": 1.8977596501353924, + "learning_rate": 3.932191446195075e-06, + "loss": 0.6797, + "num_input_tokens_seen": 39865110, + "step": 1227 + }, + { + "epoch": 0.11074536682148171, + "flos": 20383495999680.0, + "grad_norm": 1.9888840756396557, + "learning_rate": 3.9320405298418175e-06, + "loss": 0.8094, + "num_input_tokens_seen": 39892430, + "step": 1228 + }, + { + "epoch": 0.11083555034495197, + "flos": 24063137938080.0, + "grad_norm": 2.06067183418494, + "learning_rate": 3.9318894486362076e-06, + "loss": 0.7224, + "num_input_tokens_seen": 39922905, + "step": 1229 + }, + { + "epoch": 0.11092573386842224, + "flos": 23152781515200.0, + "grad_norm": 2.1576378820039284, + "learning_rate": 3.9317382025911395e-06, + "loss": 0.8775, + "num_input_tokens_seen": 39951070, + "step": 1230 + }, + { + "epoch": 0.1110159173918925, + "flos": 20929196910720.0, + "grad_norm": 1.850004933324886, + "learning_rate": 3.9315867917195145e-06, + "loss": 0.8994, + "num_input_tokens_seen": 39977830, + "step": 1231 + }, + { + "epoch": 0.11110610091536276, + "flos": 22823554518240.0, + "grad_norm": 1.5626685242370615, + "learning_rate": 3.931435216034256e-06, + "loss": 0.806, + "num_input_tokens_seen": 40007795, + "step": 1232 + }, + { + "epoch": 0.11119628443883303, + "flos": 22860618502560.0, + "grad_norm": 2.23996659185837, + "learning_rate": 3.931283475548293e-06, + "loss": 0.8059, + "num_input_tokens_seen": 40036940, + "step": 1233 + }, + { + "epoch": 0.11128646796230329, + "flos": 22539011306400.0, + "grad_norm": 1.4247768260801872, + "learning_rate": 3.931131570274576e-06, + "loss": 0.8043, + "num_input_tokens_seen": 40068410, + "step": 1234 + }, + { + "epoch": 0.11137665148577355, + "flos": 26540260440960.0, + "grad_norm": 1.5220922546097424, + "learning_rate": 3.930979500226065e-06, + "loss": 0.8386, + "num_input_tokens_seen": 40100440, + "step": 1235 + }, + { + "epoch": 0.11146683500924381, + "flos": 20346915222240.0, + "grad_norm": 2.1975238563535924, + "learning_rate": 3.930827265415736e-06, + "loss": 0.6409, + "num_input_tokens_seen": 40128205, + "step": 1236 + }, + { + "epoch": 0.11155701853271408, + "flos": 21549471827520.0, + "grad_norm": 8.582187711932871, + "learning_rate": 3.930674865856578e-06, + "loss": 0.8151, + "num_input_tokens_seen": 40154990, + "step": 1237 + }, + { + "epoch": 0.11164720205618434, + "flos": 23407099978560.0, + "grad_norm": 1.7354612643162095, + "learning_rate": 3.930522301561595e-06, + "loss": 0.8338, + "num_input_tokens_seen": 40184595, + "step": 1238 + }, + { + "epoch": 0.1117373855796546, + "flos": 21986842857120.0, + "grad_norm": 2.1654699272792794, + "learning_rate": 3.930369572543804e-06, + "loss": 0.8486, + "num_input_tokens_seen": 40212450, + "step": 1239 + }, + { + "epoch": 0.11182756910312486, + "flos": 21618433199520.0, + "grad_norm": 1.631010384335371, + "learning_rate": 3.930216678816237e-06, + "loss": 0.7478, + "num_input_tokens_seen": 40243600, + "step": 1240 + }, + { + "epoch": 0.11191775262659512, + "flos": 71294928878880.0, + "grad_norm": 0.6668669553734479, + "learning_rate": 3.930063620391941e-06, + "loss": 0.6136, + "num_input_tokens_seen": 40334185, + "step": 1241 + }, + { + "epoch": 0.11200793615006538, + "flos": 22314248535840.0, + "grad_norm": 2.084909936071177, + "learning_rate": 3.9299103972839735e-06, + "loss": 0.8101, + "num_input_tokens_seen": 40363385, + "step": 1242 + }, + { + "epoch": 0.11209811967353564, + "flos": 65990678980320.0, + "grad_norm": 0.6588383221637237, + "learning_rate": 3.92975700950541e-06, + "loss": 0.5776, + "num_input_tokens_seen": 40459670, + "step": 1243 + }, + { + "epoch": 0.1121883031970059, + "flos": 25043682335040.0, + "grad_norm": 1.9915210282242504, + "learning_rate": 3.929603457069338e-06, + "loss": 0.836, + "num_input_tokens_seen": 40488960, + "step": 1244 + }, + { + "epoch": 0.11227848672047618, + "flos": 22275734930880.0, + "grad_norm": 2.103959089055486, + "learning_rate": 3.929449739988859e-06, + "loss": 0.77, + "num_input_tokens_seen": 40517290, + "step": 1245 + }, + { + "epoch": 0.11236867024394644, + "flos": 25373615557440.0, + "grad_norm": 2.2575175478771783, + "learning_rate": 3.929295858277089e-06, + "loss": 0.7997, + "num_input_tokens_seen": 40547460, + "step": 1246 + }, + { + "epoch": 0.1124588537674167, + "flos": 20602497457440.0, + "grad_norm": 3.0239251110754344, + "learning_rate": 3.9291418119471585e-06, + "loss": 0.8006, + "num_input_tokens_seen": 40574400, + "step": 1247 + }, + { + "epoch": 0.11254903729088696, + "flos": 17870350265760.0, + "grad_norm": 2.3882887650790536, + "learning_rate": 3.928987601012212e-06, + "loss": 0.8249, + "num_input_tokens_seen": 40600665, + "step": 1248 + }, + { + "epoch": 0.11263922081435722, + "flos": 16703631042720.0, + "grad_norm": 1.8784832071706663, + "learning_rate": 3.928833225485407e-06, + "loss": 0.7947, + "num_input_tokens_seen": 40626195, + "step": 1249 + }, + { + "epoch": 0.11272940433782748, + "flos": 23548814038080.0, + "grad_norm": 2.013080854290116, + "learning_rate": 3.928678685379915e-06, + "loss": 0.7916, + "num_input_tokens_seen": 40654825, + "step": 1250 + }, + { + "epoch": 0.11281958786129774, + "flos": 25004425334880.0, + "grad_norm": 1.6349983004173287, + "learning_rate": 3.928523980708924e-06, + "loss": 0.8533, + "num_input_tokens_seen": 40682835, + "step": 1251 + }, + { + "epoch": 0.112909771384768, + "flos": 35903161050720.0, + "grad_norm": 2.3850500388664235, + "learning_rate": 3.928369111485632e-06, + "loss": 0.7604, + "num_input_tokens_seen": 40712655, + "step": 1252 + }, + { + "epoch": 0.11299995490823826, + "flos": 21984240973920.0, + "grad_norm": 2.1088683473564522, + "learning_rate": 3.928214077723255e-06, + "loss": 0.7931, + "num_input_tokens_seen": 40742450, + "step": 1253 + }, + { + "epoch": 0.11309013843170852, + "flos": 22606225699680.0, + "grad_norm": 3.387003850321444, + "learning_rate": 3.928058879435021e-06, + "loss": 0.7917, + "num_input_tokens_seen": 40769535, + "step": 1254 + }, + { + "epoch": 0.11318032195517878, + "flos": 16667868000000.0, + "grad_norm": 1.930901520037785, + "learning_rate": 3.9279035166341725e-06, + "loss": 0.7877, + "num_input_tokens_seen": 40795365, + "step": 1255 + }, + { + "epoch": 0.11327050547864904, + "flos": 30509166150720.0, + "grad_norm": 1.8304343350179546, + "learning_rate": 3.927747989333965e-06, + "loss": 0.7844, + "num_input_tokens_seen": 40826205, + "step": 1256 + }, + { + "epoch": 0.11336068900211932, + "flos": 24095815890720.0, + "grad_norm": 2.1972403427364924, + "learning_rate": 3.927592297547669e-06, + "loss": 0.7632, + "num_input_tokens_seen": 40854495, + "step": 1257 + }, + { + "epoch": 0.11345087252558958, + "flos": 19035322510080.0, + "grad_norm": 2.338670745436455, + "learning_rate": 3.927436441288571e-06, + "loss": 0.8297, + "num_input_tokens_seen": 40879485, + "step": 1258 + }, + { + "epoch": 0.11354105604905984, + "flos": 29488175321280.0, + "grad_norm": 1.7826298505613971, + "learning_rate": 3.927280420569968e-06, + "loss": 0.8206, + "num_input_tokens_seen": 40909320, + "step": 1259 + }, + { + "epoch": 0.1136312395725301, + "flos": 29856213281280.0, + "grad_norm": 1.6902402598677742, + "learning_rate": 3.927124235405171e-06, + "loss": 0.8295, + "num_input_tokens_seen": 40937920, + "step": 1260 + }, + { + "epoch": 0.11372142309600036, + "flos": 24170575745280.0, + "grad_norm": 2.2487819341541164, + "learning_rate": 3.92696788580751e-06, + "loss": 0.6613, + "num_input_tokens_seen": 40968650, + "step": 1261 + }, + { + "epoch": 0.11381160661947062, + "flos": 34847670950400.0, + "grad_norm": 1.7290222460019997, + "learning_rate": 3.9268113717903225e-06, + "loss": 0.6397, + "num_input_tokens_seen": 40998960, + "step": 1262 + }, + { + "epoch": 0.11390179014294088, + "flos": 21440324211360.0, + "grad_norm": 1.8480319715764524, + "learning_rate": 3.926654693366965e-06, + "loss": 0.8705, + "num_input_tokens_seen": 41027500, + "step": 1263 + }, + { + "epoch": 0.11399197366641114, + "flos": 21288760165440.0, + "grad_norm": 2.663836696150142, + "learning_rate": 3.926497850550805e-06, + "loss": 0.8167, + "num_input_tokens_seen": 41054710, + "step": 1264 + }, + { + "epoch": 0.1140821571898814, + "flos": 20820420992160.0, + "grad_norm": 2.550370701790768, + "learning_rate": 3.926340843355226e-06, + "loss": 0.7958, + "num_input_tokens_seen": 41083125, + "step": 1265 + }, + { + "epoch": 0.11417234071335167, + "flos": 19727160682080.0, + "grad_norm": 2.188493448555727, + "learning_rate": 3.926183671793625e-06, + "loss": 0.8145, + "num_input_tokens_seen": 41111455, + "step": 1266 + }, + { + "epoch": 0.11426252423682193, + "flos": 22205509787040.0, + "grad_norm": 1.8769959316870082, + "learning_rate": 3.926026335879412e-06, + "loss": 0.768, + "num_input_tokens_seen": 41138295, + "step": 1267 + }, + { + "epoch": 0.11435270776029219, + "flos": 70576291309920.0, + "grad_norm": 0.6283909125679903, + "learning_rate": 3.925868835626012e-06, + "loss": 0.5503, + "num_input_tokens_seen": 41231875, + "step": 1268 + }, + { + "epoch": 0.11444289128376246, + "flos": 20784397761120.0, + "grad_norm": 2.193025640029023, + "learning_rate": 3.925711171046864e-06, + "loss": 0.8114, + "num_input_tokens_seen": 41261915, + "step": 1269 + }, + { + "epoch": 0.11453307480723272, + "flos": 23078988074400.0, + "grad_norm": 1.655476227437838, + "learning_rate": 3.925553342155421e-06, + "loss": 0.7638, + "num_input_tokens_seen": 41293070, + "step": 1270 + }, + { + "epoch": 0.11462325833070298, + "flos": 60978883718880.0, + "grad_norm": 0.6776138550292737, + "learning_rate": 3.9253953489651485e-06, + "loss": 0.6248, + "num_input_tokens_seen": 41381965, + "step": 1271 + }, + { + "epoch": 0.11471344185417325, + "flos": 63014918216160.0, + "grad_norm": 0.7069609169726109, + "learning_rate": 3.925237191489529e-06, + "loss": 0.6001, + "num_input_tokens_seen": 41477040, + "step": 1272 + }, + { + "epoch": 0.1148036253776435, + "flos": 23371336935840.0, + "grad_norm": 1.7902510247485293, + "learning_rate": 3.925078869742056e-06, + "loss": 0.7695, + "num_input_tokens_seen": 41506920, + "step": 1273 + }, + { + "epoch": 0.11489380890111377, + "flos": 21731372131200.0, + "grad_norm": 2.375013150630481, + "learning_rate": 3.92492038373624e-06, + "loss": 0.7356, + "num_input_tokens_seen": 41537240, + "step": 1274 + }, + { + "epoch": 0.11498399242458403, + "flos": 22861101709440.0, + "grad_norm": 1.8532427796990312, + "learning_rate": 3.924761733485602e-06, + "loss": 0.8551, + "num_input_tokens_seen": 41566690, + "step": 1275 + }, + { + "epoch": 0.11507417594805429, + "flos": 25226734901280.0, + "grad_norm": 1.7184053942533535, + "learning_rate": 3.92460291900368e-06, + "loss": 0.7986, + "num_input_tokens_seen": 41597760, + "step": 1276 + }, + { + "epoch": 0.11516435947152455, + "flos": 21804645195360.0, + "grad_norm": 1.7451464240171268, + "learning_rate": 3.924443940304025e-06, + "loss": 0.8202, + "num_input_tokens_seen": 41625905, + "step": 1277 + }, + { + "epoch": 0.11525454299499481, + "flos": 19253915100480.0, + "grad_norm": 1.933044787227994, + "learning_rate": 3.924284797400202e-06, + "loss": 0.8491, + "num_input_tokens_seen": 41652105, + "step": 1278 + }, + { + "epoch": 0.11534472651846507, + "flos": 20455988498880.0, + "grad_norm": 2.6132661964427695, + "learning_rate": 3.924125490305789e-06, + "loss": 0.8206, + "num_input_tokens_seen": 41674525, + "step": 1279 + }, + { + "epoch": 0.11543491004193533, + "flos": 31200929983200.0, + "grad_norm": 2.6610748452688817, + "learning_rate": 3.923966019034381e-06, + "loss": 0.7825, + "num_input_tokens_seen": 41706520, + "step": 1280 + }, + { + "epoch": 0.1155250935654056, + "flos": 19326816467040.0, + "grad_norm": 1.8715013287086264, + "learning_rate": 3.923806383599583e-06, + "loss": 0.8092, + "num_input_tokens_seen": 41734305, + "step": 1281 + }, + { + "epoch": 0.11561527708887587, + "flos": 28216285646400.0, + "grad_norm": 2.1147254056761966, + "learning_rate": 3.923646584015017e-06, + "loss": 0.7684, + "num_input_tokens_seen": 41764370, + "step": 1282 + }, + { + "epoch": 0.11570546061234613, + "flos": 23216055913920.0, + "grad_norm": 3.0432260426567233, + "learning_rate": 3.923486620294316e-06, + "loss": 0.8034, + "num_input_tokens_seen": 41788965, + "step": 1283 + }, + { + "epoch": 0.11579564413581639, + "flos": 29707436967360.0, + "grad_norm": 1.9601844466284486, + "learning_rate": 3.923326492451132e-06, + "loss": 0.809, + "num_input_tokens_seen": 41815520, + "step": 1284 + }, + { + "epoch": 0.11588582765928665, + "flos": 23845288742880.0, + "grad_norm": 2.884894552171437, + "learning_rate": 3.923166200499125e-06, + "loss": 0.7815, + "num_input_tokens_seen": 41847555, + "step": 1285 + }, + { + "epoch": 0.11597601118275691, + "flos": 38634118810080.0, + "grad_norm": 1.9514609610513156, + "learning_rate": 3.923005744451975e-06, + "loss": 0.771, + "num_input_tokens_seen": 41880440, + "step": 1286 + }, + { + "epoch": 0.11606619470622717, + "flos": 20201112489120.0, + "grad_norm": 2.265204660293444, + "learning_rate": 3.9228451243233715e-06, + "loss": 0.7535, + "num_input_tokens_seen": 41906705, + "step": 1287 + }, + { + "epoch": 0.11615637822969743, + "flos": 25775260714080.0, + "grad_norm": 1.5153493728918848, + "learning_rate": 3.9226843401270195e-06, + "loss": 0.8302, + "num_input_tokens_seen": 41938110, + "step": 1288 + }, + { + "epoch": 0.1162465617531677, + "flos": 20674655428800.0, + "grad_norm": 1.9736127592839479, + "learning_rate": 3.9225233918766376e-06, + "loss": 0.8021, + "num_input_tokens_seen": 41967515, + "step": 1289 + }, + { + "epoch": 0.11633674527663795, + "flos": 27994161928800.0, + "grad_norm": 1.6689734548625657, + "learning_rate": 3.92236227958596e-06, + "loss": 0.8185, + "num_input_tokens_seen": 41996135, + "step": 1290 + }, + { + "epoch": 0.11642692880010821, + "flos": 23876331226080.0, + "grad_norm": 2.45925690926057, + "learning_rate": 3.922201003268731e-06, + "loss": 0.8205, + "num_input_tokens_seen": 42024025, + "step": 1291 + }, + { + "epoch": 0.11651711232357848, + "flos": 71612707589760.0, + "grad_norm": 0.7456358857811105, + "learning_rate": 3.922039562938715e-06, + "loss": 0.6257, + "num_input_tokens_seen": 42120780, + "step": 1292 + }, + { + "epoch": 0.11660729584704875, + "flos": 25702656705600.0, + "grad_norm": 2.1175048139893486, + "learning_rate": 3.921877958609685e-06, + "loss": 0.8519, + "num_input_tokens_seen": 42151405, + "step": 1293 + }, + { + "epoch": 0.11669747937051901, + "flos": 57940820703360.0, + "grad_norm": 0.7966285773340169, + "learning_rate": 3.921716190295431e-06, + "loss": 0.6906, + "num_input_tokens_seen": 42230740, + "step": 1294 + }, + { + "epoch": 0.11678766289398927, + "flos": 19721287860000.0, + "grad_norm": 1.8555997460954101, + "learning_rate": 3.921554258009755e-06, + "loss": 0.752, + "num_input_tokens_seen": 42257435, + "step": 1295 + }, + { + "epoch": 0.11687784641745953, + "flos": 20784732288960.0, + "grad_norm": 1.6612646374656777, + "learning_rate": 3.921392161766474e-06, + "loss": 0.7863, + "num_input_tokens_seen": 42286025, + "step": 1296 + }, + { + "epoch": 0.1169680299409298, + "flos": 34482643740960.0, + "grad_norm": 3.8644087238743587, + "learning_rate": 3.92122990157942e-06, + "loss": 0.7746, + "num_input_tokens_seen": 42317025, + "step": 1297 + }, + { + "epoch": 0.11705821346440005, + "flos": 21328388863200.0, + "grad_norm": 1.691020472275732, + "learning_rate": 3.921067477462437e-06, + "loss": 0.8431, + "num_input_tokens_seen": 42345975, + "step": 1298 + }, + { + "epoch": 0.11714839698787032, + "flos": 14954778810240.0, + "grad_norm": 2.2396709927695344, + "learning_rate": 3.920904889429385e-06, + "loss": 0.8518, + "num_input_tokens_seen": 42370800, + "step": 1299 + }, + { + "epoch": 0.11723858051134058, + "flos": 25883999462880.0, + "grad_norm": 1.7550473313863915, + "learning_rate": 3.920742137494135e-06, + "loss": 0.8363, + "num_input_tokens_seen": 42401245, + "step": 1300 + }, + { + "epoch": 0.11732876403481084, + "flos": 23443792265280.0, + "grad_norm": 1.5735906289677355, + "learning_rate": 3.920579221670575e-06, + "loss": 0.8586, + "num_input_tokens_seen": 42430225, + "step": 1301 + }, + { + "epoch": 0.1174189475582811, + "flos": 35714421511200.0, + "grad_norm": 1.9194215450524916, + "learning_rate": 3.920416141972606e-06, + "loss": 0.8036, + "num_input_tokens_seen": 42460335, + "step": 1302 + }, + { + "epoch": 0.11750913108175136, + "flos": 25629123453120.0, + "grad_norm": 1.592419012456684, + "learning_rate": 3.920252898414143e-06, + "loss": 0.7955, + "num_input_tokens_seen": 42491765, + "step": 1303 + }, + { + "epoch": 0.11759931460522162, + "flos": 21112844193120.0, + "grad_norm": 1.9487117497848134, + "learning_rate": 3.920089491009114e-06, + "loss": 0.8102, + "num_input_tokens_seen": 42519560, + "step": 1304 + }, + { + "epoch": 0.1176894981286919, + "flos": 22602434384160.0, + "grad_norm": 1.853035578250808, + "learning_rate": 3.919925919771463e-06, + "loss": 0.885, + "num_input_tokens_seen": 42546945, + "step": 1305 + }, + { + "epoch": 0.11777968165216215, + "flos": 19982891596320.0, + "grad_norm": 2.245250457352465, + "learning_rate": 3.919762184715146e-06, + "loss": 0.7147, + "num_input_tokens_seen": 42575245, + "step": 1306 + }, + { + "epoch": 0.11786986517563242, + "flos": 13970814795360.0, + "grad_norm": 3.871217050138081, + "learning_rate": 3.919598285854134e-06, + "loss": 0.8005, + "num_input_tokens_seen": 42601625, + "step": 1307 + }, + { + "epoch": 0.11796004869910268, + "flos": 22386406507200.0, + "grad_norm": 2.5026783900744234, + "learning_rate": 3.919434223202411e-06, + "loss": 0.8577, + "num_input_tokens_seen": 42628875, + "step": 1308 + }, + { + "epoch": 0.11805023222257294, + "flos": 23370667880160.0, + "grad_norm": 1.6279551704851627, + "learning_rate": 3.919269996773977e-06, + "loss": 0.8564, + "num_input_tokens_seen": 42657735, + "step": 1309 + }, + { + "epoch": 0.1181404157460432, + "flos": 22602174195840.0, + "grad_norm": 1.4130942892105909, + "learning_rate": 3.919105606582844e-06, + "loss": 0.8142, + "num_input_tokens_seen": 42687845, + "step": 1310 + }, + { + "epoch": 0.11823059926951346, + "flos": 23990348080800.0, + "grad_norm": 2.6585569349458886, + "learning_rate": 3.918941052643039e-06, + "loss": 0.7656, + "num_input_tokens_seen": 42715195, + "step": 1311 + }, + { + "epoch": 0.11832078279298372, + "flos": 20820012124800.0, + "grad_norm": 1.562500686645357, + "learning_rate": 3.918776334968602e-06, + "loss": 0.8111, + "num_input_tokens_seen": 42743885, + "step": 1312 + }, + { + "epoch": 0.11841096631645398, + "flos": 21068866633440.0, + "grad_norm": 2.0006120460517507, + "learning_rate": 3.918611453573589e-06, + "loss": 0.9091, + "num_input_tokens_seen": 42770800, + "step": 1313 + }, + { + "epoch": 0.11850114983992424, + "flos": 20054752209600.0, + "grad_norm": 1.774441607045773, + "learning_rate": 3.918446408472066e-06, + "loss": 0.736, + "num_input_tokens_seen": 42800090, + "step": 1314 + }, + { + "epoch": 0.1185913333633945, + "flos": 27305111488800.0, + "grad_norm": 2.36310651148939, + "learning_rate": 3.918281199678119e-06, + "loss": 0.8379, + "num_input_tokens_seen": 42828115, + "step": 1315 + }, + { + "epoch": 0.11868151688686476, + "flos": 68475346944480.0, + "grad_norm": 0.7530979705645425, + "learning_rate": 3.9181158272058414e-06, + "loss": 0.6091, + "num_input_tokens_seen": 42918440, + "step": 1316 + }, + { + "epoch": 0.11877170041033504, + "flos": 21439952513760.0, + "grad_norm": 1.5150862847360584, + "learning_rate": 3.9179502910693455e-06, + "loss": 0.8288, + "num_input_tokens_seen": 42946550, + "step": 1317 + }, + { + "epoch": 0.1188618839338053, + "flos": 38344185983040.0, + "grad_norm": 1.659467558784762, + "learning_rate": 3.917784591282756e-06, + "loss": 0.7607, + "num_input_tokens_seen": 42978910, + "step": 1318 + }, + { + "epoch": 0.11895206745727556, + "flos": 24789549720480.0, + "grad_norm": 1.8259723085098256, + "learning_rate": 3.9176187278602105e-06, + "loss": 0.8243, + "num_input_tokens_seen": 43009550, + "step": 1319 + }, + { + "epoch": 0.11904225098074582, + "flos": 18780446500320.0, + "grad_norm": 1.692348015411527, + "learning_rate": 3.9174527008158606e-06, + "loss": 0.7431, + "num_input_tokens_seen": 43039360, + "step": 1320 + }, + { + "epoch": 0.11913243450421608, + "flos": 23626175775840.0, + "grad_norm": 1.985432380648775, + "learning_rate": 3.917286510163874e-06, + "loss": 0.7192, + "num_input_tokens_seen": 43069065, + "step": 1321 + }, + { + "epoch": 0.11922261802768634, + "flos": 26066345803680.0, + "grad_norm": 1.1760208544763435, + "learning_rate": 3.917120155918431e-06, + "loss": 0.7733, + "num_input_tokens_seen": 43100595, + "step": 1322 + }, + { + "epoch": 0.1193128015511566, + "flos": 66698314300320.0, + "grad_norm": 0.7031930678586851, + "learning_rate": 3.916953638093725e-06, + "loss": 0.6392, + "num_input_tokens_seen": 43195775, + "step": 1323 + }, + { + "epoch": 0.11940298507462686, + "flos": 23185385128320.0, + "grad_norm": 2.615510226594873, + "learning_rate": 3.916786956703964e-06, + "loss": 0.7922, + "num_input_tokens_seen": 43224385, + "step": 1324 + }, + { + "epoch": 0.11949316859809712, + "flos": 21257383154400.0, + "grad_norm": 2.448851343368254, + "learning_rate": 3.916620111763372e-06, + "loss": 0.7889, + "num_input_tokens_seen": 43252655, + "step": 1325 + }, + { + "epoch": 0.11958335212156739, + "flos": 16120866147360.0, + "grad_norm": 1.6870051470792768, + "learning_rate": 3.916453103286183e-06, + "loss": 0.861, + "num_input_tokens_seen": 43280295, + "step": 1326 + }, + { + "epoch": 0.11967353564503765, + "flos": 20740197182880.0, + "grad_norm": 2.1169181173757825, + "learning_rate": 3.916285931286648e-06, + "loss": 0.8466, + "num_input_tokens_seen": 43306225, + "step": 1327 + }, + { + "epoch": 0.11976371916850792, + "flos": 31419076536480.0, + "grad_norm": 1.898615385284983, + "learning_rate": 3.916118595779031e-06, + "loss": 0.7437, + "num_input_tokens_seen": 43337040, + "step": 1328 + }, + { + "epoch": 0.11985390269197818, + "flos": 26139693207360.0, + "grad_norm": 1.5358710335740597, + "learning_rate": 3.915951096777611e-06, + "loss": 0.769, + "num_input_tokens_seen": 43365985, + "step": 1329 + }, + { + "epoch": 0.11994408621544844, + "flos": 18707768152320.0, + "grad_norm": 1.8927670788995197, + "learning_rate": 3.915783434296678e-06, + "loss": 0.7469, + "num_input_tokens_seen": 43395535, + "step": 1330 + }, + { + "epoch": 0.1200342697389187, + "flos": 16883747197920.0, + "grad_norm": 2.447540730496922, + "learning_rate": 3.91561560835054e-06, + "loss": 0.7794, + "num_input_tokens_seen": 43419620, + "step": 1331 + }, + { + "epoch": 0.12012445326238896, + "flos": 21477239516640.0, + "grad_norm": 1.6573540228864, + "learning_rate": 3.915447618953515e-06, + "loss": 0.8167, + "num_input_tokens_seen": 43449285, + "step": 1332 + }, + { + "epoch": 0.12021463678585922, + "flos": 24645902833440.0, + "grad_norm": 6.627634927994264, + "learning_rate": 3.915279466119937e-06, + "loss": 0.7538, + "num_input_tokens_seen": 43476975, + "step": 1333 + }, + { + "epoch": 0.12030482030932949, + "flos": 68820594108000.0, + "grad_norm": 0.7229693585045543, + "learning_rate": 3.9151111498641546e-06, + "loss": 0.6238, + "num_input_tokens_seen": 43569080, + "step": 1334 + }, + { + "epoch": 0.12039500383279975, + "flos": 38307530866080.0, + "grad_norm": 2.0739040100044175, + "learning_rate": 3.914942670200529e-06, + "loss": 0.6724, + "num_input_tokens_seen": 43597180, + "step": 1335 + }, + { + "epoch": 0.12048518735627001, + "flos": 25812064510080.0, + "grad_norm": 2.0967268776346986, + "learning_rate": 3.914774027143436e-06, + "loss": 0.8576, + "num_input_tokens_seen": 43626830, + "step": 1336 + }, + { + "epoch": 0.12057537087974027, + "flos": 16958990259360.0, + "grad_norm": 1.6901994941837961, + "learning_rate": 3.914605220707265e-06, + "loss": 0.7879, + "num_input_tokens_seen": 43650345, + "step": 1337 + }, + { + "epoch": 0.12066555440321053, + "flos": 16157038057440.0, + "grad_norm": 2.0291793833287786, + "learning_rate": 3.9144362509064194e-06, + "loss": 0.9456, + "num_input_tokens_seen": 43675870, + "step": 1338 + }, + { + "epoch": 0.12075573792668079, + "flos": 13934865903840.0, + "grad_norm": 2.3930864813896022, + "learning_rate": 3.914267117755317e-06, + "loss": 0.8657, + "num_input_tokens_seen": 43701045, + "step": 1339 + }, + { + "epoch": 0.12084592145015106, + "flos": 33899432808480.0, + "grad_norm": 1.6478007022197314, + "learning_rate": 3.914097821268389e-06, + "loss": 0.7653, + "num_input_tokens_seen": 43731235, + "step": 1340 + }, + { + "epoch": 0.12093610497362133, + "flos": 25301123058240.0, + "grad_norm": 1.9720119519442167, + "learning_rate": 3.913928361460081e-06, + "loss": 0.8127, + "num_input_tokens_seen": 43761750, + "step": 1341 + }, + { + "epoch": 0.12102628849709159, + "flos": 20887746894720.0, + "grad_norm": 1.7147992591039447, + "learning_rate": 3.913758738344851e-06, + "loss": 0.7855, + "num_input_tokens_seen": 43788180, + "step": 1342 + }, + { + "epoch": 0.12111647202056185, + "flos": 65378655750240.0, + "grad_norm": 0.730508446252286, + "learning_rate": 3.913588951937174e-06, + "loss": 0.6, + "num_input_tokens_seen": 43873345, + "step": 1343 + }, + { + "epoch": 0.12120665554403211, + "flos": 67920093405120.0, + "grad_norm": 0.7152368267356899, + "learning_rate": 3.9134190022515355e-06, + "loss": 0.6013, + "num_input_tokens_seen": 43961960, + "step": 1344 + }, + { + "epoch": 0.12129683906750237, + "flos": 20310668972640.0, + "grad_norm": 2.526629245232666, + "learning_rate": 3.913248889302438e-06, + "loss": 0.7563, + "num_input_tokens_seen": 43991560, + "step": 1345 + }, + { + "epoch": 0.12138702259097263, + "flos": 30618759804000.0, + "grad_norm": 1.7050326102038402, + "learning_rate": 3.913078613104395e-06, + "loss": 0.7321, + "num_input_tokens_seen": 44025320, + "step": 1346 + }, + { + "epoch": 0.12147720611444289, + "flos": 18780669518880.0, + "grad_norm": 2.60835812793521, + "learning_rate": 3.912908173671936e-06, + "loss": 0.6798, + "num_input_tokens_seen": 44052350, + "step": 1347 + }, + { + "epoch": 0.12156738963791315, + "flos": 23334681818880.0, + "grad_norm": 1.8871854279973088, + "learning_rate": 3.9127375710196044e-06, + "loss": 0.7258, + "num_input_tokens_seen": 44080130, + "step": 1348 + }, + { + "epoch": 0.12165757316138341, + "flos": 26358620325600.0, + "grad_norm": 1.798713075684306, + "learning_rate": 3.912566805161957e-06, + "loss": 0.7433, + "num_input_tokens_seen": 44109660, + "step": 1349 + }, + { + "epoch": 0.12174775668485367, + "flos": 19137296362560.0, + "grad_norm": 1.871892197264068, + "learning_rate": 3.912395876113564e-06, + "loss": 0.8155, + "num_input_tokens_seen": 44135555, + "step": 1350 + }, + { + "epoch": 0.12183794020832393, + "flos": 19618756461120.0, + "grad_norm": 1.6735960043844882, + "learning_rate": 3.912224783889009e-06, + "loss": 0.7997, + "num_input_tokens_seen": 44163430, + "step": 1351 + }, + { + "epoch": 0.12192812373179421, + "flos": 19720990501920.0, + "grad_norm": 1.77105889473277, + "learning_rate": 3.912053528502892e-06, + "loss": 0.8108, + "num_input_tokens_seen": 44188270, + "step": 1352 + }, + { + "epoch": 0.12201830725526447, + "flos": 43449623337120.0, + "grad_norm": 1.8881050529550156, + "learning_rate": 3.911882109969825e-06, + "loss": 0.7288, + "num_input_tokens_seen": 44225540, + "step": 1353 + }, + { + "epoch": 0.12210849077873473, + "flos": 19728647472480.0, + "grad_norm": 2.141623939157209, + "learning_rate": 3.911710528304435e-06, + "loss": 0.8068, + "num_input_tokens_seen": 44255165, + "step": 1354 + }, + { + "epoch": 0.12219867430220499, + "flos": 17906150478240.0, + "grad_norm": 2.433561811650875, + "learning_rate": 3.911538783521361e-06, + "loss": 0.7502, + "num_input_tokens_seen": 44282210, + "step": 1355 + }, + { + "epoch": 0.12228885782567525, + "flos": 25993481606880.0, + "grad_norm": 1.6724769720707469, + "learning_rate": 3.9113668756352575e-06, + "loss": 0.8098, + "num_input_tokens_seen": 44311225, + "step": 1356 + }, + { + "epoch": 0.12237904134914551, + "flos": 27995462870400.0, + "grad_norm": 1.5109894641986057, + "learning_rate": 3.911194804660793e-06, + "loss": 0.7244, + "num_input_tokens_seen": 44342735, + "step": 1357 + }, + { + "epoch": 0.12246922487261577, + "flos": 21694754184000.0, + "grad_norm": 1.7300033460154727, + "learning_rate": 3.91102257061265e-06, + "loss": 0.8802, + "num_input_tokens_seen": 44370620, + "step": 1358 + }, + { + "epoch": 0.12255940839608603, + "flos": 20891538210240.0, + "grad_norm": 1.5099082968194053, + "learning_rate": 3.910850173505524e-06, + "loss": 0.8374, + "num_input_tokens_seen": 44399555, + "step": 1359 + }, + { + "epoch": 0.1226495919195563, + "flos": 18521370307680.0, + "grad_norm": 2.931222416924876, + "learning_rate": 3.9106776133541255e-06, + "loss": 0.8723, + "num_input_tokens_seen": 44425645, + "step": 1360 + }, + { + "epoch": 0.12273977544302656, + "flos": 18483377079360.0, + "grad_norm": 1.5646705809680903, + "learning_rate": 3.9105048901731766e-06, + "loss": 0.7593, + "num_input_tokens_seen": 44451875, + "step": 1361 + }, + { + "epoch": 0.12282995896649682, + "flos": 25080002924160.0, + "grad_norm": 2.4937786894178147, + "learning_rate": 3.9103320039774165e-06, + "loss": 0.7896, + "num_input_tokens_seen": 44481580, + "step": 1362 + }, + { + "epoch": 0.12292014248996708, + "flos": 20996745831840.0, + "grad_norm": 1.6346923374169353, + "learning_rate": 3.9101589547815965e-06, + "loss": 0.8559, + "num_input_tokens_seen": 44509505, + "step": 1363 + }, + { + "epoch": 0.12301032601343735, + "flos": 22969357251360.0, + "grad_norm": 1.4144605213425956, + "learning_rate": 3.909985742600482e-06, + "loss": 0.7886, + "num_input_tokens_seen": 44538405, + "step": 1364 + }, + { + "epoch": 0.12310050953690761, + "flos": 32186938334880.0, + "grad_norm": 1.6130656913487198, + "learning_rate": 3.909812367448852e-06, + "loss": 0.7617, + "num_input_tokens_seen": 44568000, + "step": 1365 + }, + { + "epoch": 0.12319069306037787, + "flos": 25694628037440.0, + "grad_norm": 2.91404204578738, + "learning_rate": 3.909638829341501e-06, + "loss": 0.773, + "num_input_tokens_seen": 44591100, + "step": 1366 + }, + { + "epoch": 0.12328087658384813, + "flos": 21979148716800.0, + "grad_norm": 1.6916572789775777, + "learning_rate": 3.909465128293234e-06, + "loss": 0.8532, + "num_input_tokens_seen": 44618000, + "step": 1367 + }, + { + "epoch": 0.1233710601073184, + "flos": 22419865024800.0, + "grad_norm": 1.646546712217952, + "learning_rate": 3.9092912643188745e-06, + "loss": 0.8555, + "num_input_tokens_seen": 44648510, + "step": 1368 + }, + { + "epoch": 0.12346124363078866, + "flos": 23334495970080.0, + "grad_norm": 1.8916201533958334, + "learning_rate": 3.909117237433256e-06, + "loss": 0.7894, + "num_input_tokens_seen": 44675835, + "step": 1369 + }, + { + "epoch": 0.12355142715425892, + "flos": 21658693783200.0, + "grad_norm": 2.0833041252631896, + "learning_rate": 3.908943047651229e-06, + "loss": 0.8396, + "num_input_tokens_seen": 44703765, + "step": 1370 + }, + { + "epoch": 0.12364161067772918, + "flos": 21075297001920.0, + "grad_norm": 1.65090880074974, + "learning_rate": 3.908768694987655e-06, + "loss": 0.7847, + "num_input_tokens_seen": 44734660, + "step": 1371 + }, + { + "epoch": 0.12373179420119944, + "flos": 21654121902720.0, + "grad_norm": 2.6576855538902517, + "learning_rate": 3.908594179457411e-06, + "loss": 0.7207, + "num_input_tokens_seen": 44759515, + "step": 1372 + }, + { + "epoch": 0.1238219777246697, + "flos": 23619968425920.0, + "grad_norm": 2.194702824003117, + "learning_rate": 3.908419501075388e-06, + "loss": 0.8026, + "num_input_tokens_seen": 44788030, + "step": 1373 + }, + { + "epoch": 0.12391216124813996, + "flos": 28290153426720.0, + "grad_norm": 2.0147931652072066, + "learning_rate": 3.90824465985649e-06, + "loss": 0.87, + "num_input_tokens_seen": 44820615, + "step": 1374 + }, + { + "epoch": 0.12400234477161022, + "flos": 14991396757440.0, + "grad_norm": 1.8990207608149268, + "learning_rate": 3.908069655815636e-06, + "loss": 0.8398, + "num_input_tokens_seen": 44846450, + "step": 1375 + }, + { + "epoch": 0.1240925282950805, + "flos": 16193544495360.0, + "grad_norm": 2.110091250286037, + "learning_rate": 3.907894488967758e-06, + "loss": 0.7345, + "num_input_tokens_seen": 44872645, + "step": 1376 + }, + { + "epoch": 0.12418271181855076, + "flos": 22751396546880.0, + "grad_norm": 1.8314950642312267, + "learning_rate": 3.9077191593278005e-06, + "loss": 0.857, + "num_input_tokens_seen": 44900875, + "step": 1377 + }, + { + "epoch": 0.12427289534202102, + "flos": 67298926414080.0, + "grad_norm": 0.7134372653367621, + "learning_rate": 3.9075436669107265e-06, + "loss": 0.5773, + "num_input_tokens_seen": 44990410, + "step": 1378 + }, + { + "epoch": 0.12436307886549128, + "flos": 15173594419200.0, + "grad_norm": 1.7155098631532761, + "learning_rate": 3.90736801173151e-06, + "loss": 0.8064, + "num_input_tokens_seen": 45016875, + "step": 1379 + }, + { + "epoch": 0.12445326238896154, + "flos": 26504088530880.0, + "grad_norm": 1.523350092018704, + "learning_rate": 3.907192193805136e-06, + "loss": 0.8077, + "num_input_tokens_seen": 45047750, + "step": 1380 + }, + { + "epoch": 0.1245434459124318, + "flos": 25666447625760.0, + "grad_norm": 2.1797966605030243, + "learning_rate": 3.907016213146608e-06, + "loss": 0.849, + "num_input_tokens_seen": 45074830, + "step": 1381 + }, + { + "epoch": 0.12463362943590206, + "flos": 22750987679520.0, + "grad_norm": 1.8879789010137846, + "learning_rate": 3.906840069770942e-06, + "loss": 0.9099, + "num_input_tokens_seen": 45103020, + "step": 1382 + }, + { + "epoch": 0.12472381295937232, + "flos": 24317344892160.0, + "grad_norm": 1.5917323936086616, + "learning_rate": 3.906663763693167e-06, + "loss": 0.8292, + "num_input_tokens_seen": 45133210, + "step": 1383 + }, + { + "epoch": 0.12481399648284258, + "flos": 26099395453920.0, + "grad_norm": 1.7233004587285634, + "learning_rate": 3.906487294928327e-06, + "loss": 0.7785, + "num_input_tokens_seen": 45163740, + "step": 1384 + }, + { + "epoch": 0.12490418000631284, + "flos": 21729327794400.0, + "grad_norm": 1.851751454687133, + "learning_rate": 3.906310663491478e-06, + "loss": 0.8004, + "num_input_tokens_seen": 45191625, + "step": 1385 + }, + { + "epoch": 0.1249943635297831, + "flos": 21804347837280.0, + "grad_norm": 1.861986138035099, + "learning_rate": 3.906133869397692e-06, + "loss": 0.852, + "num_input_tokens_seen": 45218240, + "step": 1386 + }, + { + "epoch": 0.12508454705325336, + "flos": 35028827858880.0, + "grad_norm": 1.729927684507326, + "learning_rate": 3.905956912662054e-06, + "loss": 0.7667, + "num_input_tokens_seen": 45247410, + "step": 1387 + }, + { + "epoch": 0.12517473057672363, + "flos": 20126241125280.0, + "grad_norm": 2.1755989039389387, + "learning_rate": 3.905779793299662e-06, + "loss": 0.8625, + "num_input_tokens_seen": 45268790, + "step": 1388 + }, + { + "epoch": 0.12526491410019389, + "flos": 19472656369920.0, + "grad_norm": 1.8626494226105763, + "learning_rate": 3.905602511325631e-06, + "loss": 0.8319, + "num_input_tokens_seen": 45295905, + "step": 1389 + }, + { + "epoch": 0.12535509762366415, + "flos": 20492978143680.0, + "grad_norm": 1.6528731415677638, + "learning_rate": 3.905425066755086e-06, + "loss": 0.7602, + "num_input_tokens_seen": 45324400, + "step": 1390 + }, + { + "epoch": 0.1254452811471344, + "flos": 28505623757280.0, + "grad_norm": 8.005876766799236, + "learning_rate": 3.905247459603168e-06, + "loss": 0.7538, + "num_input_tokens_seen": 45356220, + "step": 1391 + }, + { + "epoch": 0.12553546467060467, + "flos": 24316415648160.0, + "grad_norm": 1.907638544373384, + "learning_rate": 3.905069689885031e-06, + "loss": 0.6804, + "num_input_tokens_seen": 45386375, + "step": 1392 + }, + { + "epoch": 0.12562564819407493, + "flos": 20456434536000.0, + "grad_norm": 1.6611648906321788, + "learning_rate": 3.904891757615843e-06, + "loss": 0.683, + "num_input_tokens_seen": 45415875, + "step": 1393 + }, + { + "epoch": 0.12571583171754522, + "flos": 69141718097280.0, + "grad_norm": 0.6304957756291206, + "learning_rate": 3.9047136628107874e-06, + "loss": 0.5629, + "num_input_tokens_seen": 45508200, + "step": 1394 + }, + { + "epoch": 0.12580601524101548, + "flos": 27518500312800.0, + "grad_norm": 1.759424309861077, + "learning_rate": 3.904535405485059e-06, + "loss": 0.9267, + "num_input_tokens_seen": 45537500, + "step": 1395 + }, + { + "epoch": 0.12589619876448574, + "flos": 24646683398400.0, + "grad_norm": 1.722392643765139, + "learning_rate": 3.90435698565387e-06, + "loss": 0.7472, + "num_input_tokens_seen": 45564920, + "step": 1396 + }, + { + "epoch": 0.125986382287956, + "flos": 39178890477120.0, + "grad_norm": 1.712085147366832, + "learning_rate": 3.904178403332441e-06, + "loss": 0.7612, + "num_input_tokens_seen": 45598040, + "step": 1397 + }, + { + "epoch": 0.12607656581142626, + "flos": 18926026214880.0, + "grad_norm": 2.041795674614328, + "learning_rate": 3.903999658536012e-06, + "loss": 0.8852, + "num_input_tokens_seen": 45624805, + "step": 1398 + }, + { + "epoch": 0.12616674933489652, + "flos": 24643561138560.0, + "grad_norm": 1.7540279762490203, + "learning_rate": 3.903820751279833e-06, + "loss": 0.7959, + "num_input_tokens_seen": 45654495, + "step": 1399 + }, + { + "epoch": 0.12625693285836678, + "flos": 17723655458400.0, + "grad_norm": 2.5337336094330354, + "learning_rate": 3.90364168157917e-06, + "loss": 0.833, + "num_input_tokens_seen": 45679910, + "step": 1400 + }, + { + "epoch": 0.12634711638183704, + "flos": 23006346896160.0, + "grad_norm": 2.0865933533770247, + "learning_rate": 3.903462449449302e-06, + "loss": 0.8304, + "num_input_tokens_seen": 45707615, + "step": 1401 + }, + { + "epoch": 0.1264372999053073, + "flos": 22605965511360.0, + "grad_norm": 1.5020947135391904, + "learning_rate": 3.903283054905522e-06, + "loss": 0.7688, + "num_input_tokens_seen": 45737435, + "step": 1402 + }, + { + "epoch": 0.12652748342877757, + "flos": 25812027340320.0, + "grad_norm": 1.8981363819071215, + "learning_rate": 3.9031034979631385e-06, + "loss": 0.708, + "num_input_tokens_seen": 45766840, + "step": 1403 + }, + { + "epoch": 0.12661766695224783, + "flos": 21654493600320.0, + "grad_norm": 1.5919781727899949, + "learning_rate": 3.902923778637469e-06, + "loss": 0.831, + "num_input_tokens_seen": 45794910, + "step": 1404 + }, + { + "epoch": 0.1267078504757181, + "flos": 29126196032160.0, + "grad_norm": 2.036608977408666, + "learning_rate": 3.902743896943852e-06, + "loss": 0.7297, + "num_input_tokens_seen": 45824230, + "step": 1405 + }, + { + "epoch": 0.12679803399918835, + "flos": 67781092738080.0, + "grad_norm": 0.7173128687502073, + "learning_rate": 3.902563852897633e-06, + "loss": 0.6319, + "num_input_tokens_seen": 45909875, + "step": 1406 + }, + { + "epoch": 0.1268882175226586, + "flos": 65404382957760.0, + "grad_norm": 0.6956210148466695, + "learning_rate": 3.9023836465141755e-06, + "loss": 0.6065, + "num_input_tokens_seen": 45998270, + "step": 1407 + }, + { + "epoch": 0.12697840104612887, + "flos": 23327916922560.0, + "grad_norm": 2.0285177782129193, + "learning_rate": 3.902203277808856e-06, + "loss": 0.8447, + "num_input_tokens_seen": 46023885, + "step": 1408 + }, + { + "epoch": 0.12706858456959913, + "flos": 21112695514080.0, + "grad_norm": 1.773772825560643, + "learning_rate": 3.902022746797064e-06, + "loss": 0.8536, + "num_input_tokens_seen": 46051155, + "step": 1409 + }, + { + "epoch": 0.1271587680930694, + "flos": 14371939575360.0, + "grad_norm": 3.0445767073043433, + "learning_rate": 3.9018420534942035e-06, + "loss": 0.9058, + "num_input_tokens_seen": 46076405, + "step": 1410 + }, + { + "epoch": 0.12724895161653965, + "flos": 23589148961280.0, + "grad_norm": 2.264531891304925, + "learning_rate": 3.9016611979156935e-06, + "loss": 0.7903, + "num_input_tokens_seen": 46105550, + "step": 1411 + }, + { + "epoch": 0.1273391351400099, + "flos": 24391026823680.0, + "grad_norm": 2.0568723484649305, + "learning_rate": 3.9014801800769635e-06, + "loss": 0.701, + "num_input_tokens_seen": 46133325, + "step": 1412 + }, + { + "epoch": 0.12742931866348017, + "flos": 15063926426400.0, + "grad_norm": 2.7662415760048193, + "learning_rate": 3.901298999993459e-06, + "loss": 0.8237, + "num_input_tokens_seen": 46157245, + "step": 1413 + }, + { + "epoch": 0.12751950218695043, + "flos": 13716050294880.0, + "grad_norm": 2.4759592952248037, + "learning_rate": 3.901117657680642e-06, + "loss": 0.728, + "num_input_tokens_seen": 46184445, + "step": 1414 + }, + { + "epoch": 0.1276096857104207, + "flos": 23735286222240.0, + "grad_norm": 2.725062091364972, + "learning_rate": 3.900936153153982e-06, + "loss": 0.7292, + "num_input_tokens_seen": 46211780, + "step": 1415 + }, + { + "epoch": 0.12769986923389096, + "flos": 68101659180960.0, + "grad_norm": 0.7147826705436181, + "learning_rate": 3.900754486428968e-06, + "loss": 0.5589, + "num_input_tokens_seen": 46294230, + "step": 1416 + }, + { + "epoch": 0.12779005275736122, + "flos": 23188953425280.0, + "grad_norm": 3.204455885382463, + "learning_rate": 3.900572657521102e-06, + "loss": 0.8848, + "num_input_tokens_seen": 46320940, + "step": 1417 + }, + { + "epoch": 0.1278802362808315, + "flos": 23004339729120.0, + "grad_norm": 1.4928719114380078, + "learning_rate": 3.900390666445896e-06, + "loss": 0.7809, + "num_input_tokens_seen": 46351370, + "step": 1418 + }, + { + "epoch": 0.12797041980430177, + "flos": 22820357918880.0, + "grad_norm": 1.8956488075861733, + "learning_rate": 3.9002085132188795e-06, + "loss": 0.764, + "num_input_tokens_seen": 46380080, + "step": 1419 + }, + { + "epoch": 0.12806060332777203, + "flos": 22496594876640.0, + "grad_norm": 1.7448954385292395, + "learning_rate": 3.9000261978555964e-06, + "loss": 0.8409, + "num_input_tokens_seen": 46409330, + "step": 1420 + }, + { + "epoch": 0.1281507868512423, + "flos": 28321418928480.0, + "grad_norm": 2.203860004728221, + "learning_rate": 3.8998437203716e-06, + "loss": 0.8634, + "num_input_tokens_seen": 46438655, + "step": 1421 + }, + { + "epoch": 0.12824097037471255, + "flos": 23444349811680.0, + "grad_norm": 1.893110737688627, + "learning_rate": 3.899661080782462e-06, + "loss": 0.8226, + "num_input_tokens_seen": 46467260, + "step": 1422 + }, + { + "epoch": 0.1283311538981828, + "flos": 20493015313440.0, + "grad_norm": 1.9438900936949488, + "learning_rate": 3.899478279103767e-06, + "loss": 0.8258, + "num_input_tokens_seen": 46495270, + "step": 1423 + }, + { + "epoch": 0.12842133742165307, + "flos": 23662087497600.0, + "grad_norm": 1.9891978251532965, + "learning_rate": 3.8992953153511105e-06, + "loss": 0.8256, + "num_input_tokens_seen": 46524965, + "step": 1424 + }, + { + "epoch": 0.12851152094512333, + "flos": 22310345711040.0, + "grad_norm": 1.6732027717330267, + "learning_rate": 3.899112189540106e-06, + "loss": 0.8619, + "num_input_tokens_seen": 46551900, + "step": 1425 + }, + { + "epoch": 0.1286017044685936, + "flos": 25265025487680.0, + "grad_norm": 2.4181649616850764, + "learning_rate": 3.898928901686377e-06, + "loss": 0.83, + "num_input_tokens_seen": 46578030, + "step": 1426 + }, + { + "epoch": 0.12869188799206385, + "flos": 60770481376320.0, + "grad_norm": 0.7160676162082797, + "learning_rate": 3.898745451805564e-06, + "loss": 0.5642, + "num_input_tokens_seen": 46667255, + "step": 1427 + }, + { + "epoch": 0.1287820715155341, + "flos": 25883479086240.0, + "grad_norm": 1.6725468934331953, + "learning_rate": 3.898561839913319e-06, + "loss": 0.7344, + "num_input_tokens_seen": 46700005, + "step": 1428 + }, + { + "epoch": 0.12887225503900437, + "flos": 22202907903840.0, + "grad_norm": 2.5856608908423473, + "learning_rate": 3.89837806602531e-06, + "loss": 0.6443, + "num_input_tokens_seen": 46728515, + "step": 1429 + }, + { + "epoch": 0.12896243856247463, + "flos": 19654705352640.0, + "grad_norm": 4.8085451596219855, + "learning_rate": 3.898194130157217e-06, + "loss": 0.807, + "num_input_tokens_seen": 46757275, + "step": 1430 + }, + { + "epoch": 0.1290526220859449, + "flos": 14481756247200.0, + "grad_norm": 2.2976123897720506, + "learning_rate": 3.8980100323247335e-06, + "loss": 0.7399, + "num_input_tokens_seen": 46783475, + "step": 1431 + }, + { + "epoch": 0.12914280560941516, + "flos": 22132199553120.0, + "grad_norm": 1.9639054289461073, + "learning_rate": 3.897825772543568e-06, + "loss": 0.73, + "num_input_tokens_seen": 46811195, + "step": 1432 + }, + { + "epoch": 0.12923298913288542, + "flos": 20091853363680.0, + "grad_norm": 2.991056462557508, + "learning_rate": 3.897641350829444e-06, + "loss": 0.8526, + "num_input_tokens_seen": 46838050, + "step": 1433 + }, + { + "epoch": 0.12932317265635568, + "flos": 22532357919360.0, + "grad_norm": 1.871607763609628, + "learning_rate": 3.897456767198096e-06, + "loss": 0.884, + "num_input_tokens_seen": 46866725, + "step": 1434 + }, + { + "epoch": 0.12941335617982594, + "flos": 34264162659840.0, + "grad_norm": 1.767224288004369, + "learning_rate": 3.897272021665275e-06, + "loss": 0.7367, + "num_input_tokens_seen": 46897635, + "step": 1435 + }, + { + "epoch": 0.1295035397032962, + "flos": 24791519717760.0, + "grad_norm": 1.7806684648987205, + "learning_rate": 3.897087114246743e-06, + "loss": 0.9098, + "num_input_tokens_seen": 46927650, + "step": 1436 + }, + { + "epoch": 0.12959372322676646, + "flos": 20383310150880.0, + "grad_norm": 5.193055636449673, + "learning_rate": 3.896902044958279e-06, + "loss": 0.9128, + "num_input_tokens_seen": 46955170, + "step": 1437 + }, + { + "epoch": 0.12968390675023672, + "flos": 23006235386880.0, + "grad_norm": 2.453661234372823, + "learning_rate": 3.896716813815672e-06, + "loss": 0.8518, + "num_input_tokens_seen": 46980820, + "step": 1438 + }, + { + "epoch": 0.12977409027370698, + "flos": 19067591595360.0, + "grad_norm": 2.3802848041250817, + "learning_rate": 3.896531420834728e-06, + "loss": 0.7468, + "num_input_tokens_seen": 47009715, + "step": 1439 + }, + { + "epoch": 0.12986427379717724, + "flos": 43441892027040.0, + "grad_norm": 1.781009523990618, + "learning_rate": 3.896345866031266e-06, + "loss": 0.711, + "num_input_tokens_seen": 47046045, + "step": 1440 + }, + { + "epoch": 0.1299544573206475, + "flos": 28244912095200.0, + "grad_norm": 1.7991587182853417, + "learning_rate": 3.896160149421119e-06, + "loss": 0.8622, + "num_input_tokens_seen": 47074685, + "step": 1441 + }, + { + "epoch": 0.1300446408441178, + "flos": 16995273678720.0, + "grad_norm": 1.8498428020189706, + "learning_rate": 3.8959742710201314e-06, + "loss": 0.8699, + "num_input_tokens_seen": 47098445, + "step": 1442 + }, + { + "epoch": 0.13013482436758805, + "flos": 25957049508480.0, + "grad_norm": 2.5174393833897466, + "learning_rate": 3.895788230844166e-06, + "loss": 0.6955, + "num_input_tokens_seen": 47124115, + "step": 1443 + }, + { + "epoch": 0.13022500789105831, + "flos": 29235269308800.0, + "grad_norm": 2.031680134933847, + "learning_rate": 3.895602028909095e-06, + "loss": 0.9556, + "num_input_tokens_seen": 47152995, + "step": 1444 + }, + { + "epoch": 0.13031519141452858, + "flos": 69645708804000.0, + "grad_norm": 0.6337786939622189, + "learning_rate": 3.895415665230807e-06, + "loss": 0.5969, + "num_input_tokens_seen": 47250825, + "step": 1445 + }, + { + "epoch": 0.13040537493799884, + "flos": 25080077263680.0, + "grad_norm": 1.703480464692289, + "learning_rate": 3.895229139825203e-06, + "loss": 0.8236, + "num_input_tokens_seen": 47278510, + "step": 1446 + }, + { + "epoch": 0.1304955584614691, + "flos": 29819558164320.0, + "grad_norm": 1.905130839522917, + "learning_rate": 3.895042452708198e-06, + "loss": 0.7104, + "num_input_tokens_seen": 47307640, + "step": 1447 + }, + { + "epoch": 0.13058574198493936, + "flos": 23002964448000.0, + "grad_norm": 1.9708719474230278, + "learning_rate": 3.894855603895723e-06, + "loss": 0.7083, + "num_input_tokens_seen": 47336490, + "step": 1448 + }, + { + "epoch": 0.13067592550840962, + "flos": 23116052058720.0, + "grad_norm": 1.8560939016758766, + "learning_rate": 3.894668593403718e-06, + "loss": 0.8785, + "num_input_tokens_seen": 47365030, + "step": 1449 + }, + { + "epoch": 0.13076610903187988, + "flos": 27229087862400.0, + "grad_norm": 2.0253885061313652, + "learning_rate": 3.8944814212481425e-06, + "loss": 0.8061, + "num_input_tokens_seen": 47393990, + "step": 1450 + }, + { + "epoch": 0.13085629255535014, + "flos": 18452520444960.0, + "grad_norm": 1.7846477293440246, + "learning_rate": 3.894294087444966e-06, + "loss": 0.7793, + "num_input_tokens_seen": 47421110, + "step": 1451 + }, + { + "epoch": 0.1309464760788204, + "flos": 34476064693440.0, + "grad_norm": 1.7209452394886728, + "learning_rate": 3.894106592010173e-06, + "loss": 0.689, + "num_input_tokens_seen": 47450085, + "step": 1452 + }, + { + "epoch": 0.13103665960229066, + "flos": 26467321904640.0, + "grad_norm": 1.9490879909407486, + "learning_rate": 3.893918934959762e-06, + "loss": 0.6735, + "num_input_tokens_seen": 47478910, + "step": 1453 + }, + { + "epoch": 0.13112684312576092, + "flos": 26830899493440.0, + "grad_norm": 2.127105231052509, + "learning_rate": 3.893731116309743e-06, + "loss": 0.7361, + "num_input_tokens_seen": 47504115, + "step": 1454 + }, + { + "epoch": 0.13121702664923118, + "flos": 32369805052320.0, + "grad_norm": 2.0697629955080514, + "learning_rate": 3.893543136076145e-06, + "loss": 0.8098, + "num_input_tokens_seen": 47534010, + "step": 1455 + }, + { + "epoch": 0.13130721017270144, + "flos": 16625302891200.0, + "grad_norm": 3.8133322635951465, + "learning_rate": 3.893354994275006e-06, + "loss": 0.8253, + "num_input_tokens_seen": 47558395, + "step": 1456 + }, + { + "epoch": 0.1313973936961717, + "flos": 24063546805440.0, + "grad_norm": 1.6512248261559272, + "learning_rate": 3.893166690922378e-06, + "loss": 0.767, + "num_input_tokens_seen": 47587150, + "step": 1457 + }, + { + "epoch": 0.13148757721964197, + "flos": 36231570312960.0, + "grad_norm": 2.207678559837386, + "learning_rate": 3.892978226034329e-06, + "loss": 0.8272, + "num_input_tokens_seen": 47617740, + "step": 1458 + }, + { + "epoch": 0.13157776074311223, + "flos": 21221694451200.0, + "grad_norm": 2.9751493849233626, + "learning_rate": 3.89278959962694e-06, + "loss": 0.8497, + "num_input_tokens_seen": 47644490, + "step": 1459 + }, + { + "epoch": 0.1316679442665825, + "flos": 48909494519040.0, + "grad_norm": 1.682234779868434, + "learning_rate": 3.8926008117163056e-06, + "loss": 0.7729, + "num_input_tokens_seen": 47677000, + "step": 1460 + }, + { + "epoch": 0.13175812779005275, + "flos": 26279102741760.0, + "grad_norm": 2.039672643351112, + "learning_rate": 3.892411862318535e-06, + "loss": 0.7862, + "num_input_tokens_seen": 47708135, + "step": 1461 + }, + { + "epoch": 0.131848311313523, + "flos": 21037935659520.0, + "grad_norm": 1.5167248068484516, + "learning_rate": 3.892222751449749e-06, + "loss": 0.8213, + "num_input_tokens_seen": 47735115, + "step": 1462 + }, + { + "epoch": 0.13193849483699327, + "flos": 19181088073440.0, + "grad_norm": 1.866335175519387, + "learning_rate": 3.892033479126084e-06, + "loss": 0.7446, + "num_input_tokens_seen": 47761565, + "step": 1463 + }, + { + "epoch": 0.13202867836046353, + "flos": 20164791900000.0, + "grad_norm": 2.535856975938705, + "learning_rate": 3.891844045363691e-06, + "loss": 0.7809, + "num_input_tokens_seen": 47789590, + "step": 1464 + }, + { + "epoch": 0.13211886188393382, + "flos": 21221025395520.0, + "grad_norm": 2.1549522807066523, + "learning_rate": 3.891654450178732e-06, + "loss": 0.7217, + "num_input_tokens_seen": 47817300, + "step": 1465 + }, + { + "epoch": 0.13220904540740408, + "flos": 20675435993760.0, + "grad_norm": 1.7323361346542867, + "learning_rate": 3.891464693587385e-06, + "loss": 0.8444, + "num_input_tokens_seen": 47846800, + "step": 1466 + }, + { + "epoch": 0.13229922893087434, + "flos": 27558872405760.0, + "grad_norm": 1.838334551880272, + "learning_rate": 3.89127477560584e-06, + "loss": 0.8369, + "num_input_tokens_seen": 47878320, + "step": 1467 + }, + { + "epoch": 0.1323894124543446, + "flos": 23407434506400.0, + "grad_norm": 1.7155777526908986, + "learning_rate": 3.891084696250304e-06, + "loss": 0.8359, + "num_input_tokens_seen": 47908895, + "step": 1468 + }, + { + "epoch": 0.13247959597781486, + "flos": 22788237512640.0, + "grad_norm": 1.8211148830711201, + "learning_rate": 3.890894455536993e-06, + "loss": 0.8374, + "num_input_tokens_seen": 47936830, + "step": 1469 + }, + { + "epoch": 0.13256977950128512, + "flos": 16442622022560.0, + "grad_norm": 2.669206145208286, + "learning_rate": 3.890704053482142e-06, + "loss": 0.8455, + "num_input_tokens_seen": 47959710, + "step": 1470 + }, + { + "epoch": 0.13265996302475538, + "flos": 22860767181600.0, + "grad_norm": 2.2110298056592708, + "learning_rate": 3.890513490101995e-06, + "loss": 0.7455, + "num_input_tokens_seen": 47987645, + "step": 1471 + }, + { + "epoch": 0.13275014654822564, + "flos": 65467763132160.0, + "grad_norm": 0.6217717482763064, + "learning_rate": 3.890322765412814e-06, + "loss": 0.6064, + "num_input_tokens_seen": 48086345, + "step": 1472 + }, + { + "epoch": 0.1328403300716959, + "flos": 25078739152320.0, + "grad_norm": 1.4154395885181772, + "learning_rate": 3.890131879430871e-06, + "loss": 0.8274, + "num_input_tokens_seen": 48115460, + "step": 1473 + }, + { + "epoch": 0.13293051359516617, + "flos": 22349751390240.0, + "grad_norm": 1.6394293651808725, + "learning_rate": 3.889940832172454e-06, + "loss": 0.8244, + "num_input_tokens_seen": 48143860, + "step": 1474 + }, + { + "epoch": 0.13302069711863643, + "flos": 18884055822240.0, + "grad_norm": 1.724443542180093, + "learning_rate": 3.889749623653864e-06, + "loss": 0.845, + "num_input_tokens_seen": 48169175, + "step": 1475 + }, + { + "epoch": 0.1331108806421067, + "flos": 21149053272960.0, + "grad_norm": 1.5413667885494555, + "learning_rate": 3.889558253891416e-06, + "loss": 0.7432, + "num_input_tokens_seen": 48198415, + "step": 1476 + }, + { + "epoch": 0.13320106416557695, + "flos": 67234128055200.0, + "grad_norm": 0.6227250658248039, + "learning_rate": 3.8893667229014385e-06, + "loss": 0.581, + "num_input_tokens_seen": 48296075, + "step": 1477 + }, + { + "epoch": 0.1332912476890472, + "flos": 23625395210880.0, + "grad_norm": 2.262599951290837, + "learning_rate": 3.8891750307002746e-06, + "loss": 0.7069, + "num_input_tokens_seen": 48325050, + "step": 1478 + }, + { + "epoch": 0.13338143121251747, + "flos": 28727784644640.0, + "grad_norm": 1.6317144470830238, + "learning_rate": 3.888983177304281e-06, + "loss": 0.73, + "num_input_tokens_seen": 48359000, + "step": 1479 + }, + { + "epoch": 0.13347161473598773, + "flos": 56377362732000.0, + "grad_norm": 0.7487607094790092, + "learning_rate": 3.888791162729826e-06, + "loss": 0.5802, + "num_input_tokens_seen": 48441340, + "step": 1480 + }, + { + "epoch": 0.133561798259458, + "flos": 25301085888480.0, + "grad_norm": 2.170132164160588, + "learning_rate": 3.888598986993295e-06, + "loss": 0.8204, + "num_input_tokens_seen": 48469400, + "step": 1481 + }, + { + "epoch": 0.13365198178292825, + "flos": 23152186799040.0, + "grad_norm": 1.848714126434887, + "learning_rate": 3.888406650111085e-06, + "loss": 0.8236, + "num_input_tokens_seen": 48501060, + "step": 1482 + }, + { + "epoch": 0.1337421653063985, + "flos": 26426801132640.0, + "grad_norm": 2.4229834204345444, + "learning_rate": 3.888214152099607e-06, + "loss": 0.8052, + "num_input_tokens_seen": 48529225, + "step": 1483 + }, + { + "epoch": 0.13383234882986877, + "flos": 10939367997120.0, + "grad_norm": 2.1582006946394685, + "learning_rate": 3.888021492975285e-06, + "loss": 0.8716, + "num_input_tokens_seen": 48552275, + "step": 1484 + }, + { + "epoch": 0.13392253235333904, + "flos": 20018951997120.0, + "grad_norm": 1.852183435035597, + "learning_rate": 3.88782867275456e-06, + "loss": 0.851, + "num_input_tokens_seen": 48579105, + "step": 1485 + }, + { + "epoch": 0.1340127158768093, + "flos": 33278674684800.0, + "grad_norm": 1.4571022650463565, + "learning_rate": 3.8876356914538824e-06, + "loss": 0.762, + "num_input_tokens_seen": 48611450, + "step": 1486 + }, + { + "epoch": 0.13410289940027956, + "flos": 55206820757280.0, + "grad_norm": 2.2452187069261544, + "learning_rate": 3.88744254908972e-06, + "loss": 0.753, + "num_input_tokens_seen": 48645345, + "step": 1487 + }, + { + "epoch": 0.13419308292374982, + "flos": 19946013460800.0, + "grad_norm": 2.462165550330665, + "learning_rate": 3.887249245678552e-06, + "loss": 0.7684, + "num_input_tokens_seen": 48673830, + "step": 1488 + }, + { + "epoch": 0.1342832664472201, + "flos": 23152855854720.0, + "grad_norm": 1.7857911297749072, + "learning_rate": 3.887055781236872e-06, + "loss": 0.8722, + "num_input_tokens_seen": 48703925, + "step": 1489 + }, + { + "epoch": 0.13437344997069037, + "flos": 28544025852960.0, + "grad_norm": 1.8196445926073452, + "learning_rate": 3.886862155781186e-06, + "loss": 0.6706, + "num_input_tokens_seen": 48734035, + "step": 1490 + }, + { + "epoch": 0.13446363349416063, + "flos": 68320660638720.0, + "grad_norm": 0.6692996680632344, + "learning_rate": 3.886668369328019e-06, + "loss": 0.6022, + "num_input_tokens_seen": 48823460, + "step": 1491 + }, + { + "epoch": 0.1345538170176309, + "flos": 24209312368800.0, + "grad_norm": 2.3335027860237, + "learning_rate": 3.886474421893904e-06, + "loss": 0.8256, + "num_input_tokens_seen": 48853660, + "step": 1492 + }, + { + "epoch": 0.13464400054110115, + "flos": 27051462081120.0, + "grad_norm": 1.9396770920373607, + "learning_rate": 3.886280313495388e-06, + "loss": 0.7214, + "num_input_tokens_seen": 48885950, + "step": 1493 + }, + { + "epoch": 0.1347341840645714, + "flos": 23589297640320.0, + "grad_norm": 2.0803613317694203, + "learning_rate": 3.886086044149035e-06, + "loss": 0.7667, + "num_input_tokens_seen": 48915010, + "step": 1494 + }, + { + "epoch": 0.13482436758804167, + "flos": 25446033717120.0, + "grad_norm": 1.4415399752977611, + "learning_rate": 3.885891613871421e-06, + "loss": 0.8049, + "num_input_tokens_seen": 48945080, + "step": 1495 + }, + { + "epoch": 0.13491455111151193, + "flos": 27342175473120.0, + "grad_norm": 2.3611088123185677, + "learning_rate": 3.885697022679136e-06, + "loss": 0.8412, + "num_input_tokens_seen": 48974155, + "step": 1496 + }, + { + "epoch": 0.1350047346349822, + "flos": 17432049992160.0, + "grad_norm": 1.8358699055666785, + "learning_rate": 3.885502270588784e-06, + "loss": 0.8255, + "num_input_tokens_seen": 49000390, + "step": 1497 + }, + { + "epoch": 0.13509491815845245, + "flos": 17064160711200.0, + "grad_norm": 2.0290739875114596, + "learning_rate": 3.885307357616981e-06, + "loss": 0.7878, + "num_input_tokens_seen": 49026070, + "step": 1498 + }, + { + "epoch": 0.13518510168192271, + "flos": 27815755582560.0, + "grad_norm": 1.7433088995763317, + "learning_rate": 3.885112283780359e-06, + "loss": 0.7941, + "num_input_tokens_seen": 49054860, + "step": 1499 + }, + { + "epoch": 0.13527528520539298, + "flos": 22022605899840.0, + "grad_norm": 1.6841630083912953, + "learning_rate": 3.8849170490955624e-06, + "loss": 0.6881, + "num_input_tokens_seen": 49083920, + "step": 1500 + }, + { + "epoch": 0.13536546872886324, + "flos": 30439052516160.0, + "grad_norm": 2.530880936455558, + "learning_rate": 3.88472165357925e-06, + "loss": 0.7905, + "num_input_tokens_seen": 49112265, + "step": 1501 + }, + { + "epoch": 0.1354556522523335, + "flos": 21148384217280.0, + "grad_norm": 1.8180580422145052, + "learning_rate": 3.884526097248093e-06, + "loss": 0.7509, + "num_input_tokens_seen": 49141760, + "step": 1502 + }, + { + "epoch": 0.13554583577580376, + "flos": 20018654639040.0, + "grad_norm": 1.9794701460677764, + "learning_rate": 3.884330380118779e-06, + "loss": 0.8166, + "num_input_tokens_seen": 49168645, + "step": 1503 + }, + { + "epoch": 0.13563601929927402, + "flos": 16558162837440.0, + "grad_norm": 2.3360342560699374, + "learning_rate": 3.884134502208007e-06, + "loss": 0.7488, + "num_input_tokens_seen": 49194775, + "step": 1504 + }, + { + "epoch": 0.13572620282274428, + "flos": 27705938910720.0, + "grad_norm": 1.9363761842022345, + "learning_rate": 3.88393846353249e-06, + "loss": 0.9034, + "num_input_tokens_seen": 49220760, + "step": 1505 + }, + { + "epoch": 0.13581638634621454, + "flos": 20125460560320.0, + "grad_norm": 1.6884980252756778, + "learning_rate": 3.883742264108955e-06, + "loss": 0.8487, + "num_input_tokens_seen": 49247950, + "step": 1506 + }, + { + "epoch": 0.1359065698696848, + "flos": 20565916680000.0, + "grad_norm": 2.0158541053857144, + "learning_rate": 3.883545903954145e-06, + "loss": 0.8035, + "num_input_tokens_seen": 49273695, + "step": 1507 + }, + { + "epoch": 0.13599675339315506, + "flos": 21914127339360.0, + "grad_norm": 1.6112539933231371, + "learning_rate": 3.883349383084811e-06, + "loss": 0.7892, + "num_input_tokens_seen": 49301025, + "step": 1508 + }, + { + "epoch": 0.13608693691662532, + "flos": 13274813610240.0, + "grad_norm": 2.0829917882060616, + "learning_rate": 3.883152701517723e-06, + "loss": 0.7169, + "num_input_tokens_seen": 49328090, + "step": 1509 + }, + { + "epoch": 0.13617712044009558, + "flos": 17140890563040.0, + "grad_norm": 1.7993976962853355, + "learning_rate": 3.882955859269664e-06, + "loss": 0.7859, + "num_input_tokens_seen": 49352945, + "step": 1510 + }, + { + "epoch": 0.13626730396356584, + "flos": 20966298064800.0, + "grad_norm": 1.7336516375372804, + "learning_rate": 3.882758856357428e-06, + "loss": 0.802, + "num_input_tokens_seen": 49382910, + "step": 1511 + }, + { + "epoch": 0.1363574874870361, + "flos": 15128873464320.0, + "grad_norm": 2.1286833760200783, + "learning_rate": 3.882561692797824e-06, + "loss": 0.7887, + "num_input_tokens_seen": 49405775, + "step": 1512 + }, + { + "epoch": 0.1364476710105064, + "flos": 24791928585120.0, + "grad_norm": 2.027129115979133, + "learning_rate": 3.882364368607677e-06, + "loss": 0.8021, + "num_input_tokens_seen": 49434540, + "step": 1513 + }, + { + "epoch": 0.13653785453397665, + "flos": 17464170398400.0, + "grad_norm": 1.9941168922857775, + "learning_rate": 3.8821668838038225e-06, + "loss": 0.8098, + "num_input_tokens_seen": 49460090, + "step": 1514 + }, + { + "epoch": 0.13662803805744692, + "flos": 18379358890080.0, + "grad_norm": 2.3055858865399923, + "learning_rate": 3.881969238403111e-06, + "loss": 0.7104, + "num_input_tokens_seen": 49486535, + "step": 1515 + }, + { + "epoch": 0.13671822158091718, + "flos": 13242767543520.0, + "grad_norm": 2.2326567118322074, + "learning_rate": 3.881771432422408e-06, + "loss": 0.9226, + "num_input_tokens_seen": 49512080, + "step": 1516 + }, + { + "epoch": 0.13680840510438744, + "flos": 20019100676160.0, + "grad_norm": 2.186965005030382, + "learning_rate": 3.88157346587859e-06, + "loss": 0.8515, + "num_input_tokens_seen": 49539060, + "step": 1517 + }, + { + "epoch": 0.1368985886278577, + "flos": 20157655306080.0, + "grad_norm": 1.9478839592729142, + "learning_rate": 3.881375338788549e-06, + "loss": 0.7745, + "num_input_tokens_seen": 49567405, + "step": 1518 + }, + { + "epoch": 0.13698877215132796, + "flos": 17359631832480.0, + "grad_norm": 3.4996392200354247, + "learning_rate": 3.88117705116919e-06, + "loss": 0.8071, + "num_input_tokens_seen": 49593625, + "step": 1519 + }, + { + "epoch": 0.13707895567479822, + "flos": 69079899052800.0, + "grad_norm": 0.82220154283769, + "learning_rate": 3.880978603037432e-06, + "loss": 0.6558, + "num_input_tokens_seen": 49675420, + "step": 1520 + }, + { + "epoch": 0.13716913919826848, + "flos": 26649519566400.0, + "grad_norm": 1.9604959788749414, + "learning_rate": 3.880779994410209e-06, + "loss": 0.7977, + "num_input_tokens_seen": 49705765, + "step": 1521 + }, + { + "epoch": 0.13725932272173874, + "flos": 24135816286080.0, + "grad_norm": 2.941194848395881, + "learning_rate": 3.880581225304466e-06, + "loss": 0.6727, + "num_input_tokens_seen": 49733900, + "step": 1522 + }, + { + "epoch": 0.137349506245209, + "flos": 21767321022720.0, + "grad_norm": 2.2226946514010195, + "learning_rate": 3.880382295737163e-06, + "loss": 0.8237, + "num_input_tokens_seen": 49762640, + "step": 1523 + }, + { + "epoch": 0.13743968976867926, + "flos": 16371021597600.0, + "grad_norm": 2.464360451206698, + "learning_rate": 3.880183205725274e-06, + "loss": 0.7679, + "num_input_tokens_seen": 49788450, + "step": 1524 + }, + { + "epoch": 0.13752987329214952, + "flos": 17323199734080.0, + "grad_norm": 2.195054334346009, + "learning_rate": 3.879983955285788e-06, + "loss": 0.7364, + "num_input_tokens_seen": 49815670, + "step": 1525 + }, + { + "epoch": 0.13762005681561978, + "flos": 25629532320480.0, + "grad_norm": 2.017568080467192, + "learning_rate": 3.879784544435703e-06, + "loss": 0.7684, + "num_input_tokens_seen": 49841475, + "step": 1526 + }, + { + "epoch": 0.13771024033909005, + "flos": 20528889865440.0, + "grad_norm": 2.027217677317971, + "learning_rate": 3.879584973192037e-06, + "loss": 0.7351, + "num_input_tokens_seen": 49869790, + "step": 1527 + }, + { + "epoch": 0.1378004238625603, + "flos": 13971669699840.0, + "grad_norm": 3.359755241698218, + "learning_rate": 3.8793852415718165e-06, + "loss": 0.7245, + "num_input_tokens_seen": 49897895, + "step": 1528 + }, + { + "epoch": 0.13789060738603057, + "flos": 19181050903680.0, + "grad_norm": 3.1278145894274108, + "learning_rate": 3.879185349592085e-06, + "loss": 0.8029, + "num_input_tokens_seen": 49924735, + "step": 1529 + }, + { + "epoch": 0.13798079090950083, + "flos": 19180344678240.0, + "grad_norm": 4.4923286747517785, + "learning_rate": 3.878985297269897e-06, + "loss": 0.8184, + "num_input_tokens_seen": 49954380, + "step": 1530 + }, + { + "epoch": 0.1380709744329711, + "flos": 24536532198720.0, + "grad_norm": 1.8425261104602184, + "learning_rate": 3.878785084622323e-06, + "loss": 0.8301, + "num_input_tokens_seen": 49982190, + "step": 1531 + }, + { + "epoch": 0.13816115795644135, + "flos": 14773027185600.0, + "grad_norm": 3.5001141665775917, + "learning_rate": 3.878584711666447e-06, + "loss": 0.9802, + "num_input_tokens_seen": 50007690, + "step": 1532 + }, + { + "epoch": 0.1382513414799116, + "flos": 20820941368800.0, + "grad_norm": 2.0391035404679134, + "learning_rate": 3.8783841784193635e-06, + "loss": 0.8136, + "num_input_tokens_seen": 50034230, + "step": 1533 + }, + { + "epoch": 0.13834152500338187, + "flos": 25630089866880.0, + "grad_norm": 2.4005713895270477, + "learning_rate": 3.8781834848981855e-06, + "loss": 0.7769, + "num_input_tokens_seen": 50065190, + "step": 1534 + }, + { + "epoch": 0.13843170852685213, + "flos": 28611760622880.0, + "grad_norm": 1.9487494322241934, + "learning_rate": 3.877982631120037e-06, + "loss": 0.7285, + "num_input_tokens_seen": 50097590, + "step": 1535 + }, + { + "epoch": 0.1385218920503224, + "flos": 14255543856000.0, + "grad_norm": 2.6381957487892067, + "learning_rate": 3.877781617102053e-06, + "loss": 0.8557, + "num_input_tokens_seen": 50120915, + "step": 1536 + }, + { + "epoch": 0.13861207557379268, + "flos": 26100919414080.0, + "grad_norm": 1.792810902680418, + "learning_rate": 3.877580442861389e-06, + "loss": 0.8318, + "num_input_tokens_seen": 50150900, + "step": 1537 + }, + { + "epoch": 0.13870225909726294, + "flos": 16989103498560.0, + "grad_norm": 1.8477523165804297, + "learning_rate": 3.877379108415209e-06, + "loss": 0.8204, + "num_input_tokens_seen": 50176545, + "step": 1538 + }, + { + "epoch": 0.1387924426207332, + "flos": 21659288499360.0, + "grad_norm": 1.6234134117887413, + "learning_rate": 3.8771776137806915e-06, + "loss": 0.8254, + "num_input_tokens_seen": 50205975, + "step": 1539 + }, + { + "epoch": 0.13888262614420346, + "flos": 24427087224480.0, + "grad_norm": 1.970527391566145, + "learning_rate": 3.8769759589750295e-06, + "loss": 0.7401, + "num_input_tokens_seen": 50235030, + "step": 1540 + }, + { + "epoch": 0.13897280966767372, + "flos": 21440472890400.0, + "grad_norm": 1.6945457306626117, + "learning_rate": 3.876774144015429e-06, + "loss": 0.8039, + "num_input_tokens_seen": 50262780, + "step": 1541 + }, + { + "epoch": 0.13906299319114399, + "flos": 26977594300800.0, + "grad_norm": 11.496350580082197, + "learning_rate": 3.87657216891911e-06, + "loss": 0.8337, + "num_input_tokens_seen": 50289590, + "step": 1542 + }, + { + "epoch": 0.13915317671461425, + "flos": 19215289986240.0, + "grad_norm": 1.9497956242198327, + "learning_rate": 3.876370033703307e-06, + "loss": 0.7961, + "num_input_tokens_seen": 50315610, + "step": 1543 + }, + { + "epoch": 0.1392433602380845, + "flos": 29784241158720.0, + "grad_norm": 1.5388645877518417, + "learning_rate": 3.876167738385265e-06, + "loss": 0.7025, + "num_input_tokens_seen": 50347750, + "step": 1544 + }, + { + "epoch": 0.13933354376155477, + "flos": 20346171827040.0, + "grad_norm": 1.8260508451287334, + "learning_rate": 3.875965282982247e-06, + "loss": 0.8452, + "num_input_tokens_seen": 50373415, + "step": 1545 + }, + { + "epoch": 0.13942372728502503, + "flos": 27997470037440.0, + "grad_norm": 1.950188038391698, + "learning_rate": 3.875762667511528e-06, + "loss": 0.8144, + "num_input_tokens_seen": 50404205, + "step": 1546 + }, + { + "epoch": 0.1395139108084953, + "flos": 21659065480800.0, + "grad_norm": 1.834226354655863, + "learning_rate": 3.875559891990394e-06, + "loss": 0.804, + "num_input_tokens_seen": 50434295, + "step": 1547 + }, + { + "epoch": 0.13960409433196555, + "flos": 32114668854240.0, + "grad_norm": 3.4114179691324296, + "learning_rate": 3.875356956436149e-06, + "loss": 0.7693, + "num_input_tokens_seen": 50464675, + "step": 1548 + }, + { + "epoch": 0.1396942778554358, + "flos": 19909804380960.0, + "grad_norm": 1.8293593062655265, + "learning_rate": 3.875153860866108e-06, + "loss": 0.7838, + "num_input_tokens_seen": 50490755, + "step": 1549 + }, + { + "epoch": 0.13978446137890607, + "flos": 25120003319520.0, + "grad_norm": 2.337931416728509, + "learning_rate": 3.8749506052976e-06, + "loss": 0.8439, + "num_input_tokens_seen": 50520470, + "step": 1550 + }, + { + "epoch": 0.13987464490237633, + "flos": 19145213521440.0, + "grad_norm": 1.5270733606169857, + "learning_rate": 3.874747189747968e-06, + "loss": 0.8245, + "num_input_tokens_seen": 50548165, + "step": 1551 + }, + { + "epoch": 0.1399648284258466, + "flos": 27013989229440.0, + "grad_norm": 1.9827855871131153, + "learning_rate": 3.874543614234568e-06, + "loss": 0.8079, + "num_input_tokens_seen": 50577295, + "step": 1552 + }, + { + "epoch": 0.14005501194931685, + "flos": 29452709636640.0, + "grad_norm": 1.8833873077130185, + "learning_rate": 3.874339878774771e-06, + "loss": 0.7802, + "num_input_tokens_seen": 50606795, + "step": 1553 + }, + { + "epoch": 0.14014519547278712, + "flos": 23622607478880.0, + "grad_norm": 1.923999325643092, + "learning_rate": 3.874135983385961e-06, + "loss": 0.8062, + "num_input_tokens_seen": 50633510, + "step": 1554 + }, + { + "epoch": 0.14023537899625738, + "flos": 15531856732320.0, + "grad_norm": 2.0993673779507387, + "learning_rate": 3.873931928085535e-06, + "loss": 0.7958, + "num_input_tokens_seen": 50659690, + "step": 1555 + }, + { + "epoch": 0.14032556251972764, + "flos": 19977910848480.0, + "grad_norm": 1.6778093350132466, + "learning_rate": 3.873727712890904e-06, + "loss": 0.8847, + "num_input_tokens_seen": 50685570, + "step": 1556 + }, + { + "epoch": 0.1404157460431979, + "flos": 29379473742240.0, + "grad_norm": 1.7142086025901193, + "learning_rate": 3.873523337819493e-06, + "loss": 0.7699, + "num_input_tokens_seen": 50718265, + "step": 1557 + }, + { + "epoch": 0.14050592956666816, + "flos": 23844917045280.0, + "grad_norm": 1.452907587009711, + "learning_rate": 3.873318802888739e-06, + "loss": 0.7307, + "num_input_tokens_seen": 50747365, + "step": 1558 + }, + { + "epoch": 0.14059611309013842, + "flos": 64104647399040.0, + "grad_norm": 0.5644455483531962, + "learning_rate": 3.873114108116097e-06, + "loss": 0.5305, + "num_input_tokens_seen": 50849675, + "step": 1559 + }, + { + "epoch": 0.14068629661360868, + "flos": 21330581879040.0, + "grad_norm": 1.989936604184899, + "learning_rate": 3.872909253519031e-06, + "loss": 0.8901, + "num_input_tokens_seen": 50877875, + "step": 1560 + }, + { + "epoch": 0.14077648013707897, + "flos": 21840556917120.0, + "grad_norm": 2.9434594920141044, + "learning_rate": 3.8727042391150195e-06, + "loss": 0.872, + "num_input_tokens_seen": 50906595, + "step": 1561 + }, + { + "epoch": 0.14086666366054923, + "flos": 20818153636800.0, + "grad_norm": 1.868740696708461, + "learning_rate": 3.872499064921556e-06, + "loss": 0.7389, + "num_input_tokens_seen": 50933155, + "step": 1562 + }, + { + "epoch": 0.1409568471840195, + "flos": 53937607305120.0, + "grad_norm": 1.8824858955979575, + "learning_rate": 3.872293730956149e-06, + "loss": 0.814, + "num_input_tokens_seen": 50964875, + "step": 1563 + }, + { + "epoch": 0.14104703070748975, + "flos": 23188433048640.0, + "grad_norm": 2.1625440207173883, + "learning_rate": 3.872088237236316e-06, + "loss": 0.7937, + "num_input_tokens_seen": 50992010, + "step": 1564 + }, + { + "epoch": 0.14113721423096, + "flos": 18124780238400.0, + "grad_norm": 1.8788268772419152, + "learning_rate": 3.871882583779592e-06, + "loss": 0.7045, + "num_input_tokens_seen": 51019285, + "step": 1565 + }, + { + "epoch": 0.14122739775443027, + "flos": 24683078327040.0, + "grad_norm": 1.6097690136991147, + "learning_rate": 3.871676770603525e-06, + "loss": 0.8035, + "num_input_tokens_seen": 51047655, + "step": 1566 + }, + { + "epoch": 0.14131758127790053, + "flos": 17067654668640.0, + "grad_norm": 2.0013196882781976, + "learning_rate": 3.871470797725676e-06, + "loss": 0.7904, + "num_input_tokens_seen": 51075115, + "step": 1567 + }, + { + "epoch": 0.1414077648013708, + "flos": 21075817378560.0, + "grad_norm": 2.641658512989381, + "learning_rate": 3.8712646651636185e-06, + "loss": 0.8714, + "num_input_tokens_seen": 51103525, + "step": 1568 + }, + { + "epoch": 0.14149794832484106, + "flos": 23845771949760.0, + "grad_norm": 2.1687268830312343, + "learning_rate": 3.871058372934942e-06, + "loss": 0.7988, + "num_input_tokens_seen": 51138215, + "step": 1569 + }, + { + "epoch": 0.14158813184831132, + "flos": 26393156766240.0, + "grad_norm": 1.7971577960846148, + "learning_rate": 3.8708519210572485e-06, + "loss": 0.6286, + "num_input_tokens_seen": 51166555, + "step": 1570 + }, + { + "epoch": 0.14167831537178158, + "flos": 21184741976160.0, + "grad_norm": 5.5051936856462635, + "learning_rate": 3.870645309548153e-06, + "loss": 0.7378, + "num_input_tokens_seen": 51194230, + "step": 1571 + }, + { + "epoch": 0.14176849889525184, + "flos": 59779523712960.0, + "grad_norm": 0.6220008655631001, + "learning_rate": 3.870438538425284e-06, + "loss": 0.5439, + "num_input_tokens_seen": 51288520, + "step": 1572 + }, + { + "epoch": 0.1418586824187221, + "flos": 19654779692160.0, + "grad_norm": 2.530149244045409, + "learning_rate": 3.870231607706287e-06, + "loss": 0.7454, + "num_input_tokens_seen": 51316505, + "step": 1573 + }, + { + "epoch": 0.14194886594219236, + "flos": 21002321295840.0, + "grad_norm": 1.926043138628972, + "learning_rate": 3.870024517408817e-06, + "loss": 0.8029, + "num_input_tokens_seen": 51346240, + "step": 1574 + }, + { + "epoch": 0.14203904946566262, + "flos": 20310334444800.0, + "grad_norm": 2.618850953698585, + "learning_rate": 3.8698172675505425e-06, + "loss": 0.7226, + "num_input_tokens_seen": 51372375, + "step": 1575 + }, + { + "epoch": 0.14212923298913288, + "flos": 61259115238560.0, + "grad_norm": 0.7621038392695422, + "learning_rate": 3.86960985814915e-06, + "loss": 0.5583, + "num_input_tokens_seen": 51466075, + "step": 1576 + }, + { + "epoch": 0.14221941651260314, + "flos": 40854283796640.0, + "grad_norm": 1.6903650898438747, + "learning_rate": 3.869402289222335e-06, + "loss": 0.7289, + "num_input_tokens_seen": 51499360, + "step": 1577 + }, + { + "epoch": 0.1423096000360734, + "flos": 18961900766880.0, + "grad_norm": 1.697558905544361, + "learning_rate": 3.869194560787808e-06, + "loss": 0.7805, + "num_input_tokens_seen": 51528405, + "step": 1578 + }, + { + "epoch": 0.14239978355954366, + "flos": 23513980239360.0, + "grad_norm": 1.9810246095702961, + "learning_rate": 3.868986672863296e-06, + "loss": 0.7263, + "num_input_tokens_seen": 51559115, + "step": 1579 + }, + { + "epoch": 0.14248996708301392, + "flos": 29163631714080.0, + "grad_norm": 1.6242419455423818, + "learning_rate": 3.868778625466535e-06, + "loss": 0.8998, + "num_input_tokens_seen": 51590345, + "step": 1580 + }, + { + "epoch": 0.14258015060648419, + "flos": 37360928193600.0, + "grad_norm": 2.2927212687197804, + "learning_rate": 3.868570418615278e-06, + "loss": 0.715, + "num_input_tokens_seen": 51619485, + "step": 1581 + }, + { + "epoch": 0.14267033412995445, + "flos": 67961952288480.0, + "grad_norm": 0.7299901444932537, + "learning_rate": 3.8683620523272885e-06, + "loss": 0.5666, + "num_input_tokens_seen": 51709500, + "step": 1582 + }, + { + "epoch": 0.1427605176534247, + "flos": 19545186038880.0, + "grad_norm": 1.9483153671249915, + "learning_rate": 3.8681535266203464e-06, + "loss": 0.9021, + "num_input_tokens_seen": 51736835, + "step": 1583 + }, + { + "epoch": 0.14285070117689497, + "flos": 22788051663840.0, + "grad_norm": 1.7841950547966416, + "learning_rate": 3.867944841512246e-06, + "loss": 0.8211, + "num_input_tokens_seen": 51764895, + "step": 1584 + }, + { + "epoch": 0.14294088470036526, + "flos": 22787456947680.0, + "grad_norm": 1.9093337209082233, + "learning_rate": 3.867735997020791e-06, + "loss": 0.8018, + "num_input_tokens_seen": 51789935, + "step": 1585 + }, + { + "epoch": 0.14303106822383552, + "flos": 37541081518560.0, + "grad_norm": 1.8371944588667122, + "learning_rate": 3.867526993163802e-06, + "loss": 0.6991, + "num_input_tokens_seen": 51819810, + "step": 1586 + }, + { + "epoch": 0.14312125174730578, + "flos": 33601173955200.0, + "grad_norm": 1.6310304106608462, + "learning_rate": 3.867317829959113e-06, + "loss": 0.6696, + "num_input_tokens_seen": 51852015, + "step": 1587 + }, + { + "epoch": 0.14321143527077604, + "flos": 24682632289920.0, + "grad_norm": 1.936334382417044, + "learning_rate": 3.8671085074245704e-06, + "loss": 0.594, + "num_input_tokens_seen": 51880075, + "step": 1588 + }, + { + "epoch": 0.1433016187942463, + "flos": 21767729890080.0, + "grad_norm": 2.163948248866685, + "learning_rate": 3.866899025578035e-06, + "loss": 0.8658, + "num_input_tokens_seen": 51907720, + "step": 1589 + }, + { + "epoch": 0.14339180231771656, + "flos": 22788126003360.0, + "grad_norm": 1.537479729247407, + "learning_rate": 3.86668938443738e-06, + "loss": 0.7646, + "num_input_tokens_seen": 51938620, + "step": 1590 + }, + { + "epoch": 0.14348198584118682, + "flos": 25447297488960.0, + "grad_norm": 2.468275773096115, + "learning_rate": 3.866479584020495e-06, + "loss": 0.8105, + "num_input_tokens_seen": 51965805, + "step": 1591 + }, + { + "epoch": 0.14357216936465708, + "flos": 13351580631840.0, + "grad_norm": 1.989012338345915, + "learning_rate": 3.866269624345279e-06, + "loss": 0.831, + "num_input_tokens_seen": 51991185, + "step": 1592 + }, + { + "epoch": 0.14366235288812734, + "flos": 25147626184800.0, + "grad_norm": 2.1712514956594884, + "learning_rate": 3.866059505429649e-06, + "loss": 0.8775, + "num_input_tokens_seen": 52017740, + "step": 1593 + }, + { + "epoch": 0.1437525364115976, + "flos": 66984939018720.0, + "grad_norm": 0.6573796313670877, + "learning_rate": 3.865849227291532e-06, + "loss": 0.5941, + "num_input_tokens_seen": 52105025, + "step": 1594 + }, + { + "epoch": 0.14384271993506786, + "flos": 25003310242080.0, + "grad_norm": 4.324236393369607, + "learning_rate": 3.865638789948872e-06, + "loss": 0.7909, + "num_input_tokens_seen": 52131855, + "step": 1595 + }, + { + "epoch": 0.14393290345853813, + "flos": 24281507509920.0, + "grad_norm": 2.987997841952973, + "learning_rate": 3.865428193419622e-06, + "loss": 0.747, + "num_input_tokens_seen": 52161530, + "step": 1596 + }, + { + "epoch": 0.14402308698200839, + "flos": 23334087102720.0, + "grad_norm": 2.1088738870550525, + "learning_rate": 3.865217437721753e-06, + "loss": 0.7844, + "num_input_tokens_seen": 52191135, + "step": 1597 + }, + { + "epoch": 0.14411327050547865, + "flos": 32619514465440.0, + "grad_norm": 13.852729588981, + "learning_rate": 3.865006522873249e-06, + "loss": 0.728, + "num_input_tokens_seen": 52223170, + "step": 1598 + }, + { + "epoch": 0.1442034540289489, + "flos": 20310408784320.0, + "grad_norm": 1.8355733307374698, + "learning_rate": 3.864795448892103e-06, + "loss": 0.7938, + "num_input_tokens_seen": 52251555, + "step": 1599 + }, + { + "epoch": 0.14429363755241917, + "flos": 23984289409920.0, + "grad_norm": 1.9326832486775032, + "learning_rate": 3.864584215796327e-06, + "loss": 0.8499, + "num_input_tokens_seen": 52281630, + "step": 1600 + }, + { + "epoch": 0.14438382107588943, + "flos": 15529886735040.0, + "grad_norm": 2.3351361485391715, + "learning_rate": 3.8643728236039455e-06, + "loss": 0.8178, + "num_input_tokens_seen": 52306450, + "step": 1601 + }, + { + "epoch": 0.1444740045993597, + "flos": 25374396122400.0, + "grad_norm": 1.726440287273568, + "learning_rate": 3.864161272332994e-06, + "loss": 0.7796, + "num_input_tokens_seen": 52338315, + "step": 1602 + }, + { + "epoch": 0.14456418812282995, + "flos": 25519641309120.0, + "grad_norm": 1.6971008431498902, + "learning_rate": 3.863949562001524e-06, + "loss": 0.8414, + "num_input_tokens_seen": 52369660, + "step": 1603 + }, + { + "epoch": 0.1446543716463002, + "flos": 26794578904320.0, + "grad_norm": 3.341530749695365, + "learning_rate": 3.8637376926276005e-06, + "loss": 0.7695, + "num_input_tokens_seen": 52398445, + "step": 1604 + }, + { + "epoch": 0.14474455516977047, + "flos": 19035248170560.0, + "grad_norm": 2.896441407248308, + "learning_rate": 3.8635256642293e-06, + "loss": 0.802, + "num_input_tokens_seen": 52424810, + "step": 1605 + }, + { + "epoch": 0.14483473869324073, + "flos": 18962718501600.0, + "grad_norm": 1.810597440636934, + "learning_rate": 3.863313476824714e-06, + "loss": 0.7786, + "num_input_tokens_seen": 52453320, + "step": 1606 + }, + { + "epoch": 0.144924922216711, + "flos": 14845519684800.0, + "grad_norm": 3.4685682644361653, + "learning_rate": 3.863101130431948e-06, + "loss": 0.7747, + "num_input_tokens_seen": 52479095, + "step": 1607 + }, + { + "epoch": 0.14501510574018128, + "flos": 24566868456480.0, + "grad_norm": 2.2374856277089474, + "learning_rate": 3.862888625069121e-06, + "loss": 0.84, + "num_input_tokens_seen": 52508685, + "step": 1608 + }, + { + "epoch": 0.14510528926365154, + "flos": 62864028959520.0, + "grad_norm": 1.7167268810390404, + "learning_rate": 3.8626759607543645e-06, + "loss": 0.7594, + "num_input_tokens_seen": 52544770, + "step": 1609 + }, + { + "epoch": 0.1451954727871218, + "flos": 18192626517600.0, + "grad_norm": 2.466126702035389, + "learning_rate": 3.862463137505825e-06, + "loss": 0.7515, + "num_input_tokens_seen": 52572100, + "step": 1610 + }, + { + "epoch": 0.14528565631059207, + "flos": 37543386043680.0, + "grad_norm": 1.7938003772214182, + "learning_rate": 3.862250155341659e-06, + "loss": 0.7599, + "num_input_tokens_seen": 52602850, + "step": 1611 + }, + { + "epoch": 0.14537583983406233, + "flos": 19727941247040.0, + "grad_norm": 1.6502245576813284, + "learning_rate": 3.862037014280043e-06, + "loss": 0.9248, + "num_input_tokens_seen": 52628930, + "step": 1612 + }, + { + "epoch": 0.1454660233575326, + "flos": 20381340153600.0, + "grad_norm": 1.9277942342846683, + "learning_rate": 3.861823714339162e-06, + "loss": 0.7927, + "num_input_tokens_seen": 52658555, + "step": 1613 + }, + { + "epoch": 0.14555620688100285, + "flos": 28579751725920.0, + "grad_norm": 3.7964097982115987, + "learning_rate": 3.861610255537215e-06, + "loss": 0.8017, + "num_input_tokens_seen": 52691495, + "step": 1614 + }, + { + "epoch": 0.1456463904044731, + "flos": 30256631835840.0, + "grad_norm": 1.6041756501710036, + "learning_rate": 3.8613966378924165e-06, + "loss": 0.8237, + "num_input_tokens_seen": 52720405, + "step": 1615 + }, + { + "epoch": 0.14573657392794337, + "flos": 22818202072800.0, + "grad_norm": 3.814177222364734, + "learning_rate": 3.861182861422993e-06, + "loss": 0.7422, + "num_input_tokens_seen": 52746565, + "step": 1616 + }, + { + "epoch": 0.14582675745141363, + "flos": 20419928098080.0, + "grad_norm": 1.6859290263305067, + "learning_rate": 3.860968926147185e-06, + "loss": 0.8664, + "num_input_tokens_seen": 52774735, + "step": 1617 + }, + { + "epoch": 0.1459169409748839, + "flos": 35065780333920.0, + "grad_norm": 2.0574984160888805, + "learning_rate": 3.860754832083247e-06, + "loss": 0.6292, + "num_input_tokens_seen": 52807480, + "step": 1618 + }, + { + "epoch": 0.14600712449835415, + "flos": 21440584399680.0, + "grad_norm": 2.3462348481442525, + "learning_rate": 3.8605405792494475e-06, + "loss": 0.8628, + "num_input_tokens_seen": 52836705, + "step": 1619 + }, + { + "epoch": 0.1460973080218244, + "flos": 23912688984960.0, + "grad_norm": 1.9971652206589265, + "learning_rate": 3.860326167664066e-06, + "loss": 0.81, + "num_input_tokens_seen": 52865255, + "step": 1620 + }, + { + "epoch": 0.14618749154529467, + "flos": 17140035658560.0, + "grad_norm": 2.409454589795808, + "learning_rate": 3.860111597345399e-06, + "loss": 0.7771, + "num_input_tokens_seen": 52891985, + "step": 1621 + }, + { + "epoch": 0.14627767506876493, + "flos": 24026408481600.0, + "grad_norm": 21.88648624218371, + "learning_rate": 3.859896868311753e-06, + "loss": 0.6489, + "num_input_tokens_seen": 52919770, + "step": 1622 + }, + { + "epoch": 0.1463678585922352, + "flos": 23330184277920.0, + "grad_norm": 1.655478459717241, + "learning_rate": 3.859681980581452e-06, + "loss": 0.8857, + "num_input_tokens_seen": 52947440, + "step": 1623 + }, + { + "epoch": 0.14645804211570546, + "flos": 21367051147200.0, + "grad_norm": 3.033238338897331, + "learning_rate": 3.859466934172829e-06, + "loss": 0.7713, + "num_input_tokens_seen": 52973855, + "step": 1624 + }, + { + "epoch": 0.14654822563917572, + "flos": 26578662536640.0, + "grad_norm": 1.5692414384079338, + "learning_rate": 3.859251729104235e-06, + "loss": 0.7253, + "num_input_tokens_seen": 53006805, + "step": 1625 + }, + { + "epoch": 0.14663840916264598, + "flos": 24204814827840.0, + "grad_norm": 1.7547040114810575, + "learning_rate": 3.859036365394031e-06, + "loss": 0.7415, + "num_input_tokens_seen": 53035070, + "step": 1626 + }, + { + "epoch": 0.14672859268611624, + "flos": 22273133047680.0, + "grad_norm": 2.636152463784463, + "learning_rate": 3.858820843060594e-06, + "loss": 0.7768, + "num_input_tokens_seen": 53062400, + "step": 1627 + }, + { + "epoch": 0.1468187762095865, + "flos": 20127876594720.0, + "grad_norm": 1.9977114219640324, + "learning_rate": 3.858605162122314e-06, + "loss": 0.7681, + "num_input_tokens_seen": 53091875, + "step": 1628 + }, + { + "epoch": 0.14690895973305676, + "flos": 16922706840000.0, + "grad_norm": 1.751899982101429, + "learning_rate": 3.858389322597592e-06, + "loss": 0.748, + "num_input_tokens_seen": 53118745, + "step": 1629 + }, + { + "epoch": 0.14699914325652702, + "flos": 20674915617120.0, + "grad_norm": 1.827896234267556, + "learning_rate": 3.858173324504847e-06, + "loss": 0.8067, + "num_input_tokens_seen": 53147440, + "step": 1630 + }, + { + "epoch": 0.14708932677999728, + "flos": 22569050206080.0, + "grad_norm": 1.5467075295777843, + "learning_rate": 3.857957167862508e-06, + "loss": 0.7492, + "num_input_tokens_seen": 53174255, + "step": 1631 + }, + { + "epoch": 0.14717951030346757, + "flos": 66288082929120.0, + "grad_norm": 0.6729480690808594, + "learning_rate": 3.857740852689018e-06, + "loss": 0.5788, + "num_input_tokens_seen": 53264685, + "step": 1632 + }, + { + "epoch": 0.14726969382693783, + "flos": 20490747958080.0, + "grad_norm": 2.30874132436662, + "learning_rate": 3.857524379002835e-06, + "loss": 0.7978, + "num_input_tokens_seen": 53289870, + "step": 1633 + }, + { + "epoch": 0.1473598773504081, + "flos": 24857581848480.0, + "grad_norm": 1.6477856544986023, + "learning_rate": 3.85730774682243e-06, + "loss": 0.8094, + "num_input_tokens_seen": 53318045, + "step": 1634 + }, + { + "epoch": 0.14745006087387835, + "flos": 24245484278880.0, + "grad_norm": 1.9861917664602018, + "learning_rate": 3.8570909561662875e-06, + "loss": 0.853, + "num_input_tokens_seen": 53348835, + "step": 1635 + }, + { + "epoch": 0.1475402443973486, + "flos": 23839899127680.0, + "grad_norm": 1.590938706555843, + "learning_rate": 3.8568740070529045e-06, + "loss": 0.8576, + "num_input_tokens_seen": 53378130, + "step": 1636 + }, + { + "epoch": 0.14763042792081887, + "flos": 16412954820480.0, + "grad_norm": 1.867444902498381, + "learning_rate": 3.856656899500792e-06, + "loss": 0.7554, + "num_input_tokens_seen": 53404455, + "step": 1637 + }, + { + "epoch": 0.14772061144428913, + "flos": 22022197032480.0, + "grad_norm": 1.774859089698194, + "learning_rate": 3.856439633528476e-06, + "loss": 0.7233, + "num_input_tokens_seen": 53429210, + "step": 1638 + }, + { + "epoch": 0.1478107949677594, + "flos": 25702396517280.0, + "grad_norm": 1.8631202100342221, + "learning_rate": 3.856222209154494e-06, + "loss": 0.8417, + "num_input_tokens_seen": 53458845, + "step": 1639 + }, + { + "epoch": 0.14790097849122966, + "flos": 20781275501280.0, + "grad_norm": 1.6582330312379716, + "learning_rate": 3.856004626397397e-06, + "loss": 0.8135, + "num_input_tokens_seen": 53487325, + "step": 1640 + }, + { + "epoch": 0.14799116201469992, + "flos": 25045838181120.0, + "grad_norm": 3.2347047794305435, + "learning_rate": 3.855786885275753e-06, + "loss": 0.7642, + "num_input_tokens_seen": 53515975, + "step": 1641 + }, + { + "epoch": 0.14808134553817018, + "flos": 26322485585280.0, + "grad_norm": 2.3051157408015785, + "learning_rate": 3.855568985808138e-06, + "loss": 0.7041, + "num_input_tokens_seen": 53544725, + "step": 1642 + }, + { + "epoch": 0.14817152906164044, + "flos": 33205215771840.0, + "grad_norm": 1.7027595197864878, + "learning_rate": 3.855350928013145e-06, + "loss": 0.7585, + "num_input_tokens_seen": 53576340, + "step": 1643 + }, + { + "epoch": 0.1482617125851107, + "flos": 20201967393600.0, + "grad_norm": 2.0764522847591147, + "learning_rate": 3.8551327119093825e-06, + "loss": 0.7948, + "num_input_tokens_seen": 53602735, + "step": 1644 + }, + { + "epoch": 0.14835189610858096, + "flos": 22387410090720.0, + "grad_norm": 2.285288123865792, + "learning_rate": 3.854914337515467e-06, + "loss": 0.7157, + "num_input_tokens_seen": 53631795, + "step": 1645 + }, + { + "epoch": 0.14844207963205122, + "flos": 27706050420000.0, + "grad_norm": 1.9230313427731973, + "learning_rate": 3.8546958048500324e-06, + "loss": 0.6844, + "num_input_tokens_seen": 53660425, + "step": 1646 + }, + { + "epoch": 0.14853226315552148, + "flos": 68402780105760.0, + "grad_norm": 1.2924983618189099, + "learning_rate": 3.854477113931725e-06, + "loss": 0.631, + "num_input_tokens_seen": 53750730, + "step": 1647 + }, + { + "epoch": 0.14862244667899174, + "flos": 20784583609920.0, + "grad_norm": 1.8127857509165821, + "learning_rate": 3.854258264779205e-06, + "loss": 0.8109, + "num_input_tokens_seen": 53779495, + "step": 1648 + }, + { + "epoch": 0.148712630202462, + "flos": 21108012124320.0, + "grad_norm": 1.7538896657588703, + "learning_rate": 3.854039257411145e-06, + "loss": 0.7679, + "num_input_tokens_seen": 53808350, + "step": 1649 + }, + { + "epoch": 0.14880281372593226, + "flos": 25848645287520.0, + "grad_norm": 2.068080634767865, + "learning_rate": 3.853820091846232e-06, + "loss": 0.7385, + "num_input_tokens_seen": 53838135, + "step": 1650 + }, + { + "epoch": 0.14889299724940253, + "flos": 18265304865600.0, + "grad_norm": 1.909373616469361, + "learning_rate": 3.853600768103169e-06, + "loss": 0.7885, + "num_input_tokens_seen": 53863855, + "step": 1651 + }, + { + "epoch": 0.1489831807728728, + "flos": 29815506660480.0, + "grad_norm": 2.9713104449885486, + "learning_rate": 3.853381286200667e-06, + "loss": 0.7757, + "num_input_tokens_seen": 53886890, + "step": 1652 + }, + { + "epoch": 0.14907336429634305, + "flos": 23407062808800.0, + "grad_norm": 2.080691824630859, + "learning_rate": 3.853161646157453e-06, + "loss": 0.8777, + "num_input_tokens_seen": 53914125, + "step": 1653 + }, + { + "epoch": 0.1491635478198133, + "flos": 21074925304320.0, + "grad_norm": 1.9994566298500924, + "learning_rate": 3.852941847992269e-06, + "loss": 0.7504, + "num_input_tokens_seen": 53942375, + "step": 1654 + }, + { + "epoch": 0.14925373134328357, + "flos": 19035545528640.0, + "grad_norm": 2.34601615703772, + "learning_rate": 3.852721891723871e-06, + "loss": 0.8034, + "num_input_tokens_seen": 53968880, + "step": 1655 + }, + { + "epoch": 0.14934391486675386, + "flos": 17905778780640.0, + "grad_norm": 2.691189739227692, + "learning_rate": 3.852501777371025e-06, + "loss": 0.7733, + "num_input_tokens_seen": 53995265, + "step": 1656 + }, + { + "epoch": 0.14943409839022412, + "flos": 24167341976160.0, + "grad_norm": 4.635673990143775, + "learning_rate": 3.8522815049525125e-06, + "loss": 0.8259, + "num_input_tokens_seen": 54025315, + "step": 1657 + }, + { + "epoch": 0.14952428191369438, + "flos": 22860692842080.0, + "grad_norm": 1.9023946328378591, + "learning_rate": 3.852061074487129e-06, + "loss": 0.8434, + "num_input_tokens_seen": 54052600, + "step": 1658 + }, + { + "epoch": 0.14961446543716464, + "flos": 33863818444800.0, + "grad_norm": 1.8918332739350467, + "learning_rate": 3.851840485993682e-06, + "loss": 0.6955, + "num_input_tokens_seen": 54085930, + "step": 1659 + }, + { + "epoch": 0.1497046489606349, + "flos": 19182574863840.0, + "grad_norm": 1.7512595548512053, + "learning_rate": 3.851619739490994e-06, + "loss": 0.7989, + "num_input_tokens_seen": 54114805, + "step": 1660 + }, + { + "epoch": 0.14979483248410516, + "flos": 21216230496480.0, + "grad_norm": 1.7343082586970209, + "learning_rate": 3.8513988349978996e-06, + "loss": 0.8245, + "num_input_tokens_seen": 54140960, + "step": 1661 + }, + { + "epoch": 0.14988501600757542, + "flos": 21293926762080.0, + "grad_norm": 1.671498764343444, + "learning_rate": 3.851177772533249e-06, + "loss": 0.8151, + "num_input_tokens_seen": 54170145, + "step": 1662 + }, + { + "epoch": 0.14997519953104568, + "flos": 19944377991360.0, + "grad_norm": 2.706755652992812, + "learning_rate": 3.850956552115903e-06, + "loss": 0.9631, + "num_input_tokens_seen": 54193415, + "step": 1663 + }, + { + "epoch": 0.15006538305451594, + "flos": 66503032883040.0, + "grad_norm": 0.7019559678671304, + "learning_rate": 3.850735173764738e-06, + "loss": 0.5824, + "num_input_tokens_seen": 54283935, + "step": 1664 + }, + { + "epoch": 0.1501555665779862, + "flos": 23297655004320.0, + "grad_norm": 1.7341830860529543, + "learning_rate": 3.850513637498642e-06, + "loss": 0.8071, + "num_input_tokens_seen": 54314130, + "step": 1665 + }, + { + "epoch": 0.15024575010145647, + "flos": 25192979025600.0, + "grad_norm": 2.219135089821032, + "learning_rate": 3.850291943336521e-06, + "loss": 0.8067, + "num_input_tokens_seen": 54343600, + "step": 1666 + }, + { + "epoch": 0.15033593362492673, + "flos": 23989679025120.0, + "grad_norm": 2.7988240361100627, + "learning_rate": 3.850070091297287e-06, + "loss": 0.8432, + "num_input_tokens_seen": 54371740, + "step": 1667 + }, + { + "epoch": 0.150426117148397, + "flos": 66642330908160.0, + "grad_norm": 0.7118551598815976, + "learning_rate": 3.8498480813998735e-06, + "loss": 0.6735, + "num_input_tokens_seen": 54458715, + "step": 1668 + }, + { + "epoch": 0.15051630067186725, + "flos": 24896281302240.0, + "grad_norm": 1.8529151450304537, + "learning_rate": 3.84962591366322e-06, + "loss": 0.7886, + "num_input_tokens_seen": 54488225, + "step": 1669 + }, + { + "epoch": 0.1506064841953375, + "flos": 24317530740960.0, + "grad_norm": 2.037219270428307, + "learning_rate": 3.8494035881062855e-06, + "loss": 0.8405, + "num_input_tokens_seen": 54516800, + "step": 1670 + }, + { + "epoch": 0.15069666771880777, + "flos": 14989798457760.0, + "grad_norm": 2.06318526010414, + "learning_rate": 3.84918110474804e-06, + "loss": 0.7728, + "num_input_tokens_seen": 54542755, + "step": 1671 + }, + { + "epoch": 0.15078685124227803, + "flos": 30397937028000.0, + "grad_norm": 1.5909417037593623, + "learning_rate": 3.8489584636074655e-06, + "loss": 0.7767, + "num_input_tokens_seen": 54575190, + "step": 1672 + }, + { + "epoch": 0.1508770347657483, + "flos": 69190013082720.0, + "grad_norm": 0.6860684749888342, + "learning_rate": 3.848735664703561e-06, + "loss": 0.5648, + "num_input_tokens_seen": 54670680, + "step": 1673 + }, + { + "epoch": 0.15096721828921855, + "flos": 25884854367360.0, + "grad_norm": 1.7963256659661182, + "learning_rate": 3.8485127080553346e-06, + "loss": 0.7864, + "num_input_tokens_seen": 54699840, + "step": 1674 + }, + { + "epoch": 0.1510574018126888, + "flos": 22350717804000.0, + "grad_norm": 1.8624218892353064, + "learning_rate": 3.8482895936818115e-06, + "loss": 0.7879, + "num_input_tokens_seen": 54730210, + "step": 1675 + }, + { + "epoch": 0.15114758533615907, + "flos": 21731929677600.0, + "grad_norm": 1.8752541687673296, + "learning_rate": 3.848066321602029e-06, + "loss": 0.808, + "num_input_tokens_seen": 54755285, + "step": 1676 + }, + { + "epoch": 0.15123776885962933, + "flos": 21038902073280.0, + "grad_norm": 2.914658554520738, + "learning_rate": 3.847842891835038e-06, + "loss": 0.7521, + "num_input_tokens_seen": 54782910, + "step": 1677 + }, + { + "epoch": 0.1513279523830996, + "flos": 64509451985280.0, + "grad_norm": 0.6263607708775077, + "learning_rate": 3.847619304399902e-06, + "loss": 0.5803, + "num_input_tokens_seen": 54884585, + "step": 1678 + }, + { + "epoch": 0.15141813590656986, + "flos": 34183976020320.0, + "grad_norm": 2.4584204986981355, + "learning_rate": 3.8473955593157e-06, + "loss": 0.8168, + "num_input_tokens_seen": 54913825, + "step": 1679 + }, + { + "epoch": 0.15150831943004014, + "flos": 33462693664800.0, + "grad_norm": 1.777920271043839, + "learning_rate": 3.847171656601522e-06, + "loss": 0.7535, + "num_input_tokens_seen": 54945305, + "step": 1680 + }, + { + "epoch": 0.1515985029535104, + "flos": 22315437968160.0, + "grad_norm": 2.067343717339664, + "learning_rate": 3.846947596276473e-06, + "loss": 0.8188, + "num_input_tokens_seen": 54972050, + "step": 1681 + }, + { + "epoch": 0.15168868647698067, + "flos": 32333075595840.0, + "grad_norm": 2.429116234056812, + "learning_rate": 3.846723378359672e-06, + "loss": 0.8404, + "num_input_tokens_seen": 55006755, + "step": 1682 + }, + { + "epoch": 0.15177887000045093, + "flos": 20339444100480.0, + "grad_norm": 2.746655250760153, + "learning_rate": 3.846499002870249e-06, + "loss": 0.7963, + "num_input_tokens_seen": 55030520, + "step": 1683 + }, + { + "epoch": 0.1518690535239212, + "flos": 20198845133760.0, + "grad_norm": 2.04391428435897, + "learning_rate": 3.846274469827349e-06, + "loss": 0.8114, + "num_input_tokens_seen": 55059115, + "step": 1684 + }, + { + "epoch": 0.15195923704739145, + "flos": 27050607176640.0, + "grad_norm": 1.86879906478313, + "learning_rate": 3.846049779250132e-06, + "loss": 0.8019, + "num_input_tokens_seen": 55089370, + "step": 1685 + }, + { + "epoch": 0.1520494205708617, + "flos": 65118167106720.0, + "grad_norm": 0.7861753085005617, + "learning_rate": 3.845824931157769e-06, + "loss": 0.6235, + "num_input_tokens_seen": 55185430, + "step": 1686 + }, + { + "epoch": 0.15213960409433197, + "flos": 15136307416320.0, + "grad_norm": 3.0315573624977405, + "learning_rate": 3.845599925569444e-06, + "loss": 0.7778, + "num_input_tokens_seen": 55211520, + "step": 1687 + }, + { + "epoch": 0.15222978761780223, + "flos": 26394829405440.0, + "grad_norm": 2.06671140324974, + "learning_rate": 3.845374762504357e-06, + "loss": 0.7532, + "num_input_tokens_seen": 55241565, + "step": 1688 + }, + { + "epoch": 0.1523199711412725, + "flos": 26978932412160.0, + "grad_norm": 4.261886131852763, + "learning_rate": 3.8451494419817204e-06, + "loss": 0.8409, + "num_input_tokens_seen": 55271170, + "step": 1689 + }, + { + "epoch": 0.15241015466474275, + "flos": 30110940612000.0, + "grad_norm": 2.5973589504742396, + "learning_rate": 3.8449239640207594e-06, + "loss": 0.7199, + "num_input_tokens_seen": 55302305, + "step": 1690 + }, + { + "epoch": 0.152500338188213, + "flos": 21804310667520.0, + "grad_norm": 2.3058168909138965, + "learning_rate": 3.844698328640713e-06, + "loss": 0.8362, + "num_input_tokens_seen": 55330125, + "step": 1691 + }, + { + "epoch": 0.15259052171168327, + "flos": 24316192629600.0, + "grad_norm": 1.4974296322753644, + "learning_rate": 3.844472535860833e-06, + "loss": 0.8076, + "num_input_tokens_seen": 55359825, + "step": 1692 + }, + { + "epoch": 0.15268070523515354, + "flos": 20018989166880.0, + "grad_norm": 2.9099451999813555, + "learning_rate": 3.8442465857003864e-06, + "loss": 0.8418, + "num_input_tokens_seen": 55385745, + "step": 1693 + }, + { + "epoch": 0.1527708887586238, + "flos": 22057365359040.0, + "grad_norm": 1.5327670309339785, + "learning_rate": 3.844020478178653e-06, + "loss": 0.7362, + "num_input_tokens_seen": 55414740, + "step": 1694 + }, + { + "epoch": 0.15286107228209406, + "flos": 23333975593440.0, + "grad_norm": 1.7239420756546324, + "learning_rate": 3.843794213314923e-06, + "loss": 0.8161, + "num_input_tokens_seen": 55442695, + "step": 1695 + }, + { + "epoch": 0.15295125580556432, + "flos": 24792411792000.0, + "grad_norm": 1.7445547673732533, + "learning_rate": 3.843567791128505e-06, + "loss": 0.8213, + "num_input_tokens_seen": 55471135, + "step": 1696 + }, + { + "epoch": 0.15304143932903458, + "flos": 30142391962560.0, + "grad_norm": 1.7687368062204605, + "learning_rate": 3.843341211638717e-06, + "loss": 0.7802, + "num_input_tokens_seen": 55500775, + "step": 1697 + }, + { + "epoch": 0.15313162285250484, + "flos": 23223229677600.0, + "grad_norm": 2.19272012473324, + "learning_rate": 3.843114474864894e-06, + "loss": 0.752, + "num_input_tokens_seen": 55530590, + "step": 1698 + }, + { + "epoch": 0.1532218063759751, + "flos": 25847864722560.0, + "grad_norm": 1.6444968901825472, + "learning_rate": 3.84288758082638e-06, + "loss": 0.7913, + "num_input_tokens_seen": 55561065, + "step": 1699 + }, + { + "epoch": 0.15331198989944536, + "flos": 22422875775360.0, + "grad_norm": 1.822587977249508, + "learning_rate": 3.842660529542536e-06, + "loss": 0.8759, + "num_input_tokens_seen": 55590950, + "step": 1700 + }, + { + "epoch": 0.15340217342291562, + "flos": 27046964540160.0, + "grad_norm": 1.6259504619714453, + "learning_rate": 3.842433321032736e-06, + "loss": 0.8047, + "num_input_tokens_seen": 55619075, + "step": 1701 + }, + { + "epoch": 0.15349235694638588, + "flos": 31638635540640.0, + "grad_norm": 1.7983175706688173, + "learning_rate": 3.842205955316365e-06, + "loss": 0.7161, + "num_input_tokens_seen": 55649765, + "step": 1702 + }, + { + "epoch": 0.15358254046985614, + "flos": 29273113858080.0, + "grad_norm": 2.0669328851383297, + "learning_rate": 3.8419784324128256e-06, + "loss": 0.6991, + "num_input_tokens_seen": 55680895, + "step": 1703 + }, + { + "epoch": 0.15367272399332643, + "flos": 21401624757600.0, + "grad_norm": 1.836955336208957, + "learning_rate": 3.841750752341529e-06, + "loss": 0.7549, + "num_input_tokens_seen": 55702530, + "step": 1704 + }, + { + "epoch": 0.1537629075167967, + "flos": 20966669762400.0, + "grad_norm": 1.80912578493817, + "learning_rate": 3.841522915121902e-06, + "loss": 0.7968, + "num_input_tokens_seen": 55730990, + "step": 1705 + }, + { + "epoch": 0.15385309104026695, + "flos": 23443978114080.0, + "grad_norm": 2.08183450824958, + "learning_rate": 3.841294920773387e-06, + "loss": 0.7202, + "num_input_tokens_seen": 55759710, + "step": 1706 + }, + { + "epoch": 0.15394327456373721, + "flos": 17141373769920.0, + "grad_norm": 2.2058322586654544, + "learning_rate": 3.841066769315436e-06, + "loss": 0.8049, + "num_input_tokens_seen": 55787030, + "step": 1707 + }, + { + "epoch": 0.15403345808720748, + "flos": 19109376139200.0, + "grad_norm": 1.6190655257199158, + "learning_rate": 3.840838460767518e-06, + "loss": 0.8018, + "num_input_tokens_seen": 55812810, + "step": 1708 + }, + { + "epoch": 0.15412364161067774, + "flos": 23949827308800.0, + "grad_norm": 2.095264171978898, + "learning_rate": 3.840609995149111e-06, + "loss": 0.7626, + "num_input_tokens_seen": 55839260, + "step": 1709 + }, + { + "epoch": 0.154213825134148, + "flos": 30545672588640.0, + "grad_norm": 1.7164406954851523, + "learning_rate": 3.84038137247971e-06, + "loss": 0.7947, + "num_input_tokens_seen": 55869000, + "step": 1710 + }, + { + "epoch": 0.15430400865761826, + "flos": 25149596182080.0, + "grad_norm": 1.6880965414688498, + "learning_rate": 3.840152592778823e-06, + "loss": 0.7778, + "num_input_tokens_seen": 55897900, + "step": 1711 + }, + { + "epoch": 0.15439419218108852, + "flos": 27414519293280.0, + "grad_norm": 1.3460613598859237, + "learning_rate": 3.83992365606597e-06, + "loss": 0.8376, + "num_input_tokens_seen": 55929790, + "step": 1712 + }, + { + "epoch": 0.15448437570455878, + "flos": 26246461958880.0, + "grad_norm": 2.0304572098841813, + "learning_rate": 3.8396945623606855e-06, + "loss": 0.744, + "num_input_tokens_seen": 55960250, + "step": 1713 + }, + { + "epoch": 0.15457455922802904, + "flos": 23079731469600.0, + "grad_norm": 1.5875336831191462, + "learning_rate": 3.8394653116825174e-06, + "loss": 0.7463, + "num_input_tokens_seen": 55989420, + "step": 1714 + }, + { + "epoch": 0.1546647427514993, + "flos": 24317605080480.0, + "grad_norm": 2.4739238749998824, + "learning_rate": 3.839235904051026e-06, + "loss": 0.7754, + "num_input_tokens_seen": 56018435, + "step": 1715 + }, + { + "epoch": 0.15475492627496956, + "flos": 21694754184000.0, + "grad_norm": 5.833739166674743, + "learning_rate": 3.8390063394857855e-06, + "loss": 0.7691, + "num_input_tokens_seen": 56047860, + "step": 1716 + }, + { + "epoch": 0.15484510979843982, + "flos": 26172296820480.0, + "grad_norm": 1.4896572690456624, + "learning_rate": 3.838776618006385e-06, + "loss": 0.8728, + "num_input_tokens_seen": 56078085, + "step": 1717 + }, + { + "epoch": 0.15493529332191008, + "flos": 20889902740800.0, + "grad_norm": 1.8821455537978864, + "learning_rate": 3.838546739632423e-06, + "loss": 0.7961, + "num_input_tokens_seen": 56107225, + "step": 1718 + }, + { + "epoch": 0.15502547684538034, + "flos": 22711582000320.0, + "grad_norm": 1.5560155209304125, + "learning_rate": 3.838316704383517e-06, + "loss": 0.8395, + "num_input_tokens_seen": 56134090, + "step": 1719 + }, + { + "epoch": 0.1551156603688506, + "flos": 21439952513760.0, + "grad_norm": 2.0589220628480738, + "learning_rate": 3.838086512279292e-06, + "loss": 0.7958, + "num_input_tokens_seen": 56163200, + "step": 1720 + }, + { + "epoch": 0.15520584389232087, + "flos": 28173943556160.0, + "grad_norm": 1.9738754281891528, + "learning_rate": 3.837856163339391e-06, + "loss": 0.7933, + "num_input_tokens_seen": 56191785, + "step": 1721 + }, + { + "epoch": 0.15529602741579113, + "flos": 60094929292800.0, + "grad_norm": 1.493343843922638, + "learning_rate": 3.837625657583469e-06, + "loss": 0.7189, + "num_input_tokens_seen": 56227075, + "step": 1722 + }, + { + "epoch": 0.1553862109392614, + "flos": 34409036148960.0, + "grad_norm": 1.8451733671575563, + "learning_rate": 3.837394995031193e-06, + "loss": 0.7504, + "num_input_tokens_seen": 56260525, + "step": 1723 + }, + { + "epoch": 0.15547639446273165, + "flos": 31604582306880.0, + "grad_norm": 1.6694341650061175, + "learning_rate": 3.837164175702245e-06, + "loss": 0.7906, + "num_input_tokens_seen": 56291015, + "step": 1724 + }, + { + "epoch": 0.1555665779862019, + "flos": 17651274468480.0, + "grad_norm": 2.080092452452124, + "learning_rate": 3.836933199616319e-06, + "loss": 0.71, + "num_input_tokens_seen": 56317555, + "step": 1725 + }, + { + "epoch": 0.15565676150967217, + "flos": 19144618805280.0, + "grad_norm": 1.7297620858890672, + "learning_rate": 3.836702066793124e-06, + "loss": 0.8607, + "num_input_tokens_seen": 56344365, + "step": 1726 + }, + { + "epoch": 0.15574694503314246, + "flos": 21032806232640.0, + "grad_norm": 1.8857031827554356, + "learning_rate": 3.836470777252381e-06, + "loss": 0.8527, + "num_input_tokens_seen": 56374115, + "step": 1727 + }, + { + "epoch": 0.15583712855661272, + "flos": 25483915436160.0, + "grad_norm": 2.576565542502657, + "learning_rate": 3.836239331013825e-06, + "loss": 0.8052, + "num_input_tokens_seen": 56404570, + "step": 1728 + }, + { + "epoch": 0.15592731208008298, + "flos": 24245186920800.0, + "grad_norm": 2.206597371622023, + "learning_rate": 3.836007728097205e-06, + "loss": 0.7539, + "num_input_tokens_seen": 56435110, + "step": 1729 + }, + { + "epoch": 0.15601749560355324, + "flos": 24973531530720.0, + "grad_norm": 1.624839187881253, + "learning_rate": 3.835775968522282e-06, + "loss": 0.7451, + "num_input_tokens_seen": 56465365, + "step": 1730 + }, + { + "epoch": 0.1561076791270235, + "flos": 20633093903520.0, + "grad_norm": 3.7993865321028744, + "learning_rate": 3.83554405230883e-06, + "loss": 0.8099, + "num_input_tokens_seen": 56492400, + "step": 1731 + }, + { + "epoch": 0.15619786265049376, + "flos": 31093343496960.0, + "grad_norm": 1.5720514945412118, + "learning_rate": 3.835311979476639e-06, + "loss": 0.7746, + "num_input_tokens_seen": 56526885, + "step": 1732 + }, + { + "epoch": 0.15628804617396402, + "flos": 22752251451360.0, + "grad_norm": 1.5851470029513275, + "learning_rate": 3.83507975004551e-06, + "loss": 0.6787, + "num_input_tokens_seen": 56556980, + "step": 1733 + }, + { + "epoch": 0.15637822969743428, + "flos": 16340313642240.0, + "grad_norm": 1.7236764529487758, + "learning_rate": 3.834847364035258e-06, + "loss": 0.7687, + "num_input_tokens_seen": 56584505, + "step": 1734 + }, + { + "epoch": 0.15646841322090455, + "flos": 24572964297120.0, + "grad_norm": 2.670956370058413, + "learning_rate": 3.834614821465712e-06, + "loss": 0.7342, + "num_input_tokens_seen": 56612825, + "step": 1735 + }, + { + "epoch": 0.1565585967443748, + "flos": 23116237907520.0, + "grad_norm": 4.434992336881569, + "learning_rate": 3.834382122356713e-06, + "loss": 0.8285, + "num_input_tokens_seen": 56639550, + "step": 1736 + }, + { + "epoch": 0.15664878026784507, + "flos": 25514363203200.0, + "grad_norm": 1.7110730121893112, + "learning_rate": 3.834149266728117e-06, + "loss": 0.7482, + "num_input_tokens_seen": 56667720, + "step": 1737 + }, + { + "epoch": 0.15673896379131533, + "flos": 54336384656640.0, + "grad_norm": 0.6804695234502345, + "learning_rate": 3.833916254599792e-06, + "loss": 0.6195, + "num_input_tokens_seen": 56752940, + "step": 1738 + }, + { + "epoch": 0.1568291473147856, + "flos": 67274388638880.0, + "grad_norm": 0.6344275410509088, + "learning_rate": 3.83368308599162e-06, + "loss": 0.5656, + "num_input_tokens_seen": 56844040, + "step": 1739 + }, + { + "epoch": 0.15691933083825585, + "flos": 71285822287680.0, + "grad_norm": 0.6188170974648506, + "learning_rate": 3.833449760923498e-06, + "loss": 0.5374, + "num_input_tokens_seen": 56944100, + "step": 1740 + }, + { + "epoch": 0.1570095143617261, + "flos": 19252353970560.0, + "grad_norm": 1.7024706362126008, + "learning_rate": 3.83321627941533e-06, + "loss": 0.8283, + "num_input_tokens_seen": 56970600, + "step": 1741 + }, + { + "epoch": 0.15709969788519637, + "flos": 17139924149280.0, + "grad_norm": 1.9380754416144166, + "learning_rate": 3.832982641487043e-06, + "loss": 0.8453, + "num_input_tokens_seen": 56996985, + "step": 1742 + }, + { + "epoch": 0.15718988140866663, + "flos": 19654593843360.0, + "grad_norm": 1.9631643227832207, + "learning_rate": 3.832748847158568e-06, + "loss": 0.8787, + "num_input_tokens_seen": 57025385, + "step": 1743 + }, + { + "epoch": 0.1572800649321369, + "flos": 31460266364160.0, + "grad_norm": 1.602372053278606, + "learning_rate": 3.832514896449858e-06, + "loss": 0.8124, + "num_input_tokens_seen": 57058730, + "step": 1744 + }, + { + "epoch": 0.15737024845560715, + "flos": 23841943464480.0, + "grad_norm": 1.5324945647001724, + "learning_rate": 3.832280789380871e-06, + "loss": 0.7255, + "num_input_tokens_seen": 57087090, + "step": 1745 + }, + { + "epoch": 0.15746043197907741, + "flos": 21730963263840.0, + "grad_norm": 1.6246004346971636, + "learning_rate": 3.832046525971584e-06, + "loss": 0.69, + "num_input_tokens_seen": 57115765, + "step": 1746 + }, + { + "epoch": 0.15755061550254768, + "flos": 24166375562400.0, + "grad_norm": 1.8873571101790418, + "learning_rate": 3.831812106241987e-06, + "loss": 0.8474, + "num_input_tokens_seen": 57143965, + "step": 1747 + }, + { + "epoch": 0.15764079902601794, + "flos": 18306829221120.0, + "grad_norm": 1.7668421103375191, + "learning_rate": 3.8315775302120796e-06, + "loss": 0.6943, + "num_input_tokens_seen": 57171795, + "step": 1748 + }, + { + "epoch": 0.1577309825494882, + "flos": 25082716316640.0, + "grad_norm": 1.6253973401734707, + "learning_rate": 3.831342797901878e-06, + "loss": 0.8019, + "num_input_tokens_seen": 57204295, + "step": 1749 + }, + { + "epoch": 0.15782116607295846, + "flos": 68529769206720.0, + "grad_norm": 0.6711548560925058, + "learning_rate": 3.831107909331411e-06, + "loss": 0.5785, + "num_input_tokens_seen": 57307375, + "step": 1750 + }, + { + "epoch": 0.15791134959642875, + "flos": 22674257827680.0, + "grad_norm": 1.8641333720826827, + "learning_rate": 3.830872864520721e-06, + "loss": 0.829, + "num_input_tokens_seen": 57333160, + "step": 1751 + }, + { + "epoch": 0.158001533119899, + "flos": 64132753471200.0, + "grad_norm": 0.7166601565161441, + "learning_rate": 3.830637663489862e-06, + "loss": 0.6582, + "num_input_tokens_seen": 57422165, + "step": 1752 + }, + { + "epoch": 0.15809171664336927, + "flos": 65856659061120.0, + "grad_norm": 0.7789512766657617, + "learning_rate": 3.830402306258904e-06, + "loss": 0.6886, + "num_input_tokens_seen": 57511710, + "step": 1753 + }, + { + "epoch": 0.15818190016683953, + "flos": 18598471857120.0, + "grad_norm": 3.097366992967339, + "learning_rate": 3.830166792847929e-06, + "loss": 0.7741, + "num_input_tokens_seen": 57539580, + "step": 1754 + }, + { + "epoch": 0.1582720836903098, + "flos": 32986474502400.0, + "grad_norm": 1.5305719236655244, + "learning_rate": 3.829931123277031e-06, + "loss": 0.7251, + "num_input_tokens_seen": 57572850, + "step": 1755 + }, + { + "epoch": 0.15836226721378005, + "flos": 37360853854080.0, + "grad_norm": 2.2941413233759365, + "learning_rate": 3.8296952975663204e-06, + "loss": 0.7746, + "num_input_tokens_seen": 57603575, + "step": 1756 + }, + { + "epoch": 0.1584524507372503, + "flos": 25520198855520.0, + "grad_norm": 1.554119830751072, + "learning_rate": 3.829459315735918e-06, + "loss": 0.8543, + "num_input_tokens_seen": 57632335, + "step": 1757 + }, + { + "epoch": 0.15854263426072057, + "flos": 22569087375840.0, + "grad_norm": 2.038159634583347, + "learning_rate": 3.829223177805959e-06, + "loss": 0.7105, + "num_input_tokens_seen": 57661335, + "step": 1758 + }, + { + "epoch": 0.15863281778419083, + "flos": 26825955915360.0, + "grad_norm": 1.740945073667454, + "learning_rate": 3.828986883796591e-06, + "loss": 0.8272, + "num_input_tokens_seen": 57687270, + "step": 1759 + }, + { + "epoch": 0.1587230013076611, + "flos": 32515719294720.0, + "grad_norm": 1.7331565494740344, + "learning_rate": 3.828750433727979e-06, + "loss": 0.7783, + "num_input_tokens_seen": 57715520, + "step": 1760 + }, + { + "epoch": 0.15881318483113135, + "flos": 23698965633120.0, + "grad_norm": 2.0123076826812127, + "learning_rate": 3.828513827620296e-06, + "loss": 0.8342, + "num_input_tokens_seen": 57745540, + "step": 1761 + }, + { + "epoch": 0.15890336835460162, + "flos": 24457869519360.0, + "grad_norm": 2.111563444477506, + "learning_rate": 3.82827706549373e-06, + "loss": 0.8347, + "num_input_tokens_seen": 57769845, + "step": 1762 + }, + { + "epoch": 0.15899355187807188, + "flos": 19399420475520.0, + "grad_norm": 2.0917861610515773, + "learning_rate": 3.828040147368484e-06, + "loss": 0.7768, + "num_input_tokens_seen": 57797950, + "step": 1763 + }, + { + "epoch": 0.15908373540154214, + "flos": 24755607996000.0, + "grad_norm": 1.9055409598030961, + "learning_rate": 3.827803073264774e-06, + "loss": 0.7836, + "num_input_tokens_seen": 57826685, + "step": 1764 + }, + { + "epoch": 0.1591739189250124, + "flos": 15465571583040.0, + "grad_norm": 2.2905342656545113, + "learning_rate": 3.827565843202826e-06, + "loss": 0.7847, + "num_input_tokens_seen": 57853435, + "step": 1765 + }, + { + "epoch": 0.15926410244848266, + "flos": 37069657255200.0, + "grad_norm": 1.976436264168463, + "learning_rate": 3.827328457202884e-06, + "loss": 0.6577, + "num_input_tokens_seen": 57881570, + "step": 1766 + }, + { + "epoch": 0.15935428597195292, + "flos": 34517291690880.0, + "grad_norm": 1.4550401925896725, + "learning_rate": 3.8270909152852014e-06, + "loss": 0.7166, + "num_input_tokens_seen": 57913730, + "step": 1767 + }, + { + "epoch": 0.15944446949542318, + "flos": 26759113219680.0, + "grad_norm": 1.9649306291366462, + "learning_rate": 3.826853217470048e-06, + "loss": 0.7813, + "num_input_tokens_seen": 57939690, + "step": 1768 + }, + { + "epoch": 0.15953465301889344, + "flos": 24973754549280.0, + "grad_norm": 1.7952083653725992, + "learning_rate": 3.826615363777705e-06, + "loss": 0.865, + "num_input_tokens_seen": 57969415, + "step": 1769 + }, + { + "epoch": 0.1596248365423637, + "flos": 22676525183040.0, + "grad_norm": 1.89089478388539, + "learning_rate": 3.826377354228468e-06, + "loss": 0.7646, + "num_input_tokens_seen": 57998770, + "step": 1770 + }, + { + "epoch": 0.15971502006583396, + "flos": 63914532578400.0, + "grad_norm": 0.6536775761124655, + "learning_rate": 3.826139188842643e-06, + "loss": 0.5912, + "num_input_tokens_seen": 58089690, + "step": 1771 + }, + { + "epoch": 0.15980520358930422, + "flos": 28613210243520.0, + "grad_norm": 1.5869301381397296, + "learning_rate": 3.825900867640554e-06, + "loss": 0.8397, + "num_input_tokens_seen": 58118960, + "step": 1772 + }, + { + "epoch": 0.15989538711277448, + "flos": 23766774742560.0, + "grad_norm": 1.461378224851649, + "learning_rate": 3.825662390642535e-06, + "loss": 0.8247, + "num_input_tokens_seen": 58150000, + "step": 1773 + }, + { + "epoch": 0.15998557063624474, + "flos": 20638743707040.0, + "grad_norm": 1.9200387261379352, + "learning_rate": 3.825423757868934e-06, + "loss": 0.8883, + "num_input_tokens_seen": 58177120, + "step": 1774 + }, + { + "epoch": 0.16007575415971503, + "flos": 19614556278240.0, + "grad_norm": 1.6950460479467908, + "learning_rate": 3.825184969340114e-06, + "loss": 0.8107, + "num_input_tokens_seen": 58205390, + "step": 1775 + }, + { + "epoch": 0.1601659376831853, + "flos": 28762841461920.0, + "grad_norm": 1.6508389738635787, + "learning_rate": 3.824946025076447e-06, + "loss": 0.8138, + "num_input_tokens_seen": 58238950, + "step": 1776 + }, + { + "epoch": 0.16025612120665556, + "flos": 23552977051200.0, + "grad_norm": 1.914482849052241, + "learning_rate": 3.824706925098323e-06, + "loss": 0.734, + "num_input_tokens_seen": 58267035, + "step": 1777 + }, + { + "epoch": 0.16034630473012582, + "flos": 13896649656960.0, + "grad_norm": 2.373704657420375, + "learning_rate": 3.824467669426143e-06, + "loss": 0.8448, + "num_input_tokens_seen": 58290725, + "step": 1778 + }, + { + "epoch": 0.16043648825359608, + "flos": 22606374378720.0, + "grad_norm": 1.8260407263025653, + "learning_rate": 3.824228258080321e-06, + "loss": 0.7526, + "num_input_tokens_seen": 58320300, + "step": 1779 + }, + { + "epoch": 0.16052667177706634, + "flos": 25265768882880.0, + "grad_norm": 1.761523589089693, + "learning_rate": 3.823988691081285e-06, + "loss": 0.6994, + "num_input_tokens_seen": 58348095, + "step": 1780 + }, + { + "epoch": 0.1606168553005366, + "flos": 23366133169440.0, + "grad_norm": 1.663788822216955, + "learning_rate": 3.823748968449478e-06, + "loss": 0.8172, + "num_input_tokens_seen": 58376335, + "step": 1781 + }, + { + "epoch": 0.16070703882400686, + "flos": 16047964780800.0, + "grad_norm": 1.935930970360287, + "learning_rate": 3.823509090205352e-06, + "loss": 0.6782, + "num_input_tokens_seen": 58403230, + "step": 1782 + }, + { + "epoch": 0.16079722234747712, + "flos": 27450691203360.0, + "grad_norm": 2.201563396683802, + "learning_rate": 3.823269056369376e-06, + "loss": 0.7407, + "num_input_tokens_seen": 58433080, + "step": 1783 + }, + { + "epoch": 0.16088740587094738, + "flos": 21330842067360.0, + "grad_norm": 1.8652285590755204, + "learning_rate": 3.8230288669620295e-06, + "loss": 0.8196, + "num_input_tokens_seen": 58461295, + "step": 1784 + }, + { + "epoch": 0.16097758939441764, + "flos": 20602385948160.0, + "grad_norm": 1.4895846048985457, + "learning_rate": 3.822788522003809e-06, + "loss": 0.7899, + "num_input_tokens_seen": 58492595, + "step": 1785 + }, + { + "epoch": 0.1610677729178879, + "flos": 14007432742560.0, + "grad_norm": 1.9671510530932887, + "learning_rate": 3.822548021515221e-06, + "loss": 0.8596, + "num_input_tokens_seen": 58515410, + "step": 1786 + }, + { + "epoch": 0.16115795644135816, + "flos": 22161606566880.0, + "grad_norm": 1.661723037090071, + "learning_rate": 3.822307365516787e-06, + "loss": 0.7608, + "num_input_tokens_seen": 58542965, + "step": 1787 + }, + { + "epoch": 0.16124813996482842, + "flos": 24607017530880.0, + "grad_norm": 1.905137910231365, + "learning_rate": 3.8220665540290395e-06, + "loss": 0.8035, + "num_input_tokens_seen": 58571885, + "step": 1788 + }, + { + "epoch": 0.16133832348829868, + "flos": 27087559651680.0, + "grad_norm": 1.4636196871377722, + "learning_rate": 3.8218255870725265e-06, + "loss": 0.8505, + "num_input_tokens_seen": 58601375, + "step": 1789 + }, + { + "epoch": 0.16142850701176895, + "flos": 28718975411520.0, + "grad_norm": 1.681437233067874, + "learning_rate": 3.82158446466781e-06, + "loss": 0.7353, + "num_input_tokens_seen": 58630505, + "step": 1790 + }, + { + "epoch": 0.1615186905352392, + "flos": 27231801254880.0, + "grad_norm": 1.5864328447057279, + "learning_rate": 3.821343186835462e-06, + "loss": 0.8001, + "num_input_tokens_seen": 58660550, + "step": 1791 + }, + { + "epoch": 0.16160887405870947, + "flos": 18999856825440.0, + "grad_norm": 1.5816955210174508, + "learning_rate": 3.821101753596072e-06, + "loss": 0.8531, + "num_input_tokens_seen": 58688055, + "step": 1792 + }, + { + "epoch": 0.16169905758217973, + "flos": 13971037813920.0, + "grad_norm": 2.127361556142199, + "learning_rate": 3.820860164970237e-06, + "loss": 0.7725, + "num_input_tokens_seen": 58712410, + "step": 1793 + }, + { + "epoch": 0.16178924110565, + "flos": 27268382032320.0, + "grad_norm": 1.528173899111984, + "learning_rate": 3.820618420978574e-06, + "loss": 0.8538, + "num_input_tokens_seen": 58742390, + "step": 1794 + }, + { + "epoch": 0.16187942462912025, + "flos": 20383570339200.0, + "grad_norm": 2.058694855091005, + "learning_rate": 3.820376521641708e-06, + "loss": 0.7832, + "num_input_tokens_seen": 58770865, + "step": 1795 + }, + { + "epoch": 0.1619696081525905, + "flos": 30110829102720.0, + "grad_norm": 4.480677126734093, + "learning_rate": 3.82013446698028e-06, + "loss": 0.7917, + "num_input_tokens_seen": 58800985, + "step": 1796 + }, + { + "epoch": 0.16205979167606077, + "flos": 33065508879360.0, + "grad_norm": 3.2355043954903393, + "learning_rate": 3.819892257014943e-06, + "loss": 0.7529, + "num_input_tokens_seen": 58834935, + "step": 1797 + }, + { + "epoch": 0.16214997519953103, + "flos": 25593248901120.0, + "grad_norm": 1.7215673071350466, + "learning_rate": 3.819649891766364e-06, + "loss": 0.8111, + "num_input_tokens_seen": 58864825, + "step": 1798 + }, + { + "epoch": 0.16224015872300132, + "flos": 21622521873120.0, + "grad_norm": 1.528090506634363, + "learning_rate": 3.819407371255222e-06, + "loss": 0.7798, + "num_input_tokens_seen": 58893880, + "step": 1799 + }, + { + "epoch": 0.16233034224647158, + "flos": 27852187680960.0, + "grad_norm": 1.7339393824725342, + "learning_rate": 3.819164695502212e-06, + "loss": 0.7979, + "num_input_tokens_seen": 58923280, + "step": 1800 + }, + { + "epoch": 0.16242052576994184, + "flos": 23005677840480.0, + "grad_norm": 1.6209344624773478, + "learning_rate": 3.818921864528039e-06, + "loss": 0.7286, + "num_input_tokens_seen": 58949505, + "step": 1801 + }, + { + "epoch": 0.1625107092934121, + "flos": 21439841004480.0, + "grad_norm": 2.544725224588786, + "learning_rate": 3.818678878353423e-06, + "loss": 0.8383, + "num_input_tokens_seen": 58977185, + "step": 1802 + }, + { + "epoch": 0.16260089281688236, + "flos": 20638223330400.0, + "grad_norm": 2.199257885870976, + "learning_rate": 3.818435736999097e-06, + "loss": 0.7297, + "num_input_tokens_seen": 59004615, + "step": 1803 + }, + { + "epoch": 0.16269107634035262, + "flos": 24536532198720.0, + "grad_norm": 1.913994425906014, + "learning_rate": 3.818192440485807e-06, + "loss": 0.7811, + "num_input_tokens_seen": 59033345, + "step": 1804 + }, + { + "epoch": 0.16278125986382289, + "flos": 19210457917440.0, + "grad_norm": 2.4502367644834586, + "learning_rate": 3.817948988834314e-06, + "loss": 0.7185, + "num_input_tokens_seen": 59060955, + "step": 1805 + }, + { + "epoch": 0.16287144338729315, + "flos": 21727097608800.0, + "grad_norm": 1.5498516842423353, + "learning_rate": 3.817705382065388e-06, + "loss": 0.782, + "num_input_tokens_seen": 59090975, + "step": 1806 + }, + { + "epoch": 0.1629616269107634, + "flos": 28653210638880.0, + "grad_norm": 1.2469030640008507, + "learning_rate": 3.8174616201998155e-06, + "loss": 0.7503, + "num_input_tokens_seen": 59123635, + "step": 1807 + }, + { + "epoch": 0.16305181043423367, + "flos": 17541011759520.0, + "grad_norm": 3.2547897842897613, + "learning_rate": 3.817217703258397e-06, + "loss": 0.8345, + "num_input_tokens_seen": 59148125, + "step": 1808 + }, + { + "epoch": 0.16314199395770393, + "flos": 27809622572160.0, + "grad_norm": 1.5811772054569857, + "learning_rate": 3.816973631261943e-06, + "loss": 0.7722, + "num_input_tokens_seen": 59175160, + "step": 1809 + }, + { + "epoch": 0.1632321774811742, + "flos": 19982928766080.0, + "grad_norm": 2.7323742331331804, + "learning_rate": 3.816729404231281e-06, + "loss": 0.8159, + "num_input_tokens_seen": 59203015, + "step": 1810 + }, + { + "epoch": 0.16332236100464445, + "flos": 23583350478720.0, + "grad_norm": 2.624478333726617, + "learning_rate": 3.816485022187249e-06, + "loss": 0.8527, + "num_input_tokens_seen": 59225565, + "step": 1811 + }, + { + "epoch": 0.1634125445281147, + "flos": 31234871707680.0, + "grad_norm": 2.881099864167698, + "learning_rate": 3.816240485150698e-06, + "loss": 0.7405, + "num_input_tokens_seen": 59254680, + "step": 1812 + }, + { + "epoch": 0.16350272805158497, + "flos": 22351052331840.0, + "grad_norm": 1.6347363832836843, + "learning_rate": 3.815995793142495e-06, + "loss": 0.7852, + "num_input_tokens_seen": 59284710, + "step": 1813 + }, + { + "epoch": 0.16359291157505523, + "flos": 25374544801440.0, + "grad_norm": 1.8287771643018056, + "learning_rate": 3.815750946183518e-06, + "loss": 0.7239, + "num_input_tokens_seen": 59313870, + "step": 1814 + }, + { + "epoch": 0.1636830950985255, + "flos": 15100618713120.0, + "grad_norm": 1.8654897465072624, + "learning_rate": 3.815505944294658e-06, + "loss": 0.8407, + "num_input_tokens_seen": 59340225, + "step": 1815 + }, + { + "epoch": 0.16377327862199575, + "flos": 27598017896640.0, + "grad_norm": 1.7792602937289912, + "learning_rate": 3.81526078749682e-06, + "loss": 0.7406, + "num_input_tokens_seen": 59372750, + "step": 1816 + }, + { + "epoch": 0.16386346214546602, + "flos": 20784100403040.0, + "grad_norm": 1.9189710309583807, + "learning_rate": 3.8150154758109225e-06, + "loss": 0.7232, + "num_input_tokens_seen": 59402290, + "step": 1817 + }, + { + "epoch": 0.16395364566893628, + "flos": 32332964086560.0, + "grad_norm": 1.59341023133207, + "learning_rate": 3.814770009257896e-06, + "loss": 0.7589, + "num_input_tokens_seen": 59433280, + "step": 1818 + }, + { + "epoch": 0.16404382919240654, + "flos": 28323426095520.0, + "grad_norm": 1.8282091822356161, + "learning_rate": 3.814524387858687e-06, + "loss": 0.7976, + "num_input_tokens_seen": 59463530, + "step": 1819 + }, + { + "epoch": 0.1641340127158768, + "flos": 19035694207680.0, + "grad_norm": 1.7860943022049587, + "learning_rate": 3.814278611634251e-06, + "loss": 0.8092, + "num_input_tokens_seen": 59492110, + "step": 1820 + }, + { + "epoch": 0.16422419623934706, + "flos": 23438068122240.0, + "grad_norm": 1.7809791358997937, + "learning_rate": 3.8140326806055606e-06, + "loss": 0.7585, + "num_input_tokens_seen": 59520610, + "step": 1821 + }, + { + "epoch": 0.16431437976281732, + "flos": 19581729646560.0, + "grad_norm": 3.3943796771485597, + "learning_rate": 3.8137865947935992e-06, + "loss": 0.8442, + "num_input_tokens_seen": 59548775, + "step": 1822 + }, + { + "epoch": 0.1644045632862876, + "flos": 24281470340160.0, + "grad_norm": 2.5817813723525207, + "learning_rate": 3.8135403542193646e-06, + "loss": 0.8014, + "num_input_tokens_seen": 59577515, + "step": 1823 + }, + { + "epoch": 0.16449474680975787, + "flos": 16959807994080.0, + "grad_norm": 2.0894997572103424, + "learning_rate": 3.813293958903867e-06, + "loss": 0.7201, + "num_input_tokens_seen": 59603105, + "step": 1824 + }, + { + "epoch": 0.16458493033322813, + "flos": 44137112647200.0, + "grad_norm": 1.9092371314887449, + "learning_rate": 3.8130474088681306e-06, + "loss": 0.7498, + "num_input_tokens_seen": 59634675, + "step": 1825 + }, + { + "epoch": 0.1646751138566984, + "flos": 25629272132160.0, + "grad_norm": 1.8162741336151134, + "learning_rate": 3.8128007041331927e-06, + "loss": 0.7146, + "num_input_tokens_seen": 59662955, + "step": 1826 + }, + { + "epoch": 0.16476529738016865, + "flos": 20638148990880.0, + "grad_norm": 1.581718432735331, + "learning_rate": 3.812553844720102e-06, + "loss": 0.7007, + "num_input_tokens_seen": 59693975, + "step": 1827 + }, + { + "epoch": 0.1648554809036389, + "flos": 35095596215040.0, + "grad_norm": 1.8179985040673206, + "learning_rate": 3.8123068306499236e-06, + "loss": 0.8012, + "num_input_tokens_seen": 59721815, + "step": 1828 + }, + { + "epoch": 0.16494566442710917, + "flos": 31197324516480.0, + "grad_norm": 4.879623958920273, + "learning_rate": 3.812059661943733e-06, + "loss": 0.781, + "num_input_tokens_seen": 59750325, + "step": 1829 + }, + { + "epoch": 0.16503584795057943, + "flos": 23948563536960.0, + "grad_norm": 1.4331232696935337, + "learning_rate": 3.811812338622621e-06, + "loss": 0.754, + "num_input_tokens_seen": 59777725, + "step": 1830 + }, + { + "epoch": 0.1651260314740497, + "flos": 20784360591360.0, + "grad_norm": 1.6851337231512915, + "learning_rate": 3.81156486070769e-06, + "loss": 0.7763, + "num_input_tokens_seen": 59806440, + "step": 1831 + }, + { + "epoch": 0.16521621499751996, + "flos": 21002953181760.0, + "grad_norm": 1.8034820494026826, + "learning_rate": 3.811317228220056e-06, + "loss": 0.8281, + "num_input_tokens_seen": 59835600, + "step": 1832 + }, + { + "epoch": 0.16530639852099022, + "flos": 23990013552960.0, + "grad_norm": 7.407194053261719, + "learning_rate": 3.811069441180849e-06, + "loss": 0.8291, + "num_input_tokens_seen": 59865540, + "step": 1833 + }, + { + "epoch": 0.16539658204446048, + "flos": 18307238088480.0, + "grad_norm": 2.179682030893116, + "learning_rate": 3.8108214996112107e-06, + "loss": 0.7873, + "num_input_tokens_seen": 59891930, + "step": 1834 + }, + { + "epoch": 0.16548676556793074, + "flos": 21002172616800.0, + "grad_norm": 1.9842740130822483, + "learning_rate": 3.810573403532297e-06, + "loss": 0.8224, + "num_input_tokens_seen": 59916140, + "step": 1835 + }, + { + "epoch": 0.165576949091401, + "flos": 22457932592640.0, + "grad_norm": 1.9799717386714135, + "learning_rate": 3.8103251529652774e-06, + "loss": 0.7651, + "num_input_tokens_seen": 59944890, + "step": 1836 + }, + { + "epoch": 0.16566713261487126, + "flos": 23371039577760.0, + "grad_norm": 1.6782987309323163, + "learning_rate": 3.810076747931334e-06, + "loss": 0.7613, + "num_input_tokens_seen": 59973880, + "step": 1837 + }, + { + "epoch": 0.16575731613834152, + "flos": 23625878417760.0, + "grad_norm": 2.7642461928101363, + "learning_rate": 3.809828188451662e-06, + "loss": 0.7403, + "num_input_tokens_seen": 60003025, + "step": 1838 + }, + { + "epoch": 0.16584749966181178, + "flos": 28435138425120.0, + "grad_norm": 1.6728161453443944, + "learning_rate": 3.809579474547469e-06, + "loss": 0.7793, + "num_input_tokens_seen": 60033335, + "step": 1839 + }, + { + "epoch": 0.16593768318528204, + "flos": 18010354516320.0, + "grad_norm": 1.8384356445635133, + "learning_rate": 3.809330606239977e-06, + "loss": 0.8375, + "num_input_tokens_seen": 60060295, + "step": 1840 + }, + { + "epoch": 0.1660278667087523, + "flos": 25442948627040.0, + "grad_norm": 1.693492633367689, + "learning_rate": 3.809081583550422e-06, + "loss": 0.8492, + "num_input_tokens_seen": 60091155, + "step": 1841 + }, + { + "epoch": 0.16611805023222256, + "flos": 19181199582720.0, + "grad_norm": 2.209063055524326, + "learning_rate": 3.808832406500051e-06, + "loss": 0.863, + "num_input_tokens_seen": 60119515, + "step": 1842 + }, + { + "epoch": 0.16620823375569282, + "flos": 17939200128480.0, + "grad_norm": 1.662480493302024, + "learning_rate": 3.8085830751101253e-06, + "loss": 0.8554, + "num_input_tokens_seen": 60143810, + "step": 1843 + }, + { + "epoch": 0.16629841727916309, + "flos": 18086787010080.0, + "grad_norm": 1.750588386258991, + "learning_rate": 3.808333589401919e-06, + "loss": 0.8195, + "num_input_tokens_seen": 60172245, + "step": 1844 + }, + { + "epoch": 0.16638860080263335, + "flos": 18197644435200.0, + "grad_norm": 1.7911951272549782, + "learning_rate": 3.8080839493967194e-06, + "loss": 0.714, + "num_input_tokens_seen": 60199395, + "step": 1845 + }, + { + "epoch": 0.1664787843261036, + "flos": 67325688641280.0, + "grad_norm": 0.6620265267451007, + "learning_rate": 3.807834155115828e-06, + "loss": 0.5447, + "num_input_tokens_seen": 60298835, + "step": 1846 + }, + { + "epoch": 0.1665689678495739, + "flos": 70305092041920.0, + "grad_norm": 0.8505212559945216, + "learning_rate": 3.8075842065805584e-06, + "loss": 0.6496, + "num_input_tokens_seen": 60404460, + "step": 1847 + }, + { + "epoch": 0.16665915137304416, + "flos": 23151852271200.0, + "grad_norm": 1.8952895701657446, + "learning_rate": 3.8073341038122374e-06, + "loss": 0.7701, + "num_input_tokens_seen": 60434200, + "step": 1848 + }, + { + "epoch": 0.16674933489651442, + "flos": 25662024424320.0, + "grad_norm": 1.697481657443154, + "learning_rate": 3.8070838468322048e-06, + "loss": 0.8504, + "num_input_tokens_seen": 60463630, + "step": 1849 + }, + { + "epoch": 0.16683951841998468, + "flos": 20565210454560.0, + "grad_norm": 1.8741804239060174, + "learning_rate": 3.8068334356618143e-06, + "loss": 0.7206, + "num_input_tokens_seen": 60490845, + "step": 1850 + }, + { + "epoch": 0.16692970194345494, + "flos": 23516396273760.0, + "grad_norm": 2.798524593870195, + "learning_rate": 3.8065828703224324e-06, + "loss": 0.8155, + "num_input_tokens_seen": 60519075, + "step": 1851 + }, + { + "epoch": 0.1670198854669252, + "flos": 18962644162080.0, + "grad_norm": 1.9839246343941734, + "learning_rate": 3.8063321508354386e-06, + "loss": 0.7818, + "num_input_tokens_seen": 60542685, + "step": 1852 + }, + { + "epoch": 0.16711006899039546, + "flos": 19361204228640.0, + "grad_norm": 1.6698595537301812, + "learning_rate": 3.8060812772222255e-06, + "loss": 0.7638, + "num_input_tokens_seen": 60570970, + "step": 1853 + }, + { + "epoch": 0.16720025251386572, + "flos": 21732004017120.0, + "grad_norm": 1.5614341152044384, + "learning_rate": 3.8058302495041993e-06, + "loss": 0.8514, + "num_input_tokens_seen": 60598765, + "step": 1854 + }, + { + "epoch": 0.16729043603733598, + "flos": 18051990381120.0, + "grad_norm": 3.422954789590004, + "learning_rate": 3.805579067702779e-06, + "loss": 0.8013, + "num_input_tokens_seen": 60627300, + "step": 1855 + }, + { + "epoch": 0.16738061956080624, + "flos": 21658247746080.0, + "grad_norm": 1.7981240616664353, + "learning_rate": 3.8053277318393967e-06, + "loss": 0.7826, + "num_input_tokens_seen": 60654780, + "step": 1856 + }, + { + "epoch": 0.1674708030842765, + "flos": 23551118563200.0, + "grad_norm": 4.684581203897335, + "learning_rate": 3.805076241935498e-06, + "loss": 0.7993, + "num_input_tokens_seen": 60683710, + "step": 1857 + }, + { + "epoch": 0.16756098660774676, + "flos": 20056164660480.0, + "grad_norm": 6.617151757946565, + "learning_rate": 3.804824598012541e-06, + "loss": 0.662, + "num_input_tokens_seen": 60711935, + "step": 1858 + }, + { + "epoch": 0.16765117013121703, + "flos": 29602824061920.0, + "grad_norm": 1.9954306141091143, + "learning_rate": 3.8045728000919975e-06, + "loss": 0.8248, + "num_input_tokens_seen": 60743415, + "step": 1859 + }, + { + "epoch": 0.1677413536546873, + "flos": 25008848536320.0, + "grad_norm": 2.448625848727834, + "learning_rate": 3.8043208481953524e-06, + "loss": 0.8206, + "num_input_tokens_seen": 60774485, + "step": 1860 + }, + { + "epoch": 0.16783153717815755, + "flos": 37579520784000.0, + "grad_norm": 2.392931654365795, + "learning_rate": 3.804068742344104e-06, + "loss": 0.7107, + "num_input_tokens_seen": 60806475, + "step": 1861 + }, + { + "epoch": 0.1679217207016278, + "flos": 30401802683040.0, + "grad_norm": 2.156052843008938, + "learning_rate": 3.8038164825597628e-06, + "loss": 0.7551, + "num_input_tokens_seen": 60833545, + "step": 1862 + }, + { + "epoch": 0.16801190422509807, + "flos": 21112137967680.0, + "grad_norm": 2.1588516020596638, + "learning_rate": 3.8035640688638537e-06, + "loss": 0.8655, + "num_input_tokens_seen": 60858915, + "step": 1863 + }, + { + "epoch": 0.16810208774856833, + "flos": 22752028432800.0, + "grad_norm": 1.9923059764862603, + "learning_rate": 3.8033115012779125e-06, + "loss": 0.8598, + "num_input_tokens_seen": 60885940, + "step": 1864 + }, + { + "epoch": 0.1681922712720386, + "flos": 25954744983360.0, + "grad_norm": 1.6366608456950318, + "learning_rate": 3.8030587798234915e-06, + "loss": 0.8273, + "num_input_tokens_seen": 60917100, + "step": 1865 + }, + { + "epoch": 0.16828245479550885, + "flos": 23472641732640.0, + "grad_norm": 1.7586640054573206, + "learning_rate": 3.802805904522153e-06, + "loss": 0.7183, + "num_input_tokens_seen": 60941995, + "step": 1866 + }, + { + "epoch": 0.1683726383189791, + "flos": 30693556828320.0, + "grad_norm": 1.4281382687409276, + "learning_rate": 3.8025528753954742e-06, + "loss": 0.775, + "num_input_tokens_seen": 60973690, + "step": 1867 + }, + { + "epoch": 0.16846282184244937, + "flos": 23881460652960.0, + "grad_norm": 1.967221287201624, + "learning_rate": 3.802299692465045e-06, + "loss": 0.7264, + "num_input_tokens_seen": 61001880, + "step": 1868 + }, + { + "epoch": 0.16855300536591963, + "flos": 26066011275840.0, + "grad_norm": 1.9211966705366486, + "learning_rate": 3.802046355752468e-06, + "loss": 0.7725, + "num_input_tokens_seen": 61029600, + "step": 1869 + }, + { + "epoch": 0.16864318888938992, + "flos": 23479927005600.0, + "grad_norm": 2.0928871312589825, + "learning_rate": 3.80179286527936e-06, + "loss": 0.8408, + "num_input_tokens_seen": 61055300, + "step": 1870 + }, + { + "epoch": 0.16873337241286018, + "flos": 21692189470560.0, + "grad_norm": 1.6763576414406927, + "learning_rate": 3.801539221067349e-06, + "loss": 0.6955, + "num_input_tokens_seen": 61084325, + "step": 1871 + }, + { + "epoch": 0.16882355593633044, + "flos": 22605779662560.0, + "grad_norm": 1.888590829682398, + "learning_rate": 3.801285423138079e-06, + "loss": 0.7382, + "num_input_tokens_seen": 61113205, + "step": 1872 + }, + { + "epoch": 0.1689137394598007, + "flos": 27997618716480.0, + "grad_norm": 1.9312408002615755, + "learning_rate": 3.8010314715132037e-06, + "loss": 0.7359, + "num_input_tokens_seen": 61143825, + "step": 1873 + }, + { + "epoch": 0.16900392298327097, + "flos": 31896001924320.0, + "grad_norm": 1.9817856843636699, + "learning_rate": 3.800777366214393e-06, + "loss": 0.7349, + "num_input_tokens_seen": 61175815, + "step": 1874 + }, + { + "epoch": 0.16909410650674123, + "flos": 20711905261920.0, + "grad_norm": 2.038844539057198, + "learning_rate": 3.800523107263328e-06, + "loss": 0.761, + "num_input_tokens_seen": 61204830, + "step": 1875 + }, + { + "epoch": 0.1691842900302115, + "flos": 26096236024320.0, + "grad_norm": 2.148592523703892, + "learning_rate": 3.800268694681703e-06, + "loss": 0.8383, + "num_input_tokens_seen": 61233600, + "step": 1876 + }, + { + "epoch": 0.16927447355368175, + "flos": 69869319312000.0, + "grad_norm": 0.7475490656668472, + "learning_rate": 3.800014128491227e-06, + "loss": 0.5989, + "num_input_tokens_seen": 61329760, + "step": 1877 + }, + { + "epoch": 0.169364657077152, + "flos": 23404795453440.0, + "grad_norm": 2.3078768112323447, + "learning_rate": 3.79975940871362e-06, + "loss": 0.8359, + "num_input_tokens_seen": 61359785, + "step": 1878 + }, + { + "epoch": 0.16945484060062227, + "flos": 25301531925600.0, + "grad_norm": 2.077863404294011, + "learning_rate": 3.799504535370617e-06, + "loss": 0.7877, + "num_input_tokens_seen": 61386115, + "step": 1879 + }, + { + "epoch": 0.16954502412409253, + "flos": 19325032318560.0, + "grad_norm": 5.5702275808744455, + "learning_rate": 3.799249508483964e-06, + "loss": 0.7238, + "num_input_tokens_seen": 61412340, + "step": 1880 + }, + { + "epoch": 0.1696352076475628, + "flos": 61931099025120.0, + "grad_norm": 0.7199270313428082, + "learning_rate": 3.798994328075422e-06, + "loss": 0.6588, + "num_input_tokens_seen": 61511010, + "step": 1881 + }, + { + "epoch": 0.16972539117103305, + "flos": 20383124302080.0, + "grad_norm": 2.5027921343396464, + "learning_rate": 3.798738994166765e-06, + "loss": 0.7151, + "num_input_tokens_seen": 61538640, + "step": 1882 + }, + { + "epoch": 0.1698155746945033, + "flos": 21184816315680.0, + "grad_norm": 1.7012420885068873, + "learning_rate": 3.7984835067797788e-06, + "loss": 0.7868, + "num_input_tokens_seen": 61566995, + "step": 1883 + }, + { + "epoch": 0.16990575821797357, + "flos": 27558203350080.0, + "grad_norm": 1.7015181494503477, + "learning_rate": 3.798227865936263e-06, + "loss": 0.8757, + "num_input_tokens_seen": 61597830, + "step": 1884 + }, + { + "epoch": 0.16999594174144383, + "flos": 39144614224800.0, + "grad_norm": 1.9524509945433046, + "learning_rate": 3.7979720716580297e-06, + "loss": 0.6534, + "num_input_tokens_seen": 61629550, + "step": 1885 + }, + { + "epoch": 0.1700861252649141, + "flos": 23552642523360.0, + "grad_norm": 1.6221083074502896, + "learning_rate": 3.7977161239669057e-06, + "loss": 0.7752, + "num_input_tokens_seen": 61658775, + "step": 1886 + }, + { + "epoch": 0.17017630878838436, + "flos": 24463556492640.0, + "grad_norm": 2.7076909305954144, + "learning_rate": 3.7974600228847294e-06, + "loss": 0.7625, + "num_input_tokens_seen": 61686145, + "step": 1887 + }, + { + "epoch": 0.17026649231185462, + "flos": 14512910239680.0, + "grad_norm": 2.4257488156421627, + "learning_rate": 3.7972037684333534e-06, + "loss": 0.7255, + "num_input_tokens_seen": 61710575, + "step": 1888 + }, + { + "epoch": 0.17035667583532488, + "flos": 36486632171520.0, + "grad_norm": 1.6118976118403174, + "learning_rate": 3.796947360634642e-06, + "loss": 0.7093, + "num_input_tokens_seen": 61741415, + "step": 1889 + }, + { + "epoch": 0.17044685935879514, + "flos": 25411274257920.0, + "grad_norm": 1.9382447072850744, + "learning_rate": 3.796690799510473e-06, + "loss": 0.8304, + "num_input_tokens_seen": 61770485, + "step": 1890 + }, + { + "epoch": 0.1705370428822654, + "flos": 21182771978880.0, + "grad_norm": 1.80361715175183, + "learning_rate": 3.7964340850827387e-06, + "loss": 0.7477, + "num_input_tokens_seen": 61799915, + "step": 1891 + }, + { + "epoch": 0.17062722640573566, + "flos": 59498591701440.0, + "grad_norm": 0.677707132896801, + "learning_rate": 3.7961772173733425e-06, + "loss": 0.5658, + "num_input_tokens_seen": 61885820, + "step": 1892 + }, + { + "epoch": 0.17071740992920592, + "flos": 28029999311040.0, + "grad_norm": 1.4454901508746556, + "learning_rate": 3.7959201964042024e-06, + "loss": 0.7698, + "num_input_tokens_seen": 61915885, + "step": 1893 + }, + { + "epoch": 0.1708075934526762, + "flos": 20126352634560.0, + "grad_norm": 2.311115262698251, + "learning_rate": 3.795663022197248e-06, + "loss": 0.8098, + "num_input_tokens_seen": 61945140, + "step": 1894 + }, + { + "epoch": 0.17089777697614647, + "flos": 22751805414240.0, + "grad_norm": 2.4463345765164686, + "learning_rate": 3.7954056947744242e-06, + "loss": 0.7831, + "num_input_tokens_seen": 61973650, + "step": 1895 + }, + { + "epoch": 0.17098796049961673, + "flos": 18961937936640.0, + "grad_norm": 1.8851204809849633, + "learning_rate": 3.7951482141576863e-06, + "loss": 0.7581, + "num_input_tokens_seen": 62001945, + "step": 1896 + }, + { + "epoch": 0.171078144023087, + "flos": 22788237512640.0, + "grad_norm": 1.4612034846229054, + "learning_rate": 3.794890580369004e-06, + "loss": 0.8343, + "num_input_tokens_seen": 62028950, + "step": 1897 + }, + { + "epoch": 0.17116832754655725, + "flos": 21366344921760.0, + "grad_norm": 1.840400109533627, + "learning_rate": 3.7946327934303612e-06, + "loss": 0.8132, + "num_input_tokens_seen": 62052230, + "step": 1898 + }, + { + "epoch": 0.1712585110700275, + "flos": 22967387254080.0, + "grad_norm": 2.4698011840310867, + "learning_rate": 3.794374853363752e-06, + "loss": 0.8786, + "num_input_tokens_seen": 62079035, + "step": 1899 + }, + { + "epoch": 0.17134869459349777, + "flos": 21476050084320.0, + "grad_norm": 1.7404839778733634, + "learning_rate": 3.794116760191187e-06, + "loss": 0.8066, + "num_input_tokens_seen": 62107755, + "step": 1900 + }, + { + "epoch": 0.17143887811696804, + "flos": 18999150600000.0, + "grad_norm": 2.134261425118655, + "learning_rate": 3.7938585139346877e-06, + "loss": 0.8739, + "num_input_tokens_seen": 62134880, + "step": 1901 + }, + { + "epoch": 0.1715290616404383, + "flos": 19399457645280.0, + "grad_norm": 2.0151471894983297, + "learning_rate": 3.793600114616288e-06, + "loss": 0.7635, + "num_input_tokens_seen": 62163280, + "step": 1902 + }, + { + "epoch": 0.17161924516390856, + "flos": 56823208466880.0, + "grad_norm": 0.73886997223384, + "learning_rate": 3.793341562258037e-06, + "loss": 0.5556, + "num_input_tokens_seen": 62249405, + "step": 1903 + }, + { + "epoch": 0.17170942868737882, + "flos": 24206598976320.0, + "grad_norm": 4.035514763916747, + "learning_rate": 3.7930828568819953e-06, + "loss": 0.7909, + "num_input_tokens_seen": 62279925, + "step": 1904 + }, + { + "epoch": 0.17179961221084908, + "flos": 25591427582880.0, + "grad_norm": 1.804730303368209, + "learning_rate": 3.7928239985102378e-06, + "loss": 0.7805, + "num_input_tokens_seen": 62307350, + "step": 1905 + }, + { + "epoch": 0.17188979573431934, + "flos": 19290867575520.0, + "grad_norm": 1.624832144650953, + "learning_rate": 3.7925649871648505e-06, + "loss": 0.7844, + "num_input_tokens_seen": 62335225, + "step": 1906 + }, + { + "epoch": 0.1719799792577896, + "flos": 21076337755200.0, + "grad_norm": 1.555395990468121, + "learning_rate": 3.792305822867935e-06, + "loss": 0.7026, + "num_input_tokens_seen": 62364270, + "step": 1907 + }, + { + "epoch": 0.17207016278125986, + "flos": 19983151784640.0, + "grad_norm": 1.9582202215873272, + "learning_rate": 3.792046505641604e-06, + "loss": 0.7454, + "num_input_tokens_seen": 62391465, + "step": 1908 + }, + { + "epoch": 0.17216034630473012, + "flos": 25077624059520.0, + "grad_norm": 2.7900124008832865, + "learning_rate": 3.791787035507984e-06, + "loss": 0.7435, + "num_input_tokens_seen": 62418735, + "step": 1909 + }, + { + "epoch": 0.17225052982820038, + "flos": 25151231651520.0, + "grad_norm": 1.9366488279067133, + "learning_rate": 3.7915274124892136e-06, + "loss": 0.7412, + "num_input_tokens_seen": 62445460, + "step": 1910 + }, + { + "epoch": 0.17234071335167064, + "flos": 28870948324800.0, + "grad_norm": 1.726817944992733, + "learning_rate": 3.7912676366074466e-06, + "loss": 0.7625, + "num_input_tokens_seen": 62477840, + "step": 1911 + }, + { + "epoch": 0.1724308968751409, + "flos": 24427273073280.0, + "grad_norm": 3.2727921243465143, + "learning_rate": 3.7910077078848478e-06, + "loss": 0.748, + "num_input_tokens_seen": 62506695, + "step": 1912 + }, + { + "epoch": 0.17252108039861117, + "flos": 30213174652800.0, + "grad_norm": 4.912386415833135, + "learning_rate": 3.7907476263435957e-06, + "loss": 0.7465, + "num_input_tokens_seen": 62537105, + "step": 1913 + }, + { + "epoch": 0.17261126392208143, + "flos": 39619346596800.0, + "grad_norm": 2.682873225205245, + "learning_rate": 3.7904873920058826e-06, + "loss": 0.751, + "num_input_tokens_seen": 62566330, + "step": 1914 + }, + { + "epoch": 0.1727014474455517, + "flos": 21799590108000.0, + "grad_norm": 2.8472399413841956, + "learning_rate": 3.7902270048939114e-06, + "loss": 0.6608, + "num_input_tokens_seen": 62592530, + "step": 1915 + }, + { + "epoch": 0.17279163096902195, + "flos": 27924977538240.0, + "grad_norm": 2.5551519409210535, + "learning_rate": 3.7899664650299023e-06, + "loss": 0.7456, + "num_input_tokens_seen": 62620975, + "step": 1916 + }, + { + "epoch": 0.1728818144924922, + "flos": 21403111548000.0, + "grad_norm": 1.2892842419958974, + "learning_rate": 3.7897057724360836e-06, + "loss": 0.7911, + "num_input_tokens_seen": 62650770, + "step": 1917 + }, + { + "epoch": 0.1729719980159625, + "flos": 21804831044160.0, + "grad_norm": 2.2774417391547344, + "learning_rate": 3.7894449271347004e-06, + "loss": 0.7959, + "num_input_tokens_seen": 62678350, + "step": 1918 + }, + { + "epoch": 0.17306218153943276, + "flos": 22641654214560.0, + "grad_norm": 1.6602360246228378, + "learning_rate": 3.789183929148009e-06, + "loss": 0.8415, + "num_input_tokens_seen": 62708455, + "step": 1919 + }, + { + "epoch": 0.17315236506290302, + "flos": 23659820142240.0, + "grad_norm": 1.6801454851190034, + "learning_rate": 3.7889227784982795e-06, + "loss": 0.8094, + "num_input_tokens_seen": 62738210, + "step": 1920 + }, + { + "epoch": 0.17324254858637328, + "flos": 23916666149280.0, + "grad_norm": 1.491778412510132, + "learning_rate": 3.7886614752077945e-06, + "loss": 0.7776, + "num_input_tokens_seen": 62768190, + "step": 1921 + }, + { + "epoch": 0.17333273210984354, + "flos": 20784472100640.0, + "grad_norm": 5.91326036140864, + "learning_rate": 3.7884000192988495e-06, + "loss": 0.7781, + "num_input_tokens_seen": 62795620, + "step": 1922 + }, + { + "epoch": 0.1734229156333138, + "flos": 23043150692160.0, + "grad_norm": 1.8827747008775444, + "learning_rate": 3.7881384107937546e-06, + "loss": 0.8311, + "num_input_tokens_seen": 62823835, + "step": 1923 + }, + { + "epoch": 0.17351309915678406, + "flos": 37907446839360.0, + "grad_norm": 2.5964247918096772, + "learning_rate": 3.78787664971483e-06, + "loss": 0.7713, + "num_input_tokens_seen": 62855790, + "step": 1924 + }, + { + "epoch": 0.17360328268025432, + "flos": 27705418534080.0, + "grad_norm": 1.3692027471056212, + "learning_rate": 3.7876147360844115e-06, + "loss": 0.7502, + "num_input_tokens_seen": 62888265, + "step": 1925 + }, + { + "epoch": 0.17369346620372458, + "flos": 18816246712800.0, + "grad_norm": 2.321039875731636, + "learning_rate": 3.7873526699248474e-06, + "loss": 0.8456, + "num_input_tokens_seen": 62915345, + "step": 1926 + }, + { + "epoch": 0.17378364972719484, + "flos": 19800173557920.0, + "grad_norm": 1.650069238192588, + "learning_rate": 3.7870904512584974e-06, + "loss": 0.7896, + "num_input_tokens_seen": 62945825, + "step": 1927 + }, + { + "epoch": 0.1738738332506651, + "flos": 26643609574560.0, + "grad_norm": 1.6303081380337168, + "learning_rate": 3.7868280801077368e-06, + "loss": 0.7808, + "num_input_tokens_seen": 62976500, + "step": 1928 + }, + { + "epoch": 0.17396401677413537, + "flos": 27525339548640.0, + "grad_norm": 1.87683117775025, + "learning_rate": 3.7865655564949517e-06, + "loss": 0.7725, + "num_input_tokens_seen": 63010575, + "step": 1929 + }, + { + "epoch": 0.17405420029760563, + "flos": 24935612641920.0, + "grad_norm": 1.5191601297583517, + "learning_rate": 3.786302880442542e-06, + "loss": 0.7499, + "num_input_tokens_seen": 63040185, + "step": 1930 + }, + { + "epoch": 0.1741443838210759, + "flos": 18777212731200.0, + "grad_norm": 2.221959117042899, + "learning_rate": 3.7860400519729215e-06, + "loss": 0.7779, + "num_input_tokens_seen": 63067030, + "step": 1931 + }, + { + "epoch": 0.17423456734454615, + "flos": 21038567545440.0, + "grad_norm": 2.3059289724324383, + "learning_rate": 3.7857770711085157e-06, + "loss": 0.8011, + "num_input_tokens_seen": 63092535, + "step": 1932 + }, + { + "epoch": 0.1743247508680164, + "flos": 22417783518240.0, + "grad_norm": 2.1025524431326432, + "learning_rate": 3.785513937871763e-06, + "loss": 0.8287, + "num_input_tokens_seen": 63121065, + "step": 1933 + }, + { + "epoch": 0.17441493439148667, + "flos": 24751742340960.0, + "grad_norm": 2.3623228541571004, + "learning_rate": 3.785250652285116e-06, + "loss": 0.7498, + "num_input_tokens_seen": 63149705, + "step": 1934 + }, + { + "epoch": 0.17450511791495693, + "flos": 27231280878240.0, + "grad_norm": 1.9202932035564575, + "learning_rate": 3.78498721437104e-06, + "loss": 0.8598, + "num_input_tokens_seen": 63176065, + "step": 1935 + }, + { + "epoch": 0.1745953014384272, + "flos": 14846040061440.0, + "grad_norm": 2.54056736879568, + "learning_rate": 3.784723624152012e-06, + "loss": 0.7798, + "num_input_tokens_seen": 63199935, + "step": 1936 + }, + { + "epoch": 0.17468548496189745, + "flos": 26357430893280.0, + "grad_norm": 2.574583268007902, + "learning_rate": 3.784459881650524e-06, + "loss": 0.7255, + "num_input_tokens_seen": 63229370, + "step": 1937 + }, + { + "epoch": 0.1747756684853677, + "flos": 23443978114080.0, + "grad_norm": 1.708493023673649, + "learning_rate": 3.784195986889079e-06, + "loss": 0.7501, + "num_input_tokens_seen": 63259265, + "step": 1938 + }, + { + "epoch": 0.17486585200883797, + "flos": 28690125944160.0, + "grad_norm": 1.7146379699283234, + "learning_rate": 3.7839319398901946e-06, + "loss": 0.7058, + "num_input_tokens_seen": 63287150, + "step": 1939 + }, + { + "epoch": 0.17495603553230824, + "flos": 26868409514880.0, + "grad_norm": 1.6987891344695267, + "learning_rate": 3.7836677406764013e-06, + "loss": 0.85, + "num_input_tokens_seen": 63317275, + "step": 1940 + }, + { + "epoch": 0.1750462190557785, + "flos": 25411014069600.0, + "grad_norm": 1.6291985192400569, + "learning_rate": 3.7834033892702407e-06, + "loss": 0.8245, + "num_input_tokens_seen": 63345935, + "step": 1941 + }, + { + "epoch": 0.17513640257924878, + "flos": 25593248901120.0, + "grad_norm": 1.9125582019202505, + "learning_rate": 3.783138885694269e-06, + "loss": 0.7367, + "num_input_tokens_seen": 63375375, + "step": 1942 + }, + { + "epoch": 0.17522658610271905, + "flos": 18015706961760.0, + "grad_norm": 1.98603595096437, + "learning_rate": 3.7828742299710558e-06, + "loss": 0.8561, + "num_input_tokens_seen": 63402590, + "step": 1943 + }, + { + "epoch": 0.1753167696261893, + "flos": 23876479905120.0, + "grad_norm": 1.6833537698518655, + "learning_rate": 3.782609422123183e-06, + "loss": 0.7574, + "num_input_tokens_seen": 63429540, + "step": 1944 + }, + { + "epoch": 0.17540695314965957, + "flos": 18488618015520.0, + "grad_norm": 2.224143980140893, + "learning_rate": 3.7823444621732444e-06, + "loss": 0.7794, + "num_input_tokens_seen": 63457955, + "step": 1945 + }, + { + "epoch": 0.17549713667312983, + "flos": 24318831682560.0, + "grad_norm": 1.8471284076065235, + "learning_rate": 3.782079350143849e-06, + "loss": 0.7944, + "num_input_tokens_seen": 63486700, + "step": 1946 + }, + { + "epoch": 0.1755873201966001, + "flos": 22090563688320.0, + "grad_norm": 2.1145231366789385, + "learning_rate": 3.781814086057617e-06, + "loss": 0.8791, + "num_input_tokens_seen": 63512930, + "step": 1947 + }, + { + "epoch": 0.17567750372007035, + "flos": 17140593204960.0, + "grad_norm": 2.023720032986683, + "learning_rate": 3.7815486699371826e-06, + "loss": 0.7707, + "num_input_tokens_seen": 63540100, + "step": 1948 + }, + { + "epoch": 0.1757676872435406, + "flos": 42133607423520.0, + "grad_norm": 2.032565585834261, + "learning_rate": 3.7812831018051918e-06, + "loss": 0.7526, + "num_input_tokens_seen": 63572245, + "step": 1949 + }, + { + "epoch": 0.17585787076701087, + "flos": 19763369761920.0, + "grad_norm": 1.7674367605763774, + "learning_rate": 3.7810173816843058e-06, + "loss": 0.786, + "num_input_tokens_seen": 63600060, + "step": 1950 + }, + { + "epoch": 0.17594805429048113, + "flos": 23405018472000.0, + "grad_norm": 1.9463473711787507, + "learning_rate": 3.7807515095971955e-06, + "loss": 0.8531, + "num_input_tokens_seen": 63631395, + "step": 1951 + }, + { + "epoch": 0.1760382378139514, + "flos": 21690479661600.0, + "grad_norm": 1.7890880199761123, + "learning_rate": 3.7804854855665475e-06, + "loss": 0.7787, + "num_input_tokens_seen": 63659145, + "step": 1952 + }, + { + "epoch": 0.17612842133742165, + "flos": 20344462018080.0, + "grad_norm": 1.8189466206297977, + "learning_rate": 3.7802193096150606e-06, + "loss": 0.7894, + "num_input_tokens_seen": 63686775, + "step": 1953 + }, + { + "epoch": 0.17621860486089191, + "flos": 21986285310720.0, + "grad_norm": 1.5203980471780385, + "learning_rate": 3.779952981765446e-06, + "loss": 0.7321, + "num_input_tokens_seen": 63715910, + "step": 1954 + }, + { + "epoch": 0.17630878838436218, + "flos": 28799533748640.0, + "grad_norm": 1.5504166104730959, + "learning_rate": 3.779686502040429e-06, + "loss": 0.797, + "num_input_tokens_seen": 63747305, + "step": 1955 + }, + { + "epoch": 0.17639897190783244, + "flos": 69078486601920.0, + "grad_norm": 0.7086865769119636, + "learning_rate": 3.779419870462746e-06, + "loss": 0.5741, + "num_input_tokens_seen": 63828405, + "step": 1956 + }, + { + "epoch": 0.1764891554313027, + "flos": 24354780574080.0, + "grad_norm": 2.2101120772790055, + "learning_rate": 3.779153087055148e-06, + "loss": 0.7598, + "num_input_tokens_seen": 63857380, + "step": 1957 + }, + { + "epoch": 0.17657933895477296, + "flos": 25083385372320.0, + "grad_norm": 2.516240394145673, + "learning_rate": 3.7788861518403988e-06, + "loss": 0.782, + "num_input_tokens_seen": 63884510, + "step": 1958 + }, + { + "epoch": 0.17666952247824322, + "flos": 21257048626560.0, + "grad_norm": 2.429734698754032, + "learning_rate": 3.7786190648412742e-06, + "loss": 0.6778, + "num_input_tokens_seen": 63913325, + "step": 1959 + }, + { + "epoch": 0.17675970600171348, + "flos": 60886505398080.0, + "grad_norm": 0.6353366337595725, + "learning_rate": 3.778351826080564e-06, + "loss": 0.5867, + "num_input_tokens_seen": 64000150, + "step": 1960 + }, + { + "epoch": 0.17684988952518374, + "flos": 21876580148160.0, + "grad_norm": 1.7516301600568682, + "learning_rate": 3.7780844355810704e-06, + "loss": 0.8853, + "num_input_tokens_seen": 64029480, + "step": 1961 + }, + { + "epoch": 0.176940073048654, + "flos": 19800433746240.0, + "grad_norm": 1.7306399293023462, + "learning_rate": 3.777816893365608e-06, + "loss": 0.7938, + "num_input_tokens_seen": 64059845, + "step": 1962 + }, + { + "epoch": 0.17703025657212426, + "flos": 21403929282720.0, + "grad_norm": 2.7115719627188, + "learning_rate": 3.7775491994570057e-06, + "loss": 0.8566, + "num_input_tokens_seen": 64086260, + "step": 1963 + }, + { + "epoch": 0.17712044009559452, + "flos": 20124791504640.0, + "grad_norm": 2.2444448771124863, + "learning_rate": 3.777281353878105e-06, + "loss": 0.7596, + "num_input_tokens_seen": 64114610, + "step": 1964 + }, + { + "epoch": 0.17721062361906478, + "flos": 59601117366720.0, + "grad_norm": 0.629112304671583, + "learning_rate": 3.777013356651758e-06, + "loss": 0.5923, + "num_input_tokens_seen": 64213585, + "step": 1965 + }, + { + "epoch": 0.17730080714253507, + "flos": 17214721173600.0, + "grad_norm": 2.1011336218689762, + "learning_rate": 3.776745207800834e-06, + "loss": 0.7996, + "num_input_tokens_seen": 64239565, + "step": 1966 + }, + { + "epoch": 0.17739099066600533, + "flos": 22789055247360.0, + "grad_norm": 1.9624706217516743, + "learning_rate": 3.7764769073482122e-06, + "loss": 0.8083, + "num_input_tokens_seen": 64269440, + "step": 1967 + }, + { + "epoch": 0.1774811741894756, + "flos": 23219475531840.0, + "grad_norm": 1.3808212167088931, + "learning_rate": 3.7762084553167846e-06, + "loss": 0.8467, + "num_input_tokens_seen": 64299410, + "step": 1968 + }, + { + "epoch": 0.17757135771294585, + "flos": 24382477778880.0, + "grad_norm": 1.7567389028320026, + "learning_rate": 3.775939851729458e-06, + "loss": 0.8746, + "num_input_tokens_seen": 64323710, + "step": 1969 + }, + { + "epoch": 0.17766154123641612, + "flos": 19251276047520.0, + "grad_norm": 2.3969556971447568, + "learning_rate": 3.775671096609151e-06, + "loss": 0.7401, + "num_input_tokens_seen": 64350025, + "step": 1970 + }, + { + "epoch": 0.17775172475988638, + "flos": 39542876933280.0, + "grad_norm": 1.7490608556377458, + "learning_rate": 3.775402189978795e-06, + "loss": 0.7647, + "num_input_tokens_seen": 64381915, + "step": 1971 + }, + { + "epoch": 0.17784190828335664, + "flos": 23442268305120.0, + "grad_norm": 2.452857519254966, + "learning_rate": 3.7751331318613343e-06, + "loss": 0.7645, + "num_input_tokens_seen": 64413000, + "step": 1972 + }, + { + "epoch": 0.1779320918068269, + "flos": 29812495909920.0, + "grad_norm": 1.8365639125214241, + "learning_rate": 3.774863922279727e-06, + "loss": 0.8021, + "num_input_tokens_seen": 64444550, + "step": 1973 + }, + { + "epoch": 0.17802227533029716, + "flos": 22093946136480.0, + "grad_norm": 3.2670953534201974, + "learning_rate": 3.7745945612569435e-06, + "loss": 0.8417, + "num_input_tokens_seen": 64472795, + "step": 1974 + }, + { + "epoch": 0.17811245885376742, + "flos": 27446156492640.0, + "grad_norm": 1.4645538450105897, + "learning_rate": 3.7743250488159674e-06, + "loss": 0.6604, + "num_input_tokens_seen": 64503580, + "step": 1975 + }, + { + "epoch": 0.17820264237723768, + "flos": 17869495361280.0, + "grad_norm": 2.2699270510610408, + "learning_rate": 3.774055384979794e-06, + "loss": 0.6812, + "num_input_tokens_seen": 64529685, + "step": 1976 + }, + { + "epoch": 0.17829282590070794, + "flos": 26427358679040.0, + "grad_norm": 2.2043459059220543, + "learning_rate": 3.773785569771433e-06, + "loss": 0.7621, + "num_input_tokens_seen": 64559705, + "step": 1977 + }, + { + "epoch": 0.1783830094241782, + "flos": 24209015010720.0, + "grad_norm": 1.6297245303049637, + "learning_rate": 3.7735156032139066e-06, + "loss": 0.7663, + "num_input_tokens_seen": 64590010, + "step": 1978 + }, + { + "epoch": 0.17847319294764846, + "flos": 26717700373440.0, + "grad_norm": 1.829237249079963, + "learning_rate": 3.773245485330251e-06, + "loss": 0.8813, + "num_input_tokens_seen": 64620495, + "step": 1979 + }, + { + "epoch": 0.17856337647111872, + "flos": 26430889806240.0, + "grad_norm": 2.1007011378438225, + "learning_rate": 3.7729752161435115e-06, + "loss": 0.7503, + "num_input_tokens_seen": 64649120, + "step": 1980 + }, + { + "epoch": 0.17865355999458898, + "flos": 24828323513760.0, + "grad_norm": 1.6405966620041472, + "learning_rate": 3.7727047956767514e-06, + "loss": 0.8475, + "num_input_tokens_seen": 64677705, + "step": 1981 + }, + { + "epoch": 0.17874374351805924, + "flos": 20532941369280.0, + "grad_norm": 1.6136928522819134, + "learning_rate": 3.7724342239530436e-06, + "loss": 0.9103, + "num_input_tokens_seen": 64708850, + "step": 1982 + }, + { + "epoch": 0.1788339270415295, + "flos": 34665956495520.0, + "grad_norm": 1.9856305092941378, + "learning_rate": 3.772163500995474e-06, + "loss": 0.7306, + "num_input_tokens_seen": 64741055, + "step": 1983 + }, + { + "epoch": 0.17892411056499977, + "flos": 27415039669920.0, + "grad_norm": 2.068212516549216, + "learning_rate": 3.7718926268271437e-06, + "loss": 0.8755, + "num_input_tokens_seen": 64768330, + "step": 1984 + }, + { + "epoch": 0.17901429408847003, + "flos": 22350940822560.0, + "grad_norm": 2.0858002235303474, + "learning_rate": 3.771621601471164e-06, + "loss": 0.8027, + "num_input_tokens_seen": 64795535, + "step": 1985 + }, + { + "epoch": 0.1791044776119403, + "flos": 31386026886240.0, + "grad_norm": 3.5152004409442497, + "learning_rate": 3.771350424950661e-06, + "loss": 0.8304, + "num_input_tokens_seen": 64826275, + "step": 1986 + }, + { + "epoch": 0.17919466113541055, + "flos": 26612975958720.0, + "grad_norm": 3.6821963429646294, + "learning_rate": 3.771079097288772e-06, + "loss": 0.7269, + "num_input_tokens_seen": 64853120, + "step": 1987 + }, + { + "epoch": 0.1792848446588808, + "flos": 22933482699360.0, + "grad_norm": 1.466210869204366, + "learning_rate": 3.770807618508649e-06, + "loss": 0.7985, + "num_input_tokens_seen": 64884505, + "step": 1988 + }, + { + "epoch": 0.17937502818235107, + "flos": 19467378264000.0, + "grad_norm": 1.6606390755846137, + "learning_rate": 3.7705359886334555e-06, + "loss": 0.7706, + "num_input_tokens_seen": 64911680, + "step": 1989 + }, + { + "epoch": 0.17946521170582136, + "flos": 23552679693120.0, + "grad_norm": 2.0101718918287776, + "learning_rate": 3.7702642076863694e-06, + "loss": 0.7609, + "num_input_tokens_seen": 64938505, + "step": 1990 + }, + { + "epoch": 0.17955539522929162, + "flos": 21184704806400.0, + "grad_norm": 1.6468047241041255, + "learning_rate": 3.7699922756905795e-06, + "loss": 0.8401, + "num_input_tokens_seen": 64966570, + "step": 1991 + }, + { + "epoch": 0.17964557875276188, + "flos": 27596828464320.0, + "grad_norm": 1.7262530157351548, + "learning_rate": 3.7697201926692895e-06, + "loss": 0.7333, + "num_input_tokens_seen": 64995445, + "step": 1992 + }, + { + "epoch": 0.17973576227623214, + "flos": 25847864722560.0, + "grad_norm": 2.081111167882369, + "learning_rate": 3.7694479586457144e-06, + "loss": 0.7953, + "num_input_tokens_seen": 65020950, + "step": 1993 + }, + { + "epoch": 0.1798259457997024, + "flos": 18269765236800.0, + "grad_norm": 2.723688835588219, + "learning_rate": 3.7691755736430827e-06, + "loss": 0.7983, + "num_input_tokens_seen": 65046690, + "step": 1994 + }, + { + "epoch": 0.17991612932317266, + "flos": 31131039367200.0, + "grad_norm": 2.026973036313361, + "learning_rate": 3.768903037684636e-06, + "loss": 0.8255, + "num_input_tokens_seen": 65078190, + "step": 1995 + }, + { + "epoch": 0.18000631284664292, + "flos": 25593620598720.0, + "grad_norm": 1.8319064425187892, + "learning_rate": 3.7686303507936284e-06, + "loss": 0.8335, + "num_input_tokens_seen": 65106295, + "step": 1996 + }, + { + "epoch": 0.18009649637011318, + "flos": 14481087191520.0, + "grad_norm": 2.2858638352618805, + "learning_rate": 3.7683575129933272e-06, + "loss": 0.8262, + "num_input_tokens_seen": 65131360, + "step": 1997 + }, + { + "epoch": 0.18018667989358345, + "flos": 23079285432480.0, + "grad_norm": 1.7351941073228352, + "learning_rate": 3.7680845243070128e-06, + "loss": 0.7909, + "num_input_tokens_seen": 65162910, + "step": 1998 + }, + { + "epoch": 0.1802768634170537, + "flos": 61678304521920.0, + "grad_norm": 0.7156958998827706, + "learning_rate": 3.7678113847579767e-06, + "loss": 0.5539, + "num_input_tokens_seen": 65246220, + "step": 1999 + }, + { + "epoch": 0.18036704694052397, + "flos": 23735434901280.0, + "grad_norm": 2.349162857613707, + "learning_rate": 3.7675380943695264e-06, + "loss": 0.854, + "num_input_tokens_seen": 65273650, + "step": 2000 + }, + { + "epoch": 0.18045723046399423, + "flos": 23735546410560.0, + "grad_norm": 1.9490833426555756, + "learning_rate": 3.7672646531649795e-06, + "loss": 0.7485, + "num_input_tokens_seen": 65302860, + "step": 2001 + }, + { + "epoch": 0.1805474139874645, + "flos": 64140707799840.0, + "grad_norm": 0.5768015378918298, + "learning_rate": 3.7669910611676682e-06, + "loss": 0.576, + "num_input_tokens_seen": 65400480, + "step": 2002 + }, + { + "epoch": 0.18063759751093475, + "flos": 30179344437600.0, + "grad_norm": 1.8260907976236482, + "learning_rate": 3.7667173184009356e-06, + "loss": 0.7702, + "num_input_tokens_seen": 65432500, + "step": 2003 + }, + { + "epoch": 0.180727781034405, + "flos": 22021490807040.0, + "grad_norm": 2.045087310307559, + "learning_rate": 3.7664434248881403e-06, + "loss": 0.8484, + "num_input_tokens_seen": 65462375, + "step": 2004 + }, + { + "epoch": 0.18081796455787527, + "flos": 19327150994880.0, + "grad_norm": 2.2886059406638455, + "learning_rate": 3.766169380652652e-06, + "loss": 0.7934, + "num_input_tokens_seen": 65490375, + "step": 2005 + }, + { + "epoch": 0.18090814808134553, + "flos": 19836828674880.0, + "grad_norm": 1.6870354613470362, + "learning_rate": 3.7658951857178537e-06, + "loss": 0.8281, + "num_input_tokens_seen": 65518395, + "step": 2006 + }, + { + "epoch": 0.1809983316048158, + "flos": 22203205261920.0, + "grad_norm": 1.6153888374839354, + "learning_rate": 3.7656208401071414e-06, + "loss": 0.7938, + "num_input_tokens_seen": 65546695, + "step": 2007 + }, + { + "epoch": 0.18108851512828605, + "flos": 23407099978560.0, + "grad_norm": 5.146680795587572, + "learning_rate": 3.7653463438439225e-06, + "loss": 0.7244, + "num_input_tokens_seen": 65574690, + "step": 2008 + }, + { + "epoch": 0.18117869865175631, + "flos": 20090775440640.0, + "grad_norm": 2.105440893325112, + "learning_rate": 3.7650716969516203e-06, + "loss": 0.7972, + "num_input_tokens_seen": 65600755, + "step": 2009 + }, + { + "epoch": 0.18126888217522658, + "flos": 29965472406720.0, + "grad_norm": 1.9203989206582397, + "learning_rate": 3.764796899453668e-06, + "loss": 0.6863, + "num_input_tokens_seen": 65631750, + "step": 2010 + }, + { + "epoch": 0.18135906569869684, + "flos": 31203122999040.0, + "grad_norm": 1.9053169218152506, + "learning_rate": 3.7645219513735134e-06, + "loss": 0.8253, + "num_input_tokens_seen": 65661435, + "step": 2011 + }, + { + "epoch": 0.1814492492221671, + "flos": 32625275778240.0, + "grad_norm": 1.6551847811112186, + "learning_rate": 3.764246852734617e-06, + "loss": 0.8276, + "num_input_tokens_seen": 65693150, + "step": 2012 + }, + { + "epoch": 0.18153943274563739, + "flos": 18816841428960.0, + "grad_norm": 2.1424005203565515, + "learning_rate": 3.7639716035604502e-06, + "loss": 0.76, + "num_input_tokens_seen": 65720205, + "step": 2013 + }, + { + "epoch": 0.18162961626910765, + "flos": 71752563161280.0, + "grad_norm": 0.6405210410644208, + "learning_rate": 3.7636962038745e-06, + "loss": 0.548, + "num_input_tokens_seen": 65817730, + "step": 2014 + }, + { + "epoch": 0.1817197997925779, + "flos": 28435287104160.0, + "grad_norm": 1.9120070027737885, + "learning_rate": 3.763420653700265e-06, + "loss": 0.7483, + "num_input_tokens_seen": 65846500, + "step": 2015 + }, + { + "epoch": 0.18180998331604817, + "flos": 25734479753760.0, + "grad_norm": 1.8443930199679035, + "learning_rate": 3.7631449530612565e-06, + "loss": 0.7431, + "num_input_tokens_seen": 65874775, + "step": 2016 + }, + { + "epoch": 0.18190016683951843, + "flos": 25151529009600.0, + "grad_norm": 1.5593042116782392, + "learning_rate": 3.762869101980999e-06, + "loss": 0.8748, + "num_input_tokens_seen": 65901380, + "step": 2017 + }, + { + "epoch": 0.1819903503629887, + "flos": 22496186009280.0, + "grad_norm": 1.9391835804283475, + "learning_rate": 3.7625931004830287e-06, + "loss": 0.7829, + "num_input_tokens_seen": 65931360, + "step": 2018 + }, + { + "epoch": 0.18208053388645895, + "flos": 25993444437120.0, + "grad_norm": 1.8686261241260536, + "learning_rate": 3.7623169485908966e-06, + "loss": 0.7416, + "num_input_tokens_seen": 65962435, + "step": 2019 + }, + { + "epoch": 0.1821707174099292, + "flos": 67376734188960.0, + "grad_norm": 0.7182289390941662, + "learning_rate": 3.7620406463281647e-06, + "loss": 0.5948, + "num_input_tokens_seen": 66039400, + "step": 2020 + }, + { + "epoch": 0.18226090093339947, + "flos": 28324950055680.0, + "grad_norm": 2.7171327889250034, + "learning_rate": 3.7617641937184095e-06, + "loss": 0.8084, + "num_input_tokens_seen": 66069585, + "step": 2021 + }, + { + "epoch": 0.18235108445686973, + "flos": 24026780179200.0, + "grad_norm": 5.21052167568781, + "learning_rate": 3.761487590785219e-06, + "loss": 0.8137, + "num_input_tokens_seen": 66096660, + "step": 2022 + }, + { + "epoch": 0.18244126798034, + "flos": 22489272433920.0, + "grad_norm": 2.1123593165034786, + "learning_rate": 3.7612108375521942e-06, + "loss": 0.7573, + "num_input_tokens_seen": 66125090, + "step": 2023 + }, + { + "epoch": 0.18253145150381025, + "flos": 23479964175360.0, + "grad_norm": 1.5788503811562609, + "learning_rate": 3.76093393404295e-06, + "loss": 0.8337, + "num_input_tokens_seen": 66153680, + "step": 2024 + }, + { + "epoch": 0.18262163502728052, + "flos": 22533026975040.0, + "grad_norm": 2.1687688777222913, + "learning_rate": 3.7606568802811126e-06, + "loss": 0.7777, + "num_input_tokens_seen": 66182060, + "step": 2025 + }, + { + "epoch": 0.18271181855075078, + "flos": 17206432317120.0, + "grad_norm": 5.931093313159413, + "learning_rate": 3.760379676290322e-06, + "loss": 0.7981, + "num_input_tokens_seen": 66204435, + "step": 2026 + }, + { + "epoch": 0.18280200207422104, + "flos": 20524169305920.0, + "grad_norm": 5.138355809419423, + "learning_rate": 3.760102322094231e-06, + "loss": 0.7158, + "num_input_tokens_seen": 66232095, + "step": 2027 + }, + { + "epoch": 0.1828921855976913, + "flos": 25739088804000.0, + "grad_norm": 1.8692139677927075, + "learning_rate": 3.759824817716504e-06, + "loss": 0.7483, + "num_input_tokens_seen": 66263480, + "step": 2028 + }, + { + "epoch": 0.18298236912116156, + "flos": 26430406599360.0, + "grad_norm": 1.2606168960968973, + "learning_rate": 3.759547163180821e-06, + "loss": 0.8375, + "num_input_tokens_seen": 66293645, + "step": 2029 + }, + { + "epoch": 0.18307255264463182, + "flos": 27925423575360.0, + "grad_norm": 1.7289113082156717, + "learning_rate": 3.759269358510871e-06, + "loss": 0.5874, + "num_input_tokens_seen": 66323555, + "step": 2030 + }, + { + "epoch": 0.18316273616810208, + "flos": 27670213037760.0, + "grad_norm": 1.8497667577989547, + "learning_rate": 3.75899140373036e-06, + "loss": 0.6913, + "num_input_tokens_seen": 66351760, + "step": 2031 + }, + { + "epoch": 0.18325291969157234, + "flos": 18087716254080.0, + "grad_norm": 2.7227298940060343, + "learning_rate": 3.7587132988630028e-06, + "loss": 0.8439, + "num_input_tokens_seen": 66379650, + "step": 2032 + }, + { + "epoch": 0.1833431032150426, + "flos": 14773175864640.0, + "grad_norm": 1.9724276268546133, + "learning_rate": 3.7584350439325295e-06, + "loss": 0.8192, + "num_input_tokens_seen": 66404565, + "step": 2033 + }, + { + "epoch": 0.18343328673851286, + "flos": 17322865206240.0, + "grad_norm": 1.92662837526138, + "learning_rate": 3.758156638962682e-06, + "loss": 0.7741, + "num_input_tokens_seen": 66432025, + "step": 2034 + }, + { + "epoch": 0.18352347026198312, + "flos": 27156037816800.0, + "grad_norm": 1.7891838332431897, + "learning_rate": 3.757878083977216e-06, + "loss": 0.7464, + "num_input_tokens_seen": 66462635, + "step": 2035 + }, + { + "epoch": 0.18361365378545338, + "flos": 26140510942080.0, + "grad_norm": 1.670654207681433, + "learning_rate": 3.7575993789999e-06, + "loss": 0.7742, + "num_input_tokens_seen": 66493385, + "step": 2036 + }, + { + "epoch": 0.18370383730892367, + "flos": 22824409422720.0, + "grad_norm": 1.852559654109333, + "learning_rate": 3.757320524054512e-06, + "loss": 0.8211, + "num_input_tokens_seen": 66520660, + "step": 2037 + }, + { + "epoch": 0.18379402083239393, + "flos": 22168445802720.0, + "grad_norm": 1.84867904774541, + "learning_rate": 3.757041519164848e-06, + "loss": 0.7166, + "num_input_tokens_seen": 66551500, + "step": 2038 + }, + { + "epoch": 0.1838842043558642, + "flos": 25077809908320.0, + "grad_norm": 2.385222670039203, + "learning_rate": 3.7567623643547133e-06, + "loss": 0.7929, + "num_input_tokens_seen": 66578560, + "step": 2039 + }, + { + "epoch": 0.18397438787933446, + "flos": 25338558740160.0, + "grad_norm": 1.749066103654986, + "learning_rate": 3.756483059647927e-06, + "loss": 0.784, + "num_input_tokens_seen": 66605945, + "step": 2040 + }, + { + "epoch": 0.18406457140280472, + "flos": 34914364967040.0, + "grad_norm": 1.54842999838543, + "learning_rate": 3.756203605068321e-06, + "loss": 0.7588, + "num_input_tokens_seen": 66636750, + "step": 2041 + }, + { + "epoch": 0.18415475492627498, + "flos": 25302386830080.0, + "grad_norm": 1.793821443736605, + "learning_rate": 3.7559240006397396e-06, + "loss": 0.774, + "num_input_tokens_seen": 66667395, + "step": 2042 + }, + { + "epoch": 0.18424493844974524, + "flos": 23691903378720.0, + "grad_norm": 1.8515832312344809, + "learning_rate": 3.7556442463860406e-06, + "loss": 0.9032, + "num_input_tokens_seen": 66696920, + "step": 2043 + }, + { + "epoch": 0.1843351219732155, + "flos": 23699262991200.0, + "grad_norm": 2.228404250441883, + "learning_rate": 3.7553643423310934e-06, + "loss": 0.8948, + "num_input_tokens_seen": 66724955, + "step": 2044 + }, + { + "epoch": 0.18442530549668576, + "flos": 26939155035360.0, + "grad_norm": 2.482332937836909, + "learning_rate": 3.755084288498782e-06, + "loss": 0.6392, + "num_input_tokens_seen": 66755910, + "step": 2045 + }, + { + "epoch": 0.18451548902015602, + "flos": 26358062779200.0, + "grad_norm": 1.798163441342245, + "learning_rate": 3.754804084913002e-06, + "loss": 0.7902, + "num_input_tokens_seen": 66784685, + "step": 2046 + }, + { + "epoch": 0.18460567254362628, + "flos": 21729439303680.0, + "grad_norm": 1.6807385372172141, + "learning_rate": 3.754523731597661e-06, + "loss": 0.8509, + "num_input_tokens_seen": 66813810, + "step": 2047 + }, + { + "epoch": 0.18469585606709654, + "flos": 20601754062240.0, + "grad_norm": 2.2522775990288495, + "learning_rate": 3.754243228576681e-06, + "loss": 0.6812, + "num_input_tokens_seen": 66842535, + "step": 2048 + }, + { + "epoch": 0.1847860395905668, + "flos": 27415225518720.0, + "grad_norm": 1.933837998501804, + "learning_rate": 3.753962575873996e-06, + "loss": 0.8384, + "num_input_tokens_seen": 66869350, + "step": 2049 + }, + { + "epoch": 0.18487622311403706, + "flos": 22679312915040.0, + "grad_norm": 1.716675425325043, + "learning_rate": 3.7536817735135527e-06, + "loss": 0.7463, + "num_input_tokens_seen": 66898125, + "step": 2050 + }, + { + "epoch": 0.18496640663750732, + "flos": 27955499644800.0, + "grad_norm": 1.5009343892751652, + "learning_rate": 3.753400821519311e-06, + "loss": 0.8273, + "num_input_tokens_seen": 66927860, + "step": 2051 + }, + { + "epoch": 0.18505659016097759, + "flos": 43481000348160.0, + "grad_norm": 2.881307069291864, + "learning_rate": 3.7531197199152426e-06, + "loss": 0.7161, + "num_input_tokens_seen": 66964005, + "step": 2052 + }, + { + "epoch": 0.18514677368444785, + "flos": 64781171629920.0, + "grad_norm": 0.7119367514763669, + "learning_rate": 3.7528384687253335e-06, + "loss": 0.6174, + "num_input_tokens_seen": 67053575, + "step": 2053 + }, + { + "epoch": 0.1852369572079181, + "flos": 27560024668320.0, + "grad_norm": 2.1386072157748504, + "learning_rate": 3.7525570679735815e-06, + "loss": 0.8915, + "num_input_tokens_seen": 67082830, + "step": 2054 + }, + { + "epoch": 0.18532714073138837, + "flos": 21689141550240.0, + "grad_norm": 1.7584012550264485, + "learning_rate": 3.7522755176839965e-06, + "loss": 0.8005, + "num_input_tokens_seen": 67109515, + "step": 2055 + }, + { + "epoch": 0.18541732425485863, + "flos": 21140801586240.0, + "grad_norm": 2.3835370072344078, + "learning_rate": 3.7519938178806027e-06, + "loss": 0.7967, + "num_input_tokens_seen": 67136990, + "step": 2056 + }, + { + "epoch": 0.1855075077783289, + "flos": 23807667212160.0, + "grad_norm": 2.245218282168244, + "learning_rate": 3.7517119685874358e-06, + "loss": 0.7364, + "num_input_tokens_seen": 67165250, + "step": 2057 + }, + { + "epoch": 0.18559769130179915, + "flos": 23079508451040.0, + "grad_norm": 1.7584150849753097, + "learning_rate": 3.7514299698285447e-06, + "loss": 0.7764, + "num_input_tokens_seen": 67195875, + "step": 2058 + }, + { + "epoch": 0.1856878748252694, + "flos": 18925431498720.0, + "grad_norm": 1.3917989597524425, + "learning_rate": 3.751147821627991e-06, + "loss": 0.7652, + "num_input_tokens_seen": 67224965, + "step": 2059 + }, + { + "epoch": 0.18577805834873967, + "flos": 20347175410560.0, + "grad_norm": 2.586823464373582, + "learning_rate": 3.75086552400985e-06, + "loss": 0.8046, + "num_input_tokens_seen": 67252155, + "step": 2060 + }, + { + "epoch": 0.18586824187220996, + "flos": 21840556917120.0, + "grad_norm": 2.0448937764828736, + "learning_rate": 3.750583076998208e-06, + "loss": 0.7633, + "num_input_tokens_seen": 67280115, + "step": 2061 + }, + { + "epoch": 0.18595842539568022, + "flos": 24500211609600.0, + "grad_norm": 1.8682777701199742, + "learning_rate": 3.7503004806171655e-06, + "loss": 0.764, + "num_input_tokens_seen": 67309320, + "step": 2062 + }, + { + "epoch": 0.18604860891915048, + "flos": 24095853060480.0, + "grad_norm": 2.0508284499776654, + "learning_rate": 3.7500177348908354e-06, + "loss": 0.717, + "num_input_tokens_seen": 67338045, + "step": 2063 + }, + { + "epoch": 0.18613879244262074, + "flos": 23043894087360.0, + "grad_norm": 1.7215521424944968, + "learning_rate": 3.749734839843342e-06, + "loss": 0.8073, + "num_input_tokens_seen": 67365425, + "step": 2064 + }, + { + "epoch": 0.186228975966091, + "flos": 23148246804480.0, + "grad_norm": 2.542385333240857, + "learning_rate": 3.7494517954988245e-06, + "loss": 0.785, + "num_input_tokens_seen": 67394410, + "step": 2065 + }, + { + "epoch": 0.18631915948956126, + "flos": 26832126095520.0, + "grad_norm": 1.8753739302022612, + "learning_rate": 3.749168601881433e-06, + "loss": 0.774, + "num_input_tokens_seen": 67423550, + "step": 2066 + }, + { + "epoch": 0.18640934301303153, + "flos": 20347547108160.0, + "grad_norm": 2.085243938455986, + "learning_rate": 3.7488852590153315e-06, + "loss": 0.7678, + "num_input_tokens_seen": 67449005, + "step": 2067 + }, + { + "epoch": 0.1864995265365018, + "flos": 20529373072320.0, + "grad_norm": 1.739578346047324, + "learning_rate": 3.748601766924697e-06, + "loss": 0.7839, + "num_input_tokens_seen": 67476265, + "step": 2068 + }, + { + "epoch": 0.18658971005997205, + "flos": 20092113552000.0, + "grad_norm": 1.9932245285529113, + "learning_rate": 3.7483181256337176e-06, + "loss": 0.7542, + "num_input_tokens_seen": 67504095, + "step": 2069 + }, + { + "epoch": 0.1866798935834423, + "flos": 23402862625920.0, + "grad_norm": 1.54087652973524, + "learning_rate": 3.7480343351665962e-06, + "loss": 0.7897, + "num_input_tokens_seen": 67532990, + "step": 2070 + }, + { + "epoch": 0.18677007710691257, + "flos": 18633863202240.0, + "grad_norm": 1.7004498111056119, + "learning_rate": 3.747750395547546e-06, + "loss": 0.7701, + "num_input_tokens_seen": 67562055, + "step": 2071 + }, + { + "epoch": 0.18686026063038283, + "flos": 22095024059520.0, + "grad_norm": 1.53791526210779, + "learning_rate": 3.7474663068007956e-06, + "loss": 0.7701, + "num_input_tokens_seen": 67593605, + "step": 2072 + }, + { + "epoch": 0.1869504441538531, + "flos": 35721372256320.0, + "grad_norm": 1.2817526040866412, + "learning_rate": 3.747182068950584e-06, + "loss": 0.7706, + "num_input_tokens_seen": 67626235, + "step": 2073 + }, + { + "epoch": 0.18704062767732335, + "flos": 25409378600160.0, + "grad_norm": 1.8671520022286, + "learning_rate": 3.7468976820211643e-06, + "loss": 0.7834, + "num_input_tokens_seen": 67655335, + "step": 2074 + }, + { + "epoch": 0.1871308112007936, + "flos": 20383124302080.0, + "grad_norm": 1.8310289060684974, + "learning_rate": 3.746613146036803e-06, + "loss": 0.7831, + "num_input_tokens_seen": 67683135, + "step": 2075 + }, + { + "epoch": 0.18722099472426387, + "flos": 20019398034240.0, + "grad_norm": 1.7102949773670428, + "learning_rate": 3.7463284610217766e-06, + "loss": 0.8216, + "num_input_tokens_seen": 67710625, + "step": 2076 + }, + { + "epoch": 0.18731117824773413, + "flos": 19363174225920.0, + "grad_norm": 2.0484495440492014, + "learning_rate": 3.746043627000377e-06, + "loss": 0.7175, + "num_input_tokens_seen": 67739275, + "step": 2077 + }, + { + "epoch": 0.1874013617712044, + "flos": 52585010614080.0, + "grad_norm": 1.899567507157526, + "learning_rate": 3.7457586439969076e-06, + "loss": 0.73, + "num_input_tokens_seen": 67771950, + "step": 2078 + }, + { + "epoch": 0.18749154529467466, + "flos": 26429997732000.0, + "grad_norm": 1.5353579789034766, + "learning_rate": 3.7454735120356842e-06, + "loss": 0.7703, + "num_input_tokens_seen": 67802930, + "step": 2079 + }, + { + "epoch": 0.18758172881814492, + "flos": 22896864752160.0, + "grad_norm": 1.518254152837206, + "learning_rate": 3.7451882311410373e-06, + "loss": 0.8487, + "num_input_tokens_seen": 67831940, + "step": 2080 + }, + { + "epoch": 0.18767191234161518, + "flos": 20310631802880.0, + "grad_norm": 1.5662629402034047, + "learning_rate": 3.7449028013373074e-06, + "loss": 0.8206, + "num_input_tokens_seen": 67860115, + "step": 2081 + }, + { + "epoch": 0.18776209586508544, + "flos": 26027980877760.0, + "grad_norm": 1.8229193623840745, + "learning_rate": 3.7446172226488485e-06, + "loss": 0.783, + "num_input_tokens_seen": 67889555, + "step": 2082 + }, + { + "epoch": 0.1878522793885557, + "flos": 26576581030080.0, + "grad_norm": 1.999498125526139, + "learning_rate": 3.7443314951000285e-06, + "loss": 0.8438, + "num_input_tokens_seen": 67916025, + "step": 2083 + }, + { + "epoch": 0.18794246291202596, + "flos": 24244963902240.0, + "grad_norm": 2.653613049445326, + "learning_rate": 3.7440456187152276e-06, + "loss": 0.7675, + "num_input_tokens_seen": 67945160, + "step": 2084 + }, + { + "epoch": 0.18803264643549625, + "flos": 20638074651360.0, + "grad_norm": 2.1095576454490774, + "learning_rate": 3.7437595935188377e-06, + "loss": 0.8638, + "num_input_tokens_seen": 67973500, + "step": 2085 + }, + { + "epoch": 0.1881228299589665, + "flos": 21840742765920.0, + "grad_norm": 2.521354169608525, + "learning_rate": 3.7434734195352647e-06, + "loss": 0.7807, + "num_input_tokens_seen": 68002565, + "step": 2086 + }, + { + "epoch": 0.18821301348243677, + "flos": 57444078099840.0, + "grad_norm": 0.6536846200040857, + "learning_rate": 3.743187096788926e-06, + "loss": 0.58, + "num_input_tokens_seen": 68086760, + "step": 2087 + }, + { + "epoch": 0.18830319700590703, + "flos": 14190262290240.0, + "grad_norm": 1.8028130643291254, + "learning_rate": 3.7429006253042524e-06, + "loss": 0.7812, + "num_input_tokens_seen": 68113305, + "step": 2088 + }, + { + "epoch": 0.1883933805293773, + "flos": 22350680634240.0, + "grad_norm": 1.7361061460635925, + "learning_rate": 3.7426140051056867e-06, + "loss": 0.7293, + "num_input_tokens_seen": 68140515, + "step": 2089 + }, + { + "epoch": 0.18848356405284755, + "flos": 17687632227360.0, + "grad_norm": 2.1652385944928896, + "learning_rate": 3.7423272362176856e-06, + "loss": 0.7236, + "num_input_tokens_seen": 68168615, + "step": 2090 + }, + { + "epoch": 0.1885737475763178, + "flos": 21984352483200.0, + "grad_norm": 1.8062876423540013, + "learning_rate": 3.742040318664718e-06, + "loss": 0.8269, + "num_input_tokens_seen": 68198435, + "step": 2091 + }, + { + "epoch": 0.18866393109978807, + "flos": 21911859984000.0, + "grad_norm": 1.813208605927183, + "learning_rate": 3.7417532524712643e-06, + "loss": 0.84, + "num_input_tokens_seen": 68226070, + "step": 2092 + }, + { + "epoch": 0.18875411462325833, + "flos": 22532580937920.0, + "grad_norm": 1.841406674682413, + "learning_rate": 3.7414660376618195e-06, + "loss": 0.8109, + "num_input_tokens_seen": 68253255, + "step": 2093 + }, + { + "epoch": 0.1888442981467286, + "flos": 70312302975360.0, + "grad_norm": 0.6113262237421763, + "learning_rate": 3.74117867426089e-06, + "loss": 0.5901, + "num_input_tokens_seen": 68355435, + "step": 2094 + }, + { + "epoch": 0.18893448167019886, + "flos": 25447966544640.0, + "grad_norm": 1.8333670945382354, + "learning_rate": 3.7408911622929954e-06, + "loss": 0.7589, + "num_input_tokens_seen": 68385225, + "step": 2095 + }, + { + "epoch": 0.18902466519366912, + "flos": 25483952605920.0, + "grad_norm": 2.6885206369264574, + "learning_rate": 3.740603501782668e-06, + "loss": 0.8159, + "num_input_tokens_seen": 68415700, + "step": 2096 + }, + { + "epoch": 0.18911484871713938, + "flos": 24609545074560.0, + "grad_norm": 2.124241637499101, + "learning_rate": 3.7403156927544516e-06, + "loss": 0.7219, + "num_input_tokens_seen": 68446280, + "step": 2097 + }, + { + "epoch": 0.18920503224060964, + "flos": 25044648748800.0, + "grad_norm": 2.9907972328074255, + "learning_rate": 3.740027735232904e-06, + "loss": 0.7168, + "num_input_tokens_seen": 68477125, + "step": 2098 + }, + { + "epoch": 0.1892952157640799, + "flos": 18707879661600.0, + "grad_norm": 1.9303194316327643, + "learning_rate": 3.7397396292425966e-06, + "loss": 0.7666, + "num_input_tokens_seen": 68506295, + "step": 2099 + }, + { + "epoch": 0.18938539928755016, + "flos": 21543636175200.0, + "grad_norm": 1.883136642222558, + "learning_rate": 3.7394513748081105e-06, + "loss": 0.7571, + "num_input_tokens_seen": 68533755, + "step": 2100 + }, + { + "epoch": 0.18947558281102042, + "flos": 22417746348480.0, + "grad_norm": 1.726097009662124, + "learning_rate": 3.7391629719540418e-06, + "loss": 0.8138, + "num_input_tokens_seen": 68561705, + "step": 2101 + }, + { + "epoch": 0.18956576633449068, + "flos": 21621927156960.0, + "grad_norm": 6.755592290181677, + "learning_rate": 3.7388744207049998e-06, + "loss": 0.7721, + "num_input_tokens_seen": 68590100, + "step": 2102 + }, + { + "epoch": 0.18965594985796094, + "flos": 17250521386080.0, + "grad_norm": 3.130798301172484, + "learning_rate": 3.7385857210856027e-06, + "loss": 0.7533, + "num_input_tokens_seen": 68615490, + "step": 2103 + }, + { + "epoch": 0.1897461333814312, + "flos": 28581015497760.0, + "grad_norm": 2.018048035071685, + "learning_rate": 3.738296873120486e-06, + "loss": 0.7659, + "num_input_tokens_seen": 68644280, + "step": 2104 + }, + { + "epoch": 0.18983631690490146, + "flos": 26721937726080.0, + "grad_norm": 2.0843648518511158, + "learning_rate": 3.7380078768342955e-06, + "loss": 0.7495, + "num_input_tokens_seen": 68672495, + "step": 2105 + }, + { + "epoch": 0.18992650042837173, + "flos": 25802697730560.0, + "grad_norm": 1.8152879779700648, + "learning_rate": 3.7377187322516895e-06, + "loss": 0.7982, + "num_input_tokens_seen": 68699315, + "step": 2106 + }, + { + "epoch": 0.19001668395184199, + "flos": 16267300766400.0, + "grad_norm": 1.8193728923883683, + "learning_rate": 3.7374294393973395e-06, + "loss": 0.731, + "num_input_tokens_seen": 68727325, + "step": 2107 + }, + { + "epoch": 0.19010686747531225, + "flos": 28761131652960.0, + "grad_norm": 1.6904660754857412, + "learning_rate": 3.7371399982959294e-06, + "loss": 0.8344, + "num_input_tokens_seen": 68756015, + "step": 2108 + }, + { + "epoch": 0.19019705099878254, + "flos": 41213363844480.0, + "grad_norm": 1.9429535561869804, + "learning_rate": 3.7368504089721565e-06, + "loss": 0.6975, + "num_input_tokens_seen": 68785010, + "step": 2109 + }, + { + "epoch": 0.1902872345222528, + "flos": 28690311792960.0, + "grad_norm": 1.6736705087099624, + "learning_rate": 3.73656067145073e-06, + "loss": 0.7337, + "num_input_tokens_seen": 68815970, + "step": 2110 + }, + { + "epoch": 0.19037741804572306, + "flos": 22861176048960.0, + "grad_norm": 1.9465999402842507, + "learning_rate": 3.736270785756371e-06, + "loss": 0.8288, + "num_input_tokens_seen": 68846650, + "step": 2111 + }, + { + "epoch": 0.19046760156919332, + "flos": 23216390441760.0, + "grad_norm": 3.083249391667541, + "learning_rate": 3.7359807519138156e-06, + "loss": 0.7852, + "num_input_tokens_seen": 68871895, + "step": 2112 + }, + { + "epoch": 0.19055778509266358, + "flos": 25265583034080.0, + "grad_norm": 1.6714804353108148, + "learning_rate": 3.73569056994781e-06, + "loss": 0.7792, + "num_input_tokens_seen": 68900420, + "step": 2113 + }, + { + "epoch": 0.19064796861613384, + "flos": 20492606446080.0, + "grad_norm": 2.5401455960634305, + "learning_rate": 3.7354002398831144e-06, + "loss": 0.82, + "num_input_tokens_seen": 68926745, + "step": 2114 + }, + { + "epoch": 0.1907381521396041, + "flos": 20274088195200.0, + "grad_norm": 4.02238803250119, + "learning_rate": 3.7351097617445015e-06, + "loss": 0.7776, + "num_input_tokens_seen": 68955430, + "step": 2115 + }, + { + "epoch": 0.19082833566307436, + "flos": 63737464343520.0, + "grad_norm": 0.690237641772067, + "learning_rate": 3.7348191355567567e-06, + "loss": 0.6111, + "num_input_tokens_seen": 69036905, + "step": 2116 + }, + { + "epoch": 0.19091851918654462, + "flos": 22751842584000.0, + "grad_norm": 2.112982483790777, + "learning_rate": 3.734528361344677e-06, + "loss": 0.8088, + "num_input_tokens_seen": 69065455, + "step": 2117 + }, + { + "epoch": 0.19100870271001488, + "flos": 20164531711680.0, + "grad_norm": 1.9229363092953244, + "learning_rate": 3.734237439133074e-06, + "loss": 0.7771, + "num_input_tokens_seen": 69091640, + "step": 2118 + }, + { + "epoch": 0.19109888623348514, + "flos": 25006432501920.0, + "grad_norm": 1.739994876020933, + "learning_rate": 3.7339463689467702e-06, + "loss": 0.8268, + "num_input_tokens_seen": 69118820, + "step": 2119 + }, + { + "epoch": 0.1911890697569554, + "flos": 15898891108800.0, + "grad_norm": 1.6513064760748186, + "learning_rate": 3.733655150810601e-06, + "loss": 0.7508, + "num_input_tokens_seen": 69145075, + "step": 2120 + }, + { + "epoch": 0.19127925328042567, + "flos": 24020461320000.0, + "grad_norm": 2.0081262009465997, + "learning_rate": 3.7333637847494154e-06, + "loss": 0.9123, + "num_input_tokens_seen": 69173290, + "step": 2121 + }, + { + "epoch": 0.19136943680389593, + "flos": 29015450116320.0, + "grad_norm": 1.5812051005265035, + "learning_rate": 3.7330722707880734e-06, + "loss": 0.9353, + "num_input_tokens_seen": 69202675, + "step": 2122 + }, + { + "epoch": 0.1914596203273662, + "flos": 23079062413920.0, + "grad_norm": 1.800435395034861, + "learning_rate": 3.7327806089514497e-06, + "loss": 0.6881, + "num_input_tokens_seen": 69231465, + "step": 2123 + }, + { + "epoch": 0.19154980385083645, + "flos": 16814376958560.0, + "grad_norm": 1.5874305800030235, + "learning_rate": 3.7324887992644297e-06, + "loss": 0.8767, + "num_input_tokens_seen": 69258005, + "step": 2124 + }, + { + "epoch": 0.1916399873743067, + "flos": 65553679648320.0, + "grad_norm": 0.7232654787207793, + "learning_rate": 3.7321968417519123e-06, + "loss": 0.5851, + "num_input_tokens_seen": 69350730, + "step": 2125 + }, + { + "epoch": 0.19173017089777697, + "flos": 23990162232000.0, + "grad_norm": 1.8489707480716755, + "learning_rate": 3.7319047364388097e-06, + "loss": 0.7706, + "num_input_tokens_seen": 69379525, + "step": 2126 + }, + { + "epoch": 0.19182035442124723, + "flos": 52182765007680.0, + "grad_norm": 2.874159814552316, + "learning_rate": 3.7316124833500453e-06, + "loss": 0.5711, + "num_input_tokens_seen": 69467920, + "step": 2127 + }, + { + "epoch": 0.1919105379447175, + "flos": 22424474075040.0, + "grad_norm": 1.611783124088371, + "learning_rate": 3.731320082510556e-06, + "loss": 0.766, + "num_input_tokens_seen": 69499000, + "step": 2128 + }, + { + "epoch": 0.19200072146818775, + "flos": 23080920901920.0, + "grad_norm": 2.0858254849078843, + "learning_rate": 3.7310275339452906e-06, + "loss": 0.7917, + "num_input_tokens_seen": 69529175, + "step": 2129 + }, + { + "epoch": 0.192090904991658, + "flos": 22708125212640.0, + "grad_norm": 2.074697780431394, + "learning_rate": 3.7307348376792113e-06, + "loss": 0.767, + "num_input_tokens_seen": 69558385, + "step": 2130 + }, + { + "epoch": 0.19218108851512827, + "flos": 23043002013120.0, + "grad_norm": 1.6511341474205905, + "learning_rate": 3.730441993737292e-06, + "loss": 0.7735, + "num_input_tokens_seen": 69585870, + "step": 2131 + }, + { + "epoch": 0.19227127203859856, + "flos": 22747493722080.0, + "grad_norm": 2.681043187415008, + "learning_rate": 3.7301490021445205e-06, + "loss": 0.7649, + "num_input_tokens_seen": 69612405, + "step": 2132 + }, + { + "epoch": 0.19236145556206882, + "flos": 22933185341280.0, + "grad_norm": 2.6286893077742453, + "learning_rate": 3.7298558629258966e-06, + "loss": 0.7176, + "num_input_tokens_seen": 69638520, + "step": 2133 + }, + { + "epoch": 0.19245163908553908, + "flos": 22132571250720.0, + "grad_norm": 2.0134451263960624, + "learning_rate": 3.7295625761064314e-06, + "loss": 0.8129, + "num_input_tokens_seen": 69665665, + "step": 2134 + }, + { + "epoch": 0.19254182260900934, + "flos": 28580680969920.0, + "grad_norm": 2.061064017969707, + "learning_rate": 3.7292691417111504e-06, + "loss": 0.7442, + "num_input_tokens_seen": 69696275, + "step": 2135 + }, + { + "epoch": 0.1926320061324796, + "flos": 18560701647360.0, + "grad_norm": 2.3936910477592557, + "learning_rate": 3.728975559765092e-06, + "loss": 0.7349, + "num_input_tokens_seen": 69722995, + "step": 2136 + }, + { + "epoch": 0.19272218965594987, + "flos": 23298063871680.0, + "grad_norm": 2.84090725846665, + "learning_rate": 3.728681830293305e-06, + "loss": 0.8077, + "num_input_tokens_seen": 69751170, + "step": 2137 + }, + { + "epoch": 0.19281237317942013, + "flos": 31021185525600.0, + "grad_norm": 1.789035646995015, + "learning_rate": 3.7283879533208523e-06, + "loss": 0.7345, + "num_input_tokens_seen": 69782060, + "step": 2138 + }, + { + "epoch": 0.1929025567028904, + "flos": 25338372891360.0, + "grad_norm": 1.603021840723116, + "learning_rate": 3.7280939288728094e-06, + "loss": 0.7673, + "num_input_tokens_seen": 69812015, + "step": 2139 + }, + { + "epoch": 0.19299274022636065, + "flos": 20966260895040.0, + "grad_norm": 2.0056609860737558, + "learning_rate": 3.7277997569742637e-06, + "loss": 0.7389, + "num_input_tokens_seen": 69841400, + "step": 2140 + }, + { + "epoch": 0.1930829237498309, + "flos": 19435332197280.0, + "grad_norm": 2.32714536396933, + "learning_rate": 3.7275054376503155e-06, + "loss": 0.7427, + "num_input_tokens_seen": 69868315, + "step": 2141 + }, + { + "epoch": 0.19317310727330117, + "flos": 22824037725120.0, + "grad_norm": 7.850443746088373, + "learning_rate": 3.7272109709260783e-06, + "loss": 0.7173, + "num_input_tokens_seen": 69894875, + "step": 2142 + }, + { + "epoch": 0.19326329079677143, + "flos": 65482376581440.0, + "grad_norm": 0.7735593295236183, + "learning_rate": 3.7269163568266774e-06, + "loss": 0.5869, + "num_input_tokens_seen": 69982140, + "step": 2143 + }, + { + "epoch": 0.1933534743202417, + "flos": 21877026185280.0, + "grad_norm": 1.666898147084733, + "learning_rate": 3.7266215953772512e-06, + "loss": 0.763, + "num_input_tokens_seen": 70012060, + "step": 2144 + }, + { + "epoch": 0.19344365784371195, + "flos": 26358471646560.0, + "grad_norm": 1.9279268085627121, + "learning_rate": 3.7263266866029492e-06, + "loss": 0.8111, + "num_input_tokens_seen": 70041300, + "step": 2145 + }, + { + "epoch": 0.1935338413671822, + "flos": 34408106904960.0, + "grad_norm": 1.6362115989241297, + "learning_rate": 3.726031630528936e-06, + "loss": 0.7447, + "num_input_tokens_seen": 70071265, + "step": 2146 + }, + { + "epoch": 0.19362402489065247, + "flos": 26649519566400.0, + "grad_norm": 1.4606380869810038, + "learning_rate": 3.7257364271803865e-06, + "loss": 0.6706, + "num_input_tokens_seen": 70101305, + "step": 2147 + }, + { + "epoch": 0.19371420841412273, + "flos": 21731186282400.0, + "grad_norm": 3.375683609288212, + "learning_rate": 3.7254410765824896e-06, + "loss": 0.6981, + "num_input_tokens_seen": 70127355, + "step": 2148 + }, + { + "epoch": 0.193804391937593, + "flos": 22970249325600.0, + "grad_norm": 1.8420957482601323, + "learning_rate": 3.725145578760446e-06, + "loss": 0.6936, + "num_input_tokens_seen": 70154760, + "step": 2149 + }, + { + "epoch": 0.19389457546106326, + "flos": 24864755612160.0, + "grad_norm": 1.8982572744715687, + "learning_rate": 3.7248499337394696e-06, + "loss": 0.6916, + "num_input_tokens_seen": 70183715, + "step": 2150 + }, + { + "epoch": 0.19398475898453352, + "flos": 19327485522720.0, + "grad_norm": 2.398190762132902, + "learning_rate": 3.7245541415447848e-06, + "loss": 0.8662, + "num_input_tokens_seen": 70208465, + "step": 2151 + }, + { + "epoch": 0.19407494250800378, + "flos": 18962309634240.0, + "grad_norm": 2.215256425468305, + "learning_rate": 3.724258202201633e-06, + "loss": 0.7968, + "num_input_tokens_seen": 70236095, + "step": 2152 + }, + { + "epoch": 0.19416512603147404, + "flos": 24166449901920.0, + "grad_norm": 1.7380293417144954, + "learning_rate": 3.7239621157352633e-06, + "loss": 0.7908, + "num_input_tokens_seen": 70264240, + "step": 2153 + }, + { + "epoch": 0.1942553095549443, + "flos": 29052365421600.0, + "grad_norm": 2.6313211535621406, + "learning_rate": 3.7236658821709403e-06, + "loss": 0.7637, + "num_input_tokens_seen": 70293985, + "step": 2154 + }, + { + "epoch": 0.19434549307841456, + "flos": 19361204228640.0, + "grad_norm": 1.5660133534979004, + "learning_rate": 3.7233695015339404e-06, + "loss": 0.8137, + "num_input_tokens_seen": 70321025, + "step": 2155 + }, + { + "epoch": 0.19443567660188485, + "flos": 26209063446720.0, + "grad_norm": 3.3907278018959723, + "learning_rate": 3.7230729738495513e-06, + "loss": 0.7684, + "num_input_tokens_seen": 70348205, + "step": 2156 + }, + { + "epoch": 0.1945258601253551, + "flos": 26028872952000.0, + "grad_norm": 1.5850510398276274, + "learning_rate": 3.722776299143075e-06, + "loss": 0.769, + "num_input_tokens_seen": 70378530, + "step": 2157 + }, + { + "epoch": 0.19461604364882537, + "flos": 68795058482880.0, + "grad_norm": 0.6681177073581178, + "learning_rate": 3.722479477439826e-06, + "loss": 0.586, + "num_input_tokens_seen": 70468185, + "step": 2158 + }, + { + "epoch": 0.19470622717229563, + "flos": 22459791080640.0, + "grad_norm": 1.4100439391493182, + "learning_rate": 3.7221825087651306e-06, + "loss": 0.7545, + "num_input_tokens_seen": 70497460, + "step": 2159 + }, + { + "epoch": 0.1947964106957659, + "flos": 25957495545600.0, + "grad_norm": 1.8929153944959094, + "learning_rate": 3.7218853931443274e-06, + "loss": 0.7123, + "num_input_tokens_seen": 70527010, + "step": 2160 + }, + { + "epoch": 0.19488659421923615, + "flos": 24354557555520.0, + "grad_norm": 1.6888289693330971, + "learning_rate": 3.721588130602768e-06, + "loss": 0.7137, + "num_input_tokens_seen": 70556970, + "step": 2161 + }, + { + "epoch": 0.19497677774270641, + "flos": 13024360801920.0, + "grad_norm": 2.0683245633431455, + "learning_rate": 3.7212907211658164e-06, + "loss": 0.7496, + "num_input_tokens_seen": 70584560, + "step": 2162 + }, + { + "epoch": 0.19506696126617667, + "flos": 44206557226080.0, + "grad_norm": 1.6135849932309885, + "learning_rate": 3.72099316485885e-06, + "loss": 0.743, + "num_input_tokens_seen": 70616805, + "step": 2163 + }, + { + "epoch": 0.19515714478964694, + "flos": 30141834416160.0, + "grad_norm": 1.9977268175200944, + "learning_rate": 3.720695461707256e-06, + "loss": 0.8058, + "num_input_tokens_seen": 70647695, + "step": 2164 + }, + { + "epoch": 0.1952473283131172, + "flos": 21221434262880.0, + "grad_norm": 1.6610913324755625, + "learning_rate": 3.7203976117364383e-06, + "loss": 0.7898, + "num_input_tokens_seen": 70676265, + "step": 2165 + }, + { + "epoch": 0.19533751183658746, + "flos": 23480001345120.0, + "grad_norm": 1.9284015648831354, + "learning_rate": 3.7200996149718105e-06, + "loss": 0.7585, + "num_input_tokens_seen": 70703605, + "step": 2166 + }, + { + "epoch": 0.19542769536005772, + "flos": 21840891444960.0, + "grad_norm": 1.4153852651348984, + "learning_rate": 3.7198014714387985e-06, + "loss": 0.7624, + "num_input_tokens_seen": 70734655, + "step": 2167 + }, + { + "epoch": 0.19551787888352798, + "flos": 28144462202880.0, + "grad_norm": 1.916910660458341, + "learning_rate": 3.7195031811628422e-06, + "loss": 0.7078, + "num_input_tokens_seen": 70764355, + "step": 2168 + }, + { + "epoch": 0.19560806240699824, + "flos": 22675149901920.0, + "grad_norm": 2.5872402079171417, + "learning_rate": 3.719204744169393e-06, + "loss": 0.808, + "num_input_tokens_seen": 70790265, + "step": 2169 + }, + { + "epoch": 0.1956982459304685, + "flos": 24063175107840.0, + "grad_norm": 2.0294367987055386, + "learning_rate": 3.718906160483916e-06, + "loss": 0.7482, + "num_input_tokens_seen": 70817650, + "step": 2170 + }, + { + "epoch": 0.19578842945393876, + "flos": 22676636692320.0, + "grad_norm": 1.6486944206303962, + "learning_rate": 3.7186074301318868e-06, + "loss": 0.7474, + "num_input_tokens_seen": 70846580, + "step": 2171 + }, + { + "epoch": 0.19587861297740902, + "flos": 21980932865280.0, + "grad_norm": 1.6982945526806577, + "learning_rate": 3.7183085531387957e-06, + "loss": 0.8487, + "num_input_tokens_seen": 70874225, + "step": 2172 + }, + { + "epoch": 0.19596879650087928, + "flos": 24063918503040.0, + "grad_norm": 2.111618092657449, + "learning_rate": 3.7180095295301443e-06, + "loss": 0.7446, + "num_input_tokens_seen": 70903750, + "step": 2173 + }, + { + "epoch": 0.19605898002434954, + "flos": 21146451389760.0, + "grad_norm": 1.3932991034046338, + "learning_rate": 3.7177103593314465e-06, + "loss": 0.8362, + "num_input_tokens_seen": 70934720, + "step": 2174 + }, + { + "epoch": 0.1961491635478198, + "flos": 18597802801440.0, + "grad_norm": 1.8623119848195389, + "learning_rate": 3.7174110425682297e-06, + "loss": 0.8119, + "num_input_tokens_seen": 70963300, + "step": 2175 + }, + { + "epoch": 0.19623934707129007, + "flos": 21913383944160.0, + "grad_norm": 1.7447706154839173, + "learning_rate": 3.7171115792660333e-06, + "loss": 0.8977, + "num_input_tokens_seen": 70989975, + "step": 2176 + }, + { + "epoch": 0.19632953059476033, + "flos": 18156751965600.0, + "grad_norm": 2.3354139474880733, + "learning_rate": 3.7168119694504083e-06, + "loss": 0.8395, + "num_input_tokens_seen": 71016630, + "step": 2177 + }, + { + "epoch": 0.1964197141182306, + "flos": 19979880845760.0, + "grad_norm": 1.8194002804560852, + "learning_rate": 3.71651221314692e-06, + "loss": 0.7661, + "num_input_tokens_seen": 71045885, + "step": 2178 + }, + { + "epoch": 0.19650989764170085, + "flos": 23116200737760.0, + "grad_norm": 2.531670523668389, + "learning_rate": 3.716212310381145e-06, + "loss": 0.7711, + "num_input_tokens_seen": 71074790, + "step": 2179 + }, + { + "epoch": 0.19660008116517114, + "flos": 23808782304960.0, + "grad_norm": 2.030867085677923, + "learning_rate": 3.7159122611786725e-06, + "loss": 0.8176, + "num_input_tokens_seen": 71102115, + "step": 2180 + }, + { + "epoch": 0.1966902646886414, + "flos": 26321965208640.0, + "grad_norm": 2.05380554997085, + "learning_rate": 3.7156120655651045e-06, + "loss": 0.7573, + "num_input_tokens_seen": 71131885, + "step": 2181 + }, + { + "epoch": 0.19678044821211166, + "flos": 27195703684320.0, + "grad_norm": 1.718215096207849, + "learning_rate": 3.7153117235660553e-06, + "loss": 0.756, + "num_input_tokens_seen": 71162975, + "step": 2182 + }, + { + "epoch": 0.19687063173558192, + "flos": 68602936495200.0, + "grad_norm": 0.6937140189110509, + "learning_rate": 3.7150112352071514e-06, + "loss": 0.5848, + "num_input_tokens_seen": 71257295, + "step": 2183 + }, + { + "epoch": 0.19696081525905218, + "flos": 22707939363840.0, + "grad_norm": 1.9594533706883293, + "learning_rate": 3.7147106005140326e-06, + "loss": 0.7429, + "num_input_tokens_seen": 71282405, + "step": 2184 + }, + { + "epoch": 0.19705099878252244, + "flos": 67665254565120.0, + "grad_norm": 0.6843424375617193, + "learning_rate": 3.714409819512351e-06, + "loss": 0.5937, + "num_input_tokens_seen": 71374380, + "step": 2185 + }, + { + "epoch": 0.1971411823059927, + "flos": 20668373739360.0, + "grad_norm": 2.209417352985364, + "learning_rate": 3.7141088922277695e-06, + "loss": 0.7975, + "num_input_tokens_seen": 71399425, + "step": 2186 + }, + { + "epoch": 0.19723136582946296, + "flos": 23916777658560.0, + "grad_norm": 2.042240866538031, + "learning_rate": 3.7138078186859664e-06, + "loss": 0.8177, + "num_input_tokens_seen": 71428220, + "step": 2187 + }, + { + "epoch": 0.19732154935293322, + "flos": 21106859861760.0, + "grad_norm": 1.713712071194506, + "learning_rate": 3.7135065989126303e-06, + "loss": 0.7876, + "num_input_tokens_seen": 71456060, + "step": 2188 + }, + { + "epoch": 0.19741173287640348, + "flos": 64680424379520.0, + "grad_norm": 0.735235562236923, + "learning_rate": 3.713205232933463e-06, + "loss": 0.649, + "num_input_tokens_seen": 71541775, + "step": 2189 + }, + { + "epoch": 0.19750191639987374, + "flos": 17724213004800.0, + "grad_norm": 1.7287262351829213, + "learning_rate": 3.7129037207741792e-06, + "loss": 0.7836, + "num_input_tokens_seen": 71568035, + "step": 2190 + }, + { + "epoch": 0.197592099923344, + "flos": 20851128947520.0, + "grad_norm": 2.8455970550646277, + "learning_rate": 3.7126020624605046e-06, + "loss": 0.7866, + "num_input_tokens_seen": 71595480, + "step": 2191 + }, + { + "epoch": 0.19768228344681427, + "flos": 31750459379520.0, + "grad_norm": 1.6248581164114047, + "learning_rate": 3.7123002580181785e-06, + "loss": 0.7257, + "num_input_tokens_seen": 71628200, + "step": 2192 + }, + { + "epoch": 0.19777246697028453, + "flos": 23332897670400.0, + "grad_norm": 1.3672592580219929, + "learning_rate": 3.7119983074729532e-06, + "loss": 0.8478, + "num_input_tokens_seen": 71656245, + "step": 2193 + }, + { + "epoch": 0.1978626504937548, + "flos": 20963733351360.0, + "grad_norm": 2.3916180112809706, + "learning_rate": 3.7116962108505926e-06, + "loss": 0.8289, + "num_input_tokens_seen": 71683470, + "step": 2194 + }, + { + "epoch": 0.19795283401722505, + "flos": 22641951572640.0, + "grad_norm": 1.6076218769152006, + "learning_rate": 3.711393968176873e-06, + "loss": 0.8722, + "num_input_tokens_seen": 71713115, + "step": 2195 + }, + { + "epoch": 0.1980430175406953, + "flos": 20961949202880.0, + "grad_norm": 1.4748346624004491, + "learning_rate": 3.711091579477584e-06, + "loss": 0.7845, + "num_input_tokens_seen": 71742395, + "step": 2196 + }, + { + "epoch": 0.19813320106416557, + "flos": 17613206900640.0, + "grad_norm": 2.1182617615204746, + "learning_rate": 3.7107890447785255e-06, + "loss": 0.8041, + "num_input_tokens_seen": 71768260, + "step": 2197 + }, + { + "epoch": 0.19822338458763583, + "flos": 22018405716960.0, + "grad_norm": 2.971889562400655, + "learning_rate": 3.710486364105513e-06, + "loss": 0.8299, + "num_input_tokens_seen": 71797475, + "step": 2198 + }, + { + "epoch": 0.1983135681111061, + "flos": 18124074012960.0, + "grad_norm": 2.0720350476986633, + "learning_rate": 3.7101835374843728e-06, + "loss": 0.7729, + "num_input_tokens_seen": 71823320, + "step": 2199 + }, + { + "epoch": 0.19840375163457635, + "flos": 26066754671040.0, + "grad_norm": 1.9599237139122199, + "learning_rate": 3.7098805649409427e-06, + "loss": 0.7965, + "num_input_tokens_seen": 71853625, + "step": 2200 + }, + { + "epoch": 0.1984939351580466, + "flos": 19909098155520.0, + "grad_norm": 2.381952749768586, + "learning_rate": 3.7095774465010748e-06, + "loss": 0.8437, + "num_input_tokens_seen": 71878475, + "step": 2201 + }, + { + "epoch": 0.19858411868151687, + "flos": 25883553425760.0, + "grad_norm": 1.9733487264037104, + "learning_rate": 3.7092741821906328e-06, + "loss": 0.7711, + "num_input_tokens_seen": 71908090, + "step": 2202 + }, + { + "epoch": 0.19867430220498714, + "flos": 22819279995840.0, + "grad_norm": 2.0082180934571503, + "learning_rate": 3.708970772035493e-06, + "loss": 0.7583, + "num_input_tokens_seen": 71933445, + "step": 2203 + }, + { + "epoch": 0.19876448572845742, + "flos": 24203625395520.0, + "grad_norm": 1.7461831840124007, + "learning_rate": 3.7086672160615427e-06, + "loss": 0.7441, + "num_input_tokens_seen": 71963165, + "step": 2204 + }, + { + "epoch": 0.19885466925192768, + "flos": 23479592477760.0, + "grad_norm": 1.7488400838688862, + "learning_rate": 3.7083635142946852e-06, + "loss": 0.7614, + "num_input_tokens_seen": 71992855, + "step": 2205 + }, + { + "epoch": 0.19894485277539795, + "flos": 27450839882400.0, + "grad_norm": 1.7325125652259412, + "learning_rate": 3.7080596667608327e-06, + "loss": 0.8153, + "num_input_tokens_seen": 72019340, + "step": 2206 + }, + { + "epoch": 0.1990350362988682, + "flos": 23007090291360.0, + "grad_norm": 1.8583010609423236, + "learning_rate": 3.707755673485911e-06, + "loss": 0.8215, + "num_input_tokens_seen": 72049205, + "step": 2207 + }, + { + "epoch": 0.19912521982233847, + "flos": 17504170793760.0, + "grad_norm": 2.0631029808166015, + "learning_rate": 3.7074515344958584e-06, + "loss": 0.6887, + "num_input_tokens_seen": 72073915, + "step": 2208 + }, + { + "epoch": 0.19921540334580873, + "flos": 67508481019200.0, + "grad_norm": 0.6657422083428448, + "learning_rate": 3.707147249816627e-06, + "loss": 0.6133, + "num_input_tokens_seen": 72167200, + "step": 2209 + }, + { + "epoch": 0.199305586869279, + "flos": 25520459043840.0, + "grad_norm": 1.659054379280753, + "learning_rate": 3.706842819474178e-06, + "loss": 0.7999, + "num_input_tokens_seen": 72197430, + "step": 2210 + }, + { + "epoch": 0.19939577039274925, + "flos": 13533666784320.0, + "grad_norm": 1.8121503295349137, + "learning_rate": 3.706538243494489e-06, + "loss": 0.7133, + "num_input_tokens_seen": 72225420, + "step": 2211 + }, + { + "epoch": 0.1994859539162195, + "flos": 35028790689120.0, + "grad_norm": 2.125433709258798, + "learning_rate": 3.706233521903547e-06, + "loss": 0.75, + "num_input_tokens_seen": 72252415, + "step": 2212 + }, + { + "epoch": 0.19957613743968977, + "flos": 16595152482240.0, + "grad_norm": 1.9001534952100283, + "learning_rate": 3.705928654727353e-06, + "loss": 0.8589, + "num_input_tokens_seen": 72278785, + "step": 2213 + }, + { + "epoch": 0.19966632096316003, + "flos": 22606337208960.0, + "grad_norm": 1.5033333616122788, + "learning_rate": 3.7056236419919195e-06, + "loss": 0.7532, + "num_input_tokens_seen": 72309810, + "step": 2214 + }, + { + "epoch": 0.1997565044866303, + "flos": 28795742433120.0, + "grad_norm": 2.0150985621728275, + "learning_rate": 3.705318483723273e-06, + "loss": 0.829, + "num_input_tokens_seen": 72341445, + "step": 2215 + }, + { + "epoch": 0.19984668801010055, + "flos": 23147577748800.0, + "grad_norm": 1.9099631494692044, + "learning_rate": 3.7050131799474493e-06, + "loss": 0.7527, + "num_input_tokens_seen": 72366075, + "step": 2216 + }, + { + "epoch": 0.19993687153357081, + "flos": 22895192112960.0, + "grad_norm": 1.537367841421511, + "learning_rate": 3.7047077306905e-06, + "loss": 0.7778, + "num_input_tokens_seen": 72395585, + "step": 2217 + }, + { + "epoch": 0.20002705505704108, + "flos": 26423790382080.0, + "grad_norm": 1.6358677511275406, + "learning_rate": 3.704402135978488e-06, + "loss": 0.8119, + "num_input_tokens_seen": 72425340, + "step": 2218 + }, + { + "epoch": 0.20011723858051134, + "flos": 17869309512480.0, + "grad_norm": 3.372346187898739, + "learning_rate": 3.7040963958374877e-06, + "loss": 0.7987, + "num_input_tokens_seen": 72448705, + "step": 2219 + }, + { + "epoch": 0.2002074221039816, + "flos": 46689292362720.0, + "grad_norm": 9.478947653004443, + "learning_rate": 3.7037905102935864e-06, + "loss": 0.8085, + "num_input_tokens_seen": 72483840, + "step": 2220 + }, + { + "epoch": 0.20029760562745186, + "flos": 20598445953600.0, + "grad_norm": 1.6352877312958067, + "learning_rate": 3.7034844793728837e-06, + "loss": 0.7856, + "num_input_tokens_seen": 72509365, + "step": 2221 + }, + { + "epoch": 0.20038778915092212, + "flos": 32150246048160.0, + "grad_norm": 1.5705582156025932, + "learning_rate": 3.7031783031014933e-06, + "loss": 0.6674, + "num_input_tokens_seen": 72542380, + "step": 2222 + }, + { + "epoch": 0.20047797267439238, + "flos": 21185634050400.0, + "grad_norm": 1.925137549601234, + "learning_rate": 3.702871981505538e-06, + "loss": 0.7536, + "num_input_tokens_seen": 72570270, + "step": 2223 + }, + { + "epoch": 0.20056815619786264, + "flos": 23735509240800.0, + "grad_norm": 2.114425152153667, + "learning_rate": 3.7025655146111563e-06, + "loss": 0.7191, + "num_input_tokens_seen": 72601020, + "step": 2224 + }, + { + "epoch": 0.2006583397213329, + "flos": 71027192132160.0, + "grad_norm": 0.6923068156858339, + "learning_rate": 3.702258902444497e-06, + "loss": 0.6642, + "num_input_tokens_seen": 72701640, + "step": 2225 + }, + { + "epoch": 0.20074852324480316, + "flos": 25192756007040.0, + "grad_norm": 1.711778547251461, + "learning_rate": 3.701952145031722e-06, + "loss": 0.8317, + "num_input_tokens_seen": 72732765, + "step": 2226 + }, + { + "epoch": 0.20083870676827342, + "flos": 24354929253120.0, + "grad_norm": 1.5466207430176746, + "learning_rate": 3.701645242399005e-06, + "loss": 0.8604, + "num_input_tokens_seen": 72762785, + "step": 2227 + }, + { + "epoch": 0.2009288902917437, + "flos": 19400015191680.0, + "grad_norm": 10.86723032544581, + "learning_rate": 3.701338194572533e-06, + "loss": 0.8429, + "num_input_tokens_seen": 72790735, + "step": 2228 + }, + { + "epoch": 0.20101907381521397, + "flos": 15792605564160.0, + "grad_norm": 2.074116907295583, + "learning_rate": 3.7010310015785056e-06, + "loss": 0.7586, + "num_input_tokens_seen": 72816545, + "step": 2229 + }, + { + "epoch": 0.20110925733868423, + "flos": 23954250510240.0, + "grad_norm": 1.6460801032124823, + "learning_rate": 3.700723663443134e-06, + "loss": 0.72, + "num_input_tokens_seen": 72845970, + "step": 2230 + }, + { + "epoch": 0.2011994408621545, + "flos": 55317560939520.0, + "grad_norm": 0.7142487848476479, + "learning_rate": 3.7004161801926416e-06, + "loss": 0.5907, + "num_input_tokens_seen": 72931035, + "step": 2231 + }, + { + "epoch": 0.20128962438562475, + "flos": 19394365388160.0, + "grad_norm": 1.6911318514564795, + "learning_rate": 3.7001085518532643e-06, + "loss": 0.8017, + "num_input_tokens_seen": 72957565, + "step": 2232 + }, + { + "epoch": 0.20137980790909502, + "flos": 22314620233440.0, + "grad_norm": 2.7200076822565085, + "learning_rate": 3.6998007784512515e-06, + "loss": 0.7723, + "num_input_tokens_seen": 72986075, + "step": 2233 + }, + { + "epoch": 0.20146999143256528, + "flos": 37356579331680.0, + "grad_norm": 1.623292539437781, + "learning_rate": 3.6994928600128637e-06, + "loss": 0.6545, + "num_input_tokens_seen": 73019705, + "step": 2234 + }, + { + "epoch": 0.20156017495603554, + "flos": 23804173254720.0, + "grad_norm": 1.566627619980469, + "learning_rate": 3.6991847965643742e-06, + "loss": 0.7433, + "num_input_tokens_seen": 73047230, + "step": 2235 + }, + { + "epoch": 0.2016503584795058, + "flos": 22678829708160.0, + "grad_norm": 2.2565714979553655, + "learning_rate": 3.698876588132068e-06, + "loss": 0.7348, + "num_input_tokens_seen": 73074905, + "step": 2236 + }, + { + "epoch": 0.20174054200297606, + "flos": 25738196729760.0, + "grad_norm": 1.9584928677555393, + "learning_rate": 3.6985682347422446e-06, + "loss": 0.739, + "num_input_tokens_seen": 73105740, + "step": 2237 + }, + { + "epoch": 0.20183072552644632, + "flos": 18342889621920.0, + "grad_norm": 1.9515095858581606, + "learning_rate": 3.698259736421213e-06, + "loss": 0.7906, + "num_input_tokens_seen": 73131660, + "step": 2238 + }, + { + "epoch": 0.20192090904991658, + "flos": 23694505261920.0, + "grad_norm": 1.9163915049642315, + "learning_rate": 3.697951093195297e-06, + "loss": 0.7599, + "num_input_tokens_seen": 73160475, + "step": 2239 + }, + { + "epoch": 0.20201109257338684, + "flos": 25809313947840.0, + "grad_norm": 1.7076549656308433, + "learning_rate": 3.6976423050908307e-06, + "loss": 0.7529, + "num_input_tokens_seen": 73188450, + "step": 2240 + }, + { + "epoch": 0.2021012760968571, + "flos": 27524298795360.0, + "grad_norm": 1.5769803267602198, + "learning_rate": 3.697333372134163e-06, + "loss": 0.7658, + "num_input_tokens_seen": 73216910, + "step": 2241 + }, + { + "epoch": 0.20219145962032736, + "flos": 22897682486880.0, + "grad_norm": 2.4792826542582453, + "learning_rate": 3.697024294351653e-06, + "loss": 0.8714, + "num_input_tokens_seen": 73245200, + "step": 2242 + }, + { + "epoch": 0.20228164314379762, + "flos": 24386083245600.0, + "grad_norm": 1.7709569850460234, + "learning_rate": 3.696715071769672e-06, + "loss": 0.7381, + "num_input_tokens_seen": 73271940, + "step": 2243 + }, + { + "epoch": 0.20237182666726788, + "flos": 24026631500160.0, + "grad_norm": 2.1643250233858704, + "learning_rate": 3.696405704414606e-06, + "loss": 0.7756, + "num_input_tokens_seen": 73301895, + "step": 2244 + }, + { + "epoch": 0.20246201019073815, + "flos": 27304776960960.0, + "grad_norm": 1.7952834668978388, + "learning_rate": 3.6960961923128514e-06, + "loss": 0.8801, + "num_input_tokens_seen": 73332650, + "step": 2245 + }, + { + "epoch": 0.2025521937142084, + "flos": 22351052331840.0, + "grad_norm": 2.774635077873963, + "learning_rate": 3.6957865354908177e-06, + "loss": 0.7073, + "num_input_tokens_seen": 73360300, + "step": 2246 + }, + { + "epoch": 0.20264237723767867, + "flos": 28543802834400.0, + "grad_norm": 1.7924550635798384, + "learning_rate": 3.6954767339749262e-06, + "loss": 0.7985, + "num_input_tokens_seen": 73389690, + "step": 2247 + }, + { + "epoch": 0.20273256076114893, + "flos": 34300483248960.0, + "grad_norm": 2.348350282675134, + "learning_rate": 3.6951667877916113e-06, + "loss": 0.7523, + "num_input_tokens_seen": 73419375, + "step": 2248 + }, + { + "epoch": 0.2028227442846192, + "flos": 30474443861280.0, + "grad_norm": 2.0932840996113717, + "learning_rate": 3.694856696967319e-06, + "loss": 0.784, + "num_input_tokens_seen": 73453025, + "step": 2249 + }, + { + "epoch": 0.20291292780808945, + "flos": 22277742097920.0, + "grad_norm": 1.9070996126504276, + "learning_rate": 3.6945464615285077e-06, + "loss": 0.7647, + "num_input_tokens_seen": 73482890, + "step": 2250 + }, + { + "epoch": 0.2030031113315597, + "flos": 71213404128000.0, + "grad_norm": 0.6995691668075952, + "learning_rate": 3.694236081501648e-06, + "loss": 0.6912, + "num_input_tokens_seen": 73582080, + "step": 2251 + }, + { + "epoch": 0.20309329485503, + "flos": 25627302134880.0, + "grad_norm": 1.5717301683080351, + "learning_rate": 3.6939255569132246e-06, + "loss": 0.7387, + "num_input_tokens_seen": 73612510, + "step": 2252 + }, + { + "epoch": 0.20318347837850026, + "flos": 70070516454720.0, + "grad_norm": 0.8386482068414678, + "learning_rate": 3.693614887789733e-06, + "loss": 0.6199, + "num_input_tokens_seen": 73703070, + "step": 2253 + }, + { + "epoch": 0.20327366190197052, + "flos": 68231844881280.0, + "grad_norm": 0.7810926278875646, + "learning_rate": 3.69330407415768e-06, + "loss": 0.6205, + "num_input_tokens_seen": 73793915, + "step": 2254 + }, + { + "epoch": 0.20336384542544078, + "flos": 31853436815520.0, + "grad_norm": 1.717617979925507, + "learning_rate": 3.6929931160435867e-06, + "loss": 0.7815, + "num_input_tokens_seen": 73823780, + "step": 2255 + }, + { + "epoch": 0.20345402894891104, + "flos": 18525496151040.0, + "grad_norm": 2.230511933825613, + "learning_rate": 3.6926820134739858e-06, + "loss": 0.8326, + "num_input_tokens_seen": 73851930, + "step": 2256 + }, + { + "epoch": 0.2035442124723813, + "flos": 21621666968640.0, + "grad_norm": 3.1596673273667446, + "learning_rate": 3.692370766475422e-06, + "loss": 0.7137, + "num_input_tokens_seen": 73880420, + "step": 2257 + }, + { + "epoch": 0.20363439599585156, + "flos": 19580317195680.0, + "grad_norm": 2.3840758929749604, + "learning_rate": 3.692059375074453e-06, + "loss": 0.863, + "num_input_tokens_seen": 73905050, + "step": 2258 + }, + { + "epoch": 0.20372457951932182, + "flos": 36012903383040.0, + "grad_norm": 1.596257201128484, + "learning_rate": 3.6917478392976475e-06, + "loss": 0.8432, + "num_input_tokens_seen": 73935025, + "step": 2259 + }, + { + "epoch": 0.20381476304279209, + "flos": 29490405506880.0, + "grad_norm": 1.746013255472106, + "learning_rate": 3.691436159171589e-06, + "loss": 0.8378, + "num_input_tokens_seen": 73967070, + "step": 2260 + }, + { + "epoch": 0.20390494656626235, + "flos": 16412025576480.0, + "grad_norm": 1.9555181261596353, + "learning_rate": 3.6911243347228703e-06, + "loss": 0.7557, + "num_input_tokens_seen": 73994860, + "step": 2261 + }, + { + "epoch": 0.2039951300897326, + "flos": 62942165528640.0, + "grad_norm": 0.6361243158697706, + "learning_rate": 3.690812365978099e-06, + "loss": 0.533, + "num_input_tokens_seen": 74092495, + "step": 2262 + }, + { + "epoch": 0.20408531361320287, + "flos": 23442825851520.0, + "grad_norm": 1.8455586371809798, + "learning_rate": 3.690500252963893e-06, + "loss": 0.7325, + "num_input_tokens_seen": 74122160, + "step": 2263 + }, + { + "epoch": 0.20417549713667313, + "flos": 21252365236800.0, + "grad_norm": 1.8493485231557265, + "learning_rate": 3.6901879957068846e-06, + "loss": 0.7661, + "num_input_tokens_seen": 74151625, + "step": 2264 + }, + { + "epoch": 0.2042656806601434, + "flos": 25768421478240.0, + "grad_norm": 1.8744403003741539, + "learning_rate": 3.689875594233717e-06, + "loss": 0.8652, + "num_input_tokens_seen": 74182435, + "step": 2265 + }, + { + "epoch": 0.20435586418361365, + "flos": 27851146927680.0, + "grad_norm": 1.6278283374270062, + "learning_rate": 3.689563048571046e-06, + "loss": 0.7575, + "num_input_tokens_seen": 74213550, + "step": 2266 + }, + { + "epoch": 0.2044460477070839, + "flos": 20194384762560.0, + "grad_norm": 2.392967223622442, + "learning_rate": 3.6892503587455395e-06, + "loss": 0.7708, + "num_input_tokens_seen": 74241450, + "step": 2267 + }, + { + "epoch": 0.20453623123055417, + "flos": 42353538125280.0, + "grad_norm": 1.613572507710881, + "learning_rate": 3.6889375247838766e-06, + "loss": 0.7891, + "num_input_tokens_seen": 74275920, + "step": 2268 + }, + { + "epoch": 0.20462641475402443, + "flos": 21222214827840.0, + "grad_norm": 1.5461123349237498, + "learning_rate": 3.688624546712753e-06, + "loss": 0.7783, + "num_input_tokens_seen": 74304425, + "step": 2269 + }, + { + "epoch": 0.2047165982774947, + "flos": 24823305596160.0, + "grad_norm": 1.593183603806423, + "learning_rate": 3.688311424558871e-06, + "loss": 0.7295, + "num_input_tokens_seen": 74334450, + "step": 2270 + }, + { + "epoch": 0.20480678180096495, + "flos": 22605407964960.0, + "grad_norm": 2.549884539215468, + "learning_rate": 3.6879981583489496e-06, + "loss": 0.675, + "num_input_tokens_seen": 74357840, + "step": 2271 + }, + { + "epoch": 0.20489696532443522, + "flos": 38745979818720.0, + "grad_norm": 1.7324246274287605, + "learning_rate": 3.687684748109718e-06, + "loss": 0.7919, + "num_input_tokens_seen": 74391895, + "step": 2272 + }, + { + "epoch": 0.20498714884790548, + "flos": 15137125151040.0, + "grad_norm": 2.1321661671982035, + "learning_rate": 3.6873711938679174e-06, + "loss": 0.838, + "num_input_tokens_seen": 74418670, + "step": 2273 + }, + { + "epoch": 0.20507733237137574, + "flos": 19796642430720.0, + "grad_norm": 1.5698329515941003, + "learning_rate": 3.6870574956503027e-06, + "loss": 0.7669, + "num_input_tokens_seen": 74447740, + "step": 2274 + }, + { + "epoch": 0.20516751589484603, + "flos": 20055458435040.0, + "grad_norm": 2.0536308327369768, + "learning_rate": 3.68674365348364e-06, + "loss": 0.7666, + "num_input_tokens_seen": 74471610, + "step": 2275 + }, + { + "epoch": 0.2052576994183163, + "flos": 19865789651520.0, + "grad_norm": 3.3225558253435583, + "learning_rate": 3.6864296673947086e-06, + "loss": 0.7802, + "num_input_tokens_seen": 74498440, + "step": 2276 + }, + { + "epoch": 0.20534788294178655, + "flos": 20561679327360.0, + "grad_norm": 2.058939432439988, + "learning_rate": 3.686115537410298e-06, + "loss": 0.7573, + "num_input_tokens_seen": 74525500, + "step": 2277 + }, + { + "epoch": 0.2054380664652568, + "flos": 39146398373280.0, + "grad_norm": 1.9084844782025447, + "learning_rate": 3.685801263557214e-06, + "loss": 0.7258, + "num_input_tokens_seen": 74558120, + "step": 2278 + }, + { + "epoch": 0.20552824998872707, + "flos": 18561928249440.0, + "grad_norm": 1.8884198909268675, + "learning_rate": 3.68548684586227e-06, + "loss": 0.8689, + "num_input_tokens_seen": 74582930, + "step": 2279 + }, + { + "epoch": 0.20561843351219733, + "flos": 30985348143360.0, + "grad_norm": 2.6038198214658963, + "learning_rate": 3.685172284352295e-06, + "loss": 0.7718, + "num_input_tokens_seen": 74611805, + "step": 2280 + }, + { + "epoch": 0.2057086170356676, + "flos": 28438781061600.0, + "grad_norm": 1.5797160975191038, + "learning_rate": 3.684857579054128e-06, + "loss": 0.7086, + "num_input_tokens_seen": 74643685, + "step": 2281 + }, + { + "epoch": 0.20579880055913785, + "flos": 26169063051360.0, + "grad_norm": 2.1707587358052267, + "learning_rate": 3.6845427299946233e-06, + "loss": 0.8365, + "num_input_tokens_seen": 74669935, + "step": 2282 + }, + { + "epoch": 0.2058889840826081, + "flos": 23327582394720.0, + "grad_norm": 1.8555367226447586, + "learning_rate": 3.6842277372006434e-06, + "loss": 0.8733, + "num_input_tokens_seen": 74696445, + "step": 2283 + }, + { + "epoch": 0.20597916760607837, + "flos": 26429142827520.0, + "grad_norm": 2.5099353777405007, + "learning_rate": 3.6839126006990664e-06, + "loss": 0.7467, + "num_input_tokens_seen": 74729850, + "step": 2284 + }, + { + "epoch": 0.20606935112954863, + "flos": 31490937149760.0, + "grad_norm": 1.498295928663578, + "learning_rate": 3.6835973205167818e-06, + "loss": 0.8297, + "num_input_tokens_seen": 74761885, + "step": 2285 + }, + { + "epoch": 0.2061595346530189, + "flos": 18847958251680.0, + "grad_norm": 2.0431399446899485, + "learning_rate": 3.6832818966806904e-06, + "loss": 0.8118, + "num_input_tokens_seen": 74787415, + "step": 2286 + }, + { + "epoch": 0.20624971817648916, + "flos": 24425042887680.0, + "grad_norm": 1.8540438350610486, + "learning_rate": 3.682966329217706e-06, + "loss": 0.8559, + "num_input_tokens_seen": 74816905, + "step": 2287 + }, + { + "epoch": 0.20633990169995942, + "flos": 20785029647040.0, + "grad_norm": 1.7909556537933513, + "learning_rate": 3.6826506181547543e-06, + "loss": 0.8105, + "num_input_tokens_seen": 74844105, + "step": 2288 + }, + { + "epoch": 0.20643008522342968, + "flos": 21148309877760.0, + "grad_norm": 3.362905944371034, + "learning_rate": 3.682334763518774e-06, + "loss": 0.8011, + "num_input_tokens_seen": 74871905, + "step": 2289 + }, + { + "epoch": 0.20652026874689994, + "flos": 23480447382240.0, + "grad_norm": 1.9222421683058775, + "learning_rate": 3.6820187653367158e-06, + "loss": 0.7449, + "num_input_tokens_seen": 74900935, + "step": 2290 + }, + { + "epoch": 0.2066104522703702, + "flos": 45591683190720.0, + "grad_norm": 1.6727530767218373, + "learning_rate": 3.6817026236355412e-06, + "loss": 0.6684, + "num_input_tokens_seen": 74934750, + "step": 2291 + }, + { + "epoch": 0.20670063579384046, + "flos": 28690014434880.0, + "grad_norm": 3.43134325244696, + "learning_rate": 3.681386338442227e-06, + "loss": 0.7551, + "num_input_tokens_seen": 74964445, + "step": 2292 + }, + { + "epoch": 0.20679081931731072, + "flos": 25701764631360.0, + "grad_norm": 1.5542986206136922, + "learning_rate": 3.681069909783758e-06, + "loss": 0.7417, + "num_input_tokens_seen": 74995965, + "step": 2293 + }, + { + "epoch": 0.20688100284078098, + "flos": 20492680785600.0, + "grad_norm": 1.978594673889759, + "learning_rate": 3.680753337687136e-06, + "loss": 0.7848, + "num_input_tokens_seen": 75024900, + "step": 2294 + }, + { + "epoch": 0.20697118636425124, + "flos": 24209275199040.0, + "grad_norm": 2.3435830629024843, + "learning_rate": 3.680436622179371e-06, + "loss": 0.7026, + "num_input_tokens_seen": 75052235, + "step": 2295 + }, + { + "epoch": 0.2070613698877215, + "flos": 20383756188000.0, + "grad_norm": 1.976280584321854, + "learning_rate": 3.680119763287488e-06, + "loss": 0.7616, + "num_input_tokens_seen": 75079610, + "step": 2296 + }, + { + "epoch": 0.20715155341119176, + "flos": 17979460712160.0, + "grad_norm": 2.129722396777384, + "learning_rate": 3.6798027610385227e-06, + "loss": 0.7141, + "num_input_tokens_seen": 75107910, + "step": 2297 + }, + { + "epoch": 0.20724173693466202, + "flos": 51783207091200.0, + "grad_norm": 2.3073132056844767, + "learning_rate": 3.6794856154595235e-06, + "loss": 0.8243, + "num_input_tokens_seen": 75137155, + "step": 2298 + }, + { + "epoch": 0.2073319204581323, + "flos": 19800322236960.0, + "grad_norm": 2.2538056508163278, + "learning_rate": 3.6791683265775506e-06, + "loss": 0.7827, + "num_input_tokens_seen": 75164050, + "step": 2299 + }, + { + "epoch": 0.20742210398160257, + "flos": 26321816529600.0, + "grad_norm": 2.659007088911772, + "learning_rate": 3.6788508944196773e-06, + "loss": 0.737, + "num_input_tokens_seen": 75193140, + "step": 2300 + }, + { + "epoch": 0.20751228750507283, + "flos": 25336774591680.0, + "grad_norm": 1.4234983797590055, + "learning_rate": 3.678533319012989e-06, + "loss": 0.8511, + "num_input_tokens_seen": 75224165, + "step": 2301 + }, + { + "epoch": 0.2076024710285431, + "flos": 25483469399040.0, + "grad_norm": 1.787862316838449, + "learning_rate": 3.6782156003845826e-06, + "loss": 0.7258, + "num_input_tokens_seen": 75251440, + "step": 2302 + }, + { + "epoch": 0.20769265455201336, + "flos": 65628699691200.0, + "grad_norm": 0.6422542927109643, + "learning_rate": 3.6778977385615676e-06, + "loss": 0.5952, + "num_input_tokens_seen": 75344795, + "step": 2303 + }, + { + "epoch": 0.20778283807548362, + "flos": 30037816226880.0, + "grad_norm": 1.7777716823645575, + "learning_rate": 3.6775797335710656e-06, + "loss": 0.8275, + "num_input_tokens_seen": 75372065, + "step": 2304 + }, + { + "epoch": 0.20787302159895388, + "flos": 21364597943040.0, + "grad_norm": 1.553797097877152, + "learning_rate": 3.6772615854402105e-06, + "loss": 0.725, + "num_input_tokens_seen": 75401405, + "step": 2305 + }, + { + "epoch": 0.20796320512242414, + "flos": 48144829320000.0, + "grad_norm": 1.354886284381476, + "learning_rate": 3.6769432941961487e-06, + "loss": 0.6548, + "num_input_tokens_seen": 75434465, + "step": 2306 + }, + { + "epoch": 0.2080533886458944, + "flos": 39801618598080.0, + "grad_norm": 1.8049618070793794, + "learning_rate": 3.676624859866038e-06, + "loss": 0.7742, + "num_input_tokens_seen": 75465960, + "step": 2307 + }, + { + "epoch": 0.20814357216936466, + "flos": 15646840000800.0, + "grad_norm": 2.813709253383022, + "learning_rate": 3.67630628247705e-06, + "loss": 0.7031, + "num_input_tokens_seen": 75490965, + "step": 2308 + }, + { + "epoch": 0.20823375569283492, + "flos": 21221620111680.0, + "grad_norm": 2.029494715602674, + "learning_rate": 3.675987562056367e-06, + "loss": 0.8397, + "num_input_tokens_seen": 75517715, + "step": 2309 + }, + { + "epoch": 0.20832393921630518, + "flos": 22642286100480.0, + "grad_norm": 1.8199706186291416, + "learning_rate": 3.675668698631184e-06, + "loss": 0.8113, + "num_input_tokens_seen": 75546450, + "step": 2310 + }, + { + "epoch": 0.20841412273977544, + "flos": 16151128065600.0, + "grad_norm": 2.2765052259332004, + "learning_rate": 3.675349692228708e-06, + "loss": 0.8376, + "num_input_tokens_seen": 75573640, + "step": 2311 + }, + { + "epoch": 0.2085043062632457, + "flos": 27560730893760.0, + "grad_norm": 2.1340396690292516, + "learning_rate": 3.6750305428761578e-06, + "loss": 0.7468, + "num_input_tokens_seen": 75602875, + "step": 2312 + }, + { + "epoch": 0.20859448978671596, + "flos": 14517593629440.0, + "grad_norm": 2.1029119874002595, + "learning_rate": 3.674711250600766e-06, + "loss": 0.7836, + "num_input_tokens_seen": 75627655, + "step": 2313 + }, + { + "epoch": 0.20868467331018623, + "flos": 70923805828800.0, + "grad_norm": 0.6770462368315097, + "learning_rate": 3.6743918154297765e-06, + "loss": 0.5766, + "num_input_tokens_seen": 75717175, + "step": 2314 + }, + { + "epoch": 0.20877485683365649, + "flos": 24827505779040.0, + "grad_norm": 2.022052069668022, + "learning_rate": 3.6740722373904446e-06, + "loss": 0.7652, + "num_input_tokens_seen": 75743895, + "step": 2315 + }, + { + "epoch": 0.20886504035712675, + "flos": 32295862932480.0, + "grad_norm": 1.7628759402562204, + "learning_rate": 3.6737525165100383e-06, + "loss": 0.7519, + "num_input_tokens_seen": 75776265, + "step": 2316 + }, + { + "epoch": 0.208955223880597, + "flos": 26941050693120.0, + "grad_norm": 1.760580096662437, + "learning_rate": 3.6734326528158385e-06, + "loss": 0.7928, + "num_input_tokens_seen": 75806740, + "step": 2317 + }, + { + "epoch": 0.20904540740406727, + "flos": 36376034934720.0, + "grad_norm": 1.789125799533503, + "learning_rate": 3.673112646335138e-06, + "loss": 0.764, + "num_input_tokens_seen": 75836270, + "step": 2318 + }, + { + "epoch": 0.20913559092753753, + "flos": 36741359502240.0, + "grad_norm": 2.38007374945745, + "learning_rate": 3.672792497095241e-06, + "loss": 0.7235, + "num_input_tokens_seen": 75869595, + "step": 2319 + }, + { + "epoch": 0.2092257744510078, + "flos": 17978940335520.0, + "grad_norm": 2.229446845109445, + "learning_rate": 3.672472205123464e-06, + "loss": 0.7725, + "num_input_tokens_seen": 75896240, + "step": 2320 + }, + { + "epoch": 0.20931595797447805, + "flos": 22752214281600.0, + "grad_norm": 1.8454957229907825, + "learning_rate": 3.6721517704471363e-06, + "loss": 0.7045, + "num_input_tokens_seen": 75927475, + "step": 2321 + }, + { + "epoch": 0.2094061414979483, + "flos": 22165249203360.0, + "grad_norm": 1.7544977426892183, + "learning_rate": 3.6718311930936e-06, + "loss": 0.8097, + "num_input_tokens_seen": 75956990, + "step": 2322 + }, + { + "epoch": 0.2094963250214186, + "flos": 16332024785760.0, + "grad_norm": 2.213828309891047, + "learning_rate": 3.6715104730902074e-06, + "loss": 0.7285, + "num_input_tokens_seen": 75979860, + "step": 2323 + }, + { + "epoch": 0.20958650854488886, + "flos": 19436224271520.0, + "grad_norm": 2.0696876590618687, + "learning_rate": 3.671189610464325e-06, + "loss": 0.8121, + "num_input_tokens_seen": 76005250, + "step": 2324 + }, + { + "epoch": 0.20967669206835912, + "flos": 19905604198080.0, + "grad_norm": 2.15997722083902, + "learning_rate": 3.6708686052433303e-06, + "loss": 0.7233, + "num_input_tokens_seen": 76031485, + "step": 2325 + }, + { + "epoch": 0.20976687559182938, + "flos": 27743337422880.0, + "grad_norm": 1.7769748179079534, + "learning_rate": 3.6705474574546127e-06, + "loss": 0.7898, + "num_input_tokens_seen": 76060710, + "step": 2326 + }, + { + "epoch": 0.20985705911529964, + "flos": 26357653911840.0, + "grad_norm": 3.62845972638004, + "learning_rate": 3.670226167125575e-06, + "loss": 0.7931, + "num_input_tokens_seen": 76091085, + "step": 2327 + }, + { + "epoch": 0.2099472426387699, + "flos": 17758786615200.0, + "grad_norm": 1.915621189032742, + "learning_rate": 3.6699047342836313e-06, + "loss": 0.7773, + "num_input_tokens_seen": 76118165, + "step": 2328 + }, + { + "epoch": 0.21003742616224017, + "flos": 28032601194240.0, + "grad_norm": 1.450692550749654, + "learning_rate": 3.669583158956208e-06, + "loss": 0.8097, + "num_input_tokens_seen": 76149710, + "step": 2329 + }, + { + "epoch": 0.21012760968571043, + "flos": 62034819856320.0, + "grad_norm": 0.6745523734603679, + "learning_rate": 3.669261441170743e-06, + "loss": 0.6066, + "num_input_tokens_seen": 76244345, + "step": 2330 + }, + { + "epoch": 0.2102177932091807, + "flos": 21440138362560.0, + "grad_norm": 1.7694660368433162, + "learning_rate": 3.668939580954688e-06, + "loss": 0.7852, + "num_input_tokens_seen": 76272695, + "step": 2331 + }, + { + "epoch": 0.21030797673265095, + "flos": 27268382032320.0, + "grad_norm": 2.054318471049015, + "learning_rate": 3.668617578335506e-06, + "loss": 0.7504, + "num_input_tokens_seen": 76304785, + "step": 2332 + }, + { + "epoch": 0.2103981602561212, + "flos": 27997693056000.0, + "grad_norm": 1.7774038451378846, + "learning_rate": 3.6682954333406707e-06, + "loss": 0.8044, + "num_input_tokens_seen": 76335020, + "step": 2333 + }, + { + "epoch": 0.21048834377959147, + "flos": 16376820080160.0, + "grad_norm": 2.2559592484993813, + "learning_rate": 3.6679731459976707e-06, + "loss": 0.9117, + "num_input_tokens_seen": 76358245, + "step": 2334 + }, + { + "epoch": 0.21057852730306173, + "flos": 21295450722240.0, + "grad_norm": 1.5845074901072744, + "learning_rate": 3.6676507163340046e-06, + "loss": 0.7699, + "num_input_tokens_seen": 76387960, + "step": 2335 + }, + { + "epoch": 0.210668710826532, + "flos": 23626621812960.0, + "grad_norm": 2.604699154137236, + "learning_rate": 3.6673281443771842e-06, + "loss": 0.8086, + "num_input_tokens_seen": 76416620, + "step": 2336 + }, + { + "epoch": 0.21075889435000225, + "flos": 22350643464480.0, + "grad_norm": 1.768754733992533, + "learning_rate": 3.667005430154733e-06, + "loss": 0.6689, + "num_input_tokens_seen": 76446370, + "step": 2337 + }, + { + "epoch": 0.2108490778734725, + "flos": 70805700300480.0, + "grad_norm": 0.671515346343217, + "learning_rate": 3.666682573694186e-06, + "loss": 0.6803, + "num_input_tokens_seen": 76536445, + "step": 2338 + }, + { + "epoch": 0.21093926139694277, + "flos": 65665243298880.0, + "grad_norm": 0.655764650067036, + "learning_rate": 3.6663595750230924e-06, + "loss": 0.5501, + "num_input_tokens_seen": 76626735, + "step": 2339 + }, + { + "epoch": 0.21102944492041303, + "flos": 21579882424800.0, + "grad_norm": 2.142887340060406, + "learning_rate": 3.666036434169012e-06, + "loss": 0.8277, + "num_input_tokens_seen": 76654320, + "step": 2340 + }, + { + "epoch": 0.2111196284438833, + "flos": 22711693509600.0, + "grad_norm": 1.7062937105609424, + "learning_rate": 3.665713151159516e-06, + "loss": 0.7557, + "num_input_tokens_seen": 76684755, + "step": 2341 + }, + { + "epoch": 0.21120981196735356, + "flos": 26975401284960.0, + "grad_norm": 1.652642117158832, + "learning_rate": 3.665389726022189e-06, + "loss": 0.8762, + "num_input_tokens_seen": 76714570, + "step": 2342 + }, + { + "epoch": 0.21129999549082382, + "flos": 25885040216160.0, + "grad_norm": 1.9331555419048378, + "learning_rate": 3.6650661587846283e-06, + "loss": 0.7504, + "num_input_tokens_seen": 76743680, + "step": 2343 + }, + { + "epoch": 0.21139017901429408, + "flos": 22750169944800.0, + "grad_norm": 2.7539312699716714, + "learning_rate": 3.6647424494744418e-06, + "loss": 0.7399, + "num_input_tokens_seen": 76769165, + "step": 2344 + }, + { + "epoch": 0.21148036253776434, + "flos": 21582372798720.0, + "grad_norm": 1.6288165975731164, + "learning_rate": 3.6644185981192503e-06, + "loss": 0.8257, + "num_input_tokens_seen": 76797150, + "step": 2345 + }, + { + "epoch": 0.2115705460612346, + "flos": 23225162505120.0, + "grad_norm": 1.9160672646503814, + "learning_rate": 3.6640946047466868e-06, + "loss": 0.7061, + "num_input_tokens_seen": 76826720, + "step": 2346 + }, + { + "epoch": 0.2116607295847049, + "flos": 61545039465120.0, + "grad_norm": 1.9537707672675044, + "learning_rate": 3.6637704693843953e-06, + "loss": 0.7462, + "num_input_tokens_seen": 76860430, + "step": 2347 + }, + { + "epoch": 0.21175091310817515, + "flos": 24427347412800.0, + "grad_norm": 2.4920665747504183, + "learning_rate": 3.6634461920600337e-06, + "loss": 0.7092, + "num_input_tokens_seen": 76891070, + "step": 2348 + }, + { + "epoch": 0.2118410966316454, + "flos": 11020372371360.0, + "grad_norm": 2.340017576559193, + "learning_rate": 3.66312177280127e-06, + "loss": 0.8036, + "num_input_tokens_seen": 76918800, + "step": 2349 + }, + { + "epoch": 0.21193128015511567, + "flos": 20565544982400.0, + "grad_norm": 1.4124195396665253, + "learning_rate": 3.6627972116357872e-06, + "loss": 0.816, + "num_input_tokens_seen": 76947610, + "step": 2350 + }, + { + "epoch": 0.21202146367858593, + "flos": 20018580299520.0, + "grad_norm": 1.9526862910608547, + "learning_rate": 3.662472508591278e-06, + "loss": 0.8396, + "num_input_tokens_seen": 76976230, + "step": 2351 + }, + { + "epoch": 0.2121116472020562, + "flos": 29528473074720.0, + "grad_norm": 1.855546616376003, + "learning_rate": 3.662147663695447e-06, + "loss": 0.7642, + "num_input_tokens_seen": 77006810, + "step": 2352 + }, + { + "epoch": 0.21220183072552645, + "flos": 26139693207360.0, + "grad_norm": 3.0599834456650483, + "learning_rate": 3.6618226769760127e-06, + "loss": 0.7672, + "num_input_tokens_seen": 77038590, + "step": 2353 + }, + { + "epoch": 0.2122920142489967, + "flos": 29236718929440.0, + "grad_norm": 1.7868462707622372, + "learning_rate": 3.661497548460704e-06, + "loss": 0.655, + "num_input_tokens_seen": 77069240, + "step": 2354 + }, + { + "epoch": 0.21238219777246697, + "flos": 17796891352800.0, + "grad_norm": 2.2254147432375264, + "learning_rate": 3.6611722781772635e-06, + "loss": 0.7959, + "num_input_tokens_seen": 77095135, + "step": 2355 + }, + { + "epoch": 0.21247238129593723, + "flos": 63237041933760.0, + "grad_norm": 0.8331218252386228, + "learning_rate": 3.6608468661534444e-06, + "loss": 0.6447, + "num_input_tokens_seen": 77176585, + "step": 2356 + }, + { + "epoch": 0.2125625648194075, + "flos": 15647732075040.0, + "grad_norm": 2.727310908657183, + "learning_rate": 3.660521312417013e-06, + "loss": 0.699, + "num_input_tokens_seen": 77202880, + "step": 2357 + }, + { + "epoch": 0.21265274834287776, + "flos": 22311237785280.0, + "grad_norm": 2.037624392778547, + "learning_rate": 3.660195616995747e-06, + "loss": 0.7478, + "num_input_tokens_seen": 77231305, + "step": 2358 + }, + { + "epoch": 0.21274293186634802, + "flos": 22060115921280.0, + "grad_norm": 3.2179902300357393, + "learning_rate": 3.6598697799174367e-06, + "loss": 0.8021, + "num_input_tokens_seen": 77261945, + "step": 2359 + }, + { + "epoch": 0.21283311538981828, + "flos": 17723655458400.0, + "grad_norm": 2.09858895988676, + "learning_rate": 3.6595438012098844e-06, + "loss": 0.8353, + "num_input_tokens_seen": 77288890, + "step": 2360 + }, + { + "epoch": 0.21292329891328854, + "flos": 33820955977920.0, + "grad_norm": 2.8707516005738403, + "learning_rate": 3.6592176809009045e-06, + "loss": 0.748, + "num_input_tokens_seen": 77318215, + "step": 2361 + }, + { + "epoch": 0.2130134824367588, + "flos": 24973568700480.0, + "grad_norm": 1.8414621544542347, + "learning_rate": 3.6588914190183227e-06, + "loss": 0.8145, + "num_input_tokens_seen": 77347905, + "step": 2362 + }, + { + "epoch": 0.21310366596022906, + "flos": 18816544070880.0, + "grad_norm": 1.9732978608298195, + "learning_rate": 3.658565015589978e-06, + "loss": 0.649, + "num_input_tokens_seen": 77376140, + "step": 2363 + }, + { + "epoch": 0.21319384948369932, + "flos": 20456062838400.0, + "grad_norm": 4.224022403455335, + "learning_rate": 3.6582384706437217e-06, + "loss": 0.7855, + "num_input_tokens_seen": 77403985, + "step": 2364 + }, + { + "epoch": 0.21328403300716958, + "flos": 25047324971520.0, + "grad_norm": 2.0449896129913303, + "learning_rate": 3.6579117842074156e-06, + "loss": 0.7941, + "num_input_tokens_seen": 77433395, + "step": 2365 + }, + { + "epoch": 0.21337421653063984, + "flos": 24500211609600.0, + "grad_norm": 1.7374615397416135, + "learning_rate": 3.657584956308934e-06, + "loss": 0.7158, + "num_input_tokens_seen": 77463225, + "step": 2366 + }, + { + "epoch": 0.2134644000541101, + "flos": 16556118500640.0, + "grad_norm": 2.302977053317177, + "learning_rate": 3.6572579869761648e-06, + "loss": 0.8415, + "num_input_tokens_seen": 77486300, + "step": 2367 + }, + { + "epoch": 0.21355458357758036, + "flos": 16224661318080.0, + "grad_norm": 1.906222108730701, + "learning_rate": 3.6569308762370056e-06, + "loss": 0.8325, + "num_input_tokens_seen": 77511825, + "step": 2368 + }, + { + "epoch": 0.21364476710105063, + "flos": 25483506568800.0, + "grad_norm": 2.3671569822251572, + "learning_rate": 3.6566036241193676e-06, + "loss": 0.8181, + "num_input_tokens_seen": 77540430, + "step": 2369 + }, + { + "epoch": 0.2137349506245209, + "flos": 69132388487520.0, + "grad_norm": 0.6199000421440035, + "learning_rate": 3.656276230651174e-06, + "loss": 0.608, + "num_input_tokens_seen": 77636350, + "step": 2370 + }, + { + "epoch": 0.21382513414799117, + "flos": 56710120856160.0, + "grad_norm": 0.7058945892450317, + "learning_rate": 3.65594869586036e-06, + "loss": 0.6413, + "num_input_tokens_seen": 77722835, + "step": 2371 + }, + { + "epoch": 0.21391531767146144, + "flos": 29126456220480.0, + "grad_norm": 1.7753908935815252, + "learning_rate": 3.6556210197748724e-06, + "loss": 0.6958, + "num_input_tokens_seen": 77753315, + "step": 2372 + }, + { + "epoch": 0.2140055011949317, + "flos": 20783840214720.0, + "grad_norm": 1.8801137490050963, + "learning_rate": 3.655293202422671e-06, + "loss": 0.7421, + "num_input_tokens_seen": 77781725, + "step": 2373 + }, + { + "epoch": 0.21409568471840196, + "flos": 26831308360800.0, + "grad_norm": 1.5317880113510323, + "learning_rate": 3.654965243831725e-06, + "loss": 0.7881, + "num_input_tokens_seen": 77812390, + "step": 2374 + }, + { + "epoch": 0.21418586824187222, + "flos": 20892913491360.0, + "grad_norm": 1.614346279667113, + "learning_rate": 3.65463714403002e-06, + "loss": 0.8127, + "num_input_tokens_seen": 77840015, + "step": 2375 + }, + { + "epoch": 0.21427605176534248, + "flos": 29600110669440.0, + "grad_norm": 1.533232670396057, + "learning_rate": 3.65430890304555e-06, + "loss": 0.7309, + "num_input_tokens_seen": 77871790, + "step": 2376 + }, + { + "epoch": 0.21436623528881274, + "flos": 24828397853280.0, + "grad_norm": 1.4028052205631738, + "learning_rate": 3.653980520906323e-06, + "loss": 0.8036, + "num_input_tokens_seen": 77902070, + "step": 2377 + }, + { + "epoch": 0.214456418812283, + "flos": 28216657344000.0, + "grad_norm": 3.2379784896918373, + "learning_rate": 3.653651997640358e-06, + "loss": 0.7518, + "num_input_tokens_seen": 77934565, + "step": 2378 + }, + { + "epoch": 0.21454660233575326, + "flos": 23839230072000.0, + "grad_norm": 1.8421820099338717, + "learning_rate": 3.653323333275686e-06, + "loss": 0.7798, + "num_input_tokens_seen": 77961400, + "step": 2379 + }, + { + "epoch": 0.21463678585922352, + "flos": 18850114097760.0, + "grad_norm": 2.877121515778007, + "learning_rate": 3.652994527840351e-06, + "loss": 0.7581, + "num_input_tokens_seen": 77990095, + "step": 2380 + }, + { + "epoch": 0.21472696938269378, + "flos": 13497717892800.0, + "grad_norm": 2.207075108455259, + "learning_rate": 3.6526655813624087e-06, + "loss": 0.737, + "num_input_tokens_seen": 78015690, + "step": 2381 + }, + { + "epoch": 0.21481715290616404, + "flos": 67687630760640.0, + "grad_norm": 0.5812629390629989, + "learning_rate": 3.652336493869925e-06, + "loss": 0.601, + "num_input_tokens_seen": 78116660, + "step": 2382 + }, + { + "epoch": 0.2149073364296343, + "flos": 29819260806240.0, + "grad_norm": 1.9330349201337944, + "learning_rate": 3.6520072653909823e-06, + "loss": 0.7835, + "num_input_tokens_seen": 78145880, + "step": 2383 + }, + { + "epoch": 0.21499751995310457, + "flos": 20675027126400.0, + "grad_norm": 2.0981166341488127, + "learning_rate": 3.6516778959536702e-06, + "loss": 0.7717, + "num_input_tokens_seen": 78174475, + "step": 2384 + }, + { + "epoch": 0.21508770347657483, + "flos": 15752382150240.0, + "grad_norm": 2.111982528238565, + "learning_rate": 3.6513483855860923e-06, + "loss": 0.8229, + "num_input_tokens_seen": 78200470, + "step": 2385 + }, + { + "epoch": 0.2151778870000451, + "flos": 23151926610720.0, + "grad_norm": 2.204401282339566, + "learning_rate": 3.6510187343163654e-06, + "loss": 0.7182, + "num_input_tokens_seen": 78229180, + "step": 2386 + }, + { + "epoch": 0.21526807052351535, + "flos": 21038864903520.0, + "grad_norm": 1.74089358053716, + "learning_rate": 3.650688942172616e-06, + "loss": 0.8059, + "num_input_tokens_seen": 78254260, + "step": 2387 + }, + { + "epoch": 0.2153582540469856, + "flos": 20857187618400.0, + "grad_norm": 1.9737417727994104, + "learning_rate": 3.650359009182984e-06, + "loss": 0.7146, + "num_input_tokens_seen": 78281485, + "step": 2388 + }, + { + "epoch": 0.21544843757045587, + "flos": 21075371341440.0, + "grad_norm": 1.8487308272682355, + "learning_rate": 3.650028935375622e-06, + "loss": 0.8539, + "num_input_tokens_seen": 78307720, + "step": 2389 + }, + { + "epoch": 0.21553862109392613, + "flos": 28796634507360.0, + "grad_norm": 1.599975245999537, + "learning_rate": 3.6496987207786926e-06, + "loss": 0.7392, + "num_input_tokens_seen": 78340960, + "step": 2390 + }, + { + "epoch": 0.2156288046173964, + "flos": 20670232227360.0, + "grad_norm": 1.839344386145103, + "learning_rate": 3.6493683654203724e-06, + "loss": 0.7223, + "num_input_tokens_seen": 78368010, + "step": 2391 + }, + { + "epoch": 0.21571898814086665, + "flos": 24281507509920.0, + "grad_norm": 2.812950352534067, + "learning_rate": 3.6490378693288484e-06, + "loss": 0.7632, + "num_input_tokens_seen": 78395700, + "step": 2392 + }, + { + "epoch": 0.2158091716643369, + "flos": 21650925303360.0, + "grad_norm": 1.7222029939555692, + "learning_rate": 3.648707232532321e-06, + "loss": 0.8206, + "num_input_tokens_seen": 78421575, + "step": 2393 + }, + { + "epoch": 0.2158993551878072, + "flos": 23843132896800.0, + "grad_norm": 2.473676184911407, + "learning_rate": 3.6483764550590017e-06, + "loss": 0.7749, + "num_input_tokens_seen": 78448480, + "step": 2394 + }, + { + "epoch": 0.21598953871127746, + "flos": 20930126154720.0, + "grad_norm": 2.2316069629974233, + "learning_rate": 3.6480455369371133e-06, + "loss": 0.7922, + "num_input_tokens_seen": 78477190, + "step": 2395 + }, + { + "epoch": 0.21607972223474772, + "flos": 23223229677600.0, + "grad_norm": 2.56921669463296, + "learning_rate": 3.647714478194893e-06, + "loss": 0.7818, + "num_input_tokens_seen": 78506790, + "step": 2396 + }, + { + "epoch": 0.21616990575821798, + "flos": 20347063901280.0, + "grad_norm": 3.74350290135765, + "learning_rate": 3.647383278860588e-06, + "loss": 0.7199, + "num_input_tokens_seen": 78533690, + "step": 2397 + }, + { + "epoch": 0.21626008928168824, + "flos": 14518225515360.0, + "grad_norm": 1.6382459557569355, + "learning_rate": 3.6470519389624587e-06, + "loss": 0.7616, + "num_input_tokens_seen": 78559255, + "step": 2398 + }, + { + "epoch": 0.2163502728051585, + "flos": 27050086800000.0, + "grad_norm": 1.572866535861571, + "learning_rate": 3.646720458528776e-06, + "loss": 0.8354, + "num_input_tokens_seen": 78587180, + "step": 2399 + }, + { + "epoch": 0.21644045632862877, + "flos": 27733896303840.0, + "grad_norm": 2.469748765778506, + "learning_rate": 3.6463888375878235e-06, + "loss": 0.7452, + "num_input_tokens_seen": 78610015, + "step": 2400 + }, + { + "epoch": 0.21653063985209903, + "flos": 26977668640320.0, + "grad_norm": 2.1422640797222146, + "learning_rate": 3.646057076167897e-06, + "loss": 0.7897, + "num_input_tokens_seen": 78636875, + "step": 2401 + }, + { + "epoch": 0.2166208233755693, + "flos": 18671893600320.0, + "grad_norm": 1.7913153178197956, + "learning_rate": 3.645725174297305e-06, + "loss": 0.7796, + "num_input_tokens_seen": 78664360, + "step": 2402 + }, + { + "epoch": 0.21671100689903955, + "flos": 26868744042720.0, + "grad_norm": 2.67786570198451, + "learning_rate": 3.645393132004367e-06, + "loss": 0.7564, + "num_input_tokens_seen": 78693615, + "step": 2403 + }, + { + "epoch": 0.2168011904225098, + "flos": 37834285284480.0, + "grad_norm": 3.2333335489751476, + "learning_rate": 3.6450609493174135e-06, + "loss": 0.7316, + "num_input_tokens_seen": 78724450, + "step": 2404 + }, + { + "epoch": 0.21689137394598007, + "flos": 14881431406560.0, + "grad_norm": 1.8772424639597445, + "learning_rate": 3.6447286262647896e-06, + "loss": 0.7352, + "num_input_tokens_seen": 78751045, + "step": 2405 + }, + { + "epoch": 0.21698155746945033, + "flos": 14408446013280.0, + "grad_norm": 6.470612925722067, + "learning_rate": 3.64439616287485e-06, + "loss": 0.7405, + "num_input_tokens_seen": 78777625, + "step": 2406 + }, + { + "epoch": 0.2170717409929206, + "flos": 23804247594240.0, + "grad_norm": 3.362583916620294, + "learning_rate": 3.644063559175963e-06, + "loss": 0.7375, + "num_input_tokens_seen": 78805465, + "step": 2407 + }, + { + "epoch": 0.21716192451639085, + "flos": 28943924030880.0, + "grad_norm": 1.6534730195185028, + "learning_rate": 3.6437308151965074e-06, + "loss": 0.7544, + "num_input_tokens_seen": 78835680, + "step": 2408 + }, + { + "epoch": 0.2172521080398611, + "flos": 25627859681280.0, + "grad_norm": 1.6040120070904647, + "learning_rate": 3.643397930964876e-06, + "loss": 0.7871, + "num_input_tokens_seen": 78865160, + "step": 2409 + }, + { + "epoch": 0.21734229156333137, + "flos": 27924754519680.0, + "grad_norm": 2.6342342171660955, + "learning_rate": 3.6430649065094707e-06, + "loss": 0.8448, + "num_input_tokens_seen": 78893800, + "step": 2410 + }, + { + "epoch": 0.21743247508680164, + "flos": 18664868515680.0, + "grad_norm": 2.053308756833982, + "learning_rate": 3.6427317418587086e-06, + "loss": 0.8368, + "num_input_tokens_seen": 78914970, + "step": 2411 + }, + { + "epoch": 0.2175226586102719, + "flos": 22897682486880.0, + "grad_norm": 1.774055810095608, + "learning_rate": 3.6423984370410157e-06, + "loss": 0.7136, + "num_input_tokens_seen": 78945725, + "step": 2412 + }, + { + "epoch": 0.21761284213374216, + "flos": 24826836723360.0, + "grad_norm": 1.7660675169457738, + "learning_rate": 3.6420649920848324e-06, + "loss": 0.7394, + "num_input_tokens_seen": 78976095, + "step": 2413 + }, + { + "epoch": 0.21770302565721242, + "flos": 24206413127520.0, + "grad_norm": 2.04166629038697, + "learning_rate": 3.6417314070186096e-06, + "loss": 0.7822, + "num_input_tokens_seen": 79006305, + "step": 2414 + }, + { + "epoch": 0.21779320918068268, + "flos": 28326027978720.0, + "grad_norm": 1.5872405763851958, + "learning_rate": 3.641397681870811e-06, + "loss": 0.7816, + "num_input_tokens_seen": 79036300, + "step": 2415 + }, + { + "epoch": 0.21788339270415294, + "flos": 24901187710560.0, + "grad_norm": 1.6695029999373845, + "learning_rate": 3.641063816669911e-06, + "loss": 0.7938, + "num_input_tokens_seen": 79065165, + "step": 2416 + }, + { + "epoch": 0.2179735762276232, + "flos": 25265322845760.0, + "grad_norm": 2.3449458822419142, + "learning_rate": 3.640729811444398e-06, + "loss": 0.8338, + "num_input_tokens_seen": 79094330, + "step": 2417 + }, + { + "epoch": 0.2180637597510935, + "flos": 36886270161120.0, + "grad_norm": 1.7167260477604114, + "learning_rate": 3.6403956662227706e-06, + "loss": 0.7775, + "num_input_tokens_seen": 79126955, + "step": 2418 + }, + { + "epoch": 0.21815394327456375, + "flos": 22241607357600.0, + "grad_norm": 1.878091234865245, + "learning_rate": 3.6400613810335396e-06, + "loss": 0.8271, + "num_input_tokens_seen": 79154835, + "step": 2419 + }, + { + "epoch": 0.218244126798034, + "flos": 32551407997920.0, + "grad_norm": 1.742226980814036, + "learning_rate": 3.639726955905228e-06, + "loss": 0.6903, + "num_input_tokens_seen": 79183780, + "step": 2420 + }, + { + "epoch": 0.21833431032150427, + "flos": 27415002500160.0, + "grad_norm": 1.6115016774433057, + "learning_rate": 3.639392390866372e-06, + "loss": 0.7975, + "num_input_tokens_seen": 79212245, + "step": 2421 + }, + { + "epoch": 0.21842449384497453, + "flos": 23226463446720.0, + "grad_norm": 1.7110463286325144, + "learning_rate": 3.639057685945517e-06, + "loss": 0.7511, + "num_input_tokens_seen": 79241625, + "step": 2422 + }, + { + "epoch": 0.2185146773684448, + "flos": 41291580486720.0, + "grad_norm": 1.4936325505900878, + "learning_rate": 3.638722841171223e-06, + "loss": 0.662, + "num_input_tokens_seen": 79277685, + "step": 2423 + }, + { + "epoch": 0.21860486089191505, + "flos": 23516544952800.0, + "grad_norm": 1.8551080292865658, + "learning_rate": 3.638387856572061e-06, + "loss": 0.813, + "num_input_tokens_seen": 79307340, + "step": 2424 + }, + { + "epoch": 0.21869504441538531, + "flos": 67299149432640.0, + "grad_norm": 0.6564619539484459, + "learning_rate": 3.638052732176612e-06, + "loss": 0.6071, + "num_input_tokens_seen": 79393245, + "step": 2425 + }, + { + "epoch": 0.21878522793885558, + "flos": 21327013582080.0, + "grad_norm": 1.6482050582658148, + "learning_rate": 3.637717468013472e-06, + "loss": 0.7861, + "num_input_tokens_seen": 79420230, + "step": 2426 + }, + { + "epoch": 0.21887541146232584, + "flos": 65921345910720.0, + "grad_norm": 0.5957770379910771, + "learning_rate": 3.6373820641112475e-06, + "loss": 0.559, + "num_input_tokens_seen": 79512450, + "step": 2427 + }, + { + "epoch": 0.2189655949857961, + "flos": 20128768668960.0, + "grad_norm": 1.6207266886063074, + "learning_rate": 3.6370465204985567e-06, + "loss": 0.7903, + "num_input_tokens_seen": 79537380, + "step": 2428 + }, + { + "epoch": 0.21905577850926636, + "flos": 69205698721440.0, + "grad_norm": 0.6325084520989006, + "learning_rate": 3.6367108372040304e-06, + "loss": 0.6199, + "num_input_tokens_seen": 79632080, + "step": 2429 + }, + { + "epoch": 0.21914596203273662, + "flos": 23480187193920.0, + "grad_norm": 1.9115037904023775, + "learning_rate": 3.6363750142563107e-06, + "loss": 0.7766, + "num_input_tokens_seen": 79658990, + "step": 2430 + }, + { + "epoch": 0.21923614555620688, + "flos": 64481643271200.0, + "grad_norm": 0.7198329727826721, + "learning_rate": 3.636039051684052e-06, + "loss": 0.6562, + "num_input_tokens_seen": 79760370, + "step": 2431 + }, + { + "epoch": 0.21932632907967714, + "flos": 13351729310880.0, + "grad_norm": 2.0148126902262598, + "learning_rate": 3.6357029495159203e-06, + "loss": 0.7454, + "num_input_tokens_seen": 79788120, + "step": 2432 + }, + { + "epoch": 0.2194165126031474, + "flos": 23989902043680.0, + "grad_norm": 1.6008072723909783, + "learning_rate": 3.6353667077805934e-06, + "loss": 0.7497, + "num_input_tokens_seen": 79816810, + "step": 2433 + }, + { + "epoch": 0.21950669612661766, + "flos": 21440324211360.0, + "grad_norm": 1.8600472268907686, + "learning_rate": 3.6350303265067625e-06, + "loss": 0.8106, + "num_input_tokens_seen": 79845335, + "step": 2434 + }, + { + "epoch": 0.21959687965008792, + "flos": 27124809484800.0, + "grad_norm": 2.0010084947427673, + "learning_rate": 3.6346938057231285e-06, + "loss": 0.8042, + "num_input_tokens_seen": 79877745, + "step": 2435 + }, + { + "epoch": 0.21968706317355818, + "flos": 25265917561920.0, + "grad_norm": 1.7718946829212232, + "learning_rate": 3.6343571454584047e-06, + "loss": 0.7202, + "num_input_tokens_seen": 79909720, + "step": 2436 + }, + { + "epoch": 0.21977724669702844, + "flos": 21876877506240.0, + "grad_norm": 2.9798629765082567, + "learning_rate": 3.6340203457413176e-06, + "loss": 0.8144, + "num_input_tokens_seen": 79935930, + "step": 2437 + }, + { + "epoch": 0.2198674302204987, + "flos": 14809236265440.0, + "grad_norm": 2.1899619281194247, + "learning_rate": 3.633683406600605e-06, + "loss": 0.7994, + "num_input_tokens_seen": 79961820, + "step": 2438 + }, + { + "epoch": 0.21995761374396897, + "flos": 19722328613280.0, + "grad_norm": 1.4825562427524273, + "learning_rate": 3.6333463280650165e-06, + "loss": 0.8384, + "num_input_tokens_seen": 79989320, + "step": 2439 + }, + { + "epoch": 0.22004779726743923, + "flos": 56815551496320.0, + "grad_norm": 0.703788867778278, + "learning_rate": 3.6330091101633126e-06, + "loss": 0.5694, + "num_input_tokens_seen": 80077595, + "step": 2440 + }, + { + "epoch": 0.2201379807909095, + "flos": 24646237361280.0, + "grad_norm": 2.6385863069316486, + "learning_rate": 3.632671752924267e-06, + "loss": 0.8158, + "num_input_tokens_seen": 80109225, + "step": 2441 + }, + { + "epoch": 0.22022816431437978, + "flos": 15605464324320.0, + "grad_norm": 2.2637815144460616, + "learning_rate": 3.632334256376665e-06, + "loss": 0.7839, + "num_input_tokens_seen": 80134485, + "step": 2442 + }, + { + "epoch": 0.22031834783785004, + "flos": 22897496638080.0, + "grad_norm": 1.8108054164742267, + "learning_rate": 3.6319966205493044e-06, + "loss": 0.8327, + "num_input_tokens_seen": 80163165, + "step": 2443 + }, + { + "epoch": 0.2204085313613203, + "flos": 56010737222880.0, + "grad_norm": 0.6925428498779028, + "learning_rate": 3.6316588454709922e-06, + "loss": 0.6527, + "num_input_tokens_seen": 80253000, + "step": 2444 + }, + { + "epoch": 0.22049871488479056, + "flos": 22095841794240.0, + "grad_norm": 1.8431924526513737, + "learning_rate": 3.6313209311705514e-06, + "loss": 0.7287, + "num_input_tokens_seen": 80280705, + "step": 2445 + }, + { + "epoch": 0.22058889840826082, + "flos": 16412099916000.0, + "grad_norm": 1.945476050185114, + "learning_rate": 3.6309828776768133e-06, + "loss": 0.8198, + "num_input_tokens_seen": 80306955, + "step": 2446 + }, + { + "epoch": 0.22067908193173108, + "flos": 28106468974560.0, + "grad_norm": 2.8200545681753346, + "learning_rate": 3.630644685018623e-06, + "loss": 0.7414, + "num_input_tokens_seen": 80336840, + "step": 2447 + }, + { + "epoch": 0.22076926545520134, + "flos": 22241124150720.0, + "grad_norm": 2.1085285890453402, + "learning_rate": 3.6303063532248367e-06, + "loss": 0.8565, + "num_input_tokens_seen": 80363195, + "step": 2448 + }, + { + "epoch": 0.2208594489786716, + "flos": 18668325303360.0, + "grad_norm": 2.0002405498802522, + "learning_rate": 3.6299678823243236e-06, + "loss": 0.8599, + "num_input_tokens_seen": 80390515, + "step": 2449 + }, + { + "epoch": 0.22094963250214186, + "flos": 22746564478080.0, + "grad_norm": 2.0485134410384607, + "learning_rate": 3.629629272345963e-06, + "loss": 0.6814, + "num_input_tokens_seen": 80418680, + "step": 2450 + }, + { + "epoch": 0.22103981602561212, + "flos": 20345391262080.0, + "grad_norm": 3.1976136965964375, + "learning_rate": 3.6292905233186468e-06, + "loss": 0.8991, + "num_input_tokens_seen": 80446210, + "step": 2451 + }, + { + "epoch": 0.22112999954908238, + "flos": 20923807295520.0, + "grad_norm": 2.1408638925278147, + "learning_rate": 3.6289516352712796e-06, + "loss": 0.739, + "num_input_tokens_seen": 80474460, + "step": 2452 + }, + { + "epoch": 0.22122018307255265, + "flos": 64454360667360.0, + "grad_norm": 0.7096299944095932, + "learning_rate": 3.6286126082327764e-06, + "loss": 0.6028, + "num_input_tokens_seen": 80560095, + "step": 2453 + }, + { + "epoch": 0.2213103665960229, + "flos": 19217854699680.0, + "grad_norm": 1.6084909325635504, + "learning_rate": 3.628273442232066e-06, + "loss": 0.8089, + "num_input_tokens_seen": 80588150, + "step": 2454 + }, + { + "epoch": 0.22140055011949317, + "flos": 27082913431680.0, + "grad_norm": 1.3389055620811174, + "learning_rate": 3.627934137298087e-06, + "loss": 0.7693, + "num_input_tokens_seen": 80618015, + "step": 2455 + }, + { + "epoch": 0.22149073364296343, + "flos": 24827840306880.0, + "grad_norm": 2.0823409832420583, + "learning_rate": 3.627594693459792e-06, + "loss": 0.7533, + "num_input_tokens_seen": 80646110, + "step": 2456 + }, + { + "epoch": 0.2215809171664337, + "flos": 17942694085920.0, + "grad_norm": 1.7908302469110082, + "learning_rate": 3.6272551107461424e-06, + "loss": 0.7366, + "num_input_tokens_seen": 80673570, + "step": 2457 + }, + { + "epoch": 0.22167110068990395, + "flos": 36523250118720.0, + "grad_norm": 1.7279932644456704, + "learning_rate": 3.6269153891861137e-06, + "loss": 0.7676, + "num_input_tokens_seen": 80707080, + "step": 2458 + }, + { + "epoch": 0.2217612842133742, + "flos": 34482346382880.0, + "grad_norm": 1.4557968515164237, + "learning_rate": 3.6265755288086944e-06, + "loss": 0.7473, + "num_input_tokens_seen": 80742820, + "step": 2459 + }, + { + "epoch": 0.22185146773684447, + "flos": 22168334293440.0, + "grad_norm": 2.4998884176148546, + "learning_rate": 3.626235529642881e-06, + "loss": 0.7548, + "num_input_tokens_seen": 80769715, + "step": 2460 + }, + { + "epoch": 0.22194165126031473, + "flos": 26103521297280.0, + "grad_norm": 1.8533692377486752, + "learning_rate": 3.625895391717686e-06, + "loss": 0.7117, + "num_input_tokens_seen": 80795405, + "step": 2461 + }, + { + "epoch": 0.222031834783785, + "flos": 43771899588960.0, + "grad_norm": 1.990086184029964, + "learning_rate": 3.625555115062131e-06, + "loss": 0.7457, + "num_input_tokens_seen": 80829900, + "step": 2462 + }, + { + "epoch": 0.22212201830725525, + "flos": 17099180358720.0, + "grad_norm": 7.5252190339410765, + "learning_rate": 3.6252146997052507e-06, + "loss": 0.8425, + "num_input_tokens_seen": 80856295, + "step": 2463 + }, + { + "epoch": 0.22221220183072551, + "flos": 70601752595520.0, + "grad_norm": 0.6672912865346953, + "learning_rate": 3.6248741456760898e-06, + "loss": 0.5596, + "num_input_tokens_seen": 80956960, + "step": 2464 + }, + { + "epoch": 0.22230238535419578, + "flos": 25663213856640.0, + "grad_norm": 2.326609585543673, + "learning_rate": 3.624533453003708e-06, + "loss": 0.7615, + "num_input_tokens_seen": 80988420, + "step": 2465 + }, + { + "epoch": 0.22239256887766606, + "flos": 18597877140960.0, + "grad_norm": 1.99063977725594, + "learning_rate": 3.6241926217171745e-06, + "loss": 0.7832, + "num_input_tokens_seen": 81013995, + "step": 2466 + }, + { + "epoch": 0.22248275240113632, + "flos": 30617830560000.0, + "grad_norm": 1.7763457126588134, + "learning_rate": 3.6238516518455703e-06, + "loss": 0.6948, + "num_input_tokens_seen": 81046330, + "step": 2467 + }, + { + "epoch": 0.22257293592460659, + "flos": 17249815160640.0, + "grad_norm": 3.730522179345784, + "learning_rate": 3.62351054341799e-06, + "loss": 0.899, + "num_input_tokens_seen": 81071930, + "step": 2468 + }, + { + "epoch": 0.22266311944807685, + "flos": 22749575228640.0, + "grad_norm": 1.919722553512512, + "learning_rate": 3.623169296463538e-06, + "loss": 0.8039, + "num_input_tokens_seen": 81100245, + "step": 2469 + }, + { + "epoch": 0.2227533029715471, + "flos": 22242350752800.0, + "grad_norm": 1.719662580106794, + "learning_rate": 3.6228279110113316e-06, + "loss": 0.7764, + "num_input_tokens_seen": 81128360, + "step": 2470 + }, + { + "epoch": 0.22284348649501737, + "flos": 18524678416320.0, + "grad_norm": 2.103667499014788, + "learning_rate": 3.6224863870904994e-06, + "loss": 0.7861, + "num_input_tokens_seen": 81155815, + "step": 2471 + }, + { + "epoch": 0.22293367001848763, + "flos": 25739274652800.0, + "grad_norm": 1.7532456137225478, + "learning_rate": 3.6221447247301827e-06, + "loss": 0.7922, + "num_input_tokens_seen": 81184585, + "step": 2472 + }, + { + "epoch": 0.2230238535419579, + "flos": 28872137757120.0, + "grad_norm": 2.270754354894205, + "learning_rate": 3.6218029239595332e-06, + "loss": 0.7275, + "num_input_tokens_seen": 81213565, + "step": 2473 + }, + { + "epoch": 0.22311403706542815, + "flos": 27300353759520.0, + "grad_norm": 2.128986098165712, + "learning_rate": 3.621460984807716e-06, + "loss": 0.8071, + "num_input_tokens_seen": 81241370, + "step": 2474 + }, + { + "epoch": 0.2232042205888984, + "flos": 20165237937120.0, + "grad_norm": 2.876579016907955, + "learning_rate": 3.621118907303907e-06, + "loss": 0.748, + "num_input_tokens_seen": 81270715, + "step": 2475 + }, + { + "epoch": 0.22329440411236867, + "flos": 18270768820320.0, + "grad_norm": 1.919115644277233, + "learning_rate": 3.620776691477294e-06, + "loss": 0.7721, + "num_input_tokens_seen": 81299060, + "step": 2476 + }, + { + "epoch": 0.22338458763583893, + "flos": 54218948184000.0, + "grad_norm": 0.6815797182870237, + "learning_rate": 3.6204343373570765e-06, + "loss": 0.5507, + "num_input_tokens_seen": 81382890, + "step": 2477 + }, + { + "epoch": 0.2234747711593092, + "flos": 31677966880320.0, + "grad_norm": 2.0686287420871614, + "learning_rate": 3.620091844972467e-06, + "loss": 0.8015, + "num_input_tokens_seen": 81411185, + "step": 2478 + }, + { + "epoch": 0.22356495468277945, + "flos": 19909878720480.0, + "grad_norm": 2.434852140038006, + "learning_rate": 3.619749214352688e-06, + "loss": 0.7985, + "num_input_tokens_seen": 81437975, + "step": 2479 + }, + { + "epoch": 0.22365513820624972, + "flos": 17833026093120.0, + "grad_norm": 2.1723880299114247, + "learning_rate": 3.6194064455269744e-06, + "loss": 0.8817, + "num_input_tokens_seen": 81464290, + "step": 2480 + }, + { + "epoch": 0.22374532172971998, + "flos": 22605779662560.0, + "grad_norm": 2.2487244699247135, + "learning_rate": 3.6190635385245737e-06, + "loss": 0.6995, + "num_input_tokens_seen": 81493750, + "step": 2481 + }, + { + "epoch": 0.22383550525319024, + "flos": 21475789896000.0, + "grad_norm": 1.9633947531651885, + "learning_rate": 3.618720493374745e-06, + "loss": 0.7109, + "num_input_tokens_seen": 81523705, + "step": 2482 + }, + { + "epoch": 0.2239256887766605, + "flos": 27009268669920.0, + "grad_norm": 1.9854315400608893, + "learning_rate": 3.6183773101067575e-06, + "loss": 0.6378, + "num_input_tokens_seen": 81552960, + "step": 2483 + }, + { + "epoch": 0.22401587230013076, + "flos": 28981619901120.0, + "grad_norm": 1.7373916921545853, + "learning_rate": 3.6180339887498948e-06, + "loss": 0.6992, + "num_input_tokens_seen": 81584510, + "step": 2484 + }, + { + "epoch": 0.22410605582360102, + "flos": 20995110362400.0, + "grad_norm": 2.0860279077947133, + "learning_rate": 3.61769052933345e-06, + "loss": 0.781, + "num_input_tokens_seen": 81610535, + "step": 2485 + }, + { + "epoch": 0.22419623934707128, + "flos": 12185381785440.0, + "grad_norm": 2.074460462721636, + "learning_rate": 3.6173469318867297e-06, + "loss": 0.9323, + "num_input_tokens_seen": 81632275, + "step": 2486 + }, + { + "epoch": 0.22428642287054154, + "flos": 21877509392160.0, + "grad_norm": 2.193456224730357, + "learning_rate": 3.617003196439051e-06, + "loss": 0.7959, + "num_input_tokens_seen": 81660160, + "step": 2487 + }, + { + "epoch": 0.2243766063940118, + "flos": 28690014434880.0, + "grad_norm": 1.975023597866032, + "learning_rate": 3.616659323019744e-06, + "loss": 0.7902, + "num_input_tokens_seen": 81692085, + "step": 2488 + }, + { + "epoch": 0.22446678991748206, + "flos": 23329180694400.0, + "grad_norm": 2.1804352164893332, + "learning_rate": 3.616315311658149e-06, + "loss": 0.8062, + "num_input_tokens_seen": 81720185, + "step": 2489 + }, + { + "epoch": 0.22455697344095235, + "flos": 21695125881600.0, + "grad_norm": 1.9115771292238457, + "learning_rate": 3.6159711623836195e-06, + "loss": 0.8329, + "num_input_tokens_seen": 81750440, + "step": 2490 + }, + { + "epoch": 0.2246471569644226, + "flos": 71032804765920.0, + "grad_norm": 0.6475962961974562, + "learning_rate": 3.6156268752255203e-06, + "loss": 0.5644, + "num_input_tokens_seen": 81848550, + "step": 2491 + }, + { + "epoch": 0.22473734048789287, + "flos": 29197982305920.0, + "grad_norm": 1.9271251020973992, + "learning_rate": 3.615282450213227e-06, + "loss": 0.7418, + "num_input_tokens_seen": 81878200, + "step": 2492 + }, + { + "epoch": 0.22482752401136313, + "flos": 20966186555520.0, + "grad_norm": 1.6310152082172733, + "learning_rate": 3.614937887376128e-06, + "loss": 0.792, + "num_input_tokens_seen": 81906495, + "step": 2493 + }, + { + "epoch": 0.2249177075348334, + "flos": 26248952332800.0, + "grad_norm": 1.9692872540518174, + "learning_rate": 3.614593186743625e-06, + "loss": 0.793, + "num_input_tokens_seen": 81933830, + "step": 2494 + }, + { + "epoch": 0.22500789105830366, + "flos": 19217222813760.0, + "grad_norm": 2.37882657223016, + "learning_rate": 3.614248348345128e-06, + "loss": 0.7113, + "num_input_tokens_seen": 81959565, + "step": 2495 + }, + { + "epoch": 0.22509807458177392, + "flos": 25991660288640.0, + "grad_norm": 2.261113166443653, + "learning_rate": 3.6139033722100614e-06, + "loss": 0.8277, + "num_input_tokens_seen": 81988275, + "step": 2496 + }, + { + "epoch": 0.22518825810524418, + "flos": 18634049051040.0, + "grad_norm": 2.766767535228455, + "learning_rate": 3.6135582583678596e-06, + "loss": 0.8558, + "num_input_tokens_seen": 82013115, + "step": 2497 + }, + { + "epoch": 0.22527844162871444, + "flos": 12950641700640.0, + "grad_norm": 2.0793810038914184, + "learning_rate": 3.61321300684797e-06, + "loss": 0.796, + "num_input_tokens_seen": 82039355, + "step": 2498 + }, + { + "epoch": 0.2253686251521847, + "flos": 25193164874400.0, + "grad_norm": 1.477706466036139, + "learning_rate": 3.6128676176798527e-06, + "loss": 0.8023, + "num_input_tokens_seen": 82071555, + "step": 2499 + }, + { + "epoch": 0.22545880867565496, + "flos": 27813265208640.0, + "grad_norm": 1.8855910317178148, + "learning_rate": 3.612522090892976e-06, + "loss": 0.7969, + "num_input_tokens_seen": 82102790, + "step": 2500 + }, + { + "epoch": 0.22554899219912522, + "flos": 47889024066240.0, + "grad_norm": 1.4652166680262975, + "learning_rate": 3.6121764265168232e-06, + "loss": 0.7459, + "num_input_tokens_seen": 82135155, + "step": 2501 + }, + { + "epoch": 0.22563917572259548, + "flos": 22970063476800.0, + "grad_norm": 1.534517499906577, + "learning_rate": 3.611830624580888e-06, + "loss": 0.8042, + "num_input_tokens_seen": 82164565, + "step": 2502 + }, + { + "epoch": 0.22572935924606574, + "flos": 17286321598560.0, + "grad_norm": 1.7685796271450405, + "learning_rate": 3.6114846851146767e-06, + "loss": 0.7911, + "num_input_tokens_seen": 82191345, + "step": 2503 + }, + { + "epoch": 0.225819542769536, + "flos": 37359404233440.0, + "grad_norm": 1.9313067233449615, + "learning_rate": 3.6111386081477068e-06, + "loss": 0.6855, + "num_input_tokens_seen": 82222865, + "step": 2504 + }, + { + "epoch": 0.22590972629300626, + "flos": 30981519658080.0, + "grad_norm": 1.805663336283825, + "learning_rate": 3.6107923937095066e-06, + "loss": 0.7987, + "num_input_tokens_seen": 82253950, + "step": 2505 + }, + { + "epoch": 0.22599990981647652, + "flos": 19107963688320.0, + "grad_norm": 2.422938845383436, + "learning_rate": 3.6104460418296173e-06, + "loss": 0.7003, + "num_input_tokens_seen": 82280645, + "step": 2506 + }, + { + "epoch": 0.22609009333994678, + "flos": 24244517865120.0, + "grad_norm": 1.8906985812410215, + "learning_rate": 3.6100995525375924e-06, + "loss": 0.8293, + "num_input_tokens_seen": 82306575, + "step": 2507 + }, + { + "epoch": 0.22618027686341705, + "flos": 30435632898240.0, + "grad_norm": 1.5632449091052578, + "learning_rate": 3.6097529258629952e-06, + "loss": 0.7243, + "num_input_tokens_seen": 82334005, + "step": 2508 + }, + { + "epoch": 0.2262704603868873, + "flos": 21695794937280.0, + "grad_norm": 2.341398050868558, + "learning_rate": 3.6094061618354027e-06, + "loss": 0.8072, + "num_input_tokens_seen": 82362865, + "step": 2509 + }, + { + "epoch": 0.22636064391035757, + "flos": 22855080208320.0, + "grad_norm": 1.884887122156681, + "learning_rate": 3.609059260484402e-06, + "loss": 0.7771, + "num_input_tokens_seen": 82391875, + "step": 2510 + }, + { + "epoch": 0.22645082743382783, + "flos": 21402107964480.0, + "grad_norm": 3.5680425260002426, + "learning_rate": 3.6087122218395935e-06, + "loss": 0.72, + "num_input_tokens_seen": 82414240, + "step": 2511 + }, + { + "epoch": 0.2265410109572981, + "flos": 23221111001280.0, + "grad_norm": 1.5955440670755097, + "learning_rate": 3.608365045930587e-06, + "loss": 0.7479, + "num_input_tokens_seen": 82442350, + "step": 2512 + }, + { + "epoch": 0.22663119448076835, + "flos": 22600538726400.0, + "grad_norm": 1.9895780340260605, + "learning_rate": 3.608017732787007e-06, + "loss": 0.7686, + "num_input_tokens_seen": 82469720, + "step": 2513 + }, + { + "epoch": 0.22672137800423864, + "flos": 17828454212640.0, + "grad_norm": 1.6583964995553835, + "learning_rate": 3.6076702824384875e-06, + "loss": 0.7118, + "num_input_tokens_seen": 82496260, + "step": 2514 + }, + { + "epoch": 0.2268115615277089, + "flos": 31131002197440.0, + "grad_norm": 1.7265841797723296, + "learning_rate": 3.607322694914675e-06, + "loss": 0.7965, + "num_input_tokens_seen": 82525600, + "step": 2515 + }, + { + "epoch": 0.22690174505117916, + "flos": 19509311486880.0, + "grad_norm": 1.9839922318037295, + "learning_rate": 3.606974970245227e-06, + "loss": 0.7509, + "num_input_tokens_seen": 82554965, + "step": 2516 + }, + { + "epoch": 0.22699192857464942, + "flos": 31167694484160.0, + "grad_norm": 1.8153294150335524, + "learning_rate": 3.606627108459814e-06, + "loss": 0.732, + "num_input_tokens_seen": 82584915, + "step": 2517 + }, + { + "epoch": 0.22708211209811968, + "flos": 20637925972320.0, + "grad_norm": 1.716801388401526, + "learning_rate": 3.6062791095881174e-06, + "loss": 0.7059, + "num_input_tokens_seen": 82613635, + "step": 2518 + }, + { + "epoch": 0.22717229562158994, + "flos": 26504200040160.0, + "grad_norm": 2.0548021407835244, + "learning_rate": 3.6059309736598303e-06, + "loss": 0.8519, + "num_input_tokens_seen": 82641115, + "step": 2519 + }, + { + "epoch": 0.2272624791450602, + "flos": 13605378718560.0, + "grad_norm": 2.1454701409485764, + "learning_rate": 3.605582700704657e-06, + "loss": 0.7587, + "num_input_tokens_seen": 82666580, + "step": 2520 + }, + { + "epoch": 0.22735266266853046, + "flos": 18887215251840.0, + "grad_norm": 1.7133586596436625, + "learning_rate": 3.6052342907523146e-06, + "loss": 0.7832, + "num_input_tokens_seen": 82695210, + "step": 2521 + }, + { + "epoch": 0.22744284619200072, + "flos": 22781398276800.0, + "grad_norm": 2.1812079264345856, + "learning_rate": 3.604885743832532e-06, + "loss": 0.708, + "num_input_tokens_seen": 82721340, + "step": 2522 + }, + { + "epoch": 0.22753302971547099, + "flos": 57713530389120.0, + "grad_norm": 0.6929387495735485, + "learning_rate": 3.6045370599750482e-06, + "loss": 0.58, + "num_input_tokens_seen": 82806030, + "step": 2523 + }, + { + "epoch": 0.22762321323894125, + "flos": 21002655823680.0, + "grad_norm": 1.732044682087174, + "learning_rate": 3.604188239209615e-06, + "loss": 0.7577, + "num_input_tokens_seen": 82833900, + "step": 2524 + }, + { + "epoch": 0.2277133967624115, + "flos": 49711112193120.0, + "grad_norm": 2.451650869066631, + "learning_rate": 3.603839281565996e-06, + "loss": 0.7009, + "num_input_tokens_seen": 82867690, + "step": 2525 + }, + { + "epoch": 0.22780358028588177, + "flos": 16702924817280.0, + "grad_norm": 2.1629469437214035, + "learning_rate": 3.603490187073966e-06, + "loss": 0.7964, + "num_input_tokens_seen": 82894820, + "step": 2526 + }, + { + "epoch": 0.22789376380935203, + "flos": 26030731440000.0, + "grad_norm": 2.509131159164295, + "learning_rate": 3.6031409557633117e-06, + "loss": 0.8366, + "num_input_tokens_seen": 82924450, + "step": 2527 + }, + { + "epoch": 0.2279839473328223, + "flos": 20710455641280.0, + "grad_norm": 2.892045945094088, + "learning_rate": 3.602791587663831e-06, + "loss": 0.818, + "num_input_tokens_seen": 82952500, + "step": 2528 + }, + { + "epoch": 0.22807413085629255, + "flos": 21183255185760.0, + "grad_norm": 1.6190343069549724, + "learning_rate": 3.6024420828053348e-06, + "loss": 0.8164, + "num_input_tokens_seen": 82981570, + "step": 2529 + }, + { + "epoch": 0.2281643143797628, + "flos": 16267635294240.0, + "grad_norm": 2.238565423578156, + "learning_rate": 3.6020924412176445e-06, + "loss": 0.849, + "num_input_tokens_seen": 83009265, + "step": 2530 + }, + { + "epoch": 0.22825449790323307, + "flos": 25993407267360.0, + "grad_norm": 2.922647792189045, + "learning_rate": 3.601742662930593e-06, + "loss": 0.7485, + "num_input_tokens_seen": 83040210, + "step": 2531 + }, + { + "epoch": 0.22834468142670333, + "flos": 21841448991360.0, + "grad_norm": 1.7352659669115944, + "learning_rate": 3.6013927479740248e-06, + "loss": 0.7429, + "num_input_tokens_seen": 83069035, + "step": 2532 + }, + { + "epoch": 0.2284348649501736, + "flos": 16886237571840.0, + "grad_norm": 1.9989558713048656, + "learning_rate": 3.6010426963777985e-06, + "loss": 0.8573, + "num_input_tokens_seen": 83095710, + "step": 2533 + }, + { + "epoch": 0.22852504847364385, + "flos": 17793062867520.0, + "grad_norm": 1.9783472740798393, + "learning_rate": 3.6006925081717804e-06, + "loss": 0.8237, + "num_input_tokens_seen": 83121875, + "step": 2534 + }, + { + "epoch": 0.22861523199711412, + "flos": 26976999584640.0, + "grad_norm": 1.8012249249017405, + "learning_rate": 3.600342183385852e-06, + "loss": 0.7485, + "num_input_tokens_seen": 83152885, + "step": 2535 + }, + { + "epoch": 0.22870541552058438, + "flos": 20857076109120.0, + "grad_norm": 1.9397228779431206, + "learning_rate": 3.5999917220499043e-06, + "loss": 0.7175, + "num_input_tokens_seen": 83180340, + "step": 2536 + }, + { + "epoch": 0.22879559904405466, + "flos": 23042370127200.0, + "grad_norm": 1.9560364652366713, + "learning_rate": 3.5996411241938404e-06, + "loss": 0.7816, + "num_input_tokens_seen": 83207955, + "step": 2537 + }, + { + "epoch": 0.22888578256752493, + "flos": 21401996455200.0, + "grad_norm": 1.7629147548871038, + "learning_rate": 3.5992903898475752e-06, + "loss": 0.7343, + "num_input_tokens_seen": 83237670, + "step": 2538 + }, + { + "epoch": 0.2289759660909952, + "flos": 24169683671040.0, + "grad_norm": 1.502132568751711, + "learning_rate": 3.5989395190410365e-06, + "loss": 0.8744, + "num_input_tokens_seen": 83266525, + "step": 2539 + }, + { + "epoch": 0.22906614961446545, + "flos": 24313553576640.0, + "grad_norm": 1.8400116214177873, + "learning_rate": 3.598588511804161e-06, + "loss": 0.7843, + "num_input_tokens_seen": 83296455, + "step": 2540 + }, + { + "epoch": 0.2291563331379357, + "flos": 26902462748640.0, + "grad_norm": 2.4780369169304, + "learning_rate": 3.5982373681668987e-06, + "loss": 0.7856, + "num_input_tokens_seen": 83320205, + "step": 2541 + }, + { + "epoch": 0.22924651666140597, + "flos": 14517890987520.0, + "grad_norm": 1.9724879429089712, + "learning_rate": 3.597886088159212e-06, + "loss": 0.787, + "num_input_tokens_seen": 83348550, + "step": 2542 + }, + { + "epoch": 0.22933670018487623, + "flos": 45667558138080.0, + "grad_norm": 1.6212819920826735, + "learning_rate": 3.597534671811074e-06, + "loss": 0.7094, + "num_input_tokens_seen": 83382335, + "step": 2543 + }, + { + "epoch": 0.2294268837083465, + "flos": 29017903320480.0, + "grad_norm": 13.16435459616588, + "learning_rate": 3.5971831191524684e-06, + "loss": 0.7583, + "num_input_tokens_seen": 83412010, + "step": 2544 + }, + { + "epoch": 0.22951706723181675, + "flos": 25297889289120.0, + "grad_norm": 2.1959362297874154, + "learning_rate": 3.5968314302133925e-06, + "loss": 0.8287, + "num_input_tokens_seen": 83439585, + "step": 2545 + }, + { + "epoch": 0.229607250755287, + "flos": 16958692901280.0, + "grad_norm": 2.095792542724855, + "learning_rate": 3.596479605023854e-06, + "loss": 0.7773, + "num_input_tokens_seen": 83463805, + "step": 2546 + }, + { + "epoch": 0.22969743427875727, + "flos": 11165468879040.0, + "grad_norm": 1.6464244971920892, + "learning_rate": 3.596127643613873e-06, + "loss": 0.7928, + "num_input_tokens_seen": 83488390, + "step": 2547 + }, + { + "epoch": 0.22978761780222753, + "flos": 24866205232800.0, + "grad_norm": 1.7735953239761264, + "learning_rate": 3.59577554601348e-06, + "loss": 0.7993, + "num_input_tokens_seen": 83518010, + "step": 2548 + }, + { + "epoch": 0.2298778013256978, + "flos": 21980970035040.0, + "grad_norm": 2.506291674032028, + "learning_rate": 3.595423312252719e-06, + "loss": 0.8296, + "num_input_tokens_seen": 83547990, + "step": 2549 + }, + { + "epoch": 0.22996798484916806, + "flos": 22678792538400.0, + "grad_norm": 1.7549717306919659, + "learning_rate": 3.5950709423616436e-06, + "loss": 0.7616, + "num_input_tokens_seen": 83574635, + "step": 2550 + }, + { + "epoch": 0.23005816837263832, + "flos": 29017568792640.0, + "grad_norm": 2.7321994519182264, + "learning_rate": 3.5947184363703203e-06, + "loss": 0.7572, + "num_input_tokens_seen": 83603800, + "step": 2551 + }, + { + "epoch": 0.23014835189610858, + "flos": 31162156189920.0, + "grad_norm": 2.8594266959765005, + "learning_rate": 3.5943657943088274e-06, + "loss": 0.7292, + "num_input_tokens_seen": 83635100, + "step": 2552 + }, + { + "epoch": 0.23023853541957884, + "flos": 29164263600000.0, + "grad_norm": 1.6362067175051744, + "learning_rate": 3.5940130162072525e-06, + "loss": 0.704, + "num_input_tokens_seen": 83668160, + "step": 2553 + }, + { + "epoch": 0.2303287189430491, + "flos": 20893508207520.0, + "grad_norm": 3.937930129036303, + "learning_rate": 3.5936601020956985e-06, + "loss": 0.7814, + "num_input_tokens_seen": 83697125, + "step": 2554 + }, + { + "epoch": 0.23041890246651936, + "flos": 20745846986400.0, + "grad_norm": 1.85246744052624, + "learning_rate": 3.5933070520042772e-06, + "loss": 0.8118, + "num_input_tokens_seen": 83725065, + "step": 2555 + }, + { + "epoch": 0.23050908598998962, + "flos": 52823669141280.0, + "grad_norm": 0.775512147696392, + "learning_rate": 3.5929538659631133e-06, + "loss": 0.63, + "num_input_tokens_seen": 83814390, + "step": 2556 + }, + { + "epoch": 0.23059926951345988, + "flos": 15355680571680.0, + "grad_norm": 1.8942646900909832, + "learning_rate": 3.592600544002341e-06, + "loss": 0.8108, + "num_input_tokens_seen": 83841525, + "step": 2557 + }, + { + "epoch": 0.23068945303693014, + "flos": 21331027916160.0, + "grad_norm": 1.8207646410079876, + "learning_rate": 3.5922470861521098e-06, + "loss": 0.7822, + "num_input_tokens_seen": 83868545, + "step": 2558 + }, + { + "epoch": 0.2307796365604004, + "flos": 19763852968800.0, + "grad_norm": 6.080393389975312, + "learning_rate": 3.591893492442577e-06, + "loss": 0.7486, + "num_input_tokens_seen": 83894695, + "step": 2559 + }, + { + "epoch": 0.23086982008387066, + "flos": 21871599400320.0, + "grad_norm": 1.7929594417820744, + "learning_rate": 3.591539762903914e-06, + "loss": 0.8357, + "num_input_tokens_seen": 83920430, + "step": 2560 + }, + { + "epoch": 0.23096000360734095, + "flos": 23808261928320.0, + "grad_norm": 1.7920024320637002, + "learning_rate": 3.591185897566303e-06, + "loss": 0.7077, + "num_input_tokens_seen": 83950765, + "step": 2561 + }, + { + "epoch": 0.2310501871308112, + "flos": 27524075776800.0, + "grad_norm": 1.827819016109597, + "learning_rate": 3.590831896459937e-06, + "loss": 0.7769, + "num_input_tokens_seen": 83980115, + "step": 2562 + }, + { + "epoch": 0.23114037065428147, + "flos": 26868558193920.0, + "grad_norm": 1.9344302439157763, + "learning_rate": 3.5904777596150222e-06, + "loss": 0.7832, + "num_input_tokens_seen": 84009970, + "step": 2563 + }, + { + "epoch": 0.23123055417775173, + "flos": 20601977080800.0, + "grad_norm": 2.189282154550639, + "learning_rate": 3.590123487061775e-06, + "loss": 0.7761, + "num_input_tokens_seen": 84037430, + "step": 2564 + }, + { + "epoch": 0.231320737701222, + "flos": 26977631470560.0, + "grad_norm": 2.383897778111636, + "learning_rate": 3.589769078830424e-06, + "loss": 0.8745, + "num_input_tokens_seen": 84066705, + "step": 2565 + }, + { + "epoch": 0.23141092122469226, + "flos": 35757804354720.0, + "grad_norm": 1.9537322664340782, + "learning_rate": 3.58941453495121e-06, + "loss": 0.7581, + "num_input_tokens_seen": 84096680, + "step": 2566 + }, + { + "epoch": 0.23150110474816252, + "flos": 21330321690720.0, + "grad_norm": 2.8009217039507903, + "learning_rate": 3.5890598554543834e-06, + "loss": 0.7894, + "num_input_tokens_seen": 84122735, + "step": 2567 + }, + { + "epoch": 0.23159128827163278, + "flos": 21877323543360.0, + "grad_norm": 3.0092402252725585, + "learning_rate": 3.5887050403702073e-06, + "loss": 0.8279, + "num_input_tokens_seen": 84149035, + "step": 2568 + }, + { + "epoch": 0.23168147179510304, + "flos": 18598137329280.0, + "grad_norm": 1.7531483123371676, + "learning_rate": 3.588350089728958e-06, + "loss": 0.7543, + "num_input_tokens_seen": 84176605, + "step": 2569 + }, + { + "epoch": 0.2317716553185733, + "flos": 18562114098240.0, + "grad_norm": 2.3344843500511083, + "learning_rate": 3.5879950035609204e-06, + "loss": 0.8034, + "num_input_tokens_seen": 84202780, + "step": 2570 + }, + { + "epoch": 0.23186183884204356, + "flos": 26212891932000.0, + "grad_norm": 2.2578871559551197, + "learning_rate": 3.5876397818963933e-06, + "loss": 0.777, + "num_input_tokens_seen": 84231545, + "step": 2571 + }, + { + "epoch": 0.23195202236551382, + "flos": 21658247746080.0, + "grad_norm": 1.5873083194143343, + "learning_rate": 3.5872844247656858e-06, + "loss": 0.8352, + "num_input_tokens_seen": 84261015, + "step": 2572 + }, + { + "epoch": 0.23204220588898408, + "flos": 23516024576160.0, + "grad_norm": 1.6095958669268138, + "learning_rate": 3.5869289321991195e-06, + "loss": 0.7883, + "num_input_tokens_seen": 84291405, + "step": 2573 + }, + { + "epoch": 0.23213238941245434, + "flos": 24973680209760.0, + "grad_norm": 1.9506771770271496, + "learning_rate": 3.5865733042270263e-06, + "loss": 0.7722, + "num_input_tokens_seen": 84322565, + "step": 2574 + }, + { + "epoch": 0.2322225729359246, + "flos": 23225943070080.0, + "grad_norm": 2.0579757776271443, + "learning_rate": 3.5862175408797498e-06, + "loss": 0.8075, + "num_input_tokens_seen": 84351910, + "step": 2575 + }, + { + "epoch": 0.23231275645939486, + "flos": 27087299463360.0, + "grad_norm": 1.6915034382937693, + "learning_rate": 3.585861642187647e-06, + "loss": 0.7174, + "num_input_tokens_seen": 84381620, + "step": 2576 + }, + { + "epoch": 0.23240293998286513, + "flos": 26759410577760.0, + "grad_norm": 1.5122337548343028, + "learning_rate": 3.5855056081810845e-06, + "loss": 0.7755, + "num_input_tokens_seen": 84409905, + "step": 2577 + }, + { + "epoch": 0.2324931235063354, + "flos": 24388610789280.0, + "grad_norm": 1.5254273550615474, + "learning_rate": 3.5851494388904406e-06, + "loss": 0.78, + "num_input_tokens_seen": 84440315, + "step": 2578 + }, + { + "epoch": 0.23258330702980565, + "flos": 28325693450880.0, + "grad_norm": 2.292116178873491, + "learning_rate": 3.5847931343461064e-06, + "loss": 0.7163, + "num_input_tokens_seen": 84469750, + "step": 2579 + }, + { + "epoch": 0.2326734905532759, + "flos": 17796742673760.0, + "grad_norm": 1.8721781795589985, + "learning_rate": 3.5844366945784835e-06, + "loss": 0.7055, + "num_input_tokens_seen": 84496140, + "step": 2580 + }, + { + "epoch": 0.23276367407674617, + "flos": 18196975379520.0, + "grad_norm": 2.3698854085525762, + "learning_rate": 3.5840801196179856e-06, + "loss": 0.8352, + "num_input_tokens_seen": 84520645, + "step": 2581 + }, + { + "epoch": 0.23285385760021643, + "flos": 25184578659840.0, + "grad_norm": 1.891603201026975, + "learning_rate": 3.583723409495037e-06, + "loss": 0.8339, + "num_input_tokens_seen": 84546225, + "step": 2582 + }, + { + "epoch": 0.2329440411236867, + "flos": 30146406296640.0, + "grad_norm": 2.2059214274367656, + "learning_rate": 3.5833665642400747e-06, + "loss": 0.8331, + "num_input_tokens_seen": 84574960, + "step": 2583 + }, + { + "epoch": 0.23303422464715695, + "flos": 21765759892800.0, + "grad_norm": 2.155363370525417, + "learning_rate": 3.5830095838835472e-06, + "loss": 0.705, + "num_input_tokens_seen": 84602840, + "step": 2584 + }, + { + "epoch": 0.23312440817062724, + "flos": 20963250144480.0, + "grad_norm": 1.7433624411147206, + "learning_rate": 3.5826524684559125e-06, + "loss": 0.7491, + "num_input_tokens_seen": 84631400, + "step": 2585 + }, + { + "epoch": 0.2332145916940975, + "flos": 21913383944160.0, + "grad_norm": 2.1757455271556894, + "learning_rate": 3.5822952179876433e-06, + "loss": 0.8598, + "num_input_tokens_seen": 84658845, + "step": 2586 + }, + { + "epoch": 0.23330477521756776, + "flos": 30801886709760.0, + "grad_norm": 1.5288929606831345, + "learning_rate": 3.5819378325092205e-06, + "loss": 0.749, + "num_input_tokens_seen": 84690730, + "step": 2587 + }, + { + "epoch": 0.23339495874103802, + "flos": 38266229529120.0, + "grad_norm": 4.666195482354015, + "learning_rate": 3.581580312051139e-06, + "loss": 0.8004, + "num_input_tokens_seen": 84720750, + "step": 2588 + }, + { + "epoch": 0.23348514226450828, + "flos": 17687706566880.0, + "grad_norm": 1.7255053844119181, + "learning_rate": 3.5812226566439057e-06, + "loss": 0.705, + "num_input_tokens_seen": 84748225, + "step": 2589 + }, + { + "epoch": 0.23357532578797854, + "flos": 21658730952960.0, + "grad_norm": 1.9089257893433078, + "learning_rate": 3.580864866318036e-06, + "loss": 0.7555, + "num_input_tokens_seen": 84777725, + "step": 2590 + }, + { + "epoch": 0.2336655093114488, + "flos": 21367162656480.0, + "grad_norm": 2.14428787379103, + "learning_rate": 3.580506941104059e-06, + "loss": 0.7845, + "num_input_tokens_seen": 84805715, + "step": 2591 + }, + { + "epoch": 0.23375569283491907, + "flos": 30912112248960.0, + "grad_norm": 1.744022036426332, + "learning_rate": 3.580148881032515e-06, + "loss": 0.7664, + "num_input_tokens_seen": 84835485, + "step": 2592 + }, + { + "epoch": 0.23384587635838933, + "flos": 24171616498560.0, + "grad_norm": 1.8804626363314942, + "learning_rate": 3.5797906861339556e-06, + "loss": 0.7829, + "num_input_tokens_seen": 84865865, + "step": 2593 + }, + { + "epoch": 0.2339360598818596, + "flos": 43989637274880.0, + "grad_norm": 1.9002215632702446, + "learning_rate": 3.5794323564389435e-06, + "loss": 0.8037, + "num_input_tokens_seen": 84897965, + "step": 2594 + }, + { + "epoch": 0.23402624340532985, + "flos": 22933334020320.0, + "grad_norm": 2.438689210760754, + "learning_rate": 3.579073891978055e-06, + "loss": 0.7919, + "num_input_tokens_seen": 84923815, + "step": 2595 + }, + { + "epoch": 0.2341164269288001, + "flos": 20857001769600.0, + "grad_norm": 2.8545464775777196, + "learning_rate": 3.5787152927818746e-06, + "loss": 0.7496, + "num_input_tokens_seen": 84953215, + "step": 2596 + }, + { + "epoch": 0.23420661045227037, + "flos": 26758778691840.0, + "grad_norm": 1.62557885057178, + "learning_rate": 3.5783565588810003e-06, + "loss": 0.8433, + "num_input_tokens_seen": 84983335, + "step": 2597 + }, + { + "epoch": 0.23429679397574063, + "flos": 20747482455840.0, + "grad_norm": 1.7594774965070517, + "learning_rate": 3.5779976903060412e-06, + "loss": 0.7504, + "num_input_tokens_seen": 85012275, + "step": 2598 + }, + { + "epoch": 0.2343869774992109, + "flos": 25338595909920.0, + "grad_norm": 1.679156410179651, + "learning_rate": 3.577638687087619e-06, + "loss": 0.8075, + "num_input_tokens_seen": 85041080, + "step": 2599 + }, + { + "epoch": 0.23447716102268115, + "flos": 23152298308320.0, + "grad_norm": 1.732611988736247, + "learning_rate": 3.577279549256364e-06, + "loss": 0.786, + "num_input_tokens_seen": 85069570, + "step": 2600 + }, + { + "epoch": 0.2345673445461514, + "flos": 24172359893760.0, + "grad_norm": 2.1022786901890873, + "learning_rate": 3.5769202768429213e-06, + "loss": 0.8247, + "num_input_tokens_seen": 85098260, + "step": 2601 + }, + { + "epoch": 0.23465752806962167, + "flos": 28981062354720.0, + "grad_norm": 2.365150204459442, + "learning_rate": 3.5765608698779454e-06, + "loss": 0.7609, + "num_input_tokens_seen": 85126080, + "step": 2602 + }, + { + "epoch": 0.23474771159309193, + "flos": 20893805565600.0, + "grad_norm": 1.8407189622091429, + "learning_rate": 3.5762013283921033e-06, + "loss": 0.7881, + "num_input_tokens_seen": 85155110, + "step": 2603 + }, + { + "epoch": 0.2348378951165622, + "flos": 22131567667200.0, + "grad_norm": 4.584550846110368, + "learning_rate": 3.5758416524160728e-06, + "loss": 0.7825, + "num_input_tokens_seen": 85183410, + "step": 2604 + }, + { + "epoch": 0.23492807864003246, + "flos": 21181099339680.0, + "grad_norm": 1.8725965990887363, + "learning_rate": 3.5754818419805427e-06, + "loss": 0.7723, + "num_input_tokens_seen": 85210435, + "step": 2605 + }, + { + "epoch": 0.23501826216350272, + "flos": 28580383611840.0, + "grad_norm": 1.489216428087902, + "learning_rate": 3.575121897116216e-06, + "loss": 0.7601, + "num_input_tokens_seen": 85243790, + "step": 2606 + }, + { + "epoch": 0.23510844568697298, + "flos": 30038225094240.0, + "grad_norm": 1.5135268800262154, + "learning_rate": 3.574761817853803e-06, + "loss": 0.8172, + "num_input_tokens_seen": 85271920, + "step": 2607 + }, + { + "epoch": 0.23519862921044324, + "flos": 31713544074240.0, + "grad_norm": 1.4513914370882013, + "learning_rate": 3.5744016042240287e-06, + "loss": 0.7158, + "num_input_tokens_seen": 85304635, + "step": 2608 + }, + { + "epoch": 0.23528881273391353, + "flos": 24609024697920.0, + "grad_norm": 2.125986319130977, + "learning_rate": 3.5740412562576286e-06, + "loss": 0.7154, + "num_input_tokens_seen": 85332630, + "step": 2609 + }, + { + "epoch": 0.2353789962573838, + "flos": 24425823452640.0, + "grad_norm": 1.9294731855108396, + "learning_rate": 3.573680773985349e-06, + "loss": 0.7749, + "num_input_tokens_seen": 85360660, + "step": 2610 + }, + { + "epoch": 0.23546917978085405, + "flos": 32256754611360.0, + "grad_norm": 2.636876803711941, + "learning_rate": 3.5733201574379486e-06, + "loss": 0.7741, + "num_input_tokens_seen": 85389915, + "step": 2611 + }, + { + "epoch": 0.2355593633043243, + "flos": 23735026033920.0, + "grad_norm": 1.8514578101584251, + "learning_rate": 3.5729594066461975e-06, + "loss": 0.7162, + "num_input_tokens_seen": 85419890, + "step": 2612 + }, + { + "epoch": 0.23564954682779457, + "flos": 28652950450560.0, + "grad_norm": 1.6047663950631565, + "learning_rate": 3.572598521640876e-06, + "loss": 0.7952, + "num_input_tokens_seen": 85450310, + "step": 2613 + }, + { + "epoch": 0.23573973035126483, + "flos": 61968466101120.0, + "grad_norm": 0.8007067017629348, + "learning_rate": 3.5722375024527782e-06, + "loss": 0.6342, + "num_input_tokens_seen": 85537770, + "step": 2614 + }, + { + "epoch": 0.2358299138747351, + "flos": 23622533139360.0, + "grad_norm": 2.591571766464059, + "learning_rate": 3.571876349112707e-06, + "loss": 0.8004, + "num_input_tokens_seen": 85566035, + "step": 2615 + }, + { + "epoch": 0.23592009739820535, + "flos": 27047707935360.0, + "grad_norm": 1.858172757291733, + "learning_rate": 3.5715150616514784e-06, + "loss": 0.7783, + "num_input_tokens_seen": 85594945, + "step": 2616 + }, + { + "epoch": 0.2360102809216756, + "flos": 16776532409280.0, + "grad_norm": 2.0938851611792817, + "learning_rate": 3.5711536400999196e-06, + "loss": 0.833, + "num_input_tokens_seen": 85621530, + "step": 2617 + }, + { + "epoch": 0.23610046444514587, + "flos": 57687951860640.0, + "grad_norm": 0.7103904044330204, + "learning_rate": 3.570792084488869e-06, + "loss": 0.637, + "num_input_tokens_seen": 85710805, + "step": 2618 + }, + { + "epoch": 0.23619064796861614, + "flos": 66674637163200.0, + "grad_norm": 0.733282941112767, + "learning_rate": 3.5704303948491764e-06, + "loss": 0.649, + "num_input_tokens_seen": 85791420, + "step": 2619 + }, + { + "epoch": 0.2362808314920864, + "flos": 28653879694560.0, + "grad_norm": 1.7067370128414066, + "learning_rate": 3.5700685712117035e-06, + "loss": 0.8001, + "num_input_tokens_seen": 85824235, + "step": 2620 + }, + { + "epoch": 0.23637101501555666, + "flos": 22460237117760.0, + "grad_norm": 2.1249713895779063, + "learning_rate": 3.5697066136073227e-06, + "loss": 0.7645, + "num_input_tokens_seen": 85853310, + "step": 2621 + }, + { + "epoch": 0.23646119853902692, + "flos": 20419928098080.0, + "grad_norm": 1.5249064870210682, + "learning_rate": 3.5693445220669184e-06, + "loss": 0.79, + "num_input_tokens_seen": 85882255, + "step": 2622 + }, + { + "epoch": 0.23655138206249718, + "flos": 26103707146080.0, + "grad_norm": 3.0445901764689722, + "learning_rate": 3.568982296621386e-06, + "loss": 0.7731, + "num_input_tokens_seen": 85914380, + "step": 2623 + }, + { + "epoch": 0.23664156558596744, + "flos": 21251361653280.0, + "grad_norm": 2.2669763152538933, + "learning_rate": 3.5686199373016325e-06, + "loss": 0.8559, + "num_input_tokens_seen": 85942210, + "step": 2624 + }, + { + "epoch": 0.2367317491094377, + "flos": 26358657495360.0, + "grad_norm": 1.5699472334674793, + "learning_rate": 3.568257444138577e-06, + "loss": 0.7262, + "num_input_tokens_seen": 85971970, + "step": 2625 + }, + { + "epoch": 0.23682193263290796, + "flos": 18707247775680.0, + "grad_norm": 2.514015393618947, + "learning_rate": 3.5678948171631495e-06, + "loss": 0.7794, + "num_input_tokens_seen": 85998995, + "step": 2626 + }, + { + "epoch": 0.23691211615637822, + "flos": 17323274073600.0, + "grad_norm": 1.9778840242141542, + "learning_rate": 3.5675320564062908e-06, + "loss": 0.7849, + "num_input_tokens_seen": 86025860, + "step": 2627 + }, + { + "epoch": 0.23700229967984848, + "flos": 34700827464000.0, + "grad_norm": 1.7976188778463649, + "learning_rate": 3.5671691618989533e-06, + "loss": 0.6922, + "num_input_tokens_seen": 86054355, + "step": 2628 + }, + { + "epoch": 0.23709248320331874, + "flos": 21766874985600.0, + "grad_norm": 2.3280856750034458, + "learning_rate": 3.5668061336721024e-06, + "loss": 0.7852, + "num_input_tokens_seen": 86082515, + "step": 2629 + }, + { + "epoch": 0.237182666726789, + "flos": 20274422723040.0, + "grad_norm": 2.0923669146985424, + "learning_rate": 3.5664429717567117e-06, + "loss": 0.7735, + "num_input_tokens_seen": 86108520, + "step": 2630 + }, + { + "epoch": 0.23727285025025927, + "flos": 25010781363840.0, + "grad_norm": 1.6147951130770142, + "learning_rate": 3.56607967618377e-06, + "loss": 0.8106, + "num_input_tokens_seen": 86139330, + "step": 2631 + }, + { + "epoch": 0.23736303377372953, + "flos": 22606300039200.0, + "grad_norm": 1.8503166546329601, + "learning_rate": 3.5657162469842754e-06, + "loss": 0.7931, + "num_input_tokens_seen": 86166835, + "step": 2632 + }, + { + "epoch": 0.23745321729719981, + "flos": 20485618531200.0, + "grad_norm": 1.9336451051616101, + "learning_rate": 3.5653526841892374e-06, + "loss": 0.7821, + "num_input_tokens_seen": 86195095, + "step": 2633 + }, + { + "epoch": 0.23754340082067008, + "flos": 24861521843040.0, + "grad_norm": 1.8441963302127344, + "learning_rate": 3.564988987829676e-06, + "loss": 0.8104, + "num_input_tokens_seen": 86222910, + "step": 2634 + }, + { + "epoch": 0.23763358434414034, + "flos": 24860815617600.0, + "grad_norm": 1.6318129258784804, + "learning_rate": 3.564625157936626e-06, + "loss": 0.7221, + "num_input_tokens_seen": 86252515, + "step": 2635 + }, + { + "epoch": 0.2377237678676106, + "flos": 25266103410720.0, + "grad_norm": 1.757341855939209, + "learning_rate": 3.56426119454113e-06, + "loss": 0.8181, + "num_input_tokens_seen": 86281205, + "step": 2636 + }, + { + "epoch": 0.23781395139108086, + "flos": 21621852817440.0, + "grad_norm": 6.686178772083919, + "learning_rate": 3.5638970976742436e-06, + "loss": 0.8547, + "num_input_tokens_seen": 86305480, + "step": 2637 + }, + { + "epoch": 0.23790413491455112, + "flos": 26684316195360.0, + "grad_norm": 2.5350408542508633, + "learning_rate": 3.5635328673670335e-06, + "loss": 0.8082, + "num_input_tokens_seen": 86330295, + "step": 2638 + }, + { + "epoch": 0.23799431843802138, + "flos": 29596430863200.0, + "grad_norm": 6.428931598564682, + "learning_rate": 3.5631685036505783e-06, + "loss": 0.8705, + "num_input_tokens_seen": 86360130, + "step": 2639 + }, + { + "epoch": 0.23808450196149164, + "flos": 25408932563040.0, + "grad_norm": 1.477720180184145, + "learning_rate": 3.562804006555966e-06, + "loss": 0.7934, + "num_input_tokens_seen": 86392060, + "step": 2640 + }, + { + "epoch": 0.2381746854849619, + "flos": 27633595090560.0, + "grad_norm": 1.6732314836932938, + "learning_rate": 3.5624393761143e-06, + "loss": 0.7196, + "num_input_tokens_seen": 86421660, + "step": 2641 + }, + { + "epoch": 0.23826486900843216, + "flos": 21040574712480.0, + "grad_norm": 1.940726516206126, + "learning_rate": 3.5620746123566906e-06, + "loss": 0.693, + "num_input_tokens_seen": 86449740, + "step": 2642 + }, + { + "epoch": 0.23835505253190242, + "flos": 23839899127680.0, + "grad_norm": 1.7190131939597468, + "learning_rate": 3.5617097153142623e-06, + "loss": 0.7275, + "num_input_tokens_seen": 86478340, + "step": 2643 + }, + { + "epoch": 0.23844523605537268, + "flos": 23589780847200.0, + "grad_norm": 1.703679685646031, + "learning_rate": 3.5613446850181497e-06, + "loss": 0.8218, + "num_input_tokens_seen": 86506370, + "step": 2644 + }, + { + "epoch": 0.23853541957884294, + "flos": 57928251590880.0, + "grad_norm": 0.606968235000135, + "learning_rate": 3.5609795214994996e-06, + "loss": 0.5373, + "num_input_tokens_seen": 86599100, + "step": 2645 + }, + { + "epoch": 0.2386256031023132, + "flos": 25338781758720.0, + "grad_norm": 2.0252959800391386, + "learning_rate": 3.560614224789469e-06, + "loss": 0.8384, + "num_input_tokens_seen": 86629490, + "step": 2646 + }, + { + "epoch": 0.23871578662578347, + "flos": 24136039304640.0, + "grad_norm": 1.624135227450377, + "learning_rate": 3.5602487949192285e-06, + "loss": 0.7293, + "num_input_tokens_seen": 86657010, + "step": 2647 + }, + { + "epoch": 0.23880597014925373, + "flos": 40785433933920.0, + "grad_norm": 2.062254226375027, + "learning_rate": 3.559883231919957e-06, + "loss": 0.7227, + "num_input_tokens_seen": 86688180, + "step": 2648 + }, + { + "epoch": 0.238896153672724, + "flos": 24317716589760.0, + "grad_norm": 1.590001728728692, + "learning_rate": 3.5595175358228473e-06, + "loss": 0.7536, + "num_input_tokens_seen": 86718105, + "step": 2649 + }, + { + "epoch": 0.23898633719619425, + "flos": 28325656281120.0, + "grad_norm": 1.9083250664722118, + "learning_rate": 3.5591517066591027e-06, + "loss": 0.7773, + "num_input_tokens_seen": 86747935, + "step": 2650 + }, + { + "epoch": 0.2390765207196645, + "flos": 14955299186880.0, + "grad_norm": 3.9317564475845344, + "learning_rate": 3.5587857444599364e-06, + "loss": 0.7716, + "num_input_tokens_seen": 86772965, + "step": 2651 + }, + { + "epoch": 0.23916670424313477, + "flos": 16958804410560.0, + "grad_norm": 1.5878517361954534, + "learning_rate": 3.5584196492565766e-06, + "loss": 0.7463, + "num_input_tokens_seen": 86800465, + "step": 2652 + }, + { + "epoch": 0.23925688776660503, + "flos": 17286990654240.0, + "grad_norm": 1.8112986299187679, + "learning_rate": 3.5580534210802587e-06, + "loss": 0.8428, + "num_input_tokens_seen": 86825060, + "step": 2653 + }, + { + "epoch": 0.2393470712900753, + "flos": 26394866575200.0, + "grad_norm": 1.5410422357889324, + "learning_rate": 3.557687059962232e-06, + "loss": 0.8265, + "num_input_tokens_seen": 86855275, + "step": 2654 + }, + { + "epoch": 0.23943725481354555, + "flos": 23328994845600.0, + "grad_norm": 2.4394054301282235, + "learning_rate": 3.5573205659337558e-06, + "loss": 0.7732, + "num_input_tokens_seen": 86883930, + "step": 2655 + }, + { + "epoch": 0.23952743833701584, + "flos": 21038902073280.0, + "grad_norm": 1.5625987212465513, + "learning_rate": 3.5569539390261025e-06, + "loss": 0.7681, + "num_input_tokens_seen": 86911735, + "step": 2656 + }, + { + "epoch": 0.2396176218604861, + "flos": 24245224090560.0, + "grad_norm": 1.3190431531161955, + "learning_rate": 3.5565871792705543e-06, + "loss": 0.7949, + "num_input_tokens_seen": 86942270, + "step": 2657 + }, + { + "epoch": 0.23970780538395636, + "flos": 23552791202400.0, + "grad_norm": 1.5046647970322902, + "learning_rate": 3.5562202866984045e-06, + "loss": 0.7727, + "num_input_tokens_seen": 86970120, + "step": 2658 + }, + { + "epoch": 0.23979798890742662, + "flos": 23297952362400.0, + "grad_norm": 2.2678246704969283, + "learning_rate": 3.5558532613409594e-06, + "loss": 0.7946, + "num_input_tokens_seen": 86997255, + "step": 2659 + }, + { + "epoch": 0.23988817243089688, + "flos": 29746991325600.0, + "grad_norm": 1.8342463719567497, + "learning_rate": 3.555486103229535e-06, + "loss": 0.7268, + "num_input_tokens_seen": 87028125, + "step": 2660 + }, + { + "epoch": 0.23997835595436715, + "flos": 20019360864480.0, + "grad_norm": 32.042107534827316, + "learning_rate": 3.5551188123954595e-06, + "loss": 0.8022, + "num_input_tokens_seen": 87053760, + "step": 2661 + }, + { + "epoch": 0.2400685394778374, + "flos": 20019509543520.0, + "grad_norm": 1.7787557840688868, + "learning_rate": 3.5547513888700715e-06, + "loss": 0.6763, + "num_input_tokens_seen": 87080710, + "step": 2662 + }, + { + "epoch": 0.24015872300130767, + "flos": 25952551967520.0, + "grad_norm": 2.1796796244817402, + "learning_rate": 3.5543838326847224e-06, + "loss": 0.8129, + "num_input_tokens_seen": 87108845, + "step": 2663 + }, + { + "epoch": 0.24024890652477793, + "flos": 33936050755680.0, + "grad_norm": 1.8048270695869706, + "learning_rate": 3.5540161438707744e-06, + "loss": 0.7588, + "num_input_tokens_seen": 87138675, + "step": 2664 + }, + { + "epoch": 0.2403390900482482, + "flos": 32187198523200.0, + "grad_norm": 1.7644345481625994, + "learning_rate": 3.5536483224596e-06, + "loss": 0.9394, + "num_input_tokens_seen": 87170780, + "step": 2665 + }, + { + "epoch": 0.24042927357171845, + "flos": 29054261079360.0, + "grad_norm": 3.1455102379880433, + "learning_rate": 3.553280368482584e-06, + "loss": 0.7589, + "num_input_tokens_seen": 87201265, + "step": 2666 + }, + { + "epoch": 0.2405194570951887, + "flos": 49708175782080.0, + "grad_norm": 2.0883145736050635, + "learning_rate": 3.5529122819711227e-06, + "loss": 0.6008, + "num_input_tokens_seen": 87232860, + "step": 2667 + }, + { + "epoch": 0.24060964061865897, + "flos": 24828360683520.0, + "grad_norm": 2.039117571207645, + "learning_rate": 3.5525440629566223e-06, + "loss": 0.7263, + "num_input_tokens_seen": 87263515, + "step": 2668 + }, + { + "epoch": 0.24069982414212923, + "flos": 19254323967840.0, + "grad_norm": 2.9924676430338377, + "learning_rate": 3.552175711470502e-06, + "loss": 0.8009, + "num_input_tokens_seen": 87291460, + "step": 2669 + }, + { + "epoch": 0.2407900076655995, + "flos": 24135741946560.0, + "grad_norm": 2.061395407152079, + "learning_rate": 3.5518072275441912e-06, + "loss": 0.8232, + "num_input_tokens_seen": 87319175, + "step": 2670 + }, + { + "epoch": 0.24088019118906975, + "flos": 24605456400960.0, + "grad_norm": 2.090623143399131, + "learning_rate": 3.551438611209131e-06, + "loss": 0.8057, + "num_input_tokens_seen": 87349435, + "step": 2671 + }, + { + "epoch": 0.24097037471254001, + "flos": 27810960683520.0, + "grad_norm": 1.8914797403459118, + "learning_rate": 3.551069862496774e-06, + "loss": 0.8762, + "num_input_tokens_seen": 87379125, + "step": 2672 + }, + { + "epoch": 0.24106055823601027, + "flos": 70851982385280.0, + "grad_norm": 0.654462332230996, + "learning_rate": 3.5507009814385846e-06, + "loss": 0.5981, + "num_input_tokens_seen": 87469815, + "step": 2673 + }, + { + "epoch": 0.24115074175948054, + "flos": 56562831332640.0, + "grad_norm": 0.8047028697499666, + "learning_rate": 3.550331968066036e-06, + "loss": 0.5891, + "num_input_tokens_seen": 87556830, + "step": 2674 + }, + { + "epoch": 0.2412409252829508, + "flos": 17067729008160.0, + "grad_norm": 2.7996583014846688, + "learning_rate": 3.549962822410616e-06, + "loss": 0.7955, + "num_input_tokens_seen": 87584840, + "step": 2675 + }, + { + "epoch": 0.24133110880642106, + "flos": 70807298600160.0, + "grad_norm": 0.6912672474246008, + "learning_rate": 3.5495935445038217e-06, + "loss": 0.5862, + "num_input_tokens_seen": 87686395, + "step": 2676 + }, + { + "epoch": 0.24142129232989132, + "flos": 21469062169440.0, + "grad_norm": 1.7805271355034953, + "learning_rate": 3.5492241343771612e-06, + "loss": 0.8193, + "num_input_tokens_seen": 87714030, + "step": 2677 + }, + { + "epoch": 0.24151147585336158, + "flos": 26540557799040.0, + "grad_norm": 2.210417993121479, + "learning_rate": 3.548854592062156e-06, + "loss": 0.7789, + "num_input_tokens_seen": 87740225, + "step": 2678 + }, + { + "epoch": 0.24160165937683184, + "flos": 17248179691200.0, + "grad_norm": 2.5839047466531366, + "learning_rate": 3.548484917590336e-06, + "loss": 0.7595, + "num_input_tokens_seen": 87767835, + "step": 2679 + }, + { + "epoch": 0.24169184290030213, + "flos": 43079541040320.0, + "grad_norm": 2.465867979891988, + "learning_rate": 3.5481151109932447e-06, + "loss": 0.6985, + "num_input_tokens_seen": 87795650, + "step": 2680 + }, + { + "epoch": 0.2417820264237724, + "flos": 21695274560640.0, + "grad_norm": 1.6199675911145868, + "learning_rate": 3.5477451723024364e-06, + "loss": 0.8056, + "num_input_tokens_seen": 87823575, + "step": 2681 + }, + { + "epoch": 0.24187220994724265, + "flos": 70631642816160.0, + "grad_norm": 0.5913849457666616, + "learning_rate": 3.5473751015494757e-06, + "loss": 0.5705, + "num_input_tokens_seen": 87923415, + "step": 2682 + }, + { + "epoch": 0.2419623934707129, + "flos": 21257643342720.0, + "grad_norm": 1.701760681773376, + "learning_rate": 3.547004898765939e-06, + "loss": 0.7949, + "num_input_tokens_seen": 87952655, + "step": 2683 + }, + { + "epoch": 0.24205257699418317, + "flos": 23585803682880.0, + "grad_norm": 1.3922986877772772, + "learning_rate": 3.546634563983414e-06, + "loss": 0.7979, + "num_input_tokens_seen": 87981425, + "step": 2684 + }, + { + "epoch": 0.24214276051765343, + "flos": 24859291657440.0, + "grad_norm": 1.545506266758634, + "learning_rate": 3.5462640972335002e-06, + "loss": 0.7418, + "num_input_tokens_seen": 88012340, + "step": 2685 + }, + { + "epoch": 0.2422329440411237, + "flos": 26721714707520.0, + "grad_norm": 1.633036109908179, + "learning_rate": 3.5458934985478077e-06, + "loss": 0.7992, + "num_input_tokens_seen": 88042740, + "step": 2686 + }, + { + "epoch": 0.24232312756459395, + "flos": 21912603379200.0, + "grad_norm": 2.1404805378304217, + "learning_rate": 3.5455227679579577e-06, + "loss": 0.7605, + "num_input_tokens_seen": 88069785, + "step": 2687 + }, + { + "epoch": 0.24241331108806422, + "flos": 22968576686400.0, + "grad_norm": 2.2758360070746337, + "learning_rate": 3.545151905495584e-06, + "loss": 0.8878, + "num_input_tokens_seen": 88097970, + "step": 2688 + }, + { + "epoch": 0.24250349461153448, + "flos": 21075520020480.0, + "grad_norm": 6.629203182801624, + "learning_rate": 3.544780911192329e-06, + "loss": 0.7653, + "num_input_tokens_seen": 88125615, + "step": 2689 + }, + { + "epoch": 0.24259367813500474, + "flos": 24169534992000.0, + "grad_norm": 1.6705977650372956, + "learning_rate": 3.544409785079849e-06, + "loss": 0.7306, + "num_input_tokens_seen": 88156005, + "step": 2690 + }, + { + "epoch": 0.242683861658475, + "flos": 19545409057440.0, + "grad_norm": 1.5633915455736347, + "learning_rate": 3.5440385271898103e-06, + "loss": 0.7275, + "num_input_tokens_seen": 88184795, + "step": 2691 + }, + { + "epoch": 0.24277404518194526, + "flos": 24274036388160.0, + "grad_norm": 1.7627515789149886, + "learning_rate": 3.5436671375538903e-06, + "loss": 0.7069, + "num_input_tokens_seen": 88209645, + "step": 2692 + }, + { + "epoch": 0.24286422870541552, + "flos": 24172471403040.0, + "grad_norm": 2.3097774872393795, + "learning_rate": 3.543295616203779e-06, + "loss": 0.7781, + "num_input_tokens_seen": 88237560, + "step": 2693 + }, + { + "epoch": 0.24295441222888578, + "flos": 29012625214560.0, + "grad_norm": 1.8232817438673985, + "learning_rate": 3.542923963171176e-06, + "loss": 0.8769, + "num_input_tokens_seen": 88267220, + "step": 2694 + }, + { + "epoch": 0.24304459575235604, + "flos": 29416463387040.0, + "grad_norm": 1.5172407838243254, + "learning_rate": 3.542552178487793e-06, + "loss": 0.6999, + "num_input_tokens_seen": 88299895, + "step": 2695 + }, + { + "epoch": 0.2431347792758263, + "flos": 24785646895680.0, + "grad_norm": 2.1325504575173415, + "learning_rate": 3.5421802621853523e-06, + "loss": 0.7733, + "num_input_tokens_seen": 88328440, + "step": 2696 + }, + { + "epoch": 0.24322496279929656, + "flos": 27812150115840.0, + "grad_norm": 1.892307510161257, + "learning_rate": 3.5418082142955887e-06, + "loss": 0.7765, + "num_input_tokens_seen": 88357575, + "step": 2697 + }, + { + "epoch": 0.24331514632276682, + "flos": 59079433854240.0, + "grad_norm": 0.6672907505948149, + "learning_rate": 3.5414360348502463e-06, + "loss": 0.6145, + "num_input_tokens_seen": 88440900, + "step": 2698 + }, + { + "epoch": 0.24340532984623708, + "flos": 28107063690720.0, + "grad_norm": 1.937729606559493, + "learning_rate": 3.5410637238810825e-06, + "loss": 0.6831, + "num_input_tokens_seen": 88470770, + "step": 2699 + }, + { + "epoch": 0.24349551336970734, + "flos": 27706050420000.0, + "grad_norm": 6.794530284329298, + "learning_rate": 3.5406912814198635e-06, + "loss": 0.7344, + "num_input_tokens_seen": 88499645, + "step": 2700 + }, + { + "epoch": 0.2435856968931776, + "flos": 28507445075520.0, + "grad_norm": 4.217121177171604, + "learning_rate": 3.54031870749837e-06, + "loss": 0.7464, + "num_input_tokens_seen": 88529765, + "step": 2701 + }, + { + "epoch": 0.24367588041664787, + "flos": 22569124545600.0, + "grad_norm": 1.5482693947578685, + "learning_rate": 3.539946002148391e-06, + "loss": 0.7897, + "num_input_tokens_seen": 88556965, + "step": 2702 + }, + { + "epoch": 0.24376606394011813, + "flos": 21985876443360.0, + "grad_norm": 2.638535524991901, + "learning_rate": 3.5395731654017277e-06, + "loss": 0.7843, + "num_input_tokens_seen": 88580470, + "step": 2703 + }, + { + "epoch": 0.24385624746358842, + "flos": 19325329676640.0, + "grad_norm": 2.0876703386992683, + "learning_rate": 3.5392001972901923e-06, + "loss": 0.822, + "num_input_tokens_seen": 88603955, + "step": 2704 + }, + { + "epoch": 0.24394643098705868, + "flos": 22315029100800.0, + "grad_norm": 1.5311357883755634, + "learning_rate": 3.5388270978456098e-06, + "loss": 0.742, + "num_input_tokens_seen": 88632560, + "step": 2705 + }, + { + "epoch": 0.24403661451052894, + "flos": 24389502863520.0, + "grad_norm": 4.322049169554628, + "learning_rate": 3.5384538670998137e-06, + "loss": 0.7532, + "num_input_tokens_seen": 88659795, + "step": 2706 + }, + { + "epoch": 0.2441267980339992, + "flos": 19617975896160.0, + "grad_norm": 2.0035200137192377, + "learning_rate": 3.538080505084651e-06, + "loss": 0.8022, + "num_input_tokens_seen": 88686220, + "step": 2707 + }, + { + "epoch": 0.24421698155746946, + "flos": 26066791840800.0, + "grad_norm": 1.550173623605086, + "learning_rate": 3.5377070118319788e-06, + "loss": 0.7941, + "num_input_tokens_seen": 88715565, + "step": 2708 + }, + { + "epoch": 0.24430716508093972, + "flos": 40487806966560.0, + "grad_norm": 2.6840996633238174, + "learning_rate": 3.5373333873736657e-06, + "loss": 0.7401, + "num_input_tokens_seen": 88744725, + "step": 2709 + }, + { + "epoch": 0.24439734860440998, + "flos": 25593620598720.0, + "grad_norm": 1.530729555707763, + "learning_rate": 3.536959631741591e-06, + "loss": 0.7866, + "num_input_tokens_seen": 88774395, + "step": 2710 + }, + { + "epoch": 0.24448753212788024, + "flos": 26139284340000.0, + "grad_norm": 1.4181403274346684, + "learning_rate": 3.536585744967646e-06, + "loss": 0.8046, + "num_input_tokens_seen": 88807785, + "step": 2711 + }, + { + "epoch": 0.2445777156513505, + "flos": 16995459527520.0, + "grad_norm": 2.4995012739544173, + "learning_rate": 3.5362117270837326e-06, + "loss": 0.7603, + "num_input_tokens_seen": 88835400, + "step": 2712 + }, + { + "epoch": 0.24466789917482076, + "flos": 26468362657920.0, + "grad_norm": 3.740672400101727, + "learning_rate": 3.5358375781217634e-06, + "loss": 0.6846, + "num_input_tokens_seen": 88869270, + "step": 2713 + }, + { + "epoch": 0.24475808269829102, + "flos": 26613347656320.0, + "grad_norm": 2.0057178064772647, + "learning_rate": 3.535463298113664e-06, + "loss": 0.8111, + "num_input_tokens_seen": 88896510, + "step": 2714 + }, + { + "epoch": 0.24484826622176128, + "flos": 25885226064960.0, + "grad_norm": 1.7209702456538234, + "learning_rate": 3.5350888870913697e-06, + "loss": 0.741, + "num_input_tokens_seen": 88924465, + "step": 2715 + }, + { + "epoch": 0.24493844974523155, + "flos": 25294209482880.0, + "grad_norm": 2.23428557957268, + "learning_rate": 3.5347143450868273e-06, + "loss": 0.781, + "num_input_tokens_seen": 88952260, + "step": 2716 + }, + { + "epoch": 0.2450286332687018, + "flos": 20273753667360.0, + "grad_norm": 5.191566068716607, + "learning_rate": 3.534339672131994e-06, + "loss": 0.8004, + "num_input_tokens_seen": 88980090, + "step": 2717 + }, + { + "epoch": 0.24511881679217207, + "flos": 13893304378560.0, + "grad_norm": 2.6942868257632115, + "learning_rate": 3.5339648682588397e-06, + "loss": 0.7145, + "num_input_tokens_seen": 89006380, + "step": 2718 + }, + { + "epoch": 0.24520900031564233, + "flos": 19581841155840.0, + "grad_norm": 1.7833851515588715, + "learning_rate": 3.533589933499345e-06, + "loss": 0.819, + "num_input_tokens_seen": 89034925, + "step": 2719 + }, + { + "epoch": 0.2452991838391126, + "flos": 21950150570400.0, + "grad_norm": 2.718884958008793, + "learning_rate": 3.533214867885501e-06, + "loss": 0.7168, + "num_input_tokens_seen": 89063575, + "step": 2720 + }, + { + "epoch": 0.24538936736258285, + "flos": 25443803531520.0, + "grad_norm": 2.089682770705601, + "learning_rate": 3.53283967144931e-06, + "loss": 0.7232, + "num_input_tokens_seen": 89090860, + "step": 2721 + }, + { + "epoch": 0.2454795508860531, + "flos": 22131530497440.0, + "grad_norm": 1.826072192403884, + "learning_rate": 3.532464344222787e-06, + "loss": 0.7898, + "num_input_tokens_seen": 89118465, + "step": 2722 + }, + { + "epoch": 0.24556973440952337, + "flos": 16594148898720.0, + "grad_norm": 2.1130516505718973, + "learning_rate": 3.532088886237956e-06, + "loss": 0.8185, + "num_input_tokens_seen": 89146730, + "step": 2723 + }, + { + "epoch": 0.24565991793299363, + "flos": 71012881774560.0, + "grad_norm": 0.6526146514691693, + "learning_rate": 3.5317132975268535e-06, + "loss": 0.5268, + "num_input_tokens_seen": 89237950, + "step": 2724 + }, + { + "epoch": 0.2457501014564639, + "flos": 22895786829120.0, + "grad_norm": 2.817545201532178, + "learning_rate": 3.531337578121526e-06, + "loss": 0.807, + "num_input_tokens_seen": 89267360, + "step": 2725 + }, + { + "epoch": 0.24584028497993415, + "flos": 15100916071200.0, + "grad_norm": 1.6959260454383231, + "learning_rate": 3.530961728054033e-06, + "loss": 0.8158, + "num_input_tokens_seen": 89294350, + "step": 2726 + }, + { + "epoch": 0.24593046850340441, + "flos": 21768510455040.0, + "grad_norm": 1.6119232003680843, + "learning_rate": 3.5305857473564435e-06, + "loss": 0.7425, + "num_input_tokens_seen": 89321385, + "step": 2727 + }, + { + "epoch": 0.2460206520268747, + "flos": 16120345770720.0, + "grad_norm": 1.973240590182189, + "learning_rate": 3.5302096360608385e-06, + "loss": 0.7095, + "num_input_tokens_seen": 89346650, + "step": 2728 + }, + { + "epoch": 0.24611083555034496, + "flos": 20820644010720.0, + "grad_norm": 1.6526633239923618, + "learning_rate": 3.5298333941993105e-06, + "loss": 0.8343, + "num_input_tokens_seen": 89374175, + "step": 2729 + }, + { + "epoch": 0.24620101907381522, + "flos": 23042778994560.0, + "grad_norm": 2.017680221884331, + "learning_rate": 3.529457021803962e-06, + "loss": 0.7734, + "num_input_tokens_seen": 89399975, + "step": 2730 + }, + { + "epoch": 0.24629120259728549, + "flos": 14299595755200.0, + "grad_norm": 2.0802931410290006, + "learning_rate": 3.529080518906906e-06, + "loss": 0.7953, + "num_input_tokens_seen": 89426490, + "step": 2731 + }, + { + "epoch": 0.24638138612075575, + "flos": 25046841764640.0, + "grad_norm": 1.536759643320881, + "learning_rate": 3.5287038855402696e-06, + "loss": 0.7631, + "num_input_tokens_seen": 89456685, + "step": 2732 + }, + { + "epoch": 0.246471569644226, + "flos": 18634272069600.0, + "grad_norm": 1.9777154989771428, + "learning_rate": 3.528327121736188e-06, + "loss": 0.814, + "num_input_tokens_seen": 89484705, + "step": 2733 + }, + { + "epoch": 0.24656175316769627, + "flos": 21835315980960.0, + "grad_norm": 1.9188840589380924, + "learning_rate": 3.52795022752681e-06, + "loss": 0.8585, + "num_input_tokens_seen": 89511965, + "step": 2734 + }, + { + "epoch": 0.24665193669116653, + "flos": 21585495058560.0, + "grad_norm": 2.534374522104465, + "learning_rate": 3.5275732029442925e-06, + "loss": 0.8487, + "num_input_tokens_seen": 89539845, + "step": 2735 + }, + { + "epoch": 0.2467421202146368, + "flos": 24026928858240.0, + "grad_norm": 2.5772030193277424, + "learning_rate": 3.5271960480208077e-06, + "loss": 0.7209, + "num_input_tokens_seen": 89568965, + "step": 2736 + }, + { + "epoch": 0.24683230373810705, + "flos": 67412385722400.0, + "grad_norm": 0.7905618378510312, + "learning_rate": 3.526818762788534e-06, + "loss": 0.6273, + "num_input_tokens_seen": 89660490, + "step": 2737 + }, + { + "epoch": 0.2469224872615773, + "flos": 23406691111200.0, + "grad_norm": 1.6151140714815486, + "learning_rate": 3.5264413472796653e-06, + "loss": 0.8173, + "num_input_tokens_seen": 89689805, + "step": 2738 + }, + { + "epoch": 0.24701267078504757, + "flos": 21431998185120.0, + "grad_norm": 2.453223451230291, + "learning_rate": 3.5260638015264037e-06, + "loss": 0.7571, + "num_input_tokens_seen": 89714960, + "step": 2739 + }, + { + "epoch": 0.24710285430851783, + "flos": 27638724517440.0, + "grad_norm": 1.532441356961977, + "learning_rate": 3.5256861255609644e-06, + "loss": 0.7507, + "num_input_tokens_seen": 89749165, + "step": 2740 + }, + { + "epoch": 0.2471930378319881, + "flos": 18925877535840.0, + "grad_norm": 2.2610559101640852, + "learning_rate": 3.5253083194155723e-06, + "loss": 0.7953, + "num_input_tokens_seen": 89775515, + "step": 2741 + }, + { + "epoch": 0.24728322135545835, + "flos": 23657998824000.0, + "grad_norm": 1.3673863293619295, + "learning_rate": 3.5249303831224637e-06, + "loss": 0.7987, + "num_input_tokens_seen": 89805600, + "step": 2742 + }, + { + "epoch": 0.24737340487892862, + "flos": 22018888923840.0, + "grad_norm": 1.587013518518032, + "learning_rate": 3.524552316713887e-06, + "loss": 0.7669, + "num_input_tokens_seen": 89834000, + "step": 2743 + }, + { + "epoch": 0.24746358840239888, + "flos": 20854325546880.0, + "grad_norm": 1.8361901292070857, + "learning_rate": 3.5241741202220995e-06, + "loss": 0.8369, + "num_input_tokens_seen": 89862205, + "step": 2744 + }, + { + "epoch": 0.24755377192586914, + "flos": 25003533260640.0, + "grad_norm": 2.2312369551931788, + "learning_rate": 3.5237957936793724e-06, + "loss": 0.7652, + "num_input_tokens_seen": 89892730, + "step": 2745 + }, + { + "epoch": 0.2476439554493394, + "flos": 22312018350240.0, + "grad_norm": 1.4966075204293503, + "learning_rate": 3.523417337117986e-06, + "loss": 0.7425, + "num_input_tokens_seen": 89923345, + "step": 2746 + }, + { + "epoch": 0.24773413897280966, + "flos": 27596382427200.0, + "grad_norm": 1.486526375781323, + "learning_rate": 3.523038750570232e-06, + "loss": 0.7302, + "num_input_tokens_seen": 89954230, + "step": 2747 + }, + { + "epoch": 0.24782432249627992, + "flos": 24026334142080.0, + "grad_norm": 3.099730289170428, + "learning_rate": 3.522660034068414e-06, + "loss": 0.8095, + "num_input_tokens_seen": 89983700, + "step": 2748 + }, + { + "epoch": 0.24791450601975018, + "flos": 18999039090720.0, + "grad_norm": 1.893476495412071, + "learning_rate": 3.5222811876448464e-06, + "loss": 0.7915, + "num_input_tokens_seen": 90010340, + "step": 2749 + }, + { + "epoch": 0.24800468954322044, + "flos": 32077753548960.0, + "grad_norm": 1.5571182647752952, + "learning_rate": 3.521902211331854e-06, + "loss": 0.799, + "num_input_tokens_seen": 90044685, + "step": 2750 + }, + { + "epoch": 0.2480948730666907, + "flos": 24973531530720.0, + "grad_norm": 1.8834000932907142, + "learning_rate": 3.5215231051617726e-06, + "loss": 0.7279, + "num_input_tokens_seen": 90072695, + "step": 2751 + }, + { + "epoch": 0.248185056590161, + "flos": 31489636208160.0, + "grad_norm": 1.7447076881657786, + "learning_rate": 3.521143869166951e-06, + "loss": 0.6831, + "num_input_tokens_seen": 90100340, + "step": 2752 + }, + { + "epoch": 0.24827524011363125, + "flos": 14226508539840.0, + "grad_norm": 2.2001118718226937, + "learning_rate": 3.5207645033797464e-06, + "loss": 0.812, + "num_input_tokens_seen": 90124860, + "step": 2753 + }, + { + "epoch": 0.2483654236371015, + "flos": 31640568368160.0, + "grad_norm": 2.1593079927815833, + "learning_rate": 3.5203850078325293e-06, + "loss": 0.7017, + "num_input_tokens_seen": 90152900, + "step": 2754 + }, + { + "epoch": 0.24845560716057177, + "flos": 23259698945760.0, + "grad_norm": 2.188991583057322, + "learning_rate": 3.5200053825576797e-06, + "loss": 0.6928, + "num_input_tokens_seen": 90181965, + "step": 2755 + }, + { + "epoch": 0.24854579068404203, + "flos": 26468362657920.0, + "grad_norm": 1.6708654047043459, + "learning_rate": 3.51962562758759e-06, + "loss": 0.7812, + "num_input_tokens_seen": 90213030, + "step": 2756 + }, + { + "epoch": 0.2486359742075123, + "flos": 28435212764640.0, + "grad_norm": 3.648567148392383, + "learning_rate": 3.5192457429546627e-06, + "loss": 0.7562, + "num_input_tokens_seen": 90243065, + "step": 2757 + }, + { + "epoch": 0.24872615773098256, + "flos": 36848648630400.0, + "grad_norm": 3.4791812249219802, + "learning_rate": 3.5188657286913115e-06, + "loss": 0.7462, + "num_input_tokens_seen": 90273245, + "step": 2758 + }, + { + "epoch": 0.24881634125445282, + "flos": 24244852392960.0, + "grad_norm": 2.2498691308850876, + "learning_rate": 3.518485584829961e-06, + "loss": 0.7401, + "num_input_tokens_seen": 90301285, + "step": 2759 + }, + { + "epoch": 0.24890652477792308, + "flos": 19181199582720.0, + "grad_norm": 1.8321920946516774, + "learning_rate": 3.5181053114030485e-06, + "loss": 0.7876, + "num_input_tokens_seen": 90328645, + "step": 2760 + }, + { + "epoch": 0.24899670830139334, + "flos": 30507307662720.0, + "grad_norm": 2.319888606028027, + "learning_rate": 3.5177249084430198e-06, + "loss": 0.7731, + "num_input_tokens_seen": 90359130, + "step": 2761 + }, + { + "epoch": 0.2490868918248636, + "flos": 16660285368960.0, + "grad_norm": 2.2621681939555556, + "learning_rate": 3.517344375982333e-06, + "loss": 0.811, + "num_input_tokens_seen": 90385800, + "step": 2762 + }, + { + "epoch": 0.24917707534833386, + "flos": 23328288620160.0, + "grad_norm": 1.838499448758007, + "learning_rate": 3.5169637140534565e-06, + "loss": 0.8101, + "num_input_tokens_seen": 90412580, + "step": 2763 + }, + { + "epoch": 0.24926725887180412, + "flos": 25848236420160.0, + "grad_norm": 2.0310795639133117, + "learning_rate": 3.5165829226888733e-06, + "loss": 0.7496, + "num_input_tokens_seen": 90442355, + "step": 2764 + }, + { + "epoch": 0.24935744239527438, + "flos": 23261297245440.0, + "grad_norm": 1.6356893502427154, + "learning_rate": 3.516202001921072e-06, + "loss": 0.8035, + "num_input_tokens_seen": 90471780, + "step": 2765 + }, + { + "epoch": 0.24944762591874464, + "flos": 22929914402400.0, + "grad_norm": 3.790296593487317, + "learning_rate": 3.515820951782555e-06, + "loss": 0.8045, + "num_input_tokens_seen": 90501850, + "step": 2766 + }, + { + "epoch": 0.2495378094422149, + "flos": 23297766513600.0, + "grad_norm": 2.366038833763812, + "learning_rate": 3.5154397723058366e-06, + "loss": 0.7051, + "num_input_tokens_seen": 90530215, + "step": 2767 + }, + { + "epoch": 0.24962799296568516, + "flos": 18889519776960.0, + "grad_norm": 1.8935302606888889, + "learning_rate": 3.5150584635234416e-06, + "loss": 0.8321, + "num_input_tokens_seen": 90558900, + "step": 2768 + }, + { + "epoch": 0.24971817648915542, + "flos": 24062097184800.0, + "grad_norm": 1.4763951282403274, + "learning_rate": 3.5146770254679035e-06, + "loss": 0.8201, + "num_input_tokens_seen": 90590050, + "step": 2769 + }, + { + "epoch": 0.24980836001262569, + "flos": 21401290229760.0, + "grad_norm": 3.0575679067211943, + "learning_rate": 3.51429545817177e-06, + "loss": 0.6967, + "num_input_tokens_seen": 90619390, + "step": 2770 + }, + { + "epoch": 0.24989854353609595, + "flos": 22278002286240.0, + "grad_norm": 2.8294964410409285, + "learning_rate": 3.5139137616675985e-06, + "loss": 0.7875, + "num_input_tokens_seen": 90647425, + "step": 2771 + }, + { + "epoch": 0.2499887270595662, + "flos": 25442168062080.0, + "grad_norm": 1.9974490228672288, + "learning_rate": 3.513531935987957e-06, + "loss": 0.7214, + "num_input_tokens_seen": 90677430, + "step": 2772 + }, + { + "epoch": 0.2500789105830365, + "flos": 20817521750880.0, + "grad_norm": 1.7573845236379315, + "learning_rate": 3.5131499811654253e-06, + "loss": 0.7989, + "num_input_tokens_seen": 90706385, + "step": 2773 + }, + { + "epoch": 0.25016909410650673, + "flos": 25994076323040.0, + "grad_norm": 2.081127894030017, + "learning_rate": 3.512767897232594e-06, + "loss": 0.7454, + "num_input_tokens_seen": 90736525, + "step": 2774 + }, + { + "epoch": 0.250259277629977, + "flos": 19763629950240.0, + "grad_norm": 1.9868719656154026, + "learning_rate": 3.512385684222064e-06, + "loss": 0.7721, + "num_input_tokens_seen": 90762685, + "step": 2775 + }, + { + "epoch": 0.25034946115344725, + "flos": 15027940365120.0, + "grad_norm": 2.0182309124026387, + "learning_rate": 3.512003342166449e-06, + "loss": 0.8748, + "num_input_tokens_seen": 90785090, + "step": 2776 + }, + { + "epoch": 0.25043964467691754, + "flos": 22205249598720.0, + "grad_norm": 1.4685614444553374, + "learning_rate": 3.511620871098371e-06, + "loss": 0.8161, + "num_input_tokens_seen": 90815060, + "step": 2777 + }, + { + "epoch": 0.25052982820038777, + "flos": 22892590229760.0, + "grad_norm": 2.054302455084288, + "learning_rate": 3.511238271050465e-06, + "loss": 0.8094, + "num_input_tokens_seen": 90840700, + "step": 2778 + }, + { + "epoch": 0.25062001172385806, + "flos": 38340208818720.0, + "grad_norm": 1.7121588123450453, + "learning_rate": 3.5108555420553778e-06, + "loss": 0.6931, + "num_input_tokens_seen": 90871820, + "step": 2779 + }, + { + "epoch": 0.2507101952473283, + "flos": 25739237483040.0, + "grad_norm": 2.0234098101161, + "learning_rate": 3.510472684145764e-06, + "loss": 0.5965, + "num_input_tokens_seen": 90902860, + "step": 2780 + }, + { + "epoch": 0.2508003787707986, + "flos": 64279193823840.0, + "grad_norm": 1.8275696000441053, + "learning_rate": 3.5100896973542926e-06, + "loss": 0.6684, + "num_input_tokens_seen": 90937760, + "step": 2781 + }, + { + "epoch": 0.2508905622942688, + "flos": 23334087102720.0, + "grad_norm": 2.2492332211579127, + "learning_rate": 3.509706581713642e-06, + "loss": 0.8151, + "num_input_tokens_seen": 90964245, + "step": 2782 + }, + { + "epoch": 0.2509807458177391, + "flos": 27196558588800.0, + "grad_norm": 1.6103601495942972, + "learning_rate": 3.509323337256501e-06, + "loss": 0.8057, + "num_input_tokens_seen": 90994880, + "step": 2783 + }, + { + "epoch": 0.25107092934120934, + "flos": 19071494420160.0, + "grad_norm": 1.9721792720840043, + "learning_rate": 3.5089399640155703e-06, + "loss": 0.7289, + "num_input_tokens_seen": 91021160, + "step": 2784 + }, + { + "epoch": 0.2511611128646796, + "flos": 23990348080800.0, + "grad_norm": 1.6429427713978662, + "learning_rate": 3.508556462023562e-06, + "loss": 0.7719, + "num_input_tokens_seen": 91049295, + "step": 2785 + }, + { + "epoch": 0.25125129638814986, + "flos": 17935371643200.0, + "grad_norm": 2.119530201625881, + "learning_rate": 3.5081728313131984e-06, + "loss": 0.7618, + "num_input_tokens_seen": 91074775, + "step": 2786 + }, + { + "epoch": 0.25134147991162015, + "flos": 24792411792000.0, + "grad_norm": 2.617350558994617, + "learning_rate": 3.5077890719172125e-06, + "loss": 0.7714, + "num_input_tokens_seen": 91105370, + "step": 2787 + }, + { + "epoch": 0.25143166343509044, + "flos": 25374358952640.0, + "grad_norm": 2.3211894320883575, + "learning_rate": 3.5074051838683497e-06, + "loss": 0.7764, + "num_input_tokens_seen": 91133775, + "step": 2788 + }, + { + "epoch": 0.25152184695856067, + "flos": 17940426730560.0, + "grad_norm": 1.8409752737536667, + "learning_rate": 3.5070211671993643e-06, + "loss": 0.7004, + "num_input_tokens_seen": 91161350, + "step": 2789 + }, + { + "epoch": 0.25161203048203096, + "flos": 20456137177920.0, + "grad_norm": 1.7995666247109834, + "learning_rate": 3.5066370219430238e-06, + "loss": 0.8006, + "num_input_tokens_seen": 91188310, + "step": 2790 + }, + { + "epoch": 0.2517022140055012, + "flos": 21950485098240.0, + "grad_norm": 25.422759713016628, + "learning_rate": 3.5062527481321044e-06, + "loss": 0.7722, + "num_input_tokens_seen": 91218730, + "step": 2791 + }, + { + "epoch": 0.2517923975289715, + "flos": 21840854275200.0, + "grad_norm": 2.34460230271271, + "learning_rate": 3.5058683457993954e-06, + "loss": 0.7921, + "num_input_tokens_seen": 91245970, + "step": 2792 + }, + { + "epoch": 0.2518825810524417, + "flos": 66808168141920.0, + "grad_norm": 0.7622250170543703, + "learning_rate": 3.5054838149776963e-06, + "loss": 0.5354, + "num_input_tokens_seen": 91342655, + "step": 2793 + }, + { + "epoch": 0.251972764575912, + "flos": 22059000828480.0, + "grad_norm": 1.6651531102116068, + "learning_rate": 3.505099155699816e-06, + "loss": 0.8101, + "num_input_tokens_seen": 91372000, + "step": 2794 + }, + { + "epoch": 0.25206294809938223, + "flos": 35684419781280.0, + "grad_norm": 1.6922784895311294, + "learning_rate": 3.5047143679985775e-06, + "loss": 0.6498, + "num_input_tokens_seen": 91403125, + "step": 2795 + }, + { + "epoch": 0.2521531316228525, + "flos": 21475938575040.0, + "grad_norm": 1.580793258984861, + "learning_rate": 3.5043294519068126e-06, + "loss": 0.7931, + "num_input_tokens_seen": 91431860, + "step": 2796 + }, + { + "epoch": 0.25224331514632276, + "flos": 24500025760800.0, + "grad_norm": 1.641809598958286, + "learning_rate": 3.503944407457363e-06, + "loss": 0.6971, + "num_input_tokens_seen": 91459735, + "step": 2797 + }, + { + "epoch": 0.25233349866979304, + "flos": 24026557160640.0, + "grad_norm": 2.8015647228634255, + "learning_rate": 3.5035592346830846e-06, + "loss": 0.7888, + "num_input_tokens_seen": 91491015, + "step": 2798 + }, + { + "epoch": 0.2524236821932633, + "flos": 23624986343520.0, + "grad_norm": 2.4408412432148996, + "learning_rate": 3.503173933616841e-06, + "loss": 0.8479, + "num_input_tokens_seen": 91516510, + "step": 2799 + }, + { + "epoch": 0.25251386571673357, + "flos": 19944898368000.0, + "grad_norm": 4.925625779725802, + "learning_rate": 3.50278850429151e-06, + "loss": 0.8506, + "num_input_tokens_seen": 91543295, + "step": 2800 + }, + { + "epoch": 0.2526040492402038, + "flos": 19946310818880.0, + "grad_norm": 1.932740795930207, + "learning_rate": 3.502402946739977e-06, + "loss": 0.8013, + "num_input_tokens_seen": 91569100, + "step": 2801 + }, + { + "epoch": 0.2526942327636741, + "flos": 20856853090560.0, + "grad_norm": 2.5152650184619003, + "learning_rate": 3.5020172609951405e-06, + "loss": 0.7997, + "num_input_tokens_seen": 91596750, + "step": 2802 + }, + { + "epoch": 0.2527844162871443, + "flos": 27629692265760.0, + "grad_norm": 1.4671984654870946, + "learning_rate": 3.501631447089909e-06, + "loss": 0.793, + "num_input_tokens_seen": 91627960, + "step": 2803 + }, + { + "epoch": 0.2528745998106146, + "flos": 23553088560480.0, + "grad_norm": 6.860421276892225, + "learning_rate": 3.501245505057203e-06, + "loss": 0.7493, + "num_input_tokens_seen": 91656525, + "step": 2804 + }, + { + "epoch": 0.25296478333408484, + "flos": 25228927917120.0, + "grad_norm": 2.343564242313217, + "learning_rate": 3.5008594349299526e-06, + "loss": 0.8502, + "num_input_tokens_seen": 91684255, + "step": 2805 + }, + { + "epoch": 0.25305496685755513, + "flos": 36155137819200.0, + "grad_norm": 1.7668296957667946, + "learning_rate": 3.500473236741099e-06, + "loss": 0.7613, + "num_input_tokens_seen": 91713795, + "step": 2806 + }, + { + "epoch": 0.25314515038102536, + "flos": 20092410910080.0, + "grad_norm": 1.9511430373557137, + "learning_rate": 3.500086910523596e-06, + "loss": 0.7379, + "num_input_tokens_seen": 91741430, + "step": 2807 + }, + { + "epoch": 0.25323533390449565, + "flos": 50945826374400.0, + "grad_norm": 1.4195831608452603, + "learning_rate": 3.499700456310406e-06, + "loss": 0.7289, + "num_input_tokens_seen": 91776220, + "step": 2808 + }, + { + "epoch": 0.2533255174279659, + "flos": 22970286495360.0, + "grad_norm": 1.4343273977720825, + "learning_rate": 3.499313874134504e-06, + "loss": 0.8267, + "num_input_tokens_seen": 91803965, + "step": 2809 + }, + { + "epoch": 0.2534157009514362, + "flos": 22240938301920.0, + "grad_norm": 3.350024886537036, + "learning_rate": 3.498927164028875e-06, + "loss": 0.6488, + "num_input_tokens_seen": 91829950, + "step": 2810 + }, + { + "epoch": 0.25350588447490646, + "flos": 30146592145440.0, + "grad_norm": 1.3114447665958349, + "learning_rate": 3.498540326026515e-06, + "loss": 0.8162, + "num_input_tokens_seen": 91861515, + "step": 2811 + }, + { + "epoch": 0.2535960679983767, + "flos": 24607240549440.0, + "grad_norm": 1.5806741804418696, + "learning_rate": 3.4981533601604323e-06, + "loss": 0.8444, + "num_input_tokens_seen": 91891615, + "step": 2812 + }, + { + "epoch": 0.253686251521847, + "flos": 18051841702080.0, + "grad_norm": 1.902984761801011, + "learning_rate": 3.4977662664636443e-06, + "loss": 0.7886, + "num_input_tokens_seen": 91918580, + "step": 2813 + }, + { + "epoch": 0.2537764350453172, + "flos": 19946199309600.0, + "grad_norm": 1.6342771249206232, + "learning_rate": 3.497379044969179e-06, + "loss": 0.7351, + "num_input_tokens_seen": 91946445, + "step": 2814 + }, + { + "epoch": 0.2538666185687875, + "flos": 23589186131040.0, + "grad_norm": 2.2491068656842037, + "learning_rate": 3.4969916957100777e-06, + "loss": 0.7754, + "num_input_tokens_seen": 91975920, + "step": 2815 + }, + { + "epoch": 0.25395680209225774, + "flos": 28069330650720.0, + "grad_norm": 1.5256048191334504, + "learning_rate": 3.4966042187193905e-06, + "loss": 0.7703, + "num_input_tokens_seen": 92008880, + "step": 2816 + }, + { + "epoch": 0.254046985615728, + "flos": 22309416467040.0, + "grad_norm": 1.877134887913134, + "learning_rate": 3.496216614030179e-06, + "loss": 0.7742, + "num_input_tokens_seen": 92034175, + "step": 2817 + }, + { + "epoch": 0.25413716913919826, + "flos": 20673428826720.0, + "grad_norm": 2.0121800754799377, + "learning_rate": 3.495828881675516e-06, + "loss": 0.67, + "num_input_tokens_seen": 92061150, + "step": 2818 + }, + { + "epoch": 0.25422735266266855, + "flos": 22897273619520.0, + "grad_norm": 2.150853910095249, + "learning_rate": 3.4954410216884845e-06, + "loss": 0.8416, + "num_input_tokens_seen": 92082135, + "step": 2819 + }, + { + "epoch": 0.2543175361861388, + "flos": 21440026853280.0, + "grad_norm": 2.0024733746648056, + "learning_rate": 3.49505303410218e-06, + "loss": 0.7789, + "num_input_tokens_seen": 92110695, + "step": 2820 + }, + { + "epoch": 0.25440771970960907, + "flos": 24244703713920.0, + "grad_norm": 1.681699603350195, + "learning_rate": 3.4946649189497067e-06, + "loss": 0.8037, + "num_input_tokens_seen": 92143230, + "step": 2821 + }, + { + "epoch": 0.2544979032330793, + "flos": 18299729796960.0, + "grad_norm": 5.939547215939199, + "learning_rate": 3.4942766762641805e-06, + "loss": 0.7685, + "num_input_tokens_seen": 92163580, + "step": 2822 + }, + { + "epoch": 0.2545880867565496, + "flos": 21294372799200.0, + "grad_norm": 3.4547582017831706, + "learning_rate": 3.49388830607873e-06, + "loss": 0.7984, + "num_input_tokens_seen": 92191145, + "step": 2823 + }, + { + "epoch": 0.2546782702800198, + "flos": 26032255400160.0, + "grad_norm": 1.344081261190578, + "learning_rate": 3.493499808426491e-06, + "loss": 0.7409, + "num_input_tokens_seen": 92224590, + "step": 2824 + }, + { + "epoch": 0.2547684538034901, + "flos": 24354111518400.0, + "grad_norm": 1.8955246041687654, + "learning_rate": 3.493111183340614e-06, + "loss": 0.8254, + "num_input_tokens_seen": 92251900, + "step": 2825 + }, + { + "epoch": 0.25485863732696035, + "flos": 17247250447200.0, + "grad_norm": 2.2677116519014127, + "learning_rate": 3.4927224308542576e-06, + "loss": 0.6757, + "num_input_tokens_seen": 92277680, + "step": 2826 + }, + { + "epoch": 0.25494882085043064, + "flos": 23408252241120.0, + "grad_norm": 1.671766972839714, + "learning_rate": 3.4923335510005923e-06, + "loss": 0.8589, + "num_input_tokens_seen": 92309305, + "step": 2827 + }, + { + "epoch": 0.25503900437390087, + "flos": 25411422936960.0, + "grad_norm": 1.4527555426716863, + "learning_rate": 3.4919445438128e-06, + "loss": 0.7956, + "num_input_tokens_seen": 92339800, + "step": 2828 + }, + { + "epoch": 0.25512918789737116, + "flos": 23837706111840.0, + "grad_norm": 1.7593341937846607, + "learning_rate": 3.491555409324073e-06, + "loss": 0.8545, + "num_input_tokens_seen": 92368915, + "step": 2829 + }, + { + "epoch": 0.2552193714208414, + "flos": 70942879649280.0, + "grad_norm": 0.6627703942607692, + "learning_rate": 3.4911661475676136e-06, + "loss": 0.5596, + "num_input_tokens_seen": 92459625, + "step": 2830 + }, + { + "epoch": 0.2553095549443117, + "flos": 21181248018720.0, + "grad_norm": 1.5690529549769348, + "learning_rate": 3.490776758576637e-06, + "loss": 0.7556, + "num_input_tokens_seen": 92487875, + "step": 2831 + }, + { + "epoch": 0.2553997384677819, + "flos": 19761659952960.0, + "grad_norm": 1.7786494227569092, + "learning_rate": 3.4903872423843668e-06, + "loss": 0.8248, + "num_input_tokens_seen": 92515595, + "step": 2832 + }, + { + "epoch": 0.2554899219912522, + "flos": 25447557677280.0, + "grad_norm": 4.454537793296213, + "learning_rate": 3.4899975990240396e-06, + "loss": 0.7379, + "num_input_tokens_seen": 92545725, + "step": 2833 + }, + { + "epoch": 0.25558010551472243, + "flos": 19181311092000.0, + "grad_norm": 1.7550139126131097, + "learning_rate": 3.489607828528901e-06, + "loss": 0.7989, + "num_input_tokens_seen": 92571295, + "step": 2834 + }, + { + "epoch": 0.2556702890381927, + "flos": 57027565039200.0, + "grad_norm": 0.6564385960560779, + "learning_rate": 3.4892179309322093e-06, + "loss": 0.6688, + "num_input_tokens_seen": 92663850, + "step": 2835 + }, + { + "epoch": 0.255760472561663, + "flos": 19760135992800.0, + "grad_norm": 2.7043104273627177, + "learning_rate": 3.488827906267232e-06, + "loss": 0.6913, + "num_input_tokens_seen": 92691325, + "step": 2836 + }, + { + "epoch": 0.25585065608513324, + "flos": 25518340367520.0, + "grad_norm": 2.078224696012315, + "learning_rate": 3.4884377545672485e-06, + "loss": 0.8113, + "num_input_tokens_seen": 92719215, + "step": 2837 + }, + { + "epoch": 0.25594083960860353, + "flos": 24463482153120.0, + "grad_norm": 1.6342384646042005, + "learning_rate": 3.4880474758655485e-06, + "loss": 0.7473, + "num_input_tokens_seen": 92746530, + "step": 2838 + }, + { + "epoch": 0.25603102313207377, + "flos": 23509705716960.0, + "grad_norm": 1.8390490846622662, + "learning_rate": 3.487657070195433e-06, + "loss": 0.8398, + "num_input_tokens_seen": 92775195, + "step": 2839 + }, + { + "epoch": 0.25612120665554405, + "flos": 25446293905440.0, + "grad_norm": 1.7389702620812566, + "learning_rate": 3.487266537590213e-06, + "loss": 0.8513, + "num_input_tokens_seen": 92803685, + "step": 2840 + }, + { + "epoch": 0.2562113901790143, + "flos": 39143424792480.0, + "grad_norm": 1.653923127561721, + "learning_rate": 3.4868758780832116e-06, + "loss": 0.6929, + "num_input_tokens_seen": 92837435, + "step": 2841 + }, + { + "epoch": 0.2563015737024846, + "flos": 21439915344000.0, + "grad_norm": 2.080656531780526, + "learning_rate": 3.486485091707762e-06, + "loss": 0.7878, + "num_input_tokens_seen": 92863235, + "step": 2842 + }, + { + "epoch": 0.2563917572259548, + "flos": 24573782031840.0, + "grad_norm": 1.9017625112672387, + "learning_rate": 3.4860941784972077e-06, + "loss": 0.6562, + "num_input_tokens_seen": 92891850, + "step": 2843 + }, + { + "epoch": 0.2564819407494251, + "flos": 15063963596160.0, + "grad_norm": 3.0977566138769457, + "learning_rate": 3.485703138484904e-06, + "loss": 0.8702, + "num_input_tokens_seen": 92916360, + "step": 2844 + }, + { + "epoch": 0.25657212427289533, + "flos": 32588211793920.0, + "grad_norm": 1.796919847011678, + "learning_rate": 3.485311971704216e-06, + "loss": 0.8236, + "num_input_tokens_seen": 92945655, + "step": 2845 + }, + { + "epoch": 0.2566623077963656, + "flos": 23511564204960.0, + "grad_norm": 1.6308053561646878, + "learning_rate": 3.484920678188521e-06, + "loss": 0.8055, + "num_input_tokens_seen": 92974705, + "step": 2846 + }, + { + "epoch": 0.25675249131983585, + "flos": 19655485917600.0, + "grad_norm": 1.9387493412452037, + "learning_rate": 3.4845292579712063e-06, + "loss": 0.803, + "num_input_tokens_seen": 92999945, + "step": 2847 + }, + { + "epoch": 0.25684267484330614, + "flos": 20930126154720.0, + "grad_norm": 1.950416212412583, + "learning_rate": 3.484137711085669e-06, + "loss": 0.6581, + "num_input_tokens_seen": 93027805, + "step": 2848 + }, + { + "epoch": 0.2569328583667764, + "flos": 29602266515520.0, + "grad_norm": 1.533394848930252, + "learning_rate": 3.4837460375653198e-06, + "loss": 0.7594, + "num_input_tokens_seen": 93062005, + "step": 2849 + }, + { + "epoch": 0.25702304189024666, + "flos": 24682372101600.0, + "grad_norm": 2.020411050727997, + "learning_rate": 3.483354237443576e-06, + "loss": 0.808, + "num_input_tokens_seen": 93090135, + "step": 2850 + }, + { + "epoch": 0.2571132254137169, + "flos": 32221809303360.0, + "grad_norm": 1.5410802172527198, + "learning_rate": 3.48296231075387e-06, + "loss": 0.7187, + "num_input_tokens_seen": 93122275, + "step": 2851 + }, + { + "epoch": 0.2572034089371872, + "flos": 25482019778400.0, + "grad_norm": 1.8412264348093428, + "learning_rate": 3.4825702575296433e-06, + "loss": 0.7933, + "num_input_tokens_seen": 93151675, + "step": 2852 + }, + { + "epoch": 0.2572935924606574, + "flos": 67474433519040.0, + "grad_norm": 0.6967775789940804, + "learning_rate": 3.482178077804347e-06, + "loss": 0.613, + "num_input_tokens_seen": 93239075, + "step": 2853 + }, + { + "epoch": 0.2573837759841277, + "flos": 43117125401280.0, + "grad_norm": 1.6958095683214252, + "learning_rate": 3.4817857716114443e-06, + "loss": 0.739, + "num_input_tokens_seen": 93271960, + "step": 2854 + }, + { + "epoch": 0.25747395950759794, + "flos": 39510087471360.0, + "grad_norm": 2.0460207954908607, + "learning_rate": 3.4813933389844094e-06, + "loss": 0.7278, + "num_input_tokens_seen": 93303305, + "step": 2855 + }, + { + "epoch": 0.2575641430310682, + "flos": 36996495700320.0, + "grad_norm": 1.7739661684906685, + "learning_rate": 3.4810007799567264e-06, + "loss": 0.7797, + "num_input_tokens_seen": 93336355, + "step": 2856 + }, + { + "epoch": 0.25765432655453846, + "flos": 17716147166880.0, + "grad_norm": 2.2323319491265874, + "learning_rate": 3.480608094561891e-06, + "loss": 0.7913, + "num_input_tokens_seen": 93360125, + "step": 2857 + }, + { + "epoch": 0.25774451007800875, + "flos": 29670372983040.0, + "grad_norm": 1.7830946221824246, + "learning_rate": 3.4802152828334083e-06, + "loss": 0.7965, + "num_input_tokens_seen": 93390600, + "step": 2858 + }, + { + "epoch": 0.25783469360147904, + "flos": 27159494604480.0, + "grad_norm": 1.8181931105176339, + "learning_rate": 3.479822344804796e-06, + "loss": 0.7181, + "num_input_tokens_seen": 93420420, + "step": 2859 + }, + { + "epoch": 0.25792487712494927, + "flos": 19030750629600.0, + "grad_norm": 1.8000309305712454, + "learning_rate": 3.479429280509582e-06, + "loss": 0.8313, + "num_input_tokens_seen": 93446805, + "step": 2860 + }, + { + "epoch": 0.25801506064841956, + "flos": 21075780208800.0, + "grad_norm": 1.905263551718651, + "learning_rate": 3.4790360899813038e-06, + "loss": 0.8179, + "num_input_tokens_seen": 93471980, + "step": 2861 + }, + { + "epoch": 0.2581052441718898, + "flos": 16011012305760.0, + "grad_norm": 4.757857399213432, + "learning_rate": 3.4786427732535115e-06, + "loss": 0.7089, + "num_input_tokens_seen": 93499720, + "step": 2862 + }, + { + "epoch": 0.2581954276953601, + "flos": 20818153636800.0, + "grad_norm": 1.598742724201507, + "learning_rate": 3.478249330359764e-06, + "loss": 0.8199, + "num_input_tokens_seen": 93527250, + "step": 2863 + }, + { + "epoch": 0.2582856112188303, + "flos": 21622298854560.0, + "grad_norm": 2.151807105427837, + "learning_rate": 3.4778557613336333e-06, + "loss": 0.7061, + "num_input_tokens_seen": 93555270, + "step": 2864 + }, + { + "epoch": 0.2583757947423006, + "flos": 24464262718080.0, + "grad_norm": 1.7939636528560312, + "learning_rate": 3.4774620662087004e-06, + "loss": 0.7914, + "num_input_tokens_seen": 93586185, + "step": 2865 + }, + { + "epoch": 0.25846597826577083, + "flos": 20819528917920.0, + "grad_norm": 2.4091382218323028, + "learning_rate": 3.477068245018557e-06, + "loss": 0.7769, + "num_input_tokens_seen": 93613090, + "step": 2866 + }, + { + "epoch": 0.2585561617892411, + "flos": 21731446470720.0, + "grad_norm": 1.8024766599331825, + "learning_rate": 3.476674297796807e-06, + "loss": 0.781, + "num_input_tokens_seen": 93640235, + "step": 2867 + }, + { + "epoch": 0.25864634531271136, + "flos": 24682669459680.0, + "grad_norm": 1.812486188112447, + "learning_rate": 3.4762802245770627e-06, + "loss": 0.8146, + "num_input_tokens_seen": 93668900, + "step": 2868 + }, + { + "epoch": 0.25873652883618165, + "flos": 23698928463360.0, + "grad_norm": 2.0941357613087055, + "learning_rate": 3.4758860253929497e-06, + "loss": 0.7271, + "num_input_tokens_seen": 93698260, + "step": 2869 + }, + { + "epoch": 0.2588267123596519, + "flos": 29747214344160.0, + "grad_norm": 1.6802271486172093, + "learning_rate": 3.4754917002781038e-06, + "loss": 0.796, + "num_input_tokens_seen": 93728690, + "step": 2870 + }, + { + "epoch": 0.25891689588312217, + "flos": 19398788589600.0, + "grad_norm": 5.707736056004146, + "learning_rate": 3.475097249266169e-06, + "loss": 0.8147, + "num_input_tokens_seen": 93757685, + "step": 2871 + }, + { + "epoch": 0.2590070794065924, + "flos": 37761495427200.0, + "grad_norm": 1.7131705147877245, + "learning_rate": 3.4747026723908044e-06, + "loss": 0.7599, + "num_input_tokens_seen": 93788805, + "step": 2872 + }, + { + "epoch": 0.2590972629300627, + "flos": 26030248233120.0, + "grad_norm": 2.6244745182683324, + "learning_rate": 3.474307969685676e-06, + "loss": 0.7561, + "num_input_tokens_seen": 93818280, + "step": 2873 + }, + { + "epoch": 0.2591874464535329, + "flos": 28071151968960.0, + "grad_norm": 3.844514600178588, + "learning_rate": 3.473913141184462e-06, + "loss": 0.7217, + "num_input_tokens_seen": 93847915, + "step": 2874 + }, + { + "epoch": 0.2592776299770032, + "flos": 26868521024160.0, + "grad_norm": 1.7715888113509597, + "learning_rate": 3.4735181869208523e-06, + "loss": 0.6471, + "num_input_tokens_seen": 93878930, + "step": 2875 + }, + { + "epoch": 0.25936781350047344, + "flos": 24099086829600.0, + "grad_norm": 2.1400263708040645, + "learning_rate": 3.473123106928546e-06, + "loss": 0.6546, + "num_input_tokens_seen": 93905685, + "step": 2876 + }, + { + "epoch": 0.25945799702394373, + "flos": 29381889776640.0, + "grad_norm": 1.6252707475911403, + "learning_rate": 3.4727279012412533e-06, + "loss": 0.7315, + "num_input_tokens_seen": 93939020, + "step": 2877 + }, + { + "epoch": 0.25954818054741396, + "flos": 59401226899200.0, + "grad_norm": 0.5891084192728252, + "learning_rate": 3.4723325698926953e-06, + "loss": 0.575, + "num_input_tokens_seen": 94031035, + "step": 2878 + }, + { + "epoch": 0.25963836407088425, + "flos": 24493446713280.0, + "grad_norm": 1.850738986085477, + "learning_rate": 3.4719371129166045e-06, + "loss": 0.7816, + "num_input_tokens_seen": 94060220, + "step": 2879 + }, + { + "epoch": 0.2597285475943545, + "flos": 22496520537120.0, + "grad_norm": 1.8985026156086218, + "learning_rate": 3.471541530346723e-06, + "loss": 0.7008, + "num_input_tokens_seen": 94087880, + "step": 2880 + }, + { + "epoch": 0.2598187311178248, + "flos": 28980727826880.0, + "grad_norm": 2.0429262700232176, + "learning_rate": 3.4711458222168037e-06, + "loss": 0.8437, + "num_input_tokens_seen": 94115180, + "step": 2881 + }, + { + "epoch": 0.259908914641295, + "flos": 23262263659200.0, + "grad_norm": 1.5522411018764102, + "learning_rate": 3.4707499885606114e-06, + "loss": 0.8154, + "num_input_tokens_seen": 94141850, + "step": 2882 + }, + { + "epoch": 0.2599990981647653, + "flos": 21148755914880.0, + "grad_norm": 1.733126422869115, + "learning_rate": 3.4703540294119204e-06, + "loss": 0.8127, + "num_input_tokens_seen": 94169550, + "step": 2883 + }, + { + "epoch": 0.2600892816882356, + "flos": 18889556946720.0, + "grad_norm": 3.8859057625708133, + "learning_rate": 3.4699579448045163e-06, + "loss": 0.8241, + "num_input_tokens_seen": 94197945, + "step": 2884 + }, + { + "epoch": 0.2601794652117058, + "flos": 24134552514240.0, + "grad_norm": 1.6355512365224267, + "learning_rate": 3.4695617347721947e-06, + "loss": 0.6637, + "num_input_tokens_seen": 94229940, + "step": 2885 + }, + { + "epoch": 0.2602696487351761, + "flos": 20673057129120.0, + "grad_norm": 1.9562559133812008, + "learning_rate": 3.469165399348763e-06, + "loss": 0.7461, + "num_input_tokens_seen": 94257335, + "step": 2886 + }, + { + "epoch": 0.26035983225864634, + "flos": 21688732682880.0, + "grad_norm": 1.966320294519911, + "learning_rate": 3.4687689385680384e-06, + "loss": 0.7549, + "num_input_tokens_seen": 94284290, + "step": 2887 + }, + { + "epoch": 0.26045001578211663, + "flos": 27924308482560.0, + "grad_norm": 1.5173823599803031, + "learning_rate": 3.4683723524638494e-06, + "loss": 0.7557, + "num_input_tokens_seen": 94316830, + "step": 2888 + }, + { + "epoch": 0.26054019930558686, + "flos": 16667310453600.0, + "grad_norm": 2.5769915314351306, + "learning_rate": 3.4679756410700354e-06, + "loss": 0.8454, + "num_input_tokens_seen": 94344400, + "step": 2889 + }, + { + "epoch": 0.26063038282905715, + "flos": 12987445496640.0, + "grad_norm": 2.1179352175465325, + "learning_rate": 3.4675788044204445e-06, + "loss": 0.7763, + "num_input_tokens_seen": 94370140, + "step": 2890 + }, + { + "epoch": 0.2607205663525274, + "flos": 30146703654720.0, + "grad_norm": 3.0055612833936465, + "learning_rate": 3.467181842548938e-06, + "loss": 0.7067, + "num_input_tokens_seen": 94400035, + "step": 2891 + }, + { + "epoch": 0.26081074987599767, + "flos": 24934460379360.0, + "grad_norm": 1.5822161531123609, + "learning_rate": 3.466784755489387e-06, + "loss": 0.7726, + "num_input_tokens_seen": 94429995, + "step": 2892 + }, + { + "epoch": 0.2609009333994679, + "flos": 20346803712960.0, + "grad_norm": 1.8280631160858294, + "learning_rate": 3.4663875432756726e-06, + "loss": 0.8929, + "num_input_tokens_seen": 94457245, + "step": 2893 + }, + { + "epoch": 0.2609911169229382, + "flos": 19326779297280.0, + "grad_norm": 2.0612444668539998, + "learning_rate": 3.465990205941687e-06, + "loss": 0.9435, + "num_input_tokens_seen": 94486170, + "step": 2894 + }, + { + "epoch": 0.2610813004464084, + "flos": 16735379751360.0, + "grad_norm": 1.9699928962269617, + "learning_rate": 3.465592743521335e-06, + "loss": 0.7626, + "num_input_tokens_seen": 94512780, + "step": 2895 + }, + { + "epoch": 0.2611714839698787, + "flos": 26686249022880.0, + "grad_norm": 1.9794948372731402, + "learning_rate": 3.465195156048528e-06, + "loss": 0.8252, + "num_input_tokens_seen": 94542540, + "step": 2896 + }, + { + "epoch": 0.26126166749334895, + "flos": 69459422468640.0, + "grad_norm": 0.6293566258086623, + "learning_rate": 3.464797443557191e-06, + "loss": 0.5728, + "num_input_tokens_seen": 94638505, + "step": 2897 + }, + { + "epoch": 0.26135185101681924, + "flos": 34008208727040.0, + "grad_norm": 1.7202495622339122, + "learning_rate": 3.46439960608126e-06, + "loss": 0.7507, + "num_input_tokens_seen": 94671495, + "step": 2898 + }, + { + "epoch": 0.26144203454028947, + "flos": 37579409274720.0, + "grad_norm": 1.6235301632911947, + "learning_rate": 3.4640016436546797e-06, + "loss": 0.7121, + "num_input_tokens_seen": 94703685, + "step": 2899 + }, + { + "epoch": 0.26153221806375976, + "flos": 16229827914720.0, + "grad_norm": 1.6852878623504044, + "learning_rate": 3.4636035563114065e-06, + "loss": 0.7522, + "num_input_tokens_seen": 94729510, + "step": 2900 + }, + { + "epoch": 0.26162240158723, + "flos": 25484398643040.0, + "grad_norm": 1.5220062575242965, + "learning_rate": 3.4632053440854085e-06, + "loss": 0.7028, + "num_input_tokens_seen": 94760240, + "step": 2901 + }, + { + "epoch": 0.2617125851107003, + "flos": 18525235962720.0, + "grad_norm": 2.0616615209409885, + "learning_rate": 3.462807007010662e-06, + "loss": 0.8315, + "num_input_tokens_seen": 94787510, + "step": 2902 + }, + { + "epoch": 0.2618027686341705, + "flos": 18525867848640.0, + "grad_norm": 2.069804924697023, + "learning_rate": 3.462408545121155e-06, + "loss": 0.8021, + "num_input_tokens_seen": 94814630, + "step": 2903 + }, + { + "epoch": 0.2618929521576408, + "flos": 12259026547200.0, + "grad_norm": 2.0072608516474375, + "learning_rate": 3.4620099584508883e-06, + "loss": 0.8366, + "num_input_tokens_seen": 94840650, + "step": 2904 + }, + { + "epoch": 0.26198313568111103, + "flos": 17722094328480.0, + "grad_norm": 1.60050058282748, + "learning_rate": 3.46161124703387e-06, + "loss": 0.7341, + "num_input_tokens_seen": 94869410, + "step": 2905 + }, + { + "epoch": 0.2620733192045813, + "flos": 21731000433600.0, + "grad_norm": 2.0536725108224445, + "learning_rate": 3.461212410904122e-06, + "loss": 0.7694, + "num_input_tokens_seen": 94898550, + "step": 2906 + }, + { + "epoch": 0.2621635027280516, + "flos": 18051284155680.0, + "grad_norm": 2.6335635599665226, + "learning_rate": 3.4608134500956726e-06, + "loss": 0.8321, + "num_input_tokens_seen": 94923725, + "step": 2907 + }, + { + "epoch": 0.26225368625152184, + "flos": 29670670341120.0, + "grad_norm": 1.4333688835052796, + "learning_rate": 3.4604143646425655e-06, + "loss": 0.7639, + "num_input_tokens_seen": 94954790, + "step": 2908 + }, + { + "epoch": 0.26234386977499213, + "flos": 26139581698080.0, + "grad_norm": 1.4305197372808247, + "learning_rate": 3.460015154578852e-06, + "loss": 0.8465, + "num_input_tokens_seen": 94987670, + "step": 2909 + }, + { + "epoch": 0.26243405329846237, + "flos": 24353888499840.0, + "grad_norm": 1.9947573971771955, + "learning_rate": 3.459615819938595e-06, + "loss": 0.7508, + "num_input_tokens_seen": 95019200, + "step": 2910 + }, + { + "epoch": 0.26252423682193265, + "flos": 14582280479040.0, + "grad_norm": 1.9908378068126664, + "learning_rate": 3.4592163607558684e-06, + "loss": 0.7393, + "num_input_tokens_seen": 95045070, + "step": 2911 + }, + { + "epoch": 0.2626144203454029, + "flos": 22201160925120.0, + "grad_norm": 1.5067647347281234, + "learning_rate": 3.4588167770647553e-06, + "loss": 0.7778, + "num_input_tokens_seen": 95073140, + "step": 2912 + }, + { + "epoch": 0.2627046038688732, + "flos": 20638186160640.0, + "grad_norm": 2.278998112649709, + "learning_rate": 3.458417068899351e-06, + "loss": 0.8706, + "num_input_tokens_seen": 95098975, + "step": 2913 + }, + { + "epoch": 0.2627947873923434, + "flos": 24389019656640.0, + "grad_norm": 1.7122830190818086, + "learning_rate": 3.4580172362937612e-06, + "loss": 0.7471, + "num_input_tokens_seen": 95127680, + "step": 2914 + }, + { + "epoch": 0.2628849709158137, + "flos": 25336402894080.0, + "grad_norm": 1.551569663777481, + "learning_rate": 3.457617279282101e-06, + "loss": 0.743, + "num_input_tokens_seen": 95157850, + "step": 2915 + }, + { + "epoch": 0.26297515443928393, + "flos": 19799504502240.0, + "grad_norm": 2.5341369742908086, + "learning_rate": 3.4572171978984975e-06, + "loss": 0.7571, + "num_input_tokens_seen": 95183080, + "step": 2916 + }, + { + "epoch": 0.2630653379627542, + "flos": 20637479935200.0, + "grad_norm": 2.6607451158238766, + "learning_rate": 3.456816992177088e-06, + "loss": 0.7757, + "num_input_tokens_seen": 95212360, + "step": 2917 + }, + { + "epoch": 0.26315552148622445, + "flos": 22093797457440.0, + "grad_norm": 1.5424145439625645, + "learning_rate": 3.4564166621520193e-06, + "loss": 0.746, + "num_input_tokens_seen": 95243420, + "step": 2918 + }, + { + "epoch": 0.26324570500969474, + "flos": 21618507539040.0, + "grad_norm": 1.739032300222466, + "learning_rate": 3.4560162078574507e-06, + "loss": 0.7685, + "num_input_tokens_seen": 95271640, + "step": 2919 + }, + { + "epoch": 0.263335888533165, + "flos": 21547985037120.0, + "grad_norm": 1.7190952301099458, + "learning_rate": 3.455615629327551e-06, + "loss": 0.7778, + "num_input_tokens_seen": 95299195, + "step": 2920 + }, + { + "epoch": 0.26342607205663526, + "flos": 22788088833600.0, + "grad_norm": 1.8902609608643581, + "learning_rate": 3.4552149265964994e-06, + "loss": 0.7501, + "num_input_tokens_seen": 95327750, + "step": 2921 + }, + { + "epoch": 0.2635162555801055, + "flos": 30728613645600.0, + "grad_norm": 1.7368080325564235, + "learning_rate": 3.4548140996984866e-06, + "loss": 0.7274, + "num_input_tokens_seen": 95358900, + "step": 2922 + }, + { + "epoch": 0.2636064391035758, + "flos": 30256111459200.0, + "grad_norm": 1.5227332419145168, + "learning_rate": 3.4544131486677124e-06, + "loss": 0.7963, + "num_input_tokens_seen": 95389540, + "step": 2923 + }, + { + "epoch": 0.263696622627046, + "flos": 52987845203040.0, + "grad_norm": 1.6142248760724156, + "learning_rate": 3.454012073538389e-06, + "loss": 0.7167, + "num_input_tokens_seen": 95424305, + "step": 2924 + }, + { + "epoch": 0.2637868061505163, + "flos": 20383607508960.0, + "grad_norm": 1.5511986005138363, + "learning_rate": 3.453610874344738e-06, + "loss": 0.7834, + "num_input_tokens_seen": 95452405, + "step": 2925 + }, + { + "epoch": 0.26387698967398654, + "flos": 18598694875680.0, + "grad_norm": 1.684946034866214, + "learning_rate": 3.453209551120993e-06, + "loss": 0.8179, + "num_input_tokens_seen": 95480965, + "step": 2926 + }, + { + "epoch": 0.26396717319745683, + "flos": 20347175410560.0, + "grad_norm": 2.7021019383252955, + "learning_rate": 3.452808103901395e-06, + "loss": 0.7469, + "num_input_tokens_seen": 95510195, + "step": 2927 + }, + { + "epoch": 0.26405735672092706, + "flos": 26540929496640.0, + "grad_norm": 2.0082857635478195, + "learning_rate": 3.4524065327202e-06, + "loss": 0.8505, + "num_input_tokens_seen": 95537205, + "step": 2928 + }, + { + "epoch": 0.26414754024439735, + "flos": 66422697564480.0, + "grad_norm": 0.6388440148055818, + "learning_rate": 3.4520048376116702e-06, + "loss": 0.5252, + "num_input_tokens_seen": 95628895, + "step": 2929 + }, + { + "epoch": 0.26423772376786764, + "flos": 18051284155680.0, + "grad_norm": 1.460168401464494, + "learning_rate": 3.4516030186100817e-06, + "loss": 0.8123, + "num_input_tokens_seen": 95657130, + "step": 2930 + }, + { + "epoch": 0.26432790729133787, + "flos": 40749485042400.0, + "grad_norm": 1.49533810967306, + "learning_rate": 3.4512010757497197e-06, + "loss": 0.6921, + "num_input_tokens_seen": 95687670, + "step": 2931 + }, + { + "epoch": 0.26441809081480816, + "flos": 26576692539360.0, + "grad_norm": 1.7672435127643256, + "learning_rate": 3.4507990090648804e-06, + "loss": 0.694, + "num_input_tokens_seen": 95718025, + "step": 2932 + }, + { + "epoch": 0.2645082743382784, + "flos": 27377938515840.0, + "grad_norm": 1.7367534652857697, + "learning_rate": 3.4503968185898696e-06, + "loss": 0.7943, + "num_input_tokens_seen": 95744675, + "step": 2933 + }, + { + "epoch": 0.2645984578617487, + "flos": 18088013612160.0, + "grad_norm": 1.7039110662015338, + "learning_rate": 3.4499945043590047e-06, + "loss": 0.7807, + "num_input_tokens_seen": 95770025, + "step": 2934 + }, + { + "epoch": 0.2646886413852189, + "flos": 32188016257920.0, + "grad_norm": 1.6271770272852921, + "learning_rate": 3.4495920664066137e-06, + "loss": 0.7887, + "num_input_tokens_seen": 95798340, + "step": 2935 + }, + { + "epoch": 0.2647788249086892, + "flos": 26867963477760.0, + "grad_norm": 1.975443225331515, + "learning_rate": 3.449189504767035e-06, + "loss": 0.7613, + "num_input_tokens_seen": 95824005, + "step": 2936 + }, + { + "epoch": 0.26486900843215944, + "flos": 30511693694400.0, + "grad_norm": 2.01897170322762, + "learning_rate": 3.4487868194746163e-06, + "loss": 0.7998, + "num_input_tokens_seen": 95852670, + "step": 2937 + }, + { + "epoch": 0.2649591919556297, + "flos": 19217594511360.0, + "grad_norm": 1.8551207527280127, + "learning_rate": 3.4483840105637175e-06, + "loss": 0.8178, + "num_input_tokens_seen": 95880710, + "step": 2938 + }, + { + "epoch": 0.26504937547909996, + "flos": 22970063476800.0, + "grad_norm": 1.9124429507255016, + "learning_rate": 3.4479810780687097e-06, + "loss": 0.7605, + "num_input_tokens_seen": 95909205, + "step": 2939 + }, + { + "epoch": 0.26513955900257025, + "flos": 27961855673760.0, + "grad_norm": 1.580217088923614, + "learning_rate": 3.4475780220239714e-06, + "loss": 0.765, + "num_input_tokens_seen": 95942535, + "step": 2940 + }, + { + "epoch": 0.2652297425260405, + "flos": 22278262474560.0, + "grad_norm": 1.3595529637095005, + "learning_rate": 3.4471748424638948e-06, + "loss": 0.7663, + "num_input_tokens_seen": 95973150, + "step": 2941 + }, + { + "epoch": 0.26531992604951077, + "flos": 57769327932480.0, + "grad_norm": 0.5987592255632203, + "learning_rate": 3.4467715394228803e-06, + "loss": 0.5925, + "num_input_tokens_seen": 96069910, + "step": 2942 + }, + { + "epoch": 0.265410109572981, + "flos": 26287131409920.0, + "grad_norm": 1.5036876172755764, + "learning_rate": 3.4463681129353413e-06, + "loss": 0.7764, + "num_input_tokens_seen": 96100885, + "step": 2943 + }, + { + "epoch": 0.2655002930964513, + "flos": 21404300980320.0, + "grad_norm": 1.7791774135982694, + "learning_rate": 3.4459645630357e-06, + "loss": 0.7688, + "num_input_tokens_seen": 96131525, + "step": 2944 + }, + { + "epoch": 0.2655904766199215, + "flos": 16485224301120.0, + "grad_norm": 2.046985652349643, + "learning_rate": 3.4455608897583884e-06, + "loss": 0.8093, + "num_input_tokens_seen": 96156885, + "step": 2945 + }, + { + "epoch": 0.2656806601433918, + "flos": 19072349324640.0, + "grad_norm": 2.757037867178452, + "learning_rate": 3.4451570931378514e-06, + "loss": 0.775, + "num_input_tokens_seen": 96183875, + "step": 2946 + }, + { + "epoch": 0.26577084366686204, + "flos": 28798492995360.0, + "grad_norm": 1.3874900284829832, + "learning_rate": 3.444753173208543e-06, + "loss": 0.7836, + "num_input_tokens_seen": 96214905, + "step": 2947 + }, + { + "epoch": 0.26586102719033233, + "flos": 19691063111520.0, + "grad_norm": 1.8942581451832963, + "learning_rate": 3.444349130004927e-06, + "loss": 0.7694, + "num_input_tokens_seen": 96242550, + "step": 2948 + }, + { + "epoch": 0.26595121071380257, + "flos": 26206870430880.0, + "grad_norm": 1.8578381638482797, + "learning_rate": 3.4439449635614794e-06, + "loss": 0.6963, + "num_input_tokens_seen": 96269255, + "step": 2949 + }, + { + "epoch": 0.26604139423727285, + "flos": 21622930740480.0, + "grad_norm": 2.195305114950879, + "learning_rate": 3.4435406739126854e-06, + "loss": 0.8033, + "num_input_tokens_seen": 96297610, + "step": 2950 + }, + { + "epoch": 0.2661315777607431, + "flos": 14627261622240.0, + "grad_norm": 2.6427149881698386, + "learning_rate": 3.443136261093042e-06, + "loss": 0.5938, + "num_input_tokens_seen": 96323060, + "step": 2951 + }, + { + "epoch": 0.2662217612842134, + "flos": 27665715496800.0, + "grad_norm": 1.6804578567073782, + "learning_rate": 3.4427317251370553e-06, + "loss": 0.7195, + "num_input_tokens_seen": 96352680, + "step": 2952 + }, + { + "epoch": 0.2663119448076836, + "flos": 29083073376960.0, + "grad_norm": 1.8585010686650152, + "learning_rate": 3.4423270660792422e-06, + "loss": 0.6866, + "num_input_tokens_seen": 96382320, + "step": 2953 + }, + { + "epoch": 0.2664021283311539, + "flos": 24973828888800.0, + "grad_norm": 1.7215445255013566, + "learning_rate": 3.4419222839541314e-06, + "loss": 0.6634, + "num_input_tokens_seen": 96413010, + "step": 2954 + }, + { + "epoch": 0.2664923118546242, + "flos": 29455200010560.0, + "grad_norm": 1.5135579909275478, + "learning_rate": 3.4415173787962607e-06, + "loss": 0.7462, + "num_input_tokens_seen": 96444640, + "step": 2955 + }, + { + "epoch": 0.2665824953780944, + "flos": 32406162811200.0, + "grad_norm": 10.654772882641096, + "learning_rate": 3.4411123506401783e-06, + "loss": 0.6373, + "num_input_tokens_seen": 96473360, + "step": 2956 + }, + { + "epoch": 0.2666726789015647, + "flos": 39728271194400.0, + "grad_norm": 3.7124543312423466, + "learning_rate": 3.440707199520444e-06, + "loss": 0.6279, + "num_input_tokens_seen": 96504205, + "step": 2957 + }, + { + "epoch": 0.26676286242503494, + "flos": 23589260470560.0, + "grad_norm": 1.9713407757067862, + "learning_rate": 3.440301925471628e-06, + "loss": 0.7339, + "num_input_tokens_seen": 96533685, + "step": 2958 + }, + { + "epoch": 0.26685304594850523, + "flos": 29925174653280.0, + "grad_norm": 1.5561621491787876, + "learning_rate": 3.43989652852831e-06, + "loss": 0.8094, + "num_input_tokens_seen": 96564435, + "step": 2959 + }, + { + "epoch": 0.26694322947197546, + "flos": 31459337120160.0, + "grad_norm": 1.8036805353251286, + "learning_rate": 3.4394910087250804e-06, + "loss": 0.8341, + "num_input_tokens_seen": 96595105, + "step": 2960 + }, + { + "epoch": 0.26703341299544575, + "flos": 20782056066240.0, + "grad_norm": 1.5748125570262184, + "learning_rate": 3.4390853660965405e-06, + "loss": 0.7904, + "num_input_tokens_seen": 96624965, + "step": 2961 + }, + { + "epoch": 0.267123596518916, + "flos": 23844991384800.0, + "grad_norm": 2.537359705462027, + "learning_rate": 3.438679600677302e-06, + "loss": 0.8517, + "num_input_tokens_seen": 96653715, + "step": 2962 + }, + { + "epoch": 0.2672137800423863, + "flos": 28324057981440.0, + "grad_norm": 1.7344989904124186, + "learning_rate": 3.4382737125019874e-06, + "loss": 0.7286, + "num_input_tokens_seen": 96683550, + "step": 2963 + }, + { + "epoch": 0.2673039635658565, + "flos": 21947251329120.0, + "grad_norm": 1.7076962919215186, + "learning_rate": 3.4378677016052294e-06, + "loss": 0.7415, + "num_input_tokens_seen": 96713580, + "step": 2964 + }, + { + "epoch": 0.2673941470893268, + "flos": 19472693539680.0, + "grad_norm": 2.1545064760200656, + "learning_rate": 3.43746156802167e-06, + "loss": 0.7914, + "num_input_tokens_seen": 96740025, + "step": 2965 + }, + { + "epoch": 0.267484330612797, + "flos": 21069052482240.0, + "grad_norm": 1.97565219106188, + "learning_rate": 3.4370553117859643e-06, + "loss": 0.7064, + "num_input_tokens_seen": 96767105, + "step": 2966 + }, + { + "epoch": 0.2675745141362673, + "flos": 24317084703840.0, + "grad_norm": 1.7502081610990545, + "learning_rate": 3.4366489329327754e-06, + "loss": 0.801, + "num_input_tokens_seen": 96795835, + "step": 2967 + }, + { + "epoch": 0.26766469765973755, + "flos": 22634406111360.0, + "grad_norm": 2.8744531194261644, + "learning_rate": 3.4362424314967777e-06, + "loss": 0.7918, + "num_input_tokens_seen": 96823815, + "step": 2968 + }, + { + "epoch": 0.26775488118320784, + "flos": 18342963961440.0, + "grad_norm": 2.0878090909448432, + "learning_rate": 3.4358358075126567e-06, + "loss": 0.7998, + "num_input_tokens_seen": 96850360, + "step": 2969 + }, + { + "epoch": 0.26784506470667807, + "flos": 20930126154720.0, + "grad_norm": 1.8309439419634537, + "learning_rate": 3.4354290610151077e-06, + "loss": 0.7927, + "num_input_tokens_seen": 96880825, + "step": 2970 + }, + { + "epoch": 0.26793524823014836, + "flos": 34736925034560.0, + "grad_norm": 1.3898948295517402, + "learning_rate": 3.4350221920388354e-06, + "loss": 0.7809, + "num_input_tokens_seen": 96915200, + "step": 2971 + }, + { + "epoch": 0.2680254317536186, + "flos": 24421809118560.0, + "grad_norm": 1.5180229016361932, + "learning_rate": 3.4346152006185574e-06, + "loss": 0.8741, + "num_input_tokens_seen": 96943490, + "step": 2972 + }, + { + "epoch": 0.2681156152770889, + "flos": 17468890957920.0, + "grad_norm": 1.8850447215838715, + "learning_rate": 3.4342080867890006e-06, + "loss": 0.875, + "num_input_tokens_seen": 96968625, + "step": 2973 + }, + { + "epoch": 0.2682057988005591, + "flos": 14117026395840.0, + "grad_norm": 2.063480837556208, + "learning_rate": 3.4338008505849016e-06, + "loss": 0.6915, + "num_input_tokens_seen": 96993715, + "step": 2974 + }, + { + "epoch": 0.2682959823240294, + "flos": 27194365572960.0, + "grad_norm": 2.4035063124832288, + "learning_rate": 3.433393492041008e-06, + "loss": 0.7831, + "num_input_tokens_seen": 97024615, + "step": 2975 + }, + { + "epoch": 0.26838616584749964, + "flos": 23876702923680.0, + "grad_norm": 1.9134185330542552, + "learning_rate": 3.432986011192078e-06, + "loss": 0.8218, + "num_input_tokens_seen": 97053470, + "step": 2976 + }, + { + "epoch": 0.2684763493709699, + "flos": 19763852968800.0, + "grad_norm": 1.9456819237145762, + "learning_rate": 3.4325784080728796e-06, + "loss": 0.8418, + "num_input_tokens_seen": 97079800, + "step": 2977 + }, + { + "epoch": 0.2685665328944402, + "flos": 30399275139360.0, + "grad_norm": 1.595741150121638, + "learning_rate": 3.4321706827181926e-06, + "loss": 0.8043, + "num_input_tokens_seen": 97109655, + "step": 2978 + }, + { + "epoch": 0.26865671641791045, + "flos": 21439766664960.0, + "grad_norm": 1.9213013451887364, + "learning_rate": 3.4317628351628064e-06, + "loss": 0.726, + "num_input_tokens_seen": 97138940, + "step": 2979 + }, + { + "epoch": 0.26874689994138073, + "flos": 23909232197280.0, + "grad_norm": 1.427446286715805, + "learning_rate": 3.43135486544152e-06, + "loss": 0.7583, + "num_input_tokens_seen": 97165545, + "step": 2980 + }, + { + "epoch": 0.26883708346485097, + "flos": 25586409665280.0, + "grad_norm": 2.359656108758037, + "learning_rate": 3.4309467735891442e-06, + "loss": 0.8134, + "num_input_tokens_seen": 97194910, + "step": 2981 + }, + { + "epoch": 0.26892726698832126, + "flos": 20346878052480.0, + "grad_norm": 1.7662997011614763, + "learning_rate": 3.4305385596405e-06, + "loss": 0.8603, + "num_input_tokens_seen": 97222345, + "step": 2982 + }, + { + "epoch": 0.2690174505117915, + "flos": 22022791748640.0, + "grad_norm": 1.6430418831750924, + "learning_rate": 3.4301302236304174e-06, + "loss": 0.7393, + "num_input_tokens_seen": 97251825, + "step": 2983 + }, + { + "epoch": 0.2691076340352618, + "flos": 61363616634720.0, + "grad_norm": 0.5988410724352501, + "learning_rate": 3.429721765593739e-06, + "loss": 0.5472, + "num_input_tokens_seen": 97338520, + "step": 2984 + }, + { + "epoch": 0.269197817558732, + "flos": 21984315313440.0, + "grad_norm": 1.5535206465415763, + "learning_rate": 3.4293131855653155e-06, + "loss": 0.8009, + "num_input_tokens_seen": 97367480, + "step": 2985 + }, + { + "epoch": 0.2692880010822023, + "flos": 37543237364640.0, + "grad_norm": 1.7620297853071873, + "learning_rate": 3.4289044835800102e-06, + "loss": 0.7127, + "num_input_tokens_seen": 97398355, + "step": 2986 + }, + { + "epoch": 0.26937818460567253, + "flos": 18488989713120.0, + "grad_norm": 1.5732243049689758, + "learning_rate": 3.4284956596726953e-06, + "loss": 0.7777, + "num_input_tokens_seen": 97425665, + "step": 2987 + }, + { + "epoch": 0.2694683681291428, + "flos": 27777316317120.0, + "grad_norm": 1.9745151938778522, + "learning_rate": 3.4280867138782544e-06, + "loss": 0.768, + "num_input_tokens_seen": 97457125, + "step": 2988 + }, + { + "epoch": 0.26955855165261305, + "flos": 21650627945280.0, + "grad_norm": 2.489466697695377, + "learning_rate": 3.4276776462315803e-06, + "loss": 0.7777, + "num_input_tokens_seen": 97483685, + "step": 2989 + }, + { + "epoch": 0.26964873517608334, + "flos": 18809407476960.0, + "grad_norm": 1.6500268298193017, + "learning_rate": 3.427268456767578e-06, + "loss": 0.7751, + "num_input_tokens_seen": 97509660, + "step": 2990 + }, + { + "epoch": 0.2697389186995536, + "flos": 21069870216960.0, + "grad_norm": 2.219562972586298, + "learning_rate": 3.42685914552116e-06, + "loss": 0.8133, + "num_input_tokens_seen": 97536225, + "step": 2991 + }, + { + "epoch": 0.26982910222302386, + "flos": 22023237785760.0, + "grad_norm": 1.9870878284051816, + "learning_rate": 3.426449712527253e-06, + "loss": 0.749, + "num_input_tokens_seen": 97564175, + "step": 2992 + }, + { + "epoch": 0.2699192857464941, + "flos": 25848719627040.0, + "grad_norm": 5.58799797416328, + "learning_rate": 3.4260401578207904e-06, + "loss": 0.6846, + "num_input_tokens_seen": 97596650, + "step": 2993 + }, + { + "epoch": 0.2700094692699644, + "flos": 32404192813920.0, + "grad_norm": 2.1908371674320435, + "learning_rate": 3.4256304814367185e-06, + "loss": 0.7321, + "num_input_tokens_seen": 97624050, + "step": 2994 + }, + { + "epoch": 0.2700996527934346, + "flos": 24207937087680.0, + "grad_norm": 1.8061113561727018, + "learning_rate": 3.4252206834099936e-06, + "loss": 0.7297, + "num_input_tokens_seen": 97649350, + "step": 2995 + }, + { + "epoch": 0.2701898363169049, + "flos": 26099692812000.0, + "grad_norm": 5.406224245219542, + "learning_rate": 3.424810763775581e-06, + "loss": 0.7699, + "num_input_tokens_seen": 97680335, + "step": 2996 + }, + { + "epoch": 0.27028001984037514, + "flos": 34007911368960.0, + "grad_norm": 1.409405452965973, + "learning_rate": 3.4244007225684587e-06, + "loss": 0.7149, + "num_input_tokens_seen": 97712950, + "step": 2997 + }, + { + "epoch": 0.27037020336384543, + "flos": 22423470491520.0, + "grad_norm": 1.5507066069734967, + "learning_rate": 3.4239905598236115e-06, + "loss": 0.7802, + "num_input_tokens_seen": 97742325, + "step": 2998 + }, + { + "epoch": 0.27046038688731566, + "flos": 22677082729440.0, + "grad_norm": 1.6620425290428544, + "learning_rate": 3.4235802755760386e-06, + "loss": 0.7993, + "num_input_tokens_seen": 97770625, + "step": 2999 + }, + { + "epoch": 0.27055057041078595, + "flos": 64699009259520.0, + "grad_norm": 0.8484574499998787, + "learning_rate": 3.4231698698607464e-06, + "loss": 0.6941, + "num_input_tokens_seen": 97858755, + "step": 3000 + }, + { + "epoch": 0.2706407539342562, + "flos": 25334432896800.0, + "grad_norm": 1.7491510920795188, + "learning_rate": 3.4227593427127543e-06, + "loss": 0.8292, + "num_input_tokens_seen": 97887595, + "step": 3001 + }, + { + "epoch": 0.2707309374577265, + "flos": 35867063480160.0, + "grad_norm": 1.9537309240775453, + "learning_rate": 3.42234869416709e-06, + "loss": 0.745, + "num_input_tokens_seen": 97921260, + "step": 3002 + }, + { + "epoch": 0.27082112098119676, + "flos": 17504728340160.0, + "grad_norm": 1.979957830673605, + "learning_rate": 3.421937924258792e-06, + "loss": 0.7758, + "num_input_tokens_seen": 97946045, + "step": 3003 + }, + { + "epoch": 0.270911304504667, + "flos": 29125973013600.0, + "grad_norm": 1.7956419321304826, + "learning_rate": 3.4215270330229096e-06, + "loss": 0.8154, + "num_input_tokens_seen": 97975800, + "step": 3004 + }, + { + "epoch": 0.2710014880281373, + "flos": 25155915041280.0, + "grad_norm": 2.071651270597683, + "learning_rate": 3.421116020494503e-06, + "loss": 0.7394, + "num_input_tokens_seen": 98003325, + "step": 3005 + }, + { + "epoch": 0.2710916715516075, + "flos": 49272180033600.0, + "grad_norm": 1.3530339076014775, + "learning_rate": 3.420704886708642e-06, + "loss": 0.7207, + "num_input_tokens_seen": 98036540, + "step": 3006 + }, + { + "epoch": 0.2711818550750778, + "flos": 28543988683200.0, + "grad_norm": 1.6544291725988138, + "learning_rate": 3.4202936317004056e-06, + "loss": 0.725, + "num_input_tokens_seen": 98067405, + "step": 3007 + }, + { + "epoch": 0.27127203859854804, + "flos": 20018505960000.0, + "grad_norm": 2.3571329137253865, + "learning_rate": 3.4198822555048856e-06, + "loss": 0.7895, + "num_input_tokens_seen": 98095295, + "step": 3008 + }, + { + "epoch": 0.2713622221220183, + "flos": 13825123571520.0, + "grad_norm": 2.1015074768804705, + "learning_rate": 3.419470758157182e-06, + "loss": 0.8569, + "num_input_tokens_seen": 98119985, + "step": 3009 + }, + { + "epoch": 0.27145240564548856, + "flos": 25405661624160.0, + "grad_norm": 1.7959964096951015, + "learning_rate": 3.4190591396924068e-06, + "loss": 0.7697, + "num_input_tokens_seen": 98147260, + "step": 3010 + }, + { + "epoch": 0.27154258916895885, + "flos": 56116056353760.0, + "grad_norm": 0.6741098229743854, + "learning_rate": 3.418647400145681e-06, + "loss": 0.6155, + "num_input_tokens_seen": 98227750, + "step": 3011 + }, + { + "epoch": 0.2716327726924291, + "flos": 67372050799200.0, + "grad_norm": 0.6984053841189988, + "learning_rate": 3.4182355395521367e-06, + "loss": 0.575, + "num_input_tokens_seen": 98310765, + "step": 3012 + }, + { + "epoch": 0.27172295621589937, + "flos": 28613730620160.0, + "grad_norm": 2.5339872842121904, + "learning_rate": 3.417823557946916e-06, + "loss": 0.7708, + "num_input_tokens_seen": 98339790, + "step": 3013 + }, + { + "epoch": 0.2718131397393696, + "flos": 30509240490240.0, + "grad_norm": 1.6281853047805988, + "learning_rate": 3.417411455365172e-06, + "loss": 0.6726, + "num_input_tokens_seen": 98369350, + "step": 3014 + }, + { + "epoch": 0.2719033232628399, + "flos": 22096176322080.0, + "grad_norm": 2.102108682329008, + "learning_rate": 3.416999231842066e-06, + "loss": 0.6842, + "num_input_tokens_seen": 98398670, + "step": 3015 + }, + { + "epoch": 0.2719935067863101, + "flos": 19363620263040.0, + "grad_norm": 1.9916038705457597, + "learning_rate": 3.416586887412773e-06, + "loss": 0.8246, + "num_input_tokens_seen": 98426275, + "step": 3016 + }, + { + "epoch": 0.2720836903097804, + "flos": 25373764236480.0, + "grad_norm": 2.4434342163235607, + "learning_rate": 3.416174422112476e-06, + "loss": 0.7388, + "num_input_tokens_seen": 98455945, + "step": 3017 + }, + { + "epoch": 0.27217387383325065, + "flos": 67165166683200.0, + "grad_norm": 0.6037705509566017, + "learning_rate": 3.4157618359763687e-06, + "loss": 0.539, + "num_input_tokens_seen": 98553200, + "step": 3018 + }, + { + "epoch": 0.27226405735672093, + "flos": 22635186676320.0, + "grad_norm": 1.6759352535451115, + "learning_rate": 3.4153491290396542e-06, + "loss": 0.6927, + "num_input_tokens_seen": 98581920, + "step": 3019 + }, + { + "epoch": 0.27235424088019117, + "flos": 25590535508640.0, + "grad_norm": 2.248014739840095, + "learning_rate": 3.4149363013375485e-06, + "loss": 0.7919, + "num_input_tokens_seen": 98613155, + "step": 3020 + }, + { + "epoch": 0.27244442440366146, + "flos": 66782446668000.0, + "grad_norm": 0.6954117393395278, + "learning_rate": 3.414523352905276e-06, + "loss": 0.5551, + "num_input_tokens_seen": 98698085, + "step": 3021 + }, + { + "epoch": 0.2725346079271317, + "flos": 32477354368800.0, + "grad_norm": 2.1667545251883786, + "learning_rate": 3.414110283778071e-06, + "loss": 0.7415, + "num_input_tokens_seen": 98728555, + "step": 3022 + }, + { + "epoch": 0.272624791450602, + "flos": 21877174864320.0, + "grad_norm": 2.0826304076396847, + "learning_rate": 3.4136970939911797e-06, + "loss": 0.7183, + "num_input_tokens_seen": 98756010, + "step": 3023 + }, + { + "epoch": 0.2727149749740722, + "flos": 24242250509760.0, + "grad_norm": 4.813508559724498, + "learning_rate": 3.413283783579857e-06, + "loss": 0.7393, + "num_input_tokens_seen": 98783055, + "step": 3024 + }, + { + "epoch": 0.2728051584975425, + "flos": 15465534413280.0, + "grad_norm": 2.8155717817182024, + "learning_rate": 3.412870352579369e-06, + "loss": 0.7076, + "num_input_tokens_seen": 98808295, + "step": 3025 + }, + { + "epoch": 0.2728953420210128, + "flos": 66460727962560.0, + "grad_norm": 0.646395328531818, + "learning_rate": 3.4124568010249915e-06, + "loss": 0.5967, + "num_input_tokens_seen": 98900435, + "step": 3026 + }, + { + "epoch": 0.272985525544483, + "flos": 11566705168320.0, + "grad_norm": 1.7831828689678177, + "learning_rate": 3.4120431289520124e-06, + "loss": 0.7576, + "num_input_tokens_seen": 98927285, + "step": 3027 + }, + { + "epoch": 0.2730757090679533, + "flos": 27997841735040.0, + "grad_norm": 2.2727260459549803, + "learning_rate": 3.4116293363957276e-06, + "loss": 0.7775, + "num_input_tokens_seen": 98956210, + "step": 3028 + }, + { + "epoch": 0.27316589259142354, + "flos": 20928713703840.0, + "grad_norm": 1.9362694308720168, + "learning_rate": 3.4112154233914438e-06, + "loss": 0.7804, + "num_input_tokens_seen": 98982655, + "step": 3029 + }, + { + "epoch": 0.27325607611489383, + "flos": 31783286011200.0, + "grad_norm": 2.5419516690249426, + "learning_rate": 3.410801389974479e-06, + "loss": 0.6771, + "num_input_tokens_seen": 99012105, + "step": 3030 + }, + { + "epoch": 0.27334625963836406, + "flos": 24864198065760.0, + "grad_norm": 1.7952199196579017, + "learning_rate": 3.410387236180161e-06, + "loss": 0.8058, + "num_input_tokens_seen": 99040010, + "step": 3031 + }, + { + "epoch": 0.27343644316183435, + "flos": 33134544590880.0, + "grad_norm": 1.8123937772997212, + "learning_rate": 3.409972962043826e-06, + "loss": 0.6567, + "num_input_tokens_seen": 99071205, + "step": 3032 + }, + { + "epoch": 0.2735266266853046, + "flos": 20310074256480.0, + "grad_norm": 2.3605673095182254, + "learning_rate": 3.4095585676008234e-06, + "loss": 0.8278, + "num_input_tokens_seen": 99098795, + "step": 3033 + }, + { + "epoch": 0.2736168102087749, + "flos": 24056001344160.0, + "grad_norm": 1.9197074017493652, + "learning_rate": 3.4091440528865125e-06, + "loss": 0.7671, + "num_input_tokens_seen": 99127485, + "step": 3034 + }, + { + "epoch": 0.2737069937322451, + "flos": 14591052542400.0, + "grad_norm": 1.7885644211059453, + "learning_rate": 3.4087294179362606e-06, + "loss": 0.7693, + "num_input_tokens_seen": 99153260, + "step": 3035 + }, + { + "epoch": 0.2737971772557154, + "flos": 26026159559520.0, + "grad_norm": 1.9298987639761773, + "learning_rate": 3.4083146627854474e-06, + "loss": 0.8155, + "num_input_tokens_seen": 99181345, + "step": 3036 + }, + { + "epoch": 0.27388736077918563, + "flos": 22824372252960.0, + "grad_norm": 2.4886706180835785, + "learning_rate": 3.4078997874694614e-06, + "loss": 0.6987, + "num_input_tokens_seen": 99210700, + "step": 3037 + }, + { + "epoch": 0.2739775443026559, + "flos": 24828286344000.0, + "grad_norm": 1.9363160361350684, + "learning_rate": 3.407484792023703e-06, + "loss": 0.6624, + "num_input_tokens_seen": 99238555, + "step": 3038 + }, + { + "epoch": 0.27406772782612615, + "flos": 27451397428800.0, + "grad_norm": 2.0960344975158716, + "learning_rate": 3.407069676483581e-06, + "loss": 0.7375, + "num_input_tokens_seen": 99266050, + "step": 3039 + }, + { + "epoch": 0.27415791134959644, + "flos": 28467630528960.0, + "grad_norm": 1.718252630228699, + "learning_rate": 3.406654440884516e-06, + "loss": 0.8055, + "num_input_tokens_seen": 99294945, + "step": 3040 + }, + { + "epoch": 0.2742480948730667, + "flos": 23479852666080.0, + "grad_norm": 1.8517880201818033, + "learning_rate": 3.4062390852619372e-06, + "loss": 0.795, + "num_input_tokens_seen": 99322945, + "step": 3041 + }, + { + "epoch": 0.27433827839653696, + "flos": 21111766270080.0, + "grad_norm": 2.3894851061302336, + "learning_rate": 3.4058236096512867e-06, + "loss": 0.7659, + "num_input_tokens_seen": 99348125, + "step": 3042 + }, + { + "epoch": 0.2744284619200072, + "flos": 20856518562720.0, + "grad_norm": 3.052775923921506, + "learning_rate": 3.405408014088013e-06, + "loss": 0.6729, + "num_input_tokens_seen": 99376095, + "step": 3043 + }, + { + "epoch": 0.2745186454434775, + "flos": 31822803199680.0, + "grad_norm": 2.039311887025792, + "learning_rate": 3.404992298607579e-06, + "loss": 0.658, + "num_input_tokens_seen": 99406100, + "step": 3044 + }, + { + "epoch": 0.2746088289669477, + "flos": 23771532471840.0, + "grad_norm": 2.9283831731918557, + "learning_rate": 3.4045764632454547e-06, + "loss": 0.7482, + "num_input_tokens_seen": 99433470, + "step": 3045 + }, + { + "epoch": 0.274699012490418, + "flos": 27154216498560.0, + "grad_norm": 1.9269800381856368, + "learning_rate": 3.4041605080371223e-06, + "loss": 0.7318, + "num_input_tokens_seen": 99460850, + "step": 3046 + }, + { + "epoch": 0.27478919601388824, + "flos": 17904217650720.0, + "grad_norm": 1.7408671486313825, + "learning_rate": 3.4037444330180726e-06, + "loss": 0.7835, + "num_input_tokens_seen": 99488570, + "step": 3047 + }, + { + "epoch": 0.2748793795373585, + "flos": 15933539058720.0, + "grad_norm": 2.0748307354381934, + "learning_rate": 3.403328238223808e-06, + "loss": 0.7815, + "num_input_tokens_seen": 99515290, + "step": 3048 + }, + { + "epoch": 0.27496956306082876, + "flos": 31603355704800.0, + "grad_norm": 1.9248020863043687, + "learning_rate": 3.4029119236898395e-06, + "loss": 0.7983, + "num_input_tokens_seen": 99541820, + "step": 3049 + }, + { + "epoch": 0.27505974658429905, + "flos": 69029894258400.0, + "grad_norm": 1.035799970746818, + "learning_rate": 3.4024954894516906e-06, + "loss": 0.6079, + "num_input_tokens_seen": 99638890, + "step": 3050 + }, + { + "epoch": 0.27514993010776934, + "flos": 24243811639680.0, + "grad_norm": 5.6399063962920115, + "learning_rate": 3.4020789355448933e-06, + "loss": 0.7732, + "num_input_tokens_seen": 99666515, + "step": 3051 + }, + { + "epoch": 0.27524011363123957, + "flos": 23662161837120.0, + "grad_norm": 3.0007825466591185, + "learning_rate": 3.40166226200499e-06, + "loss": 0.8431, + "num_input_tokens_seen": 99694770, + "step": 3052 + }, + { + "epoch": 0.27533029715470986, + "flos": 25119408603360.0, + "grad_norm": 2.3423428189398767, + "learning_rate": 3.401245468867534e-06, + "loss": 0.7528, + "num_input_tokens_seen": 99723030, + "step": 3053 + }, + { + "epoch": 0.2754204806781801, + "flos": 64015391338080.0, + "grad_norm": 0.7046034422346691, + "learning_rate": 3.400828556168088e-06, + "loss": 0.6322, + "num_input_tokens_seen": 99806175, + "step": 3054 + }, + { + "epoch": 0.2755106642016504, + "flos": 21689513247840.0, + "grad_norm": 3.393341804249001, + "learning_rate": 3.4004115239422255e-06, + "loss": 0.7841, + "num_input_tokens_seen": 99833295, + "step": 3055 + }, + { + "epoch": 0.2756008477251206, + "flos": 20747556795360.0, + "grad_norm": 1.8287252109226824, + "learning_rate": 3.3999943722255305e-06, + "loss": 0.7564, + "num_input_tokens_seen": 99860490, + "step": 3056 + }, + { + "epoch": 0.2756910312485909, + "flos": 48036127740960.0, + "grad_norm": 1.6483439197463898, + "learning_rate": 3.3995771010535955e-06, + "loss": 0.6653, + "num_input_tokens_seen": 99894245, + "step": 3057 + }, + { + "epoch": 0.27578121477206113, + "flos": 63024104880480.0, + "grad_norm": 0.6945117668791859, + "learning_rate": 3.3991597104620253e-06, + "loss": 0.5396, + "num_input_tokens_seen": 99972020, + "step": 3058 + }, + { + "epoch": 0.2758713982955314, + "flos": 28471644863040.0, + "grad_norm": 1.747380953487103, + "learning_rate": 3.398742200486434e-06, + "loss": 0.7507, + "num_input_tokens_seen": 100001885, + "step": 3059 + }, + { + "epoch": 0.27596158181900166, + "flos": 26613682184160.0, + "grad_norm": 2.004578832593406, + "learning_rate": 3.3983245711624453e-06, + "loss": 0.7882, + "num_input_tokens_seen": 100031220, + "step": 3060 + }, + { + "epoch": 0.27605176534247194, + "flos": 70648449281280.0, + "grad_norm": 0.7564835053919244, + "learning_rate": 3.3979068225256946e-06, + "loss": 0.6549, + "num_input_tokens_seen": 100125055, + "step": 3061 + }, + { + "epoch": 0.2761419488659422, + "flos": 21257606172960.0, + "grad_norm": 1.7780734063089934, + "learning_rate": 3.3974889546118246e-06, + "loss": 0.7312, + "num_input_tokens_seen": 100151665, + "step": 3062 + }, + { + "epoch": 0.27623213238941247, + "flos": 16558014158400.0, + "grad_norm": 1.640214051140942, + "learning_rate": 3.3970709674564918e-06, + "loss": 0.7145, + "num_input_tokens_seen": 100179995, + "step": 3063 + }, + { + "epoch": 0.2763223159128827, + "flos": 23006569914720.0, + "grad_norm": 1.6730033418173968, + "learning_rate": 3.3966528610953607e-06, + "loss": 0.7648, + "num_input_tokens_seen": 100210050, + "step": 3064 + }, + { + "epoch": 0.276412499436353, + "flos": 23404052058240.0, + "grad_norm": 2.0986625772231964, + "learning_rate": 3.3962346355641067e-06, + "loss": 0.7774, + "num_input_tokens_seen": 100239565, + "step": 3065 + }, + { + "epoch": 0.2765026829598232, + "flos": 21330804897600.0, + "grad_norm": 1.8583055514096443, + "learning_rate": 3.3958162908984146e-06, + "loss": 0.8036, + "num_input_tokens_seen": 100268050, + "step": 3066 + }, + { + "epoch": 0.2765928664832935, + "flos": 23552977051200.0, + "grad_norm": 1.864266061339955, + "learning_rate": 3.39539782713398e-06, + "loss": 0.763, + "num_input_tokens_seen": 100296585, + "step": 3067 + }, + { + "epoch": 0.27668305000676374, + "flos": 26613124637760.0, + "grad_norm": 2.2826654928655965, + "learning_rate": 3.394979244306509e-06, + "loss": 0.7936, + "num_input_tokens_seen": 100325175, + "step": 3068 + }, + { + "epoch": 0.27677323353023403, + "flos": 25917457980480.0, + "grad_norm": 1.8116863660726907, + "learning_rate": 3.3945605424517166e-06, + "loss": 0.7789, + "num_input_tokens_seen": 100356025, + "step": 3069 + }, + { + "epoch": 0.27686341705370426, + "flos": 22715150297280.0, + "grad_norm": 3.2474460837548222, + "learning_rate": 3.3941417216053294e-06, + "loss": 0.7645, + "num_input_tokens_seen": 100383575, + "step": 3070 + }, + { + "epoch": 0.27695360057717455, + "flos": 20784286251840.0, + "grad_norm": 1.9906177394860294, + "learning_rate": 3.3937227818030835e-06, + "loss": 0.7329, + "num_input_tokens_seen": 100410890, + "step": 3071 + }, + { + "epoch": 0.2770437841006448, + "flos": 19836642826080.0, + "grad_norm": 1.911546571723514, + "learning_rate": 3.393303723080725e-06, + "loss": 0.7324, + "num_input_tokens_seen": 100438800, + "step": 3072 + }, + { + "epoch": 0.2771339676241151, + "flos": 25411460106720.0, + "grad_norm": 1.6669546434837412, + "learning_rate": 3.3928845454740097e-06, + "loss": 0.7775, + "num_input_tokens_seen": 100467490, + "step": 3073 + }, + { + "epoch": 0.27722415114758536, + "flos": 21513039729120.0, + "grad_norm": 1.6758962027152489, + "learning_rate": 3.392465249018705e-06, + "loss": 0.7538, + "num_input_tokens_seen": 100494565, + "step": 3074 + }, + { + "epoch": 0.2773143346710556, + "flos": 19796233563360.0, + "grad_norm": 1.701770558873739, + "learning_rate": 3.3920458337505872e-06, + "loss": 0.7649, + "num_input_tokens_seen": 100521835, + "step": 3075 + }, + { + "epoch": 0.2774045181945259, + "flos": 21657430011360.0, + "grad_norm": 4.607685154357821, + "learning_rate": 3.391626299705443e-06, + "loss": 0.6072, + "num_input_tokens_seen": 100546885, + "step": 3076 + }, + { + "epoch": 0.2774947017179961, + "flos": 22569905110560.0, + "grad_norm": 1.732146610036838, + "learning_rate": 3.39120664691907e-06, + "loss": 0.7774, + "num_input_tokens_seen": 100576105, + "step": 3077 + }, + { + "epoch": 0.2775848852414664, + "flos": 25303576262400.0, + "grad_norm": 2.599308263018742, + "learning_rate": 3.390786875427275e-06, + "loss": 0.7471, + "num_input_tokens_seen": 100606370, + "step": 3078 + }, + { + "epoch": 0.27767506876493664, + "flos": 29781341917440.0, + "grad_norm": 1.855166697762281, + "learning_rate": 3.390366985265875e-06, + "loss": 0.7648, + "num_input_tokens_seen": 100635850, + "step": 3079 + }, + { + "epoch": 0.2777652522884069, + "flos": 24135630437280.0, + "grad_norm": 1.6710883455692909, + "learning_rate": 3.389946976470697e-06, + "loss": 0.7223, + "num_input_tokens_seen": 100667260, + "step": 3080 + }, + { + "epoch": 0.27785543581187716, + "flos": 41441620572480.0, + "grad_norm": 2.024926657584766, + "learning_rate": 3.3895268490775787e-06, + "loss": 0.8615, + "num_input_tokens_seen": 100698050, + "step": 3081 + }, + { + "epoch": 0.27794561933534745, + "flos": 17249369123520.0, + "grad_norm": 4.416819563804266, + "learning_rate": 3.3891066031223685e-06, + "loss": 0.7229, + "num_input_tokens_seen": 100726775, + "step": 3082 + }, + { + "epoch": 0.2780358028588177, + "flos": 21440472890400.0, + "grad_norm": 1.8497458773135373, + "learning_rate": 3.3886862386409237e-06, + "loss": 0.8015, + "num_input_tokens_seen": 100753505, + "step": 3083 + }, + { + "epoch": 0.27812598638228797, + "flos": 15173445740160.0, + "grad_norm": 2.1069073972697914, + "learning_rate": 3.388265755669111e-06, + "loss": 0.7328, + "num_input_tokens_seen": 100781320, + "step": 3084 + }, + { + "epoch": 0.2782161699057582, + "flos": 22715521994880.0, + "grad_norm": 1.9347690285726487, + "learning_rate": 3.3878451542428093e-06, + "loss": 0.8353, + "num_input_tokens_seen": 100809690, + "step": 3085 + }, + { + "epoch": 0.2783063534292285, + "flos": 20638743707040.0, + "grad_norm": 1.741631731416078, + "learning_rate": 3.387424434397907e-06, + "loss": 0.7524, + "num_input_tokens_seen": 100839035, + "step": 3086 + }, + { + "epoch": 0.2783965369526987, + "flos": 20234310818400.0, + "grad_norm": 1.6268576494391562, + "learning_rate": 3.3870035961703013e-06, + "loss": 0.6772, + "num_input_tokens_seen": 100868165, + "step": 3087 + }, + { + "epoch": 0.278486720476169, + "flos": 20961912033120.0, + "grad_norm": 4.900044398203971, + "learning_rate": 3.3865826395959014e-06, + "loss": 0.8914, + "num_input_tokens_seen": 100894550, + "step": 3088 + }, + { + "epoch": 0.27857690399963925, + "flos": 23589446319360.0, + "grad_norm": 1.5651660299742958, + "learning_rate": 3.3861615647106253e-06, + "loss": 0.8524, + "num_input_tokens_seen": 100923605, + "step": 3089 + }, + { + "epoch": 0.27866708752310954, + "flos": 31347661960320.0, + "grad_norm": 2.639504640806743, + "learning_rate": 3.3857403715504012e-06, + "loss": 0.78, + "num_input_tokens_seen": 100950785, + "step": 3090 + }, + { + "epoch": 0.27875727104657977, + "flos": 19545743585280.0, + "grad_norm": 2.0372553157739386, + "learning_rate": 3.385319060151167e-06, + "loss": 0.731, + "num_input_tokens_seen": 100979195, + "step": 3091 + }, + { + "epoch": 0.27884745457005006, + "flos": 21512184824640.0, + "grad_norm": 1.9995646598989545, + "learning_rate": 3.3848976305488728e-06, + "loss": 0.7609, + "num_input_tokens_seen": 101007050, + "step": 3092 + }, + { + "epoch": 0.2789376380935203, + "flos": 19653664599360.0, + "grad_norm": 2.5229189313337104, + "learning_rate": 3.384476082779476e-06, + "loss": 0.8083, + "num_input_tokens_seen": 101031530, + "step": 3093 + }, + { + "epoch": 0.2790278216169906, + "flos": 16740286159680.0, + "grad_norm": 1.7274159527219441, + "learning_rate": 3.3840544168789463e-06, + "loss": 0.7428, + "num_input_tokens_seen": 101058720, + "step": 3094 + }, + { + "epoch": 0.2791180051404608, + "flos": 21658619443680.0, + "grad_norm": 1.8909807107023304, + "learning_rate": 3.3836326328832617e-06, + "loss": 0.8735, + "num_input_tokens_seen": 101087535, + "step": 3095 + }, + { + "epoch": 0.2792081886639311, + "flos": 19176218834880.0, + "grad_norm": 2.837170545634698, + "learning_rate": 3.383210730828412e-06, + "loss": 0.7184, + "num_input_tokens_seen": 101115035, + "step": 3096 + }, + { + "epoch": 0.2792983721874014, + "flos": 25884482669760.0, + "grad_norm": 1.567609900070042, + "learning_rate": 3.3827887107503953e-06, + "loss": 0.8341, + "num_input_tokens_seen": 101146815, + "step": 3097 + }, + { + "epoch": 0.2793885557108716, + "flos": 23225311184160.0, + "grad_norm": 2.4092495540512795, + "learning_rate": 3.3823665726852216e-06, + "loss": 0.7123, + "num_input_tokens_seen": 101176315, + "step": 3098 + }, + { + "epoch": 0.2794787392343419, + "flos": 29564607815040.0, + "grad_norm": 1.4907554425439766, + "learning_rate": 3.3819443166689095e-06, + "loss": 0.8891, + "num_input_tokens_seen": 101207320, + "step": 3099 + }, + { + "epoch": 0.27956892275781214, + "flos": 19070342157600.0, + "grad_norm": 3.098206210292753, + "learning_rate": 3.3815219427374886e-06, + "loss": 0.7645, + "num_input_tokens_seen": 101233825, + "step": 3100 + }, + { + "epoch": 0.27965910628128243, + "flos": 25302721357920.0, + "grad_norm": 1.5341627483051965, + "learning_rate": 3.3810994509269975e-06, + "loss": 0.7811, + "num_input_tokens_seen": 101262495, + "step": 3101 + }, + { + "epoch": 0.27974928980475267, + "flos": 24967770217920.0, + "grad_norm": 1.559734338245491, + "learning_rate": 3.3806768412734864e-06, + "loss": 0.803, + "num_input_tokens_seen": 101291015, + "step": 3102 + }, + { + "epoch": 0.27983947332822295, + "flos": 26104190352960.0, + "grad_norm": 1.9715325809372626, + "learning_rate": 3.380254113813014e-06, + "loss": 0.8466, + "num_input_tokens_seen": 101323415, + "step": 3103 + }, + { + "epoch": 0.2799296568516932, + "flos": 27810663325440.0, + "grad_norm": 1.8094861195757364, + "learning_rate": 3.3798312685816496e-06, + "loss": 0.8023, + "num_input_tokens_seen": 101351090, + "step": 3104 + }, + { + "epoch": 0.2800198403751635, + "flos": 25804110181440.0, + "grad_norm": 1.8250859096634267, + "learning_rate": 3.3794083056154738e-06, + "loss": 0.768, + "num_input_tokens_seen": 101380290, + "step": 3105 + }, + { + "epoch": 0.2801100238986337, + "flos": 27554783732160.0, + "grad_norm": 2.036102016591151, + "learning_rate": 3.3789852249505746e-06, + "loss": 0.7749, + "num_input_tokens_seen": 101408350, + "step": 3106 + }, + { + "epoch": 0.280200207422104, + "flos": 20784434930880.0, + "grad_norm": 1.6422170271609917, + "learning_rate": 3.378562026623053e-06, + "loss": 0.8209, + "num_input_tokens_seen": 101436860, + "step": 3107 + }, + { + "epoch": 0.28029039094557423, + "flos": 26868558193920.0, + "grad_norm": 2.265380635730705, + "learning_rate": 3.3781387106690175e-06, + "loss": 0.8389, + "num_input_tokens_seen": 101465980, + "step": 3108 + }, + { + "epoch": 0.2803805744690445, + "flos": 22861064539680.0, + "grad_norm": 1.6723586479178516, + "learning_rate": 3.3777152771245885e-06, + "loss": 0.7931, + "num_input_tokens_seen": 101495265, + "step": 3109 + }, + { + "epoch": 0.28047075799251475, + "flos": 25150451086560.0, + "grad_norm": 5.601234926570811, + "learning_rate": 3.377291726025895e-06, + "loss": 0.799, + "num_input_tokens_seen": 101521755, + "step": 3110 + }, + { + "epoch": 0.28056094151598504, + "flos": 23480150024160.0, + "grad_norm": 2.620564027746097, + "learning_rate": 3.3768680574090782e-06, + "loss": 0.8863, + "num_input_tokens_seen": 101546605, + "step": 3111 + }, + { + "epoch": 0.2806511250394553, + "flos": 35211657406560.0, + "grad_norm": 1.84047958499081, + "learning_rate": 3.3764442713102857e-06, + "loss": 0.7477, + "num_input_tokens_seen": 101574735, + "step": 3112 + }, + { + "epoch": 0.28074130856292556, + "flos": 27377529648480.0, + "grad_norm": 1.6891440400971733, + "learning_rate": 3.3760203677656786e-06, + "loss": 0.7639, + "num_input_tokens_seen": 101603305, + "step": 3113 + }, + { + "epoch": 0.2808314920863958, + "flos": 20564987436000.0, + "grad_norm": 8.09918478171834, + "learning_rate": 3.3755963468114262e-06, + "loss": 0.8231, + "num_input_tokens_seen": 101629495, + "step": 3114 + }, + { + "epoch": 0.2809216756098661, + "flos": 25302349660320.0, + "grad_norm": 1.733838591233991, + "learning_rate": 3.3751722084837095e-06, + "loss": 0.6708, + "num_input_tokens_seen": 101656185, + "step": 3115 + }, + { + "epoch": 0.2810118591333363, + "flos": 20711496394560.0, + "grad_norm": 1.500097192158642, + "learning_rate": 3.3747479528187166e-06, + "loss": 0.7448, + "num_input_tokens_seen": 101684010, + "step": 3116 + }, + { + "epoch": 0.2811020426568066, + "flos": 12805805381280.0, + "grad_norm": 2.0155951919125856, + "learning_rate": 3.3743235798526485e-06, + "loss": 0.8166, + "num_input_tokens_seen": 101710550, + "step": 3117 + }, + { + "epoch": 0.28119222618027684, + "flos": 23004488408160.0, + "grad_norm": 1.4689707488698187, + "learning_rate": 3.373899089621714e-06, + "loss": 0.7031, + "num_input_tokens_seen": 101739895, + "step": 3118 + }, + { + "epoch": 0.2812824097037471, + "flos": 27415039669920.0, + "grad_norm": 2.0895986823307107, + "learning_rate": 3.373474482162134e-06, + "loss": 0.7455, + "num_input_tokens_seen": 101769320, + "step": 3119 + }, + { + "epoch": 0.28137259322721736, + "flos": 18488357827200.0, + "grad_norm": 2.1674821737187764, + "learning_rate": 3.3730497575101376e-06, + "loss": 0.7721, + "num_input_tokens_seen": 101793590, + "step": 3120 + }, + { + "epoch": 0.28146277675068765, + "flos": 24900555824640.0, + "grad_norm": 1.9533422120427395, + "learning_rate": 3.3726249157019654e-06, + "loss": 0.7647, + "num_input_tokens_seen": 101826370, + "step": 3121 + }, + { + "epoch": 0.28155296027415794, + "flos": 24859180148160.0, + "grad_norm": 1.8127181480933212, + "learning_rate": 3.372199956773866e-06, + "loss": 0.8534, + "num_input_tokens_seen": 101854525, + "step": 3122 + }, + { + "epoch": 0.28164314379762817, + "flos": 30402546078240.0, + "grad_norm": 1.795565716952873, + "learning_rate": 3.371774880762101e-06, + "loss": 0.7185, + "num_input_tokens_seen": 101885245, + "step": 3123 + }, + { + "epoch": 0.28173332732109846, + "flos": 24350283033120.0, + "grad_norm": 2.6032321830014515, + "learning_rate": 3.3713496877029392e-06, + "loss": 0.7527, + "num_input_tokens_seen": 101911870, + "step": 3124 + }, + { + "epoch": 0.2818235108445687, + "flos": 20266505564160.0, + "grad_norm": 2.7564716483543146, + "learning_rate": 3.37092437763266e-06, + "loss": 0.8341, + "num_input_tokens_seen": 101938710, + "step": 3125 + }, + { + "epoch": 0.281913694368039, + "flos": 24026073953760.0, + "grad_norm": 1.9480081900516488, + "learning_rate": 3.3704989505875537e-06, + "loss": 0.804, + "num_input_tokens_seen": 101967375, + "step": 3126 + }, + { + "epoch": 0.2820038778915092, + "flos": 21476347442400.0, + "grad_norm": 2.222999586461381, + "learning_rate": 3.3700734066039205e-06, + "loss": 0.7395, + "num_input_tokens_seen": 101996720, + "step": 3127 + }, + { + "epoch": 0.2820940614149795, + "flos": 24391287012000.0, + "grad_norm": 1.941703420572234, + "learning_rate": 3.36964774571807e-06, + "loss": 0.7994, + "num_input_tokens_seen": 102023145, + "step": 3128 + }, + { + "epoch": 0.28218424493844974, + "flos": 20383719018240.0, + "grad_norm": 2.610523268178932, + "learning_rate": 3.3692219679663206e-06, + "loss": 0.7651, + "num_input_tokens_seen": 102051740, + "step": 3129 + }, + { + "epoch": 0.28227442846192, + "flos": 31895407208160.0, + "grad_norm": 2.0088734476912955, + "learning_rate": 3.3687960733850043e-06, + "loss": 0.7808, + "num_input_tokens_seen": 102078820, + "step": 3130 + }, + { + "epoch": 0.28236461198539026, + "flos": 16590729280800.0, + "grad_norm": 1.7515633275951434, + "learning_rate": 3.3683700620104586e-06, + "loss": 0.7059, + "num_input_tokens_seen": 102104000, + "step": 3131 + }, + { + "epoch": 0.28245479550886055, + "flos": 22496074500000.0, + "grad_norm": 1.7819904996702192, + "learning_rate": 3.3679439338790347e-06, + "loss": 0.7865, + "num_input_tokens_seen": 102132665, + "step": 3132 + }, + { + "epoch": 0.2825449790323308, + "flos": 22344584793600.0, + "grad_norm": 1.856875879883919, + "learning_rate": 3.3675176890270916e-06, + "loss": 0.8915, + "num_input_tokens_seen": 102159765, + "step": 3133 + }, + { + "epoch": 0.28263516255580107, + "flos": 21764979327840.0, + "grad_norm": 1.7177749642456972, + "learning_rate": 3.367091327490998e-06, + "loss": 0.7909, + "num_input_tokens_seen": 102187405, + "step": 3134 + }, + { + "epoch": 0.2827253460792713, + "flos": 17609787282720.0, + "grad_norm": 1.7364513580908343, + "learning_rate": 3.3666648493071347e-06, + "loss": 0.7542, + "num_input_tokens_seen": 102213995, + "step": 3135 + }, + { + "epoch": 0.2828155296027416, + "flos": 21513225577920.0, + "grad_norm": 1.8710241125603124, + "learning_rate": 3.3662382545118914e-06, + "loss": 0.7474, + "num_input_tokens_seen": 102240715, + "step": 3136 + }, + { + "epoch": 0.2829057131262118, + "flos": 24238161836160.0, + "grad_norm": 1.8774877574739925, + "learning_rate": 3.3658115431416663e-06, + "loss": 0.8511, + "num_input_tokens_seen": 102268860, + "step": 3137 + }, + { + "epoch": 0.2829958966496821, + "flos": 67085872117920.0, + "grad_norm": 0.7032140463498102, + "learning_rate": 3.36538471523287e-06, + "loss": 0.6082, + "num_input_tokens_seen": 102365235, + "step": 3138 + }, + { + "epoch": 0.28308608017315234, + "flos": 18489733108320.0, + "grad_norm": 1.8322179247786015, + "learning_rate": 3.3649577708219204e-06, + "loss": 0.8297, + "num_input_tokens_seen": 102391595, + "step": 3139 + }, + { + "epoch": 0.28317626369662263, + "flos": 25920766089120.0, + "grad_norm": 1.9807199172162513, + "learning_rate": 3.3645307099452477e-06, + "loss": 0.7145, + "num_input_tokens_seen": 102420385, + "step": 3140 + }, + { + "epoch": 0.28326644722009287, + "flos": 15866324665440.0, + "grad_norm": 1.9592609305389002, + "learning_rate": 3.3641035326392907e-06, + "loss": 0.7671, + "num_input_tokens_seen": 102446875, + "step": 3141 + }, + { + "epoch": 0.28335663074356315, + "flos": 20894140093440.0, + "grad_norm": 2.000678543380133, + "learning_rate": 3.363676238940499e-06, + "loss": 0.7827, + "num_input_tokens_seen": 102473760, + "step": 3142 + }, + { + "epoch": 0.2834468142670334, + "flos": 19218152057760.0, + "grad_norm": 1.8532331952936483, + "learning_rate": 3.363248828885331e-06, + "loss": 0.8047, + "num_input_tokens_seen": 102502315, + "step": 3143 + }, + { + "epoch": 0.2835369977905037, + "flos": 22055395361760.0, + "grad_norm": 1.5817716409050826, + "learning_rate": 3.3628213025102562e-06, + "loss": 0.7932, + "num_input_tokens_seen": 102531020, + "step": 3144 + }, + { + "epoch": 0.28362718131397396, + "flos": 15756396484320.0, + "grad_norm": 2.232253981957658, + "learning_rate": 3.3623936598517536e-06, + "loss": 0.8088, + "num_input_tokens_seen": 102557835, + "step": 3145 + }, + { + "epoch": 0.2837173648374442, + "flos": 23916889167840.0, + "grad_norm": 2.2907659552323567, + "learning_rate": 3.3619659009463117e-06, + "loss": 0.6164, + "num_input_tokens_seen": 102585205, + "step": 3146 + }, + { + "epoch": 0.2838075483609145, + "flos": 26390406204000.0, + "grad_norm": 1.6023443035122917, + "learning_rate": 3.3615380258304287e-06, + "loss": 0.8241, + "num_input_tokens_seen": 102614570, + "step": 3147 + }, + { + "epoch": 0.2838977318843847, + "flos": 26212148536800.0, + "grad_norm": 1.7709316076796924, + "learning_rate": 3.3611100345406146e-06, + "loss": 0.7609, + "num_input_tokens_seen": 102642280, + "step": 3148 + }, + { + "epoch": 0.283987915407855, + "flos": 19326704957760.0, + "grad_norm": 1.790524614059993, + "learning_rate": 3.3606819271133873e-06, + "loss": 0.7598, + "num_input_tokens_seen": 102667785, + "step": 3149 + }, + { + "epoch": 0.28407809893132524, + "flos": 66143246609760.0, + "grad_norm": 0.7188738425996206, + "learning_rate": 3.360253703585275e-06, + "loss": 0.5927, + "num_input_tokens_seen": 102760070, + "step": 3150 + }, + { + "epoch": 0.28416828245479553, + "flos": 15902162047680.0, + "grad_norm": 1.7519016151978644, + "learning_rate": 3.3598253639928164e-06, + "loss": 0.7065, + "num_input_tokens_seen": 102783510, + "step": 3151 + }, + { + "epoch": 0.28425846597826576, + "flos": 39437817990720.0, + "grad_norm": 3.5290340796511663, + "learning_rate": 3.3593969083725596e-06, + "loss": 0.7225, + "num_input_tokens_seen": 102814505, + "step": 3152 + }, + { + "epoch": 0.28434864950173605, + "flos": 22271794936320.0, + "grad_norm": 2.3018204782053484, + "learning_rate": 3.358968336761063e-06, + "loss": 0.7234, + "num_input_tokens_seen": 102841820, + "step": 3153 + }, + { + "epoch": 0.2844388330252063, + "flos": 18560701647360.0, + "grad_norm": 2.0677933247071896, + "learning_rate": 3.3585396491948945e-06, + "loss": 0.7383, + "num_input_tokens_seen": 102870620, + "step": 3154 + }, + { + "epoch": 0.28452901654867657, + "flos": 23585989531680.0, + "grad_norm": 1.4748097668836762, + "learning_rate": 3.358110845710633e-06, + "loss": 0.8258, + "num_input_tokens_seen": 102899665, + "step": 3155 + }, + { + "epoch": 0.2846192000721468, + "flos": 20673874863840.0, + "grad_norm": 1.909099700110701, + "learning_rate": 3.357681926344865e-06, + "loss": 0.7586, + "num_input_tokens_seen": 102928035, + "step": 3156 + }, + { + "epoch": 0.2847093835956171, + "flos": 22636822145760.0, + "grad_norm": 1.520140302733036, + "learning_rate": 3.357252891134189e-06, + "loss": 0.7678, + "num_input_tokens_seen": 102955255, + "step": 3157 + }, + { + "epoch": 0.2847995671190873, + "flos": 20675250144960.0, + "grad_norm": 2.0202734522280252, + "learning_rate": 3.356823740115212e-06, + "loss": 0.8152, + "num_input_tokens_seen": 102982805, + "step": 3158 + }, + { + "epoch": 0.2848897506425576, + "flos": 32841935541120.0, + "grad_norm": 1.4410580263546104, + "learning_rate": 3.3563944733245525e-06, + "loss": 0.7149, + "num_input_tokens_seen": 103014240, + "step": 3159 + }, + { + "epoch": 0.28497993416602785, + "flos": 27524075776800.0, + "grad_norm": 2.017910748045824, + "learning_rate": 3.3559650907988375e-06, + "loss": 0.7629, + "num_input_tokens_seen": 103044020, + "step": 3160 + }, + { + "epoch": 0.28507011768949814, + "flos": 31636702713120.0, + "grad_norm": 5.526013325418703, + "learning_rate": 3.3555355925747045e-06, + "loss": 0.619, + "num_input_tokens_seen": 103075465, + "step": 3161 + }, + { + "epoch": 0.28516030121296837, + "flos": 26321630680800.0, + "grad_norm": 1.7474171788313746, + "learning_rate": 3.3551059786888e-06, + "loss": 0.7546, + "num_input_tokens_seen": 103104815, + "step": 3162 + }, + { + "epoch": 0.28525048473643866, + "flos": 28762952971200.0, + "grad_norm": 1.51724934790725, + "learning_rate": 3.3546762491777807e-06, + "loss": 0.8203, + "num_input_tokens_seen": 103133030, + "step": 3163 + }, + { + "epoch": 0.2853406682599089, + "flos": 19725859740480.0, + "grad_norm": 2.225524231949549, + "learning_rate": 3.3542464040783156e-06, + "loss": 0.7561, + "num_input_tokens_seen": 103162265, + "step": 3164 + }, + { + "epoch": 0.2854308517833792, + "flos": 20528927035200.0, + "grad_norm": 1.9929999396539, + "learning_rate": 3.353816443427079e-06, + "loss": 0.7818, + "num_input_tokens_seen": 103188125, + "step": 3165 + }, + { + "epoch": 0.2855210353068494, + "flos": 26394606386880.0, + "grad_norm": 2.164304533826703, + "learning_rate": 3.3533863672607597e-06, + "loss": 0.7483, + "num_input_tokens_seen": 103219575, + "step": 3166 + }, + { + "epoch": 0.2856112188303197, + "flos": 21108681180000.0, + "grad_norm": 1.6324428158101767, + "learning_rate": 3.352956175616052e-06, + "loss": 0.7975, + "num_input_tokens_seen": 103246500, + "step": 3167 + }, + { + "epoch": 0.28570140235378993, + "flos": 23989716194880.0, + "grad_norm": 2.5373644975835483, + "learning_rate": 3.352525868529664e-06, + "loss": 0.7345, + "num_input_tokens_seen": 103273785, + "step": 3168 + }, + { + "epoch": 0.2857915858772602, + "flos": 31604396458080.0, + "grad_norm": 1.8737199228429053, + "learning_rate": 3.3520954460383103e-06, + "loss": 0.7061, + "num_input_tokens_seen": 103303685, + "step": 3169 + }, + { + "epoch": 0.2858817694007305, + "flos": 18598211668800.0, + "grad_norm": 2.819413144752724, + "learning_rate": 3.3516649081787182e-06, + "loss": 0.7088, + "num_input_tokens_seen": 103327970, + "step": 3170 + }, + { + "epoch": 0.28597195292420075, + "flos": 38963271467520.0, + "grad_norm": 1.5607848099421544, + "learning_rate": 3.3512342549876236e-06, + "loss": 0.8664, + "num_input_tokens_seen": 103356610, + "step": 3171 + }, + { + "epoch": 0.28606213644767103, + "flos": 14080705806720.0, + "grad_norm": 1.7090753324009123, + "learning_rate": 3.350803486501771e-06, + "loss": 0.7258, + "num_input_tokens_seen": 103383405, + "step": 3172 + }, + { + "epoch": 0.28615231997114127, + "flos": 23986370916480.0, + "grad_norm": 1.571599935501808, + "learning_rate": 3.3503726027579175e-06, + "loss": 0.7932, + "num_input_tokens_seen": 103411655, + "step": 3173 + }, + { + "epoch": 0.28624250349461156, + "flos": 23151480573600.0, + "grad_norm": 1.529782506780342, + "learning_rate": 3.349941603792827e-06, + "loss": 0.824, + "num_input_tokens_seen": 103441490, + "step": 3174 + }, + { + "epoch": 0.2863326870180818, + "flos": 18853719564480.0, + "grad_norm": 6.101700355056809, + "learning_rate": 3.3495104896432755e-06, + "loss": 0.7653, + "num_input_tokens_seen": 103469240, + "step": 3175 + }, + { + "epoch": 0.2864228705415521, + "flos": 26613384826080.0, + "grad_norm": 1.7683763250840925, + "learning_rate": 3.3490792603460477e-06, + "loss": 0.7289, + "num_input_tokens_seen": 103498520, + "step": 3176 + }, + { + "epoch": 0.2865130540650223, + "flos": 26685691476480.0, + "grad_norm": 2.0958680783899926, + "learning_rate": 3.3486479159379393e-06, + "loss": 0.6987, + "num_input_tokens_seen": 103529065, + "step": 3177 + }, + { + "epoch": 0.2866032375884926, + "flos": 60724342236960.0, + "grad_norm": 0.665915719423021, + "learning_rate": 3.3482164564557537e-06, + "loss": 0.5762, + "num_input_tokens_seen": 103626845, + "step": 3178 + }, + { + "epoch": 0.28669342111196283, + "flos": 22350829313280.0, + "grad_norm": 1.7470098562852368, + "learning_rate": 3.3477848819363065e-06, + "loss": 0.8254, + "num_input_tokens_seen": 103654520, + "step": 3179 + }, + { + "epoch": 0.2867836046354331, + "flos": 61486740080640.0, + "grad_norm": 0.6668678491739931, + "learning_rate": 3.3473531924164213e-06, + "loss": 0.6052, + "num_input_tokens_seen": 103737505, + "step": 3180 + }, + { + "epoch": 0.28687378815890335, + "flos": 17096838663840.0, + "grad_norm": 2.7667103162639006, + "learning_rate": 3.3469213879329325e-06, + "loss": 0.7246, + "num_input_tokens_seen": 103760695, + "step": 3181 + }, + { + "epoch": 0.28696397168237364, + "flos": 14039441639520.0, + "grad_norm": 2.066535700473274, + "learning_rate": 3.3464894685226837e-06, + "loss": 0.7329, + "num_input_tokens_seen": 103785810, + "step": 3182 + }, + { + "epoch": 0.2870541552058439, + "flos": 24500731986240.0, + "grad_norm": 1.9162015834397683, + "learning_rate": 3.34605743422253e-06, + "loss": 0.845, + "num_input_tokens_seen": 103813875, + "step": 3183 + }, + { + "epoch": 0.28714433872931416, + "flos": 24281842037760.0, + "grad_norm": 1.9790501677374197, + "learning_rate": 3.345625285069333e-06, + "loss": 0.7379, + "num_input_tokens_seen": 103842795, + "step": 3184 + }, + { + "epoch": 0.2872345222527844, + "flos": 28980504808320.0, + "grad_norm": 1.6389152700739869, + "learning_rate": 3.345193021099967e-06, + "loss": 0.7474, + "num_input_tokens_seen": 103873540, + "step": 3185 + }, + { + "epoch": 0.2873247057762547, + "flos": 22532841126240.0, + "grad_norm": 1.540855563787203, + "learning_rate": 3.3447606423513157e-06, + "loss": 0.7353, + "num_input_tokens_seen": 103903010, + "step": 3186 + }, + { + "epoch": 0.2874148892997249, + "flos": 26066940519840.0, + "grad_norm": 1.7495590744402016, + "learning_rate": 3.344328148860271e-06, + "loss": 0.7352, + "num_input_tokens_seen": 103931775, + "step": 3187 + }, + { + "epoch": 0.2875050728231952, + "flos": 20311077840000.0, + "grad_norm": 1.564485585308143, + "learning_rate": 3.3438955406637365e-06, + "loss": 0.7845, + "num_input_tokens_seen": 103958685, + "step": 3188 + }, + { + "epoch": 0.28759525634666544, + "flos": 20747779813920.0, + "grad_norm": 1.855977520707901, + "learning_rate": 3.343462817798624e-06, + "loss": 0.7918, + "num_input_tokens_seen": 103987185, + "step": 3189 + }, + { + "epoch": 0.28768543987013573, + "flos": 32115263570400.0, + "grad_norm": 4.048814699009939, + "learning_rate": 3.343029980301856e-06, + "loss": 0.7341, + "num_input_tokens_seen": 104017685, + "step": 3190 + }, + { + "epoch": 0.28777562339360596, + "flos": 22459902589920.0, + "grad_norm": 2.5113036196781824, + "learning_rate": 3.342597028210365e-06, + "loss": 0.7964, + "num_input_tokens_seen": 104044080, + "step": 3191 + }, + { + "epoch": 0.28786580691707625, + "flos": 38741073410400.0, + "grad_norm": 2.433549663211719, + "learning_rate": 3.342163961561092e-06, + "loss": 0.7172, + "num_input_tokens_seen": 104075840, + "step": 3192 + }, + { + "epoch": 0.28795599044054654, + "flos": 19653887617920.0, + "grad_norm": 1.5631407377662618, + "learning_rate": 3.34173078039099e-06, + "loss": 0.7387, + "num_input_tokens_seen": 104103785, + "step": 3193 + }, + { + "epoch": 0.28804617396401677, + "flos": 15502003681440.0, + "grad_norm": 2.0588923026066492, + "learning_rate": 3.3412974847370193e-06, + "loss": 0.9329, + "num_input_tokens_seen": 104128770, + "step": 3194 + }, + { + "epoch": 0.28813635748748706, + "flos": 28577633049600.0, + "grad_norm": 1.7025823866169836, + "learning_rate": 3.3408640746361514e-06, + "loss": 0.7893, + "num_input_tokens_seen": 104159695, + "step": 3195 + }, + { + "epoch": 0.2882265410109573, + "flos": 71127344666400.0, + "grad_norm": 0.7494959727071574, + "learning_rate": 3.3404305501253663e-06, + "loss": 0.599, + "num_input_tokens_seen": 104249740, + "step": 3196 + }, + { + "epoch": 0.2883167245344276, + "flos": 21549992204160.0, + "grad_norm": 1.514385463727255, + "learning_rate": 3.3399969112416565e-06, + "loss": 0.838, + "num_input_tokens_seen": 104277720, + "step": 3197 + }, + { + "epoch": 0.2884069080578978, + "flos": 31160186192640.0, + "grad_norm": 1.8788536524399624, + "learning_rate": 3.3395631580220213e-06, + "loss": 0.7012, + "num_input_tokens_seen": 104308040, + "step": 3198 + }, + { + "epoch": 0.2884970915813681, + "flos": 27121687224960.0, + "grad_norm": 2.036291351366834, + "learning_rate": 3.3391292905034714e-06, + "loss": 0.7445, + "num_input_tokens_seen": 104336000, + "step": 3199 + }, + { + "epoch": 0.28858727510483834, + "flos": 25986828219840.0, + "grad_norm": 1.804482914941777, + "learning_rate": 3.338695308723027e-06, + "loss": 0.8304, + "num_input_tokens_seen": 104361545, + "step": 3200 + }, + { + "epoch": 0.2886774586283086, + "flos": 16375296120000.0, + "grad_norm": 2.0675723962435986, + "learning_rate": 3.338261212717716e-06, + "loss": 0.8238, + "num_input_tokens_seen": 104386335, + "step": 3201 + }, + { + "epoch": 0.28876764215177886, + "flos": 23880680088000.0, + "grad_norm": 1.8206209895479062, + "learning_rate": 3.33782700252458e-06, + "loss": 0.8052, + "num_input_tokens_seen": 104416180, + "step": 3202 + }, + { + "epoch": 0.28885782567524915, + "flos": 25483803926880.0, + "grad_norm": 1.7718113909678914, + "learning_rate": 3.337392678180668e-06, + "loss": 0.8067, + "num_input_tokens_seen": 104444780, + "step": 3203 + }, + { + "epoch": 0.2889480091987194, + "flos": 23772424546080.0, + "grad_norm": 3.668239400707555, + "learning_rate": 3.3369582397230377e-06, + "loss": 0.7722, + "num_input_tokens_seen": 104471355, + "step": 3204 + }, + { + "epoch": 0.28903819272218967, + "flos": 19581469458240.0, + "grad_norm": 2.0829314670798733, + "learning_rate": 3.336523687188759e-06, + "loss": 0.8113, + "num_input_tokens_seen": 104499050, + "step": 3205 + }, + { + "epoch": 0.2891283762456599, + "flos": 65295086662560.0, + "grad_norm": 0.7112362412125868, + "learning_rate": 3.336089020614909e-06, + "loss": 0.6119, + "num_input_tokens_seen": 104578735, + "step": 3206 + }, + { + "epoch": 0.2892185597691302, + "flos": 18161361015840.0, + "grad_norm": 1.6025346736162098, + "learning_rate": 3.3356542400385774e-06, + "loss": 0.6433, + "num_input_tokens_seen": 104607760, + "step": 3207 + }, + { + "epoch": 0.2893087432926004, + "flos": 25769610910560.0, + "grad_norm": 1.8288605343312605, + "learning_rate": 3.3352193454968607e-06, + "loss": 0.7062, + "num_input_tokens_seen": 104636600, + "step": 3208 + }, + { + "epoch": 0.2893989268160707, + "flos": 22127739181920.0, + "grad_norm": 1.7132501866368228, + "learning_rate": 3.3347843370268675e-06, + "loss": 0.8327, + "num_input_tokens_seen": 104662065, + "step": 3209 + }, + { + "epoch": 0.28948911033954094, + "flos": 22569496243200.0, + "grad_norm": 1.4195097649249495, + "learning_rate": 3.334349214665715e-06, + "loss": 0.7874, + "num_input_tokens_seen": 104689945, + "step": 3210 + }, + { + "epoch": 0.28957929386301123, + "flos": 24136559681280.0, + "grad_norm": 1.8524045035681267, + "learning_rate": 3.3339139784505293e-06, + "loss": 0.8514, + "num_input_tokens_seen": 104720240, + "step": 3211 + }, + { + "epoch": 0.28966947738648147, + "flos": 26249584218720.0, + "grad_norm": 1.5917029604676785, + "learning_rate": 3.333478628418448e-06, + "loss": 0.8355, + "num_input_tokens_seen": 104750480, + "step": 3212 + }, + { + "epoch": 0.28975966090995176, + "flos": 35210988350880.0, + "grad_norm": 1.9550690398257626, + "learning_rate": 3.333043164606618e-06, + "loss": 0.658, + "num_input_tokens_seen": 104781960, + "step": 3213 + }, + { + "epoch": 0.289849844433422, + "flos": 21836839941120.0, + "grad_norm": 1.8513361152643593, + "learning_rate": 3.3326075870521948e-06, + "loss": 0.8481, + "num_input_tokens_seen": 104810550, + "step": 3214 + }, + { + "epoch": 0.2899400279568923, + "flos": 21798363505920.0, + "grad_norm": 2.345346644468776, + "learning_rate": 3.3321718957923437e-06, + "loss": 0.7742, + "num_input_tokens_seen": 104838670, + "step": 3215 + }, + { + "epoch": 0.29003021148036257, + "flos": 27341580756960.0, + "grad_norm": 1.7851239291213865, + "learning_rate": 3.3317360908642413e-06, + "loss": 0.7701, + "num_input_tokens_seen": 104867430, + "step": 3216 + }, + { + "epoch": 0.2901203950038328, + "flos": 19836531316800.0, + "grad_norm": 2.1567521547161044, + "learning_rate": 3.331300172305072e-06, + "loss": 0.7232, + "num_input_tokens_seen": 104896230, + "step": 3217 + }, + { + "epoch": 0.2902105785273031, + "flos": 20748002832480.0, + "grad_norm": 1.8557661922364581, + "learning_rate": 3.330864140152032e-06, + "loss": 0.759, + "num_input_tokens_seen": 104926240, + "step": 3218 + }, + { + "epoch": 0.2903007620507733, + "flos": 38126225278560.0, + "grad_norm": 1.9945486998102897, + "learning_rate": 3.330427994442325e-06, + "loss": 0.7901, + "num_input_tokens_seen": 104956415, + "step": 3219 + }, + { + "epoch": 0.2903909455742436, + "flos": 24788285948640.0, + "grad_norm": 1.5313786043930075, + "learning_rate": 3.3299917352131657e-06, + "loss": 0.804, + "num_input_tokens_seen": 104987080, + "step": 3220 + }, + { + "epoch": 0.29048112909771384, + "flos": 11747527548960.0, + "grad_norm": 2.387330521564946, + "learning_rate": 3.329555362501778e-06, + "loss": 0.7353, + "num_input_tokens_seen": 105013735, + "step": 3221 + }, + { + "epoch": 0.29057131262118413, + "flos": 18452223086880.0, + "grad_norm": 1.778253075356615, + "learning_rate": 3.3291188763453954e-06, + "loss": 0.7358, + "num_input_tokens_seen": 105040295, + "step": 3222 + }, + { + "epoch": 0.29066149614465436, + "flos": 24864755612160.0, + "grad_norm": 2.077407598357124, + "learning_rate": 3.3286822767812618e-06, + "loss": 0.76, + "num_input_tokens_seen": 105069350, + "step": 3223 + }, + { + "epoch": 0.29075167966812465, + "flos": 22930248930240.0, + "grad_norm": 1.8635042644888984, + "learning_rate": 3.32824556384663e-06, + "loss": 0.7507, + "num_input_tokens_seen": 105099815, + "step": 3224 + }, + { + "epoch": 0.2908418631915949, + "flos": 20711124696960.0, + "grad_norm": 1.8110564830488007, + "learning_rate": 3.3278087375787628e-06, + "loss": 0.774, + "num_input_tokens_seen": 105127210, + "step": 3225 + }, + { + "epoch": 0.2909320467150652, + "flos": 26686360532160.0, + "grad_norm": 1.8429806284018366, + "learning_rate": 3.327371798014933e-06, + "loss": 0.7691, + "num_input_tokens_seen": 105152590, + "step": 3226 + }, + { + "epoch": 0.2910222302385354, + "flos": 25915822511040.0, + "grad_norm": 1.8823128824845083, + "learning_rate": 3.3269347451924218e-06, + "loss": 0.7713, + "num_input_tokens_seen": 105178635, + "step": 3227 + }, + { + "epoch": 0.2911124137620057, + "flos": 28471905051360.0, + "grad_norm": 1.447829163832333, + "learning_rate": 3.326497579148522e-06, + "loss": 0.8328, + "num_input_tokens_seen": 105209855, + "step": 3228 + }, + { + "epoch": 0.29120259728547593, + "flos": 21002730163200.0, + "grad_norm": 1.9358900826399206, + "learning_rate": 3.3260602999205345e-06, + "loss": 0.7672, + "num_input_tokens_seen": 105236610, + "step": 3229 + }, + { + "epoch": 0.2912927808089462, + "flos": 36668755493760.0, + "grad_norm": 3.71338298328384, + "learning_rate": 3.32562290754577e-06, + "loss": 0.7824, + "num_input_tokens_seen": 105266075, + "step": 3230 + }, + { + "epoch": 0.29138296433241645, + "flos": 16813113186720.0, + "grad_norm": 1.763942017677079, + "learning_rate": 3.3251854020615494e-06, + "loss": 0.7517, + "num_input_tokens_seen": 105293265, + "step": 3231 + }, + { + "epoch": 0.29147314785588674, + "flos": 25993964813760.0, + "grad_norm": 1.8260447085569744, + "learning_rate": 3.324747783505204e-06, + "loss": 0.7746, + "num_input_tokens_seen": 105322485, + "step": 3232 + }, + { + "epoch": 0.29156333137935697, + "flos": 32221772133600.0, + "grad_norm": 2.2156977546056242, + "learning_rate": 3.324310051914073e-06, + "loss": 0.6865, + "num_input_tokens_seen": 105353675, + "step": 3233 + }, + { + "epoch": 0.29165351490282726, + "flos": 17213680420320.0, + "grad_norm": 2.3533403640008066, + "learning_rate": 3.3238722073255056e-06, + "loss": 0.7376, + "num_input_tokens_seen": 105380335, + "step": 3234 + }, + { + "epoch": 0.2917436984262975, + "flos": 36232165029120.0, + "grad_norm": 1.7045492813760976, + "learning_rate": 3.323434249776863e-06, + "loss": 0.7309, + "num_input_tokens_seen": 105411135, + "step": 3235 + }, + { + "epoch": 0.2918338819497678, + "flos": 69076330755840.0, + "grad_norm": 0.7211417747367554, + "learning_rate": 3.3229961793055117e-06, + "loss": 0.5896, + "num_input_tokens_seen": 105498965, + "step": 3236 + }, + { + "epoch": 0.291924065473238, + "flos": 35431327920000.0, + "grad_norm": 1.724776297728613, + "learning_rate": 3.3225579959488314e-06, + "loss": 0.6963, + "num_input_tokens_seen": 105528675, + "step": 3237 + }, + { + "epoch": 0.2920142489967083, + "flos": 20747259437280.0, + "grad_norm": 2.151352448138059, + "learning_rate": 3.322119699744211e-06, + "loss": 0.7952, + "num_input_tokens_seen": 105555155, + "step": 3238 + }, + { + "epoch": 0.29210443252017854, + "flos": 19356892536480.0, + "grad_norm": 8.092534861426204, + "learning_rate": 3.3216812907290476e-06, + "loss": 0.782, + "num_input_tokens_seen": 105580980, + "step": 3239 + }, + { + "epoch": 0.2921946160436488, + "flos": 23952466361760.0, + "grad_norm": 1.6975263213194267, + "learning_rate": 3.3212427689407484e-06, + "loss": 0.7903, + "num_input_tokens_seen": 105611965, + "step": 3240 + }, + { + "epoch": 0.2922847995671191, + "flos": 61263018063360.0, + "grad_norm": 0.6526744940620888, + "learning_rate": 3.3208041344167317e-06, + "loss": 0.5667, + "num_input_tokens_seen": 105701205, + "step": 3241 + }, + { + "epoch": 0.29237498309058935, + "flos": 23843616103680.0, + "grad_norm": 1.6342930993881244, + "learning_rate": 3.3203653871944224e-06, + "loss": 0.6757, + "num_input_tokens_seen": 105728275, + "step": 3242 + }, + { + "epoch": 0.29246516661405964, + "flos": 22456594481280.0, + "grad_norm": 1.664840015136883, + "learning_rate": 3.3199265273112587e-06, + "loss": 0.7302, + "num_input_tokens_seen": 105756790, + "step": 3243 + }, + { + "epoch": 0.29255535013752987, + "flos": 17032226153760.0, + "grad_norm": 1.613980859720974, + "learning_rate": 3.3194875548046852e-06, + "loss": 0.7107, + "num_input_tokens_seen": 105785195, + "step": 3244 + }, + { + "epoch": 0.29264553366100016, + "flos": 23079917318400.0, + "grad_norm": 1.7688904669422882, + "learning_rate": 3.319048469712158e-06, + "loss": 0.7485, + "num_input_tokens_seen": 105814700, + "step": 3245 + }, + { + "epoch": 0.2927357171844704, + "flos": 24682409271360.0, + "grad_norm": 2.361003894490864, + "learning_rate": 3.3186092720711423e-06, + "loss": 0.8644, + "num_input_tokens_seen": 105842865, + "step": 3246 + }, + { + "epoch": 0.2928259007079407, + "flos": 23261408754720.0, + "grad_norm": 2.209853914360617, + "learning_rate": 3.3181699619191125e-06, + "loss": 0.6958, + "num_input_tokens_seen": 105871500, + "step": 3247 + }, + { + "epoch": 0.2929160842314109, + "flos": 19363397244480.0, + "grad_norm": 1.5835571465498681, + "learning_rate": 3.3177305392935536e-06, + "loss": 0.8104, + "num_input_tokens_seen": 105899850, + "step": 3248 + }, + { + "epoch": 0.2930062677548812, + "flos": 61796081256000.0, + "grad_norm": 0.738307407460725, + "learning_rate": 3.3172910042319595e-06, + "loss": 0.6552, + "num_input_tokens_seen": 105988695, + "step": 3249 + }, + { + "epoch": 0.29309645127835143, + "flos": 18524975774400.0, + "grad_norm": 1.6853960838365625, + "learning_rate": 3.316851356771833e-06, + "loss": 0.8033, + "num_input_tokens_seen": 106018300, + "step": 3250 + }, + { + "epoch": 0.2931866348018217, + "flos": 21984092294880.0, + "grad_norm": 1.8024258664741615, + "learning_rate": 3.3164115969506876e-06, + "loss": 0.8028, + "num_input_tokens_seen": 106047285, + "step": 3251 + }, + { + "epoch": 0.29327681832529195, + "flos": 16265776806240.0, + "grad_norm": 2.4443843164659014, + "learning_rate": 3.315971724806046e-06, + "loss": 0.8547, + "num_input_tokens_seen": 106068770, + "step": 3252 + }, + { + "epoch": 0.29336700184876224, + "flos": 32110877538720.0, + "grad_norm": 1.6144517988545168, + "learning_rate": 3.315531740375441e-06, + "loss": 0.7087, + "num_input_tokens_seen": 106100045, + "step": 3253 + }, + { + "epoch": 0.2934571853722325, + "flos": 28215876779040.0, + "grad_norm": 1.747336404152716, + "learning_rate": 3.315091643696414e-06, + "loss": 0.736, + "num_input_tokens_seen": 106132315, + "step": 3254 + }, + { + "epoch": 0.29354736889570276, + "flos": 24646274531040.0, + "grad_norm": 1.5041236621207505, + "learning_rate": 3.3146514348065164e-06, + "loss": 0.6936, + "num_input_tokens_seen": 106161790, + "step": 3255 + }, + { + "epoch": 0.293637552419173, + "flos": 20274088195200.0, + "grad_norm": 4.9624161566658715, + "learning_rate": 3.31421111374331e-06, + "loss": 0.7726, + "num_input_tokens_seen": 106190435, + "step": 3256 + }, + { + "epoch": 0.2937277359426433, + "flos": 17177768698560.0, + "grad_norm": 1.8480629987382189, + "learning_rate": 3.3137706805443647e-06, + "loss": 0.7777, + "num_input_tokens_seen": 106216620, + "step": 3257 + }, + { + "epoch": 0.2938179194661135, + "flos": 29852161777440.0, + "grad_norm": 1.6192773409230372, + "learning_rate": 3.313330135247261e-06, + "loss": 0.8187, + "num_input_tokens_seen": 106243715, + "step": 3258 + }, + { + "epoch": 0.2939081029895838, + "flos": 34702165575360.0, + "grad_norm": 2.3304093182281562, + "learning_rate": 3.312889477889588e-06, + "loss": 0.7421, + "num_input_tokens_seen": 106277560, + "step": 3259 + }, + { + "epoch": 0.29399828651305404, + "flos": 19837943767680.0, + "grad_norm": 4.416933999278175, + "learning_rate": 3.3124487085089464e-06, + "loss": 0.7806, + "num_input_tokens_seen": 106304140, + "step": 3260 + }, + { + "epoch": 0.29408847003652433, + "flos": 23116386586560.0, + "grad_norm": 1.6007777082293282, + "learning_rate": 3.312007827142943e-06, + "loss": 0.8027, + "num_input_tokens_seen": 106332325, + "step": 3261 + }, + { + "epoch": 0.29417865355999456, + "flos": 26173857950400.0, + "grad_norm": 1.434813269185799, + "learning_rate": 3.3115668338291983e-06, + "loss": 0.6889, + "num_input_tokens_seen": 106361120, + "step": 3262 + }, + { + "epoch": 0.29426883708346485, + "flos": 23329292203680.0, + "grad_norm": 1.77675221419126, + "learning_rate": 3.3111257286053394e-06, + "loss": 0.8517, + "num_input_tokens_seen": 106388615, + "step": 3263 + }, + { + "epoch": 0.29435902060693514, + "flos": 22568790017760.0, + "grad_norm": 16.62880852464237, + "learning_rate": 3.3106845115090043e-06, + "loss": 0.7654, + "num_input_tokens_seen": 106418790, + "step": 3264 + }, + { + "epoch": 0.2944492041304054, + "flos": 19399866512640.0, + "grad_norm": 1.8190501669980248, + "learning_rate": 3.310243182577839e-06, + "loss": 0.7169, + "num_input_tokens_seen": 106446660, + "step": 3265 + }, + { + "epoch": 0.29453938765387566, + "flos": 23513905899840.0, + "grad_norm": 1.7274341713139592, + "learning_rate": 3.3098017418495007e-06, + "loss": 0.6614, + "num_input_tokens_seen": 106476690, + "step": 3266 + }, + { + "epoch": 0.2946295711773459, + "flos": 29090767517280.0, + "grad_norm": 1.8463319706266412, + "learning_rate": 3.309360189361656e-06, + "loss": 0.7354, + "num_input_tokens_seen": 106507875, + "step": 3267 + }, + { + "epoch": 0.2947197547008162, + "flos": 24312847351200.0, + "grad_norm": 1.740166693847313, + "learning_rate": 3.3089185251519797e-06, + "loss": 0.8453, + "num_input_tokens_seen": 106536360, + "step": 3268 + }, + { + "epoch": 0.2948099382242864, + "flos": 28723212764160.0, + "grad_norm": 1.9355883858518526, + "learning_rate": 3.3084767492581574e-06, + "loss": 0.7475, + "num_input_tokens_seen": 106565510, + "step": 3269 + }, + { + "epoch": 0.2949001217477567, + "flos": 24063472465920.0, + "grad_norm": 1.9413886270932104, + "learning_rate": 3.3080348617178846e-06, + "loss": 0.7867, + "num_input_tokens_seen": 106592295, + "step": 3270 + }, + { + "epoch": 0.29499030527122694, + "flos": 19837163202720.0, + "grad_norm": 1.7071598798880168, + "learning_rate": 3.307592862568865e-06, + "loss": 0.7438, + "num_input_tokens_seen": 106620175, + "step": 3271 + }, + { + "epoch": 0.2950804887946972, + "flos": 29415794331360.0, + "grad_norm": 1.5936311415687783, + "learning_rate": 3.307150751848812e-06, + "loss": 0.833, + "num_input_tokens_seen": 106650240, + "step": 3272 + }, + { + "epoch": 0.29517067231816746, + "flos": 33898131866880.0, + "grad_norm": 4.9707975187500955, + "learning_rate": 3.3067085295954497e-06, + "loss": 0.6742, + "num_input_tokens_seen": 106683595, + "step": 3273 + }, + { + "epoch": 0.29526085584163775, + "flos": 28030965724800.0, + "grad_norm": 1.693911698846962, + "learning_rate": 3.3062661958465098e-06, + "loss": 0.6326, + "num_input_tokens_seen": 106713125, + "step": 3274 + }, + { + "epoch": 0.295351039365108, + "flos": 25046432897280.0, + "grad_norm": 2.0077486139367986, + "learning_rate": 3.305823750639736e-06, + "loss": 0.6578, + "num_input_tokens_seen": 106743170, + "step": 3275 + }, + { + "epoch": 0.29544122288857827, + "flos": 17214089287680.0, + "grad_norm": 1.6845270742865834, + "learning_rate": 3.3053811940128795e-06, + "loss": 0.7591, + "num_input_tokens_seen": 106769880, + "step": 3276 + }, + { + "epoch": 0.2955314064120485, + "flos": 21836988620160.0, + "grad_norm": 1.724009840580941, + "learning_rate": 3.3049385260037016e-06, + "loss": 0.7432, + "num_input_tokens_seen": 106797970, + "step": 3277 + }, + { + "epoch": 0.2956215899355188, + "flos": 36631654339680.0, + "grad_norm": 1.6766154740382255, + "learning_rate": 3.3044957466499736e-06, + "loss": 0.6597, + "num_input_tokens_seen": 106830270, + "step": 3278 + }, + { + "epoch": 0.295711773458989, + "flos": 23261891961600.0, + "grad_norm": 2.7767503321736657, + "learning_rate": 3.304052855989475e-06, + "loss": 0.684, + "num_input_tokens_seen": 106858190, + "step": 3279 + }, + { + "epoch": 0.2958019569824593, + "flos": 21694717014240.0, + "grad_norm": 1.7339468075162445, + "learning_rate": 3.3036098540599966e-06, + "loss": 0.8261, + "num_input_tokens_seen": 106888280, + "step": 3280 + }, + { + "epoch": 0.29589214050592955, + "flos": 20600936327520.0, + "grad_norm": 2.1610098009571272, + "learning_rate": 3.3031667408993373e-06, + "loss": 0.7227, + "num_input_tokens_seen": 106915895, + "step": 3281 + }, + { + "epoch": 0.29598232402939983, + "flos": 24572889957600.0, + "grad_norm": 1.874463068059242, + "learning_rate": 3.302723516545306e-06, + "loss": 0.7623, + "num_input_tokens_seen": 106945755, + "step": 3282 + }, + { + "epoch": 0.29607250755287007, + "flos": 21474674803200.0, + "grad_norm": 1.593789717235465, + "learning_rate": 3.302280181035722e-06, + "loss": 0.7912, + "num_input_tokens_seen": 106975495, + "step": 3283 + }, + { + "epoch": 0.29616269107634036, + "flos": 66651028632000.0, + "grad_norm": 0.5992928828371404, + "learning_rate": 3.3018367344084117e-06, + "loss": 0.6175, + "num_input_tokens_seen": 107073795, + "step": 3284 + }, + { + "epoch": 0.2962528745998106, + "flos": 29125378297440.0, + "grad_norm": 1.4460026098678906, + "learning_rate": 3.3013931767012125e-06, + "loss": 0.7534, + "num_input_tokens_seen": 107106080, + "step": 3285 + }, + { + "epoch": 0.2963430581232809, + "flos": 23334644649120.0, + "grad_norm": 1.685340912966846, + "learning_rate": 3.300949507951972e-06, + "loss": 0.7561, + "num_input_tokens_seen": 107134325, + "step": 3286 + }, + { + "epoch": 0.2964332416467511, + "flos": 21111506081760.0, + "grad_norm": 2.286831129497245, + "learning_rate": 3.300505728198546e-06, + "loss": 0.8086, + "num_input_tokens_seen": 107161400, + "step": 3287 + }, + { + "epoch": 0.2965234251702214, + "flos": 22569198885120.0, + "grad_norm": 1.9126304407453105, + "learning_rate": 3.3000618374788e-06, + "loss": 0.7477, + "num_input_tokens_seen": 107190065, + "step": 3288 + }, + { + "epoch": 0.2966136086936917, + "flos": 21512891050080.0, + "grad_norm": 1.627505571557981, + "learning_rate": 3.2996178358306104e-06, + "loss": 0.801, + "num_input_tokens_seen": 107217865, + "step": 3289 + }, + { + "epoch": 0.2967037922171619, + "flos": 34701124822080.0, + "grad_norm": 2.5874249654753916, + "learning_rate": 3.2991737232918606e-06, + "loss": 0.729, + "num_input_tokens_seen": 107249215, + "step": 3290 + }, + { + "epoch": 0.2967939757406322, + "flos": 18890374681440.0, + "grad_norm": 1.7391507492226568, + "learning_rate": 3.298729499900445e-06, + "loss": 0.7146, + "num_input_tokens_seen": 107276180, + "step": 3291 + }, + { + "epoch": 0.29688415926410244, + "flos": 14407219411200.0, + "grad_norm": 1.7996065689969238, + "learning_rate": 3.2982851656942677e-06, + "loss": 0.7043, + "num_input_tokens_seen": 107301095, + "step": 3292 + }, + { + "epoch": 0.29697434278757273, + "flos": 26905696517760.0, + "grad_norm": 1.5817407411943736, + "learning_rate": 3.2978407207112416e-06, + "loss": 0.7878, + "num_input_tokens_seen": 107331675, + "step": 3293 + }, + { + "epoch": 0.29706452631104296, + "flos": 19072274985120.0, + "grad_norm": 3.4078708257148334, + "learning_rate": 3.2973961649892888e-06, + "loss": 0.7685, + "num_input_tokens_seen": 107359170, + "step": 3294 + }, + { + "epoch": 0.29715470983451325, + "flos": 23480893419360.0, + "grad_norm": 2.711045496655584, + "learning_rate": 3.296951498566341e-06, + "loss": 0.8249, + "num_input_tokens_seen": 107388030, + "step": 3295 + }, + { + "epoch": 0.2972448933579835, + "flos": 17906708024640.0, + "grad_norm": 2.230325938127251, + "learning_rate": 3.2965067214803404e-06, + "loss": 0.7562, + "num_input_tokens_seen": 107415825, + "step": 3296 + }, + { + "epoch": 0.2973350768814538, + "flos": 27593222997600.0, + "grad_norm": 1.296742903634653, + "learning_rate": 3.2960618337692372e-06, + "loss": 0.7998, + "num_input_tokens_seen": 107447435, + "step": 3297 + }, + { + "epoch": 0.297425260404924, + "flos": 37432194090720.0, + "grad_norm": 1.6512713187355412, + "learning_rate": 3.2956168354709927e-06, + "loss": 0.7234, + "num_input_tokens_seen": 107480895, + "step": 3298 + }, + { + "epoch": 0.2975154439283943, + "flos": 20784695119200.0, + "grad_norm": 2.1766712826986083, + "learning_rate": 3.2951717266235754e-06, + "loss": 0.814, + "num_input_tokens_seen": 107506635, + "step": 3299 + }, + { + "epoch": 0.29760562745186453, + "flos": 26832200435040.0, + "grad_norm": 1.5127123344749187, + "learning_rate": 3.294726507264964e-06, + "loss": 0.6353, + "num_input_tokens_seen": 107537435, + "step": 3300 + }, + { + "epoch": 0.2976958109753348, + "flos": 25156323908640.0, + "grad_norm": 1.804660417097582, + "learning_rate": 3.2942811774331487e-06, + "loss": 0.7357, + "num_input_tokens_seen": 107564145, + "step": 3301 + }, + { + "epoch": 0.29778599449880505, + "flos": 21002432805120.0, + "grad_norm": 1.8545179641482998, + "learning_rate": 3.293835737166127e-06, + "loss": 0.7574, + "num_input_tokens_seen": 107590845, + "step": 3302 + }, + { + "epoch": 0.29787617802227534, + "flos": 21257531833440.0, + "grad_norm": 2.1512133612164264, + "learning_rate": 3.293390186501906e-06, + "loss": 0.7102, + "num_input_tokens_seen": 107622135, + "step": 3303 + }, + { + "epoch": 0.2979663615457456, + "flos": 14845817042880.0, + "grad_norm": 1.8684148864347425, + "learning_rate": 3.2929445254785024e-06, + "loss": 0.7716, + "num_input_tokens_seen": 107649785, + "step": 3304 + }, + { + "epoch": 0.29805654506921586, + "flos": 16448160316800.0, + "grad_norm": 2.203876015646091, + "learning_rate": 3.2924987541339423e-06, + "loss": 0.7188, + "num_input_tokens_seen": 107673780, + "step": 3305 + }, + { + "epoch": 0.2981467285926861, + "flos": 26103335448480.0, + "grad_norm": 1.9676591182091987, + "learning_rate": 3.292052872506262e-06, + "loss": 0.7838, + "num_input_tokens_seen": 107701105, + "step": 3306 + }, + { + "epoch": 0.2982369121161564, + "flos": 19947165723360.0, + "grad_norm": 1.5062428739387077, + "learning_rate": 3.291606880633506e-06, + "loss": 0.8026, + "num_input_tokens_seen": 107729665, + "step": 3307 + }, + { + "epoch": 0.2983270956396266, + "flos": 19472396181600.0, + "grad_norm": 2.445957379117176, + "learning_rate": 3.2911607785537297e-06, + "loss": 0.8182, + "num_input_tokens_seen": 107756625, + "step": 3308 + }, + { + "epoch": 0.2984172791630969, + "flos": 22456594481280.0, + "grad_norm": 1.8844928761522435, + "learning_rate": 3.290714566304997e-06, + "loss": 0.8522, + "num_input_tokens_seen": 107780680, + "step": 3309 + }, + { + "epoch": 0.29850746268656714, + "flos": 22532692447200.0, + "grad_norm": 3.6574518559612397, + "learning_rate": 3.2902682439253794e-06, + "loss": 0.8353, + "num_input_tokens_seen": 107805350, + "step": 3310 + }, + { + "epoch": 0.2985976462100374, + "flos": 18707656643040.0, + "grad_norm": 2.165283409580587, + "learning_rate": 3.289821811452961e-06, + "loss": 0.7774, + "num_input_tokens_seen": 107830140, + "step": 3311 + }, + { + "epoch": 0.2986878297335077, + "flos": 22022828918400.0, + "grad_norm": 2.604645514650415, + "learning_rate": 3.289375268925834e-06, + "loss": 0.8167, + "num_input_tokens_seen": 107858335, + "step": 3312 + }, + { + "epoch": 0.29877801325697795, + "flos": 26831345530560.0, + "grad_norm": 2.1323925904462433, + "learning_rate": 3.288928616382099e-06, + "loss": 0.7528, + "num_input_tokens_seen": 107888415, + "step": 3313 + }, + { + "epoch": 0.29886819678044824, + "flos": 21693713430720.0, + "grad_norm": 1.8184470904343708, + "learning_rate": 3.288481853859868e-06, + "loss": 0.7979, + "num_input_tokens_seen": 107917720, + "step": 3314 + }, + { + "epoch": 0.29895838030391847, + "flos": 22242016224960.0, + "grad_norm": 2.9219448468840787, + "learning_rate": 3.2880349813972604e-06, + "loss": 0.7391, + "num_input_tokens_seen": 107947360, + "step": 3315 + }, + { + "epoch": 0.29904856382738876, + "flos": 22382726700960.0, + "grad_norm": 2.4325694566084572, + "learning_rate": 3.2875879990324052e-06, + "loss": 0.7651, + "num_input_tokens_seen": 107977110, + "step": 3316 + }, + { + "epoch": 0.299138747350859, + "flos": 24354594725280.0, + "grad_norm": 1.8001572195904252, + "learning_rate": 3.287140906803443e-06, + "loss": 0.8137, + "num_input_tokens_seen": 108005420, + "step": 3317 + }, + { + "epoch": 0.2992289308743293, + "flos": 19982408389440.0, + "grad_norm": 2.045685050961724, + "learning_rate": 3.2866937047485216e-06, + "loss": 0.6817, + "num_input_tokens_seen": 108032825, + "step": 3318 + }, + { + "epoch": 0.2993191143977995, + "flos": 27450951391680.0, + "grad_norm": 2.560140872894009, + "learning_rate": 3.2862463929057985e-06, + "loss": 0.8242, + "num_input_tokens_seen": 108061500, + "step": 3319 + }, + { + "epoch": 0.2994092979212698, + "flos": 39396590993280.0, + "grad_norm": 5.821909517686522, + "learning_rate": 3.285798971313441e-06, + "loss": 0.6109, + "num_input_tokens_seen": 108097020, + "step": 3320 + }, + { + "epoch": 0.29949948144474003, + "flos": 23625358041120.0, + "grad_norm": 1.6546209709185862, + "learning_rate": 3.2853514400096248e-06, + "loss": 0.7847, + "num_input_tokens_seen": 108126895, + "step": 3321 + }, + { + "epoch": 0.2995896649682103, + "flos": 22823294329920.0, + "grad_norm": 2.150001095616261, + "learning_rate": 3.2849037990325367e-06, + "loss": 0.7885, + "num_input_tokens_seen": 108156520, + "step": 3322 + }, + { + "epoch": 0.29967984849168056, + "flos": 19218226397280.0, + "grad_norm": 1.477161911496048, + "learning_rate": 3.2844560484203717e-06, + "loss": 0.7214, + "num_input_tokens_seen": 108185260, + "step": 3323 + }, + { + "epoch": 0.29977003201515084, + "flos": 28322794209600.0, + "grad_norm": 1.5164816097222322, + "learning_rate": 3.2840081882113333e-06, + "loss": 0.8457, + "num_input_tokens_seen": 108217230, + "step": 3324 + }, + { + "epoch": 0.2998602155386211, + "flos": 28176842797440.0, + "grad_norm": 2.2713835046860993, + "learning_rate": 3.283560218443638e-06, + "loss": 0.737, + "num_input_tokens_seen": 108244505, + "step": 3325 + }, + { + "epoch": 0.29995039906209137, + "flos": 16522436964480.0, + "grad_norm": 1.644851109801777, + "learning_rate": 3.2831121391555064e-06, + "loss": 0.7777, + "num_input_tokens_seen": 108270880, + "step": 3326 + }, + { + "epoch": 0.3000405825855616, + "flos": 38744939065440.0, + "grad_norm": 1.8876115096544641, + "learning_rate": 3.2826639503851724e-06, + "loss": 0.7151, + "num_input_tokens_seen": 108307305, + "step": 3327 + }, + { + "epoch": 0.3001307661090319, + "flos": 26684204686080.0, + "grad_norm": 1.4207633567127174, + "learning_rate": 3.282215652170877e-06, + "loss": 0.8053, + "num_input_tokens_seen": 108336430, + "step": 3328 + }, + { + "epoch": 0.3002209496325021, + "flos": 23335462383840.0, + "grad_norm": 3.2534299504155215, + "learning_rate": 3.281767244550873e-06, + "loss": 0.7964, + "num_input_tokens_seen": 108366935, + "step": 3329 + }, + { + "epoch": 0.3003111331559724, + "flos": 65310623622240.0, + "grad_norm": 0.6195893932637068, + "learning_rate": 3.2813187275634193e-06, + "loss": 0.5525, + "num_input_tokens_seen": 108462005, + "step": 3330 + }, + { + "epoch": 0.30040131667944264, + "flos": 23368772222400.0, + "grad_norm": 1.4843698601884698, + "learning_rate": 3.280870101246787e-06, + "loss": 0.7876, + "num_input_tokens_seen": 108490305, + "step": 3331 + }, + { + "epoch": 0.30049150020291293, + "flos": 21253368820320.0, + "grad_norm": 1.8099231818531953, + "learning_rate": 3.280421365639255e-06, + "loss": 0.686, + "num_input_tokens_seen": 108515785, + "step": 3332 + }, + { + "epoch": 0.30058168372638316, + "flos": 18890188832640.0, + "grad_norm": 1.5803694674524706, + "learning_rate": 3.279972520779112e-06, + "loss": 0.7454, + "num_input_tokens_seen": 108543170, + "step": 3333 + }, + { + "epoch": 0.30067186724985345, + "flos": 18598992233760.0, + "grad_norm": 1.4934453007042738, + "learning_rate": 3.279523566704656e-06, + "loss": 0.8318, + "num_input_tokens_seen": 108569525, + "step": 3334 + }, + { + "epoch": 0.30076205077332374, + "flos": 20777707204320.0, + "grad_norm": 1.8450112635571057, + "learning_rate": 3.2790745034541935e-06, + "loss": 0.8158, + "num_input_tokens_seen": 108593425, + "step": 3335 + }, + { + "epoch": 0.300852234296794, + "flos": 17432161501440.0, + "grad_norm": 1.7131716977156382, + "learning_rate": 3.278625331066042e-06, + "loss": 0.8088, + "num_input_tokens_seen": 108619660, + "step": 3336 + }, + { + "epoch": 0.30094241782026426, + "flos": 21731074773120.0, + "grad_norm": 1.932695276379997, + "learning_rate": 3.278176049578527e-06, + "loss": 0.8115, + "num_input_tokens_seen": 108647705, + "step": 3337 + }, + { + "epoch": 0.3010326013437345, + "flos": 36013089231840.0, + "grad_norm": 1.7692957841448111, + "learning_rate": 3.2777266590299835e-06, + "loss": 0.7239, + "num_input_tokens_seen": 108677930, + "step": 3338 + }, + { + "epoch": 0.3011227848672048, + "flos": 34810569796320.0, + "grad_norm": 1.7251801631663548, + "learning_rate": 3.2772771594587562e-06, + "loss": 0.7001, + "num_input_tokens_seen": 108710980, + "step": 3339 + }, + { + "epoch": 0.301212968390675, + "flos": 22715001618240.0, + "grad_norm": 1.9375119362740312, + "learning_rate": 3.2768275509031988e-06, + "loss": 0.8694, + "num_input_tokens_seen": 108738285, + "step": 3340 + }, + { + "epoch": 0.3013031519141453, + "flos": 25191975442080.0, + "grad_norm": 2.690219922204852, + "learning_rate": 3.276377833401675e-06, + "loss": 0.7443, + "num_input_tokens_seen": 108765630, + "step": 3341 + }, + { + "epoch": 0.30139333543761554, + "flos": 28180262415360.0, + "grad_norm": 1.6887626868843695, + "learning_rate": 3.2759280069925557e-06, + "loss": 0.8278, + "num_input_tokens_seen": 108795045, + "step": 3342 + }, + { + "epoch": 0.30148351896108583, + "flos": 24385488529440.0, + "grad_norm": 1.4934954278682457, + "learning_rate": 3.2754780717142233e-06, + "loss": 0.8048, + "num_input_tokens_seen": 108823055, + "step": 3343 + }, + { + "epoch": 0.30157370248455606, + "flos": 19690654244160.0, + "grad_norm": 1.9770852102653296, + "learning_rate": 3.27502802760507e-06, + "loss": 0.741, + "num_input_tokens_seen": 108850920, + "step": 3344 + }, + { + "epoch": 0.30166388600802635, + "flos": 28252234537920.0, + "grad_norm": 1.7459813024168622, + "learning_rate": 3.2745778747034943e-06, + "loss": 0.6796, + "num_input_tokens_seen": 108882075, + "step": 3345 + }, + { + "epoch": 0.3017540695314966, + "flos": 24391138332960.0, + "grad_norm": 2.062078490679051, + "learning_rate": 3.274127613047906e-06, + "loss": 0.7463, + "num_input_tokens_seen": 108910880, + "step": 3346 + }, + { + "epoch": 0.30184425305496687, + "flos": 22642248930720.0, + "grad_norm": 1.9830133767727631, + "learning_rate": 3.273677242676725e-06, + "loss": 0.7834, + "num_input_tokens_seen": 108939930, + "step": 3347 + }, + { + "epoch": 0.3019344365784371, + "flos": 23808187588800.0, + "grad_norm": 2.40844606533881, + "learning_rate": 3.2732267636283782e-06, + "loss": 0.7653, + "num_input_tokens_seen": 108969380, + "step": 3348 + }, + { + "epoch": 0.3020246201019074, + "flos": 16735045223520.0, + "grad_norm": 2.6552316845455315, + "learning_rate": 3.2727761759413034e-06, + "loss": 0.7297, + "num_input_tokens_seen": 108995500, + "step": 3349 + }, + { + "epoch": 0.3021148036253776, + "flos": 20087467332000.0, + "grad_norm": 1.729027003274315, + "learning_rate": 3.2723254796539477e-06, + "loss": 0.7787, + "num_input_tokens_seen": 109023150, + "step": 3350 + }, + { + "epoch": 0.3022049871488479, + "flos": 36632174716320.0, + "grad_norm": 8.990259197432058, + "learning_rate": 3.271874674804766e-06, + "loss": 0.7552, + "num_input_tokens_seen": 109053995, + "step": 3351 + }, + { + "epoch": 0.30229517067231815, + "flos": 19581989834880.0, + "grad_norm": 3.0923228970068806, + "learning_rate": 3.2714237614322242e-06, + "loss": 0.7912, + "num_input_tokens_seen": 109082595, + "step": 3352 + }, + { + "epoch": 0.30238535419578844, + "flos": 22165843919520.0, + "grad_norm": 1.7769385242679923, + "learning_rate": 3.2709727395747974e-06, + "loss": 0.6039, + "num_input_tokens_seen": 109114065, + "step": 3353 + }, + { + "epoch": 0.30247553771925867, + "flos": 68515756207200.0, + "grad_norm": 0.6812241601855566, + "learning_rate": 3.2705216092709673e-06, + "loss": 0.5717, + "num_input_tokens_seen": 109205010, + "step": 3354 + }, + { + "epoch": 0.30256572124272896, + "flos": 17904291990240.0, + "grad_norm": 2.0635619464040156, + "learning_rate": 3.2700703705592282e-06, + "loss": 0.8533, + "num_input_tokens_seen": 109228635, + "step": 3355 + }, + { + "epoch": 0.3026559047661992, + "flos": 22496148839520.0, + "grad_norm": 1.8010923567861088, + "learning_rate": 3.269619023478082e-06, + "loss": 0.6683, + "num_input_tokens_seen": 109256625, + "step": 3356 + }, + { + "epoch": 0.3027460882896695, + "flos": 24312698672160.0, + "grad_norm": 2.0610990679315657, + "learning_rate": 3.26916756806604e-06, + "loss": 0.8651, + "num_input_tokens_seen": 109284230, + "step": 3357 + }, + { + "epoch": 0.3028362718131397, + "flos": 12987965873280.0, + "grad_norm": 1.6471617041941553, + "learning_rate": 3.268716004361623e-06, + "loss": 0.7834, + "num_input_tokens_seen": 109311410, + "step": 3358 + }, + { + "epoch": 0.30292645533661, + "flos": 65904650954880.0, + "grad_norm": 0.6562092859263414, + "learning_rate": 3.268264332403361e-06, + "loss": 0.5869, + "num_input_tokens_seen": 109412370, + "step": 3359 + }, + { + "epoch": 0.3030166388600803, + "flos": 25483618078080.0, + "grad_norm": 3.2746557913867815, + "learning_rate": 3.2678125522297933e-06, + "loss": 0.7984, + "num_input_tokens_seen": 109443160, + "step": 3360 + }, + { + "epoch": 0.3031068223835505, + "flos": 26285384431200.0, + "grad_norm": 2.281011595091559, + "learning_rate": 3.267360663879468e-06, + "loss": 0.707, + "num_input_tokens_seen": 109470535, + "step": 3361 + }, + { + "epoch": 0.3031970059070208, + "flos": 30111795516480.0, + "grad_norm": 2.5902352475881223, + "learning_rate": 3.266908667390942e-06, + "loss": 0.7564, + "num_input_tokens_seen": 109501925, + "step": 3362 + }, + { + "epoch": 0.30328718943049104, + "flos": 25847641704000.0, + "grad_norm": 1.7780329112256301, + "learning_rate": 3.2664565628027833e-06, + "loss": 0.8212, + "num_input_tokens_seen": 109527155, + "step": 3363 + }, + { + "epoch": 0.30337737295396133, + "flos": 22386703865280.0, + "grad_norm": 2.7052833892394657, + "learning_rate": 3.2660043501535675e-06, + "loss": 0.7654, + "num_input_tokens_seen": 109553090, + "step": 3364 + }, + { + "epoch": 0.30346755647743157, + "flos": 21257494663680.0, + "grad_norm": 2.3004579171151924, + "learning_rate": 3.2655520294818797e-06, + "loss": 0.8376, + "num_input_tokens_seen": 109577210, + "step": 3365 + }, + { + "epoch": 0.30355774000090185, + "flos": 22092013308960.0, + "grad_norm": 1.5333850533981552, + "learning_rate": 3.2650996008263146e-06, + "loss": 0.8367, + "num_input_tokens_seen": 109608025, + "step": 3366 + }, + { + "epoch": 0.3036479235243721, + "flos": 22930434779040.0, + "grad_norm": 2.193521223610329, + "learning_rate": 3.2646470642254756e-06, + "loss": 0.7266, + "num_input_tokens_seen": 109637105, + "step": 3367 + }, + { + "epoch": 0.3037381070478424, + "flos": 21804236328000.0, + "grad_norm": 2.026764127161431, + "learning_rate": 3.2641944197179767e-06, + "loss": 0.7883, + "num_input_tokens_seen": 109664650, + "step": 3368 + }, + { + "epoch": 0.3038282905713126, + "flos": 55394365130880.0, + "grad_norm": 0.6624349958090117, + "learning_rate": 3.2637416673424383e-06, + "loss": 0.572, + "num_input_tokens_seen": 109739905, + "step": 3369 + }, + { + "epoch": 0.3039184740947829, + "flos": 24132322328640.0, + "grad_norm": 2.7691304894777216, + "learning_rate": 3.2632888071374937e-06, + "loss": 0.7791, + "num_input_tokens_seen": 109769190, + "step": 3370 + }, + { + "epoch": 0.30400865761825313, + "flos": 15574979387520.0, + "grad_norm": 1.8303742225195676, + "learning_rate": 3.2628358391417815e-06, + "loss": 0.7978, + "num_input_tokens_seen": 109797120, + "step": 3371 + }, + { + "epoch": 0.3040988411417234, + "flos": 20784026063520.0, + "grad_norm": 1.7697877003350435, + "learning_rate": 3.2623827633939526e-06, + "loss": 0.7591, + "num_input_tokens_seen": 109825965, + "step": 3372 + }, + { + "epoch": 0.30418902466519365, + "flos": 22638308936160.0, + "grad_norm": 1.8886453021131877, + "learning_rate": 3.2619295799326657e-06, + "loss": 0.7597, + "num_input_tokens_seen": 109854420, + "step": 3373 + }, + { + "epoch": 0.30427920818866394, + "flos": 32076341098080.0, + "grad_norm": 2.3655693138351177, + "learning_rate": 3.2614762887965883e-06, + "loss": 0.6667, + "num_input_tokens_seen": 109880025, + "step": 3374 + }, + { + "epoch": 0.3043693917121342, + "flos": 13351654971360.0, + "grad_norm": 1.8603208043717145, + "learning_rate": 3.2610228900243984e-06, + "loss": 0.7555, + "num_input_tokens_seen": 109905225, + "step": 3375 + }, + { + "epoch": 0.30445957523560446, + "flos": 70044306040320.0, + "grad_norm": 0.6074888676930273, + "learning_rate": 3.260569383654783e-06, + "loss": 0.6234, + "num_input_tokens_seen": 110000745, + "step": 3376 + }, + { + "epoch": 0.3045497587590747, + "flos": 70303345063200.0, + "grad_norm": 0.6907053050651961, + "learning_rate": 3.2601157697264365e-06, + "loss": 0.5939, + "num_input_tokens_seen": 110091210, + "step": 3377 + }, + { + "epoch": 0.304639942282545, + "flos": 20347101071040.0, + "grad_norm": 1.85227611326556, + "learning_rate": 3.2596620482780647e-06, + "loss": 0.7447, + "num_input_tokens_seen": 110118245, + "step": 3378 + }, + { + "epoch": 0.3047301258060152, + "flos": 62647103274720.0, + "grad_norm": 0.5820937155159712, + "learning_rate": 3.2592082193483803e-06, + "loss": 0.5924, + "num_input_tokens_seen": 110216100, + "step": 3379 + }, + { + "epoch": 0.3048203093294855, + "flos": 24464225548320.0, + "grad_norm": 1.7550192107537588, + "learning_rate": 3.258754282976109e-06, + "loss": 0.7537, + "num_input_tokens_seen": 110246625, + "step": 3380 + }, + { + "epoch": 0.30491049285295574, + "flos": 17724621872160.0, + "grad_norm": 2.1377282137928906, + "learning_rate": 3.25830023919998e-06, + "loss": 0.8398, + "num_input_tokens_seen": 110274335, + "step": 3381 + }, + { + "epoch": 0.305000676376426, + "flos": 25695297093120.0, + "grad_norm": 1.7976296871764266, + "learning_rate": 3.2578460880587374e-06, + "loss": 0.7224, + "num_input_tokens_seen": 110303405, + "step": 3382 + }, + { + "epoch": 0.3050908598998963, + "flos": 19946273649120.0, + "grad_norm": 2.592862356496686, + "learning_rate": 3.2573918295911306e-06, + "loss": 0.821, + "num_input_tokens_seen": 110330455, + "step": 3383 + }, + { + "epoch": 0.30518104342336655, + "flos": 19581989834880.0, + "grad_norm": 2.127299803336988, + "learning_rate": 3.2569374638359196e-06, + "loss": 0.7077, + "num_input_tokens_seen": 110358145, + "step": 3384 + }, + { + "epoch": 0.30527122694683684, + "flos": 24207230862240.0, + "grad_norm": 1.3972082009485167, + "learning_rate": 3.2564829908318736e-06, + "loss": 0.7803, + "num_input_tokens_seen": 110388945, + "step": 3385 + }, + { + "epoch": 0.30536141047030707, + "flos": 18995545133280.0, + "grad_norm": 1.3359359384271072, + "learning_rate": 3.2560284106177705e-06, + "loss": 0.8259, + "num_input_tokens_seen": 110417590, + "step": 3386 + }, + { + "epoch": 0.30545159399377736, + "flos": 17688115434240.0, + "grad_norm": 1.7139304820488324, + "learning_rate": 3.2555737232323978e-06, + "loss": 0.7386, + "num_input_tokens_seen": 110443600, + "step": 3387 + }, + { + "epoch": 0.3055417775172476, + "flos": 26827256856960.0, + "grad_norm": 1.9155066200008524, + "learning_rate": 3.255118928714552e-06, + "loss": 0.7977, + "num_input_tokens_seen": 110474855, + "step": 3388 + }, + { + "epoch": 0.3056319610407179, + "flos": 19909358343840.0, + "grad_norm": 2.727575775128864, + "learning_rate": 3.2546640271030386e-06, + "loss": 0.6965, + "num_input_tokens_seen": 110501140, + "step": 3389 + }, + { + "epoch": 0.3057221445641881, + "flos": 70476770661600.0, + "grad_norm": 0.6226510970262608, + "learning_rate": 3.2542090184366717e-06, + "loss": 0.6244, + "num_input_tokens_seen": 110592975, + "step": 3390 + }, + { + "epoch": 0.3058123280876584, + "flos": 21652189075200.0, + "grad_norm": 2.3856542432880055, + "learning_rate": 3.253753902754276e-06, + "loss": 0.8504, + "num_input_tokens_seen": 110618905, + "step": 3391 + }, + { + "epoch": 0.30590251161112864, + "flos": 46471629016320.0, + "grad_norm": 1.3426099865007113, + "learning_rate": 3.253298680094685e-06, + "loss": 0.6995, + "num_input_tokens_seen": 110658605, + "step": 3392 + }, + { + "epoch": 0.3059926951345989, + "flos": 21257717682240.0, + "grad_norm": 4.979532693899034, + "learning_rate": 3.2528433504967394e-06, + "loss": 0.7132, + "num_input_tokens_seen": 110687475, + "step": 3393 + }, + { + "epoch": 0.30608287865806916, + "flos": 23225348353920.0, + "grad_norm": 2.711841091868022, + "learning_rate": 3.252387913999291e-06, + "loss": 0.7812, + "num_input_tokens_seen": 110715865, + "step": 3394 + }, + { + "epoch": 0.30617306218153945, + "flos": 26066568822240.0, + "grad_norm": 1.7243334159524129, + "learning_rate": 3.2519323706411998e-06, + "loss": 0.7702, + "num_input_tokens_seen": 110745160, + "step": 3395 + }, + { + "epoch": 0.3062632457050097, + "flos": 30504148233120.0, + "grad_norm": 1.5433775541315766, + "learning_rate": 3.251476720461336e-06, + "loss": 0.6568, + "num_input_tokens_seen": 110774795, + "step": 3396 + }, + { + "epoch": 0.30635342922847997, + "flos": 14114164324320.0, + "grad_norm": 1.8439274072156568, + "learning_rate": 3.251020963498578e-06, + "loss": 0.6967, + "num_input_tokens_seen": 110801090, + "step": 3397 + }, + { + "epoch": 0.3064436127519502, + "flos": 23114007721920.0, + "grad_norm": 1.524900702073444, + "learning_rate": 3.250565099791813e-06, + "loss": 0.812, + "num_input_tokens_seen": 110828475, + "step": 3398 + }, + { + "epoch": 0.3065337962754205, + "flos": 24607240549440.0, + "grad_norm": 1.5852231075049426, + "learning_rate": 3.2501091293799387e-06, + "loss": 0.8227, + "num_input_tokens_seen": 110858965, + "step": 3399 + }, + { + "epoch": 0.3066239797988907, + "flos": 21694754184000.0, + "grad_norm": 1.7768715702846298, + "learning_rate": 3.24965305230186e-06, + "loss": 0.7812, + "num_input_tokens_seen": 110886510, + "step": 3400 + }, + { + "epoch": 0.306714163322361, + "flos": 11676261651840.0, + "grad_norm": 2.1125440739655637, + "learning_rate": 3.249196868596492e-06, + "loss": 0.8399, + "num_input_tokens_seen": 110909435, + "step": 3401 + }, + { + "epoch": 0.30680434684583124, + "flos": 31130704839360.0, + "grad_norm": 3.453606343272492, + "learning_rate": 3.24874057830276e-06, + "loss": 0.6958, + "num_input_tokens_seen": 110938050, + "step": 3402 + }, + { + "epoch": 0.30689453036930153, + "flos": 33458196123840.0, + "grad_norm": 1.7799029024836994, + "learning_rate": 3.2482841814595954e-06, + "loss": 0.6883, + "num_input_tokens_seen": 110967790, + "step": 3403 + }, + { + "epoch": 0.30698471389277177, + "flos": 27159234416160.0, + "grad_norm": 2.425508984506827, + "learning_rate": 3.247827678105943e-06, + "loss": 0.7107, + "num_input_tokens_seen": 110997015, + "step": 3404 + }, + { + "epoch": 0.30707489741624205, + "flos": 24063769824000.0, + "grad_norm": 1.6517135103564993, + "learning_rate": 3.247371068280751e-06, + "loss": 0.7661, + "num_input_tokens_seen": 111027495, + "step": 3405 + }, + { + "epoch": 0.3071650809397123, + "flos": 23512047411840.0, + "grad_norm": 1.5372957264490368, + "learning_rate": 3.2469143520229823e-06, + "loss": 0.6262, + "num_input_tokens_seen": 111057735, + "step": 3406 + }, + { + "epoch": 0.3072552644631826, + "flos": 38709361871520.0, + "grad_norm": 2.214058657840041, + "learning_rate": 3.2464575293716054e-06, + "loss": 0.717, + "num_input_tokens_seen": 111088950, + "step": 3407 + }, + { + "epoch": 0.30734544798665286, + "flos": 22241718866880.0, + "grad_norm": 1.9732915176437404, + "learning_rate": 3.2460006003655997e-06, + "loss": 0.7076, + "num_input_tokens_seen": 111116475, + "step": 3408 + }, + { + "epoch": 0.3074356315101231, + "flos": 21875167697280.0, + "grad_norm": 2.1164853666661654, + "learning_rate": 3.245543565043952e-06, + "loss": 0.8207, + "num_input_tokens_seen": 111140025, + "step": 3409 + }, + { + "epoch": 0.3075258150335934, + "flos": 18523191625920.0, + "grad_norm": 2.159706882968848, + "learning_rate": 3.2450864234456592e-06, + "loss": 0.761, + "num_input_tokens_seen": 111166145, + "step": 3410 + }, + { + "epoch": 0.3076159985570636, + "flos": 37718298432480.0, + "grad_norm": 1.4577007647478644, + "learning_rate": 3.244629175609728e-06, + "loss": 0.7588, + "num_input_tokens_seen": 111195965, + "step": 3411 + }, + { + "epoch": 0.3077061820805339, + "flos": 22197927156000.0, + "grad_norm": 2.1990328179964127, + "learning_rate": 3.2441718215751726e-06, + "loss": 0.7509, + "num_input_tokens_seen": 111223740, + "step": 3412 + }, + { + "epoch": 0.30779636560400414, + "flos": 18889408267680.0, + "grad_norm": 1.7959682664680232, + "learning_rate": 3.2437143613810173e-06, + "loss": 0.7272, + "num_input_tokens_seen": 111249990, + "step": 3413 + }, + { + "epoch": 0.30788654912747443, + "flos": 28325396092800.0, + "grad_norm": 1.6319973749906336, + "learning_rate": 3.2432567950662947e-06, + "loss": 0.8092, + "num_input_tokens_seen": 111281660, + "step": 3414 + }, + { + "epoch": 0.30797673265094466, + "flos": 24756425730720.0, + "grad_norm": 2.1601636448435246, + "learning_rate": 3.2427991226700468e-06, + "loss": 0.7528, + "num_input_tokens_seen": 111312705, + "step": 3415 + }, + { + "epoch": 0.30806691617441495, + "flos": 20747668304640.0, + "grad_norm": 1.890979260759084, + "learning_rate": 3.2423413442313246e-06, + "loss": 0.6892, + "num_input_tokens_seen": 111340865, + "step": 3416 + }, + { + "epoch": 0.3081570996978852, + "flos": 25700426520000.0, + "grad_norm": 1.8955859207969918, + "learning_rate": 3.2418834597891904e-06, + "loss": 0.6815, + "num_input_tokens_seen": 111369030, + "step": 3417 + }, + { + "epoch": 0.3082472832213555, + "flos": 14626964264160.0, + "grad_norm": 2.3513734050775943, + "learning_rate": 3.2414254693827098e-06, + "loss": 0.854, + "num_input_tokens_seen": 111394090, + "step": 3418 + }, + { + "epoch": 0.3083374667448257, + "flos": 21948106233600.0, + "grad_norm": 1.5425285387413163, + "learning_rate": 3.2409673730509644e-06, + "loss": 0.7409, + "num_input_tokens_seen": 111423825, + "step": 3419 + }, + { + "epoch": 0.308427650268296, + "flos": 28545215285280.0, + "grad_norm": 2.074376791700122, + "learning_rate": 3.2405091708330393e-06, + "loss": 0.7402, + "num_input_tokens_seen": 111450475, + "step": 3420 + }, + { + "epoch": 0.3085178337917662, + "flos": 29268541977600.0, + "grad_norm": 1.9319170839828415, + "learning_rate": 3.2400508627680323e-06, + "loss": 0.754, + "num_input_tokens_seen": 111481685, + "step": 3421 + }, + { + "epoch": 0.3086080173152365, + "flos": 15246644464800.0, + "grad_norm": 1.8984616752937638, + "learning_rate": 3.2395924488950474e-06, + "loss": 0.858, + "num_input_tokens_seen": 111505800, + "step": 3422 + }, + { + "epoch": 0.30869820083870675, + "flos": 22124839940640.0, + "grad_norm": 2.0253680236202793, + "learning_rate": 3.2391339292532004e-06, + "loss": 0.8932, + "num_input_tokens_seen": 111532870, + "step": 3423 + }, + { + "epoch": 0.30878838436217704, + "flos": 34373421785280.0, + "grad_norm": 1.940372798337242, + "learning_rate": 3.238675303881614e-06, + "loss": 0.7171, + "num_input_tokens_seen": 111563965, + "step": 3424 + }, + { + "epoch": 0.30887856788564727, + "flos": 20710827338880.0, + "grad_norm": 1.6101011101148557, + "learning_rate": 3.2382165728194203e-06, + "loss": 0.7323, + "num_input_tokens_seen": 111591700, + "step": 3425 + }, + { + "epoch": 0.30896875140911756, + "flos": 35282774624640.0, + "grad_norm": 1.475952587611911, + "learning_rate": 3.237757736105761e-06, + "loss": 0.7449, + "num_input_tokens_seen": 111623395, + "step": 3426 + }, + { + "epoch": 0.3090589349325878, + "flos": 25848682457280.0, + "grad_norm": 1.5285269990532444, + "learning_rate": 3.2372987937797867e-06, + "loss": 0.8455, + "num_input_tokens_seen": 111654700, + "step": 3427 + }, + { + "epoch": 0.3091491184560581, + "flos": 19253134535520.0, + "grad_norm": 1.5680400290488712, + "learning_rate": 3.2368397458806573e-06, + "loss": 0.6719, + "num_input_tokens_seen": 111683215, + "step": 3428 + }, + { + "epoch": 0.3092393019795283, + "flos": 29928222573600.0, + "grad_norm": 2.08194753968824, + "learning_rate": 3.2363805924475412e-06, + "loss": 0.7973, + "num_input_tokens_seen": 111709290, + "step": 3429 + }, + { + "epoch": 0.3093294855029986, + "flos": 25659050843520.0, + "grad_norm": 2.3144834106009005, + "learning_rate": 3.2359213335196153e-06, + "loss": 0.7027, + "num_input_tokens_seen": 111736635, + "step": 3430 + }, + { + "epoch": 0.3094196690264689, + "flos": 21402888529440.0, + "grad_norm": 1.845125687305929, + "learning_rate": 3.2354619691360663e-06, + "loss": 0.7193, + "num_input_tokens_seen": 111766020, + "step": 3431 + }, + { + "epoch": 0.3095098525499391, + "flos": 26540260440960.0, + "grad_norm": 1.778165790646437, + "learning_rate": 3.2350024993360898e-06, + "loss": 0.8012, + "num_input_tokens_seen": 111796230, + "step": 3432 + }, + { + "epoch": 0.3096000360734094, + "flos": 21512891050080.0, + "grad_norm": 1.9159343052832618, + "learning_rate": 3.2345429241588902e-06, + "loss": 0.7643, + "num_input_tokens_seen": 111826335, + "step": 3433 + }, + { + "epoch": 0.30969021959687965, + "flos": 23844173650080.0, + "grad_norm": 2.017782904396229, + "learning_rate": 3.234083243643681e-06, + "loss": 0.7883, + "num_input_tokens_seen": 111853605, + "step": 3434 + }, + { + "epoch": 0.30978040312034993, + "flos": 18597914310720.0, + "grad_norm": 1.6056119015026857, + "learning_rate": 3.233623457829686e-06, + "loss": 0.8337, + "num_input_tokens_seen": 111879150, + "step": 3435 + }, + { + "epoch": 0.30987058664382017, + "flos": 22788274682400.0, + "grad_norm": 2.011067405363937, + "learning_rate": 3.2331635667561344e-06, + "loss": 0.7315, + "num_input_tokens_seen": 111907480, + "step": 3436 + }, + { + "epoch": 0.30996077016729046, + "flos": 25447483337760.0, + "grad_norm": 1.6853300200431784, + "learning_rate": 3.2327035704622674e-06, + "loss": 0.7953, + "num_input_tokens_seen": 111934770, + "step": 3437 + }, + { + "epoch": 0.3100509536907607, + "flos": 33682029650400.0, + "grad_norm": 1.4365576267387976, + "learning_rate": 3.2322434689873353e-06, + "loss": 0.6613, + "num_input_tokens_seen": 111968580, + "step": 3438 + }, + { + "epoch": 0.310141137214231, + "flos": 20048544859680.0, + "grad_norm": 2.0011322869415924, + "learning_rate": 3.2317832623705957e-06, + "loss": 0.8438, + "num_input_tokens_seen": 111995000, + "step": 3439 + }, + { + "epoch": 0.3102313207377012, + "flos": 50403136213920.0, + "grad_norm": 1.9818605126146445, + "learning_rate": 3.231322950651316e-06, + "loss": 0.7092, + "num_input_tokens_seen": 112027145, + "step": 3440 + }, + { + "epoch": 0.3103215042611715, + "flos": 24096782304480.0, + "grad_norm": 1.4815088810417956, + "learning_rate": 3.2308625338687735e-06, + "loss": 0.7905, + "num_input_tokens_seen": 112057565, + "step": 3441 + }, + { + "epoch": 0.31041168778464173, + "flos": 12250700520960.0, + "grad_norm": 2.498908758420394, + "learning_rate": 3.230402012062252e-06, + "loss": 0.8644, + "num_input_tokens_seen": 112078890, + "step": 3442 + }, + { + "epoch": 0.310501871308112, + "flos": 66401170539840.0, + "grad_norm": 0.6117467208839896, + "learning_rate": 3.2299413852710466e-06, + "loss": 0.6131, + "num_input_tokens_seen": 112174140, + "step": 3443 + }, + { + "epoch": 0.31059205483158225, + "flos": 18671336053920.0, + "grad_norm": 2.2028677499341516, + "learning_rate": 3.2294806535344606e-06, + "loss": 0.7198, + "num_input_tokens_seen": 112201685, + "step": 3444 + }, + { + "epoch": 0.31068223835505254, + "flos": 23662199006880.0, + "grad_norm": 1.9100868508929383, + "learning_rate": 3.2290198168918056e-06, + "loss": 0.8034, + "num_input_tokens_seen": 112230215, + "step": 3445 + }, + { + "epoch": 0.3107724218785228, + "flos": 22204952240640.0, + "grad_norm": 1.7696477249582272, + "learning_rate": 3.2285588753824035e-06, + "loss": 0.8008, + "num_input_tokens_seen": 112258395, + "step": 3446 + }, + { + "epoch": 0.31086260540199306, + "flos": 17283608206080.0, + "grad_norm": 1.8127426774247042, + "learning_rate": 3.228097829045584e-06, + "loss": 0.7767, + "num_input_tokens_seen": 112284925, + "step": 3447 + }, + { + "epoch": 0.3109527889254633, + "flos": 25482837513120.0, + "grad_norm": 1.8047176870488357, + "learning_rate": 3.227636677920685e-06, + "loss": 0.8046, + "num_input_tokens_seen": 112316215, + "step": 3448 + }, + { + "epoch": 0.3110429724489336, + "flos": 23844954215040.0, + "grad_norm": 1.7235476032408292, + "learning_rate": 3.2271754220470567e-06, + "loss": 0.8101, + "num_input_tokens_seen": 112344380, + "step": 3449 + }, + { + "epoch": 0.3111331559724038, + "flos": 16003764202560.0, + "grad_norm": 3.2856989173055866, + "learning_rate": 3.2267140614640547e-06, + "loss": 0.8452, + "num_input_tokens_seen": 112368610, + "step": 3450 + }, + { + "epoch": 0.3112233394958741, + "flos": 22532729616960.0, + "grad_norm": 2.145821383051018, + "learning_rate": 3.2262525962110445e-06, + "loss": 0.8081, + "num_input_tokens_seen": 112397450, + "step": 3451 + }, + { + "epoch": 0.31131352301934434, + "flos": 23880977446080.0, + "grad_norm": 1.9770643479213146, + "learning_rate": 3.2257910263274015e-06, + "loss": 0.691, + "num_input_tokens_seen": 112425880, + "step": 3452 + }, + { + "epoch": 0.31140370654281463, + "flos": 24094663628160.0, + "grad_norm": 2.183829798078389, + "learning_rate": 3.225329351852509e-06, + "loss": 0.7117, + "num_input_tokens_seen": 112454495, + "step": 3453 + }, + { + "epoch": 0.3114938900662849, + "flos": 23042964843360.0, + "grad_norm": 1.8460592905154782, + "learning_rate": 3.2248675728257596e-06, + "loss": 0.7684, + "num_input_tokens_seen": 112483970, + "step": 3454 + }, + { + "epoch": 0.31158407358975515, + "flos": 25914633078720.0, + "grad_norm": 1.649480552981916, + "learning_rate": 3.2244056892865557e-06, + "loss": 0.8386, + "num_input_tokens_seen": 112512685, + "step": 3455 + }, + { + "epoch": 0.31167425711322544, + "flos": 29196235327200.0, + "grad_norm": 1.5590927357818394, + "learning_rate": 3.2239437012743063e-06, + "loss": 0.7551, + "num_input_tokens_seen": 112540850, + "step": 3456 + }, + { + "epoch": 0.31176444063669567, + "flos": 17869792719360.0, + "grad_norm": 1.5646281436170264, + "learning_rate": 3.223481608828432e-06, + "loss": 0.7805, + "num_input_tokens_seen": 112568755, + "step": 3457 + }, + { + "epoch": 0.31185462416016596, + "flos": 17832431376960.0, + "grad_norm": 2.139948382958354, + "learning_rate": 3.223019411988361e-06, + "loss": 0.7799, + "num_input_tokens_seen": 112596620, + "step": 3458 + }, + { + "epoch": 0.3119448076836362, + "flos": 40056383098560.0, + "grad_norm": 1.8119741532038094, + "learning_rate": 3.22255711079353e-06, + "loss": 0.7642, + "num_input_tokens_seen": 112629630, + "step": 3459 + }, + { + "epoch": 0.3120349912071065, + "flos": 20853582151680.0, + "grad_norm": 1.7379650042602484, + "learning_rate": 3.222094705283385e-06, + "loss": 0.7269, + "num_input_tokens_seen": 112657420, + "step": 3460 + }, + { + "epoch": 0.3121251747305767, + "flos": 28281901740000.0, + "grad_norm": 1.8283731015081361, + "learning_rate": 3.2216321954973805e-06, + "loss": 0.7599, + "num_input_tokens_seen": 112682700, + "step": 3461 + }, + { + "epoch": 0.312215358254047, + "flos": 23297766513600.0, + "grad_norm": 3.3974111125173025, + "learning_rate": 3.2211695814749816e-06, + "loss": 0.7514, + "num_input_tokens_seen": 112713950, + "step": 3462 + }, + { + "epoch": 0.31230554177751724, + "flos": 15676470033120.0, + "grad_norm": 4.518251493400649, + "learning_rate": 3.220706863255661e-06, + "loss": 0.8695, + "num_input_tokens_seen": 112736145, + "step": 3463 + }, + { + "epoch": 0.3123957253009875, + "flos": 21877286373600.0, + "grad_norm": 3.0202386380324424, + "learning_rate": 3.2202440408788994e-06, + "loss": 0.7638, + "num_input_tokens_seen": 112764045, + "step": 3464 + }, + { + "epoch": 0.31248590882445776, + "flos": 18160803469440.0, + "grad_norm": 1.7570833855709544, + "learning_rate": 3.2197811143841883e-06, + "loss": 0.8359, + "num_input_tokens_seen": 112790135, + "step": 3465 + }, + { + "epoch": 0.31257609234792805, + "flos": 21689810605920.0, + "grad_norm": 1.6530747115737623, + "learning_rate": 3.2193180838110267e-06, + "loss": 0.762, + "num_input_tokens_seen": 112818570, + "step": 3466 + }, + { + "epoch": 0.3126662758713983, + "flos": 21691929282240.0, + "grad_norm": 2.30224800959292, + "learning_rate": 3.2188549491989225e-06, + "loss": 0.6602, + "num_input_tokens_seen": 112845340, + "step": 3467 + }, + { + "epoch": 0.31275645939486857, + "flos": 22605333625440.0, + "grad_norm": 2.2578513594190537, + "learning_rate": 3.2183917105873934e-06, + "loss": 0.7905, + "num_input_tokens_seen": 112873590, + "step": 3468 + }, + { + "epoch": 0.3128466429183388, + "flos": 22134689927040.0, + "grad_norm": 1.9558155296413637, + "learning_rate": 3.217928368015966e-06, + "loss": 0.8297, + "num_input_tokens_seen": 112906045, + "step": 3469 + }, + { + "epoch": 0.3129368264418091, + "flos": 23188730406720.0, + "grad_norm": 1.4310917029650938, + "learning_rate": 3.217464921524174e-06, + "loss": 0.794, + "num_input_tokens_seen": 112936380, + "step": 3470 + }, + { + "epoch": 0.3130270099652793, + "flos": 22456148444160.0, + "grad_norm": 4.853814086333796, + "learning_rate": 3.2170013711515616e-06, + "loss": 0.7793, + "num_input_tokens_seen": 112962270, + "step": 3471 + }, + { + "epoch": 0.3131171934887496, + "flos": 28361976870240.0, + "grad_norm": 2.983601734583205, + "learning_rate": 3.216537716937682e-06, + "loss": 0.6886, + "num_input_tokens_seen": 112993380, + "step": 3472 + }, + { + "epoch": 0.31320737701221985, + "flos": 23626882001280.0, + "grad_norm": 2.2503221069351245, + "learning_rate": 3.2160739589220968e-06, + "loss": 0.7411, + "num_input_tokens_seen": 113021675, + "step": 3473 + }, + { + "epoch": 0.31329756053569013, + "flos": 21987363233760.0, + "grad_norm": 2.036606753144938, + "learning_rate": 3.215610097144376e-06, + "loss": 0.824, + "num_input_tokens_seen": 113051305, + "step": 3474 + }, + { + "epoch": 0.31338774405916037, + "flos": 23698668275040.0, + "grad_norm": 2.006970893424522, + "learning_rate": 3.215146131644099e-06, + "loss": 0.7181, + "num_input_tokens_seen": 113079695, + "step": 3475 + }, + { + "epoch": 0.31347792758263066, + "flos": 19867610969760.0, + "grad_norm": 1.5385178120114784, + "learning_rate": 3.214682062460854e-06, + "loss": 0.8354, + "num_input_tokens_seen": 113107580, + "step": 3476 + }, + { + "epoch": 0.3135681111061009, + "flos": 23479964175360.0, + "grad_norm": 2.1858365955199974, + "learning_rate": 3.2142178896342367e-06, + "loss": 0.6818, + "num_input_tokens_seen": 113136300, + "step": 3477 + }, + { + "epoch": 0.3136582946295712, + "flos": 36337595669280.0, + "grad_norm": 1.60422395215817, + "learning_rate": 3.2137536132038552e-06, + "loss": 0.7528, + "num_input_tokens_seen": 113168390, + "step": 3478 + }, + { + "epoch": 0.31374847815304147, + "flos": 19866718895520.0, + "grad_norm": 1.7014734502352566, + "learning_rate": 3.2132892332093226e-06, + "loss": 0.7725, + "num_input_tokens_seen": 113194320, + "step": 3479 + }, + { + "epoch": 0.3138386616765117, + "flos": 26828371949760.0, + "grad_norm": 1.615783007806822, + "learning_rate": 3.2128247496902623e-06, + "loss": 0.8753, + "num_input_tokens_seen": 113224780, + "step": 3480 + }, + { + "epoch": 0.313928845199982, + "flos": 24572258071680.0, + "grad_norm": 1.4056854386352062, + "learning_rate": 3.2123601626863064e-06, + "loss": 0.7767, + "num_input_tokens_seen": 113257860, + "step": 3481 + }, + { + "epoch": 0.3140190287234522, + "flos": 64946525656800.0, + "grad_norm": 0.6359824621628541, + "learning_rate": 3.2118954722370974e-06, + "loss": 0.5705, + "num_input_tokens_seen": 113352535, + "step": 3482 + }, + { + "epoch": 0.3141092122469225, + "flos": 22241756036640.0, + "grad_norm": 1.903961872075697, + "learning_rate": 3.2114306783822837e-06, + "loss": 0.7102, + "num_input_tokens_seen": 113380060, + "step": 3483 + }, + { + "epoch": 0.31419939577039274, + "flos": 26829301193760.0, + "grad_norm": 1.7294725430976032, + "learning_rate": 3.210965781161525e-06, + "loss": 0.7442, + "num_input_tokens_seen": 113410905, + "step": 3484 + }, + { + "epoch": 0.31428957929386303, + "flos": 25739051634240.0, + "grad_norm": 1.8747784801598322, + "learning_rate": 3.2105007806144892e-06, + "loss": 0.8806, + "num_input_tokens_seen": 113438755, + "step": 3485 + }, + { + "epoch": 0.31437976281733326, + "flos": 23079694299840.0, + "grad_norm": 2.720742208250779, + "learning_rate": 3.2100356767808513e-06, + "loss": 0.7677, + "num_input_tokens_seen": 113468120, + "step": 3486 + }, + { + "epoch": 0.31446994634080355, + "flos": 20383681848480.0, + "grad_norm": 2.0195450625509874, + "learning_rate": 3.2095704697002977e-06, + "loss": 0.82, + "num_input_tokens_seen": 113493670, + "step": 3487 + }, + { + "epoch": 0.3145601298642738, + "flos": 21076151906400.0, + "grad_norm": 1.857481625595704, + "learning_rate": 3.209105159412522e-06, + "loss": 0.7651, + "num_input_tokens_seen": 113523385, + "step": 3488 + }, + { + "epoch": 0.3146503133877441, + "flos": 20960871279840.0, + "grad_norm": 2.2663939124911954, + "learning_rate": 3.208639745957228e-06, + "loss": 0.7273, + "num_input_tokens_seen": 113549440, + "step": 3489 + }, + { + "epoch": 0.3147404969112143, + "flos": 25593360410400.0, + "grad_norm": 1.4348324613429309, + "learning_rate": 3.2081742293741256e-06, + "loss": 0.7934, + "num_input_tokens_seen": 113581475, + "step": 3490 + }, + { + "epoch": 0.3148306804346846, + "flos": 22387484430240.0, + "grad_norm": 1.4434625281612572, + "learning_rate": 3.2077086097029366e-06, + "loss": 0.8296, + "num_input_tokens_seen": 113610565, + "step": 3491 + }, + { + "epoch": 0.31492086395815483, + "flos": 21148495726560.0, + "grad_norm": 1.8453711155586519, + "learning_rate": 3.2072428869833895e-06, + "loss": 0.7653, + "num_input_tokens_seen": 113639060, + "step": 3492 + }, + { + "epoch": 0.3150110474816251, + "flos": 23772238697280.0, + "grad_norm": 1.5237566491373895, + "learning_rate": 3.206777061255223e-06, + "loss": 0.7783, + "num_input_tokens_seen": 113668850, + "step": 3493 + }, + { + "epoch": 0.31510123100509535, + "flos": 24317567910720.0, + "grad_norm": 3.189177576555926, + "learning_rate": 3.206311132558183e-06, + "loss": 0.8097, + "num_input_tokens_seen": 113696790, + "step": 3494 + }, + { + "epoch": 0.31519141452856564, + "flos": 17899720109760.0, + "grad_norm": 1.8540100145779346, + "learning_rate": 3.205845100932026e-06, + "loss": 0.7814, + "num_input_tokens_seen": 113722725, + "step": 3495 + }, + { + "epoch": 0.31528159805203587, + "flos": 24901410729120.0, + "grad_norm": 1.6509510420547664, + "learning_rate": 3.205378966416516e-06, + "loss": 0.7914, + "num_input_tokens_seen": 113752895, + "step": 3496 + }, + { + "epoch": 0.31537178157550616, + "flos": 22969877628000.0, + "grad_norm": 1.6957355446079156, + "learning_rate": 3.204912729051426e-06, + "loss": 0.7637, + "num_input_tokens_seen": 113780845, + "step": 3497 + }, + { + "epoch": 0.3154619650989764, + "flos": 19800768274080.0, + "grad_norm": 1.95439460688443, + "learning_rate": 3.2044463888765384e-06, + "loss": 0.741, + "num_input_tokens_seen": 113809370, + "step": 3498 + }, + { + "epoch": 0.3155521486224467, + "flos": 38124775657920.0, + "grad_norm": 1.8021776986818872, + "learning_rate": 3.2039799459316436e-06, + "loss": 0.6622, + "num_input_tokens_seen": 113839370, + "step": 3499 + }, + { + "epoch": 0.3156423321459169, + "flos": 25445327491680.0, + "grad_norm": 5.666902574602731, + "learning_rate": 3.2035134002565407e-06, + "loss": 0.8018, + "num_input_tokens_seen": 113870200, + "step": 3500 + }, + { + "epoch": 0.3157325156693872, + "flos": 14954927489280.0, + "grad_norm": 1.8597235473154539, + "learning_rate": 3.203046751891039e-06, + "loss": 0.7765, + "num_input_tokens_seen": 113897030, + "step": 3501 + }, + { + "epoch": 0.3158226991928575, + "flos": 22604813248800.0, + "grad_norm": 1.7299592449922359, + "learning_rate": 3.2025800008749545e-06, + "loss": 0.7798, + "num_input_tokens_seen": 113921555, + "step": 3502 + }, + { + "epoch": 0.3159128827163277, + "flos": 23549185735680.0, + "grad_norm": 1.58385477347228, + "learning_rate": 3.202113147248114e-06, + "loss": 0.7408, + "num_input_tokens_seen": 113950745, + "step": 3503 + }, + { + "epoch": 0.316003066239798, + "flos": 24132508177440.0, + "grad_norm": 1.7445525124087369, + "learning_rate": 3.20164619105035e-06, + "loss": 0.7411, + "num_input_tokens_seen": 113979095, + "step": 3504 + }, + { + "epoch": 0.31609324976326825, + "flos": 23516842310880.0, + "grad_norm": 1.4769226119494623, + "learning_rate": 3.201179132321508e-06, + "loss": 0.621, + "num_input_tokens_seen": 114009790, + "step": 3505 + }, + { + "epoch": 0.31618343328673854, + "flos": 16521582060000.0, + "grad_norm": 2.021824138302321, + "learning_rate": 3.200711971101439e-06, + "loss": 0.7109, + "num_input_tokens_seen": 114036615, + "step": 3506 + }, + { + "epoch": 0.31627361681020877, + "flos": 18227051448960.0, + "grad_norm": 2.0168235822849123, + "learning_rate": 3.2002447074300047e-06, + "loss": 0.7674, + "num_input_tokens_seen": 114064050, + "step": 3507 + }, + { + "epoch": 0.31636380033367906, + "flos": 22023683822880.0, + "grad_norm": 2.2531357213372365, + "learning_rate": 3.1997773413470736e-06, + "loss": 0.7728, + "num_input_tokens_seen": 114093815, + "step": 3508 + }, + { + "epoch": 0.3164539838571493, + "flos": 23735211882720.0, + "grad_norm": 2.7015505047471735, + "learning_rate": 3.199309872892524e-06, + "loss": 0.7594, + "num_input_tokens_seen": 114123275, + "step": 3509 + }, + { + "epoch": 0.3165441673806196, + "flos": 24822041824320.0, + "grad_norm": 1.7930917469131078, + "learning_rate": 3.198842302106243e-06, + "loss": 0.8195, + "num_input_tokens_seen": 114150985, + "step": 3510 + }, + { + "epoch": 0.3166343509040898, + "flos": 19436038422720.0, + "grad_norm": 1.8761126078216022, + "learning_rate": 3.1983746290281265e-06, + "loss": 0.752, + "num_input_tokens_seen": 114178680, + "step": 3511 + }, + { + "epoch": 0.3167245344275601, + "flos": 33676342677120.0, + "grad_norm": 1.4064899663692723, + "learning_rate": 3.197906853698079e-06, + "loss": 0.7748, + "num_input_tokens_seen": 114207315, + "step": 3512 + }, + { + "epoch": 0.31681471795103033, + "flos": 39872512797600.0, + "grad_norm": 1.7722210850540918, + "learning_rate": 3.1974389761560137e-06, + "loss": 0.7166, + "num_input_tokens_seen": 114237840, + "step": 3513 + }, + { + "epoch": 0.3169049014745006, + "flos": 21731297791680.0, + "grad_norm": 1.5995112984089213, + "learning_rate": 3.1969709964418525e-06, + "loss": 0.7163, + "num_input_tokens_seen": 114264090, + "step": 3514 + }, + { + "epoch": 0.31699508499797086, + "flos": 24609024697920.0, + "grad_norm": 1.7492204701056122, + "learning_rate": 3.196502914595525e-06, + "loss": 0.7903, + "num_input_tokens_seen": 114294380, + "step": 3515 + }, + { + "epoch": 0.31708526852144114, + "flos": 17650939940640.0, + "grad_norm": 1.8016870169441601, + "learning_rate": 3.1960347306569723e-06, + "loss": 0.788, + "num_input_tokens_seen": 114320270, + "step": 3516 + }, + { + "epoch": 0.3171754520449114, + "flos": 15282742035360.0, + "grad_norm": 1.8956921353700558, + "learning_rate": 3.195566444666141e-06, + "loss": 0.7202, + "num_input_tokens_seen": 114346135, + "step": 3517 + }, + { + "epoch": 0.31726563556838167, + "flos": 23114044891680.0, + "grad_norm": 1.6850659512027872, + "learning_rate": 3.1950980566629886e-06, + "loss": 0.8752, + "num_input_tokens_seen": 114374320, + "step": 3518 + }, + { + "epoch": 0.3173558190918519, + "flos": 24098789471520.0, + "grad_norm": 1.5549645105218781, + "learning_rate": 3.1946295666874797e-06, + "loss": 0.7823, + "num_input_tokens_seen": 114405100, + "step": 3519 + }, + { + "epoch": 0.3174460026153222, + "flos": 22385254244640.0, + "grad_norm": 1.8467796133230263, + "learning_rate": 3.19416097477959e-06, + "loss": 0.7875, + "num_input_tokens_seen": 114432410, + "step": 3520 + }, + { + "epoch": 0.3175361861387924, + "flos": 24278236571040.0, + "grad_norm": 1.4159859068810388, + "learning_rate": 3.1936922809793005e-06, + "loss": 0.8251, + "num_input_tokens_seen": 114462085, + "step": 3521 + }, + { + "epoch": 0.3176263696622627, + "flos": 25661801405760.0, + "grad_norm": 1.5119300714297452, + "learning_rate": 3.193223485326604e-06, + "loss": 0.8874, + "num_input_tokens_seen": 114491130, + "step": 3522 + }, + { + "epoch": 0.31771655318573294, + "flos": 17868343098720.0, + "grad_norm": 2.0327629910186356, + "learning_rate": 3.1927545878615005e-06, + "loss": 0.8586, + "num_input_tokens_seen": 114516050, + "step": 3523 + }, + { + "epoch": 0.31780673670920323, + "flos": 21581889591840.0, + "grad_norm": 2.179291938793926, + "learning_rate": 3.192285588623999e-06, + "loss": 0.7196, + "num_input_tokens_seen": 114542430, + "step": 3524 + }, + { + "epoch": 0.31789692023267346, + "flos": 31348107997440.0, + "grad_norm": 1.440563172443138, + "learning_rate": 3.191816487654117e-06, + "loss": 0.6989, + "num_input_tokens_seen": 114574125, + "step": 3525 + }, + { + "epoch": 0.31798710375614375, + "flos": 15537841063680.0, + "grad_norm": 2.5778102798258127, + "learning_rate": 3.19134728499188e-06, + "loss": 0.732, + "num_input_tokens_seen": 114601445, + "step": 3526 + }, + { + "epoch": 0.31807728727961404, + "flos": 25921063447200.0, + "grad_norm": 1.7356110713126478, + "learning_rate": 3.1908779806773235e-06, + "loss": 0.6615, + "num_input_tokens_seen": 114628645, + "step": 3527 + }, + { + "epoch": 0.3181674708030843, + "flos": 28360267061280.0, + "grad_norm": 1.6977739176587021, + "learning_rate": 3.190408574750492e-06, + "loss": 0.8247, + "num_input_tokens_seen": 114657565, + "step": 3528 + }, + { + "epoch": 0.31825765432655456, + "flos": 28761168822720.0, + "grad_norm": 1.5495255236276082, + "learning_rate": 3.1899390672514367e-06, + "loss": 0.8388, + "num_input_tokens_seen": 114688160, + "step": 3529 + }, + { + "epoch": 0.3183478378500248, + "flos": 43003443074400.0, + "grad_norm": 2.260370725759625, + "learning_rate": 3.189469458220219e-06, + "loss": 0.6927, + "num_input_tokens_seen": 114718190, + "step": 3530 + }, + { + "epoch": 0.3184380213734951, + "flos": 24279574682400.0, + "grad_norm": 1.410701403291734, + "learning_rate": 3.1889997476969086e-06, + "loss": 0.7696, + "num_input_tokens_seen": 114749055, + "step": 3531 + }, + { + "epoch": 0.3185282048969653, + "flos": 23730677172000.0, + "grad_norm": 1.5632306488708612, + "learning_rate": 3.188529935721583e-06, + "loss": 0.7588, + "num_input_tokens_seen": 114777510, + "step": 3532 + }, + { + "epoch": 0.3186183884204356, + "flos": 25666113097920.0, + "grad_norm": 1.6230625193148336, + "learning_rate": 3.18806002233433e-06, + "loss": 0.8308, + "num_input_tokens_seen": 114807230, + "step": 3533 + }, + { + "epoch": 0.31870857194390584, + "flos": 17141819807040.0, + "grad_norm": 1.6854397241298564, + "learning_rate": 3.187590007575245e-06, + "loss": 0.8332, + "num_input_tokens_seen": 114834145, + "step": 3534 + }, + { + "epoch": 0.3187987554673761, + "flos": 34185834508320.0, + "grad_norm": 7.9373712003575605, + "learning_rate": 3.1871198914844327e-06, + "loss": 0.7072, + "num_input_tokens_seen": 114862765, + "step": 3535 + }, + { + "epoch": 0.31888893899084636, + "flos": 20746367363040.0, + "grad_norm": 1.946482540147358, + "learning_rate": 3.1866496741020057e-06, + "loss": 0.7044, + "num_input_tokens_seen": 114891220, + "step": 3536 + }, + { + "epoch": 0.31897912251431665, + "flos": 35613599921280.0, + "grad_norm": 1.3175812862300638, + "learning_rate": 3.186179355468085e-06, + "loss": 0.7382, + "num_input_tokens_seen": 114927910, + "step": 3537 + }, + { + "epoch": 0.3190693060377869, + "flos": 24420768365280.0, + "grad_norm": 3.381104494020749, + "learning_rate": 3.1857089356228015e-06, + "loss": 0.7594, + "num_input_tokens_seen": 114954465, + "step": 3538 + }, + { + "epoch": 0.31915948956125717, + "flos": 19545632076000.0, + "grad_norm": 1.5263147033922932, + "learning_rate": 3.1852384146062933e-06, + "loss": 0.7155, + "num_input_tokens_seen": 114983110, + "step": 3539 + }, + { + "epoch": 0.3192496730847274, + "flos": 70885812600480.0, + "grad_norm": 0.6279908383317712, + "learning_rate": 3.184767792458708e-06, + "loss": 0.5579, + "num_input_tokens_seen": 115083385, + "step": 3540 + }, + { + "epoch": 0.3193398566081977, + "flos": 22236923967840.0, + "grad_norm": 1.8129960400790077, + "learning_rate": 3.1842970692202023e-06, + "loss": 0.897, + "num_input_tokens_seen": 115111715, + "step": 3541 + }, + { + "epoch": 0.3194300401316679, + "flos": 23844247989600.0, + "grad_norm": 1.7127927710878144, + "learning_rate": 3.1838262449309403e-06, + "loss": 0.7491, + "num_input_tokens_seen": 115139330, + "step": 3542 + }, + { + "epoch": 0.3195202236551382, + "flos": 23115680361120.0, + "grad_norm": 2.0099410711731385, + "learning_rate": 3.1833553196310956e-06, + "loss": 0.6877, + "num_input_tokens_seen": 115167220, + "step": 3543 + }, + { + "epoch": 0.31961040717860845, + "flos": 23770380209280.0, + "grad_norm": 1.6160179737526743, + "learning_rate": 3.18288429336085e-06, + "loss": 0.6286, + "num_input_tokens_seen": 115195235, + "step": 3544 + }, + { + "epoch": 0.31970059070207874, + "flos": 23145942279360.0, + "grad_norm": 1.9511889207799908, + "learning_rate": 3.182413166160394e-06, + "loss": 0.874, + "num_input_tokens_seen": 115222175, + "step": 3545 + }, + { + "epoch": 0.31979077422554897, + "flos": 21002581484160.0, + "grad_norm": 1.8650742706769057, + "learning_rate": 3.1819419380699275e-06, + "loss": 0.8059, + "num_input_tokens_seen": 115250055, + "step": 3546 + }, + { + "epoch": 0.31988095774901926, + "flos": 26283525943200.0, + "grad_norm": 2.010492814224558, + "learning_rate": 3.181470609129658e-06, + "loss": 0.7768, + "num_input_tokens_seen": 115278925, + "step": 3547 + }, + { + "epoch": 0.3199711412724895, + "flos": 18051395664960.0, + "grad_norm": 2.6014900082864263, + "learning_rate": 3.1809991793798e-06, + "loss": 0.8143, + "num_input_tokens_seen": 115303655, + "step": 3548 + }, + { + "epoch": 0.3200613247959598, + "flos": 24859254487680.0, + "grad_norm": 1.937685127027742, + "learning_rate": 3.1805276488605806e-06, + "loss": 0.7252, + "num_input_tokens_seen": 115327930, + "step": 3549 + }, + { + "epoch": 0.32015150831943007, + "flos": 18885877140480.0, + "grad_norm": 1.5632615331715132, + "learning_rate": 3.1800560176122336e-06, + "loss": 0.7607, + "num_input_tokens_seen": 115354715, + "step": 3550 + }, + { + "epoch": 0.3202416918429003, + "flos": 21440324211360.0, + "grad_norm": 2.9828242721844864, + "learning_rate": 3.179584285675e-06, + "loss": 0.7728, + "num_input_tokens_seen": 115382660, + "step": 3551 + }, + { + "epoch": 0.3203318753663706, + "flos": 25551166999200.0, + "grad_norm": 2.5845074805330226, + "learning_rate": 3.1791124530891315e-06, + "loss": 0.6868, + "num_input_tokens_seen": 115410670, + "step": 3552 + }, + { + "epoch": 0.3204220588898408, + "flos": 19836494147040.0, + "grad_norm": 1.7353684612904692, + "learning_rate": 3.178640519894886e-06, + "loss": 0.7697, + "num_input_tokens_seen": 115438665, + "step": 3553 + }, + { + "epoch": 0.3205122424133111, + "flos": 33754336300800.0, + "grad_norm": 1.7286450008962324, + "learning_rate": 3.1781684861325324e-06, + "loss": 0.7269, + "num_input_tokens_seen": 115469595, + "step": 3554 + }, + { + "epoch": 0.32060242593678134, + "flos": 37761681276000.0, + "grad_norm": 1.8884703914309917, + "learning_rate": 3.177696351842348e-06, + "loss": 0.7766, + "num_input_tokens_seen": 115500870, + "step": 3555 + }, + { + "epoch": 0.32069260946025163, + "flos": 70969827725280.0, + "grad_norm": 0.7024766263772655, + "learning_rate": 3.1772241170646167e-06, + "loss": 0.5601, + "num_input_tokens_seen": 115598935, + "step": 3556 + }, + { + "epoch": 0.32078279298372187, + "flos": 14772209450880.0, + "grad_norm": 15.912133228662924, + "learning_rate": 3.1767517818396334e-06, + "loss": 0.8731, + "num_input_tokens_seen": 115624265, + "step": 3557 + }, + { + "epoch": 0.32087297650719215, + "flos": 20814176472480.0, + "grad_norm": 1.5223641551381881, + "learning_rate": 3.1762793462076986e-06, + "loss": 0.925, + "num_input_tokens_seen": 115652080, + "step": 3558 + }, + { + "epoch": 0.3209631600306624, + "flos": 17359780511520.0, + "grad_norm": 3.5072536235695164, + "learning_rate": 3.1758068102091236e-06, + "loss": 0.6516, + "num_input_tokens_seen": 115679820, + "step": 3559 + }, + { + "epoch": 0.3210533435541327, + "flos": 27014658285120.0, + "grad_norm": 1.9212276950353007, + "learning_rate": 3.175334173884229e-06, + "loss": 0.7372, + "num_input_tokens_seen": 115708980, + "step": 3560 + }, + { + "epoch": 0.3211435270776029, + "flos": 24172211214720.0, + "grad_norm": 1.483954761648461, + "learning_rate": 3.174861437273342e-06, + "loss": 0.769, + "num_input_tokens_seen": 115737920, + "step": 3561 + }, + { + "epoch": 0.3212337106010732, + "flos": 30620766971040.0, + "grad_norm": 2.029683609533656, + "learning_rate": 3.174388600416799e-06, + "loss": 0.7283, + "num_input_tokens_seen": 115768735, + "step": 3562 + }, + { + "epoch": 0.32132389412454343, + "flos": 69258714266400.0, + "grad_norm": 0.6573855248939917, + "learning_rate": 3.1739156633549445e-06, + "loss": 0.5699, + "num_input_tokens_seen": 115858805, + "step": 3563 + }, + { + "epoch": 0.3214140776480137, + "flos": 21216267666240.0, + "grad_norm": 1.8276287611702264, + "learning_rate": 3.173442626128133e-06, + "loss": 0.8319, + "num_input_tokens_seen": 115886790, + "step": 3564 + }, + { + "epoch": 0.32150426117148395, + "flos": 14954369942880.0, + "grad_norm": 3.7057182149218146, + "learning_rate": 3.1729694887767265e-06, + "loss": 0.8781, + "num_input_tokens_seen": 115907785, + "step": 3565 + }, + { + "epoch": 0.32159444469495424, + "flos": 19283991169920.0, + "grad_norm": 1.7403750721304632, + "learning_rate": 3.172496251341096e-06, + "loss": 0.7697, + "num_input_tokens_seen": 115933310, + "step": 3566 + }, + { + "epoch": 0.3216846282184245, + "flos": 31532461505280.0, + "grad_norm": 2.009913907081444, + "learning_rate": 3.172022913861619e-06, + "loss": 0.6233, + "num_input_tokens_seen": 115965495, + "step": 3567 + }, + { + "epoch": 0.32177481174189476, + "flos": 23844508177920.0, + "grad_norm": 1.83918319474632, + "learning_rate": 3.171549476378686e-06, + "loss": 0.7219, + "num_input_tokens_seen": 115994960, + "step": 3568 + }, + { + "epoch": 0.321864995265365, + "flos": 21112435325760.0, + "grad_norm": 1.6955001656867723, + "learning_rate": 3.1710759389326906e-06, + "loss": 0.7346, + "num_input_tokens_seen": 116025195, + "step": 3569 + }, + { + "epoch": 0.3219551787888353, + "flos": 22642434779520.0, + "grad_norm": 3.630917290630417, + "learning_rate": 3.1706023015640396e-06, + "loss": 0.7612, + "num_input_tokens_seen": 116052020, + "step": 3570 + }, + { + "epoch": 0.3220453623123055, + "flos": 27778988956320.0, + "grad_norm": 1.977478357961023, + "learning_rate": 3.1701285643131453e-06, + "loss": 0.7882, + "num_input_tokens_seen": 116080155, + "step": 3571 + }, + { + "epoch": 0.3221355458357758, + "flos": 20164903409280.0, + "grad_norm": 2.118448255089316, + "learning_rate": 3.16965472722043e-06, + "loss": 0.7767, + "num_input_tokens_seen": 116107160, + "step": 3572 + }, + { + "epoch": 0.32222572935924604, + "flos": 19071308571360.0, + "grad_norm": 2.2200996699598115, + "learning_rate": 3.169180790326324e-06, + "loss": 0.6897, + "num_input_tokens_seen": 116133435, + "step": 3573 + }, + { + "epoch": 0.3223159128827163, + "flos": 23371225426560.0, + "grad_norm": 1.457929482626734, + "learning_rate": 3.168706753671266e-06, + "loss": 0.7691, + "num_input_tokens_seen": 116161325, + "step": 3574 + }, + { + "epoch": 0.3224060964061866, + "flos": 41659209579360.0, + "grad_norm": 1.6555340497044557, + "learning_rate": 3.168232617295704e-06, + "loss": 0.6909, + "num_input_tokens_seen": 116193920, + "step": 3575 + }, + { + "epoch": 0.32249627992965685, + "flos": 18161175167040.0, + "grad_norm": 1.9565577743270923, + "learning_rate": 3.167758381240093e-06, + "loss": 0.7633, + "num_input_tokens_seen": 116220530, + "step": 3576 + }, + { + "epoch": 0.32258646345312714, + "flos": 25739311822560.0, + "grad_norm": 1.9115246822750303, + "learning_rate": 3.1672840455448978e-06, + "loss": 0.8099, + "num_input_tokens_seen": 116251690, + "step": 3577 + }, + { + "epoch": 0.32267664697659737, + "flos": 22019000433120.0, + "grad_norm": 1.593234633350835, + "learning_rate": 3.166809610250592e-06, + "loss": 0.7097, + "num_input_tokens_seen": 116281385, + "step": 3578 + }, + { + "epoch": 0.32276683050006766, + "flos": 24315635083200.0, + "grad_norm": 1.7026941996991736, + "learning_rate": 3.166335075397656e-06, + "loss": 0.6938, + "num_input_tokens_seen": 116310935, + "step": 3579 + }, + { + "epoch": 0.3228570140235379, + "flos": 23256353667360.0, + "grad_norm": 1.6967139011295855, + "learning_rate": 3.1658604410265808e-06, + "loss": 0.7315, + "num_input_tokens_seen": 116338890, + "step": 3580 + }, + { + "epoch": 0.3229471975470082, + "flos": 72399754717920.0, + "grad_norm": 0.9130831305312804, + "learning_rate": 3.1653857071778644e-06, + "loss": 0.5996, + "num_input_tokens_seen": 116430970, + "step": 3581 + }, + { + "epoch": 0.3230373810704784, + "flos": 19946050630560.0, + "grad_norm": 2.3673169191945003, + "learning_rate": 3.1649108738920133e-06, + "loss": 0.7612, + "num_input_tokens_seen": 116459910, + "step": 3582 + }, + { + "epoch": 0.3231275645939487, + "flos": 33790285192320.0, + "grad_norm": 1.886081311448201, + "learning_rate": 3.1644359412095432e-06, + "loss": 0.7671, + "num_input_tokens_seen": 116490685, + "step": 3583 + }, + { + "epoch": 0.32321774811741893, + "flos": 67084942873920.0, + "grad_norm": 0.7499716276524269, + "learning_rate": 3.163960909170978e-06, + "loss": 0.6093, + "num_input_tokens_seen": 116580970, + "step": 3584 + }, + { + "epoch": 0.3233079316408892, + "flos": 23626101436320.0, + "grad_norm": 1.7788115424677042, + "learning_rate": 3.1634857778168496e-06, + "loss": 0.6649, + "num_input_tokens_seen": 116609810, + "step": 3585 + }, + { + "epoch": 0.32339811516435946, + "flos": 27159048567360.0, + "grad_norm": 1.7197906811451606, + "learning_rate": 3.1630105471877002e-06, + "loss": 0.8025, + "num_input_tokens_seen": 116639670, + "step": 3586 + }, + { + "epoch": 0.32348829868782975, + "flos": 25811804321760.0, + "grad_norm": 1.9999959468800483, + "learning_rate": 3.162535217324077e-06, + "loss": 0.7456, + "num_input_tokens_seen": 116668730, + "step": 3587 + }, + { + "epoch": 0.3235784822113, + "flos": 21359691534720.0, + "grad_norm": 2.505011780153394, + "learning_rate": 3.1620597882665393e-06, + "loss": 0.7961, + "num_input_tokens_seen": 116690065, + "step": 3588 + }, + { + "epoch": 0.32366866573477027, + "flos": 20710976017920.0, + "grad_norm": 1.8516076963678831, + "learning_rate": 3.1615842600556535e-06, + "loss": 0.8257, + "num_input_tokens_seen": 116718785, + "step": 3589 + }, + { + "epoch": 0.3237588492582405, + "flos": 26431484522400.0, + "grad_norm": 1.692059186049063, + "learning_rate": 3.1611086327319932e-06, + "loss": 0.766, + "num_input_tokens_seen": 116747180, + "step": 3590 + }, + { + "epoch": 0.3238490327817108, + "flos": 24387123998880.0, + "grad_norm": 1.6066560462083572, + "learning_rate": 3.160632906336142e-06, + "loss": 0.8355, + "num_input_tokens_seen": 116775025, + "step": 3591 + }, + { + "epoch": 0.323939216305181, + "flos": 24243588621120.0, + "grad_norm": 2.9183401665497724, + "learning_rate": 3.160157080908692e-06, + "loss": 0.7892, + "num_input_tokens_seen": 116804390, + "step": 3592 + }, + { + "epoch": 0.3240293998286513, + "flos": 22894894754880.0, + "grad_norm": 1.6769558007063223, + "learning_rate": 3.1596811564902426e-06, + "loss": 0.8204, + "num_input_tokens_seen": 116835605, + "step": 3593 + }, + { + "epoch": 0.32411958335212154, + "flos": 19582064174400.0, + "grad_norm": 2.6191139532746637, + "learning_rate": 3.1592051331214023e-06, + "loss": 0.7056, + "num_input_tokens_seen": 116862100, + "step": 3594 + }, + { + "epoch": 0.32420976687559183, + "flos": 25590498338880.0, + "grad_norm": 1.9592863023263452, + "learning_rate": 3.158729010842789e-06, + "loss": 0.7874, + "num_input_tokens_seen": 116890930, + "step": 3595 + }, + { + "epoch": 0.32429995039906206, + "flos": 22129932197760.0, + "grad_norm": 2.261561887371238, + "learning_rate": 3.1582527896950266e-06, + "loss": 0.8197, + "num_input_tokens_seen": 116919500, + "step": 3596 + }, + { + "epoch": 0.32439013392253235, + "flos": 24062766240480.0, + "grad_norm": 1.8924755151291017, + "learning_rate": 3.157776469718749e-06, + "loss": 0.7839, + "num_input_tokens_seen": 116945705, + "step": 3597 + }, + { + "epoch": 0.32448031744600264, + "flos": 20019360864480.0, + "grad_norm": 1.80453893537154, + "learning_rate": 3.1573000509546004e-06, + "loss": 0.7568, + "num_input_tokens_seen": 116973570, + "step": 3598 + }, + { + "epoch": 0.3245705009694729, + "flos": 24973717379520.0, + "grad_norm": 1.595788736276333, + "learning_rate": 3.1568235334432296e-06, + "loss": 0.8188, + "num_input_tokens_seen": 116999995, + "step": 3599 + }, + { + "epoch": 0.32466068449294316, + "flos": 23438625668640.0, + "grad_norm": 2.167656183478586, + "learning_rate": 3.1563469172252964e-06, + "loss": 0.835, + "num_input_tokens_seen": 117027010, + "step": 3600 + }, + { + "epoch": 0.3247508680164134, + "flos": 21803976139680.0, + "grad_norm": 2.0467194942196643, + "learning_rate": 3.155870202341468e-06, + "loss": 0.8415, + "num_input_tokens_seen": 117055210, + "step": 3601 + }, + { + "epoch": 0.3248410515398837, + "flos": 25479455064960.0, + "grad_norm": 1.8837256532144933, + "learning_rate": 3.155393388832421e-06, + "loss": 0.763, + "num_input_tokens_seen": 117083135, + "step": 3602 + }, + { + "epoch": 0.3249312350633539, + "flos": 20779231164480.0, + "grad_norm": 3.991533259403621, + "learning_rate": 3.1549164767388386e-06, + "loss": 0.7948, + "num_input_tokens_seen": 117108240, + "step": 3603 + }, + { + "epoch": 0.3250214185868242, + "flos": 22272798519840.0, + "grad_norm": 2.0257363240489945, + "learning_rate": 3.1544394661014145e-06, + "loss": 0.7665, + "num_input_tokens_seen": 117135570, + "step": 3604 + }, + { + "epoch": 0.32511160211029444, + "flos": 33497378784480.0, + "grad_norm": 1.5950038689832489, + "learning_rate": 3.15396235696085e-06, + "loss": 0.6837, + "num_input_tokens_seen": 117168860, + "step": 3605 + }, + { + "epoch": 0.32520178563376473, + "flos": 20959533168480.0, + "grad_norm": 1.8335424506248577, + "learning_rate": 3.153485149357854e-06, + "loss": 0.758, + "num_input_tokens_seen": 117194585, + "step": 3606 + }, + { + "epoch": 0.32529196915723496, + "flos": 23881126125120.0, + "grad_norm": 1.6553633133786025, + "learning_rate": 3.153007843333145e-06, + "loss": 0.7526, + "num_input_tokens_seen": 117224770, + "step": 3607 + }, + { + "epoch": 0.32538215268070525, + "flos": 21986545499040.0, + "grad_norm": 1.9659130303869101, + "learning_rate": 3.152530438927449e-06, + "loss": 0.7983, + "num_input_tokens_seen": 117249605, + "step": 3608 + }, + { + "epoch": 0.3254723362041755, + "flos": 20748969246240.0, + "grad_norm": 1.9236682496228226, + "learning_rate": 3.1520529361815008e-06, + "loss": 0.8125, + "num_input_tokens_seen": 117276350, + "step": 3609 + }, + { + "epoch": 0.32556251972764577, + "flos": 28612541187840.0, + "grad_norm": 1.9691092753700563, + "learning_rate": 3.151575335136044e-06, + "loss": 0.8084, + "num_input_tokens_seen": 117305410, + "step": 3610 + }, + { + "epoch": 0.325652703251116, + "flos": 21294484308480.0, + "grad_norm": 1.9515167328696017, + "learning_rate": 3.1510976358318298e-06, + "loss": 0.7531, + "num_input_tokens_seen": 117331745, + "step": 3611 + }, + { + "epoch": 0.3257428867745863, + "flos": 26573495940000.0, + "grad_norm": 1.8370256161925804, + "learning_rate": 3.1506198383096186e-06, + "loss": 0.7152, + "num_input_tokens_seen": 117362490, + "step": 3612 + }, + { + "epoch": 0.3258330702980565, + "flos": 20674692598560.0, + "grad_norm": 1.9927058125340547, + "learning_rate": 3.150141942610178e-06, + "loss": 0.7822, + "num_input_tokens_seen": 117391070, + "step": 3613 + }, + { + "epoch": 0.3259232538215268, + "flos": 63590509347840.0, + "grad_norm": 1.0784114512390213, + "learning_rate": 3.1496639487742853e-06, + "loss": 0.5858, + "num_input_tokens_seen": 117476625, + "step": 3614 + }, + { + "epoch": 0.32601343734499705, + "flos": 26358731834880.0, + "grad_norm": 1.8763863842195512, + "learning_rate": 3.1491858568427247e-06, + "loss": 0.8185, + "num_input_tokens_seen": 117505595, + "step": 3615 + }, + { + "epoch": 0.32610362086846734, + "flos": 15137199490560.0, + "grad_norm": 1.7465987531305682, + "learning_rate": 3.1487076668562903e-06, + "loss": 0.8211, + "num_input_tokens_seen": 117533125, + "step": 3616 + }, + { + "epoch": 0.32619380439193757, + "flos": 21182028583680.0, + "grad_norm": 2.0040266510186844, + "learning_rate": 3.1482293788557847e-06, + "loss": 0.8148, + "num_input_tokens_seen": 117560820, + "step": 3617 + }, + { + "epoch": 0.32628398791540786, + "flos": 19363248565440.0, + "grad_norm": 1.6836467831926059, + "learning_rate": 3.1477509928820165e-06, + "loss": 0.7341, + "num_input_tokens_seen": 117588360, + "step": 3618 + }, + { + "epoch": 0.3263741714388781, + "flos": 38122025095680.0, + "grad_norm": 1.7627580034514947, + "learning_rate": 3.147272508975805e-06, + "loss": 0.7687, + "num_input_tokens_seen": 117618785, + "step": 3619 + }, + { + "epoch": 0.3264643549623484, + "flos": 54679141446240.0, + "grad_norm": 0.6744503299827993, + "learning_rate": 3.1467939271779775e-06, + "loss": 0.6059, + "num_input_tokens_seen": 117707135, + "step": 3620 + }, + { + "epoch": 0.32655453848581867, + "flos": 42750908759520.0, + "grad_norm": 1.471106971523291, + "learning_rate": 3.146315247529368e-06, + "loss": 0.7318, + "num_input_tokens_seen": 117741085, + "step": 3621 + }, + { + "epoch": 0.3266447220092889, + "flos": 25957755733920.0, + "grad_norm": 2.089599138721797, + "learning_rate": 3.1458364700708212e-06, + "loss": 0.7274, + "num_input_tokens_seen": 117770230, + "step": 3622 + }, + { + "epoch": 0.3267349055327592, + "flos": 26213449478400.0, + "grad_norm": 1.501211074845374, + "learning_rate": 3.1453575948431892e-06, + "loss": 0.7916, + "num_input_tokens_seen": 117801395, + "step": 3623 + }, + { + "epoch": 0.3268250890562294, + "flos": 27156595363200.0, + "grad_norm": 1.8763884172204466, + "learning_rate": 3.144878621887331e-06, + "loss": 0.7766, + "num_input_tokens_seen": 117829925, + "step": 3624 + }, + { + "epoch": 0.3269152725796997, + "flos": 22168966179360.0, + "grad_norm": 2.736755119951065, + "learning_rate": 3.1443995512441167e-06, + "loss": 0.8116, + "num_input_tokens_seen": 117858275, + "step": 3625 + }, + { + "epoch": 0.32700545610316994, + "flos": 19725748231200.0, + "grad_norm": 1.7887412369977949, + "learning_rate": 3.1439203829544224e-06, + "loss": 0.8366, + "num_input_tokens_seen": 117888865, + "step": 3626 + }, + { + "epoch": 0.32709563962664023, + "flos": 22058852149440.0, + "grad_norm": 3.5162245853724747, + "learning_rate": 3.143441117059133e-06, + "loss": 0.8505, + "num_input_tokens_seen": 117919220, + "step": 3627 + }, + { + "epoch": 0.32718582315011047, + "flos": 35466198888480.0, + "grad_norm": 1.9755352502148142, + "learning_rate": 3.142961753599143e-06, + "loss": 0.7167, + "num_input_tokens_seen": 117948915, + "step": 3628 + }, + { + "epoch": 0.32727600667358075, + "flos": 22569087375840.0, + "grad_norm": 4.2972030237009164, + "learning_rate": 3.1424822926153543e-06, + "loss": 0.7365, + "num_input_tokens_seen": 117977185, + "step": 3629 + }, + { + "epoch": 0.327366190197051, + "flos": 33932036421600.0, + "grad_norm": 1.901342610267453, + "learning_rate": 3.142002734148676e-06, + "loss": 0.813, + "num_input_tokens_seen": 118004730, + "step": 3630 + }, + { + "epoch": 0.3274563737205213, + "flos": 26358806174400.0, + "grad_norm": 1.9225664213138134, + "learning_rate": 3.141523078240028e-06, + "loss": 0.798, + "num_input_tokens_seen": 118032810, + "step": 3631 + }, + { + "epoch": 0.3275465572439915, + "flos": 23624428797120.0, + "grad_norm": 1.962494494193548, + "learning_rate": 3.1410433249303366e-06, + "loss": 0.8272, + "num_input_tokens_seen": 118056970, + "step": 3632 + }, + { + "epoch": 0.3276367407674618, + "flos": 25368374621280.0, + "grad_norm": 1.8790978474783981, + "learning_rate": 3.1405634742605366e-06, + "loss": 0.8595, + "num_input_tokens_seen": 118084040, + "step": 3633 + }, + { + "epoch": 0.32772692429093203, + "flos": 16521284701920.0, + "grad_norm": 2.1912365199978447, + "learning_rate": 3.1400835262715727e-06, + "loss": 0.8081, + "num_input_tokens_seen": 118108870, + "step": 3634 + }, + { + "epoch": 0.3278171078144023, + "flos": 22488083001600.0, + "grad_norm": 1.7193506751771663, + "learning_rate": 3.139603481004396e-06, + "loss": 0.759, + "num_input_tokens_seen": 118134090, + "step": 3635 + }, + { + "epoch": 0.32790729133787255, + "flos": 24573075806400.0, + "grad_norm": 1.6508506720587321, + "learning_rate": 3.139123338499966e-06, + "loss": 0.7875, + "num_input_tokens_seen": 118165210, + "step": 3636 + }, + { + "epoch": 0.32799747486134284, + "flos": 20019249355200.0, + "grad_norm": 2.0562827296101616, + "learning_rate": 3.1386430987992524e-06, + "loss": 0.8111, + "num_input_tokens_seen": 118193210, + "step": 3637 + }, + { + "epoch": 0.3280876583848131, + "flos": 40274678330880.0, + "grad_norm": 1.8664037104861202, + "learning_rate": 3.1381627619432307e-06, + "loss": 0.7155, + "num_input_tokens_seen": 118224650, + "step": 3638 + }, + { + "epoch": 0.32817784190828336, + "flos": 25627785341760.0, + "grad_norm": 2.191401462923693, + "learning_rate": 3.1376823279728864e-06, + "loss": 0.7814, + "num_input_tokens_seen": 118249620, + "step": 3639 + }, + { + "epoch": 0.3282680254317536, + "flos": 30396301558560.0, + "grad_norm": 1.9137701940187402, + "learning_rate": 3.1372017969292125e-06, + "loss": 0.8538, + "num_input_tokens_seen": 118278655, + "step": 3640 + }, + { + "epoch": 0.3283582089552239, + "flos": 22970398004640.0, + "grad_norm": 1.719377021857695, + "learning_rate": 3.136721168853211e-06, + "loss": 0.7547, + "num_input_tokens_seen": 118306195, + "step": 3641 + }, + { + "epoch": 0.3284483924786941, + "flos": 26722458102720.0, + "grad_norm": 2.0335789167381253, + "learning_rate": 3.1362404437858924e-06, + "loss": 0.7364, + "num_input_tokens_seen": 118333670, + "step": 3642 + }, + { + "epoch": 0.3285385760021644, + "flos": 25337741005440.0, + "grad_norm": 1.9466286615141182, + "learning_rate": 3.135759621768273e-06, + "loss": 0.8271, + "num_input_tokens_seen": 118362115, + "step": 3643 + }, + { + "epoch": 0.32862875952563464, + "flos": 16995236508960.0, + "grad_norm": 1.7721207221265383, + "learning_rate": 3.13527870284138e-06, + "loss": 0.8257, + "num_input_tokens_seen": 118389270, + "step": 3644 + }, + { + "epoch": 0.32871894304910493, + "flos": 20638148990880.0, + "grad_norm": 6.704985958258186, + "learning_rate": 3.134797687046249e-06, + "loss": 0.8275, + "num_input_tokens_seen": 118415880, + "step": 3645 + }, + { + "epoch": 0.3288091265725752, + "flos": 16849545285120.0, + "grad_norm": 3.531341281487516, + "learning_rate": 3.1343165744239218e-06, + "loss": 0.7074, + "num_input_tokens_seen": 118443785, + "step": 3646 + }, + { + "epoch": 0.32889931009604545, + "flos": 25301792113920.0, + "grad_norm": 1.7783299654803908, + "learning_rate": 3.13383536501545e-06, + "loss": 0.7951, + "num_input_tokens_seen": 118471805, + "step": 3647 + }, + { + "epoch": 0.32898949361951574, + "flos": 18996251358720.0, + "grad_norm": 1.624530944618724, + "learning_rate": 3.133354058861893e-06, + "loss": 0.7937, + "num_input_tokens_seen": 118499990, + "step": 3648 + }, + { + "epoch": 0.32907967714298597, + "flos": 23473087769760.0, + "grad_norm": 1.6704023060506856, + "learning_rate": 3.132872656004318e-06, + "loss": 0.7433, + "num_input_tokens_seen": 118527995, + "step": 3649 + }, + { + "epoch": 0.32916986066645626, + "flos": 31058249509920.0, + "grad_norm": 1.6358194361423675, + "learning_rate": 3.132391156483802e-06, + "loss": 0.6692, + "num_input_tokens_seen": 118557765, + "step": 3650 + }, + { + "epoch": 0.3292600441899265, + "flos": 24500285949120.0, + "grad_norm": 1.8482588255196972, + "learning_rate": 3.131909560341428e-06, + "loss": 0.7308, + "num_input_tokens_seen": 118585295, + "step": 3651 + }, + { + "epoch": 0.3293502277133968, + "flos": 23444461320960.0, + "grad_norm": 1.7496785141062972, + "learning_rate": 3.1314278676182893e-06, + "loss": 0.724, + "num_input_tokens_seen": 118614380, + "step": 3652 + }, + { + "epoch": 0.329440411236867, + "flos": 23844247989600.0, + "grad_norm": 1.517944841496008, + "learning_rate": 3.130946078355486e-06, + "loss": 0.8416, + "num_input_tokens_seen": 118642825, + "step": 3653 + }, + { + "epoch": 0.3295305947603373, + "flos": 22715410485600.0, + "grad_norm": 1.5587714913728998, + "learning_rate": 3.130464192594128e-06, + "loss": 0.8065, + "num_input_tokens_seen": 118672050, + "step": 3654 + }, + { + "epoch": 0.32962077828380754, + "flos": 25374581971200.0, + "grad_norm": 1.596680284785821, + "learning_rate": 3.1299822103753315e-06, + "loss": 0.7003, + "num_input_tokens_seen": 118702410, + "step": 3655 + }, + { + "epoch": 0.3297109618072778, + "flos": 18197421416640.0, + "grad_norm": 1.8188552590025666, + "learning_rate": 3.1295001317402217e-06, + "loss": 0.7757, + "num_input_tokens_seen": 118729730, + "step": 3656 + }, + { + "epoch": 0.32980114533074806, + "flos": 17212119290400.0, + "grad_norm": 2.2421782490077558, + "learning_rate": 3.1290179567299335e-06, + "loss": 0.72, + "num_input_tokens_seen": 118753075, + "step": 3657 + }, + { + "epoch": 0.32989132885421835, + "flos": 26427767546400.0, + "grad_norm": 1.6441034408428457, + "learning_rate": 3.128535685385607e-06, + "loss": 0.8032, + "num_input_tokens_seen": 118783500, + "step": 3658 + }, + { + "epoch": 0.3299815123776886, + "flos": 19144470126240.0, + "grad_norm": 2.05428504628535, + "learning_rate": 3.1280533177483935e-06, + "loss": 0.8538, + "num_input_tokens_seen": 118812100, + "step": 3659 + }, + { + "epoch": 0.33007169590115887, + "flos": 27232247292000.0, + "grad_norm": 1.5237817619837484, + "learning_rate": 3.127570853859451e-06, + "loss": 0.786, + "num_input_tokens_seen": 118841965, + "step": 3660 + }, + { + "epoch": 0.3301618794246291, + "flos": 25297666270560.0, + "grad_norm": 1.5775716585789883, + "learning_rate": 3.1270882937599456e-06, + "loss": 0.6599, + "num_input_tokens_seen": 118871560, + "step": 3661 + }, + { + "epoch": 0.3302520629480994, + "flos": 20054752209600.0, + "grad_norm": 1.97976756427005, + "learning_rate": 3.1266056374910532e-06, + "loss": 0.7951, + "num_input_tokens_seen": 118897770, + "step": 3662 + }, + { + "epoch": 0.3303422464715696, + "flos": 27771778022880.0, + "grad_norm": 1.958488728735996, + "learning_rate": 3.126122885093955e-06, + "loss": 0.67, + "num_input_tokens_seen": 118926875, + "step": 3663 + }, + { + "epoch": 0.3304324299950399, + "flos": 23516619292320.0, + "grad_norm": 2.027805048034868, + "learning_rate": 3.1256400366098427e-06, + "loss": 0.7657, + "num_input_tokens_seen": 118955005, + "step": 3664 + }, + { + "epoch": 0.33052261351851014, + "flos": 22818833958720.0, + "grad_norm": 2.094713018057901, + "learning_rate": 3.125157092079916e-06, + "loss": 0.7312, + "num_input_tokens_seen": 118981925, + "step": 3665 + }, + { + "epoch": 0.33061279704198043, + "flos": 22313876838240.0, + "grad_norm": 2.351636612950414, + "learning_rate": 3.1246740515453824e-06, + "loss": 0.74, + "num_input_tokens_seen": 119009470, + "step": 3666 + }, + { + "epoch": 0.33070298056545067, + "flos": 33532101073920.0, + "grad_norm": 1.5707283038872664, + "learning_rate": 3.124190915047457e-06, + "loss": 0.7116, + "num_input_tokens_seen": 119042645, + "step": 3667 + }, + { + "epoch": 0.33079316408892095, + "flos": 13716198973920.0, + "grad_norm": 2.0983291200328376, + "learning_rate": 3.123707682627364e-06, + "loss": 0.7434, + "num_input_tokens_seen": 119068845, + "step": 3668 + }, + { + "epoch": 0.33088334761239124, + "flos": 31127917107360.0, + "grad_norm": 1.6738649451035204, + "learning_rate": 3.1232243543263356e-06, + "loss": 0.7563, + "num_input_tokens_seen": 119100440, + "step": 3669 + }, + { + "epoch": 0.3309735311358615, + "flos": 28616778540480.0, + "grad_norm": 3.787662963382847, + "learning_rate": 3.1227409301856122e-06, + "loss": 0.6747, + "num_input_tokens_seen": 119129435, + "step": 3670 + }, + { + "epoch": 0.33106371465933176, + "flos": 24570659772000.0, + "grad_norm": 1.399020068334868, + "learning_rate": 3.1222574102464413e-06, + "loss": 0.7544, + "num_input_tokens_seen": 119159715, + "step": 3671 + }, + { + "epoch": 0.331153898182802, + "flos": 26940902014080.0, + "grad_norm": 1.8273487236147712, + "learning_rate": 3.12177379455008e-06, + "loss": 0.8005, + "num_input_tokens_seen": 119187395, + "step": 3672 + }, + { + "epoch": 0.3312440817062723, + "flos": 22532878296000.0, + "grad_norm": 2.1623050976720886, + "learning_rate": 3.121290083137794e-06, + "loss": 0.7628, + "num_input_tokens_seen": 119217485, + "step": 3673 + }, + { + "epoch": 0.3313342652297425, + "flos": 23552939881440.0, + "grad_norm": 1.73076150109937, + "learning_rate": 3.1208062760508547e-06, + "loss": 0.7403, + "num_input_tokens_seen": 119246500, + "step": 3674 + }, + { + "epoch": 0.3314244487532128, + "flos": 28726520872800.0, + "grad_norm": 1.4849667624404859, + "learning_rate": 3.1203223733305438e-06, + "loss": 0.7089, + "num_input_tokens_seen": 119277235, + "step": 3675 + }, + { + "epoch": 0.33151463227668304, + "flos": 64250004095040.0, + "grad_norm": 0.6583091129805092, + "learning_rate": 3.1198383750181512e-06, + "loss": 0.5635, + "num_input_tokens_seen": 119374265, + "step": 3676 + }, + { + "epoch": 0.33160481580015333, + "flos": 24245558618400.0, + "grad_norm": 1.8939921762451868, + "learning_rate": 3.1193542811549734e-06, + "loss": 0.6958, + "num_input_tokens_seen": 119404695, + "step": 3677 + }, + { + "epoch": 0.33169499932362356, + "flos": 57579887901120.0, + "grad_norm": 1.700099252159592, + "learning_rate": 3.1188700917823166e-06, + "loss": 0.6912, + "num_input_tokens_seen": 119442925, + "step": 3678 + }, + { + "epoch": 0.33178518284709385, + "flos": 66575005005600.0, + "grad_norm": 0.7054364041742132, + "learning_rate": 3.1183858069414937e-06, + "loss": 0.6263, + "num_input_tokens_seen": 119532115, + "step": 3679 + }, + { + "epoch": 0.3318753663705641, + "flos": 23771643981120.0, + "grad_norm": 2.7947914418322206, + "learning_rate": 3.117901426673827e-06, + "loss": 0.8198, + "num_input_tokens_seen": 119560220, + "step": 3680 + }, + { + "epoch": 0.3319655498940344, + "flos": 24500397458400.0, + "grad_norm": 8.764220099648236, + "learning_rate": 3.1174169510206466e-06, + "loss": 0.7272, + "num_input_tokens_seen": 119588755, + "step": 3681 + }, + { + "epoch": 0.3320557334175046, + "flos": 22386964053600.0, + "grad_norm": 1.8434813998851227, + "learning_rate": 3.1169323800232908e-06, + "loss": 0.772, + "num_input_tokens_seen": 119619005, + "step": 3682 + }, + { + "epoch": 0.3321459169409749, + "flos": 23297729343840.0, + "grad_norm": 2.21788760336104, + "learning_rate": 3.1164477137231054e-06, + "loss": 0.7972, + "num_input_tokens_seen": 119646640, + "step": 3683 + }, + { + "epoch": 0.3322361004644451, + "flos": 16448420505120.0, + "grad_norm": 2.253984208634714, + "learning_rate": 3.115962952161445e-06, + "loss": 0.7747, + "num_input_tokens_seen": 119670775, + "step": 3684 + }, + { + "epoch": 0.3323262839879154, + "flos": 24245744467200.0, + "grad_norm": 2.1766838790228586, + "learning_rate": 3.1154780953796727e-06, + "loss": 0.8068, + "num_input_tokens_seen": 119697590, + "step": 3685 + }, + { + "epoch": 0.33241646751138565, + "flos": 17360300888160.0, + "grad_norm": 2.0064411155488204, + "learning_rate": 3.114993143419158e-06, + "loss": 0.6721, + "num_input_tokens_seen": 119724480, + "step": 3686 + }, + { + "epoch": 0.33250665103485594, + "flos": 38417050179840.0, + "grad_norm": 1.5540597693095024, + "learning_rate": 3.1145080963212806e-06, + "loss": 0.6808, + "num_input_tokens_seen": 119759315, + "step": 3687 + }, + { + "epoch": 0.33259683455832617, + "flos": 23735174712960.0, + "grad_norm": 1.7235956031305903, + "learning_rate": 3.114022954127427e-06, + "loss": 0.7695, + "num_input_tokens_seen": 119786895, + "step": 3688 + }, + { + "epoch": 0.33268701808179646, + "flos": 23328362959680.0, + "grad_norm": 1.704243563881207, + "learning_rate": 3.1135377168789923e-06, + "loss": 0.9752, + "num_input_tokens_seen": 119814860, + "step": 3689 + }, + { + "epoch": 0.3327772016052667, + "flos": 18197309907360.0, + "grad_norm": 2.706504341137511, + "learning_rate": 3.1130523846173803e-06, + "loss": 0.8338, + "num_input_tokens_seen": 119842625, + "step": 3690 + }, + { + "epoch": 0.332867385128737, + "flos": 32004889352160.0, + "grad_norm": 2.1897671531321596, + "learning_rate": 3.1125669573840006e-06, + "loss": 0.7073, + "num_input_tokens_seen": 119874575, + "step": 3691 + }, + { + "epoch": 0.3329575686522072, + "flos": 16408531619040.0, + "grad_norm": 7.687562523564341, + "learning_rate": 3.112081435220274e-06, + "loss": 0.7143, + "num_input_tokens_seen": 119901155, + "step": 3692 + }, + { + "epoch": 0.3330477521756775, + "flos": 23188879085760.0, + "grad_norm": 1.672908071620171, + "learning_rate": 3.111595818167627e-06, + "loss": 0.7555, + "num_input_tokens_seen": 119929480, + "step": 3693 + }, + { + "epoch": 0.3331379356991478, + "flos": 21111468912000.0, + "grad_norm": 3.032528951558586, + "learning_rate": 3.1111101062674953e-06, + "loss": 0.8263, + "num_input_tokens_seen": 119956740, + "step": 3694 + }, + { + "epoch": 0.333228119222618, + "flos": 23043076352640.0, + "grad_norm": 1.570753348799127, + "learning_rate": 3.1106242995613233e-06, + "loss": 0.7573, + "num_input_tokens_seen": 119984730, + "step": 3695 + }, + { + "epoch": 0.3333183027460883, + "flos": 23953618624320.0, + "grad_norm": 1.8100769684164806, + "learning_rate": 3.1101383980905616e-06, + "loss": 0.766, + "num_input_tokens_seen": 120014380, + "step": 3696 + }, + { + "epoch": 0.33340848626955855, + "flos": 67487703123360.0, + "grad_norm": 0.6731323630692645, + "learning_rate": 3.109652401896671e-06, + "loss": 0.5513, + "num_input_tokens_seen": 120107950, + "step": 3697 + }, + { + "epoch": 0.33349866979302883, + "flos": 29742307935840.0, + "grad_norm": 1.6329893171588215, + "learning_rate": 3.109166311021119e-06, + "loss": 0.8135, + "num_input_tokens_seen": 120138735, + "step": 3698 + }, + { + "epoch": 0.33358885331649907, + "flos": 16193544495360.0, + "grad_norm": 1.9881170838338427, + "learning_rate": 3.1086801255053807e-06, + "loss": 0.8074, + "num_input_tokens_seen": 120162715, + "step": 3699 + }, + { + "epoch": 0.33367903683996936, + "flos": 25338558740160.0, + "grad_norm": 1.5463518741633928, + "learning_rate": 3.108193845390942e-06, + "loss": 0.8068, + "num_input_tokens_seen": 120192965, + "step": 3700 + }, + { + "epoch": 0.3337692203634396, + "flos": 28725480119520.0, + "grad_norm": 1.3119114509887055, + "learning_rate": 3.1077074707192933e-06, + "loss": 0.7456, + "num_input_tokens_seen": 120224480, + "step": 3701 + }, + { + "epoch": 0.3338594038869099, + "flos": 23551415921280.0, + "grad_norm": 1.479316444156731, + "learning_rate": 3.1072210015319353e-06, + "loss": 0.7881, + "num_input_tokens_seen": 120254050, + "step": 3702 + }, + { + "epoch": 0.3339495874103801, + "flos": 22787642796480.0, + "grad_norm": 2.226110472721755, + "learning_rate": 3.106734437870376e-06, + "loss": 0.7221, + "num_input_tokens_seen": 120284205, + "step": 3703 + }, + { + "epoch": 0.3340397709338504, + "flos": 24573596183040.0, + "grad_norm": 1.8459422768963223, + "learning_rate": 3.1062477797761327e-06, + "loss": 0.7137, + "num_input_tokens_seen": 120314615, + "step": 3704 + }, + { + "epoch": 0.33412995445732063, + "flos": 21112063628160.0, + "grad_norm": 1.6699083829635881, + "learning_rate": 3.105761027290729e-06, + "loss": 0.8115, + "num_input_tokens_seen": 120340610, + "step": 3705 + }, + { + "epoch": 0.3342201379807909, + "flos": 28725554459040.0, + "grad_norm": 1.553382671020162, + "learning_rate": 3.105274180455697e-06, + "loss": 0.7345, + "num_input_tokens_seen": 120371540, + "step": 3706 + }, + { + "epoch": 0.33431032150426115, + "flos": 22751991263040.0, + "grad_norm": 1.6742289006597035, + "learning_rate": 3.1047872393125775e-06, + "loss": 0.7881, + "num_input_tokens_seen": 120399225, + "step": 3707 + }, + { + "epoch": 0.33440050502773144, + "flos": 61435105550400.0, + "grad_norm": 0.6310927960132888, + "learning_rate": 3.1043002039029186e-06, + "loss": 0.5791, + "num_input_tokens_seen": 120492180, + "step": 3708 + }, + { + "epoch": 0.3344906885512017, + "flos": 23113896212640.0, + "grad_norm": 1.8220045877895108, + "learning_rate": 3.1038130742682782e-06, + "loss": 0.7596, + "num_input_tokens_seen": 120522870, + "step": 3709 + }, + { + "epoch": 0.33458087207467196, + "flos": 28762321085280.0, + "grad_norm": 1.7201724321960943, + "learning_rate": 3.103325850450219e-06, + "loss": 0.7436, + "num_input_tokens_seen": 120554725, + "step": 3710 + }, + { + "epoch": 0.3346710555981422, + "flos": 25443989380320.0, + "grad_norm": 1.8209287067055027, + "learning_rate": 3.1028385324903154e-06, + "loss": 0.8057, + "num_input_tokens_seen": 120582855, + "step": 3711 + }, + { + "epoch": 0.3347612391216125, + "flos": 25957681394400.0, + "grad_norm": 1.6636733912048811, + "learning_rate": 3.1023511204301465e-06, + "loss": 0.7734, + "num_input_tokens_seen": 120612560, + "step": 3712 + }, + { + "epoch": 0.3348514226450827, + "flos": 29996849417760.0, + "grad_norm": 1.5592493193898467, + "learning_rate": 3.1018636143113022e-06, + "loss": 0.7278, + "num_input_tokens_seen": 120641515, + "step": 3713 + }, + { + "epoch": 0.334941606168553, + "flos": 20420039607360.0, + "grad_norm": 1.5209260241685136, + "learning_rate": 3.1013760141753787e-06, + "loss": 0.6715, + "num_input_tokens_seen": 120669460, + "step": 3714 + }, + { + "epoch": 0.33503178969202324, + "flos": 23771420962560.0, + "grad_norm": 1.7570500055230556, + "learning_rate": 3.100888320063981e-06, + "loss": 0.7365, + "num_input_tokens_seen": 120696735, + "step": 3715 + }, + { + "epoch": 0.33512197321549353, + "flos": 22751842584000.0, + "grad_norm": 1.4635681294755976, + "learning_rate": 3.100400532018721e-06, + "loss": 0.7714, + "num_input_tokens_seen": 120727235, + "step": 3716 + }, + { + "epoch": 0.3352121567389638, + "flos": 16477344312000.0, + "grad_norm": 2.9618264828998893, + "learning_rate": 3.0999126500812204e-06, + "loss": 0.8353, + "num_input_tokens_seen": 120752110, + "step": 3717 + }, + { + "epoch": 0.33530234026243405, + "flos": 58904855993280.0, + "grad_norm": 0.7320224741080937, + "learning_rate": 3.0994246742931076e-06, + "loss": 0.6329, + "num_input_tokens_seen": 120842135, + "step": 3718 + }, + { + "epoch": 0.33539252378590434, + "flos": 33644928496320.0, + "grad_norm": 1.6565171512138428, + "learning_rate": 3.098936604696019e-06, + "loss": 0.7121, + "num_input_tokens_seen": 120875465, + "step": 3719 + }, + { + "epoch": 0.3354827073093746, + "flos": 26540000252640.0, + "grad_norm": 2.9116739501454427, + "learning_rate": 3.0984484413316e-06, + "loss": 0.7986, + "num_input_tokens_seen": 120902600, + "step": 3720 + }, + { + "epoch": 0.33557289083284486, + "flos": 17906299157280.0, + "grad_norm": 4.814722117865747, + "learning_rate": 3.0979601842415033e-06, + "loss": 0.7426, + "num_input_tokens_seen": 120929670, + "step": 3721 + }, + { + "epoch": 0.3356630743563151, + "flos": 28943477993760.0, + "grad_norm": 1.60976353371884, + "learning_rate": 3.0974718334673896e-06, + "loss": 0.6871, + "num_input_tokens_seen": 120961495, + "step": 3722 + }, + { + "epoch": 0.3357532578797854, + "flos": 20637591444480.0, + "grad_norm": 1.6113091854947534, + "learning_rate": 3.0969833890509282e-06, + "loss": 0.7585, + "num_input_tokens_seen": 120990470, + "step": 3723 + }, + { + "epoch": 0.3358434414032556, + "flos": 15755838937920.0, + "grad_norm": 2.0074599138297993, + "learning_rate": 3.096494851033795e-06, + "loss": 0.7569, + "num_input_tokens_seen": 121016780, + "step": 3724 + }, + { + "epoch": 0.3359336249267259, + "flos": 25338744588960.0, + "grad_norm": 1.8632147115346898, + "learning_rate": 3.0960062194576747e-06, + "loss": 0.6914, + "num_input_tokens_seen": 121045315, + "step": 3725 + }, + { + "epoch": 0.33602380845019614, + "flos": 16047778932000.0, + "grad_norm": 1.769259467737411, + "learning_rate": 3.0955174943642606e-06, + "loss": 0.7204, + "num_input_tokens_seen": 121073150, + "step": 3726 + }, + { + "epoch": 0.3361139919736664, + "flos": 32801674957440.0, + "grad_norm": 2.014255738735842, + "learning_rate": 3.0950286757952534e-06, + "loss": 0.6884, + "num_input_tokens_seen": 121102155, + "step": 3727 + }, + { + "epoch": 0.33620417549713666, + "flos": 20929605778080.0, + "grad_norm": 2.113265229621, + "learning_rate": 3.0945397637923617e-06, + "loss": 0.8279, + "num_input_tokens_seen": 121131430, + "step": 3728 + }, + { + "epoch": 0.33629435902060695, + "flos": 22859875107360.0, + "grad_norm": 9.435098708889292, + "learning_rate": 3.0940507583973025e-06, + "loss": 0.7726, + "num_input_tokens_seen": 121158680, + "step": 3729 + }, + { + "epoch": 0.3363845425440772, + "flos": 27633706599840.0, + "grad_norm": 1.8374598258680588, + "learning_rate": 3.093561659651799e-06, + "loss": 0.7371, + "num_input_tokens_seen": 121189625, + "step": 3730 + }, + { + "epoch": 0.33647472606754747, + "flos": 25811692812480.0, + "grad_norm": 1.7328652348934253, + "learning_rate": 3.093072467597586e-06, + "loss": 0.6995, + "num_input_tokens_seen": 121218500, + "step": 3731 + }, + { + "epoch": 0.3365649095910177, + "flos": 71423893710720.0, + "grad_norm": 0.6384171144028872, + "learning_rate": 3.092583182276402e-06, + "loss": 0.5839, + "num_input_tokens_seen": 121315835, + "step": 3732 + }, + { + "epoch": 0.336655093114488, + "flos": 54790370568960.0, + "grad_norm": 0.6381109965447718, + "learning_rate": 3.092093803729997e-06, + "loss": 0.5894, + "num_input_tokens_seen": 121412965, + "step": 3733 + }, + { + "epoch": 0.3367452766379582, + "flos": 28724030498880.0, + "grad_norm": 1.6084089620462434, + "learning_rate": 3.0916043320001264e-06, + "loss": 0.778, + "num_input_tokens_seen": 121442675, + "step": 3734 + }, + { + "epoch": 0.3368354601614285, + "flos": 22965194238240.0, + "grad_norm": 3.5375259128875065, + "learning_rate": 3.0911147671285557e-06, + "loss": 0.7755, + "num_input_tokens_seen": 121470980, + "step": 3735 + }, + { + "epoch": 0.33692564368489875, + "flos": 21986917196640.0, + "grad_norm": 1.8899568370149142, + "learning_rate": 3.0906251091570565e-06, + "loss": 0.7931, + "num_input_tokens_seen": 121499475, + "step": 3736 + }, + { + "epoch": 0.33701582720836903, + "flos": 27414444953760.0, + "grad_norm": 1.6005081055426096, + "learning_rate": 3.0901353581274094e-06, + "loss": 0.7766, + "num_input_tokens_seen": 121530425, + "step": 3737 + }, + { + "epoch": 0.33710601073183927, + "flos": 32915580302880.0, + "grad_norm": 2.702467204414566, + "learning_rate": 3.089645514081402e-06, + "loss": 0.6479, + "num_input_tokens_seen": 121564150, + "step": 3738 + }, + { + "epoch": 0.33719619425530956, + "flos": 16959027429120.0, + "grad_norm": 3.062208356399899, + "learning_rate": 3.0891555770608323e-06, + "loss": 0.6849, + "num_input_tokens_seen": 121589145, + "step": 3739 + }, + { + "epoch": 0.33728637777877984, + "flos": 24717949295520.0, + "grad_norm": 1.6811246889062894, + "learning_rate": 3.088665547107503e-06, + "loss": 0.7863, + "num_input_tokens_seen": 121619725, + "step": 3740 + }, + { + "epoch": 0.3373765613022501, + "flos": 24425340245760.0, + "grad_norm": 1.552802548621325, + "learning_rate": 3.0881754242632254e-06, + "loss": 0.841, + "num_input_tokens_seen": 121648905, + "step": 3741 + }, + { + "epoch": 0.33746674482572037, + "flos": 22240789622880.0, + "grad_norm": 1.39662419399483, + "learning_rate": 3.0876852085698213e-06, + "loss": 0.8102, + "num_input_tokens_seen": 121678505, + "step": 3742 + }, + { + "epoch": 0.3375569283491906, + "flos": 25191938272320.0, + "grad_norm": 1.848229414125687, + "learning_rate": 3.087194900069117e-06, + "loss": 0.7905, + "num_input_tokens_seen": 121707720, + "step": 3743 + }, + { + "epoch": 0.3376471118726609, + "flos": 24275894876160.0, + "grad_norm": 2.3900244463384595, + "learning_rate": 3.08670449880295e-06, + "loss": 0.8212, + "num_input_tokens_seen": 121734965, + "step": 3744 + }, + { + "epoch": 0.3377372953961311, + "flos": 27958844923200.0, + "grad_norm": 2.7320203833817973, + "learning_rate": 3.086214004813163e-06, + "loss": 0.7831, + "num_input_tokens_seen": 121763865, + "step": 3745 + }, + { + "epoch": 0.3378274789196014, + "flos": 22639349689440.0, + "grad_norm": 1.8568286931322011, + "learning_rate": 3.0857234181416074e-06, + "loss": 0.829, + "num_input_tokens_seen": 121792785, + "step": 3746 + }, + { + "epoch": 0.33791766244307164, + "flos": 26282373680640.0, + "grad_norm": 1.841556213772109, + "learning_rate": 3.085232738830143e-06, + "loss": 0.8153, + "num_input_tokens_seen": 121819220, + "step": 3747 + }, + { + "epoch": 0.33800784596654193, + "flos": 21440361381120.0, + "grad_norm": 2.098797535633453, + "learning_rate": 3.084741966920638e-06, + "loss": 0.8283, + "num_input_tokens_seen": 121846915, + "step": 3748 + }, + { + "epoch": 0.33809802949001216, + "flos": 22569459073440.0, + "grad_norm": 1.7872437139996329, + "learning_rate": 3.084251102454966e-06, + "loss": 0.7256, + "num_input_tokens_seen": 121876560, + "step": 3749 + }, + { + "epoch": 0.33818821301348245, + "flos": 70092972723360.0, + "grad_norm": 0.7310852876007538, + "learning_rate": 3.083760145475013e-06, + "loss": 0.6136, + "num_input_tokens_seen": 121956960, + "step": 3750 + }, + { + "epoch": 0.3382783965369527, + "flos": 19180865054880.0, + "grad_norm": 3.2470802983869005, + "learning_rate": 3.0832690960226678e-06, + "loss": 0.7094, + "num_input_tokens_seen": 121984395, + "step": 3751 + }, + { + "epoch": 0.338368580060423, + "flos": 46902197979840.0, + "grad_norm": 2.0957119986851978, + "learning_rate": 3.08277795413983e-06, + "loss": 0.7343, + "num_input_tokens_seen": 122016580, + "step": 3752 + }, + { + "epoch": 0.3384587635838932, + "flos": 25191009028320.0, + "grad_norm": 1.7523207943178472, + "learning_rate": 3.0822867198684073e-06, + "loss": 0.8166, + "num_input_tokens_seen": 122046925, + "step": 3753 + }, + { + "epoch": 0.3385489471073635, + "flos": 22060115921280.0, + "grad_norm": 1.755337341578324, + "learning_rate": 3.081795393250314e-06, + "loss": 0.7788, + "num_input_tokens_seen": 122077790, + "step": 3754 + }, + { + "epoch": 0.33863913063083373, + "flos": 20746070004960.0, + "grad_norm": 1.9975286951021123, + "learning_rate": 3.081303974327473e-06, + "loss": 0.7351, + "num_input_tokens_seen": 122104185, + "step": 3755 + }, + { + "epoch": 0.338729314154304, + "flos": 19582064174400.0, + "grad_norm": 1.7797560868682278, + "learning_rate": 3.080812463141814e-06, + "loss": 0.7636, + "num_input_tokens_seen": 122133305, + "step": 3756 + }, + { + "epoch": 0.33881949767777425, + "flos": 32624569552800.0, + "grad_norm": 1.5055177451981652, + "learning_rate": 3.080320859735276e-06, + "loss": 0.6397, + "num_input_tokens_seen": 122165075, + "step": 3757 + }, + { + "epoch": 0.33890968120124454, + "flos": 26901682183680.0, + "grad_norm": 2.1356649363029323, + "learning_rate": 3.079829164149806e-06, + "loss": 0.8219, + "num_input_tokens_seen": 122195565, + "step": 3758 + }, + { + "epoch": 0.3389998647247148, + "flos": 23989604685600.0, + "grad_norm": 1.6274118131779125, + "learning_rate": 3.0793373764273573e-06, + "loss": 0.7595, + "num_input_tokens_seen": 122225415, + "step": 3759 + }, + { + "epoch": 0.33909004824818506, + "flos": 23400037724160.0, + "grad_norm": 3.0904745622670204, + "learning_rate": 3.078845496609892e-06, + "loss": 0.7921, + "num_input_tokens_seen": 122251535, + "step": 3760 + }, + { + "epoch": 0.3391802317716553, + "flos": 24894274135200.0, + "grad_norm": 1.6906666889797373, + "learning_rate": 3.078353524739381e-06, + "loss": 0.8549, + "num_input_tokens_seen": 122278635, + "step": 3761 + }, + { + "epoch": 0.3392704152951256, + "flos": 20419630740000.0, + "grad_norm": 1.697868564915084, + "learning_rate": 3.077861460857801e-06, + "loss": 0.8524, + "num_input_tokens_seen": 122306240, + "step": 3762 + }, + { + "epoch": 0.3393605988185958, + "flos": 21768398945760.0, + "grad_norm": 1.737189680970647, + "learning_rate": 3.077369305007138e-06, + "loss": 0.7872, + "num_input_tokens_seen": 122336515, + "step": 3763 + }, + { + "epoch": 0.3394507823420661, + "flos": 17899311242400.0, + "grad_norm": 1.7982785921129363, + "learning_rate": 3.0768770572293852e-06, + "loss": 0.805, + "num_input_tokens_seen": 122364420, + "step": 3764 + }, + { + "epoch": 0.3395409658655364, + "flos": 25848236420160.0, + "grad_norm": 2.4838564349533625, + "learning_rate": 3.0763847175665437e-06, + "loss": 0.8483, + "num_input_tokens_seen": 122391620, + "step": 3765 + }, + { + "epoch": 0.3396311493890066, + "flos": 22058108754240.0, + "grad_norm": 1.940319655225957, + "learning_rate": 3.0758922860606237e-06, + "loss": 0.7097, + "num_input_tokens_seen": 122422595, + "step": 3766 + }, + { + "epoch": 0.3397213329124769, + "flos": 22168631651520.0, + "grad_norm": 3.7537017353726507, + "learning_rate": 3.0753997627536404e-06, + "loss": 0.7134, + "num_input_tokens_seen": 122451250, + "step": 3767 + }, + { + "epoch": 0.33981151643594715, + "flos": 24203811244320.0, + "grad_norm": 2.072838965513994, + "learning_rate": 3.0749071476876203e-06, + "loss": 0.7223, + "num_input_tokens_seen": 122477420, + "step": 3768 + }, + { + "epoch": 0.33990169995941744, + "flos": 29779297580640.0, + "grad_norm": 1.733059495911675, + "learning_rate": 3.0744144409045952e-06, + "loss": 0.7372, + "num_input_tokens_seen": 122507355, + "step": 3769 + }, + { + "epoch": 0.33999188348288767, + "flos": 17796631164480.0, + "grad_norm": 3.9235822753446787, + "learning_rate": 3.0739216424466056e-06, + "loss": 0.7948, + "num_input_tokens_seen": 122534135, + "step": 3770 + }, + { + "epoch": 0.34008206700635796, + "flos": 25374544801440.0, + "grad_norm": 2.4043698582545914, + "learning_rate": 3.0734287523557002e-06, + "loss": 0.7511, + "num_input_tokens_seen": 122564515, + "step": 3771 + }, + { + "epoch": 0.3401722505298282, + "flos": 20159662473120.0, + "grad_norm": 1.776451675058387, + "learning_rate": 3.0729357706739348e-06, + "loss": 0.7213, + "num_input_tokens_seen": 122590485, + "step": 3772 + }, + { + "epoch": 0.3402624340532985, + "flos": 29270177447040.0, + "grad_norm": 1.788397053060064, + "learning_rate": 3.0724426974433737e-06, + "loss": 0.7505, + "num_input_tokens_seen": 122617985, + "step": 3773 + }, + { + "epoch": 0.3403526175767687, + "flos": 26577027067200.0, + "grad_norm": 1.5359346779751781, + "learning_rate": 3.0719495327060874e-06, + "loss": 0.7464, + "num_input_tokens_seen": 122648280, + "step": 3774 + }, + { + "epoch": 0.340442801100239, + "flos": 20966075046240.0, + "grad_norm": 1.752246028439014, + "learning_rate": 3.071456276504157e-06, + "loss": 0.7757, + "num_input_tokens_seen": 122676925, + "step": 3775 + }, + { + "epoch": 0.34053298462370923, + "flos": 25371980088000.0, + "grad_norm": 1.5516361215263996, + "learning_rate": 3.070962928879669e-06, + "loss": 0.7661, + "num_input_tokens_seen": 122706940, + "step": 3776 + }, + { + "epoch": 0.3406231681471795, + "flos": 24384745134240.0, + "grad_norm": 1.6519388915806552, + "learning_rate": 3.0704694898747185e-06, + "loss": 0.7823, + "num_input_tokens_seen": 122734965, + "step": 3777 + }, + { + "epoch": 0.34071335167064976, + "flos": 17937341640480.0, + "grad_norm": 1.8022795627310872, + "learning_rate": 3.069975959531408e-06, + "loss": 0.8896, + "num_input_tokens_seen": 122760480, + "step": 3778 + }, + { + "epoch": 0.34080353519412004, + "flos": 70210186177440.0, + "grad_norm": 0.6663208650758605, + "learning_rate": 3.06948233789185e-06, + "loss": 0.6474, + "num_input_tokens_seen": 122847270, + "step": 3779 + }, + { + "epoch": 0.3408937187175903, + "flos": 19764299005920.0, + "grad_norm": 1.8900762975520955, + "learning_rate": 3.0689886249981614e-06, + "loss": 0.7858, + "num_input_tokens_seen": 122875800, + "step": 3780 + }, + { + "epoch": 0.34098390224106057, + "flos": 18635387162400.0, + "grad_norm": 2.056872812117529, + "learning_rate": 3.0684948208924693e-06, + "loss": 0.7977, + "num_input_tokens_seen": 122903060, + "step": 3781 + }, + { + "epoch": 0.3410740857645308, + "flos": 20783840214720.0, + "grad_norm": 1.9774177118438196, + "learning_rate": 3.068000925616907e-06, + "loss": 0.7871, + "num_input_tokens_seen": 122930280, + "step": 3782 + }, + { + "epoch": 0.3411642692880011, + "flos": 19800396576480.0, + "grad_norm": 1.977593315228025, + "learning_rate": 3.067506939213617e-06, + "loss": 0.7625, + "num_input_tokens_seen": 122955905, + "step": 3783 + }, + { + "epoch": 0.3412544528114713, + "flos": 26175419080320.0, + "grad_norm": 1.3780177985122917, + "learning_rate": 3.0670128617247493e-06, + "loss": 0.7735, + "num_input_tokens_seen": 122986455, + "step": 3784 + }, + { + "epoch": 0.3413446363349416, + "flos": 24172991779680.0, + "grad_norm": 2.1921654132537904, + "learning_rate": 3.06651869319246e-06, + "loss": 0.6801, + "num_input_tokens_seen": 123017375, + "step": 3785 + }, + { + "epoch": 0.34143481985841184, + "flos": 13096704622080.0, + "grad_norm": 2.2544934649919495, + "learning_rate": 3.0660244336589154e-06, + "loss": 0.7687, + "num_input_tokens_seen": 123042060, + "step": 3786 + }, + { + "epoch": 0.34152500338188213, + "flos": 36705113252640.0, + "grad_norm": 1.9348769120685576, + "learning_rate": 3.065530083166288e-06, + "loss": 0.7836, + "num_input_tokens_seen": 123073160, + "step": 3787 + }, + { + "epoch": 0.3416151869053524, + "flos": 24425451755040.0, + "grad_norm": 2.6148691413770364, + "learning_rate": 3.0650356417567586e-06, + "loss": 0.7315, + "num_input_tokens_seen": 123103515, + "step": 3788 + }, + { + "epoch": 0.34170537042882265, + "flos": 16084248200160.0, + "grad_norm": 2.148196231126085, + "learning_rate": 3.0645411094725156e-06, + "loss": 0.8192, + "num_input_tokens_seen": 123128875, + "step": 3789 + }, + { + "epoch": 0.34179555395229294, + "flos": 24718581181440.0, + "grad_norm": 1.7533637144384084, + "learning_rate": 3.0640464863557556e-06, + "loss": 0.7672, + "num_input_tokens_seen": 123156620, + "step": 3790 + }, + { + "epoch": 0.3418857374757632, + "flos": 31422793512480.0, + "grad_norm": 4.591148418296682, + "learning_rate": 3.063551772448682e-06, + "loss": 0.8064, + "num_input_tokens_seen": 123192075, + "step": 3791 + }, + { + "epoch": 0.34197592099923346, + "flos": 26532640640160.0, + "grad_norm": 1.702609202822004, + "learning_rate": 3.0630569677935075e-06, + "loss": 0.7013, + "num_input_tokens_seen": 123218915, + "step": 3792 + }, + { + "epoch": 0.3420661045227037, + "flos": 26351297882880.0, + "grad_norm": 2.3468554459124213, + "learning_rate": 3.06256207243245e-06, + "loss": 0.8594, + "num_input_tokens_seen": 123246505, + "step": 3793 + }, + { + "epoch": 0.342156288046174, + "flos": 59407917456000.0, + "grad_norm": 0.6033682288261526, + "learning_rate": 3.0620670864077385e-06, + "loss": 0.5678, + "num_input_tokens_seen": 123339065, + "step": 3794 + }, + { + "epoch": 0.3422464715696442, + "flos": 64373759426880.0, + "grad_norm": 0.6217938681734795, + "learning_rate": 3.0615720097616063e-06, + "loss": 0.571, + "num_input_tokens_seen": 123434345, + "step": 3795 + }, + { + "epoch": 0.3423366550931145, + "flos": 23619856916640.0, + "grad_norm": 1.6341519497149217, + "learning_rate": 3.0610768425362967e-06, + "loss": 0.7209, + "num_input_tokens_seen": 123462070, + "step": 3796 + }, + { + "epoch": 0.34242683861658474, + "flos": 19181496940800.0, + "grad_norm": 2.4752338106323135, + "learning_rate": 3.0605815847740603e-06, + "loss": 0.7466, + "num_input_tokens_seen": 123489340, + "step": 3797 + }, + { + "epoch": 0.342517022140055, + "flos": 20601828401760.0, + "grad_norm": 1.6888554392326935, + "learning_rate": 3.0600862365171553e-06, + "loss": 0.7353, + "num_input_tokens_seen": 123515670, + "step": 3798 + }, + { + "epoch": 0.34260720566352526, + "flos": 19035731377440.0, + "grad_norm": 1.445436214616509, + "learning_rate": 3.0595907978078474e-06, + "loss": 0.7453, + "num_input_tokens_seen": 123544545, + "step": 3799 + }, + { + "epoch": 0.34269738918699555, + "flos": 23844099310560.0, + "grad_norm": 1.553671806988782, + "learning_rate": 3.05909526868841e-06, + "loss": 0.7357, + "num_input_tokens_seen": 123572120, + "step": 3800 + }, + { + "epoch": 0.3427875727104658, + "flos": 24536383519680.0, + "grad_norm": 1.8339170335912667, + "learning_rate": 3.0585996492011243e-06, + "loss": 0.776, + "num_input_tokens_seen": 123601470, + "step": 3801 + }, + { + "epoch": 0.34287775623393607, + "flos": 20233827611520.0, + "grad_norm": 1.7078936251317367, + "learning_rate": 3.05810393938828e-06, + "loss": 0.8157, + "num_input_tokens_seen": 123626950, + "step": 3802 + }, + { + "epoch": 0.3429679397574063, + "flos": 31456809576480.0, + "grad_norm": 2.5021920131062845, + "learning_rate": 3.0576081392921723e-06, + "loss": 0.6791, + "num_input_tokens_seen": 123656765, + "step": 3803 + }, + { + "epoch": 0.3430581232808766, + "flos": 24166970278560.0, + "grad_norm": 1.7940212642766071, + "learning_rate": 3.057112248955107e-06, + "loss": 0.8161, + "num_input_tokens_seen": 123684945, + "step": 3804 + }, + { + "epoch": 0.3431483068043468, + "flos": 15714314582400.0, + "grad_norm": 2.053940089524868, + "learning_rate": 3.0566162684193963e-06, + "loss": 0.7852, + "num_input_tokens_seen": 123711275, + "step": 3805 + }, + { + "epoch": 0.3432384903278171, + "flos": 28363240642080.0, + "grad_norm": 1.6873163900445332, + "learning_rate": 3.056120197727359e-06, + "loss": 0.7174, + "num_input_tokens_seen": 123746375, + "step": 3806 + }, + { + "epoch": 0.34332867385128735, + "flos": 26431484522400.0, + "grad_norm": 1.5840950020612463, + "learning_rate": 3.0556240369213236e-06, + "loss": 0.7594, + "num_input_tokens_seen": 123776090, + "step": 3807 + }, + { + "epoch": 0.34341885737475764, + "flos": 24022208298720.0, + "grad_norm": 2.141648987376745, + "learning_rate": 3.055127786043624e-06, + "loss": 0.6814, + "num_input_tokens_seen": 123806375, + "step": 3808 + }, + { + "epoch": 0.34350904089822787, + "flos": 23443680756000.0, + "grad_norm": 2.4213223718984644, + "learning_rate": 3.054631445136604e-06, + "loss": 0.8521, + "num_input_tokens_seen": 123833500, + "step": 3809 + }, + { + "epoch": 0.34359922442169816, + "flos": 23844062140800.0, + "grad_norm": 2.529908660777828, + "learning_rate": 3.0541350142426147e-06, + "loss": 0.779, + "num_input_tokens_seen": 123858925, + "step": 3810 + }, + { + "epoch": 0.3436894079451684, + "flos": 24536829556800.0, + "grad_norm": 2.5194457047008467, + "learning_rate": 3.053638493404012e-06, + "loss": 0.7261, + "num_input_tokens_seen": 123889635, + "step": 3811 + }, + { + "epoch": 0.3437795914686387, + "flos": 24427273073280.0, + "grad_norm": 2.04225067297805, + "learning_rate": 3.0531418826631643e-06, + "loss": 0.7025, + "num_input_tokens_seen": 123918700, + "step": 3812 + }, + { + "epoch": 0.34386977499210897, + "flos": 26245904412480.0, + "grad_norm": 1.854510571879319, + "learning_rate": 3.052645182062444e-06, + "loss": 0.8471, + "num_input_tokens_seen": 123946480, + "step": 3813 + }, + { + "epoch": 0.3439599585155792, + "flos": 29449104169920.0, + "grad_norm": 2.2572844524705418, + "learning_rate": 3.0521483916442324e-06, + "loss": 0.8069, + "num_input_tokens_seen": 123975005, + "step": 3814 + }, + { + "epoch": 0.3440501420390495, + "flos": 24245632957920.0, + "grad_norm": 2.9115896087331086, + "learning_rate": 3.0516515114509183e-06, + "loss": 0.7643, + "num_input_tokens_seen": 124004050, + "step": 3815 + }, + { + "epoch": 0.3441403255625197, + "flos": 23225125335360.0, + "grad_norm": 1.5714403343998962, + "learning_rate": 3.0511545415249e-06, + "loss": 0.7176, + "num_input_tokens_seen": 124034995, + "step": 3816 + }, + { + "epoch": 0.34423050908599, + "flos": 20784397761120.0, + "grad_norm": 1.9635709428164658, + "learning_rate": 3.050657481908579e-06, + "loss": 0.8, + "num_input_tokens_seen": 124063010, + "step": 3817 + }, + { + "epoch": 0.34432069260946024, + "flos": 21694865693280.0, + "grad_norm": 1.7624953898917433, + "learning_rate": 3.0501603326443677e-06, + "loss": 0.8751, + "num_input_tokens_seen": 124090635, + "step": 3818 + }, + { + "epoch": 0.34441087613293053, + "flos": 22602694572480.0, + "grad_norm": 1.680352722447084, + "learning_rate": 3.049663093774687e-06, + "loss": 0.7532, + "num_input_tokens_seen": 124118430, + "step": 3819 + }, + { + "epoch": 0.34450105965640077, + "flos": 33608161870080.0, + "grad_norm": 3.173706578201432, + "learning_rate": 3.0491657653419643e-06, + "loss": 0.6636, + "num_input_tokens_seen": 124147410, + "step": 3820 + }, + { + "epoch": 0.34459124317987105, + "flos": 23334681818880.0, + "grad_norm": 4.037583454793095, + "learning_rate": 3.0486683473886325e-06, + "loss": 0.8136, + "num_input_tokens_seen": 124175875, + "step": 3821 + }, + { + "epoch": 0.3446814267033413, + "flos": 18999224939520.0, + "grad_norm": 2.0834393919333527, + "learning_rate": 3.0481708399571355e-06, + "loss": 0.8277, + "num_input_tokens_seen": 124203615, + "step": 3822 + }, + { + "epoch": 0.3447716102268116, + "flos": 23620823330400.0, + "grad_norm": 2.2210960342551127, + "learning_rate": 3.047673243089922e-06, + "loss": 0.7974, + "num_input_tokens_seen": 124233670, + "step": 3823 + }, + { + "epoch": 0.3448617937502818, + "flos": 23949269762400.0, + "grad_norm": 1.9198976594750217, + "learning_rate": 3.047175556829451e-06, + "loss": 0.8003, + "num_input_tokens_seen": 124262620, + "step": 3824 + }, + { + "epoch": 0.3449519772737521, + "flos": 21476310272640.0, + "grad_norm": 1.9752432299180362, + "learning_rate": 3.046677781218188e-06, + "loss": 0.8181, + "num_input_tokens_seen": 124289930, + "step": 3825 + }, + { + "epoch": 0.34504216079722233, + "flos": 20566325547360.0, + "grad_norm": 2.5922245917445905, + "learning_rate": 3.0461799162986043e-06, + "loss": 0.744, + "num_input_tokens_seen": 124319085, + "step": 3826 + }, + { + "epoch": 0.3451323443206926, + "flos": 20383421660160.0, + "grad_norm": 1.9777185730639086, + "learning_rate": 3.045681962113183e-06, + "loss": 0.6945, + "num_input_tokens_seen": 124344280, + "step": 3827 + }, + { + "epoch": 0.34522252784416285, + "flos": 23552642523360.0, + "grad_norm": 1.7709007100769856, + "learning_rate": 3.0451839187044095e-06, + "loss": 0.7171, + "num_input_tokens_seen": 124370400, + "step": 3828 + }, + { + "epoch": 0.34531271136763314, + "flos": 33098967396960.0, + "grad_norm": 1.95718791281318, + "learning_rate": 3.0446857861147816e-06, + "loss": 0.7878, + "num_input_tokens_seen": 124399735, + "step": 3829 + }, + { + "epoch": 0.3454028948911034, + "flos": 23407397336640.0, + "grad_norm": 2.0570070351502134, + "learning_rate": 3.044187564386802e-06, + "loss": 0.7447, + "num_input_tokens_seen": 124427560, + "step": 3830 + }, + { + "epoch": 0.34549307841457366, + "flos": 21658136236800.0, + "grad_norm": 1.8177424940453484, + "learning_rate": 3.0436892535629818e-06, + "loss": 0.83, + "num_input_tokens_seen": 124453915, + "step": 3831 + }, + { + "epoch": 0.3455832619380439, + "flos": 23552791202400.0, + "grad_norm": 2.948645529713584, + "learning_rate": 3.0431908536858393e-06, + "loss": 0.8502, + "num_input_tokens_seen": 124482725, + "step": 3832 + }, + { + "epoch": 0.3456734454615142, + "flos": 32297015195040.0, + "grad_norm": 2.0242075982738092, + "learning_rate": 3.0426923647979016e-06, + "loss": 0.7581, + "num_input_tokens_seen": 124510245, + "step": 3833 + }, + { + "epoch": 0.3457636289849844, + "flos": 23188581727680.0, + "grad_norm": 2.093168946673643, + "learning_rate": 3.0421937869417016e-06, + "loss": 0.7056, + "num_input_tokens_seen": 124536295, + "step": 3834 + }, + { + "epoch": 0.3458538125084547, + "flos": 26719893389280.0, + "grad_norm": 2.068353611521125, + "learning_rate": 3.041695120159782e-06, + "loss": 0.8319, + "num_input_tokens_seen": 124563055, + "step": 3835 + }, + { + "epoch": 0.345943996031925, + "flos": 22678234992000.0, + "grad_norm": 1.7817584198942384, + "learning_rate": 3.04119636449469e-06, + "loss": 0.7035, + "num_input_tokens_seen": 124591455, + "step": 3836 + }, + { + "epoch": 0.3460341795553952, + "flos": 30291168276480.0, + "grad_norm": 1.5100522972044241, + "learning_rate": 3.040697519988983e-06, + "loss": 0.7991, + "num_input_tokens_seen": 124625315, + "step": 3837 + }, + { + "epoch": 0.3461243630788655, + "flos": 59801236586400.0, + "grad_norm": 0.6431486617351904, + "learning_rate": 3.040198586685226e-06, + "loss": 0.591, + "num_input_tokens_seen": 124717160, + "step": 3838 + }, + { + "epoch": 0.34621454660233575, + "flos": 26716548110880.0, + "grad_norm": 2.28088072831967, + "learning_rate": 3.039699564625989e-06, + "loss": 0.7186, + "num_input_tokens_seen": 124744225, + "step": 3839 + }, + { + "epoch": 0.34630473012580604, + "flos": 70719938196960.0, + "grad_norm": 0.6720595439035967, + "learning_rate": 3.039200453853853e-06, + "loss": 0.5847, + "num_input_tokens_seen": 124833445, + "step": 3840 + }, + { + "epoch": 0.34639491364927627, + "flos": 23219772889920.0, + "grad_norm": 2.0278537233782887, + "learning_rate": 3.038701254411404e-06, + "loss": 0.8199, + "num_input_tokens_seen": 124859295, + "step": 3841 + }, + { + "epoch": 0.34648509717274656, + "flos": 28214798856000.0, + "grad_norm": 1.775954623806848, + "learning_rate": 3.0382019663412367e-06, + "loss": 0.7879, + "num_input_tokens_seen": 124885400, + "step": 3842 + }, + { + "epoch": 0.3465752806962168, + "flos": 16485075622080.0, + "grad_norm": 2.1634802740867096, + "learning_rate": 3.0377025896859532e-06, + "loss": 0.8069, + "num_input_tokens_seen": 124913590, + "step": 3843 + }, + { + "epoch": 0.3466654642196871, + "flos": 25155617683200.0, + "grad_norm": 1.5702594349563572, + "learning_rate": 3.0372031244881627e-06, + "loss": 0.8757, + "num_input_tokens_seen": 124940825, + "step": 3844 + }, + { + "epoch": 0.3467556477431573, + "flos": 22023312125280.0, + "grad_norm": 3.9435779951161676, + "learning_rate": 3.0367035707904826e-06, + "loss": 0.7282, + "num_input_tokens_seen": 124968620, + "step": 3845 + }, + { + "epoch": 0.3468458312666276, + "flos": 22017848170560.0, + "grad_norm": 2.100740179679713, + "learning_rate": 3.036203928635537e-06, + "loss": 0.6841, + "num_input_tokens_seen": 124994675, + "step": 3846 + }, + { + "epoch": 0.34693601479009784, + "flos": 24208085766720.0, + "grad_norm": 2.0406451751506136, + "learning_rate": 3.035704198065959e-06, + "loss": 0.8346, + "num_input_tokens_seen": 125020095, + "step": 3847 + }, + { + "epoch": 0.3470261983135681, + "flos": 25556853972480.0, + "grad_norm": 1.7884895706661847, + "learning_rate": 3.0352043791243886e-06, + "loss": 0.7563, + "num_input_tokens_seen": 125047675, + "step": 3848 + }, + { + "epoch": 0.34711638183703836, + "flos": 60858064798080.0, + "grad_norm": 0.6596098449987291, + "learning_rate": 3.034704471853472e-06, + "loss": 0.5883, + "num_input_tokens_seen": 125139500, + "step": 3849 + }, + { + "epoch": 0.34720656536050865, + "flos": 21581294875680.0, + "grad_norm": 1.5040994891578523, + "learning_rate": 3.0342044762958646e-06, + "loss": 0.8171, + "num_input_tokens_seen": 125167535, + "step": 3850 + }, + { + "epoch": 0.3472967488839789, + "flos": 21440101192800.0, + "grad_norm": 1.828479765027191, + "learning_rate": 3.0337043924942286e-06, + "loss": 0.8081, + "num_input_tokens_seen": 125194955, + "step": 3851 + }, + { + "epoch": 0.34738693240744917, + "flos": 22423544831040.0, + "grad_norm": 2.2394039939815453, + "learning_rate": 3.0332042204912343e-06, + "loss": 0.757, + "num_input_tokens_seen": 125225570, + "step": 3852 + }, + { + "epoch": 0.3474771159309194, + "flos": 67685214726240.0, + "grad_norm": 0.6646351308776928, + "learning_rate": 3.0327039603295587e-06, + "loss": 0.5923, + "num_input_tokens_seen": 125320550, + "step": 3853 + }, + { + "epoch": 0.3475672994543897, + "flos": 20492346257760.0, + "grad_norm": 2.2153323003051137, + "learning_rate": 3.032203612051887e-06, + "loss": 0.7291, + "num_input_tokens_seen": 125346180, + "step": 3854 + }, + { + "epoch": 0.3476574829778599, + "flos": 15501260286240.0, + "grad_norm": 1.9151224121132038, + "learning_rate": 3.0317031757009116e-06, + "loss": 0.7517, + "num_input_tokens_seen": 125371525, + "step": 3855 + }, + { + "epoch": 0.3477476665013302, + "flos": 28318445347680.0, + "grad_norm": 2.1371984971878106, + "learning_rate": 3.0312026513193326e-06, + "loss": 0.731, + "num_input_tokens_seen": 125399795, + "step": 3856 + }, + { + "epoch": 0.34783785002480044, + "flos": 33689575111680.0, + "grad_norm": 2.148705705027477, + "learning_rate": 3.0307020389498573e-06, + "loss": 0.748, + "num_input_tokens_seen": 125435335, + "step": 3857 + }, + { + "epoch": 0.34792803354827073, + "flos": 41105480000160.0, + "grad_norm": 1.4851162314604422, + "learning_rate": 3.0302013386352004e-06, + "loss": 0.6976, + "num_input_tokens_seen": 125467335, + "step": 3858 + }, + { + "epoch": 0.348018217071741, + "flos": 22168445802720.0, + "grad_norm": 1.5225408018060016, + "learning_rate": 3.0297005504180854e-06, + "loss": 0.7601, + "num_input_tokens_seen": 125498955, + "step": 3859 + }, + { + "epoch": 0.34810840059521125, + "flos": 70179664070880.0, + "grad_norm": 0.6028648002959177, + "learning_rate": 3.0291996743412417e-06, + "loss": 0.5119, + "num_input_tokens_seen": 125593670, + "step": 3860 + }, + { + "epoch": 0.34819858411868154, + "flos": 23002815768960.0, + "grad_norm": 1.6295594293994626, + "learning_rate": 3.0286987104474063e-06, + "loss": 0.858, + "num_input_tokens_seen": 125620255, + "step": 3861 + }, + { + "epoch": 0.3482887676421518, + "flos": 35537427615840.0, + "grad_norm": 1.3133193138184343, + "learning_rate": 3.028197658779325e-06, + "loss": 0.7413, + "num_input_tokens_seen": 125653185, + "step": 3862 + }, + { + "epoch": 0.34837895116562206, + "flos": 27155591779680.0, + "grad_norm": 1.9056553773094655, + "learning_rate": 3.0276965193797503e-06, + "loss": 0.7869, + "num_input_tokens_seen": 125683275, + "step": 3863 + }, + { + "epoch": 0.3484691346890923, + "flos": 45120333266880.0, + "grad_norm": 1.4140171022662362, + "learning_rate": 3.0271952922914423e-06, + "loss": 0.7348, + "num_input_tokens_seen": 125715095, + "step": 3864 + }, + { + "epoch": 0.3485593182125626, + "flos": 17573912730720.0, + "grad_norm": 2.495054694361718, + "learning_rate": 3.0266939775571675e-06, + "loss": 0.7468, + "num_input_tokens_seen": 125741125, + "step": 3865 + }, + { + "epoch": 0.3486495017360328, + "flos": 32547282154560.0, + "grad_norm": 2.0544652782411026, + "learning_rate": 3.026192575219701e-06, + "loss": 0.759, + "num_input_tokens_seen": 125771710, + "step": 3866 + }, + { + "epoch": 0.3487396852595031, + "flos": 27560507875200.0, + "grad_norm": 1.9901156553848476, + "learning_rate": 3.025691085321826e-06, + "loss": 0.659, + "num_input_tokens_seen": 125803625, + "step": 3867 + }, + { + "epoch": 0.34882986878297334, + "flos": 25520124516000.0, + "grad_norm": 1.8220261133250943, + "learning_rate": 3.025189507906332e-06, + "loss": 0.7531, + "num_input_tokens_seen": 125831890, + "step": 3868 + }, + { + "epoch": 0.34892005230644363, + "flos": 26321965208640.0, + "grad_norm": 1.7505873648082153, + "learning_rate": 3.0246878430160166e-06, + "loss": 0.7893, + "num_input_tokens_seen": 125862635, + "step": 3869 + }, + { + "epoch": 0.34901023582991386, + "flos": 25734814281600.0, + "grad_norm": 1.955114099436446, + "learning_rate": 3.024186090693684e-06, + "loss": 0.7115, + "num_input_tokens_seen": 125892390, + "step": 3870 + }, + { + "epoch": 0.34910041935338415, + "flos": 21112286646720.0, + "grad_norm": 1.6749935121552946, + "learning_rate": 3.023684250982147e-06, + "loss": 0.794, + "num_input_tokens_seen": 125918715, + "step": 3871 + }, + { + "epoch": 0.3491906028768544, + "flos": 19071457250400.0, + "grad_norm": 1.8397794740630176, + "learning_rate": 3.0231823239242252e-06, + "loss": 0.6884, + "num_input_tokens_seen": 125947145, + "step": 3872 + }, + { + "epoch": 0.34928078640032467, + "flos": 25228593389280.0, + "grad_norm": 1.8181013176176533, + "learning_rate": 3.0226803095627457e-06, + "loss": 0.7245, + "num_input_tokens_seen": 125976690, + "step": 3873 + }, + { + "epoch": 0.3493709699237949, + "flos": 25512727733760.0, + "grad_norm": 2.3103416267869563, + "learning_rate": 3.022178207940543e-06, + "loss": 0.7495, + "num_input_tokens_seen": 126004610, + "step": 3874 + }, + { + "epoch": 0.3494611534472652, + "flos": 25301160228000.0, + "grad_norm": 1.5506664781354382, + "learning_rate": 3.02167601910046e-06, + "loss": 0.7815, + "num_input_tokens_seen": 126032145, + "step": 3875 + }, + { + "epoch": 0.3495513369707354, + "flos": 23039359376640.0, + "grad_norm": 1.8998712219967724, + "learning_rate": 3.021173743085345e-06, + "loss": 0.7891, + "num_input_tokens_seen": 126060375, + "step": 3876 + }, + { + "epoch": 0.3496415204942057, + "flos": 70547516182080.0, + "grad_norm": 0.7987819369144395, + "learning_rate": 3.0206713799380557e-06, + "loss": 0.5786, + "num_input_tokens_seen": 126149780, + "step": 3877 + }, + { + "epoch": 0.34973170401767595, + "flos": 27633595090560.0, + "grad_norm": 1.7098407452583144, + "learning_rate": 3.0201689297014565e-06, + "loss": 0.7759, + "num_input_tokens_seen": 126177655, + "step": 3878 + }, + { + "epoch": 0.34982188754114624, + "flos": 25774665997920.0, + "grad_norm": 1.5634152592798942, + "learning_rate": 3.0196663924184187e-06, + "loss": 0.8114, + "num_input_tokens_seen": 126208145, + "step": 3879 + }, + { + "epoch": 0.34991207106461647, + "flos": 23115420172800.0, + "grad_norm": 2.6980848124968246, + "learning_rate": 3.019163768131822e-06, + "loss": 0.8029, + "num_input_tokens_seen": 126236945, + "step": 3880 + }, + { + "epoch": 0.35000225458808676, + "flos": 22131827855520.0, + "grad_norm": 8.453562726940655, + "learning_rate": 3.0186610568845533e-06, + "loss": 0.6891, + "num_input_tokens_seen": 126264630, + "step": 3881 + }, + { + "epoch": 0.350092438111557, + "flos": 24390915314400.0, + "grad_norm": 2.016552970245369, + "learning_rate": 3.018158258719507e-06, + "loss": 0.702, + "num_input_tokens_seen": 126292650, + "step": 3882 + }, + { + "epoch": 0.3501826216350273, + "flos": 24135407418720.0, + "grad_norm": 1.634862097077852, + "learning_rate": 3.0176553736795827e-06, + "loss": 0.794, + "num_input_tokens_seen": 126322775, + "step": 3883 + }, + { + "epoch": 0.35027280515849757, + "flos": 64955818096800.0, + "grad_norm": 0.5349514980256324, + "learning_rate": 3.017152401807691e-06, + "loss": 0.5168, + "num_input_tokens_seen": 126418555, + "step": 3884 + }, + { + "epoch": 0.3503629886819678, + "flos": 25048365724800.0, + "grad_norm": 1.6570424037703642, + "learning_rate": 3.0166493431467476e-06, + "loss": 0.7511, + "num_input_tokens_seen": 126452465, + "step": 3885 + }, + { + "epoch": 0.3504531722054381, + "flos": 22278411153600.0, + "grad_norm": 1.5461901294956202, + "learning_rate": 3.016146197739677e-06, + "loss": 0.7391, + "num_input_tokens_seen": 126482475, + "step": 3886 + }, + { + "epoch": 0.3505433557289083, + "flos": 23698779784320.0, + "grad_norm": 1.6978839410853823, + "learning_rate": 3.0156429656294097e-06, + "loss": 0.7008, + "num_input_tokens_seen": 126510105, + "step": 3887 + }, + { + "epoch": 0.3506335392523786, + "flos": 34445914284480.0, + "grad_norm": 1.6684626042488138, + "learning_rate": 3.0151396468588844e-06, + "loss": 0.7011, + "num_input_tokens_seen": 126541030, + "step": 3888 + }, + { + "epoch": 0.35072372277584885, + "flos": 28208405657280.0, + "grad_norm": 2.1540386636605846, + "learning_rate": 3.014636241471047e-06, + "loss": 0.717, + "num_input_tokens_seen": 126569030, + "step": 3889 + }, + { + "epoch": 0.35081390629931913, + "flos": 19070528006400.0, + "grad_norm": 1.6896674753381435, + "learning_rate": 3.0141327495088514e-06, + "loss": 0.7558, + "num_input_tokens_seen": 126596295, + "step": 3890 + }, + { + "epoch": 0.35090408982278937, + "flos": 28395138029760.0, + "grad_norm": 1.6233253653224355, + "learning_rate": 3.0136291710152566e-06, + "loss": 0.7202, + "num_input_tokens_seen": 126626770, + "step": 3891 + }, + { + "epoch": 0.35099427334625966, + "flos": 32624532383040.0, + "grad_norm": 1.8100010317594, + "learning_rate": 3.0131255060332325e-06, + "loss": 0.8028, + "num_input_tokens_seen": 126656605, + "step": 3892 + }, + { + "epoch": 0.3510844568697299, + "flos": 25191975442080.0, + "grad_norm": 1.571689210988257, + "learning_rate": 3.012621754605754e-06, + "loss": 0.783, + "num_input_tokens_seen": 126686320, + "step": 3893 + }, + { + "epoch": 0.3511746403932002, + "flos": 22674964053120.0, + "grad_norm": 1.7240318981813112, + "learning_rate": 3.0121179167758035e-06, + "loss": 0.7899, + "num_input_tokens_seen": 126712550, + "step": 3894 + }, + { + "epoch": 0.3512648239166704, + "flos": 22053202345920.0, + "grad_norm": 3.117724379576687, + "learning_rate": 3.0116139925863717e-06, + "loss": 0.8448, + "num_input_tokens_seen": 126735745, + "step": 3895 + }, + { + "epoch": 0.3513550074401407, + "flos": 22824892629600.0, + "grad_norm": 1.927885998422261, + "learning_rate": 3.011109982080456e-06, + "loss": 0.8028, + "num_input_tokens_seen": 126764010, + "step": 3896 + }, + { + "epoch": 0.35144519096361093, + "flos": 17612946712320.0, + "grad_norm": 2.164587955512753, + "learning_rate": 3.0106058853010614e-06, + "loss": 0.7084, + "num_input_tokens_seen": 126789715, + "step": 3897 + }, + { + "epoch": 0.3515353744870812, + "flos": 38815907604480.0, + "grad_norm": 1.6480350816751836, + "learning_rate": 3.010101702291201e-06, + "loss": 0.7474, + "num_input_tokens_seen": 126820355, + "step": 3898 + }, + { + "epoch": 0.35162555801055145, + "flos": 18306940730400.0, + "grad_norm": 1.8321641820900405, + "learning_rate": 3.009597433093893e-06, + "loss": 0.7615, + "num_input_tokens_seen": 126846970, + "step": 3899 + }, + { + "epoch": 0.35171574153402174, + "flos": 22751619565440.0, + "grad_norm": 1.7588426750183481, + "learning_rate": 3.009093077752165e-06, + "loss": 0.7713, + "num_input_tokens_seen": 126875835, + "step": 3900 + }, + { + "epoch": 0.351805925057492, + "flos": 22896790412640.0, + "grad_norm": 5.703196903977375, + "learning_rate": 3.008588636309052e-06, + "loss": 0.7926, + "num_input_tokens_seen": 126902670, + "step": 3901 + }, + { + "epoch": 0.35189610858096226, + "flos": 23662422025440.0, + "grad_norm": 3.923193721269621, + "learning_rate": 3.0080841088075947e-06, + "loss": 0.7297, + "num_input_tokens_seen": 126932645, + "step": 3902 + }, + { + "epoch": 0.3519862921044325, + "flos": 20601828401760.0, + "grad_norm": 5.123348249034394, + "learning_rate": 3.0075794952908436e-06, + "loss": 0.8066, + "num_input_tokens_seen": 126959100, + "step": 3903 + }, + { + "epoch": 0.3520764756279028, + "flos": 20201670035520.0, + "grad_norm": 1.551239484064329, + "learning_rate": 3.0070747958018528e-06, + "loss": 0.7795, + "num_input_tokens_seen": 126988940, + "step": 3904 + }, + { + "epoch": 0.352166659151373, + "flos": 23990310911040.0, + "grad_norm": 1.4588097566134526, + "learning_rate": 3.0065700103836894e-06, + "loss": 0.7817, + "num_input_tokens_seen": 127018515, + "step": 3905 + }, + { + "epoch": 0.3522568426748433, + "flos": 23735918108160.0, + "grad_norm": 2.2951608087054343, + "learning_rate": 3.0060651390794214e-06, + "loss": 0.7829, + "num_input_tokens_seen": 127044925, + "step": 3906 + }, + { + "epoch": 0.3523470261983136, + "flos": 35647987682880.0, + "grad_norm": 1.4865362395107575, + "learning_rate": 3.005560181932128e-06, + "loss": 0.7462, + "num_input_tokens_seen": 127079435, + "step": 3907 + }, + { + "epoch": 0.35243720972178383, + "flos": 38745199253760.0, + "grad_norm": 1.7935346653578943, + "learning_rate": 3.005055138984896e-06, + "loss": 0.7223, + "num_input_tokens_seen": 127109940, + "step": 3908 + }, + { + "epoch": 0.3525273932452541, + "flos": 29416351877760.0, + "grad_norm": 2.0304141158723246, + "learning_rate": 3.0045500102808174e-06, + "loss": 0.7363, + "num_input_tokens_seen": 127141215, + "step": 3909 + }, + { + "epoch": 0.35261757676872435, + "flos": 21469731225120.0, + "grad_norm": 1.9078191333752283, + "learning_rate": 3.0040447958629927e-06, + "loss": 0.8476, + "num_input_tokens_seen": 127165640, + "step": 3910 + }, + { + "epoch": 0.35270776029219464, + "flos": 23334495970080.0, + "grad_norm": 1.675688769938717, + "learning_rate": 3.00353949577453e-06, + "loss": 0.7836, + "num_input_tokens_seen": 127191265, + "step": 3911 + }, + { + "epoch": 0.35279794381566487, + "flos": 19508679600960.0, + "grad_norm": 3.956160755567089, + "learning_rate": 3.003034110058544e-06, + "loss": 0.603, + "num_input_tokens_seen": 127214800, + "step": 3912 + }, + { + "epoch": 0.35288812733913516, + "flos": 21177456703200.0, + "grad_norm": 1.5605370971220585, + "learning_rate": 3.002528638758157e-06, + "loss": 0.7424, + "num_input_tokens_seen": 127242300, + "step": 3913 + }, + { + "epoch": 0.3529783108626054, + "flos": 18780669518880.0, + "grad_norm": 2.607522729644751, + "learning_rate": 3.0020230819164985e-06, + "loss": 0.7561, + "num_input_tokens_seen": 127269710, + "step": 3914 + }, + { + "epoch": 0.3530684943860757, + "flos": 24206190108960.0, + "grad_norm": 1.475041974003873, + "learning_rate": 3.0015174395767064e-06, + "loss": 0.7683, + "num_input_tokens_seen": 127299175, + "step": 3915 + }, + { + "epoch": 0.3531586779095459, + "flos": 19545594906240.0, + "grad_norm": 1.8104289488673933, + "learning_rate": 3.001011711781923e-06, + "loss": 0.7826, + "num_input_tokens_seen": 127327555, + "step": 3916 + }, + { + "epoch": 0.3532488614330162, + "flos": 66427126499520.0, + "grad_norm": 1.7407279978880175, + "learning_rate": 3.0005058985753017e-06, + "loss": 0.7167, + "num_input_tokens_seen": 127358885, + "step": 3917 + }, + { + "epoch": 0.35333904495648644, + "flos": 19977910848480.0, + "grad_norm": 1.68277205309701, + "learning_rate": 3e-06, + "loss": 0.7509, + "num_input_tokens_seen": 127386350, + "step": 3918 + }, + { + "epoch": 0.3534292284799567, + "flos": 70587590916960.0, + "grad_norm": 0.6876481720168859, + "learning_rate": 2.9994940160991843e-06, + "loss": 0.6051, + "num_input_tokens_seen": 127474185, + "step": 3919 + }, + { + "epoch": 0.35351941200342696, + "flos": 63941226199680.0, + "grad_norm": 0.6412903656626504, + "learning_rate": 2.9989879469160285e-06, + "loss": 0.5628, + "num_input_tokens_seen": 127564355, + "step": 3920 + }, + { + "epoch": 0.35360959552689725, + "flos": 23516098915680.0, + "grad_norm": 2.187282769452478, + "learning_rate": 2.9984817924937124e-06, + "loss": 0.7797, + "num_input_tokens_seen": 127592055, + "step": 3921 + }, + { + "epoch": 0.3536997790503675, + "flos": 17360300888160.0, + "grad_norm": 1.6727066110853985, + "learning_rate": 2.997975552875424e-06, + "loss": 0.7018, + "num_input_tokens_seen": 127618725, + "step": 3922 + }, + { + "epoch": 0.35378996257383777, + "flos": 26576358011520.0, + "grad_norm": 1.7666309158138693, + "learning_rate": 2.997469228104358e-06, + "loss": 0.7501, + "num_input_tokens_seen": 127650575, + "step": 3923 + }, + { + "epoch": 0.353880146097308, + "flos": 35320247476320.0, + "grad_norm": 1.5118988481827185, + "learning_rate": 2.996962818223718e-06, + "loss": 0.6595, + "num_input_tokens_seen": 127679930, + "step": 3924 + }, + { + "epoch": 0.3539703296207783, + "flos": 17796817013280.0, + "grad_norm": 2.241789566509367, + "learning_rate": 2.9964563232767135e-06, + "loss": 0.7099, + "num_input_tokens_seen": 127707775, + "step": 3925 + }, + { + "epoch": 0.3540605131442485, + "flos": 20747891323200.0, + "grad_norm": 3.505795993072607, + "learning_rate": 2.9959497433065617e-06, + "loss": 0.7603, + "num_input_tokens_seen": 127733740, + "step": 3926 + }, + { + "epoch": 0.3541506966677188, + "flos": 45809569555680.0, + "grad_norm": 1.7615429437379768, + "learning_rate": 2.9954430783564848e-06, + "loss": 0.6824, + "num_input_tokens_seen": 127767340, + "step": 3927 + }, + { + "epoch": 0.35424088019118904, + "flos": 18887215251840.0, + "grad_norm": 1.7380604807357638, + "learning_rate": 2.994936328469716e-06, + "loss": 0.7944, + "num_input_tokens_seen": 127795085, + "step": 3928 + }, + { + "epoch": 0.35433106371465933, + "flos": 22708385400960.0, + "grad_norm": 1.562661887365547, + "learning_rate": 2.994429493689494e-06, + "loss": 0.7642, + "num_input_tokens_seen": 127822870, + "step": 3929 + }, + { + "epoch": 0.35442124723812957, + "flos": 22309602315840.0, + "grad_norm": 1.3734990510829652, + "learning_rate": 2.9939225740590642e-06, + "loss": 0.7857, + "num_input_tokens_seen": 127853155, + "step": 3930 + }, + { + "epoch": 0.35451143076159986, + "flos": 32807250421440.0, + "grad_norm": 1.4883706373870185, + "learning_rate": 2.99341556962168e-06, + "loss": 0.7166, + "num_input_tokens_seen": 127883655, + "step": 3931 + }, + { + "epoch": 0.35460161428507014, + "flos": 30250907692800.0, + "grad_norm": 1.8964197558346159, + "learning_rate": 2.992908480420602e-06, + "loss": 0.8037, + "num_input_tokens_seen": 127915360, + "step": 3932 + }, + { + "epoch": 0.3546917978085404, + "flos": 25625220628320.0, + "grad_norm": 1.7452323546829733, + "learning_rate": 2.9924013064990974e-06, + "loss": 0.7485, + "num_input_tokens_seen": 127945360, + "step": 3933 + }, + { + "epoch": 0.35478198133201067, + "flos": 71134890127680.0, + "grad_norm": 0.6568192329290331, + "learning_rate": 2.991894047900441e-06, + "loss": 0.6102, + "num_input_tokens_seen": 128047630, + "step": 3934 + }, + { + "epoch": 0.3548721648554809, + "flos": 21182586130080.0, + "grad_norm": 1.6823704074190675, + "learning_rate": 2.991386704667916e-06, + "loss": 0.7762, + "num_input_tokens_seen": 128078550, + "step": 3935 + }, + { + "epoch": 0.3549623483789512, + "flos": 27774491415360.0, + "grad_norm": 1.803337682282756, + "learning_rate": 2.9908792768448097e-06, + "loss": 0.8251, + "num_input_tokens_seen": 128110530, + "step": 3936 + }, + { + "epoch": 0.3550525319024214, + "flos": 26831865907200.0, + "grad_norm": 1.9605851179741667, + "learning_rate": 2.990371764474421e-06, + "loss": 0.758, + "num_input_tokens_seen": 128140985, + "step": 3937 + }, + { + "epoch": 0.3551427154258917, + "flos": 21112063628160.0, + "grad_norm": 1.6673587871964117, + "learning_rate": 2.9898641676000518e-06, + "loss": 0.7275, + "num_input_tokens_seen": 128169535, + "step": 3938 + }, + { + "epoch": 0.35523289894936194, + "flos": 23078988074400.0, + "grad_norm": 1.9079493468096722, + "learning_rate": 2.9893564862650138e-06, + "loss": 0.7613, + "num_input_tokens_seen": 128197990, + "step": 3939 + }, + { + "epoch": 0.35532308247283223, + "flos": 21913049416320.0, + "grad_norm": 2.4315941495935136, + "learning_rate": 2.9888487205126254e-06, + "loss": 0.8149, + "num_input_tokens_seen": 128224140, + "step": 3940 + }, + { + "epoch": 0.35541326599630246, + "flos": 27888359591040.0, + "grad_norm": 1.6777445356529856, + "learning_rate": 2.9883408703862115e-06, + "loss": 0.7568, + "num_input_tokens_seen": 128253040, + "step": 3941 + }, + { + "epoch": 0.35550344951977275, + "flos": 13023803255520.0, + "grad_norm": 2.106347065395421, + "learning_rate": 2.987832935929105e-06, + "loss": 0.8404, + "num_input_tokens_seen": 128279995, + "step": 3942 + }, + { + "epoch": 0.355593633043243, + "flos": 17177359831200.0, + "grad_norm": 1.6614434491709584, + "learning_rate": 2.9873249171846454e-06, + "loss": 0.8266, + "num_input_tokens_seen": 128304675, + "step": 3943 + }, + { + "epoch": 0.3556838165667133, + "flos": 20456285856960.0, + "grad_norm": 1.939622393364005, + "learning_rate": 2.98681681419618e-06, + "loss": 0.7536, + "num_input_tokens_seen": 128330915, + "step": 3944 + }, + { + "epoch": 0.3557740000901835, + "flos": 21836393904000.0, + "grad_norm": 4.583548084198652, + "learning_rate": 2.9863086270070627e-06, + "loss": 0.8173, + "num_input_tokens_seen": 128360735, + "step": 3945 + }, + { + "epoch": 0.3558641836136538, + "flos": 21331065085920.0, + "grad_norm": 1.6582210975634526, + "learning_rate": 2.985800355660655e-06, + "loss": 0.7664, + "num_input_tokens_seen": 128389740, + "step": 3946 + }, + { + "epoch": 0.35595436713712403, + "flos": 64245989760960.0, + "grad_norm": 0.6724743276964195, + "learning_rate": 2.9852920002003252e-06, + "loss": 0.6083, + "num_input_tokens_seen": 128484480, + "step": 3947 + }, + { + "epoch": 0.3560445506605943, + "flos": 20164271523360.0, + "grad_norm": 1.9835531981065262, + "learning_rate": 2.9847835606694494e-06, + "loss": 0.8007, + "num_input_tokens_seen": 128512895, + "step": 3948 + }, + { + "epoch": 0.35613473418406455, + "flos": 29527432321440.0, + "grad_norm": 2.001432144482367, + "learning_rate": 2.98427503711141e-06, + "loss": 0.6893, + "num_input_tokens_seen": 128543005, + "step": 3949 + }, + { + "epoch": 0.35622491770753484, + "flos": 23553274409280.0, + "grad_norm": 1.8229205394885386, + "learning_rate": 2.9837664295695973e-06, + "loss": 0.7923, + "num_input_tokens_seen": 128572230, + "step": 3950 + }, + { + "epoch": 0.35631510123100507, + "flos": 33316259045760.0, + "grad_norm": 1.5306748068927563, + "learning_rate": 2.983257738087408e-06, + "loss": 0.7573, + "num_input_tokens_seen": 128604545, + "step": 3951 + }, + { + "epoch": 0.35640528475447536, + "flos": 25556445105120.0, + "grad_norm": 1.857194086300967, + "learning_rate": 2.982748962708247e-06, + "loss": 0.7902, + "num_input_tokens_seen": 128630890, + "step": 3952 + }, + { + "epoch": 0.3564954682779456, + "flos": 23990459590080.0, + "grad_norm": 1.7906337984145206, + "learning_rate": 2.982240103475526e-06, + "loss": 0.8153, + "num_input_tokens_seen": 128660615, + "step": 3953 + }, + { + "epoch": 0.3565856518014159, + "flos": 21694791353760.0, + "grad_norm": 2.2607769891354814, + "learning_rate": 2.981731160432663e-06, + "loss": 0.761, + "num_input_tokens_seen": 128688095, + "step": 3954 + }, + { + "epoch": 0.35667583532488617, + "flos": 29018051999520.0, + "grad_norm": 1.7176828712934291, + "learning_rate": 2.981222133623084e-06, + "loss": 0.6594, + "num_input_tokens_seen": 128718660, + "step": 3955 + }, + { + "epoch": 0.3567660188483564, + "flos": 20748337360320.0, + "grad_norm": 2.5204021515956185, + "learning_rate": 2.980713023090222e-06, + "loss": 0.7698, + "num_input_tokens_seen": 128746630, + "step": 3956 + }, + { + "epoch": 0.3568562023718267, + "flos": 23443271888640.0, + "grad_norm": 1.6819347164307785, + "learning_rate": 2.980203828877518e-06, + "loss": 0.7376, + "num_input_tokens_seen": 128775705, + "step": 3957 + }, + { + "epoch": 0.3569463858952969, + "flos": 17395989591360.0, + "grad_norm": 9.403628164503848, + "learning_rate": 2.9796945510284182e-06, + "loss": 0.8015, + "num_input_tokens_seen": 128798960, + "step": 3958 + }, + { + "epoch": 0.3570365694187672, + "flos": 23079694299840.0, + "grad_norm": 1.7493096761169942, + "learning_rate": 2.9791851895863774e-06, + "loss": 0.8193, + "num_input_tokens_seen": 128828730, + "step": 3959 + }, + { + "epoch": 0.35712675294223745, + "flos": 19433325030240.0, + "grad_norm": 2.0702300433166787, + "learning_rate": 2.978675744594857e-06, + "loss": 0.7944, + "num_input_tokens_seen": 128854955, + "step": 3960 + }, + { + "epoch": 0.35721693646570774, + "flos": 33061829073120.0, + "grad_norm": 2.3294186678130777, + "learning_rate": 2.978166216097326e-06, + "loss": 0.7219, + "num_input_tokens_seen": 128889610, + "step": 3961 + }, + { + "epoch": 0.35730711998917797, + "flos": 27597274501440.0, + "grad_norm": 1.9103110783086648, + "learning_rate": 2.9776566041372596e-06, + "loss": 0.7525, + "num_input_tokens_seen": 128920580, + "step": 3962 + }, + { + "epoch": 0.35739730351264826, + "flos": 22387744618560.0, + "grad_norm": 3.6945753592207233, + "learning_rate": 2.977146908758141e-06, + "loss": 0.8153, + "num_input_tokens_seen": 128949840, + "step": 3963 + }, + { + "epoch": 0.3574874870361185, + "flos": 26176459833600.0, + "grad_norm": 1.863969655694495, + "learning_rate": 2.9766371300034604e-06, + "loss": 0.7776, + "num_input_tokens_seen": 128978170, + "step": 3964 + }, + { + "epoch": 0.3575776705595888, + "flos": 20456731894080.0, + "grad_norm": 1.552458425508828, + "learning_rate": 2.9761272679167142e-06, + "loss": 0.8132, + "num_input_tokens_seen": 129005340, + "step": 3965 + }, + { + "epoch": 0.357667854083059, + "flos": 25447557677280.0, + "grad_norm": 1.6618894617571185, + "learning_rate": 2.9756173225414072e-06, + "loss": 0.611, + "num_input_tokens_seen": 129035075, + "step": 3966 + }, + { + "epoch": 0.3577580376065293, + "flos": 26905250480640.0, + "grad_norm": 2.3466295990219894, + "learning_rate": 2.975107293921051e-06, + "loss": 0.939, + "num_input_tokens_seen": 129063905, + "step": 3967 + }, + { + "epoch": 0.35784822112999953, + "flos": 24572332411200.0, + "grad_norm": 4.8244939192777965, + "learning_rate": 2.9745971820991643e-06, + "loss": 0.8029, + "num_input_tokens_seen": 129087995, + "step": 3968 + }, + { + "epoch": 0.3579384046534698, + "flos": 34516771314240.0, + "grad_norm": 2.3353300293486963, + "learning_rate": 2.9740869871192715e-06, + "loss": 0.8054, + "num_input_tokens_seen": 129120180, + "step": 3969 + }, + { + "epoch": 0.35802858817694005, + "flos": 18379470399360.0, + "grad_norm": 1.577174816560548, + "learning_rate": 2.9735767090249065e-06, + "loss": 0.827, + "num_input_tokens_seen": 129147140, + "step": 3970 + }, + { + "epoch": 0.35811877170041034, + "flos": 32072289594240.0, + "grad_norm": 1.9087913957365126, + "learning_rate": 2.973066347859608e-06, + "loss": 0.6715, + "num_input_tokens_seen": 129176960, + "step": 3971 + }, + { + "epoch": 0.3582089552238806, + "flos": 23480038514880.0, + "grad_norm": 1.8744123173656173, + "learning_rate": 2.972555903666923e-06, + "loss": 0.7757, + "num_input_tokens_seen": 129205110, + "step": 3972 + }, + { + "epoch": 0.35829913874735086, + "flos": 24209015010720.0, + "grad_norm": 2.988278696894334, + "learning_rate": 2.972045376490406e-06, + "loss": 0.6815, + "num_input_tokens_seen": 129234140, + "step": 3973 + }, + { + "epoch": 0.3583893222708211, + "flos": 19180567696800.0, + "grad_norm": 1.5599985317076595, + "learning_rate": 2.9715347663736177e-06, + "loss": 0.6907, + "num_input_tokens_seen": 129263290, + "step": 3974 + }, + { + "epoch": 0.3584795057942914, + "flos": 18523972190880.0, + "grad_norm": 1.710986968418349, + "learning_rate": 2.9710240733601266e-06, + "loss": 0.7598, + "num_input_tokens_seen": 129291815, + "step": 3975 + }, + { + "epoch": 0.3585696893177616, + "flos": 23552977051200.0, + "grad_norm": 2.2474669927394637, + "learning_rate": 2.970513297493507e-06, + "loss": 0.8439, + "num_input_tokens_seen": 129317595, + "step": 3976 + }, + { + "epoch": 0.3586598728412319, + "flos": 25336663082400.0, + "grad_norm": 1.531129014821441, + "learning_rate": 2.9700024388173416e-06, + "loss": 0.8118, + "num_input_tokens_seen": 129346955, + "step": 3977 + }, + { + "epoch": 0.35875005636470214, + "flos": 19764670703520.0, + "grad_norm": 1.849023357555355, + "learning_rate": 2.969491497375219e-06, + "loss": 0.8162, + "num_input_tokens_seen": 129375540, + "step": 3978 + }, + { + "epoch": 0.35884023988817243, + "flos": 19909953060000.0, + "grad_norm": 3.396134642044831, + "learning_rate": 2.9689804732107364e-06, + "loss": 0.7864, + "num_input_tokens_seen": 129404185, + "step": 3979 + }, + { + "epoch": 0.3589304234116427, + "flos": 14627484640800.0, + "grad_norm": 2.0135030296422247, + "learning_rate": 2.9684693663674968e-06, + "loss": 0.7217, + "num_input_tokens_seen": 129431225, + "step": 3980 + }, + { + "epoch": 0.35902060693511295, + "flos": 24932601891360.0, + "grad_norm": 1.6647023148519269, + "learning_rate": 2.9679581768891115e-06, + "loss": 0.8465, + "num_input_tokens_seen": 129460350, + "step": 3981 + }, + { + "epoch": 0.35911079045858324, + "flos": 22127107296000.0, + "grad_norm": 1.6054500383832033, + "learning_rate": 2.967446904819197e-06, + "loss": 0.6864, + "num_input_tokens_seen": 129490505, + "step": 3982 + }, + { + "epoch": 0.3592009739820535, + "flos": 20492011729920.0, + "grad_norm": 1.6013324107062366, + "learning_rate": 2.966935550201378e-06, + "loss": 0.7901, + "num_input_tokens_seen": 129518930, + "step": 3983 + }, + { + "epoch": 0.35929115750552376, + "flos": 19978617073920.0, + "grad_norm": 2.0425567503274187, + "learning_rate": 2.966424113079286e-06, + "loss": 0.7605, + "num_input_tokens_seen": 129547125, + "step": 3984 + }, + { + "epoch": 0.359381341028994, + "flos": 27305631865440.0, + "grad_norm": 1.8517012403312296, + "learning_rate": 2.9659125934965596e-06, + "loss": 0.8173, + "num_input_tokens_seen": 129577055, + "step": 3985 + }, + { + "epoch": 0.3594715245524643, + "flos": 23219252513280.0, + "grad_norm": 1.7199446947609969, + "learning_rate": 2.9654009914968457e-06, + "loss": 0.8238, + "num_input_tokens_seen": 129607280, + "step": 3986 + }, + { + "epoch": 0.3595617080759345, + "flos": 24427496091840.0, + "grad_norm": 2.216015810052819, + "learning_rate": 2.9648893071237956e-06, + "loss": 0.6523, + "num_input_tokens_seen": 129635545, + "step": 3987 + }, + { + "epoch": 0.3596518915994048, + "flos": 23035828249440.0, + "grad_norm": 1.6405118630772955, + "learning_rate": 2.964377540421069e-06, + "loss": 0.7427, + "num_input_tokens_seen": 129662895, + "step": 3988 + }, + { + "epoch": 0.35974207512287504, + "flos": 16735008053760.0, + "grad_norm": 2.3610550917418753, + "learning_rate": 2.963865691432334e-06, + "loss": 0.7896, + "num_input_tokens_seen": 129687700, + "step": 3989 + }, + { + "epoch": 0.3598322586463453, + "flos": 27446342341440.0, + "grad_norm": 1.6706288765156958, + "learning_rate": 2.963353760201263e-06, + "loss": 0.679, + "num_input_tokens_seen": 129716695, + "step": 3990 + }, + { + "epoch": 0.35992244216981556, + "flos": 15278132985120.0, + "grad_norm": 1.8748780528784488, + "learning_rate": 2.962841746771537e-06, + "loss": 0.7477, + "num_input_tokens_seen": 129742110, + "step": 3991 + }, + { + "epoch": 0.36001262569328585, + "flos": 27996540793440.0, + "grad_norm": 1.8255317096322716, + "learning_rate": 2.9623296511868445e-06, + "loss": 0.7698, + "num_input_tokens_seen": 129771220, + "step": 3992 + }, + { + "epoch": 0.3601028092167561, + "flos": 22748460135840.0, + "grad_norm": 1.7432540571131914, + "learning_rate": 2.96181747349088e-06, + "loss": 0.6926, + "num_input_tokens_seen": 129800880, + "step": 3993 + }, + { + "epoch": 0.36019299274022637, + "flos": 18307126579200.0, + "grad_norm": 2.2750125423546774, + "learning_rate": 2.961305213727345e-06, + "loss": 0.7845, + "num_input_tokens_seen": 129828460, + "step": 3994 + }, + { + "epoch": 0.3602831762636966, + "flos": 70471083688320.0, + "grad_norm": 0.6454200165523957, + "learning_rate": 2.960792871939949e-06, + "loss": 0.6027, + "num_input_tokens_seen": 129921115, + "step": 3995 + }, + { + "epoch": 0.3603733597871669, + "flos": 28726112005440.0, + "grad_norm": 2.4411544792055286, + "learning_rate": 2.9602804481724064e-06, + "loss": 0.8562, + "num_input_tokens_seen": 129950105, + "step": 3996 + }, + { + "epoch": 0.3604635433106371, + "flos": 22715038788000.0, + "grad_norm": 2.4569787533826934, + "learning_rate": 2.9597679424684427e-06, + "loss": 0.7609, + "num_input_tokens_seen": 129979710, + "step": 3997 + }, + { + "epoch": 0.3605537268341074, + "flos": 23625655399200.0, + "grad_norm": 2.2262990695315557, + "learning_rate": 2.9592553548717848e-06, + "loss": 0.7816, + "num_input_tokens_seen": 130007185, + "step": 3998 + }, + { + "epoch": 0.36064391035757765, + "flos": 13127895784320.0, + "grad_norm": 2.104857403779786, + "learning_rate": 2.958742685426171e-06, + "loss": 0.8485, + "num_input_tokens_seen": 130031660, + "step": 3999 + }, + { + "epoch": 0.36073409388104793, + "flos": 22350940822560.0, + "grad_norm": 1.666833495691459, + "learning_rate": 2.9582299341753446e-06, + "loss": 0.8351, + "num_input_tokens_seen": 130060220, + "step": 4000 + }, + { + "epoch": 0.36082427740451817, + "flos": 21840148049760.0, + "grad_norm": 4.5342134453500496, + "learning_rate": 2.957717101163057e-06, + "loss": 0.708, + "num_input_tokens_seen": 130084100, + "step": 4001 + }, + { + "epoch": 0.36091446092798846, + "flos": 22350383276160.0, + "grad_norm": 1.6313149917926673, + "learning_rate": 2.9572041864330655e-06, + "loss": 0.7671, + "num_input_tokens_seen": 130111450, + "step": 4002 + }, + { + "epoch": 0.36100464445145874, + "flos": 28690720660320.0, + "grad_norm": 2.1591136547879994, + "learning_rate": 2.9566911900291346e-06, + "loss": 0.7836, + "num_input_tokens_seen": 130144570, + "step": 4003 + }, + { + "epoch": 0.361094827974929, + "flos": 24609730923360.0, + "grad_norm": 2.5425875095247576, + "learning_rate": 2.9561781119950368e-06, + "loss": 0.6818, + "num_input_tokens_seen": 130173965, + "step": 4004 + }, + { + "epoch": 0.36118501149839927, + "flos": 19794337905600.0, + "grad_norm": 1.4823128279432969, + "learning_rate": 2.9556649523745493e-06, + "loss": 0.7111, + "num_input_tokens_seen": 130201285, + "step": 4005 + }, + { + "epoch": 0.3612751950218695, + "flos": 21731557980000.0, + "grad_norm": 1.609237665038743, + "learning_rate": 2.955151711211459e-06, + "loss": 0.8262, + "num_input_tokens_seen": 130228520, + "step": 4006 + }, + { + "epoch": 0.3613653785453398, + "flos": 23951202589920.0, + "grad_norm": 1.7148336006159646, + "learning_rate": 2.9546383885495583e-06, + "loss": 0.7344, + "num_input_tokens_seen": 130258790, + "step": 4007 + }, + { + "epoch": 0.36145556206881, + "flos": 34045978936800.0, + "grad_norm": 1.9312977732456178, + "learning_rate": 2.9541249844326464e-06, + "loss": 0.7189, + "num_input_tokens_seen": 130289450, + "step": 4008 + }, + { + "epoch": 0.3615457455922803, + "flos": 24099830224800.0, + "grad_norm": 1.6554510240744456, + "learning_rate": 2.9536114989045295e-06, + "loss": 0.747, + "num_input_tokens_seen": 130318360, + "step": 4009 + }, + { + "epoch": 0.36163592911575054, + "flos": 18926397912480.0, + "grad_norm": 1.5990451288015894, + "learning_rate": 2.9530979320090216e-06, + "loss": 0.8256, + "num_input_tokens_seen": 130346010, + "step": 4010 + }, + { + "epoch": 0.36172611263922083, + "flos": 23042890503840.0, + "grad_norm": 2.2047789297198865, + "learning_rate": 2.9525842837899422e-06, + "loss": 0.8429, + "num_input_tokens_seen": 130375545, + "step": 4011 + }, + { + "epoch": 0.36181629616269106, + "flos": 20638557858240.0, + "grad_norm": 2.1814741790033745, + "learning_rate": 2.95207055429112e-06, + "loss": 0.8468, + "num_input_tokens_seen": 130401910, + "step": 4012 + }, + { + "epoch": 0.36190647968616135, + "flos": 22930137420960.0, + "grad_norm": 1.6337075586504985, + "learning_rate": 2.951556743556388e-06, + "loss": 0.7607, + "num_input_tokens_seen": 130429600, + "step": 4013 + }, + { + "epoch": 0.3619966632096316, + "flos": 17760087556800.0, + "grad_norm": 2.6946165209047215, + "learning_rate": 2.951042851629588e-06, + "loss": 0.8147, + "num_input_tokens_seen": 130455635, + "step": 4014 + }, + { + "epoch": 0.3620868467331019, + "flos": 23622718988160.0, + "grad_norm": 1.763879030889296, + "learning_rate": 2.950528878554568e-06, + "loss": 0.7217, + "num_input_tokens_seen": 130483285, + "step": 4015 + }, + { + "epoch": 0.3621770302565721, + "flos": 25812250358880.0, + "grad_norm": 1.4416320127758018, + "learning_rate": 2.950014824375183e-06, + "loss": 0.8044, + "num_input_tokens_seen": 130512845, + "step": 4016 + }, + { + "epoch": 0.3622672137800424, + "flos": 67389074549280.0, + "grad_norm": 0.632575284978129, + "learning_rate": 2.949500689135295e-06, + "loss": 0.5272, + "num_input_tokens_seen": 130610710, + "step": 4017 + }, + { + "epoch": 0.36235739730351263, + "flos": 23990013552960.0, + "grad_norm": 1.5036679562125022, + "learning_rate": 2.9489864728787722e-06, + "loss": 0.8469, + "num_input_tokens_seen": 130639545, + "step": 4018 + }, + { + "epoch": 0.3624475808269829, + "flos": 22962369336480.0, + "grad_norm": 1.6213668409375608, + "learning_rate": 2.9484721756494915e-06, + "loss": 0.8266, + "num_input_tokens_seen": 130665255, + "step": 4019 + }, + { + "epoch": 0.36253776435045315, + "flos": 24937879997280.0, + "grad_norm": 2.0778642074891236, + "learning_rate": 2.9479577974913343e-06, + "loss": 0.7718, + "num_input_tokens_seen": 130696585, + "step": 4020 + }, + { + "epoch": 0.36262794787392344, + "flos": 23841608936640.0, + "grad_norm": 1.6552592499161525, + "learning_rate": 2.9474433384481908e-06, + "loss": 0.7632, + "num_input_tokens_seen": 130727400, + "step": 4021 + }, + { + "epoch": 0.3627181313973937, + "flos": 27232804838400.0, + "grad_norm": 1.7131225011418352, + "learning_rate": 2.9469287985639577e-06, + "loss": 0.7637, + "num_input_tokens_seen": 130755235, + "step": 4022 + }, + { + "epoch": 0.36280831492086396, + "flos": 23771420962560.0, + "grad_norm": 1.7081414634449965, + "learning_rate": 2.9464141778825384e-06, + "loss": 0.8136, + "num_input_tokens_seen": 130783805, + "step": 4023 + }, + { + "epoch": 0.3628984984443342, + "flos": 27117895909440.0, + "grad_norm": 2.004054251814024, + "learning_rate": 2.9458994764478427e-06, + "loss": 0.8082, + "num_input_tokens_seen": 130813875, + "step": 4024 + }, + { + "epoch": 0.3629886819678045, + "flos": 20595695391360.0, + "grad_norm": 1.8127331419331736, + "learning_rate": 2.9453846943037883e-06, + "loss": 0.7987, + "num_input_tokens_seen": 130839350, + "step": 4025 + }, + { + "epoch": 0.36307886549127477, + "flos": 23007053121600.0, + "grad_norm": 2.6692409805241217, + "learning_rate": 2.9448698314942987e-06, + "loss": 0.7135, + "num_input_tokens_seen": 130870615, + "step": 4026 + }, + { + "epoch": 0.363169049014745, + "flos": 71164631669280.0, + "grad_norm": 0.7570621275534384, + "learning_rate": 2.944354888063305e-06, + "loss": 0.6231, + "num_input_tokens_seen": 130967870, + "step": 4027 + }, + { + "epoch": 0.3632592325382153, + "flos": 21695051542080.0, + "grad_norm": 1.957150453755629, + "learning_rate": 2.9438398640547453e-06, + "loss": 0.7439, + "num_input_tokens_seen": 130995690, + "step": 4028 + }, + { + "epoch": 0.3633494160616855, + "flos": 23007127461120.0, + "grad_norm": 1.6993266455058664, + "learning_rate": 2.943324759512564e-06, + "loss": 0.7047, + "num_input_tokens_seen": 131024365, + "step": 4029 + }, + { + "epoch": 0.3634395995851558, + "flos": 24207788408640.0, + "grad_norm": 1.7926767813790994, + "learning_rate": 2.9428095744807134e-06, + "loss": 0.7959, + "num_input_tokens_seen": 131055870, + "step": 4030 + }, + { + "epoch": 0.36352978310862605, + "flos": 35683825065120.0, + "grad_norm": 4.206564478008342, + "learning_rate": 2.942294309003151e-06, + "loss": 0.72, + "num_input_tokens_seen": 131088590, + "step": 4031 + }, + { + "epoch": 0.36361996663209634, + "flos": 28070743101600.0, + "grad_norm": 1.8001128505723, + "learning_rate": 2.941778963123843e-06, + "loss": 0.785, + "num_input_tokens_seen": 131116780, + "step": 4032 + }, + { + "epoch": 0.36371015015556657, + "flos": 31240521511200.0, + "grad_norm": 1.6825761662739402, + "learning_rate": 2.94126353688676e-06, + "loss": 0.767, + "num_input_tokens_seen": 131149320, + "step": 4033 + }, + { + "epoch": 0.36380033367903686, + "flos": 24537126914880.0, + "grad_norm": 2.578568668914997, + "learning_rate": 2.9407480303358825e-06, + "loss": 0.7173, + "num_input_tokens_seen": 131176780, + "step": 4034 + }, + { + "epoch": 0.3638905172025071, + "flos": 19757980146720.0, + "grad_norm": 1.6617387477171834, + "learning_rate": 2.940232443515195e-06, + "loss": 0.7735, + "num_input_tokens_seen": 131207505, + "step": 4035 + }, + { + "epoch": 0.3639807007259774, + "flos": 22532172070560.0, + "grad_norm": 3.2098820569008244, + "learning_rate": 2.9397167764686916e-06, + "loss": 0.789, + "num_input_tokens_seen": 131238310, + "step": 4036 + }, + { + "epoch": 0.3640708842494476, + "flos": 24828435023040.0, + "grad_norm": 2.042118749066892, + "learning_rate": 2.9392010292403714e-06, + "loss": 0.7427, + "num_input_tokens_seen": 131266710, + "step": 4037 + }, + { + "epoch": 0.3641610677729179, + "flos": 21841114463520.0, + "grad_norm": 2.310644384730538, + "learning_rate": 2.9386852018742404e-06, + "loss": 0.7607, + "num_input_tokens_seen": 131293270, + "step": 4038 + }, + { + "epoch": 0.36425125129638813, + "flos": 23734765845600.0, + "grad_norm": 15.114255004653822, + "learning_rate": 2.938169294414312e-06, + "loss": 0.74, + "num_input_tokens_seen": 131322815, + "step": 4039 + }, + { + "epoch": 0.3643414348198584, + "flos": 22090600858080.0, + "grad_norm": 2.3701832765623525, + "learning_rate": 2.9376533069046067e-06, + "loss": 0.7968, + "num_input_tokens_seen": 131351850, + "step": 4040 + }, + { + "epoch": 0.36443161834332866, + "flos": 66606419186400.0, + "grad_norm": 0.7153369057134475, + "learning_rate": 2.9371372393891514e-06, + "loss": 0.6344, + "num_input_tokens_seen": 131440865, + "step": 4041 + }, + { + "epoch": 0.36452180186679894, + "flos": 25338298551840.0, + "grad_norm": 1.2664055948414599, + "learning_rate": 2.936621091911979e-06, + "loss": 0.8352, + "num_input_tokens_seen": 131472395, + "step": 4042 + }, + { + "epoch": 0.3646119853902692, + "flos": 24135964965120.0, + "grad_norm": 2.560489610254252, + "learning_rate": 2.936104864517131e-06, + "loss": 0.6992, + "num_input_tokens_seen": 131500155, + "step": 4043 + }, + { + "epoch": 0.36470216891373947, + "flos": 38848808575680.0, + "grad_norm": 1.9391616341012334, + "learning_rate": 2.9355885572486535e-06, + "loss": 0.7233, + "num_input_tokens_seen": 131528150, + "step": 4044 + }, + { + "epoch": 0.3647923524372097, + "flos": 22129820688480.0, + "grad_norm": 1.6650527371826658, + "learning_rate": 2.9350721701506026e-06, + "loss": 0.798, + "num_input_tokens_seen": 131558840, + "step": 4045 + }, + { + "epoch": 0.36488253596068, + "flos": 29564607815040.0, + "grad_norm": 1.8201456934913969, + "learning_rate": 2.9345557032670375e-06, + "loss": 0.7577, + "num_input_tokens_seen": 131588920, + "step": 4046 + }, + { + "epoch": 0.3649727194841502, + "flos": 19035471189120.0, + "grad_norm": 1.9876779416638959, + "learning_rate": 2.934039156642027e-06, + "loss": 0.8101, + "num_input_tokens_seen": 131614840, + "step": 4047 + }, + { + "epoch": 0.3650629030076205, + "flos": 37317768368640.0, + "grad_norm": 1.87121103831, + "learning_rate": 2.9335225303196454e-06, + "loss": 0.6283, + "num_input_tokens_seen": 131644590, + "step": 4048 + }, + { + "epoch": 0.36515308653109074, + "flos": 22969989137280.0, + "grad_norm": 1.7926558344097483, + "learning_rate": 2.933005824343974e-06, + "loss": 0.7347, + "num_input_tokens_seen": 131673005, + "step": 4049 + }, + { + "epoch": 0.36524327005456103, + "flos": 18450067240800.0, + "grad_norm": 2.404849251604581, + "learning_rate": 2.932489038759101e-06, + "loss": 0.7394, + "num_input_tokens_seen": 131700565, + "step": 4050 + }, + { + "epoch": 0.3653334535780313, + "flos": 53060820909120.0, + "grad_norm": 1.4647706687499, + "learning_rate": 2.9319721736091215e-06, + "loss": 0.6574, + "num_input_tokens_seen": 131734240, + "step": 4051 + }, + { + "epoch": 0.36542363710150155, + "flos": 25005986464800.0, + "grad_norm": 1.5211273675543262, + "learning_rate": 2.9314552289381377e-06, + "loss": 0.6729, + "num_input_tokens_seen": 131765620, + "step": 4052 + }, + { + "epoch": 0.36551382062497184, + "flos": 18670852847040.0, + "grad_norm": 1.845525371731605, + "learning_rate": 2.9309382047902574e-06, + "loss": 0.7363, + "num_input_tokens_seen": 131790330, + "step": 4053 + }, + { + "epoch": 0.3656040041484421, + "flos": 24100722299040.0, + "grad_norm": 1.5423581230909058, + "learning_rate": 2.9304211012095963e-06, + "loss": 0.8106, + "num_input_tokens_seen": 131821195, + "step": 4054 + }, + { + "epoch": 0.36569418767191236, + "flos": 18483191230560.0, + "grad_norm": 1.823423878178387, + "learning_rate": 2.929903918240277e-06, + "loss": 0.823, + "num_input_tokens_seen": 131847825, + "step": 4055 + }, + { + "epoch": 0.3657843711953826, + "flos": 66725453958720.0, + "grad_norm": 0.6319203386727867, + "learning_rate": 2.9293866559264273e-06, + "loss": 0.5778, + "num_input_tokens_seen": 131938830, + "step": 4056 + }, + { + "epoch": 0.3658745547188529, + "flos": 21986508329280.0, + "grad_norm": 2.241713736322382, + "learning_rate": 2.928869314312184e-06, + "loss": 0.7652, + "num_input_tokens_seen": 131967710, + "step": 4057 + }, + { + "epoch": 0.3659647382423231, + "flos": 21695237390880.0, + "grad_norm": 1.7027788422902554, + "learning_rate": 2.9283518934416892e-06, + "loss": 0.7639, + "num_input_tokens_seen": 131994970, + "step": 4058 + }, + { + "epoch": 0.3660549217657934, + "flos": 20711050357440.0, + "grad_norm": 1.541282872377727, + "learning_rate": 2.927834393359092e-06, + "loss": 0.774, + "num_input_tokens_seen": 132022160, + "step": 4059 + }, + { + "epoch": 0.36614510528926364, + "flos": 20674543919520.0, + "grad_norm": 2.059091120648512, + "learning_rate": 2.927316814108548e-06, + "loss": 0.8348, + "num_input_tokens_seen": 132050115, + "step": 4060 + }, + { + "epoch": 0.36623528881273393, + "flos": 24792077264160.0, + "grad_norm": 2.4811346647039842, + "learning_rate": 2.92679915573422e-06, + "loss": 0.6703, + "num_input_tokens_seen": 132080380, + "step": 4061 + }, + { + "epoch": 0.36632547233620416, + "flos": 25445364661440.0, + "grad_norm": 2.1914252894045347, + "learning_rate": 2.926281418280278e-06, + "loss": 0.7055, + "num_input_tokens_seen": 132110935, + "step": 4062 + }, + { + "epoch": 0.36641565585967445, + "flos": 24391435691040.0, + "grad_norm": 2.554858933975782, + "learning_rate": 2.925763601790899e-06, + "loss": 0.7655, + "num_input_tokens_seen": 132139755, + "step": 4063 + }, + { + "epoch": 0.3665058393831447, + "flos": 21841002954240.0, + "grad_norm": 1.9903241587444278, + "learning_rate": 2.9252457063102635e-06, + "loss": 0.7098, + "num_input_tokens_seen": 132169130, + "step": 4064 + }, + { + "epoch": 0.36659602290661497, + "flos": 53608380308160.0, + "grad_norm": 1.9879461279579393, + "learning_rate": 2.9247277318825626e-06, + "loss": 0.7264, + "num_input_tokens_seen": 132204265, + "step": 4065 + }, + { + "epoch": 0.3666862064300852, + "flos": 22820803956000.0, + "grad_norm": 2.0169806837557465, + "learning_rate": 2.924209678551993e-06, + "loss": 0.8558, + "num_input_tokens_seen": 132232430, + "step": 4066 + }, + { + "epoch": 0.3667763899535555, + "flos": 20602162929600.0, + "grad_norm": 1.7259887156156308, + "learning_rate": 2.923691546362757e-06, + "loss": 0.7763, + "num_input_tokens_seen": 132258860, + "step": 4067 + }, + { + "epoch": 0.3668665734770257, + "flos": 22132199553120.0, + "grad_norm": 1.8228472305879095, + "learning_rate": 2.9231733353590663e-06, + "loss": 0.7274, + "num_input_tokens_seen": 132286030, + "step": 4068 + }, + { + "epoch": 0.366956757000496, + "flos": 23293863688800.0, + "grad_norm": 9.818585300824672, + "learning_rate": 2.922655045585136e-06, + "loss": 0.8348, + "num_input_tokens_seen": 132312705, + "step": 4069 + }, + { + "epoch": 0.36704694052396625, + "flos": 71095521618240.0, + "grad_norm": 0.6915737962037217, + "learning_rate": 2.92213667708519e-06, + "loss": 0.6199, + "num_input_tokens_seen": 132407125, + "step": 4070 + }, + { + "epoch": 0.36713712404743654, + "flos": 24387235508160.0, + "grad_norm": 2.5494334470385587, + "learning_rate": 2.921618229903457e-06, + "loss": 0.7742, + "num_input_tokens_seen": 132434325, + "step": 4071 + }, + { + "epoch": 0.36722730757090677, + "flos": 34664767063200.0, + "grad_norm": 2.001601174283922, + "learning_rate": 2.9210997040841752e-06, + "loss": 0.6849, + "num_input_tokens_seen": 132467885, + "step": 4072 + }, + { + "epoch": 0.36731749109437706, + "flos": 17465285491200.0, + "grad_norm": 1.9145038329489819, + "learning_rate": 2.9205810996715885e-06, + "loss": 0.7301, + "num_input_tokens_seen": 132494665, + "step": 4073 + }, + { + "epoch": 0.36740767461784735, + "flos": 23262523847520.0, + "grad_norm": 1.7431969562066278, + "learning_rate": 2.9200624167099456e-06, + "loss": 0.7274, + "num_input_tokens_seen": 132522040, + "step": 4074 + }, + { + "epoch": 0.3674978581413176, + "flos": 24828249174240.0, + "grad_norm": 1.4865700002356383, + "learning_rate": 2.919543655243505e-06, + "loss": 0.6805, + "num_input_tokens_seen": 132552585, + "step": 4075 + }, + { + "epoch": 0.36758804166478787, + "flos": 22531465845120.0, + "grad_norm": 2.4334633487647794, + "learning_rate": 2.919024815316529e-06, + "loss": 0.7929, + "num_input_tokens_seen": 132580640, + "step": 4076 + }, + { + "epoch": 0.3676782251882581, + "flos": 22350643464480.0, + "grad_norm": 1.8286335555660111, + "learning_rate": 2.9185058969732877e-06, + "loss": 0.8836, + "num_input_tokens_seen": 132607120, + "step": 4077 + }, + { + "epoch": 0.3677684087117284, + "flos": 24828174834720.0, + "grad_norm": 1.7232132268601672, + "learning_rate": 2.917986900258059e-06, + "loss": 0.7989, + "num_input_tokens_seen": 132635590, + "step": 4078 + }, + { + "epoch": 0.3678585922351986, + "flos": 21585941095680.0, + "grad_norm": 7.411084106462201, + "learning_rate": 2.917467825215126e-06, + "loss": 0.7566, + "num_input_tokens_seen": 132666590, + "step": 4079 + }, + { + "epoch": 0.3679487757586689, + "flos": 28253684158560.0, + "grad_norm": 1.5410596408528936, + "learning_rate": 2.9169486718887803e-06, + "loss": 0.7118, + "num_input_tokens_seen": 132698280, + "step": 4080 + }, + { + "epoch": 0.36803895928213914, + "flos": 24972193419360.0, + "grad_norm": 4.284580244160721, + "learning_rate": 2.9164294403233173e-06, + "loss": 0.814, + "num_input_tokens_seen": 132729535, + "step": 4081 + }, + { + "epoch": 0.36812914280560943, + "flos": 15246086918400.0, + "grad_norm": 2.396986730724237, + "learning_rate": 2.915910130563041e-06, + "loss": 0.7007, + "num_input_tokens_seen": 132756925, + "step": 4082 + }, + { + "epoch": 0.36821932632907967, + "flos": 22095953303520.0, + "grad_norm": 2.0245964825628366, + "learning_rate": 2.915390742652262e-06, + "loss": 0.7106, + "num_input_tokens_seen": 132785165, + "step": 4083 + }, + { + "epoch": 0.36830950985254995, + "flos": 18306606202560.0, + "grad_norm": 2.2736344399637676, + "learning_rate": 2.914871276635298e-06, + "loss": 0.6949, + "num_input_tokens_seen": 132811890, + "step": 4084 + }, + { + "epoch": 0.3683996933760202, + "flos": 28362348567840.0, + "grad_norm": 1.9959331529963267, + "learning_rate": 2.914351732556472e-06, + "loss": 0.7424, + "num_input_tokens_seen": 132839970, + "step": 4085 + }, + { + "epoch": 0.3684898768994905, + "flos": 20165237937120.0, + "grad_norm": 1.5999455174468342, + "learning_rate": 2.9138321104601144e-06, + "loss": 0.6844, + "num_input_tokens_seen": 132868160, + "step": 4086 + }, + { + "epoch": 0.3685800604229607, + "flos": 29746136421120.0, + "grad_norm": 2.3272691215953203, + "learning_rate": 2.9133124103905623e-06, + "loss": 0.6963, + "num_input_tokens_seen": 132900785, + "step": 4087 + }, + { + "epoch": 0.368670243946431, + "flos": 23042816164320.0, + "grad_norm": 2.3710987776022687, + "learning_rate": 2.9127926323921596e-06, + "loss": 0.7531, + "num_input_tokens_seen": 132928485, + "step": 4088 + }, + { + "epoch": 0.36876042746990123, + "flos": 24865275988800.0, + "grad_norm": 1.846788392077529, + "learning_rate": 2.912272776509256e-06, + "loss": 0.7813, + "num_input_tokens_seen": 132957840, + "step": 4089 + }, + { + "epoch": 0.3688506109933715, + "flos": 27480692933280.0, + "grad_norm": 1.857709186527093, + "learning_rate": 2.911752842786209e-06, + "loss": 0.7292, + "num_input_tokens_seen": 132986320, + "step": 4090 + }, + { + "epoch": 0.36894079451684175, + "flos": 21221322753600.0, + "grad_norm": 1.9710015032301405, + "learning_rate": 2.911232831267383e-06, + "loss": 0.7011, + "num_input_tokens_seen": 133015580, + "step": 4091 + }, + { + "epoch": 0.36903097804031204, + "flos": 27957060774720.0, + "grad_norm": 1.8036558827505698, + "learning_rate": 2.910712741997146e-06, + "loss": 0.7823, + "num_input_tokens_seen": 133045415, + "step": 4092 + }, + { + "epoch": 0.3691211615637823, + "flos": 27560470705440.0, + "grad_norm": 1.6358500430859655, + "learning_rate": 2.910192575019877e-06, + "loss": 0.7487, + "num_input_tokens_seen": 133074295, + "step": 4093 + }, + { + "epoch": 0.36921134508725256, + "flos": 23844433838400.0, + "grad_norm": 1.8748956651269142, + "learning_rate": 2.9096723303799583e-06, + "loss": 0.7938, + "num_input_tokens_seen": 133105310, + "step": 4094 + }, + { + "epoch": 0.3693015286107228, + "flos": 26171033048640.0, + "grad_norm": 2.385931956066793, + "learning_rate": 2.9091520081217805e-06, + "loss": 0.7699, + "num_input_tokens_seen": 133135345, + "step": 4095 + }, + { + "epoch": 0.3693917121341931, + "flos": 22678904047680.0, + "grad_norm": 1.6186984462703007, + "learning_rate": 2.908631608289741e-06, + "loss": 0.7111, + "num_input_tokens_seen": 133165010, + "step": 4096 + }, + { + "epoch": 0.3694818956576633, + "flos": 31239555097440.0, + "grad_norm": 1.6848018409733436, + "learning_rate": 2.9081111309282423e-06, + "loss": 0.7517, + "num_input_tokens_seen": 133194210, + "step": 4097 + }, + { + "epoch": 0.3695720791811336, + "flos": 68024371782720.0, + "grad_norm": 0.6420544980840045, + "learning_rate": 2.9075905760816942e-06, + "loss": 0.631, + "num_input_tokens_seen": 133290575, + "step": 4098 + }, + { + "epoch": 0.3696622627046039, + "flos": 19690505565120.0, + "grad_norm": 2.0232329394065354, + "learning_rate": 2.907069943794514e-06, + "loss": 0.7995, + "num_input_tokens_seen": 133317080, + "step": 4099 + }, + { + "epoch": 0.3697524462280741, + "flos": 39725743650720.0, + "grad_norm": 1.585783701988833, + "learning_rate": 2.906549234111125e-06, + "loss": 0.7293, + "num_input_tokens_seen": 133350605, + "step": 4100 + }, + { + "epoch": 0.3698426297515444, + "flos": 20964179388480.0, + "grad_norm": 1.483580929026314, + "learning_rate": 2.906028447075956e-06, + "loss": 0.8178, + "num_input_tokens_seen": 133379210, + "step": 4101 + }, + { + "epoch": 0.36993281327501465, + "flos": 25915934020320.0, + "grad_norm": 1.9684450428029847, + "learning_rate": 2.905507582733445e-06, + "loss": 0.845, + "num_input_tokens_seen": 133408165, + "step": 4102 + }, + { + "epoch": 0.37002299679848494, + "flos": 21476496121440.0, + "grad_norm": 1.6835043190837549, + "learning_rate": 2.904986641128033e-06, + "loss": 0.8782, + "num_input_tokens_seen": 133432500, + "step": 4103 + }, + { + "epoch": 0.37011318032195517, + "flos": 27268679390400.0, + "grad_norm": 1.424469979748807, + "learning_rate": 2.9044656223041716e-06, + "loss": 0.7512, + "num_input_tokens_seen": 133464040, + "step": 4104 + }, + { + "epoch": 0.37020336384542546, + "flos": 23516136085440.0, + "grad_norm": 1.5686702966390023, + "learning_rate": 2.9039445263063157e-06, + "loss": 0.7204, + "num_input_tokens_seen": 133492220, + "step": 4105 + }, + { + "epoch": 0.3702935473688957, + "flos": 25520161685760.0, + "grad_norm": 2.678470346269211, + "learning_rate": 2.903423353178929e-06, + "loss": 0.6578, + "num_input_tokens_seen": 133520190, + "step": 4106 + }, + { + "epoch": 0.370383730892366, + "flos": 28105316712000.0, + "grad_norm": 1.841953307352131, + "learning_rate": 2.9029021029664802e-06, + "loss": 0.7581, + "num_input_tokens_seen": 133551510, + "step": 4107 + }, + { + "epoch": 0.3704739144158362, + "flos": 24172880270400.0, + "grad_norm": 1.4620256094718262, + "learning_rate": 2.9023807757134455e-06, + "loss": 0.8154, + "num_input_tokens_seen": 133582195, + "step": 4108 + }, + { + "epoch": 0.3705640979393065, + "flos": 20015012002560.0, + "grad_norm": 1.6492622084247432, + "learning_rate": 2.901859371464307e-06, + "loss": 0.8277, + "num_input_tokens_seen": 133609775, + "step": 4109 + }, + { + "epoch": 0.37065428146277674, + "flos": 23225013826080.0, + "grad_norm": 1.76529454202376, + "learning_rate": 2.9013378902635535e-06, + "loss": 0.7997, + "num_input_tokens_seen": 133638265, + "step": 4110 + }, + { + "epoch": 0.370744464986247, + "flos": 22569087375840.0, + "grad_norm": 1.8675059282446222, + "learning_rate": 2.9008163321556823e-06, + "loss": 0.731, + "num_input_tokens_seen": 133667005, + "step": 4111 + }, + { + "epoch": 0.37083464850971726, + "flos": 21324300189600.0, + "grad_norm": 4.1049541873210815, + "learning_rate": 2.900294697185194e-06, + "loss": 0.7767, + "num_input_tokens_seen": 133691095, + "step": 4112 + }, + { + "epoch": 0.37092483203318755, + "flos": 27014695454880.0, + "grad_norm": 1.6865046355983655, + "learning_rate": 2.899772985396599e-06, + "loss": 0.7806, + "num_input_tokens_seen": 133722205, + "step": 4113 + }, + { + "epoch": 0.3710150155566578, + "flos": 26833055339520.0, + "grad_norm": 1.6484484920767992, + "learning_rate": 2.8992511968344104e-06, + "loss": 0.7512, + "num_input_tokens_seen": 133750050, + "step": 4114 + }, + { + "epoch": 0.37110519908012807, + "flos": 24712113643200.0, + "grad_norm": 1.702188987051964, + "learning_rate": 2.8987293315431523e-06, + "loss": 0.7049, + "num_input_tokens_seen": 133778895, + "step": 4115 + }, + { + "epoch": 0.3711953826035983, + "flos": 27851741643840.0, + "grad_norm": 3.287281705041535, + "learning_rate": 2.898207389567351e-06, + "loss": 0.693, + "num_input_tokens_seen": 133809640, + "step": 4116 + }, + { + "epoch": 0.3712855661270686, + "flos": 22239079813920.0, + "grad_norm": 1.8462916813344248, + "learning_rate": 2.897685370951543e-06, + "loss": 0.7529, + "num_input_tokens_seen": 133838200, + "step": 4117 + }, + { + "epoch": 0.3713757496505388, + "flos": 15829781057760.0, + "grad_norm": 3.2716674389079565, + "learning_rate": 2.89716327574027e-06, + "loss": 0.6522, + "num_input_tokens_seen": 133864560, + "step": 4118 + }, + { + "epoch": 0.3714659331740091, + "flos": 30539688257280.0, + "grad_norm": 2.023780116233556, + "learning_rate": 2.8966411039780787e-06, + "loss": 0.8481, + "num_input_tokens_seen": 133891405, + "step": 4119 + }, + { + "epoch": 0.37155611669747934, + "flos": 23618778993600.0, + "grad_norm": 1.8043242910597723, + "learning_rate": 2.8961188557095248e-06, + "loss": 0.7682, + "num_input_tokens_seen": 133918560, + "step": 4120 + }, + { + "epoch": 0.37164630022094963, + "flos": 17431975652640.0, + "grad_norm": 1.8623386774026833, + "learning_rate": 2.895596530979168e-06, + "loss": 0.705, + "num_input_tokens_seen": 133945555, + "step": 4121 + }, + { + "epoch": 0.3717364837444199, + "flos": 21398985704640.0, + "grad_norm": 1.9772033858621985, + "learning_rate": 2.895074129831578e-06, + "loss": 0.7762, + "num_input_tokens_seen": 133973450, + "step": 4122 + }, + { + "epoch": 0.37182666726789015, + "flos": 32879296883520.0, + "grad_norm": 1.7044468221160163, + "learning_rate": 2.8945516523113275e-06, + "loss": 0.678, + "num_input_tokens_seen": 134002515, + "step": 4123 + }, + { + "epoch": 0.37191685079136044, + "flos": 69369757540320.0, + "grad_norm": 0.6796944058823804, + "learning_rate": 2.894029098462998e-06, + "loss": 0.5509, + "num_input_tokens_seen": 134088330, + "step": 4124 + }, + { + "epoch": 0.3720070343148307, + "flos": 22715038788000.0, + "grad_norm": 1.7630842714902952, + "learning_rate": 2.8935064683311756e-06, + "loss": 0.7839, + "num_input_tokens_seen": 134116410, + "step": 4125 + }, + { + "epoch": 0.37209721783830096, + "flos": 65160434857440.0, + "grad_norm": 0.677547088162044, + "learning_rate": 2.8929837619604544e-06, + "loss": 0.6552, + "num_input_tokens_seen": 134207430, + "step": 4126 + }, + { + "epoch": 0.3721874013617712, + "flos": 24346603226880.0, + "grad_norm": 1.9047382801725254, + "learning_rate": 2.8924609793954346e-06, + "loss": 0.8052, + "num_input_tokens_seen": 134235290, + "step": 4127 + }, + { + "epoch": 0.3722775848852415, + "flos": 22675224241440.0, + "grad_norm": 1.8415174383518023, + "learning_rate": 2.891938120680724e-06, + "loss": 0.7149, + "num_input_tokens_seen": 134264065, + "step": 4128 + }, + { + "epoch": 0.3723677684087117, + "flos": 22930546288320.0, + "grad_norm": 1.5503302622236828, + "learning_rate": 2.8914151858609343e-06, + "loss": 0.7517, + "num_input_tokens_seen": 134294705, + "step": 4129 + }, + { + "epoch": 0.372457951932182, + "flos": 28762841461920.0, + "grad_norm": 1.8944885878331206, + "learning_rate": 2.8908921749806858e-06, + "loss": 0.7814, + "num_input_tokens_seen": 134324600, + "step": 4130 + }, + { + "epoch": 0.37254813545565224, + "flos": 21179724058560.0, + "grad_norm": 1.7606307431886687, + "learning_rate": 2.890369088084605e-06, + "loss": 0.7748, + "num_input_tokens_seen": 134351565, + "step": 4131 + }, + { + "epoch": 0.37263831897912253, + "flos": 27743263083360.0, + "grad_norm": 1.6234395531062196, + "learning_rate": 2.889845925217323e-06, + "loss": 0.7843, + "num_input_tokens_seen": 134380570, + "step": 4132 + }, + { + "epoch": 0.37272850250259276, + "flos": 28252420386720.0, + "grad_norm": 2.0494789857068563, + "learning_rate": 2.8893226864234813e-06, + "loss": 0.777, + "num_input_tokens_seen": 134411955, + "step": 4133 + }, + { + "epoch": 0.37281868602606305, + "flos": 21112286646720.0, + "grad_norm": 1.9235551514578606, + "learning_rate": 2.8887993717477236e-06, + "loss": 0.7719, + "num_input_tokens_seen": 134439255, + "step": 4134 + }, + { + "epoch": 0.3729088695495333, + "flos": 14117212244640.0, + "grad_norm": 1.8272157021582716, + "learning_rate": 2.8882759812347035e-06, + "loss": 0.651, + "num_input_tokens_seen": 134466390, + "step": 4135 + }, + { + "epoch": 0.3729990530730036, + "flos": 22569421903680.0, + "grad_norm": 1.7392516438601304, + "learning_rate": 2.887752514929078e-06, + "loss": 0.7651, + "num_input_tokens_seen": 134497080, + "step": 4136 + }, + { + "epoch": 0.3730892365964738, + "flos": 38016891813600.0, + "grad_norm": 2.002088290978154, + "learning_rate": 2.887228972875513e-06, + "loss": 0.7159, + "num_input_tokens_seen": 134526215, + "step": 4137 + }, + { + "epoch": 0.3731794201199441, + "flos": 23223341186880.0, + "grad_norm": 2.0658855972152113, + "learning_rate": 2.88670535511868e-06, + "loss": 0.7633, + "num_input_tokens_seen": 134551065, + "step": 4138 + }, + { + "epoch": 0.3732696036434143, + "flos": 22091901799680.0, + "grad_norm": 2.7078359489504202, + "learning_rate": 2.886181661703257e-06, + "loss": 0.7079, + "num_input_tokens_seen": 134578355, + "step": 4139 + }, + { + "epoch": 0.3733597871668846, + "flos": 25484287133760.0, + "grad_norm": 1.8125607052864494, + "learning_rate": 2.8856578926739285e-06, + "loss": 0.7259, + "num_input_tokens_seen": 134608015, + "step": 4140 + }, + { + "epoch": 0.37344997069035485, + "flos": 27485748020640.0, + "grad_norm": 1.9151740136030384, + "learning_rate": 2.8851340480753846e-06, + "loss": 0.719, + "num_input_tokens_seen": 134639310, + "step": 4141 + }, + { + "epoch": 0.37354015421382514, + "flos": 29054075230560.0, + "grad_norm": 2.664784442708009, + "learning_rate": 2.8846101279523232e-06, + "loss": 0.7635, + "num_input_tokens_seen": 134668435, + "step": 4142 + }, + { + "epoch": 0.37363033773729537, + "flos": 26429328676320.0, + "grad_norm": 8.31185428721501, + "learning_rate": 2.8840861323494487e-06, + "loss": 0.7378, + "num_input_tokens_seen": 134700100, + "step": 4143 + }, + { + "epoch": 0.37372052126076566, + "flos": 22310866087680.0, + "grad_norm": 1.7277081861242531, + "learning_rate": 2.88356206131147e-06, + "loss": 0.8191, + "num_input_tokens_seen": 134727695, + "step": 4144 + }, + { + "epoch": 0.37381070478423595, + "flos": 27342770189280.0, + "grad_norm": 1.8925330882033364, + "learning_rate": 2.883037914883104e-06, + "loss": 0.7491, + "num_input_tokens_seen": 134757175, + "step": 4145 + }, + { + "epoch": 0.3739008883077062, + "flos": 19691806506720.0, + "grad_norm": 1.7033895584551422, + "learning_rate": 2.882513693109075e-06, + "loss": 0.7296, + "num_input_tokens_seen": 134784945, + "step": 4146 + }, + { + "epoch": 0.37399107183117647, + "flos": 25663808572800.0, + "grad_norm": 1.3804497657597443, + "learning_rate": 2.8819893960341106e-06, + "loss": 0.7383, + "num_input_tokens_seen": 134815260, + "step": 4147 + }, + { + "epoch": 0.3740812553546467, + "flos": 19756530526080.0, + "grad_norm": 2.442525427070012, + "learning_rate": 2.881465023702948e-06, + "loss": 0.8212, + "num_input_tokens_seen": 134838675, + "step": 4148 + }, + { + "epoch": 0.374171438878117, + "flos": 24643672647840.0, + "grad_norm": 2.4354777628400908, + "learning_rate": 2.8809405761603294e-06, + "loss": 0.8441, + "num_input_tokens_seen": 134868015, + "step": 4149 + }, + { + "epoch": 0.3742616224015872, + "flos": 16448197486560.0, + "grad_norm": 2.0176277560751785, + "learning_rate": 2.880416053451003e-06, + "loss": 0.6816, + "num_input_tokens_seen": 134893110, + "step": 4150 + }, + { + "epoch": 0.3743518059250575, + "flos": 20747631134880.0, + "grad_norm": 1.7309997295068698, + "learning_rate": 2.879891455619725e-06, + "loss": 0.852, + "num_input_tokens_seen": 134921205, + "step": 4151 + }, + { + "epoch": 0.37444198944852775, + "flos": 36484550664960.0, + "grad_norm": 1.883186208272747, + "learning_rate": 2.879366782711256e-06, + "loss": 0.7683, + "num_input_tokens_seen": 134953875, + "step": 4152 + }, + { + "epoch": 0.37453217297199803, + "flos": 24792188773440.0, + "grad_norm": 2.1080114407889416, + "learning_rate": 2.8788420347703643e-06, + "loss": 0.7854, + "num_input_tokens_seen": 134983000, + "step": 4153 + }, + { + "epoch": 0.37462235649546827, + "flos": 62273415511200.0, + "grad_norm": 0.6240268999699152, + "learning_rate": 2.8783172118418244e-06, + "loss": 0.5764, + "num_input_tokens_seen": 135072025, + "step": 4154 + }, + { + "epoch": 0.37471254001893856, + "flos": 30323771889600.0, + "grad_norm": 1.6161666817958995, + "learning_rate": 2.877792313970417e-06, + "loss": 0.7694, + "num_input_tokens_seen": 135101040, + "step": 4155 + }, + { + "epoch": 0.3748027235424088, + "flos": 26794764753120.0, + "grad_norm": 1.7428738050535377, + "learning_rate": 2.8772673412009293e-06, + "loss": 0.8191, + "num_input_tokens_seen": 135131545, + "step": 4156 + }, + { + "epoch": 0.3748929070658791, + "flos": 14728715098080.0, + "grad_norm": 1.7694943994722883, + "learning_rate": 2.8767422935781545e-06, + "loss": 0.8122, + "num_input_tokens_seen": 135158430, + "step": 4157 + }, + { + "epoch": 0.3749830905893493, + "flos": 18197718774720.0, + "grad_norm": 1.4425945397772266, + "learning_rate": 2.8762171711468935e-06, + "loss": 0.799, + "num_input_tokens_seen": 135186975, + "step": 4158 + }, + { + "epoch": 0.3750732741128196, + "flos": 18706244192160.0, + "grad_norm": 1.594060082283273, + "learning_rate": 2.875691973951952e-06, + "loss": 0.7662, + "num_input_tokens_seen": 135214700, + "step": 4159 + }, + { + "epoch": 0.37516345763628983, + "flos": 67167434038560.0, + "grad_norm": 0.6373703553635959, + "learning_rate": 2.8751667020381425e-06, + "loss": 0.6476, + "num_input_tokens_seen": 135303965, + "step": 4160 + }, + { + "epoch": 0.3752536411597601, + "flos": 24244108997760.0, + "grad_norm": 2.924420069400012, + "learning_rate": 2.8746413554502837e-06, + "loss": 0.8306, + "num_input_tokens_seen": 135333655, + "step": 4161 + }, + { + "epoch": 0.37534382468323035, + "flos": 24500843495520.0, + "grad_norm": 1.7292314846651888, + "learning_rate": 2.8741159342332027e-06, + "loss": 0.6965, + "num_input_tokens_seen": 135362070, + "step": 4162 + }, + { + "epoch": 0.37543400820670064, + "flos": 24682186252800.0, + "grad_norm": 1.8123923302602036, + "learning_rate": 2.87359043843173e-06, + "loss": 0.8208, + "num_input_tokens_seen": 135392310, + "step": 4163 + }, + { + "epoch": 0.3755241917301709, + "flos": 25266437938560.0, + "grad_norm": 1.5719997773594978, + "learning_rate": 2.873064868090704e-06, + "loss": 0.8018, + "num_input_tokens_seen": 135422945, + "step": 4164 + }, + { + "epoch": 0.37561437525364116, + "flos": 17031259740000.0, + "grad_norm": 2.339190511315823, + "learning_rate": 2.8725392232549697e-06, + "loss": 0.8549, + "num_input_tokens_seen": 135450170, + "step": 4165 + }, + { + "epoch": 0.3757045587771114, + "flos": 22313690989440.0, + "grad_norm": 2.0900962035382884, + "learning_rate": 2.872013503969378e-06, + "loss": 0.8353, + "num_input_tokens_seen": 135473635, + "step": 4166 + }, + { + "epoch": 0.3757947423005817, + "flos": 26029727856480.0, + "grad_norm": 1.728976396257247, + "learning_rate": 2.8714877102787853e-06, + "loss": 0.7764, + "num_input_tokens_seen": 135504710, + "step": 4167 + }, + { + "epoch": 0.3758849258240519, + "flos": 24278794117440.0, + "grad_norm": 2.2382891323623206, + "learning_rate": 2.8709618422280564e-06, + "loss": 0.7744, + "num_input_tokens_seen": 135533280, + "step": 4168 + }, + { + "epoch": 0.3759751093475222, + "flos": 27082021357440.0, + "grad_norm": 1.6897330168034126, + "learning_rate": 2.8704358998620605e-06, + "loss": 0.7416, + "num_input_tokens_seen": 135560770, + "step": 4169 + }, + { + "epoch": 0.3760652928709925, + "flos": 21257792021760.0, + "grad_norm": 1.6142776035727202, + "learning_rate": 2.8699098832256735e-06, + "loss": 0.7945, + "num_input_tokens_seen": 135588325, + "step": 4170 + }, + { + "epoch": 0.37615547639446273, + "flos": 16230125272800.0, + "grad_norm": 2.1619897273452158, + "learning_rate": 2.86938379236378e-06, + "loss": 0.6788, + "num_input_tokens_seen": 135615785, + "step": 4171 + }, + { + "epoch": 0.376245659917933, + "flos": 43408098981600.0, + "grad_norm": 1.6877186245053764, + "learning_rate": 2.868857627321266e-06, + "loss": 0.6913, + "num_input_tokens_seen": 135645845, + "step": 4172 + }, + { + "epoch": 0.37633584344140325, + "flos": 25447669186560.0, + "grad_norm": 2.1070301727756098, + "learning_rate": 2.8683313881430296e-06, + "loss": 0.7704, + "num_input_tokens_seen": 135675095, + "step": 4173 + }, + { + "epoch": 0.37642602696487354, + "flos": 16262543037120.0, + "grad_norm": 1.934661878472539, + "learning_rate": 2.8678050748739706e-06, + "loss": 0.7605, + "num_input_tokens_seen": 135699865, + "step": 4174 + }, + { + "epoch": 0.37651621048834377, + "flos": 22601951177280.0, + "grad_norm": 1.591108564157833, + "learning_rate": 2.8672786875589976e-06, + "loss": 0.7386, + "num_input_tokens_seen": 135728690, + "step": 4175 + }, + { + "epoch": 0.37660639401181406, + "flos": 23298138211200.0, + "grad_norm": 1.9461993911229456, + "learning_rate": 2.866752226243025e-06, + "loss": 0.7203, + "num_input_tokens_seen": 135756100, + "step": 4176 + }, + { + "epoch": 0.3766965775352843, + "flos": 24864644102880.0, + "grad_norm": 5.722499264591648, + "learning_rate": 2.8662256909709733e-06, + "loss": 0.8075, + "num_input_tokens_seen": 135783835, + "step": 4177 + }, + { + "epoch": 0.3767867610587546, + "flos": 19028966481120.0, + "grad_norm": 1.9118262481708923, + "learning_rate": 2.865699081787769e-06, + "loss": 0.8047, + "num_input_tokens_seen": 135812010, + "step": 4178 + }, + { + "epoch": 0.3768769445822248, + "flos": 39510570678240.0, + "grad_norm": 1.6679278688088752, + "learning_rate": 2.8651723987383465e-06, + "loss": 0.7485, + "num_input_tokens_seen": 135842690, + "step": 4179 + }, + { + "epoch": 0.3769671281056951, + "flos": 26103409788000.0, + "grad_norm": 1.9081525844725502, + "learning_rate": 2.8646456418676437e-06, + "loss": 0.744, + "num_input_tokens_seen": 135870725, + "step": 4180 + }, + { + "epoch": 0.37705731162916534, + "flos": 59044377036960.0, + "grad_norm": 0.7253477545063101, + "learning_rate": 2.8641188112206067e-06, + "loss": 0.5961, + "num_input_tokens_seen": 135962540, + "step": 4181 + }, + { + "epoch": 0.3771474951526356, + "flos": 16449052391040.0, + "grad_norm": 2.2928266075046535, + "learning_rate": 2.863591906842189e-06, + "loss": 0.7933, + "num_input_tokens_seen": 135990415, + "step": 4182 + }, + { + "epoch": 0.37723767867610586, + "flos": 28363389321120.0, + "grad_norm": 1.885275215477051, + "learning_rate": 2.863064928777347e-06, + "loss": 0.7986, + "num_input_tokens_seen": 136022955, + "step": 4183 + }, + { + "epoch": 0.37732786219957615, + "flos": 25409675958240.0, + "grad_norm": 1.946553703725491, + "learning_rate": 2.862537877071047e-06, + "loss": 0.7424, + "num_input_tokens_seen": 136051355, + "step": 4184 + }, + { + "epoch": 0.3774180457230464, + "flos": 69780732306720.0, + "grad_norm": 0.6448858268362797, + "learning_rate": 2.8620107517682597e-06, + "loss": 0.5602, + "num_input_tokens_seen": 136139255, + "step": 4185 + }, + { + "epoch": 0.37750822924651667, + "flos": 37288584373440.0, + "grad_norm": 1.5102195698842849, + "learning_rate": 2.8614835529139618e-06, + "loss": 0.6993, + "num_input_tokens_seen": 136174330, + "step": 4186 + }, + { + "epoch": 0.3775984127699869, + "flos": 18634383578880.0, + "grad_norm": 2.4204985614182126, + "learning_rate": 2.8609562805531367e-06, + "loss": 0.7248, + "num_input_tokens_seen": 136199530, + "step": 4187 + }, + { + "epoch": 0.3776885962934572, + "flos": 19321017984480.0, + "grad_norm": 1.9504828319913678, + "learning_rate": 2.8604289347307746e-06, + "loss": 0.7977, + "num_input_tokens_seen": 136227085, + "step": 4188 + }, + { + "epoch": 0.3777787798169274, + "flos": 20420039607360.0, + "grad_norm": 2.148688839169036, + "learning_rate": 2.859901515491871e-06, + "loss": 0.7603, + "num_input_tokens_seen": 136253815, + "step": 4189 + }, + { + "epoch": 0.3778689633403977, + "flos": 26284901224320.0, + "grad_norm": 1.549025764699267, + "learning_rate": 2.8593740228814298e-06, + "loss": 0.6515, + "num_input_tokens_seen": 136284455, + "step": 4190 + }, + { + "epoch": 0.37795914686386795, + "flos": 15282370337760.0, + "grad_norm": 2.2001182654410854, + "learning_rate": 2.8588464569444574e-06, + "loss": 0.7511, + "num_input_tokens_seen": 136309115, + "step": 4191 + }, + { + "epoch": 0.37804933038733823, + "flos": 22275028705440.0, + "grad_norm": 1.7555183189737906, + "learning_rate": 2.8583188177259697e-06, + "loss": 0.7566, + "num_input_tokens_seen": 136337555, + "step": 4192 + }, + { + "epoch": 0.3781395139108085, + "flos": 21876542978400.0, + "grad_norm": 1.81659886149054, + "learning_rate": 2.857791105270988e-06, + "loss": 0.7486, + "num_input_tokens_seen": 136364520, + "step": 4193 + }, + { + "epoch": 0.37822969743427876, + "flos": 22315029100800.0, + "grad_norm": 1.649322416877444, + "learning_rate": 2.857263319624539e-06, + "loss": 0.7966, + "num_input_tokens_seen": 136393380, + "step": 4194 + }, + { + "epoch": 0.37831988095774904, + "flos": 22023237785760.0, + "grad_norm": 1.6173854835689965, + "learning_rate": 2.856735460831657e-06, + "loss": 0.7484, + "num_input_tokens_seen": 136425695, + "step": 4195 + }, + { + "epoch": 0.3784100644812193, + "flos": 19252093782240.0, + "grad_norm": 1.5034041718472158, + "learning_rate": 2.856207528937382e-06, + "loss": 0.7738, + "num_input_tokens_seen": 136452940, + "step": 4196 + }, + { + "epoch": 0.37850024800468957, + "flos": 20598074256000.0, + "grad_norm": 4.2661338195271465, + "learning_rate": 2.855679523986759e-06, + "loss": 0.7496, + "num_input_tokens_seen": 136482840, + "step": 4197 + }, + { + "epoch": 0.3785904315281598, + "flos": 21290321295360.0, + "grad_norm": 4.0196662487199895, + "learning_rate": 2.8551514460248406e-06, + "loss": 0.814, + "num_input_tokens_seen": 136510265, + "step": 4198 + }, + { + "epoch": 0.3786806150516301, + "flos": 26465128888800.0, + "grad_norm": 1.8910708571643755, + "learning_rate": 2.8546232950966868e-06, + "loss": 0.7702, + "num_input_tokens_seen": 136540505, + "step": 4199 + }, + { + "epoch": 0.3787707985751003, + "flos": 25884445500000.0, + "grad_norm": 6.432472710371785, + "learning_rate": 2.85409507124736e-06, + "loss": 0.7145, + "num_input_tokens_seen": 136569160, + "step": 4200 + }, + { + "epoch": 0.3788609820985706, + "flos": 28580532290880.0, + "grad_norm": 3.40881796813505, + "learning_rate": 2.8535667745219324e-06, + "loss": 0.7569, + "num_input_tokens_seen": 136598170, + "step": 4201 + }, + { + "epoch": 0.37895116562204084, + "flos": 32224262507520.0, + "grad_norm": 1.7258230153353589, + "learning_rate": 2.853038404965481e-06, + "loss": 0.7563, + "num_input_tokens_seen": 136628305, + "step": 4202 + }, + { + "epoch": 0.37904134914551113, + "flos": 16297934382240.0, + "grad_norm": 1.952860944064325, + "learning_rate": 2.8525099626230894e-06, + "loss": 0.8368, + "num_input_tokens_seen": 136652740, + "step": 4203 + }, + { + "epoch": 0.37913153266898136, + "flos": 21914610546240.0, + "grad_norm": 2.7065036364096993, + "learning_rate": 2.8519814475398472e-06, + "loss": 0.6959, + "num_input_tokens_seen": 136681770, + "step": 4204 + }, + { + "epoch": 0.37922171619245165, + "flos": 26866328008320.0, + "grad_norm": 1.9114781585690812, + "learning_rate": 2.8514528597608502e-06, + "loss": 0.7424, + "num_input_tokens_seen": 136713545, + "step": 4205 + }, + { + "epoch": 0.3793118997159219, + "flos": 27377975685600.0, + "grad_norm": 1.6896213337505075, + "learning_rate": 2.8509241993312004e-06, + "loss": 0.7957, + "num_input_tokens_seen": 136745000, + "step": 4206 + }, + { + "epoch": 0.3794020832393922, + "flos": 22200937906560.0, + "grad_norm": 2.470827315153241, + "learning_rate": 2.850395466296006e-06, + "loss": 0.8178, + "num_input_tokens_seen": 136772525, + "step": 4207 + }, + { + "epoch": 0.3794922667628624, + "flos": 24059012094720.0, + "grad_norm": 2.115745250265205, + "learning_rate": 2.849866660700381e-06, + "loss": 0.7949, + "num_input_tokens_seen": 136801630, + "step": 4208 + }, + { + "epoch": 0.3795824502863327, + "flos": 30760176505440.0, + "grad_norm": 1.79866979774097, + "learning_rate": 2.8493377825894464e-06, + "loss": 0.8263, + "num_input_tokens_seen": 136831685, + "step": 4209 + }, + { + "epoch": 0.37967263380980293, + "flos": 25771506568320.0, + "grad_norm": 2.170086350546758, + "learning_rate": 2.848808832008329e-06, + "loss": 0.8256, + "num_input_tokens_seen": 136861035, + "step": 4210 + }, + { + "epoch": 0.3797628173332732, + "flos": 27924234143040.0, + "grad_norm": 1.7784739495515245, + "learning_rate": 2.848279809002162e-06, + "loss": 0.7941, + "num_input_tokens_seen": 136891870, + "step": 4211 + }, + { + "epoch": 0.37985300085674345, + "flos": 22315437968160.0, + "grad_norm": 2.1336178778586996, + "learning_rate": 2.8477507136160842e-06, + "loss": 0.782, + "num_input_tokens_seen": 136922305, + "step": 4212 + }, + { + "epoch": 0.37994318438021374, + "flos": 71144925962880.0, + "grad_norm": 0.6329129221790574, + "learning_rate": 2.847221545895241e-06, + "loss": 0.5884, + "num_input_tokens_seen": 137013925, + "step": 4213 + }, + { + "epoch": 0.38003336790368397, + "flos": 23115457342560.0, + "grad_norm": 1.499056439859635, + "learning_rate": 2.846692305884785e-06, + "loss": 0.7601, + "num_input_tokens_seen": 137044380, + "step": 4214 + }, + { + "epoch": 0.38012355142715426, + "flos": 23335202195520.0, + "grad_norm": 1.5502891240109806, + "learning_rate": 2.8461629936298718e-06, + "loss": 0.8078, + "num_input_tokens_seen": 137074985, + "step": 4215 + }, + { + "epoch": 0.3802137349506245, + "flos": 21767915738880.0, + "grad_norm": 1.8256050413321212, + "learning_rate": 2.845633609175666e-06, + "loss": 0.8201, + "num_input_tokens_seen": 137102360, + "step": 4216 + }, + { + "epoch": 0.3803039184740948, + "flos": 23954027491680.0, + "grad_norm": 1.8915471752030604, + "learning_rate": 2.8451041525673383e-06, + "loss": 0.7999, + "num_input_tokens_seen": 137132580, + "step": 4217 + }, + { + "epoch": 0.38039410199756507, + "flos": 19245849262560.0, + "grad_norm": 1.9050991774446464, + "learning_rate": 2.8445746238500647e-06, + "loss": 0.7975, + "num_input_tokens_seen": 137155350, + "step": 4218 + }, + { + "epoch": 0.3804842855210353, + "flos": 20638334839680.0, + "grad_norm": 1.8710358357874013, + "learning_rate": 2.844045023069027e-06, + "loss": 0.7471, + "num_input_tokens_seen": 137182935, + "step": 4219 + }, + { + "epoch": 0.3805744690445056, + "flos": 20820532501440.0, + "grad_norm": 1.6994273792091883, + "learning_rate": 2.8435153502694136e-06, + "loss": 0.8052, + "num_input_tokens_seen": 137212940, + "step": 4220 + }, + { + "epoch": 0.3806646525679758, + "flos": 17796222297120.0, + "grad_norm": 1.8577952365698778, + "learning_rate": 2.84298560549642e-06, + "loss": 0.7023, + "num_input_tokens_seen": 137239370, + "step": 4221 + }, + { + "epoch": 0.3807548360914461, + "flos": 23216724969600.0, + "grad_norm": 2.5444645620806603, + "learning_rate": 2.8424557887952462e-06, + "loss": 0.7939, + "num_input_tokens_seen": 137264210, + "step": 4222 + }, + { + "epoch": 0.38084501961491635, + "flos": 20565656491680.0, + "grad_norm": 4.470215243761929, + "learning_rate": 2.841925900211099e-06, + "loss": 0.6956, + "num_input_tokens_seen": 137292310, + "step": 4223 + }, + { + "epoch": 0.38093520313838664, + "flos": 25193239213920.0, + "grad_norm": 1.4380070994365033, + "learning_rate": 2.841395939789192e-06, + "loss": 0.8208, + "num_input_tokens_seen": 137323430, + "step": 4224 + }, + { + "epoch": 0.38102538666185687, + "flos": 22861584916320.0, + "grad_norm": 1.871139430243048, + "learning_rate": 2.8408659075747435e-06, + "loss": 0.7859, + "num_input_tokens_seen": 137352570, + "step": 4225 + }, + { + "epoch": 0.38111557018532716, + "flos": 18197830284000.0, + "grad_norm": 1.9892673407931782, + "learning_rate": 2.8403358036129796e-06, + "loss": 0.7667, + "num_input_tokens_seen": 137381350, + "step": 4226 + }, + { + "epoch": 0.3812057537087974, + "flos": 15246793143840.0, + "grad_norm": 1.7442792984703992, + "learning_rate": 2.839805627949132e-06, + "loss": 0.8203, + "num_input_tokens_seen": 137407925, + "step": 4227 + }, + { + "epoch": 0.3812959372322677, + "flos": 18124854577920.0, + "grad_norm": 1.905128586902324, + "learning_rate": 2.8392753806284367e-06, + "loss": 0.8222, + "num_input_tokens_seen": 137434410, + "step": 4228 + }, + { + "epoch": 0.3813861207557379, + "flos": 26322039548160.0, + "grad_norm": 4.3358631574426285, + "learning_rate": 2.838745061696139e-06, + "loss": 0.7178, + "num_input_tokens_seen": 137462780, + "step": 4229 + }, + { + "epoch": 0.3814763042792082, + "flos": 24423481757760.0, + "grad_norm": 1.6803620159564394, + "learning_rate": 2.838214671197487e-06, + "loss": 0.708, + "num_input_tokens_seen": 137493365, + "step": 4230 + }, + { + "epoch": 0.38156648780267843, + "flos": 22096436510400.0, + "grad_norm": 1.4969114454447665, + "learning_rate": 2.8376842091777377e-06, + "loss": 0.8748, + "num_input_tokens_seen": 137522585, + "step": 4231 + }, + { + "epoch": 0.3816566713261487, + "flos": 25225582638720.0, + "grad_norm": 1.7375652767516103, + "learning_rate": 2.8371536756821524e-06, + "loss": 0.678, + "num_input_tokens_seen": 137551455, + "step": 4232 + }, + { + "epoch": 0.38174685484961896, + "flos": 23844619687200.0, + "grad_norm": 1.7406370508403464, + "learning_rate": 2.836623070756e-06, + "loss": 0.7503, + "num_input_tokens_seen": 137582180, + "step": 4233 + }, + { + "epoch": 0.38183703837308924, + "flos": 18780372160800.0, + "grad_norm": 1.7384333511843386, + "learning_rate": 2.8360923944445542e-06, + "loss": 0.7957, + "num_input_tokens_seen": 137608005, + "step": 4234 + }, + { + "epoch": 0.3819272218965595, + "flos": 23587327643040.0, + "grad_norm": 1.6560524876702867, + "learning_rate": 2.8355616467930947e-06, + "loss": 0.7545, + "num_input_tokens_seen": 137637570, + "step": 4235 + }, + { + "epoch": 0.38201740542002977, + "flos": 20055718623360.0, + "grad_norm": 2.5234079270652483, + "learning_rate": 2.8350308278469085e-06, + "loss": 0.7494, + "num_input_tokens_seen": 137664670, + "step": 4236 + }, + { + "epoch": 0.3821075889435, + "flos": 24099532866720.0, + "grad_norm": 1.8339972449877897, + "learning_rate": 2.8344999376512877e-06, + "loss": 0.8083, + "num_input_tokens_seen": 137693060, + "step": 4237 + }, + { + "epoch": 0.3821977724669703, + "flos": 23334384460800.0, + "grad_norm": 1.5065436682387612, + "learning_rate": 2.8339689762515307e-06, + "loss": 0.761, + "num_input_tokens_seen": 137721640, + "step": 4238 + }, + { + "epoch": 0.3822879559904405, + "flos": 25957532715360.0, + "grad_norm": 2.3871533486722014, + "learning_rate": 2.8334379436929424e-06, + "loss": 0.7044, + "num_input_tokens_seen": 137748630, + "step": 4239 + }, + { + "epoch": 0.3823781395139108, + "flos": 24500546137440.0, + "grad_norm": 1.7473029061256107, + "learning_rate": 2.832906840020833e-06, + "loss": 0.7404, + "num_input_tokens_seen": 137779155, + "step": 4240 + }, + { + "epoch": 0.3824683230373811, + "flos": 22059707053920.0, + "grad_norm": 1.54123839882153, + "learning_rate": 2.83237566528052e-06, + "loss": 0.7839, + "num_input_tokens_seen": 137806835, + "step": 4241 + }, + { + "epoch": 0.38255850656085133, + "flos": 28835742828480.0, + "grad_norm": 1.7629071137468413, + "learning_rate": 2.831844419517325e-06, + "loss": 0.7488, + "num_input_tokens_seen": 137835255, + "step": 4242 + }, + { + "epoch": 0.3826486900843216, + "flos": 21148607235840.0, + "grad_norm": 1.5041488649929875, + "learning_rate": 2.8313131027765774e-06, + "loss": 0.7985, + "num_input_tokens_seen": 137863695, + "step": 4243 + }, + { + "epoch": 0.38273887360779185, + "flos": 20602014250560.0, + "grad_norm": 1.6236982266554019, + "learning_rate": 2.8307817151036124e-06, + "loss": 0.7433, + "num_input_tokens_seen": 137890905, + "step": 4244 + }, + { + "epoch": 0.38282905713126214, + "flos": 26539888743360.0, + "grad_norm": 1.675077775316712, + "learning_rate": 2.8302502565437704e-06, + "loss": 0.7455, + "num_input_tokens_seen": 137919385, + "step": 4245 + }, + { + "epoch": 0.3829192406547324, + "flos": 21473819898720.0, + "grad_norm": 1.9743014585124716, + "learning_rate": 2.829718727142398e-06, + "loss": 0.6986, + "num_input_tokens_seen": 137946620, + "step": 4246 + }, + { + "epoch": 0.38300942417820266, + "flos": 36085247203200.0, + "grad_norm": 1.7944960395171943, + "learning_rate": 2.829187126944849e-06, + "loss": 0.7795, + "num_input_tokens_seen": 137979565, + "step": 4247 + }, + { + "epoch": 0.3830996077016729, + "flos": 22642323270240.0, + "grad_norm": 2.465099097807575, + "learning_rate": 2.8286554559964826e-06, + "loss": 0.71, + "num_input_tokens_seen": 138006925, + "step": 4248 + }, + { + "epoch": 0.3831897912251432, + "flos": 24567723360960.0, + "grad_norm": 2.6572320356947032, + "learning_rate": 2.8281237143426637e-06, + "loss": 0.7893, + "num_input_tokens_seen": 138035325, + "step": 4249 + }, + { + "epoch": 0.3832799747486134, + "flos": 32841266485440.0, + "grad_norm": 2.5069310430683385, + "learning_rate": 2.8275919020287626e-06, + "loss": 0.6689, + "num_input_tokens_seen": 138068760, + "step": 4250 + }, + { + "epoch": 0.3833701582720837, + "flos": 24501029344320.0, + "grad_norm": 1.707704040489682, + "learning_rate": 2.827060019100158e-06, + "loss": 0.7982, + "num_input_tokens_seen": 138095885, + "step": 4251 + }, + { + "epoch": 0.38346034179555394, + "flos": 22565816436960.0, + "grad_norm": 1.627722953022118, + "learning_rate": 2.8265280656022315e-06, + "loss": 0.7633, + "num_input_tokens_seen": 138126100, + "step": 4252 + }, + { + "epoch": 0.3835505253190242, + "flos": 24274408085760.0, + "grad_norm": 1.8885228473997229, + "learning_rate": 2.825996041580373e-06, + "loss": 0.6601, + "num_input_tokens_seen": 138152170, + "step": 4253 + }, + { + "epoch": 0.38364070884249446, + "flos": 23042555976000.0, + "grad_norm": 1.7396097314472496, + "learning_rate": 2.825463947079978e-06, + "loss": 0.7913, + "num_input_tokens_seen": 138182645, + "step": 4254 + }, + { + "epoch": 0.38373089236596475, + "flos": 23402639607360.0, + "grad_norm": 1.6427461263808039, + "learning_rate": 2.8249317821464483e-06, + "loss": 0.7692, + "num_input_tokens_seen": 138211440, + "step": 4255 + }, + { + "epoch": 0.383821075889435, + "flos": 23771792660160.0, + "grad_norm": 1.6311484440435633, + "learning_rate": 2.824399546825189e-06, + "loss": 0.7016, + "num_input_tokens_seen": 138241930, + "step": 4256 + }, + { + "epoch": 0.38391125941290527, + "flos": 27303438849600.0, + "grad_norm": 1.474566609349861, + "learning_rate": 2.823867241161616e-06, + "loss": 0.7224, + "num_input_tokens_seen": 138272900, + "step": 4257 + }, + { + "epoch": 0.3840014429363755, + "flos": 23590078205280.0, + "grad_norm": 1.8758499762080512, + "learning_rate": 2.8233348652011456e-06, + "loss": 0.8228, + "num_input_tokens_seen": 138301860, + "step": 4258 + }, + { + "epoch": 0.3840916264598458, + "flos": 52080202172640.0, + "grad_norm": 1.8606442519099473, + "learning_rate": 2.8228024189892057e-06, + "loss": 0.6827, + "num_input_tokens_seen": 138334115, + "step": 4259 + }, + { + "epoch": 0.384181809983316, + "flos": 23224567788960.0, + "grad_norm": 1.3487051599807331, + "learning_rate": 2.822269902571226e-06, + "loss": 0.7732, + "num_input_tokens_seen": 138364055, + "step": 4260 + }, + { + "epoch": 0.3842719935067863, + "flos": 22893222115680.0, + "grad_norm": 1.5259678099100629, + "learning_rate": 2.8217373159926446e-06, + "loss": 0.75, + "num_input_tokens_seen": 138392630, + "step": 4261 + }, + { + "epoch": 0.38436217703025655, + "flos": 23802983822400.0, + "grad_norm": 1.292309042439524, + "learning_rate": 2.8212046592989046e-06, + "loss": 0.7089, + "num_input_tokens_seen": 138421240, + "step": 4262 + }, + { + "epoch": 0.38445236055372684, + "flos": 19756567695840.0, + "grad_norm": 1.8407058801675311, + "learning_rate": 2.820671932535455e-06, + "loss": 0.7769, + "num_input_tokens_seen": 138447575, + "step": 4263 + }, + { + "epoch": 0.3845425440771971, + "flos": 21221805960480.0, + "grad_norm": 3.438389056283213, + "learning_rate": 2.8201391357477506e-06, + "loss": 0.7517, + "num_input_tokens_seen": 138475110, + "step": 4264 + }, + { + "epoch": 0.38463272760066736, + "flos": 26030694270240.0, + "grad_norm": 4.287799364321035, + "learning_rate": 2.8196062689812525e-06, + "loss": 0.7457, + "num_input_tokens_seen": 138504285, + "step": 4265 + }, + { + "epoch": 0.38472291112413765, + "flos": 23334570309600.0, + "grad_norm": 1.917600141792467, + "learning_rate": 2.819073332281429e-06, + "loss": 0.7137, + "num_input_tokens_seen": 138533175, + "step": 4266 + }, + { + "epoch": 0.3848130946476079, + "flos": 70859044639680.0, + "grad_norm": 0.6157280772709415, + "learning_rate": 2.8185403256937524e-06, + "loss": 0.5494, + "num_input_tokens_seen": 138630810, + "step": 4267 + }, + { + "epoch": 0.38490327817107817, + "flos": 23258918380800.0, + "grad_norm": 1.7748452535522659, + "learning_rate": 2.8180072492637016e-06, + "loss": 0.7292, + "num_input_tokens_seen": 138659185, + "step": 4268 + }, + { + "epoch": 0.3849934616945484, + "flos": 18087790593600.0, + "grad_norm": 1.9666841961461976, + "learning_rate": 2.817474103036762e-06, + "loss": 0.7628, + "num_input_tokens_seen": 138686935, + "step": 4269 + }, + { + "epoch": 0.3850836452180187, + "flos": 24135741946560.0, + "grad_norm": 1.8938482882232548, + "learning_rate": 2.816940887058425e-06, + "loss": 0.7973, + "num_input_tokens_seen": 138715390, + "step": 4270 + }, + { + "epoch": 0.3851738287414889, + "flos": 19978988771520.0, + "grad_norm": 1.7537943033290084, + "learning_rate": 2.816407601374186e-06, + "loss": 0.8052, + "num_input_tokens_seen": 138741765, + "step": 4271 + }, + { + "epoch": 0.3852640122649592, + "flos": 48100703081280.0, + "grad_norm": 1.6948050165695914, + "learning_rate": 2.815874246029549e-06, + "loss": 0.7027, + "num_input_tokens_seen": 138773210, + "step": 4272 + }, + { + "epoch": 0.38535419578842944, + "flos": 30656381334720.0, + "grad_norm": 1.5863094551361105, + "learning_rate": 2.815340821070023e-06, + "loss": 0.717, + "num_input_tokens_seen": 138803785, + "step": 4273 + }, + { + "epoch": 0.38544437931189973, + "flos": 29272965179040.0, + "grad_norm": 1.769527814242836, + "learning_rate": 2.814807326541122e-06, + "loss": 0.7494, + "num_input_tokens_seen": 138829610, + "step": 4274 + }, + { + "epoch": 0.38553456283536996, + "flos": 23224010242560.0, + "grad_norm": 1.5300143959265722, + "learning_rate": 2.8142737624883676e-06, + "loss": 0.7027, + "num_input_tokens_seen": 138859090, + "step": 4275 + }, + { + "epoch": 0.38562474635884025, + "flos": 28102454640480.0, + "grad_norm": 1.6212247864491283, + "learning_rate": 2.8137401289572854e-06, + "loss": 0.7676, + "num_input_tokens_seen": 138888775, + "step": 4276 + }, + { + "epoch": 0.3857149298823105, + "flos": 24318162626880.0, + "grad_norm": 1.6962806995736925, + "learning_rate": 2.8132064259934086e-06, + "loss": 0.8118, + "num_input_tokens_seen": 138916840, + "step": 4277 + }, + { + "epoch": 0.3858051134057808, + "flos": 49778661114240.0, + "grad_norm": 1.597709524297061, + "learning_rate": 2.812672653642276e-06, + "loss": 0.7648, + "num_input_tokens_seen": 138949220, + "step": 4278 + }, + { + "epoch": 0.385895296929251, + "flos": 25410976899840.0, + "grad_norm": 1.4554097252188878, + "learning_rate": 2.812138811949431e-06, + "loss": 0.78, + "num_input_tokens_seen": 138979965, + "step": 4279 + }, + { + "epoch": 0.3859854804527213, + "flos": 20674692598560.0, + "grad_norm": 2.476974501169392, + "learning_rate": 2.8116049009604247e-06, + "loss": 0.8323, + "num_input_tokens_seen": 139007635, + "step": 4280 + }, + { + "epoch": 0.38607566397619153, + "flos": 28617559105440.0, + "grad_norm": 1.6193258549996694, + "learning_rate": 2.8110709207208132e-06, + "loss": 0.6502, + "num_input_tokens_seen": 139038735, + "step": 4281 + }, + { + "epoch": 0.3861658474996618, + "flos": 20529224393280.0, + "grad_norm": 2.712031338896049, + "learning_rate": 2.810536871276158e-06, + "loss": 0.7976, + "num_input_tokens_seen": 139065515, + "step": 4282 + }, + { + "epoch": 0.38625603102313205, + "flos": 59027353286880.0, + "grad_norm": 0.7161570923816054, + "learning_rate": 2.8100027526720283e-06, + "loss": 0.6289, + "num_input_tokens_seen": 139150325, + "step": 4283 + }, + { + "epoch": 0.38634621454660234, + "flos": 30693556828320.0, + "grad_norm": 1.6525398300380116, + "learning_rate": 2.8094685649539974e-06, + "loss": 0.6955, + "num_input_tokens_seen": 139179495, + "step": 4284 + }, + { + "epoch": 0.3864363980700726, + "flos": 64148364770400.0, + "grad_norm": 0.7123922617919866, + "learning_rate": 2.8089343081676455e-06, + "loss": 0.6005, + "num_input_tokens_seen": 139270760, + "step": 4285 + }, + { + "epoch": 0.38652658159354286, + "flos": 20711793752640.0, + "grad_norm": 1.7602719204175317, + "learning_rate": 2.8083999823585577e-06, + "loss": 0.7488, + "num_input_tokens_seen": 139298825, + "step": 4286 + }, + { + "epoch": 0.3866167651170131, + "flos": 22641691384320.0, + "grad_norm": 1.4257151967255144, + "learning_rate": 2.8078655875723254e-06, + "loss": 0.7929, + "num_input_tokens_seen": 139328110, + "step": 4287 + }, + { + "epoch": 0.3867069486404834, + "flos": 12841568424000.0, + "grad_norm": 2.394768508681906, + "learning_rate": 2.807331123854547e-06, + "loss": 0.851, + "num_input_tokens_seen": 139354575, + "step": 4288 + }, + { + "epoch": 0.38679713216395367, + "flos": 25229188105440.0, + "grad_norm": 2.367354887530212, + "learning_rate": 2.806796591250826e-06, + "loss": 0.6379, + "num_input_tokens_seen": 139382690, + "step": 4289 + }, + { + "epoch": 0.3868873156874239, + "flos": 18270359952960.0, + "grad_norm": 2.070114183824125, + "learning_rate": 2.8062619898067707e-06, + "loss": 0.8323, + "num_input_tokens_seen": 139410105, + "step": 4290 + }, + { + "epoch": 0.3869774992108942, + "flos": 28945039123680.0, + "grad_norm": 2.2679457781262653, + "learning_rate": 2.8057273195679963e-06, + "loss": 0.7804, + "num_input_tokens_seen": 139440115, + "step": 4291 + }, + { + "epoch": 0.3870676827343644, + "flos": 26426652453600.0, + "grad_norm": 2.3456139845351838, + "learning_rate": 2.8051925805801253e-06, + "loss": 0.783, + "num_input_tokens_seen": 139468280, + "step": 4292 + }, + { + "epoch": 0.3871578662578347, + "flos": 22679387254560.0, + "grad_norm": 1.6566874808008776, + "learning_rate": 2.804657772888783e-06, + "loss": 0.8001, + "num_input_tokens_seen": 139497595, + "step": 4293 + }, + { + "epoch": 0.38724804978130495, + "flos": 18889742795520.0, + "grad_norm": 1.9245220656166901, + "learning_rate": 2.804122896539602e-06, + "loss": 0.7647, + "num_input_tokens_seen": 139524315, + "step": 4294 + }, + { + "epoch": 0.38733823330477524, + "flos": 33353434539360.0, + "grad_norm": 1.5584520930495638, + "learning_rate": 2.8035879515782225e-06, + "loss": 0.7195, + "num_input_tokens_seen": 139553290, + "step": 4295 + }, + { + "epoch": 0.38742841682824547, + "flos": 18925245649920.0, + "grad_norm": 2.2006402037858335, + "learning_rate": 2.803052938050288e-06, + "loss": 0.82, + "num_input_tokens_seen": 139578460, + "step": 4296 + }, + { + "epoch": 0.38751860035171576, + "flos": 26795619657600.0, + "grad_norm": 2.615780398037893, + "learning_rate": 2.802517856001449e-06, + "loss": 0.8095, + "num_input_tokens_seen": 139608260, + "step": 4297 + }, + { + "epoch": 0.387608783875186, + "flos": 35093217350400.0, + "grad_norm": 2.1944544734269615, + "learning_rate": 2.801982705477361e-06, + "loss": 0.7299, + "num_input_tokens_seen": 139634915, + "step": 4298 + }, + { + "epoch": 0.3876989673986563, + "flos": 21548282395200.0, + "grad_norm": 1.5127023262073118, + "learning_rate": 2.8014474865236867e-06, + "loss": 0.749, + "num_input_tokens_seen": 139662890, + "step": 4299 + }, + { + "epoch": 0.3877891509221265, + "flos": 28723212764160.0, + "grad_norm": 1.727619037750093, + "learning_rate": 2.800912199186094e-06, + "loss": 0.7583, + "num_input_tokens_seen": 139691915, + "step": 4300 + }, + { + "epoch": 0.3878793344455968, + "flos": 34884363237120.0, + "grad_norm": 1.749162814296765, + "learning_rate": 2.800376843510256e-06, + "loss": 0.7758, + "num_input_tokens_seen": 139723715, + "step": 4301 + }, + { + "epoch": 0.38796951796906703, + "flos": 23661864479040.0, + "grad_norm": 1.6708059012173875, + "learning_rate": 2.799841419541852e-06, + "loss": 0.788, + "num_input_tokens_seen": 139753485, + "step": 4302 + }, + { + "epoch": 0.3880597014925373, + "flos": 69565224806400.0, + "grad_norm": 0.740754883077068, + "learning_rate": 2.799305927326568e-06, + "loss": 0.5842, + "num_input_tokens_seen": 139846160, + "step": 4303 + }, + { + "epoch": 0.38814988501600756, + "flos": 70475321040960.0, + "grad_norm": 0.7164630164366538, + "learning_rate": 2.7987703669100955e-06, + "loss": 0.5939, + "num_input_tokens_seen": 139934175, + "step": 4304 + }, + { + "epoch": 0.38824006853947785, + "flos": 21694791353760.0, + "grad_norm": 1.6269708566054304, + "learning_rate": 2.79823473833813e-06, + "loss": 0.7709, + "num_input_tokens_seen": 139963905, + "step": 4305 + }, + { + "epoch": 0.3883302520629481, + "flos": 22204654882560.0, + "grad_norm": 1.969315644401513, + "learning_rate": 2.797699041656376e-06, + "loss": 0.7127, + "num_input_tokens_seen": 139993050, + "step": 4306 + }, + { + "epoch": 0.38842043558641837, + "flos": 29636505598080.0, + "grad_norm": 2.051969754195063, + "learning_rate": 2.7971632769105412e-06, + "loss": 0.7807, + "num_input_tokens_seen": 140021255, + "step": 4307 + }, + { + "epoch": 0.3885106191098886, + "flos": 31859235298080.0, + "grad_norm": 1.8929342243920564, + "learning_rate": 2.79662744414634e-06, + "loss": 0.7289, + "num_input_tokens_seen": 140048590, + "step": 4308 + }, + { + "epoch": 0.3886008026333589, + "flos": 20273716497600.0, + "grad_norm": 1.852103946887181, + "learning_rate": 2.7960915434094923e-06, + "loss": 0.7159, + "num_input_tokens_seen": 140075785, + "step": 4309 + }, + { + "epoch": 0.3886909861568291, + "flos": 22678792538400.0, + "grad_norm": 1.8105993499887847, + "learning_rate": 2.7955555747457256e-06, + "loss": 0.741, + "num_input_tokens_seen": 140103145, + "step": 4310 + }, + { + "epoch": 0.3887811696802994, + "flos": 27888694118880.0, + "grad_norm": 1.5971666252669372, + "learning_rate": 2.79501953820077e-06, + "loss": 0.7591, + "num_input_tokens_seen": 140134780, + "step": 4311 + }, + { + "epoch": 0.3888713532037697, + "flos": 21366790958880.0, + "grad_norm": 1.6311390163039547, + "learning_rate": 2.7944834338203637e-06, + "loss": 0.7813, + "num_input_tokens_seen": 140161765, + "step": 4312 + }, + { + "epoch": 0.38896153672723993, + "flos": 26722012065600.0, + "grad_norm": 1.7161143121057978, + "learning_rate": 2.79394726165025e-06, + "loss": 0.7258, + "num_input_tokens_seen": 140193235, + "step": 4313 + }, + { + "epoch": 0.3890517202507102, + "flos": 21652560772800.0, + "grad_norm": 2.2457950293884523, + "learning_rate": 2.793411021736178e-06, + "loss": 0.7415, + "num_input_tokens_seen": 140216935, + "step": 4314 + }, + { + "epoch": 0.38914190377418045, + "flos": 23985590351520.0, + "grad_norm": 1.8477369617471309, + "learning_rate": 2.7928747141239027e-06, + "loss": 0.8528, + "num_input_tokens_seen": 140247535, + "step": 4315 + }, + { + "epoch": 0.38923208729765074, + "flos": 70222377858720.0, + "grad_norm": 0.6482798373192478, + "learning_rate": 2.7923383388591856e-06, + "loss": 0.5878, + "num_input_tokens_seen": 140338480, + "step": 4316 + }, + { + "epoch": 0.389322270821121, + "flos": 27412660805280.0, + "grad_norm": 2.142247163149019, + "learning_rate": 2.7918018959877923e-06, + "loss": 0.731, + "num_input_tokens_seen": 140368090, + "step": 4317 + }, + { + "epoch": 0.38941245434459126, + "flos": 25084017258240.0, + "grad_norm": 1.6636027386020242, + "learning_rate": 2.791265385555495e-06, + "loss": 0.7912, + "num_input_tokens_seen": 140395575, + "step": 4318 + }, + { + "epoch": 0.3895026378680615, + "flos": 22351052331840.0, + "grad_norm": 1.6803327873414242, + "learning_rate": 2.790728807608072e-06, + "loss": 0.828, + "num_input_tokens_seen": 140423215, + "step": 4319 + }, + { + "epoch": 0.3895928213915318, + "flos": 20383719018240.0, + "grad_norm": 2.025702427670454, + "learning_rate": 2.790192162191307e-06, + "loss": 0.7492, + "num_input_tokens_seen": 140449195, + "step": 4320 + }, + { + "epoch": 0.389683004915002, + "flos": 28362608756160.0, + "grad_norm": 2.4197433442770024, + "learning_rate": 2.78965544935099e-06, + "loss": 0.7936, + "num_input_tokens_seen": 140479090, + "step": 4321 + }, + { + "epoch": 0.3897731884384723, + "flos": 18743419685760.0, + "grad_norm": 1.8064212810259563, + "learning_rate": 2.789118669132916e-06, + "loss": 0.7445, + "num_input_tokens_seen": 140506225, + "step": 4322 + }, + { + "epoch": 0.38986337196194254, + "flos": 30292989594720.0, + "grad_norm": 3.2684400189028295, + "learning_rate": 2.7885818215828856e-06, + "loss": 0.8271, + "num_input_tokens_seen": 140537260, + "step": 4323 + }, + { + "epoch": 0.38995355548541283, + "flos": 18301290926880.0, + "grad_norm": 2.083812760184282, + "learning_rate": 2.7880449067467064e-06, + "loss": 0.7194, + "num_input_tokens_seen": 140560925, + "step": 4324 + }, + { + "epoch": 0.39004373900888306, + "flos": 23662570704480.0, + "grad_norm": 1.988742018035657, + "learning_rate": 2.78750792467019e-06, + "loss": 0.7196, + "num_input_tokens_seen": 140589840, + "step": 4325 + }, + { + "epoch": 0.39013392253235335, + "flos": 27122653638720.0, + "grad_norm": 1.8241522243875474, + "learning_rate": 2.786970875399156e-06, + "loss": 0.6318, + "num_input_tokens_seen": 140617065, + "step": 4326 + }, + { + "epoch": 0.3902241060558236, + "flos": 34700381426880.0, + "grad_norm": 2.001834385769846, + "learning_rate": 2.7864337589794267e-06, + "loss": 0.6982, + "num_input_tokens_seen": 140652455, + "step": 4327 + }, + { + "epoch": 0.39031428957929387, + "flos": 15094522872480.0, + "grad_norm": 1.7826618823018399, + "learning_rate": 2.7858965754568335e-06, + "loss": 0.7822, + "num_input_tokens_seen": 140677210, + "step": 4328 + }, + { + "epoch": 0.3904044731027641, + "flos": 21038864903520.0, + "grad_norm": 1.7418055778605959, + "learning_rate": 2.785359324877211e-06, + "loss": 0.6996, + "num_input_tokens_seen": 140706155, + "step": 4329 + }, + { + "epoch": 0.3904946566262344, + "flos": 24714492507840.0, + "grad_norm": 1.7467544978143796, + "learning_rate": 2.7848220072864e-06, + "loss": 0.7892, + "num_input_tokens_seen": 140735680, + "step": 4330 + }, + { + "epoch": 0.3905848401497046, + "flos": 33200866909920.0, + "grad_norm": 1.8292007541594044, + "learning_rate": 2.784284622730248e-06, + "loss": 0.7525, + "num_input_tokens_seen": 140764190, + "step": 4331 + }, + { + "epoch": 0.3906750236731749, + "flos": 22168222784160.0, + "grad_norm": 1.694502606288316, + "learning_rate": 2.7837471712546073e-06, + "loss": 0.7807, + "num_input_tokens_seen": 140792930, + "step": 4332 + }, + { + "epoch": 0.39076520719664515, + "flos": 20528815525920.0, + "grad_norm": 1.4731559411587885, + "learning_rate": 2.783209652905337e-06, + "loss": 0.7365, + "num_input_tokens_seen": 140823265, + "step": 4333 + }, + { + "epoch": 0.39085539072011544, + "flos": 24061651147680.0, + "grad_norm": 1.504532007598924, + "learning_rate": 2.7826720677283e-06, + "loss": 0.7255, + "num_input_tokens_seen": 140856355, + "step": 4334 + }, + { + "epoch": 0.39094557424358567, + "flos": 33498902744640.0, + "grad_norm": 6.403906714632623, + "learning_rate": 2.782134415769367e-06, + "loss": 0.726, + "num_input_tokens_seen": 140888260, + "step": 4335 + }, + { + "epoch": 0.39103575776705596, + "flos": 31531309242720.0, + "grad_norm": 2.02697091909888, + "learning_rate": 2.7815966970744126e-06, + "loss": 0.8081, + "num_input_tokens_seen": 140917760, + "step": 4336 + }, + { + "epoch": 0.39112594129052625, + "flos": 21003176200320.0, + "grad_norm": 1.9425256215373272, + "learning_rate": 2.7810589116893184e-06, + "loss": 0.8298, + "num_input_tokens_seen": 140945600, + "step": 4337 + }, + { + "epoch": 0.3912161248139965, + "flos": 21400137967200.0, + "grad_norm": 2.903329181368896, + "learning_rate": 2.780521059659972e-06, + "loss": 0.7648, + "num_input_tokens_seen": 140976245, + "step": 4338 + }, + { + "epoch": 0.39130630833746677, + "flos": 24639621144000.0, + "grad_norm": 2.712380412757311, + "learning_rate": 2.7799831410322637e-06, + "loss": 0.7478, + "num_input_tokens_seen": 141003800, + "step": 4339 + }, + { + "epoch": 0.391396491860937, + "flos": 21836468243520.0, + "grad_norm": 1.52818630228163, + "learning_rate": 2.779445155852094e-06, + "loss": 0.7181, + "num_input_tokens_seen": 141035190, + "step": 4340 + }, + { + "epoch": 0.3914866753844073, + "flos": 70190443301280.0, + "grad_norm": 0.6547709100070972, + "learning_rate": 2.7789071041653655e-06, + "loss": 0.6252, + "num_input_tokens_seen": 141132800, + "step": 4341 + }, + { + "epoch": 0.3915768589078775, + "flos": 23401784702880.0, + "grad_norm": 1.6772116223031515, + "learning_rate": 2.7783689860179875e-06, + "loss": 0.8147, + "num_input_tokens_seen": 141162675, + "step": 4342 + }, + { + "epoch": 0.3916670424313478, + "flos": 18379544738880.0, + "grad_norm": 2.074726624462525, + "learning_rate": 2.7778308014558767e-06, + "loss": 0.6996, + "num_input_tokens_seen": 141189720, + "step": 4343 + }, + { + "epoch": 0.39175722595481804, + "flos": 40384978209600.0, + "grad_norm": 2.011231239283507, + "learning_rate": 2.7772925505249524e-06, + "loss": 0.8359, + "num_input_tokens_seen": 141221350, + "step": 4344 + }, + { + "epoch": 0.39184740947828833, + "flos": 19722663141120.0, + "grad_norm": 1.6937500394138458, + "learning_rate": 2.7767542332711417e-06, + "loss": 0.7285, + "num_input_tokens_seen": 141247735, + "step": 4345 + }, + { + "epoch": 0.39193759300175857, + "flos": 21293071857600.0, + "grad_norm": 1.6673870993258924, + "learning_rate": 2.776215849740377e-06, + "loss": 0.7368, + "num_input_tokens_seen": 141275890, + "step": 4346 + }, + { + "epoch": 0.39202777652522885, + "flos": 24427273073280.0, + "grad_norm": 1.553777226940681, + "learning_rate": 2.775677399978596e-06, + "loss": 0.7767, + "num_input_tokens_seen": 141305700, + "step": 4347 + }, + { + "epoch": 0.3921179600486991, + "flos": 66685193375040.0, + "grad_norm": 0.5701360821574185, + "learning_rate": 2.775138884031742e-06, + "loss": 0.5723, + "num_input_tokens_seen": 141401215, + "step": 4348 + }, + { + "epoch": 0.3922081435721694, + "flos": 18888887891040.0, + "grad_norm": 1.8027582466636218, + "learning_rate": 2.774600301945764e-06, + "loss": 0.72, + "num_input_tokens_seen": 141429845, + "step": 4349 + }, + { + "epoch": 0.3922983270956396, + "flos": 24496791991680.0, + "grad_norm": 1.7773101468106045, + "learning_rate": 2.774061653766618e-06, + "loss": 0.7928, + "num_input_tokens_seen": 141459935, + "step": 4350 + }, + { + "epoch": 0.3923885106191099, + "flos": 23735434901280.0, + "grad_norm": 1.7027941041128805, + "learning_rate": 2.773522939540263e-06, + "loss": 0.7534, + "num_input_tokens_seen": 141486950, + "step": 4351 + }, + { + "epoch": 0.39247869414258013, + "flos": 23332823330880.0, + "grad_norm": 2.6691675578661846, + "learning_rate": 2.7729841593126663e-06, + "loss": 0.6873, + "num_input_tokens_seen": 141515420, + "step": 4352 + }, + { + "epoch": 0.3925688776660504, + "flos": 22680725365920.0, + "grad_norm": 2.124649018866208, + "learning_rate": 2.7724453131297988e-06, + "loss": 0.8253, + "num_input_tokens_seen": 141548865, + "step": 4353 + }, + { + "epoch": 0.39265906118952065, + "flos": 26468065299840.0, + "grad_norm": 2.126288416098879, + "learning_rate": 2.771906401037637e-06, + "loss": 0.7869, + "num_input_tokens_seen": 141578140, + "step": 4354 + }, + { + "epoch": 0.39274924471299094, + "flos": 18161100827520.0, + "grad_norm": 1.8550974263520326, + "learning_rate": 2.7713674230821664e-06, + "loss": 0.7238, + "num_input_tokens_seen": 141603885, + "step": 4355 + }, + { + "epoch": 0.3928394282364612, + "flos": 22022531560320.0, + "grad_norm": 1.6238454238304196, + "learning_rate": 2.7708283793093724e-06, + "loss": 0.7159, + "num_input_tokens_seen": 141634790, + "step": 4356 + }, + { + "epoch": 0.39292961175993146, + "flos": 35575606692960.0, + "grad_norm": 1.6343475864656913, + "learning_rate": 2.7702892697652514e-06, + "loss": 0.7337, + "num_input_tokens_seen": 141664120, + "step": 4357 + }, + { + "epoch": 0.3930197952834017, + "flos": 24753489319680.0, + "grad_norm": 1.4492206419240936, + "learning_rate": 2.7697500944958024e-06, + "loss": 0.7396, + "num_input_tokens_seen": 141695875, + "step": 4358 + }, + { + "epoch": 0.393109978806872, + "flos": 19976089530240.0, + "grad_norm": 1.8962286512547362, + "learning_rate": 2.7692108535470312e-06, + "loss": 0.8598, + "num_input_tokens_seen": 141722830, + "step": 4359 + }, + { + "epoch": 0.3932001623303423, + "flos": 18995322114720.0, + "grad_norm": 1.9243795930932126, + "learning_rate": 2.768671546964948e-06, + "loss": 0.7963, + "num_input_tokens_seen": 141747070, + "step": 4360 + }, + { + "epoch": 0.3932903458538125, + "flos": 13716236143680.0, + "grad_norm": 1.8302735964219665, + "learning_rate": 2.7681321747955713e-06, + "loss": 0.7864, + "num_input_tokens_seen": 141772000, + "step": 4361 + }, + { + "epoch": 0.3933805293772828, + "flos": 25626856097760.0, + "grad_norm": 1.6280766485433371, + "learning_rate": 2.767592737084921e-06, + "loss": 0.7794, + "num_input_tokens_seen": 141802795, + "step": 4362 + }, + { + "epoch": 0.39347071290075303, + "flos": 22605407964960.0, + "grad_norm": 2.0249801116720225, + "learning_rate": 2.767053233879026e-06, + "loss": 0.8584, + "num_input_tokens_seen": 141830340, + "step": 4363 + }, + { + "epoch": 0.3935608964242233, + "flos": 26139433019040.0, + "grad_norm": 1.68592223831354, + "learning_rate": 2.76651366522392e-06, + "loss": 0.8372, + "num_input_tokens_seen": 141858510, + "step": 4364 + }, + { + "epoch": 0.39365107994769355, + "flos": 22933965906240.0, + "grad_norm": 2.6052162700065957, + "learning_rate": 2.7659740311656413e-06, + "loss": 0.7312, + "num_input_tokens_seen": 141886355, + "step": 4365 + }, + { + "epoch": 0.39374126347116384, + "flos": 20854065358560.0, + "grad_norm": 2.103710225772498, + "learning_rate": 2.7654343317502352e-06, + "loss": 0.7855, + "num_input_tokens_seen": 141915125, + "step": 4366 + }, + { + "epoch": 0.39383144699463407, + "flos": 25259673042240.0, + "grad_norm": 1.785584179361904, + "learning_rate": 2.7648945670237502e-06, + "loss": 0.7446, + "num_input_tokens_seen": 141943630, + "step": 4367 + }, + { + "epoch": 0.39392163051810436, + "flos": 21731000433600.0, + "grad_norm": 3.394162631021135, + "learning_rate": 2.7643547370322446e-06, + "loss": 0.8207, + "num_input_tokens_seen": 141970615, + "step": 4368 + }, + { + "epoch": 0.3940118140415746, + "flos": 24901336389600.0, + "grad_norm": 2.267763168052818, + "learning_rate": 2.7638148418217775e-06, + "loss": 0.8012, + "num_input_tokens_seen": 142000000, + "step": 4369 + }, + { + "epoch": 0.3941019975650449, + "flos": 21658099067040.0, + "grad_norm": 1.9108586492357327, + "learning_rate": 2.7632748814384163e-06, + "loss": 0.8037, + "num_input_tokens_seen": 142023640, + "step": 4370 + }, + { + "epoch": 0.3941921810885151, + "flos": 26063260713600.0, + "grad_norm": 1.4093720852637759, + "learning_rate": 2.7627348559282335e-06, + "loss": 0.7874, + "num_input_tokens_seen": 142055130, + "step": 4371 + }, + { + "epoch": 0.3942823646119854, + "flos": 24134106477120.0, + "grad_norm": 1.6176681172943266, + "learning_rate": 2.7621947653373075e-06, + "loss": 0.7008, + "num_input_tokens_seen": 142084745, + "step": 4372 + }, + { + "epoch": 0.39437254813545564, + "flos": 64870279011840.0, + "grad_norm": 0.660464401104455, + "learning_rate": 2.7616546097117213e-06, + "loss": 0.5637, + "num_input_tokens_seen": 142178570, + "step": 4373 + }, + { + "epoch": 0.3944627316589259, + "flos": 20894437451520.0, + "grad_norm": 1.8974323515964642, + "learning_rate": 2.761114389097564e-06, + "loss": 0.7537, + "num_input_tokens_seen": 142207200, + "step": 4374 + }, + { + "epoch": 0.39455291518239616, + "flos": 22314248535840.0, + "grad_norm": 1.6952593395142965, + "learning_rate": 2.7605741035409305e-06, + "loss": 0.7968, + "num_input_tokens_seen": 142235205, + "step": 4375 + }, + { + "epoch": 0.39464309870586645, + "flos": 33536524275360.0, + "grad_norm": 1.532615519863372, + "learning_rate": 2.76003375308792e-06, + "loss": 0.7406, + "num_input_tokens_seen": 142267720, + "step": 4376 + }, + { + "epoch": 0.3947332822293367, + "flos": 22714667090400.0, + "grad_norm": 1.4711064042870505, + "learning_rate": 2.75949333778464e-06, + "loss": 0.726, + "num_input_tokens_seen": 142298560, + "step": 4377 + }, + { + "epoch": 0.39482346575280697, + "flos": 21949704533280.0, + "grad_norm": 1.593496190224849, + "learning_rate": 2.7589528576772e-06, + "loss": 0.6464, + "num_input_tokens_seen": 142328335, + "step": 4378 + }, + { + "epoch": 0.3949136492762772, + "flos": 40348360262400.0, + "grad_norm": 1.76937912714459, + "learning_rate": 2.758412312811717e-06, + "loss": 0.7486, + "num_input_tokens_seen": 142358535, + "step": 4379 + }, + { + "epoch": 0.3950038327997475, + "flos": 63282660696480.0, + "grad_norm": 0.7466359589923777, + "learning_rate": 2.7578717032343146e-06, + "loss": 0.6294, + "num_input_tokens_seen": 142440590, + "step": 4380 + }, + { + "epoch": 0.3950940163232177, + "flos": 23438402650080.0, + "grad_norm": 1.748117456130061, + "learning_rate": 2.757331028991119e-06, + "loss": 0.7845, + "num_input_tokens_seen": 142470680, + "step": 4381 + }, + { + "epoch": 0.395184199846688, + "flos": 22278076625760.0, + "grad_norm": 2.147885560317054, + "learning_rate": 2.7567902901282642e-06, + "loss": 0.7764, + "num_input_tokens_seen": 142498345, + "step": 4382 + }, + { + "epoch": 0.3952743833701583, + "flos": 21003176200320.0, + "grad_norm": 2.0758299810728826, + "learning_rate": 2.7562494866918892e-06, + "loss": 0.7506, + "num_input_tokens_seen": 142526475, + "step": 4383 + }, + { + "epoch": 0.39536456689362853, + "flos": 18416311365120.0, + "grad_norm": 1.5823535732123712, + "learning_rate": 2.7557086187281378e-06, + "loss": 0.7332, + "num_input_tokens_seen": 142554865, + "step": 4384 + }, + { + "epoch": 0.3954547504170988, + "flos": 22824149234400.0, + "grad_norm": 1.8749254847660728, + "learning_rate": 2.75516768628316e-06, + "loss": 0.8558, + "num_input_tokens_seen": 142583150, + "step": 4385 + }, + { + "epoch": 0.39554493394056905, + "flos": 24315337725120.0, + "grad_norm": 1.5972912658290213, + "learning_rate": 2.7546266894031114e-06, + "loss": 0.8297, + "num_input_tokens_seen": 142614415, + "step": 4386 + }, + { + "epoch": 0.39563511746403934, + "flos": 30798727280160.0, + "grad_norm": 1.63238826515375, + "learning_rate": 2.7540856281341526e-06, + "loss": 0.7035, + "num_input_tokens_seen": 142643215, + "step": 4387 + }, + { + "epoch": 0.3957253009875096, + "flos": 25809351117600.0, + "grad_norm": 2.18091485801714, + "learning_rate": 2.7535445025224506e-06, + "loss": 0.7233, + "num_input_tokens_seen": 142673250, + "step": 4388 + }, + { + "epoch": 0.39581548451097986, + "flos": 33207668976000.0, + "grad_norm": 1.9577467290085937, + "learning_rate": 2.753003312614176e-06, + "loss": 0.7618, + "num_input_tokens_seen": 142702590, + "step": 4389 + }, + { + "epoch": 0.3959056680344501, + "flos": 22350866483040.0, + "grad_norm": 2.5361408972363435, + "learning_rate": 2.7524620584555065e-06, + "loss": 0.7659, + "num_input_tokens_seen": 142729950, + "step": 4390 + }, + { + "epoch": 0.3959958515579204, + "flos": 19323768546720.0, + "grad_norm": 1.6923414644626178, + "learning_rate": 2.7519207400926253e-06, + "loss": 0.7951, + "num_input_tokens_seen": 142756405, + "step": 4391 + }, + { + "epoch": 0.3960860350813906, + "flos": 20527031377440.0, + "grad_norm": 1.640510337092448, + "learning_rate": 2.751379357571721e-06, + "loss": 0.7421, + "num_input_tokens_seen": 142782540, + "step": 4392 + }, + { + "epoch": 0.3961762186048609, + "flos": 24573744862080.0, + "grad_norm": 1.564988253125842, + "learning_rate": 2.7508379109389865e-06, + "loss": 0.8266, + "num_input_tokens_seen": 142812910, + "step": 4393 + }, + { + "epoch": 0.39626640212833114, + "flos": 23329626731520.0, + "grad_norm": 7.588861092621571, + "learning_rate": 2.750296400240622e-06, + "loss": 0.7733, + "num_input_tokens_seen": 142840480, + "step": 4394 + }, + { + "epoch": 0.39635658565180143, + "flos": 27633743769600.0, + "grad_norm": 1.421689031826775, + "learning_rate": 2.7497548255228305e-06, + "loss": 0.793, + "num_input_tokens_seen": 142871475, + "step": 4395 + }, + { + "epoch": 0.39644676917527166, + "flos": 22422429738240.0, + "grad_norm": 1.7512293993155492, + "learning_rate": 2.749213186831824e-06, + "loss": 0.7097, + "num_input_tokens_seen": 142899670, + "step": 4396 + }, + { + "epoch": 0.39653695269874195, + "flos": 23516619292320.0, + "grad_norm": 1.7940553518238267, + "learning_rate": 2.7486714842138173e-06, + "loss": 0.8091, + "num_input_tokens_seen": 142929245, + "step": 4397 + }, + { + "epoch": 0.3966271362222122, + "flos": 29815618169760.0, + "grad_norm": 13.971957322818591, + "learning_rate": 2.748129717715031e-06, + "loss": 0.7115, + "num_input_tokens_seen": 142958015, + "step": 4398 + }, + { + "epoch": 0.3967173197456825, + "flos": 29017828980960.0, + "grad_norm": 1.626038219669003, + "learning_rate": 2.747587887381692e-06, + "loss": 0.7781, + "num_input_tokens_seen": 142988100, + "step": 4399 + }, + { + "epoch": 0.3968075032691527, + "flos": 24207342371520.0, + "grad_norm": 1.5401747504082268, + "learning_rate": 2.7470459932600328e-06, + "loss": 0.7049, + "num_input_tokens_seen": 143017960, + "step": 4400 + }, + { + "epoch": 0.396897686792623, + "flos": 27122802317760.0, + "grad_norm": 1.78961692781487, + "learning_rate": 2.7465040353962897e-06, + "loss": 0.7597, + "num_input_tokens_seen": 143047315, + "step": 4401 + }, + { + "epoch": 0.3969878703160932, + "flos": 26466913037280.0, + "grad_norm": 1.8590178587467643, + "learning_rate": 2.745962013836706e-06, + "loss": 0.7437, + "num_input_tokens_seen": 143076125, + "step": 4402 + }, + { + "epoch": 0.3970780538395635, + "flos": 29163074167680.0, + "grad_norm": 1.9329323610410218, + "learning_rate": 2.74541992862753e-06, + "loss": 0.7349, + "num_input_tokens_seen": 143105015, + "step": 4403 + }, + { + "epoch": 0.39716823736303375, + "flos": 16521990927360.0, + "grad_norm": 2.0349311923829645, + "learning_rate": 2.744877779815016e-06, + "loss": 0.7594, + "num_input_tokens_seen": 143130345, + "step": 4404 + }, + { + "epoch": 0.39725842088650404, + "flos": 30762109332960.0, + "grad_norm": 1.382667189037613, + "learning_rate": 2.7443355674454234e-06, + "loss": 0.7531, + "num_input_tokens_seen": 143163395, + "step": 4405 + }, + { + "epoch": 0.39734860440997427, + "flos": 15391629463200.0, + "grad_norm": 2.1712209691032336, + "learning_rate": 2.743793291565015e-06, + "loss": 0.7967, + "num_input_tokens_seen": 143190850, + "step": 4406 + }, + { + "epoch": 0.39743878793344456, + "flos": 28283834567520.0, + "grad_norm": 2.688317995005976, + "learning_rate": 2.7432509522200617e-06, + "loss": 0.6803, + "num_input_tokens_seen": 143218755, + "step": 4407 + }, + { + "epoch": 0.39752897145691485, + "flos": 74630890517280.0, + "grad_norm": 2.4156966619484384, + "learning_rate": 2.7427085494568383e-06, + "loss": 0.7052, + "num_input_tokens_seen": 143254510, + "step": 4408 + }, + { + "epoch": 0.3976191549803851, + "flos": 70176727659840.0, + "grad_norm": 0.6590927857937298, + "learning_rate": 2.742166083321628e-06, + "loss": 0.6064, + "num_input_tokens_seen": 143346915, + "step": 4409 + }, + { + "epoch": 0.39770933850385537, + "flos": 21107751936000.0, + "grad_norm": 2.435904053757567, + "learning_rate": 2.7416235538607137e-06, + "loss": 0.7991, + "num_input_tokens_seen": 143374905, + "step": 4410 + }, + { + "epoch": 0.3977995220273256, + "flos": 38484264573120.0, + "grad_norm": 1.5592474080636538, + "learning_rate": 2.7410809611203894e-06, + "loss": 0.7388, + "num_input_tokens_seen": 143408620, + "step": 4411 + }, + { + "epoch": 0.3978897055507959, + "flos": 23913878417280.0, + "grad_norm": 1.7314000088810135, + "learning_rate": 2.7405383051469507e-06, + "loss": 0.719, + "num_input_tokens_seen": 143437530, + "step": 4412 + }, + { + "epoch": 0.3979798890742661, + "flos": 20966186555520.0, + "grad_norm": 2.0355317982884134, + "learning_rate": 2.7399955859867e-06, + "loss": 0.5741, + "num_input_tokens_seen": 143462670, + "step": 4413 + }, + { + "epoch": 0.3980700725977364, + "flos": 64970951922720.0, + "grad_norm": 0.6216651878859222, + "learning_rate": 2.7394528036859465e-06, + "loss": 0.5505, + "num_input_tokens_seen": 143552750, + "step": 4414 + }, + { + "epoch": 0.39816025612120665, + "flos": 23225050995840.0, + "grad_norm": 1.5660692266214802, + "learning_rate": 2.738909958291002e-06, + "loss": 0.758, + "num_input_tokens_seen": 143580720, + "step": 4415 + }, + { + "epoch": 0.39825043964467693, + "flos": 36814335208320.0, + "grad_norm": 1.8367707654434176, + "learning_rate": 2.7383670498481863e-06, + "loss": 0.6721, + "num_input_tokens_seen": 143610975, + "step": 4416 + }, + { + "epoch": 0.39834062316814717, + "flos": 64829869749120.0, + "grad_norm": 0.5771861572332619, + "learning_rate": 2.737824078403822e-06, + "loss": 0.5092, + "num_input_tokens_seen": 143702010, + "step": 4417 + }, + { + "epoch": 0.39843080669161746, + "flos": 17140630374720.0, + "grad_norm": 2.1520250364853823, + "learning_rate": 2.737281044004239e-06, + "loss": 0.7348, + "num_input_tokens_seen": 143728755, + "step": 4418 + }, + { + "epoch": 0.3985209902150877, + "flos": 30220088228160.0, + "grad_norm": 1.6844238925271116, + "learning_rate": 2.736737946695772e-06, + "loss": 0.7407, + "num_input_tokens_seen": 143758215, + "step": 4419 + }, + { + "epoch": 0.398611173738558, + "flos": 30216073894080.0, + "grad_norm": 1.9257618127426155, + "learning_rate": 2.736194786524761e-06, + "loss": 0.7543, + "num_input_tokens_seen": 143785255, + "step": 4420 + }, + { + "epoch": 0.3987013572620282, + "flos": 13862224725600.0, + "grad_norm": 1.9980009936959333, + "learning_rate": 2.7356515635375517e-06, + "loss": 0.7574, + "num_input_tokens_seen": 143811880, + "step": 4421 + }, + { + "epoch": 0.3987915407854985, + "flos": 21076003227360.0, + "grad_norm": 1.9581894787918441, + "learning_rate": 2.735108277780495e-06, + "loss": 0.6707, + "num_input_tokens_seen": 143839935, + "step": 4422 + }, + { + "epoch": 0.39888172430896873, + "flos": 25299710607360.0, + "grad_norm": 2.354122859130884, + "learning_rate": 2.7345649292999456e-06, + "loss": 0.7461, + "num_input_tokens_seen": 143869035, + "step": 4423 + }, + { + "epoch": 0.398971907832439, + "flos": 20819974955040.0, + "grad_norm": 1.9253455359731784, + "learning_rate": 2.734021518142267e-06, + "loss": 0.7631, + "num_input_tokens_seen": 143899265, + "step": 4424 + }, + { + "epoch": 0.39906209135590925, + "flos": 34810272438240.0, + "grad_norm": 1.692158802525303, + "learning_rate": 2.733478044353825e-06, + "loss": 0.596, + "num_input_tokens_seen": 143931125, + "step": 4425 + }, + { + "epoch": 0.39915227487937954, + "flos": 30949622270400.0, + "grad_norm": 1.8018845097518776, + "learning_rate": 2.7329345079809917e-06, + "loss": 0.7631, + "num_input_tokens_seen": 143963935, + "step": 4426 + }, + { + "epoch": 0.3992424584028498, + "flos": 23479964175360.0, + "grad_norm": 1.4625361055009203, + "learning_rate": 2.7323909090701447e-06, + "loss": 0.6967, + "num_input_tokens_seen": 143993625, + "step": 4427 + }, + { + "epoch": 0.39933264192632006, + "flos": 16776420900000.0, + "grad_norm": 1.875903039272898, + "learning_rate": 2.731847247667667e-06, + "loss": 0.8111, + "num_input_tokens_seen": 144020690, + "step": 4428 + }, + { + "epoch": 0.3994228254497903, + "flos": 25811581303200.0, + "grad_norm": 1.9900639005206635, + "learning_rate": 2.731303523819947e-06, + "loss": 0.7345, + "num_input_tokens_seen": 144052275, + "step": 4429 + }, + { + "epoch": 0.3995130089732606, + "flos": 16592773617600.0, + "grad_norm": 2.737288922398189, + "learning_rate": 2.7307597375733783e-06, + "loss": 0.8189, + "num_input_tokens_seen": 144077355, + "step": 4430 + }, + { + "epoch": 0.3996031924967309, + "flos": 26393788652160.0, + "grad_norm": 1.9333291229114504, + "learning_rate": 2.7302158889743587e-06, + "loss": 0.7559, + "num_input_tokens_seen": 144106920, + "step": 4431 + }, + { + "epoch": 0.3996933760202011, + "flos": 25009926459360.0, + "grad_norm": 1.7813770851174981, + "learning_rate": 2.7296719780692937e-06, + "loss": 0.7174, + "num_input_tokens_seen": 144135185, + "step": 4432 + }, + { + "epoch": 0.3997835595436714, + "flos": 29127757162080.0, + "grad_norm": 1.5947596773726154, + "learning_rate": 2.7291280049045916e-06, + "loss": 0.6664, + "num_input_tokens_seen": 144166270, + "step": 4433 + }, + { + "epoch": 0.39987374306714163, + "flos": 24385711548000.0, + "grad_norm": 1.5799008943162245, + "learning_rate": 2.7285839695266683e-06, + "loss": 0.7, + "num_input_tokens_seen": 144194980, + "step": 4434 + }, + { + "epoch": 0.3999639265906119, + "flos": 24460285553760.0, + "grad_norm": 1.9892418120559565, + "learning_rate": 2.7280398719819423e-06, + "loss": 0.7078, + "num_input_tokens_seen": 144224565, + "step": 4435 + }, + { + "epoch": 0.40005411011408215, + "flos": 21839627673120.0, + "grad_norm": 1.9228187662856702, + "learning_rate": 2.727495712316841e-06, + "loss": 0.8225, + "num_input_tokens_seen": 144250350, + "step": 4436 + }, + { + "epoch": 0.40014429363755244, + "flos": 29090581668480.0, + "grad_norm": 2.465106738489678, + "learning_rate": 2.7269514905777945e-06, + "loss": 0.7046, + "num_input_tokens_seen": 144276135, + "step": 4437 + }, + { + "epoch": 0.4002344771610227, + "flos": 25880840033280.0, + "grad_norm": 2.167651893898967, + "learning_rate": 2.7264072068112377e-06, + "loss": 0.8772, + "num_input_tokens_seen": 144305595, + "step": 4438 + }, + { + "epoch": 0.40032466068449296, + "flos": 33455817259200.0, + "grad_norm": 2.0755948604335583, + "learning_rate": 2.7258628610636133e-06, + "loss": 0.7196, + "num_input_tokens_seen": 144334885, + "step": 4439 + }, + { + "epoch": 0.4004148442079632, + "flos": 26759038880160.0, + "grad_norm": 1.8561974951514373, + "learning_rate": 2.7253184533813667e-06, + "loss": 0.8375, + "num_input_tokens_seen": 144363620, + "step": 4440 + }, + { + "epoch": 0.4005050277314335, + "flos": 19393138786080.0, + "grad_norm": 1.7729704657825611, + "learning_rate": 2.72477398381095e-06, + "loss": 0.7664, + "num_input_tokens_seen": 144388175, + "step": 4441 + }, + { + "epoch": 0.4005952112549037, + "flos": 33681509273760.0, + "grad_norm": 1.6175767365638285, + "learning_rate": 2.724229452398821e-06, + "loss": 0.7629, + "num_input_tokens_seen": 144419325, + "step": 4442 + }, + { + "epoch": 0.400685394778374, + "flos": 20310371614560.0, + "grad_norm": 2.122006608030509, + "learning_rate": 2.7236848591914422e-06, + "loss": 0.7782, + "num_input_tokens_seen": 144447525, + "step": 4443 + }, + { + "epoch": 0.40077557830184424, + "flos": 31712986527840.0, + "grad_norm": 1.873794676876009, + "learning_rate": 2.7231402042352803e-06, + "loss": 0.7404, + "num_input_tokens_seen": 144477320, + "step": 4444 + }, + { + "epoch": 0.4008657618253145, + "flos": 24242139000480.0, + "grad_norm": 1.4207491766802893, + "learning_rate": 2.722595487576809e-06, + "loss": 0.7863, + "num_input_tokens_seen": 144507215, + "step": 4445 + }, + { + "epoch": 0.40095594534878476, + "flos": 18014591868960.0, + "grad_norm": 1.78817453795117, + "learning_rate": 2.722050709262506e-06, + "loss": 0.8047, + "num_input_tokens_seen": 144533145, + "step": 4446 + }, + { + "epoch": 0.40104612887225505, + "flos": 23587104624480.0, + "grad_norm": 1.4362857500080788, + "learning_rate": 2.7215058693388557e-06, + "loss": 0.7363, + "num_input_tokens_seen": 144564000, + "step": 4447 + }, + { + "epoch": 0.4011363123957253, + "flos": 23225199674880.0, + "grad_norm": 1.7894402650688723, + "learning_rate": 2.720960967852346e-06, + "loss": 0.6565, + "num_input_tokens_seen": 144591425, + "step": 4448 + }, + { + "epoch": 0.40122649591919557, + "flos": 17505471735360.0, + "grad_norm": 1.5304050060353858, + "learning_rate": 2.720416004849471e-06, + "loss": 0.689, + "num_input_tokens_seen": 144619400, + "step": 4449 + }, + { + "epoch": 0.4013166794426658, + "flos": 24536495028960.0, + "grad_norm": 1.9342273637330352, + "learning_rate": 2.7198709803767304e-06, + "loss": 0.7077, + "num_input_tokens_seen": 144648825, + "step": 4450 + }, + { + "epoch": 0.4014068629661361, + "flos": 39109037030880.0, + "grad_norm": 1.4178793802813332, + "learning_rate": 2.7193258944806286e-06, + "loss": 0.7263, + "num_input_tokens_seen": 144681860, + "step": 4451 + }, + { + "epoch": 0.4014970464896063, + "flos": 43480963178400.0, + "grad_norm": 1.8504417021009187, + "learning_rate": 2.718780747207675e-06, + "loss": 0.7593, + "num_input_tokens_seen": 144711500, + "step": 4452 + }, + { + "epoch": 0.4015872300130766, + "flos": 22716376899360.0, + "grad_norm": 1.9530903317235664, + "learning_rate": 2.7182355386043847e-06, + "loss": 0.8172, + "num_input_tokens_seen": 144739195, + "step": 4453 + }, + { + "epoch": 0.40167741353654685, + "flos": 21110985705120.0, + "grad_norm": 1.5546135812947204, + "learning_rate": 2.717690268717278e-06, + "loss": 0.8231, + "num_input_tokens_seen": 144767550, + "step": 4454 + }, + { + "epoch": 0.40176759706001713, + "flos": 26173226064480.0, + "grad_norm": 1.4614208054386142, + "learning_rate": 2.7171449375928803e-06, + "loss": 0.7792, + "num_input_tokens_seen": 144797305, + "step": 4455 + }, + { + "epoch": 0.4018577805834874, + "flos": 16111833895680.0, + "grad_norm": 2.1252620759806455, + "learning_rate": 2.716599545277722e-06, + "loss": 0.7843, + "num_input_tokens_seen": 144820420, + "step": 4456 + }, + { + "epoch": 0.40194796410695766, + "flos": 22058963658720.0, + "grad_norm": 1.688590015595388, + "learning_rate": 2.7160540918183394e-06, + "loss": 0.7181, + "num_input_tokens_seen": 144847070, + "step": 4457 + }, + { + "epoch": 0.40203814763042794, + "flos": 24169609331520.0, + "grad_norm": 2.2927496576347424, + "learning_rate": 2.715508577261273e-06, + "loss": 0.6775, + "num_input_tokens_seen": 144876640, + "step": 4458 + }, + { + "epoch": 0.4021283311538982, + "flos": 20707370551200.0, + "grad_norm": 1.8344836383213667, + "learning_rate": 2.7149630016530702e-06, + "loss": 0.7377, + "num_input_tokens_seen": 144903810, + "step": 4459 + }, + { + "epoch": 0.40221851467736847, + "flos": 22820395088640.0, + "grad_norm": 1.433114618795321, + "learning_rate": 2.7144173650402815e-06, + "loss": 0.7756, + "num_input_tokens_seen": 144931710, + "step": 4460 + }, + { + "epoch": 0.4023086982008387, + "flos": 32513749297440.0, + "grad_norm": 10.396182944331622, + "learning_rate": 2.7138716674694636e-06, + "loss": 0.7798, + "num_input_tokens_seen": 144960205, + "step": 4461 + }, + { + "epoch": 0.402398881724309, + "flos": 32440104535680.0, + "grad_norm": 1.9131602126342016, + "learning_rate": 2.7133259089871795e-06, + "loss": 0.7054, + "num_input_tokens_seen": 144990570, + "step": 4462 + }, + { + "epoch": 0.4024890652477792, + "flos": 16149529765920.0, + "grad_norm": 1.7516209043894435, + "learning_rate": 2.712780089639995e-06, + "loss": 0.7541, + "num_input_tokens_seen": 145015530, + "step": 4463 + }, + { + "epoch": 0.4025792487712495, + "flos": 26941831258080.0, + "grad_norm": 1.7313799730118906, + "learning_rate": 2.712234209474483e-06, + "loss": 0.6364, + "num_input_tokens_seen": 145047955, + "step": 4464 + }, + { + "epoch": 0.40266943229471974, + "flos": 25956417622560.0, + "grad_norm": 1.7236485812841014, + "learning_rate": 2.7116882685372218e-06, + "loss": 0.7967, + "num_input_tokens_seen": 145077725, + "step": 4465 + }, + { + "epoch": 0.40275961581819003, + "flos": 41222693454240.0, + "grad_norm": 2.4452877348111954, + "learning_rate": 2.7111422668747927e-06, + "loss": 0.7228, + "num_input_tokens_seen": 145108515, + "step": 4466 + }, + { + "epoch": 0.40284979934166026, + "flos": 27083247959520.0, + "grad_norm": 1.7641553585727388, + "learning_rate": 2.7105962045337846e-06, + "loss": 0.7555, + "num_input_tokens_seen": 145138600, + "step": 4467 + }, + { + "epoch": 0.40293998286513055, + "flos": 23370705049920.0, + "grad_norm": 2.119202615560759, + "learning_rate": 2.7100500815607898e-06, + "loss": 0.7511, + "num_input_tokens_seen": 145167705, + "step": 4468 + }, + { + "epoch": 0.4030301663886008, + "flos": 26868112156800.0, + "grad_norm": 1.6354882615407837, + "learning_rate": 2.709503898002407e-06, + "loss": 0.7765, + "num_input_tokens_seen": 145199665, + "step": 4469 + }, + { + "epoch": 0.4031203499120711, + "flos": 22853407569120.0, + "grad_norm": 1.7149118049342955, + "learning_rate": 2.708957653905239e-06, + "loss": 0.753, + "num_input_tokens_seen": 145225665, + "step": 4470 + }, + { + "epoch": 0.4032105334355413, + "flos": 21585829586400.0, + "grad_norm": 1.5989172699301653, + "learning_rate": 2.7084113493158956e-06, + "loss": 0.7678, + "num_input_tokens_seen": 145255410, + "step": 4471 + }, + { + "epoch": 0.4033007169590116, + "flos": 21655943220960.0, + "grad_norm": 1.5554507121827374, + "learning_rate": 2.7078649842809888e-06, + "loss": 0.7674, + "num_input_tokens_seen": 145285635, + "step": 4472 + }, + { + "epoch": 0.40339090048248183, + "flos": 20416285461600.0, + "grad_norm": 6.2388077276666385, + "learning_rate": 2.707318558847139e-06, + "loss": 0.8054, + "num_input_tokens_seen": 145310610, + "step": 4473 + }, + { + "epoch": 0.4034810840059521, + "flos": 58420942690560.0, + "grad_norm": 0.7165589730350221, + "learning_rate": 2.7067720730609697e-06, + "loss": 0.579, + "num_input_tokens_seen": 145402360, + "step": 4474 + }, + { + "epoch": 0.40357126752942235, + "flos": 27228567485760.0, + "grad_norm": 1.691626342831536, + "learning_rate": 2.70622552696911e-06, + "loss": 0.7294, + "num_input_tokens_seen": 145431650, + "step": 4475 + }, + { + "epoch": 0.40366145105289264, + "flos": 26063149204320.0, + "grad_norm": 1.853576787777584, + "learning_rate": 2.7056789206181943e-06, + "loss": 0.7074, + "num_input_tokens_seen": 145460535, + "step": 4476 + }, + { + "epoch": 0.40375163457636287, + "flos": 17978828826240.0, + "grad_norm": 1.6546970500436313, + "learning_rate": 2.7051322540548615e-06, + "loss": 0.8122, + "num_input_tokens_seen": 145487490, + "step": 4477 + }, + { + "epoch": 0.40384181809983316, + "flos": 24536829556800.0, + "grad_norm": 2.600857017561627, + "learning_rate": 2.704585527325757e-06, + "loss": 0.727, + "num_input_tokens_seen": 145517005, + "step": 4478 + }, + { + "epoch": 0.40393200162330345, + "flos": 25301346076800.0, + "grad_norm": 1.9941473203242157, + "learning_rate": 2.7040387404775303e-06, + "loss": 0.7125, + "num_input_tokens_seen": 145547260, + "step": 4479 + }, + { + "epoch": 0.4040221851467737, + "flos": 27305297337600.0, + "grad_norm": 2.41503067107696, + "learning_rate": 2.703491893556837e-06, + "loss": 0.7871, + "num_input_tokens_seen": 145577265, + "step": 4480 + }, + { + "epoch": 0.40411236867024397, + "flos": 22022977597440.0, + "grad_norm": 1.4453863176620103, + "learning_rate": 2.702944986610335e-06, + "loss": 0.7265, + "num_input_tokens_seen": 145605605, + "step": 4481 + }, + { + "epoch": 0.4042025521937142, + "flos": 29634461261280.0, + "grad_norm": 1.6751628170306347, + "learning_rate": 2.7023980196846917e-06, + "loss": 0.767, + "num_input_tokens_seen": 145634770, + "step": 4482 + }, + { + "epoch": 0.4042927357171845, + "flos": 61163980621920.0, + "grad_norm": 0.7040983456605607, + "learning_rate": 2.7018509928265763e-06, + "loss": 0.5983, + "num_input_tokens_seen": 145724470, + "step": 4483 + }, + { + "epoch": 0.4043829192406547, + "flos": 20638706537280.0, + "grad_norm": 1.6791441282490398, + "learning_rate": 2.7013039060826635e-06, + "loss": 0.7827, + "num_input_tokens_seen": 145754585, + "step": 4484 + }, + { + "epoch": 0.404473102764125, + "flos": 23917000677120.0, + "grad_norm": 2.09773115027911, + "learning_rate": 2.7007567594996347e-06, + "loss": 0.8282, + "num_input_tokens_seen": 145780590, + "step": 4485 + }, + { + "epoch": 0.40456328628759525, + "flos": 29592453698880.0, + "grad_norm": 1.640207073932004, + "learning_rate": 2.7002095531241757e-06, + "loss": 0.6601, + "num_input_tokens_seen": 145810025, + "step": 4486 + }, + { + "epoch": 0.40465346981106554, + "flos": 21911413946880.0, + "grad_norm": 1.8151595893035144, + "learning_rate": 2.6996622870029767e-06, + "loss": 0.7286, + "num_input_tokens_seen": 145839170, + "step": 4487 + }, + { + "epoch": 0.40474365333453577, + "flos": 20019063506400.0, + "grad_norm": 1.9620372226578915, + "learning_rate": 2.6991149611827335e-06, + "loss": 0.8674, + "num_input_tokens_seen": 145866000, + "step": 4488 + }, + { + "epoch": 0.40483383685800606, + "flos": 31166839579680.0, + "grad_norm": 2.064860062926244, + "learning_rate": 2.6985675757101466e-06, + "loss": 0.797, + "num_input_tokens_seen": 145894970, + "step": 4489 + }, + { + "epoch": 0.4049240203814763, + "flos": 20346729373440.0, + "grad_norm": 1.8901761996167272, + "learning_rate": 2.698020130631922e-06, + "loss": 0.7144, + "num_input_tokens_seen": 145922610, + "step": 4490 + }, + { + "epoch": 0.4050142039049466, + "flos": 25848719627040.0, + "grad_norm": 1.9187116563177808, + "learning_rate": 2.6974726259947713e-06, + "loss": 0.6901, + "num_input_tokens_seen": 145954630, + "step": 4491 + }, + { + "epoch": 0.4051043874284168, + "flos": 20820718350240.0, + "grad_norm": 2.181038167991192, + "learning_rate": 2.6969250618454106e-06, + "loss": 0.732, + "num_input_tokens_seen": 145979445, + "step": 4492 + }, + { + "epoch": 0.4051945709518871, + "flos": 25445773528800.0, + "grad_norm": 1.5055446666982226, + "learning_rate": 2.696377438230561e-06, + "loss": 0.8296, + "num_input_tokens_seen": 146009475, + "step": 4493 + }, + { + "epoch": 0.40528475447535733, + "flos": 48286320360960.0, + "grad_norm": 1.7927003214906638, + "learning_rate": 2.6958297551969484e-06, + "loss": 0.6969, + "num_input_tokens_seen": 146040450, + "step": 4494 + }, + { + "epoch": 0.4053749379988276, + "flos": 22456334292960.0, + "grad_norm": 2.15414879187049, + "learning_rate": 2.695282012791304e-06, + "loss": 0.7246, + "num_input_tokens_seen": 146066120, + "step": 4495 + }, + { + "epoch": 0.40546512152229786, + "flos": 27487866696960.0, + "grad_norm": 12.837361549886332, + "learning_rate": 2.6947342110603646e-06, + "loss": 0.813, + "num_input_tokens_seen": 146094790, + "step": 4496 + }, + { + "epoch": 0.40555530504576814, + "flos": 22788274682400.0, + "grad_norm": 2.4155185098604135, + "learning_rate": 2.6941863500508717e-06, + "loss": 0.8096, + "num_input_tokens_seen": 146124130, + "step": 4497 + }, + { + "epoch": 0.4056454885692384, + "flos": 23480596061280.0, + "grad_norm": 1.723377725458622, + "learning_rate": 2.693638429809572e-06, + "loss": 0.8088, + "num_input_tokens_seen": 146151290, + "step": 4498 + }, + { + "epoch": 0.40573567209270867, + "flos": 20820272313120.0, + "grad_norm": 1.827856060362737, + "learning_rate": 2.6930904503832167e-06, + "loss": 0.7413, + "num_input_tokens_seen": 146177815, + "step": 4499 + }, + { + "epoch": 0.4058258556161789, + "flos": 23403866209440.0, + "grad_norm": 9.404110078947031, + "learning_rate": 2.692542411818562e-06, + "loss": 0.7425, + "num_input_tokens_seen": 146206880, + "step": 4500 + }, + { + "epoch": 0.4059160391396492, + "flos": 21075073983360.0, + "grad_norm": 1.7348625941104747, + "learning_rate": 2.69199431416237e-06, + "loss": 0.7683, + "num_input_tokens_seen": 146235800, + "step": 4501 + }, + { + "epoch": 0.4060062226631194, + "flos": 23043373710720.0, + "grad_norm": 1.6773337972884959, + "learning_rate": 2.691446157461408e-06, + "loss": 0.713, + "num_input_tokens_seen": 146265325, + "step": 4502 + }, + { + "epoch": 0.4060964061865897, + "flos": 24390952484160.0, + "grad_norm": 1.7115636925086353, + "learning_rate": 2.690897941762447e-06, + "loss": 0.8169, + "num_input_tokens_seen": 146295630, + "step": 4503 + }, + { + "epoch": 0.40618658971006, + "flos": 21585792416640.0, + "grad_norm": 1.6170376325815476, + "learning_rate": 2.6903496671122642e-06, + "loss": 0.734, + "num_input_tokens_seen": 146324010, + "step": 4504 + }, + { + "epoch": 0.40627677323353023, + "flos": 63674673151680.0, + "grad_norm": 0.6351980057207626, + "learning_rate": 2.689801333557641e-06, + "loss": 0.5282, + "num_input_tokens_seen": 146410010, + "step": 4505 + }, + { + "epoch": 0.4063669567570005, + "flos": 14335358797920.0, + "grad_norm": 2.190721265389135, + "learning_rate": 2.689252941145365e-06, + "loss": 0.7576, + "num_input_tokens_seen": 146433955, + "step": 4506 + }, + { + "epoch": 0.40645714028047075, + "flos": 28214241309600.0, + "grad_norm": 1.7433601846079791, + "learning_rate": 2.6887044899222277e-06, + "loss": 0.7534, + "num_input_tokens_seen": 146465385, + "step": 4507 + }, + { + "epoch": 0.40654732380394104, + "flos": 22456222783680.0, + "grad_norm": 1.7114839421509986, + "learning_rate": 2.688155979935025e-06, + "loss": 0.8032, + "num_input_tokens_seen": 146494165, + "step": 4508 + }, + { + "epoch": 0.4066375073274113, + "flos": 22674369336960.0, + "grad_norm": 1.6688900978057475, + "learning_rate": 2.68760741123056e-06, + "loss": 0.7165, + "num_input_tokens_seen": 146523530, + "step": 4509 + }, + { + "epoch": 0.40672769085088156, + "flos": 23077426944480.0, + "grad_norm": 2.1793881009022344, + "learning_rate": 2.6870587838556394e-06, + "loss": 0.7747, + "num_input_tokens_seen": 146549960, + "step": 4510 + }, + { + "epoch": 0.4068178743743518, + "flos": 22635967241280.0, + "grad_norm": 1.9340406729791255, + "learning_rate": 2.686510097857075e-06, + "loss": 0.7765, + "num_input_tokens_seen": 146577410, + "step": 4511 + }, + { + "epoch": 0.4069080578978221, + "flos": 32299431229440.0, + "grad_norm": 1.4241628998955291, + "learning_rate": 2.685961353281683e-06, + "loss": 0.7523, + "num_input_tokens_seen": 146610425, + "step": 4512 + }, + { + "epoch": 0.4069982414212923, + "flos": 12003853179360.0, + "grad_norm": 2.0448000342388477, + "learning_rate": 2.6854125501762863e-06, + "loss": 0.735, + "num_input_tokens_seen": 146637245, + "step": 4513 + }, + { + "epoch": 0.4070884249447626, + "flos": 21731780998560.0, + "grad_norm": 1.946235774679756, + "learning_rate": 2.684863688587712e-06, + "loss": 0.7598, + "num_input_tokens_seen": 146668135, + "step": 4514 + }, + { + "epoch": 0.40717860846823284, + "flos": 28178738455200.0, + "grad_norm": 7.024882913611527, + "learning_rate": 2.6843147685627916e-06, + "loss": 0.7288, + "num_input_tokens_seen": 146700610, + "step": 4515 + }, + { + "epoch": 0.4072687919917031, + "flos": 40748853156480.0, + "grad_norm": 6.0359425054179345, + "learning_rate": 2.683765790148361e-06, + "loss": 0.73, + "num_input_tokens_seen": 146731035, + "step": 4516 + }, + { + "epoch": 0.40735897551517336, + "flos": 26431707540960.0, + "grad_norm": 1.9081730757253699, + "learning_rate": 2.6832167533912637e-06, + "loss": 0.8141, + "num_input_tokens_seen": 146759080, + "step": 4517 + }, + { + "epoch": 0.40744915903864365, + "flos": 18379210211040.0, + "grad_norm": 2.664314772667679, + "learning_rate": 2.682667658338345e-06, + "loss": 0.8314, + "num_input_tokens_seen": 146784870, + "step": 4518 + }, + { + "epoch": 0.4075393425621139, + "flos": 23877669337440.0, + "grad_norm": 1.6869890357782227, + "learning_rate": 2.682118505036458e-06, + "loss": 0.7305, + "num_input_tokens_seen": 146812850, + "step": 4519 + }, + { + "epoch": 0.40762952608558417, + "flos": 66434591887680.0, + "grad_norm": 0.6129673683059527, + "learning_rate": 2.681569293532459e-06, + "loss": 0.5487, + "num_input_tokens_seen": 146905535, + "step": 4520 + }, + { + "epoch": 0.4077197096090544, + "flos": 68627431367040.0, + "grad_norm": 0.6508891626395588, + "learning_rate": 2.6810200238732102e-06, + "loss": 0.577, + "num_input_tokens_seen": 147005110, + "step": 4521 + }, + { + "epoch": 0.4078098931325247, + "flos": 18378838513440.0, + "grad_norm": 2.7917652017844428, + "learning_rate": 2.6804706961055776e-06, + "loss": 0.8011, + "num_input_tokens_seen": 147033325, + "step": 4522 + }, + { + "epoch": 0.4079000766559949, + "flos": 27268419202080.0, + "grad_norm": 1.8711086583190653, + "learning_rate": 2.6799213102764326e-06, + "loss": 0.7894, + "num_input_tokens_seen": 147063630, + "step": 4523 + }, + { + "epoch": 0.4079902601794652, + "flos": 15465348564480.0, + "grad_norm": 1.8253190767587375, + "learning_rate": 2.679371866432653e-06, + "loss": 0.8489, + "num_input_tokens_seen": 147089170, + "step": 4524 + }, + { + "epoch": 0.40808044370293545, + "flos": 21034330192800.0, + "grad_norm": 1.6816644434288812, + "learning_rate": 2.6788223646211194e-06, + "loss": 0.728, + "num_input_tokens_seen": 147116845, + "step": 4525 + }, + { + "epoch": 0.40817062722640574, + "flos": 21075928887840.0, + "grad_norm": 1.758221659195037, + "learning_rate": 2.6782728048887183e-06, + "loss": 0.6521, + "num_input_tokens_seen": 147145890, + "step": 4526 + }, + { + "epoch": 0.408260810749876, + "flos": 42278518082400.0, + "grad_norm": 1.590216140811198, + "learning_rate": 2.6777231872823416e-06, + "loss": 0.7421, + "num_input_tokens_seen": 147179850, + "step": 4527 + }, + { + "epoch": 0.40835099427334626, + "flos": 26905287650400.0, + "grad_norm": 2.326879488202689, + "learning_rate": 2.6771735118488864e-06, + "loss": 0.7582, + "num_input_tokens_seen": 147208615, + "step": 4528 + }, + { + "epoch": 0.40844117779681655, + "flos": 15793274619840.0, + "grad_norm": 1.6743209629658757, + "learning_rate": 2.6766237786352523e-06, + "loss": 0.7942, + "num_input_tokens_seen": 147235315, + "step": 4529 + }, + { + "epoch": 0.4085313613202868, + "flos": 41585081610720.0, + "grad_norm": 1.4638609984650688, + "learning_rate": 2.676073987688347e-06, + "loss": 0.6824, + "num_input_tokens_seen": 147269170, + "step": 4530 + }, + { + "epoch": 0.40862154484375707, + "flos": 40129767672000.0, + "grad_norm": 1.720340478376809, + "learning_rate": 2.6755241390550818e-06, + "loss": 0.7598, + "num_input_tokens_seen": 147301120, + "step": 4531 + }, + { + "epoch": 0.4087117283672273, + "flos": 27522180119040.0, + "grad_norm": 2.50344658737238, + "learning_rate": 2.6749742327823716e-06, + "loss": 0.761, + "num_input_tokens_seen": 147332000, + "step": 4532 + }, + { + "epoch": 0.4088019118906976, + "flos": 23477325122400.0, + "grad_norm": 2.077785951878227, + "learning_rate": 2.674424268917138e-06, + "loss": 0.7705, + "num_input_tokens_seen": 147359035, + "step": 4533 + }, + { + "epoch": 0.4088920954141678, + "flos": 17500602496800.0, + "grad_norm": 1.8477528972228827, + "learning_rate": 2.6738742475063074e-06, + "loss": 0.8195, + "num_input_tokens_seen": 147384525, + "step": 4534 + }, + { + "epoch": 0.4089822789376381, + "flos": 12841902951840.0, + "grad_norm": 3.7863917425807503, + "learning_rate": 2.6733241685968104e-06, + "loss": 0.7988, + "num_input_tokens_seen": 147410690, + "step": 4535 + }, + { + "epoch": 0.40907246246110834, + "flos": 12987036629280.0, + "grad_norm": 2.209569501071627, + "learning_rate": 2.6727740322355826e-06, + "loss": 0.7733, + "num_input_tokens_seen": 147436705, + "step": 4536 + }, + { + "epoch": 0.40916264598457863, + "flos": 23115494512320.0, + "grad_norm": 2.7210335623267032, + "learning_rate": 2.6722238384695644e-06, + "loss": 0.7437, + "num_input_tokens_seen": 147464385, + "step": 4537 + }, + { + "epoch": 0.40925282950804887, + "flos": 22678346501280.0, + "grad_norm": 3.56325027863856, + "learning_rate": 2.671673587345702e-06, + "loss": 0.7676, + "num_input_tokens_seen": 147492910, + "step": 4538 + }, + { + "epoch": 0.40934301303151915, + "flos": 64938013781760.0, + "grad_norm": 0.6502865544816421, + "learning_rate": 2.6711232789109455e-06, + "loss": 0.5673, + "num_input_tokens_seen": 147579150, + "step": 4539 + }, + { + "epoch": 0.4094331965549894, + "flos": 27597014313120.0, + "grad_norm": 2.0768280571342794, + "learning_rate": 2.6705729132122497e-06, + "loss": 0.7625, + "num_input_tokens_seen": 147611535, + "step": 4540 + }, + { + "epoch": 0.4095233800784597, + "flos": 23297580664800.0, + "grad_norm": 2.1813640094681537, + "learning_rate": 2.670022490296576e-06, + "loss": 0.832, + "num_input_tokens_seen": 147636245, + "step": 4541 + }, + { + "epoch": 0.4096135636019299, + "flos": 21622336024320.0, + "grad_norm": 1.9016586417325017, + "learning_rate": 2.669472010210889e-06, + "loss": 0.8163, + "num_input_tokens_seen": 147664525, + "step": 4542 + }, + { + "epoch": 0.4097037471254002, + "flos": 16484629584960.0, + "grad_norm": 1.9995053991047873, + "learning_rate": 2.668921473002159e-06, + "loss": 0.7339, + "num_input_tokens_seen": 147692040, + "step": 4543 + }, + { + "epoch": 0.40979393064887043, + "flos": 38380246383840.0, + "grad_norm": 1.5466008569566796, + "learning_rate": 2.6683708787173596e-06, + "loss": 0.8123, + "num_input_tokens_seen": 147723475, + "step": 4544 + }, + { + "epoch": 0.4098841141723407, + "flos": 28177586192640.0, + "grad_norm": 1.7121022061740316, + "learning_rate": 2.6678202274034718e-06, + "loss": 0.7768, + "num_input_tokens_seen": 147754080, + "step": 4545 + }, + { + "epoch": 0.40997429769581095, + "flos": 24900853182720.0, + "grad_norm": 1.934482562844573, + "learning_rate": 2.66726951910748e-06, + "loss": 0.8455, + "num_input_tokens_seen": 147784340, + "step": 4546 + }, + { + "epoch": 0.41006448121928124, + "flos": 23037835416480.0, + "grad_norm": 1.9062074124941473, + "learning_rate": 2.6667187538763737e-06, + "loss": 0.7608, + "num_input_tokens_seen": 147813545, + "step": 4547 + }, + { + "epoch": 0.4101546647427515, + "flos": 24208754822400.0, + "grad_norm": 1.564138621365227, + "learning_rate": 2.6661679317571473e-06, + "loss": 0.7408, + "num_input_tokens_seen": 147842260, + "step": 4548 + }, + { + "epoch": 0.41024484826622176, + "flos": 26795619657600.0, + "grad_norm": 1.7035001289136975, + "learning_rate": 2.665617052796799e-06, + "loss": 0.7371, + "num_input_tokens_seen": 147873255, + "step": 4549 + }, + { + "epoch": 0.41033503178969205, + "flos": 20747928492960.0, + "grad_norm": 1.9893896464660483, + "learning_rate": 2.6650661170423346e-06, + "loss": 0.6409, + "num_input_tokens_seen": 147900460, + "step": 4550 + }, + { + "epoch": 0.4104252153131623, + "flos": 31457813160000.0, + "grad_norm": 1.8015856487444506, + "learning_rate": 2.6645151245407614e-06, + "loss": 0.8362, + "num_input_tokens_seen": 147928620, + "step": 4551 + }, + { + "epoch": 0.4105153988366326, + "flos": 20383310150880.0, + "grad_norm": 1.7137980477061856, + "learning_rate": 2.6639640753390936e-06, + "loss": 0.7529, + "num_input_tokens_seen": 147956200, + "step": 4552 + }, + { + "epoch": 0.4106055823601028, + "flos": 29599404444000.0, + "grad_norm": 1.469427074449094, + "learning_rate": 2.66341296948435e-06, + "loss": 0.661, + "num_input_tokens_seen": 147986780, + "step": 4553 + }, + { + "epoch": 0.4106957658835731, + "flos": 18234819928800.0, + "grad_norm": 1.3931286166500554, + "learning_rate": 2.6628618070235534e-06, + "loss": 0.7646, + "num_input_tokens_seen": 148016140, + "step": 4554 + }, + { + "epoch": 0.4107859494070433, + "flos": 26103149599680.0, + "grad_norm": 1.7705819418990825, + "learning_rate": 2.662310588003733e-06, + "loss": 0.7619, + "num_input_tokens_seen": 148045215, + "step": 4555 + }, + { + "epoch": 0.4108761329305136, + "flos": 11749125848640.0, + "grad_norm": 2.4066395630019164, + "learning_rate": 2.6617593124719205e-06, + "loss": 0.7342, + "num_input_tokens_seen": 148072925, + "step": 4556 + }, + { + "epoch": 0.41096631645398385, + "flos": 25408077658560.0, + "grad_norm": 1.8905780250451367, + "learning_rate": 2.661207980475155e-06, + "loss": 0.7621, + "num_input_tokens_seen": 148105735, + "step": 4557 + }, + { + "epoch": 0.41105649997745414, + "flos": 22641654214560.0, + "grad_norm": 1.6850201080888312, + "learning_rate": 2.6606565920604793e-06, + "loss": 0.7136, + "num_input_tokens_seen": 148133300, + "step": 4558 + }, + { + "epoch": 0.41114668350092437, + "flos": 22714667090400.0, + "grad_norm": 1.571788341036978, + "learning_rate": 2.66010514727494e-06, + "loss": 0.8344, + "num_input_tokens_seen": 148160460, + "step": 4559 + }, + { + "epoch": 0.41123686702439466, + "flos": 22529384338560.0, + "grad_norm": 1.8231580374636975, + "learning_rate": 2.659553646165589e-06, + "loss": 0.6537, + "num_input_tokens_seen": 148189330, + "step": 4560 + }, + { + "epoch": 0.4113270505478649, + "flos": 23079508451040.0, + "grad_norm": 1.8172598197105978, + "learning_rate": 2.659002088779485e-06, + "loss": 0.7266, + "num_input_tokens_seen": 148218455, + "step": 4561 + }, + { + "epoch": 0.4114172340713352, + "flos": 70394279496960.0, + "grad_norm": 0.5846384270696016, + "learning_rate": 2.6584504751636888e-06, + "loss": 0.5273, + "num_input_tokens_seen": 148308060, + "step": 4562 + }, + { + "epoch": 0.4115074175948054, + "flos": 24318162626880.0, + "grad_norm": 1.7433732449550141, + "learning_rate": 2.657898805365268e-06, + "loss": 0.7221, + "num_input_tokens_seen": 148339080, + "step": 4563 + }, + { + "epoch": 0.4115976011182757, + "flos": 26320181060160.0, + "grad_norm": 3.8892130913397676, + "learning_rate": 2.657347079431293e-06, + "loss": 0.7239, + "num_input_tokens_seen": 148369895, + "step": 4564 + }, + { + "epoch": 0.41168778464174594, + "flos": 31271638333920.0, + "grad_norm": 1.434107051912944, + "learning_rate": 2.6567952974088403e-06, + "loss": 0.7302, + "num_input_tokens_seen": 148403230, + "step": 4565 + }, + { + "epoch": 0.4117779681652162, + "flos": 24936393206880.0, + "grad_norm": 1.5024955176922907, + "learning_rate": 2.6562434593449917e-06, + "loss": 0.8069, + "num_input_tokens_seen": 148433475, + "step": 4566 + }, + { + "epoch": 0.41186815168868646, + "flos": 22715484825120.0, + "grad_norm": 1.8962456251381867, + "learning_rate": 2.6556915652868325e-06, + "loss": 0.782, + "num_input_tokens_seen": 148460475, + "step": 4567 + }, + { + "epoch": 0.41195833521215675, + "flos": 28140447868800.0, + "grad_norm": 1.7909382810709096, + "learning_rate": 2.6551396152814534e-06, + "loss": 0.766, + "num_input_tokens_seen": 148491855, + "step": 4568 + }, + { + "epoch": 0.412048518735627, + "flos": 22168111274880.0, + "grad_norm": 5.080776238496333, + "learning_rate": 2.65458760937595e-06, + "loss": 0.7505, + "num_input_tokens_seen": 148520300, + "step": 4569 + }, + { + "epoch": 0.41213870225909727, + "flos": 38016148418400.0, + "grad_norm": 1.446748834303984, + "learning_rate": 2.654035547617423e-06, + "loss": 0.7081, + "num_input_tokens_seen": 148553930, + "step": 4570 + }, + { + "epoch": 0.4122288857825675, + "flos": 24755459316960.0, + "grad_norm": 2.002332638379662, + "learning_rate": 2.653483430052976e-06, + "loss": 0.6593, + "num_input_tokens_seen": 148582970, + "step": 4571 + }, + { + "epoch": 0.4123190693060378, + "flos": 21112063628160.0, + "grad_norm": 1.7651631924449718, + "learning_rate": 2.6529312567297197e-06, + "loss": 0.7391, + "num_input_tokens_seen": 148611295, + "step": 4572 + }, + { + "epoch": 0.412409252829508, + "flos": 20784137572800.0, + "grad_norm": 4.927938934084058, + "learning_rate": 2.652379027694768e-06, + "loss": 0.8372, + "num_input_tokens_seen": 148638850, + "step": 4573 + }, + { + "epoch": 0.4124994363529783, + "flos": 21731706659040.0, + "grad_norm": 1.9417626650522322, + "learning_rate": 2.651826742995241e-06, + "loss": 0.8157, + "num_input_tokens_seen": 148666920, + "step": 4574 + }, + { + "epoch": 0.4125896198764486, + "flos": 21109573254240.0, + "grad_norm": 1.5450191732408252, + "learning_rate": 2.651274402678262e-06, + "loss": 0.7531, + "num_input_tokens_seen": 148695270, + "step": 4575 + }, + { + "epoch": 0.41267980339991883, + "flos": 27816015770880.0, + "grad_norm": 1.775573116781139, + "learning_rate": 2.6507220067909597e-06, + "loss": 0.7823, + "num_input_tokens_seen": 148726630, + "step": 4576 + }, + { + "epoch": 0.4127699869233891, + "flos": 30182280848640.0, + "grad_norm": 1.848286495022318, + "learning_rate": 2.650169555380468e-06, + "loss": 0.7121, + "num_input_tokens_seen": 148756110, + "step": 4577 + }, + { + "epoch": 0.41286017044685935, + "flos": 20747482455840.0, + "grad_norm": 3.766509779191878, + "learning_rate": 2.6496170484939253e-06, + "loss": 0.7509, + "num_input_tokens_seen": 148786370, + "step": 4578 + }, + { + "epoch": 0.41295035397032964, + "flos": 22824520932000.0, + "grad_norm": 1.5822350641522998, + "learning_rate": 2.6490644861784735e-06, + "loss": 0.7685, + "num_input_tokens_seen": 148815715, + "step": 4579 + }, + { + "epoch": 0.4130405374937999, + "flos": 23115754700640.0, + "grad_norm": 2.7850654234127443, + "learning_rate": 2.648511868481261e-06, + "loss": 0.778, + "num_input_tokens_seen": 148845045, + "step": 4580 + }, + { + "epoch": 0.41313072101727016, + "flos": 20965889197440.0, + "grad_norm": 1.6487469857212187, + "learning_rate": 2.6479591954494397e-06, + "loss": 0.7562, + "num_input_tokens_seen": 148873900, + "step": 4581 + }, + { + "epoch": 0.4132209045407404, + "flos": 26613384826080.0, + "grad_norm": 1.8824496097583645, + "learning_rate": 2.647406467130167e-06, + "loss": 0.6552, + "num_input_tokens_seen": 148903550, + "step": 4582 + }, + { + "epoch": 0.4133110880642107, + "flos": 20019137845920.0, + "grad_norm": 1.630704550332418, + "learning_rate": 2.646853683570605e-06, + "loss": 0.7774, + "num_input_tokens_seen": 148932505, + "step": 4583 + }, + { + "epoch": 0.4134012715876809, + "flos": 27160349508960.0, + "grad_norm": 1.9094885534805945, + "learning_rate": 2.6463008448179196e-06, + "loss": 0.6952, + "num_input_tokens_seen": 148963710, + "step": 4584 + }, + { + "epoch": 0.4134914551111512, + "flos": 50878128774240.0, + "grad_norm": 1.8083638978241452, + "learning_rate": 2.6457479509192828e-06, + "loss": 0.7071, + "num_input_tokens_seen": 148998905, + "step": 4585 + }, + { + "epoch": 0.41358163863462144, + "flos": 24787616892960.0, + "grad_norm": 4.546016651423478, + "learning_rate": 2.645195001921871e-06, + "loss": 0.7491, + "num_input_tokens_seen": 149027365, + "step": 4586 + }, + { + "epoch": 0.41367182215809173, + "flos": 25301234567520.0, + "grad_norm": 1.9383193098803544, + "learning_rate": 2.644641997872863e-06, + "loss": 0.7368, + "num_input_tokens_seen": 149056750, + "step": 4587 + }, + { + "epoch": 0.41376200568156196, + "flos": 21586201284000.0, + "grad_norm": 2.62373466876141, + "learning_rate": 2.644088938819445e-06, + "loss": 0.7585, + "num_input_tokens_seen": 149083445, + "step": 4588 + }, + { + "epoch": 0.41385218920503225, + "flos": 16776197881440.0, + "grad_norm": 1.8726953329719407, + "learning_rate": 2.6435358248088077e-06, + "loss": 0.8394, + "num_input_tokens_seen": 149106645, + "step": 4589 + }, + { + "epoch": 0.4139423727285025, + "flos": 19217780360160.0, + "grad_norm": 1.816135181176603, + "learning_rate": 2.642982655888146e-06, + "loss": 0.7964, + "num_input_tokens_seen": 149133440, + "step": 4590 + }, + { + "epoch": 0.41403255625197277, + "flos": 18015595452480.0, + "grad_norm": 2.161271592005313, + "learning_rate": 2.6424294321046585e-06, + "loss": 0.7197, + "num_input_tokens_seen": 149161075, + "step": 4591 + }, + { + "epoch": 0.414122739775443, + "flos": 70218289185120.0, + "grad_norm": 0.7322946259903202, + "learning_rate": 2.641876153505549e-06, + "loss": 0.5735, + "num_input_tokens_seen": 149246800, + "step": 4592 + }, + { + "epoch": 0.4142129232989133, + "flos": 23910198611040.0, + "grad_norm": 2.450844062678789, + "learning_rate": 2.641322820138027e-06, + "loss": 0.7671, + "num_input_tokens_seen": 149274855, + "step": 4593 + }, + { + "epoch": 0.4143031068223835, + "flos": 27086704747200.0, + "grad_norm": 1.9123613544149074, + "learning_rate": 2.640769432049306e-06, + "loss": 0.7064, + "num_input_tokens_seen": 149303920, + "step": 4594 + }, + { + "epoch": 0.4143932903458538, + "flos": 20566251207840.0, + "grad_norm": 2.315763902294851, + "learning_rate": 2.6402159892866038e-06, + "loss": 0.6752, + "num_input_tokens_seen": 149333055, + "step": 4595 + }, + { + "epoch": 0.41448347386932405, + "flos": 23953544284800.0, + "grad_norm": 1.5753565006357955, + "learning_rate": 2.639662491897143e-06, + "loss": 0.738, + "num_input_tokens_seen": 149364430, + "step": 4596 + }, + { + "epoch": 0.41457365739279434, + "flos": 21841114463520.0, + "grad_norm": 1.9668360898484765, + "learning_rate": 2.639108939928152e-06, + "loss": 0.7711, + "num_input_tokens_seen": 149393655, + "step": 4597 + }, + { + "epoch": 0.4146638409162646, + "flos": 24645456796320.0, + "grad_norm": 1.839440433169211, + "learning_rate": 2.638555333426862e-06, + "loss": 0.7699, + "num_input_tokens_seen": 149420530, + "step": 4598 + }, + { + "epoch": 0.41475402443973486, + "flos": 21767469701760.0, + "grad_norm": 1.4816775256267842, + "learning_rate": 2.6380016724405093e-06, + "loss": 0.79, + "num_input_tokens_seen": 149448270, + "step": 4599 + }, + { + "epoch": 0.41484420796320515, + "flos": 18270805990080.0, + "grad_norm": 1.7053587377821797, + "learning_rate": 2.637447957016336e-06, + "loss": 0.7139, + "num_input_tokens_seen": 149475715, + "step": 4600 + }, + { + "epoch": 0.4149343914866754, + "flos": 23589334810080.0, + "grad_norm": 1.9293088966268506, + "learning_rate": 2.636894187201589e-06, + "loss": 0.8121, + "num_input_tokens_seen": 149500250, + "step": 4601 + }, + { + "epoch": 0.41502457501014567, + "flos": 27410430619680.0, + "grad_norm": 2.8458215902224904, + "learning_rate": 2.6363403630435176e-06, + "loss": 0.6647, + "num_input_tokens_seen": 149530805, + "step": 4602 + }, + { + "epoch": 0.4151147585336159, + "flos": 24865201649280.0, + "grad_norm": 1.7320462650789035, + "learning_rate": 2.635786484589378e-06, + "loss": 0.7482, + "num_input_tokens_seen": 149561595, + "step": 4603 + }, + { + "epoch": 0.4152049420570862, + "flos": 15136864962720.0, + "grad_norm": 1.7122160432500384, + "learning_rate": 2.63523255188643e-06, + "loss": 0.6578, + "num_input_tokens_seen": 149588750, + "step": 4604 + }, + { + "epoch": 0.4152951255805564, + "flos": 25119334263840.0, + "grad_norm": 1.4558576096139433, + "learning_rate": 2.6346785649819375e-06, + "loss": 0.7821, + "num_input_tokens_seen": 149618235, + "step": 4605 + }, + { + "epoch": 0.4153853091040267, + "flos": 27997395697920.0, + "grad_norm": 1.7279013711500553, + "learning_rate": 2.6341245239231706e-06, + "loss": 0.7805, + "num_input_tokens_seen": 149647515, + "step": 4606 + }, + { + "epoch": 0.41547549262749695, + "flos": 26829189684480.0, + "grad_norm": 1.2940224217885217, + "learning_rate": 2.6335704287574024e-06, + "loss": 0.7565, + "num_input_tokens_seen": 149678335, + "step": 4607 + }, + { + "epoch": 0.41556567615096723, + "flos": 26062257130080.0, + "grad_norm": 2.5115682933589145, + "learning_rate": 2.6330162795319124e-06, + "loss": 0.8304, + "num_input_tokens_seen": 149706295, + "step": 4608 + }, + { + "epoch": 0.41565585967443747, + "flos": 24317902438560.0, + "grad_norm": 2.260632506113666, + "learning_rate": 2.632462076293983e-06, + "loss": 0.8102, + "num_input_tokens_seen": 149735255, + "step": 4609 + }, + { + "epoch": 0.41574604319790776, + "flos": 24099123999360.0, + "grad_norm": 1.5843953118160299, + "learning_rate": 2.6319078190909017e-06, + "loss": 0.7539, + "num_input_tokens_seen": 149765305, + "step": 4610 + }, + { + "epoch": 0.415836226721378, + "flos": 16959399126720.0, + "grad_norm": 2.498076939047735, + "learning_rate": 2.6313535079699606e-06, + "loss": 0.6739, + "num_input_tokens_seen": 149792710, + "step": 4611 + }, + { + "epoch": 0.4159264102448483, + "flos": 25519715648640.0, + "grad_norm": 2.0796029707813855, + "learning_rate": 2.6307991429784572e-06, + "loss": 0.7327, + "num_input_tokens_seen": 149818510, + "step": 4612 + }, + { + "epoch": 0.4160165937683185, + "flos": 24533038241280.0, + "grad_norm": 1.5436081732338571, + "learning_rate": 2.6302447241636924e-06, + "loss": 0.8163, + "num_input_tokens_seen": 149847520, + "step": 4613 + }, + { + "epoch": 0.4161067772917888, + "flos": 19397227459680.0, + "grad_norm": 1.8026688420213932, + "learning_rate": 2.629690251572973e-06, + "loss": 0.8327, + "num_input_tokens_seen": 149875095, + "step": 4614 + }, + { + "epoch": 0.41619696081525903, + "flos": 19472916558240.0, + "grad_norm": 1.6548268663828964, + "learning_rate": 2.629135725253609e-06, + "loss": 0.7316, + "num_input_tokens_seen": 149901475, + "step": 4615 + }, + { + "epoch": 0.4162871443387293, + "flos": 26430629617920.0, + "grad_norm": 1.6684903260688762, + "learning_rate": 2.6285811452529162e-06, + "loss": 0.7437, + "num_input_tokens_seen": 149930345, + "step": 4616 + }, + { + "epoch": 0.41637732786219955, + "flos": 34987043315040.0, + "grad_norm": 1.6297902148932444, + "learning_rate": 2.6280265116182136e-06, + "loss": 0.7723, + "num_input_tokens_seen": 149961380, + "step": 4617 + }, + { + "epoch": 0.41646751138566984, + "flos": 31235317744800.0, + "grad_norm": 1.5547144135105697, + "learning_rate": 2.6274718243968266e-06, + "loss": 0.7568, + "num_input_tokens_seen": 149991980, + "step": 4618 + }, + { + "epoch": 0.4165576949091401, + "flos": 23152632836160.0, + "grad_norm": 1.8314331640097106, + "learning_rate": 2.626917083636084e-06, + "loss": 0.6732, + "num_input_tokens_seen": 150019035, + "step": 4619 + }, + { + "epoch": 0.41664787843261036, + "flos": 23443197549120.0, + "grad_norm": 1.4831519458893403, + "learning_rate": 2.6263622893833183e-06, + "loss": 0.6898, + "num_input_tokens_seen": 150049055, + "step": 4620 + }, + { + "epoch": 0.4167380619560806, + "flos": 24063546805440.0, + "grad_norm": 1.5947343367079336, + "learning_rate": 2.625807441685869e-06, + "loss": 0.7014, + "num_input_tokens_seen": 150079930, + "step": 4621 + }, + { + "epoch": 0.4168282454795509, + "flos": 69703667927040.0, + "grad_norm": 0.6082362024503867, + "learning_rate": 2.625252540591078e-06, + "loss": 0.5608, + "num_input_tokens_seen": 150168685, + "step": 4622 + }, + { + "epoch": 0.4169184290030212, + "flos": 27227006355840.0, + "grad_norm": 1.7603699115510878, + "learning_rate": 2.6246975861462927e-06, + "loss": 0.8281, + "num_input_tokens_seen": 150196690, + "step": 4623 + }, + { + "epoch": 0.4170086125264914, + "flos": 22532878296000.0, + "grad_norm": 1.8989727380331427, + "learning_rate": 2.624142578398864e-06, + "loss": 0.8005, + "num_input_tokens_seen": 150223570, + "step": 4624 + }, + { + "epoch": 0.4170987960499617, + "flos": 29418210365760.0, + "grad_norm": 3.2081870669096935, + "learning_rate": 2.6235875173961498e-06, + "loss": 0.7387, + "num_input_tokens_seen": 150255570, + "step": 4625 + }, + { + "epoch": 0.41718897957343193, + "flos": 26867554610400.0, + "grad_norm": 1.773438844386192, + "learning_rate": 2.62303240318551e-06, + "loss": 0.686, + "num_input_tokens_seen": 150286780, + "step": 4626 + }, + { + "epoch": 0.4172791630969022, + "flos": 28070594422560.0, + "grad_norm": 2.3279838775260013, + "learning_rate": 2.62247723581431e-06, + "loss": 0.8316, + "num_input_tokens_seen": 150317115, + "step": 4627 + }, + { + "epoch": 0.41736934662037245, + "flos": 20347026731520.0, + "grad_norm": 3.436456001576194, + "learning_rate": 2.62192201532992e-06, + "loss": 0.7576, + "num_input_tokens_seen": 150341930, + "step": 4628 + }, + { + "epoch": 0.41745953014384274, + "flos": 23006830103040.0, + "grad_norm": 1.6791128195686684, + "learning_rate": 2.6213667417797145e-06, + "loss": 0.773, + "num_input_tokens_seen": 150372425, + "step": 4629 + }, + { + "epoch": 0.41754971366731297, + "flos": 17211561744000.0, + "grad_norm": 2.268615012448571, + "learning_rate": 2.6208114152110725e-06, + "loss": 0.7822, + "num_input_tokens_seen": 150399695, + "step": 4630 + }, + { + "epoch": 0.41763989719078326, + "flos": 16519500553440.0, + "grad_norm": 1.7563796922033128, + "learning_rate": 2.6202560356713774e-06, + "loss": 0.7709, + "num_input_tokens_seen": 150425530, + "step": 4631 + }, + { + "epoch": 0.4177300807142535, + "flos": 67156877826720.0, + "grad_norm": 0.6377957368672943, + "learning_rate": 2.619700603208017e-06, + "loss": 0.6384, + "num_input_tokens_seen": 150517955, + "step": 4632 + }, + { + "epoch": 0.4178202642377238, + "flos": 23330630315040.0, + "grad_norm": 1.6009054185138796, + "learning_rate": 2.6191451178683842e-06, + "loss": 0.6683, + "num_input_tokens_seen": 150547845, + "step": 4633 + }, + { + "epoch": 0.417910447761194, + "flos": 24570919960320.0, + "grad_norm": 1.877042420929661, + "learning_rate": 2.6185895796998764e-06, + "loss": 0.7851, + "num_input_tokens_seen": 150575835, + "step": 4634 + }, + { + "epoch": 0.4180006312846643, + "flos": 18451888559040.0, + "grad_norm": 1.8241876440234732, + "learning_rate": 2.6180339887498946e-06, + "loss": 0.8254, + "num_input_tokens_seen": 150600865, + "step": 4635 + }, + { + "epoch": 0.41809081480813454, + "flos": 24500211609600.0, + "grad_norm": 2.17266172801359, + "learning_rate": 2.617478345065846e-06, + "loss": 0.6978, + "num_input_tokens_seen": 150629415, + "step": 4636 + }, + { + "epoch": 0.4181809983316048, + "flos": 29054521267680.0, + "grad_norm": 1.8519943956862943, + "learning_rate": 2.616922648695139e-06, + "loss": 0.7765, + "num_input_tokens_seen": 150656955, + "step": 4637 + }, + { + "epoch": 0.41827118185507506, + "flos": 29345197489920.0, + "grad_norm": 1.9504893104712087, + "learning_rate": 2.61636689968519e-06, + "loss": 0.6548, + "num_input_tokens_seen": 150685815, + "step": 4638 + }, + { + "epoch": 0.41836136537854535, + "flos": 12622975833600.0, + "grad_norm": 2.0271511096607786, + "learning_rate": 2.6158110980834186e-06, + "loss": 0.8197, + "num_input_tokens_seen": 150711745, + "step": 4639 + }, + { + "epoch": 0.4184515489020156, + "flos": 18161658373920.0, + "grad_norm": 3.094348888036634, + "learning_rate": 2.615255243937249e-06, + "loss": 0.8096, + "num_input_tokens_seen": 150737930, + "step": 4640 + }, + { + "epoch": 0.41854173242548587, + "flos": 18779963293440.0, + "grad_norm": 2.5348362889744687, + "learning_rate": 2.61469933729411e-06, + "loss": 0.7822, + "num_input_tokens_seen": 150763560, + "step": 4641 + }, + { + "epoch": 0.4186319159489561, + "flos": 29309137089120.0, + "grad_norm": 1.8895362682712784, + "learning_rate": 2.614143378201433e-06, + "loss": 0.7414, + "num_input_tokens_seen": 150791305, + "step": 4642 + }, + { + "epoch": 0.4187220994724264, + "flos": 23549222905440.0, + "grad_norm": 1.6982581920429003, + "learning_rate": 2.6135873667066567e-06, + "loss": 0.7371, + "num_input_tokens_seen": 150821120, + "step": 4643 + }, + { + "epoch": 0.4188122829958966, + "flos": 22819688863200.0, + "grad_norm": 12.912612639800935, + "learning_rate": 2.613031302857224e-06, + "loss": 0.716, + "num_input_tokens_seen": 150848950, + "step": 4644 + }, + { + "epoch": 0.4189024665193669, + "flos": 20491788711360.0, + "grad_norm": 1.941828476490018, + "learning_rate": 2.6124751867005792e-06, + "loss": 0.7835, + "num_input_tokens_seen": 150875255, + "step": 4645 + }, + { + "epoch": 0.4189926500428372, + "flos": 22054726306080.0, + "grad_norm": 1.3136365601610562, + "learning_rate": 2.611919018284175e-06, + "loss": 0.807, + "num_input_tokens_seen": 150904875, + "step": 4646 + }, + { + "epoch": 0.41908283356630743, + "flos": 24026743009440.0, + "grad_norm": 2.0815031659452474, + "learning_rate": 2.611362797655466e-06, + "loss": 0.8077, + "num_input_tokens_seen": 150932955, + "step": 4647 + }, + { + "epoch": 0.4191730170897777, + "flos": 24099272678400.0, + "grad_norm": 1.9216207320564749, + "learning_rate": 2.6108065248619124e-06, + "loss": 0.8153, + "num_input_tokens_seen": 150962550, + "step": 4648 + }, + { + "epoch": 0.41926320061324795, + "flos": 27924643010400.0, + "grad_norm": 1.4293950516562386, + "learning_rate": 2.610250199950978e-06, + "loss": 0.6915, + "num_input_tokens_seen": 150994760, + "step": 4649 + }, + { + "epoch": 0.41935338413671824, + "flos": 18962346804000.0, + "grad_norm": 1.9888916277851778, + "learning_rate": 2.609693822970131e-06, + "loss": 0.728, + "num_input_tokens_seen": 151023165, + "step": 4650 + }, + { + "epoch": 0.4194435676601885, + "flos": 19104358221600.0, + "grad_norm": 2.17641452086361, + "learning_rate": 2.609137393966846e-06, + "loss": 0.7199, + "num_input_tokens_seen": 151051175, + "step": 4651 + }, + { + "epoch": 0.41953375118365877, + "flos": 23042407296960.0, + "grad_norm": 1.8370667576130377, + "learning_rate": 2.6085809129886e-06, + "loss": 0.713, + "num_input_tokens_seen": 151079205, + "step": 4652 + }, + { + "epoch": 0.419623934707129, + "flos": 17469262655520.0, + "grad_norm": 2.5379571474694678, + "learning_rate": 2.608024380082874e-06, + "loss": 0.767, + "num_input_tokens_seen": 151104720, + "step": 4653 + }, + { + "epoch": 0.4197141182305993, + "flos": 25334284217760.0, + "grad_norm": 1.6347044428561923, + "learning_rate": 2.6074677952971554e-06, + "loss": 0.7552, + "num_input_tokens_seen": 151131535, + "step": 4654 + }, + { + "epoch": 0.4198043017540695, + "flos": 34446509000640.0, + "grad_norm": 2.1062485873163848, + "learning_rate": 2.606911158678935e-06, + "loss": 0.6845, + "num_input_tokens_seen": 151164885, + "step": 4655 + }, + { + "epoch": 0.4198944852775398, + "flos": 18743754213600.0, + "grad_norm": 2.1717746629110493, + "learning_rate": 2.606354470275708e-06, + "loss": 0.7463, + "num_input_tokens_seen": 151191755, + "step": 4656 + }, + { + "epoch": 0.41998466880101004, + "flos": 39947681519520.0, + "grad_norm": 3.969831424468335, + "learning_rate": 2.6057977301349744e-06, + "loss": 0.7261, + "num_input_tokens_seen": 151222925, + "step": 4657 + }, + { + "epoch": 0.42007485232448033, + "flos": 28031560440960.0, + "grad_norm": 1.7067818535860377, + "learning_rate": 2.6052409383042383e-06, + "loss": 0.7162, + "num_input_tokens_seen": 151250860, + "step": 4658 + }, + { + "epoch": 0.42016503584795056, + "flos": 23258881211040.0, + "grad_norm": 1.6151751100306808, + "learning_rate": 2.6046840948310074e-06, + "loss": 0.7351, + "num_input_tokens_seen": 151280210, + "step": 4659 + }, + { + "epoch": 0.42025521937142085, + "flos": 28143495789120.0, + "grad_norm": 1.6258863086109197, + "learning_rate": 2.6041271997627962e-06, + "loss": 0.7566, + "num_input_tokens_seen": 151310565, + "step": 4660 + }, + { + "epoch": 0.4203454028948911, + "flos": 20493461350560.0, + "grad_norm": 2.1804055839308916, + "learning_rate": 2.6035702531471202e-06, + "loss": 0.7434, + "num_input_tokens_seen": 151338195, + "step": 4661 + }, + { + "epoch": 0.4204355864183614, + "flos": 20782502103360.0, + "grad_norm": 1.581885135637757, + "learning_rate": 2.6030132550315035e-06, + "loss": 0.7528, + "num_input_tokens_seen": 151366495, + "step": 4662 + }, + { + "epoch": 0.4205257699418316, + "flos": 58100116059360.0, + "grad_norm": 0.6480982249071569, + "learning_rate": 2.60245620546347e-06, + "loss": 0.613, + "num_input_tokens_seen": 151454865, + "step": 4663 + }, + { + "epoch": 0.4206159534653019, + "flos": 25483023361920.0, + "grad_norm": 2.603684668121926, + "learning_rate": 2.6018991044905517e-06, + "loss": 0.7337, + "num_input_tokens_seen": 151481925, + "step": 4664 + }, + { + "epoch": 0.42070613698877213, + "flos": 21221434262880.0, + "grad_norm": 1.651036604463376, + "learning_rate": 2.6013419521602825e-06, + "loss": 0.7934, + "num_input_tokens_seen": 151508030, + "step": 4665 + }, + { + "epoch": 0.4207963205122424, + "flos": 26207427977280.0, + "grad_norm": 1.887110256857361, + "learning_rate": 2.600784748520202e-06, + "loss": 0.8829, + "num_input_tokens_seen": 151537895, + "step": 4666 + }, + { + "epoch": 0.42088650403571265, + "flos": 20128471310880.0, + "grad_norm": 1.586219649749743, + "learning_rate": 2.6002274936178544e-06, + "loss": 0.6929, + "num_input_tokens_seen": 151566385, + "step": 4667 + }, + { + "epoch": 0.42097668755918294, + "flos": 17105201859840.0, + "grad_norm": 1.9155728978699382, + "learning_rate": 2.5996701875007873e-06, + "loss": 0.7705, + "num_input_tokens_seen": 151593000, + "step": 4668 + }, + { + "epoch": 0.4210668710826532, + "flos": 27451657617120.0, + "grad_norm": 1.6493892724278651, + "learning_rate": 2.5991128302165533e-06, + "loss": 0.743, + "num_input_tokens_seen": 151621310, + "step": 4669 + }, + { + "epoch": 0.42115705460612346, + "flos": 23371113917280.0, + "grad_norm": 2.291644102043266, + "learning_rate": 2.5985554218127094e-06, + "loss": 0.8196, + "num_input_tokens_seen": 151651090, + "step": 4670 + }, + { + "epoch": 0.42124723812959375, + "flos": 17686702983360.0, + "grad_norm": 2.1320736901464836, + "learning_rate": 2.597997962336816e-06, + "loss": 0.6895, + "num_input_tokens_seen": 151677990, + "step": 4671 + }, + { + "epoch": 0.421337421653064, + "flos": 21658210576320.0, + "grad_norm": 2.229783789754952, + "learning_rate": 2.5974404518364393e-06, + "loss": 0.7453, + "num_input_tokens_seen": 151707295, + "step": 4672 + }, + { + "epoch": 0.42142760517653427, + "flos": 21075817378560.0, + "grad_norm": 1.821025463183886, + "learning_rate": 2.596882890359149e-06, + "loss": 0.8763, + "num_input_tokens_seen": 151732995, + "step": 4673 + }, + { + "epoch": 0.4215177887000045, + "flos": 18451777049760.0, + "grad_norm": 2.3844771777016915, + "learning_rate": 2.5963252779525196e-06, + "loss": 0.6779, + "num_input_tokens_seen": 151758990, + "step": 4674 + }, + { + "epoch": 0.4216079722234748, + "flos": 29636505598080.0, + "grad_norm": 1.7522920176556676, + "learning_rate": 2.595767614664129e-06, + "loss": 0.7579, + "num_input_tokens_seen": 151788455, + "step": 4675 + }, + { + "epoch": 0.421698155746945, + "flos": 18015149415360.0, + "grad_norm": 1.6903744693510698, + "learning_rate": 2.5952099005415607e-06, + "loss": 0.7359, + "num_input_tokens_seen": 151815765, + "step": 4676 + }, + { + "epoch": 0.4217883392704153, + "flos": 23152000950240.0, + "grad_norm": 3.1091526277856176, + "learning_rate": 2.594652135632402e-06, + "loss": 0.7205, + "num_input_tokens_seen": 151841960, + "step": 4677 + }, + { + "epoch": 0.42187852279388555, + "flos": 25115096911200.0, + "grad_norm": 1.3617578865244495, + "learning_rate": 2.594094319984244e-06, + "loss": 0.7115, + "num_input_tokens_seen": 151871430, + "step": 4678 + }, + { + "epoch": 0.42196870631735584, + "flos": 24136485341760.0, + "grad_norm": 1.9270178362739014, + "learning_rate": 2.5935364536446825e-06, + "loss": 0.8142, + "num_input_tokens_seen": 151900920, + "step": 4679 + }, + { + "epoch": 0.42205888984082607, + "flos": 21941118318720.0, + "grad_norm": 2.2787146915749164, + "learning_rate": 2.5929785366613185e-06, + "loss": 0.7568, + "num_input_tokens_seen": 151925880, + "step": 4680 + }, + { + "epoch": 0.42214907336429636, + "flos": 27669692661120.0, + "grad_norm": 2.070030682654276, + "learning_rate": 2.592420569081756e-06, + "loss": 0.7914, + "num_input_tokens_seen": 151953920, + "step": 4681 + }, + { + "epoch": 0.4222392568877666, + "flos": 30547828434720.0, + "grad_norm": 3.040549103435567, + "learning_rate": 2.5918625509536037e-06, + "loss": 0.7141, + "num_input_tokens_seen": 151984740, + "step": 4682 + }, + { + "epoch": 0.4223294404112369, + "flos": 67684471331040.0, + "grad_norm": 0.6219766207913199, + "learning_rate": 2.591304482324475e-06, + "loss": 0.5776, + "num_input_tokens_seen": 152078360, + "step": 4683 + }, + { + "epoch": 0.4224196239347071, + "flos": 23146760014080.0, + "grad_norm": 2.1097705611460955, + "learning_rate": 2.5907463632419878e-06, + "loss": 0.7681, + "num_input_tokens_seen": 152107305, + "step": 4684 + }, + { + "epoch": 0.4225098074581774, + "flos": 26828446289280.0, + "grad_norm": 2.3058207166643903, + "learning_rate": 2.5901881937537632e-06, + "loss": 0.7199, + "num_input_tokens_seen": 152134930, + "step": 4685 + }, + { + "epoch": 0.42259999098164763, + "flos": 67126355720160.0, + "grad_norm": 0.7125988506878522, + "learning_rate": 2.589629973907428e-06, + "loss": 0.6117, + "num_input_tokens_seen": 152221295, + "step": 4686 + }, + { + "epoch": 0.4226901745051179, + "flos": 26795470978560.0, + "grad_norm": 2.2100643954669277, + "learning_rate": 2.589071703750612e-06, + "loss": 0.7083, + "num_input_tokens_seen": 152249900, + "step": 4687 + }, + { + "epoch": 0.42278035802858815, + "flos": 19542955853280.0, + "grad_norm": 1.9530887447806906, + "learning_rate": 2.5885133833309504e-06, + "loss": 0.7939, + "num_input_tokens_seen": 152278070, + "step": 4688 + }, + { + "epoch": 0.42287054155205844, + "flos": 21693564751680.0, + "grad_norm": 1.6190037502695418, + "learning_rate": 2.5879550126960814e-06, + "loss": 0.8337, + "num_input_tokens_seen": 152305740, + "step": 4689 + }, + { + "epoch": 0.4229607250755287, + "flos": 20456323026720.0, + "grad_norm": 1.514651743075455, + "learning_rate": 2.5873965918936494e-06, + "loss": 0.812, + "num_input_tokens_seen": 152333975, + "step": 4690 + }, + { + "epoch": 0.42305090859899896, + "flos": 30614187923520.0, + "grad_norm": 1.7783510811923233, + "learning_rate": 2.586838120971301e-06, + "loss": 0.7826, + "num_input_tokens_seen": 152363365, + "step": 4691 + }, + { + "epoch": 0.4231410921224692, + "flos": 25513248110400.0, + "grad_norm": 1.5208140942475792, + "learning_rate": 2.586279599976689e-06, + "loss": 0.8173, + "num_input_tokens_seen": 152391850, + "step": 4692 + }, + { + "epoch": 0.4232312756459395, + "flos": 23953432775520.0, + "grad_norm": 1.6297289191127708, + "learning_rate": 2.585721028957468e-06, + "loss": 0.7542, + "num_input_tokens_seen": 152421705, + "step": 4693 + }, + { + "epoch": 0.4233214591694098, + "flos": 24645382456800.0, + "grad_norm": 1.6208467226366163, + "learning_rate": 2.585162407961299e-06, + "loss": 0.7661, + "num_input_tokens_seen": 152450285, + "step": 4694 + }, + { + "epoch": 0.42341164269288, + "flos": 19290384368640.0, + "grad_norm": 2.025850955515638, + "learning_rate": 2.584603737035847e-06, + "loss": 0.697, + "num_input_tokens_seen": 152476915, + "step": 4695 + }, + { + "epoch": 0.4235018262163503, + "flos": 23950236176160.0, + "grad_norm": 2.059946733915667, + "learning_rate": 2.5840450162287806e-06, + "loss": 0.7506, + "num_input_tokens_seen": 152503885, + "step": 4696 + }, + { + "epoch": 0.42359200973982053, + "flos": 26358397307040.0, + "grad_norm": 8.933501089238518, + "learning_rate": 2.583486245587774e-06, + "loss": 0.669, + "num_input_tokens_seen": 152533520, + "step": 4697 + }, + { + "epoch": 0.4236821932632908, + "flos": 19654482334080.0, + "grad_norm": 1.97918274186951, + "learning_rate": 2.5829274251605023e-06, + "loss": 0.8607, + "num_input_tokens_seen": 152560855, + "step": 4698 + }, + { + "epoch": 0.42377237678676105, + "flos": 18708214189440.0, + "grad_norm": 1.7530518895644513, + "learning_rate": 2.582368554994649e-06, + "loss": 0.7318, + "num_input_tokens_seen": 152588575, + "step": 4699 + }, + { + "epoch": 0.42386256031023134, + "flos": 65856436042560.0, + "grad_norm": 0.7477637967973452, + "learning_rate": 2.5818096351378994e-06, + "loss": 0.6132, + "num_input_tokens_seen": 152670405, + "step": 4700 + }, + { + "epoch": 0.4239527438337016, + "flos": 31713804262560.0, + "grad_norm": 2.6999416239043863, + "learning_rate": 2.5812506656379435e-06, + "loss": 0.8117, + "num_input_tokens_seen": 152700175, + "step": 4701 + }, + { + "epoch": 0.42404292735717186, + "flos": 24172545742560.0, + "grad_norm": 2.2839114197558135, + "learning_rate": 2.580691646542476e-06, + "loss": 0.7059, + "num_input_tokens_seen": 152727735, + "step": 4702 + }, + { + "epoch": 0.4241331108806421, + "flos": 22966234991520.0, + "grad_norm": 1.5738810515433967, + "learning_rate": 2.5801325778991958e-06, + "loss": 0.7894, + "num_input_tokens_seen": 152754355, + "step": 4703 + }, + { + "epoch": 0.4242232944041124, + "flos": 20747296607040.0, + "grad_norm": 4.045910812458446, + "learning_rate": 2.5795734597558043e-06, + "loss": 0.613, + "num_input_tokens_seen": 152781360, + "step": 4704 + }, + { + "epoch": 0.4243134779275826, + "flos": 24501029344320.0, + "grad_norm": 3.4731586380394135, + "learning_rate": 2.579014292160011e-06, + "loss": 0.7564, + "num_input_tokens_seen": 152809930, + "step": 4705 + }, + { + "epoch": 0.4244036614510529, + "flos": 26722272253920.0, + "grad_norm": 1.46145514634923, + "learning_rate": 2.5784550751595236e-06, + "loss": 0.739, + "num_input_tokens_seen": 152842505, + "step": 4706 + }, + { + "epoch": 0.42449384497452314, + "flos": 20129326215360.0, + "grad_norm": 1.4445551357079098, + "learning_rate": 2.577895808802061e-06, + "loss": 0.8154, + "num_input_tokens_seen": 152869360, + "step": 4707 + }, + { + "epoch": 0.4245840284979934, + "flos": 28837303958400.0, + "grad_norm": 1.813468937197889, + "learning_rate": 2.577336493135341e-06, + "loss": 0.673, + "num_input_tokens_seen": 152901580, + "step": 4708 + }, + { + "epoch": 0.42467421202146366, + "flos": 21550252392480.0, + "grad_norm": 2.0295984449811426, + "learning_rate": 2.576777128207088e-06, + "loss": 0.8522, + "num_input_tokens_seen": 152930890, + "step": 4709 + }, + { + "epoch": 0.42476439554493395, + "flos": 31091596518240.0, + "grad_norm": 1.5441208799312311, + "learning_rate": 2.5762177140650306e-06, + "loss": 0.7162, + "num_input_tokens_seen": 152962895, + "step": 4710 + }, + { + "epoch": 0.4248545790684042, + "flos": 22897236449760.0, + "grad_norm": 2.1042635552285938, + "learning_rate": 2.5756582507569003e-06, + "loss": 0.7656, + "num_input_tokens_seen": 152989435, + "step": 4711 + }, + { + "epoch": 0.42494476259187447, + "flos": 27086741916960.0, + "grad_norm": 1.7842018698162658, + "learning_rate": 2.5750987383304335e-06, + "loss": 0.7807, + "num_input_tokens_seen": 153020345, + "step": 4712 + }, + { + "epoch": 0.4250349461153447, + "flos": 28507853942880.0, + "grad_norm": 2.3203442503943537, + "learning_rate": 2.574539176833372e-06, + "loss": 0.743, + "num_input_tokens_seen": 153047230, + "step": 4713 + }, + { + "epoch": 0.425125129638815, + "flos": 22384473679680.0, + "grad_norm": 1.9399689356292964, + "learning_rate": 2.5739795663134594e-06, + "loss": 0.7056, + "num_input_tokens_seen": 153075355, + "step": 4714 + }, + { + "epoch": 0.4252153131622852, + "flos": 25042418563200.0, + "grad_norm": 1.5737656161235694, + "learning_rate": 2.5734199068184454e-06, + "loss": 0.8413, + "num_input_tokens_seen": 153105125, + "step": 4715 + }, + { + "epoch": 0.4253054966857555, + "flos": 31380191233920.0, + "grad_norm": 1.8032286060523288, + "learning_rate": 2.572860198396083e-06, + "loss": 0.6793, + "num_input_tokens_seen": 153132935, + "step": 4716 + }, + { + "epoch": 0.4253956802092258, + "flos": 27742928555520.0, + "grad_norm": 1.814874409200274, + "learning_rate": 2.57230044109413e-06, + "loss": 0.7799, + "num_input_tokens_seen": 153161415, + "step": 4717 + }, + { + "epoch": 0.42548586373269603, + "flos": 20455728310560.0, + "grad_norm": 2.2221653401194352, + "learning_rate": 2.5717406349603483e-06, + "loss": 0.7395, + "num_input_tokens_seen": 153186460, + "step": 4718 + }, + { + "epoch": 0.4255760472561663, + "flos": 27848730893280.0, + "grad_norm": 1.7583365104369377, + "learning_rate": 2.5711807800425026e-06, + "loss": 0.773, + "num_input_tokens_seen": 153216010, + "step": 4719 + }, + { + "epoch": 0.42566623077963656, + "flos": 31634249508960.0, + "grad_norm": 2.3402054495974673, + "learning_rate": 2.5706208763883633e-06, + "loss": 0.5921, + "num_input_tokens_seen": 153245795, + "step": 4720 + }, + { + "epoch": 0.42575641430310684, + "flos": 22679164236000.0, + "grad_norm": 1.7841856339913562, + "learning_rate": 2.570060924045704e-06, + "loss": 0.7132, + "num_input_tokens_seen": 153273050, + "step": 4721 + }, + { + "epoch": 0.4258465978265771, + "flos": 22635930071520.0, + "grad_norm": 1.6175381929820436, + "learning_rate": 2.569500923062304e-06, + "loss": 0.788, + "num_input_tokens_seen": 153301185, + "step": 4722 + }, + { + "epoch": 0.42593678135004737, + "flos": 43590779850240.0, + "grad_norm": 2.023607990638051, + "learning_rate": 2.5689408734859445e-06, + "loss": 0.6782, + "num_input_tokens_seen": 153332010, + "step": 4723 + }, + { + "epoch": 0.4260269648735176, + "flos": 22023163446240.0, + "grad_norm": 1.8404039311704248, + "learning_rate": 2.5683807753644127e-06, + "loss": 0.7001, + "num_input_tokens_seen": 153361205, + "step": 4724 + }, + { + "epoch": 0.4261171483969879, + "flos": 16849582454880.0, + "grad_norm": 1.544817085425681, + "learning_rate": 2.5678206287454996e-06, + "loss": 0.7333, + "num_input_tokens_seen": 153387520, + "step": 4725 + }, + { + "epoch": 0.4262073319204581, + "flos": 22095581605920.0, + "grad_norm": 2.2583486586985644, + "learning_rate": 2.567260433676999e-06, + "loss": 0.7886, + "num_input_tokens_seen": 153416550, + "step": 4726 + }, + { + "epoch": 0.4262975154439284, + "flos": 27487866696960.0, + "grad_norm": 2.0143274670093105, + "learning_rate": 2.5667001902067107e-06, + "loss": 0.7291, + "num_input_tokens_seen": 153444485, + "step": 4727 + }, + { + "epoch": 0.42638769896739864, + "flos": 24646051512480.0, + "grad_norm": 2.244134675525186, + "learning_rate": 2.566139898382437e-06, + "loss": 0.8076, + "num_input_tokens_seen": 153472530, + "step": 4728 + }, + { + "epoch": 0.42647788249086893, + "flos": 20199365510400.0, + "grad_norm": 3.728627910092256, + "learning_rate": 2.5655795582519853e-06, + "loss": 0.771, + "num_input_tokens_seen": 153500885, + "step": 4729 + }, + { + "epoch": 0.42656806601433916, + "flos": 24241655793600.0, + "grad_norm": 2.0218239024572857, + "learning_rate": 2.565019169863168e-06, + "loss": 0.8283, + "num_input_tokens_seen": 153529520, + "step": 4730 + }, + { + "epoch": 0.42665824953780945, + "flos": 64977456630720.0, + "grad_norm": 0.586500177420795, + "learning_rate": 2.5644587332637994e-06, + "loss": 0.547, + "num_input_tokens_seen": 153629635, + "step": 4731 + }, + { + "epoch": 0.4267484330612797, + "flos": 26176125305760.0, + "grad_norm": 1.9331393237713324, + "learning_rate": 2.5638982485016994e-06, + "loss": 0.7865, + "num_input_tokens_seen": 153658595, + "step": 4732 + }, + { + "epoch": 0.42683861658475, + "flos": 70773691403520.0, + "grad_norm": 0.7081816361244886, + "learning_rate": 2.5633377156246917e-06, + "loss": 0.5516, + "num_input_tokens_seen": 153748455, + "step": 4733 + }, + { + "epoch": 0.4269288001082202, + "flos": 19254175288800.0, + "grad_norm": 2.533639415855223, + "learning_rate": 2.562777134680603e-06, + "loss": 0.7368, + "num_input_tokens_seen": 153776115, + "step": 4734 + }, + { + "epoch": 0.4270189836316905, + "flos": 26030545591200.0, + "grad_norm": 1.8006957855134615, + "learning_rate": 2.562216505717267e-06, + "loss": 0.7627, + "num_input_tokens_seen": 153803830, + "step": 4735 + }, + { + "epoch": 0.42710916715516073, + "flos": 20274162534720.0, + "grad_norm": 1.723495936159589, + "learning_rate": 2.561655828782518e-06, + "loss": 0.7811, + "num_input_tokens_seen": 153831430, + "step": 4736 + }, + { + "epoch": 0.427199350678631, + "flos": 23331448049760.0, + "grad_norm": 2.4370131373186363, + "learning_rate": 2.561095103924197e-06, + "loss": 0.7412, + "num_input_tokens_seen": 153857850, + "step": 4737 + }, + { + "epoch": 0.42728953420210125, + "flos": 25372240276320.0, + "grad_norm": 1.6925441094965745, + "learning_rate": 2.560534331190148e-06, + "loss": 0.7045, + "num_input_tokens_seen": 153888625, + "step": 4738 + }, + { + "epoch": 0.42737971772557154, + "flos": 70838006555520.0, + "grad_norm": 0.5834601610683537, + "learning_rate": 2.559973510628218e-06, + "loss": 0.5486, + "num_input_tokens_seen": 153984575, + "step": 4739 + }, + { + "epoch": 0.4274699012490418, + "flos": 21331325274240.0, + "grad_norm": 4.085193573603762, + "learning_rate": 2.5594126422862615e-06, + "loss": 0.7043, + "num_input_tokens_seen": 154013640, + "step": 4740 + }, + { + "epoch": 0.42756008477251206, + "flos": 24426938545440.0, + "grad_norm": 2.1872999917283176, + "learning_rate": 2.558851726212134e-06, + "loss": 0.7042, + "num_input_tokens_seen": 154041865, + "step": 4741 + }, + { + "epoch": 0.42765026829598235, + "flos": 28288629466560.0, + "grad_norm": 1.5401319476982487, + "learning_rate": 2.5582907624536953e-06, + "loss": 0.7238, + "num_input_tokens_seen": 154072345, + "step": 4742 + }, + { + "epoch": 0.4277404518194526, + "flos": 20893322358720.0, + "grad_norm": 3.009518464056675, + "learning_rate": 2.557729751058811e-06, + "loss": 0.7722, + "num_input_tokens_seen": 154099015, + "step": 4743 + }, + { + "epoch": 0.42783063534292287, + "flos": 17210595330240.0, + "grad_norm": 1.8432816379304513, + "learning_rate": 2.557168692075348e-06, + "loss": 0.7289, + "num_input_tokens_seen": 154125975, + "step": 4744 + }, + { + "epoch": 0.4279208188663931, + "flos": 26249026672320.0, + "grad_norm": 1.6132536878840715, + "learning_rate": 2.556607585551181e-06, + "loss": 0.6968, + "num_input_tokens_seen": 154156575, + "step": 4745 + }, + { + "epoch": 0.4280110023898634, + "flos": 66301947249600.0, + "grad_norm": 0.7153684014458144, + "learning_rate": 2.5560464315341844e-06, + "loss": 0.6767, + "num_input_tokens_seen": 154241995, + "step": 4746 + }, + { + "epoch": 0.4281011859133336, + "flos": 20893619716800.0, + "grad_norm": 1.8792280845445144, + "learning_rate": 2.555485230072242e-06, + "loss": 0.7472, + "num_input_tokens_seen": 154271175, + "step": 4747 + }, + { + "epoch": 0.4281913694368039, + "flos": 25484064115200.0, + "grad_norm": 1.8215900571279524, + "learning_rate": 2.5549239812132354e-06, + "loss": 0.7994, + "num_input_tokens_seen": 154297860, + "step": 4748 + }, + { + "epoch": 0.42828155296027415, + "flos": 26030619930720.0, + "grad_norm": 1.5932439860959642, + "learning_rate": 2.5543626850050556e-06, + "loss": 0.8176, + "num_input_tokens_seen": 154328115, + "step": 4749 + }, + { + "epoch": 0.42837173648374444, + "flos": 27159791962560.0, + "grad_norm": 2.5610561489994, + "learning_rate": 2.5538013414955944e-06, + "loss": 0.808, + "num_input_tokens_seen": 154357375, + "step": 4750 + }, + { + "epoch": 0.42846192000721467, + "flos": 24719473255680.0, + "grad_norm": 2.585474716219292, + "learning_rate": 2.5532399507327494e-06, + "loss": 0.7115, + "num_input_tokens_seen": 154388025, + "step": 4751 + }, + { + "epoch": 0.42855210353068496, + "flos": 67245093134400.0, + "grad_norm": 0.6006901844390559, + "learning_rate": 2.552678512764421e-06, + "loss": 0.5855, + "num_input_tokens_seen": 154488055, + "step": 4752 + }, + { + "epoch": 0.4286422870541552, + "flos": 28872026247840.0, + "grad_norm": 2.055509569610241, + "learning_rate": 2.5521170276385147e-06, + "loss": 0.8105, + "num_input_tokens_seen": 154515085, + "step": 4753 + }, + { + "epoch": 0.4287324705776255, + "flos": 24318088287360.0, + "grad_norm": 1.5773061762014617, + "learning_rate": 2.5515554954029394e-06, + "loss": 0.8035, + "num_input_tokens_seen": 154543170, + "step": 4754 + }, + { + "epoch": 0.4288226541010957, + "flos": 24974089077120.0, + "grad_norm": 1.8025958337584782, + "learning_rate": 2.550993916105608e-06, + "loss": 0.7832, + "num_input_tokens_seen": 154574360, + "step": 4755 + }, + { + "epoch": 0.428912837624566, + "flos": 24391101163200.0, + "grad_norm": 1.4691307710592494, + "learning_rate": 2.550432289794437e-06, + "loss": 0.712, + "num_input_tokens_seen": 154602965, + "step": 4756 + }, + { + "epoch": 0.42900302114803623, + "flos": 26941125032640.0, + "grad_norm": 3.2726266079650963, + "learning_rate": 2.5498706165173483e-06, + "loss": 0.7633, + "num_input_tokens_seen": 154629290, + "step": 4757 + }, + { + "epoch": 0.4290932046715065, + "flos": 24790181606400.0, + "grad_norm": 1.4508707753272199, + "learning_rate": 2.5493088963222668e-06, + "loss": 0.7292, + "num_input_tokens_seen": 154659080, + "step": 4758 + }, + { + "epoch": 0.42918338819497676, + "flos": 22314397214880.0, + "grad_norm": 1.8948942171027383, + "learning_rate": 2.548747129257121e-06, + "loss": 0.8055, + "num_input_tokens_seen": 154689535, + "step": 4759 + }, + { + "epoch": 0.42927357171844704, + "flos": 21292142613600.0, + "grad_norm": 1.7558852417666788, + "learning_rate": 2.548185315369845e-06, + "loss": 0.6767, + "num_input_tokens_seen": 154718305, + "step": 4760 + }, + { + "epoch": 0.4293637552419173, + "flos": 25120114828800.0, + "grad_norm": 1.7045331960295103, + "learning_rate": 2.5476234547083746e-06, + "loss": 0.8432, + "num_input_tokens_seen": 154748270, + "step": 4761 + }, + { + "epoch": 0.42945393876538757, + "flos": 26102889411360.0, + "grad_norm": 1.7189470351593814, + "learning_rate": 2.547061547320652e-06, + "loss": 0.8234, + "num_input_tokens_seen": 154776170, + "step": 4762 + }, + { + "epoch": 0.4295441222888578, + "flos": 24312921690720.0, + "grad_norm": 1.33224628759931, + "learning_rate": 2.5464995932546217e-06, + "loss": 0.7382, + "num_input_tokens_seen": 154806485, + "step": 4763 + }, + { + "epoch": 0.4296343058123281, + "flos": 25554846805440.0, + "grad_norm": 2.024968808722754, + "learning_rate": 2.545937592558232e-06, + "loss": 0.7077, + "num_input_tokens_seen": 154836855, + "step": 4764 + }, + { + "epoch": 0.4297244893357984, + "flos": 44392471863840.0, + "grad_norm": 1.57174882628875, + "learning_rate": 2.5453755452794374e-06, + "loss": 0.7015, + "num_input_tokens_seen": 154867815, + "step": 4765 + }, + { + "epoch": 0.4298146728592686, + "flos": 25117698794400.0, + "grad_norm": 1.487962984111766, + "learning_rate": 2.5448134514661938e-06, + "loss": 0.7967, + "num_input_tokens_seen": 154897175, + "step": 4766 + }, + { + "epoch": 0.4299048563827389, + "flos": 23836628188800.0, + "grad_norm": 1.4690001254461453, + "learning_rate": 2.5442513111664623e-06, + "loss": 0.7489, + "num_input_tokens_seen": 154924160, + "step": 4767 + }, + { + "epoch": 0.42999503990620913, + "flos": 14153049626880.0, + "grad_norm": 2.1441893591670227, + "learning_rate": 2.5436891244282084e-06, + "loss": 0.7465, + "num_input_tokens_seen": 154948990, + "step": 4768 + }, + { + "epoch": 0.4300852234296794, + "flos": 19503104136960.0, + "grad_norm": 1.7075164594178116, + "learning_rate": 2.5431268912994004e-06, + "loss": 0.8521, + "num_input_tokens_seen": 154974905, + "step": 4769 + }, + { + "epoch": 0.43017540695314965, + "flos": 20821313066400.0, + "grad_norm": 1.7347773179826453, + "learning_rate": 2.5425646118280108e-06, + "loss": 0.725, + "num_input_tokens_seen": 155002980, + "step": 4770 + }, + { + "epoch": 0.43026559047661994, + "flos": 66624409350240.0, + "grad_norm": 0.591291129675719, + "learning_rate": 2.5420022860620172e-06, + "loss": 0.5617, + "num_input_tokens_seen": 155101895, + "step": 4771 + }, + { + "epoch": 0.4303557740000902, + "flos": 15391852481760.0, + "grad_norm": 1.8522322204041324, + "learning_rate": 2.5414399140493995e-06, + "loss": 0.6852, + "num_input_tokens_seen": 155128155, + "step": 4772 + }, + { + "epoch": 0.43044595752356046, + "flos": 37615544015040.0, + "grad_norm": 3.188460411042553, + "learning_rate": 2.5408774958381436e-06, + "loss": 0.6947, + "num_input_tokens_seen": 155159160, + "step": 4773 + }, + { + "epoch": 0.4305361410470307, + "flos": 22424028037920.0, + "grad_norm": 1.8807972613543096, + "learning_rate": 2.540315031476237e-06, + "loss": 0.789, + "num_input_tokens_seen": 155189675, + "step": 4774 + }, + { + "epoch": 0.430626324570501, + "flos": 21331139425440.0, + "grad_norm": 2.0885012298499994, + "learning_rate": 2.5397525210116737e-06, + "loss": 0.7619, + "num_input_tokens_seen": 155217390, + "step": 4775 + }, + { + "epoch": 0.4307165080939712, + "flos": 35394115256640.0, + "grad_norm": 1.7324782301070336, + "learning_rate": 2.539189964492448e-06, + "loss": 0.708, + "num_input_tokens_seen": 155249060, + "step": 4776 + }, + { + "epoch": 0.4308066916174415, + "flos": 28136842402080.0, + "grad_norm": 1.7217459490094054, + "learning_rate": 2.5386273619665613e-06, + "loss": 0.7936, + "num_input_tokens_seen": 155277410, + "step": 4777 + }, + { + "epoch": 0.43089687514091174, + "flos": 56329222159200.0, + "grad_norm": 0.6791125529521442, + "learning_rate": 2.5380647134820186e-06, + "loss": 0.5833, + "num_input_tokens_seen": 155356575, + "step": 4778 + }, + { + "epoch": 0.43098705866438203, + "flos": 19288488710880.0, + "grad_norm": 1.4127474829672102, + "learning_rate": 2.5375020190868277e-06, + "loss": 0.8046, + "num_input_tokens_seen": 155384640, + "step": 4779 + }, + { + "epoch": 0.43107724218785226, + "flos": 21767172343680.0, + "grad_norm": 1.5660932042580409, + "learning_rate": 2.536939278829001e-06, + "loss": 0.7482, + "num_input_tokens_seen": 155416165, + "step": 4780 + }, + { + "epoch": 0.43116742571132255, + "flos": 22861287558240.0, + "grad_norm": 2.1894750126094302, + "learning_rate": 2.5363764927565536e-06, + "loss": 0.8153, + "num_input_tokens_seen": 155443940, + "step": 4781 + }, + { + "epoch": 0.4312576092347928, + "flos": 26102257525440.0, + "grad_norm": 2.171538621938893, + "learning_rate": 2.5358136609175064e-06, + "loss": 0.7079, + "num_input_tokens_seen": 155476015, + "step": 4782 + }, + { + "epoch": 0.43134779275826307, + "flos": 20274868760160.0, + "grad_norm": 1.4815973896584471, + "learning_rate": 2.535250783359884e-06, + "loss": 0.7567, + "num_input_tokens_seen": 155503270, + "step": 4783 + }, + { + "epoch": 0.4314379762817333, + "flos": 29600296518240.0, + "grad_norm": 1.5215562208496198, + "learning_rate": 2.5346878601317124e-06, + "loss": 0.6495, + "num_input_tokens_seen": 155533775, + "step": 4784 + }, + { + "epoch": 0.4315281598052036, + "flos": 70669227177120.0, + "grad_norm": 0.5730813627780046, + "learning_rate": 2.534124891281025e-06, + "loss": 0.5461, + "num_input_tokens_seen": 155633555, + "step": 4785 + }, + { + "epoch": 0.4316183433286738, + "flos": 22022903257920.0, + "grad_norm": 1.7786224795623964, + "learning_rate": 2.533561876855857e-06, + "loss": 0.7644, + "num_input_tokens_seen": 155661525, + "step": 4786 + }, + { + "epoch": 0.4317085268521441, + "flos": 22095916133760.0, + "grad_norm": 7.455020303679359, + "learning_rate": 2.532998816904247e-06, + "loss": 0.7038, + "num_input_tokens_seen": 155690765, + "step": 4787 + }, + { + "epoch": 0.4317987103756144, + "flos": 40858149451680.0, + "grad_norm": 1.7787754203390032, + "learning_rate": 2.53243571147424e-06, + "loss": 0.6444, + "num_input_tokens_seen": 155723360, + "step": 4788 + }, + { + "epoch": 0.43188889389908464, + "flos": 22422206719680.0, + "grad_norm": 1.9835988126181578, + "learning_rate": 2.5318725606138815e-06, + "loss": 0.6619, + "num_input_tokens_seen": 155751080, + "step": 4789 + }, + { + "epoch": 0.4319790774225549, + "flos": 25297926458880.0, + "grad_norm": 1.8769770689200194, + "learning_rate": 2.5313093643712235e-06, + "loss": 0.7614, + "num_input_tokens_seen": 155781350, + "step": 4790 + }, + { + "epoch": 0.43206926094602516, + "flos": 22351424029440.0, + "grad_norm": 1.8440934200826669, + "learning_rate": 2.530746122794321e-06, + "loss": 0.802, + "num_input_tokens_seen": 155811185, + "step": 4791 + }, + { + "epoch": 0.43215944446949545, + "flos": 35866097066400.0, + "grad_norm": 1.5393729114711903, + "learning_rate": 2.5301828359312323e-06, + "loss": 0.6852, + "num_input_tokens_seen": 155843245, + "step": 4792 + }, + { + "epoch": 0.4322496279929657, + "flos": 29229656675040.0, + "grad_norm": 1.8341135907787702, + "learning_rate": 2.529619503830021e-06, + "loss": 0.712, + "num_input_tokens_seen": 155872665, + "step": 4793 + }, + { + "epoch": 0.43233981151643597, + "flos": 25808273194560.0, + "grad_norm": 2.5796816663485487, + "learning_rate": 2.529056126538753e-06, + "loss": 0.6819, + "num_input_tokens_seen": 155901205, + "step": 4794 + }, + { + "epoch": 0.4324299950399062, + "flos": 27488089715520.0, + "grad_norm": 1.8081626955145151, + "learning_rate": 2.5284927041054995e-06, + "loss": 0.7839, + "num_input_tokens_seen": 155931455, + "step": 4795 + }, + { + "epoch": 0.4325201785633765, + "flos": 55483658361600.0, + "grad_norm": 0.610270136170064, + "learning_rate": 2.5279292365783348e-06, + "loss": 0.5244, + "num_input_tokens_seen": 156026240, + "step": 4796 + }, + { + "epoch": 0.4326103620868467, + "flos": 27342212642880.0, + "grad_norm": 1.6531046384772516, + "learning_rate": 2.527365724005336e-06, + "loss": 0.8424, + "num_input_tokens_seen": 156054105, + "step": 4797 + }, + { + "epoch": 0.432700545610317, + "flos": 21148347047520.0, + "grad_norm": 1.8808102546845566, + "learning_rate": 2.526802166434586e-06, + "loss": 0.807, + "num_input_tokens_seen": 156081845, + "step": 4798 + }, + { + "epoch": 0.43279072913378724, + "flos": 23404944132480.0, + "grad_norm": 1.6113899727713787, + "learning_rate": 2.5262385639141708e-06, + "loss": 0.7214, + "num_input_tokens_seen": 156112060, + "step": 4799 + }, + { + "epoch": 0.43288091265725753, + "flos": 26500668912960.0, + "grad_norm": 1.549844300233159, + "learning_rate": 2.525674916492179e-06, + "loss": 0.824, + "num_input_tokens_seen": 156143460, + "step": 4800 + }, + { + "epoch": 0.43297109618072777, + "flos": 29018683885440.0, + "grad_norm": 1.3667591840938889, + "learning_rate": 2.5251112242167056e-06, + "loss": 0.759, + "num_input_tokens_seen": 156175880, + "step": 4801 + }, + { + "epoch": 0.43306127970419805, + "flos": 19506746773440.0, + "grad_norm": 1.8767151934830046, + "learning_rate": 2.5245474871358464e-06, + "loss": 0.7107, + "num_input_tokens_seen": 156204320, + "step": 4802 + }, + { + "epoch": 0.4331514632276683, + "flos": 17031334079520.0, + "grad_norm": 2.831557914377374, + "learning_rate": 2.5239837052977037e-06, + "loss": 0.6682, + "num_input_tokens_seen": 156231260, + "step": 4803 + }, + { + "epoch": 0.4332416467511386, + "flos": 18958927186080.0, + "grad_norm": 2.54967424518531, + "learning_rate": 2.523419878750381e-06, + "loss": 0.8644, + "num_input_tokens_seen": 156258460, + "step": 4804 + }, + { + "epoch": 0.4333318302746088, + "flos": 23586472738560.0, + "grad_norm": 2.071636424418046, + "learning_rate": 2.522856007541989e-06, + "loss": 0.706, + "num_input_tokens_seen": 156288530, + "step": 4805 + }, + { + "epoch": 0.4334220137980791, + "flos": 20412494146080.0, + "grad_norm": 1.8550476881411189, + "learning_rate": 2.5222920917206397e-06, + "loss": 0.6643, + "num_input_tokens_seen": 156315845, + "step": 4806 + }, + { + "epoch": 0.43351219732154933, + "flos": 70608002848800.0, + "grad_norm": 0.6446617225600716, + "learning_rate": 2.5217281313344493e-06, + "loss": 0.5616, + "num_input_tokens_seen": 156402375, + "step": 4807 + }, + { + "epoch": 0.4336023808450196, + "flos": 25119631621920.0, + "grad_norm": 1.678909192408863, + "learning_rate": 2.5211641264315372e-06, + "loss": 0.7996, + "num_input_tokens_seen": 156429820, + "step": 4808 + }, + { + "epoch": 0.43369256436848985, + "flos": 26246053091520.0, + "grad_norm": 1.7356065381407206, + "learning_rate": 2.5206000770600286e-06, + "loss": 0.7331, + "num_input_tokens_seen": 156460315, + "step": 4809 + }, + { + "epoch": 0.43378274789196014, + "flos": 25921695333120.0, + "grad_norm": 2.8629178470999066, + "learning_rate": 2.520035983268051e-06, + "loss": 0.7101, + "num_input_tokens_seen": 156490385, + "step": 4810 + }, + { + "epoch": 0.4338729314154304, + "flos": 33675562112160.0, + "grad_norm": 2.5752115144245256, + "learning_rate": 2.5194718451037357e-06, + "loss": 0.7524, + "num_input_tokens_seen": 156521165, + "step": 4811 + }, + { + "epoch": 0.43396311493890066, + "flos": 24464597245920.0, + "grad_norm": 1.8058980869547758, + "learning_rate": 2.518907662615218e-06, + "loss": 0.7804, + "num_input_tokens_seen": 156548485, + "step": 4812 + }, + { + "epoch": 0.43405329846237095, + "flos": 22059223847040.0, + "grad_norm": 1.879560709519555, + "learning_rate": 2.5183434358506373e-06, + "loss": 0.7139, + "num_input_tokens_seen": 156577000, + "step": 4813 + }, + { + "epoch": 0.4341434819858412, + "flos": 25265025487680.0, + "grad_norm": 3.2424616318596633, + "learning_rate": 2.5177791648581368e-06, + "loss": 0.7359, + "num_input_tokens_seen": 156607405, + "step": 4814 + }, + { + "epoch": 0.4342336655093115, + "flos": 22234656612480.0, + "grad_norm": 1.7997324241334245, + "learning_rate": 2.517214849685863e-06, + "loss": 0.7568, + "num_input_tokens_seen": 156633545, + "step": 4815 + }, + { + "epoch": 0.4343238490327817, + "flos": 18160914978720.0, + "grad_norm": 1.7538303282348056, + "learning_rate": 2.5166504903819663e-06, + "loss": 0.7703, + "num_input_tokens_seen": 156658705, + "step": 4816 + }, + { + "epoch": 0.434414032556252, + "flos": 22897310789280.0, + "grad_norm": 1.9335970791874748, + "learning_rate": 2.5160860869946014e-06, + "loss": 0.8191, + "num_input_tokens_seen": 156685135, + "step": 4817 + }, + { + "epoch": 0.4345042160797222, + "flos": 23698593935520.0, + "grad_norm": 1.6200191015129939, + "learning_rate": 2.5155216395719253e-06, + "loss": 0.8356, + "num_input_tokens_seen": 156715640, + "step": 4818 + }, + { + "epoch": 0.4345943996031925, + "flos": 26175196061760.0, + "grad_norm": 3.838267564584166, + "learning_rate": 2.5149571481621e-06, + "loss": 0.7146, + "num_input_tokens_seen": 156745880, + "step": 4819 + }, + { + "epoch": 0.43468458312666275, + "flos": 71544489612960.0, + "grad_norm": 0.6112401490695113, + "learning_rate": 2.514392612813292e-06, + "loss": 0.5435, + "num_input_tokens_seen": 156842310, + "step": 4820 + }, + { + "epoch": 0.43477476665013304, + "flos": 22715745013440.0, + "grad_norm": 4.177368661675299, + "learning_rate": 2.5138280335736695e-06, + "loss": 0.8054, + "num_input_tokens_seen": 156872625, + "step": 4821 + }, + { + "epoch": 0.43486495017360327, + "flos": 28831133778240.0, + "grad_norm": 3.4137444598985134, + "learning_rate": 2.5132634104914064e-06, + "loss": 0.8362, + "num_input_tokens_seen": 156902300, + "step": 4822 + }, + { + "epoch": 0.43495513369707356, + "flos": 27450728373120.0, + "grad_norm": 1.5655285285877167, + "learning_rate": 2.5126987436146794e-06, + "loss": 0.7661, + "num_input_tokens_seen": 156934210, + "step": 4823 + }, + { + "epoch": 0.4350453172205438, + "flos": 22388041976640.0, + "grad_norm": 1.6935015731997936, + "learning_rate": 2.5121340329916675e-06, + "loss": 0.7332, + "num_input_tokens_seen": 156961585, + "step": 4824 + }, + { + "epoch": 0.4351355007440141, + "flos": 20164866239520.0, + "grad_norm": 1.7033971866476492, + "learning_rate": 2.5115692786705566e-06, + "loss": 0.6735, + "num_input_tokens_seen": 156988740, + "step": 4825 + }, + { + "epoch": 0.4352256842674843, + "flos": 21803976139680.0, + "grad_norm": 1.8971935325279081, + "learning_rate": 2.511004480699534e-06, + "loss": 0.7916, + "num_input_tokens_seen": 157015530, + "step": 4826 + }, + { + "epoch": 0.4353158677909546, + "flos": 69773409864000.0, + "grad_norm": 0.7210246457002589, + "learning_rate": 2.510439639126791e-06, + "loss": 0.6189, + "num_input_tokens_seen": 157104015, + "step": 4827 + }, + { + "epoch": 0.43540605131442484, + "flos": 24537312763680.0, + "grad_norm": 1.5629922472907352, + "learning_rate": 2.509874754000524e-06, + "loss": 0.7525, + "num_input_tokens_seen": 157134110, + "step": 4828 + }, + { + "epoch": 0.4354962348378951, + "flos": 39363727191840.0, + "grad_norm": 2.0694518405631372, + "learning_rate": 2.509309825368932e-06, + "loss": 0.7377, + "num_input_tokens_seen": 157161655, + "step": 4829 + }, + { + "epoch": 0.43558641836136536, + "flos": 23006235386880.0, + "grad_norm": 2.3987641251919927, + "learning_rate": 2.5087448532802173e-06, + "loss": 0.7502, + "num_input_tokens_seen": 157191010, + "step": 4830 + }, + { + "epoch": 0.43567660188483565, + "flos": 23254829707200.0, + "grad_norm": 2.0558616851265996, + "learning_rate": 2.508179837782586e-06, + "loss": 0.7324, + "num_input_tokens_seen": 157221085, + "step": 4831 + }, + { + "epoch": 0.4357667854083059, + "flos": 26248506295680.0, + "grad_norm": 1.6189314427554253, + "learning_rate": 2.5076147789242493e-06, + "loss": 0.7238, + "num_input_tokens_seen": 157249905, + "step": 4832 + }, + { + "epoch": 0.43585696893177617, + "flos": 25520644892640.0, + "grad_norm": 1.6584547950442303, + "learning_rate": 2.5070496767534202e-06, + "loss": 0.7696, + "num_input_tokens_seen": 157279500, + "step": 4833 + }, + { + "epoch": 0.4359471524552464, + "flos": 22132199553120.0, + "grad_norm": 1.8620990705675138, + "learning_rate": 2.506484531318317e-06, + "loss": 0.7777, + "num_input_tokens_seen": 157307390, + "step": 4834 + }, + { + "epoch": 0.4360373359787167, + "flos": 66476079073440.0, + "grad_norm": 0.7607958166295229, + "learning_rate": 2.5059193426671613e-06, + "loss": 0.5907, + "num_input_tokens_seen": 157404450, + "step": 4835 + }, + { + "epoch": 0.436127519502187, + "flos": 28762544103840.0, + "grad_norm": 1.746795445031337, + "learning_rate": 2.5053541108481772e-06, + "loss": 0.8849, + "num_input_tokens_seen": 157433370, + "step": 4836 + }, + { + "epoch": 0.4362177030256572, + "flos": 17577927064800.0, + "grad_norm": 2.1278941697285365, + "learning_rate": 2.5047888359095935e-06, + "loss": 0.7356, + "num_input_tokens_seen": 157460195, + "step": 4837 + }, + { + "epoch": 0.4363078865491275, + "flos": 20273902346400.0, + "grad_norm": 2.3389296744327606, + "learning_rate": 2.5042235178996436e-06, + "loss": 0.7344, + "num_input_tokens_seen": 157484915, + "step": 4838 + }, + { + "epoch": 0.43639807007259773, + "flos": 19436224271520.0, + "grad_norm": 1.708468881517354, + "learning_rate": 2.5036581568665627e-06, + "loss": 0.8157, + "num_input_tokens_seen": 157511900, + "step": 4839 + }, + { + "epoch": 0.436488253596068, + "flos": 22460534475840.0, + "grad_norm": 1.9497745310207066, + "learning_rate": 2.503092752858591e-06, + "loss": 0.8178, + "num_input_tokens_seen": 157537035, + "step": 4840 + }, + { + "epoch": 0.43657843711953825, + "flos": 69642326355840.0, + "grad_norm": 0.7781243167246075, + "learning_rate": 2.502527305923971e-06, + "loss": 0.6074, + "num_input_tokens_seen": 157626770, + "step": 4841 + }, + { + "epoch": 0.43666862064300854, + "flos": 22453323542400.0, + "grad_norm": 2.0716962688333083, + "learning_rate": 2.5019618161109506e-06, + "loss": 0.7374, + "num_input_tokens_seen": 157648065, + "step": 4842 + }, + { + "epoch": 0.4367588041664788, + "flos": 28981396882560.0, + "grad_norm": 1.801478054585274, + "learning_rate": 2.5013962834677804e-06, + "loss": 0.7712, + "num_input_tokens_seen": 157678475, + "step": 4843 + }, + { + "epoch": 0.43684898768994906, + "flos": 23844322329120.0, + "grad_norm": 4.33857012033768, + "learning_rate": 2.500830708042715e-06, + "loss": 0.7395, + "num_input_tokens_seen": 157705115, + "step": 4844 + }, + { + "epoch": 0.4369391712134193, + "flos": 23589780847200.0, + "grad_norm": 1.8920192779179883, + "learning_rate": 2.500265089884011e-06, + "loss": 0.8165, + "num_input_tokens_seen": 157734925, + "step": 4845 + }, + { + "epoch": 0.4370293547368896, + "flos": 30547382397600.0, + "grad_norm": 1.5253527217752052, + "learning_rate": 2.499699429039932e-06, + "loss": 0.7517, + "num_input_tokens_seen": 157765485, + "step": 4846 + }, + { + "epoch": 0.4371195382603598, + "flos": 27706459287360.0, + "grad_norm": 1.707131738563808, + "learning_rate": 2.4991337255587425e-06, + "loss": 0.7042, + "num_input_tokens_seen": 157795115, + "step": 4847 + }, + { + "epoch": 0.4372097217838301, + "flos": 28937828190240.0, + "grad_norm": 1.8740401036175935, + "learning_rate": 2.4985679794887106e-06, + "loss": 0.6237, + "num_input_tokens_seen": 157823740, + "step": 4848 + }, + { + "epoch": 0.43729990530730034, + "flos": 23006830103040.0, + "grad_norm": 1.9076112358287096, + "learning_rate": 2.49800219087811e-06, + "loss": 0.7874, + "num_input_tokens_seen": 157850755, + "step": 4849 + }, + { + "epoch": 0.43739008883077063, + "flos": 14954332773120.0, + "grad_norm": 2.3993165831274865, + "learning_rate": 2.4974363597752163e-06, + "loss": 0.7869, + "num_input_tokens_seen": 157875695, + "step": 4850 + }, + { + "epoch": 0.43748027235424086, + "flos": 21257903531040.0, + "grad_norm": 2.333682806092791, + "learning_rate": 2.4968704862283097e-06, + "loss": 0.7977, + "num_input_tokens_seen": 157904955, + "step": 4851 + }, + { + "epoch": 0.43757045587771115, + "flos": 24281321661120.0, + "grad_norm": 1.651173856078404, + "learning_rate": 2.4963045702856737e-06, + "loss": 0.8215, + "num_input_tokens_seen": 157935370, + "step": 4852 + }, + { + "epoch": 0.4376606394011814, + "flos": 28471310335200.0, + "grad_norm": 1.5483202107465055, + "learning_rate": 2.4957386119955946e-06, + "loss": 0.697, + "num_input_tokens_seen": 157964375, + "step": 4853 + }, + { + "epoch": 0.4377508229246517, + "flos": 68445605402880.0, + "grad_norm": 0.686778969234594, + "learning_rate": 2.495172611406364e-06, + "loss": 0.5679, + "num_input_tokens_seen": 158064570, + "step": 4854 + }, + { + "epoch": 0.4378410064481219, + "flos": 23770937755680.0, + "grad_norm": 7.2189798690065, + "learning_rate": 2.4946065685662757e-06, + "loss": 0.7789, + "num_input_tokens_seen": 158092870, + "step": 4855 + }, + { + "epoch": 0.4379311899715922, + "flos": 27192915952320.0, + "grad_norm": 2.274580939965341, + "learning_rate": 2.4940404835236283e-06, + "loss": 0.7552, + "num_input_tokens_seen": 158121580, + "step": 4856 + }, + { + "epoch": 0.4380213734950624, + "flos": 32806469856480.0, + "grad_norm": 1.7428510283311527, + "learning_rate": 2.4934743563267223e-06, + "loss": 0.7423, + "num_input_tokens_seen": 158153885, + "step": 4857 + }, + { + "epoch": 0.4381115570185327, + "flos": 21546981453600.0, + "grad_norm": 1.9945246133421215, + "learning_rate": 2.4929081870238635e-06, + "loss": 0.7158, + "num_input_tokens_seen": 158183215, + "step": 4858 + }, + { + "epoch": 0.43820174054200295, + "flos": 16411765388160.0, + "grad_norm": 1.7004429408446375, + "learning_rate": 2.49234197566336e-06, + "loss": 0.7917, + "num_input_tokens_seen": 158209735, + "step": 4859 + }, + { + "epoch": 0.43829192406547324, + "flos": 28833140945280.0, + "grad_norm": 1.9193083208650954, + "learning_rate": 2.4917757222935247e-06, + "loss": 0.7718, + "num_input_tokens_seen": 158240435, + "step": 4860 + }, + { + "epoch": 0.4383821075889435, + "flos": 21875948262240.0, + "grad_norm": 1.782030837925684, + "learning_rate": 2.4912094269626725e-06, + "loss": 0.7747, + "num_input_tokens_seen": 158269300, + "step": 4861 + }, + { + "epoch": 0.43847229111241376, + "flos": 13533852633120.0, + "grad_norm": 1.9444438419643861, + "learning_rate": 2.4906430897191245e-06, + "loss": 0.7928, + "num_input_tokens_seen": 158296000, + "step": 4862 + }, + { + "epoch": 0.43856247463588405, + "flos": 25625109119040.0, + "grad_norm": 1.641803136791384, + "learning_rate": 2.490076710611202e-06, + "loss": 0.7874, + "num_input_tokens_seen": 158323560, + "step": 4863 + }, + { + "epoch": 0.4386526581593543, + "flos": 19400163870720.0, + "grad_norm": 2.1378126395395136, + "learning_rate": 2.4895102896872326e-06, + "loss": 0.7999, + "num_input_tokens_seen": 158349350, + "step": 4864 + }, + { + "epoch": 0.43874284168282457, + "flos": 23624354457600.0, + "grad_norm": 1.9215785472458506, + "learning_rate": 2.4889438269955457e-06, + "loss": 0.6843, + "num_input_tokens_seen": 158379205, + "step": 4865 + }, + { + "epoch": 0.4388330252062948, + "flos": 17360003530080.0, + "grad_norm": 1.5009945909935094, + "learning_rate": 2.4883773225844755e-06, + "loss": 0.7327, + "num_input_tokens_seen": 158407090, + "step": 4866 + }, + { + "epoch": 0.4389232087297651, + "flos": 15902905442880.0, + "grad_norm": 1.7381248146559072, + "learning_rate": 2.48781077650236e-06, + "loss": 0.7445, + "num_input_tokens_seen": 158432555, + "step": 4867 + }, + { + "epoch": 0.4390133922532353, + "flos": 28977940094880.0, + "grad_norm": 2.905285911212718, + "learning_rate": 2.4872441887975386e-06, + "loss": 0.7242, + "num_input_tokens_seen": 158463810, + "step": 4868 + }, + { + "epoch": 0.4391035757767056, + "flos": 26905138971360.0, + "grad_norm": 1.3448609040527926, + "learning_rate": 2.486677559518356e-06, + "loss": 0.6923, + "num_input_tokens_seen": 158491760, + "step": 4869 + }, + { + "epoch": 0.43919375930017585, + "flos": 24824643707520.0, + "grad_norm": 1.3417394367567708, + "learning_rate": 2.4861108887131614e-06, + "loss": 0.7398, + "num_input_tokens_seen": 158522095, + "step": 4870 + }, + { + "epoch": 0.43928394282364613, + "flos": 33130678935840.0, + "grad_norm": 2.018286197635912, + "learning_rate": 2.485544176430305e-06, + "loss": 0.7197, + "num_input_tokens_seen": 158554235, + "step": 4871 + }, + { + "epoch": 0.43937412634711637, + "flos": 25551687375840.0, + "grad_norm": 1.7914981836751067, + "learning_rate": 2.4849774227181425e-06, + "loss": 0.8312, + "num_input_tokens_seen": 158580950, + "step": 4872 + }, + { + "epoch": 0.43946430987058666, + "flos": 25302089472000.0, + "grad_norm": 1.501382984938273, + "learning_rate": 2.484410627625032e-06, + "loss": 0.751, + "num_input_tokens_seen": 158610440, + "step": 4873 + }, + { + "epoch": 0.4395544933940569, + "flos": 17099552056320.0, + "grad_norm": 1.8870859361415688, + "learning_rate": 2.4838437911993356e-06, + "loss": 0.7907, + "num_input_tokens_seen": 158635225, + "step": 4874 + }, + { + "epoch": 0.4396446769175272, + "flos": 23147057372160.0, + "grad_norm": 4.570012635395104, + "learning_rate": 2.483276913489419e-06, + "loss": 0.7922, + "num_input_tokens_seen": 158662320, + "step": 4875 + }, + { + "epoch": 0.4397348604409974, + "flos": 37870643043360.0, + "grad_norm": 1.8587256828518344, + "learning_rate": 2.4827099945436516e-06, + "loss": 0.6506, + "num_input_tokens_seen": 158690140, + "step": 4876 + }, + { + "epoch": 0.4398250439644677, + "flos": 13460468059680.0, + "grad_norm": 2.40702371419676, + "learning_rate": 2.482143034410405e-06, + "loss": 0.8342, + "num_input_tokens_seen": 158713075, + "step": 4877 + }, + { + "epoch": 0.43991522748793793, + "flos": 24714678356640.0, + "grad_norm": 1.5738440888921454, + "learning_rate": 2.4815760331380573e-06, + "loss": 0.7816, + "num_input_tokens_seen": 158743330, + "step": 4878 + }, + { + "epoch": 0.4400054110114082, + "flos": 27160423848480.0, + "grad_norm": 1.4825516594929113, + "learning_rate": 2.481008990774987e-06, + "loss": 0.7118, + "num_input_tokens_seen": 158774000, + "step": 4879 + }, + { + "epoch": 0.44009559453487845, + "flos": 14584250476320.0, + "grad_norm": 1.7956392766037752, + "learning_rate": 2.480441907369577e-06, + "loss": 0.7505, + "num_input_tokens_seen": 158799025, + "step": 4880 + }, + { + "epoch": 0.44018577805834874, + "flos": 32368727129280.0, + "grad_norm": 1.6533608338673542, + "learning_rate": 2.479874782970214e-06, + "loss": 0.6953, + "num_input_tokens_seen": 158832335, + "step": 4881 + }, + { + "epoch": 0.440275961581819, + "flos": 21912900737280.0, + "grad_norm": 2.1877523004264536, + "learning_rate": 2.4793076176252887e-06, + "loss": 0.7706, + "num_input_tokens_seen": 158861830, + "step": 4882 + }, + { + "epoch": 0.44036614510528926, + "flos": 25156063720320.0, + "grad_norm": 1.7896959940657315, + "learning_rate": 2.478740411383195e-06, + "loss": 0.6362, + "num_input_tokens_seen": 158894485, + "step": 4883 + }, + { + "epoch": 0.44045632862875955, + "flos": 26248431956160.0, + "grad_norm": 1.7652006061225298, + "learning_rate": 2.4781731642923296e-06, + "loss": 0.7434, + "num_input_tokens_seen": 158924945, + "step": 4884 + }, + { + "epoch": 0.4405465121522298, + "flos": 24866391081600.0, + "grad_norm": 2.6608421572921137, + "learning_rate": 2.477605876401093e-06, + "loss": 0.737, + "num_input_tokens_seen": 158955045, + "step": 4885 + }, + { + "epoch": 0.4406366956757001, + "flos": 25521983004000.0, + "grad_norm": 1.7654990092198852, + "learning_rate": 2.4770385477578894e-06, + "loss": 0.7097, + "num_input_tokens_seen": 158988725, + "step": 4886 + }, + { + "epoch": 0.4407268791991703, + "flos": 26537770067040.0, + "grad_norm": 1.4008678011688411, + "learning_rate": 2.476471178411127e-06, + "loss": 0.7066, + "num_input_tokens_seen": 159020085, + "step": 4887 + }, + { + "epoch": 0.4408170627226406, + "flos": 23877037451520.0, + "grad_norm": 3.1323048472827537, + "learning_rate": 2.475903768409216e-06, + "loss": 0.7712, + "num_input_tokens_seen": 159048495, + "step": 4888 + }, + { + "epoch": 0.44090724624611083, + "flos": 20419965267840.0, + "grad_norm": 1.8252725763433388, + "learning_rate": 2.475336317800572e-06, + "loss": 0.728, + "num_input_tokens_seen": 159076850, + "step": 4889 + }, + { + "epoch": 0.4409974297695811, + "flos": 23006161047360.0, + "grad_norm": 1.670204254725284, + "learning_rate": 2.4747688266336118e-06, + "loss": 0.7386, + "num_input_tokens_seen": 159103130, + "step": 4890 + }, + { + "epoch": 0.44108761329305135, + "flos": 20887152178560.0, + "grad_norm": 2.1610307630602628, + "learning_rate": 2.4742012949567574e-06, + "loss": 0.8544, + "num_input_tokens_seen": 159130610, + "step": 4891 + }, + { + "epoch": 0.44117779681652164, + "flos": 23188507388160.0, + "grad_norm": 4.697000463539704, + "learning_rate": 2.4736337228184338e-06, + "loss": 0.7333, + "num_input_tokens_seen": 159159525, + "step": 4892 + }, + { + "epoch": 0.44126798033999187, + "flos": 22022903257920.0, + "grad_norm": 11.80182855953402, + "learning_rate": 2.4730661102670692e-06, + "loss": 0.8504, + "num_input_tokens_seen": 159186735, + "step": 4893 + }, + { + "epoch": 0.44135816386346216, + "flos": 28978386132000.0, + "grad_norm": 1.8766894041121551, + "learning_rate": 2.472498457351096e-06, + "loss": 0.7567, + "num_input_tokens_seen": 159217090, + "step": 4894 + }, + { + "epoch": 0.4414483473869324, + "flos": 25447037300640.0, + "grad_norm": 2.0834393919333527, + "learning_rate": 2.4719307641189495e-06, + "loss": 0.7386, + "num_input_tokens_seen": 159244090, + "step": 4895 + }, + { + "epoch": 0.4415385309104027, + "flos": 26644687497600.0, + "grad_norm": 2.9711181832210904, + "learning_rate": 2.4713630306190673e-06, + "loss": 0.7152, + "num_input_tokens_seen": 159271710, + "step": 4896 + }, + { + "epoch": 0.4416287144338729, + "flos": 20783803044960.0, + "grad_norm": 2.0818095420182607, + "learning_rate": 2.4707952568998923e-06, + "loss": 0.8335, + "num_input_tokens_seen": 159298645, + "step": 4897 + }, + { + "epoch": 0.4417188979573432, + "flos": 20456806233600.0, + "grad_norm": 2.5283519968321766, + "learning_rate": 2.4702274430098703e-06, + "loss": 0.7241, + "num_input_tokens_seen": 159322820, + "step": 4898 + }, + { + "epoch": 0.44180908148081344, + "flos": 19509125638080.0, + "grad_norm": 2.289043856079418, + "learning_rate": 2.4696595889974497e-06, + "loss": 0.7051, + "num_input_tokens_seen": 159350770, + "step": 4899 + }, + { + "epoch": 0.4418992650042837, + "flos": 23765399461440.0, + "grad_norm": 1.62917437279221, + "learning_rate": 2.469091694911084e-06, + "loss": 0.8085, + "num_input_tokens_seen": 159377715, + "step": 4900 + }, + { + "epoch": 0.44198944852775396, + "flos": 22491465449760.0, + "grad_norm": 1.660493847797395, + "learning_rate": 2.4685237607992276e-06, + "loss": 0.751, + "num_input_tokens_seen": 159406980, + "step": 4901 + }, + { + "epoch": 0.44207963205122425, + "flos": 21695088711840.0, + "grad_norm": 1.6821388986232602, + "learning_rate": 2.4679557867103416e-06, + "loss": 0.8163, + "num_input_tokens_seen": 159432355, + "step": 4902 + }, + { + "epoch": 0.4421698155746945, + "flos": 22312315708320.0, + "grad_norm": 3.076238683550317, + "learning_rate": 2.4673877726928865e-06, + "loss": 0.7596, + "num_input_tokens_seen": 159462470, + "step": 4903 + }, + { + "epoch": 0.44225999909816477, + "flos": 24318125457120.0, + "grad_norm": 2.161930838418232, + "learning_rate": 2.46681971879533e-06, + "loss": 0.6414, + "num_input_tokens_seen": 159491960, + "step": 4904 + }, + { + "epoch": 0.442350182621635, + "flos": 14984446012320.0, + "grad_norm": 1.7844165296114338, + "learning_rate": 2.4662516250661407e-06, + "loss": 0.6752, + "num_input_tokens_seen": 159517365, + "step": 4905 + }, + { + "epoch": 0.4424403661451053, + "flos": 20346617864160.0, + "grad_norm": 1.873961478953949, + "learning_rate": 2.465683491553792e-06, + "loss": 0.7255, + "num_input_tokens_seen": 159542230, + "step": 4906 + }, + { + "epoch": 0.4425305496685755, + "flos": 20565619321920.0, + "grad_norm": 1.80836449111371, + "learning_rate": 2.4651153183067604e-06, + "loss": 0.7565, + "num_input_tokens_seen": 159570630, + "step": 4907 + }, + { + "epoch": 0.4426207331920458, + "flos": 23842017804000.0, + "grad_norm": 1.5306012862001341, + "learning_rate": 2.4645471053735245e-06, + "loss": 0.6929, + "num_input_tokens_seen": 159600955, + "step": 4908 + }, + { + "epoch": 0.4427109167155161, + "flos": 22388116316160.0, + "grad_norm": 1.6916494569075302, + "learning_rate": 2.4639788528025684e-06, + "loss": 0.6896, + "num_input_tokens_seen": 159630040, + "step": 4909 + }, + { + "epoch": 0.44280110023898633, + "flos": 26649519566400.0, + "grad_norm": 1.2753016862922575, + "learning_rate": 2.463410560642378e-06, + "loss": 0.7561, + "num_input_tokens_seen": 159662465, + "step": 4910 + }, + { + "epoch": 0.4428912837624566, + "flos": 62651675155200.0, + "grad_norm": 0.631435332144186, + "learning_rate": 2.4628422289414448e-06, + "loss": 0.588, + "num_input_tokens_seen": 159757325, + "step": 4911 + }, + { + "epoch": 0.44298146728592686, + "flos": 70892763345600.0, + "grad_norm": 0.66563648607973, + "learning_rate": 2.4622738577482592e-06, + "loss": 0.563, + "num_input_tokens_seen": 159850890, + "step": 4912 + }, + { + "epoch": 0.44307165080939714, + "flos": 38161505114400.0, + "grad_norm": 1.550399387341586, + "learning_rate": 2.461705447111319e-06, + "loss": 0.7385, + "num_input_tokens_seen": 159884660, + "step": 4913 + }, + { + "epoch": 0.4431618343328674, + "flos": 25884742858080.0, + "grad_norm": 2.112442047407693, + "learning_rate": 2.4611369970791246e-06, + "loss": 0.7253, + "num_input_tokens_seen": 159914805, + "step": 4914 + }, + { + "epoch": 0.44325201785633767, + "flos": 19873186433760.0, + "grad_norm": 1.66048070990693, + "learning_rate": 2.460568507700179e-06, + "loss": 0.748, + "num_input_tokens_seen": 159943880, + "step": 4915 + }, + { + "epoch": 0.4433422013798079, + "flos": 19143094845120.0, + "grad_norm": 1.7454641004819225, + "learning_rate": 2.4599999790229887e-06, + "loss": 0.796, + "num_input_tokens_seen": 159969565, + "step": 4916 + }, + { + "epoch": 0.4434323849032782, + "flos": 20638966725600.0, + "grad_norm": 2.5121312492366146, + "learning_rate": 2.459431411096064e-06, + "loss": 0.7137, + "num_input_tokens_seen": 159997235, + "step": 4917 + }, + { + "epoch": 0.4435225684267484, + "flos": 29598029162880.0, + "grad_norm": 1.7368000706463997, + "learning_rate": 2.458862803967918e-06, + "loss": 0.7196, + "num_input_tokens_seen": 160027885, + "step": 4918 + }, + { + "epoch": 0.4436127519502187, + "flos": 18525533320800.0, + "grad_norm": 1.693897694134259, + "learning_rate": 2.4582941576870667e-06, + "loss": 0.7688, + "num_input_tokens_seen": 160055290, + "step": 4919 + }, + { + "epoch": 0.44370293547368894, + "flos": 20346915222240.0, + "grad_norm": 1.7402830106401668, + "learning_rate": 2.4577254723020315e-06, + "loss": 0.7328, + "num_input_tokens_seen": 160083870, + "step": 4920 + }, + { + "epoch": 0.44379311899715923, + "flos": 21982196637120.0, + "grad_norm": 1.7642125919847096, + "learning_rate": 2.457156747861335e-06, + "loss": 0.7701, + "num_input_tokens_seen": 160112250, + "step": 4921 + }, + { + "epoch": 0.44388330252062946, + "flos": 20492643615840.0, + "grad_norm": 2.8735676390076756, + "learning_rate": 2.456587984413504e-06, + "loss": 0.8179, + "num_input_tokens_seen": 160135330, + "step": 4922 + }, + { + "epoch": 0.44397348604409975, + "flos": 25228704898560.0, + "grad_norm": 1.6936694502592096, + "learning_rate": 2.4560191820070683e-06, + "loss": 0.6841, + "num_input_tokens_seen": 160167355, + "step": 4923 + }, + { + "epoch": 0.44406366956757, + "flos": 24828658041600.0, + "grad_norm": 2.1507225508046246, + "learning_rate": 2.4554503406905617e-06, + "loss": 0.7332, + "num_input_tokens_seen": 160196715, + "step": 4924 + }, + { + "epoch": 0.4441538530910403, + "flos": 28033864966080.0, + "grad_norm": 2.3808205204735633, + "learning_rate": 2.454881460512521e-06, + "loss": 0.7901, + "num_input_tokens_seen": 160223640, + "step": 4925 + }, + { + "epoch": 0.4442440366145105, + "flos": 22896790412640.0, + "grad_norm": 1.7750461357460825, + "learning_rate": 2.4543125415214856e-06, + "loss": 0.7174, + "num_input_tokens_seen": 160255785, + "step": 4926 + }, + { + "epoch": 0.4443342201379808, + "flos": 26394829405440.0, + "grad_norm": 1.726845489371295, + "learning_rate": 2.4537435837659996e-06, + "loss": 0.7769, + "num_input_tokens_seen": 160285025, + "step": 4927 + }, + { + "epoch": 0.44442440366145103, + "flos": 24347495301120.0, + "grad_norm": 2.5229477540003247, + "learning_rate": 2.4531745872946085e-06, + "loss": 0.7072, + "num_input_tokens_seen": 160313490, + "step": 4928 + }, + { + "epoch": 0.4445145871849213, + "flos": 13639023084960.0, + "grad_norm": 2.01871850429094, + "learning_rate": 2.4526055521558632e-06, + "loss": 0.8157, + "num_input_tokens_seen": 160339420, + "step": 4929 + }, + { + "epoch": 0.44460477070839155, + "flos": 22205138089440.0, + "grad_norm": 1.6992228190055447, + "learning_rate": 2.4520364783983164e-06, + "loss": 0.7698, + "num_input_tokens_seen": 160368355, + "step": 4930 + }, + { + "epoch": 0.44469495423186184, + "flos": 23298361229760.0, + "grad_norm": 1.6234422700167532, + "learning_rate": 2.451467366070525e-06, + "loss": 0.7333, + "num_input_tokens_seen": 160397370, + "step": 4931 + }, + { + "epoch": 0.4447851377553321, + "flos": 14734959617760.0, + "grad_norm": 2.852290815595162, + "learning_rate": 2.450898215221048e-06, + "loss": 0.8214, + "num_input_tokens_seen": 160422110, + "step": 4932 + }, + { + "epoch": 0.44487532127880236, + "flos": 19909395513600.0, + "grad_norm": 2.4511925928220557, + "learning_rate": 2.4503290258984498e-06, + "loss": 0.8951, + "num_input_tokens_seen": 160447540, + "step": 4933 + }, + { + "epoch": 0.44496550480227265, + "flos": 25448486921280.0, + "grad_norm": 1.969754220279453, + "learning_rate": 2.4497597981512952e-06, + "loss": 0.6164, + "num_input_tokens_seen": 160476445, + "step": 4934 + }, + { + "epoch": 0.4450556883257429, + "flos": 66213954960480.0, + "grad_norm": 0.6410584495322006, + "learning_rate": 2.4491905320281555e-06, + "loss": 0.5915, + "num_input_tokens_seen": 160570715, + "step": 4935 + }, + { + "epoch": 0.44514587184921317, + "flos": 21803864630400.0, + "grad_norm": 1.9180906921281797, + "learning_rate": 2.448621227577602e-06, + "loss": 0.7258, + "num_input_tokens_seen": 160600450, + "step": 4936 + }, + { + "epoch": 0.4452360553726834, + "flos": 26212260046080.0, + "grad_norm": 1.7897274330870225, + "learning_rate": 2.4480518848482123e-06, + "loss": 0.6979, + "num_input_tokens_seen": 160629490, + "step": 4937 + }, + { + "epoch": 0.4453262388961537, + "flos": 22749463719360.0, + "grad_norm": 1.6902626876275875, + "learning_rate": 2.447482503888565e-06, + "loss": 0.7639, + "num_input_tokens_seen": 160658800, + "step": 4938 + }, + { + "epoch": 0.4454164224196239, + "flos": 19837200372480.0, + "grad_norm": 2.1401245200392416, + "learning_rate": 2.4469130847472434e-06, + "loss": 0.728, + "num_input_tokens_seen": 160685645, + "step": 4939 + }, + { + "epoch": 0.4455066059430942, + "flos": 22460162778240.0, + "grad_norm": 1.8643633181659638, + "learning_rate": 2.4463436274728326e-06, + "loss": 0.7924, + "num_input_tokens_seen": 160714010, + "step": 4940 + }, + { + "epoch": 0.44559678946656445, + "flos": 30874490718240.0, + "grad_norm": 1.6144827370428054, + "learning_rate": 2.4457741321139227e-06, + "loss": 0.5989, + "num_input_tokens_seen": 160747190, + "step": 4941 + }, + { + "epoch": 0.44568697299003474, + "flos": 19108409725440.0, + "grad_norm": 1.9624813734980784, + "learning_rate": 2.4452045987191063e-06, + "loss": 0.7748, + "num_input_tokens_seen": 160773950, + "step": 4942 + }, + { + "epoch": 0.44577715651350497, + "flos": 21876877506240.0, + "grad_norm": 1.4784312054055209, + "learning_rate": 2.4446350273369776e-06, + "loss": 0.7242, + "num_input_tokens_seen": 160804830, + "step": 4943 + }, + { + "epoch": 0.44586734003697526, + "flos": 26212408725120.0, + "grad_norm": 1.5282294395657494, + "learning_rate": 2.4440654180161374e-06, + "loss": 0.7894, + "num_input_tokens_seen": 160834590, + "step": 4944 + }, + { + "epoch": 0.4459575235604455, + "flos": 33096588532320.0, + "grad_norm": 1.4714758718562388, + "learning_rate": 2.4434957708051875e-06, + "loss": 0.7066, + "num_input_tokens_seen": 160868905, + "step": 4945 + }, + { + "epoch": 0.4460477070839158, + "flos": 29526205719360.0, + "grad_norm": 1.827427657713773, + "learning_rate": 2.4429260857527324e-06, + "loss": 0.6942, + "num_input_tokens_seen": 160900205, + "step": 4946 + }, + { + "epoch": 0.446137890607386, + "flos": 33025731502560.0, + "grad_norm": 2.2988252210315134, + "learning_rate": 2.4423563629073815e-06, + "loss": 0.6611, + "num_input_tokens_seen": 160930805, + "step": 4947 + }, + { + "epoch": 0.4462280741308563, + "flos": 26211925518240.0, + "grad_norm": 1.5998755525829005, + "learning_rate": 2.4417866023177466e-06, + "loss": 0.7542, + "num_input_tokens_seen": 160959910, + "step": 4948 + }, + { + "epoch": 0.44631825765432653, + "flos": 22825078478400.0, + "grad_norm": 2.7342078457558276, + "learning_rate": 2.441216804032443e-06, + "loss": 0.786, + "num_input_tokens_seen": 160988635, + "step": 4949 + }, + { + "epoch": 0.4464084411777968, + "flos": 22716079541280.0, + "grad_norm": 1.7254203367190406, + "learning_rate": 2.440646968100089e-06, + "loss": 0.7662, + "num_input_tokens_seen": 161018890, + "step": 4950 + }, + { + "epoch": 0.44649862470126706, + "flos": 59170293948480.0, + "grad_norm": 0.6847879278856355, + "learning_rate": 2.4400770945693055e-06, + "loss": 0.5732, + "num_input_tokens_seen": 161105095, + "step": 4951 + }, + { + "epoch": 0.44658880822473734, + "flos": 20674246561440.0, + "grad_norm": 2.0926844604203976, + "learning_rate": 2.4395071834887177e-06, + "loss": 0.7829, + "num_input_tokens_seen": 161133205, + "step": 4952 + }, + { + "epoch": 0.4466789917482076, + "flos": 61221828235680.0, + "grad_norm": 0.6970486557135614, + "learning_rate": 2.438937234906954e-06, + "loss": 0.6346, + "num_input_tokens_seen": 161224805, + "step": 4953 + }, + { + "epoch": 0.44676917527167787, + "flos": 22708273891680.0, + "grad_norm": 1.63424240362168, + "learning_rate": 2.4383672488726447e-06, + "loss": 0.739, + "num_input_tokens_seen": 161254250, + "step": 4954 + }, + { + "epoch": 0.44685935879514815, + "flos": 18996734565600.0, + "grad_norm": 1.9836748945684897, + "learning_rate": 2.4377972254344256e-06, + "loss": 0.6735, + "num_input_tokens_seen": 161281465, + "step": 4955 + }, + { + "epoch": 0.4469495423186184, + "flos": 26248766484000.0, + "grad_norm": 1.8088695076964572, + "learning_rate": 2.437227164640932e-06, + "loss": 0.6378, + "num_input_tokens_seen": 161308870, + "step": 4956 + }, + { + "epoch": 0.4470397258420887, + "flos": 23553646106880.0, + "grad_norm": 1.6914776441205754, + "learning_rate": 2.436657066540807e-06, + "loss": 0.6937, + "num_input_tokens_seen": 161338095, + "step": 4957 + }, + { + "epoch": 0.4471299093655589, + "flos": 18853496545920.0, + "grad_norm": 2.368557675085215, + "learning_rate": 2.4360869311826927e-06, + "loss": 0.6287, + "num_input_tokens_seen": 161364535, + "step": 4958 + }, + { + "epoch": 0.4472200928890292, + "flos": 19144767484320.0, + "grad_norm": 3.5363187617240452, + "learning_rate": 2.4355167586152367e-06, + "loss": 0.8434, + "num_input_tokens_seen": 161390290, + "step": 4959 + }, + { + "epoch": 0.44731027641249943, + "flos": 25155952211040.0, + "grad_norm": 1.9781564309273207, + "learning_rate": 2.4349465488870896e-06, + "loss": 0.718, + "num_input_tokens_seen": 161419085, + "step": 4960 + }, + { + "epoch": 0.4474004599359697, + "flos": 26867443101120.0, + "grad_norm": 6.377897613823235, + "learning_rate": 2.434376302046905e-06, + "loss": 0.8393, + "num_input_tokens_seen": 161450875, + "step": 4961 + }, + { + "epoch": 0.44749064345943995, + "flos": 16813187526240.0, + "grad_norm": 1.8401403491769832, + "learning_rate": 2.433806018143339e-06, + "loss": 0.7794, + "num_input_tokens_seen": 161477940, + "step": 4962 + }, + { + "epoch": 0.44758082698291024, + "flos": 65037274241760.0, + "grad_norm": 0.6724514373631191, + "learning_rate": 2.433235697225051e-06, + "loss": 0.6153, + "num_input_tokens_seen": 161561410, + "step": 4963 + }, + { + "epoch": 0.4476710105063805, + "flos": 20165089258080.0, + "grad_norm": 2.014783106790214, + "learning_rate": 2.4326653393407048e-06, + "loss": 0.7666, + "num_input_tokens_seen": 161588845, + "step": 4964 + }, + { + "epoch": 0.44776119402985076, + "flos": 20966335234560.0, + "grad_norm": 2.16666462482454, + "learning_rate": 2.432094944538966e-06, + "loss": 0.7204, + "num_input_tokens_seen": 161618190, + "step": 4965 + }, + { + "epoch": 0.447851377553321, + "flos": 24865164479520.0, + "grad_norm": 1.4779308780359368, + "learning_rate": 2.4315245128685047e-06, + "loss": 0.7195, + "num_input_tokens_seen": 161647095, + "step": 4966 + }, + { + "epoch": 0.4479415610767913, + "flos": 23298361229760.0, + "grad_norm": 1.6355080872594947, + "learning_rate": 2.4309540443779925e-06, + "loss": 0.7517, + "num_input_tokens_seen": 161678720, + "step": 4967 + }, + { + "epoch": 0.4480317446002615, + "flos": 31197473195520.0, + "grad_norm": 1.5955836649410717, + "learning_rate": 2.4303835391161047e-06, + "loss": 0.6873, + "num_input_tokens_seen": 161709160, + "step": 4968 + }, + { + "epoch": 0.4481219281237318, + "flos": 19870584550560.0, + "grad_norm": 2.079596321291127, + "learning_rate": 2.42981299713152e-06, + "loss": 0.7011, + "num_input_tokens_seen": 161739915, + "step": 4969 + }, + { + "epoch": 0.44821211164720204, + "flos": 32110914708480.0, + "grad_norm": 6.635831220084286, + "learning_rate": 2.4292424184729204e-06, + "loss": 0.7888, + "num_input_tokens_seen": 161770490, + "step": 4970 + }, + { + "epoch": 0.4483022951706723, + "flos": 23298324060000.0, + "grad_norm": 1.7050360360909962, + "learning_rate": 2.4286718031889913e-06, + "loss": 0.7453, + "num_input_tokens_seen": 161800405, + "step": 4971 + }, + { + "epoch": 0.44839247869414256, + "flos": 27922784522400.0, + "grad_norm": 1.4332713250490592, + "learning_rate": 2.4281011513284202e-06, + "loss": 0.725, + "num_input_tokens_seen": 161830380, + "step": 4972 + }, + { + "epoch": 0.44848266221761285, + "flos": 26103261108960.0, + "grad_norm": 2.2531106427247143, + "learning_rate": 2.4275304629398985e-06, + "loss": 0.686, + "num_input_tokens_seen": 161858865, + "step": 4973 + }, + { + "epoch": 0.4485728457410831, + "flos": 25331496485760.0, + "grad_norm": 1.8206399778567801, + "learning_rate": 2.4269597380721194e-06, + "loss": 0.8269, + "num_input_tokens_seen": 161887675, + "step": 4974 + }, + { + "epoch": 0.44866302926455337, + "flos": 21510549355200.0, + "grad_norm": 1.5808511736152544, + "learning_rate": 2.426388976773782e-06, + "loss": 0.7453, + "num_input_tokens_seen": 161914920, + "step": 4975 + }, + { + "epoch": 0.4487532127880236, + "flos": 17578410271680.0, + "grad_norm": 2.541922311485576, + "learning_rate": 2.425818179093586e-06, + "loss": 0.8554, + "num_input_tokens_seen": 161941385, + "step": 4976 + }, + { + "epoch": 0.4488433963114939, + "flos": 25700612368800.0, + "grad_norm": 1.534844985739284, + "learning_rate": 2.4252473450802346e-06, + "loss": 0.6608, + "num_input_tokens_seen": 161971525, + "step": 4977 + }, + { + "epoch": 0.4489335798349641, + "flos": 23186017014240.0, + "grad_norm": 1.9311993194152215, + "learning_rate": 2.4246764747824355e-06, + "loss": 0.7818, + "num_input_tokens_seen": 162000565, + "step": 4978 + }, + { + "epoch": 0.4490237633584344, + "flos": 23516693631840.0, + "grad_norm": 1.712806203712621, + "learning_rate": 2.424105568248897e-06, + "loss": 0.7612, + "num_input_tokens_seen": 162028520, + "step": 4979 + }, + { + "epoch": 0.4491139468819047, + "flos": 28216322816160.0, + "grad_norm": 1.6005825859625908, + "learning_rate": 2.4235346255283337e-06, + "loss": 0.6906, + "num_input_tokens_seen": 162060300, + "step": 4980 + }, + { + "epoch": 0.44920413040537494, + "flos": 20347286919840.0, + "grad_norm": 2.4120845654404786, + "learning_rate": 2.42296364666946e-06, + "loss": 0.7603, + "num_input_tokens_seen": 162085955, + "step": 4981 + }, + { + "epoch": 0.4492943139288452, + "flos": 22965305747520.0, + "grad_norm": 1.7416675972213571, + "learning_rate": 2.4223926317209965e-06, + "loss": 0.6971, + "num_input_tokens_seen": 162115400, + "step": 4982 + }, + { + "epoch": 0.44938449745231546, + "flos": 31856893603200.0, + "grad_norm": 1.5727952906112173, + "learning_rate": 2.4218215807316647e-06, + "loss": 0.7521, + "num_input_tokens_seen": 162146975, + "step": 4983 + }, + { + "epoch": 0.44947468097578575, + "flos": 20274013855680.0, + "grad_norm": 1.8498889425811575, + "learning_rate": 2.4212504937501894e-06, + "loss": 0.8356, + "num_input_tokens_seen": 162173725, + "step": 4984 + }, + { + "epoch": 0.449564864499256, + "flos": 21658024727520.0, + "grad_norm": 1.4266838665546338, + "learning_rate": 2.4206793708253e-06, + "loss": 0.8081, + "num_input_tokens_seen": 162204080, + "step": 4985 + }, + { + "epoch": 0.44965504802272627, + "flos": 18525198792960.0, + "grad_norm": 2.0563382671855805, + "learning_rate": 2.420108212005726e-06, + "loss": 0.7056, + "num_input_tokens_seen": 162229605, + "step": 4986 + }, + { + "epoch": 0.4497452315461965, + "flos": 17615325576960.0, + "grad_norm": 1.9178005540987608, + "learning_rate": 2.4195370173402034e-06, + "loss": 0.7468, + "num_input_tokens_seen": 162254360, + "step": 4987 + }, + { + "epoch": 0.4498354150696668, + "flos": 20420485644480.0, + "grad_norm": 2.1316016256725994, + "learning_rate": 2.4189657868774696e-06, + "loss": 0.8713, + "num_input_tokens_seen": 162282510, + "step": 4988 + }, + { + "epoch": 0.449925598593137, + "flos": 22605147776640.0, + "grad_norm": 1.9636273420211712, + "learning_rate": 2.418394520666264e-06, + "loss": 0.6888, + "num_input_tokens_seen": 162312480, + "step": 4989 + }, + { + "epoch": 0.4500157821166073, + "flos": 24976913978880.0, + "grad_norm": 1.5816110311892404, + "learning_rate": 2.4178232187553307e-06, + "loss": 0.7362, + "num_input_tokens_seen": 162345815, + "step": 4990 + }, + { + "epoch": 0.45010596564007754, + "flos": 24281953547040.0, + "grad_norm": 1.5689044921913782, + "learning_rate": 2.417251881193417e-06, + "loss": 0.7805, + "num_input_tokens_seen": 162373780, + "step": 4991 + }, + { + "epoch": 0.45019614916354783, + "flos": 23949901648320.0, + "grad_norm": 1.6024104409030517, + "learning_rate": 2.4166805080292723e-06, + "loss": 0.8082, + "num_input_tokens_seen": 162399175, + "step": 4992 + }, + { + "epoch": 0.45028633268701806, + "flos": 25008105141120.0, + "grad_norm": 1.5779985433844936, + "learning_rate": 2.4161090993116485e-06, + "loss": 0.7022, + "num_input_tokens_seen": 162430125, + "step": 4993 + }, + { + "epoch": 0.45037651621048835, + "flos": 23620600311840.0, + "grad_norm": 2.1342034467839293, + "learning_rate": 2.4155376550893026e-06, + "loss": 0.6337, + "num_input_tokens_seen": 162456985, + "step": 4994 + }, + { + "epoch": 0.4504666997339586, + "flos": 19400089531200.0, + "grad_norm": 1.659832373049253, + "learning_rate": 2.4149661754109926e-06, + "loss": 0.763, + "num_input_tokens_seen": 162483160, + "step": 4995 + }, + { + "epoch": 0.4505568832574289, + "flos": 24245409939360.0, + "grad_norm": 1.8902012373748562, + "learning_rate": 2.41439466032548e-06, + "loss": 0.8096, + "num_input_tokens_seen": 162507650, + "step": 4996 + }, + { + "epoch": 0.4506470667808991, + "flos": 30616269430080.0, + "grad_norm": 14.891909436831957, + "learning_rate": 2.41382310988153e-06, + "loss": 0.7027, + "num_input_tokens_seen": 162536025, + "step": 4997 + }, + { + "epoch": 0.4507372503043694, + "flos": 20747222267520.0, + "grad_norm": 2.305860524772604, + "learning_rate": 2.413251524127911e-06, + "loss": 0.8079, + "num_input_tokens_seen": 162564125, + "step": 4998 + }, + { + "epoch": 0.45082743382783963, + "flos": 18816952938240.0, + "grad_norm": 1.6605252921622857, + "learning_rate": 2.412679903113393e-06, + "loss": 0.7225, + "num_input_tokens_seen": 162590325, + "step": 4999 + }, + { + "epoch": 0.4509176173513099, + "flos": 16916759678400.0, + "grad_norm": 2.3648087538318414, + "learning_rate": 2.4121082468867505e-06, + "loss": 0.7674, + "num_input_tokens_seen": 162610250, + "step": 5000 + }, + { + "epoch": 0.45100780087478015, + "flos": 37797481488480.0, + "grad_norm": 4.138424846627234, + "learning_rate": 2.4115365554967597e-06, + "loss": 0.6883, + "num_input_tokens_seen": 162637580, + "step": 5001 + }, + { + "epoch": 0.45109798439825044, + "flos": 21695237390880.0, + "grad_norm": 1.717945951177805, + "learning_rate": 2.4109648289922006e-06, + "loss": 0.752, + "num_input_tokens_seen": 162665165, + "step": 5002 + }, + { + "epoch": 0.45118816792172073, + "flos": 28249260957120.0, + "grad_norm": 1.9651849348096808, + "learning_rate": 2.4103930674218565e-06, + "loss": 0.7614, + "num_input_tokens_seen": 162696235, + "step": 5003 + }, + { + "epoch": 0.45127835144519096, + "flos": 21182511790560.0, + "grad_norm": 1.5645654378477947, + "learning_rate": 2.409821270834513e-06, + "loss": 0.7295, + "num_input_tokens_seen": 162724240, + "step": 5004 + }, + { + "epoch": 0.45136853496866125, + "flos": 25666410456000.0, + "grad_norm": 1.7950080792768566, + "learning_rate": 2.409249439278959e-06, + "loss": 0.8072, + "num_input_tokens_seen": 162754575, + "step": 5005 + }, + { + "epoch": 0.4514587184921315, + "flos": 62953836833280.0, + "grad_norm": 0.6262304829508627, + "learning_rate": 2.408677572803986e-06, + "loss": 0.5519, + "num_input_tokens_seen": 162850010, + "step": 5006 + }, + { + "epoch": 0.45154890201560177, + "flos": 23152186799040.0, + "grad_norm": 1.7645926369065619, + "learning_rate": 2.408105671458389e-06, + "loss": 0.7533, + "num_input_tokens_seen": 162877875, + "step": 5007 + }, + { + "epoch": 0.451639085539072, + "flos": 29964580332480.0, + "grad_norm": 1.3969286658711868, + "learning_rate": 2.4075337352909663e-06, + "loss": 0.6956, + "num_input_tokens_seen": 162911225, + "step": 5008 + }, + { + "epoch": 0.4517292690625423, + "flos": 23401375835520.0, + "grad_norm": 1.5354225762442015, + "learning_rate": 2.4069617643505177e-06, + "loss": 0.8801, + "num_input_tokens_seen": 162938770, + "step": 5009 + }, + { + "epoch": 0.4518194525860125, + "flos": 22678160652480.0, + "grad_norm": 1.5574033393072586, + "learning_rate": 2.406389758685848e-06, + "loss": 0.8369, + "num_input_tokens_seen": 162968280, + "step": 5010 + }, + { + "epoch": 0.4519096361094828, + "flos": 25079928584640.0, + "grad_norm": 1.6418796646406217, + "learning_rate": 2.405817718345763e-06, + "loss": 0.7894, + "num_input_tokens_seen": 162996635, + "step": 5011 + }, + { + "epoch": 0.45199981963295305, + "flos": 23553088560480.0, + "grad_norm": 2.3795978812301417, + "learning_rate": 2.4052456433790726e-06, + "loss": 0.7083, + "num_input_tokens_seen": 163028545, + "step": 5012 + }, + { + "epoch": 0.45209000315642334, + "flos": 31561794179520.0, + "grad_norm": 1.7661568844838618, + "learning_rate": 2.4046735338345897e-06, + "loss": 0.8105, + "num_input_tokens_seen": 163060040, + "step": 5013 + }, + { + "epoch": 0.45218018667989357, + "flos": 22859391900480.0, + "grad_norm": 3.122385228096347, + "learning_rate": 2.404101389761129e-06, + "loss": 0.7679, + "num_input_tokens_seen": 163087770, + "step": 5014 + }, + { + "epoch": 0.45227037020336386, + "flos": 62577026809920.0, + "grad_norm": 0.7094921889113729, + "learning_rate": 2.4035292112075097e-06, + "loss": 0.5631, + "num_input_tokens_seen": 163179625, + "step": 5015 + }, + { + "epoch": 0.4523605537268341, + "flos": 15861901464000.0, + "grad_norm": 1.5629846965506768, + "learning_rate": 2.4029569982225534e-06, + "loss": 0.811, + "num_input_tokens_seen": 163205775, + "step": 5016 + }, + { + "epoch": 0.4524507372503044, + "flos": 57732784324800.0, + "grad_norm": 0.6981271502291796, + "learning_rate": 2.402384750855084e-06, + "loss": 0.6351, + "num_input_tokens_seen": 163293765, + "step": 5017 + }, + { + "epoch": 0.4525409207737746, + "flos": 27233473894080.0, + "grad_norm": 2.5138010556476993, + "learning_rate": 2.4018124691539286e-06, + "loss": 0.7552, + "num_input_tokens_seen": 163323175, + "step": 5018 + }, + { + "epoch": 0.4526311042972449, + "flos": 25989207084480.0, + "grad_norm": 4.7413297125634815, + "learning_rate": 2.4012401531679178e-06, + "loss": 0.7409, + "num_input_tokens_seen": 163350860, + "step": 5019 + }, + { + "epoch": 0.45272128782071513, + "flos": 27086816256480.0, + "grad_norm": 2.035682888314716, + "learning_rate": 2.4006678029458847e-06, + "loss": 0.7495, + "num_input_tokens_seen": 163379380, + "step": 5020 + }, + { + "epoch": 0.4528114713441854, + "flos": 21731780998560.0, + "grad_norm": 1.626172816257722, + "learning_rate": 2.400095418536666e-06, + "loss": 0.6592, + "num_input_tokens_seen": 163406695, + "step": 5021 + }, + { + "epoch": 0.45290165486765566, + "flos": 22422615587040.0, + "grad_norm": 2.272687860493951, + "learning_rate": 2.3995229999890996e-06, + "loss": 0.73, + "num_input_tokens_seen": 163436285, + "step": 5022 + }, + { + "epoch": 0.45299183839112594, + "flos": 23116014888960.0, + "grad_norm": 2.4612708456339436, + "learning_rate": 2.398950547352028e-06, + "loss": 0.6952, + "num_input_tokens_seen": 163465160, + "step": 5023 + }, + { + "epoch": 0.4530820219145962, + "flos": 20784249082080.0, + "grad_norm": 2.2785627658780117, + "learning_rate": 2.398378060674295e-06, + "loss": 0.6794, + "num_input_tokens_seen": 163494110, + "step": 5024 + }, + { + "epoch": 0.45317220543806647, + "flos": 24318088287360.0, + "grad_norm": 1.5884660379719433, + "learning_rate": 2.39780554000475e-06, + "loss": 0.7538, + "num_input_tokens_seen": 163524515, + "step": 5025 + }, + { + "epoch": 0.4532623889615367, + "flos": 17869495361280.0, + "grad_norm": 1.9562511602526262, + "learning_rate": 2.3972329853922434e-06, + "loss": 0.782, + "num_input_tokens_seen": 163551310, + "step": 5026 + }, + { + "epoch": 0.453352572485007, + "flos": 67686255479520.0, + "grad_norm": 0.5923153711522896, + "learning_rate": 2.3966603968856278e-06, + "loss": 0.5704, + "num_input_tokens_seen": 163654700, + "step": 5027 + }, + { + "epoch": 0.4534427560084773, + "flos": 22713291809280.0, + "grad_norm": 1.7515926606901973, + "learning_rate": 2.39608777453376e-06, + "loss": 0.8243, + "num_input_tokens_seen": 163682240, + "step": 5028 + }, + { + "epoch": 0.4535329395319475, + "flos": 21330619048800.0, + "grad_norm": 1.7907806541868583, + "learning_rate": 2.3955151183854993e-06, + "loss": 0.7783, + "num_input_tokens_seen": 163708690, + "step": 5029 + }, + { + "epoch": 0.4536231230554178, + "flos": 25079928584640.0, + "grad_norm": 2.2322148905439008, + "learning_rate": 2.3949424284897073e-06, + "loss": 0.7157, + "num_input_tokens_seen": 163737270, + "step": 5030 + }, + { + "epoch": 0.45371330657888803, + "flos": 33787088592960.0, + "grad_norm": 1.9173765181031768, + "learning_rate": 2.39436970489525e-06, + "loss": 0.6223, + "num_input_tokens_seen": 163769065, + "step": 5031 + }, + { + "epoch": 0.4538034901023583, + "flos": 22532803956480.0, + "grad_norm": 1.7324220814323088, + "learning_rate": 2.3937969476509955e-06, + "loss": 0.7527, + "num_input_tokens_seen": 163797435, + "step": 5032 + }, + { + "epoch": 0.45389367362582855, + "flos": 20492569276320.0, + "grad_norm": 1.6103085523127962, + "learning_rate": 2.393224156805813e-06, + "loss": 0.826, + "num_input_tokens_seen": 163824415, + "step": 5033 + }, + { + "epoch": 0.45398385714929884, + "flos": 20747222267520.0, + "grad_norm": 1.9363163439600253, + "learning_rate": 2.392651332408578e-06, + "loss": 0.7288, + "num_input_tokens_seen": 163850850, + "step": 5034 + }, + { + "epoch": 0.4540740406727691, + "flos": 20273679327840.0, + "grad_norm": 2.5080706977431735, + "learning_rate": 2.3920784745081655e-06, + "loss": 0.7352, + "num_input_tokens_seen": 163877440, + "step": 5035 + }, + { + "epoch": 0.45416422419623936, + "flos": 13570544919840.0, + "grad_norm": 1.6435599097510616, + "learning_rate": 2.391505583153456e-06, + "loss": 0.7904, + "num_input_tokens_seen": 163904330, + "step": 5036 + }, + { + "epoch": 0.4542544077197096, + "flos": 22314360045120.0, + "grad_norm": 1.6655332128487912, + "learning_rate": 2.3909326583933315e-06, + "loss": 0.7558, + "num_input_tokens_seen": 163931095, + "step": 5037 + }, + { + "epoch": 0.4543445912431799, + "flos": 37791125459520.0, + "grad_norm": 1.8058676555888455, + "learning_rate": 2.3903597002766777e-06, + "loss": 0.7177, + "num_input_tokens_seen": 163965055, + "step": 5038 + }, + { + "epoch": 0.4544347747666501, + "flos": 25483989775680.0, + "grad_norm": 1.6155685208333976, + "learning_rate": 2.389786708852381e-06, + "loss": 0.7128, + "num_input_tokens_seen": 163993870, + "step": 5039 + }, + { + "epoch": 0.4545249582901204, + "flos": 67126169871360.0, + "grad_norm": 0.9229860238105104, + "learning_rate": 2.389213684169333e-06, + "loss": 0.6944, + "num_input_tokens_seen": 164090375, + "step": 5040 + }, + { + "epoch": 0.45461514181359064, + "flos": 22788274682400.0, + "grad_norm": 1.6310739707240849, + "learning_rate": 2.388640626276428e-06, + "loss": 0.7175, + "num_input_tokens_seen": 164121140, + "step": 5041 + }, + { + "epoch": 0.45470532533706093, + "flos": 63961372209600.0, + "grad_norm": 0.6799492825148993, + "learning_rate": 2.388067535222561e-06, + "loss": 0.5352, + "num_input_tokens_seen": 164218035, + "step": 5042 + }, + { + "epoch": 0.45479550886053116, + "flos": 28250710577760.0, + "grad_norm": 1.8462091630204325, + "learning_rate": 2.3874944110566332e-06, + "loss": 0.7671, + "num_input_tokens_seen": 164248495, + "step": 5043 + }, + { + "epoch": 0.45488569238400145, + "flos": 70356249098880.0, + "grad_norm": 0.7592672169491473, + "learning_rate": 2.3869212538275447e-06, + "loss": 0.6775, + "num_input_tokens_seen": 164334055, + "step": 5044 + }, + { + "epoch": 0.4549758759074717, + "flos": 20420597153760.0, + "grad_norm": 1.991183038939858, + "learning_rate": 2.386348063584202e-06, + "loss": 0.7689, + "num_input_tokens_seen": 164364695, + "step": 5045 + }, + { + "epoch": 0.45506605943094197, + "flos": 19618459103040.0, + "grad_norm": 1.776082289185491, + "learning_rate": 2.385774840375511e-06, + "loss": 0.8017, + "num_input_tokens_seen": 164391490, + "step": 5046 + }, + { + "epoch": 0.4551562429544122, + "flos": 19071717438720.0, + "grad_norm": 1.9378044750562358, + "learning_rate": 2.385201584250385e-06, + "loss": 0.7042, + "num_input_tokens_seen": 164419590, + "step": 5047 + }, + { + "epoch": 0.4552464264778825, + "flos": 62073667989120.0, + "grad_norm": 0.6362285013819832, + "learning_rate": 2.3846282952577346e-06, + "loss": 0.6127, + "num_input_tokens_seen": 164514555, + "step": 5048 + }, + { + "epoch": 0.4553366100013527, + "flos": 23731271888160.0, + "grad_norm": 1.6433583333252655, + "learning_rate": 2.3840549734464785e-06, + "loss": 0.8435, + "num_input_tokens_seen": 164543935, + "step": 5049 + }, + { + "epoch": 0.455426793524823, + "flos": 17979386372640.0, + "grad_norm": 2.4564060695462016, + "learning_rate": 2.3834816188655336e-06, + "loss": 0.6865, + "num_input_tokens_seen": 164572615, + "step": 5050 + }, + { + "epoch": 0.4555169770482933, + "flos": 36595891296960.0, + "grad_norm": 1.6049330065759517, + "learning_rate": 2.3829082315638224e-06, + "loss": 0.7002, + "num_input_tokens_seen": 164604080, + "step": 5051 + }, + { + "epoch": 0.45560716057176354, + "flos": 26467284734880.0, + "grad_norm": 2.025445479383144, + "learning_rate": 2.3823348115902695e-06, + "loss": 0.6123, + "num_input_tokens_seen": 164631455, + "step": 5052 + }, + { + "epoch": 0.4556973440952338, + "flos": 21549694846080.0, + "grad_norm": 1.6897062078999128, + "learning_rate": 2.3817613589938026e-06, + "loss": 0.7742, + "num_input_tokens_seen": 164657725, + "step": 5053 + }, + { + "epoch": 0.45578752761870406, + "flos": 22240640943840.0, + "grad_norm": 1.6202632385240452, + "learning_rate": 2.3811878738233517e-06, + "loss": 0.7815, + "num_input_tokens_seen": 164687615, + "step": 5054 + }, + { + "epoch": 0.45587771114217435, + "flos": 35029013707680.0, + "grad_norm": 2.0368263582014925, + "learning_rate": 2.380614356127849e-06, + "loss": 0.6979, + "num_input_tokens_seen": 164717365, + "step": 5055 + }, + { + "epoch": 0.4559678946656446, + "flos": 22420719929280.0, + "grad_norm": 1.8411804656387056, + "learning_rate": 2.3800408059562318e-06, + "loss": 0.7486, + "num_input_tokens_seen": 164746075, + "step": 5056 + }, + { + "epoch": 0.45605807818911487, + "flos": 22532469428640.0, + "grad_norm": 1.815026528345865, + "learning_rate": 2.3794672233574365e-06, + "loss": 0.7703, + "num_input_tokens_seen": 164777440, + "step": 5057 + }, + { + "epoch": 0.4561482617125851, + "flos": 23917446714240.0, + "grad_norm": 2.0733139449722766, + "learning_rate": 2.3788936083804058e-06, + "loss": 0.7674, + "num_input_tokens_seen": 164805780, + "step": 5058 + }, + { + "epoch": 0.4562384452360554, + "flos": 23331894086880.0, + "grad_norm": 1.5545921104726628, + "learning_rate": 2.378319961074083e-06, + "loss": 0.7546, + "num_input_tokens_seen": 164835290, + "step": 5059 + }, + { + "epoch": 0.4563286287595256, + "flos": 24135556097760.0, + "grad_norm": 1.7437620018986957, + "learning_rate": 2.377746281487415e-06, + "loss": 0.79, + "num_input_tokens_seen": 164864500, + "step": 5060 + }, + { + "epoch": 0.4564188122829959, + "flos": 21367497184320.0, + "grad_norm": 2.1503009275045444, + "learning_rate": 2.377172569669352e-06, + "loss": 0.8441, + "num_input_tokens_seen": 164892480, + "step": 5061 + }, + { + "epoch": 0.45650899580646614, + "flos": 23480298703200.0, + "grad_norm": 1.7333573113519147, + "learning_rate": 2.376598825668845e-06, + "loss": 0.6025, + "num_input_tokens_seen": 164922385, + "step": 5062 + }, + { + "epoch": 0.45659917932993643, + "flos": 20889493873440.0, + "grad_norm": 1.7904024395263864, + "learning_rate": 2.3760250495348495e-06, + "loss": 0.7415, + "num_input_tokens_seen": 164949985, + "step": 5063 + }, + { + "epoch": 0.45668936285340667, + "flos": 24097042492800.0, + "grad_norm": 1.7851488376188083, + "learning_rate": 2.3754512413163236e-06, + "loss": 0.7986, + "num_input_tokens_seen": 164980225, + "step": 5064 + }, + { + "epoch": 0.45677954637687695, + "flos": 28871580210720.0, + "grad_norm": 1.4704649632342028, + "learning_rate": 2.3748774010622285e-06, + "loss": 0.7319, + "num_input_tokens_seen": 165010770, + "step": 5065 + }, + { + "epoch": 0.4568697299003472, + "flos": 15465014036640.0, + "grad_norm": 1.8873072746864705, + "learning_rate": 2.3743035288215254e-06, + "loss": 0.7575, + "num_input_tokens_seen": 165037375, + "step": 5066 + }, + { + "epoch": 0.4569599134238175, + "flos": 22168334293440.0, + "grad_norm": 1.6231607151618945, + "learning_rate": 2.3737296246431815e-06, + "loss": 0.7622, + "num_input_tokens_seen": 165063380, + "step": 5067 + }, + { + "epoch": 0.4570500969472877, + "flos": 26249695728000.0, + "grad_norm": 2.2878000990165863, + "learning_rate": 2.3731556885761656e-06, + "loss": 0.7239, + "num_input_tokens_seen": 165092215, + "step": 5068 + }, + { + "epoch": 0.457140280470758, + "flos": 70419895195200.0, + "grad_norm": 0.6390624645578239, + "learning_rate": 2.372581720669449e-06, + "loss": 0.5792, + "num_input_tokens_seen": 165167985, + "step": 5069 + }, + { + "epoch": 0.45723046399422823, + "flos": 17392198275840.0, + "grad_norm": 2.125280586118611, + "learning_rate": 2.3720077209720046e-06, + "loss": 0.7815, + "num_input_tokens_seen": 165194045, + "step": 5070 + }, + { + "epoch": 0.4573206475176985, + "flos": 70690828541280.0, + "grad_norm": 0.6940605593967231, + "learning_rate": 2.3714336895328112e-06, + "loss": 0.5724, + "num_input_tokens_seen": 165285845, + "step": 5071 + }, + { + "epoch": 0.45741083104116875, + "flos": 22205175259200.0, + "grad_norm": 1.7769410735682025, + "learning_rate": 2.370859626400847e-06, + "loss": 0.7089, + "num_input_tokens_seen": 165313160, + "step": 5072 + }, + { + "epoch": 0.45750101456463904, + "flos": 19434142764960.0, + "grad_norm": 1.5177018398686222, + "learning_rate": 2.3702855316250943e-06, + "loss": 0.7296, + "num_input_tokens_seen": 165341520, + "step": 5073 + }, + { + "epoch": 0.45759119808810933, + "flos": 25337889684480.0, + "grad_norm": 1.6160530866080334, + "learning_rate": 2.369711405254539e-06, + "loss": 0.7537, + "num_input_tokens_seen": 165369130, + "step": 5074 + }, + { + "epoch": 0.45768138161157956, + "flos": 22277927946720.0, + "grad_norm": 1.7402593779988698, + "learning_rate": 2.3691372473381673e-06, + "loss": 0.8062, + "num_input_tokens_seen": 165397545, + "step": 5075 + }, + { + "epoch": 0.45777156513504985, + "flos": 24172954609920.0, + "grad_norm": 1.6744516072791202, + "learning_rate": 2.3685630579249708e-06, + "loss": 0.8122, + "num_input_tokens_seen": 165427700, + "step": 5076 + }, + { + "epoch": 0.4578617486585201, + "flos": 27520433140320.0, + "grad_norm": 1.602431196619275, + "learning_rate": 2.367988837063942e-06, + "loss": 0.7288, + "num_input_tokens_seen": 165455505, + "step": 5077 + }, + { + "epoch": 0.4579519321819904, + "flos": 23042853334080.0, + "grad_norm": 1.6853518765526927, + "learning_rate": 2.367414584804076e-06, + "loss": 0.8298, + "num_input_tokens_seen": 165482765, + "step": 5078 + }, + { + "epoch": 0.4580421157054606, + "flos": 22897124940480.0, + "grad_norm": 2.244096959737562, + "learning_rate": 2.366840301194372e-06, + "loss": 0.7871, + "num_input_tokens_seen": 165509870, + "step": 5079 + }, + { + "epoch": 0.4581322992289309, + "flos": 22058629130880.0, + "grad_norm": 1.600411499274639, + "learning_rate": 2.3662659862838308e-06, + "loss": 0.7408, + "num_input_tokens_seen": 165538895, + "step": 5080 + }, + { + "epoch": 0.45822248275240113, + "flos": 19508345073120.0, + "grad_norm": 2.2904219512665676, + "learning_rate": 2.365691640121456e-06, + "loss": 0.8112, + "num_input_tokens_seen": 165566835, + "step": 5081 + }, + { + "epoch": 0.4583126662758714, + "flos": 34118174077920.0, + "grad_norm": 1.8909540441769945, + "learning_rate": 2.365117262756254e-06, + "loss": 0.6857, + "num_input_tokens_seen": 165601860, + "step": 5082 + }, + { + "epoch": 0.45840284979934165, + "flos": 20457735477600.0, + "grad_norm": 1.7418833924780162, + "learning_rate": 2.3645428542372342e-06, + "loss": 0.7479, + "num_input_tokens_seen": 165630260, + "step": 5083 + }, + { + "epoch": 0.45849303332281194, + "flos": 70922393377920.0, + "grad_norm": 0.6169102202837902, + "learning_rate": 2.3639684146134083e-06, + "loss": 0.5315, + "num_input_tokens_seen": 165725155, + "step": 5084 + }, + { + "epoch": 0.45858321684628217, + "flos": 28906488348960.0, + "grad_norm": 2.2945123055403034, + "learning_rate": 2.3633939439337897e-06, + "loss": 0.7156, + "num_input_tokens_seen": 165753525, + "step": 5085 + }, + { + "epoch": 0.45867340036975246, + "flos": 18051693023040.0, + "grad_norm": 6.06898639191054, + "learning_rate": 2.362819442247396e-06, + "loss": 0.7266, + "num_input_tokens_seen": 165781325, + "step": 5086 + }, + { + "epoch": 0.4587635838932227, + "flos": 68469325443360.0, + "grad_norm": 0.636441995584808, + "learning_rate": 2.3622449096032477e-06, + "loss": 0.542, + "num_input_tokens_seen": 165878115, + "step": 5087 + }, + { + "epoch": 0.458853767416693, + "flos": 18014368850400.0, + "grad_norm": 3.884072051506504, + "learning_rate": 2.361670346050366e-06, + "loss": 0.7891, + "num_input_tokens_seen": 165906100, + "step": 5088 + }, + { + "epoch": 0.4589439509401632, + "flos": 16922372312160.0, + "grad_norm": 2.0139719964195284, + "learning_rate": 2.3610957516377757e-06, + "loss": 0.6906, + "num_input_tokens_seen": 165933405, + "step": 5089 + }, + { + "epoch": 0.4590341344636335, + "flos": 28796522998080.0, + "grad_norm": 2.449309084259146, + "learning_rate": 2.3605211264145048e-06, + "loss": 0.7552, + "num_input_tokens_seen": 165962635, + "step": 5090 + }, + { + "epoch": 0.45912431798710374, + "flos": 23078876565120.0, + "grad_norm": 1.6342550230026374, + "learning_rate": 2.3599464704295836e-06, + "loss": 0.7191, + "num_input_tokens_seen": 165993440, + "step": 5091 + }, + { + "epoch": 0.459214501510574, + "flos": 24099793055040.0, + "grad_norm": 1.8651585748259654, + "learning_rate": 2.359371783732045e-06, + "loss": 0.6454, + "num_input_tokens_seen": 166022635, + "step": 5092 + }, + { + "epoch": 0.45930468503404426, + "flos": 22091121234720.0, + "grad_norm": 1.5939053478449696, + "learning_rate": 2.358797066370924e-06, + "loss": 0.7331, + "num_input_tokens_seen": 166051565, + "step": 5093 + }, + { + "epoch": 0.45939486855751455, + "flos": 23662013158080.0, + "grad_norm": 1.744920033150122, + "learning_rate": 2.3582223183952594e-06, + "loss": 0.7529, + "num_input_tokens_seen": 166081445, + "step": 5094 + }, + { + "epoch": 0.4594850520809848, + "flos": 21585903925920.0, + "grad_norm": 1.701616371500003, + "learning_rate": 2.357647539854091e-06, + "loss": 0.7864, + "num_input_tokens_seen": 166109410, + "step": 5095 + }, + { + "epoch": 0.45957523560445507, + "flos": 23986742614080.0, + "grad_norm": 2.0809075792678997, + "learning_rate": 2.3570727307964624e-06, + "loss": 0.7558, + "num_input_tokens_seen": 166136905, + "step": 5096 + }, + { + "epoch": 0.4596654191279253, + "flos": 28137437118240.0, + "grad_norm": 1.8044184364452636, + "learning_rate": 2.35649789127142e-06, + "loss": 0.8002, + "num_input_tokens_seen": 166163085, + "step": 5097 + }, + { + "epoch": 0.4597556026513956, + "flos": 16957986675840.0, + "grad_norm": 1.9142762356749161, + "learning_rate": 2.3559230213280115e-06, + "loss": 0.716, + "num_input_tokens_seen": 166190460, + "step": 5098 + }, + { + "epoch": 0.4598457861748659, + "flos": 23516359104000.0, + "grad_norm": 1.7155078479672945, + "learning_rate": 2.3553481210152886e-06, + "loss": 0.6652, + "num_input_tokens_seen": 166220880, + "step": 5099 + }, + { + "epoch": 0.4599359696983361, + "flos": 55843667653440.0, + "grad_norm": 0.8452517249634999, + "learning_rate": 2.3547731903823043e-06, + "loss": 0.5969, + "num_input_tokens_seen": 166301880, + "step": 5100 + }, + { + "epoch": 0.4600261532218064, + "flos": 25265025487680.0, + "grad_norm": 1.637176740931025, + "learning_rate": 2.3541982294781155e-06, + "loss": 0.7284, + "num_input_tokens_seen": 166328330, + "step": 5101 + }, + { + "epoch": 0.46011633674527663, + "flos": 26866476687360.0, + "grad_norm": 3.247798027007061, + "learning_rate": 2.3536232383517804e-06, + "loss": 0.7515, + "num_input_tokens_seen": 166358480, + "step": 5102 + }, + { + "epoch": 0.4602065202687469, + "flos": 32224634205120.0, + "grad_norm": 2.2868700170970926, + "learning_rate": 2.3530482170523602e-06, + "loss": 0.7463, + "num_input_tokens_seen": 166388160, + "step": 5103 + }, + { + "epoch": 0.46029670379221715, + "flos": 22782513369600.0, + "grad_norm": 1.500977515553536, + "learning_rate": 2.3524731656289206e-06, + "loss": 0.8461, + "num_input_tokens_seen": 166416475, + "step": 5104 + }, + { + "epoch": 0.46038688731568744, + "flos": 25517299614240.0, + "grad_norm": 2.0229565847945485, + "learning_rate": 2.351898084130526e-06, + "loss": 0.8379, + "num_input_tokens_seen": 166446210, + "step": 5105 + }, + { + "epoch": 0.4604770708391577, + "flos": 23619448049280.0, + "grad_norm": 1.7293308900116362, + "learning_rate": 2.351322972606247e-06, + "loss": 0.8183, + "num_input_tokens_seen": 166473355, + "step": 5106 + }, + { + "epoch": 0.46056725436262796, + "flos": 24026222632800.0, + "grad_norm": 1.8752342713549124, + "learning_rate": 2.350747831105155e-06, + "loss": 0.903, + "num_input_tokens_seen": 166499005, + "step": 5107 + }, + { + "epoch": 0.4606574378860982, + "flos": 24201692568000.0, + "grad_norm": 1.897709962187567, + "learning_rate": 2.350172659676323e-06, + "loss": 0.8316, + "num_input_tokens_seen": 166525520, + "step": 5108 + }, + { + "epoch": 0.4607476214095685, + "flos": 21585569398080.0, + "grad_norm": 1.8874469875950013, + "learning_rate": 2.3495974583688306e-06, + "loss": 0.7889, + "num_input_tokens_seen": 166552610, + "step": 5109 + }, + { + "epoch": 0.4608378049330387, + "flos": 23694468092160.0, + "grad_norm": 2.123928585346182, + "learning_rate": 2.3490222272317543e-06, + "loss": 0.8239, + "num_input_tokens_seen": 166580530, + "step": 5110 + }, + { + "epoch": 0.460927988456509, + "flos": 17541346287360.0, + "grad_norm": 1.8482881719485578, + "learning_rate": 2.348446966314177e-06, + "loss": 0.7998, + "num_input_tokens_seen": 166604380, + "step": 5111 + }, + { + "epoch": 0.46101817197997924, + "flos": 23224976656320.0, + "grad_norm": 1.9967823968229088, + "learning_rate": 2.3478716756651837e-06, + "loss": 0.7262, + "num_input_tokens_seen": 166632415, + "step": 5112 + }, + { + "epoch": 0.46110835550344953, + "flos": 21222028979040.0, + "grad_norm": 1.8358532176049849, + "learning_rate": 2.347296355333861e-06, + "loss": 0.7466, + "num_input_tokens_seen": 166659490, + "step": 5113 + }, + { + "epoch": 0.46119853902691976, + "flos": 24536346349920.0, + "grad_norm": 1.8055545986205606, + "learning_rate": 2.3467210053692972e-06, + "loss": 0.7608, + "num_input_tokens_seen": 166689070, + "step": 5114 + }, + { + "epoch": 0.46128872255039005, + "flos": 21039459619680.0, + "grad_norm": 1.7739176499306137, + "learning_rate": 2.3461456258205866e-06, + "loss": 0.757, + "num_input_tokens_seen": 166716705, + "step": 5115 + }, + { + "epoch": 0.4613789060738603, + "flos": 37211928861120.0, + "grad_norm": 1.971825150768043, + "learning_rate": 2.345570216736822e-06, + "loss": 0.6935, + "num_input_tokens_seen": 166749330, + "step": 5116 + }, + { + "epoch": 0.4614690895973306, + "flos": 24755905354080.0, + "grad_norm": 1.8889158736277332, + "learning_rate": 2.3449947781671013e-06, + "loss": 0.7621, + "num_input_tokens_seen": 166779465, + "step": 5117 + }, + { + "epoch": 0.4615592731208008, + "flos": 20853693660960.0, + "grad_norm": 1.776404163883812, + "learning_rate": 2.3444193101605237e-06, + "loss": 0.7827, + "num_input_tokens_seen": 166807130, + "step": 5118 + }, + { + "epoch": 0.4616494566442711, + "flos": 26063000525280.0, + "grad_norm": 1.5930444988446093, + "learning_rate": 2.3438438127661913e-06, + "loss": 0.7719, + "num_input_tokens_seen": 166833565, + "step": 5119 + }, + { + "epoch": 0.4617396401677413, + "flos": 35390546959680.0, + "grad_norm": 1.929608115598114, + "learning_rate": 2.3432682860332096e-06, + "loss": 0.5846, + "num_input_tokens_seen": 166864605, + "step": 5120 + }, + { + "epoch": 0.4618298236912116, + "flos": 28939277810880.0, + "grad_norm": 1.8287992619667333, + "learning_rate": 2.342692730010684e-06, + "loss": 0.7047, + "num_input_tokens_seen": 166894605, + "step": 5121 + }, + { + "epoch": 0.4619200072146819, + "flos": 19799244313920.0, + "grad_norm": 2.2299530448588802, + "learning_rate": 2.342117144747726e-06, + "loss": 0.7639, + "num_input_tokens_seen": 166917000, + "step": 5122 + }, + { + "epoch": 0.46201019073815214, + "flos": 22747753910400.0, + "grad_norm": 1.8601284904603603, + "learning_rate": 2.3415415302934457e-06, + "loss": 0.7547, + "num_input_tokens_seen": 166943800, + "step": 5123 + }, + { + "epoch": 0.4621003742616224, + "flos": 25264988317920.0, + "grad_norm": 1.6193514733793704, + "learning_rate": 2.340965886696959e-06, + "loss": 0.7838, + "num_input_tokens_seen": 166970000, + "step": 5124 + }, + { + "epoch": 0.46219055778509266, + "flos": 19142760317280.0, + "grad_norm": 1.6233205920247475, + "learning_rate": 2.340390214007384e-06, + "loss": 0.7021, + "num_input_tokens_seen": 166998665, + "step": 5125 + }, + { + "epoch": 0.46228074130856295, + "flos": 21109350235680.0, + "grad_norm": 1.6684003713720326, + "learning_rate": 2.339814512273838e-06, + "loss": 0.6352, + "num_input_tokens_seen": 167025555, + "step": 5126 + }, + { + "epoch": 0.4623709248320332, + "flos": 24500620476960.0, + "grad_norm": 1.6421429832224725, + "learning_rate": 2.3392387815454447e-06, + "loss": 0.8585, + "num_input_tokens_seen": 167053175, + "step": 5127 + }, + { + "epoch": 0.46246110835550347, + "flos": 39288335451360.0, + "grad_norm": 2.168865726252592, + "learning_rate": 2.3386630218713273e-06, + "loss": 0.7486, + "num_input_tokens_seen": 167084905, + "step": 5128 + }, + { + "epoch": 0.4625512918789737, + "flos": 21840296728800.0, + "grad_norm": 1.716182108270436, + "learning_rate": 2.3380872333006135e-06, + "loss": 0.7769, + "num_input_tokens_seen": 167112505, + "step": 5129 + }, + { + "epoch": 0.462641475402444, + "flos": 23147131711680.0, + "grad_norm": 1.8410966175053454, + "learning_rate": 2.3375114158824335e-06, + "loss": 0.7985, + "num_input_tokens_seen": 167141720, + "step": 5130 + }, + { + "epoch": 0.4627316589259142, + "flos": 27670027188960.0, + "grad_norm": 1.5373617156574724, + "learning_rate": 2.3369355696659184e-06, + "loss": 0.7898, + "num_input_tokens_seen": 167172960, + "step": 5131 + }, + { + "epoch": 0.4628218424493845, + "flos": 21002916012000.0, + "grad_norm": 1.8110500323768912, + "learning_rate": 2.336359694700202e-06, + "loss": 0.7412, + "num_input_tokens_seen": 167201030, + "step": 5132 + }, + { + "epoch": 0.46291202597285475, + "flos": 18774313489920.0, + "grad_norm": 1.7973308482582127, + "learning_rate": 2.335783791034422e-06, + "loss": 0.7574, + "num_input_tokens_seen": 167226490, + "step": 5133 + }, + { + "epoch": 0.46300220949632503, + "flos": 20820644010720.0, + "grad_norm": 1.583550597225355, + "learning_rate": 2.3352078587177173e-06, + "loss": 0.7358, + "num_input_tokens_seen": 167255310, + "step": 5134 + }, + { + "epoch": 0.46309239301979527, + "flos": 13970963474400.0, + "grad_norm": 1.8704116629283403, + "learning_rate": 2.33463189779923e-06, + "loss": 0.7623, + "num_input_tokens_seen": 167279810, + "step": 5135 + }, + { + "epoch": 0.46318257654326556, + "flos": 21731260621920.0, + "grad_norm": 1.5107084145990681, + "learning_rate": 2.334055908328104e-06, + "loss": 0.8291, + "num_input_tokens_seen": 167310495, + "step": 5136 + }, + { + "epoch": 0.4632727600667358, + "flos": 22568901527040.0, + "grad_norm": 1.8967961252521275, + "learning_rate": 2.3334798903534866e-06, + "loss": 0.6629, + "num_input_tokens_seen": 167340375, + "step": 5137 + }, + { + "epoch": 0.4633629435902061, + "flos": 19545557736480.0, + "grad_norm": 1.4081776017456629, + "learning_rate": 2.3329038439245252e-06, + "loss": 0.6503, + "num_input_tokens_seen": 167368580, + "step": 5138 + }, + { + "epoch": 0.4634531271136763, + "flos": 21439506476640.0, + "grad_norm": 2.315496925590845, + "learning_rate": 2.3323277690903724e-06, + "loss": 0.7544, + "num_input_tokens_seen": 167394540, + "step": 5139 + }, + { + "epoch": 0.4635433106371466, + "flos": 33533587864320.0, + "grad_norm": 1.6921011045878975, + "learning_rate": 2.3317516659001827e-06, + "loss": 0.7304, + "num_input_tokens_seen": 167424855, + "step": 5140 + }, + { + "epoch": 0.46363349416061683, + "flos": 21072249081600.0, + "grad_norm": 1.575651742419978, + "learning_rate": 2.331175534403111e-06, + "loss": 0.7673, + "num_input_tokens_seen": 167452370, + "step": 5141 + }, + { + "epoch": 0.4637236776840871, + "flos": 28831914343200.0, + "grad_norm": 1.5070094999382229, + "learning_rate": 2.3305993746483167e-06, + "loss": 0.8383, + "num_input_tokens_seen": 167478925, + "step": 5142 + }, + { + "epoch": 0.46381386120755735, + "flos": 15683346438720.0, + "grad_norm": 2.3007561726004946, + "learning_rate": 2.3300231866849606e-06, + "loss": 0.7327, + "num_input_tokens_seen": 167503855, + "step": 5143 + }, + { + "epoch": 0.46390404473102764, + "flos": 34773542981760.0, + "grad_norm": 1.9318568587893779, + "learning_rate": 2.3294469705622067e-06, + "loss": 0.7945, + "num_input_tokens_seen": 167535730, + "step": 5144 + }, + { + "epoch": 0.4639942282544979, + "flos": 23953395605760.0, + "grad_norm": 1.8246172647340326, + "learning_rate": 2.3288707263292203e-06, + "loss": 0.7307, + "num_input_tokens_seen": 167563520, + "step": 5145 + }, + { + "epoch": 0.46408441177796816, + "flos": 32405047718400.0, + "grad_norm": 1.4694774530633825, + "learning_rate": 2.3282944540351707e-06, + "loss": 0.7712, + "num_input_tokens_seen": 167595480, + "step": 5146 + }, + { + "epoch": 0.46417459530143845, + "flos": 21913235265120.0, + "grad_norm": 2.106456744181138, + "learning_rate": 2.327718153729228e-06, + "loss": 0.7294, + "num_input_tokens_seen": 167624490, + "step": 5147 + }, + { + "epoch": 0.4642647788249087, + "flos": 24172768761120.0, + "grad_norm": 1.7558263790437674, + "learning_rate": 2.327141825460566e-06, + "loss": 0.7942, + "num_input_tokens_seen": 167652680, + "step": 5148 + }, + { + "epoch": 0.464354962348379, + "flos": 57606018242400.0, + "grad_norm": 0.6705694821958026, + "learning_rate": 2.326565469278358e-06, + "loss": 0.5844, + "num_input_tokens_seen": 167736310, + "step": 5149 + }, + { + "epoch": 0.4644451458718492, + "flos": 55101235704480.0, + "grad_norm": 0.7127051543300827, + "learning_rate": 2.3259890852317846e-06, + "loss": 0.5938, + "num_input_tokens_seen": 167816400, + "step": 5150 + }, + { + "epoch": 0.4645353293953195, + "flos": 21542149384800.0, + "grad_norm": 1.9797396850933817, + "learning_rate": 2.3254126733700246e-06, + "loss": 0.7534, + "num_input_tokens_seen": 167843545, + "step": 5151 + }, + { + "epoch": 0.46462551291878973, + "flos": 20451825485760.0, + "grad_norm": 1.6837794650642481, + "learning_rate": 2.324836233742262e-06, + "loss": 0.7852, + "num_input_tokens_seen": 167869230, + "step": 5152 + }, + { + "epoch": 0.46471569644226, + "flos": 20346803712960.0, + "grad_norm": 1.9424247909773793, + "learning_rate": 2.3242597663976793e-06, + "loss": 0.7922, + "num_input_tokens_seen": 167896315, + "step": 5153 + }, + { + "epoch": 0.46480587996573025, + "flos": 19363806111840.0, + "grad_norm": 1.8166898772597209, + "learning_rate": 2.3236832713854663e-06, + "loss": 0.7355, + "num_input_tokens_seen": 167924545, + "step": 5154 + }, + { + "epoch": 0.46489606348920054, + "flos": 21075557190240.0, + "grad_norm": 1.7936425366260513, + "learning_rate": 2.323106748754812e-06, + "loss": 0.7512, + "num_input_tokens_seen": 167952550, + "step": 5155 + }, + { + "epoch": 0.4649862470126708, + "flos": 18307275258240.0, + "grad_norm": 15.717114221235379, + "learning_rate": 2.3225301985549077e-06, + "loss": 0.8465, + "num_input_tokens_seen": 167981015, + "step": 5156 + }, + { + "epoch": 0.46507643053614106, + "flos": 22095321417600.0, + "grad_norm": 1.925625808354115, + "learning_rate": 2.321953620834948e-06, + "loss": 0.686, + "num_input_tokens_seen": 168009200, + "step": 5157 + }, + { + "epoch": 0.4651666140596113, + "flos": 23220887982720.0, + "grad_norm": 1.5024851240034167, + "learning_rate": 2.3213770156441314e-06, + "loss": 0.721, + "num_input_tokens_seen": 168037305, + "step": 5158 + }, + { + "epoch": 0.4652567975830816, + "flos": 26102963750880.0, + "grad_norm": 2.313140883433724, + "learning_rate": 2.3208003830316554e-06, + "loss": 0.8037, + "num_input_tokens_seen": 168068730, + "step": 5159 + }, + { + "epoch": 0.4653469811065518, + "flos": 18925877535840.0, + "grad_norm": 1.6610630565710844, + "learning_rate": 2.3202237230467215e-06, + "loss": 0.8435, + "num_input_tokens_seen": 168093940, + "step": 5160 + }, + { + "epoch": 0.4654371646300221, + "flos": 22203428280480.0, + "grad_norm": 2.5195448764166404, + "learning_rate": 2.3196470357385338e-06, + "loss": 0.7107, + "num_input_tokens_seen": 168123615, + "step": 5161 + }, + { + "epoch": 0.46552734815349234, + "flos": 31676851787520.0, + "grad_norm": 2.6215996289599577, + "learning_rate": 2.319070321156299e-06, + "loss": 0.7474, + "num_input_tokens_seen": 168155530, + "step": 5162 + }, + { + "epoch": 0.4656175316769626, + "flos": 27304442433120.0, + "grad_norm": 1.5060501472564005, + "learning_rate": 2.318493579349224e-06, + "loss": 0.886, + "num_input_tokens_seen": 168187475, + "step": 5163 + }, + { + "epoch": 0.46570771520043286, + "flos": 18525347472000.0, + "grad_norm": 1.9428256888307778, + "learning_rate": 2.317916810366522e-06, + "loss": 0.7433, + "num_input_tokens_seen": 168215555, + "step": 5164 + }, + { + "epoch": 0.46579789872390315, + "flos": 25916454396960.0, + "grad_norm": 1.7474497878210438, + "learning_rate": 2.317340014257404e-06, + "loss": 0.6998, + "num_input_tokens_seen": 168244645, + "step": 5165 + }, + { + "epoch": 0.4658880822473734, + "flos": 23040251450880.0, + "grad_norm": 1.4254782994979185, + "learning_rate": 2.316763191071086e-06, + "loss": 0.6825, + "num_input_tokens_seen": 168273475, + "step": 5166 + }, + { + "epoch": 0.46597826577084367, + "flos": 43700893880160.0, + "grad_norm": 1.739789398536609, + "learning_rate": 2.316186340856787e-06, + "loss": 0.7132, + "num_input_tokens_seen": 168305460, + "step": 5167 + }, + { + "epoch": 0.4660684492943139, + "flos": 25950581970240.0, + "grad_norm": 2.0750894963830517, + "learning_rate": 2.315609463663725e-06, + "loss": 0.7056, + "num_input_tokens_seen": 168332040, + "step": 5168 + }, + { + "epoch": 0.4661586328177842, + "flos": 37834285284480.0, + "grad_norm": 2.0296646974382493, + "learning_rate": 2.315032559541123e-06, + "loss": 0.7297, + "num_input_tokens_seen": 168361730, + "step": 5169 + }, + { + "epoch": 0.4662488163412545, + "flos": 28616667031200.0, + "grad_norm": 1.8825615680044228, + "learning_rate": 2.314455628538207e-06, + "loss": 0.7692, + "num_input_tokens_seen": 168388950, + "step": 5170 + }, + { + "epoch": 0.4663389998647247, + "flos": 46717435604640.0, + "grad_norm": 1.5625805643292545, + "learning_rate": 2.3138786707042023e-06, + "loss": 0.7251, + "num_input_tokens_seen": 168421265, + "step": 5171 + }, + { + "epoch": 0.466429183388195, + "flos": 24317121873600.0, + "grad_norm": 1.5786499812081, + "learning_rate": 2.3133016860883387e-06, + "loss": 0.764, + "num_input_tokens_seen": 168451895, + "step": 5172 + }, + { + "epoch": 0.46651936691166523, + "flos": 22678681029120.0, + "grad_norm": 1.7435599767041288, + "learning_rate": 2.3127246747398475e-06, + "loss": 0.7284, + "num_input_tokens_seen": 168481405, + "step": 5173 + }, + { + "epoch": 0.4666095504351355, + "flos": 31239852455520.0, + "grad_norm": 1.3674576410710484, + "learning_rate": 2.312147636707963e-06, + "loss": 0.75, + "num_input_tokens_seen": 168513230, + "step": 5174 + }, + { + "epoch": 0.46669973395860576, + "flos": 18233667666240.0, + "grad_norm": 1.764915728948183, + "learning_rate": 2.3115705720419214e-06, + "loss": 0.7369, + "num_input_tokens_seen": 168539535, + "step": 5175 + }, + { + "epoch": 0.46678991748207604, + "flos": 33754410640320.0, + "grad_norm": 1.9459028458959335, + "learning_rate": 2.31099348079096e-06, + "loss": 0.7723, + "num_input_tokens_seen": 168571170, + "step": 5176 + }, + { + "epoch": 0.4668801010055463, + "flos": 22424102377440.0, + "grad_norm": 1.4480237944054748, + "learning_rate": 2.31041636300432e-06, + "loss": 0.7956, + "num_input_tokens_seen": 168598840, + "step": 5177 + }, + { + "epoch": 0.46697028452901657, + "flos": 21805388590560.0, + "grad_norm": 1.835574499727285, + "learning_rate": 2.3098392187312445e-06, + "loss": 0.8375, + "num_input_tokens_seen": 168626260, + "step": 5178 + }, + { + "epoch": 0.4670604680524868, + "flos": 21513337087200.0, + "grad_norm": 2.0114203074288595, + "learning_rate": 2.309262048020978e-06, + "loss": 0.7307, + "num_input_tokens_seen": 168656080, + "step": 5179 + }, + { + "epoch": 0.4671506515759571, + "flos": 16446338998560.0, + "grad_norm": 1.6618675118715571, + "learning_rate": 2.308684850922769e-06, + "loss": 0.7832, + "num_input_tokens_seen": 168681865, + "step": 5180 + }, + { + "epoch": 0.4672408350994273, + "flos": 17978977505280.0, + "grad_norm": 1.8341542775649795, + "learning_rate": 2.3081076274858664e-06, + "loss": 0.7475, + "num_input_tokens_seen": 168708920, + "step": 5181 + }, + { + "epoch": 0.4673310186228976, + "flos": 19284920413920.0, + "grad_norm": 1.466643425728429, + "learning_rate": 2.307530377759522e-06, + "loss": 0.7781, + "num_input_tokens_seen": 168736525, + "step": 5182 + }, + { + "epoch": 0.46742120214636784, + "flos": 20200926640320.0, + "grad_norm": 2.1249279683190285, + "learning_rate": 2.30695310179299e-06, + "loss": 0.6837, + "num_input_tokens_seen": 168762705, + "step": 5183 + }, + { + "epoch": 0.46751138566983813, + "flos": 23478811912800.0, + "grad_norm": 1.5128693846795591, + "learning_rate": 2.3063757996355267e-06, + "loss": 0.7518, + "num_input_tokens_seen": 168791995, + "step": 5184 + }, + { + "epoch": 0.46760156919330836, + "flos": 17723915646720.0, + "grad_norm": 2.01518516768827, + "learning_rate": 2.3057984713363903e-06, + "loss": 0.783, + "num_input_tokens_seen": 168818045, + "step": 5185 + }, + { + "epoch": 0.46769175271677865, + "flos": 71176972029600.0, + "grad_norm": 0.6954944672518728, + "learning_rate": 2.3052211169448436e-06, + "loss": 0.556, + "num_input_tokens_seen": 168917620, + "step": 5186 + }, + { + "epoch": 0.4677819362402489, + "flos": 32661373348800.0, + "grad_norm": 1.7790695363436118, + "learning_rate": 2.3046437365101474e-06, + "loss": 0.6954, + "num_input_tokens_seen": 168949460, + "step": 5187 + }, + { + "epoch": 0.4678721197637192, + "flos": 21439580816160.0, + "grad_norm": 1.787639601798066, + "learning_rate": 2.3040663300815673e-06, + "loss": 0.7281, + "num_input_tokens_seen": 168976355, + "step": 5188 + }, + { + "epoch": 0.4679623032871894, + "flos": 65410404458880.0, + "grad_norm": 0.7275193999020828, + "learning_rate": 2.3034888977083723e-06, + "loss": 0.5646, + "num_input_tokens_seen": 169064405, + "step": 5189 + }, + { + "epoch": 0.4680524868106597, + "flos": 28362348567840.0, + "grad_norm": 2.180625686273661, + "learning_rate": 2.30291143943983e-06, + "loss": 0.6635, + "num_input_tokens_seen": 169094435, + "step": 5190 + }, + { + "epoch": 0.46814267033412993, + "flos": 32771227190400.0, + "grad_norm": 2.1512043839826704, + "learning_rate": 2.3023339553252145e-06, + "loss": 0.7145, + "num_input_tokens_seen": 169126230, + "step": 5191 + }, + { + "epoch": 0.4682328538576002, + "flos": 23881534992480.0, + "grad_norm": 2.366846544745927, + "learning_rate": 2.301756445413799e-06, + "loss": 0.7862, + "num_input_tokens_seen": 169155865, + "step": 5192 + }, + { + "epoch": 0.4683230373810705, + "flos": 25993704625440.0, + "grad_norm": 1.8618553991909086, + "learning_rate": 2.3011789097548585e-06, + "loss": 0.7595, + "num_input_tokens_seen": 169184640, + "step": 5193 + }, + { + "epoch": 0.46841322090454074, + "flos": 24934609058400.0, + "grad_norm": 1.8473292377930837, + "learning_rate": 2.3006013483976738e-06, + "loss": 0.7489, + "num_input_tokens_seen": 169214480, + "step": 5194 + }, + { + "epoch": 0.468503404428011, + "flos": 19685041610400.0, + "grad_norm": 2.2873831046832387, + "learning_rate": 2.300023761391524e-06, + "loss": 0.7492, + "num_input_tokens_seen": 169240275, + "step": 5195 + }, + { + "epoch": 0.46859358795148126, + "flos": 25629346471680.0, + "grad_norm": 2.09092430629396, + "learning_rate": 2.299446148785693e-06, + "loss": 0.8046, + "num_input_tokens_seen": 169266910, + "step": 5196 + }, + { + "epoch": 0.46868377147495155, + "flos": 15063703407840.0, + "grad_norm": 2.107005618246778, + "learning_rate": 2.2988685106294654e-06, + "loss": 0.7385, + "num_input_tokens_seen": 169291910, + "step": 5197 + }, + { + "epoch": 0.4687739549984218, + "flos": 32442743588640.0, + "grad_norm": 2.0399613427258063, + "learning_rate": 2.2982908469721284e-06, + "loss": 0.6514, + "num_input_tokens_seen": 169319720, + "step": 5198 + }, + { + "epoch": 0.46886413852189207, + "flos": 23989604685600.0, + "grad_norm": 1.8474079634643994, + "learning_rate": 2.2977131578629714e-06, + "loss": 0.8223, + "num_input_tokens_seen": 169348640, + "step": 5199 + }, + { + "epoch": 0.4689543220453623, + "flos": 23007164630880.0, + "grad_norm": 1.6641571304565093, + "learning_rate": 2.297135443351286e-06, + "loss": 0.7759, + "num_input_tokens_seen": 169379545, + "step": 5200 + }, + { + "epoch": 0.4690445055688326, + "flos": 25995005567040.0, + "grad_norm": 1.3696200718994274, + "learning_rate": 2.296557703486367e-06, + "loss": 0.7653, + "num_input_tokens_seen": 169410540, + "step": 5201 + }, + { + "epoch": 0.4691346890923028, + "flos": 22781435446560.0, + "grad_norm": 1.9289164373314986, + "learning_rate": 2.295979938317509e-06, + "loss": 0.7799, + "num_input_tokens_seen": 169438630, + "step": 5202 + }, + { + "epoch": 0.4692248726157731, + "flos": 15647025849600.0, + "grad_norm": 2.3053744052157508, + "learning_rate": 2.295402147894011e-06, + "loss": 0.7561, + "num_input_tokens_seen": 169467115, + "step": 5203 + }, + { + "epoch": 0.46931505613924335, + "flos": 36483472741920.0, + "grad_norm": 1.5329121501856418, + "learning_rate": 2.2948243322651723e-06, + "loss": 0.6789, + "num_input_tokens_seen": 169501380, + "step": 5204 + }, + { + "epoch": 0.46940523966271364, + "flos": 32296792176480.0, + "grad_norm": 1.5943129050299059, + "learning_rate": 2.2942464914802962e-06, + "loss": 0.7584, + "num_input_tokens_seen": 169531930, + "step": 5205 + }, + { + "epoch": 0.46949542318618387, + "flos": 17468928127680.0, + "grad_norm": 1.9328526165067115, + "learning_rate": 2.293668625588687e-06, + "loss": 0.7339, + "num_input_tokens_seen": 169557170, + "step": 5206 + }, + { + "epoch": 0.46958560670965416, + "flos": 15282816374880.0, + "grad_norm": 2.1813447729469853, + "learning_rate": 2.293090734639651e-06, + "loss": 0.799, + "num_input_tokens_seen": 169581985, + "step": 5207 + }, + { + "epoch": 0.4696757902331244, + "flos": 23406579601920.0, + "grad_norm": 2.390419745516134, + "learning_rate": 2.2925128186824983e-06, + "loss": 0.8713, + "num_input_tokens_seen": 169608980, + "step": 5208 + }, + { + "epoch": 0.4697659737565947, + "flos": 25738196729760.0, + "grad_norm": 2.312148608754386, + "learning_rate": 2.2919348777665384e-06, + "loss": 0.8568, + "num_input_tokens_seen": 169636780, + "step": 5209 + }, + { + "epoch": 0.4698561572800649, + "flos": 25046730255360.0, + "grad_norm": 1.748005888900614, + "learning_rate": 2.2913569119410856e-06, + "loss": 0.7609, + "num_input_tokens_seen": 169665205, + "step": 5210 + }, + { + "epoch": 0.4699463408035352, + "flos": 27123174015360.0, + "grad_norm": 1.966194371178006, + "learning_rate": 2.290778921255454e-06, + "loss": 0.7801, + "num_input_tokens_seen": 169694810, + "step": 5211 + }, + { + "epoch": 0.47003652432700543, + "flos": 21803307084000.0, + "grad_norm": 1.5270498632047826, + "learning_rate": 2.2902009057589613e-06, + "loss": 0.7896, + "num_input_tokens_seen": 169725625, + "step": 5212 + }, + { + "epoch": 0.4701267078504757, + "flos": 27158007814080.0, + "grad_norm": 1.4154918043787996, + "learning_rate": 2.2896228655009276e-06, + "loss": 0.711, + "num_input_tokens_seen": 169758395, + "step": 5213 + }, + { + "epoch": 0.47021689137394596, + "flos": 20565619321920.0, + "grad_norm": 5.884778589638889, + "learning_rate": 2.289044800530674e-06, + "loss": 0.7236, + "num_input_tokens_seen": 169786175, + "step": 5214 + }, + { + "epoch": 0.47030707489741624, + "flos": 13898991351840.0, + "grad_norm": 1.9902559378432003, + "learning_rate": 2.2884667108975245e-06, + "loss": 0.7205, + "num_input_tokens_seen": 169812690, + "step": 5215 + }, + { + "epoch": 0.4703972584208865, + "flos": 20966521083360.0, + "grad_norm": 1.632102118195657, + "learning_rate": 2.287888596650804e-06, + "loss": 0.7922, + "num_input_tokens_seen": 169841755, + "step": 5216 + }, + { + "epoch": 0.47048744194435677, + "flos": 22200714888000.0, + "grad_norm": 1.926502580167229, + "learning_rate": 2.287310457839841e-06, + "loss": 0.7522, + "num_input_tokens_seen": 169869880, + "step": 5217 + }, + { + "epoch": 0.47057762546782705, + "flos": 26465203228320.0, + "grad_norm": 1.881862669885049, + "learning_rate": 2.286732294513966e-06, + "loss": 0.8268, + "num_input_tokens_seen": 169896445, + "step": 5218 + }, + { + "epoch": 0.4706678089912973, + "flos": 19801139971680.0, + "grad_norm": 1.683468913607883, + "learning_rate": 2.2861541067225106e-06, + "loss": 0.7668, + "num_input_tokens_seen": 169924545, + "step": 5219 + }, + { + "epoch": 0.4707579925147676, + "flos": 21513151238400.0, + "grad_norm": 2.8319118711369233, + "learning_rate": 2.2855758945148095e-06, + "loss": 0.7396, + "num_input_tokens_seen": 169953460, + "step": 5220 + }, + { + "epoch": 0.4708481760382378, + "flos": 16594111728960.0, + "grad_norm": 2.729347446268502, + "learning_rate": 2.2849976579401977e-06, + "loss": 0.8152, + "num_input_tokens_seen": 169980315, + "step": 5221 + }, + { + "epoch": 0.4709383595617081, + "flos": 19180195999200.0, + "grad_norm": 3.332772462070513, + "learning_rate": 2.284419397048014e-06, + "loss": 0.7467, + "num_input_tokens_seen": 170008290, + "step": 5222 + }, + { + "epoch": 0.47102854308517833, + "flos": 26831977416480.0, + "grad_norm": 1.688106992838072, + "learning_rate": 2.2838411118875997e-06, + "loss": 0.7632, + "num_input_tokens_seen": 170037375, + "step": 5223 + }, + { + "epoch": 0.4711187266086486, + "flos": 24645977172960.0, + "grad_norm": 1.5898740423442417, + "learning_rate": 2.283262802508296e-06, + "loss": 0.679, + "num_input_tokens_seen": 170065765, + "step": 5224 + }, + { + "epoch": 0.47120891013211885, + "flos": 25082939335200.0, + "grad_norm": 1.3943489874546895, + "learning_rate": 2.2826844689594492e-06, + "loss": 0.7342, + "num_input_tokens_seen": 170097600, + "step": 5225 + }, + { + "epoch": 0.47129909365558914, + "flos": 19545892264320.0, + "grad_norm": 1.7677962869295516, + "learning_rate": 2.282106111290404e-06, + "loss": 0.8288, + "num_input_tokens_seen": 170127965, + "step": 5226 + }, + { + "epoch": 0.4713892771790594, + "flos": 64635554745600.0, + "grad_norm": 0.7008727771947619, + "learning_rate": 2.2815277295505098e-06, + "loss": 0.6009, + "num_input_tokens_seen": 170215995, + "step": 5227 + }, + { + "epoch": 0.47147946070252966, + "flos": 22857830770560.0, + "grad_norm": 4.365842772549651, + "learning_rate": 2.2809493237891174e-06, + "loss": 0.7009, + "num_input_tokens_seen": 170243620, + "step": 5228 + }, + { + "epoch": 0.4715696442259999, + "flos": 24282585432960.0, + "grad_norm": 1.7004045229553502, + "learning_rate": 2.2803708940555796e-06, + "loss": 0.783, + "num_input_tokens_seen": 170272805, + "step": 5229 + }, + { + "epoch": 0.4716598277494702, + "flos": 25192421479200.0, + "grad_norm": 1.4212313022910157, + "learning_rate": 2.2797924403992514e-06, + "loss": 0.8269, + "num_input_tokens_seen": 170303480, + "step": 5230 + }, + { + "epoch": 0.4717500112729404, + "flos": 19101756338400.0, + "grad_norm": 1.9961786121462093, + "learning_rate": 2.2792139628694892e-06, + "loss": 0.7747, + "num_input_tokens_seen": 170330585, + "step": 5231 + }, + { + "epoch": 0.4718401947964107, + "flos": 25264876808640.0, + "grad_norm": 1.6872535808218319, + "learning_rate": 2.2786354615156524e-06, + "loss": 0.7151, + "num_input_tokens_seen": 170358500, + "step": 5232 + }, + { + "epoch": 0.47193037831988094, + "flos": 17755143978720.0, + "grad_norm": 1.9354535647167892, + "learning_rate": 2.2780569363871016e-06, + "loss": 0.8198, + "num_input_tokens_seen": 170384960, + "step": 5233 + }, + { + "epoch": 0.4720205618433512, + "flos": 67630086238560.0, + "grad_norm": 0.6158620145732211, + "learning_rate": 2.277478387533199e-06, + "loss": 0.5719, + "num_input_tokens_seen": 170471740, + "step": 5234 + }, + { + "epoch": 0.47211074536682146, + "flos": 22201235264640.0, + "grad_norm": 1.6941274553156047, + "learning_rate": 2.276899815003311e-06, + "loss": 0.7912, + "num_input_tokens_seen": 170499720, + "step": 5235 + }, + { + "epoch": 0.47220092889029175, + "flos": 18853347866880.0, + "grad_norm": 1.5871979914267966, + "learning_rate": 2.2763212188468045e-06, + "loss": 0.7476, + "num_input_tokens_seen": 170527025, + "step": 5236 + }, + { + "epoch": 0.472291112413762, + "flos": 24280355247360.0, + "grad_norm": 1.4533651010004438, + "learning_rate": 2.2757425991130473e-06, + "loss": 0.7682, + "num_input_tokens_seen": 170557175, + "step": 5237 + }, + { + "epoch": 0.47238129593723227, + "flos": 67819978040640.0, + "grad_norm": 0.7596039587775469, + "learning_rate": 2.2751639558514117e-06, + "loss": 0.5851, + "num_input_tokens_seen": 170646095, + "step": 5238 + }, + { + "epoch": 0.4724714794607025, + "flos": 25738568427360.0, + "grad_norm": 3.1567228359661783, + "learning_rate": 2.2745852891112697e-06, + "loss": 0.7944, + "num_input_tokens_seen": 170677400, + "step": 5239 + }, + { + "epoch": 0.4725616629841728, + "flos": 23584688590080.0, + "grad_norm": 1.9957896260338357, + "learning_rate": 2.274006598941997e-06, + "loss": 0.7592, + "num_input_tokens_seen": 170707220, + "step": 5240 + }, + { + "epoch": 0.4726518465076431, + "flos": 31711871435040.0, + "grad_norm": 1.7922043843625561, + "learning_rate": 2.27342788539297e-06, + "loss": 0.7339, + "num_input_tokens_seen": 170740945, + "step": 5241 + }, + { + "epoch": 0.4727420300311133, + "flos": 29673978449760.0, + "grad_norm": 1.527449270052203, + "learning_rate": 2.2728491485135684e-06, + "loss": 0.712, + "num_input_tokens_seen": 170769385, + "step": 5242 + }, + { + "epoch": 0.4728322135545836, + "flos": 26212260046080.0, + "grad_norm": 2.040785839524558, + "learning_rate": 2.272270388353173e-06, + "loss": 0.8116, + "num_input_tokens_seen": 170797885, + "step": 5243 + }, + { + "epoch": 0.47292239707805384, + "flos": 24973828888800.0, + "grad_norm": 1.7937486070368338, + "learning_rate": 2.2716916049611666e-06, + "loss": 0.7286, + "num_input_tokens_seen": 170828350, + "step": 5244 + }, + { + "epoch": 0.4730125806015241, + "flos": 24095258344320.0, + "grad_norm": 1.8021098302168446, + "learning_rate": 2.2711127983869346e-06, + "loss": 0.7367, + "num_input_tokens_seen": 170855935, + "step": 5245 + }, + { + "epoch": 0.47310276412499436, + "flos": 27811704078720.0, + "grad_norm": 1.5090952738419885, + "learning_rate": 2.270533968679864e-06, + "loss": 0.7648, + "num_input_tokens_seen": 170884810, + "step": 5246 + }, + { + "epoch": 0.47319294764846465, + "flos": 18307052239680.0, + "grad_norm": 3.3253655254250694, + "learning_rate": 2.269955115889343e-06, + "loss": 0.7615, + "num_input_tokens_seen": 170913095, + "step": 5247 + }, + { + "epoch": 0.4732831311719349, + "flos": 20857187618400.0, + "grad_norm": 2.171977088779283, + "learning_rate": 2.269376240064763e-06, + "loss": 0.7152, + "num_input_tokens_seen": 170941290, + "step": 5248 + }, + { + "epoch": 0.47337331469540517, + "flos": 26170326823200.0, + "grad_norm": 1.7566791865260372, + "learning_rate": 2.268797341255517e-06, + "loss": 0.7739, + "num_input_tokens_seen": 170972965, + "step": 5249 + }, + { + "epoch": 0.4734634982188754, + "flos": 21621109422240.0, + "grad_norm": 1.988848232526229, + "learning_rate": 2.268218419511e-06, + "loss": 0.6712, + "num_input_tokens_seen": 170999720, + "step": 5250 + }, + { + "epoch": 0.4735536817423457, + "flos": 23657329768320.0, + "grad_norm": 1.733335234873291, + "learning_rate": 2.267639474880608e-06, + "loss": 0.6794, + "num_input_tokens_seen": 171027920, + "step": 5251 + }, + { + "epoch": 0.4736438652658159, + "flos": 24973345681920.0, + "grad_norm": 1.5152961924865895, + "learning_rate": 2.2670605074137407e-06, + "loss": 0.7356, + "num_input_tokens_seen": 171058530, + "step": 5252 + }, + { + "epoch": 0.4737340487892862, + "flos": 24209015010720.0, + "grad_norm": 1.7990738393357635, + "learning_rate": 2.2664815171597983e-06, + "loss": 0.7381, + "num_input_tokens_seen": 171085920, + "step": 5253 + }, + { + "epoch": 0.47382423231275644, + "flos": 19982594238240.0, + "grad_norm": 1.6466201237550075, + "learning_rate": 2.265902504168183e-06, + "loss": 0.6996, + "num_input_tokens_seen": 171112800, + "step": 5254 + }, + { + "epoch": 0.47391441583622673, + "flos": 19909692871680.0, + "grad_norm": 1.6402624547107887, + "learning_rate": 2.2653234684883007e-06, + "loss": 0.7099, + "num_input_tokens_seen": 171138295, + "step": 5255 + }, + { + "epoch": 0.47400459935969697, + "flos": 13420095966720.0, + "grad_norm": 2.329412731441313, + "learning_rate": 2.264744410169556e-06, + "loss": 0.6752, + "num_input_tokens_seen": 171163570, + "step": 5256 + }, + { + "epoch": 0.47409478288316725, + "flos": 30183804808800.0, + "grad_norm": 2.038013641581303, + "learning_rate": 2.264165329261359e-06, + "loss": 0.7353, + "num_input_tokens_seen": 171196865, + "step": 5257 + }, + { + "epoch": 0.4741849664066375, + "flos": 67058063403840.0, + "grad_norm": 0.5738187421702872, + "learning_rate": 2.26358622581312e-06, + "loss": 0.5746, + "num_input_tokens_seen": 171298160, + "step": 5258 + }, + { + "epoch": 0.4742751499301078, + "flos": 28285618716000.0, + "grad_norm": 1.880838649552381, + "learning_rate": 2.2630070998742504e-06, + "loss": 0.817, + "num_input_tokens_seen": 171325235, + "step": 5259 + }, + { + "epoch": 0.474365333453578, + "flos": 22787754305760.0, + "grad_norm": 1.6696840708099627, + "learning_rate": 2.262427951494165e-06, + "loss": 0.7778, + "num_input_tokens_seen": 171355660, + "step": 5260 + }, + { + "epoch": 0.4744555169770483, + "flos": 67534728603360.0, + "grad_norm": 0.6498629975822026, + "learning_rate": 2.2618487807222794e-06, + "loss": 0.6142, + "num_input_tokens_seen": 171451225, + "step": 5261 + }, + { + "epoch": 0.47454570050051853, + "flos": 26686286192640.0, + "grad_norm": 1.5236610444212135, + "learning_rate": 2.261269587608012e-06, + "loss": 0.8366, + "num_input_tokens_seen": 171480765, + "step": 5262 + }, + { + "epoch": 0.4746358840239888, + "flos": 25957607054880.0, + "grad_norm": 1.5193966999646284, + "learning_rate": 2.260690372200783e-06, + "loss": 0.7601, + "num_input_tokens_seen": 171509425, + "step": 5263 + }, + { + "epoch": 0.47472606754745905, + "flos": 21328463202720.0, + "grad_norm": 1.699267507270129, + "learning_rate": 2.2601111345500138e-06, + "loss": 0.8413, + "num_input_tokens_seen": 171539710, + "step": 5264 + }, + { + "epoch": 0.47481625107092934, + "flos": 24643300950240.0, + "grad_norm": 1.6136417309455648, + "learning_rate": 2.2595318747051286e-06, + "loss": 0.7452, + "num_input_tokens_seen": 171569675, + "step": 5265 + }, + { + "epoch": 0.47490643459439963, + "flos": 21366939637920.0, + "grad_norm": 1.6089551804003837, + "learning_rate": 2.258952592715553e-06, + "loss": 0.7775, + "num_input_tokens_seen": 171597320, + "step": 5266 + }, + { + "epoch": 0.47499661811786986, + "flos": 64023977552640.0, + "grad_norm": 0.7054054155783758, + "learning_rate": 2.2583732886307142e-06, + "loss": 0.6171, + "num_input_tokens_seen": 171689825, + "step": 5267 + }, + { + "epoch": 0.47508680164134015, + "flos": 23152186799040.0, + "grad_norm": 1.660101891917988, + "learning_rate": 2.2577939625000414e-06, + "loss": 0.703, + "num_input_tokens_seen": 171719790, + "step": 5268 + }, + { + "epoch": 0.4751769851648104, + "flos": 66473291341440.0, + "grad_norm": 0.67215292749032, + "learning_rate": 2.257214614372967e-06, + "loss": 0.5768, + "num_input_tokens_seen": 171806735, + "step": 5269 + }, + { + "epoch": 0.4752671686882807, + "flos": 23152223968800.0, + "grad_norm": 1.5566079245163549, + "learning_rate": 2.2566352442989227e-06, + "loss": 0.7796, + "num_input_tokens_seen": 171835125, + "step": 5270 + }, + { + "epoch": 0.4753573522117509, + "flos": 22423990868160.0, + "grad_norm": 1.9423189224621598, + "learning_rate": 2.256055852327344e-06, + "loss": 0.7603, + "num_input_tokens_seen": 171864715, + "step": 5271 + }, + { + "epoch": 0.4754475357352212, + "flos": 34151558256000.0, + "grad_norm": 1.5620275927706213, + "learning_rate": 2.2554764385076685e-06, + "loss": 0.7018, + "num_input_tokens_seen": 171894605, + "step": 5272 + }, + { + "epoch": 0.4755377192586914, + "flos": 35134927554720.0, + "grad_norm": 1.5183183001553224, + "learning_rate": 2.2548970028893348e-06, + "loss": 0.767, + "num_input_tokens_seen": 171925230, + "step": 5273 + }, + { + "epoch": 0.4756279027821617, + "flos": 34082039337600.0, + "grad_norm": 2.1206304841626955, + "learning_rate": 2.254317545521783e-06, + "loss": 0.7658, + "num_input_tokens_seen": 171952730, + "step": 5274 + }, + { + "epoch": 0.47571808630563195, + "flos": 20602162929600.0, + "grad_norm": 2.355367427437002, + "learning_rate": 2.253738066454457e-06, + "loss": 0.7855, + "num_input_tokens_seen": 171980090, + "step": 5275 + }, + { + "epoch": 0.47580826982910224, + "flos": 24536457859200.0, + "grad_norm": 1.6803774813771557, + "learning_rate": 2.2531585657367986e-06, + "loss": 0.7304, + "num_input_tokens_seen": 172008490, + "step": 5276 + }, + { + "epoch": 0.47589845335257247, + "flos": 25518414707040.0, + "grad_norm": 3.0721354035597415, + "learning_rate": 2.252579043418256e-06, + "loss": 0.7112, + "num_input_tokens_seen": 172038255, + "step": 5277 + }, + { + "epoch": 0.47598863687604276, + "flos": 21545569002720.0, + "grad_norm": 2.6036006973687633, + "learning_rate": 2.251999499548277e-06, + "loss": 0.7277, + "num_input_tokens_seen": 172066475, + "step": 5278 + }, + { + "epoch": 0.476078820399513, + "flos": 37797927525600.0, + "grad_norm": 1.9443594194926512, + "learning_rate": 2.251419934176311e-06, + "loss": 0.7167, + "num_input_tokens_seen": 172096630, + "step": 5279 + }, + { + "epoch": 0.4761690039229833, + "flos": 22237221325920.0, + "grad_norm": 1.4855048447269619, + "learning_rate": 2.25084034735181e-06, + "loss": 0.8036, + "num_input_tokens_seen": 172124370, + "step": 5280 + }, + { + "epoch": 0.4762591874464535, + "flos": 22783145255520.0, + "grad_norm": 1.7296471292114342, + "learning_rate": 2.2502607391242274e-06, + "loss": 0.8019, + "num_input_tokens_seen": 172154415, + "step": 5281 + }, + { + "epoch": 0.4763493709699238, + "flos": 57703191462240.0, + "grad_norm": 0.6596517045552386, + "learning_rate": 2.2496811095430182e-06, + "loss": 0.5815, + "num_input_tokens_seen": 172241715, + "step": 5282 + }, + { + "epoch": 0.47643955449339404, + "flos": 21400732683360.0, + "grad_norm": 2.385359006848389, + "learning_rate": 2.249101458657641e-06, + "loss": 0.7525, + "num_input_tokens_seen": 172268410, + "step": 5283 + }, + { + "epoch": 0.4765297380168643, + "flos": 18124297031520.0, + "grad_norm": 1.9650473517616636, + "learning_rate": 2.2485217865175526e-06, + "loss": 0.8239, + "num_input_tokens_seen": 172297040, + "step": 5284 + }, + { + "epoch": 0.47661992154033456, + "flos": 21585680907360.0, + "grad_norm": 2.5516471853411615, + "learning_rate": 2.2479420931722156e-06, + "loss": 0.8919, + "num_input_tokens_seen": 172325595, + "step": 5285 + }, + { + "epoch": 0.47671010506380485, + "flos": 64725963069120.0, + "grad_norm": 0.6980127771350337, + "learning_rate": 2.2473623786710923e-06, + "loss": 0.6186, + "num_input_tokens_seen": 172425125, + "step": 5286 + }, + { + "epoch": 0.4768002885872751, + "flos": 26650262961600.0, + "grad_norm": 1.9561233698898877, + "learning_rate": 2.2467826430636465e-06, + "loss": 0.67, + "num_input_tokens_seen": 172456115, + "step": 5287 + }, + { + "epoch": 0.47689047211074537, + "flos": 24493521052800.0, + "grad_norm": 1.4561103590128435, + "learning_rate": 2.246202886399345e-06, + "loss": 0.779, + "num_input_tokens_seen": 172486000, + "step": 5288 + }, + { + "epoch": 0.47698065563421566, + "flos": 22819428674880.0, + "grad_norm": 1.9770800851434114, + "learning_rate": 2.2456231087276556e-06, + "loss": 0.7505, + "num_input_tokens_seen": 172514390, + "step": 5289 + }, + { + "epoch": 0.4770708391576859, + "flos": 28583617380960.0, + "grad_norm": 1.5281278738919246, + "learning_rate": 2.245043310098048e-06, + "loss": 0.6834, + "num_input_tokens_seen": 172548390, + "step": 5290 + }, + { + "epoch": 0.4771610226811562, + "flos": 22387372920960.0, + "grad_norm": 1.9916247002983214, + "learning_rate": 2.244463490559995e-06, + "loss": 0.795, + "num_input_tokens_seen": 172574230, + "step": 5291 + }, + { + "epoch": 0.4772512062046264, + "flos": 21257866361280.0, + "grad_norm": 1.6188341685914496, + "learning_rate": 2.2438836501629683e-06, + "loss": 0.7657, + "num_input_tokens_seen": 172602440, + "step": 5292 + }, + { + "epoch": 0.4773413897280967, + "flos": 25410642372000.0, + "grad_norm": 1.4883456479054724, + "learning_rate": 2.2433037889564437e-06, + "loss": 0.7649, + "num_input_tokens_seen": 172635125, + "step": 5293 + }, + { + "epoch": 0.47743157325156693, + "flos": 22380756703680.0, + "grad_norm": 1.9927784361192882, + "learning_rate": 2.242723906989899e-06, + "loss": 0.6996, + "num_input_tokens_seen": 172667940, + "step": 5294 + }, + { + "epoch": 0.4775217567750372, + "flos": 27779472163200.0, + "grad_norm": 2.087637676335975, + "learning_rate": 2.2421440043128114e-06, + "loss": 0.6885, + "num_input_tokens_seen": 172699895, + "step": 5295 + }, + { + "epoch": 0.47761194029850745, + "flos": 13052504043840.0, + "grad_norm": 4.106501165456965, + "learning_rate": 2.241564080974662e-06, + "loss": 0.6711, + "num_input_tokens_seen": 172723370, + "step": 5296 + }, + { + "epoch": 0.47770212382197774, + "flos": 22606114190400.0, + "grad_norm": 1.9530010336635373, + "learning_rate": 2.2409841370249343e-06, + "loss": 0.7062, + "num_input_tokens_seen": 172751185, + "step": 5297 + }, + { + "epoch": 0.477792307345448, + "flos": 16339235719200.0, + "grad_norm": 1.9244816168811223, + "learning_rate": 2.2404041725131106e-06, + "loss": 0.8406, + "num_input_tokens_seen": 172776460, + "step": 5298 + }, + { + "epoch": 0.47788249086891826, + "flos": 21075557190240.0, + "grad_norm": 3.6448014170445475, + "learning_rate": 2.239824187488677e-06, + "loss": 0.7316, + "num_input_tokens_seen": 172802915, + "step": 5299 + }, + { + "epoch": 0.4779726743923885, + "flos": 64071857937120.0, + "grad_norm": 0.7239560371929686, + "learning_rate": 2.239244182001122e-06, + "loss": 0.6178, + "num_input_tokens_seen": 172902115, + "step": 5300 + }, + { + "epoch": 0.4780628579158588, + "flos": 28032229496640.0, + "grad_norm": 1.7560077812439518, + "learning_rate": 2.2386641560999336e-06, + "loss": 0.7714, + "num_input_tokens_seen": 172931815, + "step": 5301 + }, + { + "epoch": 0.478153041439329, + "flos": 22714964448480.0, + "grad_norm": 1.783441767857072, + "learning_rate": 2.238084109834604e-06, + "loss": 0.792, + "num_input_tokens_seen": 172959790, + "step": 5302 + }, + { + "epoch": 0.4782432249627993, + "flos": 22345253849280.0, + "grad_norm": 1.900465584479213, + "learning_rate": 2.237504043254625e-06, + "loss": 0.7596, + "num_input_tokens_seen": 172987035, + "step": 5303 + }, + { + "epoch": 0.47833340848626954, + "flos": 31746296366400.0, + "grad_norm": 1.454015889407602, + "learning_rate": 2.2369239564094915e-06, + "loss": 0.7274, + "num_input_tokens_seen": 173017595, + "step": 5304 + }, + { + "epoch": 0.47842359200973983, + "flos": 23329255033920.0, + "grad_norm": 1.6352505528668215, + "learning_rate": 2.2363438493486995e-06, + "loss": 0.6856, + "num_input_tokens_seen": 173047705, + "step": 5305 + }, + { + "epoch": 0.47851377553321006, + "flos": 23225757221280.0, + "grad_norm": 2.5010347132418653, + "learning_rate": 2.235763722121747e-06, + "loss": 0.6444, + "num_input_tokens_seen": 173076555, + "step": 5306 + }, + { + "epoch": 0.47860395905668035, + "flos": 27305854884000.0, + "grad_norm": 2.957028347401977, + "learning_rate": 2.2351835747781346e-06, + "loss": 0.7757, + "num_input_tokens_seen": 173106660, + "step": 5307 + }, + { + "epoch": 0.4786941425801506, + "flos": 23661269762880.0, + "grad_norm": 1.605409644308143, + "learning_rate": 2.234603407367362e-06, + "loss": 0.8001, + "num_input_tokens_seen": 173136135, + "step": 5308 + }, + { + "epoch": 0.47878432610362087, + "flos": 15824911819200.0, + "grad_norm": 1.7318995224790423, + "learning_rate": 2.2340232199389337e-06, + "loss": 0.7696, + "num_input_tokens_seen": 173161865, + "step": 5309 + }, + { + "epoch": 0.4788745096270911, + "flos": 24824457858720.0, + "grad_norm": 1.5927524810853122, + "learning_rate": 2.2334430125423538e-06, + "loss": 0.6379, + "num_input_tokens_seen": 173192150, + "step": 5310 + }, + { + "epoch": 0.4789646931505614, + "flos": 29448100586400.0, + "grad_norm": 1.858861323334212, + "learning_rate": 2.232862785227128e-06, + "loss": 0.804, + "num_input_tokens_seen": 173218910, + "step": 5311 + }, + { + "epoch": 0.4790548766740317, + "flos": 21111989288640.0, + "grad_norm": 1.6554106258829726, + "learning_rate": 2.232282538042766e-06, + "loss": 0.7915, + "num_input_tokens_seen": 173248245, + "step": 5312 + }, + { + "epoch": 0.4791450601975019, + "flos": 17533131770400.0, + "grad_norm": 2.1859116373447707, + "learning_rate": 2.231702271038777e-06, + "loss": 0.8045, + "num_input_tokens_seen": 173273150, + "step": 5313 + }, + { + "epoch": 0.4792352437209722, + "flos": 26794467395040.0, + "grad_norm": 2.0488210265233917, + "learning_rate": 2.231121984264673e-06, + "loss": 0.7296, + "num_input_tokens_seen": 173302510, + "step": 5314 + }, + { + "epoch": 0.47932542724444244, + "flos": 19977464811360.0, + "grad_norm": 1.694731371025253, + "learning_rate": 2.2305416777699665e-06, + "loss": 0.7872, + "num_input_tokens_seen": 173330775, + "step": 5315 + }, + { + "epoch": 0.4794156107679127, + "flos": 28981619901120.0, + "grad_norm": 1.574712406119703, + "learning_rate": 2.229961351604173e-06, + "loss": 0.755, + "num_input_tokens_seen": 173359490, + "step": 5316 + }, + { + "epoch": 0.47950579429138296, + "flos": 28068550085760.0, + "grad_norm": 1.6538611404545176, + "learning_rate": 2.2293810058168085e-06, + "loss": 0.714, + "num_input_tokens_seen": 173389725, + "step": 5317 + }, + { + "epoch": 0.47959597781485325, + "flos": 19726157098560.0, + "grad_norm": 2.2091653683933323, + "learning_rate": 2.2288006404573922e-06, + "loss": 0.7785, + "num_input_tokens_seen": 173416655, + "step": 5318 + }, + { + "epoch": 0.4796861613383235, + "flos": 20854362716640.0, + "grad_norm": 1.9553425521455516, + "learning_rate": 2.228220255575444e-06, + "loss": 0.7446, + "num_input_tokens_seen": 173444090, + "step": 5319 + }, + { + "epoch": 0.47977634486179377, + "flos": 25993853304480.0, + "grad_norm": 1.7488529669637707, + "learning_rate": 2.2276398512204847e-06, + "loss": 0.7013, + "num_input_tokens_seen": 173473140, + "step": 5320 + }, + { + "epoch": 0.479866528385264, + "flos": 22530908298720.0, + "grad_norm": 1.7969472455967208, + "learning_rate": 2.2270594274420382e-06, + "loss": 0.7632, + "num_input_tokens_seen": 173501620, + "step": 5321 + }, + { + "epoch": 0.4799567119087343, + "flos": 64154386271520.0, + "grad_norm": 0.6751350223635711, + "learning_rate": 2.22647898428963e-06, + "loss": 0.6075, + "num_input_tokens_seen": 173595145, + "step": 5322 + }, + { + "epoch": 0.4800468954322045, + "flos": 22275363233280.0, + "grad_norm": 1.6153613113287366, + "learning_rate": 2.225898521812785e-06, + "loss": 0.8007, + "num_input_tokens_seen": 173623170, + "step": 5323 + }, + { + "epoch": 0.4801370789556748, + "flos": 25812324698400.0, + "grad_norm": 1.7277156379471998, + "learning_rate": 2.2253180400610337e-06, + "loss": 0.8089, + "num_input_tokens_seen": 173651200, + "step": 5324 + }, + { + "epoch": 0.48022726247914505, + "flos": 25774591658400.0, + "grad_norm": 1.7894038245424053, + "learning_rate": 2.2247375390839037e-06, + "loss": 0.7373, + "num_input_tokens_seen": 173679485, + "step": 5325 + }, + { + "epoch": 0.48031744600261533, + "flos": 37397211612960.0, + "grad_norm": 1.737413374024377, + "learning_rate": 2.224157018930928e-06, + "loss": 0.7231, + "num_input_tokens_seen": 173707845, + "step": 5326 + }, + { + "epoch": 0.48040762952608557, + "flos": 24792002924640.0, + "grad_norm": 1.6028039354202246, + "learning_rate": 2.2235764796516395e-06, + "loss": 0.756, + "num_input_tokens_seen": 173737940, + "step": 5327 + }, + { + "epoch": 0.48049781304955586, + "flos": 25046247048480.0, + "grad_norm": 1.4861971462852237, + "learning_rate": 2.222995921295573e-06, + "loss": 0.753, + "num_input_tokens_seen": 173770155, + "step": 5328 + }, + { + "epoch": 0.4805879965730261, + "flos": 26174155308480.0, + "grad_norm": 1.7518846717483116, + "learning_rate": 2.222415343912265e-06, + "loss": 0.734, + "num_input_tokens_seen": 173798205, + "step": 5329 + }, + { + "epoch": 0.4806781800964964, + "flos": 14225542126080.0, + "grad_norm": 2.159300042934061, + "learning_rate": 2.221834747551254e-06, + "loss": 0.8192, + "num_input_tokens_seen": 173823705, + "step": 5330 + }, + { + "epoch": 0.4807683636199666, + "flos": 21658433594880.0, + "grad_norm": 2.347136936732531, + "learning_rate": 2.221254132262078e-06, + "loss": 0.7278, + "num_input_tokens_seen": 173851080, + "step": 5331 + }, + { + "epoch": 0.4808585471434369, + "flos": 28143793147200.0, + "grad_norm": 2.875568665365332, + "learning_rate": 2.2206734980942802e-06, + "loss": 0.8005, + "num_input_tokens_seen": 173879480, + "step": 5332 + }, + { + "epoch": 0.48094873066690713, + "flos": 38453259259680.0, + "grad_norm": 2.4555206264219094, + "learning_rate": 2.2200928450974024e-06, + "loss": 0.7342, + "num_input_tokens_seen": 173909015, + "step": 5333 + }, + { + "epoch": 0.4810389141903774, + "flos": 28576666635840.0, + "grad_norm": 1.7354092650085848, + "learning_rate": 2.21951217332099e-06, + "loss": 0.7779, + "num_input_tokens_seen": 173936020, + "step": 5334 + }, + { + "epoch": 0.48112909771384765, + "flos": 14439488496480.0, + "grad_norm": 1.9450639313141904, + "learning_rate": 2.2189314828145883e-06, + "loss": 0.6749, + "num_input_tokens_seen": 173960830, + "step": 5335 + }, + { + "epoch": 0.48121928123731794, + "flos": 27670361716800.0, + "grad_norm": 1.6252596721047439, + "learning_rate": 2.2183507736277453e-06, + "loss": 0.738, + "num_input_tokens_seen": 173992790, + "step": 5336 + }, + { + "epoch": 0.48130946476078823, + "flos": 24646237361280.0, + "grad_norm": 1.8980494601052895, + "learning_rate": 2.2177700458100107e-06, + "loss": 0.7729, + "num_input_tokens_seen": 174019440, + "step": 5337 + }, + { + "epoch": 0.48139964828425846, + "flos": 24241209756480.0, + "grad_norm": 1.7407894255471221, + "learning_rate": 2.2171892994109346e-06, + "loss": 0.7318, + "num_input_tokens_seen": 174047300, + "step": 5338 + }, + { + "epoch": 0.48148983180772875, + "flos": 27629394907680.0, + "grad_norm": 1.514239513651245, + "learning_rate": 2.21660853448007e-06, + "loss": 0.8213, + "num_input_tokens_seen": 174077520, + "step": 5339 + }, + { + "epoch": 0.481580015331199, + "flos": 20893694056320.0, + "grad_norm": 3.637714581534536, + "learning_rate": 2.2160277510669703e-06, + "loss": 0.7143, + "num_input_tokens_seen": 174106535, + "step": 5340 + }, + { + "epoch": 0.4816701988546693, + "flos": 30111237970080.0, + "grad_norm": 1.4543506826507941, + "learning_rate": 2.215446949221193e-06, + "loss": 0.7478, + "num_input_tokens_seen": 174139410, + "step": 5341 + }, + { + "epoch": 0.4817603823781395, + "flos": 21075222662400.0, + "grad_norm": 1.7707350385724012, + "learning_rate": 2.2148661289922924e-06, + "loss": 0.7661, + "num_input_tokens_seen": 174167040, + "step": 5342 + }, + { + "epoch": 0.4818505659016098, + "flos": 20091927703200.0, + "grad_norm": 1.7722323856175273, + "learning_rate": 2.21428529042983e-06, + "loss": 0.8191, + "num_input_tokens_seen": 174193560, + "step": 5343 + }, + { + "epoch": 0.48194074942508003, + "flos": 16776866937120.0, + "grad_norm": 1.782941450157902, + "learning_rate": 2.2137044335833647e-06, + "loss": 0.8035, + "num_input_tokens_seen": 174218275, + "step": 5344 + }, + { + "epoch": 0.4820309329485503, + "flos": 25629680999520.0, + "grad_norm": 2.2413830526912624, + "learning_rate": 2.213123558502459e-06, + "loss": 0.7411, + "num_input_tokens_seen": 174246110, + "step": 5345 + }, + { + "epoch": 0.48212111647202055, + "flos": 23731086039360.0, + "grad_norm": 1.8581023148205023, + "learning_rate": 2.2125426652366763e-06, + "loss": 0.7713, + "num_input_tokens_seen": 174274055, + "step": 5346 + }, + { + "epoch": 0.48221129999549084, + "flos": 23916889167840.0, + "grad_norm": 1.8103009399487482, + "learning_rate": 2.211961753835581e-06, + "loss": 0.8531, + "num_input_tokens_seen": 174300430, + "step": 5347 + }, + { + "epoch": 0.48230148351896107, + "flos": 23408140731840.0, + "grad_norm": 1.3525781508763959, + "learning_rate": 2.21138082434874e-06, + "loss": 0.7494, + "num_input_tokens_seen": 174332265, + "step": 5348 + }, + { + "epoch": 0.48239166704243136, + "flos": 20565842340480.0, + "grad_norm": 1.7280596287867789, + "learning_rate": 2.210799876825722e-06, + "loss": 0.7833, + "num_input_tokens_seen": 174356005, + "step": 5349 + }, + { + "epoch": 0.4824818505659016, + "flos": 18707582303520.0, + "grad_norm": 3.712805733470275, + "learning_rate": 2.210218911316096e-06, + "loss": 0.7159, + "num_input_tokens_seen": 174383280, + "step": 5350 + }, + { + "epoch": 0.4825720340893719, + "flos": 25410530862720.0, + "grad_norm": 1.9252867769920092, + "learning_rate": 2.2096379278694336e-06, + "loss": 0.7046, + "num_input_tokens_seen": 174412890, + "step": 5351 + }, + { + "epoch": 0.4826622176128421, + "flos": 23807332684320.0, + "grad_norm": 1.5984503304571582, + "learning_rate": 2.2090569265353074e-06, + "loss": 0.702, + "num_input_tokens_seen": 174441280, + "step": 5352 + }, + { + "epoch": 0.4827524011363124, + "flos": 20747705474400.0, + "grad_norm": 2.1843509350762207, + "learning_rate": 2.2084759073632912e-06, + "loss": 0.7534, + "num_input_tokens_seen": 174465410, + "step": 5353 + }, + { + "epoch": 0.48284258465978264, + "flos": 26067423726720.0, + "grad_norm": 1.7408876231219266, + "learning_rate": 2.2078948704029606e-06, + "loss": 0.7137, + "num_input_tokens_seen": 174495880, + "step": 5354 + }, + { + "epoch": 0.4829327681832529, + "flos": 36157516683840.0, + "grad_norm": 1.947062857735378, + "learning_rate": 2.2073138157038935e-06, + "loss": 0.679, + "num_input_tokens_seen": 174527515, + "step": 5355 + }, + { + "epoch": 0.48302295170672316, + "flos": 17214089287680.0, + "grad_norm": 1.68236877768582, + "learning_rate": 2.2067327433156687e-06, + "loss": 0.7422, + "num_input_tokens_seen": 174555265, + "step": 5356 + }, + { + "epoch": 0.48311313523019345, + "flos": 27302881303200.0, + "grad_norm": 1.5483252922536415, + "learning_rate": 2.2061516532878667e-06, + "loss": 0.7712, + "num_input_tokens_seen": 174585130, + "step": 5357 + }, + { + "epoch": 0.4832033187536637, + "flos": 17541643645440.0, + "grad_norm": 1.780589198632607, + "learning_rate": 2.2055705456700686e-06, + "loss": 0.8028, + "num_input_tokens_seen": 174613510, + "step": 5358 + }, + { + "epoch": 0.48329350227713397, + "flos": 26649779754720.0, + "grad_norm": 2.040878714756153, + "learning_rate": 2.204989420511858e-06, + "loss": 0.6539, + "num_input_tokens_seen": 174643920, + "step": 5359 + }, + { + "epoch": 0.48338368580060426, + "flos": 23517288348000.0, + "grad_norm": 1.50832591432605, + "learning_rate": 2.20440827786282e-06, + "loss": 0.7009, + "num_input_tokens_seen": 174674540, + "step": 5360 + }, + { + "epoch": 0.4834738693240745, + "flos": 23583350478720.0, + "grad_norm": 1.9605674850496317, + "learning_rate": 2.20382711777254e-06, + "loss": 0.76, + "num_input_tokens_seen": 174702390, + "step": 5361 + }, + { + "epoch": 0.4835640528475448, + "flos": 34404947475360.0, + "grad_norm": 1.552144256351917, + "learning_rate": 2.203245940290607e-06, + "loss": 0.7686, + "num_input_tokens_seen": 174733645, + "step": 5362 + }, + { + "epoch": 0.483654236371015, + "flos": 17505508905120.0, + "grad_norm": 2.4143738777302213, + "learning_rate": 2.2026647454666097e-06, + "loss": 0.7483, + "num_input_tokens_seen": 174761570, + "step": 5363 + }, + { + "epoch": 0.4837444198944853, + "flos": 17936561075520.0, + "grad_norm": 2.035331264592675, + "learning_rate": 2.2020835333501384e-06, + "loss": 0.7871, + "num_input_tokens_seen": 174785895, + "step": 5364 + }, + { + "epoch": 0.48383460341795553, + "flos": 23662645044000.0, + "grad_norm": 1.5611431333863373, + "learning_rate": 2.2015023039907863e-06, + "loss": 0.8133, + "num_input_tokens_seen": 174813850, + "step": 5365 + }, + { + "epoch": 0.4839247869414258, + "flos": 29564273287200.0, + "grad_norm": 1.5100517445976163, + "learning_rate": 2.2009210574381464e-06, + "loss": 0.7416, + "num_input_tokens_seen": 174845345, + "step": 5366 + }, + { + "epoch": 0.48401497046489605, + "flos": 26244900828960.0, + "grad_norm": 1.6398286612490294, + "learning_rate": 2.2003397937418134e-06, + "loss": 0.7011, + "num_input_tokens_seen": 174873715, + "step": 5367 + }, + { + "epoch": 0.48410515398836634, + "flos": 62255270934720.0, + "grad_norm": 0.6707057315195899, + "learning_rate": 2.1997585129513852e-06, + "loss": 0.6021, + "num_input_tokens_seen": 174967045, + "step": 5368 + }, + { + "epoch": 0.4841953375118366, + "flos": 24792188773440.0, + "grad_norm": 1.6411984122798342, + "learning_rate": 2.1991772151164595e-06, + "loss": 0.7815, + "num_input_tokens_seen": 174997705, + "step": 5369 + }, + { + "epoch": 0.48428552103530687, + "flos": 22891326457920.0, + "grad_norm": 1.431854356994085, + "learning_rate": 2.1985959002866346e-06, + "loss": 0.7997, + "num_input_tokens_seen": 175029565, + "step": 5370 + }, + { + "epoch": 0.4843757045587771, + "flos": 16921145710080.0, + "grad_norm": 2.841330033067529, + "learning_rate": 2.198014568511513e-06, + "loss": 0.8196, + "num_input_tokens_seen": 175051825, + "step": 5371 + }, + { + "epoch": 0.4844658880822474, + "flos": 67027504127520.0, + "grad_norm": 0.6633462296682746, + "learning_rate": 2.1974332198406965e-06, + "loss": 0.5747, + "num_input_tokens_seen": 175143690, + "step": 5372 + }, + { + "epoch": 0.4845560716057176, + "flos": 26610188226720.0, + "grad_norm": 1.6691902922800745, + "learning_rate": 2.196851854323789e-06, + "loss": 0.7384, + "num_input_tokens_seen": 175173150, + "step": 5373 + }, + { + "epoch": 0.4846462551291879, + "flos": 37907409669600.0, + "grad_norm": 1.8603488071050789, + "learning_rate": 2.196270472010396e-06, + "loss": 0.7385, + "num_input_tokens_seen": 175203325, + "step": 5374 + }, + { + "epoch": 0.48473643865265814, + "flos": 23333901253920.0, + "grad_norm": 1.7248001425139698, + "learning_rate": 2.195689072950124e-06, + "loss": 0.7501, + "num_input_tokens_seen": 175232935, + "step": 5375 + }, + { + "epoch": 0.48482662217612843, + "flos": 68779775977920.0, + "grad_norm": 0.6099664556551121, + "learning_rate": 2.195107657192581e-06, + "loss": 0.5621, + "num_input_tokens_seen": 175334085, + "step": 5376 + }, + { + "epoch": 0.48491680569959866, + "flos": 68732490309600.0, + "grad_norm": 0.5865889170677573, + "learning_rate": 2.194526224787378e-06, + "loss": 0.5164, + "num_input_tokens_seen": 175429845, + "step": 5377 + }, + { + "epoch": 0.48500698922306895, + "flos": 50730876420480.0, + "grad_norm": 1.6500356439150596, + "learning_rate": 2.1939447757841236e-06, + "loss": 0.7228, + "num_input_tokens_seen": 175462835, + "step": 5378 + }, + { + "epoch": 0.4850971727465392, + "flos": 29200100982240.0, + "grad_norm": 1.864525465537922, + "learning_rate": 2.193363310232432e-06, + "loss": 0.8148, + "num_input_tokens_seen": 175489470, + "step": 5379 + }, + { + "epoch": 0.4851873562700095, + "flos": 19431280693440.0, + "grad_norm": 1.3560435810428215, + "learning_rate": 2.192781828181917e-06, + "loss": 0.854, + "num_input_tokens_seen": 175518285, + "step": 5380 + }, + { + "epoch": 0.4852775397934797, + "flos": 27814603320000.0, + "grad_norm": 1.3644274508520455, + "learning_rate": 2.192200329682193e-06, + "loss": 0.777, + "num_input_tokens_seen": 175550090, + "step": 5381 + }, + { + "epoch": 0.48536772331695, + "flos": 21547501830240.0, + "grad_norm": 1.3936052426829744, + "learning_rate": 2.1916188147828767e-06, + "loss": 0.7517, + "num_input_tokens_seen": 175579580, + "step": 5382 + }, + { + "epoch": 0.48545790684042023, + "flos": 19727160682080.0, + "grad_norm": 1.9104884820195067, + "learning_rate": 2.191037283533587e-06, + "loss": 0.7696, + "num_input_tokens_seen": 175607695, + "step": 5383 + }, + { + "epoch": 0.4855480903638905, + "flos": 20383124302080.0, + "grad_norm": 2.4433458115942983, + "learning_rate": 2.1904557359839428e-06, + "loss": 0.7515, + "num_input_tokens_seen": 175633245, + "step": 5384 + }, + { + "epoch": 0.4856382738873608, + "flos": 19071903287520.0, + "grad_norm": 1.8470660632039124, + "learning_rate": 2.189874172183565e-06, + "loss": 0.7317, + "num_input_tokens_seen": 175660630, + "step": 5385 + }, + { + "epoch": 0.48572845741083104, + "flos": 23335313704800.0, + "grad_norm": 1.6293248066489148, + "learning_rate": 2.1892925921820763e-06, + "loss": 0.7255, + "num_input_tokens_seen": 175690475, + "step": 5386 + }, + { + "epoch": 0.4858186409343013, + "flos": 14736855275520.0, + "grad_norm": 1.9768200129852525, + "learning_rate": 2.1887109960290994e-06, + "loss": 0.7495, + "num_input_tokens_seen": 175717355, + "step": 5387 + }, + { + "epoch": 0.48590882445777156, + "flos": 18671150205120.0, + "grad_norm": 1.6717971355809356, + "learning_rate": 2.18812938377426e-06, + "loss": 0.8503, + "num_input_tokens_seen": 175744810, + "step": 5388 + }, + { + "epoch": 0.48599900798124185, + "flos": 23516730801600.0, + "grad_norm": 1.6346262663914886, + "learning_rate": 2.187547755467184e-06, + "loss": 0.7928, + "num_input_tokens_seen": 175772330, + "step": 5389 + }, + { + "epoch": 0.4860891915047121, + "flos": 34810829984640.0, + "grad_norm": 1.537348300924388, + "learning_rate": 2.1869661111574994e-06, + "loss": 0.6467, + "num_input_tokens_seen": 175802470, + "step": 5390 + }, + { + "epoch": 0.48617937502818237, + "flos": 25848236420160.0, + "grad_norm": 2.0312116179141126, + "learning_rate": 2.1863844508948353e-06, + "loss": 0.8099, + "num_input_tokens_seen": 175830610, + "step": 5391 + }, + { + "epoch": 0.4862695585516526, + "flos": 21869629403040.0, + "grad_norm": 1.699291920456743, + "learning_rate": 2.185802774728823e-06, + "loss": 0.769, + "num_input_tokens_seen": 175857205, + "step": 5392 + }, + { + "epoch": 0.4863597420751229, + "flos": 20528704016640.0, + "grad_norm": 1.5302213210802702, + "learning_rate": 2.1852210827090927e-06, + "loss": 0.7091, + "num_input_tokens_seen": 175885975, + "step": 5393 + }, + { + "epoch": 0.4864499255985931, + "flos": 24427830619680.0, + "grad_norm": 1.6774509887812472, + "learning_rate": 2.184639374885278e-06, + "loss": 0.6977, + "num_input_tokens_seen": 175914765, + "step": 5394 + }, + { + "epoch": 0.4865401091220634, + "flos": 20274385553280.0, + "grad_norm": 2.261310336276444, + "learning_rate": 2.184057651307014e-06, + "loss": 0.8461, + "num_input_tokens_seen": 175942255, + "step": 5395 + }, + { + "epoch": 0.48663029264553365, + "flos": 20820086464320.0, + "grad_norm": 3.3549930825993592, + "learning_rate": 2.183475912023937e-06, + "loss": 0.8217, + "num_input_tokens_seen": 175964860, + "step": 5396 + }, + { + "epoch": 0.48672047616900393, + "flos": 19502918288160.0, + "grad_norm": 1.7385179679936071, + "learning_rate": 2.1828941570856826e-06, + "loss": 0.7681, + "num_input_tokens_seen": 175992110, + "step": 5397 + }, + { + "epoch": 0.48681065969247417, + "flos": 68138202788640.0, + "grad_norm": 0.6352070608613191, + "learning_rate": 2.1823123865418903e-06, + "loss": 0.5718, + "num_input_tokens_seen": 176083760, + "step": 5398 + }, + { + "epoch": 0.48690084321594446, + "flos": 24499988591040.0, + "grad_norm": 2.4516362817965955, + "learning_rate": 2.1817306004422e-06, + "loss": 0.7314, + "num_input_tokens_seen": 176113315, + "step": 5399 + }, + { + "epoch": 0.4869910267394147, + "flos": 29923055976960.0, + "grad_norm": 1.7152949893034697, + "learning_rate": 2.1811487988362527e-06, + "loss": 0.7767, + "num_input_tokens_seen": 176141810, + "step": 5400 + }, + { + "epoch": 0.487081210262885, + "flos": 24645531135840.0, + "grad_norm": 2.1430971669692287, + "learning_rate": 2.1805669817736917e-06, + "loss": 0.8256, + "num_input_tokens_seen": 176166965, + "step": 5401 + }, + { + "epoch": 0.4871713937863552, + "flos": 23881126125120.0, + "grad_norm": 7.015905293937901, + "learning_rate": 2.17998514930416e-06, + "loss": 0.5797, + "num_input_tokens_seen": 176197245, + "step": 5402 + }, + { + "epoch": 0.4872615773098255, + "flos": 27087410972640.0, + "grad_norm": 2.079286867182496, + "learning_rate": 2.1794033014773025e-06, + "loss": 0.7513, + "num_input_tokens_seen": 176224290, + "step": 5403 + }, + { + "epoch": 0.48735176083329573, + "flos": 21622410363840.0, + "grad_norm": 1.7152008866524895, + "learning_rate": 2.178821438342766e-06, + "loss": 0.8215, + "num_input_tokens_seen": 176252075, + "step": 5404 + }, + { + "epoch": 0.487441944356766, + "flos": 26794393055520.0, + "grad_norm": 1.9693755109829303, + "learning_rate": 2.1782395599501996e-06, + "loss": 0.7491, + "num_input_tokens_seen": 176283200, + "step": 5405 + }, + { + "epoch": 0.48753212788023625, + "flos": 70364909652960.0, + "grad_norm": 0.648771108272896, + "learning_rate": 2.1776576663492498e-06, + "loss": 0.5841, + "num_input_tokens_seen": 176374880, + "step": 5406 + }, + { + "epoch": 0.48762231140370654, + "flos": 24057079267200.0, + "grad_norm": 2.2551417410682526, + "learning_rate": 2.177075757589569e-06, + "loss": 0.8423, + "num_input_tokens_seen": 176402115, + "step": 5407 + }, + { + "epoch": 0.48771249492717683, + "flos": 24680104746240.0, + "grad_norm": 1.5891822732293444, + "learning_rate": 2.176493833720808e-06, + "loss": 0.6782, + "num_input_tokens_seen": 176432690, + "step": 5408 + }, + { + "epoch": 0.48780267845064706, + "flos": 25083162353760.0, + "grad_norm": 3.3599066734976817, + "learning_rate": 2.1759118947926195e-06, + "loss": 0.7227, + "num_input_tokens_seen": 176461820, + "step": 5409 + }, + { + "epoch": 0.48789286197411735, + "flos": 32295788592960.0, + "grad_norm": 1.4584653612953125, + "learning_rate": 2.1753299408546587e-06, + "loss": 0.763, + "num_input_tokens_seen": 176493555, + "step": 5410 + }, + { + "epoch": 0.4879830454975876, + "flos": 33750136117920.0, + "grad_norm": 1.3788534267779267, + "learning_rate": 2.1747479719565803e-06, + "loss": 0.7591, + "num_input_tokens_seen": 176527020, + "step": 5411 + }, + { + "epoch": 0.4880732290210579, + "flos": 26176013796480.0, + "grad_norm": 1.3896701406257184, + "learning_rate": 2.174165988148042e-06, + "loss": 0.8171, + "num_input_tokens_seen": 176558135, + "step": 5412 + }, + { + "epoch": 0.4881634125445281, + "flos": 22565110211520.0, + "grad_norm": 1.5817847542584045, + "learning_rate": 2.1735839894787003e-06, + "loss": 0.7253, + "num_input_tokens_seen": 176584760, + "step": 5413 + }, + { + "epoch": 0.4882535960679984, + "flos": 64627520343840.0, + "grad_norm": 0.8039442166481799, + "learning_rate": 2.1730019759982163e-06, + "loss": 0.6286, + "num_input_tokens_seen": 176674455, + "step": 5414 + }, + { + "epoch": 0.48834377959146863, + "flos": 20565953849760.0, + "grad_norm": 1.9931006638771802, + "learning_rate": 2.172419947756249e-06, + "loss": 0.8084, + "num_input_tokens_seen": 176702365, + "step": 5415 + }, + { + "epoch": 0.4884339631149389, + "flos": 19690840092960.0, + "grad_norm": 2.045471175925725, + "learning_rate": 2.171837904802461e-06, + "loss": 0.8012, + "num_input_tokens_seen": 176729615, + "step": 5416 + }, + { + "epoch": 0.48852414663840915, + "flos": 27445821964800.0, + "grad_norm": 1.4727966889217274, + "learning_rate": 2.171255847186516e-06, + "loss": 0.7538, + "num_input_tokens_seen": 176760685, + "step": 5417 + }, + { + "epoch": 0.48861433016187944, + "flos": 18452780633280.0, + "grad_norm": 2.591569466521144, + "learning_rate": 2.1706737749580783e-06, + "loss": 0.7416, + "num_input_tokens_seen": 176786635, + "step": 5418 + }, + { + "epoch": 0.4887045136853497, + "flos": 22751545225920.0, + "grad_norm": 1.3786911837222822, + "learning_rate": 2.1700916881668127e-06, + "loss": 0.8283, + "num_input_tokens_seen": 176816840, + "step": 5419 + }, + { + "epoch": 0.48879469720881996, + "flos": 20601568213440.0, + "grad_norm": 1.733018155358907, + "learning_rate": 2.1695095868623862e-06, + "loss": 0.8457, + "num_input_tokens_seen": 176845070, + "step": 5420 + }, + { + "epoch": 0.4888848807322902, + "flos": 34992878967360.0, + "grad_norm": 1.6674349285737755, + "learning_rate": 2.168927471094467e-06, + "loss": 0.6877, + "num_input_tokens_seen": 176877590, + "step": 5421 + }, + { + "epoch": 0.4889750642557605, + "flos": 21184667636640.0, + "grad_norm": 2.3035457563184276, + "learning_rate": 2.168345340912725e-06, + "loss": 0.7759, + "num_input_tokens_seen": 176904090, + "step": 5422 + }, + { + "epoch": 0.4890652477792307, + "flos": 23806254761280.0, + "grad_norm": 1.5391064942554105, + "learning_rate": 2.1677631963668298e-06, + "loss": 0.7727, + "num_input_tokens_seen": 176936260, + "step": 5423 + }, + { + "epoch": 0.489155431302701, + "flos": 20492569276320.0, + "grad_norm": 3.273372676817513, + "learning_rate": 2.167181037506453e-06, + "loss": 0.7151, + "num_input_tokens_seen": 176963040, + "step": 5424 + }, + { + "epoch": 0.48924561482617124, + "flos": 22788163173120.0, + "grad_norm": 1.67437770727949, + "learning_rate": 2.1665988643812693e-06, + "loss": 0.6918, + "num_input_tokens_seen": 176990880, + "step": 5425 + }, + { + "epoch": 0.4893357983496415, + "flos": 26503605324000.0, + "grad_norm": 3.7608372813217525, + "learning_rate": 2.166016677040951e-06, + "loss": 0.7531, + "num_input_tokens_seen": 177019820, + "step": 5426 + }, + { + "epoch": 0.48942598187311176, + "flos": 27123285524640.0, + "grad_norm": 1.9135032615344758, + "learning_rate": 2.165434475535175e-06, + "loss": 0.7428, + "num_input_tokens_seen": 177047510, + "step": 5427 + }, + { + "epoch": 0.48951616539658205, + "flos": 22489383943200.0, + "grad_norm": 1.69906601493434, + "learning_rate": 2.1648522599136173e-06, + "loss": 0.7948, + "num_input_tokens_seen": 177077170, + "step": 5428 + }, + { + "epoch": 0.4896063489200523, + "flos": 25992180665280.0, + "grad_norm": 1.6150454280357383, + "learning_rate": 2.164270030225956e-06, + "loss": 0.7271, + "num_input_tokens_seen": 177107915, + "step": 5429 + }, + { + "epoch": 0.48969653244352257, + "flos": 22313170612800.0, + "grad_norm": 2.558827629756673, + "learning_rate": 2.16368778652187e-06, + "loss": 0.6675, + "num_input_tokens_seen": 177137215, + "step": 5430 + }, + { + "epoch": 0.4897867159669928, + "flos": 24828100495200.0, + "grad_norm": 1.7671818578817218, + "learning_rate": 2.163105528851039e-06, + "loss": 0.7929, + "num_input_tokens_seen": 177166590, + "step": 5431 + }, + { + "epoch": 0.4898768994904631, + "flos": 24790962171360.0, + "grad_norm": 1.656228407233214, + "learning_rate": 2.1625232572631448e-06, + "loss": 0.7603, + "num_input_tokens_seen": 177197730, + "step": 5432 + }, + { + "epoch": 0.4899670830139334, + "flos": 70840496929440.0, + "grad_norm": 0.7307862188607892, + "learning_rate": 2.161940971807871e-06, + "loss": 0.5901, + "num_input_tokens_seen": 177296250, + "step": 5433 + }, + { + "epoch": 0.4900572665374036, + "flos": 25301903623200.0, + "grad_norm": 2.318333870700067, + "learning_rate": 2.1613586725348994e-06, + "loss": 0.7398, + "num_input_tokens_seen": 177323770, + "step": 5434 + }, + { + "epoch": 0.4901474500608739, + "flos": 23581752179040.0, + "grad_norm": 1.5382438844992894, + "learning_rate": 2.1607763594939176e-06, + "loss": 0.8317, + "num_input_tokens_seen": 177351335, + "step": 5435 + }, + { + "epoch": 0.49023763358434413, + "flos": 31605214192800.0, + "grad_norm": 1.9271477422174623, + "learning_rate": 2.1601940327346093e-06, + "loss": 0.6682, + "num_input_tokens_seen": 177383955, + "step": 5436 + }, + { + "epoch": 0.4903278171078144, + "flos": 26032143890880.0, + "grad_norm": 1.4876496984984418, + "learning_rate": 2.159611692306663e-06, + "loss": 0.8286, + "num_input_tokens_seen": 177415750, + "step": 5437 + }, + { + "epoch": 0.49041800063128466, + "flos": 29527804019040.0, + "grad_norm": 1.9196230092083966, + "learning_rate": 2.1590293382597667e-06, + "loss": 0.7209, + "num_input_tokens_seen": 177445650, + "step": 5438 + }, + { + "epoch": 0.49050818415475494, + "flos": 26322336906240.0, + "grad_norm": 1.650461048677767, + "learning_rate": 2.1584469706436102e-06, + "loss": 0.6956, + "num_input_tokens_seen": 177475670, + "step": 5439 + }, + { + "epoch": 0.4905983676782252, + "flos": 23589371979840.0, + "grad_norm": 2.0239167695253233, + "learning_rate": 2.1578645895078855e-06, + "loss": 0.6738, + "num_input_tokens_seen": 177505220, + "step": 5440 + }, + { + "epoch": 0.49068855120169547, + "flos": 26613421995840.0, + "grad_norm": 1.7054249344150092, + "learning_rate": 2.157282194902283e-06, + "loss": 0.8126, + "num_input_tokens_seen": 177532000, + "step": 5441 + }, + { + "epoch": 0.4907787347251657, + "flos": 12550780692480.0, + "grad_norm": 1.852824943656396, + "learning_rate": 2.1566997868764965e-06, + "loss": 0.6984, + "num_input_tokens_seen": 177559220, + "step": 5442 + }, + { + "epoch": 0.490868918248636, + "flos": 18962272464480.0, + "grad_norm": 3.2538413308068503, + "learning_rate": 2.15611736548022e-06, + "loss": 0.7149, + "num_input_tokens_seen": 177584350, + "step": 5443 + }, + { + "epoch": 0.4909591017721062, + "flos": 20310222935520.0, + "grad_norm": 2.01488286039948, + "learning_rate": 2.155534930763149e-06, + "loss": 0.7608, + "num_input_tokens_seen": 177611760, + "step": 5444 + }, + { + "epoch": 0.4910492852955765, + "flos": 19836977353920.0, + "grad_norm": 1.711963781232165, + "learning_rate": 2.1549524827749804e-06, + "loss": 0.6363, + "num_input_tokens_seen": 177638210, + "step": 5445 + }, + { + "epoch": 0.49113946881904674, + "flos": 41185369281600.0, + "grad_norm": 2.747171681497105, + "learning_rate": 2.1543700215654115e-06, + "loss": 0.6051, + "num_input_tokens_seen": 177671850, + "step": 5446 + }, + { + "epoch": 0.49122965234251703, + "flos": 22205361108000.0, + "grad_norm": 1.6926226393698185, + "learning_rate": 2.153787547184141e-06, + "loss": 0.7454, + "num_input_tokens_seen": 177700650, + "step": 5447 + }, + { + "epoch": 0.49131983586598726, + "flos": 36085916258880.0, + "grad_norm": 1.984678530569926, + "learning_rate": 2.1532050596808695e-06, + "loss": 0.683, + "num_input_tokens_seen": 177732320, + "step": 5448 + }, + { + "epoch": 0.49141001938945755, + "flos": 24937694148480.0, + "grad_norm": 1.7783754142256125, + "learning_rate": 2.152622559105297e-06, + "loss": 0.8013, + "num_input_tokens_seen": 177761725, + "step": 5449 + }, + { + "epoch": 0.4915002029129278, + "flos": 22824037725120.0, + "grad_norm": 2.118940127683997, + "learning_rate": 2.152040045507126e-06, + "loss": 0.7966, + "num_input_tokens_seen": 177788595, + "step": 5450 + }, + { + "epoch": 0.4915903864363981, + "flos": 21257866361280.0, + "grad_norm": 1.8334005155680604, + "learning_rate": 2.1514575189360607e-06, + "loss": 0.7677, + "num_input_tokens_seen": 177816120, + "step": 5451 + }, + { + "epoch": 0.4916805699598683, + "flos": 17614396332960.0, + "grad_norm": 1.6095881644991226, + "learning_rate": 2.1508749794418043e-06, + "loss": 0.7113, + "num_input_tokens_seen": 177844445, + "step": 5452 + }, + { + "epoch": 0.4917707534833386, + "flos": 26282076322560.0, + "grad_norm": 1.8201722185166476, + "learning_rate": 2.1502924270740626e-06, + "loss": 0.725, + "num_input_tokens_seen": 177873370, + "step": 5453 + }, + { + "epoch": 0.49186093700680883, + "flos": 20450413034880.0, + "grad_norm": 2.2866473166904715, + "learning_rate": 2.1497098618825427e-06, + "loss": 0.7617, + "num_input_tokens_seen": 177900795, + "step": 5454 + }, + { + "epoch": 0.4919511205302791, + "flos": 22350903652800.0, + "grad_norm": 3.252700124179811, + "learning_rate": 2.1491272839169516e-06, + "loss": 0.7945, + "num_input_tokens_seen": 177928945, + "step": 5455 + }, + { + "epoch": 0.4920413040537494, + "flos": 21104369487840.0, + "grad_norm": 1.6585972545584957, + "learning_rate": 2.1485446932269986e-06, + "loss": 0.8646, + "num_input_tokens_seen": 177957085, + "step": 5456 + }, + { + "epoch": 0.49213148757721964, + "flos": 22675410090240.0, + "grad_norm": 1.5839792573707403, + "learning_rate": 2.147962089862393e-06, + "loss": 0.7909, + "num_input_tokens_seen": 177987380, + "step": 5457 + }, + { + "epoch": 0.49222167110068993, + "flos": 23370110333760.0, + "grad_norm": 2.3935780954014017, + "learning_rate": 2.1473794738728462e-06, + "loss": 0.6628, + "num_input_tokens_seen": 178015115, + "step": 5458 + }, + { + "epoch": 0.49231185462416016, + "flos": 18452631954240.0, + "grad_norm": 1.7796047040721537, + "learning_rate": 2.14679684530807e-06, + "loss": 0.7887, + "num_input_tokens_seen": 178040150, + "step": 5459 + }, + { + "epoch": 0.49240203814763045, + "flos": 25262312095200.0, + "grad_norm": 1.8514144130027497, + "learning_rate": 2.1462142042177774e-06, + "loss": 0.7364, + "num_input_tokens_seen": 178070100, + "step": 5460 + }, + { + "epoch": 0.4924922216711007, + "flos": 22423842189120.0, + "grad_norm": 1.6540217255222966, + "learning_rate": 2.145631550651683e-06, + "loss": 0.7467, + "num_input_tokens_seen": 178098305, + "step": 5461 + }, + { + "epoch": 0.49258240519457097, + "flos": 21112100797920.0, + "grad_norm": 1.564449924651455, + "learning_rate": 2.1450488846595016e-06, + "loss": 0.8562, + "num_input_tokens_seen": 178126050, + "step": 5462 + }, + { + "epoch": 0.4926725887180412, + "flos": 17100852997920.0, + "grad_norm": 2.035258753679121, + "learning_rate": 2.14446620629095e-06, + "loss": 0.732, + "num_input_tokens_seen": 178152790, + "step": 5463 + }, + { + "epoch": 0.4927627722415115, + "flos": 24388870977600.0, + "grad_norm": 1.6690355235228302, + "learning_rate": 2.1438835155957445e-06, + "loss": 0.8187, + "num_input_tokens_seen": 178182430, + "step": 5464 + }, + { + "epoch": 0.4928529557649817, + "flos": 23734505657280.0, + "grad_norm": 2.5218160516025185, + "learning_rate": 2.143300812623604e-06, + "loss": 0.7945, + "num_input_tokens_seen": 178210295, + "step": 5465 + }, + { + "epoch": 0.492943139288452, + "flos": 21513114068640.0, + "grad_norm": 1.7133011885916585, + "learning_rate": 2.1427180974242485e-06, + "loss": 0.7373, + "num_input_tokens_seen": 178237640, + "step": 5466 + }, + { + "epoch": 0.49303332281192225, + "flos": 21725759497440.0, + "grad_norm": 1.847976818059041, + "learning_rate": 2.142135370047398e-06, + "loss": 0.7811, + "num_input_tokens_seen": 178264835, + "step": 5467 + }, + { + "epoch": 0.49312350633539254, + "flos": 17577889895040.0, + "grad_norm": 1.9161460279910656, + "learning_rate": 2.1415526305427735e-06, + "loss": 0.7863, + "num_input_tokens_seen": 178289040, + "step": 5468 + }, + { + "epoch": 0.49321368985886277, + "flos": 30182912734560.0, + "grad_norm": 1.4179610154962954, + "learning_rate": 2.140969878960098e-06, + "loss": 0.6781, + "num_input_tokens_seen": 178319185, + "step": 5469 + }, + { + "epoch": 0.49330387338233306, + "flos": 26318099553600.0, + "grad_norm": 1.5589505124315433, + "learning_rate": 2.1403871153490956e-06, + "loss": 0.8103, + "num_input_tokens_seen": 178348615, + "step": 5470 + }, + { + "epoch": 0.4933940569058033, + "flos": 20527031377440.0, + "grad_norm": 1.5648212737870368, + "learning_rate": 2.13980433975949e-06, + "loss": 0.8747, + "num_input_tokens_seen": 178373110, + "step": 5471 + }, + { + "epoch": 0.4934842404292736, + "flos": 24935761320960.0, + "grad_norm": 1.937345129406839, + "learning_rate": 2.1392215522410076e-06, + "loss": 0.7979, + "num_input_tokens_seen": 178404085, + "step": 5472 + }, + { + "epoch": 0.4935744239527438, + "flos": 20165052088320.0, + "grad_norm": 2.2378625927589533, + "learning_rate": 2.1386387528433743e-06, + "loss": 0.8671, + "num_input_tokens_seen": 178430335, + "step": 5473 + }, + { + "epoch": 0.4936646074762141, + "flos": 20383310150880.0, + "grad_norm": 1.8827787530810913, + "learning_rate": 2.1380559416163186e-06, + "loss": 0.781, + "num_input_tokens_seen": 178456965, + "step": 5474 + }, + { + "epoch": 0.49375479099968433, + "flos": 17431864143360.0, + "grad_norm": 1.7808261835611945, + "learning_rate": 2.1374731186095685e-06, + "loss": 0.8, + "num_input_tokens_seen": 178484990, + "step": 5475 + }, + { + "epoch": 0.4938449745231546, + "flos": 21360397760160.0, + "grad_norm": 1.585232507517435, + "learning_rate": 2.136890283872854e-06, + "loss": 0.7084, + "num_input_tokens_seen": 178512085, + "step": 5476 + }, + { + "epoch": 0.49393515804662486, + "flos": 24567277323840.0, + "grad_norm": 2.5361922253100504, + "learning_rate": 2.136307437455906e-06, + "loss": 0.7359, + "num_input_tokens_seen": 178540495, + "step": 5477 + }, + { + "epoch": 0.49402534157009514, + "flos": 25190228463360.0, + "grad_norm": 1.4893304125490523, + "learning_rate": 2.135724579408456e-06, + "loss": 0.7019, + "num_input_tokens_seen": 178570115, + "step": 5478 + }, + { + "epoch": 0.49411552509356543, + "flos": 18926286403200.0, + "grad_norm": 1.8317612569032762, + "learning_rate": 2.1351417097802356e-06, + "loss": 0.8474, + "num_input_tokens_seen": 178598075, + "step": 5479 + }, + { + "epoch": 0.49420570861703567, + "flos": 23698445256480.0, + "grad_norm": 2.0307710523085882, + "learning_rate": 2.1345588286209798e-06, + "loss": 0.6382, + "num_input_tokens_seen": 178624545, + "step": 5480 + }, + { + "epoch": 0.49429589214050595, + "flos": 36303839793600.0, + "grad_norm": 2.009850443689718, + "learning_rate": 2.1339759359804227e-06, + "loss": 0.6036, + "num_input_tokens_seen": 178653425, + "step": 5481 + }, + { + "epoch": 0.4943860756639762, + "flos": 34407103321440.0, + "grad_norm": 1.561851519484101, + "learning_rate": 2.1333930319082997e-06, + "loss": 0.7277, + "num_input_tokens_seen": 178683720, + "step": 5482 + }, + { + "epoch": 0.4944762591874465, + "flos": 25845485857920.0, + "grad_norm": 1.340458987256387, + "learning_rate": 2.132810116454348e-06, + "loss": 0.7574, + "num_input_tokens_seen": 178713595, + "step": 5483 + }, + { + "epoch": 0.4945664427109167, + "flos": 25665518381760.0, + "grad_norm": 2.251844391923161, + "learning_rate": 2.132227189668305e-06, + "loss": 0.7502, + "num_input_tokens_seen": 178742570, + "step": 5484 + }, + { + "epoch": 0.494656626234387, + "flos": 32402891872320.0, + "grad_norm": 1.7276229708635986, + "learning_rate": 2.1316442515999096e-06, + "loss": 0.6984, + "num_input_tokens_seen": 178772505, + "step": 5485 + }, + { + "epoch": 0.49474680975785723, + "flos": 28252346047200.0, + "grad_norm": 1.4577901464385972, + "learning_rate": 2.1310613022989e-06, + "loss": 0.7848, + "num_input_tokens_seen": 178803605, + "step": 5486 + }, + { + "epoch": 0.4948369932813275, + "flos": 33353062841760.0, + "grad_norm": 2.079146858177034, + "learning_rate": 2.130478341815017e-06, + "loss": 0.6043, + "num_input_tokens_seen": 178834455, + "step": 5487 + }, + { + "epoch": 0.49492717680479775, + "flos": 25153387497600.0, + "grad_norm": 1.6600134215903888, + "learning_rate": 2.1298953701980033e-06, + "loss": 0.6478, + "num_input_tokens_seen": 178864715, + "step": 5488 + }, + { + "epoch": 0.49501736032826804, + "flos": 38744455858560.0, + "grad_norm": 2.1008313123044386, + "learning_rate": 2.1293123874976003e-06, + "loss": 0.7613, + "num_input_tokens_seen": 178892105, + "step": 5489 + }, + { + "epoch": 0.4951075438517383, + "flos": 24682037573760.0, + "grad_norm": 1.7159711394784083, + "learning_rate": 2.1287293937635513e-06, + "loss": 0.7018, + "num_input_tokens_seen": 178920435, + "step": 5490 + }, + { + "epoch": 0.49519772737520856, + "flos": 26468288318400.0, + "grad_norm": 1.7403434265424655, + "learning_rate": 2.1281463890456005e-06, + "loss": 0.8052, + "num_input_tokens_seen": 178949765, + "step": 5491 + }, + { + "epoch": 0.4952879108986788, + "flos": 19540651328160.0, + "grad_norm": 1.7586366882911912, + "learning_rate": 2.127563373393493e-06, + "loss": 0.7474, + "num_input_tokens_seen": 178976775, + "step": 5492 + }, + { + "epoch": 0.4953780944221491, + "flos": 24245075411520.0, + "grad_norm": 2.1718287840392474, + "learning_rate": 2.1269803468569756e-06, + "loss": 0.7464, + "num_input_tokens_seen": 179006650, + "step": 5493 + }, + { + "epoch": 0.4954682779456193, + "flos": 47704744897920.0, + "grad_norm": 1.6948619191013603, + "learning_rate": 2.126397309485794e-06, + "loss": 0.7899, + "num_input_tokens_seen": 179039490, + "step": 5494 + }, + { + "epoch": 0.4955584614690896, + "flos": 30330350937120.0, + "grad_norm": 1.667869118209608, + "learning_rate": 2.1258142613296983e-06, + "loss": 0.7146, + "num_input_tokens_seen": 179072545, + "step": 5495 + }, + { + "epoch": 0.49564864499255984, + "flos": 33097406267040.0, + "grad_norm": 1.4113205517550622, + "learning_rate": 2.125231202438435e-06, + "loss": 0.7278, + "num_input_tokens_seen": 179105895, + "step": 5496 + }, + { + "epoch": 0.49573882851603013, + "flos": 20346097487520.0, + "grad_norm": 1.4517620627018877, + "learning_rate": 2.1246481328617553e-06, + "loss": 0.7882, + "num_input_tokens_seen": 179135065, + "step": 5497 + }, + { + "epoch": 0.49582901203950036, + "flos": 22932999492480.0, + "grad_norm": 2.009554571074694, + "learning_rate": 2.1240650526494096e-06, + "loss": 0.7353, + "num_input_tokens_seen": 179160930, + "step": 5498 + }, + { + "epoch": 0.49591919556297065, + "flos": 24755050449600.0, + "grad_norm": 1.9573495571744326, + "learning_rate": 2.1234819618511493e-06, + "loss": 0.7219, + "num_input_tokens_seen": 179189305, + "step": 5499 + }, + { + "epoch": 0.4960093790864409, + "flos": 67518894285600.0, + "grad_norm": 0.681736477558639, + "learning_rate": 2.122898860516728e-06, + "loss": 0.6178, + "num_input_tokens_seen": 179276915, + "step": 5500 + }, + { + "epoch": 0.49609956260991117, + "flos": 24208829161920.0, + "grad_norm": 1.5290527148842905, + "learning_rate": 2.1223157486958976e-06, + "loss": 0.7384, + "num_input_tokens_seen": 179306145, + "step": 5501 + }, + { + "epoch": 0.4961897461333814, + "flos": 26794616074080.0, + "grad_norm": 1.514484408933398, + "learning_rate": 2.1217326264384127e-06, + "loss": 0.6436, + "num_input_tokens_seen": 179336470, + "step": 5502 + }, + { + "epoch": 0.4962799296568517, + "flos": 14546889133920.0, + "grad_norm": 2.1032180788208477, + "learning_rate": 2.1211494937940296e-06, + "loss": 0.7762, + "num_input_tokens_seen": 179357275, + "step": 5503 + }, + { + "epoch": 0.496370113180322, + "flos": 27159160076640.0, + "grad_norm": 1.3982821756552313, + "learning_rate": 2.1205663508125034e-06, + "loss": 0.7686, + "num_input_tokens_seen": 179386315, + "step": 5504 + }, + { + "epoch": 0.4964602967037922, + "flos": 25555404351840.0, + "grad_norm": 1.5173010456564064, + "learning_rate": 2.1199831975435914e-06, + "loss": 0.7158, + "num_input_tokens_seen": 179415145, + "step": 5505 + }, + { + "epoch": 0.4965504802272625, + "flos": 26174675685120.0, + "grad_norm": 1.5839773758852378, + "learning_rate": 2.1194000340370517e-06, + "loss": 0.8029, + "num_input_tokens_seen": 179447630, + "step": 5506 + }, + { + "epoch": 0.49664066375073274, + "flos": 24464448566880.0, + "grad_norm": 2.238891837836196, + "learning_rate": 2.1188168603426423e-06, + "loss": 0.6927, + "num_input_tokens_seen": 179475780, + "step": 5507 + }, + { + "epoch": 0.496730847274203, + "flos": 19909246834560.0, + "grad_norm": 3.065196971195979, + "learning_rate": 2.118233676510123e-06, + "loss": 0.7178, + "num_input_tokens_seen": 179503005, + "step": 5508 + }, + { + "epoch": 0.49682103079767326, + "flos": 24208829161920.0, + "grad_norm": 1.7209685832053654, + "learning_rate": 2.117650482589255e-06, + "loss": 0.7636, + "num_input_tokens_seen": 179531220, + "step": 5509 + }, + { + "epoch": 0.49691121432114355, + "flos": 13315371552000.0, + "grad_norm": 1.9360277673645256, + "learning_rate": 2.1170672786297988e-06, + "loss": 0.7929, + "num_input_tokens_seen": 179556170, + "step": 5510 + }, + { + "epoch": 0.4970013978446138, + "flos": 23731197548640.0, + "grad_norm": 1.5231978007791576, + "learning_rate": 2.1164840646815174e-06, + "loss": 0.7225, + "num_input_tokens_seen": 179586540, + "step": 5511 + }, + { + "epoch": 0.49709158136808407, + "flos": 21581480724480.0, + "grad_norm": 1.8209843521283786, + "learning_rate": 2.1159008407941726e-06, + "loss": 0.8366, + "num_input_tokens_seen": 179614225, + "step": 5512 + }, + { + "epoch": 0.4971817648915543, + "flos": 18744460439040.0, + "grad_norm": 1.633002968224224, + "learning_rate": 2.1153176070175293e-06, + "loss": 0.7428, + "num_input_tokens_seen": 179639920, + "step": 5513 + }, + { + "epoch": 0.4972719484150246, + "flos": 27597497520000.0, + "grad_norm": 1.5767798403445523, + "learning_rate": 2.114734363401352e-06, + "loss": 0.6415, + "num_input_tokens_seen": 179672375, + "step": 5514 + }, + { + "epoch": 0.4973621319384948, + "flos": 15319062624480.0, + "grad_norm": 1.9321403204229126, + "learning_rate": 2.1141511099954056e-06, + "loss": 0.7507, + "num_input_tokens_seen": 179698145, + "step": 5515 + }, + { + "epoch": 0.4974523154619651, + "flos": 25661578387200.0, + "grad_norm": 1.4188781848421732, + "learning_rate": 2.1135678468494576e-06, + "loss": 0.7348, + "num_input_tokens_seen": 179727845, + "step": 5516 + }, + { + "epoch": 0.49754249898543534, + "flos": 28507370736000.0, + "grad_norm": 1.9992178937426421, + "learning_rate": 2.112984574013275e-06, + "loss": 0.7912, + "num_input_tokens_seen": 179753800, + "step": 5517 + }, + { + "epoch": 0.49763268250890563, + "flos": 20054900888640.0, + "grad_norm": 1.5680182859220622, + "learning_rate": 2.112401291536625e-06, + "loss": 0.7127, + "num_input_tokens_seen": 179783330, + "step": 5518 + }, + { + "epoch": 0.49772286603237587, + "flos": 20711087527200.0, + "grad_norm": 1.9706548527985925, + "learning_rate": 2.111817999469278e-06, + "loss": 0.7881, + "num_input_tokens_seen": 179811860, + "step": 5519 + }, + { + "epoch": 0.49781304955584615, + "flos": 17979089014560.0, + "grad_norm": 1.636675704353588, + "learning_rate": 2.1112346978610016e-06, + "loss": 0.7037, + "num_input_tokens_seen": 179840180, + "step": 5520 + }, + { + "epoch": 0.4979032330793164, + "flos": 42715889112000.0, + "grad_norm": 1.779914957972469, + "learning_rate": 2.1106513867615678e-06, + "loss": 0.7077, + "num_input_tokens_seen": 179871550, + "step": 5521 + }, + { + "epoch": 0.4979934166027867, + "flos": 23258472343680.0, + "grad_norm": 1.6872793689183383, + "learning_rate": 2.110068066220748e-06, + "loss": 0.756, + "num_input_tokens_seen": 179901465, + "step": 5522 + }, + { + "epoch": 0.4980836001262569, + "flos": 24757392144480.0, + "grad_norm": 1.331966991496699, + "learning_rate": 2.109484736288313e-06, + "loss": 0.8197, + "num_input_tokens_seen": 179933000, + "step": 5523 + }, + { + "epoch": 0.4981737836497272, + "flos": 22933594208640.0, + "grad_norm": 1.8930522376025423, + "learning_rate": 2.108901397014037e-06, + "loss": 0.7689, + "num_input_tokens_seen": 179961825, + "step": 5524 + }, + { + "epoch": 0.49826396717319743, + "flos": 21622298854560.0, + "grad_norm": 2.1037830972589595, + "learning_rate": 2.1083180484476934e-06, + "loss": 0.7568, + "num_input_tokens_seen": 179990355, + "step": 5525 + }, + { + "epoch": 0.4983541506966677, + "flos": 14591201221440.0, + "grad_norm": 1.7411371323480977, + "learning_rate": 2.1077346906390567e-06, + "loss": 0.7734, + "num_input_tokens_seen": 180016555, + "step": 5526 + }, + { + "epoch": 0.498444334220138, + "flos": 16958321203680.0, + "grad_norm": 1.6928632068697878, + "learning_rate": 2.107151323637902e-06, + "loss": 0.7709, + "num_input_tokens_seen": 180043195, + "step": 5527 + }, + { + "epoch": 0.49853451774360824, + "flos": 20311635386400.0, + "grad_norm": 2.1231105482560175, + "learning_rate": 2.106567947494006e-06, + "loss": 0.7423, + "num_input_tokens_seen": 180071610, + "step": 5528 + }, + { + "epoch": 0.49862470126707853, + "flos": 26066494482720.0, + "grad_norm": 1.9479521341885802, + "learning_rate": 2.1059845622571447e-06, + "loss": 0.8196, + "num_input_tokens_seen": 180099190, + "step": 5529 + }, + { + "epoch": 0.49871488479054876, + "flos": 21730703075520.0, + "grad_norm": 1.6727335499091598, + "learning_rate": 2.1054011679770956e-06, + "loss": 0.7938, + "num_input_tokens_seen": 180127090, + "step": 5530 + }, + { + "epoch": 0.49880506831401905, + "flos": 29232741765120.0, + "grad_norm": 2.134852401491392, + "learning_rate": 2.104817764703638e-06, + "loss": 0.763, + "num_input_tokens_seen": 180157035, + "step": 5531 + }, + { + "epoch": 0.4988952518374893, + "flos": 23771569641600.0, + "grad_norm": 1.8090836127270415, + "learning_rate": 2.1042343524865516e-06, + "loss": 0.7582, + "num_input_tokens_seen": 180182480, + "step": 5532 + }, + { + "epoch": 0.4989854353609596, + "flos": 23403308663040.0, + "grad_norm": 1.573938235902086, + "learning_rate": 2.103650931375615e-06, + "loss": 0.862, + "num_input_tokens_seen": 180209845, + "step": 5533 + }, + { + "epoch": 0.4990756188844298, + "flos": 20455728310560.0, + "grad_norm": 1.8365646914277074, + "learning_rate": 2.1030675014206094e-06, + "loss": 0.6921, + "num_input_tokens_seen": 180234645, + "step": 5534 + }, + { + "epoch": 0.4991658024079001, + "flos": 24537089745120.0, + "grad_norm": 2.385962933178082, + "learning_rate": 2.1024840626713166e-06, + "loss": 0.6837, + "num_input_tokens_seen": 180259375, + "step": 5535 + }, + { + "epoch": 0.4992559859313703, + "flos": 22132162383360.0, + "grad_norm": 1.6540881749258978, + "learning_rate": 2.1019006151775177e-06, + "loss": 0.853, + "num_input_tokens_seen": 180287940, + "step": 5536 + }, + { + "epoch": 0.4993461694548406, + "flos": 20201261168160.0, + "grad_norm": 1.7608530831602203, + "learning_rate": 2.101317158988997e-06, + "loss": 0.7425, + "num_input_tokens_seen": 180314920, + "step": 5537 + }, + { + "epoch": 0.49943635297831085, + "flos": 19763852968800.0, + "grad_norm": 2.334454528189262, + "learning_rate": 2.1007336941555374e-06, + "loss": 0.7272, + "num_input_tokens_seen": 180341060, + "step": 5538 + }, + { + "epoch": 0.49952653650178114, + "flos": 26722569612000.0, + "grad_norm": 1.6102648746992103, + "learning_rate": 2.1001502207269238e-06, + "loss": 0.7203, + "num_input_tokens_seen": 180372125, + "step": 5539 + }, + { + "epoch": 0.49961672002525137, + "flos": 24463444983360.0, + "grad_norm": 2.072966059059564, + "learning_rate": 2.0995667387529407e-06, + "loss": 0.7105, + "num_input_tokens_seen": 180400080, + "step": 5540 + }, + { + "epoch": 0.49970690354872166, + "flos": 20489149658400.0, + "grad_norm": 1.5984782224000371, + "learning_rate": 2.098983248283375e-06, + "loss": 0.7047, + "num_input_tokens_seen": 180427650, + "step": 5541 + }, + { + "epoch": 0.4997970870721919, + "flos": 21803753121120.0, + "grad_norm": 2.039702917794375, + "learning_rate": 2.098399749368012e-06, + "loss": 0.7549, + "num_input_tokens_seen": 180456270, + "step": 5542 + }, + { + "epoch": 0.4998872705956622, + "flos": 70552719948480.0, + "grad_norm": 0.6502996615930317, + "learning_rate": 2.09781624205664e-06, + "loss": 0.5693, + "num_input_tokens_seen": 180543765, + "step": 5543 + }, + { + "epoch": 0.4999774541191324, + "flos": 24862525426560.0, + "grad_norm": 1.5355675221648717, + "learning_rate": 2.0972327263990477e-06, + "loss": 0.8014, + "num_input_tokens_seen": 180572170, + "step": 5544 + }, + { + "epoch": 0.5000676376426026, + "flos": 20602162929600.0, + "grad_norm": 1.848819229552701, + "learning_rate": 2.0966492024450226e-06, + "loss": 0.6839, + "num_input_tokens_seen": 180600010, + "step": 5545 + }, + { + "epoch": 0.500157821166073, + "flos": 29200547019360.0, + "grad_norm": 2.6679281787384643, + "learning_rate": 2.0960656702443545e-06, + "loss": 0.9011, + "num_input_tokens_seen": 180629160, + "step": 5546 + }, + { + "epoch": 0.5002480046895432, + "flos": 21148681575360.0, + "grad_norm": 2.39170832802599, + "learning_rate": 2.0954821298468343e-06, + "loss": 0.7527, + "num_input_tokens_seen": 180656555, + "step": 5547 + }, + { + "epoch": 0.5003381882130135, + "flos": 24022803014880.0, + "grad_norm": 1.7854980542191816, + "learning_rate": 2.0948985813022513e-06, + "loss": 0.8142, + "num_input_tokens_seen": 180685155, + "step": 5548 + }, + { + "epoch": 0.5004283717364837, + "flos": 24500062930560.0, + "grad_norm": 2.2311709178940005, + "learning_rate": 2.094315024660399e-06, + "loss": 0.7899, + "num_input_tokens_seen": 180715330, + "step": 5549 + }, + { + "epoch": 0.500518555259954, + "flos": 23006198217120.0, + "grad_norm": 2.048952867985159, + "learning_rate": 2.0937314599710676e-06, + "loss": 0.7754, + "num_input_tokens_seen": 180743795, + "step": 5550 + }, + { + "epoch": 0.5006087387834243, + "flos": 23917186525920.0, + "grad_norm": 2.180379668780306, + "learning_rate": 2.0931478872840526e-06, + "loss": 0.746, + "num_input_tokens_seen": 180773975, + "step": 5551 + }, + { + "epoch": 0.5006989223068945, + "flos": 21986768517600.0, + "grad_norm": 5.0955300058629245, + "learning_rate": 2.092564306649145e-06, + "loss": 0.7331, + "num_input_tokens_seen": 180802100, + "step": 5552 + }, + { + "epoch": 0.5007891058303648, + "flos": 24281358830880.0, + "grad_norm": 1.626831050107688, + "learning_rate": 2.091980718116141e-06, + "loss": 0.7903, + "num_input_tokens_seen": 180833345, + "step": 5553 + }, + { + "epoch": 0.5008792893538351, + "flos": 36268151090400.0, + "grad_norm": 2.495105243213828, + "learning_rate": 2.091397121734835e-06, + "loss": 0.7942, + "num_input_tokens_seen": 180864410, + "step": 5554 + }, + { + "epoch": 0.5009694728773053, + "flos": 24936727734720.0, + "grad_norm": 2.3333916202712834, + "learning_rate": 2.090813517555022e-06, + "loss": 0.774, + "num_input_tokens_seen": 180894305, + "step": 5555 + }, + { + "epoch": 0.5010596564007755, + "flos": 34405839549600.0, + "grad_norm": 1.822062621098871, + "learning_rate": 2.0902299056265e-06, + "loss": 0.6946, + "num_input_tokens_seen": 180922350, + "step": 5556 + }, + { + "epoch": 0.5011498399242459, + "flos": 19249714917600.0, + "grad_norm": 2.299770580127923, + "learning_rate": 2.0896462859990643e-06, + "loss": 0.734, + "num_input_tokens_seen": 180949390, + "step": 5557 + }, + { + "epoch": 0.5012400234477161, + "flos": 20856778751040.0, + "grad_norm": 1.8420702507879254, + "learning_rate": 2.089062658722513e-06, + "loss": 0.8134, + "num_input_tokens_seen": 180976425, + "step": 5558 + }, + { + "epoch": 0.5013302069711864, + "flos": 25231455460800.0, + "grad_norm": 3.227140730430941, + "learning_rate": 2.0884790238466452e-06, + "loss": 0.8015, + "num_input_tokens_seen": 181006855, + "step": 5559 + }, + { + "epoch": 0.5014203904946566, + "flos": 26321407662240.0, + "grad_norm": 2.1771714650496223, + "learning_rate": 2.087895381421259e-06, + "loss": 0.7842, + "num_input_tokens_seen": 181033045, + "step": 5560 + }, + { + "epoch": 0.5015105740181269, + "flos": 21767618380800.0, + "grad_norm": 4.231531518673856, + "learning_rate": 2.087311731496154e-06, + "loss": 0.737, + "num_input_tokens_seen": 181058715, + "step": 5561 + }, + { + "epoch": 0.5016007575415972, + "flos": 21623079419520.0, + "grad_norm": 1.7297388608452258, + "learning_rate": 2.08672807412113e-06, + "loss": 0.7455, + "num_input_tokens_seen": 181088750, + "step": 5562 + }, + { + "epoch": 0.5016909410650674, + "flos": 15902199217440.0, + "grad_norm": 2.583035226254575, + "learning_rate": 2.08614440934599e-06, + "loss": 0.8454, + "num_input_tokens_seen": 181115770, + "step": 5563 + }, + { + "epoch": 0.5017811245885376, + "flos": 29085935448480.0, + "grad_norm": 2.40346147540401, + "learning_rate": 2.0855607372205337e-06, + "loss": 0.6643, + "num_input_tokens_seen": 181144445, + "step": 5564 + }, + { + "epoch": 0.501871308112008, + "flos": 26649816924480.0, + "grad_norm": 2.9545341238058005, + "learning_rate": 2.0849770577945623e-06, + "loss": 0.767, + "num_input_tokens_seen": 181171880, + "step": 5565 + }, + { + "epoch": 0.5019614916354782, + "flos": 42676148904960.0, + "grad_norm": 1.6402668879993851, + "learning_rate": 2.084393371117881e-06, + "loss": 0.7844, + "num_input_tokens_seen": 181205690, + "step": 5566 + }, + { + "epoch": 0.5020516751589484, + "flos": 33935641888320.0, + "grad_norm": 1.8387216438477243, + "learning_rate": 2.0838096772402902e-06, + "loss": 0.6839, + "num_input_tokens_seen": 181238445, + "step": 5567 + }, + { + "epoch": 0.5021418586824187, + "flos": 44461730593920.0, + "grad_norm": 2.525321516963352, + "learning_rate": 2.0832259762115973e-06, + "loss": 0.6382, + "num_input_tokens_seen": 181271920, + "step": 5568 + }, + { + "epoch": 0.502232042205889, + "flos": 26172594178560.0, + "grad_norm": 4.58615730447737, + "learning_rate": 2.082642268081605e-06, + "loss": 0.7821, + "num_input_tokens_seen": 181302520, + "step": 5569 + }, + { + "epoch": 0.5023222257293593, + "flos": 21876059771520.0, + "grad_norm": 2.661949595122465, + "learning_rate": 2.082058552900118e-06, + "loss": 0.7089, + "num_input_tokens_seen": 181326805, + "step": 5570 + }, + { + "epoch": 0.5024124092528295, + "flos": 36194023121760.0, + "grad_norm": 1.7908530125069995, + "learning_rate": 2.081474830716944e-06, + "loss": 0.727, + "num_input_tokens_seen": 181358825, + "step": 5571 + }, + { + "epoch": 0.5025025927762997, + "flos": 20747705474400.0, + "grad_norm": 3.3666862946747043, + "learning_rate": 2.080891101581887e-06, + "loss": 0.7811, + "num_input_tokens_seen": 181387560, + "step": 5572 + }, + { + "epoch": 0.5025927762997701, + "flos": 10582815492960.0, + "grad_norm": 2.477828511496559, + "learning_rate": 2.080307365544755e-06, + "loss": 0.7945, + "num_input_tokens_seen": 181412465, + "step": 5573 + }, + { + "epoch": 0.5026829598232403, + "flos": 22276961532960.0, + "grad_norm": 2.050591741615988, + "learning_rate": 2.0797236226553567e-06, + "loss": 0.7921, + "num_input_tokens_seen": 181443230, + "step": 5574 + }, + { + "epoch": 0.5027731433467105, + "flos": 17141745467520.0, + "grad_norm": 4.088994882656449, + "learning_rate": 2.079139872963499e-06, + "loss": 0.6905, + "num_input_tokens_seen": 181470370, + "step": 5575 + }, + { + "epoch": 0.5028633268701809, + "flos": 19655151389760.0, + "grad_norm": 2.628509536897114, + "learning_rate": 2.078556116518991e-06, + "loss": 0.6547, + "num_input_tokens_seen": 181497195, + "step": 5576 + }, + { + "epoch": 0.5029535103936511, + "flos": 22021974013920.0, + "grad_norm": 1.9455931434265836, + "learning_rate": 2.077972353371642e-06, + "loss": 0.7246, + "num_input_tokens_seen": 181524855, + "step": 5577 + }, + { + "epoch": 0.5030436939171213, + "flos": 21072174742080.0, + "grad_norm": 4.149380621592417, + "learning_rate": 2.077388583571262e-06, + "loss": 0.6996, + "num_input_tokens_seen": 181554785, + "step": 5578 + }, + { + "epoch": 0.5031338774405916, + "flos": 64279708466880.0, + "grad_norm": 0.6489187545912602, + "learning_rate": 2.0768048071676608e-06, + "loss": 0.513, + "num_input_tokens_seen": 181650110, + "step": 5579 + }, + { + "epoch": 0.5032240609640619, + "flos": 23698965633120.0, + "grad_norm": 2.466745359437095, + "learning_rate": 2.0762210242106505e-06, + "loss": 0.7176, + "num_input_tokens_seen": 181679940, + "step": 5580 + }, + { + "epoch": 0.5033142444875321, + "flos": 27269051088000.0, + "grad_norm": 2.5164770732815276, + "learning_rate": 2.0756372347500424e-06, + "loss": 0.7683, + "num_input_tokens_seen": 181708885, + "step": 5581 + }, + { + "epoch": 0.5034044280110024, + "flos": 23043708238560.0, + "grad_norm": 2.6514675395373435, + "learning_rate": 2.0750534388356473e-06, + "loss": 0.8573, + "num_input_tokens_seen": 181732855, + "step": 5582 + }, + { + "epoch": 0.5034946115344726, + "flos": 32551631016480.0, + "grad_norm": 2.331020116849327, + "learning_rate": 2.07446963651728e-06, + "loss": 0.698, + "num_input_tokens_seen": 181763780, + "step": 5583 + }, + { + "epoch": 0.503584795057943, + "flos": 19982408389440.0, + "grad_norm": 2.1747584241683513, + "learning_rate": 2.0738858278447516e-06, + "loss": 0.6915, + "num_input_tokens_seen": 181790755, + "step": 5584 + }, + { + "epoch": 0.5036749785814132, + "flos": 25411088409120.0, + "grad_norm": 1.8261432829061823, + "learning_rate": 2.073302012867878e-06, + "loss": 0.8438, + "num_input_tokens_seen": 181817915, + "step": 5585 + }, + { + "epoch": 0.5037651621048834, + "flos": 20966706932160.0, + "grad_norm": 1.8816935906569014, + "learning_rate": 2.0727181916364725e-06, + "loss": 0.7873, + "num_input_tokens_seen": 181846675, + "step": 5586 + }, + { + "epoch": 0.5038553456283537, + "flos": 22350866483040.0, + "grad_norm": 1.970746012467702, + "learning_rate": 2.0721343642003493e-06, + "loss": 0.7858, + "num_input_tokens_seen": 181874040, + "step": 5587 + }, + { + "epoch": 0.503945529151824, + "flos": 19218077718240.0, + "grad_norm": 3.0748975845277307, + "learning_rate": 2.0715505306093247e-06, + "loss": 0.7813, + "num_input_tokens_seen": 181901380, + "step": 5588 + }, + { + "epoch": 0.5040357126752942, + "flos": 15751898943360.0, + "grad_norm": 2.747507092481489, + "learning_rate": 2.070966690913214e-06, + "loss": 0.7868, + "num_input_tokens_seen": 181925075, + "step": 5589 + }, + { + "epoch": 0.5041258961987645, + "flos": 22788200342880.0, + "grad_norm": 4.642276765368926, + "learning_rate": 2.0703828451618346e-06, + "loss": 0.638, + "num_input_tokens_seen": 181954645, + "step": 5590 + }, + { + "epoch": 0.5042160797222347, + "flos": 25084500465120.0, + "grad_norm": 2.993714582835011, + "learning_rate": 2.069798993405002e-06, + "loss": 0.7481, + "num_input_tokens_seen": 181986990, + "step": 5591 + }, + { + "epoch": 0.504306263245705, + "flos": 27633372072000.0, + "grad_norm": 1.9650331561370709, + "learning_rate": 2.0692151356925345e-06, + "loss": 0.7919, + "num_input_tokens_seen": 182017390, + "step": 5592 + }, + { + "epoch": 0.5043964467691753, + "flos": 25520421874080.0, + "grad_norm": 2.043201440575554, + "learning_rate": 2.068631272074251e-06, + "loss": 0.7762, + "num_input_tokens_seen": 182045495, + "step": 5593 + }, + { + "epoch": 0.5044866302926455, + "flos": 22386443676960.0, + "grad_norm": 2.38113494354788, + "learning_rate": 2.0680474025999676e-06, + "loss": 0.8793, + "num_input_tokens_seen": 182075315, + "step": 5594 + }, + { + "epoch": 0.5045768138161157, + "flos": 70612165861920.0, + "grad_norm": 0.8633369669475702, + "learning_rate": 2.0674635273195055e-06, + "loss": 0.6311, + "num_input_tokens_seen": 182151240, + "step": 5595 + }, + { + "epoch": 0.5046669973395861, + "flos": 23219884399200.0, + "grad_norm": 2.493145510421085, + "learning_rate": 2.066879646282682e-06, + "loss": 0.7513, + "num_input_tokens_seen": 182180670, + "step": 5596 + }, + { + "epoch": 0.5047571808630563, + "flos": 21622038666240.0, + "grad_norm": 2.0523166685201732, + "learning_rate": 2.0662957595393194e-06, + "loss": 0.7254, + "num_input_tokens_seen": 182208440, + "step": 5597 + }, + { + "epoch": 0.5048473643865266, + "flos": 21585680907360.0, + "grad_norm": 2.012424263802617, + "learning_rate": 2.0657118671392373e-06, + "loss": 0.778, + "num_input_tokens_seen": 182237340, + "step": 5598 + }, + { + "epoch": 0.5049375479099969, + "flos": 24864644102880.0, + "grad_norm": 5.282204626358362, + "learning_rate": 2.0651279691322558e-06, + "loss": 0.8104, + "num_input_tokens_seen": 182265050, + "step": 5599 + }, + { + "epoch": 0.5050277314334671, + "flos": 28723993329120.0, + "grad_norm": 1.8025527151708625, + "learning_rate": 2.0645440655681973e-06, + "loss": 0.707, + "num_input_tokens_seen": 182294545, + "step": 5600 + }, + { + "epoch": 0.5051179149569374, + "flos": 20747073588480.0, + "grad_norm": 4.154041822286303, + "learning_rate": 2.0639601564968826e-06, + "loss": 0.8175, + "num_input_tokens_seen": 182321200, + "step": 5601 + }, + { + "epoch": 0.5052080984804076, + "flos": 71970226507680.0, + "grad_norm": 0.6935036737431346, + "learning_rate": 2.0633762419681355e-06, + "loss": 0.6653, + "num_input_tokens_seen": 182415895, + "step": 5602 + }, + { + "epoch": 0.5052982820038779, + "flos": 38674528072800.0, + "grad_norm": 1.8886256089012279, + "learning_rate": 2.062792322031777e-06, + "loss": 0.775, + "num_input_tokens_seen": 182448420, + "step": 5603 + }, + { + "epoch": 0.5053884655273482, + "flos": 21257977870560.0, + "grad_norm": 2.8198144531080898, + "learning_rate": 2.062208396737632e-06, + "loss": 0.7126, + "num_input_tokens_seen": 182476620, + "step": 5604 + }, + { + "epoch": 0.5054786490508184, + "flos": 25739125973760.0, + "grad_norm": 1.7892473949857401, + "learning_rate": 2.0616244661355235e-06, + "loss": 0.7793, + "num_input_tokens_seen": 182507135, + "step": 5605 + }, + { + "epoch": 0.5055688325742886, + "flos": 23297803683360.0, + "grad_norm": 8.182129008236297, + "learning_rate": 2.0610405302752752e-06, + "loss": 0.7882, + "num_input_tokens_seen": 182535905, + "step": 5606 + }, + { + "epoch": 0.505659016097759, + "flos": 31162973924640.0, + "grad_norm": 1.9255543046978905, + "learning_rate": 2.060456589206713e-06, + "loss": 0.745, + "num_input_tokens_seen": 182565595, + "step": 5607 + }, + { + "epoch": 0.5057491996212292, + "flos": 26097425456640.0, + "grad_norm": 2.484259836658629, + "learning_rate": 2.0598726429796614e-06, + "loss": 0.7055, + "num_input_tokens_seen": 182594160, + "step": 5608 + }, + { + "epoch": 0.5058393831446995, + "flos": 29454679633920.0, + "grad_norm": 1.9589673361983329, + "learning_rate": 2.059288691643945e-06, + "loss": 0.717, + "num_input_tokens_seen": 182623220, + "step": 5609 + }, + { + "epoch": 0.5059295666681697, + "flos": 20855589318720.0, + "grad_norm": 2.024844354369942, + "learning_rate": 2.0587047352493913e-06, + "loss": 0.8325, + "num_input_tokens_seen": 182648970, + "step": 5610 + }, + { + "epoch": 0.50601975019164, + "flos": 23188730406720.0, + "grad_norm": 2.0114757798510143, + "learning_rate": 2.0581207738458248e-06, + "loss": 0.7324, + "num_input_tokens_seen": 182677860, + "step": 5611 + }, + { + "epoch": 0.5061099337151103, + "flos": 18306494693280.0, + "grad_norm": 2.4835784402851204, + "learning_rate": 2.0575368074830743e-06, + "loss": 0.7112, + "num_input_tokens_seen": 182703395, + "step": 5612 + }, + { + "epoch": 0.5062001172385805, + "flos": 30767090080800.0, + "grad_norm": 8.655708378672488, + "learning_rate": 2.0569528362109667e-06, + "loss": 0.6648, + "num_input_tokens_seen": 182734525, + "step": 5613 + }, + { + "epoch": 0.5062903007620507, + "flos": 34007242313280.0, + "grad_norm": 1.5903406514820488, + "learning_rate": 2.056368860079327e-06, + "loss": 0.7505, + "num_input_tokens_seen": 182766060, + "step": 5614 + }, + { + "epoch": 0.5063804842855211, + "flos": 21694642674720.0, + "grad_norm": 3.8493544495596566, + "learning_rate": 2.0557848791379874e-06, + "loss": 0.6952, + "num_input_tokens_seen": 182791465, + "step": 5615 + }, + { + "epoch": 0.5064706678089913, + "flos": 20092299400800.0, + "grad_norm": 1.856124729848818, + "learning_rate": 2.0552008934367734e-06, + "loss": 0.6556, + "num_input_tokens_seen": 182820075, + "step": 5616 + }, + { + "epoch": 0.5065608513324615, + "flos": 20018914827360.0, + "grad_norm": 2.2702272172046603, + "learning_rate": 2.0546169030255154e-06, + "loss": 0.8258, + "num_input_tokens_seen": 182844365, + "step": 5617 + }, + { + "epoch": 0.5066510348559318, + "flos": 23516024576160.0, + "grad_norm": 2.410844355308671, + "learning_rate": 2.054032907954041e-06, + "loss": 0.8272, + "num_input_tokens_seen": 182870210, + "step": 5618 + }, + { + "epoch": 0.5067412183794021, + "flos": 22205175259200.0, + "grad_norm": 1.9104560351570685, + "learning_rate": 2.053448908272182e-06, + "loss": 0.7184, + "num_input_tokens_seen": 182901155, + "step": 5619 + }, + { + "epoch": 0.5068314019028723, + "flos": 18598509026880.0, + "grad_norm": 2.32918949241122, + "learning_rate": 2.0528649040297673e-06, + "loss": 0.7887, + "num_input_tokens_seen": 182929120, + "step": 5620 + }, + { + "epoch": 0.5069215854263426, + "flos": 17542089682560.0, + "grad_norm": 1.8183382649737974, + "learning_rate": 2.0522808952766266e-06, + "loss": 0.7595, + "num_input_tokens_seen": 182955785, + "step": 5621 + }, + { + "epoch": 0.5070117689498129, + "flos": 24203997093120.0, + "grad_norm": 1.9061064744292124, + "learning_rate": 2.0516968820625925e-06, + "loss": 0.8002, + "num_input_tokens_seen": 182985100, + "step": 5622 + }, + { + "epoch": 0.5071019524732832, + "flos": 15938222448480.0, + "grad_norm": 2.1142441676450576, + "learning_rate": 2.051112864437495e-06, + "loss": 0.7462, + "num_input_tokens_seen": 183010050, + "step": 5623 + }, + { + "epoch": 0.5071921359967534, + "flos": 22059521205120.0, + "grad_norm": 2.9425471348904924, + "learning_rate": 2.050528842451166e-06, + "loss": 0.7231, + "num_input_tokens_seen": 183033555, + "step": 5624 + }, + { + "epoch": 0.5072823195202236, + "flos": 21439469306880.0, + "grad_norm": 2.1242857461951856, + "learning_rate": 2.049944816153438e-06, + "loss": 0.6357, + "num_input_tokens_seen": 183065455, + "step": 5625 + }, + { + "epoch": 0.507372503043694, + "flos": 24318831682560.0, + "grad_norm": 1.7869802959916288, + "learning_rate": 2.049360785594142e-06, + "loss": 0.7934, + "num_input_tokens_seen": 183093990, + "step": 5626 + }, + { + "epoch": 0.5074626865671642, + "flos": 67344390764160.0, + "grad_norm": 0.6163201817133498, + "learning_rate": 2.048776750823113e-06, + "loss": 0.5098, + "num_input_tokens_seen": 183193180, + "step": 5627 + }, + { + "epoch": 0.5075528700906344, + "flos": 65519403396000.0, + "grad_norm": 0.6398150976750043, + "learning_rate": 2.0481927118901817e-06, + "loss": 0.5627, + "num_input_tokens_seen": 183291965, + "step": 5628 + }, + { + "epoch": 0.5076430536141047, + "flos": 33608607907200.0, + "grad_norm": 2.6825490192296706, + "learning_rate": 2.0476086688451824e-06, + "loss": 0.6962, + "num_input_tokens_seen": 183320610, + "step": 5629 + }, + { + "epoch": 0.507733237137575, + "flos": 27812075776320.0, + "grad_norm": 2.6653616413994103, + "learning_rate": 2.04702462173795e-06, + "loss": 0.7436, + "num_input_tokens_seen": 183349725, + "step": 5630 + }, + { + "epoch": 0.5078234206610452, + "flos": 21147975349920.0, + "grad_norm": 2.686583961684189, + "learning_rate": 2.0464405706183167e-06, + "loss": 0.7318, + "num_input_tokens_seen": 183375705, + "step": 5631 + }, + { + "epoch": 0.5079136041845155, + "flos": 21104109299520.0, + "grad_norm": 1.8842990436080436, + "learning_rate": 2.045856515536118e-06, + "loss": 0.7698, + "num_input_tokens_seen": 183402885, + "step": 5632 + }, + { + "epoch": 0.5080037877079857, + "flos": 19472656369920.0, + "grad_norm": 2.7037432189658848, + "learning_rate": 2.045272456541188e-06, + "loss": 0.7693, + "num_input_tokens_seen": 183431145, + "step": 5633 + }, + { + "epoch": 0.508093971231456, + "flos": 23735137543200.0, + "grad_norm": 2.0541644572877384, + "learning_rate": 2.0446883936833635e-06, + "loss": 0.8257, + "num_input_tokens_seen": 183460040, + "step": 5634 + }, + { + "epoch": 0.5081841547549263, + "flos": 24354780574080.0, + "grad_norm": 2.4011754614457415, + "learning_rate": 2.0441043270124782e-06, + "loss": 0.7243, + "num_input_tokens_seen": 183487665, + "step": 5635 + }, + { + "epoch": 0.5082743382783965, + "flos": 29819297976000.0, + "grad_norm": 2.298509392563767, + "learning_rate": 2.0435202565783683e-06, + "loss": 0.7817, + "num_input_tokens_seen": 183518175, + "step": 5636 + }, + { + "epoch": 0.5083645218018668, + "flos": 25702545196320.0, + "grad_norm": 3.5952673321091235, + "learning_rate": 2.042936182430871e-06, + "loss": 0.7161, + "num_input_tokens_seen": 183546990, + "step": 5637 + }, + { + "epoch": 0.5084547053253371, + "flos": 24755905354080.0, + "grad_norm": 2.259284044811896, + "learning_rate": 2.0423521046198206e-06, + "loss": 0.7605, + "num_input_tokens_seen": 183575025, + "step": 5638 + }, + { + "epoch": 0.5085448888488073, + "flos": 41514930806400.0, + "grad_norm": 2.508725389360827, + "learning_rate": 2.041768023195056e-06, + "loss": 0.672, + "num_input_tokens_seen": 183609335, + "step": 5639 + }, + { + "epoch": 0.5086350723722776, + "flos": 21003176200320.0, + "grad_norm": 1.7580608277283583, + "learning_rate": 2.0411839382064126e-06, + "loss": 0.8227, + "num_input_tokens_seen": 183637370, + "step": 5640 + }, + { + "epoch": 0.5087252558957478, + "flos": 14259149322720.0, + "grad_norm": 2.0976171507645622, + "learning_rate": 2.040599849703729e-06, + "loss": 0.6489, + "num_input_tokens_seen": 183664290, + "step": 5641 + }, + { + "epoch": 0.5088154394192181, + "flos": 29346015224640.0, + "grad_norm": 2.3144814533801674, + "learning_rate": 2.040015757736843e-06, + "loss": 0.692, + "num_input_tokens_seen": 183693165, + "step": 5642 + }, + { + "epoch": 0.5089056229426884, + "flos": 23546918380320.0, + "grad_norm": 2.0679075848130624, + "learning_rate": 2.039431662355591e-06, + "loss": 0.6783, + "num_input_tokens_seen": 183720110, + "step": 5643 + }, + { + "epoch": 0.5089958064661586, + "flos": 22964636691840.0, + "grad_norm": 1.9484458109442087, + "learning_rate": 2.0388475636098126e-06, + "loss": 0.7693, + "num_input_tokens_seen": 183747735, + "step": 5644 + }, + { + "epoch": 0.5090859899896288, + "flos": 22421909361600.0, + "grad_norm": 1.9382436002175216, + "learning_rate": 2.038263461549346e-06, + "loss": 0.7854, + "num_input_tokens_seen": 183776810, + "step": 5645 + }, + { + "epoch": 0.5091761735130992, + "flos": 14002228976160.0, + "grad_norm": 2.931474227300098, + "learning_rate": 2.0376793562240297e-06, + "loss": 0.7547, + "num_input_tokens_seen": 183800370, + "step": 5646 + }, + { + "epoch": 0.5092663570365694, + "flos": 21366865298400.0, + "grad_norm": 2.8572835308593745, + "learning_rate": 2.037095247683703e-06, + "loss": 0.7662, + "num_input_tokens_seen": 183829470, + "step": 5647 + }, + { + "epoch": 0.5093565405600397, + "flos": 25338038363520.0, + "grad_norm": 2.6036191034035507, + "learning_rate": 2.0365111359782046e-06, + "loss": 0.7113, + "num_input_tokens_seen": 183859965, + "step": 5648 + }, + { + "epoch": 0.50944672408351, + "flos": 28544992266720.0, + "grad_norm": 2.1999399957276666, + "learning_rate": 2.0359270211573757e-06, + "loss": 0.7843, + "num_input_tokens_seen": 183892125, + "step": 5649 + }, + { + "epoch": 0.5095369076069802, + "flos": 15318839605920.0, + "grad_norm": 2.57002975624759, + "learning_rate": 2.0353429032710545e-06, + "loss": 0.817, + "num_input_tokens_seen": 183917305, + "step": 5650 + }, + { + "epoch": 0.5096270911304505, + "flos": 22861064539680.0, + "grad_norm": 2.6813490807070686, + "learning_rate": 2.0347587823690825e-06, + "loss": 0.7289, + "num_input_tokens_seen": 183947585, + "step": 5651 + }, + { + "epoch": 0.5097172746539207, + "flos": 28136433534720.0, + "grad_norm": 2.134903438269484, + "learning_rate": 2.034174658501299e-06, + "loss": 0.7525, + "num_input_tokens_seen": 183977760, + "step": 5652 + }, + { + "epoch": 0.509807458177391, + "flos": 31967639519040.0, + "grad_norm": 1.7201916977047695, + "learning_rate": 2.0335905317175453e-06, + "loss": 0.7872, + "num_input_tokens_seen": 184009240, + "step": 5653 + }, + { + "epoch": 0.5098976417008613, + "flos": 36194655007680.0, + "grad_norm": 2.3234295699311964, + "learning_rate": 2.033006402067663e-06, + "loss": 0.7528, + "num_input_tokens_seen": 184038605, + "step": 5654 + }, + { + "epoch": 0.5099878252243315, + "flos": 21549025790400.0, + "grad_norm": 1.7642743506322098, + "learning_rate": 2.0324222696014912e-06, + "loss": 0.6572, + "num_input_tokens_seen": 184066745, + "step": 5655 + }, + { + "epoch": 0.5100780087478017, + "flos": 22606262869440.0, + "grad_norm": 2.5454235036307207, + "learning_rate": 2.0318381343688733e-06, + "loss": 0.7917, + "num_input_tokens_seen": 184093980, + "step": 5656 + }, + { + "epoch": 0.5101681922712721, + "flos": 29928148234080.0, + "grad_norm": 2.3863794862955694, + "learning_rate": 2.0312539964196505e-06, + "loss": 0.7369, + "num_input_tokens_seen": 184123430, + "step": 5657 + }, + { + "epoch": 0.5102583757947423, + "flos": 22934003076000.0, + "grad_norm": 2.666677226601991, + "learning_rate": 2.030669855803664e-06, + "loss": 0.7312, + "num_input_tokens_seen": 184154415, + "step": 5658 + }, + { + "epoch": 0.5103485593182125, + "flos": 22056844982400.0, + "grad_norm": 4.901846884058687, + "learning_rate": 2.0300857125707563e-06, + "loss": 0.8063, + "num_input_tokens_seen": 184182980, + "step": 5659 + }, + { + "epoch": 0.5104387428416828, + "flos": 19726974833280.0, + "grad_norm": 2.684458497495197, + "learning_rate": 2.0295015667707697e-06, + "loss": 0.7816, + "num_input_tokens_seen": 184208790, + "step": 5660 + }, + { + "epoch": 0.5105289263651531, + "flos": 29162553791040.0, + "grad_norm": 2.2741832231100885, + "learning_rate": 2.0289174184535472e-06, + "loss": 0.7614, + "num_input_tokens_seen": 184238085, + "step": 5661 + }, + { + "epoch": 0.5106191098886234, + "flos": 26613421995840.0, + "grad_norm": 2.301066408474761, + "learning_rate": 2.02833326766893e-06, + "loss": 0.7258, + "num_input_tokens_seen": 184266625, + "step": 5662 + }, + { + "epoch": 0.5107092934120936, + "flos": 20050180329120.0, + "grad_norm": 2.077142468844269, + "learning_rate": 2.027749114466763e-06, + "loss": 0.8183, + "num_input_tokens_seen": 184294025, + "step": 5663 + }, + { + "epoch": 0.5107994769355638, + "flos": 22749203531040.0, + "grad_norm": 1.9129836798498696, + "learning_rate": 2.027164958896889e-06, + "loss": 0.7996, + "num_input_tokens_seen": 184324740, + "step": 5664 + }, + { + "epoch": 0.5108896604590342, + "flos": 17790014947200.0, + "grad_norm": 2.3311125768497893, + "learning_rate": 2.02658080100915e-06, + "loss": 0.8726, + "num_input_tokens_seen": 184347965, + "step": 5665 + }, + { + "epoch": 0.5109798439825044, + "flos": 27995462870400.0, + "grad_norm": 1.7580322128313062, + "learning_rate": 2.0259966408533915e-06, + "loss": 0.8263, + "num_input_tokens_seen": 184377000, + "step": 5666 + }, + { + "epoch": 0.5110700275059746, + "flos": 23771235113760.0, + "grad_norm": 2.204264677449038, + "learning_rate": 2.025412478479455e-06, + "loss": 0.8443, + "num_input_tokens_seen": 184404685, + "step": 5667 + }, + { + "epoch": 0.5111602110294449, + "flos": 37143078998400.0, + "grad_norm": 2.5044874924103517, + "learning_rate": 2.0248283139371862e-06, + "loss": 0.7966, + "num_input_tokens_seen": 184433650, + "step": 5668 + }, + { + "epoch": 0.5112503945529152, + "flos": 18853682394720.0, + "grad_norm": 1.996080730716776, + "learning_rate": 2.024244147276429e-06, + "loss": 0.7233, + "num_input_tokens_seen": 184462160, + "step": 5669 + }, + { + "epoch": 0.5113405780763854, + "flos": 26322411245760.0, + "grad_norm": 2.2062890889538624, + "learning_rate": 2.023659978547027e-06, + "loss": 0.8215, + "num_input_tokens_seen": 184492980, + "step": 5670 + }, + { + "epoch": 0.5114307615998557, + "flos": 24171765177600.0, + "grad_norm": 2.0534491342654144, + "learning_rate": 2.023075807798826e-06, + "loss": 0.7539, + "num_input_tokens_seen": 184523755, + "step": 5671 + }, + { + "epoch": 0.511520945123326, + "flos": 27268679390400.0, + "grad_norm": 1.845164193081288, + "learning_rate": 2.0224916350816696e-06, + "loss": 0.6804, + "num_input_tokens_seen": 184557005, + "step": 5672 + }, + { + "epoch": 0.5116111286467963, + "flos": 19108298216160.0, + "grad_norm": 1.702919623914739, + "learning_rate": 2.0219074604454026e-06, + "loss": 0.7785, + "num_input_tokens_seen": 184582065, + "step": 5673 + }, + { + "epoch": 0.5117013121702665, + "flos": 23801199673920.0, + "grad_norm": 2.59776757941063, + "learning_rate": 2.02132328393987e-06, + "loss": 0.7553, + "num_input_tokens_seen": 184607075, + "step": 5674 + }, + { + "epoch": 0.5117914956937367, + "flos": 19977836508960.0, + "grad_norm": 2.143290287473485, + "learning_rate": 2.0207391056149174e-06, + "loss": 0.7072, + "num_input_tokens_seen": 184635940, + "step": 5675 + }, + { + "epoch": 0.5118816792172071, + "flos": 31786222422240.0, + "grad_norm": 2.1687676684624533, + "learning_rate": 2.020154925520391e-06, + "loss": 0.7243, + "num_input_tokens_seen": 184666375, + "step": 5676 + }, + { + "epoch": 0.5119718627406773, + "flos": 28359746684640.0, + "grad_norm": 1.782296994006431, + "learning_rate": 2.0195707437061332e-06, + "loss": 0.8529, + "num_input_tokens_seen": 184694335, + "step": 5677 + }, + { + "epoch": 0.5120620462641475, + "flos": 18451554031200.0, + "grad_norm": 3.2483166957042258, + "learning_rate": 2.0189865602219934e-06, + "loss": 0.756, + "num_input_tokens_seen": 184721070, + "step": 5678 + }, + { + "epoch": 0.5121522297876178, + "flos": 19363397244480.0, + "grad_norm": 2.3698854085525762, + "learning_rate": 2.0184023751178154e-06, + "loss": 0.829, + "num_input_tokens_seen": 184748115, + "step": 5679 + }, + { + "epoch": 0.5122424133110881, + "flos": 24464597245920.0, + "grad_norm": 1.6847559199152058, + "learning_rate": 2.017818188443444e-06, + "loss": 0.7067, + "num_input_tokens_seen": 184779905, + "step": 5680 + }, + { + "epoch": 0.5123325968345583, + "flos": 16557939818880.0, + "grad_norm": 2.468784090603479, + "learning_rate": 2.017234000248728e-06, + "loss": 0.7681, + "num_input_tokens_seen": 184806825, + "step": 5681 + }, + { + "epoch": 0.5124227803580286, + "flos": 20419816588800.0, + "grad_norm": 2.0956826471215266, + "learning_rate": 2.0166498105835108e-06, + "loss": 0.6948, + "num_input_tokens_seen": 184836410, + "step": 5682 + }, + { + "epoch": 0.5125129638814988, + "flos": 31673766697440.0, + "grad_norm": 1.5940508838941514, + "learning_rate": 2.0160656194976407e-06, + "loss": 0.6924, + "num_input_tokens_seen": 184870635, + "step": 5683 + }, + { + "epoch": 0.5126031474049692, + "flos": 24900927522240.0, + "grad_norm": 1.945799801011359, + "learning_rate": 2.0154814270409634e-06, + "loss": 0.7072, + "num_input_tokens_seen": 184900185, + "step": 5684 + }, + { + "epoch": 0.5126933309284394, + "flos": 21804161988480.0, + "grad_norm": 2.0579431073820857, + "learning_rate": 2.0148972332633247e-06, + "loss": 0.7167, + "num_input_tokens_seen": 184925925, + "step": 5685 + }, + { + "epoch": 0.5127835144519096, + "flos": 23881088955360.0, + "grad_norm": 2.9719521369109154, + "learning_rate": 2.0143130382145733e-06, + "loss": 0.8144, + "num_input_tokens_seen": 184955615, + "step": 5686 + }, + { + "epoch": 0.5128736979753798, + "flos": 23006086707840.0, + "grad_norm": 3.4248404695432586, + "learning_rate": 2.0137288419445533e-06, + "loss": 0.7294, + "num_input_tokens_seen": 184983425, + "step": 5687 + }, + { + "epoch": 0.5129638814988502, + "flos": 19836568486560.0, + "grad_norm": 2.377629280305883, + "learning_rate": 2.0131446445031134e-06, + "loss": 0.7953, + "num_input_tokens_seen": 185011805, + "step": 5688 + }, + { + "epoch": 0.5130540650223204, + "flos": 26430815466720.0, + "grad_norm": 2.5403978797585407, + "learning_rate": 2.0125604459400994e-06, + "loss": 0.838, + "num_input_tokens_seen": 185039550, + "step": 5689 + }, + { + "epoch": 0.5131442485457907, + "flos": 21548468244000.0, + "grad_norm": 2.919828136559736, + "learning_rate": 2.0119762463053596e-06, + "loss": 0.8131, + "num_input_tokens_seen": 185067390, + "step": 5690 + }, + { + "epoch": 0.5132344320692609, + "flos": 19399643494080.0, + "grad_norm": 3.253261836652535, + "learning_rate": 2.0113920456487406e-06, + "loss": 0.7443, + "num_input_tokens_seen": 185094310, + "step": 5691 + }, + { + "epoch": 0.5133246155927312, + "flos": 20855366300160.0, + "grad_norm": 4.307070645339511, + "learning_rate": 2.010807844020088e-06, + "loss": 0.7493, + "num_input_tokens_seen": 185122070, + "step": 5692 + }, + { + "epoch": 0.5134147991162015, + "flos": 22459233534240.0, + "grad_norm": 2.0546109145363354, + "learning_rate": 2.0102236414692524e-06, + "loss": 0.8079, + "num_input_tokens_seen": 185151120, + "step": 5693 + }, + { + "epoch": 0.5135049826396717, + "flos": 22785821478240.0, + "grad_norm": 1.7136386816819966, + "learning_rate": 2.0096394380460777e-06, + "loss": 0.7463, + "num_input_tokens_seen": 185180375, + "step": 5694 + }, + { + "epoch": 0.513595166163142, + "flos": 27011796213600.0, + "grad_norm": 1.8369834355462527, + "learning_rate": 2.0090552338004136e-06, + "loss": 0.7867, + "num_input_tokens_seen": 185213025, + "step": 5695 + }, + { + "epoch": 0.5136853496866123, + "flos": 32511927979200.0, + "grad_norm": 1.858759417532464, + "learning_rate": 2.0084710287821077e-06, + "loss": 0.7522, + "num_input_tokens_seen": 185242200, + "step": 5696 + }, + { + "epoch": 0.5137755332100825, + "flos": 25301531925600.0, + "grad_norm": 2.041601945538924, + "learning_rate": 2.007886823041006e-06, + "loss": 0.7931, + "num_input_tokens_seen": 185270795, + "step": 5697 + }, + { + "epoch": 0.5138657167335527, + "flos": 26503977021600.0, + "grad_norm": 2.024297465211359, + "learning_rate": 2.0073026166269577e-06, + "loss": 0.7736, + "num_input_tokens_seen": 185299905, + "step": 5698 + }, + { + "epoch": 0.5139559002570231, + "flos": 23840493843840.0, + "grad_norm": 1.9819071284649, + "learning_rate": 2.0067184095898093e-06, + "loss": 0.8168, + "num_input_tokens_seen": 185330805, + "step": 5699 + }, + { + "epoch": 0.5140460837804933, + "flos": 22824558101760.0, + "grad_norm": 1.9412826406629538, + "learning_rate": 2.0061342019794094e-06, + "loss": 0.7699, + "num_input_tokens_seen": 185359400, + "step": 5700 + }, + { + "epoch": 0.5141362673039636, + "flos": 29163966241920.0, + "grad_norm": 2.998912773214669, + "learning_rate": 2.0055499938456058e-06, + "loss": 0.8351, + "num_input_tokens_seen": 185386880, + "step": 5701 + }, + { + "epoch": 0.5142264508274338, + "flos": 23189473801920.0, + "grad_norm": 2.582619650507173, + "learning_rate": 2.0049657852382464e-06, + "loss": 0.7266, + "num_input_tokens_seen": 185414790, + "step": 5702 + }, + { + "epoch": 0.5143166343509041, + "flos": 40055676873120.0, + "grad_norm": 2.6727767225832686, + "learning_rate": 2.0043815762071782e-06, + "loss": 0.6874, + "num_input_tokens_seen": 185440635, + "step": 5703 + }, + { + "epoch": 0.5144068178743744, + "flos": 24279017136000.0, + "grad_norm": 3.447795362537991, + "learning_rate": 2.0037973668022492e-06, + "loss": 0.7249, + "num_input_tokens_seen": 185470105, + "step": 5704 + }, + { + "epoch": 0.5144970013978446, + "flos": 28362199888800.0, + "grad_norm": 1.9561503058919918, + "learning_rate": 2.003213157073309e-06, + "loss": 0.637, + "num_input_tokens_seen": 185500555, + "step": 5705 + }, + { + "epoch": 0.5145871849213148, + "flos": 28326325336800.0, + "grad_norm": 2.108413703301323, + "learning_rate": 2.002628947070204e-06, + "loss": 0.7517, + "num_input_tokens_seen": 185531425, + "step": 5706 + }, + { + "epoch": 0.5146773684447852, + "flos": 27705976080480.0, + "grad_norm": 3.650542540269579, + "learning_rate": 2.002044736842783e-06, + "loss": 0.8168, + "num_input_tokens_seen": 185560345, + "step": 5707 + }, + { + "epoch": 0.5147675519682554, + "flos": 27740512521120.0, + "grad_norm": 2.2592246315542712, + "learning_rate": 2.001460526440894e-06, + "loss": 0.7377, + "num_input_tokens_seen": 185589670, + "step": 5708 + }, + { + "epoch": 0.5148577354917256, + "flos": 69764743576320.0, + "grad_norm": 0.6253240222241093, + "learning_rate": 2.0008763159143843e-06, + "loss": 0.5806, + "num_input_tokens_seen": 185679085, + "step": 5709 + }, + { + "epoch": 0.5149479190151959, + "flos": 25338075533280.0, + "grad_norm": 1.7009324826179328, + "learning_rate": 2.000292105313103e-06, + "loss": 0.7859, + "num_input_tokens_seen": 185706510, + "step": 5710 + }, + { + "epoch": 0.5150381025386662, + "flos": 19582361532480.0, + "grad_norm": 3.240947440281768, + "learning_rate": 1.999707894686897e-06, + "loss": 0.7952, + "num_input_tokens_seen": 185734120, + "step": 5711 + }, + { + "epoch": 0.5151282860621365, + "flos": 19981553484960.0, + "grad_norm": 1.677289519657076, + "learning_rate": 1.9991236840856155e-06, + "loss": 0.7433, + "num_input_tokens_seen": 185762205, + "step": 5712 + }, + { + "epoch": 0.5152184695856067, + "flos": 22641171007680.0, + "grad_norm": 2.7721507351393684, + "learning_rate": 1.9985394735591065e-06, + "loss": 0.7275, + "num_input_tokens_seen": 185793540, + "step": 5713 + }, + { + "epoch": 0.5153086531090769, + "flos": 20237990624640.0, + "grad_norm": 2.1617834987945486, + "learning_rate": 1.997955263157217e-06, + "loss": 0.7997, + "num_input_tokens_seen": 185819920, + "step": 5714 + }, + { + "epoch": 0.5153988366325473, + "flos": 23225199674880.0, + "grad_norm": 2.9661717512803034, + "learning_rate": 1.997371052929796e-06, + "loss": 0.7884, + "num_input_tokens_seen": 185846040, + "step": 5715 + }, + { + "epoch": 0.5154890201560175, + "flos": 26212334385600.0, + "grad_norm": 1.9914734402168133, + "learning_rate": 1.996786842926691e-06, + "loss": 0.7085, + "num_input_tokens_seen": 185878280, + "step": 5716 + }, + { + "epoch": 0.5155792036794877, + "flos": 25447706356320.0, + "grad_norm": 1.9013579083781025, + "learning_rate": 1.9962026331977506e-06, + "loss": 0.7664, + "num_input_tokens_seen": 185906900, + "step": 5717 + }, + { + "epoch": 0.5156693872029581, + "flos": 21330879237120.0, + "grad_norm": 2.0189583591136695, + "learning_rate": 1.9956184237928224e-06, + "loss": 0.7708, + "num_input_tokens_seen": 185936220, + "step": 5718 + }, + { + "epoch": 0.5157595707264283, + "flos": 16920253635840.0, + "grad_norm": 2.7745633271456525, + "learning_rate": 1.995034214761754e-06, + "loss": 0.7441, + "num_input_tokens_seen": 185962935, + "step": 5719 + }, + { + "epoch": 0.5158497542498985, + "flos": 21038827733760.0, + "grad_norm": 1.4698120497874685, + "learning_rate": 1.9944500061543945e-06, + "loss": 0.7293, + "num_input_tokens_seen": 185992110, + "step": 5720 + }, + { + "epoch": 0.5159399377733688, + "flos": 28398074440800.0, + "grad_norm": 3.256872686496192, + "learning_rate": 1.99386579802059e-06, + "loss": 0.7785, + "num_input_tokens_seen": 186020465, + "step": 5721 + }, + { + "epoch": 0.5160301212968391, + "flos": 23769227946720.0, + "grad_norm": 1.767988665203509, + "learning_rate": 1.993281590410191e-06, + "loss": 0.7342, + "num_input_tokens_seen": 186048365, + "step": 5722 + }, + { + "epoch": 0.5161203048203094, + "flos": 27990890989920.0, + "grad_norm": 1.712292417767506, + "learning_rate": 1.992697383373043e-06, + "loss": 0.7079, + "num_input_tokens_seen": 186077715, + "step": 5723 + }, + { + "epoch": 0.5162104883437796, + "flos": 63301208406720.0, + "grad_norm": 0.7750939250758673, + "learning_rate": 1.9921131769589937e-06, + "loss": 0.6209, + "num_input_tokens_seen": 186159385, + "step": 5724 + }, + { + "epoch": 0.5163006718672498, + "flos": 22313096273280.0, + "grad_norm": 1.7577442071152147, + "learning_rate": 1.991528971217893e-06, + "loss": 0.7559, + "num_input_tokens_seen": 186188000, + "step": 5725 + }, + { + "epoch": 0.5163908553907202, + "flos": 23553311579040.0, + "grad_norm": 1.6694836493254404, + "learning_rate": 1.9909447661995858e-06, + "loss": 0.7494, + "num_input_tokens_seen": 186216195, + "step": 5726 + }, + { + "epoch": 0.5164810389141904, + "flos": 20200443433440.0, + "grad_norm": 2.633556136431607, + "learning_rate": 1.990360561953922e-06, + "loss": 0.7715, + "num_input_tokens_seen": 186241935, + "step": 5727 + }, + { + "epoch": 0.5165712224376606, + "flos": 27666198703680.0, + "grad_norm": 7.444281555721773, + "learning_rate": 1.9897763585307483e-06, + "loss": 0.7761, + "num_input_tokens_seen": 186272950, + "step": 5728 + }, + { + "epoch": 0.5166614059611309, + "flos": 26428808299680.0, + "grad_norm": 8.726322946018163, + "learning_rate": 1.989192155979912e-06, + "loss": 0.7994, + "num_input_tokens_seen": 186302115, + "step": 5729 + }, + { + "epoch": 0.5167515894846012, + "flos": 28256286041760.0, + "grad_norm": 1.4330689511691095, + "learning_rate": 1.98860795435126e-06, + "loss": 0.7918, + "num_input_tokens_seen": 186336470, + "step": 5730 + }, + { + "epoch": 0.5168417730080714, + "flos": 26904432745920.0, + "grad_norm": 1.670219814201917, + "learning_rate": 1.9880237536946406e-06, + "loss": 0.7623, + "num_input_tokens_seen": 186367655, + "step": 5731 + }, + { + "epoch": 0.5169319565315417, + "flos": 14517816648000.0, + "grad_norm": 2.378705998152561, + "learning_rate": 1.987439554059901e-06, + "loss": 0.7233, + "num_input_tokens_seen": 186393460, + "step": 5732 + }, + { + "epoch": 0.5170221400550119, + "flos": 18634792446240.0, + "grad_norm": 2.0572250419805926, + "learning_rate": 1.9868553554968864e-06, + "loss": 0.7286, + "num_input_tokens_seen": 186420035, + "step": 5733 + }, + { + "epoch": 0.5171123235784822, + "flos": 30474778389120.0, + "grad_norm": 2.487500701597489, + "learning_rate": 1.986271158055447e-06, + "loss": 0.7052, + "num_input_tokens_seen": 186449390, + "step": 5734 + }, + { + "epoch": 0.5172025071019525, + "flos": 21473076503520.0, + "grad_norm": 2.116040922270385, + "learning_rate": 1.9856869617854273e-06, + "loss": 0.7488, + "num_input_tokens_seen": 186476915, + "step": 5735 + }, + { + "epoch": 0.5172926906254227, + "flos": 24092582121600.0, + "grad_norm": 2.7492623640342453, + "learning_rate": 1.9851027667366746e-06, + "loss": 0.7351, + "num_input_tokens_seen": 186504345, + "step": 5736 + }, + { + "epoch": 0.517382874148893, + "flos": 33710953457280.0, + "grad_norm": 2.2160160252304966, + "learning_rate": 1.984518572959037e-06, + "loss": 0.8522, + "num_input_tokens_seen": 186533150, + "step": 5737 + }, + { + "epoch": 0.5174730576723633, + "flos": 22601988347040.0, + "grad_norm": 1.91519797763399, + "learning_rate": 1.9839343805023587e-06, + "loss": 0.6815, + "num_input_tokens_seen": 186560800, + "step": 5738 + }, + { + "epoch": 0.5175632411958335, + "flos": 19181385431520.0, + "grad_norm": 5.4610567748396885, + "learning_rate": 1.9833501894164886e-06, + "loss": 0.8459, + "num_input_tokens_seen": 186587630, + "step": 5739 + }, + { + "epoch": 0.5176534247193038, + "flos": 19647494419200.0, + "grad_norm": 2.0909610221656023, + "learning_rate": 1.982765999751273e-06, + "loss": 0.7917, + "num_input_tokens_seen": 186614215, + "step": 5740 + }, + { + "epoch": 0.5177436082427741, + "flos": 21876505808640.0, + "grad_norm": 2.165665676511084, + "learning_rate": 1.9821818115565553e-06, + "loss": 0.7866, + "num_input_tokens_seen": 186639805, + "step": 5741 + }, + { + "epoch": 0.5178337917662443, + "flos": 26431558861920.0, + "grad_norm": 1.9777561247577324, + "learning_rate": 1.9815976248821853e-06, + "loss": 0.6644, + "num_input_tokens_seen": 186668245, + "step": 5742 + }, + { + "epoch": 0.5179239752897146, + "flos": 28325879299680.0, + "grad_norm": 2.845318037066448, + "learning_rate": 1.981013439778007e-06, + "loss": 0.7671, + "num_input_tokens_seen": 186697915, + "step": 5743 + }, + { + "epoch": 0.5180141588131848, + "flos": 19764075987360.0, + "grad_norm": 2.200098759428543, + "learning_rate": 1.9804292562938666e-06, + "loss": 0.7618, + "num_input_tokens_seen": 186725390, + "step": 5744 + }, + { + "epoch": 0.5181043423366551, + "flos": 23225645712000.0, + "grad_norm": 2.17812058654951, + "learning_rate": 1.97984507447961e-06, + "loss": 0.7913, + "num_input_tokens_seen": 186754705, + "step": 5745 + }, + { + "epoch": 0.5181945258601254, + "flos": 23363345437440.0, + "grad_norm": 2.3032889567883434, + "learning_rate": 1.9792608943850824e-06, + "loss": 0.6959, + "num_input_tokens_seen": 186782745, + "step": 5746 + }, + { + "epoch": 0.5182847093835956, + "flos": 24609991111680.0, + "grad_norm": 1.8316336975161904, + "learning_rate": 1.9786767160601305e-06, + "loss": 0.7358, + "num_input_tokens_seen": 186811825, + "step": 5747 + }, + { + "epoch": 0.5183748929070658, + "flos": 18744311760000.0, + "grad_norm": 1.7400617417529485, + "learning_rate": 1.9780925395545977e-06, + "loss": 0.6869, + "num_input_tokens_seen": 186838885, + "step": 5748 + }, + { + "epoch": 0.5184650764305362, + "flos": 68368689702240.0, + "grad_norm": 0.6391297078875978, + "learning_rate": 1.9775083649183306e-06, + "loss": 0.5664, + "num_input_tokens_seen": 186938240, + "step": 5749 + }, + { + "epoch": 0.5185552599540064, + "flos": 18926360742720.0, + "grad_norm": 2.2226595752872202, + "learning_rate": 1.976924192201174e-06, + "loss": 0.7957, + "num_input_tokens_seen": 186962980, + "step": 5750 + }, + { + "epoch": 0.5186454434774767, + "flos": 27665789836320.0, + "grad_norm": 1.8768945023891872, + "learning_rate": 1.9763400214529723e-06, + "loss": 0.6853, + "num_input_tokens_seen": 186992880, + "step": 5751 + }, + { + "epoch": 0.5187356270009469, + "flos": 24245335599840.0, + "grad_norm": 2.0173749793738307, + "learning_rate": 1.9757558527235713e-06, + "loss": 0.6272, + "num_input_tokens_seen": 187023420, + "step": 5752 + }, + { + "epoch": 0.5188258105244172, + "flos": 23662161837120.0, + "grad_norm": 2.401343557668916, + "learning_rate": 1.9751716860628136e-06, + "loss": 0.7427, + "num_input_tokens_seen": 187051580, + "step": 5753 + }, + { + "epoch": 0.5189159940478875, + "flos": 29341183155840.0, + "grad_norm": 1.9835554217631668, + "learning_rate": 1.974587521520545e-06, + "loss": 0.7815, + "num_input_tokens_seen": 187082360, + "step": 5754 + }, + { + "epoch": 0.5190061775713577, + "flos": 31236023970240.0, + "grad_norm": 2.914737899277459, + "learning_rate": 1.9740033591466088e-06, + "loss": 0.7895, + "num_input_tokens_seen": 187112605, + "step": 5755 + }, + { + "epoch": 0.5190963610948279, + "flos": 24645196608000.0, + "grad_norm": 2.1855173117485007, + "learning_rate": 1.97341919899085e-06, + "loss": 0.8671, + "num_input_tokens_seen": 187142385, + "step": 5756 + }, + { + "epoch": 0.5191865446182983, + "flos": 25664068761120.0, + "grad_norm": 1.821327810412216, + "learning_rate": 1.9728350411031114e-06, + "loss": 0.7817, + "num_input_tokens_seen": 187171080, + "step": 5757 + }, + { + "epoch": 0.5192767281417685, + "flos": 22423693510080.0, + "grad_norm": 3.423111226331168, + "learning_rate": 1.9722508855332367e-06, + "loss": 0.8317, + "num_input_tokens_seen": 187199310, + "step": 5758 + }, + { + "epoch": 0.5193669116652387, + "flos": 33607418474880.0, + "grad_norm": 1.5911712727695624, + "learning_rate": 1.97166673233107e-06, + "loss": 0.6892, + "num_input_tokens_seen": 187230240, + "step": 5759 + }, + { + "epoch": 0.519457095188709, + "flos": 30729468550080.0, + "grad_norm": 1.7123396192755462, + "learning_rate": 1.971082581546453e-06, + "loss": 0.7967, + "num_input_tokens_seen": 187261290, + "step": 5760 + }, + { + "epoch": 0.5195472787121793, + "flos": 19142611638240.0, + "grad_norm": 2.5350747117388672, + "learning_rate": 1.9704984332292306e-06, + "loss": 0.7831, + "num_input_tokens_seen": 187288870, + "step": 5761 + }, + { + "epoch": 0.5196374622356495, + "flos": 18598471857120.0, + "grad_norm": 2.269648064711021, + "learning_rate": 1.9699142874292444e-06, + "loss": 0.8465, + "num_input_tokens_seen": 187314065, + "step": 5762 + }, + { + "epoch": 0.5197276457591198, + "flos": 22933557038880.0, + "grad_norm": 1.9890892320681022, + "learning_rate": 1.969330144196336e-06, + "loss": 0.8451, + "num_input_tokens_seen": 187342340, + "step": 5763 + }, + { + "epoch": 0.51981782928259, + "flos": 22424139547200.0, + "grad_norm": 2.8317052614105163, + "learning_rate": 1.9687460035803497e-06, + "loss": 0.7028, + "num_input_tokens_seen": 187370290, + "step": 5764 + }, + { + "epoch": 0.5199080128060604, + "flos": 17286433107840.0, + "grad_norm": 2.2686086016852665, + "learning_rate": 1.9681618656311265e-06, + "loss": 0.7435, + "num_input_tokens_seen": 187396490, + "step": 5765 + }, + { + "epoch": 0.5199981963295306, + "flos": 15319657340640.0, + "grad_norm": 2.830886420005867, + "learning_rate": 1.9675777303985086e-06, + "loss": 0.669, + "num_input_tokens_seen": 187423965, + "step": 5766 + }, + { + "epoch": 0.5200883798530008, + "flos": 19797348656160.0, + "grad_norm": 2.07235593895591, + "learning_rate": 1.9669935979323376e-06, + "loss": 0.7877, + "num_input_tokens_seen": 187451870, + "step": 5767 + }, + { + "epoch": 0.5201785633764712, + "flos": 24973717379520.0, + "grad_norm": 2.650807358455504, + "learning_rate": 1.9664094682824545e-06, + "loss": 0.8139, + "num_input_tokens_seen": 187480400, + "step": 5768 + }, + { + "epoch": 0.5202687468999414, + "flos": 21039496789440.0, + "grad_norm": 1.6663689903818792, + "learning_rate": 1.965825341498701e-06, + "loss": 0.7205, + "num_input_tokens_seen": 187508500, + "step": 5769 + }, + { + "epoch": 0.5203589304234116, + "flos": 21510177657600.0, + "grad_norm": 3.144455125728419, + "learning_rate": 1.9652412176309177e-06, + "loss": 0.8308, + "num_input_tokens_seen": 187536400, + "step": 5770 + }, + { + "epoch": 0.5204491139468819, + "flos": 32839184978880.0, + "grad_norm": 1.8599958264909087, + "learning_rate": 1.9646570967289453e-06, + "loss": 0.6221, + "num_input_tokens_seen": 187568730, + "step": 5771 + }, + { + "epoch": 0.5205392974703522, + "flos": 21471664052640.0, + "grad_norm": 4.193960188531632, + "learning_rate": 1.9640729788426246e-06, + "loss": 0.7944, + "num_input_tokens_seen": 187596645, + "step": 5772 + }, + { + "epoch": 0.5206294809938224, + "flos": 22130489744160.0, + "grad_norm": 1.966565752134327, + "learning_rate": 1.963488864021795e-06, + "loss": 0.7855, + "num_input_tokens_seen": 187627925, + "step": 5773 + }, + { + "epoch": 0.5207196645172927, + "flos": 29738553790080.0, + "grad_norm": 2.0802752620784646, + "learning_rate": 1.962904752316298e-06, + "loss": 0.6511, + "num_input_tokens_seen": 187658630, + "step": 5774 + }, + { + "epoch": 0.5208098480407629, + "flos": 13715604257760.0, + "grad_norm": 2.5284316774008224, + "learning_rate": 1.9623206437759706e-06, + "loss": 0.6864, + "num_input_tokens_seen": 187685645, + "step": 5775 + }, + { + "epoch": 0.5209000315642333, + "flos": 25046470067040.0, + "grad_norm": 1.8676582285735646, + "learning_rate": 1.9617365384506545e-06, + "loss": 0.7171, + "num_input_tokens_seen": 187715155, + "step": 5776 + }, + { + "epoch": 0.5209902150877035, + "flos": 29162479451520.0, + "grad_norm": 1.505138021584657, + "learning_rate": 1.9611524363901872e-06, + "loss": 0.7428, + "num_input_tokens_seen": 187745305, + "step": 5777 + }, + { + "epoch": 0.5210803986111737, + "flos": 20855291960640.0, + "grad_norm": 2.0431966563320656, + "learning_rate": 1.960568337644409e-06, + "loss": 0.7504, + "num_input_tokens_seen": 187773725, + "step": 5778 + }, + { + "epoch": 0.521170582134644, + "flos": 23842835538720.0, + "grad_norm": 1.960383667568472, + "learning_rate": 1.9599842422631576e-06, + "loss": 0.7383, + "num_input_tokens_seen": 187804905, + "step": 5779 + }, + { + "epoch": 0.5212607656581143, + "flos": 32078311095360.0, + "grad_norm": 2.128852605581298, + "learning_rate": 1.9594001502962703e-06, + "loss": 0.7466, + "num_input_tokens_seen": 187834735, + "step": 5780 + }, + { + "epoch": 0.5213509491815845, + "flos": 24755756675040.0, + "grad_norm": 3.0941160351547037, + "learning_rate": 1.9588160617935868e-06, + "loss": 0.8152, + "num_input_tokens_seen": 187864330, + "step": 5781 + }, + { + "epoch": 0.5214411327050548, + "flos": 26758741522080.0, + "grad_norm": 3.451830452999875, + "learning_rate": 1.958231976804944e-06, + "loss": 0.8441, + "num_input_tokens_seen": 187890590, + "step": 5782 + }, + { + "epoch": 0.521531316228525, + "flos": 21258015040320.0, + "grad_norm": 2.3607947517830485, + "learning_rate": 1.957647895380179e-06, + "loss": 0.6802, + "num_input_tokens_seen": 187918565, + "step": 5783 + }, + { + "epoch": 0.5216214997519953, + "flos": 48073117385760.0, + "grad_norm": 1.722026337828681, + "learning_rate": 1.9570638175691297e-06, + "loss": 0.6832, + "num_input_tokens_seen": 187956435, + "step": 5784 + }, + { + "epoch": 0.5217116832754656, + "flos": 24135927795360.0, + "grad_norm": 2.374851422932782, + "learning_rate": 1.956479743421632e-06, + "loss": 0.7987, + "num_input_tokens_seen": 187984535, + "step": 5785 + }, + { + "epoch": 0.5218018667989358, + "flos": 19072349324640.0, + "grad_norm": 2.738385643377179, + "learning_rate": 1.955895672987522e-06, + "loss": 0.689, + "num_input_tokens_seen": 188011020, + "step": 5786 + }, + { + "epoch": 0.521892050322406, + "flos": 27634784522880.0, + "grad_norm": 1.615330611363751, + "learning_rate": 1.9553116063166367e-06, + "loss": 0.661, + "num_input_tokens_seen": 188043770, + "step": 5787 + }, + { + "epoch": 0.5219822338458764, + "flos": 30296557891680.0, + "grad_norm": 1.5637946296745069, + "learning_rate": 1.954727543458812e-06, + "loss": 0.6614, + "num_input_tokens_seen": 188078530, + "step": 5788 + }, + { + "epoch": 0.5220724173693466, + "flos": 22856938696320.0, + "grad_norm": 2.2096829039296995, + "learning_rate": 1.954143484463883e-06, + "loss": 0.7839, + "num_input_tokens_seen": 188107995, + "step": 5789 + }, + { + "epoch": 0.5221626008928169, + "flos": 20927189743680.0, + "grad_norm": 2.5825414571914993, + "learning_rate": 1.9535594293816836e-06, + "loss": 0.7394, + "num_input_tokens_seen": 188135125, + "step": 5790 + }, + { + "epoch": 0.5222527844162872, + "flos": 14736446408160.0, + "grad_norm": 3.660825686471028, + "learning_rate": 1.952975378262051e-06, + "loss": 0.7016, + "num_input_tokens_seen": 188162510, + "step": 5791 + }, + { + "epoch": 0.5223429679397574, + "flos": 17213829099360.0, + "grad_norm": 2.8808868510342287, + "learning_rate": 1.952391331154817e-06, + "loss": 0.7067, + "num_input_tokens_seen": 188188695, + "step": 5792 + }, + { + "epoch": 0.5224331514632277, + "flos": 26025602013120.0, + "grad_norm": 1.7425629078744274, + "learning_rate": 1.9518072881098185e-06, + "loss": 0.8122, + "num_input_tokens_seen": 188216970, + "step": 5793 + }, + { + "epoch": 0.5225233349866979, + "flos": 27343699433280.0, + "grad_norm": 4.918837999964774, + "learning_rate": 1.9512232491768867e-06, + "loss": 0.7446, + "num_input_tokens_seen": 188251140, + "step": 5794 + }, + { + "epoch": 0.5226135185101682, + "flos": 28145131258560.0, + "grad_norm": 1.5368538126623335, + "learning_rate": 1.9506392144058573e-06, + "loss": 0.7457, + "num_input_tokens_seen": 188284345, + "step": 5795 + }, + { + "epoch": 0.5227037020336385, + "flos": 20018766148320.0, + "grad_norm": 1.7718583525021934, + "learning_rate": 1.9500551838465623e-06, + "loss": 0.7475, + "num_input_tokens_seen": 188310420, + "step": 5796 + }, + { + "epoch": 0.5227938855571087, + "flos": 32439249631200.0, + "grad_norm": 1.6057675865477687, + "learning_rate": 1.9494711575488337e-06, + "loss": 0.7673, + "num_input_tokens_seen": 188341290, + "step": 5797 + }, + { + "epoch": 0.5228840690805789, + "flos": 18962123785440.0, + "grad_norm": 2.3089559048307704, + "learning_rate": 1.948887135562505e-06, + "loss": 0.6978, + "num_input_tokens_seen": 188368170, + "step": 5798 + }, + { + "epoch": 0.5229742526040493, + "flos": 34446769188960.0, + "grad_norm": 1.7572850092875023, + "learning_rate": 1.9483031179374074e-06, + "loss": 0.8011, + "num_input_tokens_seen": 188400240, + "step": 5799 + }, + { + "epoch": 0.5230644361275195, + "flos": 23003782182720.0, + "grad_norm": 1.6699430765512664, + "learning_rate": 1.9477191047233736e-06, + "loss": 0.8115, + "num_input_tokens_seen": 188429125, + "step": 5800 + }, + { + "epoch": 0.5231546196509897, + "flos": 33352505295360.0, + "grad_norm": 1.9964982013078092, + "learning_rate": 1.9471350959702334e-06, + "loss": 0.6367, + "num_input_tokens_seen": 188460810, + "step": 5801 + }, + { + "epoch": 0.52324480317446, + "flos": 21257345984640.0, + "grad_norm": 2.405318240802779, + "learning_rate": 1.9465510917278184e-06, + "loss": 0.76, + "num_input_tokens_seen": 188489630, + "step": 5802 + }, + { + "epoch": 0.5233349866979303, + "flos": 21877137694560.0, + "grad_norm": 1.8893089442894315, + "learning_rate": 1.9459670920459593e-06, + "loss": 0.7502, + "num_input_tokens_seen": 188517065, + "step": 5803 + }, + { + "epoch": 0.5234251702214006, + "flos": 50396743015200.0, + "grad_norm": 2.1712551192869496, + "learning_rate": 1.945383096974485e-06, + "loss": 0.6702, + "num_input_tokens_seen": 188548590, + "step": 5804 + }, + { + "epoch": 0.5235153537448708, + "flos": 36013014892320.0, + "grad_norm": 2.674669288085662, + "learning_rate": 1.944799106563227e-06, + "loss": 0.6746, + "num_input_tokens_seen": 188580005, + "step": 5805 + }, + { + "epoch": 0.523605537268341, + "flos": 26977334112480.0, + "grad_norm": 2.193236105521471, + "learning_rate": 1.9442151208620133e-06, + "loss": 0.7914, + "num_input_tokens_seen": 188606660, + "step": 5806 + }, + { + "epoch": 0.5236957207918114, + "flos": 22715187467040.0, + "grad_norm": 2.05863405365551, + "learning_rate": 1.943631139920672e-06, + "loss": 0.8026, + "num_input_tokens_seen": 188633975, + "step": 5807 + }, + { + "epoch": 0.5237859043152816, + "flos": 23189027764800.0, + "grad_norm": 2.142303922176007, + "learning_rate": 1.943047163789034e-06, + "loss": 0.7028, + "num_input_tokens_seen": 188662120, + "step": 5808 + }, + { + "epoch": 0.5238760878387518, + "flos": 30437900253600.0, + "grad_norm": 1.564427826828072, + "learning_rate": 1.942463192516925e-06, + "loss": 0.7104, + "num_input_tokens_seen": 188695885, + "step": 5809 + }, + { + "epoch": 0.5239662713622221, + "flos": 22712474074560.0, + "grad_norm": 1.8683923640639644, + "learning_rate": 1.9418792261541746e-06, + "loss": 0.7713, + "num_input_tokens_seen": 188726495, + "step": 5810 + }, + { + "epoch": 0.5240564548856924, + "flos": 17760124726560.0, + "grad_norm": 2.0094456543076666, + "learning_rate": 1.9412952647506094e-06, + "loss": 0.7079, + "num_input_tokens_seen": 188754070, + "step": 5811 + }, + { + "epoch": 0.5241466384091626, + "flos": 23698928463360.0, + "grad_norm": 2.117842681971405, + "learning_rate": 1.9407113083560552e-06, + "loss": 0.7285, + "num_input_tokens_seen": 188784965, + "step": 5812 + }, + { + "epoch": 0.5242368219326329, + "flos": 24317679420000.0, + "grad_norm": 2.3268864556624087, + "learning_rate": 1.940127357020339e-06, + "loss": 0.7345, + "num_input_tokens_seen": 188812945, + "step": 5813 + }, + { + "epoch": 0.5243270054561032, + "flos": 15647471886720.0, + "grad_norm": 1.8725637502984738, + "learning_rate": 1.939543410793287e-06, + "loss": 0.76, + "num_input_tokens_seen": 188840230, + "step": 5814 + }, + { + "epoch": 0.5244171889795735, + "flos": 19799839030080.0, + "grad_norm": 1.8400194606581164, + "learning_rate": 1.9389594697247246e-06, + "loss": 0.7888, + "num_input_tokens_seen": 188868210, + "step": 5815 + }, + { + "epoch": 0.5245073725030437, + "flos": 23005640670720.0, + "grad_norm": 2.2443427915288368, + "learning_rate": 1.9383755338644763e-06, + "loss": 0.8278, + "num_input_tokens_seen": 188895785, + "step": 5816 + }, + { + "epoch": 0.5245975560265139, + "flos": 18196938209760.0, + "grad_norm": 2.1522545770079593, + "learning_rate": 1.937791603262368e-06, + "loss": 0.7809, + "num_input_tokens_seen": 188921830, + "step": 5817 + }, + { + "epoch": 0.5246877395499843, + "flos": 23439443403360.0, + "grad_norm": 1.8124593532871613, + "learning_rate": 1.9372076779682235e-06, + "loss": 0.7037, + "num_input_tokens_seen": 188949075, + "step": 5818 + }, + { + "epoch": 0.5247779230734545, + "flos": 22198187344320.0, + "grad_norm": 2.4293522741512454, + "learning_rate": 1.9366237580318648e-06, + "loss": 0.7043, + "num_input_tokens_seen": 188977620, + "step": 5819 + }, + { + "epoch": 0.5248681065969247, + "flos": 19837757918880.0, + "grad_norm": 1.5566025637157923, + "learning_rate": 1.9360398435031176e-06, + "loss": 0.7579, + "num_input_tokens_seen": 189005385, + "step": 5820 + }, + { + "epoch": 0.524958290120395, + "flos": 24318385645440.0, + "grad_norm": 3.1518359147461794, + "learning_rate": 1.9354559344318025e-06, + "loss": 0.7602, + "num_input_tokens_seen": 189034875, + "step": 5821 + }, + { + "epoch": 0.5250484736438653, + "flos": 19253915100480.0, + "grad_norm": 2.044658946392226, + "learning_rate": 1.934872030867744e-06, + "loss": 0.805, + "num_input_tokens_seen": 189060820, + "step": 5822 + }, + { + "epoch": 0.5251386571673355, + "flos": 24245893146240.0, + "grad_norm": 3.027686157551329, + "learning_rate": 1.934288132860763e-06, + "loss": 0.7637, + "num_input_tokens_seen": 189090120, + "step": 5823 + }, + { + "epoch": 0.5252288406908058, + "flos": 25338744588960.0, + "grad_norm": 1.656441947322774, + "learning_rate": 1.93370424046068e-06, + "loss": 0.7504, + "num_input_tokens_seen": 189118475, + "step": 5824 + }, + { + "epoch": 0.525319024214276, + "flos": 20019732562080.0, + "grad_norm": 1.7966516107106478, + "learning_rate": 1.9331203537173177e-06, + "loss": 0.8133, + "num_input_tokens_seen": 189147845, + "step": 5825 + }, + { + "epoch": 0.5254092077377464, + "flos": 29928185403840.0, + "grad_norm": 1.4587372719962075, + "learning_rate": 1.9325364726804947e-06, + "loss": 0.8635, + "num_input_tokens_seen": 189179670, + "step": 5826 + }, + { + "epoch": 0.5254993912612166, + "flos": 17684509967520.0, + "grad_norm": 2.2818653896592274, + "learning_rate": 1.9319525974000327e-06, + "loss": 0.6802, + "num_input_tokens_seen": 189206735, + "step": 5827 + }, + { + "epoch": 0.5255895747846868, + "flos": 13927729309920.0, + "grad_norm": 3.246268551156171, + "learning_rate": 1.93136872792575e-06, + "loss": 0.8161, + "num_input_tokens_seen": 189231250, + "step": 5828 + }, + { + "epoch": 0.525679758308157, + "flos": 22678532350080.0, + "grad_norm": 2.2825641765952343, + "learning_rate": 1.9307848643074653e-06, + "loss": 0.7052, + "num_input_tokens_seen": 189261090, + "step": 5829 + }, + { + "epoch": 0.5257699418316274, + "flos": 21805797457920.0, + "grad_norm": 1.7566076600059009, + "learning_rate": 1.9302010065949985e-06, + "loss": 0.7845, + "num_input_tokens_seen": 189290970, + "step": 5830 + }, + { + "epoch": 0.5258601253550976, + "flos": 26576618199840.0, + "grad_norm": 1.7403168492837386, + "learning_rate": 1.9296171548381657e-06, + "loss": 0.7121, + "num_input_tokens_seen": 189320125, + "step": 5831 + }, + { + "epoch": 0.5259503088785679, + "flos": 24023769428640.0, + "grad_norm": 2.195870868089408, + "learning_rate": 1.9290333090867862e-06, + "loss": 0.7389, + "num_input_tokens_seen": 189348875, + "step": 5832 + }, + { + "epoch": 0.5260404924020381, + "flos": 18452446105440.0, + "grad_norm": 3.7700832456914477, + "learning_rate": 1.928449469390676e-06, + "loss": 0.7543, + "num_input_tokens_seen": 189373740, + "step": 5833 + }, + { + "epoch": 0.5261306759255084, + "flos": 20857447806720.0, + "grad_norm": 1.8997800498942172, + "learning_rate": 1.927865635799651e-06, + "loss": 0.7235, + "num_input_tokens_seen": 189401800, + "step": 5834 + }, + { + "epoch": 0.5262208594489787, + "flos": 26136013401120.0, + "grad_norm": 1.7530506655459417, + "learning_rate": 1.927281808363528e-06, + "loss": 0.6774, + "num_input_tokens_seen": 189430985, + "step": 5835 + }, + { + "epoch": 0.5263110429724489, + "flos": 28543691325120.0, + "grad_norm": 1.9684665415320692, + "learning_rate": 1.9266979871321216e-06, + "loss": 0.8274, + "num_input_tokens_seen": 189460745, + "step": 5836 + }, + { + "epoch": 0.5264012264959192, + "flos": 23183935507680.0, + "grad_norm": 2.0720948806012403, + "learning_rate": 1.9261141721552482e-06, + "loss": 0.7024, + "num_input_tokens_seen": 189488320, + "step": 5837 + }, + { + "epoch": 0.5264914100193895, + "flos": 18489398580480.0, + "grad_norm": 3.0552398884640812, + "learning_rate": 1.9255303634827204e-06, + "loss": 0.8187, + "num_input_tokens_seen": 189515905, + "step": 5838 + }, + { + "epoch": 0.5265815935428597, + "flos": 19798203560640.0, + "grad_norm": 1.9997700320591432, + "learning_rate": 1.924946561164352e-06, + "loss": 0.8186, + "num_input_tokens_seen": 189542705, + "step": 5839 + }, + { + "epoch": 0.52667177706633, + "flos": 27228976353120.0, + "grad_norm": 1.9115569862733963, + "learning_rate": 1.9243627652499582e-06, + "loss": 0.7025, + "num_input_tokens_seen": 189573370, + "step": 5840 + }, + { + "epoch": 0.5267619605898003, + "flos": 21401104380960.0, + "grad_norm": 1.8135877830898224, + "learning_rate": 1.9237789757893493e-06, + "loss": 0.8154, + "num_input_tokens_seen": 189601630, + "step": 5841 + }, + { + "epoch": 0.5268521441132705, + "flos": 14700051479520.0, + "grad_norm": 2.196187778171194, + "learning_rate": 1.9231951928323395e-06, + "loss": 0.7984, + "num_input_tokens_seen": 189626370, + "step": 5842 + }, + { + "epoch": 0.5269423276367408, + "flos": 25374470461920.0, + "grad_norm": 2.0414168406115807, + "learning_rate": 1.922611416428738e-06, + "loss": 0.7049, + "num_input_tokens_seen": 189654325, + "step": 5843 + }, + { + "epoch": 0.527032511160211, + "flos": 21360174741600.0, + "grad_norm": 2.1269442695594467, + "learning_rate": 1.922027646628358e-06, + "loss": 0.7775, + "num_input_tokens_seen": 189681730, + "step": 5844 + }, + { + "epoch": 0.5271226946836813, + "flos": 58737420459840.0, + "grad_norm": 0.6592752345032615, + "learning_rate": 1.9214438834810092e-06, + "loss": 0.6344, + "num_input_tokens_seen": 189768575, + "step": 5845 + }, + { + "epoch": 0.5272128782071516, + "flos": 21292700160000.0, + "grad_norm": 1.6172396886393352, + "learning_rate": 1.9208601270365008e-06, + "loss": 0.6813, + "num_input_tokens_seen": 189797560, + "step": 5846 + }, + { + "epoch": 0.5273030617306218, + "flos": 20638780876800.0, + "grad_norm": 1.7220304913939257, + "learning_rate": 1.9202763773446435e-06, + "loss": 0.8323, + "num_input_tokens_seen": 189826020, + "step": 5847 + }, + { + "epoch": 0.527393245254092, + "flos": 25156472587680.0, + "grad_norm": 2.0895008982418894, + "learning_rate": 1.9196926344552444e-06, + "loss": 0.6505, + "num_input_tokens_seen": 189853750, + "step": 5848 + }, + { + "epoch": 0.5274834287775624, + "flos": 25259264174880.0, + "grad_norm": 4.202629201923756, + "learning_rate": 1.919108898418113e-06, + "loss": 0.7786, + "num_input_tokens_seen": 189883910, + "step": 5849 + }, + { + "epoch": 0.5275736123010326, + "flos": 32953164663840.0, + "grad_norm": 1.4914630509396185, + "learning_rate": 1.918525169283057e-06, + "loss": 0.6482, + "num_input_tokens_seen": 189919365, + "step": 5850 + }, + { + "epoch": 0.5276637958245028, + "flos": 24496903500960.0, + "grad_norm": 3.018652152614051, + "learning_rate": 1.9179414470998817e-06, + "loss": 0.7762, + "num_input_tokens_seen": 189950805, + "step": 5851 + }, + { + "epoch": 0.5277539793479731, + "flos": 33640988501760.0, + "grad_norm": 2.874090465473059, + "learning_rate": 1.917357731918395e-06, + "loss": 0.711, + "num_input_tokens_seen": 189981365, + "step": 5852 + }, + { + "epoch": 0.5278441628714434, + "flos": 21141433472160.0, + "grad_norm": 1.6945000033138002, + "learning_rate": 1.9167740237884025e-06, + "loss": 0.8812, + "num_input_tokens_seen": 190007175, + "step": 5853 + }, + { + "epoch": 0.5279343463949137, + "flos": 21726354213600.0, + "grad_norm": 1.8383089373927592, + "learning_rate": 1.916190322759709e-06, + "loss": 0.7923, + "num_input_tokens_seen": 190034975, + "step": 5854 + }, + { + "epoch": 0.5280245299183839, + "flos": 23043410880480.0, + "grad_norm": 2.379395984169344, + "learning_rate": 1.91560662888212e-06, + "loss": 0.8519, + "num_input_tokens_seen": 190063040, + "step": 5855 + }, + { + "epoch": 0.5281147134418541, + "flos": 23071888650240.0, + "grad_norm": 1.8330502797299884, + "learning_rate": 1.915022942205438e-06, + "loss": 0.8004, + "num_input_tokens_seen": 190089630, + "step": 5856 + }, + { + "epoch": 0.5282048969653245, + "flos": 13897950598560.0, + "grad_norm": 2.2646181171159143, + "learning_rate": 1.914439262779468e-06, + "loss": 0.745, + "num_input_tokens_seen": 190114820, + "step": 5857 + }, + { + "epoch": 0.5282950804887947, + "flos": 21286827337920.0, + "grad_norm": 2.2727384246205777, + "learning_rate": 1.9138555906540103e-06, + "loss": 0.6859, + "num_input_tokens_seen": 190141440, + "step": 5858 + }, + { + "epoch": 0.5283852640122649, + "flos": 24682297762080.0, + "grad_norm": 1.6695924668252253, + "learning_rate": 1.91327192587887e-06, + "loss": 0.6607, + "num_input_tokens_seen": 190171425, + "step": 5859 + }, + { + "epoch": 0.5284754475357353, + "flos": 27159866302080.0, + "grad_norm": 2.1823933531408244, + "learning_rate": 1.912688268503846e-06, + "loss": 0.7278, + "num_input_tokens_seen": 190200040, + "step": 5860 + }, + { + "epoch": 0.5285656310592055, + "flos": 16740174650400.0, + "grad_norm": 2.4068119334217473, + "learning_rate": 1.912104618578741e-06, + "loss": 0.7146, + "num_input_tokens_seen": 190226500, + "step": 5861 + }, + { + "epoch": 0.5286558145826757, + "flos": 30327897732960.0, + "grad_norm": 1.6872957600515144, + "learning_rate": 1.9115209761533554e-06, + "loss": 0.7327, + "num_input_tokens_seen": 190260250, + "step": 5862 + }, + { + "epoch": 0.528745998106146, + "flos": 24537275593920.0, + "grad_norm": 2.4201927002723185, + "learning_rate": 1.9109373412774863e-06, + "loss": 0.6665, + "num_input_tokens_seen": 190288440, + "step": 5863 + }, + { + "epoch": 0.5288361816296163, + "flos": 24134032137600.0, + "grad_norm": 1.9225164444000595, + "learning_rate": 1.910353714000936e-06, + "loss": 0.7569, + "num_input_tokens_seen": 190318040, + "step": 5864 + }, + { + "epoch": 0.5289263651530866, + "flos": 28504025457600.0, + "grad_norm": 3.0949760088506872, + "learning_rate": 1.9097700943734997e-06, + "loss": 0.7103, + "num_input_tokens_seen": 190349635, + "step": 5865 + }, + { + "epoch": 0.5290165486765568, + "flos": 14518076836320.0, + "grad_norm": 2.11407681366557, + "learning_rate": 1.909186482444977e-06, + "loss": 0.6531, + "num_input_tokens_seen": 190376635, + "step": 5866 + }, + { + "epoch": 0.529106732200027, + "flos": 22678829708160.0, + "grad_norm": 2.0973282030062648, + "learning_rate": 1.9086028782651652e-06, + "loss": 0.6568, + "num_input_tokens_seen": 190405415, + "step": 5867 + }, + { + "epoch": 0.5291969157234974, + "flos": 19871513794560.0, + "grad_norm": 1.7687485334302666, + "learning_rate": 1.908019281883859e-06, + "loss": 0.7477, + "num_input_tokens_seen": 190432355, + "step": 5868 + }, + { + "epoch": 0.5292870992469676, + "flos": 25229002256640.0, + "grad_norm": 1.8492895411747692, + "learning_rate": 1.9074356933508545e-06, + "loss": 0.7437, + "num_input_tokens_seen": 190461165, + "step": 5869 + }, + { + "epoch": 0.5293772827704378, + "flos": 18335418500160.0, + "grad_norm": 2.8270919346145598, + "learning_rate": 1.9068521127159477e-06, + "loss": 0.7805, + "num_input_tokens_seen": 190484880, + "step": 5870 + }, + { + "epoch": 0.5294674662939081, + "flos": 24318385645440.0, + "grad_norm": 1.7413564846891054, + "learning_rate": 1.9062685400289322e-06, + "loss": 0.6985, + "num_input_tokens_seen": 190514035, + "step": 5871 + }, + { + "epoch": 0.5295576498173784, + "flos": 20311338028320.0, + "grad_norm": 3.1975694068134253, + "learning_rate": 1.9056849753396018e-06, + "loss": 0.7251, + "num_input_tokens_seen": 190542400, + "step": 5872 + }, + { + "epoch": 0.5296478333408486, + "flos": 23006941612320.0, + "grad_norm": 2.1378323792926106, + "learning_rate": 1.9051014186977485e-06, + "loss": 0.7838, + "num_input_tokens_seen": 190571505, + "step": 5873 + }, + { + "epoch": 0.5297380168643189, + "flos": 23438179631520.0, + "grad_norm": 7.058996631332396, + "learning_rate": 1.9045178701531664e-06, + "loss": 0.7501, + "num_input_tokens_seen": 190600755, + "step": 5874 + }, + { + "epoch": 0.5298282003877891, + "flos": 22495739972160.0, + "grad_norm": 1.5284508792570057, + "learning_rate": 1.903934329755645e-06, + "loss": 0.7455, + "num_input_tokens_seen": 190631375, + "step": 5875 + }, + { + "epoch": 0.5299183839112594, + "flos": 26977148263680.0, + "grad_norm": 1.945527765703855, + "learning_rate": 1.9033507975549775e-06, + "loss": 0.7873, + "num_input_tokens_seen": 190660420, + "step": 5876 + }, + { + "epoch": 0.5300085674347297, + "flos": 19689278963040.0, + "grad_norm": 1.6342961629667874, + "learning_rate": 1.9027672736009525e-06, + "loss": 0.7424, + "num_input_tokens_seen": 190687300, + "step": 5877 + }, + { + "epoch": 0.5300987509581999, + "flos": 30365110396320.0, + "grad_norm": 1.657227785360664, + "learning_rate": 1.9021837579433593e-06, + "loss": 0.6985, + "num_input_tokens_seen": 190717905, + "step": 5878 + }, + { + "epoch": 0.5301889344816701, + "flos": 35722227160800.0, + "grad_norm": 2.0048761530271766, + "learning_rate": 1.901600250631988e-06, + "loss": 0.7244, + "num_input_tokens_seen": 190749420, + "step": 5879 + }, + { + "epoch": 0.5302791180051405, + "flos": 23695285826880.0, + "grad_norm": 2.0658934449065876, + "learning_rate": 1.901016751716625e-06, + "loss": 0.7382, + "num_input_tokens_seen": 190778015, + "step": 5880 + }, + { + "epoch": 0.5303693015286107, + "flos": 23515764387840.0, + "grad_norm": 1.5561095208798448, + "learning_rate": 1.9004332612470593e-06, + "loss": 0.715, + "num_input_tokens_seen": 190808170, + "step": 5881 + }, + { + "epoch": 0.530459485052081, + "flos": 23697999219360.0, + "grad_norm": 1.3130033072976894, + "learning_rate": 1.8998497792730763e-06, + "loss": 0.7734, + "num_input_tokens_seen": 190838515, + "step": 5882 + }, + { + "epoch": 0.5305496685755512, + "flos": 25229076596160.0, + "grad_norm": 1.7110936341831258, + "learning_rate": 1.8992663058444629e-06, + "loss": 0.8023, + "num_input_tokens_seen": 190868365, + "step": 5883 + }, + { + "epoch": 0.5306398520990215, + "flos": 33061011338400.0, + "grad_norm": 2.210142821793336, + "learning_rate": 1.8986828410110032e-06, + "loss": 0.7888, + "num_input_tokens_seen": 190899650, + "step": 5884 + }, + { + "epoch": 0.5307300356224918, + "flos": 23953841642880.0, + "grad_norm": 2.1150278710948025, + "learning_rate": 1.8980993848224823e-06, + "loss": 0.7406, + "num_input_tokens_seen": 190931765, + "step": 5885 + }, + { + "epoch": 0.530820219145962, + "flos": 24500546137440.0, + "grad_norm": 1.7289506095508087, + "learning_rate": 1.8975159373286843e-06, + "loss": 0.709, + "num_input_tokens_seen": 190959490, + "step": 5886 + }, + { + "epoch": 0.5309104026694323, + "flos": 28358297064000.0, + "grad_norm": 2.241520160959639, + "learning_rate": 1.8969324985793904e-06, + "loss": 0.7878, + "num_input_tokens_seen": 190990035, + "step": 5887 + }, + { + "epoch": 0.5310005861929026, + "flos": 22751470886400.0, + "grad_norm": 2.358433011198441, + "learning_rate": 1.8963490686243851e-06, + "loss": 0.7456, + "num_input_tokens_seen": 191019465, + "step": 5888 + }, + { + "epoch": 0.5310907697163728, + "flos": 24129869124480.0, + "grad_norm": 2.3232409560211114, + "learning_rate": 1.8957656475134486e-06, + "loss": 0.6893, + "num_input_tokens_seen": 191048090, + "step": 5889 + }, + { + "epoch": 0.531180953239843, + "flos": 22605816832320.0, + "grad_norm": 2.781393754651489, + "learning_rate": 1.895182235296361e-06, + "loss": 0.7288, + "num_input_tokens_seen": 191077765, + "step": 5890 + }, + { + "epoch": 0.5312711367633134, + "flos": 30835754094720.0, + "grad_norm": 1.8347390666531431, + "learning_rate": 1.8945988320229042e-06, + "loss": 0.6722, + "num_input_tokens_seen": 191106765, + "step": 5891 + }, + { + "epoch": 0.5313613202867836, + "flos": 23479964175360.0, + "grad_norm": 1.820789913064239, + "learning_rate": 1.8940154377428553e-06, + "loss": 0.83, + "num_input_tokens_seen": 191135685, + "step": 5892 + }, + { + "epoch": 0.5314515038102539, + "flos": 24969331347840.0, + "grad_norm": 1.8298474535913962, + "learning_rate": 1.8934320525059944e-06, + "loss": 0.779, + "num_input_tokens_seen": 191164570, + "step": 5893 + }, + { + "epoch": 0.5315416873337241, + "flos": 22204803561600.0, + "grad_norm": 2.651805615224287, + "learning_rate": 1.8928486763620984e-06, + "loss": 0.6349, + "num_input_tokens_seen": 191191240, + "step": 5894 + }, + { + "epoch": 0.5316318708571944, + "flos": 18525310302240.0, + "grad_norm": 2.014907354302758, + "learning_rate": 1.892265309360943e-06, + "loss": 0.7638, + "num_input_tokens_seen": 191217395, + "step": 5895 + }, + { + "epoch": 0.5317220543806647, + "flos": 28651723848480.0, + "grad_norm": 1.8136892034855945, + "learning_rate": 1.8916819515523067e-06, + "loss": 0.736, + "num_input_tokens_seen": 191246515, + "step": 5896 + }, + { + "epoch": 0.5318122379041349, + "flos": 20675175805440.0, + "grad_norm": 1.9203393274703808, + "learning_rate": 1.891098602985963e-06, + "loss": 0.6676, + "num_input_tokens_seen": 191276280, + "step": 5897 + }, + { + "epoch": 0.5319024214276051, + "flos": 27487048962240.0, + "grad_norm": 1.8977422500791818, + "learning_rate": 1.8905152637116872e-06, + "loss": 0.7733, + "num_input_tokens_seen": 191306605, + "step": 5898 + }, + { + "epoch": 0.5319926049510755, + "flos": 18524492567520.0, + "grad_norm": 2.9506116394632547, + "learning_rate": 1.8899319337792527e-06, + "loss": 0.7908, + "num_input_tokens_seen": 191332155, + "step": 5899 + }, + { + "epoch": 0.5320827884745457, + "flos": 27998287772160.0, + "grad_norm": 1.467126395321147, + "learning_rate": 1.8893486132384325e-06, + "loss": 0.7859, + "num_input_tokens_seen": 191362685, + "step": 5900 + }, + { + "epoch": 0.5321729719980159, + "flos": 20783988893760.0, + "grad_norm": 2.420857664982897, + "learning_rate": 1.888765302138999e-06, + "loss": 0.7183, + "num_input_tokens_seen": 191390425, + "step": 5901 + }, + { + "epoch": 0.5322631555214862, + "flos": 32332926916800.0, + "grad_norm": 2.5920252753249886, + "learning_rate": 1.8881820005307224e-06, + "loss": 0.693, + "num_input_tokens_seen": 191418585, + "step": 5902 + }, + { + "epoch": 0.5323533390449565, + "flos": 24276192234240.0, + "grad_norm": 1.7626132766358655, + "learning_rate": 1.8875987084633748e-06, + "loss": 0.6816, + "num_input_tokens_seen": 191446470, + "step": 5903 + }, + { + "epoch": 0.5324435225684268, + "flos": 22496334688320.0, + "grad_norm": 3.490711693980636, + "learning_rate": 1.8870154259867246e-06, + "loss": 0.7811, + "num_input_tokens_seen": 191472705, + "step": 5904 + }, + { + "epoch": 0.532533706091897, + "flos": 21294707327040.0, + "grad_norm": 2.323949051283751, + "learning_rate": 1.886432153150542e-06, + "loss": 0.6813, + "num_input_tokens_seen": 191502695, + "step": 5905 + }, + { + "epoch": 0.5326238896153672, + "flos": 15683309268960.0, + "grad_norm": 1.7548548250933238, + "learning_rate": 1.8858488900045944e-06, + "loss": 0.7274, + "num_input_tokens_seen": 191528205, + "step": 5906 + }, + { + "epoch": 0.5327140731388376, + "flos": 24854236570080.0, + "grad_norm": 2.6845404719376447, + "learning_rate": 1.885265636598648e-06, + "loss": 0.7074, + "num_input_tokens_seen": 191548370, + "step": 5907 + }, + { + "epoch": 0.5328042566623078, + "flos": 28618042312320.0, + "grad_norm": 1.7248548114877031, + "learning_rate": 1.884682392982471e-06, + "loss": 0.7827, + "num_input_tokens_seen": 191581335, + "step": 5908 + }, + { + "epoch": 0.532894440185778, + "flos": 20055086737440.0, + "grad_norm": 3.133154186917168, + "learning_rate": 1.8840991592058274e-06, + "loss": 0.8006, + "num_input_tokens_seen": 191608040, + "step": 5909 + }, + { + "epoch": 0.5329846237092484, + "flos": 22969989137280.0, + "grad_norm": 2.123900634021809, + "learning_rate": 1.8835159353184828e-06, + "loss": 0.7665, + "num_input_tokens_seen": 191637940, + "step": 5910 + }, + { + "epoch": 0.5330748072327186, + "flos": 24247900313280.0, + "grad_norm": 2.3383648868569877, + "learning_rate": 1.8829327213702013e-06, + "loss": 0.8234, + "num_input_tokens_seen": 191668125, + "step": 5911 + }, + { + "epoch": 0.5331649907561888, + "flos": 24354594725280.0, + "grad_norm": 1.755967591953359, + "learning_rate": 1.8823495174107452e-06, + "loss": 0.8189, + "num_input_tokens_seen": 191697680, + "step": 5912 + }, + { + "epoch": 0.5332551742796591, + "flos": 20420188286400.0, + "grad_norm": 3.200518333655494, + "learning_rate": 1.8817663234898773e-06, + "loss": 0.7194, + "num_input_tokens_seen": 191726410, + "step": 5913 + }, + { + "epoch": 0.5333453578031294, + "flos": 24573893541120.0, + "grad_norm": 2.4801540385368415, + "learning_rate": 1.881183139657358e-06, + "loss": 0.7563, + "num_input_tokens_seen": 191755920, + "step": 5914 + }, + { + "epoch": 0.5334355413265996, + "flos": 37324198737120.0, + "grad_norm": 1.6716792624858892, + "learning_rate": 1.8805999659629488e-06, + "loss": 0.7623, + "num_input_tokens_seen": 191784555, + "step": 5915 + }, + { + "epoch": 0.5335257248500699, + "flos": 64853255261760.0, + "grad_norm": 0.6220766363615965, + "learning_rate": 1.880016802456409e-06, + "loss": 0.5355, + "num_input_tokens_seen": 191871975, + "step": 5916 + }, + { + "epoch": 0.5336159083735401, + "flos": 28981359712800.0, + "grad_norm": 3.040450144618257, + "learning_rate": 1.8794336491874964e-06, + "loss": 0.6864, + "num_input_tokens_seen": 191898560, + "step": 5917 + }, + { + "epoch": 0.5337060918970105, + "flos": 30475670463360.0, + "grad_norm": 2.242351778505254, + "learning_rate": 1.8788505062059708e-06, + "loss": 0.7067, + "num_input_tokens_seen": 191927275, + "step": 5918 + }, + { + "epoch": 0.5337962754204807, + "flos": 24755645165760.0, + "grad_norm": 1.5683022912683273, + "learning_rate": 1.8782673735615869e-06, + "loss": 0.732, + "num_input_tokens_seen": 191956930, + "step": 5919 + }, + { + "epoch": 0.5338864589439509, + "flos": 22163502224640.0, + "grad_norm": 2.132229344498162, + "learning_rate": 1.8776842513041026e-06, + "loss": 0.7068, + "num_input_tokens_seen": 191984755, + "step": 5920 + }, + { + "epoch": 0.5339766424674212, + "flos": 27633111883680.0, + "grad_norm": 2.2430758427770994, + "learning_rate": 1.8771011394832727e-06, + "loss": 0.7176, + "num_input_tokens_seen": 192013685, + "step": 5921 + }, + { + "epoch": 0.5340668259908915, + "flos": 22747530891840.0, + "grad_norm": 1.8553584979579618, + "learning_rate": 1.8765180381488501e-06, + "loss": 0.7303, + "num_input_tokens_seen": 192042830, + "step": 5922 + }, + { + "epoch": 0.5341570095143617, + "flos": 59011327386720.0, + "grad_norm": 0.7386453931429958, + "learning_rate": 1.8759349473505905e-06, + "loss": 0.6664, + "num_input_tokens_seen": 192133790, + "step": 5923 + }, + { + "epoch": 0.534247193037832, + "flos": 20929903136160.0, + "grad_norm": 2.3906271940732373, + "learning_rate": 1.8753518671382447e-06, + "loss": 0.6691, + "num_input_tokens_seen": 192163395, + "step": 5924 + }, + { + "epoch": 0.5343373765613022, + "flos": 30801440672640.0, + "grad_norm": 1.9837128142613591, + "learning_rate": 1.8747687975615649e-06, + "loss": 0.7289, + "num_input_tokens_seen": 192194835, + "step": 5925 + }, + { + "epoch": 0.5344275600847725, + "flos": 20893656886560.0, + "grad_norm": 2.0322072047897732, + "learning_rate": 1.874185738670302e-06, + "loss": 0.6413, + "num_input_tokens_seen": 192225120, + "step": 5926 + }, + { + "epoch": 0.5345177436082428, + "flos": 26504051361120.0, + "grad_norm": 1.8433192768331277, + "learning_rate": 1.8736026905142057e-06, + "loss": 0.7342, + "num_input_tokens_seen": 192254965, + "step": 5927 + }, + { + "epoch": 0.534607927131713, + "flos": 21800259163680.0, + "grad_norm": 1.870400764339379, + "learning_rate": 1.873019653143025e-06, + "loss": 0.7231, + "num_input_tokens_seen": 192281275, + "step": 5928 + }, + { + "epoch": 0.5346981106551832, + "flos": 21586052604960.0, + "grad_norm": 2.29717154113183, + "learning_rate": 1.8724366266065069e-06, + "loss": 0.7606, + "num_input_tokens_seen": 192311590, + "step": 5929 + }, + { + "epoch": 0.5347882941786536, + "flos": 16194585248640.0, + "grad_norm": 2.497879750475041, + "learning_rate": 1.8718536109543998e-06, + "loss": 0.823, + "num_input_tokens_seen": 192337100, + "step": 5930 + }, + { + "epoch": 0.5348784777021238, + "flos": 18197904623520.0, + "grad_norm": 2.1335511657644433, + "learning_rate": 1.8712706062364485e-06, + "loss": 0.7619, + "num_input_tokens_seen": 192366460, + "step": 5931 + }, + { + "epoch": 0.534968661225594, + "flos": 25444658436000.0, + "grad_norm": 1.810497295602013, + "learning_rate": 1.8706876125024e-06, + "loss": 0.6339, + "num_input_tokens_seen": 192395415, + "step": 5932 + }, + { + "epoch": 0.5350588447490644, + "flos": 19760135992800.0, + "grad_norm": 2.680653567545485, + "learning_rate": 1.870104629801997e-06, + "loss": 0.7881, + "num_input_tokens_seen": 192421135, + "step": 5933 + }, + { + "epoch": 0.5351490282725346, + "flos": 24390915314400.0, + "grad_norm": 3.4321708291315858, + "learning_rate": 1.8695216581849823e-06, + "loss": 0.7565, + "num_input_tokens_seen": 192450530, + "step": 5934 + }, + { + "epoch": 0.5352392117960049, + "flos": 21001763749440.0, + "grad_norm": 1.7307554399315876, + "learning_rate": 1.8689386977011003e-06, + "loss": 0.7145, + "num_input_tokens_seen": 192479120, + "step": 5935 + }, + { + "epoch": 0.5353293953194751, + "flos": 64252866166560.0, + "grad_norm": 0.6689889935342173, + "learning_rate": 1.8683557484000903e-06, + "loss": 0.5747, + "num_input_tokens_seen": 192569830, + "step": 5936 + }, + { + "epoch": 0.5354195788429454, + "flos": 23079211092960.0, + "grad_norm": 3.3700471493283723, + "learning_rate": 1.8677728103316947e-06, + "loss": 0.7828, + "num_input_tokens_seen": 192598085, + "step": 5937 + }, + { + "epoch": 0.5355097623664157, + "flos": 25006469671680.0, + "grad_norm": 1.506673352907474, + "learning_rate": 1.8671898835456518e-06, + "loss": 0.6979, + "num_input_tokens_seen": 192629650, + "step": 5938 + }, + { + "epoch": 0.5355999458898859, + "flos": 17942545406880.0, + "grad_norm": 1.8710024498822042, + "learning_rate": 1.8666069680917003e-06, + "loss": 0.8483, + "num_input_tokens_seen": 192656225, + "step": 5939 + }, + { + "epoch": 0.5356901294133561, + "flos": 20668187890560.0, + "grad_norm": 2.3427713003780894, + "learning_rate": 1.8660240640195775e-06, + "loss": 0.7019, + "num_input_tokens_seen": 192682580, + "step": 5940 + }, + { + "epoch": 0.5357803129368265, + "flos": 21292254122880.0, + "grad_norm": 1.6670073478916656, + "learning_rate": 1.8654411713790203e-06, + "loss": 0.7824, + "num_input_tokens_seen": 192711310, + "step": 5941 + }, + { + "epoch": 0.5358704964602967, + "flos": 23079359772000.0, + "grad_norm": 2.306327624164735, + "learning_rate": 1.8648582902197648e-06, + "loss": 0.7675, + "num_input_tokens_seen": 192740730, + "step": 5942 + }, + { + "epoch": 0.535960679983767, + "flos": 33644779817280.0, + "grad_norm": 2.302402307382505, + "learning_rate": 1.8642754205915452e-06, + "loss": 0.7504, + "num_input_tokens_seen": 192770135, + "step": 5943 + }, + { + "epoch": 0.5360508635072372, + "flos": 21987140215200.0, + "grad_norm": 2.3922192611928, + "learning_rate": 1.8636925625440943e-06, + "loss": 0.6938, + "num_input_tokens_seen": 192799660, + "step": 5944 + }, + { + "epoch": 0.5361410470307075, + "flos": 23808187588800.0, + "grad_norm": 2.189932097858562, + "learning_rate": 1.863109716127146e-06, + "loss": 0.6556, + "num_input_tokens_seen": 192827435, + "step": 5945 + }, + { + "epoch": 0.5362312305541778, + "flos": 23006681424000.0, + "grad_norm": 1.7099693729491034, + "learning_rate": 1.8625268813904311e-06, + "loss": 0.7891, + "num_input_tokens_seen": 192855275, + "step": 5946 + }, + { + "epoch": 0.536321414077648, + "flos": 19175772797760.0, + "grad_norm": 1.7868442693137925, + "learning_rate": 1.8619440583836814e-06, + "loss": 0.774, + "num_input_tokens_seen": 192881490, + "step": 5947 + }, + { + "epoch": 0.5364115976011182, + "flos": 22606225699680.0, + "grad_norm": 2.5187139087296755, + "learning_rate": 1.8613612471566249e-06, + "loss": 0.7524, + "num_input_tokens_seen": 192909780, + "step": 5948 + }, + { + "epoch": 0.5365017811245886, + "flos": 14663916739200.0, + "grad_norm": 2.258474074854425, + "learning_rate": 1.8607784477589922e-06, + "loss": 0.6937, + "num_input_tokens_seen": 192936325, + "step": 5949 + }, + { + "epoch": 0.5365919646480588, + "flos": 23589818016960.0, + "grad_norm": 1.9335272267005221, + "learning_rate": 1.8601956602405103e-06, + "loss": 0.7636, + "num_input_tokens_seen": 192964160, + "step": 5950 + }, + { + "epoch": 0.536682148171529, + "flos": 25007733443520.0, + "grad_norm": 1.7470716089421894, + "learning_rate": 1.8596128846509043e-06, + "loss": 0.7895, + "num_input_tokens_seen": 192993450, + "step": 5951 + }, + { + "epoch": 0.5367723316949993, + "flos": 22970063476800.0, + "grad_norm": 1.8000305994399894, + "learning_rate": 1.859030121039902e-06, + "loss": 0.7037, + "num_input_tokens_seen": 193020640, + "step": 5952 + }, + { + "epoch": 0.5368625152184696, + "flos": 29198539852320.0, + "grad_norm": 1.722547739638014, + "learning_rate": 1.8584473694572268e-06, + "loss": 0.7007, + "num_input_tokens_seen": 193051210, + "step": 5953 + }, + { + "epoch": 0.5369526987419398, + "flos": 21840631256640.0, + "grad_norm": 2.050155689561266, + "learning_rate": 1.8578646299526026e-06, + "loss": 0.7692, + "num_input_tokens_seen": 193080070, + "step": 5954 + }, + { + "epoch": 0.5370428822654101, + "flos": 20453089257600.0, + "grad_norm": 1.8828544770306053, + "learning_rate": 1.8572819025757518e-06, + "loss": 0.7199, + "num_input_tokens_seen": 193106930, + "step": 5955 + }, + { + "epoch": 0.5371330657888804, + "flos": 25265583034080.0, + "grad_norm": 2.2684519001815975, + "learning_rate": 1.8566991873763959e-06, + "loss": 0.7612, + "num_input_tokens_seen": 193137240, + "step": 5956 + }, + { + "epoch": 0.5372232493123507, + "flos": 18926100554400.0, + "grad_norm": 1.892170046183243, + "learning_rate": 1.856116484404256e-06, + "loss": 0.697, + "num_input_tokens_seen": 193165200, + "step": 5957 + }, + { + "epoch": 0.5373134328358209, + "flos": 31604991174240.0, + "grad_norm": 2.1585789141868252, + "learning_rate": 1.8555337937090502e-06, + "loss": 0.7132, + "num_input_tokens_seen": 193192835, + "step": 5958 + }, + { + "epoch": 0.5374036163592911, + "flos": 13568760771360.0, + "grad_norm": 5.032567959292013, + "learning_rate": 1.8549511153404984e-06, + "loss": 0.724, + "num_input_tokens_seen": 193216285, + "step": 5959 + }, + { + "epoch": 0.5374937998827615, + "flos": 20416173952320.0, + "grad_norm": 1.7756888597573224, + "learning_rate": 1.854368449348317e-06, + "loss": 0.8552, + "num_input_tokens_seen": 193241420, + "step": 5960 + }, + { + "epoch": 0.5375839834062317, + "flos": 19800396576480.0, + "grad_norm": 3.2994251964383268, + "learning_rate": 1.853785795782222e-06, + "loss": 0.7055, + "num_input_tokens_seen": 193268240, + "step": 5961 + }, + { + "epoch": 0.5376741669297019, + "flos": 26795917015680.0, + "grad_norm": 2.630703678635737, + "learning_rate": 1.85320315469193e-06, + "loss": 0.6131, + "num_input_tokens_seen": 193297340, + "step": 5962 + }, + { + "epoch": 0.5377643504531722, + "flos": 69964268079840.0, + "grad_norm": 0.6486461717286051, + "learning_rate": 1.8526205261271534e-06, + "loss": 0.5809, + "num_input_tokens_seen": 193381980, + "step": 5963 + }, + { + "epoch": 0.5378545339766425, + "flos": 35575755372000.0, + "grad_norm": 1.8892457202692787, + "learning_rate": 1.852037910137607e-06, + "loss": 0.6564, + "num_input_tokens_seen": 193414715, + "step": 5964 + }, + { + "epoch": 0.5379447175001127, + "flos": 25083831409440.0, + "grad_norm": 2.430999990126666, + "learning_rate": 1.851455306773002e-06, + "loss": 0.8315, + "num_input_tokens_seen": 193446700, + "step": 5965 + }, + { + "epoch": 0.538034901023583, + "flos": 31390487257440.0, + "grad_norm": 1.5215617051262542, + "learning_rate": 1.8508727160830483e-06, + "loss": 0.7218, + "num_input_tokens_seen": 193483375, + "step": 5966 + }, + { + "epoch": 0.5381250845470532, + "flos": 25630684583040.0, + "grad_norm": 1.7920845860247243, + "learning_rate": 1.8502901381174575e-06, + "loss": 0.8144, + "num_input_tokens_seen": 193512495, + "step": 5967 + }, + { + "epoch": 0.5382152680705236, + "flos": 26426838302400.0, + "grad_norm": 2.22187209417526, + "learning_rate": 1.8497075729259372e-06, + "loss": 0.6511, + "num_input_tokens_seen": 193541895, + "step": 5968 + }, + { + "epoch": 0.5383054515939938, + "flos": 58208854808160.0, + "grad_norm": 0.7627836950380852, + "learning_rate": 1.8491250205581963e-06, + "loss": 0.6408, + "num_input_tokens_seen": 193624570, + "step": 5969 + }, + { + "epoch": 0.538395635117464, + "flos": 29674275807840.0, + "grad_norm": 1.674680619607392, + "learning_rate": 1.8485424810639393e-06, + "loss": 0.7539, + "num_input_tokens_seen": 193656905, + "step": 5970 + }, + { + "epoch": 0.5384858186409343, + "flos": 22387335751200.0, + "grad_norm": 2.192060675456501, + "learning_rate": 1.847959954492874e-06, + "loss": 0.8059, + "num_input_tokens_seen": 193684725, + "step": 5971 + }, + { + "epoch": 0.5385760021644046, + "flos": 22168111274880.0, + "grad_norm": 7.7118944619741105, + "learning_rate": 1.8473774408947035e-06, + "loss": 0.8412, + "num_input_tokens_seen": 193710185, + "step": 5972 + }, + { + "epoch": 0.5386661856878748, + "flos": 67188589365600.0, + "grad_norm": 0.6911959139615941, + "learning_rate": 1.8467949403191308e-06, + "loss": 0.5527, + "num_input_tokens_seen": 193793105, + "step": 5973 + }, + { + "epoch": 0.5387563692113451, + "flos": 34847447931840.0, + "grad_norm": 1.9584591466651773, + "learning_rate": 1.8462124528158592e-06, + "loss": 0.8035, + "num_input_tokens_seen": 193823820, + "step": 5974 + }, + { + "epoch": 0.5388465527348153, + "flos": 23149250388000.0, + "grad_norm": 1.72176076575877, + "learning_rate": 1.8456299784345881e-06, + "loss": 0.7385, + "num_input_tokens_seen": 193854720, + "step": 5975 + }, + { + "epoch": 0.5389367362582856, + "flos": 22934300434080.0, + "grad_norm": 1.9621020502815327, + "learning_rate": 1.8450475172250194e-06, + "loss": 0.7676, + "num_input_tokens_seen": 193882220, + "step": 5976 + }, + { + "epoch": 0.5390269197817559, + "flos": 26325905203200.0, + "grad_norm": 1.6473285868360332, + "learning_rate": 1.844465069236851e-06, + "loss": 0.7272, + "num_input_tokens_seen": 193913945, + "step": 5977 + }, + { + "epoch": 0.5391171033052261, + "flos": 28068252727680.0, + "grad_norm": 2.490698103268635, + "learning_rate": 1.8438826345197796e-06, + "loss": 0.6541, + "num_input_tokens_seen": 193946220, + "step": 5978 + }, + { + "epoch": 0.5392072868286965, + "flos": 21986173801440.0, + "grad_norm": 2.9616992946781697, + "learning_rate": 1.8433002131235036e-06, + "loss": 0.7387, + "num_input_tokens_seen": 193972545, + "step": 5979 + }, + { + "epoch": 0.5392974703521667, + "flos": 30580840915200.0, + "grad_norm": 1.6561406657281093, + "learning_rate": 1.8427178050977167e-06, + "loss": 0.7672, + "num_input_tokens_seen": 194003540, + "step": 5980 + }, + { + "epoch": 0.5393876538756369, + "flos": 19362579509760.0, + "grad_norm": 2.5248210881366644, + "learning_rate": 1.8421354104921143e-06, + "loss": 0.8255, + "num_input_tokens_seen": 194031560, + "step": 5981 + }, + { + "epoch": 0.5394778373991072, + "flos": 69787125505440.0, + "grad_norm": 0.6034870819946857, + "learning_rate": 1.8415530293563894e-06, + "loss": 0.5487, + "num_input_tokens_seen": 194124320, + "step": 5982 + }, + { + "epoch": 0.5395680209225775, + "flos": 41368310338560.0, + "grad_norm": 1.7479059769944267, + "learning_rate": 1.8409706617402333e-06, + "loss": 0.729, + "num_input_tokens_seen": 194156910, + "step": 5983 + }, + { + "epoch": 0.5396582044460477, + "flos": 34406768793600.0, + "grad_norm": 1.6662957335003061, + "learning_rate": 1.8403883076933378e-06, + "loss": 0.7702, + "num_input_tokens_seen": 194190275, + "step": 5984 + }, + { + "epoch": 0.539748387969518, + "flos": 24681777385440.0, + "grad_norm": 1.7893471970231911, + "learning_rate": 1.839805967265391e-06, + "loss": 0.7381, + "num_input_tokens_seen": 194222535, + "step": 5985 + }, + { + "epoch": 0.5398385714929882, + "flos": 20966149385760.0, + "grad_norm": 2.4374019896536314, + "learning_rate": 1.839223640506083e-06, + "loss": 0.7998, + "num_input_tokens_seen": 194251750, + "step": 5986 + }, + { + "epoch": 0.5399287550164585, + "flos": 25775223544320.0, + "grad_norm": 2.8594575463558853, + "learning_rate": 1.8386413274650998e-06, + "loss": 0.7412, + "num_input_tokens_seen": 194279770, + "step": 5987 + }, + { + "epoch": 0.5400189385399288, + "flos": 22563400402560.0, + "grad_norm": 1.8563552184419725, + "learning_rate": 1.8380590281921294e-06, + "loss": 0.7965, + "num_input_tokens_seen": 194308025, + "step": 5988 + }, + { + "epoch": 0.540109122063399, + "flos": 34152933537120.0, + "grad_norm": 1.7112562326856957, + "learning_rate": 1.8374767427368552e-06, + "loss": 0.6552, + "num_input_tokens_seen": 194339895, + "step": 5989 + }, + { + "epoch": 0.5401993055868692, + "flos": 17614247653920.0, + "grad_norm": 2.8522292938296667, + "learning_rate": 1.8368944711489608e-06, + "loss": 0.7748, + "num_input_tokens_seen": 194367055, + "step": 5990 + }, + { + "epoch": 0.5402894891103396, + "flos": 20929754457120.0, + "grad_norm": 2.1134407702472435, + "learning_rate": 1.8363122134781304e-06, + "loss": 0.7755, + "num_input_tokens_seen": 194394465, + "step": 5991 + }, + { + "epoch": 0.5403796726338098, + "flos": 14809013246880.0, + "grad_norm": 2.5042822402290854, + "learning_rate": 1.835729969774044e-06, + "loss": 0.7737, + "num_input_tokens_seen": 194421230, + "step": 5992 + }, + { + "epoch": 0.54046985615728, + "flos": 21622633382400.0, + "grad_norm": 1.6836502525955874, + "learning_rate": 1.8351477400863823e-06, + "loss": 0.7557, + "num_input_tokens_seen": 194448955, + "step": 5993 + }, + { + "epoch": 0.5405600396807503, + "flos": 20054863718880.0, + "grad_norm": 2.0767765115717465, + "learning_rate": 1.8345655244648249e-06, + "loss": 0.7291, + "num_input_tokens_seen": 194478075, + "step": 5994 + }, + { + "epoch": 0.5406502232042206, + "flos": 46867512860160.0, + "grad_norm": 1.4309300094111856, + "learning_rate": 1.8339833229590486e-06, + "loss": 0.739, + "num_input_tokens_seen": 194510025, + "step": 5995 + }, + { + "epoch": 0.5407404067276909, + "flos": 23368363355040.0, + "grad_norm": 2.5777041929894082, + "learning_rate": 1.833401135618731e-06, + "loss": 0.8178, + "num_input_tokens_seen": 194537835, + "step": 5996 + }, + { + "epoch": 0.5408305902511611, + "flos": 16264959071520.0, + "grad_norm": 1.8979259790833434, + "learning_rate": 1.8328189624935466e-06, + "loss": 0.6752, + "num_input_tokens_seen": 194565275, + "step": 5997 + }, + { + "epoch": 0.5409207737746313, + "flos": 26794913432160.0, + "grad_norm": 2.39025517015284, + "learning_rate": 1.832236803633171e-06, + "loss": 0.7848, + "num_input_tokens_seen": 194596925, + "step": 5998 + }, + { + "epoch": 0.5410109572981017, + "flos": 25228556219520.0, + "grad_norm": 2.1055192534382394, + "learning_rate": 1.831654659087276e-06, + "loss": 0.7144, + "num_input_tokens_seen": 194624705, + "step": 5999 + }, + { + "epoch": 0.5411011408215719, + "flos": 20961577505280.0, + "grad_norm": 2.2012450943068473, + "learning_rate": 1.831072528905533e-06, + "loss": 0.763, + "num_input_tokens_seen": 194651895, + "step": 6000 + }, + { + "epoch": 0.5411913243450421, + "flos": 25628640246240.0, + "grad_norm": 1.6435361918940368, + "learning_rate": 1.8304904131376142e-06, + "loss": 0.7995, + "num_input_tokens_seen": 194681195, + "step": 6001 + }, + { + "epoch": 0.5412815078685124, + "flos": 20966446743840.0, + "grad_norm": 1.9801266466924605, + "learning_rate": 1.8299083118331874e-06, + "loss": 0.7664, + "num_input_tokens_seen": 194709105, + "step": 6002 + }, + { + "epoch": 0.5413716913919827, + "flos": 24026780179200.0, + "grad_norm": 2.1409788187360967, + "learning_rate": 1.8293262250419217e-06, + "loss": 0.8819, + "num_input_tokens_seen": 194735195, + "step": 6003 + }, + { + "epoch": 0.541461874915453, + "flos": 30110271556320.0, + "grad_norm": 2.0273001193899165, + "learning_rate": 1.828744152813484e-06, + "loss": 0.8616, + "num_input_tokens_seen": 194764470, + "step": 6004 + }, + { + "epoch": 0.5415520584389232, + "flos": 24536160501120.0, + "grad_norm": 1.8799420394485014, + "learning_rate": 1.8281620951975382e-06, + "loss": 0.7079, + "num_input_tokens_seen": 194793735, + "step": 6005 + }, + { + "epoch": 0.5416422419623935, + "flos": 20417846591520.0, + "grad_norm": 3.202295261005716, + "learning_rate": 1.827580052243751e-06, + "loss": 0.7095, + "num_input_tokens_seen": 194819605, + "step": 6006 + }, + { + "epoch": 0.5417324254858638, + "flos": 25702842554400.0, + "grad_norm": 2.4069078215067705, + "learning_rate": 1.826998024001784e-06, + "loss": 0.7804, + "num_input_tokens_seen": 194853410, + "step": 6007 + }, + { + "epoch": 0.541822609009334, + "flos": 18847995421440.0, + "grad_norm": 1.8585922773563024, + "learning_rate": 1.8264160105212995e-06, + "loss": 0.7766, + "num_input_tokens_seen": 194880640, + "step": 6008 + }, + { + "epoch": 0.5419127925328042, + "flos": 19873781149920.0, + "grad_norm": 2.838232535698366, + "learning_rate": 1.8258340118519582e-06, + "loss": 0.7166, + "num_input_tokens_seen": 194908395, + "step": 6009 + }, + { + "epoch": 0.5420029760562746, + "flos": 25149596182080.0, + "grad_norm": 2.1828332439903275, + "learning_rate": 1.82525202804342e-06, + "loss": 0.7342, + "num_input_tokens_seen": 194939175, + "step": 6010 + }, + { + "epoch": 0.5420931595797448, + "flos": 23006346896160.0, + "grad_norm": 2.2049343173738998, + "learning_rate": 1.8246700591453415e-06, + "loss": 0.7267, + "num_input_tokens_seen": 194965035, + "step": 6011 + }, + { + "epoch": 0.542183343103215, + "flos": 23737442068320.0, + "grad_norm": 2.3644273233132425, + "learning_rate": 1.8240881052073801e-06, + "loss": 0.7518, + "num_input_tokens_seen": 194998590, + "step": 6012 + }, + { + "epoch": 0.5422735266266853, + "flos": 21804161988480.0, + "grad_norm": 2.561987383672187, + "learning_rate": 1.8235061662791923e-06, + "loss": 0.717, + "num_input_tokens_seen": 195027705, + "step": 6013 + }, + { + "epoch": 0.5423637101501556, + "flos": 23545134231840.0, + "grad_norm": 2.571873709584465, + "learning_rate": 1.8229242424104309e-06, + "loss": 0.8194, + "num_input_tokens_seen": 195053665, + "step": 6014 + }, + { + "epoch": 0.5424538936736258, + "flos": 59909566467840.0, + "grad_norm": 0.613904478401158, + "learning_rate": 1.8223423336507503e-06, + "loss": 0.5592, + "num_input_tokens_seen": 195151325, + "step": 6015 + }, + { + "epoch": 0.5425440771970961, + "flos": 25261420020960.0, + "grad_norm": 2.6438207305753396, + "learning_rate": 1.8217604400498012e-06, + "loss": 0.7362, + "num_input_tokens_seen": 195179280, + "step": 6016 + }, + { + "epoch": 0.5426342607205663, + "flos": 23114193570720.0, + "grad_norm": 1.6351773599208825, + "learning_rate": 1.8211785616572333e-06, + "loss": 0.7961, + "num_input_tokens_seen": 195210115, + "step": 6017 + }, + { + "epoch": 0.5427244442440367, + "flos": 21075594360000.0, + "grad_norm": 1.993000178909868, + "learning_rate": 1.8205966985226975e-06, + "loss": 0.8158, + "num_input_tokens_seen": 195236225, + "step": 6018 + }, + { + "epoch": 0.5428146277675069, + "flos": 24099755885280.0, + "grad_norm": 2.303110805348685, + "learning_rate": 1.8200148506958397e-06, + "loss": 0.7605, + "num_input_tokens_seen": 195267200, + "step": 6019 + }, + { + "epoch": 0.5429048112909771, + "flos": 67157584052160.0, + "grad_norm": 0.6500258468110677, + "learning_rate": 1.819433018226308e-06, + "loss": 0.6053, + "num_input_tokens_seen": 195353450, + "step": 6020 + }, + { + "epoch": 0.5429949948144474, + "flos": 68585163616320.0, + "grad_norm": 0.6004625643417267, + "learning_rate": 1.8188512011637471e-06, + "loss": 0.5497, + "num_input_tokens_seen": 195449550, + "step": 6021 + }, + { + "epoch": 0.5430851783379177, + "flos": 28798827523200.0, + "grad_norm": 2.1418671728101897, + "learning_rate": 1.8182693995578e-06, + "loss": 0.7251, + "num_input_tokens_seen": 195477675, + "step": 6022 + }, + { + "epoch": 0.5431753618613879, + "flos": 23006012368320.0, + "grad_norm": 2.288397995252153, + "learning_rate": 1.8176876134581098e-06, + "loss": 0.6291, + "num_input_tokens_seen": 195509890, + "step": 6023 + }, + { + "epoch": 0.5432655453848582, + "flos": 25301866453440.0, + "grad_norm": 4.2415188706785525, + "learning_rate": 1.8171058429143176e-06, + "loss": 0.6967, + "num_input_tokens_seen": 195539145, + "step": 6024 + }, + { + "epoch": 0.5433557289083284, + "flos": 26606359741440.0, + "grad_norm": 2.4111912512852767, + "learning_rate": 1.8165240879760637e-06, + "loss": 0.7863, + "num_input_tokens_seen": 195568045, + "step": 6025 + }, + { + "epoch": 0.5434459124317987, + "flos": 28690051604640.0, + "grad_norm": 2.1042540377867858, + "learning_rate": 1.8159423486929862e-06, + "loss": 0.7785, + "num_input_tokens_seen": 195597845, + "step": 6026 + }, + { + "epoch": 0.543536095955269, + "flos": 20857224788160.0, + "grad_norm": 4.820500449005001, + "learning_rate": 1.815360625114722e-06, + "loss": 0.7502, + "num_input_tokens_seen": 195625660, + "step": 6027 + }, + { + "epoch": 0.5436262794787392, + "flos": 23371002408000.0, + "grad_norm": 1.7655451258594108, + "learning_rate": 1.814778917290908e-06, + "loss": 0.7366, + "num_input_tokens_seen": 195654850, + "step": 6028 + }, + { + "epoch": 0.5437164630022095, + "flos": 14663507871840.0, + "grad_norm": 16.44800953982778, + "learning_rate": 1.8141972252711773e-06, + "loss": 0.7652, + "num_input_tokens_seen": 195680435, + "step": 6029 + }, + { + "epoch": 0.5438066465256798, + "flos": 18888887891040.0, + "grad_norm": 2.408758663999281, + "learning_rate": 1.8136155491051645e-06, + "loss": 0.7475, + "num_input_tokens_seen": 195709115, + "step": 6030 + }, + { + "epoch": 0.54389683004915, + "flos": 58569904853280.0, + "grad_norm": 0.6662497458537978, + "learning_rate": 1.8130338888424998e-06, + "loss": 0.5933, + "num_input_tokens_seen": 195798145, + "step": 6031 + }, + { + "epoch": 0.5439870135726202, + "flos": 18453486858720.0, + "grad_norm": 2.032046587342397, + "learning_rate": 1.812452244532816e-06, + "loss": 0.7667, + "num_input_tokens_seen": 195825770, + "step": 6032 + }, + { + "epoch": 0.5440771970960906, + "flos": 28726335024000.0, + "grad_norm": 1.906061194403184, + "learning_rate": 1.8118706162257405e-06, + "loss": 0.6867, + "num_input_tokens_seen": 195855395, + "step": 6033 + }, + { + "epoch": 0.5441673806195608, + "flos": 22494253181760.0, + "grad_norm": 2.2323305606939696, + "learning_rate": 1.8112890039709002e-06, + "loss": 0.8308, + "num_input_tokens_seen": 195883610, + "step": 6034 + }, + { + "epoch": 0.5442575641430311, + "flos": 25557114160800.0, + "grad_norm": 3.6102221266864056, + "learning_rate": 1.8107074078179238e-06, + "loss": 0.7552, + "num_input_tokens_seen": 195912870, + "step": 6035 + }, + { + "epoch": 0.5443477476665013, + "flos": 34443275231520.0, + "grad_norm": 1.7714569863554255, + "learning_rate": 1.8101258278164348e-06, + "loss": 0.7022, + "num_input_tokens_seen": 195944340, + "step": 6036 + }, + { + "epoch": 0.5444379311899716, + "flos": 25301829283680.0, + "grad_norm": 1.8123117546455993, + "learning_rate": 1.8095442640160575e-06, + "loss": 0.8586, + "num_input_tokens_seen": 195972235, + "step": 6037 + }, + { + "epoch": 0.5445281147134419, + "flos": 20964774104640.0, + "grad_norm": 3.3139948800691457, + "learning_rate": 1.8089627164664132e-06, + "loss": 0.7134, + "num_input_tokens_seen": 195998100, + "step": 6038 + }, + { + "epoch": 0.5446182982369121, + "flos": 23623796911200.0, + "grad_norm": 1.5466665262326362, + "learning_rate": 1.8083811852171233e-06, + "loss": 0.6688, + "num_input_tokens_seen": 196028170, + "step": 6039 + }, + { + "epoch": 0.5447084817603823, + "flos": 20493535690080.0, + "grad_norm": 1.8194273405104284, + "learning_rate": 1.8077996703178078e-06, + "loss": 0.75, + "num_input_tokens_seen": 196058535, + "step": 6040 + }, + { + "epoch": 0.5447986652838527, + "flos": 21440138362560.0, + "grad_norm": 2.3590817300788656, + "learning_rate": 1.8072181718180833e-06, + "loss": 0.6931, + "num_input_tokens_seen": 196087960, + "step": 6041 + }, + { + "epoch": 0.5448888488073229, + "flos": 30256780514880.0, + "grad_norm": 2.6360731450642927, + "learning_rate": 1.806636689767568e-06, + "loss": 0.8342, + "num_input_tokens_seen": 196118610, + "step": 6042 + }, + { + "epoch": 0.5449790323307931, + "flos": 28580234932800.0, + "grad_norm": 1.8078526076184427, + "learning_rate": 1.8060552242158769e-06, + "loss": 0.6892, + "num_input_tokens_seen": 196149430, + "step": 6043 + }, + { + "epoch": 0.5450692158542634, + "flos": 26140325093280.0, + "grad_norm": 1.8320710720785296, + "learning_rate": 1.8054737752126224e-06, + "loss": 0.7215, + "num_input_tokens_seen": 196180640, + "step": 6044 + }, + { + "epoch": 0.5451593993777337, + "flos": 20674469580000.0, + "grad_norm": 2.289236537319289, + "learning_rate": 1.804892342807419e-06, + "loss": 0.7352, + "num_input_tokens_seen": 196209780, + "step": 6045 + }, + { + "epoch": 0.545249582901204, + "flos": 22891103439360.0, + "grad_norm": 3.8090935168895705, + "learning_rate": 1.8043109270498756e-06, + "loss": 0.8173, + "num_input_tokens_seen": 196235270, + "step": 6046 + }, + { + "epoch": 0.5453397664246742, + "flos": 24203848414080.0, + "grad_norm": 2.018056186933507, + "learning_rate": 1.803729527989604e-06, + "loss": 0.7598, + "num_input_tokens_seen": 196266940, + "step": 6047 + }, + { + "epoch": 0.5454299499481444, + "flos": 22132757099520.0, + "grad_norm": 1.837764723516276, + "learning_rate": 1.8031481456762112e-06, + "loss": 0.7405, + "num_input_tokens_seen": 196294550, + "step": 6048 + }, + { + "epoch": 0.5455201334716148, + "flos": 20674989956640.0, + "grad_norm": 1.9365585870289295, + "learning_rate": 1.8025667801593033e-06, + "loss": 0.6788, + "num_input_tokens_seen": 196323905, + "step": 6049 + }, + { + "epoch": 0.545610316995085, + "flos": 27408757980480.0, + "grad_norm": 2.8502063609585937, + "learning_rate": 1.8019854314884871e-06, + "loss": 0.6866, + "num_input_tokens_seen": 196353455, + "step": 6050 + }, + { + "epoch": 0.5457005005185552, + "flos": 28943886861120.0, + "grad_norm": 1.6751734202554367, + "learning_rate": 1.8014040997133652e-06, + "loss": 0.803, + "num_input_tokens_seen": 196385680, + "step": 6051 + }, + { + "epoch": 0.5457906840420256, + "flos": 68828293981920.0, + "grad_norm": 0.7584141967613621, + "learning_rate": 1.8008227848835414e-06, + "loss": 0.6137, + "num_input_tokens_seen": 196474990, + "step": 6052 + }, + { + "epoch": 0.5458808675654958, + "flos": 18452446105440.0, + "grad_norm": 2.5092962041243676, + "learning_rate": 1.8002414870486144e-06, + "loss": 0.7053, + "num_input_tokens_seen": 196502320, + "step": 6053 + }, + { + "epoch": 0.545971051088966, + "flos": 25663957251840.0, + "grad_norm": 1.8981623823085332, + "learning_rate": 1.7996602062581864e-06, + "loss": 0.7019, + "num_input_tokens_seen": 196534070, + "step": 6054 + }, + { + "epoch": 0.5460612346124363, + "flos": 19800062048640.0, + "grad_norm": 2.0567493609224203, + "learning_rate": 1.7990789425618544e-06, + "loss": 0.7972, + "num_input_tokens_seen": 196561130, + "step": 6055 + }, + { + "epoch": 0.5461514181359066, + "flos": 53718085470720.0, + "grad_norm": 1.5398695325700962, + "learning_rate": 1.7984976960092137e-06, + "loss": 0.7322, + "num_input_tokens_seen": 196595535, + "step": 6056 + }, + { + "epoch": 0.5462416016593769, + "flos": 29815209302400.0, + "grad_norm": 2.2142994260583158, + "learning_rate": 1.7979164666498617e-06, + "loss": 0.7153, + "num_input_tokens_seen": 196624955, + "step": 6057 + }, + { + "epoch": 0.5463317851828471, + "flos": 64840017093600.0, + "grad_norm": 0.6168461350525752, + "learning_rate": 1.7973352545333901e-06, + "loss": 0.5527, + "num_input_tokens_seen": 196714710, + "step": 6058 + }, + { + "epoch": 0.5464219687063173, + "flos": 34008097217760.0, + "grad_norm": 1.9133716815778408, + "learning_rate": 1.796754059709393e-06, + "loss": 0.667, + "num_input_tokens_seen": 196742230, + "step": 6059 + }, + { + "epoch": 0.5465121522297877, + "flos": 27232730498880.0, + "grad_norm": 2.7598757711731063, + "learning_rate": 1.7961728822274603e-06, + "loss": 0.8121, + "num_input_tokens_seen": 196770930, + "step": 6060 + }, + { + "epoch": 0.5466023357532579, + "flos": 25630164206400.0, + "grad_norm": 3.591707926525732, + "learning_rate": 1.7955917221371802e-06, + "loss": 0.7031, + "num_input_tokens_seen": 196799180, + "step": 6061 + }, + { + "epoch": 0.5466925192767281, + "flos": 24901224880320.0, + "grad_norm": 3.824753253432958, + "learning_rate": 1.7950105794881422e-06, + "loss": 0.6645, + "num_input_tokens_seen": 196827525, + "step": 6062 + }, + { + "epoch": 0.5467827028001984, + "flos": 29272965179040.0, + "grad_norm": 1.7481229115576982, + "learning_rate": 1.7944294543299317e-06, + "loss": 0.7315, + "num_input_tokens_seen": 196859825, + "step": 6063 + }, + { + "epoch": 0.5468728863236687, + "flos": 41878619904480.0, + "grad_norm": 1.8480815760826357, + "learning_rate": 1.7938483467121333e-06, + "loss": 0.8017, + "num_input_tokens_seen": 196891305, + "step": 6064 + }, + { + "epoch": 0.5469630698471389, + "flos": 17760347745120.0, + "grad_norm": 1.9871472195560476, + "learning_rate": 1.7932672566843313e-06, + "loss": 0.7315, + "num_input_tokens_seen": 196918535, + "step": 6065 + }, + { + "epoch": 0.5470532533706092, + "flos": 22897868335680.0, + "grad_norm": 2.29556683506902, + "learning_rate": 1.7926861842961065e-06, + "loss": 0.672, + "num_input_tokens_seen": 196947370, + "step": 6066 + }, + { + "epoch": 0.5471434368940794, + "flos": 26175679268640.0, + "grad_norm": 1.7272751662132164, + "learning_rate": 1.7921051295970399e-06, + "loss": 0.6691, + "num_input_tokens_seen": 196975910, + "step": 6067 + }, + { + "epoch": 0.5472336204175497, + "flos": 34986411429120.0, + "grad_norm": 1.793136090217405, + "learning_rate": 1.7915240926367092e-06, + "loss": 0.7165, + "num_input_tokens_seen": 197006610, + "step": 6068 + }, + { + "epoch": 0.54732380394102, + "flos": 68431449457920.0, + "grad_norm": 1.727335070894919, + "learning_rate": 1.7909430734646932e-06, + "loss": 0.7925, + "num_input_tokens_seen": 197040270, + "step": 6069 + }, + { + "epoch": 0.5474139874644902, + "flos": 63313214239200.0, + "grad_norm": 0.6800941916752974, + "learning_rate": 1.790362072130567e-06, + "loss": 0.5855, + "num_input_tokens_seen": 197133065, + "step": 6070 + }, + { + "epoch": 0.5475041709879604, + "flos": 24245893146240.0, + "grad_norm": 2.8215029415241695, + "learning_rate": 1.7897810886839037e-06, + "loss": 0.7324, + "num_input_tokens_seen": 197161675, + "step": 6071 + }, + { + "epoch": 0.5475943545114308, + "flos": 31893623059680.0, + "grad_norm": 3.44924709010335, + "learning_rate": 1.7892001231742782e-06, + "loss": 0.7575, + "num_input_tokens_seen": 197192245, + "step": 6072 + }, + { + "epoch": 0.547684538034901, + "flos": 26468288318400.0, + "grad_norm": 1.6642973111822734, + "learning_rate": 1.7886191756512594e-06, + "loss": 0.7574, + "num_input_tokens_seen": 197219660, + "step": 6073 + }, + { + "epoch": 0.5477747215583713, + "flos": 23915699735520.0, + "grad_norm": 2.1302201722304983, + "learning_rate": 1.7880382461644192e-06, + "loss": 0.828, + "num_input_tokens_seen": 197248955, + "step": 6074 + }, + { + "epoch": 0.5478649050818416, + "flos": 26977854489120.0, + "grad_norm": 2.182280061701567, + "learning_rate": 1.7874573347633235e-06, + "loss": 0.7268, + "num_input_tokens_seen": 197275155, + "step": 6075 + }, + { + "epoch": 0.5479550886053118, + "flos": 26431075655040.0, + "grad_norm": 1.9547169415568837, + "learning_rate": 1.7868764414975408e-06, + "loss": 0.8185, + "num_input_tokens_seen": 197305610, + "step": 6076 + }, + { + "epoch": 0.5480452721287821, + "flos": 23363419776960.0, + "grad_norm": 1.963768727524669, + "learning_rate": 1.7862955664166353e-06, + "loss": 0.6955, + "num_input_tokens_seen": 197332105, + "step": 6077 + }, + { + "epoch": 0.5481354556522523, + "flos": 22640873649600.0, + "grad_norm": 1.9120381140361657, + "learning_rate": 1.78571470957017e-06, + "loss": 0.8325, + "num_input_tokens_seen": 197361755, + "step": 6078 + }, + { + "epoch": 0.5482256391757226, + "flos": 19727235021600.0, + "grad_norm": 3.51640480717797, + "learning_rate": 1.7851338710077074e-06, + "loss": 0.816, + "num_input_tokens_seen": 197389230, + "step": 6079 + }, + { + "epoch": 0.5483158226991929, + "flos": 25405215587040.0, + "grad_norm": 2.843725518760523, + "learning_rate": 1.7845530507788076e-06, + "loss": 0.8554, + "num_input_tokens_seen": 197415395, + "step": 6080 + }, + { + "epoch": 0.5484060062226631, + "flos": 23443606416480.0, + "grad_norm": 1.861612336274679, + "learning_rate": 1.7839722489330298e-06, + "loss": 0.7198, + "num_input_tokens_seen": 197443795, + "step": 6081 + }, + { + "epoch": 0.5484961897461333, + "flos": 23078950904640.0, + "grad_norm": 1.8614876549675416, + "learning_rate": 1.7833914655199308e-06, + "loss": 0.7181, + "num_input_tokens_seen": 197473890, + "step": 6082 + }, + { + "epoch": 0.5485863732696037, + "flos": 25082716316640.0, + "grad_norm": 2.5000119209005334, + "learning_rate": 1.7828107005890658e-06, + "loss": 0.7937, + "num_input_tokens_seen": 197501725, + "step": 6083 + }, + { + "epoch": 0.5486765567930739, + "flos": 10983457066080.0, + "grad_norm": 2.892862639514471, + "learning_rate": 1.7822299541899898e-06, + "loss": 0.6671, + "num_input_tokens_seen": 197527125, + "step": 6084 + }, + { + "epoch": 0.5487667403165442, + "flos": 26466764358240.0, + "grad_norm": 2.5245537429622558, + "learning_rate": 1.7816492263722545e-06, + "loss": 0.7478, + "num_input_tokens_seen": 197559530, + "step": 6085 + }, + { + "epoch": 0.5488569238400144, + "flos": 23808113249280.0, + "grad_norm": 1.9987566182383028, + "learning_rate": 1.781068517185412e-06, + "loss": 0.7726, + "num_input_tokens_seen": 197587395, + "step": 6086 + }, + { + "epoch": 0.5489471073634847, + "flos": 20747036418720.0, + "grad_norm": 1.9461261319315146, + "learning_rate": 1.7804878266790104e-06, + "loss": 0.8142, + "num_input_tokens_seen": 197616465, + "step": 6087 + }, + { + "epoch": 0.549037290886955, + "flos": 25483729587360.0, + "grad_norm": 1.587958565510126, + "learning_rate": 1.779907154902597e-06, + "loss": 0.684, + "num_input_tokens_seen": 197646340, + "step": 6088 + }, + { + "epoch": 0.5491274744104252, + "flos": 32147012279040.0, + "grad_norm": 1.598701713256338, + "learning_rate": 1.7793265019057198e-06, + "loss": 0.6775, + "num_input_tokens_seen": 197678980, + "step": 6089 + }, + { + "epoch": 0.5492176579338954, + "flos": 70643988910080.0, + "grad_norm": 0.7379359492391215, + "learning_rate": 1.7787458677379212e-06, + "loss": 0.5946, + "num_input_tokens_seen": 197758945, + "step": 6090 + }, + { + "epoch": 0.5493078414573658, + "flos": 25884965876640.0, + "grad_norm": 1.695575254268106, + "learning_rate": 1.7781652524487463e-06, + "loss": 0.807, + "num_input_tokens_seen": 197789050, + "step": 6091 + }, + { + "epoch": 0.549398024980836, + "flos": 24063212277600.0, + "grad_norm": 1.9517408425394924, + "learning_rate": 1.777584656087735e-06, + "loss": 0.6964, + "num_input_tokens_seen": 197819295, + "step": 6092 + }, + { + "epoch": 0.5494882085043062, + "flos": 14260190076000.0, + "grad_norm": 2.1087298501617355, + "learning_rate": 1.777004078704427e-06, + "loss": 0.6682, + "num_input_tokens_seen": 197845450, + "step": 6093 + }, + { + "epoch": 0.5495783920277765, + "flos": 22966643858880.0, + "grad_norm": 2.227862169512287, + "learning_rate": 1.7764235203483603e-06, + "loss": 0.773, + "num_input_tokens_seen": 197872565, + "step": 6094 + }, + { + "epoch": 0.5496685755512468, + "flos": 22824446592480.0, + "grad_norm": 2.384221995232929, + "learning_rate": 1.775842981069072e-06, + "loss": 0.7211, + "num_input_tokens_seen": 197902980, + "step": 6095 + }, + { + "epoch": 0.549758759074717, + "flos": 26650411640640.0, + "grad_norm": 1.9138205239066803, + "learning_rate": 1.7752624609160966e-06, + "loss": 0.7602, + "num_input_tokens_seen": 197933210, + "step": 6096 + }, + { + "epoch": 0.5498489425981873, + "flos": 28398185950080.0, + "grad_norm": 2.3007381415740795, + "learning_rate": 1.7746819599389665e-06, + "loss": 0.7839, + "num_input_tokens_seen": 197962235, + "step": 6097 + }, + { + "epoch": 0.5499391261216575, + "flos": 26320478418240.0, + "grad_norm": 2.448651066992208, + "learning_rate": 1.774101478187215e-06, + "loss": 0.7535, + "num_input_tokens_seen": 197991480, + "step": 6098 + }, + { + "epoch": 0.5500293096451279, + "flos": 25482279966720.0, + "grad_norm": 2.0866768773146993, + "learning_rate": 1.773521015710371e-06, + "loss": 0.8213, + "num_input_tokens_seen": 198017780, + "step": 6099 + }, + { + "epoch": 0.5501194931685981, + "flos": 19108335385920.0, + "grad_norm": 2.7546859178630507, + "learning_rate": 1.7729405725579614e-06, + "loss": 0.8258, + "num_input_tokens_seen": 198046200, + "step": 6100 + }, + { + "epoch": 0.5502096766920683, + "flos": 25594438333440.0, + "grad_norm": 1.67292153946423, + "learning_rate": 1.7723601487795151e-06, + "loss": 0.723, + "num_input_tokens_seen": 198076775, + "step": 6101 + }, + { + "epoch": 0.5502998602155387, + "flos": 29344528434240.0, + "grad_norm": 2.152831532890762, + "learning_rate": 1.7717797444245557e-06, + "loss": 0.6971, + "num_input_tokens_seen": 198106960, + "step": 6102 + }, + { + "epoch": 0.5503900437390089, + "flos": 29892013493760.0, + "grad_norm": 1.645851819220275, + "learning_rate": 1.7711993595426076e-06, + "loss": 0.6866, + "num_input_tokens_seen": 198141085, + "step": 6103 + }, + { + "epoch": 0.5504802272624791, + "flos": 18561519382080.0, + "grad_norm": 2.0761649848486328, + "learning_rate": 1.7706189941831915e-06, + "loss": 0.8597, + "num_input_tokens_seen": 198166530, + "step": 6104 + }, + { + "epoch": 0.5505704107859494, + "flos": 41623260687840.0, + "grad_norm": 1.7952519923704944, + "learning_rate": 1.770038648395827e-06, + "loss": 0.7109, + "num_input_tokens_seen": 198200490, + "step": 6105 + }, + { + "epoch": 0.5506605943094197, + "flos": 62075154779520.0, + "grad_norm": 0.6078578206403219, + "learning_rate": 1.7694583222300336e-06, + "loss": 0.5935, + "num_input_tokens_seen": 198299740, + "step": 6106 + }, + { + "epoch": 0.55075077783289, + "flos": 21325155094080.0, + "grad_norm": 1.7223826772877373, + "learning_rate": 1.7688780157353272e-06, + "loss": 0.7292, + "num_input_tokens_seen": 198326685, + "step": 6107 + }, + { + "epoch": 0.5508409613563602, + "flos": 33498568216800.0, + "grad_norm": 1.9826394248318844, + "learning_rate": 1.768297728961223e-06, + "loss": 0.6931, + "num_input_tokens_seen": 198359795, + "step": 6108 + }, + { + "epoch": 0.5509311448798304, + "flos": 27593185827840.0, + "grad_norm": 2.075730744344352, + "learning_rate": 1.7677174619572342e-06, + "loss": 0.756, + "num_input_tokens_seen": 198387610, + "step": 6109 + }, + { + "epoch": 0.5510213284033008, + "flos": 20419630740000.0, + "grad_norm": 2.0359595062635747, + "learning_rate": 1.7671372147728717e-06, + "loss": 0.7661, + "num_input_tokens_seen": 198415190, + "step": 6110 + }, + { + "epoch": 0.551111511926771, + "flos": 18087530405280.0, + "grad_norm": 2.570414810695299, + "learning_rate": 1.7665569874576471e-06, + "loss": 0.8155, + "num_input_tokens_seen": 198440845, + "step": 6111 + }, + { + "epoch": 0.5512016954502412, + "flos": 25556705293440.0, + "grad_norm": 8.942325596440442, + "learning_rate": 1.7659767800610664e-06, + "loss": 0.7521, + "num_input_tokens_seen": 198470005, + "step": 6112 + }, + { + "epoch": 0.5512918789737115, + "flos": 22241904715680.0, + "grad_norm": 2.228710436852268, + "learning_rate": 1.7653965926326379e-06, + "loss": 0.6564, + "num_input_tokens_seen": 198499840, + "step": 6113 + }, + { + "epoch": 0.5513820624971818, + "flos": 21182994997440.0, + "grad_norm": 1.503139627429112, + "learning_rate": 1.764816425221866e-06, + "loss": 0.7069, + "num_input_tokens_seen": 198528900, + "step": 6114 + }, + { + "epoch": 0.551472246020652, + "flos": 29600259348480.0, + "grad_norm": 3.07072667247642, + "learning_rate": 1.7642362778782524e-06, + "loss": 0.6669, + "num_input_tokens_seen": 198562590, + "step": 6115 + }, + { + "epoch": 0.5515624295441223, + "flos": 20128322631840.0, + "grad_norm": 15.347760144564356, + "learning_rate": 1.7636561506513005e-06, + "loss": 0.7218, + "num_input_tokens_seen": 198591120, + "step": 6116 + }, + { + "epoch": 0.5516526130675925, + "flos": 48689972684640.0, + "grad_norm": 1.517541911946273, + "learning_rate": 1.7630760435905083e-06, + "loss": 0.7246, + "num_input_tokens_seen": 198625350, + "step": 6117 + }, + { + "epoch": 0.5517427965910628, + "flos": 16740137480640.0, + "grad_norm": 2.1915617154983646, + "learning_rate": 1.762495956745375e-06, + "loss": 0.7093, + "num_input_tokens_seen": 198652265, + "step": 6118 + }, + { + "epoch": 0.5518329801145331, + "flos": 24132247989120.0, + "grad_norm": 3.2433862147736265, + "learning_rate": 1.7619158901653962e-06, + "loss": 0.792, + "num_input_tokens_seen": 198678795, + "step": 6119 + }, + { + "epoch": 0.5519231636380033, + "flos": 24136299492960.0, + "grad_norm": 2.3470466317354086, + "learning_rate": 1.761335843900066e-06, + "loss": 0.749, + "num_input_tokens_seen": 198706805, + "step": 6120 + }, + { + "epoch": 0.5520133471614735, + "flos": 36194617837920.0, + "grad_norm": 2.001189712483637, + "learning_rate": 1.7607558179988785e-06, + "loss": 0.7334, + "num_input_tokens_seen": 198739215, + "step": 6121 + }, + { + "epoch": 0.5521035306849439, + "flos": 47485520421600.0, + "grad_norm": 1.6937077393837356, + "learning_rate": 1.760175812511323e-06, + "loss": 0.6771, + "num_input_tokens_seen": 198771985, + "step": 6122 + }, + { + "epoch": 0.5521937142084141, + "flos": 27269943162240.0, + "grad_norm": 2.1217048003712047, + "learning_rate": 1.75959582748689e-06, + "loss": 0.7073, + "num_input_tokens_seen": 198799345, + "step": 6123 + }, + { + "epoch": 0.5522838977318844, + "flos": 21039348110400.0, + "grad_norm": 2.0123131327621726, + "learning_rate": 1.7590158629750657e-06, + "loss": 0.7812, + "num_input_tokens_seen": 198825565, + "step": 6124 + }, + { + "epoch": 0.5523740812553547, + "flos": 19217520171840.0, + "grad_norm": 3.1318982091076832, + "learning_rate": 1.7584359190253376e-06, + "loss": 0.7986, + "num_input_tokens_seen": 198855235, + "step": 6125 + }, + { + "epoch": 0.5524642647788249, + "flos": 22824149234400.0, + "grad_norm": 1.974167469658399, + "learning_rate": 1.7578559956871892e-06, + "loss": 0.8112, + "num_input_tokens_seen": 198884170, + "step": 6126 + }, + { + "epoch": 0.5525544483022952, + "flos": 19982668577760.0, + "grad_norm": 1.5507338970651399, + "learning_rate": 1.7572760930101012e-06, + "loss": 0.7461, + "num_input_tokens_seen": 198913290, + "step": 6127 + }, + { + "epoch": 0.5526446318257654, + "flos": 64464545181600.0, + "grad_norm": 0.7672721893741328, + "learning_rate": 1.7566962110435563e-06, + "loss": 0.6334, + "num_input_tokens_seen": 199009470, + "step": 6128 + }, + { + "epoch": 0.5527348153492357, + "flos": 26103521297280.0, + "grad_norm": 1.9294249939071608, + "learning_rate": 1.7561163498370313e-06, + "loss": 0.7644, + "num_input_tokens_seen": 199042090, + "step": 6129 + }, + { + "epoch": 0.552824998872706, + "flos": 27452884219200.0, + "grad_norm": 1.7555736610562747, + "learning_rate": 1.755536509440005e-06, + "loss": 0.7139, + "num_input_tokens_seen": 199076435, + "step": 6130 + }, + { + "epoch": 0.5529151823961762, + "flos": 22601430800640.0, + "grad_norm": 3.2045145137825433, + "learning_rate": 1.7549566899019519e-06, + "loss": 0.7483, + "num_input_tokens_seen": 199104050, + "step": 6131 + }, + { + "epoch": 0.5530053659196464, + "flos": 24281730528480.0, + "grad_norm": 2.053999637005708, + "learning_rate": 1.754376891272344e-06, + "loss": 0.7665, + "num_input_tokens_seen": 199132555, + "step": 6132 + }, + { + "epoch": 0.5530955494431168, + "flos": 21622670552160.0, + "grad_norm": 1.5865957463199631, + "learning_rate": 1.753797113600655e-06, + "loss": 0.6979, + "num_input_tokens_seen": 199161600, + "step": 6133 + }, + { + "epoch": 0.553185732966587, + "flos": 24713340245280.0, + "grad_norm": 2.0476014689919206, + "learning_rate": 1.7532173569363535e-06, + "loss": 0.8844, + "num_input_tokens_seen": 199191125, + "step": 6134 + }, + { + "epoch": 0.5532759164900573, + "flos": 39109371558720.0, + "grad_norm": 1.9389411119801778, + "learning_rate": 1.7526376213289077e-06, + "loss": 0.5656, + "num_input_tokens_seen": 199222220, + "step": 6135 + }, + { + "epoch": 0.5533661000135275, + "flos": 22933817227200.0, + "grad_norm": 1.6162281231384261, + "learning_rate": 1.7520579068277844e-06, + "loss": 0.7319, + "num_input_tokens_seen": 199252045, + "step": 6136 + }, + { + "epoch": 0.5534562835369978, + "flos": 24682669459680.0, + "grad_norm": 1.8098606755709201, + "learning_rate": 1.7514782134824472e-06, + "loss": 0.7755, + "num_input_tokens_seen": 199282280, + "step": 6137 + }, + { + "epoch": 0.5535464670604681, + "flos": 20667779023200.0, + "grad_norm": 2.3473085980696133, + "learning_rate": 1.7508985413423599e-06, + "loss": 0.7422, + "num_input_tokens_seen": 199308975, + "step": 6138 + }, + { + "epoch": 0.5536366505839383, + "flos": 22496929404480.0, + "grad_norm": 1.8460137645749677, + "learning_rate": 1.7503188904569814e-06, + "loss": 0.8218, + "num_input_tokens_seen": 199337450, + "step": 6139 + }, + { + "epoch": 0.5537268341074085, + "flos": 21834461076480.0, + "grad_norm": 6.484711199398136, + "learning_rate": 1.7497392608757728e-06, + "loss": 0.7149, + "num_input_tokens_seen": 199367235, + "step": 6140 + }, + { + "epoch": 0.5538170176308789, + "flos": 25010149477920.0, + "grad_norm": 1.6521384830898767, + "learning_rate": 1.7491596526481897e-06, + "loss": 0.7657, + "num_input_tokens_seen": 199398660, + "step": 6141 + }, + { + "epoch": 0.5539072011543491, + "flos": 20784286251840.0, + "grad_norm": 2.3999564365565855, + "learning_rate": 1.7485800658236888e-06, + "loss": 0.6774, + "num_input_tokens_seen": 199426920, + "step": 6142 + }, + { + "epoch": 0.5539973846778193, + "flos": 22569013036320.0, + "grad_norm": 1.9265815975235785, + "learning_rate": 1.7480005004517232e-06, + "loss": 0.8178, + "num_input_tokens_seen": 199455255, + "step": 6143 + }, + { + "epoch": 0.5540875682012896, + "flos": 23842686859680.0, + "grad_norm": 2.5482058178284284, + "learning_rate": 1.7474209565817435e-06, + "loss": 0.8093, + "num_input_tokens_seen": 199483955, + "step": 6144 + }, + { + "epoch": 0.5541777517247599, + "flos": 31163494301280.0, + "grad_norm": 1.698007577283867, + "learning_rate": 1.7468414342632014e-06, + "loss": 0.7473, + "num_input_tokens_seen": 199515590, + "step": 6145 + }, + { + "epoch": 0.5542679352482301, + "flos": 22532952635520.0, + "grad_norm": 2.5249660338581688, + "learning_rate": 1.746261933545543e-06, + "loss": 0.8038, + "num_input_tokens_seen": 199544115, + "step": 6146 + }, + { + "epoch": 0.5543581187717004, + "flos": 23079768639360.0, + "grad_norm": 1.71782062719434, + "learning_rate": 1.7456824544782165e-06, + "loss": 0.7496, + "num_input_tokens_seen": 199571885, + "step": 6147 + }, + { + "epoch": 0.5544483022951707, + "flos": 24719287406880.0, + "grad_norm": 1.74928868687652, + "learning_rate": 1.7451029971106653e-06, + "loss": 0.754, + "num_input_tokens_seen": 199599795, + "step": 6148 + }, + { + "epoch": 0.554538485818641, + "flos": 21724904592960.0, + "grad_norm": 1.6587501944431862, + "learning_rate": 1.7445235614923313e-06, + "loss": 0.7782, + "num_input_tokens_seen": 199628205, + "step": 6149 + }, + { + "epoch": 0.5546286693421112, + "flos": 21513448596480.0, + "grad_norm": 2.108928152720861, + "learning_rate": 1.7439441476726556e-06, + "loss": 0.7383, + "num_input_tokens_seen": 199656490, + "step": 6150 + }, + { + "epoch": 0.5547188528655814, + "flos": 22824558101760.0, + "grad_norm": 3.554096415914763, + "learning_rate": 1.7433647557010776e-06, + "loss": 0.7573, + "num_input_tokens_seen": 199685155, + "step": 6151 + }, + { + "epoch": 0.5548090363890518, + "flos": 35612038791360.0, + "grad_norm": 1.9666926821372703, + "learning_rate": 1.7427853856270338e-06, + "loss": 0.7143, + "num_input_tokens_seen": 199714545, + "step": 6152 + }, + { + "epoch": 0.554899219912522, + "flos": 26576692539360.0, + "grad_norm": 2.3920166353266024, + "learning_rate": 1.7422060374999593e-06, + "loss": 0.7107, + "num_input_tokens_seen": 199744810, + "step": 6153 + }, + { + "epoch": 0.5549894034359922, + "flos": 21002878842240.0, + "grad_norm": 1.627883407035607, + "learning_rate": 1.7416267113692862e-06, + "loss": 0.7096, + "num_input_tokens_seen": 199774775, + "step": 6154 + }, + { + "epoch": 0.5550795869594625, + "flos": 23079173923200.0, + "grad_norm": 1.8445433510008145, + "learning_rate": 1.7410474072844475e-06, + "loss": 0.7784, + "num_input_tokens_seen": 199802785, + "step": 6155 + }, + { + "epoch": 0.5551697704829328, + "flos": 34154717685600.0, + "grad_norm": 1.8774211510407686, + "learning_rate": 1.740468125294871e-06, + "loss": 0.7327, + "num_input_tokens_seen": 199831905, + "step": 6156 + }, + { + "epoch": 0.555259954006403, + "flos": 21002321295840.0, + "grad_norm": 2.0487640050813187, + "learning_rate": 1.739888865449986e-06, + "loss": 0.6637, + "num_input_tokens_seen": 199855875, + "step": 6157 + }, + { + "epoch": 0.5553501375298733, + "flos": 19216776776640.0, + "grad_norm": 2.7378859295820415, + "learning_rate": 1.7393096277992174e-06, + "loss": 0.7748, + "num_input_tokens_seen": 199881720, + "step": 6158 + }, + { + "epoch": 0.5554403210533435, + "flos": 28976304625440.0, + "grad_norm": 1.893550658356982, + "learning_rate": 1.738730412391988e-06, + "loss": 0.7619, + "num_input_tokens_seen": 199911720, + "step": 6159 + }, + { + "epoch": 0.5555305045768139, + "flos": 22017179114880.0, + "grad_norm": 3.9641643073541597, + "learning_rate": 1.738151219277721e-06, + "loss": 0.801, + "num_input_tokens_seen": 199938545, + "step": 6160 + }, + { + "epoch": 0.5556206881002841, + "flos": 17177359831200.0, + "grad_norm": 1.977773182514772, + "learning_rate": 1.7375720485058349e-06, + "loss": 0.7052, + "num_input_tokens_seen": 199964435, + "step": 6161 + }, + { + "epoch": 0.5557108716237543, + "flos": 18197384246880.0, + "grad_norm": 2.2267780400402986, + "learning_rate": 1.7369929001257498e-06, + "loss": 0.7684, + "num_input_tokens_seen": 199992935, + "step": 6162 + }, + { + "epoch": 0.5558010551472246, + "flos": 25010223817440.0, + "grad_norm": 3.7700841310458277, + "learning_rate": 1.73641377418688e-06, + "loss": 0.6663, + "num_input_tokens_seen": 200021210, + "step": 6163 + }, + { + "epoch": 0.5558912386706949, + "flos": 21440398550880.0, + "grad_norm": 1.762188901567519, + "learning_rate": 1.7358346707386408e-06, + "loss": 0.6846, + "num_input_tokens_seen": 200048995, + "step": 6164 + }, + { + "epoch": 0.5559814221941651, + "flos": 17432161501440.0, + "grad_norm": 2.205845336977478, + "learning_rate": 1.7352555898304439e-06, + "loss": 0.799, + "num_input_tokens_seen": 200076185, + "step": 6165 + }, + { + "epoch": 0.5560716057176354, + "flos": 20085683183520.0, + "grad_norm": 2.500536479608336, + "learning_rate": 1.7346765315116996e-06, + "loss": 0.6675, + "num_input_tokens_seen": 200103840, + "step": 6166 + }, + { + "epoch": 0.5561617892411056, + "flos": 25698679541280.0, + "grad_norm": 2.0717411510103165, + "learning_rate": 1.734097495831817e-06, + "loss": 0.7114, + "num_input_tokens_seen": 200131530, + "step": 6167 + }, + { + "epoch": 0.5562519727645759, + "flos": 17505248716800.0, + "grad_norm": 2.5364985740232004, + "learning_rate": 1.7335184828402015e-06, + "loss": 0.7639, + "num_input_tokens_seen": 200158925, + "step": 6168 + }, + { + "epoch": 0.5563421562880462, + "flos": 22642583458560.0, + "grad_norm": 1.6415634377776611, + "learning_rate": 1.7329394925862595e-06, + "loss": 0.7276, + "num_input_tokens_seen": 200185555, + "step": 6169 + }, + { + "epoch": 0.5564323398115164, + "flos": 18816506901120.0, + "grad_norm": 2.2732758874244583, + "learning_rate": 1.7323605251193922e-06, + "loss": 0.7964, + "num_input_tokens_seen": 200211475, + "step": 6170 + }, + { + "epoch": 0.5565225233349868, + "flos": 14371344859200.0, + "grad_norm": 2.993974037786639, + "learning_rate": 1.7317815804890001e-06, + "loss": 0.77, + "num_input_tokens_seen": 200235680, + "step": 6171 + }, + { + "epoch": 0.556612706858457, + "flos": 25192309969920.0, + "grad_norm": 1.980744953946888, + "learning_rate": 1.731202658744483e-06, + "loss": 0.707, + "num_input_tokens_seen": 200264080, + "step": 6172 + }, + { + "epoch": 0.5567028903819272, + "flos": 24937508299680.0, + "grad_norm": 1.4720700663186408, + "learning_rate": 1.7306237599352365e-06, + "loss": 0.8092, + "num_input_tokens_seen": 200295155, + "step": 6173 + }, + { + "epoch": 0.5567930739053975, + "flos": 24061242280320.0, + "grad_norm": 1.6676656192962043, + "learning_rate": 1.730044884110657e-06, + "loss": 0.736, + "num_input_tokens_seen": 200324150, + "step": 6174 + }, + { + "epoch": 0.5568832574288678, + "flos": 31780646958240.0, + "grad_norm": 1.982808373268829, + "learning_rate": 1.7294660313201366e-06, + "loss": 0.7057, + "num_input_tokens_seen": 200352035, + "step": 6175 + }, + { + "epoch": 0.556973440952338, + "flos": 25301123058240.0, + "grad_norm": 14.560768908644128, + "learning_rate": 1.7288872016130652e-06, + "loss": 0.728, + "num_input_tokens_seen": 200374095, + "step": 6176 + }, + { + "epoch": 0.5570636244758083, + "flos": 32114706024000.0, + "grad_norm": 1.7917665261525348, + "learning_rate": 1.7283083950388334e-06, + "loss": 0.662, + "num_input_tokens_seen": 200406460, + "step": 6177 + }, + { + "epoch": 0.5571538079992785, + "flos": 43804428862560.0, + "grad_norm": 1.9136693433911096, + "learning_rate": 1.727729611646827e-06, + "loss": 0.6537, + "num_input_tokens_seen": 200439470, + "step": 6178 + }, + { + "epoch": 0.5572439915227488, + "flos": 22824781120320.0, + "grad_norm": 2.2021987807784873, + "learning_rate": 1.7271508514864318e-06, + "loss": 0.8022, + "num_input_tokens_seen": 200467750, + "step": 6179 + }, + { + "epoch": 0.5573341750462191, + "flos": 15647248868160.0, + "grad_norm": 2.4786576996132013, + "learning_rate": 1.7265721146070302e-06, + "loss": 0.7499, + "num_input_tokens_seen": 200492215, + "step": 6180 + }, + { + "epoch": 0.5574243585696893, + "flos": 20675175805440.0, + "grad_norm": 2.1830344260031143, + "learning_rate": 1.7259934010580035e-06, + "loss": 0.7485, + "num_input_tokens_seen": 200518870, + "step": 6181 + }, + { + "epoch": 0.5575145420931595, + "flos": 28868903988000.0, + "grad_norm": 3.1859929130939943, + "learning_rate": 1.725414710888731e-06, + "loss": 0.6942, + "num_input_tokens_seen": 200548485, + "step": 6182 + }, + { + "epoch": 0.5576047256166299, + "flos": 19836865844640.0, + "grad_norm": 2.058761097658654, + "learning_rate": 1.7248360441485885e-06, + "loss": 0.7243, + "num_input_tokens_seen": 200578805, + "step": 6183 + }, + { + "epoch": 0.5576949091401001, + "flos": 30325741886880.0, + "grad_norm": 2.11166359409188, + "learning_rate": 1.7242574008869528e-06, + "loss": 0.8209, + "num_input_tokens_seen": 200606930, + "step": 6184 + }, + { + "epoch": 0.5577850926635703, + "flos": 71174332976640.0, + "grad_norm": 0.672340453625399, + "learning_rate": 1.7236787811531951e-06, + "loss": 0.6291, + "num_input_tokens_seen": 200710520, + "step": 6185 + }, + { + "epoch": 0.5578752761870406, + "flos": 27047670765600.0, + "grad_norm": 1.9107644452691088, + "learning_rate": 1.7231001849966887e-06, + "loss": 0.7179, + "num_input_tokens_seen": 200741660, + "step": 6186 + }, + { + "epoch": 0.5579654597105109, + "flos": 69778279102560.0, + "grad_norm": 0.6516995475682013, + "learning_rate": 1.722521612466801e-06, + "loss": 0.6068, + "num_input_tokens_seen": 200839520, + "step": 6187 + }, + { + "epoch": 0.5580556432339812, + "flos": 18707805322080.0, + "grad_norm": 2.575567468131625, + "learning_rate": 1.7219430636128984e-06, + "loss": 0.6925, + "num_input_tokens_seen": 200866835, + "step": 6188 + }, + { + "epoch": 0.5581458267574514, + "flos": 20310111426240.0, + "grad_norm": 1.807158854622148, + "learning_rate": 1.7213645384843479e-06, + "loss": 0.6933, + "num_input_tokens_seen": 200894505, + "step": 6189 + }, + { + "epoch": 0.5582360102809216, + "flos": 35972196762240.0, + "grad_norm": 2.2225002777133764, + "learning_rate": 1.7207860371305108e-06, + "loss": 0.7396, + "num_input_tokens_seen": 200926070, + "step": 6190 + }, + { + "epoch": 0.558326193804392, + "flos": 39323094910560.0, + "grad_norm": 2.3079500543489107, + "learning_rate": 1.7202075596007487e-06, + "loss": 0.721, + "num_input_tokens_seen": 200957920, + "step": 6191 + }, + { + "epoch": 0.5584163773278622, + "flos": 48582349028640.0, + "grad_norm": 2.4136035075447, + "learning_rate": 1.7196291059444206e-06, + "loss": 0.7069, + "num_input_tokens_seen": 200990665, + "step": 6192 + }, + { + "epoch": 0.5585065608513324, + "flos": 30401802683040.0, + "grad_norm": 2.040953595984763, + "learning_rate": 1.7190506762108828e-06, + "loss": 0.7316, + "num_input_tokens_seen": 201019830, + "step": 6193 + }, + { + "epoch": 0.5585967443748028, + "flos": 20782910970720.0, + "grad_norm": 2.1694018876572634, + "learning_rate": 1.7184722704494907e-06, + "loss": 0.7467, + "num_input_tokens_seen": 201049905, + "step": 6194 + }, + { + "epoch": 0.558686927898273, + "flos": 25735520507040.0, + "grad_norm": 1.7534413560557385, + "learning_rate": 1.717893888709596e-06, + "loss": 0.7617, + "num_input_tokens_seen": 201077530, + "step": 6195 + }, + { + "epoch": 0.5587771114217432, + "flos": 20128917348000.0, + "grad_norm": 1.7756484445637433, + "learning_rate": 1.7173155310405515e-06, + "loss": 0.7767, + "num_input_tokens_seen": 201103635, + "step": 6196 + }, + { + "epoch": 0.5588672949452135, + "flos": 25116323513280.0, + "grad_norm": 1.6599113014377447, + "learning_rate": 1.7167371974917043e-06, + "loss": 0.8218, + "num_input_tokens_seen": 201133970, + "step": 6197 + }, + { + "epoch": 0.5589574784686838, + "flos": 18125263445280.0, + "grad_norm": 2.4674870604511083, + "learning_rate": 1.7161588881124003e-06, + "loss": 0.7631, + "num_input_tokens_seen": 201159615, + "step": 6198 + }, + { + "epoch": 0.559047661992154, + "flos": 29309285768160.0, + "grad_norm": 4.412544245943842, + "learning_rate": 1.7155806029519861e-06, + "loss": 0.7329, + "num_input_tokens_seen": 201188570, + "step": 6199 + }, + { + "epoch": 0.5591378455156243, + "flos": 19982259710400.0, + "grad_norm": 2.313202107011949, + "learning_rate": 1.7150023420598023e-06, + "loss": 0.7894, + "num_input_tokens_seen": 201215215, + "step": 6200 + }, + { + "epoch": 0.5592280290390945, + "flos": 32698028465760.0, + "grad_norm": 1.8305216020873991, + "learning_rate": 1.714424105485191e-06, + "loss": 0.8001, + "num_input_tokens_seen": 201247190, + "step": 6201 + }, + { + "epoch": 0.5593182125625649, + "flos": 23990385250560.0, + "grad_norm": 2.164277544595208, + "learning_rate": 1.7138458932774896e-06, + "loss": 0.7119, + "num_input_tokens_seen": 201276160, + "step": 6202 + }, + { + "epoch": 0.5594083960860351, + "flos": 60014625410400.0, + "grad_norm": 0.6501824902238165, + "learning_rate": 1.7132677054860335e-06, + "loss": 0.6061, + "num_input_tokens_seen": 201362540, + "step": 6203 + }, + { + "epoch": 0.5594985796095053, + "flos": 19397561987520.0, + "grad_norm": 2.651813257394552, + "learning_rate": 1.7126895421601586e-06, + "loss": 0.771, + "num_input_tokens_seen": 201390040, + "step": 6204 + }, + { + "epoch": 0.5595887631329756, + "flos": 29598326520960.0, + "grad_norm": 2.1809778256834815, + "learning_rate": 1.712111403349196e-06, + "loss": 0.6921, + "num_input_tokens_seen": 201422630, + "step": 6205 + }, + { + "epoch": 0.5596789466564459, + "flos": 24828658041600.0, + "grad_norm": 1.5763682038700106, + "learning_rate": 1.7115332891024757e-06, + "loss": 0.78, + "num_input_tokens_seen": 201451575, + "step": 6206 + }, + { + "epoch": 0.5597691301799161, + "flos": 58924598869440.0, + "grad_norm": 0.7135097793545142, + "learning_rate": 1.7109551994693257e-06, + "loss": 0.6563, + "num_input_tokens_seen": 201537395, + "step": 6207 + }, + { + "epoch": 0.5598593137033864, + "flos": 22096882547520.0, + "grad_norm": 1.8500832229659359, + "learning_rate": 1.7103771344990725e-06, + "loss": 0.6889, + "num_input_tokens_seen": 201566050, + "step": 6208 + }, + { + "epoch": 0.5599494972268566, + "flos": 40020285528000.0, + "grad_norm": 1.956439631231924, + "learning_rate": 1.709799094241039e-06, + "loss": 0.6632, + "num_input_tokens_seen": 201597410, + "step": 6209 + }, + { + "epoch": 0.560039680750327, + "flos": 26321519171520.0, + "grad_norm": 1.9883121270740667, + "learning_rate": 1.709221078744546e-06, + "loss": 0.781, + "num_input_tokens_seen": 201627210, + "step": 6210 + }, + { + "epoch": 0.5601298642737972, + "flos": 26937482396160.0, + "grad_norm": 1.6582808369645277, + "learning_rate": 1.7086430880589148e-06, + "loss": 0.7564, + "num_input_tokens_seen": 201658175, + "step": 6211 + }, + { + "epoch": 0.5602200477972674, + "flos": 18161546864640.0, + "grad_norm": 3.4051030178982837, + "learning_rate": 1.7080651222334612e-06, + "loss": 0.7541, + "num_input_tokens_seen": 201685555, + "step": 6212 + }, + { + "epoch": 0.5603102313207377, + "flos": 25520273195040.0, + "grad_norm": 2.125297749920672, + "learning_rate": 1.7074871813175018e-06, + "loss": 0.8065, + "num_input_tokens_seen": 201709805, + "step": 6213 + }, + { + "epoch": 0.560400414844208, + "flos": 25044983276640.0, + "grad_norm": 1.7964455796324739, + "learning_rate": 1.706909265360349e-06, + "loss": 0.6618, + "num_input_tokens_seen": 201741250, + "step": 6214 + }, + { + "epoch": 0.5604905983676782, + "flos": 68480699389920.0, + "grad_norm": 0.6983821488535672, + "learning_rate": 1.7063313744113128e-06, + "loss": 0.5955, + "num_input_tokens_seen": 201828300, + "step": 6215 + }, + { + "epoch": 0.5605807818911485, + "flos": 69190198931520.0, + "grad_norm": 0.6359456757715847, + "learning_rate": 1.7057535085197042e-06, + "loss": 0.5695, + "num_input_tokens_seen": 201920860, + "step": 6216 + }, + { + "epoch": 0.5606709654146187, + "flos": 19472619200160.0, + "grad_norm": 13.768486219209239, + "learning_rate": 1.705175667734828e-06, + "loss": 0.6691, + "num_input_tokens_seen": 201948450, + "step": 6217 + }, + { + "epoch": 0.560761148938089, + "flos": 27086964935520.0, + "grad_norm": 2.3860690517170218, + "learning_rate": 1.7045978521059894e-06, + "loss": 0.7155, + "num_input_tokens_seen": 201976215, + "step": 6218 + }, + { + "epoch": 0.5608513324615593, + "flos": 20821387405920.0, + "grad_norm": 1.575388055255893, + "learning_rate": 1.7040200616824914e-06, + "loss": 0.7473, + "num_input_tokens_seen": 202004825, + "step": 6219 + }, + { + "epoch": 0.5609415159850295, + "flos": 25339413644640.0, + "grad_norm": 1.9423570357923867, + "learning_rate": 1.7034422965136333e-06, + "loss": 0.6864, + "num_input_tokens_seen": 202033860, + "step": 6220 + }, + { + "epoch": 0.5610316995084998, + "flos": 23772127188000.0, + "grad_norm": 2.1737487175242856, + "learning_rate": 1.7028645566487137e-06, + "loss": 0.7868, + "num_input_tokens_seen": 202062115, + "step": 6221 + }, + { + "epoch": 0.5611218830319701, + "flos": 31309259864640.0, + "grad_norm": 1.8136450997848594, + "learning_rate": 1.7022868421370284e-06, + "loss": 0.7224, + "num_input_tokens_seen": 202090465, + "step": 6222 + }, + { + "epoch": 0.5612120665554403, + "flos": 21691334566080.0, + "grad_norm": 1.87720772785829, + "learning_rate": 1.701709153027872e-06, + "loss": 0.7929, + "num_input_tokens_seen": 202118450, + "step": 6223 + }, + { + "epoch": 0.5613022500789105, + "flos": 23261483094240.0, + "grad_norm": 1.8743609292983114, + "learning_rate": 1.7011314893705353e-06, + "loss": 0.7146, + "num_input_tokens_seen": 202147380, + "step": 6224 + }, + { + "epoch": 0.5613924336023809, + "flos": 20893917074880.0, + "grad_norm": 1.7351832525564534, + "learning_rate": 1.700553851214307e-06, + "loss": 0.7469, + "num_input_tokens_seen": 202176295, + "step": 6225 + }, + { + "epoch": 0.5614826171258511, + "flos": 43766844501600.0, + "grad_norm": 2.5783252985047618, + "learning_rate": 1.699976238608476e-06, + "loss": 0.6155, + "num_input_tokens_seen": 202206755, + "step": 6226 + }, + { + "epoch": 0.5615728006493214, + "flos": 25228779238080.0, + "grad_norm": 16.533550023125326, + "learning_rate": 1.699398651602326e-06, + "loss": 0.6979, + "num_input_tokens_seen": 202235045, + "step": 6227 + }, + { + "epoch": 0.5616629841727916, + "flos": 26395424121600.0, + "grad_norm": 1.9095939323413595, + "learning_rate": 1.6988210902451413e-06, + "loss": 0.7891, + "num_input_tokens_seen": 202264535, + "step": 6228 + }, + { + "epoch": 0.5617531676962619, + "flos": 20528815525920.0, + "grad_norm": 1.9348103712742826, + "learning_rate": 1.6982435545862011e-06, + "loss": 0.7271, + "num_input_tokens_seen": 202292165, + "step": 6229 + }, + { + "epoch": 0.5618433512197322, + "flos": 33019970189760.0, + "grad_norm": 2.1169138376093195, + "learning_rate": 1.6976660446747853e-06, + "loss": 0.7532, + "num_input_tokens_seen": 202322140, + "step": 6230 + }, + { + "epoch": 0.5619335347432024, + "flos": 24063063598560.0, + "grad_norm": 2.8330364258723932, + "learning_rate": 1.6970885605601696e-06, + "loss": 0.6754, + "num_input_tokens_seen": 202350580, + "step": 6231 + }, + { + "epoch": 0.5620237182666726, + "flos": 23297506325280.0, + "grad_norm": 2.1388089906816377, + "learning_rate": 1.6965111022916282e-06, + "loss": 0.8583, + "num_input_tokens_seen": 202377025, + "step": 6232 + }, + { + "epoch": 0.562113901790143, + "flos": 26066605992000.0, + "grad_norm": 1.8483260958741705, + "learning_rate": 1.6959336699184323e-06, + "loss": 0.7209, + "num_input_tokens_seen": 202405875, + "step": 6233 + }, + { + "epoch": 0.5622040853136132, + "flos": 22022197032480.0, + "grad_norm": 1.875593536529989, + "learning_rate": 1.6953562634898529e-06, + "loss": 0.7413, + "num_input_tokens_seen": 202435410, + "step": 6234 + }, + { + "epoch": 0.5622942688370834, + "flos": 17682354121440.0, + "grad_norm": 2.277175190414509, + "learning_rate": 1.6947788830551569e-06, + "loss": 0.7104, + "num_input_tokens_seen": 202459200, + "step": 6235 + }, + { + "epoch": 0.5623844523605537, + "flos": 59359776883200.0, + "grad_norm": 0.7664439822988806, + "learning_rate": 1.6942015286636093e-06, + "loss": 0.6292, + "num_input_tokens_seen": 202543200, + "step": 6236 + }, + { + "epoch": 0.562474635884024, + "flos": 22205138089440.0, + "grad_norm": 1.8845753628060893, + "learning_rate": 1.6936242003644735e-06, + "loss": 0.7315, + "num_input_tokens_seen": 202569910, + "step": 6237 + }, + { + "epoch": 0.5625648194074943, + "flos": 17650977110400.0, + "grad_norm": 3.199091609529128, + "learning_rate": 1.6930468982070106e-06, + "loss": 0.8164, + "num_input_tokens_seen": 202594580, + "step": 6238 + }, + { + "epoch": 0.5626550029309645, + "flos": 66994826174880.0, + "grad_norm": 0.5914672588289887, + "learning_rate": 1.692469622240478e-06, + "loss": 0.5178, + "num_input_tokens_seen": 202687580, + "step": 6239 + }, + { + "epoch": 0.5627451864544347, + "flos": 27591513188640.0, + "grad_norm": 1.6388984678099852, + "learning_rate": 1.6918923725141339e-06, + "loss": 0.7029, + "num_input_tokens_seen": 202716565, + "step": 6240 + }, + { + "epoch": 0.5628353699779051, + "flos": 30183172922880.0, + "grad_norm": 2.0342728643371246, + "learning_rate": 1.6913151490772312e-06, + "loss": 0.7967, + "num_input_tokens_seen": 202749115, + "step": 6241 + }, + { + "epoch": 0.5629255535013753, + "flos": 23987225820960.0, + "grad_norm": 2.2490301161090045, + "learning_rate": 1.6907379519790215e-06, + "loss": 0.7865, + "num_input_tokens_seen": 202779290, + "step": 6242 + }, + { + "epoch": 0.5630157370248455, + "flos": 38381026948800.0, + "grad_norm": 1.5852078417658895, + "learning_rate": 1.6901607812687558e-06, + "loss": 0.6546, + "num_input_tokens_seen": 202812325, + "step": 6243 + }, + { + "epoch": 0.5631059205483159, + "flos": 16994455944000.0, + "grad_norm": 2.276163570767952, + "learning_rate": 1.6895836369956794e-06, + "loss": 0.7016, + "num_input_tokens_seen": 202838160, + "step": 6244 + }, + { + "epoch": 0.5631961040717861, + "flos": 46570103177760.0, + "grad_norm": 0.6764788031841467, + "learning_rate": 1.6890065192090402e-06, + "loss": 0.5418, + "num_input_tokens_seen": 202919460, + "step": 6245 + }, + { + "epoch": 0.5632862875952563, + "flos": 20272973102400.0, + "grad_norm": 2.2477509596318446, + "learning_rate": 1.6884294279580793e-06, + "loss": 0.6461, + "num_input_tokens_seen": 202945375, + "step": 6246 + }, + { + "epoch": 0.5633764711187266, + "flos": 24646869247200.0, + "grad_norm": 2.3688280319204287, + "learning_rate": 1.6878523632920371e-06, + "loss": 0.7729, + "num_input_tokens_seen": 202973020, + "step": 6247 + }, + { + "epoch": 0.5634666546421969, + "flos": 17687483548320.0, + "grad_norm": 1.900818701083893, + "learning_rate": 1.6872753252601525e-06, + "loss": 0.8113, + "num_input_tokens_seen": 202999860, + "step": 6248 + }, + { + "epoch": 0.5635568381656672, + "flos": 19400163870720.0, + "grad_norm": 1.713256727217945, + "learning_rate": 1.6866983139116616e-06, + "loss": 0.6885, + "num_input_tokens_seen": 203029515, + "step": 6249 + }, + { + "epoch": 0.5636470216891374, + "flos": 20420076777120.0, + "grad_norm": 1.7354578984900904, + "learning_rate": 1.6861213292957981e-06, + "loss": 0.7162, + "num_input_tokens_seen": 203058315, + "step": 6250 + }, + { + "epoch": 0.5637372052126076, + "flos": 26358397307040.0, + "grad_norm": 2.0578624722170638, + "learning_rate": 1.685544371461793e-06, + "loss": 0.7791, + "num_input_tokens_seen": 203086955, + "step": 6251 + }, + { + "epoch": 0.563827388736078, + "flos": 18671224544640.0, + "grad_norm": 1.7355272744029502, + "learning_rate": 1.6849674404588767e-06, + "loss": 0.7696, + "num_input_tokens_seen": 203114645, + "step": 6252 + }, + { + "epoch": 0.5639175722595482, + "flos": 17572128582240.0, + "grad_norm": 2.170624297383835, + "learning_rate": 1.6843905363362758e-06, + "loss": 0.777, + "num_input_tokens_seen": 203139075, + "step": 6253 + }, + { + "epoch": 0.5640077557830184, + "flos": 24572518260000.0, + "grad_norm": 1.8289092900227697, + "learning_rate": 1.6838136591432136e-06, + "loss": 0.6862, + "num_input_tokens_seen": 203171345, + "step": 6254 + }, + { + "epoch": 0.5640979393064887, + "flos": 22168743160800.0, + "grad_norm": 2.9054331298023244, + "learning_rate": 1.6832368089289139e-06, + "loss": 0.772, + "num_input_tokens_seen": 203199575, + "step": 6255 + }, + { + "epoch": 0.564188122829959, + "flos": 23914473133440.0, + "grad_norm": 2.212283422354924, + "learning_rate": 1.682659985742596e-06, + "loss": 0.7636, + "num_input_tokens_seen": 203228270, + "step": 6256 + }, + { + "epoch": 0.5642783063534292, + "flos": 25374842159520.0, + "grad_norm": 2.4837951943250665, + "learning_rate": 1.6820831896334782e-06, + "loss": 0.7872, + "num_input_tokens_seen": 203256635, + "step": 6257 + }, + { + "epoch": 0.5643684898768995, + "flos": 27663485311200.0, + "grad_norm": 2.3148078743159477, + "learning_rate": 1.681506420650776e-06, + "loss": 0.7673, + "num_input_tokens_seen": 203287195, + "step": 6258 + }, + { + "epoch": 0.5644586734003697, + "flos": 22966718198400.0, + "grad_norm": 1.5975648438810435, + "learning_rate": 1.680929678843701e-06, + "loss": 0.7338, + "num_input_tokens_seen": 203315570, + "step": 6259 + }, + { + "epoch": 0.56454885692384, + "flos": 24645716984640.0, + "grad_norm": 1.7926676046417003, + "learning_rate": 1.6803529642614662e-06, + "loss": 0.7935, + "num_input_tokens_seen": 203345910, + "step": 6260 + }, + { + "epoch": 0.5646390404473103, + "flos": 28180188075840.0, + "grad_norm": 2.347061665879263, + "learning_rate": 1.6797762769532785e-06, + "loss": 0.7518, + "num_input_tokens_seen": 203376575, + "step": 6261 + }, + { + "epoch": 0.5647292239707805, + "flos": 24754790261280.0, + "grad_norm": 1.4694750193536033, + "learning_rate": 1.679199616968345e-06, + "loss": 0.7813, + "num_input_tokens_seen": 203406830, + "step": 6262 + }, + { + "epoch": 0.5648194074942507, + "flos": 20014900493280.0, + "grad_norm": 1.6673264707439077, + "learning_rate": 1.6786229843558689e-06, + "loss": 0.774, + "num_input_tokens_seen": 203435235, + "step": 6263 + }, + { + "epoch": 0.5649095910177211, + "flos": 18047938877280.0, + "grad_norm": 4.513358527432045, + "learning_rate": 1.6780463791650514e-06, + "loss": 0.7296, + "num_input_tokens_seen": 203460205, + "step": 6264 + }, + { + "epoch": 0.5649997745411913, + "flos": 28143978996000.0, + "grad_norm": 3.4724072945333497, + "learning_rate": 1.6774698014450928e-06, + "loss": 0.6934, + "num_input_tokens_seen": 203490825, + "step": 6265 + }, + { + "epoch": 0.5650899580646616, + "flos": 62923692157920.0, + "grad_norm": 0.6547868858047161, + "learning_rate": 1.6768932512451883e-06, + "loss": 0.5783, + "num_input_tokens_seen": 203575045, + "step": 6266 + }, + { + "epoch": 0.5651801415881319, + "flos": 23225757221280.0, + "grad_norm": 2.0977301273798052, + "learning_rate": 1.676316728614534e-06, + "loss": 0.6979, + "num_input_tokens_seen": 203606640, + "step": 6267 + }, + { + "epoch": 0.5652703251116021, + "flos": 68771784479520.0, + "grad_norm": 0.5717468253574214, + "learning_rate": 1.675740233602321e-06, + "loss": 0.5875, + "num_input_tokens_seen": 203704990, + "step": 6268 + }, + { + "epoch": 0.5653605086350724, + "flos": 23183377961280.0, + "grad_norm": 1.9665334424730718, + "learning_rate": 1.6751637662577385e-06, + "loss": 0.791, + "num_input_tokens_seen": 203733630, + "step": 6269 + }, + { + "epoch": 0.5654506921585426, + "flos": 58223542596960.0, + "grad_norm": 0.8107874501789454, + "learning_rate": 1.6745873266299753e-06, + "loss": 0.5314, + "num_input_tokens_seen": 203813550, + "step": 6270 + }, + { + "epoch": 0.565540875682013, + "flos": 18452669124000.0, + "grad_norm": 2.71002243384616, + "learning_rate": 1.6740109147682148e-06, + "loss": 0.708, + "num_input_tokens_seen": 203841740, + "step": 6271 + }, + { + "epoch": 0.5656310592054832, + "flos": 22423470491520.0, + "grad_norm": 2.5705249084085016, + "learning_rate": 1.6734345307216418e-06, + "loss": 0.7578, + "num_input_tokens_seen": 203871080, + "step": 6272 + }, + { + "epoch": 0.5657212427289534, + "flos": 18743865722880.0, + "grad_norm": 2.7797818541528123, + "learning_rate": 1.6728581745394346e-06, + "loss": 0.7188, + "num_input_tokens_seen": 203897750, + "step": 6273 + }, + { + "epoch": 0.5658114262524236, + "flos": 24608950358400.0, + "grad_norm": 1.6291664702406405, + "learning_rate": 1.672281846270772e-06, + "loss": 0.8206, + "num_input_tokens_seen": 203928845, + "step": 6274 + }, + { + "epoch": 0.565901609775894, + "flos": 22060301770080.0, + "grad_norm": 2.023188630930919, + "learning_rate": 1.6717055459648295e-06, + "loss": 0.7438, + "num_input_tokens_seen": 203958490, + "step": 6275 + }, + { + "epoch": 0.5659917932993642, + "flos": 26174118138720.0, + "grad_norm": 2.404876713437913, + "learning_rate": 1.6711292736707793e-06, + "loss": 0.6968, + "num_input_tokens_seen": 203986640, + "step": 6276 + }, + { + "epoch": 0.5660819768228345, + "flos": 18524901434880.0, + "grad_norm": 2.011995581680816, + "learning_rate": 1.6705530294377938e-06, + "loss": 0.8184, + "num_input_tokens_seen": 204014540, + "step": 6277 + }, + { + "epoch": 0.5661721603463047, + "flos": 26431298673600.0, + "grad_norm": 1.8925913523870195, + "learning_rate": 1.6699768133150395e-06, + "loss": 0.7352, + "num_input_tokens_seen": 204044795, + "step": 6278 + }, + { + "epoch": 0.566262343869775, + "flos": 21657132653280.0, + "grad_norm": 2.4867024585646833, + "learning_rate": 1.6694006253516837e-06, + "loss": 0.737, + "num_input_tokens_seen": 204071460, + "step": 6279 + }, + { + "epoch": 0.5663525273932453, + "flos": 21949890382080.0, + "grad_norm": 2.3916604785196705, + "learning_rate": 1.6688244655968896e-06, + "loss": 0.7409, + "num_input_tokens_seen": 204099495, + "step": 6280 + }, + { + "epoch": 0.5664427109167155, + "flos": 28908904383360.0, + "grad_norm": 3.1451993054915692, + "learning_rate": 1.6682483340998175e-06, + "loss": 0.6687, + "num_input_tokens_seen": 204131540, + "step": 6281 + }, + { + "epoch": 0.5665328944401857, + "flos": 22130489744160.0, + "grad_norm": 12.544157843238969, + "learning_rate": 1.6676722309096276e-06, + "loss": 0.7052, + "num_input_tokens_seen": 204158210, + "step": 6282 + }, + { + "epoch": 0.5666230779636561, + "flos": 65947890852960.0, + "grad_norm": 0.7431119281462893, + "learning_rate": 1.6670961560754744e-06, + "loss": 0.6462, + "num_input_tokens_seen": 204243985, + "step": 6283 + }, + { + "epoch": 0.5667132614871263, + "flos": 31969386497760.0, + "grad_norm": 1.849196777728931, + "learning_rate": 1.6665201096465138e-06, + "loss": 0.7413, + "num_input_tokens_seen": 204275950, + "step": 6284 + }, + { + "epoch": 0.5668034450105965, + "flos": 19982965935840.0, + "grad_norm": 2.06496386313974, + "learning_rate": 1.6659440916718961e-06, + "loss": 0.79, + "num_input_tokens_seen": 204304140, + "step": 6285 + }, + { + "epoch": 0.5668936285340668, + "flos": 20269888012320.0, + "grad_norm": 2.3078987120975465, + "learning_rate": 1.6653681022007696e-06, + "loss": 0.7597, + "num_input_tokens_seen": 204331660, + "step": 6286 + }, + { + "epoch": 0.5669838120575371, + "flos": 19800768274080.0, + "grad_norm": 3.7929932645561424, + "learning_rate": 1.6647921412822825e-06, + "loss": 0.6766, + "num_input_tokens_seen": 204359480, + "step": 6287 + }, + { + "epoch": 0.5670739955810074, + "flos": 22241681697120.0, + "grad_norm": 4.08244942161312, + "learning_rate": 1.6642162089655782e-06, + "loss": 0.7417, + "num_input_tokens_seen": 204387505, + "step": 6288 + }, + { + "epoch": 0.5671641791044776, + "flos": 20780606445600.0, + "grad_norm": 2.1740884822045743, + "learning_rate": 1.663640305299798e-06, + "loss": 0.7109, + "num_input_tokens_seen": 204415370, + "step": 6289 + }, + { + "epoch": 0.5672543626279479, + "flos": 64394245698240.0, + "grad_norm": 0.6616519033993423, + "learning_rate": 1.6630644303340824e-06, + "loss": 0.6098, + "num_input_tokens_seen": 204510470, + "step": 6290 + }, + { + "epoch": 0.5673445461514182, + "flos": 27268828069440.0, + "grad_norm": 2.9301054389391488, + "learning_rate": 1.662488584117567e-06, + "loss": 0.7538, + "num_input_tokens_seen": 204539820, + "step": 6291 + }, + { + "epoch": 0.5674347296748884, + "flos": 35283926887200.0, + "grad_norm": 1.8241892124047354, + "learning_rate": 1.6619127666993867e-06, + "loss": 0.8007, + "num_input_tokens_seen": 204572370, + "step": 6292 + }, + { + "epoch": 0.5675249131983586, + "flos": 20889085006080.0, + "grad_norm": 2.3502354727114394, + "learning_rate": 1.6613369781286727e-06, + "loss": 0.7466, + "num_input_tokens_seen": 204600150, + "step": 6293 + }, + { + "epoch": 0.567615096721829, + "flos": 32879594241600.0, + "grad_norm": 2.0259254507224016, + "learning_rate": 1.6607612184545562e-06, + "loss": 0.6826, + "num_input_tokens_seen": 204631265, + "step": 6294 + }, + { + "epoch": 0.5677052802452992, + "flos": 21764310272160.0, + "grad_norm": 2.0113241753138413, + "learning_rate": 1.6601854877261617e-06, + "loss": 0.8126, + "num_input_tokens_seen": 204659360, + "step": 6295 + }, + { + "epoch": 0.5677954637687694, + "flos": 19290310029120.0, + "grad_norm": 2.1826133642653875, + "learning_rate": 1.6596097859926163e-06, + "loss": 0.7431, + "num_input_tokens_seen": 204684120, + "step": 6296 + }, + { + "epoch": 0.5678856472922397, + "flos": 17978196940320.0, + "grad_norm": 2.380801693825478, + "learning_rate": 1.6590341133030407e-06, + "loss": 0.8289, + "num_input_tokens_seen": 204710310, + "step": 6297 + }, + { + "epoch": 0.56797583081571, + "flos": 26866104989760.0, + "grad_norm": 1.815026528345865, + "learning_rate": 1.658458469706554e-06, + "loss": 0.8213, + "num_input_tokens_seen": 204743710, + "step": 6298 + }, + { + "epoch": 0.5680660143391802, + "flos": 21512965389600.0, + "grad_norm": 1.7245034069714469, + "learning_rate": 1.6578828552522746e-06, + "loss": 0.6993, + "num_input_tokens_seen": 204772500, + "step": 6299 + }, + { + "epoch": 0.5681561978626505, + "flos": 15319880359200.0, + "grad_norm": 1.8882171809383441, + "learning_rate": 1.6573072699893156e-06, + "loss": 0.7923, + "num_input_tokens_seen": 204798100, + "step": 6300 + }, + { + "epoch": 0.5682463813861207, + "flos": 24937954336800.0, + "grad_norm": 3.3721412984853827, + "learning_rate": 1.6567317139667906e-06, + "loss": 0.714, + "num_input_tokens_seen": 204826915, + "step": 6301 + }, + { + "epoch": 0.5683365649095911, + "flos": 24638468881440.0, + "grad_norm": 2.2207732311177275, + "learning_rate": 1.6561561872338087e-06, + "loss": 0.7602, + "num_input_tokens_seen": 204853075, + "step": 6302 + }, + { + "epoch": 0.5684267484330613, + "flos": 57726359689920.0, + "grad_norm": 2.3116137894255613, + "learning_rate": 1.6555806898394764e-06, + "loss": 0.6585, + "num_input_tokens_seen": 204886640, + "step": 6303 + }, + { + "epoch": 0.5685169319565315, + "flos": 25812138849600.0, + "grad_norm": 2.25294471474453, + "learning_rate": 1.6550052218328987e-06, + "loss": 0.716, + "num_input_tokens_seen": 204916440, + "step": 6304 + }, + { + "epoch": 0.5686071154800018, + "flos": 24573521843520.0, + "grad_norm": 2.522908441691349, + "learning_rate": 1.6544297832631777e-06, + "loss": 0.716, + "num_input_tokens_seen": 204945825, + "step": 6305 + }, + { + "epoch": 0.5686972990034721, + "flos": 26539963082880.0, + "grad_norm": 2.820738276838551, + "learning_rate": 1.6538543741794135e-06, + "loss": 0.75, + "num_input_tokens_seen": 204971680, + "step": 6306 + }, + { + "epoch": 0.5687874825269423, + "flos": 31017654398400.0, + "grad_norm": 1.8194252438620058, + "learning_rate": 1.6532789946307028e-06, + "loss": 0.7138, + "num_input_tokens_seen": 205004650, + "step": 6307 + }, + { + "epoch": 0.5688776660504126, + "flos": 68390365405920.0, + "grad_norm": 0.6343910168284145, + "learning_rate": 1.6527036446661393e-06, + "loss": 0.6124, + "num_input_tokens_seen": 205099290, + "step": 6308 + }, + { + "epoch": 0.5689678495738828, + "flos": 18853347866880.0, + "grad_norm": 1.6396045962953973, + "learning_rate": 1.6521283243348165e-06, + "loss": 0.7771, + "num_input_tokens_seen": 205126530, + "step": 6309 + }, + { + "epoch": 0.5690580330973531, + "flos": 21181879904640.0, + "grad_norm": 2.861784873154739, + "learning_rate": 1.6515530336858227e-06, + "loss": 0.7356, + "num_input_tokens_seen": 205155170, + "step": 6310 + }, + { + "epoch": 0.5691482166208234, + "flos": 18889891474560.0, + "grad_norm": 3.194508346434464, + "learning_rate": 1.6509777727682457e-06, + "loss": 0.7918, + "num_input_tokens_seen": 205184080, + "step": 6311 + }, + { + "epoch": 0.5692384001442936, + "flos": 21767544041280.0, + "grad_norm": 1.6132954372862864, + "learning_rate": 1.65040254163117e-06, + "loss": 0.7833, + "num_input_tokens_seen": 205212315, + "step": 6312 + }, + { + "epoch": 0.569328583667764, + "flos": 16121051996160.0, + "grad_norm": 4.393892314662657, + "learning_rate": 1.649827340323676e-06, + "loss": 0.6526, + "num_input_tokens_seen": 205238755, + "step": 6313 + }, + { + "epoch": 0.5694187671912342, + "flos": 26504162870400.0, + "grad_norm": 1.9236125999594877, + "learning_rate": 1.6492521688948454e-06, + "loss": 0.7108, + "num_input_tokens_seen": 205266870, + "step": 6314 + }, + { + "epoch": 0.5695089507147044, + "flos": 22239265662720.0, + "grad_norm": 2.212260036063366, + "learning_rate": 1.6486770273937526e-06, + "loss": 0.7641, + "num_input_tokens_seen": 205297635, + "step": 6315 + }, + { + "epoch": 0.5695991342381747, + "flos": 25555627370400.0, + "grad_norm": 1.5879571391646485, + "learning_rate": 1.6481019158694738e-06, + "loss": 0.7214, + "num_input_tokens_seen": 205329095, + "step": 6316 + }, + { + "epoch": 0.569689317761645, + "flos": 66933862034880.0, + "grad_norm": 0.7581461142597574, + "learning_rate": 1.6475268343710792e-06, + "loss": 0.6513, + "num_input_tokens_seen": 205420330, + "step": 6317 + }, + { + "epoch": 0.5697795012851152, + "flos": 19035285340320.0, + "grad_norm": 2.049962983146107, + "learning_rate": 1.6469517829476396e-06, + "loss": 0.6754, + "num_input_tokens_seen": 205447085, + "step": 6318 + }, + { + "epoch": 0.5698696848085855, + "flos": 20966409574080.0, + "grad_norm": 2.866975615361447, + "learning_rate": 1.64637676164822e-06, + "loss": 0.7259, + "num_input_tokens_seen": 205474855, + "step": 6319 + }, + { + "epoch": 0.5699598683320557, + "flos": 24974906811840.0, + "grad_norm": 1.6219430192564854, + "learning_rate": 1.6458017705218848e-06, + "loss": 0.7804, + "num_input_tokens_seen": 205507115, + "step": 6320 + }, + { + "epoch": 0.570050051855526, + "flos": 22015543645440.0, + "grad_norm": 1.7318679285197607, + "learning_rate": 1.645226809617696e-06, + "loss": 0.7689, + "num_input_tokens_seen": 205537075, + "step": 6321 + }, + { + "epoch": 0.5701402353789963, + "flos": 23292265389120.0, + "grad_norm": 2.0682011040351416, + "learning_rate": 1.6446518789847112e-06, + "loss": 0.7784, + "num_input_tokens_seen": 205563860, + "step": 6322 + }, + { + "epoch": 0.5702304189024665, + "flos": 17541048929280.0, + "grad_norm": 3.231472463715642, + "learning_rate": 1.6440769786719883e-06, + "loss": 0.7642, + "num_input_tokens_seen": 205588750, + "step": 6323 + }, + { + "epoch": 0.5703206024259367, + "flos": 26283934810560.0, + "grad_norm": 1.5993672281950726, + "learning_rate": 1.6435021087285803e-06, + "loss": 0.7845, + "num_input_tokens_seen": 205618285, + "step": 6324 + }, + { + "epoch": 0.5704107859494071, + "flos": 22417486160160.0, + "grad_norm": 2.025952635494796, + "learning_rate": 1.642927269203537e-06, + "loss": 0.784, + "num_input_tokens_seen": 205646150, + "step": 6325 + }, + { + "epoch": 0.5705009694728773, + "flos": 25003012884000.0, + "grad_norm": 2.736444703585967, + "learning_rate": 1.642352460145909e-06, + "loss": 0.7955, + "num_input_tokens_seen": 205674240, + "step": 6326 + }, + { + "epoch": 0.5705911529963476, + "flos": 28908755704320.0, + "grad_norm": 1.7775159815245531, + "learning_rate": 1.6417776816047402e-06, + "loss": 0.7524, + "num_input_tokens_seen": 205703900, + "step": 6327 + }, + { + "epoch": 0.5706813365198178, + "flos": 23006830103040.0, + "grad_norm": 2.3272883812963094, + "learning_rate": 1.6412029336290755e-06, + "loss": 0.7317, + "num_input_tokens_seen": 205730875, + "step": 6328 + }, + { + "epoch": 0.5707715200432881, + "flos": 20456025668640.0, + "grad_norm": 2.8066391808831965, + "learning_rate": 1.6406282162679551e-06, + "loss": 0.7431, + "num_input_tokens_seen": 205756680, + "step": 6329 + }, + { + "epoch": 0.5708617035667584, + "flos": 30765751969440.0, + "grad_norm": 2.9638132910565003, + "learning_rate": 1.6400535295704162e-06, + "loss": 0.7179, + "num_input_tokens_seen": 205788695, + "step": 6330 + }, + { + "epoch": 0.5709518870902286, + "flos": 22860098125920.0, + "grad_norm": 1.7199625073272173, + "learning_rate": 1.6394788735854955e-06, + "loss": 0.6836, + "num_input_tokens_seen": 205818635, + "step": 6331 + }, + { + "epoch": 0.5710420706136988, + "flos": 23547847624320.0, + "grad_norm": 2.507409844798265, + "learning_rate": 1.6389042483622246e-06, + "loss": 0.8124, + "num_input_tokens_seen": 205847555, + "step": 6332 + }, + { + "epoch": 0.5711322541371692, + "flos": 36558269766240.0, + "grad_norm": 4.062998814403559, + "learning_rate": 1.638329653949635e-06, + "loss": 0.7415, + "num_input_tokens_seen": 205878665, + "step": 6333 + }, + { + "epoch": 0.5712224376606394, + "flos": 18157272342240.0, + "grad_norm": 1.7441016659545117, + "learning_rate": 1.637755090396753e-06, + "loss": 0.7094, + "num_input_tokens_seen": 205905390, + "step": 6334 + }, + { + "epoch": 0.5713126211841096, + "flos": 69712031123040.0, + "grad_norm": 0.6493957058025898, + "learning_rate": 1.6371805577526039e-06, + "loss": 0.6101, + "num_input_tokens_seen": 206005600, + "step": 6335 + }, + { + "epoch": 0.5714028047075799, + "flos": 23335796911680.0, + "grad_norm": 2.065329604733875, + "learning_rate": 1.636606056066211e-06, + "loss": 0.6951, + "num_input_tokens_seen": 206033570, + "step": 6336 + }, + { + "epoch": 0.5714929882310502, + "flos": 23001700676160.0, + "grad_norm": 2.4108047972949613, + "learning_rate": 1.636031585386592e-06, + "loss": 0.6781, + "num_input_tokens_seen": 206060920, + "step": 6337 + }, + { + "epoch": 0.5715831717545204, + "flos": 23516767971360.0, + "grad_norm": 1.617155212964607, + "learning_rate": 1.635457145762766e-06, + "loss": 0.8042, + "num_input_tokens_seen": 206090440, + "step": 6338 + }, + { + "epoch": 0.5716733552779907, + "flos": 17823807992640.0, + "grad_norm": 2.4723689920337084, + "learning_rate": 1.6348827372437456e-06, + "loss": 0.7735, + "num_input_tokens_seen": 206113265, + "step": 6339 + }, + { + "epoch": 0.571763538801461, + "flos": 26463716437920.0, + "grad_norm": 1.8168566073709385, + "learning_rate": 1.634308359878544e-06, + "loss": 0.7475, + "num_input_tokens_seen": 206144485, + "step": 6340 + }, + { + "epoch": 0.5718537223249313, + "flos": 41222470435680.0, + "grad_norm": 1.872204413110768, + "learning_rate": 1.6337340137161695e-06, + "loss": 0.8243, + "num_input_tokens_seen": 206174345, + "step": 6341 + }, + { + "epoch": 0.5719439058484015, + "flos": 25769499401280.0, + "grad_norm": 2.1063821541731085, + "learning_rate": 1.6331596988056277e-06, + "loss": 0.7467, + "num_input_tokens_seen": 206201865, + "step": 6342 + }, + { + "epoch": 0.5720340893718717, + "flos": 24969814554720.0, + "grad_norm": 1.907961155553987, + "learning_rate": 1.632585415195924e-06, + "loss": 0.7758, + "num_input_tokens_seen": 206228955, + "step": 6343 + }, + { + "epoch": 0.5721242728953421, + "flos": 25483209210720.0, + "grad_norm": 2.0332677356373745, + "learning_rate": 1.6320111629360583e-06, + "loss": 0.7735, + "num_input_tokens_seen": 206261020, + "step": 6344 + }, + { + "epoch": 0.5722144564188123, + "flos": 21402702680640.0, + "grad_norm": 2.090623143399131, + "learning_rate": 1.631436942075029e-06, + "loss": 0.7222, + "num_input_tokens_seen": 206284380, + "step": 6345 + }, + { + "epoch": 0.5723046399422825, + "flos": 19144581635520.0, + "grad_norm": 2.601966162172009, + "learning_rate": 1.630862752661833e-06, + "loss": 0.7205, + "num_input_tokens_seen": 206311340, + "step": 6346 + }, + { + "epoch": 0.5723948234657528, + "flos": 46215675083520.0, + "grad_norm": 1.7995523160943656, + "learning_rate": 1.6302885947454612e-06, + "loss": 0.6721, + "num_input_tokens_seen": 206348290, + "step": 6347 + }, + { + "epoch": 0.5724850069892231, + "flos": 33061345866240.0, + "grad_norm": 2.1059562953750453, + "learning_rate": 1.6297144683749057e-06, + "loss": 0.7595, + "num_input_tokens_seen": 206376495, + "step": 6348 + }, + { + "epoch": 0.5725751905126933, + "flos": 19436112762240.0, + "grad_norm": 2.042348618051049, + "learning_rate": 1.629140373599153e-06, + "loss": 0.7557, + "num_input_tokens_seen": 206402270, + "step": 6349 + }, + { + "epoch": 0.5726653740361636, + "flos": 27995871737760.0, + "grad_norm": 2.4136807531837747, + "learning_rate": 1.628566310467189e-06, + "loss": 0.7614, + "num_input_tokens_seen": 206432625, + "step": 6350 + }, + { + "epoch": 0.5727555575596338, + "flos": 16375519138560.0, + "grad_norm": 2.141665908674806, + "learning_rate": 1.6279922790279957e-06, + "loss": 0.7102, + "num_input_tokens_seen": 206457685, + "step": 6351 + }, + { + "epoch": 0.5728457410831042, + "flos": 22861027369920.0, + "grad_norm": 2.2081453735206353, + "learning_rate": 1.6274182793305512e-06, + "loss": 0.7617, + "num_input_tokens_seen": 206485725, + "step": 6352 + }, + { + "epoch": 0.5729359246065744, + "flos": 15282593356320.0, + "grad_norm": 3.081153580277021, + "learning_rate": 1.626844311423835e-06, + "loss": 0.77, + "num_input_tokens_seen": 206512355, + "step": 6353 + }, + { + "epoch": 0.5730261081300446, + "flos": 24245038241760.0, + "grad_norm": 2.2319433678993215, + "learning_rate": 1.6262703753568181e-06, + "loss": 0.8258, + "num_input_tokens_seen": 206539385, + "step": 6354 + }, + { + "epoch": 0.5731162916535149, + "flos": 13607088527520.0, + "grad_norm": 2.1422917914514956, + "learning_rate": 1.6256964711784747e-06, + "loss": 0.8216, + "num_input_tokens_seen": 206565190, + "step": 6355 + }, + { + "epoch": 0.5732064751769852, + "flos": 25410084825600.0, + "grad_norm": 2.3671243489217155, + "learning_rate": 1.6251225989377723e-06, + "loss": 0.7622, + "num_input_tokens_seen": 206595045, + "step": 6356 + }, + { + "epoch": 0.5732966587004554, + "flos": 24682074743520.0, + "grad_norm": 1.9078186959830732, + "learning_rate": 1.624548758683676e-06, + "loss": 0.749, + "num_input_tokens_seen": 206621875, + "step": 6357 + }, + { + "epoch": 0.5733868422239257, + "flos": 20232972707040.0, + "grad_norm": 1.9826675638625666, + "learning_rate": 1.6239749504651505e-06, + "loss": 0.8423, + "num_input_tokens_seen": 206647675, + "step": 6358 + }, + { + "epoch": 0.5734770257473959, + "flos": 23151517743360.0, + "grad_norm": 1.8096838152266101, + "learning_rate": 1.6234011743311552e-06, + "loss": 0.7278, + "num_input_tokens_seen": 206677975, + "step": 6359 + }, + { + "epoch": 0.5735672092708662, + "flos": 20995890927360.0, + "grad_norm": 2.085323629168072, + "learning_rate": 1.6228274303306483e-06, + "loss": 0.6835, + "num_input_tokens_seen": 206706520, + "step": 6360 + }, + { + "epoch": 0.5736573927943365, + "flos": 24646014342720.0, + "grad_norm": 6.085021355245915, + "learning_rate": 1.6222537185125847e-06, + "loss": 0.8343, + "num_input_tokens_seen": 206735485, + "step": 6361 + }, + { + "epoch": 0.5737475763178067, + "flos": 22201086585600.0, + "grad_norm": 1.8608924379826381, + "learning_rate": 1.6216800389259172e-06, + "loss": 0.7687, + "num_input_tokens_seen": 206764290, + "step": 6362 + }, + { + "epoch": 0.573837759841277, + "flos": 26721974895840.0, + "grad_norm": 1.5119918852492464, + "learning_rate": 1.6211063916195949e-06, + "loss": 0.7534, + "num_input_tokens_seen": 206795650, + "step": 6363 + }, + { + "epoch": 0.5739279433647473, + "flos": 26066382973440.0, + "grad_norm": 2.0416915139483396, + "learning_rate": 1.6205327766425633e-06, + "loss": 0.7594, + "num_input_tokens_seen": 206822670, + "step": 6364 + }, + { + "epoch": 0.5740181268882175, + "flos": 30761365937760.0, + "grad_norm": 1.8222992496311627, + "learning_rate": 1.6199591940437689e-06, + "loss": 0.7615, + "num_input_tokens_seen": 206853560, + "step": 6365 + }, + { + "epoch": 0.5741083104116878, + "flos": 19181348261760.0, + "grad_norm": 1.7236877951349094, + "learning_rate": 1.6193856438721505e-06, + "loss": 0.7104, + "num_input_tokens_seen": 206882290, + "step": 6366 + }, + { + "epoch": 0.5741984939351581, + "flos": 21476050084320.0, + "grad_norm": 2.374060595593097, + "learning_rate": 1.6188121261766483e-06, + "loss": 0.7587, + "num_input_tokens_seen": 206910570, + "step": 6367 + }, + { + "epoch": 0.5742886774586283, + "flos": 31598077598880.0, + "grad_norm": 1.736613504410046, + "learning_rate": 1.6182386410061976e-06, + "loss": 0.6841, + "num_input_tokens_seen": 206940785, + "step": 6368 + }, + { + "epoch": 0.5743788609820986, + "flos": 23188581727680.0, + "grad_norm": 2.870327012840964, + "learning_rate": 1.61766518840973e-06, + "loss": 0.6595, + "num_input_tokens_seen": 206970395, + "step": 6369 + }, + { + "epoch": 0.5744690445055688, + "flos": 23552902711680.0, + "grad_norm": 2.1208691557111607, + "learning_rate": 1.6170917684361779e-06, + "loss": 0.6237, + "num_input_tokens_seen": 206997210, + "step": 6370 + }, + { + "epoch": 0.5745592280290391, + "flos": 23803467029280.0, + "grad_norm": 1.8474675217226935, + "learning_rate": 1.6165183811344662e-06, + "loss": 0.8283, + "num_input_tokens_seen": 207024805, + "step": 6371 + }, + { + "epoch": 0.5746494115525094, + "flos": 24500137270080.0, + "grad_norm": 3.2681098500796297, + "learning_rate": 1.6159450265535218e-06, + "loss": 0.7472, + "num_input_tokens_seen": 207053830, + "step": 6372 + }, + { + "epoch": 0.5747395950759796, + "flos": 21507687283680.0, + "grad_norm": 2.1766682157564516, + "learning_rate": 1.6153717047422652e-06, + "loss": 0.8353, + "num_input_tokens_seen": 207082910, + "step": 6373 + }, + { + "epoch": 0.5748297785994498, + "flos": 27305669035200.0, + "grad_norm": 2.7212434055597168, + "learning_rate": 1.6147984157496155e-06, + "loss": 0.6374, + "num_input_tokens_seen": 207114490, + "step": 6374 + }, + { + "epoch": 0.5749199621229202, + "flos": 24245372769600.0, + "grad_norm": 1.8555461024172373, + "learning_rate": 1.6142251596244886e-06, + "loss": 0.6776, + "num_input_tokens_seen": 207142820, + "step": 6375 + }, + { + "epoch": 0.5750101456463904, + "flos": 18052139060160.0, + "grad_norm": 1.610363702861302, + "learning_rate": 1.6136519364157983e-06, + "loss": 0.8053, + "num_input_tokens_seen": 207170355, + "step": 6376 + }, + { + "epoch": 0.5751003291698606, + "flos": 14444692262880.0, + "grad_norm": 2.117585655020277, + "learning_rate": 1.6130787461724555e-06, + "loss": 0.6936, + "num_input_tokens_seen": 207197795, + "step": 6377 + }, + { + "epoch": 0.5751905126933309, + "flos": 22530796789440.0, + "grad_norm": 1.647106627764304, + "learning_rate": 1.6125055889433679e-06, + "loss": 0.8289, + "num_input_tokens_seen": 207227075, + "step": 6378 + }, + { + "epoch": 0.5752806962168012, + "flos": 21181731225600.0, + "grad_norm": 1.6821448514982646, + "learning_rate": 1.6119324647774386e-06, + "loss": 0.7762, + "num_input_tokens_seen": 207255100, + "step": 6379 + }, + { + "epoch": 0.5753708797402715, + "flos": 20231820444480.0, + "grad_norm": 1.9464991982921611, + "learning_rate": 1.6113593737235724e-06, + "loss": 0.7596, + "num_input_tokens_seen": 207280285, + "step": 6380 + }, + { + "epoch": 0.5754610632637417, + "flos": 22714964448480.0, + "grad_norm": 2.9738972262881602, + "learning_rate": 1.6107863158306665e-06, + "loss": 0.6989, + "num_input_tokens_seen": 207308845, + "step": 6381 + }, + { + "epoch": 0.5755512467872119, + "flos": 15683234929440.0, + "grad_norm": 2.4530603679719736, + "learning_rate": 1.610213291147619e-06, + "loss": 0.6469, + "num_input_tokens_seen": 207334510, + "step": 6382 + }, + { + "epoch": 0.5756414303106823, + "flos": 21294298459680.0, + "grad_norm": 1.6967410911145193, + "learning_rate": 1.609640299723322e-06, + "loss": 0.7713, + "num_input_tokens_seen": 207363795, + "step": 6383 + }, + { + "epoch": 0.5757316138341525, + "flos": 22787605626720.0, + "grad_norm": 1.6292293969034486, + "learning_rate": 1.609067341606668e-06, + "loss": 0.7367, + "num_input_tokens_seen": 207392350, + "step": 6384 + }, + { + "epoch": 0.5758217973576227, + "flos": 23371076747520.0, + "grad_norm": 2.9216785058845964, + "learning_rate": 1.6084944168465438e-06, + "loss": 0.7444, + "num_input_tokens_seen": 207420415, + "step": 6385 + }, + { + "epoch": 0.5759119808810931, + "flos": 22781695634880.0, + "grad_norm": 1.8560594120505414, + "learning_rate": 1.6079215254918339e-06, + "loss": 0.7724, + "num_input_tokens_seen": 207446825, + "step": 6386 + }, + { + "epoch": 0.5760021644045633, + "flos": 30147001012800.0, + "grad_norm": 2.974215966782747, + "learning_rate": 1.6073486675914222e-06, + "loss": 0.6616, + "num_input_tokens_seen": 207474460, + "step": 6387 + }, + { + "epoch": 0.5760923479280335, + "flos": 16623332893920.0, + "grad_norm": 4.460646103729787, + "learning_rate": 1.606775843194187e-06, + "loss": 0.7798, + "num_input_tokens_seen": 207497105, + "step": 6388 + }, + { + "epoch": 0.5761825314515038, + "flos": 12580968271200.0, + "grad_norm": 2.5111739307842926, + "learning_rate": 1.6062030523490053e-06, + "loss": 0.7422, + "num_input_tokens_seen": 207518475, + "step": 6389 + }, + { + "epoch": 0.5762727149749741, + "flos": 18598620536160.0, + "grad_norm": 3.1695660986711123, + "learning_rate": 1.60563029510475e-06, + "loss": 0.6747, + "num_input_tokens_seen": 207543685, + "step": 6390 + }, + { + "epoch": 0.5763628984984444, + "flos": 67565064861120.0, + "grad_norm": 0.6500387757913156, + "learning_rate": 1.6050575715102927e-06, + "loss": 0.5684, + "num_input_tokens_seen": 207645705, + "step": 6391 + }, + { + "epoch": 0.5764530820219146, + "flos": 31892061929760.0, + "grad_norm": 2.0498526078248047, + "learning_rate": 1.6044848816145014e-06, + "loss": 0.8092, + "num_input_tokens_seen": 207673795, + "step": 6392 + }, + { + "epoch": 0.5765432655453848, + "flos": 17104569973920.0, + "grad_norm": 2.564240213636386, + "learning_rate": 1.60391222546624e-06, + "loss": 0.777, + "num_input_tokens_seen": 207700805, + "step": 6393 + }, + { + "epoch": 0.5766334490688552, + "flos": 24245186920800.0, + "grad_norm": 1.7254000932305442, + "learning_rate": 1.6033396031143725e-06, + "loss": 0.7449, + "num_input_tokens_seen": 207729245, + "step": 6394 + }, + { + "epoch": 0.5767236325923254, + "flos": 23837854790880.0, + "grad_norm": 2.1035721823510323, + "learning_rate": 1.602767014607757e-06, + "loss": 0.7524, + "num_input_tokens_seen": 207756835, + "step": 6395 + }, + { + "epoch": 0.5768138161157956, + "flos": 39402761173440.0, + "grad_norm": 1.6935626021314527, + "learning_rate": 1.6021944599952493e-06, + "loss": 0.7641, + "num_input_tokens_seen": 207789430, + "step": 6396 + }, + { + "epoch": 0.5769039996392659, + "flos": 20379332986560.0, + "grad_norm": 2.12097639739246, + "learning_rate": 1.6016219393257048e-06, + "loss": 0.7647, + "num_input_tokens_seen": 207815195, + "step": 6397 + }, + { + "epoch": 0.5769941831627362, + "flos": 19763964478080.0, + "grad_norm": 1.823072771926928, + "learning_rate": 1.6010494526479722e-06, + "loss": 0.702, + "num_input_tokens_seen": 207843000, + "step": 6398 + }, + { + "epoch": 0.5770843666862064, + "flos": 23334458800320.0, + "grad_norm": 2.957789375549298, + "learning_rate": 1.6004770000109006e-06, + "loss": 0.6958, + "num_input_tokens_seen": 207870535, + "step": 6399 + }, + { + "epoch": 0.5771745502096767, + "flos": 28214687346720.0, + "grad_norm": 2.124157233573328, + "learning_rate": 1.5999045814633348e-06, + "loss": 0.7173, + "num_input_tokens_seen": 207900725, + "step": 6400 + }, + { + "epoch": 0.5772647337331469, + "flos": 25958015922240.0, + "grad_norm": 1.876654403684853, + "learning_rate": 1.5993321970541151e-06, + "loss": 0.7598, + "num_input_tokens_seen": 207929940, + "step": 6401 + }, + { + "epoch": 0.5773549172566173, + "flos": 20565396303360.0, + "grad_norm": 2.0415053660714406, + "learning_rate": 1.5987598468320825e-06, + "loss": 0.7555, + "num_input_tokens_seen": 207956210, + "step": 6402 + }, + { + "epoch": 0.5774451007800875, + "flos": 39471982733760.0, + "grad_norm": 1.8284993236771028, + "learning_rate": 1.5981875308460717e-06, + "loss": 0.7359, + "num_input_tokens_seen": 207989250, + "step": 6403 + }, + { + "epoch": 0.5775352843035577, + "flos": 22824632441280.0, + "grad_norm": 2.197835212632468, + "learning_rate": 1.5976152491449169e-06, + "loss": 0.7363, + "num_input_tokens_seen": 208016700, + "step": 6404 + }, + { + "epoch": 0.577625467827028, + "flos": 20346394845600.0, + "grad_norm": 3.361850553850266, + "learning_rate": 1.5970430017774468e-06, + "loss": 0.8112, + "num_input_tokens_seen": 208043050, + "step": 6405 + }, + { + "epoch": 0.5777156513504983, + "flos": 38198754947520.0, + "grad_norm": 2.436806213237734, + "learning_rate": 1.5964707887924904e-06, + "loss": 0.7637, + "num_input_tokens_seen": 208078210, + "step": 6406 + }, + { + "epoch": 0.5778058348739685, + "flos": 36848871648960.0, + "grad_norm": 1.571742986212909, + "learning_rate": 1.5958986102388714e-06, + "loss": 0.6364, + "num_input_tokens_seen": 208109240, + "step": 6407 + }, + { + "epoch": 0.5778960183974388, + "flos": 37178693362080.0, + "grad_norm": 1.765318177157184, + "learning_rate": 1.5953264661654104e-06, + "loss": 0.7057, + "num_input_tokens_seen": 208138595, + "step": 6408 + }, + { + "epoch": 0.5779862019209091, + "flos": 25226697731520.0, + "grad_norm": 1.819258683633046, + "learning_rate": 1.5947543566209276e-06, + "loss": 0.7856, + "num_input_tokens_seen": 208168150, + "step": 6409 + }, + { + "epoch": 0.5780763854443793, + "flos": 23626138606080.0, + "grad_norm": 2.1037604314372063, + "learning_rate": 1.5941822816542367e-06, + "loss": 0.7336, + "num_input_tokens_seen": 208197340, + "step": 6410 + }, + { + "epoch": 0.5781665689678496, + "flos": 30766272346080.0, + "grad_norm": 2.6590581075309494, + "learning_rate": 1.5936102413141519e-06, + "loss": 0.7071, + "num_input_tokens_seen": 208229700, + "step": 6411 + }, + { + "epoch": 0.5782567524913198, + "flos": 23735137543200.0, + "grad_norm": 2.5177366504226266, + "learning_rate": 1.5930382356494823e-06, + "loss": 0.6714, + "num_input_tokens_seen": 208257135, + "step": 6412 + }, + { + "epoch": 0.5783469360147901, + "flos": 23735583580320.0, + "grad_norm": 2.0960025342387247, + "learning_rate": 1.5924662647090335e-06, + "loss": 0.799, + "num_input_tokens_seen": 208285420, + "step": 6413 + }, + { + "epoch": 0.5784371195382604, + "flos": 22531056977760.0, + "grad_norm": 2.1003636908085688, + "learning_rate": 1.5918943285416108e-06, + "loss": 0.7712, + "num_input_tokens_seen": 208312720, + "step": 6414 + }, + { + "epoch": 0.5785273030617306, + "flos": 40675134055200.0, + "grad_norm": 1.748872666327467, + "learning_rate": 1.5913224271960139e-06, + "loss": 0.6859, + "num_input_tokens_seen": 208348435, + "step": 6415 + }, + { + "epoch": 0.5786174865852008, + "flos": 36117479118720.0, + "grad_norm": 2.4171089776076187, + "learning_rate": 1.590750560721041e-06, + "loss": 0.7633, + "num_input_tokens_seen": 208379435, + "step": 6416 + }, + { + "epoch": 0.5787076701086712, + "flos": 52043026679040.0, + "grad_norm": 1.4538436261218357, + "learning_rate": 1.5901787291654874e-06, + "loss": 0.7145, + "num_input_tokens_seen": 208412495, + "step": 6417 + }, + { + "epoch": 0.5787978536321414, + "flos": 46357760840640.0, + "grad_norm": 1.885444036745255, + "learning_rate": 1.5896069325781435e-06, + "loss": 0.6213, + "num_input_tokens_seen": 208447970, + "step": 6418 + }, + { + "epoch": 0.5788880371556117, + "flos": 24246822390240.0, + "grad_norm": 1.802253700391188, + "learning_rate": 1.5890351710077998e-06, + "loss": 0.7873, + "num_input_tokens_seen": 208476930, + "step": 6419 + }, + { + "epoch": 0.5789782206790819, + "flos": 22892590229760.0, + "grad_norm": 1.7909489975977708, + "learning_rate": 1.5884634445032406e-06, + "loss": 0.8082, + "num_input_tokens_seen": 208506485, + "step": 6420 + }, + { + "epoch": 0.5790684042025522, + "flos": 69683293164960.0, + "grad_norm": 0.6053217063286418, + "learning_rate": 1.5878917531132501e-06, + "loss": 0.5974, + "num_input_tokens_seen": 208612895, + "step": 6421 + }, + { + "epoch": 0.5791585877260225, + "flos": 23079285432480.0, + "grad_norm": 2.0807847519676312, + "learning_rate": 1.5873200968866077e-06, + "loss": 0.6238, + "num_input_tokens_seen": 208640285, + "step": 6422 + }, + { + "epoch": 0.5792487712494927, + "flos": 25337517986880.0, + "grad_norm": 2.9541659664983597, + "learning_rate": 1.586748475872089e-06, + "loss": 0.7223, + "num_input_tokens_seen": 208666520, + "step": 6423 + }, + { + "epoch": 0.5793389547729629, + "flos": 27815309545440.0, + "grad_norm": 2.6958837815939107, + "learning_rate": 1.58617689011847e-06, + "loss": 0.6168, + "num_input_tokens_seen": 208696500, + "step": 6424 + }, + { + "epoch": 0.5794291382964333, + "flos": 23953507115040.0, + "grad_norm": 2.17270562185453, + "learning_rate": 1.5856053396745198e-06, + "loss": 0.7845, + "num_input_tokens_seen": 208726755, + "step": 6425 + }, + { + "epoch": 0.5795193218199035, + "flos": 20343607113600.0, + "grad_norm": 2.196546757405975, + "learning_rate": 1.5850338245890078e-06, + "loss": 0.8147, + "num_input_tokens_seen": 208752600, + "step": 6426 + }, + { + "epoch": 0.5796095053433737, + "flos": 32806804384320.0, + "grad_norm": 2.700449683571193, + "learning_rate": 1.5844623449106974e-06, + "loss": 0.733, + "num_input_tokens_seen": 208780575, + "step": 6427 + }, + { + "epoch": 0.579699688866844, + "flos": 29379994118880.0, + "grad_norm": 1.9199914774109514, + "learning_rate": 1.583890900688351e-06, + "loss": 0.7908, + "num_input_tokens_seen": 208811490, + "step": 6428 + }, + { + "epoch": 0.5797898723903143, + "flos": 13643371946880.0, + "grad_norm": 2.8812261419073906, + "learning_rate": 1.583319491970728e-06, + "loss": 0.8298, + "num_input_tokens_seen": 208836070, + "step": 6429 + }, + { + "epoch": 0.5798800559137846, + "flos": 24352513218720.0, + "grad_norm": 2.092974163189388, + "learning_rate": 1.5827481188065828e-06, + "loss": 0.7651, + "num_input_tokens_seen": 208867075, + "step": 6430 + }, + { + "epoch": 0.5799702394372548, + "flos": 18741932895360.0, + "grad_norm": 2.194642531499026, + "learning_rate": 1.5821767812446689e-06, + "loss": 0.718, + "num_input_tokens_seen": 208895590, + "step": 6431 + }, + { + "epoch": 0.5800604229607251, + "flos": 20310520293600.0, + "grad_norm": 1.99571944159339, + "learning_rate": 1.581605479333736e-06, + "loss": 0.8648, + "num_input_tokens_seen": 208922705, + "step": 6432 + }, + { + "epoch": 0.5801506064841954, + "flos": 22459568062080.0, + "grad_norm": 5.775419687321197, + "learning_rate": 1.5810342131225308e-06, + "loss": 0.7569, + "num_input_tokens_seen": 208950925, + "step": 6433 + }, + { + "epoch": 0.5802407900076656, + "flos": 18705872494560.0, + "grad_norm": 3.1326846431983015, + "learning_rate": 1.580462982659797e-06, + "loss": 0.79, + "num_input_tokens_seen": 208977440, + "step": 6434 + }, + { + "epoch": 0.5803309735311358, + "flos": 11093422416960.0, + "grad_norm": 2.458100830549014, + "learning_rate": 1.5798917879942736e-06, + "loss": 0.8235, + "num_input_tokens_seen": 209003420, + "step": 6435 + }, + { + "epoch": 0.5804211570546062, + "flos": 24281098642560.0, + "grad_norm": 1.841099595960051, + "learning_rate": 1.5793206291747006e-06, + "loss": 0.6912, + "num_input_tokens_seen": 209033920, + "step": 6436 + }, + { + "epoch": 0.5805113405780764, + "flos": 22935452696640.0, + "grad_norm": 1.4672178027646448, + "learning_rate": 1.57874950624981e-06, + "loss": 0.6869, + "num_input_tokens_seen": 209063120, + "step": 6437 + }, + { + "epoch": 0.5806015241015466, + "flos": 22642917986400.0, + "grad_norm": 1.683277711169649, + "learning_rate": 1.5781784192683351e-06, + "loss": 0.7474, + "num_input_tokens_seen": 209094175, + "step": 6438 + }, + { + "epoch": 0.5806917076250169, + "flos": 57033332085600.0, + "grad_norm": 1.40156601615133, + "learning_rate": 1.5776073682790033e-06, + "loss": 0.7096, + "num_input_tokens_seen": 209130850, + "step": 6439 + }, + { + "epoch": 0.5807818911484872, + "flos": 33207668976000.0, + "grad_norm": 2.0368585477845196, + "learning_rate": 1.5770363533305393e-06, + "loss": 0.7752, + "num_input_tokens_seen": 209162175, + "step": 6440 + }, + { + "epoch": 0.5808720746719575, + "flos": 35611964451840.0, + "grad_norm": 2.0350615905889025, + "learning_rate": 1.5764653744716665e-06, + "loss": 0.7512, + "num_input_tokens_seen": 209190180, + "step": 6441 + }, + { + "epoch": 0.5809622581954277, + "flos": 22387707448800.0, + "grad_norm": 2.140287832172841, + "learning_rate": 1.575894431751103e-06, + "loss": 0.7786, + "num_input_tokens_seen": 209215165, + "step": 6442 + }, + { + "epoch": 0.5810524417188979, + "flos": 21512482182720.0, + "grad_norm": 1.9476360248077114, + "learning_rate": 1.575323525217565e-06, + "loss": 0.7779, + "num_input_tokens_seen": 209240980, + "step": 6443 + }, + { + "epoch": 0.5811426252423683, + "flos": 21216118987200.0, + "grad_norm": 2.1748714496726738, + "learning_rate": 1.574752654919766e-06, + "loss": 0.8325, + "num_input_tokens_seen": 209269740, + "step": 6444 + }, + { + "epoch": 0.5812328087658385, + "flos": 15975100584000.0, + "grad_norm": 2.209233790178565, + "learning_rate": 1.5741818209064146e-06, + "loss": 0.7212, + "num_input_tokens_seen": 209295850, + "step": 6445 + }, + { + "epoch": 0.5813229922893087, + "flos": 36661953427680.0, + "grad_norm": 2.517741479893521, + "learning_rate": 1.5736110232262183e-06, + "loss": 0.7601, + "num_input_tokens_seen": 209326075, + "step": 6446 + }, + { + "epoch": 0.581413175812779, + "flos": 21180839151360.0, + "grad_norm": 1.9892183804618426, + "learning_rate": 1.5730402619278804e-06, + "loss": 0.7939, + "num_input_tokens_seen": 209353870, + "step": 6447 + }, + { + "epoch": 0.5815033593362493, + "flos": 65962833096480.0, + "grad_norm": 0.6673342049893426, + "learning_rate": 1.5724695370601024e-06, + "loss": 0.5554, + "num_input_tokens_seen": 209446630, + "step": 6448 + }, + { + "epoch": 0.5815935428597195, + "flos": 21184630466880.0, + "grad_norm": 1.975154390215762, + "learning_rate": 1.5718988486715798e-06, + "loss": 0.6715, + "num_input_tokens_seen": 209476485, + "step": 6449 + }, + { + "epoch": 0.5816837263831898, + "flos": 25193164874400.0, + "grad_norm": 1.8795505616495123, + "learning_rate": 1.5713281968110087e-06, + "loss": 0.732, + "num_input_tokens_seen": 209505510, + "step": 6450 + }, + { + "epoch": 0.58177390990666, + "flos": 25041229130880.0, + "grad_norm": 1.9413017997045205, + "learning_rate": 1.5707575815270796e-06, + "loss": 0.7328, + "num_input_tokens_seen": 209534260, + "step": 6451 + }, + { + "epoch": 0.5818640934301303, + "flos": 20820792689760.0, + "grad_norm": 2.1429148212118085, + "learning_rate": 1.57018700286848e-06, + "loss": 0.7516, + "num_input_tokens_seen": 209562570, + "step": 6452 + }, + { + "epoch": 0.5819542769536006, + "flos": 14809013246880.0, + "grad_norm": 2.299196691792868, + "learning_rate": 1.5696164608838956e-06, + "loss": 0.7657, + "num_input_tokens_seen": 209587430, + "step": 6453 + }, + { + "epoch": 0.5820444604770708, + "flos": 26831494209600.0, + "grad_norm": 1.8173576898493633, + "learning_rate": 1.5690459556220073e-06, + "loss": 0.7705, + "num_input_tokens_seen": 209616865, + "step": 6454 + }, + { + "epoch": 0.582134644000541, + "flos": 22533249993600.0, + "grad_norm": 1.9529671567078242, + "learning_rate": 1.5684754871314949e-06, + "loss": 0.8095, + "num_input_tokens_seen": 209644995, + "step": 6455 + }, + { + "epoch": 0.5822248275240114, + "flos": 18634086220800.0, + "grad_norm": 1.9312154303854763, + "learning_rate": 1.5679050554610335e-06, + "loss": 0.7519, + "num_input_tokens_seen": 209670190, + "step": 6456 + }, + { + "epoch": 0.5823150110474816, + "flos": 25301346076800.0, + "grad_norm": 2.3273050797378465, + "learning_rate": 1.567334660659295e-06, + "loss": 0.7539, + "num_input_tokens_seen": 209700590, + "step": 6457 + }, + { + "epoch": 0.5824051945709519, + "flos": 37761086559840.0, + "grad_norm": 1.772150253095518, + "learning_rate": 1.5667643027749488e-06, + "loss": 0.7675, + "num_input_tokens_seen": 209732115, + "step": 6458 + }, + { + "epoch": 0.5824953780944222, + "flos": 28543765664640.0, + "grad_norm": 2.3933999903790704, + "learning_rate": 1.5661939818566614e-06, + "loss": 0.816, + "num_input_tokens_seen": 209760895, + "step": 6459 + }, + { + "epoch": 0.5825855616178924, + "flos": 27993901740480.0, + "grad_norm": 2.0138061360278496, + "learning_rate": 1.5656236979530956e-06, + "loss": 0.6158, + "num_input_tokens_seen": 209790770, + "step": 6460 + }, + { + "epoch": 0.5826757451413627, + "flos": 24493781241120.0, + "grad_norm": 2.0937164076202706, + "learning_rate": 1.5650534511129106e-06, + "loss": 0.6906, + "num_input_tokens_seen": 209819930, + "step": 6461 + }, + { + "epoch": 0.5827659286648329, + "flos": 20711422055040.0, + "grad_norm": 2.2537054703005874, + "learning_rate": 1.5644832413847635e-06, + "loss": 0.7334, + "num_input_tokens_seen": 209847645, + "step": 6462 + }, + { + "epoch": 0.5828561121883032, + "flos": 21766280269440.0, + "grad_norm": 2.4329839098453583, + "learning_rate": 1.5639130688173082e-06, + "loss": 0.7912, + "num_input_tokens_seen": 209876155, + "step": 6463 + }, + { + "epoch": 0.5829462957117735, + "flos": 23586175380480.0, + "grad_norm": 1.78724631530066, + "learning_rate": 1.5633429334591932e-06, + "loss": 0.7289, + "num_input_tokens_seen": 209904415, + "step": 6464 + }, + { + "epoch": 0.5830364792352437, + "flos": 18342889621920.0, + "grad_norm": 2.0599338867170625, + "learning_rate": 1.562772835359068e-06, + "loss": 0.7854, + "num_input_tokens_seen": 209929770, + "step": 6465 + }, + { + "epoch": 0.5831266627587139, + "flos": 19072237815360.0, + "grad_norm": 3.4266915759308096, + "learning_rate": 1.5622027745655753e-06, + "loss": 0.7208, + "num_input_tokens_seen": 209955960, + "step": 6466 + }, + { + "epoch": 0.5832168462821843, + "flos": 19690951602240.0, + "grad_norm": 1.8301291280537972, + "learning_rate": 1.561632751127355e-06, + "loss": 0.7064, + "num_input_tokens_seen": 209984200, + "step": 6467 + }, + { + "epoch": 0.5833070298056545, + "flos": 21002507144640.0, + "grad_norm": 2.2979700501351967, + "learning_rate": 1.561062765093046e-06, + "loss": 0.7921, + "num_input_tokens_seen": 210010535, + "step": 6468 + }, + { + "epoch": 0.5833972133291248, + "flos": 19943746105440.0, + "grad_norm": 40.550037509134164, + "learning_rate": 1.5604928165112817e-06, + "loss": 0.7404, + "num_input_tokens_seen": 210038940, + "step": 6469 + }, + { + "epoch": 0.583487396852595, + "flos": 26431187164320.0, + "grad_norm": 2.0733318839522976, + "learning_rate": 1.5599229054306945e-06, + "loss": 0.7422, + "num_input_tokens_seen": 210070180, + "step": 6470 + }, + { + "epoch": 0.5835775803760653, + "flos": 22569867940800.0, + "grad_norm": 2.099270389604601, + "learning_rate": 1.5593530318999111e-06, + "loss": 0.7451, + "num_input_tokens_seen": 210097300, + "step": 6471 + }, + { + "epoch": 0.5836677638995356, + "flos": 22787977324320.0, + "grad_norm": 1.745724769196908, + "learning_rate": 1.5587831959675572e-06, + "loss": 0.7757, + "num_input_tokens_seen": 210124575, + "step": 6472 + }, + { + "epoch": 0.5837579474230058, + "flos": 37392714072000.0, + "grad_norm": 2.574038509374267, + "learning_rate": 1.5582133976822534e-06, + "loss": 0.7346, + "num_input_tokens_seen": 210157190, + "step": 6473 + }, + { + "epoch": 0.583848130946476, + "flos": 59909869559520.0, + "grad_norm": 1.7490993634500687, + "learning_rate": 1.5576436370926185e-06, + "loss": 0.7072, + "num_input_tokens_seen": 210190525, + "step": 6474 + }, + { + "epoch": 0.5839383144699464, + "flos": 21512519352480.0, + "grad_norm": 1.6945491074025212, + "learning_rate": 1.5570739142472679e-06, + "loss": 0.7323, + "num_input_tokens_seen": 210219920, + "step": 6475 + }, + { + "epoch": 0.5840284979934166, + "flos": 24901856766240.0, + "grad_norm": 1.855964289297009, + "learning_rate": 1.5565042291948127e-06, + "loss": 0.8157, + "num_input_tokens_seen": 210250475, + "step": 6476 + }, + { + "epoch": 0.5841186815168868, + "flos": 19691955185760.0, + "grad_norm": 2.1871182789853805, + "learning_rate": 1.5559345819838624e-06, + "loss": 0.7851, + "num_input_tokens_seen": 210278515, + "step": 6477 + }, + { + "epoch": 0.5842088650403571, + "flos": 23696177901120.0, + "grad_norm": 1.6460889384412578, + "learning_rate": 1.5553649726630226e-06, + "loss": 0.7333, + "num_input_tokens_seen": 210308610, + "step": 6478 + }, + { + "epoch": 0.5842990485638274, + "flos": 20163974165280.0, + "grad_norm": 2.2064244879433543, + "learning_rate": 1.5547954012808942e-06, + "loss": 0.6826, + "num_input_tokens_seen": 210338550, + "step": 6479 + }, + { + "epoch": 0.5843892320872976, + "flos": 23763801161760.0, + "grad_norm": 2.1522588972753436, + "learning_rate": 1.5542258678860776e-06, + "loss": 0.7276, + "num_input_tokens_seen": 210367215, + "step": 6480 + }, + { + "epoch": 0.5844794156107679, + "flos": 24020349810720.0, + "grad_norm": 2.2981441390978556, + "learning_rate": 1.553656372527167e-06, + "loss": 0.8729, + "num_input_tokens_seen": 210394805, + "step": 6481 + }, + { + "epoch": 0.5845695991342382, + "flos": 22314360045120.0, + "grad_norm": 3.0372878828669903, + "learning_rate": 1.5530869152527568e-06, + "loss": 0.8458, + "num_input_tokens_seen": 210422320, + "step": 6482 + }, + { + "epoch": 0.5846597826577085, + "flos": 25738085220480.0, + "grad_norm": 1.7620960178431346, + "learning_rate": 1.5525174961114353e-06, + "loss": 0.768, + "num_input_tokens_seen": 210452570, + "step": 6483 + }, + { + "epoch": 0.5847499661811787, + "flos": 18044667938400.0, + "grad_norm": 1.8365690403150516, + "learning_rate": 1.5519481151517875e-06, + "loss": 0.7832, + "num_input_tokens_seen": 210478145, + "step": 6484 + }, + { + "epoch": 0.5848401497046489, + "flos": 27197041795680.0, + "grad_norm": 1.944156778340129, + "learning_rate": 1.551378772422398e-06, + "loss": 0.7532, + "num_input_tokens_seen": 210508615, + "step": 6485 + }, + { + "epoch": 0.5849303332281193, + "flos": 28759161655680.0, + "grad_norm": 1.9844374759480998, + "learning_rate": 1.5508094679718447e-06, + "loss": 0.7645, + "num_input_tokens_seen": 210538095, + "step": 6486 + }, + { + "epoch": 0.5850205167515895, + "flos": 27959402469600.0, + "grad_norm": 2.475968154203253, + "learning_rate": 1.5502402018487048e-06, + "loss": 0.6533, + "num_input_tokens_seen": 210570505, + "step": 6487 + }, + { + "epoch": 0.5851107002750597, + "flos": 28289149843200.0, + "grad_norm": 2.208835856523624, + "learning_rate": 1.54967097410155e-06, + "loss": 0.7552, + "num_input_tokens_seen": 210599870, + "step": 6488 + }, + { + "epoch": 0.58520088379853, + "flos": 31640568368160.0, + "grad_norm": 1.7811559685766887, + "learning_rate": 1.5491017847789519e-06, + "loss": 0.6701, + "num_input_tokens_seen": 210630380, + "step": 6489 + }, + { + "epoch": 0.5852910673220003, + "flos": 27079902681120.0, + "grad_norm": 2.3265946241918902, + "learning_rate": 1.5485326339294755e-06, + "loss": 0.6937, + "num_input_tokens_seen": 210659365, + "step": 6490 + }, + { + "epoch": 0.5853812508454705, + "flos": 21001540730880.0, + "grad_norm": 2.0917022712179802, + "learning_rate": 1.5479635216016832e-06, + "loss": 0.7191, + "num_input_tokens_seen": 210688270, + "step": 6491 + }, + { + "epoch": 0.5854714343689408, + "flos": 29382633171840.0, + "grad_norm": 1.9437411906051942, + "learning_rate": 1.547394447844137e-06, + "loss": 0.6669, + "num_input_tokens_seen": 210716785, + "step": 6492 + }, + { + "epoch": 0.585561617892411, + "flos": 54047200958400.0, + "grad_norm": 1.721048448185462, + "learning_rate": 1.546825412705391e-06, + "loss": 0.6674, + "num_input_tokens_seen": 210751405, + "step": 6493 + }, + { + "epoch": 0.5856518014158814, + "flos": 21545160135360.0, + "grad_norm": 1.993795845198227, + "learning_rate": 1.5462564162340007e-06, + "loss": 0.7787, + "num_input_tokens_seen": 210780255, + "step": 6494 + }, + { + "epoch": 0.5857419849393516, + "flos": 14546294417760.0, + "grad_norm": 2.0744934474853536, + "learning_rate": 1.5456874584785144e-06, + "loss": 0.7546, + "num_input_tokens_seen": 210804585, + "step": 6495 + }, + { + "epoch": 0.5858321684628218, + "flos": 34081890658560.0, + "grad_norm": 5.329051007210493, + "learning_rate": 1.5451185394874785e-06, + "loss": 0.7251, + "num_input_tokens_seen": 210833360, + "step": 6496 + }, + { + "epoch": 0.5859223519862921, + "flos": 22788534870720.0, + "grad_norm": 2.028656227826253, + "learning_rate": 1.5445496593094381e-06, + "loss": 0.7621, + "num_input_tokens_seen": 210863000, + "step": 6497 + }, + { + "epoch": 0.5860125355097624, + "flos": 19071643099200.0, + "grad_norm": 2.01847460495171, + "learning_rate": 1.5439808179929316e-06, + "loss": 0.7655, + "num_input_tokens_seen": 210888285, + "step": 6498 + }, + { + "epoch": 0.5861027190332326, + "flos": 19581841155840.0, + "grad_norm": 2.1418173038135255, + "learning_rate": 1.543412015586496e-06, + "loss": 0.7115, + "num_input_tokens_seen": 210917385, + "step": 6499 + }, + { + "epoch": 0.5861929025567029, + "flos": 31382904626400.0, + "grad_norm": 1.8387945791891944, + "learning_rate": 1.5428432521386655e-06, + "loss": 0.7156, + "num_input_tokens_seen": 210947885, + "step": 6500 + }, + { + "epoch": 0.5862830860801731, + "flos": 29125415467200.0, + "grad_norm": 2.247825419424502, + "learning_rate": 1.5422745276979688e-06, + "loss": 0.7861, + "num_input_tokens_seen": 210978045, + "step": 6501 + }, + { + "epoch": 0.5863732696036434, + "flos": 21433968182400.0, + "grad_norm": 1.7337041021211053, + "learning_rate": 1.5417058423129336e-06, + "loss": 0.7422, + "num_input_tokens_seen": 211004070, + "step": 6502 + }, + { + "epoch": 0.5864634531271137, + "flos": 26212631743680.0, + "grad_norm": 2.4356074688921434, + "learning_rate": 1.5411371960320822e-06, + "loss": 0.7856, + "num_input_tokens_seen": 211034495, + "step": 6503 + }, + { + "epoch": 0.5865536366505839, + "flos": 26213746836480.0, + "grad_norm": 1.6061905400889296, + "learning_rate": 1.5405685889039363e-06, + "loss": 0.7846, + "num_input_tokens_seen": 211064025, + "step": 6504 + }, + { + "epoch": 0.5866438201740543, + "flos": 30618573955200.0, + "grad_norm": 2.1695460724590436, + "learning_rate": 1.5400000209770118e-06, + "loss": 0.7482, + "num_input_tokens_seen": 211094890, + "step": 6505 + }, + { + "epoch": 0.5867340036975245, + "flos": 33091793633280.0, + "grad_norm": 2.153660532398794, + "learning_rate": 1.5394314922998208e-06, + "loss": 0.761, + "num_input_tokens_seen": 211123300, + "step": 6506 + }, + { + "epoch": 0.5868241872209947, + "flos": 24499133686560.0, + "grad_norm": 2.3329231946249416, + "learning_rate": 1.5388630029208756e-06, + "loss": 0.8002, + "num_input_tokens_seen": 211145540, + "step": 6507 + }, + { + "epoch": 0.586914370744465, + "flos": 35499768915360.0, + "grad_norm": 1.8435683080083376, + "learning_rate": 1.5382945528886806e-06, + "loss": 0.7742, + "num_input_tokens_seen": 211180120, + "step": 6508 + }, + { + "epoch": 0.5870045542679353, + "flos": 22344844981920.0, + "grad_norm": 2.2667882629578027, + "learning_rate": 1.5377261422517412e-06, + "loss": 0.737, + "num_input_tokens_seen": 211205795, + "step": 6509 + }, + { + "epoch": 0.5870947377914055, + "flos": 20092187891520.0, + "grad_norm": 1.8932564447913596, + "learning_rate": 1.5371577710585553e-06, + "loss": 0.7086, + "num_input_tokens_seen": 211233725, + "step": 6510 + }, + { + "epoch": 0.5871849213148758, + "flos": 70880571664320.0, + "grad_norm": 0.8420978901879137, + "learning_rate": 1.536589439357621e-06, + "loss": 0.5646, + "num_input_tokens_seen": 211332340, + "step": 6511 + }, + { + "epoch": 0.587275104838346, + "flos": 34078954247520.0, + "grad_norm": 1.8645580084020361, + "learning_rate": 1.5360211471974315e-06, + "loss": 0.754, + "num_input_tokens_seen": 211361720, + "step": 6512 + }, + { + "epoch": 0.5873652883618163, + "flos": 21329652635040.0, + "grad_norm": 2.4797448731523355, + "learning_rate": 1.5354528946264753e-06, + "loss": 0.8035, + "num_input_tokens_seen": 211389445, + "step": 6513 + }, + { + "epoch": 0.5874554718852866, + "flos": 37470707695680.0, + "grad_norm": 1.814972013883021, + "learning_rate": 1.5348846816932399e-06, + "loss": 0.6844, + "num_input_tokens_seen": 211419920, + "step": 6514 + }, + { + "epoch": 0.5875456554087568, + "flos": 68196230517600.0, + "grad_norm": 0.6300998284326201, + "learning_rate": 1.5343165084462077e-06, + "loss": 0.5735, + "num_input_tokens_seen": 211511895, + "step": 6515 + }, + { + "epoch": 0.587635838932227, + "flos": 21804682365120.0, + "grad_norm": 2.6616263339673925, + "learning_rate": 1.5337483749338595e-06, + "loss": 0.7021, + "num_input_tokens_seen": 211539590, + "step": 6516 + }, + { + "epoch": 0.5877260224556974, + "flos": 25367185188960.0, + "grad_norm": 2.1808522165035096, + "learning_rate": 1.5331802812046708e-06, + "loss": 0.7197, + "num_input_tokens_seen": 211566150, + "step": 6517 + }, + { + "epoch": 0.5878162059791676, + "flos": 38526755342400.0, + "grad_norm": 1.7698388242884548, + "learning_rate": 1.5326122273071133e-06, + "loss": 0.5997, + "num_input_tokens_seen": 211598505, + "step": 6518 + }, + { + "epoch": 0.5879063895026378, + "flos": 21433224787200.0, + "grad_norm": 1.8612611324266877, + "learning_rate": 1.532044213289659e-06, + "loss": 0.7875, + "num_input_tokens_seen": 211624295, + "step": 6519 + }, + { + "epoch": 0.5879965730261081, + "flos": 28544583399360.0, + "grad_norm": 1.8106070532154581, + "learning_rate": 1.5314762392007718e-06, + "loss": 0.6357, + "num_input_tokens_seen": 211655170, + "step": 6520 + }, + { + "epoch": 0.5880867565495784, + "flos": 24169014615360.0, + "grad_norm": 2.035991709551767, + "learning_rate": 1.530908305088916e-06, + "loss": 0.7992, + "num_input_tokens_seen": 211684815, + "step": 6521 + }, + { + "epoch": 0.5881769400730487, + "flos": 25410976899840.0, + "grad_norm": 2.0068707465427007, + "learning_rate": 1.5303404110025501e-06, + "loss": 0.7436, + "num_input_tokens_seen": 211714195, + "step": 6522 + }, + { + "epoch": 0.5882671235965189, + "flos": 23331708238080.0, + "grad_norm": 1.5714548994311115, + "learning_rate": 1.5297725569901293e-06, + "loss": 0.7067, + "num_input_tokens_seen": 211744165, + "step": 6523 + }, + { + "epoch": 0.5883573071199891, + "flos": 30395297975040.0, + "grad_norm": 1.8534268035348473, + "learning_rate": 1.5292047431001077e-06, + "loss": 0.6477, + "num_input_tokens_seen": 211770650, + "step": 6524 + }, + { + "epoch": 0.5884474906434595, + "flos": 62135344088160.0, + "grad_norm": 0.6711773244134581, + "learning_rate": 1.5286369693809321e-06, + "loss": 0.5882, + "num_input_tokens_seen": 211860470, + "step": 6525 + }, + { + "epoch": 0.5885376741669297, + "flos": 17650642582560.0, + "grad_norm": 3.184278917594472, + "learning_rate": 1.5280692358810506e-06, + "loss": 0.7592, + "num_input_tokens_seen": 211887805, + "step": 6526 + }, + { + "epoch": 0.5886278576903999, + "flos": 22169300707200.0, + "grad_norm": 2.0044195696327427, + "learning_rate": 1.527501542648904e-06, + "loss": 0.6991, + "num_input_tokens_seen": 211916430, + "step": 6527 + }, + { + "epoch": 0.5887180412138703, + "flos": 24858473922720.0, + "grad_norm": 1.9531401366600956, + "learning_rate": 1.5269338897329308e-06, + "loss": 0.6565, + "num_input_tokens_seen": 211943740, + "step": 6528 + }, + { + "epoch": 0.5888082247373405, + "flos": 28686743496000.0, + "grad_norm": 2.326262477900735, + "learning_rate": 1.5263662771815662e-06, + "loss": 0.7755, + "num_input_tokens_seen": 211972245, + "step": 6529 + }, + { + "epoch": 0.5888984082608107, + "flos": 30065996638560.0, + "grad_norm": 2.4110993900199733, + "learning_rate": 1.5257987050432429e-06, + "loss": 0.8242, + "num_input_tokens_seen": 212000705, + "step": 6530 + }, + { + "epoch": 0.588988591784281, + "flos": 24208903501440.0, + "grad_norm": 4.419914754675551, + "learning_rate": 1.5252311733663887e-06, + "loss": 0.6538, + "num_input_tokens_seen": 212030140, + "step": 6531 + }, + { + "epoch": 0.5890787753077513, + "flos": 16193841853440.0, + "grad_norm": 3.6019739673103746, + "learning_rate": 1.5246636821994281e-06, + "loss": 0.6921, + "num_input_tokens_seen": 212056590, + "step": 6532 + }, + { + "epoch": 0.5891689588312216, + "flos": 31820089807200.0, + "grad_norm": 1.9084776697448982, + "learning_rate": 1.524096231590784e-06, + "loss": 0.7964, + "num_input_tokens_seen": 212087880, + "step": 6533 + }, + { + "epoch": 0.5892591423546918, + "flos": 23188767576480.0, + "grad_norm": 1.885345717582783, + "learning_rate": 1.5235288215888736e-06, + "loss": 0.7939, + "num_input_tokens_seen": 212113960, + "step": 6534 + }, + { + "epoch": 0.589349325878162, + "flos": 24937136602080.0, + "grad_norm": 2.3567118993875953, + "learning_rate": 1.5229614522421102e-06, + "loss": 0.7543, + "num_input_tokens_seen": 212139170, + "step": 6535 + }, + { + "epoch": 0.5894395094016324, + "flos": 18197049719040.0, + "grad_norm": 2.0203113340535235, + "learning_rate": 1.5223941235989071e-06, + "loss": 0.5881, + "num_input_tokens_seen": 212167185, + "step": 6536 + }, + { + "epoch": 0.5895296929251026, + "flos": 24643040761920.0, + "grad_norm": 2.1303498859211114, + "learning_rate": 1.52182683570767e-06, + "loss": 0.7195, + "num_input_tokens_seen": 212197725, + "step": 6537 + }, + { + "epoch": 0.5896198764485728, + "flos": 15568065812160.0, + "grad_norm": 1.8342143311720693, + "learning_rate": 1.5212595886168046e-06, + "loss": 0.7718, + "num_input_tokens_seen": 212223375, + "step": 6538 + }, + { + "epoch": 0.5897100599720431, + "flos": 24203105018880.0, + "grad_norm": 1.7114714046388517, + "learning_rate": 1.520692382374711e-06, + "loss": 0.7629, + "num_input_tokens_seen": 212252080, + "step": 6539 + }, + { + "epoch": 0.5898002434955134, + "flos": 24676462109760.0, + "grad_norm": 2.375250150657451, + "learning_rate": 1.5201252170297854e-06, + "loss": 0.7744, + "num_input_tokens_seen": 212279505, + "step": 6540 + }, + { + "epoch": 0.5898904270189836, + "flos": 26394457707840.0, + "grad_norm": 2.2410355110742337, + "learning_rate": 1.5195580926304232e-06, + "loss": 0.6378, + "num_input_tokens_seen": 212309830, + "step": 6541 + }, + { + "epoch": 0.5899806105424539, + "flos": 17865481027200.0, + "grad_norm": 2.5322229846556867, + "learning_rate": 1.5189910092250131e-06, + "loss": 0.7684, + "num_input_tokens_seen": 212336605, + "step": 6542 + }, + { + "epoch": 0.5900707940659241, + "flos": 32187830409120.0, + "grad_norm": 2.462995558703341, + "learning_rate": 1.5184239668619427e-06, + "loss": 0.7592, + "num_input_tokens_seen": 212366515, + "step": 6543 + }, + { + "epoch": 0.5901609775893945, + "flos": 30213880878240.0, + "grad_norm": 1.781006445044669, + "learning_rate": 1.5178569655895946e-06, + "loss": 0.7495, + "num_input_tokens_seen": 212395725, + "step": 6544 + }, + { + "epoch": 0.5902511611128647, + "flos": 21658842462240.0, + "grad_norm": 1.5252531529412587, + "learning_rate": 1.5172900054563487e-06, + "loss": 0.7799, + "num_input_tokens_seen": 212425820, + "step": 6545 + }, + { + "epoch": 0.5903413446363349, + "flos": 22459791080640.0, + "grad_norm": 2.5550465929295725, + "learning_rate": 1.5167230865105814e-06, + "loss": 0.768, + "num_input_tokens_seen": 212453970, + "step": 6546 + }, + { + "epoch": 0.5904315281598052, + "flos": 31204200922080.0, + "grad_norm": 2.5136797474320685, + "learning_rate": 1.5161562088006644e-06, + "loss": 0.6667, + "num_input_tokens_seen": 212483150, + "step": 6547 + }, + { + "epoch": 0.5905217116832755, + "flos": 23116200737760.0, + "grad_norm": 2.2303046652170013, + "learning_rate": 1.5155893723749685e-06, + "loss": 0.774, + "num_input_tokens_seen": 212510615, + "step": 6548 + }, + { + "epoch": 0.5906118952067457, + "flos": 27560061838080.0, + "grad_norm": 1.8645722657058124, + "learning_rate": 1.5150225772818582e-06, + "loss": 0.7557, + "num_input_tokens_seen": 212541285, + "step": 6549 + }, + { + "epoch": 0.590702078730216, + "flos": 26359363720800.0, + "grad_norm": 1.971467277342364, + "learning_rate": 1.5144558235696949e-06, + "loss": 0.8008, + "num_input_tokens_seen": 212572485, + "step": 6550 + }, + { + "epoch": 0.5907922622536863, + "flos": 28762952971200.0, + "grad_norm": 1.7674882223122657, + "learning_rate": 1.5138891112868388e-06, + "loss": 0.6666, + "num_input_tokens_seen": 212604265, + "step": 6551 + }, + { + "epoch": 0.5908824457771565, + "flos": 71924687818080.0, + "grad_norm": 0.6940332066633197, + "learning_rate": 1.5133224404816433e-06, + "loss": 0.5781, + "num_input_tokens_seen": 212696925, + "step": 6552 + }, + { + "epoch": 0.5909726293006268, + "flos": 19466895057120.0, + "grad_norm": 1.8785262328268997, + "learning_rate": 1.5127558112024617e-06, + "loss": 0.6776, + "num_input_tokens_seen": 212722600, + "step": 6553 + }, + { + "epoch": 0.591062812824097, + "flos": 69248003641920.0, + "grad_norm": 0.6135807642328781, + "learning_rate": 1.5121892234976404e-06, + "loss": 0.5918, + "num_input_tokens_seen": 212822580, + "step": 6554 + }, + { + "epoch": 0.5911529963475673, + "flos": 20739788315520.0, + "grad_norm": 1.9393652889341733, + "learning_rate": 1.5116226774155243e-06, + "loss": 0.7359, + "num_input_tokens_seen": 212847770, + "step": 6555 + }, + { + "epoch": 0.5912431798710376, + "flos": 21366716619360.0, + "grad_norm": 3.515472001879135, + "learning_rate": 1.5110561730044547e-06, + "loss": 0.7208, + "num_input_tokens_seen": 212875880, + "step": 6556 + }, + { + "epoch": 0.5913333633945078, + "flos": 18379098701760.0, + "grad_norm": 1.8797030276246185, + "learning_rate": 1.510489710312768e-06, + "loss": 0.7508, + "num_input_tokens_seen": 212903885, + "step": 6557 + }, + { + "epoch": 0.591423546917978, + "flos": 22497003744000.0, + "grad_norm": 2.251102918938637, + "learning_rate": 1.5099232893887987e-06, + "loss": 0.7417, + "num_input_tokens_seen": 212935205, + "step": 6558 + }, + { + "epoch": 0.5915137304414484, + "flos": 31929757800000.0, + "grad_norm": 1.9542589481691068, + "learning_rate": 1.5093569102808758e-06, + "loss": 0.7195, + "num_input_tokens_seen": 212966575, + "step": 6559 + }, + { + "epoch": 0.5916039139649186, + "flos": 38854607058240.0, + "grad_norm": 2.222061909085257, + "learning_rate": 1.5087905730373275e-06, + "loss": 0.7534, + "num_input_tokens_seen": 212998070, + "step": 6560 + }, + { + "epoch": 0.5916940974883889, + "flos": 71139387668640.0, + "grad_norm": 0.617233033251035, + "learning_rate": 1.508224277706476e-06, + "loss": 0.59, + "num_input_tokens_seen": 213098825, + "step": 6561 + }, + { + "epoch": 0.5917842810118591, + "flos": 23553237239520.0, + "grad_norm": 4.990832889743916, + "learning_rate": 1.5076580243366399e-06, + "loss": 0.7036, + "num_input_tokens_seen": 213127245, + "step": 6562 + }, + { + "epoch": 0.5918744645353294, + "flos": 52588950608640.0, + "grad_norm": 2.5159749800525133, + "learning_rate": 1.507091812976137e-06, + "loss": 0.6564, + "num_input_tokens_seen": 213161175, + "step": 6563 + }, + { + "epoch": 0.5919646480587997, + "flos": 29053703532960.0, + "grad_norm": 1.9367807960804548, + "learning_rate": 1.5065256436732773e-06, + "loss": 0.6976, + "num_input_tokens_seen": 213192075, + "step": 6564 + }, + { + "epoch": 0.5920548315822699, + "flos": 15312520746720.0, + "grad_norm": 2.759559834252702, + "learning_rate": 1.5059595164763717e-06, + "loss": 0.6766, + "num_input_tokens_seen": 213217800, + "step": 6565 + }, + { + "epoch": 0.5921450151057401, + "flos": 56796960882720.0, + "grad_norm": 0.5774124238875973, + "learning_rate": 1.5053934314337243e-06, + "loss": 0.5849, + "num_input_tokens_seen": 213312885, + "step": 6566 + }, + { + "epoch": 0.5922351986292105, + "flos": 30363995303520.0, + "grad_norm": 1.6007075563207749, + "learning_rate": 1.5048273885936356e-06, + "loss": 0.737, + "num_input_tokens_seen": 213343285, + "step": 6567 + }, + { + "epoch": 0.5923253821526807, + "flos": 23479778326560.0, + "grad_norm": 1.5757099216792545, + "learning_rate": 1.5042613880044053e-06, + "loss": 0.6638, + "num_input_tokens_seen": 213376115, + "step": 6568 + }, + { + "epoch": 0.592415565676151, + "flos": 19655002710720.0, + "grad_norm": 1.7705467964673611, + "learning_rate": 1.5036954297143265e-06, + "loss": 0.7746, + "num_input_tokens_seen": 213401410, + "step": 6569 + }, + { + "epoch": 0.5925057491996212, + "flos": 39147253277760.0, + "grad_norm": 2.230277298762694, + "learning_rate": 1.50312951377169e-06, + "loss": 0.7012, + "num_input_tokens_seen": 213435305, + "step": 6570 + }, + { + "epoch": 0.5925959327230915, + "flos": 64183384417920.0, + "grad_norm": 0.5906190387485543, + "learning_rate": 1.502563640224784e-06, + "loss": 0.5874, + "num_input_tokens_seen": 213531180, + "step": 6571 + }, + { + "epoch": 0.5926861162465618, + "flos": 26064375806400.0, + "grad_norm": 4.421159774036513, + "learning_rate": 1.5019978091218903e-06, + "loss": 0.7442, + "num_input_tokens_seen": 213560215, + "step": 6572 + }, + { + "epoch": 0.592776299770032, + "flos": 23773353790080.0, + "grad_norm": 1.501200354626332, + "learning_rate": 1.50143202051129e-06, + "loss": 0.8071, + "num_input_tokens_seen": 213591030, + "step": 6573 + }, + { + "epoch": 0.5928664832935022, + "flos": 23438700008160.0, + "grad_norm": 1.8805510049286251, + "learning_rate": 1.500866274441258e-06, + "loss": 0.8002, + "num_input_tokens_seen": 213616360, + "step": 6574 + }, + { + "epoch": 0.5929566668169726, + "flos": 24968996820000.0, + "grad_norm": 2.469823579704036, + "learning_rate": 1.5003005709600682e-06, + "loss": 0.73, + "num_input_tokens_seen": 213645915, + "step": 6575 + }, + { + "epoch": 0.5930468503404428, + "flos": 20420113946880.0, + "grad_norm": 2.364075683234268, + "learning_rate": 1.4997349101159885e-06, + "loss": 0.8233, + "num_input_tokens_seen": 213673590, + "step": 6576 + }, + { + "epoch": 0.593137033863913, + "flos": 18889519776960.0, + "grad_norm": 2.3073945262935602, + "learning_rate": 1.4991692919572854e-06, + "loss": 0.7031, + "num_input_tokens_seen": 213700135, + "step": 6577 + }, + { + "epoch": 0.5932272173873834, + "flos": 23915067849600.0, + "grad_norm": 1.9661154906771534, + "learning_rate": 1.4986037165322199e-06, + "loss": 0.7633, + "num_input_tokens_seen": 213726980, + "step": 6578 + }, + { + "epoch": 0.5933174009108536, + "flos": 70298550164160.0, + "grad_norm": 0.6703705244581777, + "learning_rate": 1.498038183889049e-06, + "loss": 0.6144, + "num_input_tokens_seen": 213814760, + "step": 6579 + }, + { + "epoch": 0.5934075844343238, + "flos": 24573150145920.0, + "grad_norm": 1.8088135554833633, + "learning_rate": 1.4974726940760292e-06, + "loss": 0.6659, + "num_input_tokens_seen": 213844550, + "step": 6580 + }, + { + "epoch": 0.5934977679577941, + "flos": 21913681302240.0, + "grad_norm": 1.957350957952678, + "learning_rate": 1.496907247141409e-06, + "loss": 0.8173, + "num_input_tokens_seen": 213872075, + "step": 6581 + }, + { + "epoch": 0.5935879514812644, + "flos": 69961740536160.0, + "grad_norm": 0.7438027723807938, + "learning_rate": 1.4963418431334372e-06, + "loss": 0.656, + "num_input_tokens_seen": 213956755, + "step": 6582 + }, + { + "epoch": 0.5936781350047347, + "flos": 19763555610720.0, + "grad_norm": 2.175015710357402, + "learning_rate": 1.4957764821003566e-06, + "loss": 0.7285, + "num_input_tokens_seen": 213983905, + "step": 6583 + }, + { + "epoch": 0.5937683185282049, + "flos": 21911785644480.0, + "grad_norm": 2.0625154899246696, + "learning_rate": 1.4952111640904063e-06, + "loss": 0.7498, + "num_input_tokens_seen": 214011985, + "step": 6584 + }, + { + "epoch": 0.5938585020516751, + "flos": 18926881119360.0, + "grad_norm": 2.0287992511630555, + "learning_rate": 1.494645889151823e-06, + "loss": 0.73, + "num_input_tokens_seen": 214040685, + "step": 6585 + }, + { + "epoch": 0.5939486855751455, + "flos": 65990604640800.0, + "grad_norm": 0.652670863935218, + "learning_rate": 1.494080657332839e-06, + "loss": 0.5605, + "num_input_tokens_seen": 214128285, + "step": 6586 + }, + { + "epoch": 0.5940388690986157, + "flos": 23552419504800.0, + "grad_norm": 2.0406562744361207, + "learning_rate": 1.4935154686816832e-06, + "loss": 0.7046, + "num_input_tokens_seen": 214156535, + "step": 6587 + }, + { + "epoch": 0.5941290526220859, + "flos": 22532952635520.0, + "grad_norm": 5.1396694976589705, + "learning_rate": 1.4929503232465802e-06, + "loss": 0.8248, + "num_input_tokens_seen": 214184325, + "step": 6588 + }, + { + "epoch": 0.5942192361455562, + "flos": 19944675349440.0, + "grad_norm": 2.994879007586211, + "learning_rate": 1.492385221075751e-06, + "loss": 0.652, + "num_input_tokens_seen": 214213025, + "step": 6589 + }, + { + "epoch": 0.5943094196690265, + "flos": 19291610970720.0, + "grad_norm": 2.3681860098854126, + "learning_rate": 1.4918201622174142e-06, + "loss": 0.7796, + "num_input_tokens_seen": 214243220, + "step": 6590 + }, + { + "epoch": 0.5943996031924967, + "flos": 21182697639360.0, + "grad_norm": 1.903511519244655, + "learning_rate": 1.4912551467197827e-06, + "loss": 0.768, + "num_input_tokens_seen": 214271750, + "step": 6591 + }, + { + "epoch": 0.594489786715967, + "flos": 15829223511360.0, + "grad_norm": 3.573827703377735, + "learning_rate": 1.4906901746310678e-06, + "loss": 0.8306, + "num_input_tokens_seen": 214298535, + "step": 6592 + }, + { + "epoch": 0.5945799702394372, + "flos": 30181128586080.0, + "grad_norm": 1.7214447397785402, + "learning_rate": 1.4901252459994757e-06, + "loss": 0.7068, + "num_input_tokens_seen": 214332460, + "step": 6593 + }, + { + "epoch": 0.5946701537629075, + "flos": 20018989166880.0, + "grad_norm": 2.0035281056926286, + "learning_rate": 1.489560360873208e-06, + "loss": 0.7312, + "num_input_tokens_seen": 214358210, + "step": 6594 + }, + { + "epoch": 0.5947603372863778, + "flos": 23989121478720.0, + "grad_norm": 1.9412758244198713, + "learning_rate": 1.4889955193004659e-06, + "loss": 0.7211, + "num_input_tokens_seen": 214386815, + "step": 6595 + }, + { + "epoch": 0.594850520809848, + "flos": 58231534095360.0, + "grad_norm": 0.6360843751273196, + "learning_rate": 1.4884307213294428e-06, + "loss": 0.5746, + "num_input_tokens_seen": 214477700, + "step": 6596 + }, + { + "epoch": 0.5949407043333182, + "flos": 25519938667200.0, + "grad_norm": 3.1451788383543366, + "learning_rate": 1.4878659670083321e-06, + "loss": 0.7317, + "num_input_tokens_seen": 214504570, + "step": 6597 + }, + { + "epoch": 0.5950308878567886, + "flos": 29017940490240.0, + "grad_norm": 1.7697846018661603, + "learning_rate": 1.4873012563853208e-06, + "loss": 0.6674, + "num_input_tokens_seen": 214532920, + "step": 6598 + }, + { + "epoch": 0.5951210713802588, + "flos": 12914841488160.0, + "grad_norm": 2.609606247209033, + "learning_rate": 1.4867365895085935e-06, + "loss": 0.7537, + "num_input_tokens_seen": 214559360, + "step": 6599 + }, + { + "epoch": 0.5952112549037291, + "flos": 25590944376000.0, + "grad_norm": 1.5854921514292186, + "learning_rate": 1.4861719664263301e-06, + "loss": 0.7272, + "num_input_tokens_seen": 214589710, + "step": 6600 + }, + { + "epoch": 0.5953014384271994, + "flos": 23771866999680.0, + "grad_norm": 2.112405253435626, + "learning_rate": 1.485607387186708e-06, + "loss": 0.8516, + "num_input_tokens_seen": 214615420, + "step": 6601 + }, + { + "epoch": 0.5953916219506696, + "flos": 25041117621600.0, + "grad_norm": 2.349938119925814, + "learning_rate": 1.4850428518379001e-06, + "loss": 0.7423, + "num_input_tokens_seen": 214640850, + "step": 6602 + }, + { + "epoch": 0.5954818054741399, + "flos": 23803095331680.0, + "grad_norm": 1.9122444081772667, + "learning_rate": 1.4844783604280746e-06, + "loss": 0.7391, + "num_input_tokens_seen": 214671930, + "step": 6603 + }, + { + "epoch": 0.5955719889976101, + "flos": 33097926643680.0, + "grad_norm": 2.967974430971465, + "learning_rate": 1.483913913005399e-06, + "loss": 0.6886, + "num_input_tokens_seen": 214699645, + "step": 6604 + }, + { + "epoch": 0.5956621725210804, + "flos": 32041061262240.0, + "grad_norm": 2.0492487531087873, + "learning_rate": 1.483349509618034e-06, + "loss": 0.8072, + "num_input_tokens_seen": 214728005, + "step": 6605 + }, + { + "epoch": 0.5957523560445507, + "flos": 32806581365760.0, + "grad_norm": 2.4370567701299413, + "learning_rate": 1.4827851503141367e-06, + "loss": 0.6811, + "num_input_tokens_seen": 214757810, + "step": 6606 + }, + { + "epoch": 0.5958425395680209, + "flos": 23990719778400.0, + "grad_norm": 2.302984507195144, + "learning_rate": 1.482220835141863e-06, + "loss": 0.6729, + "num_input_tokens_seen": 214788145, + "step": 6607 + }, + { + "epoch": 0.5959327230914911, + "flos": 24755607996000.0, + "grad_norm": 2.1152959164783587, + "learning_rate": 1.481656564149362e-06, + "loss": 0.8228, + "num_input_tokens_seen": 214817360, + "step": 6608 + }, + { + "epoch": 0.5960229066149615, + "flos": 20632238999040.0, + "grad_norm": 1.9843338489020228, + "learning_rate": 1.4810923373847818e-06, + "loss": 0.7478, + "num_input_tokens_seen": 214844290, + "step": 6609 + }, + { + "epoch": 0.5961130901384317, + "flos": 20275017439200.0, + "grad_norm": 2.149616708522892, + "learning_rate": 1.4805281548962647e-06, + "loss": 0.7522, + "num_input_tokens_seen": 214873245, + "step": 6610 + }, + { + "epoch": 0.596203273661902, + "flos": 24388313431200.0, + "grad_norm": 2.4651951365581213, + "learning_rate": 1.4799640167319488e-06, + "loss": 0.6892, + "num_input_tokens_seen": 214902775, + "step": 6611 + }, + { + "epoch": 0.5962934571853722, + "flos": 34006647597120.0, + "grad_norm": 1.9194660130656993, + "learning_rate": 1.4793999229399714e-06, + "loss": 0.7312, + "num_input_tokens_seen": 214933550, + "step": 6612 + }, + { + "epoch": 0.5963836407088425, + "flos": 25261122662880.0, + "grad_norm": 3.006881292285413, + "learning_rate": 1.4788358735684626e-06, + "loss": 0.8112, + "num_input_tokens_seen": 214962570, + "step": 6613 + }, + { + "epoch": 0.5964738242323128, + "flos": 23188767576480.0, + "grad_norm": 1.8986618474571453, + "learning_rate": 1.4782718686655514e-06, + "loss": 0.8139, + "num_input_tokens_seen": 214992120, + "step": 6614 + }, + { + "epoch": 0.596564007755783, + "flos": 33207929164320.0, + "grad_norm": 2.2772157086713545, + "learning_rate": 1.4777079082793605e-06, + "loss": 0.7956, + "num_input_tokens_seen": 215021575, + "step": 6615 + }, + { + "epoch": 0.5966541912792532, + "flos": 52152137125440.0, + "grad_norm": 2.0349097514467815, + "learning_rate": 1.4771439924580108e-06, + "loss": 0.7181, + "num_input_tokens_seen": 215054545, + "step": 6616 + }, + { + "epoch": 0.5967443748027236, + "flos": 19761920141280.0, + "grad_norm": 2.496203400736142, + "learning_rate": 1.4765801212496189e-06, + "loss": 0.788, + "num_input_tokens_seen": 215082525, + "step": 6617 + }, + { + "epoch": 0.5968345583261938, + "flos": 23079545620800.0, + "grad_norm": 2.1655166092427014, + "learning_rate": 1.4760162947022968e-06, + "loss": 0.6634, + "num_input_tokens_seen": 215111760, + "step": 6618 + }, + { + "epoch": 0.596924741849664, + "flos": 26063446562400.0, + "grad_norm": 1.738382469402035, + "learning_rate": 1.475452512864154e-06, + "loss": 0.7369, + "num_input_tokens_seen": 215141550, + "step": 6619 + }, + { + "epoch": 0.5970149253731343, + "flos": 33823966728480.0, + "grad_norm": 1.6366011911931668, + "learning_rate": 1.4748887757832945e-06, + "loss": 0.7547, + "num_input_tokens_seen": 215173370, + "step": 6620 + }, + { + "epoch": 0.5971051088966046, + "flos": 38417421877440.0, + "grad_norm": 1.8703675583054848, + "learning_rate": 1.4743250835078209e-06, + "loss": 0.7805, + "num_input_tokens_seen": 215204535, + "step": 6621 + }, + { + "epoch": 0.5971952924200749, + "flos": 29121772830720.0, + "grad_norm": 1.6580809064021251, + "learning_rate": 1.4737614360858297e-06, + "loss": 0.7041, + "num_input_tokens_seen": 215237090, + "step": 6622 + }, + { + "epoch": 0.5972854759435451, + "flos": 19435815404160.0, + "grad_norm": 2.2641968533896644, + "learning_rate": 1.4731978335654133e-06, + "loss": 0.702, + "num_input_tokens_seen": 215263720, + "step": 6623 + }, + { + "epoch": 0.5973756594670154, + "flos": 30907391689440.0, + "grad_norm": 2.4981145424584583, + "learning_rate": 1.4726342759946638e-06, + "loss": 0.6852, + "num_input_tokens_seen": 215292635, + "step": 6624 + }, + { + "epoch": 0.5974658429904857, + "flos": 22387856127840.0, + "grad_norm": 2.266862412957563, + "learning_rate": 1.4720707634216653e-06, + "loss": 0.7428, + "num_input_tokens_seen": 215323090, + "step": 6625 + }, + { + "epoch": 0.5975560265139559, + "flos": 29672789017440.0, + "grad_norm": 2.3311664760650705, + "learning_rate": 1.4715072958945e-06, + "loss": 0.7528, + "num_input_tokens_seen": 215352625, + "step": 6626 + }, + { + "epoch": 0.5976462100374261, + "flos": 22600575896160.0, + "grad_norm": 2.101993105426083, + "learning_rate": 1.470943873461247e-06, + "loss": 0.7445, + "num_input_tokens_seen": 215381630, + "step": 6627 + }, + { + "epoch": 0.5977363935608965, + "flos": 20310148596000.0, + "grad_norm": 4.576874337556388, + "learning_rate": 1.470380496169979e-06, + "loss": 0.7061, + "num_input_tokens_seen": 215409295, + "step": 6628 + }, + { + "epoch": 0.5978265770843667, + "flos": 24318571494240.0, + "grad_norm": 1.8437342885527943, + "learning_rate": 1.4698171640687682e-06, + "loss": 0.7394, + "num_input_tokens_seen": 215437450, + "step": 6629 + }, + { + "epoch": 0.5979167606078369, + "flos": 24536569368480.0, + "grad_norm": 1.7213135761163298, + "learning_rate": 1.4692538772056792e-06, + "loss": 0.6806, + "num_input_tokens_seen": 215466600, + "step": 6630 + }, + { + "epoch": 0.5980069441313072, + "flos": 35648582399040.0, + "grad_norm": 1.566746929155344, + "learning_rate": 1.4686906356287772e-06, + "loss": 0.6582, + "num_input_tokens_seen": 215498645, + "step": 6631 + }, + { + "epoch": 0.5980971276547775, + "flos": 22060450449120.0, + "grad_norm": 1.6299507579503423, + "learning_rate": 1.4681274393861194e-06, + "loss": 0.7781, + "num_input_tokens_seen": 215527905, + "step": 6632 + }, + { + "epoch": 0.5981873111782477, + "flos": 18482038968000.0, + "grad_norm": 2.048380174569431, + "learning_rate": 1.4675642885257603e-06, + "loss": 0.8416, + "num_input_tokens_seen": 215553940, + "step": 6633 + }, + { + "epoch": 0.598277494701718, + "flos": 16558869062880.0, + "grad_norm": 2.1803540813674043, + "learning_rate": 1.4670011830957529e-06, + "loss": 0.8318, + "num_input_tokens_seen": 215580065, + "step": 6634 + }, + { + "epoch": 0.5983676782251882, + "flos": 22017587982240.0, + "grad_norm": 2.7497287096242755, + "learning_rate": 1.4664381231441427e-06, + "loss": 0.7097, + "num_input_tokens_seen": 215604470, + "step": 6635 + }, + { + "epoch": 0.5984578617486586, + "flos": 25447632016800.0, + "grad_norm": 2.0138042417546167, + "learning_rate": 1.4658751087189746e-06, + "loss": 0.7571, + "num_input_tokens_seen": 215633765, + "step": 6636 + }, + { + "epoch": 0.5985480452721288, + "flos": 25265062657440.0, + "grad_norm": 3.0001052202210277, + "learning_rate": 1.4653121398682874e-06, + "loss": 0.7956, + "num_input_tokens_seen": 215664115, + "step": 6637 + }, + { + "epoch": 0.598638228795599, + "flos": 25847604534240.0, + "grad_norm": 1.443666417728656, + "learning_rate": 1.4647492166401159e-06, + "loss": 0.6887, + "num_input_tokens_seen": 215697575, + "step": 6638 + }, + { + "epoch": 0.5987284123190693, + "flos": 19690988772000.0, + "grad_norm": 2.301229281106868, + "learning_rate": 1.4641863390824934e-06, + "loss": 0.7315, + "num_input_tokens_seen": 215725375, + "step": 6639 + }, + { + "epoch": 0.5988185958425396, + "flos": 29567135358720.0, + "grad_norm": 2.4487109471586015, + "learning_rate": 1.4636235072434465e-06, + "loss": 0.7355, + "num_input_tokens_seen": 215759495, + "step": 6640 + }, + { + "epoch": 0.5989087793660098, + "flos": 26826104594400.0, + "grad_norm": 1.5896513351487136, + "learning_rate": 1.4630607211709994e-06, + "loss": 0.7725, + "num_input_tokens_seen": 215788360, + "step": 6641 + }, + { + "epoch": 0.5989989628894801, + "flos": 22604032683840.0, + "grad_norm": 2.493288950737583, + "learning_rate": 1.4624979809131723e-06, + "loss": 0.7531, + "num_input_tokens_seen": 215817320, + "step": 6642 + }, + { + "epoch": 0.5990891464129503, + "flos": 19544293964640.0, + "grad_norm": 2.0262701172156548, + "learning_rate": 1.4619352865179814e-06, + "loss": 0.6557, + "num_input_tokens_seen": 215845675, + "step": 6643 + }, + { + "epoch": 0.5991793299364206, + "flos": 20310817651680.0, + "grad_norm": 2.6374339850241006, + "learning_rate": 1.4613726380334391e-06, + "loss": 0.7157, + "num_input_tokens_seen": 215873490, + "step": 6644 + }, + { + "epoch": 0.5992695134598909, + "flos": 34003153639680.0, + "grad_norm": 2.429876289515298, + "learning_rate": 1.4608100355075522e-06, + "loss": 0.6002, + "num_input_tokens_seen": 215903230, + "step": 6645 + }, + { + "epoch": 0.5993596969833611, + "flos": 66799619097120.0, + "grad_norm": 0.6215381591848617, + "learning_rate": 1.460247478988327e-06, + "loss": 0.5557, + "num_input_tokens_seen": 215995655, + "step": 6646 + }, + { + "epoch": 0.5994498805068315, + "flos": 23188470218400.0, + "grad_norm": 1.8049643167999927, + "learning_rate": 1.4596849685237623e-06, + "loss": 0.7609, + "num_input_tokens_seen": 216024715, + "step": 6647 + }, + { + "epoch": 0.5995400640303017, + "flos": 30944046806400.0, + "grad_norm": 1.5603764408700518, + "learning_rate": 1.459122504161856e-06, + "loss": 0.653, + "num_input_tokens_seen": 216056480, + "step": 6648 + }, + { + "epoch": 0.5996302475537719, + "flos": 28835928677280.0, + "grad_norm": 1.7983259893956218, + "learning_rate": 1.4585600859506001e-06, + "loss": 0.7172, + "num_input_tokens_seen": 216087450, + "step": 6649 + }, + { + "epoch": 0.5997204310772422, + "flos": 23917855581600.0, + "grad_norm": 1.6993695773129018, + "learning_rate": 1.4579977139379826e-06, + "loss": 0.803, + "num_input_tokens_seen": 216118355, + "step": 6650 + }, + { + "epoch": 0.5998106146007125, + "flos": 17140964902560.0, + "grad_norm": 2.3403701826870247, + "learning_rate": 1.4574353881719895e-06, + "loss": 0.8089, + "num_input_tokens_seen": 216142830, + "step": 6651 + }, + { + "epoch": 0.5999007981241827, + "flos": 23698519596000.0, + "grad_norm": 1.9409824576550616, + "learning_rate": 1.4568731087005998e-06, + "loss": 0.7487, + "num_input_tokens_seen": 216169365, + "step": 6652 + }, + { + "epoch": 0.599990981647653, + "flos": 43589739096960.0, + "grad_norm": 1.6960160865380665, + "learning_rate": 1.4563108755717916e-06, + "loss": 0.7718, + "num_input_tokens_seen": 216202810, + "step": 6653 + }, + { + "epoch": 0.6000811651711232, + "flos": 20926780876320.0, + "grad_norm": 3.4538705978848707, + "learning_rate": 1.455748688833538e-06, + "loss": 0.7716, + "num_input_tokens_seen": 216232310, + "step": 6654 + }, + { + "epoch": 0.6001713486945935, + "flos": 38161988321280.0, + "grad_norm": 2.657938970015839, + "learning_rate": 1.4551865485338065e-06, + "loss": 0.6603, + "num_input_tokens_seen": 216263550, + "step": 6655 + }, + { + "epoch": 0.6002615322180638, + "flos": 28799719597440.0, + "grad_norm": 1.8876982805807951, + "learning_rate": 1.4546244547205629e-06, + "loss": 0.8077, + "num_input_tokens_seen": 216293505, + "step": 6656 + }, + { + "epoch": 0.600351715741534, + "flos": 19363360074720.0, + "grad_norm": 1.98984644350095, + "learning_rate": 1.4540624074417678e-06, + "loss": 0.7716, + "num_input_tokens_seen": 216320925, + "step": 6657 + }, + { + "epoch": 0.6004418992650042, + "flos": 40637512524480.0, + "grad_norm": 9.545268971014066, + "learning_rate": 1.453500406745379e-06, + "loss": 0.8035, + "num_input_tokens_seen": 216355210, + "step": 6658 + }, + { + "epoch": 0.6005320827884746, + "flos": 18742936478880.0, + "grad_norm": 2.0171471104910887, + "learning_rate": 1.4529384526793486e-06, + "loss": 0.8228, + "num_input_tokens_seen": 216380975, + "step": 6659 + }, + { + "epoch": 0.6006222663119448, + "flos": 19833371887200.0, + "grad_norm": 1.7653141929711536, + "learning_rate": 1.4523765452916252e-06, + "loss": 0.7984, + "num_input_tokens_seen": 216407085, + "step": 6660 + }, + { + "epoch": 0.600712449835415, + "flos": 44173173048000.0, + "grad_norm": 3.2269260303305076, + "learning_rate": 1.4518146846301554e-06, + "loss": 0.7731, + "num_input_tokens_seen": 216439550, + "step": 6661 + }, + { + "epoch": 0.6008026333588853, + "flos": 18741077990880.0, + "grad_norm": 1.643510587829336, + "learning_rate": 1.4512528707428787e-06, + "loss": 0.7209, + "num_input_tokens_seen": 216466810, + "step": 6662 + }, + { + "epoch": 0.6008928168823556, + "flos": 28435398613440.0, + "grad_norm": 4.762768298824343, + "learning_rate": 1.4506911036777335e-06, + "loss": 0.7268, + "num_input_tokens_seen": 216497430, + "step": 6663 + }, + { + "epoch": 0.6009830004058259, + "flos": 53270784381600.0, + "grad_norm": 0.6735866611561642, + "learning_rate": 1.450129383482651e-06, + "loss": 0.5479, + "num_input_tokens_seen": 216591450, + "step": 6664 + }, + { + "epoch": 0.6010731839292961, + "flos": 23698816954080.0, + "grad_norm": 1.6333620889558769, + "learning_rate": 1.4495677102055629e-06, + "loss": 0.7006, + "num_input_tokens_seen": 216623950, + "step": 6665 + }, + { + "epoch": 0.6011633674527663, + "flos": 24096150418560.0, + "grad_norm": 1.9077634587963541, + "learning_rate": 1.4490060838943924e-06, + "loss": 0.7814, + "num_input_tokens_seen": 216653050, + "step": 6666 + }, + { + "epoch": 0.6012535509762367, + "flos": 22495628462880.0, + "grad_norm": 1.9434032339389704, + "learning_rate": 1.4484445045970609e-06, + "loss": 0.7483, + "num_input_tokens_seen": 216683815, + "step": 6667 + }, + { + "epoch": 0.6013437344997069, + "flos": 23224753637760.0, + "grad_norm": 2.6124716104671504, + "learning_rate": 1.447882972361485e-06, + "loss": 0.7426, + "num_input_tokens_seen": 216713475, + "step": 6668 + }, + { + "epoch": 0.6014339180231771, + "flos": 26977519961280.0, + "grad_norm": 5.086258885414986, + "learning_rate": 1.4473214872355785e-06, + "loss": 0.7327, + "num_input_tokens_seen": 216743830, + "step": 6669 + }, + { + "epoch": 0.6015241015466475, + "flos": 20741721143040.0, + "grad_norm": 2.1143057379455605, + "learning_rate": 1.4467600492672508e-06, + "loss": 0.7803, + "num_input_tokens_seen": 216765350, + "step": 6670 + }, + { + "epoch": 0.6016142850701177, + "flos": 18707545133760.0, + "grad_norm": 2.225216428538831, + "learning_rate": 1.4461986585044054e-06, + "loss": 0.726, + "num_input_tokens_seen": 216793285, + "step": 6671 + }, + { + "epoch": 0.601704468593588, + "flos": 23661306932640.0, + "grad_norm": 1.7319689720977822, + "learning_rate": 1.4456373149949446e-06, + "loss": 0.7208, + "num_input_tokens_seen": 216822340, + "step": 6672 + }, + { + "epoch": 0.6017946521170582, + "flos": 23334272951520.0, + "grad_norm": 2.0995809863633124, + "learning_rate": 1.4450760187867648e-06, + "loss": 0.7527, + "num_input_tokens_seen": 216854535, + "step": 6673 + }, + { + "epoch": 0.6018848356405285, + "flos": 24501066514080.0, + "grad_norm": 2.194119492412586, + "learning_rate": 1.4445147699277581e-06, + "loss": 0.739, + "num_input_tokens_seen": 216885420, + "step": 6674 + }, + { + "epoch": 0.6019750191639988, + "flos": 23043336540960.0, + "grad_norm": 1.9438953676479502, + "learning_rate": 1.4439535684658154e-06, + "loss": 0.7968, + "num_input_tokens_seen": 216915065, + "step": 6675 + }, + { + "epoch": 0.602065202687469, + "flos": 53463055048320.0, + "grad_norm": 0.7667477608030177, + "learning_rate": 1.44339241444882e-06, + "loss": 0.6173, + "num_input_tokens_seen": 216988430, + "step": 6676 + }, + { + "epoch": 0.6021553862109392, + "flos": 22750913340000.0, + "grad_norm": 2.54612280848666, + "learning_rate": 1.4428313079246518e-06, + "loss": 0.6976, + "num_input_tokens_seen": 217014720, + "step": 6677 + }, + { + "epoch": 0.6022455697344096, + "flos": 64027248491520.0, + "grad_norm": 0.6669247818003102, + "learning_rate": 1.4422702489411896e-06, + "loss": 0.5949, + "num_input_tokens_seen": 217100535, + "step": 6678 + }, + { + "epoch": 0.6023357532578798, + "flos": 20711682243360.0, + "grad_norm": 2.687212019284015, + "learning_rate": 1.4417092375463043e-06, + "loss": 0.6938, + "num_input_tokens_seen": 217127790, + "step": 6679 + }, + { + "epoch": 0.60242593678135, + "flos": 21982568334720.0, + "grad_norm": 2.001544237016298, + "learning_rate": 1.441148273787866e-06, + "loss": 0.7819, + "num_input_tokens_seen": 217156185, + "step": 6680 + }, + { + "epoch": 0.6025161203048203, + "flos": 67632502272960.0, + "grad_norm": 0.7833841927906458, + "learning_rate": 1.4405873577137383e-06, + "loss": 0.6013, + "num_input_tokens_seen": 217245550, + "step": 6681 + }, + { + "epoch": 0.6026063038282906, + "flos": 29673606752160.0, + "grad_norm": 2.359236757384171, + "learning_rate": 1.4400264893717816e-06, + "loss": 0.7565, + "num_input_tokens_seen": 217274505, + "step": 6682 + }, + { + "epoch": 0.6026964873517608, + "flos": 23844842705760.0, + "grad_norm": 2.1714153214996905, + "learning_rate": 1.4394656688098526e-06, + "loss": 0.7969, + "num_input_tokens_seen": 217303680, + "step": 6683 + }, + { + "epoch": 0.6027866708752311, + "flos": 19801028462400.0, + "grad_norm": 2.184482782458005, + "learning_rate": 1.4389048960758032e-06, + "loss": 0.7704, + "num_input_tokens_seen": 217332415, + "step": 6684 + }, + { + "epoch": 0.6028768543987013, + "flos": 24973791719040.0, + "grad_norm": 9.351834171335806, + "learning_rate": 1.4383441712174826e-06, + "loss": 0.7759, + "num_input_tokens_seen": 217359785, + "step": 6685 + }, + { + "epoch": 0.6029670379221717, + "flos": 19830472645920.0, + "grad_norm": 2.2357509617863665, + "learning_rate": 1.4377834942827333e-06, + "loss": 0.8135, + "num_input_tokens_seen": 217385910, + "step": 6686 + }, + { + "epoch": 0.6030572214456419, + "flos": 19618607782080.0, + "grad_norm": 2.076695689227658, + "learning_rate": 1.437222865319397e-06, + "loss": 0.6873, + "num_input_tokens_seen": 217413340, + "step": 6687 + }, + { + "epoch": 0.6031474049691121, + "flos": 21549806355360.0, + "grad_norm": 1.7297804176025748, + "learning_rate": 1.4366622843753092e-06, + "loss": 0.6725, + "num_input_tokens_seen": 217441965, + "step": 6688 + }, + { + "epoch": 0.6032375884925824, + "flos": 33680096822880.0, + "grad_norm": 2.1249184312311815, + "learning_rate": 1.4361017514983006e-06, + "loss": 0.7639, + "num_input_tokens_seen": 217474830, + "step": 6689 + }, + { + "epoch": 0.6033277720160527, + "flos": 19249826426880.0, + "grad_norm": 1.6493798766763341, + "learning_rate": 1.4355412667362006e-06, + "loss": 0.8038, + "num_input_tokens_seen": 217504600, + "step": 6690 + }, + { + "epoch": 0.6034179555395229, + "flos": 19764150326880.0, + "grad_norm": 1.778608471620906, + "learning_rate": 1.4349808301368311e-06, + "loss": 0.7046, + "num_input_tokens_seen": 217533130, + "step": 6691 + }, + { + "epoch": 0.6035081390629932, + "flos": 20710976017920.0, + "grad_norm": 2.2600738921787293, + "learning_rate": 1.4344204417480139e-06, + "loss": 0.7906, + "num_input_tokens_seen": 217558795, + "step": 6692 + }, + { + "epoch": 0.6035983225864634, + "flos": 22970026307040.0, + "grad_norm": 1.9577006339273295, + "learning_rate": 1.4338601016175628e-06, + "loss": 0.7003, + "num_input_tokens_seen": 217587960, + "step": 6693 + }, + { + "epoch": 0.6036885061099337, + "flos": 35757321147840.0, + "grad_norm": 1.8349827658028446, + "learning_rate": 1.433299809793289e-06, + "loss": 0.7465, + "num_input_tokens_seen": 217618820, + "step": 6694 + }, + { + "epoch": 0.603778689633404, + "flos": 22491093752160.0, + "grad_norm": 2.9819305155057183, + "learning_rate": 1.432739566323001e-06, + "loss": 0.62, + "num_input_tokens_seen": 217646600, + "step": 6695 + }, + { + "epoch": 0.6038688731568742, + "flos": 24756165542400.0, + "grad_norm": 2.26073596533296, + "learning_rate": 1.4321793712545004e-06, + "loss": 0.7158, + "num_input_tokens_seen": 217675795, + "step": 6696 + }, + { + "epoch": 0.6039590566803446, + "flos": 22452468637920.0, + "grad_norm": 1.9537451408067097, + "learning_rate": 1.4316192246355873e-06, + "loss": 0.7961, + "num_input_tokens_seen": 217702135, + "step": 6697 + }, + { + "epoch": 0.6040492402038148, + "flos": 64495253136960.0, + "grad_norm": 0.6029826406365123, + "learning_rate": 1.4310591265140555e-06, + "loss": 0.5784, + "num_input_tokens_seen": 217799715, + "step": 6698 + }, + { + "epoch": 0.604139423727285, + "flos": 24427050054720.0, + "grad_norm": 1.567717576448518, + "learning_rate": 1.4304990769376963e-06, + "loss": 0.7539, + "num_input_tokens_seen": 217831895, + "step": 6699 + }, + { + "epoch": 0.6042296072507553, + "flos": 30033950571840.0, + "grad_norm": 2.2990142824675948, + "learning_rate": 1.4299390759542962e-06, + "loss": 0.7312, + "num_input_tokens_seen": 217862145, + "step": 6700 + }, + { + "epoch": 0.6043197907742256, + "flos": 23807072496000.0, + "grad_norm": 1.8820247664286476, + "learning_rate": 1.4293791236116368e-06, + "loss": 0.7295, + "num_input_tokens_seen": 217889520, + "step": 6701 + }, + { + "epoch": 0.6044099742976958, + "flos": 23152186799040.0, + "grad_norm": 2.3514025053934526, + "learning_rate": 1.4288192199574978e-06, + "loss": 0.7094, + "num_input_tokens_seen": 217917645, + "step": 6702 + }, + { + "epoch": 0.6045001578211661, + "flos": 67416660244800.0, + "grad_norm": 0.640589247845677, + "learning_rate": 1.4282593650396524e-06, + "loss": 0.551, + "num_input_tokens_seen": 218014925, + "step": 6703 + }, + { + "epoch": 0.6045903413446363, + "flos": 23480224363680.0, + "grad_norm": 2.100804756007509, + "learning_rate": 1.4276995589058695e-06, + "loss": 0.8453, + "num_input_tokens_seen": 218043850, + "step": 6704 + }, + { + "epoch": 0.6046805248681066, + "flos": 66439684144800.0, + "grad_norm": 0.5956392848092142, + "learning_rate": 1.4271398016039168e-06, + "loss": 0.5478, + "num_input_tokens_seen": 218140795, + "step": 6705 + }, + { + "epoch": 0.6047707083915769, + "flos": 19581580967520.0, + "grad_norm": 2.214889498983256, + "learning_rate": 1.4265800931815542e-06, + "loss": 0.7893, + "num_input_tokens_seen": 218168385, + "step": 6706 + }, + { + "epoch": 0.6048608919150471, + "flos": 20711682243360.0, + "grad_norm": 1.6606991587007047, + "learning_rate": 1.4260204336865406e-06, + "loss": 0.7546, + "num_input_tokens_seen": 218195640, + "step": 6707 + }, + { + "epoch": 0.6049510754385173, + "flos": 23186463051360.0, + "grad_norm": 1.7497999894421528, + "learning_rate": 1.4254608231666281e-06, + "loss": 0.7439, + "num_input_tokens_seen": 218225395, + "step": 6708 + }, + { + "epoch": 0.6050412589619877, + "flos": 22310197032000.0, + "grad_norm": 1.7227128553713666, + "learning_rate": 1.4249012616695661e-06, + "loss": 0.7557, + "num_input_tokens_seen": 218252700, + "step": 6709 + }, + { + "epoch": 0.6051314424854579, + "flos": 47486338156320.0, + "grad_norm": 1.8202043754980244, + "learning_rate": 1.4243417492431e-06, + "loss": 0.7534, + "num_input_tokens_seen": 218285420, + "step": 6710 + }, + { + "epoch": 0.6052216260089281, + "flos": 16703816891520.0, + "grad_norm": 2.3011036049580063, + "learning_rate": 1.4237822859349696e-06, + "loss": 0.7248, + "num_input_tokens_seen": 218312445, + "step": 6711 + }, + { + "epoch": 0.6053118095323984, + "flos": 24568987132800.0, + "grad_norm": 1.4737302066680653, + "learning_rate": 1.423222871792912e-06, + "loss": 0.7175, + "num_input_tokens_seen": 218341610, + "step": 6712 + }, + { + "epoch": 0.6054019930558687, + "flos": 19614593448000.0, + "grad_norm": 1.6496931599922586, + "learning_rate": 1.4226635068646586e-06, + "loss": 0.8035, + "num_input_tokens_seen": 218369215, + "step": 6713 + }, + { + "epoch": 0.605492176579339, + "flos": 61724071963680.0, + "grad_norm": 0.6208487693309913, + "learning_rate": 1.4221041911979393e-06, + "loss": 0.5718, + "num_input_tokens_seen": 218460980, + "step": 6714 + }, + { + "epoch": 0.6055823601028092, + "flos": 18812789925120.0, + "grad_norm": 2.3999940872119545, + "learning_rate": 1.4215449248404765e-06, + "loss": 0.7628, + "num_input_tokens_seen": 218487470, + "step": 6715 + }, + { + "epoch": 0.6056725436262794, + "flos": 27153956310240.0, + "grad_norm": 1.6667147708944383, + "learning_rate": 1.4209857078399896e-06, + "loss": 0.7662, + "num_input_tokens_seen": 218520385, + "step": 6716 + }, + { + "epoch": 0.6057627271497498, + "flos": 13606456641600.0, + "grad_norm": 2.2198502942032636, + "learning_rate": 1.4204265402441955e-06, + "loss": 0.717, + "num_input_tokens_seen": 218546935, + "step": 6717 + }, + { + "epoch": 0.60585291067322, + "flos": 19800619595040.0, + "grad_norm": 1.6566790618790965, + "learning_rate": 1.419867422100804e-06, + "loss": 0.6251, + "num_input_tokens_seen": 218575980, + "step": 6718 + }, + { + "epoch": 0.6059430941966902, + "flos": 23552716862880.0, + "grad_norm": 2.1023187233956677, + "learning_rate": 1.4193083534575236e-06, + "loss": 0.732, + "num_input_tokens_seen": 218605525, + "step": 6719 + }, + { + "epoch": 0.6060332777201606, + "flos": 22277890776960.0, + "grad_norm": 2.4647893884531293, + "learning_rate": 1.4187493343620567e-06, + "loss": 0.7731, + "num_input_tokens_seen": 218633270, + "step": 6720 + }, + { + "epoch": 0.6061234612436308, + "flos": 21184630466880.0, + "grad_norm": 1.9701600247949838, + "learning_rate": 1.4181903648621006e-06, + "loss": 0.7797, + "num_input_tokens_seen": 218661065, + "step": 6721 + }, + { + "epoch": 0.606213644767101, + "flos": 23697999219360.0, + "grad_norm": 1.7695728827625177, + "learning_rate": 1.4176314450053512e-06, + "loss": 0.7789, + "num_input_tokens_seen": 218692070, + "step": 6722 + }, + { + "epoch": 0.6063038282905713, + "flos": 19181459771040.0, + "grad_norm": 2.1748547867279027, + "learning_rate": 1.4170725748394977e-06, + "loss": 0.7359, + "num_input_tokens_seen": 218720305, + "step": 6723 + }, + { + "epoch": 0.6063940118140416, + "flos": 23917632563040.0, + "grad_norm": 1.7159350839353666, + "learning_rate": 1.4165137544122266e-06, + "loss": 0.6582, + "num_input_tokens_seen": 218748800, + "step": 6724 + }, + { + "epoch": 0.6064841953375119, + "flos": 24675941733120.0, + "grad_norm": 2.002040180081287, + "learning_rate": 1.4159549837712194e-06, + "loss": 0.8168, + "num_input_tokens_seen": 218777870, + "step": 6725 + }, + { + "epoch": 0.6065743788609821, + "flos": 34880794940160.0, + "grad_norm": 1.5803946612986979, + "learning_rate": 1.415396262964153e-06, + "loss": 0.6657, + "num_input_tokens_seen": 218809165, + "step": 6726 + }, + { + "epoch": 0.6066645623844523, + "flos": 22491651298560.0, + "grad_norm": 1.9972547525236413, + "learning_rate": 1.4148375920387016e-06, + "loss": 0.8231, + "num_input_tokens_seen": 218837250, + "step": 6727 + }, + { + "epoch": 0.6067547459079227, + "flos": 67341268504320.0, + "grad_norm": 0.6408537014509322, + "learning_rate": 1.4142789710425325e-06, + "loss": 0.5751, + "num_input_tokens_seen": 218935630, + "step": 6728 + }, + { + "epoch": 0.6068449294313929, + "flos": 22022717409120.0, + "grad_norm": 2.110054754297052, + "learning_rate": 1.4137204000233118e-06, + "loss": 0.6849, + "num_input_tokens_seen": 218961615, + "step": 6729 + }, + { + "epoch": 0.6069351129548631, + "flos": 37069397066880.0, + "grad_norm": 2.095364076575553, + "learning_rate": 1.4131618790286987e-06, + "loss": 0.7432, + "num_input_tokens_seen": 218992570, + "step": 6730 + }, + { + "epoch": 0.6070252964783334, + "flos": 23990422420320.0, + "grad_norm": 2.761323676656836, + "learning_rate": 1.4126034081063506e-06, + "loss": 0.7331, + "num_input_tokens_seen": 219020830, + "step": 6731 + }, + { + "epoch": 0.6071154800018037, + "flos": 28143644468160.0, + "grad_norm": 1.9975601453642227, + "learning_rate": 1.4120449873039186e-06, + "loss": 0.7348, + "num_input_tokens_seen": 219049860, + "step": 6732 + }, + { + "epoch": 0.6072056635252739, + "flos": 26140325093280.0, + "grad_norm": 2.519059673800372, + "learning_rate": 1.4114866166690494e-06, + "loss": 0.7863, + "num_input_tokens_seen": 219077655, + "step": 6733 + }, + { + "epoch": 0.6072958470487442, + "flos": 46432520695200.0, + "grad_norm": 3.561700162276363, + "learning_rate": 1.4109282962493877e-06, + "loss": 0.7066, + "num_input_tokens_seen": 219113560, + "step": 6734 + }, + { + "epoch": 0.6073860305722144, + "flos": 21512816710560.0, + "grad_norm": 2.737282999574045, + "learning_rate": 1.4103700260925716e-06, + "loss": 0.7368, + "num_input_tokens_seen": 219142165, + "step": 6735 + }, + { + "epoch": 0.6074762140956848, + "flos": 27888582609600.0, + "grad_norm": 1.890570143226953, + "learning_rate": 1.4098118062462364e-06, + "loss": 0.7863, + "num_input_tokens_seen": 219171830, + "step": 6736 + }, + { + "epoch": 0.607566397619155, + "flos": 22528194906240.0, + "grad_norm": 1.7170855005045877, + "learning_rate": 1.4092536367580123e-06, + "loss": 0.7412, + "num_input_tokens_seen": 219199560, + "step": 6737 + }, + { + "epoch": 0.6076565811426252, + "flos": 24427161564000.0, + "grad_norm": 1.872595198570679, + "learning_rate": 1.4086955176755248e-06, + "loss": 0.66, + "num_input_tokens_seen": 219225500, + "step": 6738 + }, + { + "epoch": 0.6077467646660955, + "flos": 19764373345440.0, + "grad_norm": 2.3754753841665566, + "learning_rate": 1.4081374490463964e-06, + "loss": 0.6876, + "num_input_tokens_seen": 219254455, + "step": 6739 + }, + { + "epoch": 0.6078369481895658, + "flos": 15719852876640.0, + "grad_norm": 2.109790563268239, + "learning_rate": 1.4075794309182443e-06, + "loss": 0.7602, + "num_input_tokens_seen": 219282080, + "step": 6740 + }, + { + "epoch": 0.607927131713036, + "flos": 21074999643840.0, + "grad_norm": 2.6963605971416573, + "learning_rate": 1.407021463338682e-06, + "loss": 0.759, + "num_input_tokens_seen": 219309055, + "step": 6741 + }, + { + "epoch": 0.6080173152365063, + "flos": 70961836226880.0, + "grad_norm": 0.5938381581361608, + "learning_rate": 1.4064635463553177e-06, + "loss": 0.5204, + "num_input_tokens_seen": 219402485, + "step": 6742 + }, + { + "epoch": 0.6081074987599766, + "flos": 23297283306720.0, + "grad_norm": 2.942682280748659, + "learning_rate": 1.4059056800157563e-06, + "loss": 0.7766, + "num_input_tokens_seen": 219427825, + "step": 6743 + }, + { + "epoch": 0.6081976822834468, + "flos": 22605370795200.0, + "grad_norm": 1.781158913408302, + "learning_rate": 1.4053478643675982e-06, + "loss": 0.6766, + "num_input_tokens_seen": 219455130, + "step": 6744 + }, + { + "epoch": 0.6082878658069171, + "flos": 21148681575360.0, + "grad_norm": 1.7092570234519584, + "learning_rate": 1.4047900994584389e-06, + "loss": 0.6426, + "num_input_tokens_seen": 219482580, + "step": 6745 + }, + { + "epoch": 0.6083780493303873, + "flos": 26649408057120.0, + "grad_norm": 2.385124010904632, + "learning_rate": 1.404232385335871e-06, + "loss": 0.6563, + "num_input_tokens_seen": 219512435, + "step": 6746 + }, + { + "epoch": 0.6084682328538576, + "flos": 26353082031360.0, + "grad_norm": 2.0147097379943197, + "learning_rate": 1.4036747220474806e-06, + "loss": 0.7459, + "num_input_tokens_seen": 219540760, + "step": 6747 + }, + { + "epoch": 0.6085584163773279, + "flos": 24391361351520.0, + "grad_norm": 3.7714436636040656, + "learning_rate": 1.4031171096408506e-06, + "loss": 0.6961, + "num_input_tokens_seen": 219571715, + "step": 6748 + }, + { + "epoch": 0.6086485999007981, + "flos": 37215125460480.0, + "grad_norm": 6.105589022827677, + "learning_rate": 1.4025595481635607e-06, + "loss": 0.7685, + "num_input_tokens_seen": 219601185, + "step": 6749 + }, + { + "epoch": 0.6087387834242683, + "flos": 16336373647680.0, + "grad_norm": 2.0777470524647312, + "learning_rate": 1.4020020376631836e-06, + "loss": 0.7646, + "num_input_tokens_seen": 219626650, + "step": 6750 + }, + { + "epoch": 0.6088289669477387, + "flos": 25410679541760.0, + "grad_norm": 2.6009934178100043, + "learning_rate": 1.4014445781872908e-06, + "loss": 0.758, + "num_input_tokens_seen": 219654795, + "step": 6751 + }, + { + "epoch": 0.6089191504712089, + "flos": 26322299736480.0, + "grad_norm": 1.7499691415518352, + "learning_rate": 1.4008871697834465e-06, + "loss": 0.8067, + "num_input_tokens_seen": 219683745, + "step": 6752 + }, + { + "epoch": 0.6090093339946792, + "flos": 22387893297600.0, + "grad_norm": 1.6701965463070203, + "learning_rate": 1.400329812499213e-06, + "loss": 0.7817, + "num_input_tokens_seen": 219716275, + "step": 6753 + }, + { + "epoch": 0.6090995175181494, + "flos": 24204740488320.0, + "grad_norm": 1.5630776672640219, + "learning_rate": 1.3997725063821458e-06, + "loss": 0.7563, + "num_input_tokens_seen": 219745050, + "step": 6754 + }, + { + "epoch": 0.6091897010416197, + "flos": 28835296791360.0, + "grad_norm": 5.019145930006341, + "learning_rate": 1.3992152514797978e-06, + "loss": 0.7108, + "num_input_tokens_seen": 219775850, + "step": 6755 + }, + { + "epoch": 0.60927988456509, + "flos": 65146236009120.0, + "grad_norm": 0.6206992954066392, + "learning_rate": 1.398658047839718e-06, + "loss": 0.5953, + "num_input_tokens_seen": 219866185, + "step": 6756 + }, + { + "epoch": 0.6093700680885602, + "flos": 72383059762080.0, + "grad_norm": 0.6031258815304338, + "learning_rate": 1.3981008955094481e-06, + "loss": 0.5649, + "num_input_tokens_seen": 219970975, + "step": 6757 + }, + { + "epoch": 0.6094602516120304, + "flos": 40493159412000.0, + "grad_norm": 2.4700863275615874, + "learning_rate": 1.39754379453653e-06, + "loss": 0.7225, + "num_input_tokens_seen": 220004395, + "step": 6758 + }, + { + "epoch": 0.6095504351355008, + "flos": 22125211638240.0, + "grad_norm": 2.267333760923356, + "learning_rate": 1.3969867449684972e-06, + "loss": 0.7539, + "num_input_tokens_seen": 220032035, + "step": 6759 + }, + { + "epoch": 0.609640618658971, + "flos": 29852830833120.0, + "grad_norm": 3.2489983408915757, + "learning_rate": 1.396429746852879e-06, + "loss": 0.738, + "num_input_tokens_seen": 220060380, + "step": 6760 + }, + { + "epoch": 0.6097308021824412, + "flos": 23073189591840.0, + "grad_norm": 1.8083637659819933, + "learning_rate": 1.395872800237204e-06, + "loss": 0.7244, + "num_input_tokens_seen": 220087330, + "step": 6761 + }, + { + "epoch": 0.6098209857059115, + "flos": 60166040777280.0, + "grad_norm": 0.7562040188895203, + "learning_rate": 1.3953159051689918e-06, + "loss": 0.5972, + "num_input_tokens_seen": 220171350, + "step": 6762 + }, + { + "epoch": 0.6099111692293818, + "flos": 26319846532320.0, + "grad_norm": 1.8137553898884395, + "learning_rate": 1.3947590616957618e-06, + "loss": 0.7904, + "num_input_tokens_seen": 220200330, + "step": 6763 + }, + { + "epoch": 0.610001352752852, + "flos": 21585792416640.0, + "grad_norm": 2.0969405276297217, + "learning_rate": 1.3942022698650258e-06, + "loss": 0.7741, + "num_input_tokens_seen": 220228015, + "step": 6764 + }, + { + "epoch": 0.6100915362763223, + "flos": 26824543464480.0, + "grad_norm": 1.7275472741069322, + "learning_rate": 1.3936455297242917e-06, + "loss": 0.7797, + "num_input_tokens_seen": 220255785, + "step": 6765 + }, + { + "epoch": 0.6101817197997926, + "flos": 23953990321920.0, + "grad_norm": 1.7546351220320724, + "learning_rate": 1.3930888413210652e-06, + "loss": 0.7644, + "num_input_tokens_seen": 220284775, + "step": 6766 + }, + { + "epoch": 0.6102719033232629, + "flos": 22386964053600.0, + "grad_norm": 2.4961118981025514, + "learning_rate": 1.392532204702845e-06, + "loss": 0.7092, + "num_input_tokens_seen": 220313055, + "step": 6767 + }, + { + "epoch": 0.6103620868467331, + "flos": 16011867210240.0, + "grad_norm": 2.3273620379377045, + "learning_rate": 1.3919756199171266e-06, + "loss": 0.675, + "num_input_tokens_seen": 220339510, + "step": 6768 + }, + { + "epoch": 0.6104522703702033, + "flos": 22823926215840.0, + "grad_norm": 1.9335397423802205, + "learning_rate": 1.3914190870114009e-06, + "loss": 0.8195, + "num_input_tokens_seen": 220368105, + "step": 6769 + }, + { + "epoch": 0.6105424538936737, + "flos": 22715373315840.0, + "grad_norm": 1.9851881148518085, + "learning_rate": 1.3908626060331541e-06, + "loss": 0.7973, + "num_input_tokens_seen": 220397040, + "step": 6770 + }, + { + "epoch": 0.6106326374171439, + "flos": 15792940092000.0, + "grad_norm": 1.8897143606273814, + "learning_rate": 1.3903061770298693e-06, + "loss": 0.7784, + "num_input_tokens_seen": 220424615, + "step": 6771 + }, + { + "epoch": 0.6107228209406141, + "flos": 26825732896800.0, + "grad_norm": 2.082481578914871, + "learning_rate": 1.3897498000490223e-06, + "loss": 0.7206, + "num_input_tokens_seen": 220450560, + "step": 6772 + }, + { + "epoch": 0.6108130044640844, + "flos": 21805351420800.0, + "grad_norm": 2.0054703050049363, + "learning_rate": 1.3891934751380879e-06, + "loss": 0.8172, + "num_input_tokens_seen": 220479610, + "step": 6773 + }, + { + "epoch": 0.6109031879875547, + "flos": 20777558525280.0, + "grad_norm": 1.882612922189509, + "learning_rate": 1.3886372023445334e-06, + "loss": 0.766, + "num_input_tokens_seen": 220505630, + "step": 6774 + }, + { + "epoch": 0.610993371511025, + "flos": 20384127885600.0, + "grad_norm": 1.7303019630534313, + "learning_rate": 1.3880809817158246e-06, + "loss": 0.8164, + "num_input_tokens_seen": 220532110, + "step": 6775 + }, + { + "epoch": 0.6110835550344952, + "flos": 21403594754880.0, + "grad_norm": 2.3928513476010806, + "learning_rate": 1.3875248132994206e-06, + "loss": 0.7388, + "num_input_tokens_seen": 220558430, + "step": 6776 + }, + { + "epoch": 0.6111737385579654, + "flos": 18853310697120.0, + "grad_norm": 1.8664734564425867, + "learning_rate": 1.386968697142776e-06, + "loss": 0.6707, + "num_input_tokens_seen": 220587075, + "step": 6777 + }, + { + "epoch": 0.6112639220814358, + "flos": 22168929009600.0, + "grad_norm": 2.630340139781554, + "learning_rate": 1.386412633293343e-06, + "loss": 0.7923, + "num_input_tokens_seen": 220616810, + "step": 6778 + }, + { + "epoch": 0.611354105604906, + "flos": 16265219259840.0, + "grad_norm": 2.9152126139186807, + "learning_rate": 1.3858566217985672e-06, + "loss": 0.7201, + "num_input_tokens_seen": 220643250, + "step": 6779 + }, + { + "epoch": 0.6114442891283762, + "flos": 22673774620800.0, + "grad_norm": 2.1306905286206304, + "learning_rate": 1.3853006627058905e-06, + "loss": 0.8303, + "num_input_tokens_seen": 220670590, + "step": 6780 + }, + { + "epoch": 0.6115344726518465, + "flos": 25592877203520.0, + "grad_norm": 2.383841470644709, + "learning_rate": 1.3847447560627512e-06, + "loss": 0.63, + "num_input_tokens_seen": 220696200, + "step": 6781 + }, + { + "epoch": 0.6116246561753168, + "flos": 26060733169920.0, + "grad_norm": 1.9123346743464553, + "learning_rate": 1.3841889019165812e-06, + "loss": 0.7871, + "num_input_tokens_seen": 220725460, + "step": 6782 + }, + { + "epoch": 0.611714839698787, + "flos": 30474815558880.0, + "grad_norm": 2.8429465364037534, + "learning_rate": 1.3836331003148106e-06, + "loss": 0.7029, + "num_input_tokens_seen": 220755220, + "step": 6783 + }, + { + "epoch": 0.6118050232222573, + "flos": 28253200951680.0, + "grad_norm": 1.6643019669567738, + "learning_rate": 1.3830773513048612e-06, + "loss": 0.7836, + "num_input_tokens_seen": 220785245, + "step": 6784 + }, + { + "epoch": 0.6118952067457275, + "flos": 21366865298400.0, + "grad_norm": 1.751370574481642, + "learning_rate": 1.382521654934155e-06, + "loss": 0.763, + "num_input_tokens_seen": 220813030, + "step": 6785 + }, + { + "epoch": 0.6119853902691978, + "flos": 22350717804000.0, + "grad_norm": 1.9512340088994866, + "learning_rate": 1.3819660112501057e-06, + "loss": 0.6949, + "num_input_tokens_seen": 220840650, + "step": 6786 + }, + { + "epoch": 0.6120755737926681, + "flos": 24900667333920.0, + "grad_norm": 1.725811272736038, + "learning_rate": 1.3814104203001234e-06, + "loss": 0.7115, + "num_input_tokens_seen": 220868635, + "step": 6787 + }, + { + "epoch": 0.6121657573161383, + "flos": 21003510728160.0, + "grad_norm": 1.592490427775936, + "learning_rate": 1.3808548821316156e-06, + "loss": 0.7881, + "num_input_tokens_seen": 220898490, + "step": 6788 + }, + { + "epoch": 0.6122559408396087, + "flos": 25735260318720.0, + "grad_norm": 1.679915088937363, + "learning_rate": 1.3802993967919824e-06, + "loss": 0.6798, + "num_input_tokens_seen": 220928425, + "step": 6789 + }, + { + "epoch": 0.6123461243630789, + "flos": 25010223817440.0, + "grad_norm": 2.2235684793888715, + "learning_rate": 1.3797439643286227e-06, + "loss": 0.7072, + "num_input_tokens_seen": 220956770, + "step": 6790 + }, + { + "epoch": 0.6124363078865491, + "flos": 26030359742400.0, + "grad_norm": 1.5582382829342813, + "learning_rate": 1.3791885847889277e-06, + "loss": 0.7344, + "num_input_tokens_seen": 220986840, + "step": 6791 + }, + { + "epoch": 0.6125264914100194, + "flos": 17578335932160.0, + "grad_norm": 1.6512887170195818, + "learning_rate": 1.3786332582202853e-06, + "loss": 0.7848, + "num_input_tokens_seen": 221012035, + "step": 6792 + }, + { + "epoch": 0.6126166749334897, + "flos": 25666373286240.0, + "grad_norm": 1.838242273205243, + "learning_rate": 1.3780779846700799e-06, + "loss": 0.747, + "num_input_tokens_seen": 221041210, + "step": 6793 + }, + { + "epoch": 0.6127068584569599, + "flos": 29964543162720.0, + "grad_norm": 1.6106780619349754, + "learning_rate": 1.3775227641856899e-06, + "loss": 0.7572, + "num_input_tokens_seen": 221073255, + "step": 6794 + }, + { + "epoch": 0.6127970419804302, + "flos": 26649593905920.0, + "grad_norm": 1.8511391972694102, + "learning_rate": 1.37696759681449e-06, + "loss": 0.7458, + "num_input_tokens_seen": 221104765, + "step": 6795 + }, + { + "epoch": 0.6128872255039004, + "flos": 64968796076640.0, + "grad_norm": 0.6921737204529915, + "learning_rate": 1.37641248260385e-06, + "loss": 0.6061, + "num_input_tokens_seen": 221195910, + "step": 6796 + }, + { + "epoch": 0.6129774090273707, + "flos": 23151926610720.0, + "grad_norm": 1.7866152220882607, + "learning_rate": 1.375857421601136e-06, + "loss": 0.7972, + "num_input_tokens_seen": 221224585, + "step": 6797 + }, + { + "epoch": 0.613067592550841, + "flos": 27597348840960.0, + "grad_norm": 2.3275446840579925, + "learning_rate": 1.3753024138537082e-06, + "loss": 0.758, + "num_input_tokens_seen": 221254530, + "step": 6798 + }, + { + "epoch": 0.6131577760743112, + "flos": 21476607630720.0, + "grad_norm": 1.5309986764614332, + "learning_rate": 1.3747474594089221e-06, + "loss": 0.8134, + "num_input_tokens_seen": 221281130, + "step": 6799 + }, + { + "epoch": 0.6132479595977814, + "flos": 25411199918400.0, + "grad_norm": 2.5517845342195393, + "learning_rate": 1.374192558314131e-06, + "loss": 0.7283, + "num_input_tokens_seen": 221309240, + "step": 6800 + }, + { + "epoch": 0.6133381431212518, + "flos": 22707790684800.0, + "grad_norm": 2.0274128984775364, + "learning_rate": 1.373637710616681e-06, + "loss": 0.8092, + "num_input_tokens_seen": 221333510, + "step": 6801 + }, + { + "epoch": 0.613428326644722, + "flos": 15099652299360.0, + "grad_norm": 2.1899791293260185, + "learning_rate": 1.373082916363916e-06, + "loss": 0.8569, + "num_input_tokens_seen": 221356760, + "step": 6802 + }, + { + "epoch": 0.6135185101681923, + "flos": 27378458892480.0, + "grad_norm": 1.5694948411500065, + "learning_rate": 1.3725281756031732e-06, + "loss": 0.7905, + "num_input_tokens_seen": 221388135, + "step": 6803 + }, + { + "epoch": 0.6136086936916625, + "flos": 20856555732480.0, + "grad_norm": 1.707377453228122, + "learning_rate": 1.3719734883817858e-06, + "loss": 0.749, + "num_input_tokens_seen": 221416840, + "step": 6804 + }, + { + "epoch": 0.6136988772151328, + "flos": 24354483216000.0, + "grad_norm": 1.6765968454332423, + "learning_rate": 1.371418854747084e-06, + "loss": 0.8641, + "num_input_tokens_seen": 221445045, + "step": 6805 + }, + { + "epoch": 0.6137890607386031, + "flos": 27232656159360.0, + "grad_norm": 2.361787080927747, + "learning_rate": 1.3708642747463905e-06, + "loss": 0.784, + "num_input_tokens_seen": 221473385, + "step": 6806 + }, + { + "epoch": 0.6138792442620733, + "flos": 20454873406080.0, + "grad_norm": 1.7485442237224749, + "learning_rate": 1.370309748427027e-06, + "loss": 0.7785, + "num_input_tokens_seen": 221503280, + "step": 6807 + }, + { + "epoch": 0.6139694277855435, + "flos": 22496446197600.0, + "grad_norm": 1.9392948450999088, + "learning_rate": 1.3697552758363079e-06, + "loss": 0.8238, + "num_input_tokens_seen": 221531120, + "step": 6808 + }, + { + "epoch": 0.6140596113090139, + "flos": 21913644132480.0, + "grad_norm": 3.2997737633576127, + "learning_rate": 1.3692008570215432e-06, + "loss": 0.7536, + "num_input_tokens_seen": 221560715, + "step": 6809 + }, + { + "epoch": 0.6141497948324841, + "flos": 23618593144800.0, + "grad_norm": 1.7004050137002846, + "learning_rate": 1.3686464920300398e-06, + "loss": 0.7043, + "num_input_tokens_seen": 221589830, + "step": 6810 + }, + { + "epoch": 0.6142399783559543, + "flos": 28507965452160.0, + "grad_norm": 2.2481641697381085, + "learning_rate": 1.3680921809090985e-06, + "loss": 0.7257, + "num_input_tokens_seen": 221621060, + "step": 6811 + }, + { + "epoch": 0.6143301618794246, + "flos": 30875828829600.0, + "grad_norm": 1.969515212771875, + "learning_rate": 1.3675379237060175e-06, + "loss": 0.7864, + "num_input_tokens_seen": 221649290, + "step": 6812 + }, + { + "epoch": 0.6144203454028949, + "flos": 33206219355360.0, + "grad_norm": 1.67542353726926, + "learning_rate": 1.366983720468088e-06, + "loss": 0.7447, + "num_input_tokens_seen": 221680085, + "step": 6813 + }, + { + "epoch": 0.6145105289263652, + "flos": 24718581181440.0, + "grad_norm": 2.6525783519336374, + "learning_rate": 1.3664295712425972e-06, + "loss": 0.7562, + "num_input_tokens_seen": 221711265, + "step": 6814 + }, + { + "epoch": 0.6146007124498354, + "flos": 29997704322240.0, + "grad_norm": 2.0139663140624937, + "learning_rate": 1.3658754760768296e-06, + "loss": 0.6849, + "num_input_tokens_seen": 221739960, + "step": 6815 + }, + { + "epoch": 0.6146908959733057, + "flos": 20092448079840.0, + "grad_norm": 1.9529731386274096, + "learning_rate": 1.3653214350180621e-06, + "loss": 0.7865, + "num_input_tokens_seen": 221765960, + "step": 6816 + }, + { + "epoch": 0.614781079496776, + "flos": 20528778356160.0, + "grad_norm": 1.6468081263477183, + "learning_rate": 1.3647674481135703e-06, + "loss": 0.8173, + "num_input_tokens_seen": 221792505, + "step": 6817 + }, + { + "epoch": 0.6148712630202462, + "flos": 23298658587840.0, + "grad_norm": 1.4016903600984623, + "learning_rate": 1.3642135154106217e-06, + "loss": 0.6486, + "num_input_tokens_seen": 221821440, + "step": 6818 + }, + { + "epoch": 0.6149614465437164, + "flos": 24349688316960.0, + "grad_norm": 2.332293267282877, + "learning_rate": 1.363659636956482e-06, + "loss": 0.704, + "num_input_tokens_seen": 221850450, + "step": 6819 + }, + { + "epoch": 0.6150516300671868, + "flos": 36194692177440.0, + "grad_norm": 1.655981437933823, + "learning_rate": 1.3631058127984112e-06, + "loss": 0.7193, + "num_input_tokens_seen": 221881280, + "step": 6820 + }, + { + "epoch": 0.615141813590657, + "flos": 22275623421600.0, + "grad_norm": 3.0242691670469704, + "learning_rate": 1.3625520429836632e-06, + "loss": 0.7251, + "num_input_tokens_seen": 221909015, + "step": 6821 + }, + { + "epoch": 0.6152319971141272, + "flos": 27087002105280.0, + "grad_norm": 1.720170630374897, + "learning_rate": 1.361998327559491e-06, + "loss": 0.6544, + "num_input_tokens_seen": 221937540, + "step": 6822 + }, + { + "epoch": 0.6153221806375975, + "flos": 64653730758240.0, + "grad_norm": 0.6601174472939478, + "learning_rate": 1.3614446665731385e-06, + "loss": 0.5805, + "num_input_tokens_seen": 222031135, + "step": 6823 + }, + { + "epoch": 0.6154123641610678, + "flos": 22022977597440.0, + "grad_norm": 2.6899957822745346, + "learning_rate": 1.3608910600718484e-06, + "loss": 0.7572, + "num_input_tokens_seen": 222060030, + "step": 6824 + }, + { + "epoch": 0.615502547684538, + "flos": 21073103986080.0, + "grad_norm": 1.8069581118838995, + "learning_rate": 1.360337508102857e-06, + "loss": 0.7518, + "num_input_tokens_seen": 222089435, + "step": 6825 + }, + { + "epoch": 0.6155927312080083, + "flos": 19799913369600.0, + "grad_norm": 2.187913365136075, + "learning_rate": 1.3597840107133962e-06, + "loss": 0.8459, + "num_input_tokens_seen": 222114305, + "step": 6826 + }, + { + "epoch": 0.6156829147314785, + "flos": 28683584066400.0, + "grad_norm": 1.67016242899159, + "learning_rate": 1.3592305679506944e-06, + "loss": 0.7586, + "num_input_tokens_seen": 222144415, + "step": 6827 + }, + { + "epoch": 0.6157730982549489, + "flos": 23879118958080.0, + "grad_norm": 1.6267447275182116, + "learning_rate": 1.3586771798619726e-06, + "loss": 0.6587, + "num_input_tokens_seen": 222174560, + "step": 6828 + }, + { + "epoch": 0.6158632817784191, + "flos": 27087150784320.0, + "grad_norm": 1.6905566189346426, + "learning_rate": 1.358123846494451e-06, + "loss": 0.8134, + "num_input_tokens_seen": 222203535, + "step": 6829 + }, + { + "epoch": 0.6159534653018893, + "flos": 24017450569440.0, + "grad_norm": 2.454239124551835, + "learning_rate": 1.3575705678953422e-06, + "loss": 0.753, + "num_input_tokens_seen": 222225690, + "step": 6830 + }, + { + "epoch": 0.6160436488253596, + "flos": 20310483123840.0, + "grad_norm": 1.8577023844894447, + "learning_rate": 1.357017344111854e-06, + "loss": 0.788, + "num_input_tokens_seen": 222253080, + "step": 6831 + }, + { + "epoch": 0.6161338323488299, + "flos": 19872889075680.0, + "grad_norm": 1.754153703594862, + "learning_rate": 1.356464175191192e-06, + "loss": 0.7612, + "num_input_tokens_seen": 222281635, + "step": 6832 + }, + { + "epoch": 0.6162240158723001, + "flos": 28253015102880.0, + "grad_norm": 2.3677065440779668, + "learning_rate": 1.3559110611805542e-06, + "loss": 0.7767, + "num_input_tokens_seen": 222310170, + "step": 6833 + }, + { + "epoch": 0.6163141993957704, + "flos": 20015904076800.0, + "grad_norm": 2.1582350508455947, + "learning_rate": 1.3553580021271372e-06, + "loss": 0.79, + "num_input_tokens_seen": 222338115, + "step": 6834 + }, + { + "epoch": 0.6164043829192406, + "flos": 23589669337920.0, + "grad_norm": 1.7348063851755966, + "learning_rate": 1.3548049980781297e-06, + "loss": 0.6934, + "num_input_tokens_seen": 222367615, + "step": 6835 + }, + { + "epoch": 0.616494566442711, + "flos": 22605630983520.0, + "grad_norm": 1.8842269206658766, + "learning_rate": 1.3542520490807166e-06, + "loss": 0.7409, + "num_input_tokens_seen": 222394500, + "step": 6836 + }, + { + "epoch": 0.6165847499661812, + "flos": 15209060103840.0, + "grad_norm": 2.176481014569391, + "learning_rate": 1.3536991551820802e-06, + "loss": 0.7397, + "num_input_tokens_seen": 222420095, + "step": 6837 + }, + { + "epoch": 0.6166749334896514, + "flos": 37214344895520.0, + "grad_norm": 1.715029626039906, + "learning_rate": 1.3531463164293952e-06, + "loss": 0.7202, + "num_input_tokens_seen": 222455885, + "step": 6838 + }, + { + "epoch": 0.6167651170131218, + "flos": 20492234748480.0, + "grad_norm": 2.0030464573546576, + "learning_rate": 1.3525935328698332e-06, + "loss": 0.6603, + "num_input_tokens_seen": 222484030, + "step": 6839 + }, + { + "epoch": 0.616855300536592, + "flos": 23219810059680.0, + "grad_norm": 3.275597053442098, + "learning_rate": 1.3520408045505605e-06, + "loss": 0.6894, + "num_input_tokens_seen": 222512020, + "step": 6840 + }, + { + "epoch": 0.6169454840600622, + "flos": 24828546532320.0, + "grad_norm": 1.986453072098031, + "learning_rate": 1.3514881315187396e-06, + "loss": 0.7159, + "num_input_tokens_seen": 222541050, + "step": 6841 + }, + { + "epoch": 0.6170356675835325, + "flos": 24132508177440.0, + "grad_norm": 1.6641000376304602, + "learning_rate": 1.3509355138215273e-06, + "loss": 0.809, + "num_input_tokens_seen": 222568680, + "step": 6842 + }, + { + "epoch": 0.6171258511070028, + "flos": 21731632319520.0, + "grad_norm": 6.1543036932682, + "learning_rate": 1.350382951506075e-06, + "loss": 0.832, + "num_input_tokens_seen": 222596105, + "step": 6843 + }, + { + "epoch": 0.617216034630473, + "flos": 69451653988800.0, + "grad_norm": 0.6400769378192945, + "learning_rate": 1.349830444619532e-06, + "loss": 0.5669, + "num_input_tokens_seen": 222688050, + "step": 6844 + }, + { + "epoch": 0.6173062181539433, + "flos": 28617038728800.0, + "grad_norm": 3.127632100755813, + "learning_rate": 1.3492779932090397e-06, + "loss": 0.7647, + "num_input_tokens_seen": 222716850, + "step": 6845 + }, + { + "epoch": 0.6173964016774135, + "flos": 23261594603520.0, + "grad_norm": 1.863975859272868, + "learning_rate": 1.3487255973217377e-06, + "loss": 0.777, + "num_input_tokens_seen": 222747930, + "step": 6846 + }, + { + "epoch": 0.6174865852008838, + "flos": 23953730133600.0, + "grad_norm": 1.7026745961830347, + "learning_rate": 1.3481732570047592e-06, + "loss": 0.8216, + "num_input_tokens_seen": 222775960, + "step": 6847 + }, + { + "epoch": 0.6175767687243541, + "flos": 26940790504800.0, + "grad_norm": 2.9922152286370216, + "learning_rate": 1.3476209723052314e-06, + "loss": 0.6812, + "num_input_tokens_seen": 222804295, + "step": 6848 + }, + { + "epoch": 0.6176669522478243, + "flos": 23187875502240.0, + "grad_norm": 1.9338124777645103, + "learning_rate": 1.3470687432702806e-06, + "loss": 0.7753, + "num_input_tokens_seen": 222830945, + "step": 6849 + }, + { + "epoch": 0.6177571357712945, + "flos": 23152075289760.0, + "grad_norm": 2.4186508868115237, + "learning_rate": 1.346516569947024e-06, + "loss": 0.6814, + "num_input_tokens_seen": 222858835, + "step": 6850 + }, + { + "epoch": 0.6178473192947649, + "flos": 15283373921280.0, + "grad_norm": 2.0483545677919843, + "learning_rate": 1.3459644523825774e-06, + "loss": 0.7367, + "num_input_tokens_seen": 222887175, + "step": 6851 + }, + { + "epoch": 0.6179375028182351, + "flos": 16084099521120.0, + "grad_norm": 2.251370648291833, + "learning_rate": 1.34541239062405e-06, + "loss": 0.7893, + "num_input_tokens_seen": 222915485, + "step": 6852 + }, + { + "epoch": 0.6180276863417054, + "flos": 24318274136160.0, + "grad_norm": 1.543093055836341, + "learning_rate": 1.3448603847185464e-06, + "loss": 0.7504, + "num_input_tokens_seen": 222946700, + "step": 6853 + }, + { + "epoch": 0.6181178698651756, + "flos": 22819986221280.0, + "grad_norm": 1.7155610063913176, + "learning_rate": 1.344308434713168e-06, + "loss": 0.6906, + "num_input_tokens_seen": 222975075, + "step": 6854 + }, + { + "epoch": 0.6182080533886459, + "flos": 22860469823520.0, + "grad_norm": 2.2601983687177967, + "learning_rate": 1.3437565406550083e-06, + "loss": 0.6817, + "num_input_tokens_seen": 223006805, + "step": 6855 + }, + { + "epoch": 0.6182982369121162, + "flos": 13862001707040.0, + "grad_norm": 2.3066911683774993, + "learning_rate": 1.34320470259116e-06, + "loss": 0.7341, + "num_input_tokens_seen": 223033685, + "step": 6856 + }, + { + "epoch": 0.6183884204355864, + "flos": 27411657221760.0, + "grad_norm": 2.601264730372372, + "learning_rate": 1.3426529205687078e-06, + "loss": 0.6532, + "num_input_tokens_seen": 223063585, + "step": 6857 + }, + { + "epoch": 0.6184786039590566, + "flos": 24097823057760.0, + "grad_norm": 1.7787020345347562, + "learning_rate": 1.3421011946347323e-06, + "loss": 0.8061, + "num_input_tokens_seen": 223091145, + "step": 6858 + }, + { + "epoch": 0.618568787482527, + "flos": 19435518046080.0, + "grad_norm": 1.8964232131414325, + "learning_rate": 1.3415495248363113e-06, + "loss": 0.7118, + "num_input_tokens_seen": 223120735, + "step": 6859 + }, + { + "epoch": 0.6186589710059972, + "flos": 15975360772320.0, + "grad_norm": 2.638700698254632, + "learning_rate": 1.3409979112205148e-06, + "loss": 0.7733, + "num_input_tokens_seen": 223146820, + "step": 6860 + }, + { + "epoch": 0.6187491545294674, + "flos": 13279162472160.0, + "grad_norm": 2.034971496044988, + "learning_rate": 1.3404463538344107e-06, + "loss": 0.7795, + "num_input_tokens_seen": 223172565, + "step": 6861 + }, + { + "epoch": 0.6188393380529378, + "flos": 20928156157440.0, + "grad_norm": 2.171270821602733, + "learning_rate": 1.3398948527250602e-06, + "loss": 0.838, + "num_input_tokens_seen": 223194875, + "step": 6862 + }, + { + "epoch": 0.618929521576408, + "flos": 69727047706080.0, + "grad_norm": 1.458091125356416, + "learning_rate": 1.3393434079395207e-06, + "loss": 0.5909, + "num_input_tokens_seen": 223284280, + "step": 6863 + }, + { + "epoch": 0.6190197050998782, + "flos": 24931709817120.0, + "grad_norm": 1.7514468751256285, + "learning_rate": 1.3387920195248449e-06, + "loss": 0.7365, + "num_input_tokens_seen": 223314470, + "step": 6864 + }, + { + "epoch": 0.6191098886233485, + "flos": 28322459681760.0, + "grad_norm": 5.467173147387116, + "learning_rate": 1.3382406875280791e-06, + "loss": 0.7648, + "num_input_tokens_seen": 223345545, + "step": 6865 + }, + { + "epoch": 0.6192000721468188, + "flos": 17608932378240.0, + "grad_norm": 1.9576771292804986, + "learning_rate": 1.3376894119962672e-06, + "loss": 0.7741, + "num_input_tokens_seen": 223371965, + "step": 6866 + }, + { + "epoch": 0.6192902556702891, + "flos": 20966149385760.0, + "grad_norm": 2.262839979898642, + "learning_rate": 1.3371381929764464e-06, + "loss": 0.7773, + "num_input_tokens_seen": 223397085, + "step": 6867 + }, + { + "epoch": 0.6193804391937593, + "flos": 67807860698880.0, + "grad_norm": 0.6383853001104386, + "learning_rate": 1.3365870305156506e-06, + "loss": 0.5696, + "num_input_tokens_seen": 223486750, + "step": 6868 + }, + { + "epoch": 0.6194706227172295, + "flos": 22205992993920.0, + "grad_norm": 2.4023556592692614, + "learning_rate": 1.3360359246609073e-06, + "loss": 0.7535, + "num_input_tokens_seen": 223512795, + "step": 6869 + }, + { + "epoch": 0.6195608062406999, + "flos": 21184630466880.0, + "grad_norm": 1.7026204753042926, + "learning_rate": 1.3354848754592387e-06, + "loss": 0.7457, + "num_input_tokens_seen": 223541925, + "step": 6870 + }, + { + "epoch": 0.6196509897641701, + "flos": 29520853273920.0, + "grad_norm": 2.219655993674499, + "learning_rate": 1.334933882957666e-06, + "loss": 0.8385, + "num_input_tokens_seen": 223569980, + "step": 6871 + }, + { + "epoch": 0.6197411732876403, + "flos": 24317939608320.0, + "grad_norm": 1.948312980876226, + "learning_rate": 1.3343829472032004e-06, + "loss": 0.7141, + "num_input_tokens_seen": 223600295, + "step": 6872 + }, + { + "epoch": 0.6198313568111106, + "flos": 23152298308320.0, + "grad_norm": 2.637007995564415, + "learning_rate": 1.3338320682428527e-06, + "loss": 0.7354, + "num_input_tokens_seen": 223628995, + "step": 6873 + }, + { + "epoch": 0.6199215403345809, + "flos": 24677577202560.0, + "grad_norm": 2.091579621873437, + "learning_rate": 1.3332812461236263e-06, + "loss": 0.7129, + "num_input_tokens_seen": 223656880, + "step": 6874 + }, + { + "epoch": 0.6200117238580511, + "flos": 20893433868000.0, + "grad_norm": 1.5391394891217225, + "learning_rate": 1.3327304808925192e-06, + "loss": 0.7142, + "num_input_tokens_seen": 223686125, + "step": 6875 + }, + { + "epoch": 0.6201019073815214, + "flos": 30219716530560.0, + "grad_norm": 2.834246338738887, + "learning_rate": 1.332179772596528e-06, + "loss": 0.6125, + "num_input_tokens_seen": 223715500, + "step": 6876 + }, + { + "epoch": 0.6201920909049916, + "flos": 26973877324800.0, + "grad_norm": 1.838192078923896, + "learning_rate": 1.3316291212826402e-06, + "loss": 0.6751, + "num_input_tokens_seen": 223742125, + "step": 6877 + }, + { + "epoch": 0.620282274428462, + "flos": 19832405473440.0, + "grad_norm": 1.9752642925495167, + "learning_rate": 1.3310785269978413e-06, + "loss": 0.7993, + "num_input_tokens_seen": 223770450, + "step": 6878 + }, + { + "epoch": 0.6203724579519322, + "flos": 20310408784320.0, + "grad_norm": 6.405350189284211, + "learning_rate": 1.3305279897891111e-06, + "loss": 0.7902, + "num_input_tokens_seen": 223797500, + "step": 6879 + }, + { + "epoch": 0.6204626414754024, + "flos": 34955480455200.0, + "grad_norm": 2.7280762413535466, + "learning_rate": 1.329977509703424e-06, + "loss": 0.6145, + "num_input_tokens_seen": 223827655, + "step": 6880 + }, + { + "epoch": 0.6205528249988727, + "flos": 18962569822560.0, + "grad_norm": 2.1352724972571355, + "learning_rate": 1.32942708678775e-06, + "loss": 0.8371, + "num_input_tokens_seen": 223852665, + "step": 6881 + }, + { + "epoch": 0.620643008522343, + "flos": 25295250236160.0, + "grad_norm": 1.7455541813925461, + "learning_rate": 1.3288767210890548e-06, + "loss": 0.762, + "num_input_tokens_seen": 223882040, + "step": 6882 + }, + { + "epoch": 0.6207331920458132, + "flos": 23844656856960.0, + "grad_norm": 2.214882825072805, + "learning_rate": 1.3283264126542986e-06, + "loss": 0.715, + "num_input_tokens_seen": 223911150, + "step": 6883 + }, + { + "epoch": 0.6208233755692835, + "flos": 28284689472000.0, + "grad_norm": 1.7274539769138124, + "learning_rate": 1.3277761615304356e-06, + "loss": 0.7487, + "num_input_tokens_seen": 223941150, + "step": 6884 + }, + { + "epoch": 0.6209135590927538, + "flos": 24099198338880.0, + "grad_norm": 1.9859400907510085, + "learning_rate": 1.3272259677644178e-06, + "loss": 0.8494, + "num_input_tokens_seen": 223967865, + "step": 6885 + }, + { + "epoch": 0.621003742616224, + "flos": 24572964297120.0, + "grad_norm": 2.3409924370253656, + "learning_rate": 1.32667583140319e-06, + "loss": 0.6229, + "num_input_tokens_seen": 223998340, + "step": 6886 + }, + { + "epoch": 0.6210939261396943, + "flos": 13752036356160.0, + "grad_norm": 3.1346875902962523, + "learning_rate": 1.3261257524936924e-06, + "loss": 0.7892, + "num_input_tokens_seen": 224024125, + "step": 6887 + }, + { + "epoch": 0.6211841096631645, + "flos": 25590981545760.0, + "grad_norm": 1.7111445610828269, + "learning_rate": 1.3255757310828619e-06, + "loss": 0.7778, + "num_input_tokens_seen": 224054910, + "step": 6888 + }, + { + "epoch": 0.6212742931866349, + "flos": 13971112153440.0, + "grad_norm": 2.2641148236680526, + "learning_rate": 1.3250257672176282e-06, + "loss": 0.7676, + "num_input_tokens_seen": 224082005, + "step": 6889 + }, + { + "epoch": 0.6213644767101051, + "flos": 24825981818880.0, + "grad_norm": 1.982867111043189, + "learning_rate": 1.3244758609449183e-06, + "loss": 0.6564, + "num_input_tokens_seen": 224112300, + "step": 6890 + }, + { + "epoch": 0.6214546602335753, + "flos": 20447625302880.0, + "grad_norm": 1.9392902962826957, + "learning_rate": 1.323926012311653e-06, + "loss": 0.7691, + "num_input_tokens_seen": 224137325, + "step": 6891 + }, + { + "epoch": 0.6215448437570456, + "flos": 21913792811520.0, + "grad_norm": 3.8803155188722704, + "learning_rate": 1.3233762213647476e-06, + "loss": 0.7772, + "num_input_tokens_seen": 224165040, + "step": 6892 + }, + { + "epoch": 0.6216350272805159, + "flos": 27194885949600.0, + "grad_norm": 1.915212729368874, + "learning_rate": 1.3228264881511137e-06, + "loss": 0.725, + "num_input_tokens_seen": 224194335, + "step": 6893 + }, + { + "epoch": 0.6217252108039861, + "flos": 20783765875200.0, + "grad_norm": 1.722686974906696, + "learning_rate": 1.322276812717658e-06, + "loss": 0.8311, + "num_input_tokens_seen": 224223025, + "step": 6894 + }, + { + "epoch": 0.6218153943274564, + "flos": 27669878509920.0, + "grad_norm": 1.921485054960555, + "learning_rate": 1.3217271951112818e-06, + "loss": 0.6589, + "num_input_tokens_seen": 224256850, + "step": 6895 + }, + { + "epoch": 0.6219055778509266, + "flos": 21907102254720.0, + "grad_norm": 2.174932509504246, + "learning_rate": 1.321177635378881e-06, + "loss": 0.7322, + "num_input_tokens_seen": 224283900, + "step": 6896 + }, + { + "epoch": 0.6219957613743969, + "flos": 35539694971200.0, + "grad_norm": 2.4660110732214346, + "learning_rate": 1.3206281335673475e-06, + "loss": 0.642, + "num_input_tokens_seen": 224317700, + "step": 6897 + }, + { + "epoch": 0.6220859448978672, + "flos": 25264876808640.0, + "grad_norm": 2.255743537339739, + "learning_rate": 1.3200786897235677e-06, + "loss": 0.7252, + "num_input_tokens_seen": 224344865, + "step": 6898 + }, + { + "epoch": 0.6221761284213374, + "flos": 21767506871520.0, + "grad_norm": 1.8409487894489447, + "learning_rate": 1.3195293038944227e-06, + "loss": 0.7659, + "num_input_tokens_seen": 224372430, + "step": 6899 + }, + { + "epoch": 0.6222663119448076, + "flos": 34770495061440.0, + "grad_norm": 1.6539479217541149, + "learning_rate": 1.3189799761267902e-06, + "loss": 0.743, + "num_input_tokens_seen": 224406740, + "step": 6900 + }, + { + "epoch": 0.622356495468278, + "flos": 19837014523680.0, + "grad_norm": 3.511100332696952, + "learning_rate": 1.3184307064675412e-06, + "loss": 0.7457, + "num_input_tokens_seen": 224434025, + "step": 6901 + }, + { + "epoch": 0.6224466789917482, + "flos": 62763201636000.0, + "grad_norm": 0.6727708566041715, + "learning_rate": 1.3178814949635416e-06, + "loss": 0.6238, + "num_input_tokens_seen": 224524915, + "step": 6902 + }, + { + "epoch": 0.6225368625152184, + "flos": 27406787983200.0, + "grad_norm": 1.9639290411497214, + "learning_rate": 1.3173323416616549e-06, + "loss": 0.6982, + "num_input_tokens_seen": 224554220, + "step": 6903 + }, + { + "epoch": 0.6226270460386887, + "flos": 26539740064320.0, + "grad_norm": 2.4260435592187553, + "learning_rate": 1.3167832466087361e-06, + "loss": 0.804, + "num_input_tokens_seen": 224582590, + "step": 6904 + }, + { + "epoch": 0.622717229562159, + "flos": 26321147473920.0, + "grad_norm": 1.6772173794394323, + "learning_rate": 1.3162342098516388e-06, + "loss": 0.6988, + "num_input_tokens_seen": 224614030, + "step": 6905 + }, + { + "epoch": 0.6228074130856293, + "flos": 21913495453440.0, + "grad_norm": 1.7595107627511293, + "learning_rate": 1.3156852314372086e-06, + "loss": 0.6703, + "num_input_tokens_seen": 224643485, + "step": 6906 + }, + { + "epoch": 0.6228975966090995, + "flos": 68429548066560.0, + "grad_norm": 0.6769942714119163, + "learning_rate": 1.3151363114122882e-06, + "loss": 0.5595, + "num_input_tokens_seen": 224734525, + "step": 6907 + }, + { + "epoch": 0.6229877801325698, + "flos": 23953692963840.0, + "grad_norm": 1.722583518363457, + "learning_rate": 1.3145874498237133e-06, + "loss": 0.7633, + "num_input_tokens_seen": 224765535, + "step": 6908 + }, + { + "epoch": 0.6230779636560401, + "flos": 22750578812160.0, + "grad_norm": 1.6675493843280844, + "learning_rate": 1.3140386467183166e-06, + "loss": 0.7981, + "num_input_tokens_seen": 224793320, + "step": 6909 + }, + { + "epoch": 0.6231681471795103, + "flos": 24607798095840.0, + "grad_norm": 1.5380560978008246, + "learning_rate": 1.3134899021429258e-06, + "loss": 0.7909, + "num_input_tokens_seen": 224822165, + "step": 6910 + }, + { + "epoch": 0.6232583307029805, + "flos": 27268047504480.0, + "grad_norm": 2.150336075176635, + "learning_rate": 1.3129412161443604e-06, + "loss": 0.7334, + "num_input_tokens_seen": 224851335, + "step": 6911 + }, + { + "epoch": 0.6233485142264509, + "flos": 23698482426240.0, + "grad_norm": 1.7277524136027016, + "learning_rate": 1.3123925887694402e-06, + "loss": 0.7561, + "num_input_tokens_seen": 224880280, + "step": 6912 + }, + { + "epoch": 0.6234386977499211, + "flos": 24753117622080.0, + "grad_norm": 1.6631161860046444, + "learning_rate": 1.3118440200649756e-06, + "loss": 0.7605, + "num_input_tokens_seen": 224908900, + "step": 6913 + }, + { + "epoch": 0.6235288812733913, + "flos": 21439171948800.0, + "grad_norm": 1.6969455289848507, + "learning_rate": 1.3112955100777727e-06, + "loss": 0.7113, + "num_input_tokens_seen": 224939425, + "step": 6914 + }, + { + "epoch": 0.6236190647968616, + "flos": 24971264175360.0, + "grad_norm": 2.262044142011587, + "learning_rate": 1.3107470588546353e-06, + "loss": 0.7874, + "num_input_tokens_seen": 224969360, + "step": 6915 + }, + { + "epoch": 0.6237092483203319, + "flos": 24354483216000.0, + "grad_norm": 2.053895514842143, + "learning_rate": 1.3101986664423583e-06, + "loss": 0.7212, + "num_input_tokens_seen": 224997775, + "step": 6916 + }, + { + "epoch": 0.6237994318438022, + "flos": 30289272618720.0, + "grad_norm": 2.0171296174175657, + "learning_rate": 1.3096503328877358e-06, + "loss": 0.7252, + "num_input_tokens_seen": 225025580, + "step": 6917 + }, + { + "epoch": 0.6238896153672724, + "flos": 20598483123360.0, + "grad_norm": 1.7804113722791752, + "learning_rate": 1.309102058237553e-06, + "loss": 0.77, + "num_input_tokens_seen": 225054800, + "step": 6918 + }, + { + "epoch": 0.6239797988907426, + "flos": 25664886495840.0, + "grad_norm": 6.391780235876405, + "learning_rate": 1.3085538425385917e-06, + "loss": 0.801, + "num_input_tokens_seen": 225085815, + "step": 6919 + }, + { + "epoch": 0.624069982414213, + "flos": 19545780755040.0, + "grad_norm": 2.2641425182452686, + "learning_rate": 1.3080056858376298e-06, + "loss": 0.7837, + "num_input_tokens_seen": 225113960, + "step": 6920 + }, + { + "epoch": 0.6241601659376832, + "flos": 29892533870400.0, + "grad_norm": 2.4058106256604086, + "learning_rate": 1.3074575881814383e-06, + "loss": 0.7359, + "num_input_tokens_seen": 225140825, + "step": 6921 + }, + { + "epoch": 0.6242503494611534, + "flos": 29782865877600.0, + "grad_norm": 3.4364587767432, + "learning_rate": 1.3069095496167841e-06, + "loss": 0.6727, + "num_input_tokens_seen": 225170010, + "step": 6922 + }, + { + "epoch": 0.6243405329846237, + "flos": 22164171280320.0, + "grad_norm": 2.037916424094231, + "learning_rate": 1.3063615701904285e-06, + "loss": 0.6642, + "num_input_tokens_seen": 225197905, + "step": 6923 + }, + { + "epoch": 0.624430716508094, + "flos": 23407137148320.0, + "grad_norm": 2.44367794701387, + "learning_rate": 1.3058136499491283e-06, + "loss": 0.782, + "num_input_tokens_seen": 225225685, + "step": 6924 + }, + { + "epoch": 0.6245209000315642, + "flos": 20565544982400.0, + "grad_norm": 2.273775370552241, + "learning_rate": 1.3052657889396357e-06, + "loss": 0.7325, + "num_input_tokens_seen": 225254530, + "step": 6925 + }, + { + "epoch": 0.6246110835550345, + "flos": 21986917196640.0, + "grad_norm": 2.659642017196572, + "learning_rate": 1.304717987208696e-06, + "loss": 0.737, + "num_input_tokens_seen": 225280145, + "step": 6926 + }, + { + "epoch": 0.6247012670785047, + "flos": 27232953517440.0, + "grad_norm": 1.6605094982634825, + "learning_rate": 1.304170244803052e-06, + "loss": 0.7374, + "num_input_tokens_seen": 225310445, + "step": 6927 + }, + { + "epoch": 0.624791450601975, + "flos": 31204275261600.0, + "grad_norm": 2.0507524179520864, + "learning_rate": 1.3036225617694387e-06, + "loss": 0.7168, + "num_input_tokens_seen": 225342000, + "step": 6928 + }, + { + "epoch": 0.6248816341254453, + "flos": 20420113946880.0, + "grad_norm": 1.606755393112246, + "learning_rate": 1.3030749381545892e-06, + "loss": 0.7803, + "num_input_tokens_seen": 225369505, + "step": 6929 + }, + { + "epoch": 0.6249718176489155, + "flos": 15865767119040.0, + "grad_norm": 2.5240528785377667, + "learning_rate": 1.3025273740052285e-06, + "loss": 0.6865, + "num_input_tokens_seen": 225397155, + "step": 6930 + }, + { + "epoch": 0.6250620011723858, + "flos": 24681703045920.0, + "grad_norm": 1.534965134389343, + "learning_rate": 1.3019798693680774e-06, + "loss": 0.7942, + "num_input_tokens_seen": 225429335, + "step": 6931 + }, + { + "epoch": 0.6251521846958561, + "flos": 25300900039680.0, + "grad_norm": 1.9517390712638751, + "learning_rate": 1.3014324242898536e-06, + "loss": 0.8659, + "num_input_tokens_seen": 225458050, + "step": 6932 + }, + { + "epoch": 0.6252423682193263, + "flos": 21767841399360.0, + "grad_norm": 4.059896016034246, + "learning_rate": 1.3008850388172668e-06, + "loss": 0.8061, + "num_input_tokens_seen": 225485915, + "step": 6933 + }, + { + "epoch": 0.6253325517427966, + "flos": 28220931866400.0, + "grad_norm": 1.4504333506238922, + "learning_rate": 1.3003377129970233e-06, + "loss": 0.7056, + "num_input_tokens_seen": 225518585, + "step": 6934 + }, + { + "epoch": 0.6254227352662669, + "flos": 33972185496000.0, + "grad_norm": 1.4327280200609511, + "learning_rate": 1.2997904468758243e-06, + "loss": 0.668, + "num_input_tokens_seen": 225553685, + "step": 6935 + }, + { + "epoch": 0.6255129187897371, + "flos": 26432413766400.0, + "grad_norm": 1.8709345931053996, + "learning_rate": 1.299243240500365e-06, + "loss": 0.7639, + "num_input_tokens_seen": 225581025, + "step": 6936 + }, + { + "epoch": 0.6256031023132074, + "flos": 25957458375840.0, + "grad_norm": 1.8461847554883843, + "learning_rate": 1.2986960939173368e-06, + "loss": 0.7654, + "num_input_tokens_seen": 225608885, + "step": 6937 + }, + { + "epoch": 0.6256932858366776, + "flos": 15792717073440.0, + "grad_norm": 2.0770339970164096, + "learning_rate": 1.298149007173424e-06, + "loss": 0.7909, + "num_input_tokens_seen": 225633985, + "step": 6938 + }, + { + "epoch": 0.625783469360148, + "flos": 17651162959200.0, + "grad_norm": 2.1047747145766205, + "learning_rate": 1.2976019803153087e-06, + "loss": 0.7798, + "num_input_tokens_seen": 225660600, + "step": 6939 + }, + { + "epoch": 0.6258736528836182, + "flos": 30544929193440.0, + "grad_norm": 1.919343350893895, + "learning_rate": 1.2970550133896652e-06, + "loss": 0.694, + "num_input_tokens_seen": 225690795, + "step": 6940 + }, + { + "epoch": 0.6259638364070884, + "flos": 20711199036480.0, + "grad_norm": 2.1457467169702653, + "learning_rate": 1.2965081064431634e-06, + "loss": 0.8319, + "num_input_tokens_seen": 225717000, + "step": 6941 + }, + { + "epoch": 0.6260540199305586, + "flos": 19691397639360.0, + "grad_norm": 2.7542505360201415, + "learning_rate": 1.2959612595224698e-06, + "loss": 0.7704, + "num_input_tokens_seen": 225743030, + "step": 6942 + }, + { + "epoch": 0.626144203454029, + "flos": 15645501889440.0, + "grad_norm": 2.4015216692380323, + "learning_rate": 1.2954144726742424e-06, + "loss": 0.7615, + "num_input_tokens_seen": 225766075, + "step": 6943 + }, + { + "epoch": 0.6262343869774992, + "flos": 26540817987360.0, + "grad_norm": 1.7561225736282098, + "learning_rate": 1.2948677459451385e-06, + "loss": 0.7438, + "num_input_tokens_seen": 225797525, + "step": 6944 + }, + { + "epoch": 0.6263245705009695, + "flos": 23881609332000.0, + "grad_norm": 2.464922194887572, + "learning_rate": 1.2943210793818064e-06, + "loss": 0.7946, + "num_input_tokens_seen": 225824965, + "step": 6945 + }, + { + "epoch": 0.6264147540244397, + "flos": 32369730712800.0, + "grad_norm": 2.0610080293353694, + "learning_rate": 1.2937744730308899e-06, + "loss": 0.7282, + "num_input_tokens_seen": 225856305, + "step": 6946 + }, + { + "epoch": 0.62650493754791, + "flos": 26650002773280.0, + "grad_norm": 2.7612490760452846, + "learning_rate": 1.2932279269390305e-06, + "loss": 0.7444, + "num_input_tokens_seen": 225884515, + "step": 6947 + }, + { + "epoch": 0.6265951210713803, + "flos": 22278113795520.0, + "grad_norm": 1.9546138124428152, + "learning_rate": 1.292681441152861e-06, + "loss": 0.6522, + "num_input_tokens_seen": 225912110, + "step": 6948 + }, + { + "epoch": 0.6266853045948505, + "flos": 24463816680960.0, + "grad_norm": 2.1064582155811347, + "learning_rate": 1.292135015719011e-06, + "loss": 0.7635, + "num_input_tokens_seen": 225940985, + "step": 6949 + }, + { + "epoch": 0.6267754881183207, + "flos": 18889519776960.0, + "grad_norm": 1.931510897794035, + "learning_rate": 1.2915886506841046e-06, + "loss": 0.7062, + "num_input_tokens_seen": 225969400, + "step": 6950 + }, + { + "epoch": 0.6268656716417911, + "flos": 35575569523200.0, + "grad_norm": 1.6319921157403374, + "learning_rate": 1.2910423460947613e-06, + "loss": 0.6818, + "num_input_tokens_seen": 226001060, + "step": 6951 + }, + { + "epoch": 0.6269558551652613, + "flos": 24134106477120.0, + "grad_norm": 1.9627641172113939, + "learning_rate": 1.290496101997594e-06, + "loss": 0.718, + "num_input_tokens_seen": 226029105, + "step": 6952 + }, + { + "epoch": 0.6270460386887315, + "flos": 62796771662880.0, + "grad_norm": 0.7183115077704342, + "learning_rate": 1.2899499184392105e-06, + "loss": 0.5809, + "num_input_tokens_seen": 226119540, + "step": 6953 + }, + { + "epoch": 0.6271362222122018, + "flos": 24352327369920.0, + "grad_norm": 2.0072052627388057, + "learning_rate": 1.289403795466216e-06, + "loss": 0.7272, + "num_input_tokens_seen": 226148495, + "step": 6954 + }, + { + "epoch": 0.6272264057356721, + "flos": 17977936752000.0, + "grad_norm": 2.618709838850569, + "learning_rate": 1.288857733125207e-06, + "loss": 0.7637, + "num_input_tokens_seen": 226173755, + "step": 6955 + }, + { + "epoch": 0.6273165892591424, + "flos": 24864309575040.0, + "grad_norm": 2.015654807646604, + "learning_rate": 1.2883117314627785e-06, + "loss": 0.7424, + "num_input_tokens_seen": 226202820, + "step": 6956 + }, + { + "epoch": 0.6274067727826126, + "flos": 25556407935360.0, + "grad_norm": 1.7151823991142883, + "learning_rate": 1.2877657905255168e-06, + "loss": 0.688, + "num_input_tokens_seen": 226233255, + "step": 6957 + }, + { + "epoch": 0.6274969563060829, + "flos": 23110996971360.0, + "grad_norm": 1.8517033004341557, + "learning_rate": 1.2872199103600046e-06, + "loss": 0.8109, + "num_input_tokens_seen": 226261610, + "step": 6958 + }, + { + "epoch": 0.6275871398295532, + "flos": 24938103015840.0, + "grad_norm": 2.0102157755940087, + "learning_rate": 1.286674091012821e-06, + "loss": 0.8172, + "num_input_tokens_seen": 226291955, + "step": 6959 + }, + { + "epoch": 0.6276773233530234, + "flos": 19873818319680.0, + "grad_norm": 1.7921757824888036, + "learning_rate": 1.2861283325305356e-06, + "loss": 0.805, + "num_input_tokens_seen": 226317010, + "step": 6960 + }, + { + "epoch": 0.6277675068764936, + "flos": 21035668304160.0, + "grad_norm": 2.5337040625737806, + "learning_rate": 1.2855826349597185e-06, + "loss": 0.7464, + "num_input_tokens_seen": 226343745, + "step": 6961 + }, + { + "epoch": 0.627857690399964, + "flos": 19035545528640.0, + "grad_norm": 1.5593863935952594, + "learning_rate": 1.2850369983469302e-06, + "loss": 0.6046, + "num_input_tokens_seen": 226372575, + "step": 6962 + }, + { + "epoch": 0.6279478739234342, + "flos": 26314902954240.0, + "grad_norm": 1.8739635781989952, + "learning_rate": 1.2844914227387266e-06, + "loss": 0.775, + "num_input_tokens_seen": 226400760, + "step": 6963 + }, + { + "epoch": 0.6280380574469044, + "flos": 29382038455680.0, + "grad_norm": 1.5088047853903666, + "learning_rate": 1.2839459081816606e-06, + "loss": 0.7712, + "num_input_tokens_seen": 226432885, + "step": 6964 + }, + { + "epoch": 0.6281282409703747, + "flos": 21942939636960.0, + "grad_norm": 2.165705198534188, + "learning_rate": 1.283400454722278e-06, + "loss": 0.8168, + "num_input_tokens_seen": 226457660, + "step": 6965 + }, + { + "epoch": 0.628218424493845, + "flos": 21907325273280.0, + "grad_norm": 1.8520124830233111, + "learning_rate": 1.28285506240712e-06, + "loss": 0.8122, + "num_input_tokens_seen": 226485795, + "step": 6966 + }, + { + "epoch": 0.6283086080173153, + "flos": 25630610243520.0, + "grad_norm": 1.6016393875646484, + "learning_rate": 1.2823097312827225e-06, + "loss": 0.7773, + "num_input_tokens_seen": 226517045, + "step": 6967 + }, + { + "epoch": 0.6283987915407855, + "flos": 24171690838080.0, + "grad_norm": 1.7515874883007565, + "learning_rate": 1.2817644613956153e-06, + "loss": 0.7583, + "num_input_tokens_seen": 226545995, + "step": 6968 + }, + { + "epoch": 0.6284889750642557, + "flos": 22715075957760.0, + "grad_norm": 6.822885419080005, + "learning_rate": 1.2812192527923253e-06, + "loss": 0.8098, + "num_input_tokens_seen": 226572860, + "step": 6969 + }, + { + "epoch": 0.6285791585877261, + "flos": 27738170826240.0, + "grad_norm": 2.088073208625324, + "learning_rate": 1.2806741055193712e-06, + "loss": 0.7814, + "num_input_tokens_seen": 226601890, + "step": 6970 + }, + { + "epoch": 0.6286693421111963, + "flos": 13898099277600.0, + "grad_norm": 3.0577659607290517, + "learning_rate": 1.2801290196232695e-06, + "loss": 0.7059, + "num_input_tokens_seen": 226628605, + "step": 6971 + }, + { + "epoch": 0.6287595256346665, + "flos": 24318236966400.0, + "grad_norm": 2.184170505190895, + "learning_rate": 1.2795839951505282e-06, + "loss": 0.7408, + "num_input_tokens_seen": 226658180, + "step": 6972 + }, + { + "epoch": 0.6288497091581368, + "flos": 14882100462240.0, + "grad_norm": 2.715278426026909, + "learning_rate": 1.2790390321476538e-06, + "loss": 0.7692, + "num_input_tokens_seen": 226684695, + "step": 6973 + }, + { + "epoch": 0.6289398926816071, + "flos": 24499802742240.0, + "grad_norm": 2.201036694495932, + "learning_rate": 1.2784941306611446e-06, + "loss": 0.7226, + "num_input_tokens_seen": 226712750, + "step": 6974 + }, + { + "epoch": 0.6290300762050773, + "flos": 16886534929920.0, + "grad_norm": 2.2565591362596753, + "learning_rate": 1.2779492907374935e-06, + "loss": 0.8091, + "num_input_tokens_seen": 226739740, + "step": 6975 + }, + { + "epoch": 0.6291202597285476, + "flos": 20893285188960.0, + "grad_norm": 1.9189462443342202, + "learning_rate": 1.2774045124231911e-06, + "loss": 0.7359, + "num_input_tokens_seen": 226766265, + "step": 6976 + }, + { + "epoch": 0.6292104432520178, + "flos": 16339644586560.0, + "grad_norm": 1.5416232266191252, + "learning_rate": 1.2768597957647197e-06, + "loss": 0.6318, + "num_input_tokens_seen": 226792870, + "step": 6977 + }, + { + "epoch": 0.6293006267754881, + "flos": 20712053940960.0, + "grad_norm": 2.836304359772528, + "learning_rate": 1.2763151408085582e-06, + "loss": 0.7605, + "num_input_tokens_seen": 226819630, + "step": 6978 + }, + { + "epoch": 0.6293908102989584, + "flos": 26759150389440.0, + "grad_norm": 1.9369064775575937, + "learning_rate": 1.2757705476011788e-06, + "loss": 0.7176, + "num_input_tokens_seen": 226849570, + "step": 6979 + }, + { + "epoch": 0.6294809938224286, + "flos": 23480484552000.0, + "grad_norm": 2.308766831366089, + "learning_rate": 1.27522601618905e-06, + "loss": 0.695, + "num_input_tokens_seen": 226879025, + "step": 6980 + }, + { + "epoch": 0.629571177345899, + "flos": 64620160731360.0, + "grad_norm": 0.920095925357728, + "learning_rate": 1.2746815466186337e-06, + "loss": 0.6033, + "num_input_tokens_seen": 226976485, + "step": 6981 + }, + { + "epoch": 0.6296613608693692, + "flos": 29195566271520.0, + "grad_norm": 1.872614169135365, + "learning_rate": 1.274137138936387e-06, + "loss": 0.7516, + "num_input_tokens_seen": 227006190, + "step": 6982 + }, + { + "epoch": 0.6297515443928394, + "flos": 46609551760320.0, + "grad_norm": 6.446654689794787, + "learning_rate": 1.2735927931887625e-06, + "loss": 0.749, + "num_input_tokens_seen": 227041610, + "step": 6983 + }, + { + "epoch": 0.6298417279163097, + "flos": 66909621617760.0, + "grad_norm": 0.6837574789973699, + "learning_rate": 1.2730485094222061e-06, + "loss": 0.5852, + "num_input_tokens_seen": 227130895, + "step": 6984 + }, + { + "epoch": 0.62993191143978, + "flos": 22125843524160.0, + "grad_norm": 1.7887352390066793, + "learning_rate": 1.2725042876831586e-06, + "loss": 0.7868, + "num_input_tokens_seen": 227159620, + "step": 6985 + }, + { + "epoch": 0.6300220949632502, + "flos": 25514883579840.0, + "grad_norm": 2.418755768367385, + "learning_rate": 1.2719601280180573e-06, + "loss": 0.6467, + "num_input_tokens_seen": 227188910, + "step": 6986 + }, + { + "epoch": 0.6301122784867205, + "flos": 25659422541120.0, + "grad_norm": 2.2444186391175798, + "learning_rate": 1.2714160304733317e-06, + "loss": 0.8155, + "num_input_tokens_seen": 227216055, + "step": 6987 + }, + { + "epoch": 0.6302024620101907, + "flos": 24971821721760.0, + "grad_norm": 1.9278454965954315, + "learning_rate": 1.2708719950954082e-06, + "loss": 0.7622, + "num_input_tokens_seen": 227245510, + "step": 6988 + }, + { + "epoch": 0.630292645533661, + "flos": 18307275258240.0, + "grad_norm": 1.9002641594611056, + "learning_rate": 1.2703280219307065e-06, + "loss": 0.7668, + "num_input_tokens_seen": 227272845, + "step": 6989 + }, + { + "epoch": 0.6303828290571313, + "flos": 20780643615360.0, + "grad_norm": 1.9549705026064719, + "learning_rate": 1.2697841110256411e-06, + "loss": 0.6811, + "num_input_tokens_seen": 227299065, + "step": 6990 + }, + { + "epoch": 0.6304730125806015, + "flos": 21950522268000.0, + "grad_norm": 2.8844199725743884, + "learning_rate": 1.2692402624266221e-06, + "loss": 0.7946, + "num_input_tokens_seen": 227327740, + "step": 6991 + }, + { + "epoch": 0.6305631961040717, + "flos": 23079768639360.0, + "grad_norm": 2.603066224605866, + "learning_rate": 1.2686964761800529e-06, + "loss": 0.6583, + "num_input_tokens_seen": 227355865, + "step": 6992 + }, + { + "epoch": 0.6306533796275421, + "flos": 70546066561440.0, + "grad_norm": 0.6395500282123563, + "learning_rate": 1.268152752332333e-06, + "loss": 0.5833, + "num_input_tokens_seen": 227449640, + "step": 6993 + }, + { + "epoch": 0.6307435631510123, + "flos": 33863632596000.0, + "grad_norm": 5.738802746576269, + "learning_rate": 1.2676090909298549e-06, + "loss": 0.6649, + "num_input_tokens_seen": 227480610, + "step": 6994 + }, + { + "epoch": 0.6308337466744826, + "flos": 26467693602240.0, + "grad_norm": 3.0811040569043846, + "learning_rate": 1.2670654920190086e-06, + "loss": 0.795, + "num_input_tokens_seen": 227510345, + "step": 6995 + }, + { + "epoch": 0.6309239301979528, + "flos": 26065825427040.0, + "grad_norm": 1.6849540295541643, + "learning_rate": 1.2665219556461754e-06, + "loss": 0.7997, + "num_input_tokens_seen": 227541515, + "step": 6996 + }, + { + "epoch": 0.6310141137214231, + "flos": 21986880026880.0, + "grad_norm": 2.843926392838053, + "learning_rate": 1.2659784818577329e-06, + "loss": 0.848, + "num_input_tokens_seen": 227569365, + "step": 6997 + }, + { + "epoch": 0.6311042972448934, + "flos": 31712949358080.0, + "grad_norm": 1.5839935565875232, + "learning_rate": 1.2654350707000542e-06, + "loss": 0.6346, + "num_input_tokens_seen": 227601145, + "step": 6998 + }, + { + "epoch": 0.6311944807683636, + "flos": 23990236571520.0, + "grad_norm": 1.8128745251141853, + "learning_rate": 1.264891722219505e-06, + "loss": 0.7353, + "num_input_tokens_seen": 227629000, + "step": 6999 + }, + { + "epoch": 0.6312846642918338, + "flos": 23079545620800.0, + "grad_norm": 8.075643545672246, + "learning_rate": 1.2643484364624483e-06, + "loss": 0.7428, + "num_input_tokens_seen": 227657965, + "step": 7000 + }, + { + "epoch": 0.6313748478153042, + "flos": 42716000621280.0, + "grad_norm": 2.0018858363310534, + "learning_rate": 1.2638052134752393e-06, + "loss": 0.8004, + "num_input_tokens_seen": 227691325, + "step": 7001 + }, + { + "epoch": 0.6314650313387744, + "flos": 22203985826880.0, + "grad_norm": 1.8969233247486799, + "learning_rate": 1.2632620533042277e-06, + "loss": 0.8231, + "num_input_tokens_seen": 227718515, + "step": 7002 + }, + { + "epoch": 0.6315552148622446, + "flos": 21471069336480.0, + "grad_norm": 2.339135675808239, + "learning_rate": 1.2627189559957612e-06, + "loss": 0.7188, + "num_input_tokens_seen": 227745980, + "step": 7003 + }, + { + "epoch": 0.631645398385715, + "flos": 18707284945440.0, + "grad_norm": 2.955738828309971, + "learning_rate": 1.262175921596178e-06, + "loss": 0.8072, + "num_input_tokens_seen": 227771435, + "step": 7004 + }, + { + "epoch": 0.6317355819091852, + "flos": 20528109300480.0, + "grad_norm": 2.983291828854227, + "learning_rate": 1.2616329501518137e-06, + "loss": 0.6228, + "num_input_tokens_seen": 227798750, + "step": 7005 + }, + { + "epoch": 0.6318257654326555, + "flos": 27556456371360.0, + "grad_norm": 6.4075581029191735, + "learning_rate": 1.2610900417089978e-06, + "loss": 0.6714, + "num_input_tokens_seen": 227829925, + "step": 7006 + }, + { + "epoch": 0.6319159489561257, + "flos": 25703102742720.0, + "grad_norm": 1.8185827853490664, + "learning_rate": 1.2605471963140535e-06, + "loss": 0.7592, + "num_input_tokens_seen": 227860935, + "step": 7007 + }, + { + "epoch": 0.632006132479596, + "flos": 25444323908160.0, + "grad_norm": 1.9740046057062808, + "learning_rate": 1.2600044140133e-06, + "loss": 0.8232, + "num_input_tokens_seen": 227889130, + "step": 7008 + }, + { + "epoch": 0.6320963160030663, + "flos": 19326593448480.0, + "grad_norm": 1.7197030632844006, + "learning_rate": 1.2594616948530493e-06, + "loss": 0.7012, + "num_input_tokens_seen": 227917565, + "step": 7009 + }, + { + "epoch": 0.6321864995265365, + "flos": 17432756217600.0, + "grad_norm": 2.192182488411948, + "learning_rate": 1.258919038879611e-06, + "loss": 0.7719, + "num_input_tokens_seen": 227944315, + "step": 7010 + }, + { + "epoch": 0.6322766830500067, + "flos": 16266631710720.0, + "grad_norm": 2.0173852612320555, + "learning_rate": 1.2583764461392859e-06, + "loss": 0.7468, + "num_input_tokens_seen": 227972430, + "step": 7011 + }, + { + "epoch": 0.6323668665734771, + "flos": 30767238759840.0, + "grad_norm": 1.6661203760664909, + "learning_rate": 1.2578339166783724e-06, + "loss": 0.7113, + "num_input_tokens_seen": 228003960, + "step": 7012 + }, + { + "epoch": 0.6324570500969473, + "flos": 23553608937120.0, + "grad_norm": 3.7636788432002826, + "learning_rate": 1.2572914505431613e-06, + "loss": 0.6855, + "num_input_tokens_seen": 228033730, + "step": 7013 + }, + { + "epoch": 0.6325472336204175, + "flos": 35212437971520.0, + "grad_norm": 1.5633666877333452, + "learning_rate": 1.2567490477799383e-06, + "loss": 0.6477, + "num_input_tokens_seen": 228070940, + "step": 7014 + }, + { + "epoch": 0.6326374171438878, + "flos": 23915216528640.0, + "grad_norm": 2.2939865738692826, + "learning_rate": 1.2562067084349852e-06, + "loss": 0.8422, + "num_input_tokens_seen": 228100705, + "step": 7015 + }, + { + "epoch": 0.6327276006673581, + "flos": 22967870460960.0, + "grad_norm": 2.0372268774536937, + "learning_rate": 1.2556644325545764e-06, + "loss": 0.7315, + "num_input_tokens_seen": 228130795, + "step": 7016 + }, + { + "epoch": 0.6328177841908283, + "flos": 21695497579200.0, + "grad_norm": 1.8404040607172958, + "learning_rate": 1.255122220184983e-06, + "loss": 0.7303, + "num_input_tokens_seen": 228158830, + "step": 7017 + }, + { + "epoch": 0.6329079677142986, + "flos": 22350940822560.0, + "grad_norm": 2.2625281906479224, + "learning_rate": 1.2545800713724694e-06, + "loss": 0.6564, + "num_input_tokens_seen": 228185845, + "step": 7018 + }, + { + "epoch": 0.6329981512377688, + "flos": 36376741160160.0, + "grad_norm": 1.8490602993496734, + "learning_rate": 1.254037986163294e-06, + "loss": 0.6126, + "num_input_tokens_seen": 228217215, + "step": 7019 + }, + { + "epoch": 0.6330883347612392, + "flos": 53500830991680.0, + "grad_norm": 1.8059745921801054, + "learning_rate": 1.2534959646037104e-06, + "loss": 0.7042, + "num_input_tokens_seen": 228250005, + "step": 7020 + }, + { + "epoch": 0.6331785182847094, + "flos": 25739088804000.0, + "grad_norm": 1.9484519902893487, + "learning_rate": 1.2529540067399675e-06, + "loss": 0.7524, + "num_input_tokens_seen": 228279680, + "step": 7021 + }, + { + "epoch": 0.6332687018081796, + "flos": 19436633138880.0, + "grad_norm": 1.9836833679700403, + "learning_rate": 1.252412112618308e-06, + "loss": 0.68, + "num_input_tokens_seen": 228307030, + "step": 7022 + }, + { + "epoch": 0.6333588853316499, + "flos": 25994299341600.0, + "grad_norm": 1.9919464563887326, + "learning_rate": 1.2518702822849696e-06, + "loss": 0.6306, + "num_input_tokens_seen": 228334980, + "step": 7023 + }, + { + "epoch": 0.6334490688551202, + "flos": 17578633290240.0, + "grad_norm": 2.483500488680566, + "learning_rate": 1.2513285157861831e-06, + "loss": 0.6906, + "num_input_tokens_seen": 228362090, + "step": 7024 + }, + { + "epoch": 0.6335392523785904, + "flos": 25409564448960.0, + "grad_norm": 1.7044568934698165, + "learning_rate": 1.2507868131681764e-06, + "loss": 0.7418, + "num_input_tokens_seen": 228391845, + "step": 7025 + }, + { + "epoch": 0.6336294359020607, + "flos": 24098194755360.0, + "grad_norm": 1.9182830988020663, + "learning_rate": 1.250245174477169e-06, + "loss": 0.7312, + "num_input_tokens_seen": 228421115, + "step": 7026 + }, + { + "epoch": 0.6337196194255309, + "flos": 70621198113600.0, + "grad_norm": 0.6475089675326529, + "learning_rate": 1.2497035997593783e-06, + "loss": 0.5193, + "num_input_tokens_seen": 228507195, + "step": 7027 + }, + { + "epoch": 0.6338098029490012, + "flos": 30438532139520.0, + "grad_norm": 1.6365196817102072, + "learning_rate": 1.2491620890610135e-06, + "loss": 0.7452, + "num_input_tokens_seen": 228537300, + "step": 7028 + }, + { + "epoch": 0.6338999864724715, + "flos": 23366765055360.0, + "grad_norm": 1.720001320062175, + "learning_rate": 1.2486206424282788e-06, + "loss": 0.7309, + "num_input_tokens_seen": 228564240, + "step": 7029 + }, + { + "epoch": 0.6339901699959417, + "flos": 21840705596160.0, + "grad_norm": 1.8764215802365267, + "learning_rate": 1.2480792599073743e-06, + "loss": 0.8031, + "num_input_tokens_seen": 228592225, + "step": 7030 + }, + { + "epoch": 0.634080353519412, + "flos": 19980624240960.0, + "grad_norm": 1.73463032535649, + "learning_rate": 1.247537941544493e-06, + "loss": 0.7895, + "num_input_tokens_seen": 228621715, + "step": 7031 + }, + { + "epoch": 0.6341705370428823, + "flos": 22235288498400.0, + "grad_norm": 14.620170643615536, + "learning_rate": 1.2469966873858242e-06, + "loss": 0.7798, + "num_input_tokens_seen": 228650240, + "step": 7032 + }, + { + "epoch": 0.6342607205663525, + "flos": 23733873771360.0, + "grad_norm": 2.04175643983033, + "learning_rate": 1.2464554974775496e-06, + "loss": 0.6695, + "num_input_tokens_seen": 228675155, + "step": 7033 + }, + { + "epoch": 0.6343509040898228, + "flos": 17396101100640.0, + "grad_norm": 2.7092110800957667, + "learning_rate": 1.2459143718658474e-06, + "loss": 0.7464, + "num_input_tokens_seen": 228701805, + "step": 7034 + }, + { + "epoch": 0.6344410876132931, + "flos": 23698445256480.0, + "grad_norm": 2.266327847607622, + "learning_rate": 1.2453733105968886e-06, + "loss": 0.6825, + "num_input_tokens_seen": 228730775, + "step": 7035 + }, + { + "epoch": 0.6345312711367633, + "flos": 67556441476800.0, + "grad_norm": 0.627844441324307, + "learning_rate": 1.2448323137168399e-06, + "loss": 0.5845, + "num_input_tokens_seen": 228819535, + "step": 7036 + }, + { + "epoch": 0.6346214546602336, + "flos": 41113136970720.0, + "grad_norm": 1.7378370767662028, + "learning_rate": 1.2442913812718625e-06, + "loss": 0.6686, + "num_input_tokens_seen": 228852595, + "step": 7037 + }, + { + "epoch": 0.6347116381837038, + "flos": 34769974684800.0, + "grad_norm": 1.592318621356588, + "learning_rate": 1.2437505133081108e-06, + "loss": 0.7307, + "num_input_tokens_seen": 228882965, + "step": 7038 + }, + { + "epoch": 0.6348018217071741, + "flos": 20018543129760.0, + "grad_norm": 6.32139670523308, + "learning_rate": 1.2432097098717358e-06, + "loss": 0.7607, + "num_input_tokens_seen": 228908605, + "step": 7039 + }, + { + "epoch": 0.6348920052306444, + "flos": 21798735203520.0, + "grad_norm": 2.000266534211314, + "learning_rate": 1.2426689710088813e-06, + "loss": 0.7502, + "num_input_tokens_seen": 228936750, + "step": 7040 + }, + { + "epoch": 0.6349821887541146, + "flos": 17393610726720.0, + "grad_norm": 5.116971496216372, + "learning_rate": 1.2421282967656855e-06, + "loss": 0.8019, + "num_input_tokens_seen": 228963145, + "step": 7041 + }, + { + "epoch": 0.6350723722775848, + "flos": 21293480724960.0, + "grad_norm": 1.7956277250168993, + "learning_rate": 1.2415876871882827e-06, + "loss": 0.7738, + "num_input_tokens_seen": 228991915, + "step": 7042 + }, + { + "epoch": 0.6351625558010552, + "flos": 69728385817440.0, + "grad_norm": 0.674115658654854, + "learning_rate": 1.2410471423227998e-06, + "loss": 0.6487, + "num_input_tokens_seen": 229095590, + "step": 7043 + }, + { + "epoch": 0.6352527393245254, + "flos": 24937136602080.0, + "grad_norm": 1.7779567457887242, + "learning_rate": 1.24050666221536e-06, + "loss": 0.7386, + "num_input_tokens_seen": 229126765, + "step": 7044 + }, + { + "epoch": 0.6353429228479957, + "flos": 20565805170720.0, + "grad_norm": 1.982751016587942, + "learning_rate": 1.23996624691208e-06, + "loss": 0.8113, + "num_input_tokens_seen": 229154925, + "step": 7045 + }, + { + "epoch": 0.6354331063714659, + "flos": 27485487832320.0, + "grad_norm": 1.5776218329777842, + "learning_rate": 1.2394258964590693e-06, + "loss": 0.73, + "num_input_tokens_seen": 229186020, + "step": 7046 + }, + { + "epoch": 0.6355232898949362, + "flos": 24864160896000.0, + "grad_norm": 1.912453422727192, + "learning_rate": 1.238885610902436e-06, + "loss": 0.6951, + "num_input_tokens_seen": 229213365, + "step": 7047 + }, + { + "epoch": 0.6356134734184065, + "flos": 28723881819840.0, + "grad_norm": 2.624365775375343, + "learning_rate": 1.2383453902882787e-06, + "loss": 0.7576, + "num_input_tokens_seen": 229242345, + "step": 7048 + }, + { + "epoch": 0.6357036569418767, + "flos": 37282079665440.0, + "grad_norm": 1.5573757068152645, + "learning_rate": 1.2378052346626927e-06, + "loss": 0.8098, + "num_input_tokens_seen": 229276435, + "step": 7049 + }, + { + "epoch": 0.6357938404653469, + "flos": 14845296666240.0, + "grad_norm": 2.5456152298606587, + "learning_rate": 1.2372651440717665e-06, + "loss": 0.6739, + "num_input_tokens_seen": 229302800, + "step": 7050 + }, + { + "epoch": 0.6358840239888173, + "flos": 22023163446240.0, + "grad_norm": 1.8656941591156564, + "learning_rate": 1.236725118561584e-06, + "loss": 0.7795, + "num_input_tokens_seen": 229328955, + "step": 7051 + }, + { + "epoch": 0.6359742075122875, + "flos": 24498018593760.0, + "grad_norm": 1.933149128666788, + "learning_rate": 1.2361851581782232e-06, + "loss": 0.7454, + "num_input_tokens_seen": 229359215, + "step": 7052 + }, + { + "epoch": 0.6360643910357577, + "flos": 22095767454720.0, + "grad_norm": 1.7877297580550853, + "learning_rate": 1.2356452629677554e-06, + "loss": 0.8098, + "num_input_tokens_seen": 229386675, + "step": 7053 + }, + { + "epoch": 0.6361545745592281, + "flos": 34227953580000.0, + "grad_norm": 1.9366616311831293, + "learning_rate": 1.2351054329762494e-06, + "loss": 0.6778, + "num_input_tokens_seen": 229419910, + "step": 7054 + }, + { + "epoch": 0.6362447580826983, + "flos": 15356238118080.0, + "grad_norm": 1.848566262668998, + "learning_rate": 1.2345656682497648e-06, + "loss": 0.7687, + "num_input_tokens_seen": 229444965, + "step": 7055 + }, + { + "epoch": 0.6363349416061685, + "flos": 23189027764800.0, + "grad_norm": 1.9691978429628685, + "learning_rate": 1.2340259688343583e-06, + "loss": 0.7194, + "num_input_tokens_seen": 229476660, + "step": 7056 + }, + { + "epoch": 0.6364251251296388, + "flos": 24208383124800.0, + "grad_norm": 1.6460808274133465, + "learning_rate": 1.2334863347760803e-06, + "loss": 0.6468, + "num_input_tokens_seen": 229509835, + "step": 7057 + }, + { + "epoch": 0.6365153086531091, + "flos": 54270991581600.0, + "grad_norm": 0.6408276586277206, + "learning_rate": 1.2329467661209734e-06, + "loss": 0.5439, + "num_input_tokens_seen": 229600380, + "step": 7058 + }, + { + "epoch": 0.6366054921765794, + "flos": 27160684036800.0, + "grad_norm": 1.9201560047468298, + "learning_rate": 1.2324072629150788e-06, + "loss": 0.7889, + "num_input_tokens_seen": 229627965, + "step": 7059 + }, + { + "epoch": 0.6366956757000496, + "flos": 18196306323840.0, + "grad_norm": 3.053593198081984, + "learning_rate": 1.2318678252044287e-06, + "loss": 0.7971, + "num_input_tokens_seen": 229653010, + "step": 7060 + }, + { + "epoch": 0.6367858592235198, + "flos": 20491974560160.0, + "grad_norm": 2.616526141513754, + "learning_rate": 1.2313284530350512e-06, + "loss": 0.7994, + "num_input_tokens_seen": 229679875, + "step": 7061 + }, + { + "epoch": 0.6368760427469902, + "flos": 20378515251840.0, + "grad_norm": 1.9754004398439309, + "learning_rate": 1.230789146452969e-06, + "loss": 0.6822, + "num_input_tokens_seen": 229707335, + "step": 7062 + }, + { + "epoch": 0.6369662262704604, + "flos": 24463370643840.0, + "grad_norm": 2.0850778015404035, + "learning_rate": 1.2302499055041974e-06, + "loss": 0.7514, + "num_input_tokens_seen": 229735585, + "step": 7063 + }, + { + "epoch": 0.6370564097939306, + "flos": 64696110018240.0, + "grad_norm": 0.6621466207015198, + "learning_rate": 1.2297107302347488e-06, + "loss": 0.5814, + "num_input_tokens_seen": 229825195, + "step": 7064 + }, + { + "epoch": 0.6371465933174009, + "flos": 24569433169920.0, + "grad_norm": 1.8787900131340247, + "learning_rate": 1.2291716206906275e-06, + "loss": 0.8142, + "num_input_tokens_seen": 229853755, + "step": 7065 + }, + { + "epoch": 0.6372367768408712, + "flos": 31093826703840.0, + "grad_norm": 2.462204476840359, + "learning_rate": 1.2286325769178345e-06, + "loss": 0.812, + "num_input_tokens_seen": 229879935, + "step": 7066 + }, + { + "epoch": 0.6373269603643414, + "flos": 22678606689600.0, + "grad_norm": 1.7873914485595792, + "learning_rate": 1.2280935989623633e-06, + "loss": 0.7732, + "num_input_tokens_seen": 229907050, + "step": 7067 + }, + { + "epoch": 0.6374171438878117, + "flos": 24427273073280.0, + "grad_norm": 2.301151783353131, + "learning_rate": 1.2275546868702017e-06, + "loss": 0.6807, + "num_input_tokens_seen": 229934645, + "step": 7068 + }, + { + "epoch": 0.6375073274112819, + "flos": 17542052512800.0, + "grad_norm": 1.7426319325222104, + "learning_rate": 1.2270158406873341e-06, + "loss": 0.6835, + "num_input_tokens_seen": 229962705, + "step": 7069 + }, + { + "epoch": 0.6375975109347523, + "flos": 25810540549920.0, + "grad_norm": 1.895268688025735, + "learning_rate": 1.2264770604597363e-06, + "loss": 0.7871, + "num_input_tokens_seen": 229994060, + "step": 7070 + }, + { + "epoch": 0.6376876944582225, + "flos": 22019000433120.0, + "grad_norm": 1.7238461629864739, + "learning_rate": 1.2259383462333819e-06, + "loss": 0.845, + "num_input_tokens_seen": 230021920, + "step": 7071 + }, + { + "epoch": 0.6377778779816927, + "flos": 23509111000800.0, + "grad_norm": 1.8808994308753946, + "learning_rate": 1.2253996980542359e-06, + "loss": 0.7022, + "num_input_tokens_seen": 230049790, + "step": 7072 + }, + { + "epoch": 0.637868061505163, + "flos": 23115531682080.0, + "grad_norm": 1.7579078478914834, + "learning_rate": 1.2248611159682578e-06, + "loss": 0.7284, + "num_input_tokens_seen": 230076705, + "step": 7073 + }, + { + "epoch": 0.6379582450286333, + "flos": 63608016304800.0, + "grad_norm": 0.6889035679571833, + "learning_rate": 1.2243226000214044e-06, + "loss": 0.5624, + "num_input_tokens_seen": 230165935, + "step": 7074 + }, + { + "epoch": 0.6380484285521035, + "flos": 25666968002400.0, + "grad_norm": 2.097829346290114, + "learning_rate": 1.2237841502596232e-06, + "loss": 0.7499, + "num_input_tokens_seen": 230195730, + "step": 7075 + }, + { + "epoch": 0.6381386120755738, + "flos": 27487792357440.0, + "grad_norm": 2.0523787026102718, + "learning_rate": 1.2232457667288583e-06, + "loss": 0.7528, + "num_input_tokens_seen": 230226665, + "step": 7076 + }, + { + "epoch": 0.6382287955990441, + "flos": 23516470613280.0, + "grad_norm": 2.9668396274281377, + "learning_rate": 1.2227074494750476e-06, + "loss": 0.7499, + "num_input_tokens_seen": 230253840, + "step": 7077 + }, + { + "epoch": 0.6383189791225143, + "flos": 25006358162400.0, + "grad_norm": 2.2818259988255054, + "learning_rate": 1.2221691985441238e-06, + "loss": 0.7796, + "num_input_tokens_seen": 230282550, + "step": 7078 + }, + { + "epoch": 0.6384091626459846, + "flos": 23370816559200.0, + "grad_norm": 2.194738672814935, + "learning_rate": 1.2216310139820128e-06, + "loss": 0.6935, + "num_input_tokens_seen": 230311240, + "step": 7079 + }, + { + "epoch": 0.6384993461694548, + "flos": 26429254336800.0, + "grad_norm": 1.682680594854147, + "learning_rate": 1.2210928958346347e-06, + "loss": 0.7897, + "num_input_tokens_seen": 230343450, + "step": 7080 + }, + { + "epoch": 0.6385895296929252, + "flos": 25555924728480.0, + "grad_norm": 1.550822990105581, + "learning_rate": 1.2205548441479065e-06, + "loss": 0.6753, + "num_input_tokens_seen": 230374280, + "step": 7081 + }, + { + "epoch": 0.6386797132163954, + "flos": 24390134749440.0, + "grad_norm": 2.0856595086297043, + "learning_rate": 1.2200168589677357e-06, + "loss": 0.6835, + "num_input_tokens_seen": 230403590, + "step": 7082 + }, + { + "epoch": 0.6387698967398656, + "flos": 24861819201120.0, + "grad_norm": 1.6377204979522122, + "learning_rate": 1.2194789403400284e-06, + "loss": 0.7155, + "num_input_tokens_seen": 230432565, + "step": 7083 + }, + { + "epoch": 0.6388600802633358, + "flos": 29855655734880.0, + "grad_norm": 3.030979105062776, + "learning_rate": 1.2189410883106816e-06, + "loss": 0.7266, + "num_input_tokens_seen": 230461275, + "step": 7084 + }, + { + "epoch": 0.6389502637868062, + "flos": 20929271250240.0, + "grad_norm": 1.968641611173402, + "learning_rate": 1.2184033029255872e-06, + "loss": 0.8119, + "num_input_tokens_seen": 230492695, + "step": 7085 + }, + { + "epoch": 0.6390404473102764, + "flos": 69316593316320.0, + "grad_norm": 0.644458720432296, + "learning_rate": 1.2178655842306334e-06, + "loss": 0.58, + "num_input_tokens_seen": 230589815, + "step": 7086 + }, + { + "epoch": 0.6391306308337467, + "flos": 45156691025760.0, + "grad_norm": 2.1734035245213006, + "learning_rate": 1.2173279322716999e-06, + "loss": 0.7465, + "num_input_tokens_seen": 230620740, + "step": 7087 + }, + { + "epoch": 0.6392208143572169, + "flos": 20820123634080.0, + "grad_norm": 2.009389888536222, + "learning_rate": 1.216790347094663e-06, + "loss": 0.7484, + "num_input_tokens_seen": 230648515, + "step": 7088 + }, + { + "epoch": 0.6393109978806872, + "flos": 28252754914560.0, + "grad_norm": 1.8759348128447446, + "learning_rate": 1.2162528287453927e-06, + "loss": 0.7871, + "num_input_tokens_seen": 230677825, + "step": 7089 + }, + { + "epoch": 0.6394011814041575, + "flos": 21440101192800.0, + "grad_norm": 2.452752832257307, + "learning_rate": 1.215715377269752e-06, + "loss": 0.7511, + "num_input_tokens_seen": 230702650, + "step": 7090 + }, + { + "epoch": 0.6394913649276277, + "flos": 26136310759200.0, + "grad_norm": 2.1748964438504625, + "learning_rate": 1.2151779927136003e-06, + "loss": 0.7286, + "num_input_tokens_seen": 230732775, + "step": 7091 + }, + { + "epoch": 0.6395815484510979, + "flos": 22165583731200.0, + "grad_norm": 1.8295441686193856, + "learning_rate": 1.2146406751227893e-06, + "loss": 0.8182, + "num_input_tokens_seen": 230759430, + "step": 7092 + }, + { + "epoch": 0.6396717319745683, + "flos": 22605556644000.0, + "grad_norm": 1.8396329009771668, + "learning_rate": 1.214103424543167e-06, + "loss": 0.7745, + "num_input_tokens_seen": 230787025, + "step": 7093 + }, + { + "epoch": 0.6397619154980385, + "flos": 26468028130080.0, + "grad_norm": 3.3970477189246076, + "learning_rate": 1.2135662410205735e-06, + "loss": 0.6607, + "num_input_tokens_seen": 230819110, + "step": 7094 + }, + { + "epoch": 0.6398520990215087, + "flos": 15242146923840.0, + "grad_norm": 2.305632109947751, + "learning_rate": 1.2130291246008444e-06, + "loss": 0.6645, + "num_input_tokens_seen": 230843075, + "step": 7095 + }, + { + "epoch": 0.639942282544979, + "flos": 33497750482080.0, + "grad_norm": 2.631155245200656, + "learning_rate": 1.21249207532981e-06, + "loss": 0.6644, + "num_input_tokens_seen": 230875390, + "step": 7096 + }, + { + "epoch": 0.6400324660684493, + "flos": 21907882819680.0, + "grad_norm": 1.985267138245877, + "learning_rate": 1.2119550932532936e-06, + "loss": 0.6674, + "num_input_tokens_seen": 230904030, + "step": 7097 + }, + { + "epoch": 0.6401226495919196, + "flos": 22315103440320.0, + "grad_norm": 1.8687028681192248, + "learning_rate": 1.2114181784171144e-06, + "loss": 0.7219, + "num_input_tokens_seen": 230932695, + "step": 7098 + }, + { + "epoch": 0.6402128331153898, + "flos": 23697404503200.0, + "grad_norm": 1.927427999930418, + "learning_rate": 1.2108813308670837e-06, + "loss": 0.6491, + "num_input_tokens_seen": 230964590, + "step": 7099 + }, + { + "epoch": 0.6403030166388601, + "flos": 21986731347840.0, + "grad_norm": 2.3921358408168234, + "learning_rate": 1.2103445506490099e-06, + "loss": 0.7491, + "num_input_tokens_seen": 230992845, + "step": 7100 + }, + { + "epoch": 0.6403932001623304, + "flos": 71050726323840.0, + "grad_norm": 0.7203868802739352, + "learning_rate": 1.209807837808693e-06, + "loss": 0.5919, + "num_input_tokens_seen": 231078160, + "step": 7101 + }, + { + "epoch": 0.6404833836858006, + "flos": 33571432413600.0, + "grad_norm": 2.3298787929459626, + "learning_rate": 1.2092711923919282e-06, + "loss": 0.7853, + "num_input_tokens_seen": 231104890, + "step": 7102 + }, + { + "epoch": 0.6405735672092708, + "flos": 19690356886080.0, + "grad_norm": 1.60279188654945, + "learning_rate": 1.2087346144445053e-06, + "loss": 0.7062, + "num_input_tokens_seen": 231132885, + "step": 7103 + }, + { + "epoch": 0.6406637507327412, + "flos": 23880865936800.0, + "grad_norm": 2.1398101146626876, + "learning_rate": 1.2081981040122081e-06, + "loss": 0.755, + "num_input_tokens_seen": 231158950, + "step": 7104 + }, + { + "epoch": 0.6407539342562114, + "flos": 38637241069920.0, + "grad_norm": 1.692268697916281, + "learning_rate": 1.2076616611408148e-06, + "loss": 0.6959, + "num_input_tokens_seen": 231194430, + "step": 7105 + }, + { + "epoch": 0.6408441177796816, + "flos": 28616815710240.0, + "grad_norm": 2.0181712546294714, + "learning_rate": 1.2071252858760972e-06, + "loss": 0.7771, + "num_input_tokens_seen": 231224475, + "step": 7106 + }, + { + "epoch": 0.6409343013031519, + "flos": 20711793752640.0, + "grad_norm": 2.3615846707528845, + "learning_rate": 1.2065889782638218e-06, + "loss": 0.7311, + "num_input_tokens_seen": 231252115, + "step": 7107 + }, + { + "epoch": 0.6410244848266222, + "flos": 20274348383520.0, + "grad_norm": 1.7271438931598733, + "learning_rate": 1.2060527383497506e-06, + "loss": 0.7236, + "num_input_tokens_seen": 231280115, + "step": 7108 + }, + { + "epoch": 0.6411146683500925, + "flos": 21727655155200.0, + "grad_norm": 1.8462706969798577, + "learning_rate": 1.2055165661796363e-06, + "loss": 0.8578, + "num_input_tokens_seen": 231309250, + "step": 7109 + }, + { + "epoch": 0.6412048518735627, + "flos": 22897608147360.0, + "grad_norm": 2.136145589831609, + "learning_rate": 1.2049804617992303e-06, + "loss": 0.816, + "num_input_tokens_seen": 231335865, + "step": 7110 + }, + { + "epoch": 0.6412950353970329, + "flos": 54007380678240.0, + "grad_norm": 0.6189331525364683, + "learning_rate": 1.204444425254275e-06, + "loss": 0.5919, + "num_input_tokens_seen": 231422960, + "step": 7111 + }, + { + "epoch": 0.6413852189205033, + "flos": 21112026458400.0, + "grad_norm": 1.9002989759624203, + "learning_rate": 1.203908456590507e-06, + "loss": 0.7298, + "num_input_tokens_seen": 231453345, + "step": 7112 + }, + { + "epoch": 0.6414754024439735, + "flos": 20602237269120.0, + "grad_norm": 1.872285976733059, + "learning_rate": 1.20337255585366e-06, + "loss": 0.7208, + "num_input_tokens_seen": 231479435, + "step": 7113 + }, + { + "epoch": 0.6415655859674437, + "flos": 58975192646400.0, + "grad_norm": 0.7901774168583376, + "learning_rate": 1.2028367230894582e-06, + "loss": 0.5378, + "num_input_tokens_seen": 231565625, + "step": 7114 + }, + { + "epoch": 0.641655769490914, + "flos": 21839962200960.0, + "grad_norm": 2.0448937764828736, + "learning_rate": 1.2023009583436237e-06, + "loss": 0.6666, + "num_input_tokens_seen": 231594015, + "step": 7115 + }, + { + "epoch": 0.6417459530143843, + "flos": 23037129191040.0, + "grad_norm": 2.0268774065867396, + "learning_rate": 1.2017652616618698e-06, + "loss": 0.81, + "num_input_tokens_seen": 231620415, + "step": 7116 + }, + { + "epoch": 0.6418361365378545, + "flos": 22454735993280.0, + "grad_norm": 2.474972476228202, + "learning_rate": 1.2012296330899048e-06, + "loss": 0.7404, + "num_input_tokens_seen": 231648930, + "step": 7117 + }, + { + "epoch": 0.6419263200613248, + "flos": 27159568944000.0, + "grad_norm": 2.3203726096025608, + "learning_rate": 1.200694072673432e-06, + "loss": 0.777, + "num_input_tokens_seen": 231678530, + "step": 7118 + }, + { + "epoch": 0.642016503584795, + "flos": 18924613764000.0, + "grad_norm": 1.8024470967268424, + "learning_rate": 1.200158580458148e-06, + "loss": 0.7762, + "num_input_tokens_seen": 231706385, + "step": 7119 + }, + { + "epoch": 0.6421066871082654, + "flos": 27595453183200.0, + "grad_norm": 2.010348132716216, + "learning_rate": 1.1996231564897448e-06, + "loss": 0.7201, + "num_input_tokens_seen": 231736495, + "step": 7120 + }, + { + "epoch": 0.6421968706317356, + "flos": 21291213369600.0, + "grad_norm": 1.9227583181111771, + "learning_rate": 1.1990878008139062e-06, + "loss": 0.7008, + "num_input_tokens_seen": 231762715, + "step": 7121 + }, + { + "epoch": 0.6422870541552058, + "flos": 48721684223520.0, + "grad_norm": 2.5608287106228222, + "learning_rate": 1.1985525134763132e-06, + "loss": 0.7738, + "num_input_tokens_seen": 231793020, + "step": 7122 + }, + { + "epoch": 0.6423772376786762, + "flos": 29558697823200.0, + "grad_norm": 1.832728141540883, + "learning_rate": 1.1980172945226393e-06, + "loss": 0.7913, + "num_input_tokens_seen": 231823400, + "step": 7123 + }, + { + "epoch": 0.6424674212021464, + "flos": 18561742400640.0, + "grad_norm": 2.139647991784527, + "learning_rate": 1.197482143998551e-06, + "loss": 0.7503, + "num_input_tokens_seen": 231849075, + "step": 7124 + }, + { + "epoch": 0.6425576047256166, + "flos": 20274088195200.0, + "grad_norm": 1.9986047646418297, + "learning_rate": 1.196947061949712e-06, + "loss": 0.7303, + "num_input_tokens_seen": 231876765, + "step": 7125 + }, + { + "epoch": 0.6426477882490869, + "flos": 18708177019680.0, + "grad_norm": 1.7848554562299406, + "learning_rate": 1.1964120484217768e-06, + "loss": 0.8028, + "num_input_tokens_seen": 231903930, + "step": 7126 + }, + { + "epoch": 0.6427379717725572, + "flos": 62195161699200.0, + "grad_norm": 0.6986107127144472, + "learning_rate": 1.1958771034603975e-06, + "loss": 0.5622, + "num_input_tokens_seen": 231988655, + "step": 7127 + }, + { + "epoch": 0.6428281552960274, + "flos": 22749538058880.0, + "grad_norm": 2.1756613679523955, + "learning_rate": 1.1953422271112175e-06, + "loss": 0.6865, + "num_input_tokens_seen": 232018685, + "step": 7128 + }, + { + "epoch": 0.6429183388194977, + "flos": 22205063749920.0, + "grad_norm": 2.1001575410877384, + "learning_rate": 1.1948074194198748e-06, + "loss": 0.7325, + "num_input_tokens_seen": 232045595, + "step": 7129 + }, + { + "epoch": 0.6430085223429679, + "flos": 38777579848320.0, + "grad_norm": 2.4305150625095275, + "learning_rate": 1.1942726804320033e-06, + "loss": 0.6574, + "num_input_tokens_seen": 232078010, + "step": 7130 + }, + { + "epoch": 0.6430987058664382, + "flos": 17395989591360.0, + "grad_norm": 2.234951025065681, + "learning_rate": 1.1937380101932295e-06, + "loss": 0.688, + "num_input_tokens_seen": 232104095, + "step": 7131 + }, + { + "epoch": 0.6431888893899085, + "flos": 21076635113280.0, + "grad_norm": 1.5490589330654083, + "learning_rate": 1.1932034087491745e-06, + "loss": 0.7424, + "num_input_tokens_seen": 232134255, + "step": 7132 + }, + { + "epoch": 0.6432790729133787, + "flos": 32693345076000.0, + "grad_norm": 1.9322004133888389, + "learning_rate": 1.1926688761454531e-06, + "loss": 0.6602, + "num_input_tokens_seen": 232163385, + "step": 7133 + }, + { + "epoch": 0.643369256436849, + "flos": 28945745349120.0, + "grad_norm": 1.6857405779984134, + "learning_rate": 1.1921344124276746e-06, + "loss": 0.7649, + "num_input_tokens_seen": 232195270, + "step": 7134 + }, + { + "epoch": 0.6434594399603193, + "flos": 17752876623360.0, + "grad_norm": 1.8154740104989775, + "learning_rate": 1.1916000176414431e-06, + "loss": 0.7668, + "num_input_tokens_seen": 232221345, + "step": 7135 + }, + { + "epoch": 0.6435496234837895, + "flos": 25625629495680.0, + "grad_norm": 2.0468360955421323, + "learning_rate": 1.1910656918323546e-06, + "loss": 0.7825, + "num_input_tokens_seen": 232248095, + "step": 7136 + }, + { + "epoch": 0.6436398070072598, + "flos": 30220088228160.0, + "grad_norm": 2.1436948274235674, + "learning_rate": 1.1905314350460024e-06, + "loss": 0.7715, + "num_input_tokens_seen": 232279340, + "step": 7137 + }, + { + "epoch": 0.64372999053073, + "flos": 24354520385760.0, + "grad_norm": 1.7077794301285558, + "learning_rate": 1.1899972473279717e-06, + "loss": 0.8008, + "num_input_tokens_seen": 232308490, + "step": 7138 + }, + { + "epoch": 0.6438201740542003, + "flos": 24791854245600.0, + "grad_norm": 1.9699470876955838, + "learning_rate": 1.1894631287238414e-06, + "loss": 0.7359, + "num_input_tokens_seen": 232337615, + "step": 7139 + }, + { + "epoch": 0.6439103575776706, + "flos": 25514846410080.0, + "grad_norm": 1.6555804932755023, + "learning_rate": 1.188929079279187e-06, + "loss": 0.7569, + "num_input_tokens_seen": 232367800, + "step": 7140 + }, + { + "epoch": 0.6440005411011408, + "flos": 21257643342720.0, + "grad_norm": 1.806570284737095, + "learning_rate": 1.1883950990395751e-06, + "loss": 0.6835, + "num_input_tokens_seen": 232395460, + "step": 7141 + }, + { + "epoch": 0.644090724624611, + "flos": 20674729768320.0, + "grad_norm": 2.048386576213775, + "learning_rate": 1.187861188050569e-06, + "loss": 0.6815, + "num_input_tokens_seen": 232424405, + "step": 7142 + }, + { + "epoch": 0.6441809081480814, + "flos": 23079173923200.0, + "grad_norm": 1.7449387521570197, + "learning_rate": 1.187327346357724e-06, + "loss": 0.7326, + "num_input_tokens_seen": 232455705, + "step": 7143 + }, + { + "epoch": 0.6442710916715516, + "flos": 19435964083200.0, + "grad_norm": 2.1102858095812795, + "learning_rate": 1.1867935740065912e-06, + "loss": 0.7695, + "num_input_tokens_seen": 232482935, + "step": 7144 + }, + { + "epoch": 0.6443612751950218, + "flos": 28433056918560.0, + "grad_norm": 2.6091422702457328, + "learning_rate": 1.1862598710427148e-06, + "loss": 0.6347, + "num_input_tokens_seen": 232516005, + "step": 7145 + }, + { + "epoch": 0.6444514587184921, + "flos": 15100247015520.0, + "grad_norm": 3.7016382250723, + "learning_rate": 1.1857262375116328e-06, + "loss": 0.7075, + "num_input_tokens_seen": 232542140, + "step": 7146 + }, + { + "epoch": 0.6445416422419624, + "flos": 28143570128640.0, + "grad_norm": 2.188501837064004, + "learning_rate": 1.1851926734588783e-06, + "loss": 0.7468, + "num_input_tokens_seen": 232572710, + "step": 7147 + }, + { + "epoch": 0.6446318257654327, + "flos": 19873037754720.0, + "grad_norm": 2.0961129817869484, + "learning_rate": 1.184659178929977e-06, + "loss": 0.7671, + "num_input_tokens_seen": 232598580, + "step": 7148 + }, + { + "epoch": 0.6447220092889029, + "flos": 19069412913600.0, + "grad_norm": 1.7808069046099153, + "learning_rate": 1.1841257539704513e-06, + "loss": 0.7414, + "num_input_tokens_seen": 232628715, + "step": 7149 + }, + { + "epoch": 0.6448121928123732, + "flos": 23844471008160.0, + "grad_norm": 2.2108156420012683, + "learning_rate": 1.1835923986258146e-06, + "loss": 0.7642, + "num_input_tokens_seen": 232658020, + "step": 7150 + }, + { + "epoch": 0.6449023763358435, + "flos": 21147863840640.0, + "grad_norm": 1.6167718472954913, + "learning_rate": 1.1830591129415754e-06, + "loss": 0.7062, + "num_input_tokens_seen": 232687020, + "step": 7151 + }, + { + "epoch": 0.6449925598593137, + "flos": 25045801011360.0, + "grad_norm": 2.561870311501041, + "learning_rate": 1.182525896963238e-06, + "loss": 0.6908, + "num_input_tokens_seen": 232716335, + "step": 7152 + }, + { + "epoch": 0.6450827433827839, + "flos": 19180902224640.0, + "grad_norm": 4.294599450721647, + "learning_rate": 1.181992750736298e-06, + "loss": 0.6757, + "num_input_tokens_seen": 232743345, + "step": 7153 + }, + { + "epoch": 0.6451729269062543, + "flos": 21476458951680.0, + "grad_norm": 2.327897335928351, + "learning_rate": 1.1814596743062474e-06, + "loss": 0.7512, + "num_input_tokens_seen": 232769465, + "step": 7154 + }, + { + "epoch": 0.6452631104297245, + "flos": 21543338817120.0, + "grad_norm": 2.569296128629033, + "learning_rate": 1.1809266677185711e-06, + "loss": 0.7689, + "num_input_tokens_seen": 232797435, + "step": 7155 + }, + { + "epoch": 0.6453532939531947, + "flos": 67308813570240.0, + "grad_norm": 0.651454982531398, + "learning_rate": 1.180393731018747e-06, + "loss": 0.6119, + "num_input_tokens_seen": 232896200, + "step": 7156 + }, + { + "epoch": 0.645443477476665, + "flos": 53969536128960.0, + "grad_norm": 0.6502504398149398, + "learning_rate": 1.1798608642522498e-06, + "loss": 0.5509, + "num_input_tokens_seen": 232990835, + "step": 7157 + }, + { + "epoch": 0.6455336610001353, + "flos": 25447520507520.0, + "grad_norm": 3.016863158625378, + "learning_rate": 1.1793280674645454e-06, + "loss": 0.7272, + "num_input_tokens_seen": 233019345, + "step": 7158 + }, + { + "epoch": 0.6456238445236056, + "flos": 17942248048800.0, + "grad_norm": 3.133055717939575, + "learning_rate": 1.1787953407010954e-06, + "loss": 0.8186, + "num_input_tokens_seen": 233046285, + "step": 7159 + }, + { + "epoch": 0.6457140280470758, + "flos": 25702768214880.0, + "grad_norm": 1.9922447046780853, + "learning_rate": 1.1782626840073554e-06, + "loss": 0.7494, + "num_input_tokens_seen": 233075925, + "step": 7160 + }, + { + "epoch": 0.645804211570546, + "flos": 30402025701600.0, + "grad_norm": 2.2959742434183426, + "learning_rate": 1.1777300974287738e-06, + "loss": 0.6504, + "num_input_tokens_seen": 233105390, + "step": 7161 + }, + { + "epoch": 0.6458943950940164, + "flos": 18231586159680.0, + "grad_norm": 5.569416503751604, + "learning_rate": 1.1771975810107947e-06, + "loss": 0.7731, + "num_input_tokens_seen": 233132705, + "step": 7162 + }, + { + "epoch": 0.6459845786174866, + "flos": 16631027034240.0, + "grad_norm": 2.109520462106876, + "learning_rate": 1.1766651347988542e-06, + "loss": 0.8151, + "num_input_tokens_seen": 233159915, + "step": 7163 + }, + { + "epoch": 0.6460747621409568, + "flos": 17942508237120.0, + "grad_norm": 2.939325130882363, + "learning_rate": 1.1761327588383848e-06, + "loss": 0.8574, + "num_input_tokens_seen": 233184995, + "step": 7164 + }, + { + "epoch": 0.6461649456644271, + "flos": 25775706751200.0, + "grad_norm": 9.41068931073716, + "learning_rate": 1.1756004531748105e-06, + "loss": 0.7593, + "num_input_tokens_seen": 233214410, + "step": 7165 + }, + { + "epoch": 0.6462551291878974, + "flos": 26253301194720.0, + "grad_norm": 1.6776839260914074, + "learning_rate": 1.1750682178535521e-06, + "loss": 0.7813, + "num_input_tokens_seen": 233248625, + "step": 7166 + }, + { + "epoch": 0.6463453127113676, + "flos": 23043410880480.0, + "grad_norm": 1.9641789246301942, + "learning_rate": 1.1745360529200218e-06, + "loss": 0.6958, + "num_input_tokens_seen": 233279120, + "step": 7167 + }, + { + "epoch": 0.6464354962348379, + "flos": 24565158647520.0, + "grad_norm": 2.8675250860312547, + "learning_rate": 1.1740039584196261e-06, + "loss": 0.7701, + "num_input_tokens_seen": 233303565, + "step": 7168 + }, + { + "epoch": 0.6465256797583081, + "flos": 20273902346400.0, + "grad_norm": 1.7233831900290435, + "learning_rate": 1.1734719343977683e-06, + "loss": 0.776, + "num_input_tokens_seen": 233330910, + "step": 7169 + }, + { + "epoch": 0.6466158632817784, + "flos": 25517596972320.0, + "grad_norm": 4.021136705618629, + "learning_rate": 1.1729399808998416e-06, + "loss": 0.6892, + "num_input_tokens_seen": 233359505, + "step": 7170 + }, + { + "epoch": 0.6467060468052487, + "flos": 12952165660800.0, + "grad_norm": 2.754653201698391, + "learning_rate": 1.1724080979712368e-06, + "loss": 0.6374, + "num_input_tokens_seen": 233385285, + "step": 7171 + }, + { + "epoch": 0.6467962303287189, + "flos": 28799347899840.0, + "grad_norm": 2.108958676544265, + "learning_rate": 1.1718762856573365e-06, + "loss": 0.8141, + "num_input_tokens_seen": 233413785, + "step": 7172 + }, + { + "epoch": 0.6468864138521893, + "flos": 30178340854080.0, + "grad_norm": 1.6356527639969087, + "learning_rate": 1.1713445440035172e-06, + "loss": 0.7327, + "num_input_tokens_seen": 233445660, + "step": 7173 + }, + { + "epoch": 0.6469765973756595, + "flos": 27669209454240.0, + "grad_norm": 2.0265426081826017, + "learning_rate": 1.1708128730551506e-06, + "loss": 0.7833, + "num_input_tokens_seen": 233473485, + "step": 7174 + }, + { + "epoch": 0.6470667808991297, + "flos": 19363360074720.0, + "grad_norm": 1.7877367596451383, + "learning_rate": 1.1702812728576019e-06, + "loss": 0.7342, + "num_input_tokens_seen": 233502130, + "step": 7175 + }, + { + "epoch": 0.6471569644226, + "flos": 24099012490080.0, + "grad_norm": 1.970519648045751, + "learning_rate": 1.1697497434562303e-06, + "loss": 0.7751, + "num_input_tokens_seen": 233528990, + "step": 7176 + }, + { + "epoch": 0.6472471479460703, + "flos": 20383124302080.0, + "grad_norm": 2.0389633446916338, + "learning_rate": 1.1692182848963885e-06, + "loss": 0.7793, + "num_input_tokens_seen": 233554080, + "step": 7177 + }, + { + "epoch": 0.6473373314695405, + "flos": 22970360834880.0, + "grad_norm": 1.8856377515908536, + "learning_rate": 1.1686868972234227e-06, + "loss": 0.7083, + "num_input_tokens_seen": 233578720, + "step": 7178 + }, + { + "epoch": 0.6474275149930108, + "flos": 23698593935520.0, + "grad_norm": 3.228519021994205, + "learning_rate": 1.1681555804826755e-06, + "loss": 0.7558, + "num_input_tokens_seen": 233609355, + "step": 7179 + }, + { + "epoch": 0.647517698516481, + "flos": 28473949388160.0, + "grad_norm": 2.201552133962189, + "learning_rate": 1.1676243347194806e-06, + "loss": 0.6823, + "num_input_tokens_seen": 233642225, + "step": 7180 + }, + { + "epoch": 0.6476078820399513, + "flos": 24208197276000.0, + "grad_norm": 1.5725916933092825, + "learning_rate": 1.167093159979167e-06, + "loss": 0.7627, + "num_input_tokens_seen": 233672300, + "step": 7181 + }, + { + "epoch": 0.6476980655634216, + "flos": 26467619262720.0, + "grad_norm": 1.893189196752807, + "learning_rate": 1.1665620563070575e-06, + "loss": 0.7727, + "num_input_tokens_seen": 233703340, + "step": 7182 + }, + { + "epoch": 0.6477882490868918, + "flos": 17468444920800.0, + "grad_norm": 1.9606840418206029, + "learning_rate": 1.1660310237484691e-06, + "loss": 0.8574, + "num_input_tokens_seen": 233728540, + "step": 7183 + }, + { + "epoch": 0.647878432610362, + "flos": 25518303197760.0, + "grad_norm": 1.6334182126833394, + "learning_rate": 1.165500062348713e-06, + "loss": 0.6881, + "num_input_tokens_seen": 233758530, + "step": 7184 + }, + { + "epoch": 0.6479686161338324, + "flos": 31342346684640.0, + "grad_norm": 2.6422226265637034, + "learning_rate": 1.164969172153091e-06, + "loss": 0.7864, + "num_input_tokens_seen": 233786935, + "step": 7185 + }, + { + "epoch": 0.6480587996573026, + "flos": 30943303411200.0, + "grad_norm": 1.9079336017035606, + "learning_rate": 1.1644383532069055e-06, + "loss": 0.7627, + "num_input_tokens_seen": 233817010, + "step": 7186 + }, + { + "epoch": 0.6481489831807729, + "flos": 26026010880480.0, + "grad_norm": 1.7412461958143273, + "learning_rate": 1.1639076055554454e-06, + "loss": 0.7855, + "num_input_tokens_seen": 233847580, + "step": 7187 + }, + { + "epoch": 0.6482391667042431, + "flos": 24281916377280.0, + "grad_norm": 2.3343748651833858, + "learning_rate": 1.163376929244e-06, + "loss": 0.7356, + "num_input_tokens_seen": 233875325, + "step": 7188 + }, + { + "epoch": 0.6483293502277134, + "flos": 26357319384000.0, + "grad_norm": 1.6525991256000598, + "learning_rate": 1.1628463243178472e-06, + "loss": 0.7101, + "num_input_tokens_seen": 233905730, + "step": 7189 + }, + { + "epoch": 0.6484195337511837, + "flos": 27698430619200.0, + "grad_norm": 2.0105454662222453, + "learning_rate": 1.1623157908222623e-06, + "loss": 0.8026, + "num_input_tokens_seen": 233934055, + "step": 7190 + }, + { + "epoch": 0.6485097172746539, + "flos": 19691174620800.0, + "grad_norm": 1.6814326956372312, + "learning_rate": 1.1617853288025129e-06, + "loss": 0.7819, + "num_input_tokens_seen": 233962400, + "step": 7191 + }, + { + "epoch": 0.6485999007981241, + "flos": 30111052121280.0, + "grad_norm": 2.033292242583441, + "learning_rate": 1.1612549383038612e-06, + "loss": 0.7212, + "num_input_tokens_seen": 233992435, + "step": 7192 + }, + { + "epoch": 0.6486900843215945, + "flos": 29852719323840.0, + "grad_norm": 1.6873560950061401, + "learning_rate": 1.1607246193715629e-06, + "loss": 0.7231, + "num_input_tokens_seen": 234025940, + "step": 7193 + }, + { + "epoch": 0.6487802678450647, + "flos": 19176032986080.0, + "grad_norm": 3.784857298203095, + "learning_rate": 1.1601943720508684e-06, + "loss": 0.761, + "num_input_tokens_seen": 234053125, + "step": 7194 + }, + { + "epoch": 0.6488704513685349, + "flos": 21688955701440.0, + "grad_norm": 1.7040480465943233, + "learning_rate": 1.1596641963870202e-06, + "loss": 0.7958, + "num_input_tokens_seen": 234080515, + "step": 7195 + }, + { + "epoch": 0.6489606348920053, + "flos": 28213683763200.0, + "grad_norm": 1.610216531823565, + "learning_rate": 1.1591340924252561e-06, + "loss": 0.7603, + "num_input_tokens_seen": 234112220, + "step": 7196 + }, + { + "epoch": 0.6490508184154755, + "flos": 71599549494720.0, + "grad_norm": 0.7487482910900961, + "learning_rate": 1.158604060210808e-06, + "loss": 0.6231, + "num_input_tokens_seen": 234190890, + "step": 7197 + }, + { + "epoch": 0.6491410019389457, + "flos": 19654519503840.0, + "grad_norm": 4.909185418269233, + "learning_rate": 1.1580740997889008e-06, + "loss": 0.8084, + "num_input_tokens_seen": 234218515, + "step": 7198 + }, + { + "epoch": 0.649231185462416, + "flos": 23261743282560.0, + "grad_norm": 2.4483942959204708, + "learning_rate": 1.1575442112047544e-06, + "loss": 0.7492, + "num_input_tokens_seen": 234246765, + "step": 7199 + }, + { + "epoch": 0.6493213689858863, + "flos": 25993816134720.0, + "grad_norm": 2.24428298272909, + "learning_rate": 1.1570143945035797e-06, + "loss": 0.688, + "num_input_tokens_seen": 234275910, + "step": 7200 + }, + { + "epoch": 0.6494115525093566, + "flos": 25733364660960.0, + "grad_norm": 2.4038793602630584, + "learning_rate": 1.1564846497305864e-06, + "loss": 0.7017, + "num_input_tokens_seen": 234302635, + "step": 7201 + }, + { + "epoch": 0.6495017360328268, + "flos": 24390469277280.0, + "grad_norm": 2.2499576140755893, + "learning_rate": 1.1559549769309726e-06, + "loss": 0.7242, + "num_input_tokens_seen": 234330585, + "step": 7202 + }, + { + "epoch": 0.649591919556297, + "flos": 24574897124640.0, + "grad_norm": 1.6917703073758865, + "learning_rate": 1.1554253761499358e-06, + "loss": 0.7404, + "num_input_tokens_seen": 234361475, + "step": 7203 + }, + { + "epoch": 0.6496821030797674, + "flos": 23808261928320.0, + "grad_norm": 2.5186140416559293, + "learning_rate": 1.1548958474326617e-06, + "loss": 0.7824, + "num_input_tokens_seen": 234389005, + "step": 7204 + }, + { + "epoch": 0.6497722866032376, + "flos": 29016713888160.0, + "grad_norm": 1.7379096501947362, + "learning_rate": 1.154366390824334e-06, + "loss": 0.6789, + "num_input_tokens_seen": 234418915, + "step": 7205 + }, + { + "epoch": 0.6498624701267078, + "flos": 66284886329760.0, + "grad_norm": 0.9306731889083111, + "learning_rate": 1.1538370063701287e-06, + "loss": 0.5583, + "num_input_tokens_seen": 234509445, + "step": 7206 + }, + { + "epoch": 0.6499526536501781, + "flos": 23517994573440.0, + "grad_norm": 2.003724088066981, + "learning_rate": 1.1533076941152153e-06, + "loss": 0.703, + "num_input_tokens_seen": 234540130, + "step": 7207 + }, + { + "epoch": 0.6500428371736484, + "flos": 27159494604480.0, + "grad_norm": 2.7506225054763616, + "learning_rate": 1.1527784541047583e-06, + "loss": 0.6758, + "num_input_tokens_seen": 234568895, + "step": 7208 + }, + { + "epoch": 0.6501330206971186, + "flos": 20377920535680.0, + "grad_norm": 1.978207834446899, + "learning_rate": 1.1522492863839152e-06, + "loss": 0.7002, + "num_input_tokens_seen": 234594775, + "step": 7209 + }, + { + "epoch": 0.6502232042205889, + "flos": 20564913096480.0, + "grad_norm": 5.074377561085329, + "learning_rate": 1.1517201909978382e-06, + "loss": 0.7088, + "num_input_tokens_seen": 234618330, + "step": 7210 + }, + { + "epoch": 0.6503133877440591, + "flos": 58190269928160.0, + "grad_norm": 0.5998745061844933, + "learning_rate": 1.151191167991671e-06, + "loss": 0.5779, + "num_input_tokens_seen": 234715760, + "step": 7211 + }, + { + "epoch": 0.6504035712675295, + "flos": 27299982061920.0, + "grad_norm": 1.9752570504093006, + "learning_rate": 1.1506622174105536e-06, + "loss": 0.7532, + "num_input_tokens_seen": 234744505, + "step": 7212 + }, + { + "epoch": 0.6504937547909997, + "flos": 20346580694400.0, + "grad_norm": 1.9424338125492573, + "learning_rate": 1.1501333392996194e-06, + "loss": 0.7746, + "num_input_tokens_seen": 234770895, + "step": 7213 + }, + { + "epoch": 0.6505839383144699, + "flos": 59574615327840.0, + "grad_norm": 0.659842964188155, + "learning_rate": 1.1496045337039943e-06, + "loss": 0.5847, + "num_input_tokens_seen": 234858605, + "step": 7214 + }, + { + "epoch": 0.6506741218379402, + "flos": 21579993934080.0, + "grad_norm": 1.7594686886385442, + "learning_rate": 1.1490758006687995e-06, + "loss": 0.7041, + "num_input_tokens_seen": 234886320, + "step": 7215 + }, + { + "epoch": 0.6507643053614105, + "flos": 25556593784160.0, + "grad_norm": 2.2316491632817366, + "learning_rate": 1.1485471402391502e-06, + "loss": 0.7255, + "num_input_tokens_seen": 234915530, + "step": 7216 + }, + { + "epoch": 0.6508544888848807, + "flos": 19031531194560.0, + "grad_norm": 2.839077642859141, + "learning_rate": 1.1480185524601522e-06, + "loss": 0.7187, + "num_input_tokens_seen": 234943230, + "step": 7217 + }, + { + "epoch": 0.650944672408351, + "flos": 25228147352160.0, + "grad_norm": 3.1147062517659734, + "learning_rate": 1.1474900373769108e-06, + "loss": 0.7498, + "num_input_tokens_seen": 234972830, + "step": 7218 + }, + { + "epoch": 0.6510348559318213, + "flos": 69810659697120.0, + "grad_norm": 0.5682854096305262, + "learning_rate": 1.1469615950345184e-06, + "loss": 0.524, + "num_input_tokens_seen": 235064365, + "step": 7219 + }, + { + "epoch": 0.6511250394552915, + "flos": 67990653076800.0, + "grad_norm": 0.6962061507629755, + "learning_rate": 1.1464332254780678e-06, + "loss": 0.5615, + "num_input_tokens_seen": 235152815, + "step": 7220 + }, + { + "epoch": 0.6512152229787618, + "flos": 20274645741600.0, + "grad_norm": 1.643119006285649, + "learning_rate": 1.1459049287526404e-06, + "loss": 0.7315, + "num_input_tokens_seen": 235179210, + "step": 7221 + }, + { + "epoch": 0.651305406502232, + "flos": 15858147318240.0, + "grad_norm": 1.8912627980769279, + "learning_rate": 1.1453767049033137e-06, + "loss": 0.7044, + "num_input_tokens_seen": 235203825, + "step": 7222 + }, + { + "epoch": 0.6513955900257024, + "flos": 20926855215840.0, + "grad_norm": 1.8591028222658137, + "learning_rate": 1.1448485539751586e-06, + "loss": 0.7872, + "num_input_tokens_seen": 235231140, + "step": 7223 + }, + { + "epoch": 0.6514857735491726, + "flos": 15866101646880.0, + "grad_norm": 1.7514295869569627, + "learning_rate": 1.1443204760132408e-06, + "loss": 0.7181, + "num_input_tokens_seen": 235258065, + "step": 7224 + }, + { + "epoch": 0.6515759570726428, + "flos": 23004562747680.0, + "grad_norm": 2.8236243755602146, + "learning_rate": 1.1437924710626185e-06, + "loss": 0.7242, + "num_input_tokens_seen": 235288920, + "step": 7225 + }, + { + "epoch": 0.651666140596113, + "flos": 63929623500960.0, + "grad_norm": 0.6536728801425498, + "learning_rate": 1.1432645391683429e-06, + "loss": 0.5467, + "num_input_tokens_seen": 235380705, + "step": 7226 + }, + { + "epoch": 0.6517563241195834, + "flos": 23180738908320.0, + "grad_norm": 2.547878044461599, + "learning_rate": 1.1427366803754609e-06, + "loss": 0.7813, + "num_input_tokens_seen": 235409060, + "step": 7227 + }, + { + "epoch": 0.6518465076430536, + "flos": 27482105384160.0, + "grad_norm": 1.6990835552584322, + "learning_rate": 1.142208894729012e-06, + "loss": 0.76, + "num_input_tokens_seen": 235438090, + "step": 7228 + }, + { + "epoch": 0.6519366911665239, + "flos": 28325656281120.0, + "grad_norm": 1.7675978854428946, + "learning_rate": 1.1416811822740301e-06, + "loss": 0.7849, + "num_input_tokens_seen": 235467085, + "step": 7229 + }, + { + "epoch": 0.6520268746899941, + "flos": 23044414464000.0, + "grad_norm": 1.6210155789934524, + "learning_rate": 1.1411535430555428e-06, + "loss": 0.7926, + "num_input_tokens_seen": 235496935, + "step": 7230 + }, + { + "epoch": 0.6521170582134644, + "flos": 15137162320800.0, + "grad_norm": 2.051385118496385, + "learning_rate": 1.1406259771185705e-06, + "loss": 0.8009, + "num_input_tokens_seen": 235522400, + "step": 7231 + }, + { + "epoch": 0.6522072417369347, + "flos": 24753861017280.0, + "grad_norm": 1.6485597827265805, + "learning_rate": 1.1400984845081282e-06, + "loss": 0.7057, + "num_input_tokens_seen": 235554565, + "step": 7232 + }, + { + "epoch": 0.6522974252604049, + "flos": 25119631621920.0, + "grad_norm": 2.1919057892093656, + "learning_rate": 1.139571065269226e-06, + "loss": 0.8216, + "num_input_tokens_seen": 235582475, + "step": 7233 + }, + { + "epoch": 0.6523876087838751, + "flos": 43954691966880.0, + "grad_norm": 2.182120110832716, + "learning_rate": 1.139043719446863e-06, + "loss": 0.7895, + "num_input_tokens_seen": 235614850, + "step": 7234 + }, + { + "epoch": 0.6524777923073455, + "flos": 21221136904800.0, + "grad_norm": 1.7742827833505541, + "learning_rate": 1.1385164470860385e-06, + "loss": 0.8184, + "num_input_tokens_seen": 235645315, + "step": 7235 + }, + { + "epoch": 0.6525679758308157, + "flos": 18853645224960.0, + "grad_norm": 2.396690104886801, + "learning_rate": 1.1379892482317403e-06, + "loss": 0.746, + "num_input_tokens_seen": 235672735, + "step": 7236 + }, + { + "epoch": 0.652658159354286, + "flos": 20565098945280.0, + "grad_norm": 1.9307122231758682, + "learning_rate": 1.1374621229289524e-06, + "loss": 0.7292, + "num_input_tokens_seen": 235699410, + "step": 7237 + }, + { + "epoch": 0.6527483428777562, + "flos": 23294532744480.0, + "grad_norm": 1.8473032963042038, + "learning_rate": 1.1369350712226525e-06, + "loss": 0.73, + "num_input_tokens_seen": 235726440, + "step": 7238 + }, + { + "epoch": 0.6528385264012265, + "flos": 27706347778080.0, + "grad_norm": 1.9863541112829723, + "learning_rate": 1.136408093157811e-06, + "loss": 0.6945, + "num_input_tokens_seen": 235754865, + "step": 7239 + }, + { + "epoch": 0.6529287099246968, + "flos": 23443717925760.0, + "grad_norm": 2.9344596549955964, + "learning_rate": 1.1358811887793935e-06, + "loss": 0.7174, + "num_input_tokens_seen": 235782995, + "step": 7240 + }, + { + "epoch": 0.653018893448167, + "flos": 23335090686240.0, + "grad_norm": 1.9352506686151452, + "learning_rate": 1.135354358132356e-06, + "loss": 0.7857, + "num_input_tokens_seen": 235811555, + "step": 7241 + }, + { + "epoch": 0.6531090769716373, + "flos": 23625989927040.0, + "grad_norm": 3.0077950613333755, + "learning_rate": 1.1348276012616542e-06, + "loss": 0.7843, + "num_input_tokens_seen": 235840425, + "step": 7242 + }, + { + "epoch": 0.6531992604951076, + "flos": 19981702164000.0, + "grad_norm": 2.2323750969098413, + "learning_rate": 1.134300918212231e-06, + "loss": 0.7036, + "num_input_tokens_seen": 235867410, + "step": 7243 + }, + { + "epoch": 0.6532894440185778, + "flos": 35539100255040.0, + "grad_norm": 2.160325552460727, + "learning_rate": 1.133774309029027e-06, + "loss": 0.7561, + "num_input_tokens_seen": 235895515, + "step": 7244 + }, + { + "epoch": 0.653379627542048, + "flos": 23808224758560.0, + "grad_norm": 2.0078342302979597, + "learning_rate": 1.133247773756975e-06, + "loss": 0.7883, + "num_input_tokens_seen": 235923865, + "step": 7245 + }, + { + "epoch": 0.6534698110655184, + "flos": 69272392738080.0, + "grad_norm": 0.6559710590997399, + "learning_rate": 1.1327213124410024e-06, + "loss": 0.6168, + "num_input_tokens_seen": 236009930, + "step": 7246 + }, + { + "epoch": 0.6535599945889886, + "flos": 24172768761120.0, + "grad_norm": 1.8175531519656687, + "learning_rate": 1.1321949251260292e-06, + "loss": 0.7874, + "num_input_tokens_seen": 236039670, + "step": 7247 + }, + { + "epoch": 0.6536501781124588, + "flos": 27888396760800.0, + "grad_norm": 2.2873482909210594, + "learning_rate": 1.1316686118569712e-06, + "loss": 0.6275, + "num_input_tokens_seen": 236068745, + "step": 7248 + }, + { + "epoch": 0.6537403616359291, + "flos": 22605854002080.0, + "grad_norm": 1.7836744974432948, + "learning_rate": 1.1311423726787335e-06, + "loss": 0.657, + "num_input_tokens_seen": 236096510, + "step": 7249 + }, + { + "epoch": 0.6538305451593994, + "flos": 24420433837440.0, + "grad_norm": 1.7958981056067869, + "learning_rate": 1.130616207636221e-06, + "loss": 0.7174, + "num_input_tokens_seen": 236125355, + "step": 7250 + }, + { + "epoch": 0.6539207286828697, + "flos": 20492346257760.0, + "grad_norm": 2.1726567899009916, + "learning_rate": 1.1300901167743263e-06, + "loss": 0.7482, + "num_input_tokens_seen": 236152170, + "step": 7251 + }, + { + "epoch": 0.6540109122063399, + "flos": 23660675046720.0, + "grad_norm": 1.5496706120377965, + "learning_rate": 1.12956410013794e-06, + "loss": 0.7407, + "num_input_tokens_seen": 236183940, + "step": 7252 + }, + { + "epoch": 0.6541010957298101, + "flos": 21403297396800.0, + "grad_norm": 2.490347155840728, + "learning_rate": 1.1290381577719436e-06, + "loss": 0.7316, + "num_input_tokens_seen": 236208660, + "step": 7253 + }, + { + "epoch": 0.6541912792532805, + "flos": 22460051268960.0, + "grad_norm": 2.0188418010235014, + "learning_rate": 1.1285122897212143e-06, + "loss": 0.6515, + "num_input_tokens_seen": 236238365, + "step": 7254 + }, + { + "epoch": 0.6542814627767507, + "flos": 51930082013760.0, + "grad_norm": 0.7479424465982054, + "learning_rate": 1.1279864960306228e-06, + "loss": 0.6171, + "num_input_tokens_seen": 236327280, + "step": 7255 + }, + { + "epoch": 0.6543716463002209, + "flos": 22388190655680.0, + "grad_norm": 1.7995351588531434, + "learning_rate": 1.1274607767450297e-06, + "loss": 0.8227, + "num_input_tokens_seen": 236356645, + "step": 7256 + }, + { + "epoch": 0.6544618298236912, + "flos": 22059298186560.0, + "grad_norm": 1.6245318985681174, + "learning_rate": 1.126935131909296e-06, + "loss": 0.7215, + "num_input_tokens_seen": 236386625, + "step": 7257 + }, + { + "epoch": 0.6545520133471615, + "flos": 20890274438400.0, + "grad_norm": 1.5459949803042623, + "learning_rate": 1.1264095615682693e-06, + "loss": 0.6953, + "num_input_tokens_seen": 236416080, + "step": 7258 + }, + { + "epoch": 0.6546421968706317, + "flos": 20965889197440.0, + "grad_norm": 1.784786194356977, + "learning_rate": 1.1258840657667973e-06, + "loss": 0.6618, + "num_input_tokens_seen": 236444830, + "step": 7259 + }, + { + "epoch": 0.654732380394102, + "flos": 22131865025280.0, + "grad_norm": 1.6372997918659289, + "learning_rate": 1.125358644549716e-06, + "loss": 0.6726, + "num_input_tokens_seen": 236471920, + "step": 7260 + }, + { + "epoch": 0.6548225639175722, + "flos": 59438891333280.0, + "grad_norm": 2.191928631268079, + "learning_rate": 1.1248332979618578e-06, + "loss": 0.6059, + "num_input_tokens_seen": 236509525, + "step": 7261 + }, + { + "epoch": 0.6549127474410426, + "flos": 32551036300320.0, + "grad_norm": 1.5785780152600792, + "learning_rate": 1.1243080260480482e-06, + "loss": 0.6365, + "num_input_tokens_seen": 236544895, + "step": 7262 + }, + { + "epoch": 0.6550029309645128, + "flos": 20528852695680.0, + "grad_norm": 2.8041758801184677, + "learning_rate": 1.1237828288531063e-06, + "loss": 0.8005, + "num_input_tokens_seen": 236569855, + "step": 7263 + }, + { + "epoch": 0.655093114487983, + "flos": 21622559042880.0, + "grad_norm": 1.9296280083385438, + "learning_rate": 1.1232577064218449e-06, + "loss": 0.6597, + "num_input_tokens_seen": 236597700, + "step": 7264 + }, + { + "epoch": 0.6551832980114533, + "flos": 25119371433600.0, + "grad_norm": 2.1616058176731796, + "learning_rate": 1.1227326587990711e-06, + "loss": 0.8216, + "num_input_tokens_seen": 236627790, + "step": 7265 + }, + { + "epoch": 0.6552734815349236, + "flos": 25629197792640.0, + "grad_norm": 2.017678922072207, + "learning_rate": 1.1222076860295832e-06, + "loss": 0.7288, + "num_input_tokens_seen": 236657970, + "step": 7266 + }, + { + "epoch": 0.6553636650583938, + "flos": 22789389775200.0, + "grad_norm": 2.260477993978483, + "learning_rate": 1.1216827881581756e-06, + "loss": 0.7643, + "num_input_tokens_seen": 236686010, + "step": 7267 + }, + { + "epoch": 0.6554538485818641, + "flos": 18592152997920.0, + "grad_norm": 2.0857932508790507, + "learning_rate": 1.1211579652296355e-06, + "loss": 0.7403, + "num_input_tokens_seen": 236711745, + "step": 7268 + }, + { + "epoch": 0.6555440321053344, + "flos": 26394866575200.0, + "grad_norm": 1.9998077061717254, + "learning_rate": 1.1206332172887438e-06, + "loss": 0.8295, + "num_input_tokens_seen": 236740275, + "step": 7269 + }, + { + "epoch": 0.6556342156288046, + "flos": 39292201106400.0, + "grad_norm": 2.1222836579866686, + "learning_rate": 1.1201085443802756e-06, + "loss": 0.7608, + "num_input_tokens_seen": 236770750, + "step": 7270 + }, + { + "epoch": 0.6557243991522749, + "flos": 20602980664320.0, + "grad_norm": 1.744042405527234, + "learning_rate": 1.1195839465489964e-06, + "loss": 0.7126, + "num_input_tokens_seen": 236799385, + "step": 7271 + }, + { + "epoch": 0.6558145826757451, + "flos": 26503902682080.0, + "grad_norm": 2.257031473943449, + "learning_rate": 1.1190594238396708e-06, + "loss": 0.7906, + "num_input_tokens_seen": 236825620, + "step": 7272 + }, + { + "epoch": 0.6559047661992154, + "flos": 29534160048000.0, + "grad_norm": 1.500697053436988, + "learning_rate": 1.1185349762970515e-06, + "loss": 0.7177, + "num_input_tokens_seen": 236860040, + "step": 7273 + }, + { + "epoch": 0.6559949497226857, + "flos": 44756755678080.0, + "grad_norm": 1.8119346789360815, + "learning_rate": 1.1180106039658896e-06, + "loss": 0.7473, + "num_input_tokens_seen": 236895255, + "step": 7274 + }, + { + "epoch": 0.6560851332461559, + "flos": 61896971451840.0, + "grad_norm": 0.5836713328235811, + "learning_rate": 1.117486306890925e-06, + "loss": 0.558, + "num_input_tokens_seen": 236993225, + "step": 7275 + }, + { + "epoch": 0.6561753167696261, + "flos": 27632926034880.0, + "grad_norm": 2.844962060873153, + "learning_rate": 1.116962085116896e-06, + "loss": 0.696, + "num_input_tokens_seen": 237022650, + "step": 7276 + }, + { + "epoch": 0.6562655002930965, + "flos": 68432001270720.0, + "grad_norm": 0.6639760129370397, + "learning_rate": 1.1164379386885302e-06, + "loss": 0.5166, + "num_input_tokens_seen": 237114845, + "step": 7277 + }, + { + "epoch": 0.6563556838165667, + "flos": 17687743736640.0, + "grad_norm": 2.030412706788375, + "learning_rate": 1.1159138676505516e-06, + "loss": 0.7895, + "num_input_tokens_seen": 237141440, + "step": 7278 + }, + { + "epoch": 0.656445867340037, + "flos": 29052328251840.0, + "grad_norm": 1.6622196792955943, + "learning_rate": 1.1153898720476761e-06, + "loss": 0.8034, + "num_input_tokens_seen": 237170710, + "step": 7279 + }, + { + "epoch": 0.6565360508635072, + "flos": 18233295968640.0, + "grad_norm": 3.0463065375151945, + "learning_rate": 1.114865951924615e-06, + "loss": 0.7335, + "num_input_tokens_seen": 237194550, + "step": 7280 + }, + { + "epoch": 0.6566262343869775, + "flos": 19691025941760.0, + "grad_norm": 1.8707266748124534, + "learning_rate": 1.1143421073260721e-06, + "loss": 0.6883, + "num_input_tokens_seen": 237223095, + "step": 7281 + }, + { + "epoch": 0.6567164179104478, + "flos": 17431603955040.0, + "grad_norm": 2.229329100302886, + "learning_rate": 1.1138183382967432e-06, + "loss": 0.6836, + "num_input_tokens_seen": 237248795, + "step": 7282 + }, + { + "epoch": 0.656806601433918, + "flos": 20849902345440.0, + "grad_norm": 2.0010092096378926, + "learning_rate": 1.11329464488132e-06, + "loss": 0.6526, + "num_input_tokens_seen": 237275740, + "step": 7283 + }, + { + "epoch": 0.6568967849573882, + "flos": 33058483794720.0, + "grad_norm": 1.9644446147704782, + "learning_rate": 1.112771027124487e-06, + "loss": 0.7192, + "num_input_tokens_seen": 237307355, + "step": 7284 + }, + { + "epoch": 0.6569869684808586, + "flos": 24135853455840.0, + "grad_norm": 2.936958100676929, + "learning_rate": 1.112247485070922e-06, + "loss": 0.7543, + "num_input_tokens_seen": 237335955, + "step": 7285 + }, + { + "epoch": 0.6570771520043288, + "flos": 19101867847680.0, + "grad_norm": 1.7796827415020888, + "learning_rate": 1.1117240187652968e-06, + "loss": 0.7512, + "num_input_tokens_seen": 237363330, + "step": 7286 + }, + { + "epoch": 0.657167335527799, + "flos": 17614210484160.0, + "grad_norm": 2.2668726149662954, + "learning_rate": 1.1112006282522767e-06, + "loss": 0.8245, + "num_input_tokens_seen": 237390325, + "step": 7287 + }, + { + "epoch": 0.6572575190512693, + "flos": 20491156825440.0, + "grad_norm": 2.0178634870118386, + "learning_rate": 1.1106773135765183e-06, + "loss": 0.7279, + "num_input_tokens_seen": 237418555, + "step": 7288 + }, + { + "epoch": 0.6573477025747396, + "flos": 25558526611680.0, + "grad_norm": 1.9962613806164264, + "learning_rate": 1.110154074782677e-06, + "loss": 0.7206, + "num_input_tokens_seen": 237450550, + "step": 7289 + }, + { + "epoch": 0.6574378860982099, + "flos": 22314806082240.0, + "grad_norm": 2.840311507750439, + "learning_rate": 1.1096309119153948e-06, + "loss": 0.7415, + "num_input_tokens_seen": 237478155, + "step": 7290 + }, + { + "epoch": 0.6575280696216801, + "flos": 68918107589280.0, + "grad_norm": 0.6253976272295841, + "learning_rate": 1.1091078250193145e-06, + "loss": 0.6186, + "num_input_tokens_seen": 237570585, + "step": 7291 + }, + { + "epoch": 0.6576182531451504, + "flos": 18087902102880.0, + "grad_norm": 2.6686908867250176, + "learning_rate": 1.108584814139066e-06, + "loss": 0.7483, + "num_input_tokens_seen": 237596580, + "step": 7292 + }, + { + "epoch": 0.6577084366686207, + "flos": 33972780212160.0, + "grad_norm": 2.169924072484261, + "learning_rate": 1.108061879319276e-06, + "loss": 0.7439, + "num_input_tokens_seen": 237624510, + "step": 7293 + }, + { + "epoch": 0.6577986201920909, + "flos": 21726800250720.0, + "grad_norm": 1.9127755334376664, + "learning_rate": 1.1075390206045648e-06, + "loss": 0.803, + "num_input_tokens_seen": 237652345, + "step": 7294 + }, + { + "epoch": 0.6578888037155611, + "flos": 25811767152000.0, + "grad_norm": 2.1205200768374612, + "learning_rate": 1.1070162380395454e-06, + "loss": 0.7865, + "num_input_tokens_seen": 237678590, + "step": 7295 + }, + { + "epoch": 0.6579789872390315, + "flos": 20965740518400.0, + "grad_norm": 1.901981264393143, + "learning_rate": 1.1064935316688253e-06, + "loss": 0.6881, + "num_input_tokens_seen": 237706130, + "step": 7296 + }, + { + "epoch": 0.6580691707625017, + "flos": 23656809391680.0, + "grad_norm": 1.986258266424571, + "learning_rate": 1.105970901537002e-06, + "loss": 0.8296, + "num_input_tokens_seen": 237731880, + "step": 7297 + }, + { + "epoch": 0.6581593542859719, + "flos": 20199402680160.0, + "grad_norm": 1.865404243123965, + "learning_rate": 1.1054483476886727e-06, + "loss": 0.7893, + "num_input_tokens_seen": 237760970, + "step": 7298 + }, + { + "epoch": 0.6582495378094422, + "flos": 23298138211200.0, + "grad_norm": 1.84394945261971, + "learning_rate": 1.1049258701684222e-06, + "loss": 0.6653, + "num_input_tokens_seen": 237790465, + "step": 7299 + }, + { + "epoch": 0.6583397213329125, + "flos": 18707247775680.0, + "grad_norm": 1.932479753277165, + "learning_rate": 1.1044034690208315e-06, + "loss": 0.695, + "num_input_tokens_seen": 237818340, + "step": 7300 + }, + { + "epoch": 0.6584299048563828, + "flos": 17978791656480.0, + "grad_norm": 4.157415470404021, + "learning_rate": 1.1038811442904755e-06, + "loss": 0.7779, + "num_input_tokens_seen": 237846515, + "step": 7301 + }, + { + "epoch": 0.658520088379853, + "flos": 21185299522560.0, + "grad_norm": 2.096133227977156, + "learning_rate": 1.103358896021921e-06, + "loss": 0.6843, + "num_input_tokens_seen": 237873580, + "step": 7302 + }, + { + "epoch": 0.6586102719033232, + "flos": 23298584248320.0, + "grad_norm": 2.257227415946343, + "learning_rate": 1.1028367242597298e-06, + "loss": 0.7222, + "num_input_tokens_seen": 237903275, + "step": 7303 + }, + { + "epoch": 0.6587004554267936, + "flos": 18561816740160.0, + "grad_norm": 2.254567173359595, + "learning_rate": 1.102314629048457e-06, + "loss": 0.7862, + "num_input_tokens_seen": 237928605, + "step": 7304 + }, + { + "epoch": 0.6587906389502638, + "flos": 21877472222400.0, + "grad_norm": 1.7006772851950178, + "learning_rate": 1.1017926104326484e-06, + "loss": 0.797, + "num_input_tokens_seen": 237958030, + "step": 7305 + }, + { + "epoch": 0.658880822473734, + "flos": 16157484094560.0, + "grad_norm": 2.00022290894936, + "learning_rate": 1.1012706684568483e-06, + "loss": 0.799, + "num_input_tokens_seen": 237984890, + "step": 7306 + }, + { + "epoch": 0.6589710059972043, + "flos": 19727272191360.0, + "grad_norm": 1.719885572353151, + "learning_rate": 1.1007488031655894e-06, + "loss": 0.7605, + "num_input_tokens_seen": 238014630, + "step": 7307 + }, + { + "epoch": 0.6590611895206746, + "flos": 26394755065920.0, + "grad_norm": 2.501548764192897, + "learning_rate": 1.1002270146034013e-06, + "loss": 0.7008, + "num_input_tokens_seen": 238041900, + "step": 7308 + }, + { + "epoch": 0.6591513730441448, + "flos": 28651426490400.0, + "grad_norm": 1.3767386627921914, + "learning_rate": 1.0997053028148052e-06, + "loss": 0.7733, + "num_input_tokens_seen": 238072700, + "step": 7309 + }, + { + "epoch": 0.6592415565676151, + "flos": 26467136055840.0, + "grad_norm": 2.2959976077515902, + "learning_rate": 1.0991836678443173e-06, + "loss": 0.7519, + "num_input_tokens_seen": 238104415, + "step": 7310 + }, + { + "epoch": 0.6593317400910853, + "flos": 24567872040000.0, + "grad_norm": 2.0429694502119977, + "learning_rate": 1.0986621097364465e-06, + "loss": 0.6924, + "num_input_tokens_seen": 238132585, + "step": 7311 + }, + { + "epoch": 0.6594219236145556, + "flos": 21075334171680.0, + "grad_norm": 2.1753647027138308, + "learning_rate": 1.0981406285356932e-06, + "loss": 0.6597, + "num_input_tokens_seen": 238158150, + "step": 7312 + }, + { + "epoch": 0.6595121071380259, + "flos": 23116609605120.0, + "grad_norm": 2.1063502347283873, + "learning_rate": 1.0976192242865554e-06, + "loss": 0.6706, + "num_input_tokens_seen": 238189160, + "step": 7313 + }, + { + "epoch": 0.6596022906614961, + "flos": 18962123785440.0, + "grad_norm": 2.0407969380450335, + "learning_rate": 1.0970978970335202e-06, + "loss": 0.8121, + "num_input_tokens_seen": 238216760, + "step": 7314 + }, + { + "epoch": 0.6596924741849665, + "flos": 27739360258560.0, + "grad_norm": 1.8221222227203229, + "learning_rate": 1.0965766468210714e-06, + "loss": 0.7668, + "num_input_tokens_seen": 238246395, + "step": 7315 + }, + { + "epoch": 0.6597826577084367, + "flos": 22563437572320.0, + "grad_norm": 2.4355020404228385, + "learning_rate": 1.0960554736936843e-06, + "loss": 0.7547, + "num_input_tokens_seen": 238272540, + "step": 7316 + }, + { + "epoch": 0.6598728412319069, + "flos": 22604961927840.0, + "grad_norm": 2.0786940863837673, + "learning_rate": 1.0955343776958283e-06, + "loss": 0.6547, + "num_input_tokens_seen": 238298570, + "step": 7317 + }, + { + "epoch": 0.6599630247553772, + "flos": 22059632714400.0, + "grad_norm": 2.0677068471703044, + "learning_rate": 1.0950133588719665e-06, + "loss": 0.7603, + "num_input_tokens_seen": 238327600, + "step": 7318 + }, + { + "epoch": 0.6600532082788475, + "flos": 21544379570400.0, + "grad_norm": 2.4269880895522773, + "learning_rate": 1.0944924172665551e-06, + "loss": 0.7709, + "num_input_tokens_seen": 238354130, + "step": 7319 + }, + { + "epoch": 0.6601433918023177, + "flos": 19030899308640.0, + "grad_norm": 1.9352215321216977, + "learning_rate": 1.0939715529240437e-06, + "loss": 0.7854, + "num_input_tokens_seen": 238380060, + "step": 7320 + }, + { + "epoch": 0.660233575325788, + "flos": 27815309545440.0, + "grad_norm": 5.146743796829708, + "learning_rate": 1.0934507658888755e-06, + "loss": 0.777, + "num_input_tokens_seen": 238409565, + "step": 7321 + }, + { + "epoch": 0.6603237588492582, + "flos": 25338001193760.0, + "grad_norm": 1.6990431420797034, + "learning_rate": 1.092930056205486e-06, + "loss": 0.7178, + "num_input_tokens_seen": 238441710, + "step": 7322 + }, + { + "epoch": 0.6604139423727285, + "flos": 15283596939840.0, + "grad_norm": 2.624617594158045, + "learning_rate": 1.092409423918306e-06, + "loss": 0.7135, + "num_input_tokens_seen": 238468450, + "step": 7323 + }, + { + "epoch": 0.6605041258961988, + "flos": 23551081393440.0, + "grad_norm": 1.62660959195759, + "learning_rate": 1.0918888690717581e-06, + "loss": 0.6689, + "num_input_tokens_seen": 238500195, + "step": 7324 + }, + { + "epoch": 0.660594309419669, + "flos": 25658047260000.0, + "grad_norm": 1.7973522050179973, + "learning_rate": 1.091368391710259e-06, + "loss": 0.8119, + "num_input_tokens_seen": 238527505, + "step": 7325 + }, + { + "epoch": 0.6606844929431392, + "flos": 14153049626880.0, + "grad_norm": 1.7443170234014378, + "learning_rate": 1.0908479918782198e-06, + "loss": 0.6919, + "num_input_tokens_seen": 238552985, + "step": 7326 + }, + { + "epoch": 0.6607746764666096, + "flos": 22896418715040.0, + "grad_norm": 1.7395910236241956, + "learning_rate": 1.0903276696200413e-06, + "loss": 0.7673, + "num_input_tokens_seen": 238582255, + "step": 7327 + }, + { + "epoch": 0.6608648599900798, + "flos": 26031177477120.0, + "grad_norm": 2.180734581155177, + "learning_rate": 1.0898074249801234e-06, + "loss": 0.7515, + "num_input_tokens_seen": 238611775, + "step": 7328 + }, + { + "epoch": 0.6609550435135501, + "flos": 25519975836960.0, + "grad_norm": 1.5504317574440247, + "learning_rate": 1.0892872580028533e-06, + "loss": 0.7635, + "num_input_tokens_seen": 238643545, + "step": 7329 + }, + { + "epoch": 0.6610452270370203, + "flos": 25076100099360.0, + "grad_norm": 1.7678169215637445, + "learning_rate": 1.0887671687326178e-06, + "loss": 0.7793, + "num_input_tokens_seen": 238673360, + "step": 7330 + }, + { + "epoch": 0.6611354105604906, + "flos": 16303621355520.0, + "grad_norm": 2.091990286525546, + "learning_rate": 1.0882471572137908e-06, + "loss": 0.7146, + "num_input_tokens_seen": 238699885, + "step": 7331 + }, + { + "epoch": 0.6612255940839609, + "flos": 20674729768320.0, + "grad_norm": 1.886286399697392, + "learning_rate": 1.087727223490744e-06, + "loss": 0.7906, + "num_input_tokens_seen": 238729125, + "step": 7332 + }, + { + "epoch": 0.6613157776074311, + "flos": 20235202892640.0, + "grad_norm": 1.8751364976153952, + "learning_rate": 1.0872073676078405e-06, + "loss": 0.8064, + "num_input_tokens_seen": 238757115, + "step": 7333 + }, + { + "epoch": 0.6614059611309013, + "flos": 26134117743360.0, + "grad_norm": 5.00585994653479, + "learning_rate": 1.0866875896094375e-06, + "loss": 0.7004, + "num_input_tokens_seen": 238785055, + "step": 7334 + }, + { + "epoch": 0.6614961446543717, + "flos": 22896976261440.0, + "grad_norm": 1.8110982143915446, + "learning_rate": 1.0861678895398854e-06, + "loss": 0.7653, + "num_input_tokens_seen": 238813795, + "step": 7335 + }, + { + "epoch": 0.6615863281778419, + "flos": 16809173192160.0, + "grad_norm": 2.2604932874593824, + "learning_rate": 1.0856482674435286e-06, + "loss": 0.7111, + "num_input_tokens_seen": 238840830, + "step": 7336 + }, + { + "epoch": 0.6616765117013121, + "flos": 71324484571680.0, + "grad_norm": 0.6667063199608425, + "learning_rate": 1.0851287233647024e-06, + "loss": 0.5741, + "num_input_tokens_seen": 238926505, + "step": 7337 + }, + { + "epoch": 0.6617666952247825, + "flos": 21367311335520.0, + "grad_norm": 3.106660664543062, + "learning_rate": 1.084609257347738e-06, + "loss": 0.802, + "num_input_tokens_seen": 238953080, + "step": 7338 + }, + { + "epoch": 0.6618568787482527, + "flos": 25301197397760.0, + "grad_norm": 2.981814099571242, + "learning_rate": 1.0840898694369594e-06, + "loss": 0.7226, + "num_input_tokens_seen": 238984090, + "step": 7339 + }, + { + "epoch": 0.661947062271723, + "flos": 27123954580320.0, + "grad_norm": 1.9427402142592554, + "learning_rate": 1.083570559676683e-06, + "loss": 0.7369, + "num_input_tokens_seen": 239014865, + "step": 7340 + }, + { + "epoch": 0.6620372457951932, + "flos": 24974014737600.0, + "grad_norm": 1.7920837212658003, + "learning_rate": 1.08305132811122e-06, + "loss": 0.7236, + "num_input_tokens_seen": 239043425, + "step": 7341 + }, + { + "epoch": 0.6621274293186635, + "flos": 35466421907040.0, + "grad_norm": 1.7399334888242788, + "learning_rate": 1.0825321747848735e-06, + "loss": 0.749, + "num_input_tokens_seen": 239073805, + "step": 7342 + }, + { + "epoch": 0.6622176128421338, + "flos": 62742163551840.0, + "grad_norm": 0.7091818290222158, + "learning_rate": 1.0820130997419417e-06, + "loss": 0.6439, + "num_input_tokens_seen": 239155635, + "step": 7343 + }, + { + "epoch": 0.662307796365604, + "flos": 21396829858560.0, + "grad_norm": 2.174147151490494, + "learning_rate": 1.0814941030267123e-06, + "loss": 0.7372, + "num_input_tokens_seen": 239182350, + "step": 7344 + }, + { + "epoch": 0.6623979798890742, + "flos": 66033206919360.0, + "grad_norm": 0.5822494661010155, + "learning_rate": 1.080975184683472e-06, + "loss": 0.5622, + "num_input_tokens_seen": 239282460, + "step": 7345 + }, + { + "epoch": 0.6624881634125446, + "flos": 24496531803360.0, + "grad_norm": 4.86148879022119, + "learning_rate": 1.0804563447564948e-06, + "loss": 0.8058, + "num_input_tokens_seen": 239312585, + "step": 7346 + }, + { + "epoch": 0.6625783469360148, + "flos": 46031470254720.0, + "grad_norm": 2.4106714820089326, + "learning_rate": 1.0799375832900545e-06, + "loss": 0.6935, + "num_input_tokens_seen": 239341295, + "step": 7347 + }, + { + "epoch": 0.662668530459485, + "flos": 21797025394560.0, + "grad_norm": 2.0760674867557194, + "learning_rate": 1.0794189003284118e-06, + "loss": 0.7833, + "num_input_tokens_seen": 239367940, + "step": 7348 + }, + { + "epoch": 0.6627587139829553, + "flos": 31673915376480.0, + "grad_norm": 1.8914638581419165, + "learning_rate": 1.0789002959158242e-06, + "loss": 0.7597, + "num_input_tokens_seen": 239396340, + "step": 7349 + }, + { + "epoch": 0.6628488975064256, + "flos": 13970926304640.0, + "grad_norm": 6.625945689523691, + "learning_rate": 1.0783817700965428e-06, + "loss": 0.7001, + "num_input_tokens_seen": 239422925, + "step": 7350 + }, + { + "epoch": 0.6629390810298958, + "flos": 26791456644480.0, + "grad_norm": 1.7145515488389664, + "learning_rate": 1.0778633229148102e-06, + "loss": 0.7959, + "num_input_tokens_seen": 239453815, + "step": 7351 + }, + { + "epoch": 0.6630292645533661, + "flos": 28143309940320.0, + "grad_norm": 2.0255350559866905, + "learning_rate": 1.0773449544148645e-06, + "loss": 0.7508, + "num_input_tokens_seen": 239481900, + "step": 7352 + }, + { + "epoch": 0.6631194480768363, + "flos": 25338447230880.0, + "grad_norm": 1.821910042532074, + "learning_rate": 1.076826664640934e-06, + "loss": 0.7873, + "num_input_tokens_seen": 239512325, + "step": 7353 + }, + { + "epoch": 0.6632096316003067, + "flos": 50603881585920.0, + "grad_norm": 0.7443938137466656, + "learning_rate": 1.0763084536372424e-06, + "loss": 0.6033, + "num_input_tokens_seen": 239590555, + "step": 7354 + }, + { + "epoch": 0.6632998151237769, + "flos": 22236998307360.0, + "grad_norm": 1.960416200200619, + "learning_rate": 1.0757903214480068e-06, + "loss": 0.7555, + "num_input_tokens_seen": 239616765, + "step": 7355 + }, + { + "epoch": 0.6633899986472471, + "flos": 27013877720160.0, + "grad_norm": 1.796014198111166, + "learning_rate": 1.0752722681174376e-06, + "loss": 0.693, + "num_input_tokens_seen": 239645300, + "step": 7356 + }, + { + "epoch": 0.6634801821707174, + "flos": 27159866302080.0, + "grad_norm": 21.325399751482898, + "learning_rate": 1.074754293689737e-06, + "loss": 0.6698, + "num_input_tokens_seen": 239675420, + "step": 7357 + }, + { + "epoch": 0.6635703656941877, + "flos": 26248877993280.0, + "grad_norm": 2.2710456982159144, + "learning_rate": 1.0742363982091023e-06, + "loss": 0.7517, + "num_input_tokens_seen": 239704220, + "step": 7358 + }, + { + "epoch": 0.6636605492176579, + "flos": 20018877657600.0, + "grad_norm": 1.936266044712759, + "learning_rate": 1.0737185817197215e-06, + "loss": 0.7032, + "num_input_tokens_seen": 239730770, + "step": 7359 + }, + { + "epoch": 0.6637507327411282, + "flos": 21294075441120.0, + "grad_norm": 2.3779156506259214, + "learning_rate": 1.0732008442657803e-06, + "loss": 0.7093, + "num_input_tokens_seen": 239758355, + "step": 7360 + }, + { + "epoch": 0.6638409162645985, + "flos": 26322411245760.0, + "grad_norm": 1.9068103029371666, + "learning_rate": 1.0726831858914516e-06, + "loss": 0.7285, + "num_input_tokens_seen": 239789070, + "step": 7361 + }, + { + "epoch": 0.6639310997880687, + "flos": 21840594086880.0, + "grad_norm": 2.330310077745269, + "learning_rate": 1.0721656066409084e-06, + "loss": 0.7654, + "num_input_tokens_seen": 239817600, + "step": 7362 + }, + { + "epoch": 0.664021283311539, + "flos": 21439878174240.0, + "grad_norm": 2.648695221460972, + "learning_rate": 1.0716481065583108e-06, + "loss": 0.7576, + "num_input_tokens_seen": 239844950, + "step": 7363 + }, + { + "epoch": 0.6641114668350092, + "flos": 65536575825120.0, + "grad_norm": 0.6355959113112559, + "learning_rate": 1.071130685687816e-06, + "loss": 0.5735, + "num_input_tokens_seen": 239937100, + "step": 7364 + }, + { + "epoch": 0.6642016503584796, + "flos": 24936170188320.0, + "grad_norm": 1.9333909670142762, + "learning_rate": 1.0706133440735723e-06, + "loss": 0.6826, + "num_input_tokens_seen": 239968435, + "step": 7365 + }, + { + "epoch": 0.6642918338819498, + "flos": 27013952059680.0, + "grad_norm": 2.327591188408404, + "learning_rate": 1.070096081759723e-06, + "loss": 0.6938, + "num_input_tokens_seen": 239997130, + "step": 7366 + }, + { + "epoch": 0.66438201740542, + "flos": 19617864386880.0, + "grad_norm": 2.0842148886614162, + "learning_rate": 1.069578898790404e-06, + "loss": 0.8476, + "num_input_tokens_seen": 240024235, + "step": 7367 + }, + { + "epoch": 0.6644722009288903, + "flos": 25698976899360.0, + "grad_norm": 2.6106619944655653, + "learning_rate": 1.0690617952097424e-06, + "loss": 0.7289, + "num_input_tokens_seen": 240052620, + "step": 7368 + }, + { + "epoch": 0.6645623844523606, + "flos": 22383990472800.0, + "grad_norm": 1.9105373384670177, + "learning_rate": 1.068544771061863e-06, + "loss": 0.7934, + "num_input_tokens_seen": 240080125, + "step": 7369 + }, + { + "epoch": 0.6646525679758308, + "flos": 27518202954720.0, + "grad_norm": 2.4766156404491952, + "learning_rate": 1.0680278263908787e-06, + "loss": 0.7848, + "num_input_tokens_seen": 240105985, + "step": 7370 + }, + { + "epoch": 0.6647427514993011, + "flos": 67523837863680.0, + "grad_norm": 0.7308286300545538, + "learning_rate": 1.0675109612408991e-06, + "loss": 0.5836, + "num_input_tokens_seen": 240197690, + "step": 7371 + }, + { + "epoch": 0.6648329350227713, + "flos": 24828174834720.0, + "grad_norm": 2.0391685487494016, + "learning_rate": 1.0669941756560264e-06, + "loss": 0.715, + "num_input_tokens_seen": 240225940, + "step": 7372 + }, + { + "epoch": 0.6649231185462416, + "flos": 21581666573280.0, + "grad_norm": 2.8688271638885197, + "learning_rate": 1.0664774696803548e-06, + "loss": 0.6928, + "num_input_tokens_seen": 240252770, + "step": 7373 + }, + { + "epoch": 0.6650133020697119, + "flos": 23225199674880.0, + "grad_norm": 1.8632529074634556, + "learning_rate": 1.065960843357973e-06, + "loss": 0.7175, + "num_input_tokens_seen": 240281810, + "step": 7374 + }, + { + "epoch": 0.6651034855931821, + "flos": 15938333957760.0, + "grad_norm": 2.0397770239948256, + "learning_rate": 1.065444296732963e-06, + "loss": 0.7605, + "num_input_tokens_seen": 240308345, + "step": 7375 + }, + { + "epoch": 0.6651936691166523, + "flos": 65633568929760.0, + "grad_norm": 0.671305725760997, + "learning_rate": 1.064927829849397e-06, + "loss": 0.5858, + "num_input_tokens_seen": 240398335, + "step": 7376 + }, + { + "epoch": 0.6652838526401227, + "flos": 19945493084160.0, + "grad_norm": 2.276328853642803, + "learning_rate": 1.0644114427513465e-06, + "loss": 0.7404, + "num_input_tokens_seen": 240424060, + "step": 7377 + }, + { + "epoch": 0.6653740361635929, + "flos": 26248246107360.0, + "grad_norm": 2.141366426472666, + "learning_rate": 1.0638951354828693e-06, + "loss": 0.6756, + "num_input_tokens_seen": 240456640, + "step": 7378 + }, + { + "epoch": 0.6654642196870632, + "flos": 25336848931200.0, + "grad_norm": 2.8019590574781668, + "learning_rate": 1.063378908088021e-06, + "loss": 0.7999, + "num_input_tokens_seen": 240485160, + "step": 7379 + }, + { + "epoch": 0.6655544032105334, + "flos": 21112323816480.0, + "grad_norm": 2.3080520123351396, + "learning_rate": 1.0628627606108486e-06, + "loss": 0.7372, + "num_input_tokens_seen": 240512690, + "step": 7380 + }, + { + "epoch": 0.6656445867340037, + "flos": 23695843373280.0, + "grad_norm": 2.536458061787449, + "learning_rate": 1.062346693095393e-06, + "loss": 0.8095, + "num_input_tokens_seen": 240539680, + "step": 7381 + }, + { + "epoch": 0.665734770257474, + "flos": 18742304592960.0, + "grad_norm": 1.8268884242396883, + "learning_rate": 1.0618307055856882e-06, + "loss": 0.6851, + "num_input_tokens_seen": 240568120, + "step": 7382 + }, + { + "epoch": 0.6658249537809442, + "flos": 20164420202400.0, + "grad_norm": 1.722703721134115, + "learning_rate": 1.061314798125759e-06, + "loss": 0.7975, + "num_input_tokens_seen": 240593575, + "step": 7383 + }, + { + "epoch": 0.6659151373044144, + "flos": 26321593511040.0, + "grad_norm": 1.6514771987121744, + "learning_rate": 1.0607989707596293e-06, + "loss": 0.7029, + "num_input_tokens_seen": 240626000, + "step": 7384 + }, + { + "epoch": 0.6660053208278848, + "flos": 25699683124800.0, + "grad_norm": 1.7037423790966268, + "learning_rate": 1.0602832235313078e-06, + "loss": 0.8691, + "num_input_tokens_seen": 240655165, + "step": 7385 + }, + { + "epoch": 0.666095504351355, + "flos": 21872194116480.0, + "grad_norm": 2.6145930397694475, + "learning_rate": 1.0597675564848053e-06, + "loss": 0.736, + "num_input_tokens_seen": 240681565, + "step": 7386 + }, + { + "epoch": 0.6661856878748252, + "flos": 21658470764640.0, + "grad_norm": 2.584353881159224, + "learning_rate": 1.059251969664118e-06, + "loss": 0.813, + "num_input_tokens_seen": 240707900, + "step": 7387 + }, + { + "epoch": 0.6662758713982956, + "flos": 25957458375840.0, + "grad_norm": 2.122257707153806, + "learning_rate": 1.0587364631132402e-06, + "loss": 0.7553, + "num_input_tokens_seen": 240736465, + "step": 7388 + }, + { + "epoch": 0.6663660549217658, + "flos": 26935958436000.0, + "grad_norm": 1.71324740340324, + "learning_rate": 1.0582210368761573e-06, + "loss": 0.7848, + "num_input_tokens_seen": 240765155, + "step": 7389 + }, + { + "epoch": 0.666456238445236, + "flos": 24390766635360.0, + "grad_norm": 3.028764472577338, + "learning_rate": 1.0577056909968485e-06, + "loss": 0.7495, + "num_input_tokens_seen": 240792405, + "step": 7390 + }, + { + "epoch": 0.6665464219687063, + "flos": 28793029040640.0, + "grad_norm": 2.398046505150189, + "learning_rate": 1.0571904255192857e-06, + "loss": 0.7636, + "num_input_tokens_seen": 240822410, + "step": 7391 + }, + { + "epoch": 0.6666366054921766, + "flos": 14662913155680.0, + "grad_norm": 5.651171685992255, + "learning_rate": 1.0566752404874354e-06, + "loss": 0.7759, + "num_input_tokens_seen": 240848155, + "step": 7392 + }, + { + "epoch": 0.6667267890156469, + "flos": 19727569549440.0, + "grad_norm": 13.278796303672033, + "learning_rate": 1.0561601359452543e-06, + "loss": 0.7426, + "num_input_tokens_seen": 240873510, + "step": 7393 + }, + { + "epoch": 0.6668169725391171, + "flos": 35750184553920.0, + "grad_norm": 2.646215716587251, + "learning_rate": 1.0556451119366947e-06, + "loss": 0.6801, + "num_input_tokens_seen": 240903630, + "step": 7394 + }, + { + "epoch": 0.6669071560625873, + "flos": 42785668218720.0, + "grad_norm": 1.8321257934901733, + "learning_rate": 1.0551301685057011e-06, + "loss": 0.6847, + "num_input_tokens_seen": 240936350, + "step": 7395 + }, + { + "epoch": 0.6669973395860577, + "flos": 26279697457920.0, + "grad_norm": 1.8269140683465543, + "learning_rate": 1.0546153056962117e-06, + "loss": 0.8295, + "num_input_tokens_seen": 240964325, + "step": 7396 + }, + { + "epoch": 0.6670875231095279, + "flos": 20783951724000.0, + "grad_norm": 2.0297827483184236, + "learning_rate": 1.0541005235521578e-06, + "loss": 0.7417, + "num_input_tokens_seen": 240993685, + "step": 7397 + }, + { + "epoch": 0.6671777066329981, + "flos": 18962867180640.0, + "grad_norm": 2.023610582644707, + "learning_rate": 1.0535858221174614e-06, + "loss": 0.7577, + "num_input_tokens_seen": 241018955, + "step": 7398 + }, + { + "epoch": 0.6672678901564684, + "flos": 19872703226880.0, + "grad_norm": 1.5761382939615085, + "learning_rate": 1.0530712014360426e-06, + "loss": 0.7827, + "num_input_tokens_seen": 241047145, + "step": 7399 + }, + { + "epoch": 0.6673580736799387, + "flos": 22346257432800.0, + "grad_norm": 1.7832896952177182, + "learning_rate": 1.0525566615518088e-06, + "loss": 0.7967, + "num_input_tokens_seen": 241074460, + "step": 7400 + }, + { + "epoch": 0.667448257203409, + "flos": 23691866208960.0, + "grad_norm": 1.719225384252961, + "learning_rate": 1.0520422025086662e-06, + "loss": 0.7595, + "num_input_tokens_seen": 241102940, + "step": 7401 + }, + { + "epoch": 0.6675384407268792, + "flos": 27124140429120.0, + "grad_norm": 1.9787751730825587, + "learning_rate": 1.0515278243505092e-06, + "loss": 0.7982, + "num_input_tokens_seen": 241132810, + "step": 7402 + }, + { + "epoch": 0.6676286242503494, + "flos": 19254286798080.0, + "grad_norm": 1.4672866999003944, + "learning_rate": 1.0510135271212278e-06, + "loss": 0.7717, + "num_input_tokens_seen": 241161110, + "step": 7403 + }, + { + "epoch": 0.6677188077738198, + "flos": 25737155976480.0, + "grad_norm": 1.6030205765115688, + "learning_rate": 1.0504993108647052e-06, + "loss": 0.8109, + "num_input_tokens_seen": 241192110, + "step": 7404 + }, + { + "epoch": 0.66780899129729, + "flos": 47488122304800.0, + "grad_norm": 1.6162508403542821, + "learning_rate": 1.0499851756248168e-06, + "loss": 0.7168, + "num_input_tokens_seen": 241226180, + "step": 7405 + }, + { + "epoch": 0.6678991748207602, + "flos": 26392747898880.0, + "grad_norm": 2.705564952244804, + "learning_rate": 1.0494711214454316e-06, + "loss": 0.7287, + "num_input_tokens_seen": 241257785, + "step": 7406 + }, + { + "epoch": 0.6679893583442305, + "flos": 28653024790080.0, + "grad_norm": 2.1512269932404933, + "learning_rate": 1.0489571483704111e-06, + "loss": 0.6641, + "num_input_tokens_seen": 241287760, + "step": 7407 + }, + { + "epoch": 0.6680795418677008, + "flos": 29527692509760.0, + "grad_norm": 2.2036712185803964, + "learning_rate": 1.048443256443612e-06, + "loss": 0.6616, + "num_input_tokens_seen": 241316180, + "step": 7408 + }, + { + "epoch": 0.668169725391171, + "flos": 19764410515200.0, + "grad_norm": 1.9636955167228602, + "learning_rate": 1.0479294457088801e-06, + "loss": 0.7631, + "num_input_tokens_seen": 241343990, + "step": 7409 + }, + { + "epoch": 0.6682599089146413, + "flos": 25593509089440.0, + "grad_norm": 1.7384026302863707, + "learning_rate": 1.0474157162100574e-06, + "loss": 0.8679, + "num_input_tokens_seen": 241373725, + "step": 7410 + }, + { + "epoch": 0.6683500924381116, + "flos": 14335804835040.0, + "grad_norm": 1.8987054831916022, + "learning_rate": 1.0469020679909786e-06, + "loss": 0.6834, + "num_input_tokens_seen": 241397790, + "step": 7411 + }, + { + "epoch": 0.6684402759615818, + "flos": 30365890961280.0, + "grad_norm": 2.1388287212391757, + "learning_rate": 1.0463885010954705e-06, + "loss": 0.6716, + "num_input_tokens_seen": 241426580, + "step": 7412 + }, + { + "epoch": 0.6685304594850521, + "flos": 20528220809760.0, + "grad_norm": 1.9355811184376246, + "learning_rate": 1.0458750155673536e-06, + "loss": 0.8161, + "num_input_tokens_seen": 241452850, + "step": 7413 + }, + { + "epoch": 0.6686206430085223, + "flos": 31786110912960.0, + "grad_norm": 1.9985019915019429, + "learning_rate": 1.0453616114504421e-06, + "loss": 0.7008, + "num_input_tokens_seen": 241482335, + "step": 7414 + }, + { + "epoch": 0.6687108265319927, + "flos": 20419630740000.0, + "grad_norm": 2.3384160699796586, + "learning_rate": 1.0448482887885406e-06, + "loss": 0.7794, + "num_input_tokens_seen": 241510355, + "step": 7415 + }, + { + "epoch": 0.6688010100554629, + "flos": 21690554001120.0, + "grad_norm": 2.0709090614927783, + "learning_rate": 1.044335047625451e-06, + "loss": 0.7509, + "num_input_tokens_seen": 241539530, + "step": 7416 + }, + { + "epoch": 0.6688911935789331, + "flos": 34738151636640.0, + "grad_norm": 4.2125130220206275, + "learning_rate": 1.0438218880049637e-06, + "loss": 0.7567, + "num_input_tokens_seen": 241570950, + "step": 7417 + }, + { + "epoch": 0.6689813771024034, + "flos": 22642583458560.0, + "grad_norm": 2.1089885216333215, + "learning_rate": 1.0433088099708653e-06, + "loss": 0.8032, + "num_input_tokens_seen": 241596905, + "step": 7418 + }, + { + "epoch": 0.6690715606258737, + "flos": 67236060882720.0, + "grad_norm": 0.5928267780167096, + "learning_rate": 1.0427958135669346e-06, + "loss": 0.5117, + "num_input_tokens_seen": 241692270, + "step": 7419 + }, + { + "epoch": 0.6691617441493439, + "flos": 24827951816160.0, + "grad_norm": 1.961579662209633, + "learning_rate": 1.0422828988369428e-06, + "loss": 0.6805, + "num_input_tokens_seen": 241722055, + "step": 7420 + }, + { + "epoch": 0.6692519276728142, + "flos": 20893917074880.0, + "grad_norm": 1.6444679664885562, + "learning_rate": 1.041770065824655e-06, + "loss": 0.8159, + "num_input_tokens_seen": 241749695, + "step": 7421 + }, + { + "epoch": 0.6693421111962844, + "flos": 35248201014240.0, + "grad_norm": 1.8802860768766, + "learning_rate": 1.0412573145738287e-06, + "loss": 0.6007, + "num_input_tokens_seen": 241782485, + "step": 7422 + }, + { + "epoch": 0.6694322947197547, + "flos": 22823851876320.0, + "grad_norm": 2.056732668361157, + "learning_rate": 1.040744645128216e-06, + "loss": 0.7217, + "num_input_tokens_seen": 241809065, + "step": 7423 + }, + { + "epoch": 0.669522478243225, + "flos": 24464262718080.0, + "grad_norm": 1.9558263789236017, + "learning_rate": 1.040232057531558e-06, + "loss": 0.8042, + "num_input_tokens_seen": 241838130, + "step": 7424 + }, + { + "epoch": 0.6696126617666952, + "flos": 15682974741120.0, + "grad_norm": 3.4509897664354194, + "learning_rate": 1.0397195518275932e-06, + "loss": 0.7217, + "num_input_tokens_seen": 241863150, + "step": 7425 + }, + { + "epoch": 0.6697028452901654, + "flos": 49526795855040.0, + "grad_norm": 2.151423927559759, + "learning_rate": 1.0392071280600512e-06, + "loss": 0.7166, + "num_input_tokens_seen": 241897530, + "step": 7426 + }, + { + "epoch": 0.6697930288136358, + "flos": 20602162929600.0, + "grad_norm": 2.0983063952974415, + "learning_rate": 1.0386947862726549e-06, + "loss": 0.7462, + "num_input_tokens_seen": 241926005, + "step": 7427 + }, + { + "epoch": 0.669883212337106, + "flos": 41368570526880.0, + "grad_norm": 2.1510202870141613, + "learning_rate": 1.0381825265091197e-06, + "loss": 0.6684, + "num_input_tokens_seen": 241956370, + "step": 7428 + }, + { + "epoch": 0.6699733958605762, + "flos": 22569273224640.0, + "grad_norm": 2.073026670160158, + "learning_rate": 1.037670348813155e-06, + "loss": 0.6669, + "num_input_tokens_seen": 241985650, + "step": 7429 + }, + { + "epoch": 0.6700635793840465, + "flos": 27593706204480.0, + "grad_norm": 1.9230759132822761, + "learning_rate": 1.0371582532284624e-06, + "loss": 0.7768, + "num_input_tokens_seen": 242015130, + "step": 7430 + }, + { + "epoch": 0.6701537629075168, + "flos": 55095480125280.0, + "grad_norm": 1.8936008331672503, + "learning_rate": 1.0366462397987375e-06, + "loss": 0.6791, + "num_input_tokens_seen": 242048155, + "step": 7431 + }, + { + "epoch": 0.6702439464309871, + "flos": 22274471159040.0, + "grad_norm": 2.5679611999753535, + "learning_rate": 1.0361343085676665e-06, + "loss": 0.7537, + "num_input_tokens_seen": 242074750, + "step": 7432 + }, + { + "epoch": 0.6703341299544573, + "flos": 21112472495520.0, + "grad_norm": 1.8472164993707385, + "learning_rate": 1.0356224595789309e-06, + "loss": 0.8627, + "num_input_tokens_seen": 242103190, + "step": 7433 + }, + { + "epoch": 0.6704243134779276, + "flos": 22455925425600.0, + "grad_norm": 1.8688242613667683, + "learning_rate": 1.0351106928762046e-06, + "loss": 0.8038, + "num_input_tokens_seen": 242130835, + "step": 7434 + }, + { + "epoch": 0.6705144970013979, + "flos": 12805470853440.0, + "grad_norm": 2.538392057668365, + "learning_rate": 1.034599008503154e-06, + "loss": 0.818, + "num_input_tokens_seen": 242156770, + "step": 7435 + }, + { + "epoch": 0.6706046805248681, + "flos": 21367943221440.0, + "grad_norm": 2.471052134594895, + "learning_rate": 1.0340874065034406e-06, + "loss": 0.7364, + "num_input_tokens_seen": 242187135, + "step": 7436 + }, + { + "epoch": 0.6706948640483383, + "flos": 22641802893600.0, + "grad_norm": 1.9333066785005335, + "learning_rate": 1.0335758869207137e-06, + "loss": 0.6727, + "num_input_tokens_seen": 242214230, + "step": 7437 + }, + { + "epoch": 0.6707850475718087, + "flos": 20233455913920.0, + "grad_norm": 2.932943340587936, + "learning_rate": 1.0330644497986227e-06, + "loss": 0.7638, + "num_input_tokens_seen": 242242650, + "step": 7438 + }, + { + "epoch": 0.6708752310952789, + "flos": 28106543314080.0, + "grad_norm": 1.767357845174307, + "learning_rate": 1.0325530951808029e-06, + "loss": 0.7854, + "num_input_tokens_seen": 242269885, + "step": 7439 + }, + { + "epoch": 0.6709654146187491, + "flos": 33237075989760.0, + "grad_norm": 1.959117272567137, + "learning_rate": 1.0320418231108887e-06, + "loss": 0.7551, + "num_input_tokens_seen": 242300560, + "step": 7440 + }, + { + "epoch": 0.6710555981422194, + "flos": 22892478720480.0, + "grad_norm": 1.8052095260953696, + "learning_rate": 1.0315306336325028e-06, + "loss": 0.7877, + "num_input_tokens_seen": 242329955, + "step": 7441 + }, + { + "epoch": 0.6711457816656897, + "flos": 18700371370080.0, + "grad_norm": 1.9670069410922097, + "learning_rate": 1.0310195267892635e-06, + "loss": 0.7013, + "num_input_tokens_seen": 242356210, + "step": 7442 + }, + { + "epoch": 0.67123596518916, + "flos": 27705976080480.0, + "grad_norm": 1.9256516852134768, + "learning_rate": 1.030508502624781e-06, + "loss": 0.6918, + "num_input_tokens_seen": 242387210, + "step": 7443 + }, + { + "epoch": 0.6713261487126302, + "flos": 21768213096960.0, + "grad_norm": 1.8412336214350973, + "learning_rate": 1.0299975611826587e-06, + "loss": 0.6958, + "num_input_tokens_seen": 242418745, + "step": 7444 + }, + { + "epoch": 0.6714163322361004, + "flos": 20671310150400.0, + "grad_norm": 1.6329109855774313, + "learning_rate": 1.0294867025064928e-06, + "loss": 0.7545, + "num_input_tokens_seen": 242448095, + "step": 7445 + }, + { + "epoch": 0.6715065157595708, + "flos": 22094838210720.0, + "grad_norm": 1.9486064675492474, + "learning_rate": 1.028975926639874e-06, + "loss": 0.755, + "num_input_tokens_seen": 242479270, + "step": 7446 + }, + { + "epoch": 0.671596699283041, + "flos": 13096890470880.0, + "grad_norm": 2.5887363189232606, + "learning_rate": 1.0284652336263823e-06, + "loss": 0.6654, + "num_input_tokens_seen": 242503245, + "step": 7447 + }, + { + "epoch": 0.6716868828065112, + "flos": 15684424361760.0, + "grad_norm": 1.882936465665831, + "learning_rate": 1.0279546235095938e-06, + "loss": 0.674, + "num_input_tokens_seen": 242531045, + "step": 7448 + }, + { + "epoch": 0.6717770663299815, + "flos": 66412024109760.0, + "grad_norm": 0.7347430666310519, + "learning_rate": 1.0274440963330768e-06, + "loss": 0.608, + "num_input_tokens_seen": 242618995, + "step": 7449 + }, + { + "epoch": 0.6718672498534518, + "flos": 22017104775360.0, + "grad_norm": 1.9226712073992487, + "learning_rate": 1.0269336521403919e-06, + "loss": 0.7471, + "num_input_tokens_seen": 242647095, + "step": 7450 + }, + { + "epoch": 0.671957433376922, + "flos": 26066829010560.0, + "grad_norm": 1.9717726739852026, + "learning_rate": 1.0264232909750936e-06, + "loss": 0.8747, + "num_input_tokens_seen": 242675290, + "step": 7451 + }, + { + "epoch": 0.6720476169003923, + "flos": 26393268275520.0, + "grad_norm": 1.7301020868863732, + "learning_rate": 1.025913012880728e-06, + "loss": 0.8098, + "num_input_tokens_seen": 242706655, + "step": 7452 + }, + { + "epoch": 0.6721378004238625, + "flos": 20855180451360.0, + "grad_norm": 1.5913120397323843, + "learning_rate": 1.0254028179008362e-06, + "loss": 0.6561, + "num_input_tokens_seen": 242734640, + "step": 7453 + }, + { + "epoch": 0.6722279839473329, + "flos": 28432796730240.0, + "grad_norm": 2.2318134696979923, + "learning_rate": 1.0248927060789483e-06, + "loss": 0.6402, + "num_input_tokens_seen": 242762445, + "step": 7454 + }, + { + "epoch": 0.6723181674708031, + "flos": 18306866390880.0, + "grad_norm": 1.664066798244889, + "learning_rate": 1.0243826774585928e-06, + "loss": 0.7053, + "num_input_tokens_seen": 242791010, + "step": 7455 + }, + { + "epoch": 0.6724083509942733, + "flos": 25156881455040.0, + "grad_norm": 2.136122932561213, + "learning_rate": 1.0238727320832854e-06, + "loss": 0.7333, + "num_input_tokens_seen": 242822440, + "step": 7456 + }, + { + "epoch": 0.6724985345177437, + "flos": 23662273346400.0, + "grad_norm": 4.537867062016491, + "learning_rate": 1.0233628699965403e-06, + "loss": 0.7495, + "num_input_tokens_seen": 242850795, + "step": 7457 + }, + { + "epoch": 0.6725887180412139, + "flos": 23039954092800.0, + "grad_norm": 1.785090205266762, + "learning_rate": 1.0228530912418594e-06, + "loss": 0.7701, + "num_input_tokens_seen": 242878360, + "step": 7458 + }, + { + "epoch": 0.6726789015646841, + "flos": 28577187012480.0, + "grad_norm": 3.1415268850892826, + "learning_rate": 1.0223433958627404e-06, + "loss": 0.7181, + "num_input_tokens_seen": 242907180, + "step": 7459 + }, + { + "epoch": 0.6727690850881544, + "flos": 30620729801280.0, + "grad_norm": 3.077304788901149, + "learning_rate": 1.021833783902674e-06, + "loss": 0.7696, + "num_input_tokens_seen": 242934270, + "step": 7460 + }, + { + "epoch": 0.6728592686116247, + "flos": 15173557249440.0, + "grad_norm": 1.9612190682747934, + "learning_rate": 1.0213242554051427e-06, + "loss": 0.7524, + "num_input_tokens_seen": 242960615, + "step": 7461 + }, + { + "epoch": 0.6729494521350949, + "flos": 19726826154240.0, + "grad_norm": 1.7858317826988666, + "learning_rate": 1.0208148104136229e-06, + "loss": 0.6879, + "num_input_tokens_seen": 242989130, + "step": 7462 + }, + { + "epoch": 0.6730396356585652, + "flos": 14918160863040.0, + "grad_norm": 2.0213916702021884, + "learning_rate": 1.020305448971582e-06, + "loss": 0.7202, + "num_input_tokens_seen": 243015710, + "step": 7463 + }, + { + "epoch": 0.6731298191820354, + "flos": 25880654184480.0, + "grad_norm": 1.8567153116265145, + "learning_rate": 1.0197961711224824e-06, + "loss": 0.7122, + "num_input_tokens_seen": 243045525, + "step": 7464 + }, + { + "epoch": 0.6732200027055057, + "flos": 21109313065920.0, + "grad_norm": 2.0656263758298414, + "learning_rate": 1.0192869769097777e-06, + "loss": 0.7577, + "num_input_tokens_seen": 243073830, + "step": 7465 + }, + { + "epoch": 0.673310186228976, + "flos": 20274534232320.0, + "grad_norm": 2.9833707867478867, + "learning_rate": 1.018777866376916e-06, + "loss": 0.7514, + "num_input_tokens_seen": 243101060, + "step": 7466 + }, + { + "epoch": 0.6734003697524462, + "flos": 19253246044800.0, + "grad_norm": 5.055230655407339, + "learning_rate": 1.0182688395673374e-06, + "loss": 0.6159, + "num_input_tokens_seen": 243128770, + "step": 7467 + }, + { + "epoch": 0.6734905532759164, + "flos": 21512928219840.0, + "grad_norm": 1.8179875535570384, + "learning_rate": 1.017759896524475e-06, + "loss": 0.7148, + "num_input_tokens_seen": 243159155, + "step": 7468 + }, + { + "epoch": 0.6735807367993868, + "flos": 26576283672000.0, + "grad_norm": 2.2222414082122786, + "learning_rate": 1.0172510372917528e-06, + "loss": 0.7812, + "num_input_tokens_seen": 243184310, + "step": 7469 + }, + { + "epoch": 0.673670920322857, + "flos": 17317772949120.0, + "grad_norm": 2.5058250276228944, + "learning_rate": 1.0167422619125925e-06, + "loss": 0.6559, + "num_input_tokens_seen": 243209845, + "step": 7470 + }, + { + "epoch": 0.6737611038463273, + "flos": 24318683003520.0, + "grad_norm": 2.5475018437821566, + "learning_rate": 1.0162335704304026e-06, + "loss": 0.7507, + "num_input_tokens_seen": 243239290, + "step": 7471 + }, + { + "epoch": 0.6738512873697975, + "flos": 67190893890720.0, + "grad_norm": 0.6905072288728675, + "learning_rate": 1.0157249628885903e-06, + "loss": 0.5916, + "num_input_tokens_seen": 243325750, + "step": 7472 + }, + { + "epoch": 0.6739414708932678, + "flos": 27925683763680.0, + "grad_norm": 1.3371637082017065, + "learning_rate": 1.0152164393305506e-06, + "loss": 0.6553, + "num_input_tokens_seen": 243357070, + "step": 7473 + }, + { + "epoch": 0.6740316544167381, + "flos": 22566113795040.0, + "grad_norm": 2.2371563439744784, + "learning_rate": 1.0147079997996746e-06, + "loss": 0.7418, + "num_input_tokens_seen": 243384475, + "step": 7474 + }, + { + "epoch": 0.6741218379402083, + "flos": 22423730679840.0, + "grad_norm": 2.195650122260171, + "learning_rate": 1.0141996443393446e-06, + "loss": 0.7776, + "num_input_tokens_seen": 243412110, + "step": 7475 + }, + { + "epoch": 0.6742120214636785, + "flos": 15574905048000.0, + "grad_norm": 1.9738870841045284, + "learning_rate": 1.0136913729929369e-06, + "loss": 0.7315, + "num_input_tokens_seen": 243439365, + "step": 7476 + }, + { + "epoch": 0.6743022049871489, + "flos": 27232544650080.0, + "grad_norm": 1.9041088774389252, + "learning_rate": 1.0131831858038203e-06, + "loss": 0.7041, + "num_input_tokens_seen": 243467720, + "step": 7477 + }, + { + "epoch": 0.6743923885106191, + "flos": 31383722361120.0, + "grad_norm": 1.7918297301615655, + "learning_rate": 1.0126750828153538e-06, + "loss": 0.8078, + "num_input_tokens_seen": 243498795, + "step": 7478 + }, + { + "epoch": 0.6744825720340893, + "flos": 14663582211360.0, + "grad_norm": 2.1666378850736967, + "learning_rate": 1.012167064070895e-06, + "loss": 0.7872, + "num_input_tokens_seen": 243523740, + "step": 7479 + }, + { + "epoch": 0.6745727555575597, + "flos": 21580254122400.0, + "grad_norm": 1.99812330172517, + "learning_rate": 1.0116591296137885e-06, + "loss": 0.7735, + "num_input_tokens_seen": 243550260, + "step": 7480 + }, + { + "epoch": 0.6746629390810299, + "flos": 25993778964960.0, + "grad_norm": 1.8340915571424123, + "learning_rate": 1.0111512794873746e-06, + "loss": 0.7723, + "num_input_tokens_seen": 243583135, + "step": 7481 + }, + { + "epoch": 0.6747531226045002, + "flos": 27737576110080.0, + "grad_norm": 1.5302223338228258, + "learning_rate": 1.010643513734986e-06, + "loss": 0.7462, + "num_input_tokens_seen": 243613885, + "step": 7482 + }, + { + "epoch": 0.6748433061279704, + "flos": 21913383944160.0, + "grad_norm": 1.8009841374361155, + "learning_rate": 1.010135832399948e-06, + "loss": 0.7851, + "num_input_tokens_seen": 243642135, + "step": 7483 + }, + { + "epoch": 0.6749334896514407, + "flos": 20020252938720.0, + "grad_norm": 1.5345156354607903, + "learning_rate": 1.0096282355255792e-06, + "loss": 0.7517, + "num_input_tokens_seen": 243672000, + "step": 7484 + }, + { + "epoch": 0.675023673174911, + "flos": 18197309907360.0, + "grad_norm": 2.498721845047431, + "learning_rate": 1.0091207231551905e-06, + "loss": 0.8037, + "num_input_tokens_seen": 243696860, + "step": 7485 + }, + { + "epoch": 0.6751138566983812, + "flos": 23152038120000.0, + "grad_norm": 1.7839409428858535, + "learning_rate": 1.0086132953320842e-06, + "loss": 0.7114, + "num_input_tokens_seen": 243724985, + "step": 7486 + }, + { + "epoch": 0.6752040402218514, + "flos": 29634981637920.0, + "grad_norm": 2.094018719637759, + "learning_rate": 1.0081059520995591e-06, + "loss": 0.7778, + "num_input_tokens_seen": 243753430, + "step": 7487 + }, + { + "epoch": 0.6752942237453218, + "flos": 25666261776960.0, + "grad_norm": 2.0017206419889, + "learning_rate": 1.0075986935009028e-06, + "loss": 0.8177, + "num_input_tokens_seen": 243781460, + "step": 7488 + }, + { + "epoch": 0.675384407268792, + "flos": 21768361776000.0, + "grad_norm": 2.0508525145727914, + "learning_rate": 1.0070915195793982e-06, + "loss": 0.7585, + "num_input_tokens_seen": 243808790, + "step": 7489 + }, + { + "epoch": 0.6754745907922622, + "flos": 22168185614400.0, + "grad_norm": 2.7991742960205164, + "learning_rate": 1.0065844303783197e-06, + "loss": 0.7229, + "num_input_tokens_seen": 243835660, + "step": 7490 + }, + { + "epoch": 0.6755647743157325, + "flos": 23258695362240.0, + "grad_norm": 1.9279420193940646, + "learning_rate": 1.0060774259409356e-06, + "loss": 0.6645, + "num_input_tokens_seen": 243864545, + "step": 7491 + }, + { + "epoch": 0.6756549578392028, + "flos": 22166921842560.0, + "grad_norm": 2.2679230709383815, + "learning_rate": 1.0055705063105065e-06, + "loss": 0.782, + "num_input_tokens_seen": 243894810, + "step": 7492 + }, + { + "epoch": 0.675745141362673, + "flos": 19217520171840.0, + "grad_norm": 2.186083525769458, + "learning_rate": 1.0050636715302837e-06, + "loss": 0.7398, + "num_input_tokens_seen": 243923255, + "step": 7493 + }, + { + "epoch": 0.6758353248861433, + "flos": 23078988074400.0, + "grad_norm": 2.3599660777224085, + "learning_rate": 1.0045569216435157e-06, + "loss": 0.7603, + "num_input_tokens_seen": 243950790, + "step": 7494 + }, + { + "epoch": 0.6759255084096135, + "flos": 21982568334720.0, + "grad_norm": 1.7233919056371505, + "learning_rate": 1.0040502566934384e-06, + "loss": 0.8269, + "num_input_tokens_seen": 243978815, + "step": 7495 + }, + { + "epoch": 0.6760156919330839, + "flos": 22751210698080.0, + "grad_norm": 2.28139401334009, + "learning_rate": 1.0035436767232866e-06, + "loss": 0.6843, + "num_input_tokens_seen": 244005105, + "step": 7496 + }, + { + "epoch": 0.6761058754565541, + "flos": 23079173923200.0, + "grad_norm": 1.639362321601734, + "learning_rate": 1.0030371817762816e-06, + "loss": 0.7603, + "num_input_tokens_seen": 244032300, + "step": 7497 + }, + { + "epoch": 0.6761960589800243, + "flos": 20930497852320.0, + "grad_norm": 1.727523122210411, + "learning_rate": 1.0025307718956417e-06, + "loss": 0.7546, + "num_input_tokens_seen": 244061180, + "step": 7498 + }, + { + "epoch": 0.6762862425034946, + "flos": 25374730650240.0, + "grad_norm": 1.6718966045944552, + "learning_rate": 1.0020244471245765e-06, + "loss": 0.7695, + "num_input_tokens_seen": 244091370, + "step": 7499 + }, + { + "epoch": 0.6763764260269649, + "flos": 20820904199040.0, + "grad_norm": 2.374359646853227, + "learning_rate": 1.001518207506288e-06, + "loss": 0.7028, + "num_input_tokens_seen": 244116630, + "step": 7500 + }, + { + "epoch": 0.6764666095504351, + "flos": 23442193965600.0, + "grad_norm": 1.6960863026262611, + "learning_rate": 1.0010120530839717e-06, + "loss": 0.8119, + "num_input_tokens_seen": 244145615, + "step": 7501 + }, + { + "epoch": 0.6765567930739054, + "flos": 25775037695520.0, + "grad_norm": 1.9569283646818705, + "learning_rate": 1.0005059839008161e-06, + "loss": 0.7712, + "num_input_tokens_seen": 244177510, + "step": 7502 + }, + { + "epoch": 0.6766469765973756, + "flos": 22715187467040.0, + "grad_norm": 2.09672210181395, + "learning_rate": 1.0000000000000004e-06, + "loss": 0.7001, + "num_input_tokens_seen": 244206740, + "step": 7503 + }, + { + "epoch": 0.676737160120846, + "flos": 19873669640640.0, + "grad_norm": 1.6778906859300053, + "learning_rate": 9.994941014246985e-07, + "loss": 0.793, + "num_input_tokens_seen": 244234315, + "step": 7504 + }, + { + "epoch": 0.6768273436443162, + "flos": 25299636267840.0, + "grad_norm": 1.8231656876274949, + "learning_rate": 9.989882882180766e-07, + "loss": 0.6391, + "num_input_tokens_seen": 244265645, + "step": 7505 + }, + { + "epoch": 0.6769175271677864, + "flos": 18707954001120.0, + "grad_norm": 1.9145761851556753, + "learning_rate": 9.984825604232938e-07, + "loss": 0.7657, + "num_input_tokens_seen": 244291905, + "step": 7506 + }, + { + "epoch": 0.6770077106912568, + "flos": 22926903651840.0, + "grad_norm": 2.503632576636673, + "learning_rate": 9.97976918083502e-07, + "loss": 0.688, + "num_input_tokens_seen": 244320390, + "step": 7507 + }, + { + "epoch": 0.677097894214727, + "flos": 33171645744960.0, + "grad_norm": 1.7792720855761488, + "learning_rate": 9.974713612418427e-07, + "loss": 0.82, + "num_input_tokens_seen": 244353460, + "step": 7508 + }, + { + "epoch": 0.6771880777381972, + "flos": 27378719080800.0, + "grad_norm": 2.0430469388569024, + "learning_rate": 9.969658899414563e-07, + "loss": 0.7608, + "num_input_tokens_seen": 244383820, + "step": 7509 + }, + { + "epoch": 0.6772782612616675, + "flos": 22496037330240.0, + "grad_norm": 1.930188442181945, + "learning_rate": 9.964605042254696e-07, + "loss": 0.6654, + "num_input_tokens_seen": 244411690, + "step": 7510 + }, + { + "epoch": 0.6773684447851378, + "flos": 19650170641920.0, + "grad_norm": 1.895443537397169, + "learning_rate": 9.959552041370076e-07, + "loss": 0.6879, + "num_input_tokens_seen": 244439805, + "step": 7511 + }, + { + "epoch": 0.677458628308608, + "flos": 25447148809920.0, + "grad_norm": 1.9174745004443021, + "learning_rate": 9.954499897191824e-07, + "loss": 0.6661, + "num_input_tokens_seen": 244469160, + "step": 7512 + }, + { + "epoch": 0.6775488118320783, + "flos": 39801432749280.0, + "grad_norm": 4.094081020803964, + "learning_rate": 9.949448610151043e-07, + "loss": 0.7, + "num_input_tokens_seen": 244497165, + "step": 7513 + }, + { + "epoch": 0.6776389953555485, + "flos": 34409445016320.0, + "grad_norm": 8.69186577085802, + "learning_rate": 9.944398180678719e-07, + "loss": 0.7256, + "num_input_tokens_seen": 244530540, + "step": 7514 + }, + { + "epoch": 0.6777291788790188, + "flos": 28726223514720.0, + "grad_norm": 5.176530385054509, + "learning_rate": 9.939348609205789e-07, + "loss": 0.6981, + "num_input_tokens_seen": 244559815, + "step": 7515 + }, + { + "epoch": 0.6778193624024891, + "flos": 25994001983520.0, + "grad_norm": 2.1057560146172976, + "learning_rate": 9.93429989616311e-07, + "loss": 0.7556, + "num_input_tokens_seen": 244590575, + "step": 7516 + }, + { + "epoch": 0.6779095459259593, + "flos": 20820420992160.0, + "grad_norm": 1.656110865668901, + "learning_rate": 9.929252041981464e-07, + "loss": 0.663, + "num_input_tokens_seen": 244619555, + "step": 7517 + }, + { + "epoch": 0.6779997294494295, + "flos": 28283054002560.0, + "grad_norm": 2.3967778431126736, + "learning_rate": 9.924205047091572e-07, + "loss": 0.7195, + "num_input_tokens_seen": 244648860, + "step": 7518 + }, + { + "epoch": 0.6780899129728999, + "flos": 22235400007680.0, + "grad_norm": 4.70944873572366, + "learning_rate": 9.919158911924056e-07, + "loss": 0.8688, + "num_input_tokens_seen": 244675905, + "step": 7519 + }, + { + "epoch": 0.6781800964963701, + "flos": 21731669489280.0, + "grad_norm": 1.7179535841261586, + "learning_rate": 9.914113636909483e-07, + "loss": 0.7821, + "num_input_tokens_seen": 244705985, + "step": 7520 + }, + { + "epoch": 0.6782702800198404, + "flos": 58867792008960.0, + "grad_norm": 0.7343695620071682, + "learning_rate": 9.90906922247835e-07, + "loss": 0.626, + "num_input_tokens_seen": 244799865, + "step": 7521 + }, + { + "epoch": 0.6783604635433106, + "flos": 28506404322240.0, + "grad_norm": 2.345746525117605, + "learning_rate": 9.904025669061072e-07, + "loss": 0.7747, + "num_input_tokens_seen": 244832440, + "step": 7522 + }, + { + "epoch": 0.6784506470667809, + "flos": 31092711611040.0, + "grad_norm": 1.7698617251852289, + "learning_rate": 9.89898297708799e-07, + "loss": 0.7444, + "num_input_tokens_seen": 244864425, + "step": 7523 + }, + { + "epoch": 0.6785408305902512, + "flos": 25957272527040.0, + "grad_norm": 8.370622544192502, + "learning_rate": 9.893941146989388e-07, + "loss": 0.6977, + "num_input_tokens_seen": 244894490, + "step": 7524 + }, + { + "epoch": 0.6786310141137214, + "flos": 38964014862720.0, + "grad_norm": 2.24234603693523, + "learning_rate": 9.888900179195437e-07, + "loss": 0.7117, + "num_input_tokens_seen": 244925005, + "step": 7525 + }, + { + "epoch": 0.6787211976371916, + "flos": 19903968728640.0, + "grad_norm": 2.07193608949107, + "learning_rate": 9.883860074136285e-07, + "loss": 0.8157, + "num_input_tokens_seen": 244952590, + "step": 7526 + }, + { + "epoch": 0.678811381160662, + "flos": 27880107904320.0, + "grad_norm": 2.2790407953833305, + "learning_rate": 9.87882083224196e-07, + "loss": 0.7389, + "num_input_tokens_seen": 244977700, + "step": 7527 + }, + { + "epoch": 0.6789015646841322, + "flos": 68503267167840.0, + "grad_norm": 0.9637916790451236, + "learning_rate": 9.873782453942462e-07, + "loss": 0.6037, + "num_input_tokens_seen": 245072410, + "step": 7528 + }, + { + "epoch": 0.6789917482076024, + "flos": 20820346652640.0, + "grad_norm": 11.744966768142243, + "learning_rate": 9.868744939667676e-07, + "loss": 0.6733, + "num_input_tokens_seen": 245099260, + "step": 7529 + }, + { + "epoch": 0.6790819317310728, + "flos": 25447000130880.0, + "grad_norm": 2.1244259227138595, + "learning_rate": 9.863708289847432e-07, + "loss": 0.6312, + "num_input_tokens_seen": 245127085, + "step": 7530 + }, + { + "epoch": 0.679172115254543, + "flos": 25047585159840.0, + "grad_norm": 2.41730259607127, + "learning_rate": 9.85867250491149e-07, + "loss": 0.655, + "num_input_tokens_seen": 245156495, + "step": 7531 + }, + { + "epoch": 0.6792622987780133, + "flos": 19253840760960.0, + "grad_norm": 2.0694587530655566, + "learning_rate": 9.853637585289528e-07, + "loss": 0.8563, + "num_input_tokens_seen": 245183555, + "step": 7532 + }, + { + "epoch": 0.6793524823014835, + "flos": 28434989746080.0, + "grad_norm": 2.187024637070437, + "learning_rate": 9.848603531411159e-07, + "loss": 0.7342, + "num_input_tokens_seen": 245215595, + "step": 7533 + }, + { + "epoch": 0.6794426658249538, + "flos": 18416348534880.0, + "grad_norm": 2.212042758052418, + "learning_rate": 9.843570343705899e-07, + "loss": 0.757, + "num_input_tokens_seen": 245241840, + "step": 7534 + }, + { + "epoch": 0.6795328493484241, + "flos": 27194105384640.0, + "grad_norm": 1.4803577874705023, + "learning_rate": 9.83853802260323e-07, + "loss": 0.8191, + "num_input_tokens_seen": 245272550, + "step": 7535 + }, + { + "epoch": 0.6796230328718943, + "flos": 25521090929760.0, + "grad_norm": 1.9442458081587124, + "learning_rate": 9.833506568532524e-07, + "loss": 0.6887, + "num_input_tokens_seen": 245301470, + "step": 7536 + }, + { + "epoch": 0.6797132163953645, + "flos": 19363657432800.0, + "grad_norm": 2.069373151616475, + "learning_rate": 9.828475981923093e-07, + "loss": 0.5911, + "num_input_tokens_seen": 245330295, + "step": 7537 + }, + { + "epoch": 0.6798033999188349, + "flos": 20856704411520.0, + "grad_norm": 2.6433064767339878, + "learning_rate": 9.823446263204175e-07, + "loss": 0.7239, + "num_input_tokens_seen": 245356530, + "step": 7538 + }, + { + "epoch": 0.6798935834423051, + "flos": 23589818016960.0, + "grad_norm": 1.824578782684489, + "learning_rate": 9.818417412804937e-07, + "loss": 0.7596, + "num_input_tokens_seen": 245385260, + "step": 7539 + }, + { + "epoch": 0.6799837669657753, + "flos": 22242164904000.0, + "grad_norm": 1.7800884140216207, + "learning_rate": 9.813389431154463e-07, + "loss": 0.8291, + "num_input_tokens_seen": 245412950, + "step": 7540 + }, + { + "epoch": 0.6800739504892456, + "flos": 23808187588800.0, + "grad_norm": 2.218935246860697, + "learning_rate": 9.808362318681783e-07, + "loss": 0.7748, + "num_input_tokens_seen": 245439675, + "step": 7541 + }, + { + "epoch": 0.6801641340127159, + "flos": 16739171066880.0, + "grad_norm": 2.227040771515408, + "learning_rate": 9.803336075815807e-07, + "loss": 0.7689, + "num_input_tokens_seen": 245467570, + "step": 7542 + }, + { + "epoch": 0.6802543175361861, + "flos": 28070928950400.0, + "grad_norm": 1.906022542922252, + "learning_rate": 9.79831070298544e-07, + "loss": 0.6762, + "num_input_tokens_seen": 245499125, + "step": 7543 + }, + { + "epoch": 0.6803445010596564, + "flos": 30729542889600.0, + "grad_norm": 2.3478191401736677, + "learning_rate": 9.793286200619443e-07, + "loss": 0.7161, + "num_input_tokens_seen": 245527045, + "step": 7544 + }, + { + "epoch": 0.6804346845831266, + "flos": 22857682091520.0, + "grad_norm": 1.8256013193132412, + "learning_rate": 9.78826256914655e-07, + "loss": 0.8089, + "num_input_tokens_seen": 245556975, + "step": 7545 + }, + { + "epoch": 0.680524868106597, + "flos": 20529335902560.0, + "grad_norm": 1.7480182325691123, + "learning_rate": 9.7832398089954e-07, + "loss": 0.7936, + "num_input_tokens_seen": 245584420, + "step": 7546 + }, + { + "epoch": 0.6806150516300672, + "flos": 19721994085440.0, + "grad_norm": 2.0279323774421147, + "learning_rate": 9.778217920594565e-07, + "loss": 0.8521, + "num_input_tokens_seen": 245609085, + "step": 7547 + }, + { + "epoch": 0.6807052351535374, + "flos": 19909358343840.0, + "grad_norm": 2.027098418516489, + "learning_rate": 9.773196904372547e-07, + "loss": 0.7893, + "num_input_tokens_seen": 245637200, + "step": 7548 + }, + { + "epoch": 0.6807954186770077, + "flos": 21877323543360.0, + "grad_norm": 1.9969607745211713, + "learning_rate": 9.768176760757742e-07, + "loss": 0.8312, + "num_input_tokens_seen": 245664105, + "step": 7549 + }, + { + "epoch": 0.680885602200478, + "flos": 19618644951840.0, + "grad_norm": 1.8928804421854548, + "learning_rate": 9.76315749017853e-07, + "loss": 0.7285, + "num_input_tokens_seen": 245692905, + "step": 7550 + }, + { + "epoch": 0.6809757857239482, + "flos": 21069424179840.0, + "grad_norm": 1.90221998372174, + "learning_rate": 9.758139093063161e-07, + "loss": 0.751, + "num_input_tokens_seen": 245719735, + "step": 7551 + }, + { + "epoch": 0.6810659692474185, + "flos": 23333232198240.0, + "grad_norm": 2.3498773461656897, + "learning_rate": 9.753121569839834e-07, + "loss": 0.8133, + "num_input_tokens_seen": 245750445, + "step": 7552 + }, + { + "epoch": 0.6811561527708888, + "flos": 58054874727840.0, + "grad_norm": 0.7360829316949355, + "learning_rate": 9.748104920936678e-07, + "loss": 0.548, + "num_input_tokens_seen": 245825130, + "step": 7553 + }, + { + "epoch": 0.681246336294359, + "flos": 29234563083360.0, + "grad_norm": 1.9287909150743858, + "learning_rate": 9.743089146781738e-07, + "loss": 0.7257, + "num_input_tokens_seen": 245854885, + "step": 7554 + }, + { + "epoch": 0.6813365198178293, + "flos": 22861547746560.0, + "grad_norm": 1.6532245569952229, + "learning_rate": 9.738074247802988e-07, + "loss": 0.6733, + "num_input_tokens_seen": 245884235, + "step": 7555 + }, + { + "epoch": 0.6814267033412995, + "flos": 26065416559680.0, + "grad_norm": 2.0568682915002365, + "learning_rate": 9.733060224428325e-07, + "loss": 0.766, + "num_input_tokens_seen": 245911700, + "step": 7556 + }, + { + "epoch": 0.6815168868647699, + "flos": 18489547259520.0, + "grad_norm": 2.1398982464138063, + "learning_rate": 9.728047077085577e-07, + "loss": 0.753, + "num_input_tokens_seen": 245938105, + "step": 7557 + }, + { + "epoch": 0.6816070703882401, + "flos": 24063398126400.0, + "grad_norm": 2.99457695675071, + "learning_rate": 9.723034806202497e-07, + "loss": 0.6545, + "num_input_tokens_seen": 245964915, + "step": 7558 + }, + { + "epoch": 0.6816972539117103, + "flos": 17432347350240.0, + "grad_norm": 3.5154361928467086, + "learning_rate": 9.718023412206748e-07, + "loss": 0.7688, + "num_input_tokens_seen": 245991075, + "step": 7559 + }, + { + "epoch": 0.6817874374351806, + "flos": 25153275988320.0, + "grad_norm": 1.663335793722062, + "learning_rate": 9.713012895525935e-07, + "loss": 0.7689, + "num_input_tokens_seen": 246022440, + "step": 7560 + }, + { + "epoch": 0.6818776209586509, + "flos": 20092559589120.0, + "grad_norm": 2.0871452806808772, + "learning_rate": 9.708003256587584e-07, + "loss": 0.7365, + "num_input_tokens_seen": 246050270, + "step": 7561 + }, + { + "epoch": 0.6819678044821211, + "flos": 29673941280000.0, + "grad_norm": 1.9904836511656812, + "learning_rate": 9.702994495819147e-07, + "loss": 0.7964, + "num_input_tokens_seen": 246081800, + "step": 7562 + }, + { + "epoch": 0.6820579880055914, + "flos": 13743970518240.0, + "grad_norm": 2.542627830330136, + "learning_rate": 9.697986613647999e-07, + "loss": 0.6515, + "num_input_tokens_seen": 246103995, + "step": 7563 + }, + { + "epoch": 0.6821481715290616, + "flos": 63703411109760.0, + "grad_norm": 0.6973872786536859, + "learning_rate": 9.692979610501425e-07, + "loss": 0.6125, + "num_input_tokens_seen": 246192515, + "step": 7564 + }, + { + "epoch": 0.6822383550525319, + "flos": 22672659528000.0, + "grad_norm": 1.850101006817697, + "learning_rate": 9.68797348680668e-07, + "loss": 0.8205, + "num_input_tokens_seen": 246218545, + "step": 7565 + }, + { + "epoch": 0.6823285385760022, + "flos": 22239897548640.0, + "grad_norm": 1.6358051527060393, + "learning_rate": 9.682968242990878e-07, + "loss": 0.717, + "num_input_tokens_seen": 246248645, + "step": 7566 + }, + { + "epoch": 0.6824187220994724, + "flos": 28504471494720.0, + "grad_norm": 1.844321275560069, + "learning_rate": 9.677963879481132e-07, + "loss": 0.7975, + "num_input_tokens_seen": 246277815, + "step": 7567 + }, + { + "epoch": 0.6825089056229426, + "flos": 16667087435040.0, + "grad_norm": 2.23049044890797, + "learning_rate": 9.672960396704416e-07, + "loss": 0.7378, + "num_input_tokens_seen": 246303630, + "step": 7568 + }, + { + "epoch": 0.682599089146413, + "flos": 21253852027200.0, + "grad_norm": 1.9850834460459095, + "learning_rate": 9.667957795087657e-07, + "loss": 0.7522, + "num_input_tokens_seen": 246331810, + "step": 7569 + }, + { + "epoch": 0.6826892726698832, + "flos": 19764447684960.0, + "grad_norm": 2.495569690499378, + "learning_rate": 9.662956075057712e-07, + "loss": 0.8094, + "num_input_tokens_seen": 246359895, + "step": 7570 + }, + { + "epoch": 0.6827794561933535, + "flos": 20559337632480.0, + "grad_norm": 4.068183796244266, + "learning_rate": 9.657955237041354e-07, + "loss": 0.8156, + "num_input_tokens_seen": 246384480, + "step": 7571 + }, + { + "epoch": 0.6828696397168237, + "flos": 21214297668960.0, + "grad_norm": 1.748260587276442, + "learning_rate": 9.652955281465278e-07, + "loss": 0.726, + "num_input_tokens_seen": 246412820, + "step": 7572 + }, + { + "epoch": 0.682959823240294, + "flos": 17504951358720.0, + "grad_norm": 2.1142458591607824, + "learning_rate": 9.64795620875612e-07, + "loss": 0.7541, + "num_input_tokens_seen": 246440500, + "step": 7573 + }, + { + "epoch": 0.6830500067637643, + "flos": 28356921782880.0, + "grad_norm": 2.164732460952725, + "learning_rate": 9.64295801934041e-07, + "loss": 0.6944, + "num_input_tokens_seen": 246471675, + "step": 7574 + }, + { + "epoch": 0.6831401902872345, + "flos": 65082850101120.0, + "grad_norm": 0.6690801999611874, + "learning_rate": 9.63796071364463e-07, + "loss": 0.5866, + "num_input_tokens_seen": 246554080, + "step": 7575 + }, + { + "epoch": 0.6832303738107048, + "flos": 19108484064960.0, + "grad_norm": 1.9672020776040369, + "learning_rate": 9.632964292095179e-07, + "loss": 0.8345, + "num_input_tokens_seen": 246581630, + "step": 7576 + }, + { + "epoch": 0.6833205573341751, + "flos": 21440175532320.0, + "grad_norm": 2.286667127062396, + "learning_rate": 9.627968755118374e-07, + "loss": 0.7361, + "num_input_tokens_seen": 246610060, + "step": 7577 + }, + { + "epoch": 0.6834107408576453, + "flos": 56167839563040.0, + "grad_norm": 0.6677770506857864, + "learning_rate": 9.622974103140468e-07, + "loss": 0.5662, + "num_input_tokens_seen": 246691495, + "step": 7578 + }, + { + "epoch": 0.6835009243811155, + "flos": 24241915981920.0, + "grad_norm": 1.6028128604512917, + "learning_rate": 9.617980336587632e-07, + "loss": 0.8113, + "num_input_tokens_seen": 246721220, + "step": 7579 + }, + { + "epoch": 0.6835911079045859, + "flos": 26612715770400.0, + "grad_norm": 2.27708891658209, + "learning_rate": 9.612987455885964e-07, + "loss": 0.629, + "num_input_tokens_seen": 246752900, + "step": 7580 + }, + { + "epoch": 0.6836812914280561, + "flos": 30292989594720.0, + "grad_norm": 1.7760437623251206, + "learning_rate": 9.607995461461467e-07, + "loss": 0.7541, + "num_input_tokens_seen": 246782490, + "step": 7581 + }, + { + "epoch": 0.6837714749515263, + "flos": 40347914225280.0, + "grad_norm": 1.9117603392572722, + "learning_rate": 9.603004353740111e-07, + "loss": 0.6501, + "num_input_tokens_seen": 246815795, + "step": 7582 + }, + { + "epoch": 0.6838616584749966, + "flos": 62886330815520.0, + "grad_norm": 0.7629538279895786, + "learning_rate": 9.598014133147738e-07, + "loss": 0.6217, + "num_input_tokens_seen": 246895975, + "step": 7583 + }, + { + "epoch": 0.6839518419984669, + "flos": 26797589654880.0, + "grad_norm": 1.5415375457853493, + "learning_rate": 9.59302480011017e-07, + "loss": 0.7089, + "num_input_tokens_seen": 246929010, + "step": 7584 + }, + { + "epoch": 0.6840420255219372, + "flos": 63809362126560.0, + "grad_norm": 0.7029255054176702, + "learning_rate": 9.588036355053102e-07, + "loss": 0.5981, + "num_input_tokens_seen": 247011820, + "step": 7585 + }, + { + "epoch": 0.6841322090454074, + "flos": 20492755125120.0, + "grad_norm": 19.82341015900219, + "learning_rate": 9.583048798402182e-07, + "loss": 0.6152, + "num_input_tokens_seen": 247039690, + "step": 7586 + }, + { + "epoch": 0.6842223925688776, + "flos": 19327522692480.0, + "grad_norm": 1.7292730536154697, + "learning_rate": 9.57806213058298e-07, + "loss": 0.7413, + "num_input_tokens_seen": 247068345, + "step": 7587 + }, + { + "epoch": 0.684312576092348, + "flos": 60470283961920.0, + "grad_norm": 0.6886737946793517, + "learning_rate": 9.57307635202098e-07, + "loss": 0.6156, + "num_input_tokens_seen": 247162105, + "step": 7588 + }, + { + "epoch": 0.6844027596158182, + "flos": 19035248170560.0, + "grad_norm": 2.0630982427240006, + "learning_rate": 9.568091463141607e-07, + "loss": 0.8518, + "num_input_tokens_seen": 247188995, + "step": 7589 + }, + { + "epoch": 0.6844929431392884, + "flos": 22278299644320.0, + "grad_norm": 1.6909021062324479, + "learning_rate": 9.563107464370187e-07, + "loss": 0.7116, + "num_input_tokens_seen": 247216900, + "step": 7590 + }, + { + "epoch": 0.6845831266627587, + "flos": 19758091656000.0, + "grad_norm": 2.045322674054981, + "learning_rate": 9.558124356131982e-07, + "loss": 0.7662, + "num_input_tokens_seen": 247244260, + "step": 7591 + }, + { + "epoch": 0.684673310186229, + "flos": 23990310911040.0, + "grad_norm": 8.125381695511535, + "learning_rate": 9.553142138852187e-07, + "loss": 0.6803, + "num_input_tokens_seen": 247272185, + "step": 7592 + }, + { + "epoch": 0.6847634937096992, + "flos": 24682149083040.0, + "grad_norm": 7.522404077843171, + "learning_rate": 9.548160812955905e-07, + "loss": 0.6563, + "num_input_tokens_seen": 247302135, + "step": 7593 + }, + { + "epoch": 0.6848536772331695, + "flos": 27269422785600.0, + "grad_norm": 1.8855260392249524, + "learning_rate": 9.543180378868175e-07, + "loss": 0.5696, + "num_input_tokens_seen": 247332525, + "step": 7594 + }, + { + "epoch": 0.6849438607566397, + "flos": 23909975592480.0, + "grad_norm": 12.490612930395095, + "learning_rate": 9.538200837013962e-07, + "loss": 0.784, + "num_input_tokens_seen": 247359800, + "step": 7595 + }, + { + "epoch": 0.68503404428011, + "flos": 64395695318880.0, + "grad_norm": 0.6592974748279868, + "learning_rate": 9.533222187818122e-07, + "loss": 0.6132, + "num_input_tokens_seen": 247449695, + "step": 7596 + }, + { + "epoch": 0.6851242278035803, + "flos": 16849693964160.0, + "grad_norm": 2.436404935902124, + "learning_rate": 9.528244431705492e-07, + "loss": 0.7568, + "num_input_tokens_seen": 247475495, + "step": 7597 + }, + { + "epoch": 0.6852144113270505, + "flos": 23768967758400.0, + "grad_norm": 1.8280400313495409, + "learning_rate": 9.523267569100774e-07, + "loss": 0.7558, + "num_input_tokens_seen": 247505000, + "step": 7598 + }, + { + "epoch": 0.6853045948505209, + "flos": 24786464630400.0, + "grad_norm": 1.840916541541507, + "learning_rate": 9.518291600428652e-07, + "loss": 0.7017, + "num_input_tokens_seen": 247533200, + "step": 7599 + }, + { + "epoch": 0.6853947783739911, + "flos": 17833286281440.0, + "grad_norm": 2.08794143943222, + "learning_rate": 9.513316526113677e-07, + "loss": 0.7729, + "num_input_tokens_seen": 247560100, + "step": 7600 + }, + { + "epoch": 0.6854849618974613, + "flos": 21732078356640.0, + "grad_norm": 2.0967231252050262, + "learning_rate": 9.50834234658036e-07, + "loss": 0.7719, + "num_input_tokens_seen": 247586475, + "step": 7601 + }, + { + "epoch": 0.6855751454209316, + "flos": 26171070218400.0, + "grad_norm": 1.6827751700990263, + "learning_rate": 9.503369062253123e-07, + "loss": 0.762, + "num_input_tokens_seen": 247616170, + "step": 7602 + }, + { + "epoch": 0.6856653289444019, + "flos": 66368046550080.0, + "grad_norm": 0.6370317356836649, + "learning_rate": 9.498396673556317e-07, + "loss": 0.5893, + "num_input_tokens_seen": 247705415, + "step": 7603 + }, + { + "epoch": 0.6857555124678721, + "flos": 24969182668800.0, + "grad_norm": 1.9377659645884628, + "learning_rate": 9.493425180914219e-07, + "loss": 0.7828, + "num_input_tokens_seen": 247735405, + "step": 7604 + }, + { + "epoch": 0.6858456959913424, + "flos": 25702136328960.0, + "grad_norm": 1.826886662417605, + "learning_rate": 9.488454584751e-07, + "loss": 0.6786, + "num_input_tokens_seen": 247765840, + "step": 7605 + }, + { + "epoch": 0.6859358795148126, + "flos": 48109029107520.0, + "grad_norm": 1.7292427904156042, + "learning_rate": 9.483484885490813e-07, + "loss": 0.7516, + "num_input_tokens_seen": 247798695, + "step": 7606 + }, + { + "epoch": 0.686026063038283, + "flos": 25193016195360.0, + "grad_norm": 1.9131607111152935, + "learning_rate": 9.478516083557675e-07, + "loss": 0.7609, + "num_input_tokens_seen": 247824835, + "step": 7607 + }, + { + "epoch": 0.6861162465617532, + "flos": 27084548901120.0, + "grad_norm": 2.398254586368002, + "learning_rate": 9.473548179375561e-07, + "loss": 0.7652, + "num_input_tokens_seen": 247852690, + "step": 7608 + }, + { + "epoch": 0.6862064300852234, + "flos": 60742858511040.0, + "grad_norm": 0.6901738665428669, + "learning_rate": 9.468581173368358e-07, + "loss": 0.6164, + "num_input_tokens_seen": 247936530, + "step": 7609 + }, + { + "epoch": 0.6862966136086937, + "flos": 25407222754080.0, + "grad_norm": 1.5802895082744042, + "learning_rate": 9.463615065959878e-07, + "loss": 0.7553, + "num_input_tokens_seen": 247967290, + "step": 7610 + }, + { + "epoch": 0.686386797132164, + "flos": 22636078750560.0, + "grad_norm": 4.141684572934613, + "learning_rate": 9.458649857573857e-07, + "loss": 0.8046, + "num_input_tokens_seen": 247995235, + "step": 7611 + }, + { + "epoch": 0.6864769806556342, + "flos": 24719064388320.0, + "grad_norm": 1.7913058013636374, + "learning_rate": 9.453685548633963e-07, + "loss": 0.7589, + "num_input_tokens_seen": 248023160, + "step": 7612 + }, + { + "epoch": 0.6865671641791045, + "flos": 39365213982240.0, + "grad_norm": 1.9337714836046715, + "learning_rate": 9.448722139563756e-07, + "loss": 0.6967, + "num_input_tokens_seen": 248054895, + "step": 7613 + }, + { + "epoch": 0.6866573477025747, + "flos": 29965249388160.0, + "grad_norm": 1.7260226963668273, + "learning_rate": 9.443759630786769e-07, + "loss": 0.6429, + "num_input_tokens_seen": 248087815, + "step": 7614 + }, + { + "epoch": 0.686747531226045, + "flos": 23442305474880.0, + "grad_norm": 3.339950113666338, + "learning_rate": 9.438798022726408e-07, + "loss": 0.8216, + "num_input_tokens_seen": 248115170, + "step": 7615 + }, + { + "epoch": 0.6868377147495153, + "flos": 20346506354880.0, + "grad_norm": 1.7835637508213769, + "learning_rate": 9.433837315806037e-07, + "loss": 0.7008, + "num_input_tokens_seen": 248141340, + "step": 7616 + }, + { + "epoch": 0.6869278982729855, + "flos": 32114520175200.0, + "grad_norm": 3.3553078804932923, + "learning_rate": 9.428877510448925e-07, + "loss": 0.7052, + "num_input_tokens_seen": 248168170, + "step": 7617 + }, + { + "epoch": 0.6870180817964557, + "flos": 21473522540640.0, + "grad_norm": 1.776612183640245, + "learning_rate": 9.423918607078272e-07, + "loss": 0.6767, + "num_input_tokens_seen": 248197185, + "step": 7618 + }, + { + "epoch": 0.6871082653199261, + "flos": 21256491080160.0, + "grad_norm": 2.0564756553687347, + "learning_rate": 9.418960606117208e-07, + "loss": 0.76, + "num_input_tokens_seen": 248224545, + "step": 7619 + }, + { + "epoch": 0.6871984488433963, + "flos": 23297580664800.0, + "grad_norm": 2.702306722810256, + "learning_rate": 9.414003507988752e-07, + "loss": 0.6905, + "num_input_tokens_seen": 248252685, + "step": 7620 + }, + { + "epoch": 0.6872886323668665, + "flos": 14481124361280.0, + "grad_norm": 2.678939581502642, + "learning_rate": 9.409047313115904e-07, + "loss": 0.6832, + "num_input_tokens_seen": 248279390, + "step": 7621 + }, + { + "epoch": 0.6873788158903368, + "flos": 25482540155040.0, + "grad_norm": 2.671941031370653, + "learning_rate": 9.404092021921521e-07, + "loss": 0.6863, + "num_input_tokens_seen": 248309095, + "step": 7622 + }, + { + "epoch": 0.6874689994138071, + "flos": 23807704381920.0, + "grad_norm": 2.0208063287667595, + "learning_rate": 9.399137634828447e-07, + "loss": 0.73, + "num_input_tokens_seen": 248337845, + "step": 7623 + }, + { + "epoch": 0.6875591829372774, + "flos": 20129066027040.0, + "grad_norm": 1.7589342395229175, + "learning_rate": 9.394184152259396e-07, + "loss": 0.7699, + "num_input_tokens_seen": 248363840, + "step": 7624 + }, + { + "epoch": 0.6876493664607476, + "flos": 23661641460480.0, + "grad_norm": 2.832277568760072, + "learning_rate": 9.389231574637033e-07, + "loss": 0.8516, + "num_input_tokens_seen": 248394805, + "step": 7625 + }, + { + "epoch": 0.6877395499842179, + "flos": 19799578841760.0, + "grad_norm": 2.069668536530605, + "learning_rate": 9.384279902383938e-07, + "loss": 0.8366, + "num_input_tokens_seen": 248419525, + "step": 7626 + }, + { + "epoch": 0.6878297335076882, + "flos": 29814688925760.0, + "grad_norm": 1.8951716333734097, + "learning_rate": 9.379329135922615e-07, + "loss": 0.7651, + "num_input_tokens_seen": 248448620, + "step": 7627 + }, + { + "epoch": 0.6879199170311584, + "flos": 36340792268640.0, + "grad_norm": 2.8998511637937145, + "learning_rate": 9.374379275675495e-07, + "loss": 0.7047, + "num_input_tokens_seen": 248480400, + "step": 7628 + }, + { + "epoch": 0.6880101005546286, + "flos": 24422292325440.0, + "grad_norm": 2.044671306543068, + "learning_rate": 9.369430322064931e-07, + "loss": 0.7468, + "num_input_tokens_seen": 248508395, + "step": 7629 + }, + { + "epoch": 0.688100284078099, + "flos": 19545483396960.0, + "grad_norm": 3.150076692646651, + "learning_rate": 9.364482275513179e-07, + "loss": 0.8262, + "num_input_tokens_seen": 248535155, + "step": 7630 + }, + { + "epoch": 0.6881904676015692, + "flos": 20201484186720.0, + "grad_norm": 1.688735156668169, + "learning_rate": 9.359535136442444e-07, + "loss": 0.8147, + "num_input_tokens_seen": 248562285, + "step": 7631 + }, + { + "epoch": 0.6882806511250394, + "flos": 23733390564480.0, + "grad_norm": 2.2126734089110287, + "learning_rate": 9.354588905274843e-07, + "loss": 0.6836, + "num_input_tokens_seen": 248591190, + "step": 7632 + }, + { + "epoch": 0.6883708346485097, + "flos": 23443271888640.0, + "grad_norm": 2.0480637920174987, + "learning_rate": 9.349643582432414e-07, + "loss": 0.7804, + "num_input_tokens_seen": 248620035, + "step": 7633 + }, + { + "epoch": 0.68846101817198, + "flos": 23626250115360.0, + "grad_norm": 1.600183851884765, + "learning_rate": 9.344699168337127e-07, + "loss": 0.7379, + "num_input_tokens_seen": 248647180, + "step": 7634 + }, + { + "epoch": 0.6885512016954503, + "flos": 19181682789600.0, + "grad_norm": 1.8894534303650323, + "learning_rate": 9.339755663410845e-07, + "loss": 0.7094, + "num_input_tokens_seen": 248675000, + "step": 7635 + }, + { + "epoch": 0.6886413852189205, + "flos": 19217966208960.0, + "grad_norm": 1.8471574493116734, + "learning_rate": 9.334813068075405e-07, + "loss": 0.7802, + "num_input_tokens_seen": 248703170, + "step": 7636 + }, + { + "epoch": 0.6887315687423907, + "flos": 24391361351520.0, + "grad_norm": 1.57867709030005, + "learning_rate": 9.329871382752506e-07, + "loss": 0.7515, + "num_input_tokens_seen": 248731885, + "step": 7637 + }, + { + "epoch": 0.6888217522658611, + "flos": 25191009028320.0, + "grad_norm": 1.890878770679359, + "learning_rate": 9.32493060786383e-07, + "loss": 0.7379, + "num_input_tokens_seen": 248761040, + "step": 7638 + }, + { + "epoch": 0.6889119357893313, + "flos": 34333384220160.0, + "grad_norm": 1.5558357024349299, + "learning_rate": 9.31999074383093e-07, + "loss": 0.8066, + "num_input_tokens_seen": 248793120, + "step": 7639 + }, + { + "epoch": 0.6890021193128015, + "flos": 24718767030240.0, + "grad_norm": 3.141934705341114, + "learning_rate": 9.315051791075308e-07, + "loss": 0.7397, + "num_input_tokens_seen": 248821120, + "step": 7640 + }, + { + "epoch": 0.6890923028362718, + "flos": 24497089349760.0, + "grad_norm": 1.5986099194918928, + "learning_rate": 9.310113750018382e-07, + "loss": 0.7405, + "num_input_tokens_seen": 248850545, + "step": 7641 + }, + { + "epoch": 0.6891824863597421, + "flos": 16521210362400.0, + "grad_norm": 2.225315534527733, + "learning_rate": 9.305176621081496e-07, + "loss": 0.8146, + "num_input_tokens_seen": 248876710, + "step": 7642 + }, + { + "epoch": 0.6892726698832123, + "flos": 26030619930720.0, + "grad_norm": 2.426205903506988, + "learning_rate": 9.300240404685911e-07, + "loss": 0.7819, + "num_input_tokens_seen": 248902805, + "step": 7643 + }, + { + "epoch": 0.6893628534066826, + "flos": 24100053243360.0, + "grad_norm": 1.8548671028055985, + "learning_rate": 9.295305101252812e-07, + "loss": 0.7492, + "num_input_tokens_seen": 248930955, + "step": 7644 + }, + { + "epoch": 0.6894530369301528, + "flos": 22495962990720.0, + "grad_norm": 2.0060529903285467, + "learning_rate": 9.290370711203314e-07, + "loss": 0.7289, + "num_input_tokens_seen": 248956975, + "step": 7645 + }, + { + "epoch": 0.6895432204536232, + "flos": 20559858009120.0, + "grad_norm": 1.7556667540471567, + "learning_rate": 9.285437234958433e-07, + "loss": 0.7434, + "num_input_tokens_seen": 248984040, + "step": 7646 + }, + { + "epoch": 0.6896334039770934, + "flos": 27811146532320.0, + "grad_norm": 2.1497761942570413, + "learning_rate": 9.280504672939124e-07, + "loss": 0.7378, + "num_input_tokens_seen": 249011100, + "step": 7647 + }, + { + "epoch": 0.6897235875005636, + "flos": 25335696668640.0, + "grad_norm": 1.8636978931165993, + "learning_rate": 9.275573025566266e-07, + "loss": 0.788, + "num_input_tokens_seen": 249040900, + "step": 7648 + }, + { + "epoch": 0.689813771024034, + "flos": 27995500040160.0, + "grad_norm": 1.6580110939722283, + "learning_rate": 9.27064229326065e-07, + "loss": 0.7823, + "num_input_tokens_seen": 249071880, + "step": 7649 + }, + { + "epoch": 0.6899039545475042, + "flos": 24209052180480.0, + "grad_norm": 2.4519326780887845, + "learning_rate": 9.265712476442995e-07, + "loss": 0.7172, + "num_input_tokens_seen": 249099730, + "step": 7650 + }, + { + "epoch": 0.6899941380709744, + "flos": 24683264175840.0, + "grad_norm": 1.60969281067784, + "learning_rate": 9.260783575533949e-07, + "loss": 0.6765, + "num_input_tokens_seen": 249129495, + "step": 7651 + }, + { + "epoch": 0.6900843215944447, + "flos": 21804533686080.0, + "grad_norm": 2.2073202028420695, + "learning_rate": 9.255855590954045e-07, + "loss": 0.6972, + "num_input_tokens_seen": 249158030, + "step": 7652 + }, + { + "epoch": 0.690174505117915, + "flos": 22934226094560.0, + "grad_norm": 2.247511122933807, + "learning_rate": 9.250928523123802e-07, + "loss": 0.8168, + "num_input_tokens_seen": 249185785, + "step": 7653 + }, + { + "epoch": 0.6902646886413852, + "flos": 26577250085760.0, + "grad_norm": 10.949561917673458, + "learning_rate": 9.24600237246359e-07, + "loss": 0.7578, + "num_input_tokens_seen": 249217495, + "step": 7654 + }, + { + "epoch": 0.6903548721648555, + "flos": 20784620779680.0, + "grad_norm": 2.2549095631330616, + "learning_rate": 9.241077139393769e-07, + "loss": 0.755, + "num_input_tokens_seen": 249245150, + "step": 7655 + }, + { + "epoch": 0.6904450556883257, + "flos": 22387038393120.0, + "grad_norm": 2.1644906952768785, + "learning_rate": 9.236152824334564e-07, + "loss": 0.7798, + "num_input_tokens_seen": 249274770, + "step": 7656 + }, + { + "epoch": 0.690535239211796, + "flos": 26504014191360.0, + "grad_norm": 1.8104871556907054, + "learning_rate": 9.231229427706151e-07, + "loss": 0.6475, + "num_input_tokens_seen": 249303830, + "step": 7657 + }, + { + "epoch": 0.6906254227352663, + "flos": 23152372647840.0, + "grad_norm": 2.713082819612752, + "learning_rate": 9.226306949928622e-07, + "loss": 0.7548, + "num_input_tokens_seen": 249333285, + "step": 7658 + }, + { + "epoch": 0.6907156062587365, + "flos": 16593145315200.0, + "grad_norm": 2.6308908212845488, + "learning_rate": 9.221385391421988e-07, + "loss": 0.8225, + "num_input_tokens_seen": 249358285, + "step": 7659 + }, + { + "epoch": 0.6908057897822067, + "flos": 23589520658880.0, + "grad_norm": 1.8569376381472193, + "learning_rate": 9.216464752606192e-07, + "loss": 0.7639, + "num_input_tokens_seen": 249388335, + "step": 7660 + }, + { + "epoch": 0.6908959733056771, + "flos": 17755441336800.0, + "grad_norm": 1.6105636355249418, + "learning_rate": 9.211545033901078e-07, + "loss": 0.8283, + "num_input_tokens_seen": 249414815, + "step": 7661 + }, + { + "epoch": 0.6909861568291473, + "flos": 21578916011040.0, + "grad_norm": 2.3850123522296665, + "learning_rate": 9.206626235726426e-07, + "loss": 0.7761, + "num_input_tokens_seen": 249441955, + "step": 7662 + }, + { + "epoch": 0.6910763403526176, + "flos": 27050235479040.0, + "grad_norm": 4.086140556468512, + "learning_rate": 9.20170835850194e-07, + "loss": 0.686, + "num_input_tokens_seen": 249469385, + "step": 7663 + }, + { + "epoch": 0.6911665238760878, + "flos": 68590293043200.0, + "grad_norm": 0.6372196123189265, + "learning_rate": 9.196791402647237e-07, + "loss": 0.5726, + "num_input_tokens_seen": 249565480, + "step": 7664 + }, + { + "epoch": 0.6912567073995581, + "flos": 25191083367840.0, + "grad_norm": 1.7816478887174474, + "learning_rate": 9.191875368581861e-07, + "loss": 0.672, + "num_input_tokens_seen": 249596255, + "step": 7665 + }, + { + "epoch": 0.6913468909230284, + "flos": 31457032595040.0, + "grad_norm": 2.2600820150094316, + "learning_rate": 9.186960256725271e-07, + "loss": 0.7575, + "num_input_tokens_seen": 249626345, + "step": 7666 + }, + { + "epoch": 0.6914370744464986, + "flos": 22746899005920.0, + "grad_norm": 1.7254450016724343, + "learning_rate": 9.182046067496856e-07, + "loss": 0.8121, + "num_input_tokens_seen": 249655115, + "step": 7667 + }, + { + "epoch": 0.6915272579699688, + "flos": 26867926308000.0, + "grad_norm": 2.1946113525633755, + "learning_rate": 9.177132801315927e-07, + "loss": 0.834, + "num_input_tokens_seen": 249683675, + "step": 7668 + }, + { + "epoch": 0.6916174414934392, + "flos": 19139229190080.0, + "grad_norm": 1.6751595435382785, + "learning_rate": 9.172220458601692e-07, + "loss": 0.7239, + "num_input_tokens_seen": 249708490, + "step": 7669 + }, + { + "epoch": 0.6917076250169094, + "flos": 38964200711520.0, + "grad_norm": 2.401394887720218, + "learning_rate": 9.167309039773324e-07, + "loss": 0.7161, + "num_input_tokens_seen": 249737660, + "step": 7670 + }, + { + "epoch": 0.6917978085403796, + "flos": 33132128556480.0, + "grad_norm": 1.430776629397255, + "learning_rate": 9.162398545249872e-07, + "loss": 0.6476, + "num_input_tokens_seen": 249770015, + "step": 7671 + }, + { + "epoch": 0.69188799206385, + "flos": 26539888743360.0, + "grad_norm": 1.4501879142803713, + "learning_rate": 9.157488975450334e-07, + "loss": 0.8027, + "num_input_tokens_seen": 249803500, + "step": 7672 + }, + { + "epoch": 0.6919781755873202, + "flos": 41946689202240.0, + "grad_norm": 1.6865088059630589, + "learning_rate": 9.15258033079362e-07, + "loss": 0.582, + "num_input_tokens_seen": 249836790, + "step": 7673 + }, + { + "epoch": 0.6920683591107905, + "flos": 22679461594080.0, + "grad_norm": 1.7986041166676041, + "learning_rate": 9.147672611698567e-07, + "loss": 0.7378, + "num_input_tokens_seen": 249867045, + "step": 7674 + }, + { + "epoch": 0.6921585426342607, + "flos": 17576477444160.0, + "grad_norm": 3.135777162293274, + "learning_rate": 9.142765818583933e-07, + "loss": 0.7376, + "num_input_tokens_seen": 249893505, + "step": 7675 + }, + { + "epoch": 0.692248726157731, + "flos": 26103818655360.0, + "grad_norm": 1.6195478674941906, + "learning_rate": 9.13785995186837e-07, + "loss": 0.7266, + "num_input_tokens_seen": 249922725, + "step": 7676 + }, + { + "epoch": 0.6923389096812013, + "flos": 25010372496480.0, + "grad_norm": 1.8005087557354396, + "learning_rate": 9.132955011970502e-07, + "loss": 0.7139, + "num_input_tokens_seen": 249951710, + "step": 7677 + }, + { + "epoch": 0.6924290932046715, + "flos": 39614960565120.0, + "grad_norm": 1.3906817478216809, + "learning_rate": 9.128050999308827e-07, + "loss": 0.738, + "num_input_tokens_seen": 249983275, + "step": 7678 + }, + { + "epoch": 0.6925192767281417, + "flos": 24755310637920.0, + "grad_norm": 2.2909131169455796, + "learning_rate": 9.123147914301789e-07, + "loss": 0.8386, + "num_input_tokens_seen": 250009940, + "step": 7679 + }, + { + "epoch": 0.6926094602516121, + "flos": 16265776806240.0, + "grad_norm": 1.7671933255828975, + "learning_rate": 9.118245757367745e-07, + "loss": 0.7918, + "num_input_tokens_seen": 250036300, + "step": 7680 + }, + { + "epoch": 0.6926996437750823, + "flos": 65434384687680.0, + "grad_norm": 0.6222168227393834, + "learning_rate": 9.113344528924973e-07, + "loss": 0.5592, + "num_input_tokens_seen": 250118870, + "step": 7681 + }, + { + "epoch": 0.6927898272985525, + "flos": 38598318597600.0, + "grad_norm": 2.208657750705939, + "learning_rate": 9.108444229391676e-07, + "loss": 0.73, + "num_input_tokens_seen": 250150710, + "step": 7682 + }, + { + "epoch": 0.6928800108220228, + "flos": 18671261714400.0, + "grad_norm": 2.1614978342391775, + "learning_rate": 9.103544859185979e-07, + "loss": 0.7234, + "num_input_tokens_seen": 250177285, + "step": 7683 + }, + { + "epoch": 0.6929701943454931, + "flos": 23662793723040.0, + "grad_norm": 1.8472391508596688, + "learning_rate": 9.098646418725902e-07, + "loss": 0.8359, + "num_input_tokens_seen": 250204840, + "step": 7684 + }, + { + "epoch": 0.6930603778689634, + "flos": 64853329601280.0, + "grad_norm": 0.6363070040125346, + "learning_rate": 9.093748908429437e-07, + "loss": 0.6025, + "num_input_tokens_seen": 250293995, + "step": 7685 + }, + { + "epoch": 0.6931505613924336, + "flos": 20747259437280.0, + "grad_norm": 2.205968874562175, + "learning_rate": 9.088852328714444e-07, + "loss": 0.737, + "num_input_tokens_seen": 250321535, + "step": 7686 + }, + { + "epoch": 0.6932407449159038, + "flos": 35244930075360.0, + "grad_norm": 1.6383781670030222, + "learning_rate": 9.083956679998735e-07, + "loss": 0.6561, + "num_input_tokens_seen": 250351290, + "step": 7687 + }, + { + "epoch": 0.6933309284393742, + "flos": 19618830800640.0, + "grad_norm": 2.3733894509028093, + "learning_rate": 9.079061962700032e-07, + "loss": 0.7181, + "num_input_tokens_seen": 250381030, + "step": 7688 + }, + { + "epoch": 0.6934211119628444, + "flos": 25375436875680.0, + "grad_norm": 1.5545133560259656, + "learning_rate": 9.074168177235979e-07, + "loss": 0.7703, + "num_input_tokens_seen": 250411020, + "step": 7689 + }, + { + "epoch": 0.6935112954863146, + "flos": 23043002013120.0, + "grad_norm": 2.8032040724132514, + "learning_rate": 9.069275324024151e-07, + "loss": 0.7628, + "num_input_tokens_seen": 250439225, + "step": 7690 + }, + { + "epoch": 0.6936014790097849, + "flos": 23990125062240.0, + "grad_norm": 2.468591226228416, + "learning_rate": 9.064383403482005e-07, + "loss": 0.6814, + "num_input_tokens_seen": 250467550, + "step": 7691 + }, + { + "epoch": 0.6936916625332552, + "flos": 24973791719040.0, + "grad_norm": 2.068252748040977, + "learning_rate": 9.059492416026983e-07, + "loss": 0.7124, + "num_input_tokens_seen": 250497255, + "step": 7692 + }, + { + "epoch": 0.6937818460567254, + "flos": 34880460412320.0, + "grad_norm": 2.378778463605292, + "learning_rate": 9.054602362076378e-07, + "loss": 0.7452, + "num_input_tokens_seen": 250529560, + "step": 7693 + }, + { + "epoch": 0.6938720295801957, + "flos": 23223824393760.0, + "grad_norm": 2.5100924861632126, + "learning_rate": 9.049713242047468e-07, + "loss": 0.7387, + "num_input_tokens_seen": 250556575, + "step": 7694 + }, + { + "epoch": 0.693962213103666, + "flos": 18671113035360.0, + "grad_norm": 1.9044878585119949, + "learning_rate": 9.044825056357395e-07, + "loss": 0.7223, + "num_input_tokens_seen": 250584460, + "step": 7695 + }, + { + "epoch": 0.6940523966271362, + "flos": 21548988620640.0, + "grad_norm": 1.7831269798441638, + "learning_rate": 9.039937805423255e-07, + "loss": 0.6998, + "num_input_tokens_seen": 250610590, + "step": 7696 + }, + { + "epoch": 0.6941425801506065, + "flos": 69654741055680.0, + "grad_norm": 0.5926014180845248, + "learning_rate": 9.035051489662051e-07, + "loss": 0.5167, + "num_input_tokens_seen": 250703210, + "step": 7697 + }, + { + "epoch": 0.6942327636740767, + "flos": 61437229960320.0, + "grad_norm": 1.9666083055123562, + "learning_rate": 9.030166109490718e-07, + "loss": 0.6948, + "num_input_tokens_seen": 250736275, + "step": 7698 + }, + { + "epoch": 0.6943229471975471, + "flos": 14700051479520.0, + "grad_norm": 3.508232108594156, + "learning_rate": 9.025281665326099e-07, + "loss": 0.7511, + "num_input_tokens_seen": 250760280, + "step": 7699 + }, + { + "epoch": 0.6944131307210173, + "flos": 22096325001120.0, + "grad_norm": 2.7257828349193316, + "learning_rate": 9.020398157584967e-07, + "loss": 0.7555, + "num_input_tokens_seen": 250789635, + "step": 7700 + }, + { + "epoch": 0.6945033142444875, + "flos": 20237321568960.0, + "grad_norm": 3.1002935701232013, + "learning_rate": 9.015515586684002e-07, + "loss": 0.6503, + "num_input_tokens_seen": 250816675, + "step": 7701 + }, + { + "epoch": 0.6945934977679578, + "flos": 69819580439520.0, + "grad_norm": 0.6525135618751652, + "learning_rate": 9.010633953039812e-07, + "loss": 0.592, + "num_input_tokens_seen": 250908505, + "step": 7702 + }, + { + "epoch": 0.6946836812914281, + "flos": 26935252210560.0, + "grad_norm": 1.7362826741537942, + "learning_rate": 9.005753257068929e-07, + "loss": 0.7612, + "num_input_tokens_seen": 250935885, + "step": 7703 + }, + { + "epoch": 0.6947738648148983, + "flos": 19210383577920.0, + "grad_norm": 2.009461434553549, + "learning_rate": 9.000873499187797e-07, + "loss": 0.7645, + "num_input_tokens_seen": 250963310, + "step": 7704 + }, + { + "epoch": 0.6948640483383686, + "flos": 20930163324480.0, + "grad_norm": 1.8300314198074032, + "learning_rate": 8.995994679812797e-07, + "loss": 0.7803, + "num_input_tokens_seen": 250991485, + "step": 7705 + }, + { + "epoch": 0.6949542318618388, + "flos": 26503382305440.0, + "grad_norm": 1.6997592110365338, + "learning_rate": 8.991116799360192e-07, + "loss": 0.7689, + "num_input_tokens_seen": 251021390, + "step": 7706 + }, + { + "epoch": 0.6950444153853091, + "flos": 22711247472480.0, + "grad_norm": 2.5856476128584998, + "learning_rate": 8.986239858246217e-07, + "loss": 0.7399, + "num_input_tokens_seen": 251050655, + "step": 7707 + }, + { + "epoch": 0.6951345989087794, + "flos": 27159680453280.0, + "grad_norm": 2.27013490301232, + "learning_rate": 8.981363856886972e-07, + "loss": 0.8276, + "num_input_tokens_seen": 251077405, + "step": 7708 + }, + { + "epoch": 0.6952247824322496, + "flos": 29124746411520.0, + "grad_norm": 5.5031658077947325, + "learning_rate": 8.976488795698533e-07, + "loss": 0.7266, + "num_input_tokens_seen": 251108995, + "step": 7709 + }, + { + "epoch": 0.6953149659557198, + "flos": 22931326853280.0, + "grad_norm": 2.2144570526072065, + "learning_rate": 8.971614675096841e-07, + "loss": 0.7622, + "num_input_tokens_seen": 251137105, + "step": 7710 + }, + { + "epoch": 0.6954051494791902, + "flos": 23261780452320.0, + "grad_norm": 2.4297890595924145, + "learning_rate": 8.966741495497807e-07, + "loss": 0.8119, + "num_input_tokens_seen": 251163570, + "step": 7711 + }, + { + "epoch": 0.6954953330026604, + "flos": 26176943040480.0, + "grad_norm": 1.56553675237235, + "learning_rate": 8.961869257317218e-07, + "loss": 0.7122, + "num_input_tokens_seen": 251194120, + "step": 7712 + }, + { + "epoch": 0.6955855165261307, + "flos": 18669812093760.0, + "grad_norm": 1.8588810112246565, + "learning_rate": 8.956997960970809e-07, + "loss": 0.7298, + "num_input_tokens_seen": 251220400, + "step": 7713 + }, + { + "epoch": 0.6956757000496009, + "flos": 30438755158080.0, + "grad_norm": 1.878505798936046, + "learning_rate": 8.952127606874224e-07, + "loss": 0.6535, + "num_input_tokens_seen": 251250745, + "step": 7714 + }, + { + "epoch": 0.6957658835730712, + "flos": 20092113552000.0, + "grad_norm": 1.726396923842708, + "learning_rate": 8.947258195443028e-07, + "loss": 0.7499, + "num_input_tokens_seen": 251279170, + "step": 7715 + }, + { + "epoch": 0.6958560670965415, + "flos": 18263632226400.0, + "grad_norm": 1.8799625845493362, + "learning_rate": 8.942389727092716e-07, + "loss": 0.7237, + "num_input_tokens_seen": 251306745, + "step": 7716 + }, + { + "epoch": 0.6959462506200117, + "flos": 58740096682560.0, + "grad_norm": 0.6782881703053399, + "learning_rate": 8.937522202238677e-07, + "loss": 0.5637, + "num_input_tokens_seen": 251395855, + "step": 7717 + }, + { + "epoch": 0.696036434143482, + "flos": 23111703196800.0, + "grad_norm": 2.641306681929721, + "learning_rate": 8.932655621296239e-07, + "loss": 0.7653, + "num_input_tokens_seen": 251423355, + "step": 7718 + }, + { + "epoch": 0.6961266176669523, + "flos": 46901751942720.0, + "grad_norm": 1.5067213785033267, + "learning_rate": 8.927789984680649e-07, + "loss": 0.7417, + "num_input_tokens_seen": 251458915, + "step": 7719 + }, + { + "epoch": 0.6962168011904225, + "flos": 44827687047360.0, + "grad_norm": 1.603608996406073, + "learning_rate": 8.922925292807068e-07, + "loss": 0.6148, + "num_input_tokens_seen": 251495750, + "step": 7720 + }, + { + "epoch": 0.6963069847138927, + "flos": 22386741035040.0, + "grad_norm": 1.9995672830725906, + "learning_rate": 8.91806154609058e-07, + "loss": 0.8314, + "num_input_tokens_seen": 251524455, + "step": 7721 + }, + { + "epoch": 0.6963971682373631, + "flos": 21877360713120.0, + "grad_norm": 2.0840742955207925, + "learning_rate": 8.913198744946195e-07, + "loss": 0.7171, + "num_input_tokens_seen": 251551355, + "step": 7722 + }, + { + "epoch": 0.6964873517608333, + "flos": 23879156127840.0, + "grad_norm": 1.7838822708329203, + "learning_rate": 8.908336889788807e-07, + "loss": 0.7707, + "num_input_tokens_seen": 251580955, + "step": 7723 + }, + { + "epoch": 0.6965775352843036, + "flos": 28066840276800.0, + "grad_norm": 2.3250377077715996, + "learning_rate": 8.903475981033293e-07, + "loss": 0.628, + "num_input_tokens_seen": 251612215, + "step": 7724 + }, + { + "epoch": 0.6966677188077738, + "flos": 58556560909440.0, + "grad_norm": 0.7390559738548492, + "learning_rate": 8.898616019094376e-07, + "loss": 0.632, + "num_input_tokens_seen": 251691535, + "step": 7725 + }, + { + "epoch": 0.6967579023312441, + "flos": 20529224393280.0, + "grad_norm": 2.1601868225472733, + "learning_rate": 8.89375700438677e-07, + "loss": 0.783, + "num_input_tokens_seen": 251719425, + "step": 7726 + }, + { + "epoch": 0.6968480858547144, + "flos": 20783765875200.0, + "grad_norm": 1.9878335085140835, + "learning_rate": 8.888898937325047e-07, + "loss": 0.7114, + "num_input_tokens_seen": 251746865, + "step": 7727 + }, + { + "epoch": 0.6969382693781846, + "flos": 24281098642560.0, + "grad_norm": 3.2657409355354914, + "learning_rate": 8.884041818323733e-07, + "loss": 0.7125, + "num_input_tokens_seen": 251773360, + "step": 7728 + }, + { + "epoch": 0.6970284529016548, + "flos": 27305148658560.0, + "grad_norm": 1.7520153838357537, + "learning_rate": 8.879185647797262e-07, + "loss": 0.7825, + "num_input_tokens_seen": 251804325, + "step": 7729 + }, + { + "epoch": 0.6971186364251252, + "flos": 23261520264000.0, + "grad_norm": 1.8740504085474476, + "learning_rate": 8.874330426159993e-07, + "loss": 0.7267, + "num_input_tokens_seen": 251833885, + "step": 7730 + }, + { + "epoch": 0.6972088199485954, + "flos": 20343384095040.0, + "grad_norm": 1.759015769263233, + "learning_rate": 8.869476153826205e-07, + "loss": 0.7702, + "num_input_tokens_seen": 251861880, + "step": 7731 + }, + { + "epoch": 0.6972990034720656, + "flos": 20812169305440.0, + "grad_norm": 1.6901965319331471, + "learning_rate": 8.864622831210071e-07, + "loss": 0.7487, + "num_input_tokens_seen": 251888385, + "step": 7732 + }, + { + "epoch": 0.6973891869955359, + "flos": 24208494634080.0, + "grad_norm": 2.0595493595944023, + "learning_rate": 8.85977045872573e-07, + "loss": 0.7534, + "num_input_tokens_seen": 251914190, + "step": 7733 + }, + { + "epoch": 0.6974793705190062, + "flos": 26831977416480.0, + "grad_norm": 1.7344735607288546, + "learning_rate": 8.854919036787194e-07, + "loss": 0.7228, + "num_input_tokens_seen": 251944300, + "step": 7734 + }, + { + "epoch": 0.6975695540424764, + "flos": 21509285583360.0, + "grad_norm": 1.9421301860132416, + "learning_rate": 8.850068565808417e-07, + "loss": 0.7105, + "num_input_tokens_seen": 251974235, + "step": 7735 + }, + { + "epoch": 0.6976597375659467, + "flos": 26322002378400.0, + "grad_norm": 2.21124815359087, + "learning_rate": 8.845219046203271e-07, + "loss": 0.7968, + "num_input_tokens_seen": 252001175, + "step": 7736 + }, + { + "epoch": 0.6977499210894169, + "flos": 17541680815200.0, + "grad_norm": 2.8120221897731117, + "learning_rate": 8.840370478385544e-07, + "loss": 0.7746, + "num_input_tokens_seen": 252027475, + "step": 7737 + }, + { + "epoch": 0.6978401046128873, + "flos": 26828074591680.0, + "grad_norm": 4.396540556268649, + "learning_rate": 8.83552286276894e-07, + "loss": 0.7689, + "num_input_tokens_seen": 252059295, + "step": 7738 + }, + { + "epoch": 0.6979302881363575, + "flos": 18233444647680.0, + "grad_norm": 1.9014370302481172, + "learning_rate": 8.830676199767095e-07, + "loss": 0.7653, + "num_input_tokens_seen": 252085755, + "step": 7739 + }, + { + "epoch": 0.6980204716598277, + "flos": 24354817743840.0, + "grad_norm": 1.7379359213031378, + "learning_rate": 8.825830489793527e-07, + "loss": 0.7296, + "num_input_tokens_seen": 252115865, + "step": 7740 + }, + { + "epoch": 0.698110655183298, + "flos": 28724773894080.0, + "grad_norm": 1.7268862873803348, + "learning_rate": 8.820985733261732e-07, + "loss": 0.6973, + "num_input_tokens_seen": 252146915, + "step": 7741 + }, + { + "epoch": 0.6982008387067683, + "flos": 23660675046720.0, + "grad_norm": 3.145519030366523, + "learning_rate": 8.816141930585066e-07, + "loss": 0.657, + "num_input_tokens_seen": 252174240, + "step": 7742 + }, + { + "epoch": 0.6982910222302385, + "flos": 67206021983040.0, + "grad_norm": 0.6834460507639838, + "learning_rate": 8.811299082176837e-07, + "loss": 0.5886, + "num_input_tokens_seen": 252273790, + "step": 7743 + }, + { + "epoch": 0.6983812057537088, + "flos": 24425005717920.0, + "grad_norm": 2.3937897522018474, + "learning_rate": 8.806457188450265e-07, + "loss": 0.709, + "num_input_tokens_seen": 252304860, + "step": 7744 + }, + { + "epoch": 0.6984713892771791, + "flos": 22752214281600.0, + "grad_norm": 1.6687970215534027, + "learning_rate": 8.801616249818487e-07, + "loss": 0.7835, + "num_input_tokens_seen": 252333525, + "step": 7745 + }, + { + "epoch": 0.6985615728006493, + "flos": 20638520688480.0, + "grad_norm": 1.9095153981736903, + "learning_rate": 8.796776266694564e-07, + "loss": 0.7291, + "num_input_tokens_seen": 252361850, + "step": 7746 + }, + { + "epoch": 0.6986517563241196, + "flos": 33462842343840.0, + "grad_norm": 1.4232152972685301, + "learning_rate": 8.79193723949145e-07, + "loss": 0.692, + "num_input_tokens_seen": 252393000, + "step": 7747 + }, + { + "epoch": 0.6987419398475898, + "flos": 35605199555520.0, + "grad_norm": 1.9297017085366959, + "learning_rate": 8.787099168622063e-07, + "loss": 0.6438, + "num_input_tokens_seen": 252422925, + "step": 7748 + }, + { + "epoch": 0.6988321233710602, + "flos": 31746668064000.0, + "grad_norm": 2.1641444662387843, + "learning_rate": 8.782262054499199e-07, + "loss": 0.6877, + "num_input_tokens_seen": 252452090, + "step": 7749 + }, + { + "epoch": 0.6989223068945304, + "flos": 23261148566400.0, + "grad_norm": 1.4952327950180893, + "learning_rate": 8.777425897535588e-07, + "loss": 0.7755, + "num_input_tokens_seen": 252481365, + "step": 7750 + }, + { + "epoch": 0.6990124904180006, + "flos": 22751247867840.0, + "grad_norm": 2.0743543792462855, + "learning_rate": 8.77259069814388e-07, + "loss": 0.751, + "num_input_tokens_seen": 252509825, + "step": 7751 + }, + { + "epoch": 0.6991026739414709, + "flos": 28107732746400.0, + "grad_norm": 1.915913212612445, + "learning_rate": 8.767756456736641e-07, + "loss": 0.7462, + "num_input_tokens_seen": 252542500, + "step": 7752 + }, + { + "epoch": 0.6991928574649412, + "flos": 22823740367040.0, + "grad_norm": 1.7342553999620018, + "learning_rate": 8.762923173726358e-07, + "loss": 0.6711, + "num_input_tokens_seen": 252572830, + "step": 7753 + }, + { + "epoch": 0.6992830409884114, + "flos": 18304338847200.0, + "grad_norm": 1.6477374719286155, + "learning_rate": 8.758090849525428e-07, + "loss": 0.8405, + "num_input_tokens_seen": 252598430, + "step": 7754 + }, + { + "epoch": 0.6993732245118817, + "flos": 19654965540960.0, + "grad_norm": 2.0063868110570047, + "learning_rate": 8.753259484546174e-07, + "loss": 0.6842, + "num_input_tokens_seen": 252623470, + "step": 7755 + }, + { + "epoch": 0.6994634080353519, + "flos": 12113372493120.0, + "grad_norm": 2.0940466144040633, + "learning_rate": 8.748429079200841e-07, + "loss": 0.7028, + "num_input_tokens_seen": 252648745, + "step": 7756 + }, + { + "epoch": 0.6995535915588222, + "flos": 19581841155840.0, + "grad_norm": 2.5644206780780046, + "learning_rate": 8.743599633901575e-07, + "loss": 0.8025, + "num_input_tokens_seen": 252674335, + "step": 7757 + }, + { + "epoch": 0.6996437750822925, + "flos": 23838746865120.0, + "grad_norm": 1.8087916090897607, + "learning_rate": 8.738771149060453e-07, + "loss": 0.7637, + "num_input_tokens_seen": 252700420, + "step": 7758 + }, + { + "epoch": 0.6997339586057627, + "flos": 25225173771360.0, + "grad_norm": 1.696733011448735, + "learning_rate": 8.73394362508947e-07, + "loss": 0.7105, + "num_input_tokens_seen": 252728910, + "step": 7759 + }, + { + "epoch": 0.6998241421292329, + "flos": 53719683770400.0, + "grad_norm": 1.9927098804819496, + "learning_rate": 8.72911706240054e-07, + "loss": 0.709, + "num_input_tokens_seen": 252765070, + "step": 7760 + }, + { + "epoch": 0.6999143256527033, + "flos": 42673955889120.0, + "grad_norm": 2.001363527888964, + "learning_rate": 8.724291461405493e-07, + "loss": 0.6199, + "num_input_tokens_seen": 252798245, + "step": 7761 + }, + { + "epoch": 0.7000045091761735, + "flos": 25295064387360.0, + "grad_norm": 1.918816093772718, + "learning_rate": 8.71946682251606e-07, + "loss": 0.7117, + "num_input_tokens_seen": 252825975, + "step": 7762 + }, + { + "epoch": 0.7000946926996438, + "flos": 24750961776000.0, + "grad_norm": 2.62579324634993, + "learning_rate": 8.714643146143932e-07, + "loss": 0.7424, + "num_input_tokens_seen": 252854725, + "step": 7763 + }, + { + "epoch": 0.700184876223114, + "flos": 23698742614560.0, + "grad_norm": 2.153773558095368, + "learning_rate": 8.709820432700663e-07, + "loss": 0.7015, + "num_input_tokens_seen": 252882150, + "step": 7764 + }, + { + "epoch": 0.7002750597465843, + "flos": 24681851724960.0, + "grad_norm": 1.683382662752222, + "learning_rate": 8.704998682597784e-07, + "loss": 0.6607, + "num_input_tokens_seen": 252911960, + "step": 7765 + }, + { + "epoch": 0.7003652432700546, + "flos": 22970249325600.0, + "grad_norm": 1.987174274936274, + "learning_rate": 8.700177896246688e-07, + "loss": 0.763, + "num_input_tokens_seen": 252941580, + "step": 7766 + }, + { + "epoch": 0.7004554267935248, + "flos": 19728089926080.0, + "grad_norm": 2.068792051135707, + "learning_rate": 8.695358074058721e-07, + "loss": 0.6727, + "num_input_tokens_seen": 252967370, + "step": 7767 + }, + { + "epoch": 0.7005456103169951, + "flos": 16157744282880.0, + "grad_norm": 2.0557853753545103, + "learning_rate": 8.690539216445136e-07, + "loss": 0.7831, + "num_input_tokens_seen": 252994240, + "step": 7768 + }, + { + "epoch": 0.7006357938404654, + "flos": 23844879875520.0, + "grad_norm": 19.96496565275345, + "learning_rate": 8.685721323817106e-07, + "loss": 0.6502, + "num_input_tokens_seen": 253024410, + "step": 7769 + }, + { + "epoch": 0.7007259773639356, + "flos": 24675755884320.0, + "grad_norm": 2.3136744223378893, + "learning_rate": 8.680904396585718e-07, + "loss": 0.7829, + "num_input_tokens_seen": 253053000, + "step": 7770 + }, + { + "epoch": 0.7008161608874058, + "flos": 36887199405120.0, + "grad_norm": 2.136746422067157, + "learning_rate": 8.676088435161988e-07, + "loss": 0.6871, + "num_input_tokens_seen": 253082765, + "step": 7771 + }, + { + "epoch": 0.7009063444108762, + "flos": 22200194511360.0, + "grad_norm": 2.0845982970999994, + "learning_rate": 8.671273439956824e-07, + "loss": 0.7636, + "num_input_tokens_seen": 253107740, + "step": 7772 + }, + { + "epoch": 0.7009965279343464, + "flos": 24574005050400.0, + "grad_norm": 1.7233961942535307, + "learning_rate": 8.666459411381075e-07, + "loss": 0.7453, + "num_input_tokens_seen": 253136685, + "step": 7773 + }, + { + "epoch": 0.7010867114578166, + "flos": 28614176657280.0, + "grad_norm": 2.2372701601438783, + "learning_rate": 8.661646349845501e-07, + "loss": 0.7727, + "num_input_tokens_seen": 253167335, + "step": 7774 + }, + { + "epoch": 0.7011768949812869, + "flos": 13279162472160.0, + "grad_norm": 1.9686314986153957, + "learning_rate": 8.656834255760783e-07, + "loss": 0.8002, + "num_input_tokens_seen": 253191010, + "step": 7775 + }, + { + "epoch": 0.7012670785047572, + "flos": 22094875380480.0, + "grad_norm": 1.9287194051435956, + "learning_rate": 8.652023129537509e-07, + "loss": 0.7999, + "num_input_tokens_seen": 253220125, + "step": 7776 + }, + { + "epoch": 0.7013572620282275, + "flos": 25775483732640.0, + "grad_norm": 2.1059844848626446, + "learning_rate": 8.647212971586195e-07, + "loss": 0.7735, + "num_input_tokens_seen": 253250435, + "step": 7777 + }, + { + "epoch": 0.7014474455516977, + "flos": 17432904896640.0, + "grad_norm": 2.089565707810961, + "learning_rate": 8.642403782317275e-07, + "loss": 0.6952, + "num_input_tokens_seen": 253279410, + "step": 7778 + }, + { + "epoch": 0.7015376290751679, + "flos": 21505717286400.0, + "grad_norm": 2.3251390189991046, + "learning_rate": 8.637595562141075e-07, + "loss": 0.7502, + "num_input_tokens_seen": 253306185, + "step": 7779 + }, + { + "epoch": 0.7016278125986383, + "flos": 25994559529920.0, + "grad_norm": 2.283394524062969, + "learning_rate": 8.632788311467889e-07, + "loss": 0.7129, + "num_input_tokens_seen": 253339910, + "step": 7780 + }, + { + "epoch": 0.7017179961221085, + "flos": 23697999219360.0, + "grad_norm": 2.102489508059492, + "learning_rate": 8.627982030707867e-07, + "loss": 0.7409, + "num_input_tokens_seen": 253372265, + "step": 7781 + }, + { + "epoch": 0.7018081796455787, + "flos": 20602385948160.0, + "grad_norm": 2.069810684134653, + "learning_rate": 8.623176720271139e-07, + "loss": 0.8094, + "num_input_tokens_seen": 253397355, + "step": 7782 + }, + { + "epoch": 0.701898363169049, + "flos": 19501134139680.0, + "grad_norm": 2.172503483355751, + "learning_rate": 8.618372380567696e-07, + "loss": 0.8145, + "num_input_tokens_seen": 253419945, + "step": 7783 + }, + { + "epoch": 0.7019885466925193, + "flos": 25115319929760.0, + "grad_norm": 2.2185740871271395, + "learning_rate": 8.613569012007478e-07, + "loss": 0.699, + "num_input_tokens_seen": 253447425, + "step": 7784 + }, + { + "epoch": 0.7020787302159895, + "flos": 16594669275360.0, + "grad_norm": 1.9933281120287196, + "learning_rate": 8.608766615000338e-07, + "loss": 0.7836, + "num_input_tokens_seen": 253474785, + "step": 7785 + }, + { + "epoch": 0.7021689137394598, + "flos": 17759790198720.0, + "grad_norm": 2.0217464258740927, + "learning_rate": 8.603965189956039e-07, + "loss": 0.7661, + "num_input_tokens_seen": 253500605, + "step": 7786 + }, + { + "epoch": 0.70225909726293, + "flos": 25848236420160.0, + "grad_norm": 2.178809423199861, + "learning_rate": 8.599164737284276e-07, + "loss": 0.7502, + "num_input_tokens_seen": 253526365, + "step": 7787 + }, + { + "epoch": 0.7023492807864004, + "flos": 22605891171840.0, + "grad_norm": 1.7970218764275327, + "learning_rate": 8.594365257394634e-07, + "loss": 0.7042, + "num_input_tokens_seen": 253554565, + "step": 7788 + }, + { + "epoch": 0.7024394643098706, + "flos": 26139581698080.0, + "grad_norm": 2.092650167347708, + "learning_rate": 8.589566750696637e-07, + "loss": 0.6901, + "num_input_tokens_seen": 253584255, + "step": 7789 + }, + { + "epoch": 0.7025296478333408, + "flos": 32478729649920.0, + "grad_norm": 2.0290236963850234, + "learning_rate": 8.584769217599721e-07, + "loss": 0.6782, + "num_input_tokens_seen": 253614845, + "step": 7790 + }, + { + "epoch": 0.7026198313568112, + "flos": 17759864538240.0, + "grad_norm": 3.7419367250264353, + "learning_rate": 8.579972658513239e-07, + "loss": 0.694, + "num_input_tokens_seen": 253640785, + "step": 7791 + }, + { + "epoch": 0.7027100148802814, + "flos": 21293294876160.0, + "grad_norm": 2.3445808463871236, + "learning_rate": 8.57517707384646e-07, + "loss": 0.7518, + "num_input_tokens_seen": 253667425, + "step": 7792 + }, + { + "epoch": 0.7028001984037516, + "flos": 24609656583840.0, + "grad_norm": 2.252554608867491, + "learning_rate": 8.570382464008574e-07, + "loss": 0.7844, + "num_input_tokens_seen": 253697225, + "step": 7793 + }, + { + "epoch": 0.7028903819272219, + "flos": 24063100768320.0, + "grad_norm": 2.8600313027542854, + "learning_rate": 8.565588829408665e-07, + "loss": 0.7087, + "num_input_tokens_seen": 253726040, + "step": 7794 + }, + { + "epoch": 0.7029805654506922, + "flos": 23954436359040.0, + "grad_norm": 2.0747437468251917, + "learning_rate": 8.560796170455782e-07, + "loss": 0.722, + "num_input_tokens_seen": 253756430, + "step": 7795 + }, + { + "epoch": 0.7030707489741624, + "flos": 22788832228800.0, + "grad_norm": 1.9301072254932483, + "learning_rate": 8.556004487558828e-07, + "loss": 0.7846, + "num_input_tokens_seen": 253786885, + "step": 7796 + }, + { + "epoch": 0.7031609324976327, + "flos": 42278443742880.0, + "grad_norm": 1.86758029269787, + "learning_rate": 8.55121378112669e-07, + "loss": 0.66, + "num_input_tokens_seen": 253819295, + "step": 7797 + }, + { + "epoch": 0.7032511160211029, + "flos": 20930051815200.0, + "grad_norm": 2.2652394986213933, + "learning_rate": 8.546424051568111e-07, + "loss": 0.723, + "num_input_tokens_seen": 253848640, + "step": 7798 + }, + { + "epoch": 0.7033412995445733, + "flos": 15787364628000.0, + "grad_norm": 2.210655167416681, + "learning_rate": 8.541635299291785e-07, + "loss": 0.7959, + "num_input_tokens_seen": 253873005, + "step": 7799 + }, + { + "epoch": 0.7034314830680435, + "flos": 70959643211040.0, + "grad_norm": 0.597721395494976, + "learning_rate": 8.536847524706317e-07, + "loss": 0.5991, + "num_input_tokens_seen": 253972645, + "step": 7800 + }, + { + "epoch": 0.7035216665915137, + "flos": 32660852972160.0, + "grad_norm": 1.6940974086872482, + "learning_rate": 8.532060728220225e-07, + "loss": 0.6743, + "num_input_tokens_seen": 254002825, + "step": 7801 + }, + { + "epoch": 0.703611850114984, + "flos": 23188767576480.0, + "grad_norm": 2.3957843830802434, + "learning_rate": 8.527274910241955e-07, + "loss": 0.7555, + "num_input_tokens_seen": 254032065, + "step": 7802 + }, + { + "epoch": 0.7037020336384543, + "flos": 22456334292960.0, + "grad_norm": 2.00585818638901, + "learning_rate": 8.522490071179833e-07, + "loss": 0.7396, + "num_input_tokens_seen": 254060525, + "step": 7803 + }, + { + "epoch": 0.7037922171619245, + "flos": 41294814255840.0, + "grad_norm": 1.6553306907272192, + "learning_rate": 8.517706211442159e-07, + "loss": 0.6239, + "num_input_tokens_seen": 254094825, + "step": 7804 + }, + { + "epoch": 0.7038824006853948, + "flos": 23986073558400.0, + "grad_norm": 1.8679696204719127, + "learning_rate": 8.512923331437097e-07, + "loss": 0.8239, + "num_input_tokens_seen": 254122375, + "step": 7805 + }, + { + "epoch": 0.703972584208865, + "flos": 23298324060000.0, + "grad_norm": 1.5820293661977596, + "learning_rate": 8.508141431572755e-07, + "loss": 0.7112, + "num_input_tokens_seen": 254151790, + "step": 7806 + }, + { + "epoch": 0.7040627677323353, + "flos": 22314322875360.0, + "grad_norm": 2.485199701432413, + "learning_rate": 8.503360512257152e-07, + "loss": 0.731, + "num_input_tokens_seen": 254180075, + "step": 7807 + }, + { + "epoch": 0.7041529512558056, + "flos": 30328492449120.0, + "grad_norm": 2.0355624856974326, + "learning_rate": 8.498580573898219e-07, + "loss": 0.7115, + "num_input_tokens_seen": 254209595, + "step": 7808 + }, + { + "epoch": 0.7042431347792758, + "flos": 16484109208320.0, + "grad_norm": 2.7138347699294045, + "learning_rate": 8.493801616903813e-07, + "loss": 0.751, + "num_input_tokens_seen": 254233945, + "step": 7809 + }, + { + "epoch": 0.704333318302746, + "flos": 18125486463840.0, + "grad_norm": 1.7070062491465121, + "learning_rate": 8.489023641681705e-07, + "loss": 0.7627, + "num_input_tokens_seen": 254261755, + "step": 7810 + }, + { + "epoch": 0.7044235018262164, + "flos": 27742928555520.0, + "grad_norm": 1.8687808209687615, + "learning_rate": 8.484246648639555e-07, + "loss": 0.7927, + "num_input_tokens_seen": 254293600, + "step": 7811 + }, + { + "epoch": 0.7045136853496866, + "flos": 16230199612320.0, + "grad_norm": 2.579427031982873, + "learning_rate": 8.479470638184994e-07, + "loss": 0.6191, + "num_input_tokens_seen": 254321455, + "step": 7812 + }, + { + "epoch": 0.7046038688731568, + "flos": 24172136875200.0, + "grad_norm": 2.8303935189163765, + "learning_rate": 8.474695610725513e-07, + "loss": 0.6625, + "num_input_tokens_seen": 254348220, + "step": 7813 + }, + { + "epoch": 0.7046940523966272, + "flos": 37175422423200.0, + "grad_norm": 1.5874564878252648, + "learning_rate": 8.469921566668552e-07, + "loss": 0.8174, + "num_input_tokens_seen": 254377800, + "step": 7814 + }, + { + "epoch": 0.7047842359200974, + "flos": 24713600433600.0, + "grad_norm": 4.576643876706417, + "learning_rate": 8.46514850642146e-07, + "loss": 0.7926, + "num_input_tokens_seen": 254408840, + "step": 7815 + }, + { + "epoch": 0.7048744194435677, + "flos": 22970174986080.0, + "grad_norm": 2.137014197752669, + "learning_rate": 8.460376430391499e-07, + "loss": 0.7282, + "num_input_tokens_seen": 254437795, + "step": 7816 + }, + { + "epoch": 0.7049646029670379, + "flos": 24464225548320.0, + "grad_norm": 1.7408581096492382, + "learning_rate": 8.455605338985858e-07, + "loss": 0.7723, + "num_input_tokens_seen": 254470785, + "step": 7817 + }, + { + "epoch": 0.7050547864905082, + "flos": 37434758804160.0, + "grad_norm": 1.7611548627251632, + "learning_rate": 8.45083523261161e-07, + "loss": 0.5993, + "num_input_tokens_seen": 254503725, + "step": 7818 + }, + { + "epoch": 0.7051449700139785, + "flos": 25702805384640.0, + "grad_norm": 1.873609790404074, + "learning_rate": 8.446066111675796e-07, + "loss": 0.7495, + "num_input_tokens_seen": 254531695, + "step": 7819 + }, + { + "epoch": 0.7052351535374487, + "flos": 27922598673600.0, + "grad_norm": 2.01413559407372, + "learning_rate": 8.441297976585314e-07, + "loss": 0.6362, + "num_input_tokens_seen": 254561405, + "step": 7820 + }, + { + "epoch": 0.7053253370609189, + "flos": 21221434262880.0, + "grad_norm": 1.814041403057167, + "learning_rate": 8.436530827747037e-07, + "loss": 0.8295, + "num_input_tokens_seen": 254589365, + "step": 7821 + }, + { + "epoch": 0.7054155205843893, + "flos": 21652486433280.0, + "grad_norm": 2.310577830915835, + "learning_rate": 8.431764665567704e-07, + "loss": 0.8798, + "num_input_tokens_seen": 254615105, + "step": 7822 + }, + { + "epoch": 0.7055057041078595, + "flos": 21039645468480.0, + "grad_norm": 2.5327212920289757, + "learning_rate": 8.426999490453996e-07, + "loss": 0.6866, + "num_input_tokens_seen": 254643245, + "step": 7823 + }, + { + "epoch": 0.7055958876313297, + "flos": 18816506901120.0, + "grad_norm": 1.9646887898247491, + "learning_rate": 8.422235302812504e-07, + "loss": 0.7532, + "num_input_tokens_seen": 254669310, + "step": 7824 + }, + { + "epoch": 0.7056860711548, + "flos": 25406851056480.0, + "grad_norm": 1.7664462179920308, + "learning_rate": 8.417472103049734e-07, + "loss": 0.8165, + "num_input_tokens_seen": 254697550, + "step": 7825 + }, + { + "epoch": 0.7057762546782703, + "flos": 21726540062400.0, + "grad_norm": 1.9404064651274606, + "learning_rate": 8.412709891572112e-07, + "loss": 0.7771, + "num_input_tokens_seen": 254726155, + "step": 7826 + }, + { + "epoch": 0.7058664382017406, + "flos": 19290087010560.0, + "grad_norm": 2.1792626188997484, + "learning_rate": 8.407948668785978e-07, + "loss": 0.78, + "num_input_tokens_seen": 254753635, + "step": 7827 + }, + { + "epoch": 0.7059566217252108, + "flos": 23735026033920.0, + "grad_norm": 5.464863295175913, + "learning_rate": 8.403188435097576e-07, + "loss": 0.7375, + "num_input_tokens_seen": 254782755, + "step": 7828 + }, + { + "epoch": 0.706046805248681, + "flos": 25005689106720.0, + "grad_norm": 1.705505107888765, + "learning_rate": 8.398429190913081e-07, + "loss": 0.8152, + "num_input_tokens_seen": 254812290, + "step": 7829 + }, + { + "epoch": 0.7061369887721514, + "flos": 26540632138560.0, + "grad_norm": 1.8326734381146232, + "learning_rate": 8.393670936638578e-07, + "loss": 0.7451, + "num_input_tokens_seen": 254840540, + "step": 7830 + }, + { + "epoch": 0.7062271722956216, + "flos": 22496520537120.0, + "grad_norm": 2.125693376400946, + "learning_rate": 8.388913672680067e-07, + "loss": 0.7412, + "num_input_tokens_seen": 254870650, + "step": 7831 + }, + { + "epoch": 0.7063173558190918, + "flos": 24676945316640.0, + "grad_norm": 2.2666051388785875, + "learning_rate": 8.384157399443472e-07, + "loss": 0.6901, + "num_input_tokens_seen": 254898680, + "step": 7832 + }, + { + "epoch": 0.7064075393425621, + "flos": 19909730041440.0, + "grad_norm": 1.6734699947789005, + "learning_rate": 8.379402117334601e-07, + "loss": 0.7672, + "num_input_tokens_seen": 254925180, + "step": 7833 + }, + { + "epoch": 0.7064977228660324, + "flos": 22782736388160.0, + "grad_norm": 1.7760285258622717, + "learning_rate": 8.374647826759232e-07, + "loss": 0.7432, + "num_input_tokens_seen": 254954055, + "step": 7834 + }, + { + "epoch": 0.7065879063895026, + "flos": 24492703318080.0, + "grad_norm": 3.2259120331682554, + "learning_rate": 8.369894528122998e-07, + "loss": 0.7437, + "num_input_tokens_seen": 254982435, + "step": 7835 + }, + { + "epoch": 0.7066780899129729, + "flos": 24245632957920.0, + "grad_norm": 2.1217653675999344, + "learning_rate": 8.365142221831505e-07, + "loss": 0.6075, + "num_input_tokens_seen": 255012800, + "step": 7836 + }, + { + "epoch": 0.7067682734364432, + "flos": 23407025639040.0, + "grad_norm": 2.26431394329929, + "learning_rate": 8.360390908290222e-07, + "loss": 0.7055, + "num_input_tokens_seen": 255040720, + "step": 7837 + }, + { + "epoch": 0.7068584569599135, + "flos": 21986880026880.0, + "grad_norm": 2.28415348841992, + "learning_rate": 8.355640587904569e-07, + "loss": 0.7246, + "num_input_tokens_seen": 255067640, + "step": 7838 + }, + { + "epoch": 0.7069486404833837, + "flos": 20237061380640.0, + "grad_norm": 2.5973873142622623, + "learning_rate": 8.350891261079866e-07, + "loss": 0.754, + "num_input_tokens_seen": 255093035, + "step": 7839 + }, + { + "epoch": 0.7070388240068539, + "flos": 21148347047520.0, + "grad_norm": 2.0207734115473337, + "learning_rate": 8.346142928221356e-07, + "loss": 0.7866, + "num_input_tokens_seen": 255121085, + "step": 7840 + }, + { + "epoch": 0.7071290075303243, + "flos": 18743717043840.0, + "grad_norm": 2.182180203039519, + "learning_rate": 8.341395589734189e-07, + "loss": 0.5863, + "num_input_tokens_seen": 255148800, + "step": 7841 + }, + { + "epoch": 0.7072191910537945, + "flos": 15647211698400.0, + "grad_norm": 2.5336377220325614, + "learning_rate": 8.336649246023433e-07, + "loss": 0.6957, + "num_input_tokens_seen": 255173900, + "step": 7842 + }, + { + "epoch": 0.7073093745772647, + "flos": 21185299522560.0, + "grad_norm": 1.7477085915692359, + "learning_rate": 8.331903897494084e-07, + "loss": 0.6452, + "num_input_tokens_seen": 255202455, + "step": 7843 + }, + { + "epoch": 0.707399558100735, + "flos": 20747668304640.0, + "grad_norm": 2.0986545112521204, + "learning_rate": 8.327159544551024e-07, + "loss": 0.757, + "num_input_tokens_seen": 255229265, + "step": 7844 + }, + { + "epoch": 0.7074897416242053, + "flos": 26572232168160.0, + "grad_norm": 2.061385576128163, + "learning_rate": 8.322416187599073e-07, + "loss": 0.7048, + "num_input_tokens_seen": 255257220, + "step": 7845 + }, + { + "epoch": 0.7075799251476755, + "flos": 25264728129600.0, + "grad_norm": 2.4478083052222432, + "learning_rate": 8.317673827042963e-07, + "loss": 0.7664, + "num_input_tokens_seen": 255283535, + "step": 7846 + }, + { + "epoch": 0.7076701086711458, + "flos": 39905339429280.0, + "grad_norm": 1.9501477552335442, + "learning_rate": 8.312932463287339e-07, + "loss": 0.7292, + "num_input_tokens_seen": 255311855, + "step": 7847 + }, + { + "epoch": 0.707760292194616, + "flos": 27816015770880.0, + "grad_norm": 2.7182133462862983, + "learning_rate": 8.308192096736759e-07, + "loss": 0.7134, + "num_input_tokens_seen": 255341365, + "step": 7848 + }, + { + "epoch": 0.7078504757180863, + "flos": 20237358738720.0, + "grad_norm": 1.7229400885334165, + "learning_rate": 8.303452727795703e-07, + "loss": 0.7225, + "num_input_tokens_seen": 255367470, + "step": 7849 + }, + { + "epoch": 0.7079406592415566, + "flos": 18889408267680.0, + "grad_norm": 2.3275451962255023, + "learning_rate": 8.298714356868542e-07, + "loss": 0.8212, + "num_input_tokens_seen": 255394225, + "step": 7850 + }, + { + "epoch": 0.7080308427650268, + "flos": 29673792600960.0, + "grad_norm": 1.8761224565861723, + "learning_rate": 8.293976984359605e-07, + "loss": 0.7592, + "num_input_tokens_seen": 255424395, + "step": 7851 + }, + { + "epoch": 0.708121026288497, + "flos": 21731595149760.0, + "grad_norm": 3.643636481216393, + "learning_rate": 8.289240610673092e-07, + "loss": 0.7479, + "num_input_tokens_seen": 255453165, + "step": 7852 + }, + { + "epoch": 0.7082112098119674, + "flos": 20560415555520.0, + "grad_norm": 2.083756467446472, + "learning_rate": 8.284505236213144e-07, + "loss": 0.7753, + "num_input_tokens_seen": 255481570, + "step": 7853 + }, + { + "epoch": 0.7083013933354376, + "flos": 27775941036000.0, + "grad_norm": 1.7956211525291281, + "learning_rate": 8.279770861383806e-07, + "loss": 0.7371, + "num_input_tokens_seen": 255511520, + "step": 7854 + }, + { + "epoch": 0.7083915768589079, + "flos": 21258721265760.0, + "grad_norm": 1.8230744720456127, + "learning_rate": 8.275037486589042e-07, + "loss": 0.7394, + "num_input_tokens_seen": 255540080, + "step": 7855 + }, + { + "epoch": 0.7084817603823781, + "flos": 63376042600800.0, + "grad_norm": 0.67175555830893, + "learning_rate": 8.270305112232739e-07, + "loss": 0.5582, + "num_input_tokens_seen": 255629015, + "step": 7856 + }, + { + "epoch": 0.7085719439058484, + "flos": 25663585554240.0, + "grad_norm": 2.1820156557131627, + "learning_rate": 8.265573738718665e-07, + "loss": 0.7417, + "num_input_tokens_seen": 255659005, + "step": 7857 + }, + { + "epoch": 0.7086621274293187, + "flos": 23949975987840.0, + "grad_norm": 2.038683625900221, + "learning_rate": 8.260843366450559e-07, + "loss": 0.7031, + "num_input_tokens_seen": 255686680, + "step": 7858 + }, + { + "epoch": 0.7087523109527889, + "flos": 22606039850880.0, + "grad_norm": 1.8669893806086038, + "learning_rate": 8.256113995832017e-07, + "loss": 0.74, + "num_input_tokens_seen": 255715195, + "step": 7859 + }, + { + "epoch": 0.7088424944762591, + "flos": 25813922998080.0, + "grad_norm": 1.837802410542361, + "learning_rate": 8.251385627266583e-07, + "loss": 0.7875, + "num_input_tokens_seen": 255748685, + "step": 7860 + }, + { + "epoch": 0.7089326779997295, + "flos": 23589074621760.0, + "grad_norm": 2.33278675716681, + "learning_rate": 8.24665826115771e-07, + "loss": 0.7564, + "num_input_tokens_seen": 255776895, + "step": 7861 + }, + { + "epoch": 0.7090228615231997, + "flos": 23403754700160.0, + "grad_norm": 1.753149060306918, + "learning_rate": 8.241931897908763e-07, + "loss": 0.7602, + "num_input_tokens_seen": 255806135, + "step": 7862 + }, + { + "epoch": 0.7091130450466699, + "flos": 24932787740160.0, + "grad_norm": 1.7207870203148492, + "learning_rate": 8.237206537923016e-07, + "loss": 0.7655, + "num_input_tokens_seen": 255834785, + "step": 7863 + }, + { + "epoch": 0.7092032285701403, + "flos": 19290198519840.0, + "grad_norm": 4.650567600514803, + "learning_rate": 8.232482181603671e-07, + "loss": 0.734, + "num_input_tokens_seen": 255862190, + "step": 7864 + }, + { + "epoch": 0.7092934120936105, + "flos": 21219612944640.0, + "grad_norm": 1.9648080143797693, + "learning_rate": 8.227758829353828e-07, + "loss": 0.7219, + "num_input_tokens_seen": 255892505, + "step": 7865 + }, + { + "epoch": 0.7093835956170808, + "flos": 22423879358880.0, + "grad_norm": 1.8893407447932502, + "learning_rate": 8.223036481576522e-07, + "loss": 0.7129, + "num_input_tokens_seen": 255920725, + "step": 7866 + }, + { + "epoch": 0.709473779140551, + "flos": 37645248386880.0, + "grad_norm": 2.302097119136171, + "learning_rate": 8.218315138674672e-07, + "loss": 0.7466, + "num_input_tokens_seen": 255950725, + "step": 7867 + }, + { + "epoch": 0.7095639626640213, + "flos": 60895909347360.0, + "grad_norm": 0.6184238171328773, + "learning_rate": 8.21359480105114e-07, + "loss": 0.5758, + "num_input_tokens_seen": 256037025, + "step": 7868 + }, + { + "epoch": 0.7096541461874916, + "flos": 27045923786880.0, + "grad_norm": 1.9473934456857465, + "learning_rate": 8.208875469108689e-07, + "loss": 0.7763, + "num_input_tokens_seen": 256065790, + "step": 7869 + }, + { + "epoch": 0.7097443297109618, + "flos": 27013654701600.0, + "grad_norm": 3.060671980880725, + "learning_rate": 8.204157143249997e-07, + "loss": 0.7295, + "num_input_tokens_seen": 256093345, + "step": 7870 + }, + { + "epoch": 0.709834513234432, + "flos": 31092600101760.0, + "grad_norm": 1.7380671337142324, + "learning_rate": 8.199439823877668e-07, + "loss": 0.7179, + "num_input_tokens_seen": 256125310, + "step": 7871 + }, + { + "epoch": 0.7099246967579024, + "flos": 27779434993440.0, + "grad_norm": 1.8056729751292682, + "learning_rate": 8.194723511394186e-07, + "loss": 0.7269, + "num_input_tokens_seen": 256155645, + "step": 7872 + }, + { + "epoch": 0.7100148802813726, + "flos": 14044125029280.0, + "grad_norm": 2.2363533918127767, + "learning_rate": 8.190008206202002e-07, + "loss": 0.6901, + "num_input_tokens_seen": 256181695, + "step": 7873 + }, + { + "epoch": 0.7101050638048428, + "flos": 12294863929440.0, + "grad_norm": 2.3449118214114817, + "learning_rate": 8.185293908703423e-07, + "loss": 0.7912, + "num_input_tokens_seen": 256206720, + "step": 7874 + }, + { + "epoch": 0.7101952473283131, + "flos": 17360152209120.0, + "grad_norm": 1.7202731232809694, + "learning_rate": 8.180580619300727e-07, + "loss": 0.7059, + "num_input_tokens_seen": 256233615, + "step": 7875 + }, + { + "epoch": 0.7102854308517834, + "flos": 43882125128160.0, + "grad_norm": 1.56184396322919, + "learning_rate": 8.175868338396057e-07, + "loss": 0.6948, + "num_input_tokens_seen": 256269635, + "step": 7876 + }, + { + "epoch": 0.7103756143752537, + "flos": 24282139395840.0, + "grad_norm": 2.168896395884396, + "learning_rate": 8.171157066391499e-07, + "loss": 0.7137, + "num_input_tokens_seen": 256299585, + "step": 7877 + }, + { + "epoch": 0.7104657978987239, + "flos": 23553200069760.0, + "grad_norm": 2.42688297432516, + "learning_rate": 8.166446803689045e-07, + "loss": 0.7642, + "num_input_tokens_seen": 256327680, + "step": 7878 + }, + { + "epoch": 0.7105559814221941, + "flos": 65276607558240.0, + "grad_norm": 0.6696469156959705, + "learning_rate": 8.161737550690595e-07, + "loss": 0.5826, + "num_input_tokens_seen": 256420500, + "step": 7879 + }, + { + "epoch": 0.7106461649456645, + "flos": 23516359104000.0, + "grad_norm": 4.7528699938143255, + "learning_rate": 8.157029307797976e-07, + "loss": 0.7657, + "num_input_tokens_seen": 256450760, + "step": 7880 + }, + { + "epoch": 0.7107363484691347, + "flos": 23512381939680.0, + "grad_norm": 1.5420489309837402, + "learning_rate": 8.152322075412925e-07, + "loss": 0.7971, + "num_input_tokens_seen": 256479430, + "step": 7881 + }, + { + "epoch": 0.7108265319926049, + "flos": 26026382578080.0, + "grad_norm": 1.5706936957155262, + "learning_rate": 8.147615853937073e-07, + "loss": 0.6949, + "num_input_tokens_seen": 256511220, + "step": 7882 + }, + { + "epoch": 0.7109167155160752, + "flos": 23918004260640.0, + "grad_norm": 1.947739033751034, + "learning_rate": 8.142910643771992e-07, + "loss": 0.6656, + "num_input_tokens_seen": 256541705, + "step": 7883 + }, + { + "epoch": 0.7110068990395455, + "flos": 21913309604640.0, + "grad_norm": 2.1017133807507293, + "learning_rate": 8.138206445319152e-07, + "loss": 0.7024, + "num_input_tokens_seen": 256567760, + "step": 7884 + }, + { + "epoch": 0.7110970825630157, + "flos": 27888248081760.0, + "grad_norm": 1.7620884408047666, + "learning_rate": 8.133503258979944e-07, + "loss": 0.6422, + "num_input_tokens_seen": 256595680, + "step": 7885 + }, + { + "epoch": 0.711187266086486, + "flos": 20636773709760.0, + "grad_norm": 5.384783689781994, + "learning_rate": 8.12880108515567e-07, + "loss": 0.6672, + "num_input_tokens_seen": 256624645, + "step": 7886 + }, + { + "epoch": 0.7112774496099563, + "flos": 24901224880320.0, + "grad_norm": 1.6595419790734451, + "learning_rate": 8.124099924247543e-07, + "loss": 0.7118, + "num_input_tokens_seen": 256653475, + "step": 7887 + }, + { + "epoch": 0.7113676331334265, + "flos": 25549791718080.0, + "grad_norm": 1.6778599222479549, + "learning_rate": 8.119399776656701e-07, + "loss": 0.6466, + "num_input_tokens_seen": 256681915, + "step": 7888 + }, + { + "epoch": 0.7114578166568968, + "flos": 26866179329280.0, + "grad_norm": 1.7435754285270766, + "learning_rate": 8.114700642784167e-07, + "loss": 0.7382, + "num_input_tokens_seen": 256711110, + "step": 7889 + }, + { + "epoch": 0.711548000180367, + "flos": 25769164873440.0, + "grad_norm": 1.668810451141285, + "learning_rate": 8.110002523030921e-07, + "loss": 0.7997, + "num_input_tokens_seen": 256739620, + "step": 7890 + }, + { + "epoch": 0.7116381837038374, + "flos": 22786825061760.0, + "grad_norm": 1.7967719007022223, + "learning_rate": 8.105305417797808e-07, + "loss": 0.7615, + "num_input_tokens_seen": 256768985, + "step": 7891 + }, + { + "epoch": 0.7117283672273076, + "flos": 26686137513600.0, + "grad_norm": 2.104815493219785, + "learning_rate": 8.100609327485635e-07, + "loss": 0.7799, + "num_input_tokens_seen": 256797520, + "step": 7892 + }, + { + "epoch": 0.7118185507507778, + "flos": 22854522661920.0, + "grad_norm": 4.008081378803495, + "learning_rate": 8.095914252495082e-07, + "loss": 0.8075, + "num_input_tokens_seen": 256823450, + "step": 7893 + }, + { + "epoch": 0.7119087342742481, + "flos": 20849976684960.0, + "grad_norm": 2.2685096003846565, + "learning_rate": 8.091220193226762e-07, + "loss": 0.7095, + "num_input_tokens_seen": 256849945, + "step": 7894 + }, + { + "epoch": 0.7119989177977184, + "flos": 28108885008960.0, + "grad_norm": 1.622035476889255, + "learning_rate": 8.0865271500812e-07, + "loss": 0.7271, + "num_input_tokens_seen": 256880180, + "step": 7895 + }, + { + "epoch": 0.7120891013211886, + "flos": 21221211244320.0, + "grad_norm": 3.362909063819745, + "learning_rate": 8.081835123458831e-07, + "loss": 0.7786, + "num_input_tokens_seen": 256908520, + "step": 7896 + }, + { + "epoch": 0.7121792848446589, + "flos": 22824520932000.0, + "grad_norm": 2.2069757235133434, + "learning_rate": 8.077144113760013e-07, + "loss": 0.7244, + "num_input_tokens_seen": 256936880, + "step": 7897 + }, + { + "epoch": 0.7122694683681291, + "flos": 35975393361600.0, + "grad_norm": 1.562323674743392, + "learning_rate": 8.072454121384995e-07, + "loss": 0.7201, + "num_input_tokens_seen": 256968690, + "step": 7898 + }, + { + "epoch": 0.7123596518915994, + "flos": 28799050541760.0, + "grad_norm": 2.071565414474999, + "learning_rate": 8.067765146733958e-07, + "loss": 0.7863, + "num_input_tokens_seen": 256998480, + "step": 7899 + }, + { + "epoch": 0.7124498354150697, + "flos": 24171876686880.0, + "grad_norm": 4.241850950263685, + "learning_rate": 8.063077190206993e-07, + "loss": 0.816, + "num_input_tokens_seen": 257026825, + "step": 7900 + }, + { + "epoch": 0.7125400189385399, + "flos": 60137637347040.0, + "grad_norm": 0.6050419195091585, + "learning_rate": 8.058390252204101e-07, + "loss": 0.4932, + "num_input_tokens_seen": 257122420, + "step": 7901 + }, + { + "epoch": 0.7126302024620101, + "flos": 27303810547200.0, + "grad_norm": 2.0516342859427006, + "learning_rate": 8.0537043331252e-07, + "loss": 0.7346, + "num_input_tokens_seen": 257155320, + "step": 7902 + }, + { + "epoch": 0.7127203859854805, + "flos": 30255888440640.0, + "grad_norm": 2.0857756476884646, + "learning_rate": 8.049019433370121e-07, + "loss": 0.6787, + "num_input_tokens_seen": 257183760, + "step": 7903 + }, + { + "epoch": 0.7128105695089507, + "flos": 35862380090400.0, + "grad_norm": 1.8556133655608649, + "learning_rate": 8.044335553338588e-07, + "loss": 0.743, + "num_input_tokens_seen": 257213490, + "step": 7904 + }, + { + "epoch": 0.712900753032421, + "flos": 21840854275200.0, + "grad_norm": 1.976528061879614, + "learning_rate": 8.039652693430281e-07, + "loss": 0.7534, + "num_input_tokens_seen": 257243250, + "step": 7905 + }, + { + "epoch": 0.7129909365558912, + "flos": 25623956856480.0, + "grad_norm": 3.1739986432558753, + "learning_rate": 8.034970854044742e-07, + "loss": 0.7176, + "num_input_tokens_seen": 257272035, + "step": 7906 + }, + { + "epoch": 0.7130811200793615, + "flos": 24313181879040.0, + "grad_norm": 6.026670308294856, + "learning_rate": 8.03029003558148e-07, + "loss": 0.8381, + "num_input_tokens_seen": 257300515, + "step": 7907 + }, + { + "epoch": 0.7131713036028318, + "flos": 13095998396640.0, + "grad_norm": 2.200860449111025, + "learning_rate": 8.025610238439864e-07, + "loss": 0.805, + "num_input_tokens_seen": 257322890, + "step": 7908 + }, + { + "epoch": 0.713261487126302, + "flos": 20164494541920.0, + "grad_norm": 2.4486694693455426, + "learning_rate": 8.020931463019207e-07, + "loss": 0.8178, + "num_input_tokens_seen": 257349955, + "step": 7909 + }, + { + "epoch": 0.7133516706497723, + "flos": 16448717863200.0, + "grad_norm": 1.9993039350409958, + "learning_rate": 8.016253709718732e-07, + "loss": 0.6875, + "num_input_tokens_seen": 257377435, + "step": 7910 + }, + { + "epoch": 0.7134418541732426, + "flos": 29382633171840.0, + "grad_norm": 5.264738557271008, + "learning_rate": 8.011576978937567e-07, + "loss": 0.65, + "num_input_tokens_seen": 257404635, + "step": 7911 + }, + { + "epoch": 0.7135320376967128, + "flos": 17577369518400.0, + "grad_norm": 1.8562444410658292, + "learning_rate": 8.006901271074764e-07, + "loss": 0.7388, + "num_input_tokens_seen": 257431815, + "step": 7912 + }, + { + "epoch": 0.713622221220183, + "flos": 21652300584480.0, + "grad_norm": 3.1231238264440835, + "learning_rate": 8.002226586529261e-07, + "loss": 0.7806, + "num_input_tokens_seen": 257460510, + "step": 7913 + }, + { + "epoch": 0.7137124047436534, + "flos": 28979575564320.0, + "grad_norm": 1.9581067447925498, + "learning_rate": 7.997552925699956e-07, + "loss": 0.82, + "num_input_tokens_seen": 257490140, + "step": 7914 + }, + { + "epoch": 0.7138025882671236, + "flos": 25483878266400.0, + "grad_norm": 2.057642215527661, + "learning_rate": 7.992880288985606e-07, + "loss": 0.8358, + "num_input_tokens_seen": 257518685, + "step": 7915 + }, + { + "epoch": 0.7138927717905938, + "flos": 23328325789920.0, + "grad_norm": 1.834363676747819, + "learning_rate": 7.988208676784918e-07, + "loss": 0.7512, + "num_input_tokens_seen": 257547520, + "step": 7916 + }, + { + "epoch": 0.7139829553140641, + "flos": 36450237242880.0, + "grad_norm": 4.076245806708544, + "learning_rate": 7.983538089496497e-07, + "loss": 0.7373, + "num_input_tokens_seen": 257576825, + "step": 7917 + }, + { + "epoch": 0.7140731388375344, + "flos": 29091696761280.0, + "grad_norm": 1.9234003976052558, + "learning_rate": 7.978868527518864e-07, + "loss": 0.7001, + "num_input_tokens_seen": 257607735, + "step": 7918 + }, + { + "epoch": 0.7141633223610047, + "flos": 27378867759840.0, + "grad_norm": 1.7057609807510425, + "learning_rate": 7.974199991250455e-07, + "loss": 0.717, + "num_input_tokens_seen": 257638960, + "step": 7919 + }, + { + "epoch": 0.7142535058844749, + "flos": 23371151087040.0, + "grad_norm": 2.4614369687834676, + "learning_rate": 7.969532481089616e-07, + "loss": 0.8399, + "num_input_tokens_seen": 257666965, + "step": 7920 + }, + { + "epoch": 0.7143436894079451, + "flos": 23224902316800.0, + "grad_norm": 1.8240317140549323, + "learning_rate": 7.964865997434589e-07, + "loss": 0.7564, + "num_input_tokens_seen": 257694935, + "step": 7921 + }, + { + "epoch": 0.7144338729314155, + "flos": 27123620052480.0, + "grad_norm": 1.8294021188428633, + "learning_rate": 7.96020054068357e-07, + "loss": 0.8132, + "num_input_tokens_seen": 257723625, + "step": 7922 + }, + { + "epoch": 0.7145240564548857, + "flos": 34337621572800.0, + "grad_norm": 1.5606634409199105, + "learning_rate": 7.95553611123462e-07, + "loss": 0.6979, + "num_input_tokens_seen": 257759715, + "step": 7923 + }, + { + "epoch": 0.7146142399783559, + "flos": 22460237117760.0, + "grad_norm": 3.9596798798225326, + "learning_rate": 7.950872709485741e-07, + "loss": 0.9198, + "num_input_tokens_seen": 257786680, + "step": 7924 + }, + { + "epoch": 0.7147044235018262, + "flos": 20820458161920.0, + "grad_norm": 2.9317746471697155, + "learning_rate": 7.946210335834842e-07, + "loss": 0.7973, + "num_input_tokens_seen": 257813975, + "step": 7925 + }, + { + "epoch": 0.7147946070252965, + "flos": 21913606962720.0, + "grad_norm": 1.5598699178012971, + "learning_rate": 7.94154899067974e-07, + "loss": 0.667, + "num_input_tokens_seen": 257844845, + "step": 7926 + }, + { + "epoch": 0.7148847905487667, + "flos": 24500694816480.0, + "grad_norm": 1.9340936188036495, + "learning_rate": 7.936888674418177e-07, + "loss": 0.6701, + "num_input_tokens_seen": 257875000, + "step": 7927 + }, + { + "epoch": 0.714974974072237, + "flos": 16995385188000.0, + "grad_norm": 2.3285763706218354, + "learning_rate": 7.932229387447771e-07, + "loss": 0.7455, + "num_input_tokens_seen": 257900655, + "step": 7928 + }, + { + "epoch": 0.7150651575957072, + "flos": 20747333776800.0, + "grad_norm": 2.814642704568792, + "learning_rate": 7.927571130166109e-07, + "loss": 0.7326, + "num_input_tokens_seen": 257925010, + "step": 7929 + }, + { + "epoch": 0.7151553411191776, + "flos": 67449003669600.0, + "grad_norm": 0.6266213605825653, + "learning_rate": 7.922913902970632e-07, + "loss": 0.5669, + "num_input_tokens_seen": 258024305, + "step": 7930 + }, + { + "epoch": 0.7152455246426478, + "flos": 21293815252800.0, + "grad_norm": 2.066186442519814, + "learning_rate": 7.918257706258744e-07, + "loss": 0.7323, + "num_input_tokens_seen": 258052565, + "step": 7931 + }, + { + "epoch": 0.715335708166118, + "flos": 22896567394080.0, + "grad_norm": 3.6655621743570452, + "learning_rate": 7.913602540427724e-07, + "loss": 0.7448, + "num_input_tokens_seen": 258080800, + "step": 7932 + }, + { + "epoch": 0.7154258916895884, + "flos": 23079471281280.0, + "grad_norm": 1.781530993868377, + "learning_rate": 7.908948405874775e-07, + "loss": 0.7386, + "num_input_tokens_seen": 258108090, + "step": 7933 + }, + { + "epoch": 0.7155160752130586, + "flos": 34956075171360.0, + "grad_norm": 1.7452406879565139, + "learning_rate": 7.904295302997019e-07, + "loss": 0.6671, + "num_input_tokens_seen": 258136290, + "step": 7934 + }, + { + "epoch": 0.7156062587365288, + "flos": 21872305625760.0, + "grad_norm": 1.9698810658642598, + "learning_rate": 7.899643232191484e-07, + "loss": 0.7546, + "num_input_tokens_seen": 258163135, + "step": 7935 + }, + { + "epoch": 0.7156964422599991, + "flos": 22715187467040.0, + "grad_norm": 1.6460767718844034, + "learning_rate": 7.894992193855108e-07, + "loss": 0.7262, + "num_input_tokens_seen": 258191560, + "step": 7936 + }, + { + "epoch": 0.7157866257834694, + "flos": 22493249598240.0, + "grad_norm": 2.151159497251933, + "learning_rate": 7.890342188384751e-07, + "loss": 0.6912, + "num_input_tokens_seen": 258220770, + "step": 7937 + }, + { + "epoch": 0.7158768093069396, + "flos": 21838475410560.0, + "grad_norm": 2.4521364783721222, + "learning_rate": 7.885693216177165e-07, + "loss": 0.6829, + "num_input_tokens_seen": 258250055, + "step": 7938 + }, + { + "epoch": 0.7159669928304099, + "flos": 23150476990080.0, + "grad_norm": 1.7488042969962376, + "learning_rate": 7.88104527762903e-07, + "loss": 0.7373, + "num_input_tokens_seen": 258279690, + "step": 7939 + }, + { + "epoch": 0.7160571763538801, + "flos": 20528778356160.0, + "grad_norm": 2.563783370776815, + "learning_rate": 7.876398373136936e-07, + "loss": 0.6975, + "num_input_tokens_seen": 258308615, + "step": 7940 + }, + { + "epoch": 0.7161473598773505, + "flos": 28362385737600.0, + "grad_norm": 1.7588806297955513, + "learning_rate": 7.87175250309738e-07, + "loss": 0.7389, + "num_input_tokens_seen": 258339580, + "step": 7941 + }, + { + "epoch": 0.7162375434008207, + "flos": 22343952907680.0, + "grad_norm": 1.805153724591379, + "learning_rate": 7.867107667906785e-07, + "loss": 0.8142, + "num_input_tokens_seen": 258366475, + "step": 7942 + }, + { + "epoch": 0.7163277269242909, + "flos": 24789326701920.0, + "grad_norm": 2.653270261851962, + "learning_rate": 7.862463867961446e-07, + "loss": 0.7406, + "num_input_tokens_seen": 258397625, + "step": 7943 + }, + { + "epoch": 0.7164179104477612, + "flos": 25186176959520.0, + "grad_norm": 1.518195420724965, + "learning_rate": 7.857821103657632e-07, + "loss": 0.8043, + "num_input_tokens_seen": 258425950, + "step": 7944 + }, + { + "epoch": 0.7165080939712315, + "flos": 22093165571520.0, + "grad_norm": 1.7085251894269788, + "learning_rate": 7.853179375391459e-07, + "loss": 0.7725, + "num_input_tokens_seen": 258454490, + "step": 7945 + }, + { + "epoch": 0.7165982774947017, + "flos": 26613793693440.0, + "grad_norm": 2.2269766439130385, + "learning_rate": 7.848538683559012e-07, + "loss": 0.7296, + "num_input_tokens_seen": 258483825, + "step": 7946 + }, + { + "epoch": 0.716688461018172, + "flos": 23585989531680.0, + "grad_norm": 2.152758328202963, + "learning_rate": 7.843899028556238e-07, + "loss": 0.7819, + "num_input_tokens_seen": 258511260, + "step": 7947 + }, + { + "epoch": 0.7167786445416422, + "flos": 23735657919840.0, + "grad_norm": 3.5413581863396817, + "learning_rate": 7.839260410779029e-07, + "loss": 0.7341, + "num_input_tokens_seen": 258543525, + "step": 7948 + }, + { + "epoch": 0.7168688280651125, + "flos": 24719324576640.0, + "grad_norm": 1.7227467623436497, + "learning_rate": 7.834622830623175e-07, + "loss": 0.7621, + "num_input_tokens_seen": 258572730, + "step": 7949 + }, + { + "epoch": 0.7169590115885828, + "flos": 22496334688320.0, + "grad_norm": 2.1143228780764702, + "learning_rate": 7.82998628848438e-07, + "loss": 0.7126, + "num_input_tokens_seen": 258600705, + "step": 7950 + }, + { + "epoch": 0.717049195112053, + "flos": 16996202922720.0, + "grad_norm": 1.5692286760365146, + "learning_rate": 7.825350784758261e-07, + "loss": 0.7314, + "num_input_tokens_seen": 258629260, + "step": 7951 + }, + { + "epoch": 0.7171393786355232, + "flos": 24609061867680.0, + "grad_norm": 1.6679007094216542, + "learning_rate": 7.820716319840342e-07, + "loss": 0.7055, + "num_input_tokens_seen": 258658640, + "step": 7952 + }, + { + "epoch": 0.7172295621589936, + "flos": 25301940792960.0, + "grad_norm": 2.8173918672677285, + "learning_rate": 7.816082894126069e-07, + "loss": 0.8265, + "num_input_tokens_seen": 258688140, + "step": 7953 + }, + { + "epoch": 0.7173197456824638, + "flos": 24718990048800.0, + "grad_norm": 2.158277470615684, + "learning_rate": 7.811450508010778e-07, + "loss": 0.7133, + "num_input_tokens_seen": 258715385, + "step": 7954 + }, + { + "epoch": 0.717409929205934, + "flos": 25010223817440.0, + "grad_norm": 4.785335615260814, + "learning_rate": 7.806819161889737e-07, + "loss": 0.7128, + "num_input_tokens_seen": 258744355, + "step": 7955 + }, + { + "epoch": 0.7175001127294043, + "flos": 20012484458880.0, + "grad_norm": 2.0164639405219456, + "learning_rate": 7.802188856158119e-07, + "loss": 0.774, + "num_input_tokens_seen": 258771725, + "step": 7956 + }, + { + "epoch": 0.7175902962528746, + "flos": 21549360318240.0, + "grad_norm": 2.0594363722968736, + "learning_rate": 7.797559591211002e-07, + "loss": 0.7275, + "num_input_tokens_seen": 258799210, + "step": 7957 + }, + { + "epoch": 0.7176804797763449, + "flos": 27779509332960.0, + "grad_norm": 1.8862479749988483, + "learning_rate": 7.79293136744339e-07, + "loss": 0.7061, + "num_input_tokens_seen": 258828780, + "step": 7958 + }, + { + "epoch": 0.7177706632998151, + "flos": 18889371097920.0, + "grad_norm": 1.9830157813801887, + "learning_rate": 7.788304185250185e-07, + "loss": 0.7375, + "num_input_tokens_seen": 258855400, + "step": 7959 + }, + { + "epoch": 0.7178608468232854, + "flos": 21184147260000.0, + "grad_norm": 1.9191273221968599, + "learning_rate": 7.78367804502619e-07, + "loss": 0.7677, + "num_input_tokens_seen": 258882585, + "step": 7960 + }, + { + "epoch": 0.7179510303467557, + "flos": 12586134867840.0, + "grad_norm": 2.449998843912903, + "learning_rate": 7.779052947166156e-07, + "loss": 0.749, + "num_input_tokens_seen": 258907385, + "step": 7961 + }, + { + "epoch": 0.7180412138702259, + "flos": 26212706083200.0, + "grad_norm": 3.990254451124238, + "learning_rate": 7.774428892064697e-07, + "loss": 0.6959, + "num_input_tokens_seen": 258936090, + "step": 7962 + }, + { + "epoch": 0.7181313973936961, + "flos": 24208717652640.0, + "grad_norm": 1.7550522218852034, + "learning_rate": 7.769805880116391e-07, + "loss": 0.7333, + "num_input_tokens_seen": 258963900, + "step": 7963 + }, + { + "epoch": 0.7182215809171665, + "flos": 13716050294880.0, + "grad_norm": 2.393434556540126, + "learning_rate": 7.765183911715678e-07, + "loss": 0.7187, + "num_input_tokens_seen": 258989075, + "step": 7964 + }, + { + "epoch": 0.7183117644406367, + "flos": 70548042292320.0, + "grad_norm": 4.287431250374118, + "learning_rate": 7.760562987256933e-07, + "loss": 0.6411, + "num_input_tokens_seen": 259023990, + "step": 7965 + }, + { + "epoch": 0.718401947964107, + "flos": 17250707234880.0, + "grad_norm": 1.8812417090350486, + "learning_rate": 7.755943107134444e-07, + "loss": 0.7462, + "num_input_tokens_seen": 259051785, + "step": 7966 + }, + { + "epoch": 0.7184921314875772, + "flos": 59095980131040.0, + "grad_norm": 0.633055863856241, + "learning_rate": 7.751324271742401e-07, + "loss": 0.5674, + "num_input_tokens_seen": 259135375, + "step": 7967 + }, + { + "epoch": 0.7185823150110475, + "flos": 25665741400320.0, + "grad_norm": 1.8083047656543632, + "learning_rate": 7.746706481474916e-07, + "loss": 0.7526, + "num_input_tokens_seen": 259166260, + "step": 7968 + }, + { + "epoch": 0.7186724985345178, + "flos": 68629141176000.0, + "grad_norm": 0.5984691999918862, + "learning_rate": 7.742089736725992e-07, + "loss": 0.5361, + "num_input_tokens_seen": 259262425, + "step": 7969 + }, + { + "epoch": 0.718762682057988, + "flos": 24573001466880.0, + "grad_norm": 6.74366180107697, + "learning_rate": 7.737474037889559e-07, + "loss": 0.6595, + "num_input_tokens_seen": 259290800, + "step": 7970 + }, + { + "epoch": 0.7188528655814582, + "flos": 31820015467680.0, + "grad_norm": 1.7614607865286207, + "learning_rate": 7.732859385359458e-07, + "loss": 0.7124, + "num_input_tokens_seen": 259320710, + "step": 7971 + }, + { + "epoch": 0.7189430491049286, + "flos": 26503828342560.0, + "grad_norm": 2.836399513631275, + "learning_rate": 7.728245779529434e-07, + "loss": 0.6609, + "num_input_tokens_seen": 259348615, + "step": 7972 + }, + { + "epoch": 0.7190332326283988, + "flos": 19436298611040.0, + "grad_norm": 1.8218234753397686, + "learning_rate": 7.723633220793146e-07, + "loss": 0.7683, + "num_input_tokens_seen": 259374630, + "step": 7973 + }, + { + "epoch": 0.719123416151869, + "flos": 27087150784320.0, + "grad_norm": 2.389154218630855, + "learning_rate": 7.719021709544162e-07, + "loss": 0.6324, + "num_input_tokens_seen": 259402895, + "step": 7974 + }, + { + "epoch": 0.7192135996753393, + "flos": 23297729343840.0, + "grad_norm": 1.7795527888649254, + "learning_rate": 7.714411246175964e-07, + "loss": 0.8291, + "num_input_tokens_seen": 259431290, + "step": 7975 + }, + { + "epoch": 0.7193037831988096, + "flos": 24753861017280.0, + "grad_norm": 1.7782426175066548, + "learning_rate": 7.709801831081946e-07, + "loss": 0.6586, + "num_input_tokens_seen": 259460200, + "step": 7976 + }, + { + "epoch": 0.7193939667222798, + "flos": 28179853548000.0, + "grad_norm": 2.3914442466252055, + "learning_rate": 7.705193464655391e-07, + "loss": 0.6971, + "num_input_tokens_seen": 259489925, + "step": 7977 + }, + { + "epoch": 0.7194841502457501, + "flos": 23367396941280.0, + "grad_norm": 1.4510629441051572, + "learning_rate": 7.700586147289534e-07, + "loss": 0.7176, + "num_input_tokens_seen": 259519520, + "step": 7978 + }, + { + "epoch": 0.7195743337692203, + "flos": 20055978811680.0, + "grad_norm": 1.734751961097428, + "learning_rate": 7.695979879377481e-07, + "loss": 0.698, + "num_input_tokens_seen": 259546930, + "step": 7979 + }, + { + "epoch": 0.7196645172926907, + "flos": 26359623909120.0, + "grad_norm": 1.8280684633560416, + "learning_rate": 7.691374661312266e-07, + "loss": 0.7593, + "num_input_tokens_seen": 259578210, + "step": 7980 + }, + { + "epoch": 0.7197547008161609, + "flos": 37649485739520.0, + "grad_norm": 1.776510860930638, + "learning_rate": 7.686770493486834e-07, + "loss": 0.7138, + "num_input_tokens_seen": 259610165, + "step": 7981 + }, + { + "epoch": 0.7198448843396311, + "flos": 22641394026240.0, + "grad_norm": 2.408569803035351, + "learning_rate": 7.68216737629404e-07, + "loss": 0.6806, + "num_input_tokens_seen": 259638720, + "step": 7982 + }, + { + "epoch": 0.7199350678631015, + "flos": 26613570674880.0, + "grad_norm": 1.6009702750549422, + "learning_rate": 7.67756531012665e-07, + "loss": 0.8078, + "num_input_tokens_seen": 259669095, + "step": 7983 + }, + { + "epoch": 0.7200252513865717, + "flos": 25008402499200.0, + "grad_norm": 3.0179996924378054, + "learning_rate": 7.67296429537732e-07, + "loss": 0.7974, + "num_input_tokens_seen": 259699150, + "step": 7984 + }, + { + "epoch": 0.7201154349100419, + "flos": 19582287192960.0, + "grad_norm": 2.1752300009299708, + "learning_rate": 7.668364332438661e-07, + "loss": 0.8035, + "num_input_tokens_seen": 259727265, + "step": 7985 + }, + { + "epoch": 0.7202056184335122, + "flos": 22386815374560.0, + "grad_norm": 2.025170839426522, + "learning_rate": 7.663765421703145e-07, + "loss": 0.7514, + "num_input_tokens_seen": 259758020, + "step": 7986 + }, + { + "epoch": 0.7202958019569825, + "flos": 23808745135200.0, + "grad_norm": 1.9583940767106691, + "learning_rate": 7.659167563563187e-07, + "loss": 0.6767, + "num_input_tokens_seen": 259788690, + "step": 7987 + }, + { + "epoch": 0.7203859854804527, + "flos": 19728052756320.0, + "grad_norm": 2.574582341960472, + "learning_rate": 7.654570758411096e-07, + "loss": 0.7621, + "num_input_tokens_seen": 259817825, + "step": 7988 + }, + { + "epoch": 0.720476169003923, + "flos": 19982817256800.0, + "grad_norm": 1.9778608196960228, + "learning_rate": 7.649975006639103e-07, + "loss": 0.8, + "num_input_tokens_seen": 259845465, + "step": 7989 + }, + { + "epoch": 0.7205663525273932, + "flos": 66426934917120.0, + "grad_norm": 0.6229004882276565, + "learning_rate": 7.645380308639337e-07, + "loss": 0.5501, + "num_input_tokens_seen": 259928775, + "step": 7990 + }, + { + "epoch": 0.7206565360508635, + "flos": 23334458800320.0, + "grad_norm": 1.7581434150325352, + "learning_rate": 7.640786664803853e-07, + "loss": 0.7857, + "num_input_tokens_seen": 259956540, + "step": 7991 + }, + { + "epoch": 0.7207467195743338, + "flos": 53210080429920.0, + "grad_norm": 2.0543255505298603, + "learning_rate": 7.636194075524587e-07, + "loss": 0.6622, + "num_input_tokens_seen": 259992640, + "step": 7992 + }, + { + "epoch": 0.720836903097804, + "flos": 29965435236960.0, + "grad_norm": 1.985420912597642, + "learning_rate": 7.631602541193429e-07, + "loss": 0.6942, + "num_input_tokens_seen": 260024095, + "step": 7993 + }, + { + "epoch": 0.7209270866212742, + "flos": 25183909604160.0, + "grad_norm": 2.870725688567946, + "learning_rate": 7.627012062202132e-07, + "loss": 0.7945, + "num_input_tokens_seen": 260049885, + "step": 7994 + }, + { + "epoch": 0.7210172701447446, + "flos": 14517296271360.0, + "grad_norm": 2.34814427741278, + "learning_rate": 7.622422638942391e-07, + "loss": 0.7388, + "num_input_tokens_seen": 260077300, + "step": 7995 + }, + { + "epoch": 0.7211074536682148, + "flos": 24134961381600.0, + "grad_norm": 2.065627068361294, + "learning_rate": 7.617834271805801e-07, + "loss": 0.8043, + "num_input_tokens_seen": 260108095, + "step": 7996 + }, + { + "epoch": 0.7211976371916851, + "flos": 69946197842880.0, + "grad_norm": 0.6901836901256198, + "learning_rate": 7.613246961183863e-07, + "loss": 0.5645, + "num_input_tokens_seen": 260196985, + "step": 7997 + }, + { + "epoch": 0.7212878207151553, + "flos": 22497301102080.0, + "grad_norm": 1.8031859653684932, + "learning_rate": 7.608660707468002e-07, + "loss": 0.6978, + "num_input_tokens_seen": 260227785, + "step": 7998 + }, + { + "epoch": 0.7213780042386256, + "flos": 30982783429920.0, + "grad_norm": 2.567763342793632, + "learning_rate": 7.604075511049522e-07, + "loss": 0.6334, + "num_input_tokens_seen": 260260280, + "step": 7999 + }, + { + "epoch": 0.7214681877620959, + "flos": 32332666728480.0, + "grad_norm": 1.844750278496143, + "learning_rate": 7.599491372319682e-07, + "loss": 0.6851, + "num_input_tokens_seen": 260291985, + "step": 8000 + }, + { + "epoch": 0.7215583712855661, + "flos": 26501337968640.0, + "grad_norm": 1.3779698723570992, + "learning_rate": 7.594908291669601e-07, + "loss": 0.7376, + "num_input_tokens_seen": 260324810, + "step": 8001 + }, + { + "epoch": 0.7216485548090363, + "flos": 19506746773440.0, + "grad_norm": 3.0935530166337273, + "learning_rate": 7.590326269490359e-07, + "loss": 0.7746, + "num_input_tokens_seen": 260353540, + "step": 8002 + }, + { + "epoch": 0.7217387383325067, + "flos": 22679275745280.0, + "grad_norm": 1.572311419655285, + "learning_rate": 7.585745306172899e-07, + "loss": 0.6987, + "num_input_tokens_seen": 260383690, + "step": 8003 + }, + { + "epoch": 0.7218289218559769, + "flos": 21003287709600.0, + "grad_norm": 3.8584952027302286, + "learning_rate": 7.5811654021081e-07, + "loss": 0.7183, + "num_input_tokens_seen": 260411765, + "step": 8004 + }, + { + "epoch": 0.7219191053794471, + "flos": 24821707296480.0, + "grad_norm": 1.6213512537858896, + "learning_rate": 7.576586557686748e-07, + "loss": 0.7692, + "num_input_tokens_seen": 260440095, + "step": 8005 + }, + { + "epoch": 0.7220092889029175, + "flos": 30731401377600.0, + "grad_norm": 2.078900529523145, + "learning_rate": 7.572008773299531e-07, + "loss": 0.7117, + "num_input_tokens_seen": 260471085, + "step": 8006 + }, + { + "epoch": 0.7220994724263877, + "flos": 29709295455360.0, + "grad_norm": 1.7407524459451336, + "learning_rate": 7.567432049337055e-07, + "loss": 0.7757, + "num_input_tokens_seen": 260502345, + "step": 8007 + }, + { + "epoch": 0.722189655949858, + "flos": 21440249871840.0, + "grad_norm": 2.121977283108105, + "learning_rate": 7.562856386189834e-07, + "loss": 0.6758, + "num_input_tokens_seen": 260530410, + "step": 8008 + }, + { + "epoch": 0.7222798394733282, + "flos": 21841263142560.0, + "grad_norm": 3.112882543135214, + "learning_rate": 7.558281784248275e-07, + "loss": 0.7512, + "num_input_tokens_seen": 260558375, + "step": 8009 + }, + { + "epoch": 0.7223700229967985, + "flos": 19654222145760.0, + "grad_norm": 1.907248751291018, + "learning_rate": 7.553708243902721e-07, + "loss": 0.7267, + "num_input_tokens_seen": 260584320, + "step": 8010 + }, + { + "epoch": 0.7224602065202688, + "flos": 31350524031840.0, + "grad_norm": 1.7266430728115356, + "learning_rate": 7.549135765543404e-07, + "loss": 0.6278, + "num_input_tokens_seen": 260617350, + "step": 8011 + }, + { + "epoch": 0.722550390043739, + "flos": 20894028584160.0, + "grad_norm": 3.1353366034710577, + "learning_rate": 7.544564349560481e-07, + "loss": 0.7624, + "num_input_tokens_seen": 260642580, + "step": 8012 + }, + { + "epoch": 0.7226405735672092, + "flos": 28036057981920.0, + "grad_norm": 13.157665475554603, + "learning_rate": 7.539993996344009e-07, + "loss": 0.723, + "num_input_tokens_seen": 260674490, + "step": 8013 + }, + { + "epoch": 0.7227307570906796, + "flos": 21659028311040.0, + "grad_norm": 1.815392586676269, + "learning_rate": 7.535424706283941e-07, + "loss": 0.7169, + "num_input_tokens_seen": 260702605, + "step": 8014 + }, + { + "epoch": 0.7228209406141498, + "flos": 26136831135840.0, + "grad_norm": 2.2580223118006604, + "learning_rate": 7.530856479770181e-07, + "loss": 0.6987, + "num_input_tokens_seen": 260732400, + "step": 8015 + }, + { + "epoch": 0.72291112413762, + "flos": 27340577173440.0, + "grad_norm": 1.6431127669211678, + "learning_rate": 7.526289317192484e-07, + "loss": 0.8172, + "num_input_tokens_seen": 260759960, + "step": 8016 + }, + { + "epoch": 0.7230013076610903, + "flos": 18780037632960.0, + "grad_norm": 2.1931397896260103, + "learning_rate": 7.521723218940579e-07, + "loss": 0.735, + "num_input_tokens_seen": 260785050, + "step": 8017 + }, + { + "epoch": 0.7230914911845606, + "flos": 22460125608480.0, + "grad_norm": 2.8212516256656452, + "learning_rate": 7.517158185404038e-07, + "loss": 0.799, + "num_input_tokens_seen": 260812365, + "step": 8018 + }, + { + "epoch": 0.7231816747080309, + "flos": 26758927370880.0, + "grad_norm": 2.1717152708308274, + "learning_rate": 7.512594216972403e-07, + "loss": 0.7626, + "num_input_tokens_seen": 260843900, + "step": 8019 + }, + { + "epoch": 0.7232718582315011, + "flos": 22021193448960.0, + "grad_norm": 1.638474790158588, + "learning_rate": 7.508031314035078e-07, + "loss": 0.7736, + "num_input_tokens_seen": 260873330, + "step": 8020 + }, + { + "epoch": 0.7233620417549713, + "flos": 22860804351360.0, + "grad_norm": 1.8323100514345307, + "learning_rate": 7.503469476981401e-07, + "loss": 0.7589, + "num_input_tokens_seen": 260901490, + "step": 8021 + }, + { + "epoch": 0.7234522252784417, + "flos": 34190963935200.0, + "grad_norm": 2.9218919722936207, + "learning_rate": 7.498908706200613e-07, + "loss": 0.6112, + "num_input_tokens_seen": 260932195, + "step": 8022 + }, + { + "epoch": 0.7235424088019119, + "flos": 22935080999040.0, + "grad_norm": 1.7031116310084877, + "learning_rate": 7.494349002081866e-07, + "loss": 0.7865, + "num_input_tokens_seen": 260962045, + "step": 8023 + }, + { + "epoch": 0.7236325923253821, + "flos": 20126129616000.0, + "grad_norm": 2.391038646939281, + "learning_rate": 7.489790365014224e-07, + "loss": 0.766, + "num_input_tokens_seen": 260990360, + "step": 8024 + }, + { + "epoch": 0.7237227758488524, + "flos": 27743597611200.0, + "grad_norm": 2.659491502157664, + "learning_rate": 7.485232795386642e-07, + "loss": 0.757, + "num_input_tokens_seen": 261021300, + "step": 8025 + }, + { + "epoch": 0.7238129593723227, + "flos": 22858202468160.0, + "grad_norm": 1.9256146031946841, + "learning_rate": 7.480676293588002e-07, + "loss": 0.762, + "num_input_tokens_seen": 261049125, + "step": 8026 + }, + { + "epoch": 0.7239031428957929, + "flos": 67559526566880.0, + "grad_norm": 0.7625747956659631, + "learning_rate": 7.476120860007093e-07, + "loss": 0.5958, + "num_input_tokens_seen": 261133175, + "step": 8027 + }, + { + "epoch": 0.7239933264192632, + "flos": 22932070248480.0, + "grad_norm": 2.1994467256324404, + "learning_rate": 7.471566495032608e-07, + "loss": 0.7803, + "num_input_tokens_seen": 261156875, + "step": 8028 + }, + { + "epoch": 0.7240835099427335, + "flos": 37979344622400.0, + "grad_norm": 2.0235967978439136, + "learning_rate": 7.467013199053152e-07, + "loss": 0.6737, + "num_input_tokens_seen": 261187110, + "step": 8029 + }, + { + "epoch": 0.7241736934662037, + "flos": 19071866117760.0, + "grad_norm": 1.9934662548239195, + "learning_rate": 7.46246097245724e-07, + "loss": 0.7359, + "num_input_tokens_seen": 261213825, + "step": 8030 + }, + { + "epoch": 0.724263876989674, + "flos": 21870707326080.0, + "grad_norm": 2.061836945274215, + "learning_rate": 7.457909815633276e-07, + "loss": 0.7419, + "num_input_tokens_seen": 261240155, + "step": 8031 + }, + { + "epoch": 0.7243540605131442, + "flos": 31386918960480.0, + "grad_norm": 2.29383254149558, + "learning_rate": 7.453359728969618e-07, + "loss": 0.7314, + "num_input_tokens_seen": 261269005, + "step": 8032 + }, + { + "epoch": 0.7244442440366146, + "flos": 35794385132160.0, + "grad_norm": 1.6843917619370128, + "learning_rate": 7.448810712854475e-07, + "loss": 0.6654, + "num_input_tokens_seen": 261301810, + "step": 8033 + }, + { + "epoch": 0.7245344275600848, + "flos": 21804533686080.0, + "grad_norm": 1.7518928372866251, + "learning_rate": 7.444262767676022e-07, + "loss": 0.7786, + "num_input_tokens_seen": 261330235, + "step": 8034 + }, + { + "epoch": 0.724624611083555, + "flos": 22781026579200.0, + "grad_norm": 1.8967663352044686, + "learning_rate": 7.439715893822296e-07, + "loss": 0.812, + "num_input_tokens_seen": 261357205, + "step": 8035 + }, + { + "epoch": 0.7247147946070253, + "flos": 18562969002720.0, + "grad_norm": 7.292207127477871, + "learning_rate": 7.435170091681264e-07, + "loss": 0.7617, + "num_input_tokens_seen": 261386780, + "step": 8036 + }, + { + "epoch": 0.7248049781304956, + "flos": 24969257008320.0, + "grad_norm": 2.3890800719927885, + "learning_rate": 7.430625361640803e-07, + "loss": 0.6967, + "num_input_tokens_seen": 261415345, + "step": 8037 + }, + { + "epoch": 0.7248951616539658, + "flos": 20638334839680.0, + "grad_norm": 2.3097093600783296, + "learning_rate": 7.426081704088694e-07, + "loss": 0.7823, + "num_input_tokens_seen": 261442500, + "step": 8038 + }, + { + "epoch": 0.7249853451774361, + "flos": 18962495483040.0, + "grad_norm": 1.9188226170420228, + "learning_rate": 7.42153911941263e-07, + "loss": 0.8096, + "num_input_tokens_seen": 261469350, + "step": 8039 + }, + { + "epoch": 0.7250755287009063, + "flos": 18121583639040.0, + "grad_norm": 1.7709874777930061, + "learning_rate": 7.416997608000192e-07, + "loss": 0.78, + "num_input_tokens_seen": 261496165, + "step": 8040 + }, + { + "epoch": 0.7251657122243766, + "flos": 19108000858080.0, + "grad_norm": 2.06325297626968, + "learning_rate": 7.412457170238918e-07, + "loss": 0.7215, + "num_input_tokens_seen": 261522760, + "step": 8041 + }, + { + "epoch": 0.7252558957478469, + "flos": 21549917864640.0, + "grad_norm": 1.831421122188309, + "learning_rate": 7.407917806516193e-07, + "loss": 0.6859, + "num_input_tokens_seen": 261551405, + "step": 8042 + }, + { + "epoch": 0.7253460792713171, + "flos": 22379753120160.0, + "grad_norm": 2.3179540496761173, + "learning_rate": 7.403379517219354e-07, + "loss": 0.7318, + "num_input_tokens_seen": 261579260, + "step": 8043 + }, + { + "epoch": 0.7254362627947873, + "flos": 24317976778080.0, + "grad_norm": 2.215365878126795, + "learning_rate": 7.398842302735636e-07, + "loss": 0.7668, + "num_input_tokens_seen": 261604595, + "step": 8044 + }, + { + "epoch": 0.7255264463182577, + "flos": 67731242356320.0, + "grad_norm": 0.6675913229631562, + "learning_rate": 7.394306163452171e-07, + "loss": 0.5453, + "num_input_tokens_seen": 261695115, + "step": 8045 + }, + { + "epoch": 0.7256166298417279, + "flos": 24317902438560.0, + "grad_norm": 2.195710169928989, + "learning_rate": 7.38977109975601e-07, + "loss": 0.7169, + "num_input_tokens_seen": 261725155, + "step": 8046 + }, + { + "epoch": 0.7257068133651982, + "flos": 28470715619040.0, + "grad_norm": 2.444936705162024, + "learning_rate": 7.385237112034119e-07, + "loss": 0.6507, + "num_input_tokens_seen": 261753800, + "step": 8047 + }, + { + "epoch": 0.7257969968886684, + "flos": 17723804137440.0, + "grad_norm": 2.2163236003602163, + "learning_rate": 7.380704200673342e-07, + "loss": 0.8031, + "num_input_tokens_seen": 261781375, + "step": 8048 + }, + { + "epoch": 0.7258871804121387, + "flos": 24535194087360.0, + "grad_norm": 1.5131364823054998, + "learning_rate": 7.376172366060478e-07, + "loss": 0.7636, + "num_input_tokens_seen": 261810865, + "step": 8049 + }, + { + "epoch": 0.725977363935609, + "flos": 53154128473920.0, + "grad_norm": 0.6638582252016444, + "learning_rate": 7.371641608582187e-07, + "loss": 0.5114, + "num_input_tokens_seen": 261896210, + "step": 8050 + }, + { + "epoch": 0.7260675474590792, + "flos": 20272527065280.0, + "grad_norm": 2.682925211827706, + "learning_rate": 7.367111928625067e-07, + "loss": 0.6066, + "num_input_tokens_seen": 261923995, + "step": 8051 + }, + { + "epoch": 0.7261577309825495, + "flos": 27888508270080.0, + "grad_norm": 1.9278460531148482, + "learning_rate": 7.362583326575613e-07, + "loss": 0.6515, + "num_input_tokens_seen": 261953365, + "step": 8052 + }, + { + "epoch": 0.7262479145060198, + "flos": 23808336267840.0, + "grad_norm": 2.2242131202661346, + "learning_rate": 7.358055802820234e-07, + "loss": 0.7285, + "num_input_tokens_seen": 261981395, + "step": 8053 + }, + { + "epoch": 0.72633809802949, + "flos": 27043470582720.0, + "grad_norm": 2.8907257887306304, + "learning_rate": 7.353529357745245e-07, + "loss": 0.7082, + "num_input_tokens_seen": 262009570, + "step": 8054 + }, + { + "epoch": 0.7264282815529602, + "flos": 20419779419040.0, + "grad_norm": 2.1027948673057213, + "learning_rate": 7.349003991736851e-07, + "loss": 0.7781, + "num_input_tokens_seen": 262035300, + "step": 8055 + }, + { + "epoch": 0.7265184650764306, + "flos": 29090395819680.0, + "grad_norm": 1.8589822049738465, + "learning_rate": 7.344479705181206e-07, + "loss": 0.5889, + "num_input_tokens_seen": 262063025, + "step": 8056 + }, + { + "epoch": 0.7266086485999008, + "flos": 22934114585280.0, + "grad_norm": 2.87380840652004, + "learning_rate": 7.339956498464322e-07, + "loss": 0.6895, + "num_input_tokens_seen": 262092790, + "step": 8057 + }, + { + "epoch": 0.726698832123371, + "flos": 20565582152160.0, + "grad_norm": 1.9471793045608583, + "learning_rate": 7.335434371972169e-07, + "loss": 0.7859, + "num_input_tokens_seen": 262120850, + "step": 8058 + }, + { + "epoch": 0.7267890156468413, + "flos": 23626027096800.0, + "grad_norm": 1.7277607621884554, + "learning_rate": 7.33091332609058e-07, + "loss": 0.7279, + "num_input_tokens_seen": 262149015, + "step": 8059 + }, + { + "epoch": 0.7268791991703116, + "flos": 18407948169120.0, + "grad_norm": 2.26963209758728, + "learning_rate": 7.326393361205323e-07, + "loss": 0.7556, + "num_input_tokens_seen": 262175205, + "step": 8060 + }, + { + "epoch": 0.7269693826937819, + "flos": 14955224847360.0, + "grad_norm": 1.9652831418608605, + "learning_rate": 7.321874477702068e-07, + "loss": 0.8191, + "num_input_tokens_seen": 262201715, + "step": 8061 + }, + { + "epoch": 0.7270595662172521, + "flos": 69163170782400.0, + "grad_norm": 0.7537350633194565, + "learning_rate": 7.317356675966386e-07, + "loss": 0.5911, + "num_input_tokens_seen": 262291740, + "step": 8062 + }, + { + "epoch": 0.7271497497407223, + "flos": 20237247229440.0, + "grad_norm": 1.8143129161130014, + "learning_rate": 7.312839956383765e-07, + "loss": 0.7718, + "num_input_tokens_seen": 262322020, + "step": 8063 + }, + { + "epoch": 0.7272399332641927, + "flos": 22021639486080.0, + "grad_norm": 1.8763098909602547, + "learning_rate": 7.308324319339603e-07, + "loss": 0.7182, + "num_input_tokens_seen": 262351970, + "step": 8064 + }, + { + "epoch": 0.7273301167876629, + "flos": 31890129102240.0, + "grad_norm": 2.8594271962555777, + "learning_rate": 7.303809765219182e-07, + "loss": 0.741, + "num_input_tokens_seen": 262382480, + "step": 8065 + }, + { + "epoch": 0.7274203003111331, + "flos": 24536978235840.0, + "grad_norm": 1.7386090944569015, + "learning_rate": 7.299296294407719e-07, + "loss": 0.7857, + "num_input_tokens_seen": 262410250, + "step": 8066 + }, + { + "epoch": 0.7275104838346034, + "flos": 25081080847200.0, + "grad_norm": 1.994046947883108, + "learning_rate": 7.294783907290327e-07, + "loss": 0.7802, + "num_input_tokens_seen": 262439105, + "step": 8067 + }, + { + "epoch": 0.7276006673580737, + "flos": 37287692299200.0, + "grad_norm": 2.3220741443279955, + "learning_rate": 7.290272604252028e-07, + "loss": 0.6061, + "num_input_tokens_seen": 262472730, + "step": 8068 + }, + { + "epoch": 0.727690850881544, + "flos": 37941946110240.0, + "grad_norm": 1.9729050281131688, + "learning_rate": 7.285762385677758e-07, + "loss": 0.6448, + "num_input_tokens_seen": 262501640, + "step": 8069 + }, + { + "epoch": 0.7277810344050142, + "flos": 20490413430240.0, + "grad_norm": 1.8885371762737986, + "learning_rate": 7.281253251952335e-07, + "loss": 0.7784, + "num_input_tokens_seen": 262530375, + "step": 8070 + }, + { + "epoch": 0.7278712179284844, + "flos": 22492803561120.0, + "grad_norm": 1.6540811121008956, + "learning_rate": 7.276745203460526e-07, + "loss": 0.7194, + "num_input_tokens_seen": 262559950, + "step": 8071 + }, + { + "epoch": 0.7279614014519548, + "flos": 69868650256320.0, + "grad_norm": 0.7226999218091003, + "learning_rate": 7.272238240586959e-07, + "loss": 0.6315, + "num_input_tokens_seen": 262653920, + "step": 8072 + }, + { + "epoch": 0.728051584975425, + "flos": 31204349601120.0, + "grad_norm": 2.3669836378285436, + "learning_rate": 7.267732363716219e-07, + "loss": 0.6744, + "num_input_tokens_seen": 262687305, + "step": 8073 + }, + { + "epoch": 0.7281417684988952, + "flos": 15501408965280.0, + "grad_norm": 1.6204953479505864, + "learning_rate": 7.263227573232753e-07, + "loss": 0.6938, + "num_input_tokens_seen": 262715670, + "step": 8074 + }, + { + "epoch": 0.7282319520223655, + "flos": 26577175746240.0, + "grad_norm": 3.0549330356582214, + "learning_rate": 7.258723869520937e-07, + "loss": 0.828, + "num_input_tokens_seen": 262742790, + "step": 8075 + }, + { + "epoch": 0.7283221355458358, + "flos": 19796642430720.0, + "grad_norm": 1.6539567870292093, + "learning_rate": 7.254221252965059e-07, + "loss": 0.7837, + "num_input_tokens_seen": 262770620, + "step": 8076 + }, + { + "epoch": 0.728412319069306, + "flos": 25374210273600.0, + "grad_norm": 2.224265322316575, + "learning_rate": 7.249719723949301e-07, + "loss": 0.6559, + "num_input_tokens_seen": 262799485, + "step": 8077 + }, + { + "epoch": 0.7285025025927763, + "flos": 23295759346560.0, + "grad_norm": 1.7303293830774071, + "learning_rate": 7.245219282857761e-07, + "loss": 0.7614, + "num_input_tokens_seen": 262829110, + "step": 8078 + }, + { + "epoch": 0.7285926861162466, + "flos": 27044585675520.0, + "grad_norm": 2.011326308998804, + "learning_rate": 7.240719930074442e-07, + "loss": 0.7796, + "num_input_tokens_seen": 262857910, + "step": 8079 + }, + { + "epoch": 0.7286828696397168, + "flos": 23655731468640.0, + "grad_norm": 10.465400729391963, + "learning_rate": 7.236221665983257e-07, + "loss": 0.7674, + "num_input_tokens_seen": 262885275, + "step": 8080 + }, + { + "epoch": 0.7287730531631871, + "flos": 61491274791360.0, + "grad_norm": 0.7242963171183117, + "learning_rate": 7.231724490968012e-07, + "loss": 0.5566, + "num_input_tokens_seen": 262973625, + "step": 8081 + }, + { + "epoch": 0.7288632366866573, + "flos": 23079954488160.0, + "grad_norm": 1.8857501527133047, + "learning_rate": 7.227228405412438e-07, + "loss": 0.7088, + "num_input_tokens_seen": 263001905, + "step": 8082 + }, + { + "epoch": 0.7289534202101277, + "flos": 19830286797120.0, + "grad_norm": 1.9251631234152113, + "learning_rate": 7.222733409700165e-07, + "loss": 0.6974, + "num_input_tokens_seen": 263029795, + "step": 8083 + }, + { + "epoch": 0.7290436037335979, + "flos": 26650411640640.0, + "grad_norm": 1.5750304748977684, + "learning_rate": 7.21823950421473e-07, + "loss": 0.7498, + "num_input_tokens_seen": 263062200, + "step": 8084 + }, + { + "epoch": 0.7291337872570681, + "flos": 21148161198720.0, + "grad_norm": 2.001013022407582, + "learning_rate": 7.213746689339577e-07, + "loss": 0.7076, + "num_input_tokens_seen": 263089845, + "step": 8085 + }, + { + "epoch": 0.7292239707805384, + "flos": 29747363023200.0, + "grad_norm": 2.506195783147597, + "learning_rate": 7.20925496545807e-07, + "loss": 0.7422, + "num_input_tokens_seen": 263120215, + "step": 8086 + }, + { + "epoch": 0.7293141543040087, + "flos": 58710026346720.0, + "grad_norm": 1.73136516972941, + "learning_rate": 7.20476433295344e-07, + "loss": 0.6469, + "num_input_tokens_seen": 263154360, + "step": 8087 + }, + { + "epoch": 0.7294043378274789, + "flos": 25374842159520.0, + "grad_norm": 4.390786164607717, + "learning_rate": 7.200274792208882e-07, + "loss": 0.7786, + "num_input_tokens_seen": 263185160, + "step": 8088 + }, + { + "epoch": 0.7294945213509492, + "flos": 27050718685920.0, + "grad_norm": 2.0128154246820276, + "learning_rate": 7.195786343607444e-07, + "loss": 0.6292, + "num_input_tokens_seen": 263214310, + "step": 8089 + }, + { + "epoch": 0.7295847048744194, + "flos": 21658284915840.0, + "grad_norm": 1.9146461686818714, + "learning_rate": 7.191298987532131e-07, + "loss": 0.7325, + "num_input_tokens_seen": 263243065, + "step": 8090 + }, + { + "epoch": 0.7296748883978897, + "flos": 27779360653920.0, + "grad_norm": 1.8724196162369158, + "learning_rate": 7.186812724365805e-07, + "loss": 0.7449, + "num_input_tokens_seen": 263275955, + "step": 8091 + }, + { + "epoch": 0.72976507192136, + "flos": 24755384977440.0, + "grad_norm": 4.692664594654598, + "learning_rate": 7.182327554491272e-07, + "loss": 0.7466, + "num_input_tokens_seen": 263303475, + "step": 8092 + }, + { + "epoch": 0.7298552554448302, + "flos": 62473566167040.0, + "grad_norm": 0.6877508789425735, + "learning_rate": 7.177843478291225e-07, + "loss": 0.5863, + "num_input_tokens_seen": 263398145, + "step": 8093 + }, + { + "epoch": 0.7299454389683004, + "flos": 39255620328960.0, + "grad_norm": 2.1062483609246914, + "learning_rate": 7.173360496148276e-07, + "loss": 0.7647, + "num_input_tokens_seen": 263428570, + "step": 8094 + }, + { + "epoch": 0.7300356224917708, + "flos": 27524112946560.0, + "grad_norm": 1.9959386477893393, + "learning_rate": 7.168878608444939e-07, + "loss": 0.7531, + "num_input_tokens_seen": 263458525, + "step": 8095 + }, + { + "epoch": 0.730125806015241, + "flos": 21877583731680.0, + "grad_norm": 2.6171503320588414, + "learning_rate": 7.164397815563623e-07, + "loss": 0.7435, + "num_input_tokens_seen": 263485125, + "step": 8096 + }, + { + "epoch": 0.7302159895387113, + "flos": 25775074865280.0, + "grad_norm": 2.3123433859089118, + "learning_rate": 7.159918117886661e-07, + "loss": 0.7458, + "num_input_tokens_seen": 263512430, + "step": 8097 + }, + { + "epoch": 0.7303061730621815, + "flos": 16121126335680.0, + "grad_norm": 4.246062755032918, + "learning_rate": 7.155439515796284e-07, + "loss": 0.7566, + "num_input_tokens_seen": 263538945, + "step": 8098 + }, + { + "epoch": 0.7303963565856518, + "flos": 15647397547200.0, + "grad_norm": 2.4837892429609414, + "learning_rate": 7.150962009674633e-07, + "loss": 0.7407, + "num_input_tokens_seen": 263567520, + "step": 8099 + }, + { + "epoch": 0.7304865401091221, + "flos": 26680673558880.0, + "grad_norm": 76.53733961582901, + "learning_rate": 7.146485599903751e-07, + "loss": 0.7889, + "num_input_tokens_seen": 263594290, + "step": 8100 + }, + { + "epoch": 0.7305767236325923, + "flos": 23078579207040.0, + "grad_norm": 1.7550833305724882, + "learning_rate": 7.142010286865592e-07, + "loss": 0.8122, + "num_input_tokens_seen": 263623695, + "step": 8101 + }, + { + "epoch": 0.7306669071560626, + "flos": 66227633432160.0, + "grad_norm": 0.5496424542350075, + "learning_rate": 7.137536070942012e-07, + "loss": 0.5469, + "num_input_tokens_seen": 263712975, + "step": 8102 + }, + { + "epoch": 0.7307570906795329, + "flos": 25374767820000.0, + "grad_norm": 2.045631205511194, + "learning_rate": 7.133062952514786e-07, + "loss": 0.7348, + "num_input_tokens_seen": 263745725, + "step": 8103 + }, + { + "epoch": 0.7308472742030031, + "flos": 65313337014720.0, + "grad_norm": 0.6448699292543493, + "learning_rate": 7.128590931965562e-07, + "loss": 0.6023, + "num_input_tokens_seen": 263843400, + "step": 8104 + }, + { + "epoch": 0.7309374577264733, + "flos": 22022977597440.0, + "grad_norm": 2.1354789414668183, + "learning_rate": 7.124120009675945e-07, + "loss": 0.8031, + "num_input_tokens_seen": 263873160, + "step": 8105 + }, + { + "epoch": 0.7310276412499437, + "flos": 24683003987520.0, + "grad_norm": 1.719290284288776, + "learning_rate": 7.119650186027399e-07, + "loss": 0.7664, + "num_input_tokens_seen": 263900630, + "step": 8106 + }, + { + "epoch": 0.7311178247734139, + "flos": 24528875228160.0, + "grad_norm": 2.1047086741537084, + "learning_rate": 7.11518146140132e-07, + "loss": 0.8605, + "num_input_tokens_seen": 263927335, + "step": 8107 + }, + { + "epoch": 0.7312080082968841, + "flos": 17067691838400.0, + "grad_norm": 2.903234414460468, + "learning_rate": 7.110713836179007e-07, + "loss": 0.7005, + "num_input_tokens_seen": 263954310, + "step": 8108 + }, + { + "epoch": 0.7312981918203544, + "flos": 20311077840000.0, + "grad_norm": 1.8269556984711002, + "learning_rate": 7.106247310741659e-07, + "loss": 0.6669, + "num_input_tokens_seen": 263982050, + "step": 8109 + }, + { + "epoch": 0.7313883753438247, + "flos": 23698519596000.0, + "grad_norm": 1.905157119899699, + "learning_rate": 7.101781885470393e-07, + "loss": 0.8029, + "num_input_tokens_seen": 264010380, + "step": 8110 + }, + { + "epoch": 0.731478558867295, + "flos": 23730045286080.0, + "grad_norm": 2.4987185054776977, + "learning_rate": 7.097317560746203e-07, + "loss": 0.7792, + "num_input_tokens_seen": 264038710, + "step": 8111 + }, + { + "epoch": 0.7315687423907652, + "flos": 26025713522400.0, + "grad_norm": 3.243930871878419, + "learning_rate": 7.092854336950036e-07, + "loss": 0.7993, + "num_input_tokens_seen": 264067640, + "step": 8112 + }, + { + "epoch": 0.7316589259142354, + "flos": 24860443920000.0, + "grad_norm": 1.6718490455965835, + "learning_rate": 7.0883922144627e-07, + "loss": 0.7965, + "num_input_tokens_seen": 264094985, + "step": 8113 + }, + { + "epoch": 0.7317491094377058, + "flos": 21767581211040.0, + "grad_norm": 1.7966668713321228, + "learning_rate": 7.083931193664934e-07, + "loss": 0.7798, + "num_input_tokens_seen": 264120610, + "step": 8114 + }, + { + "epoch": 0.731839292961176, + "flos": 23258844041280.0, + "grad_norm": 2.1133940661314834, + "learning_rate": 7.079471274937378e-07, + "loss": 0.789, + "num_input_tokens_seen": 264149870, + "step": 8115 + }, + { + "epoch": 0.7319294764846462, + "flos": 28503096213600.0, + "grad_norm": 3.15229065577383, + "learning_rate": 7.075012458660574e-07, + "loss": 0.7269, + "num_input_tokens_seen": 264178770, + "step": 8116 + }, + { + "epoch": 0.7320196600081165, + "flos": 60761592070080.0, + "grad_norm": 0.6566699137784904, + "learning_rate": 7.070554745214976e-07, + "loss": 0.5747, + "num_input_tokens_seen": 264276625, + "step": 8117 + }, + { + "epoch": 0.7321098435315868, + "flos": 27415708725600.0, + "grad_norm": 2.256023926121067, + "learning_rate": 7.066098134980947e-07, + "loss": 0.6995, + "num_input_tokens_seen": 264307265, + "step": 8118 + }, + { + "epoch": 0.732200027055057, + "flos": 22821324332640.0, + "grad_norm": 2.7104406423678715, + "learning_rate": 7.061642628338727e-07, + "loss": 0.7436, + "num_input_tokens_seen": 264336700, + "step": 8119 + }, + { + "epoch": 0.7322902105785273, + "flos": 23585023117920.0, + "grad_norm": 2.2366456984869783, + "learning_rate": 7.057188225668513e-07, + "loss": 0.7398, + "num_input_tokens_seen": 264367315, + "step": 8120 + }, + { + "epoch": 0.7323803941019975, + "flos": 26248840823520.0, + "grad_norm": 2.0332863797045633, + "learning_rate": 7.052734927350358e-07, + "loss": 0.89, + "num_input_tokens_seen": 264397675, + "step": 8121 + }, + { + "epoch": 0.7324705776254679, + "flos": 25042418563200.0, + "grad_norm": 3.218869994065664, + "learning_rate": 7.048282733764252e-07, + "loss": 0.7647, + "num_input_tokens_seen": 264426900, + "step": 8122 + }, + { + "epoch": 0.7325607611489381, + "flos": 22131753516000.0, + "grad_norm": 2.264631698171479, + "learning_rate": 7.043831645290077e-07, + "loss": 0.8224, + "num_input_tokens_seen": 264454415, + "step": 8123 + }, + { + "epoch": 0.7326509446724083, + "flos": 17396509968000.0, + "grad_norm": 2.7435829546223025, + "learning_rate": 7.039381662307624e-07, + "loss": 0.712, + "num_input_tokens_seen": 264480710, + "step": 8124 + }, + { + "epoch": 0.7327411281958787, + "flos": 22856678508000.0, + "grad_norm": 2.549344042014664, + "learning_rate": 7.034932785196601e-07, + "loss": 0.7235, + "num_input_tokens_seen": 264511155, + "step": 8125 + }, + { + "epoch": 0.7328313117193489, + "flos": 21768659134080.0, + "grad_norm": 5.624735847204745, + "learning_rate": 7.030485014336585e-07, + "loss": 0.7262, + "num_input_tokens_seen": 264538895, + "step": 8126 + }, + { + "epoch": 0.7329214952428191, + "flos": 16666864416480.0, + "grad_norm": 2.1951899918553286, + "learning_rate": 7.026038350107118e-07, + "loss": 0.7002, + "num_input_tokens_seen": 264565140, + "step": 8127 + }, + { + "epoch": 0.7330116787662894, + "flos": 29416277538240.0, + "grad_norm": 1.6615261753625175, + "learning_rate": 7.021592792887579e-07, + "loss": 0.8711, + "num_input_tokens_seen": 264594930, + "step": 8128 + }, + { + "epoch": 0.7331018622897597, + "flos": 23698110728640.0, + "grad_norm": 1.7449011773051193, + "learning_rate": 7.01714834305732e-07, + "loss": 0.7944, + "num_input_tokens_seen": 264626195, + "step": 8129 + }, + { + "epoch": 0.7331920458132299, + "flos": 28720648050720.0, + "grad_norm": 2.228010382787321, + "learning_rate": 7.012705000995544e-07, + "loss": 0.7007, + "num_input_tokens_seen": 264656390, + "step": 8130 + }, + { + "epoch": 0.7332822293367002, + "flos": 22600204198560.0, + "grad_norm": 1.7386529760762703, + "learning_rate": 7.008262767081392e-07, + "loss": 0.7361, + "num_input_tokens_seen": 264685630, + "step": 8131 + }, + { + "epoch": 0.7333724128601704, + "flos": 22677974803680.0, + "grad_norm": 4.028868689000426, + "learning_rate": 7.003821641693892e-07, + "loss": 0.7967, + "num_input_tokens_seen": 264710905, + "step": 8132 + }, + { + "epoch": 0.7334625963836408, + "flos": 21440918927520.0, + "grad_norm": 1.753925824640211, + "learning_rate": 6.999381625211993e-07, + "loss": 0.7312, + "num_input_tokens_seen": 264740025, + "step": 8133 + }, + { + "epoch": 0.733552779907111, + "flos": 28101934263840.0, + "grad_norm": 1.893361341925202, + "learning_rate": 6.994942718014536e-07, + "loss": 0.6644, + "num_input_tokens_seen": 264772805, + "step": 8134 + }, + { + "epoch": 0.7336429634305812, + "flos": 15901269973440.0, + "grad_norm": 2.5281086019354335, + "learning_rate": 6.990504920480282e-07, + "loss": 0.6545, + "num_input_tokens_seen": 264799795, + "step": 8135 + }, + { + "epoch": 0.7337331469540515, + "flos": 15902124877920.0, + "grad_norm": 4.048043218660898, + "learning_rate": 6.986068232987879e-07, + "loss": 0.7392, + "num_input_tokens_seen": 264827185, + "step": 8136 + }, + { + "epoch": 0.7338233304775218, + "flos": 24899700920160.0, + "grad_norm": 1.5176867589767702, + "learning_rate": 6.981632655915888e-07, + "loss": 0.6961, + "num_input_tokens_seen": 264856880, + "step": 8137 + }, + { + "epoch": 0.733913514000992, + "flos": 25847939062080.0, + "grad_norm": 1.837737933473564, + "learning_rate": 6.977198189642783e-07, + "loss": 0.8043, + "num_input_tokens_seen": 264886685, + "step": 8138 + }, + { + "epoch": 0.7340036975244623, + "flos": 21440138362560.0, + "grad_norm": 2.061599766173403, + "learning_rate": 6.972764834546935e-07, + "loss": 0.7697, + "num_input_tokens_seen": 264914930, + "step": 8139 + }, + { + "epoch": 0.7340938810479325, + "flos": 23079954488160.0, + "grad_norm": 1.846627334022067, + "learning_rate": 6.96833259100663e-07, + "loss": 0.759, + "num_input_tokens_seen": 264944560, + "step": 8140 + }, + { + "epoch": 0.7341840645714028, + "flos": 19690542734880.0, + "grad_norm": 4.531589048131733, + "learning_rate": 6.96390145940003e-07, + "loss": 0.7171, + "num_input_tokens_seen": 264971335, + "step": 8141 + }, + { + "epoch": 0.7342742480948731, + "flos": 23333603895840.0, + "grad_norm": 2.025617801944193, + "learning_rate": 6.959471440105253e-07, + "loss": 0.6955, + "num_input_tokens_seen": 265002045, + "step": 8142 + }, + { + "epoch": 0.7343644316183433, + "flos": 19217557341600.0, + "grad_norm": 2.9260167433285083, + "learning_rate": 6.955042533500261e-07, + "loss": 0.7748, + "num_input_tokens_seen": 265029120, + "step": 8143 + }, + { + "epoch": 0.7344546151418135, + "flos": 22305141944640.0, + "grad_norm": 1.9028954303253376, + "learning_rate": 6.950614739962986e-07, + "loss": 0.7003, + "num_input_tokens_seen": 265054765, + "step": 8144 + }, + { + "epoch": 0.7345447986652839, + "flos": 24751444982880.0, + "grad_norm": 1.8866063433979388, + "learning_rate": 6.946188059871198e-07, + "loss": 0.6315, + "num_input_tokens_seen": 265084565, + "step": 8145 + }, + { + "epoch": 0.7346349821887541, + "flos": 20816964204480.0, + "grad_norm": 1.9184524951627782, + "learning_rate": 6.941762493602638e-07, + "loss": 0.8622, + "num_input_tokens_seen": 265113160, + "step": 8146 + }, + { + "epoch": 0.7347251657122243, + "flos": 25119891810240.0, + "grad_norm": 2.483904140164603, + "learning_rate": 6.937338041534899e-07, + "loss": 0.6824, + "num_input_tokens_seen": 265143110, + "step": 8147 + }, + { + "epoch": 0.7348153492356947, + "flos": 26573644619040.0, + "grad_norm": 1.6896089162273387, + "learning_rate": 6.932914704045505e-07, + "loss": 0.7401, + "num_input_tokens_seen": 265174220, + "step": 8148 + }, + { + "epoch": 0.7349055327591649, + "flos": 25662507631200.0, + "grad_norm": 2.044352717288366, + "learning_rate": 6.928492481511878e-07, + "loss": 0.7698, + "num_input_tokens_seen": 265202005, + "step": 8149 + }, + { + "epoch": 0.7349957162826352, + "flos": 21257606172960.0, + "grad_norm": 3.0375241502343377, + "learning_rate": 6.924071374311349e-07, + "loss": 0.6932, + "num_input_tokens_seen": 265229845, + "step": 8150 + }, + { + "epoch": 0.7350858998061054, + "flos": 52519240107840.0, + "grad_norm": 0.5987537504568484, + "learning_rate": 6.919651382821157e-07, + "loss": 0.5579, + "num_input_tokens_seen": 265317645, + "step": 8151 + }, + { + "epoch": 0.7351760833295757, + "flos": 24496197275520.0, + "grad_norm": 2.2317047168696713, + "learning_rate": 6.915232507418425e-07, + "loss": 0.7126, + "num_input_tokens_seen": 265347885, + "step": 8152 + }, + { + "epoch": 0.735266266853046, + "flos": 22310234201760.0, + "grad_norm": 2.1726763228141097, + "learning_rate": 6.910814748480204e-07, + "loss": 0.6257, + "num_input_tokens_seen": 265377200, + "step": 8153 + }, + { + "epoch": 0.7353564503765162, + "flos": 30651326247360.0, + "grad_norm": 2.2945123055403034, + "learning_rate": 6.906398106383445e-07, + "loss": 0.7582, + "num_input_tokens_seen": 265403655, + "step": 8154 + }, + { + "epoch": 0.7354466338999864, + "flos": 20821015708320.0, + "grad_norm": 1.6305061368946674, + "learning_rate": 6.901982581504994e-07, + "loss": 0.7362, + "num_input_tokens_seen": 265432205, + "step": 8155 + }, + { + "epoch": 0.7355368174234568, + "flos": 45735664605600.0, + "grad_norm": 3.1418651959926667, + "learning_rate": 6.897568174221611e-07, + "loss": 0.6123, + "num_input_tokens_seen": 265465225, + "step": 8156 + }, + { + "epoch": 0.735627000946927, + "flos": 20632387678080.0, + "grad_norm": 2.0797812379824396, + "learning_rate": 6.893154884909966e-07, + "loss": 0.8108, + "num_input_tokens_seen": 265493405, + "step": 8157 + }, + { + "epoch": 0.7357171844703972, + "flos": 68631557210400.0, + "grad_norm": 0.6048125125482527, + "learning_rate": 6.888742713946602e-07, + "loss": 0.5831, + "num_input_tokens_seen": 265589210, + "step": 8158 + }, + { + "epoch": 0.7358073679938675, + "flos": 21585903925920.0, + "grad_norm": 2.4654159244520977, + "learning_rate": 6.884331661708018e-07, + "loss": 0.7428, + "num_input_tokens_seen": 265617060, + "step": 8159 + }, + { + "epoch": 0.7358975515173378, + "flos": 20926892385600.0, + "grad_norm": 1.7046643221474405, + "learning_rate": 6.879921728570561e-07, + "loss": 0.7388, + "num_input_tokens_seen": 265646970, + "step": 8160 + }, + { + "epoch": 0.735987735040808, + "flos": 22749538058880.0, + "grad_norm": 1.6568642682764363, + "learning_rate": 6.875512914910539e-07, + "loss": 0.8182, + "num_input_tokens_seen": 265676470, + "step": 8161 + }, + { + "epoch": 0.7360779185642783, + "flos": 24098343434400.0, + "grad_norm": 1.9481176044909954, + "learning_rate": 6.871105221104119e-07, + "loss": 0.6893, + "num_input_tokens_seen": 265707050, + "step": 8162 + }, + { + "epoch": 0.7361681020877485, + "flos": 23550226488960.0, + "grad_norm": 2.29363723186641, + "learning_rate": 6.866698647527391e-07, + "loss": 0.7234, + "num_input_tokens_seen": 265737415, + "step": 8163 + }, + { + "epoch": 0.7362582856112189, + "flos": 20853210454080.0, + "grad_norm": 2.4014721289404717, + "learning_rate": 6.862293194556353e-07, + "loss": 0.7631, + "num_input_tokens_seen": 265765385, + "step": 8164 + }, + { + "epoch": 0.7363484691346891, + "flos": 19873297943040.0, + "grad_norm": 2.384827809065896, + "learning_rate": 6.857888862566896e-07, + "loss": 0.6863, + "num_input_tokens_seen": 265795080, + "step": 8165 + }, + { + "epoch": 0.7364386526581593, + "flos": 26244937998720.0, + "grad_norm": 1.8418444468563042, + "learning_rate": 6.853485651934836e-07, + "loss": 0.7056, + "num_input_tokens_seen": 265824675, + "step": 8166 + }, + { + "epoch": 0.7365288361816296, + "flos": 35949443135520.0, + "grad_norm": 2.692621275172662, + "learning_rate": 6.849083563035855e-07, + "loss": 0.5948, + "num_input_tokens_seen": 265862780, + "step": 8167 + }, + { + "epoch": 0.7366190197050999, + "flos": 65718253110240.0, + "grad_norm": 0.7100073629991456, + "learning_rate": 6.844682596245592e-07, + "loss": 0.6606, + "num_input_tokens_seen": 265948935, + "step": 8168 + }, + { + "epoch": 0.7367092032285701, + "flos": 20778004562400.0, + "grad_norm": 1.9239202641918765, + "learning_rate": 6.840282751939539e-07, + "loss": 0.7996, + "num_input_tokens_seen": 265975010, + "step": 8169 + }, + { + "epoch": 0.7367993867520404, + "flos": 19910399097120.0, + "grad_norm": 2.6279535480324716, + "learning_rate": 6.835884030493126e-07, + "loss": 0.7874, + "num_input_tokens_seen": 266001430, + "step": 8170 + }, + { + "epoch": 0.7368895702755107, + "flos": 26430889806240.0, + "grad_norm": 2.051512030324264, + "learning_rate": 6.831486432281672e-07, + "loss": 0.7654, + "num_input_tokens_seen": 266030860, + "step": 8171 + }, + { + "epoch": 0.736979753798981, + "flos": 17724361683840.0, + "grad_norm": 2.2564881344395986, + "learning_rate": 6.827089957680407e-07, + "loss": 0.7526, + "num_input_tokens_seen": 266058275, + "step": 8172 + }, + { + "epoch": 0.7370699373224512, + "flos": 29268764996160.0, + "grad_norm": 1.8086390314551175, + "learning_rate": 6.822694607064461e-07, + "loss": 0.8505, + "num_input_tokens_seen": 266087535, + "step": 8173 + }, + { + "epoch": 0.7371601208459214, + "flos": 25703362931040.0, + "grad_norm": 1.98513076687805, + "learning_rate": 6.818300380808877e-07, + "loss": 0.7646, + "num_input_tokens_seen": 266116870, + "step": 8174 + }, + { + "epoch": 0.7372503043693918, + "flos": 24026705839680.0, + "grad_norm": 1.7991232697184298, + "learning_rate": 6.813907279288574e-07, + "loss": 0.7329, + "num_input_tokens_seen": 266146240, + "step": 8175 + }, + { + "epoch": 0.737340487892862, + "flos": 17354539575360.0, + "grad_norm": 1.9322606278743049, + "learning_rate": 6.809515302878422e-07, + "loss": 0.7691, + "num_input_tokens_seen": 266172495, + "step": 8176 + }, + { + "epoch": 0.7374306714163322, + "flos": 26831680058400.0, + "grad_norm": 3.141222996502688, + "learning_rate": 6.80512445195315e-07, + "loss": 0.7351, + "num_input_tokens_seen": 266200710, + "step": 8177 + }, + { + "epoch": 0.7375208549398025, + "flos": 25262795302080.0, + "grad_norm": 3.0455109501743443, + "learning_rate": 6.800734726887416e-07, + "loss": 0.7601, + "num_input_tokens_seen": 266232025, + "step": 8178 + }, + { + "epoch": 0.7376110384632728, + "flos": 16994976320640.0, + "grad_norm": 1.911108984625666, + "learning_rate": 6.796346128055775e-07, + "loss": 0.7461, + "num_input_tokens_seen": 266257455, + "step": 8179 + }, + { + "epoch": 0.737701221986743, + "flos": 22781658465120.0, + "grad_norm": 2.077799032933297, + "learning_rate": 6.791958655832684e-07, + "loss": 0.7118, + "num_input_tokens_seen": 266285430, + "step": 8180 + }, + { + "epoch": 0.7377914055102133, + "flos": 33607418474880.0, + "grad_norm": 2.1173579861737526, + "learning_rate": 6.787572310592518e-07, + "loss": 0.7063, + "num_input_tokens_seen": 266314860, + "step": 8181 + }, + { + "epoch": 0.7378815890336835, + "flos": 26394755065920.0, + "grad_norm": 2.104877226051265, + "learning_rate": 6.783187092709521e-07, + "loss": 0.8255, + "num_input_tokens_seen": 266343260, + "step": 8182 + }, + { + "epoch": 0.7379717725571538, + "flos": 18343261319520.0, + "grad_norm": 5.761922996909491, + "learning_rate": 6.778803002557891e-07, + "loss": 0.7528, + "num_input_tokens_seen": 266369040, + "step": 8183 + }, + { + "epoch": 0.7380619560806241, + "flos": 24209126520000.0, + "grad_norm": 2.0601615364663637, + "learning_rate": 6.774420040511686e-07, + "loss": 0.7785, + "num_input_tokens_seen": 266399295, + "step": 8184 + }, + { + "epoch": 0.7381521396040943, + "flos": 24464857434240.0, + "grad_norm": 2.679876559958953, + "learning_rate": 6.770038206944886e-07, + "loss": 0.7449, + "num_input_tokens_seen": 266428800, + "step": 8185 + }, + { + "epoch": 0.7382423231275645, + "flos": 26977780149600.0, + "grad_norm": 1.819230572633334, + "learning_rate": 6.765657502231375e-07, + "loss": 0.7571, + "num_input_tokens_seen": 266459870, + "step": 8186 + }, + { + "epoch": 0.7383325066510349, + "flos": 22096101982560.0, + "grad_norm": 6.8000410751897356, + "learning_rate": 6.761277926744939e-07, + "loss": 0.8093, + "num_input_tokens_seen": 266490655, + "step": 8187 + }, + { + "epoch": 0.7384226901745051, + "flos": 26243079510720.0, + "grad_norm": 2.0599408311586904, + "learning_rate": 6.756899480859268e-07, + "loss": 0.732, + "num_input_tokens_seen": 266519920, + "step": 8188 + }, + { + "epoch": 0.7385128736979754, + "flos": 22824186404160.0, + "grad_norm": 2.4482884442616095, + "learning_rate": 6.752522164947956e-07, + "loss": 0.7858, + "num_input_tokens_seen": 266546330, + "step": 8189 + }, + { + "epoch": 0.7386030572214456, + "flos": 24863826368160.0, + "grad_norm": 1.4939145783614256, + "learning_rate": 6.748145979384498e-07, + "loss": 0.7982, + "num_input_tokens_seen": 266576990, + "step": 8190 + }, + { + "epoch": 0.7386932407449159, + "flos": 24281842037760.0, + "grad_norm": 1.882743232687536, + "learning_rate": 6.743770924542303e-07, + "loss": 0.6847, + "num_input_tokens_seen": 266605525, + "step": 8191 + }, + { + "epoch": 0.7387834242683862, + "flos": 58716711169920.0, + "grad_norm": 0.6361054819552996, + "learning_rate": 6.739397000794658e-07, + "loss": 0.5688, + "num_input_tokens_seen": 266690805, + "step": 8192 + }, + { + "epoch": 0.7388736077918564, + "flos": 21145076108640.0, + "grad_norm": 1.9667074719199067, + "learning_rate": 6.735024208514782e-07, + "loss": 0.7187, + "num_input_tokens_seen": 266717585, + "step": 8193 + }, + { + "epoch": 0.7389637913153266, + "flos": 35062843922880.0, + "grad_norm": 1.515044897692047, + "learning_rate": 6.73065254807578e-07, + "loss": 0.6604, + "num_input_tokens_seen": 266749885, + "step": 8194 + }, + { + "epoch": 0.739053974838797, + "flos": 22677417257280.0, + "grad_norm": 1.749139165324369, + "learning_rate": 6.726282019850669e-07, + "loss": 0.7959, + "num_input_tokens_seen": 266777840, + "step": 8195 + }, + { + "epoch": 0.7391441583622672, + "flos": 24100053243360.0, + "grad_norm": 1.8777619682238378, + "learning_rate": 6.721912624212376e-07, + "loss": 0.8125, + "num_input_tokens_seen": 266810010, + "step": 8196 + }, + { + "epoch": 0.7392343418857374, + "flos": 36262687135680.0, + "grad_norm": 1.8647885413881133, + "learning_rate": 6.717544361533696e-07, + "loss": 0.6438, + "num_input_tokens_seen": 266839685, + "step": 8197 + }, + { + "epoch": 0.7393245254092078, + "flos": 23990756948160.0, + "grad_norm": 2.922234987328533, + "learning_rate": 6.713177232187386e-07, + "loss": 0.7244, + "num_input_tokens_seen": 266868555, + "step": 8198 + }, + { + "epoch": 0.739414708932678, + "flos": 23150402650560.0, + "grad_norm": 1.5441289861246712, + "learning_rate": 6.708811236546041e-07, + "loss": 0.7355, + "num_input_tokens_seen": 266897930, + "step": 8199 + }, + { + "epoch": 0.7395048924561483, + "flos": 24201023512320.0, + "grad_norm": 1.988836664293209, + "learning_rate": 6.704446374982224e-07, + "loss": 0.7092, + "num_input_tokens_seen": 266924195, + "step": 8200 + }, + { + "epoch": 0.7395950759796185, + "flos": 16813001677440.0, + "grad_norm": 2.115380335850155, + "learning_rate": 6.700082647868346e-07, + "loss": 0.776, + "num_input_tokens_seen": 266950685, + "step": 8201 + }, + { + "epoch": 0.7396852595030888, + "flos": 26322485585280.0, + "grad_norm": 2.0734334202534397, + "learning_rate": 6.695720055576751e-07, + "loss": 0.7971, + "num_input_tokens_seen": 266980220, + "step": 8202 + }, + { + "epoch": 0.7397754430265591, + "flos": 24172025365920.0, + "grad_norm": 1.9737727565311156, + "learning_rate": 6.691358598479679e-07, + "loss": 0.7529, + "num_input_tokens_seen": 267008105, + "step": 8203 + }, + { + "epoch": 0.7398656265500293, + "flos": 22420162382880.0, + "grad_norm": 1.5509133078513637, + "learning_rate": 6.686998276949276e-07, + "loss": 0.7334, + "num_input_tokens_seen": 267037130, + "step": 8204 + }, + { + "epoch": 0.7399558100734995, + "flos": 65869296779520.0, + "grad_norm": 0.6359791819341195, + "learning_rate": 6.682639091357587e-07, + "loss": 0.5853, + "num_input_tokens_seen": 267132235, + "step": 8205 + }, + { + "epoch": 0.7400459935969699, + "flos": 20857336297440.0, + "grad_norm": 1.9099516651768358, + "learning_rate": 6.678281042076568e-07, + "loss": 0.7355, + "num_input_tokens_seen": 267161530, + "step": 8206 + }, + { + "epoch": 0.7401361771204401, + "flos": 24172136875200.0, + "grad_norm": 4.95959281419033, + "learning_rate": 6.673924129478059e-07, + "loss": 0.7473, + "num_input_tokens_seen": 267191795, + "step": 8207 + }, + { + "epoch": 0.7402263606439103, + "flos": 27268976748480.0, + "grad_norm": 1.6676247306739447, + "learning_rate": 6.669568353933824e-07, + "loss": 0.6647, + "num_input_tokens_seen": 267223040, + "step": 8208 + }, + { + "epoch": 0.7403165441673806, + "flos": 19509274317120.0, + "grad_norm": 1.5584225668152503, + "learning_rate": 6.665213715815519e-07, + "loss": 0.6783, + "num_input_tokens_seen": 267252315, + "step": 8209 + }, + { + "epoch": 0.7404067276908509, + "flos": 36843928070880.0, + "grad_norm": 2.434132328106848, + "learning_rate": 6.660860215494706e-07, + "loss": 0.7565, + "num_input_tokens_seen": 267280370, + "step": 8210 + }, + { + "epoch": 0.7404969112143212, + "flos": 27774751603680.0, + "grad_norm": 1.9945564098093644, + "learning_rate": 6.656507853342852e-07, + "loss": 0.6969, + "num_input_tokens_seen": 267309405, + "step": 8211 + }, + { + "epoch": 0.7405870947377914, + "flos": 26868335175360.0, + "grad_norm": 1.631680694080773, + "learning_rate": 6.652156629731323e-07, + "loss": 0.7549, + "num_input_tokens_seen": 267341670, + "step": 8212 + }, + { + "epoch": 0.7406772782612616, + "flos": 21689699096640.0, + "grad_norm": 2.02527573203539, + "learning_rate": 6.647806545031396e-07, + "loss": 0.8086, + "num_input_tokens_seen": 267369760, + "step": 8213 + }, + { + "epoch": 0.740767461784732, + "flos": 22861547746560.0, + "grad_norm": 1.7810176898649654, + "learning_rate": 6.643457599614224e-07, + "loss": 0.6669, + "num_input_tokens_seen": 267395940, + "step": 8214 + }, + { + "epoch": 0.7408576453082022, + "flos": 24755050449600.0, + "grad_norm": 3.5818979951955416, + "learning_rate": 6.63910979385091e-07, + "loss": 0.8017, + "num_input_tokens_seen": 267424410, + "step": 8215 + }, + { + "epoch": 0.7409478288316724, + "flos": 19254435477120.0, + "grad_norm": 1.817932340997514, + "learning_rate": 6.634763128112409e-07, + "loss": 0.8266, + "num_input_tokens_seen": 267451295, + "step": 8216 + }, + { + "epoch": 0.7410380123551427, + "flos": 23261520264000.0, + "grad_norm": 2.1483500514873235, + "learning_rate": 6.630417602769622e-07, + "loss": 0.7295, + "num_input_tokens_seen": 267478820, + "step": 8217 + }, + { + "epoch": 0.741128195878613, + "flos": 35827843649760.0, + "grad_norm": 1.9173170174995218, + "learning_rate": 6.62607321819332e-07, + "loss": 0.7483, + "num_input_tokens_seen": 267509195, + "step": 8218 + }, + { + "epoch": 0.7412183794020832, + "flos": 20346617864160.0, + "grad_norm": 1.4849358553149776, + "learning_rate": 6.621729974754196e-07, + "loss": 0.741, + "num_input_tokens_seen": 267539660, + "step": 8219 + }, + { + "epoch": 0.7413085629255535, + "flos": 22455144860640.0, + "grad_norm": 2.2417991381599047, + "learning_rate": 6.617387872822835e-07, + "loss": 0.6928, + "num_input_tokens_seen": 267568440, + "step": 8220 + }, + { + "epoch": 0.7413987464490238, + "flos": 23516433443520.0, + "grad_norm": 2.404467628545078, + "learning_rate": 6.613046912769731e-07, + "loss": 0.7138, + "num_input_tokens_seen": 267596120, + "step": 8221 + }, + { + "epoch": 0.741488929972494, + "flos": 25011227400960.0, + "grad_norm": 2.4793732394112906, + "learning_rate": 6.608707094965289e-07, + "loss": 0.8119, + "num_input_tokens_seen": 267623675, + "step": 8222 + }, + { + "epoch": 0.7415791134959643, + "flos": 15537766724160.0, + "grad_norm": 2.825495725364874, + "learning_rate": 6.604368419779787e-07, + "loss": 0.6799, + "num_input_tokens_seen": 267649110, + "step": 8223 + }, + { + "epoch": 0.7416692970194345, + "flos": 25700538029280.0, + "grad_norm": 2.0375559418844413, + "learning_rate": 6.600030887583434e-07, + "loss": 0.7547, + "num_input_tokens_seen": 267677260, + "step": 8224 + }, + { + "epoch": 0.7417594805429049, + "flos": 27811183702080.0, + "grad_norm": 1.9799993135470346, + "learning_rate": 6.595694498746336e-07, + "loss": 0.8491, + "num_input_tokens_seen": 267706180, + "step": 8225 + }, + { + "epoch": 0.7418496640663751, + "flos": 31420117289760.0, + "grad_norm": 2.2468819736165577, + "learning_rate": 6.591359253638491e-07, + "loss": 0.7118, + "num_input_tokens_seen": 267739600, + "step": 8226 + }, + { + "epoch": 0.7419398475898453, + "flos": 21731000433600.0, + "grad_norm": 4.782809333588363, + "learning_rate": 6.587025152629808e-07, + "loss": 0.6343, + "num_input_tokens_seen": 267767750, + "step": 8227 + }, + { + "epoch": 0.7420300311133156, + "flos": 15829112002080.0, + "grad_norm": 3.064134900970454, + "learning_rate": 6.582692196090107e-07, + "loss": 0.7826, + "num_input_tokens_seen": 267793865, + "step": 8228 + }, + { + "epoch": 0.7421202146367859, + "flos": 27773822359680.0, + "grad_norm": 1.7556795870482547, + "learning_rate": 6.578360384389074e-07, + "loss": 0.7572, + "num_input_tokens_seen": 267821460, + "step": 8229 + }, + { + "epoch": 0.7422103981602561, + "flos": 27633557920800.0, + "grad_norm": 2.66985365048623, + "learning_rate": 6.574029717896355e-07, + "loss": 0.724, + "num_input_tokens_seen": 267851935, + "step": 8230 + }, + { + "epoch": 0.7423005816837264, + "flos": 21731223452160.0, + "grad_norm": 2.6819961434435444, + "learning_rate": 6.569700196981436e-07, + "loss": 0.7341, + "num_input_tokens_seen": 267880205, + "step": 8231 + }, + { + "epoch": 0.7423907652071966, + "flos": 24609656583840.0, + "grad_norm": 3.6102609579177427, + "learning_rate": 6.565371822013763e-07, + "loss": 0.6461, + "num_input_tokens_seen": 267909200, + "step": 8232 + }, + { + "epoch": 0.742480948730667, + "flos": 19035954396000.0, + "grad_norm": 2.136189899168339, + "learning_rate": 6.561044593362636e-07, + "loss": 0.7986, + "num_input_tokens_seen": 267936450, + "step": 8233 + }, + { + "epoch": 0.7425711322541372, + "flos": 23334458800320.0, + "grad_norm": 2.0485644175350033, + "learning_rate": 6.556718511397288e-07, + "loss": 0.7272, + "num_input_tokens_seen": 267964045, + "step": 8234 + }, + { + "epoch": 0.7426613157776074, + "flos": 20197209664320.0, + "grad_norm": 3.067678627340962, + "learning_rate": 6.552393576486843e-07, + "loss": 0.8617, + "num_input_tokens_seen": 267990965, + "step": 8235 + }, + { + "epoch": 0.7427514993010776, + "flos": 12951236416800.0, + "grad_norm": 2.0900362015285645, + "learning_rate": 6.548069789000325e-07, + "loss": 0.7572, + "num_input_tokens_seen": 268017645, + "step": 8236 + }, + { + "epoch": 0.742841682824548, + "flos": 30114323060160.0, + "grad_norm": 2.017778886996864, + "learning_rate": 6.543747149306673e-07, + "loss": 0.7329, + "num_input_tokens_seen": 268051965, + "step": 8237 + }, + { + "epoch": 0.7429318663480182, + "flos": 30182206509120.0, + "grad_norm": 1.7019258529792967, + "learning_rate": 6.5394256577747e-07, + "loss": 0.7919, + "num_input_tokens_seen": 268081620, + "step": 8238 + }, + { + "epoch": 0.7430220498714885, + "flos": 18445346681280.0, + "grad_norm": 1.701337243505148, + "learning_rate": 6.535105314773161e-07, + "loss": 0.7737, + "num_input_tokens_seen": 268106400, + "step": 8239 + }, + { + "epoch": 0.7431122333949587, + "flos": 31312159105920.0, + "grad_norm": 2.0811329347604817, + "learning_rate": 6.530786120670677e-07, + "loss": 0.6757, + "num_input_tokens_seen": 268136360, + "step": 8240 + }, + { + "epoch": 0.743202416918429, + "flos": 20995147532160.0, + "grad_norm": 2.142087673362674, + "learning_rate": 6.526468075835787e-07, + "loss": 0.7981, + "num_input_tokens_seen": 268160870, + "step": 8241 + }, + { + "epoch": 0.7432926004418993, + "flos": 24057079267200.0, + "grad_norm": 2.335096329002865, + "learning_rate": 6.522151180636937e-07, + "loss": 0.7911, + "num_input_tokens_seen": 268188690, + "step": 8242 + }, + { + "epoch": 0.7433827839653695, + "flos": 29928148234080.0, + "grad_norm": 1.8295727727219337, + "learning_rate": 6.517835435442461e-07, + "loss": 0.7077, + "num_input_tokens_seen": 268219350, + "step": 8243 + }, + { + "epoch": 0.7434729674888398, + "flos": 22787271098880.0, + "grad_norm": 3.6335566589128994, + "learning_rate": 6.513520840620606e-07, + "loss": 0.7694, + "num_input_tokens_seen": 268244115, + "step": 8244 + }, + { + "epoch": 0.7435631510123101, + "flos": 22205769975360.0, + "grad_norm": 4.001469580581776, + "learning_rate": 6.509207396539525e-07, + "loss": 0.7065, + "num_input_tokens_seen": 268273875, + "step": 8245 + }, + { + "epoch": 0.7436533345357803, + "flos": 62457694679520.0, + "grad_norm": 0.620401320356259, + "learning_rate": 6.50489510356724e-07, + "loss": 0.5516, + "num_input_tokens_seen": 268364685, + "step": 8246 + }, + { + "epoch": 0.7437435180592505, + "flos": 17869235172960.0, + "grad_norm": 2.3808267292291174, + "learning_rate": 6.500583962071732e-07, + "loss": 0.6714, + "num_input_tokens_seen": 268392595, + "step": 8247 + }, + { + "epoch": 0.7438337015827209, + "flos": 23996109393600.0, + "grad_norm": 1.5749391180714618, + "learning_rate": 6.496273972420827e-07, + "loss": 0.7722, + "num_input_tokens_seen": 268424990, + "step": 8248 + }, + { + "epoch": 0.7439238851061911, + "flos": 20855031772320.0, + "grad_norm": 1.890998488176973, + "learning_rate": 6.491965134982287e-07, + "loss": 0.7921, + "num_input_tokens_seen": 268447610, + "step": 8249 + }, + { + "epoch": 0.7440140686296614, + "flos": 22675707448320.0, + "grad_norm": 2.0964946691844224, + "learning_rate": 6.487657450123765e-07, + "loss": 0.7148, + "num_input_tokens_seen": 268476360, + "step": 8250 + }, + { + "epoch": 0.7441042521531316, + "flos": 22386666695520.0, + "grad_norm": 1.5583913571304095, + "learning_rate": 6.483350918212814e-07, + "loss": 0.7588, + "num_input_tokens_seen": 268504775, + "step": 8251 + }, + { + "epoch": 0.7441944356766019, + "flos": 67407033276960.0, + "grad_norm": 0.7031043791396244, + "learning_rate": 6.479045539616898e-07, + "loss": 0.5836, + "num_input_tokens_seen": 268595940, + "step": 8252 + }, + { + "epoch": 0.7442846192000722, + "flos": 23763169275840.0, + "grad_norm": 2.101780876527282, + "learning_rate": 6.474741314703358e-07, + "loss": 0.7472, + "num_input_tokens_seen": 268618595, + "step": 8253 + }, + { + "epoch": 0.7443748027235424, + "flos": 18779777444640.0, + "grad_norm": 3.91435341125119, + "learning_rate": 6.47043824383948e-07, + "loss": 0.8044, + "num_input_tokens_seen": 268644260, + "step": 8254 + }, + { + "epoch": 0.7444649862470126, + "flos": 25623399310080.0, + "grad_norm": 2.150501937826313, + "learning_rate": 6.466136327392399e-07, + "loss": 0.7849, + "num_input_tokens_seen": 268667360, + "step": 8255 + }, + { + "epoch": 0.744555169770483, + "flos": 18742527611520.0, + "grad_norm": 3.6086466000591764, + "learning_rate": 6.461835565729206e-07, + "loss": 0.7798, + "num_input_tokens_seen": 268695555, + "step": 8256 + }, + { + "epoch": 0.7446453532939532, + "flos": 18702267027840.0, + "grad_norm": 1.8411726960880308, + "learning_rate": 6.457535959216844e-07, + "loss": 0.7751, + "num_input_tokens_seen": 268722365, + "step": 8257 + }, + { + "epoch": 0.7447355368174234, + "flos": 23006309726400.0, + "grad_norm": 2.205065258383908, + "learning_rate": 6.453237508222186e-07, + "loss": 0.7056, + "num_input_tokens_seen": 268750240, + "step": 8258 + }, + { + "epoch": 0.7448257203408937, + "flos": 23441339061120.0, + "grad_norm": 1.5947953330242222, + "learning_rate": 6.448940213112e-07, + "loss": 0.7668, + "num_input_tokens_seen": 268781360, + "step": 8259 + }, + { + "epoch": 0.744915903864364, + "flos": 22202052999360.0, + "grad_norm": 1.6720715701331637, + "learning_rate": 6.444644074252954e-07, + "loss": 0.8281, + "num_input_tokens_seen": 268809520, + "step": 8260 + }, + { + "epoch": 0.7450060873878342, + "flos": 22096845377760.0, + "grad_norm": 1.702375544565804, + "learning_rate": 6.440349092011628e-07, + "loss": 0.8155, + "num_input_tokens_seen": 268838145, + "step": 8261 + }, + { + "epoch": 0.7450962709113045, + "flos": 19437413703840.0, + "grad_norm": 1.7672636816736613, + "learning_rate": 6.436055266754475e-07, + "loss": 0.7364, + "num_input_tokens_seen": 268865355, + "step": 8262 + }, + { + "epoch": 0.7451864544347747, + "flos": 26642122784160.0, + "grad_norm": 2.218456947423046, + "learning_rate": 6.431762598847879e-07, + "loss": 0.7246, + "num_input_tokens_seen": 268896445, + "step": 8263 + }, + { + "epoch": 0.7452766379582451, + "flos": 27411508542720.0, + "grad_norm": 2.0412382597623617, + "learning_rate": 6.427471088658111e-07, + "loss": 0.7218, + "num_input_tokens_seen": 268925100, + "step": 8264 + }, + { + "epoch": 0.7453668214817153, + "flos": 36231087106080.0, + "grad_norm": 2.167236534394528, + "learning_rate": 6.42318073655135e-07, + "loss": 0.7444, + "num_input_tokens_seen": 268954735, + "step": 8265 + }, + { + "epoch": 0.7454570050051855, + "flos": 26684836572000.0, + "grad_norm": 1.9042953099953506, + "learning_rate": 6.41889154289367e-07, + "loss": 0.7089, + "num_input_tokens_seen": 268986260, + "step": 8266 + }, + { + "epoch": 0.7455471885286559, + "flos": 18525050113920.0, + "grad_norm": 2.432458113557692, + "learning_rate": 6.414603508051055e-07, + "loss": 0.8032, + "num_input_tokens_seen": 269012530, + "step": 8267 + }, + { + "epoch": 0.7456373720521261, + "flos": 59433681833280.0, + "grad_norm": 0.637159464145909, + "learning_rate": 6.410316632389365e-07, + "loss": 0.6036, + "num_input_tokens_seen": 269105605, + "step": 8268 + }, + { + "epoch": 0.7457275555755963, + "flos": 22715038788000.0, + "grad_norm": 2.4914071227509056, + "learning_rate": 6.406030916274406e-07, + "loss": 0.7548, + "num_input_tokens_seen": 269133405, + "step": 8269 + }, + { + "epoch": 0.7458177390990666, + "flos": 21944723785440.0, + "grad_norm": 2.30059865541098, + "learning_rate": 6.401746360071831e-07, + "loss": 0.7046, + "num_input_tokens_seen": 269161630, + "step": 8270 + }, + { + "epoch": 0.7459079226225369, + "flos": 28832732077920.0, + "grad_norm": 2.380103600755562, + "learning_rate": 6.397462964147251e-07, + "loss": 0.7095, + "num_input_tokens_seen": 269192270, + "step": 8271 + }, + { + "epoch": 0.7459981061460071, + "flos": 22095730284960.0, + "grad_norm": 1.8602001378867206, + "learning_rate": 6.393180728866128e-07, + "loss": 0.7355, + "num_input_tokens_seen": 269218960, + "step": 8272 + }, + { + "epoch": 0.7460882896694774, + "flos": 26060324302560.0, + "grad_norm": 1.9595526558434069, + "learning_rate": 6.388899654593853e-07, + "loss": 0.7809, + "num_input_tokens_seen": 269245555, + "step": 8273 + }, + { + "epoch": 0.7461784731929476, + "flos": 23769153607200.0, + "grad_norm": 2.242818604017575, + "learning_rate": 6.384619741695709e-07, + "loss": 0.7726, + "num_input_tokens_seen": 269275430, + "step": 8274 + }, + { + "epoch": 0.746268656716418, + "flos": 23880568578720.0, + "grad_norm": 1.295799728808262, + "learning_rate": 6.380340990536883e-07, + "loss": 0.7458, + "num_input_tokens_seen": 269307655, + "step": 8275 + }, + { + "epoch": 0.7463588402398882, + "flos": 27268605050880.0, + "grad_norm": 2.790273195242601, + "learning_rate": 6.37606340148247e-07, + "loss": 0.7513, + "num_input_tokens_seen": 269337070, + "step": 8276 + }, + { + "epoch": 0.7464490237633584, + "flos": 23297469155520.0, + "grad_norm": 2.1916615818482024, + "learning_rate": 6.371786974897433e-07, + "loss": 0.6566, + "num_input_tokens_seen": 269361570, + "step": 8277 + }, + { + "epoch": 0.7465392072868287, + "flos": 22199302437120.0, + "grad_norm": 1.6084936747238352, + "learning_rate": 6.367511711146691e-07, + "loss": 0.7761, + "num_input_tokens_seen": 269389225, + "step": 8278 + }, + { + "epoch": 0.746629390810299, + "flos": 25010595515040.0, + "grad_norm": 2.007114392938363, + "learning_rate": 6.363237610595014e-07, + "loss": 0.6904, + "num_input_tokens_seen": 269417175, + "step": 8279 + }, + { + "epoch": 0.7467195743337692, + "flos": 20413200371520.0, + "grad_norm": 2.132093706680399, + "learning_rate": 6.358964673607094e-07, + "loss": 0.7278, + "num_input_tokens_seen": 269444120, + "step": 8280 + }, + { + "epoch": 0.7468097578572395, + "flos": 24829587285600.0, + "grad_norm": 2.4909012204573036, + "learning_rate": 6.354692900547525e-07, + "loss": 0.7795, + "num_input_tokens_seen": 269474215, + "step": 8281 + }, + { + "epoch": 0.7468999413807097, + "flos": 20820346652640.0, + "grad_norm": 1.8091891731936838, + "learning_rate": 6.350422291780797e-07, + "loss": 0.7207, + "num_input_tokens_seen": 269502340, + "step": 8282 + }, + { + "epoch": 0.74699012490418, + "flos": 19070342157600.0, + "grad_norm": 2.3899750667532387, + "learning_rate": 6.346152847671302e-07, + "loss": 0.8453, + "num_input_tokens_seen": 269526210, + "step": 8283 + }, + { + "epoch": 0.7470803084276503, + "flos": 21003659407200.0, + "grad_norm": 2.0796573123704185, + "learning_rate": 6.34188456858334e-07, + "loss": 0.7377, + "num_input_tokens_seen": 269554965, + "step": 8284 + }, + { + "epoch": 0.7471704919511205, + "flos": 25187961108000.0, + "grad_norm": 2.2204177391243887, + "learning_rate": 6.337617454881081e-07, + "loss": 0.6989, + "num_input_tokens_seen": 269584190, + "step": 8285 + }, + { + "epoch": 0.7472606754745907, + "flos": 40858372470240.0, + "grad_norm": 1.7556712354232782, + "learning_rate": 6.333351506928651e-07, + "loss": 0.6567, + "num_input_tokens_seen": 269615975, + "step": 8286 + }, + { + "epoch": 0.7473508589980611, + "flos": 35794682490240.0, + "grad_norm": 1.8498628436615232, + "learning_rate": 6.329086725090018e-07, + "loss": 0.717, + "num_input_tokens_seen": 269646645, + "step": 8287 + }, + { + "epoch": 0.7474410425215313, + "flos": 23225980239840.0, + "grad_norm": 2.441121565433172, + "learning_rate": 6.324823109729087e-07, + "loss": 0.7103, + "num_input_tokens_seen": 269675710, + "step": 8288 + }, + { + "epoch": 0.7475312260450016, + "flos": 34151260897920.0, + "grad_norm": 4.029029412020926, + "learning_rate": 6.320560661209653e-07, + "loss": 0.6809, + "num_input_tokens_seen": 269707795, + "step": 8289 + }, + { + "epoch": 0.7476214095684719, + "flos": 22492692051840.0, + "grad_norm": 1.836028790740621, + "learning_rate": 6.316299379895411e-07, + "loss": 0.6302, + "num_input_tokens_seen": 269736650, + "step": 8290 + }, + { + "epoch": 0.7477115930919421, + "flos": 18525310302240.0, + "grad_norm": 1.8941451160946448, + "learning_rate": 6.312039266149965e-07, + "loss": 0.7821, + "num_input_tokens_seen": 269764015, + "step": 8291 + }, + { + "epoch": 0.7478017766154124, + "flos": 25663920082080.0, + "grad_norm": 1.7654347275017255, + "learning_rate": 6.307780320336789e-07, + "loss": 0.7774, + "num_input_tokens_seen": 269794480, + "step": 8292 + }, + { + "epoch": 0.7478919601388826, + "flos": 19072237815360.0, + "grad_norm": 2.069761267638635, + "learning_rate": 6.303522542819306e-07, + "loss": 0.8043, + "num_input_tokens_seen": 269822110, + "step": 8293 + }, + { + "epoch": 0.7479821436623529, + "flos": 19909135325280.0, + "grad_norm": 3.1290773875930884, + "learning_rate": 6.299265933960796e-07, + "loss": 0.7175, + "num_input_tokens_seen": 269849745, + "step": 8294 + }, + { + "epoch": 0.7480723271858232, + "flos": 31311044013120.0, + "grad_norm": 2.3447224951807706, + "learning_rate": 6.295010494124462e-07, + "loss": 0.7002, + "num_input_tokens_seen": 269874730, + "step": 8295 + }, + { + "epoch": 0.7481625107092934, + "flos": 23477027764320.0, + "grad_norm": 1.8515286341922024, + "learning_rate": 6.290756223673399e-07, + "loss": 0.723, + "num_input_tokens_seen": 269904290, + "step": 8296 + }, + { + "epoch": 0.7482526942327636, + "flos": 28470641279520.0, + "grad_norm": 1.6964836470189335, + "learning_rate": 6.28650312297061e-07, + "loss": 0.7768, + "num_input_tokens_seen": 269935165, + "step": 8297 + }, + { + "epoch": 0.748342877756234, + "flos": 11747787737280.0, + "grad_norm": 4.101748507686996, + "learning_rate": 6.282251192378987e-07, + "loss": 0.6629, + "num_input_tokens_seen": 269959230, + "step": 8298 + }, + { + "epoch": 0.7484330612797042, + "flos": 25739088804000.0, + "grad_norm": 1.7571147975165458, + "learning_rate": 6.278000432261334e-07, + "loss": 0.6733, + "num_input_tokens_seen": 269991475, + "step": 8299 + }, + { + "epoch": 0.7485232448031744, + "flos": 16771923359040.0, + "grad_norm": 2.0728935994409534, + "learning_rate": 6.273750842980345e-07, + "loss": 0.7671, + "num_input_tokens_seen": 270018105, + "step": 8300 + }, + { + "epoch": 0.7486134283266447, + "flos": 21221248414080.0, + "grad_norm": 6.341860635458657, + "learning_rate": 6.269502424898625e-07, + "loss": 0.7693, + "num_input_tokens_seen": 270045685, + "step": 8301 + }, + { + "epoch": 0.748703611850115, + "flos": 18488580845760.0, + "grad_norm": 2.074942309944932, + "learning_rate": 6.265255178378663e-07, + "loss": 0.7876, + "num_input_tokens_seen": 270071405, + "step": 8302 + }, + { + "epoch": 0.7487937953735853, + "flos": 28471347504960.0, + "grad_norm": 1.8116033243419614, + "learning_rate": 6.261009103782861e-07, + "loss": 0.6409, + "num_input_tokens_seen": 270102690, + "step": 8303 + }, + { + "epoch": 0.7488839788970555, + "flos": 18452037238080.0, + "grad_norm": 6.331219705281359, + "learning_rate": 6.256764201473519e-07, + "loss": 0.7913, + "num_input_tokens_seen": 270129455, + "step": 8304 + }, + { + "epoch": 0.7489741624205257, + "flos": 58970657935680.0, + "grad_norm": 0.62644255576919, + "learning_rate": 6.252520471812835e-07, + "loss": 0.5714, + "num_input_tokens_seen": 270217705, + "step": 8305 + }, + { + "epoch": 0.7490643459439961, + "flos": 23188470218400.0, + "grad_norm": 2.110800706340418, + "learning_rate": 6.248277915162912e-07, + "loss": 0.7602, + "num_input_tokens_seen": 270246885, + "step": 8306 + }, + { + "epoch": 0.7491545294674663, + "flos": 23808261928320.0, + "grad_norm": 1.9737310219952158, + "learning_rate": 6.244036531885731e-07, + "loss": 0.5992, + "num_input_tokens_seen": 270276460, + "step": 8307 + }, + { + "epoch": 0.7492447129909365, + "flos": 45411938733120.0, + "grad_norm": 1.928884176851827, + "learning_rate": 6.239796322343216e-07, + "loss": 0.6785, + "num_input_tokens_seen": 270310125, + "step": 8308 + }, + { + "epoch": 0.7493348965144068, + "flos": 20930237664000.0, + "grad_norm": 2.3026292823837413, + "learning_rate": 6.235557286897137e-07, + "loss": 0.6833, + "num_input_tokens_seen": 270339545, + "step": 8309 + }, + { + "epoch": 0.7494250800378771, + "flos": 23880828767040.0, + "grad_norm": 1.8552995785338249, + "learning_rate": 6.231319425909223e-07, + "loss": 0.743, + "num_input_tokens_seen": 270369280, + "step": 8310 + }, + { + "epoch": 0.7495152635613473, + "flos": 22934151755040.0, + "grad_norm": 1.7643485392343996, + "learning_rate": 6.227082739741045e-07, + "loss": 0.7777, + "num_input_tokens_seen": 270398540, + "step": 8311 + }, + { + "epoch": 0.7496054470848176, + "flos": 48031370011680.0, + "grad_norm": 1.8126052299733006, + "learning_rate": 6.222847228754113e-07, + "loss": 0.7719, + "num_input_tokens_seen": 270431265, + "step": 8312 + }, + { + "epoch": 0.7496956306082878, + "flos": 20961651844800.0, + "grad_norm": 42.22585640598364, + "learning_rate": 6.218612893309823e-07, + "loss": 0.8403, + "num_input_tokens_seen": 270455585, + "step": 8313 + }, + { + "epoch": 0.7497858141317582, + "flos": 25152792781440.0, + "grad_norm": 1.8924368381416568, + "learning_rate": 6.214379733769468e-07, + "loss": 0.6864, + "num_input_tokens_seen": 270487530, + "step": 8314 + }, + { + "epoch": 0.7498759976552284, + "flos": 20347249750080.0, + "grad_norm": 1.744385841961923, + "learning_rate": 6.21014775049425e-07, + "loss": 0.7756, + "num_input_tokens_seen": 270515220, + "step": 8315 + }, + { + "epoch": 0.7499661811786986, + "flos": 65103330638880.0, + "grad_norm": 0.5761569263652704, + "learning_rate": 6.205916943845267e-07, + "loss": 0.5638, + "num_input_tokens_seen": 270620125, + "step": 8316 + }, + { + "epoch": 0.750056364702169, + "flos": 15136567604640.0, + "grad_norm": 2.0667734667141495, + "learning_rate": 6.201687314183504e-07, + "loss": 0.7907, + "num_input_tokens_seen": 270644675, + "step": 8317 + }, + { + "epoch": 0.7501465482256392, + "flos": 70516733887200.0, + "grad_norm": 0.6516566284029316, + "learning_rate": 6.197458861869862e-07, + "loss": 0.6107, + "num_input_tokens_seen": 270734215, + "step": 8318 + }, + { + "epoch": 0.7502367317491094, + "flos": 28542278874240.0, + "grad_norm": 1.9703109245082235, + "learning_rate": 6.193231587265138e-07, + "loss": 0.7908, + "num_input_tokens_seen": 270765880, + "step": 8319 + }, + { + "epoch": 0.7503269152725797, + "flos": 20310297275040.0, + "grad_norm": 2.2809209716967347, + "learning_rate": 6.189005490730024e-07, + "loss": 0.7266, + "num_input_tokens_seen": 270790735, + "step": 8320 + }, + { + "epoch": 0.75041709879605, + "flos": 33787757648640.0, + "grad_norm": 2.4454929663027216, + "learning_rate": 6.184780572625115e-07, + "loss": 0.7677, + "num_input_tokens_seen": 270821745, + "step": 8321 + }, + { + "epoch": 0.7505072823195202, + "flos": 39619383766560.0, + "grad_norm": 1.9910619211305924, + "learning_rate": 6.180556833310902e-07, + "loss": 0.6469, + "num_input_tokens_seen": 270852385, + "step": 8322 + }, + { + "epoch": 0.7505974658429905, + "flos": 20669117134560.0, + "grad_norm": 2.4438312175552266, + "learning_rate": 6.176334273147788e-07, + "loss": 0.7829, + "num_input_tokens_seen": 270878040, + "step": 8323 + }, + { + "epoch": 0.7506876493664607, + "flos": 24026482821120.0, + "grad_norm": 2.6072001273626606, + "learning_rate": 6.172112892496042e-07, + "loss": 0.6627, + "num_input_tokens_seen": 270903445, + "step": 8324 + }, + { + "epoch": 0.750777832889931, + "flos": 22420050873600.0, + "grad_norm": 1.9207514758538202, + "learning_rate": 6.167892691715883e-07, + "loss": 0.7886, + "num_input_tokens_seen": 270931330, + "step": 8325 + }, + { + "epoch": 0.7508680164134013, + "flos": 25302349660320.0, + "grad_norm": 1.9594631047494901, + "learning_rate": 6.163673671167378e-07, + "loss": 0.7257, + "num_input_tokens_seen": 270960740, + "step": 8326 + }, + { + "epoch": 0.7509581999368715, + "flos": 23181854001120.0, + "grad_norm": 1.7153482238196365, + "learning_rate": 6.15945583121054e-07, + "loss": 0.7848, + "num_input_tokens_seen": 270989345, + "step": 8327 + }, + { + "epoch": 0.7510483834603418, + "flos": 22896790412640.0, + "grad_norm": 2.461613928458722, + "learning_rate": 6.15523917220524e-07, + "loss": 0.695, + "num_input_tokens_seen": 271014690, + "step": 8328 + }, + { + "epoch": 0.7511385669838121, + "flos": 17574135749280.0, + "grad_norm": 1.843594884004817, + "learning_rate": 6.151023694511273e-07, + "loss": 0.8109, + "num_input_tokens_seen": 271041205, + "step": 8329 + }, + { + "epoch": 0.7512287505072823, + "flos": 21877435052640.0, + "grad_norm": 1.728264707467687, + "learning_rate": 6.146809398488328e-07, + "loss": 0.6912, + "num_input_tokens_seen": 271069515, + "step": 8330 + }, + { + "epoch": 0.7513189340307526, + "flos": 23404721113920.0, + "grad_norm": 1.8767094766579702, + "learning_rate": 6.142596284495989e-07, + "loss": 0.7803, + "num_input_tokens_seen": 271099915, + "step": 8331 + }, + { + "epoch": 0.7514091175542228, + "flos": 37506545077920.0, + "grad_norm": 1.8011368340463465, + "learning_rate": 6.138384352893751e-07, + "loss": 0.7016, + "num_input_tokens_seen": 271133365, + "step": 8332 + }, + { + "epoch": 0.7514993010776931, + "flos": 17104346955360.0, + "grad_norm": 2.6103043414581757, + "learning_rate": 6.134173604040987e-07, + "loss": 0.7362, + "num_input_tokens_seen": 271160320, + "step": 8333 + }, + { + "epoch": 0.7515894846011634, + "flos": 19472396181600.0, + "grad_norm": 2.03515214974545, + "learning_rate": 6.129964038296984e-07, + "loss": 0.7489, + "num_input_tokens_seen": 271188020, + "step": 8334 + }, + { + "epoch": 0.7516796681246336, + "flos": 23807890230720.0, + "grad_norm": 2.083996044294297, + "learning_rate": 6.12575565602093e-07, + "loss": 0.8148, + "num_input_tokens_seen": 271217725, + "step": 8335 + }, + { + "epoch": 0.7517698516481038, + "flos": 25048254215520.0, + "grad_norm": 1.6482337717292777, + "learning_rate": 6.121548457571905e-07, + "loss": 0.8172, + "num_input_tokens_seen": 271251790, + "step": 8336 + }, + { + "epoch": 0.7518600351715742, + "flos": 35244855735840.0, + "grad_norm": 2.8056154548956833, + "learning_rate": 6.11734244330889e-07, + "loss": 0.6907, + "num_input_tokens_seen": 271285620, + "step": 8337 + }, + { + "epoch": 0.7519502186950444, + "flos": 27487197641280.0, + "grad_norm": 3.2202753693255284, + "learning_rate": 6.113137613590773e-07, + "loss": 0.7565, + "num_input_tokens_seen": 271314735, + "step": 8338 + }, + { + "epoch": 0.7520404022185146, + "flos": 21185076504000.0, + "grad_norm": 1.71613515137593, + "learning_rate": 6.108933968776313e-07, + "loss": 0.8013, + "num_input_tokens_seen": 271341805, + "step": 8339 + }, + { + "epoch": 0.752130585741985, + "flos": 23732758678560.0, + "grad_norm": 2.1343707865147032, + "learning_rate": 6.104731509224212e-07, + "loss": 0.7444, + "num_input_tokens_seen": 271370305, + "step": 8340 + }, + { + "epoch": 0.7522207692654552, + "flos": 36086325126240.0, + "grad_norm": 1.6827368447414803, + "learning_rate": 6.100530235293027e-07, + "loss": 0.6769, + "num_input_tokens_seen": 271401380, + "step": 8341 + }, + { + "epoch": 0.7523109527889255, + "flos": 28835817168000.0, + "grad_norm": 4.044733489773572, + "learning_rate": 6.096330147341253e-07, + "loss": 0.7679, + "num_input_tokens_seen": 271430660, + "step": 8342 + }, + { + "epoch": 0.7524011363123957, + "flos": 19173877140000.0, + "grad_norm": 2.881165072498935, + "learning_rate": 6.09213124572725e-07, + "loss": 0.8022, + "num_input_tokens_seen": 271452985, + "step": 8343 + }, + { + "epoch": 0.752491319835866, + "flos": 21367794542400.0, + "grad_norm": 1.6873758764888511, + "learning_rate": 6.087933530809297e-07, + "loss": 0.8205, + "num_input_tokens_seen": 271482140, + "step": 8344 + }, + { + "epoch": 0.7525815033593363, + "flos": 13206335445120.0, + "grad_norm": 1.7744005587981972, + "learning_rate": 6.083737002945566e-07, + "loss": 0.7307, + "num_input_tokens_seen": 271508845, + "step": 8345 + }, + { + "epoch": 0.7526716868828065, + "flos": 45521904084000.0, + "grad_norm": 2.118298003559642, + "learning_rate": 6.079541662494126e-07, + "loss": 0.6961, + "num_input_tokens_seen": 271540500, + "step": 8346 + }, + { + "epoch": 0.7527618704062767, + "flos": 23622421630080.0, + "grad_norm": 1.8131605457396969, + "learning_rate": 6.075347509812954e-07, + "loss": 0.7491, + "num_input_tokens_seen": 271569065, + "step": 8347 + }, + { + "epoch": 0.7528520539297471, + "flos": 22601244951840.0, + "grad_norm": 1.6832555444451307, + "learning_rate": 6.0711545452599e-07, + "loss": 0.6986, + "num_input_tokens_seen": 271595315, + "step": 8348 + }, + { + "epoch": 0.7529422374532173, + "flos": 37360036119360.0, + "grad_norm": 4.662745099618214, + "learning_rate": 6.066962769192756e-07, + "loss": 0.6698, + "num_input_tokens_seen": 271629420, + "step": 8349 + }, + { + "epoch": 0.7530324209766875, + "flos": 18231995027040.0, + "grad_norm": 3.7446799369913273, + "learning_rate": 6.062772181969167e-07, + "loss": 0.6801, + "num_input_tokens_seen": 271654025, + "step": 8350 + }, + { + "epoch": 0.7531226045001578, + "flos": 22638977991840.0, + "grad_norm": 1.7524895670712226, + "learning_rate": 6.058582783946706e-07, + "loss": 0.8134, + "num_input_tokens_seen": 271681815, + "step": 8351 + }, + { + "epoch": 0.7532127880236281, + "flos": 22569347564160.0, + "grad_norm": 2.2259854823870375, + "learning_rate": 6.054394575482833e-07, + "loss": 0.7351, + "num_input_tokens_seen": 271711350, + "step": 8352 + }, + { + "epoch": 0.7533029715470984, + "flos": 21075222662400.0, + "grad_norm": 1.8882399719044725, + "learning_rate": 6.05020755693491e-07, + "loss": 0.66, + "num_input_tokens_seen": 271737915, + "step": 8353 + }, + { + "epoch": 0.7533931550705686, + "flos": 32879482732320.0, + "grad_norm": 1.8578851324694565, + "learning_rate": 6.046021728660198e-07, + "loss": 0.6571, + "num_input_tokens_seen": 271767845, + "step": 8354 + }, + { + "epoch": 0.7534833385940388, + "flos": 21002804502720.0, + "grad_norm": 2.476810671529033, + "learning_rate": 6.041837091015858e-07, + "loss": 0.7819, + "num_input_tokens_seen": 271796435, + "step": 8355 + }, + { + "epoch": 0.7535735221175092, + "flos": 34154680515840.0, + "grad_norm": 1.67519199352774, + "learning_rate": 6.037653644358931e-07, + "loss": 0.7953, + "num_input_tokens_seen": 271826815, + "step": 8356 + }, + { + "epoch": 0.7536637056409794, + "flos": 23079396941760.0, + "grad_norm": 2.6424612849782005, + "learning_rate": 6.033471389046393e-07, + "loss": 0.7553, + "num_input_tokens_seen": 271857775, + "step": 8357 + }, + { + "epoch": 0.7537538891644496, + "flos": 25083608390880.0, + "grad_norm": 1.7928551199483125, + "learning_rate": 6.029290325435084e-07, + "loss": 0.7006, + "num_input_tokens_seen": 271889190, + "step": 8358 + }, + { + "epoch": 0.7538440726879199, + "flos": 22314954761280.0, + "grad_norm": 2.9248971252608826, + "learning_rate": 6.025110453881756e-07, + "loss": 0.8065, + "num_input_tokens_seen": 271915105, + "step": 8359 + }, + { + "epoch": 0.7539342562113902, + "flos": 28762321085280.0, + "grad_norm": 2.2080013337476823, + "learning_rate": 6.020931774743061e-07, + "loss": 0.6807, + "num_input_tokens_seen": 271943725, + "step": 8360 + }, + { + "epoch": 0.7540244397348604, + "flos": 24646311700800.0, + "grad_norm": 2.558490966596045, + "learning_rate": 6.016754288375546e-07, + "loss": 0.6684, + "num_input_tokens_seen": 271970305, + "step": 8361 + }, + { + "epoch": 0.7541146232583307, + "flos": 20966446743840.0, + "grad_norm": 2.0151539333413306, + "learning_rate": 6.012577995135665e-07, + "loss": 0.7375, + "num_input_tokens_seen": 271998520, + "step": 8362 + }, + { + "epoch": 0.754204806781801, + "flos": 18416348534880.0, + "grad_norm": 2.2895100126828556, + "learning_rate": 6.008402895379743e-07, + "loss": 0.6904, + "num_input_tokens_seen": 272024835, + "step": 8363 + }, + { + "epoch": 0.7542949903052713, + "flos": 23880754427520.0, + "grad_norm": 4.14037647581113, + "learning_rate": 6.004228989464047e-07, + "loss": 0.8318, + "num_input_tokens_seen": 272049870, + "step": 8364 + }, + { + "epoch": 0.7543851738287415, + "flos": 18416199855840.0, + "grad_norm": 2.2938659056605415, + "learning_rate": 6.000056277744692e-07, + "loss": 0.7677, + "num_input_tokens_seen": 272079220, + "step": 8365 + }, + { + "epoch": 0.7544753573522117, + "flos": 21401327399520.0, + "grad_norm": 1.6896180177163198, + "learning_rate": 5.995884760577745e-07, + "loss": 0.7144, + "num_input_tokens_seen": 272109685, + "step": 8366 + }, + { + "epoch": 0.7545655408756821, + "flos": 64537632396960.0, + "grad_norm": 0.622662967085654, + "learning_rate": 5.99171443831912e-07, + "loss": 0.5643, + "num_input_tokens_seen": 272203145, + "step": 8367 + }, + { + "epoch": 0.7546557243991523, + "flos": 23990794117920.0, + "grad_norm": 1.5711318578134141, + "learning_rate": 5.98754531132466e-07, + "loss": 0.743, + "num_input_tokens_seen": 272234515, + "step": 8368 + }, + { + "epoch": 0.7547459079226225, + "flos": 25083534051360.0, + "grad_norm": 2.078721957404667, + "learning_rate": 5.983377379950099e-07, + "loss": 0.7457, + "num_input_tokens_seen": 272262530, + "step": 8369 + }, + { + "epoch": 0.7548360914460928, + "flos": 21804496516320.0, + "grad_norm": 2.8875557766105766, + "learning_rate": 5.979210644551067e-07, + "loss": 0.6824, + "num_input_tokens_seen": 272290570, + "step": 8370 + }, + { + "epoch": 0.7549262749695631, + "flos": 20820272313120.0, + "grad_norm": 2.6368403653955514, + "learning_rate": 5.975045105483091e-07, + "loss": 0.7431, + "num_input_tokens_seen": 272314635, + "step": 8371 + }, + { + "epoch": 0.7550164584930333, + "flos": 69322912175520.0, + "grad_norm": 0.60221767664684, + "learning_rate": 5.970880763101607e-07, + "loss": 0.587, + "num_input_tokens_seen": 272413170, + "step": 8372 + }, + { + "epoch": 0.7551066420165036, + "flos": 17936523905760.0, + "grad_norm": 2.3984668245293532, + "learning_rate": 5.966717617761925e-07, + "loss": 0.7686, + "num_input_tokens_seen": 272438790, + "step": 8373 + }, + { + "epoch": 0.7551968255399738, + "flos": 22059669884160.0, + "grad_norm": 4.159074784241453, + "learning_rate": 5.962555669819276e-07, + "loss": 0.7259, + "num_input_tokens_seen": 272469395, + "step": 8374 + }, + { + "epoch": 0.7552870090634441, + "flos": 27705158345760.0, + "grad_norm": 1.4553255216488175, + "learning_rate": 5.958394919628777e-07, + "loss": 0.7654, + "num_input_tokens_seen": 272501060, + "step": 8375 + }, + { + "epoch": 0.7553771925869144, + "flos": 25301569095360.0, + "grad_norm": 1.4719015357992724, + "learning_rate": 5.954235367545451e-07, + "loss": 0.6841, + "num_input_tokens_seen": 272533425, + "step": 8376 + }, + { + "epoch": 0.7554673761103846, + "flos": 22565147381280.0, + "grad_norm": 2.0380332950945776, + "learning_rate": 5.950077013924213e-07, + "loss": 0.7397, + "num_input_tokens_seen": 272562615, + "step": 8377 + }, + { + "epoch": 0.7555575596338548, + "flos": 19654891201440.0, + "grad_norm": 2.158720950321607, + "learning_rate": 5.945919859119865e-07, + "loss": 0.8285, + "num_input_tokens_seen": 272589100, + "step": 8378 + }, + { + "epoch": 0.7556477431573252, + "flos": 21586089774720.0, + "grad_norm": 1.928869467871666, + "learning_rate": 5.94176390348714e-07, + "loss": 0.6335, + "num_input_tokens_seen": 272618220, + "step": 8379 + }, + { + "epoch": 0.7557379266807954, + "flos": 24427979298720.0, + "grad_norm": 1.4138051425827898, + "learning_rate": 5.937609147380622e-07, + "loss": 0.8506, + "num_input_tokens_seen": 272649635, + "step": 8380 + }, + { + "epoch": 0.7558281102042657, + "flos": 21622744891680.0, + "grad_norm": 1.72982121530526, + "learning_rate": 5.933455591154844e-07, + "loss": 0.6887, + "num_input_tokens_seen": 272676840, + "step": 8381 + }, + { + "epoch": 0.7559182937277359, + "flos": 21913644132480.0, + "grad_norm": 2.5767140544598517, + "learning_rate": 5.929303235164191e-07, + "loss": 0.7165, + "num_input_tokens_seen": 272703840, + "step": 8382 + }, + { + "epoch": 0.7560084772512062, + "flos": 23370965238240.0, + "grad_norm": 2.2211029041811443, + "learning_rate": 5.92515207976297e-07, + "loss": 0.7165, + "num_input_tokens_seen": 272731765, + "step": 8383 + }, + { + "epoch": 0.7560986607746765, + "flos": 28253684158560.0, + "grad_norm": 1.724509490113374, + "learning_rate": 5.921002125305383e-07, + "loss": 0.735, + "num_input_tokens_seen": 272761255, + "step": 8384 + }, + { + "epoch": 0.7561888442981467, + "flos": 25409155581600.0, + "grad_norm": 2.552998769153936, + "learning_rate": 5.916853372145525e-07, + "loss": 0.6656, + "num_input_tokens_seen": 272788210, + "step": 8385 + }, + { + "epoch": 0.756279027821617, + "flos": 25264765299360.0, + "grad_norm": 1.895389574815643, + "learning_rate": 5.912705820637389e-07, + "loss": 0.6737, + "num_input_tokens_seen": 272816560, + "step": 8386 + }, + { + "epoch": 0.7563692113450873, + "flos": 30619540368960.0, + "grad_norm": 1.6556499041840365, + "learning_rate": 5.908559471134871e-07, + "loss": 0.8113, + "num_input_tokens_seen": 272848745, + "step": 8387 + }, + { + "epoch": 0.7564593948685575, + "flos": 23840419504320.0, + "grad_norm": 1.8330087880216837, + "learning_rate": 5.904414323991764e-07, + "loss": 0.6827, + "num_input_tokens_seen": 272875625, + "step": 8388 + }, + { + "epoch": 0.7565495783920277, + "flos": 23293975198080.0, + "grad_norm": 1.8420772399778553, + "learning_rate": 5.900270379561743e-07, + "loss": 0.7513, + "num_input_tokens_seen": 272903345, + "step": 8389 + }, + { + "epoch": 0.7566397619154981, + "flos": 20711050357440.0, + "grad_norm": 2.005830253827983, + "learning_rate": 5.896127638198399e-07, + "loss": 0.7635, + "num_input_tokens_seen": 272931010, + "step": 8390 + }, + { + "epoch": 0.7567299454389683, + "flos": 17686926001920.0, + "grad_norm": 2.030633451350863, + "learning_rate": 5.89198610025521e-07, + "loss": 0.7475, + "num_input_tokens_seen": 272955005, + "step": 8391 + }, + { + "epoch": 0.7568201289624386, + "flos": 25666224607200.0, + "grad_norm": 2.212262407035062, + "learning_rate": 5.887845766085559e-07, + "loss": 0.8116, + "num_input_tokens_seen": 272986640, + "step": 8392 + }, + { + "epoch": 0.7569103124859088, + "flos": 20378738270400.0, + "grad_norm": 2.203275959443277, + "learning_rate": 5.883706636042722e-07, + "loss": 0.7067, + "num_input_tokens_seen": 273014425, + "step": 8393 + }, + { + "epoch": 0.7570004960093791, + "flos": 23516433443520.0, + "grad_norm": 3.1477292307793463, + "learning_rate": 5.879568710479879e-07, + "loss": 0.6861, + "num_input_tokens_seen": 273041105, + "step": 8394 + }, + { + "epoch": 0.7570906795328494, + "flos": 27335633595360.0, + "grad_norm": 1.7698079076077529, + "learning_rate": 5.875431989750078e-07, + "loss": 0.8344, + "num_input_tokens_seen": 273072730, + "step": 8395 + }, + { + "epoch": 0.7571808630563196, + "flos": 20856927430080.0, + "grad_norm": 1.5820673432177128, + "learning_rate": 5.871296474206313e-07, + "loss": 0.6808, + "num_input_tokens_seen": 273101790, + "step": 8396 + }, + { + "epoch": 0.7572710465797898, + "flos": 21877323543360.0, + "grad_norm": 1.6223459577981687, + "learning_rate": 5.867162164201427e-07, + "loss": 0.7504, + "num_input_tokens_seen": 273130445, + "step": 8397 + }, + { + "epoch": 0.7573612301032602, + "flos": 21512482182720.0, + "grad_norm": 1.898334139298203, + "learning_rate": 5.863029060088205e-07, + "loss": 0.7748, + "num_input_tokens_seen": 273159285, + "step": 8398 + }, + { + "epoch": 0.7574514136267304, + "flos": 20420559984000.0, + "grad_norm": 1.7710416633879087, + "learning_rate": 5.858897162219289e-07, + "loss": 0.6433, + "num_input_tokens_seen": 273185100, + "step": 8399 + }, + { + "epoch": 0.7575415971502006, + "flos": 24026445651360.0, + "grad_norm": 2.0066521404961457, + "learning_rate": 5.854766470947238e-07, + "loss": 0.7927, + "num_input_tokens_seen": 273212410, + "step": 8400 + }, + { + "epoch": 0.7576317806736709, + "flos": 29231217804960.0, + "grad_norm": 1.6700825574878246, + "learning_rate": 5.850636986624511e-07, + "loss": 0.7362, + "num_input_tokens_seen": 273242095, + "step": 8401 + }, + { + "epoch": 0.7577219641971412, + "flos": 23406653941440.0, + "grad_norm": 2.3936896533181966, + "learning_rate": 5.846508709603453e-07, + "loss": 0.7553, + "num_input_tokens_seen": 273268580, + "step": 8402 + }, + { + "epoch": 0.7578121477206115, + "flos": 24242101830720.0, + "grad_norm": 1.7411408295266149, + "learning_rate": 5.842381640236318e-07, + "loss": 0.7215, + "num_input_tokens_seen": 273300680, + "step": 8403 + }, + { + "epoch": 0.7579023312440817, + "flos": 24792263112960.0, + "grad_norm": 1.6616272636162266, + "learning_rate": 5.838255778875242e-07, + "loss": 0.7442, + "num_input_tokens_seen": 273330705, + "step": 8404 + }, + { + "epoch": 0.7579925147675519, + "flos": 28362014040000.0, + "grad_norm": 2.395323978404035, + "learning_rate": 5.83413112587227e-07, + "loss": 0.7633, + "num_input_tokens_seen": 273358650, + "step": 8405 + }, + { + "epoch": 0.7580826982910223, + "flos": 36923668673280.0, + "grad_norm": 2.100810316969269, + "learning_rate": 5.830007681579338e-07, + "loss": 0.7319, + "num_input_tokens_seen": 273389885, + "step": 8406 + }, + { + "epoch": 0.7581728818144925, + "flos": 18116714400480.0, + "grad_norm": 1.9007941794749452, + "learning_rate": 5.825885446348284e-07, + "loss": 0.7603, + "num_input_tokens_seen": 273412375, + "step": 8407 + }, + { + "epoch": 0.7582630653379627, + "flos": 35682784311840.0, + "grad_norm": 1.7784065171222305, + "learning_rate": 5.821764420530842e-07, + "loss": 0.6809, + "num_input_tokens_seen": 273445370, + "step": 8408 + }, + { + "epoch": 0.7583532488614331, + "flos": 22022940427680.0, + "grad_norm": 1.6822527791381683, + "learning_rate": 5.817644604478633e-07, + "loss": 0.7625, + "num_input_tokens_seen": 273476405, + "step": 8409 + }, + { + "epoch": 0.7584434323849033, + "flos": 21549062960160.0, + "grad_norm": 1.8165695937349733, + "learning_rate": 5.81352599854319e-07, + "loss": 0.6909, + "num_input_tokens_seen": 273505360, + "step": 8410 + }, + { + "epoch": 0.7585336159083735, + "flos": 21289838088480.0, + "grad_norm": 1.6147218788345374, + "learning_rate": 5.809408603075938e-07, + "loss": 0.7764, + "num_input_tokens_seen": 273531660, + "step": 8411 + }, + { + "epoch": 0.7586237994318438, + "flos": 23917669732800.0, + "grad_norm": 2.2679627032434246, + "learning_rate": 5.805292418428176e-07, + "loss": 0.7257, + "num_input_tokens_seen": 273559955, + "step": 8412 + }, + { + "epoch": 0.7587139829553141, + "flos": 22933185341280.0, + "grad_norm": 2.4988653469131092, + "learning_rate": 5.801177444951148e-07, + "loss": 0.7353, + "num_input_tokens_seen": 273585930, + "step": 8413 + }, + { + "epoch": 0.7588041664787843, + "flos": 23954324849760.0, + "grad_norm": 1.7188413422328068, + "learning_rate": 5.797063682995944e-07, + "loss": 0.7754, + "num_input_tokens_seen": 273616660, + "step": 8414 + }, + { + "epoch": 0.7588943500022546, + "flos": 23698110728640.0, + "grad_norm": 1.99392606382358, + "learning_rate": 5.792951132913584e-07, + "loss": 0.8103, + "num_input_tokens_seen": 273643380, + "step": 8415 + }, + { + "epoch": 0.7589845335257248, + "flos": 22059409695840.0, + "grad_norm": 2.3270909618046605, + "learning_rate": 5.788839795054968e-07, + "loss": 0.7981, + "num_input_tokens_seen": 273672095, + "step": 8416 + }, + { + "epoch": 0.7590747170491952, + "flos": 21694531165440.0, + "grad_norm": 1.7280849459043472, + "learning_rate": 5.784729669770898e-07, + "loss": 0.6615, + "num_input_tokens_seen": 273702550, + "step": 8417 + }, + { + "epoch": 0.7591649005726654, + "flos": 21692895696000.0, + "grad_norm": 1.5019778880842238, + "learning_rate": 5.780620757412084e-07, + "loss": 0.742, + "num_input_tokens_seen": 273732730, + "step": 8418 + }, + { + "epoch": 0.7592550840961356, + "flos": 21986508329280.0, + "grad_norm": 1.9647833206257828, + "learning_rate": 5.776513058329098e-07, + "loss": 0.7636, + "num_input_tokens_seen": 273759255, + "step": 8419 + }, + { + "epoch": 0.7593452676196059, + "flos": 18994838907840.0, + "grad_norm": 1.7259770432113595, + "learning_rate": 5.772406572872459e-07, + "loss": 0.7687, + "num_input_tokens_seen": 273786900, + "step": 8420 + }, + { + "epoch": 0.7594354511430762, + "flos": 13534298670240.0, + "grad_norm": 2.334236367640053, + "learning_rate": 5.768301301392535e-07, + "loss": 0.7736, + "num_input_tokens_seen": 273812305, + "step": 8421 + }, + { + "epoch": 0.7595256346665464, + "flos": 26176088136000.0, + "grad_norm": 3.8712291212692116, + "learning_rate": 5.764197244239615e-07, + "loss": 0.7544, + "num_input_tokens_seen": 273841520, + "step": 8422 + }, + { + "epoch": 0.7596158181900167, + "flos": 26686880908800.0, + "grad_norm": 1.854471295938637, + "learning_rate": 5.760094401763884e-07, + "loss": 0.7836, + "num_input_tokens_seen": 273871815, + "step": 8423 + }, + { + "epoch": 0.7597060017134869, + "flos": 32474826825120.0, + "grad_norm": 4.50646550566625, + "learning_rate": 5.755992774315414e-07, + "loss": 0.7325, + "num_input_tokens_seen": 273905185, + "step": 8424 + }, + { + "epoch": 0.7597961852369572, + "flos": 21913123755840.0, + "grad_norm": 2.987391999020307, + "learning_rate": 5.751892362244183e-07, + "loss": 0.7345, + "num_input_tokens_seen": 273935525, + "step": 8425 + }, + { + "epoch": 0.7598863687604275, + "flos": 18379842096960.0, + "grad_norm": 2.0834931757325186, + "learning_rate": 5.747793165900065e-07, + "loss": 0.8011, + "num_input_tokens_seen": 273961700, + "step": 8426 + }, + { + "epoch": 0.7599765522838977, + "flos": 19509237147360.0, + "grad_norm": 1.8606092700030648, + "learning_rate": 5.743695185632806e-07, + "loss": 0.8017, + "num_input_tokens_seen": 273989665, + "step": 8427 + }, + { + "epoch": 0.7600667358073679, + "flos": 26321965208640.0, + "grad_norm": 2.015837902091774, + "learning_rate": 5.739598421792098e-07, + "loss": 0.7846, + "num_input_tokens_seen": 274014205, + "step": 8428 + }, + { + "epoch": 0.7601569193308383, + "flos": 22096473680160.0, + "grad_norm": 2.187652364601645, + "learning_rate": 5.735502874727474e-07, + "loss": 0.7278, + "num_input_tokens_seen": 274044195, + "step": 8429 + }, + { + "epoch": 0.7602471028543085, + "flos": 29194153820640.0, + "grad_norm": 1.7361466577917462, + "learning_rate": 5.731408544788398e-07, + "loss": 0.8749, + "num_input_tokens_seen": 274073325, + "step": 8430 + }, + { + "epoch": 0.7603372863777788, + "flos": 26683572800160.0, + "grad_norm": 2.1640057607986574, + "learning_rate": 5.727315432324225e-07, + "loss": 0.6616, + "num_input_tokens_seen": 274106830, + "step": 8431 + }, + { + "epoch": 0.760427469901249, + "flos": 20309776898400.0, + "grad_norm": 2.1021971471073773, + "learning_rate": 5.723223537684196e-07, + "loss": 0.7312, + "num_input_tokens_seen": 274133935, + "step": 8432 + }, + { + "epoch": 0.7605176534247193, + "flos": 18481667270400.0, + "grad_norm": 2.1465754537376194, + "learning_rate": 5.719132861217462e-07, + "loss": 0.8245, + "num_input_tokens_seen": 274160890, + "step": 8433 + }, + { + "epoch": 0.7606078369481896, + "flos": 44355073351680.0, + "grad_norm": 4.016762182806886, + "learning_rate": 5.715043403273044e-07, + "loss": 0.7125, + "num_input_tokens_seen": 274196410, + "step": 8434 + }, + { + "epoch": 0.7606980204716598, + "flos": 24093920232960.0, + "grad_norm": 3.3961901019012397, + "learning_rate": 5.710955164199902e-07, + "loss": 0.699, + "num_input_tokens_seen": 274219720, + "step": 8435 + }, + { + "epoch": 0.7607882039951301, + "flos": 20015569548960.0, + "grad_norm": 1.8973927704036542, + "learning_rate": 5.706868144346841e-07, + "loss": 0.7658, + "num_input_tokens_seen": 274246830, + "step": 8436 + }, + { + "epoch": 0.7608783875186004, + "flos": 23881609332000.0, + "grad_norm": 1.8016156230945313, + "learning_rate": 5.702782344062613e-07, + "loss": 0.7388, + "num_input_tokens_seen": 274276500, + "step": 8437 + }, + { + "epoch": 0.7609685710420706, + "flos": 28216397155680.0, + "grad_norm": 1.9705382808402159, + "learning_rate": 5.698697763695826e-07, + "loss": 0.7231, + "num_input_tokens_seen": 274306380, + "step": 8438 + }, + { + "epoch": 0.7610587545655408, + "flos": 21840817105440.0, + "grad_norm": 1.9275699997596427, + "learning_rate": 5.694614403595002e-07, + "loss": 0.7672, + "num_input_tokens_seen": 274334090, + "step": 8439 + }, + { + "epoch": 0.7611489380890112, + "flos": 34440487499520.0, + "grad_norm": 2.0118026567756515, + "learning_rate": 5.690532264108554e-07, + "loss": 0.6558, + "num_input_tokens_seen": 274364860, + "step": 8440 + }, + { + "epoch": 0.7612391216124814, + "flos": 22346480451360.0, + "grad_norm": 1.9443077955216341, + "learning_rate": 5.686451345584795e-07, + "loss": 0.7784, + "num_input_tokens_seen": 274393910, + "step": 8441 + }, + { + "epoch": 0.7613293051359517, + "flos": 25592802864000.0, + "grad_norm": 2.3429479625424645, + "learning_rate": 5.682371648371933e-07, + "loss": 0.842, + "num_input_tokens_seen": 274420970, + "step": 8442 + }, + { + "epoch": 0.7614194886594219, + "flos": 22017067605600.0, + "grad_norm": 1.9462967186761417, + "learning_rate": 5.678293172818074e-07, + "loss": 0.7059, + "num_input_tokens_seen": 274448655, + "step": 8443 + }, + { + "epoch": 0.7615096721828922, + "flos": 26206238544960.0, + "grad_norm": 1.9948501445541982, + "learning_rate": 5.674215919271204e-07, + "loss": 0.7594, + "num_input_tokens_seen": 274475885, + "step": 8444 + }, + { + "epoch": 0.7615998557063625, + "flos": 18233667666240.0, + "grad_norm": 2.0443718433734777, + "learning_rate": 5.670139888079224e-07, + "loss": 0.6717, + "num_input_tokens_seen": 274503955, + "step": 8445 + }, + { + "epoch": 0.7616900392298327, + "flos": 24570994299840.0, + "grad_norm": 1.9979009341938812, + "learning_rate": 5.666065079589924e-07, + "loss": 0.8027, + "num_input_tokens_seen": 274535095, + "step": 8446 + }, + { + "epoch": 0.7617802227533029, + "flos": 25047250632000.0, + "grad_norm": 1.7691300971117392, + "learning_rate": 5.661991494150986e-07, + "loss": 0.6979, + "num_input_tokens_seen": 274564485, + "step": 8447 + }, + { + "epoch": 0.7618704062767733, + "flos": 19324809300000.0, + "grad_norm": 1.928234588130182, + "learning_rate": 5.657919132109999e-07, + "loss": 0.7917, + "num_input_tokens_seen": 274593665, + "step": 8448 + }, + { + "epoch": 0.7619605898002435, + "flos": 24679435690560.0, + "grad_norm": 1.858462326066156, + "learning_rate": 5.653847993814421e-07, + "loss": 0.7781, + "num_input_tokens_seen": 274620995, + "step": 8449 + }, + { + "epoch": 0.7620507733237137, + "flos": 22860432653760.0, + "grad_norm": 3.0729861666476332, + "learning_rate": 5.649778079611647e-07, + "loss": 0.7551, + "num_input_tokens_seen": 274650455, + "step": 8450 + }, + { + "epoch": 0.762140956847184, + "flos": 22819688863200.0, + "grad_norm": 1.807019465136331, + "learning_rate": 5.645709389848923e-07, + "loss": 0.7263, + "num_input_tokens_seen": 274678110, + "step": 8451 + }, + { + "epoch": 0.7622311403706543, + "flos": 17760533593920.0, + "grad_norm": 1.9554221720373801, + "learning_rate": 5.641641924873435e-07, + "loss": 0.7081, + "num_input_tokens_seen": 274706130, + "step": 8452 + }, + { + "epoch": 0.7623213238941245, + "flos": 21549323148480.0, + "grad_norm": 2.006473554948827, + "learning_rate": 5.637575685032217e-07, + "loss": 0.7429, + "num_input_tokens_seen": 274734175, + "step": 8453 + }, + { + "epoch": 0.7624115074175948, + "flos": 29200063812480.0, + "grad_norm": 2.246922719817437, + "learning_rate": 5.633510670672246e-07, + "loss": 0.8237, + "num_input_tokens_seen": 274763445, + "step": 8454 + }, + { + "epoch": 0.762501690941065, + "flos": 65860450376640.0, + "grad_norm": 0.6231935621400398, + "learning_rate": 5.629446882140354e-07, + "loss": 0.6044, + "num_input_tokens_seen": 274851425, + "step": 8455 + }, + { + "epoch": 0.7625918744645354, + "flos": 26101848658080.0, + "grad_norm": 1.5529482520251656, + "learning_rate": 5.625384319783295e-07, + "loss": 0.7559, + "num_input_tokens_seen": 274880260, + "step": 8456 + }, + { + "epoch": 0.7626820579880056, + "flos": 24645716984640.0, + "grad_norm": 8.289257165922226, + "learning_rate": 5.621322983947705e-07, + "loss": 0.6941, + "num_input_tokens_seen": 274907560, + "step": 8457 + }, + { + "epoch": 0.7627722415114758, + "flos": 26500111366560.0, + "grad_norm": 2.3912446459377747, + "learning_rate": 5.617262874980122e-07, + "loss": 0.7806, + "num_input_tokens_seen": 274934815, + "step": 8458 + }, + { + "epoch": 0.7628624250349462, + "flos": 27011350176480.0, + "grad_norm": 1.9390732315344406, + "learning_rate": 5.613203993226981e-07, + "loss": 0.7904, + "num_input_tokens_seen": 274968660, + "step": 8459 + }, + { + "epoch": 0.7629526085584164, + "flos": 18298837722720.0, + "grad_norm": 2.7359037648838394, + "learning_rate": 5.609146339034599e-07, + "loss": 0.7999, + "num_input_tokens_seen": 274992070, + "step": 8460 + }, + { + "epoch": 0.7630427920818866, + "flos": 17760161896320.0, + "grad_norm": 2.5750333802365115, + "learning_rate": 5.605089912749199e-07, + "loss": 0.6839, + "num_input_tokens_seen": 275020325, + "step": 8461 + }, + { + "epoch": 0.7631329756053569, + "flos": 23875625000640.0, + "grad_norm": 2.5582201502943067, + "learning_rate": 5.601034714716901e-07, + "loss": 0.6899, + "num_input_tokens_seen": 275048195, + "step": 8462 + }, + { + "epoch": 0.7632231591288272, + "flos": 24791556887520.0, + "grad_norm": 1.7776855140162346, + "learning_rate": 5.59698074528372e-07, + "loss": 0.7176, + "num_input_tokens_seen": 275080270, + "step": 8463 + }, + { + "epoch": 0.7633133426522974, + "flos": 22714592750880.0, + "grad_norm": 1.6711398495971608, + "learning_rate": 5.592928004795555e-07, + "loss": 0.7168, + "num_input_tokens_seen": 275109710, + "step": 8464 + }, + { + "epoch": 0.7634035261757677, + "flos": 17942359558080.0, + "grad_norm": 2.3430795346354425, + "learning_rate": 5.58887649359822e-07, + "loss": 0.7505, + "num_input_tokens_seen": 275135835, + "step": 8465 + }, + { + "epoch": 0.7634937096992379, + "flos": 34520116592640.0, + "grad_norm": 2.0207271613581668, + "learning_rate": 5.584826212037393e-07, + "loss": 0.5988, + "num_input_tokens_seen": 275165945, + "step": 8466 + }, + { + "epoch": 0.7635838932227083, + "flos": 39910543195680.0, + "grad_norm": 1.6529710088588758, + "learning_rate": 5.580777160458689e-07, + "loss": 0.6645, + "num_input_tokens_seen": 275199635, + "step": 8467 + }, + { + "epoch": 0.7636740767461785, + "flos": 26030842949280.0, + "grad_norm": 1.6348049291263396, + "learning_rate": 5.576729339207574e-07, + "loss": 0.7669, + "num_input_tokens_seen": 275229275, + "step": 8468 + }, + { + "epoch": 0.7637642602696487, + "flos": 26904841613280.0, + "grad_norm": 1.33797754158536, + "learning_rate": 5.572682748629449e-07, + "loss": 0.8034, + "num_input_tokens_seen": 275261900, + "step": 8469 + }, + { + "epoch": 0.763854443793119, + "flos": 29346089564160.0, + "grad_norm": 1.6325980597775525, + "learning_rate": 5.568637389069582e-07, + "loss": 0.7298, + "num_input_tokens_seen": 275292625, + "step": 8470 + }, + { + "epoch": 0.7639446273165893, + "flos": 20198176078080.0, + "grad_norm": 2.1170692551591546, + "learning_rate": 5.564593260873145e-07, + "loss": 0.755, + "num_input_tokens_seen": 275319065, + "step": 8471 + }, + { + "epoch": 0.7640348108400595, + "flos": 21547501830240.0, + "grad_norm": 1.6777500778014054, + "learning_rate": 5.560550364385206e-07, + "loss": 0.7812, + "num_input_tokens_seen": 275348345, + "step": 8472 + }, + { + "epoch": 0.7641249943635298, + "flos": 22751619565440.0, + "grad_norm": 1.897552221029454, + "learning_rate": 5.556508699950728e-07, + "loss": 0.8167, + "num_input_tokens_seen": 275374790, + "step": 8473 + }, + { + "epoch": 0.764215177887, + "flos": 17978271279840.0, + "grad_norm": 1.925680904598381, + "learning_rate": 5.552468267914577e-07, + "loss": 0.6966, + "num_input_tokens_seen": 275398680, + "step": 8474 + }, + { + "epoch": 0.7643053614104703, + "flos": 22459976929440.0, + "grad_norm": 2.3839225809788434, + "learning_rate": 5.548429068621481e-07, + "loss": 0.6581, + "num_input_tokens_seen": 275427940, + "step": 8475 + }, + { + "epoch": 0.7643955449339406, + "flos": 61970690553120.0, + "grad_norm": 0.6516548448069194, + "learning_rate": 5.544391102416115e-07, + "loss": 0.5738, + "num_input_tokens_seen": 275517440, + "step": 8476 + }, + { + "epoch": 0.7644857284574108, + "flos": 17614507842240.0, + "grad_norm": 3.1214768958739083, + "learning_rate": 5.540354369643003e-07, + "loss": 0.7997, + "num_input_tokens_seen": 275543225, + "step": 8477 + }, + { + "epoch": 0.764575911980881, + "flos": 27487123301760.0, + "grad_norm": 3.675261798278957, + "learning_rate": 5.536318870646586e-07, + "loss": 0.5956, + "num_input_tokens_seen": 275572645, + "step": 8478 + }, + { + "epoch": 0.7646660955043514, + "flos": 25629383641440.0, + "grad_norm": 2.738504136791939, + "learning_rate": 5.532284605771194e-07, + "loss": 0.6496, + "num_input_tokens_seen": 275600200, + "step": 8479 + }, + { + "epoch": 0.7647562790278216, + "flos": 32660704293120.0, + "grad_norm": 1.4329042361645494, + "learning_rate": 5.528251575361052e-07, + "loss": 0.7122, + "num_input_tokens_seen": 275631320, + "step": 8480 + }, + { + "epoch": 0.7648464625512919, + "flos": 22860878690880.0, + "grad_norm": 1.70033630803131, + "learning_rate": 5.524219779760284e-07, + "loss": 0.7409, + "num_input_tokens_seen": 275660700, + "step": 8481 + }, + { + "epoch": 0.7649366460747622, + "flos": 24609619414080.0, + "grad_norm": 1.9037285059255906, + "learning_rate": 5.520189219312907e-07, + "loss": 0.7646, + "num_input_tokens_seen": 275690625, + "step": 8482 + }, + { + "epoch": 0.7650268295982324, + "flos": 23151926610720.0, + "grad_norm": 2.2555143810871647, + "learning_rate": 5.516159894362817e-07, + "loss": 0.6302, + "num_input_tokens_seen": 275722165, + "step": 8483 + }, + { + "epoch": 0.7651170131217027, + "flos": 23330518805760.0, + "grad_norm": 1.7200999420538583, + "learning_rate": 5.512131805253839e-07, + "loss": 0.6796, + "num_input_tokens_seen": 275750625, + "step": 8484 + }, + { + "epoch": 0.7652071966451729, + "flos": 27632703016320.0, + "grad_norm": 1.6060205703928654, + "learning_rate": 5.508104952329653e-07, + "loss": 0.7581, + "num_input_tokens_seen": 275781940, + "step": 8485 + }, + { + "epoch": 0.7652973801686432, + "flos": 25775372223360.0, + "grad_norm": 2.664325510952858, + "learning_rate": 5.504079335933862e-07, + "loss": 0.7713, + "num_input_tokens_seen": 275808015, + "step": 8486 + }, + { + "epoch": 0.7653875636921135, + "flos": 16301725697760.0, + "grad_norm": 2.4539916821861913, + "learning_rate": 5.500054956409952e-07, + "loss": 0.7746, + "num_input_tokens_seen": 275834255, + "step": 8487 + }, + { + "epoch": 0.7654777472155837, + "flos": 21803976139680.0, + "grad_norm": 1.9997523869775509, + "learning_rate": 5.496031814101303e-07, + "loss": 0.7044, + "num_input_tokens_seen": 275859205, + "step": 8488 + }, + { + "epoch": 0.7655679307390539, + "flos": 20485358342880.0, + "grad_norm": 1.7610646321286865, + "learning_rate": 5.492009909351203e-07, + "loss": 0.7502, + "num_input_tokens_seen": 275887060, + "step": 8489 + }, + { + "epoch": 0.7656581142625243, + "flos": 23626027096800.0, + "grad_norm": 2.1821320201470007, + "learning_rate": 5.4879892425028e-07, + "loss": 0.6689, + "num_input_tokens_seen": 275916210, + "step": 8490 + }, + { + "epoch": 0.7657482977859945, + "flos": 21944760955200.0, + "grad_norm": 1.8412571234477864, + "learning_rate": 5.483969813899184e-07, + "loss": 0.8803, + "num_input_tokens_seen": 275944855, + "step": 8491 + }, + { + "epoch": 0.7658384813094647, + "flos": 32259988380480.0, + "grad_norm": 1.5109220864930915, + "learning_rate": 5.479951623883299e-07, + "loss": 0.8248, + "num_input_tokens_seen": 275976340, + "step": 8492 + }, + { + "epoch": 0.765928664832935, + "flos": 27336414160320.0, + "grad_norm": 2.3950286398431535, + "learning_rate": 5.475934672798004e-07, + "loss": 0.6786, + "num_input_tokens_seen": 276005245, + "step": 8493 + }, + { + "epoch": 0.7660188483564053, + "flos": 19833409056960.0, + "grad_norm": 1.904837601014749, + "learning_rate": 5.471918960986047e-07, + "loss": 0.7061, + "num_input_tokens_seen": 276032760, + "step": 8494 + }, + { + "epoch": 0.7661090318798756, + "flos": 26176013796480.0, + "grad_norm": 1.9540863112770035, + "learning_rate": 5.467904488790071e-07, + "loss": 0.8397, + "num_input_tokens_seen": 276060005, + "step": 8495 + }, + { + "epoch": 0.7661992154033458, + "flos": 26392450540800.0, + "grad_norm": 2.2372189009565244, + "learning_rate": 5.463891256552615e-07, + "loss": 0.7478, + "num_input_tokens_seen": 276089610, + "step": 8496 + }, + { + "epoch": 0.766289398926816, + "flos": 21620923573440.0, + "grad_norm": 1.9258006870612028, + "learning_rate": 5.459879264616107e-07, + "loss": 0.6511, + "num_input_tokens_seen": 276117330, + "step": 8497 + }, + { + "epoch": 0.7663795824502864, + "flos": 27706310608320.0, + "grad_norm": 2.225761189074105, + "learning_rate": 5.455868513322874e-07, + "loss": 0.6284, + "num_input_tokens_seen": 276145240, + "step": 8498 + }, + { + "epoch": 0.7664697659737566, + "flos": 24828212004480.0, + "grad_norm": 1.887899278257442, + "learning_rate": 5.451859003015143e-07, + "loss": 0.6323, + "num_input_tokens_seen": 276174005, + "step": 8499 + }, + { + "epoch": 0.7665599494972268, + "flos": 25808719231680.0, + "grad_norm": 1.8629387434769915, + "learning_rate": 5.447850734035009e-07, + "loss": 0.7805, + "num_input_tokens_seen": 276201610, + "step": 8500 + }, + { + "epoch": 0.7666501330206971, + "flos": 19509125638080.0, + "grad_norm": 9.517915797608797, + "learning_rate": 5.443843706724494e-07, + "loss": 0.7247, + "num_input_tokens_seen": 276229425, + "step": 8501 + }, + { + "epoch": 0.7667403165441674, + "flos": 22492134505440.0, + "grad_norm": 2.865417433015596, + "learning_rate": 5.439837921425494e-07, + "loss": 0.6748, + "num_input_tokens_seen": 276256905, + "step": 8502 + }, + { + "epoch": 0.7668305000676376, + "flos": 24932787740160.0, + "grad_norm": 2.3360354808046315, + "learning_rate": 5.435833378479807e-07, + "loss": 0.7351, + "num_input_tokens_seen": 276287320, + "step": 8503 + }, + { + "epoch": 0.7669206835911079, + "flos": 25884965876640.0, + "grad_norm": 1.609460328441843, + "learning_rate": 5.431830078229128e-07, + "loss": 0.6, + "num_input_tokens_seen": 276318110, + "step": 8504 + }, + { + "epoch": 0.7670108671145782, + "flos": 19978133867040.0, + "grad_norm": 3.2556045331804113, + "learning_rate": 5.427828021015022e-07, + "loss": 0.76, + "num_input_tokens_seen": 276346145, + "step": 8505 + }, + { + "epoch": 0.7671010506380485, + "flos": 29375756766240.0, + "grad_norm": 1.9045364933378321, + "learning_rate": 5.42382720717899e-07, + "loss": 0.773, + "num_input_tokens_seen": 276374490, + "step": 8506 + }, + { + "epoch": 0.7671912341615187, + "flos": 24646125852000.0, + "grad_norm": 1.6988517277161896, + "learning_rate": 5.419827637062384e-07, + "loss": 0.7741, + "num_input_tokens_seen": 276404295, + "step": 8507 + }, + { + "epoch": 0.7672814176849889, + "flos": 17572611789120.0, + "grad_norm": 2.6691178937067153, + "learning_rate": 5.415829311006487e-07, + "loss": 0.8852, + "num_input_tokens_seen": 276428910, + "step": 8508 + }, + { + "epoch": 0.7673716012084593, + "flos": 22018926093600.0, + "grad_norm": 1.9809294697059723, + "learning_rate": 5.411832229352447e-07, + "loss": 0.8256, + "num_input_tokens_seen": 276455130, + "step": 8509 + }, + { + "epoch": 0.7674617847319295, + "flos": 22059186677280.0, + "grad_norm": 1.9000626503503373, + "learning_rate": 5.407836392441319e-07, + "loss": 0.8396, + "num_input_tokens_seen": 276482650, + "step": 8510 + }, + { + "epoch": 0.7675519682553997, + "flos": 27815532564000.0, + "grad_norm": 1.6787668233271356, + "learning_rate": 5.403841800614049e-07, + "loss": 0.7612, + "num_input_tokens_seen": 276513440, + "step": 8511 + }, + { + "epoch": 0.76764215177887, + "flos": 22206699219360.0, + "grad_norm": 2.8942381675697795, + "learning_rate": 5.39984845421148e-07, + "loss": 0.7005, + "num_input_tokens_seen": 276545465, + "step": 8512 + }, + { + "epoch": 0.7677323353023403, + "flos": 25918981940640.0, + "grad_norm": 2.609806322095354, + "learning_rate": 5.395856353574344e-07, + "loss": 0.7288, + "num_input_tokens_seen": 276574460, + "step": 8513 + }, + { + "epoch": 0.7678225188258105, + "flos": 70055456968320.0, + "grad_norm": 0.5835035507172566, + "learning_rate": 5.391865499043275e-07, + "loss": 0.566, + "num_input_tokens_seen": 276677965, + "step": 8514 + }, + { + "epoch": 0.7679127023492808, + "flos": 24245707297440.0, + "grad_norm": 1.888214718739943, + "learning_rate": 5.387875890958788e-07, + "loss": 0.7483, + "num_input_tokens_seen": 276706830, + "step": 8515 + }, + { + "epoch": 0.768002885872751, + "flos": 25156769945760.0, + "grad_norm": 1.66704896674868, + "learning_rate": 5.383887529661298e-07, + "loss": 0.7688, + "num_input_tokens_seen": 276736450, + "step": 8516 + }, + { + "epoch": 0.7680930693962214, + "flos": 43335271954560.0, + "grad_norm": 2.6596760813598275, + "learning_rate": 5.379900415491116e-07, + "loss": 0.6981, + "num_input_tokens_seen": 276766860, + "step": 8517 + }, + { + "epoch": 0.7681832529196916, + "flos": 26722272253920.0, + "grad_norm": 2.0135387890974923, + "learning_rate": 5.375914548788447e-07, + "loss": 0.6958, + "num_input_tokens_seen": 276795240, + "step": 8518 + }, + { + "epoch": 0.7682734364431618, + "flos": 20635881635520.0, + "grad_norm": 2.1136664922586124, + "learning_rate": 5.371929929893384e-07, + "loss": 0.6266, + "num_input_tokens_seen": 276822565, + "step": 8519 + }, + { + "epoch": 0.768363619966632, + "flos": 38453779636320.0, + "grad_norm": 2.0384278460737044, + "learning_rate": 5.367946559145917e-07, + "loss": 0.7077, + "num_input_tokens_seen": 276852600, + "step": 8520 + }, + { + "epoch": 0.7684538034901024, + "flos": 23590041035520.0, + "grad_norm": 2.109807740080541, + "learning_rate": 5.363964436885935e-07, + "loss": 0.7616, + "num_input_tokens_seen": 276883285, + "step": 8521 + }, + { + "epoch": 0.7685439870135726, + "flos": 21658322085600.0, + "grad_norm": 1.5292317850382315, + "learning_rate": 5.359983563453199e-07, + "loss": 0.8264, + "num_input_tokens_seen": 276914830, + "step": 8522 + }, + { + "epoch": 0.7686341705370429, + "flos": 19618198914720.0, + "grad_norm": 2.074675831119102, + "learning_rate": 5.356003939187402e-07, + "loss": 0.8587, + "num_input_tokens_seen": 276943755, + "step": 8523 + }, + { + "epoch": 0.7687243540605131, + "flos": 27013766210880.0, + "grad_norm": 2.093557092690334, + "learning_rate": 5.352025564428082e-07, + "loss": 0.7215, + "num_input_tokens_seen": 276971420, + "step": 8524 + }, + { + "epoch": 0.7688145375839834, + "flos": 26028092387040.0, + "grad_norm": 1.9561939389907048, + "learning_rate": 5.348048439514723e-07, + "loss": 0.7821, + "num_input_tokens_seen": 277002200, + "step": 8525 + }, + { + "epoch": 0.7689047211074537, + "flos": 26826885159360.0, + "grad_norm": 2.2274966773148575, + "learning_rate": 5.344072564786653e-07, + "loss": 0.6233, + "num_input_tokens_seen": 277031790, + "step": 8526 + }, + { + "epoch": 0.7689949046309239, + "flos": 25445513340480.0, + "grad_norm": 1.6839604166548066, + "learning_rate": 5.340097940583123e-07, + "loss": 0.7563, + "num_input_tokens_seen": 277062610, + "step": 8527 + }, + { + "epoch": 0.7690850881543942, + "flos": 17868826305600.0, + "grad_norm": 2.8553780725278313, + "learning_rate": 5.336124567243275e-07, + "loss": 0.7063, + "num_input_tokens_seen": 277086645, + "step": 8528 + }, + { + "epoch": 0.7691752716778645, + "flos": 18416460044160.0, + "grad_norm": 2.230100156989609, + "learning_rate": 5.33215244510613e-07, + "loss": 0.7494, + "num_input_tokens_seen": 277113745, + "step": 8529 + }, + { + "epoch": 0.7692654552013347, + "flos": 14444506414080.0, + "grad_norm": 2.1239542912595515, + "learning_rate": 5.328181574510624e-07, + "loss": 0.7382, + "num_input_tokens_seen": 277139245, + "step": 8530 + }, + { + "epoch": 0.769355638724805, + "flos": 27519503896320.0, + "grad_norm": 2.339410553890561, + "learning_rate": 5.324211955795559e-07, + "loss": 0.7984, + "num_input_tokens_seen": 277167410, + "step": 8531 + }, + { + "epoch": 0.7694458222482753, + "flos": 23079396941760.0, + "grad_norm": 1.8833207833313623, + "learning_rate": 5.320243589299651e-07, + "loss": 0.6704, + "num_input_tokens_seen": 277196295, + "step": 8532 + }, + { + "epoch": 0.7695360057717455, + "flos": 23516842310880.0, + "grad_norm": 2.2597451567650166, + "learning_rate": 5.316276475361505e-07, + "loss": 0.6965, + "num_input_tokens_seen": 277224380, + "step": 8533 + }, + { + "epoch": 0.7696261892952158, + "flos": 20856741581280.0, + "grad_norm": 2.664183583122848, + "learning_rate": 5.312310614319613e-07, + "loss": 0.7811, + "num_input_tokens_seen": 277249475, + "step": 8534 + }, + { + "epoch": 0.769716372818686, + "flos": 19071457250400.0, + "grad_norm": 2.322853315598588, + "learning_rate": 5.308346006512367e-07, + "loss": 0.7134, + "num_input_tokens_seen": 277272540, + "step": 8535 + }, + { + "epoch": 0.7698065563421563, + "flos": 23476024180800.0, + "grad_norm": 1.8753431006275654, + "learning_rate": 5.30438265227805e-07, + "loss": 0.6752, + "num_input_tokens_seen": 277302260, + "step": 8536 + }, + { + "epoch": 0.7698967398656266, + "flos": 70555544850240.0, + "grad_norm": 0.602712399174197, + "learning_rate": 5.300420551954837e-07, + "loss": 0.559, + "num_input_tokens_seen": 277405985, + "step": 8537 + }, + { + "epoch": 0.7699869233890968, + "flos": 30001755826080.0, + "grad_norm": 1.5938132778396346, + "learning_rate": 5.296459705880798e-07, + "loss": 0.6689, + "num_input_tokens_seen": 277436180, + "step": 8538 + }, + { + "epoch": 0.770077106912567, + "flos": 23698779784320.0, + "grad_norm": 1.6793372920253284, + "learning_rate": 5.292500114393881e-07, + "loss": 0.7983, + "num_input_tokens_seen": 277466625, + "step": 8539 + }, + { + "epoch": 0.7701672904360374, + "flos": 20418961684320.0, + "grad_norm": 1.901392955634808, + "learning_rate": 5.288541777831963e-07, + "loss": 0.7638, + "num_input_tokens_seen": 277492220, + "step": 8540 + }, + { + "epoch": 0.7702574739595076, + "flos": 25119557282400.0, + "grad_norm": 1.7845398487698543, + "learning_rate": 5.284584696532772e-07, + "loss": 0.7981, + "num_input_tokens_seen": 277524095, + "step": 8541 + }, + { + "epoch": 0.7703476574829778, + "flos": 18524232379200.0, + "grad_norm": 1.9536663068245894, + "learning_rate": 5.280628870833954e-07, + "loss": 0.7503, + "num_input_tokens_seen": 277552445, + "step": 8542 + }, + { + "epoch": 0.7704378410064481, + "flos": 22350606294720.0, + "grad_norm": 1.944356537905685, + "learning_rate": 5.276674301073045e-07, + "loss": 0.7739, + "num_input_tokens_seen": 277580825, + "step": 8543 + }, + { + "epoch": 0.7705280245299184, + "flos": 21618804897120.0, + "grad_norm": 1.530021641896091, + "learning_rate": 5.272720987587467e-07, + "loss": 0.7576, + "num_input_tokens_seen": 277610405, + "step": 8544 + }, + { + "epoch": 0.7706182080533887, + "flos": 22023200616000.0, + "grad_norm": 2.100565961231142, + "learning_rate": 5.268768930714545e-07, + "loss": 0.6956, + "num_input_tokens_seen": 277637990, + "step": 8545 + }, + { + "epoch": 0.7707083915768589, + "flos": 25957644224640.0, + "grad_norm": 3.3647179985063618, + "learning_rate": 5.264818130791473e-07, + "loss": 0.7732, + "num_input_tokens_seen": 277668610, + "step": 8546 + }, + { + "epoch": 0.7707985751003291, + "flos": 16703928400800.0, + "grad_norm": 2.2662779820729866, + "learning_rate": 5.260868588155378e-07, + "loss": 0.7467, + "num_input_tokens_seen": 277695220, + "step": 8547 + }, + { + "epoch": 0.7708887586237995, + "flos": 23443717925760.0, + "grad_norm": 1.8686918319784116, + "learning_rate": 5.256920303143242e-07, + "loss": 0.7727, + "num_input_tokens_seen": 277725340, + "step": 8548 + }, + { + "epoch": 0.7709789421472697, + "flos": 22820878295520.0, + "grad_norm": 1.9386895742536565, + "learning_rate": 5.252973276091956e-07, + "loss": 0.8571, + "num_input_tokens_seen": 277755680, + "step": 8549 + }, + { + "epoch": 0.7710691256707399, + "flos": 21801783123840.0, + "grad_norm": 1.8705152282217323, + "learning_rate": 5.249027507338307e-07, + "loss": 0.8134, + "num_input_tokens_seen": 277784630, + "step": 8550 + }, + { + "epoch": 0.7711593091942102, + "flos": 22060376109600.0, + "grad_norm": 1.61599090092524, + "learning_rate": 5.245082997218966e-07, + "loss": 0.7164, + "num_input_tokens_seen": 277814835, + "step": 8551 + }, + { + "epoch": 0.7712494927176805, + "flos": 23475615313440.0, + "grad_norm": 2.9248522924965568, + "learning_rate": 5.241139746070499e-07, + "loss": 0.7677, + "num_input_tokens_seen": 277842840, + "step": 8552 + }, + { + "epoch": 0.7713396762411507, + "flos": 20638706537280.0, + "grad_norm": 1.72620633279317, + "learning_rate": 5.237197754229376e-07, + "loss": 0.6269, + "num_input_tokens_seen": 277872020, + "step": 8553 + }, + { + "epoch": 0.771429859764621, + "flos": 35937437303040.0, + "grad_norm": 2.1206166554541745, + "learning_rate": 5.233257022031931e-07, + "loss": 0.6541, + "num_input_tokens_seen": 277906915, + "step": 8554 + }, + { + "epoch": 0.7715200432880913, + "flos": 19472544860640.0, + "grad_norm": 1.6936808526257676, + "learning_rate": 5.229317549814432e-07, + "loss": 0.755, + "num_input_tokens_seen": 277933660, + "step": 8555 + }, + { + "epoch": 0.7716102268115616, + "flos": 20164457372160.0, + "grad_norm": 2.091070139755293, + "learning_rate": 5.225379337912998e-07, + "loss": 0.826, + "num_input_tokens_seen": 277960325, + "step": 8556 + }, + { + "epoch": 0.7717004103350318, + "flos": 23516544952800.0, + "grad_norm": 19.812244305329532, + "learning_rate": 5.221442386663663e-07, + "loss": 0.7606, + "num_input_tokens_seen": 277988860, + "step": 8557 + }, + { + "epoch": 0.771790593858502, + "flos": 26467805111520.0, + "grad_norm": 1.989347820090317, + "learning_rate": 5.217506696402354e-07, + "loss": 0.7832, + "num_input_tokens_seen": 278018745, + "step": 8558 + }, + { + "epoch": 0.7718807773819724, + "flos": 20711124696960.0, + "grad_norm": 2.7122492625991224, + "learning_rate": 5.213572267464883e-07, + "loss": 0.7589, + "num_input_tokens_seen": 278046065, + "step": 8559 + }, + { + "epoch": 0.7719709609054426, + "flos": 16700620292160.0, + "grad_norm": 1.9205783718915286, + "learning_rate": 5.209639100186965e-07, + "loss": 0.7252, + "num_input_tokens_seen": 278070655, + "step": 8560 + }, + { + "epoch": 0.7720611444289128, + "flos": 67822208226240.0, + "grad_norm": 0.6630264502739643, + "learning_rate": 5.205707194904179e-07, + "loss": 0.6401, + "num_input_tokens_seen": 278159715, + "step": 8561 + }, + { + "epoch": 0.7721513279523831, + "flos": 26026717105920.0, + "grad_norm": 2.0289396792225807, + "learning_rate": 5.201776551952042e-07, + "loss": 0.6707, + "num_input_tokens_seen": 278187320, + "step": 8562 + }, + { + "epoch": 0.7722415114758534, + "flos": 30395483823840.0, + "grad_norm": 3.111289043863545, + "learning_rate": 5.197847171665914e-07, + "loss": 0.7082, + "num_input_tokens_seen": 278213395, + "step": 8563 + }, + { + "epoch": 0.7723316949993236, + "flos": 20964253728000.0, + "grad_norm": 1.4929872935853719, + "learning_rate": 5.193919054381095e-07, + "loss": 0.7305, + "num_input_tokens_seen": 278241300, + "step": 8564 + }, + { + "epoch": 0.7724218785227939, + "flos": 20524949870880.0, + "grad_norm": 1.9678056056419977, + "learning_rate": 5.189992200432738e-07, + "loss": 0.7091, + "num_input_tokens_seen": 278269365, + "step": 8565 + }, + { + "epoch": 0.7725120620462641, + "flos": 15640595481120.0, + "grad_norm": 2.889550746595546, + "learning_rate": 5.186066610155906e-07, + "loss": 0.7583, + "num_input_tokens_seen": 278294735, + "step": 8566 + }, + { + "epoch": 0.7726022455697344, + "flos": 23297655004320.0, + "grad_norm": 2.267059608601286, + "learning_rate": 5.182142283885555e-07, + "loss": 0.7992, + "num_input_tokens_seen": 278322490, + "step": 8567 + }, + { + "epoch": 0.7726924290932047, + "flos": 62982983658720.0, + "grad_norm": 0.6539515936476762, + "learning_rate": 5.178219221956528e-07, + "loss": 0.5303, + "num_input_tokens_seen": 278404685, + "step": 8568 + }, + { + "epoch": 0.7727826126166749, + "flos": 21076040397120.0, + "grad_norm": 2.250568635929168, + "learning_rate": 5.174297424703565e-07, + "loss": 0.7648, + "num_input_tokens_seen": 278433195, + "step": 8569 + }, + { + "epoch": 0.7728727961401451, + "flos": 26828706477600.0, + "grad_norm": 1.8320939758884685, + "learning_rate": 5.170376892461299e-07, + "loss": 0.6782, + "num_input_tokens_seen": 278463925, + "step": 8570 + }, + { + "epoch": 0.7729629796636155, + "flos": 66574521798720.0, + "grad_norm": 0.969200121995601, + "learning_rate": 5.16645762556424e-07, + "loss": 0.6302, + "num_input_tokens_seen": 278549750, + "step": 8571 + }, + { + "epoch": 0.7730531631870857, + "flos": 41803153824480.0, + "grad_norm": 1.9612507968801287, + "learning_rate": 5.162539624346809e-07, + "loss": 0.7923, + "num_input_tokens_seen": 278581280, + "step": 8572 + }, + { + "epoch": 0.773143346710556, + "flos": 23914733321760.0, + "grad_norm": 1.9447551986089353, + "learning_rate": 5.158622889143309e-07, + "loss": 0.8118, + "num_input_tokens_seen": 278608115, + "step": 8573 + }, + { + "epoch": 0.7732335302340262, + "flos": 23480038514880.0, + "grad_norm": 2.4739408365266695, + "learning_rate": 5.154707420287939e-07, + "loss": 0.7047, + "num_input_tokens_seen": 278635350, + "step": 8574 + }, + { + "epoch": 0.7733237137574965, + "flos": 27377009271840.0, + "grad_norm": 1.8669595619158057, + "learning_rate": 5.150793218114793e-07, + "loss": 0.6726, + "num_input_tokens_seen": 278665945, + "step": 8575 + }, + { + "epoch": 0.7734138972809668, + "flos": 23990348080800.0, + "grad_norm": 1.7764373144701227, + "learning_rate": 5.146880282957837e-07, + "loss": 0.7323, + "num_input_tokens_seen": 278693670, + "step": 8576 + }, + { + "epoch": 0.773504080804437, + "flos": 25265434355040.0, + "grad_norm": 2.220378117192404, + "learning_rate": 5.142968615150964e-07, + "loss": 0.7324, + "num_input_tokens_seen": 278720695, + "step": 8577 + }, + { + "epoch": 0.7735942643279073, + "flos": 29017159925280.0, + "grad_norm": 2.0430567414278653, + "learning_rate": 5.139058215027921e-07, + "loss": 0.6753, + "num_input_tokens_seen": 278750695, + "step": 8578 + }, + { + "epoch": 0.7736844478513776, + "flos": 26756771524800.0, + "grad_norm": 1.4951005233344272, + "learning_rate": 5.135149082922383e-07, + "loss": 0.7335, + "num_input_tokens_seen": 278780800, + "step": 8579 + }, + { + "epoch": 0.7737746313748478, + "flos": 33746976688320.0, + "grad_norm": 1.9706841913075537, + "learning_rate": 5.131241219167879e-07, + "loss": 0.6728, + "num_input_tokens_seen": 278809940, + "step": 8580 + }, + { + "epoch": 0.773864814898318, + "flos": 59032928750880.0, + "grad_norm": 0.7776721279921592, + "learning_rate": 5.127334624097869e-07, + "loss": 0.6214, + "num_input_tokens_seen": 278897945, + "step": 8581 + }, + { + "epoch": 0.7739549984217884, + "flos": 31563950025600.0, + "grad_norm": 1.550655715435212, + "learning_rate": 5.123429298045672e-07, + "loss": 0.7994, + "num_input_tokens_seen": 278927730, + "step": 8582 + }, + { + "epoch": 0.7740451819452586, + "flos": 35794310792640.0, + "grad_norm": 2.072029409509153, + "learning_rate": 5.119525241344515e-07, + "loss": 0.6341, + "num_input_tokens_seen": 278958445, + "step": 8583 + }, + { + "epoch": 0.7741353654687289, + "flos": 25010707024320.0, + "grad_norm": 5.430220543473058, + "learning_rate": 5.115622454327515e-07, + "loss": 0.7539, + "num_input_tokens_seen": 278985950, + "step": 8584 + }, + { + "epoch": 0.7742255489921991, + "flos": 21253443159840.0, + "grad_norm": 4.132974718800626, + "learning_rate": 5.11172093732768e-07, + "loss": 0.7306, + "num_input_tokens_seen": 279011560, + "step": 8585 + }, + { + "epoch": 0.7743157325156694, + "flos": 23953284096480.0, + "grad_norm": 2.28540225538812, + "learning_rate": 5.107820690677911e-07, + "loss": 0.7722, + "num_input_tokens_seen": 279041560, + "step": 8586 + }, + { + "epoch": 0.7744059160391397, + "flos": 20747408116320.0, + "grad_norm": 1.842401884512656, + "learning_rate": 5.103921714710991e-07, + "loss": 0.6928, + "num_input_tokens_seen": 279069945, + "step": 8587 + }, + { + "epoch": 0.7744960995626099, + "flos": 19728052756320.0, + "grad_norm": 1.8679433912648344, + "learning_rate": 5.100024009759605e-07, + "loss": 0.7601, + "num_input_tokens_seen": 279097330, + "step": 8588 + }, + { + "epoch": 0.7745862830860801, + "flos": 16666790076960.0, + "grad_norm": 2.7410609873931824, + "learning_rate": 5.09612757615633e-07, + "loss": 0.7843, + "num_input_tokens_seen": 279123110, + "step": 8589 + }, + { + "epoch": 0.7746764666095505, + "flos": 21768138757440.0, + "grad_norm": 1.6504687048144535, + "learning_rate": 5.092232414233628e-07, + "loss": 0.6987, + "num_input_tokens_seen": 279151425, + "step": 8590 + }, + { + "epoch": 0.7747666501330207, + "flos": 17102265448800.0, + "grad_norm": 1.9489566114517618, + "learning_rate": 5.088338524323858e-07, + "loss": 0.7111, + "num_input_tokens_seen": 279175780, + "step": 8591 + }, + { + "epoch": 0.7748568336564909, + "flos": 19983226124160.0, + "grad_norm": 2.9668640571104743, + "learning_rate": 5.084445906759271e-07, + "loss": 0.8017, + "num_input_tokens_seen": 279203790, + "step": 8592 + }, + { + "epoch": 0.7749470171799612, + "flos": 20053674286560.0, + "grad_norm": 1.7978376380758303, + "learning_rate": 5.080554561871995e-07, + "loss": 0.8077, + "num_input_tokens_seen": 279231540, + "step": 8593 + }, + { + "epoch": 0.7750372007034315, + "flos": 22678569519840.0, + "grad_norm": 1.68927029188162, + "learning_rate": 5.076664489994078e-07, + "loss": 0.7725, + "num_input_tokens_seen": 279260830, + "step": 8594 + }, + { + "epoch": 0.7751273842269017, + "flos": 21105075713280.0, + "grad_norm": 2.1348693766602236, + "learning_rate": 5.07277569145742e-07, + "loss": 0.7135, + "num_input_tokens_seen": 279287180, + "step": 8595 + }, + { + "epoch": 0.775217567750372, + "flos": 20601977080800.0, + "grad_norm": 2.979737838375923, + "learning_rate": 5.068888166593861e-07, + "loss": 0.6599, + "num_input_tokens_seen": 279314905, + "step": 8596 + }, + { + "epoch": 0.7753077512738422, + "flos": 33717644014080.0, + "grad_norm": 2.0125114582754406, + "learning_rate": 5.065001915735087e-07, + "loss": 0.7319, + "num_input_tokens_seen": 279343385, + "step": 8597 + }, + { + "epoch": 0.7753979347973126, + "flos": 22933668548160.0, + "grad_norm": 2.1126126908260323, + "learning_rate": 5.061116939212702e-07, + "loss": 0.7122, + "num_input_tokens_seen": 279371170, + "step": 8598 + }, + { + "epoch": 0.7754881183207828, + "flos": 29855655734880.0, + "grad_norm": 2.2402415055125733, + "learning_rate": 5.05723323735819e-07, + "loss": 0.7434, + "num_input_tokens_seen": 279399370, + "step": 8599 + }, + { + "epoch": 0.775578301844253, + "flos": 61812919157280.0, + "grad_norm": 0.6272170560037615, + "learning_rate": 5.053350810502932e-07, + "loss": 0.5823, + "num_input_tokens_seen": 279487675, + "step": 8600 + }, + { + "epoch": 0.7756684853677234, + "flos": 22897868335680.0, + "grad_norm": 1.7114082281107683, + "learning_rate": 5.049469658978202e-07, + "loss": 0.7398, + "num_input_tokens_seen": 279517695, + "step": 8601 + }, + { + "epoch": 0.7757586688911936, + "flos": 68811636195840.0, + "grad_norm": 0.6280614023608057, + "learning_rate": 5.045589783115147e-07, + "loss": 0.5313, + "num_input_tokens_seen": 279617875, + "step": 8602 + }, + { + "epoch": 0.7758488524146638, + "flos": 20128173952800.0, + "grad_norm": 1.9808518981491583, + "learning_rate": 5.041711183244842e-07, + "loss": 0.7819, + "num_input_tokens_seen": 279644335, + "step": 8603 + }, + { + "epoch": 0.7759390359381341, + "flos": 21184258769280.0, + "grad_norm": 2.641199805545063, + "learning_rate": 5.037833859698211e-07, + "loss": 0.7636, + "num_input_tokens_seen": 279671670, + "step": 8604 + }, + { + "epoch": 0.7760292194616044, + "flos": 27087187954080.0, + "grad_norm": 2.2601132401647313, + "learning_rate": 5.033957812806096e-07, + "loss": 0.8107, + "num_input_tokens_seen": 279700390, + "step": 8605 + }, + { + "epoch": 0.7761194029850746, + "flos": 26685765816000.0, + "grad_norm": 2.2927622401549557, + "learning_rate": 5.030083042899223e-07, + "loss": 0.7956, + "num_input_tokens_seen": 279726670, + "step": 8606 + }, + { + "epoch": 0.7762095865085449, + "flos": 24025962444480.0, + "grad_norm": 2.3576219121452535, + "learning_rate": 5.026209550308207e-07, + "loss": 0.743, + "num_input_tokens_seen": 279754530, + "step": 8607 + }, + { + "epoch": 0.7762997700320151, + "flos": 23626361624640.0, + "grad_norm": 2.1159219372849836, + "learning_rate": 5.022337335363558e-07, + "loss": 0.8001, + "num_input_tokens_seen": 279783600, + "step": 8608 + }, + { + "epoch": 0.7763899535554855, + "flos": 22637342522400.0, + "grad_norm": 1.8249028820168307, + "learning_rate": 5.018466398395677e-07, + "loss": 0.7696, + "num_input_tokens_seen": 279809775, + "step": 8609 + }, + { + "epoch": 0.7764801370789557, + "flos": 25156435417920.0, + "grad_norm": 3.928138746239062, + "learning_rate": 5.01459673973484e-07, + "loss": 0.6774, + "num_input_tokens_seen": 279841130, + "step": 8610 + }, + { + "epoch": 0.7765703206024259, + "flos": 20237730436320.0, + "grad_norm": 2.417264919113846, + "learning_rate": 5.01072835971125e-07, + "loss": 0.8076, + "num_input_tokens_seen": 279866525, + "step": 8611 + }, + { + "epoch": 0.7766605041258962, + "flos": 30507047474400.0, + "grad_norm": 1.7666104697640288, + "learning_rate": 5.006861258654959e-07, + "loss": 0.6499, + "num_input_tokens_seen": 279898530, + "step": 8612 + }, + { + "epoch": 0.7767506876493665, + "flos": 20237841945600.0, + "grad_norm": 5.421812557333603, + "learning_rate": 5.002995436895938e-07, + "loss": 0.7729, + "num_input_tokens_seen": 279927975, + "step": 8613 + }, + { + "epoch": 0.7768408711728367, + "flos": 23625841248000.0, + "grad_norm": 1.8039636522678646, + "learning_rate": 4.999130894764039e-07, + "loss": 0.7023, + "num_input_tokens_seen": 279956700, + "step": 8614 + }, + { + "epoch": 0.776931054696307, + "flos": 23370593540640.0, + "grad_norm": 1.8036452417228566, + "learning_rate": 4.995267632589006e-07, + "loss": 0.6855, + "num_input_tokens_seen": 279985010, + "step": 8615 + }, + { + "epoch": 0.7770212382197772, + "flos": 59391116724480.0, + "grad_norm": 0.6843413706141976, + "learning_rate": 4.99140565070048e-07, + "loss": 0.5802, + "num_input_tokens_seen": 280074430, + "step": 8616 + }, + { + "epoch": 0.7771114217432475, + "flos": 18744014401920.0, + "grad_norm": 1.7840809995754858, + "learning_rate": 4.987544949427969e-07, + "loss": 0.7654, + "num_input_tokens_seen": 280101975, + "step": 8617 + }, + { + "epoch": 0.7772016052667178, + "flos": 67441978584960.0, + "grad_norm": 0.6649810048667348, + "learning_rate": 4.98368552910091e-07, + "loss": 0.5871, + "num_input_tokens_seen": 280190955, + "step": 8618 + }, + { + "epoch": 0.777291788790188, + "flos": 22782922236960.0, + "grad_norm": 2.2646255919705274, + "learning_rate": 4.979827390048596e-07, + "loss": 0.8588, + "num_input_tokens_seen": 280216875, + "step": 8619 + }, + { + "epoch": 0.7773819723136582, + "flos": 20055421265280.0, + "grad_norm": 1.5007039643483453, + "learning_rate": 4.975970532600231e-07, + "loss": 0.7085, + "num_input_tokens_seen": 280246535, + "step": 8620 + }, + { + "epoch": 0.7774721558371286, + "flos": 17797374559680.0, + "grad_norm": 1.992598547323238, + "learning_rate": 4.972114957084901e-07, + "loss": 0.7658, + "num_input_tokens_seen": 280274105, + "step": 8621 + }, + { + "epoch": 0.7775623393605988, + "flos": 22932516285600.0, + "grad_norm": 1.8580057568209418, + "learning_rate": 4.968260663831585e-07, + "loss": 0.7245, + "num_input_tokens_seen": 280304245, + "step": 8622 + }, + { + "epoch": 0.777652522884069, + "flos": 23515652878560.0, + "grad_norm": 1.6255825905729318, + "learning_rate": 4.964407653169154e-07, + "loss": 0.8549, + "num_input_tokens_seen": 280334060, + "step": 8623 + }, + { + "epoch": 0.7777427064075394, + "flos": 23917446714240.0, + "grad_norm": 1.5747226258917737, + "learning_rate": 4.960555925426366e-07, + "loss": 0.7124, + "num_input_tokens_seen": 280362475, + "step": 8624 + }, + { + "epoch": 0.7778328899310096, + "flos": 15137125151040.0, + "grad_norm": 2.1857473709280866, + "learning_rate": 4.956705480931876e-07, + "loss": 0.8058, + "num_input_tokens_seen": 280387625, + "step": 8625 + }, + { + "epoch": 0.7779230734544799, + "flos": 25953221023200.0, + "grad_norm": 1.5467354297287714, + "learning_rate": 4.952856320014225e-07, + "loss": 0.7246, + "num_input_tokens_seen": 280417930, + "step": 8626 + }, + { + "epoch": 0.7780132569779501, + "flos": 38379800346720.0, + "grad_norm": 1.5232525835539663, + "learning_rate": 4.949008443001838e-07, + "loss": 0.6803, + "num_input_tokens_seen": 280453610, + "step": 8627 + }, + { + "epoch": 0.7781034405014204, + "flos": 19834635659040.0, + "grad_norm": 2.052788034131564, + "learning_rate": 4.945161850223041e-07, + "loss": 0.758, + "num_input_tokens_seen": 280482380, + "step": 8628 + }, + { + "epoch": 0.7781936240248907, + "flos": 32988147141600.0, + "grad_norm": 2.848556815707212, + "learning_rate": 4.941316542006044e-07, + "loss": 0.6774, + "num_input_tokens_seen": 280510035, + "step": 8629 + }, + { + "epoch": 0.7782838075483609, + "flos": 18927215647200.0, + "grad_norm": 1.56363705270623, + "learning_rate": 4.937472518678956e-07, + "loss": 0.7697, + "num_input_tokens_seen": 280538695, + "step": 8630 + }, + { + "epoch": 0.7783739910718311, + "flos": 19685859345120.0, + "grad_norm": 2.072498362394452, + "learning_rate": 4.93362978056977e-07, + "loss": 0.7434, + "num_input_tokens_seen": 280566510, + "step": 8631 + }, + { + "epoch": 0.7784641745953015, + "flos": 32078831472000.0, + "grad_norm": 2.154035343124604, + "learning_rate": 4.929788328006355e-07, + "loss": 0.6957, + "num_input_tokens_seen": 280601015, + "step": 8632 + }, + { + "epoch": 0.7785543581187717, + "flos": 19618050235680.0, + "grad_norm": 2.0961399387498596, + "learning_rate": 4.925948161316506e-07, + "loss": 0.8171, + "num_input_tokens_seen": 280626440, + "step": 8633 + }, + { + "epoch": 0.778644541642242, + "flos": 23550486677280.0, + "grad_norm": 1.9329369248204848, + "learning_rate": 4.922109280827868e-07, + "loss": 0.7326, + "num_input_tokens_seen": 280655345, + "step": 8634 + }, + { + "epoch": 0.7787347251657122, + "flos": 21950076230880.0, + "grad_norm": 1.9026603670240947, + "learning_rate": 4.918271686868016e-07, + "loss": 0.7377, + "num_input_tokens_seen": 280683810, + "step": 8635 + }, + { + "epoch": 0.7788249086891825, + "flos": 20638260500160.0, + "grad_norm": 8.738558018359457, + "learning_rate": 4.914435379764379e-07, + "loss": 0.7871, + "num_input_tokens_seen": 280712225, + "step": 8636 + }, + { + "epoch": 0.7789150922126528, + "flos": 27157933474560.0, + "grad_norm": 1.9985432684067588, + "learning_rate": 4.910600359844294e-07, + "loss": 0.7361, + "num_input_tokens_seen": 280742455, + "step": 8637 + }, + { + "epoch": 0.779005275736123, + "flos": 24974572284000.0, + "grad_norm": 3.1093368144542035, + "learning_rate": 4.90676662743499e-07, + "loss": 0.6533, + "num_input_tokens_seen": 280769840, + "step": 8638 + }, + { + "epoch": 0.7790954592595932, + "flos": 22314917591520.0, + "grad_norm": 2.4937453229281896, + "learning_rate": 4.902934182863581e-07, + "loss": 0.7385, + "num_input_tokens_seen": 280797410, + "step": 8639 + }, + { + "epoch": 0.7791856427830636, + "flos": 23224567788960.0, + "grad_norm": 2.0349696214737567, + "learning_rate": 4.899103026457069e-07, + "loss": 0.6159, + "num_input_tokens_seen": 280826725, + "step": 8640 + }, + { + "epoch": 0.7792758263065338, + "flos": 66005509714560.0, + "grad_norm": 0.5866647911403038, + "learning_rate": 4.895273158542361e-07, + "loss": 0.5295, + "num_input_tokens_seen": 280918770, + "step": 8641 + }, + { + "epoch": 0.779366009830004, + "flos": 23915551056480.0, + "grad_norm": 2.024254357846669, + "learning_rate": 4.891444579446227e-07, + "loss": 0.76, + "num_input_tokens_seen": 280948760, + "step": 8642 + }, + { + "epoch": 0.7794561933534743, + "flos": 23917112186400.0, + "grad_norm": 1.973755362215302, + "learning_rate": 4.887617289495349e-07, + "loss": 0.701, + "num_input_tokens_seen": 280977120, + "step": 8643 + }, + { + "epoch": 0.7795463768769446, + "flos": 62116976493120.0, + "grad_norm": 0.5891150716701405, + "learning_rate": 4.883791289016292e-07, + "loss": 0.5101, + "num_input_tokens_seen": 281062655, + "step": 8644 + }, + { + "epoch": 0.7796365604004148, + "flos": 19728127095840.0, + "grad_norm": 2.0011029778347664, + "learning_rate": 4.879966578335514e-07, + "loss": 0.754, + "num_input_tokens_seen": 281092455, + "step": 8645 + }, + { + "epoch": 0.7797267439238851, + "flos": 23730416983680.0, + "grad_norm": 3.1167746057705665, + "learning_rate": 4.876143157779358e-07, + "loss": 0.6411, + "num_input_tokens_seen": 281120155, + "step": 8646 + }, + { + "epoch": 0.7798169274473554, + "flos": 34767632989920.0, + "grad_norm": 2.3820474365725355, + "learning_rate": 4.872321027674058e-07, + "loss": 0.631, + "num_input_tokens_seen": 281152285, + "step": 8647 + }, + { + "epoch": 0.7799071109708257, + "flos": 23625989927040.0, + "grad_norm": 1.7085764719678582, + "learning_rate": 4.868500188345748e-07, + "loss": 0.8148, + "num_input_tokens_seen": 281178110, + "step": 8648 + }, + { + "epoch": 0.7799972944942959, + "flos": 21658582273920.0, + "grad_norm": 2.007624635924747, + "learning_rate": 4.864680640120425e-07, + "loss": 0.8058, + "num_input_tokens_seen": 281207300, + "step": 8649 + }, + { + "epoch": 0.7800874780177661, + "flos": 14772469639200.0, + "grad_norm": 2.6604345234924094, + "learning_rate": 4.860862383324016e-07, + "loss": 0.7512, + "num_input_tokens_seen": 281231410, + "step": 8650 + }, + { + "epoch": 0.7801776615412365, + "flos": 22497078083520.0, + "grad_norm": 1.922851810484367, + "learning_rate": 4.857045418282295e-07, + "loss": 0.7402, + "num_input_tokens_seen": 281261000, + "step": 8651 + }, + { + "epoch": 0.7802678450647067, + "flos": 23881311973920.0, + "grad_norm": 5.713553981251922, + "learning_rate": 4.853229745320966e-07, + "loss": 0.6957, + "num_input_tokens_seen": 281288090, + "step": 8652 + }, + { + "epoch": 0.7803580285881769, + "flos": 23916889167840.0, + "grad_norm": 1.7400185122546517, + "learning_rate": 4.849415364765587e-07, + "loss": 0.728, + "num_input_tokens_seen": 281316800, + "step": 8653 + }, + { + "epoch": 0.7804482121116472, + "flos": 28507556584800.0, + "grad_norm": 2.049490502474375, + "learning_rate": 4.845602276941631e-07, + "loss": 0.808, + "num_input_tokens_seen": 281347495, + "step": 8654 + }, + { + "epoch": 0.7805383956351175, + "flos": 23552568183840.0, + "grad_norm": 1.8326481347884551, + "learning_rate": 4.841790482174449e-07, + "loss": 0.7806, + "num_input_tokens_seen": 281377695, + "step": 8655 + }, + { + "epoch": 0.7806285791585877, + "flos": 30836237301600.0, + "grad_norm": 2.0969953294955785, + "learning_rate": 4.837979980789282e-07, + "loss": 0.7334, + "num_input_tokens_seen": 281407970, + "step": 8656 + }, + { + "epoch": 0.780718762682058, + "flos": 29710856585280.0, + "grad_norm": 1.8436334217522419, + "learning_rate": 4.834170773111273e-07, + "loss": 0.6857, + "num_input_tokens_seen": 281437395, + "step": 8657 + }, + { + "epoch": 0.7808089462055282, + "flos": 20669005625280.0, + "grad_norm": 1.8519194053761598, + "learning_rate": 4.830362859465431e-07, + "loss": 0.7648, + "num_input_tokens_seen": 281465520, + "step": 8658 + }, + { + "epoch": 0.7808991297289986, + "flos": 22788237512640.0, + "grad_norm": 1.9521846491177772, + "learning_rate": 4.826556240176675e-07, + "loss": 0.8043, + "num_input_tokens_seen": 281495405, + "step": 8659 + }, + { + "epoch": 0.7809893132524688, + "flos": 20930274833760.0, + "grad_norm": 2.3514547228273566, + "learning_rate": 4.822750915569807e-07, + "loss": 0.7568, + "num_input_tokens_seen": 281521570, + "step": 8660 + }, + { + "epoch": 0.781079496775939, + "flos": 16449052391040.0, + "grad_norm": 2.082887398359692, + "learning_rate": 4.818946885969514e-07, + "loss": 0.82, + "num_input_tokens_seen": 281547925, + "step": 8661 + }, + { + "epoch": 0.7811696802994093, + "flos": 21586201284000.0, + "grad_norm": 2.092139920298343, + "learning_rate": 4.815144151700383e-07, + "loss": 0.7851, + "num_input_tokens_seen": 281578680, + "step": 8662 + }, + { + "epoch": 0.7812598638228796, + "flos": 20014454456160.0, + "grad_norm": 2.1822050042497176, + "learning_rate": 4.811342713086885e-07, + "loss": 0.7502, + "num_input_tokens_seen": 281606595, + "step": 8663 + }, + { + "epoch": 0.7813500473463498, + "flos": 15137496848640.0, + "grad_norm": 2.2639284292979314, + "learning_rate": 4.807542570453367e-07, + "loss": 0.7806, + "num_input_tokens_seen": 281632050, + "step": 8664 + }, + { + "epoch": 0.7814402308698201, + "flos": 21400769853120.0, + "grad_norm": 1.7536968284589132, + "learning_rate": 4.803743724124098e-07, + "loss": 0.7075, + "num_input_tokens_seen": 281662195, + "step": 8665 + }, + { + "epoch": 0.7815304143932903, + "flos": 36155100649440.0, + "grad_norm": 2.3502156909410066, + "learning_rate": 4.799946174423192e-07, + "loss": 0.749, + "num_input_tokens_seen": 281691400, + "step": 8666 + }, + { + "epoch": 0.7816205979167606, + "flos": 70991837956800.0, + "grad_norm": 0.7304067738398574, + "learning_rate": 4.796149921674706e-07, + "loss": 0.5316, + "num_input_tokens_seen": 281782290, + "step": 8667 + }, + { + "epoch": 0.7817107814402309, + "flos": 27597014313120.0, + "grad_norm": 1.7021173418523678, + "learning_rate": 4.792354966202534e-07, + "loss": 0.7018, + "num_input_tokens_seen": 281811145, + "step": 8668 + }, + { + "epoch": 0.7818009649637011, + "flos": 21184816315680.0, + "grad_norm": 1.9952993585188141, + "learning_rate": 4.788561308330489e-07, + "loss": 0.5712, + "num_input_tokens_seen": 281838625, + "step": 8669 + }, + { + "epoch": 0.7818911484871713, + "flos": 25265322845760.0, + "grad_norm": 2.6302199461022777, + "learning_rate": 4.784768948382272e-07, + "loss": 0.6874, + "num_input_tokens_seen": 281866270, + "step": 8670 + }, + { + "epoch": 0.7819813320106417, + "flos": 23407322997120.0, + "grad_norm": 1.8093681243080648, + "learning_rate": 4.780977886681461e-07, + "loss": 0.752, + "num_input_tokens_seen": 281895235, + "step": 8671 + }, + { + "epoch": 0.7820715155341119, + "flos": 25338484400640.0, + "grad_norm": 2.080016343896183, + "learning_rate": 4.777188123551541e-07, + "loss": 0.6906, + "num_input_tokens_seen": 281924650, + "step": 8672 + }, + { + "epoch": 0.7821616990575821, + "flos": 23590003865760.0, + "grad_norm": 2.0051412067489625, + "learning_rate": 4.773399659315856e-07, + "loss": 0.775, + "num_input_tokens_seen": 281953680, + "step": 8673 + }, + { + "epoch": 0.7822518825810525, + "flos": 16223174527680.0, + "grad_norm": 2.486303576567736, + "learning_rate": 4.769612494297681e-07, + "loss": 0.7269, + "num_input_tokens_seen": 281979620, + "step": 8674 + }, + { + "epoch": 0.7823420661045227, + "flos": 65842200024480.0, + "grad_norm": 0.6413882988335537, + "learning_rate": 4.765826628820142e-07, + "loss": 0.5544, + "num_input_tokens_seen": 282069110, + "step": 8675 + }, + { + "epoch": 0.782432249627993, + "flos": 24464002529760.0, + "grad_norm": 2.1684581857088516, + "learning_rate": 4.7620420632062775e-07, + "loss": 0.6577, + "num_input_tokens_seen": 282096370, + "step": 8676 + }, + { + "epoch": 0.7825224331514632, + "flos": 28508337149760.0, + "grad_norm": 1.9183934007725223, + "learning_rate": 4.758258797779002e-07, + "loss": 0.6995, + "num_input_tokens_seen": 282126860, + "step": 8677 + }, + { + "epoch": 0.7826126166749335, + "flos": 63906280891680.0, + "grad_norm": 0.6125996683733196, + "learning_rate": 4.7544768328611317e-07, + "loss": 0.5639, + "num_input_tokens_seen": 282213245, + "step": 8678 + }, + { + "epoch": 0.7827028001984038, + "flos": 24281693358720.0, + "grad_norm": 1.7515671388835203, + "learning_rate": 4.750696168775359e-07, + "loss": 0.7271, + "num_input_tokens_seen": 282245195, + "step": 8679 + }, + { + "epoch": 0.782792983721874, + "flos": 69944265015360.0, + "grad_norm": 0.5986500872908959, + "learning_rate": 4.746916805844279e-07, + "loss": 0.5509, + "num_input_tokens_seen": 282344220, + "step": 8680 + }, + { + "epoch": 0.7828831672453442, + "flos": 21439878174240.0, + "grad_norm": 2.6312234763940237, + "learning_rate": 4.743138744390356e-07, + "loss": 0.772, + "num_input_tokens_seen": 282373490, + "step": 8681 + }, + { + "epoch": 0.7829733507688146, + "flos": 22861101709440.0, + "grad_norm": 1.758359086569745, + "learning_rate": 4.739361984735959e-07, + "loss": 0.8165, + "num_input_tokens_seen": 282402020, + "step": 8682 + }, + { + "epoch": 0.7830635342922848, + "flos": 22824149234400.0, + "grad_norm": 2.0585885382627933, + "learning_rate": 4.7355865272033455e-07, + "loss": 0.7009, + "num_input_tokens_seen": 282430765, + "step": 8683 + }, + { + "epoch": 0.783153717815755, + "flos": 22060413279360.0, + "grad_norm": 1.8311141917414713, + "learning_rate": 4.7318123721146563e-07, + "loss": 0.7343, + "num_input_tokens_seen": 282459200, + "step": 8684 + }, + { + "epoch": 0.7832439013392253, + "flos": 25410493692960.0, + "grad_norm": 1.686396273185224, + "learning_rate": 4.728039519791924e-07, + "loss": 0.73, + "num_input_tokens_seen": 282488365, + "step": 8685 + }, + { + "epoch": 0.7833340848626956, + "flos": 67902469205280.0, + "grad_norm": 0.674695128892011, + "learning_rate": 4.72426797055707e-07, + "loss": 0.5982, + "num_input_tokens_seen": 282579205, + "step": 8686 + }, + { + "epoch": 0.7834242683861659, + "flos": 23623239364800.0, + "grad_norm": 1.8823411280247313, + "learning_rate": 4.720497724731904e-07, + "loss": 0.7115, + "num_input_tokens_seen": 282606205, + "step": 8687 + }, + { + "epoch": 0.7835144519096361, + "flos": 22460014099200.0, + "grad_norm": 2.4296726827384303, + "learning_rate": 4.7167287826381153e-07, + "loss": 0.7807, + "num_input_tokens_seen": 282634640, + "step": 8688 + }, + { + "epoch": 0.7836046354331063, + "flos": 28616778540480.0, + "grad_norm": 1.7025167796558616, + "learning_rate": 4.712961144597307e-07, + "loss": 0.7515, + "num_input_tokens_seen": 282666950, + "step": 8689 + }, + { + "epoch": 0.7836948189565767, + "flos": 23553794785920.0, + "grad_norm": 5.151499276298228, + "learning_rate": 4.7091948109309343e-07, + "loss": 0.7803, + "num_input_tokens_seen": 282694285, + "step": 8690 + }, + { + "epoch": 0.7837850024800469, + "flos": 18488543676000.0, + "grad_norm": 2.869417825062282, + "learning_rate": 4.705429781960384e-07, + "loss": 0.7536, + "num_input_tokens_seen": 282719660, + "step": 8691 + }, + { + "epoch": 0.7838751860035171, + "flos": 20748486039360.0, + "grad_norm": 2.4554021677620224, + "learning_rate": 4.7016660580068923e-07, + "loss": 0.7404, + "num_input_tokens_seen": 282748145, + "step": 8692 + }, + { + "epoch": 0.7839653695269874, + "flos": 26715581697120.0, + "grad_norm": 2.2738957418464896, + "learning_rate": 4.6979036393916093e-07, + "loss": 0.7768, + "num_input_tokens_seen": 282774365, + "step": 8693 + }, + { + "epoch": 0.7840555530504577, + "flos": 20524094966400.0, + "grad_norm": 1.6462528885082488, + "learning_rate": 4.6941425264355603e-07, + "loss": 0.8241, + "num_input_tokens_seen": 282801910, + "step": 8694 + }, + { + "epoch": 0.7841457365739279, + "flos": 20673428826720.0, + "grad_norm": 1.8052825607239313, + "learning_rate": 4.6903827194596666e-07, + "loss": 0.7549, + "num_input_tokens_seen": 282833155, + "step": 8695 + }, + { + "epoch": 0.7842359200973982, + "flos": 19800173557920.0, + "grad_norm": 1.5821950568648384, + "learning_rate": 4.686624218784743e-07, + "loss": 0.8013, + "num_input_tokens_seen": 282861290, + "step": 8696 + }, + { + "epoch": 0.7843261036208685, + "flos": 20303086341600.0, + "grad_norm": 2.2765741372841934, + "learning_rate": 4.6828670247314696e-07, + "loss": 0.7344, + "num_input_tokens_seen": 282885605, + "step": 8697 + }, + { + "epoch": 0.7844162871443388, + "flos": 31238848872000.0, + "grad_norm": 3.430222315353841, + "learning_rate": 4.679111137620442e-07, + "loss": 0.7616, + "num_input_tokens_seen": 282913905, + "step": 8698 + }, + { + "epoch": 0.784506470667809, + "flos": 27488275564320.0, + "grad_norm": 1.9595289300689336, + "learning_rate": 4.67535655777213e-07, + "loss": 0.7052, + "num_input_tokens_seen": 282943005, + "step": 8699 + }, + { + "epoch": 0.7845966541912792, + "flos": 22423656340320.0, + "grad_norm": 1.8424773268941517, + "learning_rate": 4.6716032855068956e-07, + "loss": 0.7525, + "num_input_tokens_seen": 282969545, + "step": 8700 + }, + { + "epoch": 0.7846868377147496, + "flos": 11092976379840.0, + "grad_norm": 2.0055205923586303, + "learning_rate": 4.6678513211449867e-07, + "loss": 0.7888, + "num_input_tokens_seen": 282993295, + "step": 8701 + }, + { + "epoch": 0.7847770212382198, + "flos": 17497740425280.0, + "grad_norm": 6.050534109704309, + "learning_rate": 4.6641006650065516e-07, + "loss": 0.7107, + "num_input_tokens_seen": 283012390, + "step": 8702 + }, + { + "epoch": 0.78486720476169, + "flos": 31422310305600.0, + "grad_norm": 1.9283793103938989, + "learning_rate": 4.6603513174115973e-07, + "loss": 0.7331, + "num_input_tokens_seen": 283041450, + "step": 8703 + }, + { + "epoch": 0.7849573882851603, + "flos": 24245670127680.0, + "grad_norm": 2.218205667266609, + "learning_rate": 4.6566032786800625e-07, + "loss": 0.7596, + "num_input_tokens_seen": 283069875, + "step": 8704 + }, + { + "epoch": 0.7850475718086306, + "flos": 22132422571680.0, + "grad_norm": 2.552776590244855, + "learning_rate": 4.6528565491317274e-07, + "loss": 0.7236, + "num_input_tokens_seen": 283097340, + "step": 8705 + }, + { + "epoch": 0.7851377553321008, + "flos": 31313051180160.0, + "grad_norm": 3.7369341520817074, + "learning_rate": 4.649111129086305e-07, + "loss": 0.812, + "num_input_tokens_seen": 283127900, + "step": 8706 + }, + { + "epoch": 0.7852279388555711, + "flos": 24752671584960.0, + "grad_norm": 1.8061803282925981, + "learning_rate": 4.6453670188633596e-07, + "loss": 0.7672, + "num_input_tokens_seen": 283157880, + "step": 8707 + }, + { + "epoch": 0.7853181223790413, + "flos": 24172025365920.0, + "grad_norm": 2.5946014109942963, + "learning_rate": 4.641624218782365e-07, + "loss": 0.7618, + "num_input_tokens_seen": 283185760, + "step": 8708 + }, + { + "epoch": 0.7854083059025116, + "flos": 21875167697280.0, + "grad_norm": 1.9315492861051515, + "learning_rate": 4.6378827291626765e-07, + "loss": 0.7104, + "num_input_tokens_seen": 283216455, + "step": 8709 + }, + { + "epoch": 0.7854984894259819, + "flos": 27851927492640.0, + "grad_norm": 1.3530660656082893, + "learning_rate": 4.634142550323541e-07, + "loss": 0.7967, + "num_input_tokens_seen": 283247510, + "step": 8710 + }, + { + "epoch": 0.7855886729494521, + "flos": 15208948594560.0, + "grad_norm": 2.2008752902315605, + "learning_rate": 4.6304036825840943e-07, + "loss": 0.7935, + "num_input_tokens_seen": 283273220, + "step": 8711 + }, + { + "epoch": 0.7856788564729223, + "flos": 22205249598720.0, + "grad_norm": 2.450620405313961, + "learning_rate": 4.626666126263341e-07, + "loss": 0.6514, + "num_input_tokens_seen": 283302080, + "step": 8712 + }, + { + "epoch": 0.7857690399963927, + "flos": 22055246682720.0, + "grad_norm": 2.0388343654182886, + "learning_rate": 4.622929881680213e-07, + "loss": 0.7706, + "num_input_tokens_seen": 283331010, + "step": 8713 + }, + { + "epoch": 0.7858592235198629, + "flos": 14946898821120.0, + "grad_norm": 3.200020378763159, + "learning_rate": 4.6191949491534887e-07, + "loss": 0.6646, + "num_input_tokens_seen": 283357260, + "step": 8714 + }, + { + "epoch": 0.7859494070433332, + "flos": 18525607660320.0, + "grad_norm": 1.8917502215357453, + "learning_rate": 4.6154613290018617e-07, + "loss": 0.7576, + "num_input_tokens_seen": 283385020, + "step": 8715 + }, + { + "epoch": 0.7860395905668034, + "flos": 20019249355200.0, + "grad_norm": 2.219214699665003, + "learning_rate": 4.6117290215439043e-07, + "loss": 0.6918, + "num_input_tokens_seen": 283413220, + "step": 8716 + }, + { + "epoch": 0.7861297740902737, + "flos": 21148607235840.0, + "grad_norm": 2.315476435169698, + "learning_rate": 4.6079980270980744e-07, + "loss": 0.6085, + "num_input_tokens_seen": 283441915, + "step": 8717 + }, + { + "epoch": 0.786219957613744, + "flos": 19800173557920.0, + "grad_norm": 2.0604033939974213, + "learning_rate": 4.6042683459827245e-07, + "loss": 0.7164, + "num_input_tokens_seen": 283470030, + "step": 8718 + }, + { + "epoch": 0.7863101411372142, + "flos": 70283496411360.0, + "grad_norm": 0.7025657336952429, + "learning_rate": 4.600539978516098e-07, + "loss": 0.5627, + "num_input_tokens_seen": 283564030, + "step": 8719 + }, + { + "epoch": 0.7864003246606845, + "flos": 21799590108000.0, + "grad_norm": 1.8967264887130775, + "learning_rate": 4.5968129250163004e-07, + "loss": 0.7082, + "num_input_tokens_seen": 283593790, + "step": 8720 + }, + { + "epoch": 0.7864905081841548, + "flos": 30729951756960.0, + "grad_norm": 2.1320419316847206, + "learning_rate": 4.5930871858013653e-07, + "loss": 0.6757, + "num_input_tokens_seen": 283623705, + "step": 8721 + }, + { + "epoch": 0.786580691707625, + "flos": 63967393710720.0, + "grad_norm": 0.6562919603284596, + "learning_rate": 4.589362761189182e-07, + "loss": 0.621, + "num_input_tokens_seen": 283720415, + "step": 8722 + }, + { + "epoch": 0.7866708752310952, + "flos": 24901299219840.0, + "grad_norm": 1.781773540454255, + "learning_rate": 4.585639651497539e-07, + "loss": 0.8096, + "num_input_tokens_seen": 283749805, + "step": 8723 + }, + { + "epoch": 0.7867610587545656, + "flos": 39838608242880.0, + "grad_norm": 1.9210801380446239, + "learning_rate": 4.581917857044115e-07, + "loss": 0.6972, + "num_input_tokens_seen": 283779835, + "step": 8724 + }, + { + "epoch": 0.7868512422780358, + "flos": 21841783519200.0, + "grad_norm": 1.8949356119495058, + "learning_rate": 4.5781973781464734e-07, + "loss": 0.7822, + "num_input_tokens_seen": 283811140, + "step": 8725 + }, + { + "epoch": 0.7869414258015061, + "flos": 23004414068640.0, + "grad_norm": 2.845077037338123, + "learning_rate": 4.574478215122073e-07, + "loss": 0.7858, + "num_input_tokens_seen": 283840950, + "step": 8726 + }, + { + "epoch": 0.7870316093249763, + "flos": 21658879632000.0, + "grad_norm": 5.61928742992319, + "learning_rate": 4.5707603682882357e-07, + "loss": 0.7246, + "num_input_tokens_seen": 283866915, + "step": 8727 + }, + { + "epoch": 0.7871217928484466, + "flos": 25041563658720.0, + "grad_norm": 2.8245818984950946, + "learning_rate": 4.56704383796221e-07, + "loss": 0.6101, + "num_input_tokens_seen": 283897005, + "step": 8728 + }, + { + "epoch": 0.7872119763719169, + "flos": 21003324879360.0, + "grad_norm": 2.1305091353289747, + "learning_rate": 4.5633286244610956e-07, + "loss": 0.7814, + "num_input_tokens_seen": 283926995, + "step": 8729 + }, + { + "epoch": 0.7873021598953871, + "flos": 23187652483680.0, + "grad_norm": 1.9024486473661986, + "learning_rate": 4.5596147281018993e-07, + "loss": 0.8122, + "num_input_tokens_seen": 283956775, + "step": 8730 + }, + { + "epoch": 0.7873923434188573, + "flos": 20635212579840.0, + "grad_norm": 3.576720915919493, + "learning_rate": 4.5559021492015137e-07, + "loss": 0.8443, + "num_input_tokens_seen": 283984915, + "step": 8731 + }, + { + "epoch": 0.7874825269423277, + "flos": 36484327646400.0, + "grad_norm": 1.6658351572372478, + "learning_rate": 4.552190888076712e-07, + "loss": 0.6537, + "num_input_tokens_seen": 284015290, + "step": 8732 + }, + { + "epoch": 0.7875727104657979, + "flos": 24864569763360.0, + "grad_norm": 1.9072840028112648, + "learning_rate": 4.548480945044164e-07, + "loss": 0.6928, + "num_input_tokens_seen": 284046880, + "step": 8733 + }, + { + "epoch": 0.7876628939892681, + "flos": 34295093633760.0, + "grad_norm": 1.8128375035142972, + "learning_rate": 4.54477232042042e-07, + "loss": 0.7405, + "num_input_tokens_seen": 284077550, + "step": 8734 + }, + { + "epoch": 0.7877530775127384, + "flos": 29200732868160.0, + "grad_norm": 1.9340760525382534, + "learning_rate": 4.541065014521921e-07, + "loss": 0.7364, + "num_input_tokens_seen": 284108200, + "step": 8735 + }, + { + "epoch": 0.7878432610362087, + "flos": 19399680663840.0, + "grad_norm": 2.940701789137123, + "learning_rate": 4.5373590276649996e-07, + "loss": 0.6929, + "num_input_tokens_seen": 284134005, + "step": 8736 + }, + { + "epoch": 0.787933444559679, + "flos": 35537836483200.0, + "grad_norm": 1.5817044144210382, + "learning_rate": 4.533654360165862e-07, + "loss": 0.7114, + "num_input_tokens_seen": 284165735, + "step": 8737 + }, + { + "epoch": 0.7880236280831492, + "flos": 17141225090880.0, + "grad_norm": 1.9658329257854166, + "learning_rate": 4.5299510123406115e-07, + "loss": 0.7998, + "num_input_tokens_seen": 284193100, + "step": 8738 + }, + { + "epoch": 0.7881138116066194, + "flos": 34373235936480.0, + "grad_norm": 1.6981274136612818, + "learning_rate": 4.5262489845052456e-07, + "loss": 0.7096, + "num_input_tokens_seen": 284222390, + "step": 8739 + }, + { + "epoch": 0.7882039951300898, + "flos": 16556601707520.0, + "grad_norm": 2.0341649194585143, + "learning_rate": 4.5225482769756353e-07, + "loss": 0.7123, + "num_input_tokens_seen": 284248265, + "step": 8740 + }, + { + "epoch": 0.78829417865356, + "flos": 26175493419840.0, + "grad_norm": 2.8177339807684034, + "learning_rate": 4.5188488900675545e-07, + "loss": 0.6732, + "num_input_tokens_seen": 284273170, + "step": 8741 + }, + { + "epoch": 0.7883843621770302, + "flos": 24387867394080.0, + "grad_norm": 1.99031577350948, + "learning_rate": 4.5151508240966363e-07, + "loss": 0.7731, + "num_input_tokens_seen": 284301465, + "step": 8742 + }, + { + "epoch": 0.7884745457005006, + "flos": 23371039577760.0, + "grad_norm": 1.8090662822990264, + "learning_rate": 4.511454079378445e-07, + "loss": 0.7914, + "num_input_tokens_seen": 284329645, + "step": 8743 + }, + { + "epoch": 0.7885647292239708, + "flos": 47051940707520.0, + "grad_norm": 3.3072120636899456, + "learning_rate": 4.507758656228382e-07, + "loss": 0.647, + "num_input_tokens_seen": 284360905, + "step": 8744 + }, + { + "epoch": 0.788654912747441, + "flos": 25879910789280.0, + "grad_norm": 1.9408116495054848, + "learning_rate": 4.5040645549617864e-07, + "loss": 0.8065, + "num_input_tokens_seen": 284391245, + "step": 8745 + }, + { + "epoch": 0.7887450962709113, + "flos": 25301308907040.0, + "grad_norm": 3.237580116488417, + "learning_rate": 4.5003717758938384e-07, + "loss": 0.6612, + "num_input_tokens_seen": 284422115, + "step": 8746 + }, + { + "epoch": 0.7888352797943816, + "flos": 26650002773280.0, + "grad_norm": 1.9039064603796816, + "learning_rate": 4.4966803193396365e-07, + "loss": 0.6846, + "num_input_tokens_seen": 284451600, + "step": 8747 + }, + { + "epoch": 0.7889254633178518, + "flos": 13825235080800.0, + "grad_norm": 1.8119101386661527, + "learning_rate": 4.492990185614154e-07, + "loss": 0.7502, + "num_input_tokens_seen": 284479140, + "step": 8748 + }, + { + "epoch": 0.7890156468413221, + "flos": 27229608239040.0, + "grad_norm": 2.000758384922489, + "learning_rate": 4.489301375032255e-07, + "loss": 0.7325, + "num_input_tokens_seen": 284507895, + "step": 8749 + }, + { + "epoch": 0.7891058303647923, + "flos": 26213746836480.0, + "grad_norm": 2.0372695932994414, + "learning_rate": 4.4856138879086857e-07, + "loss": 0.7493, + "num_input_tokens_seen": 284538415, + "step": 8750 + }, + { + "epoch": 0.7891960138882627, + "flos": 31858677751680.0, + "grad_norm": 1.423599371306025, + "learning_rate": 4.481927724558092e-07, + "loss": 0.6953, + "num_input_tokens_seen": 284572405, + "step": 8751 + }, + { + "epoch": 0.7892861974117329, + "flos": 20419705079520.0, + "grad_norm": 2.142212884420055, + "learning_rate": 4.478242885294985e-07, + "loss": 0.7157, + "num_input_tokens_seen": 284602750, + "step": 8752 + }, + { + "epoch": 0.7893763809352031, + "flos": 39324210003360.0, + "grad_norm": 3.117124781539315, + "learning_rate": 4.474559370433779e-07, + "loss": 0.7061, + "num_input_tokens_seen": 284634685, + "step": 8753 + }, + { + "epoch": 0.7894665644586734, + "flos": 17250632895360.0, + "grad_norm": 2.4054515554291167, + "learning_rate": 4.470877180288777e-07, + "loss": 0.6904, + "num_input_tokens_seen": 284661115, + "step": 8754 + }, + { + "epoch": 0.7895567479821437, + "flos": 26140956979200.0, + "grad_norm": 1.6127340575686664, + "learning_rate": 4.4671963151741574e-07, + "loss": 0.7124, + "num_input_tokens_seen": 284692115, + "step": 8755 + }, + { + "epoch": 0.7896469315056139, + "flos": 25667376869760.0, + "grad_norm": 5.50256825218871, + "learning_rate": 4.4635167754039973e-07, + "loss": 0.721, + "num_input_tokens_seen": 284724590, + "step": 8756 + }, + { + "epoch": 0.7897371150290842, + "flos": 23879118958080.0, + "grad_norm": 1.5693407230743384, + "learning_rate": 4.459838561292253e-07, + "loss": 0.7721, + "num_input_tokens_seen": 284754660, + "step": 8757 + }, + { + "epoch": 0.7898272985525544, + "flos": 23910161441280.0, + "grad_norm": 2.2973725694717926, + "learning_rate": 4.456161673152774e-07, + "loss": 0.6561, + "num_input_tokens_seen": 284782000, + "step": 8758 + }, + { + "epoch": 0.7899174820760247, + "flos": 21731780998560.0, + "grad_norm": 1.9405470852406572, + "learning_rate": 4.4524861112992806e-07, + "loss": 0.7334, + "num_input_tokens_seen": 284810260, + "step": 8759 + }, + { + "epoch": 0.790007665599495, + "flos": 40238692269600.0, + "grad_norm": 2.226563892029444, + "learning_rate": 4.448811876045411e-07, + "loss": 0.7422, + "num_input_tokens_seen": 284840280, + "step": 8760 + }, + { + "epoch": 0.7900978491229652, + "flos": 18998927581440.0, + "grad_norm": 2.4959514260945346, + "learning_rate": 4.445138967704647e-07, + "loss": 0.8112, + "num_input_tokens_seen": 284866960, + "step": 8761 + }, + { + "epoch": 0.7901880326464354, + "flos": 29123222451360.0, + "grad_norm": 2.0347307166687854, + "learning_rate": 4.4414673865904075e-07, + "loss": 0.7163, + "num_input_tokens_seen": 284896950, + "step": 8762 + }, + { + "epoch": 0.7902782161699058, + "flos": 24608876018880.0, + "grad_norm": 1.565169076530923, + "learning_rate": 4.437797133015955e-07, + "loss": 0.7591, + "num_input_tokens_seen": 284926915, + "step": 8763 + }, + { + "epoch": 0.790368399693376, + "flos": 22130824272000.0, + "grad_norm": 2.3170792886851563, + "learning_rate": 4.4341282072944586e-07, + "loss": 0.6575, + "num_input_tokens_seen": 284957680, + "step": 8764 + }, + { + "epoch": 0.7904585832168463, + "flos": 25520384704320.0, + "grad_norm": 1.8094669483388575, + "learning_rate": 4.430460609738973e-07, + "loss": 0.7643, + "num_input_tokens_seen": 284987310, + "step": 8765 + }, + { + "epoch": 0.7905487667403166, + "flos": 67751165347680.0, + "grad_norm": 0.7082207300403977, + "learning_rate": 4.4267943406624386e-07, + "loss": 0.6095, + "num_input_tokens_seen": 285077640, + "step": 8766 + }, + { + "epoch": 0.7906389502637868, + "flos": 15355606232160.0, + "grad_norm": 5.5411686303007155, + "learning_rate": 4.4231294003776853e-07, + "loss": 0.8533, + "num_input_tokens_seen": 285104575, + "step": 8767 + }, + { + "epoch": 0.7907291337872571, + "flos": 24864569763360.0, + "grad_norm": 2.3913216168077867, + "learning_rate": 4.419465789197416e-07, + "loss": 0.7826, + "num_input_tokens_seen": 285131055, + "step": 8768 + }, + { + "epoch": 0.7908193173107273, + "flos": 21768213096960.0, + "grad_norm": 1.9377232699804066, + "learning_rate": 4.415803507434237e-07, + "loss": 0.74, + "num_input_tokens_seen": 285158930, + "step": 8769 + }, + { + "epoch": 0.7909095008341976, + "flos": 23696066391840.0, + "grad_norm": 2.048837434440242, + "learning_rate": 4.4121425554006307e-07, + "loss": 0.766, + "num_input_tokens_seen": 285188485, + "step": 8770 + }, + { + "epoch": 0.7909996843576679, + "flos": 23509148170560.0, + "grad_norm": 2.173797415270503, + "learning_rate": 4.4084829334089744e-07, + "loss": 0.7258, + "num_input_tokens_seen": 285215310, + "step": 8771 + }, + { + "epoch": 0.7910898678811381, + "flos": 37397620480320.0, + "grad_norm": 1.8249891071778344, + "learning_rate": 4.404824641771525e-07, + "loss": 0.6823, + "num_input_tokens_seen": 285249415, + "step": 8772 + }, + { + "epoch": 0.7911800514046083, + "flos": 23729896607040.0, + "grad_norm": 1.5117138608710505, + "learning_rate": 4.4011676808004327e-07, + "loss": 0.81, + "num_input_tokens_seen": 285277925, + "step": 8773 + }, + { + "epoch": 0.7912702349280787, + "flos": 23550263658720.0, + "grad_norm": 1.7559542858368993, + "learning_rate": 4.3975120508077145e-07, + "loss": 0.7513, + "num_input_tokens_seen": 285305875, + "step": 8774 + }, + { + "epoch": 0.7913604184515489, + "flos": 26977965998400.0, + "grad_norm": 2.0663295219613103, + "learning_rate": 4.39385775210531e-07, + "loss": 0.7916, + "num_input_tokens_seen": 285334770, + "step": 8775 + }, + { + "epoch": 0.7914506019750192, + "flos": 18744237420480.0, + "grad_norm": 2.1869792863434405, + "learning_rate": 4.390204785005003e-07, + "loss": 0.7581, + "num_input_tokens_seen": 285362800, + "step": 8776 + }, + { + "epoch": 0.7915407854984894, + "flos": 24171616498560.0, + "grad_norm": 1.920206291421414, + "learning_rate": 4.386553149818504e-07, + "loss": 0.7477, + "num_input_tokens_seen": 285394720, + "step": 8777 + }, + { + "epoch": 0.7916309690219597, + "flos": 33243617867520.0, + "grad_norm": 2.86780643279178, + "learning_rate": 4.3829028468573793e-07, + "loss": 0.6106, + "num_input_tokens_seen": 285425035, + "step": 8778 + }, + { + "epoch": 0.79172115254543, + "flos": 29342446927680.0, + "grad_norm": 1.875368336420159, + "learning_rate": 4.3792538764330935e-07, + "loss": 0.6834, + "num_input_tokens_seen": 285454100, + "step": 8779 + }, + { + "epoch": 0.7918113360689002, + "flos": 17104569973920.0, + "grad_norm": 2.395060295667993, + "learning_rate": 4.3756062388569994e-07, + "loss": 0.7151, + "num_input_tokens_seen": 285478660, + "step": 8780 + }, + { + "epoch": 0.7919015195923704, + "flos": 52963524712800.0, + "grad_norm": 0.7095002958580728, + "learning_rate": 4.3719599344403346e-07, + "loss": 0.5793, + "num_input_tokens_seen": 285573365, + "step": 8781 + }, + { + "epoch": 0.7919917031158408, + "flos": 32368615620000.0, + "grad_norm": 1.5915694185410305, + "learning_rate": 4.3683149634942243e-07, + "loss": 0.7229, + "num_input_tokens_seen": 285604820, + "step": 8782 + }, + { + "epoch": 0.792081886639311, + "flos": 37249476052320.0, + "grad_norm": 1.9596033307022471, + "learning_rate": 4.364671326329663e-07, + "loss": 0.6253, + "num_input_tokens_seen": 285640620, + "step": 8783 + }, + { + "epoch": 0.7921720701627812, + "flos": 49492816960800.0, + "grad_norm": 2.405880193655866, + "learning_rate": 4.3610290232575673e-07, + "loss": 0.764, + "num_input_tokens_seen": 285673495, + "step": 8784 + }, + { + "epoch": 0.7922622536862515, + "flos": 22752400130400.0, + "grad_norm": 1.7330817134718948, + "learning_rate": 4.357388054588702e-07, + "loss": 0.7404, + "num_input_tokens_seen": 285701220, + "step": 8785 + }, + { + "epoch": 0.7923524372097218, + "flos": 22533175654080.0, + "grad_norm": 1.6783434081067194, + "learning_rate": 4.3537484206337405e-07, + "loss": 0.6877, + "num_input_tokens_seen": 285729230, + "step": 8786 + }, + { + "epoch": 0.792442620733192, + "flos": 15902236387200.0, + "grad_norm": 1.9438313433440098, + "learning_rate": 4.3501101217032366e-07, + "loss": 0.7737, + "num_input_tokens_seen": 285754470, + "step": 8787 + }, + { + "epoch": 0.7925328042566623, + "flos": 34518964330080.0, + "grad_norm": 2.088516184460318, + "learning_rate": 4.346473158107629e-07, + "loss": 0.668, + "num_input_tokens_seen": 285784305, + "step": 8788 + }, + { + "epoch": 0.7926229877801325, + "flos": 24063212277600.0, + "grad_norm": 2.1898188698124486, + "learning_rate": 4.342837530157244e-07, + "loss": 0.7463, + "num_input_tokens_seen": 285811475, + "step": 8789 + }, + { + "epoch": 0.7927131713036029, + "flos": 40456206936960.0, + "grad_norm": 2.5617483013434468, + "learning_rate": 4.3392032381622987e-07, + "loss": 0.6451, + "num_input_tokens_seen": 285841615, + "step": 8790 + }, + { + "epoch": 0.7928033548270731, + "flos": 19580911911840.0, + "grad_norm": 1.5637225899249447, + "learning_rate": 4.3355702824328765e-07, + "loss": 0.7035, + "num_input_tokens_seen": 285870190, + "step": 8791 + }, + { + "epoch": 0.7928935383505433, + "flos": 25556036237760.0, + "grad_norm": 1.8119844163725587, + "learning_rate": 4.3319386632789823e-07, + "loss": 0.7839, + "num_input_tokens_seen": 285901540, + "step": 8792 + }, + { + "epoch": 0.7929837218740137, + "flos": 20563686494400.0, + "grad_norm": 2.0713431617905527, + "learning_rate": 4.328308381010466e-07, + "loss": 0.7376, + "num_input_tokens_seen": 285929375, + "step": 8793 + }, + { + "epoch": 0.7930739053974839, + "flos": 22169412216480.0, + "grad_norm": 1.8095891532596482, + "learning_rate": 4.3246794359370933e-07, + "loss": 0.747, + "num_input_tokens_seen": 285955945, + "step": 8794 + }, + { + "epoch": 0.7931640889209541, + "flos": 21439989683520.0, + "grad_norm": 2.1244282794857905, + "learning_rate": 4.3210518283685025e-07, + "loss": 0.7173, + "num_input_tokens_seen": 285984975, + "step": 8795 + }, + { + "epoch": 0.7932542724444244, + "flos": 19763815799040.0, + "grad_norm": 2.1846528189046306, + "learning_rate": 4.317425558614225e-07, + "loss": 0.7397, + "num_input_tokens_seen": 286012175, + "step": 8796 + }, + { + "epoch": 0.7933444559678947, + "flos": 21364003226880.0, + "grad_norm": 1.4683237471828676, + "learning_rate": 4.3138006269836744e-07, + "loss": 0.7063, + "num_input_tokens_seen": 286039775, + "step": 8797 + }, + { + "epoch": 0.793434639491365, + "flos": 26026531257120.0, + "grad_norm": 2.579796543970565, + "learning_rate": 4.3101770337861376e-07, + "loss": 0.7299, + "num_input_tokens_seen": 286069990, + "step": 8798 + }, + { + "epoch": 0.7935248230148352, + "flos": 62678108588160.0, + "grad_norm": 0.6656728404945843, + "learning_rate": 4.30655477933082e-07, + "loss": 0.5928, + "num_input_tokens_seen": 286165170, + "step": 8799 + }, + { + "epoch": 0.7936150065383054, + "flos": 27887876384160.0, + "grad_norm": 1.749225240595686, + "learning_rate": 4.30293386392677e-07, + "loss": 0.7659, + "num_input_tokens_seen": 286198805, + "step": 8800 + }, + { + "epoch": 0.7937051900617758, + "flos": 25741616347680.0, + "grad_norm": 1.5809255998491203, + "learning_rate": 4.299314287882967e-07, + "loss": 0.7617, + "num_input_tokens_seen": 286232635, + "step": 8801 + }, + { + "epoch": 0.793795373585246, + "flos": 30219121814400.0, + "grad_norm": 2.1233777978853623, + "learning_rate": 4.2956960515082353e-07, + "loss": 0.7745, + "num_input_tokens_seen": 286264700, + "step": 8802 + }, + { + "epoch": 0.7938855571087162, + "flos": 16769470154880.0, + "grad_norm": 1.917401449364134, + "learning_rate": 4.29207915511131e-07, + "loss": 0.6973, + "num_input_tokens_seen": 286292240, + "step": 8803 + }, + { + "epoch": 0.7939757406321865, + "flos": 19654073466720.0, + "grad_norm": 2.0134203778749216, + "learning_rate": 4.2884635990008024e-07, + "loss": 0.8002, + "num_input_tokens_seen": 286319045, + "step": 8804 + }, + { + "epoch": 0.7940659241556568, + "flos": 21294112610880.0, + "grad_norm": 2.5511603314698266, + "learning_rate": 4.284849383485214e-07, + "loss": 0.7066, + "num_input_tokens_seen": 286348300, + "step": 8805 + }, + { + "epoch": 0.794156107679127, + "flos": 21512891050080.0, + "grad_norm": 2.9356751962763568, + "learning_rate": 4.2812365088729296e-07, + "loss": 0.7676, + "num_input_tokens_seen": 286376695, + "step": 8806 + }, + { + "epoch": 0.7942462912025973, + "flos": 21076337755200.0, + "grad_norm": 3.275064651200562, + "learning_rate": 4.2776249754722227e-07, + "loss": 0.7623, + "num_input_tokens_seen": 286406945, + "step": 8807 + }, + { + "epoch": 0.7943364747260675, + "flos": 17970725818560.0, + "grad_norm": 3.0163113302231452, + "learning_rate": 4.27401478359124e-07, + "loss": 0.8126, + "num_input_tokens_seen": 286426605, + "step": 8808 + }, + { + "epoch": 0.7944266582495378, + "flos": 23404795453440.0, + "grad_norm": 1.9822699126832164, + "learning_rate": 4.2704059335380283e-07, + "loss": 0.8367, + "num_input_tokens_seen": 286455270, + "step": 8809 + }, + { + "epoch": 0.7945168417730081, + "flos": 25922996274720.0, + "grad_norm": 1.646398069171867, + "learning_rate": 4.266798425620515e-07, + "loss": 0.8451, + "num_input_tokens_seen": 286487270, + "step": 8810 + }, + { + "epoch": 0.7946070252964783, + "flos": 50734296038400.0, + "grad_norm": 1.9736921857371572, + "learning_rate": 4.263192260146511e-07, + "loss": 0.7274, + "num_input_tokens_seen": 286523730, + "step": 8811 + }, + { + "epoch": 0.7946972088199485, + "flos": 22131976534560.0, + "grad_norm": 1.9469579761634956, + "learning_rate": 4.2595874374237216e-07, + "loss": 0.7906, + "num_input_tokens_seen": 286551105, + "step": 8812 + }, + { + "epoch": 0.7947873923434189, + "flos": 22747939759200.0, + "grad_norm": 1.4636883464178614, + "learning_rate": 4.255983957759712e-07, + "loss": 0.7376, + "num_input_tokens_seen": 286581480, + "step": 8813 + }, + { + "epoch": 0.7948775758668891, + "flos": 23698259407680.0, + "grad_norm": 1.6506933893886788, + "learning_rate": 4.2523818214619745e-07, + "loss": 0.6793, + "num_input_tokens_seen": 286612745, + "step": 8814 + }, + { + "epoch": 0.7949677593903594, + "flos": 26902685767200.0, + "grad_norm": 1.696339520289769, + "learning_rate": 4.24878102883784e-07, + "loss": 0.7882, + "num_input_tokens_seen": 286642610, + "step": 8815 + }, + { + "epoch": 0.7950579429138297, + "flos": 21984092294880.0, + "grad_norm": 1.6525496406428517, + "learning_rate": 4.24518158019457e-07, + "loss": 0.7487, + "num_input_tokens_seen": 286671235, + "step": 8816 + }, + { + "epoch": 0.7951481264372999, + "flos": 27414779481600.0, + "grad_norm": 2.756796587596648, + "learning_rate": 4.241583475839274e-07, + "loss": 0.6821, + "num_input_tokens_seen": 286699560, + "step": 8817 + }, + { + "epoch": 0.7952383099607702, + "flos": 22678718198880.0, + "grad_norm": 1.826923986581429, + "learning_rate": 4.237986716078965e-07, + "loss": 0.7865, + "num_input_tokens_seen": 286727735, + "step": 8818 + }, + { + "epoch": 0.7953284934842404, + "flos": 26066643161760.0, + "grad_norm": 1.8403171327169559, + "learning_rate": 4.2343913012205433e-07, + "loss": 0.7285, + "num_input_tokens_seen": 286757745, + "step": 8819 + }, + { + "epoch": 0.7954186770077107, + "flos": 25480718836800.0, + "grad_norm": 1.937535870127729, + "learning_rate": 4.230797231570784e-07, + "loss": 0.7168, + "num_input_tokens_seen": 286788570, + "step": 8820 + }, + { + "epoch": 0.795508860531181, + "flos": 22423358982240.0, + "grad_norm": 3.2585044510210843, + "learning_rate": 4.227204507436357e-07, + "loss": 0.7442, + "num_input_tokens_seen": 286817840, + "step": 8821 + }, + { + "epoch": 0.7955990440546512, + "flos": 74626801843680.0, + "grad_norm": 1.758516636856251, + "learning_rate": 4.223613129123811e-07, + "loss": 0.6599, + "num_input_tokens_seen": 286856415, + "step": 8822 + }, + { + "epoch": 0.7956892275781214, + "flos": 24643635478080.0, + "grad_norm": 1.7440020088569201, + "learning_rate": 4.220023096939589e-07, + "loss": 0.6472, + "num_input_tokens_seen": 286885515, + "step": 8823 + }, + { + "epoch": 0.7957794111015918, + "flos": 19107740669760.0, + "grad_norm": 2.0523383922819174, + "learning_rate": 4.21643441119e-07, + "loss": 0.7365, + "num_input_tokens_seen": 286910440, + "step": 8824 + }, + { + "epoch": 0.795869594625062, + "flos": 66915977646720.0, + "grad_norm": 0.6293727258377886, + "learning_rate": 4.212847072181256e-07, + "loss": 0.629, + "num_input_tokens_seen": 287001595, + "step": 8825 + }, + { + "epoch": 0.7959597781485322, + "flos": 21367385675040.0, + "grad_norm": 1.9387433769265436, + "learning_rate": 4.2092610802194505e-07, + "loss": 0.7159, + "num_input_tokens_seen": 287030295, + "step": 8826 + }, + { + "epoch": 0.7960499616720025, + "flos": 24243885979200.0, + "grad_norm": 2.1873791252527655, + "learning_rate": 4.2056764356105587e-07, + "loss": 0.652, + "num_input_tokens_seen": 287059545, + "step": 8827 + }, + { + "epoch": 0.7961401451954728, + "flos": 24719027218560.0, + "grad_norm": 1.8114792152621184, + "learning_rate": 4.202093138660443e-07, + "loss": 0.7878, + "num_input_tokens_seen": 287087560, + "step": 8828 + }, + { + "epoch": 0.7962303287189431, + "flos": 25044946106880.0, + "grad_norm": 1.902507673090175, + "learning_rate": 4.198511189674854e-07, + "loss": 0.7735, + "num_input_tokens_seen": 287117715, + "step": 8829 + }, + { + "epoch": 0.7963205122424133, + "flos": 28507816773120.0, + "grad_norm": 2.6354930600237965, + "learning_rate": 4.1949305889594066e-07, + "loss": 0.761, + "num_input_tokens_seen": 287146620, + "step": 8830 + }, + { + "epoch": 0.7964106957658835, + "flos": 27303476019360.0, + "grad_norm": 1.6035412729421499, + "learning_rate": 4.191351336819642e-07, + "loss": 0.7332, + "num_input_tokens_seen": 287178175, + "step": 8831 + }, + { + "epoch": 0.7965008792893539, + "flos": 34628260625280.0, + "grad_norm": 1.733207447114533, + "learning_rate": 4.187773433560939e-07, + "loss": 0.6782, + "num_input_tokens_seen": 287209490, + "step": 8832 + }, + { + "epoch": 0.7965910628128241, + "flos": 25810057343040.0, + "grad_norm": 3.043573556912645, + "learning_rate": 4.184196879488604e-07, + "loss": 0.7377, + "num_input_tokens_seen": 287237920, + "step": 8833 + }, + { + "epoch": 0.7966812463362943, + "flos": 27086556068160.0, + "grad_norm": 1.7734424742238197, + "learning_rate": 4.1806216749077936e-07, + "loss": 0.8028, + "num_input_tokens_seen": 287267705, + "step": 8834 + }, + { + "epoch": 0.7967714298597646, + "flos": 30689951361600.0, + "grad_norm": 1.586115634479, + "learning_rate": 4.177047820123569e-07, + "loss": 0.7832, + "num_input_tokens_seen": 287297645, + "step": 8835 + }, + { + "epoch": 0.7968616133832349, + "flos": 22713477658080.0, + "grad_norm": 1.8669334462240492, + "learning_rate": 4.1734753154408733e-07, + "loss": 0.6129, + "num_input_tokens_seen": 287326215, + "step": 8836 + }, + { + "epoch": 0.7969517969067051, + "flos": 20558742916320.0, + "grad_norm": 2.0834004836386666, + "learning_rate": 4.169904161164528e-07, + "loss": 0.7237, + "num_input_tokens_seen": 287353265, + "step": 8837 + }, + { + "epoch": 0.7970419804301754, + "flos": 18561147684480.0, + "grad_norm": 2.3214938563294942, + "learning_rate": 4.1663343575992526e-07, + "loss": 0.8594, + "num_input_tokens_seen": 287377675, + "step": 8838 + }, + { + "epoch": 0.7971321639536457, + "flos": 24790962171360.0, + "grad_norm": 1.9143951127019811, + "learning_rate": 4.1627659050496275e-07, + "loss": 0.7585, + "num_input_tokens_seen": 287402915, + "step": 8839 + }, + { + "epoch": 0.797222347477116, + "flos": 26358992023200.0, + "grad_norm": 1.9628767781694436, + "learning_rate": 4.1591988038201453e-07, + "loss": 0.6571, + "num_input_tokens_seen": 287433930, + "step": 8840 + }, + { + "epoch": 0.7973125310005862, + "flos": 17360300888160.0, + "grad_norm": 1.764480017017334, + "learning_rate": 4.155633054215164e-07, + "loss": 0.7859, + "num_input_tokens_seen": 287461365, + "step": 8841 + }, + { + "epoch": 0.7974027145240564, + "flos": 27050049630240.0, + "grad_norm": 2.233161436649001, + "learning_rate": 4.152068656538934e-07, + "loss": 0.6507, + "num_input_tokens_seen": 287488980, + "step": 8842 + }, + { + "epoch": 0.7974928980475268, + "flos": 34628483643840.0, + "grad_norm": 1.661742478072576, + "learning_rate": 4.148505611095594e-07, + "loss": 0.7974, + "num_input_tokens_seen": 287519625, + "step": 8843 + }, + { + "epoch": 0.797583081570997, + "flos": 26102852241600.0, + "grad_norm": 3.0375475404888697, + "learning_rate": 4.1449439181891563e-07, + "loss": 0.6155, + "num_input_tokens_seen": 287548570, + "step": 8844 + }, + { + "epoch": 0.7976732650944672, + "flos": 23188693236960.0, + "grad_norm": 2.0229651883006223, + "learning_rate": 4.14138357812353e-07, + "loss": 0.7603, + "num_input_tokens_seen": 287576195, + "step": 8845 + }, + { + "epoch": 0.7977634486179375, + "flos": 27487903866720.0, + "grad_norm": 1.639419330633088, + "learning_rate": 4.137824591202506e-07, + "loss": 0.7966, + "num_input_tokens_seen": 287608180, + "step": 8846 + }, + { + "epoch": 0.7978536321414078, + "flos": 22459605231840.0, + "grad_norm": 1.7344429758858408, + "learning_rate": 4.134266957729737e-07, + "loss": 0.7301, + "num_input_tokens_seen": 287637915, + "step": 8847 + }, + { + "epoch": 0.797943815664878, + "flos": 20165646804480.0, + "grad_norm": 1.8092072272384716, + "learning_rate": 4.1307106780088065e-07, + "loss": 0.8359, + "num_input_tokens_seen": 287664555, + "step": 8848 + }, + { + "epoch": 0.7980339991883483, + "flos": 24938028676320.0, + "grad_norm": 1.920147437256158, + "learning_rate": 4.1271557523431387e-07, + "loss": 0.751, + "num_input_tokens_seen": 287693355, + "step": 8849 + }, + { + "epoch": 0.7981241827118185, + "flos": 16521842248320.0, + "grad_norm": 2.0077796072436405, + "learning_rate": 4.1236021810360634e-07, + "loss": 0.7251, + "num_input_tokens_seen": 287719710, + "step": 8850 + }, + { + "epoch": 0.7982143662352889, + "flos": 20966112216000.0, + "grad_norm": 2.1220018891048382, + "learning_rate": 4.120049964390793e-07, + "loss": 0.6849, + "num_input_tokens_seen": 287748135, + "step": 8851 + }, + { + "epoch": 0.7983045497587591, + "flos": 23843616103680.0, + "grad_norm": 3.0715891583813324, + "learning_rate": 4.116499102710418e-07, + "loss": 0.69, + "num_input_tokens_seen": 287773740, + "step": 8852 + }, + { + "epoch": 0.7983947332822293, + "flos": 26795842676160.0, + "grad_norm": 2.7016551701816676, + "learning_rate": 4.112949596297928e-07, + "loss": 0.7132, + "num_input_tokens_seen": 287803385, + "step": 8853 + }, + { + "epoch": 0.7984849168056996, + "flos": 24714269489280.0, + "grad_norm": 2.0116026020778865, + "learning_rate": 4.1094014454561664e-07, + "loss": 0.6865, + "num_input_tokens_seen": 287832130, + "step": 8854 + }, + { + "epoch": 0.7985751003291699, + "flos": 17468742278880.0, + "grad_norm": 2.334539192686107, + "learning_rate": 4.1058546504879057e-07, + "loss": 0.6464, + "num_input_tokens_seen": 287857925, + "step": 8855 + }, + { + "epoch": 0.7986652838526401, + "flos": 24937805657760.0, + "grad_norm": 2.0068828642247643, + "learning_rate": 4.1023092116957583e-07, + "loss": 0.7974, + "num_input_tokens_seen": 287884775, + "step": 8856 + }, + { + "epoch": 0.7987554673761104, + "flos": 34155646929600.0, + "grad_norm": 1.6578644854909574, + "learning_rate": 4.098765129382249e-07, + "loss": 0.7794, + "num_input_tokens_seen": 287916690, + "step": 8857 + }, + { + "epoch": 0.7988456508995806, + "flos": 24859105808640.0, + "grad_norm": 1.922772515853172, + "learning_rate": 4.0952224038497764e-07, + "loss": 0.8124, + "num_input_tokens_seen": 287942750, + "step": 8858 + }, + { + "epoch": 0.7989358344230509, + "flos": 13236894721440.0, + "grad_norm": 2.3040058744408634, + "learning_rate": 4.091681035400627e-07, + "loss": 0.7358, + "num_input_tokens_seen": 287967825, + "step": 8859 + }, + { + "epoch": 0.7990260179465212, + "flos": 21403631924640.0, + "grad_norm": 1.9098085426319278, + "learning_rate": 4.088141024336971e-07, + "loss": 0.7767, + "num_input_tokens_seen": 287996215, + "step": 8860 + }, + { + "epoch": 0.7991162014699914, + "flos": 22748757493920.0, + "grad_norm": 1.5744458358960107, + "learning_rate": 4.0846023709608636e-07, + "loss": 0.8207, + "num_input_tokens_seen": 288025610, + "step": 8861 + }, + { + "epoch": 0.7992063849934617, + "flos": 23443606416480.0, + "grad_norm": 2.133372809621029, + "learning_rate": 4.081065075574226e-07, + "loss": 0.8155, + "num_input_tokens_seen": 288055005, + "step": 8862 + }, + { + "epoch": 0.799296568516932, + "flos": 24457311972960.0, + "grad_norm": 1.6906301642883736, + "learning_rate": 4.077529138478906e-07, + "loss": 0.6708, + "num_input_tokens_seen": 288083235, + "step": 8863 + }, + { + "epoch": 0.7993867520404022, + "flos": 67530379741440.0, + "grad_norm": 0.6803063173587, + "learning_rate": 4.073994559976588e-07, + "loss": 0.5405, + "num_input_tokens_seen": 288166660, + "step": 8864 + }, + { + "epoch": 0.7994769355638724, + "flos": 27232804838400.0, + "grad_norm": 2.02930192698932, + "learning_rate": 4.0704613403688716e-07, + "loss": 0.7051, + "num_input_tokens_seen": 288196770, + "step": 8865 + }, + { + "epoch": 0.7995671190873428, + "flos": 23477696820000.0, + "grad_norm": 2.1206419518038033, + "learning_rate": 4.0669294799572264e-07, + "loss": 0.7992, + "num_input_tokens_seen": 288225385, + "step": 8866 + }, + { + "epoch": 0.799657302610813, + "flos": 14626518227040.0, + "grad_norm": 2.3821316105815473, + "learning_rate": 4.0633989790430113e-07, + "loss": 0.8413, + "num_input_tokens_seen": 288248775, + "step": 8867 + }, + { + "epoch": 0.7997474861342833, + "flos": 22860395484000.0, + "grad_norm": 1.994444940644288, + "learning_rate": 4.059869837927477e-07, + "loss": 0.7494, + "num_input_tokens_seen": 288275510, + "step": 8868 + }, + { + "epoch": 0.7998376696577535, + "flos": 22751285037600.0, + "grad_norm": 2.55182910094044, + "learning_rate": 4.056342056911728e-07, + "loss": 0.7249, + "num_input_tokens_seen": 288303985, + "step": 8869 + }, + { + "epoch": 0.7999278531812238, + "flos": 24865052970240.0, + "grad_norm": 2.344077837585793, + "learning_rate": 4.052815636296798e-07, + "loss": 0.7613, + "num_input_tokens_seen": 288334830, + "step": 8870 + }, + { + "epoch": 0.8000180367046941, + "flos": 21730814584800.0, + "grad_norm": 2.0129175262459866, + "learning_rate": 4.0492905763835593e-07, + "loss": 0.7742, + "num_input_tokens_seen": 288363750, + "step": 8871 + }, + { + "epoch": 0.8001082202281643, + "flos": 23587922359200.0, + "grad_norm": 1.8798869977631982, + "learning_rate": 4.0457668774728115e-07, + "loss": 0.7723, + "num_input_tokens_seen": 288391060, + "step": 8872 + }, + { + "epoch": 0.8001984037516345, + "flos": 23692981301760.0, + "grad_norm": 2.478442227815172, + "learning_rate": 4.0422445398651985e-07, + "loss": 0.7601, + "num_input_tokens_seen": 288419250, + "step": 8873 + }, + { + "epoch": 0.8002885872751049, + "flos": 40566804173760.0, + "grad_norm": 1.8602947236436833, + "learning_rate": 4.0387235638612706e-07, + "loss": 0.6651, + "num_input_tokens_seen": 288453070, + "step": 8874 + }, + { + "epoch": 0.8003787707985751, + "flos": 19144432956480.0, + "grad_norm": 2.1831013733431446, + "learning_rate": 4.0352039497614586e-07, + "loss": 0.7487, + "num_input_tokens_seen": 288479835, + "step": 8875 + }, + { + "epoch": 0.8004689543220453, + "flos": 17942322388320.0, + "grad_norm": 2.352883813663689, + "learning_rate": 4.031685697866074e-07, + "loss": 0.7313, + "num_input_tokens_seen": 288507325, + "step": 8876 + }, + { + "epoch": 0.8005591378455156, + "flos": 22234953970560.0, + "grad_norm": 1.5766923652447142, + "learning_rate": 4.0281688084753165e-07, + "loss": 0.782, + "num_input_tokens_seen": 288536880, + "step": 8877 + }, + { + "epoch": 0.8006493213689859, + "flos": 21694568335200.0, + "grad_norm": 2.42961213709488, + "learning_rate": 4.0246532818892675e-07, + "loss": 0.7918, + "num_input_tokens_seen": 288563795, + "step": 8878 + }, + { + "epoch": 0.8007395048924562, + "flos": 20779193994720.0, + "grad_norm": 2.0657248283883005, + "learning_rate": 4.0211391184078814e-07, + "loss": 0.7779, + "num_input_tokens_seen": 288590765, + "step": 8879 + }, + { + "epoch": 0.8008296884159264, + "flos": 22640464782240.0, + "grad_norm": 2.1327794110871903, + "learning_rate": 4.0176263183310135e-07, + "loss": 0.7749, + "num_input_tokens_seen": 288619735, + "step": 8880 + }, + { + "epoch": 0.8009198719393966, + "flos": 23295945195360.0, + "grad_norm": 1.7377463900183272, + "learning_rate": 4.0141148819583925e-07, + "loss": 0.7976, + "num_input_tokens_seen": 288648225, + "step": 8881 + }, + { + "epoch": 0.801010055462867, + "flos": 64078845852000.0, + "grad_norm": 0.6034001607222654, + "learning_rate": 4.010604809589637e-07, + "loss": 0.569, + "num_input_tokens_seen": 288742820, + "step": 8882 + }, + { + "epoch": 0.8011002389863372, + "flos": 27778988956320.0, + "grad_norm": 1.7747644509923979, + "learning_rate": 4.0070961015242475e-07, + "loss": 0.7452, + "num_input_tokens_seen": 288770290, + "step": 8883 + }, + { + "epoch": 0.8011904225098074, + "flos": 22816566603360.0, + "grad_norm": 2.3119109150906283, + "learning_rate": 4.0035887580615933e-07, + "loss": 0.8138, + "num_input_tokens_seen": 288797900, + "step": 8884 + }, + { + "epoch": 0.8012806060332777, + "flos": 26538030255360.0, + "grad_norm": 2.323129914991564, + "learning_rate": 4.0000827795009594e-07, + "loss": 0.7914, + "num_input_tokens_seen": 288828090, + "step": 8885 + }, + { + "epoch": 0.801370789556748, + "flos": 19035025152000.0, + "grad_norm": 3.6251085199357873, + "learning_rate": 3.996578166141475e-07, + "loss": 0.7084, + "num_input_tokens_seen": 288854350, + "step": 8886 + }, + { + "epoch": 0.8014609730802182, + "flos": 19800693934560.0, + "grad_norm": 4.61327649535371, + "learning_rate": 3.9930749182821955e-07, + "loss": 0.7511, + "num_input_tokens_seen": 288882255, + "step": 8887 + }, + { + "epoch": 0.8015511566036885, + "flos": 58899800905920.0, + "grad_norm": 0.5984872513451388, + "learning_rate": 3.9895730362220116e-07, + "loss": 0.5944, + "num_input_tokens_seen": 288973515, + "step": 8888 + }, + { + "epoch": 0.8016413401271588, + "flos": 15209877838560.0, + "grad_norm": 1.9301356980446247, + "learning_rate": 3.986072520259749e-07, + "loss": 0.7843, + "num_input_tokens_seen": 288998805, + "step": 8889 + }, + { + "epoch": 0.801731523650629, + "flos": 22569384733920.0, + "grad_norm": 1.9693400997264294, + "learning_rate": 3.9825733706940736e-07, + "loss": 0.7321, + "num_input_tokens_seen": 289028125, + "step": 8890 + }, + { + "epoch": 0.8018217071740993, + "flos": 25957161017760.0, + "grad_norm": 1.9198545054105507, + "learning_rate": 3.979075587823557e-07, + "loss": 0.76, + "num_input_tokens_seen": 289057575, + "step": 8891 + }, + { + "epoch": 0.8019118906975695, + "flos": 26387246774400.0, + "grad_norm": 5.402980278728952, + "learning_rate": 3.9755791719466504e-07, + "loss": 0.6213, + "num_input_tokens_seen": 289084745, + "step": 8892 + }, + { + "epoch": 0.8020020742210399, + "flos": 23042816164320.0, + "grad_norm": 2.10978457395212, + "learning_rate": 3.9720841233616875e-07, + "loss": 0.7013, + "num_input_tokens_seen": 289113315, + "step": 8893 + }, + { + "epoch": 0.8020922577445101, + "flos": 17936821263840.0, + "grad_norm": 1.9693920360180632, + "learning_rate": 3.968590442366888e-07, + "loss": 0.7203, + "num_input_tokens_seen": 289138520, + "step": 8894 + }, + { + "epoch": 0.8021824412679803, + "flos": 22896641733600.0, + "grad_norm": 6.486740967552732, + "learning_rate": 3.9650981292603423e-07, + "loss": 0.7196, + "num_input_tokens_seen": 289167905, + "step": 8895 + }, + { + "epoch": 0.8022726247914506, + "flos": 21366642279840.0, + "grad_norm": 2.2151575152187766, + "learning_rate": 3.961607184340041e-07, + "loss": 0.7206, + "num_input_tokens_seen": 289192955, + "step": 8896 + }, + { + "epoch": 0.8023628083149209, + "flos": 38672743924320.0, + "grad_norm": 1.5800204083778484, + "learning_rate": 3.9581176079038505e-07, + "loss": 0.6414, + "num_input_tokens_seen": 289226080, + "step": 8897 + }, + { + "epoch": 0.8024529918383911, + "flos": 41404147720800.0, + "grad_norm": 1.7310734848212175, + "learning_rate": 3.954629400249516e-07, + "loss": 0.6898, + "num_input_tokens_seen": 289259925, + "step": 8898 + }, + { + "epoch": 0.8025431753618614, + "flos": 23845325912640.0, + "grad_norm": 1.7263586692112223, + "learning_rate": 3.9511425616746787e-07, + "loss": 0.8118, + "num_input_tokens_seen": 289287725, + "step": 8899 + }, + { + "epoch": 0.8026333588853316, + "flos": 20820235143360.0, + "grad_norm": 3.768061818921741, + "learning_rate": 3.947657092476853e-07, + "loss": 0.81, + "num_input_tokens_seen": 289315845, + "step": 8900 + }, + { + "epoch": 0.802723542408802, + "flos": 18084965691840.0, + "grad_norm": 3.7081164750258857, + "learning_rate": 3.944172992953425e-07, + "loss": 0.7556, + "num_input_tokens_seen": 289342360, + "step": 8901 + }, + { + "epoch": 0.8028137259322722, + "flos": 31239666606720.0, + "grad_norm": 1.7752183887292559, + "learning_rate": 3.9406902634017e-07, + "loss": 0.6973, + "num_input_tokens_seen": 289373715, + "step": 8902 + }, + { + "epoch": 0.8029039094557424, + "flos": 16630915524960.0, + "grad_norm": 1.6440721899550796, + "learning_rate": 3.9372089041188227e-07, + "loss": 0.7781, + "num_input_tokens_seen": 289400840, + "step": 8903 + }, + { + "epoch": 0.8029940929792126, + "flos": 37944324974880.0, + "grad_norm": 2.0762403159286027, + "learning_rate": 3.9337289154018593e-07, + "loss": 0.6604, + "num_input_tokens_seen": 289433075, + "step": 8904 + }, + { + "epoch": 0.803084276502683, + "flos": 27997990414080.0, + "grad_norm": 1.9951611633981172, + "learning_rate": 3.930250297547728e-07, + "loss": 0.7457, + "num_input_tokens_seen": 289462750, + "step": 8905 + }, + { + "epoch": 0.8031744600261532, + "flos": 24094143251520.0, + "grad_norm": 4.260520927564932, + "learning_rate": 3.9267730508532513e-07, + "loss": 0.6257, + "num_input_tokens_seen": 289492730, + "step": 8906 + }, + { + "epoch": 0.8032646435496235, + "flos": 20962878446880.0, + "grad_norm": 2.352055596142181, + "learning_rate": 3.923297175615121e-07, + "loss": 0.7359, + "num_input_tokens_seen": 289520135, + "step": 8907 + }, + { + "epoch": 0.8033548270730937, + "flos": 28285470036960.0, + "grad_norm": 2.270177857409516, + "learning_rate": 3.9198226721299243e-07, + "loss": 0.7856, + "num_input_tokens_seen": 289552015, + "step": 8908 + }, + { + "epoch": 0.803445010596564, + "flos": 16266185673600.0, + "grad_norm": 3.3589771123834877, + "learning_rate": 3.916349540694128e-07, + "loss": 0.7243, + "num_input_tokens_seen": 289578665, + "step": 8909 + }, + { + "epoch": 0.8035351941200343, + "flos": 28690423302240.0, + "grad_norm": 2.10542866344898, + "learning_rate": 3.912877781604063e-07, + "loss": 0.7551, + "num_input_tokens_seen": 289607535, + "step": 8910 + }, + { + "epoch": 0.8036253776435045, + "flos": 23581417651200.0, + "grad_norm": 4.074652702650472, + "learning_rate": 3.909407395155977e-07, + "loss": 0.7736, + "num_input_tokens_seen": 289633635, + "step": 8911 + }, + { + "epoch": 0.8037155611669748, + "flos": 20783617196160.0, + "grad_norm": 2.4165606859141886, + "learning_rate": 3.9059383816459725e-07, + "loss": 0.6875, + "num_input_tokens_seen": 289661145, + "step": 8912 + }, + { + "epoch": 0.8038057446904451, + "flos": 25224579055200.0, + "grad_norm": 2.465062634985788, + "learning_rate": 3.902470741370045e-07, + "loss": 0.8052, + "num_input_tokens_seen": 289690900, + "step": 8913 + }, + { + "epoch": 0.8038959282139153, + "flos": 26758667182560.0, + "grad_norm": 1.4913289261401668, + "learning_rate": 3.8990044746240746e-07, + "loss": 0.7329, + "num_input_tokens_seen": 289722585, + "step": 8914 + }, + { + "epoch": 0.8039861117373855, + "flos": 25046358557760.0, + "grad_norm": 2.327240231744983, + "learning_rate": 3.8955395817038237e-07, + "loss": 0.8422, + "num_input_tokens_seen": 289748775, + "step": 8915 + }, + { + "epoch": 0.8040762952608559, + "flos": 25192793176800.0, + "grad_norm": 1.7143836305674318, + "learning_rate": 3.892076062904934e-07, + "loss": 0.712, + "num_input_tokens_seen": 289779190, + "step": 8916 + }, + { + "epoch": 0.8041664787843261, + "flos": 15209840668800.0, + "grad_norm": 2.1936506721637277, + "learning_rate": 3.8886139185229384e-07, + "loss": 0.7041, + "num_input_tokens_seen": 289806325, + "step": 8917 + }, + { + "epoch": 0.8042566623077964, + "flos": 20086761106560.0, + "grad_norm": 1.8587873155225516, + "learning_rate": 3.8851531488532284e-07, + "loss": 0.6618, + "num_input_tokens_seen": 289833515, + "step": 8918 + }, + { + "epoch": 0.8043468458312666, + "flos": 21511069731840.0, + "grad_norm": 1.7795201652436847, + "learning_rate": 3.88169375419112e-07, + "loss": 0.7673, + "num_input_tokens_seen": 289863215, + "step": 8919 + }, + { + "epoch": 0.8044370293547369, + "flos": 43079541040320.0, + "grad_norm": 3.732278147374946, + "learning_rate": 3.8782357348317717e-07, + "loss": 0.7182, + "num_input_tokens_seen": 289895160, + "step": 8920 + }, + { + "epoch": 0.8045272128782072, + "flos": 13789434868320.0, + "grad_norm": 2.7262496987420937, + "learning_rate": 3.8747790910702437e-07, + "loss": 0.7904, + "num_input_tokens_seen": 289919850, + "step": 8921 + }, + { + "epoch": 0.8046173964016774, + "flos": 21727134778560.0, + "grad_norm": 1.987322083193213, + "learning_rate": 3.8713238232014776e-07, + "loss": 0.7444, + "num_input_tokens_seen": 289950020, + "step": 8922 + }, + { + "epoch": 0.8047075799251476, + "flos": 19363508753760.0, + "grad_norm": 1.8728362632299222, + "learning_rate": 3.867869931520296e-07, + "loss": 0.8035, + "num_input_tokens_seen": 289977200, + "step": 8923 + }, + { + "epoch": 0.804797763448618, + "flos": 27701106841920.0, + "grad_norm": 1.749899520714211, + "learning_rate": 3.864417416321406e-07, + "loss": 0.757, + "num_input_tokens_seen": 290005290, + "step": 8924 + }, + { + "epoch": 0.8048879469720882, + "flos": 23297097457920.0, + "grad_norm": 1.7547439579700015, + "learning_rate": 3.8609662778993847e-07, + "loss": 0.7267, + "num_input_tokens_seen": 290033680, + "step": 8925 + }, + { + "epoch": 0.8049781304955584, + "flos": 68111880864960.0, + "grad_norm": 0.5902273181465653, + "learning_rate": 3.85751651654872e-07, + "loss": 0.5568, + "num_input_tokens_seen": 290135140, + "step": 8926 + }, + { + "epoch": 0.8050683140190287, + "flos": 15865581270240.0, + "grad_norm": 1.978794571548644, + "learning_rate": 3.8540681325637505e-07, + "loss": 0.8564, + "num_input_tokens_seen": 290158180, + "step": 8927 + }, + { + "epoch": 0.805158497542499, + "flos": 24790925001600.0, + "grad_norm": 2.955999680742342, + "learning_rate": 3.8506211262387155e-07, + "loss": 0.7554, + "num_input_tokens_seen": 290190830, + "step": 8928 + }, + { + "epoch": 0.8052486810659693, + "flos": 26613124637760.0, + "grad_norm": 3.8983232561698533, + "learning_rate": 3.847175497867732e-07, + "loss": 0.753, + "num_input_tokens_seen": 290220255, + "step": 8929 + }, + { + "epoch": 0.8053388645894395, + "flos": 20893694056320.0, + "grad_norm": 1.8945788190709731, + "learning_rate": 3.843731247744801e-07, + "loss": 0.6943, + "num_input_tokens_seen": 290249305, + "step": 8930 + }, + { + "epoch": 0.8054290481129097, + "flos": 24277641854880.0, + "grad_norm": 1.8724736041460976, + "learning_rate": 3.8402883761638047e-07, + "loss": 0.7858, + "num_input_tokens_seen": 290276025, + "step": 8931 + }, + { + "epoch": 0.8055192316363801, + "flos": 66110085450240.0, + "grad_norm": 0.6627996666658216, + "learning_rate": 3.8368468834185076e-07, + "loss": 0.6134, + "num_input_tokens_seen": 290373585, + "step": 8932 + }, + { + "epoch": 0.8056094151598503, + "flos": 21039013582560.0, + "grad_norm": 1.8118338840710584, + "learning_rate": 3.8334067698025583e-07, + "loss": 0.7927, + "num_input_tokens_seen": 290402760, + "step": 8933 + }, + { + "epoch": 0.8056995986833205, + "flos": 21257866361280.0, + "grad_norm": 1.8109360886602028, + "learning_rate": 3.8299680356094897e-07, + "loss": 0.6779, + "num_input_tokens_seen": 290430545, + "step": 8934 + }, + { + "epoch": 0.8057897822067909, + "flos": 18525979357920.0, + "grad_norm": 3.0598221573854025, + "learning_rate": 3.8265306811327024e-07, + "loss": 0.6339, + "num_input_tokens_seen": 290459550, + "step": 8935 + }, + { + "epoch": 0.8058799657302611, + "flos": 14042340880800.0, + "grad_norm": 2.1850369346866483, + "learning_rate": 3.8230947066654994e-07, + "loss": 0.718, + "num_input_tokens_seen": 290483220, + "step": 8936 + }, + { + "epoch": 0.8059701492537313, + "flos": 20091630345120.0, + "grad_norm": 1.8675302486191345, + "learning_rate": 3.819660112501053e-07, + "loss": 0.8042, + "num_input_tokens_seen": 290512290, + "step": 8937 + }, + { + "epoch": 0.8060603327772016, + "flos": 67378258149120.0, + "grad_norm": 0.6216740566313849, + "learning_rate": 3.816226898932422e-07, + "loss": 0.5969, + "num_input_tokens_seen": 290608130, + "step": 8938 + }, + { + "epoch": 0.8061505163006719, + "flos": 24282659772480.0, + "grad_norm": 1.786206827356055, + "learning_rate": 3.812795066252557e-07, + "loss": 0.7335, + "num_input_tokens_seen": 290638980, + "step": 8939 + }, + { + "epoch": 0.8062406998241421, + "flos": 13169271460800.0, + "grad_norm": 2.295825432996317, + "learning_rate": 3.8093646147542577e-07, + "loss": 0.7604, + "num_input_tokens_seen": 290663585, + "step": 8940 + }, + { + "epoch": 0.8063308833476124, + "flos": 18954875682240.0, + "grad_norm": 2.1561400896891523, + "learning_rate": 3.805935544730259e-07, + "loss": 0.7151, + "num_input_tokens_seen": 290689675, + "step": 8941 + }, + { + "epoch": 0.8064210668710826, + "flos": 25921546654080.0, + "grad_norm": 2.5879924250319406, + "learning_rate": 3.802507856473118e-07, + "loss": 0.6837, + "num_input_tokens_seen": 290719905, + "step": 8942 + }, + { + "epoch": 0.806511250394553, + "flos": 19216999795200.0, + "grad_norm": 2.423436209249737, + "learning_rate": 3.7990815502753317e-07, + "loss": 0.7501, + "num_input_tokens_seen": 290744980, + "step": 8943 + }, + { + "epoch": 0.8066014339180232, + "flos": 19866904744320.0, + "grad_norm": 1.8049371059587298, + "learning_rate": 3.795656626429231e-07, + "loss": 0.7517, + "num_input_tokens_seen": 290772435, + "step": 8944 + }, + { + "epoch": 0.8066916174414934, + "flos": 25775483732640.0, + "grad_norm": 17.55103468059291, + "learning_rate": 3.792233085227059e-07, + "loss": 0.6649, + "num_input_tokens_seen": 290800985, + "step": 8945 + }, + { + "epoch": 0.8067818009649637, + "flos": 22125992203200.0, + "grad_norm": 2.3746230177590184, + "learning_rate": 3.788810926960928e-07, + "loss": 0.7567, + "num_input_tokens_seen": 290827075, + "step": 8946 + }, + { + "epoch": 0.806871984488434, + "flos": 24496308784800.0, + "grad_norm": 1.8625625087822137, + "learning_rate": 3.785390151922836e-07, + "loss": 0.7325, + "num_input_tokens_seen": 290855910, + "step": 8947 + }, + { + "epoch": 0.8069621680119042, + "flos": 20492643615840.0, + "grad_norm": 2.2292427929390297, + "learning_rate": 3.781970760404665e-07, + "loss": 0.7293, + "num_input_tokens_seen": 290881945, + "step": 8948 + }, + { + "epoch": 0.8070523515353745, + "flos": 29600816894880.0, + "grad_norm": 2.098781063901882, + "learning_rate": 3.778552752698176e-07, + "loss": 0.823, + "num_input_tokens_seen": 290910410, + "step": 8949 + }, + { + "epoch": 0.8071425350588447, + "flos": 39692619660960.0, + "grad_norm": 1.8222856428512644, + "learning_rate": 3.775136129095007e-07, + "loss": 0.6592, + "num_input_tokens_seen": 290943535, + "step": 8950 + }, + { + "epoch": 0.807232718582315, + "flos": 23371225426560.0, + "grad_norm": 2.567707074663474, + "learning_rate": 3.771720889886685e-07, + "loss": 0.7178, + "num_input_tokens_seen": 290973790, + "step": 8951 + }, + { + "epoch": 0.8073229021057853, + "flos": 61910804336160.0, + "grad_norm": 0.6916243430099525, + "learning_rate": 3.7683070353646194e-07, + "loss": 0.5954, + "num_input_tokens_seen": 291057440, + "step": 8952 + }, + { + "epoch": 0.8074130856292555, + "flos": 26285495940480.0, + "grad_norm": 1.9643439990227225, + "learning_rate": 3.7648945658200983e-07, + "loss": 0.7068, + "num_input_tokens_seen": 291085630, + "step": 8953 + }, + { + "epoch": 0.8075032691527257, + "flos": 39473766882240.0, + "grad_norm": 1.8746958803859781, + "learning_rate": 3.761483481544292e-07, + "loss": 0.634, + "num_input_tokens_seen": 291117695, + "step": 8954 + }, + { + "epoch": 0.8075934526761961, + "flos": 20417883761280.0, + "grad_norm": 1.9112208857292898, + "learning_rate": 3.7580737828282525e-07, + "loss": 0.7146, + "num_input_tokens_seen": 291145995, + "step": 8955 + }, + { + "epoch": 0.8076836361996663, + "flos": 20419853758560.0, + "grad_norm": 1.7081921294246858, + "learning_rate": 3.754665469962921e-07, + "loss": 0.7246, + "num_input_tokens_seen": 291174250, + "step": 8956 + }, + { + "epoch": 0.8077738197231366, + "flos": 21147752331360.0, + "grad_norm": 2.1526913232431872, + "learning_rate": 3.7512585432390973e-07, + "loss": 0.7013, + "num_input_tokens_seen": 291202510, + "step": 8957 + }, + { + "epoch": 0.8078640032466069, + "flos": 14772581148480.0, + "grad_norm": 1.8130700760484222, + "learning_rate": 3.7478530029474987e-07, + "loss": 0.6831, + "num_input_tokens_seen": 291228070, + "step": 8958 + }, + { + "epoch": 0.8079541867700771, + "flos": 70567476343200.0, + "grad_norm": 0.6191898651278103, + "learning_rate": 3.7444488493786854e-07, + "loss": 0.6273, + "num_input_tokens_seen": 291317280, + "step": 8959 + }, + { + "epoch": 0.8080443702935474, + "flos": 19035136661280.0, + "grad_norm": 2.0347063442242868, + "learning_rate": 3.7410460828231405e-07, + "loss": 0.7947, + "num_input_tokens_seen": 291344460, + "step": 8960 + }, + { + "epoch": 0.8081345538170176, + "flos": 31936002319680.0, + "grad_norm": 1.4849482985207887, + "learning_rate": 3.737644703571188e-07, + "loss": 0.6805, + "num_input_tokens_seen": 291378930, + "step": 8961 + }, + { + "epoch": 0.8082247373404879, + "flos": 23079062413920.0, + "grad_norm": 4.971991099711007, + "learning_rate": 3.734244711913059e-07, + "loss": 0.8158, + "num_input_tokens_seen": 291404700, + "step": 8962 + }, + { + "epoch": 0.8083149208639582, + "flos": 25520198855520.0, + "grad_norm": 2.1285333989367903, + "learning_rate": 3.7308461081388584e-07, + "loss": 0.7208, + "num_input_tokens_seen": 291433930, + "step": 8963 + }, + { + "epoch": 0.8084051043874284, + "flos": 24791519717760.0, + "grad_norm": 1.6093748518564102, + "learning_rate": 3.727448892538576e-07, + "loss": 0.6744, + "num_input_tokens_seen": 291465505, + "step": 8964 + }, + { + "epoch": 0.8084952879108986, + "flos": 25739237483040.0, + "grad_norm": 1.9961279700046508, + "learning_rate": 3.724053065402086e-07, + "loss": 0.8307, + "num_input_tokens_seen": 291496870, + "step": 8965 + }, + { + "epoch": 0.808585471434369, + "flos": 22751582395680.0, + "grad_norm": 1.6499276318719038, + "learning_rate": 3.7206586270191285e-07, + "loss": 0.7111, + "num_input_tokens_seen": 291528210, + "step": 8966 + }, + { + "epoch": 0.8086756549578392, + "flos": 24245261260320.0, + "grad_norm": 1.9329734963397402, + "learning_rate": 3.7172655776793385e-07, + "loss": 0.7908, + "num_input_tokens_seen": 291557155, + "step": 8967 + }, + { + "epoch": 0.8087658384813095, + "flos": 22715224636800.0, + "grad_norm": 2.1567521547161044, + "learning_rate": 3.7138739176722323e-07, + "loss": 0.6377, + "num_input_tokens_seen": 291585125, + "step": 8968 + }, + { + "epoch": 0.8088560220047797, + "flos": 25192309969920.0, + "grad_norm": 2.131006279487255, + "learning_rate": 3.710483647287206e-07, + "loss": 0.7124, + "num_input_tokens_seen": 291614090, + "step": 8969 + }, + { + "epoch": 0.80894620552825, + "flos": 17906633685120.0, + "grad_norm": 1.7117180981410642, + "learning_rate": 3.707094766813532e-07, + "loss": 0.6423, + "num_input_tokens_seen": 291640855, + "step": 8970 + }, + { + "epoch": 0.8090363890517203, + "flos": 21221731620960.0, + "grad_norm": 2.1215179187845954, + "learning_rate": 3.7037072765403754e-07, + "loss": 0.7092, + "num_input_tokens_seen": 291669665, + "step": 8971 + }, + { + "epoch": 0.8091265725751905, + "flos": 22785970157280.0, + "grad_norm": 4.269509806670762, + "learning_rate": 3.700321176756762e-07, + "loss": 0.698, + "num_input_tokens_seen": 291697925, + "step": 8972 + }, + { + "epoch": 0.8092167560986607, + "flos": 21658433594880.0, + "grad_norm": 1.8528663133053311, + "learning_rate": 3.69693646775163e-07, + "loss": 0.7879, + "num_input_tokens_seen": 291726095, + "step": 8973 + }, + { + "epoch": 0.8093069396221311, + "flos": 22824706780800.0, + "grad_norm": 2.4505488968045266, + "learning_rate": 3.693553149813764e-07, + "loss": 0.727, + "num_input_tokens_seen": 291754030, + "step": 8974 + }, + { + "epoch": 0.8093971231456013, + "flos": 60679807130880.0, + "grad_norm": 0.5775547565989227, + "learning_rate": 3.690171223231866e-07, + "loss": 0.5509, + "num_input_tokens_seen": 291847200, + "step": 8975 + }, + { + "epoch": 0.8094873066690715, + "flos": 26283154245600.0, + "grad_norm": 2.112992074752077, + "learning_rate": 3.6867906882944854e-07, + "loss": 0.8344, + "num_input_tokens_seen": 291877430, + "step": 8976 + }, + { + "epoch": 0.8095774901925418, + "flos": 21580328461920.0, + "grad_norm": 2.551536086179925, + "learning_rate": 3.6834115452900737e-07, + "loss": 0.7447, + "num_input_tokens_seen": 291906680, + "step": 8977 + }, + { + "epoch": 0.8096676737160121, + "flos": 20746999248960.0, + "grad_norm": 1.8693026130662496, + "learning_rate": 3.680033794506958e-07, + "loss": 0.6446, + "num_input_tokens_seen": 291934735, + "step": 8978 + }, + { + "epoch": 0.8097578572394823, + "flos": 15646728491520.0, + "grad_norm": 2.121978519032555, + "learning_rate": 3.676657436233346e-07, + "loss": 0.5761, + "num_input_tokens_seen": 291962690, + "step": 8979 + }, + { + "epoch": 0.8098480407629526, + "flos": 18007418105280.0, + "grad_norm": 1.9030249157475723, + "learning_rate": 3.6732824707573305e-07, + "loss": 0.7398, + "num_input_tokens_seen": 291987890, + "step": 8980 + }, + { + "epoch": 0.8099382242864229, + "flos": 24135890625600.0, + "grad_norm": 2.364458279656842, + "learning_rate": 3.6699088983668716e-07, + "loss": 0.7803, + "num_input_tokens_seen": 292015195, + "step": 8981 + }, + { + "epoch": 0.8100284078098932, + "flos": 25698010485600.0, + "grad_norm": 2.365462074061223, + "learning_rate": 3.6665367193498376e-07, + "loss": 0.6986, + "num_input_tokens_seen": 292044210, + "step": 8982 + }, + { + "epoch": 0.8101185913333634, + "flos": 18124594389600.0, + "grad_norm": 2.495364469155426, + "learning_rate": 3.663165933993948e-07, + "loss": 0.7476, + "num_input_tokens_seen": 292070805, + "step": 8983 + }, + { + "epoch": 0.8102087748568336, + "flos": 23475987011040.0, + "grad_norm": 3.370156096966418, + "learning_rate": 3.659796542586822e-07, + "loss": 0.7689, + "num_input_tokens_seen": 292098950, + "step": 8984 + }, + { + "epoch": 0.810298958380304, + "flos": 24240020324160.0, + "grad_norm": 1.7881254049098687, + "learning_rate": 3.6564285454159526e-07, + "loss": 0.702, + "num_input_tokens_seen": 292127125, + "step": 8985 + }, + { + "epoch": 0.8103891419037742, + "flos": 20963882030400.0, + "grad_norm": 1.6819083502449403, + "learning_rate": 3.653061942768718e-07, + "loss": 0.7902, + "num_input_tokens_seen": 292156695, + "step": 8986 + }, + { + "epoch": 0.8104793254272444, + "flos": 54119098741440.0, + "grad_norm": 3.4256987088741377, + "learning_rate": 3.649696734932375e-07, + "loss": 0.6975, + "num_input_tokens_seen": 292189015, + "step": 8987 + }, + { + "epoch": 0.8105695089507147, + "flos": 26319995211360.0, + "grad_norm": 2.249107077695977, + "learning_rate": 3.646332922194064e-07, + "loss": 0.6743, + "num_input_tokens_seen": 292219705, + "step": 8988 + }, + { + "epoch": 0.810659692474185, + "flos": 23587513491840.0, + "grad_norm": 1.9088211230184446, + "learning_rate": 3.6429705048407943e-07, + "loss": 0.7289, + "num_input_tokens_seen": 292245060, + "step": 8989 + }, + { + "epoch": 0.8107498759976552, + "flos": 34117765210560.0, + "grad_norm": 2.7288692302579083, + "learning_rate": 3.6396094831594804e-07, + "loss": 0.7353, + "num_input_tokens_seen": 292273760, + "step": 8990 + }, + { + "epoch": 0.8108400595211255, + "flos": 29673606752160.0, + "grad_norm": 2.0981057252018442, + "learning_rate": 3.6362498574368926e-07, + "loss": 0.696, + "num_input_tokens_seen": 292301710, + "step": 8991 + }, + { + "epoch": 0.8109302430445957, + "flos": 62241220765440.0, + "grad_norm": 0.7054061338025508, + "learning_rate": 3.6328916279596935e-07, + "loss": 0.6223, + "num_input_tokens_seen": 292389975, + "step": 8992 + }, + { + "epoch": 0.811020426568066, + "flos": 13971223662720.0, + "grad_norm": 1.9784764006361424, + "learning_rate": 3.6295347950144305e-07, + "loss": 0.7848, + "num_input_tokens_seen": 292417010, + "step": 8993 + }, + { + "epoch": 0.8111106100915363, + "flos": 31671499342080.0, + "grad_norm": 1.9855637481530872, + "learning_rate": 3.626179358887522e-07, + "loss": 0.5931, + "num_input_tokens_seen": 292446560, + "step": 8994 + }, + { + "epoch": 0.8112007936150065, + "flos": 22350383276160.0, + "grad_norm": 2.0555972561887894, + "learning_rate": 3.6228253198652816e-07, + "loss": 0.8017, + "num_input_tokens_seen": 292477230, + "step": 8995 + }, + { + "epoch": 0.8112909771384768, + "flos": 22423507661280.0, + "grad_norm": 1.8874404190461798, + "learning_rate": 3.6194726782338767e-07, + "loss": 0.6929, + "num_input_tokens_seen": 292506590, + "step": 8996 + }, + { + "epoch": 0.8113811606619471, + "flos": 17171933046240.0, + "grad_norm": 2.756166912738197, + "learning_rate": 3.6161214342793953e-07, + "loss": 0.7153, + "num_input_tokens_seen": 292534450, + "step": 8997 + }, + { + "epoch": 0.8114713441854173, + "flos": 21547873527840.0, + "grad_norm": 1.9380562506780437, + "learning_rate": 3.612771588287764e-07, + "loss": 0.7711, + "num_input_tokens_seen": 292562250, + "step": 8998 + }, + { + "epoch": 0.8115615277088876, + "flos": 15534161257440.0, + "grad_norm": 2.477920785501419, + "learning_rate": 3.609423140544827e-07, + "loss": 0.8065, + "num_input_tokens_seen": 292582425, + "step": 8999 + }, + { + "epoch": 0.8116517112323578, + "flos": 22459642401600.0, + "grad_norm": 2.701830427167021, + "learning_rate": 3.6060760913362787e-07, + "loss": 0.7588, + "num_input_tokens_seen": 292608105, + "step": 9000 + }, + { + "epoch": 0.8117418947558281, + "flos": 34263121906560.0, + "grad_norm": 2.9979703713264203, + "learning_rate": 3.6027304409477146e-07, + "loss": 0.7498, + "num_input_tokens_seen": 292638730, + "step": 9001 + }, + { + "epoch": 0.8118320782792984, + "flos": 22605296455680.0, + "grad_norm": 1.5267556075184763, + "learning_rate": 3.599386189664604e-07, + "loss": 0.7175, + "num_input_tokens_seen": 292668055, + "step": 9002 + }, + { + "epoch": 0.8119222618027686, + "flos": 17869383852000.0, + "grad_norm": 2.746185518253507, + "learning_rate": 3.5960433377722945e-07, + "loss": 0.7272, + "num_input_tokens_seen": 292694660, + "step": 9003 + }, + { + "epoch": 0.8120124453262388, + "flos": 26795024941440.0, + "grad_norm": 2.096237185671633, + "learning_rate": 3.5927018855560174e-07, + "loss": 0.8012, + "num_input_tokens_seen": 292722840, + "step": 9004 + }, + { + "epoch": 0.8121026288497092, + "flos": 34477811672160.0, + "grad_norm": 1.9865748909525867, + "learning_rate": 3.5893618333008904e-07, + "loss": 0.7356, + "num_input_tokens_seen": 292754130, + "step": 9005 + }, + { + "epoch": 0.8121928123731794, + "flos": 25331719504320.0, + "grad_norm": 2.002587313802286, + "learning_rate": 3.586023181291893e-07, + "loss": 0.8313, + "num_input_tokens_seen": 292782010, + "step": 9006 + }, + { + "epoch": 0.8122829958966497, + "flos": 69928393527840.0, + "grad_norm": 0.7490126945988479, + "learning_rate": 3.5826859298139044e-07, + "loss": 0.6096, + "num_input_tokens_seen": 292878480, + "step": 9007 + }, + { + "epoch": 0.81237317942012, + "flos": 30067817956800.0, + "grad_norm": 1.7162071143763609, + "learning_rate": 3.5793500791516773e-07, + "loss": 0.7241, + "num_input_tokens_seen": 292904895, + "step": 9008 + }, + { + "epoch": 0.8124633629435902, + "flos": 21768398945760.0, + "grad_norm": 1.7578085666188632, + "learning_rate": 3.5760156295898415e-07, + "loss": 0.7122, + "num_input_tokens_seen": 292930960, + "step": 9009 + }, + { + "epoch": 0.8125535464670605, + "flos": 28143644468160.0, + "grad_norm": 2.1708045352287426, + "learning_rate": 3.5726825814129203e-07, + "loss": 0.7382, + "num_input_tokens_seen": 292959220, + "step": 9010 + }, + { + "epoch": 0.8126437299905307, + "flos": 69362912570880.0, + "grad_norm": 0.6965447126387632, + "learning_rate": 3.5693509349052886e-07, + "loss": 0.6123, + "num_input_tokens_seen": 293044040, + "step": 9011 + }, + { + "epoch": 0.812733913514001, + "flos": 23874324059040.0, + "grad_norm": 1.6969382932882466, + "learning_rate": 3.5660206903512433e-07, + "loss": 0.8127, + "num_input_tokens_seen": 293071420, + "step": 9012 + }, + { + "epoch": 0.8128240970374713, + "flos": 17687632227360.0, + "grad_norm": 1.8459325900060146, + "learning_rate": 3.56269184803492e-07, + "loss": 0.7971, + "num_input_tokens_seen": 293097955, + "step": 9013 + }, + { + "epoch": 0.8129142805609415, + "flos": 22390681029600.0, + "grad_norm": 1.5404662880383069, + "learning_rate": 3.5593644082403727e-07, + "loss": 0.7645, + "num_input_tokens_seen": 293128770, + "step": 9014 + }, + { + "epoch": 0.8130044640844117, + "flos": 21112695514080.0, + "grad_norm": 2.149787173715796, + "learning_rate": 3.5560383712514994e-07, + "loss": 0.8457, + "num_input_tokens_seen": 293156865, + "step": 9015 + }, + { + "epoch": 0.8130946476078821, + "flos": 23476024180800.0, + "grad_norm": 1.6956669309488526, + "learning_rate": 3.5527137373521066e-07, + "loss": 0.7458, + "num_input_tokens_seen": 293185240, + "step": 9016 + }, + { + "epoch": 0.8131848311313523, + "flos": 26941236541920.0, + "grad_norm": 1.5324201200243264, + "learning_rate": 3.5493905068258645e-07, + "loss": 0.7012, + "num_input_tokens_seen": 293214045, + "step": 9017 + }, + { + "epoch": 0.8132750146548225, + "flos": 29928631440960.0, + "grad_norm": 1.6347569473382664, + "learning_rate": 3.546068679956333e-07, + "loss": 0.7575, + "num_input_tokens_seen": 293246780, + "step": 9018 + }, + { + "epoch": 0.8133651981782928, + "flos": 22532172070560.0, + "grad_norm": 1.6435913879685085, + "learning_rate": 3.5427482570269487e-07, + "loss": 0.8049, + "num_input_tokens_seen": 293277355, + "step": 9019 + }, + { + "epoch": 0.8134553817017631, + "flos": 22641914402880.0, + "grad_norm": 3.640336708802441, + "learning_rate": 3.539429238321026e-07, + "loss": 0.5895, + "num_input_tokens_seen": 293302675, + "step": 9020 + }, + { + "epoch": 0.8135455652252334, + "flos": 23480075684640.0, + "grad_norm": 1.7172676630026704, + "learning_rate": 3.536111624121769e-07, + "loss": 0.5939, + "num_input_tokens_seen": 293329965, + "step": 9021 + }, + { + "epoch": 0.8136357487487036, + "flos": 42240190326240.0, + "grad_norm": 1.9955672971445642, + "learning_rate": 3.532795414712244e-07, + "loss": 0.7065, + "num_input_tokens_seen": 293361160, + "step": 9022 + }, + { + "epoch": 0.8137259322721738, + "flos": 24865201649280.0, + "grad_norm": 1.767984012768817, + "learning_rate": 3.5294806103754124e-07, + "loss": 0.7599, + "num_input_tokens_seen": 293389685, + "step": 9023 + }, + { + "epoch": 0.8138161157956442, + "flos": 23953692963840.0, + "grad_norm": 2.4213123283182343, + "learning_rate": 3.526167211394115e-07, + "loss": 0.7563, + "num_input_tokens_seen": 293418820, + "step": 9024 + }, + { + "epoch": 0.8139062993191144, + "flos": 38237565910560.0, + "grad_norm": 1.6904337071415596, + "learning_rate": 3.522855218051066e-07, + "loss": 0.6838, + "num_input_tokens_seen": 293457085, + "step": 9025 + }, + { + "epoch": 0.8139964828425846, + "flos": 27187600676640.0, + "grad_norm": 1.8169295674866974, + "learning_rate": 3.5195446306288633e-07, + "loss": 0.7448, + "num_input_tokens_seen": 293480350, + "step": 9026 + }, + { + "epoch": 0.8140866663660549, + "flos": 34590155887680.0, + "grad_norm": 1.6141774641603763, + "learning_rate": 3.51623544940999e-07, + "loss": 0.6864, + "num_input_tokens_seen": 293514585, + "step": 9027 + }, + { + "epoch": 0.8141768498895252, + "flos": 23734951694400.0, + "grad_norm": 1.777812384559975, + "learning_rate": 3.5129276746767886e-07, + "loss": 0.7296, + "num_input_tokens_seen": 293545555, + "step": 9028 + }, + { + "epoch": 0.8142670334129954, + "flos": 22932962322720.0, + "grad_norm": 2.0047330879385745, + "learning_rate": 3.5096213067115165e-07, + "loss": 0.7167, + "num_input_tokens_seen": 293575170, + "step": 9029 + }, + { + "epoch": 0.8143572169364657, + "flos": 24639026427840.0, + "grad_norm": 2.059545539428418, + "learning_rate": 3.506316345796272e-07, + "loss": 0.7606, + "num_input_tokens_seen": 293602280, + "step": 9030 + }, + { + "epoch": 0.814447400459936, + "flos": 72160936044480.0, + "grad_norm": 0.5943065343949142, + "learning_rate": 3.5030127922130714e-07, + "loss": 0.5352, + "num_input_tokens_seen": 293701520, + "step": 9031 + }, + { + "epoch": 0.8145375839834063, + "flos": 18884687708160.0, + "grad_norm": 1.6715222280646418, + "learning_rate": 3.4997106462437784e-07, + "loss": 0.6328, + "num_input_tokens_seen": 293729055, + "step": 9032 + }, + { + "epoch": 0.8146277675068765, + "flos": 69737498142240.0, + "grad_norm": 0.7279646275279348, + "learning_rate": 3.496409908170157e-07, + "loss": 0.5606, + "num_input_tokens_seen": 293812365, + "step": 9033 + }, + { + "epoch": 0.8147179510303467, + "flos": 22641914402880.0, + "grad_norm": 1.8834342721316786, + "learning_rate": 3.493110578273839e-07, + "loss": 0.7086, + "num_input_tokens_seen": 293843230, + "step": 9034 + }, + { + "epoch": 0.8148081345538171, + "flos": 21366605110080.0, + "grad_norm": 3.0528373090751417, + "learning_rate": 3.489812656836346e-07, + "loss": 0.7976, + "num_input_tokens_seen": 293870810, + "step": 9035 + }, + { + "epoch": 0.8148983180772873, + "flos": 28690274623200.0, + "grad_norm": 1.6289684549067545, + "learning_rate": 3.486516144139078e-07, + "loss": 0.6578, + "num_input_tokens_seen": 293902085, + "step": 9036 + }, + { + "epoch": 0.8149885016007575, + "flos": 26100584886240.0, + "grad_norm": 1.870401975296844, + "learning_rate": 3.4832210404632957e-07, + "loss": 0.7134, + "num_input_tokens_seen": 293932145, + "step": 9037 + }, + { + "epoch": 0.8150786851242278, + "flos": 22857421903200.0, + "grad_norm": 1.5163706881929035, + "learning_rate": 3.479927346090179e-07, + "loss": 0.7497, + "num_input_tokens_seen": 293962735, + "step": 9038 + }, + { + "epoch": 0.8151688686476981, + "flos": 14553505351200.0, + "grad_norm": 1.9341533429126416, + "learning_rate": 3.4766350613007455e-07, + "loss": 0.7285, + "num_input_tokens_seen": 293987475, + "step": 9039 + }, + { + "epoch": 0.8152590521711683, + "flos": 24644230194240.0, + "grad_norm": 1.717462994150455, + "learning_rate": 3.4733441863759173e-07, + "loss": 0.73, + "num_input_tokens_seen": 294015565, + "step": 9040 + }, + { + "epoch": 0.8153492356946386, + "flos": 21874647320640.0, + "grad_norm": 2.170993104450199, + "learning_rate": 3.4700547215964916e-07, + "loss": 0.6323, + "num_input_tokens_seen": 294047050, + "step": 9041 + }, + { + "epoch": 0.8154394192181088, + "flos": 29417466970560.0, + "grad_norm": 1.86848589711817, + "learning_rate": 3.46676666724314e-07, + "loss": 0.7231, + "num_input_tokens_seen": 294078690, + "step": 9042 + }, + { + "epoch": 0.8155296027415792, + "flos": 19545000190080.0, + "grad_norm": 2.5131549911398157, + "learning_rate": 3.463480023596421e-07, + "loss": 0.7324, + "num_input_tokens_seen": 294107100, + "step": 9043 + }, + { + "epoch": 0.8156197862650494, + "flos": 25625555156160.0, + "grad_norm": 2.02234764647534, + "learning_rate": 3.460194790936772e-07, + "loss": 0.8142, + "num_input_tokens_seen": 294135010, + "step": 9044 + }, + { + "epoch": 0.8157099697885196, + "flos": 22494550539840.0, + "grad_norm": 1.6573471178344448, + "learning_rate": 3.456910969544495e-07, + "loss": 0.7314, + "num_input_tokens_seen": 294164590, + "step": 9045 + }, + { + "epoch": 0.8158001533119899, + "flos": 17796519655200.0, + "grad_norm": 2.199741877238566, + "learning_rate": 3.4536285596997994e-07, + "loss": 0.7737, + "num_input_tokens_seen": 294190480, + "step": 9046 + }, + { + "epoch": 0.8158903368354602, + "flos": 39801618598080.0, + "grad_norm": 1.9153390169976618, + "learning_rate": 3.450347561682747e-07, + "loss": 0.7038, + "num_input_tokens_seen": 294219685, + "step": 9047 + }, + { + "epoch": 0.8159805203589304, + "flos": 31160446380960.0, + "grad_norm": 2.0646648027787107, + "learning_rate": 3.4470679757732945e-07, + "loss": 0.686, + "num_input_tokens_seen": 294246925, + "step": 9048 + }, + { + "epoch": 0.8160707038824007, + "flos": 22605370795200.0, + "grad_norm": 13.868887989927254, + "learning_rate": 3.4437898022512735e-07, + "loss": 0.8226, + "num_input_tokens_seen": 294271970, + "step": 9049 + }, + { + "epoch": 0.8161608874058709, + "flos": 23370816559200.0, + "grad_norm": 2.0792634756704893, + "learning_rate": 3.4405130413963977e-07, + "loss": 0.7441, + "num_input_tokens_seen": 294299935, + "step": 9050 + }, + { + "epoch": 0.8162510709293412, + "flos": 21147454973280.0, + "grad_norm": 2.7027769360786564, + "learning_rate": 3.437237693488262e-07, + "loss": 0.8133, + "num_input_tokens_seen": 294329140, + "step": 9051 + }, + { + "epoch": 0.8163412544528115, + "flos": 21329838483840.0, + "grad_norm": 1.6681350517899118, + "learning_rate": 3.433963758806322e-07, + "loss": 0.758, + "num_input_tokens_seen": 294357960, + "step": 9052 + }, + { + "epoch": 0.8164314379762817, + "flos": 15789334625280.0, + "grad_norm": 1.84867904774541, + "learning_rate": 3.430691237629948e-07, + "loss": 0.6096, + "num_input_tokens_seen": 294384395, + "step": 9053 + }, + { + "epoch": 0.816521621499752, + "flos": 48762762541920.0, + "grad_norm": 1.9010031360919415, + "learning_rate": 3.427420130238354e-07, + "loss": 0.6902, + "num_input_tokens_seen": 294416805, + "step": 9054 + }, + { + "epoch": 0.8166118050232223, + "flos": 36845117503200.0, + "grad_norm": 2.214461789843473, + "learning_rate": 3.424150436910658e-07, + "loss": 0.8269, + "num_input_tokens_seen": 294447675, + "step": 9055 + }, + { + "epoch": 0.8167019885466925, + "flos": 23769042097920.0, + "grad_norm": 2.2471239994537235, + "learning_rate": 3.420882157925842e-07, + "loss": 0.7998, + "num_input_tokens_seen": 294470105, + "step": 9056 + }, + { + "epoch": 0.8167921720701627, + "flos": 18816804259200.0, + "grad_norm": 2.5680896922042677, + "learning_rate": 3.417615293562777e-07, + "loss": 0.7202, + "num_input_tokens_seen": 294498465, + "step": 9057 + }, + { + "epoch": 0.8168823555936331, + "flos": 26758109636160.0, + "grad_norm": 1.674365176705667, + "learning_rate": 3.4143498441002105e-07, + "loss": 0.6535, + "num_input_tokens_seen": 294529970, + "step": 9058 + }, + { + "epoch": 0.8169725391171033, + "flos": 25046804594880.0, + "grad_norm": 1.7302745425949244, + "learning_rate": 3.411085809816767e-07, + "loss": 0.7489, + "num_input_tokens_seen": 294558455, + "step": 9059 + }, + { + "epoch": 0.8170627226405736, + "flos": 17833397790720.0, + "grad_norm": 2.0818732167685527, + "learning_rate": 3.407823190990953e-07, + "loss": 0.8547, + "num_input_tokens_seen": 294584815, + "step": 9060 + }, + { + "epoch": 0.8171529061640438, + "flos": 22313728159200.0, + "grad_norm": 5.5912084506371755, + "learning_rate": 3.4045619879011577e-07, + "loss": 0.8239, + "num_input_tokens_seen": 294608975, + "step": 9061 + }, + { + "epoch": 0.8172430896875141, + "flos": 31636702713120.0, + "grad_norm": 1.9559730211827377, + "learning_rate": 3.4013022008256334e-07, + "loss": 0.6732, + "num_input_tokens_seen": 294636920, + "step": 9062 + }, + { + "epoch": 0.8173332732109844, + "flos": 22059669884160.0, + "grad_norm": 1.8782106884595056, + "learning_rate": 3.398043830042532e-07, + "loss": 0.6818, + "num_input_tokens_seen": 294666175, + "step": 9063 + }, + { + "epoch": 0.8174234567344546, + "flos": 19874190017280.0, + "grad_norm": 2.0432386639410205, + "learning_rate": 3.394786875829871e-07, + "loss": 0.741, + "num_input_tokens_seen": 294693820, + "step": 9064 + }, + { + "epoch": 0.8175136402579248, + "flos": 36264842981760.0, + "grad_norm": 1.8756581740733085, + "learning_rate": 3.3915313384655564e-07, + "loss": 0.6236, + "num_input_tokens_seen": 294728055, + "step": 9065 + }, + { + "epoch": 0.8176038237813952, + "flos": 26211516650880.0, + "grad_norm": 1.6951393003809223, + "learning_rate": 3.388277218227369e-07, + "loss": 0.7582, + "num_input_tokens_seen": 294758815, + "step": 9066 + }, + { + "epoch": 0.8176940073048654, + "flos": 18014963566560.0, + "grad_norm": 2.5824412887413613, + "learning_rate": 3.3850245153929557e-07, + "loss": 0.762, + "num_input_tokens_seen": 294783425, + "step": 9067 + }, + { + "epoch": 0.8177841908283356, + "flos": 22675707448320.0, + "grad_norm": 2.297810993087733, + "learning_rate": 3.381773230239875e-07, + "loss": 0.6989, + "num_input_tokens_seen": 294810895, + "step": 9068 + }, + { + "epoch": 0.8178743743518059, + "flos": 22350792143520.0, + "grad_norm": 1.686353718004735, + "learning_rate": 3.3785233630455247e-07, + "loss": 0.772, + "num_input_tokens_seen": 294839470, + "step": 9069 + }, + { + "epoch": 0.8179645578752762, + "flos": 24094514949120.0, + "grad_norm": 2.657850254694458, + "learning_rate": 3.375274914087221e-07, + "loss": 0.749, + "num_input_tokens_seen": 294865775, + "step": 9070 + }, + { + "epoch": 0.8180547413987465, + "flos": 34301821360320.0, + "grad_norm": 1.7381509453442958, + "learning_rate": 3.3720278836421234e-07, + "loss": 0.7148, + "num_input_tokens_seen": 294899165, + "step": 9071 + }, + { + "epoch": 0.8181449249222167, + "flos": 18343149810240.0, + "grad_norm": 2.752077358411665, + "learning_rate": 3.368782271987294e-07, + "loss": 0.7659, + "num_input_tokens_seen": 294926660, + "step": 9072 + }, + { + "epoch": 0.8182351084456869, + "flos": 19181125243200.0, + "grad_norm": 1.690613241383951, + "learning_rate": 3.3655380793996636e-07, + "loss": 0.7463, + "num_input_tokens_seen": 294952995, + "step": 9073 + }, + { + "epoch": 0.8183252919691573, + "flos": 21730740245280.0, + "grad_norm": 1.833497069011891, + "learning_rate": 3.362295306156047e-07, + "loss": 0.6626, + "num_input_tokens_seen": 294981255, + "step": 9074 + }, + { + "epoch": 0.8184154754926275, + "flos": 26025787861920.0, + "grad_norm": 1.8206329063717908, + "learning_rate": 3.3590539525331327e-07, + "loss": 0.7567, + "num_input_tokens_seen": 295012455, + "step": 9075 + }, + { + "epoch": 0.8185056590160977, + "flos": 20746404532800.0, + "grad_norm": 1.6631741727134677, + "learning_rate": 3.3558140188074967e-07, + "loss": 0.6926, + "num_input_tokens_seen": 295040835, + "step": 9076 + }, + { + "epoch": 0.8185958425395681, + "flos": 41368198829280.0, + "grad_norm": 2.1463744089353374, + "learning_rate": 3.3525755052555817e-07, + "loss": 0.7724, + "num_input_tokens_seen": 295070535, + "step": 9077 + }, + { + "epoch": 0.8186860260630383, + "flos": 61858792374720.0, + "grad_norm": 0.7289750437762238, + "learning_rate": 3.3493384121537147e-07, + "loss": 0.6477, + "num_input_tokens_seen": 295156840, + "step": 9078 + }, + { + "epoch": 0.8187762095865085, + "flos": 21945615859680.0, + "grad_norm": 1.8887083568816907, + "learning_rate": 3.3461027397781075e-07, + "loss": 0.7429, + "num_input_tokens_seen": 295184215, + "step": 9079 + }, + { + "epoch": 0.8188663931099788, + "flos": 21804124818720.0, + "grad_norm": 2.248192060805047, + "learning_rate": 3.3428684884048397e-07, + "loss": 0.7241, + "num_input_tokens_seen": 295212875, + "step": 9080 + }, + { + "epoch": 0.8189565766334491, + "flos": 17505137207520.0, + "grad_norm": 1.6239174391332503, + "learning_rate": 3.3396356583098826e-07, + "loss": 0.7409, + "num_input_tokens_seen": 295239660, + "step": 9081 + }, + { + "epoch": 0.8190467601569194, + "flos": 22350234597120.0, + "grad_norm": 1.8027744474741083, + "learning_rate": 3.3364042497690736e-07, + "loss": 0.775, + "num_input_tokens_seen": 295269720, + "step": 9082 + }, + { + "epoch": 0.8191369436803896, + "flos": 28911506266560.0, + "grad_norm": 1.7650111658969658, + "learning_rate": 3.3331742630581405e-07, + "loss": 0.7985, + "num_input_tokens_seen": 295303520, + "step": 9083 + }, + { + "epoch": 0.8192271272038598, + "flos": 33748909515840.0, + "grad_norm": 1.4906873901873832, + "learning_rate": 3.3299456984526717e-07, + "loss": 0.6906, + "num_input_tokens_seen": 295334395, + "step": 9084 + }, + { + "epoch": 0.8193173107273302, + "flos": 21838884277920.0, + "grad_norm": 1.9798522230353415, + "learning_rate": 3.3267185562281605e-07, + "loss": 0.777, + "num_input_tokens_seen": 295364380, + "step": 9085 + }, + { + "epoch": 0.8194074942508004, + "flos": 27706607966400.0, + "grad_norm": 2.0412890675456756, + "learning_rate": 3.3234928366599514e-07, + "loss": 0.7362, + "num_input_tokens_seen": 295395395, + "step": 9086 + }, + { + "epoch": 0.8194976777742706, + "flos": 19945716102720.0, + "grad_norm": 2.465003732365071, + "learning_rate": 3.3202685400232946e-07, + "loss": 0.7745, + "num_input_tokens_seen": 295422890, + "step": 9087 + }, + { + "epoch": 0.8195878612977409, + "flos": 18634569427680.0, + "grad_norm": 2.5778667898631413, + "learning_rate": 3.317045666593292e-07, + "loss": 0.7686, + "num_input_tokens_seen": 295449500, + "step": 9088 + }, + { + "epoch": 0.8196780448212112, + "flos": 17905927459680.0, + "grad_norm": 1.9674777809837016, + "learning_rate": 3.3138242166449426e-07, + "loss": 0.7508, + "num_input_tokens_seen": 295477565, + "step": 9089 + }, + { + "epoch": 0.8197682283446814, + "flos": 24355263780960.0, + "grad_norm": 1.5659274085212043, + "learning_rate": 3.310604190453117e-07, + "loss": 0.805, + "num_input_tokens_seen": 295506670, + "step": 9090 + }, + { + "epoch": 0.8198584118681517, + "flos": 46466239401120.0, + "grad_norm": 1.713665394437078, + "learning_rate": 3.307385588292566e-07, + "loss": 0.6921, + "num_input_tokens_seen": 295540390, + "step": 9091 + }, + { + "epoch": 0.8199485953916219, + "flos": 33134024214240.0, + "grad_norm": 2.015936893905953, + "learning_rate": 3.304168410437924e-07, + "loss": 0.7357, + "num_input_tokens_seen": 295568105, + "step": 9092 + }, + { + "epoch": 0.8200387789150922, + "flos": 69823408924800.0, + "grad_norm": 0.6252417573656257, + "learning_rate": 3.300952657163687e-07, + "loss": 0.603, + "num_input_tokens_seen": 295657610, + "step": 9093 + }, + { + "epoch": 0.8201289624385625, + "flos": 26795582487840.0, + "grad_norm": 2.833173934342247, + "learning_rate": 3.297738328744248e-07, + "loss": 0.7256, + "num_input_tokens_seen": 295686940, + "step": 9094 + }, + { + "epoch": 0.8202191459620327, + "flos": 25229150935680.0, + "grad_norm": 1.8369273013378593, + "learning_rate": 3.2945254254538714e-07, + "loss": 0.7354, + "num_input_tokens_seen": 295715135, + "step": 9095 + }, + { + "epoch": 0.820309329485503, + "flos": 26972762232000.0, + "grad_norm": 1.9187255112268824, + "learning_rate": 3.2913139475666963e-07, + "loss": 0.7102, + "num_input_tokens_seen": 295742645, + "step": 9096 + }, + { + "epoch": 0.8203995130089733, + "flos": 21949630193760.0, + "grad_norm": 2.3945767521229846, + "learning_rate": 3.288103895356749e-07, + "loss": 0.6909, + "num_input_tokens_seen": 295770935, + "step": 9097 + }, + { + "epoch": 0.8204896965324435, + "flos": 25848868306080.0, + "grad_norm": 2.046676160167003, + "learning_rate": 3.284895269097927e-07, + "loss": 0.6762, + "num_input_tokens_seen": 295800385, + "step": 9098 + }, + { + "epoch": 0.8205798800559138, + "flos": 64369447734720.0, + "grad_norm": 0.6630456431847002, + "learning_rate": 3.281688069063999e-07, + "loss": 0.5807, + "num_input_tokens_seen": 295895100, + "step": 9099 + }, + { + "epoch": 0.8206700635793841, + "flos": 23698854123840.0, + "grad_norm": 2.048041091611017, + "learning_rate": 3.2784822955286396e-07, + "loss": 0.5675, + "num_input_tokens_seen": 295922605, + "step": 9100 + }, + { + "epoch": 0.8207602471028543, + "flos": 22747753910400.0, + "grad_norm": 1.5926223954575482, + "learning_rate": 3.275277948765365e-07, + "loss": 0.8128, + "num_input_tokens_seen": 295952605, + "step": 9101 + }, + { + "epoch": 0.8208504306263246, + "flos": 29272742160480.0, + "grad_norm": 3.252984803739762, + "learning_rate": 3.2720750290475964e-07, + "loss": 0.7855, + "num_input_tokens_seen": 295982105, + "step": 9102 + }, + { + "epoch": 0.8209406141497948, + "flos": 20856704411520.0, + "grad_norm": 2.1247501226376437, + "learning_rate": 3.268873536648622e-07, + "loss": 0.7464, + "num_input_tokens_seen": 296006335, + "step": 9103 + }, + { + "epoch": 0.8210307976732651, + "flos": 31641051575040.0, + "grad_norm": 2.42873714787199, + "learning_rate": 3.265673471841612e-07, + "loss": 0.7135, + "num_input_tokens_seen": 296039730, + "step": 9104 + }, + { + "epoch": 0.8211209811967354, + "flos": 62234790396960.0, + "grad_norm": 0.6366036433920715, + "learning_rate": 3.262474834899616e-07, + "loss": 0.5399, + "num_input_tokens_seen": 296126155, + "step": 9105 + }, + { + "epoch": 0.8212111647202056, + "flos": 63442364919840.0, + "grad_norm": 0.6730266725205869, + "learning_rate": 3.2592776260955534e-07, + "loss": 0.6162, + "num_input_tokens_seen": 296225540, + "step": 9106 + }, + { + "epoch": 0.8213013482436758, + "flos": 22569273224640.0, + "grad_norm": 4.30211550105571, + "learning_rate": 3.256081845702239e-07, + "loss": 0.7809, + "num_input_tokens_seen": 296251975, + "step": 9107 + }, + { + "epoch": 0.8213915317671462, + "flos": 18451665540480.0, + "grad_norm": 2.772393946360643, + "learning_rate": 3.2528874939923335e-07, + "loss": 0.7922, + "num_input_tokens_seen": 296279395, + "step": 9108 + }, + { + "epoch": 0.8214817152906164, + "flos": 22169189197920.0, + "grad_norm": 2.0889942189415556, + "learning_rate": 3.2496945712384217e-07, + "loss": 0.6955, + "num_input_tokens_seen": 296307910, + "step": 9109 + }, + { + "epoch": 0.8215718988140867, + "flos": 26351074864320.0, + "grad_norm": 1.972328929372494, + "learning_rate": 3.246503077712923e-07, + "loss": 0.7221, + "num_input_tokens_seen": 296332530, + "step": 9110 + }, + { + "epoch": 0.8216620823375569, + "flos": 23293826519040.0, + "grad_norm": 2.477014542488835, + "learning_rate": 3.2433130136881625e-07, + "loss": 0.8329, + "num_input_tokens_seen": 296360915, + "step": 9111 + }, + { + "epoch": 0.8217522658610272, + "flos": 15822830312640.0, + "grad_norm": 2.379466524841351, + "learning_rate": 3.2401243794363287e-07, + "loss": 0.7348, + "num_input_tokens_seen": 296383060, + "step": 9112 + }, + { + "epoch": 0.8218424493844975, + "flos": 26649928433760.0, + "grad_norm": 1.7354116692381394, + "learning_rate": 3.236937175229495e-07, + "loss": 0.7961, + "num_input_tokens_seen": 296412590, + "step": 9113 + }, + { + "epoch": 0.8219326329079677, + "flos": 17282604622560.0, + "grad_norm": 1.9571587983557188, + "learning_rate": 3.233751401339615e-07, + "loss": 0.7098, + "num_input_tokens_seen": 296438920, + "step": 9114 + }, + { + "epoch": 0.8220228164314379, + "flos": 17177248321920.0, + "grad_norm": 3.1756551231873056, + "learning_rate": 3.2305670580385157e-07, + "loss": 0.8248, + "num_input_tokens_seen": 296465880, + "step": 9115 + }, + { + "epoch": 0.8221129999549083, + "flos": 27888136572480.0, + "grad_norm": 1.7577128742384376, + "learning_rate": 3.227384145597898e-07, + "loss": 0.7425, + "num_input_tokens_seen": 296495345, + "step": 9116 + }, + { + "epoch": 0.8222031834783785, + "flos": 23480596061280.0, + "grad_norm": 2.223092785490059, + "learning_rate": 3.224202664289346e-07, + "loss": 0.7293, + "num_input_tokens_seen": 296523735, + "step": 9117 + }, + { + "epoch": 0.8222933670018487, + "flos": 33134916288480.0, + "grad_norm": 1.891425034415157, + "learning_rate": 3.2210226143843257e-07, + "loss": 0.6841, + "num_input_tokens_seen": 296552440, + "step": 9118 + }, + { + "epoch": 0.822383550525319, + "flos": 59620042508160.0, + "grad_norm": 0.6550763852955619, + "learning_rate": 3.217843996154173e-07, + "loss": 0.6124, + "num_input_tokens_seen": 296640285, + "step": 9119 + }, + { + "epoch": 0.8224737340487893, + "flos": 66370648433280.0, + "grad_norm": 1.4343769314489057, + "learning_rate": 3.2146668098701055e-07, + "loss": 0.5282, + "num_input_tokens_seen": 296734555, + "step": 9120 + }, + { + "epoch": 0.8225639175722596, + "flos": 24347123603520.0, + "grad_norm": 2.0115035154884184, + "learning_rate": 3.2114910558032215e-07, + "loss": 0.8704, + "num_input_tokens_seen": 296761665, + "step": 9121 + }, + { + "epoch": 0.8226541010957298, + "flos": 17614953879360.0, + "grad_norm": 2.330266901700404, + "learning_rate": 3.2083167342244945e-07, + "loss": 0.7769, + "num_input_tokens_seen": 296789195, + "step": 9122 + }, + { + "epoch": 0.8227442846192, + "flos": 14371493538240.0, + "grad_norm": 2.085547249967774, + "learning_rate": 3.205143845404763e-07, + "loss": 0.8419, + "num_input_tokens_seen": 296813080, + "step": 9123 + }, + { + "epoch": 0.8228344681426704, + "flos": 20164606051200.0, + "grad_norm": 2.251237634839842, + "learning_rate": 3.201972389614773e-07, + "loss": 0.7171, + "num_input_tokens_seen": 296842165, + "step": 9124 + }, + { + "epoch": 0.8229246516661406, + "flos": 13351580631840.0, + "grad_norm": 2.9635816056298396, + "learning_rate": 3.198802367125115e-07, + "loss": 0.736, + "num_input_tokens_seen": 296865250, + "step": 9125 + }, + { + "epoch": 0.8230148351896108, + "flos": 25447371828480.0, + "grad_norm": 1.76819059604232, + "learning_rate": 3.195633778206288e-07, + "loss": 0.7554, + "num_input_tokens_seen": 296894275, + "step": 9126 + }, + { + "epoch": 0.8231050187130812, + "flos": 20383198641600.0, + "grad_norm": 2.1647197950841695, + "learning_rate": 3.19246662312864e-07, + "loss": 0.8188, + "num_input_tokens_seen": 296920380, + "step": 9127 + }, + { + "epoch": 0.8231952022365514, + "flos": 70075645881600.0, + "grad_norm": 0.6911355259080489, + "learning_rate": 3.189300902162417e-07, + "loss": 0.577, + "num_input_tokens_seen": 297015365, + "step": 9128 + }, + { + "epoch": 0.8232853857600216, + "flos": 19253729251680.0, + "grad_norm": 2.2952173180837248, + "learning_rate": 3.1861366155777327e-07, + "loss": 0.754, + "num_input_tokens_seen": 297043270, + "step": 9129 + }, + { + "epoch": 0.8233755692834919, + "flos": 26941087862880.0, + "grad_norm": 1.7123579982718582, + "learning_rate": 3.182973763644583e-07, + "loss": 0.753, + "num_input_tokens_seen": 297077550, + "step": 9130 + }, + { + "epoch": 0.8234657528069622, + "flos": 13606716829920.0, + "grad_norm": 2.557313928034604, + "learning_rate": 3.1798123466328463e-07, + "loss": 0.6688, + "num_input_tokens_seen": 297105610, + "step": 9131 + }, + { + "epoch": 0.8235559363304324, + "flos": 31052376687840.0, + "grad_norm": 2.596048554570343, + "learning_rate": 3.17665236481226e-07, + "loss": 0.7449, + "num_input_tokens_seen": 297134185, + "step": 9132 + }, + { + "epoch": 0.8236461198539027, + "flos": 21251807690400.0, + "grad_norm": 2.1951529556891587, + "learning_rate": 3.1734938184524576e-07, + "loss": 0.7028, + "num_input_tokens_seen": 297161250, + "step": 9133 + }, + { + "epoch": 0.8237363033773729, + "flos": 20159699642880.0, + "grad_norm": 1.796401517101489, + "learning_rate": 3.1703367078229427e-07, + "loss": 0.7506, + "num_input_tokens_seen": 297190745, + "step": 9134 + }, + { + "epoch": 0.8238264869008433, + "flos": 35175225308160.0, + "grad_norm": 2.5479450435311075, + "learning_rate": 3.167181033193096e-07, + "loss": 0.7061, + "num_input_tokens_seen": 297219795, + "step": 9135 + }, + { + "epoch": 0.8239166704243135, + "flos": 22782104502240.0, + "grad_norm": 1.8862148583477458, + "learning_rate": 3.16402679483218e-07, + "loss": 0.7557, + "num_input_tokens_seen": 297248255, + "step": 9136 + }, + { + "epoch": 0.8240068539477837, + "flos": 25160598431040.0, + "grad_norm": 1.5061888971241921, + "learning_rate": 3.1608739930093366e-07, + "loss": 0.6758, + "num_input_tokens_seen": 297282130, + "step": 9137 + }, + { + "epoch": 0.824097037471254, + "flos": 33387116075520.0, + "grad_norm": 1.9951039825913657, + "learning_rate": 3.157722627993562e-07, + "loss": 0.657, + "num_input_tokens_seen": 297314085, + "step": 9138 + }, + { + "epoch": 0.8241872209947243, + "flos": 32147681334720.0, + "grad_norm": 1.8713526536691398, + "learning_rate": 3.1545727000537727e-07, + "loss": 0.7057, + "num_input_tokens_seen": 297345250, + "step": 9139 + }, + { + "epoch": 0.8242774045181945, + "flos": 22639163840640.0, + "grad_norm": 1.9696036410618027, + "learning_rate": 3.151424209458713e-07, + "loss": 0.7419, + "num_input_tokens_seen": 297372820, + "step": 9140 + }, + { + "epoch": 0.8243675880416648, + "flos": 12987854364000.0, + "grad_norm": 2.1691048059815543, + "learning_rate": 3.148277156477053e-07, + "loss": 0.7571, + "num_input_tokens_seen": 297399935, + "step": 9141 + }, + { + "epoch": 0.824457771565135, + "flos": 17249852330400.0, + "grad_norm": 2.619245306621227, + "learning_rate": 3.145131541377299e-07, + "loss": 0.7453, + "num_input_tokens_seen": 297426140, + "step": 9142 + }, + { + "epoch": 0.8245479550886053, + "flos": 12185790652800.0, + "grad_norm": 2.092484504523836, + "learning_rate": 3.1419873644278606e-07, + "loss": 0.7101, + "num_input_tokens_seen": 297451325, + "step": 9143 + }, + { + "epoch": 0.8246381386120756, + "flos": 25225359620160.0, + "grad_norm": 1.789729098274021, + "learning_rate": 3.1388446258970147e-07, + "loss": 0.7625, + "num_input_tokens_seen": 297479955, + "step": 9144 + }, + { + "epoch": 0.8247283221355458, + "flos": 32799407602080.0, + "grad_norm": 1.8814947811827814, + "learning_rate": 3.1357033260529145e-07, + "loss": 0.7282, + "num_input_tokens_seen": 297509405, + "step": 9145 + }, + { + "epoch": 0.824818505659016, + "flos": 24460582911840.0, + "grad_norm": 2.1988289794269638, + "learning_rate": 3.1325634651636025e-07, + "loss": 0.8174, + "num_input_tokens_seen": 297539520, + "step": 9146 + }, + { + "epoch": 0.8249086891824864, + "flos": 21366530770560.0, + "grad_norm": 2.114782790625524, + "learning_rate": 3.1294250434969694e-07, + "loss": 0.7823, + "num_input_tokens_seen": 297568035, + "step": 9147 + }, + { + "epoch": 0.8249988727059566, + "flos": 17213680420320.0, + "grad_norm": 1.985514396298278, + "learning_rate": 3.1262880613208274e-07, + "loss": 0.7827, + "num_input_tokens_seen": 297591185, + "step": 9148 + }, + { + "epoch": 0.8250890562294269, + "flos": 30184399524960.0, + "grad_norm": 2.0105562573397995, + "learning_rate": 3.123152518902823e-07, + "loss": 0.7834, + "num_input_tokens_seen": 297620865, + "step": 9149 + }, + { + "epoch": 0.8251792397528972, + "flos": 26320812946080.0, + "grad_norm": 1.580047343063885, + "learning_rate": 3.1200184165105017e-07, + "loss": 0.698, + "num_input_tokens_seen": 297651030, + "step": 9150 + }, + { + "epoch": 0.8252694232763674, + "flos": 17501160043200.0, + "grad_norm": 4.7584037726556, + "learning_rate": 3.116885754411287e-07, + "loss": 0.7555, + "num_input_tokens_seen": 297676820, + "step": 9151 + }, + { + "epoch": 0.8253596067998377, + "flos": 27559095424320.0, + "grad_norm": 2.2890331279445557, + "learning_rate": 3.1137545328724703e-07, + "loss": 0.6767, + "num_input_tokens_seen": 297705655, + "step": 9152 + }, + { + "epoch": 0.8254497903233079, + "flos": 21330767727840.0, + "grad_norm": 2.089153767705577, + "learning_rate": 3.1106247521612285e-07, + "loss": 0.7455, + "num_input_tokens_seen": 297734335, + "step": 9153 + }, + { + "epoch": 0.8255399738467782, + "flos": 29921643526080.0, + "grad_norm": 2.395596987358743, + "learning_rate": 3.107496412544612e-07, + "loss": 0.7604, + "num_input_tokens_seen": 297761010, + "step": 9154 + }, + { + "epoch": 0.8256301573702485, + "flos": 19181719959360.0, + "grad_norm": 1.5546488541443637, + "learning_rate": 3.1043695142895397e-07, + "loss": 0.7706, + "num_input_tokens_seen": 297788800, + "step": 9155 + }, + { + "epoch": 0.8257203408937187, + "flos": 28579788895680.0, + "grad_norm": 1.6397005655659838, + "learning_rate": 3.101244057662828e-07, + "loss": 0.6649, + "num_input_tokens_seen": 297820505, + "step": 9156 + }, + { + "epoch": 0.8258105244171889, + "flos": 22751768244480.0, + "grad_norm": 1.9522608561966859, + "learning_rate": 3.098120042931152e-07, + "loss": 0.7309, + "num_input_tokens_seen": 297849085, + "step": 9157 + }, + { + "epoch": 0.8259007079406593, + "flos": 24937396790400.0, + "grad_norm": 2.6289551320852422, + "learning_rate": 3.0949974703610647e-07, + "loss": 0.7778, + "num_input_tokens_seen": 297877175, + "step": 9158 + }, + { + "epoch": 0.8259908914641295, + "flos": 36010450178880.0, + "grad_norm": 1.7458467244778029, + "learning_rate": 3.0918763402190107e-07, + "loss": 0.7326, + "num_input_tokens_seen": 297906040, + "step": 9159 + }, + { + "epoch": 0.8260810749875998, + "flos": 15971197759200.0, + "grad_norm": 2.963269284317474, + "learning_rate": 3.088756652771296e-07, + "loss": 0.6978, + "num_input_tokens_seen": 297930390, + "step": 9160 + }, + { + "epoch": 0.82617125851107, + "flos": 23291856521760.0, + "grad_norm": 2.6237428470421382, + "learning_rate": 3.0856384082841147e-07, + "loss": 0.8366, + "num_input_tokens_seen": 297957320, + "step": 9161 + }, + { + "epoch": 0.8262614420345403, + "flos": 29272704990720.0, + "grad_norm": 1.6948764081924208, + "learning_rate": 3.0825216070235207e-07, + "loss": 0.6913, + "num_input_tokens_seen": 297987640, + "step": 9162 + }, + { + "epoch": 0.8263516255580106, + "flos": 16885419837120.0, + "grad_norm": 2.3060364993962805, + "learning_rate": 3.0794062492554764e-07, + "loss": 0.8516, + "num_input_tokens_seen": 298013035, + "step": 9163 + }, + { + "epoch": 0.8264418090814808, + "flos": 21833234474400.0, + "grad_norm": 1.9736270744017588, + "learning_rate": 3.076292335245783e-07, + "loss": 0.7551, + "num_input_tokens_seen": 298040520, + "step": 9164 + }, + { + "epoch": 0.826531992604951, + "flos": 20820718350240.0, + "grad_norm": 2.0094515867454676, + "learning_rate": 3.073179865260145e-07, + "loss": 0.811, + "num_input_tokens_seen": 298068340, + "step": 9165 + }, + { + "epoch": 0.8266221761284214, + "flos": 24790107266880.0, + "grad_norm": 1.9729482906697935, + "learning_rate": 3.070068839564135e-07, + "loss": 0.7497, + "num_input_tokens_seen": 298096840, + "step": 9166 + }, + { + "epoch": 0.8267123596518916, + "flos": 22059372526080.0, + "grad_norm": 2.5265126107220306, + "learning_rate": 3.0669592584232006e-07, + "loss": 0.7818, + "num_input_tokens_seen": 298122855, + "step": 9167 + }, + { + "epoch": 0.8268025431753618, + "flos": 40457396369280.0, + "grad_norm": 1.886024363112628, + "learning_rate": 3.063851122102672e-07, + "loss": 0.6547, + "num_input_tokens_seen": 298155595, + "step": 9168 + }, + { + "epoch": 0.8268927266988321, + "flos": 31820981881440.0, + "grad_norm": 1.891555305037466, + "learning_rate": 3.06074443086775e-07, + "loss": 0.6618, + "num_input_tokens_seen": 298189405, + "step": 9169 + }, + { + "epoch": 0.8269829102223024, + "flos": 69936719554080.0, + "grad_norm": 0.6011022813755291, + "learning_rate": 3.057639184983514e-07, + "loss": 0.5305, + "num_input_tokens_seen": 298290565, + "step": 9170 + }, + { + "epoch": 0.8270730937457726, + "flos": 22787456947680.0, + "grad_norm": 1.495803924455632, + "learning_rate": 3.054535384714927e-07, + "loss": 0.746, + "num_input_tokens_seen": 298322060, + "step": 9171 + }, + { + "epoch": 0.8271632772692429, + "flos": 25957755733920.0, + "grad_norm": 1.8071572714607147, + "learning_rate": 3.0514330303268135e-07, + "loss": 0.9097, + "num_input_tokens_seen": 298349850, + "step": 9172 + }, + { + "epoch": 0.8272534607927132, + "flos": 18379953606240.0, + "grad_norm": 1.9194447107800994, + "learning_rate": 3.0483321220838876e-07, + "loss": 0.6973, + "num_input_tokens_seen": 298376620, + "step": 9173 + }, + { + "epoch": 0.8273436443161835, + "flos": 43954914985440.0, + "grad_norm": 6.649716031314355, + "learning_rate": 3.045232660250734e-07, + "loss": 0.6001, + "num_input_tokens_seen": 298411045, + "step": 9174 + }, + { + "epoch": 0.8274338278396537, + "flos": 24753303470880.0, + "grad_norm": 1.8498537572845708, + "learning_rate": 3.0421346450918185e-07, + "loss": 0.7693, + "num_input_tokens_seen": 298441570, + "step": 9175 + }, + { + "epoch": 0.8275240113631239, + "flos": 26905585008480.0, + "grad_norm": 1.4851235359517954, + "learning_rate": 3.039038076871485e-07, + "loss": 0.6923, + "num_input_tokens_seen": 298470335, + "step": 9176 + }, + { + "epoch": 0.8276141948865943, + "flos": 29121326793600.0, + "grad_norm": 1.8098577115734729, + "learning_rate": 3.035942955853934e-07, + "loss": 0.7608, + "num_input_tokens_seen": 298498865, + "step": 9177 + }, + { + "epoch": 0.8277043784100645, + "flos": 23152186799040.0, + "grad_norm": 2.352826865308554, + "learning_rate": 3.0328492823032804e-07, + "loss": 0.7182, + "num_input_tokens_seen": 298526755, + "step": 9178 + }, + { + "epoch": 0.8277945619335347, + "flos": 18522113702880.0, + "grad_norm": 2.286057723163449, + "learning_rate": 3.029757056483471e-07, + "loss": 0.6819, + "num_input_tokens_seen": 298555630, + "step": 9179 + }, + { + "epoch": 0.827884745457005, + "flos": 25300082304960.0, + "grad_norm": 2.0668173022388165, + "learning_rate": 3.026666278658372e-07, + "loss": 0.6333, + "num_input_tokens_seen": 298585055, + "step": 9180 + }, + { + "epoch": 0.8279749289804753, + "flos": 20235165722880.0, + "grad_norm": 1.4731650042857707, + "learning_rate": 3.023576949091691e-07, + "loss": 0.7425, + "num_input_tokens_seen": 298612965, + "step": 9181 + }, + { + "epoch": 0.8280651125039455, + "flos": 20493089652960.0, + "grad_norm": 1.9626232667345218, + "learning_rate": 3.020489068047032e-07, + "loss": 0.667, + "num_input_tokens_seen": 298639170, + "step": 9182 + }, + { + "epoch": 0.8281552960274158, + "flos": 30547270888320.0, + "grad_norm": 1.8830702355971145, + "learning_rate": 3.017402635787869e-07, + "loss": 0.7022, + "num_input_tokens_seen": 298670695, + "step": 9183 + }, + { + "epoch": 0.828245479550886, + "flos": 21659139820320.0, + "grad_norm": 1.8838256389414998, + "learning_rate": 3.0143176525775537e-07, + "loss": 0.7162, + "num_input_tokens_seen": 298699030, + "step": 9184 + }, + { + "epoch": 0.8283356630743564, + "flos": 22752251451360.0, + "grad_norm": 1.8935688523878715, + "learning_rate": 3.0112341186793155e-07, + "loss": 0.6894, + "num_input_tokens_seen": 298727895, + "step": 9185 + }, + { + "epoch": 0.8284258465978266, + "flos": 21871227702720.0, + "grad_norm": 1.7010858910397375, + "learning_rate": 3.008152034356264e-07, + "loss": 0.6815, + "num_input_tokens_seen": 298756140, + "step": 9186 + }, + { + "epoch": 0.8285160301212968, + "flos": 66407712417600.0, + "grad_norm": 0.6226044282683427, + "learning_rate": 3.005071399871366e-07, + "loss": 0.563, + "num_input_tokens_seen": 298850485, + "step": 9187 + }, + { + "epoch": 0.828606213644767, + "flos": 27596791294560.0, + "grad_norm": 2.0250629980389774, + "learning_rate": 3.0019922154874853e-07, + "loss": 0.8014, + "num_input_tokens_seen": 298879490, + "step": 9188 + }, + { + "epoch": 0.8286963971682374, + "flos": 20601828401760.0, + "grad_norm": 1.9624701965401536, + "learning_rate": 2.998914481467356e-07, + "loss": 0.6986, + "num_input_tokens_seen": 298906300, + "step": 9189 + }, + { + "epoch": 0.8287865806917076, + "flos": 19436707478400.0, + "grad_norm": 2.7089121175644024, + "learning_rate": 2.9958381980735837e-07, + "loss": 0.6903, + "num_input_tokens_seen": 298933495, + "step": 9190 + }, + { + "epoch": 0.8288767642151779, + "flos": 36231495973440.0, + "grad_norm": 1.694233564202826, + "learning_rate": 2.992763365568658e-07, + "loss": 0.6734, + "num_input_tokens_seen": 298964565, + "step": 9191 + }, + { + "epoch": 0.8289669477386481, + "flos": 23808745135200.0, + "grad_norm": 1.8369736365190081, + "learning_rate": 2.98968998421494e-07, + "loss": 0.743, + "num_input_tokens_seen": 298995190, + "step": 9192 + }, + { + "epoch": 0.8290571312621184, + "flos": 20709043190400.0, + "grad_norm": 3.512056294174703, + "learning_rate": 2.98661805427467e-07, + "loss": 0.7329, + "num_input_tokens_seen": 299020925, + "step": 9193 + }, + { + "epoch": 0.8291473147855887, + "flos": 45556217506080.0, + "grad_norm": 1.5870688025517996, + "learning_rate": 2.9835475760099483e-07, + "loss": 0.7586, + "num_input_tokens_seen": 299055560, + "step": 9194 + }, + { + "epoch": 0.8292374983090589, + "flos": 31198662627840.0, + "grad_norm": 1.9542272280463935, + "learning_rate": 2.9804785496827856e-07, + "loss": 0.7587, + "num_input_tokens_seen": 299085565, + "step": 9195 + }, + { + "epoch": 0.8293276818325293, + "flos": 15063629068320.0, + "grad_norm": 2.419622641374901, + "learning_rate": 2.977410975555028e-07, + "loss": 0.7334, + "num_input_tokens_seen": 299111850, + "step": 9196 + }, + { + "epoch": 0.8294178653559995, + "flos": 25155394664640.0, + "grad_norm": 1.9273521716950934, + "learning_rate": 2.9743448538884376e-07, + "loss": 0.8022, + "num_input_tokens_seen": 299135300, + "step": 9197 + }, + { + "epoch": 0.8295080488794697, + "flos": 21439097609280.0, + "grad_norm": 1.6665473974785723, + "learning_rate": 2.9712801849446154e-07, + "loss": 0.7304, + "num_input_tokens_seen": 299165945, + "step": 9198 + }, + { + "epoch": 0.82959823240294, + "flos": 29018312187840.0, + "grad_norm": 4.3238120503755315, + "learning_rate": 2.9682169689850665e-07, + "loss": 0.6945, + "num_input_tokens_seen": 299195865, + "step": 9199 + }, + { + "epoch": 0.8296884159264103, + "flos": 63543706886400.0, + "grad_norm": 0.6706734936788866, + "learning_rate": 2.9651552062711573e-07, + "loss": 0.5266, + "num_input_tokens_seen": 299287015, + "step": 9200 + }, + { + "epoch": 0.8297785994498805, + "flos": 20487030982080.0, + "grad_norm": 2.313996320624299, + "learning_rate": 2.9620948970641333e-07, + "loss": 0.8225, + "num_input_tokens_seen": 299317380, + "step": 9201 + }, + { + "epoch": 0.8298687829733508, + "flos": 23733985280640.0, + "grad_norm": 1.5454250437424701, + "learning_rate": 2.959036041625125e-07, + "loss": 0.7608, + "num_input_tokens_seen": 299348025, + "step": 9202 + }, + { + "epoch": 0.829958966496821, + "flos": 28070445743520.0, + "grad_norm": 1.8233493473086644, + "learning_rate": 2.95597864021512e-07, + "loss": 0.6525, + "num_input_tokens_seen": 299378925, + "step": 9203 + }, + { + "epoch": 0.8300491500202913, + "flos": 16079639149920.0, + "grad_norm": 2.4170403246469094, + "learning_rate": 2.9529226930949966e-07, + "loss": 0.7936, + "num_input_tokens_seen": 299401080, + "step": 9204 + }, + { + "epoch": 0.8301393335437616, + "flos": 22710132379680.0, + "grad_norm": 1.8371617557326532, + "learning_rate": 2.949868200525505e-07, + "loss": 0.7541, + "num_input_tokens_seen": 299429245, + "step": 9205 + }, + { + "epoch": 0.8302295170672318, + "flos": 23407880543520.0, + "grad_norm": 2.4087505476376396, + "learning_rate": 2.9468151627672734e-07, + "loss": 0.7364, + "num_input_tokens_seen": 299459400, + "step": 9206 + }, + { + "epoch": 0.830319700590702, + "flos": 27485599341600.0, + "grad_norm": 1.8414304334357765, + "learning_rate": 2.9437635800808026e-07, + "loss": 0.7003, + "num_input_tokens_seen": 299488620, + "step": 9207 + }, + { + "epoch": 0.8304098841141724, + "flos": 22059409695840.0, + "grad_norm": 6.331104622590556, + "learning_rate": 2.940713452726473e-07, + "loss": 0.7322, + "num_input_tokens_seen": 299514460, + "step": 9208 + }, + { + "epoch": 0.8305000676376426, + "flos": 19217780360160.0, + "grad_norm": 2.0025701697696072, + "learning_rate": 2.937664780964526e-07, + "loss": 0.7378, + "num_input_tokens_seen": 299542295, + "step": 9209 + }, + { + "epoch": 0.8305902511611128, + "flos": 27887950723680.0, + "grad_norm": 2.0573365283050418, + "learning_rate": 2.9346175650551133e-07, + "loss": 0.7483, + "num_input_tokens_seen": 299569090, + "step": 9210 + }, + { + "epoch": 0.8306804346845831, + "flos": 23218546287840.0, + "grad_norm": 2.0037860797149176, + "learning_rate": 2.931571805258215e-07, + "loss": 0.7628, + "num_input_tokens_seen": 299598230, + "step": 9211 + }, + { + "epoch": 0.8307706182080534, + "flos": 25119482942880.0, + "grad_norm": 5.939656398095887, + "learning_rate": 2.9285275018337353e-07, + "loss": 0.792, + "num_input_tokens_seen": 299626045, + "step": 9212 + }, + { + "epoch": 0.8308608017315237, + "flos": 23217617043840.0, + "grad_norm": 2.614589939387716, + "learning_rate": 2.9254846550414146e-07, + "loss": 0.8431, + "num_input_tokens_seen": 299652045, + "step": 9213 + }, + { + "epoch": 0.8309509852549939, + "flos": 19945567423680.0, + "grad_norm": 2.67747971584925, + "learning_rate": 2.922443265140893e-07, + "loss": 0.7916, + "num_input_tokens_seen": 299678750, + "step": 9214 + }, + { + "epoch": 0.8310411687784641, + "flos": 24281990716800.0, + "grad_norm": 1.9421682416677684, + "learning_rate": 2.919403332391674e-07, + "loss": 0.7425, + "num_input_tokens_seen": 299706265, + "step": 9215 + }, + { + "epoch": 0.8311313523019345, + "flos": 31457887499520.0, + "grad_norm": 1.844284174203942, + "learning_rate": 2.9163648570531464e-07, + "loss": 0.7653, + "num_input_tokens_seen": 299740365, + "step": 9216 + }, + { + "epoch": 0.8312215358254047, + "flos": 38563187440800.0, + "grad_norm": 1.9324696365478033, + "learning_rate": 2.9133278393845717e-07, + "loss": 0.681, + "num_input_tokens_seen": 299774075, + "step": 9217 + }, + { + "epoch": 0.8313117193488749, + "flos": 24718432502400.0, + "grad_norm": 2.3026309390533246, + "learning_rate": 2.9102922796450703e-07, + "loss": 0.8117, + "num_input_tokens_seen": 299803520, + "step": 9218 + }, + { + "epoch": 0.8314019028723453, + "flos": 41843860445280.0, + "grad_norm": 1.6731884511915218, + "learning_rate": 2.907258178093672e-07, + "loss": 0.6999, + "num_input_tokens_seen": 299838155, + "step": 9219 + }, + { + "epoch": 0.8314920863958155, + "flos": 24135667607040.0, + "grad_norm": 1.6989233702947126, + "learning_rate": 2.904225534989251e-07, + "loss": 0.6664, + "num_input_tokens_seen": 299867475, + "step": 9220 + }, + { + "epoch": 0.8315822699192857, + "flos": 52367904814080.0, + "grad_norm": 2.1072272775655705, + "learning_rate": 2.901194350590572e-07, + "loss": 0.6439, + "num_input_tokens_seen": 299900450, + "step": 9221 + }, + { + "epoch": 0.831672453442756, + "flos": 34044752334720.0, + "grad_norm": 1.705615261632978, + "learning_rate": 2.898164625156274e-07, + "loss": 0.6203, + "num_input_tokens_seen": 299933665, + "step": 9222 + }, + { + "epoch": 0.8317626369662263, + "flos": 23261706112800.0, + "grad_norm": 6.247896069217865, + "learning_rate": 2.8951363589448676e-07, + "loss": 0.6984, + "num_input_tokens_seen": 299963810, + "step": 9223 + }, + { + "epoch": 0.8318528204896966, + "flos": 18671187374880.0, + "grad_norm": 3.111716457541366, + "learning_rate": 2.8921095522147434e-07, + "loss": 0.7371, + "num_input_tokens_seen": 299991515, + "step": 9224 + }, + { + "epoch": 0.8319430040131668, + "flos": 16412583122880.0, + "grad_norm": 1.9942540837963765, + "learning_rate": 2.8890842052241683e-07, + "loss": 0.7752, + "num_input_tokens_seen": 300017090, + "step": 9225 + }, + { + "epoch": 0.832033187536637, + "flos": 65738516363040.0, + "grad_norm": 0.5908165545383456, + "learning_rate": 2.886060318231267e-07, + "loss": 0.5257, + "num_input_tokens_seen": 300108605, + "step": 9226 + }, + { + "epoch": 0.8321233710601074, + "flos": 25811953000800.0, + "grad_norm": 1.694588572551331, + "learning_rate": 2.8830378914940755e-07, + "loss": 0.7654, + "num_input_tokens_seen": 300138775, + "step": 9227 + }, + { + "epoch": 0.8322135545835776, + "flos": 17287325182080.0, + "grad_norm": 2.0823385788361075, + "learning_rate": 2.8800169252704675e-07, + "loss": 0.7294, + "num_input_tokens_seen": 300163990, + "step": 9228 + }, + { + "epoch": 0.8323037381070478, + "flos": 22714890108960.0, + "grad_norm": 1.6878577665499985, + "learning_rate": 2.8769974198182143e-07, + "loss": 0.7633, + "num_input_tokens_seen": 300192535, + "step": 9229 + }, + { + "epoch": 0.8323939216305181, + "flos": 30876460715520.0, + "grad_norm": 1.7761987372456707, + "learning_rate": 2.873979375394955e-07, + "loss": 0.7412, + "num_input_tokens_seen": 300222255, + "step": 9230 + }, + { + "epoch": 0.8324841051539884, + "flos": 24756016863360.0, + "grad_norm": 1.6386090199506629, + "learning_rate": 2.870962792258209e-07, + "loss": 0.7756, + "num_input_tokens_seen": 300249985, + "step": 9231 + }, + { + "epoch": 0.8325742886774586, + "flos": 22531056977760.0, + "grad_norm": 1.6459977594965245, + "learning_rate": 2.8679476706653716e-07, + "loss": 0.7533, + "num_input_tokens_seen": 300279185, + "step": 9232 + }, + { + "epoch": 0.8326644722009289, + "flos": 24500434628160.0, + "grad_norm": 2.861459274671409, + "learning_rate": 2.864934010873692e-07, + "loss": 0.7616, + "num_input_tokens_seen": 300307225, + "step": 9233 + }, + { + "epoch": 0.8327546557243991, + "flos": 25880431165920.0, + "grad_norm": 1.514870052492933, + "learning_rate": 2.8619218131403357e-07, + "loss": 0.7165, + "num_input_tokens_seen": 300338130, + "step": 9234 + }, + { + "epoch": 0.8328448392478695, + "flos": 18561965419200.0, + "grad_norm": 1.9363466336964663, + "learning_rate": 2.858911077722299e-07, + "loss": 0.7676, + "num_input_tokens_seen": 300365840, + "step": 9235 + }, + { + "epoch": 0.8329350227713397, + "flos": 28906042311840.0, + "grad_norm": 2.0907120941677895, + "learning_rate": 2.855901804876493e-07, + "loss": 0.5915, + "num_input_tokens_seen": 300397255, + "step": 9236 + }, + { + "epoch": 0.8330252062948099, + "flos": 26391632806080.0, + "grad_norm": 1.8326852115226149, + "learning_rate": 2.852893994859673e-07, + "loss": 0.7773, + "num_input_tokens_seen": 300428270, + "step": 9237 + }, + { + "epoch": 0.8331153898182801, + "flos": 32110431501600.0, + "grad_norm": 1.9682192616862466, + "learning_rate": 2.849887647928484e-07, + "loss": 0.6866, + "num_input_tokens_seen": 300460660, + "step": 9238 + }, + { + "epoch": 0.8332055733417505, + "flos": 22314322875360.0, + "grad_norm": 3.57048865620374, + "learning_rate": 2.8468827643394465e-07, + "loss": 0.6423, + "num_input_tokens_seen": 300490450, + "step": 9239 + }, + { + "epoch": 0.8332957568652207, + "flos": 26641862595840.0, + "grad_norm": 1.9869981983984895, + "learning_rate": 2.843879344348954e-07, + "loss": 0.7536, + "num_input_tokens_seen": 300518220, + "step": 9240 + }, + { + "epoch": 0.833385940388691, + "flos": 60385382496480.0, + "grad_norm": 12.50011352487511, + "learning_rate": 2.840877388213272e-07, + "loss": 0.7142, + "num_input_tokens_seen": 300554720, + "step": 9241 + }, + { + "epoch": 0.8334761239121612, + "flos": 25119445773120.0, + "grad_norm": 2.415009050687965, + "learning_rate": 2.8378768961885515e-07, + "loss": 0.7456, + "num_input_tokens_seen": 300579880, + "step": 9242 + }, + { + "epoch": 0.8335663074356315, + "flos": 16411505199840.0, + "grad_norm": 1.8057195181451766, + "learning_rate": 2.8348778685307983e-07, + "loss": 0.7757, + "num_input_tokens_seen": 300606015, + "step": 9243 + }, + { + "epoch": 0.8336564909591018, + "flos": 24063546805440.0, + "grad_norm": 3.5439841971809303, + "learning_rate": 2.831880305495915e-07, + "loss": 0.8195, + "num_input_tokens_seen": 300634930, + "step": 9244 + }, + { + "epoch": 0.833746674482572, + "flos": 64792471236960.0, + "grad_norm": 0.6506707673552671, + "learning_rate": 2.828884207339668e-07, + "loss": 0.6111, + "num_input_tokens_seen": 300728600, + "step": 9245 + }, + { + "epoch": 0.8338368580060423, + "flos": 17212044950880.0, + "grad_norm": 2.5103414268351365, + "learning_rate": 2.8258895743177014e-07, + "loss": 0.8196, + "num_input_tokens_seen": 300753655, + "step": 9246 + }, + { + "epoch": 0.8339270415295126, + "flos": 22532952635520.0, + "grad_norm": 1.8691457272004552, + "learning_rate": 2.8228964066855356e-07, + "loss": 0.8158, + "num_input_tokens_seen": 300779730, + "step": 9247 + }, + { + "epoch": 0.8340172250529828, + "flos": 27885869217120.0, + "grad_norm": 1.6852326172001648, + "learning_rate": 2.819904704698555e-07, + "loss": 0.7202, + "num_input_tokens_seen": 300809345, + "step": 9248 + }, + { + "epoch": 0.834107408576453, + "flos": 59527063737600.0, + "grad_norm": 0.633798772564112, + "learning_rate": 2.8169144686120437e-07, + "loss": 0.5268, + "num_input_tokens_seen": 300899160, + "step": 9249 + }, + { + "epoch": 0.8341975920999234, + "flos": 28508411489280.0, + "grad_norm": 2.202553316438758, + "learning_rate": 2.8139256986811254e-07, + "loss": 0.6546, + "num_input_tokens_seen": 300929165, + "step": 9250 + }, + { + "epoch": 0.8342877756233936, + "flos": 20601642552960.0, + "grad_norm": 2.089778378613965, + "learning_rate": 2.8109383951608424e-07, + "loss": 0.7945, + "num_input_tokens_seen": 300954885, + "step": 9251 + }, + { + "epoch": 0.8343779591468639, + "flos": 29527246472640.0, + "grad_norm": 1.9288512359968795, + "learning_rate": 2.8079525583060683e-07, + "loss": 0.8079, + "num_input_tokens_seen": 300983620, + "step": 9252 + }, + { + "epoch": 0.8344681426703341, + "flos": 25265248506240.0, + "grad_norm": 1.5682927137675922, + "learning_rate": 2.804968188371577e-07, + "loss": 0.7488, + "num_input_tokens_seen": 301013955, + "step": 9253 + }, + { + "epoch": 0.8345583261938044, + "flos": 30182912734560.0, + "grad_norm": 1.9129346366120812, + "learning_rate": 2.801985285612014e-07, + "loss": 0.7444, + "num_input_tokens_seen": 301045445, + "step": 9254 + }, + { + "epoch": 0.8346485097172747, + "flos": 25119557282400.0, + "grad_norm": 1.7863228156142053, + "learning_rate": 2.7990038502818934e-07, + "loss": 0.7158, + "num_input_tokens_seen": 301074160, + "step": 9255 + }, + { + "epoch": 0.8347386932407449, + "flos": 22309267788000.0, + "grad_norm": 1.7077776152319009, + "learning_rate": 2.796023882635612e-07, + "loss": 0.7184, + "num_input_tokens_seen": 301103150, + "step": 9256 + }, + { + "epoch": 0.8348288767642151, + "flos": 20196912306240.0, + "grad_norm": 2.055814020844191, + "learning_rate": 2.7930453829274323e-07, + "loss": 0.7912, + "num_input_tokens_seen": 301131580, + "step": 9257 + }, + { + "epoch": 0.8349190602876855, + "flos": 21767729890080.0, + "grad_norm": 2.5186989525859027, + "learning_rate": 2.7900683514115054e-07, + "loss": 0.7848, + "num_input_tokens_seen": 301160295, + "step": 9258 + }, + { + "epoch": 0.8350092438111557, + "flos": 22314174196320.0, + "grad_norm": 3.057164741407227, + "learning_rate": 2.787092788341836e-07, + "loss": 0.6786, + "num_input_tokens_seen": 301188105, + "step": 9259 + }, + { + "epoch": 0.8350994273346259, + "flos": 21658768122720.0, + "grad_norm": 3.0784036732891633, + "learning_rate": 2.7841186939723195e-07, + "loss": 0.7419, + "num_input_tokens_seen": 301213950, + "step": 9260 + }, + { + "epoch": 0.8351896108580962, + "flos": 18489324240960.0, + "grad_norm": 1.8573594268345182, + "learning_rate": 2.7811460685567255e-07, + "loss": 0.7189, + "num_input_tokens_seen": 301240875, + "step": 9261 + }, + { + "epoch": 0.8352797943815665, + "flos": 27774156887520.0, + "grad_norm": 2.1570231944938545, + "learning_rate": 2.778174912348692e-07, + "loss": 0.7377, + "num_input_tokens_seen": 301267990, + "step": 9262 + }, + { + "epoch": 0.8353699779050368, + "flos": 24244963902240.0, + "grad_norm": 1.9235080510826972, + "learning_rate": 2.7752052256017354e-07, + "loss": 0.7045, + "num_input_tokens_seen": 301294935, + "step": 9263 + }, + { + "epoch": 0.835460161428507, + "flos": 17468221902240.0, + "grad_norm": 2.0316817778394896, + "learning_rate": 2.7722370085692493e-07, + "loss": 0.7123, + "num_input_tokens_seen": 301322265, + "step": 9264 + }, + { + "epoch": 0.8355503449519772, + "flos": 22533473012160.0, + "grad_norm": 2.676215678123675, + "learning_rate": 2.769270261504486e-07, + "loss": 0.7145, + "num_input_tokens_seen": 301349490, + "step": 9265 + }, + { + "epoch": 0.8356405284754476, + "flos": 29596207844640.0, + "grad_norm": 1.663177254768724, + "learning_rate": 2.7663049846606015e-07, + "loss": 0.702, + "num_input_tokens_seen": 301381955, + "step": 9266 + }, + { + "epoch": 0.8357307119989178, + "flos": 34810198098720.0, + "grad_norm": 1.66174914964884, + "learning_rate": 2.763341178290592e-07, + "loss": 0.7538, + "num_input_tokens_seen": 301412500, + "step": 9267 + }, + { + "epoch": 0.835820895522388, + "flos": 21476570460960.0, + "grad_norm": 2.3472921435032124, + "learning_rate": 2.7603788426473663e-07, + "loss": 0.7666, + "num_input_tokens_seen": 301442830, + "step": 9268 + }, + { + "epoch": 0.8359110790458584, + "flos": 24167527824960.0, + "grad_norm": 3.412410519722788, + "learning_rate": 2.7574179779836695e-07, + "loss": 0.6751, + "num_input_tokens_seen": 301473255, + "step": 9269 + }, + { + "epoch": 0.8360012625693286, + "flos": 42934481702400.0, + "grad_norm": 1.8252775399273014, + "learning_rate": 2.754458584552146e-07, + "loss": 0.6997, + "num_input_tokens_seen": 301504645, + "step": 9270 + }, + { + "epoch": 0.8360914460927988, + "flos": 28616704200960.0, + "grad_norm": 2.5909564129524796, + "learning_rate": 2.751500662605308e-07, + "loss": 0.7262, + "num_input_tokens_seen": 301532115, + "step": 9271 + }, + { + "epoch": 0.8361816296162691, + "flos": 24425340245760.0, + "grad_norm": 6.789188923772543, + "learning_rate": 2.7485442123955383e-07, + "loss": 0.6955, + "num_input_tokens_seen": 301561935, + "step": 9272 + }, + { + "epoch": 0.8362718131397394, + "flos": 28580197763040.0, + "grad_norm": 3.530427279886169, + "learning_rate": 2.7455892341751075e-07, + "loss": 0.6262, + "num_input_tokens_seen": 301591550, + "step": 9273 + }, + { + "epoch": 0.8363619966632097, + "flos": 23079582790560.0, + "grad_norm": 1.996712545792281, + "learning_rate": 2.7426357281961365e-07, + "loss": 0.6166, + "num_input_tokens_seen": 301617570, + "step": 9274 + }, + { + "epoch": 0.8364521801866799, + "flos": 25154130892800.0, + "grad_norm": 2.5108232815205036, + "learning_rate": 2.7396836947106416e-07, + "loss": 0.7503, + "num_input_tokens_seen": 301641845, + "step": 9275 + }, + { + "epoch": 0.8365423637101501, + "flos": 24641665480800.0, + "grad_norm": 2.4772203215591078, + "learning_rate": 2.736733133970506e-07, + "loss": 0.839, + "num_input_tokens_seen": 301670495, + "step": 9276 + }, + { + "epoch": 0.8366325472336205, + "flos": 24680327764800.0, + "grad_norm": 1.4681766283645186, + "learning_rate": 2.7337840462274896e-07, + "loss": 0.7816, + "num_input_tokens_seen": 301702420, + "step": 9277 + }, + { + "epoch": 0.8367227307570907, + "flos": 24097005323040.0, + "grad_norm": 2.3672494408085822, + "learning_rate": 2.730836431733221e-07, + "loss": 0.7095, + "num_input_tokens_seen": 301731280, + "step": 9278 + }, + { + "epoch": 0.8368129142805609, + "flos": 22087478598240.0, + "grad_norm": 1.8269173961813634, + "learning_rate": 2.727890290739212e-07, + "loss": 0.7894, + "num_input_tokens_seen": 301757855, + "step": 9279 + }, + { + "epoch": 0.8369030978040312, + "flos": 24244703713920.0, + "grad_norm": 1.718366406724005, + "learning_rate": 2.7249456234968395e-07, + "loss": 0.7705, + "num_input_tokens_seen": 301787890, + "step": 9280 + }, + { + "epoch": 0.8369932813275015, + "flos": 21257977870560.0, + "grad_norm": 1.6843242432372727, + "learning_rate": 2.722002430257364e-07, + "loss": 0.8095, + "num_input_tokens_seen": 301815865, + "step": 9281 + }, + { + "epoch": 0.8370834648509717, + "flos": 27742854216000.0, + "grad_norm": 1.9661551435304887, + "learning_rate": 2.7190607112719035e-07, + "loss": 0.7055, + "num_input_tokens_seen": 301846265, + "step": 9282 + }, + { + "epoch": 0.837173648374442, + "flos": 29090804687040.0, + "grad_norm": 2.4664561564169554, + "learning_rate": 2.716120466791476e-07, + "loss": 0.6616, + "num_input_tokens_seen": 301875700, + "step": 9283 + }, + { + "epoch": 0.8372638318979122, + "flos": 23771569641600.0, + "grad_norm": 2.1045487186070266, + "learning_rate": 2.7131816970669483e-07, + "loss": 0.7193, + "num_input_tokens_seen": 301903920, + "step": 9284 + }, + { + "epoch": 0.8373540154213825, + "flos": 48145015168800.0, + "grad_norm": 3.2717432266996562, + "learning_rate": 2.7102444023490777e-07, + "loss": 0.7478, + "num_input_tokens_seen": 301934565, + "step": 9285 + }, + { + "epoch": 0.8374441989448528, + "flos": 23950533534240.0, + "grad_norm": 1.6822915406911072, + "learning_rate": 2.70730858288849e-07, + "loss": 0.7961, + "num_input_tokens_seen": 301964985, + "step": 9286 + }, + { + "epoch": 0.837534382468323, + "flos": 35353594484640.0, + "grad_norm": 1.9196592754789834, + "learning_rate": 2.704374238935685e-07, + "loss": 0.6769, + "num_input_tokens_seen": 301995605, + "step": 9287 + }, + { + "epoch": 0.8376245659917932, + "flos": 51796136434080.0, + "grad_norm": 0.6721696872736321, + "learning_rate": 2.70144137074104e-07, + "loss": 0.5735, + "num_input_tokens_seen": 302082475, + "step": 9288 + }, + { + "epoch": 0.8377147495152636, + "flos": 26796400222560.0, + "grad_norm": 1.8687064404976632, + "learning_rate": 2.6985099785547926e-07, + "loss": 0.7611, + "num_input_tokens_seen": 302113320, + "step": 9289 + }, + { + "epoch": 0.8378049330387338, + "flos": 23079136753440.0, + "grad_norm": 1.9000191712667087, + "learning_rate": 2.695580062627083e-07, + "loss": 0.7674, + "num_input_tokens_seen": 302144065, + "step": 9290 + }, + { + "epoch": 0.8378951165622041, + "flos": 21512705201280.0, + "grad_norm": 3.0436847906940923, + "learning_rate": 2.692651623207891e-07, + "loss": 0.7382, + "num_input_tokens_seen": 302169640, + "step": 9291 + }, + { + "epoch": 0.8379853000856744, + "flos": 34226057922240.0, + "grad_norm": 1.7602695501428627, + "learning_rate": 2.689724660547097e-07, + "loss": 0.7268, + "num_input_tokens_seen": 302199855, + "step": 9292 + }, + { + "epoch": 0.8380754836091446, + "flos": 65227723590240.0, + "grad_norm": 0.6165622489607735, + "learning_rate": 2.686799174894441e-07, + "loss": 0.5702, + "num_input_tokens_seen": 302290655, + "step": 9293 + }, + { + "epoch": 0.8381656671326149, + "flos": 65708886330720.0, + "grad_norm": 0.6399388942381049, + "learning_rate": 2.683875166499545e-07, + "loss": 0.5705, + "num_input_tokens_seen": 302388260, + "step": 9294 + }, + { + "epoch": 0.8382558506560851, + "flos": 15828665964960.0, + "grad_norm": 1.8832084904143063, + "learning_rate": 2.680952635611899e-07, + "loss": 0.7478, + "num_input_tokens_seen": 302411475, + "step": 9295 + }, + { + "epoch": 0.8383460341795554, + "flos": 18707507964000.0, + "grad_norm": 2.1916742008016876, + "learning_rate": 2.678031582480875e-07, + "loss": 0.6684, + "num_input_tokens_seen": 302438715, + "step": 9296 + }, + { + "epoch": 0.8384362177030257, + "flos": 24753377810400.0, + "grad_norm": 1.8142944529358336, + "learning_rate": 2.6751120073557e-07, + "loss": 0.7811, + "num_input_tokens_seen": 302467890, + "step": 9297 + }, + { + "epoch": 0.8385264012264959, + "flos": 17869123663680.0, + "grad_norm": 3.751273384059488, + "learning_rate": 2.672193910485505e-07, + "loss": 0.7081, + "num_input_tokens_seen": 302495495, + "step": 9298 + }, + { + "epoch": 0.8386165847499661, + "flos": 21804199158240.0, + "grad_norm": 1.8667920703430991, + "learning_rate": 2.669277292119265e-07, + "loss": 0.7401, + "num_input_tokens_seen": 302524770, + "step": 9299 + }, + { + "epoch": 0.8387067682734365, + "flos": 24136522511520.0, + "grad_norm": 3.1215679395725457, + "learning_rate": 2.666362152505848e-07, + "loss": 0.7427, + "num_input_tokens_seen": 302550880, + "step": 9300 + }, + { + "epoch": 0.8387969517969067, + "flos": 23036943342240.0, + "grad_norm": 2.438515671617695, + "learning_rate": 2.663448491893989e-07, + "loss": 0.6746, + "num_input_tokens_seen": 302580670, + "step": 9301 + }, + { + "epoch": 0.838887135320377, + "flos": 24128977050240.0, + "grad_norm": 1.7448331306047595, + "learning_rate": 2.6605363105322974e-07, + "loss": 0.7146, + "num_input_tokens_seen": 302609130, + "step": 9302 + }, + { + "epoch": 0.8389773188438472, + "flos": 28143904656480.0, + "grad_norm": 1.8806874799761035, + "learning_rate": 2.657625608669263e-07, + "loss": 0.7383, + "num_input_tokens_seen": 302636080, + "step": 9303 + }, + { + "epoch": 0.8390675023673175, + "flos": 43444270891680.0, + "grad_norm": 1.6279001033042202, + "learning_rate": 2.654716386553224e-07, + "loss": 0.6535, + "num_input_tokens_seen": 302668215, + "step": 9304 + }, + { + "epoch": 0.8391576858907878, + "flos": 34628260625280.0, + "grad_norm": 2.3277140004968677, + "learning_rate": 2.651808644432436e-07, + "loss": 0.7371, + "num_input_tokens_seen": 302700575, + "step": 9305 + }, + { + "epoch": 0.839247869414258, + "flos": 22934114585280.0, + "grad_norm": 1.990579592008715, + "learning_rate": 2.6489023825549807e-07, + "loss": 0.8717, + "num_input_tokens_seen": 302724650, + "step": 9306 + }, + { + "epoch": 0.8393380529377282, + "flos": 27305222998080.0, + "grad_norm": 2.6277341681319135, + "learning_rate": 2.6459976011688547e-07, + "loss": 0.6612, + "num_input_tokens_seen": 302750200, + "step": 9307 + }, + { + "epoch": 0.8394282364611986, + "flos": 21476942158560.0, + "grad_norm": 3.5136184052178105, + "learning_rate": 2.6430943005219e-07, + "loss": 0.7371, + "num_input_tokens_seen": 302778305, + "step": 9308 + }, + { + "epoch": 0.8395184199846688, + "flos": 25774294300320.0, + "grad_norm": 4.173098495465503, + "learning_rate": 2.6401924808618447e-07, + "loss": 0.7674, + "num_input_tokens_seen": 302806705, + "step": 9309 + }, + { + "epoch": 0.839608603508139, + "flos": 20671719017760.0, + "grad_norm": 2.0639508086486074, + "learning_rate": 2.637292142436287e-07, + "loss": 0.7347, + "num_input_tokens_seen": 302835315, + "step": 9310 + }, + { + "epoch": 0.8396987870316093, + "flos": 21585309209760.0, + "grad_norm": 2.4747099581225784, + "learning_rate": 2.6343932854927e-07, + "loss": 0.7095, + "num_input_tokens_seen": 302863365, + "step": 9311 + }, + { + "epoch": 0.8397889705550796, + "flos": 69888430302240.0, + "grad_norm": 0.6835868399134565, + "learning_rate": 2.6314959102784316e-07, + "loss": 0.5692, + "num_input_tokens_seen": 302953995, + "step": 9312 + }, + { + "epoch": 0.8398791540785498, + "flos": 16558051328160.0, + "grad_norm": 3.462799329248103, + "learning_rate": 2.6286000170407074e-07, + "loss": 0.7312, + "num_input_tokens_seen": 302979230, + "step": 9313 + }, + { + "epoch": 0.8399693376020201, + "flos": 23735323392000.0, + "grad_norm": 2.4060633203001838, + "learning_rate": 2.625705606026607e-07, + "loss": 0.7144, + "num_input_tokens_seen": 303008280, + "step": 9314 + }, + { + "epoch": 0.8400595211254904, + "flos": 23254792537440.0, + "grad_norm": 1.5485059059037833, + "learning_rate": 2.622812677483106e-07, + "loss": 0.8201, + "num_input_tokens_seen": 303035220, + "step": 9315 + }, + { + "epoch": 0.8401497046489607, + "flos": 14444655093120.0, + "grad_norm": 3.9747679732638708, + "learning_rate": 2.6199212316570453e-07, + "loss": 0.7921, + "num_input_tokens_seen": 303059495, + "step": 9316 + }, + { + "epoch": 0.8402398881724309, + "flos": 21111840609600.0, + "grad_norm": 2.4414368162149067, + "learning_rate": 2.617031268795138e-07, + "loss": 0.7199, + "num_input_tokens_seen": 303090055, + "step": 9317 + }, + { + "epoch": 0.8403300716959011, + "flos": 24570734111520.0, + "grad_norm": 1.4902660521494022, + "learning_rate": 2.614142789143976e-07, + "loss": 0.7681, + "num_input_tokens_seen": 303121020, + "step": 9318 + }, + { + "epoch": 0.8404202552193715, + "flos": 19108595574240.0, + "grad_norm": 1.7221611851298209, + "learning_rate": 2.6112557929500047e-07, + "loss": 0.7231, + "num_input_tokens_seen": 303149210, + "step": 9319 + }, + { + "epoch": 0.8405104387428417, + "flos": 21076226245920.0, + "grad_norm": 1.8628167107145257, + "learning_rate": 2.6083702804595817e-07, + "loss": 0.7626, + "num_input_tokens_seen": 303177675, + "step": 9320 + }, + { + "epoch": 0.8406006222663119, + "flos": 22897533807840.0, + "grad_norm": 1.742121947556333, + "learning_rate": 2.6054862519188915e-07, + "loss": 0.7238, + "num_input_tokens_seen": 303207150, + "step": 9321 + }, + { + "epoch": 0.8406908057897822, + "flos": 18634383578880.0, + "grad_norm": 1.7973906067520093, + "learning_rate": 2.6026037075740357e-07, + "loss": 0.6741, + "num_input_tokens_seen": 303234590, + "step": 9322 + }, + { + "epoch": 0.8407809893132525, + "flos": 30365036056800.0, + "grad_norm": 1.995591789181919, + "learning_rate": 2.5997226476709524e-07, + "loss": 0.7067, + "num_input_tokens_seen": 303262470, + "step": 9323 + }, + { + "epoch": 0.8408711728367227, + "flos": 18087790593600.0, + "grad_norm": 1.9702239195887614, + "learning_rate": 2.5968430724554856e-07, + "loss": 0.7418, + "num_input_tokens_seen": 303288800, + "step": 9324 + }, + { + "epoch": 0.840961356360193, + "flos": 36850804476480.0, + "grad_norm": 2.033337151674948, + "learning_rate": 2.5939649821733225e-07, + "loss": 0.6731, + "num_input_tokens_seen": 303318540, + "step": 9325 + }, + { + "epoch": 0.8410515398836632, + "flos": 20562980268960.0, + "grad_norm": 6.422432756914613, + "learning_rate": 2.5910883770700433e-07, + "loss": 0.7015, + "num_input_tokens_seen": 303347935, + "step": 9326 + }, + { + "epoch": 0.8411417234071336, + "flos": 37870866061920.0, + "grad_norm": 3.103851626664703, + "learning_rate": 2.5882132573910965e-07, + "loss": 0.5819, + "num_input_tokens_seen": 303381380, + "step": 9327 + }, + { + "epoch": 0.8412319069306038, + "flos": 24353851330080.0, + "grad_norm": 1.3774678884595395, + "learning_rate": 2.585339623381801e-07, + "loss": 0.7783, + "num_input_tokens_seen": 303413075, + "step": 9328 + }, + { + "epoch": 0.841322090454074, + "flos": 27305780544480.0, + "grad_norm": 1.9331686766432816, + "learning_rate": 2.582467475287358e-07, + "loss": 0.7173, + "num_input_tokens_seen": 303444145, + "step": 9329 + }, + { + "epoch": 0.8414122739775443, + "flos": 24427867789440.0, + "grad_norm": 1.9914128611332553, + "learning_rate": 2.5795968133528224e-07, + "loss": 0.7487, + "num_input_tokens_seen": 303471530, + "step": 9330 + }, + { + "epoch": 0.8415024575010146, + "flos": 22642286100480.0, + "grad_norm": 1.790297835521863, + "learning_rate": 2.576727637823144e-07, + "loss": 0.7409, + "num_input_tokens_seen": 303499350, + "step": 9331 + }, + { + "epoch": 0.8415926410244848, + "flos": 21031170763200.0, + "grad_norm": 2.9608033484513427, + "learning_rate": 2.5738599489431335e-07, + "loss": 0.7796, + "num_input_tokens_seen": 303525980, + "step": 9332 + }, + { + "epoch": 0.8416828245479551, + "flos": 27045180391680.0, + "grad_norm": 1.9877163247475347, + "learning_rate": 2.5709937469574794e-07, + "loss": 0.8006, + "num_input_tokens_seen": 303555145, + "step": 9333 + }, + { + "epoch": 0.8417730080714253, + "flos": 25156100890080.0, + "grad_norm": 2.80874994132802, + "learning_rate": 2.568129032110742e-07, + "loss": 0.7778, + "num_input_tokens_seen": 303584480, + "step": 9334 + }, + { + "epoch": 0.8418631915948956, + "flos": 19910064569280.0, + "grad_norm": 1.638334291806463, + "learning_rate": 2.5652658046473565e-07, + "loss": 0.8154, + "num_input_tokens_seen": 303612575, + "step": 9335 + }, + { + "epoch": 0.8419533751183659, + "flos": 28140670887360.0, + "grad_norm": 1.7363476918408536, + "learning_rate": 2.5624040648116184e-07, + "loss": 0.7391, + "num_input_tokens_seen": 303641135, + "step": 9336 + }, + { + "epoch": 0.8420435586418361, + "flos": 15829372190400.0, + "grad_norm": 2.0988347953493016, + "learning_rate": 2.5595438128477245e-07, + "loss": 0.6771, + "num_input_tokens_seen": 303668570, + "step": 9337 + }, + { + "epoch": 0.8421337421653065, + "flos": 22532469428640.0, + "grad_norm": 1.700218660653333, + "learning_rate": 2.5566850489997096e-07, + "loss": 0.7457, + "num_input_tokens_seen": 303694825, + "step": 9338 + }, + { + "epoch": 0.8422239256887767, + "flos": 22861027369920.0, + "grad_norm": 2.210908168422926, + "learning_rate": 2.5538277735115166e-07, + "loss": 0.7632, + "num_input_tokens_seen": 303725180, + "step": 9339 + }, + { + "epoch": 0.8423141092122469, + "flos": 28688936511840.0, + "grad_norm": 1.5591206436799792, + "learning_rate": 2.5509719866269306e-07, + "loss": 0.7922, + "num_input_tokens_seen": 303754960, + "step": 9340 + }, + { + "epoch": 0.8424042927357172, + "flos": 18744460439040.0, + "grad_norm": 2.0511939223932933, + "learning_rate": 2.548117688589628e-07, + "loss": 0.8015, + "num_input_tokens_seen": 303781680, + "step": 9341 + }, + { + "epoch": 0.8424944762591875, + "flos": 24682483610880.0, + "grad_norm": 2.0652213205697327, + "learning_rate": 2.545264879643152e-07, + "loss": 0.7237, + "num_input_tokens_seen": 303811835, + "step": 9342 + }, + { + "epoch": 0.8425846597826577, + "flos": 23407099978560.0, + "grad_norm": 1.817842502251318, + "learning_rate": 2.542413560030923e-07, + "loss": 0.697, + "num_input_tokens_seen": 303838970, + "step": 9343 + }, + { + "epoch": 0.842674843306128, + "flos": 23658928068000.0, + "grad_norm": 1.7814041204868902, + "learning_rate": 2.53956372999623e-07, + "loss": 0.6332, + "num_input_tokens_seen": 303868435, + "step": 9344 + }, + { + "epoch": 0.8427650268295982, + "flos": 69899729909280.0, + "grad_norm": 0.6372181858533366, + "learning_rate": 2.5367153897822293e-07, + "loss": 0.6082, + "num_input_tokens_seen": 303964635, + "step": 9345 + }, + { + "epoch": 0.8428552103530685, + "flos": 25264244922720.0, + "grad_norm": 2.1267211899698895, + "learning_rate": 2.5338685396319715e-07, + "loss": 0.7323, + "num_input_tokens_seen": 303994945, + "step": 9346 + }, + { + "epoch": 0.8429453938765388, + "flos": 28471161656160.0, + "grad_norm": 2.5211007367850873, + "learning_rate": 2.531023179788352e-07, + "loss": 0.6762, + "num_input_tokens_seen": 304023190, + "step": 9347 + }, + { + "epoch": 0.843035577400009, + "flos": 26613161807520.0, + "grad_norm": 2.926057647158064, + "learning_rate": 2.528179310494158e-07, + "loss": 0.7113, + "num_input_tokens_seen": 304053525, + "step": 9348 + }, + { + "epoch": 0.8431257609234792, + "flos": 38271210276960.0, + "grad_norm": 1.7248776876236653, + "learning_rate": 2.5253369319920436e-07, + "loss": 0.7594, + "num_input_tokens_seen": 304087310, + "step": 9349 + }, + { + "epoch": 0.8432159444469496, + "flos": 21731892507840.0, + "grad_norm": 2.3475909485835342, + "learning_rate": 2.522496044524538e-07, + "loss": 0.7404, + "num_input_tokens_seen": 304114745, + "step": 9350 + }, + { + "epoch": 0.8433061279704198, + "flos": 12586841093280.0, + "grad_norm": 2.4899844773509594, + "learning_rate": 2.5196566483340386e-07, + "loss": 0.8292, + "num_input_tokens_seen": 304139595, + "step": 9351 + }, + { + "epoch": 0.84339631149389, + "flos": 26756920203840.0, + "grad_norm": 5.777721090935307, + "learning_rate": 2.516818743662825e-07, + "loss": 0.7546, + "num_input_tokens_seen": 304169000, + "step": 9352 + }, + { + "epoch": 0.8434864950173603, + "flos": 12871161286560.0, + "grad_norm": 2.136466002640396, + "learning_rate": 2.5139823307530285e-07, + "loss": 0.7374, + "num_input_tokens_seen": 304194060, + "step": 9353 + }, + { + "epoch": 0.8435766785408306, + "flos": 25629680999520.0, + "grad_norm": 2.3220236277597945, + "learning_rate": 2.5111474098466836e-07, + "loss": 0.7762, + "num_input_tokens_seen": 304221610, + "step": 9354 + }, + { + "epoch": 0.8436668620643009, + "flos": 22273393236000.0, + "grad_norm": 1.7229814633320464, + "learning_rate": 2.50831398118567e-07, + "loss": 0.7849, + "num_input_tokens_seen": 304250130, + "step": 9355 + }, + { + "epoch": 0.8437570455877711, + "flos": 17577815555520.0, + "grad_norm": 1.969867933852127, + "learning_rate": 2.5054820450117576e-07, + "loss": 0.6855, + "num_input_tokens_seen": 304276250, + "step": 9356 + }, + { + "epoch": 0.8438472291112413, + "flos": 25627339304640.0, + "grad_norm": 2.517476034646998, + "learning_rate": 2.502651601566579e-07, + "loss": 0.7031, + "num_input_tokens_seen": 304303825, + "step": 9357 + }, + { + "epoch": 0.8439374126347117, + "flos": 25593100222080.0, + "grad_norm": 2.332910011128941, + "learning_rate": 2.499822651091645e-07, + "loss": 0.6907, + "num_input_tokens_seen": 304331575, + "step": 9358 + }, + { + "epoch": 0.8440275961581819, + "flos": 17793248716320.0, + "grad_norm": 2.303295581561615, + "learning_rate": 2.496995193828344e-07, + "loss": 0.7595, + "num_input_tokens_seen": 304358295, + "step": 9359 + }, + { + "epoch": 0.8441177796816521, + "flos": 24893493570240.0, + "grad_norm": 2.0309169789733352, + "learning_rate": 2.494169230017913e-07, + "loss": 0.738, + "num_input_tokens_seen": 304388865, + "step": 9360 + }, + { + "epoch": 0.8442079632051224, + "flos": 24132136479840.0, + "grad_norm": 1.7068402425718465, + "learning_rate": 2.491344759901499e-07, + "loss": 0.6964, + "num_input_tokens_seen": 304417655, + "step": 9361 + }, + { + "epoch": 0.8442981467285927, + "flos": 27086853426240.0, + "grad_norm": 2.1183178126108957, + "learning_rate": 2.488521783720088e-07, + "loss": 0.7314, + "num_input_tokens_seen": 304451125, + "step": 9362 + }, + { + "epoch": 0.844388330252063, + "flos": 30548311641600.0, + "grad_norm": 2.0957274706618274, + "learning_rate": 2.4857003017145526e-07, + "loss": 0.7742, + "num_input_tokens_seen": 304479625, + "step": 9363 + }, + { + "epoch": 0.8444785137755332, + "flos": 18343186980000.0, + "grad_norm": 1.984530825766934, + "learning_rate": 2.482880314125644e-07, + "loss": 0.7782, + "num_input_tokens_seen": 304506190, + "step": 9364 + }, + { + "epoch": 0.8445686972990035, + "flos": 30110494574880.0, + "grad_norm": 1.7433244220051232, + "learning_rate": 2.4800618211939726e-07, + "loss": 0.7312, + "num_input_tokens_seen": 304535010, + "step": 9365 + }, + { + "epoch": 0.8446588808224738, + "flos": 18671224544640.0, + "grad_norm": 1.9756862220823985, + "learning_rate": 2.477244823160034e-07, + "loss": 0.7487, + "num_input_tokens_seen": 304561820, + "step": 9366 + }, + { + "epoch": 0.844749064345944, + "flos": 23953507115040.0, + "grad_norm": 2.2651505598338977, + "learning_rate": 2.474429320264184e-07, + "loss": 0.7561, + "num_input_tokens_seen": 304591235, + "step": 9367 + }, + { + "epoch": 0.8448392478694142, + "flos": 17687706566880.0, + "grad_norm": 1.6248005964739187, + "learning_rate": 2.47161531274666e-07, + "loss": 0.6972, + "num_input_tokens_seen": 304619455, + "step": 9368 + }, + { + "epoch": 0.8449294313928846, + "flos": 43950528953760.0, + "grad_norm": 1.8446778371261536, + "learning_rate": 2.4688028008475714e-07, + "loss": 0.6714, + "num_input_tokens_seen": 304651760, + "step": 9369 + }, + { + "epoch": 0.8450196149163548, + "flos": 23438216801280.0, + "grad_norm": 2.2249696622344697, + "learning_rate": 2.465991784806891e-07, + "loss": 0.7317, + "num_input_tokens_seen": 304677195, + "step": 9370 + }, + { + "epoch": 0.845109798439825, + "flos": 16228452633600.0, + "grad_norm": 2.1095797969158845, + "learning_rate": 2.463182264864472e-07, + "loss": 0.8167, + "num_input_tokens_seen": 304703555, + "step": 9371 + }, + { + "epoch": 0.8451999819632953, + "flos": 21622038666240.0, + "grad_norm": 2.245268615421217, + "learning_rate": 2.460374241260039e-07, + "loss": 0.6666, + "num_input_tokens_seen": 304731125, + "step": 9372 + }, + { + "epoch": 0.8452901654867656, + "flos": 17687372039040.0, + "grad_norm": 2.309719372832508, + "learning_rate": 2.4575677142331884e-07, + "loss": 0.8054, + "num_input_tokens_seen": 304759685, + "step": 9373 + }, + { + "epoch": 0.8453803490102358, + "flos": 23735583580320.0, + "grad_norm": 2.0519642932529094, + "learning_rate": 2.454762684023395e-07, + "loss": 0.6807, + "num_input_tokens_seen": 304790925, + "step": 9374 + }, + { + "epoch": 0.8454705325337061, + "flos": 25630015527360.0, + "grad_norm": 2.492792133899907, + "learning_rate": 2.4519591508699823e-07, + "loss": 0.6862, + "num_input_tokens_seen": 304819450, + "step": 9375 + }, + { + "epoch": 0.8455607160571763, + "flos": 17323125394560.0, + "grad_norm": 2.069508061579698, + "learning_rate": 2.4491571150121815e-07, + "loss": 0.785, + "num_input_tokens_seen": 304845560, + "step": 9376 + }, + { + "epoch": 0.8456508995806467, + "flos": 27085626824160.0, + "grad_norm": 2.0966314728678523, + "learning_rate": 2.446356576689062e-07, + "loss": 0.737, + "num_input_tokens_seen": 304875475, + "step": 9377 + }, + { + "epoch": 0.8457410831041169, + "flos": 26900269732800.0, + "grad_norm": 2.988168272986906, + "learning_rate": 2.4435575361395976e-07, + "loss": 0.7461, + "num_input_tokens_seen": 304903895, + "step": 9378 + }, + { + "epoch": 0.8458312666275871, + "flos": 21731743828800.0, + "grad_norm": 1.9935472223847137, + "learning_rate": 2.440759993602606e-07, + "loss": 0.7154, + "num_input_tokens_seen": 304933990, + "step": 9379 + }, + { + "epoch": 0.8459214501510574, + "flos": 21732524393760.0, + "grad_norm": 2.6289258392193635, + "learning_rate": 2.437963949316793e-07, + "loss": 0.6802, + "num_input_tokens_seen": 304963665, + "step": 9380 + }, + { + "epoch": 0.8460116336745277, + "flos": 23366727885600.0, + "grad_norm": 2.097182463661346, + "learning_rate": 2.435169403520729e-07, + "loss": 0.7326, + "num_input_tokens_seen": 304989890, + "step": 9381 + }, + { + "epoch": 0.8461018171979979, + "flos": 25587227400000.0, + "grad_norm": 1.9916036311220502, + "learning_rate": 2.4323763564528653e-07, + "loss": 0.7847, + "num_input_tokens_seen": 305017660, + "step": 9382 + }, + { + "epoch": 0.8461920007214682, + "flos": 24825164084160.0, + "grad_norm": 2.131566057234389, + "learning_rate": 2.429584808351517e-07, + "loss": 0.7568, + "num_input_tokens_seen": 305046885, + "step": 9383 + }, + { + "epoch": 0.8462821842449384, + "flos": 17388667148640.0, + "grad_norm": 2.6174776286541834, + "learning_rate": 2.42679475945488e-07, + "loss": 0.862, + "num_input_tokens_seen": 305069860, + "step": 9384 + }, + { + "epoch": 0.8463723677684087, + "flos": 39947086803360.0, + "grad_norm": 1.5914007337546685, + "learning_rate": 2.424006210001008e-07, + "loss": 0.7367, + "num_input_tokens_seen": 305104275, + "step": 9385 + }, + { + "epoch": 0.846462551291879, + "flos": 30801849540000.0, + "grad_norm": 1.8172148186687591, + "learning_rate": 2.421219160227839e-07, + "loss": 0.7631, + "num_input_tokens_seen": 305135350, + "step": 9386 + }, + { + "epoch": 0.8465527348153492, + "flos": 24317642250240.0, + "grad_norm": 1.7596752554227497, + "learning_rate": 2.4184336103731785e-07, + "loss": 0.7323, + "num_input_tokens_seen": 305167265, + "step": 9387 + }, + { + "epoch": 0.8466429183388196, + "flos": 19181942977920.0, + "grad_norm": 2.4624402498848434, + "learning_rate": 2.4156495606747065e-07, + "loss": 0.7597, + "num_input_tokens_seen": 305193625, + "step": 9388 + }, + { + "epoch": 0.8467331018622898, + "flos": 23370259012800.0, + "grad_norm": 2.268842740435938, + "learning_rate": 2.412867011369972e-07, + "loss": 0.8566, + "num_input_tokens_seen": 305221190, + "step": 9389 + }, + { + "epoch": 0.84682328538576, + "flos": 26503568154240.0, + "grad_norm": 1.8290498138618683, + "learning_rate": 2.4100859626963997e-07, + "loss": 0.7464, + "num_input_tokens_seen": 305250140, + "step": 9390 + }, + { + "epoch": 0.8469134689092302, + "flos": 26941794088320.0, + "grad_norm": 6.128427908655841, + "learning_rate": 2.407306414891288e-07, + "loss": 0.7376, + "num_input_tokens_seen": 305282330, + "step": 9391 + }, + { + "epoch": 0.8470036524327006, + "flos": 31311564389760.0, + "grad_norm": 1.5375485792471475, + "learning_rate": 2.4045283681917893e-07, + "loss": 0.8406, + "num_input_tokens_seen": 305313215, + "step": 9392 + }, + { + "epoch": 0.8470938359561708, + "flos": 25957086678240.0, + "grad_norm": 1.8794372189025477, + "learning_rate": 2.4017518228349586e-07, + "loss": 0.8529, + "num_input_tokens_seen": 305342265, + "step": 9393 + }, + { + "epoch": 0.8471840194796411, + "flos": 66797314572000.0, + "grad_norm": 0.622417335192046, + "learning_rate": 2.3989767790576887e-07, + "loss": 0.5908, + "num_input_tokens_seen": 305437885, + "step": 9394 + }, + { + "epoch": 0.8472742030031113, + "flos": 22860767181600.0, + "grad_norm": 2.106314692649856, + "learning_rate": 2.396203237096781e-07, + "loss": 0.7081, + "num_input_tokens_seen": 305466850, + "step": 9395 + }, + { + "epoch": 0.8473643865265816, + "flos": 30329384523360.0, + "grad_norm": 2.0734725155780738, + "learning_rate": 2.393431197188873e-07, + "loss": 0.7321, + "num_input_tokens_seen": 305496725, + "step": 9396 + }, + { + "epoch": 0.8474545700500519, + "flos": 18413597972640.0, + "grad_norm": 2.598018213738313, + "learning_rate": 2.3906606595705004e-07, + "loss": 0.8117, + "num_input_tokens_seen": 305522050, + "step": 9397 + }, + { + "epoch": 0.8475447535735221, + "flos": 22383284247360.0, + "grad_norm": 1.9812661395934819, + "learning_rate": 2.387891624478056e-07, + "loss": 0.7933, + "num_input_tokens_seen": 305550325, + "step": 9398 + }, + { + "epoch": 0.8476349370969923, + "flos": 22059335356320.0, + "grad_norm": 2.0020787641596844, + "learning_rate": 2.3851240921478075e-07, + "loss": 0.7589, + "num_input_tokens_seen": 305577070, + "step": 9399 + }, + { + "epoch": 0.8477251206204627, + "flos": 26795693997120.0, + "grad_norm": 2.1382472065176232, + "learning_rate": 2.3823580628159057e-07, + "loss": 0.7193, + "num_input_tokens_seen": 305606425, + "step": 9400 + }, + { + "epoch": 0.8478153041439329, + "flos": 21622744891680.0, + "grad_norm": 2.2469313146248155, + "learning_rate": 2.3795935367183517e-07, + "loss": 0.7212, + "num_input_tokens_seen": 305632790, + "step": 9401 + }, + { + "epoch": 0.8479054876674031, + "flos": 25921658163360.0, + "grad_norm": 1.9545287923929309, + "learning_rate": 2.376830514091035e-07, + "loss": 0.7604, + "num_input_tokens_seen": 305663415, + "step": 9402 + }, + { + "epoch": 0.8479956711908734, + "flos": 25375139517600.0, + "grad_norm": 1.8037270636630227, + "learning_rate": 2.3740689951697135e-07, + "loss": 0.8222, + "num_input_tokens_seen": 305694070, + "step": 9403 + }, + { + "epoch": 0.8480858547143437, + "flos": 21870670156320.0, + "grad_norm": 1.6816479265185202, + "learning_rate": 2.371308980190012e-07, + "loss": 0.7384, + "num_input_tokens_seen": 305721820, + "step": 9404 + }, + { + "epoch": 0.848176038237814, + "flos": 61731803273760.0, + "grad_norm": 0.622340432631577, + "learning_rate": 2.3685504693874337e-07, + "loss": 0.5971, + "num_input_tokens_seen": 305814660, + "step": 9405 + }, + { + "epoch": 0.8482662217612842, + "flos": 27414667972320.0, + "grad_norm": 2.1919462521221598, + "learning_rate": 2.3657934629973497e-07, + "loss": 0.7748, + "num_input_tokens_seen": 305842200, + "step": 9406 + }, + { + "epoch": 0.8483564052847544, + "flos": 12441930434400.0, + "grad_norm": 2.0534780445545975, + "learning_rate": 2.3630379612549944e-07, + "loss": 0.7464, + "num_input_tokens_seen": 305868990, + "step": 9407 + }, + { + "epoch": 0.8484465888082248, + "flos": 69642698053440.0, + "grad_norm": 0.7352139976271016, + "learning_rate": 2.3602839643954997e-07, + "loss": 0.6201, + "num_input_tokens_seen": 305964645, + "step": 9408 + }, + { + "epoch": 0.848536772331695, + "flos": 22387893297600.0, + "grad_norm": 1.9395531266506891, + "learning_rate": 2.3575314726538308e-07, + "loss": 0.7428, + "num_input_tokens_seen": 305994720, + "step": 9409 + }, + { + "epoch": 0.8486269558551652, + "flos": 23549520263520.0, + "grad_norm": 2.273819933838697, + "learning_rate": 2.3547804862648645e-07, + "loss": 0.7754, + "num_input_tokens_seen": 306023655, + "step": 9410 + }, + { + "epoch": 0.8487171393786356, + "flos": 28107138030240.0, + "grad_norm": 1.9471019802374765, + "learning_rate": 2.3520310054633174e-07, + "loss": 0.7859, + "num_input_tokens_seen": 306055575, + "step": 9411 + }, + { + "epoch": 0.8488073229021058, + "flos": 25046841764640.0, + "grad_norm": 1.8735567260168633, + "learning_rate": 2.3492830304837973e-07, + "loss": 0.7017, + "num_input_tokens_seen": 306083300, + "step": 9412 + }, + { + "epoch": 0.848897506425576, + "flos": 23915142189120.0, + "grad_norm": 2.0324981376010673, + "learning_rate": 2.3465365615607723e-07, + "loss": 0.6855, + "num_input_tokens_seen": 306115485, + "step": 9413 + }, + { + "epoch": 0.8489876899490463, + "flos": 22350569124960.0, + "grad_norm": 2.121937733145619, + "learning_rate": 2.3437915989285884e-07, + "loss": 0.7325, + "num_input_tokens_seen": 306144845, + "step": 9414 + }, + { + "epoch": 0.8490778734725166, + "flos": 17177025303360.0, + "grad_norm": 2.0571723098576524, + "learning_rate": 2.3410481428214647e-07, + "loss": 0.8027, + "num_input_tokens_seen": 306171520, + "step": 9415 + }, + { + "epoch": 0.8491680569959869, + "flos": 25441759194720.0, + "grad_norm": 1.7421159943416766, + "learning_rate": 2.338306193473476e-07, + "loss": 0.7662, + "num_input_tokens_seen": 306199260, + "step": 9416 + }, + { + "epoch": 0.8492582405194571, + "flos": 39729386287200.0, + "grad_norm": 1.8400491328521134, + "learning_rate": 2.3355657511185957e-07, + "loss": 0.7416, + "num_input_tokens_seen": 306231345, + "step": 9417 + }, + { + "epoch": 0.8493484240429273, + "flos": 63884159150880.0, + "grad_norm": 0.659718159334042, + "learning_rate": 2.3328268159906428e-07, + "loss": 0.5621, + "num_input_tokens_seen": 306318585, + "step": 9418 + }, + { + "epoch": 0.8494386075663977, + "flos": 20310260105280.0, + "grad_norm": 22.73205797900704, + "learning_rate": 2.330089388323322e-07, + "loss": 0.7891, + "num_input_tokens_seen": 306346760, + "step": 9419 + }, + { + "epoch": 0.8495287910898679, + "flos": 23620191444480.0, + "grad_norm": 1.9103165697715587, + "learning_rate": 2.327353468350204e-07, + "loss": 0.7794, + "num_input_tokens_seen": 306374730, + "step": 9420 + }, + { + "epoch": 0.8496189746133381, + "flos": 33608459228160.0, + "grad_norm": 1.900110708324107, + "learning_rate": 2.3246190563047352e-07, + "loss": 0.7319, + "num_input_tokens_seen": 306404675, + "step": 9421 + }, + { + "epoch": 0.8497091581368084, + "flos": 23407360166880.0, + "grad_norm": 1.782744867912569, + "learning_rate": 2.3218861524202293e-07, + "loss": 0.726, + "num_input_tokens_seen": 306433035, + "step": 9422 + }, + { + "epoch": 0.8497993416602787, + "flos": 26682420537600.0, + "grad_norm": 2.5247771778557104, + "learning_rate": 2.3191547569298775e-07, + "loss": 0.7507, + "num_input_tokens_seen": 306462090, + "step": 9423 + }, + { + "epoch": 0.8498895251837489, + "flos": 22308896090400.0, + "grad_norm": 4.21550937095291, + "learning_rate": 2.3164248700667245e-07, + "loss": 0.7265, + "num_input_tokens_seen": 306488655, + "step": 9424 + }, + { + "epoch": 0.8499797087072192, + "flos": 23077538453760.0, + "grad_norm": 1.807611449988362, + "learning_rate": 2.313696492063717e-07, + "loss": 0.7745, + "num_input_tokens_seen": 306519385, + "step": 9425 + }, + { + "epoch": 0.8500698922306894, + "flos": 18780595179360.0, + "grad_norm": 2.388765796427495, + "learning_rate": 2.3109696231536401e-07, + "loss": 0.7292, + "num_input_tokens_seen": 306546880, + "step": 9426 + }, + { + "epoch": 0.8501600757541597, + "flos": 24682409271360.0, + "grad_norm": 1.9693440343419288, + "learning_rate": 2.3082442635691722e-07, + "loss": 0.7, + "num_input_tokens_seen": 306576240, + "step": 9427 + }, + { + "epoch": 0.85025025927763, + "flos": 19213654516800.0, + "grad_norm": 2.965472530388911, + "learning_rate": 2.305520413542854e-07, + "loss": 0.7544, + "num_input_tokens_seen": 306604315, + "step": 9428 + }, + { + "epoch": 0.8503404428011002, + "flos": 27159866302080.0, + "grad_norm": 1.6281436443176573, + "learning_rate": 2.3027980733071018e-07, + "loss": 0.6645, + "num_input_tokens_seen": 306635890, + "step": 9429 + }, + { + "epoch": 0.8504306263245704, + "flos": 65965657998240.0, + "grad_norm": 0.5947750178299054, + "learning_rate": 2.3000772430942027e-07, + "loss": 0.5444, + "num_input_tokens_seen": 306737310, + "step": 9430 + }, + { + "epoch": 0.8505208098480408, + "flos": 21913458283680.0, + "grad_norm": 2.7390022580213036, + "learning_rate": 2.2973579231363028e-07, + "loss": 0.7546, + "num_input_tokens_seen": 306764905, + "step": 9431 + }, + { + "epoch": 0.850610993371511, + "flos": 17505546074880.0, + "grad_norm": 2.0052225351862814, + "learning_rate": 2.2946401136654446e-07, + "loss": 0.7651, + "num_input_tokens_seen": 306793630, + "step": 9432 + }, + { + "epoch": 0.8507011768949813, + "flos": 21258238058880.0, + "grad_norm": 2.5640442079848174, + "learning_rate": 2.2919238149135077e-07, + "loss": 0.7681, + "num_input_tokens_seen": 306821800, + "step": 9433 + }, + { + "epoch": 0.8507913604184516, + "flos": 24937508299680.0, + "grad_norm": 1.7696458387851548, + "learning_rate": 2.289209027112282e-07, + "loss": 0.7432, + "num_input_tokens_seen": 306852595, + "step": 9434 + }, + { + "epoch": 0.8508815439419218, + "flos": 25119334263840.0, + "grad_norm": 4.375418288806953, + "learning_rate": 2.2864957504933934e-07, + "loss": 0.7992, + "num_input_tokens_seen": 306878155, + "step": 9435 + }, + { + "epoch": 0.8509717274653921, + "flos": 23876591414400.0, + "grad_norm": 2.328252084834958, + "learning_rate": 2.2837839852883589e-07, + "loss": 0.7221, + "num_input_tokens_seen": 306908040, + "step": 9436 + }, + { + "epoch": 0.8510619109888623, + "flos": 24571365997440.0, + "grad_norm": 1.820279951991571, + "learning_rate": 2.2810737317285623e-07, + "loss": 0.736, + "num_input_tokens_seen": 306936080, + "step": 9437 + }, + { + "epoch": 0.8511520945123326, + "flos": 37506619417440.0, + "grad_norm": 2.4048621398686625, + "learning_rate": 2.278364990045254e-07, + "loss": 0.6803, + "num_input_tokens_seen": 306967040, + "step": 9438 + }, + { + "epoch": 0.8512422780358029, + "flos": 22642360440000.0, + "grad_norm": 2.2320463410355695, + "learning_rate": 2.2756577604695625e-07, + "loss": 0.7303, + "num_input_tokens_seen": 306994890, + "step": 9439 + }, + { + "epoch": 0.8513324615592731, + "flos": 22821844709280.0, + "grad_norm": 4.212290247554016, + "learning_rate": 2.2729520432324855e-07, + "loss": 0.7574, + "num_input_tokens_seen": 307024890, + "step": 9440 + }, + { + "epoch": 0.8514226450827433, + "flos": 28071263478240.0, + "grad_norm": 2.3562565082490523, + "learning_rate": 2.2702478385648826e-07, + "loss": 0.6554, + "num_input_tokens_seen": 307053765, + "step": 9441 + }, + { + "epoch": 0.8515128286062137, + "flos": 25593880787040.0, + "grad_norm": 1.8916434075134152, + "learning_rate": 2.2675451466974938e-07, + "loss": 0.7355, + "num_input_tokens_seen": 307082875, + "step": 9442 + }, + { + "epoch": 0.8516030121296839, + "flos": 33135064967520.0, + "grad_norm": 2.2130687133560496, + "learning_rate": 2.26484396786093e-07, + "loss": 0.6724, + "num_input_tokens_seen": 307112205, + "step": 9443 + }, + { + "epoch": 0.8516931956531542, + "flos": 57359282410080.0, + "grad_norm": 0.6470181891775076, + "learning_rate": 2.2621443022856667e-07, + "loss": 0.6019, + "num_input_tokens_seen": 307195395, + "step": 9444 + }, + { + "epoch": 0.8517833791766244, + "flos": 24499356705120.0, + "grad_norm": 1.8523586830879049, + "learning_rate": 2.2594461502020646e-07, + "loss": 0.7818, + "num_input_tokens_seen": 307225650, + "step": 9445 + }, + { + "epoch": 0.8518735627000947, + "flos": 30470652545760.0, + "grad_norm": 2.008263563260625, + "learning_rate": 2.2567495118403278e-07, + "loss": 0.619, + "num_input_tokens_seen": 307255160, + "step": 9446 + }, + { + "epoch": 0.851963746223565, + "flos": 19363434414240.0, + "grad_norm": 1.6499243083122905, + "learning_rate": 2.254054387430566e-07, + "loss": 0.7531, + "num_input_tokens_seen": 307283590, + "step": 9447 + }, + { + "epoch": 0.8520539297470352, + "flos": 21986396820000.0, + "grad_norm": 2.1601868225472733, + "learning_rate": 2.2513607772027243e-07, + "loss": 0.7257, + "num_input_tokens_seen": 307312420, + "step": 9448 + }, + { + "epoch": 0.8521441132705054, + "flos": 21140727246720.0, + "grad_norm": 2.4707374781455607, + "learning_rate": 2.2486686813866562e-07, + "loss": 0.7015, + "num_input_tokens_seen": 307338110, + "step": 9449 + }, + { + "epoch": 0.8522342967939758, + "flos": 21076263415680.0, + "grad_norm": 1.9637842070710132, + "learning_rate": 2.2459781002120514e-07, + "loss": 0.7912, + "num_input_tokens_seen": 307365920, + "step": 9450 + }, + { + "epoch": 0.852324480317446, + "flos": 28981657070880.0, + "grad_norm": 1.5995103295372195, + "learning_rate": 2.243289033908491e-07, + "loss": 0.8081, + "num_input_tokens_seen": 307396130, + "step": 9451 + }, + { + "epoch": 0.8524146638409162, + "flos": 20420522814240.0, + "grad_norm": 2.192637704420007, + "learning_rate": 2.2406014827054176e-07, + "loss": 0.7522, + "num_input_tokens_seen": 307423515, + "step": 9452 + }, + { + "epoch": 0.8525048473643865, + "flos": 22642100251680.0, + "grad_norm": 4.139137778312237, + "learning_rate": 2.2379154468321525e-07, + "loss": 0.8042, + "num_input_tokens_seen": 307450195, + "step": 9453 + }, + { + "epoch": 0.8525950308878568, + "flos": 26941310881440.0, + "grad_norm": 1.6148476743165987, + "learning_rate": 2.2352309265178793e-07, + "loss": 0.7743, + "num_input_tokens_seen": 307481615, + "step": 9454 + }, + { + "epoch": 0.852685214411327, + "flos": 22751508056160.0, + "grad_norm": 2.3059740516545926, + "learning_rate": 2.2325479219916565e-07, + "loss": 0.8117, + "num_input_tokens_seen": 307501815, + "step": 9455 + }, + { + "epoch": 0.8527753979347973, + "flos": 27597646199040.0, + "grad_norm": 1.861397612836747, + "learning_rate": 2.229866433482419e-07, + "loss": 0.7735, + "num_input_tokens_seen": 307529790, + "step": 9456 + }, + { + "epoch": 0.8528655814582676, + "flos": 19472805048960.0, + "grad_norm": 1.909806233109124, + "learning_rate": 2.2271864612189552e-07, + "loss": 0.7914, + "num_input_tokens_seen": 307558105, + "step": 9457 + }, + { + "epoch": 0.8529557649817379, + "flos": 26098837907520.0, + "grad_norm": 1.81492005941929, + "learning_rate": 2.2245080054299415e-07, + "loss": 0.7718, + "num_input_tokens_seen": 307586720, + "step": 9458 + }, + { + "epoch": 0.8530459485052081, + "flos": 42455697826560.0, + "grad_norm": 1.9450303451406783, + "learning_rate": 2.2218310663439198e-07, + "loss": 0.6904, + "num_input_tokens_seen": 307616510, + "step": 9459 + }, + { + "epoch": 0.8531361320286783, + "flos": 45193717840320.0, + "grad_norm": 1.948932145019393, + "learning_rate": 2.2191556441892968e-07, + "loss": 0.7648, + "num_input_tokens_seen": 307647890, + "step": 9460 + }, + { + "epoch": 0.8532263155521487, + "flos": 24424373832000.0, + "grad_norm": 1.7684262764630687, + "learning_rate": 2.216481739194358e-07, + "loss": 0.7448, + "num_input_tokens_seen": 307679275, + "step": 9461 + }, + { + "epoch": 0.8533164990756189, + "flos": 25002455337600.0, + "grad_norm": 2.218685202928438, + "learning_rate": 2.2138093515872592e-07, + "loss": 0.6912, + "num_input_tokens_seen": 307707260, + "step": 9462 + }, + { + "epoch": 0.8534066825990891, + "flos": 24024847351680.0, + "grad_norm": 1.5675670099612948, + "learning_rate": 2.2111384815960087e-07, + "loss": 0.8211, + "num_input_tokens_seen": 307737550, + "step": 9463 + }, + { + "epoch": 0.8534968661225594, + "flos": 18561928249440.0, + "grad_norm": 1.8630395888515885, + "learning_rate": 2.208469129448518e-07, + "loss": 0.6662, + "num_input_tokens_seen": 307765640, + "step": 9464 + }, + { + "epoch": 0.8535870496460297, + "flos": 65636133643200.0, + "grad_norm": 0.6223126093768662, + "learning_rate": 2.2058012953725357e-07, + "loss": 0.5676, + "num_input_tokens_seen": 307856635, + "step": 9465 + }, + { + "epoch": 0.8536772331695, + "flos": 21363668699040.0, + "grad_norm": 2.222438593768131, + "learning_rate": 2.203134979595711e-07, + "loss": 0.7099, + "num_input_tokens_seen": 307882960, + "step": 9466 + }, + { + "epoch": 0.8537674166929702, + "flos": 26390369034240.0, + "grad_norm": 1.7708801113756873, + "learning_rate": 2.2004701823455374e-07, + "loss": 0.6901, + "num_input_tokens_seen": 307913460, + "step": 9467 + }, + { + "epoch": 0.8538576002164404, + "flos": 22860841521120.0, + "grad_norm": 1.9251359396161059, + "learning_rate": 2.1978069038493906e-07, + "loss": 0.7718, + "num_input_tokens_seen": 307942215, + "step": 9468 + }, + { + "epoch": 0.8539477837399108, + "flos": 20638520688480.0, + "grad_norm": 1.4742535496553042, + "learning_rate": 2.1951451443345225e-07, + "loss": 0.7427, + "num_input_tokens_seen": 307970425, + "step": 9469 + }, + { + "epoch": 0.854037967263381, + "flos": 24463853850720.0, + "grad_norm": 2.5020547052588418, + "learning_rate": 2.1924849040280425e-07, + "loss": 0.7279, + "num_input_tokens_seen": 307998510, + "step": 9470 + }, + { + "epoch": 0.8541281507868512, + "flos": 42861914863680.0, + "grad_norm": 3.0484114465426266, + "learning_rate": 2.1898261831569465e-07, + "loss": 0.8254, + "num_input_tokens_seen": 308024590, + "step": 9471 + }, + { + "epoch": 0.8542183343103215, + "flos": 24974312095680.0, + "grad_norm": 1.7504412231078825, + "learning_rate": 2.1871689819480798e-07, + "loss": 0.7518, + "num_input_tokens_seen": 308055205, + "step": 9472 + }, + { + "epoch": 0.8543085178337918, + "flos": 18670927186560.0, + "grad_norm": 8.163553176397345, + "learning_rate": 2.1845133006281745e-07, + "loss": 0.7268, + "num_input_tokens_seen": 308084380, + "step": 9473 + }, + { + "epoch": 0.854398701357262, + "flos": 59181673628640.0, + "grad_norm": 1.5028147514074146, + "learning_rate": 2.1818591394238294e-07, + "loss": 0.6019, + "num_input_tokens_seen": 308120465, + "step": 9474 + }, + { + "epoch": 0.8544888848807323, + "flos": 28285023999840.0, + "grad_norm": 3.3928995545683542, + "learning_rate": 2.1792064985615076e-07, + "loss": 0.7238, + "num_input_tokens_seen": 308148695, + "step": 9475 + }, + { + "epoch": 0.8545790684042025, + "flos": 27411917410080.0, + "grad_norm": 1.6017170156979867, + "learning_rate": 2.1765553782675528e-07, + "loss": 0.7117, + "num_input_tokens_seen": 308178010, + "step": 9476 + }, + { + "epoch": 0.8546692519276728, + "flos": 24824234840160.0, + "grad_norm": 2.018563545993806, + "learning_rate": 2.1739057787681703e-07, + "loss": 0.8154, + "num_input_tokens_seen": 308208070, + "step": 9477 + }, + { + "epoch": 0.8547594354511431, + "flos": 20419370551680.0, + "grad_norm": 1.889751579367781, + "learning_rate": 2.1712577002894372e-07, + "loss": 0.7624, + "num_input_tokens_seen": 308235360, + "step": 9478 + }, + { + "epoch": 0.8548496189746133, + "flos": 30469574622720.0, + "grad_norm": 1.7536332697503747, + "learning_rate": 2.1686111430573105e-07, + "loss": 0.7151, + "num_input_tokens_seen": 308266215, + "step": 9479 + }, + { + "epoch": 0.8549398024980835, + "flos": 19035694207680.0, + "grad_norm": 1.9126928917842143, + "learning_rate": 2.165966107297592e-07, + "loss": 0.794, + "num_input_tokens_seen": 308295685, + "step": 9480 + }, + { + "epoch": 0.8550299860215539, + "flos": 21914127339360.0, + "grad_norm": 2.6269827121358906, + "learning_rate": 2.16332259323599e-07, + "loss": 0.7669, + "num_input_tokens_seen": 308324115, + "step": 9481 + }, + { + "epoch": 0.8551201695450241, + "flos": 16885531346400.0, + "grad_norm": 3.8551001715602244, + "learning_rate": 2.1606806010980504e-07, + "loss": 0.6891, + "num_input_tokens_seen": 308348355, + "step": 9482 + }, + { + "epoch": 0.8552103530684944, + "flos": 21475827065760.0, + "grad_norm": 1.9607016128976218, + "learning_rate": 2.1580401311092067e-07, + "loss": 0.74, + "num_input_tokens_seen": 308376230, + "step": 9483 + }, + { + "epoch": 0.8553005365919647, + "flos": 15100618713120.0, + "grad_norm": 1.6949667863770626, + "learning_rate": 2.1554011834947604e-07, + "loss": 0.7295, + "num_input_tokens_seen": 308402945, + "step": 9484 + }, + { + "epoch": 0.8553907201154349, + "flos": 27524298795360.0, + "grad_norm": 2.3072121450757956, + "learning_rate": 2.1527637584798764e-07, + "loss": 0.771, + "num_input_tokens_seen": 308433155, + "step": 9485 + }, + { + "epoch": 0.8554809036389052, + "flos": 24605456400960.0, + "grad_norm": 1.840751407861592, + "learning_rate": 2.150127856289603e-07, + "loss": 0.8067, + "num_input_tokens_seen": 308459035, + "step": 9486 + }, + { + "epoch": 0.8555710871623754, + "flos": 67923698871840.0, + "grad_norm": 0.6143778576013151, + "learning_rate": 2.1474934771488363e-07, + "loss": 0.5562, + "num_input_tokens_seen": 308559220, + "step": 9487 + }, + { + "epoch": 0.8556612706858457, + "flos": 22970769702240.0, + "grad_norm": 1.733327050686723, + "learning_rate": 2.1448606212823715e-07, + "loss": 0.7845, + "num_input_tokens_seen": 308586590, + "step": 9488 + }, + { + "epoch": 0.855751454209316, + "flos": 24426938545440.0, + "grad_norm": 1.9718429853013781, + "learning_rate": 2.1422292889148452e-07, + "loss": 0.7247, + "num_input_tokens_seen": 308616995, + "step": 9489 + }, + { + "epoch": 0.8558416377327862, + "flos": 27123434203680.0, + "grad_norm": 1.7186168272270945, + "learning_rate": 2.139599480270784e-07, + "loss": 0.761, + "num_input_tokens_seen": 308648060, + "step": 9490 + }, + { + "epoch": 0.8559318212562564, + "flos": 23079731469600.0, + "grad_norm": 2.580245746766865, + "learning_rate": 2.1369711955745773e-07, + "loss": 0.7302, + "num_input_tokens_seen": 308674025, + "step": 9491 + }, + { + "epoch": 0.8560220047797268, + "flos": 32733159622560.0, + "grad_norm": 3.500066756565514, + "learning_rate": 2.1343444350504813e-07, + "loss": 0.6932, + "num_input_tokens_seen": 308704825, + "step": 9492 + }, + { + "epoch": 0.856112188303197, + "flos": 15683941154880.0, + "grad_norm": 1.9201705321436473, + "learning_rate": 2.1317191989226302e-07, + "loss": 0.7427, + "num_input_tokens_seen": 308732405, + "step": 9493 + }, + { + "epoch": 0.8562023718266673, + "flos": 22059744223680.0, + "grad_norm": 2.104777319902448, + "learning_rate": 2.129095487415027e-07, + "loss": 0.7248, + "num_input_tokens_seen": 308761390, + "step": 9494 + }, + { + "epoch": 0.8562925553501375, + "flos": 20127876594720.0, + "grad_norm": 2.1439378261305135, + "learning_rate": 2.1264733007515257e-07, + "loss": 0.7506, + "num_input_tokens_seen": 308789325, + "step": 9495 + }, + { + "epoch": 0.8563827388736078, + "flos": 40421075780160.0, + "grad_norm": 3.144790695576088, + "learning_rate": 2.1238526391558852e-07, + "loss": 0.7255, + "num_input_tokens_seen": 308820645, + "step": 9496 + }, + { + "epoch": 0.8564729223970781, + "flos": 17867599703520.0, + "grad_norm": 2.048615509143426, + "learning_rate": 2.1212335028517003e-07, + "loss": 0.8798, + "num_input_tokens_seen": 308844005, + "step": 9497 + }, + { + "epoch": 0.8565631059205483, + "flos": 21840073710240.0, + "grad_norm": 2.0743141511576475, + "learning_rate": 2.1186158920624563e-07, + "loss": 0.7503, + "num_input_tokens_seen": 308872025, + "step": 9498 + }, + { + "epoch": 0.8566532894440185, + "flos": 21765685553280.0, + "grad_norm": 1.84109053108292, + "learning_rate": 2.1159998070115015e-07, + "loss": 0.7406, + "num_input_tokens_seen": 308899225, + "step": 9499 + }, + { + "epoch": 0.8567434729674889, + "flos": 19470723542400.0, + "grad_norm": 2.254446510341987, + "learning_rate": 2.113385247922055e-07, + "loss": 0.7273, + "num_input_tokens_seen": 308927860, + "step": 9500 + }, + { + "epoch": 0.8568336564909591, + "flos": 16332061955520.0, + "grad_norm": 1.7534530495916314, + "learning_rate": 2.1107722150172068e-07, + "loss": 0.7485, + "num_input_tokens_seen": 308951375, + "step": 9501 + }, + { + "epoch": 0.8569238400144293, + "flos": 30911777721120.0, + "grad_norm": 2.5315367924734433, + "learning_rate": 2.108160708519906e-07, + "loss": 0.7789, + "num_input_tokens_seen": 308983115, + "step": 9502 + }, + { + "epoch": 0.8570140235378996, + "flos": 23330593145280.0, + "grad_norm": 2.3607492044874308, + "learning_rate": 2.1055507286529984e-07, + "loss": 0.7197, + "num_input_tokens_seen": 309010445, + "step": 9503 + }, + { + "epoch": 0.8571042070613699, + "flos": 26280998399520.0, + "grad_norm": 2.123676112047653, + "learning_rate": 2.1029422756391612e-07, + "loss": 0.7626, + "num_input_tokens_seen": 309040280, + "step": 9504 + }, + { + "epoch": 0.8571943905848401, + "flos": 40637140826880.0, + "grad_norm": 1.7700673484920375, + "learning_rate": 2.1003353497009812e-07, + "loss": 0.6542, + "num_input_tokens_seen": 309072365, + "step": 9505 + }, + { + "epoch": 0.8572845741083104, + "flos": 24062766240480.0, + "grad_norm": 2.0302732026364705, + "learning_rate": 2.0977299510608825e-07, + "loss": 0.6635, + "num_input_tokens_seen": 309101115, + "step": 9506 + }, + { + "epoch": 0.8573747576317807, + "flos": 23443940944320.0, + "grad_norm": 1.718261094349311, + "learning_rate": 2.0951260799411784e-07, + "loss": 0.7813, + "num_input_tokens_seen": 309130110, + "step": 9507 + }, + { + "epoch": 0.857464941155251, + "flos": 21840779935680.0, + "grad_norm": 1.804255247988263, + "learning_rate": 2.0925237365640424e-07, + "loss": 0.7333, + "num_input_tokens_seen": 309159310, + "step": 9508 + }, + { + "epoch": 0.8575551246787212, + "flos": 70466920675200.0, + "grad_norm": 0.6383063293908788, + "learning_rate": 2.0899229211515211e-07, + "loss": 0.575, + "num_input_tokens_seen": 309252775, + "step": 9509 + }, + { + "epoch": 0.8576453082021914, + "flos": 22708459740480.0, + "grad_norm": 4.642484247512674, + "learning_rate": 2.0873236339255306e-07, + "loss": 0.6679, + "num_input_tokens_seen": 309278095, + "step": 9510 + }, + { + "epoch": 0.8577354917256618, + "flos": 19252205291520.0, + "grad_norm": 2.0941342812505925, + "learning_rate": 2.0847258751078644e-07, + "loss": 0.6901, + "num_input_tokens_seen": 309306915, + "step": 9511 + }, + { + "epoch": 0.857825675249132, + "flos": 19327225334400.0, + "grad_norm": 2.555570863825333, + "learning_rate": 2.082129644920163e-07, + "loss": 0.7186, + "num_input_tokens_seen": 309334955, + "step": 9512 + }, + { + "epoch": 0.8579158587726022, + "flos": 24281581849440.0, + "grad_norm": 1.6149290965669287, + "learning_rate": 2.0795349435839605e-07, + "loss": 0.7518, + "num_input_tokens_seen": 309365380, + "step": 9513 + }, + { + "epoch": 0.8580060422960725, + "flos": 11821135140960.0, + "grad_norm": 2.0021662900994137, + "learning_rate": 2.0769417713206484e-07, + "loss": 0.714, + "num_input_tokens_seen": 309390020, + "step": 9514 + }, + { + "epoch": 0.8580962258195428, + "flos": 24025516407360.0, + "grad_norm": 1.4251022603335652, + "learning_rate": 2.074350128351492e-07, + "loss": 0.8043, + "num_input_tokens_seen": 309421015, + "step": 9515 + }, + { + "epoch": 0.858186409343013, + "flos": 30074843041440.0, + "grad_norm": 1.6351108710585, + "learning_rate": 2.0717600148976256e-07, + "loss": 0.6609, + "num_input_tokens_seen": 309448820, + "step": 9516 + }, + { + "epoch": 0.8582765928664833, + "flos": 24244666544160.0, + "grad_norm": 2.362721879823456, + "learning_rate": 2.0691714311800436e-07, + "loss": 0.6723, + "num_input_tokens_seen": 309476985, + "step": 9517 + }, + { + "epoch": 0.8583667763899535, + "flos": 20091890533440.0, + "grad_norm": 1.9763761892452067, + "learning_rate": 2.066584377419631e-07, + "loss": 0.8048, + "num_input_tokens_seen": 309504930, + "step": 9518 + }, + { + "epoch": 0.8584569599134239, + "flos": 21002692993440.0, + "grad_norm": 1.6825712068178789, + "learning_rate": 2.0639988538371167e-07, + "loss": 0.7643, + "num_input_tokens_seen": 309531875, + "step": 9519 + }, + { + "epoch": 0.8585471434368941, + "flos": 32224225337760.0, + "grad_norm": 1.9930644180981008, + "learning_rate": 2.0614148606531258e-07, + "loss": 0.707, + "num_input_tokens_seen": 309563080, + "step": 9520 + }, + { + "epoch": 0.8586373269603643, + "flos": 32733940187520.0, + "grad_norm": 2.475114368917362, + "learning_rate": 2.0588323980881285e-07, + "loss": 0.7721, + "num_input_tokens_seen": 309590640, + "step": 9521 + }, + { + "epoch": 0.8587275104838346, + "flos": 20817447411360.0, + "grad_norm": 1.8407407869924166, + "learning_rate": 2.0562514663624752e-07, + "loss": 0.7301, + "num_input_tokens_seen": 309618210, + "step": 9522 + }, + { + "epoch": 0.8588176940073049, + "flos": 25994410850880.0, + "grad_norm": 1.81285157575679, + "learning_rate": 2.0536720656963902e-07, + "loss": 0.8071, + "num_input_tokens_seen": 309649675, + "step": 9523 + }, + { + "epoch": 0.8589078775307751, + "flos": 34659042920160.0, + "grad_norm": 2.7010032838408113, + "learning_rate": 2.051094196309957e-07, + "loss": 0.7673, + "num_input_tokens_seen": 309675950, + "step": 9524 + }, + { + "epoch": 0.8589980610542454, + "flos": 22496855064960.0, + "grad_norm": 2.8481518558778824, + "learning_rate": 2.0485178584231378e-07, + "loss": 0.7018, + "num_input_tokens_seen": 309706280, + "step": 9525 + }, + { + "epoch": 0.8590882445777156, + "flos": 63474373816800.0, + "grad_norm": 0.6734855112543644, + "learning_rate": 2.0459430522557587e-07, + "loss": 0.5768, + "num_input_tokens_seen": 309789660, + "step": 9526 + }, + { + "epoch": 0.8591784281011859, + "flos": 24318571494240.0, + "grad_norm": 2.857822684743846, + "learning_rate": 2.0433697780275195e-07, + "loss": 0.7427, + "num_input_tokens_seen": 309818715, + "step": 9527 + }, + { + "epoch": 0.8592686116246562, + "flos": 25487112035520.0, + "grad_norm": 1.8088030766079233, + "learning_rate": 2.040798035957978e-07, + "loss": 0.7579, + "num_input_tokens_seen": 309850115, + "step": 9528 + }, + { + "epoch": 0.8593587951481264, + "flos": 70776367626240.0, + "grad_norm": 0.7240020180747848, + "learning_rate": 2.038227826266574e-07, + "loss": 0.5926, + "num_input_tokens_seen": 309947390, + "step": 9529 + }, + { + "epoch": 0.8594489786715968, + "flos": 22278225304800.0, + "grad_norm": 2.014304623133406, + "learning_rate": 2.0356591491726126e-07, + "loss": 0.7506, + "num_input_tokens_seen": 309974585, + "step": 9530 + }, + { + "epoch": 0.859539162195067, + "flos": 27779137635360.0, + "grad_norm": 1.8029757882564426, + "learning_rate": 2.033092004895267e-07, + "loss": 0.7831, + "num_input_tokens_seen": 310006380, + "step": 9531 + }, + { + "epoch": 0.8596293457185372, + "flos": 20201595696000.0, + "grad_norm": 1.8618368952534423, + "learning_rate": 2.03052639365358e-07, + "loss": 0.7569, + "num_input_tokens_seen": 310034415, + "step": 9532 + }, + { + "epoch": 0.8597195292420075, + "flos": 26869004231040.0, + "grad_norm": 1.9612994220990891, + "learning_rate": 2.0279623156664694e-07, + "loss": 0.7614, + "num_input_tokens_seen": 310064760, + "step": 9533 + }, + { + "epoch": 0.8598097127654778, + "flos": 23042964843360.0, + "grad_norm": 2.214803920703936, + "learning_rate": 2.0253997711527005e-07, + "loss": 0.8004, + "num_input_tokens_seen": 310094200, + "step": 9534 + }, + { + "epoch": 0.859899896288948, + "flos": 18670890016800.0, + "grad_norm": 1.7936189424174676, + "learning_rate": 2.0228387603309428e-07, + "loss": 0.7913, + "num_input_tokens_seen": 310122610, + "step": 9535 + }, + { + "epoch": 0.8599900798124183, + "flos": 24536532198720.0, + "grad_norm": 1.793457762408815, + "learning_rate": 2.0202792834197035e-07, + "loss": 0.7629, + "num_input_tokens_seen": 310150915, + "step": 9536 + }, + { + "epoch": 0.8600802633358885, + "flos": 18375641914080.0, + "grad_norm": 2.2177712874274036, + "learning_rate": 2.017721340637375e-07, + "loss": 0.6727, + "num_input_tokens_seen": 310176730, + "step": 9537 + }, + { + "epoch": 0.8601704468593588, + "flos": 21476347442400.0, + "grad_norm": 1.7538554772587374, + "learning_rate": 2.0151649322022134e-07, + "loss": 0.7265, + "num_input_tokens_seen": 310204750, + "step": 9538 + }, + { + "epoch": 0.8602606303828291, + "flos": 27851146927680.0, + "grad_norm": 2.131144000239275, + "learning_rate": 2.012610058332349e-07, + "loss": 0.5506, + "num_input_tokens_seen": 310232500, + "step": 9539 + }, + { + "epoch": 0.8603508139062993, + "flos": 71460437318400.0, + "grad_norm": 0.6848923038831697, + "learning_rate": 2.010056719245774e-07, + "loss": 0.573, + "num_input_tokens_seen": 310320190, + "step": 9540 + }, + { + "epoch": 0.8604409974297695, + "flos": 35718324336000.0, + "grad_norm": 2.0364284529449157, + "learning_rate": 2.0075049151603563e-07, + "loss": 0.7664, + "num_input_tokens_seen": 310350355, + "step": 9541 + }, + { + "epoch": 0.8605311809532399, + "flos": 17031148230720.0, + "grad_norm": 1.9098830076407183, + "learning_rate": 2.0049546462938326e-07, + "loss": 0.748, + "num_input_tokens_seen": 310376575, + "step": 9542 + }, + { + "epoch": 0.8606213644767101, + "flos": 30547085039520.0, + "grad_norm": 1.732762317257595, + "learning_rate": 2.0024059128637961e-07, + "loss": 0.7642, + "num_input_tokens_seen": 310407600, + "step": 9543 + }, + { + "epoch": 0.8607115480001803, + "flos": 29745355856160.0, + "grad_norm": 1.6087235548903236, + "learning_rate": 1.99985871508773e-07, + "loss": 0.6076, + "num_input_tokens_seen": 310439075, + "step": 9544 + }, + { + "epoch": 0.8608017315236506, + "flos": 27160275169440.0, + "grad_norm": 2.320727070465116, + "learning_rate": 1.9973130531829674e-07, + "loss": 0.7686, + "num_input_tokens_seen": 310471205, + "step": 9545 + }, + { + "epoch": 0.8608919150471209, + "flos": 19618793630880.0, + "grad_norm": 2.1629872870069446, + "learning_rate": 1.994768927366721e-07, + "loss": 0.6913, + "num_input_tokens_seen": 310499575, + "step": 9546 + }, + { + "epoch": 0.8609820985705912, + "flos": 15536019745440.0, + "grad_norm": 2.7312054383856736, + "learning_rate": 1.992226337856069e-07, + "loss": 0.7116, + "num_input_tokens_seen": 310526315, + "step": 9547 + }, + { + "epoch": 0.8610722820940614, + "flos": 23625766908480.0, + "grad_norm": 1.7542974974470409, + "learning_rate": 1.9896852848679592e-07, + "loss": 0.698, + "num_input_tokens_seen": 310555370, + "step": 9548 + }, + { + "epoch": 0.8611624656175316, + "flos": 23735026033920.0, + "grad_norm": 1.8008985501892467, + "learning_rate": 1.9871457686192094e-07, + "loss": 0.751, + "num_input_tokens_seen": 310585330, + "step": 9549 + }, + { + "epoch": 0.861252649141002, + "flos": 20303978415840.0, + "grad_norm": 2.8702288304460213, + "learning_rate": 1.984607789326509e-07, + "loss": 0.735, + "num_input_tokens_seen": 310612540, + "step": 9550 + }, + { + "epoch": 0.8613428326644722, + "flos": 25010223817440.0, + "grad_norm": 1.640177275101431, + "learning_rate": 1.982071347206402e-07, + "loss": 0.6894, + "num_input_tokens_seen": 310641920, + "step": 9551 + }, + { + "epoch": 0.8614330161879424, + "flos": 43878333812640.0, + "grad_norm": 1.7425217927858865, + "learning_rate": 1.9795364424753202e-07, + "loss": 0.7208, + "num_input_tokens_seen": 310674445, + "step": 9552 + }, + { + "epoch": 0.8615231997114128, + "flos": 20889419533920.0, + "grad_norm": 10.047327959875739, + "learning_rate": 1.9770030753495505e-07, + "loss": 0.7117, + "num_input_tokens_seen": 310699740, + "step": 9553 + }, + { + "epoch": 0.861613383234883, + "flos": 20745698307360.0, + "grad_norm": 1.833264356818348, + "learning_rate": 1.9744712460452573e-07, + "loss": 0.7022, + "num_input_tokens_seen": 310727765, + "step": 9554 + }, + { + "epoch": 0.8617035667583532, + "flos": 24573744862080.0, + "grad_norm": 2.0901546068820283, + "learning_rate": 1.9719409547784703e-07, + "loss": 0.723, + "num_input_tokens_seen": 310756550, + "step": 9555 + }, + { + "epoch": 0.8617937502818235, + "flos": 23074862231040.0, + "grad_norm": 2.754088916537168, + "learning_rate": 1.9694122017650837e-07, + "loss": 0.7652, + "num_input_tokens_seen": 310783510, + "step": 9556 + }, + { + "epoch": 0.8618839338052938, + "flos": 30217486344960.0, + "grad_norm": 1.7926312962200626, + "learning_rate": 1.9668849872208738e-07, + "loss": 0.7522, + "num_input_tokens_seen": 310812705, + "step": 9557 + }, + { + "epoch": 0.8619741173287641, + "flos": 19907537025600.0, + "grad_norm": 1.8128970796436887, + "learning_rate": 1.9643593113614632e-07, + "loss": 0.5897, + "num_input_tokens_seen": 310840905, + "step": 9558 + }, + { + "epoch": 0.8620643008522343, + "flos": 25082864995680.0, + "grad_norm": 1.5962701208030945, + "learning_rate": 1.961835174402371e-07, + "loss": 0.7838, + "num_input_tokens_seen": 310870250, + "step": 9559 + }, + { + "epoch": 0.8621544843757045, + "flos": 23626027096800.0, + "grad_norm": 3.5989880411087705, + "learning_rate": 1.9593125765589535e-07, + "loss": 0.7072, + "num_input_tokens_seen": 310897715, + "step": 9560 + }, + { + "epoch": 0.8622446678991749, + "flos": 23115643191360.0, + "grad_norm": 1.8323260560169066, + "learning_rate": 1.9567915180464721e-07, + "loss": 0.6867, + "num_input_tokens_seen": 310926105, + "step": 9561 + }, + { + "epoch": 0.8623348514226451, + "flos": 26140027735200.0, + "grad_norm": 2.3826775184295994, + "learning_rate": 1.9542719990800217e-07, + "loss": 0.7259, + "num_input_tokens_seen": 310954555, + "step": 9562 + }, + { + "epoch": 0.8624250349461153, + "flos": 17936152208160.0, + "grad_norm": 1.9060919023835212, + "learning_rate": 1.9517540198745896e-07, + "loss": 0.8109, + "num_input_tokens_seen": 310981460, + "step": 9563 + }, + { + "epoch": 0.8625152184695856, + "flos": 21767506871520.0, + "grad_norm": 2.570940213529472, + "learning_rate": 1.94923758064502e-07, + "loss": 0.7389, + "num_input_tokens_seen": 311008710, + "step": 9564 + }, + { + "epoch": 0.8626054019930559, + "flos": 20416545649920.0, + "grad_norm": 1.7643054994143366, + "learning_rate": 1.9467226816060322e-07, + "loss": 0.708, + "num_input_tokens_seen": 311036950, + "step": 9565 + }, + { + "epoch": 0.8626955855165261, + "flos": 21622373194080.0, + "grad_norm": 1.8600766436875245, + "learning_rate": 1.9442093229722122e-07, + "loss": 0.6838, + "num_input_tokens_seen": 311065675, + "step": 9566 + }, + { + "epoch": 0.8627857690399964, + "flos": 14845333836000.0, + "grad_norm": 2.820326532978852, + "learning_rate": 1.9416975049580085e-07, + "loss": 0.8151, + "num_input_tokens_seen": 311089250, + "step": 9567 + }, + { + "epoch": 0.8628759525634666, + "flos": 28325247413760.0, + "grad_norm": 2.088786832985785, + "learning_rate": 1.9391872277777456e-07, + "loss": 0.7242, + "num_input_tokens_seen": 311116430, + "step": 9568 + }, + { + "epoch": 0.862966136086937, + "flos": 22459419383040.0, + "grad_norm": 2.5371599317157765, + "learning_rate": 1.9366784916456158e-07, + "loss": 0.7302, + "num_input_tokens_seen": 311141785, + "step": 9569 + }, + { + "epoch": 0.8630563196104072, + "flos": 24536829556800.0, + "grad_norm": 2.027029494550423, + "learning_rate": 1.9341712967756774e-07, + "loss": 0.6942, + "num_input_tokens_seen": 311172240, + "step": 9570 + }, + { + "epoch": 0.8631465031338774, + "flos": 23115977719200.0, + "grad_norm": 3.4248840479470726, + "learning_rate": 1.9316656433818566e-07, + "loss": 0.6508, + "num_input_tokens_seen": 311202570, + "step": 9571 + }, + { + "epoch": 0.8632366866573477, + "flos": 29673866940480.0, + "grad_norm": 2.787694686454468, + "learning_rate": 1.929161531677954e-07, + "loss": 0.6414, + "num_input_tokens_seen": 311231375, + "step": 9572 + }, + { + "epoch": 0.863326870180818, + "flos": 23547067059360.0, + "grad_norm": 1.7961657243787343, + "learning_rate": 1.9266589618776251e-07, + "loss": 0.6972, + "num_input_tokens_seen": 311259010, + "step": 9573 + }, + { + "epoch": 0.8634170537042882, + "flos": 48723468372000.0, + "grad_norm": 1.9058056923291165, + "learning_rate": 1.924157934194417e-07, + "loss": 0.792, + "num_input_tokens_seen": 311290865, + "step": 9574 + }, + { + "epoch": 0.8635072372277585, + "flos": 26358062779200.0, + "grad_norm": 1.9494245951790026, + "learning_rate": 1.9216584488417142e-07, + "loss": 0.6324, + "num_input_tokens_seen": 311321090, + "step": 9575 + }, + { + "epoch": 0.8635974207512288, + "flos": 14553393841920.0, + "grad_norm": 2.275658849534106, + "learning_rate": 1.919160506032802e-07, + "loss": 0.7976, + "num_input_tokens_seen": 311347185, + "step": 9576 + }, + { + "epoch": 0.863687604274699, + "flos": 20128434141120.0, + "grad_norm": 2.615422989166755, + "learning_rate": 1.916664105980812e-07, + "loss": 0.8022, + "num_input_tokens_seen": 311373060, + "step": 9577 + }, + { + "epoch": 0.8637777877981693, + "flos": 35211322878720.0, + "grad_norm": 4.939901805936523, + "learning_rate": 1.914169248898747e-07, + "loss": 0.7742, + "num_input_tokens_seen": 311402050, + "step": 9578 + }, + { + "epoch": 0.8638679713216395, + "flos": 21767246683200.0, + "grad_norm": 1.6561856527244412, + "learning_rate": 1.9116759349994882e-07, + "loss": 0.6551, + "num_input_tokens_seen": 311431480, + "step": 9579 + }, + { + "epoch": 0.8639581548451098, + "flos": 20197990229280.0, + "grad_norm": 1.8976642933561592, + "learning_rate": 1.9091841644957763e-07, + "loss": 0.7746, + "num_input_tokens_seen": 311459695, + "step": 9580 + }, + { + "epoch": 0.8640483383685801, + "flos": 23336354458080.0, + "grad_norm": 1.4436147254884084, + "learning_rate": 1.9066939376002278e-07, + "loss": 0.8248, + "num_input_tokens_seen": 311491845, + "step": 9581 + }, + { + "epoch": 0.8641385218920503, + "flos": 19873186433760.0, + "grad_norm": 2.4992416185225323, + "learning_rate": 1.9042052545253085e-07, + "loss": 0.6561, + "num_input_tokens_seen": 311516985, + "step": 9582 + }, + { + "epoch": 0.8642287054155205, + "flos": 21548691262560.0, + "grad_norm": 2.560739051712766, + "learning_rate": 1.901718115483384e-07, + "loss": 0.7179, + "num_input_tokens_seen": 311544585, + "step": 9583 + }, + { + "epoch": 0.8643188889389909, + "flos": 22235585856480.0, + "grad_norm": 2.0709478590962194, + "learning_rate": 1.8992325206866598e-07, + "loss": 0.7259, + "num_input_tokens_seen": 311573170, + "step": 9584 + }, + { + "epoch": 0.8644090724624611, + "flos": 27632851695360.0, + "grad_norm": 1.4803750202230812, + "learning_rate": 1.8967484703472225e-07, + "loss": 0.714, + "num_input_tokens_seen": 311606095, + "step": 9585 + }, + { + "epoch": 0.8644992559859314, + "flos": 24170464236000.0, + "grad_norm": 1.5152149235577583, + "learning_rate": 1.8942659646770288e-07, + "loss": 0.7877, + "num_input_tokens_seen": 311636380, + "step": 9586 + }, + { + "epoch": 0.8645894395094016, + "flos": 24791928585120.0, + "grad_norm": 2.0759137083726227, + "learning_rate": 1.8917850038878936e-07, + "loss": 0.8058, + "num_input_tokens_seen": 311663360, + "step": 9587 + }, + { + "epoch": 0.8646796230328719, + "flos": 29637509181600.0, + "grad_norm": 1.8377920320924934, + "learning_rate": 1.8893055881915121e-07, + "loss": 0.6525, + "num_input_tokens_seen": 311693685, + "step": 9588 + }, + { + "epoch": 0.8647698065563422, + "flos": 19290347198880.0, + "grad_norm": 2.5768208297302113, + "learning_rate": 1.886827717799442e-07, + "loss": 0.6874, + "num_input_tokens_seen": 311719545, + "step": 9589 + }, + { + "epoch": 0.8648599900798124, + "flos": 17759827368480.0, + "grad_norm": 2.1353688552590286, + "learning_rate": 1.884351392923096e-07, + "loss": 0.7557, + "num_input_tokens_seen": 311747000, + "step": 9590 + }, + { + "epoch": 0.8649501736032826, + "flos": 22234507933440.0, + "grad_norm": 2.50953249312405, + "learning_rate": 1.8818766137737896e-07, + "loss": 0.7812, + "num_input_tokens_seen": 311771720, + "step": 9591 + }, + { + "epoch": 0.865040357126753, + "flos": 18298466025120.0, + "grad_norm": 2.2620465662014064, + "learning_rate": 1.8794033805626653e-07, + "loss": 0.7156, + "num_input_tokens_seen": 311795215, + "step": 9592 + }, + { + "epoch": 0.8651305406502232, + "flos": 38342178816000.0, + "grad_norm": 2.3430328289146582, + "learning_rate": 1.876931693500763e-07, + "loss": 0.735, + "num_input_tokens_seen": 311829100, + "step": 9593 + }, + { + "epoch": 0.8652207241736934, + "flos": 36516485222400.0, + "grad_norm": 2.3688123307293325, + "learning_rate": 1.8744615527989783e-07, + "loss": 0.588, + "num_input_tokens_seen": 311859925, + "step": 9594 + }, + { + "epoch": 0.8653109076971637, + "flos": 21585532228320.0, + "grad_norm": 2.333556630031237, + "learning_rate": 1.871992958668076e-07, + "loss": 0.8277, + "num_input_tokens_seen": 311888075, + "step": 9595 + }, + { + "epoch": 0.865401091220634, + "flos": 18306680542080.0, + "grad_norm": 2.3328931484091875, + "learning_rate": 1.8695259113186944e-07, + "loss": 0.7828, + "num_input_tokens_seen": 311912215, + "step": 9596 + }, + { + "epoch": 0.8654912747441043, + "flos": 18999039090720.0, + "grad_norm": 1.6332627547189547, + "learning_rate": 1.8670604109613252e-07, + "loss": 0.6997, + "num_input_tokens_seen": 311940110, + "step": 9597 + }, + { + "epoch": 0.8655814582675745, + "flos": 26210959104480.0, + "grad_norm": 2.020378127068463, + "learning_rate": 1.8645964578063533e-07, + "loss": 0.7841, + "num_input_tokens_seen": 311970825, + "step": 9598 + }, + { + "epoch": 0.8656716417910447, + "flos": 24646237361280.0, + "grad_norm": 1.5502912770660062, + "learning_rate": 1.862134052064006e-07, + "loss": 0.7987, + "num_input_tokens_seen": 312000040, + "step": 9599 + }, + { + "epoch": 0.8657618253145151, + "flos": 18707507964000.0, + "grad_norm": 7.693639769024009, + "learning_rate": 1.8596731939443932e-07, + "loss": 0.784, + "num_input_tokens_seen": 312025020, + "step": 9600 + }, + { + "epoch": 0.8658520088379853, + "flos": 16483923359520.0, + "grad_norm": 1.7031840042952442, + "learning_rate": 1.857213883657487e-07, + "loss": 0.7694, + "num_input_tokens_seen": 312052025, + "step": 9601 + }, + { + "epoch": 0.8659421923614555, + "flos": 56332716116640.0, + "grad_norm": 0.6366782146258713, + "learning_rate": 1.8547561214131303e-07, + "loss": 0.5497, + "num_input_tokens_seen": 312140245, + "step": 9602 + }, + { + "epoch": 0.8660323758849259, + "flos": 23407546015680.0, + "grad_norm": 2.0797948796661645, + "learning_rate": 1.8522999074210355e-07, + "loss": 0.8178, + "num_input_tokens_seen": 312167235, + "step": 9603 + }, + { + "epoch": 0.8661225594083961, + "flos": 19546078113120.0, + "grad_norm": 2.1903495756917226, + "learning_rate": 1.849845241890775e-07, + "loss": 0.7815, + "num_input_tokens_seen": 312194230, + "step": 9604 + }, + { + "epoch": 0.8662127429318663, + "flos": 25192941855840.0, + "grad_norm": 2.3717580050235543, + "learning_rate": 1.8473921250317992e-07, + "loss": 0.7456, + "num_input_tokens_seen": 312224620, + "step": 9605 + }, + { + "epoch": 0.8663029264553366, + "flos": 31783880727360.0, + "grad_norm": 1.7938226399350916, + "learning_rate": 1.8449405570534225e-07, + "loss": 0.6708, + "num_input_tokens_seen": 312251220, + "step": 9606 + }, + { + "epoch": 0.8663931099788069, + "flos": 22314174196320.0, + "grad_norm": 1.8378160321688686, + "learning_rate": 1.8424905381648204e-07, + "loss": 0.7865, + "num_input_tokens_seen": 312280305, + "step": 9607 + }, + { + "epoch": 0.8664832935022772, + "flos": 25920617410080.0, + "grad_norm": 1.9396883388767607, + "learning_rate": 1.8400420685750452e-07, + "loss": 0.696, + "num_input_tokens_seen": 312313220, + "step": 9608 + }, + { + "epoch": 0.8665734770257474, + "flos": 23807778721440.0, + "grad_norm": 1.9209625435135775, + "learning_rate": 1.8375951484930142e-07, + "loss": 0.6758, + "num_input_tokens_seen": 312341665, + "step": 9609 + }, + { + "epoch": 0.8666636605492176, + "flos": 22202536206240.0, + "grad_norm": 1.8098038318404188, + "learning_rate": 1.8351497781275094e-07, + "loss": 0.7448, + "num_input_tokens_seen": 312370455, + "step": 9610 + }, + { + "epoch": 0.866753844072688, + "flos": 19399940852160.0, + "grad_norm": 1.9079151072859892, + "learning_rate": 1.8327059576871907e-07, + "loss": 0.7282, + "num_input_tokens_seen": 312397165, + "step": 9611 + }, + { + "epoch": 0.8668440275961582, + "flos": 22895489471040.0, + "grad_norm": 1.5845428498271141, + "learning_rate": 1.8302636873805665e-07, + "loss": 0.7906, + "num_input_tokens_seen": 312427530, + "step": 9612 + }, + { + "epoch": 0.8669342111196284, + "flos": 25338261382080.0, + "grad_norm": 2.039379109908372, + "learning_rate": 1.8278229674160373e-07, + "loss": 0.8174, + "num_input_tokens_seen": 312453770, + "step": 9613 + }, + { + "epoch": 0.8670243946430987, + "flos": 43115229743520.0, + "grad_norm": 1.8756277305128084, + "learning_rate": 1.825383798001845e-07, + "loss": 0.6736, + "num_input_tokens_seen": 312486905, + "step": 9614 + }, + { + "epoch": 0.867114578166569, + "flos": 22386555186240.0, + "grad_norm": 1.7024046048005288, + "learning_rate": 1.8229461793461297e-07, + "loss": 0.684, + "num_input_tokens_seen": 312517965, + "step": 9615 + }, + { + "epoch": 0.8672047616900392, + "flos": 46287163999200.0, + "grad_norm": 1.8712932822726662, + "learning_rate": 1.8205101116568698e-07, + "loss": 0.7535, + "num_input_tokens_seen": 312548555, + "step": 9616 + }, + { + "epoch": 0.8672949452135095, + "flos": 18889556946720.0, + "grad_norm": 2.2004237287114647, + "learning_rate": 1.818075595141928e-07, + "loss": 0.7914, + "num_input_tokens_seen": 312575310, + "step": 9617 + }, + { + "epoch": 0.8673851287369797, + "flos": 22640799310080.0, + "grad_norm": 3.0569221927522956, + "learning_rate": 1.8156426300090288e-07, + "loss": 0.7382, + "num_input_tokens_seen": 312603405, + "step": 9618 + }, + { + "epoch": 0.86747531226045, + "flos": 21731706659040.0, + "grad_norm": 1.7870077802619995, + "learning_rate": 1.8132112164657686e-07, + "loss": 0.732, + "num_input_tokens_seen": 312633110, + "step": 9619 + }, + { + "epoch": 0.8675654957839203, + "flos": 28289112673440.0, + "grad_norm": 2.839993185653033, + "learning_rate": 1.8107813547196106e-07, + "loss": 0.6338, + "num_input_tokens_seen": 312661925, + "step": 9620 + }, + { + "epoch": 0.8676556793073905, + "flos": 16922520991200.0, + "grad_norm": 1.9365645580764035, + "learning_rate": 1.8083530449778817e-07, + "loss": 0.7788, + "num_input_tokens_seen": 312689310, + "step": 9621 + }, + { + "epoch": 0.8677458628308607, + "flos": 36923408484960.0, + "grad_norm": 1.7426315220764885, + "learning_rate": 1.8059262874477787e-07, + "loss": 0.6616, + "num_input_tokens_seen": 312718495, + "step": 9622 + }, + { + "epoch": 0.8678360463543311, + "flos": 26759484917280.0, + "grad_norm": 1.753013468575841, + "learning_rate": 1.8035010823363627e-07, + "loss": 0.7248, + "num_input_tokens_seen": 312751030, + "step": 9623 + }, + { + "epoch": 0.8679262298778013, + "flos": 25732918623840.0, + "grad_norm": 1.9959287929867597, + "learning_rate": 1.8010774298505705e-07, + "loss": 0.7787, + "num_input_tokens_seen": 312781065, + "step": 9624 + }, + { + "epoch": 0.8680164134012716, + "flos": 16630915524960.0, + "grad_norm": 1.9711758034433267, + "learning_rate": 1.7986553301972007e-07, + "loss": 0.8323, + "num_input_tokens_seen": 312802770, + "step": 9625 + }, + { + "epoch": 0.8681065969247419, + "flos": 20269293296160.0, + "grad_norm": 1.7703258086307796, + "learning_rate": 1.7962347835829171e-07, + "loss": 0.7305, + "num_input_tokens_seen": 312830415, + "step": 9626 + }, + { + "epoch": 0.8681967804482121, + "flos": 22164394298880.0, + "grad_norm": 2.1343581638817124, + "learning_rate": 1.793815790214257e-07, + "loss": 0.7462, + "num_input_tokens_seen": 312857765, + "step": 9627 + }, + { + "epoch": 0.8682869639716824, + "flos": 16011904380000.0, + "grad_norm": 2.3603563099546276, + "learning_rate": 1.791398350297626e-07, + "loss": 0.7018, + "num_input_tokens_seen": 312883385, + "step": 9628 + }, + { + "epoch": 0.8683771474951526, + "flos": 25950990837600.0, + "grad_norm": 1.569715775560369, + "learning_rate": 1.7889824640392813e-07, + "loss": 0.7465, + "num_input_tokens_seen": 312912795, + "step": 9629 + }, + { + "epoch": 0.868467331018623, + "flos": 19654519503840.0, + "grad_norm": 1.8963209370550853, + "learning_rate": 1.7865681316453741e-07, + "loss": 0.7216, + "num_input_tokens_seen": 312939145, + "step": 9630 + }, + { + "epoch": 0.8685575145420932, + "flos": 26285607449760.0, + "grad_norm": 1.9547094403323138, + "learning_rate": 1.7841553533218968e-07, + "loss": 0.738, + "num_input_tokens_seen": 312970260, + "step": 9631 + }, + { + "epoch": 0.8686476980655634, + "flos": 18306420353760.0, + "grad_norm": 2.183861021769092, + "learning_rate": 1.7817441292747292e-07, + "loss": 0.843, + "num_input_tokens_seen": 312996980, + "step": 9632 + }, + { + "epoch": 0.8687378815890336, + "flos": 24714789865920.0, + "grad_norm": 2.321047990643844, + "learning_rate": 1.779334459709607e-07, + "loss": 0.7377, + "num_input_tokens_seen": 313025905, + "step": 9633 + }, + { + "epoch": 0.868828065112504, + "flos": 22421760682560.0, + "grad_norm": 3.342919077084071, + "learning_rate": 1.7769263448321347e-07, + "loss": 0.7561, + "num_input_tokens_seen": 313050795, + "step": 9634 + }, + { + "epoch": 0.8689182486359742, + "flos": 21876580148160.0, + "grad_norm": 2.1205349180750677, + "learning_rate": 1.7745197848477879e-07, + "loss": 0.7323, + "num_input_tokens_seen": 313079350, + "step": 9635 + }, + { + "epoch": 0.8690084321594445, + "flos": 20492086069440.0, + "grad_norm": 2.3033414368917047, + "learning_rate": 1.7721147799619063e-07, + "loss": 0.842, + "num_input_tokens_seen": 313103775, + "step": 9636 + }, + { + "epoch": 0.8690986156829147, + "flos": 70013077708320.0, + "grad_norm": 0.592849600862142, + "learning_rate": 1.769711330379704e-07, + "loss": 0.5469, + "num_input_tokens_seen": 313208110, + "step": 9637 + }, + { + "epoch": 0.869188799206385, + "flos": 24317270552640.0, + "grad_norm": 1.992548950915034, + "learning_rate": 1.767309436306248e-07, + "loss": 0.7926, + "num_input_tokens_seen": 313234500, + "step": 9638 + }, + { + "epoch": 0.8692789827298553, + "flos": 23371708633440.0, + "grad_norm": 2.0594003678700035, + "learning_rate": 1.764909097946483e-07, + "loss": 0.8046, + "num_input_tokens_seen": 313262915, + "step": 9639 + }, + { + "epoch": 0.8693691662533255, + "flos": 21616351692960.0, + "grad_norm": 2.023518800022725, + "learning_rate": 1.7625103155052236e-07, + "loss": 0.8072, + "num_input_tokens_seen": 313289695, + "step": 9640 + }, + { + "epoch": 0.8694593497767957, + "flos": 20121929433120.0, + "grad_norm": 2.6760144208682024, + "learning_rate": 1.760113089187143e-07, + "loss": 0.7534, + "num_input_tokens_seen": 313316280, + "step": 9641 + }, + { + "epoch": 0.8695495333002661, + "flos": 27412437786720.0, + "grad_norm": 1.8545108932829308, + "learning_rate": 1.7577174191967868e-07, + "loss": 0.7676, + "num_input_tokens_seen": 313345655, + "step": 9642 + }, + { + "epoch": 0.8696397168237363, + "flos": 23480373042720.0, + "grad_norm": 2.027081246553022, + "learning_rate": 1.755323305738574e-07, + "loss": 0.7515, + "num_input_tokens_seen": 313373460, + "step": 9643 + }, + { + "epoch": 0.8697299003472065, + "flos": 22351052331840.0, + "grad_norm": 1.884224896122403, + "learning_rate": 1.7529307490167677e-07, + "loss": 0.7728, + "num_input_tokens_seen": 313404835, + "step": 9644 + }, + { + "epoch": 0.8698200838706768, + "flos": 25006469671680.0, + "grad_norm": 2.733299959568256, + "learning_rate": 1.7505397492355288e-07, + "loss": 0.7231, + "num_input_tokens_seen": 313432950, + "step": 9645 + }, + { + "epoch": 0.8699102673941471, + "flos": 22059855732960.0, + "grad_norm": 1.9029300107488756, + "learning_rate": 1.7481503065988589e-07, + "loss": 0.7784, + "num_input_tokens_seen": 313461760, + "step": 9646 + }, + { + "epoch": 0.8700004509176174, + "flos": 28072155552480.0, + "grad_norm": 1.5331105198197734, + "learning_rate": 1.7457624213106526e-07, + "loss": 0.7083, + "num_input_tokens_seen": 313495505, + "step": 9647 + }, + { + "epoch": 0.8700906344410876, + "flos": 19363174225920.0, + "grad_norm": 2.971683699009369, + "learning_rate": 1.7433760935746465e-07, + "loss": 0.7406, + "num_input_tokens_seen": 313520625, + "step": 9648 + }, + { + "epoch": 0.8701808179645579, + "flos": 20091927703200.0, + "grad_norm": 1.962779543945706, + "learning_rate": 1.740991323594456e-07, + "loss": 0.6849, + "num_input_tokens_seen": 313550825, + "step": 9649 + }, + { + "epoch": 0.8702710014880282, + "flos": 22386629525760.0, + "grad_norm": 1.9632119897844695, + "learning_rate": 1.7386081115735651e-07, + "loss": 0.7143, + "num_input_tokens_seen": 313578955, + "step": 9650 + }, + { + "epoch": 0.8703611850114984, + "flos": 25411088409120.0, + "grad_norm": 2.950995267661943, + "learning_rate": 1.736226457715324e-07, + "loss": 0.7006, + "num_input_tokens_seen": 313608285, + "step": 9651 + }, + { + "epoch": 0.8704513685349686, + "flos": 24901931105760.0, + "grad_norm": 1.5217910868796423, + "learning_rate": 1.7338463622229505e-07, + "loss": 0.6936, + "num_input_tokens_seen": 313641765, + "step": 9652 + }, + { + "epoch": 0.870541552058439, + "flos": 35205004019520.0, + "grad_norm": 2.118678282477624, + "learning_rate": 1.7314678252995152e-07, + "loss": 0.6923, + "num_input_tokens_seen": 313669900, + "step": 9653 + }, + { + "epoch": 0.8706317355819092, + "flos": 29707362627840.0, + "grad_norm": 1.758547209738594, + "learning_rate": 1.7290908471479847e-07, + "loss": 0.6747, + "num_input_tokens_seen": 313699690, + "step": 9654 + }, + { + "epoch": 0.8707219191053794, + "flos": 25338707419200.0, + "grad_norm": 1.8980992019703895, + "learning_rate": 1.7267154279711637e-07, + "loss": 0.7394, + "num_input_tokens_seen": 313729625, + "step": 9655 + }, + { + "epoch": 0.8708121026288497, + "flos": 22714518411360.0, + "grad_norm": 1.5659708002858874, + "learning_rate": 1.724341567971741e-07, + "loss": 0.7854, + "num_input_tokens_seen": 313759435, + "step": 9656 + }, + { + "epoch": 0.87090228615232, + "flos": 21836282394720.0, + "grad_norm": 1.7467019474795513, + "learning_rate": 1.7219692673522657e-07, + "loss": 0.8401, + "num_input_tokens_seen": 313786380, + "step": 9657 + }, + { + "epoch": 0.8709924696757902, + "flos": 24683450024640.0, + "grad_norm": 1.668693152898847, + "learning_rate": 1.7195985263151558e-07, + "loss": 0.7604, + "num_input_tokens_seen": 313815920, + "step": 9658 + }, + { + "epoch": 0.8710826531992605, + "flos": 27813228038880.0, + "grad_norm": 2.0432974731426703, + "learning_rate": 1.7172293450626985e-07, + "loss": 0.7899, + "num_input_tokens_seen": 313846440, + "step": 9659 + }, + { + "epoch": 0.8711728367227307, + "flos": 24749214797280.0, + "grad_norm": 1.9312976497956815, + "learning_rate": 1.7148617237970475e-07, + "loss": 0.7424, + "num_input_tokens_seen": 313875245, + "step": 9660 + }, + { + "epoch": 0.8712630202462011, + "flos": 27999737392800.0, + "grad_norm": 1.5761970603093562, + "learning_rate": 1.7124956627202102e-07, + "loss": 0.7065, + "num_input_tokens_seen": 313910080, + "step": 9661 + }, + { + "epoch": 0.8713532037696713, + "flos": 22128742765440.0, + "grad_norm": 2.240052486043189, + "learning_rate": 1.7101311620340852e-07, + "loss": 0.755, + "num_input_tokens_seen": 313931100, + "step": 9662 + }, + { + "epoch": 0.8714433872931415, + "flos": 11706003193440.0, + "grad_norm": 2.0352759737907156, + "learning_rate": 1.7077682219404155e-07, + "loss": 0.6833, + "num_input_tokens_seen": 313955405, + "step": 9663 + }, + { + "epoch": 0.8715335708166118, + "flos": 14363241851520.0, + "grad_norm": 2.383428075103311, + "learning_rate": 1.705406842640824e-07, + "loss": 0.8707, + "num_input_tokens_seen": 313973275, + "step": 9664 + }, + { + "epoch": 0.8716237543400821, + "flos": 21841225972800.0, + "grad_norm": 3.41939493125053, + "learning_rate": 1.7030470243367946e-07, + "loss": 0.7317, + "num_input_tokens_seen": 314003435, + "step": 9665 + }, + { + "epoch": 0.8717139378635523, + "flos": 17031780116640.0, + "grad_norm": 1.5601810031201342, + "learning_rate": 1.7006887672296834e-07, + "loss": 0.6677, + "num_input_tokens_seen": 314029420, + "step": 9666 + }, + { + "epoch": 0.8718041213870226, + "flos": 24427050054720.0, + "grad_norm": 2.188102530195444, + "learning_rate": 1.6983320715207094e-07, + "loss": 0.6857, + "num_input_tokens_seen": 314058920, + "step": 9667 + }, + { + "epoch": 0.8718943049104928, + "flos": 27704266271520.0, + "grad_norm": 1.8622299203331694, + "learning_rate": 1.6959769374109523e-07, + "loss": 0.7583, + "num_input_tokens_seen": 314088730, + "step": 9668 + }, + { + "epoch": 0.8719844884339631, + "flos": 22893073436640.0, + "grad_norm": 1.6963796465135954, + "learning_rate": 1.6936233651013754e-07, + "loss": 0.6752, + "num_input_tokens_seen": 314116045, + "step": 9669 + }, + { + "epoch": 0.8720746719574334, + "flos": 24097191171840.0, + "grad_norm": 1.8490918894784172, + "learning_rate": 1.691271354792787e-07, + "loss": 0.7702, + "num_input_tokens_seen": 314144835, + "step": 9670 + }, + { + "epoch": 0.8721648554809036, + "flos": 28361642342400.0, + "grad_norm": 2.0534536624098805, + "learning_rate": 1.6889209066858866e-07, + "loss": 0.6216, + "num_input_tokens_seen": 314176685, + "step": 9671 + }, + { + "epoch": 0.872255039004374, + "flos": 21220988225760.0, + "grad_norm": 2.120162506532852, + "learning_rate": 1.6865720209812185e-07, + "loss": 0.7833, + "num_input_tokens_seen": 314205515, + "step": 9672 + }, + { + "epoch": 0.8723452225278442, + "flos": 22606114190400.0, + "grad_norm": 2.787238627954007, + "learning_rate": 1.684224697879204e-07, + "loss": 0.7063, + "num_input_tokens_seen": 314232390, + "step": 9673 + }, + { + "epoch": 0.8724354060513144, + "flos": 21402405322560.0, + "grad_norm": 2.0109450546869674, + "learning_rate": 1.6818789375801302e-07, + "loss": 0.8157, + "num_input_tokens_seen": 314258620, + "step": 9674 + }, + { + "epoch": 0.8725255895747847, + "flos": 20784286251840.0, + "grad_norm": 1.6066890635980478, + "learning_rate": 1.679534740284152e-07, + "loss": 0.7003, + "num_input_tokens_seen": 314288420, + "step": 9675 + }, + { + "epoch": 0.872615773098255, + "flos": 19692066695040.0, + "grad_norm": 1.9929385097337329, + "learning_rate": 1.6771921061912853e-07, + "loss": 0.6829, + "num_input_tokens_seen": 314315295, + "step": 9676 + }, + { + "epoch": 0.8727059566217252, + "flos": 21038976412800.0, + "grad_norm": 2.0994322236312115, + "learning_rate": 1.6748510355014234e-07, + "loss": 0.6805, + "num_input_tokens_seen": 314342510, + "step": 9677 + }, + { + "epoch": 0.8727961401451955, + "flos": 23805808724160.0, + "grad_norm": 1.6291290789444606, + "learning_rate": 1.6725115284143132e-07, + "loss": 0.7549, + "num_input_tokens_seen": 314372900, + "step": 9678 + }, + { + "epoch": 0.8728863236686657, + "flos": 41434037941440.0, + "grad_norm": 4.131931838471375, + "learning_rate": 1.670173585129575e-07, + "loss": 0.7144, + "num_input_tokens_seen": 314399965, + "step": 9679 + }, + { + "epoch": 0.872976507192136, + "flos": 21730703075520.0, + "grad_norm": 1.956521399931026, + "learning_rate": 1.667837205846696e-07, + "loss": 0.8002, + "num_input_tokens_seen": 314430050, + "step": 9680 + }, + { + "epoch": 0.8730666907156063, + "flos": 20602646136480.0, + "grad_norm": 2.166740880820016, + "learning_rate": 1.6655023907650278e-07, + "loss": 0.7499, + "num_input_tokens_seen": 314458105, + "step": 9681 + }, + { + "epoch": 0.8731568742390765, + "flos": 24754307054400.0, + "grad_norm": 2.150862445400264, + "learning_rate": 1.6631691400837954e-07, + "loss": 0.7818, + "num_input_tokens_seen": 314485805, + "step": 9682 + }, + { + "epoch": 0.8732470577625467, + "flos": 19691174620800.0, + "grad_norm": 1.9404790187606684, + "learning_rate": 1.6608374540020752e-07, + "loss": 0.7883, + "num_input_tokens_seen": 314512390, + "step": 9683 + }, + { + "epoch": 0.8733372412860171, + "flos": 66561357970080.0, + "grad_norm": 0.692920986795941, + "learning_rate": 1.658507332718828e-07, + "loss": 0.5265, + "num_input_tokens_seen": 314602645, + "step": 9684 + }, + { + "epoch": 0.8734274248094873, + "flos": 29124486223200.0, + "grad_norm": 1.7779165832982007, + "learning_rate": 1.656178776432864e-07, + "loss": 0.676, + "num_input_tokens_seen": 314635135, + "step": 9685 + }, + { + "epoch": 0.8735176083329576, + "flos": 19983040275360.0, + "grad_norm": 2.638975361751303, + "learning_rate": 1.6538517853428814e-07, + "loss": 0.7178, + "num_input_tokens_seen": 314661085, + "step": 9686 + }, + { + "epoch": 0.8736077918564278, + "flos": 25196807510880.0, + "grad_norm": 1.526306658793657, + "learning_rate": 1.6515263596474194e-07, + "loss": 0.7233, + "num_input_tokens_seen": 314693515, + "step": 9687 + }, + { + "epoch": 0.8736979753798981, + "flos": 22168743160800.0, + "grad_norm": 1.9655506841684118, + "learning_rate": 1.6492024995449017e-07, + "loss": 0.7227, + "num_input_tokens_seen": 314719585, + "step": 9688 + }, + { + "epoch": 0.8737881589033684, + "flos": 21221136904800.0, + "grad_norm": 1.8963211885088962, + "learning_rate": 1.6468802052336116e-07, + "loss": 0.7081, + "num_input_tokens_seen": 314747855, + "step": 9689 + }, + { + "epoch": 0.8738783424268386, + "flos": 37725360686880.0, + "grad_norm": 2.2257578684207027, + "learning_rate": 1.6445594769116998e-07, + "loss": 0.7616, + "num_input_tokens_seen": 314780075, + "step": 9690 + }, + { + "epoch": 0.8739685259503088, + "flos": 24718952879040.0, + "grad_norm": 1.8541609160373391, + "learning_rate": 1.6422403147771836e-07, + "loss": 0.8089, + "num_input_tokens_seen": 314809765, + "step": 9691 + }, + { + "epoch": 0.8740587094737792, + "flos": 22752214281600.0, + "grad_norm": 2.1333768328563094, + "learning_rate": 1.6399227190279485e-07, + "loss": 0.7036, + "num_input_tokens_seen": 314838750, + "step": 9692 + }, + { + "epoch": 0.8741488929972494, + "flos": 27854046168960.0, + "grad_norm": 1.9508420495521992, + "learning_rate": 1.637606689861748e-07, + "loss": 0.6695, + "num_input_tokens_seen": 314872100, + "step": 9693 + }, + { + "epoch": 0.8742390765207196, + "flos": 21330284520960.0, + "grad_norm": 2.8658440792159214, + "learning_rate": 1.6352922274761883e-07, + "loss": 0.6991, + "num_input_tokens_seen": 314899205, + "step": 9694 + }, + { + "epoch": 0.87432926004419, + "flos": 34446174472800.0, + "grad_norm": 1.6841312975270606, + "learning_rate": 1.6329793320687602e-07, + "loss": 0.6073, + "num_input_tokens_seen": 314933380, + "step": 9695 + }, + { + "epoch": 0.8744194435676602, + "flos": 24496606142880.0, + "grad_norm": 1.8855377355100775, + "learning_rate": 1.630668003836808e-07, + "loss": 0.7758, + "num_input_tokens_seen": 314963765, + "step": 9696 + }, + { + "epoch": 0.8745096270911304, + "flos": 19942854031200.0, + "grad_norm": 1.9307075923938535, + "learning_rate": 1.62835824297755e-07, + "loss": 0.7472, + "num_input_tokens_seen": 314992850, + "step": 9697 + }, + { + "epoch": 0.8745998106146007, + "flos": 22059112337760.0, + "grad_norm": 1.6103138083525366, + "learning_rate": 1.626050049688066e-07, + "loss": 0.7487, + "num_input_tokens_seen": 315020905, + "step": 9698 + }, + { + "epoch": 0.874689994138071, + "flos": 20162821902720.0, + "grad_norm": 1.7502187183485083, + "learning_rate": 1.623743424165309e-07, + "loss": 0.7173, + "num_input_tokens_seen": 315050795, + "step": 9699 + }, + { + "epoch": 0.8747801776615413, + "flos": 16886051723040.0, + "grad_norm": 1.883244888231882, + "learning_rate": 1.6214383666060826e-07, + "loss": 0.688, + "num_input_tokens_seen": 315076095, + "step": 9700 + }, + { + "epoch": 0.8748703611850115, + "flos": 20092336570560.0, + "grad_norm": 1.7227558963526923, + "learning_rate": 1.619134877207078e-07, + "loss": 0.7602, + "num_input_tokens_seen": 315104700, + "step": 9701 + }, + { + "epoch": 0.8749605447084817, + "flos": 23661009574560.0, + "grad_norm": 1.7458925408616792, + "learning_rate": 1.616832956164831e-07, + "loss": 0.7394, + "num_input_tokens_seen": 315133510, + "step": 9702 + }, + { + "epoch": 0.8750507282319521, + "flos": 31494951483840.0, + "grad_norm": 1.7187579761666842, + "learning_rate": 1.6145326036757667e-07, + "loss": 0.7124, + "num_input_tokens_seen": 315166285, + "step": 9703 + }, + { + "epoch": 0.8751409117554223, + "flos": 25374470461920.0, + "grad_norm": 1.8641634278261472, + "learning_rate": 1.612233819936155e-07, + "loss": 0.7066, + "num_input_tokens_seen": 315195015, + "step": 9704 + }, + { + "epoch": 0.8752310952788925, + "flos": 32220025154880.0, + "grad_norm": 1.8530262501026087, + "learning_rate": 1.6099366051421414e-07, + "loss": 0.7332, + "num_input_tokens_seen": 315223405, + "step": 9705 + }, + { + "epoch": 0.8753212788023628, + "flos": 23877000281760.0, + "grad_norm": 1.839111182282327, + "learning_rate": 1.6076409594897378e-07, + "loss": 0.7997, + "num_input_tokens_seen": 315251770, + "step": 9706 + }, + { + "epoch": 0.8754114623258331, + "flos": 69743928510720.0, + "grad_norm": 0.6584454777926344, + "learning_rate": 1.605346883174823e-07, + "loss": 0.5269, + "num_input_tokens_seen": 315341515, + "step": 9707 + }, + { + "epoch": 0.8755016458493033, + "flos": 32005372559040.0, + "grad_norm": 2.0058007756395937, + "learning_rate": 1.6030543763931427e-07, + "loss": 0.7112, + "num_input_tokens_seen": 315370715, + "step": 9708 + }, + { + "epoch": 0.8755918293727736, + "flos": 14881617255360.0, + "grad_norm": 1.95856414273453, + "learning_rate": 1.600763439340298e-07, + "loss": 0.7607, + "num_input_tokens_seen": 315394900, + "step": 9709 + }, + { + "epoch": 0.8756820128962438, + "flos": 27809027856000.0, + "grad_norm": 1.904602902593607, + "learning_rate": 1.5984740722117707e-07, + "loss": 0.7949, + "num_input_tokens_seen": 315424785, + "step": 9710 + }, + { + "epoch": 0.8757721964197142, + "flos": 24608318472480.0, + "grad_norm": 1.4311984889993212, + "learning_rate": 1.5961862752028998e-07, + "loss": 0.7541, + "num_input_tokens_seen": 315454655, + "step": 9711 + }, + { + "epoch": 0.8758623799431844, + "flos": 24499839912000.0, + "grad_norm": 2.426235678550436, + "learning_rate": 1.5939000485088937e-07, + "loss": 0.6482, + "num_input_tokens_seen": 315486400, + "step": 9712 + }, + { + "epoch": 0.8759525634666546, + "flos": 21359468516160.0, + "grad_norm": 1.8497676600368393, + "learning_rate": 1.5916153923248254e-07, + "loss": 0.8343, + "num_input_tokens_seen": 315513530, + "step": 9713 + }, + { + "epoch": 0.8760427469901249, + "flos": 21147975349920.0, + "grad_norm": 1.7059750289967748, + "learning_rate": 1.5893323068456342e-07, + "loss": 0.787, + "num_input_tokens_seen": 315541105, + "step": 9714 + }, + { + "epoch": 0.8761329305135952, + "flos": 25302163811520.0, + "grad_norm": 1.600462223878735, + "learning_rate": 1.5870507922661248e-07, + "loss": 0.7866, + "num_input_tokens_seen": 315569385, + "step": 9715 + }, + { + "epoch": 0.8762231140370654, + "flos": 69382692616800.0, + "grad_norm": 0.6188690119256984, + "learning_rate": 1.5847708487809763e-07, + "loss": 0.5659, + "num_input_tokens_seen": 315667190, + "step": 9716 + }, + { + "epoch": 0.8763132975605357, + "flos": 22459828250400.0, + "grad_norm": 2.230520912535851, + "learning_rate": 1.5824924765847113e-07, + "loss": 0.8163, + "num_input_tokens_seen": 315695850, + "step": 9717 + }, + { + "epoch": 0.8764034810840059, + "flos": 23331150691680.0, + "grad_norm": 1.530463191709215, + "learning_rate": 1.5802156758717478e-07, + "loss": 0.7173, + "num_input_tokens_seen": 315725555, + "step": 9718 + }, + { + "epoch": 0.8764936646074762, + "flos": 23079508451040.0, + "grad_norm": 1.7453889725886764, + "learning_rate": 1.5779404468363433e-07, + "loss": 0.7398, + "num_input_tokens_seen": 315753755, + "step": 9719 + }, + { + "epoch": 0.8765838481309465, + "flos": 23079768639360.0, + "grad_norm": 2.0151748746046922, + "learning_rate": 1.5756667896726405e-07, + "loss": 0.7973, + "num_input_tokens_seen": 315782445, + "step": 9720 + }, + { + "epoch": 0.8766740316544167, + "flos": 28903589107680.0, + "grad_norm": 2.3770074892997313, + "learning_rate": 1.5733947045746377e-07, + "loss": 0.7591, + "num_input_tokens_seen": 315810070, + "step": 9721 + }, + { + "epoch": 0.876764215177887, + "flos": 25702916893920.0, + "grad_norm": 2.0169399023828696, + "learning_rate": 1.5711241917362018e-07, + "loss": 0.8087, + "num_input_tokens_seen": 315840330, + "step": 9722 + }, + { + "epoch": 0.8768543987013573, + "flos": 21768213096960.0, + "grad_norm": 2.3811047046618903, + "learning_rate": 1.5688552513510688e-07, + "loss": 0.6802, + "num_input_tokens_seen": 315867970, + "step": 9723 + }, + { + "epoch": 0.8769445822248275, + "flos": 26502192873120.0, + "grad_norm": 2.4691601183217924, + "learning_rate": 1.5665878836128266e-07, + "loss": 0.7657, + "num_input_tokens_seen": 315896180, + "step": 9724 + }, + { + "epoch": 0.8770347657482978, + "flos": 34117690871040.0, + "grad_norm": 12.045232084389113, + "learning_rate": 1.5643220887149554e-07, + "loss": 0.7898, + "num_input_tokens_seen": 315929955, + "step": 9725 + }, + { + "epoch": 0.8771249492717681, + "flos": 69536747036640.0, + "grad_norm": 0.7003248406156578, + "learning_rate": 1.562057866850772e-07, + "loss": 0.5881, + "num_input_tokens_seen": 316015265, + "step": 9726 + }, + { + "epoch": 0.8772151327952383, + "flos": 38927396915520.0, + "grad_norm": 2.742981450706898, + "learning_rate": 1.5597952182134777e-07, + "loss": 0.6163, + "num_input_tokens_seen": 316044405, + "step": 9727 + }, + { + "epoch": 0.8773053163187086, + "flos": 25007621934240.0, + "grad_norm": 1.6238262632578184, + "learning_rate": 1.557534142996133e-07, + "loss": 0.7955, + "num_input_tokens_seen": 316074455, + "step": 9728 + }, + { + "epoch": 0.8773954998421788, + "flos": 23042741824800.0, + "grad_norm": 2.3502764559521667, + "learning_rate": 1.5552746413916662e-07, + "loss": 0.7743, + "num_input_tokens_seen": 316102825, + "step": 9729 + }, + { + "epoch": 0.8774856833656491, + "flos": 23115791870400.0, + "grad_norm": 1.9319708284462174, + "learning_rate": 1.5530167135928697e-07, + "loss": 0.7336, + "num_input_tokens_seen": 316126930, + "step": 9730 + }, + { + "epoch": 0.8775758668891194, + "flos": 27159829132320.0, + "grad_norm": 2.500164026601477, + "learning_rate": 1.5507603597924068e-07, + "loss": 0.6803, + "num_input_tokens_seen": 316156770, + "step": 9731 + }, + { + "epoch": 0.8776660504125896, + "flos": 25484212794240.0, + "grad_norm": 1.8756277305128084, + "learning_rate": 1.548505580182793e-07, + "loss": 0.6999, + "num_input_tokens_seen": 316185460, + "step": 9732 + }, + { + "epoch": 0.8777562339360598, + "flos": 23262226489440.0, + "grad_norm": 1.7802462510352188, + "learning_rate": 1.5462523749564271e-07, + "loss": 0.7384, + "num_input_tokens_seen": 316214940, + "step": 9733 + }, + { + "epoch": 0.8778464174595302, + "flos": 24714752696160.0, + "grad_norm": 2.300620832825493, + "learning_rate": 1.5440007443055602e-07, + "loss": 0.7389, + "num_input_tokens_seen": 316243395, + "step": 9734 + }, + { + "epoch": 0.8779366009830004, + "flos": 26172408329760.0, + "grad_norm": 2.056407484151729, + "learning_rate": 1.541750688422314e-07, + "loss": 0.6739, + "num_input_tokens_seen": 316272415, + "step": 9735 + }, + { + "epoch": 0.8780267845064706, + "flos": 24534450692160.0, + "grad_norm": 2.308230195570231, + "learning_rate": 1.5395022074986797e-07, + "loss": 0.7412, + "num_input_tokens_seen": 316301155, + "step": 9736 + }, + { + "epoch": 0.8781169680299409, + "flos": 24828174834720.0, + "grad_norm": 1.6654505425390316, + "learning_rate": 1.5372553017265033e-07, + "loss": 0.6845, + "num_input_tokens_seen": 316332475, + "step": 9737 + }, + { + "epoch": 0.8782071515534112, + "flos": 21940932469920.0, + "grad_norm": 3.1310491379446015, + "learning_rate": 1.5350099712975116e-07, + "loss": 0.7973, + "num_input_tokens_seen": 316354610, + "step": 9738 + }, + { + "epoch": 0.8782973350768815, + "flos": 19363248565440.0, + "grad_norm": 1.601597334320106, + "learning_rate": 1.5327662164032785e-07, + "loss": 0.7313, + "num_input_tokens_seen": 316382395, + "step": 9739 + }, + { + "epoch": 0.8783875186003517, + "flos": 27961335297120.0, + "grad_norm": 1.9206660740560253, + "learning_rate": 1.5305240372352656e-07, + "loss": 0.5957, + "num_input_tokens_seen": 316413445, + "step": 9740 + }, + { + "epoch": 0.8784777021238219, + "flos": 23917186525920.0, + "grad_norm": 1.7783577175071488, + "learning_rate": 1.5282834339847738e-07, + "loss": 0.7512, + "num_input_tokens_seen": 316442375, + "step": 9741 + }, + { + "epoch": 0.8785678856472923, + "flos": 31713915771840.0, + "grad_norm": 1.7564988038112999, + "learning_rate": 1.526044406842999e-07, + "loss": 0.7201, + "num_input_tokens_seen": 316474675, + "step": 9742 + }, + { + "epoch": 0.8786580691707625, + "flos": 28727375777280.0, + "grad_norm": 1.9700996979789644, + "learning_rate": 1.523806956000977e-07, + "loss": 0.6683, + "num_input_tokens_seen": 316506585, + "step": 9743 + }, + { + "epoch": 0.8787482526942327, + "flos": 29675651088960.0, + "grad_norm": 1.5327723973245742, + "learning_rate": 1.5215710816496197e-07, + "loss": 0.7918, + "num_input_tokens_seen": 316538140, + "step": 9744 + }, + { + "epoch": 0.8788384362177031, + "flos": 29418842251680.0, + "grad_norm": 1.9784548299172329, + "learning_rate": 1.5193367839797077e-07, + "loss": 0.687, + "num_input_tokens_seen": 316566965, + "step": 9745 + }, + { + "epoch": 0.8789286197411733, + "flos": 20565396303360.0, + "grad_norm": 1.7282490497786809, + "learning_rate": 1.5171040631818842e-07, + "loss": 0.7397, + "num_input_tokens_seen": 316594780, + "step": 9746 + }, + { + "epoch": 0.8790188032646435, + "flos": 36267630713760.0, + "grad_norm": 1.8898979239171987, + "learning_rate": 1.5148729194466547e-07, + "loss": 0.6743, + "num_input_tokens_seen": 316625265, + "step": 9747 + }, + { + "epoch": 0.8791089867881138, + "flos": 19727012003040.0, + "grad_norm": 1.7568430622119637, + "learning_rate": 1.5126433529643956e-07, + "loss": 0.8104, + "num_input_tokens_seen": 316652325, + "step": 9748 + }, + { + "epoch": 0.8791991703115841, + "flos": 25812250358880.0, + "grad_norm": 1.747041244648835, + "learning_rate": 1.5104153639253436e-07, + "loss": 0.7407, + "num_input_tokens_seen": 316681220, + "step": 9749 + }, + { + "epoch": 0.8792893538350544, + "flos": 57799329662400.0, + "grad_norm": 0.6126381727965053, + "learning_rate": 1.5081889525196002e-07, + "loss": 0.6291, + "num_input_tokens_seen": 316772465, + "step": 9750 + }, + { + "epoch": 0.8793795373585246, + "flos": 21912863567520.0, + "grad_norm": 2.1031361187078397, + "learning_rate": 1.5059641189371398e-07, + "loss": 0.799, + "num_input_tokens_seen": 316800460, + "step": 9751 + }, + { + "epoch": 0.8794697208819948, + "flos": 28617075898560.0, + "grad_norm": 1.7431248764634388, + "learning_rate": 1.503740863367795e-07, + "loss": 0.756, + "num_input_tokens_seen": 316830805, + "step": 9752 + }, + { + "epoch": 0.8795599044054652, + "flos": 20455876989600.0, + "grad_norm": 2.2495543780379608, + "learning_rate": 1.50151918600127e-07, + "loss": 0.7692, + "num_input_tokens_seen": 316856765, + "step": 9753 + }, + { + "epoch": 0.8796500879289354, + "flos": 28395621236640.0, + "grad_norm": 3.9326993954034846, + "learning_rate": 1.4992990870271217e-07, + "loss": 0.7967, + "num_input_tokens_seen": 316886240, + "step": 9754 + }, + { + "epoch": 0.8797402714524056, + "flos": 18816098033760.0, + "grad_norm": 2.236960136027686, + "learning_rate": 1.497080566634794e-07, + "loss": 0.7448, + "num_input_tokens_seen": 316911935, + "step": 9755 + }, + { + "epoch": 0.8798304549758759, + "flos": 20420374135200.0, + "grad_norm": 2.063625144393299, + "learning_rate": 1.4948636250135693e-07, + "loss": 0.7502, + "num_input_tokens_seen": 316938410, + "step": 9756 + }, + { + "epoch": 0.8799206384993462, + "flos": 19618124575200.0, + "grad_norm": 2.2368752955404267, + "learning_rate": 1.4926482623526249e-07, + "loss": 0.8023, + "num_input_tokens_seen": 316966190, + "step": 9757 + }, + { + "epoch": 0.8800108220228164, + "flos": 22350420445920.0, + "grad_norm": 2.170185996667529, + "learning_rate": 1.4904344788409694e-07, + "loss": 0.7431, + "num_input_tokens_seen": 316991845, + "step": 9758 + }, + { + "epoch": 0.8801010055462867, + "flos": 20238473831520.0, + "grad_norm": 2.440802562081622, + "learning_rate": 1.4882222746675143e-07, + "loss": 0.7032, + "num_input_tokens_seen": 317019760, + "step": 9759 + }, + { + "epoch": 0.8801911890697569, + "flos": 22678681029120.0, + "grad_norm": 2.4067924185399043, + "learning_rate": 1.4860116500210018e-07, + "loss": 0.7852, + "num_input_tokens_seen": 317046295, + "step": 9760 + }, + { + "epoch": 0.8802813725932273, + "flos": 18926323572960.0, + "grad_norm": 2.4064724311573085, + "learning_rate": 1.4838026050900632e-07, + "loss": 0.8135, + "num_input_tokens_seen": 317072405, + "step": 9761 + }, + { + "epoch": 0.8803715561166975, + "flos": 22533101314560.0, + "grad_norm": 1.61125621288395, + "learning_rate": 1.481595140063181e-07, + "loss": 0.79, + "num_input_tokens_seen": 317101165, + "step": 9762 + }, + { + "epoch": 0.8804617396401677, + "flos": 23953544284800.0, + "grad_norm": 2.2787561240979723, + "learning_rate": 1.4793892551287136e-07, + "loss": 0.8318, + "num_input_tokens_seen": 317128395, + "step": 9763 + }, + { + "epoch": 0.880551923163638, + "flos": 26573421600480.0, + "grad_norm": 2.334996981404414, + "learning_rate": 1.4771849504748768e-07, + "loss": 0.6754, + "num_input_tokens_seen": 317157785, + "step": 9764 + }, + { + "epoch": 0.8806421066871083, + "flos": 24355152271680.0, + "grad_norm": 1.618873123140759, + "learning_rate": 1.4749822262897517e-07, + "loss": 0.8198, + "num_input_tokens_seen": 317186735, + "step": 9765 + }, + { + "epoch": 0.8807322902105785, + "flos": 53499976087200.0, + "grad_norm": 1.9839362913562524, + "learning_rate": 1.4727810827612895e-07, + "loss": 0.6839, + "num_input_tokens_seen": 317219445, + "step": 9766 + }, + { + "epoch": 0.8808224737340488, + "flos": 25738456918080.0, + "grad_norm": 2.1480149841993588, + "learning_rate": 1.470581520077303e-07, + "loss": 0.7517, + "num_input_tokens_seen": 317247485, + "step": 9767 + }, + { + "epoch": 0.8809126572575191, + "flos": 13206038087040.0, + "grad_norm": 5.354623361943079, + "learning_rate": 1.4683835384254705e-07, + "loss": 0.7373, + "num_input_tokens_seen": 317272390, + "step": 9768 + }, + { + "epoch": 0.8810028407809893, + "flos": 21653043979680.0, + "grad_norm": 1.9070001595731607, + "learning_rate": 1.4661871379933376e-07, + "loss": 0.8229, + "num_input_tokens_seen": 317299625, + "step": 9769 + }, + { + "epoch": 0.8810930243044596, + "flos": 17030590684320.0, + "grad_norm": 2.034326541634068, + "learning_rate": 1.4639923189683169e-07, + "loss": 0.7922, + "num_input_tokens_seen": 317326270, + "step": 9770 + }, + { + "epoch": 0.8811832078279298, + "flos": 21221248414080.0, + "grad_norm": 1.642808460319883, + "learning_rate": 1.461799081537669e-07, + "loss": 0.7928, + "num_input_tokens_seen": 317357795, + "step": 9771 + }, + { + "epoch": 0.8812733913514001, + "flos": 21512853880320.0, + "grad_norm": 2.4158861500270614, + "learning_rate": 1.4596074258885514e-07, + "loss": 0.8086, + "num_input_tokens_seen": 317386680, + "step": 9772 + }, + { + "epoch": 0.8813635748748704, + "flos": 25116472192320.0, + "grad_norm": 1.6888653212221894, + "learning_rate": 1.4574173522079502e-07, + "loss": 0.7159, + "num_input_tokens_seen": 317415130, + "step": 9773 + }, + { + "epoch": 0.8814537583983406, + "flos": 23438811517440.0, + "grad_norm": 4.625234030911111, + "learning_rate": 1.4552288606827513e-07, + "loss": 0.7193, + "num_input_tokens_seen": 317444395, + "step": 9774 + }, + { + "epoch": 0.8815439419218108, + "flos": 20201372677440.0, + "grad_norm": 4.340061735074492, + "learning_rate": 1.4530419514996761e-07, + "loss": 0.8202, + "num_input_tokens_seen": 317472345, + "step": 9775 + }, + { + "epoch": 0.8816341254452812, + "flos": 71353296869280.0, + "grad_norm": 0.6964347868771269, + "learning_rate": 1.4508566248453291e-07, + "loss": 0.5878, + "num_input_tokens_seen": 317559230, + "step": 9776 + }, + { + "epoch": 0.8817243089687514, + "flos": 34045570069440.0, + "grad_norm": 3.2634775502254025, + "learning_rate": 1.448672880906172e-07, + "loss": 0.7382, + "num_input_tokens_seen": 317586960, + "step": 9777 + }, + { + "epoch": 0.8818144924922217, + "flos": 41914234268160.0, + "grad_norm": 2.5459978898512374, + "learning_rate": 1.4464907198685382e-07, + "loss": 0.6982, + "num_input_tokens_seen": 317619555, + "step": 9778 + }, + { + "epoch": 0.8819046760156919, + "flos": 35536721390400.0, + "grad_norm": 2.0767840885052435, + "learning_rate": 1.444310141918621e-07, + "loss": 0.7467, + "num_input_tokens_seen": 317650030, + "step": 9779 + }, + { + "epoch": 0.8819948595391622, + "flos": 21505048230720.0, + "grad_norm": 3.268689191847085, + "learning_rate": 1.4421311472424735e-07, + "loss": 0.7728, + "num_input_tokens_seen": 317677370, + "step": 9780 + }, + { + "epoch": 0.8820850430626325, + "flos": 18707545133760.0, + "grad_norm": 1.857229004161392, + "learning_rate": 1.4399537360260273e-07, + "loss": 0.7798, + "num_input_tokens_seen": 317702835, + "step": 9781 + }, + { + "epoch": 0.8821752265861027, + "flos": 24244592204640.0, + "grad_norm": 1.9797364335007757, + "learning_rate": 1.4377779084550645e-07, + "loss": 0.7702, + "num_input_tokens_seen": 317728320, + "step": 9782 + }, + { + "epoch": 0.8822654101095729, + "flos": 23188916255520.0, + "grad_norm": 2.673441199205917, + "learning_rate": 1.4356036647152413e-07, + "loss": 0.7364, + "num_input_tokens_seen": 317757870, + "step": 9783 + }, + { + "epoch": 0.8823555936330433, + "flos": 34699749540960.0, + "grad_norm": 1.5969604618621898, + "learning_rate": 1.4334310049920785e-07, + "loss": 0.7204, + "num_input_tokens_seen": 317787470, + "step": 9784 + }, + { + "epoch": 0.8824457771565135, + "flos": 70088952655680.0, + "grad_norm": 0.607752571976122, + "learning_rate": 1.431259929470956e-07, + "loss": 0.5284, + "num_input_tokens_seen": 317878585, + "step": 9785 + }, + { + "epoch": 0.8825359606799837, + "flos": 32073776384640.0, + "grad_norm": 2.2111668552015127, + "learning_rate": 1.4290904383371237e-07, + "loss": 0.8192, + "num_input_tokens_seen": 317909840, + "step": 9786 + }, + { + "epoch": 0.882626144203454, + "flos": 31458779573760.0, + "grad_norm": 2.0770733689273535, + "learning_rate": 1.4269225317756961e-07, + "loss": 0.7272, + "num_input_tokens_seen": 317940370, + "step": 9787 + }, + { + "epoch": 0.8827163277269243, + "flos": 18634792446240.0, + "grad_norm": 2.3005113945363544, + "learning_rate": 1.424756209971645e-07, + "loss": 0.7065, + "num_input_tokens_seen": 317967000, + "step": 9788 + }, + { + "epoch": 0.8828065112503946, + "flos": 17322642187680.0, + "grad_norm": 2.2695265226593944, + "learning_rate": 1.4225914731098199e-07, + "loss": 0.7413, + "num_input_tokens_seen": 317993520, + "step": 9789 + }, + { + "epoch": 0.8828966947738648, + "flos": 23516544952800.0, + "grad_norm": 1.997439056154289, + "learning_rate": 1.4204283213749248e-07, + "loss": 0.6742, + "num_input_tokens_seen": 318023975, + "step": 9790 + }, + { + "epoch": 0.8829868782973351, + "flos": 21913495453440.0, + "grad_norm": 1.8087896978296651, + "learning_rate": 1.4182667549515315e-07, + "loss": 0.8025, + "num_input_tokens_seen": 318052145, + "step": 9791 + }, + { + "epoch": 0.8830770618208054, + "flos": 25811767152000.0, + "grad_norm": 2.1453743477704617, + "learning_rate": 1.4161067740240752e-07, + "loss": 0.7263, + "num_input_tokens_seen": 318080610, + "step": 9792 + }, + { + "epoch": 0.8831672453442756, + "flos": 28398260289600.0, + "grad_norm": 1.7126118031781195, + "learning_rate": 1.4139483787768614e-07, + "loss": 0.736, + "num_input_tokens_seen": 318110210, + "step": 9793 + }, + { + "epoch": 0.8832574288677458, + "flos": 29417281121760.0, + "grad_norm": 2.4809785093816656, + "learning_rate": 1.4117915693940584e-07, + "loss": 0.6531, + "num_input_tokens_seen": 318141680, + "step": 9794 + }, + { + "epoch": 0.8833476123912162, + "flos": 36377781913440.0, + "grad_norm": 1.735080815942287, + "learning_rate": 1.409636346059684e-07, + "loss": 0.7399, + "num_input_tokens_seen": 318174045, + "step": 9795 + }, + { + "epoch": 0.8834377959146864, + "flos": 21723640821120.0, + "grad_norm": 2.745952575440775, + "learning_rate": 1.4074827089576501e-07, + "loss": 0.7753, + "num_input_tokens_seen": 318200565, + "step": 9796 + }, + { + "epoch": 0.8835279794381566, + "flos": 19873000584960.0, + "grad_norm": 7.336636088525789, + "learning_rate": 1.4053306582717085e-07, + "loss": 0.7588, + "num_input_tokens_seen": 318228380, + "step": 9797 + }, + { + "epoch": 0.8836181629616269, + "flos": 23042853334080.0, + "grad_norm": 1.9882188948350976, + "learning_rate": 1.4031801941854827e-07, + "loss": 0.7701, + "num_input_tokens_seen": 318258130, + "step": 9798 + }, + { + "epoch": 0.8837083464850972, + "flos": 26505017774880.0, + "grad_norm": 1.91077754676105, + "learning_rate": 1.401031316882466e-07, + "loss": 0.7109, + "num_input_tokens_seen": 318290685, + "step": 9799 + }, + { + "epoch": 0.8837985300085675, + "flos": 22308450053280.0, + "grad_norm": 2.6276831764664874, + "learning_rate": 1.39888402654601e-07, + "loss": 0.7582, + "num_input_tokens_seen": 318317585, + "step": 9800 + }, + { + "epoch": 0.8838887135320377, + "flos": 60906316880160.0, + "grad_norm": 0.6078664740988802, + "learning_rate": 1.3967383233593344e-07, + "loss": 0.5866, + "num_input_tokens_seen": 318415355, + "step": 9801 + }, + { + "epoch": 0.8839788970555079, + "flos": 22934337603840.0, + "grad_norm": 2.0218682408404094, + "learning_rate": 1.3945942075055218e-07, + "loss": 0.704, + "num_input_tokens_seen": 318443110, + "step": 9802 + }, + { + "epoch": 0.8840690805789783, + "flos": 18958778507040.0, + "grad_norm": 2.4933733854119824, + "learning_rate": 1.3924516791675212e-07, + "loss": 0.769, + "num_input_tokens_seen": 318470955, + "step": 9803 + }, + { + "epoch": 0.8841592641024485, + "flos": 22933036662240.0, + "grad_norm": 1.6759498351335083, + "learning_rate": 1.3903107385281487e-07, + "loss": 0.654, + "num_input_tokens_seen": 318503655, + "step": 9804 + }, + { + "epoch": 0.8842494476259187, + "flos": 19724633138400.0, + "grad_norm": 1.9679053780595879, + "learning_rate": 1.3881713857700717e-07, + "loss": 0.7533, + "num_input_tokens_seen": 318533115, + "step": 9805 + }, + { + "epoch": 0.884339631149389, + "flos": 28726558042560.0, + "grad_norm": 2.4144680830599086, + "learning_rate": 1.3860336210758372e-07, + "loss": 0.7223, + "num_input_tokens_seen": 318560690, + "step": 9806 + }, + { + "epoch": 0.8844298146728593, + "flos": 24240280512480.0, + "grad_norm": 1.6639641387954078, + "learning_rate": 1.3838974446278506e-07, + "loss": 0.7024, + "num_input_tokens_seen": 318588935, + "step": 9807 + }, + { + "epoch": 0.8845199981963295, + "flos": 46724720877600.0, + "grad_norm": 1.821521406719652, + "learning_rate": 1.3817628566083817e-07, + "loss": 0.7203, + "num_input_tokens_seen": 318621115, + "step": 9808 + }, + { + "epoch": 0.8846101817197998, + "flos": 21513076898880.0, + "grad_norm": 1.6664838770130241, + "learning_rate": 1.3796298571995712e-07, + "loss": 0.754, + "num_input_tokens_seen": 318649350, + "step": 9809 + }, + { + "epoch": 0.88470036524327, + "flos": 19545929434080.0, + "grad_norm": 2.315624909394503, + "learning_rate": 1.377498446583405e-07, + "loss": 0.642, + "num_input_tokens_seen": 318677290, + "step": 9810 + }, + { + "epoch": 0.8847905487667403, + "flos": 25337778175200.0, + "grad_norm": 1.910709729972267, + "learning_rate": 1.3753686249417596e-07, + "loss": 0.7168, + "num_input_tokens_seen": 318703815, + "step": 9811 + }, + { + "epoch": 0.8848807322902106, + "flos": 19068149141760.0, + "grad_norm": 2.415381901242807, + "learning_rate": 1.373240392456354e-07, + "loss": 0.7011, + "num_input_tokens_seen": 318731240, + "step": 9812 + }, + { + "epoch": 0.8849709158136808, + "flos": 23225013826080.0, + "grad_norm": 1.8004621442286124, + "learning_rate": 1.37111374930879e-07, + "loss": 0.6786, + "num_input_tokens_seen": 318758670, + "step": 9813 + }, + { + "epoch": 0.885061099337151, + "flos": 27415522876800.0, + "grad_norm": 1.6587626992424418, + "learning_rate": 1.3689886956805176e-07, + "loss": 0.6734, + "num_input_tokens_seen": 318787900, + "step": 9814 + }, + { + "epoch": 0.8851512828606214, + "flos": 22310828917920.0, + "grad_norm": 1.7064915552430504, + "learning_rate": 1.3668652317528585e-07, + "loss": 0.7343, + "num_input_tokens_seen": 318815990, + "step": 9815 + }, + { + "epoch": 0.8852414663840916, + "flos": 21549285978720.0, + "grad_norm": 1.7637302731195492, + "learning_rate": 1.3647433577070012e-07, + "loss": 0.7587, + "num_input_tokens_seen": 318846135, + "step": 9816 + }, + { + "epoch": 0.8853316499075619, + "flos": 67963476248640.0, + "grad_norm": 0.6508471287312576, + "learning_rate": 1.3626230737239942e-07, + "loss": 0.5486, + "num_input_tokens_seen": 318930970, + "step": 9817 + }, + { + "epoch": 0.8854218334310322, + "flos": 69961034310720.0, + "grad_norm": 0.7043890928129621, + "learning_rate": 1.3605043799847527e-07, + "loss": 0.5553, + "num_input_tokens_seen": 319026065, + "step": 9818 + }, + { + "epoch": 0.8855120169545024, + "flos": 46979634057120.0, + "grad_norm": 1.54486014415229, + "learning_rate": 1.3583872766700567e-07, + "loss": 0.6798, + "num_input_tokens_seen": 319059655, + "step": 9819 + }, + { + "epoch": 0.8856022004779727, + "flos": 41912264270880.0, + "grad_norm": 3.267325656620584, + "learning_rate": 1.3562717639605437e-07, + "loss": 0.7227, + "num_input_tokens_seen": 319092065, + "step": 9820 + }, + { + "epoch": 0.8856923840014429, + "flos": 21435863840160.0, + "grad_norm": 2.6490497612835338, + "learning_rate": 1.3541578420367229e-07, + "loss": 0.7127, + "num_input_tokens_seen": 319120930, + "step": 9821 + }, + { + "epoch": 0.8857825675249132, + "flos": 20419482060960.0, + "grad_norm": 3.4122555487972837, + "learning_rate": 1.3520455110789697e-07, + "loss": 0.7578, + "num_input_tokens_seen": 319149110, + "step": 9822 + }, + { + "epoch": 0.8858727510483835, + "flos": 18233333138400.0, + "grad_norm": 2.1185597832288674, + "learning_rate": 1.3499347712675158e-07, + "loss": 0.7203, + "num_input_tokens_seen": 319176860, + "step": 9823 + }, + { + "epoch": 0.8859629345718537, + "flos": 31636219506240.0, + "grad_norm": 3.9161047566228038, + "learning_rate": 1.3478256227824635e-07, + "loss": 0.7039, + "num_input_tokens_seen": 319206315, + "step": 9824 + }, + { + "epoch": 0.8860531180953239, + "flos": 30183581790240.0, + "grad_norm": 1.7614335804137573, + "learning_rate": 1.3457180658037759e-07, + "loss": 0.6521, + "num_input_tokens_seen": 319237595, + "step": 9825 + }, + { + "epoch": 0.8861433016187943, + "flos": 20203528523520.0, + "grad_norm": 1.7721916225320151, + "learning_rate": 1.3436121005112843e-07, + "loss": 0.7745, + "num_input_tokens_seen": 319267520, + "step": 9826 + }, + { + "epoch": 0.8862334851422645, + "flos": 26540780817600.0, + "grad_norm": 2.9895387085741967, + "learning_rate": 1.3415077270846719e-07, + "loss": 0.7133, + "num_input_tokens_seen": 319296875, + "step": 9827 + }, + { + "epoch": 0.8863236686657348, + "flos": 15173371400640.0, + "grad_norm": 3.4543914543851852, + "learning_rate": 1.3394049457035105e-07, + "loss": 0.8519, + "num_input_tokens_seen": 319321705, + "step": 9828 + }, + { + "epoch": 0.886413852189205, + "flos": 20237693266560.0, + "grad_norm": 1.8291070370813693, + "learning_rate": 1.3373037565472034e-07, + "loss": 0.7507, + "num_input_tokens_seen": 319349200, + "step": 9829 + }, + { + "epoch": 0.8865040357126753, + "flos": 17104421294880.0, + "grad_norm": 2.739484536211038, + "learning_rate": 1.3352041597950537e-07, + "loss": 0.7189, + "num_input_tokens_seen": 319376155, + "step": 9830 + }, + { + "epoch": 0.8865942192361456, + "flos": 21214111820160.0, + "grad_norm": 2.011236218108285, + "learning_rate": 1.333106155626196e-07, + "loss": 0.7321, + "num_input_tokens_seen": 319403310, + "step": 9831 + }, + { + "epoch": 0.8866844027596158, + "flos": 23840902711200.0, + "grad_norm": 1.6729089979832876, + "learning_rate": 1.331009744219651e-07, + "loss": 0.6846, + "num_input_tokens_seen": 319432305, + "step": 9832 + }, + { + "epoch": 0.886774586283086, + "flos": 69530130819360.0, + "grad_norm": 0.6682199819939071, + "learning_rate": 1.3289149257542943e-07, + "loss": 0.6104, + "num_input_tokens_seen": 319519790, + "step": 9833 + }, + { + "epoch": 0.8868647698065564, + "flos": 31969089139680.0, + "grad_norm": 3.0766207509800956, + "learning_rate": 1.3268217004088666e-07, + "loss": 0.6473, + "num_input_tokens_seen": 319550980, + "step": 9834 + }, + { + "epoch": 0.8869549533300266, + "flos": 29309880484320.0, + "grad_norm": 2.037619127406794, + "learning_rate": 1.3247300683619788e-07, + "loss": 0.6307, + "num_input_tokens_seen": 319581370, + "step": 9835 + }, + { + "epoch": 0.8870451368534968, + "flos": 24062952089280.0, + "grad_norm": 3.202841546147196, + "learning_rate": 1.3226400297920903e-07, + "loss": 0.7187, + "num_input_tokens_seen": 319610245, + "step": 9836 + }, + { + "epoch": 0.8871353203769671, + "flos": 26504014191360.0, + "grad_norm": 1.866094357029614, + "learning_rate": 1.3205515848775428e-07, + "loss": 0.7225, + "num_input_tokens_seen": 319637355, + "step": 9837 + }, + { + "epoch": 0.8872255039004374, + "flos": 16631212883040.0, + "grad_norm": 2.0863177349415385, + "learning_rate": 1.3184647337965316e-07, + "loss": 0.683, + "num_input_tokens_seen": 319664925, + "step": 9838 + }, + { + "epoch": 0.8873156874239077, + "flos": 36887013556320.0, + "grad_norm": 1.5888736398820276, + "learning_rate": 1.3163794767271163e-07, + "loss": 0.6899, + "num_input_tokens_seen": 319698770, + "step": 9839 + }, + { + "epoch": 0.8874058709473779, + "flos": 42747823669440.0, + "grad_norm": 1.7212171706848263, + "learning_rate": 1.314295813847226e-07, + "loss": 0.7131, + "num_input_tokens_seen": 319731450, + "step": 9840 + }, + { + "epoch": 0.8874960544708482, + "flos": 24062952089280.0, + "grad_norm": 2.0842096265991676, + "learning_rate": 1.3122137453346515e-07, + "loss": 0.8443, + "num_input_tokens_seen": 319760915, + "step": 9841 + }, + { + "epoch": 0.8875862379943185, + "flos": 20125757918400.0, + "grad_norm": 1.9396982335831836, + "learning_rate": 1.3101332713670376e-07, + "loss": 0.798, + "num_input_tokens_seen": 319790160, + "step": 9842 + }, + { + "epoch": 0.8876764215177887, + "flos": 62957373694080.0, + "grad_norm": 0.7024855567085415, + "learning_rate": 1.3080543921219133e-07, + "loss": 0.6061, + "num_input_tokens_seen": 319873960, + "step": 9843 + }, + { + "epoch": 0.8877666050412589, + "flos": 18780260651520.0, + "grad_norm": 7.221661661678821, + "learning_rate": 1.3059771077766478e-07, + "loss": 0.7344, + "num_input_tokens_seen": 319899795, + "step": 9844 + }, + { + "epoch": 0.8878567885647293, + "flos": 20381600341920.0, + "grad_norm": 2.1757922078101535, + "learning_rate": 1.3039014185085018e-07, + "loss": 0.7822, + "num_input_tokens_seen": 319922625, + "step": 9845 + }, + { + "epoch": 0.8879469720881995, + "flos": 26832423453600.0, + "grad_norm": 2.0000972724147927, + "learning_rate": 1.301827324494571e-07, + "loss": 0.7492, + "num_input_tokens_seen": 319949780, + "step": 9846 + }, + { + "epoch": 0.8880371556116697, + "flos": 35932493724960.0, + "grad_norm": 2.556403187682961, + "learning_rate": 1.2997548259118342e-07, + "loss": 0.7804, + "num_input_tokens_seen": 319975855, + "step": 9847 + }, + { + "epoch": 0.88812733913514, + "flos": 24754604412480.0, + "grad_norm": 2.257545006557903, + "learning_rate": 1.2976839229371272e-07, + "loss": 0.7692, + "num_input_tokens_seen": 320005090, + "step": 9848 + }, + { + "epoch": 0.8882175226586103, + "flos": 60760960184160.0, + "grad_norm": 0.6200972906810456, + "learning_rate": 1.2956146157471515e-07, + "loss": 0.5947, + "num_input_tokens_seen": 320100520, + "step": 9849 + }, + { + "epoch": 0.8883077061820805, + "flos": 28143495789120.0, + "grad_norm": 1.539307628029273, + "learning_rate": 1.2935469045184745e-07, + "loss": 0.7619, + "num_input_tokens_seen": 320129860, + "step": 9850 + }, + { + "epoch": 0.8883978897055508, + "flos": 21290469974400.0, + "grad_norm": 2.4970324069216723, + "learning_rate": 1.291480789427517e-07, + "loss": 0.7216, + "num_input_tokens_seen": 320156020, + "step": 9851 + }, + { + "epoch": 0.888488073229021, + "flos": 27157821965280.0, + "grad_norm": 2.0552355829421605, + "learning_rate": 1.2894162706505807e-07, + "loss": 0.7842, + "num_input_tokens_seen": 320186075, + "step": 9852 + }, + { + "epoch": 0.8885782567524914, + "flos": 20528852695680.0, + "grad_norm": 1.6983940841895209, + "learning_rate": 1.2873533483638155e-07, + "loss": 0.7342, + "num_input_tokens_seen": 320215180, + "step": 9853 + }, + { + "epoch": 0.8886684402759616, + "flos": 23227950237120.0, + "grad_norm": 2.3124039088724433, + "learning_rate": 1.285292022743243e-07, + "loss": 0.6911, + "num_input_tokens_seen": 320246860, + "step": 9854 + }, + { + "epoch": 0.8887586237994318, + "flos": 26937816924000.0, + "grad_norm": 1.8806820921568972, + "learning_rate": 1.2832322939647467e-07, + "loss": 0.6886, + "num_input_tokens_seen": 320276200, + "step": 9855 + }, + { + "epoch": 0.8888488073229021, + "flos": 22023051936960.0, + "grad_norm": 1.9193998696278969, + "learning_rate": 1.281174162204075e-07, + "loss": 0.7276, + "num_input_tokens_seen": 320306920, + "step": 9856 + }, + { + "epoch": 0.8889389908463724, + "flos": 24569581848960.0, + "grad_norm": 1.5914420826566582, + "learning_rate": 1.2791176276368366e-07, + "loss": 0.6209, + "num_input_tokens_seen": 320337865, + "step": 9857 + }, + { + "epoch": 0.8890291743698426, + "flos": 29999116773120.0, + "grad_norm": 1.594502589235928, + "learning_rate": 1.2770626904385128e-07, + "loss": 0.7551, + "num_input_tokens_seen": 320370850, + "step": 9858 + }, + { + "epoch": 0.8891193578933129, + "flos": 33899395638720.0, + "grad_norm": 2.002423725175689, + "learning_rate": 1.2750093507844306e-07, + "loss": 0.6047, + "num_input_tokens_seen": 320398260, + "step": 9859 + }, + { + "epoch": 0.8892095414167831, + "flos": 25263873225120.0, + "grad_norm": 1.526418185953251, + "learning_rate": 1.272957608849805e-07, + "loss": 0.7798, + "num_input_tokens_seen": 320428890, + "step": 9860 + }, + { + "epoch": 0.8892997249402534, + "flos": 22095024059520.0, + "grad_norm": 1.694482908003426, + "learning_rate": 1.270907464809694e-07, + "loss": 0.7163, + "num_input_tokens_seen": 320459150, + "step": 9861 + }, + { + "epoch": 0.8893899084637237, + "flos": 24571997883360.0, + "grad_norm": 1.751984628927831, + "learning_rate": 1.2688589188390285e-07, + "loss": 0.7453, + "num_input_tokens_seen": 320488775, + "step": 9862 + }, + { + "epoch": 0.8894800919871939, + "flos": 45594731111040.0, + "grad_norm": 2.0221505214686246, + "learning_rate": 1.2668119711126023e-07, + "loss": 0.7506, + "num_input_tokens_seen": 320521965, + "step": 9863 + }, + { + "epoch": 0.8895702755106643, + "flos": 28507482245280.0, + "grad_norm": 1.7108795295544446, + "learning_rate": 1.2647666218050735e-07, + "loss": 0.7448, + "num_input_tokens_seen": 320552175, + "step": 9864 + }, + { + "epoch": 0.8896604590341345, + "flos": 14663619381120.0, + "grad_norm": 2.163956622346487, + "learning_rate": 1.2627228710909643e-07, + "loss": 0.7059, + "num_input_tokens_seen": 320579340, + "step": 9865 + }, + { + "epoch": 0.8897506425576047, + "flos": 26686063174080.0, + "grad_norm": 1.7608238366312179, + "learning_rate": 1.260680719144649e-07, + "loss": 0.7082, + "num_input_tokens_seen": 320609340, + "step": 9866 + }, + { + "epoch": 0.889840826081075, + "flos": 13453071277440.0, + "grad_norm": 3.137510814876079, + "learning_rate": 1.2586401661403877e-07, + "loss": 0.6759, + "num_input_tokens_seen": 320631655, + "step": 9867 + }, + { + "epoch": 0.8899310096045453, + "flos": 26577621783360.0, + "grad_norm": 2.786214105949048, + "learning_rate": 1.2566012122522817e-07, + "loss": 0.7134, + "num_input_tokens_seen": 320660960, + "step": 9868 + }, + { + "epoch": 0.8900211931280155, + "flos": 25220973588480.0, + "grad_norm": 2.7276737604640644, + "learning_rate": 1.254563857654316e-07, + "loss": 0.7141, + "num_input_tokens_seen": 320685865, + "step": 9869 + }, + { + "epoch": 0.8901113766514858, + "flos": 18774610848000.0, + "grad_norm": 3.6183469821339243, + "learning_rate": 1.2525281025203205e-07, + "loss": 0.6309, + "num_input_tokens_seen": 320712825, + "step": 9870 + }, + { + "epoch": 0.890201560174956, + "flos": 26795954185440.0, + "grad_norm": 2.6358661996981527, + "learning_rate": 1.2504939470240006e-07, + "loss": 0.7222, + "num_input_tokens_seen": 320741840, + "step": 9871 + }, + { + "epoch": 0.8902917436984263, + "flos": 26394755065920.0, + "grad_norm": 1.6447695742913804, + "learning_rate": 1.2484613913389196e-07, + "loss": 0.8076, + "num_input_tokens_seen": 320772665, + "step": 9872 + }, + { + "epoch": 0.8903819272218966, + "flos": 28216917532320.0, + "grad_norm": 2.0989498646252143, + "learning_rate": 1.2464304356385057e-07, + "loss": 0.6544, + "num_input_tokens_seen": 320803485, + "step": 9873 + }, + { + "epoch": 0.8904721107453668, + "flos": 27046815861120.0, + "grad_norm": 2.083969845493223, + "learning_rate": 1.2444010800960558e-07, + "loss": 0.7998, + "num_input_tokens_seen": 320831070, + "step": 9874 + }, + { + "epoch": 0.890562294268837, + "flos": 35096859986880.0, + "grad_norm": 2.4924559254322167, + "learning_rate": 1.2423733248847267e-07, + "loss": 0.7269, + "num_input_tokens_seen": 320861005, + "step": 9875 + }, + { + "epoch": 0.8906524777923074, + "flos": 32253186314400.0, + "grad_norm": 3.481145893505838, + "learning_rate": 1.2403471701775293e-07, + "loss": 0.6806, + "num_input_tokens_seen": 320886670, + "step": 9876 + }, + { + "epoch": 0.8907426613157776, + "flos": 17286879144960.0, + "grad_norm": 2.13223314625827, + "learning_rate": 1.2383226161473515e-07, + "loss": 0.7777, + "num_input_tokens_seen": 320913650, + "step": 9877 + }, + { + "epoch": 0.8908328448392479, + "flos": 25848719627040.0, + "grad_norm": 1.9857317273901987, + "learning_rate": 1.2362996629669376e-07, + "loss": 0.8164, + "num_input_tokens_seen": 320941640, + "step": 9878 + }, + { + "epoch": 0.8909230283627181, + "flos": 23662347685920.0, + "grad_norm": 1.902575844895823, + "learning_rate": 1.2342783108089007e-07, + "loss": 0.7544, + "num_input_tokens_seen": 320972070, + "step": 9879 + }, + { + "epoch": 0.8910132118861884, + "flos": 25484361473280.0, + "grad_norm": 1.9722441289995438, + "learning_rate": 1.2322585598457135e-07, + "loss": 0.7985, + "num_input_tokens_seen": 321001025, + "step": 9880 + }, + { + "epoch": 0.8911033954096587, + "flos": 25410270674400.0, + "grad_norm": 1.3895456647366533, + "learning_rate": 1.2302404102497034e-07, + "loss": 0.7139, + "num_input_tokens_seen": 321032755, + "step": 9881 + }, + { + "epoch": 0.8911935789331289, + "flos": 16157892961920.0, + "grad_norm": 2.64662364860698, + "learning_rate": 1.228223862193083e-07, + "loss": 0.6678, + "num_input_tokens_seen": 321059295, + "step": 9882 + }, + { + "epoch": 0.8912837624565991, + "flos": 13422958038240.0, + "grad_norm": 2.496155739553271, + "learning_rate": 1.2262089158479038e-07, + "loss": 0.7508, + "num_input_tokens_seen": 321082995, + "step": 9883 + }, + { + "epoch": 0.8913739459800695, + "flos": 18416460044160.0, + "grad_norm": 2.01764725366355, + "learning_rate": 1.2241955713861042e-07, + "loss": 0.7851, + "num_input_tokens_seen": 321110370, + "step": 9884 + }, + { + "epoch": 0.8914641295035397, + "flos": 25592505505920.0, + "grad_norm": 1.5164893920054514, + "learning_rate": 1.222183828979464e-07, + "loss": 0.6806, + "num_input_tokens_seen": 321142525, + "step": 9885 + }, + { + "epoch": 0.8915543130270099, + "flos": 74632302968160.0, + "grad_norm": 1.6259324259681573, + "learning_rate": 1.2201736887996372e-07, + "loss": 0.7392, + "num_input_tokens_seen": 321180835, + "step": 9886 + }, + { + "epoch": 0.8916444965504803, + "flos": 20747556795360.0, + "grad_norm": 2.4625150922128825, + "learning_rate": 1.2181651510181444e-07, + "loss": 0.8199, + "num_input_tokens_seen": 321206290, + "step": 9887 + }, + { + "epoch": 0.8917346800739505, + "flos": 18379879266720.0, + "grad_norm": 1.8777607620127912, + "learning_rate": 1.2161582158063622e-07, + "loss": 0.7695, + "num_input_tokens_seen": 321233835, + "step": 9888 + }, + { + "epoch": 0.8918248635974207, + "flos": 24281172982080.0, + "grad_norm": 1.8970190329755676, + "learning_rate": 1.214152883335533e-07, + "loss": 0.6474, + "num_input_tokens_seen": 321265095, + "step": 9889 + }, + { + "epoch": 0.891915047120891, + "flos": 18416422874400.0, + "grad_norm": 2.3722445917476915, + "learning_rate": 1.2121491537767648e-07, + "loss": 0.6961, + "num_input_tokens_seen": 321292625, + "step": 9890 + }, + { + "epoch": 0.8920052306443613, + "flos": 19762477687680.0, + "grad_norm": 1.6380500568477796, + "learning_rate": 1.2101470273010294e-07, + "loss": 0.7768, + "num_input_tokens_seen": 321319935, + "step": 9891 + }, + { + "epoch": 0.8920954141678316, + "flos": 15465088376160.0, + "grad_norm": 2.2877943672927867, + "learning_rate": 1.2081465040791528e-07, + "loss": 0.7108, + "num_input_tokens_seen": 321344880, + "step": 9892 + }, + { + "epoch": 0.8921855976913018, + "flos": 30832483155840.0, + "grad_norm": 1.9619803524258477, + "learning_rate": 1.2061475842818335e-07, + "loss": 0.7046, + "num_input_tokens_seen": 321376315, + "step": 9893 + }, + { + "epoch": 0.892275781214772, + "flos": 29670261473760.0, + "grad_norm": 1.57747878639237, + "learning_rate": 1.2041502680796313e-07, + "loss": 0.7182, + "num_input_tokens_seen": 321406340, + "step": 9894 + }, + { + "epoch": 0.8923659647382424, + "flos": 26500222875840.0, + "grad_norm": 1.5381235269464277, + "learning_rate": 1.2021545556429648e-07, + "loss": 0.7468, + "num_input_tokens_seen": 321433925, + "step": 9895 + }, + { + "epoch": 0.8924561482617126, + "flos": 23808001740000.0, + "grad_norm": 1.487021887099313, + "learning_rate": 1.2001604471421245e-07, + "loss": 0.7403, + "num_input_tokens_seen": 321465320, + "step": 9896 + }, + { + "epoch": 0.8925463317851828, + "flos": 24538948233120.0, + "grad_norm": 2.2033069075568976, + "learning_rate": 1.1981679427472567e-07, + "loss": 0.7922, + "num_input_tokens_seen": 321495845, + "step": 9897 + }, + { + "epoch": 0.8926365153086531, + "flos": 25042084035360.0, + "grad_norm": 1.8656120517096073, + "learning_rate": 1.196177042628368e-07, + "loss": 0.7938, + "num_input_tokens_seen": 321525025, + "step": 9898 + }, + { + "epoch": 0.8927266988321234, + "flos": 15720224574240.0, + "grad_norm": 3.090352823058216, + "learning_rate": 1.194187746955344e-07, + "loss": 0.7461, + "num_input_tokens_seen": 321548225, + "step": 9899 + }, + { + "epoch": 0.8928168823555936, + "flos": 29339510516640.0, + "grad_norm": 1.6444995723038347, + "learning_rate": 1.1922000558979094e-07, + "loss": 0.7163, + "num_input_tokens_seen": 321575030, + "step": 9900 + }, + { + "epoch": 0.8929070658790639, + "flos": 22526708115840.0, + "grad_norm": 1.4754976338309715, + "learning_rate": 1.1902139696256752e-07, + "loss": 0.7724, + "num_input_tokens_seen": 321602865, + "step": 9901 + }, + { + "epoch": 0.8929972494025341, + "flos": 20602200099360.0, + "grad_norm": 3.9633343609379965, + "learning_rate": 1.188229488308099e-07, + "loss": 0.6794, + "num_input_tokens_seen": 321630935, + "step": 9902 + }, + { + "epoch": 0.8930874329260045, + "flos": 57560622498240.0, + "grad_norm": 0.5601170184284627, + "learning_rate": 1.1862466121145098e-07, + "loss": 0.5564, + "num_input_tokens_seen": 321730300, + "step": 9903 + }, + { + "epoch": 0.8931776164494747, + "flos": 16193804683680.0, + "grad_norm": 2.2437419753077252, + "learning_rate": 1.184265341214099e-07, + "loss": 0.6835, + "num_input_tokens_seen": 321757210, + "step": 9904 + }, + { + "epoch": 0.8932677999729449, + "flos": 31815555096480.0, + "grad_norm": 1.8714315790362301, + "learning_rate": 1.182285675775918e-07, + "loss": 0.7655, + "num_input_tokens_seen": 321788425, + "step": 9905 + }, + { + "epoch": 0.8933579834964152, + "flos": 18270025425120.0, + "grad_norm": 2.4093723598916004, + "learning_rate": 1.1803076159688851e-07, + "loss": 0.7238, + "num_input_tokens_seen": 321814170, + "step": 9906 + }, + { + "epoch": 0.8934481670198855, + "flos": 25811432624160.0, + "grad_norm": 1.8889666764130593, + "learning_rate": 1.1783311619617741e-07, + "loss": 0.7757, + "num_input_tokens_seen": 321843460, + "step": 9907 + }, + { + "epoch": 0.8935383505433557, + "flos": 30729803077920.0, + "grad_norm": 2.316856404866094, + "learning_rate": 1.1763563139232302e-07, + "loss": 0.708, + "num_input_tokens_seen": 321872235, + "step": 9908 + }, + { + "epoch": 0.893628534066826, + "flos": 24682334931840.0, + "grad_norm": 1.8517262833020123, + "learning_rate": 1.1743830720217562e-07, + "loss": 0.6768, + "num_input_tokens_seen": 321897880, + "step": 9909 + }, + { + "epoch": 0.8937187175902963, + "flos": 26357765421120.0, + "grad_norm": 2.561495048897921, + "learning_rate": 1.1724114364257243e-07, + "loss": 0.7173, + "num_input_tokens_seen": 321926400, + "step": 9910 + }, + { + "epoch": 0.8938089011137665, + "flos": 27924940368480.0, + "grad_norm": 5.544837426983302, + "learning_rate": 1.1704414073033619e-07, + "loss": 0.7479, + "num_input_tokens_seen": 321956700, + "step": 9911 + }, + { + "epoch": 0.8938990846372368, + "flos": 19618867970400.0, + "grad_norm": 2.033062405072683, + "learning_rate": 1.1684729848227636e-07, + "loss": 0.7817, + "num_input_tokens_seen": 321985615, + "step": 9912 + }, + { + "epoch": 0.893989268160707, + "flos": 26680896577440.0, + "grad_norm": 2.367805727560954, + "learning_rate": 1.1665061691518884e-07, + "loss": 0.7976, + "num_input_tokens_seen": 322014020, + "step": 9913 + }, + { + "epoch": 0.8940794516841774, + "flos": 23254495179360.0, + "grad_norm": 2.167013421799129, + "learning_rate": 1.1645409604585532e-07, + "loss": 0.7409, + "num_input_tokens_seen": 322042165, + "step": 9914 + }, + { + "epoch": 0.8941696352076476, + "flos": 23149696425120.0, + "grad_norm": 1.8031852381542874, + "learning_rate": 1.162577358910437e-07, + "loss": 0.7496, + "num_input_tokens_seen": 322070795, + "step": 9915 + }, + { + "epoch": 0.8942598187311178, + "flos": 30001644316800.0, + "grad_norm": 2.120309140140174, + "learning_rate": 1.160615364675095e-07, + "loss": 0.723, + "num_input_tokens_seen": 322100215, + "step": 9916 + }, + { + "epoch": 0.894350002254588, + "flos": 65824352806080.0, + "grad_norm": 0.6099896387877467, + "learning_rate": 1.1586549779199262e-07, + "loss": 0.5583, + "num_input_tokens_seen": 322190040, + "step": 9917 + }, + { + "epoch": 0.8944401857780584, + "flos": 21731000433600.0, + "grad_norm": 1.6455330957117336, + "learning_rate": 1.1566961988122037e-07, + "loss": 0.799, + "num_input_tokens_seen": 322220735, + "step": 9918 + }, + { + "epoch": 0.8945303693015286, + "flos": 21546609756000.0, + "grad_norm": 2.0975617968216325, + "learning_rate": 1.1547390275190627e-07, + "loss": 0.705, + "num_input_tokens_seen": 322249295, + "step": 9919 + }, + { + "epoch": 0.8946205528249989, + "flos": 12768964415520.0, + "grad_norm": 1.762514260416981, + "learning_rate": 1.1527834642075007e-07, + "loss": 0.6802, + "num_input_tokens_seen": 322275880, + "step": 9920 + }, + { + "epoch": 0.8947107363484691, + "flos": 27450988561440.0, + "grad_norm": 2.4365516186997516, + "learning_rate": 1.1508295090443797e-07, + "loss": 0.7241, + "num_input_tokens_seen": 322304005, + "step": 9921 + }, + { + "epoch": 0.8948009198719394, + "flos": 24642408876000.0, + "grad_norm": 2.0764398841450387, + "learning_rate": 1.148877162196411e-07, + "loss": 0.6914, + "num_input_tokens_seen": 322334390, + "step": 9922 + }, + { + "epoch": 0.8948911033954097, + "flos": 24680959650720.0, + "grad_norm": 2.4276852714010397, + "learning_rate": 1.1469264238301924e-07, + "loss": 0.6705, + "num_input_tokens_seen": 322358275, + "step": 9923 + }, + { + "epoch": 0.8949812869188799, + "flos": 16667459132640.0, + "grad_norm": 1.9488052204718298, + "learning_rate": 1.1449772941121638e-07, + "loss": 0.7667, + "num_input_tokens_seen": 322385830, + "step": 9924 + }, + { + "epoch": 0.8950714704423501, + "flos": 17068323724320.0, + "grad_norm": 11.741559832276199, + "learning_rate": 1.1430297732086369e-07, + "loss": 0.7102, + "num_input_tokens_seen": 322413725, + "step": 9925 + }, + { + "epoch": 0.8951616539658205, + "flos": 28180150906080.0, + "grad_norm": 1.7489854051298364, + "learning_rate": 1.1410838612857876e-07, + "loss": 0.7305, + "num_input_tokens_seen": 322445130, + "step": 9926 + }, + { + "epoch": 0.8952518374892907, + "flos": 24208866331680.0, + "grad_norm": 2.9933050474035423, + "learning_rate": 1.1391395585096497e-07, + "loss": 0.8403, + "num_input_tokens_seen": 322472290, + "step": 9927 + }, + { + "epoch": 0.895342021012761, + "flos": 29740895484960.0, + "grad_norm": 1.8099149487993094, + "learning_rate": 1.1371968650461216e-07, + "loss": 0.6691, + "num_input_tokens_seen": 322503350, + "step": 9928 + }, + { + "epoch": 0.8954322045362312, + "flos": 66492662520000.0, + "grad_norm": 0.691155555233186, + "learning_rate": 1.1352557810609687e-07, + "loss": 0.5719, + "num_input_tokens_seen": 322593240, + "step": 9929 + }, + { + "epoch": 0.8955223880597015, + "flos": 22970063476800.0, + "grad_norm": 2.806452713677258, + "learning_rate": 1.1333163067198048e-07, + "loss": 0.7245, + "num_input_tokens_seen": 322620185, + "step": 9930 + }, + { + "epoch": 0.8956125715831718, + "flos": 25626335721120.0, + "grad_norm": 1.57893011121354, + "learning_rate": 1.1313784421881311e-07, + "loss": 0.8037, + "num_input_tokens_seen": 322647600, + "step": 9931 + }, + { + "epoch": 0.895702755106642, + "flos": 20346766543200.0, + "grad_norm": 1.7985917224927286, + "learning_rate": 1.1294421876312865e-07, + "loss": 0.7763, + "num_input_tokens_seen": 322675460, + "step": 9932 + }, + { + "epoch": 0.8957929386301122, + "flos": 24172211214720.0, + "grad_norm": 2.2903955113048853, + "learning_rate": 1.1275075432144831e-07, + "loss": 0.7168, + "num_input_tokens_seen": 322702050, + "step": 9933 + }, + { + "epoch": 0.8958831221535826, + "flos": 22241904715680.0, + "grad_norm": 2.5320083989068856, + "learning_rate": 1.1255745091028002e-07, + "loss": 0.7535, + "num_input_tokens_seen": 322728315, + "step": 9934 + }, + { + "epoch": 0.8959733056770528, + "flos": 24536718047520.0, + "grad_norm": 2.1497699836292785, + "learning_rate": 1.1236430854611723e-07, + "loss": 0.7512, + "num_input_tokens_seen": 322755600, + "step": 9935 + }, + { + "epoch": 0.896063489200523, + "flos": 26905138971360.0, + "grad_norm": 2.3244769297583403, + "learning_rate": 1.1217132724544032e-07, + "loss": 0.7043, + "num_input_tokens_seen": 322783075, + "step": 9936 + }, + { + "epoch": 0.8961536727239934, + "flos": 16084471218720.0, + "grad_norm": 2.3922097930766433, + "learning_rate": 1.1197850702471434e-07, + "loss": 0.7893, + "num_input_tokens_seen": 322808670, + "step": 9937 + }, + { + "epoch": 0.8962438562474636, + "flos": 23735137543200.0, + "grad_norm": 1.8964193158132963, + "learning_rate": 1.1178584790039348e-07, + "loss": 0.8059, + "num_input_tokens_seen": 322837160, + "step": 9938 + }, + { + "epoch": 0.8963340397709338, + "flos": 27633409241760.0, + "grad_norm": 2.02902463641751, + "learning_rate": 1.1159334988891478e-07, + "loss": 0.6844, + "num_input_tokens_seen": 322866030, + "step": 9939 + }, + { + "epoch": 0.8964242232944041, + "flos": 24864606933120.0, + "grad_norm": 2.1569327779949616, + "learning_rate": 1.1140101300670446e-07, + "loss": 0.719, + "num_input_tokens_seen": 322896440, + "step": 9940 + }, + { + "epoch": 0.8965144068178744, + "flos": 22387187072160.0, + "grad_norm": 1.7610994252693981, + "learning_rate": 1.1120883727017338e-07, + "loss": 0.6076, + "num_input_tokens_seen": 322922700, + "step": 9941 + }, + { + "epoch": 0.8966045903413447, + "flos": 20529075714240.0, + "grad_norm": 2.4907608014348948, + "learning_rate": 1.1101682269571889e-07, + "loss": 0.794, + "num_input_tokens_seen": 322951000, + "step": 9942 + }, + { + "epoch": 0.8966947738648149, + "flos": 25226400373440.0, + "grad_norm": 2.3393836484520394, + "learning_rate": 1.1082496929972473e-07, + "loss": 0.7192, + "num_input_tokens_seen": 322980255, + "step": 9943 + }, + { + "epoch": 0.8967849573882851, + "flos": 26905287650400.0, + "grad_norm": 2.0344615490107145, + "learning_rate": 1.1063327709856096e-07, + "loss": 0.7092, + "num_input_tokens_seen": 323010675, + "step": 9944 + }, + { + "epoch": 0.8968751409117555, + "flos": 39109334388960.0, + "grad_norm": 2.0991912283475718, + "learning_rate": 1.1044174610858403e-07, + "loss": 0.6247, + "num_input_tokens_seen": 323038880, + "step": 9945 + }, + { + "epoch": 0.8969653244352257, + "flos": 25301346076800.0, + "grad_norm": 2.024939020439634, + "learning_rate": 1.1025037634613643e-07, + "loss": 0.6962, + "num_input_tokens_seen": 323067705, + "step": 9946 + }, + { + "epoch": 0.8970555079586959, + "flos": 24172471403040.0, + "grad_norm": 2.0008239241059314, + "learning_rate": 1.1005916782754643e-07, + "loss": 0.7665, + "num_input_tokens_seen": 323098610, + "step": 9947 + }, + { + "epoch": 0.8971456914821662, + "flos": 22714741429920.0, + "grad_norm": 2.033163020813053, + "learning_rate": 1.0986812056912919e-07, + "loss": 0.7265, + "num_input_tokens_seen": 323125730, + "step": 9948 + }, + { + "epoch": 0.8972358750056365, + "flos": 33893931684000.0, + "grad_norm": 2.284559488217871, + "learning_rate": 1.0967723458718613e-07, + "loss": 0.62, + "num_input_tokens_seen": 323155440, + "step": 9949 + }, + { + "epoch": 0.8973260585291067, + "flos": 20374129220160.0, + "grad_norm": 2.6636584022987937, + "learning_rate": 1.0948650989800445e-07, + "loss": 0.7082, + "num_input_tokens_seen": 323176980, + "step": 9950 + }, + { + "epoch": 0.897416242052577, + "flos": 25336923270720.0, + "grad_norm": 1.5962430117699848, + "learning_rate": 1.0929594651785823e-07, + "loss": 0.7205, + "num_input_tokens_seen": 323208325, + "step": 9951 + }, + { + "epoch": 0.8975064255760472, + "flos": 23037203530560.0, + "grad_norm": 1.8774290245649907, + "learning_rate": 1.0910554446300646e-07, + "loss": 0.7917, + "num_input_tokens_seen": 323236355, + "step": 9952 + }, + { + "epoch": 0.8975966090995176, + "flos": 24971933231040.0, + "grad_norm": 2.0266280187928905, + "learning_rate": 1.089153037496966e-07, + "loss": 0.638, + "num_input_tokens_seen": 323266260, + "step": 9953 + }, + { + "epoch": 0.8976867926229878, + "flos": 14225950993440.0, + "grad_norm": 4.137939503540492, + "learning_rate": 1.0872522439415966e-07, + "loss": 0.7603, + "num_input_tokens_seen": 323290780, + "step": 9954 + }, + { + "epoch": 0.897776976146458, + "flos": 20125237541760.0, + "grad_norm": 2.242529440934284, + "learning_rate": 1.0853530641261554e-07, + "loss": 0.6874, + "num_input_tokens_seen": 323318275, + "step": 9955 + }, + { + "epoch": 0.8978671596699282, + "flos": 22569347564160.0, + "grad_norm": 1.6803585397643008, + "learning_rate": 1.083455498212682e-07, + "loss": 0.7521, + "num_input_tokens_seen": 323346415, + "step": 9956 + }, + { + "epoch": 0.8979573431933986, + "flos": 24609879602400.0, + "grad_norm": 2.010353588109528, + "learning_rate": 1.0815595463630911e-07, + "loss": 0.777, + "num_input_tokens_seen": 323373700, + "step": 9957 + }, + { + "epoch": 0.8980475267168688, + "flos": 45157620269760.0, + "grad_norm": 2.1382934792962995, + "learning_rate": 1.0796652087391556e-07, + "loss": 0.6108, + "num_input_tokens_seen": 323405525, + "step": 9958 + }, + { + "epoch": 0.8981377102403391, + "flos": 27958324546560.0, + "grad_norm": 1.6694061732264756, + "learning_rate": 1.0777724855025083e-07, + "loss": 0.8347, + "num_input_tokens_seen": 323436945, + "step": 9959 + }, + { + "epoch": 0.8982278937638094, + "flos": 24937545469440.0, + "grad_norm": 1.7977337982546824, + "learning_rate": 1.075881376814649e-07, + "loss": 0.7401, + "num_input_tokens_seen": 323465375, + "step": 9960 + }, + { + "epoch": 0.8983180772872796, + "flos": 21549546167040.0, + "grad_norm": 1.906464486484424, + "learning_rate": 1.0739918828369377e-07, + "loss": 0.7956, + "num_input_tokens_seen": 323493870, + "step": 9961 + }, + { + "epoch": 0.8984082608107499, + "flos": 19691546318400.0, + "grad_norm": 2.3350705990951943, + "learning_rate": 1.0721040037305983e-07, + "loss": 0.7472, + "num_input_tokens_seen": 323521380, + "step": 9962 + }, + { + "epoch": 0.8984984443342201, + "flos": 18233370308160.0, + "grad_norm": 1.809102788026165, + "learning_rate": 1.0702177396567114e-07, + "loss": 0.6881, + "num_input_tokens_seen": 323548475, + "step": 9963 + }, + { + "epoch": 0.8985886278576904, + "flos": 69710693011680.0, + "grad_norm": 0.6221573318149374, + "learning_rate": 1.0683330907762233e-07, + "loss": 0.5618, + "num_input_tokens_seen": 323642675, + "step": 9964 + }, + { + "epoch": 0.8986788113811607, + "flos": 41186781732480.0, + "grad_norm": 1.7362210870454247, + "learning_rate": 1.0664500572499435e-07, + "loss": 0.6689, + "num_input_tokens_seen": 323674355, + "step": 9965 + }, + { + "epoch": 0.8987689949046309, + "flos": 33860956373280.0, + "grad_norm": 2.257877546911192, + "learning_rate": 1.0645686392385455e-07, + "loss": 0.6802, + "num_input_tokens_seen": 323704160, + "step": 9966 + }, + { + "epoch": 0.8988591784281011, + "flos": 27488535752640.0, + "grad_norm": 1.8283623435339422, + "learning_rate": 1.0626888369025588e-07, + "loss": 0.6699, + "num_input_tokens_seen": 323736940, + "step": 9967 + }, + { + "epoch": 0.8989493619515715, + "flos": 24534636540960.0, + "grad_norm": 1.7663777864939143, + "learning_rate": 1.0608106504023817e-07, + "loss": 0.7699, + "num_input_tokens_seen": 323767310, + "step": 9968 + }, + { + "epoch": 0.8990395454750417, + "flos": 22750541642400.0, + "grad_norm": 1.6556877045441383, + "learning_rate": 1.0589340798982637e-07, + "loss": 0.727, + "num_input_tokens_seen": 323797480, + "step": 9969 + }, + { + "epoch": 0.899129728998512, + "flos": 21877286373600.0, + "grad_norm": 1.7905284089937519, + "learning_rate": 1.057059125550337e-07, + "loss": 0.6816, + "num_input_tokens_seen": 323827415, + "step": 9970 + }, + { + "epoch": 0.8992199125219822, + "flos": 26353230710400.0, + "grad_norm": 1.7356134752626535, + "learning_rate": 1.0551857875185732e-07, + "loss": 0.7512, + "num_input_tokens_seen": 323856900, + "step": 9971 + }, + { + "epoch": 0.8993100960454525, + "flos": 25447223149440.0, + "grad_norm": 1.6816415465526293, + "learning_rate": 1.0533140659628181e-07, + "loss": 0.6272, + "num_input_tokens_seen": 323888565, + "step": 9972 + }, + { + "epoch": 0.8994002795689228, + "flos": 21688286645760.0, + "grad_norm": 2.503986137175043, + "learning_rate": 1.0514439610427772e-07, + "loss": 0.7968, + "num_input_tokens_seen": 323915890, + "step": 9973 + }, + { + "epoch": 0.899490463092393, + "flos": 21258386737920.0, + "grad_norm": 2.1416433098059926, + "learning_rate": 1.0495754729180206e-07, + "loss": 0.7677, + "num_input_tokens_seen": 323944770, + "step": 9974 + }, + { + "epoch": 0.8995806466158632, + "flos": 19250606991840.0, + "grad_norm": 1.94405903705386, + "learning_rate": 1.0477086017479741e-07, + "loss": 0.8381, + "num_input_tokens_seen": 323974625, + "step": 9975 + }, + { + "epoch": 0.8996708301393336, + "flos": 18343893205440.0, + "grad_norm": 2.721830617759614, + "learning_rate": 1.0458433476919327e-07, + "loss": 0.6969, + "num_input_tokens_seen": 324001850, + "step": 9976 + }, + { + "epoch": 0.8997610136628038, + "flos": 26758072466400.0, + "grad_norm": 1.4925359349102088, + "learning_rate": 1.0439797109090509e-07, + "loss": 0.768, + "num_input_tokens_seen": 324033640, + "step": 9977 + }, + { + "epoch": 0.899851197186274, + "flos": 22059744223680.0, + "grad_norm": 2.385729096010182, + "learning_rate": 1.0421176915583396e-07, + "loss": 0.7116, + "num_input_tokens_seen": 324061695, + "step": 9978 + }, + { + "epoch": 0.8999413807097443, + "flos": 19361724605280.0, + "grad_norm": 2.1015657899965534, + "learning_rate": 1.0402572897986828e-07, + "loss": 0.7696, + "num_input_tokens_seen": 324089055, + "step": 9979 + }, + { + "epoch": 0.9000315642332146, + "flos": 65095710838080.0, + "grad_norm": 0.5852510372952765, + "learning_rate": 1.0383985057888134e-07, + "loss": 0.5673, + "num_input_tokens_seen": 324193250, + "step": 9980 + }, + { + "epoch": 0.9001217477566849, + "flos": 33168114617760.0, + "grad_norm": 2.0360136074967943, + "learning_rate": 1.036541339687338e-07, + "loss": 0.6436, + "num_input_tokens_seen": 324223265, + "step": 9981 + }, + { + "epoch": 0.9002119312801551, + "flos": 70325578313280.0, + "grad_norm": 0.6406340714719528, + "learning_rate": 1.0346857916527186e-07, + "loss": 0.5373, + "num_input_tokens_seen": 324313750, + "step": 9982 + }, + { + "epoch": 0.9003021148036254, + "flos": 16959436296480.0, + "grad_norm": 1.9299168624167349, + "learning_rate": 1.0328318618432819e-07, + "loss": 0.7356, + "num_input_tokens_seen": 324342500, + "step": 9983 + }, + { + "epoch": 0.9003922983270957, + "flos": 26175530589600.0, + "grad_norm": 2.05647020639197, + "learning_rate": 1.0309795504172148e-07, + "loss": 0.7993, + "num_input_tokens_seen": 324370595, + "step": 9984 + }, + { + "epoch": 0.9004824818505659, + "flos": 15902533745280.0, + "grad_norm": 2.064856714445384, + "learning_rate": 1.0291288575325685e-07, + "loss": 0.7153, + "num_input_tokens_seen": 324397910, + "step": 9985 + }, + { + "epoch": 0.9005726653740361, + "flos": 31022114769600.0, + "grad_norm": 2.003133345897505, + "learning_rate": 1.0272797833472502e-07, + "loss": 0.6994, + "num_input_tokens_seen": 324429265, + "step": 9986 + }, + { + "epoch": 0.9006628488975065, + "flos": 70342899421440.0, + "grad_norm": 0.6835972159161579, + "learning_rate": 1.0254323280190335e-07, + "loss": 0.5469, + "num_input_tokens_seen": 324522280, + "step": 9987 + }, + { + "epoch": 0.9007530324209767, + "flos": 17869904228640.0, + "grad_norm": 2.6042215773833237, + "learning_rate": 1.023586491705557e-07, + "loss": 0.7143, + "num_input_tokens_seen": 324549070, + "step": 9988 + }, + { + "epoch": 0.9008432159444469, + "flos": 22707865024320.0, + "grad_norm": 2.298941895449636, + "learning_rate": 1.0217422745643145e-07, + "loss": 0.6746, + "num_input_tokens_seen": 324578175, + "step": 9989 + }, + { + "epoch": 0.9009333994679172, + "flos": 22278411153600.0, + "grad_norm": 1.7189459949063381, + "learning_rate": 1.0198996767526691e-07, + "loss": 0.6613, + "num_input_tokens_seen": 324607105, + "step": 9990 + }, + { + "epoch": 0.9010235829913875, + "flos": 37907260990560.0, + "grad_norm": 3.1668875433201182, + "learning_rate": 1.018058698427835e-07, + "loss": 0.7209, + "num_input_tokens_seen": 324637060, + "step": 9991 + }, + { + "epoch": 0.9011137665148578, + "flos": 23479666817280.0, + "grad_norm": 1.7346966419449814, + "learning_rate": 1.0162193397469021e-07, + "loss": 0.7875, + "num_input_tokens_seen": 324666330, + "step": 9992 + }, + { + "epoch": 0.901203950038328, + "flos": 24935649811680.0, + "grad_norm": 2.0046958631785556, + "learning_rate": 1.0143816008668049e-07, + "loss": 0.6757, + "num_input_tokens_seen": 324692980, + "step": 9993 + }, + { + "epoch": 0.9012941335617982, + "flos": 60182506980960.0, + "grad_norm": 0.6366522818666257, + "learning_rate": 1.0125454819443624e-07, + "loss": 0.5853, + "num_input_tokens_seen": 324785610, + "step": 9994 + }, + { + "epoch": 0.9013843170852686, + "flos": 40564127951040.0, + "grad_norm": 1.7521043795101388, + "learning_rate": 1.0107109831362315e-07, + "loss": 0.6442, + "num_input_tokens_seen": 324819570, + "step": 9995 + }, + { + "epoch": 0.9014745006087388, + "flos": 26393751482400.0, + "grad_norm": 1.823673338295255, + "learning_rate": 1.0088781045989447e-07, + "loss": 0.7457, + "num_input_tokens_seen": 324850015, + "step": 9996 + }, + { + "epoch": 0.901564684132209, + "flos": 19287819655200.0, + "grad_norm": 2.073664416937974, + "learning_rate": 1.0070468464888926e-07, + "loss": 0.7882, + "num_input_tokens_seen": 324876650, + "step": 9997 + }, + { + "epoch": 0.9016548676556793, + "flos": 25587599097600.0, + "grad_norm": 2.0082082395519993, + "learning_rate": 1.0052172089623324e-07, + "loss": 0.7565, + "num_input_tokens_seen": 324904355, + "step": 9998 + }, + { + "epoch": 0.9017450511791496, + "flos": 23845028554560.0, + "grad_norm": 2.3259332568104183, + "learning_rate": 1.0033891921753746e-07, + "loss": 0.7782, + "num_input_tokens_seen": 324935105, + "step": 9999 + }, + { + "epoch": 0.9018352347026198, + "flos": 13242581694720.0, + "grad_norm": 1.903720490695931, + "learning_rate": 1.0015627962839968e-07, + "loss": 0.6466, + "num_input_tokens_seen": 324961375, + "step": 10000 + }, + { + "epoch": 0.9019254182260901, + "flos": 26571674621760.0, + "grad_norm": 2.3799443227201866, + "learning_rate": 9.99738021444041e-08, + "loss": 0.7125, + "num_input_tokens_seen": 324990770, + "step": 10001 + }, + { + "epoch": 0.9020156017495603, + "flos": 32916286528320.0, + "grad_norm": 1.7101264320168454, + "learning_rate": 9.979148678112003e-08, + "loss": 0.6673, + "num_input_tokens_seen": 325022040, + "step": 10002 + }, + { + "epoch": 0.9021057852730306, + "flos": 16083653484000.0, + "grad_norm": 2.453382733649309, + "learning_rate": 9.960933355410417e-08, + "loss": 0.8368, + "num_input_tokens_seen": 325046800, + "step": 10003 + }, + { + "epoch": 0.9021959687965009, + "flos": 18888181665600.0, + "grad_norm": 4.4049743537286155, + "learning_rate": 9.942734247889828e-08, + "loss": 0.6526, + "num_input_tokens_seen": 325075770, + "step": 10004 + }, + { + "epoch": 0.9022861523199711, + "flos": 19909953060000.0, + "grad_norm": 1.742145554931407, + "learning_rate": 9.92455135710315e-08, + "loss": 0.7423, + "num_input_tokens_seen": 325103715, + "step": 10005 + }, + { + "epoch": 0.9023763358434415, + "flos": 25411051239360.0, + "grad_norm": 2.5193624277257785, + "learning_rate": 9.906384684601787e-08, + "loss": 0.6694, + "num_input_tokens_seen": 325133285, + "step": 10006 + }, + { + "epoch": 0.9024665193669117, + "flos": 23621826913920.0, + "grad_norm": 1.8216312199547064, + "learning_rate": 9.8882342319359e-08, + "loss": 0.7842, + "num_input_tokens_seen": 325162835, + "step": 10007 + }, + { + "epoch": 0.9025567028903819, + "flos": 28033047231360.0, + "grad_norm": 1.6742854344931768, + "learning_rate": 9.870100000654048e-08, + "loss": 0.6876, + "num_input_tokens_seen": 325194685, + "step": 10008 + }, + { + "epoch": 0.9026468864138522, + "flos": 28031709120000.0, + "grad_norm": 1.5297763506419486, + "learning_rate": 9.851981992303704e-08, + "loss": 0.7083, + "num_input_tokens_seen": 325225120, + "step": 10009 + }, + { + "epoch": 0.9027370699373225, + "flos": 25336440063840.0, + "grad_norm": 1.5900062271811943, + "learning_rate": 9.833880208430678e-08, + "loss": 0.7396, + "num_input_tokens_seen": 325256725, + "step": 10010 + }, + { + "epoch": 0.9028272534607927, + "flos": 22023423634560.0, + "grad_norm": 2.7311625765638343, + "learning_rate": 9.815794650579601e-08, + "loss": 0.7306, + "num_input_tokens_seen": 325284320, + "step": 10011 + }, + { + "epoch": 0.902917436984263, + "flos": 21579956764320.0, + "grad_norm": 2.7170610114342217, + "learning_rate": 9.797725320293548e-08, + "loss": 0.7153, + "num_input_tokens_seen": 325311750, + "step": 10012 + }, + { + "epoch": 0.9030076205077332, + "flos": 19982891596320.0, + "grad_norm": 2.296829949800015, + "learning_rate": 9.779672219114354e-08, + "loss": 0.8148, + "num_input_tokens_seen": 325335970, + "step": 10013 + }, + { + "epoch": 0.9030978040312035, + "flos": 31200929983200.0, + "grad_norm": 1.980840584122875, + "learning_rate": 9.761635348582386e-08, + "loss": 0.715, + "num_input_tokens_seen": 325369825, + "step": 10014 + }, + { + "epoch": 0.9031879875546738, + "flos": 18015558282720.0, + "grad_norm": 1.8776815947829704, + "learning_rate": 9.743614710236658e-08, + "loss": 0.7676, + "num_input_tokens_seen": 325397540, + "step": 10015 + }, + { + "epoch": 0.903278171078144, + "flos": 15064558312320.0, + "grad_norm": 1.832192289842919, + "learning_rate": 9.725610305614806e-08, + "loss": 0.6994, + "num_input_tokens_seen": 325425110, + "step": 10016 + }, + { + "epoch": 0.9033683546016142, + "flos": 35064888259680.0, + "grad_norm": 2.4108045006074064, + "learning_rate": 9.707622136253002e-08, + "loss": 0.7324, + "num_input_tokens_seen": 325456410, + "step": 10017 + }, + { + "epoch": 0.9034585381250846, + "flos": 23261483094240.0, + "grad_norm": 1.7803214480506835, + "learning_rate": 9.689650203686173e-08, + "loss": 0.7601, + "num_input_tokens_seen": 325485055, + "step": 10018 + }, + { + "epoch": 0.9035487216485548, + "flos": 22496706385920.0, + "grad_norm": 2.184626845002207, + "learning_rate": 9.671694509447715e-08, + "loss": 0.7118, + "num_input_tokens_seen": 325514215, + "step": 10019 + }, + { + "epoch": 0.903638905172025, + "flos": 25039519321920.0, + "grad_norm": 3.1558077662899366, + "learning_rate": 9.653755055069757e-08, + "loss": 0.7513, + "num_input_tokens_seen": 325542000, + "step": 10020 + }, + { + "epoch": 0.9037290886954953, + "flos": 39514399163520.0, + "grad_norm": 1.432368032558176, + "learning_rate": 9.635831842082987e-08, + "loss": 0.7146, + "num_input_tokens_seen": 325577720, + "step": 10021 + }, + { + "epoch": 0.9038192722189656, + "flos": 23625952757280.0, + "grad_norm": 1.8074639830951424, + "learning_rate": 9.617924872016691e-08, + "loss": 0.755, + "num_input_tokens_seen": 325603400, + "step": 10022 + }, + { + "epoch": 0.9039094557424359, + "flos": 21767432532000.0, + "grad_norm": 1.9263202152908998, + "learning_rate": 9.600034146398806e-08, + "loss": 0.7753, + "num_input_tokens_seen": 325632105, + "step": 10023 + }, + { + "epoch": 0.9039996392659061, + "flos": 71615606831040.0, + "grad_norm": 0.696580352417519, + "learning_rate": 9.582159666755863e-08, + "loss": 0.5799, + "num_input_tokens_seen": 325733400, + "step": 10024 + }, + { + "epoch": 0.9040898227893763, + "flos": 33059115680640.0, + "grad_norm": 1.6178075366563947, + "learning_rate": 9.564301434612976e-08, + "loss": 0.7078, + "num_input_tokens_seen": 325761750, + "step": 10025 + }, + { + "epoch": 0.9041800063128467, + "flos": 20050812215040.0, + "grad_norm": 1.7272505964111677, + "learning_rate": 9.546459451494015e-08, + "loss": 0.7794, + "num_input_tokens_seen": 325785560, + "step": 10026 + }, + { + "epoch": 0.9042701898363169, + "flos": 67044124743840.0, + "grad_norm": 0.6320357323712757, + "learning_rate": 9.528633718921231e-08, + "loss": 0.6046, + "num_input_tokens_seen": 325872395, + "step": 10027 + }, + { + "epoch": 0.9043603733597871, + "flos": 24024921691200.0, + "grad_norm": 1.9161085131543931, + "learning_rate": 9.510824238415672e-08, + "loss": 0.7863, + "num_input_tokens_seen": 325901845, + "step": 10028 + }, + { + "epoch": 0.9044505568832575, + "flos": 26974286192160.0, + "grad_norm": 1.8311658169401943, + "learning_rate": 9.493031011496944e-08, + "loss": 0.7706, + "num_input_tokens_seen": 325931770, + "step": 10029 + }, + { + "epoch": 0.9045407404067277, + "flos": 23953730133600.0, + "grad_norm": 15.978158825354134, + "learning_rate": 9.475254039683234e-08, + "loss": 0.5934, + "num_input_tokens_seen": 325960420, + "step": 10030 + }, + { + "epoch": 0.904630923930198, + "flos": 28508634507840.0, + "grad_norm": 1.6756196199734126, + "learning_rate": 9.45749332449144e-08, + "loss": 0.7018, + "num_input_tokens_seen": 325992380, + "step": 10031 + }, + { + "epoch": 0.9047211074536682, + "flos": 62354828752800.0, + "grad_norm": 0.6431258397889285, + "learning_rate": 9.439748867436903e-08, + "loss": 0.5886, + "num_input_tokens_seen": 326082890, + "step": 10032 + }, + { + "epoch": 0.9048112909771385, + "flos": 20783096819520.0, + "grad_norm": 1.7280800480713023, + "learning_rate": 9.42202067003377e-08, + "loss": 0.698, + "num_input_tokens_seen": 326112385, + "step": 10033 + }, + { + "epoch": 0.9049014745006088, + "flos": 21694940032800.0, + "grad_norm": 1.963276414299811, + "learning_rate": 9.404308733794652e-08, + "loss": 0.8746, + "num_input_tokens_seen": 326137730, + "step": 10034 + }, + { + "epoch": 0.904991658024079, + "flos": 24606051117120.0, + "grad_norm": 1.8297490787975474, + "learning_rate": 9.38661306023083e-08, + "loss": 0.7558, + "num_input_tokens_seen": 326168145, + "step": 10035 + }, + { + "epoch": 0.9050818415475492, + "flos": 18525533320800.0, + "grad_norm": 3.3007917957024584, + "learning_rate": 9.368933650852229e-08, + "loss": 0.729, + "num_input_tokens_seen": 326194605, + "step": 10036 + }, + { + "epoch": 0.9051720250710196, + "flos": 66296148767040.0, + "grad_norm": 0.6902151030617133, + "learning_rate": 9.351270507167352e-08, + "loss": 0.5928, + "num_input_tokens_seen": 326290925, + "step": 10037 + }, + { + "epoch": 0.9052622085944898, + "flos": 37579520784000.0, + "grad_norm": 2.089174309579477, + "learning_rate": 9.333623630683285e-08, + "loss": 0.672, + "num_input_tokens_seen": 326321615, + "step": 10038 + }, + { + "epoch": 0.90535239211796, + "flos": 27919810941600.0, + "grad_norm": 2.381486667676629, + "learning_rate": 9.315993022905799e-08, + "loss": 0.588, + "num_input_tokens_seen": 326350495, + "step": 10039 + }, + { + "epoch": 0.9054425756414303, + "flos": 22168557312000.0, + "grad_norm": 1.914628237203606, + "learning_rate": 9.298378685339158e-08, + "loss": 0.7161, + "num_input_tokens_seen": 326377975, + "step": 10040 + }, + { + "epoch": 0.9055327591649006, + "flos": 22204357524480.0, + "grad_norm": 2.556824236775945, + "learning_rate": 9.280780619486406e-08, + "loss": 0.8014, + "num_input_tokens_seen": 326405785, + "step": 10041 + }, + { + "epoch": 0.9056229426883708, + "flos": 25229373954240.0, + "grad_norm": 1.9795771591399633, + "learning_rate": 9.26319882684905e-08, + "loss": 0.818, + "num_input_tokens_seen": 326433290, + "step": 10042 + }, + { + "epoch": 0.9057131262118411, + "flos": 32368652789760.0, + "grad_norm": 1.7814146266954782, + "learning_rate": 9.245633308927293e-08, + "loss": 0.7082, + "num_input_tokens_seen": 326464955, + "step": 10043 + }, + { + "epoch": 0.9058033097353113, + "flos": 29782568519520.0, + "grad_norm": 1.8830034467814498, + "learning_rate": 9.228084067219888e-08, + "loss": 0.7178, + "num_input_tokens_seen": 326495220, + "step": 10044 + }, + { + "epoch": 0.9058934932587817, + "flos": 22897496638080.0, + "grad_norm": 1.9394502361233834, + "learning_rate": 9.210551103224284e-08, + "loss": 0.6825, + "num_input_tokens_seen": 326523740, + "step": 10045 + }, + { + "epoch": 0.9059836767822519, + "flos": 25445513340480.0, + "grad_norm": 1.664312925882266, + "learning_rate": 9.193034418436463e-08, + "loss": 0.7889, + "num_input_tokens_seen": 326553810, + "step": 10046 + }, + { + "epoch": 0.9060738603057221, + "flos": 63983828478240.0, + "grad_norm": 0.653668936413597, + "learning_rate": 9.175534014351005e-08, + "loss": 0.6126, + "num_input_tokens_seen": 326649375, + "step": 10047 + }, + { + "epoch": 0.9061640438291924, + "flos": 19472098823520.0, + "grad_norm": 3.1161994611097823, + "learning_rate": 9.158049892461228e-08, + "loss": 0.6798, + "num_input_tokens_seen": 326675665, + "step": 10048 + }, + { + "epoch": 0.9062542273526627, + "flos": 44428606604160.0, + "grad_norm": 2.2881810701320586, + "learning_rate": 9.140582054258871e-08, + "loss": 0.6756, + "num_input_tokens_seen": 326709535, + "step": 10049 + }, + { + "epoch": 0.9063444108761329, + "flos": 33494145015360.0, + "grad_norm": 1.7135055986529457, + "learning_rate": 9.123130501234499e-08, + "loss": 0.7458, + "num_input_tokens_seen": 326741040, + "step": 10050 + }, + { + "epoch": 0.9064345943996032, + "flos": 22963335750240.0, + "grad_norm": 2.367228491903483, + "learning_rate": 9.105695234877098e-08, + "loss": 0.7857, + "num_input_tokens_seen": 326769540, + "step": 10051 + }, + { + "epoch": 0.9065247779230734, + "flos": 21291213369600.0, + "grad_norm": 3.444868063428013, + "learning_rate": 9.088276256674344e-08, + "loss": 0.7638, + "num_input_tokens_seen": 326798465, + "step": 10052 + }, + { + "epoch": 0.9066149614465437, + "flos": 23880903106560.0, + "grad_norm": 1.693093671709531, + "learning_rate": 9.070873568112536e-08, + "loss": 0.7234, + "num_input_tokens_seen": 326826600, + "step": 10053 + }, + { + "epoch": 0.906705144970014, + "flos": 25994039153280.0, + "grad_norm": 1.8704503492196403, + "learning_rate": 9.053487170676577e-08, + "loss": 0.7961, + "num_input_tokens_seen": 326857005, + "step": 10054 + }, + { + "epoch": 0.9067953284934842, + "flos": 20053562777280.0, + "grad_norm": 1.9338153750632279, + "learning_rate": 9.036117065849968e-08, + "loss": 0.7934, + "num_input_tokens_seen": 326885730, + "step": 10055 + }, + { + "epoch": 0.9068855120169546, + "flos": 24858213734400.0, + "grad_norm": 1.8453204691793152, + "learning_rate": 9.018763255114837e-08, + "loss": 0.721, + "num_input_tokens_seen": 326915275, + "step": 10056 + }, + { + "epoch": 0.9069756955404248, + "flos": 20199068152320.0, + "grad_norm": 1.7362274037724648, + "learning_rate": 9.00142573995184e-08, + "loss": 0.7055, + "num_input_tokens_seen": 326942455, + "step": 10057 + }, + { + "epoch": 0.907065879063895, + "flos": 32187533051040.0, + "grad_norm": 2.688701893169046, + "learning_rate": 8.984104521840375e-08, + "loss": 0.8164, + "num_input_tokens_seen": 326972680, + "step": 10058 + }, + { + "epoch": 0.9071560625873653, + "flos": 22569979450080.0, + "grad_norm": 2.1564846187602886, + "learning_rate": 8.966799602258346e-08, + "loss": 0.7582, + "num_input_tokens_seen": 326999890, + "step": 10059 + }, + { + "epoch": 0.9072462461108356, + "flos": 16704077079840.0, + "grad_norm": 1.896059658528373, + "learning_rate": 8.949510982682329e-08, + "loss": 0.7789, + "num_input_tokens_seen": 327026230, + "step": 10060 + }, + { + "epoch": 0.9073364296343058, + "flos": 32288168792160.0, + "grad_norm": 2.66898488567785, + "learning_rate": 8.932238664587499e-08, + "loss": 0.8055, + "num_input_tokens_seen": 327050915, + "step": 10061 + }, + { + "epoch": 0.9074266131577761, + "flos": 22707530496480.0, + "grad_norm": 2.0600872376867643, + "learning_rate": 8.914982649447567e-08, + "loss": 0.829, + "num_input_tokens_seen": 327077835, + "step": 10062 + }, + { + "epoch": 0.9075167966812463, + "flos": 22496632046400.0, + "grad_norm": 1.6739752581282197, + "learning_rate": 8.897742938734975e-08, + "loss": 0.8034, + "num_input_tokens_seen": 327106400, + "step": 10063 + }, + { + "epoch": 0.9076069802047166, + "flos": 21364932470880.0, + "grad_norm": 1.9134020853261908, + "learning_rate": 8.880519533920661e-08, + "loss": 0.7645, + "num_input_tokens_seen": 327134030, + "step": 10064 + }, + { + "epoch": 0.9076971637281869, + "flos": 23259327248160.0, + "grad_norm": 1.5442751991077315, + "learning_rate": 8.863312436474268e-08, + "loss": 0.7384, + "num_input_tokens_seen": 327166240, + "step": 10065 + }, + { + "epoch": 0.9077873472516571, + "flos": 18372408144960.0, + "grad_norm": 2.3365245068777334, + "learning_rate": 8.846121647863936e-08, + "loss": 0.7433, + "num_input_tokens_seen": 327193190, + "step": 10066 + }, + { + "epoch": 0.9078775307751273, + "flos": 25375176687360.0, + "grad_norm": 1.6369158277921962, + "learning_rate": 8.828947169556555e-08, + "loss": 0.7399, + "num_input_tokens_seen": 327221900, + "step": 10067 + }, + { + "epoch": 0.9079677142985977, + "flos": 24750887436480.0, + "grad_norm": 1.721385253379312, + "learning_rate": 8.81178900301749e-08, + "loss": 0.7067, + "num_input_tokens_seen": 327252060, + "step": 10068 + }, + { + "epoch": 0.9080578978220679, + "flos": 37792240552320.0, + "grad_norm": 1.593660838307107, + "learning_rate": 8.794647149710787e-08, + "loss": 0.6551, + "num_input_tokens_seen": 327283880, + "step": 10069 + }, + { + "epoch": 0.9081480813455381, + "flos": 24899589410880.0, + "grad_norm": 1.922105046592118, + "learning_rate": 8.777521611099081e-08, + "loss": 0.7304, + "num_input_tokens_seen": 327308650, + "step": 10070 + }, + { + "epoch": 0.9082382648690084, + "flos": 19290384368640.0, + "grad_norm": 1.8506541539001884, + "learning_rate": 8.760412388643624e-08, + "loss": 0.6771, + "num_input_tokens_seen": 327338245, + "step": 10071 + }, + { + "epoch": 0.9083284483924787, + "flos": 19870398701760.0, + "grad_norm": 4.22413009626721, + "learning_rate": 8.74331948380429e-08, + "loss": 0.7525, + "num_input_tokens_seen": 327365420, + "step": 10072 + }, + { + "epoch": 0.908418631915949, + "flos": 18051990381120.0, + "grad_norm": 1.8617420036602141, + "learning_rate": 8.726242898039516e-08, + "loss": 0.7387, + "num_input_tokens_seen": 327392520, + "step": 10073 + }, + { + "epoch": 0.9085088154394192, + "flos": 16521619229760.0, + "grad_norm": 1.9620861929257463, + "learning_rate": 8.709182632806334e-08, + "loss": 0.7472, + "num_input_tokens_seen": 327420180, + "step": 10074 + }, + { + "epoch": 0.9085989989628894, + "flos": 27377938515840.0, + "grad_norm": 2.060064322569215, + "learning_rate": 8.692138689560469e-08, + "loss": 0.7462, + "num_input_tokens_seen": 327448880, + "step": 10075 + }, + { + "epoch": 0.9086891824863598, + "flos": 23553869125440.0, + "grad_norm": 1.9973057242056922, + "learning_rate": 8.675111069756203e-08, + "loss": 0.719, + "num_input_tokens_seen": 327480835, + "step": 10076 + }, + { + "epoch": 0.90877936600983, + "flos": 36044837940480.0, + "grad_norm": 1.7392987991292894, + "learning_rate": 8.658099774846395e-08, + "loss": 0.7087, + "num_input_tokens_seen": 327512315, + "step": 10077 + }, + { + "epoch": 0.9088695495333002, + "flos": 32078199586080.0, + "grad_norm": 1.482937729333916, + "learning_rate": 8.641104806282595e-08, + "loss": 0.6816, + "num_input_tokens_seen": 327545885, + "step": 10078 + }, + { + "epoch": 0.9089597330567706, + "flos": 20856964599840.0, + "grad_norm": 2.088928935104483, + "learning_rate": 8.624126165514845e-08, + "loss": 0.7476, + "num_input_tokens_seen": 327573545, + "step": 10079 + }, + { + "epoch": 0.9090499165802408, + "flos": 29491446260160.0, + "grad_norm": 1.9762785334400208, + "learning_rate": 8.607163853991917e-08, + "loss": 0.7012, + "num_input_tokens_seen": 327603870, + "step": 10080 + }, + { + "epoch": 0.909140100103711, + "flos": 36960063601920.0, + "grad_norm": 1.9929674603896566, + "learning_rate": 8.590217873161054e-08, + "loss": 0.5849, + "num_input_tokens_seen": 327637385, + "step": 10081 + }, + { + "epoch": 0.9092302836271813, + "flos": 26937854093760.0, + "grad_norm": 1.7548854617702248, + "learning_rate": 8.573288224468255e-08, + "loss": 0.755, + "num_input_tokens_seen": 327666420, + "step": 10082 + }, + { + "epoch": 0.9093204671506516, + "flos": 20927672950560.0, + "grad_norm": 2.3408848861426095, + "learning_rate": 8.556374909358011e-08, + "loss": 0.6668, + "num_input_tokens_seen": 327696195, + "step": 10083 + }, + { + "epoch": 0.9094106506741219, + "flos": 53278558595040.0, + "grad_norm": 2.1092569953971556, + "learning_rate": 8.539477929273476e-08, + "loss": 0.716, + "num_input_tokens_seen": 327729975, + "step": 10084 + }, + { + "epoch": 0.9095008341975921, + "flos": 27592999979040.0, + "grad_norm": 1.7278640467267519, + "learning_rate": 8.522597285656386e-08, + "loss": 0.7745, + "num_input_tokens_seen": 327758455, + "step": 10085 + }, + { + "epoch": 0.9095910177210623, + "flos": 71581813785600.0, + "grad_norm": 3.19751721267899, + "learning_rate": 8.505732979947078e-08, + "loss": 0.5888, + "num_input_tokens_seen": 327843120, + "step": 10086 + }, + { + "epoch": 0.9096812012445327, + "flos": 22207070916960.0, + "grad_norm": 1.3283598299546773, + "learning_rate": 8.488885013584557e-08, + "loss": 0.7535, + "num_input_tokens_seen": 327873185, + "step": 10087 + }, + { + "epoch": 0.9097713847680029, + "flos": 24971598703200.0, + "grad_norm": 1.8126057561078182, + "learning_rate": 8.472053388006295e-08, + "loss": 0.6785, + "num_input_tokens_seen": 327904275, + "step": 10088 + }, + { + "epoch": 0.9098615682914731, + "flos": 23115791870400.0, + "grad_norm": 2.6802718968322012, + "learning_rate": 8.455238104648565e-08, + "loss": 0.8259, + "num_input_tokens_seen": 327932705, + "step": 10089 + }, + { + "epoch": 0.9099517518149434, + "flos": 47889470103360.0, + "grad_norm": 3.2928673851910704, + "learning_rate": 8.438439164946043e-08, + "loss": 0.6113, + "num_input_tokens_seen": 327965300, + "step": 10090 + }, + { + "epoch": 0.9100419353384137, + "flos": 24317382061920.0, + "grad_norm": 1.9556926481763077, + "learning_rate": 8.42165657033218e-08, + "loss": 0.7512, + "num_input_tokens_seen": 327995225, + "step": 10091 + }, + { + "epoch": 0.9101321188618839, + "flos": 24570882790560.0, + "grad_norm": 2.0501217317491744, + "learning_rate": 8.4048903222389e-08, + "loss": 0.7717, + "num_input_tokens_seen": 328023530, + "step": 10092 + }, + { + "epoch": 0.9102223023853542, + "flos": 24099123999360.0, + "grad_norm": 2.0703513087827785, + "learning_rate": 8.388140422096856e-08, + "loss": 0.7883, + "num_input_tokens_seen": 328052795, + "step": 10093 + }, + { + "epoch": 0.9103124859088244, + "flos": 24317939608320.0, + "grad_norm": 2.2083635747986143, + "learning_rate": 8.371406871335173e-08, + "loss": 0.751, + "num_input_tokens_seen": 328082375, + "step": 10094 + }, + { + "epoch": 0.9104026694322948, + "flos": 22093277080800.0, + "grad_norm": 1.886860591382322, + "learning_rate": 8.354689671381732e-08, + "loss": 0.7401, + "num_input_tokens_seen": 328112000, + "step": 10095 + }, + { + "epoch": 0.910492852955765, + "flos": 38052803535360.0, + "grad_norm": 1.638704392811856, + "learning_rate": 8.337988823662834e-08, + "loss": 0.7339, + "num_input_tokens_seen": 328145685, + "step": 10096 + }, + { + "epoch": 0.9105830364792352, + "flos": 19290607387200.0, + "grad_norm": 1.7264288941209507, + "learning_rate": 8.321304329603607e-08, + "loss": 0.6926, + "num_input_tokens_seen": 328172765, + "step": 10097 + }, + { + "epoch": 0.9106732200027055, + "flos": 70995146065440.0, + "grad_norm": 0.6275951152692261, + "learning_rate": 8.304636190627557e-08, + "loss": 0.586, + "num_input_tokens_seen": 328270340, + "step": 10098 + }, + { + "epoch": 0.9107634035261758, + "flos": 20273902346400.0, + "grad_norm": 2.7867743672391865, + "learning_rate": 8.287984408156945e-08, + "loss": 0.7536, + "num_input_tokens_seen": 328296230, + "step": 10099 + }, + { + "epoch": 0.910853587049646, + "flos": 23734877354880.0, + "grad_norm": 11.955047333167746, + "learning_rate": 8.271348983612591e-08, + "loss": 0.64, + "num_input_tokens_seen": 328323645, + "step": 10100 + }, + { + "epoch": 0.9109437705731163, + "flos": 32988184311360.0, + "grad_norm": 1.9688197532285998, + "learning_rate": 8.254729918413938e-08, + "loss": 0.6986, + "num_input_tokens_seen": 328356440, + "step": 10101 + }, + { + "epoch": 0.9110339540965866, + "flos": 21695311730400.0, + "grad_norm": 1.7599504143058775, + "learning_rate": 8.238127213979006e-08, + "loss": 0.7055, + "num_input_tokens_seen": 328385890, + "step": 10102 + }, + { + "epoch": 0.9111241376200568, + "flos": 32582599160160.0, + "grad_norm": 2.0423493184756567, + "learning_rate": 8.221540871724398e-08, + "loss": 0.81, + "num_input_tokens_seen": 328414370, + "step": 10103 + }, + { + "epoch": 0.9112143211435271, + "flos": 27414853821120.0, + "grad_norm": 1.7267537421040562, + "learning_rate": 8.2049708930654e-08, + "loss": 0.6257, + "num_input_tokens_seen": 328447020, + "step": 10104 + }, + { + "epoch": 0.9113045046669973, + "flos": 29891790475200.0, + "grad_norm": 2.2302887371262377, + "learning_rate": 8.188417279415793e-08, + "loss": 0.7024, + "num_input_tokens_seen": 328477685, + "step": 10105 + }, + { + "epoch": 0.9113946881904677, + "flos": 21731037603360.0, + "grad_norm": 2.224708508031834, + "learning_rate": 8.171880032188117e-08, + "loss": 0.8311, + "num_input_tokens_seen": 328504310, + "step": 10106 + }, + { + "epoch": 0.9114848717139379, + "flos": 24864792781920.0, + "grad_norm": 1.5724246874606096, + "learning_rate": 8.155359152793351e-08, + "loss": 0.7184, + "num_input_tokens_seen": 328534410, + "step": 10107 + }, + { + "epoch": 0.9115750552374081, + "flos": 18523488984000.0, + "grad_norm": 2.1299697879206723, + "learning_rate": 8.138854642641147e-08, + "loss": 0.7262, + "num_input_tokens_seen": 328564075, + "step": 10108 + }, + { + "epoch": 0.9116652387608783, + "flos": 34846370008800.0, + "grad_norm": 2.118851911856381, + "learning_rate": 8.122366503139777e-08, + "loss": 0.6637, + "num_input_tokens_seen": 328595330, + "step": 10109 + }, + { + "epoch": 0.9117554222843487, + "flos": 23875104624000.0, + "grad_norm": 13.447102650837035, + "learning_rate": 8.105894735696117e-08, + "loss": 0.6783, + "num_input_tokens_seen": 328625020, + "step": 10110 + }, + { + "epoch": 0.9118456058078189, + "flos": 20675027126400.0, + "grad_norm": 2.0543524755456444, + "learning_rate": 8.089439341715576e-08, + "loss": 0.719, + "num_input_tokens_seen": 328652365, + "step": 10111 + }, + { + "epoch": 0.9119357893312892, + "flos": 18087902102880.0, + "grad_norm": 2.0742123131373758, + "learning_rate": 8.073000322602319e-08, + "loss": 0.7691, + "num_input_tokens_seen": 328680240, + "step": 10112 + }, + { + "epoch": 0.9120259728547594, + "flos": 21836951450400.0, + "grad_norm": 1.6571881498291998, + "learning_rate": 8.056577679758891e-08, + "loss": 0.7529, + "num_input_tokens_seen": 328709700, + "step": 10113 + }, + { + "epoch": 0.9121161563782297, + "flos": 34373310276000.0, + "grad_norm": 2.0120090904713033, + "learning_rate": 8.040171414586638e-08, + "loss": 0.6835, + "num_input_tokens_seen": 328736670, + "step": 10114 + }, + { + "epoch": 0.9122063399017, + "flos": 27196595758560.0, + "grad_norm": 1.6306229654155422, + "learning_rate": 8.023781528485419e-08, + "loss": 0.7009, + "num_input_tokens_seen": 328767585, + "step": 10115 + }, + { + "epoch": 0.9122965234251702, + "flos": 21366270582240.0, + "grad_norm": 2.3721997668205796, + "learning_rate": 8.00740802285369e-08, + "loss": 0.7635, + "num_input_tokens_seen": 328795480, + "step": 10116 + }, + { + "epoch": 0.9123867069486404, + "flos": 24318497154720.0, + "grad_norm": 1.7164922971344283, + "learning_rate": 7.99105089908858e-08, + "loss": 0.6028, + "num_input_tokens_seen": 328823910, + "step": 10117 + }, + { + "epoch": 0.9124768904721108, + "flos": 22932033078720.0, + "grad_norm": 1.749794130477627, + "learning_rate": 7.974710158585685e-08, + "loss": 0.7104, + "num_input_tokens_seen": 328854125, + "step": 10118 + }, + { + "epoch": 0.912567073995581, + "flos": 23626175775840.0, + "grad_norm": 1.83496464053395, + "learning_rate": 7.958385802739375e-08, + "loss": 0.7629, + "num_input_tokens_seen": 328883020, + "step": 10119 + }, + { + "epoch": 0.9126572575190512, + "flos": 22206030163680.0, + "grad_norm": 1.860804673330223, + "learning_rate": 7.942077832942452e-08, + "loss": 0.799, + "num_input_tokens_seen": 328913150, + "step": 10120 + }, + { + "epoch": 0.9127474410425215, + "flos": 68367500269920.0, + "grad_norm": 0.7874255054298216, + "learning_rate": 7.925786250586508e-08, + "loss": 0.58, + "num_input_tokens_seen": 329004735, + "step": 10121 + }, + { + "epoch": 0.9128376245659918, + "flos": 62324604004320.0, + "grad_norm": 0.6527559725408916, + "learning_rate": 7.909511057061524e-08, + "loss": 0.5795, + "num_input_tokens_seen": 329089945, + "step": 10122 + }, + { + "epoch": 0.9129278080894621, + "flos": 20420299795680.0, + "grad_norm": 2.083277688872772, + "learning_rate": 7.893252253756234e-08, + "loss": 0.7257, + "num_input_tokens_seen": 329118680, + "step": 10123 + }, + { + "epoch": 0.9130179916129323, + "flos": 23989976383200.0, + "grad_norm": 2.1780319215871553, + "learning_rate": 7.877009842057925e-08, + "loss": 0.6654, + "num_input_tokens_seen": 329148240, + "step": 10124 + }, + { + "epoch": 0.9131081751364026, + "flos": 19289975501280.0, + "grad_norm": 3.08859486182957, + "learning_rate": 7.860783823352512e-08, + "loss": 0.7655, + "num_input_tokens_seen": 329174340, + "step": 10125 + }, + { + "epoch": 0.9131983586598729, + "flos": 28576555126560.0, + "grad_norm": 1.9004568027684403, + "learning_rate": 7.844574199024445e-08, + "loss": 0.7451, + "num_input_tokens_seen": 329203925, + "step": 10126 + }, + { + "epoch": 0.9132885421833431, + "flos": 23006607084480.0, + "grad_norm": 1.8786039049676342, + "learning_rate": 7.82838097045686e-08, + "loss": 0.7794, + "num_input_tokens_seen": 329233225, + "step": 10127 + }, + { + "epoch": 0.9133787257068133, + "flos": 21257792021760.0, + "grad_norm": 2.2260272537167083, + "learning_rate": 7.812204139031454e-08, + "loss": 0.7242, + "num_input_tokens_seen": 329260020, + "step": 10128 + }, + { + "epoch": 0.9134689092302837, + "flos": 16193247137280.0, + "grad_norm": 4.960505909806538, + "learning_rate": 7.796043706128474e-08, + "loss": 0.7547, + "num_input_tokens_seen": 329285740, + "step": 10129 + }, + { + "epoch": 0.9135590927537539, + "flos": 19327002315840.0, + "grad_norm": 1.9882823292067922, + "learning_rate": 7.779899673126844e-08, + "loss": 0.7653, + "num_input_tokens_seen": 329312790, + "step": 10130 + }, + { + "epoch": 0.9136492762772241, + "flos": 33640282276320.0, + "grad_norm": 1.6909420794916894, + "learning_rate": 7.76377204140406e-08, + "loss": 0.7819, + "num_input_tokens_seen": 329342170, + "step": 10131 + }, + { + "epoch": 0.9137394598006944, + "flos": 23188433048640.0, + "grad_norm": 1.7947561625465052, + "learning_rate": 7.74766081233622e-08, + "loss": 0.7297, + "num_input_tokens_seen": 329371195, + "step": 10132 + }, + { + "epoch": 0.9138296433241647, + "flos": 24682260592320.0, + "grad_norm": 1.6760461415605887, + "learning_rate": 7.73156598729805e-08, + "loss": 0.8039, + "num_input_tokens_seen": 329398630, + "step": 10133 + }, + { + "epoch": 0.913919826847635, + "flos": 24209275199040.0, + "grad_norm": 2.1893541380040507, + "learning_rate": 7.715487567662849e-08, + "loss": 0.7899, + "num_input_tokens_seen": 329428160, + "step": 10134 + }, + { + "epoch": 0.9140100103711052, + "flos": 21148718745120.0, + "grad_norm": 1.9544060130109193, + "learning_rate": 7.69942555480243e-08, + "loss": 0.7075, + "num_input_tokens_seen": 329454895, + "step": 10135 + }, + { + "epoch": 0.9141001938945754, + "flos": 17141113581600.0, + "grad_norm": 2.920838197567128, + "learning_rate": 7.68337995008741e-08, + "loss": 0.8411, + "num_input_tokens_seen": 329482595, + "step": 10136 + }, + { + "epoch": 0.9141903774180458, + "flos": 28362460077120.0, + "grad_norm": 1.8899748133439405, + "learning_rate": 7.667350754886803e-08, + "loss": 0.596, + "num_input_tokens_seen": 329511680, + "step": 10137 + }, + { + "epoch": 0.914280560941516, + "flos": 22016695908000.0, + "grad_norm": 2.410752283028971, + "learning_rate": 7.651337970568361e-08, + "loss": 0.7299, + "num_input_tokens_seen": 329539680, + "step": 10138 + }, + { + "epoch": 0.9143707444649862, + "flos": 27670659074880.0, + "grad_norm": 2.0280258413094243, + "learning_rate": 7.635341598498368e-08, + "loss": 0.7012, + "num_input_tokens_seen": 329569840, + "step": 10139 + }, + { + "epoch": 0.9144609279884565, + "flos": 24282102226080.0, + "grad_norm": 1.818337084902988, + "learning_rate": 7.61936164004171e-08, + "loss": 0.7375, + "num_input_tokens_seen": 329597600, + "step": 10140 + }, + { + "epoch": 0.9145511115119268, + "flos": 20926260499680.0, + "grad_norm": 1.5886386362298435, + "learning_rate": 7.603398096561875e-08, + "loss": 0.7803, + "num_input_tokens_seen": 329626490, + "step": 10141 + }, + { + "epoch": 0.914641295035397, + "flos": 21804161988480.0, + "grad_norm": 2.3753947883548143, + "learning_rate": 7.587450969420994e-08, + "loss": 0.7404, + "num_input_tokens_seen": 329656060, + "step": 10142 + }, + { + "epoch": 0.9147314785588673, + "flos": 67836704432640.0, + "grad_norm": 0.6774469035270504, + "learning_rate": 7.571520259979757e-08, + "loss": 0.5604, + "num_input_tokens_seen": 329743580, + "step": 10143 + }, + { + "epoch": 0.9148216620823375, + "flos": 20309962747200.0, + "grad_norm": 2.145162631896509, + "learning_rate": 7.555605969597455e-08, + "loss": 0.5779, + "num_input_tokens_seen": 329770240, + "step": 10144 + }, + { + "epoch": 0.9149118456058078, + "flos": 30033876232320.0, + "grad_norm": 2.48415253800814, + "learning_rate": 7.539708099631959e-08, + "loss": 0.8216, + "num_input_tokens_seen": 329802220, + "step": 10145 + }, + { + "epoch": 0.9150020291292781, + "flos": 26575763295360.0, + "grad_norm": 2.4476568422338096, + "learning_rate": 7.52382665143978e-08, + "loss": 0.7447, + "num_input_tokens_seen": 329832315, + "step": 10146 + }, + { + "epoch": 0.9150922126527483, + "flos": 21985541915520.0, + "grad_norm": 2.004830130708541, + "learning_rate": 7.507961626376014e-08, + "loss": 0.7927, + "num_input_tokens_seen": 329860185, + "step": 10147 + }, + { + "epoch": 0.9151823961762187, + "flos": 28690125944160.0, + "grad_norm": 2.213315729316349, + "learning_rate": 7.492113025794378e-08, + "loss": 0.7755, + "num_input_tokens_seen": 329890335, + "step": 10148 + }, + { + "epoch": 0.9152725796996889, + "flos": 19873483791840.0, + "grad_norm": 2.310860645597298, + "learning_rate": 7.476280851047101e-08, + "loss": 0.684, + "num_input_tokens_seen": 329919535, + "step": 10149 + }, + { + "epoch": 0.9153627632231591, + "flos": 24281247321600.0, + "grad_norm": 1.5018902630294322, + "learning_rate": 7.460465103485125e-08, + "loss": 0.7613, + "num_input_tokens_seen": 329949160, + "step": 10150 + }, + { + "epoch": 0.9154529467466294, + "flos": 23152744345440.0, + "grad_norm": 2.0963718453050837, + "learning_rate": 7.444665784457948e-08, + "loss": 0.7268, + "num_input_tokens_seen": 329974995, + "step": 10151 + }, + { + "epoch": 0.9155431302700997, + "flos": 24427124394240.0, + "grad_norm": 16.45599322844919, + "learning_rate": 7.42888289531356e-08, + "loss": 0.6996, + "num_input_tokens_seen": 330002095, + "step": 10152 + }, + { + "epoch": 0.9156333137935699, + "flos": 33200420872800.0, + "grad_norm": 1.8304860445150706, + "learning_rate": 7.41311643739877e-08, + "loss": 0.661, + "num_input_tokens_seen": 330030535, + "step": 10153 + }, + { + "epoch": 0.9157234973170402, + "flos": 23152112459520.0, + "grad_norm": 1.7153465559225294, + "learning_rate": 7.39736641205877e-08, + "loss": 0.7758, + "num_input_tokens_seen": 330058210, + "step": 10154 + }, + { + "epoch": 0.9158136808405104, + "flos": 17797300220160.0, + "grad_norm": 1.8413855051806982, + "learning_rate": 7.381632820637462e-08, + "loss": 0.7644, + "num_input_tokens_seen": 330085515, + "step": 10155 + }, + { + "epoch": 0.9159038643639807, + "flos": 20746776230400.0, + "grad_norm": 2.0080800156551732, + "learning_rate": 7.365915664477352e-08, + "loss": 0.7255, + "num_input_tokens_seen": 330114920, + "step": 10156 + }, + { + "epoch": 0.915994047887451, + "flos": 30906053578080.0, + "grad_norm": 1.955931516322174, + "learning_rate": 7.350214944919474e-08, + "loss": 0.7424, + "num_input_tokens_seen": 330143610, + "step": 10157 + }, + { + "epoch": 0.9160842314109212, + "flos": 26972799401760.0, + "grad_norm": 11.546234071775714, + "learning_rate": 7.334530663303539e-08, + "loss": 0.7675, + "num_input_tokens_seen": 330171910, + "step": 10158 + }, + { + "epoch": 0.9161744149343914, + "flos": 24354929253120.0, + "grad_norm": 2.022478855998605, + "learning_rate": 7.318862820967742e-08, + "loss": 0.7438, + "num_input_tokens_seen": 330200250, + "step": 10159 + }, + { + "epoch": 0.9162645984578618, + "flos": 21513039729120.0, + "grad_norm": 2.800124853620248, + "learning_rate": 7.303211419249056e-08, + "loss": 0.7722, + "num_input_tokens_seen": 330227980, + "step": 10160 + }, + { + "epoch": 0.916354781981332, + "flos": 30441580059840.0, + "grad_norm": 1.8516800597651013, + "learning_rate": 7.287576459482858e-08, + "loss": 0.7551, + "num_input_tokens_seen": 330261820, + "step": 10161 + }, + { + "epoch": 0.9164449655048023, + "flos": 21110651177280.0, + "grad_norm": 4.587324566306132, + "learning_rate": 7.271957943003259e-08, + "loss": 0.7786, + "num_input_tokens_seen": 330290875, + "step": 10162 + }, + { + "epoch": 0.9165351490282725, + "flos": 17833249111680.0, + "grad_norm": 2.0257749275670838, + "learning_rate": 7.256355871142883e-08, + "loss": 0.7362, + "num_input_tokens_seen": 330318965, + "step": 10163 + }, + { + "epoch": 0.9166253325517428, + "flos": 25957681394400.0, + "grad_norm": 1.8818130689734962, + "learning_rate": 7.240770245233019e-08, + "loss": 0.7887, + "num_input_tokens_seen": 330346930, + "step": 10164 + }, + { + "epoch": 0.9167155160752131, + "flos": 26754243981120.0, + "grad_norm": 2.1579945459164342, + "learning_rate": 7.225201066603492e-08, + "loss": 0.655, + "num_input_tokens_seen": 330375945, + "step": 10165 + }, + { + "epoch": 0.9168056995986833, + "flos": 24752114038560.0, + "grad_norm": 3.0176433864807803, + "learning_rate": 7.209648336582774e-08, + "loss": 0.8482, + "num_input_tokens_seen": 330405570, + "step": 10166 + }, + { + "epoch": 0.9168958831221535, + "flos": 23407694694720.0, + "grad_norm": 1.956330438489663, + "learning_rate": 7.19411205649787e-08, + "loss": 0.7264, + "num_input_tokens_seen": 330435765, + "step": 10167 + }, + { + "epoch": 0.9169860666456239, + "flos": 23219549871360.0, + "grad_norm": 2.018539214570898, + "learning_rate": 7.178592227674474e-08, + "loss": 0.7605, + "num_input_tokens_seen": 330464760, + "step": 10168 + }, + { + "epoch": 0.9170762501690941, + "flos": 21476719140000.0, + "grad_norm": 1.9880809870443819, + "learning_rate": 7.163088851436771e-08, + "loss": 0.6565, + "num_input_tokens_seen": 330493340, + "step": 10169 + }, + { + "epoch": 0.9171664336925643, + "flos": 24391547200320.0, + "grad_norm": 1.8427256308592472, + "learning_rate": 7.147601929107639e-08, + "loss": 0.7068, + "num_input_tokens_seen": 330524175, + "step": 10170 + }, + { + "epoch": 0.9172566172160346, + "flos": 21731111942880.0, + "grad_norm": 1.6124646796564719, + "learning_rate": 7.132131462008461e-08, + "loss": 0.7831, + "num_input_tokens_seen": 330552975, + "step": 10171 + }, + { + "epoch": 0.9173468007395049, + "flos": 22896678903360.0, + "grad_norm": 2.0729271842140204, + "learning_rate": 7.116677451459297e-08, + "loss": 0.7455, + "num_input_tokens_seen": 330581860, + "step": 10172 + }, + { + "epoch": 0.9174369842629752, + "flos": 25045429313760.0, + "grad_norm": 1.356202380491417, + "learning_rate": 7.101239898778799e-08, + "loss": 0.7629, + "num_input_tokens_seen": 330614365, + "step": 10173 + }, + { + "epoch": 0.9175271677864454, + "flos": 27997284188640.0, + "grad_norm": 2.129461989354406, + "learning_rate": 7.085818805284094e-08, + "loss": 0.7747, + "num_input_tokens_seen": 330644720, + "step": 10174 + }, + { + "epoch": 0.9176173513099157, + "flos": 18307200918720.0, + "grad_norm": 6.132663836772996, + "learning_rate": 7.070414172291083e-08, + "loss": 0.8175, + "num_input_tokens_seen": 330669755, + "step": 10175 + }, + { + "epoch": 0.917707534833386, + "flos": 63804238433280.0, + "grad_norm": 1.51133033040414, + "learning_rate": 7.055026001114095e-08, + "loss": 0.6801, + "num_input_tokens_seen": 330704540, + "step": 10176 + }, + { + "epoch": 0.9177977183568562, + "flos": 57816433485600.0, + "grad_norm": 0.6508064887933156, + "learning_rate": 7.039654293066211e-08, + "loss": 0.5647, + "num_input_tokens_seen": 330783235, + "step": 10177 + }, + { + "epoch": 0.9178879018803264, + "flos": 21220988225760.0, + "grad_norm": 2.6047279871933013, + "learning_rate": 7.024299049459003e-08, + "loss": 0.6578, + "num_input_tokens_seen": 330811985, + "step": 10178 + }, + { + "epoch": 0.9179780854037968, + "flos": 24354743404320.0, + "grad_norm": 3.3636875593896027, + "learning_rate": 7.008960271602627e-08, + "loss": 0.6494, + "num_input_tokens_seen": 330838610, + "step": 10179 + }, + { + "epoch": 0.918068268927267, + "flos": 29417652819360.0, + "grad_norm": 1.8499738098558882, + "learning_rate": 6.993637960805921e-08, + "loss": 0.7021, + "num_input_tokens_seen": 330869155, + "step": 10180 + }, + { + "epoch": 0.9181584524507372, + "flos": 20959124301120.0, + "grad_norm": 1.6332458213146597, + "learning_rate": 6.97833211837624e-08, + "loss": 0.7657, + "num_input_tokens_seen": 330896495, + "step": 10181 + }, + { + "epoch": 0.9182486359742075, + "flos": 17322642187680.0, + "grad_norm": 2.018532009578706, + "learning_rate": 6.963042745619562e-08, + "loss": 0.6877, + "num_input_tokens_seen": 330921930, + "step": 10182 + }, + { + "epoch": 0.9183388194976778, + "flos": 16267003408320.0, + "grad_norm": 1.77417245835553, + "learning_rate": 6.947769843840511e-08, + "loss": 0.7071, + "num_input_tokens_seen": 330949540, + "step": 10183 + }, + { + "epoch": 0.918429003021148, + "flos": 26791307965440.0, + "grad_norm": 3.5022248962099525, + "learning_rate": 6.9325134143422e-08, + "loss": 0.7692, + "num_input_tokens_seen": 330977490, + "step": 10184 + }, + { + "epoch": 0.9185191865446183, + "flos": 47489348906880.0, + "grad_norm": 4.072794854699509, + "learning_rate": 6.917273458426387e-08, + "loss": 0.7177, + "num_input_tokens_seen": 331009945, + "step": 10185 + }, + { + "epoch": 0.9186093700680885, + "flos": 62861086814880.0, + "grad_norm": 0.6838085599994458, + "learning_rate": 6.902049977393476e-08, + "loss": 0.5529, + "num_input_tokens_seen": 331100860, + "step": 10186 + }, + { + "epoch": 0.9186995535915589, + "flos": 28872509454720.0, + "grad_norm": 1.7107141078922714, + "learning_rate": 6.886842972542362e-08, + "loss": 0.745, + "num_input_tokens_seen": 331131430, + "step": 10187 + }, + { + "epoch": 0.9187897371150291, + "flos": 37360668005280.0, + "grad_norm": 1.7546288715734812, + "learning_rate": 6.871652445170672e-08, + "loss": 0.6284, + "num_input_tokens_seen": 331162910, + "step": 10188 + }, + { + "epoch": 0.9188799206384993, + "flos": 30401839852800.0, + "grad_norm": 2.8864752598384036, + "learning_rate": 6.856478396574416e-08, + "loss": 0.821, + "num_input_tokens_seen": 331193315, + "step": 10189 + }, + { + "epoch": 0.9189701041619696, + "flos": 20857373467200.0, + "grad_norm": 1.9374193974618343, + "learning_rate": 6.841320828048491e-08, + "loss": 0.7999, + "num_input_tokens_seen": 331220390, + "step": 10190 + }, + { + "epoch": 0.9190602876854399, + "flos": 21949444344960.0, + "grad_norm": 1.659386956985412, + "learning_rate": 6.826179740886062e-08, + "loss": 0.7396, + "num_input_tokens_seen": 331248555, + "step": 10191 + }, + { + "epoch": 0.9191504712089101, + "flos": 18958741337280.0, + "grad_norm": 2.5681243208742486, + "learning_rate": 6.811055136379184e-08, + "loss": 0.8002, + "num_input_tokens_seen": 331275835, + "step": 10192 + }, + { + "epoch": 0.9192406547323804, + "flos": 22350717804000.0, + "grad_norm": 1.9669753053372685, + "learning_rate": 6.79594701581827e-08, + "loss": 0.7174, + "num_input_tokens_seen": 331303125, + "step": 10193 + }, + { + "epoch": 0.9193308382558506, + "flos": 19362876867840.0, + "grad_norm": 2.0488381326465546, + "learning_rate": 6.780855380492511e-08, + "loss": 0.7462, + "num_input_tokens_seen": 331330375, + "step": 10194 + }, + { + "epoch": 0.919421021779321, + "flos": 24239648626560.0, + "grad_norm": 1.8981861842961258, + "learning_rate": 6.765780231689544e-08, + "loss": 0.7548, + "num_input_tokens_seen": 331360150, + "step": 10195 + }, + { + "epoch": 0.9195112053027912, + "flos": 14043864840960.0, + "grad_norm": 2.1900847287096, + "learning_rate": 6.750721570695695e-08, + "loss": 0.796, + "num_input_tokens_seen": 331383840, + "step": 10196 + }, + { + "epoch": 0.9196013888262614, + "flos": 25190971858560.0, + "grad_norm": 2.7893344949546033, + "learning_rate": 6.735679398795868e-08, + "loss": 0.7127, + "num_input_tokens_seen": 331408060, + "step": 10197 + }, + { + "epoch": 0.9196915723497318, + "flos": 27633149053440.0, + "grad_norm": 2.5218720200909828, + "learning_rate": 6.720653717273506e-08, + "loss": 0.6743, + "num_input_tokens_seen": 331438255, + "step": 10198 + }, + { + "epoch": 0.919781755873202, + "flos": 21947474347680.0, + "grad_norm": 2.8284137220166086, + "learning_rate": 6.705644527410714e-08, + "loss": 0.7401, + "num_input_tokens_seen": 331466495, + "step": 10199 + }, + { + "epoch": 0.9198719393966722, + "flos": 15792679903680.0, + "grad_norm": 2.301079256353104, + "learning_rate": 6.690651830488136e-08, + "loss": 0.7058, + "num_input_tokens_seen": 331492985, + "step": 10200 + }, + { + "epoch": 0.9199621229201425, + "flos": 12259063716960.0, + "grad_norm": 3.6236667154247972, + "learning_rate": 6.675675627785037e-08, + "loss": 0.736, + "num_input_tokens_seen": 331519705, + "step": 10201 + }, + { + "epoch": 0.9200523064436128, + "flos": 22166364296160.0, + "grad_norm": 1.952068073398977, + "learning_rate": 6.660715920579263e-08, + "loss": 0.7574, + "num_input_tokens_seen": 331548265, + "step": 10202 + }, + { + "epoch": 0.920142489967083, + "flos": 68083328755680.0, + "grad_norm": 0.5902961612780808, + "learning_rate": 6.645772710147279e-08, + "loss": 0.5769, + "num_input_tokens_seen": 331644795, + "step": 10203 + }, + { + "epoch": 0.9202326734905533, + "flos": 30984902106240.0, + "grad_norm": 1.4387094137000327, + "learning_rate": 6.630845997764112e-08, + "loss": 0.7106, + "num_input_tokens_seen": 331679450, + "step": 10204 + }, + { + "epoch": 0.9203228570140235, + "flos": 24974089077120.0, + "grad_norm": 1.6066608690154598, + "learning_rate": 6.615935784703409e-08, + "loss": 0.6791, + "num_input_tokens_seen": 331710815, + "step": 10205 + }, + { + "epoch": 0.9204130405374938, + "flos": 26501969854560.0, + "grad_norm": 2.2104132472187694, + "learning_rate": 6.601042072237328e-08, + "loss": 0.6647, + "num_input_tokens_seen": 331740690, + "step": 10206 + }, + { + "epoch": 0.9205032240609641, + "flos": 24645642645120.0, + "grad_norm": 2.3106691484513173, + "learning_rate": 6.586164861636767e-08, + "loss": 0.7711, + "num_input_tokens_seen": 331770630, + "step": 10207 + }, + { + "epoch": 0.9205934075844343, + "flos": 20748374530080.0, + "grad_norm": 2.4266366724983204, + "learning_rate": 6.571304154171065e-08, + "loss": 0.6536, + "num_input_tokens_seen": 331798735, + "step": 10208 + }, + { + "epoch": 0.9206835911079045, + "flos": 26683758648960.0, + "grad_norm": 1.9875327389497806, + "learning_rate": 6.556459951108273e-08, + "loss": 0.7648, + "num_input_tokens_seen": 331828515, + "step": 10209 + }, + { + "epoch": 0.9207737746313749, + "flos": 19946496667680.0, + "grad_norm": 2.8845751989450963, + "learning_rate": 6.541632253714957e-08, + "loss": 0.7157, + "num_input_tokens_seen": 331854755, + "step": 10210 + }, + { + "epoch": 0.9208639581548451, + "flos": 22168929009600.0, + "grad_norm": 2.1680030613838905, + "learning_rate": 6.526821063256261e-08, + "loss": 0.6673, + "num_input_tokens_seen": 331883840, + "step": 10211 + }, + { + "epoch": 0.9209541416783154, + "flos": 19947128553600.0, + "grad_norm": 2.101390391963677, + "learning_rate": 6.512026380996016e-08, + "loss": 0.6473, + "num_input_tokens_seen": 331911010, + "step": 10212 + }, + { + "epoch": 0.9210443252017856, + "flos": 35357088442080.0, + "grad_norm": 1.8873574891511302, + "learning_rate": 6.49724820819657e-08, + "loss": 0.7084, + "num_input_tokens_seen": 331941440, + "step": 10213 + }, + { + "epoch": 0.9211345087252559, + "flos": 22278113795520.0, + "grad_norm": 1.9219841654717706, + "learning_rate": 6.48248654611887e-08, + "loss": 0.7719, + "num_input_tokens_seen": 331970770, + "step": 10214 + }, + { + "epoch": 0.9212246922487262, + "flos": 28616518352160.0, + "grad_norm": 1.5137247974619337, + "learning_rate": 6.467741396022419e-08, + "loss": 0.7859, + "num_input_tokens_seen": 332003690, + "step": 10215 + }, + { + "epoch": 0.9213148757721964, + "flos": 24828658041600.0, + "grad_norm": 1.8983261640891702, + "learning_rate": 6.453012759165455e-08, + "loss": 0.7458, + "num_input_tokens_seen": 332032830, + "step": 10216 + }, + { + "epoch": 0.9214050592956666, + "flos": 25522094513280.0, + "grad_norm": 1.6656231792504412, + "learning_rate": 6.438300636804639e-08, + "loss": 0.7057, + "num_input_tokens_seen": 332062120, + "step": 10217 + }, + { + "epoch": 0.921495242819137, + "flos": 24974906811840.0, + "grad_norm": 1.7374480918979407, + "learning_rate": 6.423605030195278e-08, + "loss": 0.7718, + "num_input_tokens_seen": 332092240, + "step": 10218 + }, + { + "epoch": 0.9215854263426072, + "flos": 17614061805120.0, + "grad_norm": 2.055947037382048, + "learning_rate": 6.408925940591304e-08, + "loss": 0.8105, + "num_input_tokens_seen": 332118770, + "step": 10219 + }, + { + "epoch": 0.9216756098660774, + "flos": 36845563540320.0, + "grad_norm": 2.040883270790835, + "learning_rate": 6.394263369245222e-08, + "loss": 0.6741, + "num_input_tokens_seen": 332152500, + "step": 10220 + }, + { + "epoch": 0.9217657933895478, + "flos": 74623419395520.0, + "grad_norm": 2.0063297719337063, + "learning_rate": 6.379617317408126e-08, + "loss": 0.7423, + "num_input_tokens_seen": 332186715, + "step": 10221 + }, + { + "epoch": 0.921855976913018, + "flos": 26103112429920.0, + "grad_norm": 2.3921046446361607, + "learning_rate": 6.364987786329723e-08, + "loss": 0.7349, + "num_input_tokens_seen": 332216115, + "step": 10222 + }, + { + "epoch": 0.9219461604364882, + "flos": 26869152910080.0, + "grad_norm": 2.4305284032364667, + "learning_rate": 6.350374777258193e-08, + "loss": 0.7324, + "num_input_tokens_seen": 332246920, + "step": 10223 + }, + { + "epoch": 0.9220363439599585, + "flos": 22569273224640.0, + "grad_norm": 1.7126495992097923, + "learning_rate": 6.335778291440519e-08, + "loss": 0.7181, + "num_input_tokens_seen": 332276130, + "step": 10224 + }, + { + "epoch": 0.9221265274834288, + "flos": 25372834992480.0, + "grad_norm": 1.8849135582933663, + "learning_rate": 6.321198330122057e-08, + "loss": 0.6433, + "num_input_tokens_seen": 332305935, + "step": 10225 + }, + { + "epoch": 0.9222167110068991, + "flos": 28721763143520.0, + "grad_norm": 1.7890289836559279, + "learning_rate": 6.306634894546902e-08, + "loss": 0.6757, + "num_input_tokens_seen": 332338425, + "step": 10226 + }, + { + "epoch": 0.9223068945303693, + "flos": 23880419899680.0, + "grad_norm": 1.589154518186359, + "learning_rate": 6.292087985957661e-08, + "loss": 0.6794, + "num_input_tokens_seen": 332369190, + "step": 10227 + }, + { + "epoch": 0.9223970780538395, + "flos": 20128694329440.0, + "grad_norm": 1.996121281323865, + "learning_rate": 6.277557605595585e-08, + "loss": 0.7769, + "num_input_tokens_seen": 332396245, + "step": 10228 + }, + { + "epoch": 0.9224872615773099, + "flos": 23735323392000.0, + "grad_norm": 1.5776355097585124, + "learning_rate": 6.263043754700481e-08, + "loss": 0.6605, + "num_input_tokens_seen": 332426015, + "step": 10229 + }, + { + "epoch": 0.9225774451007801, + "flos": 18088496819040.0, + "grad_norm": 2.229018934351722, + "learning_rate": 6.248546434510671e-08, + "loss": 0.7391, + "num_input_tokens_seen": 332454265, + "step": 10230 + }, + { + "epoch": 0.9226676286242503, + "flos": 35830928739840.0, + "grad_norm": 2.077579742260022, + "learning_rate": 6.234065646263298e-08, + "loss": 0.6698, + "num_input_tokens_seen": 332486115, + "step": 10231 + }, + { + "epoch": 0.9227578121477206, + "flos": 20273976685920.0, + "grad_norm": 2.0211348816616335, + "learning_rate": 6.219601391193796e-08, + "loss": 0.6501, + "num_input_tokens_seen": 332514685, + "step": 10232 + }, + { + "epoch": 0.9228479956711909, + "flos": 44026738428960.0, + "grad_norm": 1.6972104185408086, + "learning_rate": 6.205153670536423e-08, + "loss": 0.6652, + "num_input_tokens_seen": 332550860, + "step": 10233 + }, + { + "epoch": 0.9229381791946611, + "flos": 27267341279040.0, + "grad_norm": 1.7663595646421437, + "learning_rate": 6.190722485523902e-08, + "loss": 0.775, + "num_input_tokens_seen": 332583135, + "step": 10234 + }, + { + "epoch": 0.9230283627181314, + "flos": 23116572435360.0, + "grad_norm": 1.8788509241789477, + "learning_rate": 6.176307837387607e-08, + "loss": 0.716, + "num_input_tokens_seen": 332610275, + "step": 10235 + }, + { + "epoch": 0.9231185462416016, + "flos": 27557459954880.0, + "grad_norm": 1.509881847887469, + "learning_rate": 6.16190972735744e-08, + "loss": 0.6874, + "num_input_tokens_seen": 332639745, + "step": 10236 + }, + { + "epoch": 0.923208729765072, + "flos": 34409853883680.0, + "grad_norm": 2.4592737780514344, + "learning_rate": 6.147528156661974e-08, + "loss": 0.7636, + "num_input_tokens_seen": 332670200, + "step": 10237 + }, + { + "epoch": 0.9232989132885422, + "flos": 25009926459360.0, + "grad_norm": 2.003118944084547, + "learning_rate": 6.133163126528273e-08, + "loss": 0.6618, + "num_input_tokens_seen": 332700460, + "step": 10238 + }, + { + "epoch": 0.9233890968120124, + "flos": 20268103863840.0, + "grad_norm": 1.9789649325170167, + "learning_rate": 6.11881463818209e-08, + "loss": 0.7471, + "num_input_tokens_seen": 332727585, + "step": 10239 + }, + { + "epoch": 0.9234792803354827, + "flos": 21105930617760.0, + "grad_norm": 1.708210971761041, + "learning_rate": 6.104482692847668e-08, + "loss": 0.7249, + "num_input_tokens_seen": 332754645, + "step": 10240 + }, + { + "epoch": 0.923569463858953, + "flos": 22533101314560.0, + "grad_norm": 1.7783854020685181, + "learning_rate": 6.090167291747917e-08, + "loss": 0.7063, + "num_input_tokens_seen": 332784690, + "step": 10241 + }, + { + "epoch": 0.9236596473824232, + "flos": 24609470735040.0, + "grad_norm": 2.087214047154171, + "learning_rate": 6.075868436104303e-08, + "loss": 0.6392, + "num_input_tokens_seen": 332814090, + "step": 10242 + }, + { + "epoch": 0.9237498309058935, + "flos": 27305074319040.0, + "grad_norm": 2.1051442987995648, + "learning_rate": 6.061586127136875e-08, + "loss": 0.7235, + "num_input_tokens_seen": 332845605, + "step": 10243 + }, + { + "epoch": 0.9238400144293638, + "flos": 23696846956800.0, + "grad_norm": 1.7603443815600581, + "learning_rate": 6.047320366064324e-08, + "loss": 0.7048, + "num_input_tokens_seen": 332876895, + "step": 10244 + }, + { + "epoch": 0.923930197952834, + "flos": 20488554942240.0, + "grad_norm": 1.8473711176867016, + "learning_rate": 6.033071154103786e-08, + "loss": 0.7456, + "num_input_tokens_seen": 332905860, + "step": 10245 + }, + { + "epoch": 0.9240203814763043, + "flos": 21695720597760.0, + "grad_norm": 2.2135653027039246, + "learning_rate": 6.018838492471178e-08, + "loss": 0.7594, + "num_input_tokens_seen": 332933560, + "step": 10246 + }, + { + "epoch": 0.9241105649997745, + "flos": 70333532641920.0, + "grad_norm": 0.6206978069675374, + "learning_rate": 6.00462238238082e-08, + "loss": 0.5791, + "num_input_tokens_seen": 333030060, + "step": 10247 + }, + { + "epoch": 0.9242007485232449, + "flos": 25552802468640.0, + "grad_norm": 2.520560499406958, + "learning_rate": 5.990422825045827e-08, + "loss": 0.6992, + "num_input_tokens_seen": 333058255, + "step": 10248 + }, + { + "epoch": 0.9242909320467151, + "flos": 20381897700000.0, + "grad_norm": 3.0085689237566324, + "learning_rate": 5.976239821677675e-08, + "loss": 0.715, + "num_input_tokens_seen": 333086645, + "step": 10249 + }, + { + "epoch": 0.9243811155701853, + "flos": 26904990292320.0, + "grad_norm": 3.3294725152671605, + "learning_rate": 5.962073373486598e-08, + "loss": 0.6806, + "num_input_tokens_seen": 333117380, + "step": 10250 + }, + { + "epoch": 0.9244712990936556, + "flos": 34919866091520.0, + "grad_norm": 1.8455624481359287, + "learning_rate": 5.947923481681316e-08, + "loss": 0.7004, + "num_input_tokens_seen": 333149170, + "step": 10251 + }, + { + "epoch": 0.9245614826171259, + "flos": 26465314737600.0, + "grad_norm": 1.8355914499959456, + "learning_rate": 5.933790147469198e-08, + "loss": 0.7091, + "num_input_tokens_seen": 333177955, + "step": 10252 + }, + { + "epoch": 0.9246516661405961, + "flos": 22929914402400.0, + "grad_norm": 2.175019218096837, + "learning_rate": 5.9196733720561665e-08, + "loss": 0.6887, + "num_input_tokens_seen": 333204710, + "step": 10253 + }, + { + "epoch": 0.9247418496640664, + "flos": 19977799339200.0, + "grad_norm": 2.032014086869739, + "learning_rate": 5.905573156646793e-08, + "loss": 0.8208, + "num_input_tokens_seen": 333230990, + "step": 10254 + }, + { + "epoch": 0.9248320331875366, + "flos": 23370556370880.0, + "grad_norm": 2.428774548699538, + "learning_rate": 5.8914895024441134e-08, + "loss": 0.8138, + "num_input_tokens_seen": 333257450, + "step": 10255 + }, + { + "epoch": 0.9249222167110069, + "flos": 18810782758080.0, + "grad_norm": 2.3666532310140287, + "learning_rate": 5.877422410649857e-08, + "loss": 0.7725, + "num_input_tokens_seen": 333280985, + "step": 10256 + }, + { + "epoch": 0.9250124002344772, + "flos": 17395878082080.0, + "grad_norm": 2.4670868111974444, + "learning_rate": 5.863371882464285e-08, + "loss": 0.7535, + "num_input_tokens_seen": 333307550, + "step": 10257 + }, + { + "epoch": 0.9251025837579474, + "flos": 19472916558240.0, + "grad_norm": 1.949400318075514, + "learning_rate": 5.849337919086283e-08, + "loss": 0.7221, + "num_input_tokens_seen": 333333880, + "step": 10258 + }, + { + "epoch": 0.9251927672814176, + "flos": 26613124637760.0, + "grad_norm": 2.1006045515795444, + "learning_rate": 5.835320521713316e-08, + "loss": 0.7158, + "num_input_tokens_seen": 333362885, + "step": 10259 + }, + { + "epoch": 0.925282950804888, + "flos": 21475789896000.0, + "grad_norm": 1.5849688348203206, + "learning_rate": 5.8213196915414264e-08, + "loss": 0.7214, + "num_input_tokens_seen": 333391290, + "step": 10260 + }, + { + "epoch": 0.9253731343283582, + "flos": 23260999887360.0, + "grad_norm": 3.264255948540648, + "learning_rate": 5.807335429765237e-08, + "loss": 0.7941, + "num_input_tokens_seen": 333418195, + "step": 10261 + }, + { + "epoch": 0.9254633178518284, + "flos": 22164840336000.0, + "grad_norm": 1.921900865334236, + "learning_rate": 5.7933677375779034e-08, + "loss": 0.7864, + "num_input_tokens_seen": 333447745, + "step": 10262 + }, + { + "epoch": 0.9255535013752987, + "flos": 23331039182400.0, + "grad_norm": 2.5316801294400353, + "learning_rate": 5.77941661617134e-08, + "loss": 0.7877, + "num_input_tokens_seen": 333478495, + "step": 10263 + }, + { + "epoch": 0.925643684898769, + "flos": 26174861533920.0, + "grad_norm": 2.080269531613083, + "learning_rate": 5.765482066735816e-08, + "loss": 0.6706, + "num_input_tokens_seen": 333507235, + "step": 10264 + }, + { + "epoch": 0.9257338684222393, + "flos": 18051470004480.0, + "grad_norm": 2.1332032680598743, + "learning_rate": 5.7515640904604256e-08, + "loss": 0.7426, + "num_input_tokens_seen": 333534270, + "step": 10265 + }, + { + "epoch": 0.9258240519457095, + "flos": 23951537117760.0, + "grad_norm": 1.9237155939486916, + "learning_rate": 5.7376626885326187e-08, + "loss": 0.7218, + "num_input_tokens_seen": 333563960, + "step": 10266 + }, + { + "epoch": 0.9259142354691798, + "flos": 21767841399360.0, + "grad_norm": 1.8200344151607093, + "learning_rate": 5.723777862138601e-08, + "loss": 0.8124, + "num_input_tokens_seen": 333592700, + "step": 10267 + }, + { + "epoch": 0.9260044189926501, + "flos": 16558051328160.0, + "grad_norm": 1.820344916452073, + "learning_rate": 5.7099096124630705e-08, + "loss": 0.8181, + "num_input_tokens_seen": 333618385, + "step": 10268 + }, + { + "epoch": 0.9260946025161203, + "flos": 24936987923040.0, + "grad_norm": 1.3917133915456366, + "learning_rate": 5.696057940689347e-08, + "loss": 0.7474, + "num_input_tokens_seen": 333649165, + "step": 10269 + }, + { + "epoch": 0.9261847860395905, + "flos": 22532692447200.0, + "grad_norm": 2.222020170509952, + "learning_rate": 5.6822228479993736e-08, + "loss": 0.7571, + "num_input_tokens_seen": 333678055, + "step": 10270 + }, + { + "epoch": 0.9262749695630609, + "flos": 20672388073440.0, + "grad_norm": 2.0303249893872946, + "learning_rate": 5.668404335573584e-08, + "loss": 0.7806, + "num_input_tokens_seen": 333702030, + "step": 10271 + }, + { + "epoch": 0.9263651530865311, + "flos": 27013952059680.0, + "grad_norm": 1.7758549417762017, + "learning_rate": 5.654602404591058e-08, + "loss": 0.683, + "num_input_tokens_seen": 333732655, + "step": 10272 + }, + { + "epoch": 0.9264553366100013, + "flos": 29673904110240.0, + "grad_norm": 1.8588953120697091, + "learning_rate": 5.640817056229474e-08, + "loss": 0.7962, + "num_input_tokens_seen": 333762340, + "step": 10273 + }, + { + "epoch": 0.9265455201334716, + "flos": 22387187072160.0, + "grad_norm": 1.73100434361028, + "learning_rate": 5.6270482916650706e-08, + "loss": 0.7793, + "num_input_tokens_seen": 333791420, + "step": 10274 + }, + { + "epoch": 0.9266357036569419, + "flos": 27925535084640.0, + "grad_norm": 2.0426423083704766, + "learning_rate": 5.613296112072663e-08, + "loss": 0.8132, + "num_input_tokens_seen": 333822790, + "step": 10275 + }, + { + "epoch": 0.9267258871804122, + "flos": 20710641490080.0, + "grad_norm": 2.4511709023279944, + "learning_rate": 5.59956051862569e-08, + "loss": 0.7423, + "num_input_tokens_seen": 333848125, + "step": 10276 + }, + { + "epoch": 0.9268160707038824, + "flos": 22569682092000.0, + "grad_norm": 1.8062830885658034, + "learning_rate": 5.585841512496081e-08, + "loss": 0.8033, + "num_input_tokens_seen": 333878090, + "step": 10277 + }, + { + "epoch": 0.9269062542273526, + "flos": 26175530589600.0, + "grad_norm": 2.0707200135119983, + "learning_rate": 5.5721390948545e-08, + "loss": 0.8475, + "num_input_tokens_seen": 333904770, + "step": 10278 + }, + { + "epoch": 0.926996437750823, + "flos": 25302572678880.0, + "grad_norm": 2.5353832642845076, + "learning_rate": 5.558453266870056e-08, + "loss": 0.784, + "num_input_tokens_seen": 333935095, + "step": 10279 + }, + { + "epoch": 0.9270866212742932, + "flos": 28799868276480.0, + "grad_norm": 1.8494066008420207, + "learning_rate": 5.544784029710525e-08, + "loss": 0.6793, + "num_input_tokens_seen": 333964960, + "step": 10280 + }, + { + "epoch": 0.9271768047977634, + "flos": 16411988406720.0, + "grad_norm": 2.1620734262622343, + "learning_rate": 5.531131384542242e-08, + "loss": 0.8106, + "num_input_tokens_seen": 333991190, + "step": 10281 + }, + { + "epoch": 0.9272669883212337, + "flos": 25303130225280.0, + "grad_norm": 2.589273013003913, + "learning_rate": 5.51749533253012e-08, + "loss": 0.6889, + "num_input_tokens_seen": 334021830, + "step": 10282 + }, + { + "epoch": 0.927357171844704, + "flos": 35644419385920.0, + "grad_norm": 1.706427705334904, + "learning_rate": 5.503875874837649e-08, + "loss": 0.6822, + "num_input_tokens_seen": 334050575, + "step": 10283 + }, + { + "epoch": 0.9274473553681742, + "flos": 25010112308160.0, + "grad_norm": 2.316437745161625, + "learning_rate": 5.4902730126269225e-08, + "loss": 0.706, + "num_input_tokens_seen": 334079210, + "step": 10284 + }, + { + "epoch": 0.9275375388916445, + "flos": 26832014586240.0, + "grad_norm": 1.8928227537172482, + "learning_rate": 5.476686747058656e-08, + "loss": 0.7685, + "num_input_tokens_seen": 334108560, + "step": 10285 + }, + { + "epoch": 0.9276277224151147, + "flos": 22715745013440.0, + "grad_norm": 2.1897213148007943, + "learning_rate": 5.4631170792920124e-08, + "loss": 0.7139, + "num_input_tokens_seen": 334136195, + "step": 10286 + }, + { + "epoch": 0.927717905938585, + "flos": 20893508207520.0, + "grad_norm": 2.1389348395339716, + "learning_rate": 5.449564010484953e-08, + "loss": 0.7201, + "num_input_tokens_seen": 334162385, + "step": 10287 + }, + { + "epoch": 0.9278080894620553, + "flos": 24755533656480.0, + "grad_norm": 2.1474580508876686, + "learning_rate": 5.436027541793775e-08, + "loss": 0.7626, + "num_input_tokens_seen": 334190755, + "step": 10288 + }, + { + "epoch": 0.9278982729855255, + "flos": 25552616619840.0, + "grad_norm": 1.7337012142036543, + "learning_rate": 5.4225076743735554e-08, + "loss": 0.7135, + "num_input_tokens_seen": 334222160, + "step": 10289 + }, + { + "epoch": 0.9279884565089958, + "flos": 21768287436480.0, + "grad_norm": 1.925449242128143, + "learning_rate": 5.409004409377882e-08, + "loss": 0.6581, + "num_input_tokens_seen": 334250115, + "step": 10290 + }, + { + "epoch": 0.9280786400324661, + "flos": 48837782584800.0, + "grad_norm": 2.9129405789454896, + "learning_rate": 5.3955177479589e-08, + "loss": 0.6643, + "num_input_tokens_seen": 334285805, + "step": 10291 + }, + { + "epoch": 0.9281688235559363, + "flos": 21987363233760.0, + "grad_norm": 1.8431016460916982, + "learning_rate": 5.3820476912674e-08, + "loss": 0.7476, + "num_input_tokens_seen": 334316025, + "step": 10292 + }, + { + "epoch": 0.9282590070794066, + "flos": 28727301437760.0, + "grad_norm": 1.5065090890503234, + "learning_rate": 5.3685942404527063e-08, + "loss": 0.7674, + "num_input_tokens_seen": 334347520, + "step": 10293 + }, + { + "epoch": 0.9283491906028769, + "flos": 23189325122880.0, + "grad_norm": 1.9055544100029818, + "learning_rate": 5.355157396662702e-08, + "loss": 0.7235, + "num_input_tokens_seen": 334375255, + "step": 10294 + }, + { + "epoch": 0.9284393741263471, + "flos": 26281295757600.0, + "grad_norm": 1.7871640723357645, + "learning_rate": 5.34173716104398e-08, + "loss": 0.7229, + "num_input_tokens_seen": 334405020, + "step": 10295 + }, + { + "epoch": 0.9285295576498174, + "flos": 20377623177600.0, + "grad_norm": 1.82750430533219, + "learning_rate": 5.328333534741536e-08, + "loss": 0.7474, + "num_input_tokens_seen": 334429905, + "step": 10296 + }, + { + "epoch": 0.9286197411732876, + "flos": 23334495970080.0, + "grad_norm": 2.1084263681785327, + "learning_rate": 5.314946518899099e-08, + "loss": 0.713, + "num_input_tokens_seen": 334455135, + "step": 10297 + }, + { + "epoch": 0.928709924696758, + "flos": 25155729192480.0, + "grad_norm": 2.02913426471927, + "learning_rate": 5.301576114658912e-08, + "loss": 0.8125, + "num_input_tokens_seen": 334484225, + "step": 10298 + }, + { + "epoch": 0.9288001082202282, + "flos": 22095247078080.0, + "grad_norm": 1.924780595339298, + "learning_rate": 5.288222323161795e-08, + "loss": 0.7667, + "num_input_tokens_seen": 334511515, + "step": 10299 + }, + { + "epoch": 0.9288902917436984, + "flos": 12804504439680.0, + "grad_norm": 5.382435291289343, + "learning_rate": 5.274885145547214e-08, + "loss": 0.7842, + "num_input_tokens_seen": 334537320, + "step": 10300 + }, + { + "epoch": 0.9289804752671686, + "flos": 25263613036800.0, + "grad_norm": 1.5832524613590646, + "learning_rate": 5.261564582953082e-08, + "loss": 0.732, + "num_input_tokens_seen": 334570210, + "step": 10301 + }, + { + "epoch": 0.929070658790639, + "flos": 22569236054880.0, + "grad_norm": 2.3946425643860576, + "learning_rate": 5.248260636516066e-08, + "loss": 0.7187, + "num_input_tokens_seen": 334598095, + "step": 10302 + }, + { + "epoch": 0.9291608423141092, + "flos": 21002655823680.0, + "grad_norm": 1.995279164584709, + "learning_rate": 5.2349733073712824e-08, + "loss": 0.683, + "num_input_tokens_seen": 334626795, + "step": 10303 + }, + { + "epoch": 0.9292510258375795, + "flos": 26941013523360.0, + "grad_norm": 1.787351031198118, + "learning_rate": 5.221702596652533e-08, + "loss": 0.8286, + "num_input_tokens_seen": 334654825, + "step": 10304 + }, + { + "epoch": 0.9293412093610497, + "flos": 12950939058720.0, + "grad_norm": 2.77985904500923, + "learning_rate": 5.208448505492091e-08, + "loss": 0.6794, + "num_input_tokens_seen": 334679835, + "step": 10305 + }, + { + "epoch": 0.92943139288452, + "flos": 29819483824800.0, + "grad_norm": 2.029628752224576, + "learning_rate": 5.1952110350208965e-08, + "loss": 0.7227, + "num_input_tokens_seen": 334707495, + "step": 10306 + }, + { + "epoch": 0.9295215764079903, + "flos": 25338595909920.0, + "grad_norm": 1.750122679088801, + "learning_rate": 5.181990186368446e-08, + "loss": 0.6974, + "num_input_tokens_seen": 334736870, + "step": 10307 + }, + { + "epoch": 0.9296117599314605, + "flos": 21764310272160.0, + "grad_norm": 1.954112665315472, + "learning_rate": 5.1687859606627915e-08, + "loss": 0.8376, + "num_input_tokens_seen": 334765550, + "step": 10308 + }, + { + "epoch": 0.9297019434549307, + "flos": 36413730804960.0, + "grad_norm": 1.8463636719997487, + "learning_rate": 5.1555983590306327e-08, + "loss": 0.6507, + "num_input_tokens_seen": 334796980, + "step": 10309 + }, + { + "epoch": 0.9297921269784011, + "flos": 70912580561280.0, + "grad_norm": 0.6540854041130395, + "learning_rate": 5.1424273825971806e-08, + "loss": 0.5768, + "num_input_tokens_seen": 334885840, + "step": 10310 + }, + { + "epoch": 0.9298823105018713, + "flos": 34264906055040.0, + "grad_norm": 3.233841856723497, + "learning_rate": 5.1292730324862475e-08, + "loss": 0.7037, + "num_input_tokens_seen": 334916060, + "step": 10311 + }, + { + "epoch": 0.9299724940253415, + "flos": 20782056066240.0, + "grad_norm": 1.6071254229735297, + "learning_rate": 5.116135309820224e-08, + "loss": 0.7539, + "num_input_tokens_seen": 334944410, + "step": 10312 + }, + { + "epoch": 0.9300626775488118, + "flos": 21471664052640.0, + "grad_norm": 1.728250291362723, + "learning_rate": 5.103014215720147e-08, + "loss": 0.7751, + "num_input_tokens_seen": 334972900, + "step": 10313 + }, + { + "epoch": 0.9301528610722821, + "flos": 19216813946400.0, + "grad_norm": 2.496643101471063, + "learning_rate": 5.0899097513055214e-08, + "loss": 0.7955, + "num_input_tokens_seen": 334999965, + "step": 10314 + }, + { + "epoch": 0.9302430445957524, + "flos": 28143086921760.0, + "grad_norm": 2.126068856194577, + "learning_rate": 5.076821917694563e-08, + "loss": 0.7277, + "num_input_tokens_seen": 335028110, + "step": 10315 + }, + { + "epoch": 0.9303332281192226, + "flos": 19654593843360.0, + "grad_norm": 2.0674013795591026, + "learning_rate": 5.063750716003889e-08, + "loss": 0.7883, + "num_input_tokens_seen": 335054545, + "step": 10316 + }, + { + "epoch": 0.9304234116426929, + "flos": 20674246561440.0, + "grad_norm": 1.856472346286102, + "learning_rate": 5.050696147348921e-08, + "loss": 0.7262, + "num_input_tokens_seen": 335080910, + "step": 10317 + }, + { + "epoch": 0.9305135951661632, + "flos": 35758027373280.0, + "grad_norm": 2.110742535400927, + "learning_rate": 5.037658212843454e-08, + "loss": 0.7135, + "num_input_tokens_seen": 335112275, + "step": 10318 + }, + { + "epoch": 0.9306037786896334, + "flos": 34956261020160.0, + "grad_norm": 1.6826583494176315, + "learning_rate": 5.0246369136000444e-08, + "loss": 0.6305, + "num_input_tokens_seen": 335142560, + "step": 10319 + }, + { + "epoch": 0.9306939622131036, + "flos": 19101124452480.0, + "grad_norm": 2.0122666881209095, + "learning_rate": 5.011632250729691e-08, + "loss": 0.8108, + "num_input_tokens_seen": 335169365, + "step": 10320 + }, + { + "epoch": 0.930784145736574, + "flos": 23295387648960.0, + "grad_norm": 2.0011758923792757, + "learning_rate": 4.998644225342019e-08, + "loss": 0.8295, + "num_input_tokens_seen": 335196700, + "step": 10321 + }, + { + "epoch": 0.9308743292600442, + "flos": 21768844982880.0, + "grad_norm": 1.6913488787993072, + "learning_rate": 4.9856728385452296e-08, + "loss": 0.7427, + "num_input_tokens_seen": 335224685, + "step": 10322 + }, + { + "epoch": 0.9309645127835144, + "flos": 21035816983200.0, + "grad_norm": 1.8307633231726714, + "learning_rate": 4.9727180914461485e-08, + "loss": 0.8228, + "num_input_tokens_seen": 335253050, + "step": 10323 + }, + { + "epoch": 0.9310546963069847, + "flos": 26352264296640.0, + "grad_norm": 1.9277904003779318, + "learning_rate": 4.959779985150137e-08, + "loss": 0.749, + "num_input_tokens_seen": 335281375, + "step": 10324 + }, + { + "epoch": 0.931144879830455, + "flos": 16084025181600.0, + "grad_norm": 2.7371194200425335, + "learning_rate": 4.9468585207611105e-08, + "loss": 0.6862, + "num_input_tokens_seen": 335306935, + "step": 10325 + }, + { + "epoch": 0.9312350633539253, + "flos": 20488480602720.0, + "grad_norm": 2.6756322666097163, + "learning_rate": 4.9339536993816764e-08, + "loss": 0.7151, + "num_input_tokens_seen": 335332845, + "step": 10326 + }, + { + "epoch": 0.9313252468773955, + "flos": 26537807236800.0, + "grad_norm": 1.9056712663488353, + "learning_rate": 4.921065522112844e-08, + "loss": 0.6909, + "num_input_tokens_seen": 335361575, + "step": 10327 + }, + { + "epoch": 0.9314154304008657, + "flos": 30581101103520.0, + "grad_norm": 1.8194424100998756, + "learning_rate": 4.908193990054377e-08, + "loss": 0.7839, + "num_input_tokens_seen": 335390195, + "step": 10328 + }, + { + "epoch": 0.9315056139243361, + "flos": 29086827522720.0, + "grad_norm": 1.5723377282331445, + "learning_rate": 4.89533910430453e-08, + "loss": 0.7551, + "num_input_tokens_seen": 335418990, + "step": 10329 + }, + { + "epoch": 0.9315957974478063, + "flos": 23080065997440.0, + "grad_norm": 5.03420630373056, + "learning_rate": 4.8825008659601376e-08, + "loss": 0.7303, + "num_input_tokens_seen": 335448435, + "step": 10330 + }, + { + "epoch": 0.9316859809712765, + "flos": 19544851511040.0, + "grad_norm": 2.2621436370589465, + "learning_rate": 4.869679276116634e-08, + "loss": 0.7564, + "num_input_tokens_seen": 335472445, + "step": 10331 + }, + { + "epoch": 0.9317761644947468, + "flos": 24535156917600.0, + "grad_norm": 1.903633448138219, + "learning_rate": 4.856874335868055e-08, + "loss": 0.758, + "num_input_tokens_seen": 335503325, + "step": 10332 + }, + { + "epoch": 0.9318663480182171, + "flos": 24063249447360.0, + "grad_norm": 4.440472104516178, + "learning_rate": 4.844086046306928e-08, + "loss": 0.8432, + "num_input_tokens_seen": 335529155, + "step": 10333 + }, + { + "epoch": 0.9319565315416873, + "flos": 29783795121600.0, + "grad_norm": 1.8076292559995222, + "learning_rate": 4.8313144085244896e-08, + "loss": 0.7413, + "num_input_tokens_seen": 335559445, + "step": 10334 + }, + { + "epoch": 0.9320467150651576, + "flos": 62621561916000.0, + "grad_norm": 0.6401165828840487, + "learning_rate": 4.818559423610424e-08, + "loss": 0.5739, + "num_input_tokens_seen": 335649605, + "step": 10335 + }, + { + "epoch": 0.9321368985886278, + "flos": 22568641338720.0, + "grad_norm": 1.9186074617108546, + "learning_rate": 4.8058210926531284e-08, + "loss": 0.6816, + "num_input_tokens_seen": 335677965, + "step": 10336 + }, + { + "epoch": 0.9322270821120981, + "flos": 23807964570240.0, + "grad_norm": 1.754485512552704, + "learning_rate": 4.7930994167394435e-08, + "loss": 0.7724, + "num_input_tokens_seen": 335707630, + "step": 10337 + }, + { + "epoch": 0.9323172656355684, + "flos": 20530859862720.0, + "grad_norm": 1.5444872373755685, + "learning_rate": 4.7803943969548786e-08, + "loss": 0.7816, + "num_input_tokens_seen": 335738835, + "step": 10338 + }, + { + "epoch": 0.9324074491590386, + "flos": 22787382608160.0, + "grad_norm": 2.029389336162056, + "learning_rate": 4.7677060343834784e-08, + "loss": 0.6874, + "num_input_tokens_seen": 335768185, + "step": 10339 + }, + { + "epoch": 0.932497632682509, + "flos": 15355383213600.0, + "grad_norm": 1.9363672575344884, + "learning_rate": 4.75503433010791e-08, + "loss": 0.7339, + "num_input_tokens_seen": 335794550, + "step": 10340 + }, + { + "epoch": 0.9325878162059792, + "flos": 28099889927040.0, + "grad_norm": 2.241990137219748, + "learning_rate": 4.742379285209419e-08, + "loss": 0.7223, + "num_input_tokens_seen": 335821655, + "step": 10341 + }, + { + "epoch": 0.9326779997294494, + "flos": 37616361749760.0, + "grad_norm": 1.8539931684036888, + "learning_rate": 4.72974090076772e-08, + "loss": 0.8416, + "num_input_tokens_seen": 335853405, + "step": 10342 + }, + { + "epoch": 0.9327681832529197, + "flos": 26317207479360.0, + "grad_norm": 1.9496300517571532, + "learning_rate": 4.717119177861262e-08, + "loss": 0.6965, + "num_input_tokens_seen": 335883760, + "step": 10343 + }, + { + "epoch": 0.93285836677639, + "flos": 23550189319200.0, + "grad_norm": 1.825720811920049, + "learning_rate": 4.70451411756696e-08, + "loss": 0.7552, + "num_input_tokens_seen": 335911365, + "step": 10344 + }, + { + "epoch": 0.9329485502998602, + "flos": 21400807022880.0, + "grad_norm": 1.53961860950132, + "learning_rate": 4.691925720960355e-08, + "loss": 0.7795, + "num_input_tokens_seen": 335940035, + "step": 10345 + }, + { + "epoch": 0.9330387338233305, + "flos": 16953823662720.0, + "grad_norm": 1.9664275382294918, + "learning_rate": 4.6793539891155645e-08, + "loss": 0.7621, + "num_input_tokens_seen": 335966580, + "step": 10346 + }, + { + "epoch": 0.9331289173468007, + "flos": 36085321542720.0, + "grad_norm": 2.1872425472800963, + "learning_rate": 4.6667989231052864e-08, + "loss": 0.7194, + "num_input_tokens_seen": 335997530, + "step": 10347 + }, + { + "epoch": 0.933219100870271, + "flos": 22642100251680.0, + "grad_norm": 2.3297608024378658, + "learning_rate": 4.654260524000797e-08, + "loss": 0.7491, + "num_input_tokens_seen": 336025340, + "step": 10348 + }, + { + "epoch": 0.9333092843937413, + "flos": 23550895544640.0, + "grad_norm": 1.61231385568638, + "learning_rate": 4.6417387928719076e-08, + "loss": 0.7284, + "num_input_tokens_seen": 336054480, + "step": 10349 + }, + { + "epoch": 0.9333994679172115, + "flos": 25483952605920.0, + "grad_norm": 1.6846072516330441, + "learning_rate": 4.629233730787052e-08, + "loss": 0.6146, + "num_input_tokens_seen": 336084115, + "step": 10350 + }, + { + "epoch": 0.9334896514406817, + "flos": 19545260378400.0, + "grad_norm": 2.751900709639894, + "learning_rate": 4.616745338813266e-08, + "loss": 0.7036, + "num_input_tokens_seen": 336110710, + "step": 10351 + }, + { + "epoch": 0.9335798349641521, + "flos": 21984315313440.0, + "grad_norm": 1.621687520681431, + "learning_rate": 4.6042736180160744e-08, + "loss": 0.757, + "num_input_tokens_seen": 336140250, + "step": 10352 + }, + { + "epoch": 0.9336700184876223, + "flos": 19690988772000.0, + "grad_norm": 1.8600918325703957, + "learning_rate": 4.591818569459671e-08, + "loss": 0.8586, + "num_input_tokens_seen": 336168375, + "step": 10353 + }, + { + "epoch": 0.9337602020110926, + "flos": 21694493995680.0, + "grad_norm": 4.476188604109043, + "learning_rate": 4.5793801942067614e-08, + "loss": 0.7614, + "num_input_tokens_seen": 336197430, + "step": 10354 + }, + { + "epoch": 0.9338503855345628, + "flos": 29378098461120.0, + "grad_norm": 1.9107937675313742, + "learning_rate": 4.566958493318673e-08, + "loss": 0.61, + "num_input_tokens_seen": 336225940, + "step": 10355 + }, + { + "epoch": 0.9339405690580331, + "flos": 67994481562080.0, + "grad_norm": 0.6220249416986793, + "learning_rate": 4.554553467855316e-08, + "loss": 0.5986, + "num_input_tokens_seen": 336314730, + "step": 10356 + }, + { + "epoch": 0.9340307525815034, + "flos": 23772127188000.0, + "grad_norm": 2.5613534735000822, + "learning_rate": 4.5421651188751074e-08, + "loss": 0.7576, + "num_input_tokens_seen": 336342825, + "step": 10357 + }, + { + "epoch": 0.9341209361049736, + "flos": 22168705991040.0, + "grad_norm": 1.755988840797611, + "learning_rate": 4.529793447435137e-08, + "loss": 0.7534, + "num_input_tokens_seen": 336371675, + "step": 10358 + }, + { + "epoch": 0.9342111196284438, + "flos": 19472544860640.0, + "grad_norm": 1.7498634830087405, + "learning_rate": 4.5174384545909824e-08, + "loss": 0.741, + "num_input_tokens_seen": 336400155, + "step": 10359 + }, + { + "epoch": 0.9343013031519142, + "flos": 26175716438400.0, + "grad_norm": 2.309047906034182, + "learning_rate": 4.505100141396867e-08, + "loss": 0.7768, + "num_input_tokens_seen": 336426035, + "step": 10360 + }, + { + "epoch": 0.9343914866753844, + "flos": 13752519563040.0, + "grad_norm": 2.4028050917859627, + "learning_rate": 4.492778508905548e-08, + "loss": 0.7727, + "num_input_tokens_seen": 336451990, + "step": 10361 + }, + { + "epoch": 0.9344816701988546, + "flos": 33899692996800.0, + "grad_norm": 2.0259017961431316, + "learning_rate": 4.480473558168385e-08, + "loss": 0.6788, + "num_input_tokens_seen": 336481710, + "step": 10362 + }, + { + "epoch": 0.934571853722325, + "flos": 26208468730560.0, + "grad_norm": 3.3472971417566746, + "learning_rate": 4.4681852902352936e-08, + "loss": 0.6449, + "num_input_tokens_seen": 336511250, + "step": 10363 + }, + { + "epoch": 0.9346620372457952, + "flos": 24463853850720.0, + "grad_norm": 2.2848729667159926, + "learning_rate": 4.455913706154812e-08, + "loss": 0.654, + "num_input_tokens_seen": 336539370, + "step": 10364 + }, + { + "epoch": 0.9347522207692655, + "flos": 22350234597120.0, + "grad_norm": 2.4893913248203505, + "learning_rate": 4.443658806973949e-08, + "loss": 0.7659, + "num_input_tokens_seen": 336566585, + "step": 10365 + }, + { + "epoch": 0.9348424042927357, + "flos": 62092778979360.0, + "grad_norm": 0.661689467609781, + "learning_rate": 4.431420593738444e-08, + "loss": 0.6002, + "num_input_tokens_seen": 336657390, + "step": 10366 + }, + { + "epoch": 0.934932587816206, + "flos": 21948440761440.0, + "grad_norm": 1.8904486960184432, + "learning_rate": 4.419199067492485e-08, + "loss": 0.6524, + "num_input_tokens_seen": 336683915, + "step": 10367 + }, + { + "epoch": 0.9350227713396763, + "flos": 21840705596160.0, + "grad_norm": 1.7952382469924029, + "learning_rate": 4.4069942292788596e-08, + "loss": 0.7496, + "num_input_tokens_seen": 336712435, + "step": 10368 + }, + { + "epoch": 0.9351129548631465, + "flos": 29268839335680.0, + "grad_norm": 1.9005655049448902, + "learning_rate": 4.39480608013898e-08, + "loss": 0.7462, + "num_input_tokens_seen": 336744545, + "step": 10369 + }, + { + "epoch": 0.9352031383866167, + "flos": 30911740551360.0, + "grad_norm": 1.8387416770430314, + "learning_rate": 4.3826346211128126e-08, + "loss": 0.8312, + "num_input_tokens_seen": 336773910, + "step": 10370 + }, + { + "epoch": 0.9352933219100871, + "flos": 20343978811200.0, + "grad_norm": 1.5609706260366083, + "learning_rate": 4.370479853238884e-08, + "loss": 0.7444, + "num_input_tokens_seen": 336802645, + "step": 10371 + }, + { + "epoch": 0.9353835054335573, + "flos": 22022531560320.0, + "grad_norm": 2.5349796274784846, + "learning_rate": 4.3583417775542756e-08, + "loss": 0.7055, + "num_input_tokens_seen": 336831835, + "step": 10372 + }, + { + "epoch": 0.9354736889570275, + "flos": 36123723638400.0, + "grad_norm": 1.6054102383460218, + "learning_rate": 4.3462203950947575e-08, + "loss": 0.6944, + "num_input_tokens_seen": 336866585, + "step": 10373 + }, + { + "epoch": 0.9355638724804978, + "flos": 23407285827360.0, + "grad_norm": 1.6741332729661835, + "learning_rate": 4.3341157068944814e-08, + "loss": 0.7259, + "num_input_tokens_seen": 336894260, + "step": 10374 + }, + { + "epoch": 0.9356540560039681, + "flos": 27372511730880.0, + "grad_norm": 1.9881210412493007, + "learning_rate": 4.322027713986376e-08, + "loss": 0.7546, + "num_input_tokens_seen": 336922730, + "step": 10375 + }, + { + "epoch": 0.9357442395274383, + "flos": 24898846015680.0, + "grad_norm": 2.709791221459267, + "learning_rate": 4.309956417401816e-08, + "loss": 0.8103, + "num_input_tokens_seen": 336947925, + "step": 10376 + }, + { + "epoch": 0.9358344230509086, + "flos": 29161252849440.0, + "grad_norm": 1.8624983768328052, + "learning_rate": 4.297901818170801e-08, + "loss": 0.7105, + "num_input_tokens_seen": 336980095, + "step": 10377 + }, + { + "epoch": 0.9359246065743788, + "flos": 30403661171040.0, + "grad_norm": 1.8280359230184584, + "learning_rate": 4.285863917321886e-08, + "loss": 0.6749, + "num_input_tokens_seen": 337013240, + "step": 10378 + }, + { + "epoch": 0.9360147900978492, + "flos": 21434079691680.0, + "grad_norm": 1.753685816312898, + "learning_rate": 4.2738427158822253e-08, + "loss": 0.6765, + "num_input_tokens_seen": 337040650, + "step": 10379 + }, + { + "epoch": 0.9361049736213194, + "flos": 16411765388160.0, + "grad_norm": 2.215395688779169, + "learning_rate": 4.261838214877511e-08, + "loss": 0.7497, + "num_input_tokens_seen": 337067525, + "step": 10380 + }, + { + "epoch": 0.9361951571447896, + "flos": 28469303168160.0, + "grad_norm": 2.303964585542036, + "learning_rate": 4.249850415332079e-08, + "loss": 0.7468, + "num_input_tokens_seen": 337097500, + "step": 10381 + }, + { + "epoch": 0.9362853406682599, + "flos": 25917532320000.0, + "grad_norm": 1.890706525573913, + "learning_rate": 4.237879318268756e-08, + "loss": 0.8007, + "num_input_tokens_seen": 337127135, + "step": 10382 + }, + { + "epoch": 0.9363755241917302, + "flos": 25593620598720.0, + "grad_norm": 1.9407521302971824, + "learning_rate": 4.225924924708968e-08, + "loss": 0.6722, + "num_input_tokens_seen": 337157040, + "step": 10383 + }, + { + "epoch": 0.9364657077152004, + "flos": 29563901589600.0, + "grad_norm": 2.138436751309088, + "learning_rate": 4.2139872356727665e-08, + "loss": 0.7491, + "num_input_tokens_seen": 337186525, + "step": 10384 + }, + { + "epoch": 0.9365558912386707, + "flos": 27631141886400.0, + "grad_norm": 1.7071609273228914, + "learning_rate": 4.202066252178738e-08, + "loss": 0.7044, + "num_input_tokens_seen": 337217670, + "step": 10385 + }, + { + "epoch": 0.936646074762141, + "flos": 28617001559040.0, + "grad_norm": 1.775767941976778, + "learning_rate": 4.1901619752440445e-08, + "loss": 0.8322, + "num_input_tokens_seen": 337246800, + "step": 10386 + }, + { + "epoch": 0.9367362582856112, + "flos": 23474723239200.0, + "grad_norm": 1.9388550357662604, + "learning_rate": 4.178274405884363e-08, + "loss": 0.7401, + "num_input_tokens_seen": 337273775, + "step": 10387 + }, + { + "epoch": 0.9368264418090815, + "flos": 51277878273120.0, + "grad_norm": 1.8275927559500098, + "learning_rate": 4.166403545114105e-08, + "loss": 0.7498, + "num_input_tokens_seen": 337306810, + "step": 10388 + }, + { + "epoch": 0.9369166253325517, + "flos": 25119259924320.0, + "grad_norm": 2.383413570449479, + "learning_rate": 4.154549393946083e-08, + "loss": 0.7944, + "num_input_tokens_seen": 337334535, + "step": 10389 + }, + { + "epoch": 0.937006808856022, + "flos": 21768621964320.0, + "grad_norm": 2.1191124980132225, + "learning_rate": 4.14271195339182e-08, + "loss": 0.767, + "num_input_tokens_seen": 337364015, + "step": 10390 + }, + { + "epoch": 0.9370969923794923, + "flos": 36047588502720.0, + "grad_norm": 1.6089980785863391, + "learning_rate": 4.1308912244613084e-08, + "loss": 0.7074, + "num_input_tokens_seen": 337397320, + "step": 10391 + }, + { + "epoch": 0.9371871759029625, + "flos": 23514203257920.0, + "grad_norm": 1.5621838822068235, + "learning_rate": 4.1190872081631636e-08, + "loss": 0.698, + "num_input_tokens_seen": 337426515, + "step": 10392 + }, + { + "epoch": 0.9372773594264328, + "flos": 18561482212320.0, + "grad_norm": 3.8714010692136376, + "learning_rate": 4.107299905504558e-08, + "loss": 0.8076, + "num_input_tokens_seen": 337452230, + "step": 10393 + }, + { + "epoch": 0.9373675429499031, + "flos": 30074434174080.0, + "grad_norm": 1.7580053605137884, + "learning_rate": 4.095529317491286e-08, + "loss": 0.7335, + "num_input_tokens_seen": 337482840, + "step": 10394 + }, + { + "epoch": 0.9374577264733733, + "flos": 19106737086240.0, + "grad_norm": 2.1939444300900144, + "learning_rate": 4.0837754451276575e-08, + "loss": 0.7443, + "num_input_tokens_seen": 337510850, + "step": 10395 + }, + { + "epoch": 0.9375479099968436, + "flos": 25883404746720.0, + "grad_norm": 1.5614434293918298, + "learning_rate": 4.072038289416557e-08, + "loss": 0.7269, + "num_input_tokens_seen": 337543260, + "step": 10396 + }, + { + "epoch": 0.9376380935203138, + "flos": 22677417257280.0, + "grad_norm": 2.8604639199936517, + "learning_rate": 4.0603178513595185e-08, + "loss": 0.7452, + "num_input_tokens_seen": 337572855, + "step": 10397 + }, + { + "epoch": 0.9377282770437841, + "flos": 16776420900000.0, + "grad_norm": 3.105517577741137, + "learning_rate": 4.0486141319565624e-08, + "loss": 0.7489, + "num_input_tokens_seen": 337599315, + "step": 10398 + }, + { + "epoch": 0.9378184605672544, + "flos": 24864718442400.0, + "grad_norm": 2.0289239329616646, + "learning_rate": 4.0369271322062916e-08, + "loss": 0.6847, + "num_input_tokens_seen": 337625305, + "step": 10399 + }, + { + "epoch": 0.9379086440907246, + "flos": 17652017863680.0, + "grad_norm": 1.7983904874673933, + "learning_rate": 4.0252568531059295e-08, + "loss": 0.759, + "num_input_tokens_seen": 337653615, + "step": 10400 + }, + { + "epoch": 0.9379988276141948, + "flos": 20930200494240.0, + "grad_norm": 2.1524350238700407, + "learning_rate": 4.013603295651235e-08, + "loss": 0.7424, + "num_input_tokens_seen": 337682140, + "step": 10401 + }, + { + "epoch": 0.9380890111376652, + "flos": 20019063506400.0, + "grad_norm": 2.1462859876900353, + "learning_rate": 4.001966460836592e-08, + "loss": 0.7681, + "num_input_tokens_seen": 337708540, + "step": 10402 + }, + { + "epoch": 0.9381791946611354, + "flos": 24318348475680.0, + "grad_norm": 3.521069325018943, + "learning_rate": 3.990346349654894e-08, + "loss": 0.8143, + "num_input_tokens_seen": 337737770, + "step": 10403 + }, + { + "epoch": 0.9382693781846057, + "flos": 30872855248800.0, + "grad_norm": 1.726714390766543, + "learning_rate": 3.9787429630975924e-08, + "loss": 0.7173, + "num_input_tokens_seen": 337768330, + "step": 10404 + }, + { + "epoch": 0.9383595617080759, + "flos": 20019175015680.0, + "grad_norm": 2.507203310360698, + "learning_rate": 3.967156302154828e-08, + "loss": 0.7275, + "num_input_tokens_seen": 337797275, + "step": 10405 + }, + { + "epoch": 0.9384497452315462, + "flos": 26139841886400.0, + "grad_norm": 2.372404788712573, + "learning_rate": 3.955586367815189e-08, + "loss": 0.6966, + "num_input_tokens_seen": 337826645, + "step": 10406 + }, + { + "epoch": 0.9385399287550165, + "flos": 26832534962880.0, + "grad_norm": 1.9040325587798648, + "learning_rate": 3.944033161065907e-08, + "loss": 0.8105, + "num_input_tokens_seen": 337856510, + "step": 10407 + }, + { + "epoch": 0.9386301122784867, + "flos": 36775412736000.0, + "grad_norm": 1.8412692304281602, + "learning_rate": 3.93249668289275e-08, + "loss": 0.6263, + "num_input_tokens_seen": 337887590, + "step": 10408 + }, + { + "epoch": 0.9387202958019569, + "flos": 23589892356480.0, + "grad_norm": 2.7349195537556743, + "learning_rate": 3.920976934280063e-08, + "loss": 0.6668, + "num_input_tokens_seen": 337917795, + "step": 10409 + }, + { + "epoch": 0.9388104793254273, + "flos": 19654445164320.0, + "grad_norm": 1.7891260660531199, + "learning_rate": 3.909473916210815e-08, + "loss": 0.8271, + "num_input_tokens_seen": 337944670, + "step": 10410 + }, + { + "epoch": 0.9389006628488975, + "flos": 17322976715520.0, + "grad_norm": 2.4899251110044744, + "learning_rate": 3.897987629666488e-08, + "loss": 0.703, + "num_input_tokens_seen": 337971030, + "step": 10411 + }, + { + "epoch": 0.9389908463723677, + "flos": 23807741551680.0, + "grad_norm": 1.6579679540291254, + "learning_rate": 3.886518075627143e-08, + "loss": 0.8145, + "num_input_tokens_seen": 338002745, + "step": 10412 + }, + { + "epoch": 0.9390810298958381, + "flos": 21840222389280.0, + "grad_norm": 2.5632737084916966, + "learning_rate": 3.875065255071419e-08, + "loss": 0.7408, + "num_input_tokens_seen": 338027405, + "step": 10413 + }, + { + "epoch": 0.9391712134193083, + "flos": 21221248414080.0, + "grad_norm": 1.8106694020576959, + "learning_rate": 3.863629168976579e-08, + "loss": 0.7105, + "num_input_tokens_seen": 338054975, + "step": 10414 + }, + { + "epoch": 0.9392613969427785, + "flos": 19326890806560.0, + "grad_norm": 1.7563898729983243, + "learning_rate": 3.852209818318375e-08, + "loss": 0.798, + "num_input_tokens_seen": 338082415, + "step": 10415 + }, + { + "epoch": 0.9393515804662488, + "flos": 26358360137280.0, + "grad_norm": 2.0883316987262894, + "learning_rate": 3.840807204071161e-08, + "loss": 0.6484, + "num_input_tokens_seen": 338112570, + "step": 10416 + }, + { + "epoch": 0.9394417639897191, + "flos": 28361642342400.0, + "grad_norm": 1.6661832664731018, + "learning_rate": 3.829421327207894e-08, + "loss": 0.7272, + "num_input_tokens_seen": 338140370, + "step": 10417 + }, + { + "epoch": 0.9395319475131894, + "flos": 20927152573920.0, + "grad_norm": 1.9790360725550202, + "learning_rate": 3.8180521887000825e-08, + "loss": 0.6929, + "num_input_tokens_seen": 338170220, + "step": 10418 + }, + { + "epoch": 0.9396221310366596, + "flos": 68289320797440.0, + "grad_norm": 0.5876508458240725, + "learning_rate": 3.806699789517775e-08, + "loss": 0.6094, + "num_input_tokens_seen": 338264105, + "step": 10419 + }, + { + "epoch": 0.9397123145601298, + "flos": 20456062838400.0, + "grad_norm": 1.6900696439895706, + "learning_rate": 3.7953641306296635e-08, + "loss": 0.7262, + "num_input_tokens_seen": 338291340, + "step": 10420 + }, + { + "epoch": 0.9398024980836002, + "flos": 21621072252480.0, + "grad_norm": 2.8058061414382354, + "learning_rate": 3.784045213002951e-08, + "loss": 0.5846, + "num_input_tokens_seen": 338317915, + "step": 10421 + }, + { + "epoch": 0.9398926816070704, + "flos": 22275920779680.0, + "grad_norm": 1.983663416356893, + "learning_rate": 3.7727430376033986e-08, + "loss": 0.7564, + "num_input_tokens_seen": 338345635, + "step": 10422 + }, + { + "epoch": 0.9399828651305406, + "flos": 19472433351360.0, + "grad_norm": 2.262284756532898, + "learning_rate": 3.7614576053954126e-08, + "loss": 0.7448, + "num_input_tokens_seen": 338373295, + "step": 10423 + }, + { + "epoch": 0.9400730486540109, + "flos": 35063661657600.0, + "grad_norm": 1.9888195815462735, + "learning_rate": 3.75018891734189e-08, + "loss": 0.739, + "num_input_tokens_seen": 338403830, + "step": 10424 + }, + { + "epoch": 0.9401632321774812, + "flos": 31851615497280.0, + "grad_norm": 2.380040191343089, + "learning_rate": 3.738936974404372e-08, + "loss": 0.7888, + "num_input_tokens_seen": 338430765, + "step": 10425 + }, + { + "epoch": 0.9402534157009514, + "flos": 19399569154560.0, + "grad_norm": 1.3639781668025923, + "learning_rate": 3.7277017775429354e-08, + "loss": 0.8063, + "num_input_tokens_seen": 338459680, + "step": 10426 + }, + { + "epoch": 0.9403435992244217, + "flos": 17760830952000.0, + "grad_norm": 1.6237592361810078, + "learning_rate": 3.7164833277162136e-08, + "loss": 0.7655, + "num_input_tokens_seen": 338486540, + "step": 10427 + }, + { + "epoch": 0.9404337827478919, + "flos": 25229596972800.0, + "grad_norm": 1.4686812526765298, + "learning_rate": 3.705281625881418e-08, + "loss": 0.7237, + "num_input_tokens_seen": 338518055, + "step": 10428 + }, + { + "epoch": 0.9405239662713623, + "flos": 44062278453120.0, + "grad_norm": 2.1102135016009322, + "learning_rate": 3.694096672994362e-08, + "loss": 0.6511, + "num_input_tokens_seen": 338552875, + "step": 10429 + }, + { + "epoch": 0.9406141497948325, + "flos": 22491911486880.0, + "grad_norm": 1.8368201548671808, + "learning_rate": 3.682928470009394e-08, + "loss": 0.7974, + "num_input_tokens_seen": 338580060, + "step": 10430 + }, + { + "epoch": 0.9407043333183027, + "flos": 14845519684800.0, + "grad_norm": 2.4075281290884836, + "learning_rate": 3.6717770178794406e-08, + "loss": 0.6982, + "num_input_tokens_seen": 338602745, + "step": 10431 + }, + { + "epoch": 0.940794516841773, + "flos": 25190079784320.0, + "grad_norm": 4.65913496313815, + "learning_rate": 3.6606423175560287e-08, + "loss": 0.8463, + "num_input_tokens_seen": 338633445, + "step": 10432 + }, + { + "epoch": 0.9408847003652433, + "flos": 68815836378720.0, + "grad_norm": 0.7853976964978466, + "learning_rate": 3.649524369989221e-08, + "loss": 0.5447, + "num_input_tokens_seen": 338729240, + "step": 10433 + }, + { + "epoch": 0.9409748838887135, + "flos": 22930211760480.0, + "grad_norm": 1.8278874299970809, + "learning_rate": 3.638423176127636e-08, + "loss": 0.775, + "num_input_tokens_seen": 338758540, + "step": 10434 + }, + { + "epoch": 0.9410650674121838, + "flos": 24135518928000.0, + "grad_norm": 1.495047340281443, + "learning_rate": 3.6273387369185396e-08, + "loss": 0.7783, + "num_input_tokens_seen": 338787675, + "step": 10435 + }, + { + "epoch": 0.9411552509356541, + "flos": 31494505446720.0, + "grad_norm": 1.948723311519837, + "learning_rate": 3.616271053307685e-08, + "loss": 0.6885, + "num_input_tokens_seen": 338820425, + "step": 10436 + }, + { + "epoch": 0.9412454344591243, + "flos": 22714927278720.0, + "grad_norm": 2.128180311831389, + "learning_rate": 3.6052201262394275e-08, + "loss": 0.7635, + "num_input_tokens_seen": 338850500, + "step": 10437 + }, + { + "epoch": 0.9413356179825946, + "flos": 24536532198720.0, + "grad_norm": 1.8437005375838527, + "learning_rate": 3.5941859566566816e-08, + "loss": 0.798, + "num_input_tokens_seen": 338880245, + "step": 10438 + }, + { + "epoch": 0.9414258015060648, + "flos": 57023290516800.0, + "grad_norm": 0.6927483028458532, + "learning_rate": 3.583168545500981e-08, + "loss": 0.5666, + "num_input_tokens_seen": 338969715, + "step": 10439 + }, + { + "epoch": 0.9415159850295352, + "flos": 20638929555840.0, + "grad_norm": 2.044107677646885, + "learning_rate": 3.5721678937123746e-08, + "loss": 0.6883, + "num_input_tokens_seen": 338997570, + "step": 10440 + }, + { + "epoch": 0.9416061685530054, + "flos": 25594587012480.0, + "grad_norm": 1.9229151074481345, + "learning_rate": 3.561184002229467e-08, + "loss": 0.8261, + "num_input_tokens_seen": 339026170, + "step": 10441 + }, + { + "epoch": 0.9416963520764756, + "flos": 30074620022880.0, + "grad_norm": 2.0898106652608948, + "learning_rate": 3.550216871989531e-08, + "loss": 0.6898, + "num_input_tokens_seen": 339056605, + "step": 10442 + }, + { + "epoch": 0.9417865355999459, + "flos": 70415429090400.0, + "grad_norm": 0.6385620209362708, + "learning_rate": 3.539266503928262e-08, + "loss": 0.6454, + "num_input_tokens_seen": 339157000, + "step": 10443 + }, + { + "epoch": 0.9418767191234162, + "flos": 23516359104000.0, + "grad_norm": 1.6947568343165837, + "learning_rate": 3.528332898980091e-08, + "loss": 0.7537, + "num_input_tokens_seen": 339186675, + "step": 10444 + }, + { + "epoch": 0.9419669026468864, + "flos": 38490880790400.0, + "grad_norm": 2.467086521278267, + "learning_rate": 3.517416058077849e-08, + "loss": 0.7098, + "num_input_tokens_seen": 339218520, + "step": 10445 + }, + { + "epoch": 0.9420570861703567, + "flos": 18889779965280.0, + "grad_norm": 1.9688813907436502, + "learning_rate": 3.506515982153102e-08, + "loss": 0.6455, + "num_input_tokens_seen": 339245695, + "step": 10446 + }, + { + "epoch": 0.9421472696938269, + "flos": 30183544620480.0, + "grad_norm": 1.9374225354859769, + "learning_rate": 3.495632672135862e-08, + "loss": 0.7477, + "num_input_tokens_seen": 339276690, + "step": 10447 + }, + { + "epoch": 0.9422374532172972, + "flos": 26248246107360.0, + "grad_norm": 1.5591705708103991, + "learning_rate": 3.4847661289547417e-08, + "loss": 0.7716, + "num_input_tokens_seen": 339308845, + "step": 10448 + }, + { + "epoch": 0.9423276367407675, + "flos": 19836828674880.0, + "grad_norm": 2.4374497237277906, + "learning_rate": 3.473916353536932e-08, + "loss": 0.6993, + "num_input_tokens_seen": 339336550, + "step": 10449 + }, + { + "epoch": 0.9424178202642377, + "flos": 21510028978560.0, + "grad_norm": 1.8091935219946664, + "learning_rate": 3.463083346808249e-08, + "loss": 0.7089, + "num_input_tokens_seen": 339365435, + "step": 10450 + }, + { + "epoch": 0.9425080037877079, + "flos": 23115866209920.0, + "grad_norm": 1.7358599654907942, + "learning_rate": 3.452267109692975e-08, + "loss": 0.6387, + "num_input_tokens_seen": 339394115, + "step": 10451 + }, + { + "epoch": 0.9425981873111783, + "flos": 21834238057920.0, + "grad_norm": 3.3970255406912804, + "learning_rate": 3.441467643114016e-08, + "loss": 0.7434, + "num_input_tokens_seen": 339420425, + "step": 10452 + }, + { + "epoch": 0.9426883708346485, + "flos": 66687981107040.0, + "grad_norm": 0.5859988371215843, + "learning_rate": 3.430684947992857e-08, + "loss": 0.5565, + "num_input_tokens_seen": 339506505, + "step": 10453 + }, + { + "epoch": 0.9427785543581187, + "flos": 21876096941280.0, + "grad_norm": 2.007867122159488, + "learning_rate": 3.419919025249518e-08, + "loss": 0.7579, + "num_input_tokens_seen": 339535825, + "step": 10454 + }, + { + "epoch": 0.942868737881589, + "flos": 22461315040800.0, + "grad_norm": 1.7926207892502317, + "learning_rate": 3.40916987580262e-08, + "loss": 0.7637, + "num_input_tokens_seen": 339566200, + "step": 10455 + }, + { + "epoch": 0.9429589214050593, + "flos": 21836839941120.0, + "grad_norm": 1.8230533512279172, + "learning_rate": 3.398437500569362e-08, + "loss": 0.7207, + "num_input_tokens_seen": 339593890, + "step": 10456 + }, + { + "epoch": 0.9430491049285296, + "flos": 61770508461120.0, + "grad_norm": 2.0846945956565013, + "learning_rate": 3.3877219004654347e-08, + "loss": 0.6236, + "num_input_tokens_seen": 339626525, + "step": 10457 + }, + { + "epoch": 0.9431392884519998, + "flos": 22569273224640.0, + "grad_norm": 3.1114174733620175, + "learning_rate": 3.3770230764051946e-08, + "loss": 0.776, + "num_input_tokens_seen": 339654650, + "step": 10458 + }, + { + "epoch": 0.9432294719754701, + "flos": 25335101952480.0, + "grad_norm": 2.217589598762593, + "learning_rate": 3.366341029301534e-08, + "loss": 0.7604, + "num_input_tokens_seen": 339682525, + "step": 10459 + }, + { + "epoch": 0.9433196554989404, + "flos": 25483766757120.0, + "grad_norm": 2.53280949530638, + "learning_rate": 3.355675760065857e-08, + "loss": 0.7484, + "num_input_tokens_seen": 339709875, + "step": 10460 + }, + { + "epoch": 0.9434098390224106, + "flos": 62433194074080.0, + "grad_norm": 0.5778571358868833, + "learning_rate": 3.345027269608236e-08, + "loss": 0.525, + "num_input_tokens_seen": 339808940, + "step": 10461 + }, + { + "epoch": 0.9435000225458808, + "flos": 70542312415680.0, + "grad_norm": 0.6636037925210023, + "learning_rate": 3.334395558837211e-08, + "loss": 0.5915, + "num_input_tokens_seen": 339896430, + "step": 10462 + }, + { + "epoch": 0.9435902060693512, + "flos": 21508096151040.0, + "grad_norm": 1.6135380058008986, + "learning_rate": 3.3237806286599667e-08, + "loss": 0.6536, + "num_input_tokens_seen": 339924110, + "step": 10463 + }, + { + "epoch": 0.9436803895928214, + "flos": 20783914554240.0, + "grad_norm": 1.713721740289321, + "learning_rate": 3.313182479982224e-08, + "loss": 0.7292, + "num_input_tokens_seen": 339951910, + "step": 10464 + }, + { + "epoch": 0.9437705731162916, + "flos": 28434655218240.0, + "grad_norm": 1.8892884377282992, + "learning_rate": 3.302601113708259e-08, + "loss": 0.8089, + "num_input_tokens_seen": 339979910, + "step": 10465 + }, + { + "epoch": 0.9438607566397619, + "flos": 29559292539360.0, + "grad_norm": 1.3735215735280908, + "learning_rate": 3.292036530740972e-08, + "loss": 0.7877, + "num_input_tokens_seen": 340010485, + "step": 10466 + }, + { + "epoch": 0.9439509401632322, + "flos": 24936987923040.0, + "grad_norm": 2.1770995168373397, + "learning_rate": 3.2814887319817294e-08, + "loss": 0.7339, + "num_input_tokens_seen": 340039070, + "step": 10467 + }, + { + "epoch": 0.9440411236867025, + "flos": 21439915344000.0, + "grad_norm": 1.7874480713457295, + "learning_rate": 3.270957718330591e-08, + "loss": 0.7827, + "num_input_tokens_seen": 340067200, + "step": 10468 + }, + { + "epoch": 0.9441313072101727, + "flos": 21439357797600.0, + "grad_norm": 2.124403477135874, + "learning_rate": 3.260443490686082e-08, + "loss": 0.7854, + "num_input_tokens_seen": 340096735, + "step": 10469 + }, + { + "epoch": 0.9442214907336429, + "flos": 19545706415520.0, + "grad_norm": 1.9448963624070437, + "learning_rate": 3.249946049945351e-08, + "loss": 0.7572, + "num_input_tokens_seen": 340124100, + "step": 10470 + }, + { + "epoch": 0.9443116742571133, + "flos": 21803009725920.0, + "grad_norm": 2.0276043380814315, + "learning_rate": 3.239465397004082e-08, + "loss": 0.7567, + "num_input_tokens_seen": 340150960, + "step": 10471 + }, + { + "epoch": 0.9444018577805835, + "flos": 29637286163040.0, + "grad_norm": 3.1264176776979693, + "learning_rate": 3.229001532756559e-08, + "loss": 0.7022, + "num_input_tokens_seen": 340183370, + "step": 10472 + }, + { + "epoch": 0.9444920413040537, + "flos": 20637888802560.0, + "grad_norm": 2.2274499028284778, + "learning_rate": 3.218554458095602e-08, + "loss": 0.738, + "num_input_tokens_seen": 340211335, + "step": 10473 + }, + { + "epoch": 0.944582224827524, + "flos": 12076680206400.0, + "grad_norm": 2.5582695442555843, + "learning_rate": 3.20812417391263e-08, + "loss": 0.7492, + "num_input_tokens_seen": 340237825, + "step": 10474 + }, + { + "epoch": 0.9446724083509943, + "flos": 23077389774720.0, + "grad_norm": 1.4361622432129897, + "learning_rate": 3.1977106810975764e-08, + "loss": 0.7353, + "num_input_tokens_seen": 340269660, + "step": 10475 + }, + { + "epoch": 0.9447625918744645, + "flos": 54315866949120.0, + "grad_norm": 0.7835653328425601, + "learning_rate": 3.187313980539042e-08, + "loss": 0.6209, + "num_input_tokens_seen": 340338870, + "step": 10476 + }, + { + "epoch": 0.9448527753979348, + "flos": 22751210698080.0, + "grad_norm": 1.7127725871529864, + "learning_rate": 3.176934073124071e-08, + "loss": 0.823, + "num_input_tokens_seen": 340370305, + "step": 10477 + }, + { + "epoch": 0.944942958921405, + "flos": 20967227308800.0, + "grad_norm": 2.1936865381696133, + "learning_rate": 3.166570959738357e-08, + "loss": 0.7397, + "num_input_tokens_seen": 340398655, + "step": 10478 + }, + { + "epoch": 0.9450331424448754, + "flos": 41004546900960.0, + "grad_norm": 1.9064577958712736, + "learning_rate": 3.1562246412661476e-08, + "loss": 0.6795, + "num_input_tokens_seen": 340428240, + "step": 10479 + }, + { + "epoch": 0.9451233259683456, + "flos": 22606411548480.0, + "grad_norm": 1.7716004523927202, + "learning_rate": 3.145895118590225e-08, + "loss": 0.7744, + "num_input_tokens_seen": 340457710, + "step": 10480 + }, + { + "epoch": 0.9452135094918158, + "flos": 32110208483040.0, + "grad_norm": 1.7739026640129418, + "learning_rate": 3.135582392591996e-08, + "loss": 0.6525, + "num_input_tokens_seen": 340486985, + "step": 10481 + }, + { + "epoch": 0.9453036930152862, + "flos": 60248680620960.0, + "grad_norm": 0.5906035222583741, + "learning_rate": 3.125286464151333e-08, + "loss": 0.535, + "num_input_tokens_seen": 340584035, + "step": 10482 + }, + { + "epoch": 0.9453938765387564, + "flos": 24208531803840.0, + "grad_norm": 2.6839184516024877, + "learning_rate": 3.115007334146824e-08, + "loss": 0.7151, + "num_input_tokens_seen": 340611375, + "step": 10483 + }, + { + "epoch": 0.9454840600622266, + "flos": 59183229024960.0, + "grad_norm": 0.630492206879265, + "learning_rate": 3.104745003455478e-08, + "loss": 0.5564, + "num_input_tokens_seen": 340706240, + "step": 10484 + }, + { + "epoch": 0.9455742435856969, + "flos": 21293480724960.0, + "grad_norm": 1.8131837541698574, + "learning_rate": 3.094499472952972e-08, + "loss": 0.765, + "num_input_tokens_seen": 340736400, + "step": 10485 + }, + { + "epoch": 0.9456644271091672, + "flos": 24715942128480.0, + "grad_norm": 9.262785864162684, + "learning_rate": 3.084270743513495e-08, + "loss": 0.7099, + "num_input_tokens_seen": 340766285, + "step": 10486 + }, + { + "epoch": 0.9457546106326374, + "flos": 25374210273600.0, + "grad_norm": 2.498683964523211, + "learning_rate": 3.074058816009817e-08, + "loss": 0.6301, + "num_input_tokens_seen": 340794925, + "step": 10487 + }, + { + "epoch": 0.9458447941561077, + "flos": 25156509757440.0, + "grad_norm": 1.9081982521768004, + "learning_rate": 3.063863691313284e-08, + "loss": 0.691, + "num_input_tokens_seen": 340823905, + "step": 10488 + }, + { + "epoch": 0.9459349776795779, + "flos": 36045060959040.0, + "grad_norm": 1.874623769524823, + "learning_rate": 3.0536853702937794e-08, + "loss": 0.6811, + "num_input_tokens_seen": 340854765, + "step": 10489 + }, + { + "epoch": 0.9460251612030482, + "flos": 22568901527040.0, + "grad_norm": 2.3930690464144537, + "learning_rate": 3.043523853819807e-08, + "loss": 0.7443, + "num_input_tokens_seen": 340881790, + "step": 10490 + }, + { + "epoch": 0.9461153447265185, + "flos": 17095872250080.0, + "grad_norm": 2.3788056250526703, + "learning_rate": 3.0333791427583855e-08, + "loss": 0.7661, + "num_input_tokens_seen": 340907125, + "step": 10491 + }, + { + "epoch": 0.9462055282499887, + "flos": 19941144222240.0, + "grad_norm": 1.6834764196309644, + "learning_rate": 3.023251237975111e-08, + "loss": 0.8647, + "num_input_tokens_seen": 340934535, + "step": 10492 + }, + { + "epoch": 0.946295711773459, + "flos": 66951592010400.0, + "grad_norm": 0.753442532725477, + "learning_rate": 3.0131401403341584e-08, + "loss": 0.5931, + "num_input_tokens_seen": 341030860, + "step": 10493 + }, + { + "epoch": 0.9463858952969293, + "flos": 24064178691360.0, + "grad_norm": 1.587707884385965, + "learning_rate": 3.00304585069826e-08, + "loss": 0.7191, + "num_input_tokens_seen": 341061475, + "step": 10494 + }, + { + "epoch": 0.9464760788203995, + "flos": 20783951724000.0, + "grad_norm": 3.1751597266682436, + "learning_rate": 2.992968369928728e-08, + "loss": 0.7305, + "num_input_tokens_seen": 341090450, + "step": 10495 + }, + { + "epoch": 0.9465662623438698, + "flos": 18812195208960.0, + "grad_norm": 2.021983091598736, + "learning_rate": 2.982907698885429e-08, + "loss": 0.7575, + "num_input_tokens_seen": 341116265, + "step": 10496 + }, + { + "epoch": 0.94665644586734, + "flos": 23005380482400.0, + "grad_norm": 2.1706760308011375, + "learning_rate": 2.9728638384267645e-08, + "loss": 0.759, + "num_input_tokens_seen": 341140450, + "step": 10497 + }, + { + "epoch": 0.9467466293908103, + "flos": 18634160560320.0, + "grad_norm": 2.106474174545837, + "learning_rate": 2.962836789409784e-08, + "loss": 0.7492, + "num_input_tokens_seen": 341166710, + "step": 10498 + }, + { + "epoch": 0.9468368129142806, + "flos": 23480967758880.0, + "grad_norm": 1.670075919201644, + "learning_rate": 2.95282655268998e-08, + "loss": 0.835, + "num_input_tokens_seen": 341193200, + "step": 10499 + }, + { + "epoch": 0.9469269964377508, + "flos": 27342175473120.0, + "grad_norm": 1.8336032394209136, + "learning_rate": 2.942833129121558e-08, + "loss": 0.7436, + "num_input_tokens_seen": 341223690, + "step": 10500 + }, + { + "epoch": 0.947017179961221, + "flos": 25191640914240.0, + "grad_norm": 1.8642583880016463, + "learning_rate": 2.9328565195571475e-08, + "loss": 0.6687, + "num_input_tokens_seen": 341254330, + "step": 10501 + }, + { + "epoch": 0.9471073634846914, + "flos": 26794913432160.0, + "grad_norm": 1.7142957363516902, + "learning_rate": 2.9228967248480675e-08, + "loss": 0.8138, + "num_input_tokens_seen": 341282780, + "step": 10502 + }, + { + "epoch": 0.9471975470081616, + "flos": 15720336083520.0, + "grad_norm": 2.120176338203529, + "learning_rate": 2.912953745844082e-08, + "loss": 0.7867, + "num_input_tokens_seen": 341308775, + "step": 10503 + }, + { + "epoch": 0.9472877305316318, + "flos": 26389179601920.0, + "grad_norm": 1.7556865127559071, + "learning_rate": 2.9030275833936247e-08, + "loss": 0.732, + "num_input_tokens_seen": 341337860, + "step": 10504 + }, + { + "epoch": 0.9473779140551022, + "flos": 19653961957440.0, + "grad_norm": 2.600977376485286, + "learning_rate": 2.893118238343617e-08, + "loss": 0.6475, + "num_input_tokens_seen": 341362570, + "step": 10505 + }, + { + "epoch": 0.9474680975785724, + "flos": 23990459590080.0, + "grad_norm": 2.191954083567394, + "learning_rate": 2.8832257115396052e-08, + "loss": 0.6576, + "num_input_tokens_seen": 341391390, + "step": 10506 + }, + { + "epoch": 0.9475582811020427, + "flos": 21257383154400.0, + "grad_norm": 2.8578779126539424, + "learning_rate": 2.873350003825692e-08, + "loss": 0.725, + "num_input_tokens_seen": 341419875, + "step": 10507 + }, + { + "epoch": 0.9476484646255129, + "flos": 66514295320320.0, + "grad_norm": 0.5610645309739537, + "learning_rate": 2.8634911160444696e-08, + "loss": 0.4713, + "num_input_tokens_seen": 341511080, + "step": 10508 + }, + { + "epoch": 0.9477386481489832, + "flos": 25077847078080.0, + "grad_norm": 1.807612900855094, + "learning_rate": 2.853649049037199e-08, + "loss": 0.8017, + "num_input_tokens_seen": 341538485, + "step": 10509 + }, + { + "epoch": 0.9478288316724535, + "flos": 17906038968960.0, + "grad_norm": 2.20126339875302, + "learning_rate": 2.8438238036436525e-08, + "loss": 0.791, + "num_input_tokens_seen": 341564215, + "step": 10510 + }, + { + "epoch": 0.9479190151959237, + "flos": 30833152211520.0, + "grad_norm": 1.9499217506750093, + "learning_rate": 2.834015380702137e-08, + "loss": 0.7403, + "num_input_tokens_seen": 341592535, + "step": 10511 + }, + { + "epoch": 0.9480091987193939, + "flos": 61581465829920.0, + "grad_norm": 0.5801718279747698, + "learning_rate": 2.824223781049606e-08, + "loss": 0.5126, + "num_input_tokens_seen": 341688370, + "step": 10512 + }, + { + "epoch": 0.9480993822428643, + "flos": 17464393416960.0, + "grad_norm": 2.845056589968303, + "learning_rate": 2.8144490055215465e-08, + "loss": 0.8016, + "num_input_tokens_seen": 341714690, + "step": 10513 + }, + { + "epoch": 0.9481895657663345, + "flos": 23368735052640.0, + "grad_norm": 1.820823041187008, + "learning_rate": 2.8046910549519355e-08, + "loss": 0.7776, + "num_input_tokens_seen": 341743410, + "step": 10514 + }, + { + "epoch": 0.9482797492898047, + "flos": 26577473104320.0, + "grad_norm": 2.0651617503145263, + "learning_rate": 2.794949930173418e-08, + "loss": 0.7497, + "num_input_tokens_seen": 341775425, + "step": 10515 + }, + { + "epoch": 0.948369932813275, + "flos": 18594197334720.0, + "grad_norm": 2.0303389633635915, + "learning_rate": 2.7852256320171296e-08, + "loss": 0.7699, + "num_input_tokens_seen": 341801990, + "step": 10516 + }, + { + "epoch": 0.9484601163367453, + "flos": 29891976324000.0, + "grad_norm": 3.740682343033085, + "learning_rate": 2.775518161312851e-08, + "loss": 0.604, + "num_input_tokens_seen": 341831880, + "step": 10517 + }, + { + "epoch": 0.9485502998602156, + "flos": 26212111367040.0, + "grad_norm": 1.5794688630979243, + "learning_rate": 2.76582751888883e-08, + "loss": 0.7705, + "num_input_tokens_seen": 341863460, + "step": 10518 + }, + { + "epoch": 0.9486404833836858, + "flos": 17723766967680.0, + "grad_norm": 2.7252487908978216, + "learning_rate": 2.756153705571962e-08, + "loss": 0.7648, + "num_input_tokens_seen": 341888605, + "step": 10519 + }, + { + "epoch": 0.948730666907156, + "flos": 25559121327840.0, + "grad_norm": 1.4617251960290123, + "learning_rate": 2.74649672218763e-08, + "loss": 0.7264, + "num_input_tokens_seen": 341921065, + "step": 10520 + }, + { + "epoch": 0.9488208504306264, + "flos": 31422235966080.0, + "grad_norm": 1.7664608622582088, + "learning_rate": 2.7368565695598424e-08, + "loss": 0.7285, + "num_input_tokens_seen": 341951010, + "step": 10521 + }, + { + "epoch": 0.9489110339540966, + "flos": 27338793024960.0, + "grad_norm": 2.3548621666616945, + "learning_rate": 2.727233248511185e-08, + "loss": 0.671, + "num_input_tokens_seen": 341980090, + "step": 10522 + }, + { + "epoch": 0.9490012174775668, + "flos": 23844285159360.0, + "grad_norm": 2.337831475729406, + "learning_rate": 2.71762675986269e-08, + "loss": 0.6979, + "num_input_tokens_seen": 342006145, + "step": 10523 + }, + { + "epoch": 0.9490914010010371, + "flos": 26100845074560.0, + "grad_norm": 1.7854789592450706, + "learning_rate": 2.7080371044341242e-08, + "loss": 0.7809, + "num_input_tokens_seen": 342036480, + "step": 10524 + }, + { + "epoch": 0.9491815845245074, + "flos": 23005268973120.0, + "grad_norm": 1.928320767497161, + "learning_rate": 2.6984642830436556e-08, + "loss": 0.7737, + "num_input_tokens_seen": 342064765, + "step": 10525 + }, + { + "epoch": 0.9492717680479776, + "flos": 37761569766720.0, + "grad_norm": 1.7514370739800897, + "learning_rate": 2.688908296508141e-08, + "loss": 0.7279, + "num_input_tokens_seen": 342095740, + "step": 10526 + }, + { + "epoch": 0.9493619515714479, + "flos": 23443234718880.0, + "grad_norm": 2.005516788158466, + "learning_rate": 2.679369145642929e-08, + "loss": 0.7936, + "num_input_tokens_seen": 342123040, + "step": 10527 + }, + { + "epoch": 0.9494521350949181, + "flos": 23078913734880.0, + "grad_norm": 2.3178534530183716, + "learning_rate": 2.669846831261946e-08, + "loss": 0.7632, + "num_input_tokens_seen": 342149190, + "step": 10528 + }, + { + "epoch": 0.9495423186183884, + "flos": 17575176502560.0, + "grad_norm": 2.1750612009141648, + "learning_rate": 2.6603413541776976e-08, + "loss": 0.7341, + "num_input_tokens_seen": 342175900, + "step": 10529 + }, + { + "epoch": 0.9496325021418587, + "flos": 26537993085600.0, + "grad_norm": 1.792468298255298, + "learning_rate": 2.6508527152012683e-08, + "loss": 0.696, + "num_input_tokens_seen": 342206805, + "step": 10530 + }, + { + "epoch": 0.9497226856653289, + "flos": 31709381061120.0, + "grad_norm": 1.555541757492147, + "learning_rate": 2.641380915142233e-08, + "loss": 0.6453, + "num_input_tokens_seen": 342238325, + "step": 10531 + }, + { + "epoch": 0.9498128691887993, + "flos": 22643066665440.0, + "grad_norm": 1.8899290837501321, + "learning_rate": 2.6319259548088334e-08, + "loss": 0.7675, + "num_input_tokens_seen": 342267145, + "step": 10532 + }, + { + "epoch": 0.9499030527122695, + "flos": 20966669762400.0, + "grad_norm": 2.188760993446266, + "learning_rate": 2.6224878350077585e-08, + "loss": 0.7412, + "num_input_tokens_seen": 342297765, + "step": 10533 + }, + { + "epoch": 0.9499932362357397, + "flos": 31386361414080.0, + "grad_norm": 2.3445212557515136, + "learning_rate": 2.6130665565443633e-08, + "loss": 0.6599, + "num_input_tokens_seen": 342327135, + "step": 10534 + }, + { + "epoch": 0.95008341975921, + "flos": 19946236479360.0, + "grad_norm": 1.5792631067840528, + "learning_rate": 2.603662120222494e-08, + "loss": 0.7474, + "num_input_tokens_seen": 342355805, + "step": 10535 + }, + { + "epoch": 0.9501736032826803, + "flos": 31640679877440.0, + "grad_norm": 1.8728815190712838, + "learning_rate": 2.59427452684462e-08, + "loss": 0.5874, + "num_input_tokens_seen": 342383605, + "step": 10536 + }, + { + "epoch": 0.9502637868061505, + "flos": 27595378843680.0, + "grad_norm": 1.8408703703613254, + "learning_rate": 2.5849037772117443e-08, + "loss": 0.7189, + "num_input_tokens_seen": 342414050, + "step": 10537 + }, + { + "epoch": 0.9503539703296208, + "flos": 21873383548800.0, + "grad_norm": 2.0675087423266603, + "learning_rate": 2.575549872123384e-08, + "loss": 0.7654, + "num_input_tokens_seen": 342443300, + "step": 10538 + }, + { + "epoch": 0.950444153853091, + "flos": 22530536601120.0, + "grad_norm": 1.8099592751194549, + "learning_rate": 2.5662128123776994e-08, + "loss": 0.749, + "num_input_tokens_seen": 342470575, + "step": 10539 + }, + { + "epoch": 0.9505343373765613, + "flos": 22676934050400.0, + "grad_norm": 2.61313052804796, + "learning_rate": 2.5568925987713875e-08, + "loss": 0.7381, + "num_input_tokens_seen": 342502395, + "step": 10540 + }, + { + "epoch": 0.9506245209000316, + "flos": 18305862807360.0, + "grad_norm": 2.3313290866394607, + "learning_rate": 2.5475892320996785e-08, + "loss": 0.7142, + "num_input_tokens_seen": 342527500, + "step": 10541 + }, + { + "epoch": 0.9507147044235018, + "flos": 59869194374880.0, + "grad_norm": 0.6681935785281317, + "learning_rate": 2.5383027131564038e-08, + "loss": 0.56, + "num_input_tokens_seen": 342617755, + "step": 10542 + }, + { + "epoch": 0.950804887946972, + "flos": 21184927824960.0, + "grad_norm": 1.8298345544163972, + "learning_rate": 2.52903304273393e-08, + "loss": 0.7119, + "num_input_tokens_seen": 342646510, + "step": 10543 + }, + { + "epoch": 0.9508950714704424, + "flos": 20601828401760.0, + "grad_norm": 1.6101926929952948, + "learning_rate": 2.519780221623202e-08, + "loss": 0.7259, + "num_input_tokens_seen": 342676175, + "step": 10544 + }, + { + "epoch": 0.9509852549939126, + "flos": 31969126309440.0, + "grad_norm": 3.9901339928636306, + "learning_rate": 2.510544250613722e-08, + "loss": 0.6991, + "num_input_tokens_seen": 342706740, + "step": 10545 + }, + { + "epoch": 0.9510754385173829, + "flos": 21839887861440.0, + "grad_norm": 1.942843175986119, + "learning_rate": 2.501325130493548e-08, + "loss": 0.8406, + "num_input_tokens_seen": 342732370, + "step": 10546 + }, + { + "epoch": 0.9511656220408531, + "flos": 61348042505280.0, + "grad_norm": 0.7203260226025077, + "learning_rate": 2.4921228620493395e-08, + "loss": 0.5777, + "num_input_tokens_seen": 342821115, + "step": 10547 + }, + { + "epoch": 0.9512558055643234, + "flos": 61921955264160.0, + "grad_norm": 0.7316980472807134, + "learning_rate": 2.4829374460662244e-08, + "loss": 0.614, + "num_input_tokens_seen": 342914740, + "step": 10548 + }, + { + "epoch": 0.9513459890877937, + "flos": 24755756675040.0, + "grad_norm": 1.8933728638851823, + "learning_rate": 2.473768883327976e-08, + "loss": 0.7602, + "num_input_tokens_seen": 342944830, + "step": 10549 + }, + { + "epoch": 0.9514361726112639, + "flos": 28581201346560.0, + "grad_norm": 1.9752603697268623, + "learning_rate": 2.464617174616923e-08, + "loss": 0.7039, + "num_input_tokens_seen": 342975385, + "step": 10550 + }, + { + "epoch": 0.9515263561347341, + "flos": 23987560348800.0, + "grad_norm": 1.9786007590773362, + "learning_rate": 2.455482320713953e-08, + "loss": 0.7444, + "num_input_tokens_seen": 343004185, + "step": 10551 + }, + { + "epoch": 0.9516165396582045, + "flos": 28249223787360.0, + "grad_norm": 1.9620881371295174, + "learning_rate": 2.4463643223984643e-08, + "loss": 0.8011, + "num_input_tokens_seen": 343033370, + "step": 10552 + }, + { + "epoch": 0.9517067231816747, + "flos": 28106543314080.0, + "grad_norm": 2.3719518073591477, + "learning_rate": 2.4372631804484567e-08, + "loss": 0.8169, + "num_input_tokens_seen": 343056525, + "step": 10553 + }, + { + "epoch": 0.9517969067051449, + "flos": 25411125578880.0, + "grad_norm": 1.8450650196972465, + "learning_rate": 2.4281788956405313e-08, + "loss": 0.8259, + "num_input_tokens_seen": 343085760, + "step": 10554 + }, + { + "epoch": 0.9518870902286153, + "flos": 26280180664800.0, + "grad_norm": 2.6355545751703136, + "learning_rate": 2.4191114687497572e-08, + "loss": 0.6716, + "num_input_tokens_seen": 343113935, + "step": 10555 + }, + { + "epoch": 0.9519772737520855, + "flos": 24572815618080.0, + "grad_norm": 3.037822558697648, + "learning_rate": 2.4100609005498706e-08, + "loss": 0.7024, + "num_input_tokens_seen": 343143165, + "step": 10556 + }, + { + "epoch": 0.9520674572755558, + "flos": 25885263234720.0, + "grad_norm": 7.1747323122498035, + "learning_rate": 2.4010271918130764e-08, + "loss": 0.6739, + "num_input_tokens_seen": 343173940, + "step": 10557 + }, + { + "epoch": 0.952157640799026, + "flos": 18197458586400.0, + "grad_norm": 1.9807278014131613, + "learning_rate": 2.39201034331018e-08, + "loss": 0.705, + "num_input_tokens_seen": 343201920, + "step": 10558 + }, + { + "epoch": 0.9522478243224963, + "flos": 19399866512640.0, + "grad_norm": 7.302476844215443, + "learning_rate": 2.3830103558105663e-08, + "loss": 0.7348, + "num_input_tokens_seen": 343229935, + "step": 10559 + }, + { + "epoch": 0.9523380078459666, + "flos": 22314211366080.0, + "grad_norm": 2.165983153393043, + "learning_rate": 2.374027230082154e-08, + "loss": 0.7488, + "num_input_tokens_seen": 343255620, + "step": 10560 + }, + { + "epoch": 0.9524281913694368, + "flos": 67315466957280.0, + "grad_norm": 0.6881414585414668, + "learning_rate": 2.365060966891441e-08, + "loss": 0.5941, + "num_input_tokens_seen": 343339570, + "step": 10561 + }, + { + "epoch": 0.952518374892907, + "flos": 26066271464160.0, + "grad_norm": 1.7245598134644895, + "learning_rate": 2.3561115670034827e-08, + "loss": 0.7369, + "num_input_tokens_seen": 343372430, + "step": 10562 + }, + { + "epoch": 0.9526085584163774, + "flos": 21221657281440.0, + "grad_norm": 1.6649904484517828, + "learning_rate": 2.3471790311818675e-08, + "loss": 0.7634, + "num_input_tokens_seen": 343401300, + "step": 10563 + }, + { + "epoch": 0.9526987419398476, + "flos": 24493744071360.0, + "grad_norm": 1.6644077569686822, + "learning_rate": 2.338263360188808e-08, + "loss": 0.7325, + "num_input_tokens_seen": 343429675, + "step": 10564 + }, + { + "epoch": 0.9527889254633178, + "flos": 22381537268640.0, + "grad_norm": 2.2720303195085476, + "learning_rate": 2.329364554784985e-08, + "loss": 0.68, + "num_input_tokens_seen": 343457420, + "step": 10565 + }, + { + "epoch": 0.9528791089867881, + "flos": 19357710271200.0, + "grad_norm": 1.8894990452897507, + "learning_rate": 2.3204826157297465e-08, + "loss": 0.8026, + "num_input_tokens_seen": 343484770, + "step": 10566 + }, + { + "epoch": 0.9529692925102584, + "flos": 21804422176800.0, + "grad_norm": 1.7575564049559156, + "learning_rate": 2.3116175437809082e-08, + "loss": 0.6829, + "num_input_tokens_seen": 343513935, + "step": 10567 + }, + { + "epoch": 0.9530594760337286, + "flos": 69848467076640.0, + "grad_norm": 0.6462404649573399, + "learning_rate": 2.30276933969491e-08, + "loss": 0.5954, + "num_input_tokens_seen": 343603415, + "step": 10568 + }, + { + "epoch": 0.9531496595571989, + "flos": 25193164874400.0, + "grad_norm": 1.746349819446056, + "learning_rate": 2.2939380042267255e-08, + "loss": 0.7979, + "num_input_tokens_seen": 343633360, + "step": 10569 + }, + { + "epoch": 0.9532398430806691, + "flos": 19654928371200.0, + "grad_norm": 1.9695917176978146, + "learning_rate": 2.2851235381298627e-08, + "loss": 0.649, + "num_input_tokens_seen": 343659210, + "step": 10570 + }, + { + "epoch": 0.9533300266041395, + "flos": 23918004260640.0, + "grad_norm": 1.8736412211189404, + "learning_rate": 2.2763259421564986e-08, + "loss": 0.6697, + "num_input_tokens_seen": 343687610, + "step": 10571 + }, + { + "epoch": 0.9534202101276097, + "flos": 28725963326400.0, + "grad_norm": 1.955353038258484, + "learning_rate": 2.2675452170571873e-08, + "loss": 0.7154, + "num_input_tokens_seen": 343717685, + "step": 10572 + }, + { + "epoch": 0.9535103936510799, + "flos": 32114668854240.0, + "grad_norm": 1.9817279371580858, + "learning_rate": 2.2587813635812414e-08, + "loss": 0.7381, + "num_input_tokens_seen": 343747920, + "step": 10573 + }, + { + "epoch": 0.9536005771745502, + "flos": 21476198763360.0, + "grad_norm": 2.548658156300009, + "learning_rate": 2.2500343824763958e-08, + "loss": 0.8284, + "num_input_tokens_seen": 343777095, + "step": 10574 + }, + { + "epoch": 0.9536907606980205, + "flos": 27956912095680.0, + "grad_norm": 1.9723332206700683, + "learning_rate": 2.2413042744890088e-08, + "loss": 0.6926, + "num_input_tokens_seen": 343804935, + "step": 10575 + }, + { + "epoch": 0.9537809442214907, + "flos": 25556928312000.0, + "grad_norm": 2.2526388588293744, + "learning_rate": 2.2325910403639514e-08, + "loss": 0.7277, + "num_input_tokens_seen": 343832145, + "step": 10576 + }, + { + "epoch": 0.953871127744961, + "flos": 30618982822560.0, + "grad_norm": 1.758646517097105, + "learning_rate": 2.223894680844718e-08, + "loss": 0.6982, + "num_input_tokens_seen": 343863655, + "step": 10577 + }, + { + "epoch": 0.9539613112684313, + "flos": 33971850968160.0, + "grad_norm": 1.6057986920508094, + "learning_rate": 2.2152151966733146e-08, + "loss": 0.5776, + "num_input_tokens_seen": 343893730, + "step": 10578 + }, + { + "epoch": 0.9540514947919015, + "flos": 23115011305440.0, + "grad_norm": 1.5639250551057973, + "learning_rate": 2.2065525885903267e-08, + "loss": 0.7782, + "num_input_tokens_seen": 343924165, + "step": 10579 + }, + { + "epoch": 0.9541416783153718, + "flos": 21549174469440.0, + "grad_norm": 1.9893396105121817, + "learning_rate": 2.1979068573348747e-08, + "loss": 0.7876, + "num_input_tokens_seen": 343951315, + "step": 10580 + }, + { + "epoch": 0.954231861838842, + "flos": 25114353516000.0, + "grad_norm": 1.7586960669993839, + "learning_rate": 2.1892780036447013e-08, + "loss": 0.7045, + "num_input_tokens_seen": 343979740, + "step": 10581 + }, + { + "epoch": 0.9543220453623124, + "flos": 15902273556960.0, + "grad_norm": 2.2020740572596567, + "learning_rate": 2.1806660282560175e-08, + "loss": 0.8879, + "num_input_tokens_seen": 344003040, + "step": 10582 + }, + { + "epoch": 0.9544122288857826, + "flos": 26028129556800.0, + "grad_norm": 1.5558450501376548, + "learning_rate": 2.1720709319037024e-08, + "loss": 0.7648, + "num_input_tokens_seen": 344033455, + "step": 10583 + }, + { + "epoch": 0.9545024124092528, + "flos": 29309099919360.0, + "grad_norm": 1.887309169596186, + "learning_rate": 2.1634927153211023e-08, + "loss": 0.7076, + "num_input_tokens_seen": 344063595, + "step": 10584 + }, + { + "epoch": 0.954592595932723, + "flos": 21767432532000.0, + "grad_norm": 2.4713185143319314, + "learning_rate": 2.1549313792401437e-08, + "loss": 0.806, + "num_input_tokens_seen": 344086390, + "step": 10585 + }, + { + "epoch": 0.9546827794561934, + "flos": 24460025365440.0, + "grad_norm": 3.583919979837022, + "learning_rate": 2.1463869243913746e-08, + "loss": 0.835, + "num_input_tokens_seen": 344115035, + "step": 10586 + }, + { + "epoch": 0.9547729629796636, + "flos": 22933817227200.0, + "grad_norm": 2.0268096514297196, + "learning_rate": 2.1378593515037902e-08, + "loss": 0.7209, + "num_input_tokens_seen": 344143785, + "step": 10587 + }, + { + "epoch": 0.9548631465031339, + "flos": 23334793328160.0, + "grad_norm": 1.8516022239318994, + "learning_rate": 2.129348661305075e-08, + "loss": 0.7994, + "num_input_tokens_seen": 344171905, + "step": 10588 + }, + { + "epoch": 0.9549533300266041, + "flos": 25190488651680.0, + "grad_norm": 1.8893799899152406, + "learning_rate": 2.1208548545213813e-08, + "loss": 0.7756, + "num_input_tokens_seen": 344199185, + "step": 10589 + }, + { + "epoch": 0.9550435135500744, + "flos": 23990273741280.0, + "grad_norm": 2.8016581632298627, + "learning_rate": 2.1123779318774404e-08, + "loss": 0.6961, + "num_input_tokens_seen": 344227035, + "step": 10590 + }, + { + "epoch": 0.9551336970735447, + "flos": 23152335478080.0, + "grad_norm": 1.7034132827215829, + "learning_rate": 2.1039178940965408e-08, + "loss": 0.6879, + "num_input_tokens_seen": 344255680, + "step": 10591 + }, + { + "epoch": 0.9552238805970149, + "flos": 24898288469280.0, + "grad_norm": 1.5783795444178097, + "learning_rate": 2.0954747419005712e-08, + "loss": 0.8042, + "num_input_tokens_seen": 344283430, + "step": 10592 + }, + { + "epoch": 0.9553140641204851, + "flos": 19581729646560.0, + "grad_norm": 1.8511646986220758, + "learning_rate": 2.087048476009934e-08, + "loss": 0.7262, + "num_input_tokens_seen": 344311845, + "step": 10593 + }, + { + "epoch": 0.9554042476439555, + "flos": 23626101436320.0, + "grad_norm": 2.0298068275056433, + "learning_rate": 2.0786390971435862e-08, + "loss": 0.7175, + "num_input_tokens_seen": 344342095, + "step": 10594 + }, + { + "epoch": 0.9554944311674257, + "flos": 19509125638080.0, + "grad_norm": 2.4866385075131285, + "learning_rate": 2.070246606019088e-08, + "loss": 0.7125, + "num_input_tokens_seen": 344368735, + "step": 10595 + }, + { + "epoch": 0.955584614690896, + "flos": 20456100008160.0, + "grad_norm": 2.8143284258461008, + "learning_rate": 2.0618710033525112e-08, + "loss": 0.6554, + "num_input_tokens_seen": 344395065, + "step": 10596 + }, + { + "epoch": 0.9556747982143662, + "flos": 18379284550560.0, + "grad_norm": 1.9488339703705995, + "learning_rate": 2.053512289858528e-08, + "loss": 0.708, + "num_input_tokens_seen": 344422820, + "step": 10597 + }, + { + "epoch": 0.9557649817378365, + "flos": 21840445407840.0, + "grad_norm": 2.3067367494921305, + "learning_rate": 2.0451704662503456e-08, + "loss": 0.7831, + "num_input_tokens_seen": 344448995, + "step": 10598 + }, + { + "epoch": 0.9558551652613068, + "flos": 19873074924480.0, + "grad_norm": 1.9782201277050102, + "learning_rate": 2.0368455332397282e-08, + "loss": 0.8228, + "num_input_tokens_seen": 344473570, + "step": 10599 + }, + { + "epoch": 0.955945348784777, + "flos": 20711570734080.0, + "grad_norm": 2.1477811053393308, + "learning_rate": 2.0285374915369967e-08, + "loss": 0.7602, + "num_input_tokens_seen": 344501105, + "step": 10600 + }, + { + "epoch": 0.9560355323082473, + "flos": 25844333595360.0, + "grad_norm": 1.900045773306151, + "learning_rate": 2.020246341851073e-08, + "loss": 0.7843, + "num_input_tokens_seen": 344528260, + "step": 10601 + }, + { + "epoch": 0.9561257158317176, + "flos": 20347175410560.0, + "grad_norm": 1.6362735989634107, + "learning_rate": 2.0119720848893463e-08, + "loss": 0.786, + "num_input_tokens_seen": 344554040, + "step": 10602 + }, + { + "epoch": 0.9562158993551878, + "flos": 23951648627040.0, + "grad_norm": 1.8016073520764473, + "learning_rate": 2.0037147213578964e-08, + "loss": 0.6934, + "num_input_tokens_seen": 344581815, + "step": 10603 + }, + { + "epoch": 0.956306082878658, + "flos": 31349074411200.0, + "grad_norm": 1.6611881415085552, + "learning_rate": 1.9954742519612265e-08, + "loss": 0.769, + "num_input_tokens_seen": 344613645, + "step": 10604 + }, + { + "epoch": 0.9563962664021284, + "flos": 21622373194080.0, + "grad_norm": 1.8312934740133453, + "learning_rate": 1.9872506774024633e-08, + "loss": 0.7207, + "num_input_tokens_seen": 344640940, + "step": 10605 + }, + { + "epoch": 0.9564864499255986, + "flos": 15610593751200.0, + "grad_norm": 1.895050292341644, + "learning_rate": 1.979043998383334e-08, + "loss": 0.7514, + "num_input_tokens_seen": 344668100, + "step": 10606 + }, + { + "epoch": 0.9565766334490688, + "flos": 33355441706400.0, + "grad_norm": 1.3848906596434918, + "learning_rate": 1.970854215604034e-08, + "loss": 0.6851, + "num_input_tokens_seen": 344702400, + "step": 10607 + }, + { + "epoch": 0.9566668169725391, + "flos": 23873878021920.0, + "grad_norm": 7.571346184829823, + "learning_rate": 1.9626813297633826e-08, + "loss": 0.7122, + "num_input_tokens_seen": 344727170, + "step": 10608 + }, + { + "epoch": 0.9567570004960094, + "flos": 20419853758560.0, + "grad_norm": 2.483536680801243, + "learning_rate": 1.954525341558688e-08, + "loss": 0.6444, + "num_input_tokens_seen": 344754610, + "step": 10609 + }, + { + "epoch": 0.9568471840194797, + "flos": 21294112610880.0, + "grad_norm": 2.0802984129979003, + "learning_rate": 1.9463862516859498e-08, + "loss": 0.7072, + "num_input_tokens_seen": 344781550, + "step": 10610 + }, + { + "epoch": 0.9569373675429499, + "flos": 19797497335200.0, + "grad_norm": 1.8250062210813074, + "learning_rate": 1.938264060839545e-08, + "loss": 0.8276, + "num_input_tokens_seen": 344810055, + "step": 10611 + }, + { + "epoch": 0.9570275510664201, + "flos": 23254309330560.0, + "grad_norm": 2.0349369333603207, + "learning_rate": 1.9301587697126086e-08, + "loss": 0.7159, + "num_input_tokens_seen": 344833650, + "step": 10612 + }, + { + "epoch": 0.9571177345898905, + "flos": 24864309575040.0, + "grad_norm": 2.1619298458944813, + "learning_rate": 1.9220703789966318e-08, + "loss": 0.6954, + "num_input_tokens_seen": 344862080, + "step": 10613 + }, + { + "epoch": 0.9572079181133607, + "flos": 23407434506400.0, + "grad_norm": 1.8744123173656173, + "learning_rate": 1.913998889381818e-08, + "loss": 0.6987, + "num_input_tokens_seen": 344890910, + "step": 10614 + }, + { + "epoch": 0.9572981016368309, + "flos": 19545780755040.0, + "grad_norm": 1.8670739813736914, + "learning_rate": 1.9059443015568387e-08, + "loss": 0.756, + "num_input_tokens_seen": 344917090, + "step": 10615 + }, + { + "epoch": 0.9573882851603012, + "flos": 22970286495360.0, + "grad_norm": 2.147740809447036, + "learning_rate": 1.8979066162089884e-08, + "loss": 0.7549, + "num_input_tokens_seen": 344945390, + "step": 10616 + }, + { + "epoch": 0.9574784686837715, + "flos": 24646200191520.0, + "grad_norm": 2.0781402587330433, + "learning_rate": 1.889885834024052e-08, + "loss": 0.7476, + "num_input_tokens_seen": 344973345, + "step": 10617 + }, + { + "epoch": 0.9575686522072417, + "flos": 26284975563840.0, + "grad_norm": 1.8206626326144848, + "learning_rate": 1.8818819556864374e-08, + "loss": 0.693, + "num_input_tokens_seen": 345001090, + "step": 10618 + }, + { + "epoch": 0.957658835730712, + "flos": 13461248624640.0, + "grad_norm": 2.535018000190828, + "learning_rate": 1.873894981879065e-08, + "loss": 0.7226, + "num_input_tokens_seen": 345025695, + "step": 10619 + }, + { + "epoch": 0.9577490192541822, + "flos": 70560116730720.0, + "grad_norm": 0.6272558034245997, + "learning_rate": 1.8659249132834342e-08, + "loss": 0.5899, + "num_input_tokens_seen": 345123095, + "step": 10620 + }, + { + "epoch": 0.9578392027776526, + "flos": 21913978660320.0, + "grad_norm": 1.6811357508983646, + "learning_rate": 1.857971750579579e-08, + "loss": 0.7235, + "num_input_tokens_seen": 345152750, + "step": 10621 + }, + { + "epoch": 0.9579293863011228, + "flos": 33019226794560.0, + "grad_norm": 2.801073811891487, + "learning_rate": 1.8500354944461116e-08, + "loss": 0.6235, + "num_input_tokens_seen": 345182475, + "step": 10622 + }, + { + "epoch": 0.958019569824593, + "flos": 24171281970720.0, + "grad_norm": 1.7028356883849871, + "learning_rate": 1.8421161455602242e-08, + "loss": 0.7487, + "num_input_tokens_seen": 345214180, + "step": 10623 + }, + { + "epoch": 0.9581097533480634, + "flos": 18124371371040.0, + "grad_norm": 1.8997846933141997, + "learning_rate": 1.834213704597598e-08, + "loss": 0.7668, + "num_input_tokens_seen": 345241860, + "step": 10624 + }, + { + "epoch": 0.9581999368715336, + "flos": 21798512184960.0, + "grad_norm": 4.2381185614656935, + "learning_rate": 1.8263281722325385e-08, + "loss": 0.7532, + "num_input_tokens_seen": 345271470, + "step": 10625 + }, + { + "epoch": 0.9582901203950038, + "flos": 24354594725280.0, + "grad_norm": 2.036251542176845, + "learning_rate": 1.818459549137885e-08, + "loss": 0.7512, + "num_input_tokens_seen": 345298945, + "step": 10626 + }, + { + "epoch": 0.9583803039184741, + "flos": 16885903044000.0, + "grad_norm": 1.9530105557290736, + "learning_rate": 1.8106078359850117e-08, + "loss": 0.7314, + "num_input_tokens_seen": 345324600, + "step": 10627 + }, + { + "epoch": 0.9584704874419444, + "flos": 27597088652640.0, + "grad_norm": 2.244726677209766, + "learning_rate": 1.802773033443894e-08, + "loss": 0.7589, + "num_input_tokens_seen": 345353965, + "step": 10628 + }, + { + "epoch": 0.9585606709654146, + "flos": 27888768458400.0, + "grad_norm": 3.3903555345471905, + "learning_rate": 1.7949551421830413e-08, + "loss": 0.7085, + "num_input_tokens_seen": 345383060, + "step": 10629 + }, + { + "epoch": 0.9586508544888849, + "flos": 27050904534720.0, + "grad_norm": 1.7267349640338108, + "learning_rate": 1.7871541628694752e-08, + "loss": 0.7096, + "num_input_tokens_seen": 345411985, + "step": 10630 + }, + { + "epoch": 0.9587410380123551, + "flos": 25993890474240.0, + "grad_norm": 6.159849092691035, + "learning_rate": 1.779370096168864e-08, + "loss": 0.652, + "num_input_tokens_seen": 345441220, + "step": 10631 + }, + { + "epoch": 0.9588312215358255, + "flos": 24093994572480.0, + "grad_norm": 2.810297718363416, + "learning_rate": 1.771602942745387e-08, + "loss": 0.7052, + "num_input_tokens_seen": 345470160, + "step": 10632 + }, + { + "epoch": 0.9589214050592957, + "flos": 22237593023520.0, + "grad_norm": 3.2519661751333495, + "learning_rate": 1.763852703261759e-08, + "loss": 0.7281, + "num_input_tokens_seen": 345495170, + "step": 10633 + }, + { + "epoch": 0.9590115885827659, + "flos": 30730063266240.0, + "grad_norm": 2.774582403568269, + "learning_rate": 1.756119378379295e-08, + "loss": 0.8043, + "num_input_tokens_seen": 345527040, + "step": 10634 + }, + { + "epoch": 0.9591017721062362, + "flos": 22715521994880.0, + "grad_norm": 1.7992252404397764, + "learning_rate": 1.7484029687578005e-08, + "loss": 0.8344, + "num_input_tokens_seen": 345555870, + "step": 10635 + }, + { + "epoch": 0.9591919556297065, + "flos": 22168631651520.0, + "grad_norm": 1.861790154384815, + "learning_rate": 1.740703475055727e-08, + "loss": 0.7159, + "num_input_tokens_seen": 345585240, + "step": 10636 + }, + { + "epoch": 0.9592821391531767, + "flos": 21730480056960.0, + "grad_norm": 1.7119518739362114, + "learning_rate": 1.7330208979300153e-08, + "loss": 0.7799, + "num_input_tokens_seen": 345613430, + "step": 10637 + }, + { + "epoch": 0.959372322676647, + "flos": 25957384036320.0, + "grad_norm": 1.6038234475637274, + "learning_rate": 1.725355238036208e-08, + "loss": 0.784, + "num_input_tokens_seen": 345644155, + "step": 10638 + }, + { + "epoch": 0.9594625062001172, + "flos": 26321370492480.0, + "grad_norm": 2.0651022937921706, + "learning_rate": 1.7177064960283594e-08, + "loss": 0.7851, + "num_input_tokens_seen": 345670765, + "step": 10639 + }, + { + "epoch": 0.9595526897235875, + "flos": 18670220961120.0, + "grad_norm": 2.32958006991701, + "learning_rate": 1.7100746725591253e-08, + "loss": 0.7384, + "num_input_tokens_seen": 345697290, + "step": 10640 + }, + { + "epoch": 0.9596428732470578, + "flos": 37943395730880.0, + "grad_norm": 3.1896536040502212, + "learning_rate": 1.7024597682796517e-08, + "loss": 0.7352, + "num_input_tokens_seen": 345725515, + "step": 10641 + }, + { + "epoch": 0.959733056770528, + "flos": 24245261260320.0, + "grad_norm": 1.911771064419766, + "learning_rate": 1.6948617838397293e-08, + "loss": 0.8163, + "num_input_tokens_seen": 345755000, + "step": 10642 + }, + { + "epoch": 0.9598232402939982, + "flos": 21184630466880.0, + "grad_norm": 2.3678100573036387, + "learning_rate": 1.6872807198876404e-08, + "loss": 0.7711, + "num_input_tokens_seen": 345785665, + "step": 10643 + }, + { + "epoch": 0.9599134238174686, + "flos": 19763964478080.0, + "grad_norm": 2.0587487062914547, + "learning_rate": 1.679716577070245e-08, + "loss": 0.8407, + "num_input_tokens_seen": 345813980, + "step": 10644 + }, + { + "epoch": 0.9600036073409388, + "flos": 17432793387360.0, + "grad_norm": 2.047715458601549, + "learning_rate": 1.6721693560329596e-08, + "loss": 0.829, + "num_input_tokens_seen": 345839880, + "step": 10645 + }, + { + "epoch": 0.960093790864409, + "flos": 24244480695360.0, + "grad_norm": 17.146930991627645, + "learning_rate": 1.6646390574197366e-08, + "loss": 0.6528, + "num_input_tokens_seen": 345869135, + "step": 10646 + }, + { + "epoch": 0.9601839743878793, + "flos": 60923978249760.0, + "grad_norm": 0.6865080700127028, + "learning_rate": 1.6571256818731504e-08, + "loss": 0.5961, + "num_input_tokens_seen": 345940045, + "step": 10647 + }, + { + "epoch": 0.9602741579113496, + "flos": 26248320446880.0, + "grad_norm": 1.8636396211396906, + "learning_rate": 1.6496292300342218e-08, + "loss": 0.677, + "num_input_tokens_seen": 345968105, + "step": 10648 + }, + { + "epoch": 0.9603643414348199, + "flos": 28726706721600.0, + "grad_norm": 2.1894721813886906, + "learning_rate": 1.642149702542639e-08, + "loss": 0.6808, + "num_input_tokens_seen": 345996555, + "step": 10649 + }, + { + "epoch": 0.9604545249582901, + "flos": 22276366816800.0, + "grad_norm": 1.5737713729461498, + "learning_rate": 1.634687100036558e-08, + "loss": 0.7042, + "num_input_tokens_seen": 346025255, + "step": 10650 + }, + { + "epoch": 0.9605447084817604, + "flos": 25082418958560.0, + "grad_norm": 2.0780236929386313, + "learning_rate": 1.627241423152781e-08, + "loss": 0.6798, + "num_input_tokens_seen": 346052665, + "step": 10651 + }, + { + "epoch": 0.9606348920052307, + "flos": 20521195725120.0, + "grad_norm": 6.949795660953199, + "learning_rate": 1.619812672526555e-08, + "loss": 0.7945, + "num_input_tokens_seen": 346078560, + "step": 10652 + }, + { + "epoch": 0.9607250755287009, + "flos": 16594855124160.0, + "grad_norm": 2.189317220915879, + "learning_rate": 1.6124008487917727e-08, + "loss": 0.7271, + "num_input_tokens_seen": 346104095, + "step": 10653 + }, + { + "epoch": 0.9608152590521711, + "flos": 70655994742560.0, + "grad_norm": 0.6397527944452359, + "learning_rate": 1.6050059525808623e-08, + "loss": 0.5937, + "num_input_tokens_seen": 346189615, + "step": 10654 + }, + { + "epoch": 0.9609054425756415, + "flos": 65701266529920.0, + "grad_norm": 0.6269279070859841, + "learning_rate": 1.597627984524763e-08, + "loss": 0.589, + "num_input_tokens_seen": 346284535, + "step": 10655 + }, + { + "epoch": 0.9609956260991117, + "flos": 23735695089600.0, + "grad_norm": 2.5781225955836327, + "learning_rate": 1.590266945253038e-08, + "loss": 0.7787, + "num_input_tokens_seen": 346314125, + "step": 10656 + }, + { + "epoch": 0.9610858096225819, + "flos": 18741263839680.0, + "grad_norm": 1.5596151040941804, + "learning_rate": 1.582922835393763e-08, + "loss": 0.7566, + "num_input_tokens_seen": 346341655, + "step": 10657 + }, + { + "epoch": 0.9611759931460522, + "flos": 18015818471040.0, + "grad_norm": 1.940727928982547, + "learning_rate": 1.5755956555735473e-08, + "loss": 0.7203, + "num_input_tokens_seen": 346368530, + "step": 10658 + }, + { + "epoch": 0.9612661766695225, + "flos": 56500634856960.0, + "grad_norm": 0.7311955880029906, + "learning_rate": 1.5682854064176244e-08, + "loss": 0.6643, + "num_input_tokens_seen": 346451655, + "step": 10659 + }, + { + "epoch": 0.9613563601929928, + "flos": 21585755246880.0, + "grad_norm": 1.7170521065787558, + "learning_rate": 1.5609920885497395e-08, + "loss": 0.6761, + "num_input_tokens_seen": 346480530, + "step": 10660 + }, + { + "epoch": 0.961446543716463, + "flos": 25515366786720.0, + "grad_norm": 2.343695372898544, + "learning_rate": 1.5537157025921732e-08, + "loss": 0.7634, + "num_input_tokens_seen": 346510105, + "step": 10661 + }, + { + "epoch": 0.9615367272399332, + "flos": 23589148961280.0, + "grad_norm": 2.3188871893180676, + "learning_rate": 1.5464562491658285e-08, + "loss": 0.7268, + "num_input_tokens_seen": 346538150, + "step": 10662 + }, + { + "epoch": 0.9616269107634036, + "flos": 21804756704640.0, + "grad_norm": 1.5672735158300357, + "learning_rate": 1.5392137288900764e-08, + "loss": 0.7795, + "num_input_tokens_seen": 346566330, + "step": 10663 + }, + { + "epoch": 0.9617170942868738, + "flos": 70482160276800.0, + "grad_norm": 0.592364602291686, + "learning_rate": 1.531988142382934e-08, + "loss": 0.5793, + "num_input_tokens_seen": 346662455, + "step": 10664 + }, + { + "epoch": 0.961807277810344, + "flos": 46431294093120.0, + "grad_norm": 3.1707012842447226, + "learning_rate": 1.5247794902608634e-08, + "loss": 0.6743, + "num_input_tokens_seen": 346695700, + "step": 10665 + }, + { + "epoch": 0.9618974613338143, + "flos": 22744854669120.0, + "grad_norm": 2.56609343737273, + "learning_rate": 1.5175877731390398e-08, + "loss": 0.7013, + "num_input_tokens_seen": 346724570, + "step": 10666 + }, + { + "epoch": 0.9619876448572846, + "flos": 30872557890720.0, + "grad_norm": 2.0484509407260107, + "learning_rate": 1.510412991631016e-08, + "loss": 0.7701, + "num_input_tokens_seen": 346752665, + "step": 10667 + }, + { + "epoch": 0.9620778283807548, + "flos": 30329310183840.0, + "grad_norm": 1.5510493510380219, + "learning_rate": 1.503255146349014e-08, + "loss": 0.7685, + "num_input_tokens_seen": 346784400, + "step": 10668 + }, + { + "epoch": 0.9621680119042251, + "flos": 24937136602080.0, + "grad_norm": 2.680988051772394, + "learning_rate": 1.4961142379037893e-08, + "loss": 0.7054, + "num_input_tokens_seen": 346813835, + "step": 10669 + }, + { + "epoch": 0.9622581954276953, + "flos": 22460162778240.0, + "grad_norm": 1.7386765619890558, + "learning_rate": 1.4889902669046327e-08, + "loss": 0.782, + "num_input_tokens_seen": 346844180, + "step": 10670 + }, + { + "epoch": 0.9623483789511657, + "flos": 15829558039200.0, + "grad_norm": 2.4299376135144932, + "learning_rate": 1.4818832339594135e-08, + "loss": 0.7418, + "num_input_tokens_seen": 346868995, + "step": 10671 + }, + { + "epoch": 0.9624385624746359, + "flos": 21766280269440.0, + "grad_norm": 1.8294267502826282, + "learning_rate": 1.474793139674535e-08, + "loss": 0.6941, + "num_input_tokens_seen": 346898950, + "step": 10672 + }, + { + "epoch": 0.9625287459981061, + "flos": 29705950176960.0, + "grad_norm": 2.0601932457215746, + "learning_rate": 1.4677199846549581e-08, + "loss": 0.7397, + "num_input_tokens_seen": 346929955, + "step": 10673 + }, + { + "epoch": 0.9626189295215765, + "flos": 22387372920960.0, + "grad_norm": 1.9557974270533067, + "learning_rate": 1.4606637695042224e-08, + "loss": 0.7377, + "num_input_tokens_seen": 346957905, + "step": 10674 + }, + { + "epoch": 0.9627091130450467, + "flos": 23225050995840.0, + "grad_norm": 2.29427912375388, + "learning_rate": 1.4536244948243793e-08, + "loss": 0.8329, + "num_input_tokens_seen": 346982835, + "step": 10675 + }, + { + "epoch": 0.9627992965685169, + "flos": 25080263112480.0, + "grad_norm": 2.3427188893972684, + "learning_rate": 1.4466021612160595e-08, + "loss": 0.7118, + "num_input_tokens_seen": 347012550, + "step": 10676 + }, + { + "epoch": 0.9628894800919872, + "flos": 15465088376160.0, + "grad_norm": 1.8647685962327143, + "learning_rate": 1.4395967692784505e-08, + "loss": 0.7112, + "num_input_tokens_seen": 347040060, + "step": 10677 + }, + { + "epoch": 0.9629796636154575, + "flos": 33681434934240.0, + "grad_norm": 1.9049759030757945, + "learning_rate": 1.4326083196092963e-08, + "loss": 0.6728, + "num_input_tokens_seen": 347070090, + "step": 10678 + }, + { + "epoch": 0.9630698471389277, + "flos": 24281804868000.0, + "grad_norm": 1.6807091023662404, + "learning_rate": 1.42563681280492e-08, + "loss": 0.7501, + "num_input_tokens_seen": 347099000, + "step": 10679 + }, + { + "epoch": 0.963160030662398, + "flos": 21330619048800.0, + "grad_norm": 1.8770649665275476, + "learning_rate": 1.4186822494600902e-08, + "loss": 0.7156, + "num_input_tokens_seen": 347127210, + "step": 10680 + }, + { + "epoch": 0.9632502141858682, + "flos": 21948923968320.0, + "grad_norm": 2.107400379654625, + "learning_rate": 1.4117446301682877e-08, + "loss": 0.7287, + "num_input_tokens_seen": 347159370, + "step": 10681 + }, + { + "epoch": 0.9633403977093385, + "flos": 30146443466400.0, + "grad_norm": 2.0253534267618325, + "learning_rate": 1.4048239555214392e-08, + "loss": 0.6952, + "num_input_tokens_seen": 347191060, + "step": 10682 + }, + { + "epoch": 0.9634305812328088, + "flos": 24537349933440.0, + "grad_norm": 1.671099759341185, + "learning_rate": 1.3979202261100497e-08, + "loss": 0.7729, + "num_input_tokens_seen": 347220650, + "step": 10683 + }, + { + "epoch": 0.963520764756279, + "flos": 23261371584960.0, + "grad_norm": 4.254992525876358, + "learning_rate": 1.3910334425231817e-08, + "loss": 0.8034, + "num_input_tokens_seen": 347248360, + "step": 10684 + }, + { + "epoch": 0.9636109482797492, + "flos": 24715458921600.0, + "grad_norm": 1.8858734195824225, + "learning_rate": 1.384163605348454e-08, + "loss": 0.7381, + "num_input_tokens_seen": 347279190, + "step": 10685 + }, + { + "epoch": 0.9637011318032196, + "flos": 17537480632320.0, + "grad_norm": 2.216630917693484, + "learning_rate": 1.3773107151720642e-08, + "loss": 0.7966, + "num_input_tokens_seen": 347305820, + "step": 10686 + }, + { + "epoch": 0.9637913153266898, + "flos": 22529458678080.0, + "grad_norm": 1.882684093861164, + "learning_rate": 1.3704747725787003e-08, + "loss": 0.8037, + "num_input_tokens_seen": 347333080, + "step": 10687 + }, + { + "epoch": 0.9638814988501601, + "flos": 57607616542080.0, + "grad_norm": 0.7310131716526517, + "learning_rate": 1.3636557781516512e-08, + "loss": 0.5849, + "num_input_tokens_seen": 347427990, + "step": 10688 + }, + { + "epoch": 0.9639716823736303, + "flos": 34847039064480.0, + "grad_norm": 1.801622239881659, + "learning_rate": 1.3568537324727847e-08, + "loss": 0.7432, + "num_input_tokens_seen": 347456150, + "step": 10689 + }, + { + "epoch": 0.9640618658971006, + "flos": 32989448083200.0, + "grad_norm": 1.781438382543772, + "learning_rate": 1.3500686361224589e-08, + "loss": 0.7382, + "num_input_tokens_seen": 347485125, + "step": 10690 + }, + { + "epoch": 0.9641520494205709, + "flos": 16339644586560.0, + "grad_norm": 2.332994220705228, + "learning_rate": 1.3433004896796108e-08, + "loss": 0.7343, + "num_input_tokens_seen": 347510990, + "step": 10691 + }, + { + "epoch": 0.9642422329440411, + "flos": 24714157980000.0, + "grad_norm": 1.6323093204805459, + "learning_rate": 1.336549293721756e-08, + "loss": 0.6323, + "num_input_tokens_seen": 347538520, + "step": 10692 + }, + { + "epoch": 0.9643324164675113, + "flos": 63280684965600.0, + "grad_norm": 0.6604011126671191, + "learning_rate": 1.3298150488249227e-08, + "loss": 0.5715, + "num_input_tokens_seen": 347625785, + "step": 10693 + }, + { + "epoch": 0.9644225999909817, + "flos": 18743791383360.0, + "grad_norm": 2.5303082068059304, + "learning_rate": 1.3230977555637401e-08, + "loss": 0.8143, + "num_input_tokens_seen": 347651880, + "step": 10694 + }, + { + "epoch": 0.9645127835144519, + "flos": 27122690808480.0, + "grad_norm": 1.5245039117349426, + "learning_rate": 1.3163974145113499e-08, + "loss": 0.6959, + "num_input_tokens_seen": 347683365, + "step": 10695 + }, + { + "epoch": 0.9646029670379221, + "flos": 35538988745760.0, + "grad_norm": 2.379323036502196, + "learning_rate": 1.3097140262394723e-08, + "loss": 0.6411, + "num_input_tokens_seen": 347712900, + "step": 10696 + }, + { + "epoch": 0.9646931505613925, + "flos": 16375593478080.0, + "grad_norm": 2.326252741332901, + "learning_rate": 1.303047591318318e-08, + "loss": 0.774, + "num_input_tokens_seen": 347737335, + "step": 10697 + }, + { + "epoch": 0.9647833340848627, + "flos": 35502630986880.0, + "grad_norm": 1.77804611916046, + "learning_rate": 1.2963981103167875e-08, + "loss": 0.6722, + "num_input_tokens_seen": 347765760, + "step": 10698 + }, + { + "epoch": 0.964873517608333, + "flos": 19904451935520.0, + "grad_norm": 1.5172560262869939, + "learning_rate": 1.2897655838021825e-08, + "loss": 0.724, + "num_input_tokens_seen": 347793270, + "step": 10699 + }, + { + "epoch": 0.9649637011318032, + "flos": 42935113588320.0, + "grad_norm": 2.3393361555212735, + "learning_rate": 1.2831500123404726e-08, + "loss": 0.7482, + "num_input_tokens_seen": 347826640, + "step": 10700 + }, + { + "epoch": 0.9650538846552735, + "flos": 25702396517280.0, + "grad_norm": 2.4340264439333534, + "learning_rate": 1.2765513964961172e-08, + "loss": 0.8895, + "num_input_tokens_seen": 347852240, + "step": 10701 + }, + { + "epoch": 0.9651440681787438, + "flos": 22095841794240.0, + "grad_norm": 1.8959244786524758, + "learning_rate": 1.2699697368321549e-08, + "loss": 0.7365, + "num_input_tokens_seen": 347880945, + "step": 10702 + }, + { + "epoch": 0.965234251702214, + "flos": 25810280361600.0, + "grad_norm": 1.9190040792925773, + "learning_rate": 1.2634050339101366e-08, + "loss": 0.7266, + "num_input_tokens_seen": 347912675, + "step": 10703 + }, + { + "epoch": 0.9653244352256842, + "flos": 27814937847840.0, + "grad_norm": 1.5351133073619743, + "learning_rate": 1.2568572882902361e-08, + "loss": 0.718, + "num_input_tokens_seen": 347944535, + "step": 10704 + }, + { + "epoch": 0.9654146187491546, + "flos": 26135827552320.0, + "grad_norm": 1.6311732190065078, + "learning_rate": 1.2503265005311402e-08, + "loss": 0.7131, + "num_input_tokens_seen": 347973890, + "step": 10705 + }, + { + "epoch": 0.9655048022726248, + "flos": 21658284915840.0, + "grad_norm": 2.4973485237443036, + "learning_rate": 1.2438126711900698e-08, + "loss": 0.7073, + "num_input_tokens_seen": 348000205, + "step": 10706 + }, + { + "epoch": 0.965594985796095, + "flos": 44029117293600.0, + "grad_norm": 1.5894573970051953, + "learning_rate": 1.2373158008228247e-08, + "loss": 0.5899, + "num_input_tokens_seen": 348034140, + "step": 10707 + }, + { + "epoch": 0.9656851693195653, + "flos": 24135816286080.0, + "grad_norm": 1.7036579942944645, + "learning_rate": 1.2308358899837833e-08, + "loss": 0.7138, + "num_input_tokens_seen": 348064360, + "step": 10708 + }, + { + "epoch": 0.9657753528430356, + "flos": 20669191474080.0, + "grad_norm": 2.0160100997792916, + "learning_rate": 1.224372939225815e-08, + "loss": 0.7205, + "num_input_tokens_seen": 348092180, + "step": 10709 + }, + { + "epoch": 0.9658655363665059, + "flos": 24646163021760.0, + "grad_norm": 1.9583451967353498, + "learning_rate": 1.2179269491003674e-08, + "loss": 0.8225, + "num_input_tokens_seen": 348121495, + "step": 10710 + }, + { + "epoch": 0.9659557198899761, + "flos": 22424362565760.0, + "grad_norm": 1.6861793153735904, + "learning_rate": 1.2114979201574894e-08, + "loss": 0.6146, + "num_input_tokens_seen": 348149755, + "step": 10711 + }, + { + "epoch": 0.9660459034134463, + "flos": 54519777484320.0, + "grad_norm": 1.515980472368473, + "learning_rate": 1.2050858529456975e-08, + "loss": 0.7568, + "num_input_tokens_seen": 348185990, + "step": 10712 + }, + { + "epoch": 0.9661360869369167, + "flos": 23554017804480.0, + "grad_norm": 2.1616651566303346, + "learning_rate": 1.1986907480121545e-08, + "loss": 0.7419, + "num_input_tokens_seen": 348217410, + "step": 10713 + }, + { + "epoch": 0.9662262704603869, + "flos": 27123397033920.0, + "grad_norm": 1.8000383478954116, + "learning_rate": 1.192312605902468e-08, + "loss": 0.7615, + "num_input_tokens_seen": 348248360, + "step": 10714 + }, + { + "epoch": 0.9663164539838571, + "flos": 70783509953760.0, + "grad_norm": 0.796101288015749, + "learning_rate": 1.1859514271608917e-08, + "loss": 0.6018, + "num_input_tokens_seen": 348325990, + "step": 10715 + }, + { + "epoch": 0.9664066375073274, + "flos": 16877353999200.0, + "grad_norm": 2.82529446915412, + "learning_rate": 1.1796072123301914e-08, + "loss": 0.7761, + "num_input_tokens_seen": 348348765, + "step": 10716 + }, + { + "epoch": 0.9664968210307977, + "flos": 65884096077600.0, + "grad_norm": 0.6355299353613384, + "learning_rate": 1.1732799619516897e-08, + "loss": 0.5807, + "num_input_tokens_seen": 348435870, + "step": 10717 + }, + { + "epoch": 0.9665870045542679, + "flos": 12440926850880.0, + "grad_norm": 2.0968723075383235, + "learning_rate": 1.1669696765652659e-08, + "loss": 0.7315, + "num_input_tokens_seen": 348462675, + "step": 10718 + }, + { + "epoch": 0.9666771880777382, + "flos": 22784186008800.0, + "grad_norm": 1.8458674927845116, + "learning_rate": 1.1606763567093336e-08, + "loss": 0.7699, + "num_input_tokens_seen": 348490885, + "step": 10719 + }, + { + "epoch": 0.9667673716012085, + "flos": 59716775424480.0, + "grad_norm": 0.6897121731657702, + "learning_rate": 1.1544000029208857e-08, + "loss": 0.571, + "num_input_tokens_seen": 348575700, + "step": 10720 + }, + { + "epoch": 0.9668575551246787, + "flos": 25411162748640.0, + "grad_norm": 1.818100071824993, + "learning_rate": 1.148140615735449e-08, + "loss": 0.6055, + "num_input_tokens_seen": 348605280, + "step": 10721 + }, + { + "epoch": 0.966947738648149, + "flos": 22966011972960.0, + "grad_norm": 1.8406096401124221, + "learning_rate": 1.1418981956871076e-08, + "loss": 0.715, + "num_input_tokens_seen": 348633040, + "step": 10722 + }, + { + "epoch": 0.9670379221716192, + "flos": 17906076138720.0, + "grad_norm": 2.7628369259668735, + "learning_rate": 1.1356727433085245e-08, + "loss": 0.7249, + "num_input_tokens_seen": 348659655, + "step": 10723 + }, + { + "epoch": 0.9671281056950896, + "flos": 21804496516320.0, + "grad_norm": 1.5286485800861938, + "learning_rate": 1.1294642591308524e-08, + "loss": 0.7677, + "num_input_tokens_seen": 348688605, + "step": 10724 + }, + { + "epoch": 0.9672182892185598, + "flos": 24646869247200.0, + "grad_norm": 1.565207005664444, + "learning_rate": 1.1232727436838452e-08, + "loss": 0.7344, + "num_input_tokens_seen": 348719495, + "step": 10725 + }, + { + "epoch": 0.96730847274203, + "flos": 18956213793600.0, + "grad_norm": 2.0449919447238893, + "learning_rate": 1.1170981974958138e-08, + "loss": 0.7866, + "num_input_tokens_seen": 348746285, + "step": 10726 + }, + { + "epoch": 0.9673986562655003, + "flos": 19761362594880.0, + "grad_norm": 1.6440645765397677, + "learning_rate": 1.1109406210936035e-08, + "loss": 0.8034, + "num_input_tokens_seen": 348774325, + "step": 10727 + }, + { + "epoch": 0.9674888397889706, + "flos": 17906261987520.0, + "grad_norm": 3.8317759017561683, + "learning_rate": 1.1048000150025939e-08, + "loss": 0.862, + "num_input_tokens_seen": 348802345, + "step": 10728 + }, + { + "epoch": 0.9675790233124408, + "flos": 22127776351680.0, + "grad_norm": 3.9633982461586936, + "learning_rate": 1.0986763797467213e-08, + "loss": 0.6359, + "num_input_tokens_seen": 348829440, + "step": 10729 + }, + { + "epoch": 0.9676692068359111, + "flos": 58181008924320.0, + "grad_norm": 0.5840994680021816, + "learning_rate": 1.0925697158485459e-08, + "loss": 0.5425, + "num_input_tokens_seen": 348926050, + "step": 10730 + }, + { + "epoch": 0.9677593903593813, + "flos": 31492052242560.0, + "grad_norm": 1.75506533107174, + "learning_rate": 1.0864800238290727e-08, + "loss": 0.7438, + "num_input_tokens_seen": 348957950, + "step": 10731 + }, + { + "epoch": 0.9678495738828516, + "flos": 16630692506400.0, + "grad_norm": 2.1010265021343493, + "learning_rate": 1.0804073042079309e-08, + "loss": 0.7048, + "num_input_tokens_seen": 348984000, + "step": 10732 + }, + { + "epoch": 0.9679397574063219, + "flos": 65377800845760.0, + "grad_norm": 0.6380737490346655, + "learning_rate": 1.0743515575032392e-08, + "loss": 0.5886, + "num_input_tokens_seen": 349072685, + "step": 10733 + }, + { + "epoch": 0.9680299409297921, + "flos": 32625015589920.0, + "grad_norm": 2.091350261977636, + "learning_rate": 1.0683127842317619e-08, + "loss": 0.7507, + "num_input_tokens_seen": 349103090, + "step": 10734 + }, + { + "epoch": 0.9681201244532623, + "flos": 28324912885920.0, + "grad_norm": 2.175363825918774, + "learning_rate": 1.0622909849087314e-08, + "loss": 0.7721, + "num_input_tokens_seen": 349130805, + "step": 10735 + }, + { + "epoch": 0.9682103079767327, + "flos": 21360211911360.0, + "grad_norm": 1.7887882872082004, + "learning_rate": 1.0562861600479588e-08, + "loss": 0.8034, + "num_input_tokens_seen": 349158915, + "step": 10736 + }, + { + "epoch": 0.9683004915002029, + "flos": 22970174986080.0, + "grad_norm": 2.45796105965916, + "learning_rate": 1.0502983101618345e-08, + "loss": 0.8358, + "num_input_tokens_seen": 349187030, + "step": 10737 + }, + { + "epoch": 0.9683906750236732, + "flos": 34883471162880.0, + "grad_norm": 1.986212953062365, + "learning_rate": 1.0443274357612386e-08, + "loss": 0.6239, + "num_input_tokens_seen": 349218130, + "step": 10738 + }, + { + "epoch": 0.9684808585471434, + "flos": 25484287133760.0, + "grad_norm": 1.8250265354683406, + "learning_rate": 1.0383735373556524e-08, + "loss": 0.7574, + "num_input_tokens_seen": 349247270, + "step": 10739 + }, + { + "epoch": 0.9685710420706137, + "flos": 25592914373280.0, + "grad_norm": 1.5416339750518142, + "learning_rate": 1.0324366154531139e-08, + "loss": 0.7701, + "num_input_tokens_seen": 349278780, + "step": 10740 + }, + { + "epoch": 0.968661225594084, + "flos": 12878149201440.0, + "grad_norm": 2.5490596271483272, + "learning_rate": 1.0265166705601735e-08, + "loss": 0.7037, + "num_input_tokens_seen": 349304145, + "step": 10741 + }, + { + "epoch": 0.9687514091175542, + "flos": 71137268992320.0, + "grad_norm": 0.6809525391597228, + "learning_rate": 1.0206137031819606e-08, + "loss": 0.6101, + "num_input_tokens_seen": 349397265, + "step": 10742 + }, + { + "epoch": 0.9688415926410244, + "flos": 25229373954240.0, + "grad_norm": 1.7495878960529576, + "learning_rate": 1.0147277138221388e-08, + "loss": 0.7529, + "num_input_tokens_seen": 349427695, + "step": 10743 + }, + { + "epoch": 0.9689317761644948, + "flos": 21985913613120.0, + "grad_norm": 2.2694599186361812, + "learning_rate": 1.0088587029829287e-08, + "loss": 0.7757, + "num_input_tokens_seen": 349454770, + "step": 10744 + }, + { + "epoch": 0.969021959687965, + "flos": 23298472739040.0, + "grad_norm": 1.6556566003107245, + "learning_rate": 1.003006671165152e-08, + "loss": 0.8456, + "num_input_tokens_seen": 349484270, + "step": 10745 + }, + { + "epoch": 0.9691121432114352, + "flos": 18962495483040.0, + "grad_norm": 1.9515177102366532, + "learning_rate": 9.971716188680978e-09, + "loss": 0.7373, + "num_input_tokens_seen": 349512075, + "step": 10746 + }, + { + "epoch": 0.9692023267349056, + "flos": 22022977597440.0, + "grad_norm": 2.011438324280726, + "learning_rate": 9.91353546589635e-09, + "loss": 0.7051, + "num_input_tokens_seen": 349540760, + "step": 10747 + }, + { + "epoch": 0.9692925102583758, + "flos": 29454828312960.0, + "grad_norm": 3.2255375968505295, + "learning_rate": 9.855524548262106e-09, + "loss": 0.6791, + "num_input_tokens_seen": 349574380, + "step": 10748 + }, + { + "epoch": 0.969382693781846, + "flos": 18160914978720.0, + "grad_norm": 1.7016526603807285, + "learning_rate": 9.797683440728288e-09, + "loss": 0.6857, + "num_input_tokens_seen": 349602300, + "step": 10749 + }, + { + "epoch": 0.9694728773053163, + "flos": 21144109694880.0, + "grad_norm": 1.7828488452359348, + "learning_rate": 9.740012148229836e-09, + "loss": 0.6541, + "num_input_tokens_seen": 349631070, + "step": 10750 + }, + { + "epoch": 0.9695630608287866, + "flos": 27889734872160.0, + "grad_norm": 1.6013299540585435, + "learning_rate": 9.682510675687705e-09, + "loss": 0.6906, + "num_input_tokens_seen": 349660500, + "step": 10751 + }, + { + "epoch": 0.9696532443522569, + "flos": 23662050327840.0, + "grad_norm": 2.144112301120082, + "learning_rate": 9.625179028008191e-09, + "loss": 0.6648, + "num_input_tokens_seen": 349689665, + "step": 10752 + }, + { + "epoch": 0.9697434278757271, + "flos": 28866487953600.0, + "grad_norm": 2.03416280972955, + "learning_rate": 9.568017210083379e-09, + "loss": 0.5866, + "num_input_tokens_seen": 349720135, + "step": 10753 + }, + { + "epoch": 0.9698336113991973, + "flos": 29120546228640.0, + "grad_norm": 2.278879894042722, + "learning_rate": 9.511025226790259e-09, + "loss": 0.7176, + "num_input_tokens_seen": 349749920, + "step": 10754 + }, + { + "epoch": 0.9699237949226677, + "flos": 22095767454720.0, + "grad_norm": 1.6253817183313108, + "learning_rate": 9.454203082992052e-09, + "loss": 0.715, + "num_input_tokens_seen": 349776995, + "step": 10755 + }, + { + "epoch": 0.9700139784461379, + "flos": 21075668699520.0, + "grad_norm": 2.269610352705215, + "learning_rate": 9.3975507835371e-09, + "loss": 0.6477, + "num_input_tokens_seen": 349804420, + "step": 10756 + }, + { + "epoch": 0.9701041619696081, + "flos": 70114159486560.0, + "grad_norm": 0.6219678521925354, + "learning_rate": 9.341068333259094e-09, + "loss": 0.5646, + "num_input_tokens_seen": 349904840, + "step": 10757 + }, + { + "epoch": 0.9701943454930784, + "flos": 22605779662560.0, + "grad_norm": 1.6496807309687649, + "learning_rate": 9.28475573697729e-09, + "loss": 0.715, + "num_input_tokens_seen": 349932395, + "step": 10758 + }, + { + "epoch": 0.9702845290165487, + "flos": 19218672434400.0, + "grad_norm": 1.6900937668333929, + "learning_rate": 9.228612999497177e-09, + "loss": 0.7426, + "num_input_tokens_seen": 349960050, + "step": 10759 + }, + { + "epoch": 0.970374712540019, + "flos": 23948600706720.0, + "grad_norm": 2.4712219417293184, + "learning_rate": 9.172640125608478e-09, + "loss": 0.6832, + "num_input_tokens_seen": 349988450, + "step": 10760 + }, + { + "epoch": 0.9704648960634892, + "flos": 23080177506720.0, + "grad_norm": 1.3696764715030518, + "learning_rate": 9.116837120087817e-09, + "loss": 0.754, + "num_input_tokens_seen": 350020825, + "step": 10761 + }, + { + "epoch": 0.9705550795869594, + "flos": 18124631559360.0, + "grad_norm": 2.2824173122440703, + "learning_rate": 9.061203987695832e-09, + "loss": 0.807, + "num_input_tokens_seen": 350047495, + "step": 10762 + }, + { + "epoch": 0.9706452631104298, + "flos": 21840631256640.0, + "grad_norm": 2.4405393968875013, + "learning_rate": 9.005740733180055e-09, + "loss": 0.6879, + "num_input_tokens_seen": 350076535, + "step": 10763 + }, + { + "epoch": 0.9707354466339, + "flos": 20201335507680.0, + "grad_norm": 1.8610728870407778, + "learning_rate": 8.950447361272483e-09, + "loss": 0.8252, + "num_input_tokens_seen": 350103430, + "step": 10764 + }, + { + "epoch": 0.9708256301573702, + "flos": 19836196788960.0, + "grad_norm": 2.5709817588986983, + "learning_rate": 8.895323876691784e-09, + "loss": 0.7455, + "num_input_tokens_seen": 350129425, + "step": 10765 + }, + { + "epoch": 0.9709158136808405, + "flos": 14079590713920.0, + "grad_norm": 5.194397901756043, + "learning_rate": 8.840370284140419e-09, + "loss": 0.7918, + "num_input_tokens_seen": 350152995, + "step": 10766 + }, + { + "epoch": 0.9710059972043108, + "flos": 21549137299680.0, + "grad_norm": 2.303940577621563, + "learning_rate": 8.78558658830797e-09, + "loss": 0.6584, + "num_input_tokens_seen": 350179750, + "step": 10767 + }, + { + "epoch": 0.971096180727781, + "flos": 67828341236640.0, + "grad_norm": 0.6931295954394576, + "learning_rate": 8.730972793868696e-09, + "loss": 0.5766, + "num_input_tokens_seen": 350268855, + "step": 10768 + }, + { + "epoch": 0.9711863642512513, + "flos": 27336860197440.0, + "grad_norm": 2.0532977259426075, + "learning_rate": 8.67652890548265e-09, + "loss": 0.6996, + "num_input_tokens_seen": 350297830, + "step": 10769 + }, + { + "epoch": 0.9712765477747216, + "flos": 24900741673440.0, + "grad_norm": 1.5756966821146448, + "learning_rate": 8.622254927795004e-09, + "loss": 0.7234, + "num_input_tokens_seen": 350330795, + "step": 10770 + }, + { + "epoch": 0.9713667312981918, + "flos": 36627268308000.0, + "grad_norm": 1.847173906075245, + "learning_rate": 8.568150865436941e-09, + "loss": 0.6912, + "num_input_tokens_seen": 350360635, + "step": 10771 + }, + { + "epoch": 0.9714569148216621, + "flos": 42462611401920.0, + "grad_norm": 3.321386545035517, + "learning_rate": 8.514216723024991e-09, + "loss": 0.7528, + "num_input_tokens_seen": 350395705, + "step": 10772 + }, + { + "epoch": 0.9715470983451323, + "flos": 23661938818560.0, + "grad_norm": 2.795465209951873, + "learning_rate": 8.460452505161031e-09, + "loss": 0.6627, + "num_input_tokens_seen": 350422715, + "step": 10773 + }, + { + "epoch": 0.9716372818686027, + "flos": 21622150175520.0, + "grad_norm": 1.8852982317009779, + "learning_rate": 8.4068582164325e-09, + "loss": 0.7632, + "num_input_tokens_seen": 350451285, + "step": 10774 + }, + { + "epoch": 0.9717274653920729, + "flos": 19470797881920.0, + "grad_norm": 1.8819626279033033, + "learning_rate": 8.353433861412406e-09, + "loss": 0.722, + "num_input_tokens_seen": 350478780, + "step": 10775 + }, + { + "epoch": 0.9718176489155431, + "flos": 30507976718400.0, + "grad_norm": 1.6676879932402977, + "learning_rate": 8.300179444658883e-09, + "loss": 0.6947, + "num_input_tokens_seen": 350510810, + "step": 10776 + }, + { + "epoch": 0.9719078324390134, + "flos": 19144878993600.0, + "grad_norm": 2.284552809113454, + "learning_rate": 8.247094970716296e-09, + "loss": 0.8315, + "num_input_tokens_seen": 350537785, + "step": 10777 + }, + { + "epoch": 0.9719980159624837, + "flos": 23402230740000.0, + "grad_norm": 1.8445348847033094, + "learning_rate": 8.19418044411413e-09, + "loss": 0.7757, + "num_input_tokens_seen": 350568300, + "step": 10778 + }, + { + "epoch": 0.9720881994859539, + "flos": 23327359376160.0, + "grad_norm": 1.8517832564082626, + "learning_rate": 8.141435869367219e-09, + "loss": 0.7415, + "num_input_tokens_seen": 350596440, + "step": 10779 + }, + { + "epoch": 0.9721783830094242, + "flos": 22601096272800.0, + "grad_norm": 1.918063035189683, + "learning_rate": 8.088861250975742e-09, + "loss": 0.776, + "num_input_tokens_seen": 350623900, + "step": 10780 + }, + { + "epoch": 0.9722685665328944, + "flos": 22054057250400.0, + "grad_norm": 2.2084004973748352, + "learning_rate": 8.036456593426111e-09, + "loss": 0.7095, + "num_input_tokens_seen": 350650730, + "step": 10781 + }, + { + "epoch": 0.9723587500563647, + "flos": 15355457553120.0, + "grad_norm": 2.0735651915498527, + "learning_rate": 7.984221901189415e-09, + "loss": 0.7736, + "num_input_tokens_seen": 350675960, + "step": 10782 + }, + { + "epoch": 0.972448933579835, + "flos": 21908812063680.0, + "grad_norm": 1.9707996658557185, + "learning_rate": 7.932157178722976e-09, + "loss": 0.6617, + "num_input_tokens_seen": 350704975, + "step": 10783 + }, + { + "epoch": 0.9725391171033052, + "flos": 29272965179040.0, + "grad_norm": 1.8509827684067182, + "learning_rate": 7.880262430468799e-09, + "loss": 0.7062, + "num_input_tokens_seen": 350734150, + "step": 10784 + }, + { + "epoch": 0.9726293006267754, + "flos": 21221285583840.0, + "grad_norm": 2.1424715194678683, + "learning_rate": 7.828537660855339e-09, + "loss": 0.6712, + "num_input_tokens_seen": 350762025, + "step": 10785 + }, + { + "epoch": 0.9727194841502458, + "flos": 23149584915840.0, + "grad_norm": 3.062648224649579, + "learning_rate": 7.776982874295512e-09, + "loss": 0.7557, + "num_input_tokens_seen": 350791945, + "step": 10786 + }, + { + "epoch": 0.972809667673716, + "flos": 29266832168640.0, + "grad_norm": 2.189525102832538, + "learning_rate": 7.725598075188688e-09, + "loss": 0.647, + "num_input_tokens_seen": 350821270, + "step": 10787 + }, + { + "epoch": 0.9728998511971862, + "flos": 20673837694080.0, + "grad_norm": 1.9282225326031868, + "learning_rate": 7.674383267918916e-09, + "loss": 0.6905, + "num_input_tokens_seen": 350851145, + "step": 10788 + }, + { + "epoch": 0.9729900347206565, + "flos": 22132162383360.0, + "grad_norm": 1.8480455823082234, + "learning_rate": 7.623338456856476e-09, + "loss": 0.7255, + "num_input_tokens_seen": 350880800, + "step": 10789 + }, + { + "epoch": 0.9730802182441268, + "flos": 21840631256640.0, + "grad_norm": 2.322975659726105, + "learning_rate": 7.572463646356554e-09, + "loss": 0.6316, + "num_input_tokens_seen": 350907300, + "step": 10790 + }, + { + "epoch": 0.9731704017675971, + "flos": 20055384095520.0, + "grad_norm": 3.7629836066317504, + "learning_rate": 7.521758840760339e-09, + "loss": 0.7088, + "num_input_tokens_seen": 350936485, + "step": 10791 + }, + { + "epoch": 0.9732605852910673, + "flos": 18707470794240.0, + "grad_norm": 2.582483572302949, + "learning_rate": 7.471224044393931e-09, + "loss": 0.6958, + "num_input_tokens_seen": 350959700, + "step": 10792 + }, + { + "epoch": 0.9733507688145376, + "flos": 26973691476000.0, + "grad_norm": 1.7542657632389995, + "learning_rate": 7.420859261569434e-09, + "loss": 0.7971, + "num_input_tokens_seen": 350989520, + "step": 10793 + }, + { + "epoch": 0.9734409523380079, + "flos": 27734491020000.0, + "grad_norm": 2.0906434427539824, + "learning_rate": 7.370664496584078e-09, + "loss": 0.7975, + "num_input_tokens_seen": 351015665, + "step": 10794 + }, + { + "epoch": 0.9735311358614781, + "flos": 35320656343680.0, + "grad_norm": 2.5916466514902106, + "learning_rate": 7.3206397537211026e-09, + "loss": 0.6632, + "num_input_tokens_seen": 351049335, + "step": 10795 + }, + { + "epoch": 0.9736213193849483, + "flos": 35867174989440.0, + "grad_norm": 2.0827915886003066, + "learning_rate": 7.270785037248428e-09, + "loss": 0.7044, + "num_input_tokens_seen": 351079760, + "step": 10796 + }, + { + "epoch": 0.9737115029084187, + "flos": 23079917318400.0, + "grad_norm": 2.09416798077608, + "learning_rate": 7.221100351420428e-09, + "loss": 0.7001, + "num_input_tokens_seen": 351108635, + "step": 10797 + }, + { + "epoch": 0.9738016864318889, + "flos": 38048491843200.0, + "grad_norm": 1.9819215641187644, + "learning_rate": 7.171585700475935e-09, + "loss": 0.6654, + "num_input_tokens_seen": 351142030, + "step": 10798 + }, + { + "epoch": 0.9738918699553591, + "flos": 22054057250400.0, + "grad_norm": 2.06430668162598, + "learning_rate": 7.122241088640235e-09, + "loss": 0.7934, + "num_input_tokens_seen": 351170125, + "step": 10799 + }, + { + "epoch": 0.9739820534788294, + "flos": 27341878115040.0, + "grad_norm": 2.1369903224499236, + "learning_rate": 7.073066520123516e-09, + "loss": 0.8613, + "num_input_tokens_seen": 351198685, + "step": 10800 + }, + { + "epoch": 0.9740722370022997, + "flos": 20638372009440.0, + "grad_norm": 2.573119604676293, + "learning_rate": 7.0240619991217555e-09, + "loss": 0.6324, + "num_input_tokens_seen": 351224855, + "step": 10801 + }, + { + "epoch": 0.97416242052577, + "flos": 19983486312480.0, + "grad_norm": 5.1864556789551015, + "learning_rate": 6.975227529816052e-09, + "loss": 0.8135, + "num_input_tokens_seen": 351251455, + "step": 10802 + }, + { + "epoch": 0.9742526040492402, + "flos": 19509014128800.0, + "grad_norm": 2.00253730996339, + "learning_rate": 6.926563116373296e-09, + "loss": 0.7822, + "num_input_tokens_seen": 351280050, + "step": 10803 + }, + { + "epoch": 0.9743427875727104, + "flos": 23075791475040.0, + "grad_norm": 1.9741337143850537, + "learning_rate": 6.878068762945943e-09, + "loss": 0.6736, + "num_input_tokens_seen": 351308090, + "step": 10804 + }, + { + "epoch": 0.9744329710961808, + "flos": 29417987347200.0, + "grad_norm": 1.7913569101292388, + "learning_rate": 6.829744473671794e-09, + "loss": 0.7732, + "num_input_tokens_seen": 351338770, + "step": 10805 + }, + { + "epoch": 0.974523154619651, + "flos": 63117226596480.0, + "grad_norm": 0.5850808047868563, + "learning_rate": 6.781590252674219e-09, + "loss": 0.477, + "num_input_tokens_seen": 351432795, + "step": 10806 + }, + { + "epoch": 0.9746133381431212, + "flos": 33719688350880.0, + "grad_norm": 1.6081103198274282, + "learning_rate": 6.733606104061484e-09, + "loss": 0.7452, + "num_input_tokens_seen": 351470125, + "step": 10807 + }, + { + "epoch": 0.9747035216665915, + "flos": 25957421206080.0, + "grad_norm": 1.6705119914012971, + "learning_rate": 6.6857920319283165e-09, + "loss": 0.6807, + "num_input_tokens_seen": 351502885, + "step": 10808 + }, + { + "epoch": 0.9747937051900618, + "flos": 24572332411200.0, + "grad_norm": 1.9546103360857445, + "learning_rate": 6.638148040354563e-09, + "loss": 0.7114, + "num_input_tokens_seen": 351533750, + "step": 10809 + }, + { + "epoch": 0.974883888713532, + "flos": 22204729222080.0, + "grad_norm": 1.6776852761523782, + "learning_rate": 6.590674133405194e-09, + "loss": 0.6961, + "num_input_tokens_seen": 351560180, + "step": 10810 + }, + { + "epoch": 0.9749740722370023, + "flos": 30584855249280.0, + "grad_norm": 1.815289160023017, + "learning_rate": 6.5433703151311914e-09, + "loss": 0.7501, + "num_input_tokens_seen": 351591040, + "step": 10811 + }, + { + "epoch": 0.9750642557604725, + "flos": 35903644257600.0, + "grad_norm": 1.80239868344912, + "learning_rate": 6.49623658956866e-09, + "loss": 0.7383, + "num_input_tokens_seen": 351621290, + "step": 10812 + }, + { + "epoch": 0.9751544392839429, + "flos": 14590643675040.0, + "grad_norm": 2.7079925811812458, + "learning_rate": 6.44927296073905e-09, + "loss": 0.6166, + "num_input_tokens_seen": 351648110, + "step": 10813 + }, + { + "epoch": 0.9752446228074131, + "flos": 22459939759680.0, + "grad_norm": 1.7985131798802936, + "learning_rate": 6.402479432649821e-09, + "loss": 0.8237, + "num_input_tokens_seen": 351674760, + "step": 10814 + }, + { + "epoch": 0.9753348063308833, + "flos": 23113450175520.0, + "grad_norm": 1.861464152214065, + "learning_rate": 6.355856009293781e-09, + "loss": 0.7313, + "num_input_tokens_seen": 351701350, + "step": 10815 + }, + { + "epoch": 0.9754249898543537, + "flos": 25812250358880.0, + "grad_norm": 3.6917132860487905, + "learning_rate": 6.3094026946488575e-09, + "loss": 0.7105, + "num_input_tokens_seen": 351727825, + "step": 10816 + }, + { + "epoch": 0.9755151733778239, + "flos": 26322411245760.0, + "grad_norm": 2.2888636584059, + "learning_rate": 6.2631194926787704e-09, + "loss": 0.7373, + "num_input_tokens_seen": 351756410, + "step": 10817 + }, + { + "epoch": 0.9756053569012941, + "flos": 27123434203680.0, + "grad_norm": 1.7017843586981052, + "learning_rate": 6.217006407332581e-09, + "loss": 0.6353, + "num_input_tokens_seen": 351784950, + "step": 10818 + }, + { + "epoch": 0.9756955404247644, + "flos": 25737564843840.0, + "grad_norm": 2.3425765087942043, + "learning_rate": 6.1710634425453654e-09, + "loss": 0.6846, + "num_input_tokens_seen": 351814955, + "step": 10819 + }, + { + "epoch": 0.9757857239482347, + "flos": 25224393206400.0, + "grad_norm": 1.8068773599119026, + "learning_rate": 6.1252906022366544e-09, + "loss": 0.7577, + "num_input_tokens_seen": 351843490, + "step": 10820 + }, + { + "epoch": 0.9758759074717049, + "flos": 17760570763680.0, + "grad_norm": 6.204904142587633, + "learning_rate": 6.079687890312213e-09, + "loss": 0.7152, + "num_input_tokens_seen": 351870760, + "step": 10821 + }, + { + "epoch": 0.9759660909951752, + "flos": 26139879056160.0, + "grad_norm": 1.7427454854589073, + "learning_rate": 6.034255310663372e-09, + "loss": 0.7172, + "num_input_tokens_seen": 351902340, + "step": 10822 + }, + { + "epoch": 0.9760562745186454, + "flos": 29819037787680.0, + "grad_norm": 1.8316557607190131, + "learning_rate": 5.988992867166143e-09, + "loss": 0.8297, + "num_input_tokens_seen": 351932115, + "step": 10823 + }, + { + "epoch": 0.9761464580421157, + "flos": 24900964692000.0, + "grad_norm": 2.7493846378213855, + "learning_rate": 5.943900563682991e-09, + "loss": 0.7779, + "num_input_tokens_seen": 351959930, + "step": 10824 + }, + { + "epoch": 0.976236641565586, + "flos": 66262169872800.0, + "grad_norm": 0.7435119857129842, + "learning_rate": 5.898978404061506e-09, + "loss": 0.655, + "num_input_tokens_seen": 352050680, + "step": 10825 + }, + { + "epoch": 0.9763268250890562, + "flos": 21908143008000.0, + "grad_norm": 2.8708595647476263, + "learning_rate": 5.85422639213462e-09, + "loss": 0.7335, + "num_input_tokens_seen": 352079050, + "step": 10826 + }, + { + "epoch": 0.9764170086125264, + "flos": 30141722906880.0, + "grad_norm": 1.7367012301036218, + "learning_rate": 5.809644531720614e-09, + "loss": 0.7479, + "num_input_tokens_seen": 352110275, + "step": 10827 + }, + { + "epoch": 0.9765071921359968, + "flos": 23441450570400.0, + "grad_norm": 1.7541013069692082, + "learning_rate": 5.765232826623556e-09, + "loss": 0.7646, + "num_input_tokens_seen": 352141315, + "step": 10828 + }, + { + "epoch": 0.976597375659467, + "flos": 24169758010560.0, + "grad_norm": 1.8604485754314606, + "learning_rate": 5.720991280633081e-09, + "loss": 0.7174, + "num_input_tokens_seen": 352168770, + "step": 10829 + }, + { + "epoch": 0.9766875591829373, + "flos": 19654965540960.0, + "grad_norm": 1.725927382810254, + "learning_rate": 5.676919897523724e-09, + "loss": 0.7231, + "num_input_tokens_seen": 352196615, + "step": 10830 + }, + { + "epoch": 0.9767777427064075, + "flos": 45412310430720.0, + "grad_norm": 1.4764344497944346, + "learning_rate": 5.633018681056256e-09, + "loss": 0.7401, + "num_input_tokens_seen": 352229610, + "step": 10831 + }, + { + "epoch": 0.9768679262298778, + "flos": 26062294299840.0, + "grad_norm": 2.4846854345835787, + "learning_rate": 5.589287634976569e-09, + "loss": 0.7396, + "num_input_tokens_seen": 352256600, + "step": 10832 + }, + { + "epoch": 0.9769581097533481, + "flos": 56559528957600.0, + "grad_norm": 1.6634709555977187, + "learning_rate": 5.5457267630159014e-09, + "loss": 0.7086, + "num_input_tokens_seen": 352292935, + "step": 10833 + }, + { + "epoch": 0.9770482932768183, + "flos": 23803244010720.0, + "grad_norm": 1.8096826295123452, + "learning_rate": 5.5023360688910555e-09, + "loss": 0.7713, + "num_input_tokens_seen": 352318695, + "step": 10834 + }, + { + "epoch": 0.9771384768002885, + "flos": 22423396152000.0, + "grad_norm": 1.67150190236779, + "learning_rate": 5.459115556304183e-09, + "loss": 0.7842, + "num_input_tokens_seen": 352346175, + "step": 10835 + }, + { + "epoch": 0.9772286603237589, + "flos": 37106572560480.0, + "grad_norm": 2.31182335929194, + "learning_rate": 5.416065228943889e-09, + "loss": 0.6699, + "num_input_tokens_seen": 352375970, + "step": 10836 + }, + { + "epoch": 0.9773188438472291, + "flos": 22934003076000.0, + "grad_norm": 2.867844176386112, + "learning_rate": 5.373185090482568e-09, + "loss": 0.6884, + "num_input_tokens_seen": 352403920, + "step": 10837 + }, + { + "epoch": 0.9774090273706993, + "flos": 18562114098240.0, + "grad_norm": 1.891870009910674, + "learning_rate": 5.330475144579516e-09, + "loss": 0.7119, + "num_input_tokens_seen": 352431775, + "step": 10838 + }, + { + "epoch": 0.9774992108941697, + "flos": 25884110972160.0, + "grad_norm": 1.6764593997479618, + "learning_rate": 5.2879353948787065e-09, + "loss": 0.8197, + "num_input_tokens_seen": 352459060, + "step": 10839 + }, + { + "epoch": 0.9775893944176399, + "flos": 20164940579040.0, + "grad_norm": 2.18136553963869, + "learning_rate": 5.245565845010125e-09, + "loss": 0.7985, + "num_input_tokens_seen": 352485975, + "step": 10840 + }, + { + "epoch": 0.9776795779411102, + "flos": 29195343252960.0, + "grad_norm": 1.7055726268271545, + "learning_rate": 5.2033664985886575e-09, + "loss": 0.6676, + "num_input_tokens_seen": 352516820, + "step": 10841 + }, + { + "epoch": 0.9777697614645804, + "flos": 20449037753760.0, + "grad_norm": 1.9110373120819897, + "learning_rate": 5.161337359215201e-09, + "loss": 0.8813, + "num_input_tokens_seen": 352543995, + "step": 10842 + }, + { + "epoch": 0.9778599449880507, + "flos": 22096287831360.0, + "grad_norm": 1.9729366292081187, + "learning_rate": 5.119478430475999e-09, + "loss": 0.7476, + "num_input_tokens_seen": 352571775, + "step": 10843 + }, + { + "epoch": 0.977950128511521, + "flos": 23224790807520.0, + "grad_norm": 2.696126444553963, + "learning_rate": 5.077789715942416e-09, + "loss": 0.6892, + "num_input_tokens_seen": 352601345, + "step": 10844 + }, + { + "epoch": 0.9780403120349912, + "flos": 18962123785440.0, + "grad_norm": 2.102227882075649, + "learning_rate": 5.036271219171606e-09, + "loss": 0.7595, + "num_input_tokens_seen": 352627310, + "step": 10845 + }, + { + "epoch": 0.9781304955584614, + "flos": 19208004713280.0, + "grad_norm": 2.1528644243530004, + "learning_rate": 4.994922943706514e-09, + "loss": 0.7727, + "num_input_tokens_seen": 352649960, + "step": 10846 + }, + { + "epoch": 0.9782206790819318, + "flos": 19542138118560.0, + "grad_norm": 1.7114684792061354, + "learning_rate": 4.953744893074763e-09, + "loss": 0.72, + "num_input_tokens_seen": 352675815, + "step": 10847 + }, + { + "epoch": 0.978310862605402, + "flos": 18488803864320.0, + "grad_norm": 2.1039964839992953, + "learning_rate": 4.912737070789985e-09, + "loss": 0.6632, + "num_input_tokens_seen": 352702635, + "step": 10848 + }, + { + "epoch": 0.9784010461288722, + "flos": 22241904715680.0, + "grad_norm": 2.1386494674553886, + "learning_rate": 4.871899480351604e-09, + "loss": 0.807, + "num_input_tokens_seen": 352731750, + "step": 10849 + }, + { + "epoch": 0.9784912296523425, + "flos": 23735806598880.0, + "grad_norm": 1.7230462910770732, + "learning_rate": 4.831232125243501e-09, + "loss": 0.7499, + "num_input_tokens_seen": 352761905, + "step": 10850 + }, + { + "epoch": 0.9785814131758128, + "flos": 16485001282560.0, + "grad_norm": 2.142811792892867, + "learning_rate": 4.7907350089360086e-09, + "loss": 0.8381, + "num_input_tokens_seen": 352788860, + "step": 10851 + }, + { + "epoch": 0.978671596699283, + "flos": 17615028218880.0, + "grad_norm": 2.0799774861951303, + "learning_rate": 4.750408134884365e-09, + "loss": 0.6305, + "num_input_tokens_seen": 352816930, + "step": 10852 + }, + { + "epoch": 0.9787617802227533, + "flos": 29892125003040.0, + "grad_norm": 1.6705523811994938, + "learning_rate": 4.710251506529816e-09, + "loss": 0.6474, + "num_input_tokens_seen": 352848285, + "step": 10853 + }, + { + "epoch": 0.9788519637462235, + "flos": 31240001134560.0, + "grad_norm": 1.8068083484222854, + "learning_rate": 4.6702651272982894e-09, + "loss": 0.7502, + "num_input_tokens_seen": 352878940, + "step": 10854 + }, + { + "epoch": 0.9789421472696939, + "flos": 22670615191200.0, + "grad_norm": 2.3401803871464195, + "learning_rate": 4.630449000602166e-09, + "loss": 0.7087, + "num_input_tokens_seen": 352904920, + "step": 10855 + }, + { + "epoch": 0.9790323307931641, + "flos": 19355442915840.0, + "grad_norm": 3.2148780125661305, + "learning_rate": 4.590803129838283e-09, + "loss": 0.8005, + "num_input_tokens_seen": 352929320, + "step": 10856 + }, + { + "epoch": 0.9791225143166343, + "flos": 19945641763200.0, + "grad_norm": 3.4651989330087107, + "learning_rate": 4.551327518389714e-09, + "loss": 0.7435, + "num_input_tokens_seen": 352955020, + "step": 10857 + }, + { + "epoch": 0.9792126978401046, + "flos": 20269516314720.0, + "grad_norm": 2.607353203830027, + "learning_rate": 4.512022169624652e-09, + "loss": 0.7316, + "num_input_tokens_seen": 352980530, + "step": 10858 + }, + { + "epoch": 0.9793028813635749, + "flos": 20493349841280.0, + "grad_norm": 1.8633539917380166, + "learning_rate": 4.472887086896637e-09, + "loss": 0.7102, + "num_input_tokens_seen": 353009170, + "step": 10859 + }, + { + "epoch": 0.9793930648870451, + "flos": 27921929617920.0, + "grad_norm": 2.1397798080890738, + "learning_rate": 4.433922273545443e-09, + "loss": 0.6825, + "num_input_tokens_seen": 353038010, + "step": 10860 + }, + { + "epoch": 0.9794832484105154, + "flos": 69785973242880.0, + "grad_norm": 0.5986798316200826, + "learning_rate": 4.395127732895299e-09, + "loss": 0.5797, + "num_input_tokens_seen": 353129920, + "step": 10861 + }, + { + "epoch": 0.9795734319339856, + "flos": 25847195666880.0, + "grad_norm": 50.17446125769563, + "learning_rate": 4.356503468256445e-09, + "loss": 0.8166, + "num_input_tokens_seen": 353157735, + "step": 10862 + }, + { + "epoch": 0.979663615457456, + "flos": 19071866117760.0, + "grad_norm": 2.523688997662229, + "learning_rate": 4.318049482924913e-09, + "loss": 0.7508, + "num_input_tokens_seen": 353182395, + "step": 10863 + }, + { + "epoch": 0.9797537989809262, + "flos": 20165052088320.0, + "grad_norm": 5.81805023673128, + "learning_rate": 4.279765780181188e-09, + "loss": 0.705, + "num_input_tokens_seen": 353206270, + "step": 10864 + }, + { + "epoch": 0.9798439825043964, + "flos": 24900927522240.0, + "grad_norm": 2.2085113693542295, + "learning_rate": 4.241652363291992e-09, + "loss": 0.7516, + "num_input_tokens_seen": 353234695, + "step": 10865 + }, + { + "epoch": 0.9799341660278668, + "flos": 13789026000960.0, + "grad_norm": 3.942335883034403, + "learning_rate": 4.203709235509834e-09, + "loss": 0.7127, + "num_input_tokens_seen": 353259755, + "step": 10866 + }, + { + "epoch": 0.980024349551337, + "flos": 23443978114080.0, + "grad_norm": 1.7940192043845806, + "learning_rate": 4.165936400071679e-09, + "loss": 0.7598, + "num_input_tokens_seen": 353289070, + "step": 10867 + }, + { + "epoch": 0.9801145330748072, + "flos": 21476719140000.0, + "grad_norm": 1.7496535775762208, + "learning_rate": 4.12833386020095e-09, + "loss": 0.6863, + "num_input_tokens_seen": 353317870, + "step": 10868 + }, + { + "epoch": 0.9802047165982775, + "flos": 17940947107200.0, + "grad_norm": 2.21619579887918, + "learning_rate": 4.090901619105746e-09, + "loss": 0.7725, + "num_input_tokens_seen": 353344480, + "step": 10869 + }, + { + "epoch": 0.9802949001217478, + "flos": 13678205745600.0, + "grad_norm": 1.8617329752607101, + "learning_rate": 4.053639679980181e-09, + "loss": 0.6294, + "num_input_tokens_seen": 353370600, + "step": 10870 + }, + { + "epoch": 0.980385083645218, + "flos": 22861213218720.0, + "grad_norm": 1.9050079426637407, + "learning_rate": 4.01654804600371e-09, + "loss": 0.7446, + "num_input_tokens_seen": 353397435, + "step": 10871 + }, + { + "epoch": 0.9804752671686883, + "flos": 20456025668640.0, + "grad_norm": 1.7110213167745483, + "learning_rate": 3.9796267203409114e-09, + "loss": 0.7459, + "num_input_tokens_seen": 353424905, + "step": 10872 + }, + { + "epoch": 0.9805654506921585, + "flos": 25916565906240.0, + "grad_norm": 1.6531902336417301, + "learning_rate": 3.942875706142379e-09, + "loss": 0.7161, + "num_input_tokens_seen": 353452735, + "step": 10873 + }, + { + "epoch": 0.9806556342156288, + "flos": 20016201434880.0, + "grad_norm": 2.335553600188668, + "learning_rate": 3.906295006543825e-09, + "loss": 0.7982, + "num_input_tokens_seen": 353481650, + "step": 10874 + }, + { + "epoch": 0.9807458177390991, + "flos": 24752002529280.0, + "grad_norm": 1.9316362433002945, + "learning_rate": 3.8698846246665305e-09, + "loss": 0.7144, + "num_input_tokens_seen": 353510710, + "step": 10875 + }, + { + "epoch": 0.9808360012625693, + "flos": 19507229980320.0, + "grad_norm": 1.8326635509953404, + "learning_rate": 3.833644563617344e-09, + "loss": 0.7428, + "num_input_tokens_seen": 353537900, + "step": 10876 + }, + { + "epoch": 0.9809261847860395, + "flos": 23297655004320.0, + "grad_norm": 2.7348239203196587, + "learning_rate": 3.797574826488237e-09, + "loss": 0.7482, + "num_input_tokens_seen": 353567515, + "step": 10877 + }, + { + "epoch": 0.9810163683095099, + "flos": 29491446260160.0, + "grad_norm": 1.7304389210491689, + "learning_rate": 3.761675416356969e-09, + "loss": 0.6674, + "num_input_tokens_seen": 353596820, + "step": 10878 + }, + { + "epoch": 0.9811065518329801, + "flos": 34588334569440.0, + "grad_norm": 2.426141242253233, + "learning_rate": 3.725946336286867e-09, + "loss": 0.6456, + "num_input_tokens_seen": 353627125, + "step": 10879 + }, + { + "epoch": 0.9811967353564504, + "flos": 25014795697920.0, + "grad_norm": 1.7127073705376865, + "learning_rate": 3.6903875893261604e-09, + "loss": 0.7108, + "num_input_tokens_seen": 353660825, + "step": 10880 + }, + { + "epoch": 0.9812869188799206, + "flos": 24063509635680.0, + "grad_norm": 2.1480197569721686, + "learning_rate": 3.6549991785093105e-09, + "loss": 0.668, + "num_input_tokens_seen": 353690425, + "step": 10881 + }, + { + "epoch": 0.9813771024033909, + "flos": 27086295879840.0, + "grad_norm": 1.8875834692884137, + "learning_rate": 3.6197811068554575e-09, + "loss": 0.726, + "num_input_tokens_seen": 353719645, + "step": 10882 + }, + { + "epoch": 0.9814672859268612, + "flos": 23443048870080.0, + "grad_norm": 1.709332553817722, + "learning_rate": 3.584733377369975e-09, + "loss": 0.8214, + "num_input_tokens_seen": 353749520, + "step": 10883 + }, + { + "epoch": 0.9815574694503314, + "flos": 21804459346560.0, + "grad_norm": 2.0638096437205435, + "learning_rate": 3.549855993043138e-09, + "loss": 0.7163, + "num_input_tokens_seen": 353776300, + "step": 10884 + }, + { + "epoch": 0.9816476529738016, + "flos": 22380199157280.0, + "grad_norm": 1.645020473318037, + "learning_rate": 3.5151489568507887e-09, + "loss": 0.7238, + "num_input_tokens_seen": 353804210, + "step": 10885 + }, + { + "epoch": 0.981737836497272, + "flos": 26061811092960.0, + "grad_norm": 2.2806033955599023, + "learning_rate": 3.4806122717545572e-09, + "loss": 0.7503, + "num_input_tokens_seen": 353834270, + "step": 10886 + }, + { + "epoch": 0.9818280200207422, + "flos": 24610214130240.0, + "grad_norm": 3.0397438521135958, + "learning_rate": 3.446245940701198e-09, + "loss": 0.6967, + "num_input_tokens_seen": 353864130, + "step": 10887 + }, + { + "epoch": 0.9819182035442124, + "flos": 21804050479200.0, + "grad_norm": 1.8505661613595363, + "learning_rate": 3.41204996662281e-09, + "loss": 0.6776, + "num_input_tokens_seen": 353892350, + "step": 10888 + }, + { + "epoch": 0.9820083870676828, + "flos": 31131559743840.0, + "grad_norm": 2.049142644285565, + "learning_rate": 3.3780243524375028e-09, + "loss": 0.736, + "num_input_tokens_seen": 353924760, + "step": 10889 + }, + { + "epoch": 0.982098570591153, + "flos": 25442651268960.0, + "grad_norm": 2.8878525093542593, + "learning_rate": 3.3441691010485107e-09, + "loss": 0.7658, + "num_input_tokens_seen": 353953105, + "step": 10890 + }, + { + "epoch": 0.9821887541146233, + "flos": 22054205929440.0, + "grad_norm": 1.78400074897203, + "learning_rate": 3.3104842153444113e-09, + "loss": 0.8151, + "num_input_tokens_seen": 353977630, + "step": 10891 + }, + { + "epoch": 0.9822789376380935, + "flos": 26686100343840.0, + "grad_norm": 2.094870082678055, + "learning_rate": 3.27696969819935e-09, + "loss": 0.7312, + "num_input_tokens_seen": 354005290, + "step": 10892 + }, + { + "epoch": 0.9823691211615638, + "flos": 22241718866880.0, + "grad_norm": 1.6070083698133075, + "learning_rate": 3.2436255524732615e-09, + "loss": 0.6951, + "num_input_tokens_seen": 354033510, + "step": 10893 + }, + { + "epoch": 0.9824593046850341, + "flos": 22387298581440.0, + "grad_norm": 1.9367758105059896, + "learning_rate": 3.210451781010759e-09, + "loss": 0.721, + "num_input_tokens_seen": 354061695, + "step": 10894 + }, + { + "epoch": 0.9825494882085043, + "flos": 30071943800160.0, + "grad_norm": 1.7292571293098082, + "learning_rate": 3.1774483866426895e-09, + "loss": 0.6008, + "num_input_tokens_seen": 354091650, + "step": 10895 + }, + { + "epoch": 0.9826396717319745, + "flos": 41367641282880.0, + "grad_norm": 1.6235010863436492, + "learning_rate": 3.144615372185244e-09, + "loss": 0.7031, + "num_input_tokens_seen": 354123335, + "step": 10896 + }, + { + "epoch": 0.9827298552554449, + "flos": 25521871494720.0, + "grad_norm": 1.4255134226455746, + "learning_rate": 3.1119527404399604e-09, + "loss": 0.8285, + "num_input_tokens_seen": 354156290, + "step": 10897 + }, + { + "epoch": 0.9828200387789151, + "flos": 22933371190080.0, + "grad_norm": 2.2754797922107595, + "learning_rate": 3.0794604941932754e-09, + "loss": 0.7407, + "num_input_tokens_seen": 354182560, + "step": 10898 + }, + { + "epoch": 0.9829102223023853, + "flos": 34628260625280.0, + "grad_norm": 1.8806252338768055, + "learning_rate": 3.0471386362180827e-09, + "loss": 0.6371, + "num_input_tokens_seen": 354215595, + "step": 10899 + }, + { + "epoch": 0.9830004058258556, + "flos": 22638160257120.0, + "grad_norm": 1.9780251741456258, + "learning_rate": 3.0149871692719542e-09, + "loss": 0.786, + "num_input_tokens_seen": 354244945, + "step": 10900 + }, + { + "epoch": 0.9830905893493259, + "flos": 23589632168160.0, + "grad_norm": 1.6993142287516219, + "learning_rate": 2.9830060960984728e-09, + "loss": 0.7228, + "num_input_tokens_seen": 354276515, + "step": 10901 + }, + { + "epoch": 0.9831807728727961, + "flos": 62593981954080.0, + "grad_norm": 0.6665705497316443, + "learning_rate": 2.9511954194263442e-09, + "loss": 0.5834, + "num_input_tokens_seen": 354363285, + "step": 10902 + }, + { + "epoch": 0.9832709563962664, + "flos": 26504200040160.0, + "grad_norm": 2.0978154809283645, + "learning_rate": 2.9195551419698426e-09, + "loss": 0.7346, + "num_input_tokens_seen": 354395155, + "step": 10903 + }, + { + "epoch": 0.9833611399197366, + "flos": 16228118105760.0, + "grad_norm": 2.206787636063972, + "learning_rate": 2.888085266428808e-09, + "loss": 0.6997, + "num_input_tokens_seen": 354421480, + "step": 10904 + }, + { + "epoch": 0.983451323443207, + "flos": 23512865146560.0, + "grad_norm": 1.9951474807836924, + "learning_rate": 2.8567857954882037e-09, + "loss": 0.694, + "num_input_tokens_seen": 354448195, + "step": 10905 + }, + { + "epoch": 0.9835415069666772, + "flos": 25440681271680.0, + "grad_norm": 1.9924882848723084, + "learning_rate": 2.82565673181856e-09, + "loss": 0.7914, + "num_input_tokens_seen": 354475545, + "step": 10906 + }, + { + "epoch": 0.9836316904901474, + "flos": 25520682062400.0, + "grad_norm": 1.8711077026611498, + "learning_rate": 2.7946980780764184e-09, + "loss": 0.7558, + "num_input_tokens_seen": 354506375, + "step": 10907 + }, + { + "epoch": 0.9837218740136177, + "flos": 19145399370240.0, + "grad_norm": 2.7024695864226764, + "learning_rate": 2.763909836903e-09, + "loss": 0.7923, + "num_input_tokens_seen": 354532825, + "step": 10908 + }, + { + "epoch": 0.983812057537088, + "flos": 22605742492800.0, + "grad_norm": 1.609650597539055, + "learning_rate": 2.7332920109255364e-09, + "loss": 0.6527, + "num_input_tokens_seen": 354562070, + "step": 10909 + }, + { + "epoch": 0.9839022410605582, + "flos": 22894969094400.0, + "grad_norm": 6.547884114502419, + "learning_rate": 2.702844602756382e-09, + "loss": 0.7183, + "num_input_tokens_seen": 354592070, + "step": 10910 + }, + { + "epoch": 0.9839924245840285, + "flos": 20347138240800.0, + "grad_norm": 2.9453008433953682, + "learning_rate": 2.6725676149936814e-09, + "loss": 0.7148, + "num_input_tokens_seen": 354620310, + "step": 10911 + }, + { + "epoch": 0.9840826081074988, + "flos": 26133039820320.0, + "grad_norm": 1.9952650048107914, + "learning_rate": 2.642461050220479e-09, + "loss": 0.8033, + "num_input_tokens_seen": 354647640, + "step": 10912 + }, + { + "epoch": 0.984172791630969, + "flos": 33717235146720.0, + "grad_norm": 1.8451743362474562, + "learning_rate": 2.612524911005609e-09, + "loss": 0.7155, + "num_input_tokens_seen": 354679500, + "step": 10913 + }, + { + "epoch": 0.9842629751544393, + "flos": 29965175048640.0, + "grad_norm": 1.669598250234004, + "learning_rate": 2.582759199903917e-09, + "loss": 0.6145, + "num_input_tokens_seen": 354711450, + "step": 10914 + }, + { + "epoch": 0.9843531586779095, + "flos": 32588063114880.0, + "grad_norm": 2.127653820963446, + "learning_rate": 2.553163919454704e-09, + "loss": 0.6977, + "num_input_tokens_seen": 354740280, + "step": 10915 + }, + { + "epoch": 0.9844433422013799, + "flos": 37689486134880.0, + "grad_norm": 1.9816697671778845, + "learning_rate": 2.523739072183506e-09, + "loss": 0.7194, + "num_input_tokens_seen": 354771405, + "step": 10916 + }, + { + "epoch": 0.9845335257248501, + "flos": 30328641128160.0, + "grad_norm": 1.7391847469826094, + "learning_rate": 2.4944846606007597e-09, + "loss": 0.6978, + "num_input_tokens_seen": 354803075, + "step": 10917 + }, + { + "epoch": 0.9846237092483203, + "flos": 53824513960800.0, + "grad_norm": 0.6102139249106758, + "learning_rate": 2.46540068720269e-09, + "loss": 0.614, + "num_input_tokens_seen": 354894070, + "step": 10918 + }, + { + "epoch": 0.9847138927717906, + "flos": 22168408632960.0, + "grad_norm": 1.5738868836857176, + "learning_rate": 2.4364871544708674e-09, + "loss": 0.7727, + "num_input_tokens_seen": 354923035, + "step": 10919 + }, + { + "epoch": 0.9848040762952609, + "flos": 28617075898560.0, + "grad_norm": 1.639860865343254, + "learning_rate": 2.4077440648726523e-09, + "loss": 0.6865, + "num_input_tokens_seen": 354951480, + "step": 10920 + }, + { + "epoch": 0.9848942598187311, + "flos": 22860432653760.0, + "grad_norm": 1.8354012862855382, + "learning_rate": 2.379171420860082e-09, + "loss": 0.7988, + "num_input_tokens_seen": 354981540, + "step": 10921 + }, + { + "epoch": 0.9849844433422014, + "flos": 21476198763360.0, + "grad_norm": 1.9315891548296238, + "learning_rate": 2.3507692248714296e-09, + "loss": 0.6599, + "num_input_tokens_seen": 355010930, + "step": 10922 + }, + { + "epoch": 0.9850746268656716, + "flos": 29452077750720.0, + "grad_norm": 2.109325154033449, + "learning_rate": 2.322537479330089e-09, + "loss": 0.7489, + "num_input_tokens_seen": 355037495, + "step": 10923 + }, + { + "epoch": 0.9851648103891419, + "flos": 25046655915840.0, + "grad_norm": 18.874128511027504, + "learning_rate": 2.2944761866450223e-09, + "loss": 0.8142, + "num_input_tokens_seen": 355066305, + "step": 10924 + }, + { + "epoch": 0.9852549939126122, + "flos": 18124966087200.0, + "grad_norm": 2.0353708575067815, + "learning_rate": 2.266585349210315e-09, + "loss": 0.7173, + "num_input_tokens_seen": 355094370, + "step": 10925 + }, + { + "epoch": 0.9853451774360824, + "flos": 23334310121280.0, + "grad_norm": 2.342108494989296, + "learning_rate": 2.2388649694060623e-09, + "loss": 0.6961, + "num_input_tokens_seen": 355122020, + "step": 10926 + }, + { + "epoch": 0.9854353609595526, + "flos": 24316006780800.0, + "grad_norm": 1.8962487055674129, + "learning_rate": 2.211315049597262e-09, + "loss": 0.7974, + "num_input_tokens_seen": 355151475, + "step": 10927 + }, + { + "epoch": 0.985525544483023, + "flos": 28940764601280.0, + "grad_norm": 2.760500886220156, + "learning_rate": 2.1839355921349224e-09, + "loss": 0.7717, + "num_input_tokens_seen": 355179285, + "step": 10928 + }, + { + "epoch": 0.9856157280064932, + "flos": 23115568851840.0, + "grad_norm": 2.462859647406296, + "learning_rate": 2.156726599354952e-09, + "loss": 0.7335, + "num_input_tokens_seen": 355207185, + "step": 10929 + }, + { + "epoch": 0.9857059115299635, + "flos": 29017680301920.0, + "grad_norm": 1.9316865396780198, + "learning_rate": 2.129688073578828e-09, + "loss": 0.7608, + "num_input_tokens_seen": 355235490, + "step": 10930 + }, + { + "epoch": 0.9857960950534337, + "flos": 22056696303360.0, + "grad_norm": 2.681314224861127, + "learning_rate": 2.1028200171142597e-09, + "loss": 0.8671, + "num_input_tokens_seen": 355264245, + "step": 10931 + }, + { + "epoch": 0.985886278576904, + "flos": 22970026307040.0, + "grad_norm": 2.3733076792126977, + "learning_rate": 2.076122432253191e-09, + "loss": 0.7157, + "num_input_tokens_seen": 355292460, + "step": 10932 + }, + { + "epoch": 0.9859764621003743, + "flos": 29524495910400.0, + "grad_norm": 2.342407961312559, + "learning_rate": 2.0495953212738005e-09, + "loss": 0.7774, + "num_input_tokens_seen": 355320680, + "step": 10933 + }, + { + "epoch": 0.9860666456238445, + "flos": 33751548568800.0, + "grad_norm": 3.4963588848486586, + "learning_rate": 2.0232386864396102e-09, + "loss": 0.817, + "num_input_tokens_seen": 355353655, + "step": 10934 + }, + { + "epoch": 0.9861568291473148, + "flos": 26206981940160.0, + "grad_norm": 2.2700061399738103, + "learning_rate": 1.9970525299992656e-09, + "loss": 0.7613, + "num_input_tokens_seen": 355382745, + "step": 10935 + }, + { + "epoch": 0.9862470126707851, + "flos": 25441015799520.0, + "grad_norm": 2.0321874949317262, + "learning_rate": 1.9710368541874245e-09, + "loss": 0.697, + "num_input_tokens_seen": 355412640, + "step": 10936 + }, + { + "epoch": 0.9863371961942553, + "flos": 71215820162400.0, + "grad_norm": 0.5674758020779256, + "learning_rate": 1.945191661223644e-09, + "loss": 0.585, + "num_input_tokens_seen": 355515510, + "step": 10937 + }, + { + "epoch": 0.9864273797177255, + "flos": 27852150511200.0, + "grad_norm": 1.9430583471508875, + "learning_rate": 1.9195169533132714e-09, + "loss": 0.7834, + "num_input_tokens_seen": 355546160, + "step": 10938 + }, + { + "epoch": 0.9865175632411959, + "flos": 62937928176000.0, + "grad_norm": 0.5850714322844662, + "learning_rate": 1.894012732646999e-09, + "loss": 0.5923, + "num_input_tokens_seen": 355642920, + "step": 10939 + }, + { + "epoch": 0.9866077467646661, + "flos": 25666410456000.0, + "grad_norm": 3.2216064222780965, + "learning_rate": 1.8686790014010854e-09, + "loss": 0.684, + "num_input_tokens_seen": 355671980, + "step": 10940 + }, + { + "epoch": 0.9866979302881363, + "flos": 55971071355360.0, + "grad_norm": 0.6020519692810663, + "learning_rate": 1.8435157617369134e-09, + "loss": 0.5581, + "num_input_tokens_seen": 355759495, + "step": 10941 + }, + { + "epoch": 0.9867881138116066, + "flos": 22168668821280.0, + "grad_norm": 1.90320851059767, + "learning_rate": 1.818523015801876e-09, + "loss": 0.8791, + "num_input_tokens_seen": 355787025, + "step": 10942 + }, + { + "epoch": 0.9868782973350769, + "flos": 15938519806560.0, + "grad_norm": 3.0889649039534985, + "learning_rate": 1.7937007657282677e-09, + "loss": 0.698, + "num_input_tokens_seen": 355812020, + "step": 10943 + }, + { + "epoch": 0.9869684808585472, + "flos": 23587773680160.0, + "grad_norm": 1.776760600912777, + "learning_rate": 1.7690490136341718e-09, + "loss": 0.6781, + "num_input_tokens_seen": 355840520, + "step": 10944 + }, + { + "epoch": 0.9870586643820174, + "flos": 34883062295520.0, + "grad_norm": 2.233371962977105, + "learning_rate": 1.744567761622795e-09, + "loss": 0.7322, + "num_input_tokens_seen": 355874030, + "step": 10945 + }, + { + "epoch": 0.9871488479054876, + "flos": 29560704990240.0, + "grad_norm": 1.9924435320119427, + "learning_rate": 1.7202570117831327e-09, + "loss": 0.7308, + "num_input_tokens_seen": 355903775, + "step": 10946 + }, + { + "epoch": 0.987239031428958, + "flos": 21220170491040.0, + "grad_norm": 2.4567248906965786, + "learning_rate": 1.696116766189526e-09, + "loss": 0.7141, + "num_input_tokens_seen": 355930505, + "step": 10947 + }, + { + "epoch": 0.9873292149524282, + "flos": 26831048172480.0, + "grad_norm": 1.8784980568448033, + "learning_rate": 1.6721470269021042e-09, + "loss": 0.7749, + "num_input_tokens_seen": 355961490, + "step": 10948 + }, + { + "epoch": 0.9874193984758984, + "flos": 21184556127360.0, + "grad_norm": 2.2347118417222047, + "learning_rate": 1.6483477959654546e-09, + "loss": 0.7074, + "num_input_tokens_seen": 355989275, + "step": 10949 + }, + { + "epoch": 0.9875095819993687, + "flos": 40600745898240.0, + "grad_norm": 2.0461466417908354, + "learning_rate": 1.6247190754106187e-09, + "loss": 0.72, + "num_input_tokens_seen": 356023935, + "step": 10950 + }, + { + "epoch": 0.987599765522839, + "flos": 29960603168160.0, + "grad_norm": 1.566971522288522, + "learning_rate": 1.6012608672537619e-09, + "loss": 0.7002, + "num_input_tokens_seen": 356054085, + "step": 10951 + }, + { + "epoch": 0.9876899490463092, + "flos": 24573298824960.0, + "grad_norm": 2.9410902768015106, + "learning_rate": 1.5779731734963942e-09, + "loss": 0.7446, + "num_input_tokens_seen": 356080345, + "step": 10952 + }, + { + "epoch": 0.9877801325697795, + "flos": 26357839760640.0, + "grad_norm": 2.9559285414528027, + "learning_rate": 1.5548559961253705e-09, + "loss": 0.7968, + "num_input_tokens_seen": 356108350, + "step": 10953 + }, + { + "epoch": 0.9878703160932497, + "flos": 23077575623520.0, + "grad_norm": 1.9582780532609936, + "learning_rate": 1.5319093371135573e-09, + "loss": 0.6613, + "num_input_tokens_seen": 356137460, + "step": 10954 + }, + { + "epoch": 0.9879604996167201, + "flos": 28867826064960.0, + "grad_norm": 1.7597515346966421, + "learning_rate": 1.5091331984184997e-09, + "loss": 0.8264, + "num_input_tokens_seen": 356168770, + "step": 10955 + }, + { + "epoch": 0.9880506831401903, + "flos": 27153621782400.0, + "grad_norm": 1.7860231527513009, + "learning_rate": 1.486527581983754e-09, + "loss": 0.7356, + "num_input_tokens_seen": 356197475, + "step": 10956 + }, + { + "epoch": 0.9881408666636605, + "flos": 22022828918400.0, + "grad_norm": 1.9008350068380016, + "learning_rate": 1.4640924897382223e-09, + "loss": 0.7969, + "num_input_tokens_seen": 356225080, + "step": 10957 + }, + { + "epoch": 0.9882310501871309, + "flos": 23696289410400.0, + "grad_norm": 2.6305234109735522, + "learning_rate": 1.4418279235961506e-09, + "loss": 0.7773, + "num_input_tokens_seen": 356254910, + "step": 10958 + }, + { + "epoch": 0.9883212337106011, + "flos": 27086741916960.0, + "grad_norm": 1.9209762580750795, + "learning_rate": 1.4197338854573526e-09, + "loss": 0.6413, + "num_input_tokens_seen": 356284665, + "step": 10959 + }, + { + "epoch": 0.9884114172340713, + "flos": 17935669001280.0, + "grad_norm": 1.9416123708683735, + "learning_rate": 1.3978103772067651e-09, + "loss": 0.6931, + "num_input_tokens_seen": 356311090, + "step": 10960 + }, + { + "epoch": 0.9885016007575416, + "flos": 33536115408000.0, + "grad_norm": 1.7069818066173286, + "learning_rate": 1.3760574007153358e-09, + "loss": 0.6584, + "num_input_tokens_seen": 356341060, + "step": 10961 + }, + { + "epoch": 0.9885917842810119, + "flos": 20965926367200.0, + "grad_norm": 2.1666648449034245, + "learning_rate": 1.3544749578389137e-09, + "loss": 0.8187, + "num_input_tokens_seen": 356367475, + "step": 10962 + }, + { + "epoch": 0.9886819678044821, + "flos": 19797869032800.0, + "grad_norm": 1.9494618357904552, + "learning_rate": 1.3330630504189143e-09, + "loss": 0.7852, + "num_input_tokens_seen": 356393540, + "step": 10963 + }, + { + "epoch": 0.9887721513279524, + "flos": 35793939095040.0, + "grad_norm": 1.6671824690252188, + "learning_rate": 1.3118216802827652e-09, + "loss": 0.7049, + "num_input_tokens_seen": 356423635, + "step": 10964 + }, + { + "epoch": 0.9888623348514226, + "flos": 30876014678400.0, + "grad_norm": 2.2262177853017873, + "learning_rate": 1.2907508492425722e-09, + "loss": 0.7021, + "num_input_tokens_seen": 356452530, + "step": 10965 + }, + { + "epoch": 0.988952518374893, + "flos": 42461087441760.0, + "grad_norm": 1.9000970313743704, + "learning_rate": 1.2698505590962305e-09, + "loss": 0.6983, + "num_input_tokens_seen": 356483540, + "step": 10966 + }, + { + "epoch": 0.9890427018983632, + "flos": 25411497276480.0, + "grad_norm": 1.8019901611703073, + "learning_rate": 1.2491208116272022e-09, + "loss": 0.7289, + "num_input_tokens_seen": 356513790, + "step": 10967 + }, + { + "epoch": 0.9891328854218334, + "flos": 18342740942880.0, + "grad_norm": 1.9544767051122447, + "learning_rate": 1.2285616086040728e-09, + "loss": 0.7691, + "num_input_tokens_seen": 356539760, + "step": 10968 + }, + { + "epoch": 0.9892230689453037, + "flos": 24973791719040.0, + "grad_norm": 1.4698658215015354, + "learning_rate": 1.2081729517812167e-09, + "loss": 0.8025, + "num_input_tokens_seen": 356572735, + "step": 10969 + }, + { + "epoch": 0.989313252468774, + "flos": 28689494058240.0, + "grad_norm": 1.6827833876221212, + "learning_rate": 1.1879548428983533e-09, + "loss": 0.7309, + "num_input_tokens_seen": 356603645, + "step": 10970 + }, + { + "epoch": 0.9894034359922442, + "flos": 26030768609760.0, + "grad_norm": 2.266185717420588, + "learning_rate": 1.167907283680547e-09, + "loss": 0.6919, + "num_input_tokens_seen": 356633685, + "step": 10971 + }, + { + "epoch": 0.9894936195157145, + "flos": 26905027462080.0, + "grad_norm": 1.9992700675290171, + "learning_rate": 1.1480302758382077e-09, + "loss": 0.6781, + "num_input_tokens_seen": 356661165, + "step": 10972 + }, + { + "epoch": 0.9895838030391847, + "flos": 24135890625600.0, + "grad_norm": 3.55701854694326, + "learning_rate": 1.1283238210675338e-09, + "loss": 0.728, + "num_input_tokens_seen": 356692630, + "step": 10973 + }, + { + "epoch": 0.989673986562655, + "flos": 19981627824480.0, + "grad_norm": 2.511314632482177, + "learning_rate": 1.1087879210498474e-09, + "loss": 0.666, + "num_input_tokens_seen": 356720285, + "step": 10974 + }, + { + "epoch": 0.9897641700861253, + "flos": 14991991473600.0, + "grad_norm": 1.9564832578781612, + "learning_rate": 1.0894225774522592e-09, + "loss": 0.7834, + "num_input_tokens_seen": 356746730, + "step": 10975 + }, + { + "epoch": 0.9898543536095955, + "flos": 67494691038240.0, + "grad_norm": 0.6478090343023801, + "learning_rate": 1.070227791927003e-09, + "loss": 0.5941, + "num_input_tokens_seen": 356832875, + "step": 10976 + }, + { + "epoch": 0.9899445371330657, + "flos": 21476496121440.0, + "grad_norm": 2.9556507430686803, + "learning_rate": 1.0512035661118802e-09, + "loss": 0.8359, + "num_input_tokens_seen": 356859020, + "step": 10977 + }, + { + "epoch": 0.9900347206565361, + "flos": 22276887193440.0, + "grad_norm": 2.213432602532955, + "learning_rate": 1.0323499016300364e-09, + "loss": 0.7579, + "num_input_tokens_seen": 356886195, + "step": 10978 + }, + { + "epoch": 0.9901249041800063, + "flos": 22386852544320.0, + "grad_norm": 3.597059235548251, + "learning_rate": 1.013666800090407e-09, + "loss": 0.6909, + "num_input_tokens_seen": 356912935, + "step": 10979 + }, + { + "epoch": 0.9902150877034765, + "flos": 28034422512480.0, + "grad_norm": 1.7919541357506306, + "learning_rate": 9.951542630870502e-10, + "loss": 0.7198, + "num_input_tokens_seen": 356942975, + "step": 10980 + }, + { + "epoch": 0.9903052712269468, + "flos": 23841274408800.0, + "grad_norm": 2.0166969701628727, + "learning_rate": 9.768122921995915e-10, + "loss": 0.7754, + "num_input_tokens_seen": 356970560, + "step": 10981 + }, + { + "epoch": 0.9903954547504171, + "flos": 22641728554080.0, + "grad_norm": 2.5362763597407394, + "learning_rate": 9.58640888992779e-10, + "loss": 0.6756, + "num_input_tokens_seen": 356996120, + "step": 10982 + }, + { + "epoch": 0.9904856382738874, + "flos": 26977928828640.0, + "grad_norm": 2.0286638669610904, + "learning_rate": 9.40640055017594e-10, + "loss": 0.654, + "num_input_tokens_seen": 357026490, + "step": 10983 + }, + { + "epoch": 0.9905758217973576, + "flos": 24755050449600.0, + "grad_norm": 2.0367812919305908, + "learning_rate": 9.228097918094757e-10, + "loss": 0.7366, + "num_input_tokens_seen": 357056925, + "step": 10984 + }, + { + "epoch": 0.9906660053208279, + "flos": 28213312065600.0, + "grad_norm": 2.7593157510239523, + "learning_rate": 9.051501008900952e-10, + "loss": 0.7076, + "num_input_tokens_seen": 357089780, + "step": 10985 + }, + { + "epoch": 0.9907561888442982, + "flos": 25119371433600.0, + "grad_norm": 7.588243033965721, + "learning_rate": 8.876609837662475e-10, + "loss": 0.7545, + "num_input_tokens_seen": 357119240, + "step": 10986 + }, + { + "epoch": 0.9908463723677684, + "flos": 25700946896640.0, + "grad_norm": 1.6916599567950492, + "learning_rate": 8.70342441930294e-10, + "loss": 0.8233, + "num_input_tokens_seen": 357151345, + "step": 10987 + }, + { + "epoch": 0.9909365558912386, + "flos": 24822785219520.0, + "grad_norm": 2.6266002318731663, + "learning_rate": 8.531944768594979e-10, + "loss": 0.7087, + "num_input_tokens_seen": 357181070, + "step": 10988 + }, + { + "epoch": 0.991026739414709, + "flos": 24608429981760.0, + "grad_norm": 1.6372114728121605, + "learning_rate": 8.362170900175769e-10, + "loss": 0.818, + "num_input_tokens_seen": 357210720, + "step": 10989 + }, + { + "epoch": 0.9911169229381792, + "flos": 28981471222080.0, + "grad_norm": 1.8052610336450428, + "learning_rate": 8.194102828527061e-10, + "loss": 0.69, + "num_input_tokens_seen": 357239270, + "step": 10990 + }, + { + "epoch": 0.9912071064616494, + "flos": 17979089014560.0, + "grad_norm": 5.272574980099986, + "learning_rate": 8.027740567992936e-10, + "loss": 0.7812, + "num_input_tokens_seen": 357265965, + "step": 10991 + }, + { + "epoch": 0.9912972899851197, + "flos": 28325730620640.0, + "grad_norm": 1.7679030989031062, + "learning_rate": 7.863084132766484e-10, + "loss": 0.7324, + "num_input_tokens_seen": 357296820, + "step": 10992 + }, + { + "epoch": 0.99138747350859, + "flos": 25702359347520.0, + "grad_norm": 2.1717565490476365, + "learning_rate": 7.700133536896469e-10, + "loss": 0.7535, + "num_input_tokens_seen": 357325350, + "step": 10993 + }, + { + "epoch": 0.9914776570320603, + "flos": 24536495028960.0, + "grad_norm": 1.8357804900864325, + "learning_rate": 7.538888794287324e-10, + "loss": 0.7049, + "num_input_tokens_seen": 357355045, + "step": 10994 + }, + { + "epoch": 0.9915678405555305, + "flos": 25886341157760.0, + "grad_norm": 1.750901603051618, + "learning_rate": 7.379349918696931e-10, + "loss": 0.7548, + "num_input_tokens_seen": 357387635, + "step": 10995 + }, + { + "epoch": 0.9916580240790007, + "flos": 21439841004480.0, + "grad_norm": 2.390591390535548, + "learning_rate": 7.221516923738846e-10, + "loss": 0.7097, + "num_input_tokens_seen": 357414535, + "step": 10996 + }, + { + "epoch": 0.9917482076024711, + "flos": 20566065359040.0, + "grad_norm": 2.644564877723096, + "learning_rate": 7.065389822880075e-10, + "loss": 0.84, + "num_input_tokens_seen": 357441175, + "step": 10997 + }, + { + "epoch": 0.9918383911259413, + "flos": 22387484430240.0, + "grad_norm": 2.051613833038706, + "learning_rate": 6.910968629443292e-10, + "loss": 0.8132, + "num_input_tokens_seen": 357469735, + "step": 10998 + }, + { + "epoch": 0.9919285746494115, + "flos": 15282742035360.0, + "grad_norm": 2.0355048585785482, + "learning_rate": 6.758253356602406e-10, + "loss": 0.7343, + "num_input_tokens_seen": 357496855, + "step": 10999 + }, + { + "epoch": 0.9920187581728818, + "flos": 25520087346240.0, + "grad_norm": 1.7993602648996616, + "learning_rate": 6.607244017389213e-10, + "loss": 0.8333, + "num_input_tokens_seen": 357524210, + "step": 11000 + }, + { + "epoch": 0.9921089416963521, + "flos": 24390952484160.0, + "grad_norm": 2.299201254432989, + "learning_rate": 6.457940624686742e-10, + "loss": 0.7563, + "num_input_tokens_seen": 357553780, + "step": 11001 + }, + { + "epoch": 0.9921991252198223, + "flos": 25739274652800.0, + "grad_norm": 2.029058947305229, + "learning_rate": 6.310343191238132e-10, + "loss": 0.8232, + "num_input_tokens_seen": 357584435, + "step": 11002 + }, + { + "epoch": 0.9922893087432926, + "flos": 18853682394720.0, + "grad_norm": 1.889004856489087, + "learning_rate": 6.164451729635534e-10, + "loss": 0.6323, + "num_input_tokens_seen": 357612595, + "step": 11003 + }, + { + "epoch": 0.9923794922667628, + "flos": 17943474650880.0, + "grad_norm": 2.027042667912805, + "learning_rate": 6.020266252324546e-10, + "loss": 0.6073, + "num_input_tokens_seen": 357640345, + "step": 11004 + }, + { + "epoch": 0.9924696757902332, + "flos": 21986657008320.0, + "grad_norm": 2.026558608240348, + "learning_rate": 5.877786771610882e-10, + "loss": 0.7097, + "num_input_tokens_seen": 357669675, + "step": 11005 + }, + { + "epoch": 0.9925598593137034, + "flos": 67243234646400.0, + "grad_norm": 0.7083605078552148, + "learning_rate": 5.737013299651483e-10, + "loss": 0.6456, + "num_input_tokens_seen": 357765540, + "step": 11006 + }, + { + "epoch": 0.9926500428371736, + "flos": 20345279752800.0, + "grad_norm": 2.079487289547627, + "learning_rate": 5.597945848458963e-10, + "loss": 0.7404, + "num_input_tokens_seen": 357789685, + "step": 11007 + }, + { + "epoch": 0.992740226360644, + "flos": 19435964083200.0, + "grad_norm": 2.1542842584484365, + "learning_rate": 5.460584429894944e-10, + "loss": 0.7107, + "num_input_tokens_seen": 357817100, + "step": 11008 + }, + { + "epoch": 0.9928304098841142, + "flos": 20814771188640.0, + "grad_norm": 2.161111851386712, + "learning_rate": 5.32492905568338e-10, + "loss": 0.7383, + "num_input_tokens_seen": 357845610, + "step": 11009 + }, + { + "epoch": 0.9929205934075844, + "flos": 23480484552000.0, + "grad_norm": 1.7471437305528017, + "learning_rate": 5.190979737399459e-10, + "loss": 0.7682, + "num_input_tokens_seen": 357874510, + "step": 11010 + }, + { + "epoch": 0.9930107769310547, + "flos": 34625844590880.0, + "grad_norm": 1.9754522771494745, + "learning_rate": 5.058736486469594e-10, + "loss": 0.6275, + "num_input_tokens_seen": 357907545, + "step": 11011 + }, + { + "epoch": 0.993100960454525, + "flos": 20092262231040.0, + "grad_norm": 3.092048099839521, + "learning_rate": 4.928199314180314e-10, + "loss": 0.7696, + "num_input_tokens_seen": 357933965, + "step": 11012 + }, + { + "epoch": 0.9931911439779952, + "flos": 23734617166560.0, + "grad_norm": 1.9263622962652087, + "learning_rate": 4.799368231669376e-10, + "loss": 0.655, + "num_input_tokens_seen": 357964775, + "step": 11013 + }, + { + "epoch": 0.9932813275014655, + "flos": 23262040640640.0, + "grad_norm": 1.869529627836347, + "learning_rate": 4.672243249927988e-10, + "loss": 0.7233, + "num_input_tokens_seen": 357993000, + "step": 11014 + }, + { + "epoch": 0.9933715110249357, + "flos": 28945150632960.0, + "grad_norm": 1.757529002746642, + "learning_rate": 4.546824379803027e-10, + "loss": 0.7676, + "num_input_tokens_seen": 358023150, + "step": 11015 + }, + { + "epoch": 0.993461694548406, + "flos": 19727346530880.0, + "grad_norm": 2.2772491068991654, + "learning_rate": 4.4231116319970454e-10, + "loss": 0.7664, + "num_input_tokens_seen": 358049295, + "step": 11016 + }, + { + "epoch": 0.9935518780718763, + "flos": 24974237756160.0, + "grad_norm": 1.739443683919893, + "learning_rate": 4.3011050170660423e-10, + "loss": 0.7411, + "num_input_tokens_seen": 358078045, + "step": 11017 + }, + { + "epoch": 0.9936420615953465, + "flos": 25848979815360.0, + "grad_norm": 1.8014825543467587, + "learning_rate": 4.18080454542169e-10, + "loss": 0.7288, + "num_input_tokens_seen": 358107950, + "step": 11018 + }, + { + "epoch": 0.9937322451188167, + "flos": 63563592708000.0, + "grad_norm": 0.6567407771066653, + "learning_rate": 4.0622102273246694e-10, + "loss": 0.5751, + "num_input_tokens_seen": 358204455, + "step": 11019 + }, + { + "epoch": 0.9938224286422871, + "flos": 21838289561760.0, + "grad_norm": 1.6108977881789897, + "learning_rate": 3.945322072897994e-10, + "loss": 0.7967, + "num_input_tokens_seen": 358233065, + "step": 11020 + }, + { + "epoch": 0.9939126121657573, + "flos": 18853087678560.0, + "grad_norm": 2.808869116260185, + "learning_rate": 3.830140092111467e-10, + "loss": 0.7962, + "num_input_tokens_seen": 358260175, + "step": 11021 + }, + { + "epoch": 0.9940027956892276, + "flos": 19727718228480.0, + "grad_norm": 2.160777769991413, + "learning_rate": 3.7166642947972225e-10, + "loss": 0.7338, + "num_input_tokens_seen": 358286475, + "step": 11022 + }, + { + "epoch": 0.9940929792126978, + "flos": 24897322055520.0, + "grad_norm": 2.1832847308313563, + "learning_rate": 3.604894690634186e-10, + "loss": 0.7128, + "num_input_tokens_seen": 358314845, + "step": 11023 + }, + { + "epoch": 0.9941831627361681, + "flos": 24968253424800.0, + "grad_norm": 4.344943534387599, + "learning_rate": 3.494831289161393e-10, + "loss": 0.8224, + "num_input_tokens_seen": 358342060, + "step": 11024 + }, + { + "epoch": 0.9942733462596384, + "flos": 21258498247200.0, + "grad_norm": 1.9801051541543921, + "learning_rate": 3.3864740997668897e-10, + "loss": 0.7716, + "num_input_tokens_seen": 358370775, + "step": 11025 + }, + { + "epoch": 0.9943635297831086, + "flos": 23553534597600.0, + "grad_norm": 2.3185407765224793, + "learning_rate": 3.279823131701054e-10, + "loss": 0.7636, + "num_input_tokens_seen": 358403200, + "step": 11026 + }, + { + "epoch": 0.9944537133065788, + "flos": 27667536815040.0, + "grad_norm": 1.5497095358855555, + "learning_rate": 3.1748783940610536e-10, + "loss": 0.6932, + "num_input_tokens_seen": 358433815, + "step": 11027 + }, + { + "epoch": 0.9945438968300492, + "flos": 20674878447360.0, + "grad_norm": 1.7172188614540502, + "learning_rate": 3.071639895801947e-10, + "loss": 0.6981, + "num_input_tokens_seen": 358461645, + "step": 11028 + }, + { + "epoch": 0.9946340803535194, + "flos": 21148272708000.0, + "grad_norm": 2.2805330377856587, + "learning_rate": 2.9701076457322447e-10, + "loss": 0.8013, + "num_input_tokens_seen": 358487995, + "step": 11029 + }, + { + "epoch": 0.9947242638769896, + "flos": 18889333928160.0, + "grad_norm": 2.219947813984366, + "learning_rate": 2.870281652513906e-10, + "loss": 0.7794, + "num_input_tokens_seen": 358514315, + "step": 11030 + }, + { + "epoch": 0.99481444740046, + "flos": 24683003987520.0, + "grad_norm": 1.9019608317853847, + "learning_rate": 2.772161924669003e-10, + "loss": 0.7268, + "num_input_tokens_seen": 358543570, + "step": 11031 + }, + { + "epoch": 0.9949046309239302, + "flos": 27414816651360.0, + "grad_norm": 2.369506606399268, + "learning_rate": 2.6757484705641765e-10, + "loss": 0.7241, + "num_input_tokens_seen": 358574550, + "step": 11032 + }, + { + "epoch": 0.9949948144474005, + "flos": 22969877628000.0, + "grad_norm": 2.248758291300257, + "learning_rate": 2.58104129843062e-10, + "loss": 0.7363, + "num_input_tokens_seen": 358601585, + "step": 11033 + }, + { + "epoch": 0.9950849979708707, + "flos": 11923592200320.0, + "grad_norm": 2.3663656992569067, + "learning_rate": 2.4880404163463154e-10, + "loss": 0.6944, + "num_input_tokens_seen": 358626790, + "step": 11034 + }, + { + "epoch": 0.995175181494341, + "flos": 24169423482720.0, + "grad_norm": 2.693194632578376, + "learning_rate": 2.3967458322471377e-10, + "loss": 0.7266, + "num_input_tokens_seen": 358654255, + "step": 11035 + }, + { + "epoch": 0.9952653650178113, + "flos": 24132619686720.0, + "grad_norm": 1.723314985414938, + "learning_rate": 2.307157553922412e-10, + "loss": 0.75, + "num_input_tokens_seen": 358683510, + "step": 11036 + }, + { + "epoch": 0.9953555485412815, + "flos": 18671521902720.0, + "grad_norm": 1.71637902149243, + "learning_rate": 2.2192755890193538e-10, + "loss": 0.7834, + "num_input_tokens_seen": 358710705, + "step": 11037 + }, + { + "epoch": 0.9954457320647517, + "flos": 53517477310560.0, + "grad_norm": 0.6269417402237961, + "learning_rate": 2.133099945034189e-10, + "loss": 0.5444, + "num_input_tokens_seen": 358810460, + "step": 11038 + }, + { + "epoch": 0.9955359155882221, + "flos": 24675755884320.0, + "grad_norm": 2.1406120801104174, + "learning_rate": 2.048630629318815e-10, + "loss": 0.7603, + "num_input_tokens_seen": 358835520, + "step": 11039 + }, + { + "epoch": 0.9956260991116923, + "flos": 23880122541600.0, + "grad_norm": 1.672449815279459, + "learning_rate": 1.965867649080799e-10, + "loss": 0.7342, + "num_input_tokens_seen": 358866595, + "step": 11040 + }, + { + "epoch": 0.9957162826351625, + "flos": 65891901727200.0, + "grad_norm": 0.6070418318878011, + "learning_rate": 1.8848110113856008e-10, + "loss": 0.5626, + "num_input_tokens_seen": 358966380, + "step": 11041 + }, + { + "epoch": 0.9958064661586328, + "flos": 19144878993600.0, + "grad_norm": 2.3189193704857365, + "learning_rate": 1.8054607231454687e-10, + "loss": 0.7772, + "num_input_tokens_seen": 358995200, + "step": 11042 + }, + { + "epoch": 0.9958966496821031, + "flos": 20383719018240.0, + "grad_norm": 1.7637879934348437, + "learning_rate": 1.7278167911327635e-10, + "loss": 0.6914, + "num_input_tokens_seen": 359022925, + "step": 11043 + }, + { + "epoch": 0.9959868332055734, + "flos": 19763555610720.0, + "grad_norm": 1.9102982232315242, + "learning_rate": 1.6518792219710753e-10, + "loss": 0.7712, + "num_input_tokens_seen": 359050015, + "step": 11044 + }, + { + "epoch": 0.9960770167290436, + "flos": 21693416072640.0, + "grad_norm": 1.9606675041925032, + "learning_rate": 1.5776480221418865e-10, + "loss": 0.7423, + "num_input_tokens_seen": 359079395, + "step": 11045 + }, + { + "epoch": 0.9961672002525138, + "flos": 14735368485120.0, + "grad_norm": 3.291971474228443, + "learning_rate": 1.505123197977909e-10, + "loss": 0.7385, + "num_input_tokens_seen": 359102750, + "step": 11046 + }, + { + "epoch": 0.9962573837759842, + "flos": 41403813192960.0, + "grad_norm": 2.0158886404389094, + "learning_rate": 1.4343047556675258e-10, + "loss": 0.6725, + "num_input_tokens_seen": 359137715, + "step": 11047 + }, + { + "epoch": 0.9963475672994544, + "flos": 16083541974720.0, + "grad_norm": 2.332407960864596, + "learning_rate": 1.3651927012503506e-10, + "loss": 0.7417, + "num_input_tokens_seen": 359163620, + "step": 11048 + }, + { + "epoch": 0.9964377508229246, + "flos": 23183043433440.0, + "grad_norm": 2.029794494299094, + "learning_rate": 1.297787040630549e-10, + "loss": 0.7437, + "num_input_tokens_seen": 359191785, + "step": 11049 + }, + { + "epoch": 0.9965279343463949, + "flos": 19436187101760.0, + "grad_norm": 6.203679672540498, + "learning_rate": 1.2320877795524153e-10, + "loss": 0.7613, + "num_input_tokens_seen": 359216480, + "step": 11050 + }, + { + "epoch": 0.9966181178698652, + "flos": 52479802992480.0, + "grad_norm": 1.9803859234406103, + "learning_rate": 1.1680949236247962e-10, + "loss": 0.7198, + "num_input_tokens_seen": 359249600, + "step": 11051 + }, + { + "epoch": 0.9967083013933354, + "flos": 21731409300960.0, + "grad_norm": 1.8136060562695655, + "learning_rate": 1.1058084783099886e-10, + "loss": 0.7431, + "num_input_tokens_seen": 359278065, + "step": 11052 + }, + { + "epoch": 0.9967984849168057, + "flos": 25411980483360.0, + "grad_norm": 2.2352092759296944, + "learning_rate": 1.0452284489170793e-10, + "loss": 0.7498, + "num_input_tokens_seen": 359306610, + "step": 11053 + }, + { + "epoch": 0.996888668440276, + "flos": 21841300312320.0, + "grad_norm": 2.2542552975231382, + "learning_rate": 9.86354840621928e-11, + "loss": 0.7219, + "num_input_tokens_seen": 359337930, + "step": 11054 + }, + { + "epoch": 0.9969788519637462, + "flos": 29999265452160.0, + "grad_norm": 1.9884679440606743, + "learning_rate": 9.291876584427427e-11, + "loss": 0.6123, + "num_input_tokens_seen": 359368320, + "step": 11055 + }, + { + "epoch": 0.9970690354872165, + "flos": 21759701221920.0, + "grad_norm": 2.0935160947499427, + "learning_rate": 8.737269072578435e-11, + "loss": 0.7088, + "num_input_tokens_seen": 359392155, + "step": 11056 + }, + { + "epoch": 0.9971592190106867, + "flos": 23844285159360.0, + "grad_norm": 1.8144593007749448, + "learning_rate": 8.199725918012212e-11, + "loss": 0.7333, + "num_input_tokens_seen": 359420390, + "step": 11057 + }, + { + "epoch": 0.9972494025341571, + "flos": 19946868365280.0, + "grad_norm": 3.31867951571515, + "learning_rate": 7.679247166603175e-11, + "loss": 0.7463, + "num_input_tokens_seen": 359448290, + "step": 11058 + }, + { + "epoch": 0.9973395860576273, + "flos": 21840965784480.0, + "grad_norm": 2.0517823309978978, + "learning_rate": 7.17583286273804e-11, + "loss": 0.746, + "num_input_tokens_seen": 359476190, + "step": 11059 + }, + { + "epoch": 0.9974297695810975, + "flos": 25589383246080.0, + "grad_norm": 2.251914481421202, + "learning_rate": 6.689483049360233e-11, + "loss": 0.758, + "num_input_tokens_seen": 359504125, + "step": 11060 + }, + { + "epoch": 0.9975199531045678, + "flos": 16703928400800.0, + "grad_norm": 2.1777186098084065, + "learning_rate": 6.220197768014302e-11, + "loss": 0.7168, + "num_input_tokens_seen": 359531210, + "step": 11061 + }, + { + "epoch": 0.9976101366280381, + "flos": 24752076868800.0, + "grad_norm": 2.271595946414835, + "learning_rate": 5.7679770587126806e-11, + "loss": 0.7587, + "num_input_tokens_seen": 359559610, + "step": 11062 + }, + { + "epoch": 0.9977003201515083, + "flos": 68318653471680.0, + "grad_norm": 0.6123780567071877, + "learning_rate": 5.33282096002452e-11, + "loss": 0.6644, + "num_input_tokens_seen": 359648480, + "step": 11063 + }, + { + "epoch": 0.9977905036749786, + "flos": 22236403591200.0, + "grad_norm": 5.118487802946883, + "learning_rate": 4.914729509120086e-11, + "loss": 0.7083, + "num_input_tokens_seen": 359673525, + "step": 11064 + }, + { + "epoch": 0.9978806871984488, + "flos": 38378796763200.0, + "grad_norm": 1.6440046831096764, + "learning_rate": 4.513702741637537e-11, + "loss": 0.6615, + "num_input_tokens_seen": 359705890, + "step": 11065 + }, + { + "epoch": 0.9979708707219191, + "flos": 23844210819840.0, + "grad_norm": 2.6832053890488066, + "learning_rate": 4.129740691816153e-11, + "loss": 0.7991, + "num_input_tokens_seen": 359734380, + "step": 11066 + }, + { + "epoch": 0.9980610542453894, + "flos": 25521016590240.0, + "grad_norm": 1.7522433071457668, + "learning_rate": 3.762843392429715e-11, + "loss": 0.7572, + "num_input_tokens_seen": 359764085, + "step": 11067 + }, + { + "epoch": 0.9981512377688596, + "flos": 30875828829600.0, + "grad_norm": 2.6723332597693594, + "learning_rate": 3.413010874742106e-11, + "loss": 0.7175, + "num_input_tokens_seen": 359794820, + "step": 11068 + }, + { + "epoch": 0.9982414212923298, + "flos": 20165052088320.0, + "grad_norm": 2.406554462938073, + "learning_rate": 3.080243168618324e-11, + "loss": 0.7346, + "num_input_tokens_seen": 359823615, + "step": 11069 + }, + { + "epoch": 0.9983316048158002, + "flos": 24166635750720.0, + "grad_norm": 1.8152421399897132, + "learning_rate": 2.7645403024800783e-11, + "loss": 0.6857, + "num_input_tokens_seen": 359852190, + "step": 11070 + }, + { + "epoch": 0.9984217883392704, + "flos": 25771580907840.0, + "grad_norm": 1.4948916075079262, + "learning_rate": 2.4659023032391756e-11, + "loss": 0.8096, + "num_input_tokens_seen": 359881845, + "step": 11071 + }, + { + "epoch": 0.9985119718627407, + "flos": 31749715984320.0, + "grad_norm": 1.6260407122996736, + "learning_rate": 2.1843291963863364e-11, + "loss": 0.7854, + "num_input_tokens_seen": 359916640, + "step": 11072 + }, + { + "epoch": 0.9986021553862109, + "flos": 27997953244320.0, + "grad_norm": 2.355921055020005, + "learning_rate": 1.9198210059245822e-11, + "loss": 0.7577, + "num_input_tokens_seen": 359947255, + "step": 11073 + }, + { + "epoch": 0.9986923389096812, + "flos": 18889594116480.0, + "grad_norm": 2.141431559065539, + "learning_rate": 1.672377754458054e-11, + "loss": 0.7446, + "num_input_tokens_seen": 359973920, + "step": 11074 + }, + { + "epoch": 0.9987825224331515, + "flos": 27556270522560.0, + "grad_norm": 1.741781007703433, + "learning_rate": 1.4419994630809895e-11, + "loss": 0.7335, + "num_input_tokens_seen": 360002265, + "step": 11075 + }, + { + "epoch": 0.9988727059566217, + "flos": 28216397155680.0, + "grad_norm": 2.405348571768546, + "learning_rate": 1.2286861514443358e-11, + "loss": 0.6911, + "num_input_tokens_seen": 360033125, + "step": 11076 + }, + { + "epoch": 0.998962889480092, + "flos": 24717131560800.0, + "grad_norm": 1.5553175066652167, + "learning_rate": 1.0324378377779553e-11, + "loss": 0.777, + "num_input_tokens_seen": 360062410, + "step": 11077 + }, + { + "epoch": 0.9990530730035623, + "flos": 21589323543840.0, + "grad_norm": 1.836122544073976, + "learning_rate": 8.532545388018064e-12, + "loss": 0.7374, + "num_input_tokens_seen": 360095190, + "step": 11078 + }, + { + "epoch": 0.9991432565270325, + "flos": 28617410426400.0, + "grad_norm": 1.7601712825182325, + "learning_rate": 6.911362697925582e-12, + "loss": 0.575, + "num_input_tokens_seen": 360129270, + "step": 11079 + }, + { + "epoch": 0.9992334400505027, + "flos": 26030657100480.0, + "grad_norm": 1.8474848145517704, + "learning_rate": 5.46083044605794e-12, + "loss": 0.7094, + "num_input_tokens_seen": 360157605, + "step": 11080 + }, + { + "epoch": 0.9993236235739731, + "flos": 25479306385920.0, + "grad_norm": 1.8049620712606609, + "learning_rate": 4.1809487563160276e-12, + "loss": 0.7883, + "num_input_tokens_seen": 360186810, + "step": 11081 + }, + { + "epoch": 0.9994138070974433, + "flos": 24645531135840.0, + "grad_norm": 1.7686915141973738, + "learning_rate": 3.0717177375017e-12, + "loss": 0.7457, + "num_input_tokens_seen": 360217850, + "step": 11082 + }, + { + "epoch": 0.9995039906209136, + "flos": 24536643708000.0, + "grad_norm": 2.491157534156632, + "learning_rate": 2.1331374846500495e-12, + "loss": 0.7382, + "num_input_tokens_seen": 360244735, + "step": 11083 + }, + { + "epoch": 0.9995941741443838, + "flos": 21148086859200.0, + "grad_norm": 2.1108236354158825, + "learning_rate": 1.3652080774750885e-12, + "loss": 0.8361, + "num_input_tokens_seen": 360273275, + "step": 11084 + }, + { + "epoch": 0.9996843576678541, + "flos": 22496520537120.0, + "grad_norm": 1.4947101779041232, + "learning_rate": 7.679295817020204e-13, + "loss": 0.6566, + "num_input_tokens_seen": 360303005, + "step": 11085 + }, + { + "epoch": 0.9997745411913244, + "flos": 27520321631040.0, + "grad_norm": 1.884501795512068, + "learning_rate": 3.413020484011042e-13, + "loss": 0.7152, + "num_input_tokens_seen": 360331720, + "step": 11086 + }, + { + "epoch": 0.9998647247147946, + "flos": 17098845830880.0, + "grad_norm": 1.9920263487019265, + "learning_rate": 8.53255139876552e-14, + "loss": 0.7791, + "num_input_tokens_seen": 360355640, + "step": 11087 + }, + { + "epoch": 0.9999549082382648, + "flos": 22386034809600.0, + "grad_norm": 1.827759274147829, + "learning_rate": 0.0, + "loss": 0.6795, + "num_input_tokens_seen": 360384005, + "step": 11088 + }, + { + "epoch": 0.9999549082382648, + "num_input_tokens_seen": 360384005, + "step": 11088, + "total_flos": 1.3974821586704794e+18, + "train_loss": 0.7512187881860211, + "train_runtime": 130584.3731, + "train_samples_per_second": 5.095, + "train_steps_per_second": 0.085 + } + ], + "logging_steps": 1.0, + "max_steps": 11088, + "num_input_tokens_seen": 360384005, + "num_train_epochs": 1, + "save_steps": 832, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.3974821586704794e+18, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +} diff --git a/sft_full/hyperrouter/training_args.bin b/sft_full/hyperrouter/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..6c98ccaa1b9ecd812195aa23b1089087d7775b29 --- /dev/null +++ b/sft_full/hyperrouter/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ce64b5e180ddbd78a6d2dbb50c07989eae0c5fc9f607e2a2025709f4528ef98d +size 8184 diff --git a/sft_full/smoe/added_tokens.json b/sft_full/smoe/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d3d3a1b74d87e381e471f7b33784015d2dc0ea --- /dev/null +++ b/sft_full/smoe/added_tokens.json @@ -0,0 +1,13 @@ +{ + "<|assistant|>": 32001, + "<|endoftext|>": 32000, + "<|end|>": 32007, + "<|placeholder1|>": 32002, + "<|placeholder2|>": 32003, + "<|placeholder3|>": 32004, + "<|placeholder4|>": 32005, + "<|placeholder5|>": 32008, + "<|placeholder6|>": 32009, + "<|system|>": 32006, + "<|user|>": 32010 +} diff --git a/sft_full/smoe/config.json b/sft_full/smoe/config.json new file mode 100644 index 0000000000000000000000000000000000000000..c5a4ef8c399fb23737d8ab94239edd824a01a75d --- /dev/null +++ b/sft_full/smoe/config.json @@ -0,0 +1,168 @@ +{ + "_name_or_path": "/cm/archive/namnv78/checkpoints/phi35-siglip224/pft", + "architectures": [ + "LlavaPhiForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "auto_map": { + "AutoConfig": "configuration_phi3.Phi3Config", + "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM" + }, + "balance_loss_coef": 0.1, + "bos_token_id": 1, + "clip_smoe": true, + "dropout": false, + "embd_pdrop": 0.0, + "eos_token_id": 32000, + "freeze_mm_mlp_adapter": false, + "hidden_act": "silu", + "hidden_size": 3072, + "image_aspect_ratio": "pad", + "initializer_range": 0.02, + "intermediate_size": 8192, + "local_rank": 0, + "max_position_embeddings": 131072, + "mlp_smoe": true, + "mm_hidden_size": 1152, + "mm_patch_merge_type": "flat", + "mm_projector_lr": null, + "mm_projector_type": "moe", + "mm_use_im_patch_token": false, + "mm_use_im_start_end": false, + "mm_vision_select_feature": "patch", + "mm_vision_select_layer": -2, + "mm_vision_tower": "google/siglip-so400m-patch14-224", + "model_type": "llava_phi", + "moe_name": "smoe", + "num_attention_heads": 32, + "num_experts": 4, + "num_hidden_layers": 32, + "num_key_value_heads": 32, + "num_layers": 3, + "num_selected": 2, + "original_max_position_embeddings": 4096, + "pad_token_id": 32000, + "resid_pdrop": 0.0, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "long_factor": [ + 1.0800000429153442, + 1.1100000143051147, + 1.1399999856948853, + 1.340000033378601, + 1.5899999141693115, + 1.600000023841858, + 1.6200000047683716, + 2.620000123977661, + 3.2300000190734863, + 3.2300000190734863, + 4.789999961853027, + 7.400000095367432, + 7.700000286102295, + 9.09000015258789, + 12.199999809265137, + 17.670000076293945, + 24.46000099182129, + 28.57000160217285, + 30.420001983642578, + 30.840002059936523, + 32.590003967285156, + 32.93000411987305, + 42.320003509521484, + 44.96000289916992, + 50.340003967285156, + 50.45000457763672, + 57.55000305175781, + 57.93000411987305, + 58.21000289916992, + 60.1400032043457, + 62.61000442504883, + 62.62000274658203, + 62.71000289916992, + 63.1400032043457, + 63.1400032043457, + 63.77000427246094, + 63.93000411987305, + 63.96000289916992, + 63.970001220703125, + 64.02999877929688, + 64.06999969482422, + 64.08000183105469, + 64.12000274658203, + 64.41000366210938, + 64.4800033569336, + 64.51000213623047, + 64.52999877929688, + 64.83999633789062 + ], + "short_factor": [ + 1.0, + 1.0199999809265137, + 1.0299999713897705, + 1.0299999713897705, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0699999332427979, + 1.0999999046325684, + 1.1099998950958252, + 1.1599998474121094, + 1.1599998474121094, + 1.1699998378753662, + 1.2899998426437378, + 1.339999794960022, + 1.679999828338623, + 1.7899998426437378, + 1.8199998140335083, + 1.8499997854232788, + 1.8799997568130493, + 1.9099997282028198, + 1.9399996995925903, + 1.9899996519088745, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0799996852874756, + 2.0899996757507324, + 2.189999580383301, + 2.2199995517730713, + 2.5899994373321533, + 2.729999542236328, + 2.749999523162842, + 2.8399994373321533 + ], + "type": "longrope" + }, + "rope_theta": 10000.0, + "router_z_loss_coef": 0.01, + "scales": [ + 1, + 3 + ], + "sliding_window": 262144, + "tie_word_embeddings": false, + "tokenizer_model_max_length": 2048, + "tokenizer_padding_side": "right", + "torch_dtype": "bfloat16", + "training": true, + "transformers_version": "4.43.0", + "tune_mm_mlp_adapter": false, + "use_cache": true, + "use_mm_proj": true, + "vocab_size": 32064 +} diff --git a/sft_full/smoe/generation_config.json b/sft_full/smoe/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..dad5c4578f0dc5969b38755d095fc30c368bb54a --- /dev/null +++ b/sft_full/smoe/generation_config.json @@ -0,0 +1,12 @@ +{ + "_from_model_config": true, + "bos_token_id": 1, + "do_sample": true, + "eos_token_id": [ + 32007, + 32001, + 32000 + ], + "pad_token_id": 32000, + "transformers_version": "4.43.0" +} diff --git a/sft_full/smoe/model-00001-of-00003.safetensors b/sft_full/smoe/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9ea32f403a7bb71c7eda1aa6576164fe048b4c17 --- /dev/null +++ b/sft_full/smoe/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3060186020e81e85f75d680f7a6446b1397f8e60fcf713fae963647075b186f6 +size 4972489328 diff --git a/sft_full/smoe/model-00002-of-00003.safetensors b/sft_full/smoe/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e7e9408c40f7316be0ed0d72cf851cbfed044019 --- /dev/null +++ b/sft_full/smoe/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8cb877ce803d3f1c50c57760d319415451c7d7a4e2221b564707681bffa8c6fb +size 4985529648 diff --git a/sft_full/smoe/model-00003-of-00003.safetensors b/sft_full/smoe/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2ce04c07ef9cd70dfb553effc477bb6ab40d18ae --- /dev/null +++ b/sft_full/smoe/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e0e2b038bcb482466796bb79ce90e466efe95f6772a7318cd0cc19be3e9a828 +size 248943552 diff --git a/sft_full/smoe/model.safetensors.index.json b/sft_full/smoe/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..aa54419fc0a3eab502aa7c4ad974dca52ed10803 --- /dev/null +++ b/sft_full/smoe/model.safetensors.index.json @@ -0,0 +1,1005 @@ +{ + "metadata": { + "total_size": 10206819456 + }, + "weight_map": { + "lm_head.weight": "model-00003-of-00003.safetensors", + "model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.gate.weight": "model-00003-of-00003.safetensors", + "model.norm.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.position_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors" + } +} diff --git a/sft_full/smoe/special_tokens_map.json b/sft_full/smoe/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..3e4d5a5bc1cb51753cc9ae0305ece0da60052b10 --- /dev/null +++ b/sft_full/smoe/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/sft_full/smoe/tokenizer.model b/sft_full/smoe/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/sft_full/smoe/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/sft_full/smoe/tokenizer_config.json b/sft_full/smoe/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d579bb0b91b24b214ea3c2e487e27a65017cdc4a --- /dev/null +++ b/sft_full/smoe/tokenizer_config.json @@ -0,0 +1,132 @@ +{ + "add_bos_token": false, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": false + }, + "32000": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "32001": { + "content": "<|assistant|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32002": { + "content": "<|placeholder1|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32003": { + "content": "<|placeholder2|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32004": { + "content": "<|placeholder3|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32005": { + "content": "<|placeholder4|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32006": { + "content": "<|system|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32007": { + "content": "<|end|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32008": { + "content": "<|placeholder5|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32009": { + "content": "<|placeholder6|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32010": { + "content": "<|user|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "chat_template": "{% for message in messages %}{% if message['role'] == 'system' and message['content'] %}{{'<|system|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'user' %}{{'<|user|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>\n' + message['content'] + '<|end|>\n'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}", + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "legacy": false, + "model_max_length": 2048, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +} diff --git a/sft_full/smoe/trainer_state.json b/sft_full/smoe/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..0eff8ad7a8ba2ab52b20e4bd39ba15a50721e1db --- /dev/null +++ b/sft_full/smoe/trainer_state.json @@ -0,0 +1,77658 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999549082382648, + "eval_steps": 500, + "global_step": 11088, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 9.018352347026198e-05, + "grad_norm": 56.64920409166178, + "learning_rate": 0.0, + "loss": 2.4703, + "step": 1 + }, + { + "epoch": 0.00018036704694052397, + "grad_norm": 54.464072608262825, + "learning_rate": 4.773623799730706e-07, + "loss": 2.3977, + "step": 2 + }, + { + "epoch": 0.0002705505704107859, + "grad_norm": 35.02037458026666, + "learning_rate": 7.566014715123208e-07, + "loss": 2.1225, + "step": 3 + }, + { + "epoch": 0.00036073409388104793, + "grad_norm": 52.21865415535788, + "learning_rate": 9.547247599461412e-07, + "loss": 2.4372, + "step": 4 + }, + { + "epoch": 0.0004509176173513099, + "grad_norm": 40.66762187232, + "learning_rate": 1.108401121501769e-06, + "loss": 2.0398, + "step": 5 + }, + { + "epoch": 0.0005411011408215718, + "grad_norm": 44.58863763836379, + "learning_rate": 1.2339638514853914e-06, + "loss": 2.218, + "step": 6 + }, + { + "epoch": 0.0006312846642918339, + "grad_norm": 35.671086212807744, + "learning_rate": 1.3401256270225321e-06, + "loss": 1.9481, + "step": 7 + }, + { + "epoch": 0.0007214681877620959, + "grad_norm": 28.583305811150503, + "learning_rate": 1.4320871399192119e-06, + "loss": 1.8786, + "step": 8 + }, + { + "epoch": 0.0008116517112323579, + "grad_norm": 16.710180521366027, + "learning_rate": 1.5132029430246416e-06, + "loss": 1.7486, + "step": 9 + }, + { + "epoch": 0.0009018352347026198, + "grad_norm": 17.07217613539782, + "learning_rate": 1.5857635014748399e-06, + "loss": 1.8546, + "step": 10 + }, + { + "epoch": 0.0009920187581728818, + "grad_norm": 19.026155638716716, + "learning_rate": 1.6514025108267924e-06, + "loss": 1.9309, + "step": 11 + }, + { + "epoch": 0.0010822022816431437, + "grad_norm": 12.23888656039761, + "learning_rate": 1.711326231458462e-06, + "loss": 1.8125, + "step": 12 + }, + { + "epoch": 0.0011723858051134058, + "grad_norm": 8.94255083256821, + "learning_rate": 1.7664507107987104e-06, + "loss": 1.637, + "step": 13 + }, + { + "epoch": 0.0012625693285836677, + "grad_norm": 7.025282841211693, + "learning_rate": 1.8174880069956024e-06, + "loss": 1.5969, + "step": 14 + }, + { + "epoch": 0.0013527528520539298, + "grad_norm": 6.310701926194069, + "learning_rate": 1.8650025930140899e-06, + "loss": 1.5514, + "step": 15 + }, + { + "epoch": 0.0014429363755241917, + "grad_norm": 7.002087690478211, + "learning_rate": 1.9094495198922823e-06, + "loss": 1.664, + "step": 16 + }, + { + "epoch": 0.0015331198989944536, + "grad_norm": 6.573082810451778, + "learning_rate": 1.9512009899507514e-06, + "loss": 1.5061, + "step": 17 + }, + { + "epoch": 0.0016233034224647158, + "grad_norm": 6.22790038167459, + "learning_rate": 1.990565322997712e-06, + "loss": 1.557, + "step": 18 + }, + { + "epoch": 0.0017134869459349777, + "grad_norm": 5.391468457563881, + "learning_rate": 2.027800787770518e-06, + "loss": 1.4645, + "step": 19 + }, + { + "epoch": 0.0018036704694052396, + "grad_norm": 5.827101548563864, + "learning_rate": 2.06312588144791e-06, + "loss": 1.5568, + "step": 20 + }, + { + "epoch": 0.0018938539928755017, + "grad_norm": 4.350142570877751, + "learning_rate": 2.0967270985348526e-06, + "loss": 1.4321, + "step": 21 + }, + { + "epoch": 0.0019840375163457636, + "grad_norm": 5.709926447887328, + "learning_rate": 2.128764890799863e-06, + "loss": 1.3351, + "step": 22 + }, + { + "epoch": 0.0020742210398160257, + "grad_norm": 4.2569617305126, + "learning_rate": 2.1593783012990145e-06, + "loss": 1.3301, + "step": 23 + }, + { + "epoch": 0.0021644045632862874, + "grad_norm": 5.058537003482389, + "learning_rate": 2.188688611431533e-06, + "loss": 1.474, + "step": 24 + }, + { + "epoch": 0.0022545880867565495, + "grad_norm": 3.8526785512216004, + "learning_rate": 2.216802243003538e-06, + "loss": 1.4478, + "step": 25 + }, + { + "epoch": 0.0023447716102268116, + "grad_norm": 4.914108688549823, + "learning_rate": 2.243813090771781e-06, + "loss": 1.2846, + "step": 26 + }, + { + "epoch": 0.0024349551336970737, + "grad_norm": 4.144134991512808, + "learning_rate": 2.269804414536962e-06, + "loss": 1.4398, + "step": 27 + }, + { + "epoch": 0.0025251386571673354, + "grad_norm": 3.62140055882338, + "learning_rate": 2.2948503869686733e-06, + "loss": 1.3482, + "step": 28 + }, + { + "epoch": 0.0026153221806375975, + "grad_norm": 4.343548228016634, + "learning_rate": 2.3190173696980436e-06, + "loss": 1.0648, + "step": 29 + }, + { + "epoch": 0.0027055057041078597, + "grad_norm": 3.2509177452718543, + "learning_rate": 2.3423649729871604e-06, + "loss": 1.3449, + "step": 30 + }, + { + "epoch": 0.0027956892275781214, + "grad_norm": 3.1359391057077355, + "learning_rate": 2.364946941580084e-06, + "loss": 1.3367, + "step": 31 + }, + { + "epoch": 0.0028858727510483835, + "grad_norm": 3.6359059934476408, + "learning_rate": 2.3868118998653532e-06, + "loss": 1.3923, + "step": 32 + }, + { + "epoch": 0.0029760562745186456, + "grad_norm": 3.1675421525157117, + "learning_rate": 2.408003982339113e-06, + "loss": 0.9589, + "step": 33 + }, + { + "epoch": 0.0030662397979889073, + "grad_norm": 2.3285810804710314, + "learning_rate": 2.4285633699238223e-06, + "loss": 1.2136, + "step": 34 + }, + { + "epoch": 0.0031564233214591694, + "grad_norm": 2.977156452182432, + "learning_rate": 2.4485267485243007e-06, + "loss": 1.2449, + "step": 35 + }, + { + "epoch": 0.0032466068449294315, + "grad_norm": 2.255907251419716, + "learning_rate": 2.467927702970783e-06, + "loss": 1.2798, + "step": 36 + }, + { + "epoch": 0.003336790368399693, + "grad_norm": 2.558847941810267, + "learning_rate": 2.4867970569753584e-06, + "loss": 1.2983, + "step": 37 + }, + { + "epoch": 0.0034269738918699553, + "grad_norm": 3.3051271258906727, + "learning_rate": 2.5051631677435883e-06, + "loss": 1.1356, + "step": 38 + }, + { + "epoch": 0.0035171574153402174, + "grad_norm": 2.1611199048911307, + "learning_rate": 2.523052182311031e-06, + "loss": 1.1071, + "step": 39 + }, + { + "epoch": 0.003607340938810479, + "grad_norm": 2.3379228505245595, + "learning_rate": 2.540488261420981e-06, + "loss": 1.305, + "step": 40 + }, + { + "epoch": 0.0036975244622807412, + "grad_norm": 2.20321936439088, + "learning_rate": 2.557493775753984e-06, + "loss": 1.1024, + "step": 41 + }, + { + "epoch": 0.0037877079857510034, + "grad_norm": 1.7532399022600156, + "learning_rate": 2.5740894785079235e-06, + "loss": 1.2185, + "step": 42 + }, + { + "epoch": 0.0038778915092212655, + "grad_norm": 2.014298231536251, + "learning_rate": 2.5902946576685834e-06, + "loss": 1.2179, + "step": 43 + }, + { + "epoch": 0.003968075032691527, + "grad_norm": 2.549582978350552, + "learning_rate": 2.606127270772933e-06, + "loss": 1.1424, + "step": 44 + }, + { + "epoch": 0.004058258556161789, + "grad_norm": 2.0445788368106723, + "learning_rate": 2.62160406452641e-06, + "loss": 1.1152, + "step": 45 + }, + { + "epoch": 0.004148442079632051, + "grad_norm": 1.988114505513473, + "learning_rate": 2.636740681272085e-06, + "loss": 1.2772, + "step": 46 + }, + { + "epoch": 0.004238625603102313, + "grad_norm": 2.168445101805877, + "learning_rate": 2.651551754008722e-06, + "loss": 1.3112, + "step": 47 + }, + { + "epoch": 0.004328809126572575, + "grad_norm": 2.3089189381072357, + "learning_rate": 2.6660509914046035e-06, + "loss": 1.2442, + "step": 48 + }, + { + "epoch": 0.004418992650042837, + "grad_norm": 2.319143803719451, + "learning_rate": 2.6802512540450642e-06, + "loss": 1.2022, + "step": 49 + }, + { + "epoch": 0.004509176173513099, + "grad_norm": 5.725244388611387, + "learning_rate": 2.694164622976609e-06, + "loss": 1.1873, + "step": 50 + }, + { + "epoch": 0.0045993596969833616, + "grad_norm": 2.1017503618984428, + "learning_rate": 2.707802461463072e-06, + "loss": 1.1239, + "step": 51 + }, + { + "epoch": 0.004689543220453623, + "grad_norm": 1.9285863908560692, + "learning_rate": 2.7211754707448516e-06, + "loss": 1.1667, + "step": 52 + }, + { + "epoch": 0.004779726743923885, + "grad_norm": 1.681287775005279, + "learning_rate": 2.734293740486721e-06, + "loss": 1.1466, + "step": 53 + }, + { + "epoch": 0.0048699102673941475, + "grad_norm": 1.9075363696669516, + "learning_rate": 2.747166794510033e-06, + "loss": 1.1896, + "step": 54 + }, + { + "epoch": 0.004960093790864409, + "grad_norm": 1.9111114797049666, + "learning_rate": 2.759803632328562e-06, + "loss": 1.159, + "step": 55 + }, + { + "epoch": 0.005050277314334671, + "grad_norm": 2.5384335721971856, + "learning_rate": 2.772212766941744e-06, + "loss": 1.275, + "step": 56 + }, + { + "epoch": 0.005140460837804933, + "grad_norm": 3.6776375213365284, + "learning_rate": 2.7844022592828385e-06, + "loss": 1.2596, + "step": 57 + }, + { + "epoch": 0.005230644361275195, + "grad_norm": 2.6712186129791804, + "learning_rate": 2.7963797496711145e-06, + "loss": 1.1128, + "step": 58 + }, + { + "epoch": 0.005320827884745457, + "grad_norm": 2.152958887610426, + "learning_rate": 2.80815248657541e-06, + "loss": 1.1999, + "step": 59 + }, + { + "epoch": 0.005411011408215719, + "grad_norm": 2.3271827583167357, + "learning_rate": 2.819727352960231e-06, + "loss": 0.8915, + "step": 60 + }, + { + "epoch": 0.005501194931685981, + "grad_norm": 2.2017116477158174, + "learning_rate": 2.8311108904541717e-06, + "loss": 1.1594, + "step": 61 + }, + { + "epoch": 0.005591378455156243, + "grad_norm": 2.045029601836535, + "learning_rate": 2.842309321553155e-06, + "loss": 1.1069, + "step": 62 + }, + { + "epoch": 0.005681561978626505, + "grad_norm": 1.8134101687075441, + "learning_rate": 2.8533285700471737e-06, + "loss": 1.0902, + "step": 63 + }, + { + "epoch": 0.005771745502096767, + "grad_norm": 2.2701861541309687, + "learning_rate": 2.8641742798384237e-06, + "loss": 1.2303, + "step": 64 + }, + { + "epoch": 0.005861929025567029, + "grad_norm": 1.7784303131521357, + "learning_rate": 2.874851832300479e-06, + "loss": 1.184, + "step": 65 + }, + { + "epoch": 0.005952112549037291, + "grad_norm": 2.0509026191764264, + "learning_rate": 2.8853663623121834e-06, + "loss": 1.1866, + "step": 66 + }, + { + "epoch": 0.006042296072507553, + "grad_norm": 2.0545832966536506, + "learning_rate": 2.895722773085839e-06, + "loss": 1.2126, + "step": 67 + }, + { + "epoch": 0.0061324795959778146, + "grad_norm": 1.572959983515541, + "learning_rate": 2.905925749896893e-06, + "loss": 1.1843, + "step": 68 + }, + { + "epoch": 0.006222663119448077, + "grad_norm": 1.7096675531795704, + "learning_rate": 2.915979772811335e-06, + "loss": 1.1692, + "step": 69 + }, + { + "epoch": 0.006312846642918339, + "grad_norm": 2.6150672636990917, + "learning_rate": 2.925889128497372e-06, + "loss": 1.1298, + "step": 70 + }, + { + "epoch": 0.0064030301663886005, + "grad_norm": 2.1027217347999816, + "learning_rate": 2.9356579211992906e-06, + "loss": 1.2285, + "step": 71 + }, + { + "epoch": 0.006493213689858863, + "grad_norm": 1.8564652186514952, + "learning_rate": 2.9452900829438533e-06, + "loss": 1.0983, + "step": 72 + }, + { + "epoch": 0.006583397213329125, + "grad_norm": 2.212408971260902, + "learning_rate": 2.954789383042727e-06, + "loss": 1.1294, + "step": 73 + }, + { + "epoch": 0.006673580736799386, + "grad_norm": 1.5978050255931702, + "learning_rate": 2.9641594369484293e-06, + "loss": 0.9326, + "step": 74 + }, + { + "epoch": 0.006763764260269649, + "grad_norm": 2.233586312195228, + "learning_rate": 2.9734037145158586e-06, + "loss": 1.1519, + "step": 75 + }, + { + "epoch": 0.006853947783739911, + "grad_norm": 2.064219711519473, + "learning_rate": 2.982525547716659e-06, + "loss": 1.1502, + "step": 76 + }, + { + "epoch": 0.006944131307210172, + "grad_norm": 2.4651641879232282, + "learning_rate": 2.9915281378493246e-06, + "loss": 1.1245, + "step": 77 + }, + { + "epoch": 0.007034314830680435, + "grad_norm": 2.0622329539032567, + "learning_rate": 3.000414562284102e-06, + "loss": 1.1465, + "step": 78 + }, + { + "epoch": 0.0071244983541506966, + "grad_norm": 1.876205184167479, + "learning_rate": 3.009187780778246e-06, + "loss": 1.1379, + "step": 79 + }, + { + "epoch": 0.007214681877620958, + "grad_norm": 2.52440376796136, + "learning_rate": 3.017850641394051e-06, + "loss": 1.1321, + "step": 80 + }, + { + "epoch": 0.007304865401091221, + "grad_norm": 1.9985456543289508, + "learning_rate": 3.0264058860492832e-06, + "loss": 1.0009, + "step": 81 + }, + { + "epoch": 0.0073950489245614825, + "grad_norm": 1.7924105039038922, + "learning_rate": 3.0348561557270548e-06, + "loss": 1.1861, + "step": 82 + }, + { + "epoch": 0.007485232448031745, + "grad_norm": 1.2237282827630405, + "learning_rate": 3.043203995369939e-06, + "loss": 0.8845, + "step": 83 + }, + { + "epoch": 0.007575415971502007, + "grad_norm": 2.9533786563825632, + "learning_rate": 3.051451858480994e-06, + "loss": 1.2309, + "step": 84 + }, + { + "epoch": 0.007665599494972268, + "grad_norm": 1.9626250281871165, + "learning_rate": 3.05960211145252e-06, + "loss": 1.06, + "step": 85 + }, + { + "epoch": 0.007755783018442531, + "grad_norm": 1.9263416271956406, + "learning_rate": 3.0676570376416543e-06, + "loss": 1.058, + "step": 86 + }, + { + "epoch": 0.007845966541912792, + "grad_norm": 2.047703233257725, + "learning_rate": 3.0756188412103647e-06, + "loss": 1.1567, + "step": 87 + }, + { + "epoch": 0.007936150065383054, + "grad_norm": 1.7136932893436347, + "learning_rate": 3.083489650746004e-06, + "loss": 1.1334, + "step": 88 + }, + { + "epoch": 0.008026333588853317, + "grad_norm": 2.221744826483259, + "learning_rate": 3.0912715226772975e-06, + "loss": 1.1582, + "step": 89 + }, + { + "epoch": 0.008116517112323578, + "grad_norm": 1.40803194526716, + "learning_rate": 3.098966444499481e-06, + "loss": 1.0717, + "step": 90 + }, + { + "epoch": 0.00820670063579384, + "grad_norm": 2.5191125802827217, + "learning_rate": 3.1065763378212426e-06, + "loss": 1.1417, + "step": 91 + }, + { + "epoch": 0.008296884159264103, + "grad_norm": 1.5085555701459539, + "learning_rate": 3.1141030612451554e-06, + "loss": 1.1422, + "step": 92 + }, + { + "epoch": 0.008387067682734364, + "grad_norm": 1.097846723979175, + "learning_rate": 3.1215484130924052e-06, + "loss": 0.9073, + "step": 93 + }, + { + "epoch": 0.008477251206204626, + "grad_norm": 2.4924349766612814, + "learning_rate": 3.128914133981793e-06, + "loss": 1.147, + "step": 94 + }, + { + "epoch": 0.008567434729674889, + "grad_norm": 2.118637545591757, + "learning_rate": 3.136201909272287e-06, + "loss": 1.1391, + "step": 95 + }, + { + "epoch": 0.00865761825314515, + "grad_norm": 1.9457923266754842, + "learning_rate": 3.1434133713776735e-06, + "loss": 1.0993, + "step": 96 + }, + { + "epoch": 0.008747801776615412, + "grad_norm": 7.2524738695002, + "learning_rate": 3.15055010196128e-06, + "loss": 1.1286, + "step": 97 + }, + { + "epoch": 0.008837985300085675, + "grad_norm": 1.9767556077311461, + "learning_rate": 3.157613634018135e-06, + "loss": 1.1186, + "step": 98 + }, + { + "epoch": 0.008928168823555935, + "grad_norm": 2.038076929895943, + "learning_rate": 3.1646054538514336e-06, + "loss": 1.1445, + "step": 99 + }, + { + "epoch": 0.009018352347026198, + "grad_norm": 2.327401170281561, + "learning_rate": 3.1715270029496797e-06, + "loss": 1.1748, + "step": 100 + }, + { + "epoch": 0.00910853587049646, + "grad_norm": 1.9174066718315983, + "learning_rate": 3.1783796797704243e-06, + "loss": 1.1604, + "step": 101 + }, + { + "epoch": 0.009198719393966723, + "grad_norm": 2.0515134249161235, + "learning_rate": 3.185164841436142e-06, + "loss": 1.0842, + "step": 102 + }, + { + "epoch": 0.009288902917436984, + "grad_norm": 2.0850927807040835, + "learning_rate": 3.1918838053473723e-06, + "loss": 1.1197, + "step": 103 + }, + { + "epoch": 0.009379086440907246, + "grad_norm": 2.5629341874143656, + "learning_rate": 3.198537850717922e-06, + "loss": 1.0253, + "step": 104 + }, + { + "epoch": 0.009469269964377509, + "grad_norm": 2.071614787934727, + "learning_rate": 3.205128220036622e-06, + "loss": 1.0801, + "step": 105 + }, + { + "epoch": 0.00955945348784777, + "grad_norm": 1.813325496661325, + "learning_rate": 3.2116561204597917e-06, + "loss": 1.0778, + "step": 106 + }, + { + "epoch": 0.009649637011318032, + "grad_norm": 1.0818503879367827, + "learning_rate": 3.218122725138335e-06, + "loss": 0.8918, + "step": 107 + }, + { + "epoch": 0.009739820534788295, + "grad_norm": 2.1649152817570156, + "learning_rate": 3.224529174483104e-06, + "loss": 1.1126, + "step": 108 + }, + { + "epoch": 0.009830004058258556, + "grad_norm": 2.9198730464667313, + "learning_rate": 3.2308765773719435e-06, + "loss": 1.0566, + "step": 109 + }, + { + "epoch": 0.009920187581728818, + "grad_norm": 1.7724800382063104, + "learning_rate": 3.2371660123016323e-06, + "loss": 1.1071, + "step": 110 + }, + { + "epoch": 0.010010371105199081, + "grad_norm": 1.926917307807138, + "learning_rate": 3.2433985284876787e-06, + "loss": 1.0315, + "step": 111 + }, + { + "epoch": 0.010100554628669342, + "grad_norm": 1.9054205293641964, + "learning_rate": 3.2495751469148143e-06, + "loss": 1.1058, + "step": 112 + }, + { + "epoch": 0.010190738152139604, + "grad_norm": 2.512805472256089, + "learning_rate": 3.2556968613407816e-06, + "loss": 1.0814, + "step": 113 + }, + { + "epoch": 0.010280921675609867, + "grad_norm": 2.426516312493061, + "learning_rate": 3.2617646392559094e-06, + "loss": 1.1473, + "step": 114 + }, + { + "epoch": 0.010371105199080128, + "grad_norm": 1.964613307536787, + "learning_rate": 3.2677794228007836e-06, + "loss": 1.1563, + "step": 115 + }, + { + "epoch": 0.01046128872255039, + "grad_norm": 1.9175051499835987, + "learning_rate": 3.273742129644185e-06, + "loss": 1.0933, + "step": 116 + }, + { + "epoch": 0.010551472246020653, + "grad_norm": 2.4801295251246986, + "learning_rate": 3.279653653823352e-06, + "loss": 1.1118, + "step": 117 + }, + { + "epoch": 0.010641655769490914, + "grad_norm": 1.841821923191783, + "learning_rate": 3.285514866548481e-06, + "loss": 1.109, + "step": 118 + }, + { + "epoch": 0.010731839292961176, + "grad_norm": 1.8719579178367587, + "learning_rate": 3.2913266169732838e-06, + "loss": 1.166, + "step": 119 + }, + { + "epoch": 0.010822022816431439, + "grad_norm": 2.1686787310496407, + "learning_rate": 3.2970897329333017e-06, + "loss": 1.0831, + "step": 120 + }, + { + "epoch": 0.0109122063399017, + "grad_norm": 1.8298140979625273, + "learning_rate": 3.302805021653585e-06, + "loss": 1.0825, + "step": 121 + }, + { + "epoch": 0.011002389863371962, + "grad_norm": 2.3318319826598195, + "learning_rate": 3.3084732704272426e-06, + "loss": 1.0952, + "step": 122 + }, + { + "epoch": 0.011092573386842225, + "grad_norm": 2.32111547685262, + "learning_rate": 3.314095247266304e-06, + "loss": 1.0959, + "step": 123 + }, + { + "epoch": 0.011182756910312485, + "grad_norm": 13.039153778140665, + "learning_rate": 3.3196717015262255e-06, + "loss": 1.0576, + "step": 124 + }, + { + "epoch": 0.011272940433782748, + "grad_norm": 1.8759671577881125, + "learning_rate": 3.325203364505307e-06, + "loss": 1.086, + "step": 125 + }, + { + "epoch": 0.01136312395725301, + "grad_norm": 2.246792838725935, + "learning_rate": 3.3306909500202442e-06, + "loss": 1.1317, + "step": 126 + }, + { + "epoch": 0.011453307480723271, + "grad_norm": 2.333592389121998, + "learning_rate": 3.3361351549589145e-06, + "loss": 1.161, + "step": 127 + }, + { + "epoch": 0.011543491004193534, + "grad_norm": 2.112485386724546, + "learning_rate": 3.341536659811494e-06, + "loss": 1.0347, + "step": 128 + }, + { + "epoch": 0.011633674527663796, + "grad_norm": 2.1650896079392403, + "learning_rate": 3.346896129180904e-06, + "loss": 1.1221, + "step": 129 + }, + { + "epoch": 0.011723858051134057, + "grad_norm": 1.738323356889946, + "learning_rate": 3.35221421227355e-06, + "loss": 1.1045, + "step": 130 + }, + { + "epoch": 0.01181404157460432, + "grad_norm": 2.3393669343116805, + "learning_rate": 3.357491543371255e-06, + "loss": 1.1213, + "step": 131 + }, + { + "epoch": 0.011904225098074582, + "grad_norm": 1.6012086803099028, + "learning_rate": 3.3627287422852543e-06, + "loss": 1.0495, + "step": 132 + }, + { + "epoch": 0.011994408621544843, + "grad_norm": 1.633522426921832, + "learning_rate": 3.3679264147930497e-06, + "loss": 1.1227, + "step": 133 + }, + { + "epoch": 0.012084592145015106, + "grad_norm": 1.9456806370726651, + "learning_rate": 3.37308515305891e-06, + "loss": 1.0435, + "step": 134 + }, + { + "epoch": 0.012174775668485368, + "grad_norm": 1.5688187815872996, + "learning_rate": 3.3782055360387313e-06, + "loss": 1.1631, + "step": 135 + }, + { + "epoch": 0.012264959191955629, + "grad_norm": 1.835817438623321, + "learning_rate": 3.3832881298699633e-06, + "loss": 1.054, + "step": 136 + }, + { + "epoch": 0.012355142715425892, + "grad_norm": 1.7907022349692119, + "learning_rate": 3.388333488247249e-06, + "loss": 1.0681, + "step": 137 + }, + { + "epoch": 0.012445326238896154, + "grad_norm": 1.6171926599687456, + "learning_rate": 3.393342152784406e-06, + "loss": 1.1497, + "step": 138 + }, + { + "epoch": 0.012535509762366415, + "grad_norm": 1.8655855337597909, + "learning_rate": 3.3983146533633376e-06, + "loss": 1.1075, + "step": 139 + }, + { + "epoch": 0.012625693285836678, + "grad_norm": 2.479223031650788, + "learning_rate": 3.403251508470442e-06, + "loss": 1.1585, + "step": 140 + }, + { + "epoch": 0.01271587680930694, + "grad_norm": 1.7390015213550056, + "learning_rate": 3.408153225521043e-06, + "loss": 1.0554, + "step": 141 + }, + { + "epoch": 0.012806060332777201, + "grad_norm": 2.119452696995676, + "learning_rate": 3.413020301172361e-06, + "loss": 0.9882, + "step": 142 + }, + { + "epoch": 0.012896243856247463, + "grad_norm": 1.7060528008942417, + "learning_rate": 3.4178532216255024e-06, + "loss": 1.079, + "step": 143 + }, + { + "epoch": 0.012986427379717726, + "grad_norm": 2.258100022748779, + "learning_rate": 3.422652462916924e-06, + "loss": 1.1333, + "step": 144 + }, + { + "epoch": 0.013076610903187987, + "grad_norm": 2.4600879958937005, + "learning_rate": 3.4274184911998124e-06, + "loss": 0.9979, + "step": 145 + }, + { + "epoch": 0.01316679442665825, + "grad_norm": 2.5672886462262072, + "learning_rate": 3.4321517630157976e-06, + "loss": 1.1854, + "step": 146 + }, + { + "epoch": 0.013256977950128512, + "grad_norm": 1.8687339986160734, + "learning_rate": 3.4368527255573845e-06, + "loss": 1.1562, + "step": 147 + }, + { + "epoch": 0.013347161473598773, + "grad_norm": 2.0556559437933504, + "learning_rate": 3.4415218169214994e-06, + "loss": 1.0651, + "step": 148 + }, + { + "epoch": 0.013437344997069035, + "grad_norm": 1.5928321888353967, + "learning_rate": 3.4461594663544882e-06, + "loss": 1.0712, + "step": 149 + }, + { + "epoch": 0.013527528520539298, + "grad_norm": 1.9998971197370305, + "learning_rate": 3.450766094488929e-06, + "loss": 1.0365, + "step": 150 + }, + { + "epoch": 0.013617712044009559, + "grad_norm": 1.9126906480696277, + "learning_rate": 3.4553421135725735e-06, + "loss": 1.1405, + "step": 151 + }, + { + "epoch": 0.013707895567479821, + "grad_norm": 1.789927710666445, + "learning_rate": 3.45988792768973e-06, + "loss": 1.1034, + "step": 152 + }, + { + "epoch": 0.013798079090950084, + "grad_norm": 1.6443818448082654, + "learning_rate": 3.464403932975393e-06, + "loss": 1.1539, + "step": 153 + }, + { + "epoch": 0.013888262614420345, + "grad_norm": 1.675401124323194, + "learning_rate": 3.468890517822395e-06, + "loss": 1.1438, + "step": 154 + }, + { + "epoch": 0.013978446137890607, + "grad_norm": 3.423363906020573, + "learning_rate": 3.473348063081853e-06, + "loss": 1.0821, + "step": 155 + }, + { + "epoch": 0.01406862966136087, + "grad_norm": 2.100887828434344, + "learning_rate": 3.4777769422571727e-06, + "loss": 1.0396, + "step": 156 + }, + { + "epoch": 0.01415881318483113, + "grad_norm": 1.6263532872279993, + "learning_rate": 3.4821775216918497e-06, + "loss": 1.0857, + "step": 157 + }, + { + "epoch": 0.014248996708301393, + "grad_norm": 1.8074096362687064, + "learning_rate": 3.4865501607513164e-06, + "loss": 1.0333, + "step": 158 + }, + { + "epoch": 0.014339180231771656, + "grad_norm": 1.8651604922376517, + "learning_rate": 3.4908952119990423e-06, + "loss": 1.1569, + "step": 159 + }, + { + "epoch": 0.014429363755241916, + "grad_norm": 1.4080420625510535, + "learning_rate": 3.495213021367122e-06, + "loss": 0.8468, + "step": 160 + }, + { + "epoch": 0.014519547278712179, + "grad_norm": 11.312502023264667, + "learning_rate": 3.4995039283215464e-06, + "loss": 1.1122, + "step": 161 + }, + { + "epoch": 0.014609730802182442, + "grad_norm": 1.6840372939567383, + "learning_rate": 3.5037682660223533e-06, + "loss": 1.1018, + "step": 162 + }, + { + "epoch": 0.014699914325652702, + "grad_norm": 2.0987026794079133, + "learning_rate": 3.508006361478857e-06, + "loss": 1.1179, + "step": 163 + }, + { + "epoch": 0.014790097849122965, + "grad_norm": 1.7079907825944889, + "learning_rate": 3.5122185357001253e-06, + "loss": 1.0698, + "step": 164 + }, + { + "epoch": 0.014880281372593228, + "grad_norm": 2.3000755587896147, + "learning_rate": 3.5164051038408817e-06, + "loss": 1.1464, + "step": 165 + }, + { + "epoch": 0.01497046489606349, + "grad_norm": 1.973273394142587, + "learning_rate": 3.5205663753430093e-06, + "loss": 1.1301, + "step": 166 + }, + { + "epoch": 0.015060648419533751, + "grad_norm": 1.568234335355456, + "learning_rate": 3.5247026540727915e-06, + "loss": 1.1814, + "step": 167 + }, + { + "epoch": 0.015150831943004013, + "grad_norm": 9.676328850698258, + "learning_rate": 3.5288142384540645e-06, + "loss": 1.0325, + "step": 168 + }, + { + "epoch": 0.015241015466474276, + "grad_norm": 1.9582828014699094, + "learning_rate": 3.532901421597421e-06, + "loss": 1.1161, + "step": 169 + }, + { + "epoch": 0.015331198989944537, + "grad_norm": 2.280497596717711, + "learning_rate": 3.5369644914255915e-06, + "loss": 1.0339, + "step": 170 + }, + { + "epoch": 0.0154213825134148, + "grad_norm": 4.756795339111549, + "learning_rate": 3.5410037307951596e-06, + "loss": 1.0823, + "step": 171 + }, + { + "epoch": 0.015511566036885062, + "grad_norm": 1.553008586734533, + "learning_rate": 3.545019417614725e-06, + "loss": 1.132, + "step": 172 + }, + { + "epoch": 0.015601749560355323, + "grad_norm": 3.0442388623693226, + "learning_rate": 3.5490118249596387e-06, + "loss": 1.1265, + "step": 173 + }, + { + "epoch": 0.015691933083825584, + "grad_norm": 1.8298230232587396, + "learning_rate": 3.5529812211834352e-06, + "loss": 1.1392, + "step": 174 + }, + { + "epoch": 0.015782116607295848, + "grad_norm": 1.80109368052995, + "learning_rate": 3.5569278700260707e-06, + "loss": 1.0886, + "step": 175 + }, + { + "epoch": 0.01587230013076611, + "grad_norm": 3.2154983836641375, + "learning_rate": 3.5608520307190746e-06, + "loss": 1.074, + "step": 176 + }, + { + "epoch": 0.01596248365423637, + "grad_norm": 1.7181810304255511, + "learning_rate": 3.564753958087731e-06, + "loss": 1.0229, + "step": 177 + }, + { + "epoch": 0.016052667177706634, + "grad_norm": 2.413726289416372, + "learning_rate": 3.5686339026503684e-06, + "loss": 1.094, + "step": 178 + }, + { + "epoch": 0.016142850701176895, + "grad_norm": 1.844487640757218, + "learning_rate": 3.5724921107148806e-06, + "loss": 1.1699, + "step": 179 + }, + { + "epoch": 0.016233034224647155, + "grad_norm": 1.921057922848302, + "learning_rate": 3.576328824472552e-06, + "loss": 1.1445, + "step": 180 + }, + { + "epoch": 0.01632321774811742, + "grad_norm": 1.764355633609439, + "learning_rate": 3.5801442820892838e-06, + "loss": 1.1787, + "step": 181 + }, + { + "epoch": 0.01641340127158768, + "grad_norm": 1.4537658918829055, + "learning_rate": 3.583938717794313e-06, + "loss": 1.1092, + "step": 182 + }, + { + "epoch": 0.01650358479505794, + "grad_norm": 1.9978034354997318, + "learning_rate": 3.5877123619664928e-06, + "loss": 1.1294, + "step": 183 + }, + { + "epoch": 0.016593768318528206, + "grad_norm": 1.6559596257010327, + "learning_rate": 3.5914654412182268e-06, + "loss": 1.1047, + "step": 184 + }, + { + "epoch": 0.016683951841998466, + "grad_norm": 1.891107670987192, + "learning_rate": 3.595198178477127e-06, + "loss": 1.1169, + "step": 185 + }, + { + "epoch": 0.016774135365468727, + "grad_norm": 1.56452574067292, + "learning_rate": 3.5989107930654757e-06, + "loss": 1.0167, + "step": 186 + }, + { + "epoch": 0.01686431888893899, + "grad_norm": 1.881069151844827, + "learning_rate": 3.6026035007775437e-06, + "loss": 1.1531, + "step": 187 + }, + { + "epoch": 0.016954502412409252, + "grad_norm": 2.204427023247532, + "learning_rate": 3.6062765139548636e-06, + "loss": 1.0565, + "step": 188 + }, + { + "epoch": 0.017044685935879513, + "grad_norm": 1.7253057816976958, + "learning_rate": 3.6099300415594945e-06, + "loss": 0.9485, + "step": 189 + }, + { + "epoch": 0.017134869459349777, + "grad_norm": 1.6730692510825267, + "learning_rate": 3.6135642892453575e-06, + "loss": 0.9642, + "step": 190 + }, + { + "epoch": 0.01722505298282004, + "grad_norm": 1.5554218187441835, + "learning_rate": 3.6171794594277004e-06, + "loss": 1.1003, + "step": 191 + }, + { + "epoch": 0.0173152365062903, + "grad_norm": 1.8289438353810965, + "learning_rate": 3.620775751350745e-06, + "loss": 1.0317, + "step": 192 + }, + { + "epoch": 0.017405420029760563, + "grad_norm": 2.365193146929852, + "learning_rate": 3.6243533611535794e-06, + "loss": 1.1245, + "step": 193 + }, + { + "epoch": 0.017495603553230824, + "grad_norm": 1.6466400326313608, + "learning_rate": 3.627912481934351e-06, + "loss": 1.0783, + "step": 194 + }, + { + "epoch": 0.017585787076701085, + "grad_norm": 2.141927726459503, + "learning_rate": 3.6314533038128e-06, + "loss": 1.0607, + "step": 195 + }, + { + "epoch": 0.01767597060017135, + "grad_norm": 1.0407083999159903, + "learning_rate": 3.6349760139912048e-06, + "loss": 0.8271, + "step": 196 + }, + { + "epoch": 0.01776615412364161, + "grad_norm": 1.9491090915695903, + "learning_rate": 3.638480796813769e-06, + "loss": 1.1449, + "step": 197 + }, + { + "epoch": 0.01785633764711187, + "grad_norm": 2.0287759826462977, + "learning_rate": 3.641967833824504e-06, + "loss": 1.0798, + "step": 198 + }, + { + "epoch": 0.017946521170582135, + "grad_norm": 1.8901778393792978, + "learning_rate": 3.645437303823663e-06, + "loss": 1.0756, + "step": 199 + }, + { + "epoch": 0.018036704694052396, + "grad_norm": 2.7752447553041626, + "learning_rate": 3.64888938292275e-06, + "loss": 1.0687, + "step": 200 + }, + { + "epoch": 0.01812688821752266, + "grad_norm": 1.7106886033821875, + "learning_rate": 3.6523242445981603e-06, + "loss": 1.0894, + "step": 201 + }, + { + "epoch": 0.01821707174099292, + "grad_norm": 1.5884394711833627, + "learning_rate": 3.655742059743495e-06, + "loss": 1.0489, + "step": 202 + }, + { + "epoch": 0.018307255264463182, + "grad_norm": 1.7655006972586769, + "learning_rate": 3.659142996720576e-06, + "loss": 1.1403, + "step": 203 + }, + { + "epoch": 0.018397438787933446, + "grad_norm": 2.2327952101981037, + "learning_rate": 3.6625272214092135e-06, + "loss": 1.0531, + "step": 204 + }, + { + "epoch": 0.018487622311403707, + "grad_norm": 0.861735517904151, + "learning_rate": 3.6658948972557535e-06, + "loss": 0.7994, + "step": 205 + }, + { + "epoch": 0.018577805834873968, + "grad_norm": 1.7201643239858428, + "learning_rate": 3.6692461853204432e-06, + "loss": 1.0018, + "step": 206 + }, + { + "epoch": 0.018667989358344232, + "grad_norm": 1.81747647822506, + "learning_rate": 3.672581244323656e-06, + "loss": 1.0208, + "step": 207 + }, + { + "epoch": 0.018758172881814493, + "grad_norm": 2.0421378960747174, + "learning_rate": 3.6759002306909926e-06, + "loss": 0.9607, + "step": 208 + }, + { + "epoch": 0.018848356405284754, + "grad_norm": 1.570252754251524, + "learning_rate": 3.67920329859731e-06, + "loss": 1.0872, + "step": 209 + }, + { + "epoch": 0.018938539928755018, + "grad_norm": 1.7523491623378646, + "learning_rate": 3.6824906000096923e-06, + "loss": 1.0811, + "step": 210 + }, + { + "epoch": 0.01902872345222528, + "grad_norm": 1.9530302101021517, + "learning_rate": 3.6857622847294067e-06, + "loss": 1.0875, + "step": 211 + }, + { + "epoch": 0.01911890697569554, + "grad_norm": 4.318497619762518, + "learning_rate": 3.6890185004328626e-06, + "loss": 1.0756, + "step": 212 + }, + { + "epoch": 0.019209090499165804, + "grad_norm": 2.1413981579833314, + "learning_rate": 3.6922593927116113e-06, + "loss": 1.0147, + "step": 213 + }, + { + "epoch": 0.019299274022636065, + "grad_norm": 2.3179223694000592, + "learning_rate": 3.695485105111406e-06, + "loss": 1.0646, + "step": 214 + }, + { + "epoch": 0.019389457546106326, + "grad_norm": 1.5265457917688108, + "learning_rate": 3.698695779170352e-06, + "loss": 1.047, + "step": 215 + }, + { + "epoch": 0.01947964106957659, + "grad_norm": 1.784645324646871, + "learning_rate": 3.7018915544561744e-06, + "loss": 1.0666, + "step": 216 + }, + { + "epoch": 0.01956982459304685, + "grad_norm": 2.4842286456682565, + "learning_rate": 3.7050725686026164e-06, + "loss": 1.0266, + "step": 217 + }, + { + "epoch": 0.01966000811651711, + "grad_norm": 1.99926243533443, + "learning_rate": 3.708238957345014e-06, + "loss": 1.1349, + "step": 218 + }, + { + "epoch": 0.019750191639987376, + "grad_norm": 1.8294732754307008, + "learning_rate": 3.7113908545550482e-06, + "loss": 1.1131, + "step": 219 + }, + { + "epoch": 0.019840375163457637, + "grad_norm": 1.7029216539972443, + "learning_rate": 3.7145283922747028e-06, + "loss": 1.0675, + "step": 220 + }, + { + "epoch": 0.019930558686927898, + "grad_norm": 1.6295562106068087, + "learning_rate": 3.7176517007494612e-06, + "loss": 1.1608, + "step": 221 + }, + { + "epoch": 0.020020742210398162, + "grad_norm": 1.7506131051860294, + "learning_rate": 3.7207609084607496e-06, + "loss": 1.0157, + "step": 222 + }, + { + "epoch": 0.020110925733868423, + "grad_norm": 2.74319456904428, + "learning_rate": 3.723856142157645e-06, + "loss": 1.1236, + "step": 223 + }, + { + "epoch": 0.020201109257338683, + "grad_norm": 1.8707324099205043, + "learning_rate": 3.726937526887885e-06, + "loss": 1.0828, + "step": 224 + }, + { + "epoch": 0.020291292780808948, + "grad_norm": 1.9308165054455086, + "learning_rate": 3.7300051860281798e-06, + "loss": 1.039, + "step": 225 + }, + { + "epoch": 0.02038147630427921, + "grad_norm": 2.107592133016161, + "learning_rate": 3.733059241313852e-06, + "loss": 1.1548, + "step": 226 + }, + { + "epoch": 0.02047165982774947, + "grad_norm": 1.8601166984906576, + "learning_rate": 3.736099812867827e-06, + "loss": 1.059, + "step": 227 + }, + { + "epoch": 0.020561843351219734, + "grad_norm": 2.0812675406601775, + "learning_rate": 3.73912701922898e-06, + "loss": 1.1331, + "step": 228 + }, + { + "epoch": 0.020652026874689994, + "grad_norm": 1.1447764579912785, + "learning_rate": 3.742140977379868e-06, + "loss": 0.8294, + "step": 229 + }, + { + "epoch": 0.020742210398160255, + "grad_norm": 1.7335502106750524, + "learning_rate": 3.745141802773854e-06, + "loss": 1.1034, + "step": 230 + }, + { + "epoch": 0.02083239392163052, + "grad_norm": 2.0218608118706207, + "learning_rate": 3.748129609361645e-06, + "loss": 1.0628, + "step": 231 + }, + { + "epoch": 0.02092257744510078, + "grad_norm": 1.3963804406389702, + "learning_rate": 3.7511045096172555e-06, + "loss": 1.0446, + "step": 232 + }, + { + "epoch": 0.02101276096857104, + "grad_norm": 2.0150997453333614, + "learning_rate": 3.7540666145634137e-06, + "loss": 1.0612, + "step": 233 + }, + { + "epoch": 0.021102944492041305, + "grad_norm": 1.642872315686217, + "learning_rate": 3.7570160337964225e-06, + "loss": 1.0472, + "step": 234 + }, + { + "epoch": 0.021193128015511566, + "grad_norm": 2.023751959796826, + "learning_rate": 3.7599528755104913e-06, + "loss": 1.1043, + "step": 235 + }, + { + "epoch": 0.021283311538981827, + "grad_norm": 1.5480882914891676, + "learning_rate": 3.7628772465215515e-06, + "loss": 1.0681, + "step": 236 + }, + { + "epoch": 0.02137349506245209, + "grad_norm": 1.7927458048984064, + "learning_rate": 3.7657892522905666e-06, + "loss": 1.0605, + "step": 237 + }, + { + "epoch": 0.021463678585922352, + "grad_norm": 2.4780535616685713, + "learning_rate": 3.7686889969463542e-06, + "loss": 0.984, + "step": 238 + }, + { + "epoch": 0.021553862109392613, + "grad_norm": 1.9409843615806477, + "learning_rate": 3.771576583307928e-06, + "loss": 1.0462, + "step": 239 + }, + { + "epoch": 0.021644045632862877, + "grad_norm": 2.159557845990573, + "learning_rate": 3.7744521129063722e-06, + "loss": 1.1042, + "step": 240 + }, + { + "epoch": 0.021734229156333138, + "grad_norm": 1.7521008415420887, + "learning_rate": 3.7773156860062653e-06, + "loss": 1.0915, + "step": 241 + }, + { + "epoch": 0.0218244126798034, + "grad_norm": 2.0174192973543925, + "learning_rate": 3.7801674016266554e-06, + "loss": 1.1686, + "step": 242 + }, + { + "epoch": 0.021914596203273663, + "grad_norm": 0.7322277574791446, + "learning_rate": 3.7830073575616035e-06, + "loss": 0.7725, + "step": 243 + }, + { + "epoch": 0.022004779726743924, + "grad_norm": 1.8753063587405367, + "learning_rate": 3.785835650400313e-06, + "loss": 1.1052, + "step": 244 + }, + { + "epoch": 0.022094963250214185, + "grad_norm": 1.891780594687362, + "learning_rate": 3.7886523755468334e-06, + "loss": 0.9752, + "step": 245 + }, + { + "epoch": 0.02218514677368445, + "grad_norm": 1.5617737417381208, + "learning_rate": 3.7914576272393746e-06, + "loss": 1.0606, + "step": 246 + }, + { + "epoch": 0.02227533029715471, + "grad_norm": 2.181933923743569, + "learning_rate": 3.7942514985692284e-06, + "loss": 1.1719, + "step": 247 + }, + { + "epoch": 0.02236551382062497, + "grad_norm": 1.598109397886617, + "learning_rate": 3.797034081499296e-06, + "loss": 1.06, + "step": 248 + }, + { + "epoch": 0.022455697344095235, + "grad_norm": 1.8686302708063998, + "learning_rate": 3.7998054668822595e-06, + "loss": 1.0443, + "step": 249 + }, + { + "epoch": 0.022545880867565496, + "grad_norm": 2.0742909334485637, + "learning_rate": 3.8025657444783776e-06, + "loss": 1.0441, + "step": 250 + }, + { + "epoch": 0.022636064391035757, + "grad_norm": 2.053693988057138, + "learning_rate": 3.80531500297293e-06, + "loss": 0.9785, + "step": 251 + }, + { + "epoch": 0.02272624791450602, + "grad_norm": 2.184529494687398, + "learning_rate": 3.8080533299933147e-06, + "loss": 1.0424, + "step": 252 + }, + { + "epoch": 0.022816431437976282, + "grad_norm": 1.8580460487424515, + "learning_rate": 3.8107808121258067e-06, + "loss": 1.0318, + "step": 253 + }, + { + "epoch": 0.022906614961446543, + "grad_norm": 1.6742310367334527, + "learning_rate": 3.813497534931985e-06, + "loss": 1.126, + "step": 254 + }, + { + "epoch": 0.022996798484916807, + "grad_norm": 1.762774571884953, + "learning_rate": 3.816203582964841e-06, + "loss": 1.0917, + "step": 255 + }, + { + "epoch": 0.023086982008387068, + "grad_norm": 1.7415659527048781, + "learning_rate": 3.818899039784565e-06, + "loss": 1.1233, + "step": 256 + }, + { + "epoch": 0.02317716553185733, + "grad_norm": 1.7097765323673604, + "learning_rate": 3.821583987974031e-06, + "loss": 1.1251, + "step": 257 + }, + { + "epoch": 0.023267349055327593, + "grad_norm": 7.2757149151719425, + "learning_rate": 3.8242585091539755e-06, + "loss": 1.0091, + "step": 258 + }, + { + "epoch": 0.023357532578797854, + "grad_norm": 1.6293448536918749, + "learning_rate": 3.8269226839978895e-06, + "loss": 1.1501, + "step": 259 + }, + { + "epoch": 0.023447716102268115, + "grad_norm": 2.162755468195598, + "learning_rate": 3.82957659224662e-06, + "loss": 1.023, + "step": 260 + }, + { + "epoch": 0.02353789962573838, + "grad_norm": 2.177429889553204, + "learning_rate": 3.8322203127226855e-06, + "loss": 0.9546, + "step": 261 + }, + { + "epoch": 0.02362808314920864, + "grad_norm": 1.851376552338774, + "learning_rate": 3.834853923344326e-06, + "loss": 1.1812, + "step": 262 + }, + { + "epoch": 0.0237182666726789, + "grad_norm": 1.6719982779142646, + "learning_rate": 3.837477501139285e-06, + "loss": 1.0707, + "step": 263 + }, + { + "epoch": 0.023808450196149165, + "grad_norm": 4.333372604974666, + "learning_rate": 3.840091122258324e-06, + "loss": 1.0801, + "step": 264 + }, + { + "epoch": 0.023898633719619426, + "grad_norm": 1.7724709586708791, + "learning_rate": 3.84269486198849e-06, + "loss": 0.9903, + "step": 265 + }, + { + "epoch": 0.023988817243089686, + "grad_norm": 1.6317538247032222, + "learning_rate": 3.845288794766121e-06, + "loss": 1.0905, + "step": 266 + }, + { + "epoch": 0.02407900076655995, + "grad_norm": 2.0564758872397664, + "learning_rate": 3.847872994189619e-06, + "loss": 1.0637, + "step": 267 + }, + { + "epoch": 0.02416918429003021, + "grad_norm": 1.7802867627338321, + "learning_rate": 3.8504475330319805e-06, + "loss": 1.1282, + "step": 268 + }, + { + "epoch": 0.024259367813500472, + "grad_norm": 1.82472088129444, + "learning_rate": 3.853012483253093e-06, + "loss": 1.0468, + "step": 269 + }, + { + "epoch": 0.024349551336970737, + "grad_norm": 2.6424040811068887, + "learning_rate": 3.855567916011802e-06, + "loss": 1.0746, + "step": 270 + }, + { + "epoch": 0.024439734860440997, + "grad_norm": 2.2249944151047467, + "learning_rate": 3.858113901677755e-06, + "loss": 1.0771, + "step": 271 + }, + { + "epoch": 0.024529918383911258, + "grad_norm": 1.7012739008584676, + "learning_rate": 3.860650509843034e-06, + "loss": 1.0294, + "step": 272 + }, + { + "epoch": 0.024620101907381522, + "grad_norm": 2.436314123179813, + "learning_rate": 3.863177809333563e-06, + "loss": 1.0534, + "step": 273 + }, + { + "epoch": 0.024710285430851783, + "grad_norm": 1.5618891475627494, + "learning_rate": 3.86569586822032e-06, + "loss": 1.0418, + "step": 274 + }, + { + "epoch": 0.024800468954322044, + "grad_norm": 2.1240251212277808, + "learning_rate": 3.868204753830331e-06, + "loss": 1.0325, + "step": 275 + }, + { + "epoch": 0.02489065247779231, + "grad_norm": 1.5924441653873138, + "learning_rate": 3.870704532757476e-06, + "loss": 0.9825, + "step": 276 + }, + { + "epoch": 0.02498083600126257, + "grad_norm": 1.8941629897402337, + "learning_rate": 3.8731952708730974e-06, + "loss": 1.0963, + "step": 277 + }, + { + "epoch": 0.02507101952473283, + "grad_norm": 2.0524020520348887, + "learning_rate": 3.8756770333364085e-06, + "loss": 1.0886, + "step": 278 + }, + { + "epoch": 0.025161203048203094, + "grad_norm": 2.276546489211565, + "learning_rate": 3.878149884604725e-06, + "loss": 0.9874, + "step": 279 + }, + { + "epoch": 0.025251386571673355, + "grad_norm": 2.8562833821622653, + "learning_rate": 3.8806138884435125e-06, + "loss": 1.0437, + "step": 280 + }, + { + "epoch": 0.025341570095143616, + "grad_norm": 1.6299081186715032, + "learning_rate": 3.883069107936248e-06, + "loss": 0.9927, + "step": 281 + }, + { + "epoch": 0.02543175361861388, + "grad_norm": 3.2159823768664264, + "learning_rate": 3.885515605494114e-06, + "loss": 1.1225, + "step": 282 + }, + { + "epoch": 0.02552193714208414, + "grad_norm": 2.6841163626118694, + "learning_rate": 3.8879534428655145e-06, + "loss": 1.1248, + "step": 283 + }, + { + "epoch": 0.025612120665554402, + "grad_norm": 1.6755871782655098, + "learning_rate": 3.890382681145432e-06, + "loss": 1.106, + "step": 284 + }, + { + "epoch": 0.025702304189024666, + "grad_norm": 2.573014992342227, + "learning_rate": 3.892803380784608e-06, + "loss": 1.1024, + "step": 285 + }, + { + "epoch": 0.025792487712494927, + "grad_norm": 1.0238172103375731, + "learning_rate": 3.8952156015985725e-06, + "loss": 0.8502, + "step": 286 + }, + { + "epoch": 0.025882671235965188, + "grad_norm": 1.8604741414053603, + "learning_rate": 3.897619402776516e-06, + "loss": 1.0701, + "step": 287 + }, + { + "epoch": 0.025972854759435452, + "grad_norm": 2.1966732057272447, + "learning_rate": 3.900014842889995e-06, + "loss": 1.1389, + "step": 288 + }, + { + "epoch": 0.026063038282905713, + "grad_norm": 1.6520611316139975, + "learning_rate": 3.902401979901503e-06, + "loss": 0.9843, + "step": 289 + }, + { + "epoch": 0.026153221806375974, + "grad_norm": 1.9812648158918378, + "learning_rate": 3.904780871172884e-06, + "loss": 1.111, + "step": 290 + }, + { + "epoch": 0.026243405329846238, + "grad_norm": 2.4198599030793098, + "learning_rate": 3.907151573473601e-06, + "loss": 0.9484, + "step": 291 + }, + { + "epoch": 0.0263335888533165, + "grad_norm": 1.3679175471336806, + "learning_rate": 3.909514142988868e-06, + "loss": 1.0714, + "step": 292 + }, + { + "epoch": 0.02642377237678676, + "grad_norm": 1.745052564862182, + "learning_rate": 3.911868635327639e-06, + "loss": 1.0276, + "step": 293 + }, + { + "epoch": 0.026513955900257024, + "grad_norm": 1.431513053341536, + "learning_rate": 3.914215105530455e-06, + "loss": 1.0381, + "step": 294 + }, + { + "epoch": 0.026604139423727285, + "grad_norm": 1.6377495408107747, + "learning_rate": 3.916553608077179e-06, + "loss": 1.0547, + "step": 295 + }, + { + "epoch": 0.026694322947197546, + "grad_norm": 1.793878793910849, + "learning_rate": 3.91888419689457e-06, + "loss": 1.0498, + "step": 296 + }, + { + "epoch": 0.02678450647066781, + "grad_norm": 1.573432896596923, + "learning_rate": 3.921206925363754e-06, + "loss": 0.9941, + "step": 297 + }, + { + "epoch": 0.02687468999413807, + "grad_norm": 1.7522539881975023, + "learning_rate": 3.923521846327559e-06, + "loss": 0.977, + "step": 298 + }, + { + "epoch": 0.02696487351760833, + "grad_norm": 1.7382450400020697, + "learning_rate": 3.925829012097725e-06, + "loss": 1.0167, + "step": 299 + }, + { + "epoch": 0.027055057041078596, + "grad_norm": 2.346582455948373, + "learning_rate": 3.928128474462e-06, + "loss": 0.9796, + "step": 300 + }, + { + "epoch": 0.027145240564548857, + "grad_norm": 2.138634194516805, + "learning_rate": 3.930420284691115e-06, + "loss": 1.0694, + "step": 301 + }, + { + "epoch": 0.027235424088019117, + "grad_norm": 2.63339706858833, + "learning_rate": 3.932704493545644e-06, + "loss": 1.0896, + "step": 302 + }, + { + "epoch": 0.02732560761148938, + "grad_norm": 1.9204660849304056, + "learning_rate": 3.934981151282745e-06, + "loss": 1.1412, + "step": 303 + }, + { + "epoch": 0.027415791134959643, + "grad_norm": 1.808036438203428, + "learning_rate": 3.9372503076628006e-06, + "loss": 1.0555, + "step": 304 + }, + { + "epoch": 0.027505974658429903, + "grad_norm": 1.672289716675802, + "learning_rate": 3.939512011955941e-06, + "loss": 1.0807, + "step": 305 + }, + { + "epoch": 0.027596158181900168, + "grad_norm": 1.7001779939929627, + "learning_rate": 3.941766312948463e-06, + "loss": 1.0546, + "step": 306 + }, + { + "epoch": 0.02768634170537043, + "grad_norm": 1.7051919416301184, + "learning_rate": 3.944013258949147e-06, + "loss": 1.027, + "step": 307 + }, + { + "epoch": 0.02777652522884069, + "grad_norm": 1.9372787349242109, + "learning_rate": 3.946252897795465e-06, + "loss": 0.8795, + "step": 308 + }, + { + "epoch": 0.027866708752310954, + "grad_norm": 2.91900461136028, + "learning_rate": 3.9484852768596935e-06, + "loss": 1.0684, + "step": 309 + }, + { + "epoch": 0.027956892275781214, + "grad_norm": 1.7418833924780162, + "learning_rate": 3.950710443054923e-06, + "loss": 0.9572, + "step": 310 + }, + { + "epoch": 0.028047075799251475, + "grad_norm": 1.976296991299902, + "learning_rate": 3.952928442840981e-06, + "loss": 0.9722, + "step": 311 + }, + { + "epoch": 0.02813725932272174, + "grad_norm": 1.9092385675663193, + "learning_rate": 3.955139322230243e-06, + "loss": 1.0748, + "step": 312 + }, + { + "epoch": 0.028227442846192, + "grad_norm": 1.8821855823013693, + "learning_rate": 3.957343126793365e-06, + "loss": 0.993, + "step": 313 + }, + { + "epoch": 0.02831762636966226, + "grad_norm": 1.5836468018206806, + "learning_rate": 3.959539901664921e-06, + "loss": 1.0369, + "step": 314 + }, + { + "epoch": 0.028407809893132525, + "grad_norm": 2.4000723668950084, + "learning_rate": 3.9617296915489425e-06, + "loss": 1.1521, + "step": 315 + }, + { + "epoch": 0.028497993416602786, + "grad_norm": 2.0965283307691145, + "learning_rate": 3.963912540724387e-06, + "loss": 0.9963, + "step": 316 + }, + { + "epoch": 0.028588176940073047, + "grad_norm": 2.3535099520159846, + "learning_rate": 3.966088493050501e-06, + "loss": 1.0595, + "step": 317 + }, + { + "epoch": 0.02867836046354331, + "grad_norm": 1.6287860780293333, + "learning_rate": 3.968257591972113e-06, + "loss": 1.0375, + "step": 318 + }, + { + "epoch": 0.028768543987013572, + "grad_norm": 1.537915804702626, + "learning_rate": 3.970419880524835e-06, + "loss": 1.1336, + "step": 319 + }, + { + "epoch": 0.028858727510483833, + "grad_norm": 1.8871742472694155, + "learning_rate": 3.972575401340192e-06, + "loss": 1.0179, + "step": 320 + }, + { + "epoch": 0.028948911033954097, + "grad_norm": 2.5242603011682845, + "learning_rate": 3.974724196650656e-06, + "loss": 1.0865, + "step": 321 + }, + { + "epoch": 0.029039094557424358, + "grad_norm": 1.9201140360941678, + "learning_rate": 3.976866308294617e-06, + "loss": 1.0387, + "step": 322 + }, + { + "epoch": 0.02912927808089462, + "grad_norm": 1.498102657070341, + "learning_rate": 3.979001777721269e-06, + "loss": 1.005, + "step": 323 + }, + { + "epoch": 0.029219461604364883, + "grad_norm": 4.62421080583295, + "learning_rate": 3.981130645995424e-06, + "loss": 1.0167, + "step": 324 + }, + { + "epoch": 0.029309645127835144, + "grad_norm": 8.494402669345524, + "learning_rate": 3.983252953802248e-06, + "loss": 1.1228, + "step": 325 + }, + { + "epoch": 0.029399828651305405, + "grad_norm": 0.8413408287634376, + "learning_rate": 3.9853687414519285e-06, + "loss": 0.8415, + "step": 326 + }, + { + "epoch": 0.02949001217477567, + "grad_norm": 2.0062558563242017, + "learning_rate": 3.987478048884265e-06, + "loss": 1.0934, + "step": 327 + }, + { + "epoch": 0.02958019569824593, + "grad_norm": 1.9334288863980864, + "learning_rate": 3.989580915673196e-06, + "loss": 1.0249, + "step": 328 + }, + { + "epoch": 0.02967037922171619, + "grad_norm": 2.0736031346111976, + "learning_rate": 3.991677381031255e-06, + "loss": 1.0211, + "step": 329 + }, + { + "epoch": 0.029760562745186455, + "grad_norm": 2.301214258350511, + "learning_rate": 3.993767483813953e-06, + "loss": 1.1202, + "step": 330 + }, + { + "epoch": 0.029850746268656716, + "grad_norm": 1.6566209917404973, + "learning_rate": 3.995851262524104e-06, + "loss": 0.9611, + "step": 331 + }, + { + "epoch": 0.02994092979212698, + "grad_norm": 2.234842744965865, + "learning_rate": 3.997928755316079e-06, + "loss": 1.0402, + "step": 332 + }, + { + "epoch": 0.03003111331559724, + "grad_norm": 1.1287046470275373, + "learning_rate": 4e-06, + "loss": 0.8274, + "step": 333 + }, + { + "epoch": 0.030121296839067502, + "grad_norm": 1.9461711535839694, + "learning_rate": 3.999999914674486e-06, + "loss": 1.0102, + "step": 334 + }, + { + "epoch": 0.030211480362537766, + "grad_norm": 1.597159012174902, + "learning_rate": 3.999999658697952e-06, + "loss": 1.0272, + "step": 335 + }, + { + "epoch": 0.030301663886008027, + "grad_norm": 0.9094151766576742, + "learning_rate": 3.9999992320704185e-06, + "loss": 0.7655, + "step": 336 + }, + { + "epoch": 0.030391847409478288, + "grad_norm": 2.7856121358901804, + "learning_rate": 3.999998634791922e-06, + "loss": 1.0593, + "step": 337 + }, + { + "epoch": 0.030482030932948552, + "grad_norm": 1.7452694442752177, + "learning_rate": 3.999997866862515e-06, + "loss": 1.0258, + "step": 338 + }, + { + "epoch": 0.030572214456418813, + "grad_norm": 1.6769554452726785, + "learning_rate": 3.999996928282262e-06, + "loss": 1.1322, + "step": 339 + }, + { + "epoch": 0.030662397979889074, + "grad_norm": 2.200011522089523, + "learning_rate": 3.999995819051244e-06, + "loss": 1.0353, + "step": 340 + }, + { + "epoch": 0.030752581503359338, + "grad_norm": 1.7759554292954178, + "learning_rate": 3.9999945391695536e-06, + "loss": 1.0027, + "step": 341 + }, + { + "epoch": 0.0308427650268296, + "grad_norm": 1.7491979804771658, + "learning_rate": 3.999993088637302e-06, + "loss": 1.1271, + "step": 342 + }, + { + "epoch": 0.03093294855029986, + "grad_norm": 1.687277531971077, + "learning_rate": 3.999991467454612e-06, + "loss": 1.1065, + "step": 343 + }, + { + "epoch": 0.031023132073770124, + "grad_norm": 1.9227717098706254, + "learning_rate": 3.999989675621622e-06, + "loss": 0.9583, + "step": 344 + }, + { + "epoch": 0.031113315597240385, + "grad_norm": 1.6255925638672692, + "learning_rate": 3.999987713138485e-06, + "loss": 1.0429, + "step": 345 + }, + { + "epoch": 0.031203499120710645, + "grad_norm": 0.9822616598185316, + "learning_rate": 3.999985580005369e-06, + "loss": 0.8746, + "step": 346 + }, + { + "epoch": 0.031293682644180906, + "grad_norm": 2.2978170111020613, + "learning_rate": 3.999983276222455e-06, + "loss": 1.1118, + "step": 347 + }, + { + "epoch": 0.03138386616765117, + "grad_norm": 1.7216960280882236, + "learning_rate": 3.999980801789941e-06, + "loss": 1.0516, + "step": 348 + }, + { + "epoch": 0.031474049691121435, + "grad_norm": 2.012295479185534, + "learning_rate": 3.999978156708036e-06, + "loss": 1.0502, + "step": 349 + }, + { + "epoch": 0.031564233214591696, + "grad_norm": 2.0261323282571864, + "learning_rate": 3.9999753409769675e-06, + "loss": 1.1521, + "step": 350 + }, + { + "epoch": 0.031654416738061956, + "grad_norm": 1.0002308817406709, + "learning_rate": 3.999972354596975e-06, + "loss": 0.8163, + "step": 351 + }, + { + "epoch": 0.03174460026153222, + "grad_norm": 1.7810514238998845, + "learning_rate": 3.999969197568314e-06, + "loss": 0.9938, + "step": 352 + }, + { + "epoch": 0.03183478378500248, + "grad_norm": 2.025732087032695, + "learning_rate": 3.999965869891253e-06, + "loss": 1.0983, + "step": 353 + }, + { + "epoch": 0.03192496730847274, + "grad_norm": 1.6138722812646529, + "learning_rate": 3.999962371566075e-06, + "loss": 1.0682, + "step": 354 + }, + { + "epoch": 0.03201515083194301, + "grad_norm": 1.6781097809227539, + "learning_rate": 3.999958702593082e-06, + "loss": 1.1098, + "step": 355 + }, + { + "epoch": 0.03210533435541327, + "grad_norm": 1.49318178614195, + "learning_rate": 3.999954862972583e-06, + "loss": 1.0672, + "step": 356 + }, + { + "epoch": 0.03219551787888353, + "grad_norm": 1.5963541334773481, + "learning_rate": 3.999950852704908e-06, + "loss": 0.9886, + "step": 357 + }, + { + "epoch": 0.03228570140235379, + "grad_norm": 1.5995952153772721, + "learning_rate": 3.9999466717903995e-06, + "loss": 1.1006, + "step": 358 + }, + { + "epoch": 0.03237588492582405, + "grad_norm": 1.0281254220515257, + "learning_rate": 3.999942320229413e-06, + "loss": 0.8034, + "step": 359 + }, + { + "epoch": 0.03246606844929431, + "grad_norm": 2.110555135007203, + "learning_rate": 3.99993779802232e-06, + "loss": 1.1746, + "step": 360 + }, + { + "epoch": 0.03255625197276458, + "grad_norm": 1.9519703618285373, + "learning_rate": 3.999933105169506e-06, + "loss": 1.0336, + "step": 361 + }, + { + "epoch": 0.03264643549623484, + "grad_norm": 1.549857222226131, + "learning_rate": 3.999928241671373e-06, + "loss": 1.0424, + "step": 362 + }, + { + "epoch": 0.0327366190197051, + "grad_norm": 2.004422424345354, + "learning_rate": 3.999923207528334e-06, + "loss": 0.9412, + "step": 363 + }, + { + "epoch": 0.03282680254317536, + "grad_norm": 1.5153075210133902, + "learning_rate": 3.9999180027408196e-06, + "loss": 1.04, + "step": 364 + }, + { + "epoch": 0.03291698606664562, + "grad_norm": 2.0327170136270847, + "learning_rate": 3.9999126273092735e-06, + "loss": 1.0229, + "step": 365 + }, + { + "epoch": 0.03300716959011588, + "grad_norm": 1.8936384791406218, + "learning_rate": 3.999907081234156e-06, + "loss": 1.1332, + "step": 366 + }, + { + "epoch": 0.03309735311358615, + "grad_norm": 1.8052068846429499, + "learning_rate": 3.999901364515938e-06, + "loss": 1.0114, + "step": 367 + }, + { + "epoch": 0.03318753663705641, + "grad_norm": 1.7830688827139936, + "learning_rate": 3.999895477155108e-06, + "loss": 1.0895, + "step": 368 + }, + { + "epoch": 0.03327772016052667, + "grad_norm": 0.8938087977562154, + "learning_rate": 3.999889419152169e-06, + "loss": 0.782, + "step": 369 + }, + { + "epoch": 0.03336790368399693, + "grad_norm": 1.7130484600706113, + "learning_rate": 3.999883190507638e-06, + "loss": 1.0413, + "step": 370 + }, + { + "epoch": 0.033458087207467194, + "grad_norm": 2.205979358185714, + "learning_rate": 3.999876791222044e-06, + "loss": 1.0285, + "step": 371 + }, + { + "epoch": 0.033548270730937454, + "grad_norm": 2.005394693277105, + "learning_rate": 3.999870221295936e-06, + "loss": 0.9913, + "step": 372 + }, + { + "epoch": 0.03363845425440772, + "grad_norm": 1.7272548754467705, + "learning_rate": 3.999863480729875e-06, + "loss": 1.0079, + "step": 373 + }, + { + "epoch": 0.03372863777787798, + "grad_norm": 2.068754250329022, + "learning_rate": 3.999856569524433e-06, + "loss": 0.9789, + "step": 374 + }, + { + "epoch": 0.033818821301348244, + "grad_norm": 1.5870105138958028, + "learning_rate": 3.999849487680202e-06, + "loss": 0.9813, + "step": 375 + }, + { + "epoch": 0.033909004824818505, + "grad_norm": 2.859944771453491, + "learning_rate": 3.999842235197786e-06, + "loss": 0.9744, + "step": 376 + }, + { + "epoch": 0.033999188348288766, + "grad_norm": 1.9128273852011743, + "learning_rate": 3.999834812077803e-06, + "loss": 0.9743, + "step": 377 + }, + { + "epoch": 0.034089371871759026, + "grad_norm": 1.699600843242886, + "learning_rate": 3.999827218320886e-06, + "loss": 1.0918, + "step": 378 + }, + { + "epoch": 0.034179555395229294, + "grad_norm": 2.05685901840803, + "learning_rate": 3.999819453927685e-06, + "loss": 0.9484, + "step": 379 + }, + { + "epoch": 0.034269738918699555, + "grad_norm": 1.739450126009724, + "learning_rate": 3.999811518898861e-06, + "loss": 1.0431, + "step": 380 + }, + { + "epoch": 0.034359922442169816, + "grad_norm": 2.702675842718736, + "learning_rate": 3.999803413235092e-06, + "loss": 1.0519, + "step": 381 + }, + { + "epoch": 0.03445010596564008, + "grad_norm": 1.7265099564924478, + "learning_rate": 3.999795136937068e-06, + "loss": 1.1518, + "step": 382 + }, + { + "epoch": 0.03454028948911034, + "grad_norm": 2.3970937537047425, + "learning_rate": 3.999786690005496e-06, + "loss": 1.0314, + "step": 383 + }, + { + "epoch": 0.0346304730125806, + "grad_norm": 1.706370699459528, + "learning_rate": 3.999778072441098e-06, + "loss": 1.0225, + "step": 384 + }, + { + "epoch": 0.034720656536050866, + "grad_norm": 2.410579699841582, + "learning_rate": 3.999769284244608e-06, + "loss": 1.068, + "step": 385 + }, + { + "epoch": 0.03481084005952113, + "grad_norm": 1.4561479360807814, + "learning_rate": 3.999760325416775e-06, + "loss": 1.0533, + "step": 386 + }, + { + "epoch": 0.03490102358299139, + "grad_norm": 1.8822613934473418, + "learning_rate": 3.999751195958366e-06, + "loss": 1.1273, + "step": 387 + }, + { + "epoch": 0.03499120710646165, + "grad_norm": 1.0791818166035132, + "learning_rate": 3.999741895870157e-06, + "loss": 0.8062, + "step": 388 + }, + { + "epoch": 0.03508139062993191, + "grad_norm": 2.6066780086507984, + "learning_rate": 3.999732425152944e-06, + "loss": 1.0817, + "step": 389 + }, + { + "epoch": 0.03517157415340217, + "grad_norm": 1.8049212547941504, + "learning_rate": 3.999722783807533e-06, + "loss": 1.1005, + "step": 390 + }, + { + "epoch": 0.03526175767687244, + "grad_norm": 2.0010778383808034, + "learning_rate": 3.999712971834748e-06, + "loss": 0.9718, + "step": 391 + }, + { + "epoch": 0.0353519412003427, + "grad_norm": 2.5813848982099317, + "learning_rate": 3.999702989235427e-06, + "loss": 1.0506, + "step": 392 + }, + { + "epoch": 0.03544212472381296, + "grad_norm": 1.705263877092982, + "learning_rate": 3.999692836010419e-06, + "loss": 1.1258, + "step": 393 + }, + { + "epoch": 0.03553230824728322, + "grad_norm": 2.011618365403479, + "learning_rate": 3.999682512160593e-06, + "loss": 1.1802, + "step": 394 + }, + { + "epoch": 0.03562249177075348, + "grad_norm": 0.9579417665757022, + "learning_rate": 3.99967201768683e-06, + "loss": 0.7841, + "step": 395 + }, + { + "epoch": 0.03571267529422374, + "grad_norm": 1.8503413297896973, + "learning_rate": 3.999661352590023e-06, + "loss": 1.0251, + "step": 396 + }, + { + "epoch": 0.03580285881769401, + "grad_norm": 1.7032280287070893, + "learning_rate": 3.999650516871083e-06, + "loss": 1.0419, + "step": 397 + }, + { + "epoch": 0.03589304234116427, + "grad_norm": 1.6635819578536597, + "learning_rate": 3.9996395105309365e-06, + "loss": 1.0667, + "step": 398 + }, + { + "epoch": 0.03598322586463453, + "grad_norm": 1.5323123749821104, + "learning_rate": 3.99962833357052e-06, + "loss": 1.0629, + "step": 399 + }, + { + "epoch": 0.03607340938810479, + "grad_norm": 1.6789559123991402, + "learning_rate": 3.999616985990789e-06, + "loss": 1.0236, + "step": 400 + }, + { + "epoch": 0.03616359291157505, + "grad_norm": 1.537440185623928, + "learning_rate": 3.9996054677927104e-06, + "loss": 0.9809, + "step": 401 + }, + { + "epoch": 0.03625377643504532, + "grad_norm": 2.54975297890766, + "learning_rate": 3.9995937789772675e-06, + "loss": 0.9837, + "step": 402 + }, + { + "epoch": 0.03634395995851558, + "grad_norm": 0.9097370606749965, + "learning_rate": 3.999581919545458e-06, + "loss": 0.8089, + "step": 403 + }, + { + "epoch": 0.03643414348198584, + "grad_norm": 2.345209099855682, + "learning_rate": 3.9995698894982935e-06, + "loss": 1.0677, + "step": 404 + }, + { + "epoch": 0.0365243270054561, + "grad_norm": 2.053179285269486, + "learning_rate": 3.9995576888368e-06, + "loss": 1.0942, + "step": 405 + }, + { + "epoch": 0.036614510528926364, + "grad_norm": 1.9788328257303875, + "learning_rate": 3.9995453175620194e-06, + "loss": 1.033, + "step": 406 + }, + { + "epoch": 0.036704694052396625, + "grad_norm": 2.152196418718248, + "learning_rate": 3.999532775675007e-06, + "loss": 0.8983, + "step": 407 + }, + { + "epoch": 0.03679487757586689, + "grad_norm": 2.1190865083455876, + "learning_rate": 3.9995200631768326e-06, + "loss": 1.0161, + "step": 408 + }, + { + "epoch": 0.03688506109933715, + "grad_norm": 1.9324737696082706, + "learning_rate": 3.9995071800685815e-06, + "loss": 1.1717, + "step": 409 + }, + { + "epoch": 0.036975244622807414, + "grad_norm": 2.373735894537423, + "learning_rate": 3.999494126351352e-06, + "loss": 1.0114, + "step": 410 + }, + { + "epoch": 0.037065428146277675, + "grad_norm": 2.2160577693047063, + "learning_rate": 3.99948090202626e-06, + "loss": 1.0505, + "step": 411 + }, + { + "epoch": 0.037155611669747936, + "grad_norm": 1.8184369948524604, + "learning_rate": 3.999467507094431e-06, + "loss": 1.0441, + "step": 412 + }, + { + "epoch": 0.0372457951932182, + "grad_norm": 1.5783575660888929, + "learning_rate": 3.999453941557011e-06, + "loss": 0.9462, + "step": 413 + }, + { + "epoch": 0.037335978716688464, + "grad_norm": 0.9250047864016855, + "learning_rate": 3.999440205415154e-06, + "loss": 0.7613, + "step": 414 + }, + { + "epoch": 0.037426162240158725, + "grad_norm": 2.0132348134817977, + "learning_rate": 3.999426298670035e-06, + "loss": 1.088, + "step": 415 + }, + { + "epoch": 0.037516345763628986, + "grad_norm": 1.5978876146894239, + "learning_rate": 3.9994122213228385e-06, + "loss": 1.0824, + "step": 416 + }, + { + "epoch": 0.03760652928709925, + "grad_norm": 2.148816106554965, + "learning_rate": 3.9993979733747675e-06, + "loss": 1.073, + "step": 417 + }, + { + "epoch": 0.03769671281056951, + "grad_norm": 1.715908198095155, + "learning_rate": 3.999383554827037e-06, + "loss": 1.0394, + "step": 418 + }, + { + "epoch": 0.03778689633403977, + "grad_norm": 1.7364827313584643, + "learning_rate": 3.999368965680876e-06, + "loss": 0.9996, + "step": 419 + }, + { + "epoch": 0.037877079857510036, + "grad_norm": 1.740499251530154, + "learning_rate": 3.999354205937531e-06, + "loss": 1.0848, + "step": 420 + }, + { + "epoch": 0.0379672633809803, + "grad_norm": 1.7744370387116604, + "learning_rate": 3.999339275598261e-06, + "loss": 1.0876, + "step": 421 + }, + { + "epoch": 0.03805744690445056, + "grad_norm": 1.6058075262007596, + "learning_rate": 3.99932417466434e-06, + "loss": 1.0198, + "step": 422 + }, + { + "epoch": 0.03814763042792082, + "grad_norm": 2.7392446697824195, + "learning_rate": 3.999308903137056e-06, + "loss": 0.9516, + "step": 423 + }, + { + "epoch": 0.03823781395139108, + "grad_norm": 1.8205423496499324, + "learning_rate": 3.999293461017711e-06, + "loss": 1.0916, + "step": 424 + }, + { + "epoch": 0.03832799747486134, + "grad_norm": 1.8916913012722456, + "learning_rate": 3.9992778483076255e-06, + "loss": 1.0332, + "step": 425 + }, + { + "epoch": 0.03841818099833161, + "grad_norm": 1.6295333131028538, + "learning_rate": 3.99926206500813e-06, + "loss": 1.0937, + "step": 426 + }, + { + "epoch": 0.03850836452180187, + "grad_norm": 2.0019426447896413, + "learning_rate": 3.999246111120571e-06, + "loss": 0.9702, + "step": 427 + }, + { + "epoch": 0.03859854804527213, + "grad_norm": 1.6032753321783144, + "learning_rate": 3.999229986646311e-06, + "loss": 1.1095, + "step": 428 + }, + { + "epoch": 0.03868873156874239, + "grad_norm": 1.6572552635211717, + "learning_rate": 3.999213691586723e-06, + "loss": 1.0065, + "step": 429 + }, + { + "epoch": 0.03877891509221265, + "grad_norm": 1.785949130283243, + "learning_rate": 3.9991972259432e-06, + "loss": 1.0828, + "step": 430 + }, + { + "epoch": 0.03886909861568291, + "grad_norm": 1.499480952422693, + "learning_rate": 3.999180589717147e-06, + "loss": 1.123, + "step": 431 + }, + { + "epoch": 0.03895928213915318, + "grad_norm": 1.8028503579039457, + "learning_rate": 3.999163782909983e-06, + "loss": 0.9504, + "step": 432 + }, + { + "epoch": 0.03904946566262344, + "grad_norm": 1.8474455182994483, + "learning_rate": 3.99914680552314e-06, + "loss": 1.0699, + "step": 433 + }, + { + "epoch": 0.0391396491860937, + "grad_norm": 1.684692166109156, + "learning_rate": 3.999129657558069e-06, + "loss": 0.9719, + "step": 434 + }, + { + "epoch": 0.03922983270956396, + "grad_norm": 0.8831148895635226, + "learning_rate": 3.999112339016234e-06, + "loss": 0.7744, + "step": 435 + }, + { + "epoch": 0.03932001623303422, + "grad_norm": 1.7287494048740248, + "learning_rate": 3.999094849899109e-06, + "loss": 1.0553, + "step": 436 + }, + { + "epoch": 0.039410199756504484, + "grad_norm": 1.7996828780037901, + "learning_rate": 3.99907719020819e-06, + "loss": 0.9587, + "step": 437 + }, + { + "epoch": 0.03950038327997475, + "grad_norm": 1.6259990702036395, + "learning_rate": 3.999059359944982e-06, + "loss": 1.0669, + "step": 438 + }, + { + "epoch": 0.03959056680344501, + "grad_norm": 1.592202576031498, + "learning_rate": 3.999041359111007e-06, + "loss": 1.098, + "step": 439 + }, + { + "epoch": 0.03968075032691527, + "grad_norm": 1.6926799673980428, + "learning_rate": 3.999023187707801e-06, + "loss": 1.1151, + "step": 440 + }, + { + "epoch": 0.039770933850385534, + "grad_norm": 1.6330750176788265, + "learning_rate": 3.999004845736913e-06, + "loss": 1.103, + "step": 441 + }, + { + "epoch": 0.039861117373855795, + "grad_norm": 1.6843885063804716, + "learning_rate": 3.9989863331999096e-06, + "loss": 1.1148, + "step": 442 + }, + { + "epoch": 0.039951300897326056, + "grad_norm": 1.8867101570391793, + "learning_rate": 3.99896765009837e-06, + "loss": 0.9739, + "step": 443 + }, + { + "epoch": 0.040041484420796324, + "grad_norm": 1.4914134949098028, + "learning_rate": 3.998948796433888e-06, + "loss": 1.0075, + "step": 444 + }, + { + "epoch": 0.040131667944266584, + "grad_norm": 3.5936273968556676, + "learning_rate": 3.998929772208073e-06, + "loss": 1.0805, + "step": 445 + }, + { + "epoch": 0.040221851467736845, + "grad_norm": 0.8758375723131595, + "learning_rate": 3.998910577422547e-06, + "loss": 0.8011, + "step": 446 + }, + { + "epoch": 0.040312034991207106, + "grad_norm": 1.5039618940091506, + "learning_rate": 3.99889121207895e-06, + "loss": 1.0815, + "step": 447 + }, + { + "epoch": 0.04040221851467737, + "grad_norm": 1.9683326399907186, + "learning_rate": 3.9988716761789324e-06, + "loss": 1.0628, + "step": 448 + }, + { + "epoch": 0.04049240203814763, + "grad_norm": 1.6912461837195443, + "learning_rate": 3.998851969724161e-06, + "loss": 0.9796, + "step": 449 + }, + { + "epoch": 0.040582585561617895, + "grad_norm": 1.8044159259653976, + "learning_rate": 3.998832092716319e-06, + "loss": 1.0965, + "step": 450 + }, + { + "epoch": 0.040672769085088156, + "grad_norm": 1.452050663030041, + "learning_rate": 3.998812045157102e-06, + "loss": 1.0224, + "step": 451 + }, + { + "epoch": 0.04076295260855842, + "grad_norm": 2.838452613670214, + "learning_rate": 3.998791827048219e-06, + "loss": 1.0865, + "step": 452 + }, + { + "epoch": 0.04085313613202868, + "grad_norm": 1.7364955001847442, + "learning_rate": 3.998771438391396e-06, + "loss": 0.9854, + "step": 453 + }, + { + "epoch": 0.04094331965549894, + "grad_norm": 1.0358071062466279, + "learning_rate": 3.9987508791883725e-06, + "loss": 0.7798, + "step": 454 + }, + { + "epoch": 0.0410335031789692, + "grad_norm": 1.8529063309670812, + "learning_rate": 3.998730149440904e-06, + "loss": 1.1525, + "step": 455 + }, + { + "epoch": 0.04112368670243947, + "grad_norm": 2.180066803774806, + "learning_rate": 3.998709249150758e-06, + "loss": 1.0935, + "step": 456 + }, + { + "epoch": 0.04121387022590973, + "grad_norm": 1.762830497651177, + "learning_rate": 3.998688178319717e-06, + "loss": 1.0655, + "step": 457 + }, + { + "epoch": 0.04130405374937999, + "grad_norm": 1.7379834552236404, + "learning_rate": 3.9986669369495805e-06, + "loss": 1.0313, + "step": 458 + }, + { + "epoch": 0.04139423727285025, + "grad_norm": 1.9148889742506392, + "learning_rate": 3.998645525042161e-06, + "loss": 1.0946, + "step": 459 + }, + { + "epoch": 0.04148442079632051, + "grad_norm": 1.9094298059867187, + "learning_rate": 3.998623942599284e-06, + "loss": 1.1081, + "step": 460 + }, + { + "epoch": 0.04157460431979077, + "grad_norm": 1.9643892099932774, + "learning_rate": 3.998602189622793e-06, + "loss": 1.0357, + "step": 461 + }, + { + "epoch": 0.04166478784326104, + "grad_norm": 2.283844191305776, + "learning_rate": 3.998580266114542e-06, + "loss": 1.0772, + "step": 462 + }, + { + "epoch": 0.0417549713667313, + "grad_norm": 1.5586195252851864, + "learning_rate": 3.998558172076404e-06, + "loss": 0.9451, + "step": 463 + }, + { + "epoch": 0.04184515489020156, + "grad_norm": 2.171939327467438, + "learning_rate": 3.998535907510262e-06, + "loss": 0.977, + "step": 464 + }, + { + "epoch": 0.04193533841367182, + "grad_norm": 1.6363265630587556, + "learning_rate": 3.998513472418016e-06, + "loss": 1.0547, + "step": 465 + }, + { + "epoch": 0.04202552193714208, + "grad_norm": 1.6691230871206313, + "learning_rate": 3.998490866801582e-06, + "loss": 1.0733, + "step": 466 + }, + { + "epoch": 0.04211570546061234, + "grad_norm": 1.5023714234594951, + "learning_rate": 3.998468090662886e-06, + "loss": 1.0309, + "step": 467 + }, + { + "epoch": 0.04220588898408261, + "grad_norm": 1.5395517104243153, + "learning_rate": 3.998445144003874e-06, + "loss": 0.9344, + "step": 468 + }, + { + "epoch": 0.04229607250755287, + "grad_norm": 2.137068083553703, + "learning_rate": 3.998422026826504e-06, + "loss": 1.0657, + "step": 469 + }, + { + "epoch": 0.04238625603102313, + "grad_norm": 2.1037990765165304, + "learning_rate": 3.998398739132746e-06, + "loss": 1.0719, + "step": 470 + }, + { + "epoch": 0.04247643955449339, + "grad_norm": 1.5776631652069943, + "learning_rate": 3.99837528092459e-06, + "loss": 0.9943, + "step": 471 + }, + { + "epoch": 0.042566623077963654, + "grad_norm": 1.8432718723851138, + "learning_rate": 3.998351652204034e-06, + "loss": 1.088, + "step": 472 + }, + { + "epoch": 0.042656806601433915, + "grad_norm": 1.8727871234408764, + "learning_rate": 3.998327852973098e-06, + "loss": 1.0834, + "step": 473 + }, + { + "epoch": 0.04274699012490418, + "grad_norm": 1.9284080557274519, + "learning_rate": 3.99830388323381e-06, + "loss": 0.9868, + "step": 474 + }, + { + "epoch": 0.042837173648374444, + "grad_norm": 1.6488818975273543, + "learning_rate": 3.998279742988216e-06, + "loss": 1.0752, + "step": 475 + }, + { + "epoch": 0.042927357171844704, + "grad_norm": 1.5454991705209462, + "learning_rate": 3.998255432238377e-06, + "loss": 1.0086, + "step": 476 + }, + { + "epoch": 0.043017540695314965, + "grad_norm": 2.04852845479594, + "learning_rate": 3.9982309509863656e-06, + "loss": 1.1171, + "step": 477 + }, + { + "epoch": 0.043107724218785226, + "grad_norm": 2.0109722048499825, + "learning_rate": 3.998206299234272e-06, + "loss": 0.9895, + "step": 478 + }, + { + "epoch": 0.04319790774225549, + "grad_norm": 1.6071796443301445, + "learning_rate": 3.998181476984198e-06, + "loss": 1.0568, + "step": 479 + }, + { + "epoch": 0.043288091265725755, + "grad_norm": 1.7222428636783746, + "learning_rate": 3.998156484238263e-06, + "loss": 0.9692, + "step": 480 + }, + { + "epoch": 0.043378274789196015, + "grad_norm": 1.9010290972847166, + "learning_rate": 3.998131320998599e-06, + "loss": 0.9808, + "step": 481 + }, + { + "epoch": 0.043468458312666276, + "grad_norm": 1.8061304310224542, + "learning_rate": 3.998105987267353e-06, + "loss": 1.084, + "step": 482 + }, + { + "epoch": 0.04355864183613654, + "grad_norm": 1.7037141113531569, + "learning_rate": 3.998080483046687e-06, + "loss": 0.9682, + "step": 483 + }, + { + "epoch": 0.0436488253596068, + "grad_norm": 1.5098075515432898, + "learning_rate": 3.998054808338776e-06, + "loss": 0.988, + "step": 484 + }, + { + "epoch": 0.04373900888307706, + "grad_norm": 1.7120883502578788, + "learning_rate": 3.998028963145812e-06, + "loss": 0.9993, + "step": 485 + }, + { + "epoch": 0.043829192406547327, + "grad_norm": 1.7022562173412463, + "learning_rate": 3.99800294747e-06, + "loss": 1.0565, + "step": 486 + }, + { + "epoch": 0.04391937593001759, + "grad_norm": 2.3334187991520787, + "learning_rate": 3.99797676131356e-06, + "loss": 1.0742, + "step": 487 + }, + { + "epoch": 0.04400955945348785, + "grad_norm": 1.9676714167877114, + "learning_rate": 3.997950404678726e-06, + "loss": 0.8992, + "step": 488 + }, + { + "epoch": 0.04409974297695811, + "grad_norm": 1.7849529659795738, + "learning_rate": 3.997923877567746e-06, + "loss": 1.1552, + "step": 489 + }, + { + "epoch": 0.04418992650042837, + "grad_norm": 1.5381742906171036, + "learning_rate": 3.9978971799828855e-06, + "loss": 0.8696, + "step": 490 + }, + { + "epoch": 0.04428011002389863, + "grad_norm": 1.9472822152275662, + "learning_rate": 3.997870311926421e-06, + "loss": 1.032, + "step": 491 + }, + { + "epoch": 0.0443702935473689, + "grad_norm": 1.7119328638710138, + "learning_rate": 3.997843273400645e-06, + "loss": 0.9982, + "step": 492 + }, + { + "epoch": 0.04446047707083916, + "grad_norm": 2.0870496663872284, + "learning_rate": 3.997816064407865e-06, + "loss": 1.0216, + "step": 493 + }, + { + "epoch": 0.04455066059430942, + "grad_norm": 1.5343608008696923, + "learning_rate": 3.997788684950402e-06, + "loss": 1.0608, + "step": 494 + }, + { + "epoch": 0.04464084411777968, + "grad_norm": 1.9176254431364288, + "learning_rate": 3.997761135030593e-06, + "loss": 1.079, + "step": 495 + }, + { + "epoch": 0.04473102764124994, + "grad_norm": 1.6190191391301936, + "learning_rate": 3.997733414650789e-06, + "loss": 0.9364, + "step": 496 + }, + { + "epoch": 0.0448212111647202, + "grad_norm": 2.4475226118935, + "learning_rate": 3.9977055238133554e-06, + "loss": 1.0016, + "step": 497 + }, + { + "epoch": 0.04491139468819047, + "grad_norm": 1.4775747230598328, + "learning_rate": 3.99767746252067e-06, + "loss": 0.9183, + "step": 498 + }, + { + "epoch": 0.04500157821166073, + "grad_norm": 1.6186271561954333, + "learning_rate": 3.997649230775129e-06, + "loss": 1.0376, + "step": 499 + }, + { + "epoch": 0.04509176173513099, + "grad_norm": 1.4996173688512031, + "learning_rate": 3.9976208285791395e-06, + "loss": 0.9759, + "step": 500 + }, + { + "epoch": 0.04518194525860125, + "grad_norm": 0.8410160488157715, + "learning_rate": 3.997592255935127e-06, + "loss": 0.8753, + "step": 501 + }, + { + "epoch": 0.045272128782071513, + "grad_norm": 3.146878066909038, + "learning_rate": 3.997563512845529e-06, + "loss": 1.069, + "step": 502 + }, + { + "epoch": 0.045362312305541774, + "grad_norm": 1.8502810264762124, + "learning_rate": 3.9975345993127975e-06, + "loss": 0.9981, + "step": 503 + }, + { + "epoch": 0.04545249582901204, + "grad_norm": 1.7363912873424552, + "learning_rate": 3.9975055153393985e-06, + "loss": 1.0513, + "step": 504 + }, + { + "epoch": 0.0455426793524823, + "grad_norm": 1.5752925570814198, + "learning_rate": 3.997476260927816e-06, + "loss": 1.0587, + "step": 505 + }, + { + "epoch": 0.045632862875952564, + "grad_norm": 3.755256402091178, + "learning_rate": 3.997446836080545e-06, + "loss": 1.0145, + "step": 506 + }, + { + "epoch": 0.045723046399422825, + "grad_norm": 2.165439869727797, + "learning_rate": 3.997417240800095e-06, + "loss": 1.0916, + "step": 507 + }, + { + "epoch": 0.045813229922893085, + "grad_norm": 1.6108023230654096, + "learning_rate": 3.997387475088994e-06, + "loss": 1.1391, + "step": 508 + }, + { + "epoch": 0.045903413446363346, + "grad_norm": 2.191306589617744, + "learning_rate": 3.99735753894978e-06, + "loss": 1.1409, + "step": 509 + }, + { + "epoch": 0.045993596969833614, + "grad_norm": 1.673408660497694, + "learning_rate": 3.997327432385006e-06, + "loss": 1.0384, + "step": 510 + }, + { + "epoch": 0.046083780493303875, + "grad_norm": 2.026984445708884, + "learning_rate": 3.997297155397244e-06, + "loss": 0.9725, + "step": 511 + }, + { + "epoch": 0.046173964016774136, + "grad_norm": 2.3440615637639675, + "learning_rate": 3.997266707989074e-06, + "loss": 0.953, + "step": 512 + }, + { + "epoch": 0.046264147540244396, + "grad_norm": 2.0085110766423764, + "learning_rate": 3.997236090163097e-06, + "loss": 0.9567, + "step": 513 + }, + { + "epoch": 0.04635433106371466, + "grad_norm": 1.5094582070740477, + "learning_rate": 3.9972053019219235e-06, + "loss": 1.1683, + "step": 514 + }, + { + "epoch": 0.04644451458718492, + "grad_norm": 3.1781900614902763, + "learning_rate": 3.997174343268181e-06, + "loss": 1.0515, + "step": 515 + }, + { + "epoch": 0.046534698110655186, + "grad_norm": 1.9248424663736055, + "learning_rate": 3.9971432142045115e-06, + "loss": 1.042, + "step": 516 + }, + { + "epoch": 0.04662488163412545, + "grad_norm": 0.8592987026677593, + "learning_rate": 3.99711191473357e-06, + "loss": 0.867, + "step": 517 + }, + { + "epoch": 0.04671506515759571, + "grad_norm": 1.6838725626969413, + "learning_rate": 3.99708044485803e-06, + "loss": 1.0561, + "step": 518 + }, + { + "epoch": 0.04680524868106597, + "grad_norm": 1.7415723869444202, + "learning_rate": 3.997048804580574e-06, + "loss": 1.0482, + "step": 519 + }, + { + "epoch": 0.04689543220453623, + "grad_norm": 1.9271496598110418, + "learning_rate": 3.997016993903901e-06, + "loss": 0.9912, + "step": 520 + }, + { + "epoch": 0.04698561572800649, + "grad_norm": 1.5853691398056928, + "learning_rate": 3.996985012830728e-06, + "loss": 1.1155, + "step": 521 + }, + { + "epoch": 0.04707579925147676, + "grad_norm": 1.489849075763947, + "learning_rate": 3.996952861363782e-06, + "loss": 0.9778, + "step": 522 + }, + { + "epoch": 0.04716598277494702, + "grad_norm": 1.7463734379300555, + "learning_rate": 3.9969205395058064e-06, + "loss": 1.0442, + "step": 523 + }, + { + "epoch": 0.04725616629841728, + "grad_norm": 1.9231585426616293, + "learning_rate": 3.99688804725956e-06, + "loss": 1.0343, + "step": 524 + }, + { + "epoch": 0.04734634982188754, + "grad_norm": 1.2329244661917074, + "learning_rate": 3.996855384627815e-06, + "loss": 0.989, + "step": 525 + }, + { + "epoch": 0.0474365333453578, + "grad_norm": 1.739087824195809, + "learning_rate": 3.996822551613357e-06, + "loss": 1.0723, + "step": 526 + }, + { + "epoch": 0.04752671686882806, + "grad_norm": 1.689570816438958, + "learning_rate": 3.996789548218989e-06, + "loss": 0.9672, + "step": 527 + }, + { + "epoch": 0.04761690039229833, + "grad_norm": 1.6650132879236736, + "learning_rate": 3.996756374447526e-06, + "loss": 0.9243, + "step": 528 + }, + { + "epoch": 0.04770708391576859, + "grad_norm": 2.557207830003995, + "learning_rate": 3.9967230303018005e-06, + "loss": 1.0432, + "step": 529 + }, + { + "epoch": 0.04779726743923885, + "grad_norm": 1.6888259340914298, + "learning_rate": 3.996689515784655e-06, + "loss": 1.0162, + "step": 530 + }, + { + "epoch": 0.04788745096270911, + "grad_norm": 2.2747922446158793, + "learning_rate": 3.996655830898951e-06, + "loss": 1.0123, + "step": 531 + }, + { + "epoch": 0.04797763448617937, + "grad_norm": 1.8798501544444102, + "learning_rate": 3.996621975647562e-06, + "loss": 1.0388, + "step": 532 + }, + { + "epoch": 0.04806781800964964, + "grad_norm": 1.4339811960474502, + "learning_rate": 3.996587950033377e-06, + "loss": 0.9687, + "step": 533 + }, + { + "epoch": 0.0481580015331199, + "grad_norm": 2.2719146767028406, + "learning_rate": 3.996553754059299e-06, + "loss": 0.9688, + "step": 534 + }, + { + "epoch": 0.04824818505659016, + "grad_norm": 2.0809762081562475, + "learning_rate": 3.996519387728245e-06, + "loss": 1.0384, + "step": 535 + }, + { + "epoch": 0.04833836858006042, + "grad_norm": 1.525953748164562, + "learning_rate": 3.9964848510431495e-06, + "loss": 1.0157, + "step": 536 + }, + { + "epoch": 0.048428552103530684, + "grad_norm": 1.8955675574610054, + "learning_rate": 3.996450144006957e-06, + "loss": 1.0925, + "step": 537 + }, + { + "epoch": 0.048518735627000945, + "grad_norm": 2.0387696973081852, + "learning_rate": 3.99641526662263e-06, + "loss": 1.0149, + "step": 538 + }, + { + "epoch": 0.04860891915047121, + "grad_norm": 1.8065999784278932, + "learning_rate": 3.996380218893145e-06, + "loss": 1.121, + "step": 539 + }, + { + "epoch": 0.04869910267394147, + "grad_norm": 1.8649697001146108, + "learning_rate": 3.996345000821491e-06, + "loss": 1.0611, + "step": 540 + }, + { + "epoch": 0.048789286197411734, + "grad_norm": 1.9284290117328267, + "learning_rate": 3.996309612410674e-06, + "loss": 1.1441, + "step": 541 + }, + { + "epoch": 0.048879469720881995, + "grad_norm": 1.9898541118110906, + "learning_rate": 3.996274053663713e-06, + "loss": 0.9713, + "step": 542 + }, + { + "epoch": 0.048969653244352256, + "grad_norm": 0.8419116798232296, + "learning_rate": 3.996238324583643e-06, + "loss": 0.8232, + "step": 543 + }, + { + "epoch": 0.049059836767822516, + "grad_norm": 1.870047449948095, + "learning_rate": 3.996202425173512e-06, + "loss": 1.1039, + "step": 544 + }, + { + "epoch": 0.049150020291292784, + "grad_norm": 2.240763461252041, + "learning_rate": 3.996166355436383e-06, + "loss": 0.9952, + "step": 545 + }, + { + "epoch": 0.049240203814763045, + "grad_norm": 1.5640442655645606, + "learning_rate": 3.996130115375333e-06, + "loss": 1.0443, + "step": 546 + }, + { + "epoch": 0.049330387338233306, + "grad_norm": 0.9595922963337635, + "learning_rate": 3.996093704993456e-06, + "loss": 0.8376, + "step": 547 + }, + { + "epoch": 0.04942057086170357, + "grad_norm": 1.6287608276147814, + "learning_rate": 3.996057124293857e-06, + "loss": 1.1189, + "step": 548 + }, + { + "epoch": 0.04951075438517383, + "grad_norm": 1.7871049724587165, + "learning_rate": 3.996020373279659e-06, + "loss": 1.0874, + "step": 549 + }, + { + "epoch": 0.04960093790864409, + "grad_norm": 1.9870787695988743, + "learning_rate": 3.995983451953996e-06, + "loss": 1.0482, + "step": 550 + }, + { + "epoch": 0.049691121432114356, + "grad_norm": 1.720570311253983, + "learning_rate": 3.99594636032002e-06, + "loss": 1.0565, + "step": 551 + }, + { + "epoch": 0.04978130495558462, + "grad_norm": 2.3901894365834537, + "learning_rate": 3.995909098380894e-06, + "loss": 0.9238, + "step": 552 + }, + { + "epoch": 0.04987148847905488, + "grad_norm": 1.995079424813939, + "learning_rate": 3.995871666139799e-06, + "loss": 0.9157, + "step": 553 + }, + { + "epoch": 0.04996167200252514, + "grad_norm": 1.9265639627641382, + "learning_rate": 3.995834063599928e-06, + "loss": 1.0115, + "step": 554 + }, + { + "epoch": 0.0500518555259954, + "grad_norm": 1.5302349540969158, + "learning_rate": 3.99579629076449e-06, + "loss": 1.0617, + "step": 555 + }, + { + "epoch": 0.05014203904946566, + "grad_norm": 1.8151204470137396, + "learning_rate": 3.9957583476367084e-06, + "loss": 1.0073, + "step": 556 + }, + { + "epoch": 0.05023222257293593, + "grad_norm": 1.7483042266478666, + "learning_rate": 3.995720234219819e-06, + "loss": 1.037, + "step": 557 + }, + { + "epoch": 0.05032240609640619, + "grad_norm": 1.7993264103563475, + "learning_rate": 3.995681950517075e-06, + "loss": 0.9391, + "step": 558 + }, + { + "epoch": 0.05041258961987645, + "grad_norm": 1.9592255190056913, + "learning_rate": 3.995643496531743e-06, + "loss": 1.0561, + "step": 559 + }, + { + "epoch": 0.05050277314334671, + "grad_norm": 2.276474016154814, + "learning_rate": 3.9956048722671044e-06, + "loss": 1.047, + "step": 560 + }, + { + "epoch": 0.05059295666681697, + "grad_norm": 1.714249840429153, + "learning_rate": 3.995566077726454e-06, + "loss": 1.0129, + "step": 561 + }, + { + "epoch": 0.05068314019028723, + "grad_norm": 1.6795828143476805, + "learning_rate": 3.995527112913103e-06, + "loss": 0.9953, + "step": 562 + }, + { + "epoch": 0.0507733237137575, + "grad_norm": 1.7333197605384152, + "learning_rate": 3.995487977830375e-06, + "loss": 0.9911, + "step": 563 + }, + { + "epoch": 0.05086350723722776, + "grad_norm": 1.7072906646198018, + "learning_rate": 3.9954486724816105e-06, + "loss": 1.1292, + "step": 564 + }, + { + "epoch": 0.05095369076069802, + "grad_norm": 1.8708680719606834, + "learning_rate": 3.995409196870161e-06, + "loss": 1.0751, + "step": 565 + }, + { + "epoch": 0.05104387428416828, + "grad_norm": 0.9041091361545491, + "learning_rate": 3.995369550999398e-06, + "loss": 0.7833, + "step": 566 + }, + { + "epoch": 0.05113405780763854, + "grad_norm": 2.226551042493821, + "learning_rate": 3.995329734872702e-06, + "loss": 1.0565, + "step": 567 + }, + { + "epoch": 0.051224241331108804, + "grad_norm": 1.6270417810947506, + "learning_rate": 3.9952897484934706e-06, + "loss": 1.1491, + "step": 568 + }, + { + "epoch": 0.05131442485457907, + "grad_norm": 2.046308948783954, + "learning_rate": 3.995249591865115e-06, + "loss": 1.039, + "step": 569 + }, + { + "epoch": 0.05140460837804933, + "grad_norm": 1.8024304961455986, + "learning_rate": 3.995209264991063e-06, + "loss": 0.9945, + "step": 570 + }, + { + "epoch": 0.05149479190151959, + "grad_norm": 1.9454615279261178, + "learning_rate": 3.995168767874756e-06, + "loss": 1.083, + "step": 571 + }, + { + "epoch": 0.051584975424989854, + "grad_norm": 2.166581873579463, + "learning_rate": 3.995128100519648e-06, + "loss": 1.0211, + "step": 572 + }, + { + "epoch": 0.051675158948460115, + "grad_norm": 1.829549902507215, + "learning_rate": 3.995087262929209e-06, + "loss": 1.0504, + "step": 573 + }, + { + "epoch": 0.051765342471930376, + "grad_norm": 1.6545709700431916, + "learning_rate": 3.995046255106925e-06, + "loss": 1.0454, + "step": 574 + }, + { + "epoch": 0.05185552599540064, + "grad_norm": 1.8782452156069935, + "learning_rate": 3.995005077056293e-06, + "loss": 1.1008, + "step": 575 + }, + { + "epoch": 0.051945709518870904, + "grad_norm": 2.1028490630877212, + "learning_rate": 3.9949637287808284e-06, + "loss": 0.9288, + "step": 576 + }, + { + "epoch": 0.052035893042341165, + "grad_norm": 1.5069864016237688, + "learning_rate": 3.994922210284057e-06, + "loss": 0.9788, + "step": 577 + }, + { + "epoch": 0.052126076565811426, + "grad_norm": 2.613409977181496, + "learning_rate": 3.994880521569524e-06, + "loss": 1.066, + "step": 578 + }, + { + "epoch": 0.05221626008928169, + "grad_norm": 1.6736492831501144, + "learning_rate": 3.994838662640785e-06, + "loss": 1.0448, + "step": 579 + }, + { + "epoch": 0.05230644361275195, + "grad_norm": 2.3724515191642825, + "learning_rate": 3.9947966335014116e-06, + "loss": 1.0695, + "step": 580 + }, + { + "epoch": 0.052396627136222215, + "grad_norm": 2.245364181783857, + "learning_rate": 3.99475443415499e-06, + "loss": 1.0171, + "step": 581 + }, + { + "epoch": 0.052486810659692476, + "grad_norm": 1.7596330497499582, + "learning_rate": 3.994712064605121e-06, + "loss": 1.0331, + "step": 582 + }, + { + "epoch": 0.05257699418316274, + "grad_norm": 1.9045217215014636, + "learning_rate": 3.99466952485542e-06, + "loss": 1.1835, + "step": 583 + }, + { + "epoch": 0.052667177706633, + "grad_norm": 1.9500608288007195, + "learning_rate": 3.994626814909518e-06, + "loss": 1.0526, + "step": 584 + }, + { + "epoch": 0.05275736123010326, + "grad_norm": 2.613551834307633, + "learning_rate": 3.994583934771056e-06, + "loss": 1.0137, + "step": 585 + }, + { + "epoch": 0.05284754475357352, + "grad_norm": 1.4680468722715911, + "learning_rate": 3.9945408844436955e-06, + "loss": 1.0966, + "step": 586 + }, + { + "epoch": 0.05293772827704379, + "grad_norm": 1.9312450594051829, + "learning_rate": 3.994497663931109e-06, + "loss": 1.0588, + "step": 587 + }, + { + "epoch": 0.05302791180051405, + "grad_norm": 2.0905177660526792, + "learning_rate": 3.994454273236984e-06, + "loss": 1.0522, + "step": 588 + }, + { + "epoch": 0.05311809532398431, + "grad_norm": 1.665463569648088, + "learning_rate": 3.994410712365023e-06, + "loss": 1.0431, + "step": 589 + }, + { + "epoch": 0.05320827884745457, + "grad_norm": 2.3048119462892345, + "learning_rate": 3.994366981318943e-06, + "loss": 1.1383, + "step": 590 + }, + { + "epoch": 0.05329846237092483, + "grad_norm": 1.8041113391994146, + "learning_rate": 3.9943230801024765e-06, + "loss": 1.0375, + "step": 591 + }, + { + "epoch": 0.05338864589439509, + "grad_norm": 1.7977758388537648, + "learning_rate": 3.9942790087193666e-06, + "loss": 1.078, + "step": 592 + }, + { + "epoch": 0.05347882941786536, + "grad_norm": 1.7221522556298687, + "learning_rate": 3.994234767173376e-06, + "loss": 1.0306, + "step": 593 + }, + { + "epoch": 0.05356901294133562, + "grad_norm": 1.9364614471963186, + "learning_rate": 3.994190355468279e-06, + "loss": 0.9257, + "step": 594 + }, + { + "epoch": 0.05365919646480588, + "grad_norm": 2.4502835674059162, + "learning_rate": 3.994145773607865e-06, + "loss": 1.039, + "step": 595 + }, + { + "epoch": 0.05374937998827614, + "grad_norm": 1.651503689818688, + "learning_rate": 3.994101021595938e-06, + "loss": 1.0065, + "step": 596 + }, + { + "epoch": 0.0538395635117464, + "grad_norm": 1.656847935832904, + "learning_rate": 3.9940560994363165e-06, + "loss": 1.0512, + "step": 597 + }, + { + "epoch": 0.05392974703521666, + "grad_norm": 2.0281977095303336, + "learning_rate": 3.994011007132833e-06, + "loss": 0.9116, + "step": 598 + }, + { + "epoch": 0.05401993055868693, + "grad_norm": 0.8649042675206331, + "learning_rate": 3.993965744689337e-06, + "loss": 0.8529, + "step": 599 + }, + { + "epoch": 0.05411011408215719, + "grad_norm": 1.8138265687960722, + "learning_rate": 3.993920312109687e-06, + "loss": 1.0204, + "step": 600 + }, + { + "epoch": 0.05420029760562745, + "grad_norm": 1.6614376373755904, + "learning_rate": 3.993874709397764e-06, + "loss": 1.0314, + "step": 601 + }, + { + "epoch": 0.05429048112909771, + "grad_norm": 2.2609672286076794, + "learning_rate": 3.993828936557454e-06, + "loss": 1.02, + "step": 602 + }, + { + "epoch": 0.054380664652567974, + "grad_norm": 1.6141365499580265, + "learning_rate": 3.993782993592667e-06, + "loss": 0.9965, + "step": 603 + }, + { + "epoch": 0.054470848176038235, + "grad_norm": 1.8446070085581179, + "learning_rate": 3.993736880507321e-06, + "loss": 1.0718, + "step": 604 + }, + { + "epoch": 0.0545610316995085, + "grad_norm": 1.9834812584643806, + "learning_rate": 3.99369059730535e-06, + "loss": 0.9827, + "step": 605 + }, + { + "epoch": 0.05465121522297876, + "grad_norm": 1.3363567386029256, + "learning_rate": 3.993644143990706e-06, + "loss": 1.0368, + "step": 606 + }, + { + "epoch": 0.054741398746449024, + "grad_norm": 1.9775928329881096, + "learning_rate": 3.99359752056735e-06, + "loss": 0.954, + "step": 607 + }, + { + "epoch": 0.054831582269919285, + "grad_norm": 2.059121572735848, + "learning_rate": 3.993550727039261e-06, + "loss": 0.8915, + "step": 608 + }, + { + "epoch": 0.054921765793389546, + "grad_norm": 2.117616616993303, + "learning_rate": 3.993503763410431e-06, + "loss": 1.0541, + "step": 609 + }, + { + "epoch": 0.05501194931685981, + "grad_norm": 1.8253211013285642, + "learning_rate": 3.9934566296848686e-06, + "loss": 0.9468, + "step": 610 + }, + { + "epoch": 0.055102132840330074, + "grad_norm": 1.8564442208661638, + "learning_rate": 3.993409325866595e-06, + "loss": 1.1188, + "step": 611 + }, + { + "epoch": 0.055192316363800335, + "grad_norm": 0.7417190679690522, + "learning_rate": 3.993361851959645e-06, + "loss": 0.8016, + "step": 612 + }, + { + "epoch": 0.055282499887270596, + "grad_norm": 1.4848743602636665, + "learning_rate": 3.993314207968071e-06, + "loss": 1.0786, + "step": 613 + }, + { + "epoch": 0.05537268341074086, + "grad_norm": 1.7083683320468903, + "learning_rate": 3.993266393895938e-06, + "loss": 1.0212, + "step": 614 + }, + { + "epoch": 0.05546286693421112, + "grad_norm": 2.0306325120631383, + "learning_rate": 3.993218409747326e-06, + "loss": 1.0425, + "step": 615 + }, + { + "epoch": 0.05555305045768138, + "grad_norm": 1.5848005005588732, + "learning_rate": 3.993170255526328e-06, + "loss": 1.03, + "step": 616 + }, + { + "epoch": 0.055643233981151646, + "grad_norm": 1.573826061710417, + "learning_rate": 3.993121931237054e-06, + "loss": 1.0626, + "step": 617 + }, + { + "epoch": 0.05573341750462191, + "grad_norm": 1.7814314231167991, + "learning_rate": 3.993073436883627e-06, + "loss": 1.0541, + "step": 618 + }, + { + "epoch": 0.05582360102809217, + "grad_norm": 1.6341539922777457, + "learning_rate": 3.993024772470184e-06, + "loss": 1.0179, + "step": 619 + }, + { + "epoch": 0.05591378455156243, + "grad_norm": 1.9478243501652504, + "learning_rate": 3.992975938000878e-06, + "loss": 1.0195, + "step": 620 + }, + { + "epoch": 0.05600396807503269, + "grad_norm": 0.7372463986427513, + "learning_rate": 3.992926933479876e-06, + "loss": 0.7911, + "step": 621 + }, + { + "epoch": 0.05609415159850295, + "grad_norm": 1.7182779530882697, + "learning_rate": 3.9928777589113595e-06, + "loss": 1.0579, + "step": 622 + }, + { + "epoch": 0.05618433512197322, + "grad_norm": 1.724588638129151, + "learning_rate": 3.992828414299524e-06, + "loss": 0.9953, + "step": 623 + }, + { + "epoch": 0.05627451864544348, + "grad_norm": 2.970780170351997, + "learning_rate": 3.992778899648579e-06, + "loss": 1.0342, + "step": 624 + }, + { + "epoch": 0.05636470216891374, + "grad_norm": 1.6471758891524906, + "learning_rate": 3.992729214962751e-06, + "loss": 1.0545, + "step": 625 + }, + { + "epoch": 0.056454885692384, + "grad_norm": 1.8689032935100076, + "learning_rate": 3.992679360246279e-06, + "loss": 1.0761, + "step": 626 + }, + { + "epoch": 0.05654506921585426, + "grad_norm": 1.7225767363797395, + "learning_rate": 3.992629335503416e-06, + "loss": 1.0666, + "step": 627 + }, + { + "epoch": 0.05663525273932452, + "grad_norm": 1.7585922207310931, + "learning_rate": 3.9925791407384304e-06, + "loss": 1.0146, + "step": 628 + }, + { + "epoch": 0.05672543626279479, + "grad_norm": 2.1246218625385684, + "learning_rate": 3.992528775955606e-06, + "loss": 1.0795, + "step": 629 + }, + { + "epoch": 0.05681561978626505, + "grad_norm": 1.6137672412757957, + "learning_rate": 3.992478241159239e-06, + "loss": 1.0029, + "step": 630 + }, + { + "epoch": 0.05690580330973531, + "grad_norm": 1.8365478799603792, + "learning_rate": 3.992427536353643e-06, + "loss": 1.0538, + "step": 631 + }, + { + "epoch": 0.05699598683320557, + "grad_norm": 2.4714457605805262, + "learning_rate": 3.992376661543143e-06, + "loss": 1.0216, + "step": 632 + }, + { + "epoch": 0.05708617035667583, + "grad_norm": 2.3627407496263886, + "learning_rate": 3.992325616732081e-06, + "loss": 1.013, + "step": 633 + }, + { + "epoch": 0.057176353880146094, + "grad_norm": 1.735840942554153, + "learning_rate": 3.992274401924811e-06, + "loss": 1.142, + "step": 634 + }, + { + "epoch": 0.05726653740361636, + "grad_norm": 1.637130284977975, + "learning_rate": 3.992223017125704e-06, + "loss": 0.953, + "step": 635 + }, + { + "epoch": 0.05735672092708662, + "grad_norm": 1.7928253315929694, + "learning_rate": 3.992171462339145e-06, + "loss": 1.0603, + "step": 636 + }, + { + "epoch": 0.057446904450556883, + "grad_norm": 1.9134883720218085, + "learning_rate": 3.992119737569532e-06, + "loss": 1.1015, + "step": 637 + }, + { + "epoch": 0.057537087974027144, + "grad_norm": 1.926218412661388, + "learning_rate": 3.992067842821277e-06, + "loss": 1.0363, + "step": 638 + }, + { + "epoch": 0.057627271497497405, + "grad_norm": 2.1456666035984266, + "learning_rate": 3.99201577809881e-06, + "loss": 1.1647, + "step": 639 + }, + { + "epoch": 0.057717455020967666, + "grad_norm": 2.2621502769328226, + "learning_rate": 3.991963543406574e-06, + "loss": 1.0495, + "step": 640 + }, + { + "epoch": 0.057807638544437934, + "grad_norm": 1.8810670605295035, + "learning_rate": 3.991911138749024e-06, + "loss": 1.0308, + "step": 641 + }, + { + "epoch": 0.057897822067908195, + "grad_norm": 1.7277405461221182, + "learning_rate": 3.991858564130633e-06, + "loss": 1.058, + "step": 642 + }, + { + "epoch": 0.057988005591378455, + "grad_norm": 1.9914828381901533, + "learning_rate": 3.991805819555885e-06, + "loss": 1.0135, + "step": 643 + }, + { + "epoch": 0.058078189114848716, + "grad_norm": 1.7981584029133177, + "learning_rate": 3.991752905029283e-06, + "loss": 1.0356, + "step": 644 + }, + { + "epoch": 0.05816837263831898, + "grad_norm": 1.8552589699461894, + "learning_rate": 3.991699820555341e-06, + "loss": 0.9045, + "step": 645 + }, + { + "epoch": 0.05825855616178924, + "grad_norm": 1.797167745875319, + "learning_rate": 3.991646566138588e-06, + "loss": 1.0333, + "step": 646 + }, + { + "epoch": 0.058348739685259506, + "grad_norm": 1.8435245958037154, + "learning_rate": 3.991593141783567e-06, + "loss": 0.9545, + "step": 647 + }, + { + "epoch": 0.058438923208729766, + "grad_norm": 2.570774303564384, + "learning_rate": 3.991539547494839e-06, + "loss": 1.0545, + "step": 648 + }, + { + "epoch": 0.05852910673220003, + "grad_norm": 1.9781639637657704, + "learning_rate": 3.991485783276974e-06, + "loss": 1.0336, + "step": 649 + }, + { + "epoch": 0.05861929025567029, + "grad_norm": 1.7530213568475077, + "learning_rate": 3.991431849134563e-06, + "loss": 1.086, + "step": 650 + }, + { + "epoch": 0.05870947377914055, + "grad_norm": 1.955090624667366, + "learning_rate": 3.991377745072205e-06, + "loss": 0.9268, + "step": 651 + }, + { + "epoch": 0.05879965730261081, + "grad_norm": 1.681922880049015, + "learning_rate": 3.991323471094517e-06, + "loss": 1.0601, + "step": 652 + }, + { + "epoch": 0.05888984082608108, + "grad_norm": 2.099613462932067, + "learning_rate": 3.991269027206131e-06, + "loss": 1.0047, + "step": 653 + }, + { + "epoch": 0.05898002434955134, + "grad_norm": 1.0091818444590905, + "learning_rate": 3.9912144134116916e-06, + "loss": 0.8436, + "step": 654 + }, + { + "epoch": 0.0590702078730216, + "grad_norm": 1.8388865060140966, + "learning_rate": 3.99115962971586e-06, + "loss": 0.9579, + "step": 655 + }, + { + "epoch": 0.05916039139649186, + "grad_norm": 2.0826749397061417, + "learning_rate": 3.991104676123308e-06, + "loss": 1.0707, + "step": 656 + }, + { + "epoch": 0.05925057491996212, + "grad_norm": 0.8590701169272648, + "learning_rate": 3.991049552638727e-06, + "loss": 0.7777, + "step": 657 + }, + { + "epoch": 0.05934075844343238, + "grad_norm": 2.0560183548237876, + "learning_rate": 3.99099425926682e-06, + "loss": 0.9979, + "step": 658 + }, + { + "epoch": 0.05943094196690265, + "grad_norm": 1.7835223778084632, + "learning_rate": 3.990938796012304e-06, + "loss": 0.9986, + "step": 659 + }, + { + "epoch": 0.05952112549037291, + "grad_norm": 1.8560572925554564, + "learning_rate": 3.990883162879912e-06, + "loss": 1.187, + "step": 660 + }, + { + "epoch": 0.05961130901384317, + "grad_norm": 2.196943769830127, + "learning_rate": 3.990827359874391e-06, + "loss": 1.1004, + "step": 661 + }, + { + "epoch": 0.05970149253731343, + "grad_norm": 1.5885423587318122, + "learning_rate": 3.990771387000503e-06, + "loss": 1.0311, + "step": 662 + }, + { + "epoch": 0.05979167606078369, + "grad_norm": 2.6406950630960337, + "learning_rate": 3.990715244263023e-06, + "loss": 1.0594, + "step": 663 + }, + { + "epoch": 0.05988185958425396, + "grad_norm": 1.2111003643155978, + "learning_rate": 3.990658931666741e-06, + "loss": 0.8118, + "step": 664 + }, + { + "epoch": 0.05997204310772422, + "grad_norm": 1.687800415990267, + "learning_rate": 3.990602449216463e-06, + "loss": 0.9524, + "step": 665 + }, + { + "epoch": 0.06006222663119448, + "grad_norm": 1.7304449144362248, + "learning_rate": 3.990545796917008e-06, + "loss": 0.9313, + "step": 666 + }, + { + "epoch": 0.06015241015466474, + "grad_norm": 1.520794027508602, + "learning_rate": 3.99048897477321e-06, + "loss": 1.0809, + "step": 667 + }, + { + "epoch": 0.060242593678135004, + "grad_norm": 1.8304455366954318, + "learning_rate": 3.990431982789917e-06, + "loss": 1.0548, + "step": 668 + }, + { + "epoch": 0.060332777201605264, + "grad_norm": 1.7627181709477011, + "learning_rate": 3.990374820971992e-06, + "loss": 0.9665, + "step": 669 + }, + { + "epoch": 0.06042296072507553, + "grad_norm": 1.820969688087831, + "learning_rate": 3.990317489324312e-06, + "loss": 0.9212, + "step": 670 + }, + { + "epoch": 0.06051314424854579, + "grad_norm": 1.9163404961186332, + "learning_rate": 3.99025998785177e-06, + "loss": 1.0146, + "step": 671 + }, + { + "epoch": 0.060603327772016054, + "grad_norm": 1.6077170118442299, + "learning_rate": 3.990202316559271e-06, + "loss": 0.9737, + "step": 672 + }, + { + "epoch": 0.060693511295486315, + "grad_norm": 3.2722191921954833, + "learning_rate": 3.990144475451738e-06, + "loss": 1.0823, + "step": 673 + }, + { + "epoch": 0.060783694818956575, + "grad_norm": 1.4553828592393734, + "learning_rate": 3.9900864645341036e-06, + "loss": 1.029, + "step": 674 + }, + { + "epoch": 0.060873878342426836, + "grad_norm": 1.8689237686049276, + "learning_rate": 3.990028283811319e-06, + "loss": 1.04, + "step": 675 + }, + { + "epoch": 0.060964061865897104, + "grad_norm": 1.6530980041261096, + "learning_rate": 3.989969933288348e-06, + "loss": 1.0271, + "step": 676 + }, + { + "epoch": 0.061054245389367365, + "grad_norm": 0.8558767382254652, + "learning_rate": 3.98991141297017e-06, + "loss": 0.7547, + "step": 677 + }, + { + "epoch": 0.061144428912837626, + "grad_norm": 1.60151515983769, + "learning_rate": 3.989852722861778e-06, + "loss": 0.9786, + "step": 678 + }, + { + "epoch": 0.061234612436307886, + "grad_norm": 0.9122896853507695, + "learning_rate": 3.98979386296818e-06, + "loss": 0.7829, + "step": 679 + }, + { + "epoch": 0.06132479595977815, + "grad_norm": 1.796145813640767, + "learning_rate": 3.989734833294398e-06, + "loss": 1.0084, + "step": 680 + }, + { + "epoch": 0.06141497948324841, + "grad_norm": 1.7691874391136195, + "learning_rate": 3.989675633845469e-06, + "loss": 1.095, + "step": 681 + }, + { + "epoch": 0.061505163006718676, + "grad_norm": 2.140848635387068, + "learning_rate": 3.989616264626443e-06, + "loss": 1.0068, + "step": 682 + }, + { + "epoch": 0.06159534653018894, + "grad_norm": 1.5559908517063563, + "learning_rate": 3.989556725642388e-06, + "loss": 1.0645, + "step": 683 + }, + { + "epoch": 0.0616855300536592, + "grad_norm": 1.9567198972052118, + "learning_rate": 3.989497016898382e-06, + "loss": 1.0827, + "step": 684 + }, + { + "epoch": 0.06177571357712946, + "grad_norm": 1.603970016051793, + "learning_rate": 3.98943713839952e-06, + "loss": 1.0252, + "step": 685 + }, + { + "epoch": 0.06186589710059972, + "grad_norm": 1.057862764908224, + "learning_rate": 3.9893770901509125e-06, + "loss": 0.8123, + "step": 686 + }, + { + "epoch": 0.06195608062406998, + "grad_norm": 1.660230639417913, + "learning_rate": 3.989316872157682e-06, + "loss": 1.0185, + "step": 687 + }, + { + "epoch": 0.06204626414754025, + "grad_norm": 1.7399589757358225, + "learning_rate": 3.989256484424968e-06, + "loss": 1.1064, + "step": 688 + }, + { + "epoch": 0.06213644767101051, + "grad_norm": 2.170740174133001, + "learning_rate": 3.98919592695792e-06, + "loss": 1.0008, + "step": 689 + }, + { + "epoch": 0.06222663119448077, + "grad_norm": 1.6440840088061452, + "learning_rate": 3.9891351997617096e-06, + "loss": 1.0618, + "step": 690 + }, + { + "epoch": 0.06231681471795103, + "grad_norm": 2.295398471123337, + "learning_rate": 3.989074302841514e-06, + "loss": 0.9722, + "step": 691 + }, + { + "epoch": 0.06240699824142129, + "grad_norm": 1.8376484791458634, + "learning_rate": 3.989013236202533e-06, + "loss": 1.1435, + "step": 692 + }, + { + "epoch": 0.06249718176489155, + "grad_norm": 1.4180145676621847, + "learning_rate": 3.988951999849974e-06, + "loss": 1.0717, + "step": 693 + }, + { + "epoch": 0.06258736528836181, + "grad_norm": 2.0724923803467794, + "learning_rate": 3.988890593789064e-06, + "loss": 0.9493, + "step": 694 + }, + { + "epoch": 0.06267754881183207, + "grad_norm": 1.613634195582732, + "learning_rate": 3.9888290180250415e-06, + "loss": 1.0185, + "step": 695 + }, + { + "epoch": 0.06276773233530233, + "grad_norm": 1.7585151454825403, + "learning_rate": 3.988767272563161e-06, + "loss": 0.9952, + "step": 696 + }, + { + "epoch": 0.06285791585877261, + "grad_norm": 1.8836454076656965, + "learning_rate": 3.988705357408691e-06, + "loss": 0.9727, + "step": 697 + }, + { + "epoch": 0.06294809938224287, + "grad_norm": 1.9778692577341146, + "learning_rate": 3.9886432725669146e-06, + "loss": 0.9903, + "step": 698 + }, + { + "epoch": 0.06303828290571313, + "grad_norm": 1.6134913125337702, + "learning_rate": 3.988581018043128e-06, + "loss": 0.9923, + "step": 699 + }, + { + "epoch": 0.06312846642918339, + "grad_norm": 1.828557867985494, + "learning_rate": 3.988518593842645e-06, + "loss": 1.042, + "step": 700 + }, + { + "epoch": 0.06321864995265365, + "grad_norm": 2.0920040765288626, + "learning_rate": 3.9884559999707906e-06, + "loss": 1.0288, + "step": 701 + }, + { + "epoch": 0.06330883347612391, + "grad_norm": 1.7481544163192533, + "learning_rate": 3.988393236432906e-06, + "loss": 1.0399, + "step": 702 + }, + { + "epoch": 0.06339901699959417, + "grad_norm": 1.7915466918993128, + "learning_rate": 3.988330303234347e-06, + "loss": 1.1063, + "step": 703 + }, + { + "epoch": 0.06348920052306443, + "grad_norm": 1.6509486592406828, + "learning_rate": 3.988267200380483e-06, + "loss": 0.98, + "step": 704 + }, + { + "epoch": 0.0635793840465347, + "grad_norm": 1.4196813240563495, + "learning_rate": 3.988203927876698e-06, + "loss": 1.0802, + "step": 705 + }, + { + "epoch": 0.06366956757000496, + "grad_norm": 1.5677743773209856, + "learning_rate": 3.988140485728391e-06, + "loss": 0.9144, + "step": 706 + }, + { + "epoch": 0.06375975109347522, + "grad_norm": 1.6164376552180375, + "learning_rate": 3.988076873940975e-06, + "loss": 0.9791, + "step": 707 + }, + { + "epoch": 0.06384993461694548, + "grad_norm": 3.303319047986144, + "learning_rate": 3.9880130925198786e-06, + "loss": 1.0283, + "step": 708 + }, + { + "epoch": 0.06394011814041575, + "grad_norm": 2.2913789973181187, + "learning_rate": 3.987949141470543e-06, + "loss": 1.043, + "step": 709 + }, + { + "epoch": 0.06403030166388601, + "grad_norm": 1.553427794508082, + "learning_rate": 3.987885020798425e-06, + "loss": 1.0273, + "step": 710 + }, + { + "epoch": 0.06412048518735627, + "grad_norm": 1.933715939755072, + "learning_rate": 3.987820730508996e-06, + "loss": 0.9973, + "step": 711 + }, + { + "epoch": 0.06421066871082654, + "grad_norm": 1.8295235786013262, + "learning_rate": 3.987756270607742e-06, + "loss": 1.1295, + "step": 712 + }, + { + "epoch": 0.0643008522342968, + "grad_norm": 0.971329822524993, + "learning_rate": 3.987691641100162e-06, + "loss": 0.7695, + "step": 713 + }, + { + "epoch": 0.06439103575776706, + "grad_norm": 1.952464793201664, + "learning_rate": 3.987626841991771e-06, + "loss": 1.1163, + "step": 714 + }, + { + "epoch": 0.06448121928123732, + "grad_norm": 1.9619156423043043, + "learning_rate": 3.987561873288099e-06, + "loss": 1.0689, + "step": 715 + }, + { + "epoch": 0.06457140280470758, + "grad_norm": 1.8700015517855646, + "learning_rate": 3.987496734994688e-06, + "loss": 0.981, + "step": 716 + }, + { + "epoch": 0.06466158632817784, + "grad_norm": 2.0542097225151883, + "learning_rate": 3.987431427117097e-06, + "loss": 1.0616, + "step": 717 + }, + { + "epoch": 0.0647517698516481, + "grad_norm": 1.990254021155487, + "learning_rate": 3.9873659496608985e-06, + "loss": 1.0421, + "step": 718 + }, + { + "epoch": 0.06484195337511836, + "grad_norm": 2.7125234224916692, + "learning_rate": 3.987300302631678e-06, + "loss": 0.9804, + "step": 719 + }, + { + "epoch": 0.06493213689858862, + "grad_norm": 1.9595744346347928, + "learning_rate": 3.987234486035039e-06, + "loss": 1.0644, + "step": 720 + }, + { + "epoch": 0.0650223204220589, + "grad_norm": 2.032644526749634, + "learning_rate": 3.987168499876595e-06, + "loss": 1.0791, + "step": 721 + }, + { + "epoch": 0.06511250394552916, + "grad_norm": 1.936412444519167, + "learning_rate": 3.987102344161978e-06, + "loss": 1.0383, + "step": 722 + }, + { + "epoch": 0.06520268746899942, + "grad_norm": 1.800264098078446, + "learning_rate": 3.987036018896832e-06, + "loss": 1.064, + "step": 723 + }, + { + "epoch": 0.06529287099246968, + "grad_norm": 1.68100144204572, + "learning_rate": 3.986969524086817e-06, + "loss": 1.093, + "step": 724 + }, + { + "epoch": 0.06538305451593994, + "grad_norm": 1.9527394639017757, + "learning_rate": 3.986902859737605e-06, + "loss": 1.0047, + "step": 725 + }, + { + "epoch": 0.0654732380394102, + "grad_norm": 2.6340303858551817, + "learning_rate": 3.986836025854886e-06, + "loss": 0.9621, + "step": 726 + }, + { + "epoch": 0.06556342156288046, + "grad_norm": 1.8691843759165798, + "learning_rate": 3.986769022444362e-06, + "loss": 0.9865, + "step": 727 + }, + { + "epoch": 0.06565360508635072, + "grad_norm": 2.007886358305333, + "learning_rate": 3.986701849511751e-06, + "loss": 1.0245, + "step": 728 + }, + { + "epoch": 0.06574378860982098, + "grad_norm": 1.9962953588865826, + "learning_rate": 3.986634507062782e-06, + "loss": 1.0095, + "step": 729 + }, + { + "epoch": 0.06583397213329124, + "grad_norm": 1.9309767767459436, + "learning_rate": 3.986566995103204e-06, + "loss": 1.0471, + "step": 730 + }, + { + "epoch": 0.0659241556567615, + "grad_norm": 1.7491624053835322, + "learning_rate": 3.986499313638776e-06, + "loss": 0.966, + "step": 731 + }, + { + "epoch": 0.06601433918023177, + "grad_norm": 1.5113315924365351, + "learning_rate": 3.986431462675272e-06, + "loss": 0.9586, + "step": 732 + }, + { + "epoch": 0.06610452270370204, + "grad_norm": 1.3546713646584496, + "learning_rate": 3.9863634422184835e-06, + "loss": 1.0451, + "step": 733 + }, + { + "epoch": 0.0661947062271723, + "grad_norm": 1.685338720241118, + "learning_rate": 3.986295252274213e-06, + "loss": 1.0624, + "step": 734 + }, + { + "epoch": 0.06628488975064256, + "grad_norm": 1.7434234340693129, + "learning_rate": 3.9862268928482796e-06, + "loss": 1.0716, + "step": 735 + }, + { + "epoch": 0.06637507327411282, + "grad_norm": 1.8139023453020386, + "learning_rate": 3.986158363946515e-06, + "loss": 0.9381, + "step": 736 + }, + { + "epoch": 0.06646525679758308, + "grad_norm": 1.622002036938876, + "learning_rate": 3.9860896655747685e-06, + "loss": 1.0298, + "step": 737 + }, + { + "epoch": 0.06655544032105334, + "grad_norm": 1.6503728560896267, + "learning_rate": 3.9860207977388994e-06, + "loss": 1.1017, + "step": 738 + }, + { + "epoch": 0.0666456238445236, + "grad_norm": 1.6135866924363154, + "learning_rate": 3.9859517604447854e-06, + "loss": 0.9631, + "step": 739 + }, + { + "epoch": 0.06673580736799387, + "grad_norm": 1.5948638763458391, + "learning_rate": 3.985882553698317e-06, + "loss": 0.9971, + "step": 740 + }, + { + "epoch": 0.06682599089146413, + "grad_norm": 1.9387206877528411, + "learning_rate": 3.985813177505399e-06, + "loss": 1.0899, + "step": 741 + }, + { + "epoch": 0.06691617441493439, + "grad_norm": 1.8186120862550175, + "learning_rate": 3.985743631871951e-06, + "loss": 0.9759, + "step": 742 + }, + { + "epoch": 0.06700635793840465, + "grad_norm": 1.7533334591119718, + "learning_rate": 3.985673916803907e-06, + "loss": 1.0237, + "step": 743 + }, + { + "epoch": 0.06709654146187491, + "grad_norm": 2.0266563705550884, + "learning_rate": 3.985604032307215e-06, + "loss": 1.0472, + "step": 744 + }, + { + "epoch": 0.06718672498534518, + "grad_norm": 1.7341806801201194, + "learning_rate": 3.985533978387839e-06, + "loss": 1.0392, + "step": 745 + }, + { + "epoch": 0.06727690850881544, + "grad_norm": 2.5382573655854865, + "learning_rate": 3.985463755051756e-06, + "loss": 1.0693, + "step": 746 + }, + { + "epoch": 0.0673670920322857, + "grad_norm": 1.7678119989441297, + "learning_rate": 3.9853933623049575e-06, + "loss": 1.0283, + "step": 747 + }, + { + "epoch": 0.06745727555575597, + "grad_norm": 1.7969913610846169, + "learning_rate": 3.98532280015345e-06, + "loss": 1.0223, + "step": 748 + }, + { + "epoch": 0.06754745907922623, + "grad_norm": 1.722678878531756, + "learning_rate": 3.985252068603254e-06, + "loss": 0.9549, + "step": 749 + }, + { + "epoch": 0.06763764260269649, + "grad_norm": 1.5873967114585945, + "learning_rate": 3.985181167660406e-06, + "loss": 1.0573, + "step": 750 + }, + { + "epoch": 0.06772782612616675, + "grad_norm": 1.7414034460497838, + "learning_rate": 3.985110097330953e-06, + "loss": 1.0532, + "step": 751 + }, + { + "epoch": 0.06781800964963701, + "grad_norm": 1.602540699029862, + "learning_rate": 3.985038857620962e-06, + "loss": 1.0203, + "step": 752 + }, + { + "epoch": 0.06790819317310727, + "grad_norm": 0.7966181023079268, + "learning_rate": 3.9849674485365094e-06, + "loss": 0.7935, + "step": 753 + }, + { + "epoch": 0.06799837669657753, + "grad_norm": 1.8219373270304082, + "learning_rate": 3.98489587008369e-06, + "loss": 0.9578, + "step": 754 + }, + { + "epoch": 0.06808856022004779, + "grad_norm": 2.5255279857691293, + "learning_rate": 3.98482412226861e-06, + "loss": 0.8726, + "step": 755 + }, + { + "epoch": 0.06817874374351805, + "grad_norm": 2.013026255163974, + "learning_rate": 3.984752205097391e-06, + "loss": 1.0627, + "step": 756 + }, + { + "epoch": 0.06826892726698833, + "grad_norm": 1.8985816053162656, + "learning_rate": 3.984680118576171e-06, + "loss": 1.0313, + "step": 757 + }, + { + "epoch": 0.06835911079045859, + "grad_norm": 1.643178206518601, + "learning_rate": 3.984607862711099e-06, + "loss": 0.9826, + "step": 758 + }, + { + "epoch": 0.06844929431392885, + "grad_norm": 2.1761016335874914, + "learning_rate": 3.984535437508341e-06, + "loss": 1.0604, + "step": 759 + }, + { + "epoch": 0.06853947783739911, + "grad_norm": 0.7332234383372087, + "learning_rate": 3.984462842974078e-06, + "loss": 0.7981, + "step": 760 + }, + { + "epoch": 0.06862966136086937, + "grad_norm": 1.6330754556597114, + "learning_rate": 3.984390079114502e-06, + "loss": 1.0427, + "step": 761 + }, + { + "epoch": 0.06871984488433963, + "grad_norm": 1.5192528006192159, + "learning_rate": 3.984317145935824e-06, + "loss": 1.1474, + "step": 762 + }, + { + "epoch": 0.06881002840780989, + "grad_norm": 1.6667646220349572, + "learning_rate": 3.984244043444264e-06, + "loss": 1.0007, + "step": 763 + }, + { + "epoch": 0.06890021193128015, + "grad_norm": 1.6094661057249062, + "learning_rate": 3.984170771646062e-06, + "loss": 0.9767, + "step": 764 + }, + { + "epoch": 0.06899039545475041, + "grad_norm": 1.8075906101377137, + "learning_rate": 3.9840973305474695e-06, + "loss": 1.0696, + "step": 765 + }, + { + "epoch": 0.06908057897822067, + "grad_norm": 1.9541437772645287, + "learning_rate": 3.984023720154752e-06, + "loss": 1.0786, + "step": 766 + }, + { + "epoch": 0.06917076250169094, + "grad_norm": 1.618373491251908, + "learning_rate": 3.9839499404741915e-06, + "loss": 1.038, + "step": 767 + }, + { + "epoch": 0.0692609460251612, + "grad_norm": 0.8933623560299861, + "learning_rate": 3.983875991512082e-06, + "loss": 0.8129, + "step": 768 + }, + { + "epoch": 0.06935112954863147, + "grad_norm": 1.5355061139674744, + "learning_rate": 3.9838018732747345e-06, + "loss": 1.0408, + "step": 769 + }, + { + "epoch": 0.06944131307210173, + "grad_norm": 1.7002431302957586, + "learning_rate": 3.9837275857684716e-06, + "loss": 1.0143, + "step": 770 + }, + { + "epoch": 0.06953149659557199, + "grad_norm": 2.2148625879310773, + "learning_rate": 3.983653128999634e-06, + "loss": 1.1557, + "step": 771 + }, + { + "epoch": 0.06962168011904225, + "grad_norm": 1.543120712329938, + "learning_rate": 3.983578502974574e-06, + "loss": 1.0565, + "step": 772 + }, + { + "epoch": 0.06971186364251251, + "grad_norm": 2.043953127784569, + "learning_rate": 3.983503707699658e-06, + "loss": 1.0199, + "step": 773 + }, + { + "epoch": 0.06980204716598278, + "grad_norm": 1.5299179355578605, + "learning_rate": 3.983428743181268e-06, + "loss": 1.055, + "step": 774 + }, + { + "epoch": 0.06989223068945304, + "grad_norm": 1.7128187314542886, + "learning_rate": 3.983353609425802e-06, + "loss": 1.1, + "step": 775 + }, + { + "epoch": 0.0699824142129233, + "grad_norm": 1.7218791566995895, + "learning_rate": 3.983278306439671e-06, + "loss": 1.0249, + "step": 776 + }, + { + "epoch": 0.07007259773639356, + "grad_norm": 1.5128747428535143, + "learning_rate": 3.983202834229297e-06, + "loss": 1.0602, + "step": 777 + }, + { + "epoch": 0.07016278125986382, + "grad_norm": 1.5472439412651575, + "learning_rate": 3.983127192801123e-06, + "loss": 1.1121, + "step": 778 + }, + { + "epoch": 0.07025296478333408, + "grad_norm": 2.5907952817533784, + "learning_rate": 3.983051382161602e-06, + "loss": 1.0612, + "step": 779 + }, + { + "epoch": 0.07034314830680434, + "grad_norm": 1.6629396330565664, + "learning_rate": 3.982975402317203e-06, + "loss": 1.0476, + "step": 780 + }, + { + "epoch": 0.07043333183027461, + "grad_norm": 1.8255478389409474, + "learning_rate": 3.982899253274409e-06, + "loss": 1.0531, + "step": 781 + }, + { + "epoch": 0.07052351535374488, + "grad_norm": 1.9944894095192505, + "learning_rate": 3.982822935039717e-06, + "loss": 1.0108, + "step": 782 + }, + { + "epoch": 0.07061369887721514, + "grad_norm": 1.7508098227175037, + "learning_rate": 3.982746447619638e-06, + "loss": 0.9609, + "step": 783 + }, + { + "epoch": 0.0707038824006854, + "grad_norm": 0.8618417882018312, + "learning_rate": 3.9826697910207e-06, + "loss": 0.7873, + "step": 784 + }, + { + "epoch": 0.07079406592415566, + "grad_norm": 1.5211118504055157, + "learning_rate": 3.982592965249442e-06, + "loss": 1.0071, + "step": 785 + }, + { + "epoch": 0.07088424944762592, + "grad_norm": 1.772326957746174, + "learning_rate": 3.982515970312422e-06, + "loss": 1.0732, + "step": 786 + }, + { + "epoch": 0.07097443297109618, + "grad_norm": 2.2928395015545497, + "learning_rate": 3.982438806216207e-06, + "loss": 1.0002, + "step": 787 + }, + { + "epoch": 0.07106461649456644, + "grad_norm": 1.8883662326614523, + "learning_rate": 3.982361472967382e-06, + "loss": 1.0101, + "step": 788 + }, + { + "epoch": 0.0711548000180367, + "grad_norm": 1.7554059997398692, + "learning_rate": 3.982283970572546e-06, + "loss": 1.1014, + "step": 789 + }, + { + "epoch": 0.07124498354150696, + "grad_norm": 2.1192991417167373, + "learning_rate": 3.982206299038311e-06, + "loss": 1.0021, + "step": 790 + }, + { + "epoch": 0.07133516706497722, + "grad_norm": 1.6729909434600696, + "learning_rate": 3.9821284583713054e-06, + "loss": 1.1029, + "step": 791 + }, + { + "epoch": 0.07142535058844748, + "grad_norm": 1.6924376833781887, + "learning_rate": 3.98205044857817e-06, + "loss": 1.0124, + "step": 792 + }, + { + "epoch": 0.07151553411191776, + "grad_norm": 0.8499898825772596, + "learning_rate": 3.981972269665561e-06, + "loss": 0.8164, + "step": 793 + }, + { + "epoch": 0.07160571763538802, + "grad_norm": 2.2707596046541134, + "learning_rate": 3.98189392164015e-06, + "loss": 1.0178, + "step": 794 + }, + { + "epoch": 0.07169590115885828, + "grad_norm": 1.7036541458000174, + "learning_rate": 3.981815404508621e-06, + "loss": 1.0702, + "step": 795 + }, + { + "epoch": 0.07178608468232854, + "grad_norm": 1.7021442354186513, + "learning_rate": 3.981736718277674e-06, + "loss": 1.033, + "step": 796 + }, + { + "epoch": 0.0718762682057988, + "grad_norm": 1.8576522026496, + "learning_rate": 3.9816578629540235e-06, + "loss": 1.0599, + "step": 797 + }, + { + "epoch": 0.07196645172926906, + "grad_norm": 1.864126337689764, + "learning_rate": 3.981578838544398e-06, + "loss": 1.0356, + "step": 798 + }, + { + "epoch": 0.07205663525273932, + "grad_norm": 0.7712637670391843, + "learning_rate": 3.981499645055539e-06, + "loss": 0.7697, + "step": 799 + }, + { + "epoch": 0.07214681877620958, + "grad_norm": 0.732006026152002, + "learning_rate": 3.981420282494204e-06, + "loss": 0.8042, + "step": 800 + }, + { + "epoch": 0.07223700229967984, + "grad_norm": 0.8341416293326914, + "learning_rate": 3.981340750867166e-06, + "loss": 0.8026, + "step": 801 + }, + { + "epoch": 0.0723271858231501, + "grad_norm": 1.5975435772177293, + "learning_rate": 3.981261050181209e-06, + "loss": 1.015, + "step": 802 + }, + { + "epoch": 0.07241736934662037, + "grad_norm": 1.6447897954748558, + "learning_rate": 3.9811811804431355e-06, + "loss": 0.9781, + "step": 803 + }, + { + "epoch": 0.07250755287009064, + "grad_norm": 1.6699667048989904, + "learning_rate": 3.981101141659759e-06, + "loss": 1.0709, + "step": 804 + }, + { + "epoch": 0.0725977363935609, + "grad_norm": 1.7610685582373373, + "learning_rate": 3.98102093383791e-06, + "loss": 1.0521, + "step": 805 + }, + { + "epoch": 0.07268791991703116, + "grad_norm": 1.6992589485412126, + "learning_rate": 3.9809405569844315e-06, + "loss": 0.9499, + "step": 806 + }, + { + "epoch": 0.07277810344050142, + "grad_norm": 2.0014815803269186, + "learning_rate": 3.980860011106182e-06, + "loss": 0.9916, + "step": 807 + }, + { + "epoch": 0.07286828696397168, + "grad_norm": 1.4434046344603921, + "learning_rate": 3.980779296210033e-06, + "loss": 0.963, + "step": 808 + }, + { + "epoch": 0.07295847048744195, + "grad_norm": 1.7268013076991118, + "learning_rate": 3.980698412302874e-06, + "loss": 1.0992, + "step": 809 + }, + { + "epoch": 0.0730486540109122, + "grad_norm": 1.9196010876543796, + "learning_rate": 3.980617359391604e-06, + "loss": 1.0396, + "step": 810 + }, + { + "epoch": 0.07313883753438247, + "grad_norm": 1.8418786848877289, + "learning_rate": 3.98053613748314e-06, + "loss": 0.9815, + "step": 811 + }, + { + "epoch": 0.07322902105785273, + "grad_norm": 1.436406507472332, + "learning_rate": 3.980454746584413e-06, + "loss": 0.8864, + "step": 812 + }, + { + "epoch": 0.07331920458132299, + "grad_norm": 1.9722837190785916, + "learning_rate": 3.9803731867023665e-06, + "loss": 1.1092, + "step": 813 + }, + { + "epoch": 0.07340938810479325, + "grad_norm": 1.7703155059654256, + "learning_rate": 3.9802914578439596e-06, + "loss": 1.0103, + "step": 814 + }, + { + "epoch": 0.07349957162826351, + "grad_norm": 1.7195906923896243, + "learning_rate": 3.980209560016167e-06, + "loss": 1.1265, + "step": 815 + }, + { + "epoch": 0.07358975515173379, + "grad_norm": 1.5931531499124394, + "learning_rate": 3.980127493225975e-06, + "loss": 1.1151, + "step": 816 + }, + { + "epoch": 0.07367993867520405, + "grad_norm": 1.7518237692411853, + "learning_rate": 3.980045257480387e-06, + "loss": 1.054, + "step": 817 + }, + { + "epoch": 0.0737701221986743, + "grad_norm": 1.9322557540257073, + "learning_rate": 3.9799628527864205e-06, + "loss": 0.9806, + "step": 818 + }, + { + "epoch": 0.07386030572214457, + "grad_norm": 1.6366753401723928, + "learning_rate": 3.979880279151106e-06, + "loss": 1.005, + "step": 819 + }, + { + "epoch": 0.07395048924561483, + "grad_norm": 1.7471133675128987, + "learning_rate": 3.979797536581489e-06, + "loss": 0.9678, + "step": 820 + }, + { + "epoch": 0.07404067276908509, + "grad_norm": 2.0222713689223286, + "learning_rate": 3.97971462508463e-06, + "loss": 0.8712, + "step": 821 + }, + { + "epoch": 0.07413085629255535, + "grad_norm": 1.6430024133154895, + "learning_rate": 3.979631544667603e-06, + "loss": 1.0125, + "step": 822 + }, + { + "epoch": 0.07422103981602561, + "grad_norm": 1.7450987436507863, + "learning_rate": 3.979548295337496e-06, + "loss": 1.0365, + "step": 823 + }, + { + "epoch": 0.07431122333949587, + "grad_norm": 1.8102747312451883, + "learning_rate": 3.9794648771014146e-06, + "loss": 0.9881, + "step": 824 + }, + { + "epoch": 0.07440140686296613, + "grad_norm": 1.576465602955626, + "learning_rate": 3.9793812899664745e-06, + "loss": 1.0278, + "step": 825 + }, + { + "epoch": 0.0744915903864364, + "grad_norm": 2.4645388455431996, + "learning_rate": 3.979297533939809e-06, + "loss": 0.9867, + "step": 826 + }, + { + "epoch": 0.07458177390990665, + "grad_norm": 1.9542038646267423, + "learning_rate": 3.979213609028564e-06, + "loss": 1.0002, + "step": 827 + }, + { + "epoch": 0.07467195743337693, + "grad_norm": 1.6652435982556342, + "learning_rate": 3.979129515239901e-06, + "loss": 1.0364, + "step": 828 + }, + { + "epoch": 0.07476214095684719, + "grad_norm": 1.518066877412278, + "learning_rate": 3.979045252580994e-06, + "loss": 1.0341, + "step": 829 + }, + { + "epoch": 0.07485232448031745, + "grad_norm": 1.6343312478272303, + "learning_rate": 3.978960821059034e-06, + "loss": 0.9669, + "step": 830 + }, + { + "epoch": 0.07494250800378771, + "grad_norm": 1.651390648511087, + "learning_rate": 3.978876220681225e-06, + "loss": 1.0646, + "step": 831 + }, + { + "epoch": 0.07503269152725797, + "grad_norm": 1.6104879142262747, + "learning_rate": 3.978791451454786e-06, + "loss": 0.9486, + "step": 832 + }, + { + "epoch": 0.07512287505072823, + "grad_norm": 1.736665810873415, + "learning_rate": 3.978706513386949e-06, + "loss": 0.9827, + "step": 833 + }, + { + "epoch": 0.0752130585741985, + "grad_norm": 1.9858807235273201, + "learning_rate": 3.978621406484962e-06, + "loss": 1.0371, + "step": 834 + }, + { + "epoch": 0.07530324209766875, + "grad_norm": 1.8192978024133286, + "learning_rate": 3.978536130756086e-06, + "loss": 1.0161, + "step": 835 + }, + { + "epoch": 0.07539342562113902, + "grad_norm": 1.4883308302262728, + "learning_rate": 3.978450686207599e-06, + "loss": 1.0836, + "step": 836 + }, + { + "epoch": 0.07548360914460928, + "grad_norm": 1.9346412364546801, + "learning_rate": 3.978365072846789e-06, + "loss": 1.0047, + "step": 837 + }, + { + "epoch": 0.07557379266807954, + "grad_norm": 1.6715995347731856, + "learning_rate": 3.9782792906809625e-06, + "loss": 0.9768, + "step": 838 + }, + { + "epoch": 0.0756639761915498, + "grad_norm": 1.777716897213441, + "learning_rate": 3.97819333971744e-06, + "loss": 1.0119, + "step": 839 + }, + { + "epoch": 0.07575415971502007, + "grad_norm": 1.9501179243768387, + "learning_rate": 3.978107219963553e-06, + "loss": 1.0402, + "step": 840 + }, + { + "epoch": 0.07584434323849033, + "grad_norm": 1.5186172521526833, + "learning_rate": 3.978020931426651e-06, + "loss": 0.9282, + "step": 841 + }, + { + "epoch": 0.0759345267619606, + "grad_norm": 1.601896669834361, + "learning_rate": 3.977934474114096e-06, + "loss": 0.9643, + "step": 842 + }, + { + "epoch": 0.07602471028543085, + "grad_norm": 0.7170672829128586, + "learning_rate": 3.977847848033267e-06, + "loss": 0.7809, + "step": 843 + }, + { + "epoch": 0.07611489380890112, + "grad_norm": 1.7674266434140495, + "learning_rate": 3.977761053191553e-06, + "loss": 1.0662, + "step": 844 + }, + { + "epoch": 0.07620507733237138, + "grad_norm": 1.5117567584954563, + "learning_rate": 3.977674089596361e-06, + "loss": 1.0044, + "step": 845 + }, + { + "epoch": 0.07629526085584164, + "grad_norm": 0.7863392055467934, + "learning_rate": 3.97758695725511e-06, + "loss": 0.7791, + "step": 846 + }, + { + "epoch": 0.0763854443793119, + "grad_norm": 1.5657733961120976, + "learning_rate": 3.977499656175236e-06, + "loss": 1.0196, + "step": 847 + }, + { + "epoch": 0.07647562790278216, + "grad_norm": 1.612387790767898, + "learning_rate": 3.977412186364187e-06, + "loss": 0.9306, + "step": 848 + }, + { + "epoch": 0.07656581142625242, + "grad_norm": 1.8038985604526654, + "learning_rate": 3.977324547829428e-06, + "loss": 1.0505, + "step": 849 + }, + { + "epoch": 0.07665599494972268, + "grad_norm": 1.948066324886924, + "learning_rate": 3.977236740578435e-06, + "loss": 1.0441, + "step": 850 + }, + { + "epoch": 0.07674617847319294, + "grad_norm": 2.1988973975642083, + "learning_rate": 3.9771487646187015e-06, + "loss": 1.0105, + "step": 851 + }, + { + "epoch": 0.07683636199666322, + "grad_norm": 1.7962312955154567, + "learning_rate": 3.9770606199577325e-06, + "loss": 1.0568, + "step": 852 + }, + { + "epoch": 0.07692654552013348, + "grad_norm": 2.016142076511639, + "learning_rate": 3.9769723066030505e-06, + "loss": 0.9605, + "step": 853 + }, + { + "epoch": 0.07701672904360374, + "grad_norm": 1.9423488117232157, + "learning_rate": 3.976883824562191e-06, + "loss": 1.077, + "step": 854 + }, + { + "epoch": 0.077106912567074, + "grad_norm": 1.6460268735766064, + "learning_rate": 3.976795173842703e-06, + "loss": 1.0222, + "step": 855 + }, + { + "epoch": 0.07719709609054426, + "grad_norm": 2.002914212900363, + "learning_rate": 3.97670635445215e-06, + "loss": 0.9996, + "step": 856 + }, + { + "epoch": 0.07728727961401452, + "grad_norm": 1.7599884129272856, + "learning_rate": 3.976617366398112e-06, + "loss": 0.9914, + "step": 857 + }, + { + "epoch": 0.07737746313748478, + "grad_norm": 1.348036427707475, + "learning_rate": 3.976528209688181e-06, + "loss": 1.0545, + "step": 858 + }, + { + "epoch": 0.07746764666095504, + "grad_norm": 1.5660205090228467, + "learning_rate": 3.976438884329965e-06, + "loss": 1.0241, + "step": 859 + }, + { + "epoch": 0.0775578301844253, + "grad_norm": 1.7291761957235399, + "learning_rate": 3.976349390331085e-06, + "loss": 1.1535, + "step": 860 + }, + { + "epoch": 0.07764801370789556, + "grad_norm": 1.7107937547751229, + "learning_rate": 3.976259727699178e-06, + "loss": 0.9906, + "step": 861 + }, + { + "epoch": 0.07773819723136582, + "grad_norm": 5.287250608510285, + "learning_rate": 3.976169896441895e-06, + "loss": 0.9774, + "step": 862 + }, + { + "epoch": 0.07782838075483609, + "grad_norm": 2.166612905666748, + "learning_rate": 3.976079896566898e-06, + "loss": 1.0578, + "step": 863 + }, + { + "epoch": 0.07791856427830636, + "grad_norm": 5.323449194659887, + "learning_rate": 3.97598972808187e-06, + "loss": 1.0864, + "step": 864 + }, + { + "epoch": 0.07800874780177662, + "grad_norm": 1.683064459424789, + "learning_rate": 3.975899390994501e-06, + "loss": 1.0155, + "step": 865 + }, + { + "epoch": 0.07809893132524688, + "grad_norm": 1.6407740207386077, + "learning_rate": 3.975808885312502e-06, + "loss": 1.0848, + "step": 866 + }, + { + "epoch": 0.07818911484871714, + "grad_norm": 1.5016355339969416, + "learning_rate": 3.975718211043594e-06, + "loss": 1.069, + "step": 867 + }, + { + "epoch": 0.0782792983721874, + "grad_norm": 1.5579923076415256, + "learning_rate": 3.975627368195515e-06, + "loss": 0.9329, + "step": 868 + }, + { + "epoch": 0.07836948189565766, + "grad_norm": 1.385637404369534, + "learning_rate": 3.975536356776015e-06, + "loss": 0.9884, + "step": 869 + }, + { + "epoch": 0.07845966541912792, + "grad_norm": 1.1980693167023335, + "learning_rate": 3.975445176792861e-06, + "loss": 0.8558, + "step": 870 + }, + { + "epoch": 0.07854984894259819, + "grad_norm": 1.6056450888511846, + "learning_rate": 3.975353828253831e-06, + "loss": 1.0285, + "step": 871 + }, + { + "epoch": 0.07864003246606845, + "grad_norm": 1.5818984741253737, + "learning_rate": 3.97526231116672e-06, + "loss": 1.0237, + "step": 872 + }, + { + "epoch": 0.07873021598953871, + "grad_norm": 1.6039836168094623, + "learning_rate": 3.975170625539338e-06, + "loss": 0.9618, + "step": 873 + }, + { + "epoch": 0.07882039951300897, + "grad_norm": 1.4011278326867103, + "learning_rate": 3.975078771379507e-06, + "loss": 1.0185, + "step": 874 + }, + { + "epoch": 0.07891058303647923, + "grad_norm": 1.8940660672452383, + "learning_rate": 3.974986748695064e-06, + "loss": 0.9309, + "step": 875 + }, + { + "epoch": 0.0790007665599495, + "grad_norm": 0.7987488425161365, + "learning_rate": 3.974894557493862e-06, + "loss": 0.7937, + "step": 876 + }, + { + "epoch": 0.07909095008341976, + "grad_norm": 1.541665214675357, + "learning_rate": 3.974802197783768e-06, + "loss": 1.0683, + "step": 877 + }, + { + "epoch": 0.07918113360689003, + "grad_norm": 1.8775346154934398, + "learning_rate": 3.974709669572661e-06, + "loss": 1.0764, + "step": 878 + }, + { + "epoch": 0.07927131713036029, + "grad_norm": 1.6238615743290428, + "learning_rate": 3.974616972868436e-06, + "loss": 1.0151, + "step": 879 + }, + { + "epoch": 0.07936150065383055, + "grad_norm": 1.9225230171204797, + "learning_rate": 3.974524107679003e-06, + "loss": 1.0259, + "step": 880 + }, + { + "epoch": 0.07945168417730081, + "grad_norm": 1.4289049185052658, + "learning_rate": 3.974431074012286e-06, + "loss": 1.0142, + "step": 881 + }, + { + "epoch": 0.07954186770077107, + "grad_norm": 2.1724577197492017, + "learning_rate": 3.974337871876223e-06, + "loss": 1.0524, + "step": 882 + }, + { + "epoch": 0.07963205122424133, + "grad_norm": 1.5588326713217733, + "learning_rate": 3.974244501278766e-06, + "loss": 0.9907, + "step": 883 + }, + { + "epoch": 0.07972223474771159, + "grad_norm": 1.5064238641834984, + "learning_rate": 3.974150962227883e-06, + "loss": 0.9668, + "step": 884 + }, + { + "epoch": 0.07981241827118185, + "grad_norm": 1.7320410343181156, + "learning_rate": 3.9740572547315535e-06, + "loss": 0.9514, + "step": 885 + }, + { + "epoch": 0.07990260179465211, + "grad_norm": 2.1285353031183423, + "learning_rate": 3.973963378797775e-06, + "loss": 1.0524, + "step": 886 + }, + { + "epoch": 0.07999278531812237, + "grad_norm": 1.542007339915261, + "learning_rate": 3.973869334434556e-06, + "loss": 1.0332, + "step": 887 + }, + { + "epoch": 0.08008296884159265, + "grad_norm": 1.447258623649352, + "learning_rate": 3.973775121649922e-06, + "loss": 1.033, + "step": 888 + }, + { + "epoch": 0.08017315236506291, + "grad_norm": 1.534003526273253, + "learning_rate": 3.973680740451911e-06, + "loss": 0.93, + "step": 889 + }, + { + "epoch": 0.08026333588853317, + "grad_norm": 1.722641925362955, + "learning_rate": 3.9735861908485776e-06, + "loss": 0.9841, + "step": 890 + }, + { + "epoch": 0.08035351941200343, + "grad_norm": 1.5694766880575446, + "learning_rate": 3.973491472847987e-06, + "loss": 1.0407, + "step": 891 + }, + { + "epoch": 0.08044370293547369, + "grad_norm": 1.5061266076922077, + "learning_rate": 3.973396586458222e-06, + "loss": 0.9489, + "step": 892 + }, + { + "epoch": 0.08053388645894395, + "grad_norm": 1.5034304968623837, + "learning_rate": 3.97330153168738e-06, + "loss": 1.0051, + "step": 893 + }, + { + "epoch": 0.08062406998241421, + "grad_norm": 1.921222545003204, + "learning_rate": 3.973206308543571e-06, + "loss": 0.9621, + "step": 894 + }, + { + "epoch": 0.08071425350588447, + "grad_norm": 1.6711199472593377, + "learning_rate": 3.973110917034918e-06, + "loss": 0.9908, + "step": 895 + }, + { + "epoch": 0.08080443702935473, + "grad_norm": 1.6560245576410557, + "learning_rate": 3.973015357169563e-06, + "loss": 0.9261, + "step": 896 + }, + { + "epoch": 0.080894620552825, + "grad_norm": 1.7521060804537738, + "learning_rate": 3.972919628955659e-06, + "loss": 0.9113, + "step": 897 + }, + { + "epoch": 0.08098480407629526, + "grad_norm": 2.174937003966304, + "learning_rate": 3.972823732401373e-06, + "loss": 0.9991, + "step": 898 + }, + { + "epoch": 0.08107498759976552, + "grad_norm": 1.5180891789261688, + "learning_rate": 3.972727667514888e-06, + "loss": 1.0109, + "step": 899 + }, + { + "epoch": 0.08116517112323579, + "grad_norm": 1.563827484796971, + "learning_rate": 3.972631434304402e-06, + "loss": 0.9771, + "step": 900 + }, + { + "epoch": 0.08125535464670605, + "grad_norm": 1.9677641686157323, + "learning_rate": 3.972535032778124e-06, + "loss": 1.1285, + "step": 901 + }, + { + "epoch": 0.08134553817017631, + "grad_norm": 1.4207740965247477, + "learning_rate": 3.97243846294428e-06, + "loss": 1.1031, + "step": 902 + }, + { + "epoch": 0.08143572169364657, + "grad_norm": 1.5632372833276449, + "learning_rate": 3.972341724811111e-06, + "loss": 1.149, + "step": 903 + }, + { + "epoch": 0.08152590521711683, + "grad_norm": 1.7171699890965535, + "learning_rate": 3.972244818386872e-06, + "loss": 0.9599, + "step": 904 + }, + { + "epoch": 0.0816160887405871, + "grad_norm": 1.7074197636933583, + "learning_rate": 3.972147743679828e-06, + "loss": 0.9564, + "step": 905 + }, + { + "epoch": 0.08170627226405736, + "grad_norm": 1.6613040322270625, + "learning_rate": 3.972050500698265e-06, + "loss": 0.9787, + "step": 906 + }, + { + "epoch": 0.08179645578752762, + "grad_norm": 1.8669423856217628, + "learning_rate": 3.971953089450481e-06, + "loss": 1.01, + "step": 907 + }, + { + "epoch": 0.08188663931099788, + "grad_norm": 1.888164148262619, + "learning_rate": 3.971855509944784e-06, + "loss": 1.088, + "step": 908 + }, + { + "epoch": 0.08197682283446814, + "grad_norm": 1.8494366380862846, + "learning_rate": 3.971757762189504e-06, + "loss": 1.0019, + "step": 909 + }, + { + "epoch": 0.0820670063579384, + "grad_norm": 2.267634165200339, + "learning_rate": 3.9716598461929785e-06, + "loss": 1.007, + "step": 910 + }, + { + "epoch": 0.08215718988140866, + "grad_norm": 1.883970039340026, + "learning_rate": 3.971561761963563e-06, + "loss": 0.9909, + "step": 911 + }, + { + "epoch": 0.08224737340487893, + "grad_norm": 1.8672659630031192, + "learning_rate": 3.971463509509628e-06, + "loss": 1.0897, + "step": 912 + }, + { + "epoch": 0.0823375569283492, + "grad_norm": 2.863287744866014, + "learning_rate": 3.9713650888395555e-06, + "loss": 1.0061, + "step": 913 + }, + { + "epoch": 0.08242774045181946, + "grad_norm": 1.7059590968497784, + "learning_rate": 3.9712664999617425e-06, + "loss": 1.0035, + "step": 914 + }, + { + "epoch": 0.08251792397528972, + "grad_norm": 1.5482022535001427, + "learning_rate": 3.971167742884603e-06, + "loss": 0.9869, + "step": 915 + }, + { + "epoch": 0.08260810749875998, + "grad_norm": 1.8056041156726963, + "learning_rate": 3.971068817616564e-06, + "loss": 1.0687, + "step": 916 + }, + { + "epoch": 0.08269829102223024, + "grad_norm": 1.6557897252082538, + "learning_rate": 3.970969724166064e-06, + "loss": 1.0389, + "step": 917 + }, + { + "epoch": 0.0827884745457005, + "grad_norm": 0.9289925807044787, + "learning_rate": 3.970870462541559e-06, + "loss": 0.7816, + "step": 918 + }, + { + "epoch": 0.08287865806917076, + "grad_norm": 1.3371941974469495, + "learning_rate": 3.97077103275152e-06, + "loss": 1.0339, + "step": 919 + }, + { + "epoch": 0.08296884159264102, + "grad_norm": 1.5000237621968364, + "learning_rate": 3.970671434804428e-06, + "loss": 0.9979, + "step": 920 + }, + { + "epoch": 0.08305902511611128, + "grad_norm": 1.7684158953209657, + "learning_rate": 3.970571668708784e-06, + "loss": 1.0102, + "step": 921 + }, + { + "epoch": 0.08314920863958154, + "grad_norm": 0.8193936242402778, + "learning_rate": 3.9704717344731e-06, + "loss": 0.7914, + "step": 922 + }, + { + "epoch": 0.0832393921630518, + "grad_norm": 0.7348498168928472, + "learning_rate": 3.9703716321059026e-06, + "loss": 0.8067, + "step": 923 + }, + { + "epoch": 0.08332957568652208, + "grad_norm": 1.589859271149474, + "learning_rate": 3.9702713616157325e-06, + "loss": 1.0998, + "step": 924 + }, + { + "epoch": 0.08341975920999234, + "grad_norm": 1.8895179092313548, + "learning_rate": 3.9701709230111455e-06, + "loss": 1.0295, + "step": 925 + }, + { + "epoch": 0.0835099427334626, + "grad_norm": 1.5453111854910202, + "learning_rate": 3.970070316300713e-06, + "loss": 0.9788, + "step": 926 + }, + { + "epoch": 0.08360012625693286, + "grad_norm": 2.213353861901247, + "learning_rate": 3.969969541493017e-06, + "loss": 1.0317, + "step": 927 + }, + { + "epoch": 0.08369030978040312, + "grad_norm": 1.6042061920581074, + "learning_rate": 3.969868598596658e-06, + "loss": 1.0656, + "step": 928 + }, + { + "epoch": 0.08378049330387338, + "grad_norm": 1.395229944604117, + "learning_rate": 3.969767487620249e-06, + "loss": 0.9546, + "step": 929 + }, + { + "epoch": 0.08387067682734364, + "grad_norm": 1.3969666828559812, + "learning_rate": 3.969666208572416e-06, + "loss": 1.0598, + "step": 930 + }, + { + "epoch": 0.0839608603508139, + "grad_norm": 1.8253218197237693, + "learning_rate": 3.969564761461802e-06, + "loss": 1.0786, + "step": 931 + }, + { + "epoch": 0.08405104387428416, + "grad_norm": 1.6384969806660785, + "learning_rate": 3.969463146297062e-06, + "loss": 0.9933, + "step": 932 + }, + { + "epoch": 0.08414122739775443, + "grad_norm": 1.9770833411047573, + "learning_rate": 3.969361363086867e-06, + "loss": 1.0409, + "step": 933 + }, + { + "epoch": 0.08423141092122469, + "grad_norm": 1.7098337732913096, + "learning_rate": 3.9692594118399014e-06, + "loss": 0.986, + "step": 934 + }, + { + "epoch": 0.08432159444469496, + "grad_norm": 1.7506071127509197, + "learning_rate": 3.969157292564865e-06, + "loss": 1.0505, + "step": 935 + }, + { + "epoch": 0.08441177796816522, + "grad_norm": 1.8030253761312587, + "learning_rate": 3.96905500527047e-06, + "loss": 1.0008, + "step": 936 + }, + { + "epoch": 0.08450196149163548, + "grad_norm": 1.6628106651589964, + "learning_rate": 3.968952549965445e-06, + "loss": 1.0026, + "step": 937 + }, + { + "epoch": 0.08459214501510574, + "grad_norm": 1.597562530678756, + "learning_rate": 3.968849926658532e-06, + "loss": 1.0503, + "step": 938 + }, + { + "epoch": 0.084682328538576, + "grad_norm": 1.8765638823111062, + "learning_rate": 3.9687471353584866e-06, + "loss": 0.9306, + "step": 939 + }, + { + "epoch": 0.08477251206204627, + "grad_norm": 1.5464936373925389, + "learning_rate": 3.9686441760740795e-06, + "loss": 1.0332, + "step": 940 + }, + { + "epoch": 0.08486269558551653, + "grad_norm": 1.684878679882938, + "learning_rate": 3.968541048814098e-06, + "loss": 0.957, + "step": 941 + }, + { + "epoch": 0.08495287910898679, + "grad_norm": 2.0571005688211117, + "learning_rate": 3.968437753587339e-06, + "loss": 0.9429, + "step": 942 + }, + { + "epoch": 0.08504306263245705, + "grad_norm": 1.4619609491427044, + "learning_rate": 3.968334290402616e-06, + "loss": 1.0588, + "step": 943 + }, + { + "epoch": 0.08513324615592731, + "grad_norm": 1.52457334772499, + "learning_rate": 3.968230659268759e-06, + "loss": 0.9097, + "step": 944 + }, + { + "epoch": 0.08522342967939757, + "grad_norm": 1.9293180413175073, + "learning_rate": 3.968126860194609e-06, + "loss": 1.0849, + "step": 945 + }, + { + "epoch": 0.08531361320286783, + "grad_norm": 1.5512565441746657, + "learning_rate": 3.968022893189025e-06, + "loss": 1.0679, + "step": 946 + }, + { + "epoch": 0.0854037967263381, + "grad_norm": 1.5329312028720166, + "learning_rate": 3.967918758260874e-06, + "loss": 1.0565, + "step": 947 + }, + { + "epoch": 0.08549398024980837, + "grad_norm": 2.8339653619171825, + "learning_rate": 3.967814455419044e-06, + "loss": 1.0676, + "step": 948 + }, + { + "epoch": 0.08558416377327863, + "grad_norm": 1.6682838303153997, + "learning_rate": 3.967709984672434e-06, + "loss": 1.0636, + "step": 949 + }, + { + "epoch": 0.08567434729674889, + "grad_norm": 1.6077358453715063, + "learning_rate": 3.967605346029959e-06, + "loss": 1.05, + "step": 950 + }, + { + "epoch": 0.08576453082021915, + "grad_norm": 1.4359228355480798, + "learning_rate": 3.9675005395005466e-06, + "loss": 1.0981, + "step": 951 + }, + { + "epoch": 0.08585471434368941, + "grad_norm": 1.7780773619051096, + "learning_rate": 3.967395565093139e-06, + "loss": 1.0399, + "step": 952 + }, + { + "epoch": 0.08594489786715967, + "grad_norm": 1.5151441932270113, + "learning_rate": 3.967290422816693e-06, + "loss": 1.0266, + "step": 953 + }, + { + "epoch": 0.08603508139062993, + "grad_norm": 1.8479720444608367, + "learning_rate": 3.967185112680183e-06, + "loss": 0.9815, + "step": 954 + }, + { + "epoch": 0.08612526491410019, + "grad_norm": 1.6726429681727242, + "learning_rate": 3.96707963469259e-06, + "loss": 1.1018, + "step": 955 + }, + { + "epoch": 0.08621544843757045, + "grad_norm": 1.6701001880767845, + "learning_rate": 3.966973988862917e-06, + "loss": 1.0156, + "step": 956 + }, + { + "epoch": 0.08630563196104071, + "grad_norm": 2.2653958961143177, + "learning_rate": 3.966868175200178e-06, + "loss": 1.0719, + "step": 957 + }, + { + "epoch": 0.08639581548451097, + "grad_norm": 1.865619272194509, + "learning_rate": 3.9667621937134e-06, + "loss": 1.0834, + "step": 958 + }, + { + "epoch": 0.08648599900798125, + "grad_norm": 2.2314443508955804, + "learning_rate": 3.966656044411627e-06, + "loss": 0.9369, + "step": 959 + }, + { + "epoch": 0.08657618253145151, + "grad_norm": 1.4885700576384717, + "learning_rate": 3.966549727303918e-06, + "loss": 0.9931, + "step": 960 + }, + { + "epoch": 0.08666636605492177, + "grad_norm": 1.8497037933967357, + "learning_rate": 3.966443242399341e-06, + "loss": 1.0545, + "step": 961 + }, + { + "epoch": 0.08675654957839203, + "grad_norm": 0.9127560269426621, + "learning_rate": 3.966336589706985e-06, + "loss": 0.8028, + "step": 962 + }, + { + "epoch": 0.08684673310186229, + "grad_norm": 1.7298659400595189, + "learning_rate": 3.966229769235948e-06, + "loss": 1.0323, + "step": 963 + }, + { + "epoch": 0.08693691662533255, + "grad_norm": 1.7659034846966564, + "learning_rate": 3.966122780995345e-06, + "loss": 1.1005, + "step": 964 + }, + { + "epoch": 0.08702710014880281, + "grad_norm": 2.0269465709733967, + "learning_rate": 3.966015624994306e-06, + "loss": 1.0607, + "step": 965 + }, + { + "epoch": 0.08711728367227307, + "grad_norm": 1.7966641509699621, + "learning_rate": 3.9659083012419735e-06, + "loss": 0.9931, + "step": 966 + }, + { + "epoch": 0.08720746719574334, + "grad_norm": 1.7032858396404802, + "learning_rate": 3.965800809747505e-06, + "loss": 0.9565, + "step": 967 + }, + { + "epoch": 0.0872976507192136, + "grad_norm": 1.6134017640699656, + "learning_rate": 3.965693150520071e-06, + "loss": 0.9855, + "step": 968 + }, + { + "epoch": 0.08738783424268386, + "grad_norm": 1.8289650837113602, + "learning_rate": 3.96558532356886e-06, + "loss": 0.9517, + "step": 969 + }, + { + "epoch": 0.08747801776615412, + "grad_norm": 1.9758333210628043, + "learning_rate": 3.9654773289030704e-06, + "loss": 1.0687, + "step": 970 + }, + { + "epoch": 0.08756820128962439, + "grad_norm": 1.7963138533400944, + "learning_rate": 3.9653691665319176e-06, + "loss": 1.0335, + "step": 971 + }, + { + "epoch": 0.08765838481309465, + "grad_norm": 1.638228637257977, + "learning_rate": 3.96526083646463e-06, + "loss": 0.951, + "step": 972 + }, + { + "epoch": 0.08774856833656491, + "grad_norm": 1.7162613624813423, + "learning_rate": 3.9651523387104526e-06, + "loss": 0.9693, + "step": 973 + }, + { + "epoch": 0.08783875186003517, + "grad_norm": 1.9771199401532182, + "learning_rate": 3.965043673278641e-06, + "loss": 1.034, + "step": 974 + }, + { + "epoch": 0.08792893538350544, + "grad_norm": 1.5939855214541916, + "learning_rate": 3.964934840178469e-06, + "loss": 1.0265, + "step": 975 + }, + { + "epoch": 0.0880191189069757, + "grad_norm": 1.7292192137060856, + "learning_rate": 3.964825839419221e-06, + "loss": 1.0715, + "step": 976 + }, + { + "epoch": 0.08810930243044596, + "grad_norm": 2.087091933699993, + "learning_rate": 3.964716671010199e-06, + "loss": 1.0075, + "step": 977 + }, + { + "epoch": 0.08819948595391622, + "grad_norm": 1.52025988841593, + "learning_rate": 3.9646073349607165e-06, + "loss": 0.9872, + "step": 978 + }, + { + "epoch": 0.08828966947738648, + "grad_norm": 1.8885786473363195, + "learning_rate": 3.964497831280105e-06, + "loss": 1.1535, + "step": 979 + }, + { + "epoch": 0.08837985300085674, + "grad_norm": 2.3021220783638268, + "learning_rate": 3.964388159977705e-06, + "loss": 1.0267, + "step": 980 + }, + { + "epoch": 0.088470036524327, + "grad_norm": 2.5196092223884485, + "learning_rate": 3.964278321062876e-06, + "loss": 0.9317, + "step": 981 + }, + { + "epoch": 0.08856022004779726, + "grad_norm": 1.590186679605534, + "learning_rate": 3.96416831454499e-06, + "loss": 1.0446, + "step": 982 + }, + { + "epoch": 0.08865040357126754, + "grad_norm": 0.7682949145787163, + "learning_rate": 3.964058140433434e-06, + "loss": 0.833, + "step": 983 + }, + { + "epoch": 0.0887405870947378, + "grad_norm": 1.7198484205388935, + "learning_rate": 3.963947798737606e-06, + "loss": 1.0773, + "step": 984 + }, + { + "epoch": 0.08883077061820806, + "grad_norm": 1.7447359749935012, + "learning_rate": 3.963837289466923e-06, + "loss": 1.0451, + "step": 985 + }, + { + "epoch": 0.08892095414167832, + "grad_norm": 1.57532713989405, + "learning_rate": 3.9637266126308145e-06, + "loss": 0.9905, + "step": 986 + }, + { + "epoch": 0.08901113766514858, + "grad_norm": 0.9081609900760295, + "learning_rate": 3.963615768238724e-06, + "loss": 0.8235, + "step": 987 + }, + { + "epoch": 0.08910132118861884, + "grad_norm": 1.5763241908184136, + "learning_rate": 3.963504756300107e-06, + "loss": 0.9103, + "step": 988 + }, + { + "epoch": 0.0891915047120891, + "grad_norm": 1.580393303556745, + "learning_rate": 3.96339357682444e-06, + "loss": 1.0594, + "step": 989 + }, + { + "epoch": 0.08928168823555936, + "grad_norm": 1.3587883746809026, + "learning_rate": 3.963282229821206e-06, + "loss": 1.0419, + "step": 990 + }, + { + "epoch": 0.08937187175902962, + "grad_norm": 1.5521357311987778, + "learning_rate": 3.963170715299906e-06, + "loss": 1.0336, + "step": 991 + }, + { + "epoch": 0.08946205528249988, + "grad_norm": 1.514788603307594, + "learning_rate": 3.963059033270056e-06, + "loss": 0.9459, + "step": 992 + }, + { + "epoch": 0.08955223880597014, + "grad_norm": 1.7217805673836166, + "learning_rate": 3.9629471837411855e-06, + "loss": 0.968, + "step": 993 + }, + { + "epoch": 0.0896424223294404, + "grad_norm": 3.737845240949401, + "learning_rate": 3.962835166722838e-06, + "loss": 0.9667, + "step": 994 + }, + { + "epoch": 0.08973260585291068, + "grad_norm": 1.5627686078933811, + "learning_rate": 3.96272298222457e-06, + "loss": 1.0271, + "step": 995 + }, + { + "epoch": 0.08982278937638094, + "grad_norm": 1.6425832777815657, + "learning_rate": 3.962610630255956e-06, + "loss": 1.145, + "step": 996 + }, + { + "epoch": 0.0899129728998512, + "grad_norm": 1.8172686098714725, + "learning_rate": 3.96249811082658e-06, + "loss": 0.9916, + "step": 997 + }, + { + "epoch": 0.09000315642332146, + "grad_norm": 0.8626432866478627, + "learning_rate": 3.962385423946046e-06, + "loss": 0.8767, + "step": 998 + }, + { + "epoch": 0.09009333994679172, + "grad_norm": 1.4728317358556458, + "learning_rate": 3.962272569623966e-06, + "loss": 0.9927, + "step": 999 + }, + { + "epoch": 0.09018352347026198, + "grad_norm": 1.967246435129453, + "learning_rate": 3.9621595478699704e-06, + "loss": 0.9131, + "step": 1000 + }, + { + "epoch": 0.09027370699373224, + "grad_norm": 1.6116298693559172, + "learning_rate": 3.962046358693703e-06, + "loss": 1.0137, + "step": 1001 + }, + { + "epoch": 0.0903638905172025, + "grad_norm": 1.7367805089021164, + "learning_rate": 3.961933002104822e-06, + "loss": 1.1022, + "step": 1002 + }, + { + "epoch": 0.09045407404067277, + "grad_norm": 1.6423495719921886, + "learning_rate": 3.961819478112999e-06, + "loss": 0.9443, + "step": 1003 + }, + { + "epoch": 0.09054425756414303, + "grad_norm": 1.6248190118795867, + "learning_rate": 3.961705786727921e-06, + "loss": 1.0427, + "step": 1004 + }, + { + "epoch": 0.09063444108761329, + "grad_norm": 1.494134162440068, + "learning_rate": 3.961591927959288e-06, + "loss": 0.9447, + "step": 1005 + }, + { + "epoch": 0.09072462461108355, + "grad_norm": 0.858610402606118, + "learning_rate": 3.961477901816816e-06, + "loss": 0.7978, + "step": 1006 + }, + { + "epoch": 0.09081480813455382, + "grad_norm": 7.895834836397724, + "learning_rate": 3.961363708310233e-06, + "loss": 1.087, + "step": 1007 + }, + { + "epoch": 0.09090499165802408, + "grad_norm": 1.6599039043050352, + "learning_rate": 3.961249347449286e-06, + "loss": 1.0482, + "step": 1008 + }, + { + "epoch": 0.09099517518149434, + "grad_norm": 1.8642809603149735, + "learning_rate": 3.961134819243728e-06, + "loss": 1.0119, + "step": 1009 + }, + { + "epoch": 0.0910853587049646, + "grad_norm": 1.6652602062866102, + "learning_rate": 3.961020123703335e-06, + "loss": 1.0275, + "step": 1010 + }, + { + "epoch": 0.09117554222843487, + "grad_norm": 1.7478674428902352, + "learning_rate": 3.960905260837892e-06, + "loss": 1.0366, + "step": 1011 + }, + { + "epoch": 0.09126572575190513, + "grad_norm": 1.6772495763928545, + "learning_rate": 3.960790230657199e-06, + "loss": 0.9955, + "step": 1012 + }, + { + "epoch": 0.09135590927537539, + "grad_norm": 1.3257051750660473, + "learning_rate": 3.960675033171072e-06, + "loss": 0.9723, + "step": 1013 + }, + { + "epoch": 0.09144609279884565, + "grad_norm": 1.9241907697742697, + "learning_rate": 3.960559668389341e-06, + "loss": 0.9592, + "step": 1014 + }, + { + "epoch": 0.09153627632231591, + "grad_norm": 1.6398752588422658, + "learning_rate": 3.960444136321847e-06, + "loss": 1.0208, + "step": 1015 + }, + { + "epoch": 0.09162645984578617, + "grad_norm": 2.182569232729619, + "learning_rate": 3.960328436978451e-06, + "loss": 1.1121, + "step": 1016 + }, + { + "epoch": 0.09171664336925643, + "grad_norm": 1.0676177691996513, + "learning_rate": 3.960212570369024e-06, + "loss": 0.7995, + "step": 1017 + }, + { + "epoch": 0.09180682689272669, + "grad_norm": 1.697098525132515, + "learning_rate": 3.9600965365034515e-06, + "loss": 0.9528, + "step": 1018 + }, + { + "epoch": 0.09189701041619697, + "grad_norm": 1.9602956747351288, + "learning_rate": 3.959980335391634e-06, + "loss": 1.0748, + "step": 1019 + }, + { + "epoch": 0.09198719393966723, + "grad_norm": 1.707592834696924, + "learning_rate": 3.959863967043487e-06, + "loss": 0.9625, + "step": 1020 + }, + { + "epoch": 0.09207737746313749, + "grad_norm": 1.662868949277517, + "learning_rate": 3.9597474314689405e-06, + "loss": 1.0551, + "step": 1021 + }, + { + "epoch": 0.09216756098660775, + "grad_norm": 1.6240073620154847, + "learning_rate": 3.959630728677937e-06, + "loss": 1.0408, + "step": 1022 + }, + { + "epoch": 0.09225774451007801, + "grad_norm": 1.7082894792587962, + "learning_rate": 3.959513858680434e-06, + "loss": 1.0008, + "step": 1023 + }, + { + "epoch": 0.09234792803354827, + "grad_norm": 1.8714905001628266, + "learning_rate": 3.959396821486405e-06, + "loss": 1.1145, + "step": 1024 + }, + { + "epoch": 0.09243811155701853, + "grad_norm": 1.0850409230192495, + "learning_rate": 3.959279617105835e-06, + "loss": 0.8297, + "step": 1025 + }, + { + "epoch": 0.09252829508048879, + "grad_norm": 2.041441950465793, + "learning_rate": 3.9591622455487235e-06, + "loss": 1.0321, + "step": 1026 + }, + { + "epoch": 0.09261847860395905, + "grad_norm": 1.6814552408750105, + "learning_rate": 3.959044706825087e-06, + "loss": 1.0936, + "step": 1027 + }, + { + "epoch": 0.09270866212742931, + "grad_norm": 1.6404585980995856, + "learning_rate": 3.958927000944954e-06, + "loss": 0.9254, + "step": 1028 + }, + { + "epoch": 0.09279884565089958, + "grad_norm": 1.777083761461869, + "learning_rate": 3.958809127918368e-06, + "loss": 1.1055, + "step": 1029 + }, + { + "epoch": 0.09288902917436984, + "grad_norm": 1.5033561195766878, + "learning_rate": 3.958691087755387e-06, + "loss": 0.9558, + "step": 1030 + }, + { + "epoch": 0.09297921269784011, + "grad_norm": 1.6985470902251731, + "learning_rate": 3.958572880466081e-06, + "loss": 0.8928, + "step": 1031 + }, + { + "epoch": 0.09306939622131037, + "grad_norm": 1.432663784816489, + "learning_rate": 3.9584545060605385e-06, + "loss": 1.0053, + "step": 1032 + }, + { + "epoch": 0.09315957974478063, + "grad_norm": 1.5822440298651612, + "learning_rate": 3.958335964548859e-06, + "loss": 0.9836, + "step": 1033 + }, + { + "epoch": 0.0932497632682509, + "grad_norm": 1.7843047600865933, + "learning_rate": 3.958217255941156e-06, + "loss": 1.0702, + "step": 1034 + }, + { + "epoch": 0.09333994679172115, + "grad_norm": 1.5516238289781554, + "learning_rate": 3.95809838024756e-06, + "loss": 0.9717, + "step": 1035 + }, + { + "epoch": 0.09343013031519141, + "grad_norm": 1.6886565871983934, + "learning_rate": 3.957979337478212e-06, + "loss": 0.9837, + "step": 1036 + }, + { + "epoch": 0.09352031383866168, + "grad_norm": 1.8520366206534182, + "learning_rate": 3.957860127643272e-06, + "loss": 1.0314, + "step": 1037 + }, + { + "epoch": 0.09361049736213194, + "grad_norm": 1.864671488405953, + "learning_rate": 3.95774075075291e-06, + "loss": 1.0264, + "step": 1038 + }, + { + "epoch": 0.0937006808856022, + "grad_norm": 1.5562857076079681, + "learning_rate": 3.957621206817312e-06, + "loss": 1.0799, + "step": 1039 + }, + { + "epoch": 0.09379086440907246, + "grad_norm": 1.6872720917693798, + "learning_rate": 3.957501495846679e-06, + "loss": 1.006, + "step": 1040 + }, + { + "epoch": 0.09388104793254272, + "grad_norm": 0.951608357945282, + "learning_rate": 3.957381617851225e-06, + "loss": 0.803, + "step": 1041 + }, + { + "epoch": 0.09397123145601298, + "grad_norm": 1.6081387855125715, + "learning_rate": 3.9572615728411776e-06, + "loss": 0.9438, + "step": 1042 + }, + { + "epoch": 0.09406141497948325, + "grad_norm": 1.8952498184242257, + "learning_rate": 3.957141360826781e-06, + "loss": 0.9009, + "step": 1043 + }, + { + "epoch": 0.09415159850295352, + "grad_norm": 1.7173009139035718, + "learning_rate": 3.957020981818292e-06, + "loss": 1.0043, + "step": 1044 + }, + { + "epoch": 0.09424178202642378, + "grad_norm": 1.4453693275620771, + "learning_rate": 3.956900435825982e-06, + "loss": 1.0094, + "step": 1045 + }, + { + "epoch": 0.09433196554989404, + "grad_norm": 1.5231657128103397, + "learning_rate": 3.9567797228601364e-06, + "loss": 0.9718, + "step": 1046 + }, + { + "epoch": 0.0944221490733643, + "grad_norm": 1.5493942707643737, + "learning_rate": 3.956658842931055e-06, + "loss": 1.0973, + "step": 1047 + }, + { + "epoch": 0.09451233259683456, + "grad_norm": 1.4665477556373059, + "learning_rate": 3.956537796049052e-06, + "loss": 1.0317, + "step": 1048 + }, + { + "epoch": 0.09460251612030482, + "grad_norm": 0.9801701854832269, + "learning_rate": 3.956416582224457e-06, + "loss": 0.8226, + "step": 1049 + }, + { + "epoch": 0.09469269964377508, + "grad_norm": 1.7897835823379444, + "learning_rate": 3.956295201467611e-06, + "loss": 1.0561, + "step": 1050 + }, + { + "epoch": 0.09478288316724534, + "grad_norm": 1.407308476834968, + "learning_rate": 3.956173653788872e-06, + "loss": 1.0511, + "step": 1051 + }, + { + "epoch": 0.0948730666907156, + "grad_norm": 2.2537522288145873, + "learning_rate": 3.95605193919861e-06, + "loss": 1.0596, + "step": 1052 + }, + { + "epoch": 0.09496325021418586, + "grad_norm": 1.6310898303911094, + "learning_rate": 3.955930057707211e-06, + "loss": 1.0079, + "step": 1053 + }, + { + "epoch": 0.09505343373765612, + "grad_norm": 1.645247206157636, + "learning_rate": 3.955808009325075e-06, + "loss": 1.0292, + "step": 1054 + }, + { + "epoch": 0.0951436172611264, + "grad_norm": 1.7253401213654398, + "learning_rate": 3.955685794062615e-06, + "loss": 0.9909, + "step": 1055 + }, + { + "epoch": 0.09523380078459666, + "grad_norm": 1.870033616870485, + "learning_rate": 3.95556341193026e-06, + "loss": 1.1221, + "step": 1056 + }, + { + "epoch": 0.09532398430806692, + "grad_norm": 3.72852611190576, + "learning_rate": 3.955440862938452e-06, + "loss": 1.1151, + "step": 1057 + }, + { + "epoch": 0.09541416783153718, + "grad_norm": 2.1466105513415465, + "learning_rate": 3.955318147097647e-06, + "loss": 1.02, + "step": 1058 + }, + { + "epoch": 0.09550435135500744, + "grad_norm": 1.9865879725331888, + "learning_rate": 3.955195264418316e-06, + "loss": 0.9844, + "step": 1059 + }, + { + "epoch": 0.0955945348784777, + "grad_norm": 1.5619133420148081, + "learning_rate": 3.955072214910944e-06, + "loss": 0.8777, + "step": 1060 + }, + { + "epoch": 0.09568471840194796, + "grad_norm": 1.8270854765800801, + "learning_rate": 3.954948998586032e-06, + "loss": 0.9425, + "step": 1061 + }, + { + "epoch": 0.09577490192541822, + "grad_norm": 1.519167270559141, + "learning_rate": 3.954825615454089e-06, + "loss": 1.0274, + "step": 1062 + }, + { + "epoch": 0.09586508544888848, + "grad_norm": 1.7361953392789744, + "learning_rate": 3.954702065525649e-06, + "loss": 1.015, + "step": 1063 + }, + { + "epoch": 0.09595526897235875, + "grad_norm": 1.9519880113428953, + "learning_rate": 3.954578348811248e-06, + "loss": 1.0474, + "step": 1064 + }, + { + "epoch": 0.096045452495829, + "grad_norm": 1.7726415116566199, + "learning_rate": 3.954454465321447e-06, + "loss": 0.9938, + "step": 1065 + }, + { + "epoch": 0.09613563601929928, + "grad_norm": 1.9025374358976659, + "learning_rate": 3.954330415066813e-06, + "loss": 1.0991, + "step": 1066 + }, + { + "epoch": 0.09622581954276954, + "grad_norm": 1.745185632969791, + "learning_rate": 3.954206198057932e-06, + "loss": 1.0061, + "step": 1067 + }, + { + "epoch": 0.0963160030662398, + "grad_norm": 1.6329747169625786, + "learning_rate": 3.954081814305403e-06, + "loss": 1.0769, + "step": 1068 + }, + { + "epoch": 0.09640618658971006, + "grad_norm": 1.6358039138306841, + "learning_rate": 3.953957263819839e-06, + "loss": 0.9565, + "step": 1069 + }, + { + "epoch": 0.09649637011318032, + "grad_norm": 1.6490850397116226, + "learning_rate": 3.953832546611867e-06, + "loss": 0.9527, + "step": 1070 + }, + { + "epoch": 0.09658655363665059, + "grad_norm": 1.783156729655548, + "learning_rate": 3.953707662692129e-06, + "loss": 1.0458, + "step": 1071 + }, + { + "epoch": 0.09667673716012085, + "grad_norm": 1.7297522997068293, + "learning_rate": 3.95358261207128e-06, + "loss": 0.9506, + "step": 1072 + }, + { + "epoch": 0.0967669206835911, + "grad_norm": 1.966760386911722, + "learning_rate": 3.953457394759992e-06, + "loss": 1.0111, + "step": 1073 + }, + { + "epoch": 0.09685710420706137, + "grad_norm": 1.6164227580227297, + "learning_rate": 3.953332010768947e-06, + "loss": 1.0364, + "step": 1074 + }, + { + "epoch": 0.09694728773053163, + "grad_norm": 1.9255176541755592, + "learning_rate": 3.9532064601088436e-06, + "loss": 0.9822, + "step": 1075 + }, + { + "epoch": 0.09703747125400189, + "grad_norm": 2.071847253213763, + "learning_rate": 3.953080742790396e-06, + "loss": 1.1735, + "step": 1076 + }, + { + "epoch": 0.09712765477747215, + "grad_norm": 0.8965076784250808, + "learning_rate": 3.95295485882433e-06, + "loss": 0.864, + "step": 1077 + }, + { + "epoch": 0.09721783830094242, + "grad_norm": 0.7138684816779266, + "learning_rate": 3.952828808221387e-06, + "loss": 0.8317, + "step": 1078 + }, + { + "epoch": 0.09730802182441269, + "grad_norm": 2.4184324346840023, + "learning_rate": 3.9527025909923225e-06, + "loss": 0.9956, + "step": 1079 + }, + { + "epoch": 0.09739820534788295, + "grad_norm": 1.7048779486994217, + "learning_rate": 3.952576207147906e-06, + "loss": 1.1099, + "step": 1080 + }, + { + "epoch": 0.09748838887135321, + "grad_norm": 1.723322248712206, + "learning_rate": 3.95244965669892e-06, + "loss": 1.0562, + "step": 1081 + }, + { + "epoch": 0.09757857239482347, + "grad_norm": 2.176883000795314, + "learning_rate": 3.952322939656165e-06, + "loss": 0.9775, + "step": 1082 + }, + { + "epoch": 0.09766875591829373, + "grad_norm": 1.5199830626497743, + "learning_rate": 3.952196056030451e-06, + "loss": 1.0116, + "step": 1083 + }, + { + "epoch": 0.09775893944176399, + "grad_norm": 1.611375176869845, + "learning_rate": 3.952069005832605e-06, + "loss": 0.9862, + "step": 1084 + }, + { + "epoch": 0.09784912296523425, + "grad_norm": 1.7494468495945439, + "learning_rate": 3.951941789073468e-06, + "loss": 1.0444, + "step": 1085 + }, + { + "epoch": 0.09793930648870451, + "grad_norm": 1.7424269712625278, + "learning_rate": 3.9518144057638955e-06, + "loss": 1.0465, + "step": 1086 + }, + { + "epoch": 0.09802949001217477, + "grad_norm": 1.769560554716473, + "learning_rate": 3.951686855914755e-06, + "loss": 1.0132, + "step": 1087 + }, + { + "epoch": 0.09811967353564503, + "grad_norm": 1.8506686471457607, + "learning_rate": 3.9515591395369305e-06, + "loss": 1.0459, + "step": 1088 + }, + { + "epoch": 0.0982098570591153, + "grad_norm": 1.5183829943382952, + "learning_rate": 3.95143125664132e-06, + "loss": 1.0028, + "step": 1089 + }, + { + "epoch": 0.09830004058258557, + "grad_norm": 1.8362338861702363, + "learning_rate": 3.951303207238833e-06, + "loss": 1.0386, + "step": 1090 + }, + { + "epoch": 0.09839022410605583, + "grad_norm": 1.5413912277920416, + "learning_rate": 3.951174991340399e-06, + "loss": 0.9888, + "step": 1091 + }, + { + "epoch": 0.09848040762952609, + "grad_norm": 1.526433258676769, + "learning_rate": 3.9510466089569546e-06, + "loss": 0.9766, + "step": 1092 + }, + { + "epoch": 0.09857059115299635, + "grad_norm": 1.5806489156329362, + "learning_rate": 3.950918060099456e-06, + "loss": 0.9823, + "step": 1093 + }, + { + "epoch": 0.09866077467646661, + "grad_norm": 1.5100797694022279, + "learning_rate": 3.950789344778871e-06, + "loss": 1.0677, + "step": 1094 + }, + { + "epoch": 0.09875095819993687, + "grad_norm": 1.5896187137984312, + "learning_rate": 3.950660463006184e-06, + "loss": 0.9612, + "step": 1095 + }, + { + "epoch": 0.09884114172340713, + "grad_norm": 1.4592273514940668, + "learning_rate": 3.950531414792389e-06, + "loss": 1.0853, + "step": 1096 + }, + { + "epoch": 0.0989313252468774, + "grad_norm": 1.5044990461149326, + "learning_rate": 3.950402200148498e-06, + "loss": 1.0248, + "step": 1097 + }, + { + "epoch": 0.09902150877034765, + "grad_norm": 1.820414397044072, + "learning_rate": 3.950272819085538e-06, + "loss": 0.9967, + "step": 1098 + }, + { + "epoch": 0.09911169229381792, + "grad_norm": 1.9388678244740252, + "learning_rate": 3.9501432716145474e-06, + "loss": 1.1042, + "step": 1099 + }, + { + "epoch": 0.09920187581728818, + "grad_norm": 2.5052278217871518, + "learning_rate": 3.950013557746579e-06, + "loss": 0.9125, + "step": 1100 + }, + { + "epoch": 0.09929205934075844, + "grad_norm": 1.5431981942240907, + "learning_rate": 3.949883677492703e-06, + "loss": 1.056, + "step": 1101 + }, + { + "epoch": 0.09938224286422871, + "grad_norm": 1.5985545125811642, + "learning_rate": 3.9497536308639994e-06, + "loss": 1.0319, + "step": 1102 + }, + { + "epoch": 0.09947242638769897, + "grad_norm": 1.9299707243482815, + "learning_rate": 3.949623417871565e-06, + "loss": 1.068, + "step": 1103 + }, + { + "epoch": 0.09956260991116923, + "grad_norm": 3.1659781226090864, + "learning_rate": 3.949493038526511e-06, + "loss": 0.9996, + "step": 1104 + }, + { + "epoch": 0.0996527934346395, + "grad_norm": 1.7923780478008549, + "learning_rate": 3.949362492839961e-06, + "loss": 0.9717, + "step": 1105 + }, + { + "epoch": 0.09974297695810976, + "grad_norm": 1.7220453056950427, + "learning_rate": 3.949231780823054e-06, + "loss": 0.9794, + "step": 1106 + }, + { + "epoch": 0.09983316048158002, + "grad_norm": 1.6825195567565947, + "learning_rate": 3.949100902486945e-06, + "loss": 0.959, + "step": 1107 + }, + { + "epoch": 0.09992334400505028, + "grad_norm": 1.5959435868917464, + "learning_rate": 3.948969857842799e-06, + "loss": 0.9842, + "step": 1108 + }, + { + "epoch": 0.10001352752852054, + "grad_norm": 1.7993069983641432, + "learning_rate": 3.948838646901798e-06, + "loss": 1.0076, + "step": 1109 + }, + { + "epoch": 0.1001037110519908, + "grad_norm": 1.9054077664196705, + "learning_rate": 3.948707269675138e-06, + "loss": 1.006, + "step": 1110 + }, + { + "epoch": 0.10019389457546106, + "grad_norm": 1.6950181978087167, + "learning_rate": 3.948575726174028e-06, + "loss": 1.003, + "step": 1111 + }, + { + "epoch": 0.10028407809893132, + "grad_norm": 1.9896836047511124, + "learning_rate": 3.9484440164096935e-06, + "loss": 1.0937, + "step": 1112 + }, + { + "epoch": 0.10037426162240158, + "grad_norm": 1.8249950513380706, + "learning_rate": 3.948312140393372e-06, + "loss": 1.0413, + "step": 1113 + }, + { + "epoch": 0.10046444514587186, + "grad_norm": 1.8742771026699263, + "learning_rate": 3.948180098136316e-06, + "loss": 0.986, + "step": 1114 + }, + { + "epoch": 0.10055462866934212, + "grad_norm": 1.8726763632294596, + "learning_rate": 3.948047889649791e-06, + "loss": 0.9858, + "step": 1115 + }, + { + "epoch": 0.10064481219281238, + "grad_norm": 1.5962761698633583, + "learning_rate": 3.947915514945079e-06, + "loss": 0.9951, + "step": 1116 + }, + { + "epoch": 0.10073499571628264, + "grad_norm": 2.06307073845935, + "learning_rate": 3.947782974033474e-06, + "loss": 1.082, + "step": 1117 + }, + { + "epoch": 0.1008251792397529, + "grad_norm": 2.084038030332301, + "learning_rate": 3.9476502669262866e-06, + "loss": 0.9784, + "step": 1118 + }, + { + "epoch": 0.10091536276322316, + "grad_norm": 1.7112967057468724, + "learning_rate": 3.947517393634839e-06, + "loss": 0.9813, + "step": 1119 + }, + { + "epoch": 0.10100554628669342, + "grad_norm": 1.635592198114033, + "learning_rate": 3.947384354170469e-06, + "loss": 1.0564, + "step": 1120 + }, + { + "epoch": 0.10109572981016368, + "grad_norm": 1.6423378858297368, + "learning_rate": 3.947251148544528e-06, + "loss": 1.0543, + "step": 1121 + }, + { + "epoch": 0.10118591333363394, + "grad_norm": 1.678234305631302, + "learning_rate": 3.947117776768382e-06, + "loss": 1.0434, + "step": 1122 + }, + { + "epoch": 0.1012760968571042, + "grad_norm": 1.4242163025651369, + "learning_rate": 3.9469842388534105e-06, + "loss": 0.9951, + "step": 1123 + }, + { + "epoch": 0.10136628038057446, + "grad_norm": 1.4966144502249303, + "learning_rate": 3.946850534811009e-06, + "loss": 1.0252, + "step": 1124 + }, + { + "epoch": 0.10145646390404472, + "grad_norm": 1.432320260427177, + "learning_rate": 3.946716664652585e-06, + "loss": 1.0022, + "step": 1125 + }, + { + "epoch": 0.101546647427515, + "grad_norm": 2.1901883635663877, + "learning_rate": 3.94658262838956e-06, + "loss": 0.9189, + "step": 1126 + }, + { + "epoch": 0.10163683095098526, + "grad_norm": 1.4325910590557838, + "learning_rate": 3.946448426033373e-06, + "loss": 0.934, + "step": 1127 + }, + { + "epoch": 0.10172701447445552, + "grad_norm": 1.5155173194623048, + "learning_rate": 3.946314057595473e-06, + "loss": 1.0478, + "step": 1128 + }, + { + "epoch": 0.10181719799792578, + "grad_norm": 1.753361674769892, + "learning_rate": 3.946179523087326e-06, + "loss": 1.0794, + "step": 1129 + }, + { + "epoch": 0.10190738152139604, + "grad_norm": 1.4233573480645352, + "learning_rate": 3.9460448225204104e-06, + "loss": 0.9303, + "step": 1130 + }, + { + "epoch": 0.1019975650448663, + "grad_norm": 1.745555274081651, + "learning_rate": 3.945909955906221e-06, + "loss": 1.0077, + "step": 1131 + }, + { + "epoch": 0.10208774856833656, + "grad_norm": 1.6194272955191327, + "learning_rate": 3.945774923256264e-06, + "loss": 1.0145, + "step": 1132 + }, + { + "epoch": 0.10217793209180683, + "grad_norm": 2.4300479928132055, + "learning_rate": 3.945639724582062e-06, + "loss": 1.0832, + "step": 1133 + }, + { + "epoch": 0.10226811561527709, + "grad_norm": 1.8526187893649049, + "learning_rate": 3.94550435989515e-06, + "loss": 0.9769, + "step": 1134 + }, + { + "epoch": 0.10235829913874735, + "grad_norm": 1.36388910508958, + "learning_rate": 3.945368829207079e-06, + "loss": 1.0368, + "step": 1135 + }, + { + "epoch": 0.10244848266221761, + "grad_norm": 1.6099487319033796, + "learning_rate": 3.945233132529414e-06, + "loss": 0.9766, + "step": 1136 + }, + { + "epoch": 0.10253866618568787, + "grad_norm": 1.6444309956126664, + "learning_rate": 3.9450972698737304e-06, + "loss": 1.0193, + "step": 1137 + }, + { + "epoch": 0.10262884970915814, + "grad_norm": 1.7107671365444062, + "learning_rate": 3.944961241251623e-06, + "loss": 0.9927, + "step": 1138 + }, + { + "epoch": 0.1027190332326284, + "grad_norm": 1.5255333202402792, + "learning_rate": 3.9448250466746985e-06, + "loss": 0.9816, + "step": 1139 + }, + { + "epoch": 0.10280921675609866, + "grad_norm": 1.815730799942576, + "learning_rate": 3.944688686154578e-06, + "loss": 1.045, + "step": 1140 + }, + { + "epoch": 0.10289940027956893, + "grad_norm": 1.7273109160312943, + "learning_rate": 3.944552159702894e-06, + "loss": 0.9757, + "step": 1141 + }, + { + "epoch": 0.10298958380303919, + "grad_norm": 1.5782938196512373, + "learning_rate": 3.944415467331299e-06, + "loss": 1.0055, + "step": 1142 + }, + { + "epoch": 0.10307976732650945, + "grad_norm": 1.3603201025223048, + "learning_rate": 3.944278609051455e-06, + "loss": 0.9685, + "step": 1143 + }, + { + "epoch": 0.10316995084997971, + "grad_norm": 1.5416293354497739, + "learning_rate": 3.944141584875039e-06, + "loss": 0.9486, + "step": 1144 + }, + { + "epoch": 0.10326013437344997, + "grad_norm": 1.5284554028754327, + "learning_rate": 3.944004394813743e-06, + "loss": 1.0162, + "step": 1145 + }, + { + "epoch": 0.10335031789692023, + "grad_norm": 1.4815454116025688, + "learning_rate": 3.943867038879273e-06, + "loss": 1.0387, + "step": 1146 + }, + { + "epoch": 0.10344050142039049, + "grad_norm": 1.7650986958862902, + "learning_rate": 3.943729517083349e-06, + "loss": 1.0481, + "step": 1147 + }, + { + "epoch": 0.10353068494386075, + "grad_norm": 1.6472643252705257, + "learning_rate": 3.943591829437705e-06, + "loss": 1.0399, + "step": 1148 + }, + { + "epoch": 0.10362086846733101, + "grad_norm": 1.6690744491803648, + "learning_rate": 3.9434539759540895e-06, + "loss": 0.9576, + "step": 1149 + }, + { + "epoch": 0.10371105199080129, + "grad_norm": 1.5872314886857357, + "learning_rate": 3.943315956644264e-06, + "loss": 0.9504, + "step": 1150 + }, + { + "epoch": 0.10380123551427155, + "grad_norm": 1.8411032866472405, + "learning_rate": 3.943177771520006e-06, + "loss": 1.0844, + "step": 1151 + }, + { + "epoch": 0.10389141903774181, + "grad_norm": 1.7680423358836888, + "learning_rate": 3.9430394205931065e-06, + "loss": 1.0561, + "step": 1152 + }, + { + "epoch": 0.10398160256121207, + "grad_norm": 1.4814275289133252, + "learning_rate": 3.942900903875369e-06, + "loss": 0.8873, + "step": 1153 + }, + { + "epoch": 0.10407178608468233, + "grad_norm": 1.5418006048357944, + "learning_rate": 3.942762221378614e-06, + "loss": 1.0199, + "step": 1154 + }, + { + "epoch": 0.10416196960815259, + "grad_norm": 1.4193051768589424, + "learning_rate": 3.942623373114673e-06, + "loss": 1.0471, + "step": 1155 + }, + { + "epoch": 0.10425215313162285, + "grad_norm": 1.7738058237078056, + "learning_rate": 3.942484359095396e-06, + "loss": 1.0052, + "step": 1156 + }, + { + "epoch": 0.10434233665509311, + "grad_norm": 1.5411996477724172, + "learning_rate": 3.942345179332642e-06, + "loss": 0.9452, + "step": 1157 + }, + { + "epoch": 0.10443252017856337, + "grad_norm": 1.3683458489292095, + "learning_rate": 3.942205833838287e-06, + "loss": 0.9634, + "step": 1158 + }, + { + "epoch": 0.10452270370203363, + "grad_norm": 1.799164151461906, + "learning_rate": 3.9420663226242204e-06, + "loss": 1.1089, + "step": 1159 + }, + { + "epoch": 0.1046128872255039, + "grad_norm": 1.672810658112354, + "learning_rate": 3.941926645702348e-06, + "loss": 1.0632, + "step": 1160 + }, + { + "epoch": 0.10470307074897416, + "grad_norm": 3.960851183069289, + "learning_rate": 3.941786803084586e-06, + "loss": 1.0009, + "step": 1161 + }, + { + "epoch": 0.10479325427244443, + "grad_norm": 1.849520688211691, + "learning_rate": 3.941646794782867e-06, + "loss": 0.9644, + "step": 1162 + }, + { + "epoch": 0.10488343779591469, + "grad_norm": 1.6836277367509218, + "learning_rate": 3.941506620809137e-06, + "loss": 1.0254, + "step": 1163 + }, + { + "epoch": 0.10497362131938495, + "grad_norm": 1.7538953071371812, + "learning_rate": 3.941366281175357e-06, + "loss": 1.0037, + "step": 1164 + }, + { + "epoch": 0.10506380484285521, + "grad_norm": 1.4247769934176682, + "learning_rate": 3.941225775893502e-06, + "loss": 0.9483, + "step": 1165 + }, + { + "epoch": 0.10515398836632547, + "grad_norm": 1.4862403793601082, + "learning_rate": 3.941085104975559e-06, + "loss": 0.992, + "step": 1166 + }, + { + "epoch": 0.10524417188979573, + "grad_norm": 0.8395447664576765, + "learning_rate": 3.9409442684335325e-06, + "loss": 0.8272, + "step": 1167 + }, + { + "epoch": 0.105334355413266, + "grad_norm": 1.7462986904775268, + "learning_rate": 3.940803266279438e-06, + "loss": 1.0115, + "step": 1168 + }, + { + "epoch": 0.10542453893673626, + "grad_norm": 2.166062405272844, + "learning_rate": 3.9406620985253076e-06, + "loss": 0.9504, + "step": 1169 + }, + { + "epoch": 0.10551472246020652, + "grad_norm": 2.4872970191528827, + "learning_rate": 3.940520765183187e-06, + "loss": 0.9874, + "step": 1170 + }, + { + "epoch": 0.10560490598367678, + "grad_norm": 1.951643236748808, + "learning_rate": 3.940379266265134e-06, + "loss": 1.0462, + "step": 1171 + }, + { + "epoch": 0.10569508950714704, + "grad_norm": 1.926778786453728, + "learning_rate": 3.940237601783223e-06, + "loss": 0.9746, + "step": 1172 + }, + { + "epoch": 0.1057852730306173, + "grad_norm": 1.5342333011498537, + "learning_rate": 3.940095771749542e-06, + "loss": 1.013, + "step": 1173 + }, + { + "epoch": 0.10587545655408757, + "grad_norm": 1.8298218505945683, + "learning_rate": 3.939953776176192e-06, + "loss": 1.0711, + "step": 1174 + }, + { + "epoch": 0.10596564007755783, + "grad_norm": 1.4786981548078793, + "learning_rate": 3.939811615075288e-06, + "loss": 1.0404, + "step": 1175 + }, + { + "epoch": 0.1060558236010281, + "grad_norm": 1.6936726879760362, + "learning_rate": 3.9396692884589616e-06, + "loss": 1.0807, + "step": 1176 + }, + { + "epoch": 0.10614600712449836, + "grad_norm": 1.8202774633874999, + "learning_rate": 3.9395267963393565e-06, + "loss": 0.9972, + "step": 1177 + }, + { + "epoch": 0.10623619064796862, + "grad_norm": 1.6734001832250927, + "learning_rate": 3.939384138728631e-06, + "loss": 1.0537, + "step": 1178 + }, + { + "epoch": 0.10632637417143888, + "grad_norm": 1.723462666442784, + "learning_rate": 3.939241315638956e-06, + "loss": 0.9985, + "step": 1179 + }, + { + "epoch": 0.10641655769490914, + "grad_norm": 1.5702801601439003, + "learning_rate": 3.93909832708252e-06, + "loss": 1.0009, + "step": 1180 + }, + { + "epoch": 0.1065067412183794, + "grad_norm": 1.485682102722016, + "learning_rate": 3.938955173071523e-06, + "loss": 1.0179, + "step": 1181 + }, + { + "epoch": 0.10659692474184966, + "grad_norm": 0.6922932339548268, + "learning_rate": 3.938811853618179e-06, + "loss": 0.8127, + "step": 1182 + }, + { + "epoch": 0.10668710826531992, + "grad_norm": 2.011966904824938, + "learning_rate": 3.938668368734717e-06, + "loss": 1.0481, + "step": 1183 + }, + { + "epoch": 0.10677729178879018, + "grad_norm": 1.7209513352078514, + "learning_rate": 3.93852471843338e-06, + "loss": 0.89, + "step": 1184 + }, + { + "epoch": 0.10686747531226044, + "grad_norm": 2.099705212202278, + "learning_rate": 3.9383809027264254e-06, + "loss": 0.9981, + "step": 1185 + }, + { + "epoch": 0.10695765883573072, + "grad_norm": 1.5559878637924163, + "learning_rate": 3.938236921626124e-06, + "loss": 1.0495, + "step": 1186 + }, + { + "epoch": 0.10704784235920098, + "grad_norm": 2.033225522023315, + "learning_rate": 3.938092775144761e-06, + "loss": 1.0198, + "step": 1187 + }, + { + "epoch": 0.10713802588267124, + "grad_norm": 2.2616711020186058, + "learning_rate": 3.9379484632946355e-06, + "loss": 0.9333, + "step": 1188 + }, + { + "epoch": 0.1072282094061415, + "grad_norm": 1.606553428281479, + "learning_rate": 3.937803986088062e-06, + "loss": 1.0584, + "step": 1189 + }, + { + "epoch": 0.10731839292961176, + "grad_norm": 1.9619963321554292, + "learning_rate": 3.937659343537367e-06, + "loss": 0.9712, + "step": 1190 + }, + { + "epoch": 0.10740857645308202, + "grad_norm": 1.7626178079827237, + "learning_rate": 3.937514535654893e-06, + "loss": 1.083, + "step": 1191 + }, + { + "epoch": 0.10749875997655228, + "grad_norm": 2.1753068334820496, + "learning_rate": 3.937369562452996e-06, + "loss": 1.0257, + "step": 1192 + }, + { + "epoch": 0.10758894350002254, + "grad_norm": 1.6752046602390178, + "learning_rate": 3.937224423944044e-06, + "loss": 1.0689, + "step": 1193 + }, + { + "epoch": 0.1076791270234928, + "grad_norm": 1.6886088648734061, + "learning_rate": 3.937079120140423e-06, + "loss": 0.9425, + "step": 1194 + }, + { + "epoch": 0.10776931054696307, + "grad_norm": 1.4964533839231768, + "learning_rate": 3.936933651054531e-06, + "loss": 1.0461, + "step": 1195 + }, + { + "epoch": 0.10785949407043333, + "grad_norm": 1.4531624081626493, + "learning_rate": 3.936788016698779e-06, + "loss": 1.1286, + "step": 1196 + }, + { + "epoch": 0.1079496775939036, + "grad_norm": 1.5018043791652858, + "learning_rate": 3.936642217085594e-06, + "loss": 1.043, + "step": 1197 + }, + { + "epoch": 0.10803986111737386, + "grad_norm": 2.0695109417136006, + "learning_rate": 3.936496252227417e-06, + "loss": 0.9226, + "step": 1198 + }, + { + "epoch": 0.10813004464084412, + "grad_norm": 1.6884993666900991, + "learning_rate": 3.936350122136703e-06, + "loss": 0.9937, + "step": 1199 + }, + { + "epoch": 0.10822022816431438, + "grad_norm": 0.8026089535487312, + "learning_rate": 3.936203826825919e-06, + "loss": 0.834, + "step": 1200 + }, + { + "epoch": 0.10831041168778464, + "grad_norm": 2.3521480401607446, + "learning_rate": 3.9360573663075475e-06, + "loss": 0.9752, + "step": 1201 + }, + { + "epoch": 0.1084005952112549, + "grad_norm": 5.87025584766465, + "learning_rate": 3.935910740594087e-06, + "loss": 1.0444, + "step": 1202 + }, + { + "epoch": 0.10849077873472517, + "grad_norm": 1.6868317305169647, + "learning_rate": 3.935763949698047e-06, + "loss": 1.0258, + "step": 1203 + }, + { + "epoch": 0.10858096225819543, + "grad_norm": 1.6522705206189072, + "learning_rate": 3.935616993631954e-06, + "loss": 1.0767, + "step": 1204 + }, + { + "epoch": 0.10867114578166569, + "grad_norm": 1.6609475796699602, + "learning_rate": 3.935469872408345e-06, + "loss": 0.9997, + "step": 1205 + }, + { + "epoch": 0.10876132930513595, + "grad_norm": 1.4869494148480285, + "learning_rate": 3.935322586039776e-06, + "loss": 1.0395, + "step": 1206 + }, + { + "epoch": 0.10885151282860621, + "grad_norm": 1.8336894455307193, + "learning_rate": 3.935175134538811e-06, + "loss": 0.9499, + "step": 1207 + }, + { + "epoch": 0.10894169635207647, + "grad_norm": 2.1499061697068784, + "learning_rate": 3.935027517918034e-06, + "loss": 0.9582, + "step": 1208 + }, + { + "epoch": 0.10903187987554674, + "grad_norm": 2.171902224143673, + "learning_rate": 3.93487973619004e-06, + "loss": 1.0471, + "step": 1209 + }, + { + "epoch": 0.109122063399017, + "grad_norm": 2.2771460838315525, + "learning_rate": 3.934731789367438e-06, + "loss": 1.0225, + "step": 1210 + }, + { + "epoch": 0.10921224692248727, + "grad_norm": 0.9648815873483712, + "learning_rate": 3.9345836774628505e-06, + "loss": 0.8126, + "step": 1211 + }, + { + "epoch": 0.10930243044595753, + "grad_norm": 1.7726965209098742, + "learning_rate": 3.934435400488917e-06, + "loss": 0.9995, + "step": 1212 + }, + { + "epoch": 0.10939261396942779, + "grad_norm": 1.695510360492628, + "learning_rate": 3.934286958458289e-06, + "loss": 1.047, + "step": 1213 + }, + { + "epoch": 0.10948279749289805, + "grad_norm": 1.5023986393764945, + "learning_rate": 3.934138351383632e-06, + "loss": 1.0761, + "step": 1214 + }, + { + "epoch": 0.10957298101636831, + "grad_norm": 1.6251814080729936, + "learning_rate": 3.933989579277626e-06, + "loss": 0.9482, + "step": 1215 + }, + { + "epoch": 0.10966316453983857, + "grad_norm": 1.5732878779590795, + "learning_rate": 3.933840642152966e-06, + "loss": 1.0032, + "step": 1216 + }, + { + "epoch": 0.10975334806330883, + "grad_norm": 2.0330207735605557, + "learning_rate": 3.933691540022359e-06, + "loss": 1.0411, + "step": 1217 + }, + { + "epoch": 0.10984353158677909, + "grad_norm": 1.796624672079856, + "learning_rate": 3.933542272898527e-06, + "loss": 0.9819, + "step": 1218 + }, + { + "epoch": 0.10993371511024935, + "grad_norm": 0.8987378489133169, + "learning_rate": 3.933392840794207e-06, + "loss": 0.7699, + "step": 1219 + }, + { + "epoch": 0.11002389863371961, + "grad_norm": 1.8347780502967403, + "learning_rate": 3.93324324372215e-06, + "loss": 1.1201, + "step": 1220 + }, + { + "epoch": 0.11011408215718989, + "grad_norm": 1.572012213904613, + "learning_rate": 3.9330934816951185e-06, + "loss": 1.0376, + "step": 1221 + }, + { + "epoch": 0.11020426568066015, + "grad_norm": 1.7261998412781785, + "learning_rate": 3.932943554725893e-06, + "loss": 1.0382, + "step": 1222 + }, + { + "epoch": 0.11029444920413041, + "grad_norm": 0.9303467921253122, + "learning_rate": 3.932793462827265e-06, + "loss": 0.8441, + "step": 1223 + }, + { + "epoch": 0.11038463272760067, + "grad_norm": 1.4537922957508649, + "learning_rate": 3.932643206012041e-06, + "loss": 1.0723, + "step": 1224 + }, + { + "epoch": 0.11047481625107093, + "grad_norm": 1.7303476398875752, + "learning_rate": 3.932492784293043e-06, + "loss": 0.9356, + "step": 1225 + }, + { + "epoch": 0.11056499977454119, + "grad_norm": 1.5608874583122512, + "learning_rate": 3.932342197683104e-06, + "loss": 1.0034, + "step": 1226 + }, + { + "epoch": 0.11065518329801145, + "grad_norm": 1.7407017004348708, + "learning_rate": 3.932191446195075e-06, + "loss": 0.8894, + "step": 1227 + }, + { + "epoch": 0.11074536682148171, + "grad_norm": 1.9248064835615986, + "learning_rate": 3.9320405298418175e-06, + "loss": 1.0163, + "step": 1228 + }, + { + "epoch": 0.11083555034495197, + "grad_norm": 1.7165446611523012, + "learning_rate": 3.9318894486362076e-06, + "loss": 0.9385, + "step": 1229 + }, + { + "epoch": 0.11092573386842224, + "grad_norm": 1.6106124120715528, + "learning_rate": 3.9317382025911395e-06, + "loss": 1.0867, + "step": 1230 + }, + { + "epoch": 0.1110159173918925, + "grad_norm": 1.698977328283363, + "learning_rate": 3.9315867917195145e-06, + "loss": 1.1113, + "step": 1231 + }, + { + "epoch": 0.11110610091536276, + "grad_norm": 1.497023331016313, + "learning_rate": 3.931435216034256e-06, + "loss": 1.0179, + "step": 1232 + }, + { + "epoch": 0.11119628443883303, + "grad_norm": 1.8240877222876273, + "learning_rate": 3.931283475548293e-06, + "loss": 1.0183, + "step": 1233 + }, + { + "epoch": 0.11128646796230329, + "grad_norm": 1.3766978791572515, + "learning_rate": 3.931131570274576e-06, + "loss": 1.0115, + "step": 1234 + }, + { + "epoch": 0.11137665148577355, + "grad_norm": 1.7017004372250004, + "learning_rate": 3.930979500226065e-06, + "loss": 1.0497, + "step": 1235 + }, + { + "epoch": 0.11146683500924381, + "grad_norm": 1.8087109389763487, + "learning_rate": 3.930827265415736e-06, + "loss": 0.8464, + "step": 1236 + }, + { + "epoch": 0.11155701853271408, + "grad_norm": 1.6905272140004861, + "learning_rate": 3.930674865856578e-06, + "loss": 1.0252, + "step": 1237 + }, + { + "epoch": 0.11164720205618434, + "grad_norm": 1.7770693388995709, + "learning_rate": 3.930522301561595e-06, + "loss": 1.0477, + "step": 1238 + }, + { + "epoch": 0.1117373855796546, + "grad_norm": 2.2910592285585785, + "learning_rate": 3.930369572543804e-06, + "loss": 1.0568, + "step": 1239 + }, + { + "epoch": 0.11182756910312486, + "grad_norm": 1.5690169422652378, + "learning_rate": 3.930216678816237e-06, + "loss": 0.9563, + "step": 1240 + }, + { + "epoch": 0.11191775262659512, + "grad_norm": 1.0549263012712744, + "learning_rate": 3.930063620391941e-06, + "loss": 0.8187, + "step": 1241 + }, + { + "epoch": 0.11200793615006538, + "grad_norm": 1.8159307041937123, + "learning_rate": 3.9299103972839735e-06, + "loss": 1.0189, + "step": 1242 + }, + { + "epoch": 0.11209811967353564, + "grad_norm": 1.0203664581378704, + "learning_rate": 3.92975700950541e-06, + "loss": 0.7823, + "step": 1243 + }, + { + "epoch": 0.1121883031970059, + "grad_norm": 1.8644270662716111, + "learning_rate": 3.929603457069338e-06, + "loss": 1.0546, + "step": 1244 + }, + { + "epoch": 0.11227848672047618, + "grad_norm": 1.4479986025819305, + "learning_rate": 3.929449739988859e-06, + "loss": 0.9774, + "step": 1245 + }, + { + "epoch": 0.11236867024394644, + "grad_norm": 1.383833427498724, + "learning_rate": 3.929295858277089e-06, + "loss": 1.0094, + "step": 1246 + }, + { + "epoch": 0.1124588537674167, + "grad_norm": 1.7403222606657158, + "learning_rate": 3.9291418119471585e-06, + "loss": 1.013, + "step": 1247 + }, + { + "epoch": 0.11254903729088696, + "grad_norm": 1.8733907787615296, + "learning_rate": 3.928987601012212e-06, + "loss": 1.0403, + "step": 1248 + }, + { + "epoch": 0.11263922081435722, + "grad_norm": 2.3502695578329558, + "learning_rate": 3.928833225485407e-06, + "loss": 1.0023, + "step": 1249 + }, + { + "epoch": 0.11272940433782748, + "grad_norm": 1.5474405507145033, + "learning_rate": 3.928678685379915e-06, + "loss": 1.0054, + "step": 1250 + }, + { + "epoch": 0.11281958786129774, + "grad_norm": 1.646054249032904, + "learning_rate": 3.928523980708924e-06, + "loss": 1.0591, + "step": 1251 + }, + { + "epoch": 0.112909771384768, + "grad_norm": 1.8437503879352743, + "learning_rate": 3.928369111485632e-06, + "loss": 0.9727, + "step": 1252 + }, + { + "epoch": 0.11299995490823826, + "grad_norm": 1.453709966579012, + "learning_rate": 3.928214077723255e-06, + "loss": 1.0064, + "step": 1253 + }, + { + "epoch": 0.11309013843170852, + "grad_norm": 1.8348868102834366, + "learning_rate": 3.928058879435021e-06, + "loss": 0.994, + "step": 1254 + }, + { + "epoch": 0.11318032195517878, + "grad_norm": 1.7517106686893036, + "learning_rate": 3.9279035166341725e-06, + "loss": 1.0021, + "step": 1255 + }, + { + "epoch": 0.11327050547864904, + "grad_norm": 1.7504112577850428, + "learning_rate": 3.927747989333965e-06, + "loss": 0.9942, + "step": 1256 + }, + { + "epoch": 0.11336068900211932, + "grad_norm": 1.7284965212353567, + "learning_rate": 3.927592297547669e-06, + "loss": 0.9757, + "step": 1257 + }, + { + "epoch": 0.11345087252558958, + "grad_norm": 1.8156343522216711, + "learning_rate": 3.927436441288571e-06, + "loss": 1.0415, + "step": 1258 + }, + { + "epoch": 0.11354105604905984, + "grad_norm": 1.9499256633381068, + "learning_rate": 3.927280420569968e-06, + "loss": 1.0307, + "step": 1259 + }, + { + "epoch": 0.1136312395725301, + "grad_norm": 1.6724005612861221, + "learning_rate": 3.927124235405171e-06, + "loss": 1.0399, + "step": 1260 + }, + { + "epoch": 0.11372142309600036, + "grad_norm": 1.7626831392412063, + "learning_rate": 3.92696788580751e-06, + "loss": 0.8715, + "step": 1261 + }, + { + "epoch": 0.11381160661947062, + "grad_norm": 1.546667836506719, + "learning_rate": 3.9268113717903225e-06, + "loss": 0.8515, + "step": 1262 + }, + { + "epoch": 0.11390179014294088, + "grad_norm": 1.725060616685931, + "learning_rate": 3.926654693366965e-06, + "loss": 1.0803, + "step": 1263 + }, + { + "epoch": 0.11399197366641114, + "grad_norm": 2.0461236870706885, + "learning_rate": 3.926497850550805e-06, + "loss": 1.0235, + "step": 1264 + }, + { + "epoch": 0.1140821571898814, + "grad_norm": 1.87332650842928, + "learning_rate": 3.926340843355226e-06, + "loss": 1.0082, + "step": 1265 + }, + { + "epoch": 0.11417234071335167, + "grad_norm": 1.733530268424402, + "learning_rate": 3.926183671793625e-06, + "loss": 1.0285, + "step": 1266 + }, + { + "epoch": 0.11426252423682193, + "grad_norm": 1.6080860049890144, + "learning_rate": 3.926026335879412e-06, + "loss": 0.9822, + "step": 1267 + }, + { + "epoch": 0.11435270776029219, + "grad_norm": 0.763181094545652, + "learning_rate": 3.925868835626012e-06, + "loss": 0.7535, + "step": 1268 + }, + { + "epoch": 0.11444289128376246, + "grad_norm": 1.9279201305988436, + "learning_rate": 3.925711171046864e-06, + "loss": 1.0201, + "step": 1269 + }, + { + "epoch": 0.11453307480723272, + "grad_norm": 1.4725897088435873, + "learning_rate": 3.925553342155421e-06, + "loss": 0.974, + "step": 1270 + }, + { + "epoch": 0.11462325833070298, + "grad_norm": 0.8242533888927951, + "learning_rate": 3.9253953489651485e-06, + "loss": 0.8287, + "step": 1271 + }, + { + "epoch": 0.11471344185417325, + "grad_norm": 0.7485712270388025, + "learning_rate": 3.925237191489529e-06, + "loss": 0.8029, + "step": 1272 + }, + { + "epoch": 0.1148036253776435, + "grad_norm": 1.7521277843495946, + "learning_rate": 3.925078869742056e-06, + "loss": 0.9839, + "step": 1273 + }, + { + "epoch": 0.11489380890111377, + "grad_norm": 1.6266964346996868, + "learning_rate": 3.92492038373624e-06, + "loss": 0.9407, + "step": 1274 + }, + { + "epoch": 0.11498399242458403, + "grad_norm": 1.6314823262890874, + "learning_rate": 3.924761733485602e-06, + "loss": 1.0658, + "step": 1275 + }, + { + "epoch": 0.11507417594805429, + "grad_norm": 1.4766538077082298, + "learning_rate": 3.92460291900368e-06, + "loss": 1.0082, + "step": 1276 + }, + { + "epoch": 0.11516435947152455, + "grad_norm": 1.568555617732961, + "learning_rate": 3.924443940304025e-06, + "loss": 1.0315, + "step": 1277 + }, + { + "epoch": 0.11525454299499481, + "grad_norm": 1.696041811688571, + "learning_rate": 3.924284797400202e-06, + "loss": 1.0561, + "step": 1278 + }, + { + "epoch": 0.11534472651846507, + "grad_norm": 2.046033613523604, + "learning_rate": 3.924125490305789e-06, + "loss": 1.0356, + "step": 1279 + }, + { + "epoch": 0.11543491004193533, + "grad_norm": 1.4883754430084453, + "learning_rate": 3.923966019034381e-06, + "loss": 0.99, + "step": 1280 + }, + { + "epoch": 0.1155250935654056, + "grad_norm": 1.6390701191945989, + "learning_rate": 3.923806383599583e-06, + "loss": 1.0188, + "step": 1281 + }, + { + "epoch": 0.11561527708887587, + "grad_norm": 2.0377972064287477, + "learning_rate": 3.923646584015017e-06, + "loss": 0.9777, + "step": 1282 + }, + { + "epoch": 0.11570546061234613, + "grad_norm": 1.8069288858924375, + "learning_rate": 3.923486620294316e-06, + "loss": 1.0084, + "step": 1283 + }, + { + "epoch": 0.11579564413581639, + "grad_norm": 1.6165563116392083, + "learning_rate": 3.923326492451132e-06, + "loss": 1.0211, + "step": 1284 + }, + { + "epoch": 0.11588582765928665, + "grad_norm": 1.5715317429077513, + "learning_rate": 3.923166200499125e-06, + "loss": 0.9874, + "step": 1285 + }, + { + "epoch": 0.11597601118275691, + "grad_norm": 1.486526455974509, + "learning_rate": 3.923005744451975e-06, + "loss": 0.9819, + "step": 1286 + }, + { + "epoch": 0.11606619470622717, + "grad_norm": 1.9036710209534804, + "learning_rate": 3.9228451243233715e-06, + "loss": 0.9617, + "step": 1287 + }, + { + "epoch": 0.11615637822969743, + "grad_norm": 1.5572328673354352, + "learning_rate": 3.9226843401270195e-06, + "loss": 1.0395, + "step": 1288 + }, + { + "epoch": 0.1162465617531677, + "grad_norm": 2.2590131372253004, + "learning_rate": 3.9225233918766376e-06, + "loss": 1.0135, + "step": 1289 + }, + { + "epoch": 0.11633674527663795, + "grad_norm": 1.4057028235725246, + "learning_rate": 3.92236227958596e-06, + "loss": 1.0248, + "step": 1290 + }, + { + "epoch": 0.11642692880010821, + "grad_norm": 1.950592168735197, + "learning_rate": 3.922201003268731e-06, + "loss": 1.0306, + "step": 1291 + }, + { + "epoch": 0.11651711232357848, + "grad_norm": 0.794077769511359, + "learning_rate": 3.922039562938715e-06, + "loss": 0.8294, + "step": 1292 + }, + { + "epoch": 0.11660729584704875, + "grad_norm": 1.5888155675004378, + "learning_rate": 3.921877958609685e-06, + "loss": 1.0612, + "step": 1293 + }, + { + "epoch": 0.11669747937051901, + "grad_norm": 0.838950742528232, + "learning_rate": 3.921716190295431e-06, + "loss": 0.8934, + "step": 1294 + }, + { + "epoch": 0.11678766289398927, + "grad_norm": 1.7022823383936443, + "learning_rate": 3.921554258009755e-06, + "loss": 0.9592, + "step": 1295 + }, + { + "epoch": 0.11687784641745953, + "grad_norm": 25.819154191114226, + "learning_rate": 3.921392161766474e-06, + "loss": 0.9979, + "step": 1296 + }, + { + "epoch": 0.1169680299409298, + "grad_norm": 2.597746103209847, + "learning_rate": 3.92122990157942e-06, + "loss": 0.9884, + "step": 1297 + }, + { + "epoch": 0.11705821346440005, + "grad_norm": 1.7253140039338954, + "learning_rate": 3.921067477462437e-06, + "loss": 1.0524, + "step": 1298 + }, + { + "epoch": 0.11714839698787032, + "grad_norm": 1.8298406782784917, + "learning_rate": 3.920904889429385e-06, + "loss": 1.0614, + "step": 1299 + }, + { + "epoch": 0.11723858051134058, + "grad_norm": 1.7585052481522285, + "learning_rate": 3.920742137494135e-06, + "loss": 1.0495, + "step": 1300 + }, + { + "epoch": 0.11732876403481084, + "grad_norm": 1.5372331465217424, + "learning_rate": 3.920579221670575e-06, + "loss": 1.0664, + "step": 1301 + }, + { + "epoch": 0.1174189475582811, + "grad_norm": 1.7615187165573172, + "learning_rate": 3.920416141972606e-06, + "loss": 1.0176, + "step": 1302 + }, + { + "epoch": 0.11750913108175136, + "grad_norm": 1.7985923190057753, + "learning_rate": 3.920252898414143e-06, + "loss": 1.0062, + "step": 1303 + }, + { + "epoch": 0.11759931460522162, + "grad_norm": 1.8836394587270762, + "learning_rate": 3.920089491009114e-06, + "loss": 1.0238, + "step": 1304 + }, + { + "epoch": 0.1176894981286919, + "grad_norm": 1.5262160565025185, + "learning_rate": 3.919925919771463e-06, + "loss": 1.095, + "step": 1305 + }, + { + "epoch": 0.11777968165216215, + "grad_norm": 1.755238186575762, + "learning_rate": 3.919762184715146e-06, + "loss": 0.9246, + "step": 1306 + }, + { + "epoch": 0.11786986517563242, + "grad_norm": 1.786480301999029, + "learning_rate": 3.919598285854134e-06, + "loss": 1.0039, + "step": 1307 + }, + { + "epoch": 0.11796004869910268, + "grad_norm": 1.8049703929512368, + "learning_rate": 3.919434223202411e-06, + "loss": 1.0675, + "step": 1308 + }, + { + "epoch": 0.11805023222257294, + "grad_norm": 2.034714429346408, + "learning_rate": 3.919269996773977e-06, + "loss": 1.0651, + "step": 1309 + }, + { + "epoch": 0.1181404157460432, + "grad_norm": 1.4015737560760078, + "learning_rate": 3.919105606582844e-06, + "loss": 1.0228, + "step": 1310 + }, + { + "epoch": 0.11823059926951346, + "grad_norm": 1.690733108307524, + "learning_rate": 3.918941052643039e-06, + "loss": 0.977, + "step": 1311 + }, + { + "epoch": 0.11832078279298372, + "grad_norm": 1.4229853559961594, + "learning_rate": 3.918776334968602e-06, + "loss": 1.0235, + "step": 1312 + }, + { + "epoch": 0.11841096631645398, + "grad_norm": 1.7856955350163886, + "learning_rate": 3.918611453573589e-06, + "loss": 1.115, + "step": 1313 + }, + { + "epoch": 0.11850114983992424, + "grad_norm": 2.022837664092449, + "learning_rate": 3.918446408472066e-06, + "loss": 0.9452, + "step": 1314 + }, + { + "epoch": 0.1185913333633945, + "grad_norm": 1.6046117272543878, + "learning_rate": 3.918281199678119e-06, + "loss": 1.0482, + "step": 1315 + }, + { + "epoch": 0.11868151688686476, + "grad_norm": 0.8915825431274236, + "learning_rate": 3.9181158272058414e-06, + "loss": 0.812, + "step": 1316 + }, + { + "epoch": 0.11877170041033504, + "grad_norm": 1.5866535994181303, + "learning_rate": 3.9179502910693455e-06, + "loss": 1.0383, + "step": 1317 + }, + { + "epoch": 0.1188618839338053, + "grad_norm": 1.5743511892034798, + "learning_rate": 3.917784591282756e-06, + "loss": 0.9684, + "step": 1318 + }, + { + "epoch": 0.11895206745727556, + "grad_norm": 1.5816668055339682, + "learning_rate": 3.9176187278602105e-06, + "loss": 1.0376, + "step": 1319 + }, + { + "epoch": 0.11904225098074582, + "grad_norm": 1.6774443085887376, + "learning_rate": 3.9174527008158606e-06, + "loss": 0.952, + "step": 1320 + }, + { + "epoch": 0.11913243450421608, + "grad_norm": 2.00249718693263, + "learning_rate": 3.917286510163874e-06, + "loss": 0.9298, + "step": 1321 + }, + { + "epoch": 0.11922261802768634, + "grad_norm": 1.1848574901147833, + "learning_rate": 3.917120155918431e-06, + "loss": 0.9838, + "step": 1322 + }, + { + "epoch": 0.1193128015511566, + "grad_norm": 1.0932796011685102, + "learning_rate": 3.916953638093725e-06, + "loss": 0.8432, + "step": 1323 + }, + { + "epoch": 0.11940298507462686, + "grad_norm": 2.1020913291857215, + "learning_rate": 3.916786956703964e-06, + "loss": 1.0037, + "step": 1324 + }, + { + "epoch": 0.11949316859809712, + "grad_norm": 2.1970615139572356, + "learning_rate": 3.916620111763372e-06, + "loss": 1.0026, + "step": 1325 + }, + { + "epoch": 0.11958335212156739, + "grad_norm": 1.722565386877869, + "learning_rate": 3.916453103286183e-06, + "loss": 1.069, + "step": 1326 + }, + { + "epoch": 0.11967353564503765, + "grad_norm": 1.9131001447117946, + "learning_rate": 3.916285931286648e-06, + "loss": 1.0547, + "step": 1327 + }, + { + "epoch": 0.11976371916850792, + "grad_norm": 1.566419339779316, + "learning_rate": 3.916118595779031e-06, + "loss": 0.9557, + "step": 1328 + }, + { + "epoch": 0.11985390269197818, + "grad_norm": 1.571658264749884, + "learning_rate": 3.915951096777611e-06, + "loss": 0.9803, + "step": 1329 + }, + { + "epoch": 0.11994408621544844, + "grad_norm": 1.5034070263892143, + "learning_rate": 3.915783434296678e-06, + "loss": 0.9539, + "step": 1330 + }, + { + "epoch": 0.1200342697389187, + "grad_norm": 1.891708630952012, + "learning_rate": 3.91561560835054e-06, + "loss": 0.9912, + "step": 1331 + }, + { + "epoch": 0.12012445326238896, + "grad_norm": 1.5408036505627192, + "learning_rate": 3.915447618953515e-06, + "loss": 1.0275, + "step": 1332 + }, + { + "epoch": 0.12021463678585922, + "grad_norm": 1.6935937140737798, + "learning_rate": 3.915279466119937e-06, + "loss": 0.9575, + "step": 1333 + }, + { + "epoch": 0.12030482030932949, + "grad_norm": 0.8757565497421066, + "learning_rate": 3.9151111498641546e-06, + "loss": 0.8274, + "step": 1334 + }, + { + "epoch": 0.12039500383279975, + "grad_norm": 2.3431970579840624, + "learning_rate": 3.914942670200529e-06, + "loss": 0.8797, + "step": 1335 + }, + { + "epoch": 0.12048518735627001, + "grad_norm": 2.3233913968610223, + "learning_rate": 3.914774027143436e-06, + "loss": 1.0642, + "step": 1336 + }, + { + "epoch": 0.12057537087974027, + "grad_norm": 1.7659204286510615, + "learning_rate": 3.914605220707265e-06, + "loss": 0.9949, + "step": 1337 + }, + { + "epoch": 0.12066555440321053, + "grad_norm": 1.7918491566300905, + "learning_rate": 3.9144362509064194e-06, + "loss": 1.1559, + "step": 1338 + }, + { + "epoch": 0.12075573792668079, + "grad_norm": 1.6832851472259054, + "learning_rate": 3.914267117755317e-06, + "loss": 1.0715, + "step": 1339 + }, + { + "epoch": 0.12084592145015106, + "grad_norm": 1.5472622781556993, + "learning_rate": 3.914097821268389e-06, + "loss": 0.971, + "step": 1340 + }, + { + "epoch": 0.12093610497362133, + "grad_norm": 1.3501567696569357, + "learning_rate": 3.913928361460081e-06, + "loss": 1.021, + "step": 1341 + }, + { + "epoch": 0.12102628849709159, + "grad_norm": 1.5699703925675854, + "learning_rate": 3.913758738344851e-06, + "loss": 0.9971, + "step": 1342 + }, + { + "epoch": 0.12111647202056185, + "grad_norm": 0.8785367061651949, + "learning_rate": 3.913588951937174e-06, + "loss": 0.8041, + "step": 1343 + }, + { + "epoch": 0.12120665554403211, + "grad_norm": 0.7733242067532933, + "learning_rate": 3.9134190022515355e-06, + "loss": 0.8037, + "step": 1344 + }, + { + "epoch": 0.12129683906750237, + "grad_norm": 1.8366553668432068, + "learning_rate": 3.913248889302438e-06, + "loss": 0.9742, + "step": 1345 + }, + { + "epoch": 0.12138702259097263, + "grad_norm": 1.694416494955439, + "learning_rate": 3.913078613104395e-06, + "loss": 0.942, + "step": 1346 + }, + { + "epoch": 0.12147720611444289, + "grad_norm": 2.1515879333810184, + "learning_rate": 3.912908173671936e-06, + "loss": 0.8897, + "step": 1347 + }, + { + "epoch": 0.12156738963791315, + "grad_norm": 3.0435701101671584, + "learning_rate": 3.9127375710196044e-06, + "loss": 0.9358, + "step": 1348 + }, + { + "epoch": 0.12165757316138341, + "grad_norm": 1.7640446705081876, + "learning_rate": 3.912566805161957e-06, + "loss": 0.9509, + "step": 1349 + }, + { + "epoch": 0.12174775668485367, + "grad_norm": 3.4104936395688457, + "learning_rate": 3.912395876113564e-06, + "loss": 1.0256, + "step": 1350 + }, + { + "epoch": 0.12183794020832393, + "grad_norm": 1.4367049133065952, + "learning_rate": 3.912224783889009e-06, + "loss": 1.0105, + "step": 1351 + }, + { + "epoch": 0.12192812373179421, + "grad_norm": 1.8260150046468497, + "learning_rate": 3.912053528502892e-06, + "loss": 1.0217, + "step": 1352 + }, + { + "epoch": 0.12201830725526447, + "grad_norm": 1.4760978007854089, + "learning_rate": 3.911882109969825e-06, + "loss": 0.9345, + "step": 1353 + }, + { + "epoch": 0.12210849077873473, + "grad_norm": 1.8599503212108748, + "learning_rate": 3.911710528304435e-06, + "loss": 1.014, + "step": 1354 + }, + { + "epoch": 0.12219867430220499, + "grad_norm": 2.200106886954786, + "learning_rate": 3.911538783521361e-06, + "loss": 0.9542, + "step": 1355 + }, + { + "epoch": 0.12228885782567525, + "grad_norm": 1.6551875899567328, + "learning_rate": 3.9113668756352575e-06, + "loss": 1.0189, + "step": 1356 + }, + { + "epoch": 0.12237904134914551, + "grad_norm": 1.4748931813758976, + "learning_rate": 3.911194804660793e-06, + "loss": 0.9335, + "step": 1357 + }, + { + "epoch": 0.12246922487261577, + "grad_norm": 1.6369555900575925, + "learning_rate": 3.91102257061265e-06, + "loss": 1.0939, + "step": 1358 + }, + { + "epoch": 0.12255940839608603, + "grad_norm": 1.574625951764409, + "learning_rate": 3.910850173505524e-06, + "loss": 1.0491, + "step": 1359 + }, + { + "epoch": 0.1226495919195563, + "grad_norm": 1.8432344265363598, + "learning_rate": 3.9106776133541255e-06, + "loss": 1.0841, + "step": 1360 + }, + { + "epoch": 0.12273977544302656, + "grad_norm": 2.875009122087685, + "learning_rate": 3.9105048901731766e-06, + "loss": 0.9684, + "step": 1361 + }, + { + "epoch": 0.12282995896649682, + "grad_norm": 1.6598892535759615, + "learning_rate": 3.9103320039774165e-06, + "loss": 0.9969, + "step": 1362 + }, + { + "epoch": 0.12292014248996708, + "grad_norm": 1.7250299340567177, + "learning_rate": 3.9101589547815965e-06, + "loss": 1.0668, + "step": 1363 + }, + { + "epoch": 0.12301032601343735, + "grad_norm": 1.3806675800683945, + "learning_rate": 3.909985742600482e-06, + "loss": 0.9959, + "step": 1364 + }, + { + "epoch": 0.12310050953690761, + "grad_norm": 2.0445220468945884, + "learning_rate": 3.909812367448852e-06, + "loss": 0.9687, + "step": 1365 + }, + { + "epoch": 0.12319069306037787, + "grad_norm": 2.10128169683165, + "learning_rate": 3.909638829341501e-06, + "loss": 0.9871, + "step": 1366 + }, + { + "epoch": 0.12328087658384813, + "grad_norm": 1.5851115815029613, + "learning_rate": 3.909465128293234e-06, + "loss": 1.0656, + "step": 1367 + }, + { + "epoch": 0.1233710601073184, + "grad_norm": 1.7893433329624482, + "learning_rate": 3.9092912643188745e-06, + "loss": 1.066, + "step": 1368 + }, + { + "epoch": 0.12346124363078866, + "grad_norm": 1.7309257644390477, + "learning_rate": 3.909117237433256e-06, + "loss": 1.0026, + "step": 1369 + }, + { + "epoch": 0.12355142715425892, + "grad_norm": 1.8481750405135324, + "learning_rate": 3.908943047651229e-06, + "loss": 1.0526, + "step": 1370 + }, + { + "epoch": 0.12364161067772918, + "grad_norm": 2.638865680480983, + "learning_rate": 3.908768694987655e-06, + "loss": 0.9956, + "step": 1371 + }, + { + "epoch": 0.12373179420119944, + "grad_norm": 2.160247083413206, + "learning_rate": 3.908594179457411e-06, + "loss": 0.9346, + "step": 1372 + }, + { + "epoch": 0.1238219777246697, + "grad_norm": 1.7679863052743983, + "learning_rate": 3.908419501075388e-06, + "loss": 1.0061, + "step": 1373 + }, + { + "epoch": 0.12391216124813996, + "grad_norm": 1.566198625384173, + "learning_rate": 3.90824465985649e-06, + "loss": 1.0838, + "step": 1374 + }, + { + "epoch": 0.12400234477161022, + "grad_norm": 1.8423402213353273, + "learning_rate": 3.908069655815636e-06, + "loss": 1.047, + "step": 1375 + }, + { + "epoch": 0.1240925282950805, + "grad_norm": 2.170655271679886, + "learning_rate": 3.907894488967758e-06, + "loss": 0.9459, + "step": 1376 + }, + { + "epoch": 0.12418271181855076, + "grad_norm": 1.687176850072093, + "learning_rate": 3.9077191593278005e-06, + "loss": 1.0634, + "step": 1377 + }, + { + "epoch": 0.12427289534202102, + "grad_norm": 0.8787161049200173, + "learning_rate": 3.9075436669107265e-06, + "loss": 0.7808, + "step": 1378 + }, + { + "epoch": 0.12436307886549128, + "grad_norm": 1.8241251690731886, + "learning_rate": 3.90736801173151e-06, + "loss": 1.0156, + "step": 1379 + }, + { + "epoch": 0.12445326238896154, + "grad_norm": 1.4954117697168532, + "learning_rate": 3.907192193805136e-06, + "loss": 1.0191, + "step": 1380 + }, + { + "epoch": 0.1245434459124318, + "grad_norm": 1.854984517405591, + "learning_rate": 3.907016213146608e-06, + "loss": 1.0619, + "step": 1381 + }, + { + "epoch": 0.12463362943590206, + "grad_norm": 1.873376715837674, + "learning_rate": 3.906840069770942e-06, + "loss": 1.1213, + "step": 1382 + }, + { + "epoch": 0.12472381295937232, + "grad_norm": 1.4767183897691851, + "learning_rate": 3.906663763693167e-06, + "loss": 1.0394, + "step": 1383 + }, + { + "epoch": 0.12481399648284258, + "grad_norm": 1.687658655690973, + "learning_rate": 3.906487294928327e-06, + "loss": 0.9882, + "step": 1384 + }, + { + "epoch": 0.12490418000631284, + "grad_norm": 1.6439018585491598, + "learning_rate": 3.906310663491478e-06, + "loss": 1.0068, + "step": 1385 + }, + { + "epoch": 0.1249943635297831, + "grad_norm": 1.6574639244238776, + "learning_rate": 3.906133869397692e-06, + "loss": 1.0617, + "step": 1386 + }, + { + "epoch": 0.12508454705325336, + "grad_norm": 1.764018180026876, + "learning_rate": 3.905956912662054e-06, + "loss": 0.9766, + "step": 1387 + }, + { + "epoch": 0.12517473057672363, + "grad_norm": 2.1710283564012, + "learning_rate": 3.905779793299662e-06, + "loss": 1.0713, + "step": 1388 + }, + { + "epoch": 0.12526491410019389, + "grad_norm": 1.6798660538272894, + "learning_rate": 3.905602511325631e-06, + "loss": 1.0366, + "step": 1389 + }, + { + "epoch": 0.12535509762366415, + "grad_norm": 1.7698694710097216, + "learning_rate": 3.905425066755086e-06, + "loss": 0.9675, + "step": 1390 + }, + { + "epoch": 0.1254452811471344, + "grad_norm": 1.4754817176030046, + "learning_rate": 3.905247459603168e-06, + "loss": 0.9667, + "step": 1391 + }, + { + "epoch": 0.12553546467060467, + "grad_norm": 5.672276170062852, + "learning_rate": 3.905069689885031e-06, + "loss": 0.8886, + "step": 1392 + }, + { + "epoch": 0.12562564819407493, + "grad_norm": 1.6372765657998898, + "learning_rate": 3.904891757615843e-06, + "loss": 0.8889, + "step": 1393 + }, + { + "epoch": 0.12571583171754522, + "grad_norm": 0.7768700848252825, + "learning_rate": 3.9047136628107874e-06, + "loss": 0.766, + "step": 1394 + }, + { + "epoch": 0.12580601524101548, + "grad_norm": 1.5764877588950454, + "learning_rate": 3.904535405485059e-06, + "loss": 1.139, + "step": 1395 + }, + { + "epoch": 0.12589619876448574, + "grad_norm": 1.6937763620274633, + "learning_rate": 3.90435698565387e-06, + "loss": 0.9549, + "step": 1396 + }, + { + "epoch": 0.125986382287956, + "grad_norm": 1.5943497669910434, + "learning_rate": 3.904178403332441e-06, + "loss": 0.9726, + "step": 1397 + }, + { + "epoch": 0.12607656581142626, + "grad_norm": 2.622567912077911, + "learning_rate": 3.903999658536012e-06, + "loss": 1.0959, + "step": 1398 + }, + { + "epoch": 0.12616674933489652, + "grad_norm": 2.5117909849704763, + "learning_rate": 3.903820751279833e-06, + "loss": 1.007, + "step": 1399 + }, + { + "epoch": 0.12625693285836678, + "grad_norm": 2.075722704123863, + "learning_rate": 3.90364168157917e-06, + "loss": 1.0384, + "step": 1400 + }, + { + "epoch": 0.12634711638183704, + "grad_norm": 2.7576611182834334, + "learning_rate": 3.903462449449302e-06, + "loss": 1.036, + "step": 1401 + }, + { + "epoch": 0.1264372999053073, + "grad_norm": 1.5866314351687583, + "learning_rate": 3.903283054905522e-06, + "loss": 0.9759, + "step": 1402 + }, + { + "epoch": 0.12652748342877757, + "grad_norm": 1.8546926059668685, + "learning_rate": 3.9031034979631385e-06, + "loss": 0.9195, + "step": 1403 + }, + { + "epoch": 0.12661766695224783, + "grad_norm": 1.7237200922926803, + "learning_rate": 3.902923778637469e-06, + "loss": 1.041, + "step": 1404 + }, + { + "epoch": 0.1267078504757181, + "grad_norm": 1.6256961798419876, + "learning_rate": 3.902743896943852e-06, + "loss": 0.94, + "step": 1405 + }, + { + "epoch": 0.12679803399918835, + "grad_norm": 0.7888551949602174, + "learning_rate": 3.902563852897633e-06, + "loss": 0.8353, + "step": 1406 + }, + { + "epoch": 0.1268882175226586, + "grad_norm": 0.7126150774463172, + "learning_rate": 3.9023836465141755e-06, + "loss": 0.8084, + "step": 1407 + }, + { + "epoch": 0.12697840104612887, + "grad_norm": 1.8921217865247215, + "learning_rate": 3.902203277808856e-06, + "loss": 1.0526, + "step": 1408 + }, + { + "epoch": 0.12706858456959913, + "grad_norm": 3.1791190874198265, + "learning_rate": 3.902022746797064e-06, + "loss": 1.0616, + "step": 1409 + }, + { + "epoch": 0.1271587680930694, + "grad_norm": 2.398224364512016, + "learning_rate": 3.9018420534942035e-06, + "loss": 1.1182, + "step": 1410 + }, + { + "epoch": 0.12724895161653965, + "grad_norm": 3.097415640531235, + "learning_rate": 3.9016611979156935e-06, + "loss": 1.0033, + "step": 1411 + }, + { + "epoch": 0.1273391351400099, + "grad_norm": 2.2034788456285037, + "learning_rate": 3.9014801800769635e-06, + "loss": 0.905, + "step": 1412 + }, + { + "epoch": 0.12742931866348017, + "grad_norm": 2.060952617598494, + "learning_rate": 3.901298999993459e-06, + "loss": 1.0311, + "step": 1413 + }, + { + "epoch": 0.12751950218695043, + "grad_norm": 1.789279708003788, + "learning_rate": 3.901117657680642e-06, + "loss": 0.9364, + "step": 1414 + }, + { + "epoch": 0.1276096857104207, + "grad_norm": 1.8480482270359406, + "learning_rate": 3.900936153153982e-06, + "loss": 0.9371, + "step": 1415 + }, + { + "epoch": 0.12769986923389096, + "grad_norm": 1.0807619327342621, + "learning_rate": 3.900754486428968e-06, + "loss": 0.7628, + "step": 1416 + }, + { + "epoch": 0.12779005275736122, + "grad_norm": 2.2265591805416594, + "learning_rate": 3.900572657521102e-06, + "loss": 1.097, + "step": 1417 + }, + { + "epoch": 0.1278802362808315, + "grad_norm": 1.5236277926024535, + "learning_rate": 3.900390666445896e-06, + "loss": 0.9852, + "step": 1418 + }, + { + "epoch": 0.12797041980430177, + "grad_norm": 1.9188490826499864, + "learning_rate": 3.9002085132188795e-06, + "loss": 0.9695, + "step": 1419 + }, + { + "epoch": 0.12806060332777203, + "grad_norm": 1.5034339063938844, + "learning_rate": 3.9000261978555964e-06, + "loss": 1.0497, + "step": 1420 + }, + { + "epoch": 0.1281507868512423, + "grad_norm": 1.7944097442080063, + "learning_rate": 3.8998437203716e-06, + "loss": 1.0711, + "step": 1421 + }, + { + "epoch": 0.12824097037471255, + "grad_norm": 1.5649360739963758, + "learning_rate": 3.899661080782462e-06, + "loss": 1.0331, + "step": 1422 + }, + { + "epoch": 0.1283311538981828, + "grad_norm": 1.789022120390722, + "learning_rate": 3.899478279103767e-06, + "loss": 1.0386, + "step": 1423 + }, + { + "epoch": 0.12842133742165307, + "grad_norm": 1.777396534565603, + "learning_rate": 3.8992953153511105e-06, + "loss": 1.0352, + "step": 1424 + }, + { + "epoch": 0.12851152094512333, + "grad_norm": 1.9235147443642067, + "learning_rate": 3.899112189540106e-06, + "loss": 1.071, + "step": 1425 + }, + { + "epoch": 0.1286017044685936, + "grad_norm": 1.8064765814459771, + "learning_rate": 3.898928901686377e-06, + "loss": 1.0446, + "step": 1426 + }, + { + "epoch": 0.12869188799206385, + "grad_norm": 1.271032578321669, + "learning_rate": 3.898745451805564e-06, + "loss": 0.7685, + "step": 1427 + }, + { + "epoch": 0.1287820715155341, + "grad_norm": 1.7531381807154771, + "learning_rate": 3.898561839913319e-06, + "loss": 0.938, + "step": 1428 + }, + { + "epoch": 0.12887225503900437, + "grad_norm": 2.0085477558865765, + "learning_rate": 3.89837806602531e-06, + "loss": 0.8497, + "step": 1429 + }, + { + "epoch": 0.12896243856247463, + "grad_norm": 1.78179106943845, + "learning_rate": 3.898194130157217e-06, + "loss": 1.0146, + "step": 1430 + }, + { + "epoch": 0.1290526220859449, + "grad_norm": 2.201443943922357, + "learning_rate": 3.8980100323247335e-06, + "loss": 0.9457, + "step": 1431 + }, + { + "epoch": 0.12914280560941516, + "grad_norm": 1.7782584383327376, + "learning_rate": 3.897825772543568e-06, + "loss": 0.9386, + "step": 1432 + }, + { + "epoch": 0.12923298913288542, + "grad_norm": 1.7773919067565986, + "learning_rate": 3.897641350829444e-06, + "loss": 1.0618, + "step": 1433 + }, + { + "epoch": 0.12932317265635568, + "grad_norm": 1.8883482410193735, + "learning_rate": 3.897456767198096e-06, + "loss": 1.0893, + "step": 1434 + }, + { + "epoch": 0.12941335617982594, + "grad_norm": 1.633824814920005, + "learning_rate": 3.897272021665275e-06, + "loss": 0.9458, + "step": 1435 + }, + { + "epoch": 0.1295035397032962, + "grad_norm": 1.4580035608821131, + "learning_rate": 3.897087114246743e-06, + "loss": 1.1144, + "step": 1436 + }, + { + "epoch": 0.12959372322676646, + "grad_norm": 2.144789494319913, + "learning_rate": 3.896902044958279e-06, + "loss": 1.1237, + "step": 1437 + }, + { + "epoch": 0.12968390675023672, + "grad_norm": 1.8336748830937937, + "learning_rate": 3.896716813815672e-06, + "loss": 1.0637, + "step": 1438 + }, + { + "epoch": 0.12977409027370698, + "grad_norm": 1.635664716565623, + "learning_rate": 3.896531420834728e-06, + "loss": 0.9538, + "step": 1439 + }, + { + "epoch": 0.12986427379717724, + "grad_norm": 1.8731609862634857, + "learning_rate": 3.896345866031266e-06, + "loss": 0.9175, + "step": 1440 + }, + { + "epoch": 0.1299544573206475, + "grad_norm": 1.749316899628517, + "learning_rate": 3.896160149421119e-06, + "loss": 1.0673, + "step": 1441 + }, + { + "epoch": 0.1300446408441178, + "grad_norm": 2.6378929857841964, + "learning_rate": 3.8959742710201314e-06, + "loss": 1.0767, + "step": 1442 + }, + { + "epoch": 0.13013482436758805, + "grad_norm": 2.308550580907192, + "learning_rate": 3.895788230844166e-06, + "loss": 0.9063, + "step": 1443 + }, + { + "epoch": 0.13022500789105831, + "grad_norm": 1.7424669255493204, + "learning_rate": 3.895602028909095e-06, + "loss": 1.1689, + "step": 1444 + }, + { + "epoch": 0.13031519141452858, + "grad_norm": 1.1146866343510644, + "learning_rate": 3.895415665230807e-06, + "loss": 0.8019, + "step": 1445 + }, + { + "epoch": 0.13040537493799884, + "grad_norm": 1.4276947976196266, + "learning_rate": 3.895229139825203e-06, + "loss": 1.0337, + "step": 1446 + }, + { + "epoch": 0.1304955584614691, + "grad_norm": 1.7519014791065528, + "learning_rate": 3.895042452708198e-06, + "loss": 0.9255, + "step": 1447 + }, + { + "epoch": 0.13058574198493936, + "grad_norm": 1.5161937157860346, + "learning_rate": 3.894855603895723e-06, + "loss": 0.9126, + "step": 1448 + }, + { + "epoch": 0.13067592550840962, + "grad_norm": 1.7146943533433512, + "learning_rate": 3.894668593403718e-06, + "loss": 1.0893, + "step": 1449 + }, + { + "epoch": 0.13076610903187988, + "grad_norm": 1.9381803118044214, + "learning_rate": 3.8944814212481425e-06, + "loss": 1.0129, + "step": 1450 + }, + { + "epoch": 0.13085629255535014, + "grad_norm": 1.718683831068279, + "learning_rate": 3.894294087444966e-06, + "loss": 0.9893, + "step": 1451 + }, + { + "epoch": 0.1309464760788204, + "grad_norm": 1.6039215577700845, + "learning_rate": 3.894106592010173e-06, + "loss": 0.8997, + "step": 1452 + }, + { + "epoch": 0.13103665960229066, + "grad_norm": 1.6051741652304483, + "learning_rate": 3.893918934959762e-06, + "loss": 0.8869, + "step": 1453 + }, + { + "epoch": 0.13112684312576092, + "grad_norm": 2.030195931060264, + "learning_rate": 3.893731116309743e-06, + "loss": 0.9478, + "step": 1454 + }, + { + "epoch": 0.13121702664923118, + "grad_norm": 1.8406227876037877, + "learning_rate": 3.893543136076145e-06, + "loss": 1.0181, + "step": 1455 + }, + { + "epoch": 0.13130721017270144, + "grad_norm": 1.6604339008402136, + "learning_rate": 3.893354994275006e-06, + "loss": 1.0373, + "step": 1456 + }, + { + "epoch": 0.1313973936961717, + "grad_norm": 1.7424080200219163, + "learning_rate": 3.893166690922378e-06, + "loss": 0.9782, + "step": 1457 + }, + { + "epoch": 0.13148757721964197, + "grad_norm": 1.5689350368600623, + "learning_rate": 3.892978226034329e-06, + "loss": 1.0392, + "step": 1458 + }, + { + "epoch": 0.13157776074311223, + "grad_norm": 3.5348232049180135, + "learning_rate": 3.89278959962694e-06, + "loss": 1.0608, + "step": 1459 + }, + { + "epoch": 0.1316679442665825, + "grad_norm": 1.5978531471303337, + "learning_rate": 3.8926008117163056e-06, + "loss": 0.9832, + "step": 1460 + }, + { + "epoch": 0.13175812779005275, + "grad_norm": 1.7243139893155741, + "learning_rate": 3.892411862318535e-06, + "loss": 0.9914, + "step": 1461 + }, + { + "epoch": 0.131848311313523, + "grad_norm": 1.5240218386799973, + "learning_rate": 3.892222751449749e-06, + "loss": 1.0272, + "step": 1462 + }, + { + "epoch": 0.13193849483699327, + "grad_norm": 1.6100743820352013, + "learning_rate": 3.892033479126084e-06, + "loss": 0.9561, + "step": 1463 + }, + { + "epoch": 0.13202867836046353, + "grad_norm": 2.2248527520993764, + "learning_rate": 3.891844045363691e-06, + "loss": 0.9942, + "step": 1464 + }, + { + "epoch": 0.13211886188393382, + "grad_norm": 1.6915078077595587, + "learning_rate": 3.891654450178732e-06, + "loss": 0.9285, + "step": 1465 + }, + { + "epoch": 0.13220904540740408, + "grad_norm": 1.5683466054508173, + "learning_rate": 3.891464693587385e-06, + "loss": 1.0509, + "step": 1466 + }, + { + "epoch": 0.13229922893087434, + "grad_norm": 1.7005244708253553, + "learning_rate": 3.89127477560584e-06, + "loss": 1.0443, + "step": 1467 + }, + { + "epoch": 0.1323894124543446, + "grad_norm": 1.5929160086453884, + "learning_rate": 3.891084696250304e-06, + "loss": 1.0448, + "step": 1468 + }, + { + "epoch": 0.13247959597781486, + "grad_norm": 1.7856349845716568, + "learning_rate": 3.890894455536993e-06, + "loss": 1.0463, + "step": 1469 + }, + { + "epoch": 0.13256977950128512, + "grad_norm": 2.419784825271961, + "learning_rate": 3.890704053482142e-06, + "loss": 1.0536, + "step": 1470 + }, + { + "epoch": 0.13265996302475538, + "grad_norm": 1.816197012058132, + "learning_rate": 3.890513490101995e-06, + "loss": 0.9468, + "step": 1471 + }, + { + "epoch": 0.13275014654822564, + "grad_norm": 0.7576602507564485, + "learning_rate": 3.890322765412814e-06, + "loss": 0.8098, + "step": 1472 + }, + { + "epoch": 0.1328403300716959, + "grad_norm": 1.3326457555471491, + "learning_rate": 3.890131879430871e-06, + "loss": 1.0347, + "step": 1473 + }, + { + "epoch": 0.13293051359516617, + "grad_norm": 1.6435221205877353, + "learning_rate": 3.889940832172454e-06, + "loss": 1.0379, + "step": 1474 + }, + { + "epoch": 0.13302069711863643, + "grad_norm": 1.7886673270153335, + "learning_rate": 3.889749623653864e-06, + "loss": 1.0558, + "step": 1475 + }, + { + "epoch": 0.1331108806421067, + "grad_norm": 1.407901048997951, + "learning_rate": 3.889558253891416e-06, + "loss": 0.9543, + "step": 1476 + }, + { + "epoch": 0.13320106416557695, + "grad_norm": 0.7214045638968808, + "learning_rate": 3.8893667229014385e-06, + "loss": 0.7842, + "step": 1477 + }, + { + "epoch": 0.1332912476890472, + "grad_norm": 2.3905169892687788, + "learning_rate": 3.8891750307002746e-06, + "loss": 0.9112, + "step": 1478 + }, + { + "epoch": 0.13338143121251747, + "grad_norm": 1.4497934523173721, + "learning_rate": 3.888983177304281e-06, + "loss": 0.9401, + "step": 1479 + }, + { + "epoch": 0.13347161473598773, + "grad_norm": 0.9327163723435935, + "learning_rate": 3.888791162729826e-06, + "loss": 0.784, + "step": 1480 + }, + { + "epoch": 0.133561798259458, + "grad_norm": 1.8596709841272223, + "learning_rate": 3.888598986993295e-06, + "loss": 1.0249, + "step": 1481 + }, + { + "epoch": 0.13365198178292825, + "grad_norm": 1.5790295416379894, + "learning_rate": 3.888406650111085e-06, + "loss": 1.0373, + "step": 1482 + }, + { + "epoch": 0.1337421653063985, + "grad_norm": 1.7209261902270745, + "learning_rate": 3.888214152099607e-06, + "loss": 1.0133, + "step": 1483 + }, + { + "epoch": 0.13383234882986877, + "grad_norm": 1.7904593665724917, + "learning_rate": 3.888021492975285e-06, + "loss": 1.0734, + "step": 1484 + }, + { + "epoch": 0.13392253235333904, + "grad_norm": 1.829854033438571, + "learning_rate": 3.88782867275456e-06, + "loss": 1.0596, + "step": 1485 + }, + { + "epoch": 0.1340127158768093, + "grad_norm": 1.4670024783601738, + "learning_rate": 3.8876356914538824e-06, + "loss": 0.9722, + "step": 1486 + }, + { + "epoch": 0.13410289940027956, + "grad_norm": 1.6613437848280286, + "learning_rate": 3.88744254908972e-06, + "loss": 0.962, + "step": 1487 + }, + { + "epoch": 0.13419308292374982, + "grad_norm": 2.048746316474736, + "learning_rate": 3.887249245678552e-06, + "loss": 0.9832, + "step": 1488 + }, + { + "epoch": 0.1342832664472201, + "grad_norm": 1.7769671701415832, + "learning_rate": 3.887055781236872e-06, + "loss": 1.0841, + "step": 1489 + }, + { + "epoch": 0.13437344997069037, + "grad_norm": 1.732060924890094, + "learning_rate": 3.886862155781186e-06, + "loss": 0.8784, + "step": 1490 + }, + { + "epoch": 0.13446363349416063, + "grad_norm": 0.7421614391369245, + "learning_rate": 3.886668369328019e-06, + "loss": 0.8063, + "step": 1491 + }, + { + "epoch": 0.1345538170176309, + "grad_norm": 1.842943047267053, + "learning_rate": 3.886474421893904e-06, + "loss": 1.0329, + "step": 1492 + }, + { + "epoch": 0.13464400054110115, + "grad_norm": 1.2400606769667377, + "learning_rate": 3.886280313495388e-06, + "loss": 0.9305, + "step": 1493 + }, + { + "epoch": 0.1347341840645714, + "grad_norm": 1.9850969578198947, + "learning_rate": 3.886086044149035e-06, + "loss": 0.9741, + "step": 1494 + }, + { + "epoch": 0.13482436758804167, + "grad_norm": 1.4531839830466664, + "learning_rate": 3.885891613871421e-06, + "loss": 1.0138, + "step": 1495 + }, + { + "epoch": 0.13491455111151193, + "grad_norm": 1.7349344674302445, + "learning_rate": 3.885697022679136e-06, + "loss": 1.0559, + "step": 1496 + }, + { + "epoch": 0.1350047346349822, + "grad_norm": 2.687068194808428, + "learning_rate": 3.885502270588784e-06, + "loss": 1.0378, + "step": 1497 + }, + { + "epoch": 0.13509491815845245, + "grad_norm": 2.463399182662463, + "learning_rate": 3.885307357616981e-06, + "loss": 1.0015, + "step": 1498 + }, + { + "epoch": 0.13518510168192271, + "grad_norm": 1.7230517566986219, + "learning_rate": 3.885112283780359e-06, + "loss": 1.0016, + "step": 1499 + }, + { + "epoch": 0.13527528520539298, + "grad_norm": 1.6480028926444361, + "learning_rate": 3.8849170490955624e-06, + "loss": 0.8967, + "step": 1500 + }, + { + "epoch": 0.13536546872886324, + "grad_norm": 1.9915151620910705, + "learning_rate": 3.88472165357925e-06, + "loss": 0.9986, + "step": 1501 + }, + { + "epoch": 0.1354556522523335, + "grad_norm": 1.7368278685355885, + "learning_rate": 3.884526097248093e-06, + "loss": 0.961, + "step": 1502 + }, + { + "epoch": 0.13554583577580376, + "grad_norm": 1.7768399708851133, + "learning_rate": 3.884330380118779e-06, + "loss": 1.0271, + "step": 1503 + }, + { + "epoch": 0.13563601929927402, + "grad_norm": 1.8874011967548385, + "learning_rate": 3.884134502208007e-06, + "loss": 0.9577, + "step": 1504 + }, + { + "epoch": 0.13572620282274428, + "grad_norm": 1.7862345904786108, + "learning_rate": 3.88393846353249e-06, + "loss": 1.1131, + "step": 1505 + }, + { + "epoch": 0.13581638634621454, + "grad_norm": 1.4663531448755618, + "learning_rate": 3.883742264108955e-06, + "loss": 1.0563, + "step": 1506 + }, + { + "epoch": 0.1359065698696848, + "grad_norm": 1.5074216504083224, + "learning_rate": 3.883545903954145e-06, + "loss": 1.0101, + "step": 1507 + }, + { + "epoch": 0.13599675339315506, + "grad_norm": 1.6715113164054805, + "learning_rate": 3.883349383084811e-06, + "loss": 0.9998, + "step": 1508 + }, + { + "epoch": 0.13608693691662532, + "grad_norm": 2.315140737123864, + "learning_rate": 3.883152701517723e-06, + "loss": 0.9242, + "step": 1509 + }, + { + "epoch": 0.13617712044009558, + "grad_norm": 2.1639127713954305, + "learning_rate": 3.882955859269664e-06, + "loss": 0.9969, + "step": 1510 + }, + { + "epoch": 0.13626730396356584, + "grad_norm": 1.4977590192339538, + "learning_rate": 3.882758856357428e-06, + "loss": 1.0141, + "step": 1511 + }, + { + "epoch": 0.1363574874870361, + "grad_norm": 1.8998510804290996, + "learning_rate": 3.882561692797824e-06, + "loss": 0.9968, + "step": 1512 + }, + { + "epoch": 0.1364476710105064, + "grad_norm": 1.5558455864795502, + "learning_rate": 3.882364368607677e-06, + "loss": 1.0044, + "step": 1513 + }, + { + "epoch": 0.13653785453397665, + "grad_norm": 1.6361553524377785, + "learning_rate": 3.8821668838038225e-06, + "loss": 1.0146, + "step": 1514 + }, + { + "epoch": 0.13662803805744692, + "grad_norm": 1.8123466821477932, + "learning_rate": 3.881969238403111e-06, + "loss": 0.9196, + "step": 1515 + }, + { + "epoch": 0.13671822158091718, + "grad_norm": 2.0052694996674103, + "learning_rate": 3.881771432422408e-06, + "loss": 1.1245, + "step": 1516 + }, + { + "epoch": 0.13680840510438744, + "grad_norm": 1.7478192911164483, + "learning_rate": 3.88157346587859e-06, + "loss": 1.0579, + "step": 1517 + }, + { + "epoch": 0.1368985886278577, + "grad_norm": 1.399040347961961, + "learning_rate": 3.881375338788549e-06, + "loss": 0.9849, + "step": 1518 + }, + { + "epoch": 0.13698877215132796, + "grad_norm": 2.5107571910408386, + "learning_rate": 3.88117705116919e-06, + "loss": 1.0234, + "step": 1519 + }, + { + "epoch": 0.13707895567479822, + "grad_norm": 0.9228220455555701, + "learning_rate": 3.880978603037432e-06, + "loss": 0.859, + "step": 1520 + }, + { + "epoch": 0.13716913919826848, + "grad_norm": 1.5859906417888079, + "learning_rate": 3.880779994410209e-06, + "loss": 1.0104, + "step": 1521 + }, + { + "epoch": 0.13725932272173874, + "grad_norm": 2.008910476360873, + "learning_rate": 3.880581225304466e-06, + "loss": 0.8814, + "step": 1522 + }, + { + "epoch": 0.137349506245209, + "grad_norm": 1.7348190289089451, + "learning_rate": 3.880382295737163e-06, + "loss": 1.0298, + "step": 1523 + }, + { + "epoch": 0.13743968976867926, + "grad_norm": 1.9532698310559165, + "learning_rate": 3.880183205725274e-06, + "loss": 0.9737, + "step": 1524 + }, + { + "epoch": 0.13752987329214952, + "grad_norm": 1.9819020158122755, + "learning_rate": 3.879983955285788e-06, + "loss": 0.9421, + "step": 1525 + }, + { + "epoch": 0.13762005681561978, + "grad_norm": 2.2249479094847713, + "learning_rate": 3.879784544435703e-06, + "loss": 0.9849, + "step": 1526 + }, + { + "epoch": 0.13771024033909005, + "grad_norm": 2.1319677894907385, + "learning_rate": 3.879584973192037e-06, + "loss": 0.9515, + "step": 1527 + }, + { + "epoch": 0.1378004238625603, + "grad_norm": 2.4002294152601595, + "learning_rate": 3.8793852415718165e-06, + "loss": 0.9352, + "step": 1528 + }, + { + "epoch": 0.13789060738603057, + "grad_norm": 2.1895384963284608, + "learning_rate": 3.879185349592085e-06, + "loss": 1.0131, + "step": 1529 + }, + { + "epoch": 0.13798079090950083, + "grad_norm": 1.513109538311688, + "learning_rate": 3.878985297269897e-06, + "loss": 1.026, + "step": 1530 + }, + { + "epoch": 0.1380709744329711, + "grad_norm": 1.73481655514228, + "learning_rate": 3.878785084622323e-06, + "loss": 1.0427, + "step": 1531 + }, + { + "epoch": 0.13816115795644135, + "grad_norm": 2.0089892786935217, + "learning_rate": 3.878584711666447e-06, + "loss": 1.1914, + "step": 1532 + }, + { + "epoch": 0.1382513414799116, + "grad_norm": 1.658587551599493, + "learning_rate": 3.8783841784193635e-06, + "loss": 1.0212, + "step": 1533 + }, + { + "epoch": 0.13834152500338187, + "grad_norm": 2.001047098714064, + "learning_rate": 3.8781834848981855e-06, + "loss": 0.9861, + "step": 1534 + }, + { + "epoch": 0.13843170852685213, + "grad_norm": 1.6808854200582442, + "learning_rate": 3.877982631120037e-06, + "loss": 0.9382, + "step": 1535 + }, + { + "epoch": 0.1385218920503224, + "grad_norm": 2.0484438409423453, + "learning_rate": 3.877781617102053e-06, + "loss": 1.0688, + "step": 1536 + }, + { + "epoch": 0.13861207557379268, + "grad_norm": 1.6205574343448164, + "learning_rate": 3.877580442861389e-06, + "loss": 1.0404, + "step": 1537 + }, + { + "epoch": 0.13870225909726294, + "grad_norm": 2.0364878100305024, + "learning_rate": 3.877379108415209e-06, + "loss": 1.0335, + "step": 1538 + }, + { + "epoch": 0.1387924426207332, + "grad_norm": 1.472165701438967, + "learning_rate": 3.8771776137806915e-06, + "loss": 1.0339, + "step": 1539 + }, + { + "epoch": 0.13888262614420346, + "grad_norm": 1.6223333927423893, + "learning_rate": 3.8769759589750295e-06, + "loss": 0.9489, + "step": 1540 + }, + { + "epoch": 0.13897280966767372, + "grad_norm": 1.5320480758155428, + "learning_rate": 3.876774144015429e-06, + "loss": 1.0109, + "step": 1541 + }, + { + "epoch": 0.13906299319114399, + "grad_norm": 1.9413501263187338, + "learning_rate": 3.87657216891911e-06, + "loss": 1.0442, + "step": 1542 + }, + { + "epoch": 0.13915317671461425, + "grad_norm": 2.1883988849654483, + "learning_rate": 3.876370033703307e-06, + "loss": 1.0075, + "step": 1543 + }, + { + "epoch": 0.1392433602380845, + "grad_norm": 1.4393773466580568, + "learning_rate": 3.876167738385265e-06, + "loss": 0.9093, + "step": 1544 + }, + { + "epoch": 0.13933354376155477, + "grad_norm": 2.194195554866431, + "learning_rate": 3.875965282982247e-06, + "loss": 1.0511, + "step": 1545 + }, + { + "epoch": 0.13942372728502503, + "grad_norm": 1.7144694343394342, + "learning_rate": 3.875762667511528e-06, + "loss": 1.0257, + "step": 1546 + }, + { + "epoch": 0.1395139108084953, + "grad_norm": 1.4296077851965654, + "learning_rate": 3.875559891990394e-06, + "loss": 1.0152, + "step": 1547 + }, + { + "epoch": 0.13960409433196555, + "grad_norm": 1.6763478279440354, + "learning_rate": 3.875356956436149e-06, + "loss": 0.9718, + "step": 1548 + }, + { + "epoch": 0.1396942778554358, + "grad_norm": 1.8554351482360392, + "learning_rate": 3.875153860866108e-06, + "loss": 0.9888, + "step": 1549 + }, + { + "epoch": 0.13978446137890607, + "grad_norm": 1.5671190271603768, + "learning_rate": 3.8749506052976e-06, + "loss": 1.0527, + "step": 1550 + }, + { + "epoch": 0.13987464490237633, + "grad_norm": 1.3956278701937213, + "learning_rate": 3.874747189747968e-06, + "loss": 1.0311, + "step": 1551 + }, + { + "epoch": 0.1399648284258466, + "grad_norm": 1.6624728207768622, + "learning_rate": 3.874543614234568e-06, + "loss": 1.019, + "step": 1552 + }, + { + "epoch": 0.14005501194931685, + "grad_norm": 1.6075514304326957, + "learning_rate": 3.874339878774771e-06, + "loss": 0.988, + "step": 1553 + }, + { + "epoch": 0.14014519547278712, + "grad_norm": 1.6129178799455726, + "learning_rate": 3.874135983385961e-06, + "loss": 1.0104, + "step": 1554 + }, + { + "epoch": 0.14023537899625738, + "grad_norm": 1.9155872110992562, + "learning_rate": 3.873931928085535e-06, + "loss": 1.0126, + "step": 1555 + }, + { + "epoch": 0.14032556251972764, + "grad_norm": 1.4790893409110237, + "learning_rate": 3.873727712890904e-06, + "loss": 1.0971, + "step": 1556 + }, + { + "epoch": 0.1404157460431979, + "grad_norm": 1.4970384608076064, + "learning_rate": 3.873523337819493e-06, + "loss": 0.9785, + "step": 1557 + }, + { + "epoch": 0.14050592956666816, + "grad_norm": 1.5045298682937513, + "learning_rate": 3.873318802888739e-06, + "loss": 0.9367, + "step": 1558 + }, + { + "epoch": 0.14059611309013842, + "grad_norm": 0.6754987269431203, + "learning_rate": 3.873114108116097e-06, + "loss": 0.734, + "step": 1559 + }, + { + "epoch": 0.14068629661360868, + "grad_norm": 1.837435107480109, + "learning_rate": 3.872909253519031e-06, + "loss": 1.0973, + "step": 1560 + }, + { + "epoch": 0.14077648013707897, + "grad_norm": 1.9535017947092788, + "learning_rate": 3.8727042391150195e-06, + "loss": 1.0819, + "step": 1561 + }, + { + "epoch": 0.14086666366054923, + "grad_norm": 1.70733500207067, + "learning_rate": 3.872499064921556e-06, + "loss": 0.9515, + "step": 1562 + }, + { + "epoch": 0.1409568471840195, + "grad_norm": 1.7693159296597525, + "learning_rate": 3.872293730956149e-06, + "loss": 1.0258, + "step": 1563 + }, + { + "epoch": 0.14104703070748975, + "grad_norm": 1.715747499522205, + "learning_rate": 3.872088237236316e-06, + "loss": 1.0032, + "step": 1564 + }, + { + "epoch": 0.14113721423096, + "grad_norm": 1.8111450787279746, + "learning_rate": 3.871882583779592e-06, + "loss": 0.915, + "step": 1565 + }, + { + "epoch": 0.14122739775443027, + "grad_norm": 1.5269566506388772, + "learning_rate": 3.871676770603525e-06, + "loss": 1.0116, + "step": 1566 + }, + { + "epoch": 0.14131758127790053, + "grad_norm": 2.7218198435611463, + "learning_rate": 3.871470797725676e-06, + "loss": 1.0042, + "step": 1567 + }, + { + "epoch": 0.1414077648013708, + "grad_norm": 1.6028695332388494, + "learning_rate": 3.8712646651636185e-06, + "loss": 1.0847, + "step": 1568 + }, + { + "epoch": 0.14149794832484106, + "grad_norm": 1.343134273250301, + "learning_rate": 3.871058372934942e-06, + "loss": 1.006, + "step": 1569 + }, + { + "epoch": 0.14158813184831132, + "grad_norm": 1.8965699857173415, + "learning_rate": 3.8708519210572485e-06, + "loss": 0.8415, + "step": 1570 + }, + { + "epoch": 0.14167831537178158, + "grad_norm": 1.6815252142634118, + "learning_rate": 3.870645309548153e-06, + "loss": 0.9438, + "step": 1571 + }, + { + "epoch": 0.14176849889525184, + "grad_norm": 0.6713163803708688, + "learning_rate": 3.870438538425284e-06, + "loss": 0.7475, + "step": 1572 + }, + { + "epoch": 0.1418586824187221, + "grad_norm": 2.507935042670199, + "learning_rate": 3.870231607706287e-06, + "loss": 0.9527, + "step": 1573 + }, + { + "epoch": 0.14194886594219236, + "grad_norm": 1.6651994286775582, + "learning_rate": 3.870024517408817e-06, + "loss": 1.0152, + "step": 1574 + }, + { + "epoch": 0.14203904946566262, + "grad_norm": 1.8700466212412579, + "learning_rate": 3.8698172675505425e-06, + "loss": 0.9299, + "step": 1575 + }, + { + "epoch": 0.14212923298913288, + "grad_norm": 0.8565043969472961, + "learning_rate": 3.86960985814915e-06, + "loss": 0.7613, + "step": 1576 + }, + { + "epoch": 0.14221941651260314, + "grad_norm": 1.4944005403525868, + "learning_rate": 3.869402289222335e-06, + "loss": 0.9364, + "step": 1577 + }, + { + "epoch": 0.1423096000360734, + "grad_norm": 2.191231950162196, + "learning_rate": 3.869194560787808e-06, + "loss": 0.9876, + "step": 1578 + }, + { + "epoch": 0.14239978355954366, + "grad_norm": 1.614747496432147, + "learning_rate": 3.868986672863296e-06, + "loss": 0.9381, + "step": 1579 + }, + { + "epoch": 0.14248996708301392, + "grad_norm": 1.60050058282748, + "learning_rate": 3.868778625466535e-06, + "loss": 1.1092, + "step": 1580 + }, + { + "epoch": 0.14258015060648419, + "grad_norm": 1.7679704599545005, + "learning_rate": 3.868570418615278e-06, + "loss": 0.932, + "step": 1581 + }, + { + "epoch": 0.14267033412995445, + "grad_norm": 0.8571149759072224, + "learning_rate": 3.8683620523272885e-06, + "loss": 0.7701, + "step": 1582 + }, + { + "epoch": 0.1427605176534247, + "grad_norm": 1.7736785321559565, + "learning_rate": 3.8681535266203464e-06, + "loss": 1.1102, + "step": 1583 + }, + { + "epoch": 0.14285070117689497, + "grad_norm": 1.68861712460309, + "learning_rate": 3.867944841512246e-06, + "loss": 1.0308, + "step": 1584 + }, + { + "epoch": 0.14294088470036526, + "grad_norm": 3.9741053456136095, + "learning_rate": 3.867735997020791e-06, + "loss": 1.0164, + "step": 1585 + }, + { + "epoch": 0.14303106822383552, + "grad_norm": 2.474576520814611, + "learning_rate": 3.867526993163802e-06, + "loss": 0.9161, + "step": 1586 + }, + { + "epoch": 0.14312125174730578, + "grad_norm": 1.6426909019046887, + "learning_rate": 3.867317829959113e-06, + "loss": 0.8824, + "step": 1587 + }, + { + "epoch": 0.14321143527077604, + "grad_norm": 1.8446371239567934, + "learning_rate": 3.8671085074245704e-06, + "loss": 0.8077, + "step": 1588 + }, + { + "epoch": 0.1433016187942463, + "grad_norm": 2.1462607714161104, + "learning_rate": 3.866899025578035e-06, + "loss": 1.0721, + "step": 1589 + }, + { + "epoch": 0.14339180231771656, + "grad_norm": 1.4147721506809092, + "learning_rate": 3.86668938443738e-06, + "loss": 0.9745, + "step": 1590 + }, + { + "epoch": 0.14348198584118682, + "grad_norm": 2.111297635432603, + "learning_rate": 3.866479584020495e-06, + "loss": 1.0213, + "step": 1591 + }, + { + "epoch": 0.14357216936465708, + "grad_norm": 1.7508719179506642, + "learning_rate": 3.866269624345279e-06, + "loss": 1.038, + "step": 1592 + }, + { + "epoch": 0.14366235288812734, + "grad_norm": 1.7826251025926434, + "learning_rate": 3.866059505429649e-06, + "loss": 1.0897, + "step": 1593 + }, + { + "epoch": 0.1437525364115976, + "grad_norm": 0.9486255126320398, + "learning_rate": 3.865849227291532e-06, + "loss": 0.7969, + "step": 1594 + }, + { + "epoch": 0.14384271993506786, + "grad_norm": 1.8665757074787719, + "learning_rate": 3.865638789948872e-06, + "loss": 1.0047, + "step": 1595 + }, + { + "epoch": 0.14393290345853813, + "grad_norm": 1.9378432923410982, + "learning_rate": 3.865428193419622e-06, + "loss": 0.9612, + "step": 1596 + }, + { + "epoch": 0.14402308698200839, + "grad_norm": 1.9774222332387892, + "learning_rate": 3.865217437721753e-06, + "loss": 0.9933, + "step": 1597 + }, + { + "epoch": 0.14411327050547865, + "grad_norm": 1.9922987364586666, + "learning_rate": 3.865006522873249e-06, + "loss": 0.9356, + "step": 1598 + }, + { + "epoch": 0.1442034540289489, + "grad_norm": 1.6678421722911834, + "learning_rate": 3.864795448892103e-06, + "loss": 1.0071, + "step": 1599 + }, + { + "epoch": 0.14429363755241917, + "grad_norm": 1.598214721070738, + "learning_rate": 3.864584215796327e-06, + "loss": 1.0549, + "step": 1600 + }, + { + "epoch": 0.14438382107588943, + "grad_norm": 1.7208301787389824, + "learning_rate": 3.8643728236039455e-06, + "loss": 1.0264, + "step": 1601 + }, + { + "epoch": 0.1444740045993597, + "grad_norm": 1.6670468929814326, + "learning_rate": 3.864161272332994e-06, + "loss": 0.9906, + "step": 1602 + }, + { + "epoch": 0.14456418812282995, + "grad_norm": 1.5572848452755685, + "learning_rate": 3.863949562001524e-06, + "loss": 1.0531, + "step": 1603 + }, + { + "epoch": 0.1446543716463002, + "grad_norm": 1.693931192651287, + "learning_rate": 3.8637376926276005e-06, + "loss": 0.9792, + "step": 1604 + }, + { + "epoch": 0.14474455516977047, + "grad_norm": 2.2548146128098576, + "learning_rate": 3.8635256642293e-06, + "loss": 1.0119, + "step": 1605 + }, + { + "epoch": 0.14483473869324073, + "grad_norm": 1.8438551274733108, + "learning_rate": 3.863313476824714e-06, + "loss": 0.985, + "step": 1606 + }, + { + "epoch": 0.144924922216711, + "grad_norm": 1.9973073953863585, + "learning_rate": 3.863101130431948e-06, + "loss": 0.9873, + "step": 1607 + }, + { + "epoch": 0.14501510574018128, + "grad_norm": 1.3637971967437572, + "learning_rate": 3.862888625069121e-06, + "loss": 1.0485, + "step": 1608 + }, + { + "epoch": 0.14510528926365154, + "grad_norm": 1.4219243806082273, + "learning_rate": 3.8626759607543645e-06, + "loss": 0.9675, + "step": 1609 + }, + { + "epoch": 0.1451954727871218, + "grad_norm": 1.9974633461999325, + "learning_rate": 3.862463137505825e-06, + "loss": 0.9607, + "step": 1610 + }, + { + "epoch": 0.14528565631059207, + "grad_norm": 1.5472911698742184, + "learning_rate": 3.862250155341659e-06, + "loss": 0.9645, + "step": 1611 + }, + { + "epoch": 0.14537583983406233, + "grad_norm": 1.6087116244584436, + "learning_rate": 3.862037014280043e-06, + "loss": 1.1317, + "step": 1612 + }, + { + "epoch": 0.1454660233575326, + "grad_norm": 1.6705087087952846, + "learning_rate": 3.861823714339162e-06, + "loss": 0.9988, + "step": 1613 + }, + { + "epoch": 0.14555620688100285, + "grad_norm": 1.4805977395262642, + "learning_rate": 3.861610255537215e-06, + "loss": 1.0102, + "step": 1614 + }, + { + "epoch": 0.1456463904044731, + "grad_norm": 1.8110121177592315, + "learning_rate": 3.8613966378924165e-06, + "loss": 1.0364, + "step": 1615 + }, + { + "epoch": 0.14573657392794337, + "grad_norm": 1.573462823021926, + "learning_rate": 3.861182861422993e-06, + "loss": 0.9484, + "step": 1616 + }, + { + "epoch": 0.14582675745141363, + "grad_norm": 1.4971034375412255, + "learning_rate": 3.860968926147185e-06, + "loss": 1.0751, + "step": 1617 + }, + { + "epoch": 0.1459169409748839, + "grad_norm": 1.6325125532850584, + "learning_rate": 3.860754832083247e-06, + "loss": 0.8393, + "step": 1618 + }, + { + "epoch": 0.14600712449835415, + "grad_norm": 1.7137223663439392, + "learning_rate": 3.8605405792494475e-06, + "loss": 1.0721, + "step": 1619 + }, + { + "epoch": 0.1460973080218244, + "grad_norm": 1.9203681931516874, + "learning_rate": 3.860326167664066e-06, + "loss": 1.0213, + "step": 1620 + }, + { + "epoch": 0.14618749154529467, + "grad_norm": 1.7264263392824675, + "learning_rate": 3.860111597345399e-06, + "loss": 0.9928, + "step": 1621 + }, + { + "epoch": 0.14627767506876493, + "grad_norm": 2.063976106416663, + "learning_rate": 3.859896868311753e-06, + "loss": 0.8573, + "step": 1622 + }, + { + "epoch": 0.1463678585922352, + "grad_norm": 3.946215837757913, + "learning_rate": 3.859681980581452e-06, + "loss": 1.0962, + "step": 1623 + }, + { + "epoch": 0.14645804211570546, + "grad_norm": 1.6771831206336862, + "learning_rate": 3.859466934172829e-06, + "loss": 0.9782, + "step": 1624 + }, + { + "epoch": 0.14654822563917572, + "grad_norm": 1.5856344750962914, + "learning_rate": 3.859251729104235e-06, + "loss": 0.9362, + "step": 1625 + }, + { + "epoch": 0.14663840916264598, + "grad_norm": 1.9485207571246026, + "learning_rate": 3.859036365394031e-06, + "loss": 0.95, + "step": 1626 + }, + { + "epoch": 0.14672859268611624, + "grad_norm": 2.2148270648415287, + "learning_rate": 3.858820843060594e-06, + "loss": 0.9915, + "step": 1627 + }, + { + "epoch": 0.1468187762095865, + "grad_norm": 1.490397473130718, + "learning_rate": 3.858605162122314e-06, + "loss": 0.9777, + "step": 1628 + }, + { + "epoch": 0.14690895973305676, + "grad_norm": 1.6355738311215304, + "learning_rate": 3.858389322597592e-06, + "loss": 0.9548, + "step": 1629 + }, + { + "epoch": 0.14699914325652702, + "grad_norm": 1.6177095316996888, + "learning_rate": 3.858173324504847e-06, + "loss": 1.0152, + "step": 1630 + }, + { + "epoch": 0.14708932677999728, + "grad_norm": 1.4929844989674557, + "learning_rate": 3.857957167862508e-06, + "loss": 0.9608, + "step": 1631 + }, + { + "epoch": 0.14717951030346757, + "grad_norm": 0.8814235327016429, + "learning_rate": 3.857740852689018e-06, + "loss": 0.7825, + "step": 1632 + }, + { + "epoch": 0.14726969382693783, + "grad_norm": 1.8662588451709894, + "learning_rate": 3.857524379002835e-06, + "loss": 1.0051, + "step": 1633 + }, + { + "epoch": 0.1473598773504081, + "grad_norm": 1.6081907489143814, + "learning_rate": 3.85730774682243e-06, + "loss": 1.0206, + "step": 1634 + }, + { + "epoch": 0.14745006087387835, + "grad_norm": 1.6525191987231003, + "learning_rate": 3.8570909561662875e-06, + "loss": 1.065, + "step": 1635 + }, + { + "epoch": 0.1475402443973486, + "grad_norm": 1.6173265922549516, + "learning_rate": 3.8568740070529045e-06, + "loss": 1.067, + "step": 1636 + }, + { + "epoch": 0.14763042792081887, + "grad_norm": 1.975045206059599, + "learning_rate": 3.856656899500792e-06, + "loss": 0.9691, + "step": 1637 + }, + { + "epoch": 0.14772061144428913, + "grad_norm": 1.647479389044382, + "learning_rate": 3.856439633528476e-06, + "loss": 0.9329, + "step": 1638 + }, + { + "epoch": 0.1478107949677594, + "grad_norm": 3.072494546809374, + "learning_rate": 3.856222209154494e-06, + "loss": 1.0572, + "step": 1639 + }, + { + "epoch": 0.14790097849122966, + "grad_norm": 1.6037115054070894, + "learning_rate": 3.856004626397397e-06, + "loss": 1.025, + "step": 1640 + }, + { + "epoch": 0.14799116201469992, + "grad_norm": 1.6798177979077018, + "learning_rate": 3.855786885275753e-06, + "loss": 0.9725, + "step": 1641 + }, + { + "epoch": 0.14808134553817018, + "grad_norm": 2.265647204060107, + "learning_rate": 3.855568985808138e-06, + "loss": 0.9173, + "step": 1642 + }, + { + "epoch": 0.14817152906164044, + "grad_norm": 1.6643841213272903, + "learning_rate": 3.855350928013145e-06, + "loss": 0.9669, + "step": 1643 + }, + { + "epoch": 0.1482617125851107, + "grad_norm": 1.6982421875, + "learning_rate": 3.8551327119093825e-06, + "loss": 1.0043, + "step": 1644 + }, + { + "epoch": 0.14835189610858096, + "grad_norm": 1.723534254330746, + "learning_rate": 3.854914337515467e-06, + "loss": 0.9289, + "step": 1645 + }, + { + "epoch": 0.14844207963205122, + "grad_norm": 1.5296501833111493, + "learning_rate": 3.8546958048500324e-06, + "loss": 0.8919, + "step": 1646 + }, + { + "epoch": 0.14853226315552148, + "grad_norm": 0.8200517149062595, + "learning_rate": 3.854477113931725e-06, + "loss": 0.8345, + "step": 1647 + }, + { + "epoch": 0.14862244667899174, + "grad_norm": 1.695784683217034, + "learning_rate": 3.854258264779205e-06, + "loss": 1.0224, + "step": 1648 + }, + { + "epoch": 0.148712630202462, + "grad_norm": 1.6590274339032136, + "learning_rate": 3.854039257411145e-06, + "loss": 0.9803, + "step": 1649 + }, + { + "epoch": 0.14880281372593226, + "grad_norm": 1.8201756241709006, + "learning_rate": 3.853820091846232e-06, + "loss": 0.9513, + "step": 1650 + }, + { + "epoch": 0.14889299724940253, + "grad_norm": 1.9874816965663864, + "learning_rate": 3.853600768103169e-06, + "loss": 1.0037, + "step": 1651 + }, + { + "epoch": 0.1489831807728728, + "grad_norm": 2.110036901540017, + "learning_rate": 3.853381286200667e-06, + "loss": 0.9851, + "step": 1652 + }, + { + "epoch": 0.14907336429634305, + "grad_norm": 1.9240593011286433, + "learning_rate": 3.853161646157453e-06, + "loss": 1.0901, + "step": 1653 + }, + { + "epoch": 0.1491635478198133, + "grad_norm": 1.734156758092057, + "learning_rate": 3.852941847992269e-06, + "loss": 0.961, + "step": 1654 + }, + { + "epoch": 0.14925373134328357, + "grad_norm": 1.9591643687172149, + "learning_rate": 3.852721891723871e-06, + "loss": 1.0128, + "step": 1655 + }, + { + "epoch": 0.14934391486675386, + "grad_norm": 1.7624616388786103, + "learning_rate": 3.852501777371025e-06, + "loss": 0.978, + "step": 1656 + }, + { + "epoch": 0.14943409839022412, + "grad_norm": 1.9692802320685017, + "learning_rate": 3.8522815049525125e-06, + "loss": 1.0378, + "step": 1657 + }, + { + "epoch": 0.14952428191369438, + "grad_norm": 1.742858757402445, + "learning_rate": 3.852061074487129e-06, + "loss": 1.0487, + "step": 1658 + }, + { + "epoch": 0.14961446543716464, + "grad_norm": 1.6321008034690523, + "learning_rate": 3.851840485993682e-06, + "loss": 0.9107, + "step": 1659 + }, + { + "epoch": 0.1497046489606349, + "grad_norm": 1.5335075707884611, + "learning_rate": 3.851619739490994e-06, + "loss": 1.0051, + "step": 1660 + }, + { + "epoch": 0.14979483248410516, + "grad_norm": 1.767219093910531, + "learning_rate": 3.8513988349978996e-06, + "loss": 1.0326, + "step": 1661 + }, + { + "epoch": 0.14988501600757542, + "grad_norm": 1.4474095982074011, + "learning_rate": 3.851177772533249e-06, + "loss": 1.023, + "step": 1662 + }, + { + "epoch": 0.14997519953104568, + "grad_norm": 2.0844068304602597, + "learning_rate": 3.850956552115903e-06, + "loss": 1.175, + "step": 1663 + }, + { + "epoch": 0.15006538305451594, + "grad_norm": 0.8282607705002921, + "learning_rate": 3.850735173764738e-06, + "loss": 0.785, + "step": 1664 + }, + { + "epoch": 0.1501555665779862, + "grad_norm": 1.6004706405811713, + "learning_rate": 3.850513637498642e-06, + "loss": 1.017, + "step": 1665 + }, + { + "epoch": 0.15024575010145647, + "grad_norm": 1.8241710452391295, + "learning_rate": 3.850291943336521e-06, + "loss": 1.0204, + "step": 1666 + }, + { + "epoch": 0.15033593362492673, + "grad_norm": 1.8509428379665203, + "learning_rate": 3.850070091297287e-06, + "loss": 1.0523, + "step": 1667 + }, + { + "epoch": 0.150426117148397, + "grad_norm": 0.8669141390105201, + "learning_rate": 3.8498480813998735e-06, + "loss": 0.8769, + "step": 1668 + }, + { + "epoch": 0.15051630067186725, + "grad_norm": 1.5515907154418136, + "learning_rate": 3.84962591366322e-06, + "loss": 0.9933, + "step": 1669 + }, + { + "epoch": 0.1506064841953375, + "grad_norm": 1.8474217723756552, + "learning_rate": 3.8494035881062855e-06, + "loss": 1.042, + "step": 1670 + }, + { + "epoch": 0.15069666771880777, + "grad_norm": 1.8539696992416508, + "learning_rate": 3.84918110474804e-06, + "loss": 0.9803, + "step": 1671 + }, + { + "epoch": 0.15078685124227803, + "grad_norm": 1.6572953290277468, + "learning_rate": 3.8489584636074655e-06, + "loss": 0.9906, + "step": 1672 + }, + { + "epoch": 0.1508770347657483, + "grad_norm": 1.283510075614682, + "learning_rate": 3.848735664703561e-06, + "loss": 0.7683, + "step": 1673 + }, + { + "epoch": 0.15096721828921855, + "grad_norm": 1.4319808160253404, + "learning_rate": 3.8485127080553346e-06, + "loss": 0.9963, + "step": 1674 + }, + { + "epoch": 0.1510574018126888, + "grad_norm": 1.7811647361292062, + "learning_rate": 3.8482895936818115e-06, + "loss": 0.9976, + "step": 1675 + }, + { + "epoch": 0.15114758533615907, + "grad_norm": 2.6426872913172947, + "learning_rate": 3.848066321602029e-06, + "loss": 1.0155, + "step": 1676 + }, + { + "epoch": 0.15123776885962933, + "grad_norm": 1.8347670050151492, + "learning_rate": 3.847842891835038e-06, + "loss": 0.9638, + "step": 1677 + }, + { + "epoch": 0.1513279523830996, + "grad_norm": 1.123862963505398, + "learning_rate": 3.847619304399902e-06, + "loss": 0.7845, + "step": 1678 + }, + { + "epoch": 0.15141813590656986, + "grad_norm": 1.9293774189638841, + "learning_rate": 3.8473955593157e-06, + "loss": 1.0262, + "step": 1679 + }, + { + "epoch": 0.15150831943004014, + "grad_norm": 1.7068931820549518, + "learning_rate": 3.847171656601522e-06, + "loss": 0.9668, + "step": 1680 + }, + { + "epoch": 0.1515985029535104, + "grad_norm": 1.7344201571797617, + "learning_rate": 3.846947596276473e-06, + "loss": 1.0284, + "step": 1681 + }, + { + "epoch": 0.15168868647698067, + "grad_norm": 1.8538151811495653, + "learning_rate": 3.846723378359672e-06, + "loss": 1.0528, + "step": 1682 + }, + { + "epoch": 0.15177887000045093, + "grad_norm": 1.9163824230008277, + "learning_rate": 3.846499002870249e-06, + "loss": 1.0032, + "step": 1683 + }, + { + "epoch": 0.1518690535239212, + "grad_norm": 1.8949615933595112, + "learning_rate": 3.846274469827349e-06, + "loss": 1.0209, + "step": 1684 + }, + { + "epoch": 0.15195923704739145, + "grad_norm": 1.6962147785935067, + "learning_rate": 3.846049779250132e-06, + "loss": 1.0063, + "step": 1685 + }, + { + "epoch": 0.1520494205708617, + "grad_norm": 0.8818845006808126, + "learning_rate": 3.845824931157769e-06, + "loss": 0.8252, + "step": 1686 + }, + { + "epoch": 0.15213960409433197, + "grad_norm": 1.4475957208321941, + "learning_rate": 3.845599925569444e-06, + "loss": 0.9833, + "step": 1687 + }, + { + "epoch": 0.15222978761780223, + "grad_norm": 1.9273795097877728, + "learning_rate": 3.845374762504357e-06, + "loss": 0.9632, + "step": 1688 + }, + { + "epoch": 0.1523199711412725, + "grad_norm": 2.0324022986095924, + "learning_rate": 3.8451494419817204e-06, + "loss": 1.0539, + "step": 1689 + }, + { + "epoch": 0.15241015466474275, + "grad_norm": 2.1429872496025797, + "learning_rate": 3.8449239640207594e-06, + "loss": 0.9352, + "step": 1690 + }, + { + "epoch": 0.152500338188213, + "grad_norm": 2.0546994517163686, + "learning_rate": 3.844698328640713e-06, + "loss": 1.0479, + "step": 1691 + }, + { + "epoch": 0.15259052171168327, + "grad_norm": 1.8349795175599377, + "learning_rate": 3.844472535860833e-06, + "loss": 1.0174, + "step": 1692 + }, + { + "epoch": 0.15268070523515354, + "grad_norm": 2.0169436850376563, + "learning_rate": 3.8442465857003864e-06, + "loss": 1.0488, + "step": 1693 + }, + { + "epoch": 0.1527708887586238, + "grad_norm": 1.5720993427937915, + "learning_rate": 3.844020478178653e-06, + "loss": 0.9507, + "step": 1694 + }, + { + "epoch": 0.15286107228209406, + "grad_norm": 1.7332945198824172, + "learning_rate": 3.843794213314923e-06, + "loss": 1.026, + "step": 1695 + }, + { + "epoch": 0.15295125580556432, + "grad_norm": 1.7386537988460558, + "learning_rate": 3.843567791128505e-06, + "loss": 1.0312, + "step": 1696 + }, + { + "epoch": 0.15304143932903458, + "grad_norm": 1.4299884750328609, + "learning_rate": 3.843341211638717e-06, + "loss": 0.9875, + "step": 1697 + }, + { + "epoch": 0.15313162285250484, + "grad_norm": 3.4553632429878065, + "learning_rate": 3.843114474864894e-06, + "loss": 0.9611, + "step": 1698 + }, + { + "epoch": 0.1532218063759751, + "grad_norm": 1.5579100521845386, + "learning_rate": 3.84288758082638e-06, + "loss": 0.9988, + "step": 1699 + }, + { + "epoch": 0.15331198989944536, + "grad_norm": 1.5308428631802409, + "learning_rate": 3.842660529542536e-06, + "loss": 1.0874, + "step": 1700 + }, + { + "epoch": 0.15340217342291562, + "grad_norm": 1.5847110610080009, + "learning_rate": 3.842433321032736e-06, + "loss": 1.0154, + "step": 1701 + }, + { + "epoch": 0.15349235694638588, + "grad_norm": 1.5946684789971566, + "learning_rate": 3.842205955316365e-06, + "loss": 0.924, + "step": 1702 + }, + { + "epoch": 0.15358254046985614, + "grad_norm": 1.875929856202965, + "learning_rate": 3.8419784324128256e-06, + "loss": 0.9081, + "step": 1703 + }, + { + "epoch": 0.15367272399332643, + "grad_norm": 1.7921736539590198, + "learning_rate": 3.841750752341529e-06, + "loss": 0.9679, + "step": 1704 + }, + { + "epoch": 0.1537629075167967, + "grad_norm": 1.5904043647917292, + "learning_rate": 3.841522915121902e-06, + "loss": 1.0044, + "step": 1705 + }, + { + "epoch": 0.15385309104026695, + "grad_norm": 2.6362095319916663, + "learning_rate": 3.841294920773387e-06, + "loss": 0.9332, + "step": 1706 + }, + { + "epoch": 0.15394327456373721, + "grad_norm": 1.9447626156332167, + "learning_rate": 3.841066769315436e-06, + "loss": 1.0153, + "step": 1707 + }, + { + "epoch": 0.15403345808720748, + "grad_norm": 1.601886400174026, + "learning_rate": 3.840838460767518e-06, + "loss": 1.0095, + "step": 1708 + }, + { + "epoch": 0.15412364161067774, + "grad_norm": 1.7596584545699492, + "learning_rate": 3.840609995149111e-06, + "loss": 0.9719, + "step": 1709 + }, + { + "epoch": 0.154213825134148, + "grad_norm": 1.6042995232726727, + "learning_rate": 3.84038137247971e-06, + "loss": 1.0067, + "step": 1710 + }, + { + "epoch": 0.15430400865761826, + "grad_norm": 2.0582623730781853, + "learning_rate": 3.840152592778823e-06, + "loss": 0.9896, + "step": 1711 + }, + { + "epoch": 0.15439419218108852, + "grad_norm": 1.277552129797918, + "learning_rate": 3.83992365606597e-06, + "loss": 1.0472, + "step": 1712 + }, + { + "epoch": 0.15448437570455878, + "grad_norm": 1.6776442764478827, + "learning_rate": 3.8396945623606855e-06, + "loss": 0.9551, + "step": 1713 + }, + { + "epoch": 0.15457455922802904, + "grad_norm": 1.495029000858988, + "learning_rate": 3.8394653116825174e-06, + "loss": 0.9527, + "step": 1714 + }, + { + "epoch": 0.1546647427514993, + "grad_norm": 1.770688316540804, + "learning_rate": 3.839235904051026e-06, + "loss": 0.987, + "step": 1715 + }, + { + "epoch": 0.15475492627496956, + "grad_norm": 1.7658924811650907, + "learning_rate": 3.8390063394857855e-06, + "loss": 0.9766, + "step": 1716 + }, + { + "epoch": 0.15484510979843982, + "grad_norm": 1.4399559767934398, + "learning_rate": 3.838776618006385e-06, + "loss": 1.087, + "step": 1717 + }, + { + "epoch": 0.15493529332191008, + "grad_norm": 1.541477226096519, + "learning_rate": 3.838546739632423e-06, + "loss": 1.0104, + "step": 1718 + }, + { + "epoch": 0.15502547684538034, + "grad_norm": 1.4786565554737212, + "learning_rate": 3.838316704383517e-06, + "loss": 1.0465, + "step": 1719 + }, + { + "epoch": 0.1551156603688506, + "grad_norm": 1.6916646077311095, + "learning_rate": 3.838086512279292e-06, + "loss": 0.9931, + "step": 1720 + }, + { + "epoch": 0.15520584389232087, + "grad_norm": 1.939546242867028, + "learning_rate": 3.837856163339391e-06, + "loss": 0.9909, + "step": 1721 + }, + { + "epoch": 0.15529602741579113, + "grad_norm": 1.5346737945572289, + "learning_rate": 3.837625657583469e-06, + "loss": 0.9294, + "step": 1722 + }, + { + "epoch": 0.1553862109392614, + "grad_norm": 2.0738105444092003, + "learning_rate": 3.837394995031193e-06, + "loss": 0.9574, + "step": 1723 + }, + { + "epoch": 0.15547639446273165, + "grad_norm": 1.6749380698071223, + "learning_rate": 3.837164175702245e-06, + "loss": 0.9968, + "step": 1724 + }, + { + "epoch": 0.1555665779862019, + "grad_norm": 1.922202849918325, + "learning_rate": 3.836933199616319e-06, + "loss": 0.9131, + "step": 1725 + }, + { + "epoch": 0.15565676150967217, + "grad_norm": 1.639642475746681, + "learning_rate": 3.836702066793124e-06, + "loss": 1.0662, + "step": 1726 + }, + { + "epoch": 0.15574694503314246, + "grad_norm": 1.5893223162824994, + "learning_rate": 3.836470777252381e-06, + "loss": 1.0601, + "step": 1727 + }, + { + "epoch": 0.15583712855661272, + "grad_norm": 2.000576293886317, + "learning_rate": 3.836239331013825e-06, + "loss": 1.0137, + "step": 1728 + }, + { + "epoch": 0.15592731208008298, + "grad_norm": 1.7111625349208497, + "learning_rate": 3.836007728097205e-06, + "loss": 0.9601, + "step": 1729 + }, + { + "epoch": 0.15601749560355324, + "grad_norm": 1.4869988791455764, + "learning_rate": 3.835775968522282e-06, + "loss": 0.951, + "step": 1730 + }, + { + "epoch": 0.1561076791270235, + "grad_norm": 1.814522765499544, + "learning_rate": 3.83554405230883e-06, + "loss": 1.0179, + "step": 1731 + }, + { + "epoch": 0.15619786265049376, + "grad_norm": 1.5816829345048757, + "learning_rate": 3.835311979476639e-06, + "loss": 0.9847, + "step": 1732 + }, + { + "epoch": 0.15628804617396402, + "grad_norm": 4.595231628225178, + "learning_rate": 3.83507975004551e-06, + "loss": 0.8899, + "step": 1733 + }, + { + "epoch": 0.15637822969743428, + "grad_norm": 1.6384190579194438, + "learning_rate": 3.834847364035258e-06, + "loss": 0.9792, + "step": 1734 + }, + { + "epoch": 0.15646841322090455, + "grad_norm": 2.0969598562521194, + "learning_rate": 3.834614821465712e-06, + "loss": 0.9366, + "step": 1735 + }, + { + "epoch": 0.1565585967443748, + "grad_norm": 1.610069865612269, + "learning_rate": 3.834382122356713e-06, + "loss": 1.0337, + "step": 1736 + }, + { + "epoch": 0.15664878026784507, + "grad_norm": 1.5378723965107863, + "learning_rate": 3.834149266728117e-06, + "loss": 0.955, + "step": 1737 + }, + { + "epoch": 0.15673896379131533, + "grad_norm": 0.7615262301780994, + "learning_rate": 3.833916254599792e-06, + "loss": 0.823, + "step": 1738 + }, + { + "epoch": 0.1568291473147856, + "grad_norm": 0.6573876102831636, + "learning_rate": 3.83368308599162e-06, + "loss": 0.769, + "step": 1739 + }, + { + "epoch": 0.15691933083825585, + "grad_norm": 0.6711779904589245, + "learning_rate": 3.833449760923498e-06, + "loss": 0.7403, + "step": 1740 + }, + { + "epoch": 0.1570095143617261, + "grad_norm": 1.6615023552335197, + "learning_rate": 3.83321627941533e-06, + "loss": 1.036, + "step": 1741 + }, + { + "epoch": 0.15709969788519637, + "grad_norm": 1.800344948035038, + "learning_rate": 3.832982641487043e-06, + "loss": 1.0518, + "step": 1742 + }, + { + "epoch": 0.15718988140866663, + "grad_norm": 1.6442454762558247, + "learning_rate": 3.832748847158568e-06, + "loss": 1.0868, + "step": 1743 + }, + { + "epoch": 0.1572800649321369, + "grad_norm": 1.4739883833159577, + "learning_rate": 3.832514896449858e-06, + "loss": 1.021, + "step": 1744 + }, + { + "epoch": 0.15737024845560715, + "grad_norm": 1.7147281407899668, + "learning_rate": 3.832280789380871e-06, + "loss": 0.9372, + "step": 1745 + }, + { + "epoch": 0.15746043197907741, + "grad_norm": 1.4788405994770022, + "learning_rate": 3.832046525971584e-06, + "loss": 0.8979, + "step": 1746 + }, + { + "epoch": 0.15755061550254768, + "grad_norm": 2.371969900130017, + "learning_rate": 3.831812106241987e-06, + "loss": 1.0576, + "step": 1747 + }, + { + "epoch": 0.15764079902601794, + "grad_norm": 1.775815604431835, + "learning_rate": 3.8315775302120796e-06, + "loss": 0.9057, + "step": 1748 + }, + { + "epoch": 0.1577309825494882, + "grad_norm": 2.081414967090482, + "learning_rate": 3.831342797901878e-06, + "loss": 1.0115, + "step": 1749 + }, + { + "epoch": 0.15782116607295846, + "grad_norm": 1.1713892120551048, + "learning_rate": 3.831107909331411e-06, + "loss": 0.7817, + "step": 1750 + }, + { + "epoch": 0.15791134959642875, + "grad_norm": 1.802057637231749, + "learning_rate": 3.830872864520721e-06, + "loss": 1.0378, + "step": 1751 + }, + { + "epoch": 0.158001533119899, + "grad_norm": 0.7755593757624876, + "learning_rate": 3.830637663489862e-06, + "loss": 0.8604, + "step": 1752 + }, + { + "epoch": 0.15809171664336927, + "grad_norm": 0.8524318422271969, + "learning_rate": 3.830402306258904e-06, + "loss": 0.8913, + "step": 1753 + }, + { + "epoch": 0.15818190016683953, + "grad_norm": 1.6604503415983456, + "learning_rate": 3.830166792847929e-06, + "loss": 0.9849, + "step": 1754 + }, + { + "epoch": 0.1582720836903098, + "grad_norm": 1.4387591279350942, + "learning_rate": 3.829931123277031e-06, + "loss": 0.9364, + "step": 1755 + }, + { + "epoch": 0.15836226721378005, + "grad_norm": 1.643074387053426, + "learning_rate": 3.8296952975663204e-06, + "loss": 0.9795, + "step": 1756 + }, + { + "epoch": 0.1584524507372503, + "grad_norm": 3.79676963126202, + "learning_rate": 3.829459315735918e-06, + "loss": 1.0635, + "step": 1757 + }, + { + "epoch": 0.15854263426072057, + "grad_norm": 1.6481522589039743, + "learning_rate": 3.829223177805959e-06, + "loss": 0.92, + "step": 1758 + }, + { + "epoch": 0.15863281778419083, + "grad_norm": 1.6178864521221659, + "learning_rate": 3.828986883796591e-06, + "loss": 1.0364, + "step": 1759 + }, + { + "epoch": 0.1587230013076611, + "grad_norm": 1.9821425969382647, + "learning_rate": 3.828750433727979e-06, + "loss": 0.9888, + "step": 1760 + }, + { + "epoch": 0.15881318483113135, + "grad_norm": 1.8797611187274872, + "learning_rate": 3.828513827620296e-06, + "loss": 1.0422, + "step": 1761 + }, + { + "epoch": 0.15890336835460162, + "grad_norm": 1.8136421419722175, + "learning_rate": 3.82827706549373e-06, + "loss": 1.0507, + "step": 1762 + }, + { + "epoch": 0.15899355187807188, + "grad_norm": 1.7127954854609055, + "learning_rate": 3.828040147368484e-06, + "loss": 0.9883, + "step": 1763 + }, + { + "epoch": 0.15908373540154214, + "grad_norm": 1.52199967831221, + "learning_rate": 3.827803073264774e-06, + "loss": 0.9902, + "step": 1764 + }, + { + "epoch": 0.1591739189250124, + "grad_norm": 1.5867997254483193, + "learning_rate": 3.827565843202826e-06, + "loss": 0.9935, + "step": 1765 + }, + { + "epoch": 0.15926410244848266, + "grad_norm": 1.681955908307419, + "learning_rate": 3.827328457202884e-06, + "loss": 0.8677, + "step": 1766 + }, + { + "epoch": 0.15935428597195292, + "grad_norm": 1.5462252090181345, + "learning_rate": 3.8270909152852014e-06, + "loss": 0.9292, + "step": 1767 + }, + { + "epoch": 0.15944446949542318, + "grad_norm": 1.8757547131194081, + "learning_rate": 3.826853217470048e-06, + "loss": 0.9909, + "step": 1768 + }, + { + "epoch": 0.15953465301889344, + "grad_norm": 1.7665779445370522, + "learning_rate": 3.826615363777705e-06, + "loss": 1.0708, + "step": 1769 + }, + { + "epoch": 0.1596248365423637, + "grad_norm": 1.4355028625799886, + "learning_rate": 3.826377354228468e-06, + "loss": 0.9694, + "step": 1770 + }, + { + "epoch": 0.15971502006583396, + "grad_norm": 0.8466703948992648, + "learning_rate": 3.826139188842643e-06, + "loss": 0.7945, + "step": 1771 + }, + { + "epoch": 0.15980520358930422, + "grad_norm": 1.4202082160291958, + "learning_rate": 3.825900867640554e-06, + "loss": 1.049, + "step": 1772 + }, + { + "epoch": 0.15989538711277448, + "grad_norm": 1.4448951737287294, + "learning_rate": 3.825662390642535e-06, + "loss": 1.036, + "step": 1773 + }, + { + "epoch": 0.15998557063624474, + "grad_norm": 1.7436025030725044, + "learning_rate": 3.825423757868934e-06, + "loss": 1.0991, + "step": 1774 + }, + { + "epoch": 0.16007575415971503, + "grad_norm": 1.9284346988690397, + "learning_rate": 3.825184969340114e-06, + "loss": 1.0209, + "step": 1775 + }, + { + "epoch": 0.1601659376831853, + "grad_norm": 1.4682275370408029, + "learning_rate": 3.824946025076447e-06, + "loss": 1.0245, + "step": 1776 + }, + { + "epoch": 0.16025612120665556, + "grad_norm": 1.7094167282339292, + "learning_rate": 3.824706925098323e-06, + "loss": 0.9397, + "step": 1777 + }, + { + "epoch": 0.16034630473012582, + "grad_norm": 2.34304402209161, + "learning_rate": 3.824467669426143e-06, + "loss": 1.0523, + "step": 1778 + }, + { + "epoch": 0.16043648825359608, + "grad_norm": 1.5372767278556558, + "learning_rate": 3.824228258080321e-06, + "loss": 0.9638, + "step": 1779 + }, + { + "epoch": 0.16052667177706634, + "grad_norm": 1.761381603384252, + "learning_rate": 3.823988691081285e-06, + "loss": 0.9103, + "step": 1780 + }, + { + "epoch": 0.1606168553005366, + "grad_norm": 1.883345469151559, + "learning_rate": 3.823748968449478e-06, + "loss": 1.0282, + "step": 1781 + }, + { + "epoch": 0.16070703882400686, + "grad_norm": 1.8319670903860605, + "learning_rate": 3.823509090205352e-06, + "loss": 0.8861, + "step": 1782 + }, + { + "epoch": 0.16079722234747712, + "grad_norm": 1.645739186487393, + "learning_rate": 3.823269056369376e-06, + "loss": 0.9493, + "step": 1783 + }, + { + "epoch": 0.16088740587094738, + "grad_norm": 2.4227106437001806, + "learning_rate": 3.8230288669620295e-06, + "loss": 1.0275, + "step": 1784 + }, + { + "epoch": 0.16097758939441764, + "grad_norm": 1.4607962473753229, + "learning_rate": 3.822788522003809e-06, + "loss": 0.9959, + "step": 1785 + }, + { + "epoch": 0.1610677729178879, + "grad_norm": 2.3391006130255376, + "learning_rate": 3.822548021515221e-06, + "loss": 1.0705, + "step": 1786 + }, + { + "epoch": 0.16115795644135816, + "grad_norm": 1.688201334702256, + "learning_rate": 3.822307365516787e-06, + "loss": 0.9767, + "step": 1787 + }, + { + "epoch": 0.16124813996482842, + "grad_norm": 1.739831057820405, + "learning_rate": 3.8220665540290395e-06, + "loss": 1.0125, + "step": 1788 + }, + { + "epoch": 0.16133832348829868, + "grad_norm": 2.2872819973817995, + "learning_rate": 3.8218255870725265e-06, + "loss": 1.0622, + "step": 1789 + }, + { + "epoch": 0.16142850701176895, + "grad_norm": 1.5458528628833639, + "learning_rate": 3.82158446466781e-06, + "loss": 0.9451, + "step": 1790 + }, + { + "epoch": 0.1615186905352392, + "grad_norm": 1.4949006186021432, + "learning_rate": 3.821343186835462e-06, + "loss": 1.0036, + "step": 1791 + }, + { + "epoch": 0.16160887405870947, + "grad_norm": 1.6117971766404415, + "learning_rate": 3.821101753596072e-06, + "loss": 1.0651, + "step": 1792 + }, + { + "epoch": 0.16169905758217973, + "grad_norm": 1.857780670586351, + "learning_rate": 3.820860164970237e-06, + "loss": 0.981, + "step": 1793 + }, + { + "epoch": 0.16178924110565, + "grad_norm": 1.4939459380922666, + "learning_rate": 3.820618420978574e-06, + "loss": 1.0638, + "step": 1794 + }, + { + "epoch": 0.16187942462912025, + "grad_norm": 1.7705687456118788, + "learning_rate": 3.820376521641708e-06, + "loss": 0.9907, + "step": 1795 + }, + { + "epoch": 0.1619696081525905, + "grad_norm": 1.6622631391859968, + "learning_rate": 3.82013446698028e-06, + "loss": 1.0063, + "step": 1796 + }, + { + "epoch": 0.16205979167606077, + "grad_norm": 1.7280088554658746, + "learning_rate": 3.819892257014943e-06, + "loss": 0.9591, + "step": 1797 + }, + { + "epoch": 0.16214997519953103, + "grad_norm": 1.6516643597419607, + "learning_rate": 3.819649891766364e-06, + "loss": 1.0213, + "step": 1798 + }, + { + "epoch": 0.16224015872300132, + "grad_norm": 1.605371328400623, + "learning_rate": 3.819407371255222e-06, + "loss": 0.9868, + "step": 1799 + }, + { + "epoch": 0.16233034224647158, + "grad_norm": 1.7251353169316468, + "learning_rate": 3.819164695502212e-06, + "loss": 1.0047, + "step": 1800 + }, + { + "epoch": 0.16242052576994184, + "grad_norm": 1.552142029064174, + "learning_rate": 3.818921864528039e-06, + "loss": 0.9415, + "step": 1801 + }, + { + "epoch": 0.1625107092934121, + "grad_norm": 2.038838341329193, + "learning_rate": 3.818678878353423e-06, + "loss": 1.0499, + "step": 1802 + }, + { + "epoch": 0.16260089281688236, + "grad_norm": 1.722221752221375, + "learning_rate": 3.818435736999097e-06, + "loss": 0.937, + "step": 1803 + }, + { + "epoch": 0.16269107634035262, + "grad_norm": 1.678422247278104, + "learning_rate": 3.818192440485807e-06, + "loss": 0.9897, + "step": 1804 + }, + { + "epoch": 0.16278125986382289, + "grad_norm": 1.8642372861584422, + "learning_rate": 3.817948988834314e-06, + "loss": 0.9282, + "step": 1805 + }, + { + "epoch": 0.16287144338729315, + "grad_norm": 1.3630720267656666, + "learning_rate": 3.817705382065388e-06, + "loss": 0.9892, + "step": 1806 + }, + { + "epoch": 0.1629616269107634, + "grad_norm": 1.2533460655140602, + "learning_rate": 3.8174616201998155e-06, + "loss": 0.9634, + "step": 1807 + }, + { + "epoch": 0.16305181043423367, + "grad_norm": 1.6928719387707005, + "learning_rate": 3.817217703258397e-06, + "loss": 1.0437, + "step": 1808 + }, + { + "epoch": 0.16314199395770393, + "grad_norm": 1.5605269372308572, + "learning_rate": 3.816973631261943e-06, + "loss": 0.9839, + "step": 1809 + }, + { + "epoch": 0.1632321774811742, + "grad_norm": 1.873226853237206, + "learning_rate": 3.816729404231281e-06, + "loss": 1.0263, + "step": 1810 + }, + { + "epoch": 0.16332236100464445, + "grad_norm": 2.0011959076718524, + "learning_rate": 3.816485022187249e-06, + "loss": 1.0621, + "step": 1811 + }, + { + "epoch": 0.1634125445281147, + "grad_norm": 2.2585804831010603, + "learning_rate": 3.816240485150698e-06, + "loss": 0.9521, + "step": 1812 + }, + { + "epoch": 0.16350272805158497, + "grad_norm": 1.5003103888762388, + "learning_rate": 3.815995793142495e-06, + "loss": 0.9905, + "step": 1813 + }, + { + "epoch": 0.16359291157505523, + "grad_norm": 5.557240946811463, + "learning_rate": 3.815750946183518e-06, + "loss": 0.9338, + "step": 1814 + }, + { + "epoch": 0.1636830950985255, + "grad_norm": 1.805452919349612, + "learning_rate": 3.815505944294658e-06, + "loss": 1.0492, + "step": 1815 + }, + { + "epoch": 0.16377327862199575, + "grad_norm": 1.495993110319582, + "learning_rate": 3.81526078749682e-06, + "loss": 0.9546, + "step": 1816 + }, + { + "epoch": 0.16386346214546602, + "grad_norm": 2.3954241085241015, + "learning_rate": 3.8150154758109225e-06, + "loss": 0.9275, + "step": 1817 + }, + { + "epoch": 0.16395364566893628, + "grad_norm": 1.5346621429206555, + "learning_rate": 3.814770009257896e-06, + "loss": 0.9675, + "step": 1818 + }, + { + "epoch": 0.16404382919240654, + "grad_norm": 1.503039618147001, + "learning_rate": 3.814524387858687e-06, + "loss": 1.0047, + "step": 1819 + }, + { + "epoch": 0.1641340127158768, + "grad_norm": 2.871818108560712, + "learning_rate": 3.814278611634251e-06, + "loss": 1.0122, + "step": 1820 + }, + { + "epoch": 0.16422419623934706, + "grad_norm": 1.5987737426903665, + "learning_rate": 3.8140326806055606e-06, + "loss": 0.9629, + "step": 1821 + }, + { + "epoch": 0.16431437976281732, + "grad_norm": 1.8007526519556734, + "learning_rate": 3.8137865947935992e-06, + "loss": 1.0526, + "step": 1822 + }, + { + "epoch": 0.1644045632862876, + "grad_norm": 1.965999377187512, + "learning_rate": 3.8135403542193646e-06, + "loss": 1.0099, + "step": 1823 + }, + { + "epoch": 0.16449474680975787, + "grad_norm": 2.0895352429998115, + "learning_rate": 3.813293958903867e-06, + "loss": 0.9255, + "step": 1824 + }, + { + "epoch": 0.16458493033322813, + "grad_norm": 1.739644268438188, + "learning_rate": 3.8130474088681306e-06, + "loss": 0.9568, + "step": 1825 + }, + { + "epoch": 0.1646751138566984, + "grad_norm": 2.2516963709809152, + "learning_rate": 3.8128007041331927e-06, + "loss": 0.9222, + "step": 1826 + }, + { + "epoch": 0.16476529738016865, + "grad_norm": 1.6261153428245967, + "learning_rate": 3.812553844720102e-06, + "loss": 0.9058, + "step": 1827 + }, + { + "epoch": 0.1648554809036389, + "grad_norm": 1.7739006479605444, + "learning_rate": 3.8123068306499236e-06, + "loss": 1.0113, + "step": 1828 + }, + { + "epoch": 0.16494566442710917, + "grad_norm": 10.764057283349004, + "learning_rate": 3.812059661943733e-06, + "loss": 0.9851, + "step": 1829 + }, + { + "epoch": 0.16503584795057943, + "grad_norm": 1.589518820810504, + "learning_rate": 3.811812338622621e-06, + "loss": 0.9623, + "step": 1830 + }, + { + "epoch": 0.1651260314740497, + "grad_norm": 1.4520290713066586, + "learning_rate": 3.81156486070769e-06, + "loss": 0.9807, + "step": 1831 + }, + { + "epoch": 0.16521621499751996, + "grad_norm": 1.5567300691837302, + "learning_rate": 3.811317228220056e-06, + "loss": 1.0341, + "step": 1832 + }, + { + "epoch": 0.16530639852099022, + "grad_norm": 2.032312203490412, + "learning_rate": 3.811069441180849e-06, + "loss": 1.0402, + "step": 1833 + }, + { + "epoch": 0.16539658204446048, + "grad_norm": 1.749047633290728, + "learning_rate": 3.8108214996112107e-06, + "loss": 0.9931, + "step": 1834 + }, + { + "epoch": 0.16548676556793074, + "grad_norm": 1.9889796740959107, + "learning_rate": 3.810573403532297e-06, + "loss": 1.0282, + "step": 1835 + }, + { + "epoch": 0.165576949091401, + "grad_norm": 1.5187133313685592, + "learning_rate": 3.8103251529652774e-06, + "loss": 0.969, + "step": 1836 + }, + { + "epoch": 0.16566713261487126, + "grad_norm": 1.5613686089400538, + "learning_rate": 3.810076747931334e-06, + "loss": 0.9661, + "step": 1837 + }, + { + "epoch": 0.16575731613834152, + "grad_norm": 1.745360969546622, + "learning_rate": 3.809828188451662e-06, + "loss": 0.952, + "step": 1838 + }, + { + "epoch": 0.16584749966181178, + "grad_norm": 1.5954703227707763, + "learning_rate": 3.809579474547469e-06, + "loss": 0.9885, + "step": 1839 + }, + { + "epoch": 0.16593768318528204, + "grad_norm": 1.53018797810491, + "learning_rate": 3.809330606239977e-06, + "loss": 1.0396, + "step": 1840 + }, + { + "epoch": 0.1660278667087523, + "grad_norm": 1.4225607256003625, + "learning_rate": 3.809081583550422e-06, + "loss": 1.057, + "step": 1841 + }, + { + "epoch": 0.16611805023222256, + "grad_norm": 2.1046490887349005, + "learning_rate": 3.808832406500051e-06, + "loss": 1.0718, + "step": 1842 + }, + { + "epoch": 0.16620823375569282, + "grad_norm": 1.675427166003755, + "learning_rate": 3.8085830751101253e-06, + "loss": 1.0634, + "step": 1843 + }, + { + "epoch": 0.16629841727916309, + "grad_norm": 1.4648427734371745, + "learning_rate": 3.808333589401919e-06, + "loss": 1.0266, + "step": 1844 + }, + { + "epoch": 0.16638860080263335, + "grad_norm": 2.179755643921082, + "learning_rate": 3.8080839493967194e-06, + "loss": 0.926, + "step": 1845 + }, + { + "epoch": 0.1664787843261036, + "grad_norm": 0.7272993883480467, + "learning_rate": 3.807834155115828e-06, + "loss": 0.7477, + "step": 1846 + }, + { + "epoch": 0.1665689678495739, + "grad_norm": 0.6794562329996149, + "learning_rate": 3.8075842065805584e-06, + "loss": 0.852, + "step": 1847 + }, + { + "epoch": 0.16665915137304416, + "grad_norm": 2.380287909203566, + "learning_rate": 3.8073341038122374e-06, + "loss": 0.982, + "step": 1848 + }, + { + "epoch": 0.16674933489651442, + "grad_norm": 1.5510034666188677, + "learning_rate": 3.8070838468322048e-06, + "loss": 1.0617, + "step": 1849 + }, + { + "epoch": 0.16683951841998468, + "grad_norm": 1.7922806759652905, + "learning_rate": 3.8068334356618143e-06, + "loss": 0.9338, + "step": 1850 + }, + { + "epoch": 0.16692970194345494, + "grad_norm": 1.9490136170476737, + "learning_rate": 3.8065828703224324e-06, + "loss": 1.0231, + "step": 1851 + }, + { + "epoch": 0.1670198854669252, + "grad_norm": 1.7363239369808534, + "learning_rate": 3.8063321508354386e-06, + "loss": 0.9917, + "step": 1852 + }, + { + "epoch": 0.16711006899039546, + "grad_norm": 1.5274002572067782, + "learning_rate": 3.8060812772222255e-06, + "loss": 0.9718, + "step": 1853 + }, + { + "epoch": 0.16720025251386572, + "grad_norm": 1.3375053352178246, + "learning_rate": 3.8058302495041993e-06, + "loss": 1.0596, + "step": 1854 + }, + { + "epoch": 0.16729043603733598, + "grad_norm": 1.893501741358639, + "learning_rate": 3.805579067702779e-06, + "loss": 1.0053, + "step": 1855 + }, + { + "epoch": 0.16738061956080624, + "grad_norm": 1.7002751716732705, + "learning_rate": 3.8053277318393967e-06, + "loss": 0.995, + "step": 1856 + }, + { + "epoch": 0.1674708030842765, + "grad_norm": 1.4672389272303534, + "learning_rate": 3.805076241935498e-06, + "loss": 1.005, + "step": 1857 + }, + { + "epoch": 0.16756098660774676, + "grad_norm": 1.6133072599295541, + "learning_rate": 3.804824598012541e-06, + "loss": 0.8675, + "step": 1858 + }, + { + "epoch": 0.16765117013121703, + "grad_norm": 1.4199548688620716, + "learning_rate": 3.8045728000919975e-06, + "loss": 1.0303, + "step": 1859 + }, + { + "epoch": 0.1677413536546873, + "grad_norm": 1.525485886885164, + "learning_rate": 3.8043208481953524e-06, + "loss": 1.0276, + "step": 1860 + }, + { + "epoch": 0.16783153717815755, + "grad_norm": 1.797670337633407, + "learning_rate": 3.804068742344104e-06, + "loss": 0.924, + "step": 1861 + }, + { + "epoch": 0.1679217207016278, + "grad_norm": 1.595031298028369, + "learning_rate": 3.8038164825597628e-06, + "loss": 0.9618, + "step": 1862 + }, + { + "epoch": 0.16801190422509807, + "grad_norm": 1.7422976608687173, + "learning_rate": 3.8035640688638537e-06, + "loss": 1.0725, + "step": 1863 + }, + { + "epoch": 0.16810208774856833, + "grad_norm": 1.7597185439777479, + "learning_rate": 3.8033115012779125e-06, + "loss": 1.0725, + "step": 1864 + }, + { + "epoch": 0.1681922712720386, + "grad_norm": 1.4360084258817882, + "learning_rate": 3.8030587798234915e-06, + "loss": 1.0321, + "step": 1865 + }, + { + "epoch": 0.16828245479550885, + "grad_norm": 1.7129342607531637, + "learning_rate": 3.802805904522153e-06, + "loss": 0.9262, + "step": 1866 + }, + { + "epoch": 0.1683726383189791, + "grad_norm": 1.4353944866236648, + "learning_rate": 3.8025528753954742e-06, + "loss": 0.9863, + "step": 1867 + }, + { + "epoch": 0.16846282184244937, + "grad_norm": 2.3618702609278874, + "learning_rate": 3.802299692465045e-06, + "loss": 0.9388, + "step": 1868 + }, + { + "epoch": 0.16855300536591963, + "grad_norm": 1.7403535641527779, + "learning_rate": 3.802046355752468e-06, + "loss": 0.9803, + "step": 1869 + }, + { + "epoch": 0.16864318888938992, + "grad_norm": 3.0462436119641874, + "learning_rate": 3.80179286527936e-06, + "loss": 1.0493, + "step": 1870 + }, + { + "epoch": 0.16873337241286018, + "grad_norm": 1.8359991185013727, + "learning_rate": 3.801539221067349e-06, + "loss": 0.9082, + "step": 1871 + }, + { + "epoch": 0.16882355593633044, + "grad_norm": 1.7509978038310032, + "learning_rate": 3.801285423138079e-06, + "loss": 0.9557, + "step": 1872 + }, + { + "epoch": 0.1689137394598007, + "grad_norm": 2.451824645401232, + "learning_rate": 3.8010314715132037e-06, + "loss": 0.9479, + "step": 1873 + }, + { + "epoch": 0.16900392298327097, + "grad_norm": 1.7025286829208526, + "learning_rate": 3.800777366214393e-06, + "loss": 0.9464, + "step": 1874 + }, + { + "epoch": 0.16909410650674123, + "grad_norm": 2.172629794687066, + "learning_rate": 3.800523107263328e-06, + "loss": 0.9711, + "step": 1875 + }, + { + "epoch": 0.1691842900302115, + "grad_norm": 1.6498099304311005, + "learning_rate": 3.800268694681703e-06, + "loss": 1.0445, + "step": 1876 + }, + { + "epoch": 0.16927447355368175, + "grad_norm": 0.8066439501529906, + "learning_rate": 3.800014128491227e-06, + "loss": 0.8021, + "step": 1877 + }, + { + "epoch": 0.169364657077152, + "grad_norm": 1.5378492191351782, + "learning_rate": 3.79975940871362e-06, + "loss": 1.0427, + "step": 1878 + }, + { + "epoch": 0.16945484060062227, + "grad_norm": 2.8092272048110907, + "learning_rate": 3.799504535370617e-06, + "loss": 0.9946, + "step": 1879 + }, + { + "epoch": 0.16954502412409253, + "grad_norm": 2.004847968941924, + "learning_rate": 3.799249508483964e-06, + "loss": 0.9272, + "step": 1880 + }, + { + "epoch": 0.1696352076475628, + "grad_norm": 0.8991031005623473, + "learning_rate": 3.798994328075422e-06, + "loss": 0.8618, + "step": 1881 + }, + { + "epoch": 0.16972539117103305, + "grad_norm": 2.1194837441834533, + "learning_rate": 3.798738994166765e-06, + "loss": 0.9252, + "step": 1882 + }, + { + "epoch": 0.1698155746945033, + "grad_norm": 1.6226283152291978, + "learning_rate": 3.7984835067797788e-06, + "loss": 0.9951, + "step": 1883 + }, + { + "epoch": 0.16990575821797357, + "grad_norm": 1.8093147570870616, + "learning_rate": 3.798227865936263e-06, + "loss": 1.0846, + "step": 1884 + }, + { + "epoch": 0.16999594174144383, + "grad_norm": 1.5782348292440558, + "learning_rate": 3.7979720716580297e-06, + "loss": 0.8576, + "step": 1885 + }, + { + "epoch": 0.1700861252649141, + "grad_norm": 1.6053504621826518, + "learning_rate": 3.7977161239669057e-06, + "loss": 0.9824, + "step": 1886 + }, + { + "epoch": 0.17017630878838436, + "grad_norm": 4.003951981446152, + "learning_rate": 3.7974600228847294e-06, + "loss": 0.9718, + "step": 1887 + }, + { + "epoch": 0.17026649231185462, + "grad_norm": 3.0753512980445463, + "learning_rate": 3.7972037684333534e-06, + "loss": 0.9266, + "step": 1888 + }, + { + "epoch": 0.17035667583532488, + "grad_norm": 1.5160686983934109, + "learning_rate": 3.796947360634642e-06, + "loss": 0.9171, + "step": 1889 + }, + { + "epoch": 0.17044685935879514, + "grad_norm": 1.6604146599209033, + "learning_rate": 3.796690799510473e-06, + "loss": 1.0414, + "step": 1890 + }, + { + "epoch": 0.1705370428822654, + "grad_norm": 1.4639226433427173, + "learning_rate": 3.7964340850827387e-06, + "loss": 0.9524, + "step": 1891 + }, + { + "epoch": 0.17062722640573566, + "grad_norm": 0.7744098292808715, + "learning_rate": 3.7961772173733425e-06, + "loss": 0.7687, + "step": 1892 + }, + { + "epoch": 0.17071740992920592, + "grad_norm": 1.5252830086428117, + "learning_rate": 3.7959201964042024e-06, + "loss": 0.9781, + "step": 1893 + }, + { + "epoch": 0.1708075934526762, + "grad_norm": 1.6218084391494911, + "learning_rate": 3.795663022197248e-06, + "loss": 1.018, + "step": 1894 + }, + { + "epoch": 0.17089777697614647, + "grad_norm": 2.12578848627633, + "learning_rate": 3.7954056947744242e-06, + "loss": 0.9882, + "step": 1895 + }, + { + "epoch": 0.17098796049961673, + "grad_norm": 1.702512718522883, + "learning_rate": 3.7951482141576863e-06, + "loss": 0.9659, + "step": 1896 + }, + { + "epoch": 0.171078144023087, + "grad_norm": 1.43299058911893, + "learning_rate": 3.794890580369004e-06, + "loss": 1.0399, + "step": 1897 + }, + { + "epoch": 0.17116832754655725, + "grad_norm": 1.8955727771944677, + "learning_rate": 3.7946327934303612e-06, + "loss": 1.0283, + "step": 1898 + }, + { + "epoch": 0.1712585110700275, + "grad_norm": 1.6536411395469393, + "learning_rate": 3.794374853363752e-06, + "loss": 1.0873, + "step": 1899 + }, + { + "epoch": 0.17134869459349777, + "grad_norm": 1.4367036686946355, + "learning_rate": 3.794116760191187e-06, + "loss": 1.0152, + "step": 1900 + }, + { + "epoch": 0.17143887811696804, + "grad_norm": 1.6953151017270143, + "learning_rate": 3.7938585139346877e-06, + "loss": 1.0836, + "step": 1901 + }, + { + "epoch": 0.1715290616404383, + "grad_norm": 1.601818529459052, + "learning_rate": 3.793600114616288e-06, + "loss": 0.9728, + "step": 1902 + }, + { + "epoch": 0.17161924516390856, + "grad_norm": 0.8325414033852475, + "learning_rate": 3.793341562258037e-06, + "loss": 0.759, + "step": 1903 + }, + { + "epoch": 0.17170942868737882, + "grad_norm": 1.8882246306472863, + "learning_rate": 3.7930828568819953e-06, + "loss": 1.003, + "step": 1904 + }, + { + "epoch": 0.17179961221084908, + "grad_norm": 1.8410209242891984, + "learning_rate": 3.7928239985102378e-06, + "loss": 0.9894, + "step": 1905 + }, + { + "epoch": 0.17188979573431934, + "grad_norm": 1.7754315160694212, + "learning_rate": 3.7925649871648505e-06, + "loss": 0.991, + "step": 1906 + }, + { + "epoch": 0.1719799792577896, + "grad_norm": 1.487632950683559, + "learning_rate": 3.792305822867935e-06, + "loss": 0.9158, + "step": 1907 + }, + { + "epoch": 0.17207016278125986, + "grad_norm": 1.6465955812088895, + "learning_rate": 3.792046505641604e-06, + "loss": 0.9509, + "step": 1908 + }, + { + "epoch": 0.17216034630473012, + "grad_norm": 1.5809908236069, + "learning_rate": 3.791787035507984e-06, + "loss": 0.9517, + "step": 1909 + }, + { + "epoch": 0.17225052982820038, + "grad_norm": 4.055956926617463, + "learning_rate": 3.7915274124892136e-06, + "loss": 0.9507, + "step": 1910 + }, + { + "epoch": 0.17234071335167064, + "grad_norm": 1.4287831932877817, + "learning_rate": 3.7912676366074466e-06, + "loss": 0.9722, + "step": 1911 + }, + { + "epoch": 0.1724308968751409, + "grad_norm": 2.1345496177455505, + "learning_rate": 3.7910077078848478e-06, + "loss": 0.9517, + "step": 1912 + }, + { + "epoch": 0.17252108039861117, + "grad_norm": 1.996028294395239, + "learning_rate": 3.7907476263435957e-06, + "loss": 0.9552, + "step": 1913 + }, + { + "epoch": 0.17261126392208143, + "grad_norm": 1.7623327838763847, + "learning_rate": 3.7904873920058826e-06, + "loss": 0.9608, + "step": 1914 + }, + { + "epoch": 0.1727014474455517, + "grad_norm": 1.8443944419009355, + "learning_rate": 3.7902270048939114e-06, + "loss": 0.8635, + "step": 1915 + }, + { + "epoch": 0.17279163096902195, + "grad_norm": 2.2805797689713905, + "learning_rate": 3.7899664650299023e-06, + "loss": 0.9516, + "step": 1916 + }, + { + "epoch": 0.1728818144924922, + "grad_norm": 1.3648586383743395, + "learning_rate": 3.7897057724360836e-06, + "loss": 0.9993, + "step": 1917 + }, + { + "epoch": 0.1729719980159625, + "grad_norm": 2.10245231311256, + "learning_rate": 3.7894449271347004e-06, + "loss": 1.0077, + "step": 1918 + }, + { + "epoch": 0.17306218153943276, + "grad_norm": 1.4731865289888908, + "learning_rate": 3.789183929148009e-06, + "loss": 1.0463, + "step": 1919 + }, + { + "epoch": 0.17315236506290302, + "grad_norm": 1.5408811716039825, + "learning_rate": 3.7889227784982795e-06, + "loss": 1.013, + "step": 1920 + }, + { + "epoch": 0.17324254858637328, + "grad_norm": 1.4450741133113714, + "learning_rate": 3.7886614752077945e-06, + "loss": 0.9871, + "step": 1921 + }, + { + "epoch": 0.17333273210984354, + "grad_norm": 2.0398659712575675, + "learning_rate": 3.7884000192988495e-06, + "loss": 0.9919, + "step": 1922 + }, + { + "epoch": 0.1734229156333138, + "grad_norm": 2.0225935542061726, + "learning_rate": 3.7881384107937546e-06, + "loss": 1.0322, + "step": 1923 + }, + { + "epoch": 0.17351309915678406, + "grad_norm": 1.789908662948475, + "learning_rate": 3.78787664971483e-06, + "loss": 0.9875, + "step": 1924 + }, + { + "epoch": 0.17360328268025432, + "grad_norm": 1.4748530104506357, + "learning_rate": 3.7876147360844115e-06, + "loss": 0.9574, + "step": 1925 + }, + { + "epoch": 0.17369346620372458, + "grad_norm": 1.7418499264590914, + "learning_rate": 3.7873526699248474e-06, + "loss": 1.0604, + "step": 1926 + }, + { + "epoch": 0.17378364972719484, + "grad_norm": 6.212364100115377, + "learning_rate": 3.7870904512584974e-06, + "loss": 0.998, + "step": 1927 + }, + { + "epoch": 0.1738738332506651, + "grad_norm": 1.6018212086207395, + "learning_rate": 3.7868280801077368e-06, + "loss": 0.9876, + "step": 1928 + }, + { + "epoch": 0.17396401677413537, + "grad_norm": 1.3818568143851397, + "learning_rate": 3.7865655564949517e-06, + "loss": 0.9828, + "step": 1929 + }, + { + "epoch": 0.17405420029760563, + "grad_norm": 1.6240708555821743, + "learning_rate": 3.786302880442542e-06, + "loss": 0.9578, + "step": 1930 + }, + { + "epoch": 0.1741443838210759, + "grad_norm": 1.5627745577869465, + "learning_rate": 3.7860400519729215e-06, + "loss": 0.9868, + "step": 1931 + }, + { + "epoch": 0.17423456734454615, + "grad_norm": 1.7883864545561297, + "learning_rate": 3.7857770711085157e-06, + "loss": 1.0118, + "step": 1932 + }, + { + "epoch": 0.1743247508680164, + "grad_norm": 1.4828734181374856, + "learning_rate": 3.785513937871763e-06, + "loss": 1.0359, + "step": 1933 + }, + { + "epoch": 0.17441493439148667, + "grad_norm": 3.8980498646756123, + "learning_rate": 3.785250652285116e-06, + "loss": 0.9597, + "step": 1934 + }, + { + "epoch": 0.17450511791495693, + "grad_norm": 1.646822676288884, + "learning_rate": 3.78498721437104e-06, + "loss": 1.0636, + "step": 1935 + }, + { + "epoch": 0.1745953014384272, + "grad_norm": 1.9486492907601733, + "learning_rate": 3.784723624152012e-06, + "loss": 0.9826, + "step": 1936 + }, + { + "epoch": 0.17468548496189745, + "grad_norm": 1.9744321175694162, + "learning_rate": 3.784459881650524e-06, + "loss": 0.9332, + "step": 1937 + }, + { + "epoch": 0.1747756684853677, + "grad_norm": 1.6572840359758096, + "learning_rate": 3.784195986889079e-06, + "loss": 0.9637, + "step": 1938 + }, + { + "epoch": 0.17486585200883797, + "grad_norm": 1.724974855986075, + "learning_rate": 3.7839319398901946e-06, + "loss": 0.9102, + "step": 1939 + }, + { + "epoch": 0.17495603553230824, + "grad_norm": 1.9811328624670654, + "learning_rate": 3.7836677406764013e-06, + "loss": 1.0568, + "step": 1940 + }, + { + "epoch": 0.1750462190557785, + "grad_norm": 1.541527028676163, + "learning_rate": 3.7834033892702407e-06, + "loss": 1.0384, + "step": 1941 + }, + { + "epoch": 0.17513640257924878, + "grad_norm": 1.7490156675131012, + "learning_rate": 3.783138885694269e-06, + "loss": 0.952, + "step": 1942 + }, + { + "epoch": 0.17522658610271905, + "grad_norm": 2.0244173601034405, + "learning_rate": 3.7828742299710558e-06, + "loss": 1.065, + "step": 1943 + }, + { + "epoch": 0.1753167696261893, + "grad_norm": 1.7294025107105993, + "learning_rate": 3.782609422123183e-06, + "loss": 0.9659, + "step": 1944 + }, + { + "epoch": 0.17540695314965957, + "grad_norm": 2.1479489412543864, + "learning_rate": 3.7823444621732444e-06, + "loss": 0.9854, + "step": 1945 + }, + { + "epoch": 0.17549713667312983, + "grad_norm": 1.9608454587696325, + "learning_rate": 3.782079350143849e-06, + "loss": 1.0093, + "step": 1946 + }, + { + "epoch": 0.1755873201966001, + "grad_norm": 1.9633029485703397, + "learning_rate": 3.781814086057617e-06, + "loss": 1.0923, + "step": 1947 + }, + { + "epoch": 0.17567750372007035, + "grad_norm": 1.653848310274687, + "learning_rate": 3.7815486699371826e-06, + "loss": 0.979, + "step": 1948 + }, + { + "epoch": 0.1757676872435406, + "grad_norm": 1.5475486290459777, + "learning_rate": 3.7812831018051918e-06, + "loss": 0.9634, + "step": 1949 + }, + { + "epoch": 0.17585787076701087, + "grad_norm": 1.5859199861207751, + "learning_rate": 3.7810173816843058e-06, + "loss": 0.9967, + "step": 1950 + }, + { + "epoch": 0.17594805429048113, + "grad_norm": 1.5852978548636385, + "learning_rate": 3.7807515095971955e-06, + "loss": 1.0637, + "step": 1951 + }, + { + "epoch": 0.1760382378139514, + "grad_norm": 1.555492786813784, + "learning_rate": 3.7804854855665475e-06, + "loss": 0.9838, + "step": 1952 + }, + { + "epoch": 0.17612842133742165, + "grad_norm": 2.7152378592086692, + "learning_rate": 3.7802193096150606e-06, + "loss": 0.9972, + "step": 1953 + }, + { + "epoch": 0.17621860486089191, + "grad_norm": 1.476189864029589, + "learning_rate": 3.779952981765446e-06, + "loss": 0.9443, + "step": 1954 + }, + { + "epoch": 0.17630878838436218, + "grad_norm": 1.5456694712336105, + "learning_rate": 3.779686502040429e-06, + "loss": 1.0049, + "step": 1955 + }, + { + "epoch": 0.17639897190783244, + "grad_norm": 0.7874155135474467, + "learning_rate": 3.779419870462746e-06, + "loss": 0.777, + "step": 1956 + }, + { + "epoch": 0.1764891554313027, + "grad_norm": 2.061867472440329, + "learning_rate": 3.779153087055148e-06, + "loss": 0.9697, + "step": 1957 + }, + { + "epoch": 0.17657933895477296, + "grad_norm": 2.0889050808858416, + "learning_rate": 3.7788861518403988e-06, + "loss": 0.9916, + "step": 1958 + }, + { + "epoch": 0.17666952247824322, + "grad_norm": 1.742182504963236, + "learning_rate": 3.7786190648412742e-06, + "loss": 0.8848, + "step": 1959 + }, + { + "epoch": 0.17675970600171348, + "grad_norm": 0.8068326855355935, + "learning_rate": 3.778351826080564e-06, + "loss": 0.7891, + "step": 1960 + }, + { + "epoch": 0.17684988952518374, + "grad_norm": 1.4177577174337714, + "learning_rate": 3.7780844355810704e-06, + "loss": 1.0924, + "step": 1961 + }, + { + "epoch": 0.176940073048654, + "grad_norm": 1.5385387323343356, + "learning_rate": 3.777816893365608e-06, + "loss": 1.0012, + "step": 1962 + }, + { + "epoch": 0.17703025657212426, + "grad_norm": 1.4543128952057962, + "learning_rate": 3.7775491994570057e-06, + "loss": 1.0651, + "step": 1963 + }, + { + "epoch": 0.17712044009559452, + "grad_norm": 1.5788678979533475, + "learning_rate": 3.777281353878105e-06, + "loss": 0.9646, + "step": 1964 + }, + { + "epoch": 0.17721062361906478, + "grad_norm": 0.9836852669654483, + "learning_rate": 3.777013356651758e-06, + "loss": 0.796, + "step": 1965 + }, + { + "epoch": 0.17730080714253507, + "grad_norm": 1.6398831824822493, + "learning_rate": 3.776745207800834e-06, + "loss": 1.0044, + "step": 1966 + }, + { + "epoch": 0.17739099066600533, + "grad_norm": 1.8870464537926357, + "learning_rate": 3.7764769073482122e-06, + "loss": 1.0185, + "step": 1967 + }, + { + "epoch": 0.1774811741894756, + "grad_norm": 1.4714305036428943, + "learning_rate": 3.7762084553167846e-06, + "loss": 1.0549, + "step": 1968 + }, + { + "epoch": 0.17757135771294585, + "grad_norm": 1.7198275569480266, + "learning_rate": 3.775939851729458e-06, + "loss": 1.0847, + "step": 1969 + }, + { + "epoch": 0.17766154123641612, + "grad_norm": 1.971554045998416, + "learning_rate": 3.775671096609151e-06, + "loss": 0.9529, + "step": 1970 + }, + { + "epoch": 0.17775172475988638, + "grad_norm": 1.4905403992898962, + "learning_rate": 3.775402189978795e-06, + "loss": 0.9684, + "step": 1971 + }, + { + "epoch": 0.17784190828335664, + "grad_norm": 1.6655527525042584, + "learning_rate": 3.7751331318613343e-06, + "loss": 0.9722, + "step": 1972 + }, + { + "epoch": 0.1779320918068269, + "grad_norm": 1.5661689407306685, + "learning_rate": 3.774863922279727e-06, + "loss": 1.0141, + "step": 1973 + }, + { + "epoch": 0.17802227533029716, + "grad_norm": 1.4104794372309215, + "learning_rate": 3.7745945612569435e-06, + "loss": 1.0527, + "step": 1974 + }, + { + "epoch": 0.17811245885376742, + "grad_norm": 1.595404644882061, + "learning_rate": 3.7743250488159674e-06, + "loss": 0.8685, + "step": 1975 + }, + { + "epoch": 0.17820264237723768, + "grad_norm": 1.7381916151617682, + "learning_rate": 3.774055384979794e-06, + "loss": 0.8904, + "step": 1976 + }, + { + "epoch": 0.17829282590070794, + "grad_norm": 1.523585933398837, + "learning_rate": 3.773785569771433e-06, + "loss": 0.9756, + "step": 1977 + }, + { + "epoch": 0.1783830094241782, + "grad_norm": 1.900721051775509, + "learning_rate": 3.7735156032139066e-06, + "loss": 0.9737, + "step": 1978 + }, + { + "epoch": 0.17847319294764846, + "grad_norm": 1.5632071611423901, + "learning_rate": 3.773245485330251e-06, + "loss": 1.0938, + "step": 1979 + }, + { + "epoch": 0.17856337647111872, + "grad_norm": 1.530713201359312, + "learning_rate": 3.7729752161435115e-06, + "loss": 0.9583, + "step": 1980 + }, + { + "epoch": 0.17865355999458898, + "grad_norm": 1.690682624215611, + "learning_rate": 3.7727047956767514e-06, + "loss": 1.0544, + "step": 1981 + }, + { + "epoch": 0.17874374351805924, + "grad_norm": 1.4640455181883494, + "learning_rate": 3.7724342239530436e-06, + "loss": 1.122, + "step": 1982 + }, + { + "epoch": 0.1788339270415295, + "grad_norm": 1.5941842927305623, + "learning_rate": 3.772163500995474e-06, + "loss": 0.942, + "step": 1983 + }, + { + "epoch": 0.17892411056499977, + "grad_norm": 1.9900979846045337, + "learning_rate": 3.7718926268271437e-06, + "loss": 1.0838, + "step": 1984 + }, + { + "epoch": 0.17901429408847003, + "grad_norm": 1.8045615280244405, + "learning_rate": 3.771621601471164e-06, + "loss": 1.0093, + "step": 1985 + }, + { + "epoch": 0.1791044776119403, + "grad_norm": 1.776798240031436, + "learning_rate": 3.771350424950661e-06, + "loss": 1.0381, + "step": 1986 + }, + { + "epoch": 0.17919466113541055, + "grad_norm": 4.819350284141513, + "learning_rate": 3.771079097288772e-06, + "loss": 0.935, + "step": 1987 + }, + { + "epoch": 0.1792848446588808, + "grad_norm": 1.3370833362893573, + "learning_rate": 3.770807618508649e-06, + "loss": 1.0081, + "step": 1988 + }, + { + "epoch": 0.17937502818235107, + "grad_norm": 1.6717937128859004, + "learning_rate": 3.7705359886334555e-06, + "loss": 0.9809, + "step": 1989 + }, + { + "epoch": 0.17946521170582136, + "grad_norm": 1.9153786629981178, + "learning_rate": 3.7702642076863694e-06, + "loss": 0.971, + "step": 1990 + }, + { + "epoch": 0.17955539522929162, + "grad_norm": 1.6340771026163123, + "learning_rate": 3.7699922756905795e-06, + "loss": 1.0511, + "step": 1991 + }, + { + "epoch": 0.17964557875276188, + "grad_norm": 2.363368715291316, + "learning_rate": 3.7697201926692895e-06, + "loss": 0.9426, + "step": 1992 + }, + { + "epoch": 0.17973576227623214, + "grad_norm": 2.153245685002171, + "learning_rate": 3.7694479586457144e-06, + "loss": 1.0061, + "step": 1993 + }, + { + "epoch": 0.1798259457997024, + "grad_norm": 3.6080002581814083, + "learning_rate": 3.7691755736430827e-06, + "loss": 1.004, + "step": 1994 + }, + { + "epoch": 0.17991612932317266, + "grad_norm": 1.469641719156178, + "learning_rate": 3.768903037684636e-06, + "loss": 1.0371, + "step": 1995 + }, + { + "epoch": 0.18000631284664292, + "grad_norm": 1.5231659476028105, + "learning_rate": 3.7686303507936284e-06, + "loss": 1.0382, + "step": 1996 + }, + { + "epoch": 0.18009649637011318, + "grad_norm": 1.7676234455515891, + "learning_rate": 3.7683575129933272e-06, + "loss": 1.03, + "step": 1997 + }, + { + "epoch": 0.18018667989358345, + "grad_norm": 2.3838152667163786, + "learning_rate": 3.7680845243070128e-06, + "loss": 0.9948, + "step": 1998 + }, + { + "epoch": 0.1802768634170537, + "grad_norm": 0.8023268136054227, + "learning_rate": 3.7678113847579767e-06, + "loss": 0.7564, + "step": 1999 + }, + { + "epoch": 0.18036704694052397, + "grad_norm": 1.8373538134286562, + "learning_rate": 3.7675380943695264e-06, + "loss": 1.0619, + "step": 2000 + }, + { + "epoch": 0.18045723046399423, + "grad_norm": 1.8138364271503087, + "learning_rate": 3.7672646531649795e-06, + "loss": 0.9611, + "step": 2001 + }, + { + "epoch": 0.1805474139874645, + "grad_norm": 0.6794320426653672, + "learning_rate": 3.7669910611676682e-06, + "loss": 0.7792, + "step": 2002 + }, + { + "epoch": 0.18063759751093475, + "grad_norm": 1.6294534253618862, + "learning_rate": 3.7667173184009356e-06, + "loss": 0.98, + "step": 2003 + }, + { + "epoch": 0.180727781034405, + "grad_norm": 1.7725380788783096, + "learning_rate": 3.7664434248881403e-06, + "loss": 1.058, + "step": 2004 + }, + { + "epoch": 0.18081796455787527, + "grad_norm": 1.615999310516928, + "learning_rate": 3.766169380652652e-06, + "loss": 1.0057, + "step": 2005 + }, + { + "epoch": 0.18090814808134553, + "grad_norm": 1.5832733176384763, + "learning_rate": 3.7658951857178537e-06, + "loss": 1.038, + "step": 2006 + }, + { + "epoch": 0.1809983316048158, + "grad_norm": 1.554516040030909, + "learning_rate": 3.7656208401071414e-06, + "loss": 1.0054, + "step": 2007 + }, + { + "epoch": 0.18108851512828605, + "grad_norm": 1.9907061885164778, + "learning_rate": 3.7653463438439225e-06, + "loss": 0.9317, + "step": 2008 + }, + { + "epoch": 0.18117869865175631, + "grad_norm": 1.9249995813740857, + "learning_rate": 3.7650716969516203e-06, + "loss": 1.0073, + "step": 2009 + }, + { + "epoch": 0.18126888217522658, + "grad_norm": 1.8526071426357689, + "learning_rate": 3.764796899453668e-06, + "loss": 0.8971, + "step": 2010 + }, + { + "epoch": 0.18135906569869684, + "grad_norm": 1.9852739835894169, + "learning_rate": 3.7645219513735134e-06, + "loss": 1.0344, + "step": 2011 + }, + { + "epoch": 0.1814492492221671, + "grad_norm": 1.6653725924882135, + "learning_rate": 3.764246852734617e-06, + "loss": 1.0333, + "step": 2012 + }, + { + "epoch": 0.18153943274563739, + "grad_norm": 2.025347658619415, + "learning_rate": 3.7639716035604502e-06, + "loss": 0.9655, + "step": 2013 + }, + { + "epoch": 0.18162961626910765, + "grad_norm": 0.7464941657924898, + "learning_rate": 3.7636962038745e-06, + "loss": 0.751, + "step": 2014 + }, + { + "epoch": 0.1817197997925779, + "grad_norm": 1.7214528419579087, + "learning_rate": 3.763420653700265e-06, + "loss": 0.9525, + "step": 2015 + }, + { + "epoch": 0.18180998331604817, + "grad_norm": 1.6190186973467595, + "learning_rate": 3.7631449530612565e-06, + "loss": 0.953, + "step": 2016 + }, + { + "epoch": 0.18190016683951843, + "grad_norm": 1.6047465609757134, + "learning_rate": 3.762869101980999e-06, + "loss": 1.082, + "step": 2017 + }, + { + "epoch": 0.1819903503629887, + "grad_norm": 1.7594255294454924, + "learning_rate": 3.7625931004830287e-06, + "loss": 0.9957, + "step": 2018 + }, + { + "epoch": 0.18208053388645895, + "grad_norm": 1.6383460792881335, + "learning_rate": 3.7623169485908966e-06, + "loss": 0.9534, + "step": 2019 + }, + { + "epoch": 0.1821707174099292, + "grad_norm": 0.8420420774130737, + "learning_rate": 3.7620406463281647e-06, + "loss": 0.7978, + "step": 2020 + }, + { + "epoch": 0.18226090093339947, + "grad_norm": 1.8136895978506589, + "learning_rate": 3.7617641937184095e-06, + "loss": 1.016, + "step": 2021 + }, + { + "epoch": 0.18235108445686973, + "grad_norm": 1.777315646777302, + "learning_rate": 3.761487590785219e-06, + "loss": 1.026, + "step": 2022 + }, + { + "epoch": 0.18244126798034, + "grad_norm": 1.843369978194996, + "learning_rate": 3.7612108375521942e-06, + "loss": 0.9688, + "step": 2023 + }, + { + "epoch": 0.18253145150381025, + "grad_norm": 1.5230430679293274, + "learning_rate": 3.76093393404295e-06, + "loss": 1.04, + "step": 2024 + }, + { + "epoch": 0.18262163502728052, + "grad_norm": 1.7383124530863414, + "learning_rate": 3.7606568802811126e-06, + "loss": 0.9896, + "step": 2025 + }, + { + "epoch": 0.18271181855075078, + "grad_norm": 2.0727293482607316, + "learning_rate": 3.760379676290322e-06, + "loss": 1.0045, + "step": 2026 + }, + { + "epoch": 0.18280200207422104, + "grad_norm": 2.8190193064536984, + "learning_rate": 3.760102322094231e-06, + "loss": 0.9292, + "step": 2027 + }, + { + "epoch": 0.1828921855976913, + "grad_norm": 1.3910657205749628, + "learning_rate": 3.759824817716504e-06, + "loss": 0.9543, + "step": 2028 + }, + { + "epoch": 0.18298236912116156, + "grad_norm": 1.3387876747165781, + "learning_rate": 3.759547163180821e-06, + "loss": 1.0432, + "step": 2029 + }, + { + "epoch": 0.18307255264463182, + "grad_norm": 1.4412467889013854, + "learning_rate": 3.759269358510871e-06, + "loss": 0.7961, + "step": 2030 + }, + { + "epoch": 0.18316273616810208, + "grad_norm": 1.6543121517336246, + "learning_rate": 3.75899140373036e-06, + "loss": 0.8993, + "step": 2031 + }, + { + "epoch": 0.18325291969157234, + "grad_norm": 2.4563520065752584, + "learning_rate": 3.7587132988630028e-06, + "loss": 1.0634, + "step": 2032 + }, + { + "epoch": 0.1833431032150426, + "grad_norm": 2.063332591922033, + "learning_rate": 3.7584350439325295e-06, + "loss": 1.0296, + "step": 2033 + }, + { + "epoch": 0.18343328673851286, + "grad_norm": 1.6814996214975957, + "learning_rate": 3.758156638962682e-06, + "loss": 0.983, + "step": 2034 + }, + { + "epoch": 0.18352347026198312, + "grad_norm": 1.782200810494601, + "learning_rate": 3.757878083977216e-06, + "loss": 0.9601, + "step": 2035 + }, + { + "epoch": 0.18361365378545338, + "grad_norm": 1.4992115809759712, + "learning_rate": 3.7575993789999e-06, + "loss": 0.9749, + "step": 2036 + }, + { + "epoch": 0.18370383730892367, + "grad_norm": 1.5274843897544759, + "learning_rate": 3.757320524054512e-06, + "loss": 1.0295, + "step": 2037 + }, + { + "epoch": 0.18379402083239393, + "grad_norm": 1.5593163672300667, + "learning_rate": 3.757041519164848e-06, + "loss": 0.9245, + "step": 2038 + }, + { + "epoch": 0.1838842043558642, + "grad_norm": 2.269861509156881, + "learning_rate": 3.7567623643547133e-06, + "loss": 1.0008, + "step": 2039 + }, + { + "epoch": 0.18397438787933446, + "grad_norm": 1.5132982937429544, + "learning_rate": 3.756483059647927e-06, + "loss": 0.9925, + "step": 2040 + }, + { + "epoch": 0.18406457140280472, + "grad_norm": 1.6913535305907794, + "learning_rate": 3.756203605068321e-06, + "loss": 0.969, + "step": 2041 + }, + { + "epoch": 0.18415475492627498, + "grad_norm": 2.2631206483122233, + "learning_rate": 3.7559240006397396e-06, + "loss": 0.9865, + "step": 2042 + }, + { + "epoch": 0.18424493844974524, + "grad_norm": 1.5873044892314845, + "learning_rate": 3.7556442463860406e-06, + "loss": 1.1091, + "step": 2043 + }, + { + "epoch": 0.1843351219732155, + "grad_norm": 1.700446586292693, + "learning_rate": 3.7553643423310934e-06, + "loss": 1.1027, + "step": 2044 + }, + { + "epoch": 0.18442530549668576, + "grad_norm": 1.8603223422902067, + "learning_rate": 3.755084288498782e-06, + "loss": 0.8518, + "step": 2045 + }, + { + "epoch": 0.18451548902015602, + "grad_norm": 1.5100579022049037, + "learning_rate": 3.754804084913002e-06, + "loss": 0.9965, + "step": 2046 + }, + { + "epoch": 0.18460567254362628, + "grad_norm": 1.5603884352747583, + "learning_rate": 3.754523731597661e-06, + "loss": 1.06, + "step": 2047 + }, + { + "epoch": 0.18469585606709654, + "grad_norm": 1.8526035392130284, + "learning_rate": 3.754243228576681e-06, + "loss": 0.892, + "step": 2048 + }, + { + "epoch": 0.1847860395905668, + "grad_norm": 1.656754758681507, + "learning_rate": 3.753962575873996e-06, + "loss": 1.0481, + "step": 2049 + }, + { + "epoch": 0.18487622311403706, + "grad_norm": 1.6527687049353124, + "learning_rate": 3.7536817735135527e-06, + "loss": 0.9515, + "step": 2050 + }, + { + "epoch": 0.18496640663750732, + "grad_norm": 1.680273548176323, + "learning_rate": 3.753400821519311e-06, + "loss": 1.0357, + "step": 2051 + }, + { + "epoch": 0.18505659016097759, + "grad_norm": 1.3985112772825474, + "learning_rate": 3.7531197199152426e-06, + "loss": 0.9255, + "step": 2052 + }, + { + "epoch": 0.18514677368444785, + "grad_norm": 0.8056601784352398, + "learning_rate": 3.7528384687253335e-06, + "loss": 0.8196, + "step": 2053 + }, + { + "epoch": 0.1852369572079181, + "grad_norm": 1.7079879209953421, + "learning_rate": 3.7525570679735815e-06, + "loss": 1.1039, + "step": 2054 + }, + { + "epoch": 0.18532714073138837, + "grad_norm": 1.6140101820377584, + "learning_rate": 3.7522755176839965e-06, + "loss": 1.0083, + "step": 2055 + }, + { + "epoch": 0.18541732425485863, + "grad_norm": 1.77408792944848, + "learning_rate": 3.7519938178806027e-06, + "loss": 1.0007, + "step": 2056 + }, + { + "epoch": 0.1855075077783289, + "grad_norm": 2.2440628556204896, + "learning_rate": 3.7517119685874358e-06, + "loss": 0.9426, + "step": 2057 + }, + { + "epoch": 0.18559769130179915, + "grad_norm": 1.5396546130097037, + "learning_rate": 3.7514299698285447e-06, + "loss": 0.9818, + "step": 2058 + }, + { + "epoch": 0.1856878748252694, + "grad_norm": 1.4597496058376465, + "learning_rate": 3.751147821627991e-06, + "loss": 0.971, + "step": 2059 + }, + { + "epoch": 0.18577805834873967, + "grad_norm": 1.9654867587578755, + "learning_rate": 3.75086552400985e-06, + "loss": 1.0073, + "step": 2060 + }, + { + "epoch": 0.18586824187220996, + "grad_norm": 9.842128850262823, + "learning_rate": 3.750583076998208e-06, + "loss": 0.9669, + "step": 2061 + }, + { + "epoch": 0.18595842539568022, + "grad_norm": 1.7593621099348797, + "learning_rate": 3.7503004806171655e-06, + "loss": 0.9705, + "step": 2062 + }, + { + "epoch": 0.18604860891915048, + "grad_norm": 3.0236709070919954, + "learning_rate": 3.7500177348908354e-06, + "loss": 0.9241, + "step": 2063 + }, + { + "epoch": 0.18613879244262074, + "grad_norm": 2.456957793236954, + "learning_rate": 3.749734839843342e-06, + "loss": 1.018, + "step": 2064 + }, + { + "epoch": 0.186228975966091, + "grad_norm": 1.6251451720901866, + "learning_rate": 3.7494517954988245e-06, + "loss": 0.9956, + "step": 2065 + }, + { + "epoch": 0.18631915948956126, + "grad_norm": 1.6105083437836416, + "learning_rate": 3.749168601881433e-06, + "loss": 0.984, + "step": 2066 + }, + { + "epoch": 0.18640934301303153, + "grad_norm": 1.869360453435226, + "learning_rate": 3.7488852590153315e-06, + "loss": 0.9803, + "step": 2067 + }, + { + "epoch": 0.1864995265365018, + "grad_norm": 1.9139808481667333, + "learning_rate": 3.748601766924697e-06, + "loss": 0.9923, + "step": 2068 + }, + { + "epoch": 0.18658971005997205, + "grad_norm": 1.9501004413266019, + "learning_rate": 3.7483181256337176e-06, + "loss": 0.9608, + "step": 2069 + }, + { + "epoch": 0.1866798935834423, + "grad_norm": 1.5819714946999945, + "learning_rate": 3.7480343351665962e-06, + "loss": 0.9994, + "step": 2070 + }, + { + "epoch": 0.18677007710691257, + "grad_norm": 1.8616437774430443, + "learning_rate": 3.747750395547546e-06, + "loss": 0.9762, + "step": 2071 + }, + { + "epoch": 0.18686026063038283, + "grad_norm": 1.3271849053925404, + "learning_rate": 3.7474663068007956e-06, + "loss": 0.9764, + "step": 2072 + }, + { + "epoch": 0.1869504441538531, + "grad_norm": 1.3245769905129663, + "learning_rate": 3.747182068950584e-06, + "loss": 0.9739, + "step": 2073 + }, + { + "epoch": 0.18704062767732335, + "grad_norm": 1.5484754201670663, + "learning_rate": 3.7468976820211643e-06, + "loss": 0.9929, + "step": 2074 + }, + { + "epoch": 0.1871308112007936, + "grad_norm": 1.6171770325842554, + "learning_rate": 3.746613146036803e-06, + "loss": 0.9888, + "step": 2075 + }, + { + "epoch": 0.18722099472426387, + "grad_norm": 1.820302218297646, + "learning_rate": 3.7463284610217766e-06, + "loss": 1.0304, + "step": 2076 + }, + { + "epoch": 0.18731117824773413, + "grad_norm": 1.7367013673860543, + "learning_rate": 3.746043627000377e-06, + "loss": 0.9285, + "step": 2077 + }, + { + "epoch": 0.1874013617712044, + "grad_norm": 1.6565809099246283, + "learning_rate": 3.7457586439969076e-06, + "loss": 0.9475, + "step": 2078 + }, + { + "epoch": 0.18749154529467466, + "grad_norm": 1.3607162348272368, + "learning_rate": 3.7454735120356842e-06, + "loss": 0.9797, + "step": 2079 + }, + { + "epoch": 0.18758172881814492, + "grad_norm": 1.5201108165550115, + "learning_rate": 3.7451882311410373e-06, + "loss": 1.0551, + "step": 2080 + }, + { + "epoch": 0.18767191234161518, + "grad_norm": 1.4862530522680102, + "learning_rate": 3.7449028013373074e-06, + "loss": 1.0267, + "step": 2081 + }, + { + "epoch": 0.18776209586508544, + "grad_norm": 1.6830047497605807, + "learning_rate": 3.7446172226488485e-06, + "loss": 0.9906, + "step": 2082 + }, + { + "epoch": 0.1878522793885557, + "grad_norm": 1.7538877626392022, + "learning_rate": 3.7443314951000285e-06, + "loss": 1.0598, + "step": 2083 + }, + { + "epoch": 0.18794246291202596, + "grad_norm": 1.766757837977984, + "learning_rate": 3.7440456187152276e-06, + "loss": 0.9759, + "step": 2084 + }, + { + "epoch": 0.18803264643549625, + "grad_norm": 2.092557652999776, + "learning_rate": 3.7437595935188377e-06, + "loss": 1.0748, + "step": 2085 + }, + { + "epoch": 0.1881228299589665, + "grad_norm": 1.9974496196748088, + "learning_rate": 3.7434734195352647e-06, + "loss": 0.9922, + "step": 2086 + }, + { + "epoch": 0.18821301348243677, + "grad_norm": 0.8429753668298254, + "learning_rate": 3.743187096788926e-06, + "loss": 0.784, + "step": 2087 + }, + { + "epoch": 0.18830319700590703, + "grad_norm": 1.8215859996924417, + "learning_rate": 3.7429006253042524e-06, + "loss": 0.9918, + "step": 2088 + }, + { + "epoch": 0.1883933805293773, + "grad_norm": 3.9518553613611846, + "learning_rate": 3.7426140051056867e-06, + "loss": 0.9368, + "step": 2089 + }, + { + "epoch": 0.18848356405284755, + "grad_norm": 2.0241815679176267, + "learning_rate": 3.7423272362176856e-06, + "loss": 0.9315, + "step": 2090 + }, + { + "epoch": 0.1885737475763178, + "grad_norm": 1.6627137355081778, + "learning_rate": 3.742040318664718e-06, + "loss": 1.042, + "step": 2091 + }, + { + "epoch": 0.18866393109978807, + "grad_norm": 2.0847297057813017, + "learning_rate": 3.7417532524712643e-06, + "loss": 1.0467, + "step": 2092 + }, + { + "epoch": 0.18875411462325833, + "grad_norm": 1.697268013025147, + "learning_rate": 3.7414660376618195e-06, + "loss": 1.016, + "step": 2093 + }, + { + "epoch": 0.1888442981467286, + "grad_norm": 0.7830823771940557, + "learning_rate": 3.74117867426089e-06, + "loss": 0.7926, + "step": 2094 + }, + { + "epoch": 0.18893448167019886, + "grad_norm": 2.2024980018078404, + "learning_rate": 3.7408911622929954e-06, + "loss": 0.9689, + "step": 2095 + }, + { + "epoch": 0.18902466519366912, + "grad_norm": 1.7065505828469267, + "learning_rate": 3.740603501782668e-06, + "loss": 1.0268, + "step": 2096 + }, + { + "epoch": 0.18911484871713938, + "grad_norm": 1.68775436462281, + "learning_rate": 3.7403156927544516e-06, + "loss": 0.9272, + "step": 2097 + }, + { + "epoch": 0.18920503224060964, + "grad_norm": 2.1194799195544483, + "learning_rate": 3.740027735232904e-06, + "loss": 0.9232, + "step": 2098 + }, + { + "epoch": 0.1892952157640799, + "grad_norm": 1.9857538794170215, + "learning_rate": 3.7397396292425966e-06, + "loss": 0.9751, + "step": 2099 + }, + { + "epoch": 0.18938539928755016, + "grad_norm": 1.9001304079777728, + "learning_rate": 3.7394513748081105e-06, + "loss": 0.9658, + "step": 2100 + }, + { + "epoch": 0.18947558281102042, + "grad_norm": 1.5700267322961452, + "learning_rate": 3.7391629719540418e-06, + "loss": 1.0231, + "step": 2101 + }, + { + "epoch": 0.18956576633449068, + "grad_norm": 1.6090576451551166, + "learning_rate": 3.7388744207049998e-06, + "loss": 0.9784, + "step": 2102 + }, + { + "epoch": 0.18965594985796094, + "grad_norm": 2.4232678745704215, + "learning_rate": 3.7385857210856027e-06, + "loss": 0.9667, + "step": 2103 + }, + { + "epoch": 0.1897461333814312, + "grad_norm": 1.7345691606594884, + "learning_rate": 3.738296873120486e-06, + "loss": 0.9669, + "step": 2104 + }, + { + "epoch": 0.18983631690490146, + "grad_norm": 7.697976517778014, + "learning_rate": 3.7380078768342955e-06, + "loss": 0.9634, + "step": 2105 + }, + { + "epoch": 0.18992650042837173, + "grad_norm": 1.7647444968580221, + "learning_rate": 3.7377187322516895e-06, + "loss": 1.0147, + "step": 2106 + }, + { + "epoch": 0.19001668395184199, + "grad_norm": 1.9983159366048737, + "learning_rate": 3.7374294393973395e-06, + "loss": 0.9413, + "step": 2107 + }, + { + "epoch": 0.19010686747531225, + "grad_norm": 1.8950972822477319, + "learning_rate": 3.7371399982959294e-06, + "loss": 1.0423, + "step": 2108 + }, + { + "epoch": 0.19019705099878254, + "grad_norm": 2.1769516705820102, + "learning_rate": 3.7368504089721565e-06, + "loss": 0.9027, + "step": 2109 + }, + { + "epoch": 0.1902872345222528, + "grad_norm": 1.7978017656100982, + "learning_rate": 3.73656067145073e-06, + "loss": 0.9418, + "step": 2110 + }, + { + "epoch": 0.19037741804572306, + "grad_norm": 1.5622206628968613, + "learning_rate": 3.736270785756371e-06, + "loss": 1.0379, + "step": 2111 + }, + { + "epoch": 0.19046760156919332, + "grad_norm": 2.5334624993707275, + "learning_rate": 3.7359807519138156e-06, + "loss": 0.998, + "step": 2112 + }, + { + "epoch": 0.19055778509266358, + "grad_norm": 1.8634417002949004, + "learning_rate": 3.73569056994781e-06, + "loss": 0.9861, + "step": 2113 + }, + { + "epoch": 0.19064796861613384, + "grad_norm": 2.366358948792433, + "learning_rate": 3.7354002398831144e-06, + "loss": 1.027, + "step": 2114 + }, + { + "epoch": 0.1907381521396041, + "grad_norm": 3.627200806870137, + "learning_rate": 3.7351097617445015e-06, + "loss": 0.9911, + "step": 2115 + }, + { + "epoch": 0.19082833566307436, + "grad_norm": 0.8677911118264312, + "learning_rate": 3.7348191355567567e-06, + "loss": 0.8149, + "step": 2116 + }, + { + "epoch": 0.19091851918654462, + "grad_norm": 1.8504190253924384, + "learning_rate": 3.734528361344677e-06, + "loss": 1.0155, + "step": 2117 + }, + { + "epoch": 0.19100870271001488, + "grad_norm": 2.074529189964766, + "learning_rate": 3.734237439133074e-06, + "loss": 0.9865, + "step": 2118 + }, + { + "epoch": 0.19109888623348514, + "grad_norm": 1.8029297694885082, + "learning_rate": 3.7339463689467702e-06, + "loss": 1.0357, + "step": 2119 + }, + { + "epoch": 0.1911890697569554, + "grad_norm": 2.1247301491329487, + "learning_rate": 3.733655150810601e-06, + "loss": 0.9592, + "step": 2120 + }, + { + "epoch": 0.19127925328042567, + "grad_norm": 1.9044817868956982, + "learning_rate": 3.7333637847494154e-06, + "loss": 1.1234, + "step": 2121 + }, + { + "epoch": 0.19136943680389593, + "grad_norm": 1.7437172917405546, + "learning_rate": 3.7330722707880734e-06, + "loss": 1.1431, + "step": 2122 + }, + { + "epoch": 0.1914596203273662, + "grad_norm": 1.4449680227153547, + "learning_rate": 3.7327806089514497e-06, + "loss": 0.8974, + "step": 2123 + }, + { + "epoch": 0.19154980385083645, + "grad_norm": 1.7284305185174582, + "learning_rate": 3.7324887992644297e-06, + "loss": 1.0865, + "step": 2124 + }, + { + "epoch": 0.1916399873743067, + "grad_norm": 1.0289666878597366, + "learning_rate": 3.7321968417519123e-06, + "loss": 0.7877, + "step": 2125 + }, + { + "epoch": 0.19173017089777697, + "grad_norm": 1.6296216093734577, + "learning_rate": 3.7319047364388097e-06, + "loss": 0.9817, + "step": 2126 + }, + { + "epoch": 0.19182035442124723, + "grad_norm": 0.7143013569277036, + "learning_rate": 3.7316124833500453e-06, + "loss": 0.774, + "step": 2127 + }, + { + "epoch": 0.1919105379447175, + "grad_norm": 1.7575869267157622, + "learning_rate": 3.731320082510556e-06, + "loss": 0.9727, + "step": 2128 + }, + { + "epoch": 0.19200072146818775, + "grad_norm": 1.632503279467994, + "learning_rate": 3.7310275339452906e-06, + "loss": 1.0013, + "step": 2129 + }, + { + "epoch": 0.192090904991658, + "grad_norm": 1.9680552392045674, + "learning_rate": 3.7307348376792113e-06, + "loss": 0.9761, + "step": 2130 + }, + { + "epoch": 0.19218108851512827, + "grad_norm": 1.8024395570395713, + "learning_rate": 3.730441993737292e-06, + "loss": 0.9796, + "step": 2131 + }, + { + "epoch": 0.19227127203859856, + "grad_norm": 2.7914504826284405, + "learning_rate": 3.7301490021445205e-06, + "loss": 0.9736, + "step": 2132 + }, + { + "epoch": 0.19236145556206882, + "grad_norm": 2.1175193385802817, + "learning_rate": 3.7298558629258966e-06, + "loss": 0.9257, + "step": 2133 + }, + { + "epoch": 0.19245163908553908, + "grad_norm": 1.529034003688615, + "learning_rate": 3.7295625761064314e-06, + "loss": 1.0211, + "step": 2134 + }, + { + "epoch": 0.19254182260900934, + "grad_norm": 3.4402790280222058, + "learning_rate": 3.7292691417111504e-06, + "loss": 0.9536, + "step": 2135 + }, + { + "epoch": 0.1926320061324796, + "grad_norm": 1.8186084154738749, + "learning_rate": 3.728975559765092e-06, + "loss": 0.9457, + "step": 2136 + }, + { + "epoch": 0.19272218965594987, + "grad_norm": 2.3591385274448817, + "learning_rate": 3.728681830293305e-06, + "loss": 1.0172, + "step": 2137 + }, + { + "epoch": 0.19281237317942013, + "grad_norm": 2.212272645292753, + "learning_rate": 3.7283879533208523e-06, + "loss": 0.9499, + "step": 2138 + }, + { + "epoch": 0.1929025567028904, + "grad_norm": 1.7526305046955504, + "learning_rate": 3.7280939288728094e-06, + "loss": 0.9728, + "step": 2139 + }, + { + "epoch": 0.19299274022636065, + "grad_norm": 1.8544657033884302, + "learning_rate": 3.7277997569742637e-06, + "loss": 0.9482, + "step": 2140 + }, + { + "epoch": 0.1930829237498309, + "grad_norm": 2.411149029157697, + "learning_rate": 3.7275054376503155e-06, + "loss": 0.9501, + "step": 2141 + }, + { + "epoch": 0.19317310727330117, + "grad_norm": 2.8824639665348086, + "learning_rate": 3.7272109709260783e-06, + "loss": 0.9291, + "step": 2142 + }, + { + "epoch": 0.19326329079677143, + "grad_norm": 0.9432597611238454, + "learning_rate": 3.7269163568266774e-06, + "loss": 0.7904, + "step": 2143 + }, + { + "epoch": 0.1933534743202417, + "grad_norm": 1.8368325506901555, + "learning_rate": 3.7266215953772512e-06, + "loss": 0.9708, + "step": 2144 + }, + { + "epoch": 0.19344365784371195, + "grad_norm": 2.280551646739456, + "learning_rate": 3.7263266866029492e-06, + "loss": 1.0223, + "step": 2145 + }, + { + "epoch": 0.1935338413671822, + "grad_norm": 1.7390475180804006, + "learning_rate": 3.726031630528936e-06, + "loss": 0.9535, + "step": 2146 + }, + { + "epoch": 0.19362402489065247, + "grad_norm": 1.501203451586363, + "learning_rate": 3.7257364271803865e-06, + "loss": 0.8786, + "step": 2147 + }, + { + "epoch": 0.19371420841412273, + "grad_norm": 2.0406630508123293, + "learning_rate": 3.7254410765824896e-06, + "loss": 0.9102, + "step": 2148 + }, + { + "epoch": 0.193804391937593, + "grad_norm": 1.815245357879978, + "learning_rate": 3.725145578760446e-06, + "loss": 0.9053, + "step": 2149 + }, + { + "epoch": 0.19389457546106326, + "grad_norm": 2.4520471233675023, + "learning_rate": 3.7248499337394696e-06, + "loss": 0.9013, + "step": 2150 + }, + { + "epoch": 0.19398475898453352, + "grad_norm": 1.7626718450870327, + "learning_rate": 3.7245541415447848e-06, + "loss": 1.0719, + "step": 2151 + }, + { + "epoch": 0.19407494250800378, + "grad_norm": 1.8382964218298226, + "learning_rate": 3.724258202201633e-06, + "loss": 1.0061, + "step": 2152 + }, + { + "epoch": 0.19416512603147404, + "grad_norm": 1.9854948233251397, + "learning_rate": 3.7239621157352633e-06, + "loss": 0.9975, + "step": 2153 + }, + { + "epoch": 0.1942553095549443, + "grad_norm": 1.8983196959816477, + "learning_rate": 3.7236658821709403e-06, + "loss": 0.9723, + "step": 2154 + }, + { + "epoch": 0.19434549307841456, + "grad_norm": 1.7000498876544523, + "learning_rate": 3.7233695015339404e-06, + "loss": 1.0232, + "step": 2155 + }, + { + "epoch": 0.19443567660188485, + "grad_norm": 1.798347201760843, + "learning_rate": 3.7230729738495513e-06, + "loss": 0.9783, + "step": 2156 + }, + { + "epoch": 0.1945258601253551, + "grad_norm": 1.777671431632278, + "learning_rate": 3.722776299143075e-06, + "loss": 0.9803, + "step": 2157 + }, + { + "epoch": 0.19461604364882537, + "grad_norm": 1.0829676353218793, + "learning_rate": 3.722479477439826e-06, + "loss": 0.7893, + "step": 2158 + }, + { + "epoch": 0.19470622717229563, + "grad_norm": 1.8362209669379683, + "learning_rate": 3.7221825087651306e-06, + "loss": 0.9602, + "step": 2159 + }, + { + "epoch": 0.1947964106957659, + "grad_norm": 2.145716160905297, + "learning_rate": 3.7218853931443274e-06, + "loss": 0.9207, + "step": 2160 + }, + { + "epoch": 0.19488659421923615, + "grad_norm": 2.619769380043644, + "learning_rate": 3.721588130602768e-06, + "loss": 0.9255, + "step": 2161 + }, + { + "epoch": 0.19497677774270641, + "grad_norm": 3.867135095722902, + "learning_rate": 3.7212907211658164e-06, + "loss": 0.9556, + "step": 2162 + }, + { + "epoch": 0.19506696126617667, + "grad_norm": 1.6194313441743255, + "learning_rate": 3.72099316485885e-06, + "loss": 0.9525, + "step": 2163 + }, + { + "epoch": 0.19515714478964694, + "grad_norm": 1.7772782869125643, + "learning_rate": 3.720695461707256e-06, + "loss": 1.0162, + "step": 2164 + }, + { + "epoch": 0.1952473283131172, + "grad_norm": 1.5235712237233725, + "learning_rate": 3.7203976117364383e-06, + "loss": 0.998, + "step": 2165 + }, + { + "epoch": 0.19533751183658746, + "grad_norm": 2.5426714323419572, + "learning_rate": 3.7200996149718105e-06, + "loss": 0.963, + "step": 2166 + }, + { + "epoch": 0.19542769536005772, + "grad_norm": 1.4986174092569466, + "learning_rate": 3.7198014714387985e-06, + "loss": 0.971, + "step": 2167 + }, + { + "epoch": 0.19551787888352798, + "grad_norm": 1.982087987584644, + "learning_rate": 3.7195031811628422e-06, + "loss": 0.9151, + "step": 2168 + }, + { + "epoch": 0.19560806240699824, + "grad_norm": 2.0698668954044273, + "learning_rate": 3.719204744169393e-06, + "loss": 1.0128, + "step": 2169 + }, + { + "epoch": 0.1956982459304685, + "grad_norm": 2.6632859337145307, + "learning_rate": 3.718906160483916e-06, + "loss": 0.9569, + "step": 2170 + }, + { + "epoch": 0.19578842945393876, + "grad_norm": 1.640462013504755, + "learning_rate": 3.7186074301318868e-06, + "loss": 0.9534, + "step": 2171 + }, + { + "epoch": 0.19587861297740902, + "grad_norm": 1.4593569932108534, + "learning_rate": 3.7183085531387957e-06, + "loss": 1.0542, + "step": 2172 + }, + { + "epoch": 0.19596879650087928, + "grad_norm": 1.954262242152341, + "learning_rate": 3.7180095295301443e-06, + "loss": 0.9593, + "step": 2173 + }, + { + "epoch": 0.19605898002434954, + "grad_norm": 1.3606704590071879, + "learning_rate": 3.7177103593314465e-06, + "loss": 1.0454, + "step": 2174 + }, + { + "epoch": 0.1961491635478198, + "grad_norm": 7.270000887402424, + "learning_rate": 3.7174110425682297e-06, + "loss": 1.0166, + "step": 2175 + }, + { + "epoch": 0.19623934707129007, + "grad_norm": 1.6209507155224399, + "learning_rate": 3.7171115792660333e-06, + "loss": 1.1095, + "step": 2176 + }, + { + "epoch": 0.19632953059476033, + "grad_norm": 1.6103699210596951, + "learning_rate": 3.7168119694504083e-06, + "loss": 1.044, + "step": 2177 + }, + { + "epoch": 0.1964197141182306, + "grad_norm": 1.7100059028434373, + "learning_rate": 3.71651221314692e-06, + "loss": 0.9773, + "step": 2178 + }, + { + "epoch": 0.19650989764170085, + "grad_norm": 1.656095389581195, + "learning_rate": 3.716212310381145e-06, + "loss": 0.9763, + "step": 2179 + }, + { + "epoch": 0.19660008116517114, + "grad_norm": 1.829797485068975, + "learning_rate": 3.7159122611786725e-06, + "loss": 1.0286, + "step": 2180 + }, + { + "epoch": 0.1966902646886414, + "grad_norm": 1.803859702493153, + "learning_rate": 3.7156120655651045e-06, + "loss": 0.9663, + "step": 2181 + }, + { + "epoch": 0.19678044821211166, + "grad_norm": 1.4618632601445574, + "learning_rate": 3.7153117235660553e-06, + "loss": 0.9638, + "step": 2182 + }, + { + "epoch": 0.19687063173558192, + "grad_norm": 0.8099004166811733, + "learning_rate": 3.7150112352071514e-06, + "loss": 0.7871, + "step": 2183 + }, + { + "epoch": 0.19696081525905218, + "grad_norm": 6.824258717953222, + "learning_rate": 3.7147106005140326e-06, + "loss": 0.9549, + "step": 2184 + }, + { + "epoch": 0.19705099878252244, + "grad_norm": 0.7347311414366744, + "learning_rate": 3.714409819512351e-06, + "loss": 0.7965, + "step": 2185 + }, + { + "epoch": 0.1971411823059927, + "grad_norm": 1.6074915114159365, + "learning_rate": 3.7141088922277695e-06, + "loss": 1.0105, + "step": 2186 + }, + { + "epoch": 0.19723136582946296, + "grad_norm": 1.8014709078821676, + "learning_rate": 3.7138078186859664e-06, + "loss": 1.029, + "step": 2187 + }, + { + "epoch": 0.19732154935293322, + "grad_norm": 1.5791182461094935, + "learning_rate": 3.7135065989126303e-06, + "loss": 0.9963, + "step": 2188 + }, + { + "epoch": 0.19741173287640348, + "grad_norm": 0.8774889560015345, + "learning_rate": 3.713205232933463e-06, + "loss": 0.8509, + "step": 2189 + }, + { + "epoch": 0.19750191639987374, + "grad_norm": 1.581415654729161, + "learning_rate": 3.7129037207741792e-06, + "loss": 0.9901, + "step": 2190 + }, + { + "epoch": 0.197592099923344, + "grad_norm": 1.6584055566955014, + "learning_rate": 3.7126020624605046e-06, + "loss": 0.9942, + "step": 2191 + }, + { + "epoch": 0.19768228344681427, + "grad_norm": 1.5333868414761431, + "learning_rate": 3.7123002580181785e-06, + "loss": 0.9389, + "step": 2192 + }, + { + "epoch": 0.19777246697028453, + "grad_norm": 1.3040379260609278, + "learning_rate": 3.7119983074729532e-06, + "loss": 1.0542, + "step": 2193 + }, + { + "epoch": 0.1978626504937548, + "grad_norm": 1.8918432929082354, + "learning_rate": 3.7116962108505926e-06, + "loss": 1.0383, + "step": 2194 + }, + { + "epoch": 0.19795283401722505, + "grad_norm": 1.5078668930441679, + "learning_rate": 3.711393968176873e-06, + "loss": 1.0822, + "step": 2195 + }, + { + "epoch": 0.1980430175406953, + "grad_norm": 1.3160249061517626, + "learning_rate": 3.711091579477584e-06, + "loss": 0.9937, + "step": 2196 + }, + { + "epoch": 0.19813320106416557, + "grad_norm": 1.9234018231080514, + "learning_rate": 3.7107890447785255e-06, + "loss": 1.0144, + "step": 2197 + }, + { + "epoch": 0.19822338458763583, + "grad_norm": 1.7318355768095162, + "learning_rate": 3.710486364105513e-06, + "loss": 1.036, + "step": 2198 + }, + { + "epoch": 0.1983135681111061, + "grad_norm": 1.7602741552450105, + "learning_rate": 3.7101835374843728e-06, + "loss": 0.9817, + "step": 2199 + }, + { + "epoch": 0.19840375163457635, + "grad_norm": 1.599845747664693, + "learning_rate": 3.7098805649409427e-06, + "loss": 1.0035, + "step": 2200 + }, + { + "epoch": 0.1984939351580466, + "grad_norm": 1.8013428579799124, + "learning_rate": 3.7095774465010748e-06, + "loss": 1.0519, + "step": 2201 + }, + { + "epoch": 0.19858411868151687, + "grad_norm": 1.5785908530778006, + "learning_rate": 3.7092741821906328e-06, + "loss": 0.984, + "step": 2202 + }, + { + "epoch": 0.19867430220498714, + "grad_norm": 1.761754003878315, + "learning_rate": 3.708970772035493e-06, + "loss": 0.9676, + "step": 2203 + }, + { + "epoch": 0.19876448572845742, + "grad_norm": 1.468290703566376, + "learning_rate": 3.7086672160615427e-06, + "loss": 0.9515, + "step": 2204 + }, + { + "epoch": 0.19885466925192768, + "grad_norm": 1.4534711528024244, + "learning_rate": 3.7083635142946852e-06, + "loss": 0.969, + "step": 2205 + }, + { + "epoch": 0.19894485277539795, + "grad_norm": 1.5157549300984232, + "learning_rate": 3.7080596667608327e-06, + "loss": 1.0164, + "step": 2206 + }, + { + "epoch": 0.1990350362988682, + "grad_norm": 1.5428494564248418, + "learning_rate": 3.707755673485911e-06, + "loss": 1.0288, + "step": 2207 + }, + { + "epoch": 0.19912521982233847, + "grad_norm": 2.233345061124308, + "learning_rate": 3.7074515344958584e-06, + "loss": 0.8968, + "step": 2208 + }, + { + "epoch": 0.19921540334580873, + "grad_norm": 0.7672091075084158, + "learning_rate": 3.707147249816627e-06, + "loss": 0.8164, + "step": 2209 + }, + { + "epoch": 0.199305586869279, + "grad_norm": 1.4381358357551381, + "learning_rate": 3.706842819474178e-06, + "loss": 1.0084, + "step": 2210 + }, + { + "epoch": 0.19939577039274925, + "grad_norm": 1.6492233211249787, + "learning_rate": 3.706538243494489e-06, + "loss": 0.9219, + "step": 2211 + }, + { + "epoch": 0.1994859539162195, + "grad_norm": 1.7191740033203815, + "learning_rate": 3.706233521903547e-06, + "loss": 0.9586, + "step": 2212 + }, + { + "epoch": 0.19957613743968977, + "grad_norm": 1.6992637189855984, + "learning_rate": 3.705928654727353e-06, + "loss": 1.0645, + "step": 2213 + }, + { + "epoch": 0.19966632096316003, + "grad_norm": 1.4375467707447342, + "learning_rate": 3.7056236419919195e-06, + "loss": 0.961, + "step": 2214 + }, + { + "epoch": 0.1997565044866303, + "grad_norm": 1.3518271655659735, + "learning_rate": 3.705318483723273e-06, + "loss": 1.0357, + "step": 2215 + }, + { + "epoch": 0.19984668801010055, + "grad_norm": 1.9158072479826171, + "learning_rate": 3.7050131799474493e-06, + "loss": 0.9617, + "step": 2216 + }, + { + "epoch": 0.19993687153357081, + "grad_norm": 1.805247562581662, + "learning_rate": 3.7047077306905e-06, + "loss": 0.9812, + "step": 2217 + }, + { + "epoch": 0.20002705505704108, + "grad_norm": 1.5350207397600628, + "learning_rate": 3.704402135978488e-06, + "loss": 1.0247, + "step": 2218 + }, + { + "epoch": 0.20011723858051134, + "grad_norm": 2.293461761242328, + "learning_rate": 3.7040963958374877e-06, + "loss": 1.0172, + "step": 2219 + }, + { + "epoch": 0.2002074221039816, + "grad_norm": 1.4928655232385277, + "learning_rate": 3.7037905102935864e-06, + "loss": 1.017, + "step": 2220 + }, + { + "epoch": 0.20029760562745186, + "grad_norm": 1.4849147417659105, + "learning_rate": 3.7034844793728837e-06, + "loss": 0.9926, + "step": 2221 + }, + { + "epoch": 0.20038778915092212, + "grad_norm": 1.7511131969443248, + "learning_rate": 3.7031783031014933e-06, + "loss": 0.8732, + "step": 2222 + }, + { + "epoch": 0.20047797267439238, + "grad_norm": 1.5481201708968195, + "learning_rate": 3.702871981505538e-06, + "loss": 0.9641, + "step": 2223 + }, + { + "epoch": 0.20056815619786264, + "grad_norm": 1.694145539329179, + "learning_rate": 3.7025655146111563e-06, + "loss": 0.926, + "step": 2224 + }, + { + "epoch": 0.2006583397213329, + "grad_norm": 0.8020119490791638, + "learning_rate": 3.702258902444497e-06, + "loss": 0.8671, + "step": 2225 + }, + { + "epoch": 0.20074852324480316, + "grad_norm": 1.5126371526404128, + "learning_rate": 3.701952145031722e-06, + "loss": 1.0413, + "step": 2226 + }, + { + "epoch": 0.20083870676827342, + "grad_norm": 1.4543238790762754, + "learning_rate": 3.701645242399005e-06, + "loss": 1.069, + "step": 2227 + }, + { + "epoch": 0.2009288902917437, + "grad_norm": 1.4240138138197476, + "learning_rate": 3.701338194572533e-06, + "loss": 1.0549, + "step": 2228 + }, + { + "epoch": 0.20101907381521397, + "grad_norm": 1.457930463820276, + "learning_rate": 3.7010310015785056e-06, + "loss": 0.9633, + "step": 2229 + }, + { + "epoch": 0.20110925733868423, + "grad_norm": 1.4471741929304287, + "learning_rate": 3.700723663443134e-06, + "loss": 0.9297, + "step": 2230 + }, + { + "epoch": 0.2011994408621545, + "grad_norm": 1.138897410544378, + "learning_rate": 3.7004161801926416e-06, + "loss": 0.7942, + "step": 2231 + }, + { + "epoch": 0.20128962438562475, + "grad_norm": 1.582764255583004, + "learning_rate": 3.7001085518532643e-06, + "loss": 1.0095, + "step": 2232 + }, + { + "epoch": 0.20137980790909502, + "grad_norm": 1.5909008663757562, + "learning_rate": 3.6998007784512515e-06, + "loss": 0.983, + "step": 2233 + }, + { + "epoch": 0.20146999143256528, + "grad_norm": 1.3441649838551541, + "learning_rate": 3.6994928600128637e-06, + "loss": 0.8651, + "step": 2234 + }, + { + "epoch": 0.20156017495603554, + "grad_norm": 1.431621223516004, + "learning_rate": 3.6991847965643742e-06, + "loss": 0.9526, + "step": 2235 + }, + { + "epoch": 0.2016503584795058, + "grad_norm": 1.5141916477799846, + "learning_rate": 3.698876588132068e-06, + "loss": 0.9405, + "step": 2236 + }, + { + "epoch": 0.20174054200297606, + "grad_norm": 1.4267562250001864, + "learning_rate": 3.6985682347422446e-06, + "loss": 0.9473, + "step": 2237 + }, + { + "epoch": 0.20183072552644632, + "grad_norm": 1.7851233281077712, + "learning_rate": 3.698259736421213e-06, + "loss": 0.9965, + "step": 2238 + }, + { + "epoch": 0.20192090904991658, + "grad_norm": 1.9555996895575867, + "learning_rate": 3.697951093195297e-06, + "loss": 0.9641, + "step": 2239 + }, + { + "epoch": 0.20201109257338684, + "grad_norm": 1.462853221135013, + "learning_rate": 3.6976423050908307e-06, + "loss": 0.9629, + "step": 2240 + }, + { + "epoch": 0.2021012760968571, + "grad_norm": 1.3880035895540286, + "learning_rate": 3.697333372134163e-06, + "loss": 0.9712, + "step": 2241 + }, + { + "epoch": 0.20219145962032736, + "grad_norm": 1.6413207259268543, + "learning_rate": 3.697024294351653e-06, + "loss": 1.0765, + "step": 2242 + }, + { + "epoch": 0.20228164314379762, + "grad_norm": 1.5580035552595166, + "learning_rate": 3.696715071769672e-06, + "loss": 0.9428, + "step": 2243 + }, + { + "epoch": 0.20237182666726788, + "grad_norm": 1.6688853834005068, + "learning_rate": 3.696405704414606e-06, + "loss": 0.9826, + "step": 2244 + }, + { + "epoch": 0.20246201019073815, + "grad_norm": 1.4411979049174009, + "learning_rate": 3.6960961923128514e-06, + "loss": 1.0893, + "step": 2245 + }, + { + "epoch": 0.2025521937142084, + "grad_norm": 1.953299735835995, + "learning_rate": 3.6957865354908177e-06, + "loss": 0.9173, + "step": 2246 + }, + { + "epoch": 0.20264237723767867, + "grad_norm": 1.6016787603095717, + "learning_rate": 3.6954767339749262e-06, + "loss": 1.0001, + "step": 2247 + }, + { + "epoch": 0.20273256076114893, + "grad_norm": 1.8479890745384961, + "learning_rate": 3.6951667877916113e-06, + "loss": 0.9601, + "step": 2248 + }, + { + "epoch": 0.2028227442846192, + "grad_norm": 1.4519904845375193, + "learning_rate": 3.694856696967319e-06, + "loss": 0.9893, + "step": 2249 + }, + { + "epoch": 0.20291292780808945, + "grad_norm": 1.8290697574650727, + "learning_rate": 3.6945464615285077e-06, + "loss": 0.9695, + "step": 2250 + }, + { + "epoch": 0.2030031113315597, + "grad_norm": 0.7759615624238438, + "learning_rate": 3.694236081501648e-06, + "loss": 0.8945, + "step": 2251 + }, + { + "epoch": 0.20309329485503, + "grad_norm": 1.444976355151326, + "learning_rate": 3.6939255569132246e-06, + "loss": 0.9454, + "step": 2252 + }, + { + "epoch": 0.20318347837850026, + "grad_norm": 0.893020320013971, + "learning_rate": 3.693614887789733e-06, + "loss": 0.8218, + "step": 2253 + }, + { + "epoch": 0.20327366190197052, + "grad_norm": 0.8347411820184995, + "learning_rate": 3.69330407415768e-06, + "loss": 0.823, + "step": 2254 + }, + { + "epoch": 0.20336384542544078, + "grad_norm": 1.704421014000267, + "learning_rate": 3.6929931160435867e-06, + "loss": 0.9924, + "step": 2255 + }, + { + "epoch": 0.20345402894891104, + "grad_norm": 1.6920229027778495, + "learning_rate": 3.6926820134739858e-06, + "loss": 1.0448, + "step": 2256 + }, + { + "epoch": 0.2035442124723813, + "grad_norm": 1.6388118352215446, + "learning_rate": 3.692370766475422e-06, + "loss": 0.9191, + "step": 2257 + }, + { + "epoch": 0.20363439599585156, + "grad_norm": 6.329517955293809, + "learning_rate": 3.692059375074453e-06, + "loss": 1.0713, + "step": 2258 + }, + { + "epoch": 0.20372457951932182, + "grad_norm": 1.4873513347802225, + "learning_rate": 3.6917478392976475e-06, + "loss": 1.0492, + "step": 2259 + }, + { + "epoch": 0.20381476304279209, + "grad_norm": 1.6621833901337493, + "learning_rate": 3.691436159171589e-06, + "loss": 1.0511, + "step": 2260 + }, + { + "epoch": 0.20390494656626235, + "grad_norm": 1.612228530388294, + "learning_rate": 3.6911243347228703e-06, + "loss": 0.9655, + "step": 2261 + }, + { + "epoch": 0.2039951300897326, + "grad_norm": 0.7478444594979672, + "learning_rate": 3.690812365978099e-06, + "loss": 0.7359, + "step": 2262 + }, + { + "epoch": 0.20408531361320287, + "grad_norm": 1.5936321888183342, + "learning_rate": 3.690500252963893e-06, + "loss": 0.9436, + "step": 2263 + }, + { + "epoch": 0.20417549713667313, + "grad_norm": 1.7485391104918986, + "learning_rate": 3.6901879957068846e-06, + "loss": 0.978, + "step": 2264 + }, + { + "epoch": 0.2042656806601434, + "grad_norm": 1.5550215472479458, + "learning_rate": 3.689875594233717e-06, + "loss": 1.0732, + "step": 2265 + }, + { + "epoch": 0.20435586418361365, + "grad_norm": 1.4366608533867027, + "learning_rate": 3.689563048571046e-06, + "loss": 0.9671, + "step": 2266 + }, + { + "epoch": 0.2044460477070839, + "grad_norm": 1.5928845767335222, + "learning_rate": 3.6892503587455395e-06, + "loss": 0.9796, + "step": 2267 + }, + { + "epoch": 0.20453623123055417, + "grad_norm": 1.4347319281464046, + "learning_rate": 3.6889375247838766e-06, + "loss": 0.9941, + "step": 2268 + }, + { + "epoch": 0.20462641475402443, + "grad_norm": 1.525629432731168, + "learning_rate": 3.688624546712753e-06, + "loss": 0.9852, + "step": 2269 + }, + { + "epoch": 0.2047165982774947, + "grad_norm": 1.409617059394193, + "learning_rate": 3.688311424558871e-06, + "loss": 0.9367, + "step": 2270 + }, + { + "epoch": 0.20480678180096495, + "grad_norm": 2.2861745277323955, + "learning_rate": 3.6879981583489496e-06, + "loss": 0.8909, + "step": 2271 + }, + { + "epoch": 0.20489696532443522, + "grad_norm": 1.4785875431443682, + "learning_rate": 3.687684748109718e-06, + "loss": 1.002, + "step": 2272 + }, + { + "epoch": 0.20498714884790548, + "grad_norm": 1.6967453768346665, + "learning_rate": 3.6873711938679174e-06, + "loss": 1.0502, + "step": 2273 + }, + { + "epoch": 0.20507733237137574, + "grad_norm": 1.4087184857124275, + "learning_rate": 3.6870574956503027e-06, + "loss": 0.9754, + "step": 2274 + }, + { + "epoch": 0.20516751589484603, + "grad_norm": 1.7320930659198335, + "learning_rate": 3.68674365348364e-06, + "loss": 0.9707, + "step": 2275 + }, + { + "epoch": 0.2052576994183163, + "grad_norm": 1.6232905566449662, + "learning_rate": 3.6864296673947086e-06, + "loss": 0.9919, + "step": 2276 + }, + { + "epoch": 0.20534788294178655, + "grad_norm": 1.5300759466183513, + "learning_rate": 3.686115537410298e-06, + "loss": 0.9695, + "step": 2277 + }, + { + "epoch": 0.2054380664652568, + "grad_norm": 1.7742905770615378, + "learning_rate": 3.685801263557214e-06, + "loss": 0.93, + "step": 2278 + }, + { + "epoch": 0.20552824998872707, + "grad_norm": 1.7664277269018611, + "learning_rate": 3.68548684586227e-06, + "loss": 1.0752, + "step": 2279 + }, + { + "epoch": 0.20561843351219733, + "grad_norm": 1.7646247258411423, + "learning_rate": 3.685172284352295e-06, + "loss": 0.9805, + "step": 2280 + }, + { + "epoch": 0.2057086170356676, + "grad_norm": 1.2256386065469451, + "learning_rate": 3.684857579054128e-06, + "loss": 0.919, + "step": 2281 + }, + { + "epoch": 0.20579880055913785, + "grad_norm": 1.570523224161883, + "learning_rate": 3.6845427299946233e-06, + "loss": 1.0503, + "step": 2282 + }, + { + "epoch": 0.2058889840826081, + "grad_norm": 1.5408064358173774, + "learning_rate": 3.6842277372006434e-06, + "loss": 1.0838, + "step": 2283 + }, + { + "epoch": 0.20597916760607837, + "grad_norm": 1.5223603181701872, + "learning_rate": 3.6839126006990664e-06, + "loss": 0.9583, + "step": 2284 + }, + { + "epoch": 0.20606935112954863, + "grad_norm": 1.3555044999478856, + "learning_rate": 3.6835973205167818e-06, + "loss": 1.0385, + "step": 2285 + }, + { + "epoch": 0.2061595346530189, + "grad_norm": 1.7894547211598903, + "learning_rate": 3.6832818966806904e-06, + "loss": 1.0217, + "step": 2286 + }, + { + "epoch": 0.20624971817648916, + "grad_norm": 1.6466403222132409, + "learning_rate": 3.682966329217706e-06, + "loss": 1.0656, + "step": 2287 + }, + { + "epoch": 0.20633990169995942, + "grad_norm": 1.4584345464551864, + "learning_rate": 3.6826506181547543e-06, + "loss": 1.0232, + "step": 2288 + }, + { + "epoch": 0.20643008522342968, + "grad_norm": 1.6571612460426843, + "learning_rate": 3.682334763518774e-06, + "loss": 1.0103, + "step": 2289 + }, + { + "epoch": 0.20652026874689994, + "grad_norm": 1.4999586735436636, + "learning_rate": 3.6820187653367158e-06, + "loss": 0.9511, + "step": 2290 + }, + { + "epoch": 0.2066104522703702, + "grad_norm": 1.6622002438944028, + "learning_rate": 3.6817026236355412e-06, + "loss": 0.8808, + "step": 2291 + }, + { + "epoch": 0.20670063579384046, + "grad_norm": 2.0488052003230686, + "learning_rate": 3.681386338442227e-06, + "loss": 0.9664, + "step": 2292 + }, + { + "epoch": 0.20679081931731072, + "grad_norm": 1.424681440517478, + "learning_rate": 3.681069909783758e-06, + "loss": 0.9507, + "step": 2293 + }, + { + "epoch": 0.20688100284078098, + "grad_norm": 1.6405029251458558, + "learning_rate": 3.680753337687136e-06, + "loss": 0.998, + "step": 2294 + }, + { + "epoch": 0.20697118636425124, + "grad_norm": 1.7755358545308195, + "learning_rate": 3.680436622179371e-06, + "loss": 0.9147, + "step": 2295 + }, + { + "epoch": 0.2070613698877215, + "grad_norm": 1.6506053536577299, + "learning_rate": 3.680119763287488e-06, + "loss": 0.9654, + "step": 2296 + }, + { + "epoch": 0.20715155341119176, + "grad_norm": 1.6859289556221375, + "learning_rate": 3.6798027610385227e-06, + "loss": 0.9241, + "step": 2297 + }, + { + "epoch": 0.20724173693466202, + "grad_norm": 1.9194009254565427, + "learning_rate": 3.6794856154595235e-06, + "loss": 1.0296, + "step": 2298 + }, + { + "epoch": 0.2073319204581323, + "grad_norm": 1.9242987507317753, + "learning_rate": 3.6791683265775506e-06, + "loss": 0.9971, + "step": 2299 + }, + { + "epoch": 0.20742210398160257, + "grad_norm": 1.728228701468668, + "learning_rate": 3.6788508944196773e-06, + "loss": 0.9449, + "step": 2300 + }, + { + "epoch": 0.20751228750507283, + "grad_norm": 1.3491278975070478, + "learning_rate": 3.678533319012989e-06, + "loss": 1.0592, + "step": 2301 + }, + { + "epoch": 0.2076024710285431, + "grad_norm": 1.5508615775647818, + "learning_rate": 3.6782156003845826e-06, + "loss": 0.9343, + "step": 2302 + }, + { + "epoch": 0.20769265455201336, + "grad_norm": 0.7652909854280736, + "learning_rate": 3.6778977385615676e-06, + "loss": 0.7977, + "step": 2303 + }, + { + "epoch": 0.20778283807548362, + "grad_norm": 1.5157228418604116, + "learning_rate": 3.6775797335710656e-06, + "loss": 1.0314, + "step": 2304 + }, + { + "epoch": 0.20787302159895388, + "grad_norm": 1.4238135568634862, + "learning_rate": 3.6772615854402105e-06, + "loss": 0.9341, + "step": 2305 + }, + { + "epoch": 0.20796320512242414, + "grad_norm": 1.321472425911397, + "learning_rate": 3.6769432941961487e-06, + "loss": 0.8668, + "step": 2306 + }, + { + "epoch": 0.2080533886458944, + "grad_norm": 1.5019189957212247, + "learning_rate": 3.676624859866038e-06, + "loss": 0.9807, + "step": 2307 + }, + { + "epoch": 0.20814357216936466, + "grad_norm": 2.3416456821175644, + "learning_rate": 3.67630628247705e-06, + "loss": 0.9164, + "step": 2308 + }, + { + "epoch": 0.20823375569283492, + "grad_norm": 1.7971597860471624, + "learning_rate": 3.675987562056367e-06, + "loss": 1.0486, + "step": 2309 + }, + { + "epoch": 0.20832393921630518, + "grad_norm": 1.581714965851962, + "learning_rate": 3.675668698631184e-06, + "loss": 1.0199, + "step": 2310 + }, + { + "epoch": 0.20841412273977544, + "grad_norm": 1.4303624795453898, + "learning_rate": 3.675349692228708e-06, + "loss": 1.0487, + "step": 2311 + }, + { + "epoch": 0.2085043062632457, + "grad_norm": 1.6095647607426855, + "learning_rate": 3.6750305428761578e-06, + "loss": 0.9587, + "step": 2312 + }, + { + "epoch": 0.20859448978671596, + "grad_norm": 1.9383723848344565, + "learning_rate": 3.674711250600766e-06, + "loss": 0.9924, + "step": 2313 + }, + { + "epoch": 0.20868467331018623, + "grad_norm": 0.8118347599079919, + "learning_rate": 3.6743918154297765e-06, + "loss": 0.778, + "step": 2314 + }, + { + "epoch": 0.20877485683365649, + "grad_norm": 1.7118720023890444, + "learning_rate": 3.6740722373904446e-06, + "loss": 0.9764, + "step": 2315 + }, + { + "epoch": 0.20886504035712675, + "grad_norm": 1.6210925734182478, + "learning_rate": 3.6737525165100383e-06, + "loss": 0.9617, + "step": 2316 + }, + { + "epoch": 0.208955223880597, + "grad_norm": 1.6384811199405416, + "learning_rate": 3.6734326528158385e-06, + "loss": 1.0085, + "step": 2317 + }, + { + "epoch": 0.20904540740406727, + "grad_norm": 1.4064351277882523, + "learning_rate": 3.673112646335138e-06, + "loss": 0.9718, + "step": 2318 + }, + { + "epoch": 0.20913559092753753, + "grad_norm": 1.5191786486888499, + "learning_rate": 3.672792497095241e-06, + "loss": 0.9322, + "step": 2319 + }, + { + "epoch": 0.2092257744510078, + "grad_norm": 1.7557445654075452, + "learning_rate": 3.672472205123464e-06, + "loss": 0.9863, + "step": 2320 + }, + { + "epoch": 0.20931595797447805, + "grad_norm": 1.5164584198536066, + "learning_rate": 3.6721517704471363e-06, + "loss": 0.9106, + "step": 2321 + }, + { + "epoch": 0.2094061414979483, + "grad_norm": 1.5123058825052993, + "learning_rate": 3.6718311930936e-06, + "loss": 1.0178, + "step": 2322 + }, + { + "epoch": 0.2094963250214186, + "grad_norm": 1.6596954783364768, + "learning_rate": 3.6715104730902074e-06, + "loss": 0.9379, + "step": 2323 + }, + { + "epoch": 0.20958650854488886, + "grad_norm": 1.8627671146223517, + "learning_rate": 3.671189610464325e-06, + "loss": 1.0186, + "step": 2324 + }, + { + "epoch": 0.20967669206835912, + "grad_norm": 1.6219591152010173, + "learning_rate": 3.6708686052433303e-06, + "loss": 0.9313, + "step": 2325 + }, + { + "epoch": 0.20976687559182938, + "grad_norm": 1.9904234611976077, + "learning_rate": 3.6705474574546127e-06, + "loss": 0.9981, + "step": 2326 + }, + { + "epoch": 0.20985705911529964, + "grad_norm": 1.4896638310518475, + "learning_rate": 3.670226167125575e-06, + "loss": 1.0042, + "step": 2327 + }, + { + "epoch": 0.2099472426387699, + "grad_norm": 1.7668199799549453, + "learning_rate": 3.6699047342836313e-06, + "loss": 0.9856, + "step": 2328 + }, + { + "epoch": 0.21003742616224017, + "grad_norm": 1.440502431293463, + "learning_rate": 3.669583158956208e-06, + "loss": 1.0187, + "step": 2329 + }, + { + "epoch": 0.21012760968571043, + "grad_norm": 0.9121396633567677, + "learning_rate": 3.669261441170743e-06, + "loss": 0.8097, + "step": 2330 + }, + { + "epoch": 0.2102177932091807, + "grad_norm": 1.8186746850428939, + "learning_rate": 3.668939580954688e-06, + "loss": 0.9944, + "step": 2331 + }, + { + "epoch": 0.21030797673265095, + "grad_norm": 1.597694303250595, + "learning_rate": 3.668617578335506e-06, + "loss": 0.956, + "step": 2332 + }, + { + "epoch": 0.2103981602561212, + "grad_norm": 1.7606735343052886, + "learning_rate": 3.6682954333406707e-06, + "loss": 1.0132, + "step": 2333 + }, + { + "epoch": 0.21048834377959147, + "grad_norm": 1.9988175711503182, + "learning_rate": 3.6679731459976707e-06, + "loss": 1.1199, + "step": 2334 + }, + { + "epoch": 0.21057852730306173, + "grad_norm": 1.4682467795735665, + "learning_rate": 3.6676507163340046e-06, + "loss": 0.9833, + "step": 2335 + }, + { + "epoch": 0.210668710826532, + "grad_norm": 3.1530254246212643, + "learning_rate": 3.6673281443771842e-06, + "loss": 1.0184, + "step": 2336 + }, + { + "epoch": 0.21075889435000225, + "grad_norm": 1.5363784079587999, + "learning_rate": 3.667005430154733e-06, + "loss": 0.8714, + "step": 2337 + }, + { + "epoch": 0.2108490778734725, + "grad_norm": 0.7810344780236426, + "learning_rate": 3.666682573694186e-06, + "loss": 0.8838, + "step": 2338 + }, + { + "epoch": 0.21093926139694277, + "grad_norm": 0.7023975211703132, + "learning_rate": 3.6663595750230924e-06, + "loss": 0.7523, + "step": 2339 + }, + { + "epoch": 0.21102944492041303, + "grad_norm": 1.83595048615753, + "learning_rate": 3.666036434169012e-06, + "loss": 1.0369, + "step": 2340 + }, + { + "epoch": 0.2111196284438833, + "grad_norm": 1.5440795762847783, + "learning_rate": 3.665713151159516e-06, + "loss": 0.9678, + "step": 2341 + }, + { + "epoch": 0.21120981196735356, + "grad_norm": 2.0807652730916337, + "learning_rate": 3.665389726022189e-06, + "loss": 1.0892, + "step": 2342 + }, + { + "epoch": 0.21129999549082382, + "grad_norm": 1.6716322678094422, + "learning_rate": 3.6650661587846283e-06, + "loss": 0.9559, + "step": 2343 + }, + { + "epoch": 0.21139017901429408, + "grad_norm": 1.7757500178774661, + "learning_rate": 3.6647424494744418e-06, + "loss": 0.9473, + "step": 2344 + }, + { + "epoch": 0.21148036253776434, + "grad_norm": 1.4310439715290746, + "learning_rate": 3.6644185981192503e-06, + "loss": 1.0352, + "step": 2345 + }, + { + "epoch": 0.2115705460612346, + "grad_norm": 2.0878627621796144, + "learning_rate": 3.6640946047466868e-06, + "loss": 0.9154, + "step": 2346 + }, + { + "epoch": 0.2116607295847049, + "grad_norm": 1.5832230211612979, + "learning_rate": 3.6637704693843953e-06, + "loss": 0.9493, + "step": 2347 + }, + { + "epoch": 0.21175091310817515, + "grad_norm": 1.4829972147076225, + "learning_rate": 3.6634461920600337e-06, + "loss": 0.9175, + "step": 2348 + }, + { + "epoch": 0.2118410966316454, + "grad_norm": 1.7632688482183914, + "learning_rate": 3.66312177280127e-06, + "loss": 1.0044, + "step": 2349 + }, + { + "epoch": 0.21193128015511567, + "grad_norm": 1.5411099210445909, + "learning_rate": 3.6627972116357872e-06, + "loss": 1.0226, + "step": 2350 + }, + { + "epoch": 0.21202146367858593, + "grad_norm": 1.6238591517644894, + "learning_rate": 3.662472508591278e-06, + "loss": 1.0472, + "step": 2351 + }, + { + "epoch": 0.2121116472020562, + "grad_norm": 1.7058753807552145, + "learning_rate": 3.662147663695447e-06, + "loss": 0.9722, + "step": 2352 + }, + { + "epoch": 0.21220183072552645, + "grad_norm": 1.6657209415251404, + "learning_rate": 3.6618226769760127e-06, + "loss": 0.9742, + "step": 2353 + }, + { + "epoch": 0.2122920142489967, + "grad_norm": 1.7403576054809669, + "learning_rate": 3.661497548460704e-06, + "loss": 0.869, + "step": 2354 + }, + { + "epoch": 0.21238219777246697, + "grad_norm": 1.838211274831587, + "learning_rate": 3.6611722781772635e-06, + "loss": 1.002, + "step": 2355 + }, + { + "epoch": 0.21247238129593723, + "grad_norm": 0.9970480143924089, + "learning_rate": 3.6608468661534444e-06, + "loss": 0.8475, + "step": 2356 + }, + { + "epoch": 0.2125625648194075, + "grad_norm": 2.4238297971954847, + "learning_rate": 3.660521312417013e-06, + "loss": 0.9137, + "step": 2357 + }, + { + "epoch": 0.21265274834287776, + "grad_norm": 2.2422295011514186, + "learning_rate": 3.660195616995747e-06, + "loss": 0.9501, + "step": 2358 + }, + { + "epoch": 0.21274293186634802, + "grad_norm": 1.6471055421394256, + "learning_rate": 3.6598697799174367e-06, + "loss": 1.0142, + "step": 2359 + }, + { + "epoch": 0.21283311538981828, + "grad_norm": 1.9080860488569507, + "learning_rate": 3.6595438012098844e-06, + "loss": 1.0475, + "step": 2360 + }, + { + "epoch": 0.21292329891328854, + "grad_norm": 1.6993917442919473, + "learning_rate": 3.6592176809009045e-06, + "loss": 0.9551, + "step": 2361 + }, + { + "epoch": 0.2130134824367588, + "grad_norm": 1.4686703761817699, + "learning_rate": 3.6588914190183227e-06, + "loss": 1.0198, + "step": 2362 + }, + { + "epoch": 0.21310366596022906, + "grad_norm": 1.954183673039055, + "learning_rate": 3.658565015589978e-06, + "loss": 0.8583, + "step": 2363 + }, + { + "epoch": 0.21319384948369932, + "grad_norm": 1.8727353725022504, + "learning_rate": 3.6582384706437217e-06, + "loss": 0.9923, + "step": 2364 + }, + { + "epoch": 0.21328403300716958, + "grad_norm": 1.6765354833354809, + "learning_rate": 3.6579117842074156e-06, + "loss": 0.9978, + "step": 2365 + }, + { + "epoch": 0.21337421653063984, + "grad_norm": 1.565769665519816, + "learning_rate": 3.657584956308934e-06, + "loss": 0.9228, + "step": 2366 + }, + { + "epoch": 0.2134644000541101, + "grad_norm": 1.9720040329007893, + "learning_rate": 3.6572579869761648e-06, + "loss": 1.0503, + "step": 2367 + }, + { + "epoch": 0.21355458357758036, + "grad_norm": 7.218171596061259, + "learning_rate": 3.6569308762370056e-06, + "loss": 1.039, + "step": 2368 + }, + { + "epoch": 0.21364476710105063, + "grad_norm": 1.8385343326353882, + "learning_rate": 3.6566036241193676e-06, + "loss": 1.0277, + "step": 2369 + }, + { + "epoch": 0.2137349506245209, + "grad_norm": 0.7643051838591336, + "learning_rate": 3.656276230651174e-06, + "loss": 0.8106, + "step": 2370 + }, + { + "epoch": 0.21382513414799117, + "grad_norm": 0.7372468837285403, + "learning_rate": 3.65594869586036e-06, + "loss": 0.844, + "step": 2371 + }, + { + "epoch": 0.21391531767146144, + "grad_norm": 1.4018109938031633, + "learning_rate": 3.6556210197748724e-06, + "loss": 0.9067, + "step": 2372 + }, + { + "epoch": 0.2140055011949317, + "grad_norm": 1.5708199814988741, + "learning_rate": 3.655293202422671e-06, + "loss": 0.956, + "step": 2373 + }, + { + "epoch": 0.21409568471840196, + "grad_norm": 1.5148635209806447, + "learning_rate": 3.654965243831725e-06, + "loss": 0.9956, + "step": 2374 + }, + { + "epoch": 0.21418586824187222, + "grad_norm": 1.41056166973611, + "learning_rate": 3.65463714403002e-06, + "loss": 1.027, + "step": 2375 + }, + { + "epoch": 0.21427605176534248, + "grad_norm": 1.5252674556146626, + "learning_rate": 3.65430890304555e-06, + "loss": 0.9393, + "step": 2376 + }, + { + "epoch": 0.21436623528881274, + "grad_norm": 1.3625159096445163, + "learning_rate": 3.653980520906323e-06, + "loss": 1.0116, + "step": 2377 + }, + { + "epoch": 0.214456418812283, + "grad_norm": 1.3875223329825908, + "learning_rate": 3.653651997640358e-06, + "loss": 0.9573, + "step": 2378 + }, + { + "epoch": 0.21454660233575326, + "grad_norm": 1.560997278483607, + "learning_rate": 3.653323333275686e-06, + "loss": 0.9855, + "step": 2379 + }, + { + "epoch": 0.21463678585922352, + "grad_norm": 1.4647857247622136, + "learning_rate": 3.652994527840351e-06, + "loss": 0.9699, + "step": 2380 + }, + { + "epoch": 0.21472696938269378, + "grad_norm": 1.584019896912118, + "learning_rate": 3.6526655813624087e-06, + "loss": 0.9466, + "step": 2381 + }, + { + "epoch": 0.21481715290616404, + "grad_norm": 0.7130151509257545, + "learning_rate": 3.652336493869925e-06, + "loss": 0.804, + "step": 2382 + }, + { + "epoch": 0.2149073364296343, + "grad_norm": 1.5811288025739647, + "learning_rate": 3.6520072653909823e-06, + "loss": 0.9906, + "step": 2383 + }, + { + "epoch": 0.21499751995310457, + "grad_norm": 1.790001368815149, + "learning_rate": 3.6516778959536702e-06, + "loss": 0.9792, + "step": 2384 + }, + { + "epoch": 0.21508770347657483, + "grad_norm": 1.8975101922003987, + "learning_rate": 3.6513483855860923e-06, + "loss": 1.0318, + "step": 2385 + }, + { + "epoch": 0.2151778870000451, + "grad_norm": 1.5227629904524054, + "learning_rate": 3.6510187343163654e-06, + "loss": 0.9266, + "step": 2386 + }, + { + "epoch": 0.21526807052351535, + "grad_norm": 1.6383946835636518, + "learning_rate": 3.650688942172616e-06, + "loss": 1.0158, + "step": 2387 + }, + { + "epoch": 0.2153582540469856, + "grad_norm": 1.7298649752850352, + "learning_rate": 3.650359009182984e-06, + "loss": 0.9253, + "step": 2388 + }, + { + "epoch": 0.21544843757045587, + "grad_norm": 1.798515566035494, + "learning_rate": 3.650028935375622e-06, + "loss": 1.0689, + "step": 2389 + }, + { + "epoch": 0.21553862109392613, + "grad_norm": 1.4181149409993608, + "learning_rate": 3.6496987207786926e-06, + "loss": 0.9469, + "step": 2390 + }, + { + "epoch": 0.2156288046173964, + "grad_norm": 1.6432626502186922, + "learning_rate": 3.6493683654203724e-06, + "loss": 0.9289, + "step": 2391 + }, + { + "epoch": 0.21571898814086665, + "grad_norm": 1.6312934343540633, + "learning_rate": 3.6490378693288484e-06, + "loss": 0.9689, + "step": 2392 + }, + { + "epoch": 0.2158091716643369, + "grad_norm": 1.7045333359027437, + "learning_rate": 3.648707232532321e-06, + "loss": 1.0259, + "step": 2393 + }, + { + "epoch": 0.2158993551878072, + "grad_norm": 1.8264270950917538, + "learning_rate": 3.6483764550590017e-06, + "loss": 0.9801, + "step": 2394 + }, + { + "epoch": 0.21598953871127746, + "grad_norm": 1.502844418293628, + "learning_rate": 3.6480455369371133e-06, + "loss": 0.9982, + "step": 2395 + }, + { + "epoch": 0.21607972223474772, + "grad_norm": 1.4425045470855289, + "learning_rate": 3.647714478194893e-06, + "loss": 0.9871, + "step": 2396 + }, + { + "epoch": 0.21616990575821798, + "grad_norm": 1.8083813009038272, + "learning_rate": 3.647383278860588e-06, + "loss": 0.9261, + "step": 2397 + }, + { + "epoch": 0.21626008928168824, + "grad_norm": 1.7769789772056697, + "learning_rate": 3.6470519389624587e-06, + "loss": 0.9741, + "step": 2398 + }, + { + "epoch": 0.2163502728051585, + "grad_norm": 1.5261925458911758, + "learning_rate": 3.646720458528776e-06, + "loss": 1.0414, + "step": 2399 + }, + { + "epoch": 0.21644045632862877, + "grad_norm": 2.4622586049690516, + "learning_rate": 3.6463888375878235e-06, + "loss": 0.9584, + "step": 2400 + }, + { + "epoch": 0.21653063985209903, + "grad_norm": 1.9016884805139762, + "learning_rate": 3.646057076167897e-06, + "loss": 0.9974, + "step": 2401 + }, + { + "epoch": 0.2166208233755693, + "grad_norm": 2.4678747822791824, + "learning_rate": 3.645725174297305e-06, + "loss": 0.9879, + "step": 2402 + }, + { + "epoch": 0.21671100689903955, + "grad_norm": 3.1409937703757946, + "learning_rate": 3.645393132004367e-06, + "loss": 0.9622, + "step": 2403 + }, + { + "epoch": 0.2168011904225098, + "grad_norm": 1.5777318483631353, + "learning_rate": 3.6450609493174135e-06, + "loss": 0.9408, + "step": 2404 + }, + { + "epoch": 0.21689137394598007, + "grad_norm": 1.6030530738682827, + "learning_rate": 3.6447286262647896e-06, + "loss": 0.9476, + "step": 2405 + }, + { + "epoch": 0.21698155746945033, + "grad_norm": 1.655638599908715, + "learning_rate": 3.64439616287485e-06, + "loss": 0.9506, + "step": 2406 + }, + { + "epoch": 0.2170717409929206, + "grad_norm": 1.4362767031638255, + "learning_rate": 3.644063559175963e-06, + "loss": 0.9474, + "step": 2407 + }, + { + "epoch": 0.21716192451639085, + "grad_norm": 1.6546018064965127, + "learning_rate": 3.6437308151965074e-06, + "loss": 0.9647, + "step": 2408 + }, + { + "epoch": 0.2172521080398611, + "grad_norm": 1.472734606047075, + "learning_rate": 3.643397930964876e-06, + "loss": 0.9959, + "step": 2409 + }, + { + "epoch": 0.21734229156333137, + "grad_norm": 1.8198390885368145, + "learning_rate": 3.6430649065094707e-06, + "loss": 1.0472, + "step": 2410 + }, + { + "epoch": 0.21743247508680164, + "grad_norm": 1.930073255522375, + "learning_rate": 3.6427317418587086e-06, + "loss": 1.0504, + "step": 2411 + }, + { + "epoch": 0.2175226586102719, + "grad_norm": 2.2684088079979405, + "learning_rate": 3.6423984370410157e-06, + "loss": 0.9252, + "step": 2412 + }, + { + "epoch": 0.21761284213374216, + "grad_norm": 1.562781117899749, + "learning_rate": 3.6420649920848324e-06, + "loss": 0.948, + "step": 2413 + }, + { + "epoch": 0.21770302565721242, + "grad_norm": 1.4534821430338964, + "learning_rate": 3.6417314070186096e-06, + "loss": 0.9875, + "step": 2414 + }, + { + "epoch": 0.21779320918068268, + "grad_norm": 1.780243974317358, + "learning_rate": 3.641397681870811e-06, + "loss": 0.9909, + "step": 2415 + }, + { + "epoch": 0.21788339270415294, + "grad_norm": 1.6906020298664828, + "learning_rate": 3.641063816669911e-06, + "loss": 1.0008, + "step": 2416 + }, + { + "epoch": 0.2179735762276232, + "grad_norm": 1.567037250774707, + "learning_rate": 3.640729811444398e-06, + "loss": 1.0438, + "step": 2417 + }, + { + "epoch": 0.2180637597510935, + "grad_norm": 1.4867268452346276, + "learning_rate": 3.6403956662227706e-06, + "loss": 0.9858, + "step": 2418 + }, + { + "epoch": 0.21815394327456375, + "grad_norm": 1.6687135842453962, + "learning_rate": 3.6400613810335396e-06, + "loss": 1.0346, + "step": 2419 + }, + { + "epoch": 0.218244126798034, + "grad_norm": 1.6989798542330399, + "learning_rate": 3.639726955905228e-06, + "loss": 0.8965, + "step": 2420 + }, + { + "epoch": 0.21833431032150427, + "grad_norm": 1.8668321087727655, + "learning_rate": 3.639392390866372e-06, + "loss": 1.0076, + "step": 2421 + }, + { + "epoch": 0.21842449384497453, + "grad_norm": 1.6914317633486422, + "learning_rate": 3.639057685945517e-06, + "loss": 0.9634, + "step": 2422 + }, + { + "epoch": 0.2185146773684448, + "grad_norm": 1.548466181946535, + "learning_rate": 3.638722841171223e-06, + "loss": 0.8738, + "step": 2423 + }, + { + "epoch": 0.21860486089191505, + "grad_norm": 1.528975763626568, + "learning_rate": 3.638387856572061e-06, + "loss": 1.0215, + "step": 2424 + }, + { + "epoch": 0.21869504441538531, + "grad_norm": 0.7576895545252568, + "learning_rate": 3.638052732176612e-06, + "loss": 0.8101, + "step": 2425 + }, + { + "epoch": 0.21878522793885558, + "grad_norm": 1.6765880287694084, + "learning_rate": 3.637717468013472e-06, + "loss": 0.9967, + "step": 2426 + }, + { + "epoch": 0.21887541146232584, + "grad_norm": 0.6887940236264876, + "learning_rate": 3.6373820641112475e-06, + "loss": 0.7621, + "step": 2427 + }, + { + "epoch": 0.2189655949857961, + "grad_norm": 1.7379510115774808, + "learning_rate": 3.6370465204985567e-06, + "loss": 1.001, + "step": 2428 + }, + { + "epoch": 0.21905577850926636, + "grad_norm": 0.6850823243993764, + "learning_rate": 3.6367108372040304e-06, + "loss": 0.8228, + "step": 2429 + }, + { + "epoch": 0.21914596203273662, + "grad_norm": 1.8684829623185901, + "learning_rate": 3.6363750142563107e-06, + "loss": 0.9851, + "step": 2430 + }, + { + "epoch": 0.21923614555620688, + "grad_norm": 0.7695847313590605, + "learning_rate": 3.636039051684052e-06, + "loss": 0.8586, + "step": 2431 + }, + { + "epoch": 0.21932632907967714, + "grad_norm": 1.7045537572716307, + "learning_rate": 3.6357029495159203e-06, + "loss": 0.9498, + "step": 2432 + }, + { + "epoch": 0.2194165126031474, + "grad_norm": 1.6509254085712466, + "learning_rate": 3.6353667077805934e-06, + "loss": 0.9595, + "step": 2433 + }, + { + "epoch": 0.21950669612661766, + "grad_norm": 1.707142213056298, + "learning_rate": 3.6350303265067625e-06, + "loss": 1.0206, + "step": 2434 + }, + { + "epoch": 0.21959687965008792, + "grad_norm": 1.5094905864299486, + "learning_rate": 3.6346938057231285e-06, + "loss": 1.012, + "step": 2435 + }, + { + "epoch": 0.21968706317355818, + "grad_norm": 1.5896507352218774, + "learning_rate": 3.6343571454584047e-06, + "loss": 0.9308, + "step": 2436 + }, + { + "epoch": 0.21977724669702844, + "grad_norm": 1.6402849844138012, + "learning_rate": 3.6340203457413176e-06, + "loss": 1.0223, + "step": 2437 + }, + { + "epoch": 0.2198674302204987, + "grad_norm": 1.5477494359546065, + "learning_rate": 3.633683406600605e-06, + "loss": 1.0062, + "step": 2438 + }, + { + "epoch": 0.21995761374396897, + "grad_norm": 1.668098462530804, + "learning_rate": 3.6333463280650165e-06, + "loss": 1.0451, + "step": 2439 + }, + { + "epoch": 0.22004779726743923, + "grad_norm": 0.8675930003024973, + "learning_rate": 3.6330091101633126e-06, + "loss": 0.7713, + "step": 2440 + }, + { + "epoch": 0.2201379807909095, + "grad_norm": 1.530770674440509, + "learning_rate": 3.632671752924267e-06, + "loss": 1.0205, + "step": 2441 + }, + { + "epoch": 0.22022816431437978, + "grad_norm": 1.916222797357486, + "learning_rate": 3.632334256376665e-06, + "loss": 0.9897, + "step": 2442 + }, + { + "epoch": 0.22031834783785004, + "grad_norm": 1.6101262829697507, + "learning_rate": 3.6319966205493044e-06, + "loss": 1.0461, + "step": 2443 + }, + { + "epoch": 0.2204085313613203, + "grad_norm": 0.930934486448309, + "learning_rate": 3.6316588454709922e-06, + "loss": 0.8551, + "step": 2444 + }, + { + "epoch": 0.22049871488479056, + "grad_norm": 1.4526848895637, + "learning_rate": 3.6313209311705514e-06, + "loss": 0.9344, + "step": 2445 + }, + { + "epoch": 0.22058889840826082, + "grad_norm": 1.7988657662660505, + "learning_rate": 3.6309828776768133e-06, + "loss": 1.0288, + "step": 2446 + }, + { + "epoch": 0.22067908193173108, + "grad_norm": 1.7400312551344026, + "learning_rate": 3.630644685018623e-06, + "loss": 0.9428, + "step": 2447 + }, + { + "epoch": 0.22076926545520134, + "grad_norm": 2.4413948241920136, + "learning_rate": 3.6303063532248367e-06, + "loss": 1.0693, + "step": 2448 + }, + { + "epoch": 0.2208594489786716, + "grad_norm": 1.695455518745963, + "learning_rate": 3.6299678823243236e-06, + "loss": 1.0669, + "step": 2449 + }, + { + "epoch": 0.22094963250214186, + "grad_norm": 1.651489614215547, + "learning_rate": 3.629629272345963e-06, + "loss": 0.8871, + "step": 2450 + }, + { + "epoch": 0.22103981602561212, + "grad_norm": 1.6534290401896246, + "learning_rate": 3.6292905233186468e-06, + "loss": 1.1089, + "step": 2451 + }, + { + "epoch": 0.22112999954908238, + "grad_norm": 1.5639819937154837, + "learning_rate": 3.6289516352712796e-06, + "loss": 0.9497, + "step": 2452 + }, + { + "epoch": 0.22122018307255265, + "grad_norm": 0.8756846746440371, + "learning_rate": 3.6286126082327764e-06, + "loss": 0.8054, + "step": 2453 + }, + { + "epoch": 0.2213103665960229, + "grad_norm": 1.577329076100718, + "learning_rate": 3.628273442232066e-06, + "loss": 1.0196, + "step": 2454 + }, + { + "epoch": 0.22140055011949317, + "grad_norm": 1.7195728759718787, + "learning_rate": 3.627934137298087e-06, + "loss": 0.9796, + "step": 2455 + }, + { + "epoch": 0.22149073364296343, + "grad_norm": 2.1721744468142004, + "learning_rate": 3.627594693459792e-06, + "loss": 0.964, + "step": 2456 + }, + { + "epoch": 0.2215809171664337, + "grad_norm": 1.990810621581007, + "learning_rate": 3.6272551107461424e-06, + "loss": 0.9455, + "step": 2457 + }, + { + "epoch": 0.22167110068990395, + "grad_norm": 2.1527748299308476, + "learning_rate": 3.6269153891861137e-06, + "loss": 0.9809, + "step": 2458 + }, + { + "epoch": 0.2217612842133742, + "grad_norm": 1.3337935259508757, + "learning_rate": 3.6265755288086944e-06, + "loss": 0.9516, + "step": 2459 + }, + { + "epoch": 0.22185146773684447, + "grad_norm": 6.412372919500256, + "learning_rate": 3.626235529642881e-06, + "loss": 0.9642, + "step": 2460 + }, + { + "epoch": 0.22194165126031473, + "grad_norm": 1.6369582117114805, + "learning_rate": 3.625895391717686e-06, + "loss": 0.9227, + "step": 2461 + }, + { + "epoch": 0.222031834783785, + "grad_norm": 1.4382939012293268, + "learning_rate": 3.625555115062131e-06, + "loss": 0.9536, + "step": 2462 + }, + { + "epoch": 0.22212201830725525, + "grad_norm": 1.603905578107735, + "learning_rate": 3.6252146997052507e-06, + "loss": 1.0516, + "step": 2463 + }, + { + "epoch": 0.22221220183072551, + "grad_norm": 0.8278324762153937, + "learning_rate": 3.6248741456760898e-06, + "loss": 0.763, + "step": 2464 + }, + { + "epoch": 0.22230238535419578, + "grad_norm": 1.6810202345847156, + "learning_rate": 3.624533453003708e-06, + "loss": 0.973, + "step": 2465 + }, + { + "epoch": 0.22239256887766606, + "grad_norm": 1.642860415556118, + "learning_rate": 3.6241926217171745e-06, + "loss": 0.9938, + "step": 2466 + }, + { + "epoch": 0.22248275240113632, + "grad_norm": 1.9872455410040497, + "learning_rate": 3.6238516518455703e-06, + "loss": 0.9012, + "step": 2467 + }, + { + "epoch": 0.22257293592460659, + "grad_norm": 2.0212204028282224, + "learning_rate": 3.62351054341799e-06, + "loss": 1.1043, + "step": 2468 + }, + { + "epoch": 0.22266311944807685, + "grad_norm": 1.762607324998413, + "learning_rate": 3.623169296463538e-06, + "loss": 1.014, + "step": 2469 + }, + { + "epoch": 0.2227533029715471, + "grad_norm": 1.6501402535217196, + "learning_rate": 3.6228279110113316e-06, + "loss": 0.986, + "step": 2470 + }, + { + "epoch": 0.22284348649501737, + "grad_norm": 1.8076630869375148, + "learning_rate": 3.6224863870904994e-06, + "loss": 0.9933, + "step": 2471 + }, + { + "epoch": 0.22293367001848763, + "grad_norm": 1.715014473100544, + "learning_rate": 3.6221447247301827e-06, + "loss": 0.9973, + "step": 2472 + }, + { + "epoch": 0.2230238535419579, + "grad_norm": 1.3874299710446352, + "learning_rate": 3.6218029239595332e-06, + "loss": 0.9347, + "step": 2473 + }, + { + "epoch": 0.22311403706542815, + "grad_norm": 1.7915715111000665, + "learning_rate": 3.621460984807716e-06, + "loss": 1.0139, + "step": 2474 + }, + { + "epoch": 0.2232042205888984, + "grad_norm": 1.5745962745691235, + "learning_rate": 3.621118907303907e-06, + "loss": 0.9611, + "step": 2475 + }, + { + "epoch": 0.22329440411236867, + "grad_norm": 1.4637297196127559, + "learning_rate": 3.620776691477294e-06, + "loss": 0.9795, + "step": 2476 + }, + { + "epoch": 0.22338458763583893, + "grad_norm": 0.7870013823151718, + "learning_rate": 3.6204343373570765e-06, + "loss": 0.7529, + "step": 2477 + }, + { + "epoch": 0.2234747711593092, + "grad_norm": 1.5722005700341708, + "learning_rate": 3.620091844972467e-06, + "loss": 1.0063, + "step": 2478 + }, + { + "epoch": 0.22356495468277945, + "grad_norm": 1.6581102396872067, + "learning_rate": 3.619749214352688e-06, + "loss": 1.0106, + "step": 2479 + }, + { + "epoch": 0.22365513820624972, + "grad_norm": 1.6743458822864716, + "learning_rate": 3.6194064455269744e-06, + "loss": 1.0865, + "step": 2480 + }, + { + "epoch": 0.22374532172971998, + "grad_norm": 1.6801815991798428, + "learning_rate": 3.6190635385245737e-06, + "loss": 0.9057, + "step": 2481 + }, + { + "epoch": 0.22383550525319024, + "grad_norm": 1.6084978250115902, + "learning_rate": 3.618720493374745e-06, + "loss": 0.916, + "step": 2482 + }, + { + "epoch": 0.2239256887766605, + "grad_norm": 1.625095144567571, + "learning_rate": 3.6183773101067575e-06, + "loss": 0.8463, + "step": 2483 + }, + { + "epoch": 0.22401587230013076, + "grad_norm": 1.6795532882654152, + "learning_rate": 3.6180339887498948e-06, + "loss": 0.9085, + "step": 2484 + }, + { + "epoch": 0.22410605582360102, + "grad_norm": 1.7611061265930386, + "learning_rate": 3.61769052933345e-06, + "loss": 0.9916, + "step": 2485 + }, + { + "epoch": 0.22419623934707128, + "grad_norm": 1.8629175626951262, + "learning_rate": 3.6173469318867297e-06, + "loss": 1.1382, + "step": 2486 + }, + { + "epoch": 0.22428642287054154, + "grad_norm": 2.2991383099394436, + "learning_rate": 3.617003196439051e-06, + "loss": 1.0049, + "step": 2487 + }, + { + "epoch": 0.2243766063940118, + "grad_norm": 1.3719124808437708, + "learning_rate": 3.616659323019744e-06, + "loss": 0.9974, + "step": 2488 + }, + { + "epoch": 0.22446678991748206, + "grad_norm": 1.4308755243146636, + "learning_rate": 3.616315311658149e-06, + "loss": 1.0158, + "step": 2489 + }, + { + "epoch": 0.22455697344095235, + "grad_norm": 1.811689590278657, + "learning_rate": 3.6159711623836195e-06, + "loss": 1.0469, + "step": 2490 + }, + { + "epoch": 0.2246471569644226, + "grad_norm": 0.7797247302506488, + "learning_rate": 3.6156268752255203e-06, + "loss": 0.7667, + "step": 2491 + }, + { + "epoch": 0.22473734048789287, + "grad_norm": 1.537591066222183, + "learning_rate": 3.615282450213227e-06, + "loss": 0.9517, + "step": 2492 + }, + { + "epoch": 0.22482752401136313, + "grad_norm": 1.8026214927965645, + "learning_rate": 3.614937887376128e-06, + "loss": 0.9997, + "step": 2493 + }, + { + "epoch": 0.2249177075348334, + "grad_norm": 1.7483037493481335, + "learning_rate": 3.614593186743625e-06, + "loss": 1.001, + "step": 2494 + }, + { + "epoch": 0.22500789105830366, + "grad_norm": 1.839682472757372, + "learning_rate": 3.614248348345128e-06, + "loss": 0.9271, + "step": 2495 + }, + { + "epoch": 0.22509807458177392, + "grad_norm": 1.63793543202317, + "learning_rate": 3.6139033722100614e-06, + "loss": 1.0378, + "step": 2496 + }, + { + "epoch": 0.22518825810524418, + "grad_norm": 1.8556350151657726, + "learning_rate": 3.6135582583678596e-06, + "loss": 1.0577, + "step": 2497 + }, + { + "epoch": 0.22527844162871444, + "grad_norm": 3.1476637881130265, + "learning_rate": 3.61321300684797e-06, + "loss": 1.0068, + "step": 2498 + }, + { + "epoch": 0.2253686251521847, + "grad_norm": 1.2310138766438605, + "learning_rate": 3.6128676176798527e-06, + "loss": 1.0103, + "step": 2499 + }, + { + "epoch": 0.22545880867565496, + "grad_norm": 2.0070071256441704, + "learning_rate": 3.612522090892976e-06, + "loss": 1.0033, + "step": 2500 + }, + { + "epoch": 0.22554899219912522, + "grad_norm": 1.4826200852801101, + "learning_rate": 3.6121764265168232e-06, + "loss": 0.9526, + "step": 2501 + }, + { + "epoch": 0.22563917572259548, + "grad_norm": 1.5134819846869012, + "learning_rate": 3.611830624580888e-06, + "loss": 1.0121, + "step": 2502 + }, + { + "epoch": 0.22572935924606574, + "grad_norm": 1.8843778062992487, + "learning_rate": 3.6114846851146767e-06, + "loss": 0.9985, + "step": 2503 + }, + { + "epoch": 0.225819542769536, + "grad_norm": 1.543169380374847, + "learning_rate": 3.6111386081477068e-06, + "loss": 0.8909, + "step": 2504 + }, + { + "epoch": 0.22590972629300626, + "grad_norm": 1.7819991293240356, + "learning_rate": 3.6107923937095066e-06, + "loss": 1.0067, + "step": 2505 + }, + { + "epoch": 0.22599990981647652, + "grad_norm": 1.9804756721210082, + "learning_rate": 3.6104460418296173e-06, + "loss": 0.9158, + "step": 2506 + }, + { + "epoch": 0.22609009333994678, + "grad_norm": 2.1524670352693307, + "learning_rate": 3.6100995525375924e-06, + "loss": 1.0375, + "step": 2507 + }, + { + "epoch": 0.22618027686341705, + "grad_norm": 1.4667359198490502, + "learning_rate": 3.6097529258629952e-06, + "loss": 0.93, + "step": 2508 + }, + { + "epoch": 0.2262704603868873, + "grad_norm": 1.588135722670974, + "learning_rate": 3.6094061618354027e-06, + "loss": 1.0128, + "step": 2509 + }, + { + "epoch": 0.22636064391035757, + "grad_norm": 1.7564584900174958, + "learning_rate": 3.609059260484402e-06, + "loss": 0.9893, + "step": 2510 + }, + { + "epoch": 0.22645082743382783, + "grad_norm": 2.2022537781415252, + "learning_rate": 3.6087122218395935e-06, + "loss": 0.9328, + "step": 2511 + }, + { + "epoch": 0.2265410109572981, + "grad_norm": 1.8440890242959553, + "learning_rate": 3.608365045930587e-06, + "loss": 0.9554, + "step": 2512 + }, + { + "epoch": 0.22663119448076835, + "grad_norm": 1.757034807892789, + "learning_rate": 3.608017732787007e-06, + "loss": 0.9751, + "step": 2513 + }, + { + "epoch": 0.22672137800423864, + "grad_norm": 1.586071816755819, + "learning_rate": 3.6076702824384875e-06, + "loss": 0.9201, + "step": 2514 + }, + { + "epoch": 0.2268115615277089, + "grad_norm": 1.5165420589331051, + "learning_rate": 3.607322694914675e-06, + "loss": 1.0018, + "step": 2515 + }, + { + "epoch": 0.22690174505117916, + "grad_norm": 1.4651857510916582, + "learning_rate": 3.606974970245227e-06, + "loss": 0.9569, + "step": 2516 + }, + { + "epoch": 0.22699192857464942, + "grad_norm": 1.7319331119387193, + "learning_rate": 3.606627108459814e-06, + "loss": 0.9331, + "step": 2517 + }, + { + "epoch": 0.22708211209811968, + "grad_norm": 1.8978617230644195, + "learning_rate": 3.6062791095881174e-06, + "loss": 0.9111, + "step": 2518 + }, + { + "epoch": 0.22717229562158994, + "grad_norm": 1.6534164950518726, + "learning_rate": 3.6059309736598303e-06, + "loss": 1.0593, + "step": 2519 + }, + { + "epoch": 0.2272624791450602, + "grad_norm": 1.86708081311725, + "learning_rate": 3.605582700704657e-06, + "loss": 0.9613, + "step": 2520 + }, + { + "epoch": 0.22735266266853046, + "grad_norm": 1.6162959050550596, + "learning_rate": 3.6052342907523146e-06, + "loss": 0.9899, + "step": 2521 + }, + { + "epoch": 0.22744284619200072, + "grad_norm": 1.6105203349248784, + "learning_rate": 3.604885743832532e-06, + "loss": 0.9197, + "step": 2522 + }, + { + "epoch": 0.22753302971547099, + "grad_norm": 0.7668099867194786, + "learning_rate": 3.6045370599750482e-06, + "loss": 0.783, + "step": 2523 + }, + { + "epoch": 0.22762321323894125, + "grad_norm": 2.0415127235556123, + "learning_rate": 3.604188239209615e-06, + "loss": 0.9724, + "step": 2524 + }, + { + "epoch": 0.2277133967624115, + "grad_norm": 1.5749362417970156, + "learning_rate": 3.603839281565996e-06, + "loss": 0.9114, + "step": 2525 + }, + { + "epoch": 0.22780358028588177, + "grad_norm": 1.8309520804584314, + "learning_rate": 3.603490187073966e-06, + "loss": 1.007, + "step": 2526 + }, + { + "epoch": 0.22789376380935203, + "grad_norm": 1.6608643222309705, + "learning_rate": 3.6031409557633117e-06, + "loss": 1.0405, + "step": 2527 + }, + { + "epoch": 0.2279839473328223, + "grad_norm": 1.6464948733030147, + "learning_rate": 3.602791587663831e-06, + "loss": 1.03, + "step": 2528 + }, + { + "epoch": 0.22807413085629255, + "grad_norm": 1.4722997906881188, + "learning_rate": 3.6024420828053348e-06, + "loss": 1.024, + "step": 2529 + }, + { + "epoch": 0.2281643143797628, + "grad_norm": 1.5621274885542558, + "learning_rate": 3.6020924412176445e-06, + "loss": 1.0562, + "step": 2530 + }, + { + "epoch": 0.22825449790323307, + "grad_norm": 1.5484037455277706, + "learning_rate": 3.601742662930593e-06, + "loss": 0.9582, + "step": 2531 + }, + { + "epoch": 0.22834468142670333, + "grad_norm": 2.0747659252459703, + "learning_rate": 3.6013927479740248e-06, + "loss": 0.9501, + "step": 2532 + }, + { + "epoch": 0.2284348649501736, + "grad_norm": 1.6685308362507945, + "learning_rate": 3.6010426963777985e-06, + "loss": 1.0652, + "step": 2533 + }, + { + "epoch": 0.22852504847364385, + "grad_norm": 3.874556362229017, + "learning_rate": 3.6006925081717804e-06, + "loss": 1.0313, + "step": 2534 + }, + { + "epoch": 0.22861523199711412, + "grad_norm": 1.4413864837366337, + "learning_rate": 3.600342183385852e-06, + "loss": 0.9523, + "step": 2535 + }, + { + "epoch": 0.22870541552058438, + "grad_norm": 1.4781766273563515, + "learning_rate": 3.5999917220499043e-06, + "loss": 0.9226, + "step": 2536 + }, + { + "epoch": 0.22879559904405466, + "grad_norm": 1.6281162605856494, + "learning_rate": 3.5996411241938404e-06, + "loss": 0.9849, + "step": 2537 + }, + { + "epoch": 0.22888578256752493, + "grad_norm": 1.5951599915003123, + "learning_rate": 3.5992903898475752e-06, + "loss": 0.9446, + "step": 2538 + }, + { + "epoch": 0.2289759660909952, + "grad_norm": 1.515283683806527, + "learning_rate": 3.5989395190410365e-06, + "loss": 1.0845, + "step": 2539 + }, + { + "epoch": 0.22906614961446545, + "grad_norm": 1.659959058831226, + "learning_rate": 3.598588511804161e-06, + "loss": 0.9891, + "step": 2540 + }, + { + "epoch": 0.2291563331379357, + "grad_norm": 2.0292239134727152, + "learning_rate": 3.5982373681668987e-06, + "loss": 0.9916, + "step": 2541 + }, + { + "epoch": 0.22924651666140597, + "grad_norm": 2.0302856503661055, + "learning_rate": 3.597886088159212e-06, + "loss": 0.9955, + "step": 2542 + }, + { + "epoch": 0.22933670018487623, + "grad_norm": 1.638152593782694, + "learning_rate": 3.597534671811074e-06, + "loss": 0.918, + "step": 2543 + }, + { + "epoch": 0.2294268837083465, + "grad_norm": 1.4455600835768512, + "learning_rate": 3.5971831191524684e-06, + "loss": 0.9655, + "step": 2544 + }, + { + "epoch": 0.22951706723181675, + "grad_norm": 2.4228198761657356, + "learning_rate": 3.5968314302133925e-06, + "loss": 1.0377, + "step": 2545 + }, + { + "epoch": 0.229607250755287, + "grad_norm": 1.8074003364659763, + "learning_rate": 3.596479605023854e-06, + "loss": 0.9885, + "step": 2546 + }, + { + "epoch": 0.22969743427875727, + "grad_norm": 1.935463173112128, + "learning_rate": 3.596127643613873e-06, + "loss": 1.001, + "step": 2547 + }, + { + "epoch": 0.22978761780222753, + "grad_norm": 2.198331555685617, + "learning_rate": 3.59577554601348e-06, + "loss": 1.0061, + "step": 2548 + }, + { + "epoch": 0.2298778013256978, + "grad_norm": 1.419859914988848, + "learning_rate": 3.595423312252719e-06, + "loss": 1.036, + "step": 2549 + }, + { + "epoch": 0.22996798484916806, + "grad_norm": 1.7920963600085538, + "learning_rate": 3.5950709423616436e-06, + "loss": 0.9735, + "step": 2550 + }, + { + "epoch": 0.23005816837263832, + "grad_norm": 1.7896642880527271, + "learning_rate": 3.5947184363703203e-06, + "loss": 0.967, + "step": 2551 + }, + { + "epoch": 0.23014835189610858, + "grad_norm": 1.8216141397849672, + "learning_rate": 3.5943657943088274e-06, + "loss": 0.9352, + "step": 2552 + }, + { + "epoch": 0.23023853541957884, + "grad_norm": 1.3412444686420073, + "learning_rate": 3.5940130162072525e-06, + "loss": 0.909, + "step": 2553 + }, + { + "epoch": 0.2303287189430491, + "grad_norm": 1.7325612112133864, + "learning_rate": 3.5936601020956985e-06, + "loss": 0.9841, + "step": 2554 + }, + { + "epoch": 0.23041890246651936, + "grad_norm": 1.6677719107269187, + "learning_rate": 3.5933070520042772e-06, + "loss": 1.0232, + "step": 2555 + }, + { + "epoch": 0.23050908598998962, + "grad_norm": 0.863161035513336, + "learning_rate": 3.5929538659631133e-06, + "loss": 0.8331, + "step": 2556 + }, + { + "epoch": 0.23059926951345988, + "grad_norm": 1.6363396763251565, + "learning_rate": 3.592600544002341e-06, + "loss": 1.018, + "step": 2557 + }, + { + "epoch": 0.23068945303693014, + "grad_norm": 3.462157712467902, + "learning_rate": 3.5922470861521098e-06, + "loss": 0.9899, + "step": 2558 + }, + { + "epoch": 0.2307796365604004, + "grad_norm": 1.637402302979811, + "learning_rate": 3.591893492442577e-06, + "loss": 0.9593, + "step": 2559 + }, + { + "epoch": 0.23086982008387066, + "grad_norm": 1.7271922762400813, + "learning_rate": 3.591539762903914e-06, + "loss": 1.0504, + "step": 2560 + }, + { + "epoch": 0.23096000360734095, + "grad_norm": 1.6669655849700704, + "learning_rate": 3.591185897566303e-06, + "loss": 0.9193, + "step": 2561 + }, + { + "epoch": 0.2310501871308112, + "grad_norm": 1.5747259567661174, + "learning_rate": 3.590831896459937e-06, + "loss": 0.9874, + "step": 2562 + }, + { + "epoch": 0.23114037065428147, + "grad_norm": 1.7830582525393983, + "learning_rate": 3.5904777596150222e-06, + "loss": 0.9911, + "step": 2563 + }, + { + "epoch": 0.23123055417775173, + "grad_norm": 1.9273220498634291, + "learning_rate": 3.590123487061775e-06, + "loss": 0.9867, + "step": 2564 + }, + { + "epoch": 0.231320737701222, + "grad_norm": 1.8240406676451901, + "learning_rate": 3.589769078830424e-06, + "loss": 1.0872, + "step": 2565 + }, + { + "epoch": 0.23141092122469226, + "grad_norm": 1.6321265865253065, + "learning_rate": 3.58941453495121e-06, + "loss": 0.9682, + "step": 2566 + }, + { + "epoch": 0.23150110474816252, + "grad_norm": 1.5573629237514404, + "learning_rate": 3.5890598554543834e-06, + "loss": 0.9974, + "step": 2567 + }, + { + "epoch": 0.23159128827163278, + "grad_norm": 1.6313345759306117, + "learning_rate": 3.5887050403702073e-06, + "loss": 1.0384, + "step": 2568 + }, + { + "epoch": 0.23168147179510304, + "grad_norm": 1.6379996229131728, + "learning_rate": 3.588350089728958e-06, + "loss": 0.9603, + "step": 2569 + }, + { + "epoch": 0.2317716553185733, + "grad_norm": 1.484690100704317, + "learning_rate": 3.5879950035609204e-06, + "loss": 1.0125, + "step": 2570 + }, + { + "epoch": 0.23186183884204356, + "grad_norm": 1.7591875589642925, + "learning_rate": 3.5876397818963933e-06, + "loss": 0.983, + "step": 2571 + }, + { + "epoch": 0.23195202236551382, + "grad_norm": 1.708499163820834, + "learning_rate": 3.5872844247656858e-06, + "loss": 1.0417, + "step": 2572 + }, + { + "epoch": 0.23204220588898408, + "grad_norm": 1.4112954227513568, + "learning_rate": 3.5869289321991195e-06, + "loss": 0.9912, + "step": 2573 + }, + { + "epoch": 0.23213238941245434, + "grad_norm": 1.3755495533754878, + "learning_rate": 3.5865733042270263e-06, + "loss": 0.9767, + "step": 2574 + }, + { + "epoch": 0.2322225729359246, + "grad_norm": 1.5020125715895365, + "learning_rate": 3.5862175408797498e-06, + "loss": 1.0137, + "step": 2575 + }, + { + "epoch": 0.23231275645939486, + "grad_norm": 1.411925075634225, + "learning_rate": 3.585861642187647e-06, + "loss": 0.9247, + "step": 2576 + }, + { + "epoch": 0.23240293998286513, + "grad_norm": 1.426641836811886, + "learning_rate": 3.5855056081810845e-06, + "loss": 0.981, + "step": 2577 + }, + { + "epoch": 0.2324931235063354, + "grad_norm": 1.3979684053597325, + "learning_rate": 3.5851494388904406e-06, + "loss": 0.9877, + "step": 2578 + }, + { + "epoch": 0.23258330702980565, + "grad_norm": 1.8741039042171275, + "learning_rate": 3.5847931343461064e-06, + "loss": 0.9267, + "step": 2579 + }, + { + "epoch": 0.2326734905532759, + "grad_norm": 1.767393728511916, + "learning_rate": 3.5844366945784835e-06, + "loss": 0.9179, + "step": 2580 + }, + { + "epoch": 0.23276367407674617, + "grad_norm": 1.76723447378043, + "learning_rate": 3.5840801196179856e-06, + "loss": 1.0423, + "step": 2581 + }, + { + "epoch": 0.23285385760021643, + "grad_norm": 3.2265582142355766, + "learning_rate": 3.583723409495037e-06, + "loss": 1.043, + "step": 2582 + }, + { + "epoch": 0.2329440411236867, + "grad_norm": 1.5523175909778506, + "learning_rate": 3.5833665642400747e-06, + "loss": 1.0429, + "step": 2583 + }, + { + "epoch": 0.23303422464715695, + "grad_norm": 1.8042756638701667, + "learning_rate": 3.5830095838835472e-06, + "loss": 0.9126, + "step": 2584 + }, + { + "epoch": 0.23312440817062724, + "grad_norm": 1.5646077245173189, + "learning_rate": 3.5826524684559125e-06, + "loss": 0.9586, + "step": 2585 + }, + { + "epoch": 0.2332145916940975, + "grad_norm": 1.8151899305648562, + "learning_rate": 3.5822952179876433e-06, + "loss": 1.0666, + "step": 2586 + }, + { + "epoch": 0.23330477521756776, + "grad_norm": 1.4324909510834312, + "learning_rate": 3.5819378325092205e-06, + "loss": 0.959, + "step": 2587 + }, + { + "epoch": 0.23339495874103802, + "grad_norm": 1.4298489172025801, + "learning_rate": 3.581580312051139e-06, + "loss": 1.0092, + "step": 2588 + }, + { + "epoch": 0.23348514226450828, + "grad_norm": 2.2087232737281615, + "learning_rate": 3.5812226566439057e-06, + "loss": 0.9133, + "step": 2589 + }, + { + "epoch": 0.23357532578797854, + "grad_norm": 1.8751675848772855, + "learning_rate": 3.580864866318036e-06, + "loss": 0.9613, + "step": 2590 + }, + { + "epoch": 0.2336655093114488, + "grad_norm": 1.6800734675328883, + "learning_rate": 3.580506941104059e-06, + "loss": 0.9913, + "step": 2591 + }, + { + "epoch": 0.23375569283491907, + "grad_norm": 1.518038371831243, + "learning_rate": 3.580148881032515e-06, + "loss": 0.9797, + "step": 2592 + }, + { + "epoch": 0.23384587635838933, + "grad_norm": 1.441430564613266, + "learning_rate": 3.5797906861339556e-06, + "loss": 0.9906, + "step": 2593 + }, + { + "epoch": 0.2339360598818596, + "grad_norm": 1.5400917755374992, + "learning_rate": 3.5794323564389435e-06, + "loss": 1.0119, + "step": 2594 + }, + { + "epoch": 0.23402624340532985, + "grad_norm": 1.8908730966761795, + "learning_rate": 3.579073891978055e-06, + "loss": 0.9997, + "step": 2595 + }, + { + "epoch": 0.2341164269288001, + "grad_norm": 1.9039493497995614, + "learning_rate": 3.5787152927818746e-06, + "loss": 0.9514, + "step": 2596 + }, + { + "epoch": 0.23420661045227037, + "grad_norm": 1.4480582883810555, + "learning_rate": 3.5783565588810003e-06, + "loss": 1.0489, + "step": 2597 + }, + { + "epoch": 0.23429679397574063, + "grad_norm": 2.2324087387859906, + "learning_rate": 3.5779976903060412e-06, + "loss": 0.9641, + "step": 2598 + }, + { + "epoch": 0.2343869774992109, + "grad_norm": 1.909491050677268, + "learning_rate": 3.577638687087619e-06, + "loss": 1.0196, + "step": 2599 + }, + { + "epoch": 0.23447716102268115, + "grad_norm": 1.5851409114210222, + "learning_rate": 3.577279549256364e-06, + "loss": 0.9959, + "step": 2600 + }, + { + "epoch": 0.2345673445461514, + "grad_norm": 1.6141788673366486, + "learning_rate": 3.5769202768429213e-06, + "loss": 1.0353, + "step": 2601 + }, + { + "epoch": 0.23465752806962167, + "grad_norm": 1.7403024645156564, + "learning_rate": 3.5765608698779454e-06, + "loss": 0.9638, + "step": 2602 + }, + { + "epoch": 0.23474771159309193, + "grad_norm": 1.553040748936046, + "learning_rate": 3.5762013283921033e-06, + "loss": 0.9969, + "step": 2603 + }, + { + "epoch": 0.2348378951165622, + "grad_norm": 1.8559863844158349, + "learning_rate": 3.5758416524160728e-06, + "loss": 0.994, + "step": 2604 + }, + { + "epoch": 0.23492807864003246, + "grad_norm": 1.6124849363111342, + "learning_rate": 3.5754818419805427e-06, + "loss": 0.9801, + "step": 2605 + }, + { + "epoch": 0.23501826216350272, + "grad_norm": 1.528272965645399, + "learning_rate": 3.575121897116216e-06, + "loss": 0.9642, + "step": 2606 + }, + { + "epoch": 0.23510844568697298, + "grad_norm": 1.4336045600005414, + "learning_rate": 3.574761817853803e-06, + "loss": 1.0255, + "step": 2607 + }, + { + "epoch": 0.23519862921044324, + "grad_norm": 1.4120025805430836, + "learning_rate": 3.5744016042240287e-06, + "loss": 0.9239, + "step": 2608 + }, + { + "epoch": 0.23528881273391353, + "grad_norm": 1.8268240840053476, + "learning_rate": 3.5740412562576286e-06, + "loss": 0.9209, + "step": 2609 + }, + { + "epoch": 0.2353789962573838, + "grad_norm": 1.6042609579274723, + "learning_rate": 3.573680773985349e-06, + "loss": 0.9799, + "step": 2610 + }, + { + "epoch": 0.23546917978085405, + "grad_norm": 1.4358393987408342, + "learning_rate": 3.5733201574379486e-06, + "loss": 0.9822, + "step": 2611 + }, + { + "epoch": 0.2355593633043243, + "grad_norm": 1.6883301988643509, + "learning_rate": 3.5729594066461975e-06, + "loss": 0.9288, + "step": 2612 + }, + { + "epoch": 0.23564954682779457, + "grad_norm": 1.5218994984622998, + "learning_rate": 3.572598521640876e-06, + "loss": 0.9983, + "step": 2613 + }, + { + "epoch": 0.23573973035126483, + "grad_norm": 0.8634490803657506, + "learning_rate": 3.5722375024527782e-06, + "loss": 0.8376, + "step": 2614 + }, + { + "epoch": 0.2358299138747351, + "grad_norm": 1.3628842888048203, + "learning_rate": 3.571876349112707e-06, + "loss": 1.0081, + "step": 2615 + }, + { + "epoch": 0.23592009739820535, + "grad_norm": 1.6214454361318809, + "learning_rate": 3.5715150616514784e-06, + "loss": 0.9851, + "step": 2616 + }, + { + "epoch": 0.2360102809216756, + "grad_norm": 6.572581367787062, + "learning_rate": 3.5711536400999196e-06, + "loss": 1.04, + "step": 2617 + }, + { + "epoch": 0.23610046444514587, + "grad_norm": 0.8079082882312125, + "learning_rate": 3.570792084488869e-06, + "loss": 0.8396, + "step": 2618 + }, + { + "epoch": 0.23619064796861614, + "grad_norm": 0.7405811393401676, + "learning_rate": 3.5704303948491764e-06, + "loss": 0.8515, + "step": 2619 + }, + { + "epoch": 0.2362808314920864, + "grad_norm": 1.6415142737333603, + "learning_rate": 3.5700685712117035e-06, + "loss": 1.0101, + "step": 2620 + }, + { + "epoch": 0.23637101501555666, + "grad_norm": 2.0084595583850966, + "learning_rate": 3.5697066136073227e-06, + "loss": 0.9746, + "step": 2621 + }, + { + "epoch": 0.23646119853902692, + "grad_norm": 1.5854948581791428, + "learning_rate": 3.5693445220669184e-06, + "loss": 1.0025, + "step": 2622 + }, + { + "epoch": 0.23655138206249718, + "grad_norm": 1.3851913421884634, + "learning_rate": 3.568982296621386e-06, + "loss": 0.9797, + "step": 2623 + }, + { + "epoch": 0.23664156558596744, + "grad_norm": 1.7267798377596553, + "learning_rate": 3.5686199373016325e-06, + "loss": 1.0627, + "step": 2624 + }, + { + "epoch": 0.2367317491094377, + "grad_norm": 1.7789802813967643, + "learning_rate": 3.568257444138577e-06, + "loss": 0.9344, + "step": 2625 + }, + { + "epoch": 0.23682193263290796, + "grad_norm": 1.7451716981815952, + "learning_rate": 3.5678948171631495e-06, + "loss": 0.9858, + "step": 2626 + }, + { + "epoch": 0.23691211615637822, + "grad_norm": 1.592300878199751, + "learning_rate": 3.5675320564062908e-06, + "loss": 0.9954, + "step": 2627 + }, + { + "epoch": 0.23700229967984848, + "grad_norm": 1.7305193133749253, + "learning_rate": 3.5671691618989533e-06, + "loss": 0.8994, + "step": 2628 + }, + { + "epoch": 0.23709248320331874, + "grad_norm": 1.5513061096470442, + "learning_rate": 3.5668061336721024e-06, + "loss": 0.9978, + "step": 2629 + }, + { + "epoch": 0.237182666726789, + "grad_norm": 10.173588715327194, + "learning_rate": 3.5664429717567117e-06, + "loss": 0.9805, + "step": 2630 + }, + { + "epoch": 0.23727285025025927, + "grad_norm": 1.7675726621764687, + "learning_rate": 3.56607967618377e-06, + "loss": 1.0117, + "step": 2631 + }, + { + "epoch": 0.23736303377372953, + "grad_norm": 1.7747406058099777, + "learning_rate": 3.5657162469842754e-06, + "loss": 1.0016, + "step": 2632 + }, + { + "epoch": 0.23745321729719981, + "grad_norm": 1.5242894064883696, + "learning_rate": 3.5653526841892374e-06, + "loss": 0.9885, + "step": 2633 + }, + { + "epoch": 0.23754340082067008, + "grad_norm": 1.8677102478437368, + "learning_rate": 3.564988987829676e-06, + "loss": 1.021, + "step": 2634 + }, + { + "epoch": 0.23763358434414034, + "grad_norm": 1.532782818956985, + "learning_rate": 3.564625157936626e-06, + "loss": 0.9315, + "step": 2635 + }, + { + "epoch": 0.2377237678676106, + "grad_norm": 1.5061456034717964, + "learning_rate": 3.56426119454113e-06, + "loss": 1.0274, + "step": 2636 + }, + { + "epoch": 0.23781395139108086, + "grad_norm": 1.8539796013352894, + "learning_rate": 3.5638970976742436e-06, + "loss": 1.0631, + "step": 2637 + }, + { + "epoch": 0.23790413491455112, + "grad_norm": 1.885806666831987, + "learning_rate": 3.5635328673670335e-06, + "loss": 1.0173, + "step": 2638 + }, + { + "epoch": 0.23799431843802138, + "grad_norm": 1.4138626884792522, + "learning_rate": 3.5631685036505783e-06, + "loss": 1.0833, + "step": 2639 + }, + { + "epoch": 0.23808450196149164, + "grad_norm": 1.4383429667893206, + "learning_rate": 3.562804006555966e-06, + "loss": 1.002, + "step": 2640 + }, + { + "epoch": 0.2381746854849619, + "grad_norm": 1.6362317801583823, + "learning_rate": 3.5624393761143e-06, + "loss": 0.9293, + "step": 2641 + }, + { + "epoch": 0.23826486900843216, + "grad_norm": 1.6546907102501467, + "learning_rate": 3.5620746123566906e-06, + "loss": 0.9029, + "step": 2642 + }, + { + "epoch": 0.23835505253190242, + "grad_norm": 1.857990871686244, + "learning_rate": 3.5617097153142623e-06, + "loss": 0.9389, + "step": 2643 + }, + { + "epoch": 0.23844523605537268, + "grad_norm": 1.7716396141797937, + "learning_rate": 3.5613446850181497e-06, + "loss": 1.0319, + "step": 2644 + }, + { + "epoch": 0.23853541957884294, + "grad_norm": 0.7594241671873239, + "learning_rate": 3.5609795214994996e-06, + "loss": 0.7389, + "step": 2645 + }, + { + "epoch": 0.2386256031023132, + "grad_norm": 1.419062841264574, + "learning_rate": 3.560614224789469e-06, + "loss": 1.0437, + "step": 2646 + }, + { + "epoch": 0.23871578662578347, + "grad_norm": 1.5501474002769702, + "learning_rate": 3.5602487949192285e-06, + "loss": 0.937, + "step": 2647 + }, + { + "epoch": 0.23880597014925373, + "grad_norm": 1.753006804318668, + "learning_rate": 3.559883231919957e-06, + "loss": 0.9269, + "step": 2648 + }, + { + "epoch": 0.238896153672724, + "grad_norm": 1.3944919270760143, + "learning_rate": 3.5595175358228473e-06, + "loss": 0.9641, + "step": 2649 + }, + { + "epoch": 0.23898633719619425, + "grad_norm": 1.752622274576387, + "learning_rate": 3.5591517066591027e-06, + "loss": 0.9899, + "step": 2650 + }, + { + "epoch": 0.2390765207196645, + "grad_norm": 2.010688354898009, + "learning_rate": 3.5587857444599364e-06, + "loss": 0.9819, + "step": 2651 + }, + { + "epoch": 0.23916670424313477, + "grad_norm": 1.4661396448830273, + "learning_rate": 3.5584196492565766e-06, + "loss": 0.9574, + "step": 2652 + }, + { + "epoch": 0.23925688776660503, + "grad_norm": 1.6107921841899975, + "learning_rate": 3.5580534210802587e-06, + "loss": 1.0543, + "step": 2653 + }, + { + "epoch": 0.2393470712900753, + "grad_norm": 1.4195267286212907, + "learning_rate": 3.557687059962232e-06, + "loss": 1.0331, + "step": 2654 + }, + { + "epoch": 0.23943725481354555, + "grad_norm": 1.405033814848025, + "learning_rate": 3.5573205659337558e-06, + "loss": 0.9819, + "step": 2655 + }, + { + "epoch": 0.23952743833701584, + "grad_norm": 1.4358034488507434, + "learning_rate": 3.5569539390261025e-06, + "loss": 0.9723, + "step": 2656 + }, + { + "epoch": 0.2396176218604861, + "grad_norm": 1.2487450976270762, + "learning_rate": 3.5565871792705543e-06, + "loss": 1.0016, + "step": 2657 + }, + { + "epoch": 0.23970780538395636, + "grad_norm": 1.6345676315892774, + "learning_rate": 3.5562202866984045e-06, + "loss": 0.9794, + "step": 2658 + }, + { + "epoch": 0.23979798890742662, + "grad_norm": 1.6865249925351111, + "learning_rate": 3.5558532613409594e-06, + "loss": 1.0021, + "step": 2659 + }, + { + "epoch": 0.23988817243089688, + "grad_norm": 1.613162426579037, + "learning_rate": 3.555486103229535e-06, + "loss": 0.9304, + "step": 2660 + }, + { + "epoch": 0.23997835595436715, + "grad_norm": 1.6560827206761648, + "learning_rate": 3.5551188123954595e-06, + "loss": 1.0092, + "step": 2661 + }, + { + "epoch": 0.2400685394778374, + "grad_norm": 1.5394309904979877, + "learning_rate": 3.5547513888700715e-06, + "loss": 0.8834, + "step": 2662 + }, + { + "epoch": 0.24015872300130767, + "grad_norm": 1.4875867129143534, + "learning_rate": 3.5543838326847224e-06, + "loss": 1.0219, + "step": 2663 + }, + { + "epoch": 0.24024890652477793, + "grad_norm": 1.759273549165003, + "learning_rate": 3.5540161438707744e-06, + "loss": 0.9703, + "step": 2664 + }, + { + "epoch": 0.2403390900482482, + "grad_norm": 1.583936659991488, + "learning_rate": 3.5536483224596e-06, + "loss": 1.1448, + "step": 2665 + }, + { + "epoch": 0.24042927357171845, + "grad_norm": 1.5918931803209946, + "learning_rate": 3.553280368482584e-06, + "loss": 0.9643, + "step": 2666 + }, + { + "epoch": 0.2405194570951887, + "grad_norm": 1.7989696070132473, + "learning_rate": 3.5529122819711227e-06, + "loss": 0.8102, + "step": 2667 + }, + { + "epoch": 0.24060964061865897, + "grad_norm": 4.3620986360349585, + "learning_rate": 3.5525440629566223e-06, + "loss": 0.9353, + "step": 2668 + }, + { + "epoch": 0.24069982414212923, + "grad_norm": 1.4718650898528918, + "learning_rate": 3.552175711470502e-06, + "loss": 1.0046, + "step": 2669 + }, + { + "epoch": 0.2407900076655995, + "grad_norm": 1.6313079765218816, + "learning_rate": 3.5518072275441912e-06, + "loss": 1.0352, + "step": 2670 + }, + { + "epoch": 0.24088019118906975, + "grad_norm": 1.4236984298516266, + "learning_rate": 3.551438611209131e-06, + "loss": 1.0108, + "step": 2671 + }, + { + "epoch": 0.24097037471254001, + "grad_norm": 1.6544695947550296, + "learning_rate": 3.551069862496774e-06, + "loss": 1.0841, + "step": 2672 + }, + { + "epoch": 0.24106055823601027, + "grad_norm": 0.7502369109133432, + "learning_rate": 3.5507009814385846e-06, + "loss": 0.8006, + "step": 2673 + }, + { + "epoch": 0.24115074175948054, + "grad_norm": 0.6809848593164465, + "learning_rate": 3.550331968066036e-06, + "loss": 0.7916, + "step": 2674 + }, + { + "epoch": 0.2412409252829508, + "grad_norm": 1.6737880991330893, + "learning_rate": 3.549962822410616e-06, + "loss": 1.0028, + "step": 2675 + }, + { + "epoch": 0.24133110880642106, + "grad_norm": 0.9025534514134806, + "learning_rate": 3.5495935445038217e-06, + "loss": 0.7894, + "step": 2676 + }, + { + "epoch": 0.24142129232989132, + "grad_norm": 1.5872678391713928, + "learning_rate": 3.5492241343771612e-06, + "loss": 1.0256, + "step": 2677 + }, + { + "epoch": 0.24151147585336158, + "grad_norm": 1.608987113355736, + "learning_rate": 3.548854592062156e-06, + "loss": 0.9893, + "step": 2678 + }, + { + "epoch": 0.24160165937683184, + "grad_norm": 1.5072903692648083, + "learning_rate": 3.548484917590336e-06, + "loss": 0.967, + "step": 2679 + }, + { + "epoch": 0.24169184290030213, + "grad_norm": 1.8893170206585688, + "learning_rate": 3.5481151109932447e-06, + "loss": 0.9079, + "step": 2680 + }, + { + "epoch": 0.2417820264237724, + "grad_norm": 1.602540699029862, + "learning_rate": 3.5477451723024364e-06, + "loss": 1.0131, + "step": 2681 + }, + { + "epoch": 0.24187220994724265, + "grad_norm": 0.8554975948957638, + "learning_rate": 3.5473751015494757e-06, + "loss": 0.7737, + "step": 2682 + }, + { + "epoch": 0.2419623934707129, + "grad_norm": 1.757554302326278, + "learning_rate": 3.547004898765939e-06, + "loss": 1.0063, + "step": 2683 + }, + { + "epoch": 0.24205257699418317, + "grad_norm": 1.3577892657049093, + "learning_rate": 3.546634563983414e-06, + "loss": 1.0075, + "step": 2684 + }, + { + "epoch": 0.24214276051765343, + "grad_norm": 1.5419740971765787, + "learning_rate": 3.5462640972335002e-06, + "loss": 0.9543, + "step": 2685 + }, + { + "epoch": 0.2422329440411237, + "grad_norm": 1.3998920824827754, + "learning_rate": 3.5458934985478077e-06, + "loss": 1.0054, + "step": 2686 + }, + { + "epoch": 0.24232312756459395, + "grad_norm": 2.0451507294561613, + "learning_rate": 3.5455227679579577e-06, + "loss": 0.9678, + "step": 2687 + }, + { + "epoch": 0.24241331108806422, + "grad_norm": 2.138260586701117, + "learning_rate": 3.545151905495584e-06, + "loss": 1.0935, + "step": 2688 + }, + { + "epoch": 0.24250349461153448, + "grad_norm": 1.802753353328969, + "learning_rate": 3.544780911192329e-06, + "loss": 0.9661, + "step": 2689 + }, + { + "epoch": 0.24259367813500474, + "grad_norm": 1.5064168212443736, + "learning_rate": 3.544409785079849e-06, + "loss": 0.9374, + "step": 2690 + }, + { + "epoch": 0.242683861658475, + "grad_norm": 1.4798364421320473, + "learning_rate": 3.5440385271898103e-06, + "loss": 0.9353, + "step": 2691 + }, + { + "epoch": 0.24277404518194526, + "grad_norm": 1.7630006309114346, + "learning_rate": 3.5436671375538903e-06, + "loss": 0.9178, + "step": 2692 + }, + { + "epoch": 0.24286422870541552, + "grad_norm": 1.793874407983129, + "learning_rate": 3.543295616203779e-06, + "loss": 0.9881, + "step": 2693 + }, + { + "epoch": 0.24295441222888578, + "grad_norm": 1.4152906785168393, + "learning_rate": 3.542923963171176e-06, + "loss": 1.0806, + "step": 2694 + }, + { + "epoch": 0.24304459575235604, + "grad_norm": 1.3511291867714585, + "learning_rate": 3.542552178487793e-06, + "loss": 0.9087, + "step": 2695 + }, + { + "epoch": 0.2431347792758263, + "grad_norm": 1.6274461674613447, + "learning_rate": 3.5421802621853523e-06, + "loss": 0.9797, + "step": 2696 + }, + { + "epoch": 0.24322496279929656, + "grad_norm": 1.5634073293345507, + "learning_rate": 3.5418082142955887e-06, + "loss": 0.9858, + "step": 2697 + }, + { + "epoch": 0.24331514632276682, + "grad_norm": 0.8229934117388386, + "learning_rate": 3.5414360348502463e-06, + "loss": 0.8173, + "step": 2698 + }, + { + "epoch": 0.24340532984623708, + "grad_norm": 1.4213411775328397, + "learning_rate": 3.5410637238810825e-06, + "loss": 0.8923, + "step": 2699 + }, + { + "epoch": 0.24349551336970734, + "grad_norm": 1.8020050458128267, + "learning_rate": 3.5406912814198635e-06, + "loss": 0.9421, + "step": 2700 + }, + { + "epoch": 0.2435856968931776, + "grad_norm": 2.0230343685879477, + "learning_rate": 3.54031870749837e-06, + "loss": 0.9503, + "step": 2701 + }, + { + "epoch": 0.24367588041664787, + "grad_norm": 1.5001756247385885, + "learning_rate": 3.539946002148391e-06, + "loss": 1.0033, + "step": 2702 + }, + { + "epoch": 0.24376606394011813, + "grad_norm": 1.943847042985334, + "learning_rate": 3.5395731654017277e-06, + "loss": 0.988, + "step": 2703 + }, + { + "epoch": 0.24385624746358842, + "grad_norm": 1.8050621931821333, + "learning_rate": 3.5392001972901923e-06, + "loss": 1.0315, + "step": 2704 + }, + { + "epoch": 0.24394643098705868, + "grad_norm": 3.1030718688411585, + "learning_rate": 3.5388270978456098e-06, + "loss": 0.9523, + "step": 2705 + }, + { + "epoch": 0.24403661451052894, + "grad_norm": 3.2555623505412576, + "learning_rate": 3.5384538670998137e-06, + "loss": 0.9593, + "step": 2706 + }, + { + "epoch": 0.2441267980339992, + "grad_norm": 1.6100169262677497, + "learning_rate": 3.538080505084651e-06, + "loss": 1.0078, + "step": 2707 + }, + { + "epoch": 0.24421698155746946, + "grad_norm": 1.444226656088051, + "learning_rate": 3.5377070118319788e-06, + "loss": 0.9994, + "step": 2708 + }, + { + "epoch": 0.24430716508093972, + "grad_norm": 2.0542147132356274, + "learning_rate": 3.5373333873736657e-06, + "loss": 0.945, + "step": 2709 + }, + { + "epoch": 0.24439734860440998, + "grad_norm": 1.3817843046752671, + "learning_rate": 3.536959631741591e-06, + "loss": 0.9944, + "step": 2710 + }, + { + "epoch": 0.24448753212788024, + "grad_norm": 1.1976826425370357, + "learning_rate": 3.536585744967646e-06, + "loss": 1.0133, + "step": 2711 + }, + { + "epoch": 0.2445777156513505, + "grad_norm": 1.744153747850968, + "learning_rate": 3.5362117270837326e-06, + "loss": 0.9723, + "step": 2712 + }, + { + "epoch": 0.24466789917482076, + "grad_norm": 1.2986000721281505, + "learning_rate": 3.5358375781217634e-06, + "loss": 0.8919, + "step": 2713 + }, + { + "epoch": 0.24475808269829102, + "grad_norm": 1.5504668947712545, + "learning_rate": 3.535463298113664e-06, + "loss": 1.017, + "step": 2714 + }, + { + "epoch": 0.24484826622176128, + "grad_norm": 1.7056657230915568, + "learning_rate": 3.5350888870913697e-06, + "loss": 0.9467, + "step": 2715 + }, + { + "epoch": 0.24493844974523155, + "grad_norm": 1.7096986509561558, + "learning_rate": 3.5347143450868273e-06, + "loss": 0.9945, + "step": 2716 + }, + { + "epoch": 0.2450286332687018, + "grad_norm": 1.6459642993953523, + "learning_rate": 3.534339672131994e-06, + "loss": 1.007, + "step": 2717 + }, + { + "epoch": 0.24511881679217207, + "grad_norm": 1.4409360041018027, + "learning_rate": 3.5339648682588397e-06, + "loss": 0.9177, + "step": 2718 + }, + { + "epoch": 0.24520900031564233, + "grad_norm": 1.7329171041659461, + "learning_rate": 3.533589933499345e-06, + "loss": 1.0327, + "step": 2719 + }, + { + "epoch": 0.2452991838391126, + "grad_norm": 1.8057964930394792, + "learning_rate": 3.533214867885501e-06, + "loss": 0.9203, + "step": 2720 + }, + { + "epoch": 0.24538936736258285, + "grad_norm": 1.8171999274080646, + "learning_rate": 3.53283967144931e-06, + "loss": 0.9252, + "step": 2721 + }, + { + "epoch": 0.2454795508860531, + "grad_norm": 1.6563372319119893, + "learning_rate": 3.532464344222787e-06, + "loss": 0.9992, + "step": 2722 + }, + { + "epoch": 0.24556973440952337, + "grad_norm": 2.253003552850122, + "learning_rate": 3.532088886237956e-06, + "loss": 1.0283, + "step": 2723 + }, + { + "epoch": 0.24565991793299363, + "grad_norm": 0.773210395839702, + "learning_rate": 3.5317132975268535e-06, + "loss": 0.7293, + "step": 2724 + }, + { + "epoch": 0.2457501014564639, + "grad_norm": 1.6839637438298507, + "learning_rate": 3.531337578121526e-06, + "loss": 1.0093, + "step": 2725 + }, + { + "epoch": 0.24584028497993415, + "grad_norm": 1.6453533518937173, + "learning_rate": 3.530961728054033e-06, + "loss": 1.0224, + "step": 2726 + }, + { + "epoch": 0.24593046850340441, + "grad_norm": 1.5079371741738756, + "learning_rate": 3.5305857473564435e-06, + "loss": 0.9479, + "step": 2727 + }, + { + "epoch": 0.2460206520268747, + "grad_norm": 2.2971270156970243, + "learning_rate": 3.5302096360608385e-06, + "loss": 0.9162, + "step": 2728 + }, + { + "epoch": 0.24611083555034496, + "grad_norm": 1.6363779227519377, + "learning_rate": 3.5298333941993105e-06, + "loss": 1.0374, + "step": 2729 + }, + { + "epoch": 0.24620101907381522, + "grad_norm": 1.8908295323734516, + "learning_rate": 3.529457021803962e-06, + "loss": 0.9797, + "step": 2730 + }, + { + "epoch": 0.24629120259728549, + "grad_norm": 1.9645166447129243, + "learning_rate": 3.529080518906906e-06, + "loss": 1.0036, + "step": 2731 + }, + { + "epoch": 0.24638138612075575, + "grad_norm": 1.3030258378828752, + "learning_rate": 3.5287038855402696e-06, + "loss": 0.9701, + "step": 2732 + }, + { + "epoch": 0.246471569644226, + "grad_norm": 1.9760363334566364, + "learning_rate": 3.528327121736188e-06, + "loss": 1.0235, + "step": 2733 + }, + { + "epoch": 0.24656175316769627, + "grad_norm": 1.5684528631752648, + "learning_rate": 3.52795022752681e-06, + "loss": 1.0698, + "step": 2734 + }, + { + "epoch": 0.24665193669116653, + "grad_norm": 2.018408575800134, + "learning_rate": 3.5275732029442925e-06, + "loss": 1.0594, + "step": 2735 + }, + { + "epoch": 0.2467421202146368, + "grad_norm": 2.216819998330667, + "learning_rate": 3.5271960480208077e-06, + "loss": 0.9297, + "step": 2736 + }, + { + "epoch": 0.24683230373810705, + "grad_norm": 0.7669695511584373, + "learning_rate": 3.526818762788534e-06, + "loss": 0.8304, + "step": 2737 + }, + { + "epoch": 0.2469224872615773, + "grad_norm": 1.6491073043083178, + "learning_rate": 3.5264413472796653e-06, + "loss": 1.0259, + "step": 2738 + }, + { + "epoch": 0.24701267078504757, + "grad_norm": 2.537359235645704, + "learning_rate": 3.5260638015264037e-06, + "loss": 0.9674, + "step": 2739 + }, + { + "epoch": 0.24710285430851783, + "grad_norm": 1.3179837518278033, + "learning_rate": 3.5256861255609644e-06, + "loss": 0.9588, + "step": 2740 + }, + { + "epoch": 0.2471930378319881, + "grad_norm": 1.6347592808316638, + "learning_rate": 3.5253083194155723e-06, + "loss": 0.9995, + "step": 2741 + }, + { + "epoch": 0.24728322135545835, + "grad_norm": 1.5264914395554305, + "learning_rate": 3.5249303831224637e-06, + "loss": 1.0075, + "step": 2742 + }, + { + "epoch": 0.24737340487892862, + "grad_norm": 1.409399151647991, + "learning_rate": 3.524552316713887e-06, + "loss": 0.9746, + "step": 2743 + }, + { + "epoch": 0.24746358840239888, + "grad_norm": 1.6584374000573725, + "learning_rate": 3.5241741202220995e-06, + "loss": 1.0477, + "step": 2744 + }, + { + "epoch": 0.24755377192586914, + "grad_norm": 1.4011109014750105, + "learning_rate": 3.5237957936793724e-06, + "loss": 0.9753, + "step": 2745 + }, + { + "epoch": 0.2476439554493394, + "grad_norm": 1.435800792010214, + "learning_rate": 3.523417337117986e-06, + "loss": 0.9491, + "step": 2746 + }, + { + "epoch": 0.24773413897280966, + "grad_norm": 1.37965333924389, + "learning_rate": 3.523038750570232e-06, + "loss": 0.9376, + "step": 2747 + }, + { + "epoch": 0.24782432249627992, + "grad_norm": 3.1926010304642953, + "learning_rate": 3.522660034068414e-06, + "loss": 1.0186, + "step": 2748 + }, + { + "epoch": 0.24791450601975018, + "grad_norm": 1.8014191596119276, + "learning_rate": 3.5222811876448464e-06, + "loss": 1.0008, + "step": 2749 + }, + { + "epoch": 0.24800468954322044, + "grad_norm": 1.386398665330996, + "learning_rate": 3.521902211331854e-06, + "loss": 1.004, + "step": 2750 + }, + { + "epoch": 0.2480948730666907, + "grad_norm": 1.5493975022110331, + "learning_rate": 3.5215231051617726e-06, + "loss": 0.937, + "step": 2751 + }, + { + "epoch": 0.248185056590161, + "grad_norm": 1.6458285689787977, + "learning_rate": 3.521143869166951e-06, + "loss": 0.8894, + "step": 2752 + }, + { + "epoch": 0.24827524011363125, + "grad_norm": 2.0078905142385604, + "learning_rate": 3.5207645033797464e-06, + "loss": 1.0193, + "step": 2753 + }, + { + "epoch": 0.2483654236371015, + "grad_norm": 1.6600491118921883, + "learning_rate": 3.5203850078325293e-06, + "loss": 0.9125, + "step": 2754 + }, + { + "epoch": 0.24845560716057177, + "grad_norm": 1.8065285807739317, + "learning_rate": 3.5200053825576797e-06, + "loss": 0.8979, + "step": 2755 + }, + { + "epoch": 0.24854579068404203, + "grad_norm": 1.5083242546075275, + "learning_rate": 3.51962562758759e-06, + "loss": 0.9855, + "step": 2756 + }, + { + "epoch": 0.2486359742075123, + "grad_norm": 1.4779308780359368, + "learning_rate": 3.5192457429546627e-06, + "loss": 0.9644, + "step": 2757 + }, + { + "epoch": 0.24872615773098256, + "grad_norm": 1.5433908392214766, + "learning_rate": 3.5188657286913115e-06, + "loss": 0.9534, + "step": 2758 + }, + { + "epoch": 0.24881634125445282, + "grad_norm": 1.9904340619509016, + "learning_rate": 3.518485584829961e-06, + "loss": 0.9514, + "step": 2759 + }, + { + "epoch": 0.24890652477792308, + "grad_norm": 1.506383030505431, + "learning_rate": 3.5181053114030485e-06, + "loss": 0.9933, + "step": 2760 + }, + { + "epoch": 0.24899670830139334, + "grad_norm": 1.6611783819210655, + "learning_rate": 3.5177249084430198e-06, + "loss": 0.9816, + "step": 2761 + }, + { + "epoch": 0.2490868918248636, + "grad_norm": 1.8866204972678156, + "learning_rate": 3.517344375982333e-06, + "loss": 1.019, + "step": 2762 + }, + { + "epoch": 0.24917707534833386, + "grad_norm": 2.117734155780495, + "learning_rate": 3.5169637140534565e-06, + "loss": 1.0219, + "step": 2763 + }, + { + "epoch": 0.24926725887180412, + "grad_norm": 1.8543548774206016, + "learning_rate": 3.5165829226888733e-06, + "loss": 0.9567, + "step": 2764 + }, + { + "epoch": 0.24935744239527438, + "grad_norm": 1.4116509037571467, + "learning_rate": 3.516202001921072e-06, + "loss": 1.0112, + "step": 2765 + }, + { + "epoch": 0.24944762591874464, + "grad_norm": 1.539876732165761, + "learning_rate": 3.515820951782555e-06, + "loss": 1.0106, + "step": 2766 + }, + { + "epoch": 0.2495378094422149, + "grad_norm": 1.754201273754224, + "learning_rate": 3.5154397723058366e-06, + "loss": 0.907, + "step": 2767 + }, + { + "epoch": 0.24962799296568516, + "grad_norm": 1.7857176317456096, + "learning_rate": 3.5150584635234416e-06, + "loss": 1.0444, + "step": 2768 + }, + { + "epoch": 0.24971817648915542, + "grad_norm": 1.390203947981386, + "learning_rate": 3.5146770254679035e-06, + "loss": 1.0288, + "step": 2769 + }, + { + "epoch": 0.24980836001262569, + "grad_norm": 1.580137951900638, + "learning_rate": 3.51429545817177e-06, + "loss": 0.9066, + "step": 2770 + }, + { + "epoch": 0.24989854353609595, + "grad_norm": 1.6112372725334028, + "learning_rate": 3.5139137616675985e-06, + "loss": 0.9898, + "step": 2771 + }, + { + "epoch": 0.2499887270595662, + "grad_norm": 1.5819313300332205, + "learning_rate": 3.513531935987957e-06, + "loss": 0.9243, + "step": 2772 + }, + { + "epoch": 0.2500789105830365, + "grad_norm": 1.5004730273488396, + "learning_rate": 3.5131499811654253e-06, + "loss": 1.0031, + "step": 2773 + }, + { + "epoch": 0.25016909410650673, + "grad_norm": 2.7270052749812184, + "learning_rate": 3.512767897232594e-06, + "loss": 0.9501, + "step": 2774 + }, + { + "epoch": 0.250259277629977, + "grad_norm": 1.6148780881850857, + "learning_rate": 3.512385684222064e-06, + "loss": 0.9824, + "step": 2775 + }, + { + "epoch": 0.25034946115344725, + "grad_norm": 1.883728121280498, + "learning_rate": 3.512003342166449e-06, + "loss": 1.0832, + "step": 2776 + }, + { + "epoch": 0.25043964467691754, + "grad_norm": 1.4311498448673168, + "learning_rate": 3.511620871098371e-06, + "loss": 1.0201, + "step": 2777 + }, + { + "epoch": 0.25052982820038777, + "grad_norm": 1.6054061543311648, + "learning_rate": 3.511238271050465e-06, + "loss": 1.0138, + "step": 2778 + }, + { + "epoch": 0.25062001172385806, + "grad_norm": 1.5781861850472751, + "learning_rate": 3.5108555420553778e-06, + "loss": 0.9024, + "step": 2779 + }, + { + "epoch": 0.2507101952473283, + "grad_norm": 1.6296961490624695, + "learning_rate": 3.510472684145764e-06, + "loss": 0.8001, + "step": 2780 + }, + { + "epoch": 0.2508003787707986, + "grad_norm": 1.6021913363889029, + "learning_rate": 3.5100896973542926e-06, + "loss": 0.8756, + "step": 2781 + }, + { + "epoch": 0.2508905622942688, + "grad_norm": 1.7355982959796095, + "learning_rate": 3.509706581713642e-06, + "loss": 1.0204, + "step": 2782 + }, + { + "epoch": 0.2509807458177391, + "grad_norm": 1.805395870809417, + "learning_rate": 3.509323337256501e-06, + "loss": 1.0144, + "step": 2783 + }, + { + "epoch": 0.25107092934120934, + "grad_norm": 1.635200761607491, + "learning_rate": 3.5089399640155703e-06, + "loss": 0.9371, + "step": 2784 + }, + { + "epoch": 0.2511611128646796, + "grad_norm": 1.333749060590247, + "learning_rate": 3.508556462023562e-06, + "loss": 0.9776, + "step": 2785 + }, + { + "epoch": 0.25125129638814986, + "grad_norm": 1.7311624553043177, + "learning_rate": 3.5081728313131984e-06, + "loss": 0.9646, + "step": 2786 + }, + { + "epoch": 0.25134147991162015, + "grad_norm": 1.9572733655083776, + "learning_rate": 3.5077890719172125e-06, + "loss": 0.9834, + "step": 2787 + }, + { + "epoch": 0.25143166343509044, + "grad_norm": 1.5845815941460604, + "learning_rate": 3.5074051838683497e-06, + "loss": 0.9845, + "step": 2788 + }, + { + "epoch": 0.25152184695856067, + "grad_norm": 1.6652865497201474, + "learning_rate": 3.5070211671993643e-06, + "loss": 0.9091, + "step": 2789 + }, + { + "epoch": 0.25161203048203096, + "grad_norm": 1.6792506159897276, + "learning_rate": 3.5066370219430238e-06, + "loss": 1.0083, + "step": 2790 + }, + { + "epoch": 0.2517022140055012, + "grad_norm": 1.854530305957803, + "learning_rate": 3.5062527481321044e-06, + "loss": 0.9813, + "step": 2791 + }, + { + "epoch": 0.2517923975289715, + "grad_norm": 1.8677366718015311, + "learning_rate": 3.5058683457993954e-06, + "loss": 0.9972, + "step": 2792 + }, + { + "epoch": 0.2518825810524417, + "grad_norm": 0.8017108266911387, + "learning_rate": 3.5054838149776963e-06, + "loss": 0.7379, + "step": 2793 + }, + { + "epoch": 0.251972764575912, + "grad_norm": 1.6127063382633342, + "learning_rate": 3.505099155699816e-06, + "loss": 1.021, + "step": 2794 + }, + { + "epoch": 0.25206294809938223, + "grad_norm": 1.582475161967678, + "learning_rate": 3.5047143679985775e-06, + "loss": 0.8602, + "step": 2795 + }, + { + "epoch": 0.2521531316228525, + "grad_norm": 1.3595955329992226, + "learning_rate": 3.5043294519068126e-06, + "loss": 0.9943, + "step": 2796 + }, + { + "epoch": 0.25224331514632276, + "grad_norm": 1.6436288128987835, + "learning_rate": 3.503944407457363e-06, + "loss": 0.9032, + "step": 2797 + }, + { + "epoch": 0.25233349866979304, + "grad_norm": 1.857924079592211, + "learning_rate": 3.5035592346830846e-06, + "loss": 0.9988, + "step": 2798 + }, + { + "epoch": 0.2524236821932633, + "grad_norm": 1.7652534161372424, + "learning_rate": 3.503173933616841e-06, + "loss": 1.0545, + "step": 2799 + }, + { + "epoch": 0.25251386571673357, + "grad_norm": 1.6146885089047933, + "learning_rate": 3.50278850429151e-06, + "loss": 1.0527, + "step": 2800 + }, + { + "epoch": 0.2526040492402038, + "grad_norm": 1.7561996181225663, + "learning_rate": 3.502402946739977e-06, + "loss": 1.0075, + "step": 2801 + }, + { + "epoch": 0.2526942327636741, + "grad_norm": 1.5087787911766077, + "learning_rate": 3.5020172609951405e-06, + "loss": 1.0047, + "step": 2802 + }, + { + "epoch": 0.2527844162871443, + "grad_norm": 1.3944997917495194, + "learning_rate": 3.501631447089909e-06, + "loss": 0.9998, + "step": 2803 + }, + { + "epoch": 0.2528745998106146, + "grad_norm": 1.9986369733600293, + "learning_rate": 3.501245505057203e-06, + "loss": 0.9566, + "step": 2804 + }, + { + "epoch": 0.25296478333408484, + "grad_norm": 1.4561731506502524, + "learning_rate": 3.5008594349299526e-06, + "loss": 1.0559, + "step": 2805 + }, + { + "epoch": 0.25305496685755513, + "grad_norm": 1.8864282103952044, + "learning_rate": 3.500473236741099e-06, + "loss": 0.9723, + "step": 2806 + }, + { + "epoch": 0.25314515038102536, + "grad_norm": 1.47244714522262, + "learning_rate": 3.500086910523596e-06, + "loss": 0.9423, + "step": 2807 + }, + { + "epoch": 0.25323533390449565, + "grad_norm": 1.3979371950070154, + "learning_rate": 3.499700456310406e-06, + "loss": 0.9365, + "step": 2808 + }, + { + "epoch": 0.2533255174279659, + "grad_norm": 1.339375628223352, + "learning_rate": 3.499313874134504e-06, + "loss": 1.0341, + "step": 2809 + }, + { + "epoch": 0.2534157009514362, + "grad_norm": 1.7201328609961382, + "learning_rate": 3.498927164028875e-06, + "loss": 0.8571, + "step": 2810 + }, + { + "epoch": 0.25350588447490646, + "grad_norm": 1.2429249331977028, + "learning_rate": 3.498540326026515e-06, + "loss": 1.0218, + "step": 2811 + }, + { + "epoch": 0.2535960679983767, + "grad_norm": 1.3755611228410805, + "learning_rate": 3.4981533601604323e-06, + "loss": 1.0533, + "step": 2812 + }, + { + "epoch": 0.253686251521847, + "grad_norm": 1.5560994086884228, + "learning_rate": 3.4977662664636443e-06, + "loss": 0.9914, + "step": 2813 + }, + { + "epoch": 0.2537764350453172, + "grad_norm": 1.632079110327172, + "learning_rate": 3.497379044969179e-06, + "loss": 0.9432, + "step": 2814 + }, + { + "epoch": 0.2538666185687875, + "grad_norm": 1.641335033994321, + "learning_rate": 3.4969916957100777e-06, + "loss": 0.9804, + "step": 2815 + }, + { + "epoch": 0.25395680209225774, + "grad_norm": 1.4131922704923103, + "learning_rate": 3.4966042187193905e-06, + "loss": 0.9758, + "step": 2816 + }, + { + "epoch": 0.254046985615728, + "grad_norm": 1.5335854606384958, + "learning_rate": 3.496216614030179e-06, + "loss": 0.9806, + "step": 2817 + }, + { + "epoch": 0.25413716913919826, + "grad_norm": 1.8262552334124955, + "learning_rate": 3.495828881675516e-06, + "loss": 0.8843, + "step": 2818 + }, + { + "epoch": 0.25422735266266855, + "grad_norm": 3.354305256572881, + "learning_rate": 3.4954410216884845e-06, + "loss": 1.0562, + "step": 2819 + }, + { + "epoch": 0.2543175361861388, + "grad_norm": 1.8001642390662345, + "learning_rate": 3.49505303410218e-06, + "loss": 0.9912, + "step": 2820 + }, + { + "epoch": 0.25440771970960907, + "grad_norm": 1.6044719042340505, + "learning_rate": 3.4946649189497067e-06, + "loss": 1.0116, + "step": 2821 + }, + { + "epoch": 0.2544979032330793, + "grad_norm": 2.7660052366779486, + "learning_rate": 3.4942766762641805e-06, + "loss": 0.974, + "step": 2822 + }, + { + "epoch": 0.2545880867565496, + "grad_norm": 1.7818180316233174, + "learning_rate": 3.49388830607873e-06, + "loss": 0.9993, + "step": 2823 + }, + { + "epoch": 0.2546782702800198, + "grad_norm": 1.2496731330749729, + "learning_rate": 3.493499808426491e-06, + "loss": 0.9481, + "step": 2824 + }, + { + "epoch": 0.2547684538034901, + "grad_norm": 2.9593703015138093, + "learning_rate": 3.493111183340614e-06, + "loss": 1.0374, + "step": 2825 + }, + { + "epoch": 0.25485863732696035, + "grad_norm": 2.093825609350752, + "learning_rate": 3.4927224308542576e-06, + "loss": 0.8839, + "step": 2826 + }, + { + "epoch": 0.25494882085043064, + "grad_norm": 1.4686012192710813, + "learning_rate": 3.4923335510005923e-06, + "loss": 1.0605, + "step": 2827 + }, + { + "epoch": 0.25503900437390087, + "grad_norm": 1.4242328754002878, + "learning_rate": 3.4919445438128e-06, + "loss": 1.0038, + "step": 2828 + }, + { + "epoch": 0.25512918789737116, + "grad_norm": 2.0489785836451713, + "learning_rate": 3.491555409324073e-06, + "loss": 1.0634, + "step": 2829 + }, + { + "epoch": 0.2552193714208414, + "grad_norm": 0.7747676470292495, + "learning_rate": 3.4911661475676136e-06, + "loss": 0.7614, + "step": 2830 + }, + { + "epoch": 0.2553095549443117, + "grad_norm": 1.4664956506091709, + "learning_rate": 3.490776758576637e-06, + "loss": 0.9591, + "step": 2831 + }, + { + "epoch": 0.2553997384677819, + "grad_norm": 1.4991562377555636, + "learning_rate": 3.4903872423843668e-06, + "loss": 1.0281, + "step": 2832 + }, + { + "epoch": 0.2554899219912522, + "grad_norm": 1.5711828448405956, + "learning_rate": 3.4899975990240396e-06, + "loss": 0.946, + "step": 2833 + }, + { + "epoch": 0.25558010551472243, + "grad_norm": 2.040002131367486, + "learning_rate": 3.489607828528901e-06, + "loss": 1.0056, + "step": 2834 + }, + { + "epoch": 0.2556702890381927, + "grad_norm": 0.7866039309769327, + "learning_rate": 3.4892179309322093e-06, + "loss": 0.8712, + "step": 2835 + }, + { + "epoch": 0.255760472561663, + "grad_norm": 1.5250860440103475, + "learning_rate": 3.488827906267232e-06, + "loss": 0.8972, + "step": 2836 + }, + { + "epoch": 0.25585065608513324, + "grad_norm": 1.7111952078095434, + "learning_rate": 3.4884377545672485e-06, + "loss": 1.0234, + "step": 2837 + }, + { + "epoch": 0.25594083960860353, + "grad_norm": 1.5180791275801813, + "learning_rate": 3.4880474758655485e-06, + "loss": 0.9538, + "step": 2838 + }, + { + "epoch": 0.25603102313207377, + "grad_norm": 1.6474679563471186, + "learning_rate": 3.487657070195433e-06, + "loss": 1.0508, + "step": 2839 + }, + { + "epoch": 0.25612120665554405, + "grad_norm": 1.5216845477426473, + "learning_rate": 3.487266537590213e-06, + "loss": 1.0592, + "step": 2840 + }, + { + "epoch": 0.2562113901790143, + "grad_norm": 1.4703932963923085, + "learning_rate": 3.4868758780832116e-06, + "loss": 0.9017, + "step": 2841 + }, + { + "epoch": 0.2563015737024846, + "grad_norm": 1.5831583411572132, + "learning_rate": 3.486485091707762e-06, + "loss": 0.9949, + "step": 2842 + }, + { + "epoch": 0.2563917572259548, + "grad_norm": 1.7757222922288982, + "learning_rate": 3.4860941784972077e-06, + "loss": 0.8649, + "step": 2843 + }, + { + "epoch": 0.2564819407494251, + "grad_norm": 3.075095917981284, + "learning_rate": 3.485703138484904e-06, + "loss": 1.0766, + "step": 2844 + }, + { + "epoch": 0.25657212427289533, + "grad_norm": 1.5423820828912735, + "learning_rate": 3.485311971704216e-06, + "loss": 1.0345, + "step": 2845 + }, + { + "epoch": 0.2566623077963656, + "grad_norm": 1.4610252150044447, + "learning_rate": 3.484920678188521e-06, + "loss": 1.0126, + "step": 2846 + }, + { + "epoch": 0.25675249131983585, + "grad_norm": 1.7432173349484064, + "learning_rate": 3.4845292579712063e-06, + "loss": 1.0091, + "step": 2847 + }, + { + "epoch": 0.25684267484330614, + "grad_norm": 1.7021029844359754, + "learning_rate": 3.484137711085669e-06, + "loss": 0.8648, + "step": 2848 + }, + { + "epoch": 0.2569328583667764, + "grad_norm": 1.4074876213545742, + "learning_rate": 3.4837460375653198e-06, + "loss": 0.9716, + "step": 2849 + }, + { + "epoch": 0.25702304189024666, + "grad_norm": 1.386242164059946, + "learning_rate": 3.483354237443576e-06, + "loss": 1.0139, + "step": 2850 + }, + { + "epoch": 0.2571132254137169, + "grad_norm": 1.3055003940255876, + "learning_rate": 3.48296231075387e-06, + "loss": 0.9228, + "step": 2851 + }, + { + "epoch": 0.2572034089371872, + "grad_norm": 1.3651535599655749, + "learning_rate": 3.4825702575296433e-06, + "loss": 1.0006, + "step": 2852 + }, + { + "epoch": 0.2572935924606574, + "grad_norm": 0.7789285884244256, + "learning_rate": 3.482178077804347e-06, + "loss": 0.8153, + "step": 2853 + }, + { + "epoch": 0.2573837759841277, + "grad_norm": 1.6266688067521156, + "learning_rate": 3.4817857716114443e-06, + "loss": 0.9442, + "step": 2854 + }, + { + "epoch": 0.25747395950759794, + "grad_norm": 1.5677393997848676, + "learning_rate": 3.4813933389844094e-06, + "loss": 0.9372, + "step": 2855 + }, + { + "epoch": 0.2575641430310682, + "grad_norm": 1.594464908399119, + "learning_rate": 3.4810007799567264e-06, + "loss": 0.9916, + "step": 2856 + }, + { + "epoch": 0.25765432655453846, + "grad_norm": 1.7200511515574177, + "learning_rate": 3.480608094561891e-06, + "loss": 1.0025, + "step": 2857 + }, + { + "epoch": 0.25774451007800875, + "grad_norm": 1.523434604739837, + "learning_rate": 3.4802152828334083e-06, + "loss": 1.0046, + "step": 2858 + }, + { + "epoch": 0.25783469360147904, + "grad_norm": 1.813378943063708, + "learning_rate": 3.479822344804796e-06, + "loss": 0.9243, + "step": 2859 + }, + { + "epoch": 0.25792487712494927, + "grad_norm": 1.8044190310320884, + "learning_rate": 3.479429280509582e-06, + "loss": 1.0409, + "step": 2860 + }, + { + "epoch": 0.25801506064841956, + "grad_norm": 1.9646071183459473, + "learning_rate": 3.4790360899813038e-06, + "loss": 1.0258, + "step": 2861 + }, + { + "epoch": 0.2581052441718898, + "grad_norm": 1.9055715510503628, + "learning_rate": 3.4786427732535115e-06, + "loss": 0.9178, + "step": 2862 + }, + { + "epoch": 0.2581954276953601, + "grad_norm": 1.7169009579646226, + "learning_rate": 3.478249330359764e-06, + "loss": 1.025, + "step": 2863 + }, + { + "epoch": 0.2582856112188303, + "grad_norm": 1.5968131753259462, + "learning_rate": 3.4778557613336333e-06, + "loss": 0.9174, + "step": 2864 + }, + { + "epoch": 0.2583757947423006, + "grad_norm": 1.5606396090152572, + "learning_rate": 3.4774620662087004e-06, + "loss": 0.9974, + "step": 2865 + }, + { + "epoch": 0.25846597826577083, + "grad_norm": 1.7755432399038298, + "learning_rate": 3.477068245018557e-06, + "loss": 0.9832, + "step": 2866 + }, + { + "epoch": 0.2585561617892411, + "grad_norm": 1.6674813266377735, + "learning_rate": 3.476674297796807e-06, + "loss": 0.99, + "step": 2867 + }, + { + "epoch": 0.25864634531271136, + "grad_norm": 1.6244497101039288, + "learning_rate": 3.4762802245770627e-06, + "loss": 1.0217, + "step": 2868 + }, + { + "epoch": 0.25873652883618165, + "grad_norm": 1.2543988077425643, + "learning_rate": 3.4758860253929497e-06, + "loss": 0.9324, + "step": 2869 + }, + { + "epoch": 0.2588267123596519, + "grad_norm": 1.486532711029698, + "learning_rate": 3.4754917002781038e-06, + "loss": 0.9998, + "step": 2870 + }, + { + "epoch": 0.25891689588312217, + "grad_norm": 1.7885199644549077, + "learning_rate": 3.475097249266169e-06, + "loss": 1.0256, + "step": 2871 + }, + { + "epoch": 0.2590070794065924, + "grad_norm": 1.6082972649333107, + "learning_rate": 3.4747026723908044e-06, + "loss": 0.9608, + "step": 2872 + }, + { + "epoch": 0.2590972629300627, + "grad_norm": 1.538912770690024, + "learning_rate": 3.474307969685676e-06, + "loss": 0.9636, + "step": 2873 + }, + { + "epoch": 0.2591874464535329, + "grad_norm": 1.6110667253643003, + "learning_rate": 3.473913141184462e-06, + "loss": 0.9269, + "step": 2874 + }, + { + "epoch": 0.2592776299770032, + "grad_norm": 1.4443592949165696, + "learning_rate": 3.4735181869208523e-06, + "loss": 0.856, + "step": 2875 + }, + { + "epoch": 0.25936781350047344, + "grad_norm": 2.132448158304831, + "learning_rate": 3.473123106928546e-06, + "loss": 0.8707, + "step": 2876 + }, + { + "epoch": 0.25945799702394373, + "grad_norm": 1.5239174111383003, + "learning_rate": 3.4727279012412533e-06, + "loss": 0.9411, + "step": 2877 + }, + { + "epoch": 0.25954818054741396, + "grad_norm": 0.6635952596016266, + "learning_rate": 3.4723325698926953e-06, + "loss": 0.777, + "step": 2878 + }, + { + "epoch": 0.25963836407088425, + "grad_norm": 1.4771715149519948, + "learning_rate": 3.4719371129166045e-06, + "loss": 0.9865, + "step": 2879 + }, + { + "epoch": 0.2597285475943545, + "grad_norm": 1.8357202928984457, + "learning_rate": 3.471541530346723e-06, + "loss": 0.9104, + "step": 2880 + }, + { + "epoch": 0.2598187311178248, + "grad_norm": 1.8837210967764506, + "learning_rate": 3.4711458222168037e-06, + "loss": 1.0439, + "step": 2881 + }, + { + "epoch": 0.259908914641295, + "grad_norm": 1.6083835399611754, + "learning_rate": 3.4707499885606114e-06, + "loss": 1.0251, + "step": 2882 + }, + { + "epoch": 0.2599990981647653, + "grad_norm": 1.5566917803506353, + "learning_rate": 3.4703540294119204e-06, + "loss": 1.0192, + "step": 2883 + }, + { + "epoch": 0.2600892816882356, + "grad_norm": 1.6859128340365395, + "learning_rate": 3.4699579448045163e-06, + "loss": 1.035, + "step": 2884 + }, + { + "epoch": 0.2601794652117058, + "grad_norm": 1.45723225943067, + "learning_rate": 3.4695617347721947e-06, + "loss": 0.8727, + "step": 2885 + }, + { + "epoch": 0.2602696487351761, + "grad_norm": 1.6803631510380295, + "learning_rate": 3.469165399348763e-06, + "loss": 0.9468, + "step": 2886 + }, + { + "epoch": 0.26035983225864634, + "grad_norm": 1.8459456995854107, + "learning_rate": 3.4687689385680384e-06, + "loss": 0.9638, + "step": 2887 + }, + { + "epoch": 0.26045001578211663, + "grad_norm": 1.419677377505598, + "learning_rate": 3.4683723524638494e-06, + "loss": 0.9598, + "step": 2888 + }, + { + "epoch": 0.26054019930558686, + "grad_norm": 1.5524664113807989, + "learning_rate": 3.4679756410700354e-06, + "loss": 1.0497, + "step": 2889 + }, + { + "epoch": 0.26063038282905715, + "grad_norm": 2.028768578971363, + "learning_rate": 3.4675788044204445e-06, + "loss": 0.9825, + "step": 2890 + }, + { + "epoch": 0.2607205663525274, + "grad_norm": 1.58288490905528, + "learning_rate": 3.467181842548938e-06, + "loss": 0.9167, + "step": 2891 + }, + { + "epoch": 0.26081074987599767, + "grad_norm": 1.46073642917788, + "learning_rate": 3.466784755489387e-06, + "loss": 0.9792, + "step": 2892 + }, + { + "epoch": 0.2609009333994679, + "grad_norm": 1.8125984559777477, + "learning_rate": 3.4663875432756726e-06, + "loss": 1.0978, + "step": 2893 + }, + { + "epoch": 0.2609911169229382, + "grad_norm": 1.7181769369357867, + "learning_rate": 3.465990205941687e-06, + "loss": 1.1527, + "step": 2894 + }, + { + "epoch": 0.2610813004464084, + "grad_norm": 1.6252854903400398, + "learning_rate": 3.465592743521335e-06, + "loss": 0.9697, + "step": 2895 + }, + { + "epoch": 0.2611714839698787, + "grad_norm": 1.6622609877330068, + "learning_rate": 3.465195156048528e-06, + "loss": 1.0348, + "step": 2896 + }, + { + "epoch": 0.26126166749334895, + "grad_norm": 0.6300015359814403, + "learning_rate": 3.464797443557191e-06, + "loss": 0.775, + "step": 2897 + }, + { + "epoch": 0.26135185101681924, + "grad_norm": 1.5222179519905903, + "learning_rate": 3.46439960608126e-06, + "loss": 0.9568, + "step": 2898 + }, + { + "epoch": 0.26144203454028947, + "grad_norm": 1.6368936158498466, + "learning_rate": 3.4640016436546797e-06, + "loss": 0.921, + "step": 2899 + }, + { + "epoch": 0.26153221806375976, + "grad_norm": 1.6998200152991554, + "learning_rate": 3.4636035563114065e-06, + "loss": 0.957, + "step": 2900 + }, + { + "epoch": 0.26162240158723, + "grad_norm": 1.4151640758485446, + "learning_rate": 3.4632053440854085e-06, + "loss": 0.9119, + "step": 2901 + }, + { + "epoch": 0.2617125851107003, + "grad_norm": 1.8074993999897027, + "learning_rate": 3.462807007010662e-06, + "loss": 1.0384, + "step": 2902 + }, + { + "epoch": 0.2618027686341705, + "grad_norm": 1.620155081095988, + "learning_rate": 3.462408545121155e-06, + "loss": 1.0145, + "step": 2903 + }, + { + "epoch": 0.2618929521576408, + "grad_norm": 1.7145685135798452, + "learning_rate": 3.4620099584508883e-06, + "loss": 1.0369, + "step": 2904 + }, + { + "epoch": 0.26198313568111103, + "grad_norm": 1.742389684329397, + "learning_rate": 3.46161124703387e-06, + "loss": 0.9409, + "step": 2905 + }, + { + "epoch": 0.2620733192045813, + "grad_norm": 1.6832107143624955, + "learning_rate": 3.461212410904122e-06, + "loss": 0.9784, + "step": 2906 + }, + { + "epoch": 0.2621635027280516, + "grad_norm": 1.8884698233078312, + "learning_rate": 3.4608134500956726e-06, + "loss": 1.0348, + "step": 2907 + }, + { + "epoch": 0.26225368625152184, + "grad_norm": 1.391644479190203, + "learning_rate": 3.4604143646425655e-06, + "loss": 0.9701, + "step": 2908 + }, + { + "epoch": 0.26234386977499213, + "grad_norm": 1.297312237672595, + "learning_rate": 3.460015154578852e-06, + "loss": 1.0495, + "step": 2909 + }, + { + "epoch": 0.26243405329846237, + "grad_norm": 1.5208061773235917, + "learning_rate": 3.459615819938595e-06, + "loss": 0.9582, + "step": 2910 + }, + { + "epoch": 0.26252423682193265, + "grad_norm": 4.613915434553069, + "learning_rate": 3.4592163607558684e-06, + "loss": 0.9424, + "step": 2911 + }, + { + "epoch": 0.2626144203454029, + "grad_norm": 1.4529069306193525, + "learning_rate": 3.4588167770647553e-06, + "loss": 0.988, + "step": 2912 + }, + { + "epoch": 0.2627046038688732, + "grad_norm": 1.7230904306646024, + "learning_rate": 3.458417068899351e-06, + "loss": 1.0765, + "step": 2913 + }, + { + "epoch": 0.2627947873923434, + "grad_norm": 1.5496049162115633, + "learning_rate": 3.4580172362937612e-06, + "loss": 0.9542, + "step": 2914 + }, + { + "epoch": 0.2628849709158137, + "grad_norm": 1.4540708191480147, + "learning_rate": 3.457617279282101e-06, + "loss": 0.9499, + "step": 2915 + }, + { + "epoch": 0.26297515443928393, + "grad_norm": 1.8279199729781128, + "learning_rate": 3.4572171978984975e-06, + "loss": 0.9617, + "step": 2916 + }, + { + "epoch": 0.2630653379627542, + "grad_norm": 1.4837357600234988, + "learning_rate": 3.456816992177088e-06, + "loss": 0.9839, + "step": 2917 + }, + { + "epoch": 0.26315552148622445, + "grad_norm": 1.440184698445726, + "learning_rate": 3.4564166621520193e-06, + "loss": 0.952, + "step": 2918 + }, + { + "epoch": 0.26324570500969474, + "grad_norm": 1.4715324991456593, + "learning_rate": 3.4560162078574507e-06, + "loss": 0.9746, + "step": 2919 + }, + { + "epoch": 0.263335888533165, + "grad_norm": 1.5059603844000427, + "learning_rate": 3.455615629327551e-06, + "loss": 0.9846, + "step": 2920 + }, + { + "epoch": 0.26342607205663526, + "grad_norm": 1.683869022957133, + "learning_rate": 3.4552149265964994e-06, + "loss": 0.9571, + "step": 2921 + }, + { + "epoch": 0.2635162555801055, + "grad_norm": 1.4010446210449532, + "learning_rate": 3.4548140996984866e-06, + "loss": 0.9316, + "step": 2922 + }, + { + "epoch": 0.2636064391035758, + "grad_norm": 1.4832360967990417, + "learning_rate": 3.4544131486677124e-06, + "loss": 1.0035, + "step": 2923 + }, + { + "epoch": 0.263696622627046, + "grad_norm": 1.422082571451762, + "learning_rate": 3.454012073538389e-06, + "loss": 0.9197, + "step": 2924 + }, + { + "epoch": 0.2637868061505163, + "grad_norm": 1.4783582313846446, + "learning_rate": 3.453610874344738e-06, + "loss": 0.9926, + "step": 2925 + }, + { + "epoch": 0.26387698967398654, + "grad_norm": 1.6182163669786762, + "learning_rate": 3.453209551120993e-06, + "loss": 1.0211, + "step": 2926 + }, + { + "epoch": 0.26396717319745683, + "grad_norm": 1.6441409266229825, + "learning_rate": 3.452808103901395e-06, + "loss": 0.958, + "step": 2927 + }, + { + "epoch": 0.26405735672092706, + "grad_norm": 1.6956039389157558, + "learning_rate": 3.4524065327202e-06, + "loss": 1.06, + "step": 2928 + }, + { + "epoch": 0.26414754024439735, + "grad_norm": 0.7500116426835781, + "learning_rate": 3.4520048376116702e-06, + "loss": 0.7275, + "step": 2929 + }, + { + "epoch": 0.26423772376786764, + "grad_norm": 1.4413775516123852, + "learning_rate": 3.4516030186100817e-06, + "loss": 1.0199, + "step": 2930 + }, + { + "epoch": 0.26432790729133787, + "grad_norm": 1.423528276013249, + "learning_rate": 3.4512010757497197e-06, + "loss": 0.8982, + "step": 2931 + }, + { + "epoch": 0.26441809081480816, + "grad_norm": 1.5549633605663695, + "learning_rate": 3.4507990090648804e-06, + "loss": 0.8981, + "step": 2932 + }, + { + "epoch": 0.2645082743382784, + "grad_norm": 1.8156778823583635, + "learning_rate": 3.4503968185898696e-06, + "loss": 1.0095, + "step": 2933 + }, + { + "epoch": 0.2645984578617487, + "grad_norm": 1.6662789927061474, + "learning_rate": 3.4499945043590047e-06, + "loss": 0.9889, + "step": 2934 + }, + { + "epoch": 0.2646886413852189, + "grad_norm": 1.6087966175927753, + "learning_rate": 3.4495920664066137e-06, + "loss": 1.0003, + "step": 2935 + }, + { + "epoch": 0.2647788249086892, + "grad_norm": 4.355451014187544, + "learning_rate": 3.449189504767035e-06, + "loss": 0.9745, + "step": 2936 + }, + { + "epoch": 0.26486900843215944, + "grad_norm": 1.8541085809188835, + "learning_rate": 3.4487868194746163e-06, + "loss": 1.0103, + "step": 2937 + }, + { + "epoch": 0.2649591919556297, + "grad_norm": 1.5902777601942628, + "learning_rate": 3.4483840105637175e-06, + "loss": 1.0209, + "step": 2938 + }, + { + "epoch": 0.26504937547909996, + "grad_norm": 2.5005042521244745, + "learning_rate": 3.4479810780687097e-06, + "loss": 0.9671, + "step": 2939 + }, + { + "epoch": 0.26513955900257025, + "grad_norm": 1.4335445218897926, + "learning_rate": 3.4475780220239714e-06, + "loss": 0.9721, + "step": 2940 + }, + { + "epoch": 0.2652297425260405, + "grad_norm": 1.5273113588766816, + "learning_rate": 3.4471748424638948e-06, + "loss": 0.9723, + "step": 2941 + }, + { + "epoch": 0.26531992604951077, + "grad_norm": 0.6598342697181555, + "learning_rate": 3.4467715394228803e-06, + "loss": 0.7947, + "step": 2942 + }, + { + "epoch": 0.265410109572981, + "grad_norm": 1.3330088856451725, + "learning_rate": 3.4463681129353413e-06, + "loss": 0.9865, + "step": 2943 + }, + { + "epoch": 0.2655002930964513, + "grad_norm": 1.5468148884508777, + "learning_rate": 3.4459645630357e-06, + "loss": 0.9751, + "step": 2944 + }, + { + "epoch": 0.2655904766199215, + "grad_norm": 1.7724805089957683, + "learning_rate": 3.4455608897583884e-06, + "loss": 1.0171, + "step": 2945 + }, + { + "epoch": 0.2656806601433918, + "grad_norm": 1.5702660397161494, + "learning_rate": 3.4451570931378514e-06, + "loss": 0.9843, + "step": 2946 + }, + { + "epoch": 0.26577084366686204, + "grad_norm": 1.3211562490991333, + "learning_rate": 3.444753173208543e-06, + "loss": 0.988, + "step": 2947 + }, + { + "epoch": 0.26586102719033233, + "grad_norm": 1.8248294567740913, + "learning_rate": 3.444349130004927e-06, + "loss": 0.9784, + "step": 2948 + }, + { + "epoch": 0.26595121071380257, + "grad_norm": 1.6616710259633225, + "learning_rate": 3.4439449635614794e-06, + "loss": 0.9036, + "step": 2949 + }, + { + "epoch": 0.26604139423727285, + "grad_norm": 1.4106070941672562, + "learning_rate": 3.4435406739126854e-06, + "loss": 1.0099, + "step": 2950 + }, + { + "epoch": 0.2661315777607431, + "grad_norm": 1.9984577908652754, + "learning_rate": 3.443136261093042e-06, + "loss": 0.7957, + "step": 2951 + }, + { + "epoch": 0.2662217612842134, + "grad_norm": 1.846612228057081, + "learning_rate": 3.4427317251370553e-06, + "loss": 0.9264, + "step": 2952 + }, + { + "epoch": 0.2663119448076836, + "grad_norm": 1.8645801295531754, + "learning_rate": 3.4423270660792422e-06, + "loss": 0.8967, + "step": 2953 + }, + { + "epoch": 0.2664021283311539, + "grad_norm": 1.5661614814215907, + "learning_rate": 3.4419222839541314e-06, + "loss": 0.8737, + "step": 2954 + }, + { + "epoch": 0.2664923118546242, + "grad_norm": 1.36835800201321, + "learning_rate": 3.4415173787962607e-06, + "loss": 0.9542, + "step": 2955 + }, + { + "epoch": 0.2665824953780944, + "grad_norm": 1.8803984175217134, + "learning_rate": 3.4411123506401783e-06, + "loss": 0.8458, + "step": 2956 + }, + { + "epoch": 0.2666726789015647, + "grad_norm": 1.7892769764104097, + "learning_rate": 3.440707199520444e-06, + "loss": 0.8353, + "step": 2957 + }, + { + "epoch": 0.26676286242503494, + "grad_norm": 1.4852940173918345, + "learning_rate": 3.440301925471628e-06, + "loss": 0.9449, + "step": 2958 + }, + { + "epoch": 0.26685304594850523, + "grad_norm": 1.5228596691871328, + "learning_rate": 3.43989652852831e-06, + "loss": 1.0154, + "step": 2959 + }, + { + "epoch": 0.26694322947197546, + "grad_norm": 1.5738226531864579, + "learning_rate": 3.4394910087250804e-06, + "loss": 1.0455, + "step": 2960 + }, + { + "epoch": 0.26703341299544575, + "grad_norm": 1.7015564721371148, + "learning_rate": 3.4390853660965405e-06, + "loss": 0.9949, + "step": 2961 + }, + { + "epoch": 0.267123596518916, + "grad_norm": 1.651793981406243, + "learning_rate": 3.438679600677302e-06, + "loss": 1.0564, + "step": 2962 + }, + { + "epoch": 0.2672137800423863, + "grad_norm": 1.3724026857768423, + "learning_rate": 3.4382737125019874e-06, + "loss": 0.9373, + "step": 2963 + }, + { + "epoch": 0.2673039635658565, + "grad_norm": 1.486348015430314, + "learning_rate": 3.4378677016052294e-06, + "loss": 0.9488, + "step": 2964 + }, + { + "epoch": 0.2673941470893268, + "grad_norm": 1.6697022767709815, + "learning_rate": 3.43746156802167e-06, + "loss": 0.9979, + "step": 2965 + }, + { + "epoch": 0.267484330612797, + "grad_norm": 1.5540342252342652, + "learning_rate": 3.4370553117859643e-06, + "loss": 0.9166, + "step": 2966 + }, + { + "epoch": 0.2675745141362673, + "grad_norm": 1.8227438999231438, + "learning_rate": 3.4366489329327754e-06, + "loss": 1.0091, + "step": 2967 + }, + { + "epoch": 0.26766469765973755, + "grad_norm": 1.7794401779001912, + "learning_rate": 3.4362424314967777e-06, + "loss": 1.005, + "step": 2968 + }, + { + "epoch": 0.26775488118320784, + "grad_norm": 1.7103312912084154, + "learning_rate": 3.4358358075126567e-06, + "loss": 1.0034, + "step": 2969 + }, + { + "epoch": 0.26784506470667807, + "grad_norm": 1.6018804467175982, + "learning_rate": 3.4354290610151077e-06, + "loss": 0.9988, + "step": 2970 + }, + { + "epoch": 0.26793524823014836, + "grad_norm": 1.49123691818498, + "learning_rate": 3.4350221920388354e-06, + "loss": 0.9894, + "step": 2971 + }, + { + "epoch": 0.2680254317536186, + "grad_norm": 1.4938341412329086, + "learning_rate": 3.4346152006185574e-06, + "loss": 1.0784, + "step": 2972 + }, + { + "epoch": 0.2681156152770889, + "grad_norm": 1.6727709642224566, + "learning_rate": 3.4342080867890006e-06, + "loss": 1.0813, + "step": 2973 + }, + { + "epoch": 0.2682057988005591, + "grad_norm": 1.8531476410227277, + "learning_rate": 3.4338008505849016e-06, + "loss": 0.9052, + "step": 2974 + }, + { + "epoch": 0.2682959823240294, + "grad_norm": 1.5707378665371046, + "learning_rate": 3.433393492041008e-06, + "loss": 0.9895, + "step": 2975 + }, + { + "epoch": 0.26838616584749964, + "grad_norm": 1.3625861201935652, + "learning_rate": 3.432986011192078e-06, + "loss": 1.0319, + "step": 2976 + }, + { + "epoch": 0.2684763493709699, + "grad_norm": 1.7508128866786867, + "learning_rate": 3.4325784080728796e-06, + "loss": 1.0485, + "step": 2977 + }, + { + "epoch": 0.2685665328944402, + "grad_norm": 1.6844100920183487, + "learning_rate": 3.4321706827181926e-06, + "loss": 1.0073, + "step": 2978 + }, + { + "epoch": 0.26865671641791045, + "grad_norm": 1.2830542444638713, + "learning_rate": 3.4317628351628064e-06, + "loss": 0.9336, + "step": 2979 + }, + { + "epoch": 0.26874689994138073, + "grad_norm": 1.4435683165009952, + "learning_rate": 3.43135486544152e-06, + "loss": 0.9625, + "step": 2980 + }, + { + "epoch": 0.26883708346485097, + "grad_norm": 1.4153997515743006, + "learning_rate": 3.4309467735891442e-06, + "loss": 1.0221, + "step": 2981 + }, + { + "epoch": 0.26892726698832126, + "grad_norm": 1.6878295859150259, + "learning_rate": 3.4305385596405e-06, + "loss": 1.0679, + "step": 2982 + }, + { + "epoch": 0.2690174505117915, + "grad_norm": 1.3945552278909574, + "learning_rate": 3.4301302236304174e-06, + "loss": 0.9432, + "step": 2983 + }, + { + "epoch": 0.2691076340352618, + "grad_norm": 0.7001857145069269, + "learning_rate": 3.429721765593739e-06, + "loss": 0.7502, + "step": 2984 + }, + { + "epoch": 0.269197817558732, + "grad_norm": 2.540764059505132, + "learning_rate": 3.4293131855653155e-06, + "loss": 1.0047, + "step": 2985 + }, + { + "epoch": 0.2692880010822023, + "grad_norm": 1.5011909842980975, + "learning_rate": 3.4289044835800102e-06, + "loss": 0.9142, + "step": 2986 + }, + { + "epoch": 0.26937818460567253, + "grad_norm": 1.7422080958222412, + "learning_rate": 3.4284956596726953e-06, + "loss": 0.9789, + "step": 2987 + }, + { + "epoch": 0.2694683681291428, + "grad_norm": 1.447731344680863, + "learning_rate": 3.4280867138782544e-06, + "loss": 0.9753, + "step": 2988 + }, + { + "epoch": 0.26955855165261305, + "grad_norm": 2.043261534387239, + "learning_rate": 3.4276776462315803e-06, + "loss": 0.983, + "step": 2989 + }, + { + "epoch": 0.26964873517608334, + "grad_norm": 1.4026844598976604, + "learning_rate": 3.427268456767578e-06, + "loss": 0.9819, + "step": 2990 + }, + { + "epoch": 0.2697389186995536, + "grad_norm": 1.7041208697867571, + "learning_rate": 3.42685914552116e-06, + "loss": 1.0183, + "step": 2991 + }, + { + "epoch": 0.26982910222302386, + "grad_norm": 1.4242241705008187, + "learning_rate": 3.426449712527253e-06, + "loss": 0.9525, + "step": 2992 + }, + { + "epoch": 0.2699192857464941, + "grad_norm": 1.3978830442445294, + "learning_rate": 3.4260401578207904e-06, + "loss": 0.8941, + "step": 2993 + }, + { + "epoch": 0.2700094692699644, + "grad_norm": 1.9370215040440069, + "learning_rate": 3.4256304814367185e-06, + "loss": 0.9497, + "step": 2994 + }, + { + "epoch": 0.2700996527934346, + "grad_norm": 1.892557401907797, + "learning_rate": 3.4252206834099936e-06, + "loss": 0.9429, + "step": 2995 + }, + { + "epoch": 0.2701898363169049, + "grad_norm": 1.5497219205411852, + "learning_rate": 3.424810763775581e-06, + "loss": 0.9775, + "step": 2996 + }, + { + "epoch": 0.27028001984037514, + "grad_norm": 1.3382999879824635, + "learning_rate": 3.4244007225684587e-06, + "loss": 0.9189, + "step": 2997 + }, + { + "epoch": 0.27037020336384543, + "grad_norm": 1.661680208735762, + "learning_rate": 3.4239905598236115e-06, + "loss": 0.9877, + "step": 2998 + }, + { + "epoch": 0.27046038688731566, + "grad_norm": 1.4690426474055271, + "learning_rate": 3.4235802755760386e-06, + "loss": 1.0094, + "step": 2999 + }, + { + "epoch": 0.27055057041078595, + "grad_norm": 0.8958509613490984, + "learning_rate": 3.4231698698607464e-06, + "loss": 0.8959, + "step": 3000 + }, + { + "epoch": 0.2706407539342562, + "grad_norm": 1.6872074438740308, + "learning_rate": 3.4227593427127543e-06, + "loss": 1.0395, + "step": 3001 + }, + { + "epoch": 0.2707309374577265, + "grad_norm": 1.7675602527368208, + "learning_rate": 3.42234869416709e-06, + "loss": 0.9506, + "step": 3002 + }, + { + "epoch": 0.27082112098119676, + "grad_norm": 1.7568134096491983, + "learning_rate": 3.421937924258792e-06, + "loss": 0.9873, + "step": 3003 + }, + { + "epoch": 0.270911304504667, + "grad_norm": 1.5025178123091094, + "learning_rate": 3.4215270330229096e-06, + "loss": 1.0249, + "step": 3004 + }, + { + "epoch": 0.2710014880281373, + "grad_norm": 1.5273961987477966, + "learning_rate": 3.421116020494503e-06, + "loss": 0.9415, + "step": 3005 + }, + { + "epoch": 0.2710916715516075, + "grad_norm": 1.3012024252015528, + "learning_rate": 3.420704886708642e-06, + "loss": 0.9256, + "step": 3006 + }, + { + "epoch": 0.2711818550750778, + "grad_norm": 1.3319407681248783, + "learning_rate": 3.4202936317004056e-06, + "loss": 0.9305, + "step": 3007 + }, + { + "epoch": 0.27127203859854804, + "grad_norm": 1.646333916103221, + "learning_rate": 3.4198822555048856e-06, + "loss": 0.9948, + "step": 3008 + }, + { + "epoch": 0.2713622221220183, + "grad_norm": 3.6331041444592946, + "learning_rate": 3.419470758157182e-06, + "loss": 1.0633, + "step": 3009 + }, + { + "epoch": 0.27145240564548856, + "grad_norm": 1.6447548612310257, + "learning_rate": 3.4190591396924068e-06, + "loss": 0.9763, + "step": 3010 + }, + { + "epoch": 0.27154258916895885, + "grad_norm": 0.7673110303076259, + "learning_rate": 3.418647400145681e-06, + "loss": 0.8176, + "step": 3011 + }, + { + "epoch": 0.2716327726924291, + "grad_norm": 0.7236505627224775, + "learning_rate": 3.4182355395521367e-06, + "loss": 0.7778, + "step": 3012 + }, + { + "epoch": 0.27172295621589937, + "grad_norm": 1.4788960580933066, + "learning_rate": 3.417823557946916e-06, + "loss": 0.9778, + "step": 3013 + }, + { + "epoch": 0.2718131397393696, + "grad_norm": 1.392726767205659, + "learning_rate": 3.417411455365172e-06, + "loss": 0.8715, + "step": 3014 + }, + { + "epoch": 0.2719033232628399, + "grad_norm": 1.639807724613774, + "learning_rate": 3.416999231842066e-06, + "loss": 0.8918, + "step": 3015 + }, + { + "epoch": 0.2719935067863101, + "grad_norm": 1.8772419559409543, + "learning_rate": 3.416586887412773e-06, + "loss": 1.0359, + "step": 3016 + }, + { + "epoch": 0.2720836903097804, + "grad_norm": 3.9465148906477587, + "learning_rate": 3.416174422112476e-06, + "loss": 0.9412, + "step": 3017 + }, + { + "epoch": 0.27217387383325065, + "grad_norm": 0.6994590133373121, + "learning_rate": 3.4157618359763687e-06, + "loss": 0.7417, + "step": 3018 + }, + { + "epoch": 0.27226405735672093, + "grad_norm": 1.597882392379777, + "learning_rate": 3.4153491290396542e-06, + "loss": 0.9012, + "step": 3019 + }, + { + "epoch": 0.27235424088019117, + "grad_norm": 1.5882591204736398, + "learning_rate": 3.4149363013375485e-06, + "loss": 0.9917, + "step": 3020 + }, + { + "epoch": 0.27244442440366146, + "grad_norm": 0.7823924294776365, + "learning_rate": 3.414523352905276e-06, + "loss": 0.7576, + "step": 3021 + }, + { + "epoch": 0.2725346079271317, + "grad_norm": 1.5450849868954575, + "learning_rate": 3.414110283778071e-06, + "loss": 0.9554, + "step": 3022 + }, + { + "epoch": 0.272624791450602, + "grad_norm": 1.9767352243577785, + "learning_rate": 3.4136970939911797e-06, + "loss": 0.9298, + "step": 3023 + }, + { + "epoch": 0.2727149749740722, + "grad_norm": 1.5768251258924924, + "learning_rate": 3.413283783579857e-06, + "loss": 0.9469, + "step": 3024 + }, + { + "epoch": 0.2728051584975425, + "grad_norm": 1.8678206004923972, + "learning_rate": 3.412870352579369e-06, + "loss": 0.9176, + "step": 3025 + }, + { + "epoch": 0.2728953420210128, + "grad_norm": 1.0582772658107076, + "learning_rate": 3.4124568010249915e-06, + "loss": 0.7987, + "step": 3026 + }, + { + "epoch": 0.272985525544483, + "grad_norm": 2.444498529943663, + "learning_rate": 3.4120431289520124e-06, + "loss": 0.9644, + "step": 3027 + }, + { + "epoch": 0.2730757090679533, + "grad_norm": 1.5663493995148317, + "learning_rate": 3.4116293363957276e-06, + "loss": 0.9881, + "step": 3028 + }, + { + "epoch": 0.27316589259142354, + "grad_norm": 1.7673810480172014, + "learning_rate": 3.4112154233914438e-06, + "loss": 0.987, + "step": 3029 + }, + { + "epoch": 0.27325607611489383, + "grad_norm": 1.7416800542557211, + "learning_rate": 3.410801389974479e-06, + "loss": 0.8842, + "step": 3030 + }, + { + "epoch": 0.27334625963836406, + "grad_norm": 1.8079999768797275, + "learning_rate": 3.410387236180161e-06, + "loss": 1.0132, + "step": 3031 + }, + { + "epoch": 0.27343644316183435, + "grad_norm": 1.851611752398235, + "learning_rate": 3.409972962043826e-06, + "loss": 0.8537, + "step": 3032 + }, + { + "epoch": 0.2735266266853046, + "grad_norm": 1.7082968064411683, + "learning_rate": 3.4095585676008234e-06, + "loss": 1.0324, + "step": 3033 + }, + { + "epoch": 0.2736168102087749, + "grad_norm": 1.3323113826498818, + "learning_rate": 3.4091440528865125e-06, + "loss": 0.9692, + "step": 3034 + }, + { + "epoch": 0.2737069937322451, + "grad_norm": 1.7771397735589103, + "learning_rate": 3.4087294179362606e-06, + "loss": 0.9779, + "step": 3035 + }, + { + "epoch": 0.2737971772557154, + "grad_norm": 1.5660925191878052, + "learning_rate": 3.4083146627854474e-06, + "loss": 1.0264, + "step": 3036 + }, + { + "epoch": 0.27388736077918563, + "grad_norm": 1.9035762734148285, + "learning_rate": 3.4078997874694614e-06, + "loss": 0.9025, + "step": 3037 + }, + { + "epoch": 0.2739775443026559, + "grad_norm": 1.7113615581744384, + "learning_rate": 3.407484792023703e-06, + "loss": 0.8717, + "step": 3038 + }, + { + "epoch": 0.27406772782612615, + "grad_norm": 1.8180445348322865, + "learning_rate": 3.407069676483581e-06, + "loss": 0.9474, + "step": 3039 + }, + { + "epoch": 0.27415791134959644, + "grad_norm": 1.4151288643110214, + "learning_rate": 3.406654440884516e-06, + "loss": 1.0154, + "step": 3040 + }, + { + "epoch": 0.2742480948730667, + "grad_norm": 1.7050233113325752, + "learning_rate": 3.4062390852619372e-06, + "loss": 1.0078, + "step": 3041 + }, + { + "epoch": 0.27433827839653696, + "grad_norm": 1.8646708491013317, + "learning_rate": 3.4058236096512867e-06, + "loss": 0.9749, + "step": 3042 + }, + { + "epoch": 0.2744284619200072, + "grad_norm": 2.8727901092456762, + "learning_rate": 3.405408014088013e-06, + "loss": 0.8782, + "step": 3043 + }, + { + "epoch": 0.2745186454434775, + "grad_norm": 1.9013254937928323, + "learning_rate": 3.404992298607579e-06, + "loss": 0.864, + "step": 3044 + }, + { + "epoch": 0.2746088289669477, + "grad_norm": 2.2082517236948815, + "learning_rate": 3.4045764632454547e-06, + "loss": 0.9574, + "step": 3045 + }, + { + "epoch": 0.274699012490418, + "grad_norm": 1.5968293006165741, + "learning_rate": 3.4041605080371223e-06, + "loss": 0.9376, + "step": 3046 + }, + { + "epoch": 0.27478919601388824, + "grad_norm": 1.6636233041215662, + "learning_rate": 3.4037444330180726e-06, + "loss": 0.9861, + "step": 3047 + }, + { + "epoch": 0.2748793795373585, + "grad_norm": 1.8833802819378278, + "learning_rate": 3.403328238223808e-06, + "loss": 0.9929, + "step": 3048 + }, + { + "epoch": 0.27496956306082876, + "grad_norm": 2.0972402149408254, + "learning_rate": 3.4029119236898395e-06, + "loss": 1.0106, + "step": 3049 + }, + { + "epoch": 0.27505974658429905, + "grad_norm": 1.0385018463816273, + "learning_rate": 3.4024954894516906e-06, + "loss": 0.8102, + "step": 3050 + }, + { + "epoch": 0.27514993010776934, + "grad_norm": 1.7092575116552515, + "learning_rate": 3.4020789355448933e-06, + "loss": 0.9782, + "step": 3051 + }, + { + "epoch": 0.27524011363123957, + "grad_norm": 1.4435005996433303, + "learning_rate": 3.40166226200499e-06, + "loss": 1.0517, + "step": 3052 + }, + { + "epoch": 0.27533029715470986, + "grad_norm": 1.7721104298964216, + "learning_rate": 3.401245468867534e-06, + "loss": 0.9659, + "step": 3053 + }, + { + "epoch": 0.2754204806781801, + "grad_norm": 0.8795078846316166, + "learning_rate": 3.400828556168088e-06, + "loss": 0.835, + "step": 3054 + }, + { + "epoch": 0.2755106642016504, + "grad_norm": 1.6328918583763388, + "learning_rate": 3.4004115239422255e-06, + "loss": 0.9943, + "step": 3055 + }, + { + "epoch": 0.2756008477251206, + "grad_norm": 1.749795288646589, + "learning_rate": 3.3999943722255305e-06, + "loss": 0.9658, + "step": 3056 + }, + { + "epoch": 0.2756910312485909, + "grad_norm": 1.5206260365687614, + "learning_rate": 3.3995771010535955e-06, + "loss": 0.8753, + "step": 3057 + }, + { + "epoch": 0.27578121477206113, + "grad_norm": 0.9163151883592661, + "learning_rate": 3.3991597104620253e-06, + "loss": 0.7416, + "step": 3058 + }, + { + "epoch": 0.2758713982955314, + "grad_norm": 1.4662278617104691, + "learning_rate": 3.398742200486434e-06, + "loss": 0.9574, + "step": 3059 + }, + { + "epoch": 0.27596158181900166, + "grad_norm": 1.6144871672837708, + "learning_rate": 3.3983245711624453e-06, + "loss": 0.9985, + "step": 3060 + }, + { + "epoch": 0.27605176534247194, + "grad_norm": 1.0275916899078168, + "learning_rate": 3.3979068225256946e-06, + "loss": 0.8572, + "step": 3061 + }, + { + "epoch": 0.2761419488659422, + "grad_norm": 1.630903816759739, + "learning_rate": 3.3974889546118246e-06, + "loss": 0.9413, + "step": 3062 + }, + { + "epoch": 0.27623213238941247, + "grad_norm": 1.4744106537767756, + "learning_rate": 3.3970709674564918e-06, + "loss": 0.9214, + "step": 3063 + }, + { + "epoch": 0.2763223159128827, + "grad_norm": 1.6151478016541874, + "learning_rate": 3.3966528610953607e-06, + "loss": 0.9732, + "step": 3064 + }, + { + "epoch": 0.276412499436353, + "grad_norm": 1.6258581536511933, + "learning_rate": 3.3962346355641067e-06, + "loss": 0.9835, + "step": 3065 + }, + { + "epoch": 0.2765026829598232, + "grad_norm": 1.514930015181301, + "learning_rate": 3.3958162908984146e-06, + "loss": 1.011, + "step": 3066 + }, + { + "epoch": 0.2765928664832935, + "grad_norm": 1.4901561391549811, + "learning_rate": 3.39539782713398e-06, + "loss": 0.9709, + "step": 3067 + }, + { + "epoch": 0.27668305000676374, + "grad_norm": 2.231610274947328, + "learning_rate": 3.394979244306509e-06, + "loss": 1.0012, + "step": 3068 + }, + { + "epoch": 0.27677323353023403, + "grad_norm": 1.5189627635728835, + "learning_rate": 3.3945605424517166e-06, + "loss": 0.9852, + "step": 3069 + }, + { + "epoch": 0.27686341705370426, + "grad_norm": 1.7758064076915216, + "learning_rate": 3.3941417216053294e-06, + "loss": 0.973, + "step": 3070 + }, + { + "epoch": 0.27695360057717455, + "grad_norm": 1.7232800519861078, + "learning_rate": 3.3937227818030835e-06, + "loss": 0.9369, + "step": 3071 + }, + { + "epoch": 0.2770437841006448, + "grad_norm": 1.7749065079043311, + "learning_rate": 3.393303723080725e-06, + "loss": 0.935, + "step": 3072 + }, + { + "epoch": 0.2771339676241151, + "grad_norm": 1.9767612161372423, + "learning_rate": 3.3928845454740097e-06, + "loss": 0.9853, + "step": 3073 + }, + { + "epoch": 0.27722415114758536, + "grad_norm": 1.7181713864256132, + "learning_rate": 3.392465249018705e-06, + "loss": 0.9634, + "step": 3074 + }, + { + "epoch": 0.2773143346710556, + "grad_norm": 2.031206101163255, + "learning_rate": 3.3920458337505872e-06, + "loss": 0.9707, + "step": 3075 + }, + { + "epoch": 0.2774045181945259, + "grad_norm": 5.17915389820754, + "learning_rate": 3.391626299705443e-06, + "loss": 0.8154, + "step": 3076 + }, + { + "epoch": 0.2774947017179961, + "grad_norm": 2.0190690063126118, + "learning_rate": 3.39120664691907e-06, + "loss": 0.9844, + "step": 3077 + }, + { + "epoch": 0.2775848852414664, + "grad_norm": 1.4992889467377315, + "learning_rate": 3.390786875427275e-06, + "loss": 0.9511, + "step": 3078 + }, + { + "epoch": 0.27767506876493664, + "grad_norm": 1.5908358990565465, + "learning_rate": 3.390366985265875e-06, + "loss": 0.9697, + "step": 3079 + }, + { + "epoch": 0.2777652522884069, + "grad_norm": 1.742860672565254, + "learning_rate": 3.389946976470697e-06, + "loss": 0.9279, + "step": 3080 + }, + { + "epoch": 0.27785543581187716, + "grad_norm": 1.4000942709382245, + "learning_rate": 3.3895268490775787e-06, + "loss": 1.0639, + "step": 3081 + }, + { + "epoch": 0.27794561933534745, + "grad_norm": 1.7168415918699975, + "learning_rate": 3.3891066031223685e-06, + "loss": 0.9294, + "step": 3082 + }, + { + "epoch": 0.2780358028588177, + "grad_norm": 1.8511305035461594, + "learning_rate": 3.3886862386409237e-06, + "loss": 1.0039, + "step": 3083 + }, + { + "epoch": 0.27812598638228797, + "grad_norm": 1.8801406326856342, + "learning_rate": 3.388265755669111e-06, + "loss": 0.9356, + "step": 3084 + }, + { + "epoch": 0.2782161699057582, + "grad_norm": 1.6278530164580365, + "learning_rate": 3.3878451542428093e-06, + "loss": 1.0374, + "step": 3085 + }, + { + "epoch": 0.2783063534292285, + "grad_norm": 1.9479890969693905, + "learning_rate": 3.387424434397907e-06, + "loss": 0.9593, + "step": 3086 + }, + { + "epoch": 0.2783965369526987, + "grad_norm": 1.7828934432969816, + "learning_rate": 3.3870035961703013e-06, + "loss": 0.8857, + "step": 3087 + }, + { + "epoch": 0.278486720476169, + "grad_norm": 1.9100040304930337, + "learning_rate": 3.3865826395959014e-06, + "loss": 1.0967, + "step": 3088 + }, + { + "epoch": 0.27857690399963925, + "grad_norm": 1.4766819819796408, + "learning_rate": 3.3861615647106253e-06, + "loss": 1.0636, + "step": 3089 + }, + { + "epoch": 0.27866708752310954, + "grad_norm": 2.0212686469657584, + "learning_rate": 3.3857403715504012e-06, + "loss": 0.9823, + "step": 3090 + }, + { + "epoch": 0.27875727104657977, + "grad_norm": 1.6263879570585962, + "learning_rate": 3.385319060151167e-06, + "loss": 0.9378, + "step": 3091 + }, + { + "epoch": 0.27884745457005006, + "grad_norm": 1.8529561909923005, + "learning_rate": 3.3848976305488728e-06, + "loss": 0.9628, + "step": 3092 + }, + { + "epoch": 0.2789376380935203, + "grad_norm": 2.0473809235802296, + "learning_rate": 3.384476082779476e-06, + "loss": 1.0146, + "step": 3093 + }, + { + "epoch": 0.2790278216169906, + "grad_norm": 1.780292722228243, + "learning_rate": 3.3840544168789463e-06, + "loss": 0.951, + "step": 3094 + }, + { + "epoch": 0.2791180051404608, + "grad_norm": 1.4945978001295939, + "learning_rate": 3.3836326328832617e-06, + "loss": 1.0816, + "step": 3095 + }, + { + "epoch": 0.2792081886639311, + "grad_norm": 1.6216890643782635, + "learning_rate": 3.383210730828412e-06, + "loss": 0.9203, + "step": 3096 + }, + { + "epoch": 0.2792983721874014, + "grad_norm": 1.4467594636139902, + "learning_rate": 3.3827887107503953e-06, + "loss": 1.0389, + "step": 3097 + }, + { + "epoch": 0.2793885557108716, + "grad_norm": 1.5375536963537704, + "learning_rate": 3.3823665726852216e-06, + "loss": 0.9226, + "step": 3098 + }, + { + "epoch": 0.2794787392343419, + "grad_norm": 1.3898151911545524, + "learning_rate": 3.3819443166689095e-06, + "loss": 1.0971, + "step": 3099 + }, + { + "epoch": 0.27956892275781214, + "grad_norm": 1.5158853209005707, + "learning_rate": 3.3815219427374886e-06, + "loss": 0.9684, + "step": 3100 + }, + { + "epoch": 0.27965910628128243, + "grad_norm": 1.5564904421700676, + "learning_rate": 3.3810994509269975e-06, + "loss": 0.989, + "step": 3101 + }, + { + "epoch": 0.27974928980475267, + "grad_norm": 1.5088324383176404, + "learning_rate": 3.3806768412734864e-06, + "loss": 1.007, + "step": 3102 + }, + { + "epoch": 0.27983947332822295, + "grad_norm": 1.4582906807611185, + "learning_rate": 3.380254113813014e-06, + "loss": 1.0516, + "step": 3103 + }, + { + "epoch": 0.2799296568516932, + "grad_norm": 1.571240506746661, + "learning_rate": 3.3798312685816496e-06, + "loss": 1.0074, + "step": 3104 + }, + { + "epoch": 0.2800198403751635, + "grad_norm": 1.654010265973485, + "learning_rate": 3.3794083056154738e-06, + "loss": 0.9739, + "step": 3105 + }, + { + "epoch": 0.2801100238986337, + "grad_norm": 1.4859540069978903, + "learning_rate": 3.3789852249505746e-06, + "loss": 0.9853, + "step": 3106 + }, + { + "epoch": 0.280200207422104, + "grad_norm": 1.6286453294239864, + "learning_rate": 3.378562026623053e-06, + "loss": 1.0314, + "step": 3107 + }, + { + "epoch": 0.28029039094557423, + "grad_norm": 1.7142379490124324, + "learning_rate": 3.3781387106690175e-06, + "loss": 1.0434, + "step": 3108 + }, + { + "epoch": 0.2803805744690445, + "grad_norm": 1.3961490255128317, + "learning_rate": 3.3777152771245885e-06, + "loss": 1.0005, + "step": 3109 + }, + { + "epoch": 0.28047075799251475, + "grad_norm": 1.6584010281316255, + "learning_rate": 3.377291726025895e-06, + "loss": 1.0074, + "step": 3110 + }, + { + "epoch": 0.28056094151598504, + "grad_norm": 2.1432926235020955, + "learning_rate": 3.3768680574090782e-06, + "loss": 1.0897, + "step": 3111 + }, + { + "epoch": 0.2806511250394553, + "grad_norm": 1.772601699494119, + "learning_rate": 3.3764442713102857e-06, + "loss": 0.9483, + "step": 3112 + }, + { + "epoch": 0.28074130856292556, + "grad_norm": 1.5281652402357684, + "learning_rate": 3.3760203677656786e-06, + "loss": 0.968, + "step": 3113 + }, + { + "epoch": 0.2808314920863958, + "grad_norm": 2.254694914723006, + "learning_rate": 3.3755963468114262e-06, + "loss": 1.0292, + "step": 3114 + }, + { + "epoch": 0.2809216756098661, + "grad_norm": 1.5968019024566218, + "learning_rate": 3.3751722084837095e-06, + "loss": 0.8843, + "step": 3115 + }, + { + "epoch": 0.2810118591333363, + "grad_norm": 1.6556433520338585, + "learning_rate": 3.3747479528187166e-06, + "loss": 0.9514, + "step": 3116 + }, + { + "epoch": 0.2811020426568066, + "grad_norm": 1.6767280233821886, + "learning_rate": 3.3743235798526485e-06, + "loss": 1.0271, + "step": 3117 + }, + { + "epoch": 0.28119222618027684, + "grad_norm": 1.47028635713238, + "learning_rate": 3.373899089621714e-06, + "loss": 0.9126, + "step": 3118 + }, + { + "epoch": 0.2812824097037471, + "grad_norm": 1.6306571789421576, + "learning_rate": 3.373474482162134e-06, + "loss": 0.9538, + "step": 3119 + }, + { + "epoch": 0.28137259322721736, + "grad_norm": 1.7886849883799103, + "learning_rate": 3.3730497575101376e-06, + "loss": 0.9783, + "step": 3120 + }, + { + "epoch": 0.28146277675068765, + "grad_norm": 1.5942537596157107, + "learning_rate": 3.3726249157019654e-06, + "loss": 0.9675, + "step": 3121 + }, + { + "epoch": 0.28155296027415794, + "grad_norm": 1.4999146437201167, + "learning_rate": 3.372199956773866e-06, + "loss": 1.0601, + "step": 3122 + }, + { + "epoch": 0.28164314379762817, + "grad_norm": 1.5271895152703734, + "learning_rate": 3.371774880762101e-06, + "loss": 0.932, + "step": 3123 + }, + { + "epoch": 0.28173332732109846, + "grad_norm": 1.8736163438152555, + "learning_rate": 3.3713496877029392e-06, + "loss": 0.968, + "step": 3124 + }, + { + "epoch": 0.2818235108445687, + "grad_norm": 1.9648461161726247, + "learning_rate": 3.37092437763266e-06, + "loss": 1.0369, + "step": 3125 + }, + { + "epoch": 0.281913694368039, + "grad_norm": 1.7388117634315887, + "learning_rate": 3.3704989505875537e-06, + "loss": 1.004, + "step": 3126 + }, + { + "epoch": 0.2820038778915092, + "grad_norm": 3.3082430855254947, + "learning_rate": 3.3700734066039205e-06, + "loss": 0.9477, + "step": 3127 + }, + { + "epoch": 0.2820940614149795, + "grad_norm": 1.9323914151695298, + "learning_rate": 3.36964774571807e-06, + "loss": 1.0088, + "step": 3128 + }, + { + "epoch": 0.28218424493844974, + "grad_norm": 1.7655603852728459, + "learning_rate": 3.3692219679663206e-06, + "loss": 0.9715, + "step": 3129 + }, + { + "epoch": 0.28227442846192, + "grad_norm": 1.9360770876409674, + "learning_rate": 3.3687960733850043e-06, + "loss": 0.9961, + "step": 3130 + }, + { + "epoch": 0.28236461198539026, + "grad_norm": 1.697330240966505, + "learning_rate": 3.3683700620104586e-06, + "loss": 0.9139, + "step": 3131 + }, + { + "epoch": 0.28245479550886055, + "grad_norm": 1.859756815199807, + "learning_rate": 3.3679439338790347e-06, + "loss": 0.9924, + "step": 3132 + }, + { + "epoch": 0.2825449790323308, + "grad_norm": 1.7305848920093252, + "learning_rate": 3.3675176890270916e-06, + "loss": 1.1004, + "step": 3133 + }, + { + "epoch": 0.28263516255580107, + "grad_norm": 1.433073609500001, + "learning_rate": 3.367091327490998e-06, + "loss": 0.9993, + "step": 3134 + }, + { + "epoch": 0.2827253460792713, + "grad_norm": 2.3202590679348685, + "learning_rate": 3.3666648493071347e-06, + "loss": 0.9592, + "step": 3135 + }, + { + "epoch": 0.2828155296027416, + "grad_norm": 1.6940810833636324, + "learning_rate": 3.3662382545118914e-06, + "loss": 0.9558, + "step": 3136 + }, + { + "epoch": 0.2829057131262118, + "grad_norm": 1.6992807661994789, + "learning_rate": 3.3658115431416663e-06, + "loss": 1.0606, + "step": 3137 + }, + { + "epoch": 0.2829958966496821, + "grad_norm": 0.8263657795500297, + "learning_rate": 3.36538471523287e-06, + "loss": 0.8104, + "step": 3138 + }, + { + "epoch": 0.28308608017315234, + "grad_norm": 1.537268507983302, + "learning_rate": 3.3649577708219204e-06, + "loss": 1.0367, + "step": 3139 + }, + { + "epoch": 0.28317626369662263, + "grad_norm": 1.8206488171743929, + "learning_rate": 3.3645307099452477e-06, + "loss": 0.9215, + "step": 3140 + }, + { + "epoch": 0.28326644722009287, + "grad_norm": 1.7552752684858175, + "learning_rate": 3.3641035326392907e-06, + "loss": 0.9749, + "step": 3141 + }, + { + "epoch": 0.28335663074356315, + "grad_norm": 2.0189766629643056, + "learning_rate": 3.363676238940499e-06, + "loss": 0.9806, + "step": 3142 + }, + { + "epoch": 0.2834468142670334, + "grad_norm": 1.801175816932258, + "learning_rate": 3.363248828885331e-06, + "loss": 1.0101, + "step": 3143 + }, + { + "epoch": 0.2835369977905037, + "grad_norm": 1.4711408432337447, + "learning_rate": 3.3628213025102562e-06, + "loss": 0.9985, + "step": 3144 + }, + { + "epoch": 0.28362718131397396, + "grad_norm": 1.6562688574527158, + "learning_rate": 3.3623936598517536e-06, + "loss": 1.0167, + "step": 3145 + }, + { + "epoch": 0.2837173648374442, + "grad_norm": 1.8489906702213708, + "learning_rate": 3.3619659009463117e-06, + "loss": 0.8262, + "step": 3146 + }, + { + "epoch": 0.2838075483609145, + "grad_norm": 1.4590448369541889, + "learning_rate": 3.3615380258304287e-06, + "loss": 1.0304, + "step": 3147 + }, + { + "epoch": 0.2838977318843847, + "grad_norm": 1.7436595223310294, + "learning_rate": 3.3611100345406146e-06, + "loss": 0.9646, + "step": 3148 + }, + { + "epoch": 0.283987915407855, + "grad_norm": 1.6138058010532692, + "learning_rate": 3.3606819271133873e-06, + "loss": 0.9686, + "step": 3149 + }, + { + "epoch": 0.28407809893132524, + "grad_norm": 0.8270538528210578, + "learning_rate": 3.360253703585275e-06, + "loss": 0.7951, + "step": 3150 + }, + { + "epoch": 0.28416828245479553, + "grad_norm": 1.5519934079875757, + "learning_rate": 3.3598253639928164e-06, + "loss": 0.9147, + "step": 3151 + }, + { + "epoch": 0.28425846597826576, + "grad_norm": 1.6678960636927038, + "learning_rate": 3.3593969083725596e-06, + "loss": 0.932, + "step": 3152 + }, + { + "epoch": 0.28434864950173605, + "grad_norm": 2.683060638914045, + "learning_rate": 3.358968336761063e-06, + "loss": 0.9291, + "step": 3153 + }, + { + "epoch": 0.2844388330252063, + "grad_norm": 1.780743844915611, + "learning_rate": 3.3585396491948945e-06, + "loss": 0.9408, + "step": 3154 + }, + { + "epoch": 0.28452901654867657, + "grad_norm": 1.3847874476942095, + "learning_rate": 3.358110845710633e-06, + "loss": 1.0314, + "step": 3155 + }, + { + "epoch": 0.2846192000721468, + "grad_norm": 1.7699063137282154, + "learning_rate": 3.357681926344865e-06, + "loss": 0.9659, + "step": 3156 + }, + { + "epoch": 0.2847093835956171, + "grad_norm": 1.5026629175857535, + "learning_rate": 3.357252891134189e-06, + "loss": 0.9749, + "step": 3157 + }, + { + "epoch": 0.2847995671190873, + "grad_norm": 2.2544443952446467, + "learning_rate": 3.356823740115212e-06, + "loss": 1.0238, + "step": 3158 + }, + { + "epoch": 0.2848897506425576, + "grad_norm": 1.295134652033953, + "learning_rate": 3.3563944733245525e-06, + "loss": 0.9181, + "step": 3159 + }, + { + "epoch": 0.28497993416602785, + "grad_norm": 1.9058275223634926, + "learning_rate": 3.3559650907988375e-06, + "loss": 0.9687, + "step": 3160 + }, + { + "epoch": 0.28507011768949814, + "grad_norm": 1.8513278088132918, + "learning_rate": 3.3555355925747045e-06, + "loss": 0.8302, + "step": 3161 + }, + { + "epoch": 0.28516030121296837, + "grad_norm": 1.7620945971509225, + "learning_rate": 3.3551059786888e-06, + "loss": 0.9618, + "step": 3162 + }, + { + "epoch": 0.28525048473643866, + "grad_norm": 1.4181668902380653, + "learning_rate": 3.3546762491777807e-06, + "loss": 1.0283, + "step": 3163 + }, + { + "epoch": 0.2853406682599089, + "grad_norm": 1.595237934936754, + "learning_rate": 3.3542464040783156e-06, + "loss": 0.9679, + "step": 3164 + }, + { + "epoch": 0.2854308517833792, + "grad_norm": 1.7458671405708466, + "learning_rate": 3.353816443427079e-06, + "loss": 0.9867, + "step": 3165 + }, + { + "epoch": 0.2855210353068494, + "grad_norm": 1.8634647303038567, + "learning_rate": 3.3533863672607597e-06, + "loss": 0.9614, + "step": 3166 + }, + { + "epoch": 0.2856112188303197, + "grad_norm": 1.5323933595079822, + "learning_rate": 3.352956175616052e-06, + "loss": 1.01, + "step": 3167 + }, + { + "epoch": 0.28570140235378993, + "grad_norm": 1.5947732071791354, + "learning_rate": 3.352525868529664e-06, + "loss": 0.9432, + "step": 3168 + }, + { + "epoch": 0.2857915858772602, + "grad_norm": 1.7125654444754304, + "learning_rate": 3.3520954460383103e-06, + "loss": 0.9128, + "step": 3169 + }, + { + "epoch": 0.2858817694007305, + "grad_norm": 1.852892755940748, + "learning_rate": 3.3516649081787182e-06, + "loss": 0.9186, + "step": 3170 + }, + { + "epoch": 0.28597195292420075, + "grad_norm": 1.4416603753424875, + "learning_rate": 3.3512342549876236e-06, + "loss": 1.0739, + "step": 3171 + }, + { + "epoch": 0.28606213644767103, + "grad_norm": 1.5929017895239888, + "learning_rate": 3.350803486501771e-06, + "loss": 0.9255, + "step": 3172 + }, + { + "epoch": 0.28615231997114127, + "grad_norm": 1.510772566684092, + "learning_rate": 3.3503726027579175e-06, + "loss": 0.9957, + "step": 3173 + }, + { + "epoch": 0.28624250349461156, + "grad_norm": 1.3948610914794761, + "learning_rate": 3.349941603792827e-06, + "loss": 1.0316, + "step": 3174 + }, + { + "epoch": 0.2863326870180818, + "grad_norm": 1.5992648879767488, + "learning_rate": 3.3495104896432755e-06, + "loss": 0.971, + "step": 3175 + }, + { + "epoch": 0.2864228705415521, + "grad_norm": 1.4792976679571912, + "learning_rate": 3.3490792603460477e-06, + "loss": 0.9375, + "step": 3176 + }, + { + "epoch": 0.2865130540650223, + "grad_norm": 1.3225009545393096, + "learning_rate": 3.3486479159379393e-06, + "loss": 0.9056, + "step": 3177 + }, + { + "epoch": 0.2866032375884926, + "grad_norm": 0.7763550187602746, + "learning_rate": 3.3482164564557537e-06, + "loss": 0.7783, + "step": 3178 + }, + { + "epoch": 0.28669342111196283, + "grad_norm": 1.6951762201759295, + "learning_rate": 3.3477848819363065e-06, + "loss": 1.0303, + "step": 3179 + }, + { + "epoch": 0.2867836046354331, + "grad_norm": 0.7337133389336025, + "learning_rate": 3.3473531924164213e-06, + "loss": 0.8067, + "step": 3180 + }, + { + "epoch": 0.28687378815890335, + "grad_norm": 1.85036278337918, + "learning_rate": 3.3469213879329325e-06, + "loss": 0.9323, + "step": 3181 + }, + { + "epoch": 0.28696397168237364, + "grad_norm": 1.5898177311148327, + "learning_rate": 3.3464894685226837e-06, + "loss": 0.9343, + "step": 3182 + }, + { + "epoch": 0.2870541552058439, + "grad_norm": 1.5308367113059866, + "learning_rate": 3.34605743422253e-06, + "loss": 1.0522, + "step": 3183 + }, + { + "epoch": 0.28714433872931416, + "grad_norm": 1.818929908479953, + "learning_rate": 3.345625285069333e-06, + "loss": 0.9442, + "step": 3184 + }, + { + "epoch": 0.2872345222527844, + "grad_norm": 1.4752167720236369, + "learning_rate": 3.345193021099967e-06, + "loss": 0.9551, + "step": 3185 + }, + { + "epoch": 0.2873247057762547, + "grad_norm": 1.4023430274387718, + "learning_rate": 3.3447606423513157e-06, + "loss": 0.944, + "step": 3186 + }, + { + "epoch": 0.2874148892997249, + "grad_norm": 1.4457213777112918, + "learning_rate": 3.344328148860271e-06, + "loss": 0.9481, + "step": 3187 + }, + { + "epoch": 0.2875050728231952, + "grad_norm": 8.702951426119224, + "learning_rate": 3.3438955406637365e-06, + "loss": 0.9875, + "step": 3188 + }, + { + "epoch": 0.28759525634666544, + "grad_norm": 1.4475216040703809, + "learning_rate": 3.343462817798624e-06, + "loss": 1.0001, + "step": 3189 + }, + { + "epoch": 0.28768543987013573, + "grad_norm": 1.715902223406788, + "learning_rate": 3.343029980301856e-06, + "loss": 0.9438, + "step": 3190 + }, + { + "epoch": 0.28777562339360596, + "grad_norm": 2.154234898247211, + "learning_rate": 3.342597028210365e-06, + "loss": 0.9989, + "step": 3191 + }, + { + "epoch": 0.28786580691707625, + "grad_norm": 1.9456652585725371, + "learning_rate": 3.342163961561092e-06, + "loss": 0.9233, + "step": 3192 + }, + { + "epoch": 0.28795599044054654, + "grad_norm": 1.3823052461939784, + "learning_rate": 3.34173078039099e-06, + "loss": 0.9442, + "step": 3193 + }, + { + "epoch": 0.28804617396401677, + "grad_norm": 1.818985942729636, + "learning_rate": 3.3412974847370193e-06, + "loss": 1.14, + "step": 3194 + }, + { + "epoch": 0.28813635748748706, + "grad_norm": 1.447982466412394, + "learning_rate": 3.3408640746361514e-06, + "loss": 0.9962, + "step": 3195 + }, + { + "epoch": 0.2882265410109573, + "grad_norm": 0.806102767334894, + "learning_rate": 3.3404305501253663e-06, + "loss": 0.8006, + "step": 3196 + }, + { + "epoch": 0.2883167245344276, + "grad_norm": 1.3937178945906337, + "learning_rate": 3.3399969112416565e-06, + "loss": 1.0427, + "step": 3197 + }, + { + "epoch": 0.2884069080578978, + "grad_norm": 1.6289805297015831, + "learning_rate": 3.3395631580220213e-06, + "loss": 0.9137, + "step": 3198 + }, + { + "epoch": 0.2884970915813681, + "grad_norm": 2.232432127601129, + "learning_rate": 3.3391292905034714e-06, + "loss": 0.9513, + "step": 3199 + }, + { + "epoch": 0.28858727510483834, + "grad_norm": 1.6482065047998686, + "learning_rate": 3.338695308723027e-06, + "loss": 1.0384, + "step": 3200 + }, + { + "epoch": 0.2886774586283086, + "grad_norm": 1.7295429862007863, + "learning_rate": 3.338261212717716e-06, + "loss": 1.0394, + "step": 3201 + }, + { + "epoch": 0.28876764215177886, + "grad_norm": 1.4417957306937175, + "learning_rate": 3.33782700252458e-06, + "loss": 1.0037, + "step": 3202 + }, + { + "epoch": 0.28885782567524915, + "grad_norm": 1.5469803437091991, + "learning_rate": 3.337392678180668e-06, + "loss": 1.0176, + "step": 3203 + }, + { + "epoch": 0.2889480091987194, + "grad_norm": 1.6103400142665063, + "learning_rate": 3.3369582397230377e-06, + "loss": 0.9796, + "step": 3204 + }, + { + "epoch": 0.28903819272218967, + "grad_norm": 1.6380285880466734, + "learning_rate": 3.336523687188759e-06, + "loss": 1.0186, + "step": 3205 + }, + { + "epoch": 0.2891283762456599, + "grad_norm": 0.8005055022565843, + "learning_rate": 3.336089020614909e-06, + "loss": 0.815, + "step": 3206 + }, + { + "epoch": 0.2892185597691302, + "grad_norm": 1.6407316627343793, + "learning_rate": 3.3356542400385774e-06, + "loss": 0.8575, + "step": 3207 + }, + { + "epoch": 0.2893087432926004, + "grad_norm": 1.588614548642409, + "learning_rate": 3.3352193454968607e-06, + "loss": 0.9104, + "step": 3208 + }, + { + "epoch": 0.2893989268160707, + "grad_norm": 1.643150783146448, + "learning_rate": 3.3347843370268675e-06, + "loss": 1.0389, + "step": 3209 + }, + { + "epoch": 0.28948911033954094, + "grad_norm": 1.3465788584273484, + "learning_rate": 3.334349214665715e-06, + "loss": 0.997, + "step": 3210 + }, + { + "epoch": 0.28957929386301123, + "grad_norm": 1.693683034551359, + "learning_rate": 3.3339139784505293e-06, + "loss": 1.0563, + "step": 3211 + }, + { + "epoch": 0.28966947738648147, + "grad_norm": 1.4385702875684525, + "learning_rate": 3.333478628418448e-06, + "loss": 1.0448, + "step": 3212 + }, + { + "epoch": 0.28975966090995176, + "grad_norm": 1.4641546230584352, + "learning_rate": 3.333043164606618e-06, + "loss": 0.8639, + "step": 3213 + }, + { + "epoch": 0.289849844433422, + "grad_norm": 1.6612986504841518, + "learning_rate": 3.3326075870521948e-06, + "loss": 1.0581, + "step": 3214 + }, + { + "epoch": 0.2899400279568923, + "grad_norm": 1.6386202234577887, + "learning_rate": 3.3321718957923437e-06, + "loss": 0.9786, + "step": 3215 + }, + { + "epoch": 0.29003021148036257, + "grad_norm": 1.6679535983467264, + "learning_rate": 3.3317360908642413e-06, + "loss": 0.9741, + "step": 3216 + }, + { + "epoch": 0.2901203950038328, + "grad_norm": 2.163959927658542, + "learning_rate": 3.331300172305072e-06, + "loss": 0.9301, + "step": 3217 + }, + { + "epoch": 0.2902105785273031, + "grad_norm": 1.4337201387576646, + "learning_rate": 3.330864140152032e-06, + "loss": 0.9673, + "step": 3218 + }, + { + "epoch": 0.2903007620507733, + "grad_norm": 1.599642835926166, + "learning_rate": 3.330427994442325e-06, + "loss": 0.9915, + "step": 3219 + }, + { + "epoch": 0.2903909455742436, + "grad_norm": 1.6724382681637255, + "learning_rate": 3.3299917352131657e-06, + "loss": 1.0081, + "step": 3220 + }, + { + "epoch": 0.29048112909771384, + "grad_norm": 1.9276276996629018, + "learning_rate": 3.329555362501778e-06, + "loss": 0.9453, + "step": 3221 + }, + { + "epoch": 0.29057131262118413, + "grad_norm": 1.5797953311328157, + "learning_rate": 3.3291188763453954e-06, + "loss": 0.9413, + "step": 3222 + }, + { + "epoch": 0.29066149614465436, + "grad_norm": 1.6233854342382539, + "learning_rate": 3.3286822767812618e-06, + "loss": 0.9643, + "step": 3223 + }, + { + "epoch": 0.29075167966812465, + "grad_norm": 1.7336827176659078, + "learning_rate": 3.32824556384663e-06, + "loss": 0.9577, + "step": 3224 + }, + { + "epoch": 0.2908418631915949, + "grad_norm": 1.988553970908986, + "learning_rate": 3.3278087375787628e-06, + "loss": 0.9767, + "step": 3225 + }, + { + "epoch": 0.2909320467150652, + "grad_norm": 2.028000800469295, + "learning_rate": 3.327371798014933e-06, + "loss": 0.9809, + "step": 3226 + }, + { + "epoch": 0.2910222302385354, + "grad_norm": 1.7663291944714574, + "learning_rate": 3.3269347451924218e-06, + "loss": 0.9796, + "step": 3227 + }, + { + "epoch": 0.2911124137620057, + "grad_norm": 1.506787678210919, + "learning_rate": 3.326497579148522e-06, + "loss": 1.0431, + "step": 3228 + }, + { + "epoch": 0.29120259728547593, + "grad_norm": 1.6408172494782296, + "learning_rate": 3.3260602999205345e-06, + "loss": 0.9748, + "step": 3229 + }, + { + "epoch": 0.2912927808089462, + "grad_norm": 1.5581088351479615, + "learning_rate": 3.32562290754577e-06, + "loss": 0.992, + "step": 3230 + }, + { + "epoch": 0.29138296433241645, + "grad_norm": 1.5721590183675518, + "learning_rate": 3.3251854020615494e-06, + "loss": 0.958, + "step": 3231 + }, + { + "epoch": 0.29147314785588674, + "grad_norm": 1.574182476827948, + "learning_rate": 3.324747783505204e-06, + "loss": 0.9761, + "step": 3232 + }, + { + "epoch": 0.29156333137935697, + "grad_norm": 1.683856704604595, + "learning_rate": 3.324310051914073e-06, + "loss": 0.8897, + "step": 3233 + }, + { + "epoch": 0.29165351490282726, + "grad_norm": 1.8038076262280174, + "learning_rate": 3.3238722073255056e-06, + "loss": 0.9455, + "step": 3234 + }, + { + "epoch": 0.2917436984262975, + "grad_norm": 1.4680899090428097, + "learning_rate": 3.323434249776863e-06, + "loss": 0.9353, + "step": 3235 + }, + { + "epoch": 0.2918338819497678, + "grad_norm": 0.7689208887478418, + "learning_rate": 3.3229961793055117e-06, + "loss": 0.7911, + "step": 3236 + }, + { + "epoch": 0.291924065473238, + "grad_norm": 1.5654161325344988, + "learning_rate": 3.3225579959488314e-06, + "loss": 0.8974, + "step": 3237 + }, + { + "epoch": 0.2920142489967083, + "grad_norm": 1.7847989515673381, + "learning_rate": 3.322119699744211e-06, + "loss": 1.0009, + "step": 3238 + }, + { + "epoch": 0.29210443252017854, + "grad_norm": 1.539452827645239, + "learning_rate": 3.3216812907290476e-06, + "loss": 0.9904, + "step": 3239 + }, + { + "epoch": 0.2921946160436488, + "grad_norm": 1.5173349075130438, + "learning_rate": 3.3212427689407484e-06, + "loss": 0.9955, + "step": 3240 + }, + { + "epoch": 0.2922847995671191, + "grad_norm": 0.764918254149825, + "learning_rate": 3.3208041344167317e-06, + "loss": 0.7691, + "step": 3241 + }, + { + "epoch": 0.29237498309058935, + "grad_norm": 1.665441666867905, + "learning_rate": 3.3203653871944224e-06, + "loss": 0.8841, + "step": 3242 + }, + { + "epoch": 0.29246516661405964, + "grad_norm": 1.681423408045702, + "learning_rate": 3.3199265273112587e-06, + "loss": 0.94, + "step": 3243 + }, + { + "epoch": 0.29255535013752987, + "grad_norm": 1.6394350368546258, + "learning_rate": 3.3194875548046852e-06, + "loss": 0.9156, + "step": 3244 + }, + { + "epoch": 0.29264553366100016, + "grad_norm": 1.5436767498970136, + "learning_rate": 3.319048469712158e-06, + "loss": 0.957, + "step": 3245 + }, + { + "epoch": 0.2927357171844704, + "grad_norm": 1.8077073365849468, + "learning_rate": 3.3186092720711423e-06, + "loss": 1.0699, + "step": 3246 + }, + { + "epoch": 0.2928259007079407, + "grad_norm": 1.8098544182372827, + "learning_rate": 3.3181699619191125e-06, + "loss": 0.9046, + "step": 3247 + }, + { + "epoch": 0.2929160842314109, + "grad_norm": 1.5473541134303594, + "learning_rate": 3.3177305392935536e-06, + "loss": 1.0138, + "step": 3248 + }, + { + "epoch": 0.2930062677548812, + "grad_norm": 0.8074959937312183, + "learning_rate": 3.3172910042319595e-06, + "loss": 0.8573, + "step": 3249 + }, + { + "epoch": 0.29309645127835143, + "grad_norm": 1.6687950213927691, + "learning_rate": 3.316851356771833e-06, + "loss": 1.0105, + "step": 3250 + }, + { + "epoch": 0.2931866348018217, + "grad_norm": 1.3800851251082906, + "learning_rate": 3.3164115969506876e-06, + "loss": 1.0078, + "step": 3251 + }, + { + "epoch": 0.29327681832529195, + "grad_norm": 2.3682565824388395, + "learning_rate": 3.315971724806046e-06, + "loss": 1.0649, + "step": 3252 + }, + { + "epoch": 0.29336700184876224, + "grad_norm": 1.6886942134652996, + "learning_rate": 3.315531740375441e-06, + "loss": 0.9153, + "step": 3253 + }, + { + "epoch": 0.2934571853722325, + "grad_norm": 1.4823373561884892, + "learning_rate": 3.315091643696414e-06, + "loss": 0.9439, + "step": 3254 + }, + { + "epoch": 0.29354736889570276, + "grad_norm": 1.5533964844720007, + "learning_rate": 3.3146514348065164e-06, + "loss": 0.9035, + "step": 3255 + }, + { + "epoch": 0.293637552419173, + "grad_norm": 1.643212303723978, + "learning_rate": 3.31421111374331e-06, + "loss": 0.9803, + "step": 3256 + }, + { + "epoch": 0.2937277359426433, + "grad_norm": 1.5030468355344688, + "learning_rate": 3.3137706805443647e-06, + "loss": 0.9876, + "step": 3257 + }, + { + "epoch": 0.2938179194661135, + "grad_norm": 1.5179010977912168, + "learning_rate": 3.313330135247261e-06, + "loss": 1.023, + "step": 3258 + }, + { + "epoch": 0.2939081029895838, + "grad_norm": 1.4467653138210623, + "learning_rate": 3.312889477889588e-06, + "loss": 0.9471, + "step": 3259 + }, + { + "epoch": 0.29399828651305404, + "grad_norm": 1.5268243164944422, + "learning_rate": 3.3124487085089464e-06, + "loss": 0.9888, + "step": 3260 + }, + { + "epoch": 0.29408847003652433, + "grad_norm": 1.6786724483389415, + "learning_rate": 3.312007827142943e-06, + "loss": 1.011, + "step": 3261 + }, + { + "epoch": 0.29417865355999456, + "grad_norm": 1.4264901684978144, + "learning_rate": 3.3115668338291983e-06, + "loss": 0.8934, + "step": 3262 + }, + { + "epoch": 0.29426883708346485, + "grad_norm": 1.576211354495109, + "learning_rate": 3.3111257286053394e-06, + "loss": 1.0592, + "step": 3263 + }, + { + "epoch": 0.29435902060693514, + "grad_norm": 2.0432629346103237, + "learning_rate": 3.3106845115090043e-06, + "loss": 0.9772, + "step": 3264 + }, + { + "epoch": 0.2944492041304054, + "grad_norm": 1.6193674478535138, + "learning_rate": 3.310243182577839e-06, + "loss": 0.9299, + "step": 3265 + }, + { + "epoch": 0.29453938765387566, + "grad_norm": 1.7490031945672149, + "learning_rate": 3.3098017418495007e-06, + "loss": 0.8675, + "step": 3266 + }, + { + "epoch": 0.2946295711773459, + "grad_norm": 1.797760985505203, + "learning_rate": 3.309360189361656e-06, + "loss": 0.9388, + "step": 3267 + }, + { + "epoch": 0.2947197547008162, + "grad_norm": 1.505996796785674, + "learning_rate": 3.3089185251519797e-06, + "loss": 1.0544, + "step": 3268 + }, + { + "epoch": 0.2948099382242864, + "grad_norm": 1.7503272159382097, + "learning_rate": 3.3084767492581574e-06, + "loss": 0.9516, + "step": 3269 + }, + { + "epoch": 0.2949001217477567, + "grad_norm": 1.675483090212058, + "learning_rate": 3.3080348617178846e-06, + "loss": 0.9874, + "step": 3270 + }, + { + "epoch": 0.29499030527122694, + "grad_norm": 1.5609828449799752, + "learning_rate": 3.307592862568865e-06, + "loss": 0.9454, + "step": 3271 + }, + { + "epoch": 0.2950804887946972, + "grad_norm": 1.3272050700900804, + "learning_rate": 3.307150751848812e-06, + "loss": 1.0404, + "step": 3272 + }, + { + "epoch": 0.29517067231816746, + "grad_norm": 1.4464304131793855, + "learning_rate": 3.3067085295954497e-06, + "loss": 0.8797, + "step": 3273 + }, + { + "epoch": 0.29526085584163775, + "grad_norm": 1.6554819161439263, + "learning_rate": 3.3062661958465098e-06, + "loss": 0.8357, + "step": 3274 + }, + { + "epoch": 0.295351039365108, + "grad_norm": 1.7865191376091047, + "learning_rate": 3.305823750639736e-06, + "loss": 0.8662, + "step": 3275 + }, + { + "epoch": 0.29544122288857827, + "grad_norm": 1.624074966062788, + "learning_rate": 3.3053811940128795e-06, + "loss": 0.9668, + "step": 3276 + }, + { + "epoch": 0.2955314064120485, + "grad_norm": 1.4610535274676737, + "learning_rate": 3.3049385260037016e-06, + "loss": 0.9519, + "step": 3277 + }, + { + "epoch": 0.2956215899355188, + "grad_norm": 1.373262174176307, + "learning_rate": 3.3044957466499736e-06, + "loss": 0.8636, + "step": 3278 + }, + { + "epoch": 0.295711773458989, + "grad_norm": 1.5491663598355623, + "learning_rate": 3.304052855989475e-06, + "loss": 0.8944, + "step": 3279 + }, + { + "epoch": 0.2958019569824593, + "grad_norm": 1.515760671304659, + "learning_rate": 3.3036098540599966e-06, + "loss": 1.031, + "step": 3280 + }, + { + "epoch": 0.29589214050592955, + "grad_norm": 1.8596384839492615, + "learning_rate": 3.3031667408993373e-06, + "loss": 0.9285, + "step": 3281 + }, + { + "epoch": 0.29598232402939983, + "grad_norm": 1.6929815766289165, + "learning_rate": 3.302723516545306e-06, + "loss": 0.9682, + "step": 3282 + }, + { + "epoch": 0.29607250755287007, + "grad_norm": 1.5366850173886077, + "learning_rate": 3.302280181035722e-06, + "loss": 0.9965, + "step": 3283 + }, + { + "epoch": 0.29616269107634036, + "grad_norm": 0.6842657219722056, + "learning_rate": 3.3018367344084117e-06, + "loss": 0.8205, + "step": 3284 + }, + { + "epoch": 0.2962528745998106, + "grad_norm": 1.3250464269313584, + "learning_rate": 3.3013931767012125e-06, + "loss": 0.964, + "step": 3285 + }, + { + "epoch": 0.2963430581232809, + "grad_norm": 1.5529514760765515, + "learning_rate": 3.300949507951972e-06, + "loss": 0.9623, + "step": 3286 + }, + { + "epoch": 0.2964332416467511, + "grad_norm": 4.515268701138554, + "learning_rate": 3.300505728198546e-06, + "loss": 1.0104, + "step": 3287 + }, + { + "epoch": 0.2965234251702214, + "grad_norm": 1.672961586063716, + "learning_rate": 3.3000618374788e-06, + "loss": 0.9561, + "step": 3288 + }, + { + "epoch": 0.2966136086936917, + "grad_norm": 1.459949996039686, + "learning_rate": 3.2996178358306104e-06, + "loss": 1.0086, + "step": 3289 + }, + { + "epoch": 0.2967037922171619, + "grad_norm": 1.541463847179049, + "learning_rate": 3.2991737232918606e-06, + "loss": 0.9341, + "step": 3290 + }, + { + "epoch": 0.2967939757406322, + "grad_norm": 1.8356213885938417, + "learning_rate": 3.298729499900445e-06, + "loss": 0.9228, + "step": 3291 + }, + { + "epoch": 0.29688415926410244, + "grad_norm": 1.6743136294951821, + "learning_rate": 3.2982851656942677e-06, + "loss": 0.9185, + "step": 3292 + }, + { + "epoch": 0.29697434278757273, + "grad_norm": 1.3671189645210384, + "learning_rate": 3.2978407207112416e-06, + "loss": 0.9943, + "step": 3293 + }, + { + "epoch": 0.29706452631104296, + "grad_norm": 2.14629287490277, + "learning_rate": 3.2973961649892888e-06, + "loss": 0.975, + "step": 3294 + }, + { + "epoch": 0.29715470983451325, + "grad_norm": 1.5147897050649237, + "learning_rate": 3.296951498566341e-06, + "loss": 1.0288, + "step": 3295 + }, + { + "epoch": 0.2972448933579835, + "grad_norm": 1.5991327975478193, + "learning_rate": 3.2965067214803404e-06, + "loss": 0.962, + "step": 3296 + }, + { + "epoch": 0.2973350768814538, + "grad_norm": 1.250546097675653, + "learning_rate": 3.2960618337692372e-06, + "loss": 1.0053, + "step": 3297 + }, + { + "epoch": 0.297425260404924, + "grad_norm": 1.5811802965565014, + "learning_rate": 3.2956168354709927e-06, + "loss": 0.9348, + "step": 3298 + }, + { + "epoch": 0.2975154439283943, + "grad_norm": 2.1586663901124856, + "learning_rate": 3.2951717266235754e-06, + "loss": 1.0198, + "step": 3299 + }, + { + "epoch": 0.29760562745186453, + "grad_norm": 1.545122637526335, + "learning_rate": 3.294726507264964e-06, + "loss": 0.8444, + "step": 3300 + }, + { + "epoch": 0.2976958109753348, + "grad_norm": 1.7807868890889769, + "learning_rate": 3.2942811774331487e-06, + "loss": 0.9387, + "step": 3301 + }, + { + "epoch": 0.29778599449880505, + "grad_norm": 1.867176518747128, + "learning_rate": 3.293835737166127e-06, + "loss": 0.9711, + "step": 3302 + }, + { + "epoch": 0.29787617802227534, + "grad_norm": 1.7296493346872037, + "learning_rate": 3.293390186501906e-06, + "loss": 0.9215, + "step": 3303 + }, + { + "epoch": 0.2979663615457456, + "grad_norm": 1.6746806907906886, + "learning_rate": 3.2929445254785024e-06, + "loss": 0.9755, + "step": 3304 + }, + { + "epoch": 0.29805654506921586, + "grad_norm": 2.00693323498169, + "learning_rate": 3.2924987541339423e-06, + "loss": 0.9244, + "step": 3305 + }, + { + "epoch": 0.2981467285926861, + "grad_norm": 1.8176753379064159, + "learning_rate": 3.292052872506262e-06, + "loss": 0.9922, + "step": 3306 + }, + { + "epoch": 0.2982369121161564, + "grad_norm": 1.5227463156856444, + "learning_rate": 3.291606880633506e-06, + "loss": 1.0118, + "step": 3307 + }, + { + "epoch": 0.2983270956396266, + "grad_norm": 2.01787282115377, + "learning_rate": 3.2911607785537297e-06, + "loss": 1.0155, + "step": 3308 + }, + { + "epoch": 0.2984172791630969, + "grad_norm": 1.6868222606395902, + "learning_rate": 3.290714566304997e-06, + "loss": 1.0575, + "step": 3309 + }, + { + "epoch": 0.29850746268656714, + "grad_norm": 2.9420570575567417, + "learning_rate": 3.2902682439253794e-06, + "loss": 1.0403, + "step": 3310 + }, + { + "epoch": 0.2985976462100374, + "grad_norm": 1.959848656358153, + "learning_rate": 3.289821811452961e-06, + "loss": 0.9829, + "step": 3311 + }, + { + "epoch": 0.2986878297335077, + "grad_norm": 1.7698758697087336, + "learning_rate": 3.289375268925834e-06, + "loss": 1.0266, + "step": 3312 + }, + { + "epoch": 0.29877801325697795, + "grad_norm": 1.8833543939453772, + "learning_rate": 3.288928616382099e-06, + "loss": 0.9606, + "step": 3313 + }, + { + "epoch": 0.29886819678044824, + "grad_norm": 1.3603478819830863, + "learning_rate": 3.288481853859868e-06, + "loss": 1.0079, + "step": 3314 + }, + { + "epoch": 0.29895838030391847, + "grad_norm": 1.4154046365029964, + "learning_rate": 3.2880349813972604e-06, + "loss": 0.946, + "step": 3315 + }, + { + "epoch": 0.29904856382738876, + "grad_norm": 1.9935256353432609, + "learning_rate": 3.2875879990324052e-06, + "loss": 0.9699, + "step": 3316 + }, + { + "epoch": 0.299138747350859, + "grad_norm": 1.7937805731424148, + "learning_rate": 3.287140906803443e-06, + "loss": 1.0239, + "step": 3317 + }, + { + "epoch": 0.2992289308743293, + "grad_norm": 1.77750753130169, + "learning_rate": 3.2866937047485216e-06, + "loss": 0.8908, + "step": 3318 + }, + { + "epoch": 0.2993191143977995, + "grad_norm": 2.188238182856169, + "learning_rate": 3.2862463929057985e-06, + "loss": 1.0335, + "step": 3319 + }, + { + "epoch": 0.2994092979212698, + "grad_norm": 1.6199271910989301, + "learning_rate": 3.285798971313441e-06, + "loss": 0.8187, + "step": 3320 + }, + { + "epoch": 0.29949948144474003, + "grad_norm": 1.488546032517662, + "learning_rate": 3.2853514400096248e-06, + "loss": 0.9925, + "step": 3321 + }, + { + "epoch": 0.2995896649682103, + "grad_norm": 1.551869354176324, + "learning_rate": 3.2849037990325367e-06, + "loss": 0.9972, + "step": 3322 + }, + { + "epoch": 0.29967984849168056, + "grad_norm": 1.276426306641802, + "learning_rate": 3.2844560484203717e-06, + "loss": 0.9262, + "step": 3323 + }, + { + "epoch": 0.29977003201515084, + "grad_norm": 1.49793100717159, + "learning_rate": 3.2840081882113333e-06, + "loss": 1.0532, + "step": 3324 + }, + { + "epoch": 0.2998602155386211, + "grad_norm": 1.5478537959615195, + "learning_rate": 3.283560218443638e-06, + "loss": 0.9464, + "step": 3325 + }, + { + "epoch": 0.29995039906209137, + "grad_norm": 1.4975314490784506, + "learning_rate": 3.2831121391555064e-06, + "loss": 0.9863, + "step": 3326 + }, + { + "epoch": 0.3000405825855616, + "grad_norm": 1.4843440002416781, + "learning_rate": 3.2826639503851724e-06, + "loss": 0.9257, + "step": 3327 + }, + { + "epoch": 0.3001307661090319, + "grad_norm": 1.5968155642681663, + "learning_rate": 3.282215652170877e-06, + "loss": 1.0113, + "step": 3328 + }, + { + "epoch": 0.3002209496325021, + "grad_norm": 1.597117214160317, + "learning_rate": 3.281767244550873e-06, + "loss": 1.0052, + "step": 3329 + }, + { + "epoch": 0.3003111331559724, + "grad_norm": 0.6875628746065676, + "learning_rate": 3.2813187275634193e-06, + "loss": 0.7544, + "step": 3330 + }, + { + "epoch": 0.30040131667944264, + "grad_norm": 1.6196531950055464, + "learning_rate": 3.280870101246787e-06, + "loss": 0.9955, + "step": 3331 + }, + { + "epoch": 0.30049150020291293, + "grad_norm": 1.6408182666113302, + "learning_rate": 3.280421365639255e-06, + "loss": 0.8858, + "step": 3332 + }, + { + "epoch": 0.30058168372638316, + "grad_norm": 1.731711120340842, + "learning_rate": 3.279972520779112e-06, + "loss": 0.9603, + "step": 3333 + }, + { + "epoch": 0.30067186724985345, + "grad_norm": 1.3726061443287494, + "learning_rate": 3.279523566704656e-06, + "loss": 1.0322, + "step": 3334 + }, + { + "epoch": 0.30076205077332374, + "grad_norm": 1.7278614940103043, + "learning_rate": 3.2790745034541935e-06, + "loss": 1.0237, + "step": 3335 + }, + { + "epoch": 0.300852234296794, + "grad_norm": 1.5991308593451814, + "learning_rate": 3.278625331066042e-06, + "loss": 1.0173, + "step": 3336 + }, + { + "epoch": 0.30094241782026426, + "grad_norm": 1.6897947462139522, + "learning_rate": 3.278176049578527e-06, + "loss": 1.0206, + "step": 3337 + }, + { + "epoch": 0.3010326013437345, + "grad_norm": 1.6006750858190861, + "learning_rate": 3.2777266590299835e-06, + "loss": 0.9314, + "step": 3338 + }, + { + "epoch": 0.3011227848672048, + "grad_norm": 1.6715479022758102, + "learning_rate": 3.2772771594587562e-06, + "loss": 0.907, + "step": 3339 + }, + { + "epoch": 0.301212968390675, + "grad_norm": 1.8539469370856263, + "learning_rate": 3.2768275509031988e-06, + "loss": 1.0757, + "step": 3340 + }, + { + "epoch": 0.3013031519141453, + "grad_norm": 1.8375600584426592, + "learning_rate": 3.276377833401675e-06, + "loss": 0.9552, + "step": 3341 + }, + { + "epoch": 0.30139333543761554, + "grad_norm": 1.6043971585570278, + "learning_rate": 3.2759280069925557e-06, + "loss": 1.034, + "step": 3342 + }, + { + "epoch": 0.30148351896108583, + "grad_norm": 1.4937446019541993, + "learning_rate": 3.2754780717142233e-06, + "loss": 1.0147, + "step": 3343 + }, + { + "epoch": 0.30157370248455606, + "grad_norm": 1.6980271644755855, + "learning_rate": 3.27502802760507e-06, + "loss": 0.9429, + "step": 3344 + }, + { + "epoch": 0.30166388600802635, + "grad_norm": 2.125594112146024, + "learning_rate": 3.2745778747034943e-06, + "loss": 0.8847, + "step": 3345 + }, + { + "epoch": 0.3017540695314966, + "grad_norm": 2.1980991257659683, + "learning_rate": 3.274127613047906e-06, + "loss": 0.9499, + "step": 3346 + }, + { + "epoch": 0.30184425305496687, + "grad_norm": 1.5621517556258677, + "learning_rate": 3.273677242676725e-06, + "loss": 0.9905, + "step": 3347 + }, + { + "epoch": 0.3019344365784371, + "grad_norm": 1.6006460405461567, + "learning_rate": 3.2732267636283782e-06, + "loss": 0.9667, + "step": 3348 + }, + { + "epoch": 0.3020246201019074, + "grad_norm": 1.7727554959271297, + "learning_rate": 3.2727761759413034e-06, + "loss": 0.9418, + "step": 3349 + }, + { + "epoch": 0.3021148036253776, + "grad_norm": 1.676220745084558, + "learning_rate": 3.2723254796539477e-06, + "loss": 0.9779, + "step": 3350 + }, + { + "epoch": 0.3022049871488479, + "grad_norm": 1.7181611179345222, + "learning_rate": 3.271874674804766e-06, + "loss": 0.9559, + "step": 3351 + }, + { + "epoch": 0.30229517067231815, + "grad_norm": 1.7789498586423391, + "learning_rate": 3.2714237614322242e-06, + "loss": 0.9986, + "step": 3352 + }, + { + "epoch": 0.30238535419578844, + "grad_norm": 1.3887566434630938, + "learning_rate": 3.2709727395747974e-06, + "loss": 0.8028, + "step": 3353 + }, + { + "epoch": 0.30247553771925867, + "grad_norm": 0.761381338656464, + "learning_rate": 3.2705216092709673e-06, + "loss": 0.7736, + "step": 3354 + }, + { + "epoch": 0.30256572124272896, + "grad_norm": 1.833889472301603, + "learning_rate": 3.2700703705592282e-06, + "loss": 1.0576, + "step": 3355 + }, + { + "epoch": 0.3026559047661992, + "grad_norm": 1.4881056634446936, + "learning_rate": 3.269619023478082e-06, + "loss": 0.8713, + "step": 3356 + }, + { + "epoch": 0.3027460882896695, + "grad_norm": 1.698743591521065, + "learning_rate": 3.26916756806604e-06, + "loss": 1.0747, + "step": 3357 + }, + { + "epoch": 0.3028362718131397, + "grad_norm": 1.6947915819477168, + "learning_rate": 3.268716004361623e-06, + "loss": 0.9902, + "step": 3358 + }, + { + "epoch": 0.30292645533661, + "grad_norm": 0.7570153989598826, + "learning_rate": 3.268264332403361e-06, + "loss": 0.7898, + "step": 3359 + }, + { + "epoch": 0.3030166388600803, + "grad_norm": 1.461391113359478, + "learning_rate": 3.2678125522297933e-06, + "loss": 1.0047, + "step": 3360 + }, + { + "epoch": 0.3031068223835505, + "grad_norm": 1.781983876884296, + "learning_rate": 3.267360663879468e-06, + "loss": 0.918, + "step": 3361 + }, + { + "epoch": 0.3031970059070208, + "grad_norm": 2.0128174383360204, + "learning_rate": 3.266908667390942e-06, + "loss": 0.9636, + "step": 3362 + }, + { + "epoch": 0.30328718943049104, + "grad_norm": 1.771070741185112, + "learning_rate": 3.2664565628027833e-06, + "loss": 1.0288, + "step": 3363 + }, + { + "epoch": 0.30337737295396133, + "grad_norm": 1.9071003002400742, + "learning_rate": 3.2660043501535675e-06, + "loss": 0.9723, + "step": 3364 + }, + { + "epoch": 0.30346755647743157, + "grad_norm": 1.7073840861261307, + "learning_rate": 3.2655520294818797e-06, + "loss": 1.0436, + "step": 3365 + }, + { + "epoch": 0.30355774000090185, + "grad_norm": 1.3971322644124966, + "learning_rate": 3.2650996008263146e-06, + "loss": 1.0475, + "step": 3366 + }, + { + "epoch": 0.3036479235243721, + "grad_norm": 1.8379394005881629, + "learning_rate": 3.2646470642254756e-06, + "loss": 0.931, + "step": 3367 + }, + { + "epoch": 0.3037381070478424, + "grad_norm": 1.6924303579723166, + "learning_rate": 3.2641944197179767e-06, + "loss": 0.9983, + "step": 3368 + }, + { + "epoch": 0.3038282905713126, + "grad_norm": 0.7956883439876139, + "learning_rate": 3.2637416673424383e-06, + "loss": 0.7752, + "step": 3369 + }, + { + "epoch": 0.3039184740947829, + "grad_norm": 1.5492297658369938, + "learning_rate": 3.2632888071374937e-06, + "loss": 0.9867, + "step": 3370 + }, + { + "epoch": 0.30400865761825313, + "grad_norm": 1.8144162013112677, + "learning_rate": 3.2628358391417815e-06, + "loss": 1.0047, + "step": 3371 + }, + { + "epoch": 0.3040988411417234, + "grad_norm": 1.5729204224126916, + "learning_rate": 3.2623827633939526e-06, + "loss": 0.9604, + "step": 3372 + }, + { + "epoch": 0.30418902466519365, + "grad_norm": 1.6731055891345257, + "learning_rate": 3.2619295799326657e-06, + "loss": 0.9698, + "step": 3373 + }, + { + "epoch": 0.30427920818866394, + "grad_norm": 1.7960747305056757, + "learning_rate": 3.2614762887965883e-06, + "loss": 0.8702, + "step": 3374 + }, + { + "epoch": 0.3043693917121342, + "grad_norm": 1.4985697603378767, + "learning_rate": 3.2610228900243984e-06, + "loss": 0.9644, + "step": 3375 + }, + { + "epoch": 0.30445957523560446, + "grad_norm": 0.7652197952789921, + "learning_rate": 3.260569383654783e-06, + "loss": 0.8254, + "step": 3376 + }, + { + "epoch": 0.3045497587590747, + "grad_norm": 0.7570840145969492, + "learning_rate": 3.2601157697264365e-06, + "loss": 0.7958, + "step": 3377 + }, + { + "epoch": 0.304639942282545, + "grad_norm": 2.020783440147681, + "learning_rate": 3.2596620482780647e-06, + "loss": 0.9472, + "step": 3378 + }, + { + "epoch": 0.3047301258060152, + "grad_norm": 0.8382978539009022, + "learning_rate": 3.2592082193483803e-06, + "loss": 0.7954, + "step": 3379 + }, + { + "epoch": 0.3048203093294855, + "grad_norm": 1.5045423871475043, + "learning_rate": 3.258754282976109e-06, + "loss": 0.96, + "step": 3380 + }, + { + "epoch": 0.30491049285295574, + "grad_norm": 1.6556632244094036, + "learning_rate": 3.25830023919998e-06, + "loss": 1.0455, + "step": 3381 + }, + { + "epoch": 0.305000676376426, + "grad_norm": 1.8298468672672326, + "learning_rate": 3.2578460880587374e-06, + "loss": 0.9334, + "step": 3382 + }, + { + "epoch": 0.3050908598998963, + "grad_norm": 1.7848999373655294, + "learning_rate": 3.2573918295911306e-06, + "loss": 1.0317, + "step": 3383 + }, + { + "epoch": 0.30518104342336655, + "grad_norm": 1.6741989953410485, + "learning_rate": 3.2569374638359196e-06, + "loss": 0.9096, + "step": 3384 + }, + { + "epoch": 0.30527122694683684, + "grad_norm": 1.460913754846511, + "learning_rate": 3.2564829908318736e-06, + "loss": 0.9921, + "step": 3385 + }, + { + "epoch": 0.30536141047030707, + "grad_norm": 1.3021774359713072, + "learning_rate": 3.2560284106177705e-06, + "loss": 1.0309, + "step": 3386 + }, + { + "epoch": 0.30545159399377736, + "grad_norm": 1.8840789967313134, + "learning_rate": 3.2555737232323978e-06, + "loss": 0.9493, + "step": 3387 + }, + { + "epoch": 0.3055417775172476, + "grad_norm": 1.4195572963544152, + "learning_rate": 3.255118928714552e-06, + "loss": 1.0039, + "step": 3388 + }, + { + "epoch": 0.3056319610407179, + "grad_norm": 1.7857355893090463, + "learning_rate": 3.2546640271030386e-06, + "loss": 0.902, + "step": 3389 + }, + { + "epoch": 0.3057221445641881, + "grad_norm": 0.7479377447857715, + "learning_rate": 3.2542090184366717e-06, + "loss": 0.8269, + "step": 3390 + }, + { + "epoch": 0.3058123280876584, + "grad_norm": 2.1121044443198875, + "learning_rate": 3.253753902754276e-06, + "loss": 1.0555, + "step": 3391 + }, + { + "epoch": 0.30590251161112864, + "grad_norm": 1.3398270675465849, + "learning_rate": 3.253298680094685e-06, + "loss": 0.9092, + "step": 3392 + }, + { + "epoch": 0.3059926951345989, + "grad_norm": 1.8244290281726805, + "learning_rate": 3.2528433504967394e-06, + "loss": 0.9175, + "step": 3393 + }, + { + "epoch": 0.30608287865806916, + "grad_norm": 1.604484832059563, + "learning_rate": 3.252387913999291e-06, + "loss": 0.9839, + "step": 3394 + }, + { + "epoch": 0.30617306218153945, + "grad_norm": 1.5176743485471167, + "learning_rate": 3.2519323706411998e-06, + "loss": 0.976, + "step": 3395 + }, + { + "epoch": 0.3062632457050097, + "grad_norm": 1.721154005577693, + "learning_rate": 3.251476720461336e-06, + "loss": 0.8653, + "step": 3396 + }, + { + "epoch": 0.30635342922847997, + "grad_norm": 1.9094002130470145, + "learning_rate": 3.251020963498578e-06, + "loss": 0.9015, + "step": 3397 + }, + { + "epoch": 0.3064436127519502, + "grad_norm": 1.5791785623040306, + "learning_rate": 3.250565099791813e-06, + "loss": 1.0182, + "step": 3398 + }, + { + "epoch": 0.3065337962754205, + "grad_norm": 1.3382060549265946, + "learning_rate": 3.2501091293799387e-06, + "loss": 1.0307, + "step": 3399 + }, + { + "epoch": 0.3066239797988907, + "grad_norm": 1.4769305219638378, + "learning_rate": 3.24965305230186e-06, + "loss": 0.9894, + "step": 3400 + }, + { + "epoch": 0.306714163322361, + "grad_norm": 2.126076930300479, + "learning_rate": 3.249196868596492e-06, + "loss": 1.0504, + "step": 3401 + }, + { + "epoch": 0.30680434684583124, + "grad_norm": 1.722997791574472, + "learning_rate": 3.24874057830276e-06, + "loss": 0.9047, + "step": 3402 + }, + { + "epoch": 0.30689453036930153, + "grad_norm": 1.5118719609810736, + "learning_rate": 3.2482841814595954e-06, + "loss": 0.8958, + "step": 3403 + }, + { + "epoch": 0.30698471389277177, + "grad_norm": 1.7386488622215028, + "learning_rate": 3.247827678105943e-06, + "loss": 0.9203, + "step": 3404 + }, + { + "epoch": 0.30707489741624205, + "grad_norm": 1.5325843291238022, + "learning_rate": 3.247371068280751e-06, + "loss": 0.971, + "step": 3405 + }, + { + "epoch": 0.3071650809397123, + "grad_norm": 1.2301781205169222, + "learning_rate": 3.2469143520229823e-06, + "loss": 0.8326, + "step": 3406 + }, + { + "epoch": 0.3072552644631826, + "grad_norm": 1.509067862717008, + "learning_rate": 3.2464575293716054e-06, + "loss": 0.9298, + "step": 3407 + }, + { + "epoch": 0.30734544798665286, + "grad_norm": 1.5768859077889787, + "learning_rate": 3.2460006003655997e-06, + "loss": 0.9129, + "step": 3408 + }, + { + "epoch": 0.3074356315101231, + "grad_norm": 1.677639160296589, + "learning_rate": 3.245543565043952e-06, + "loss": 1.0339, + "step": 3409 + }, + { + "epoch": 0.3075258150335934, + "grad_norm": 1.7496302077732417, + "learning_rate": 3.2450864234456592e-06, + "loss": 0.9677, + "step": 3410 + }, + { + "epoch": 0.3076159985570636, + "grad_norm": 1.4108405315413073, + "learning_rate": 3.244629175609728e-06, + "loss": 0.9642, + "step": 3411 + }, + { + "epoch": 0.3077061820805339, + "grad_norm": 1.4581416503635798, + "learning_rate": 3.2441718215751726e-06, + "loss": 0.9568, + "step": 3412 + }, + { + "epoch": 0.30779636560400414, + "grad_norm": 1.7136386816819966, + "learning_rate": 3.2437143613810173e-06, + "loss": 0.9335, + "step": 3413 + }, + { + "epoch": 0.30788654912747443, + "grad_norm": 1.5089408329056786, + "learning_rate": 3.2432567950662947e-06, + "loss": 1.0145, + "step": 3414 + }, + { + "epoch": 0.30797673265094466, + "grad_norm": 1.4520473791384416, + "learning_rate": 3.2427991226700468e-06, + "loss": 0.9581, + "step": 3415 + }, + { + "epoch": 0.30806691617441495, + "grad_norm": 1.5209755591511107, + "learning_rate": 3.2423413442313246e-06, + "loss": 0.896, + "step": 3416 + }, + { + "epoch": 0.3081570996978852, + "grad_norm": 2.3413552829514446, + "learning_rate": 3.2418834597891904e-06, + "loss": 0.8825, + "step": 3417 + }, + { + "epoch": 0.3082472832213555, + "grad_norm": 1.7494217053359622, + "learning_rate": 3.2414254693827098e-06, + "loss": 1.0577, + "step": 3418 + }, + { + "epoch": 0.3083374667448257, + "grad_norm": 1.3843724562920823, + "learning_rate": 3.2409673730509644e-06, + "loss": 0.9523, + "step": 3419 + }, + { + "epoch": 0.308427650268296, + "grad_norm": 1.8167341695137549, + "learning_rate": 3.2405091708330393e-06, + "loss": 0.9485, + "step": 3420 + }, + { + "epoch": 0.3085178337917662, + "grad_norm": 1.8013056655531243, + "learning_rate": 3.2400508627680323e-06, + "loss": 0.9679, + "step": 3421 + }, + { + "epoch": 0.3086080173152365, + "grad_norm": 1.7034659088234798, + "learning_rate": 3.2395924488950474e-06, + "loss": 1.0649, + "step": 3422 + }, + { + "epoch": 0.30869820083870675, + "grad_norm": 1.8213841635903074, + "learning_rate": 3.2391339292532004e-06, + "loss": 1.0967, + "step": 3423 + }, + { + "epoch": 0.30878838436217704, + "grad_norm": 1.8021313949379416, + "learning_rate": 3.238675303881614e-06, + "loss": 0.9233, + "step": 3424 + }, + { + "epoch": 0.30887856788564727, + "grad_norm": 1.5518707368740685, + "learning_rate": 3.2382165728194203e-06, + "loss": 0.9392, + "step": 3425 + }, + { + "epoch": 0.30896875140911756, + "grad_norm": 2.038152732906096, + "learning_rate": 3.237757736105761e-06, + "loss": 0.9516, + "step": 3426 + }, + { + "epoch": 0.3090589349325878, + "grad_norm": 1.3463671282126504, + "learning_rate": 3.2372987937797867e-06, + "loss": 1.0516, + "step": 3427 + }, + { + "epoch": 0.3091491184560581, + "grad_norm": 1.4321512141865538, + "learning_rate": 3.2368397458806573e-06, + "loss": 0.8803, + "step": 3428 + }, + { + "epoch": 0.3092393019795283, + "grad_norm": 1.7106624713176284, + "learning_rate": 3.2363805924475412e-06, + "loss": 1.0071, + "step": 3429 + }, + { + "epoch": 0.3093294855029986, + "grad_norm": 2.194017890592951, + "learning_rate": 3.2359213335196153e-06, + "loss": 0.9149, + "step": 3430 + }, + { + "epoch": 0.3094196690264689, + "grad_norm": 1.5118597393677087, + "learning_rate": 3.2354619691360663e-06, + "loss": 0.9198, + "step": 3431 + }, + { + "epoch": 0.3095098525499391, + "grad_norm": 1.4380094206054423, + "learning_rate": 3.2350024993360898e-06, + "loss": 1.0085, + "step": 3432 + }, + { + "epoch": 0.3096000360734094, + "grad_norm": 1.5328370258573987, + "learning_rate": 3.2345429241588902e-06, + "loss": 0.9716, + "step": 3433 + }, + { + "epoch": 0.30969021959687965, + "grad_norm": 1.7140948978446726, + "learning_rate": 3.234083243643681e-06, + "loss": 0.9937, + "step": 3434 + }, + { + "epoch": 0.30978040312034993, + "grad_norm": 1.6167927135737472, + "learning_rate": 3.233623457829686e-06, + "loss": 1.0387, + "step": 3435 + }, + { + "epoch": 0.30987058664382017, + "grad_norm": 1.9981876745050615, + "learning_rate": 3.2331635667561344e-06, + "loss": 0.935, + "step": 3436 + }, + { + "epoch": 0.30996077016729046, + "grad_norm": 1.56366327851844, + "learning_rate": 3.2327035704622674e-06, + "loss": 0.9985, + "step": 3437 + }, + { + "epoch": 0.3100509536907607, + "grad_norm": 1.361904246772068, + "learning_rate": 3.2322434689873353e-06, + "loss": 0.8665, + "step": 3438 + }, + { + "epoch": 0.310141137214231, + "grad_norm": 1.552786196437724, + "learning_rate": 3.2317832623705957e-06, + "loss": 1.0542, + "step": 3439 + }, + { + "epoch": 0.3102313207377012, + "grad_norm": 1.6363735517769775, + "learning_rate": 3.231322950651316e-06, + "loss": 0.9133, + "step": 3440 + }, + { + "epoch": 0.3103215042611715, + "grad_norm": 1.243526480828015, + "learning_rate": 3.2308625338687735e-06, + "loss": 0.9978, + "step": 3441 + }, + { + "epoch": 0.31041168778464173, + "grad_norm": 2.2318305619985366, + "learning_rate": 3.230402012062252e-06, + "loss": 1.0832, + "step": 3442 + }, + { + "epoch": 0.310501871308112, + "grad_norm": 0.7541814904486044, + "learning_rate": 3.2299413852710466e-06, + "loss": 0.8157, + "step": 3443 + }, + { + "epoch": 0.31059205483158225, + "grad_norm": 1.819719930311013, + "learning_rate": 3.2294806535344606e-06, + "loss": 0.9282, + "step": 3444 + }, + { + "epoch": 0.31068223835505254, + "grad_norm": 1.7270745946327928, + "learning_rate": 3.2290198168918056e-06, + "loss": 1.0046, + "step": 3445 + }, + { + "epoch": 0.3107724218785228, + "grad_norm": 1.5775195178292027, + "learning_rate": 3.2285588753824035e-06, + "loss": 1.0031, + "step": 3446 + }, + { + "epoch": 0.31086260540199306, + "grad_norm": 1.5424960800689895, + "learning_rate": 3.228097829045584e-06, + "loss": 0.9835, + "step": 3447 + }, + { + "epoch": 0.3109527889254633, + "grad_norm": 1.4593018540016076, + "learning_rate": 3.227636677920685e-06, + "loss": 1.0105, + "step": 3448 + }, + { + "epoch": 0.3110429724489336, + "grad_norm": 1.54338466012412, + "learning_rate": 3.2271754220470567e-06, + "loss": 1.0158, + "step": 3449 + }, + { + "epoch": 0.3111331559724038, + "grad_norm": 1.6995295097929257, + "learning_rate": 3.2267140614640547e-06, + "loss": 1.0509, + "step": 3450 + }, + { + "epoch": 0.3112233394958741, + "grad_norm": 1.5891327638948798, + "learning_rate": 3.2262525962110445e-06, + "loss": 1.0149, + "step": 3451 + }, + { + "epoch": 0.31131352301934434, + "grad_norm": 1.7241757151146708, + "learning_rate": 3.2257910263274015e-06, + "loss": 0.8976, + "step": 3452 + }, + { + "epoch": 0.31140370654281463, + "grad_norm": 1.7301940699966825, + "learning_rate": 3.225329351852509e-06, + "loss": 0.9196, + "step": 3453 + }, + { + "epoch": 0.3114938900662849, + "grad_norm": 2.1152236670507474, + "learning_rate": 3.2248675728257596e-06, + "loss": 0.9761, + "step": 3454 + }, + { + "epoch": 0.31158407358975515, + "grad_norm": 1.5041685197700787, + "learning_rate": 3.2244056892865557e-06, + "loss": 1.0458, + "step": 3455 + }, + { + "epoch": 0.31167425711322544, + "grad_norm": 1.544391295016322, + "learning_rate": 3.2239437012743063e-06, + "loss": 0.9619, + "step": 3456 + }, + { + "epoch": 0.31176444063669567, + "grad_norm": 1.3770348923666484, + "learning_rate": 3.223481608828432e-06, + "loss": 0.9867, + "step": 3457 + }, + { + "epoch": 0.31185462416016596, + "grad_norm": 1.7735734122875189, + "learning_rate": 3.223019411988361e-06, + "loss": 0.9831, + "step": 3458 + }, + { + "epoch": 0.3119448076836362, + "grad_norm": 1.6604275829509962, + "learning_rate": 3.22255711079353e-06, + "loss": 0.9722, + "step": 3459 + }, + { + "epoch": 0.3120349912071065, + "grad_norm": 1.3666037484937006, + "learning_rate": 3.222094705283385e-06, + "loss": 0.9322, + "step": 3460 + }, + { + "epoch": 0.3121251747305767, + "grad_norm": 1.6945577602678261, + "learning_rate": 3.2216321954973805e-06, + "loss": 0.9674, + "step": 3461 + }, + { + "epoch": 0.312215358254047, + "grad_norm": 1.5957107450380328, + "learning_rate": 3.2211695814749816e-06, + "loss": 0.9603, + "step": 3462 + }, + { + "epoch": 0.31230554177751724, + "grad_norm": 1.9722988900280392, + "learning_rate": 3.220706863255661e-06, + "loss": 1.0746, + "step": 3463 + }, + { + "epoch": 0.3123957253009875, + "grad_norm": 1.6027042691970854, + "learning_rate": 3.2202440408788994e-06, + "loss": 0.9714, + "step": 3464 + }, + { + "epoch": 0.31248590882445776, + "grad_norm": 1.5722539486658824, + "learning_rate": 3.2197811143841883e-06, + "loss": 1.0429, + "step": 3465 + }, + { + "epoch": 0.31257609234792805, + "grad_norm": 1.3495609699848847, + "learning_rate": 3.2193180838110267e-06, + "loss": 0.9661, + "step": 3466 + }, + { + "epoch": 0.3126662758713983, + "grad_norm": 3.31851154627364, + "learning_rate": 3.2188549491989225e-06, + "loss": 0.8686, + "step": 3467 + }, + { + "epoch": 0.31275645939486857, + "grad_norm": 1.607209239073193, + "learning_rate": 3.2183917105873934e-06, + "loss": 0.9958, + "step": 3468 + }, + { + "epoch": 0.3128466429183388, + "grad_norm": 1.389453051215581, + "learning_rate": 3.217928368015966e-06, + "loss": 1.0292, + "step": 3469 + }, + { + "epoch": 0.3129368264418091, + "grad_norm": 1.3904250783478178, + "learning_rate": 3.217464921524174e-06, + "loss": 0.9995, + "step": 3470 + }, + { + "epoch": 0.3130270099652793, + "grad_norm": 1.5291938990663585, + "learning_rate": 3.2170013711515616e-06, + "loss": 0.9887, + "step": 3471 + }, + { + "epoch": 0.3131171934887496, + "grad_norm": 1.736704799443344, + "learning_rate": 3.216537716937682e-06, + "loss": 0.8966, + "step": 3472 + }, + { + "epoch": 0.31320737701221985, + "grad_norm": 1.8682498862318702, + "learning_rate": 3.2160739589220968e-06, + "loss": 0.9472, + "step": 3473 + }, + { + "epoch": 0.31329756053569013, + "grad_norm": 1.565031670929366, + "learning_rate": 3.215610097144376e-06, + "loss": 1.0296, + "step": 3474 + }, + { + "epoch": 0.31338774405916037, + "grad_norm": 1.5512543924593774, + "learning_rate": 3.215146131644099e-06, + "loss": 0.9168, + "step": 3475 + }, + { + "epoch": 0.31347792758263066, + "grad_norm": 1.503165164022304, + "learning_rate": 3.214682062460854e-06, + "loss": 1.0441, + "step": 3476 + }, + { + "epoch": 0.3135681111061009, + "grad_norm": 1.60232004997176, + "learning_rate": 3.2142178896342367e-06, + "loss": 0.886, + "step": 3477 + }, + { + "epoch": 0.3136582946295712, + "grad_norm": 1.4409029115402707, + "learning_rate": 3.2137536132038552e-06, + "loss": 0.9588, + "step": 3478 + }, + { + "epoch": 0.31374847815304147, + "grad_norm": 1.500450702349352, + "learning_rate": 3.2132892332093226e-06, + "loss": 0.9726, + "step": 3479 + }, + { + "epoch": 0.3138386616765117, + "grad_norm": 1.5384834090842403, + "learning_rate": 3.2128247496902623e-06, + "loss": 1.0833, + "step": 3480 + }, + { + "epoch": 0.313928845199982, + "grad_norm": 1.488194500655205, + "learning_rate": 3.2123601626863064e-06, + "loss": 0.9875, + "step": 3481 + }, + { + "epoch": 0.3140190287234522, + "grad_norm": 0.7120854677674648, + "learning_rate": 3.2118954722370974e-06, + "loss": 0.7731, + "step": 3482 + }, + { + "epoch": 0.3141092122469225, + "grad_norm": 1.569250858186928, + "learning_rate": 3.2114306783822837e-06, + "loss": 0.9172, + "step": 3483 + }, + { + "epoch": 0.31419939577039274, + "grad_norm": 1.5068450353895388, + "learning_rate": 3.210965781161525e-06, + "loss": 0.9504, + "step": 3484 + }, + { + "epoch": 0.31428957929386303, + "grad_norm": 1.49780295324596, + "learning_rate": 3.2105007806144892e-06, + "loss": 1.0841, + "step": 3485 + }, + { + "epoch": 0.31437976281733326, + "grad_norm": 1.9231648652436095, + "learning_rate": 3.2100356767808513e-06, + "loss": 0.976, + "step": 3486 + }, + { + "epoch": 0.31446994634080355, + "grad_norm": 1.6752006752165824, + "learning_rate": 3.2095704697002977e-06, + "loss": 1.0269, + "step": 3487 + }, + { + "epoch": 0.3145601298642738, + "grad_norm": 1.4458036670701104, + "learning_rate": 3.209105159412522e-06, + "loss": 0.9753, + "step": 3488 + }, + { + "epoch": 0.3146503133877441, + "grad_norm": 1.8206374897448232, + "learning_rate": 3.208639745957228e-06, + "loss": 0.9373, + "step": 3489 + }, + { + "epoch": 0.3147404969112143, + "grad_norm": 1.3118865532295911, + "learning_rate": 3.2081742293741256e-06, + "loss": 1.0, + "step": 3490 + }, + { + "epoch": 0.3148306804346846, + "grad_norm": 1.3921457611917125, + "learning_rate": 3.2077086097029366e-06, + "loss": 1.0353, + "step": 3491 + }, + { + "epoch": 0.31492086395815483, + "grad_norm": 1.4612532493744448, + "learning_rate": 3.2072428869833895e-06, + "loss": 0.972, + "step": 3492 + }, + { + "epoch": 0.3150110474816251, + "grad_norm": 2.9981275118939963, + "learning_rate": 3.206777061255223e-06, + "loss": 0.9844, + "step": 3493 + }, + { + "epoch": 0.31510123100509535, + "grad_norm": 1.7518967159039913, + "learning_rate": 3.206311132558183e-06, + "loss": 1.009, + "step": 3494 + }, + { + "epoch": 0.31519141452856564, + "grad_norm": 3.0511379057888552, + "learning_rate": 3.205845100932026e-06, + "loss": 0.9867, + "step": 3495 + }, + { + "epoch": 0.31528159805203587, + "grad_norm": 1.6673418425859292, + "learning_rate": 3.205378966416516e-06, + "loss": 1.0054, + "step": 3496 + }, + { + "epoch": 0.31537178157550616, + "grad_norm": 1.8946939614808822, + "learning_rate": 3.204912729051426e-06, + "loss": 0.9753, + "step": 3497 + }, + { + "epoch": 0.3154619650989764, + "grad_norm": 1.8709331913432201, + "learning_rate": 3.2044463888765384e-06, + "loss": 0.9464, + "step": 3498 + }, + { + "epoch": 0.3155521486224467, + "grad_norm": 1.8627754340590186, + "learning_rate": 3.2039799459316436e-06, + "loss": 0.8722, + "step": 3499 + }, + { + "epoch": 0.3156423321459169, + "grad_norm": 1.7591025810669918, + "learning_rate": 3.2035134002565407e-06, + "loss": 1.0088, + "step": 3500 + }, + { + "epoch": 0.3157325156693872, + "grad_norm": 1.642308924112217, + "learning_rate": 3.203046751891039e-06, + "loss": 0.9866, + "step": 3501 + }, + { + "epoch": 0.3158226991928575, + "grad_norm": 1.7584722340517374, + "learning_rate": 3.2025800008749545e-06, + "loss": 0.9885, + "step": 3502 + }, + { + "epoch": 0.3159128827163277, + "grad_norm": 1.584051956259066, + "learning_rate": 3.202113147248114e-06, + "loss": 0.9501, + "step": 3503 + }, + { + "epoch": 0.316003066239798, + "grad_norm": 1.9334089711071207, + "learning_rate": 3.20164619105035e-06, + "loss": 0.9455, + "step": 3504 + }, + { + "epoch": 0.31609324976326825, + "grad_norm": 1.5671627662661811, + "learning_rate": 3.201179132321508e-06, + "loss": 0.8259, + "step": 3505 + }, + { + "epoch": 0.31618343328673854, + "grad_norm": 1.5975425325322365, + "learning_rate": 3.200711971101439e-06, + "loss": 0.9124, + "step": 3506 + }, + { + "epoch": 0.31627361681020877, + "grad_norm": 1.919816255817328, + "learning_rate": 3.2002447074300047e-06, + "loss": 0.973, + "step": 3507 + }, + { + "epoch": 0.31636380033367906, + "grad_norm": 1.4797320382364911, + "learning_rate": 3.1997773413470736e-06, + "loss": 0.9787, + "step": 3508 + }, + { + "epoch": 0.3164539838571493, + "grad_norm": 1.9147091143125738, + "learning_rate": 3.199309872892524e-06, + "loss": 0.9681, + "step": 3509 + }, + { + "epoch": 0.3165441673806196, + "grad_norm": 1.776880358977393, + "learning_rate": 3.198842302106243e-06, + "loss": 1.0274, + "step": 3510 + }, + { + "epoch": 0.3166343509040898, + "grad_norm": 1.6616208785216948, + "learning_rate": 3.1983746290281265e-06, + "loss": 0.9576, + "step": 3511 + }, + { + "epoch": 0.3167245344275601, + "grad_norm": 1.3677353987251712, + "learning_rate": 3.197906853698079e-06, + "loss": 0.9791, + "step": 3512 + }, + { + "epoch": 0.31681471795103033, + "grad_norm": 1.5267475652430116, + "learning_rate": 3.1974389761560137e-06, + "loss": 0.9244, + "step": 3513 + }, + { + "epoch": 0.3169049014745006, + "grad_norm": 1.8185006485256898, + "learning_rate": 3.1969709964418525e-06, + "loss": 0.9264, + "step": 3514 + }, + { + "epoch": 0.31699508499797086, + "grad_norm": 1.7073950478063598, + "learning_rate": 3.196502914595525e-06, + "loss": 0.9966, + "step": 3515 + }, + { + "epoch": 0.31708526852144114, + "grad_norm": 1.663700906214072, + "learning_rate": 3.1960347306569723e-06, + "loss": 1.0005, + "step": 3516 + }, + { + "epoch": 0.3171754520449114, + "grad_norm": 1.7281031574434766, + "learning_rate": 3.195566444666141e-06, + "loss": 0.9237, + "step": 3517 + }, + { + "epoch": 0.31726563556838167, + "grad_norm": 1.4932160352864212, + "learning_rate": 3.1950980566629886e-06, + "loss": 1.0844, + "step": 3518 + }, + { + "epoch": 0.3173558190918519, + "grad_norm": 1.3446696926660364, + "learning_rate": 3.1946295666874797e-06, + "loss": 0.988, + "step": 3519 + }, + { + "epoch": 0.3174460026153222, + "grad_norm": 1.467871626503047, + "learning_rate": 3.19416097477959e-06, + "loss": 0.9938, + "step": 3520 + }, + { + "epoch": 0.3175361861387924, + "grad_norm": 1.4227086233001978, + "learning_rate": 3.1936922809793005e-06, + "loss": 1.0294, + "step": 3521 + }, + { + "epoch": 0.3176263696622627, + "grad_norm": 1.564049981950327, + "learning_rate": 3.193223485326604e-06, + "loss": 1.0956, + "step": 3522 + }, + { + "epoch": 0.31771655318573294, + "grad_norm": 1.7877410939490486, + "learning_rate": 3.1927545878615005e-06, + "loss": 1.0606, + "step": 3523 + }, + { + "epoch": 0.31780673670920323, + "grad_norm": 1.5982541035946012, + "learning_rate": 3.192285588623999e-06, + "loss": 0.9277, + "step": 3524 + }, + { + "epoch": 0.31789692023267346, + "grad_norm": 1.2303621760642476, + "learning_rate": 3.191816487654117e-06, + "loss": 0.9068, + "step": 3525 + }, + { + "epoch": 0.31798710375614375, + "grad_norm": 2.957908349113916, + "learning_rate": 3.19134728499188e-06, + "loss": 0.9439, + "step": 3526 + }, + { + "epoch": 0.31807728727961404, + "grad_norm": 1.675909148628525, + "learning_rate": 3.1908779806773235e-06, + "loss": 0.8625, + "step": 3527 + }, + { + "epoch": 0.3181674708030843, + "grad_norm": 1.411483393748124, + "learning_rate": 3.190408574750492e-06, + "loss": 1.0336, + "step": 3528 + }, + { + "epoch": 0.31825765432655456, + "grad_norm": 1.4483421121440248, + "learning_rate": 3.1899390672514367e-06, + "loss": 1.0458, + "step": 3529 + }, + { + "epoch": 0.3183478378500248, + "grad_norm": 1.519484962994005, + "learning_rate": 3.189469458220219e-06, + "loss": 0.8919, + "step": 3530 + }, + { + "epoch": 0.3184380213734951, + "grad_norm": 1.2558579511619552, + "learning_rate": 3.1889997476969086e-06, + "loss": 0.9731, + "step": 3531 + }, + { + "epoch": 0.3185282048969653, + "grad_norm": 1.4735968947536293, + "learning_rate": 3.188529935721583e-06, + "loss": 0.9664, + "step": 3532 + }, + { + "epoch": 0.3186183884204356, + "grad_norm": 1.5948300161636484, + "learning_rate": 3.18806002233433e-06, + "loss": 1.0359, + "step": 3533 + }, + { + "epoch": 0.31870857194390584, + "grad_norm": 1.4071729386415144, + "learning_rate": 3.187590007575245e-06, + "loss": 1.0387, + "step": 3534 + }, + { + "epoch": 0.3187987554673761, + "grad_norm": 1.7202095770125876, + "learning_rate": 3.1871198914844327e-06, + "loss": 0.9122, + "step": 3535 + }, + { + "epoch": 0.31888893899084636, + "grad_norm": 1.7037486763023333, + "learning_rate": 3.1866496741020057e-06, + "loss": 0.9113, + "step": 3536 + }, + { + "epoch": 0.31897912251431665, + "grad_norm": 1.219998380550498, + "learning_rate": 3.186179355468085e-06, + "loss": 0.944, + "step": 3537 + }, + { + "epoch": 0.3190693060377869, + "grad_norm": 1.5219706197852914, + "learning_rate": 3.1857089356228015e-06, + "loss": 0.9732, + "step": 3538 + }, + { + "epoch": 0.31915948956125717, + "grad_norm": 1.3978208322359837, + "learning_rate": 3.1852384146062933e-06, + "loss": 0.9188, + "step": 3539 + }, + { + "epoch": 0.3192496730847274, + "grad_norm": 0.733326386469391, + "learning_rate": 3.184767792458708e-06, + "loss": 0.7594, + "step": 3540 + }, + { + "epoch": 0.3193398566081977, + "grad_norm": 1.5559490970042245, + "learning_rate": 3.1842970692202023e-06, + "loss": 1.1047, + "step": 3541 + }, + { + "epoch": 0.3194300401316679, + "grad_norm": 1.6524793059745693, + "learning_rate": 3.1838262449309403e-06, + "loss": 0.9546, + "step": 3542 + }, + { + "epoch": 0.3195202236551382, + "grad_norm": 1.598126852775159, + "learning_rate": 3.1833553196310956e-06, + "loss": 0.894, + "step": 3543 + }, + { + "epoch": 0.31961040717860845, + "grad_norm": 1.4935109924892682, + "learning_rate": 3.18288429336085e-06, + "loss": 0.8354, + "step": 3544 + }, + { + "epoch": 0.31970059070207874, + "grad_norm": 1.6371135372108045, + "learning_rate": 3.182413166160394e-06, + "loss": 1.0831, + "step": 3545 + }, + { + "epoch": 0.31979077422554897, + "grad_norm": 1.6268162480875041, + "learning_rate": 3.1819419380699275e-06, + "loss": 1.0097, + "step": 3546 + }, + { + "epoch": 0.31988095774901926, + "grad_norm": 1.4047277052305494, + "learning_rate": 3.181470609129658e-06, + "loss": 0.9835, + "step": 3547 + }, + { + "epoch": 0.3199711412724895, + "grad_norm": 2.057664578276332, + "learning_rate": 3.1809991793798e-06, + "loss": 1.0262, + "step": 3548 + }, + { + "epoch": 0.3200613247959598, + "grad_norm": 1.614810025287924, + "learning_rate": 3.1805276488605806e-06, + "loss": 0.9328, + "step": 3549 + }, + { + "epoch": 0.32015150831943007, + "grad_norm": 1.4615856507633112, + "learning_rate": 3.1800560176122336e-06, + "loss": 0.9732, + "step": 3550 + }, + { + "epoch": 0.3202416918429003, + "grad_norm": 1.542545463349539, + "learning_rate": 3.179584285675e-06, + "loss": 0.9822, + "step": 3551 + }, + { + "epoch": 0.3203318753663706, + "grad_norm": 1.632366794917556, + "learning_rate": 3.1791124530891315e-06, + "loss": 0.8896, + "step": 3552 + }, + { + "epoch": 0.3204220588898408, + "grad_norm": 2.0744573595984064, + "learning_rate": 3.178640519894886e-06, + "loss": 0.9763, + "step": 3553 + }, + { + "epoch": 0.3205122424133111, + "grad_norm": 1.479925775728722, + "learning_rate": 3.1781684861325324e-06, + "loss": 0.9296, + "step": 3554 + }, + { + "epoch": 0.32060242593678134, + "grad_norm": 1.5335856938356998, + "learning_rate": 3.177696351842348e-06, + "loss": 0.9756, + "step": 3555 + }, + { + "epoch": 0.32069260946025163, + "grad_norm": 0.8243402689803845, + "learning_rate": 3.1772241170646167e-06, + "loss": 0.7615, + "step": 3556 + }, + { + "epoch": 0.32078279298372187, + "grad_norm": 1.7416838187265973, + "learning_rate": 3.1767517818396334e-06, + "loss": 1.0741, + "step": 3557 + }, + { + "epoch": 0.32087297650719215, + "grad_norm": 1.44839380027584, + "learning_rate": 3.1762793462076986e-06, + "loss": 1.1304, + "step": 3558 + }, + { + "epoch": 0.3209631600306624, + "grad_norm": 1.9407437151534757, + "learning_rate": 3.1758068102091236e-06, + "loss": 0.8575, + "step": 3559 + }, + { + "epoch": 0.3210533435541327, + "grad_norm": 1.6263188365585457, + "learning_rate": 3.175334173884229e-06, + "loss": 0.9413, + "step": 3560 + }, + { + "epoch": 0.3211435270776029, + "grad_norm": 1.2534722263316223, + "learning_rate": 3.174861437273342e-06, + "loss": 0.9705, + "step": 3561 + }, + { + "epoch": 0.3212337106010732, + "grad_norm": 1.5355211751180249, + "learning_rate": 3.174388600416799e-06, + "loss": 0.9325, + "step": 3562 + }, + { + "epoch": 0.32132389412454343, + "grad_norm": 0.7098139593685941, + "learning_rate": 3.1739156633549445e-06, + "loss": 0.7718, + "step": 3563 + }, + { + "epoch": 0.3214140776480137, + "grad_norm": 1.5127671029557175, + "learning_rate": 3.173442626128133e-06, + "loss": 1.0414, + "step": 3564 + }, + { + "epoch": 0.32150426117148395, + "grad_norm": 2.0399888079467132, + "learning_rate": 3.1729694887767265e-06, + "loss": 1.0923, + "step": 3565 + }, + { + "epoch": 0.32159444469495424, + "grad_norm": 1.7610251674814104, + "learning_rate": 3.172496251341096e-06, + "loss": 0.9732, + "step": 3566 + }, + { + "epoch": 0.3216846282184245, + "grad_norm": 1.6749424824874177, + "learning_rate": 3.172022913861619e-06, + "loss": 0.828, + "step": 3567 + }, + { + "epoch": 0.32177481174189476, + "grad_norm": 1.7063250794153633, + "learning_rate": 3.171549476378686e-06, + "loss": 0.9342, + "step": 3568 + }, + { + "epoch": 0.321864995265365, + "grad_norm": 1.5960144710410558, + "learning_rate": 3.1710759389326906e-06, + "loss": 0.9426, + "step": 3569 + }, + { + "epoch": 0.3219551787888353, + "grad_norm": 1.8328997214816658, + "learning_rate": 3.1706023015640396e-06, + "loss": 0.9675, + "step": 3570 + }, + { + "epoch": 0.3220453623123055, + "grad_norm": 1.625057439522504, + "learning_rate": 3.1701285643131453e-06, + "loss": 0.9966, + "step": 3571 + }, + { + "epoch": 0.3221355458357758, + "grad_norm": 1.8538803851111334, + "learning_rate": 3.16965472722043e-06, + "loss": 0.9775, + "step": 3572 + }, + { + "epoch": 0.32222572935924604, + "grad_norm": 1.9667019560713546, + "learning_rate": 3.169180790326324e-06, + "loss": 0.8949, + "step": 3573 + }, + { + "epoch": 0.3223159128827163, + "grad_norm": 1.5205731191067158, + "learning_rate": 3.168706753671266e-06, + "loss": 0.9758, + "step": 3574 + }, + { + "epoch": 0.3224060964061866, + "grad_norm": 1.347678673599113, + "learning_rate": 3.168232617295704e-06, + "loss": 0.8987, + "step": 3575 + }, + { + "epoch": 0.32249627992965685, + "grad_norm": 1.7163848556136692, + "learning_rate": 3.167758381240093e-06, + "loss": 0.9671, + "step": 3576 + }, + { + "epoch": 0.32258646345312714, + "grad_norm": 1.583385500132172, + "learning_rate": 3.1672840455448978e-06, + "loss": 1.0177, + "step": 3577 + }, + { + "epoch": 0.32267664697659737, + "grad_norm": 1.708081374157393, + "learning_rate": 3.166809610250592e-06, + "loss": 0.9123, + "step": 3578 + }, + { + "epoch": 0.32276683050006766, + "grad_norm": 1.8114592096093727, + "learning_rate": 3.166335075397656e-06, + "loss": 0.9022, + "step": 3579 + }, + { + "epoch": 0.3228570140235379, + "grad_norm": 1.7510465489882423, + "learning_rate": 3.1658604410265808e-06, + "loss": 0.9366, + "step": 3580 + }, + { + "epoch": 0.3229471975470082, + "grad_norm": 0.8989716310427417, + "learning_rate": 3.1653857071778644e-06, + "loss": 0.8016, + "step": 3581 + }, + { + "epoch": 0.3230373810704784, + "grad_norm": 1.7307276822670299, + "learning_rate": 3.1649108738920133e-06, + "loss": 0.9623, + "step": 3582 + }, + { + "epoch": 0.3231275645939487, + "grad_norm": 1.6024631107982226, + "learning_rate": 3.1644359412095432e-06, + "loss": 0.9766, + "step": 3583 + }, + { + "epoch": 0.32321774811741893, + "grad_norm": 0.8380391093826277, + "learning_rate": 3.163960909170978e-06, + "loss": 0.8118, + "step": 3584 + }, + { + "epoch": 0.3233079316408892, + "grad_norm": 1.4487679903824024, + "learning_rate": 3.1634857778168496e-06, + "loss": 0.8706, + "step": 3585 + }, + { + "epoch": 0.32339811516435946, + "grad_norm": 1.5498877576980865, + "learning_rate": 3.1630105471877002e-06, + "loss": 1.0105, + "step": 3586 + }, + { + "epoch": 0.32348829868782975, + "grad_norm": 1.8169749692453487, + "learning_rate": 3.162535217324077e-06, + "loss": 0.9542, + "step": 3587 + }, + { + "epoch": 0.3235784822113, + "grad_norm": 1.9567216639720553, + "learning_rate": 3.1620597882665393e-06, + "loss": 1.0041, + "step": 3588 + }, + { + "epoch": 0.32366866573477027, + "grad_norm": 1.6722547866266892, + "learning_rate": 3.1615842600556535e-06, + "loss": 1.0301, + "step": 3589 + }, + { + "epoch": 0.3237588492582405, + "grad_norm": 1.5082308332204508, + "learning_rate": 3.1611086327319932e-06, + "loss": 0.9697, + "step": 3590 + }, + { + "epoch": 0.3238490327817108, + "grad_norm": 1.545894736058424, + "learning_rate": 3.160632906336142e-06, + "loss": 1.0452, + "step": 3591 + }, + { + "epoch": 0.323939216305181, + "grad_norm": 1.8294376323210277, + "learning_rate": 3.160157080908692e-06, + "loss": 0.9977, + "step": 3592 + }, + { + "epoch": 0.3240293998286513, + "grad_norm": 1.3758896637138014, + "learning_rate": 3.1596811564902426e-06, + "loss": 1.03, + "step": 3593 + }, + { + "epoch": 0.32411958335212154, + "grad_norm": 1.8911924535211886, + "learning_rate": 3.1592051331214023e-06, + "loss": 0.9146, + "step": 3594 + }, + { + "epoch": 0.32420976687559183, + "grad_norm": 1.3823460368675098, + "learning_rate": 3.158729010842789e-06, + "loss": 0.9961, + "step": 3595 + }, + { + "epoch": 0.32429995039906206, + "grad_norm": 1.432386258787553, + "learning_rate": 3.1582527896950266e-06, + "loss": 1.0267, + "step": 3596 + }, + { + "epoch": 0.32439013392253235, + "grad_norm": 1.750184321914423, + "learning_rate": 3.157776469718749e-06, + "loss": 0.9916, + "step": 3597 + }, + { + "epoch": 0.32448031744600264, + "grad_norm": 1.6802597135794588, + "learning_rate": 3.1573000509546004e-06, + "loss": 0.965, + "step": 3598 + }, + { + "epoch": 0.3245705009694729, + "grad_norm": 1.5641542451609525, + "learning_rate": 3.1568235334432296e-06, + "loss": 1.027, + "step": 3599 + }, + { + "epoch": 0.32466068449294316, + "grad_norm": 1.746802951916995, + "learning_rate": 3.1563469172252964e-06, + "loss": 1.0389, + "step": 3600 + }, + { + "epoch": 0.3247508680164134, + "grad_norm": 2.183940934916086, + "learning_rate": 3.155870202341468e-06, + "loss": 1.0479, + "step": 3601 + }, + { + "epoch": 0.3248410515398837, + "grad_norm": 1.6552169024960057, + "learning_rate": 3.155393388832421e-06, + "loss": 0.9674, + "step": 3602 + }, + { + "epoch": 0.3249312350633539, + "grad_norm": 1.763964859969274, + "learning_rate": 3.1549164767388386e-06, + "loss": 1.0002, + "step": 3603 + }, + { + "epoch": 0.3250214185868242, + "grad_norm": 1.6221710669857836, + "learning_rate": 3.1544394661014145e-06, + "loss": 0.9792, + "step": 3604 + }, + { + "epoch": 0.32511160211029444, + "grad_norm": 1.6018122780643538, + "learning_rate": 3.15396235696085e-06, + "loss": 0.8907, + "step": 3605 + }, + { + "epoch": 0.32520178563376473, + "grad_norm": 1.3791772073686748, + "learning_rate": 3.153485149357854e-06, + "loss": 0.965, + "step": 3606 + }, + { + "epoch": 0.32529196915723496, + "grad_norm": 1.468023566927384, + "learning_rate": 3.153007843333145e-06, + "loss": 0.9604, + "step": 3607 + }, + { + "epoch": 0.32538215268070525, + "grad_norm": 1.6296552587229378, + "learning_rate": 3.152530438927449e-06, + "loss": 1.0016, + "step": 3608 + }, + { + "epoch": 0.3254723362041755, + "grad_norm": 1.5812008032100229, + "learning_rate": 3.1520529361815008e-06, + "loss": 1.0158, + "step": 3609 + }, + { + "epoch": 0.32556251972764577, + "grad_norm": 1.5405202245774487, + "learning_rate": 3.151575335136044e-06, + "loss": 1.0135, + "step": 3610 + }, + { + "epoch": 0.325652703251116, + "grad_norm": 1.7307198301421833, + "learning_rate": 3.1510976358318298e-06, + "loss": 0.9575, + "step": 3611 + }, + { + "epoch": 0.3257428867745863, + "grad_norm": 1.613435530039774, + "learning_rate": 3.1506198383096186e-06, + "loss": 0.9264, + "step": 3612 + }, + { + "epoch": 0.3258330702980565, + "grad_norm": 1.628860216839618, + "learning_rate": 3.150141942610178e-06, + "loss": 0.9901, + "step": 3613 + }, + { + "epoch": 0.3259232538215268, + "grad_norm": 0.7814958567006112, + "learning_rate": 3.1496639487742853e-06, + "loss": 0.7893, + "step": 3614 + }, + { + "epoch": 0.32601343734499705, + "grad_norm": 1.5066914714751871, + "learning_rate": 3.1491858568427247e-06, + "loss": 1.0217, + "step": 3615 + }, + { + "epoch": 0.32610362086846734, + "grad_norm": 1.5847106848848493, + "learning_rate": 3.1487076668562903e-06, + "loss": 1.0267, + "step": 3616 + }, + { + "epoch": 0.32619380439193757, + "grad_norm": 2.0231921662172554, + "learning_rate": 3.1482293788557847e-06, + "loss": 1.0196, + "step": 3617 + }, + { + "epoch": 0.32628398791540786, + "grad_norm": 1.5816782616423768, + "learning_rate": 3.1477509928820165e-06, + "loss": 0.9414, + "step": 3618 + }, + { + "epoch": 0.3263741714388781, + "grad_norm": 1.7960324509668937, + "learning_rate": 3.147272508975805e-06, + "loss": 0.9732, + "step": 3619 + }, + { + "epoch": 0.3264643549623484, + "grad_norm": 0.7719712664539198, + "learning_rate": 3.1467939271779775e-06, + "loss": 0.8081, + "step": 3620 + }, + { + "epoch": 0.32655453848581867, + "grad_norm": 1.441154395949892, + "learning_rate": 3.146315247529368e-06, + "loss": 0.9438, + "step": 3621 + }, + { + "epoch": 0.3266447220092889, + "grad_norm": 2.0236336749210957, + "learning_rate": 3.1458364700708212e-06, + "loss": 0.9379, + "step": 3622 + }, + { + "epoch": 0.3267349055327592, + "grad_norm": 1.5120881489333395, + "learning_rate": 3.1453575948431892e-06, + "loss": 1.0072, + "step": 3623 + }, + { + "epoch": 0.3268250890562294, + "grad_norm": 1.6999460352017846, + "learning_rate": 3.144878621887331e-06, + "loss": 0.9849, + "step": 3624 + }, + { + "epoch": 0.3269152725796997, + "grad_norm": 1.608448020851734, + "learning_rate": 3.1443995512441167e-06, + "loss": 1.0171, + "step": 3625 + }, + { + "epoch": 0.32700545610316994, + "grad_norm": 1.4809040647468574, + "learning_rate": 3.1439203829544224e-06, + "loss": 1.0461, + "step": 3626 + }, + { + "epoch": 0.32709563962664023, + "grad_norm": 1.3513059427712635, + "learning_rate": 3.143441117059133e-06, + "loss": 1.0546, + "step": 3627 + }, + { + "epoch": 0.32718582315011047, + "grad_norm": 1.5325499485706466, + "learning_rate": 3.142961753599143e-06, + "loss": 0.9222, + "step": 3628 + }, + { + "epoch": 0.32727600667358075, + "grad_norm": 1.4687238650330188, + "learning_rate": 3.1424822926153543e-06, + "loss": 0.9429, + "step": 3629 + }, + { + "epoch": 0.327366190197051, + "grad_norm": 1.585386509385019, + "learning_rate": 3.142002734148676e-06, + "loss": 1.0238, + "step": 3630 + }, + { + "epoch": 0.3274563737205213, + "grad_norm": 1.7839077980908187, + "learning_rate": 3.141523078240028e-06, + "loss": 1.0062, + "step": 3631 + }, + { + "epoch": 0.3275465572439915, + "grad_norm": 1.6715567455239053, + "learning_rate": 3.1410433249303366e-06, + "loss": 1.0333, + "step": 3632 + }, + { + "epoch": 0.3276367407674618, + "grad_norm": 1.5552581813188067, + "learning_rate": 3.1405634742605366e-06, + "loss": 1.069, + "step": 3633 + }, + { + "epoch": 0.32772692429093203, + "grad_norm": 1.6742444939354055, + "learning_rate": 3.1400835262715727e-06, + "loss": 1.021, + "step": 3634 + }, + { + "epoch": 0.3278171078144023, + "grad_norm": 1.700297817621848, + "learning_rate": 3.139603481004396e-06, + "loss": 0.9658, + "step": 3635 + }, + { + "epoch": 0.32790729133787255, + "grad_norm": 1.386626463492204, + "learning_rate": 3.139123338499966e-06, + "loss": 0.9926, + "step": 3636 + }, + { + "epoch": 0.32799747486134284, + "grad_norm": 1.7792884332349825, + "learning_rate": 3.1386430987992524e-06, + "loss": 1.0184, + "step": 3637 + }, + { + "epoch": 0.3280876583848131, + "grad_norm": 1.924108866234685, + "learning_rate": 3.1381627619432307e-06, + "loss": 0.9239, + "step": 3638 + }, + { + "epoch": 0.32817784190828336, + "grad_norm": 1.931664693314096, + "learning_rate": 3.1376823279728864e-06, + "loss": 0.985, + "step": 3639 + }, + { + "epoch": 0.3282680254317536, + "grad_norm": 1.5606973549260266, + "learning_rate": 3.1372017969292125e-06, + "loss": 1.0613, + "step": 3640 + }, + { + "epoch": 0.3283582089552239, + "grad_norm": 1.4634332401752077, + "learning_rate": 3.136721168853211e-06, + "loss": 0.9579, + "step": 3641 + }, + { + "epoch": 0.3284483924786941, + "grad_norm": 1.9142525092032892, + "learning_rate": 3.1362404437858924e-06, + "loss": 0.9426, + "step": 3642 + }, + { + "epoch": 0.3285385760021644, + "grad_norm": 1.5298220145725292, + "learning_rate": 3.135759621768273e-06, + "loss": 1.0322, + "step": 3643 + }, + { + "epoch": 0.32862875952563464, + "grad_norm": 1.4959390027365578, + "learning_rate": 3.13527870284138e-06, + "loss": 1.0299, + "step": 3644 + }, + { + "epoch": 0.32871894304910493, + "grad_norm": 1.8717925293957678, + "learning_rate": 3.134797687046249e-06, + "loss": 1.0401, + "step": 3645 + }, + { + "epoch": 0.3288091265725752, + "grad_norm": 1.9936380528319233, + "learning_rate": 3.1343165744239218e-06, + "loss": 0.9127, + "step": 3646 + }, + { + "epoch": 0.32889931009604545, + "grad_norm": 2.9944831667134113, + "learning_rate": 3.13383536501545e-06, + "loss": 1.0006, + "step": 3647 + }, + { + "epoch": 0.32898949361951574, + "grad_norm": 1.3820761320202344, + "learning_rate": 3.133354058861893e-06, + "loss": 0.9996, + "step": 3648 + }, + { + "epoch": 0.32907967714298597, + "grad_norm": 1.6079656856063957, + "learning_rate": 3.132872656004318e-06, + "loss": 0.9505, + "step": 3649 + }, + { + "epoch": 0.32916986066645626, + "grad_norm": 1.6033852230301353, + "learning_rate": 3.132391156483802e-06, + "loss": 0.8762, + "step": 3650 + }, + { + "epoch": 0.3292600441899265, + "grad_norm": 1.6828040729336498, + "learning_rate": 3.131909560341428e-06, + "loss": 0.939, + "step": 3651 + }, + { + "epoch": 0.3293502277133968, + "grad_norm": 1.5189424369307036, + "learning_rate": 3.1314278676182893e-06, + "loss": 0.9275, + "step": 3652 + }, + { + "epoch": 0.329440411236867, + "grad_norm": 1.431239718604328, + "learning_rate": 3.130946078355486e-06, + "loss": 1.0478, + "step": 3653 + }, + { + "epoch": 0.3295305947603373, + "grad_norm": 1.561708097054204, + "learning_rate": 3.130464192594128e-06, + "loss": 1.0106, + "step": 3654 + }, + { + "epoch": 0.32962077828380754, + "grad_norm": 1.773498870191061, + "learning_rate": 3.1299822103753315e-06, + "loss": 0.9079, + "step": 3655 + }, + { + "epoch": 0.3297109618072778, + "grad_norm": 1.5887878810829499, + "learning_rate": 3.1295001317402217e-06, + "loss": 0.9794, + "step": 3656 + }, + { + "epoch": 0.32980114533074806, + "grad_norm": 2.5037949369209778, + "learning_rate": 3.1290179567299335e-06, + "loss": 0.9264, + "step": 3657 + }, + { + "epoch": 0.32989132885421835, + "grad_norm": 1.4164088051925983, + "learning_rate": 3.128535685385607e-06, + "loss": 1.0096, + "step": 3658 + }, + { + "epoch": 0.3299815123776886, + "grad_norm": 1.4744681385435903, + "learning_rate": 3.1280533177483935e-06, + "loss": 1.0613, + "step": 3659 + }, + { + "epoch": 0.33007169590115887, + "grad_norm": 1.435836824991088, + "learning_rate": 3.127570853859451e-06, + "loss": 0.9926, + "step": 3660 + }, + { + "epoch": 0.3301618794246291, + "grad_norm": 1.5156346743560036, + "learning_rate": 3.1270882937599456e-06, + "loss": 0.8664, + "step": 3661 + }, + { + "epoch": 0.3302520629480994, + "grad_norm": 1.6294130410573986, + "learning_rate": 3.1266056374910532e-06, + "loss": 0.9994, + "step": 3662 + }, + { + "epoch": 0.3303422464715696, + "grad_norm": 1.618005665386075, + "learning_rate": 3.126122885093955e-06, + "loss": 0.8768, + "step": 3663 + }, + { + "epoch": 0.3304324299950399, + "grad_norm": 2.563639131749895, + "learning_rate": 3.1256400366098427e-06, + "loss": 0.9725, + "step": 3664 + }, + { + "epoch": 0.33052261351851014, + "grad_norm": 1.7439562111615574, + "learning_rate": 3.125157092079916e-06, + "loss": 0.9356, + "step": 3665 + }, + { + "epoch": 0.33061279704198043, + "grad_norm": 1.7723672468996627, + "learning_rate": 3.1246740515453824e-06, + "loss": 0.9481, + "step": 3666 + }, + { + "epoch": 0.33070298056545067, + "grad_norm": 1.5627355779440952, + "learning_rate": 3.124190915047457e-06, + "loss": 0.9162, + "step": 3667 + }, + { + "epoch": 0.33079316408892095, + "grad_norm": 1.867917161689447, + "learning_rate": 3.123707682627364e-06, + "loss": 0.9494, + "step": 3668 + }, + { + "epoch": 0.33088334761239124, + "grad_norm": 1.7340567356934316, + "learning_rate": 3.1232243543263356e-06, + "loss": 0.9625, + "step": 3669 + }, + { + "epoch": 0.3309735311358615, + "grad_norm": 1.6145050357989341, + "learning_rate": 3.1227409301856122e-06, + "loss": 0.8795, + "step": 3670 + }, + { + "epoch": 0.33106371465933176, + "grad_norm": 1.4778290014702884, + "learning_rate": 3.1222574102464413e-06, + "loss": 0.9617, + "step": 3671 + }, + { + "epoch": 0.331153898182802, + "grad_norm": 1.4431691541232163, + "learning_rate": 3.12177379455008e-06, + "loss": 1.007, + "step": 3672 + }, + { + "epoch": 0.3312440817062723, + "grad_norm": 1.578195475902558, + "learning_rate": 3.121290083137794e-06, + "loss": 0.9744, + "step": 3673 + }, + { + "epoch": 0.3313342652297425, + "grad_norm": 1.6280057691382837, + "learning_rate": 3.1208062760508547e-06, + "loss": 0.949, + "step": 3674 + }, + { + "epoch": 0.3314244487532128, + "grad_norm": 1.371615536077847, + "learning_rate": 3.1203223733305438e-06, + "loss": 0.9132, + "step": 3675 + }, + { + "epoch": 0.33151463227668304, + "grad_norm": 0.8039372103617254, + "learning_rate": 3.1198383750181512e-06, + "loss": 0.7659, + "step": 3676 + }, + { + "epoch": 0.33160481580015333, + "grad_norm": 1.65391455041445, + "learning_rate": 3.1193542811549734e-06, + "loss": 0.8988, + "step": 3677 + }, + { + "epoch": 0.33169499932362356, + "grad_norm": 1.5267002476964915, + "learning_rate": 3.1188700917823166e-06, + "loss": 0.8999, + "step": 3678 + }, + { + "epoch": 0.33178518284709385, + "grad_norm": 0.8262732693833676, + "learning_rate": 3.1183858069414937e-06, + "loss": 0.8286, + "step": 3679 + }, + { + "epoch": 0.3318753663705641, + "grad_norm": 3.118837307183823, + "learning_rate": 3.117901426673827e-06, + "loss": 1.0238, + "step": 3680 + }, + { + "epoch": 0.3319655498940344, + "grad_norm": 1.699177287681104, + "learning_rate": 3.1174169510206466e-06, + "loss": 0.935, + "step": 3681 + }, + { + "epoch": 0.3320557334175046, + "grad_norm": 1.4595095750077094, + "learning_rate": 3.1169323800232908e-06, + "loss": 0.9722, + "step": 3682 + }, + { + "epoch": 0.3321459169409749, + "grad_norm": 1.7755712369025562, + "learning_rate": 3.1164477137231054e-06, + "loss": 1.0048, + "step": 3683 + }, + { + "epoch": 0.3322361004644451, + "grad_norm": 1.7261167615425526, + "learning_rate": 3.115962952161445e-06, + "loss": 0.9785, + "step": 3684 + }, + { + "epoch": 0.3323262839879154, + "grad_norm": 2.3278086402544584, + "learning_rate": 3.1154780953796727e-06, + "loss": 1.0153, + "step": 3685 + }, + { + "epoch": 0.33241646751138565, + "grad_norm": 1.548496359929557, + "learning_rate": 3.114993143419158e-06, + "loss": 0.8801, + "step": 3686 + }, + { + "epoch": 0.33250665103485594, + "grad_norm": 1.323524000275214, + "learning_rate": 3.1145080963212806e-06, + "loss": 0.8859, + "step": 3687 + }, + { + "epoch": 0.33259683455832617, + "grad_norm": 1.3541656738668861, + "learning_rate": 3.114022954127427e-06, + "loss": 0.9765, + "step": 3688 + }, + { + "epoch": 0.33268701808179646, + "grad_norm": 1.48080376128446, + "learning_rate": 3.1135377168789923e-06, + "loss": 1.1819, + "step": 3689 + }, + { + "epoch": 0.3327772016052667, + "grad_norm": 1.4672601326381167, + "learning_rate": 3.1130523846173803e-06, + "loss": 1.0424, + "step": 3690 + }, + { + "epoch": 0.332867385128737, + "grad_norm": 1.7924263326917598, + "learning_rate": 3.1125669573840006e-06, + "loss": 0.919, + "step": 3691 + }, + { + "epoch": 0.3329575686522072, + "grad_norm": 1.8130473921694494, + "learning_rate": 3.112081435220274e-06, + "loss": 0.9166, + "step": 3692 + }, + { + "epoch": 0.3330477521756775, + "grad_norm": 1.5548300366261258, + "learning_rate": 3.111595818167627e-06, + "loss": 0.9635, + "step": 3693 + }, + { + "epoch": 0.3331379356991478, + "grad_norm": 1.6500993640920394, + "learning_rate": 3.1111101062674953e-06, + "loss": 1.0318, + "step": 3694 + }, + { + "epoch": 0.333228119222618, + "grad_norm": 1.5610369126563859, + "learning_rate": 3.1106242995613233e-06, + "loss": 0.9643, + "step": 3695 + }, + { + "epoch": 0.3333183027460883, + "grad_norm": 1.7314353292815337, + "learning_rate": 3.1101383980905616e-06, + "loss": 0.9704, + "step": 3696 + }, + { + "epoch": 0.33340848626955855, + "grad_norm": 0.8144160932359779, + "learning_rate": 3.109652401896671e-06, + "loss": 0.7535, + "step": 3697 + }, + { + "epoch": 0.33349866979302883, + "grad_norm": 1.4160518246989247, + "learning_rate": 3.109166311021119e-06, + "loss": 1.0182, + "step": 3698 + }, + { + "epoch": 0.33358885331649907, + "grad_norm": 1.8002772091288781, + "learning_rate": 3.1086801255053807e-06, + "loss": 1.0087, + "step": 3699 + }, + { + "epoch": 0.33367903683996936, + "grad_norm": 1.6080891184949455, + "learning_rate": 3.108193845390942e-06, + "loss": 1.0129, + "step": 3700 + }, + { + "epoch": 0.3337692203634396, + "grad_norm": 1.2740838686377027, + "learning_rate": 3.1077074707192933e-06, + "loss": 0.9515, + "step": 3701 + }, + { + "epoch": 0.3338594038869099, + "grad_norm": 1.4016446892190542, + "learning_rate": 3.1072210015319353e-06, + "loss": 0.993, + "step": 3702 + }, + { + "epoch": 0.3339495874103801, + "grad_norm": 1.5742469955447167, + "learning_rate": 3.106734437870376e-06, + "loss": 0.9319, + "step": 3703 + }, + { + "epoch": 0.3340397709338504, + "grad_norm": 1.6667386595753757, + "learning_rate": 3.1062477797761327e-06, + "loss": 0.9116, + "step": 3704 + }, + { + "epoch": 0.33412995445732063, + "grad_norm": 1.6577538194976238, + "learning_rate": 3.105761027290729e-06, + "loss": 1.0214, + "step": 3705 + }, + { + "epoch": 0.3342201379807909, + "grad_norm": 1.428725372262393, + "learning_rate": 3.105274180455697e-06, + "loss": 0.9442, + "step": 3706 + }, + { + "epoch": 0.33431032150426115, + "grad_norm": 1.3547641120903517, + "learning_rate": 3.1047872393125775e-06, + "loss": 0.9985, + "step": 3707 + }, + { + "epoch": 0.33440050502773144, + "grad_norm": 0.7302465253810632, + "learning_rate": 3.1043002039029186e-06, + "loss": 0.7809, + "step": 3708 + }, + { + "epoch": 0.3344906885512017, + "grad_norm": 1.3820873449660465, + "learning_rate": 3.1038130742682782e-06, + "loss": 0.9686, + "step": 3709 + }, + { + "epoch": 0.33458087207467196, + "grad_norm": 1.4219072778412642, + "learning_rate": 3.103325850450219e-06, + "loss": 0.9516, + "step": 3710 + }, + { + "epoch": 0.3346710555981422, + "grad_norm": 1.4727235166531754, + "learning_rate": 3.1028385324903154e-06, + "loss": 1.0166, + "step": 3711 + }, + { + "epoch": 0.3347612391216125, + "grad_norm": 2.185898766982275, + "learning_rate": 3.1023511204301465e-06, + "loss": 0.9815, + "step": 3712 + }, + { + "epoch": 0.3348514226450827, + "grad_norm": 1.3487290245171608, + "learning_rate": 3.1018636143113022e-06, + "loss": 0.9362, + "step": 3713 + }, + { + "epoch": 0.334941606168553, + "grad_norm": 1.4933506292096523, + "learning_rate": 3.1013760141753787e-06, + "loss": 0.8776, + "step": 3714 + }, + { + "epoch": 0.33503178969202324, + "grad_norm": 1.797336485936846, + "learning_rate": 3.100888320063981e-06, + "loss": 0.9385, + "step": 3715 + }, + { + "epoch": 0.33512197321549353, + "grad_norm": 1.407985675757869, + "learning_rate": 3.100400532018721e-06, + "loss": 0.9758, + "step": 3716 + }, + { + "epoch": 0.3352121567389638, + "grad_norm": 2.0340729098010413, + "learning_rate": 3.0999126500812204e-06, + "loss": 1.0426, + "step": 3717 + }, + { + "epoch": 0.33530234026243405, + "grad_norm": 0.759687797267013, + "learning_rate": 3.0994246742931076e-06, + "loss": 0.8352, + "step": 3718 + }, + { + "epoch": 0.33539252378590434, + "grad_norm": 1.4790257489772431, + "learning_rate": 3.098936604696019e-06, + "loss": 0.9152, + "step": 3719 + }, + { + "epoch": 0.3354827073093746, + "grad_norm": 1.811034432187608, + "learning_rate": 3.0984484413316e-06, + "loss": 1.0028, + "step": 3720 + }, + { + "epoch": 0.33557289083284486, + "grad_norm": 1.7633315187792225, + "learning_rate": 3.0979601842415033e-06, + "loss": 0.9496, + "step": 3721 + }, + { + "epoch": 0.3356630743563151, + "grad_norm": 1.3224481318637913, + "learning_rate": 3.0974718334673896e-06, + "loss": 0.8948, + "step": 3722 + }, + { + "epoch": 0.3357532578797854, + "grad_norm": 1.4494671467337399, + "learning_rate": 3.0969833890509282e-06, + "loss": 0.9657, + "step": 3723 + }, + { + "epoch": 0.3358434414032556, + "grad_norm": 1.706913924386251, + "learning_rate": 3.096494851033795e-06, + "loss": 0.9624, + "step": 3724 + }, + { + "epoch": 0.3359336249267259, + "grad_norm": 1.9374735430480041, + "learning_rate": 3.0960062194576747e-06, + "loss": 0.8985, + "step": 3725 + }, + { + "epoch": 0.33602380845019614, + "grad_norm": 1.5873037382141557, + "learning_rate": 3.0955174943642606e-06, + "loss": 0.9223, + "step": 3726 + }, + { + "epoch": 0.3361139919736664, + "grad_norm": 1.4582195600906291, + "learning_rate": 3.0950286757952534e-06, + "loss": 0.8909, + "step": 3727 + }, + { + "epoch": 0.33620417549713666, + "grad_norm": 1.4477406493121958, + "learning_rate": 3.0945397637923617e-06, + "loss": 1.0324, + "step": 3728 + }, + { + "epoch": 0.33629435902060695, + "grad_norm": 1.7600092745666622, + "learning_rate": 3.0940507583973025e-06, + "loss": 0.9813, + "step": 3729 + }, + { + "epoch": 0.3363845425440772, + "grad_norm": 1.592388394216269, + "learning_rate": 3.093561659651799e-06, + "loss": 0.9394, + "step": 3730 + }, + { + "epoch": 0.33647472606754747, + "grad_norm": 1.5786915130534023, + "learning_rate": 3.093072467597586e-06, + "loss": 0.9077, + "step": 3731 + }, + { + "epoch": 0.3365649095910177, + "grad_norm": 0.7204574538408217, + "learning_rate": 3.092583182276402e-06, + "loss": 0.7857, + "step": 3732 + }, + { + "epoch": 0.336655093114488, + "grad_norm": 0.6565120718689365, + "learning_rate": 3.092093803729997e-06, + "loss": 0.7918, + "step": 3733 + }, + { + "epoch": 0.3367452766379582, + "grad_norm": 1.5105564272006127, + "learning_rate": 3.0916043320001264e-06, + "loss": 0.9822, + "step": 3734 + }, + { + "epoch": 0.3368354601614285, + "grad_norm": 1.6794600221827936, + "learning_rate": 3.0911147671285557e-06, + "loss": 0.9805, + "step": 3735 + }, + { + "epoch": 0.33692564368489875, + "grad_norm": 1.5923487168920456, + "learning_rate": 3.0906251091570565e-06, + "loss": 1.0034, + "step": 3736 + }, + { + "epoch": 0.33701582720836903, + "grad_norm": 1.7094244690036093, + "learning_rate": 3.0901353581274094e-06, + "loss": 0.9811, + "step": 3737 + }, + { + "epoch": 0.33710601073183927, + "grad_norm": 1.4102380364283695, + "learning_rate": 3.089645514081402e-06, + "loss": 0.8453, + "step": 3738 + }, + { + "epoch": 0.33719619425530956, + "grad_norm": 2.574688001823168, + "learning_rate": 3.0891555770608323e-06, + "loss": 0.9022, + "step": 3739 + }, + { + "epoch": 0.33728637777877984, + "grad_norm": 1.6682400985234143, + "learning_rate": 3.088665547107503e-06, + "loss": 0.9958, + "step": 3740 + }, + { + "epoch": 0.3373765613022501, + "grad_norm": 1.4317588601202071, + "learning_rate": 3.0881754242632254e-06, + "loss": 1.0538, + "step": 3741 + }, + { + "epoch": 0.33746674482572037, + "grad_norm": 1.2967755842595516, + "learning_rate": 3.0876852085698213e-06, + "loss": 1.0168, + "step": 3742 + }, + { + "epoch": 0.3375569283491906, + "grad_norm": 1.7328041454832694, + "learning_rate": 3.087194900069117e-06, + "loss": 1.0003, + "step": 3743 + }, + { + "epoch": 0.3376471118726609, + "grad_norm": 1.4559625792200956, + "learning_rate": 3.08670449880295e-06, + "loss": 1.0296, + "step": 3744 + }, + { + "epoch": 0.3377372953961311, + "grad_norm": 1.7350228963874712, + "learning_rate": 3.086214004813163e-06, + "loss": 0.9948, + "step": 3745 + }, + { + "epoch": 0.3378274789196014, + "grad_norm": 1.6162203785600384, + "learning_rate": 3.0857234181416074e-06, + "loss": 1.0347, + "step": 3746 + }, + { + "epoch": 0.33791766244307164, + "grad_norm": 1.615449422891705, + "learning_rate": 3.085232738830143e-06, + "loss": 1.0238, + "step": 3747 + }, + { + "epoch": 0.33800784596654193, + "grad_norm": 1.820299991679295, + "learning_rate": 3.084741966920638e-06, + "loss": 1.0326, + "step": 3748 + }, + { + "epoch": 0.33809802949001216, + "grad_norm": 1.4606491049008674, + "learning_rate": 3.084251102454966e-06, + "loss": 0.9357, + "step": 3749 + }, + { + "epoch": 0.33818821301348245, + "grad_norm": 0.9035624741391122, + "learning_rate": 3.083760145475013e-06, + "loss": 0.8156, + "step": 3750 + }, + { + "epoch": 0.3382783965369527, + "grad_norm": 1.5025989746145088, + "learning_rate": 3.0832690960226678e-06, + "loss": 0.9168, + "step": 3751 + }, + { + "epoch": 0.338368580060423, + "grad_norm": 1.955352550533548, + "learning_rate": 3.08277795413983e-06, + "loss": 0.9428, + "step": 3752 + }, + { + "epoch": 0.3384587635838932, + "grad_norm": 1.5444705655993998, + "learning_rate": 3.0822867198684073e-06, + "loss": 1.0251, + "step": 3753 + }, + { + "epoch": 0.3385489471073635, + "grad_norm": 1.5554949326661038, + "learning_rate": 3.081795393250314e-06, + "loss": 0.9845, + "step": 3754 + }, + { + "epoch": 0.33863913063083373, + "grad_norm": 1.7876153948689675, + "learning_rate": 3.081303974327473e-06, + "loss": 0.9423, + "step": 3755 + }, + { + "epoch": 0.338729314154304, + "grad_norm": 1.441105756993978, + "learning_rate": 3.080812463141814e-06, + "loss": 0.973, + "step": 3756 + }, + { + "epoch": 0.33881949767777425, + "grad_norm": 1.377564769198344, + "learning_rate": 3.080320859735276e-06, + "loss": 0.8488, + "step": 3757 + }, + { + "epoch": 0.33890968120124454, + "grad_norm": 1.6943540896451994, + "learning_rate": 3.079829164149806e-06, + "loss": 1.0283, + "step": 3758 + }, + { + "epoch": 0.3389998647247148, + "grad_norm": 1.5523878562585947, + "learning_rate": 3.0793373764273573e-06, + "loss": 0.9636, + "step": 3759 + }, + { + "epoch": 0.33909004824818506, + "grad_norm": 1.6550557851465375, + "learning_rate": 3.078845496609892e-06, + "loss": 0.9999, + "step": 3760 + }, + { + "epoch": 0.3391802317716553, + "grad_norm": 1.4553960465656792, + "learning_rate": 3.078353524739381e-06, + "loss": 1.0631, + "step": 3761 + }, + { + "epoch": 0.3392704152951256, + "grad_norm": 1.5197835286240449, + "learning_rate": 3.077861460857801e-06, + "loss": 1.0588, + "step": 3762 + }, + { + "epoch": 0.3393605988185958, + "grad_norm": 1.5556363224829797, + "learning_rate": 3.077369305007138e-06, + "loss": 0.9903, + "step": 3763 + }, + { + "epoch": 0.3394507823420661, + "grad_norm": 1.4735574974799988, + "learning_rate": 3.0768770572293852e-06, + "loss": 1.0105, + "step": 3764 + }, + { + "epoch": 0.3395409658655364, + "grad_norm": 1.916803569327233, + "learning_rate": 3.0763847175665437e-06, + "loss": 1.0541, + "step": 3765 + }, + { + "epoch": 0.3396311493890066, + "grad_norm": 1.4130074796269443, + "learning_rate": 3.0758922860606237e-06, + "loss": 0.9147, + "step": 3766 + }, + { + "epoch": 0.3397213329124769, + "grad_norm": 1.7100466146306055, + "learning_rate": 3.0753997627536404e-06, + "loss": 0.912, + "step": 3767 + }, + { + "epoch": 0.33981151643594715, + "grad_norm": 2.372854669267299, + "learning_rate": 3.0749071476876203e-06, + "loss": 0.9274, + "step": 3768 + }, + { + "epoch": 0.33990169995941744, + "grad_norm": 1.4786917053753628, + "learning_rate": 3.0744144409045952e-06, + "loss": 0.9477, + "step": 3769 + }, + { + "epoch": 0.33999188348288767, + "grad_norm": 1.8420422290471898, + "learning_rate": 3.0739216424466056e-06, + "loss": 0.9984, + "step": 3770 + }, + { + "epoch": 0.34008206700635796, + "grad_norm": 1.8905328775456227, + "learning_rate": 3.0734287523557002e-06, + "loss": 0.9625, + "step": 3771 + }, + { + "epoch": 0.3401722505298282, + "grad_norm": 1.7309594416635192, + "learning_rate": 3.0729357706739348e-06, + "loss": 0.9289, + "step": 3772 + }, + { + "epoch": 0.3402624340532985, + "grad_norm": 1.4690079969703922, + "learning_rate": 3.0724426974433737e-06, + "loss": 0.9628, + "step": 3773 + }, + { + "epoch": 0.3403526175767687, + "grad_norm": 1.3426680533930093, + "learning_rate": 3.0719495327060874e-06, + "loss": 0.9439, + "step": 3774 + }, + { + "epoch": 0.340442801100239, + "grad_norm": 1.4743973939625254, + "learning_rate": 3.071456276504157e-06, + "loss": 0.9856, + "step": 3775 + }, + { + "epoch": 0.34053298462370923, + "grad_norm": 1.4507301038067648, + "learning_rate": 3.070962928879669e-06, + "loss": 0.9731, + "step": 3776 + }, + { + "epoch": 0.3406231681471795, + "grad_norm": 1.563770921721186, + "learning_rate": 3.0704694898747185e-06, + "loss": 0.9873, + "step": 3777 + }, + { + "epoch": 0.34071335167064976, + "grad_norm": 1.5094078992334414, + "learning_rate": 3.069975959531408e-06, + "loss": 1.0978, + "step": 3778 + }, + { + "epoch": 0.34080353519412004, + "grad_norm": 0.7542697642831154, + "learning_rate": 3.06948233789185e-06, + "loss": 0.8508, + "step": 3779 + }, + { + "epoch": 0.3408937187175903, + "grad_norm": 1.5507081444562147, + "learning_rate": 3.0689886249981614e-06, + "loss": 0.9931, + "step": 3780 + }, + { + "epoch": 0.34098390224106057, + "grad_norm": 1.3908228786998207, + "learning_rate": 3.0684948208924693e-06, + "loss": 1.0049, + "step": 3781 + }, + { + "epoch": 0.3410740857645308, + "grad_norm": 1.5641404505005974, + "learning_rate": 3.068000925616907e-06, + "loss": 0.9954, + "step": 3782 + }, + { + "epoch": 0.3411642692880011, + "grad_norm": 1.7863455719203203, + "learning_rate": 3.067506939213617e-06, + "loss": 0.9652, + "step": 3783 + }, + { + "epoch": 0.3412544528114713, + "grad_norm": 1.3525303809372777, + "learning_rate": 3.0670128617247493e-06, + "loss": 0.9816, + "step": 3784 + }, + { + "epoch": 0.3413446363349416, + "grad_norm": 1.352265941159886, + "learning_rate": 3.06651869319246e-06, + "loss": 0.8896, + "step": 3785 + }, + { + "epoch": 0.34143481985841184, + "grad_norm": 1.6980575627880663, + "learning_rate": 3.0660244336589154e-06, + "loss": 0.9722, + "step": 3786 + }, + { + "epoch": 0.34152500338188213, + "grad_norm": 1.4388833231595346, + "learning_rate": 3.065530083166288e-06, + "loss": 0.99, + "step": 3787 + }, + { + "epoch": 0.3416151869053524, + "grad_norm": 1.4065948911105728, + "learning_rate": 3.0650356417567586e-06, + "loss": 0.939, + "step": 3788 + }, + { + "epoch": 0.34170537042882265, + "grad_norm": 2.275966430631035, + "learning_rate": 3.0645411094725156e-06, + "loss": 1.0265, + "step": 3789 + }, + { + "epoch": 0.34179555395229294, + "grad_norm": 1.489658629463954, + "learning_rate": 3.0640464863557556e-06, + "loss": 0.9683, + "step": 3790 + }, + { + "epoch": 0.3418857374757632, + "grad_norm": 1.4087680736286115, + "learning_rate": 3.063551772448682e-06, + "loss": 1.0118, + "step": 3791 + }, + { + "epoch": 0.34197592099923346, + "grad_norm": 1.7458115590755021, + "learning_rate": 3.0630569677935075e-06, + "loss": 0.9057, + "step": 3792 + }, + { + "epoch": 0.3420661045227037, + "grad_norm": 1.7455218784618791, + "learning_rate": 3.06256207243245e-06, + "loss": 1.063, + "step": 3793 + }, + { + "epoch": 0.342156288046174, + "grad_norm": 0.6924509464200981, + "learning_rate": 3.0620670864077385e-06, + "loss": 0.7699, + "step": 3794 + }, + { + "epoch": 0.3422464715696442, + "grad_norm": 0.6558358156972371, + "learning_rate": 3.0615720097616063e-06, + "loss": 0.7726, + "step": 3795 + }, + { + "epoch": 0.3423366550931145, + "grad_norm": 1.4947226194945054, + "learning_rate": 3.0610768425362967e-06, + "loss": 0.9315, + "step": 3796 + }, + { + "epoch": 0.34242683861658474, + "grad_norm": 1.4672355960848056, + "learning_rate": 3.0605815847740603e-06, + "loss": 0.9563, + "step": 3797 + }, + { + "epoch": 0.342517022140055, + "grad_norm": 1.6347689793779057, + "learning_rate": 3.0600862365171553e-06, + "loss": 0.94, + "step": 3798 + }, + { + "epoch": 0.34260720566352526, + "grad_norm": 1.3981647331993519, + "learning_rate": 3.0595907978078474e-06, + "loss": 0.9522, + "step": 3799 + }, + { + "epoch": 0.34269738918699555, + "grad_norm": 1.7105680441303273, + "learning_rate": 3.05909526868841e-06, + "loss": 0.9489, + "step": 3800 + }, + { + "epoch": 0.3427875727104658, + "grad_norm": 1.4602054659089208, + "learning_rate": 3.0585996492011243e-06, + "loss": 0.9811, + "step": 3801 + }, + { + "epoch": 0.34287775623393607, + "grad_norm": 1.7440587416476623, + "learning_rate": 3.05810393938828e-06, + "loss": 1.022, + "step": 3802 + }, + { + "epoch": 0.3429679397574063, + "grad_norm": 1.4478770829239855, + "learning_rate": 3.0576081392921723e-06, + "loss": 0.886, + "step": 3803 + }, + { + "epoch": 0.3430581232808766, + "grad_norm": 1.6968499169506766, + "learning_rate": 3.057112248955107e-06, + "loss": 1.0213, + "step": 3804 + }, + { + "epoch": 0.3431483068043468, + "grad_norm": 2.188102203311406, + "learning_rate": 3.0566162684193963e-06, + "loss": 0.9893, + "step": 3805 + }, + { + "epoch": 0.3432384903278171, + "grad_norm": 1.4032802589344195, + "learning_rate": 3.056120197727359e-06, + "loss": 0.925, + "step": 3806 + }, + { + "epoch": 0.34332867385128735, + "grad_norm": 1.3992364402006285, + "learning_rate": 3.0556240369213236e-06, + "loss": 0.9613, + "step": 3807 + }, + { + "epoch": 0.34341885737475764, + "grad_norm": 1.8449760498518613, + "learning_rate": 3.055127786043624e-06, + "loss": 0.8839, + "step": 3808 + }, + { + "epoch": 0.34350904089822787, + "grad_norm": 1.3534272840770951, + "learning_rate": 3.054631445136604e-06, + "loss": 1.0632, + "step": 3809 + }, + { + "epoch": 0.34359922442169816, + "grad_norm": 1.8656751820829593, + "learning_rate": 3.0541350142426147e-06, + "loss": 0.9883, + "step": 3810 + }, + { + "epoch": 0.3436894079451684, + "grad_norm": 1.440356940032684, + "learning_rate": 3.053638493404012e-06, + "loss": 0.9315, + "step": 3811 + }, + { + "epoch": 0.3437795914686387, + "grad_norm": 1.446790197536134, + "learning_rate": 3.0531418826631643e-06, + "loss": 0.9127, + "step": 3812 + }, + { + "epoch": 0.34386977499210897, + "grad_norm": 1.5473736816518966, + "learning_rate": 3.052645182062444e-06, + "loss": 1.0535, + "step": 3813 + }, + { + "epoch": 0.3439599585155792, + "grad_norm": 1.5906407563454013, + "learning_rate": 3.0521483916442324e-06, + "loss": 1.0213, + "step": 3814 + }, + { + "epoch": 0.3440501420390495, + "grad_norm": 1.495903939340238, + "learning_rate": 3.0516515114509183e-06, + "loss": 0.9658, + "step": 3815 + }, + { + "epoch": 0.3441403255625197, + "grad_norm": 1.3419288448196578, + "learning_rate": 3.0511545415249e-06, + "loss": 0.926, + "step": 3816 + }, + { + "epoch": 0.34423050908599, + "grad_norm": 1.7277038392126913, + "learning_rate": 3.050657481908579e-06, + "loss": 1.0158, + "step": 3817 + }, + { + "epoch": 0.34432069260946024, + "grad_norm": 1.8341133307964037, + "learning_rate": 3.0501603326443677e-06, + "loss": 1.0826, + "step": 3818 + }, + { + "epoch": 0.34441087613293053, + "grad_norm": 1.5168048945865045, + "learning_rate": 3.049663093774687e-06, + "loss": 0.9634, + "step": 3819 + }, + { + "epoch": 0.34450105965640077, + "grad_norm": 1.9130459947296223, + "learning_rate": 3.0491657653419643e-06, + "loss": 0.8708, + "step": 3820 + }, + { + "epoch": 0.34459124317987105, + "grad_norm": 1.623840725473386, + "learning_rate": 3.0486683473886325e-06, + "loss": 1.0221, + "step": 3821 + }, + { + "epoch": 0.3446814267033413, + "grad_norm": 2.361660690174424, + "learning_rate": 3.0481708399571355e-06, + "loss": 1.0347, + "step": 3822 + }, + { + "epoch": 0.3447716102268116, + "grad_norm": 1.5961799796212675, + "learning_rate": 3.047673243089922e-06, + "loss": 1.0029, + "step": 3823 + }, + { + "epoch": 0.3448617937502818, + "grad_norm": 1.5797736743397266, + "learning_rate": 3.047175556829451e-06, + "loss": 1.0041, + "step": 3824 + }, + { + "epoch": 0.3449519772737521, + "grad_norm": 1.6534449738193178, + "learning_rate": 3.046677781218188e-06, + "loss": 1.0272, + "step": 3825 + }, + { + "epoch": 0.34504216079722233, + "grad_norm": 1.4526973628360185, + "learning_rate": 3.0461799162986043e-06, + "loss": 0.9544, + "step": 3826 + }, + { + "epoch": 0.3451323443206926, + "grad_norm": 1.7447482051465715, + "learning_rate": 3.045681962113183e-06, + "loss": 0.9033, + "step": 3827 + }, + { + "epoch": 0.34522252784416285, + "grad_norm": 1.4821652478022154, + "learning_rate": 3.0451839187044095e-06, + "loss": 0.9237, + "step": 3828 + }, + { + "epoch": 0.34531271136763314, + "grad_norm": 1.4913716107877417, + "learning_rate": 3.0446857861147816e-06, + "loss": 0.9921, + "step": 3829 + }, + { + "epoch": 0.3454028948911034, + "grad_norm": 1.797351608093376, + "learning_rate": 3.044187564386802e-06, + "loss": 0.945, + "step": 3830 + }, + { + "epoch": 0.34549307841457366, + "grad_norm": 1.7241869848770985, + "learning_rate": 3.0436892535629818e-06, + "loss": 1.0386, + "step": 3831 + }, + { + "epoch": 0.3455832619380439, + "grad_norm": 1.6036596198716424, + "learning_rate": 3.0431908536858393e-06, + "loss": 1.0605, + "step": 3832 + }, + { + "epoch": 0.3456734454615142, + "grad_norm": 1.8447228790618004, + "learning_rate": 3.0426923647979016e-06, + "loss": 0.9648, + "step": 3833 + }, + { + "epoch": 0.3457636289849844, + "grad_norm": 1.6218378404538936, + "learning_rate": 3.0421937869417016e-06, + "loss": 0.9153, + "step": 3834 + }, + { + "epoch": 0.3458538125084547, + "grad_norm": 1.568241024658751, + "learning_rate": 3.041695120159782e-06, + "loss": 1.0357, + "step": 3835 + }, + { + "epoch": 0.345943996031925, + "grad_norm": 1.5020080476994802, + "learning_rate": 3.04119636449469e-06, + "loss": 0.9104, + "step": 3836 + }, + { + "epoch": 0.3460341795553952, + "grad_norm": 1.367643487365856, + "learning_rate": 3.040697519988983e-06, + "loss": 1.0044, + "step": 3837 + }, + { + "epoch": 0.3461243630788655, + "grad_norm": 0.6937735596514043, + "learning_rate": 3.040198586685226e-06, + "loss": 0.793, + "step": 3838 + }, + { + "epoch": 0.34621454660233575, + "grad_norm": 1.5971541606757524, + "learning_rate": 3.039699564625989e-06, + "loss": 0.9275, + "step": 3839 + }, + { + "epoch": 0.34630473012580604, + "grad_norm": 0.725649762844125, + "learning_rate": 3.039200453853853e-06, + "loss": 0.7867, + "step": 3840 + }, + { + "epoch": 0.34639491364927627, + "grad_norm": 1.73846324867985, + "learning_rate": 3.038701254411404e-06, + "loss": 1.0214, + "step": 3841 + }, + { + "epoch": 0.34648509717274656, + "grad_norm": 1.6735692219044844, + "learning_rate": 3.0382019663412367e-06, + "loss": 0.9913, + "step": 3842 + }, + { + "epoch": 0.3465752806962168, + "grad_norm": 1.9280957286226508, + "learning_rate": 3.0377025896859532e-06, + "loss": 1.0155, + "step": 3843 + }, + { + "epoch": 0.3466654642196871, + "grad_norm": 1.5318924470901685, + "learning_rate": 3.0372031244881627e-06, + "loss": 1.0841, + "step": 3844 + }, + { + "epoch": 0.3467556477431573, + "grad_norm": 1.5750102663462449, + "learning_rate": 3.0367035707904826e-06, + "loss": 0.936, + "step": 3845 + }, + { + "epoch": 0.3468458312666276, + "grad_norm": 1.5596648625309528, + "learning_rate": 3.036203928635537e-06, + "loss": 0.8879, + "step": 3846 + }, + { + "epoch": 0.34693601479009784, + "grad_norm": 1.7303875975169207, + "learning_rate": 3.035704198065959e-06, + "loss": 1.0448, + "step": 3847 + }, + { + "epoch": 0.3470261983135681, + "grad_norm": 1.6609867666407485, + "learning_rate": 3.0352043791243886e-06, + "loss": 0.9639, + "step": 3848 + }, + { + "epoch": 0.34711638183703836, + "grad_norm": 0.7982506191778126, + "learning_rate": 3.034704471853472e-06, + "loss": 0.79, + "step": 3849 + }, + { + "epoch": 0.34720656536050865, + "grad_norm": 1.5203861293233274, + "learning_rate": 3.0342044762958646e-06, + "loss": 1.0198, + "step": 3850 + }, + { + "epoch": 0.3472967488839789, + "grad_norm": 1.5989702785513789, + "learning_rate": 3.0337043924942286e-06, + "loss": 1.0122, + "step": 3851 + }, + { + "epoch": 0.34738693240744917, + "grad_norm": 1.671122158641912, + "learning_rate": 3.0332042204912343e-06, + "loss": 0.9671, + "step": 3852 + }, + { + "epoch": 0.3474771159309194, + "grad_norm": 0.848324701106652, + "learning_rate": 3.0327039603295587e-06, + "loss": 0.7954, + "step": 3853 + }, + { + "epoch": 0.3475672994543897, + "grad_norm": 2.103012775754192, + "learning_rate": 3.032203612051887e-06, + "loss": 0.9427, + "step": 3854 + }, + { + "epoch": 0.3476574829778599, + "grad_norm": 1.5999669757057484, + "learning_rate": 3.0317031757009116e-06, + "loss": 0.9559, + "step": 3855 + }, + { + "epoch": 0.3477476665013302, + "grad_norm": 1.5194407928921938, + "learning_rate": 3.0312026513193326e-06, + "loss": 0.9347, + "step": 3856 + }, + { + "epoch": 0.34783785002480044, + "grad_norm": 1.5124621079167277, + "learning_rate": 3.0307020389498573e-06, + "loss": 0.955, + "step": 3857 + }, + { + "epoch": 0.34792803354827073, + "grad_norm": 1.4648616535364234, + "learning_rate": 3.0302013386352004e-06, + "loss": 0.9033, + "step": 3858 + }, + { + "epoch": 0.348018217071741, + "grad_norm": 1.4677155280856256, + "learning_rate": 3.0297005504180854e-06, + "loss": 0.9663, + "step": 3859 + }, + { + "epoch": 0.34810840059521125, + "grad_norm": 0.7131435833849108, + "learning_rate": 3.0291996743412417e-06, + "loss": 0.7139, + "step": 3860 + }, + { + "epoch": 0.34819858411868154, + "grad_norm": 1.4761641838073236, + "learning_rate": 3.0286987104474063e-06, + "loss": 1.0682, + "step": 3861 + }, + { + "epoch": 0.3482887676421518, + "grad_norm": 1.277760615556059, + "learning_rate": 3.028197658779325e-06, + "loss": 0.9484, + "step": 3862 + }, + { + "epoch": 0.34837895116562206, + "grad_norm": 1.792669067602984, + "learning_rate": 3.0276965193797503e-06, + "loss": 0.9977, + "step": 3863 + }, + { + "epoch": 0.3484691346890923, + "grad_norm": 1.3786947720378793, + "learning_rate": 3.0271952922914423e-06, + "loss": 0.9372, + "step": 3864 + }, + { + "epoch": 0.3485593182125626, + "grad_norm": 1.5091488307416394, + "learning_rate": 3.0266939775571675e-06, + "loss": 0.9549, + "step": 3865 + }, + { + "epoch": 0.3486495017360328, + "grad_norm": 1.7667002821669178, + "learning_rate": 3.026192575219701e-06, + "loss": 0.9633, + "step": 3866 + }, + { + "epoch": 0.3487396852595031, + "grad_norm": 1.466471589004097, + "learning_rate": 3.025691085321826e-06, + "loss": 0.8682, + "step": 3867 + }, + { + "epoch": 0.34882986878297334, + "grad_norm": 1.5298892611900106, + "learning_rate": 3.025189507906332e-06, + "loss": 0.9584, + "step": 3868 + }, + { + "epoch": 0.34892005230644363, + "grad_norm": 1.4105403303122885, + "learning_rate": 3.0246878430160166e-06, + "loss": 0.9924, + "step": 3869 + }, + { + "epoch": 0.34901023582991386, + "grad_norm": 1.5322774437357285, + "learning_rate": 3.024186090693684e-06, + "loss": 0.9159, + "step": 3870 + }, + { + "epoch": 0.34910041935338415, + "grad_norm": 1.7091508010482506, + "learning_rate": 3.023684250982147e-06, + "loss": 1.0, + "step": 3871 + }, + { + "epoch": 0.3491906028768544, + "grad_norm": 1.6241849909545154, + "learning_rate": 3.0231823239242252e-06, + "loss": 0.8938, + "step": 3872 + }, + { + "epoch": 0.34928078640032467, + "grad_norm": 1.5186853088918215, + "learning_rate": 3.0226803095627457e-06, + "loss": 0.9283, + "step": 3873 + }, + { + "epoch": 0.3493709699237949, + "grad_norm": 1.792059042200628, + "learning_rate": 3.022178207940543e-06, + "loss": 0.9491, + "step": 3874 + }, + { + "epoch": 0.3494611534472652, + "grad_norm": 1.5876626089122838, + "learning_rate": 3.02167601910046e-06, + "loss": 0.9885, + "step": 3875 + }, + { + "epoch": 0.3495513369707354, + "grad_norm": 1.6588298210940604, + "learning_rate": 3.021173743085345e-06, + "loss": 1.0003, + "step": 3876 + }, + { + "epoch": 0.3496415204942057, + "grad_norm": 0.7760273123486753, + "learning_rate": 3.0206713799380557e-06, + "loss": 0.7806, + "step": 3877 + }, + { + "epoch": 0.34973170401767595, + "grad_norm": 1.6017347289730253, + "learning_rate": 3.0201689297014565e-06, + "loss": 0.9801, + "step": 3878 + }, + { + "epoch": 0.34982188754114624, + "grad_norm": 1.2933184101941388, + "learning_rate": 3.0196663924184187e-06, + "loss": 1.0166, + "step": 3879 + }, + { + "epoch": 0.34991207106461647, + "grad_norm": 2.0318312766921816, + "learning_rate": 3.019163768131822e-06, + "loss": 1.0029, + "step": 3880 + }, + { + "epoch": 0.35000225458808676, + "grad_norm": 1.6169299966477981, + "learning_rate": 3.0186610568845533e-06, + "loss": 0.8997, + "step": 3881 + }, + { + "epoch": 0.350092438111557, + "grad_norm": 2.1638250668273353, + "learning_rate": 3.018158258719507e-06, + "loss": 0.906, + "step": 3882 + }, + { + "epoch": 0.3501826216350273, + "grad_norm": 1.4764614171542767, + "learning_rate": 3.0176553736795827e-06, + "loss": 1.0011, + "step": 3883 + }, + { + "epoch": 0.35027280515849757, + "grad_norm": 0.6323282013403178, + "learning_rate": 3.017152401807691e-06, + "loss": 0.719, + "step": 3884 + }, + { + "epoch": 0.3503629886819678, + "grad_norm": 1.4493890955257258, + "learning_rate": 3.0166493431467476e-06, + "loss": 0.9609, + "step": 3885 + }, + { + "epoch": 0.3504531722054381, + "grad_norm": 1.3695044367461742, + "learning_rate": 3.016146197739677e-06, + "loss": 0.9434, + "step": 3886 + }, + { + "epoch": 0.3505433557289083, + "grad_norm": 1.5073622589699018, + "learning_rate": 3.0156429656294097e-06, + "loss": 0.9042, + "step": 3887 + }, + { + "epoch": 0.3506335392523786, + "grad_norm": 1.3844153818135132, + "learning_rate": 3.0151396468588844e-06, + "loss": 0.9151, + "step": 3888 + }, + { + "epoch": 0.35072372277584885, + "grad_norm": 2.133078310545222, + "learning_rate": 3.014636241471047e-06, + "loss": 0.9263, + "step": 3889 + }, + { + "epoch": 0.35081390629931913, + "grad_norm": 2.005900381647845, + "learning_rate": 3.0141327495088514e-06, + "loss": 0.9672, + "step": 3890 + }, + { + "epoch": 0.35090408982278937, + "grad_norm": 1.5223369829320514, + "learning_rate": 3.0136291710152566e-06, + "loss": 0.9301, + "step": 3891 + }, + { + "epoch": 0.35099427334625966, + "grad_norm": 1.5392398635210798, + "learning_rate": 3.0131255060332325e-06, + "loss": 1.011, + "step": 3892 + }, + { + "epoch": 0.3510844568697299, + "grad_norm": 1.5107695682463402, + "learning_rate": 3.012621754605754e-06, + "loss": 0.987, + "step": 3893 + }, + { + "epoch": 0.3511746403932002, + "grad_norm": 1.5595072504812502, + "learning_rate": 3.0121179167758035e-06, + "loss": 1.0, + "step": 3894 + }, + { + "epoch": 0.3512648239166704, + "grad_norm": 1.7579494846798174, + "learning_rate": 3.0116139925863717e-06, + "loss": 1.0488, + "step": 3895 + }, + { + "epoch": 0.3513550074401407, + "grad_norm": 1.4325359714112615, + "learning_rate": 3.011109982080456e-06, + "loss": 1.0088, + "step": 3896 + }, + { + "epoch": 0.35144519096361093, + "grad_norm": 1.7252626661369965, + "learning_rate": 3.0106058853010614e-06, + "loss": 0.9181, + "step": 3897 + }, + { + "epoch": 0.3515353744870812, + "grad_norm": 1.4340232600689227, + "learning_rate": 3.010101702291201e-06, + "loss": 0.9558, + "step": 3898 + }, + { + "epoch": 0.35162555801055145, + "grad_norm": 1.5005890961037465, + "learning_rate": 3.009597433093893e-06, + "loss": 0.9674, + "step": 3899 + }, + { + "epoch": 0.35171574153402174, + "grad_norm": 1.7654349975978747, + "learning_rate": 3.009093077752165e-06, + "loss": 0.9795, + "step": 3900 + }, + { + "epoch": 0.351805925057492, + "grad_norm": 1.566263701309714, + "learning_rate": 3.008588636309052e-06, + "loss": 1.0031, + "step": 3901 + }, + { + "epoch": 0.35189610858096226, + "grad_norm": 1.4895554742230863, + "learning_rate": 3.0080841088075947e-06, + "loss": 0.9327, + "step": 3902 + }, + { + "epoch": 0.3519862921044325, + "grad_norm": 1.740380757315818, + "learning_rate": 3.0075794952908436e-06, + "loss": 1.0104, + "step": 3903 + }, + { + "epoch": 0.3520764756279028, + "grad_norm": 1.356936800343121, + "learning_rate": 3.0070747958018528e-06, + "loss": 0.9889, + "step": 3904 + }, + { + "epoch": 0.352166659151373, + "grad_norm": 1.3621435364054175, + "learning_rate": 3.0065700103836894e-06, + "loss": 0.9824, + "step": 3905 + }, + { + "epoch": 0.3522568426748433, + "grad_norm": 1.6879108070203397, + "learning_rate": 3.0060651390794214e-06, + "loss": 0.9896, + "step": 3906 + }, + { + "epoch": 0.3523470261983136, + "grad_norm": 1.4953986164037127, + "learning_rate": 3.005560181932128e-06, + "loss": 0.952, + "step": 3907 + }, + { + "epoch": 0.35243720972178383, + "grad_norm": 1.6118116728289327, + "learning_rate": 3.005055138984896e-06, + "loss": 0.9233, + "step": 3908 + }, + { + "epoch": 0.3525273932452541, + "grad_norm": 1.6392554970457553, + "learning_rate": 3.0045500102808174e-06, + "loss": 0.9416, + "step": 3909 + }, + { + "epoch": 0.35261757676872435, + "grad_norm": 1.884725588042214, + "learning_rate": 3.0040447958629927e-06, + "loss": 1.0522, + "step": 3910 + }, + { + "epoch": 0.35270776029219464, + "grad_norm": 1.6073333233791516, + "learning_rate": 3.00353949577453e-06, + "loss": 0.9923, + "step": 3911 + }, + { + "epoch": 0.35279794381566487, + "grad_norm": 5.7122645890959785, + "learning_rate": 3.003034110058544e-06, + "loss": 0.8095, + "step": 3912 + }, + { + "epoch": 0.35288812733913516, + "grad_norm": 1.5001839683887508, + "learning_rate": 3.002528638758157e-06, + "loss": 0.9493, + "step": 3913 + }, + { + "epoch": 0.3529783108626054, + "grad_norm": 2.1235677717050074, + "learning_rate": 3.0020230819164985e-06, + "loss": 0.9673, + "step": 3914 + }, + { + "epoch": 0.3530684943860757, + "grad_norm": 1.4289280276049425, + "learning_rate": 3.0015174395767064e-06, + "loss": 0.9768, + "step": 3915 + }, + { + "epoch": 0.3531586779095459, + "grad_norm": 1.9435462746770176, + "learning_rate": 3.001011711781923e-06, + "loss": 0.9918, + "step": 3916 + }, + { + "epoch": 0.3532488614330162, + "grad_norm": 1.4457376215729874, + "learning_rate": 3.0005058985753017e-06, + "loss": 0.9258, + "step": 3917 + }, + { + "epoch": 0.35333904495648644, + "grad_norm": 1.5604883595979413, + "learning_rate": 3e-06, + "loss": 0.9602, + "step": 3918 + }, + { + "epoch": 0.3534292284799567, + "grad_norm": 0.7197402683645587, + "learning_rate": 2.9994940160991843e-06, + "loss": 0.808, + "step": 3919 + }, + { + "epoch": 0.35351941200342696, + "grad_norm": 0.6643173626276633, + "learning_rate": 2.9989879469160285e-06, + "loss": 0.7646, + "step": 3920 + }, + { + "epoch": 0.35360959552689725, + "grad_norm": 1.5112717236139948, + "learning_rate": 2.9984817924937124e-06, + "loss": 0.9906, + "step": 3921 + }, + { + "epoch": 0.3536997790503675, + "grad_norm": 1.5191107709976772, + "learning_rate": 2.997975552875424e-06, + "loss": 0.9123, + "step": 3922 + }, + { + "epoch": 0.35378996257383777, + "grad_norm": 1.6015129267766053, + "learning_rate": 2.997469228104358e-06, + "loss": 0.9569, + "step": 3923 + }, + { + "epoch": 0.353880146097308, + "grad_norm": 1.5209621566488785, + "learning_rate": 2.996962818223718e-06, + "loss": 0.8631, + "step": 3924 + }, + { + "epoch": 0.3539703296207783, + "grad_norm": 1.9557179443470847, + "learning_rate": 2.9964563232767135e-06, + "loss": 0.917, + "step": 3925 + }, + { + "epoch": 0.3540605131442485, + "grad_norm": 1.6584928908657173, + "learning_rate": 2.9959497433065617e-06, + "loss": 0.9695, + "step": 3926 + }, + { + "epoch": 0.3541506966677188, + "grad_norm": 1.4293000576987829, + "learning_rate": 2.9954430783564848e-06, + "loss": 0.8867, + "step": 3927 + }, + { + "epoch": 0.35424088019118904, + "grad_norm": 1.5328224827308163, + "learning_rate": 2.994936328469716e-06, + "loss": 0.995, + "step": 3928 + }, + { + "epoch": 0.35433106371465933, + "grad_norm": 1.3884945956143189, + "learning_rate": 2.994429493689494e-06, + "loss": 0.9718, + "step": 3929 + }, + { + "epoch": 0.35442124723812957, + "grad_norm": 1.2525854076268246, + "learning_rate": 2.9939225740590642e-06, + "loss": 0.9907, + "step": 3930 + }, + { + "epoch": 0.35451143076159986, + "grad_norm": 1.3815563984037043, + "learning_rate": 2.99341556962168e-06, + "loss": 0.9254, + "step": 3931 + }, + { + "epoch": 0.35460161428507014, + "grad_norm": 1.3803812410503178, + "learning_rate": 2.992908480420602e-06, + "loss": 1.0095, + "step": 3932 + }, + { + "epoch": 0.3546917978085404, + "grad_norm": 1.5616738996606256, + "learning_rate": 2.9924013064990974e-06, + "loss": 0.9592, + "step": 3933 + }, + { + "epoch": 0.35478198133201067, + "grad_norm": 0.691069246047873, + "learning_rate": 2.991894047900441e-06, + "loss": 0.8124, + "step": 3934 + }, + { + "epoch": 0.3548721648554809, + "grad_norm": 1.4375980177923515, + "learning_rate": 2.991386704667916e-06, + "loss": 0.9845, + "step": 3935 + }, + { + "epoch": 0.3549623483789512, + "grad_norm": 1.774639579285833, + "learning_rate": 2.9908792768448097e-06, + "loss": 1.0311, + "step": 3936 + }, + { + "epoch": 0.3550525319024214, + "grad_norm": 1.6727413891679133, + "learning_rate": 2.990371764474421e-06, + "loss": 0.9583, + "step": 3937 + }, + { + "epoch": 0.3551427154258917, + "grad_norm": 1.5522091535052827, + "learning_rate": 2.9898641676000518e-06, + "loss": 0.9388, + "step": 3938 + }, + { + "epoch": 0.35523289894936194, + "grad_norm": 1.5076249554470174, + "learning_rate": 2.9893564862650138e-06, + "loss": 0.9683, + "step": 3939 + }, + { + "epoch": 0.35532308247283223, + "grad_norm": 1.8057226869877463, + "learning_rate": 2.9888487205126254e-06, + "loss": 1.0188, + "step": 3940 + }, + { + "epoch": 0.35541326599630246, + "grad_norm": 1.4146673268404424, + "learning_rate": 2.9883408703862115e-06, + "loss": 0.9622, + "step": 3941 + }, + { + "epoch": 0.35550344951977275, + "grad_norm": 1.9025697671777722, + "learning_rate": 2.987832935929105e-06, + "loss": 1.0494, + "step": 3942 + }, + { + "epoch": 0.355593633043243, + "grad_norm": 1.5368707997628084, + "learning_rate": 2.9873249171846454e-06, + "loss": 1.0296, + "step": 3943 + }, + { + "epoch": 0.3556838165667133, + "grad_norm": 2.0909400417464656, + "learning_rate": 2.98681681419618e-06, + "loss": 0.9608, + "step": 3944 + }, + { + "epoch": 0.3557740000901835, + "grad_norm": 1.9675048644843065, + "learning_rate": 2.9863086270070627e-06, + "loss": 1.0167, + "step": 3945 + }, + { + "epoch": 0.3558641836136538, + "grad_norm": 1.6374175189262576, + "learning_rate": 2.985800355660655e-06, + "loss": 0.9745, + "step": 3946 + }, + { + "epoch": 0.35595436713712403, + "grad_norm": 0.7101157752862947, + "learning_rate": 2.9852920002003252e-06, + "loss": 0.8107, + "step": 3947 + }, + { + "epoch": 0.3560445506605943, + "grad_norm": 1.7615539744388902, + "learning_rate": 2.9847835606694494e-06, + "loss": 1.0069, + "step": 3948 + }, + { + "epoch": 0.35613473418406455, + "grad_norm": 1.9988568734611505, + "learning_rate": 2.98427503711141e-06, + "loss": 0.8981, + "step": 3949 + }, + { + "epoch": 0.35622491770753484, + "grad_norm": 1.6362870040684268, + "learning_rate": 2.9837664295695973e-06, + "loss": 1.0026, + "step": 3950 + }, + { + "epoch": 0.35631510123100507, + "grad_norm": 1.273152840281359, + "learning_rate": 2.983257738087408e-06, + "loss": 0.9617, + "step": 3951 + }, + { + "epoch": 0.35640528475447536, + "grad_norm": 1.619327327331319, + "learning_rate": 2.982748962708247e-06, + "loss": 0.9981, + "step": 3952 + }, + { + "epoch": 0.3564954682779456, + "grad_norm": 1.5230950386315694, + "learning_rate": 2.982240103475526e-06, + "loss": 1.0217, + "step": 3953 + }, + { + "epoch": 0.3565856518014159, + "grad_norm": 1.9741152362986623, + "learning_rate": 2.981731160432663e-06, + "loss": 0.9711, + "step": 3954 + }, + { + "epoch": 0.35667583532488617, + "grad_norm": 1.5989652834398482, + "learning_rate": 2.981222133623084e-06, + "loss": 0.866, + "step": 3955 + }, + { + "epoch": 0.3567660188483564, + "grad_norm": 1.482040979557546, + "learning_rate": 2.980713023090222e-06, + "loss": 0.9761, + "step": 3956 + }, + { + "epoch": 0.3568562023718267, + "grad_norm": 1.4533361466115151, + "learning_rate": 2.980203828877518e-06, + "loss": 0.9383, + "step": 3957 + }, + { + "epoch": 0.3569463858952969, + "grad_norm": 1.7997604422633955, + "learning_rate": 2.9796945510284182e-06, + "loss": 1.0104, + "step": 3958 + }, + { + "epoch": 0.3570365694187672, + "grad_norm": 1.9364535674525056, + "learning_rate": 2.9791851895863774e-06, + "loss": 1.0255, + "step": 3959 + }, + { + "epoch": 0.35712675294223745, + "grad_norm": 1.5877691506330207, + "learning_rate": 2.978675744594857e-06, + "loss": 0.9969, + "step": 3960 + }, + { + "epoch": 0.35721693646570774, + "grad_norm": 1.4089292223029402, + "learning_rate": 2.978166216097326e-06, + "loss": 0.9252, + "step": 3961 + }, + { + "epoch": 0.35730711998917797, + "grad_norm": 1.617804736597166, + "learning_rate": 2.9776566041372596e-06, + "loss": 0.9549, + "step": 3962 + }, + { + "epoch": 0.35739730351264826, + "grad_norm": 1.4899677323041245, + "learning_rate": 2.977146908758141e-06, + "loss": 1.0255, + "step": 3963 + }, + { + "epoch": 0.3574874870361185, + "grad_norm": 1.431604986020423, + "learning_rate": 2.9766371300034604e-06, + "loss": 0.9825, + "step": 3964 + }, + { + "epoch": 0.3575776705595888, + "grad_norm": 1.4372503644379098, + "learning_rate": 2.9761272679167142e-06, + "loss": 1.0174, + "step": 3965 + }, + { + "epoch": 0.357667854083059, + "grad_norm": 1.5386781938843896, + "learning_rate": 2.9756173225414072e-06, + "loss": 0.8191, + "step": 3966 + }, + { + "epoch": 0.3577580376065293, + "grad_norm": 1.705144542261348, + "learning_rate": 2.975107293921051e-06, + "loss": 1.1439, + "step": 3967 + }, + { + "epoch": 0.35784822112999953, + "grad_norm": 1.5720464897270172, + "learning_rate": 2.9745971820991643e-06, + "loss": 1.0081, + "step": 3968 + }, + { + "epoch": 0.3579384046534698, + "grad_norm": 1.518771650999866, + "learning_rate": 2.9740869871192715e-06, + "loss": 1.0112, + "step": 3969 + }, + { + "epoch": 0.35802858817694005, + "grad_norm": 1.4107344015456653, + "learning_rate": 2.9735767090249065e-06, + "loss": 1.0287, + "step": 3970 + }, + { + "epoch": 0.35811877170041034, + "grad_norm": 1.631110221164979, + "learning_rate": 2.973066347859608e-06, + "loss": 0.8783, + "step": 3971 + }, + { + "epoch": 0.3582089552238806, + "grad_norm": 1.958485380993202, + "learning_rate": 2.972555903666923e-06, + "loss": 0.9833, + "step": 3972 + }, + { + "epoch": 0.35829913874735086, + "grad_norm": 2.1843140425792456, + "learning_rate": 2.972045376490406e-06, + "loss": 0.8842, + "step": 3973 + }, + { + "epoch": 0.3583893222708211, + "grad_norm": 1.405486471797036, + "learning_rate": 2.9715347663736177e-06, + "loss": 0.8935, + "step": 3974 + }, + { + "epoch": 0.3584795057942914, + "grad_norm": 1.4756688234186939, + "learning_rate": 2.9710240733601266e-06, + "loss": 0.969, + "step": 3975 + }, + { + "epoch": 0.3585696893177616, + "grad_norm": 1.8041305673466164, + "learning_rate": 2.970513297493507e-06, + "loss": 1.0531, + "step": 3976 + }, + { + "epoch": 0.3586598728412319, + "grad_norm": 1.4322302047951607, + "learning_rate": 2.9700024388173416e-06, + "loss": 1.0199, + "step": 3977 + }, + { + "epoch": 0.35875005636470214, + "grad_norm": 1.6058942319167204, + "learning_rate": 2.969491497375219e-06, + "loss": 1.024, + "step": 3978 + }, + { + "epoch": 0.35884023988817243, + "grad_norm": 1.428106882994614, + "learning_rate": 2.9689804732107364e-06, + "loss": 0.9915, + "step": 3979 + }, + { + "epoch": 0.3589304234116427, + "grad_norm": 1.7131056612545283, + "learning_rate": 2.9684693663674968e-06, + "loss": 0.9319, + "step": 3980 + }, + { + "epoch": 0.35902060693511295, + "grad_norm": 1.4942899741534335, + "learning_rate": 2.9679581768891115e-06, + "loss": 1.0529, + "step": 3981 + }, + { + "epoch": 0.35911079045858324, + "grad_norm": 1.4948701561004734, + "learning_rate": 2.967446904819197e-06, + "loss": 0.8938, + "step": 3982 + }, + { + "epoch": 0.3592009739820535, + "grad_norm": 1.5014457887510384, + "learning_rate": 2.966935550201378e-06, + "loss": 1.0001, + "step": 3983 + }, + { + "epoch": 0.35929115750552376, + "grad_norm": 1.796559115300662, + "learning_rate": 2.966424113079286e-06, + "loss": 0.9646, + "step": 3984 + }, + { + "epoch": 0.359381341028994, + "grad_norm": 1.4053969444944152, + "learning_rate": 2.9659125934965596e-06, + "loss": 1.0211, + "step": 3985 + }, + { + "epoch": 0.3594715245524643, + "grad_norm": 1.6192297088449505, + "learning_rate": 2.9654009914968457e-06, + "loss": 1.0311, + "step": 3986 + }, + { + "epoch": 0.3595617080759345, + "grad_norm": 1.911516200825904, + "learning_rate": 2.9648893071237956e-06, + "loss": 0.8645, + "step": 3987 + }, + { + "epoch": 0.3596518915994048, + "grad_norm": 1.5162208408404274, + "learning_rate": 2.964377540421069e-06, + "loss": 0.9441, + "step": 3988 + }, + { + "epoch": 0.35974207512287504, + "grad_norm": 1.774200208223546, + "learning_rate": 2.963865691432334e-06, + "loss": 0.9911, + "step": 3989 + }, + { + "epoch": 0.3598322586463453, + "grad_norm": 1.4506518740221177, + "learning_rate": 2.963353760201263e-06, + "loss": 0.8869, + "step": 3990 + }, + { + "epoch": 0.35992244216981556, + "grad_norm": 1.698574040319343, + "learning_rate": 2.962841746771537e-06, + "loss": 0.9568, + "step": 3991 + }, + { + "epoch": 0.36001262569328585, + "grad_norm": 1.4910789486293248, + "learning_rate": 2.9623296511868445e-06, + "loss": 0.9727, + "step": 3992 + }, + { + "epoch": 0.3601028092167561, + "grad_norm": 1.3580032530704913, + "learning_rate": 2.96181747349088e-06, + "loss": 0.8973, + "step": 3993 + }, + { + "epoch": 0.36019299274022637, + "grad_norm": 1.63425859726046, + "learning_rate": 2.961305213727345e-06, + "loss": 0.9884, + "step": 3994 + }, + { + "epoch": 0.3602831762636966, + "grad_norm": 0.7298350313678974, + "learning_rate": 2.960792871939949e-06, + "loss": 0.805, + "step": 3995 + }, + { + "epoch": 0.3603733597871669, + "grad_norm": 1.598560180132614, + "learning_rate": 2.9602804481724064e-06, + "loss": 1.0632, + "step": 3996 + }, + { + "epoch": 0.3604635433106371, + "grad_norm": 2.3608190904432966, + "learning_rate": 2.9597679424684427e-06, + "loss": 0.9606, + "step": 3997 + }, + { + "epoch": 0.3605537268341074, + "grad_norm": 1.6089509572010647, + "learning_rate": 2.9592553548717848e-06, + "loss": 0.9806, + "step": 3998 + }, + { + "epoch": 0.36064391035757765, + "grad_norm": 1.5610189666719916, + "learning_rate": 2.958742685426171e-06, + "loss": 1.0519, + "step": 3999 + }, + { + "epoch": 0.36073409388104793, + "grad_norm": 1.7170857087806737, + "learning_rate": 2.9582299341753446e-06, + "loss": 1.0429, + "step": 4000 + }, + { + "epoch": 0.36082427740451817, + "grad_norm": 1.630088176097547, + "learning_rate": 2.957717101163057e-06, + "loss": 0.9121, + "step": 4001 + }, + { + "epoch": 0.36091446092798846, + "grad_norm": 1.4198254076315608, + "learning_rate": 2.9572041864330655e-06, + "loss": 0.9701, + "step": 4002 + }, + { + "epoch": 0.36100464445145874, + "grad_norm": 1.5654950999814072, + "learning_rate": 2.9566911900291346e-06, + "loss": 0.9911, + "step": 4003 + }, + { + "epoch": 0.361094827974929, + "grad_norm": 1.5677755178796773, + "learning_rate": 2.9561781119950368e-06, + "loss": 0.8894, + "step": 4004 + }, + { + "epoch": 0.36118501149839927, + "grad_norm": 1.321435213992829, + "learning_rate": 2.9556649523745493e-06, + "loss": 0.9144, + "step": 4005 + }, + { + "epoch": 0.3612751950218695, + "grad_norm": 1.6684773225920777, + "learning_rate": 2.955151711211459e-06, + "loss": 1.0335, + "step": 4006 + }, + { + "epoch": 0.3613653785453398, + "grad_norm": 1.3408664548176121, + "learning_rate": 2.9546383885495583e-06, + "loss": 0.9404, + "step": 4007 + }, + { + "epoch": 0.36145556206881, + "grad_norm": 1.5384795348325278, + "learning_rate": 2.9541249844326464e-06, + "loss": 0.9227, + "step": 4008 + }, + { + "epoch": 0.3615457455922803, + "grad_norm": 1.5065267348223115, + "learning_rate": 2.9536114989045295e-06, + "loss": 0.9536, + "step": 4009 + }, + { + "epoch": 0.36163592911575054, + "grad_norm": 1.4831681013053566, + "learning_rate": 2.9530979320090216e-06, + "loss": 1.0316, + "step": 4010 + }, + { + "epoch": 0.36172611263922083, + "grad_norm": 1.7863593189911795, + "learning_rate": 2.9525842837899422e-06, + "loss": 1.0595, + "step": 4011 + }, + { + "epoch": 0.36181629616269106, + "grad_norm": 1.9819841174038795, + "learning_rate": 2.95207055429112e-06, + "loss": 1.0572, + "step": 4012 + }, + { + "epoch": 0.36190647968616135, + "grad_norm": 1.5219883995967511, + "learning_rate": 2.951556743556388e-06, + "loss": 0.9638, + "step": 4013 + }, + { + "epoch": 0.3619966632096316, + "grad_norm": 1.7242488635483788, + "learning_rate": 2.951042851629588e-06, + "loss": 1.0143, + "step": 4014 + }, + { + "epoch": 0.3620868467331019, + "grad_norm": 1.5619566163303698, + "learning_rate": 2.950528878554568e-06, + "loss": 0.9275, + "step": 4015 + }, + { + "epoch": 0.3621770302565721, + "grad_norm": 1.26381275741856, + "learning_rate": 2.950014824375183e-06, + "loss": 1.0095, + "step": 4016 + }, + { + "epoch": 0.3622672137800424, + "grad_norm": 0.7041337512485061, + "learning_rate": 2.949500689135295e-06, + "loss": 0.7292, + "step": 4017 + }, + { + "epoch": 0.36235739730351263, + "grad_norm": 1.414052847307675, + "learning_rate": 2.9489864728787722e-06, + "loss": 1.0515, + "step": 4018 + }, + { + "epoch": 0.3624475808269829, + "grad_norm": 1.523618012517651, + "learning_rate": 2.9484721756494915e-06, + "loss": 1.0337, + "step": 4019 + }, + { + "epoch": 0.36253776435045315, + "grad_norm": 1.6869244477019825, + "learning_rate": 2.9479577974913343e-06, + "loss": 0.9745, + "step": 4020 + }, + { + "epoch": 0.36262794787392344, + "grad_norm": 1.5033142509997406, + "learning_rate": 2.9474433384481908e-06, + "loss": 0.9695, + "step": 4021 + }, + { + "epoch": 0.3627181313973937, + "grad_norm": 1.4444250733144428, + "learning_rate": 2.9469287985639577e-06, + "loss": 0.9687, + "step": 4022 + }, + { + "epoch": 0.36280831492086396, + "grad_norm": 1.6437773437546415, + "learning_rate": 2.9464141778825384e-06, + "loss": 1.0156, + "step": 4023 + }, + { + "epoch": 0.3628984984443342, + "grad_norm": 1.6466475617436889, + "learning_rate": 2.9458994764478427e-06, + "loss": 1.0101, + "step": 4024 + }, + { + "epoch": 0.3629886819678045, + "grad_norm": 1.7726814574210035, + "learning_rate": 2.9453846943037883e-06, + "loss": 1.0065, + "step": 4025 + }, + { + "epoch": 0.36307886549127477, + "grad_norm": 1.9650949124231998, + "learning_rate": 2.9448698314942987e-06, + "loss": 0.9204, + "step": 4026 + }, + { + "epoch": 0.363169049014745, + "grad_norm": 0.9340084224398778, + "learning_rate": 2.944354888063305e-06, + "loss": 0.8259, + "step": 4027 + }, + { + "epoch": 0.3632592325382153, + "grad_norm": 1.4908147759061756, + "learning_rate": 2.9438398640547453e-06, + "loss": 0.9442, + "step": 4028 + }, + { + "epoch": 0.3633494160616855, + "grad_norm": 1.67174080283372, + "learning_rate": 2.943324759512564e-06, + "loss": 0.9173, + "step": 4029 + }, + { + "epoch": 0.3634395995851558, + "grad_norm": 1.654147991520209, + "learning_rate": 2.9428095744807134e-06, + "loss": 0.9983, + "step": 4030 + }, + { + "epoch": 0.36352978310862605, + "grad_norm": 1.4657755011939968, + "learning_rate": 2.942294309003151e-06, + "loss": 0.9303, + "step": 4031 + }, + { + "epoch": 0.36361996663209634, + "grad_norm": 2.5505197107614967, + "learning_rate": 2.941778963123843e-06, + "loss": 0.9923, + "step": 4032 + }, + { + "epoch": 0.36371015015556657, + "grad_norm": 1.4732587072767955, + "learning_rate": 2.94126353688676e-06, + "loss": 0.9744, + "step": 4033 + }, + { + "epoch": 0.36380033367903686, + "grad_norm": 2.031654552374573, + "learning_rate": 2.9407480303358825e-06, + "loss": 0.9202, + "step": 4034 + }, + { + "epoch": 0.3638905172025071, + "grad_norm": 1.4409984642407125, + "learning_rate": 2.940232443515195e-06, + "loss": 0.9794, + "step": 4035 + }, + { + "epoch": 0.3639807007259774, + "grad_norm": 1.5251965661861364, + "learning_rate": 2.9397167764686916e-06, + "loss": 0.9986, + "step": 4036 + }, + { + "epoch": 0.3640708842494476, + "grad_norm": 1.4528348079223314, + "learning_rate": 2.9392010292403714e-06, + "loss": 0.9479, + "step": 4037 + }, + { + "epoch": 0.3641610677729179, + "grad_norm": 1.650075812476556, + "learning_rate": 2.9386852018742404e-06, + "loss": 0.9636, + "step": 4038 + }, + { + "epoch": 0.36425125129638813, + "grad_norm": 1.480322032895356, + "learning_rate": 2.938169294414312e-06, + "loss": 0.9494, + "step": 4039 + }, + { + "epoch": 0.3643414348198584, + "grad_norm": 1.4932281699929595, + "learning_rate": 2.9376533069046067e-06, + "loss": 1.005, + "step": 4040 + }, + { + "epoch": 0.36443161834332866, + "grad_norm": 0.8128491165144129, + "learning_rate": 2.9371372393891514e-06, + "loss": 0.8368, + "step": 4041 + }, + { + "epoch": 0.36452180186679894, + "grad_norm": 1.2157502058414549, + "learning_rate": 2.936621091911979e-06, + "loss": 1.044, + "step": 4042 + }, + { + "epoch": 0.3646119853902692, + "grad_norm": 1.5602286805754815, + "learning_rate": 2.936104864517131e-06, + "loss": 0.902, + "step": 4043 + }, + { + "epoch": 0.36470216891373947, + "grad_norm": 1.6563235572459392, + "learning_rate": 2.9355885572486535e-06, + "loss": 0.9338, + "step": 4044 + }, + { + "epoch": 0.3647923524372097, + "grad_norm": 1.4659671798820924, + "learning_rate": 2.9350721701506026e-06, + "loss": 1.0039, + "step": 4045 + }, + { + "epoch": 0.36488253596068, + "grad_norm": 1.7176005594724357, + "learning_rate": 2.9345557032670375e-06, + "loss": 0.9636, + "step": 4046 + }, + { + "epoch": 0.3649727194841502, + "grad_norm": 1.9815762224189983, + "learning_rate": 2.934039156642027e-06, + "loss": 1.022, + "step": 4047 + }, + { + "epoch": 0.3650629030076205, + "grad_norm": 1.6387591696801997, + "learning_rate": 2.9335225303196454e-06, + "loss": 0.8282, + "step": 4048 + }, + { + "epoch": 0.36515308653109074, + "grad_norm": 1.6107878918001572, + "learning_rate": 2.933005824343974e-06, + "loss": 0.9406, + "step": 4049 + }, + { + "epoch": 0.36524327005456103, + "grad_norm": 1.6237897033770898, + "learning_rate": 2.932489038759101e-06, + "loss": 0.9457, + "step": 4050 + }, + { + "epoch": 0.3653334535780313, + "grad_norm": 1.3677129553169352, + "learning_rate": 2.9319721736091215e-06, + "loss": 0.8655, + "step": 4051 + }, + { + "epoch": 0.36542363710150155, + "grad_norm": 1.349621520146488, + "learning_rate": 2.9314552289381377e-06, + "loss": 0.8757, + "step": 4052 + }, + { + "epoch": 0.36551382062497184, + "grad_norm": 1.8068747208976728, + "learning_rate": 2.9309382047902574e-06, + "loss": 0.9404, + "step": 4053 + }, + { + "epoch": 0.3656040041484421, + "grad_norm": 1.2844180970026933, + "learning_rate": 2.9304211012095963e-06, + "loss": 1.014, + "step": 4054 + }, + { + "epoch": 0.36569418767191236, + "grad_norm": 1.624853274250262, + "learning_rate": 2.929903918240277e-06, + "loss": 1.0286, + "step": 4055 + }, + { + "epoch": 0.3657843711953826, + "grad_norm": 0.7548460168001018, + "learning_rate": 2.9293866559264273e-06, + "loss": 0.7805, + "step": 4056 + }, + { + "epoch": 0.3658745547188529, + "grad_norm": 1.6330901279514776, + "learning_rate": 2.928869314312184e-06, + "loss": 0.9749, + "step": 4057 + }, + { + "epoch": 0.3659647382423231, + "grad_norm": 1.6244044312896442, + "learning_rate": 2.9283518934416892e-06, + "loss": 0.9723, + "step": 4058 + }, + { + "epoch": 0.3660549217657934, + "grad_norm": 1.555202686310634, + "learning_rate": 2.927834393359092e-06, + "loss": 0.9841, + "step": 4059 + }, + { + "epoch": 0.36614510528926364, + "grad_norm": 1.678441778932869, + "learning_rate": 2.927316814108548e-06, + "loss": 1.0426, + "step": 4060 + }, + { + "epoch": 0.36623528881273393, + "grad_norm": 1.49041578955976, + "learning_rate": 2.92679915573422e-06, + "loss": 0.8744, + "step": 4061 + }, + { + "epoch": 0.36632547233620416, + "grad_norm": 1.4693586021945977, + "learning_rate": 2.926281418280278e-06, + "loss": 0.9112, + "step": 4062 + }, + { + "epoch": 0.36641565585967445, + "grad_norm": 1.512844405882883, + "learning_rate": 2.925763601790899e-06, + "loss": 0.9767, + "step": 4063 + }, + { + "epoch": 0.3665058393831447, + "grad_norm": 1.4138911865322294, + "learning_rate": 2.9252457063102635e-06, + "loss": 0.9143, + "step": 4064 + }, + { + "epoch": 0.36659602290661497, + "grad_norm": 1.3899198737480367, + "learning_rate": 2.9247277318825626e-06, + "loss": 0.9344, + "step": 4065 + }, + { + "epoch": 0.3666862064300852, + "grad_norm": 1.5258292960685238, + "learning_rate": 2.924209678551993e-06, + "loss": 1.0634, + "step": 4066 + }, + { + "epoch": 0.3667763899535555, + "grad_norm": 2.0286912500849867, + "learning_rate": 2.923691546362757e-06, + "loss": 0.9794, + "step": 4067 + }, + { + "epoch": 0.3668665734770257, + "grad_norm": 2.0503961692033066, + "learning_rate": 2.9231733353590663e-06, + "loss": 0.9278, + "step": 4068 + }, + { + "epoch": 0.366956757000496, + "grad_norm": 1.6431885808495923, + "learning_rate": 2.922655045585136e-06, + "loss": 1.0432, + "step": 4069 + }, + { + "epoch": 0.36704694052396625, + "grad_norm": 0.7503458258914817, + "learning_rate": 2.92213667708519e-06, + "loss": 0.8228, + "step": 4070 + }, + { + "epoch": 0.36713712404743654, + "grad_norm": 1.5277097609821209, + "learning_rate": 2.921618229903457e-06, + "loss": 0.98, + "step": 4071 + }, + { + "epoch": 0.36722730757090677, + "grad_norm": 2.164480010703689, + "learning_rate": 2.9210997040841752e-06, + "loss": 0.8929, + "step": 4072 + }, + { + "epoch": 0.36731749109437706, + "grad_norm": 1.6439711824429901, + "learning_rate": 2.9205810996715885e-06, + "loss": 0.936, + "step": 4073 + }, + { + "epoch": 0.36740767461784735, + "grad_norm": 1.445121133868277, + "learning_rate": 2.9200624167099456e-06, + "loss": 0.9351, + "step": 4074 + }, + { + "epoch": 0.3674978581413176, + "grad_norm": 1.3428896877945335, + "learning_rate": 2.919543655243505e-06, + "loss": 0.882, + "step": 4075 + }, + { + "epoch": 0.36758804166478787, + "grad_norm": 1.6177014257607383, + "learning_rate": 2.919024815316529e-06, + "loss": 0.9976, + "step": 4076 + }, + { + "epoch": 0.3676782251882581, + "grad_norm": 1.799606105303832, + "learning_rate": 2.9185058969732877e-06, + "loss": 1.0925, + "step": 4077 + }, + { + "epoch": 0.3677684087117284, + "grad_norm": 1.4747829309858524, + "learning_rate": 2.917986900258059e-06, + "loss": 1.0015, + "step": 4078 + }, + { + "epoch": 0.3678585922351986, + "grad_norm": 1.664499503051097, + "learning_rate": 2.917467825215126e-06, + "loss": 0.9652, + "step": 4079 + }, + { + "epoch": 0.3679487757586689, + "grad_norm": 1.4631575581522407, + "learning_rate": 2.9169486718887803e-06, + "loss": 0.9199, + "step": 4080 + }, + { + "epoch": 0.36803895928213914, + "grad_norm": 1.5673344399661808, + "learning_rate": 2.9164294403233173e-06, + "loss": 1.0215, + "step": 4081 + }, + { + "epoch": 0.36812914280560943, + "grad_norm": 1.7171150754558613, + "learning_rate": 2.915910130563041e-06, + "loss": 0.9025, + "step": 4082 + }, + { + "epoch": 0.36821932632907967, + "grad_norm": 1.8740065167862345, + "learning_rate": 2.915390742652262e-06, + "loss": 0.9144, + "step": 4083 + }, + { + "epoch": 0.36830950985254995, + "grad_norm": 1.8285896819238199, + "learning_rate": 2.914871276635298e-06, + "loss": 0.9006, + "step": 4084 + }, + { + "epoch": 0.3683996933760202, + "grad_norm": 1.7638719345973382, + "learning_rate": 2.914351732556472e-06, + "loss": 0.9476, + "step": 4085 + }, + { + "epoch": 0.3684898768994905, + "grad_norm": 1.3531876864460248, + "learning_rate": 2.9138321104601144e-06, + "loss": 0.8871, + "step": 4086 + }, + { + "epoch": 0.3685800604229607, + "grad_norm": 1.497027551447157, + "learning_rate": 2.9133124103905623e-06, + "loss": 0.9002, + "step": 4087 + }, + { + "epoch": 0.368670243946431, + "grad_norm": 1.8536805219227928, + "learning_rate": 2.9127926323921596e-06, + "loss": 0.9517, + "step": 4088 + }, + { + "epoch": 0.36876042746990123, + "grad_norm": 1.5409554396021967, + "learning_rate": 2.912272776509256e-06, + "loss": 0.9853, + "step": 4089 + }, + { + "epoch": 0.3688506109933715, + "grad_norm": 1.6520990140859446, + "learning_rate": 2.911752842786209e-06, + "loss": 0.9363, + "step": 4090 + }, + { + "epoch": 0.36894079451684175, + "grad_norm": 1.688841110245217, + "learning_rate": 2.911232831267383e-06, + "loss": 0.9133, + "step": 4091 + }, + { + "epoch": 0.36903097804031204, + "grad_norm": 1.5579055375732718, + "learning_rate": 2.910712741997146e-06, + "loss": 0.9863, + "step": 4092 + }, + { + "epoch": 0.3691211615637823, + "grad_norm": 1.4529199763221046, + "learning_rate": 2.910192575019877e-06, + "loss": 0.9592, + "step": 4093 + }, + { + "epoch": 0.36921134508725256, + "grad_norm": 1.375371232636848, + "learning_rate": 2.9096723303799583e-06, + "loss": 1.0005, + "step": 4094 + }, + { + "epoch": 0.3693015286107228, + "grad_norm": 1.5074044105247626, + "learning_rate": 2.9091520081217805e-06, + "loss": 0.976, + "step": 4095 + }, + { + "epoch": 0.3693917121341931, + "grad_norm": 1.4837374472468743, + "learning_rate": 2.908631608289741e-06, + "loss": 0.9159, + "step": 4096 + }, + { + "epoch": 0.3694818956576633, + "grad_norm": 1.8011185005468366, + "learning_rate": 2.9081111309282423e-06, + "loss": 0.9586, + "step": 4097 + }, + { + "epoch": 0.3695720791811336, + "grad_norm": 0.7143575547058146, + "learning_rate": 2.9075905760816942e-06, + "loss": 0.833, + "step": 4098 + }, + { + "epoch": 0.3696622627046039, + "grad_norm": 1.6781493485316414, + "learning_rate": 2.907069943794514e-06, + "loss": 1.0098, + "step": 4099 + }, + { + "epoch": 0.3697524462280741, + "grad_norm": 1.6755725223757039, + "learning_rate": 2.906549234111125e-06, + "loss": 0.9401, + "step": 4100 + }, + { + "epoch": 0.3698426297515444, + "grad_norm": 1.5954331131043673, + "learning_rate": 2.906028447075956e-06, + "loss": 1.0265, + "step": 4101 + }, + { + "epoch": 0.36993281327501465, + "grad_norm": 1.4726277558582885, + "learning_rate": 2.905507582733445e-06, + "loss": 1.0463, + "step": 4102 + }, + { + "epoch": 0.37002299679848494, + "grad_norm": 1.446672531693226, + "learning_rate": 2.904986641128033e-06, + "loss": 1.0842, + "step": 4103 + }, + { + "epoch": 0.37011318032195517, + "grad_norm": 1.36197553917406, + "learning_rate": 2.9044656223041716e-06, + "loss": 0.9556, + "step": 4104 + }, + { + "epoch": 0.37020336384542546, + "grad_norm": 1.3857209820297651, + "learning_rate": 2.9039445263063157e-06, + "loss": 0.9271, + "step": 4105 + }, + { + "epoch": 0.3702935473688957, + "grad_norm": 1.928467894101196, + "learning_rate": 2.903423353178929e-06, + "loss": 0.8606, + "step": 4106 + }, + { + "epoch": 0.370383730892366, + "grad_norm": 1.3986601838703183, + "learning_rate": 2.9029021029664802e-06, + "loss": 0.9636, + "step": 4107 + }, + { + "epoch": 0.3704739144158362, + "grad_norm": 1.4110340968256154, + "learning_rate": 2.9023807757134455e-06, + "loss": 1.0207, + "step": 4108 + }, + { + "epoch": 0.3705640979393065, + "grad_norm": 1.4674620967564613, + "learning_rate": 2.901859371464307e-06, + "loss": 1.0372, + "step": 4109 + }, + { + "epoch": 0.37065428146277674, + "grad_norm": 1.6913192057071522, + "learning_rate": 2.9013378902635535e-06, + "loss": 1.0057, + "step": 4110 + }, + { + "epoch": 0.370744464986247, + "grad_norm": 2.1596532308415, + "learning_rate": 2.9008163321556823e-06, + "loss": 0.9354, + "step": 4111 + }, + { + "epoch": 0.37083464850971726, + "grad_norm": 1.6247395526804256, + "learning_rate": 2.900294697185194e-06, + "loss": 0.9783, + "step": 4112 + }, + { + "epoch": 0.37092483203318755, + "grad_norm": 1.4211520358689866, + "learning_rate": 2.899772985396599e-06, + "loss": 0.9877, + "step": 4113 + }, + { + "epoch": 0.3710150155566578, + "grad_norm": 1.4443686212706672, + "learning_rate": 2.8992511968344104e-06, + "loss": 0.9628, + "step": 4114 + }, + { + "epoch": 0.37110519908012807, + "grad_norm": 1.4119585940744208, + "learning_rate": 2.8987293315431523e-06, + "loss": 0.9094, + "step": 4115 + }, + { + "epoch": 0.3711953826035983, + "grad_norm": 1.3917651271063074, + "learning_rate": 2.898207389567351e-06, + "loss": 0.8989, + "step": 4116 + }, + { + "epoch": 0.3712855661270686, + "grad_norm": 1.5316100962212686, + "learning_rate": 2.897685370951543e-06, + "loss": 0.9587, + "step": 4117 + }, + { + "epoch": 0.3713757496505388, + "grad_norm": 1.9018601071028163, + "learning_rate": 2.89716327574027e-06, + "loss": 0.8619, + "step": 4118 + }, + { + "epoch": 0.3714659331740091, + "grad_norm": 1.4310318093379888, + "learning_rate": 2.8966411039780787e-06, + "loss": 1.0549, + "step": 4119 + }, + { + "epoch": 0.37155611669747934, + "grad_norm": 1.575640772097861, + "learning_rate": 2.8961188557095248e-06, + "loss": 0.9713, + "step": 4120 + }, + { + "epoch": 0.37164630022094963, + "grad_norm": 1.759234925244931, + "learning_rate": 2.895596530979168e-06, + "loss": 0.908, + "step": 4121 + }, + { + "epoch": 0.3717364837444199, + "grad_norm": 1.6240150694594244, + "learning_rate": 2.895074129831578e-06, + "loss": 0.9776, + "step": 4122 + }, + { + "epoch": 0.37182666726789015, + "grad_norm": 1.6757655998868577, + "learning_rate": 2.8945516523113275e-06, + "loss": 0.8794, + "step": 4123 + }, + { + "epoch": 0.37191685079136044, + "grad_norm": 2.140981157286755, + "learning_rate": 2.894029098462998e-06, + "loss": 0.7534, + "step": 4124 + }, + { + "epoch": 0.3720070343148307, + "grad_norm": 1.3868430928105144, + "learning_rate": 2.8935064683311756e-06, + "loss": 0.9906, + "step": 4125 + }, + { + "epoch": 0.37209721783830096, + "grad_norm": 0.704225907476636, + "learning_rate": 2.8929837619604544e-06, + "loss": 0.8577, + "step": 4126 + }, + { + "epoch": 0.3721874013617712, + "grad_norm": 1.5882447095458914, + "learning_rate": 2.8924609793954346e-06, + "loss": 1.0162, + "step": 4127 + }, + { + "epoch": 0.3722775848852415, + "grad_norm": 1.8273081462469136, + "learning_rate": 2.891938120680724e-06, + "loss": 0.9269, + "step": 4128 + }, + { + "epoch": 0.3723677684087117, + "grad_norm": 1.6178608105442576, + "learning_rate": 2.8914151858609343e-06, + "loss": 0.9597, + "step": 4129 + }, + { + "epoch": 0.372457951932182, + "grad_norm": 2.870654263238235, + "learning_rate": 2.8908921749806858e-06, + "loss": 0.9858, + "step": 4130 + }, + { + "epoch": 0.37254813545565224, + "grad_norm": 1.3940453631218492, + "learning_rate": 2.890369088084605e-06, + "loss": 0.9799, + "step": 4131 + }, + { + "epoch": 0.37263831897912253, + "grad_norm": 1.5715623123675537, + "learning_rate": 2.889845925217323e-06, + "loss": 0.9927, + "step": 4132 + }, + { + "epoch": 0.37272850250259276, + "grad_norm": 1.5846205632044676, + "learning_rate": 2.8893226864234813e-06, + "loss": 0.9843, + "step": 4133 + }, + { + "epoch": 0.37281868602606305, + "grad_norm": 1.5351611421230495, + "learning_rate": 2.8887993717477236e-06, + "loss": 0.9808, + "step": 4134 + }, + { + "epoch": 0.3729088695495333, + "grad_norm": 1.6453560326188728, + "learning_rate": 2.8882759812347035e-06, + "loss": 0.8547, + "step": 4135 + }, + { + "epoch": 0.3729990530730036, + "grad_norm": 1.5448172397601776, + "learning_rate": 2.887752514929078e-06, + "loss": 0.9702, + "step": 4136 + }, + { + "epoch": 0.3730892365964738, + "grad_norm": 1.732464055490764, + "learning_rate": 2.887228972875513e-06, + "loss": 0.9219, + "step": 4137 + }, + { + "epoch": 0.3731794201199441, + "grad_norm": 1.6916178159127804, + "learning_rate": 2.88670535511868e-06, + "loss": 0.9692, + "step": 4138 + }, + { + "epoch": 0.3732696036434143, + "grad_norm": 1.869839815531063, + "learning_rate": 2.886181661703257e-06, + "loss": 0.919, + "step": 4139 + }, + { + "epoch": 0.3733597871668846, + "grad_norm": 1.6903278533920942, + "learning_rate": 2.8856578926739285e-06, + "loss": 0.933, + "step": 4140 + }, + { + "epoch": 0.37344997069035485, + "grad_norm": 1.707865705298278, + "learning_rate": 2.8851340480753846e-06, + "loss": 0.9232, + "step": 4141 + }, + { + "epoch": 0.37354015421382514, + "grad_norm": 1.7714991794964352, + "learning_rate": 2.8846101279523232e-06, + "loss": 0.9751, + "step": 4142 + }, + { + "epoch": 0.37363033773729537, + "grad_norm": 1.5511222867933514, + "learning_rate": 2.8840861323494487e-06, + "loss": 0.9416, + "step": 4143 + }, + { + "epoch": 0.37372052126076566, + "grad_norm": 2.762779021520348, + "learning_rate": 2.88356206131147e-06, + "loss": 1.0256, + "step": 4144 + }, + { + "epoch": 0.37381070478423595, + "grad_norm": 1.6407943638167721, + "learning_rate": 2.883037914883104e-06, + "loss": 0.9519, + "step": 4145 + }, + { + "epoch": 0.3739008883077062, + "grad_norm": 1.4560904648783983, + "learning_rate": 2.882513693109075e-06, + "loss": 0.9396, + "step": 4146 + }, + { + "epoch": 0.37399107183117647, + "grad_norm": 1.3717193480456766, + "learning_rate": 2.8819893960341106e-06, + "loss": 0.9448, + "step": 4147 + }, + { + "epoch": 0.3740812553546467, + "grad_norm": 2.211783203565665, + "learning_rate": 2.881465023702948e-06, + "loss": 1.025, + "step": 4148 + }, + { + "epoch": 0.374171438878117, + "grad_norm": 1.6351484171847899, + "learning_rate": 2.8809405761603294e-06, + "loss": 1.0556, + "step": 4149 + }, + { + "epoch": 0.3742616224015872, + "grad_norm": 1.7172992479029812, + "learning_rate": 2.880416053451003e-06, + "loss": 0.8888, + "step": 4150 + }, + { + "epoch": 0.3743518059250575, + "grad_norm": 1.674299959250475, + "learning_rate": 2.879891455619725e-06, + "loss": 1.0615, + "step": 4151 + }, + { + "epoch": 0.37444198944852775, + "grad_norm": 1.392331351609327, + "learning_rate": 2.879366782711256e-06, + "loss": 0.9754, + "step": 4152 + }, + { + "epoch": 0.37453217297199803, + "grad_norm": 1.4406209320871508, + "learning_rate": 2.8788420347703643e-06, + "loss": 0.9909, + "step": 4153 + }, + { + "epoch": 0.37462235649546827, + "grad_norm": 0.693480553581955, + "learning_rate": 2.8783172118418244e-06, + "loss": 0.7784, + "step": 4154 + }, + { + "epoch": 0.37471254001893856, + "grad_norm": 1.482661332202024, + "learning_rate": 2.877792313970417e-06, + "loss": 0.9772, + "step": 4155 + }, + { + "epoch": 0.3748027235424088, + "grad_norm": 1.3553758748801128, + "learning_rate": 2.8772673412009293e-06, + "loss": 1.0273, + "step": 4156 + }, + { + "epoch": 0.3748929070658791, + "grad_norm": 1.855537429341597, + "learning_rate": 2.8767422935781545e-06, + "loss": 1.0198, + "step": 4157 + }, + { + "epoch": 0.3749830905893493, + "grad_norm": 1.419307444622281, + "learning_rate": 2.8762171711468935e-06, + "loss": 1.0032, + "step": 4158 + }, + { + "epoch": 0.3750732741128196, + "grad_norm": 1.4675104307879503, + "learning_rate": 2.875691973951952e-06, + "loss": 0.9726, + "step": 4159 + }, + { + "epoch": 0.37516345763628983, + "grad_norm": 0.7058072744878849, + "learning_rate": 2.8751667020381425e-06, + "loss": 0.8495, + "step": 4160 + }, + { + "epoch": 0.3752536411597601, + "grad_norm": 1.6850365211988871, + "learning_rate": 2.8746413554502837e-06, + "loss": 1.0374, + "step": 4161 + }, + { + "epoch": 0.37534382468323035, + "grad_norm": 1.5610507347420195, + "learning_rate": 2.8741159342332027e-06, + "loss": 0.904, + "step": 4162 + }, + { + "epoch": 0.37543400820670064, + "grad_norm": 1.611395225283748, + "learning_rate": 2.87359043843173e-06, + "loss": 1.0305, + "step": 4163 + }, + { + "epoch": 0.3755241917301709, + "grad_norm": 1.875566969027501, + "learning_rate": 2.873064868090704e-06, + "loss": 1.0071, + "step": 4164 + }, + { + "epoch": 0.37561437525364116, + "grad_norm": 1.5342587087035042, + "learning_rate": 2.8725392232549697e-06, + "loss": 1.0626, + "step": 4165 + }, + { + "epoch": 0.3757045587771114, + "grad_norm": 1.7296815894490252, + "learning_rate": 2.872013503969378e-06, + "loss": 1.0447, + "step": 4166 + }, + { + "epoch": 0.3757947423005817, + "grad_norm": 1.3889154362790472, + "learning_rate": 2.8714877102787853e-06, + "loss": 0.982, + "step": 4167 + }, + { + "epoch": 0.3758849258240519, + "grad_norm": 1.7893037591703587, + "learning_rate": 2.8709618422280564e-06, + "loss": 0.9814, + "step": 4168 + }, + { + "epoch": 0.3759751093475222, + "grad_norm": 1.3520675773858548, + "learning_rate": 2.8704358998620605e-06, + "loss": 0.947, + "step": 4169 + }, + { + "epoch": 0.3760652928709925, + "grad_norm": 1.67228722169649, + "learning_rate": 2.8699098832256735e-06, + "loss": 1.0025, + "step": 4170 + }, + { + "epoch": 0.37615547639446273, + "grad_norm": 2.021028240544493, + "learning_rate": 2.86938379236378e-06, + "loss": 0.8847, + "step": 4171 + }, + { + "epoch": 0.376245659917933, + "grad_norm": 1.4475638510896738, + "learning_rate": 2.868857627321266e-06, + "loss": 0.8988, + "step": 4172 + }, + { + "epoch": 0.37633584344140325, + "grad_norm": 1.578946570136889, + "learning_rate": 2.8683313881430296e-06, + "loss": 0.9754, + "step": 4173 + }, + { + "epoch": 0.37642602696487354, + "grad_norm": 1.5477485887228044, + "learning_rate": 2.8678050748739706e-06, + "loss": 0.9653, + "step": 4174 + }, + { + "epoch": 0.37651621048834377, + "grad_norm": 1.4717409238877264, + "learning_rate": 2.8672786875589976e-06, + "loss": 0.945, + "step": 4175 + }, + { + "epoch": 0.37660639401181406, + "grad_norm": 1.7143609620791012, + "learning_rate": 2.866752226243025e-06, + "loss": 0.9244, + "step": 4176 + }, + { + "epoch": 0.3766965775352843, + "grad_norm": 1.7378441421764348, + "learning_rate": 2.8662256909709733e-06, + "loss": 1.0176, + "step": 4177 + }, + { + "epoch": 0.3767867610587546, + "grad_norm": 1.5473760698824632, + "learning_rate": 2.865699081787769e-06, + "loss": 1.0118, + "step": 4178 + }, + { + "epoch": 0.3768769445822248, + "grad_norm": 1.4092924478718176, + "learning_rate": 2.8651723987383465e-06, + "loss": 0.9566, + "step": 4179 + }, + { + "epoch": 0.3769671281056951, + "grad_norm": 1.7443156565703553, + "learning_rate": 2.8646456418676437e-06, + "loss": 0.9485, + "step": 4180 + }, + { + "epoch": 0.37705731162916534, + "grad_norm": 0.7096373018478878, + "learning_rate": 2.8641188112206067e-06, + "loss": 0.7985, + "step": 4181 + }, + { + "epoch": 0.3771474951526356, + "grad_norm": 1.8477864451436723, + "learning_rate": 2.863591906842189e-06, + "loss": 1.001, + "step": 4182 + }, + { + "epoch": 0.37723767867610586, + "grad_norm": 1.5811797688082314, + "learning_rate": 2.863064928777347e-06, + "loss": 1.0036, + "step": 4183 + }, + { + "epoch": 0.37732786219957615, + "grad_norm": 5.957170848774137, + "learning_rate": 2.862537877071047e-06, + "loss": 0.9444, + "step": 4184 + }, + { + "epoch": 0.3774180457230464, + "grad_norm": 0.7564410748548962, + "learning_rate": 2.8620107517682597e-06, + "loss": 0.7615, + "step": 4185 + }, + { + "epoch": 0.37750822924651667, + "grad_norm": 1.3867741963221998, + "learning_rate": 2.8614835529139618e-06, + "loss": 0.9057, + "step": 4186 + }, + { + "epoch": 0.3775984127699869, + "grad_norm": 1.5189870924008846, + "learning_rate": 2.8609562805531367e-06, + "loss": 0.9304, + "step": 4187 + }, + { + "epoch": 0.3776885962934572, + "grad_norm": 1.589128112938751, + "learning_rate": 2.8604289347307746e-06, + "loss": 1.0069, + "step": 4188 + }, + { + "epoch": 0.3777787798169274, + "grad_norm": 1.7534027396390304, + "learning_rate": 2.859901515491871e-06, + "loss": 0.9652, + "step": 4189 + }, + { + "epoch": 0.3778689633403977, + "grad_norm": 1.3831845661201334, + "learning_rate": 2.8593740228814298e-06, + "loss": 0.8597, + "step": 4190 + }, + { + "epoch": 0.37795914686386795, + "grad_norm": 1.7260887911557774, + "learning_rate": 2.8588464569444574e-06, + "loss": 0.9537, + "step": 4191 + }, + { + "epoch": 0.37804933038733823, + "grad_norm": 1.4963489444303841, + "learning_rate": 2.8583188177259697e-06, + "loss": 0.9646, + "step": 4192 + }, + { + "epoch": 0.3781395139108085, + "grad_norm": 1.760559648053149, + "learning_rate": 2.857791105270988e-06, + "loss": 0.9569, + "step": 4193 + }, + { + "epoch": 0.37822969743427876, + "grad_norm": 1.567692483038211, + "learning_rate": 2.857263319624539e-06, + "loss": 0.9999, + "step": 4194 + }, + { + "epoch": 0.37831988095774904, + "grad_norm": 1.4651482430925071, + "learning_rate": 2.856735460831657e-06, + "loss": 0.951, + "step": 4195 + }, + { + "epoch": 0.3784100644812193, + "grad_norm": 1.4398216904074777, + "learning_rate": 2.856207528937382e-06, + "loss": 0.9789, + "step": 4196 + }, + { + "epoch": 0.37850024800468957, + "grad_norm": 1.5803920212438296, + "learning_rate": 2.855679523986759e-06, + "loss": 0.9577, + "step": 4197 + }, + { + "epoch": 0.3785904315281598, + "grad_norm": 1.5080782008903286, + "learning_rate": 2.8551514460248406e-06, + "loss": 1.0173, + "step": 4198 + }, + { + "epoch": 0.3786806150516301, + "grad_norm": 1.7044550050955118, + "learning_rate": 2.8546232950966868e-06, + "loss": 0.9713, + "step": 4199 + }, + { + "epoch": 0.3787707985751003, + "grad_norm": 1.7470488869476464, + "learning_rate": 2.85409507124736e-06, + "loss": 0.918, + "step": 4200 + }, + { + "epoch": 0.3788609820985706, + "grad_norm": 1.477912568196989, + "learning_rate": 2.8535667745219324e-06, + "loss": 0.9693, + "step": 4201 + }, + { + "epoch": 0.37895116562204084, + "grad_norm": 1.557057400267068, + "learning_rate": 2.853038404965481e-06, + "loss": 0.9595, + "step": 4202 + }, + { + "epoch": 0.37904134914551113, + "grad_norm": 1.6321710668178817, + "learning_rate": 2.8525099626230894e-06, + "loss": 1.0382, + "step": 4203 + }, + { + "epoch": 0.37913153266898136, + "grad_norm": 1.5227299538825563, + "learning_rate": 2.8519814475398472e-06, + "loss": 0.8973, + "step": 4204 + }, + { + "epoch": 0.37922171619245165, + "grad_norm": 1.3832309326892955, + "learning_rate": 2.8514528597608502e-06, + "loss": 0.9471, + "step": 4205 + }, + { + "epoch": 0.3793118997159219, + "grad_norm": 1.2484179498815149, + "learning_rate": 2.8509241993312004e-06, + "loss": 1.002, + "step": 4206 + }, + { + "epoch": 0.3794020832393922, + "grad_norm": 1.5207548338954193, + "learning_rate": 2.850395466296006e-06, + "loss": 1.0264, + "step": 4207 + }, + { + "epoch": 0.3794922667628624, + "grad_norm": 3.4737754365663025, + "learning_rate": 2.849866660700381e-06, + "loss": 0.9982, + "step": 4208 + }, + { + "epoch": 0.3795824502863327, + "grad_norm": 1.730611549840414, + "learning_rate": 2.8493377825894464e-06, + "loss": 1.0339, + "step": 4209 + }, + { + "epoch": 0.37967263380980293, + "grad_norm": 2.162738822141747, + "learning_rate": 2.848808832008329e-06, + "loss": 1.0326, + "step": 4210 + }, + { + "epoch": 0.3797628173332732, + "grad_norm": 1.4460941160130463, + "learning_rate": 2.848279809002162e-06, + "loss": 1.0006, + "step": 4211 + }, + { + "epoch": 0.37985300085674345, + "grad_norm": 1.840989130915312, + "learning_rate": 2.8477507136160842e-06, + "loss": 0.9889, + "step": 4212 + }, + { + "epoch": 0.37994318438021374, + "grad_norm": 0.7007075608038963, + "learning_rate": 2.847221545895241e-06, + "loss": 0.7913, + "step": 4213 + }, + { + "epoch": 0.38003336790368397, + "grad_norm": 1.4160765746580974, + "learning_rate": 2.846692305884785e-06, + "loss": 0.9655, + "step": 4214 + }, + { + "epoch": 0.38012355142715426, + "grad_norm": 1.5600524042570878, + "learning_rate": 2.8461629936298718e-06, + "loss": 1.0134, + "step": 4215 + }, + { + "epoch": 0.3802137349506245, + "grad_norm": 1.8937977422858898, + "learning_rate": 2.845633609175666e-06, + "loss": 1.0284, + "step": 4216 + }, + { + "epoch": 0.3803039184740948, + "grad_norm": 2.3681246976431893, + "learning_rate": 2.8451041525673383e-06, + "loss": 1.01, + "step": 4217 + }, + { + "epoch": 0.38039410199756507, + "grad_norm": 1.7255866283335488, + "learning_rate": 2.8445746238500647e-06, + "loss": 1.0039, + "step": 4218 + }, + { + "epoch": 0.3804842855210353, + "grad_norm": 1.7786876250942925, + "learning_rate": 2.844045023069027e-06, + "loss": 0.9508, + "step": 4219 + }, + { + "epoch": 0.3805744690445056, + "grad_norm": 2.1421112692504685, + "learning_rate": 2.8435153502694136e-06, + "loss": 1.0149, + "step": 4220 + }, + { + "epoch": 0.3806646525679758, + "grad_norm": 1.6609465748636805, + "learning_rate": 2.84298560549642e-06, + "loss": 0.9065, + "step": 4221 + }, + { + "epoch": 0.3807548360914461, + "grad_norm": 2.693122836785146, + "learning_rate": 2.8424557887952462e-06, + "loss": 1.0003, + "step": 4222 + }, + { + "epoch": 0.38084501961491635, + "grad_norm": 1.6709124922927665, + "learning_rate": 2.841925900211099e-06, + "loss": 0.9022, + "step": 4223 + }, + { + "epoch": 0.38093520313838664, + "grad_norm": 1.3776040128288352, + "learning_rate": 2.841395939789192e-06, + "loss": 1.0249, + "step": 4224 + }, + { + "epoch": 0.38102538666185687, + "grad_norm": 1.5435401341416999, + "learning_rate": 2.8408659075747435e-06, + "loss": 0.9943, + "step": 4225 + }, + { + "epoch": 0.38111557018532716, + "grad_norm": 2.0918717571286978, + "learning_rate": 2.8403358036129796e-06, + "loss": 0.9774, + "step": 4226 + }, + { + "epoch": 0.3812057537087974, + "grad_norm": 1.676414032376796, + "learning_rate": 2.839805627949132e-06, + "loss": 1.0288, + "step": 4227 + }, + { + "epoch": 0.3812959372322677, + "grad_norm": 1.6047595608734417, + "learning_rate": 2.8392753806284367e-06, + "loss": 1.0289, + "step": 4228 + }, + { + "epoch": 0.3813861207557379, + "grad_norm": 2.3712170237338084, + "learning_rate": 2.838745061696139e-06, + "loss": 0.9255, + "step": 4229 + }, + { + "epoch": 0.3814763042792082, + "grad_norm": 2.612905523776947, + "learning_rate": 2.838214671197487e-06, + "loss": 0.917, + "step": 4230 + }, + { + "epoch": 0.38156648780267843, + "grad_norm": 1.5368371356458836, + "learning_rate": 2.8376842091777377e-06, + "loss": 1.0845, + "step": 4231 + }, + { + "epoch": 0.3816566713261487, + "grad_norm": 1.5211545613676811, + "learning_rate": 2.8371536756821524e-06, + "loss": 0.8799, + "step": 4232 + }, + { + "epoch": 0.38174685484961896, + "grad_norm": 1.5729440682489635, + "learning_rate": 2.836623070756e-06, + "loss": 0.9601, + "step": 4233 + }, + { + "epoch": 0.38183703837308924, + "grad_norm": 1.5918301257071643, + "learning_rate": 2.8360923944445542e-06, + "loss": 0.999, + "step": 4234 + }, + { + "epoch": 0.3819272218965595, + "grad_norm": 1.4337103273977498, + "learning_rate": 2.8355616467930947e-06, + "loss": 0.9614, + "step": 4235 + }, + { + "epoch": 0.38201740542002977, + "grad_norm": 1.7624087452091866, + "learning_rate": 2.8350308278469085e-06, + "loss": 0.9591, + "step": 4236 + }, + { + "epoch": 0.3821075889435, + "grad_norm": 1.5165563651557623, + "learning_rate": 2.8344999376512877e-06, + "loss": 1.0109, + "step": 4237 + }, + { + "epoch": 0.3821977724669703, + "grad_norm": 1.622891525380691, + "learning_rate": 2.8339689762515307e-06, + "loss": 0.9681, + "step": 4238 + }, + { + "epoch": 0.3822879559904405, + "grad_norm": 1.4922244881244826, + "learning_rate": 2.8334379436929424e-06, + "loss": 0.9075, + "step": 4239 + }, + { + "epoch": 0.3823781395139108, + "grad_norm": 1.5308863928246261, + "learning_rate": 2.832906840020833e-06, + "loss": 0.9466, + "step": 4240 + }, + { + "epoch": 0.3824683230373811, + "grad_norm": 1.7836295179004675, + "learning_rate": 2.83237566528052e-06, + "loss": 0.9856, + "step": 4241 + }, + { + "epoch": 0.38255850656085133, + "grad_norm": 1.4457390233189955, + "learning_rate": 2.831844419517325e-06, + "loss": 0.9562, + "step": 4242 + }, + { + "epoch": 0.3826486900843216, + "grad_norm": 1.4784099183284514, + "learning_rate": 2.8313131027765774e-06, + "loss": 1.0068, + "step": 4243 + }, + { + "epoch": 0.38273887360779185, + "grad_norm": 2.686808563710136, + "learning_rate": 2.8307817151036124e-06, + "loss": 0.947, + "step": 4244 + }, + { + "epoch": 0.38282905713126214, + "grad_norm": 1.5736883514667428, + "learning_rate": 2.8302502565437704e-06, + "loss": 0.9546, + "step": 4245 + }, + { + "epoch": 0.3829192406547324, + "grad_norm": 1.780634991399196, + "learning_rate": 2.829718727142398e-06, + "loss": 0.9024, + "step": 4246 + }, + { + "epoch": 0.38300942417820266, + "grad_norm": 1.8933168904280013, + "learning_rate": 2.829187126944849e-06, + "loss": 0.9885, + "step": 4247 + }, + { + "epoch": 0.3830996077016729, + "grad_norm": 1.7944019050135642, + "learning_rate": 2.8286554559964826e-06, + "loss": 0.9207, + "step": 4248 + }, + { + "epoch": 0.3831897912251432, + "grad_norm": 1.5848642861540962, + "learning_rate": 2.8281237143426637e-06, + "loss": 0.9918, + "step": 4249 + }, + { + "epoch": 0.3832799747486134, + "grad_norm": 1.3346413464918445, + "learning_rate": 2.8275919020287626e-06, + "loss": 0.8801, + "step": 4250 + }, + { + "epoch": 0.3833701582720837, + "grad_norm": 1.675549542251565, + "learning_rate": 2.827060019100158e-06, + "loss": 1.0059, + "step": 4251 + }, + { + "epoch": 0.38346034179555394, + "grad_norm": 1.4824632078497508, + "learning_rate": 2.8265280656022315e-06, + "loss": 0.967, + "step": 4252 + }, + { + "epoch": 0.3835505253190242, + "grad_norm": 1.8298188537827156, + "learning_rate": 2.825996041580373e-06, + "loss": 0.8609, + "step": 4253 + }, + { + "epoch": 0.38364070884249446, + "grad_norm": 1.5930333489700665, + "learning_rate": 2.825463947079978e-06, + "loss": 0.9959, + "step": 4254 + }, + { + "epoch": 0.38373089236596475, + "grad_norm": 1.364287259351359, + "learning_rate": 2.8249317821464483e-06, + "loss": 0.9798, + "step": 4255 + }, + { + "epoch": 0.383821075889435, + "grad_norm": 1.7616609618865402, + "learning_rate": 2.824399546825189e-06, + "loss": 0.9067, + "step": 4256 + }, + { + "epoch": 0.38391125941290527, + "grad_norm": 1.592452399891375, + "learning_rate": 2.823867241161616e-06, + "loss": 0.9305, + "step": 4257 + }, + { + "epoch": 0.3840014429363755, + "grad_norm": 1.7547983054743095, + "learning_rate": 2.8233348652011456e-06, + "loss": 1.0276, + "step": 4258 + }, + { + "epoch": 0.3840916264598458, + "grad_norm": 1.7606536284405347, + "learning_rate": 2.8228024189892057e-06, + "loss": 0.8863, + "step": 4259 + }, + { + "epoch": 0.384181809983316, + "grad_norm": 1.298084017198963, + "learning_rate": 2.822269902571226e-06, + "loss": 0.977, + "step": 4260 + }, + { + "epoch": 0.3842719935067863, + "grad_norm": 1.5711425561214627, + "learning_rate": 2.8217373159926446e-06, + "loss": 0.955, + "step": 4261 + }, + { + "epoch": 0.38436217703025655, + "grad_norm": 1.2903450798564946, + "learning_rate": 2.8212046592989046e-06, + "loss": 0.9108, + "step": 4262 + }, + { + "epoch": 0.38445236055372684, + "grad_norm": 1.5276421062840497, + "learning_rate": 2.820671932535455e-06, + "loss": 0.9843, + "step": 4263 + }, + { + "epoch": 0.3845425440771971, + "grad_norm": 1.73859551835658, + "learning_rate": 2.8201391357477506e-06, + "loss": 0.9624, + "step": 4264 + }, + { + "epoch": 0.38463272760066736, + "grad_norm": 1.3229643244932328, + "learning_rate": 2.8196062689812525e-06, + "loss": 0.9496, + "step": 4265 + }, + { + "epoch": 0.38472291112413765, + "grad_norm": 1.5038109687247871, + "learning_rate": 2.819073332281429e-06, + "loss": 0.9215, + "step": 4266 + }, + { + "epoch": 0.3848130946476079, + "grad_norm": 0.7456782517198978, + "learning_rate": 2.8185403256937524e-06, + "loss": 0.7511, + "step": 4267 + }, + { + "epoch": 0.38490327817107817, + "grad_norm": 1.8067588643740318, + "learning_rate": 2.8180072492637016e-06, + "loss": 0.9379, + "step": 4268 + }, + { + "epoch": 0.3849934616945484, + "grad_norm": 1.5903126917767911, + "learning_rate": 2.817474103036762e-06, + "loss": 0.9633, + "step": 4269 + }, + { + "epoch": 0.3850836452180187, + "grad_norm": 1.6035414959659349, + "learning_rate": 2.816940887058425e-06, + "loss": 1.0023, + "step": 4270 + }, + { + "epoch": 0.3851738287414889, + "grad_norm": 1.5863922669782506, + "learning_rate": 2.816407601374186e-06, + "loss": 1.0161, + "step": 4271 + }, + { + "epoch": 0.3852640122649592, + "grad_norm": 1.5006051432333682, + "learning_rate": 2.815874246029549e-06, + "loss": 0.9066, + "step": 4272 + }, + { + "epoch": 0.38535419578842944, + "grad_norm": 1.4078927088167938, + "learning_rate": 2.815340821070023e-06, + "loss": 0.9237, + "step": 4273 + }, + { + "epoch": 0.38544437931189973, + "grad_norm": 2.199877245686142, + "learning_rate": 2.814807326541122e-06, + "loss": 0.9543, + "step": 4274 + }, + { + "epoch": 0.38553456283536996, + "grad_norm": 1.3923006142368672, + "learning_rate": 2.8142737624883676e-06, + "loss": 0.9047, + "step": 4275 + }, + { + "epoch": 0.38562474635884025, + "grad_norm": 1.425220703300315, + "learning_rate": 2.8137401289572854e-06, + "loss": 0.9753, + "step": 4276 + }, + { + "epoch": 0.3857149298823105, + "grad_norm": 1.6171764428686761, + "learning_rate": 2.8132064259934086e-06, + "loss": 1.0195, + "step": 4277 + }, + { + "epoch": 0.3858051134057808, + "grad_norm": 1.6037139584066056, + "learning_rate": 2.812672653642276e-06, + "loss": 0.9676, + "step": 4278 + }, + { + "epoch": 0.385895296929251, + "grad_norm": 1.4799897317272395, + "learning_rate": 2.812138811949431e-06, + "loss": 0.9818, + "step": 4279 + }, + { + "epoch": 0.3859854804527213, + "grad_norm": 1.5077021266881014, + "learning_rate": 2.8116049009604247e-06, + "loss": 1.0384, + "step": 4280 + }, + { + "epoch": 0.38607566397619153, + "grad_norm": 1.4739867658085701, + "learning_rate": 2.8110709207208132e-06, + "loss": 0.8535, + "step": 4281 + }, + { + "epoch": 0.3861658474996618, + "grad_norm": 1.6536641357690602, + "learning_rate": 2.810536871276158e-06, + "loss": 1.0012, + "step": 4282 + }, + { + "epoch": 0.38625603102313205, + "grad_norm": 0.8834522933450037, + "learning_rate": 2.8100027526720283e-06, + "loss": 0.8319, + "step": 4283 + }, + { + "epoch": 0.38634621454660234, + "grad_norm": 1.5266092783262766, + "learning_rate": 2.8094685649539974e-06, + "loss": 0.8976, + "step": 4284 + }, + { + "epoch": 0.3864363980700726, + "grad_norm": 0.8600485156480075, + "learning_rate": 2.8089343081676455e-06, + "loss": 0.8038, + "step": 4285 + }, + { + "epoch": 0.38652658159354286, + "grad_norm": 1.7859196939857316, + "learning_rate": 2.8083999823585577e-06, + "loss": 0.9555, + "step": 4286 + }, + { + "epoch": 0.3866167651170131, + "grad_norm": 1.4025343656494473, + "learning_rate": 2.8078655875723254e-06, + "loss": 0.996, + "step": 4287 + }, + { + "epoch": 0.3867069486404834, + "grad_norm": 2.1744529737771883, + "learning_rate": 2.807331123854547e-06, + "loss": 1.0567, + "step": 4288 + }, + { + "epoch": 0.38679713216395367, + "grad_norm": 1.944701930057982, + "learning_rate": 2.806796591250826e-06, + "loss": 0.8419, + "step": 4289 + }, + { + "epoch": 0.3868873156874239, + "grad_norm": 1.6939135989749408, + "learning_rate": 2.8062619898067707e-06, + "loss": 1.037, + "step": 4290 + }, + { + "epoch": 0.3869774992108942, + "grad_norm": 1.5494654379698616, + "learning_rate": 2.8057273195679963e-06, + "loss": 0.9876, + "step": 4291 + }, + { + "epoch": 0.3870676827343644, + "grad_norm": 1.6084816684739596, + "learning_rate": 2.8051925805801253e-06, + "loss": 0.9902, + "step": 4292 + }, + { + "epoch": 0.3871578662578347, + "grad_norm": 1.533960939932118, + "learning_rate": 2.804657772888783e-06, + "loss": 1.0061, + "step": 4293 + }, + { + "epoch": 0.38724804978130495, + "grad_norm": 1.6958750832906948, + "learning_rate": 2.804122896539602e-06, + "loss": 0.9693, + "step": 4294 + }, + { + "epoch": 0.38733823330477524, + "grad_norm": 1.470539950251279, + "learning_rate": 2.8035879515782225e-06, + "loss": 0.9257, + "step": 4295 + }, + { + "epoch": 0.38742841682824547, + "grad_norm": 1.7749253808039016, + "learning_rate": 2.803052938050288e-06, + "loss": 1.0241, + "step": 4296 + }, + { + "epoch": 0.38751860035171576, + "grad_norm": 1.9350388184431433, + "learning_rate": 2.802517856001449e-06, + "loss": 1.0215, + "step": 4297 + }, + { + "epoch": 0.387608783875186, + "grad_norm": 1.8426410120053622, + "learning_rate": 2.801982705477361e-06, + "loss": 0.9408, + "step": 4298 + }, + { + "epoch": 0.3876989673986563, + "grad_norm": 1.3455891554009907, + "learning_rate": 2.8014474865236867e-06, + "loss": 0.9588, + "step": 4299 + }, + { + "epoch": 0.3877891509221265, + "grad_norm": 1.5172000841272977, + "learning_rate": 2.800912199186094e-06, + "loss": 0.9646, + "step": 4300 + }, + { + "epoch": 0.3878793344455968, + "grad_norm": 1.4662608705135556, + "learning_rate": 2.800376843510256e-06, + "loss": 0.9795, + "step": 4301 + }, + { + "epoch": 0.38796951796906703, + "grad_norm": 1.4713132689426889, + "learning_rate": 2.799841419541852e-06, + "loss": 0.9922, + "step": 4302 + }, + { + "epoch": 0.3880597014925373, + "grad_norm": 0.7120832077466729, + "learning_rate": 2.799305927326568e-06, + "loss": 0.7858, + "step": 4303 + }, + { + "epoch": 0.38814988501600756, + "grad_norm": 0.7822274006852172, + "learning_rate": 2.7987703669100955e-06, + "loss": 0.796, + "step": 4304 + }, + { + "epoch": 0.38824006853947785, + "grad_norm": 1.5706998432754125, + "learning_rate": 2.79823473833813e-06, + "loss": 0.977, + "step": 4305 + }, + { + "epoch": 0.3883302520629481, + "grad_norm": 2.7846379593689194, + "learning_rate": 2.797699041656376e-06, + "loss": 0.9189, + "step": 4306 + }, + { + "epoch": 0.38842043558641837, + "grad_norm": 1.6839875294003888, + "learning_rate": 2.7971632769105412e-06, + "loss": 0.9881, + "step": 4307 + }, + { + "epoch": 0.3885106191098886, + "grad_norm": 1.7789651370950568, + "learning_rate": 2.79662744414634e-06, + "loss": 0.9361, + "step": 4308 + }, + { + "epoch": 0.3886008026333589, + "grad_norm": 1.7423682012202981, + "learning_rate": 2.7960915434094923e-06, + "loss": 0.9248, + "step": 4309 + }, + { + "epoch": 0.3886909861568291, + "grad_norm": 2.24658293331255, + "learning_rate": 2.7955555747457256e-06, + "loss": 0.9461, + "step": 4310 + }, + { + "epoch": 0.3887811696802994, + "grad_norm": 1.4276821894051568, + "learning_rate": 2.79501953820077e-06, + "loss": 0.9729, + "step": 4311 + }, + { + "epoch": 0.3888713532037697, + "grad_norm": 1.4663594046880324, + "learning_rate": 2.7944834338203637e-06, + "loss": 0.9884, + "step": 4312 + }, + { + "epoch": 0.38896153672723993, + "grad_norm": 1.6313782009449656, + "learning_rate": 2.79394726165025e-06, + "loss": 0.9314, + "step": 4313 + }, + { + "epoch": 0.3890517202507102, + "grad_norm": 1.999838107232473, + "learning_rate": 2.793411021736178e-06, + "loss": 0.9478, + "step": 4314 + }, + { + "epoch": 0.38914190377418045, + "grad_norm": 1.6386903527619572, + "learning_rate": 2.7928747141239027e-06, + "loss": 1.0596, + "step": 4315 + }, + { + "epoch": 0.38923208729765074, + "grad_norm": 0.73235200675866, + "learning_rate": 2.7923383388591856e-06, + "loss": 0.7893, + "step": 4316 + }, + { + "epoch": 0.389322270821121, + "grad_norm": 1.4542663357889212, + "learning_rate": 2.7918018959877923e-06, + "loss": 0.9354, + "step": 4317 + }, + { + "epoch": 0.38941245434459126, + "grad_norm": 1.4045940822136087, + "learning_rate": 2.791265385555495e-06, + "loss": 0.9959, + "step": 4318 + }, + { + "epoch": 0.3895026378680615, + "grad_norm": 1.4585033680926311, + "learning_rate": 2.790728807608072e-06, + "loss": 1.0345, + "step": 4319 + }, + { + "epoch": 0.3895928213915318, + "grad_norm": 2.0795204238787504, + "learning_rate": 2.790192162191307e-06, + "loss": 0.956, + "step": 4320 + }, + { + "epoch": 0.389683004915002, + "grad_norm": 1.8049209245600748, + "learning_rate": 2.78965544935099e-06, + "loss": 1.0035, + "step": 4321 + }, + { + "epoch": 0.3897731884384723, + "grad_norm": 1.7749884457279297, + "learning_rate": 2.789118669132916e-06, + "loss": 0.9515, + "step": 4322 + }, + { + "epoch": 0.38986337196194254, + "grad_norm": 1.774011594637683, + "learning_rate": 2.7885818215828856e-06, + "loss": 1.0372, + "step": 4323 + }, + { + "epoch": 0.38995355548541283, + "grad_norm": 1.8297300547171913, + "learning_rate": 2.7880449067467064e-06, + "loss": 0.9288, + "step": 4324 + }, + { + "epoch": 0.39004373900888306, + "grad_norm": 2.3075880516172718, + "learning_rate": 2.78750792467019e-06, + "loss": 0.9209, + "step": 4325 + }, + { + "epoch": 0.39013392253235335, + "grad_norm": 1.761348372474378, + "learning_rate": 2.786970875399156e-06, + "loss": 0.8349, + "step": 4326 + }, + { + "epoch": 0.3902241060558236, + "grad_norm": 1.36653064752597, + "learning_rate": 2.7864337589794267e-06, + "loss": 0.9023, + "step": 4327 + }, + { + "epoch": 0.39031428957929387, + "grad_norm": 1.5103336267554517, + "learning_rate": 2.7858965754568335e-06, + "loss": 0.9875, + "step": 4328 + }, + { + "epoch": 0.3904044731027641, + "grad_norm": 4.914196406785436, + "learning_rate": 2.785359324877211e-06, + "loss": 0.9095, + "step": 4329 + }, + { + "epoch": 0.3904946566262344, + "grad_norm": 1.430617483376771, + "learning_rate": 2.7848220072864e-06, + "loss": 0.9947, + "step": 4330 + }, + { + "epoch": 0.3905848401497046, + "grad_norm": 1.4941973507154709, + "learning_rate": 2.784284622730248e-06, + "loss": 0.9591, + "step": 4331 + }, + { + "epoch": 0.3906750236731749, + "grad_norm": 1.6391534654596094, + "learning_rate": 2.7837471712546073e-06, + "loss": 0.982, + "step": 4332 + }, + { + "epoch": 0.39076520719664515, + "grad_norm": 1.3660750304458678, + "learning_rate": 2.783209652905337e-06, + "loss": 0.9428, + "step": 4333 + }, + { + "epoch": 0.39085539072011544, + "grad_norm": 1.4782362236775806, + "learning_rate": 2.7826720677283e-06, + "loss": 0.9333, + "step": 4334 + }, + { + "epoch": 0.39094557424358567, + "grad_norm": 1.4219696516454934, + "learning_rate": 2.782134415769367e-06, + "loss": 0.9315, + "step": 4335 + }, + { + "epoch": 0.39103575776705596, + "grad_norm": 1.4282385148692198, + "learning_rate": 2.7815966970744126e-06, + "loss": 1.0105, + "step": 4336 + }, + { + "epoch": 0.39112594129052625, + "grad_norm": 1.7053011369317883, + "learning_rate": 2.7810589116893184e-06, + "loss": 1.0339, + "step": 4337 + }, + { + "epoch": 0.3912161248139965, + "grad_norm": 1.4485554372432026, + "learning_rate": 2.780521059659972e-06, + "loss": 0.9753, + "step": 4338 + }, + { + "epoch": 0.39130630833746677, + "grad_norm": 1.624978872308566, + "learning_rate": 2.7799831410322637e-06, + "loss": 0.9527, + "step": 4339 + }, + { + "epoch": 0.391396491860937, + "grad_norm": 1.375845649074159, + "learning_rate": 2.779445155852094e-06, + "loss": 0.9209, + "step": 4340 + }, + { + "epoch": 0.3914866753844073, + "grad_norm": 0.7495743020237924, + "learning_rate": 2.7789071041653655e-06, + "loss": 0.828, + "step": 4341 + }, + { + "epoch": 0.3915768589078775, + "grad_norm": 1.5413978015728067, + "learning_rate": 2.7783689860179875e-06, + "loss": 1.0241, + "step": 4342 + }, + { + "epoch": 0.3916670424313478, + "grad_norm": 1.4985312582681871, + "learning_rate": 2.7778308014558767e-06, + "loss": 0.9019, + "step": 4343 + }, + { + "epoch": 0.39175722595481804, + "grad_norm": 1.5224224132192947, + "learning_rate": 2.7772925505249524e-06, + "loss": 1.0396, + "step": 4344 + }, + { + "epoch": 0.39184740947828833, + "grad_norm": 1.3715762949328247, + "learning_rate": 2.7767542332711417e-06, + "loss": 0.9389, + "step": 4345 + }, + { + "epoch": 0.39193759300175857, + "grad_norm": 1.4165211584124766, + "learning_rate": 2.776215849740377e-06, + "loss": 0.9391, + "step": 4346 + }, + { + "epoch": 0.39202777652522885, + "grad_norm": 1.5086217892459564, + "learning_rate": 2.775677399978596e-06, + "loss": 0.9808, + "step": 4347 + }, + { + "epoch": 0.3921179600486991, + "grad_norm": 0.6557466302131401, + "learning_rate": 2.775138884031742e-06, + "loss": 0.7745, + "step": 4348 + }, + { + "epoch": 0.3922081435721694, + "grad_norm": 1.3224575517616994, + "learning_rate": 2.774600301945764e-06, + "loss": 0.9244, + "step": 4349 + }, + { + "epoch": 0.3922983270956396, + "grad_norm": 1.6387465849840093, + "learning_rate": 2.774061653766618e-06, + "loss": 0.9966, + "step": 4350 + }, + { + "epoch": 0.3923885106191099, + "grad_norm": 1.5955720845348922, + "learning_rate": 2.773522939540263e-06, + "loss": 0.9639, + "step": 4351 + }, + { + "epoch": 0.39247869414258013, + "grad_norm": 1.7003779524110316, + "learning_rate": 2.7729841593126663e-06, + "loss": 0.8921, + "step": 4352 + }, + { + "epoch": 0.3925688776660504, + "grad_norm": 1.39738919482489, + "learning_rate": 2.7724453131297988e-06, + "loss": 1.0335, + "step": 4353 + }, + { + "epoch": 0.39265906118952065, + "grad_norm": 1.5379173549725316, + "learning_rate": 2.771906401037637e-06, + "loss": 0.9928, + "step": 4354 + }, + { + "epoch": 0.39274924471299094, + "grad_norm": 1.6301781973526395, + "learning_rate": 2.7713674230821664e-06, + "loss": 0.9269, + "step": 4355 + }, + { + "epoch": 0.3928394282364612, + "grad_norm": 1.4109706904329602, + "learning_rate": 2.7708283793093724e-06, + "loss": 0.9207, + "step": 4356 + }, + { + "epoch": 0.39292961175993146, + "grad_norm": 1.4951075398478328, + "learning_rate": 2.7702892697652514e-06, + "loss": 0.9378, + "step": 4357 + }, + { + "epoch": 0.3930197952834017, + "grad_norm": 1.6632387513285332, + "learning_rate": 2.7697500944958024e-06, + "loss": 0.943, + "step": 4358 + }, + { + "epoch": 0.393109978806872, + "grad_norm": 1.7972621998114844, + "learning_rate": 2.7692108535470312e-06, + "loss": 1.065, + "step": 4359 + }, + { + "epoch": 0.3932001623303423, + "grad_norm": 2.1101424339997426, + "learning_rate": 2.768671546964948e-06, + "loss": 1.0043, + "step": 4360 + }, + { + "epoch": 0.3932903458538125, + "grad_norm": 1.6005293328154397, + "learning_rate": 2.7681321747955713e-06, + "loss": 0.9921, + "step": 4361 + }, + { + "epoch": 0.3933805293772828, + "grad_norm": 1.3873401807551586, + "learning_rate": 2.767592737084921e-06, + "loss": 0.9826, + "step": 4362 + }, + { + "epoch": 0.39347071290075303, + "grad_norm": 1.6773726014426356, + "learning_rate": 2.767053233879026e-06, + "loss": 1.0516, + "step": 4363 + }, + { + "epoch": 0.3935608964242233, + "grad_norm": 1.5045205187021142, + "learning_rate": 2.76651366522392e-06, + "loss": 1.0447, + "step": 4364 + }, + { + "epoch": 0.39365107994769355, + "grad_norm": 1.6749887437584556, + "learning_rate": 2.7659740311656413e-06, + "loss": 0.9389, + "step": 4365 + }, + { + "epoch": 0.39374126347116384, + "grad_norm": 1.7737046767312135, + "learning_rate": 2.7654343317502352e-06, + "loss": 0.9911, + "step": 4366 + }, + { + "epoch": 0.39383144699463407, + "grad_norm": 1.5163063012446845, + "learning_rate": 2.7648945670237502e-06, + "loss": 0.9474, + "step": 4367 + }, + { + "epoch": 0.39392163051810436, + "grad_norm": 2.1864273711240534, + "learning_rate": 2.7643547370322446e-06, + "loss": 1.0236, + "step": 4368 + }, + { + "epoch": 0.3940118140415746, + "grad_norm": 1.6643424358326562, + "learning_rate": 2.7638148418217775e-06, + "loss": 1.0022, + "step": 4369 + }, + { + "epoch": 0.3941019975650449, + "grad_norm": 1.6322077311065926, + "learning_rate": 2.7632748814384163e-06, + "loss": 1.0101, + "step": 4370 + }, + { + "epoch": 0.3941921810885151, + "grad_norm": 1.2713385760601452, + "learning_rate": 2.7627348559282335e-06, + "loss": 0.9933, + "step": 4371 + }, + { + "epoch": 0.3942823646119854, + "grad_norm": 1.552282111562113, + "learning_rate": 2.7621947653373075e-06, + "loss": 0.9056, + "step": 4372 + }, + { + "epoch": 0.39437254813545564, + "grad_norm": 0.706484672920074, + "learning_rate": 2.7616546097117213e-06, + "loss": 0.7651, + "step": 4373 + }, + { + "epoch": 0.3944627316589259, + "grad_norm": 1.579296319448641, + "learning_rate": 2.761114389097564e-06, + "loss": 0.956, + "step": 4374 + }, + { + "epoch": 0.39455291518239616, + "grad_norm": 1.5098056565834044, + "learning_rate": 2.7605741035409305e-06, + "loss": 1.0034, + "step": 4375 + }, + { + "epoch": 0.39464309870586645, + "grad_norm": 1.4475574276457557, + "learning_rate": 2.76003375308792e-06, + "loss": 0.9424, + "step": 4376 + }, + { + "epoch": 0.3947332822293367, + "grad_norm": 1.3382224012664954, + "learning_rate": 2.75949333778464e-06, + "loss": 0.9308, + "step": 4377 + }, + { + "epoch": 0.39482346575280697, + "grad_norm": 1.4725725468897182, + "learning_rate": 2.7589528576772e-06, + "loss": 0.8475, + "step": 4378 + }, + { + "epoch": 0.3949136492762772, + "grad_norm": 1.713443262430395, + "learning_rate": 2.758412312811717e-06, + "loss": 0.9543, + "step": 4379 + }, + { + "epoch": 0.3950038327997475, + "grad_norm": 0.7887014186062345, + "learning_rate": 2.7578717032343146e-06, + "loss": 0.8318, + "step": 4380 + }, + { + "epoch": 0.3950940163232177, + "grad_norm": 1.5911777907271993, + "learning_rate": 2.757331028991119e-06, + "loss": 0.9912, + "step": 4381 + }, + { + "epoch": 0.395184199846688, + "grad_norm": 2.3602502475217566, + "learning_rate": 2.7567902901282642e-06, + "loss": 0.9757, + "step": 4382 + }, + { + "epoch": 0.3952743833701583, + "grad_norm": 1.9313959134481602, + "learning_rate": 2.7562494866918892e-06, + "loss": 0.9582, + "step": 4383 + }, + { + "epoch": 0.39536456689362853, + "grad_norm": 1.366066696700192, + "learning_rate": 2.7557086187281378e-06, + "loss": 0.9396, + "step": 4384 + }, + { + "epoch": 0.3954547504170988, + "grad_norm": 1.5788483425508717, + "learning_rate": 2.75516768628316e-06, + "loss": 1.0639, + "step": 4385 + }, + { + "epoch": 0.39554493394056905, + "grad_norm": 1.4928311063434638, + "learning_rate": 2.7546266894031114e-06, + "loss": 1.0396, + "step": 4386 + }, + { + "epoch": 0.39563511746403934, + "grad_norm": 2.1075520868528037, + "learning_rate": 2.7540856281341526e-06, + "loss": 0.9147, + "step": 4387 + }, + { + "epoch": 0.3957253009875096, + "grad_norm": 1.5579111234463086, + "learning_rate": 2.7535445025224506e-06, + "loss": 0.9324, + "step": 4388 + }, + { + "epoch": 0.39581548451097986, + "grad_norm": 1.821011519570027, + "learning_rate": 2.753003312614176e-06, + "loss": 0.9673, + "step": 4389 + }, + { + "epoch": 0.3959056680344501, + "grad_norm": 1.68745853231837, + "learning_rate": 2.7524620584555065e-06, + "loss": 0.9667, + "step": 4390 + }, + { + "epoch": 0.3959958515579204, + "grad_norm": 2.0765785830917576, + "learning_rate": 2.7519207400926253e-06, + "loss": 1.0013, + "step": 4391 + }, + { + "epoch": 0.3960860350813906, + "grad_norm": 1.4681804445465902, + "learning_rate": 2.751379357571721e-06, + "loss": 0.949, + "step": 4392 + }, + { + "epoch": 0.3961762186048609, + "grad_norm": 1.4668658731743072, + "learning_rate": 2.7508379109389865e-06, + "loss": 1.0293, + "step": 4393 + }, + { + "epoch": 0.39626640212833114, + "grad_norm": 1.529448247168037, + "learning_rate": 2.750296400240622e-06, + "loss": 0.9798, + "step": 4394 + }, + { + "epoch": 0.39635658565180143, + "grad_norm": 2.648693601215765, + "learning_rate": 2.7497548255228305e-06, + "loss": 0.9971, + "step": 4395 + }, + { + "epoch": 0.39644676917527166, + "grad_norm": 1.484163048318732, + "learning_rate": 2.749213186831824e-06, + "loss": 0.9141, + "step": 4396 + }, + { + "epoch": 0.39653695269874195, + "grad_norm": 1.5388689258273673, + "learning_rate": 2.7486714842138173e-06, + "loss": 1.0123, + "step": 4397 + }, + { + "epoch": 0.3966271362222122, + "grad_norm": 1.5896649084328764, + "learning_rate": 2.748129717715031e-06, + "loss": 0.9163, + "step": 4398 + }, + { + "epoch": 0.3967173197456825, + "grad_norm": 1.6112937969828198, + "learning_rate": 2.747587887381692e-06, + "loss": 0.9823, + "step": 4399 + }, + { + "epoch": 0.3968075032691527, + "grad_norm": 1.3712098596864932, + "learning_rate": 2.7470459932600328e-06, + "loss": 0.9119, + "step": 4400 + }, + { + "epoch": 0.396897686792623, + "grad_norm": 1.6248884162739257, + "learning_rate": 2.7465040353962897e-06, + "loss": 0.9654, + "step": 4401 + }, + { + "epoch": 0.3969878703160932, + "grad_norm": 1.6622549636498207, + "learning_rate": 2.745962013836706e-06, + "loss": 0.9568, + "step": 4402 + }, + { + "epoch": 0.3970780538395635, + "grad_norm": 1.5432832421535863, + "learning_rate": 2.74541992862753e-06, + "loss": 0.9391, + "step": 4403 + }, + { + "epoch": 0.39716823736303375, + "grad_norm": 1.8898941392871167, + "learning_rate": 2.744877779815016e-06, + "loss": 0.9656, + "step": 4404 + }, + { + "epoch": 0.39725842088650404, + "grad_norm": 1.4247612636085454, + "learning_rate": 2.7443355674454234e-06, + "loss": 0.9611, + "step": 4405 + }, + { + "epoch": 0.39734860440997427, + "grad_norm": 1.8360206098210063, + "learning_rate": 2.743793291565015e-06, + "loss": 1.0076, + "step": 4406 + }, + { + "epoch": 0.39743878793344456, + "grad_norm": 1.964208359867224, + "learning_rate": 2.7432509522200617e-06, + "loss": 0.8927, + "step": 4407 + }, + { + "epoch": 0.39752897145691485, + "grad_norm": 1.8008336785684043, + "learning_rate": 2.7427085494568383e-06, + "loss": 0.9095, + "step": 4408 + }, + { + "epoch": 0.3976191549803851, + "grad_norm": 0.7430658864603419, + "learning_rate": 2.742166083321628e-06, + "loss": 0.8085, + "step": 4409 + }, + { + "epoch": 0.39770933850385537, + "grad_norm": 1.59643552820696, + "learning_rate": 2.7416235538607137e-06, + "loss": 1.0087, + "step": 4410 + }, + { + "epoch": 0.3977995220273256, + "grad_norm": 1.4278474233309189, + "learning_rate": 2.7410809611203894e-06, + "loss": 0.9448, + "step": 4411 + }, + { + "epoch": 0.3978897055507959, + "grad_norm": 1.5937616684430391, + "learning_rate": 2.7405383051469507e-06, + "loss": 0.9234, + "step": 4412 + }, + { + "epoch": 0.3979798890742661, + "grad_norm": 1.517337343023485, + "learning_rate": 2.7399955859867e-06, + "loss": 0.7798, + "step": 4413 + }, + { + "epoch": 0.3980700725977364, + "grad_norm": 0.7517184200853073, + "learning_rate": 2.7394528036859465e-06, + "loss": 0.7524, + "step": 4414 + }, + { + "epoch": 0.39816025612120665, + "grad_norm": 1.52872929082585, + "learning_rate": 2.738909958291002e-06, + "loss": 0.9667, + "step": 4415 + }, + { + "epoch": 0.39825043964467693, + "grad_norm": 1.3111855873455083, + "learning_rate": 2.7383670498481863e-06, + "loss": 0.8785, + "step": 4416 + }, + { + "epoch": 0.39834062316814717, + "grad_norm": 0.7420811627152955, + "learning_rate": 2.737824078403822e-06, + "loss": 0.7113, + "step": 4417 + }, + { + "epoch": 0.39843080669161746, + "grad_norm": 1.6426593338603503, + "learning_rate": 2.737281044004239e-06, + "loss": 0.9411, + "step": 4418 + }, + { + "epoch": 0.3985209902150877, + "grad_norm": 1.560631130269102, + "learning_rate": 2.736737946695772e-06, + "loss": 0.9496, + "step": 4419 + }, + { + "epoch": 0.398611173738558, + "grad_norm": 1.7560607319455115, + "learning_rate": 2.736194786524761e-06, + "loss": 0.9615, + "step": 4420 + }, + { + "epoch": 0.3987013572620282, + "grad_norm": 1.4901166196922988, + "learning_rate": 2.7356515635375517e-06, + "loss": 0.9582, + "step": 4421 + }, + { + "epoch": 0.3987915407854985, + "grad_norm": 2.1281313544583225, + "learning_rate": 2.735108277780495e-06, + "loss": 0.8769, + "step": 4422 + }, + { + "epoch": 0.39888172430896873, + "grad_norm": 1.5218486619115459, + "learning_rate": 2.7345649292999456e-06, + "loss": 0.9535, + "step": 4423 + }, + { + "epoch": 0.398971907832439, + "grad_norm": 1.5881575656811184, + "learning_rate": 2.734021518142267e-06, + "loss": 0.9686, + "step": 4424 + }, + { + "epoch": 0.39906209135590925, + "grad_norm": 1.6407513524401123, + "learning_rate": 2.733478044353825e-06, + "loss": 0.7998, + "step": 4425 + }, + { + "epoch": 0.39915227487937954, + "grad_norm": 1.6970907281417449, + "learning_rate": 2.7329345079809917e-06, + "loss": 0.9713, + "step": 4426 + }, + { + "epoch": 0.3992424584028498, + "grad_norm": 1.4444201214690577, + "learning_rate": 2.7323909090701447e-06, + "loss": 0.902, + "step": 4427 + }, + { + "epoch": 0.39933264192632006, + "grad_norm": 1.570977293276577, + "learning_rate": 2.731847247667667e-06, + "loss": 1.0161, + "step": 4428 + }, + { + "epoch": 0.3994228254497903, + "grad_norm": 1.4523285354547848, + "learning_rate": 2.731303523819947e-06, + "loss": 0.9364, + "step": 4429 + }, + { + "epoch": 0.3995130089732606, + "grad_norm": 1.77137098191298, + "learning_rate": 2.7307597375733783e-06, + "loss": 1.0255, + "step": 4430 + }, + { + "epoch": 0.3996031924967309, + "grad_norm": 1.5790487173246335, + "learning_rate": 2.7302158889743587e-06, + "loss": 0.9661, + "step": 4431 + }, + { + "epoch": 0.3996933760202011, + "grad_norm": 1.6279435274466059, + "learning_rate": 2.7296719780692937e-06, + "loss": 0.9226, + "step": 4432 + }, + { + "epoch": 0.3997835595436714, + "grad_norm": 1.5010658292217656, + "learning_rate": 2.7291280049045916e-06, + "loss": 0.8745, + "step": 4433 + }, + { + "epoch": 0.39987374306714163, + "grad_norm": 1.3383637642795396, + "learning_rate": 2.7285839695266683e-06, + "loss": 0.9012, + "step": 4434 + }, + { + "epoch": 0.3999639265906119, + "grad_norm": 1.4698080756309864, + "learning_rate": 2.7280398719819423e-06, + "loss": 0.9128, + "step": 4435 + }, + { + "epoch": 0.40005411011408215, + "grad_norm": 1.7879877318701707, + "learning_rate": 2.727495712316841e-06, + "loss": 1.0291, + "step": 4436 + }, + { + "epoch": 0.40014429363755244, + "grad_norm": 2.2316143347501827, + "learning_rate": 2.7269514905777945e-06, + "loss": 0.9058, + "step": 4437 + }, + { + "epoch": 0.4002344771610227, + "grad_norm": 2.0965547137955243, + "learning_rate": 2.7264072068112377e-06, + "loss": 1.0842, + "step": 4438 + }, + { + "epoch": 0.40032466068449296, + "grad_norm": 1.7252149890221096, + "learning_rate": 2.7258628610636133e-06, + "loss": 0.9245, + "step": 4439 + }, + { + "epoch": 0.4004148442079632, + "grad_norm": 1.5410864055899067, + "learning_rate": 2.7253184533813667e-06, + "loss": 1.0386, + "step": 4440 + }, + { + "epoch": 0.4005050277314335, + "grad_norm": 1.6080927509109126, + "learning_rate": 2.72477398381095e-06, + "loss": 0.9721, + "step": 4441 + }, + { + "epoch": 0.4005952112549037, + "grad_norm": 1.3300040920452756, + "learning_rate": 2.724229452398821e-06, + "loss": 0.972, + "step": 4442 + }, + { + "epoch": 0.400685394778374, + "grad_norm": 2.0164181826806336, + "learning_rate": 2.7236848591914422e-06, + "loss": 0.9841, + "step": 4443 + }, + { + "epoch": 0.40077557830184424, + "grad_norm": 1.540015917732907, + "learning_rate": 2.7231402042352803e-06, + "loss": 0.9446, + "step": 4444 + }, + { + "epoch": 0.4008657618253145, + "grad_norm": 1.288966782223995, + "learning_rate": 2.722595487576809e-06, + "loss": 0.9928, + "step": 4445 + }, + { + "epoch": 0.40095594534878476, + "grad_norm": 1.6714394260468572, + "learning_rate": 2.722050709262506e-06, + "loss": 1.009, + "step": 4446 + }, + { + "epoch": 0.40104612887225505, + "grad_norm": 1.4372157769553668, + "learning_rate": 2.7215058693388557e-06, + "loss": 0.9455, + "step": 4447 + }, + { + "epoch": 0.4011363123957253, + "grad_norm": 1.7135716892692692, + "learning_rate": 2.720960967852346e-06, + "loss": 0.866, + "step": 4448 + }, + { + "epoch": 0.40122649591919557, + "grad_norm": 1.6218241689136454, + "learning_rate": 2.720416004849471e-06, + "loss": 0.8931, + "step": 4449 + }, + { + "epoch": 0.4013166794426658, + "grad_norm": 1.4308239531760363, + "learning_rate": 2.7198709803767304e-06, + "loss": 0.9135, + "step": 4450 + }, + { + "epoch": 0.4014068629661361, + "grad_norm": 1.4449759426556261, + "learning_rate": 2.7193258944806286e-06, + "loss": 0.9342, + "step": 4451 + }, + { + "epoch": 0.4014970464896063, + "grad_norm": 1.81880066266892, + "learning_rate": 2.718780747207675e-06, + "loss": 0.9656, + "step": 4452 + }, + { + "epoch": 0.4015872300130766, + "grad_norm": 1.7513158483330638, + "learning_rate": 2.7182355386043847e-06, + "loss": 1.0272, + "step": 4453 + }, + { + "epoch": 0.40167741353654685, + "grad_norm": 1.458486939469514, + "learning_rate": 2.717690268717278e-06, + "loss": 1.0228, + "step": 4454 + }, + { + "epoch": 0.40176759706001713, + "grad_norm": 1.3570151179293088, + "learning_rate": 2.7171449375928803e-06, + "loss": 0.9863, + "step": 4455 + }, + { + "epoch": 0.4018577805834874, + "grad_norm": 1.696322232692381, + "learning_rate": 2.716599545277722e-06, + "loss": 0.993, + "step": 4456 + }, + { + "epoch": 0.40194796410695766, + "grad_norm": 1.5905550928248784, + "learning_rate": 2.7160540918183394e-06, + "loss": 0.9239, + "step": 4457 + }, + { + "epoch": 0.40203814763042794, + "grad_norm": 1.7777992419430604, + "learning_rate": 2.715508577261273e-06, + "loss": 0.8813, + "step": 4458 + }, + { + "epoch": 0.4021283311538982, + "grad_norm": 1.7510835154303992, + "learning_rate": 2.7149630016530702e-06, + "loss": 0.943, + "step": 4459 + }, + { + "epoch": 0.40221851467736847, + "grad_norm": 1.3670437546503773, + "learning_rate": 2.7144173650402815e-06, + "loss": 0.9792, + "step": 4460 + }, + { + "epoch": 0.4023086982008387, + "grad_norm": 1.5264027225899957, + "learning_rate": 2.7138716674694636e-06, + "loss": 0.9827, + "step": 4461 + }, + { + "epoch": 0.402398881724309, + "grad_norm": 1.771132327998634, + "learning_rate": 2.7133259089871795e-06, + "loss": 0.9125, + "step": 4462 + }, + { + "epoch": 0.4024890652477792, + "grad_norm": 1.8546909348312233, + "learning_rate": 2.712780089639995e-06, + "loss": 0.9608, + "step": 4463 + }, + { + "epoch": 0.4025792487712495, + "grad_norm": 1.4837479722636635, + "learning_rate": 2.712234209474483e-06, + "loss": 0.8404, + "step": 4464 + }, + { + "epoch": 0.40266943229471974, + "grad_norm": 1.5324513918960594, + "learning_rate": 2.7116882685372218e-06, + "loss": 1.0028, + "step": 4465 + }, + { + "epoch": 0.40275961581819003, + "grad_norm": 1.7935347318240127, + "learning_rate": 2.7111422668747927e-06, + "loss": 0.9292, + "step": 4466 + }, + { + "epoch": 0.40284979934166026, + "grad_norm": 1.6198531585829297, + "learning_rate": 2.7105962045337846e-06, + "loss": 0.9586, + "step": 4467 + }, + { + "epoch": 0.40293998286513055, + "grad_norm": 1.5180677412094081, + "learning_rate": 2.7100500815607898e-06, + "loss": 0.9547, + "step": 4468 + }, + { + "epoch": 0.4030301663886008, + "grad_norm": 1.2821844926427408, + "learning_rate": 2.709503898002407e-06, + "loss": 0.9829, + "step": 4469 + }, + { + "epoch": 0.4031203499120711, + "grad_norm": 1.6345849888767503, + "learning_rate": 2.708957653905239e-06, + "loss": 0.9577, + "step": 4470 + }, + { + "epoch": 0.4032105334355413, + "grad_norm": 1.4486060479623137, + "learning_rate": 2.7084113493158956e-06, + "loss": 0.9761, + "step": 4471 + }, + { + "epoch": 0.4033007169590116, + "grad_norm": 2.0288581263696175, + "learning_rate": 2.7078649842809888e-06, + "loss": 0.9727, + "step": 4472 + }, + { + "epoch": 0.40339090048248183, + "grad_norm": 1.4505131536950495, + "learning_rate": 2.707318558847139e-06, + "loss": 1.0103, + "step": 4473 + }, + { + "epoch": 0.4034810840059521, + "grad_norm": 0.7377361727652944, + "learning_rate": 2.7067720730609697e-06, + "loss": 0.7811, + "step": 4474 + }, + { + "epoch": 0.40357126752942235, + "grad_norm": 1.5271529056509565, + "learning_rate": 2.70622552696911e-06, + "loss": 0.9356, + "step": 4475 + }, + { + "epoch": 0.40366145105289264, + "grad_norm": 1.4549495768248828, + "learning_rate": 2.7056789206181943e-06, + "loss": 0.9144, + "step": 4476 + }, + { + "epoch": 0.40375163457636287, + "grad_norm": 1.4767462399113414, + "learning_rate": 2.7051322540548615e-06, + "loss": 1.0163, + "step": 4477 + }, + { + "epoch": 0.40384181809983316, + "grad_norm": 1.61368191895306, + "learning_rate": 2.704585527325757e-06, + "loss": 0.933, + "step": 4478 + }, + { + "epoch": 0.40393200162330345, + "grad_norm": 1.4317243897625571, + "learning_rate": 2.7040387404775303e-06, + "loss": 0.9204, + "step": 4479 + }, + { + "epoch": 0.4040221851467737, + "grad_norm": 1.5583119532024556, + "learning_rate": 2.703491893556837e-06, + "loss": 0.9916, + "step": 4480 + }, + { + "epoch": 0.40411236867024397, + "grad_norm": 1.3345534980258444, + "learning_rate": 2.702944986610335e-06, + "loss": 0.935, + "step": 4481 + }, + { + "epoch": 0.4042025521937142, + "grad_norm": 1.9840598269102252, + "learning_rate": 2.7023980196846917e-06, + "loss": 0.9815, + "step": 4482 + }, + { + "epoch": 0.4042927357171845, + "grad_norm": 0.7849352321511502, + "learning_rate": 2.7018509928265763e-06, + "loss": 0.7995, + "step": 4483 + }, + { + "epoch": 0.4043829192406547, + "grad_norm": 1.4707221730329814, + "learning_rate": 2.7013039060826635e-06, + "loss": 0.9884, + "step": 4484 + }, + { + "epoch": 0.404473102764125, + "grad_norm": 1.7820588667894108, + "learning_rate": 2.7007567594996347e-06, + "loss": 1.0324, + "step": 4485 + }, + { + "epoch": 0.40456328628759525, + "grad_norm": 1.4767863593377677, + "learning_rate": 2.7002095531241757e-06, + "loss": 0.863, + "step": 4486 + }, + { + "epoch": 0.40465346981106554, + "grad_norm": 1.7363176206050013, + "learning_rate": 2.6996622870029767e-06, + "loss": 0.9363, + "step": 4487 + }, + { + "epoch": 0.40474365333453577, + "grad_norm": 1.717934848646932, + "learning_rate": 2.6991149611827335e-06, + "loss": 1.0738, + "step": 4488 + }, + { + "epoch": 0.40483383685800606, + "grad_norm": 1.58048962491138, + "learning_rate": 2.6985675757101466e-06, + "loss": 1.0012, + "step": 4489 + }, + { + "epoch": 0.4049240203814763, + "grad_norm": 1.8251484823571908, + "learning_rate": 2.698020130631922e-06, + "loss": 0.9231, + "step": 4490 + }, + { + "epoch": 0.4050142039049466, + "grad_norm": 1.5090626490172825, + "learning_rate": 2.6974726259947713e-06, + "loss": 0.8913, + "step": 4491 + }, + { + "epoch": 0.4051043874284168, + "grad_norm": 2.1264729724119893, + "learning_rate": 2.6969250618454106e-06, + "loss": 0.941, + "step": 4492 + }, + { + "epoch": 0.4051945709518871, + "grad_norm": 1.3789297215352523, + "learning_rate": 2.696377438230561e-06, + "loss": 1.0341, + "step": 4493 + }, + { + "epoch": 0.40528475447535733, + "grad_norm": 1.7205113489186326, + "learning_rate": 2.6958297551969484e-06, + "loss": 0.8973, + "step": 4494 + }, + { + "epoch": 0.4053749379988276, + "grad_norm": 1.5614069357826317, + "learning_rate": 2.695282012791304e-06, + "loss": 0.9298, + "step": 4495 + }, + { + "epoch": 0.40546512152229786, + "grad_norm": 1.5391505647046648, + "learning_rate": 2.6947342110603646e-06, + "loss": 1.0177, + "step": 4496 + }, + { + "epoch": 0.40555530504576814, + "grad_norm": 2.1662163877878644, + "learning_rate": 2.6941863500508717e-06, + "loss": 1.0157, + "step": 4497 + }, + { + "epoch": 0.4056454885692384, + "grad_norm": 1.6545728433020785, + "learning_rate": 2.693638429809572e-06, + "loss": 1.0155, + "step": 4498 + }, + { + "epoch": 0.40573567209270867, + "grad_norm": 1.7174979764879237, + "learning_rate": 2.6930904503832167e-06, + "loss": 0.9447, + "step": 4499 + }, + { + "epoch": 0.4058258556161789, + "grad_norm": 1.5143034374222986, + "learning_rate": 2.692542411818562e-06, + "loss": 0.9452, + "step": 4500 + }, + { + "epoch": 0.4059160391396492, + "grad_norm": 1.7246445579871388, + "learning_rate": 2.69199431416237e-06, + "loss": 0.9767, + "step": 4501 + }, + { + "epoch": 0.4060062226631194, + "grad_norm": 1.491550329491598, + "learning_rate": 2.691446157461408e-06, + "loss": 0.9188, + "step": 4502 + }, + { + "epoch": 0.4060964061865897, + "grad_norm": 1.3804811124134755, + "learning_rate": 2.690897941762447e-06, + "loss": 1.0237, + "step": 4503 + }, + { + "epoch": 0.40618658971006, + "grad_norm": 1.424913617912315, + "learning_rate": 2.6903496671122642e-06, + "loss": 0.9402, + "step": 4504 + }, + { + "epoch": 0.40627677323353023, + "grad_norm": 0.7072150060114556, + "learning_rate": 2.689801333557641e-06, + "loss": 0.7303, + "step": 4505 + }, + { + "epoch": 0.4063669567570005, + "grad_norm": 2.013310722872527, + "learning_rate": 2.689252941145365e-06, + "loss": 0.9624, + "step": 4506 + }, + { + "epoch": 0.40645714028047075, + "grad_norm": 1.4467699280522064, + "learning_rate": 2.6887044899222277e-06, + "loss": 0.9579, + "step": 4507 + }, + { + "epoch": 0.40654732380394104, + "grad_norm": 1.606741519044366, + "learning_rate": 2.688155979935025e-06, + "loss": 1.0061, + "step": 4508 + }, + { + "epoch": 0.4066375073274113, + "grad_norm": 1.4563287672752863, + "learning_rate": 2.68760741123056e-06, + "loss": 0.9235, + "step": 4509 + }, + { + "epoch": 0.40672769085088156, + "grad_norm": 3.7950472730021807, + "learning_rate": 2.6870587838556394e-06, + "loss": 0.9829, + "step": 4510 + }, + { + "epoch": 0.4068178743743518, + "grad_norm": 1.7618479882179525, + "learning_rate": 2.686510097857075e-06, + "loss": 0.9801, + "step": 4511 + }, + { + "epoch": 0.4069080578978221, + "grad_norm": 1.3070710845914888, + "learning_rate": 2.685961353281683e-06, + "loss": 0.9547, + "step": 4512 + }, + { + "epoch": 0.4069982414212923, + "grad_norm": 1.6256403028331812, + "learning_rate": 2.6854125501762863e-06, + "loss": 0.9468, + "step": 4513 + }, + { + "epoch": 0.4070884249447626, + "grad_norm": 1.5732189250152062, + "learning_rate": 2.684863688587712e-06, + "loss": 0.9665, + "step": 4514 + }, + { + "epoch": 0.40717860846823284, + "grad_norm": 2.1925235286829023, + "learning_rate": 2.6843147685627916e-06, + "loss": 0.9352, + "step": 4515 + }, + { + "epoch": 0.4072687919917031, + "grad_norm": 1.630038519679157, + "learning_rate": 2.683765790148361e-06, + "loss": 0.946, + "step": 4516 + }, + { + "epoch": 0.40735897551517336, + "grad_norm": 1.8986407512604384, + "learning_rate": 2.6832167533912637e-06, + "loss": 1.0149, + "step": 4517 + }, + { + "epoch": 0.40744915903864365, + "grad_norm": 1.9417656732729422, + "learning_rate": 2.682667658338345e-06, + "loss": 1.0391, + "step": 4518 + }, + { + "epoch": 0.4075393425621139, + "grad_norm": 1.5246275337500308, + "learning_rate": 2.682118505036458e-06, + "loss": 0.936, + "step": 4519 + }, + { + "epoch": 0.40762952608558417, + "grad_norm": 0.6848121901422057, + "learning_rate": 2.681569293532459e-06, + "loss": 0.7506, + "step": 4520 + }, + { + "epoch": 0.4077197096090544, + "grad_norm": 0.6814410073124386, + "learning_rate": 2.6810200238732102e-06, + "loss": 0.778, + "step": 4521 + }, + { + "epoch": 0.4078098931325247, + "grad_norm": 1.443195008454498, + "learning_rate": 2.6804706961055776e-06, + "loss": 1.009, + "step": 4522 + }, + { + "epoch": 0.4079000766559949, + "grad_norm": 1.4245891447228591, + "learning_rate": 2.6799213102764326e-06, + "loss": 0.9964, + "step": 4523 + }, + { + "epoch": 0.4079902601794652, + "grad_norm": 2.644273032537576, + "learning_rate": 2.679371866432653e-06, + "loss": 1.0611, + "step": 4524 + }, + { + "epoch": 0.40808044370293545, + "grad_norm": 1.5631755894188928, + "learning_rate": 2.6788223646211194e-06, + "loss": 0.9363, + "step": 4525 + }, + { + "epoch": 0.40817062722640574, + "grad_norm": 1.4850346755786203, + "learning_rate": 2.6782728048887183e-06, + "loss": 0.8548, + "step": 4526 + }, + { + "epoch": 0.408260810749876, + "grad_norm": 1.26007913636148, + "learning_rate": 2.6777231872823416e-06, + "loss": 0.9484, + "step": 4527 + }, + { + "epoch": 0.40835099427334626, + "grad_norm": 1.875023587396395, + "learning_rate": 2.6771735118488864e-06, + "loss": 0.959, + "step": 4528 + }, + { + "epoch": 0.40844117779681655, + "grad_norm": 1.5565310335530405, + "learning_rate": 2.6766237786352523e-06, + "loss": 1.0017, + "step": 4529 + }, + { + "epoch": 0.4085313613202868, + "grad_norm": 1.3889670185790073, + "learning_rate": 2.676073987688347e-06, + "loss": 0.8907, + "step": 4530 + }, + { + "epoch": 0.40862154484375707, + "grad_norm": 1.4470893453921092, + "learning_rate": 2.6755241390550818e-06, + "loss": 0.9674, + "step": 4531 + }, + { + "epoch": 0.4087117283672273, + "grad_norm": 1.530346896046174, + "learning_rate": 2.6749742327823716e-06, + "loss": 0.9636, + "step": 4532 + }, + { + "epoch": 0.4088019118906976, + "grad_norm": 2.3813204737369804, + "learning_rate": 2.674424268917138e-06, + "loss": 0.9781, + "step": 4533 + }, + { + "epoch": 0.4088920954141678, + "grad_norm": 1.525873281239632, + "learning_rate": 2.6738742475063074e-06, + "loss": 1.0255, + "step": 4534 + }, + { + "epoch": 0.4089822789376381, + "grad_norm": 1.689327380842537, + "learning_rate": 2.6733241685968104e-06, + "loss": 1.0007, + "step": 4535 + }, + { + "epoch": 0.40907246246110834, + "grad_norm": 1.7890233864319152, + "learning_rate": 2.6727740322355826e-06, + "loss": 0.9794, + "step": 4536 + }, + { + "epoch": 0.40916264598457863, + "grad_norm": 1.7457928494271484, + "learning_rate": 2.6722238384695644e-06, + "loss": 0.9473, + "step": 4537 + }, + { + "epoch": 0.40925282950804887, + "grad_norm": 1.454248465757723, + "learning_rate": 2.671673587345702e-06, + "loss": 0.9727, + "step": 4538 + }, + { + "epoch": 0.40934301303151915, + "grad_norm": 0.7561955456004982, + "learning_rate": 2.6711232789109455e-06, + "loss": 0.7693, + "step": 4539 + }, + { + "epoch": 0.4094331965549894, + "grad_norm": 1.447868849515922, + "learning_rate": 2.6705729132122497e-06, + "loss": 0.9639, + "step": 4540 + }, + { + "epoch": 0.4095233800784597, + "grad_norm": 1.8568116157265382, + "learning_rate": 2.670022490296576e-06, + "loss": 1.0422, + "step": 4541 + }, + { + "epoch": 0.4096135636019299, + "grad_norm": 1.3790489317474814, + "learning_rate": 2.669472010210889e-06, + "loss": 1.0161, + "step": 4542 + }, + { + "epoch": 0.4097037471254002, + "grad_norm": 1.6767627891506665, + "learning_rate": 2.668921473002159e-06, + "loss": 0.9374, + "step": 4543 + }, + { + "epoch": 0.40979393064887043, + "grad_norm": 1.5846989497976796, + "learning_rate": 2.6683708787173596e-06, + "loss": 1.0172, + "step": 4544 + }, + { + "epoch": 0.4098841141723407, + "grad_norm": 1.4481723020003967, + "learning_rate": 2.6678202274034718e-06, + "loss": 0.9827, + "step": 4545 + }, + { + "epoch": 0.40997429769581095, + "grad_norm": 1.6486910222793658, + "learning_rate": 2.66726951910748e-06, + "loss": 1.0545, + "step": 4546 + }, + { + "epoch": 0.41006448121928124, + "grad_norm": 1.5183365153011699, + "learning_rate": 2.6667187538763737e-06, + "loss": 0.9693, + "step": 4547 + }, + { + "epoch": 0.4101546647427515, + "grad_norm": 1.484990684916761, + "learning_rate": 2.6661679317571473e-06, + "loss": 0.9452, + "step": 4548 + }, + { + "epoch": 0.41024484826622176, + "grad_norm": 1.4134535769303738, + "learning_rate": 2.665617052796799e-06, + "loss": 0.9459, + "step": 4549 + }, + { + "epoch": 0.41033503178969205, + "grad_norm": 1.6636048883197982, + "learning_rate": 2.6650661170423346e-06, + "loss": 0.8484, + "step": 4550 + }, + { + "epoch": 0.4104252153131623, + "grad_norm": 1.7024161587319349, + "learning_rate": 2.6645151245407614e-06, + "loss": 1.0467, + "step": 4551 + }, + { + "epoch": 0.4105153988366326, + "grad_norm": 1.4433031290577951, + "learning_rate": 2.6639640753390936e-06, + "loss": 0.9596, + "step": 4552 + }, + { + "epoch": 0.4106055823601028, + "grad_norm": 1.4090475865105943, + "learning_rate": 2.66341296948435e-06, + "loss": 0.8693, + "step": 4553 + }, + { + "epoch": 0.4106957658835731, + "grad_norm": 1.3130758021569369, + "learning_rate": 2.6628618070235534e-06, + "loss": 0.9749, + "step": 4554 + }, + { + "epoch": 0.4107859494070433, + "grad_norm": 1.5970567543823042, + "learning_rate": 2.662310588003733e-06, + "loss": 0.9667, + "step": 4555 + }, + { + "epoch": 0.4108761329305136, + "grad_norm": 1.7202467903248075, + "learning_rate": 2.6617593124719205e-06, + "loss": 0.9345, + "step": 4556 + }, + { + "epoch": 0.41096631645398385, + "grad_norm": 1.302394382199677, + "learning_rate": 2.661207980475155e-06, + "loss": 0.97, + "step": 4557 + }, + { + "epoch": 0.41105649997745414, + "grad_norm": 1.5297401925695406, + "learning_rate": 2.6606565920604793e-06, + "loss": 0.9167, + "step": 4558 + }, + { + "epoch": 0.41114668350092437, + "grad_norm": 1.5801439118335052, + "learning_rate": 2.66010514727494e-06, + "loss": 1.0429, + "step": 4559 + }, + { + "epoch": 0.41123686702439466, + "grad_norm": 1.882159487867209, + "learning_rate": 2.659553646165589e-06, + "loss": 0.8599, + "step": 4560 + }, + { + "epoch": 0.4113270505478649, + "grad_norm": 1.5137770093277003, + "learning_rate": 2.659002088779485e-06, + "loss": 0.9338, + "step": 4561 + }, + { + "epoch": 0.4114172340713352, + "grad_norm": 0.7653418036661161, + "learning_rate": 2.6584504751636888e-06, + "loss": 0.7291, + "step": 4562 + }, + { + "epoch": 0.4115074175948054, + "grad_norm": 1.474019277366347, + "learning_rate": 2.657898805365268e-06, + "loss": 0.9252, + "step": 4563 + }, + { + "epoch": 0.4115976011182757, + "grad_norm": 1.674282515286552, + "learning_rate": 2.657347079431293e-06, + "loss": 0.929, + "step": 4564 + }, + { + "epoch": 0.41168778464174594, + "grad_norm": 1.2584868808982497, + "learning_rate": 2.6567952974088403e-06, + "loss": 0.9361, + "step": 4565 + }, + { + "epoch": 0.4117779681652162, + "grad_norm": 1.3708330216257656, + "learning_rate": 2.6562434593449917e-06, + "loss": 1.0125, + "step": 4566 + }, + { + "epoch": 0.41186815168868646, + "grad_norm": 1.6600020034042586, + "learning_rate": 2.6556915652868325e-06, + "loss": 0.9862, + "step": 4567 + }, + { + "epoch": 0.41195833521215675, + "grad_norm": 1.5049539458873629, + "learning_rate": 2.6551396152814534e-06, + "loss": 0.9677, + "step": 4568 + }, + { + "epoch": 0.412048518735627, + "grad_norm": 1.6855296181583297, + "learning_rate": 2.65458760937595e-06, + "loss": 0.9589, + "step": 4569 + }, + { + "epoch": 0.41213870225909727, + "grad_norm": 1.3550473308390696, + "learning_rate": 2.654035547617423e-06, + "loss": 0.9141, + "step": 4570 + }, + { + "epoch": 0.4122288857825675, + "grad_norm": 1.95653772888929, + "learning_rate": 2.653483430052976e-06, + "loss": 0.863, + "step": 4571 + }, + { + "epoch": 0.4123190693060378, + "grad_norm": 1.6865780042381326, + "learning_rate": 2.6529312567297197e-06, + "loss": 0.94, + "step": 4572 + }, + { + "epoch": 0.412409252829508, + "grad_norm": 1.672131028580761, + "learning_rate": 2.652379027694768e-06, + "loss": 1.0398, + "step": 4573 + }, + { + "epoch": 0.4124994363529783, + "grad_norm": 2.507628437073158, + "learning_rate": 2.651826742995241e-06, + "loss": 1.0214, + "step": 4574 + }, + { + "epoch": 0.4125896198764486, + "grad_norm": 1.3754874579143126, + "learning_rate": 2.651274402678262e-06, + "loss": 0.9585, + "step": 4575 + }, + { + "epoch": 0.41267980339991883, + "grad_norm": 1.424146661069181, + "learning_rate": 2.6507220067909597e-06, + "loss": 0.9903, + "step": 4576 + }, + { + "epoch": 0.4127699869233891, + "grad_norm": 1.8235352111630085, + "learning_rate": 2.650169555380468e-06, + "loss": 0.9196, + "step": 4577 + }, + { + "epoch": 0.41286017044685935, + "grad_norm": 1.3442598750620973, + "learning_rate": 2.6496170484939253e-06, + "loss": 0.9558, + "step": 4578 + }, + { + "epoch": 0.41295035397032964, + "grad_norm": 1.7615308978629858, + "learning_rate": 2.6490644861784735e-06, + "loss": 0.9759, + "step": 4579 + }, + { + "epoch": 0.4130405374937999, + "grad_norm": 1.4440372924321638, + "learning_rate": 2.648511868481261e-06, + "loss": 0.9835, + "step": 4580 + }, + { + "epoch": 0.41313072101727016, + "grad_norm": 1.5272848991119568, + "learning_rate": 2.6479591954494397e-06, + "loss": 0.9632, + "step": 4581 + }, + { + "epoch": 0.4132209045407404, + "grad_norm": 1.4689201905686966, + "learning_rate": 2.647406467130167e-06, + "loss": 0.8607, + "step": 4582 + }, + { + "epoch": 0.4133110880642107, + "grad_norm": 1.5074978040486247, + "learning_rate": 2.646853683570605e-06, + "loss": 0.9819, + "step": 4583 + }, + { + "epoch": 0.4134012715876809, + "grad_norm": 1.4699855234120973, + "learning_rate": 2.6463008448179196e-06, + "loss": 0.9012, + "step": 4584 + }, + { + "epoch": 0.4134914551111512, + "grad_norm": 1.4567284972615475, + "learning_rate": 2.6457479509192828e-06, + "loss": 0.9123, + "step": 4585 + }, + { + "epoch": 0.41358163863462144, + "grad_norm": 1.9622684536543804, + "learning_rate": 2.645195001921871e-06, + "loss": 0.954, + "step": 4586 + }, + { + "epoch": 0.41367182215809173, + "grad_norm": 1.5203957733840223, + "learning_rate": 2.644641997872863e-06, + "loss": 0.94, + "step": 4587 + }, + { + "epoch": 0.41376200568156196, + "grad_norm": 1.8821524575087627, + "learning_rate": 2.644088938819445e-06, + "loss": 0.9598, + "step": 4588 + }, + { + "epoch": 0.41385218920503225, + "grad_norm": 1.5975832001417398, + "learning_rate": 2.6435358248088077e-06, + "loss": 1.045, + "step": 4589 + }, + { + "epoch": 0.4139423727285025, + "grad_norm": 1.4779479777687863, + "learning_rate": 2.642982655888146e-06, + "loss": 1.0029, + "step": 4590 + }, + { + "epoch": 0.41403255625197277, + "grad_norm": 1.4610306817187597, + "learning_rate": 2.6424294321046585e-06, + "loss": 0.9216, + "step": 4591 + }, + { + "epoch": 0.414122739775443, + "grad_norm": 0.784744686229029, + "learning_rate": 2.641876153505549e-06, + "loss": 0.7753, + "step": 4592 + }, + { + "epoch": 0.4142129232989133, + "grad_norm": 1.9095140871625602, + "learning_rate": 2.641322820138027e-06, + "loss": 0.9752, + "step": 4593 + }, + { + "epoch": 0.4143031068223835, + "grad_norm": 1.732329872552643, + "learning_rate": 2.640769432049306e-06, + "loss": 0.9164, + "step": 4594 + }, + { + "epoch": 0.4143932903458538, + "grad_norm": 1.6713266632140733, + "learning_rate": 2.6402159892866038e-06, + "loss": 0.8873, + "step": 4595 + }, + { + "epoch": 0.41448347386932405, + "grad_norm": 1.2892957996446446, + "learning_rate": 2.639662491897143e-06, + "loss": 0.9485, + "step": 4596 + }, + { + "epoch": 0.41457365739279434, + "grad_norm": 1.494676361776233, + "learning_rate": 2.639108939928152e-06, + "loss": 0.9706, + "step": 4597 + }, + { + "epoch": 0.4146638409162646, + "grad_norm": 1.6959486088793967, + "learning_rate": 2.638555333426862e-06, + "loss": 0.9761, + "step": 4598 + }, + { + "epoch": 0.41475402443973486, + "grad_norm": 1.3863954408951478, + "learning_rate": 2.6380016724405093e-06, + "loss": 0.9905, + "step": 4599 + }, + { + "epoch": 0.41484420796320515, + "grad_norm": 1.5577219578057568, + "learning_rate": 2.637447957016336e-06, + "loss": 0.9192, + "step": 4600 + }, + { + "epoch": 0.4149343914866754, + "grad_norm": 1.6051182422280308, + "learning_rate": 2.636894187201589e-06, + "loss": 1.0196, + "step": 4601 + }, + { + "epoch": 0.41502457501014567, + "grad_norm": 1.5181936147550268, + "learning_rate": 2.6363403630435176e-06, + "loss": 0.8631, + "step": 4602 + }, + { + "epoch": 0.4151147585336159, + "grad_norm": 1.6041813721746532, + "learning_rate": 2.635786484589378e-06, + "loss": 0.9552, + "step": 4603 + }, + { + "epoch": 0.4152049420570862, + "grad_norm": 1.7303920065781173, + "learning_rate": 2.63523255188643e-06, + "loss": 0.8566, + "step": 4604 + }, + { + "epoch": 0.4152951255805564, + "grad_norm": 1.3957952332871795, + "learning_rate": 2.6346785649819375e-06, + "loss": 0.9847, + "step": 4605 + }, + { + "epoch": 0.4153853091040267, + "grad_norm": 1.5585480112042618, + "learning_rate": 2.6341245239231706e-06, + "loss": 0.9846, + "step": 4606 + }, + { + "epoch": 0.41547549262749695, + "grad_norm": 1.2275652825734915, + "learning_rate": 2.6335704287574024e-06, + "loss": 0.9608, + "step": 4607 + }, + { + "epoch": 0.41556567615096723, + "grad_norm": 1.6794268028750792, + "learning_rate": 2.6330162795319124e-06, + "loss": 1.0331, + "step": 4608 + }, + { + "epoch": 0.41565585967443747, + "grad_norm": 1.6837452689757855, + "learning_rate": 2.632462076293983e-06, + "loss": 1.0185, + "step": 4609 + }, + { + "epoch": 0.41574604319790776, + "grad_norm": 1.2525162642863927, + "learning_rate": 2.6319078190909017e-06, + "loss": 0.9578, + "step": 4610 + }, + { + "epoch": 0.415836226721378, + "grad_norm": 2.654843575131452, + "learning_rate": 2.6313535079699606e-06, + "loss": 0.8815, + "step": 4611 + }, + { + "epoch": 0.4159264102448483, + "grad_norm": 1.539169462718121, + "learning_rate": 2.6307991429784572e-06, + "loss": 0.9371, + "step": 4612 + }, + { + "epoch": 0.4160165937683185, + "grad_norm": 1.6470484372620922, + "learning_rate": 2.6302447241636924e-06, + "loss": 1.0231, + "step": 4613 + }, + { + "epoch": 0.4161067772917888, + "grad_norm": 1.6039292874100906, + "learning_rate": 2.629690251572973e-06, + "loss": 1.0395, + "step": 4614 + }, + { + "epoch": 0.41619696081525903, + "grad_norm": 1.5822386805804076, + "learning_rate": 2.629135725253609e-06, + "loss": 0.9362, + "step": 4615 + }, + { + "epoch": 0.4162871443387293, + "grad_norm": 1.5505110267411903, + "learning_rate": 2.6285811452529162e-06, + "loss": 0.9452, + "step": 4616 + }, + { + "epoch": 0.41637732786219955, + "grad_norm": 1.4644096850375286, + "learning_rate": 2.6280265116182136e-06, + "loss": 0.9802, + "step": 4617 + }, + { + "epoch": 0.41646751138566984, + "grad_norm": 1.514467564303183, + "learning_rate": 2.6274718243968266e-06, + "loss": 0.9674, + "step": 4618 + }, + { + "epoch": 0.4165576949091401, + "grad_norm": 1.4871853375404125, + "learning_rate": 2.626917083636084e-06, + "loss": 0.881, + "step": 4619 + }, + { + "epoch": 0.41664787843261036, + "grad_norm": 1.2946476940947877, + "learning_rate": 2.6263622893833183e-06, + "loss": 0.8944, + "step": 4620 + }, + { + "epoch": 0.4167380619560806, + "grad_norm": 1.6571008188336933, + "learning_rate": 2.625807441685869e-06, + "loss": 0.907, + "step": 4621 + }, + { + "epoch": 0.4168282454795509, + "grad_norm": 0.6784571989793748, + "learning_rate": 2.625252540591078e-06, + "loss": 0.7626, + "step": 4622 + }, + { + "epoch": 0.4169184290030212, + "grad_norm": 1.6353892022355707, + "learning_rate": 2.6246975861462927e-06, + "loss": 1.0357, + "step": 4623 + }, + { + "epoch": 0.4170086125264914, + "grad_norm": 1.9502631621193525, + "learning_rate": 2.624142578398864e-06, + "loss": 1.0072, + "step": 4624 + }, + { + "epoch": 0.4170987960499617, + "grad_norm": 1.3937392349853808, + "learning_rate": 2.6235875173961498e-06, + "loss": 0.9423, + "step": 4625 + }, + { + "epoch": 0.41718897957343193, + "grad_norm": 1.5858907457798121, + "learning_rate": 2.62303240318551e-06, + "loss": 0.8907, + "step": 4626 + }, + { + "epoch": 0.4172791630969022, + "grad_norm": 1.8085563112013219, + "learning_rate": 2.62247723581431e-06, + "loss": 1.0379, + "step": 4627 + }, + { + "epoch": 0.41736934662037245, + "grad_norm": 1.6419823254320522, + "learning_rate": 2.62192201532992e-06, + "loss": 0.9626, + "step": 4628 + }, + { + "epoch": 0.41745953014384274, + "grad_norm": 1.5512554683173945, + "learning_rate": 2.6213667417797145e-06, + "loss": 0.9786, + "step": 4629 + }, + { + "epoch": 0.41754971366731297, + "grad_norm": 1.3564153878060505, + "learning_rate": 2.6208114152110725e-06, + "loss": 0.9845, + "step": 4630 + }, + { + "epoch": 0.41763989719078326, + "grad_norm": 1.7143394058765784, + "learning_rate": 2.6202560356713774e-06, + "loss": 0.9763, + "step": 4631 + }, + { + "epoch": 0.4177300807142535, + "grad_norm": 0.723883375682933, + "learning_rate": 2.619700603208017e-06, + "loss": 0.8402, + "step": 4632 + }, + { + "epoch": 0.4178202642377238, + "grad_norm": 1.506861569674246, + "learning_rate": 2.6191451178683842e-06, + "loss": 0.876, + "step": 4633 + }, + { + "epoch": 0.417910447761194, + "grad_norm": 1.4807328363570573, + "learning_rate": 2.6185895796998764e-06, + "loss": 0.993, + "step": 4634 + }, + { + "epoch": 0.4180006312846643, + "grad_norm": 1.7813940157310024, + "learning_rate": 2.6180339887498946e-06, + "loss": 1.0343, + "step": 4635 + }, + { + "epoch": 0.41809081480813454, + "grad_norm": 1.7858890556913891, + "learning_rate": 2.617478345065846e-06, + "loss": 0.9004, + "step": 4636 + }, + { + "epoch": 0.4181809983316048, + "grad_norm": 1.576775530967545, + "learning_rate": 2.616922648695139e-06, + "loss": 0.9798, + "step": 4637 + }, + { + "epoch": 0.41827118185507506, + "grad_norm": 1.661254304258857, + "learning_rate": 2.61636689968519e-06, + "loss": 0.8577, + "step": 4638 + }, + { + "epoch": 0.41836136537854535, + "grad_norm": 1.7805764782106608, + "learning_rate": 2.6158110980834186e-06, + "loss": 1.0244, + "step": 4639 + }, + { + "epoch": 0.4184515489020156, + "grad_norm": 1.6894222190543782, + "learning_rate": 2.615255243937249e-06, + "loss": 1.0116, + "step": 4640 + }, + { + "epoch": 0.41854173242548587, + "grad_norm": 1.7686251242401212, + "learning_rate": 2.61469933729411e-06, + "loss": 0.9856, + "step": 4641 + }, + { + "epoch": 0.4186319159489561, + "grad_norm": 1.7595327818131916, + "learning_rate": 2.614143378201433e-06, + "loss": 0.9448, + "step": 4642 + }, + { + "epoch": 0.4187220994724264, + "grad_norm": 1.5582456271671585, + "learning_rate": 2.6135873667066567e-06, + "loss": 0.9428, + "step": 4643 + }, + { + "epoch": 0.4188122829958966, + "grad_norm": 1.4106411932208904, + "learning_rate": 2.613031302857224e-06, + "loss": 0.9191, + "step": 4644 + }, + { + "epoch": 0.4189024665193669, + "grad_norm": 1.6920650335714296, + "learning_rate": 2.6124751867005792e-06, + "loss": 0.9877, + "step": 4645 + }, + { + "epoch": 0.4189926500428372, + "grad_norm": 1.2345132871015498, + "learning_rate": 2.611919018284175e-06, + "loss": 1.0113, + "step": 4646 + }, + { + "epoch": 0.41908283356630743, + "grad_norm": 1.5988721627574158, + "learning_rate": 2.611362797655466e-06, + "loss": 1.0114, + "step": 4647 + }, + { + "epoch": 0.4191730170897777, + "grad_norm": 1.4006580219861442, + "learning_rate": 2.6108065248619124e-06, + "loss": 1.0187, + "step": 4648 + }, + { + "epoch": 0.41926320061324795, + "grad_norm": 1.2729070295769247, + "learning_rate": 2.610250199950978e-06, + "loss": 0.896, + "step": 4649 + }, + { + "epoch": 0.41935338413671824, + "grad_norm": 1.7205355299562746, + "learning_rate": 2.609693822970131e-06, + "loss": 0.9329, + "step": 4650 + }, + { + "epoch": 0.4194435676601885, + "grad_norm": 1.5718963378914068, + "learning_rate": 2.609137393966846e-06, + "loss": 0.9288, + "step": 4651 + }, + { + "epoch": 0.41953375118365877, + "grad_norm": 1.6800943281167993, + "learning_rate": 2.6085809129886e-06, + "loss": 0.9132, + "step": 4652 + }, + { + "epoch": 0.419623934707129, + "grad_norm": 1.9437551737529428, + "learning_rate": 2.608024380082874e-06, + "loss": 0.9682, + "step": 4653 + }, + { + "epoch": 0.4197141182305993, + "grad_norm": 1.607302173407927, + "learning_rate": 2.6074677952971554e-06, + "loss": 0.9573, + "step": 4654 + }, + { + "epoch": 0.4198043017540695, + "grad_norm": 1.4546338523207918, + "learning_rate": 2.606911158678935e-06, + "loss": 0.8815, + "step": 4655 + }, + { + "epoch": 0.4198944852775398, + "grad_norm": 2.794057865802423, + "learning_rate": 2.606354470275708e-06, + "loss": 0.9567, + "step": 4656 + }, + { + "epoch": 0.41998466880101004, + "grad_norm": 1.5403464910928548, + "learning_rate": 2.6057977301349744e-06, + "loss": 0.9337, + "step": 4657 + }, + { + "epoch": 0.42007485232448033, + "grad_norm": 1.5545101352139161, + "learning_rate": 2.6052409383042383e-06, + "loss": 0.9214, + "step": 4658 + }, + { + "epoch": 0.42016503584795056, + "grad_norm": 1.7882574677292087, + "learning_rate": 2.6046840948310074e-06, + "loss": 0.9371, + "step": 4659 + }, + { + "epoch": 0.42025521937142085, + "grad_norm": 1.463823700832146, + "learning_rate": 2.6041271997627962e-06, + "loss": 0.9618, + "step": 4660 + }, + { + "epoch": 0.4203454028948911, + "grad_norm": 1.775895352270391, + "learning_rate": 2.6035702531471202e-06, + "loss": 0.9483, + "step": 4661 + }, + { + "epoch": 0.4204355864183614, + "grad_norm": 1.408216711817365, + "learning_rate": 2.6030132550315035e-06, + "loss": 0.9573, + "step": 4662 + }, + { + "epoch": 0.4205257699418316, + "grad_norm": 0.7399532619743382, + "learning_rate": 2.60245620546347e-06, + "loss": 0.8148, + "step": 4663 + }, + { + "epoch": 0.4206159534653019, + "grad_norm": 1.6800333066318096, + "learning_rate": 2.6018991044905517e-06, + "loss": 0.9382, + "step": 4664 + }, + { + "epoch": 0.42070613698877213, + "grad_norm": 1.4689996385463533, + "learning_rate": 2.6013419521602825e-06, + "loss": 0.9973, + "step": 4665 + }, + { + "epoch": 0.4207963205122424, + "grad_norm": 3.2821080130517304, + "learning_rate": 2.600784748520202e-06, + "loss": 1.0881, + "step": 4666 + }, + { + "epoch": 0.42088650403571265, + "grad_norm": 1.40809023506469, + "learning_rate": 2.6002274936178544e-06, + "loss": 0.8975, + "step": 4667 + }, + { + "epoch": 0.42097668755918294, + "grad_norm": 1.599144575805644, + "learning_rate": 2.5996701875007873e-06, + "loss": 0.9724, + "step": 4668 + }, + { + "epoch": 0.4210668710826532, + "grad_norm": 1.5469956013839516, + "learning_rate": 2.5991128302165533e-06, + "loss": 0.9545, + "step": 4669 + }, + { + "epoch": 0.42115705460612346, + "grad_norm": 1.3652923091444098, + "learning_rate": 2.5985554218127094e-06, + "loss": 1.0256, + "step": 4670 + }, + { + "epoch": 0.42124723812959375, + "grad_norm": 1.8725316648772639, + "learning_rate": 2.597997962336816e-06, + "loss": 0.8948, + "step": 4671 + }, + { + "epoch": 0.421337421653064, + "grad_norm": 1.68154917611212, + "learning_rate": 2.5974404518364393e-06, + "loss": 0.9547, + "step": 4672 + }, + { + "epoch": 0.42142760517653427, + "grad_norm": 1.675421260412231, + "learning_rate": 2.596882890359149e-06, + "loss": 1.0883, + "step": 4673 + }, + { + "epoch": 0.4215177887000045, + "grad_norm": 3.482279331201783, + "learning_rate": 2.5963252779525196e-06, + "loss": 0.8772, + "step": 4674 + }, + { + "epoch": 0.4216079722234748, + "grad_norm": 1.839522023833589, + "learning_rate": 2.595767614664129e-06, + "loss": 0.9627, + "step": 4675 + }, + { + "epoch": 0.421698155746945, + "grad_norm": 1.5563520405725486, + "learning_rate": 2.5952099005415607e-06, + "loss": 0.9407, + "step": 4676 + }, + { + "epoch": 0.4217883392704153, + "grad_norm": 2.020718076375191, + "learning_rate": 2.594652135632402e-06, + "loss": 0.9263, + "step": 4677 + }, + { + "epoch": 0.42187852279388555, + "grad_norm": 1.356913343719785, + "learning_rate": 2.594094319984244e-06, + "loss": 0.9177, + "step": 4678 + }, + { + "epoch": 0.42196870631735584, + "grad_norm": 1.4853141624462247, + "learning_rate": 2.5935364536446825e-06, + "loss": 1.0187, + "step": 4679 + }, + { + "epoch": 0.42205888984082607, + "grad_norm": 1.7179114637065553, + "learning_rate": 2.5929785366613185e-06, + "loss": 0.9658, + "step": 4680 + }, + { + "epoch": 0.42214907336429636, + "grad_norm": 1.6944309175017407, + "learning_rate": 2.592420569081756e-06, + "loss": 0.9983, + "step": 4681 + }, + { + "epoch": 0.4222392568877666, + "grad_norm": 1.7802260952847089, + "learning_rate": 2.5918625509536037e-06, + "loss": 0.9194, + "step": 4682 + }, + { + "epoch": 0.4223294404112369, + "grad_norm": 0.7046216083855757, + "learning_rate": 2.591304482324475e-06, + "loss": 0.7792, + "step": 4683 + }, + { + "epoch": 0.4224196239347071, + "grad_norm": 1.4541239432514121, + "learning_rate": 2.5907463632419878e-06, + "loss": 0.9749, + "step": 4684 + }, + { + "epoch": 0.4225098074581774, + "grad_norm": 1.450824105446859, + "learning_rate": 2.5901881937537632e-06, + "loss": 0.9244, + "step": 4685 + }, + { + "epoch": 0.42259999098164763, + "grad_norm": 0.7853663149275594, + "learning_rate": 2.589629973907428e-06, + "loss": 0.8131, + "step": 4686 + }, + { + "epoch": 0.4226901745051179, + "grad_norm": 1.6821152286499443, + "learning_rate": 2.589071703750612e-06, + "loss": 0.9138, + "step": 4687 + }, + { + "epoch": 0.42278035802858815, + "grad_norm": 1.339129332540485, + "learning_rate": 2.5885133833309504e-06, + "loss": 0.9986, + "step": 4688 + }, + { + "epoch": 0.42287054155205844, + "grad_norm": 1.4489846258934012, + "learning_rate": 2.5879550126960814e-06, + "loss": 1.0419, + "step": 4689 + }, + { + "epoch": 0.4229607250755287, + "grad_norm": 1.4424519867732577, + "learning_rate": 2.5873965918936494e-06, + "loss": 1.0153, + "step": 4690 + }, + { + "epoch": 0.42305090859899896, + "grad_norm": 2.4111895703247743, + "learning_rate": 2.586838120971301e-06, + "loss": 0.9859, + "step": 4691 + }, + { + "epoch": 0.4231410921224692, + "grad_norm": 1.7852361147855675, + "learning_rate": 2.586279599976689e-06, + "loss": 1.0212, + "step": 4692 + }, + { + "epoch": 0.4232312756459395, + "grad_norm": 1.4431199222308502, + "learning_rate": 2.585721028957468e-06, + "loss": 0.957, + "step": 4693 + }, + { + "epoch": 0.4233214591694098, + "grad_norm": 1.4140546176753939, + "learning_rate": 2.585162407961299e-06, + "loss": 0.9739, + "step": 4694 + }, + { + "epoch": 0.42341164269288, + "grad_norm": 2.4093082364303458, + "learning_rate": 2.584603737035847e-06, + "loss": 0.9036, + "step": 4695 + }, + { + "epoch": 0.4235018262163503, + "grad_norm": 1.6244188883577768, + "learning_rate": 2.5840450162287806e-06, + "loss": 0.9542, + "step": 4696 + }, + { + "epoch": 0.42359200973982053, + "grad_norm": 1.66711947330023, + "learning_rate": 2.583486245587774e-06, + "loss": 0.8723, + "step": 4697 + }, + { + "epoch": 0.4236821932632908, + "grad_norm": 1.7141010874714124, + "learning_rate": 2.5829274251605023e-06, + "loss": 1.0676, + "step": 4698 + }, + { + "epoch": 0.42377237678676105, + "grad_norm": 1.4851193619611356, + "learning_rate": 2.582368554994649e-06, + "loss": 0.9356, + "step": 4699 + }, + { + "epoch": 0.42386256031023134, + "grad_norm": 0.8117934970016324, + "learning_rate": 2.5818096351378994e-06, + "loss": 0.8148, + "step": 4700 + }, + { + "epoch": 0.4239527438337016, + "grad_norm": 1.6028764498591848, + "learning_rate": 2.5812506656379435e-06, + "loss": 1.0223, + "step": 4701 + }, + { + "epoch": 0.42404292735717186, + "grad_norm": 1.6799673157873183, + "learning_rate": 2.580691646542476e-06, + "loss": 0.912, + "step": 4702 + }, + { + "epoch": 0.4241331108806421, + "grad_norm": 1.66659317649306, + "learning_rate": 2.5801325778991958e-06, + "loss": 0.9981, + "step": 4703 + }, + { + "epoch": 0.4242232944041124, + "grad_norm": 3.330262693061519, + "learning_rate": 2.5795734597558043e-06, + "loss": 0.8129, + "step": 4704 + }, + { + "epoch": 0.4243134779275826, + "grad_norm": 1.8214130267380177, + "learning_rate": 2.579014292160011e-06, + "loss": 0.9607, + "step": 4705 + }, + { + "epoch": 0.4244036614510529, + "grad_norm": 1.4032384202140575, + "learning_rate": 2.5784550751595236e-06, + "loss": 0.9412, + "step": 4706 + }, + { + "epoch": 0.42449384497452314, + "grad_norm": 1.3564689090398319, + "learning_rate": 2.577895808802061e-06, + "loss": 1.0222, + "step": 4707 + }, + { + "epoch": 0.4245840284979934, + "grad_norm": 4.569491320791203, + "learning_rate": 2.577336493135341e-06, + "loss": 0.8796, + "step": 4708 + }, + { + "epoch": 0.42467421202146366, + "grad_norm": 1.9418188382005734, + "learning_rate": 2.576777128207088e-06, + "loss": 1.0592, + "step": 4709 + }, + { + "epoch": 0.42476439554493395, + "grad_norm": 1.4618776121927566, + "learning_rate": 2.5762177140650306e-06, + "loss": 0.9179, + "step": 4710 + }, + { + "epoch": 0.4248545790684042, + "grad_norm": 1.6614377808769554, + "learning_rate": 2.5756582507569003e-06, + "loss": 0.9665, + "step": 4711 + }, + { + "epoch": 0.42494476259187447, + "grad_norm": 1.41696938851942, + "learning_rate": 2.5750987383304335e-06, + "loss": 0.9859, + "step": 4712 + }, + { + "epoch": 0.4250349461153447, + "grad_norm": 1.8515080311163465, + "learning_rate": 2.574539176833372e-06, + "loss": 0.9453, + "step": 4713 + }, + { + "epoch": 0.425125129638815, + "grad_norm": 1.672266976583853, + "learning_rate": 2.5739795663134594e-06, + "loss": 0.9064, + "step": 4714 + }, + { + "epoch": 0.4252153131622852, + "grad_norm": 1.4096409075273173, + "learning_rate": 2.5734199068184454e-06, + "loss": 1.0476, + "step": 4715 + }, + { + "epoch": 0.4253054966857555, + "grad_norm": 1.529467732712971, + "learning_rate": 2.572860198396083e-06, + "loss": 0.8877, + "step": 4716 + }, + { + "epoch": 0.4253956802092258, + "grad_norm": 1.4909409514055219, + "learning_rate": 2.57230044109413e-06, + "loss": 0.9842, + "step": 4717 + }, + { + "epoch": 0.42548586373269603, + "grad_norm": 1.8959560424441204, + "learning_rate": 2.5717406349603483e-06, + "loss": 0.9391, + "step": 4718 + }, + { + "epoch": 0.4255760472561663, + "grad_norm": 1.4198231406955206, + "learning_rate": 2.5711807800425026e-06, + "loss": 0.9821, + "step": 4719 + }, + { + "epoch": 0.42566623077963656, + "grad_norm": 1.9438979432942156, + "learning_rate": 2.5706208763883633e-06, + "loss": 0.8001, + "step": 4720 + }, + { + "epoch": 0.42575641430310684, + "grad_norm": 1.5529957677239437, + "learning_rate": 2.570060924045704e-06, + "loss": 0.9202, + "step": 4721 + }, + { + "epoch": 0.4258465978265771, + "grad_norm": 1.4729656835418699, + "learning_rate": 2.569500923062304e-06, + "loss": 0.9922, + "step": 4722 + }, + { + "epoch": 0.42593678135004737, + "grad_norm": 1.5461198909606335, + "learning_rate": 2.5689408734859445e-06, + "loss": 0.8823, + "step": 4723 + }, + { + "epoch": 0.4260269648735176, + "grad_norm": 1.5713829074145096, + "learning_rate": 2.5683807753644127e-06, + "loss": 0.9083, + "step": 4724 + }, + { + "epoch": 0.4261171483969879, + "grad_norm": 1.400565760260614, + "learning_rate": 2.5678206287454996e-06, + "loss": 0.9392, + "step": 4725 + }, + { + "epoch": 0.4262073319204581, + "grad_norm": 1.901763702254956, + "learning_rate": 2.567260433676999e-06, + "loss": 0.9993, + "step": 4726 + }, + { + "epoch": 0.4262975154439284, + "grad_norm": 2.4072804412073863, + "learning_rate": 2.5667001902067107e-06, + "loss": 0.9326, + "step": 4727 + }, + { + "epoch": 0.42638769896739864, + "grad_norm": 1.6930130513111203, + "learning_rate": 2.566139898382437e-06, + "loss": 1.0143, + "step": 4728 + }, + { + "epoch": 0.42647788249086893, + "grad_norm": 1.524101699409466, + "learning_rate": 2.5655795582519853e-06, + "loss": 0.9753, + "step": 4729 + }, + { + "epoch": 0.42656806601433916, + "grad_norm": 1.507783800553659, + "learning_rate": 2.565019169863168e-06, + "loss": 1.0289, + "step": 4730 + }, + { + "epoch": 0.42665824953780945, + "grad_norm": 0.666292731398054, + "learning_rate": 2.5644587332637994e-06, + "loss": 0.7488, + "step": 4731 + }, + { + "epoch": 0.4267484330612797, + "grad_norm": 1.566557765111154, + "learning_rate": 2.5638982485016994e-06, + "loss": 0.9928, + "step": 4732 + }, + { + "epoch": 0.42683861658475, + "grad_norm": 0.7350829344300307, + "learning_rate": 2.5633377156246917e-06, + "loss": 0.7543, + "step": 4733 + }, + { + "epoch": 0.4269288001082202, + "grad_norm": 1.671361683984836, + "learning_rate": 2.562777134680603e-06, + "loss": 0.9461, + "step": 4734 + }, + { + "epoch": 0.4270189836316905, + "grad_norm": 1.8253305057525098, + "learning_rate": 2.562216505717267e-06, + "loss": 0.9712, + "step": 4735 + }, + { + "epoch": 0.42710916715516073, + "grad_norm": 1.6067369932529938, + "learning_rate": 2.561655828782518e-06, + "loss": 0.9833, + "step": 4736 + }, + { + "epoch": 0.427199350678631, + "grad_norm": 1.7194194530313558, + "learning_rate": 2.561095103924197e-06, + "loss": 0.9412, + "step": 4737 + }, + { + "epoch": 0.42728953420210125, + "grad_norm": 1.4952831014296324, + "learning_rate": 2.560534331190148e-06, + "loss": 0.9129, + "step": 4738 + }, + { + "epoch": 0.42737971772557154, + "grad_norm": 0.7211100352257551, + "learning_rate": 2.559973510628218e-06, + "loss": 0.7504, + "step": 4739 + }, + { + "epoch": 0.4274699012490418, + "grad_norm": 1.5239885165114802, + "learning_rate": 2.5594126422862615e-06, + "loss": 0.913, + "step": 4740 + }, + { + "epoch": 0.42756008477251206, + "grad_norm": 1.664163505819901, + "learning_rate": 2.558851726212134e-06, + "loss": 0.9122, + "step": 4741 + }, + { + "epoch": 0.42765026829598235, + "grad_norm": 1.5470842932710034, + "learning_rate": 2.5582907624536953e-06, + "loss": 0.9346, + "step": 4742 + }, + { + "epoch": 0.4277404518194526, + "grad_norm": 1.6486277538267595, + "learning_rate": 2.557729751058811e-06, + "loss": 0.9732, + "step": 4743 + }, + { + "epoch": 0.42783063534292287, + "grad_norm": 1.6226032628612796, + "learning_rate": 2.557168692075348e-06, + "loss": 0.9376, + "step": 4744 + }, + { + "epoch": 0.4279208188663931, + "grad_norm": 1.4487816493318777, + "learning_rate": 2.556607585551181e-06, + "loss": 0.9054, + "step": 4745 + }, + { + "epoch": 0.4280110023898634, + "grad_norm": 0.8505075552177014, + "learning_rate": 2.5560464315341844e-06, + "loss": 0.8791, + "step": 4746 + }, + { + "epoch": 0.4281011859133336, + "grad_norm": 1.684694501196781, + "learning_rate": 2.555485230072242e-06, + "loss": 0.9542, + "step": 4747 + }, + { + "epoch": 0.4281913694368039, + "grad_norm": 1.7571928839590027, + "learning_rate": 2.5549239812132354e-06, + "loss": 1.008, + "step": 4748 + }, + { + "epoch": 0.42828155296027415, + "grad_norm": 1.5659601427708651, + "learning_rate": 2.5543626850050556e-06, + "loss": 1.0308, + "step": 4749 + }, + { + "epoch": 0.42837173648374444, + "grad_norm": 1.4414137759955539, + "learning_rate": 2.5538013414955944e-06, + "loss": 1.0123, + "step": 4750 + }, + { + "epoch": 0.42846192000721467, + "grad_norm": 1.5881089251656884, + "learning_rate": 2.5532399507327494e-06, + "loss": 0.911, + "step": 4751 + }, + { + "epoch": 0.42855210353068496, + "grad_norm": 0.7602447724819626, + "learning_rate": 2.552678512764421e-06, + "loss": 0.7883, + "step": 4752 + }, + { + "epoch": 0.4286422870541552, + "grad_norm": 1.821512440746042, + "learning_rate": 2.5521170276385147e-06, + "loss": 1.0106, + "step": 4753 + }, + { + "epoch": 0.4287324705776255, + "grad_norm": 1.5538089129080968, + "learning_rate": 2.5515554954029394e-06, + "loss": 1.0071, + "step": 4754 + }, + { + "epoch": 0.4288226541010957, + "grad_norm": 1.5033938637337618, + "learning_rate": 2.550993916105608e-06, + "loss": 0.9876, + "step": 4755 + }, + { + "epoch": 0.428912837624566, + "grad_norm": 1.3891067360527174, + "learning_rate": 2.550432289794437e-06, + "loss": 0.9181, + "step": 4756 + }, + { + "epoch": 0.42900302114803623, + "grad_norm": 1.4942918090128139, + "learning_rate": 2.5498706165173483e-06, + "loss": 0.9726, + "step": 4757 + }, + { + "epoch": 0.4290932046715065, + "grad_norm": 1.3340009517952243, + "learning_rate": 2.5493088963222668e-06, + "loss": 0.9361, + "step": 4758 + }, + { + "epoch": 0.42918338819497676, + "grad_norm": 1.4739829646592226, + "learning_rate": 2.548747129257121e-06, + "loss": 1.0083, + "step": 4759 + }, + { + "epoch": 0.42927357171844704, + "grad_norm": 1.5019134397187193, + "learning_rate": 2.548185315369845e-06, + "loss": 0.8836, + "step": 4760 + }, + { + "epoch": 0.4293637552419173, + "grad_norm": 1.3910715479162525, + "learning_rate": 2.5476234547083746e-06, + "loss": 1.0507, + "step": 4761 + }, + { + "epoch": 0.42945393876538757, + "grad_norm": 1.5657536771665028, + "learning_rate": 2.547061547320652e-06, + "loss": 1.0323, + "step": 4762 + }, + { + "epoch": 0.4295441222888578, + "grad_norm": 1.2227746095740997, + "learning_rate": 2.5464995932546217e-06, + "loss": 0.9425, + "step": 4763 + }, + { + "epoch": 0.4296343058123281, + "grad_norm": 1.669213074286436, + "learning_rate": 2.545937592558232e-06, + "loss": 0.9124, + "step": 4764 + }, + { + "epoch": 0.4297244893357984, + "grad_norm": 1.4469121380317866, + "learning_rate": 2.5453755452794374e-06, + "loss": 0.911, + "step": 4765 + }, + { + "epoch": 0.4298146728592686, + "grad_norm": 1.3450275823891849, + "learning_rate": 2.5448134514661938e-06, + "loss": 1.0011, + "step": 4766 + }, + { + "epoch": 0.4299048563827389, + "grad_norm": 1.3353551744337167, + "learning_rate": 2.5442513111664623e-06, + "loss": 0.9573, + "step": 4767 + }, + { + "epoch": 0.42999503990620913, + "grad_norm": 1.8575421443794355, + "learning_rate": 2.5436891244282084e-06, + "loss": 0.9541, + "step": 4768 + }, + { + "epoch": 0.4300852234296794, + "grad_norm": 1.7374279199372673, + "learning_rate": 2.5431268912994004e-06, + "loss": 1.0592, + "step": 4769 + }, + { + "epoch": 0.43017540695314965, + "grad_norm": 1.491195348871738, + "learning_rate": 2.5425646118280108e-06, + "loss": 0.9327, + "step": 4770 + }, + { + "epoch": 0.43026559047661994, + "grad_norm": 0.6347242605963581, + "learning_rate": 2.5420022860620172e-06, + "loss": 0.7638, + "step": 4771 + }, + { + "epoch": 0.4303557740000902, + "grad_norm": 1.7501740369087881, + "learning_rate": 2.5414399140493995e-06, + "loss": 0.8836, + "step": 4772 + }, + { + "epoch": 0.43044595752356046, + "grad_norm": 1.8315994632299453, + "learning_rate": 2.5408774958381436e-06, + "loss": 0.8976, + "step": 4773 + }, + { + "epoch": 0.4305361410470307, + "grad_norm": 1.4982545552037279, + "learning_rate": 2.540315031476237e-06, + "loss": 0.997, + "step": 4774 + }, + { + "epoch": 0.430626324570501, + "grad_norm": 1.5080218392164322, + "learning_rate": 2.5397525210116737e-06, + "loss": 0.9668, + "step": 4775 + }, + { + "epoch": 0.4307165080939712, + "grad_norm": 1.497043158974842, + "learning_rate": 2.539189964492448e-06, + "loss": 0.9152, + "step": 4776 + }, + { + "epoch": 0.4308066916174415, + "grad_norm": 1.5843126965674297, + "learning_rate": 2.5386273619665613e-06, + "loss": 0.998, + "step": 4777 + }, + { + "epoch": 0.43089687514091174, + "grad_norm": 0.7449417807725447, + "learning_rate": 2.5380647134820186e-06, + "loss": 0.7849, + "step": 4778 + }, + { + "epoch": 0.43098705866438203, + "grad_norm": 1.3519417115010495, + "learning_rate": 2.5375020190868277e-06, + "loss": 1.014, + "step": 4779 + }, + { + "epoch": 0.43107724218785226, + "grad_norm": 1.4990095206245269, + "learning_rate": 2.536939278829001e-06, + "loss": 0.9505, + "step": 4780 + }, + { + "epoch": 0.43116742571132255, + "grad_norm": 1.685722262384417, + "learning_rate": 2.5363764927565536e-06, + "loss": 1.0199, + "step": 4781 + }, + { + "epoch": 0.4312576092347928, + "grad_norm": 1.6870351080370294, + "learning_rate": 2.5358136609175064e-06, + "loss": 0.9137, + "step": 4782 + }, + { + "epoch": 0.43134779275826307, + "grad_norm": 1.3192838465122476, + "learning_rate": 2.535250783359884e-06, + "loss": 0.9576, + "step": 4783 + }, + { + "epoch": 0.4314379762817333, + "grad_norm": 1.7705811339659576, + "learning_rate": 2.5346878601317124e-06, + "loss": 0.8526, + "step": 4784 + }, + { + "epoch": 0.4315281598052036, + "grad_norm": 0.6904315867526034, + "learning_rate": 2.534124891281025e-06, + "loss": 0.7483, + "step": 4785 + }, + { + "epoch": 0.4316183433286738, + "grad_norm": 1.6012124027838204, + "learning_rate": 2.533561876855857e-06, + "loss": 0.9707, + "step": 4786 + }, + { + "epoch": 0.4317085268521441, + "grad_norm": 1.6351319407384368, + "learning_rate": 2.532998816904247e-06, + "loss": 0.9127, + "step": 4787 + }, + { + "epoch": 0.4317987103756144, + "grad_norm": 1.5263017382900974, + "learning_rate": 2.53243571147424e-06, + "loss": 0.8522, + "step": 4788 + }, + { + "epoch": 0.43188889389908464, + "grad_norm": 1.9280315506359855, + "learning_rate": 2.5318725606138815e-06, + "loss": 0.8656, + "step": 4789 + }, + { + "epoch": 0.4319790774225549, + "grad_norm": 1.6763106356925066, + "learning_rate": 2.5313093643712235e-06, + "loss": 0.9659, + "step": 4790 + }, + { + "epoch": 0.43206926094602516, + "grad_norm": 1.4258681179391204, + "learning_rate": 2.530746122794321e-06, + "loss": 1.0071, + "step": 4791 + }, + { + "epoch": 0.43215944446949545, + "grad_norm": 1.3243309131033387, + "learning_rate": 2.5301828359312323e-06, + "loss": 0.8847, + "step": 4792 + }, + { + "epoch": 0.4322496279929657, + "grad_norm": 1.5971642368498131, + "learning_rate": 2.529619503830021e-06, + "loss": 0.9163, + "step": 4793 + }, + { + "epoch": 0.43233981151643597, + "grad_norm": 1.7152815761687126, + "learning_rate": 2.529056126538753e-06, + "loss": 0.891, + "step": 4794 + }, + { + "epoch": 0.4324299950399062, + "grad_norm": 1.4345379871857968, + "learning_rate": 2.5284927041054995e-06, + "loss": 0.9886, + "step": 4795 + }, + { + "epoch": 0.4325201785633765, + "grad_norm": 0.7409080720911505, + "learning_rate": 2.5279292365783348e-06, + "loss": 0.7259, + "step": 4796 + }, + { + "epoch": 0.4326103620868467, + "grad_norm": 1.4762531743673617, + "learning_rate": 2.527365724005336e-06, + "loss": 1.0464, + "step": 4797 + }, + { + "epoch": 0.432700545610317, + "grad_norm": 1.667059049511293, + "learning_rate": 2.526802166434586e-06, + "loss": 1.0111, + "step": 4798 + }, + { + "epoch": 0.43279072913378724, + "grad_norm": 1.351872447201474, + "learning_rate": 2.5262385639141708e-06, + "loss": 0.9214, + "step": 4799 + }, + { + "epoch": 0.43288091265725753, + "grad_norm": 1.6787999846855854, + "learning_rate": 2.525674916492179e-06, + "loss": 1.0299, + "step": 4800 + }, + { + "epoch": 0.43297109618072777, + "grad_norm": 1.3575035454860072, + "learning_rate": 2.5251112242167056e-06, + "loss": 0.9639, + "step": 4801 + }, + { + "epoch": 0.43306127970419805, + "grad_norm": 1.741655619219676, + "learning_rate": 2.5245474871358464e-06, + "loss": 0.9234, + "step": 4802 + }, + { + "epoch": 0.4331514632276683, + "grad_norm": 1.7087365194150304, + "learning_rate": 2.5239837052977037e-06, + "loss": 0.8756, + "step": 4803 + }, + { + "epoch": 0.4332416467511386, + "grad_norm": 1.7779018992897007, + "learning_rate": 2.523419878750381e-06, + "loss": 1.0691, + "step": 4804 + }, + { + "epoch": 0.4333318302746088, + "grad_norm": 1.457368131662451, + "learning_rate": 2.522856007541989e-06, + "loss": 0.9114, + "step": 4805 + }, + { + "epoch": 0.4334220137980791, + "grad_norm": 1.49898788956731, + "learning_rate": 2.5222920917206397e-06, + "loss": 0.8718, + "step": 4806 + }, + { + "epoch": 0.43351219732154933, + "grad_norm": 0.7565552418615509, + "learning_rate": 2.5217281313344493e-06, + "loss": 0.7638, + "step": 4807 + }, + { + "epoch": 0.4336023808450196, + "grad_norm": 1.548346311037818, + "learning_rate": 2.5211641264315372e-06, + "loss": 0.9997, + "step": 4808 + }, + { + "epoch": 0.43369256436848985, + "grad_norm": 1.4380097522006992, + "learning_rate": 2.5206000770600286e-06, + "loss": 0.9379, + "step": 4809 + }, + { + "epoch": 0.43378274789196014, + "grad_norm": 1.6323063262040975, + "learning_rate": 2.520035983268051e-06, + "loss": 0.9156, + "step": 4810 + }, + { + "epoch": 0.4338729314154304, + "grad_norm": 1.452453376034857, + "learning_rate": 2.5194718451037357e-06, + "loss": 0.9577, + "step": 4811 + }, + { + "epoch": 0.43396311493890066, + "grad_norm": 1.5567228709549898, + "learning_rate": 2.518907662615218e-06, + "loss": 0.9842, + "step": 4812 + }, + { + "epoch": 0.43405329846237095, + "grad_norm": 2.3236215545466226, + "learning_rate": 2.5183434358506373e-06, + "loss": 0.9163, + "step": 4813 + }, + { + "epoch": 0.4341434819858412, + "grad_norm": 1.54848943136302, + "learning_rate": 2.5177791648581368e-06, + "loss": 0.9414, + "step": 4814 + }, + { + "epoch": 0.4342336655093115, + "grad_norm": 1.5326282759929255, + "learning_rate": 2.517214849685863e-06, + "loss": 0.9629, + "step": 4815 + }, + { + "epoch": 0.4343238490327817, + "grad_norm": 1.5116263270156602, + "learning_rate": 2.5166504903819663e-06, + "loss": 0.97, + "step": 4816 + }, + { + "epoch": 0.434414032556252, + "grad_norm": 1.9288821373780827, + "learning_rate": 2.5160860869946014e-06, + "loss": 1.0329, + "step": 4817 + }, + { + "epoch": 0.4345042160797222, + "grad_norm": 1.4341491122430265, + "learning_rate": 2.5155216395719253e-06, + "loss": 1.04, + "step": 4818 + }, + { + "epoch": 0.4345943996031925, + "grad_norm": 1.2971515590437057, + "learning_rate": 2.5149571481621e-06, + "loss": 0.9187, + "step": 4819 + }, + { + "epoch": 0.43468458312666275, + "grad_norm": 0.6554797966154694, + "learning_rate": 2.514392612813292e-06, + "loss": 0.7458, + "step": 4820 + }, + { + "epoch": 0.43477476665013304, + "grad_norm": 1.575981119139605, + "learning_rate": 2.5138280335736695e-06, + "loss": 1.0135, + "step": 4821 + }, + { + "epoch": 0.43486495017360327, + "grad_norm": 1.5952559443358432, + "learning_rate": 2.5132634104914064e-06, + "loss": 1.0396, + "step": 4822 + }, + { + "epoch": 0.43495513369707356, + "grad_norm": 1.3341365967522367, + "learning_rate": 2.5126987436146794e-06, + "loss": 0.9705, + "step": 4823 + }, + { + "epoch": 0.4350453172205438, + "grad_norm": 1.6064738076880052, + "learning_rate": 2.5121340329916675e-06, + "loss": 0.9381, + "step": 4824 + }, + { + "epoch": 0.4351355007440141, + "grad_norm": 1.5257374934471928, + "learning_rate": 2.5115692786705566e-06, + "loss": 0.8809, + "step": 4825 + }, + { + "epoch": 0.4352256842674843, + "grad_norm": 1.8272598045087123, + "learning_rate": 2.511004480699534e-06, + "loss": 0.998, + "step": 4826 + }, + { + "epoch": 0.4353158677909546, + "grad_norm": 0.8263395603712655, + "learning_rate": 2.510439639126791e-06, + "loss": 0.8203, + "step": 4827 + }, + { + "epoch": 0.43540605131442484, + "grad_norm": 1.5184721013515552, + "learning_rate": 2.509874754000524e-06, + "loss": 0.9571, + "step": 4828 + }, + { + "epoch": 0.4354962348378951, + "grad_norm": 1.7077263785074288, + "learning_rate": 2.509309825368932e-06, + "loss": 0.9432, + "step": 4829 + }, + { + "epoch": 0.43558641836136536, + "grad_norm": 1.9378506127878639, + "learning_rate": 2.5087448532802173e-06, + "loss": 0.9609, + "step": 4830 + }, + { + "epoch": 0.43567660188483565, + "grad_norm": 1.5228985737962186, + "learning_rate": 2.508179837782586e-06, + "loss": 0.9384, + "step": 4831 + }, + { + "epoch": 0.4357667854083059, + "grad_norm": 1.6126129761382073, + "learning_rate": 2.5076147789242493e-06, + "loss": 0.9284, + "step": 4832 + }, + { + "epoch": 0.43585696893177617, + "grad_norm": 2.1978075503582803, + "learning_rate": 2.5070496767534202e-06, + "loss": 0.9737, + "step": 4833 + }, + { + "epoch": 0.4359471524552464, + "grad_norm": 1.6585136635015958, + "learning_rate": 2.506484531318317e-06, + "loss": 0.9798, + "step": 4834 + }, + { + "epoch": 0.4360373359787167, + "grad_norm": 0.8833511532179658, + "learning_rate": 2.5059193426671613e-06, + "loss": 0.7926, + "step": 4835 + }, + { + "epoch": 0.436127519502187, + "grad_norm": 1.6496159829132244, + "learning_rate": 2.5053541108481772e-06, + "loss": 1.0873, + "step": 4836 + }, + { + "epoch": 0.4362177030256572, + "grad_norm": 1.6132995752212853, + "learning_rate": 2.5047888359095935e-06, + "loss": 0.9443, + "step": 4837 + }, + { + "epoch": 0.4363078865491275, + "grad_norm": 1.4906525231465526, + "learning_rate": 2.5042235178996436e-06, + "loss": 0.9392, + "step": 4838 + }, + { + "epoch": 0.43639807007259773, + "grad_norm": 1.5598511941558868, + "learning_rate": 2.5036581568665627e-06, + "loss": 1.0188, + "step": 4839 + }, + { + "epoch": 0.436488253596068, + "grad_norm": 1.4384093518168481, + "learning_rate": 2.503092752858591e-06, + "loss": 1.0188, + "step": 4840 + }, + { + "epoch": 0.43657843711953825, + "grad_norm": 0.8418955380002644, + "learning_rate": 2.502527305923971e-06, + "loss": 0.8097, + "step": 4841 + }, + { + "epoch": 0.43666862064300854, + "grad_norm": 1.9106681153508667, + "learning_rate": 2.5019618161109506e-06, + "loss": 0.9474, + "step": 4842 + }, + { + "epoch": 0.4367588041664788, + "grad_norm": 1.5018230327325914, + "learning_rate": 2.5013962834677804e-06, + "loss": 0.9763, + "step": 4843 + }, + { + "epoch": 0.43684898768994906, + "grad_norm": 1.6522885577135278, + "learning_rate": 2.500830708042715e-06, + "loss": 0.9445, + "step": 4844 + }, + { + "epoch": 0.4369391712134193, + "grad_norm": 1.6922645417460884, + "learning_rate": 2.500265089884011e-06, + "loss": 1.0264, + "step": 4845 + }, + { + "epoch": 0.4370293547368896, + "grad_norm": 1.5402895301039514, + "learning_rate": 2.499699429039932e-06, + "loss": 0.958, + "step": 4846 + }, + { + "epoch": 0.4371195382603598, + "grad_norm": 1.7027982345826396, + "learning_rate": 2.4991337255587425e-06, + "loss": 0.9107, + "step": 4847 + }, + { + "epoch": 0.4372097217838301, + "grad_norm": 2.0073068182359717, + "learning_rate": 2.4985679794887106e-06, + "loss": 0.8295, + "step": 4848 + }, + { + "epoch": 0.43729990530730034, + "grad_norm": 2.277720189594183, + "learning_rate": 2.49800219087811e-06, + "loss": 0.9947, + "step": 4849 + }, + { + "epoch": 0.43739008883077063, + "grad_norm": 1.8776641991304102, + "learning_rate": 2.4974363597752163e-06, + "loss": 0.9914, + "step": 4850 + }, + { + "epoch": 0.43748027235424086, + "grad_norm": 1.6890976724835456, + "learning_rate": 2.4968704862283097e-06, + "loss": 1.0057, + "step": 4851 + }, + { + "epoch": 0.43757045587771115, + "grad_norm": 2.402880402457772, + "learning_rate": 2.4963045702856737e-06, + "loss": 1.0274, + "step": 4852 + }, + { + "epoch": 0.4376606394011814, + "grad_norm": 1.4667272233743653, + "learning_rate": 2.4957386119955946e-06, + "loss": 0.9016, + "step": 4853 + }, + { + "epoch": 0.4377508229246517, + "grad_norm": 0.7483977846825423, + "learning_rate": 2.495172611406364e-06, + "loss": 0.7698, + "step": 4854 + }, + { + "epoch": 0.4378410064481219, + "grad_norm": 1.7695859517448687, + "learning_rate": 2.4946065685662757e-06, + "loss": 0.9864, + "step": 4855 + }, + { + "epoch": 0.4379311899715922, + "grad_norm": 1.8021325194715334, + "learning_rate": 2.4940404835236283e-06, + "loss": 0.958, + "step": 4856 + }, + { + "epoch": 0.4380213734950624, + "grad_norm": 1.3923177810146172, + "learning_rate": 2.4934743563267223e-06, + "loss": 0.9487, + "step": 4857 + }, + { + "epoch": 0.4381115570185327, + "grad_norm": 1.5947920440666736, + "learning_rate": 2.4929081870238635e-06, + "loss": 0.9263, + "step": 4858 + }, + { + "epoch": 0.43820174054200295, + "grad_norm": 1.656899306940296, + "learning_rate": 2.49234197566336e-06, + "loss": 0.9968, + "step": 4859 + }, + { + "epoch": 0.43829192406547324, + "grad_norm": 1.5710262365645333, + "learning_rate": 2.4917757222935247e-06, + "loss": 0.9747, + "step": 4860 + }, + { + "epoch": 0.4383821075889435, + "grad_norm": 1.6983915573686514, + "learning_rate": 2.4912094269626725e-06, + "loss": 0.9747, + "step": 4861 + }, + { + "epoch": 0.43847229111241376, + "grad_norm": 1.839530966831692, + "learning_rate": 2.4906430897191245e-06, + "loss": 0.9946, + "step": 4862 + }, + { + "epoch": 0.43856247463588405, + "grad_norm": 1.3514763247850359, + "learning_rate": 2.490076710611202e-06, + "loss": 0.9943, + "step": 4863 + }, + { + "epoch": 0.4386526581593543, + "grad_norm": 1.6979281731276101, + "learning_rate": 2.4895102896872326e-06, + "loss": 1.0061, + "step": 4864 + }, + { + "epoch": 0.43874284168282457, + "grad_norm": 1.4912116569712164, + "learning_rate": 2.4889438269955457e-06, + "loss": 0.8893, + "step": 4865 + }, + { + "epoch": 0.4388330252062948, + "grad_norm": 1.5815515615118532, + "learning_rate": 2.4883773225844755e-06, + "loss": 0.944, + "step": 4866 + }, + { + "epoch": 0.4389232087297651, + "grad_norm": 1.5569628450227224, + "learning_rate": 2.48781077650236e-06, + "loss": 0.9462, + "step": 4867 + }, + { + "epoch": 0.4390133922532353, + "grad_norm": 1.61571446656186, + "learning_rate": 2.4872441887975386e-06, + "loss": 0.9198, + "step": 4868 + }, + { + "epoch": 0.4391035757767056, + "grad_norm": 1.2868613205229447, + "learning_rate": 2.486677559518356e-06, + "loss": 0.8988, + "step": 4869 + }, + { + "epoch": 0.43919375930017585, + "grad_norm": 1.2633631239140652, + "learning_rate": 2.4861108887131614e-06, + "loss": 0.9466, + "step": 4870 + }, + { + "epoch": 0.43928394282364613, + "grad_norm": 1.6928736992291387, + "learning_rate": 2.485544176430305e-06, + "loss": 0.9243, + "step": 4871 + }, + { + "epoch": 0.43937412634711637, + "grad_norm": 1.612088702480393, + "learning_rate": 2.4849774227181425e-06, + "loss": 1.035, + "step": 4872 + }, + { + "epoch": 0.43946430987058666, + "grad_norm": 1.4973146720673347, + "learning_rate": 2.484410627625032e-06, + "loss": 0.9544, + "step": 4873 + }, + { + "epoch": 0.4395544933940569, + "grad_norm": 1.6681349088646797, + "learning_rate": 2.4838437911993356e-06, + "loss": 1.0042, + "step": 4874 + }, + { + "epoch": 0.4396446769175272, + "grad_norm": 1.943309564766808, + "learning_rate": 2.483276913489419e-06, + "loss": 0.9978, + "step": 4875 + }, + { + "epoch": 0.4397348604409974, + "grad_norm": 2.020238638086024, + "learning_rate": 2.4827099945436516e-06, + "loss": 0.8581, + "step": 4876 + }, + { + "epoch": 0.4398250439644677, + "grad_norm": 1.7848472411241696, + "learning_rate": 2.482143034410405e-06, + "loss": 1.0385, + "step": 4877 + }, + { + "epoch": 0.43991522748793793, + "grad_norm": 1.5406221339448651, + "learning_rate": 2.4815760331380573e-06, + "loss": 0.9866, + "step": 4878 + }, + { + "epoch": 0.4400054110114082, + "grad_norm": 1.415183871423906, + "learning_rate": 2.481008990774987e-06, + "loss": 0.9165, + "step": 4879 + }, + { + "epoch": 0.44009559453487845, + "grad_norm": 1.538397398400252, + "learning_rate": 2.480441907369577e-06, + "loss": 0.9556, + "step": 4880 + }, + { + "epoch": 0.44018577805834874, + "grad_norm": 1.8265655256495938, + "learning_rate": 2.479874782970214e-06, + "loss": 0.9026, + "step": 4881 + }, + { + "epoch": 0.440275961581819, + "grad_norm": 2.164809225544299, + "learning_rate": 2.4793076176252887e-06, + "loss": 0.98, + "step": 4882 + }, + { + "epoch": 0.44036614510528926, + "grad_norm": 1.5409207042861042, + "learning_rate": 2.478740411383195e-06, + "loss": 0.8415, + "step": 4883 + }, + { + "epoch": 0.44045632862875955, + "grad_norm": 1.3201960224486395, + "learning_rate": 2.4781731642923296e-06, + "loss": 0.9499, + "step": 4884 + }, + { + "epoch": 0.4405465121522298, + "grad_norm": 1.9245717426861415, + "learning_rate": 2.477605876401093e-06, + "loss": 0.9409, + "step": 4885 + }, + { + "epoch": 0.4406366956757001, + "grad_norm": 1.381374365614767, + "learning_rate": 2.4770385477578894e-06, + "loss": 0.9114, + "step": 4886 + }, + { + "epoch": 0.4407268791991703, + "grad_norm": 1.3051292904937752, + "learning_rate": 2.476471178411127e-06, + "loss": 0.9163, + "step": 4887 + }, + { + "epoch": 0.4408170627226406, + "grad_norm": 1.3697641996039993, + "learning_rate": 2.475903768409216e-06, + "loss": 0.9773, + "step": 4888 + }, + { + "epoch": 0.44090724624611083, + "grad_norm": 1.9667110481211412, + "learning_rate": 2.475336317800572e-06, + "loss": 0.9288, + "step": 4889 + }, + { + "epoch": 0.4409974297695811, + "grad_norm": 1.7625715471179366, + "learning_rate": 2.4747688266336118e-06, + "loss": 0.9452, + "step": 4890 + }, + { + "epoch": 0.44108761329305135, + "grad_norm": 1.6890239898215622, + "learning_rate": 2.4742012949567574e-06, + "loss": 1.0628, + "step": 4891 + }, + { + "epoch": 0.44117779681652164, + "grad_norm": 1.743030292924016, + "learning_rate": 2.4736337228184338e-06, + "loss": 0.9419, + "step": 4892 + }, + { + "epoch": 0.44126798033999187, + "grad_norm": 1.8173582802028405, + "learning_rate": 2.4730661102670692e-06, + "loss": 1.0609, + "step": 4893 + }, + { + "epoch": 0.44135816386346216, + "grad_norm": 1.5684473148456122, + "learning_rate": 2.472498457351096e-06, + "loss": 0.9605, + "step": 4894 + }, + { + "epoch": 0.4414483473869324, + "grad_norm": 1.9490955137998427, + "learning_rate": 2.4719307641189495e-06, + "loss": 0.9454, + "step": 4895 + }, + { + "epoch": 0.4415385309104027, + "grad_norm": 1.8601751449118775, + "learning_rate": 2.4713630306190673e-06, + "loss": 0.9235, + "step": 4896 + }, + { + "epoch": 0.4416287144338729, + "grad_norm": 1.7381377085879883, + "learning_rate": 2.4707952568998923e-06, + "loss": 1.0419, + "step": 4897 + }, + { + "epoch": 0.4417188979573432, + "grad_norm": 1.8536910043435908, + "learning_rate": 2.4702274430098703e-06, + "loss": 0.932, + "step": 4898 + }, + { + "epoch": 0.44180908148081344, + "grad_norm": 2.283368838050444, + "learning_rate": 2.4696595889974497e-06, + "loss": 0.9174, + "step": 4899 + }, + { + "epoch": 0.4418992650042837, + "grad_norm": 1.4337172286155413, + "learning_rate": 2.469091694911084e-06, + "loss": 1.012, + "step": 4900 + }, + { + "epoch": 0.44198944852775396, + "grad_norm": 1.850903421912971, + "learning_rate": 2.4685237607992276e-06, + "loss": 0.9551, + "step": 4901 + }, + { + "epoch": 0.44207963205122425, + "grad_norm": 1.6664114836205102, + "learning_rate": 2.4679557867103416e-06, + "loss": 1.0194, + "step": 4902 + }, + { + "epoch": 0.4421698155746945, + "grad_norm": 1.3980580672918985, + "learning_rate": 2.4673877726928865e-06, + "loss": 0.9638, + "step": 4903 + }, + { + "epoch": 0.44225999909816477, + "grad_norm": 1.8043671691111502, + "learning_rate": 2.46681971879533e-06, + "loss": 0.8531, + "step": 4904 + }, + { + "epoch": 0.442350182621635, + "grad_norm": 1.7131783081801195, + "learning_rate": 2.4662516250661407e-06, + "loss": 0.8839, + "step": 4905 + }, + { + "epoch": 0.4424403661451053, + "grad_norm": 1.6869475555363316, + "learning_rate": 2.465683491553792e-06, + "loss": 0.933, + "step": 4906 + }, + { + "epoch": 0.4425305496685755, + "grad_norm": 1.4705941817219506, + "learning_rate": 2.4651153183067604e-06, + "loss": 0.9642, + "step": 4907 + }, + { + "epoch": 0.4426207331920458, + "grad_norm": 1.836418381067249, + "learning_rate": 2.4645471053735245e-06, + "loss": 0.8954, + "step": 4908 + }, + { + "epoch": 0.4427109167155161, + "grad_norm": 1.4965296974179652, + "learning_rate": 2.4639788528025684e-06, + "loss": 0.8934, + "step": 4909 + }, + { + "epoch": 0.44280110023898633, + "grad_norm": 1.2633985079273558, + "learning_rate": 2.463410560642378e-06, + "loss": 0.9626, + "step": 4910 + }, + { + "epoch": 0.4428912837624566, + "grad_norm": 0.6954105393802813, + "learning_rate": 2.4628422289414448e-06, + "loss": 0.7902, + "step": 4911 + }, + { + "epoch": 0.44298146728592686, + "grad_norm": 0.705333061307297, + "learning_rate": 2.4622738577482592e-06, + "loss": 0.766, + "step": 4912 + }, + { + "epoch": 0.44307165080939714, + "grad_norm": 1.5056799795358728, + "learning_rate": 2.461705447111319e-06, + "loss": 0.9445, + "step": 4913 + }, + { + "epoch": 0.4431618343328674, + "grad_norm": 1.684559839188457, + "learning_rate": 2.4611369970791246e-06, + "loss": 0.9301, + "step": 4914 + }, + { + "epoch": 0.44325201785633767, + "grad_norm": 1.7929867679131095, + "learning_rate": 2.460568507700179e-06, + "loss": 0.9625, + "step": 4915 + }, + { + "epoch": 0.4433422013798079, + "grad_norm": 1.6363499483104373, + "learning_rate": 2.4599999790229887e-06, + "loss": 0.9977, + "step": 4916 + }, + { + "epoch": 0.4434323849032782, + "grad_norm": 1.997365109946359, + "learning_rate": 2.459431411096064e-06, + "loss": 0.9157, + "step": 4917 + }, + { + "epoch": 0.4435225684267484, + "grad_norm": 1.658688172120228, + "learning_rate": 2.458862803967918e-06, + "loss": 0.928, + "step": 4918 + }, + { + "epoch": 0.4436127519502187, + "grad_norm": 1.433678897335126, + "learning_rate": 2.4582941576870667e-06, + "loss": 0.9727, + "step": 4919 + }, + { + "epoch": 0.44370293547368894, + "grad_norm": 1.618557777922, + "learning_rate": 2.4577254723020315e-06, + "loss": 0.9405, + "step": 4920 + }, + { + "epoch": 0.44379311899715923, + "grad_norm": 1.6223483091428268, + "learning_rate": 2.457156747861335e-06, + "loss": 0.9767, + "step": 4921 + }, + { + "epoch": 0.44388330252062946, + "grad_norm": 1.6459992803940526, + "learning_rate": 2.456587984413504e-06, + "loss": 1.0237, + "step": 4922 + }, + { + "epoch": 0.44397348604409975, + "grad_norm": 1.407525056714203, + "learning_rate": 2.4560191820070683e-06, + "loss": 0.8886, + "step": 4923 + }, + { + "epoch": 0.44406366956757, + "grad_norm": 1.6489300466566463, + "learning_rate": 2.4554503406905617e-06, + "loss": 0.9407, + "step": 4924 + }, + { + "epoch": 0.4441538530910403, + "grad_norm": 1.4960163783691525, + "learning_rate": 2.454881460512521e-06, + "loss": 0.9968, + "step": 4925 + }, + { + "epoch": 0.4442440366145105, + "grad_norm": 1.2836389367664531, + "learning_rate": 2.4543125415214856e-06, + "loss": 0.9236, + "step": 4926 + }, + { + "epoch": 0.4443342201379808, + "grad_norm": 1.7683516522224783, + "learning_rate": 2.4537435837659996e-06, + "loss": 0.9806, + "step": 4927 + }, + { + "epoch": 0.44442440366145103, + "grad_norm": 1.729444489125568, + "learning_rate": 2.4531745872946085e-06, + "loss": 0.9099, + "step": 4928 + }, + { + "epoch": 0.4445145871849213, + "grad_norm": 1.7458311561929, + "learning_rate": 2.4526055521558632e-06, + "loss": 1.0197, + "step": 4929 + }, + { + "epoch": 0.44460477070839155, + "grad_norm": 1.5249739503980717, + "learning_rate": 2.4520364783983164e-06, + "loss": 0.9748, + "step": 4930 + }, + { + "epoch": 0.44469495423186184, + "grad_norm": 1.4253594597639483, + "learning_rate": 2.451467366070525e-06, + "loss": 0.9414, + "step": 4931 + }, + { + "epoch": 0.4447851377553321, + "grad_norm": 1.7686975127611106, + "learning_rate": 2.450898215221048e-06, + "loss": 1.0277, + "step": 4932 + }, + { + "epoch": 0.44487532127880236, + "grad_norm": 1.8131375342550244, + "learning_rate": 2.4503290258984498e-06, + "loss": 1.1048, + "step": 4933 + }, + { + "epoch": 0.44496550480227265, + "grad_norm": 1.7638101618595061, + "learning_rate": 2.4497597981512952e-06, + "loss": 0.8234, + "step": 4934 + }, + { + "epoch": 0.4450556883257429, + "grad_norm": 0.7163529853613216, + "learning_rate": 2.4491905320281555e-06, + "loss": 0.7934, + "step": 4935 + }, + { + "epoch": 0.44514587184921317, + "grad_norm": 1.5369053163520758, + "learning_rate": 2.448621227577602e-06, + "loss": 0.9339, + "step": 4936 + }, + { + "epoch": 0.4452360553726834, + "grad_norm": 1.6762166913596888, + "learning_rate": 2.4480518848482123e-06, + "loss": 0.9003, + "step": 4937 + }, + { + "epoch": 0.4453262388961537, + "grad_norm": 2.4568703603300404, + "learning_rate": 2.447482503888565e-06, + "loss": 0.9717, + "step": 4938 + }, + { + "epoch": 0.4454164224196239, + "grad_norm": 2.3937969233044756, + "learning_rate": 2.4469130847472434e-06, + "loss": 0.9385, + "step": 4939 + }, + { + "epoch": 0.4455066059430942, + "grad_norm": 1.631569128865101, + "learning_rate": 2.4463436274728326e-06, + "loss": 0.9968, + "step": 4940 + }, + { + "epoch": 0.44559678946656445, + "grad_norm": 1.5498847580184203, + "learning_rate": 2.4457741321139227e-06, + "loss": 0.8037, + "step": 4941 + }, + { + "epoch": 0.44568697299003474, + "grad_norm": 1.8196333899519774, + "learning_rate": 2.4452045987191063e-06, + "loss": 0.9775, + "step": 4942 + }, + { + "epoch": 0.44577715651350497, + "grad_norm": 1.589356218797826, + "learning_rate": 2.4446350273369776e-06, + "loss": 0.9297, + "step": 4943 + }, + { + "epoch": 0.44586734003697526, + "grad_norm": 1.3804549038963894, + "learning_rate": 2.4440654180161374e-06, + "loss": 0.9929, + "step": 4944 + }, + { + "epoch": 0.4459575235604455, + "grad_norm": 1.3875598774267432, + "learning_rate": 2.4434957708051875e-06, + "loss": 0.9145, + "step": 4945 + }, + { + "epoch": 0.4460477070839158, + "grad_norm": 1.6453866070670742, + "learning_rate": 2.4429260857527324e-06, + "loss": 0.8982, + "step": 4946 + }, + { + "epoch": 0.446137890607386, + "grad_norm": 1.639479973268479, + "learning_rate": 2.4423563629073815e-06, + "loss": 0.8612, + "step": 4947 + }, + { + "epoch": 0.4462280741308563, + "grad_norm": 1.3348702076885646, + "learning_rate": 2.4417866023177466e-06, + "loss": 0.956, + "step": 4948 + }, + { + "epoch": 0.44631825765432653, + "grad_norm": 1.6616811413582493, + "learning_rate": 2.441216804032443e-06, + "loss": 0.9849, + "step": 4949 + }, + { + "epoch": 0.4464084411777968, + "grad_norm": 1.5082807851387776, + "learning_rate": 2.440646968100089e-06, + "loss": 0.9717, + "step": 4950 + }, + { + "epoch": 0.44649862470126706, + "grad_norm": 0.7618289183151622, + "learning_rate": 2.4400770945693055e-06, + "loss": 0.7755, + "step": 4951 + }, + { + "epoch": 0.44658880822473734, + "grad_norm": 1.6027912171650844, + "learning_rate": 2.4395071834887177e-06, + "loss": 0.9931, + "step": 4952 + }, + { + "epoch": 0.4466789917482076, + "grad_norm": 1.4391039317376537, + "learning_rate": 2.438937234906954e-06, + "loss": 0.8363, + "step": 4953 + }, + { + "epoch": 0.44676917527167787, + "grad_norm": 1.5461336150964446, + "learning_rate": 2.4383672488726447e-06, + "loss": 0.9452, + "step": 4954 + }, + { + "epoch": 0.44685935879514815, + "grad_norm": 1.6585743267343331, + "learning_rate": 2.4377972254344256e-06, + "loss": 0.8829, + "step": 4955 + }, + { + "epoch": 0.4469495423186184, + "grad_norm": 1.709068636315609, + "learning_rate": 2.437227164640932e-06, + "loss": 0.8406, + "step": 4956 + }, + { + "epoch": 0.4470397258420887, + "grad_norm": 1.4904335458830775, + "learning_rate": 2.436657066540807e-06, + "loss": 0.8995, + "step": 4957 + }, + { + "epoch": 0.4471299093655589, + "grad_norm": 1.8831045470344034, + "learning_rate": 2.4360869311826927e-06, + "loss": 0.8314, + "step": 4958 + }, + { + "epoch": 0.4472200928890292, + "grad_norm": 1.847334722887111, + "learning_rate": 2.4355167586152367e-06, + "loss": 1.0491, + "step": 4959 + }, + { + "epoch": 0.44731027641249943, + "grad_norm": 1.723470344126742, + "learning_rate": 2.4349465488870896e-06, + "loss": 0.9263, + "step": 4960 + }, + { + "epoch": 0.4474004599359697, + "grad_norm": 1.4475177334241092, + "learning_rate": 2.434376302046905e-06, + "loss": 1.0433, + "step": 4961 + }, + { + "epoch": 0.44749064345943995, + "grad_norm": 1.65983819046533, + "learning_rate": 2.433806018143339e-06, + "loss": 0.9854, + "step": 4962 + }, + { + "epoch": 0.44758082698291024, + "grad_norm": 0.7478102267366665, + "learning_rate": 2.433235697225051e-06, + "loss": 0.8176, + "step": 4963 + }, + { + "epoch": 0.4476710105063805, + "grad_norm": 1.631884662624404, + "learning_rate": 2.4326653393407048e-06, + "loss": 0.9718, + "step": 4964 + }, + { + "epoch": 0.44776119402985076, + "grad_norm": 1.4089609506392753, + "learning_rate": 2.432094944538966e-06, + "loss": 0.9232, + "step": 4965 + }, + { + "epoch": 0.447851377553321, + "grad_norm": 1.3476141881598382, + "learning_rate": 2.4315245128685047e-06, + "loss": 0.9222, + "step": 4966 + }, + { + "epoch": 0.4479415610767913, + "grad_norm": 1.762278939310268, + "learning_rate": 2.4309540443779925e-06, + "loss": 0.9584, + "step": 4967 + }, + { + "epoch": 0.4480317446002615, + "grad_norm": 1.5125167278279035, + "learning_rate": 2.4303835391161047e-06, + "loss": 0.8909, + "step": 4968 + }, + { + "epoch": 0.4481219281237318, + "grad_norm": 1.4279787451981458, + "learning_rate": 2.42981299713152e-06, + "loss": 0.9031, + "step": 4969 + }, + { + "epoch": 0.44821211164720204, + "grad_norm": 1.5759116788795324, + "learning_rate": 2.4292424184729204e-06, + "loss": 0.9949, + "step": 4970 + }, + { + "epoch": 0.4483022951706723, + "grad_norm": 1.4454280497781864, + "learning_rate": 2.4286718031889913e-06, + "loss": 0.9504, + "step": 4971 + }, + { + "epoch": 0.44839247869414256, + "grad_norm": 1.346180779847885, + "learning_rate": 2.4281011513284202e-06, + "loss": 0.9314, + "step": 4972 + }, + { + "epoch": 0.44848266221761285, + "grad_norm": 4.470439031490906, + "learning_rate": 2.4275304629398985e-06, + "loss": 0.8919, + "step": 4973 + }, + { + "epoch": 0.4485728457410831, + "grad_norm": 1.437607139243611, + "learning_rate": 2.4269597380721194e-06, + "loss": 1.0296, + "step": 4974 + }, + { + "epoch": 0.44866302926455337, + "grad_norm": 1.4271377590224215, + "learning_rate": 2.426388976773782e-06, + "loss": 0.9493, + "step": 4975 + }, + { + "epoch": 0.4487532127880236, + "grad_norm": 5.716985378148596, + "learning_rate": 2.425818179093586e-06, + "loss": 1.0593, + "step": 4976 + }, + { + "epoch": 0.4488433963114939, + "grad_norm": 1.4008872456400436, + "learning_rate": 2.4252473450802346e-06, + "loss": 0.8643, + "step": 4977 + }, + { + "epoch": 0.4489335798349641, + "grad_norm": 1.5315528881469567, + "learning_rate": 2.4246764747824355e-06, + "loss": 0.9862, + "step": 4978 + }, + { + "epoch": 0.4490237633584344, + "grad_norm": 1.2831158426871285, + "learning_rate": 2.424105568248897e-06, + "loss": 0.9675, + "step": 4979 + }, + { + "epoch": 0.4491139468819047, + "grad_norm": 1.56925777105328, + "learning_rate": 2.4235346255283337e-06, + "loss": 0.8949, + "step": 4980 + }, + { + "epoch": 0.44920413040537494, + "grad_norm": 1.567432552510998, + "learning_rate": 2.42296364666946e-06, + "loss": 0.9667, + "step": 4981 + }, + { + "epoch": 0.4492943139288452, + "grad_norm": 1.6194526178507425, + "learning_rate": 2.4223926317209965e-06, + "loss": 0.9018, + "step": 4982 + }, + { + "epoch": 0.44938449745231546, + "grad_norm": 1.392645236373762, + "learning_rate": 2.4218215807316647e-06, + "loss": 0.9581, + "step": 4983 + }, + { + "epoch": 0.44947468097578575, + "grad_norm": 1.6642498931678438, + "learning_rate": 2.4212504937501894e-06, + "loss": 1.0436, + "step": 4984 + }, + { + "epoch": 0.449564864499256, + "grad_norm": 1.4258427854871591, + "learning_rate": 2.4206793708253e-06, + "loss": 1.0124, + "step": 4985 + }, + { + "epoch": 0.44965504802272627, + "grad_norm": 2.1706542831453506, + "learning_rate": 2.420108212005726e-06, + "loss": 0.9077, + "step": 4986 + }, + { + "epoch": 0.4497452315461965, + "grad_norm": 1.8355349484916923, + "learning_rate": 2.4195370173402034e-06, + "loss": 0.9551, + "step": 4987 + }, + { + "epoch": 0.4498354150696668, + "grad_norm": 1.6828270248310457, + "learning_rate": 2.4189657868774696e-06, + "loss": 1.0792, + "step": 4988 + }, + { + "epoch": 0.449925598593137, + "grad_norm": 1.803731557813914, + "learning_rate": 2.418394520666264e-06, + "loss": 0.8927, + "step": 4989 + }, + { + "epoch": 0.4500157821166073, + "grad_norm": 1.5108166745938956, + "learning_rate": 2.4178232187553307e-06, + "loss": 0.9417, + "step": 4990 + }, + { + "epoch": 0.45010596564007754, + "grad_norm": 1.3605172630765376, + "learning_rate": 2.417251881193417e-06, + "loss": 0.984, + "step": 4991 + }, + { + "epoch": 0.45019614916354783, + "grad_norm": 1.6104181112831548, + "learning_rate": 2.4166805080292723e-06, + "loss": 1.0142, + "step": 4992 + }, + { + "epoch": 0.45028633268701806, + "grad_norm": 1.5351840494759514, + "learning_rate": 2.4161090993116485e-06, + "loss": 0.9057, + "step": 4993 + }, + { + "epoch": 0.45037651621048835, + "grad_norm": 2.6587662391184606, + "learning_rate": 2.4155376550893026e-06, + "loss": 0.8395, + "step": 4994 + }, + { + "epoch": 0.4504666997339586, + "grad_norm": 1.6415062127229616, + "learning_rate": 2.4149661754109926e-06, + "loss": 0.9678, + "step": 4995 + }, + { + "epoch": 0.4505568832574289, + "grad_norm": 1.7159522434116026, + "learning_rate": 2.41439466032548e-06, + "loss": 1.0144, + "step": 4996 + }, + { + "epoch": 0.4506470667808991, + "grad_norm": 1.5554790686164994, + "learning_rate": 2.41382310988153e-06, + "loss": 0.9043, + "step": 4997 + }, + { + "epoch": 0.4507372503043694, + "grad_norm": 1.6253748974798978, + "learning_rate": 2.413251524127911e-06, + "loss": 1.0188, + "step": 4998 + }, + { + "epoch": 0.45082743382783963, + "grad_norm": 1.5275044466444059, + "learning_rate": 2.412679903113393e-06, + "loss": 0.9177, + "step": 4999 + }, + { + "epoch": 0.4509176173513099, + "grad_norm": 1.806583284035129, + "learning_rate": 2.4121082468867505e-06, + "loss": 0.9677, + "step": 5000 + }, + { + "epoch": 0.45100780087478015, + "grad_norm": 1.7665834104365588, + "learning_rate": 2.4115365554967597e-06, + "loss": 0.8962, + "step": 5001 + }, + { + "epoch": 0.45109798439825044, + "grad_norm": 1.506088852868982, + "learning_rate": 2.4109648289922006e-06, + "loss": 0.9569, + "step": 5002 + }, + { + "epoch": 0.45118816792172073, + "grad_norm": 1.6285526614553865, + "learning_rate": 2.4103930674218565e-06, + "loss": 0.9648, + "step": 5003 + }, + { + "epoch": 0.45127835144519096, + "grad_norm": 1.5537923411510162, + "learning_rate": 2.409821270834513e-06, + "loss": 0.9335, + "step": 5004 + }, + { + "epoch": 0.45136853496866125, + "grad_norm": 1.8712678004089014, + "learning_rate": 2.409249439278959e-06, + "loss": 1.0144, + "step": 5005 + }, + { + "epoch": 0.4514587184921315, + "grad_norm": 0.6748047537465484, + "learning_rate": 2.408677572803986e-06, + "loss": 0.7534, + "step": 5006 + }, + { + "epoch": 0.45154890201560177, + "grad_norm": 1.6381770445100083, + "learning_rate": 2.408105671458389e-06, + "loss": 0.9631, + "step": 5007 + }, + { + "epoch": 0.451639085539072, + "grad_norm": 1.3314170875820044, + "learning_rate": 2.4075337352909663e-06, + "loss": 0.9004, + "step": 5008 + }, + { + "epoch": 0.4517292690625423, + "grad_norm": 1.4289366204239426, + "learning_rate": 2.4069617643505177e-06, + "loss": 1.0854, + "step": 5009 + }, + { + "epoch": 0.4518194525860125, + "grad_norm": 1.5604965335522658, + "learning_rate": 2.406389758685848e-06, + "loss": 1.0453, + "step": 5010 + }, + { + "epoch": 0.4519096361094828, + "grad_norm": 1.3759309044999788, + "learning_rate": 2.405817718345763e-06, + "loss": 0.9963, + "step": 5011 + }, + { + "epoch": 0.45199981963295305, + "grad_norm": 1.4325750821926873, + "learning_rate": 2.4052456433790726e-06, + "loss": 0.9099, + "step": 5012 + }, + { + "epoch": 0.45209000315642334, + "grad_norm": 1.546655889849004, + "learning_rate": 2.4046735338345897e-06, + "loss": 1.0104, + "step": 5013 + }, + { + "epoch": 0.45218018667989357, + "grad_norm": 1.6347361645157557, + "learning_rate": 2.404101389761129e-06, + "loss": 0.9726, + "step": 5014 + }, + { + "epoch": 0.45227037020336386, + "grad_norm": 0.7065322337359904, + "learning_rate": 2.4035292112075097e-06, + "loss": 0.7651, + "step": 5015 + }, + { + "epoch": 0.4523605537268341, + "grad_norm": 1.4507696279656996, + "learning_rate": 2.4029569982225534e-06, + "loss": 1.0185, + "step": 5016 + }, + { + "epoch": 0.4524507372503044, + "grad_norm": 0.7674748785648022, + "learning_rate": 2.402384750855084e-06, + "loss": 0.8379, + "step": 5017 + }, + { + "epoch": 0.4525409207737746, + "grad_norm": 1.638671875, + "learning_rate": 2.4018124691539286e-06, + "loss": 0.9586, + "step": 5018 + }, + { + "epoch": 0.4526311042972449, + "grad_norm": 2.4279940183210766, + "learning_rate": 2.4012401531679178e-06, + "loss": 0.9502, + "step": 5019 + }, + { + "epoch": 0.45272128782071513, + "grad_norm": 1.7877545635648981, + "learning_rate": 2.4006678029458847e-06, + "loss": 0.9569, + "step": 5020 + }, + { + "epoch": 0.4528114713441854, + "grad_norm": 1.4714584538683284, + "learning_rate": 2.400095418536666e-06, + "loss": 0.866, + "step": 5021 + }, + { + "epoch": 0.45290165486765566, + "grad_norm": 1.599588135518905, + "learning_rate": 2.3995229999890996e-06, + "loss": 0.9374, + "step": 5022 + }, + { + "epoch": 0.45299183839112594, + "grad_norm": 1.6874580378083521, + "learning_rate": 2.398950547352028e-06, + "loss": 0.9022, + "step": 5023 + }, + { + "epoch": 0.4530820219145962, + "grad_norm": 2.229929416206582, + "learning_rate": 2.398378060674295e-06, + "loss": 0.8742, + "step": 5024 + }, + { + "epoch": 0.45317220543806647, + "grad_norm": 1.4022792705705287, + "learning_rate": 2.39780554000475e-06, + "loss": 0.9593, + "step": 5025 + }, + { + "epoch": 0.4532623889615367, + "grad_norm": 1.5356349053255542, + "learning_rate": 2.3972329853922434e-06, + "loss": 0.9892, + "step": 5026 + }, + { + "epoch": 0.453352572485007, + "grad_norm": 0.6730942198605914, + "learning_rate": 2.3966603968856278e-06, + "loss": 0.7719, + "step": 5027 + }, + { + "epoch": 0.4534427560084773, + "grad_norm": 1.4787744172144908, + "learning_rate": 2.39608777453376e-06, + "loss": 1.0304, + "step": 5028 + }, + { + "epoch": 0.4535329395319475, + "grad_norm": 2.095305477076516, + "learning_rate": 2.3955151183854993e-06, + "loss": 0.9822, + "step": 5029 + }, + { + "epoch": 0.4536231230554178, + "grad_norm": 1.6194601997509142, + "learning_rate": 2.3949424284897073e-06, + "loss": 0.922, + "step": 5030 + }, + { + "epoch": 0.45371330657888803, + "grad_norm": 1.4316648556687104, + "learning_rate": 2.39436970489525e-06, + "loss": 0.8257, + "step": 5031 + }, + { + "epoch": 0.4538034901023583, + "grad_norm": 1.6438278905015264, + "learning_rate": 2.3937969476509955e-06, + "loss": 0.9631, + "step": 5032 + }, + { + "epoch": 0.45389367362582855, + "grad_norm": 1.6072932733052652, + "learning_rate": 2.393224156805813e-06, + "loss": 1.0336, + "step": 5033 + }, + { + "epoch": 0.45398385714929884, + "grad_norm": 1.6818555457536257, + "learning_rate": 2.392651332408578e-06, + "loss": 0.9358, + "step": 5034 + }, + { + "epoch": 0.4540740406727691, + "grad_norm": 1.9519860570792662, + "learning_rate": 2.3920784745081655e-06, + "loss": 0.9396, + "step": 5035 + }, + { + "epoch": 0.45416422419623936, + "grad_norm": 1.5009027784351427, + "learning_rate": 2.391505583153456e-06, + "loss": 0.9879, + "step": 5036 + }, + { + "epoch": 0.4542544077197096, + "grad_norm": 1.5267485022093237, + "learning_rate": 2.3909326583933315e-06, + "loss": 0.965, + "step": 5037 + }, + { + "epoch": 0.4543445912431799, + "grad_norm": 1.5119647631702133, + "learning_rate": 2.3903597002766777e-06, + "loss": 0.9283, + "step": 5038 + }, + { + "epoch": 0.4544347747666501, + "grad_norm": 1.4561262413777376, + "learning_rate": 2.389786708852381e-06, + "loss": 0.9249, + "step": 5039 + }, + { + "epoch": 0.4545249582901204, + "grad_norm": 0.9951887020809841, + "learning_rate": 2.389213684169333e-06, + "loss": 0.8957, + "step": 5040 + }, + { + "epoch": 0.45461514181359064, + "grad_norm": 1.4575728568029778, + "learning_rate": 2.388640626276428e-06, + "loss": 0.9186, + "step": 5041 + }, + { + "epoch": 0.45470532533706093, + "grad_norm": 0.7336106482236359, + "learning_rate": 2.388067535222561e-06, + "loss": 0.7372, + "step": 5042 + }, + { + "epoch": 0.45479550886053116, + "grad_norm": 1.7710817125447307, + "learning_rate": 2.3874944110566332e-06, + "loss": 0.9726, + "step": 5043 + }, + { + "epoch": 0.45488569238400145, + "grad_norm": 0.8436809087651678, + "learning_rate": 2.3869212538275447e-06, + "loss": 0.8789, + "step": 5044 + }, + { + "epoch": 0.4549758759074717, + "grad_norm": 1.6236925000118485, + "learning_rate": 2.386348063584202e-06, + "loss": 0.9754, + "step": 5045 + }, + { + "epoch": 0.45506605943094197, + "grad_norm": 1.918933260736646, + "learning_rate": 2.385774840375511e-06, + "loss": 1.0073, + "step": 5046 + }, + { + "epoch": 0.4551562429544122, + "grad_norm": 1.4753830654767697, + "learning_rate": 2.385201584250385e-06, + "loss": 0.9099, + "step": 5047 + }, + { + "epoch": 0.4552464264778825, + "grad_norm": 0.8114683130227628, + "learning_rate": 2.3846282952577346e-06, + "loss": 0.814, + "step": 5048 + }, + { + "epoch": 0.4553366100013527, + "grad_norm": 1.6070871479733233, + "learning_rate": 2.3840549734464785e-06, + "loss": 1.0486, + "step": 5049 + }, + { + "epoch": 0.455426793524823, + "grad_norm": 1.6074465706876095, + "learning_rate": 2.3834816188655336e-06, + "loss": 0.8955, + "step": 5050 + }, + { + "epoch": 0.4555169770482933, + "grad_norm": 1.4912283645940594, + "learning_rate": 2.3829082315638224e-06, + "loss": 0.9099, + "step": 5051 + }, + { + "epoch": 0.45560716057176354, + "grad_norm": 1.6600452341098333, + "learning_rate": 2.3823348115902695e-06, + "loss": 0.8171, + "step": 5052 + }, + { + "epoch": 0.4556973440952338, + "grad_norm": 2.6156630901812776, + "learning_rate": 2.3817613589938026e-06, + "loss": 0.977, + "step": 5053 + }, + { + "epoch": 0.45578752761870406, + "grad_norm": 1.5054826673449366, + "learning_rate": 2.3811878738233517e-06, + "loss": 0.9944, + "step": 5054 + }, + { + "epoch": 0.45587771114217435, + "grad_norm": 1.8471877812717865, + "learning_rate": 2.380614356127849e-06, + "loss": 0.9028, + "step": 5055 + }, + { + "epoch": 0.4559678946656446, + "grad_norm": 2.6180682621879963, + "learning_rate": 2.3800408059562318e-06, + "loss": 0.9567, + "step": 5056 + }, + { + "epoch": 0.45605807818911487, + "grad_norm": 1.5685976448394514, + "learning_rate": 2.3794672233574365e-06, + "loss": 0.9745, + "step": 5057 + }, + { + "epoch": 0.4561482617125851, + "grad_norm": 1.5632462817412776, + "learning_rate": 2.3788936083804058e-06, + "loss": 0.9696, + "step": 5058 + }, + { + "epoch": 0.4562384452360554, + "grad_norm": 1.4808606758806275, + "learning_rate": 2.378319961074083e-06, + "loss": 0.9596, + "step": 5059 + }, + { + "epoch": 0.4563286287595256, + "grad_norm": 1.6053339026778555, + "learning_rate": 2.377746281487415e-06, + "loss": 0.9989, + "step": 5060 + }, + { + "epoch": 0.4564188122829959, + "grad_norm": 1.4745963594986444, + "learning_rate": 2.377172569669352e-06, + "loss": 1.0473, + "step": 5061 + }, + { + "epoch": 0.45650899580646614, + "grad_norm": 1.5810854498242644, + "learning_rate": 2.376598825668845e-06, + "loss": 0.8128, + "step": 5062 + }, + { + "epoch": 0.45659917932993643, + "grad_norm": 1.623833898149092, + "learning_rate": 2.3760250495348495e-06, + "loss": 0.9473, + "step": 5063 + }, + { + "epoch": 0.45668936285340667, + "grad_norm": 1.5760317223691724, + "learning_rate": 2.3754512413163236e-06, + "loss": 1.0033, + "step": 5064 + }, + { + "epoch": 0.45677954637687695, + "grad_norm": 1.3909525967603376, + "learning_rate": 2.3748774010622285e-06, + "loss": 0.9383, + "step": 5065 + }, + { + "epoch": 0.4568697299003472, + "grad_norm": 1.3841146168270688, + "learning_rate": 2.3743035288215254e-06, + "loss": 0.9655, + "step": 5066 + }, + { + "epoch": 0.4569599134238175, + "grad_norm": 1.4374154936788095, + "learning_rate": 2.3737296246431815e-06, + "loss": 0.9651, + "step": 5067 + }, + { + "epoch": 0.4570500969472877, + "grad_norm": 1.3802201708399973, + "learning_rate": 2.3731556885761656e-06, + "loss": 0.9286, + "step": 5068 + }, + { + "epoch": 0.457140280470758, + "grad_norm": 0.764638537736242, + "learning_rate": 2.372581720669449e-06, + "loss": 0.7815, + "step": 5069 + }, + { + "epoch": 0.45723046399422823, + "grad_norm": 1.5448119923786439, + "learning_rate": 2.3720077209720046e-06, + "loss": 0.9853, + "step": 5070 + }, + { + "epoch": 0.4573206475176985, + "grad_norm": 0.7765531113362125, + "learning_rate": 2.3714336895328112e-06, + "loss": 0.7743, + "step": 5071 + }, + { + "epoch": 0.45741083104116875, + "grad_norm": 1.6063673191790808, + "learning_rate": 2.370859626400847e-06, + "loss": 0.9122, + "step": 5072 + }, + { + "epoch": 0.45750101456463904, + "grad_norm": 1.3989405659531777, + "learning_rate": 2.3702855316250943e-06, + "loss": 0.9354, + "step": 5073 + }, + { + "epoch": 0.45759119808810933, + "grad_norm": 1.6368078967742008, + "learning_rate": 2.369711405254539e-06, + "loss": 0.9643, + "step": 5074 + }, + { + "epoch": 0.45768138161157956, + "grad_norm": 1.6673386252353726, + "learning_rate": 2.3691372473381673e-06, + "loss": 1.0134, + "step": 5075 + }, + { + "epoch": 0.45777156513504985, + "grad_norm": 1.5189598597835114, + "learning_rate": 2.3685630579249708e-06, + "loss": 1.0195, + "step": 5076 + }, + { + "epoch": 0.4578617486585201, + "grad_norm": 1.5592952669663755, + "learning_rate": 2.367988837063942e-06, + "loss": 0.9328, + "step": 5077 + }, + { + "epoch": 0.4579519321819904, + "grad_norm": 1.462024794100864, + "learning_rate": 2.367414584804076e-06, + "loss": 1.031, + "step": 5078 + }, + { + "epoch": 0.4580421157054606, + "grad_norm": 1.5485517873458547, + "learning_rate": 2.366840301194372e-06, + "loss": 0.995, + "step": 5079 + }, + { + "epoch": 0.4581322992289309, + "grad_norm": 1.6045609852395089, + "learning_rate": 2.3662659862838308e-06, + "loss": 0.9443, + "step": 5080 + }, + { + "epoch": 0.45822248275240113, + "grad_norm": 1.6481703410676485, + "learning_rate": 2.365691640121456e-06, + "loss": 1.0108, + "step": 5081 + }, + { + "epoch": 0.4583126662758714, + "grad_norm": 1.6109674223406554, + "learning_rate": 2.365117262756254e-06, + "loss": 0.888, + "step": 5082 + }, + { + "epoch": 0.45840284979934165, + "grad_norm": 1.7713287184189384, + "learning_rate": 2.3645428542372342e-06, + "loss": 0.9526, + "step": 5083 + }, + { + "epoch": 0.45849303332281194, + "grad_norm": 0.7147101605771407, + "learning_rate": 2.3639684146134083e-06, + "loss": 0.733, + "step": 5084 + }, + { + "epoch": 0.45858321684628217, + "grad_norm": 1.679595093070428, + "learning_rate": 2.3633939439337897e-06, + "loss": 0.9216, + "step": 5085 + }, + { + "epoch": 0.45867340036975246, + "grad_norm": 2.850464324604805, + "learning_rate": 2.362819442247396e-06, + "loss": 0.9326, + "step": 5086 + }, + { + "epoch": 0.4587635838932227, + "grad_norm": 0.7646316779822413, + "learning_rate": 2.3622449096032477e-06, + "loss": 0.7431, + "step": 5087 + }, + { + "epoch": 0.458853767416693, + "grad_norm": 1.637266663677827, + "learning_rate": 2.361670346050366e-06, + "loss": 0.9937, + "step": 5088 + }, + { + "epoch": 0.4589439509401632, + "grad_norm": 1.637831789720876, + "learning_rate": 2.3610957516377757e-06, + "loss": 0.8918, + "step": 5089 + }, + { + "epoch": 0.4590341344636335, + "grad_norm": 1.5116153652117856, + "learning_rate": 2.3605211264145048e-06, + "loss": 0.9634, + "step": 5090 + }, + { + "epoch": 0.45912431798710374, + "grad_norm": 1.288280134044416, + "learning_rate": 2.3599464704295836e-06, + "loss": 0.9196, + "step": 5091 + }, + { + "epoch": 0.459214501510574, + "grad_norm": 1.6171496842980153, + "learning_rate": 2.359371783732045e-06, + "loss": 0.8527, + "step": 5092 + }, + { + "epoch": 0.45930468503404426, + "grad_norm": 1.598755623790403, + "learning_rate": 2.358797066370924e-06, + "loss": 0.9387, + "step": 5093 + }, + { + "epoch": 0.45939486855751455, + "grad_norm": 1.5931206750941993, + "learning_rate": 2.3582223183952594e-06, + "loss": 0.9551, + "step": 5094 + }, + { + "epoch": 0.4594850520809848, + "grad_norm": 1.566103480275616, + "learning_rate": 2.357647539854091e-06, + "loss": 0.9916, + "step": 5095 + }, + { + "epoch": 0.45957523560445507, + "grad_norm": 1.632666841425691, + "learning_rate": 2.3570727307964624e-06, + "loss": 0.964, + "step": 5096 + }, + { + "epoch": 0.4596654191279253, + "grad_norm": 1.547489159935868, + "learning_rate": 2.35649789127142e-06, + "loss": 1.0084, + "step": 5097 + }, + { + "epoch": 0.4597556026513956, + "grad_norm": 1.6601374367040633, + "learning_rate": 2.3559230213280115e-06, + "loss": 0.9201, + "step": 5098 + }, + { + "epoch": 0.4598457861748659, + "grad_norm": 1.37388491932104, + "learning_rate": 2.3553481210152886e-06, + "loss": 0.8723, + "step": 5099 + }, + { + "epoch": 0.4599359696983361, + "grad_norm": 0.9144078441882487, + "learning_rate": 2.3547731903823043e-06, + "loss": 0.7982, + "step": 5100 + }, + { + "epoch": 0.4600261532218064, + "grad_norm": 1.6941874062462867, + "learning_rate": 2.3541982294781155e-06, + "loss": 0.9367, + "step": 5101 + }, + { + "epoch": 0.46011633674527663, + "grad_norm": 1.384644496338625, + "learning_rate": 2.3536232383517804e-06, + "loss": 0.9543, + "step": 5102 + }, + { + "epoch": 0.4602065202687469, + "grad_norm": 1.3518863797072596, + "learning_rate": 2.3530482170523602e-06, + "loss": 0.9518, + "step": 5103 + }, + { + "epoch": 0.46029670379221715, + "grad_norm": 1.4396827545650863, + "learning_rate": 2.3524731656289206e-06, + "loss": 1.0508, + "step": 5104 + }, + { + "epoch": 0.46038688731568744, + "grad_norm": 1.5797232664381444, + "learning_rate": 2.351898084130526e-06, + "loss": 1.0464, + "step": 5105 + }, + { + "epoch": 0.4604770708391577, + "grad_norm": 1.5488076515335836, + "learning_rate": 2.351322972606247e-06, + "loss": 1.0232, + "step": 5106 + }, + { + "epoch": 0.46056725436262796, + "grad_norm": 1.5746996123856247, + "learning_rate": 2.350747831105155e-06, + "loss": 1.1066, + "step": 5107 + }, + { + "epoch": 0.4606574378860982, + "grad_norm": 2.079788001517487, + "learning_rate": 2.350172659676323e-06, + "loss": 1.0358, + "step": 5108 + }, + { + "epoch": 0.4607476214095685, + "grad_norm": 1.6517949196106083, + "learning_rate": 2.3495974583688306e-06, + "loss": 0.9955, + "step": 5109 + }, + { + "epoch": 0.4608378049330387, + "grad_norm": 1.6009264171305333, + "learning_rate": 2.3490222272317543e-06, + "loss": 1.0281, + "step": 5110 + }, + { + "epoch": 0.460927988456509, + "grad_norm": 1.8005454349648615, + "learning_rate": 2.348446966314177e-06, + "loss": 1.0079, + "step": 5111 + }, + { + "epoch": 0.46101817197997924, + "grad_norm": 1.6263456641640155, + "learning_rate": 2.3478716756651837e-06, + "loss": 0.9302, + "step": 5112 + }, + { + "epoch": 0.46110835550344953, + "grad_norm": 1.7006274467879656, + "learning_rate": 2.347296355333861e-06, + "loss": 0.9514, + "step": 5113 + }, + { + "epoch": 0.46119853902691976, + "grad_norm": 1.759939170349731, + "learning_rate": 2.3467210053692972e-06, + "loss": 0.9638, + "step": 5114 + }, + { + "epoch": 0.46128872255039005, + "grad_norm": 1.7198785717927985, + "learning_rate": 2.3461456258205866e-06, + "loss": 0.9646, + "step": 5115 + }, + { + "epoch": 0.4613789060738603, + "grad_norm": 1.5740492660877972, + "learning_rate": 2.345570216736822e-06, + "loss": 0.9011, + "step": 5116 + }, + { + "epoch": 0.4614690895973306, + "grad_norm": 1.7646688387292615, + "learning_rate": 2.3449947781671013e-06, + "loss": 0.9709, + "step": 5117 + }, + { + "epoch": 0.4615592731208008, + "grad_norm": 1.5367555320072965, + "learning_rate": 2.3444193101605237e-06, + "loss": 0.9883, + "step": 5118 + }, + { + "epoch": 0.4616494566442711, + "grad_norm": 1.542209718615671, + "learning_rate": 2.3438438127661913e-06, + "loss": 0.979, + "step": 5119 + }, + { + "epoch": 0.4617396401677413, + "grad_norm": 1.7416098282709747, + "learning_rate": 2.3432682860332096e-06, + "loss": 0.791, + "step": 5120 + }, + { + "epoch": 0.4618298236912116, + "grad_norm": 1.5289542446488105, + "learning_rate": 2.342692730010684e-06, + "loss": 0.9096, + "step": 5121 + }, + { + "epoch": 0.4619200072146819, + "grad_norm": 1.9620021650280999, + "learning_rate": 2.342117144747726e-06, + "loss": 0.97, + "step": 5122 + }, + { + "epoch": 0.46201019073815214, + "grad_norm": 1.5248057945289468, + "learning_rate": 2.3415415302934457e-06, + "loss": 0.96, + "step": 5123 + }, + { + "epoch": 0.4621003742616224, + "grad_norm": 2.318197500206438, + "learning_rate": 2.340965886696959e-06, + "loss": 0.9916, + "step": 5124 + }, + { + "epoch": 0.46219055778509266, + "grad_norm": 1.5673742181179873, + "learning_rate": 2.340390214007384e-06, + "loss": 0.9103, + "step": 5125 + }, + { + "epoch": 0.46228074130856295, + "grad_norm": 1.4818193621027045, + "learning_rate": 2.339814512273838e-06, + "loss": 0.8427, + "step": 5126 + }, + { + "epoch": 0.4623709248320332, + "grad_norm": 1.5224304000405173, + "learning_rate": 2.3392387815454447e-06, + "loss": 1.0616, + "step": 5127 + }, + { + "epoch": 0.46246110835550347, + "grad_norm": 1.571065390097095, + "learning_rate": 2.3386630218713273e-06, + "loss": 0.9516, + "step": 5128 + }, + { + "epoch": 0.4625512918789737, + "grad_norm": 1.584653588355092, + "learning_rate": 2.3380872333006135e-06, + "loss": 0.9858, + "step": 5129 + }, + { + "epoch": 0.462641475402444, + "grad_norm": 1.6205216834651768, + "learning_rate": 2.3375114158824335e-06, + "loss": 1.0028, + "step": 5130 + }, + { + "epoch": 0.4627316589259142, + "grad_norm": 1.4121368954246885, + "learning_rate": 2.3369355696659184e-06, + "loss": 0.9953, + "step": 5131 + }, + { + "epoch": 0.4628218424493845, + "grad_norm": 1.6169413503651362, + "learning_rate": 2.336359694700202e-06, + "loss": 0.9479, + "step": 5132 + }, + { + "epoch": 0.46291202597285475, + "grad_norm": 1.6435776797866286, + "learning_rate": 2.335783791034422e-06, + "loss": 0.962, + "step": 5133 + }, + { + "epoch": 0.46300220949632503, + "grad_norm": 1.5946161497055438, + "learning_rate": 2.3352078587177173e-06, + "loss": 0.9393, + "step": 5134 + }, + { + "epoch": 0.46309239301979527, + "grad_norm": 1.5167838316854443, + "learning_rate": 2.33463189779923e-06, + "loss": 0.9663, + "step": 5135 + }, + { + "epoch": 0.46318257654326556, + "grad_norm": 1.6388913395247304, + "learning_rate": 2.334055908328104e-06, + "loss": 1.0362, + "step": 5136 + }, + { + "epoch": 0.4632727600667358, + "grad_norm": 1.3783902075279642, + "learning_rate": 2.3334798903534866e-06, + "loss": 0.8728, + "step": 5137 + }, + { + "epoch": 0.4633629435902061, + "grad_norm": 1.3864240736231075, + "learning_rate": 2.3329038439245252e-06, + "loss": 0.8562, + "step": 5138 + }, + { + "epoch": 0.4634531271136763, + "grad_norm": 1.7949953282117483, + "learning_rate": 2.3323277690903724e-06, + "loss": 0.9581, + "step": 5139 + }, + { + "epoch": 0.4635433106371466, + "grad_norm": 1.5215899096652785, + "learning_rate": 2.3317516659001827e-06, + "loss": 0.9322, + "step": 5140 + }, + { + "epoch": 0.46363349416061683, + "grad_norm": 1.3953385353863217, + "learning_rate": 2.331175534403111e-06, + "loss": 0.9725, + "step": 5141 + }, + { + "epoch": 0.4637236776840871, + "grad_norm": 1.4424408298567881, + "learning_rate": 2.3305993746483167e-06, + "loss": 1.0474, + "step": 5142 + }, + { + "epoch": 0.46381386120755735, + "grad_norm": 1.8593460208974677, + "learning_rate": 2.3300231866849606e-06, + "loss": 0.9391, + "step": 5143 + }, + { + "epoch": 0.46390404473102764, + "grad_norm": 1.5989767647176916, + "learning_rate": 2.3294469705622067e-06, + "loss": 0.9999, + "step": 5144 + }, + { + "epoch": 0.4639942282544979, + "grad_norm": 1.6871949379296511, + "learning_rate": 2.3288707263292203e-06, + "loss": 0.9342, + "step": 5145 + }, + { + "epoch": 0.46408441177796816, + "grad_norm": 1.3611197401063007, + "learning_rate": 2.3282944540351707e-06, + "loss": 0.9786, + "step": 5146 + }, + { + "epoch": 0.46417459530143845, + "grad_norm": 1.7978790131917775, + "learning_rate": 2.327718153729228e-06, + "loss": 0.933, + "step": 5147 + }, + { + "epoch": 0.4642647788249087, + "grad_norm": 2.234591920366586, + "learning_rate": 2.327141825460566e-06, + "loss": 1.0015, + "step": 5148 + }, + { + "epoch": 0.464354962348379, + "grad_norm": 0.7194555799889949, + "learning_rate": 2.326565469278358e-06, + "loss": 0.7865, + "step": 5149 + }, + { + "epoch": 0.4644451458718492, + "grad_norm": 0.7964099200042353, + "learning_rate": 2.3259890852317846e-06, + "loss": 0.7953, + "step": 5150 + }, + { + "epoch": 0.4645353293953195, + "grad_norm": 1.7734214345061483, + "learning_rate": 2.3254126733700246e-06, + "loss": 0.9584, + "step": 5151 + }, + { + "epoch": 0.46462551291878973, + "grad_norm": 1.5662628640927534, + "learning_rate": 2.324836233742262e-06, + "loss": 0.9875, + "step": 5152 + }, + { + "epoch": 0.46471569644226, + "grad_norm": 2.0443120154507026, + "learning_rate": 2.3242597663976793e-06, + "loss": 0.9947, + "step": 5153 + }, + { + "epoch": 0.46480587996573025, + "grad_norm": 1.5729395210003636, + "learning_rate": 2.3236832713854663e-06, + "loss": 0.9379, + "step": 5154 + }, + { + "epoch": 0.46489606348920054, + "grad_norm": 1.5925771848457337, + "learning_rate": 2.323106748754812e-06, + "loss": 0.956, + "step": 5155 + }, + { + "epoch": 0.4649862470126708, + "grad_norm": 1.4955436950340513, + "learning_rate": 2.3225301985549077e-06, + "loss": 1.051, + "step": 5156 + }, + { + "epoch": 0.46507643053614106, + "grad_norm": 1.5771976427840888, + "learning_rate": 2.321953620834948e-06, + "loss": 0.8912, + "step": 5157 + }, + { + "epoch": 0.4651666140596113, + "grad_norm": 1.4681672096599316, + "learning_rate": 2.3213770156441314e-06, + "loss": 0.9271, + "step": 5158 + }, + { + "epoch": 0.4652567975830816, + "grad_norm": 2.007252655943561, + "learning_rate": 2.3208003830316554e-06, + "loss": 1.0067, + "step": 5159 + }, + { + "epoch": 0.4653469811065518, + "grad_norm": 1.6503477192555223, + "learning_rate": 2.3202237230467215e-06, + "loss": 1.0498, + "step": 5160 + }, + { + "epoch": 0.4654371646300221, + "grad_norm": 1.480842482768087, + "learning_rate": 2.3196470357385338e-06, + "loss": 0.9192, + "step": 5161 + }, + { + "epoch": 0.46552734815349234, + "grad_norm": 1.5493415664567904, + "learning_rate": 2.319070321156299e-06, + "loss": 0.9531, + "step": 5162 + }, + { + "epoch": 0.4656175316769626, + "grad_norm": 1.4140686962350217, + "learning_rate": 2.318493579349224e-06, + "loss": 1.0918, + "step": 5163 + }, + { + "epoch": 0.46570771520043286, + "grad_norm": 1.720775451179307, + "learning_rate": 2.317916810366522e-06, + "loss": 0.9435, + "step": 5164 + }, + { + "epoch": 0.46579789872390315, + "grad_norm": 1.5319594471945106, + "learning_rate": 2.317340014257404e-06, + "loss": 0.902, + "step": 5165 + }, + { + "epoch": 0.4658880822473734, + "grad_norm": 1.420600571495778, + "learning_rate": 2.316763191071086e-06, + "loss": 0.8867, + "step": 5166 + }, + { + "epoch": 0.46597826577084367, + "grad_norm": 1.4192180752062262, + "learning_rate": 2.316186340856787e-06, + "loss": 0.9199, + "step": 5167 + }, + { + "epoch": 0.4660684492943139, + "grad_norm": 4.4762239710975935, + "learning_rate": 2.315609463663725e-06, + "loss": 0.9072, + "step": 5168 + }, + { + "epoch": 0.4661586328177842, + "grad_norm": 1.8980767806092278, + "learning_rate": 2.315032559541123e-06, + "loss": 0.9302, + "step": 5169 + }, + { + "epoch": 0.4662488163412545, + "grad_norm": 1.6558637348539706, + "learning_rate": 2.314455628538207e-06, + "loss": 0.9792, + "step": 5170 + }, + { + "epoch": 0.4663389998647247, + "grad_norm": 1.470465044303313, + "learning_rate": 2.3138786707042023e-06, + "loss": 0.9301, + "step": 5171 + }, + { + "epoch": 0.466429183388195, + "grad_norm": 1.4219345247992827, + "learning_rate": 2.3133016860883387e-06, + "loss": 0.9702, + "step": 5172 + }, + { + "epoch": 0.46651936691166523, + "grad_norm": 1.7571825042930387, + "learning_rate": 2.3127246747398475e-06, + "loss": 0.938, + "step": 5173 + }, + { + "epoch": 0.4666095504351355, + "grad_norm": 1.2828631596662339, + "learning_rate": 2.312147636707963e-06, + "loss": 0.957, + "step": 5174 + }, + { + "epoch": 0.46669973395860576, + "grad_norm": 1.7133100946436155, + "learning_rate": 2.3115705720419214e-06, + "loss": 0.9368, + "step": 5175 + }, + { + "epoch": 0.46678991748207604, + "grad_norm": 1.5376096732868372, + "learning_rate": 2.31099348079096e-06, + "loss": 0.9812, + "step": 5176 + }, + { + "epoch": 0.4668801010055463, + "grad_norm": 1.3318395392764577, + "learning_rate": 2.31041636300432e-06, + "loss": 1.0016, + "step": 5177 + }, + { + "epoch": 0.46697028452901657, + "grad_norm": 1.5661136801078637, + "learning_rate": 2.3098392187312445e-06, + "loss": 1.038, + "step": 5178 + }, + { + "epoch": 0.4670604680524868, + "grad_norm": 1.5363671572082134, + "learning_rate": 2.309262048020978e-06, + "loss": 0.9339, + "step": 5179 + }, + { + "epoch": 0.4671506515759571, + "grad_norm": 1.5742566882690041, + "learning_rate": 2.308684850922769e-06, + "loss": 0.9846, + "step": 5180 + }, + { + "epoch": 0.4672408350994273, + "grad_norm": 1.63853059325586, + "learning_rate": 2.3081076274858664e-06, + "loss": 0.9572, + "step": 5181 + }, + { + "epoch": 0.4673310186228976, + "grad_norm": 1.4205067517483665, + "learning_rate": 2.307530377759522e-06, + "loss": 0.9838, + "step": 5182 + }, + { + "epoch": 0.46742120214636784, + "grad_norm": 1.8172063562492364, + "learning_rate": 2.30695310179299e-06, + "loss": 0.8816, + "step": 5183 + }, + { + "epoch": 0.46751138566983813, + "grad_norm": 1.714916809889243, + "learning_rate": 2.3063757996355267e-06, + "loss": 0.9578, + "step": 5184 + }, + { + "epoch": 0.46760156919330836, + "grad_norm": 1.674608367001315, + "learning_rate": 2.3057984713363903e-06, + "loss": 0.9913, + "step": 5185 + }, + { + "epoch": 0.46769175271677865, + "grad_norm": 0.7766148586007068, + "learning_rate": 2.3052211169448436e-06, + "loss": 0.7585, + "step": 5186 + }, + { + "epoch": 0.4677819362402489, + "grad_norm": 1.7548989118496705, + "learning_rate": 2.3046437365101474e-06, + "loss": 0.9019, + "step": 5187 + }, + { + "epoch": 0.4678721197637192, + "grad_norm": 1.6504998345893562, + "learning_rate": 2.3040663300815673e-06, + "loss": 0.9384, + "step": 5188 + }, + { + "epoch": 0.4679623032871894, + "grad_norm": 0.829233004122415, + "learning_rate": 2.3034888977083723e-06, + "loss": 0.7671, + "step": 5189 + }, + { + "epoch": 0.4680524868106597, + "grad_norm": 1.8401047183382482, + "learning_rate": 2.30291143943983e-06, + "loss": 0.867, + "step": 5190 + }, + { + "epoch": 0.46814267033412993, + "grad_norm": 1.5647963624769576, + "learning_rate": 2.3023339553252145e-06, + "loss": 0.9216, + "step": 5191 + }, + { + "epoch": 0.4682328538576002, + "grad_norm": 1.4936637566221789, + "learning_rate": 2.301756445413799e-06, + "loss": 0.9933, + "step": 5192 + }, + { + "epoch": 0.4683230373810705, + "grad_norm": 1.7113041593732137, + "learning_rate": 2.3011789097548585e-06, + "loss": 0.9577, + "step": 5193 + }, + { + "epoch": 0.46841322090454074, + "grad_norm": 1.4104435593686224, + "learning_rate": 2.3006013483976738e-06, + "loss": 0.957, + "step": 5194 + }, + { + "epoch": 0.468503404428011, + "grad_norm": 2.2869849036768906, + "learning_rate": 2.300023761391524e-06, + "loss": 0.9539, + "step": 5195 + }, + { + "epoch": 0.46859358795148126, + "grad_norm": 1.7684026828701511, + "learning_rate": 2.299446148785693e-06, + "loss": 1.0135, + "step": 5196 + }, + { + "epoch": 0.46868377147495155, + "grad_norm": 1.8385153994606505, + "learning_rate": 2.2988685106294654e-06, + "loss": 0.9412, + "step": 5197 + }, + { + "epoch": 0.4687739549984218, + "grad_norm": 1.6547190950455308, + "learning_rate": 2.2982908469721284e-06, + "loss": 0.8579, + "step": 5198 + }, + { + "epoch": 0.46886413852189207, + "grad_norm": 1.6897843052835377, + "learning_rate": 2.2977131578629714e-06, + "loss": 1.0257, + "step": 5199 + }, + { + "epoch": 0.4689543220453623, + "grad_norm": 1.5619242561086881, + "learning_rate": 2.297135443351286e-06, + "loss": 0.9798, + "step": 5200 + }, + { + "epoch": 0.4690445055688326, + "grad_norm": 1.353462691599994, + "learning_rate": 2.296557703486367e-06, + "loss": 0.9732, + "step": 5201 + }, + { + "epoch": 0.4691346890923028, + "grad_norm": 1.670427712016778, + "learning_rate": 2.295979938317509e-06, + "loss": 0.9829, + "step": 5202 + }, + { + "epoch": 0.4692248726157731, + "grad_norm": 1.609578980787164, + "learning_rate": 2.295402147894011e-06, + "loss": 0.9613, + "step": 5203 + }, + { + "epoch": 0.46931505613924335, + "grad_norm": 1.3460688433166692, + "learning_rate": 2.2948243322651723e-06, + "loss": 0.8811, + "step": 5204 + }, + { + "epoch": 0.46940523966271364, + "grad_norm": 1.5616893954473736, + "learning_rate": 2.2942464914802962e-06, + "loss": 0.9651, + "step": 5205 + }, + { + "epoch": 0.46949542318618387, + "grad_norm": 1.610648530929559, + "learning_rate": 2.293668625588687e-06, + "loss": 0.9355, + "step": 5206 + }, + { + "epoch": 0.46958560670965416, + "grad_norm": 1.6508921205521492, + "learning_rate": 2.293090734639651e-06, + "loss": 0.9994, + "step": 5207 + }, + { + "epoch": 0.4696757902331244, + "grad_norm": 1.51974509335388, + "learning_rate": 2.2925128186824983e-06, + "loss": 1.0749, + "step": 5208 + }, + { + "epoch": 0.4697659737565947, + "grad_norm": 1.6080268472309178, + "learning_rate": 2.2919348777665384e-06, + "loss": 1.064, + "step": 5209 + }, + { + "epoch": 0.4698561572800649, + "grad_norm": 1.693631019495434, + "learning_rate": 2.2913569119410856e-06, + "loss": 0.9633, + "step": 5210 + }, + { + "epoch": 0.4699463408035352, + "grad_norm": 1.5509215322699028, + "learning_rate": 2.290778921255454e-06, + "loss": 0.9839, + "step": 5211 + }, + { + "epoch": 0.47003652432700543, + "grad_norm": 1.4046723735194306, + "learning_rate": 2.2902009057589613e-06, + "loss": 0.9982, + "step": 5212 + }, + { + "epoch": 0.4701267078504757, + "grad_norm": 1.2308227029977143, + "learning_rate": 2.2896228655009276e-06, + "loss": 0.9141, + "step": 5213 + }, + { + "epoch": 0.47021689137394596, + "grad_norm": 1.4396003637488801, + "learning_rate": 2.289044800530674e-06, + "loss": 0.9263, + "step": 5214 + }, + { + "epoch": 0.47030707489741624, + "grad_norm": 2.2987440121413503, + "learning_rate": 2.2884667108975245e-06, + "loss": 0.924, + "step": 5215 + }, + { + "epoch": 0.4703972584208865, + "grad_norm": 1.729003078888181, + "learning_rate": 2.287888596650804e-06, + "loss": 0.9991, + "step": 5216 + }, + { + "epoch": 0.47048744194435677, + "grad_norm": 1.695688021508447, + "learning_rate": 2.287310457839841e-06, + "loss": 0.959, + "step": 5217 + }, + { + "epoch": 0.47057762546782705, + "grad_norm": 1.8333748321460301, + "learning_rate": 2.286732294513966e-06, + "loss": 1.0341, + "step": 5218 + }, + { + "epoch": 0.4706678089912973, + "grad_norm": 1.6650569612765969, + "learning_rate": 2.2861541067225106e-06, + "loss": 0.9701, + "step": 5219 + }, + { + "epoch": 0.4707579925147676, + "grad_norm": 1.8917249522143336, + "learning_rate": 2.2855758945148095e-06, + "loss": 0.9394, + "step": 5220 + }, + { + "epoch": 0.4708481760382378, + "grad_norm": 1.5529203099661053, + "learning_rate": 2.2849976579401977e-06, + "loss": 1.0148, + "step": 5221 + }, + { + "epoch": 0.4709383595617081, + "grad_norm": 1.8392225378987708, + "learning_rate": 2.284419397048014e-06, + "loss": 0.9514, + "step": 5222 + }, + { + "epoch": 0.47102854308517833, + "grad_norm": 1.4357584479507046, + "learning_rate": 2.2838411118875997e-06, + "loss": 0.9689, + "step": 5223 + }, + { + "epoch": 0.4711187266086486, + "grad_norm": 1.5490329217385939, + "learning_rate": 2.283262802508296e-06, + "loss": 0.8838, + "step": 5224 + }, + { + "epoch": 0.47120891013211885, + "grad_norm": 1.2139753077162447, + "learning_rate": 2.2826844689594492e-06, + "loss": 0.9369, + "step": 5225 + }, + { + "epoch": 0.47129909365558914, + "grad_norm": 1.5052352779665086, + "learning_rate": 2.282106111290404e-06, + "loss": 1.0367, + "step": 5226 + }, + { + "epoch": 0.4713892771790594, + "grad_norm": 0.7437168226134944, + "learning_rate": 2.2815277295505098e-06, + "loss": 0.8021, + "step": 5227 + }, + { + "epoch": 0.47147946070252966, + "grad_norm": 1.6352788377093175, + "learning_rate": 2.2809493237891174e-06, + "loss": 0.9099, + "step": 5228 + }, + { + "epoch": 0.4715696442259999, + "grad_norm": 1.4594427611734142, + "learning_rate": 2.2803708940555796e-06, + "loss": 0.9905, + "step": 5229 + }, + { + "epoch": 0.4716598277494702, + "grad_norm": 1.3217845137267437, + "learning_rate": 2.2797924403992514e-06, + "loss": 1.0296, + "step": 5230 + }, + { + "epoch": 0.4717500112729404, + "grad_norm": 1.685139028773343, + "learning_rate": 2.2792139628694892e-06, + "loss": 0.9724, + "step": 5231 + }, + { + "epoch": 0.4718401947964107, + "grad_norm": 1.5051628114709166, + "learning_rate": 2.2786354615156524e-06, + "loss": 0.9211, + "step": 5232 + }, + { + "epoch": 0.47193037831988094, + "grad_norm": 1.665777334722661, + "learning_rate": 2.2780569363871016e-06, + "loss": 1.0217, + "step": 5233 + }, + { + "epoch": 0.4720205618433512, + "grad_norm": 0.6836158094934017, + "learning_rate": 2.277478387533199e-06, + "loss": 0.7742, + "step": 5234 + }, + { + "epoch": 0.47211074536682146, + "grad_norm": 2.3634028127632405, + "learning_rate": 2.276899815003311e-06, + "loss": 0.9966, + "step": 5235 + }, + { + "epoch": 0.47220092889029175, + "grad_norm": 2.0462241266347436, + "learning_rate": 2.2763212188468045e-06, + "loss": 0.9525, + "step": 5236 + }, + { + "epoch": 0.472291112413762, + "grad_norm": 2.419192791608385, + "learning_rate": 2.2757425991130473e-06, + "loss": 0.9781, + "step": 5237 + }, + { + "epoch": 0.47238129593723227, + "grad_norm": 0.882148965932457, + "learning_rate": 2.2751639558514117e-06, + "loss": 0.7872, + "step": 5238 + }, + { + "epoch": 0.4724714794607025, + "grad_norm": 1.3922561337456965, + "learning_rate": 2.2745852891112697e-06, + "loss": 1.0019, + "step": 5239 + }, + { + "epoch": 0.4725616629841728, + "grad_norm": 1.6966985143559106, + "learning_rate": 2.274006598941997e-06, + "loss": 0.9692, + "step": 5240 + }, + { + "epoch": 0.4726518465076431, + "grad_norm": 1.4909404716709949, + "learning_rate": 2.27342788539297e-06, + "loss": 0.9447, + "step": 5241 + }, + { + "epoch": 0.4727420300311133, + "grad_norm": 1.444729661699689, + "learning_rate": 2.2728491485135684e-06, + "loss": 0.9176, + "step": 5242 + }, + { + "epoch": 0.4728322135545836, + "grad_norm": 1.5680066539856772, + "learning_rate": 2.272270388353173e-06, + "loss": 1.0155, + "step": 5243 + }, + { + "epoch": 0.47292239707805384, + "grad_norm": 1.724504582125537, + "learning_rate": 2.2716916049611666e-06, + "loss": 0.9386, + "step": 5244 + }, + { + "epoch": 0.4730125806015241, + "grad_norm": 1.35591377626138, + "learning_rate": 2.2711127983869346e-06, + "loss": 0.9416, + "step": 5245 + }, + { + "epoch": 0.47310276412499436, + "grad_norm": 1.3645796836435369, + "learning_rate": 2.270533968679864e-06, + "loss": 0.9743, + "step": 5246 + }, + { + "epoch": 0.47319294764846465, + "grad_norm": 1.4720763828010206, + "learning_rate": 2.269955115889343e-06, + "loss": 0.9658, + "step": 5247 + }, + { + "epoch": 0.4732831311719349, + "grad_norm": 1.674589858443029, + "learning_rate": 2.269376240064763e-06, + "loss": 0.9235, + "step": 5248 + }, + { + "epoch": 0.47337331469540517, + "grad_norm": 1.5713643209525274, + "learning_rate": 2.268797341255517e-06, + "loss": 0.9737, + "step": 5249 + }, + { + "epoch": 0.4734634982188754, + "grad_norm": 1.6238250886559746, + "learning_rate": 2.268218419511e-06, + "loss": 0.8783, + "step": 5250 + }, + { + "epoch": 0.4735536817423457, + "grad_norm": 1.6626788193753885, + "learning_rate": 2.267639474880608e-06, + "loss": 0.879, + "step": 5251 + }, + { + "epoch": 0.4736438652658159, + "grad_norm": 2.3850426415364194, + "learning_rate": 2.2670605074137407e-06, + "loss": 0.939, + "step": 5252 + }, + { + "epoch": 0.4737340487892862, + "grad_norm": 1.535699102890648, + "learning_rate": 2.2664815171597983e-06, + "loss": 0.9427, + "step": 5253 + }, + { + "epoch": 0.47382423231275644, + "grad_norm": 1.722583518363457, + "learning_rate": 2.265902504168183e-06, + "loss": 0.9059, + "step": 5254 + }, + { + "epoch": 0.47391441583622673, + "grad_norm": 1.5643918600480724, + "learning_rate": 2.2653234684883007e-06, + "loss": 0.9156, + "step": 5255 + }, + { + "epoch": 0.47400459935969697, + "grad_norm": 1.6372834098786673, + "learning_rate": 2.264744410169556e-06, + "loss": 0.8814, + "step": 5256 + }, + { + "epoch": 0.47409478288316725, + "grad_norm": 1.5584465091155013, + "learning_rate": 2.264165329261359e-06, + "loss": 0.9403, + "step": 5257 + }, + { + "epoch": 0.4741849664066375, + "grad_norm": 0.7584480687911194, + "learning_rate": 2.26358622581312e-06, + "loss": 0.7778, + "step": 5258 + }, + { + "epoch": 0.4742751499301078, + "grad_norm": 1.48966359097896, + "learning_rate": 2.2630070998742504e-06, + "loss": 1.0261, + "step": 5259 + }, + { + "epoch": 0.474365333453578, + "grad_norm": 1.2916238480310556, + "learning_rate": 2.262427951494165e-06, + "loss": 0.9829, + "step": 5260 + }, + { + "epoch": 0.4744555169770483, + "grad_norm": 0.8522341118358546, + "learning_rate": 2.2618487807222794e-06, + "loss": 0.8162, + "step": 5261 + }, + { + "epoch": 0.47454570050051853, + "grad_norm": 1.7519056979322922, + "learning_rate": 2.261269587608012e-06, + "loss": 1.042, + "step": 5262 + }, + { + "epoch": 0.4746358840239888, + "grad_norm": 1.320822352093958, + "learning_rate": 2.260690372200783e-06, + "loss": 0.9626, + "step": 5263 + }, + { + "epoch": 0.47472606754745905, + "grad_norm": 1.3848608761675003, + "learning_rate": 2.2601111345500138e-06, + "loss": 1.0475, + "step": 5264 + }, + { + "epoch": 0.47481625107092934, + "grad_norm": 1.3959176574212226, + "learning_rate": 2.2595318747051286e-06, + "loss": 0.9506, + "step": 5265 + }, + { + "epoch": 0.47490643459439963, + "grad_norm": 1.6584111634717384, + "learning_rate": 2.258952592715553e-06, + "loss": 0.983, + "step": 5266 + }, + { + "epoch": 0.47499661811786986, + "grad_norm": 1.0294829748671814, + "learning_rate": 2.2583732886307142e-06, + "loss": 0.819, + "step": 5267 + }, + { + "epoch": 0.47508680164134015, + "grad_norm": 1.3233342100007608, + "learning_rate": 2.2577939625000414e-06, + "loss": 0.9073, + "step": 5268 + }, + { + "epoch": 0.4751769851648104, + "grad_norm": 0.794468706008015, + "learning_rate": 2.257214614372967e-06, + "loss": 0.7787, + "step": 5269 + }, + { + "epoch": 0.4752671686882807, + "grad_norm": 1.7633298962685175, + "learning_rate": 2.2566352442989227e-06, + "loss": 0.9859, + "step": 5270 + }, + { + "epoch": 0.4753573522117509, + "grad_norm": 1.5008903880750224, + "learning_rate": 2.256055852327344e-06, + "loss": 0.9649, + "step": 5271 + }, + { + "epoch": 0.4754475357352212, + "grad_norm": 1.497362042474256, + "learning_rate": 2.2554764385076685e-06, + "loss": 0.9119, + "step": 5272 + }, + { + "epoch": 0.4755377192586914, + "grad_norm": 1.437634834931882, + "learning_rate": 2.2548970028893348e-06, + "loss": 0.9739, + "step": 5273 + }, + { + "epoch": 0.4756279027821617, + "grad_norm": 1.6586159415788653, + "learning_rate": 2.254317545521783e-06, + "loss": 0.9756, + "step": 5274 + }, + { + "epoch": 0.47571808630563195, + "grad_norm": 1.4856208792867232, + "learning_rate": 2.253738066454457e-06, + "loss": 0.9938, + "step": 5275 + }, + { + "epoch": 0.47580826982910224, + "grad_norm": 1.5013759977524608, + "learning_rate": 2.2531585657367986e-06, + "loss": 0.936, + "step": 5276 + }, + { + "epoch": 0.47589845335257247, + "grad_norm": 1.4258710441009446, + "learning_rate": 2.252579043418256e-06, + "loss": 0.9141, + "step": 5277 + }, + { + "epoch": 0.47598863687604276, + "grad_norm": 1.5087641741727604, + "learning_rate": 2.251999499548277e-06, + "loss": 0.9329, + "step": 5278 + }, + { + "epoch": 0.476078820399513, + "grad_norm": 2.1598274298007643, + "learning_rate": 2.251419934176311e-06, + "loss": 0.9228, + "step": 5279 + }, + { + "epoch": 0.4761690039229833, + "grad_norm": 1.3458989724764483, + "learning_rate": 2.25084034735181e-06, + "loss": 1.0118, + "step": 5280 + }, + { + "epoch": 0.4762591874464535, + "grad_norm": 1.7256038299997312, + "learning_rate": 2.2502607391242274e-06, + "loss": 1.011, + "step": 5281 + }, + { + "epoch": 0.4763493709699238, + "grad_norm": 0.7224494019448794, + "learning_rate": 2.2496811095430182e-06, + "loss": 0.7834, + "step": 5282 + }, + { + "epoch": 0.47643955449339404, + "grad_norm": 1.5578108042379655, + "learning_rate": 2.249101458657641e-06, + "loss": 0.9617, + "step": 5283 + }, + { + "epoch": 0.4765297380168643, + "grad_norm": 1.63670374606227, + "learning_rate": 2.2485217865175526e-06, + "loss": 1.0325, + "step": 5284 + }, + { + "epoch": 0.47661992154033456, + "grad_norm": 1.6720045520820892, + "learning_rate": 2.2479420931722156e-06, + "loss": 1.0947, + "step": 5285 + }, + { + "epoch": 0.47671010506380485, + "grad_norm": 0.7158144837437485, + "learning_rate": 2.2473623786710923e-06, + "loss": 0.8206, + "step": 5286 + }, + { + "epoch": 0.4768002885872751, + "grad_norm": 2.4073378840892663, + "learning_rate": 2.2467826430636465e-06, + "loss": 0.8747, + "step": 5287 + }, + { + "epoch": 0.47689047211074537, + "grad_norm": 1.8008970939131947, + "learning_rate": 2.246202886399345e-06, + "loss": 0.9849, + "step": 5288 + }, + { + "epoch": 0.47698065563421566, + "grad_norm": 1.3996065676866207, + "learning_rate": 2.2456231087276556e-06, + "loss": 0.9584, + "step": 5289 + }, + { + "epoch": 0.4770708391576859, + "grad_norm": 1.35628592623754, + "learning_rate": 2.245043310098048e-06, + "loss": 0.8867, + "step": 5290 + }, + { + "epoch": 0.4771610226811562, + "grad_norm": 1.8437873060687184, + "learning_rate": 2.244463490559995e-06, + "loss": 1.0058, + "step": 5291 + }, + { + "epoch": 0.4772512062046264, + "grad_norm": 1.5546386557722545, + "learning_rate": 2.2438836501629683e-06, + "loss": 0.9753, + "step": 5292 + }, + { + "epoch": 0.4773413897280967, + "grad_norm": 1.5346004654508651, + "learning_rate": 2.2433037889564437e-06, + "loss": 0.9679, + "step": 5293 + }, + { + "epoch": 0.47743157325156693, + "grad_norm": 1.6111205921040286, + "learning_rate": 2.242723906989899e-06, + "loss": 0.9039, + "step": 5294 + }, + { + "epoch": 0.4775217567750372, + "grad_norm": 1.3379111183991574, + "learning_rate": 2.2421440043128114e-06, + "loss": 0.8929, + "step": 5295 + }, + { + "epoch": 0.47761194029850745, + "grad_norm": 1.7095939203142623, + "learning_rate": 2.241564080974662e-06, + "loss": 0.8727, + "step": 5296 + }, + { + "epoch": 0.47770212382197774, + "grad_norm": 1.7657803492437876, + "learning_rate": 2.2409841370249343e-06, + "loss": 0.9108, + "step": 5297 + }, + { + "epoch": 0.477792307345448, + "grad_norm": 1.8434561802431222, + "learning_rate": 2.2404041725131106e-06, + "loss": 1.0461, + "step": 5298 + }, + { + "epoch": 0.47788249086891826, + "grad_norm": 1.87309244396797, + "learning_rate": 2.239824187488677e-06, + "loss": 0.9385, + "step": 5299 + }, + { + "epoch": 0.4779726743923885, + "grad_norm": 0.7740883063895357, + "learning_rate": 2.239244182001122e-06, + "loss": 0.8202, + "step": 5300 + }, + { + "epoch": 0.4780628579158588, + "grad_norm": 1.4533286003304036, + "learning_rate": 2.2386641560999336e-06, + "loss": 0.9748, + "step": 5301 + }, + { + "epoch": 0.478153041439329, + "grad_norm": 1.6018963721639736, + "learning_rate": 2.238084109834604e-06, + "loss": 0.9942, + "step": 5302 + }, + { + "epoch": 0.4782432249627993, + "grad_norm": 1.7778225096730744, + "learning_rate": 2.237504043254625e-06, + "loss": 0.9618, + "step": 5303 + }, + { + "epoch": 0.47833340848626954, + "grad_norm": 1.370342733643387, + "learning_rate": 2.2369239564094915e-06, + "loss": 0.9332, + "step": 5304 + }, + { + "epoch": 0.47842359200973983, + "grad_norm": 1.5223894475628301, + "learning_rate": 2.2363438493486995e-06, + "loss": 0.8907, + "step": 5305 + }, + { + "epoch": 0.47851377553321006, + "grad_norm": 1.9465782612927862, + "learning_rate": 2.235763722121747e-06, + "loss": 0.8446, + "step": 5306 + }, + { + "epoch": 0.47860395905668035, + "grad_norm": 1.4058918814837345, + "learning_rate": 2.2351835747781346e-06, + "loss": 0.9775, + "step": 5307 + }, + { + "epoch": 0.4786941425801506, + "grad_norm": 1.525814998661696, + "learning_rate": 2.234603407367362e-06, + "loss": 1.0011, + "step": 5308 + }, + { + "epoch": 0.47878432610362087, + "grad_norm": 1.513786065501236, + "learning_rate": 2.2340232199389337e-06, + "loss": 0.9764, + "step": 5309 + }, + { + "epoch": 0.4788745096270911, + "grad_norm": 1.441058605418634, + "learning_rate": 2.2334430125423538e-06, + "loss": 0.8428, + "step": 5310 + }, + { + "epoch": 0.4789646931505614, + "grad_norm": 1.564679875889529, + "learning_rate": 2.232862785227128e-06, + "loss": 1.0083, + "step": 5311 + }, + { + "epoch": 0.4790548766740317, + "grad_norm": 1.5886108716888316, + "learning_rate": 2.232282538042766e-06, + "loss": 0.9922, + "step": 5312 + }, + { + "epoch": 0.4791450601975019, + "grad_norm": 1.6953748700319682, + "learning_rate": 2.231702271038777e-06, + "loss": 1.0076, + "step": 5313 + }, + { + "epoch": 0.4792352437209722, + "grad_norm": 1.6795972932945005, + "learning_rate": 2.231121984264673e-06, + "loss": 0.9421, + "step": 5314 + }, + { + "epoch": 0.47932542724444244, + "grad_norm": 1.4084390344774322, + "learning_rate": 2.2305416777699665e-06, + "loss": 0.9948, + "step": 5315 + }, + { + "epoch": 0.4794156107679127, + "grad_norm": 1.4878520997643938, + "learning_rate": 2.229961351604173e-06, + "loss": 0.9587, + "step": 5316 + }, + { + "epoch": 0.47950579429138296, + "grad_norm": 1.3681415389914182, + "learning_rate": 2.2293810058168085e-06, + "loss": 0.9172, + "step": 5317 + }, + { + "epoch": 0.47959597781485325, + "grad_norm": 1.4053374402473586, + "learning_rate": 2.2288006404573922e-06, + "loss": 0.9892, + "step": 5318 + }, + { + "epoch": 0.4796861613383235, + "grad_norm": 1.7512479147398334, + "learning_rate": 2.228220255575444e-06, + "loss": 0.9409, + "step": 5319 + }, + { + "epoch": 0.47977634486179377, + "grad_norm": 1.5156770677569695, + "learning_rate": 2.2276398512204847e-06, + "loss": 0.9079, + "step": 5320 + }, + { + "epoch": 0.479866528385264, + "grad_norm": 1.8298699943556882, + "learning_rate": 2.2270594274420382e-06, + "loss": 0.9657, + "step": 5321 + }, + { + "epoch": 0.4799567119087343, + "grad_norm": 0.7780712691297137, + "learning_rate": 2.22647898428963e-06, + "loss": 0.809, + "step": 5322 + }, + { + "epoch": 0.4800468954322045, + "grad_norm": 1.3049487006675402, + "learning_rate": 2.225898521812785e-06, + "loss": 1.0073, + "step": 5323 + }, + { + "epoch": 0.4801370789556748, + "grad_norm": 1.543418876564942, + "learning_rate": 2.2253180400610337e-06, + "loss": 1.009, + "step": 5324 + }, + { + "epoch": 0.48022726247914505, + "grad_norm": 1.7310179792469857, + "learning_rate": 2.2247375390839037e-06, + "loss": 0.9463, + "step": 5325 + }, + { + "epoch": 0.48031744600261533, + "grad_norm": 1.563657865668804, + "learning_rate": 2.224157018930928e-06, + "loss": 0.9325, + "step": 5326 + }, + { + "epoch": 0.48040762952608557, + "grad_norm": 1.5270949841619106, + "learning_rate": 2.2235764796516395e-06, + "loss": 0.9631, + "step": 5327 + }, + { + "epoch": 0.48049781304955586, + "grad_norm": 1.6966389332036664, + "learning_rate": 2.222995921295573e-06, + "loss": 0.9569, + "step": 5328 + }, + { + "epoch": 0.4805879965730261, + "grad_norm": 1.6029471017982253, + "learning_rate": 2.222415343912265e-06, + "loss": 0.9379, + "step": 5329 + }, + { + "epoch": 0.4806781800964964, + "grad_norm": 2.04683807572701, + "learning_rate": 2.221834747551254e-06, + "loss": 1.0314, + "step": 5330 + }, + { + "epoch": 0.4807683636199666, + "grad_norm": 1.5319803014489808, + "learning_rate": 2.221254132262078e-06, + "loss": 0.9363, + "step": 5331 + }, + { + "epoch": 0.4808585471434369, + "grad_norm": 1.5785265119064373, + "learning_rate": 2.2206734980942802e-06, + "loss": 1.001, + "step": 5332 + }, + { + "epoch": 0.48094873066690713, + "grad_norm": 1.8812861290184066, + "learning_rate": 2.2200928450974024e-06, + "loss": 0.9354, + "step": 5333 + }, + { + "epoch": 0.4810389141903774, + "grad_norm": 2.082974619182902, + "learning_rate": 2.21951217332099e-06, + "loss": 0.9825, + "step": 5334 + }, + { + "epoch": 0.48112909771384765, + "grad_norm": 1.875949174322621, + "learning_rate": 2.2189314828145883e-06, + "loss": 0.876, + "step": 5335 + }, + { + "epoch": 0.48121928123731794, + "grad_norm": 1.3914241851730231, + "learning_rate": 2.2183507736277453e-06, + "loss": 0.9479, + "step": 5336 + }, + { + "epoch": 0.48130946476078823, + "grad_norm": 1.440622835302485, + "learning_rate": 2.2177700458100107e-06, + "loss": 0.9743, + "step": 5337 + }, + { + "epoch": 0.48139964828425846, + "grad_norm": 1.5594478551935063, + "learning_rate": 2.2171892994109346e-06, + "loss": 0.9388, + "step": 5338 + }, + { + "epoch": 0.48148983180772875, + "grad_norm": 1.503283879692427, + "learning_rate": 2.21660853448007e-06, + "loss": 1.028, + "step": 5339 + }, + { + "epoch": 0.481580015331199, + "grad_norm": 1.798799959805119, + "learning_rate": 2.2160277510669703e-06, + "loss": 0.9136, + "step": 5340 + }, + { + "epoch": 0.4816701988546693, + "grad_norm": 1.2899445261597329, + "learning_rate": 2.215446949221193e-06, + "loss": 0.9497, + "step": 5341 + }, + { + "epoch": 0.4817603823781395, + "grad_norm": 1.6576509129102743, + "learning_rate": 2.2148661289922924e-06, + "loss": 0.9724, + "step": 5342 + }, + { + "epoch": 0.4818505659016098, + "grad_norm": 1.7431371179202564, + "learning_rate": 2.21428529042983e-06, + "loss": 1.03, + "step": 5343 + }, + { + "epoch": 0.48194074942508003, + "grad_norm": 1.6071187471670312, + "learning_rate": 2.2137044335833647e-06, + "loss": 1.009, + "step": 5344 + }, + { + "epoch": 0.4820309329485503, + "grad_norm": 1.8515210368346637, + "learning_rate": 2.213123558502459e-06, + "loss": 0.9494, + "step": 5345 + }, + { + "epoch": 0.48212111647202055, + "grad_norm": 1.539492629308299, + "learning_rate": 2.2125426652366763e-06, + "loss": 0.9766, + "step": 5346 + }, + { + "epoch": 0.48221129999549084, + "grad_norm": 1.5569883410507899, + "learning_rate": 2.211961753835581e-06, + "loss": 1.0598, + "step": 5347 + }, + { + "epoch": 0.48230148351896107, + "grad_norm": 1.282670512919961, + "learning_rate": 2.21138082434874e-06, + "loss": 0.9556, + "step": 5348 + }, + { + "epoch": 0.48239166704243136, + "grad_norm": 1.6886335733461355, + "learning_rate": 2.210799876825722e-06, + "loss": 0.993, + "step": 5349 + }, + { + "epoch": 0.4824818505659016, + "grad_norm": 1.615585934666578, + "learning_rate": 2.210218911316096e-06, + "loss": 0.9219, + "step": 5350 + }, + { + "epoch": 0.4825720340893719, + "grad_norm": 1.5310436810442023, + "learning_rate": 2.2096379278694336e-06, + "loss": 0.9069, + "step": 5351 + }, + { + "epoch": 0.4826622176128421, + "grad_norm": 1.515297451215972, + "learning_rate": 2.2090569265353074e-06, + "loss": 0.9041, + "step": 5352 + }, + { + "epoch": 0.4827524011363124, + "grad_norm": 1.937384263550396, + "learning_rate": 2.2084759073632912e-06, + "loss": 0.9612, + "step": 5353 + }, + { + "epoch": 0.48284258465978264, + "grad_norm": 1.5221491134022962, + "learning_rate": 2.2078948704029606e-06, + "loss": 0.9191, + "step": 5354 + }, + { + "epoch": 0.4829327681832529, + "grad_norm": 1.6479529079819817, + "learning_rate": 2.2073138157038935e-06, + "loss": 0.884, + "step": 5355 + }, + { + "epoch": 0.48302295170672316, + "grad_norm": 1.4506743079947257, + "learning_rate": 2.2067327433156687e-06, + "loss": 0.9505, + "step": 5356 + }, + { + "epoch": 0.48311313523019345, + "grad_norm": 1.4503902008935465, + "learning_rate": 2.2061516532878667e-06, + "loss": 0.9769, + "step": 5357 + }, + { + "epoch": 0.4832033187536637, + "grad_norm": 1.5029369847329346, + "learning_rate": 2.2055705456700686e-06, + "loss": 1.0086, + "step": 5358 + }, + { + "epoch": 0.48329350227713397, + "grad_norm": 1.5736930480541682, + "learning_rate": 2.204989420511858e-06, + "loss": 0.8565, + "step": 5359 + }, + { + "epoch": 0.48338368580060426, + "grad_norm": 1.533561907538238, + "learning_rate": 2.20440827786282e-06, + "loss": 0.9087, + "step": 5360 + }, + { + "epoch": 0.4834738693240745, + "grad_norm": 1.4157118478033381, + "learning_rate": 2.20382711777254e-06, + "loss": 0.9641, + "step": 5361 + }, + { + "epoch": 0.4835640528475448, + "grad_norm": 1.7224455205000528, + "learning_rate": 2.203245940290607e-06, + "loss": 0.9731, + "step": 5362 + }, + { + "epoch": 0.483654236371015, + "grad_norm": 1.5850323128038892, + "learning_rate": 2.2026647454666097e-06, + "loss": 0.9517, + "step": 5363 + }, + { + "epoch": 0.4837444198944853, + "grad_norm": 1.863220661705734, + "learning_rate": 2.2020835333501384e-06, + "loss": 0.9956, + "step": 5364 + }, + { + "epoch": 0.48383460341795553, + "grad_norm": 1.593091492089123, + "learning_rate": 2.2015023039907863e-06, + "loss": 1.0196, + "step": 5365 + }, + { + "epoch": 0.4839247869414258, + "grad_norm": 1.4526480435482927, + "learning_rate": 2.2009210574381464e-06, + "loss": 0.9416, + "step": 5366 + }, + { + "epoch": 0.48401497046489605, + "grad_norm": 1.5157051458503084, + "learning_rate": 2.2003397937418134e-06, + "loss": 0.9031, + "step": 5367 + }, + { + "epoch": 0.48410515398836634, + "grad_norm": 0.7387940175195813, + "learning_rate": 2.1997585129513852e-06, + "loss": 0.8044, + "step": 5368 + }, + { + "epoch": 0.4841953375118366, + "grad_norm": 1.481443300792759, + "learning_rate": 2.1991772151164595e-06, + "loss": 0.987, + "step": 5369 + }, + { + "epoch": 0.48428552103530687, + "grad_norm": 1.3417646690179472, + "learning_rate": 2.1985959002866346e-06, + "loss": 1.0044, + "step": 5370 + }, + { + "epoch": 0.4843757045587771, + "grad_norm": 1.8266114057874174, + "learning_rate": 2.198014568511513e-06, + "loss": 1.023, + "step": 5371 + }, + { + "epoch": 0.4844658880822474, + "grad_norm": 0.7330142867841453, + "learning_rate": 2.1974332198406965e-06, + "loss": 0.7768, + "step": 5372 + }, + { + "epoch": 0.4845560716057176, + "grad_norm": 2.186643596310372, + "learning_rate": 2.196851854323789e-06, + "loss": 0.943, + "step": 5373 + }, + { + "epoch": 0.4846462551291879, + "grad_norm": 1.663033825634117, + "learning_rate": 2.196270472010396e-06, + "loss": 0.9441, + "step": 5374 + }, + { + "epoch": 0.48473643865265814, + "grad_norm": 1.5954711446612424, + "learning_rate": 2.195689072950124e-06, + "loss": 0.955, + "step": 5375 + }, + { + "epoch": 0.48482662217612843, + "grad_norm": 0.7699649189103953, + "learning_rate": 2.195107657192581e-06, + "loss": 0.7649, + "step": 5376 + }, + { + "epoch": 0.48491680569959866, + "grad_norm": 0.643358585655965, + "learning_rate": 2.194526224787378e-06, + "loss": 0.7178, + "step": 5377 + }, + { + "epoch": 0.48500698922306895, + "grad_norm": 1.5759439788146603, + "learning_rate": 2.1939447757841236e-06, + "loss": 0.9252, + "step": 5378 + }, + { + "epoch": 0.4850971727465392, + "grad_norm": 1.6472021599516897, + "learning_rate": 2.193363310232432e-06, + "loss": 1.0219, + "step": 5379 + }, + { + "epoch": 0.4851873562700095, + "grad_norm": 1.3008218020054918, + "learning_rate": 2.192781828181917e-06, + "loss": 1.0621, + "step": 5380 + }, + { + "epoch": 0.4852775397934797, + "grad_norm": 1.4645023202353527, + "learning_rate": 2.192200329682193e-06, + "loss": 0.9846, + "step": 5381 + }, + { + "epoch": 0.48536772331695, + "grad_norm": 1.2827251130423765, + "learning_rate": 2.1916188147828767e-06, + "loss": 0.9588, + "step": 5382 + }, + { + "epoch": 0.48545790684042023, + "grad_norm": 1.5451242577173199, + "learning_rate": 2.191037283533587e-06, + "loss": 0.9707, + "step": 5383 + }, + { + "epoch": 0.4855480903638905, + "grad_norm": 1.9502483698810347, + "learning_rate": 2.1904557359839428e-06, + "loss": 0.9582, + "step": 5384 + }, + { + "epoch": 0.4856382738873608, + "grad_norm": 1.618711702342985, + "learning_rate": 2.189874172183565e-06, + "loss": 0.9377, + "step": 5385 + }, + { + "epoch": 0.48572845741083104, + "grad_norm": 1.4426545317862818, + "learning_rate": 2.1892925921820763e-06, + "loss": 0.9334, + "step": 5386 + }, + { + "epoch": 0.4858186409343013, + "grad_norm": 1.6901831312105824, + "learning_rate": 2.1887109960290994e-06, + "loss": 0.9564, + "step": 5387 + }, + { + "epoch": 0.48590882445777156, + "grad_norm": 1.4753987403749453, + "learning_rate": 2.18812938377426e-06, + "loss": 1.0559, + "step": 5388 + }, + { + "epoch": 0.48599900798124185, + "grad_norm": 2.650650495187538, + "learning_rate": 2.187547755467184e-06, + "loss": 0.9924, + "step": 5389 + }, + { + "epoch": 0.4860891915047121, + "grad_norm": 1.4197040794860478, + "learning_rate": 2.1869661111574994e-06, + "loss": 0.8509, + "step": 5390 + }, + { + "epoch": 0.48617937502818237, + "grad_norm": 1.7227176300670188, + "learning_rate": 2.1863844508948353e-06, + "loss": 1.0088, + "step": 5391 + }, + { + "epoch": 0.4862695585516526, + "grad_norm": 1.5470302003009866, + "learning_rate": 2.185802774728823e-06, + "loss": 0.9764, + "step": 5392 + }, + { + "epoch": 0.4863597420751229, + "grad_norm": 1.3291733138696484, + "learning_rate": 2.1852210827090927e-06, + "loss": 0.912, + "step": 5393 + }, + { + "epoch": 0.4864499255985931, + "grad_norm": 1.8511705587317606, + "learning_rate": 2.184639374885278e-06, + "loss": 0.9053, + "step": 5394 + }, + { + "epoch": 0.4865401091220634, + "grad_norm": 1.9901223043496954, + "learning_rate": 2.184057651307014e-06, + "loss": 1.0521, + "step": 5395 + }, + { + "epoch": 0.48663029264553365, + "grad_norm": 4.478152327311071, + "learning_rate": 2.183475912023937e-06, + "loss": 1.0296, + "step": 5396 + }, + { + "epoch": 0.48672047616900393, + "grad_norm": 1.5232776264519694, + "learning_rate": 2.1828941570856826e-06, + "loss": 0.9749, + "step": 5397 + }, + { + "epoch": 0.48681065969247417, + "grad_norm": 0.7497617422585467, + "learning_rate": 2.1823123865418903e-06, + "loss": 0.7733, + "step": 5398 + }, + { + "epoch": 0.48690084321594446, + "grad_norm": 2.0903809044267603, + "learning_rate": 2.1817306004422e-06, + "loss": 0.9331, + "step": 5399 + }, + { + "epoch": 0.4869910267394147, + "grad_norm": 1.5800901961838039, + "learning_rate": 2.1811487988362527e-06, + "loss": 0.9821, + "step": 5400 + }, + { + "epoch": 0.487081210262885, + "grad_norm": 1.7382663683307853, + "learning_rate": 2.1805669817736917e-06, + "loss": 1.0301, + "step": 5401 + }, + { + "epoch": 0.4871713937863552, + "grad_norm": 1.8985579966541832, + "learning_rate": 2.17998514930416e-06, + "loss": 0.7845, + "step": 5402 + }, + { + "epoch": 0.4872615773098255, + "grad_norm": 1.8483582790034039, + "learning_rate": 2.1794033014773025e-06, + "loss": 0.9572, + "step": 5403 + }, + { + "epoch": 0.48735176083329573, + "grad_norm": 1.5790498497397911, + "learning_rate": 2.178821438342766e-06, + "loss": 1.0249, + "step": 5404 + }, + { + "epoch": 0.487441944356766, + "grad_norm": 1.561196355577653, + "learning_rate": 2.1782395599501996e-06, + "loss": 0.9511, + "step": 5405 + }, + { + "epoch": 0.48753212788023625, + "grad_norm": 0.7509876344655808, + "learning_rate": 2.1776576663492498e-06, + "loss": 0.7859, + "step": 5406 + }, + { + "epoch": 0.48762231140370654, + "grad_norm": 1.9874589639901805, + "learning_rate": 2.177075757589569e-06, + "loss": 1.0452, + "step": 5407 + }, + { + "epoch": 0.48771249492717683, + "grad_norm": 1.383295179706493, + "learning_rate": 2.176493833720808e-06, + "loss": 0.888, + "step": 5408 + }, + { + "epoch": 0.48780267845064706, + "grad_norm": 1.6607166735607037, + "learning_rate": 2.1759118947926195e-06, + "loss": 0.9285, + "step": 5409 + }, + { + "epoch": 0.48789286197411735, + "grad_norm": 1.444457424953204, + "learning_rate": 2.1753299408546587e-06, + "loss": 0.9714, + "step": 5410 + }, + { + "epoch": 0.4879830454975876, + "grad_norm": 3.30133059292414, + "learning_rate": 2.1747479719565803e-06, + "loss": 0.9666, + "step": 5411 + }, + { + "epoch": 0.4880732290210579, + "grad_norm": 1.2886053488477536, + "learning_rate": 2.174165988148042e-06, + "loss": 1.0205, + "step": 5412 + }, + { + "epoch": 0.4881634125445281, + "grad_norm": 1.5011234050294051, + "learning_rate": 2.1735839894787003e-06, + "loss": 0.9259, + "step": 5413 + }, + { + "epoch": 0.4882535960679984, + "grad_norm": 0.8756953609931325, + "learning_rate": 2.1730019759982163e-06, + "loss": 0.8311, + "step": 5414 + }, + { + "epoch": 0.48834377959146863, + "grad_norm": 1.4296245456863939, + "learning_rate": 2.172419947756249e-06, + "loss": 1.011, + "step": 5415 + }, + { + "epoch": 0.4884339631149389, + "grad_norm": 1.5897388470506404, + "learning_rate": 2.171837904802461e-06, + "loss": 1.0089, + "step": 5416 + }, + { + "epoch": 0.48852414663840915, + "grad_norm": 1.5039416816798186, + "learning_rate": 2.171255847186516e-06, + "loss": 0.9556, + "step": 5417 + }, + { + "epoch": 0.48861433016187944, + "grad_norm": 1.644740800350847, + "learning_rate": 2.1706737749580783e-06, + "loss": 0.9474, + "step": 5418 + }, + { + "epoch": 0.4887045136853497, + "grad_norm": 1.4044247968237644, + "learning_rate": 2.1700916881668127e-06, + "loss": 1.0326, + "step": 5419 + }, + { + "epoch": 0.48879469720881996, + "grad_norm": 1.6329731109330226, + "learning_rate": 2.1695095868623862e-06, + "loss": 1.0496, + "step": 5420 + }, + { + "epoch": 0.4888848807322902, + "grad_norm": 1.5047749178043783, + "learning_rate": 2.168927471094467e-06, + "loss": 0.8921, + "step": 5421 + }, + { + "epoch": 0.4889750642557605, + "grad_norm": 1.9375394386461413, + "learning_rate": 2.168345340912725e-06, + "loss": 0.9804, + "step": 5422 + }, + { + "epoch": 0.4890652477792307, + "grad_norm": 1.417353893467038, + "learning_rate": 2.1677631963668298e-06, + "loss": 0.9779, + "step": 5423 + }, + { + "epoch": 0.489155431302701, + "grad_norm": 1.4791071526101778, + "learning_rate": 2.167181037506453e-06, + "loss": 0.9118, + "step": 5424 + }, + { + "epoch": 0.48924561482617124, + "grad_norm": 1.5307098525902199, + "learning_rate": 2.1665988643812693e-06, + "loss": 0.8955, + "step": 5425 + }, + { + "epoch": 0.4893357983496415, + "grad_norm": 1.8136290618095896, + "learning_rate": 2.166016677040951e-06, + "loss": 0.9625, + "step": 5426 + }, + { + "epoch": 0.48942598187311176, + "grad_norm": 1.6511443995655113, + "learning_rate": 2.165434475535175e-06, + "loss": 0.9471, + "step": 5427 + }, + { + "epoch": 0.48951616539658205, + "grad_norm": 1.488764246464874, + "learning_rate": 2.1648522599136173e-06, + "loss": 1.0013, + "step": 5428 + }, + { + "epoch": 0.4896063489200523, + "grad_norm": 1.839250667412717, + "learning_rate": 2.164270030225956e-06, + "loss": 0.9277, + "step": 5429 + }, + { + "epoch": 0.48969653244352257, + "grad_norm": 1.3732144726932576, + "learning_rate": 2.16368778652187e-06, + "loss": 0.8725, + "step": 5430 + }, + { + "epoch": 0.4897867159669928, + "grad_norm": 1.5642306851899652, + "learning_rate": 2.163105528851039e-06, + "loss": 1.0009, + "step": 5431 + }, + { + "epoch": 0.4898768994904631, + "grad_norm": 1.4469670902878118, + "learning_rate": 2.1625232572631448e-06, + "loss": 0.9647, + "step": 5432 + }, + { + "epoch": 0.4899670830139334, + "grad_norm": 0.8437141834710364, + "learning_rate": 2.161940971807871e-06, + "loss": 0.7926, + "step": 5433 + }, + { + "epoch": 0.4900572665374036, + "grad_norm": 1.7416460367590192, + "learning_rate": 2.1613586725348994e-06, + "loss": 0.9461, + "step": 5434 + }, + { + "epoch": 0.4901474500608739, + "grad_norm": 1.5450977172319194, + "learning_rate": 2.1607763594939176e-06, + "loss": 1.0396, + "step": 5435 + }, + { + "epoch": 0.49023763358434413, + "grad_norm": 1.5577217282218487, + "learning_rate": 2.1601940327346093e-06, + "loss": 0.869, + "step": 5436 + }, + { + "epoch": 0.4903278171078144, + "grad_norm": 1.4355953701417614, + "learning_rate": 2.159611692306663e-06, + "loss": 1.03, + "step": 5437 + }, + { + "epoch": 0.49041800063128466, + "grad_norm": 1.5926415571839099, + "learning_rate": 2.1590293382597667e-06, + "loss": 0.9195, + "step": 5438 + }, + { + "epoch": 0.49050818415475494, + "grad_norm": 1.51784989170939, + "learning_rate": 2.1584469706436102e-06, + "loss": 0.9002, + "step": 5439 + }, + { + "epoch": 0.4905983676782252, + "grad_norm": 1.4413913632996502, + "learning_rate": 2.1578645895078855e-06, + "loss": 0.8732, + "step": 5440 + }, + { + "epoch": 0.49068855120169547, + "grad_norm": 1.4715694393789036, + "learning_rate": 2.157282194902283e-06, + "loss": 1.0167, + "step": 5441 + }, + { + "epoch": 0.4907787347251657, + "grad_norm": 1.9114802789770904, + "learning_rate": 2.1566997868764965e-06, + "loss": 0.897, + "step": 5442 + }, + { + "epoch": 0.490868918248636, + "grad_norm": 1.708553586888099, + "learning_rate": 2.15611736548022e-06, + "loss": 0.9177, + "step": 5443 + }, + { + "epoch": 0.4909591017721062, + "grad_norm": 1.596935305417831, + "learning_rate": 2.155534930763149e-06, + "loss": 0.9698, + "step": 5444 + }, + { + "epoch": 0.4910492852955765, + "grad_norm": 1.5765856798810076, + "learning_rate": 2.1549524827749804e-06, + "loss": 0.845, + "step": 5445 + }, + { + "epoch": 0.49113946881904674, + "grad_norm": 1.343819372471188, + "learning_rate": 2.1543700215654115e-06, + "loss": 0.8099, + "step": 5446 + }, + { + "epoch": 0.49122965234251703, + "grad_norm": 1.5967587512678194, + "learning_rate": 2.153787547184141e-06, + "loss": 0.9516, + "step": 5447 + }, + { + "epoch": 0.49131983586598726, + "grad_norm": 1.5771542575598883, + "learning_rate": 2.1532050596808695e-06, + "loss": 0.8864, + "step": 5448 + }, + { + "epoch": 0.49141001938945755, + "grad_norm": 2.069428453092071, + "learning_rate": 2.152622559105297e-06, + "loss": 1.0045, + "step": 5449 + }, + { + "epoch": 0.4915002029129278, + "grad_norm": 1.8288106692113897, + "learning_rate": 2.152040045507126e-06, + "loss": 1.0084, + "step": 5450 + }, + { + "epoch": 0.4915903864363981, + "grad_norm": 1.943931487708929, + "learning_rate": 2.1514575189360607e-06, + "loss": 0.9766, + "step": 5451 + }, + { + "epoch": 0.4916805699598683, + "grad_norm": 1.5651350309234675, + "learning_rate": 2.1508749794418043e-06, + "loss": 0.9166, + "step": 5452 + }, + { + "epoch": 0.4917707534833386, + "grad_norm": 1.5809760448177954, + "learning_rate": 2.1502924270740626e-06, + "loss": 0.9303, + "step": 5453 + }, + { + "epoch": 0.49186093700680883, + "grad_norm": 1.6328646272798393, + "learning_rate": 2.1497098618825427e-06, + "loss": 0.9627, + "step": 5454 + }, + { + "epoch": 0.4919511205302791, + "grad_norm": 1.6872811352118957, + "learning_rate": 2.1491272839169516e-06, + "loss": 1.0019, + "step": 5455 + }, + { + "epoch": 0.4920413040537494, + "grad_norm": 1.4856987922903364, + "learning_rate": 2.1485446932269986e-06, + "loss": 1.0683, + "step": 5456 + }, + { + "epoch": 0.49213148757721964, + "grad_norm": 1.5797339820248957, + "learning_rate": 2.147962089862393e-06, + "loss": 1.0016, + "step": 5457 + }, + { + "epoch": 0.49222167110068993, + "grad_norm": 1.8856984414429967, + "learning_rate": 2.1473794738728462e-06, + "loss": 0.8672, + "step": 5458 + }, + { + "epoch": 0.49231185462416016, + "grad_norm": 1.8244549028672743, + "learning_rate": 2.14679684530807e-06, + "loss": 0.9874, + "step": 5459 + }, + { + "epoch": 0.49240203814763045, + "grad_norm": 1.6330354529209787, + "learning_rate": 2.1462142042177774e-06, + "loss": 0.9423, + "step": 5460 + }, + { + "epoch": 0.4924922216711007, + "grad_norm": 1.5294286054875028, + "learning_rate": 2.145631550651683e-06, + "loss": 0.9576, + "step": 5461 + }, + { + "epoch": 0.49258240519457097, + "grad_norm": 1.5650350224290328, + "learning_rate": 2.1450488846595016e-06, + "loss": 1.0645, + "step": 5462 + }, + { + "epoch": 0.4926725887180412, + "grad_norm": 2.159254772411115, + "learning_rate": 2.14446620629095e-06, + "loss": 0.9368, + "step": 5463 + }, + { + "epoch": 0.4927627722415115, + "grad_norm": 1.4481075993440384, + "learning_rate": 2.1438835155957445e-06, + "loss": 1.0269, + "step": 5464 + }, + { + "epoch": 0.4928529557649817, + "grad_norm": 1.5641255124991433, + "learning_rate": 2.143300812623604e-06, + "loss": 1.0006, + "step": 5465 + }, + { + "epoch": 0.492943139288452, + "grad_norm": 1.5457175962953091, + "learning_rate": 2.1427180974242485e-06, + "loss": 0.9462, + "step": 5466 + }, + { + "epoch": 0.49303332281192225, + "grad_norm": 1.5060937602580093, + "learning_rate": 2.142135370047398e-06, + "loss": 0.9851, + "step": 5467 + }, + { + "epoch": 0.49312350633539254, + "grad_norm": 1.8357141237115708, + "learning_rate": 2.1415526305427735e-06, + "loss": 0.9882, + "step": 5468 + }, + { + "epoch": 0.49321368985886277, + "grad_norm": 1.303344720845992, + "learning_rate": 2.140969878960098e-06, + "loss": 0.8818, + "step": 5469 + }, + { + "epoch": 0.49330387338233306, + "grad_norm": 1.5420106641497136, + "learning_rate": 2.1403871153490956e-06, + "loss": 1.0147, + "step": 5470 + }, + { + "epoch": 0.4933940569058033, + "grad_norm": 1.4825071930992249, + "learning_rate": 2.13980433975949e-06, + "loss": 1.0816, + "step": 5471 + }, + { + "epoch": 0.4934842404292736, + "grad_norm": 1.4239447484441607, + "learning_rate": 2.1392215522410076e-06, + "loss": 1.0056, + "step": 5472 + }, + { + "epoch": 0.4935744239527438, + "grad_norm": 2.134664548835125, + "learning_rate": 2.1386387528433743e-06, + "loss": 1.0759, + "step": 5473 + }, + { + "epoch": 0.4936646074762141, + "grad_norm": 1.8712900970586719, + "learning_rate": 2.1380559416163186e-06, + "loss": 0.9897, + "step": 5474 + }, + { + "epoch": 0.49375479099968433, + "grad_norm": 1.4489114029298131, + "learning_rate": 2.1374731186095685e-06, + "loss": 1.0064, + "step": 5475 + }, + { + "epoch": 0.4938449745231546, + "grad_norm": 1.3402565634349206, + "learning_rate": 2.136890283872854e-06, + "loss": 0.9079, + "step": 5476 + }, + { + "epoch": 0.49393515804662486, + "grad_norm": 1.3940652448006763, + "learning_rate": 2.136307437455906e-06, + "loss": 0.9365, + "step": 5477 + }, + { + "epoch": 0.49402534157009514, + "grad_norm": 1.3194130536857165, + "learning_rate": 2.135724579408456e-06, + "loss": 0.9076, + "step": 5478 + }, + { + "epoch": 0.49411552509356543, + "grad_norm": 1.6683775068158355, + "learning_rate": 2.1351417097802356e-06, + "loss": 1.0547, + "step": 5479 + }, + { + "epoch": 0.49420570861703567, + "grad_norm": 1.6191839895664981, + "learning_rate": 2.1345588286209798e-06, + "loss": 0.8424, + "step": 5480 + }, + { + "epoch": 0.49429589214050595, + "grad_norm": 2.087886285732745, + "learning_rate": 2.1339759359804227e-06, + "loss": 0.8079, + "step": 5481 + }, + { + "epoch": 0.4943860756639762, + "grad_norm": 1.570959916145535, + "learning_rate": 2.1333930319082997e-06, + "loss": 0.9367, + "step": 5482 + }, + { + "epoch": 0.4944762591874465, + "grad_norm": 1.646184819349698, + "learning_rate": 2.132810116454348e-06, + "loss": 0.9599, + "step": 5483 + }, + { + "epoch": 0.4945664427109167, + "grad_norm": 1.5593715629463312, + "learning_rate": 2.132227189668305e-06, + "loss": 0.9541, + "step": 5484 + }, + { + "epoch": 0.494656626234387, + "grad_norm": 1.4963207421274767, + "learning_rate": 2.1316442515999096e-06, + "loss": 0.9018, + "step": 5485 + }, + { + "epoch": 0.49474680975785723, + "grad_norm": 1.3733193790473741, + "learning_rate": 2.1310613022989e-06, + "loss": 0.9915, + "step": 5486 + }, + { + "epoch": 0.4948369932813275, + "grad_norm": 1.3865014134264682, + "learning_rate": 2.130478341815017e-06, + "loss": 0.8092, + "step": 5487 + }, + { + "epoch": 0.49492717680479775, + "grad_norm": 1.4367522907335066, + "learning_rate": 2.1298953701980033e-06, + "loss": 0.8562, + "step": 5488 + }, + { + "epoch": 0.49501736032826804, + "grad_norm": 1.8171845768177666, + "learning_rate": 2.1293123874976003e-06, + "loss": 0.9684, + "step": 5489 + }, + { + "epoch": 0.4951075438517383, + "grad_norm": 1.4626868071622707, + "learning_rate": 2.1287293937635513e-06, + "loss": 0.9112, + "step": 5490 + }, + { + "epoch": 0.49519772737520856, + "grad_norm": 1.8292298198158736, + "learning_rate": 2.1281463890456005e-06, + "loss": 1.0119, + "step": 5491 + }, + { + "epoch": 0.4952879108986788, + "grad_norm": 1.4522571228031287, + "learning_rate": 2.127563373393493e-06, + "loss": 0.9519, + "step": 5492 + }, + { + "epoch": 0.4953780944221491, + "grad_norm": 1.6123724864842828, + "learning_rate": 2.1269803468569756e-06, + "loss": 0.9533, + "step": 5493 + }, + { + "epoch": 0.4954682779456193, + "grad_norm": 1.3532924715851289, + "learning_rate": 2.126397309485794e-06, + "loss": 0.995, + "step": 5494 + }, + { + "epoch": 0.4955584614690896, + "grad_norm": 1.5816426117692124, + "learning_rate": 2.1258142613296983e-06, + "loss": 0.9225, + "step": 5495 + }, + { + "epoch": 0.49564864499255984, + "grad_norm": 1.339515579115949, + "learning_rate": 2.125231202438435e-06, + "loss": 0.9324, + "step": 5496 + }, + { + "epoch": 0.49573882851603013, + "grad_norm": 1.3163156449919873, + "learning_rate": 2.1246481328617553e-06, + "loss": 0.9924, + "step": 5497 + }, + { + "epoch": 0.49582901203950036, + "grad_norm": 1.9672765516224084, + "learning_rate": 2.1240650526494096e-06, + "loss": 0.9394, + "step": 5498 + }, + { + "epoch": 0.49591919556297065, + "grad_norm": 1.4588029741018953, + "learning_rate": 2.1234819618511493e-06, + "loss": 0.9283, + "step": 5499 + }, + { + "epoch": 0.4960093790864409, + "grad_norm": 0.7866181007105729, + "learning_rate": 2.122898860516728e-06, + "loss": 0.8205, + "step": 5500 + }, + { + "epoch": 0.49609956260991117, + "grad_norm": 1.513040994410261, + "learning_rate": 2.1223157486958976e-06, + "loss": 0.9473, + "step": 5501 + }, + { + "epoch": 0.4961897461333814, + "grad_norm": 1.407300133142009, + "learning_rate": 2.1217326264384127e-06, + "loss": 0.8483, + "step": 5502 + }, + { + "epoch": 0.4962799296568517, + "grad_norm": 2.109760051480499, + "learning_rate": 2.1211494937940296e-06, + "loss": 0.9784, + "step": 5503 + }, + { + "epoch": 0.496370113180322, + "grad_norm": 1.3235498500006362, + "learning_rate": 2.1205663508125034e-06, + "loss": 0.9704, + "step": 5504 + }, + { + "epoch": 0.4964602967037922, + "grad_norm": 1.3721048914125809, + "learning_rate": 2.1199831975435914e-06, + "loss": 0.922, + "step": 5505 + }, + { + "epoch": 0.4965504802272625, + "grad_norm": 1.4770422262148941, + "learning_rate": 2.1194000340370517e-06, + "loss": 1.0024, + "step": 5506 + }, + { + "epoch": 0.49664066375073274, + "grad_norm": 1.7310410493978232, + "learning_rate": 2.1188168603426423e-06, + "loss": 0.9013, + "step": 5507 + }, + { + "epoch": 0.496730847274203, + "grad_norm": 2.173391347980628, + "learning_rate": 2.118233676510123e-06, + "loss": 0.9237, + "step": 5508 + }, + { + "epoch": 0.49682103079767326, + "grad_norm": 1.5234036172864691, + "learning_rate": 2.117650482589255e-06, + "loss": 0.9709, + "step": 5509 + }, + { + "epoch": 0.49691121432114355, + "grad_norm": 1.6310885148485579, + "learning_rate": 2.1170672786297988e-06, + "loss": 0.995, + "step": 5510 + }, + { + "epoch": 0.4970013978446138, + "grad_norm": 1.2466391204061822, + "learning_rate": 2.1164840646815174e-06, + "loss": 0.9261, + "step": 5511 + }, + { + "epoch": 0.49709158136808407, + "grad_norm": 1.733656863438583, + "learning_rate": 2.1159008407941726e-06, + "loss": 1.0398, + "step": 5512 + }, + { + "epoch": 0.4971817648915543, + "grad_norm": 1.5310886843041136, + "learning_rate": 2.1153176070175293e-06, + "loss": 0.9498, + "step": 5513 + }, + { + "epoch": 0.4972719484150246, + "grad_norm": 1.4213689385589239, + "learning_rate": 2.114734363401352e-06, + "loss": 0.8523, + "step": 5514 + }, + { + "epoch": 0.4973621319384948, + "grad_norm": 1.7108513100570433, + "learning_rate": 2.1141511099954056e-06, + "loss": 0.9567, + "step": 5515 + }, + { + "epoch": 0.4974523154619651, + "grad_norm": 1.3655602935196332, + "learning_rate": 2.1135678468494576e-06, + "loss": 0.9393, + "step": 5516 + }, + { + "epoch": 0.49754249898543534, + "grad_norm": 1.9109131107349084, + "learning_rate": 2.112984574013275e-06, + "loss": 1.0012, + "step": 5517 + }, + { + "epoch": 0.49763268250890563, + "grad_norm": 1.3889295979997576, + "learning_rate": 2.112401291536625e-06, + "loss": 0.9184, + "step": 5518 + }, + { + "epoch": 0.49772286603237587, + "grad_norm": 1.472015484303427, + "learning_rate": 2.111817999469278e-06, + "loss": 0.9917, + "step": 5519 + }, + { + "epoch": 0.49781304955584615, + "grad_norm": 1.3886352620080007, + "learning_rate": 2.1112346978610016e-06, + "loss": 0.9061, + "step": 5520 + }, + { + "epoch": 0.4979032330793164, + "grad_norm": 1.5624573511025133, + "learning_rate": 2.1106513867615678e-06, + "loss": 0.9132, + "step": 5521 + }, + { + "epoch": 0.4979934166027867, + "grad_norm": 1.2488256660272554, + "learning_rate": 2.110068066220748e-06, + "loss": 0.9572, + "step": 5522 + }, + { + "epoch": 0.4980836001262569, + "grad_norm": 1.2478650935262334, + "learning_rate": 2.109484736288313e-06, + "loss": 1.023, + "step": 5523 + }, + { + "epoch": 0.4981737836497272, + "grad_norm": 1.5935556629849785, + "learning_rate": 2.108901397014037e-06, + "loss": 0.9753, + "step": 5524 + }, + { + "epoch": 0.49826396717319743, + "grad_norm": 1.422615612943265, + "learning_rate": 2.1083180484476934e-06, + "loss": 0.9633, + "step": 5525 + }, + { + "epoch": 0.4983541506966677, + "grad_norm": 1.4573799104765695, + "learning_rate": 2.1077346906390567e-06, + "loss": 0.9812, + "step": 5526 + }, + { + "epoch": 0.498444334220138, + "grad_norm": 1.6756056758075093, + "learning_rate": 2.107151323637902e-06, + "loss": 0.9747, + "step": 5527 + }, + { + "epoch": 0.49853451774360824, + "grad_norm": 1.678656328077549, + "learning_rate": 2.106567947494006e-06, + "loss": 0.9404, + "step": 5528 + }, + { + "epoch": 0.49862470126707853, + "grad_norm": 1.5779629614705242, + "learning_rate": 2.1059845622571447e-06, + "loss": 1.0256, + "step": 5529 + }, + { + "epoch": 0.49871488479054876, + "grad_norm": 1.3859752542370158, + "learning_rate": 2.1054011679770956e-06, + "loss": 0.9938, + "step": 5530 + }, + { + "epoch": 0.49880506831401905, + "grad_norm": 1.557453015324455, + "learning_rate": 2.104817764703638e-06, + "loss": 0.9665, + "step": 5531 + }, + { + "epoch": 0.4988952518374893, + "grad_norm": 1.6820424449386875, + "learning_rate": 2.1042343524865516e-06, + "loss": 0.9622, + "step": 5532 + }, + { + "epoch": 0.4989854353609596, + "grad_norm": 1.5159512984623524, + "learning_rate": 2.103650931375615e-06, + "loss": 1.0673, + "step": 5533 + }, + { + "epoch": 0.4990756188844298, + "grad_norm": 1.7934397492275833, + "learning_rate": 2.1030675014206094e-06, + "loss": 0.9008, + "step": 5534 + }, + { + "epoch": 0.4991658024079001, + "grad_norm": 1.990148420789422, + "learning_rate": 2.1024840626713166e-06, + "loss": 0.8994, + "step": 5535 + }, + { + "epoch": 0.4992559859313703, + "grad_norm": 1.4930708103238484, + "learning_rate": 2.1019006151775177e-06, + "loss": 1.06, + "step": 5536 + }, + { + "epoch": 0.4993461694548406, + "grad_norm": 1.6650523076131911, + "learning_rate": 2.101317158988997e-06, + "loss": 0.9487, + "step": 5537 + }, + { + "epoch": 0.49943635297831085, + "grad_norm": 1.8485625868817692, + "learning_rate": 2.1007336941555374e-06, + "loss": 0.9298, + "step": 5538 + }, + { + "epoch": 0.49952653650178114, + "grad_norm": 1.4682326521690579, + "learning_rate": 2.1001502207269238e-06, + "loss": 0.9206, + "step": 5539 + }, + { + "epoch": 0.49961672002525137, + "grad_norm": 1.8996625173902382, + "learning_rate": 2.0995667387529407e-06, + "loss": 0.9103, + "step": 5540 + }, + { + "epoch": 0.49970690354872166, + "grad_norm": 1.5135419236051897, + "learning_rate": 2.098983248283375e-06, + "loss": 0.907, + "step": 5541 + }, + { + "epoch": 0.4997970870721919, + "grad_norm": 2.183560557104604, + "learning_rate": 2.098399749368012e-06, + "loss": 0.9571, + "step": 5542 + }, + { + "epoch": 0.4998872705956622, + "grad_norm": 0.7457508277722227, + "learning_rate": 2.09781624205664e-06, + "loss": 0.7715, + "step": 5543 + }, + { + "epoch": 0.4999774541191324, + "grad_norm": 1.546788454042979, + "learning_rate": 2.0972327263990477e-06, + "loss": 1.0063, + "step": 5544 + }, + { + "epoch": 0.5000676376426026, + "grad_norm": 1.6647037470507917, + "learning_rate": 2.0966492024450226e-06, + "loss": 0.896, + "step": 5545 + }, + { + "epoch": 0.500157821166073, + "grad_norm": 2.0594978446384236, + "learning_rate": 2.0960656702443545e-06, + "loss": 1.1044, + "step": 5546 + }, + { + "epoch": 0.5002480046895432, + "grad_norm": 1.5476529255676768, + "learning_rate": 2.0954821298468343e-06, + "loss": 0.9543, + "step": 5547 + }, + { + "epoch": 0.5003381882130135, + "grad_norm": 1.4392898032134729, + "learning_rate": 2.0948985813022513e-06, + "loss": 1.0203, + "step": 5548 + }, + { + "epoch": 0.5004283717364837, + "grad_norm": 1.4575356436243, + "learning_rate": 2.094315024660399e-06, + "loss": 0.994, + "step": 5549 + }, + { + "epoch": 0.500518555259954, + "grad_norm": 1.6368258857714284, + "learning_rate": 2.0937314599710676e-06, + "loss": 0.9841, + "step": 5550 + }, + { + "epoch": 0.5006087387834243, + "grad_norm": 1.8377847671427132, + "learning_rate": 2.0931478872840526e-06, + "loss": 0.949, + "step": 5551 + }, + { + "epoch": 0.5006989223068945, + "grad_norm": 1.7237390415211231, + "learning_rate": 2.092564306649145e-06, + "loss": 0.9378, + "step": 5552 + }, + { + "epoch": 0.5007891058303648, + "grad_norm": 1.40479372702786, + "learning_rate": 2.091980718116141e-06, + "loss": 0.9942, + "step": 5553 + }, + { + "epoch": 0.5008792893538351, + "grad_norm": 1.5327213769672006, + "learning_rate": 2.091397121734835e-06, + "loss": 0.9952, + "step": 5554 + }, + { + "epoch": 0.5009694728773053, + "grad_norm": 1.7516301600568682, + "learning_rate": 2.090813517555022e-06, + "loss": 0.9856, + "step": 5555 + }, + { + "epoch": 0.5010596564007755, + "grad_norm": 1.652476781084662, + "learning_rate": 2.0902299056265e-06, + "loss": 0.9017, + "step": 5556 + }, + { + "epoch": 0.5011498399242459, + "grad_norm": 1.867865531108561, + "learning_rate": 2.0896462859990643e-06, + "loss": 0.9428, + "step": 5557 + }, + { + "epoch": 0.5012400234477161, + "grad_norm": 1.4122801871292052, + "learning_rate": 2.089062658722513e-06, + "loss": 1.018, + "step": 5558 + }, + { + "epoch": 0.5013302069711864, + "grad_norm": 1.6106298794892677, + "learning_rate": 2.0884790238466452e-06, + "loss": 1.0062, + "step": 5559 + }, + { + "epoch": 0.5014203904946566, + "grad_norm": 1.8224043716647838, + "learning_rate": 2.087895381421259e-06, + "loss": 0.9854, + "step": 5560 + }, + { + "epoch": 0.5015105740181269, + "grad_norm": 2.0829191050394704, + "learning_rate": 2.087311731496154e-06, + "loss": 0.9424, + "step": 5561 + }, + { + "epoch": 0.5016007575415972, + "grad_norm": 1.3171894675315923, + "learning_rate": 2.08672807412113e-06, + "loss": 0.9475, + "step": 5562 + }, + { + "epoch": 0.5016909410650674, + "grad_norm": 1.9050969873604862, + "learning_rate": 2.08614440934599e-06, + "loss": 1.0486, + "step": 5563 + }, + { + "epoch": 0.5017811245885376, + "grad_norm": 1.6371198722559244, + "learning_rate": 2.0855607372205337e-06, + "loss": 0.8745, + "step": 5564 + }, + { + "epoch": 0.501871308112008, + "grad_norm": 1.611556417303199, + "learning_rate": 2.0849770577945623e-06, + "loss": 0.9789, + "step": 5565 + }, + { + "epoch": 0.5019614916354782, + "grad_norm": 1.3198775077107412, + "learning_rate": 2.084393371117881e-06, + "loss": 0.9923, + "step": 5566 + }, + { + "epoch": 0.5020516751589484, + "grad_norm": 1.3512216921089932, + "learning_rate": 2.0838096772402902e-06, + "loss": 0.8922, + "step": 5567 + }, + { + "epoch": 0.5021418586824187, + "grad_norm": 1.7200968233845522, + "learning_rate": 2.0832259762115973e-06, + "loss": 0.8485, + "step": 5568 + }, + { + "epoch": 0.502232042205889, + "grad_norm": 1.3671165665890201, + "learning_rate": 2.082642268081605e-06, + "loss": 0.9882, + "step": 5569 + }, + { + "epoch": 0.5023222257293593, + "grad_norm": 1.8686952767923677, + "learning_rate": 2.082058552900118e-06, + "loss": 0.9118, + "step": 5570 + }, + { + "epoch": 0.5024124092528295, + "grad_norm": 1.5683186336940564, + "learning_rate": 2.081474830716944e-06, + "loss": 0.9367, + "step": 5571 + }, + { + "epoch": 0.5025025927762997, + "grad_norm": 1.9621895368370763, + "learning_rate": 2.080891101581887e-06, + "loss": 0.9907, + "step": 5572 + }, + { + "epoch": 0.5025927762997701, + "grad_norm": 1.8023567506795015, + "learning_rate": 2.080307365544755e-06, + "loss": 1.0015, + "step": 5573 + }, + { + "epoch": 0.5026829598232403, + "grad_norm": 1.6154269157886767, + "learning_rate": 2.0797236226553567e-06, + "loss": 0.9983, + "step": 5574 + }, + { + "epoch": 0.5027731433467105, + "grad_norm": 1.5545518518820205, + "learning_rate": 2.079139872963499e-06, + "loss": 0.899, + "step": 5575 + }, + { + "epoch": 0.5028633268701809, + "grad_norm": 1.839677677636989, + "learning_rate": 2.078556116518991e-06, + "loss": 0.8603, + "step": 5576 + }, + { + "epoch": 0.5029535103936511, + "grad_norm": 1.3640271526358068, + "learning_rate": 2.077972353371642e-06, + "loss": 0.931, + "step": 5577 + }, + { + "epoch": 0.5030436939171213, + "grad_norm": 1.4404389565379738, + "learning_rate": 2.077388583571262e-06, + "loss": 0.9069, + "step": 5578 + }, + { + "epoch": 0.5031338774405916, + "grad_norm": 0.7565329061941735, + "learning_rate": 2.0768048071676608e-06, + "loss": 0.7153, + "step": 5579 + }, + { + "epoch": 0.5032240609640619, + "grad_norm": 1.768740041320587, + "learning_rate": 2.0762210242106505e-06, + "loss": 0.9188, + "step": 5580 + }, + { + "epoch": 0.5033142444875321, + "grad_norm": 1.66518260527246, + "learning_rate": 2.0756372347500424e-06, + "loss": 0.9715, + "step": 5581 + }, + { + "epoch": 0.5034044280110024, + "grad_norm": 1.531912601901261, + "learning_rate": 2.0750534388356473e-06, + "loss": 1.0635, + "step": 5582 + }, + { + "epoch": 0.5034946115344726, + "grad_norm": 1.5769836528658914, + "learning_rate": 2.07446963651728e-06, + "loss": 0.9045, + "step": 5583 + }, + { + "epoch": 0.503584795057943, + "grad_norm": 1.4883188958800935, + "learning_rate": 2.0738858278447516e-06, + "loss": 0.8937, + "step": 5584 + }, + { + "epoch": 0.5036749785814132, + "grad_norm": 1.467156621346013, + "learning_rate": 2.073302012867878e-06, + "loss": 1.0464, + "step": 5585 + }, + { + "epoch": 0.5037651621048834, + "grad_norm": 1.4282757401993895, + "learning_rate": 2.0727181916364725e-06, + "loss": 0.9941, + "step": 5586 + }, + { + "epoch": 0.5038553456283537, + "grad_norm": 1.4224923439020973, + "learning_rate": 2.0721343642003493e-06, + "loss": 0.994, + "step": 5587 + }, + { + "epoch": 0.503945529151824, + "grad_norm": 1.6890448104397893, + "learning_rate": 2.0715505306093247e-06, + "loss": 0.9847, + "step": 5588 + }, + { + "epoch": 0.5040357126752942, + "grad_norm": 1.8648177555415588, + "learning_rate": 2.070966690913214e-06, + "loss": 0.9918, + "step": 5589 + }, + { + "epoch": 0.5041258961987645, + "grad_norm": 2.086013049639868, + "learning_rate": 2.0703828451618346e-06, + "loss": 0.8414, + "step": 5590 + }, + { + "epoch": 0.5042160797222347, + "grad_norm": 1.3073618541989958, + "learning_rate": 2.069798993405002e-06, + "loss": 0.9525, + "step": 5591 + }, + { + "epoch": 0.504306263245705, + "grad_norm": 1.458566220171329, + "learning_rate": 2.0692151356925345e-06, + "loss": 1.0041, + "step": 5592 + }, + { + "epoch": 0.5043964467691753, + "grad_norm": 1.5497279205250931, + "learning_rate": 2.068631272074251e-06, + "loss": 0.9809, + "step": 5593 + }, + { + "epoch": 0.5044866302926455, + "grad_norm": 1.5230308576878009, + "learning_rate": 2.0680474025999676e-06, + "loss": 1.0893, + "step": 5594 + }, + { + "epoch": 0.5045768138161157, + "grad_norm": 0.9211744782691932, + "learning_rate": 2.0674635273195055e-06, + "loss": 0.8328, + "step": 5595 + }, + { + "epoch": 0.5046669973395861, + "grad_norm": 1.7684537120452795, + "learning_rate": 2.066879646282682e-06, + "loss": 0.9578, + "step": 5596 + }, + { + "epoch": 0.5047571808630563, + "grad_norm": 1.8063071773151405, + "learning_rate": 2.0662957595393194e-06, + "loss": 0.9307, + "step": 5597 + }, + { + "epoch": 0.5048473643865266, + "grad_norm": 2.059186643747735, + "learning_rate": 2.0657118671392373e-06, + "loss": 0.9863, + "step": 5598 + }, + { + "epoch": 0.5049375479099969, + "grad_norm": 1.6749520195309215, + "learning_rate": 2.0651279691322558e-06, + "loss": 1.0164, + "step": 5599 + }, + { + "epoch": 0.5050277314334671, + "grad_norm": 1.4739417983926055, + "learning_rate": 2.0645440655681973e-06, + "loss": 0.9136, + "step": 5600 + }, + { + "epoch": 0.5051179149569374, + "grad_norm": 2.922610756347426, + "learning_rate": 2.0639601564968826e-06, + "loss": 1.0273, + "step": 5601 + }, + { + "epoch": 0.5052080984804076, + "grad_norm": 0.7848927829196299, + "learning_rate": 2.0633762419681355e-06, + "loss": 0.8685, + "step": 5602 + }, + { + "epoch": 0.5052982820038779, + "grad_norm": 1.1657208400730386, + "learning_rate": 2.062792322031777e-06, + "loss": 0.978, + "step": 5603 + }, + { + "epoch": 0.5053884655273482, + "grad_norm": 1.4881870510297215, + "learning_rate": 2.062208396737632e-06, + "loss": 0.9146, + "step": 5604 + }, + { + "epoch": 0.5054786490508184, + "grad_norm": 1.4186749341826597, + "learning_rate": 2.0616244661355235e-06, + "loss": 0.9799, + "step": 5605 + }, + { + "epoch": 0.5055688325742886, + "grad_norm": 1.5878684780176398, + "learning_rate": 2.0610405302752752e-06, + "loss": 0.9929, + "step": 5606 + }, + { + "epoch": 0.505659016097759, + "grad_norm": 1.5095131725938178, + "learning_rate": 2.060456589206713e-06, + "loss": 0.9543, + "step": 5607 + }, + { + "epoch": 0.5057491996212292, + "grad_norm": 1.634896732300504, + "learning_rate": 2.0598726429796614e-06, + "loss": 0.9088, + "step": 5608 + }, + { + "epoch": 0.5058393831446995, + "grad_norm": 2.0522749628999395, + "learning_rate": 2.059288691643945e-06, + "loss": 0.922, + "step": 5609 + }, + { + "epoch": 0.5059295666681697, + "grad_norm": 1.6697018483977086, + "learning_rate": 2.0587047352493913e-06, + "loss": 1.0427, + "step": 5610 + }, + { + "epoch": 0.50601975019164, + "grad_norm": 1.7188511211819162, + "learning_rate": 2.0581207738458248e-06, + "loss": 0.9364, + "step": 5611 + }, + { + "epoch": 0.5061099337151103, + "grad_norm": 2.116952693066474, + "learning_rate": 2.0575368074830743e-06, + "loss": 0.9185, + "step": 5612 + }, + { + "epoch": 0.5062001172385805, + "grad_norm": 1.5334336417324363, + "learning_rate": 2.0569528362109667e-06, + "loss": 0.8751, + "step": 5613 + }, + { + "epoch": 0.5062903007620507, + "grad_norm": 1.4243392550179224, + "learning_rate": 2.056368860079327e-06, + "loss": 0.9587, + "step": 5614 + }, + { + "epoch": 0.5063804842855211, + "grad_norm": 1.832428911399302, + "learning_rate": 2.0557848791379874e-06, + "loss": 0.9035, + "step": 5615 + }, + { + "epoch": 0.5064706678089913, + "grad_norm": 1.4834066344305292, + "learning_rate": 2.0552008934367734e-06, + "loss": 0.8629, + "step": 5616 + }, + { + "epoch": 0.5065608513324615, + "grad_norm": 1.7262849886785325, + "learning_rate": 2.0546169030255154e-06, + "loss": 1.0372, + "step": 5617 + }, + { + "epoch": 0.5066510348559318, + "grad_norm": 1.6222417604163972, + "learning_rate": 2.054032907954041e-06, + "loss": 1.0323, + "step": 5618 + }, + { + "epoch": 0.5067412183794021, + "grad_norm": 1.6806851285404405, + "learning_rate": 2.053448908272182e-06, + "loss": 0.9267, + "step": 5619 + }, + { + "epoch": 0.5068314019028723, + "grad_norm": 1.6087972103800363, + "learning_rate": 2.0528649040297673e-06, + "loss": 0.9959, + "step": 5620 + }, + { + "epoch": 0.5069215854263426, + "grad_norm": 1.422496701657405, + "learning_rate": 2.0522808952766266e-06, + "loss": 0.962, + "step": 5621 + }, + { + "epoch": 0.5070117689498129, + "grad_norm": 2.1525930824489747, + "learning_rate": 2.0516968820625925e-06, + "loss": 1.0055, + "step": 5622 + }, + { + "epoch": 0.5071019524732832, + "grad_norm": 1.67506418247522, + "learning_rate": 2.051112864437495e-06, + "loss": 0.9474, + "step": 5623 + }, + { + "epoch": 0.5071921359967534, + "grad_norm": 1.8894203698972707, + "learning_rate": 2.050528842451166e-06, + "loss": 0.9278, + "step": 5624 + }, + { + "epoch": 0.5072823195202236, + "grad_norm": 1.3862008000578612, + "learning_rate": 2.049944816153438e-06, + "loss": 0.8374, + "step": 5625 + }, + { + "epoch": 0.507372503043694, + "grad_norm": 1.4716033004447922, + "learning_rate": 2.049360785594142e-06, + "loss": 1.0016, + "step": 5626 + }, + { + "epoch": 0.5074626865671642, + "grad_norm": 0.692232554155626, + "learning_rate": 2.048776750823113e-06, + "loss": 0.7119, + "step": 5627 + }, + { + "epoch": 0.5075528700906344, + "grad_norm": 0.6632488427035246, + "learning_rate": 2.0481927118901817e-06, + "loss": 0.765, + "step": 5628 + }, + { + "epoch": 0.5076430536141047, + "grad_norm": 1.5847250527258085, + "learning_rate": 2.0476086688451824e-06, + "loss": 0.9013, + "step": 5629 + }, + { + "epoch": 0.507733237137575, + "grad_norm": 1.5293897110980168, + "learning_rate": 2.04702462173795e-06, + "loss": 0.9516, + "step": 5630 + }, + { + "epoch": 0.5078234206610452, + "grad_norm": 1.9176496251900281, + "learning_rate": 2.0464405706183167e-06, + "loss": 0.9422, + "step": 5631 + }, + { + "epoch": 0.5079136041845155, + "grad_norm": 1.5249865359352965, + "learning_rate": 2.045856515536118e-06, + "loss": 0.9757, + "step": 5632 + }, + { + "epoch": 0.5080037877079857, + "grad_norm": 1.6531551163639715, + "learning_rate": 2.045272456541188e-06, + "loss": 0.9774, + "step": 5633 + }, + { + "epoch": 0.508093971231456, + "grad_norm": 1.47416767300747, + "learning_rate": 2.0446883936833635e-06, + "loss": 1.0285, + "step": 5634 + }, + { + "epoch": 0.5081841547549263, + "grad_norm": 1.4477385084288674, + "learning_rate": 2.0441043270124782e-06, + "loss": 0.932, + "step": 5635 + }, + { + "epoch": 0.5082743382783965, + "grad_norm": 1.6275270326774751, + "learning_rate": 2.0435202565783683e-06, + "loss": 0.9898, + "step": 5636 + }, + { + "epoch": 0.5083645218018668, + "grad_norm": 1.4719382238071643, + "learning_rate": 2.042936182430871e-06, + "loss": 0.9199, + "step": 5637 + }, + { + "epoch": 0.5084547053253371, + "grad_norm": 1.5547588082833987, + "learning_rate": 2.0423521046198206e-06, + "loss": 0.9677, + "step": 5638 + }, + { + "epoch": 0.5085448888488073, + "grad_norm": 1.393442150723314, + "learning_rate": 2.041768023195056e-06, + "loss": 0.8734, + "step": 5639 + }, + { + "epoch": 0.5086350723722776, + "grad_norm": 1.5072002057119906, + "learning_rate": 2.0411839382064126e-06, + "loss": 1.0301, + "step": 5640 + }, + { + "epoch": 0.5087252558957478, + "grad_norm": 1.585142941933725, + "learning_rate": 2.040599849703729e-06, + "loss": 0.8589, + "step": 5641 + }, + { + "epoch": 0.5088154394192181, + "grad_norm": 1.830017219108076, + "learning_rate": 2.040015757736843e-06, + "loss": 0.8976, + "step": 5642 + }, + { + "epoch": 0.5089056229426884, + "grad_norm": 1.6224452256653916, + "learning_rate": 2.039431662355591e-06, + "loss": 0.8864, + "step": 5643 + }, + { + "epoch": 0.5089958064661586, + "grad_norm": 1.572313770007237, + "learning_rate": 2.0388475636098126e-06, + "loss": 0.9761, + "step": 5644 + }, + { + "epoch": 0.5090859899896288, + "grad_norm": 1.4048855413874164, + "learning_rate": 2.038263461549346e-06, + "loss": 0.9926, + "step": 5645 + }, + { + "epoch": 0.5091761735130992, + "grad_norm": 1.8438941123374784, + "learning_rate": 2.0376793562240297e-06, + "loss": 0.9621, + "step": 5646 + }, + { + "epoch": 0.5092663570365694, + "grad_norm": 1.4931244629722833, + "learning_rate": 2.037095247683703e-06, + "loss": 0.9679, + "step": 5647 + }, + { + "epoch": 0.5093565405600397, + "grad_norm": 1.5895492693383022, + "learning_rate": 2.0365111359782046e-06, + "loss": 0.9153, + "step": 5648 + }, + { + "epoch": 0.50944672408351, + "grad_norm": 1.4810582097145082, + "learning_rate": 2.0359270211573757e-06, + "loss": 0.9881, + "step": 5649 + }, + { + "epoch": 0.5095369076069802, + "grad_norm": 1.7189910719608048, + "learning_rate": 2.0353429032710545e-06, + "loss": 1.0275, + "step": 5650 + }, + { + "epoch": 0.5096270911304505, + "grad_norm": 1.3437678757299152, + "learning_rate": 2.0347587823690825e-06, + "loss": 0.933, + "step": 5651 + }, + { + "epoch": 0.5097172746539207, + "grad_norm": 1.41645761892229, + "learning_rate": 2.034174658501299e-06, + "loss": 0.9562, + "step": 5652 + }, + { + "epoch": 0.509807458177391, + "grad_norm": 1.3924967138775552, + "learning_rate": 2.0335905317175453e-06, + "loss": 0.9933, + "step": 5653 + }, + { + "epoch": 0.5098976417008613, + "grad_norm": 2.189313844985895, + "learning_rate": 2.033006402067663e-06, + "loss": 0.9621, + "step": 5654 + }, + { + "epoch": 0.5099878252243315, + "grad_norm": 1.5137457454253052, + "learning_rate": 2.0324222696014912e-06, + "loss": 0.8615, + "step": 5655 + }, + { + "epoch": 0.5100780087478017, + "grad_norm": 1.477827549496979, + "learning_rate": 2.0318381343688733e-06, + "loss": 0.9942, + "step": 5656 + }, + { + "epoch": 0.5101681922712721, + "grad_norm": 1.803725609670631, + "learning_rate": 2.0312539964196505e-06, + "loss": 0.9398, + "step": 5657 + }, + { + "epoch": 0.5102583757947423, + "grad_norm": 1.4480062589668181, + "learning_rate": 2.030669855803664e-06, + "loss": 0.9368, + "step": 5658 + }, + { + "epoch": 0.5103485593182125, + "grad_norm": 1.4148489098373178, + "learning_rate": 2.0300857125707563e-06, + "loss": 1.0071, + "step": 5659 + }, + { + "epoch": 0.5104387428416828, + "grad_norm": 1.8435944313754806, + "learning_rate": 2.0295015667707697e-06, + "loss": 0.9903, + "step": 5660 + }, + { + "epoch": 0.5105289263651531, + "grad_norm": 1.4701405699334944, + "learning_rate": 2.0289174184535472e-06, + "loss": 0.9657, + "step": 5661 + }, + { + "epoch": 0.5106191098886234, + "grad_norm": 1.8583488517877358, + "learning_rate": 2.02833326766893e-06, + "loss": 0.9314, + "step": 5662 + }, + { + "epoch": 0.5107092934120936, + "grad_norm": 1.6490639314397686, + "learning_rate": 2.027749114466763e-06, + "loss": 1.0268, + "step": 5663 + }, + { + "epoch": 0.5107994769355638, + "grad_norm": 1.5608349129661527, + "learning_rate": 2.027164958896889e-06, + "loss": 1.001, + "step": 5664 + }, + { + "epoch": 0.5108896604590342, + "grad_norm": 1.8878810295873016, + "learning_rate": 2.02658080100915e-06, + "loss": 1.0781, + "step": 5665 + }, + { + "epoch": 0.5109798439825044, + "grad_norm": 1.3151311840140043, + "learning_rate": 2.0259966408533915e-06, + "loss": 1.0344, + "step": 5666 + }, + { + "epoch": 0.5110700275059746, + "grad_norm": 1.5068070612258664, + "learning_rate": 2.025412478479455e-06, + "loss": 1.0475, + "step": 5667 + }, + { + "epoch": 0.5111602110294449, + "grad_norm": 2.0935778191213235, + "learning_rate": 2.0248283139371862e-06, + "loss": 1.0051, + "step": 5668 + }, + { + "epoch": 0.5112503945529152, + "grad_norm": 1.7211743682667349, + "learning_rate": 2.024244147276429e-06, + "loss": 0.9322, + "step": 5669 + }, + { + "epoch": 0.5113405780763854, + "grad_norm": 1.305896541467082, + "learning_rate": 2.023659978547027e-06, + "loss": 1.0264, + "step": 5670 + }, + { + "epoch": 0.5114307615998557, + "grad_norm": 1.4082994573608212, + "learning_rate": 2.023075807798826e-06, + "loss": 0.9604, + "step": 5671 + }, + { + "epoch": 0.511520945123326, + "grad_norm": 1.4234757687295405, + "learning_rate": 2.0224916350816696e-06, + "loss": 0.8858, + "step": 5672 + }, + { + "epoch": 0.5116111286467963, + "grad_norm": 1.356706257733731, + "learning_rate": 2.0219074604454026e-06, + "loss": 0.9835, + "step": 5673 + }, + { + "epoch": 0.5117013121702665, + "grad_norm": 1.8065571533715248, + "learning_rate": 2.02132328393987e-06, + "loss": 0.9667, + "step": 5674 + }, + { + "epoch": 0.5117914956937367, + "grad_norm": 1.5487346837454072, + "learning_rate": 2.0207391056149174e-06, + "loss": 0.9134, + "step": 5675 + }, + { + "epoch": 0.5118816792172071, + "grad_norm": 1.668260249589589, + "learning_rate": 2.020154925520391e-06, + "loss": 0.9362, + "step": 5676 + }, + { + "epoch": 0.5119718627406773, + "grad_norm": 1.549013759268719, + "learning_rate": 2.0195707437061332e-06, + "loss": 1.0581, + "step": 5677 + }, + { + "epoch": 0.5120620462641475, + "grad_norm": 2.184112760113168, + "learning_rate": 2.0189865602219934e-06, + "loss": 0.9632, + "step": 5678 + }, + { + "epoch": 0.5121522297876178, + "grad_norm": 1.5331231163190613, + "learning_rate": 2.0184023751178154e-06, + "loss": 1.0333, + "step": 5679 + }, + { + "epoch": 0.5122424133110881, + "grad_norm": 1.3453984462385986, + "learning_rate": 2.017818188443444e-06, + "loss": 0.9136, + "step": 5680 + }, + { + "epoch": 0.5123325968345583, + "grad_norm": 1.6657514284400592, + "learning_rate": 2.017234000248728e-06, + "loss": 0.9694, + "step": 5681 + }, + { + "epoch": 0.5124227803580286, + "grad_norm": 1.4984622066239028, + "learning_rate": 2.0166498105835108e-06, + "loss": 0.8995, + "step": 5682 + }, + { + "epoch": 0.5125129638814988, + "grad_norm": 1.3107795792988601, + "learning_rate": 2.0160656194976407e-06, + "loss": 0.8931, + "step": 5683 + }, + { + "epoch": 0.5126031474049692, + "grad_norm": 1.490957102377859, + "learning_rate": 2.0154814270409634e-06, + "loss": 0.9139, + "step": 5684 + }, + { + "epoch": 0.5126933309284394, + "grad_norm": 1.7774348308994292, + "learning_rate": 2.0148972332633247e-06, + "loss": 0.9214, + "step": 5685 + }, + { + "epoch": 0.5127835144519096, + "grad_norm": 2.1603896720276325, + "learning_rate": 2.0143130382145733e-06, + "loss": 1.0213, + "step": 5686 + }, + { + "epoch": 0.5128736979753798, + "grad_norm": 1.7134927280784873, + "learning_rate": 2.0137288419445533e-06, + "loss": 0.9317, + "step": 5687 + }, + { + "epoch": 0.5129638814988502, + "grad_norm": 1.6566599932228525, + "learning_rate": 2.0131446445031134e-06, + "loss": 1.0018, + "step": 5688 + }, + { + "epoch": 0.5130540650223204, + "grad_norm": 1.7259400225193888, + "learning_rate": 2.0125604459400994e-06, + "loss": 1.0441, + "step": 5689 + }, + { + "epoch": 0.5131442485457907, + "grad_norm": 2.045509057330078, + "learning_rate": 2.0119762463053596e-06, + "loss": 1.0169, + "step": 5690 + }, + { + "epoch": 0.5132344320692609, + "grad_norm": 1.5335997633347307, + "learning_rate": 2.0113920456487406e-06, + "loss": 0.9532, + "step": 5691 + }, + { + "epoch": 0.5133246155927312, + "grad_norm": 1.4905627127695273, + "learning_rate": 2.010807844020088e-06, + "loss": 0.9542, + "step": 5692 + }, + { + "epoch": 0.5134147991162015, + "grad_norm": 1.4915318671469013, + "learning_rate": 2.0102236414692524e-06, + "loss": 1.013, + "step": 5693 + }, + { + "epoch": 0.5135049826396717, + "grad_norm": 1.4397093339129277, + "learning_rate": 2.0096394380460777e-06, + "loss": 0.9513, + "step": 5694 + }, + { + "epoch": 0.513595166163142, + "grad_norm": 1.5057620797595066, + "learning_rate": 2.0090552338004136e-06, + "loss": 0.9907, + "step": 5695 + }, + { + "epoch": 0.5136853496866123, + "grad_norm": 1.400534863139857, + "learning_rate": 2.0084710287821077e-06, + "loss": 0.9609, + "step": 5696 + }, + { + "epoch": 0.5137755332100825, + "grad_norm": 1.6803066797991044, + "learning_rate": 2.007886823041006e-06, + "loss": 1.001, + "step": 5697 + }, + { + "epoch": 0.5138657167335527, + "grad_norm": 1.595477794486716, + "learning_rate": 2.0073026166269577e-06, + "loss": 0.9781, + "step": 5698 + }, + { + "epoch": 0.5139559002570231, + "grad_norm": 1.4005428641125603, + "learning_rate": 2.0067184095898093e-06, + "loss": 1.0188, + "step": 5699 + }, + { + "epoch": 0.5140460837804933, + "grad_norm": 1.5960906549200595, + "learning_rate": 2.0061342019794094e-06, + "loss": 0.9732, + "step": 5700 + }, + { + "epoch": 0.5141362673039636, + "grad_norm": 2.0454905246691393, + "learning_rate": 2.0055499938456058e-06, + "loss": 1.0358, + "step": 5701 + }, + { + "epoch": 0.5142264508274338, + "grad_norm": 1.5181917302623658, + "learning_rate": 2.0049657852382464e-06, + "loss": 0.9308, + "step": 5702 + }, + { + "epoch": 0.5143166343509041, + "grad_norm": 1.9823986033309253, + "learning_rate": 2.0043815762071782e-06, + "loss": 0.8895, + "step": 5703 + }, + { + "epoch": 0.5144068178743744, + "grad_norm": 2.2704276909586496, + "learning_rate": 2.0037973668022492e-06, + "loss": 0.9305, + "step": 5704 + }, + { + "epoch": 0.5144970013978446, + "grad_norm": 1.7320810217066134, + "learning_rate": 2.003213157073309e-06, + "loss": 0.8391, + "step": 5705 + }, + { + "epoch": 0.5145871849213148, + "grad_norm": 1.6255778239032919, + "learning_rate": 2.002628947070204e-06, + "loss": 0.9582, + "step": 5706 + }, + { + "epoch": 0.5146773684447852, + "grad_norm": 1.703371432612986, + "learning_rate": 2.002044736842783e-06, + "loss": 1.0232, + "step": 5707 + }, + { + "epoch": 0.5147675519682554, + "grad_norm": 1.4379872035490444, + "learning_rate": 2.001460526440894e-06, + "loss": 0.9425, + "step": 5708 + }, + { + "epoch": 0.5148577354917256, + "grad_norm": 0.7119452491909835, + "learning_rate": 2.0008763159143843e-06, + "loss": 0.7827, + "step": 5709 + }, + { + "epoch": 0.5149479190151959, + "grad_norm": 1.4909136062911057, + "learning_rate": 2.000292105313103e-06, + "loss": 0.9872, + "step": 5710 + }, + { + "epoch": 0.5150381025386662, + "grad_norm": 1.4650524753637524, + "learning_rate": 1.999707894686897e-06, + "loss": 0.998, + "step": 5711 + }, + { + "epoch": 0.5151282860621365, + "grad_norm": 1.296429293233347, + "learning_rate": 1.9991236840856155e-06, + "loss": 0.9521, + "step": 5712 + }, + { + "epoch": 0.5152184695856067, + "grad_norm": 1.4914111769261615, + "learning_rate": 1.9985394735591065e-06, + "loss": 0.9349, + "step": 5713 + }, + { + "epoch": 0.5153086531090769, + "grad_norm": 1.8592770975448258, + "learning_rate": 1.997955263157217e-06, + "loss": 1.0078, + "step": 5714 + }, + { + "epoch": 0.5153988366325473, + "grad_norm": 1.5086751259063578, + "learning_rate": 1.997371052929796e-06, + "loss": 0.9949, + "step": 5715 + }, + { + "epoch": 0.5154890201560175, + "grad_norm": 1.8633806693946051, + "learning_rate": 1.996786842926691e-06, + "loss": 0.9193, + "step": 5716 + }, + { + "epoch": 0.5155792036794877, + "grad_norm": 1.796273238550062, + "learning_rate": 1.9962026331977506e-06, + "loss": 0.9718, + "step": 5717 + }, + { + "epoch": 0.5156693872029581, + "grad_norm": 1.5161104506314156, + "learning_rate": 1.9956184237928224e-06, + "loss": 0.9763, + "step": 5718 + }, + { + "epoch": 0.5157595707264283, + "grad_norm": 1.6143004959288112, + "learning_rate": 1.995034214761754e-06, + "loss": 0.9521, + "step": 5719 + }, + { + "epoch": 0.5158497542498985, + "grad_norm": 1.273148673600111, + "learning_rate": 1.9944500061543945e-06, + "loss": 0.9362, + "step": 5720 + }, + { + "epoch": 0.5159399377733688, + "grad_norm": 1.6073015059019369, + "learning_rate": 1.99386579802059e-06, + "loss": 0.9861, + "step": 5721 + }, + { + "epoch": 0.5160301212968391, + "grad_norm": 1.7756167562602874, + "learning_rate": 1.993281590410191e-06, + "loss": 0.9403, + "step": 5722 + }, + { + "epoch": 0.5161203048203094, + "grad_norm": 1.4465673824086989, + "learning_rate": 1.992697383373043e-06, + "loss": 0.9132, + "step": 5723 + }, + { + "epoch": 0.5162104883437796, + "grad_norm": 0.8396304192287923, + "learning_rate": 1.9921131769589937e-06, + "loss": 0.8234, + "step": 5724 + }, + { + "epoch": 0.5163006718672498, + "grad_norm": 1.6152198357256382, + "learning_rate": 1.991528971217893e-06, + "loss": 0.9599, + "step": 5725 + }, + { + "epoch": 0.5163908553907202, + "grad_norm": 1.3474813596607482, + "learning_rate": 1.9909447661995858e-06, + "loss": 0.9587, + "step": 5726 + }, + { + "epoch": 0.5164810389141904, + "grad_norm": 1.7475738737893733, + "learning_rate": 1.990360561953922e-06, + "loss": 0.9803, + "step": 5727 + }, + { + "epoch": 0.5165712224376606, + "grad_norm": 1.3491311226526221, + "learning_rate": 1.9897763585307483e-06, + "loss": 0.9847, + "step": 5728 + }, + { + "epoch": 0.5166614059611309, + "grad_norm": 1.4365206990063548, + "learning_rate": 1.989192155979912e-06, + "loss": 1.0007, + "step": 5729 + }, + { + "epoch": 0.5167515894846012, + "grad_norm": 1.3583049728303649, + "learning_rate": 1.98860795435126e-06, + "loss": 0.9955, + "step": 5730 + }, + { + "epoch": 0.5168417730080714, + "grad_norm": 1.4451383743558146, + "learning_rate": 1.9880237536946406e-06, + "loss": 0.9686, + "step": 5731 + }, + { + "epoch": 0.5169319565315417, + "grad_norm": 1.6568457053986054, + "learning_rate": 1.987439554059901e-06, + "loss": 0.9347, + "step": 5732 + }, + { + "epoch": 0.5170221400550119, + "grad_norm": 1.7027604299093395, + "learning_rate": 1.9868553554968864e-06, + "loss": 0.9359, + "step": 5733 + }, + { + "epoch": 0.5171123235784822, + "grad_norm": 1.7822691026601103, + "learning_rate": 1.986271158055447e-06, + "loss": 0.91, + "step": 5734 + }, + { + "epoch": 0.5172025071019525, + "grad_norm": 1.5048761584254253, + "learning_rate": 1.9856869617854273e-06, + "loss": 0.9548, + "step": 5735 + }, + { + "epoch": 0.5172926906254227, + "grad_norm": 1.6105253682168195, + "learning_rate": 1.9851027667366746e-06, + "loss": 0.9403, + "step": 5736 + }, + { + "epoch": 0.517382874148893, + "grad_norm": 1.4695473799360235, + "learning_rate": 1.984518572959037e-06, + "loss": 1.0577, + "step": 5737 + }, + { + "epoch": 0.5174730576723633, + "grad_norm": 1.6601574707026763, + "learning_rate": 1.9839343805023587e-06, + "loss": 0.8823, + "step": 5738 + }, + { + "epoch": 0.5175632411958335, + "grad_norm": 1.6735836104177646, + "learning_rate": 1.9833501894164886e-06, + "loss": 1.0542, + "step": 5739 + }, + { + "epoch": 0.5176534247193038, + "grad_norm": 1.59749850587969, + "learning_rate": 1.982765999751273e-06, + "loss": 0.9982, + "step": 5740 + }, + { + "epoch": 0.5177436082427741, + "grad_norm": 1.6886375972631744, + "learning_rate": 1.9821818115565553e-06, + "loss": 0.9943, + "step": 5741 + }, + { + "epoch": 0.5178337917662443, + "grad_norm": 1.7896620899259625, + "learning_rate": 1.9815976248821853e-06, + "loss": 0.8728, + "step": 5742 + }, + { + "epoch": 0.5179239752897146, + "grad_norm": 1.4500772652922267, + "learning_rate": 1.981013439778007e-06, + "loss": 0.9761, + "step": 5743 + }, + { + "epoch": 0.5180141588131848, + "grad_norm": 1.6201413953557988, + "learning_rate": 1.9804292562938666e-06, + "loss": 0.9741, + "step": 5744 + }, + { + "epoch": 0.5181043423366551, + "grad_norm": 1.515412541865365, + "learning_rate": 1.97984507447961e-06, + "loss": 1.0019, + "step": 5745 + }, + { + "epoch": 0.5181945258601254, + "grad_norm": 1.6749251163881365, + "learning_rate": 1.9792608943850824e-06, + "loss": 0.9039, + "step": 5746 + }, + { + "epoch": 0.5182847093835956, + "grad_norm": 1.4814114349857102, + "learning_rate": 1.9786767160601305e-06, + "loss": 0.9394, + "step": 5747 + }, + { + "epoch": 0.5183748929070658, + "grad_norm": 1.5798265707224926, + "learning_rate": 1.9780925395545977e-06, + "loss": 0.896, + "step": 5748 + }, + { + "epoch": 0.5184650764305362, + "grad_norm": 0.6884242910467716, + "learning_rate": 1.9775083649183306e-06, + "loss": 0.768, + "step": 5749 + }, + { + "epoch": 0.5185552599540064, + "grad_norm": 1.638679877205953, + "learning_rate": 1.976924192201174e-06, + "loss": 1.0037, + "step": 5750 + }, + { + "epoch": 0.5186454434774767, + "grad_norm": 1.4853645640148434, + "learning_rate": 1.9763400214529723e-06, + "loss": 0.891, + "step": 5751 + }, + { + "epoch": 0.5187356270009469, + "grad_norm": 1.8989032440927645, + "learning_rate": 1.9757558527235713e-06, + "loss": 0.8322, + "step": 5752 + }, + { + "epoch": 0.5188258105244172, + "grad_norm": 1.5732507498263661, + "learning_rate": 1.9751716860628136e-06, + "loss": 0.9485, + "step": 5753 + }, + { + "epoch": 0.5189159940478875, + "grad_norm": 1.750630469512839, + "learning_rate": 1.974587521520545e-06, + "loss": 0.989, + "step": 5754 + }, + { + "epoch": 0.5190061775713577, + "grad_norm": 1.4316866712464127, + "learning_rate": 1.9740033591466088e-06, + "loss": 0.9973, + "step": 5755 + }, + { + "epoch": 0.5190963610948279, + "grad_norm": 1.764282458828007, + "learning_rate": 1.97341919899085e-06, + "loss": 1.071, + "step": 5756 + }, + { + "epoch": 0.5191865446182983, + "grad_norm": 1.7213917630505247, + "learning_rate": 1.9728350411031114e-06, + "loss": 0.9864, + "step": 5757 + }, + { + "epoch": 0.5192767281417685, + "grad_norm": 1.92570925688279, + "learning_rate": 1.9722508855332367e-06, + "loss": 1.0363, + "step": 5758 + }, + { + "epoch": 0.5193669116652387, + "grad_norm": 1.4165613845898226, + "learning_rate": 1.97166673233107e-06, + "loss": 0.8906, + "step": 5759 + }, + { + "epoch": 0.519457095188709, + "grad_norm": 1.3491893948452314, + "learning_rate": 1.971082581546453e-06, + "loss": 1.0056, + "step": 5760 + }, + { + "epoch": 0.5195472787121793, + "grad_norm": 1.6592600119884868, + "learning_rate": 1.9704984332292306e-06, + "loss": 0.9868, + "step": 5761 + }, + { + "epoch": 0.5196374622356495, + "grad_norm": 1.5350178663475236, + "learning_rate": 1.9699142874292444e-06, + "loss": 1.0495, + "step": 5762 + }, + { + "epoch": 0.5197276457591198, + "grad_norm": 1.585125419275514, + "learning_rate": 1.969330144196336e-06, + "loss": 1.0525, + "step": 5763 + }, + { + "epoch": 0.51981782928259, + "grad_norm": 1.8122122305945378, + "learning_rate": 1.9687460035803497e-06, + "loss": 0.9038, + "step": 5764 + }, + { + "epoch": 0.5199080128060604, + "grad_norm": 2.2462408664456115, + "learning_rate": 1.9681618656311265e-06, + "loss": 0.9463, + "step": 5765 + }, + { + "epoch": 0.5199981963295306, + "grad_norm": 1.36453473633094, + "learning_rate": 1.9675777303985086e-06, + "loss": 0.8739, + "step": 5766 + }, + { + "epoch": 0.5200883798530008, + "grad_norm": 1.5026982199214283, + "learning_rate": 1.9669935979323376e-06, + "loss": 0.9923, + "step": 5767 + }, + { + "epoch": 0.5201785633764712, + "grad_norm": 1.553494709920578, + "learning_rate": 1.9664094682824545e-06, + "loss": 1.0193, + "step": 5768 + }, + { + "epoch": 0.5202687468999414, + "grad_norm": 1.4510363262632324, + "learning_rate": 1.965825341498701e-06, + "loss": 0.9243, + "step": 5769 + }, + { + "epoch": 0.5203589304234116, + "grad_norm": 1.6948746498141871, + "learning_rate": 1.9652412176309177e-06, + "loss": 1.0407, + "step": 5770 + }, + { + "epoch": 0.5204491139468819, + "grad_norm": 1.3360015229170794, + "learning_rate": 1.9646570967289453e-06, + "loss": 0.8278, + "step": 5771 + }, + { + "epoch": 0.5205392974703522, + "grad_norm": 1.6569930515086062, + "learning_rate": 1.9640729788426246e-06, + "loss": 0.9995, + "step": 5772 + }, + { + "epoch": 0.5206294809938224, + "grad_norm": 1.4048247426236085, + "learning_rate": 1.963488864021795e-06, + "loss": 0.9916, + "step": 5773 + }, + { + "epoch": 0.5207196645172927, + "grad_norm": 1.5188661507355077, + "learning_rate": 1.962904752316298e-06, + "loss": 0.8554, + "step": 5774 + }, + { + "epoch": 0.5208098480407629, + "grad_norm": 2.0040207025216215, + "learning_rate": 1.9623206437759706e-06, + "loss": 0.8903, + "step": 5775 + }, + { + "epoch": 0.5209000315642333, + "grad_norm": 1.4437753287071107, + "learning_rate": 1.9617365384506545e-06, + "loss": 0.9263, + "step": 5776 + }, + { + "epoch": 0.5209902150877035, + "grad_norm": 1.4318369565561269, + "learning_rate": 1.9611524363901872e-06, + "loss": 0.9485, + "step": 5777 + }, + { + "epoch": 0.5210803986111737, + "grad_norm": 1.5408633003326055, + "learning_rate": 1.960568337644409e-06, + "loss": 0.9552, + "step": 5778 + }, + { + "epoch": 0.521170582134644, + "grad_norm": 1.4311482622393608, + "learning_rate": 1.9599842422631576e-06, + "loss": 0.945, + "step": 5779 + }, + { + "epoch": 0.5212607656581143, + "grad_norm": 1.5676815330605642, + "learning_rate": 1.9594001502962703e-06, + "loss": 0.9543, + "step": 5780 + }, + { + "epoch": 0.5213509491815845, + "grad_norm": 1.387512109033481, + "learning_rate": 1.9588160617935868e-06, + "loss": 1.0236, + "step": 5781 + }, + { + "epoch": 0.5214411327050548, + "grad_norm": 1.622723451929959, + "learning_rate": 1.958231976804944e-06, + "loss": 1.0487, + "step": 5782 + }, + { + "epoch": 0.521531316228525, + "grad_norm": 1.8232245339681636, + "learning_rate": 1.957647895380179e-06, + "loss": 0.8842, + "step": 5783 + }, + { + "epoch": 0.5216214997519953, + "grad_norm": 1.312841779622529, + "learning_rate": 1.9570638175691297e-06, + "loss": 0.89, + "step": 5784 + }, + { + "epoch": 0.5217116832754656, + "grad_norm": 1.9120734019495038, + "learning_rate": 1.956479743421632e-06, + "loss": 1.0051, + "step": 5785 + }, + { + "epoch": 0.5218018667989358, + "grad_norm": 1.818555384969593, + "learning_rate": 1.955895672987522e-06, + "loss": 0.9037, + "step": 5786 + }, + { + "epoch": 0.521892050322406, + "grad_norm": 1.359614603257995, + "learning_rate": 1.9553116063166367e-06, + "loss": 0.8654, + "step": 5787 + }, + { + "epoch": 0.5219822338458764, + "grad_norm": 1.4526773398994774, + "learning_rate": 1.954727543458812e-06, + "loss": 0.8685, + "step": 5788 + }, + { + "epoch": 0.5220724173693466, + "grad_norm": 1.655743143511259, + "learning_rate": 1.954143484463883e-06, + "loss": 0.993, + "step": 5789 + }, + { + "epoch": 0.5221626008928169, + "grad_norm": 1.9488955671301604, + "learning_rate": 1.9535594293816836e-06, + "loss": 0.9496, + "step": 5790 + }, + { + "epoch": 0.5222527844162872, + "grad_norm": 1.8868237578746005, + "learning_rate": 1.952975378262051e-06, + "loss": 0.9091, + "step": 5791 + }, + { + "epoch": 0.5223429679397574, + "grad_norm": 1.8195215561552434, + "learning_rate": 1.952391331154817e-06, + "loss": 0.911, + "step": 5792 + }, + { + "epoch": 0.5224331514632277, + "grad_norm": 1.5529008116380558, + "learning_rate": 1.9518072881098185e-06, + "loss": 1.0192, + "step": 5793 + }, + { + "epoch": 0.5225233349866979, + "grad_norm": 1.5048675239276523, + "learning_rate": 1.9512232491768867e-06, + "loss": 0.9486, + "step": 5794 + }, + { + "epoch": 0.5226135185101682, + "grad_norm": 1.6115802359589775, + "learning_rate": 1.9506392144058573e-06, + "loss": 0.9525, + "step": 5795 + }, + { + "epoch": 0.5227037020336385, + "grad_norm": 1.9296889208583774, + "learning_rate": 1.9500551838465623e-06, + "loss": 0.9559, + "step": 5796 + }, + { + "epoch": 0.5227938855571087, + "grad_norm": 1.360689514200718, + "learning_rate": 1.9494711575488337e-06, + "loss": 0.9708, + "step": 5797 + }, + { + "epoch": 0.5228840690805789, + "grad_norm": 1.7104878291923937, + "learning_rate": 1.948887135562505e-06, + "loss": 0.9038, + "step": 5798 + }, + { + "epoch": 0.5229742526040493, + "grad_norm": 1.5246692861699358, + "learning_rate": 1.9483031179374074e-06, + "loss": 1.0135, + "step": 5799 + }, + { + "epoch": 0.5230644361275195, + "grad_norm": 1.382208827089438, + "learning_rate": 1.9477191047233736e-06, + "loss": 1.0146, + "step": 5800 + }, + { + "epoch": 0.5231546196509897, + "grad_norm": 1.6429316699260192, + "learning_rate": 1.9471350959702334e-06, + "loss": 0.8404, + "step": 5801 + }, + { + "epoch": 0.52324480317446, + "grad_norm": 1.9249352382318525, + "learning_rate": 1.9465510917278184e-06, + "loss": 0.9652, + "step": 5802 + }, + { + "epoch": 0.5233349866979303, + "grad_norm": 1.5924733602550305, + "learning_rate": 1.9459670920459593e-06, + "loss": 0.9536, + "step": 5803 + }, + { + "epoch": 0.5234251702214006, + "grad_norm": 1.6910971696119954, + "learning_rate": 1.945383096974485e-06, + "loss": 0.8757, + "step": 5804 + }, + { + "epoch": 0.5235153537448708, + "grad_norm": 2.240440831403176, + "learning_rate": 1.944799106563227e-06, + "loss": 0.8826, + "step": 5805 + }, + { + "epoch": 0.523605537268341, + "grad_norm": 1.6864670311662464, + "learning_rate": 1.9442151208620133e-06, + "loss": 0.9971, + "step": 5806 + }, + { + "epoch": 0.5236957207918114, + "grad_norm": 1.4919371869311138, + "learning_rate": 1.943631139920672e-06, + "loss": 1.0031, + "step": 5807 + }, + { + "epoch": 0.5237859043152816, + "grad_norm": 1.6558051321581149, + "learning_rate": 1.943047163789034e-06, + "loss": 0.9092, + "step": 5808 + }, + { + "epoch": 0.5238760878387518, + "grad_norm": 1.494013522856765, + "learning_rate": 1.942463192516925e-06, + "loss": 0.9169, + "step": 5809 + }, + { + "epoch": 0.5239662713622221, + "grad_norm": 1.4847686245673957, + "learning_rate": 1.9418792261541746e-06, + "loss": 0.9784, + "step": 5810 + }, + { + "epoch": 0.5240564548856924, + "grad_norm": 1.6138128924062967, + "learning_rate": 1.9412952647506094e-06, + "loss": 0.9148, + "step": 5811 + }, + { + "epoch": 0.5241466384091626, + "grad_norm": 1.4119592695008054, + "learning_rate": 1.9407113083560552e-06, + "loss": 0.9383, + "step": 5812 + }, + { + "epoch": 0.5242368219326329, + "grad_norm": 1.6387129040788777, + "learning_rate": 1.940127357020339e-06, + "loss": 0.9392, + "step": 5813 + }, + { + "epoch": 0.5243270054561032, + "grad_norm": 1.5747001423056677, + "learning_rate": 1.939543410793287e-06, + "loss": 0.9677, + "step": 5814 + }, + { + "epoch": 0.5244171889795735, + "grad_norm": 1.8728350538464447, + "learning_rate": 1.9389594697247246e-06, + "loss": 0.9932, + "step": 5815 + }, + { + "epoch": 0.5245073725030437, + "grad_norm": 1.7202049339543894, + "learning_rate": 1.9383755338644763e-06, + "loss": 1.0316, + "step": 5816 + }, + { + "epoch": 0.5245975560265139, + "grad_norm": 1.7158784633936943, + "learning_rate": 1.937791603262368e-06, + "loss": 0.9888, + "step": 5817 + }, + { + "epoch": 0.5246877395499843, + "grad_norm": 1.5193934831707223, + "learning_rate": 1.9372076779682235e-06, + "loss": 0.9092, + "step": 5818 + }, + { + "epoch": 0.5247779230734545, + "grad_norm": 1.6658287882715037, + "learning_rate": 1.9366237580318648e-06, + "loss": 0.9106, + "step": 5819 + }, + { + "epoch": 0.5248681065969247, + "grad_norm": 1.2730636515979463, + "learning_rate": 1.9360398435031176e-06, + "loss": 0.9613, + "step": 5820 + }, + { + "epoch": 0.524958290120395, + "grad_norm": 1.670092907451272, + "learning_rate": 1.9354559344318025e-06, + "loss": 0.9653, + "step": 5821 + }, + { + "epoch": 0.5250484736438653, + "grad_norm": 1.7086952880494357, + "learning_rate": 1.934872030867744e-06, + "loss": 1.0128, + "step": 5822 + }, + { + "epoch": 0.5251386571673355, + "grad_norm": 1.575054013370586, + "learning_rate": 1.934288132860763e-06, + "loss": 0.9679, + "step": 5823 + }, + { + "epoch": 0.5252288406908058, + "grad_norm": 1.4330977327570165, + "learning_rate": 1.93370424046068e-06, + "loss": 0.9551, + "step": 5824 + }, + { + "epoch": 0.525319024214276, + "grad_norm": 1.3692493259525085, + "learning_rate": 1.9331203537173177e-06, + "loss": 1.022, + "step": 5825 + }, + { + "epoch": 0.5254092077377464, + "grad_norm": 1.311561475984702, + "learning_rate": 1.9325364726804947e-06, + "loss": 1.0702, + "step": 5826 + }, + { + "epoch": 0.5254993912612166, + "grad_norm": 1.5757469918683653, + "learning_rate": 1.9319525974000327e-06, + "loss": 0.8865, + "step": 5827 + }, + { + "epoch": 0.5255895747846868, + "grad_norm": 2.0269083426500556, + "learning_rate": 1.93136872792575e-06, + "loss": 1.0224, + "step": 5828 + }, + { + "epoch": 0.525679758308157, + "grad_norm": 1.5231920093420948, + "learning_rate": 1.9307848643074653e-06, + "loss": 0.9105, + "step": 5829 + }, + { + "epoch": 0.5257699418316274, + "grad_norm": 1.4607494865650303, + "learning_rate": 1.9302010065949985e-06, + "loss": 0.9925, + "step": 5830 + }, + { + "epoch": 0.5258601253550976, + "grad_norm": 1.4453572859140063, + "learning_rate": 1.9296171548381657e-06, + "loss": 0.917, + "step": 5831 + }, + { + "epoch": 0.5259503088785679, + "grad_norm": 1.4637361535281506, + "learning_rate": 1.9290333090867862e-06, + "loss": 0.944, + "step": 5832 + }, + { + "epoch": 0.5260404924020381, + "grad_norm": 1.6220906696730177, + "learning_rate": 1.928449469390676e-06, + "loss": 0.9622, + "step": 5833 + }, + { + "epoch": 0.5261306759255084, + "grad_norm": 1.5174088352980393, + "learning_rate": 1.927865635799651e-06, + "loss": 0.9305, + "step": 5834 + }, + { + "epoch": 0.5262208594489787, + "grad_norm": 1.3570923768959193, + "learning_rate": 1.927281808363528e-06, + "loss": 0.8857, + "step": 5835 + }, + { + "epoch": 0.5263110429724489, + "grad_norm": 1.5721710745299002, + "learning_rate": 1.9266979871321216e-06, + "loss": 1.0296, + "step": 5836 + }, + { + "epoch": 0.5264012264959192, + "grad_norm": 1.4302320433464428, + "learning_rate": 1.9261141721552482e-06, + "loss": 0.9082, + "step": 5837 + }, + { + "epoch": 0.5264914100193895, + "grad_norm": 1.5380309855014855, + "learning_rate": 1.9255303634827204e-06, + "loss": 1.0258, + "step": 5838 + }, + { + "epoch": 0.5265815935428597, + "grad_norm": 1.5479571477937106, + "learning_rate": 1.924946561164352e-06, + "loss": 1.0224, + "step": 5839 + }, + { + "epoch": 0.52667177706633, + "grad_norm": 1.5177883164645165, + "learning_rate": 1.9243627652499582e-06, + "loss": 0.9091, + "step": 5840 + }, + { + "epoch": 0.5267619605898003, + "grad_norm": 1.5969191811976133, + "learning_rate": 1.9237789757893493e-06, + "loss": 1.0244, + "step": 5841 + }, + { + "epoch": 0.5268521441132705, + "grad_norm": 1.6392173904822012, + "learning_rate": 1.9231951928323395e-06, + "loss": 1.0011, + "step": 5842 + }, + { + "epoch": 0.5269423276367408, + "grad_norm": 1.7047160005503406, + "learning_rate": 1.922611416428738e-06, + "loss": 0.9124, + "step": 5843 + }, + { + "epoch": 0.527032511160211, + "grad_norm": 1.4958523787168851, + "learning_rate": 1.922027646628358e-06, + "loss": 0.9837, + "step": 5844 + }, + { + "epoch": 0.5271226946836813, + "grad_norm": 0.7068755046798834, + "learning_rate": 1.9214438834810092e-06, + "loss": 0.8366, + "step": 5845 + }, + { + "epoch": 0.5272128782071516, + "grad_norm": 1.4416449124379933, + "learning_rate": 1.9208601270365008e-06, + "loss": 0.8867, + "step": 5846 + }, + { + "epoch": 0.5273030617306218, + "grad_norm": 1.8363445723494316, + "learning_rate": 1.9202763773446435e-06, + "loss": 1.0401, + "step": 5847 + }, + { + "epoch": 0.527393245254092, + "grad_norm": 1.5515595988297397, + "learning_rate": 1.9196926344552444e-06, + "loss": 0.8519, + "step": 5848 + }, + { + "epoch": 0.5274834287775624, + "grad_norm": 1.4658351137556345, + "learning_rate": 1.919108898418113e-06, + "loss": 0.9824, + "step": 5849 + }, + { + "epoch": 0.5275736123010326, + "grad_norm": 1.5302446139894754, + "learning_rate": 1.918525169283057e-06, + "loss": 0.8542, + "step": 5850 + }, + { + "epoch": 0.5276637958245028, + "grad_norm": 1.537291383934107, + "learning_rate": 1.9179414470998817e-06, + "loss": 0.9855, + "step": 5851 + }, + { + "epoch": 0.5277539793479731, + "grad_norm": 1.4505995267367884, + "learning_rate": 1.917357731918395e-06, + "loss": 0.9142, + "step": 5852 + }, + { + "epoch": 0.5278441628714434, + "grad_norm": 1.3589222471943263, + "learning_rate": 1.9167740237884025e-06, + "loss": 1.0886, + "step": 5853 + }, + { + "epoch": 0.5279343463949137, + "grad_norm": 1.479732602166207, + "learning_rate": 1.916190322759709e-06, + "loss": 1.0003, + "step": 5854 + }, + { + "epoch": 0.5280245299183839, + "grad_norm": 1.5030005167530667, + "learning_rate": 1.91560662888212e-06, + "loss": 1.0559, + "step": 5855 + }, + { + "epoch": 0.5281147134418541, + "grad_norm": 1.389539015986434, + "learning_rate": 1.915022942205438e-06, + "loss": 1.0021, + "step": 5856 + }, + { + "epoch": 0.5282048969653245, + "grad_norm": 1.7395136547029664, + "learning_rate": 1.914439262779468e-06, + "loss": 0.9514, + "step": 5857 + }, + { + "epoch": 0.5282950804887947, + "grad_norm": 1.670540606889203, + "learning_rate": 1.9138555906540103e-06, + "loss": 0.89, + "step": 5858 + }, + { + "epoch": 0.5283852640122649, + "grad_norm": 1.3960785816395405, + "learning_rate": 1.91327192587887e-06, + "loss": 0.8672, + "step": 5859 + }, + { + "epoch": 0.5284754475357353, + "grad_norm": 1.9087310029628985, + "learning_rate": 1.912688268503846e-06, + "loss": 0.9337, + "step": 5860 + }, + { + "epoch": 0.5285656310592055, + "grad_norm": 1.6525363674576918, + "learning_rate": 1.912104618578741e-06, + "loss": 0.9207, + "step": 5861 + }, + { + "epoch": 0.5286558145826757, + "grad_norm": 1.3462451568488063, + "learning_rate": 1.9115209761533554e-06, + "loss": 0.9343, + "step": 5862 + }, + { + "epoch": 0.528745998106146, + "grad_norm": 1.7332171450089366, + "learning_rate": 1.9109373412774863e-06, + "loss": 0.8719, + "step": 5863 + }, + { + "epoch": 0.5288361816296163, + "grad_norm": 1.3235818237012351, + "learning_rate": 1.910353714000936e-06, + "loss": 0.9602, + "step": 5864 + }, + { + "epoch": 0.5289263651530866, + "grad_norm": 1.366041433359444, + "learning_rate": 1.9097700943734997e-06, + "loss": 0.9184, + "step": 5865 + }, + { + "epoch": 0.5290165486765568, + "grad_norm": 1.506653651684837, + "learning_rate": 1.909186482444977e-06, + "loss": 0.8552, + "step": 5866 + }, + { + "epoch": 0.529106732200027, + "grad_norm": 1.7590087888930732, + "learning_rate": 1.9086028782651652e-06, + "loss": 0.8628, + "step": 5867 + }, + { + "epoch": 0.5291969157234974, + "grad_norm": 1.4483933064489132, + "learning_rate": 1.908019281883859e-06, + "loss": 0.9523, + "step": 5868 + }, + { + "epoch": 0.5292870992469676, + "grad_norm": 1.3931681064048058, + "learning_rate": 1.9074356933508545e-06, + "loss": 0.948, + "step": 5869 + }, + { + "epoch": 0.5293772827704378, + "grad_norm": 2.052661898102053, + "learning_rate": 1.9068521127159477e-06, + "loss": 0.9945, + "step": 5870 + }, + { + "epoch": 0.5294674662939081, + "grad_norm": 1.3853705071047118, + "learning_rate": 1.9062685400289322e-06, + "loss": 0.9093, + "step": 5871 + }, + { + "epoch": 0.5295576498173784, + "grad_norm": 1.432499522530308, + "learning_rate": 1.9056849753396018e-06, + "loss": 0.9315, + "step": 5872 + }, + { + "epoch": 0.5296478333408486, + "grad_norm": 2.1603691451444673, + "learning_rate": 1.9051014186977485e-06, + "loss": 0.993, + "step": 5873 + }, + { + "epoch": 0.5297380168643189, + "grad_norm": 1.4275945133113412, + "learning_rate": 1.9045178701531664e-06, + "loss": 0.9583, + "step": 5874 + }, + { + "epoch": 0.5298282003877891, + "grad_norm": 1.2403423592723346, + "learning_rate": 1.903934329755645e-06, + "loss": 0.9502, + "step": 5875 + }, + { + "epoch": 0.5299183839112594, + "grad_norm": 1.479966775547013, + "learning_rate": 1.9033507975549775e-06, + "loss": 0.9913, + "step": 5876 + }, + { + "epoch": 0.5300085674347297, + "grad_norm": 1.6169407605636013, + "learning_rate": 1.9027672736009525e-06, + "loss": 0.9485, + "step": 5877 + }, + { + "epoch": 0.5300987509581999, + "grad_norm": 1.3667206757426038, + "learning_rate": 1.9021837579433593e-06, + "loss": 0.9035, + "step": 5878 + }, + { + "epoch": 0.5301889344816701, + "grad_norm": 1.4542137087378706, + "learning_rate": 1.901600250631988e-06, + "loss": 0.9289, + "step": 5879 + }, + { + "epoch": 0.5302791180051405, + "grad_norm": 1.6309857530796084, + "learning_rate": 1.901016751716625e-06, + "loss": 0.9434, + "step": 5880 + }, + { + "epoch": 0.5303693015286107, + "grad_norm": 2.536498762013505, + "learning_rate": 1.9004332612470593e-06, + "loss": 0.9223, + "step": 5881 + }, + { + "epoch": 0.530459485052081, + "grad_norm": 1.1775224029936764, + "learning_rate": 1.8998497792730763e-06, + "loss": 0.9825, + "step": 5882 + }, + { + "epoch": 0.5305496685755512, + "grad_norm": 1.5346889415525247, + "learning_rate": 1.8992663058444629e-06, + "loss": 1.0047, + "step": 5883 + }, + { + "epoch": 0.5306398520990215, + "grad_norm": 1.5040516968780409, + "learning_rate": 1.8986828410110032e-06, + "loss": 0.9938, + "step": 5884 + }, + { + "epoch": 0.5307300356224918, + "grad_norm": 1.630129567478855, + "learning_rate": 1.8980993848224823e-06, + "loss": 0.9506, + "step": 5885 + }, + { + "epoch": 0.530820219145962, + "grad_norm": 1.4312855278836232, + "learning_rate": 1.8975159373286843e-06, + "loss": 0.9111, + "step": 5886 + }, + { + "epoch": 0.5309104026694323, + "grad_norm": 1.5161759466467557, + "learning_rate": 1.8969324985793904e-06, + "loss": 0.9881, + "step": 5887 + }, + { + "epoch": 0.5310005861929026, + "grad_norm": 1.609886310425403, + "learning_rate": 1.8963490686243851e-06, + "loss": 0.9509, + "step": 5888 + }, + { + "epoch": 0.5310907697163728, + "grad_norm": 1.7377532499982955, + "learning_rate": 1.8957656475134486e-06, + "loss": 0.8939, + "step": 5889 + }, + { + "epoch": 0.531180953239843, + "grad_norm": 1.5805159482069837, + "learning_rate": 1.895182235296361e-06, + "loss": 0.9337, + "step": 5890 + }, + { + "epoch": 0.5312711367633134, + "grad_norm": 1.6438392759961715, + "learning_rate": 1.8945988320229042e-06, + "loss": 0.8799, + "step": 5891 + }, + { + "epoch": 0.5313613202867836, + "grad_norm": 1.5504597443512553, + "learning_rate": 1.8940154377428553e-06, + "loss": 1.0355, + "step": 5892 + }, + { + "epoch": 0.5314515038102539, + "grad_norm": 1.3762068653882882, + "learning_rate": 1.8934320525059944e-06, + "loss": 0.9833, + "step": 5893 + }, + { + "epoch": 0.5315416873337241, + "grad_norm": 2.381000968346525, + "learning_rate": 1.8928486763620984e-06, + "loss": 0.8407, + "step": 5894 + }, + { + "epoch": 0.5316318708571944, + "grad_norm": 3.2169944642279904, + "learning_rate": 1.892265309360943e-06, + "loss": 0.9697, + "step": 5895 + }, + { + "epoch": 0.5317220543806647, + "grad_norm": 12.262510178743545, + "learning_rate": 1.8916819515523067e-06, + "loss": 0.9402, + "step": 5896 + }, + { + "epoch": 0.5318122379041349, + "grad_norm": 1.3863614763820098, + "learning_rate": 1.891098602985963e-06, + "loss": 0.8734, + "step": 5897 + }, + { + "epoch": 0.5319024214276051, + "grad_norm": 1.4297371944274915, + "learning_rate": 1.8905152637116872e-06, + "loss": 0.9824, + "step": 5898 + }, + { + "epoch": 0.5319926049510755, + "grad_norm": 1.6183480047129915, + "learning_rate": 1.8899319337792527e-06, + "loss": 0.9931, + "step": 5899 + }, + { + "epoch": 0.5320827884745457, + "grad_norm": 1.2287432000407665, + "learning_rate": 1.8893486132384325e-06, + "loss": 0.9956, + "step": 5900 + }, + { + "epoch": 0.5321729719980159, + "grad_norm": 1.547383465669868, + "learning_rate": 1.888765302138999e-06, + "loss": 0.9232, + "step": 5901 + }, + { + "epoch": 0.5322631555214862, + "grad_norm": 1.5535911643367284, + "learning_rate": 1.8881820005307224e-06, + "loss": 0.8976, + "step": 5902 + }, + { + "epoch": 0.5323533390449565, + "grad_norm": 1.3681033745211506, + "learning_rate": 1.8875987084633748e-06, + "loss": 0.8893, + "step": 5903 + }, + { + "epoch": 0.5324435225684268, + "grad_norm": 1.7759675115801274, + "learning_rate": 1.8870154259867246e-06, + "loss": 0.9818, + "step": 5904 + }, + { + "epoch": 0.532533706091897, + "grad_norm": 1.607159172459697, + "learning_rate": 1.886432153150542e-06, + "loss": 0.8882, + "step": 5905 + }, + { + "epoch": 0.5326238896153672, + "grad_norm": 1.4553872004220958, + "learning_rate": 1.8858488900045944e-06, + "loss": 0.9319, + "step": 5906 + }, + { + "epoch": 0.5327140731388376, + "grad_norm": 2.2362227902321608, + "learning_rate": 1.885265636598648e-06, + "loss": 0.9127, + "step": 5907 + }, + { + "epoch": 0.5328042566623078, + "grad_norm": 1.3940235570835442, + "learning_rate": 1.884682392982471e-06, + "loss": 0.989, + "step": 5908 + }, + { + "epoch": 0.532894440185778, + "grad_norm": 2.0796573123704185, + "learning_rate": 1.8840991592058274e-06, + "loss": 1.0079, + "step": 5909 + }, + { + "epoch": 0.5329846237092484, + "grad_norm": 1.4099046054656914, + "learning_rate": 1.8835159353184828e-06, + "loss": 0.9706, + "step": 5910 + }, + { + "epoch": 0.5330748072327186, + "grad_norm": 1.3529947883381528, + "learning_rate": 1.8829327213702013e-06, + "loss": 1.0284, + "step": 5911 + }, + { + "epoch": 0.5331649907561888, + "grad_norm": 1.4472797923353473, + "learning_rate": 1.8823495174107452e-06, + "loss": 1.0283, + "step": 5912 + }, + { + "epoch": 0.5332551742796591, + "grad_norm": 1.5612028459646647, + "learning_rate": 1.8817663234898773e-06, + "loss": 0.9242, + "step": 5913 + }, + { + "epoch": 0.5333453578031294, + "grad_norm": 2.0511106971243436, + "learning_rate": 1.881183139657358e-06, + "loss": 0.9644, + "step": 5914 + }, + { + "epoch": 0.5334355413265996, + "grad_norm": 1.8385561833515478, + "learning_rate": 1.8805999659629488e-06, + "loss": 0.9695, + "step": 5915 + }, + { + "epoch": 0.5335257248500699, + "grad_norm": 0.7205033270886788, + "learning_rate": 1.880016802456409e-06, + "loss": 0.737, + "step": 5916 + }, + { + "epoch": 0.5336159083735401, + "grad_norm": 1.574488160055834, + "learning_rate": 1.8794336491874964e-06, + "loss": 0.8919, + "step": 5917 + }, + { + "epoch": 0.5337060918970105, + "grad_norm": 1.7525274553746772, + "learning_rate": 1.8788505062059708e-06, + "loss": 0.9185, + "step": 5918 + }, + { + "epoch": 0.5337962754204807, + "grad_norm": 1.6050612032145233, + "learning_rate": 1.8782673735615869e-06, + "loss": 0.937, + "step": 5919 + }, + { + "epoch": 0.5338864589439509, + "grad_norm": 1.5771240232485337, + "learning_rate": 1.8776842513041026e-06, + "loss": 0.9086, + "step": 5920 + }, + { + "epoch": 0.5339766424674212, + "grad_norm": 2.488153906517503, + "learning_rate": 1.8771011394832727e-06, + "loss": 0.9184, + "step": 5921 + }, + { + "epoch": 0.5340668259908915, + "grad_norm": 1.4715746238988368, + "learning_rate": 1.8765180381488501e-06, + "loss": 0.9351, + "step": 5922 + }, + { + "epoch": 0.5341570095143617, + "grad_norm": 0.7801444432812811, + "learning_rate": 1.8759349473505905e-06, + "loss": 0.869, + "step": 5923 + }, + { + "epoch": 0.534247193037832, + "grad_norm": 1.4813229152583205, + "learning_rate": 1.8753518671382447e-06, + "loss": 0.8784, + "step": 5924 + }, + { + "epoch": 0.5343373765613022, + "grad_norm": 1.4242111130518555, + "learning_rate": 1.8747687975615649e-06, + "loss": 0.9328, + "step": 5925 + }, + { + "epoch": 0.5344275600847725, + "grad_norm": 1.6041593757893071, + "learning_rate": 1.874185738670302e-06, + "loss": 0.8468, + "step": 5926 + }, + { + "epoch": 0.5345177436082428, + "grad_norm": 1.6870584263387605, + "learning_rate": 1.8736026905142057e-06, + "loss": 0.9406, + "step": 5927 + }, + { + "epoch": 0.534607927131713, + "grad_norm": 1.435103035128679, + "learning_rate": 1.873019653143025e-06, + "loss": 0.929, + "step": 5928 + }, + { + "epoch": 0.5346981106551832, + "grad_norm": 1.4594679188196782, + "learning_rate": 1.8724366266065069e-06, + "loss": 0.9617, + "step": 5929 + }, + { + "epoch": 0.5347882941786536, + "grad_norm": 1.4640933136358916, + "learning_rate": 1.8718536109543998e-06, + "loss": 1.0263, + "step": 5930 + }, + { + "epoch": 0.5348784777021238, + "grad_norm": 1.5549334614246386, + "learning_rate": 1.8712706062364485e-06, + "loss": 0.9648, + "step": 5931 + }, + { + "epoch": 0.534968661225594, + "grad_norm": 1.6885721544748746, + "learning_rate": 1.8706876125024e-06, + "loss": 0.8395, + "step": 5932 + }, + { + "epoch": 0.5350588447490644, + "grad_norm": 1.6751837387654727, + "learning_rate": 1.870104629801997e-06, + "loss": 0.9956, + "step": 5933 + }, + { + "epoch": 0.5351490282725346, + "grad_norm": 1.6395076761720424, + "learning_rate": 1.8695216581849823e-06, + "loss": 0.9679, + "step": 5934 + }, + { + "epoch": 0.5352392117960049, + "grad_norm": 1.4029839204749088, + "learning_rate": 1.8689386977011003e-06, + "loss": 0.9195, + "step": 5935 + }, + { + "epoch": 0.5353293953194751, + "grad_norm": 0.7399190669005964, + "learning_rate": 1.8683557484000903e-06, + "loss": 0.7768, + "step": 5936 + }, + { + "epoch": 0.5354195788429454, + "grad_norm": 1.6702704171855731, + "learning_rate": 1.8677728103316947e-06, + "loss": 0.982, + "step": 5937 + }, + { + "epoch": 0.5355097623664157, + "grad_norm": 1.3251709791745085, + "learning_rate": 1.8671898835456518e-06, + "loss": 0.9068, + "step": 5938 + }, + { + "epoch": 0.5355999458898859, + "grad_norm": 1.5957675953498147, + "learning_rate": 1.8666069680917003e-06, + "loss": 1.0563, + "step": 5939 + }, + { + "epoch": 0.5356901294133561, + "grad_norm": 1.723370531567343, + "learning_rate": 1.8660240640195775e-06, + "loss": 0.9148, + "step": 5940 + }, + { + "epoch": 0.5357803129368265, + "grad_norm": 1.2940354571341441, + "learning_rate": 1.8654411713790203e-06, + "loss": 0.9906, + "step": 5941 + }, + { + "epoch": 0.5358704964602967, + "grad_norm": 1.666272911606796, + "learning_rate": 1.8648582902197648e-06, + "loss": 0.972, + "step": 5942 + }, + { + "epoch": 0.535960679983767, + "grad_norm": 1.7315476199593147, + "learning_rate": 1.8642754205915452e-06, + "loss": 0.9575, + "step": 5943 + }, + { + "epoch": 0.5360508635072372, + "grad_norm": 1.4644490842033393, + "learning_rate": 1.8636925625440943e-06, + "loss": 0.8997, + "step": 5944 + }, + { + "epoch": 0.5361410470307075, + "grad_norm": 1.7105549423818525, + "learning_rate": 1.863109716127146e-06, + "loss": 0.8587, + "step": 5945 + }, + { + "epoch": 0.5362312305541778, + "grad_norm": 1.3043263101012534, + "learning_rate": 1.8625268813904311e-06, + "loss": 0.9932, + "step": 5946 + }, + { + "epoch": 0.536321414077648, + "grad_norm": 1.3477006103291198, + "learning_rate": 1.8619440583836814e-06, + "loss": 0.9804, + "step": 5947 + }, + { + "epoch": 0.5364115976011182, + "grad_norm": 2.0019697979488638, + "learning_rate": 1.8613612471566249e-06, + "loss": 0.9597, + "step": 5948 + }, + { + "epoch": 0.5365017811245886, + "grad_norm": 1.507634522997763, + "learning_rate": 1.8607784477589922e-06, + "loss": 0.9006, + "step": 5949 + }, + { + "epoch": 0.5365919646480588, + "grad_norm": 2.218812216302813, + "learning_rate": 1.8601956602405103e-06, + "loss": 0.9698, + "step": 5950 + }, + { + "epoch": 0.536682148171529, + "grad_norm": 1.61595740966099, + "learning_rate": 1.8596128846509043e-06, + "loss": 0.9963, + "step": 5951 + }, + { + "epoch": 0.5367723316949993, + "grad_norm": 1.3355252706292682, + "learning_rate": 1.859030121039902e-06, + "loss": 0.9046, + "step": 5952 + }, + { + "epoch": 0.5368625152184696, + "grad_norm": 1.421384957514582, + "learning_rate": 1.8584473694572268e-06, + "loss": 0.9037, + "step": 5953 + }, + { + "epoch": 0.5369526987419398, + "grad_norm": 1.5207720008223151, + "learning_rate": 1.8578646299526026e-06, + "loss": 0.9735, + "step": 5954 + }, + { + "epoch": 0.5370428822654101, + "grad_norm": 1.5182025660632221, + "learning_rate": 1.8572819025757518e-06, + "loss": 0.9245, + "step": 5955 + }, + { + "epoch": 0.5371330657888804, + "grad_norm": 1.6150015832757423, + "learning_rate": 1.8566991873763959e-06, + "loss": 0.963, + "step": 5956 + }, + { + "epoch": 0.5372232493123507, + "grad_norm": 1.626611570704596, + "learning_rate": 1.856116484404256e-06, + "loss": 0.8995, + "step": 5957 + }, + { + "epoch": 0.5373134328358209, + "grad_norm": 1.6891200448395922, + "learning_rate": 1.8555337937090502e-06, + "loss": 0.9181, + "step": 5958 + }, + { + "epoch": 0.5374036163592911, + "grad_norm": 1.8442687904959216, + "learning_rate": 1.8549511153404984e-06, + "loss": 0.9413, + "step": 5959 + }, + { + "epoch": 0.5374937998827615, + "grad_norm": 1.4731447740149368, + "learning_rate": 1.854368449348317e-06, + "loss": 1.0649, + "step": 5960 + }, + { + "epoch": 0.5375839834062317, + "grad_norm": 1.6162953150179955, + "learning_rate": 1.853785795782222e-06, + "loss": 0.9129, + "step": 5961 + }, + { + "epoch": 0.5376741669297019, + "grad_norm": 1.6447830551079863, + "learning_rate": 1.85320315469193e-06, + "loss": 0.8205, + "step": 5962 + }, + { + "epoch": 0.5377643504531722, + "grad_norm": 0.6983969136847745, + "learning_rate": 1.8526205261271534e-06, + "loss": 0.783, + "step": 5963 + }, + { + "epoch": 0.5378545339766425, + "grad_norm": 1.415910388061428, + "learning_rate": 1.852037910137607e-06, + "loss": 0.8669, + "step": 5964 + }, + { + "epoch": 0.5379447175001127, + "grad_norm": 1.3541286120814213, + "learning_rate": 1.851455306773002e-06, + "loss": 1.042, + "step": 5965 + }, + { + "epoch": 0.538034901023583, + "grad_norm": 1.5575992021150118, + "learning_rate": 1.8508727160830483e-06, + "loss": 0.9271, + "step": 5966 + }, + { + "epoch": 0.5381250845470532, + "grad_norm": 1.4975175979420876, + "learning_rate": 1.8502901381174575e-06, + "loss": 1.0264, + "step": 5967 + }, + { + "epoch": 0.5382152680705236, + "grad_norm": 2.0291979474653425, + "learning_rate": 1.8497075729259372e-06, + "loss": 0.8566, + "step": 5968 + }, + { + "epoch": 0.5383054515939938, + "grad_norm": 0.8331264397965715, + "learning_rate": 1.8491250205581963e-06, + "loss": 0.8429, + "step": 5969 + }, + { + "epoch": 0.538395635117464, + "grad_norm": 1.3803492875969285, + "learning_rate": 1.8485424810639393e-06, + "loss": 0.956, + "step": 5970 + }, + { + "epoch": 0.5384858186409343, + "grad_norm": 1.7768235336079747, + "learning_rate": 1.847959954492874e-06, + "loss": 1.0141, + "step": 5971 + }, + { + "epoch": 0.5385760021644046, + "grad_norm": 1.8284499702817254, + "learning_rate": 1.8473774408947035e-06, + "loss": 1.0468, + "step": 5972 + }, + { + "epoch": 0.5386661856878748, + "grad_norm": 0.774152290578951, + "learning_rate": 1.8467949403191308e-06, + "loss": 0.7547, + "step": 5973 + }, + { + "epoch": 0.5387563692113451, + "grad_norm": 1.5898528227676088, + "learning_rate": 1.8462124528158592e-06, + "loss": 1.009, + "step": 5974 + }, + { + "epoch": 0.5388465527348153, + "grad_norm": 1.4113053054717415, + "learning_rate": 1.8456299784345881e-06, + "loss": 0.9388, + "step": 5975 + }, + { + "epoch": 0.5389367362582856, + "grad_norm": 1.5049654314753922, + "learning_rate": 1.8450475172250194e-06, + "loss": 0.969, + "step": 5976 + }, + { + "epoch": 0.5390269197817559, + "grad_norm": 1.384703168012012, + "learning_rate": 1.844465069236851e-06, + "loss": 0.9343, + "step": 5977 + }, + { + "epoch": 0.5391171033052261, + "grad_norm": 1.5266697169499694, + "learning_rate": 1.8438826345197796e-06, + "loss": 0.8631, + "step": 5978 + }, + { + "epoch": 0.5392072868286965, + "grad_norm": 1.958295463304067, + "learning_rate": 1.8433002131235036e-06, + "loss": 0.9428, + "step": 5979 + }, + { + "epoch": 0.5392974703521667, + "grad_norm": 1.341470603754441, + "learning_rate": 1.8427178050977167e-06, + "loss": 0.9728, + "step": 5980 + }, + { + "epoch": 0.5393876538756369, + "grad_norm": 1.3667317530012484, + "learning_rate": 1.8421354104921143e-06, + "loss": 1.0308, + "step": 5981 + }, + { + "epoch": 0.5394778373991072, + "grad_norm": 0.7462134539858246, + "learning_rate": 1.8415530293563894e-06, + "loss": 0.7504, + "step": 5982 + }, + { + "epoch": 0.5395680209225775, + "grad_norm": 1.3871573171062568, + "learning_rate": 1.8409706617402333e-06, + "loss": 0.9335, + "step": 5983 + }, + { + "epoch": 0.5396582044460477, + "grad_norm": 1.6506479638266995, + "learning_rate": 1.8403883076933378e-06, + "loss": 0.973, + "step": 5984 + }, + { + "epoch": 0.539748387969518, + "grad_norm": 1.4200718103975556, + "learning_rate": 1.839805967265391e-06, + "loss": 0.9418, + "step": 5985 + }, + { + "epoch": 0.5398385714929882, + "grad_norm": 1.475895887600398, + "learning_rate": 1.839223640506083e-06, + "loss": 1.0132, + "step": 5986 + }, + { + "epoch": 0.5399287550164585, + "grad_norm": 1.8696702866397081, + "learning_rate": 1.8386413274650998e-06, + "loss": 0.9423, + "step": 5987 + }, + { + "epoch": 0.5400189385399288, + "grad_norm": 1.475458368105096, + "learning_rate": 1.8380590281921294e-06, + "loss": 1.0005, + "step": 5988 + }, + { + "epoch": 0.540109122063399, + "grad_norm": 1.5179172760407849, + "learning_rate": 1.8374767427368552e-06, + "loss": 0.8589, + "step": 5989 + }, + { + "epoch": 0.5401993055868692, + "grad_norm": 1.5672179138335927, + "learning_rate": 1.8368944711489608e-06, + "loss": 0.984, + "step": 5990 + }, + { + "epoch": 0.5402894891103396, + "grad_norm": 1.842270726466647, + "learning_rate": 1.8363122134781304e-06, + "loss": 0.9836, + "step": 5991 + }, + { + "epoch": 0.5403796726338098, + "grad_norm": 1.5197094810319114, + "learning_rate": 1.835729969774044e-06, + "loss": 0.9768, + "step": 5992 + }, + { + "epoch": 0.54046985615728, + "grad_norm": 1.3655457584951176, + "learning_rate": 1.8351477400863823e-06, + "loss": 0.9621, + "step": 5993 + }, + { + "epoch": 0.5405600396807503, + "grad_norm": 1.4665401147661743, + "learning_rate": 1.8345655244648249e-06, + "loss": 0.9351, + "step": 5994 + }, + { + "epoch": 0.5406502232042206, + "grad_norm": 1.2535761698802794, + "learning_rate": 1.8339833229590486e-06, + "loss": 0.9415, + "step": 5995 + }, + { + "epoch": 0.5407404067276909, + "grad_norm": 1.7816306259654187, + "learning_rate": 1.833401135618731e-06, + "loss": 1.0227, + "step": 5996 + }, + { + "epoch": 0.5408305902511611, + "grad_norm": 1.429341675670557, + "learning_rate": 1.8328189624935466e-06, + "loss": 0.8811, + "step": 5997 + }, + { + "epoch": 0.5409207737746313, + "grad_norm": 1.5697570121937798, + "learning_rate": 1.832236803633171e-06, + "loss": 0.9869, + "step": 5998 + }, + { + "epoch": 0.5410109572981017, + "grad_norm": 1.7403934290476721, + "learning_rate": 1.831654659087276e-06, + "loss": 0.918, + "step": 5999 + }, + { + "epoch": 0.5411011408215719, + "grad_norm": 1.4967906314909387, + "learning_rate": 1.831072528905533e-06, + "loss": 0.9677, + "step": 6000 + }, + { + "epoch": 0.5411913243450421, + "grad_norm": 1.3475768577299474, + "learning_rate": 1.8304904131376142e-06, + "loss": 1.0064, + "step": 6001 + }, + { + "epoch": 0.5412815078685124, + "grad_norm": 1.6612170611259762, + "learning_rate": 1.8299083118331874e-06, + "loss": 0.9712, + "step": 6002 + }, + { + "epoch": 0.5413716913919827, + "grad_norm": 2.0361413603312783, + "learning_rate": 1.8293262250419217e-06, + "loss": 1.087, + "step": 6003 + }, + { + "epoch": 0.541461874915453, + "grad_norm": 1.5882105581272974, + "learning_rate": 1.828744152813484e-06, + "loss": 1.0713, + "step": 6004 + }, + { + "epoch": 0.5415520584389232, + "grad_norm": 1.583581311064592, + "learning_rate": 1.8281620951975382e-06, + "loss": 0.9067, + "step": 6005 + }, + { + "epoch": 0.5416422419623935, + "grad_norm": 1.5890125098003098, + "learning_rate": 1.827580052243751e-06, + "loss": 0.9185, + "step": 6006 + }, + { + "epoch": 0.5417324254858638, + "grad_norm": 2.0837542935094326, + "learning_rate": 1.826998024001784e-06, + "loss": 0.9849, + "step": 6007 + }, + { + "epoch": 0.541822609009334, + "grad_norm": 1.6901954739852285, + "learning_rate": 1.8264160105212995e-06, + "loss": 0.9843, + "step": 6008 + }, + { + "epoch": 0.5419127925328042, + "grad_norm": 1.5004989271099725, + "learning_rate": 1.8258340118519582e-06, + "loss": 0.9218, + "step": 6009 + }, + { + "epoch": 0.5420029760562746, + "grad_norm": 1.6530463706572454, + "learning_rate": 1.82525202804342e-06, + "loss": 0.9403, + "step": 6010 + }, + { + "epoch": 0.5420931595797448, + "grad_norm": 1.9374321956460385, + "learning_rate": 1.8246700591453415e-06, + "loss": 0.9384, + "step": 6011 + }, + { + "epoch": 0.542183343103215, + "grad_norm": 1.3519080718038707, + "learning_rate": 1.8240881052073801e-06, + "loss": 0.9547, + "step": 6012 + }, + { + "epoch": 0.5422735266266853, + "grad_norm": 1.526140446335634, + "learning_rate": 1.8235061662791923e-06, + "loss": 0.9247, + "step": 6013 + }, + { + "epoch": 0.5423637101501556, + "grad_norm": 1.5065946255961635, + "learning_rate": 1.8229242424104309e-06, + "loss": 1.0236, + "step": 6014 + }, + { + "epoch": 0.5424538936736258, + "grad_norm": 0.6648113067994954, + "learning_rate": 1.8223423336507503e-06, + "loss": 0.7601, + "step": 6015 + }, + { + "epoch": 0.5425440771970961, + "grad_norm": 1.623517019987987, + "learning_rate": 1.8217604400498012e-06, + "loss": 0.9382, + "step": 6016 + }, + { + "epoch": 0.5426342607205663, + "grad_norm": 1.3314676294950807, + "learning_rate": 1.8211785616572333e-06, + "loss": 1.0045, + "step": 6017 + }, + { + "epoch": 0.5427244442440367, + "grad_norm": 1.8617357926383045, + "learning_rate": 1.8205966985226975e-06, + "loss": 1.0207, + "step": 6018 + }, + { + "epoch": 0.5428146277675069, + "grad_norm": 1.2681804806909502, + "learning_rate": 1.8200148506958397e-06, + "loss": 0.9645, + "step": 6019 + }, + { + "epoch": 0.5429048112909771, + "grad_norm": 0.6805361568731048, + "learning_rate": 1.819433018226308e-06, + "loss": 0.8079, + "step": 6020 + }, + { + "epoch": 0.5429949948144474, + "grad_norm": 0.63032431100154, + "learning_rate": 1.8188512011637471e-06, + "loss": 0.7516, + "step": 6021 + }, + { + "epoch": 0.5430851783379177, + "grad_norm": 1.8199118635823786, + "learning_rate": 1.8182693995578e-06, + "loss": 0.9313, + "step": 6022 + }, + { + "epoch": 0.5431753618613879, + "grad_norm": 1.3323852874018003, + "learning_rate": 1.8176876134581098e-06, + "loss": 0.8312, + "step": 6023 + }, + { + "epoch": 0.5432655453848582, + "grad_norm": 1.663053394655548, + "learning_rate": 1.8171058429143176e-06, + "loss": 0.9063, + "step": 6024 + }, + { + "epoch": 0.5433557289083284, + "grad_norm": 1.3896685965410172, + "learning_rate": 1.8165240879760637e-06, + "loss": 0.9915, + "step": 6025 + }, + { + "epoch": 0.5434459124317987, + "grad_norm": 1.516356144371594, + "learning_rate": 1.8159423486929862e-06, + "loss": 0.9831, + "step": 6026 + }, + { + "epoch": 0.543536095955269, + "grad_norm": 1.5392455171312558, + "learning_rate": 1.815360625114722e-06, + "loss": 0.9544, + "step": 6027 + }, + { + "epoch": 0.5436262794787392, + "grad_norm": 1.5239574620976004, + "learning_rate": 1.814778917290908e-06, + "loss": 0.9383, + "step": 6028 + }, + { + "epoch": 0.5437164630022095, + "grad_norm": 1.6922288264391119, + "learning_rate": 1.8141972252711773e-06, + "loss": 0.971, + "step": 6029 + }, + { + "epoch": 0.5438066465256798, + "grad_norm": 1.6120589754708474, + "learning_rate": 1.8136155491051645e-06, + "loss": 0.952, + "step": 6030 + }, + { + "epoch": 0.54389683004915, + "grad_norm": 0.7826551008834826, + "learning_rate": 1.8130338888424998e-06, + "loss": 0.7958, + "step": 6031 + }, + { + "epoch": 0.5439870135726202, + "grad_norm": 1.6367260334153277, + "learning_rate": 1.812452244532816e-06, + "loss": 0.9749, + "step": 6032 + }, + { + "epoch": 0.5440771970960906, + "grad_norm": 1.5742771336635903, + "learning_rate": 1.8118706162257405e-06, + "loss": 0.8946, + "step": 6033 + }, + { + "epoch": 0.5441673806195608, + "grad_norm": 1.4417636500612503, + "learning_rate": 1.8112890039709002e-06, + "loss": 1.0371, + "step": 6034 + }, + { + "epoch": 0.5442575641430311, + "grad_norm": 1.6223950414109682, + "learning_rate": 1.8107074078179238e-06, + "loss": 0.9579, + "step": 6035 + }, + { + "epoch": 0.5443477476665013, + "grad_norm": 1.3812463371953256, + "learning_rate": 1.8101258278164348e-06, + "loss": 0.9083, + "step": 6036 + }, + { + "epoch": 0.5444379311899716, + "grad_norm": 1.5384688418471772, + "learning_rate": 1.8095442640160575e-06, + "loss": 1.0616, + "step": 6037 + }, + { + "epoch": 0.5445281147134419, + "grad_norm": 1.8607797525393293, + "learning_rate": 1.8089627164664132e-06, + "loss": 0.9195, + "step": 6038 + }, + { + "epoch": 0.5446182982369121, + "grad_norm": 1.2632895691738677, + "learning_rate": 1.8083811852171233e-06, + "loss": 0.874, + "step": 6039 + }, + { + "epoch": 0.5447084817603823, + "grad_norm": 1.6783343875329888, + "learning_rate": 1.8077996703178078e-06, + "loss": 0.9574, + "step": 6040 + }, + { + "epoch": 0.5447986652838527, + "grad_norm": 1.3186641990627606, + "learning_rate": 1.8072181718180833e-06, + "loss": 0.9022, + "step": 6041 + }, + { + "epoch": 0.5448888488073229, + "grad_norm": 1.6878215342186123, + "learning_rate": 1.806636689767568e-06, + "loss": 1.0392, + "step": 6042 + }, + { + "epoch": 0.5449790323307931, + "grad_norm": 1.4417528185524997, + "learning_rate": 1.8060552242158769e-06, + "loss": 0.8947, + "step": 6043 + }, + { + "epoch": 0.5450692158542634, + "grad_norm": 1.3873992110065505, + "learning_rate": 1.8054737752126224e-06, + "loss": 0.9246, + "step": 6044 + }, + { + "epoch": 0.5451593993777337, + "grad_norm": 1.4708043604402536, + "learning_rate": 1.804892342807419e-06, + "loss": 0.9426, + "step": 6045 + }, + { + "epoch": 0.545249582901204, + "grad_norm": 1.9347245425925208, + "learning_rate": 1.8043109270498756e-06, + "loss": 1.0271, + "step": 6046 + }, + { + "epoch": 0.5453397664246742, + "grad_norm": 1.3452818367506258, + "learning_rate": 1.803729527989604e-06, + "loss": 0.9659, + "step": 6047 + }, + { + "epoch": 0.5454299499481444, + "grad_norm": 1.5845440535483695, + "learning_rate": 1.8031481456762112e-06, + "loss": 0.9491, + "step": 6048 + }, + { + "epoch": 0.5455201334716148, + "grad_norm": 1.6229856917948704, + "learning_rate": 1.8025667801593033e-06, + "loss": 0.8851, + "step": 6049 + }, + { + "epoch": 0.545610316995085, + "grad_norm": 1.6172240617092872, + "learning_rate": 1.8019854314884871e-06, + "loss": 0.8907, + "step": 6050 + }, + { + "epoch": 0.5457005005185552, + "grad_norm": 1.3652316244858962, + "learning_rate": 1.8014040997133652e-06, + "loss": 1.0056, + "step": 6051 + }, + { + "epoch": 0.5457906840420256, + "grad_norm": 0.7924179268706371, + "learning_rate": 1.8008227848835414e-06, + "loss": 0.8162, + "step": 6052 + }, + { + "epoch": 0.5458808675654958, + "grad_norm": 1.9208803161480008, + "learning_rate": 1.8002414870486144e-06, + "loss": 0.9067, + "step": 6053 + }, + { + "epoch": 0.545971051088966, + "grad_norm": 1.2668684514257087, + "learning_rate": 1.7996602062581864e-06, + "loss": 0.907, + "step": 6054 + }, + { + "epoch": 0.5460612346124363, + "grad_norm": 1.538488755535582, + "learning_rate": 1.7990789425618544e-06, + "loss": 1.0025, + "step": 6055 + }, + { + "epoch": 0.5461514181359066, + "grad_norm": 1.3150728983671265, + "learning_rate": 1.7984976960092137e-06, + "loss": 0.9404, + "step": 6056 + }, + { + "epoch": 0.5462416016593769, + "grad_norm": 1.8020399746244957, + "learning_rate": 1.7979164666498617e-06, + "loss": 0.9231, + "step": 6057 + }, + { + "epoch": 0.5463317851828471, + "grad_norm": 0.7034316559815297, + "learning_rate": 1.7973352545333901e-06, + "loss": 0.7553, + "step": 6058 + }, + { + "epoch": 0.5464219687063173, + "grad_norm": 1.6766456206551112, + "learning_rate": 1.796754059709393e-06, + "loss": 0.8752, + "step": 6059 + }, + { + "epoch": 0.5465121522297877, + "grad_norm": 1.561245910907946, + "learning_rate": 1.7961728822274603e-06, + "loss": 1.0163, + "step": 6060 + }, + { + "epoch": 0.5466023357532579, + "grad_norm": 1.429060584858859, + "learning_rate": 1.7955917221371802e-06, + "loss": 0.9093, + "step": 6061 + }, + { + "epoch": 0.5466925192767281, + "grad_norm": 2.2154262523528643, + "learning_rate": 1.7950105794881422e-06, + "loss": 0.873, + "step": 6062 + }, + { + "epoch": 0.5467827028001984, + "grad_norm": 1.649036389076041, + "learning_rate": 1.7944294543299317e-06, + "loss": 0.9404, + "step": 6063 + }, + { + "epoch": 0.5468728863236687, + "grad_norm": 1.6121164324055755, + "learning_rate": 1.7938483467121333e-06, + "loss": 1.0111, + "step": 6064 + }, + { + "epoch": 0.5469630698471389, + "grad_norm": 1.6222626298214962, + "learning_rate": 1.7932672566843313e-06, + "loss": 0.9403, + "step": 6065 + }, + { + "epoch": 0.5470532533706092, + "grad_norm": 1.5595648090071963, + "learning_rate": 1.7926861842961065e-06, + "loss": 0.8801, + "step": 6066 + }, + { + "epoch": 0.5471434368940794, + "grad_norm": 1.4727350107679007, + "learning_rate": 1.7921051295970399e-06, + "loss": 0.8758, + "step": 6067 + }, + { + "epoch": 0.5472336204175497, + "grad_norm": 1.6941486354055348, + "learning_rate": 1.7915240926367092e-06, + "loss": 0.917, + "step": 6068 + }, + { + "epoch": 0.54732380394102, + "grad_norm": 2.1311502651471383, + "learning_rate": 1.7909430734646932e-06, + "loss": 0.9976, + "step": 6069 + }, + { + "epoch": 0.5474139874644902, + "grad_norm": 0.7876934222693621, + "learning_rate": 1.790362072130567e-06, + "loss": 0.7882, + "step": 6070 + }, + { + "epoch": 0.5475041709879604, + "grad_norm": 1.7788644845220287, + "learning_rate": 1.7897810886839037e-06, + "loss": 0.9444, + "step": 6071 + }, + { + "epoch": 0.5475943545114308, + "grad_norm": 1.522552467907871, + "learning_rate": 1.7892001231742782e-06, + "loss": 0.9666, + "step": 6072 + }, + { + "epoch": 0.547684538034901, + "grad_norm": 1.4821801271200934, + "learning_rate": 1.7886191756512594e-06, + "loss": 0.9607, + "step": 6073 + }, + { + "epoch": 0.5477747215583713, + "grad_norm": 1.3946583153032133, + "learning_rate": 1.7880382461644192e-06, + "loss": 1.0309, + "step": 6074 + }, + { + "epoch": 0.5478649050818416, + "grad_norm": 1.4240356628956343, + "learning_rate": 1.7874573347633235e-06, + "loss": 0.9311, + "step": 6075 + }, + { + "epoch": 0.5479550886053118, + "grad_norm": 1.5866738850699176, + "learning_rate": 1.7868764414975408e-06, + "loss": 1.0284, + "step": 6076 + }, + { + "epoch": 0.5480452721287821, + "grad_norm": 1.7000102687974339, + "learning_rate": 1.7862955664166353e-06, + "loss": 0.8999, + "step": 6077 + }, + { + "epoch": 0.5481354556522523, + "grad_norm": 1.3425725613240982, + "learning_rate": 1.78571470957017e-06, + "loss": 1.038, + "step": 6078 + }, + { + "epoch": 0.5482256391757226, + "grad_norm": 1.6278567512342916, + "learning_rate": 1.7851338710077074e-06, + "loss": 1.0204, + "step": 6079 + }, + { + "epoch": 0.5483158226991929, + "grad_norm": 1.6026973518336851, + "learning_rate": 1.7845530507788076e-06, + "loss": 1.0643, + "step": 6080 + }, + { + "epoch": 0.5484060062226631, + "grad_norm": 1.4143984063706978, + "learning_rate": 1.7839722489330298e-06, + "loss": 0.9257, + "step": 6081 + }, + { + "epoch": 0.5484961897461333, + "grad_norm": 1.429599363184714, + "learning_rate": 1.7833914655199308e-06, + "loss": 0.926, + "step": 6082 + }, + { + "epoch": 0.5485863732696037, + "grad_norm": 1.6658829594837947, + "learning_rate": 1.7828107005890658e-06, + "loss": 0.998, + "step": 6083 + }, + { + "epoch": 0.5486765567930739, + "grad_norm": 1.965277803987908, + "learning_rate": 1.7822299541899898e-06, + "loss": 0.874, + "step": 6084 + }, + { + "epoch": 0.5487667403165442, + "grad_norm": 1.3351366643211289, + "learning_rate": 1.7816492263722545e-06, + "loss": 0.9536, + "step": 6085 + }, + { + "epoch": 0.5488569238400144, + "grad_norm": 1.6006861824637035, + "learning_rate": 1.781068517185412e-06, + "loss": 0.9801, + "step": 6086 + }, + { + "epoch": 0.5489471073634847, + "grad_norm": 1.4158877400495018, + "learning_rate": 1.7804878266790104e-06, + "loss": 1.0236, + "step": 6087 + }, + { + "epoch": 0.549037290886955, + "grad_norm": 1.416854462611473, + "learning_rate": 1.779907154902597e-06, + "loss": 0.8897, + "step": 6088 + }, + { + "epoch": 0.5491274744104252, + "grad_norm": 1.300678237639368, + "learning_rate": 1.7793265019057198e-06, + "loss": 0.8792, + "step": 6089 + }, + { + "epoch": 0.5492176579338954, + "grad_norm": 0.8444397544172894, + "learning_rate": 1.7787458677379212e-06, + "loss": 0.7961, + "step": 6090 + }, + { + "epoch": 0.5493078414573658, + "grad_norm": 1.4385344059250127, + "learning_rate": 1.7781652524487463e-06, + "loss": 1.0132, + "step": 6091 + }, + { + "epoch": 0.549398024980836, + "grad_norm": 1.6193173890667059, + "learning_rate": 1.777584656087735e-06, + "loss": 0.9031, + "step": 6092 + }, + { + "epoch": 0.5494882085043062, + "grad_norm": 1.5631035211392275, + "learning_rate": 1.777004078704427e-06, + "loss": 0.8713, + "step": 6093 + }, + { + "epoch": 0.5495783920277765, + "grad_norm": 1.5538502648089285, + "learning_rate": 1.7764235203483603e-06, + "loss": 0.9806, + "step": 6094 + }, + { + "epoch": 0.5496685755512468, + "grad_norm": 1.7026114433213206, + "learning_rate": 1.775842981069072e-06, + "loss": 0.9291, + "step": 6095 + }, + { + "epoch": 0.549758759074717, + "grad_norm": 1.5952733557192553, + "learning_rate": 1.7752624609160966e-06, + "loss": 0.9649, + "step": 6096 + }, + { + "epoch": 0.5498489425981873, + "grad_norm": 1.835163099228252, + "learning_rate": 1.7746819599389665e-06, + "loss": 0.9933, + "step": 6097 + }, + { + "epoch": 0.5499391261216575, + "grad_norm": 1.5593856291322676, + "learning_rate": 1.774101478187215e-06, + "loss": 0.9609, + "step": 6098 + }, + { + "epoch": 0.5500293096451279, + "grad_norm": 1.8366033766908125, + "learning_rate": 1.773521015710371e-06, + "loss": 1.0289, + "step": 6099 + }, + { + "epoch": 0.5501194931685981, + "grad_norm": 1.6859700366849968, + "learning_rate": 1.7729405725579614e-06, + "loss": 1.0358, + "step": 6100 + }, + { + "epoch": 0.5502096766920683, + "grad_norm": 1.7539612351530338, + "learning_rate": 1.7723601487795151e-06, + "loss": 0.9308, + "step": 6101 + }, + { + "epoch": 0.5502998602155387, + "grad_norm": 1.3763901012597946, + "learning_rate": 1.7717797444245557e-06, + "loss": 0.9034, + "step": 6102 + }, + { + "epoch": 0.5503900437390089, + "grad_norm": 1.3289760331317073, + "learning_rate": 1.7711993595426076e-06, + "loss": 0.886, + "step": 6103 + }, + { + "epoch": 0.5504802272624791, + "grad_norm": 1.6304938540607514, + "learning_rate": 1.7706189941831915e-06, + "loss": 1.0671, + "step": 6104 + }, + { + "epoch": 0.5505704107859494, + "grad_norm": 1.53550921937143, + "learning_rate": 1.770038648395827e-06, + "loss": 0.9174, + "step": 6105 + }, + { + "epoch": 0.5506605943094197, + "grad_norm": 0.671833037019647, + "learning_rate": 1.7694583222300336e-06, + "loss": 0.7958, + "step": 6106 + }, + { + "epoch": 0.55075077783289, + "grad_norm": 1.5461050101588685, + "learning_rate": 1.7688780157353272e-06, + "loss": 0.935, + "step": 6107 + }, + { + "epoch": 0.5508409613563602, + "grad_norm": 1.368226664579224, + "learning_rate": 1.768297728961223e-06, + "loss": 0.8929, + "step": 6108 + }, + { + "epoch": 0.5509311448798304, + "grad_norm": 1.8793498128326565, + "learning_rate": 1.7677174619572342e-06, + "loss": 0.9647, + "step": 6109 + }, + { + "epoch": 0.5510213284033008, + "grad_norm": 1.6276342607894907, + "learning_rate": 1.7671372147728717e-06, + "loss": 0.9742, + "step": 6110 + }, + { + "epoch": 0.551111511926771, + "grad_norm": 1.7561677825254194, + "learning_rate": 1.7665569874576471e-06, + "loss": 1.0211, + "step": 6111 + }, + { + "epoch": 0.5512016954502412, + "grad_norm": 1.622107058114044, + "learning_rate": 1.7659767800610664e-06, + "loss": 0.9583, + "step": 6112 + }, + { + "epoch": 0.5512918789737115, + "grad_norm": 1.623153958420333, + "learning_rate": 1.7653965926326379e-06, + "loss": 0.8642, + "step": 6113 + }, + { + "epoch": 0.5513820624971818, + "grad_norm": 1.1973886348889071, + "learning_rate": 1.764816425221866e-06, + "loss": 0.9103, + "step": 6114 + }, + { + "epoch": 0.551472246020652, + "grad_norm": 1.393728372380389, + "learning_rate": 1.7642362778782524e-06, + "loss": 0.8735, + "step": 6115 + }, + { + "epoch": 0.5515624295441223, + "grad_norm": 1.6232622832246706, + "learning_rate": 1.7636561506513005e-06, + "loss": 0.927, + "step": 6116 + }, + { + "epoch": 0.5516526130675925, + "grad_norm": 1.3307273501429193, + "learning_rate": 1.7630760435905083e-06, + "loss": 0.9307, + "step": 6117 + }, + { + "epoch": 0.5517427965910628, + "grad_norm": 1.8003117741677839, + "learning_rate": 1.762495956745375e-06, + "loss": 0.9178, + "step": 6118 + }, + { + "epoch": 0.5518329801145331, + "grad_norm": 2.273591760599507, + "learning_rate": 1.7619158901653962e-06, + "loss": 0.999, + "step": 6119 + }, + { + "epoch": 0.5519231636380033, + "grad_norm": 1.5144578037775127, + "learning_rate": 1.761335843900066e-06, + "loss": 0.9538, + "step": 6120 + }, + { + "epoch": 0.5520133471614735, + "grad_norm": 1.38376903317533, + "learning_rate": 1.7607558179988785e-06, + "loss": 0.9382, + "step": 6121 + }, + { + "epoch": 0.5521035306849439, + "grad_norm": 1.5149515759877907, + "learning_rate": 1.760175812511323e-06, + "loss": 0.8866, + "step": 6122 + }, + { + "epoch": 0.5521937142084141, + "grad_norm": 1.5852946965958512, + "learning_rate": 1.75959582748689e-06, + "loss": 0.9107, + "step": 6123 + }, + { + "epoch": 0.5522838977318844, + "grad_norm": 1.4632681956526108, + "learning_rate": 1.7590158629750657e-06, + "loss": 0.985, + "step": 6124 + }, + { + "epoch": 0.5523740812553547, + "grad_norm": 1.3392132312597302, + "learning_rate": 1.7584359190253376e-06, + "loss": 1.0079, + "step": 6125 + }, + { + "epoch": 0.5524642647788249, + "grad_norm": 1.7708793708950354, + "learning_rate": 1.7578559956871892e-06, + "loss": 1.0184, + "step": 6126 + }, + { + "epoch": 0.5525544483022952, + "grad_norm": 1.3388743549936932, + "learning_rate": 1.7572760930101012e-06, + "loss": 0.953, + "step": 6127 + }, + { + "epoch": 0.5526446318257654, + "grad_norm": 0.730557319045104, + "learning_rate": 1.7566962110435563e-06, + "loss": 0.8345, + "step": 6128 + }, + { + "epoch": 0.5527348153492357, + "grad_norm": 1.3689607292544834, + "learning_rate": 1.7561163498370313e-06, + "loss": 0.9737, + "step": 6129 + }, + { + "epoch": 0.552824998872706, + "grad_norm": 2.3564572513813804, + "learning_rate": 1.755536509440005e-06, + "loss": 0.9214, + "step": 6130 + }, + { + "epoch": 0.5529151823961762, + "grad_norm": 1.8237334755664638, + "learning_rate": 1.7549566899019519e-06, + "loss": 0.9535, + "step": 6131 + }, + { + "epoch": 0.5530053659196464, + "grad_norm": 1.5692758507136133, + "learning_rate": 1.754376891272344e-06, + "loss": 0.9688, + "step": 6132 + }, + { + "epoch": 0.5530955494431168, + "grad_norm": 2.232094514595792, + "learning_rate": 1.753797113600655e-06, + "loss": 0.906, + "step": 6133 + }, + { + "epoch": 0.553185732966587, + "grad_norm": 1.5803168155429446, + "learning_rate": 1.7532173569363535e-06, + "loss": 1.0903, + "step": 6134 + }, + { + "epoch": 0.5532759164900573, + "grad_norm": 2.602613906786143, + "learning_rate": 1.7526376213289077e-06, + "loss": 0.7715, + "step": 6135 + }, + { + "epoch": 0.5533661000135275, + "grad_norm": 1.4416397856629948, + "learning_rate": 1.7520579068277844e-06, + "loss": 0.9375, + "step": 6136 + }, + { + "epoch": 0.5534562835369978, + "grad_norm": 1.3879596155156562, + "learning_rate": 1.7514782134824472e-06, + "loss": 0.9856, + "step": 6137 + }, + { + "epoch": 0.5535464670604681, + "grad_norm": 1.5922942151066675, + "learning_rate": 1.7508985413423599e-06, + "loss": 0.949, + "step": 6138 + }, + { + "epoch": 0.5536366505839383, + "grad_norm": 1.599010686193356, + "learning_rate": 1.7503188904569814e-06, + "loss": 1.0239, + "step": 6139 + }, + { + "epoch": 0.5537268341074085, + "grad_norm": 1.5996030404474315, + "learning_rate": 1.7497392608757728e-06, + "loss": 0.9228, + "step": 6140 + }, + { + "epoch": 0.5538170176308789, + "grad_norm": 1.477258185335091, + "learning_rate": 1.7491596526481897e-06, + "loss": 0.9732, + "step": 6141 + }, + { + "epoch": 0.5539072011543491, + "grad_norm": 1.4416206014399537, + "learning_rate": 1.7485800658236888e-06, + "loss": 0.882, + "step": 6142 + }, + { + "epoch": 0.5539973846778193, + "grad_norm": 2.113501574300699, + "learning_rate": 1.7480005004517232e-06, + "loss": 1.0279, + "step": 6143 + }, + { + "epoch": 0.5540875682012896, + "grad_norm": 2.078519855359466, + "learning_rate": 1.7474209565817435e-06, + "loss": 1.0204, + "step": 6144 + }, + { + "epoch": 0.5541777517247599, + "grad_norm": 1.4873078935454611, + "learning_rate": 1.7468414342632014e-06, + "loss": 0.9553, + "step": 6145 + }, + { + "epoch": 0.5542679352482301, + "grad_norm": 1.6015242409206898, + "learning_rate": 1.746261933545543e-06, + "loss": 1.0095, + "step": 6146 + }, + { + "epoch": 0.5543581187717004, + "grad_norm": 1.9598804679739343, + "learning_rate": 1.7456824544782165e-06, + "loss": 0.9512, + "step": 6147 + }, + { + "epoch": 0.5544483022951707, + "grad_norm": 1.3428076611405708, + "learning_rate": 1.7451029971106653e-06, + "loss": 0.9587, + "step": 6148 + }, + { + "epoch": 0.554538485818641, + "grad_norm": 1.635533452166827, + "learning_rate": 1.7445235614923313e-06, + "loss": 0.9814, + "step": 6149 + }, + { + "epoch": 0.5546286693421112, + "grad_norm": 1.4166825611494416, + "learning_rate": 1.7439441476726556e-06, + "loss": 0.946, + "step": 6150 + }, + { + "epoch": 0.5547188528655814, + "grad_norm": 1.5787285887413263, + "learning_rate": 1.7433647557010776e-06, + "loss": 0.9666, + "step": 6151 + }, + { + "epoch": 0.5548090363890518, + "grad_norm": 1.4602874288727223, + "learning_rate": 1.7427853856270338e-06, + "loss": 0.9181, + "step": 6152 + }, + { + "epoch": 0.554899219912522, + "grad_norm": 1.5680680057754812, + "learning_rate": 1.7422060374999593e-06, + "loss": 0.9201, + "step": 6153 + }, + { + "epoch": 0.5549894034359922, + "grad_norm": 4.3032736005313765, + "learning_rate": 1.7416267113692862e-06, + "loss": 0.913, + "step": 6154 + }, + { + "epoch": 0.5550795869594625, + "grad_norm": 1.4650604494662363, + "learning_rate": 1.7410474072844475e-06, + "loss": 0.9855, + "step": 6155 + }, + { + "epoch": 0.5551697704829328, + "grad_norm": 1.4758458088556283, + "learning_rate": 1.740468125294871e-06, + "loss": 0.938, + "step": 6156 + }, + { + "epoch": 0.555259954006403, + "grad_norm": 4.548952657027649, + "learning_rate": 1.739888865449986e-06, + "loss": 0.8662, + "step": 6157 + }, + { + "epoch": 0.5553501375298733, + "grad_norm": 1.669422525624139, + "learning_rate": 1.7393096277992174e-06, + "loss": 0.9773, + "step": 6158 + }, + { + "epoch": 0.5554403210533435, + "grad_norm": 1.4976992446228148, + "learning_rate": 1.738730412391988e-06, + "loss": 0.9669, + "step": 6159 + }, + { + "epoch": 0.5555305045768139, + "grad_norm": 1.372115012964582, + "learning_rate": 1.738151219277721e-06, + "loss": 1.005, + "step": 6160 + }, + { + "epoch": 0.5556206881002841, + "grad_norm": 1.7326741169329858, + "learning_rate": 1.7375720485058349e-06, + "loss": 0.9112, + "step": 6161 + }, + { + "epoch": 0.5557108716237543, + "grad_norm": 1.744442357259799, + "learning_rate": 1.7369929001257498e-06, + "loss": 0.9675, + "step": 6162 + }, + { + "epoch": 0.5558010551472246, + "grad_norm": 2.108888584144087, + "learning_rate": 1.73641377418688e-06, + "loss": 0.8726, + "step": 6163 + }, + { + "epoch": 0.5558912386706949, + "grad_norm": 1.6830274864305, + "learning_rate": 1.7358346707386408e-06, + "loss": 0.8914, + "step": 6164 + }, + { + "epoch": 0.5559814221941651, + "grad_norm": 1.7155006905634664, + "learning_rate": 1.7352555898304439e-06, + "loss": 1.0066, + "step": 6165 + }, + { + "epoch": 0.5560716057176354, + "grad_norm": 7.821176090632011, + "learning_rate": 1.7346765315116996e-06, + "loss": 0.8765, + "step": 6166 + }, + { + "epoch": 0.5561617892411056, + "grad_norm": 1.582995989482954, + "learning_rate": 1.734097495831817e-06, + "loss": 0.917, + "step": 6167 + }, + { + "epoch": 0.5562519727645759, + "grad_norm": 1.5314579258095145, + "learning_rate": 1.7335184828402015e-06, + "loss": 0.9697, + "step": 6168 + }, + { + "epoch": 0.5563421562880462, + "grad_norm": 1.4738925429400054, + "learning_rate": 1.7329394925862595e-06, + "loss": 0.9331, + "step": 6169 + }, + { + "epoch": 0.5564323398115164, + "grad_norm": 2.494078872143219, + "learning_rate": 1.7323605251193922e-06, + "loss": 1.0043, + "step": 6170 + }, + { + "epoch": 0.5565225233349868, + "grad_norm": 1.7147215363002828, + "learning_rate": 1.7317815804890001e-06, + "loss": 0.973, + "step": 6171 + }, + { + "epoch": 0.556612706858457, + "grad_norm": 1.6228168935145144, + "learning_rate": 1.731202658744483e-06, + "loss": 0.9087, + "step": 6172 + }, + { + "epoch": 0.5567028903819272, + "grad_norm": 1.3102314280878897, + "learning_rate": 1.7306237599352365e-06, + "loss": 1.019, + "step": 6173 + }, + { + "epoch": 0.5567930739053975, + "grad_norm": 1.3403943768229936, + "learning_rate": 1.730044884110657e-06, + "loss": 0.9422, + "step": 6174 + }, + { + "epoch": 0.5568832574288678, + "grad_norm": 1.505843779684768, + "learning_rate": 1.7294660313201366e-06, + "loss": 0.9141, + "step": 6175 + }, + { + "epoch": 0.556973440952338, + "grad_norm": 2.5271284669104266, + "learning_rate": 1.7288872016130652e-06, + "loss": 0.9324, + "step": 6176 + }, + { + "epoch": 0.5570636244758083, + "grad_norm": 1.5048020110324236, + "learning_rate": 1.7283083950388334e-06, + "loss": 0.8701, + "step": 6177 + }, + { + "epoch": 0.5571538079992785, + "grad_norm": 1.5020751109892423, + "learning_rate": 1.727729611646827e-06, + "loss": 0.861, + "step": 6178 + }, + { + "epoch": 0.5572439915227488, + "grad_norm": 1.5732563569882936, + "learning_rate": 1.7271508514864318e-06, + "loss": 1.0085, + "step": 6179 + }, + { + "epoch": 0.5573341750462191, + "grad_norm": 1.9245032350706728, + "learning_rate": 1.7265721146070302e-06, + "loss": 0.9579, + "step": 6180 + }, + { + "epoch": 0.5574243585696893, + "grad_norm": 1.4927931749715286, + "learning_rate": 1.7259934010580035e-06, + "loss": 0.9517, + "step": 6181 + }, + { + "epoch": 0.5575145420931595, + "grad_norm": 1.4072647248387389, + "learning_rate": 1.725414710888731e-06, + "loss": 0.9039, + "step": 6182 + }, + { + "epoch": 0.5576047256166299, + "grad_norm": 1.3762382220631328, + "learning_rate": 1.7248360441485885e-06, + "loss": 0.9308, + "step": 6183 + }, + { + "epoch": 0.5576949091401001, + "grad_norm": 1.6010419045991873, + "learning_rate": 1.7242574008869528e-06, + "loss": 1.0274, + "step": 6184 + }, + { + "epoch": 0.5577850926635703, + "grad_norm": 0.7269136134196531, + "learning_rate": 1.7236787811531951e-06, + "loss": 0.8308, + "step": 6185 + }, + { + "epoch": 0.5578752761870406, + "grad_norm": 1.586176962226872, + "learning_rate": 1.7231001849966887e-06, + "loss": 0.9248, + "step": 6186 + }, + { + "epoch": 0.5579654597105109, + "grad_norm": 0.6705049251913148, + "learning_rate": 1.722521612466801e-06, + "loss": 0.8091, + "step": 6187 + }, + { + "epoch": 0.5580556432339812, + "grad_norm": 1.6032699043564718, + "learning_rate": 1.7219430636128984e-06, + "loss": 0.896, + "step": 6188 + }, + { + "epoch": 0.5581458267574514, + "grad_norm": 1.4484696007127118, + "learning_rate": 1.7213645384843479e-06, + "loss": 0.8958, + "step": 6189 + }, + { + "epoch": 0.5582360102809216, + "grad_norm": 1.4903031679526377, + "learning_rate": 1.7207860371305108e-06, + "loss": 0.9461, + "step": 6190 + }, + { + "epoch": 0.558326193804392, + "grad_norm": 1.5584956164526338, + "learning_rate": 1.7202075596007487e-06, + "loss": 0.9162, + "step": 6191 + }, + { + "epoch": 0.5584163773278622, + "grad_norm": 1.722958216075846, + "learning_rate": 1.7196291059444206e-06, + "loss": 0.9155, + "step": 6192 + }, + { + "epoch": 0.5585065608513324, + "grad_norm": 1.5551658929483516, + "learning_rate": 1.7190506762108828e-06, + "loss": 0.9353, + "step": 6193 + }, + { + "epoch": 0.5585967443748028, + "grad_norm": 1.5768180949995525, + "learning_rate": 1.7184722704494907e-06, + "loss": 0.9502, + "step": 6194 + }, + { + "epoch": 0.558686927898273, + "grad_norm": 1.51224258376132, + "learning_rate": 1.717893888709596e-06, + "loss": 0.9656, + "step": 6195 + }, + { + "epoch": 0.5587771114217432, + "grad_norm": 1.574014276519749, + "learning_rate": 1.7173155310405515e-06, + "loss": 0.9814, + "step": 6196 + }, + { + "epoch": 0.5588672949452135, + "grad_norm": 1.3530935094397896, + "learning_rate": 1.7167371974917043e-06, + "loss": 1.0286, + "step": 6197 + }, + { + "epoch": 0.5589574784686838, + "grad_norm": 1.6940945236199503, + "learning_rate": 1.7161588881124003e-06, + "loss": 0.975, + "step": 6198 + }, + { + "epoch": 0.559047661992154, + "grad_norm": 1.6084826319417629, + "learning_rate": 1.7155806029519861e-06, + "loss": 0.9382, + "step": 6199 + }, + { + "epoch": 0.5591378455156243, + "grad_norm": 1.7012474840357197, + "learning_rate": 1.7150023420598023e-06, + "loss": 0.993, + "step": 6200 + }, + { + "epoch": 0.5592280290390945, + "grad_norm": 1.380757674771069, + "learning_rate": 1.714424105485191e-06, + "loss": 1.0089, + "step": 6201 + }, + { + "epoch": 0.5593182125625649, + "grad_norm": 1.8467938787779021, + "learning_rate": 1.7138458932774896e-06, + "loss": 0.9157, + "step": 6202 + }, + { + "epoch": 0.5594083960860351, + "grad_norm": 0.7264376297161056, + "learning_rate": 1.7132677054860335e-06, + "loss": 0.8084, + "step": 6203 + }, + { + "epoch": 0.5594985796095053, + "grad_norm": 1.5131820181401123, + "learning_rate": 1.7126895421601586e-06, + "loss": 0.9798, + "step": 6204 + }, + { + "epoch": 0.5595887631329756, + "grad_norm": 1.3130669504631591, + "learning_rate": 1.712111403349196e-06, + "loss": 0.8944, + "step": 6205 + }, + { + "epoch": 0.5596789466564459, + "grad_norm": 1.387603262761718, + "learning_rate": 1.7115332891024757e-06, + "loss": 0.9876, + "step": 6206 + }, + { + "epoch": 0.5597691301799161, + "grad_norm": 0.7442138315471166, + "learning_rate": 1.7109551994693257e-06, + "loss": 0.858, + "step": 6207 + }, + { + "epoch": 0.5598593137033864, + "grad_norm": 1.6727115997891973, + "learning_rate": 1.7103771344990725e-06, + "loss": 0.8929, + "step": 6208 + }, + { + "epoch": 0.5599494972268566, + "grad_norm": 1.6026617232202551, + "learning_rate": 1.709799094241039e-06, + "loss": 0.8702, + "step": 6209 + }, + { + "epoch": 0.560039680750327, + "grad_norm": 1.5398973243718375, + "learning_rate": 1.709221078744546e-06, + "loss": 0.9877, + "step": 6210 + }, + { + "epoch": 0.5601298642737972, + "grad_norm": 1.317160008551984, + "learning_rate": 1.7086430880589148e-06, + "loss": 0.9595, + "step": 6211 + }, + { + "epoch": 0.5602200477972674, + "grad_norm": 1.555795782515006, + "learning_rate": 1.7080651222334612e-06, + "loss": 0.9584, + "step": 6212 + }, + { + "epoch": 0.5603102313207377, + "grad_norm": 1.6185239715073714, + "learning_rate": 1.7074871813175018e-06, + "loss": 1.0142, + "step": 6213 + }, + { + "epoch": 0.560400414844208, + "grad_norm": 1.510260694496545, + "learning_rate": 1.706909265360349e-06, + "loss": 0.8646, + "step": 6214 + }, + { + "epoch": 0.5604905983676782, + "grad_norm": 0.7718978353358079, + "learning_rate": 1.7063313744113128e-06, + "loss": 0.7978, + "step": 6215 + }, + { + "epoch": 0.5605807818911485, + "grad_norm": 0.8414959292140513, + "learning_rate": 1.7057535085197042e-06, + "loss": 0.7707, + "step": 6216 + }, + { + "epoch": 0.5606709654146187, + "grad_norm": 1.6449350327993921, + "learning_rate": 1.705175667734828e-06, + "loss": 0.8762, + "step": 6217 + }, + { + "epoch": 0.560761148938089, + "grad_norm": 1.9066278051982068, + "learning_rate": 1.7045978521059894e-06, + "loss": 0.9238, + "step": 6218 + }, + { + "epoch": 0.5608513324615593, + "grad_norm": 1.4382782363652822, + "learning_rate": 1.7040200616824914e-06, + "loss": 0.9508, + "step": 6219 + }, + { + "epoch": 0.5609415159850295, + "grad_norm": 1.5511992921618085, + "learning_rate": 1.7034422965136333e-06, + "loss": 0.8968, + "step": 6220 + }, + { + "epoch": 0.5610316995084998, + "grad_norm": 1.7592751754167182, + "learning_rate": 1.7028645566487137e-06, + "loss": 0.9954, + "step": 6221 + }, + { + "epoch": 0.5611218830319701, + "grad_norm": 1.448253052787125, + "learning_rate": 1.7022868421370284e-06, + "loss": 0.9255, + "step": 6222 + }, + { + "epoch": 0.5612120665554403, + "grad_norm": 1.522840881897304, + "learning_rate": 1.701709153027872e-06, + "loss": 0.9988, + "step": 6223 + }, + { + "epoch": 0.5613022500789105, + "grad_norm": 1.7489991050573974, + "learning_rate": 1.7011314893705353e-06, + "loss": 0.9186, + "step": 6224 + }, + { + "epoch": 0.5613924336023809, + "grad_norm": 1.5532215817545796, + "learning_rate": 1.700553851214307e-06, + "loss": 0.9546, + "step": 6225 + }, + { + "epoch": 0.5614826171258511, + "grad_norm": 1.5733180344505893, + "learning_rate": 1.699976238608476e-06, + "loss": 0.8199, + "step": 6226 + }, + { + "epoch": 0.5615728006493214, + "grad_norm": 1.738297914575141, + "learning_rate": 1.699398651602326e-06, + "loss": 0.9014, + "step": 6227 + }, + { + "epoch": 0.5616629841727916, + "grad_norm": 1.4859424546858115, + "learning_rate": 1.6988210902451413e-06, + "loss": 0.9968, + "step": 6228 + }, + { + "epoch": 0.5617531676962619, + "grad_norm": 1.6575670583747233, + "learning_rate": 1.6982435545862011e-06, + "loss": 0.9323, + "step": 6229 + }, + { + "epoch": 0.5618433512197322, + "grad_norm": 1.6926501063578085, + "learning_rate": 1.6976660446747853e-06, + "loss": 0.9586, + "step": 6230 + }, + { + "epoch": 0.5619335347432024, + "grad_norm": 2.149542285191812, + "learning_rate": 1.6970885605601696e-06, + "loss": 0.8777, + "step": 6231 + }, + { + "epoch": 0.5620237182666726, + "grad_norm": 1.6226962705846137, + "learning_rate": 1.6965111022916282e-06, + "loss": 1.0605, + "step": 6232 + }, + { + "epoch": 0.562113901790143, + "grad_norm": 1.6268295112900801, + "learning_rate": 1.6959336699184323e-06, + "loss": 0.9262, + "step": 6233 + }, + { + "epoch": 0.5622040853136132, + "grad_norm": 1.3439278595784443, + "learning_rate": 1.6953562634898529e-06, + "loss": 0.9465, + "step": 6234 + }, + { + "epoch": 0.5622942688370834, + "grad_norm": 1.9165202651570925, + "learning_rate": 1.6947788830551569e-06, + "loss": 0.9191, + "step": 6235 + }, + { + "epoch": 0.5623844523605537, + "grad_norm": 0.8340621661817641, + "learning_rate": 1.6942015286636093e-06, + "loss": 0.8313, + "step": 6236 + }, + { + "epoch": 0.562474635884024, + "grad_norm": 1.4838273493748138, + "learning_rate": 1.6936242003644735e-06, + "loss": 0.9372, + "step": 6237 + }, + { + "epoch": 0.5625648194074943, + "grad_norm": 1.6295793272492805, + "learning_rate": 1.6930468982070106e-06, + "loss": 1.0241, + "step": 6238 + }, + { + "epoch": 0.5626550029309645, + "grad_norm": 0.6543140921533714, + "learning_rate": 1.692469622240478e-06, + "loss": 0.7195, + "step": 6239 + }, + { + "epoch": 0.5627451864544347, + "grad_norm": 1.5442424683648888, + "learning_rate": 1.6918923725141339e-06, + "loss": 0.9116, + "step": 6240 + }, + { + "epoch": 0.5628353699779051, + "grad_norm": 1.5315057190661576, + "learning_rate": 1.6913151490772312e-06, + "loss": 1.0073, + "step": 6241 + }, + { + "epoch": 0.5629255535013753, + "grad_norm": 1.3208261878765986, + "learning_rate": 1.6907379519790215e-06, + "loss": 0.992, + "step": 6242 + }, + { + "epoch": 0.5630157370248455, + "grad_norm": 1.297292251566362, + "learning_rate": 1.6901607812687558e-06, + "loss": 0.8618, + "step": 6243 + }, + { + "epoch": 0.5631059205483159, + "grad_norm": 1.3561538591896283, + "learning_rate": 1.6895836369956794e-06, + "loss": 0.9069, + "step": 6244 + }, + { + "epoch": 0.5631961040717861, + "grad_norm": 0.748747415426694, + "learning_rate": 1.6890065192090402e-06, + "loss": 0.7438, + "step": 6245 + }, + { + "epoch": 0.5632862875952563, + "grad_norm": 1.7823599318676615, + "learning_rate": 1.6884294279580793e-06, + "loss": 0.8561, + "step": 6246 + }, + { + "epoch": 0.5633764711187266, + "grad_norm": 1.5160138919213957, + "learning_rate": 1.6878523632920371e-06, + "loss": 0.9764, + "step": 6247 + }, + { + "epoch": 0.5634666546421969, + "grad_norm": 1.509236350315782, + "learning_rate": 1.6872753252601525e-06, + "loss": 1.0172, + "step": 6248 + }, + { + "epoch": 0.5635568381656672, + "grad_norm": 1.5239968080111739, + "learning_rate": 1.6866983139116616e-06, + "loss": 0.8929, + "step": 6249 + }, + { + "epoch": 0.5636470216891374, + "grad_norm": 1.3721451599775236, + "learning_rate": 1.6861213292957981e-06, + "loss": 0.9199, + "step": 6250 + }, + { + "epoch": 0.5637372052126076, + "grad_norm": 1.4916348057901454, + "learning_rate": 1.685544371461793e-06, + "loss": 0.9847, + "step": 6251 + }, + { + "epoch": 0.563827388736078, + "grad_norm": 1.8346813693404653, + "learning_rate": 1.6849674404588767e-06, + "loss": 0.9734, + "step": 6252 + }, + { + "epoch": 0.5639175722595482, + "grad_norm": 1.7653647712076794, + "learning_rate": 1.6843905363362758e-06, + "loss": 0.9844, + "step": 6253 + }, + { + "epoch": 0.5640077557830184, + "grad_norm": 1.4370108891859652, + "learning_rate": 1.6838136591432136e-06, + "loss": 0.8914, + "step": 6254 + }, + { + "epoch": 0.5640979393064887, + "grad_norm": 1.5882700786957642, + "learning_rate": 1.6832368089289139e-06, + "loss": 0.9789, + "step": 6255 + }, + { + "epoch": 0.564188122829959, + "grad_norm": 1.832577817027041, + "learning_rate": 1.682659985742596e-06, + "loss": 0.9672, + "step": 6256 + }, + { + "epoch": 0.5642783063534292, + "grad_norm": 1.624456021151247, + "learning_rate": 1.6820831896334782e-06, + "loss": 0.9938, + "step": 6257 + }, + { + "epoch": 0.5643684898768995, + "grad_norm": 1.5471128030496377, + "learning_rate": 1.681506420650776e-06, + "loss": 0.9722, + "step": 6258 + }, + { + "epoch": 0.5644586734003697, + "grad_norm": 1.3357436725481582, + "learning_rate": 1.680929678843701e-06, + "loss": 0.9373, + "step": 6259 + }, + { + "epoch": 0.56454885692384, + "grad_norm": 1.414498530651744, + "learning_rate": 1.6803529642614662e-06, + "loss": 0.9979, + "step": 6260 + }, + { + "epoch": 0.5646390404473103, + "grad_norm": 1.510162498749318, + "learning_rate": 1.6797762769532785e-06, + "loss": 0.9591, + "step": 6261 + }, + { + "epoch": 0.5647292239707805, + "grad_norm": 1.3321753728763364, + "learning_rate": 1.679199616968345e-06, + "loss": 0.9833, + "step": 6262 + }, + { + "epoch": 0.5648194074942507, + "grad_norm": 1.4129880753950106, + "learning_rate": 1.6786229843558689e-06, + "loss": 0.9761, + "step": 6263 + }, + { + "epoch": 0.5649095910177211, + "grad_norm": 1.8592229187903668, + "learning_rate": 1.6780463791650514e-06, + "loss": 0.9284, + "step": 6264 + }, + { + "epoch": 0.5649997745411913, + "grad_norm": 1.4881868107186011, + "learning_rate": 1.6774698014450928e-06, + "loss": 0.8947, + "step": 6265 + }, + { + "epoch": 0.5650899580646616, + "grad_norm": 0.7066479270815044, + "learning_rate": 1.6768932512451883e-06, + "loss": 0.7802, + "step": 6266 + }, + { + "epoch": 0.5651801415881319, + "grad_norm": 1.8465940878247198, + "learning_rate": 1.676316728614534e-06, + "loss": 0.9033, + "step": 6267 + }, + { + "epoch": 0.5652703251116021, + "grad_norm": 0.6190531816315591, + "learning_rate": 1.675740233602321e-06, + "loss": 0.7894, + "step": 6268 + }, + { + "epoch": 0.5653605086350724, + "grad_norm": 1.4327968284556922, + "learning_rate": 1.6751637662577385e-06, + "loss": 0.9981, + "step": 6269 + }, + { + "epoch": 0.5654506921585426, + "grad_norm": 0.842865939937247, + "learning_rate": 1.6745873266299753e-06, + "loss": 0.7327, + "step": 6270 + }, + { + "epoch": 0.565540875682013, + "grad_norm": 1.4211188181490109, + "learning_rate": 1.6740109147682148e-06, + "loss": 0.9149, + "step": 6271 + }, + { + "epoch": 0.5656310592054832, + "grad_norm": 1.5541616345989726, + "learning_rate": 1.6734345307216418e-06, + "loss": 0.9679, + "step": 6272 + }, + { + "epoch": 0.5657212427289534, + "grad_norm": 1.8138759257522101, + "learning_rate": 1.6728581745394346e-06, + "loss": 0.9249, + "step": 6273 + }, + { + "epoch": 0.5658114262524236, + "grad_norm": 5.104885170118683, + "learning_rate": 1.672281846270772e-06, + "loss": 1.0278, + "step": 6274 + }, + { + "epoch": 0.565901609775894, + "grad_norm": 1.5114242700602625, + "learning_rate": 1.6717055459648295e-06, + "loss": 0.9422, + "step": 6275 + }, + { + "epoch": 0.5659917932993642, + "grad_norm": 1.6095235070902503, + "learning_rate": 1.6711292736707793e-06, + "loss": 0.9024, + "step": 6276 + }, + { + "epoch": 0.5660819768228345, + "grad_norm": 1.6592183414512252, + "learning_rate": 1.6705530294377938e-06, + "loss": 1.0182, + "step": 6277 + }, + { + "epoch": 0.5661721603463047, + "grad_norm": 1.6322977815313489, + "learning_rate": 1.6699768133150395e-06, + "loss": 0.9456, + "step": 6278 + }, + { + "epoch": 0.566262343869775, + "grad_norm": 1.6080797038276133, + "learning_rate": 1.6694006253516837e-06, + "loss": 0.9393, + "step": 6279 + }, + { + "epoch": 0.5663525273932453, + "grad_norm": 1.6475977634487915, + "learning_rate": 1.6688244655968896e-06, + "loss": 0.9488, + "step": 6280 + }, + { + "epoch": 0.5664427109167155, + "grad_norm": 1.7272104971914328, + "learning_rate": 1.6682483340998175e-06, + "loss": 0.8717, + "step": 6281 + }, + { + "epoch": 0.5665328944401857, + "grad_norm": 1.6199557434912724, + "learning_rate": 1.6676722309096276e-06, + "loss": 0.9115, + "step": 6282 + }, + { + "epoch": 0.5666230779636561, + "grad_norm": 0.7718086430370386, + "learning_rate": 1.6670961560754744e-06, + "loss": 0.8492, + "step": 6283 + }, + { + "epoch": 0.5667132614871263, + "grad_norm": 1.5432430747782258, + "learning_rate": 1.6665201096465138e-06, + "loss": 0.9468, + "step": 6284 + }, + { + "epoch": 0.5668034450105965, + "grad_norm": 1.651209087687997, + "learning_rate": 1.6659440916718961e-06, + "loss": 0.9933, + "step": 6285 + }, + { + "epoch": 0.5668936285340668, + "grad_norm": 1.8243710701388158, + "learning_rate": 1.6653681022007696e-06, + "loss": 0.9612, + "step": 6286 + }, + { + "epoch": 0.5669838120575371, + "grad_norm": 1.7230495427780583, + "learning_rate": 1.6647921412822825e-06, + "loss": 0.8825, + "step": 6287 + }, + { + "epoch": 0.5670739955810074, + "grad_norm": 1.5207073298821665, + "learning_rate": 1.6642162089655782e-06, + "loss": 0.9486, + "step": 6288 + }, + { + "epoch": 0.5671641791044776, + "grad_norm": 1.5416736430792144, + "learning_rate": 1.663640305299798e-06, + "loss": 0.9137, + "step": 6289 + }, + { + "epoch": 0.5672543626279479, + "grad_norm": 0.761754939857944, + "learning_rate": 1.6630644303340824e-06, + "loss": 0.8117, + "step": 6290 + }, + { + "epoch": 0.5673445461514182, + "grad_norm": 1.6014449657949064, + "learning_rate": 1.662488584117567e-06, + "loss": 0.9576, + "step": 6291 + }, + { + "epoch": 0.5674347296748884, + "grad_norm": 1.2971612085937207, + "learning_rate": 1.6619127666993867e-06, + "loss": 1.0076, + "step": 6292 + }, + { + "epoch": 0.5675249131983586, + "grad_norm": 1.6355884081165961, + "learning_rate": 1.6613369781286727e-06, + "loss": 0.9523, + "step": 6293 + }, + { + "epoch": 0.567615096721829, + "grad_norm": 1.858108666300397, + "learning_rate": 1.6607612184545562e-06, + "loss": 0.8888, + "step": 6294 + }, + { + "epoch": 0.5677052802452992, + "grad_norm": 1.7542123506204532, + "learning_rate": 1.6601854877261617e-06, + "loss": 1.0254, + "step": 6295 + }, + { + "epoch": 0.5677954637687694, + "grad_norm": 1.8952897588587319, + "learning_rate": 1.6596097859926163e-06, + "loss": 0.9507, + "step": 6296 + }, + { + "epoch": 0.5678856472922397, + "grad_norm": 1.6820627141367055, + "learning_rate": 1.6590341133030407e-06, + "loss": 1.0407, + "step": 6297 + }, + { + "epoch": 0.56797583081571, + "grad_norm": 1.8349736707082156, + "learning_rate": 1.658458469706554e-06, + "loss": 1.024, + "step": 6298 + }, + { + "epoch": 0.5680660143391802, + "grad_norm": 1.4475195452172696, + "learning_rate": 1.6578828552522746e-06, + "loss": 0.907, + "step": 6299 + }, + { + "epoch": 0.5681561978626505, + "grad_norm": 1.6375738200081096, + "learning_rate": 1.6573072699893156e-06, + "loss": 0.9972, + "step": 6300 + }, + { + "epoch": 0.5682463813861207, + "grad_norm": 1.708777331208302, + "learning_rate": 1.6567317139667906e-06, + "loss": 0.9209, + "step": 6301 + }, + { + "epoch": 0.5683365649095911, + "grad_norm": 1.9256358372260454, + "learning_rate": 1.6561561872338087e-06, + "loss": 0.9655, + "step": 6302 + }, + { + "epoch": 0.5684267484330613, + "grad_norm": 1.6182545260809995, + "learning_rate": 1.6555806898394764e-06, + "loss": 0.861, + "step": 6303 + }, + { + "epoch": 0.5685169319565315, + "grad_norm": 1.5548131690680387, + "learning_rate": 1.6550052218328987e-06, + "loss": 0.9289, + "step": 6304 + }, + { + "epoch": 0.5686071154800018, + "grad_norm": 1.471540438133666, + "learning_rate": 1.6544297832631777e-06, + "loss": 0.9219, + "step": 6305 + }, + { + "epoch": 0.5686972990034721, + "grad_norm": 1.8087374998766634, + "learning_rate": 1.6538543741794135e-06, + "loss": 0.955, + "step": 6306 + }, + { + "epoch": 0.5687874825269423, + "grad_norm": 1.5002801156755137, + "learning_rate": 1.6532789946307028e-06, + "loss": 0.92, + "step": 6307 + }, + { + "epoch": 0.5688776660504126, + "grad_norm": 1.2309211504805075, + "learning_rate": 1.6527036446661393e-06, + "loss": 0.8141, + "step": 6308 + }, + { + "epoch": 0.5689678495738828, + "grad_norm": 1.4612202906137788, + "learning_rate": 1.6521283243348165e-06, + "loss": 0.9828, + "step": 6309 + }, + { + "epoch": 0.5690580330973531, + "grad_norm": 1.6895841514626186, + "learning_rate": 1.6515530336858227e-06, + "loss": 0.9436, + "step": 6310 + }, + { + "epoch": 0.5691482166208234, + "grad_norm": 1.4371566362378596, + "learning_rate": 1.6509777727682457e-06, + "loss": 0.9979, + "step": 6311 + }, + { + "epoch": 0.5692384001442936, + "grad_norm": 1.4379446751665106, + "learning_rate": 1.65040254163117e-06, + "loss": 0.9892, + "step": 6312 + }, + { + "epoch": 0.569328583667764, + "grad_norm": 2.014580744439572, + "learning_rate": 1.649827340323676e-06, + "loss": 0.8569, + "step": 6313 + }, + { + "epoch": 0.5694187671912342, + "grad_norm": 1.8134673595451731, + "learning_rate": 1.6492521688948454e-06, + "loss": 0.9164, + "step": 6314 + }, + { + "epoch": 0.5695089507147044, + "grad_norm": 1.3851412544794266, + "learning_rate": 1.6486770273937526e-06, + "loss": 0.9704, + "step": 6315 + }, + { + "epoch": 0.5695991342381747, + "grad_norm": 1.352930291947926, + "learning_rate": 1.6481019158694738e-06, + "loss": 0.9262, + "step": 6316 + }, + { + "epoch": 0.569689317761645, + "grad_norm": 0.811764604376602, + "learning_rate": 1.6475268343710792e-06, + "loss": 0.8536, + "step": 6317 + }, + { + "epoch": 0.5697795012851152, + "grad_norm": 1.6485505991821923, + "learning_rate": 1.6469517829476396e-06, + "loss": 0.8795, + "step": 6318 + }, + { + "epoch": 0.5698696848085855, + "grad_norm": 2.557950237389631, + "learning_rate": 1.64637676164822e-06, + "loss": 0.9297, + "step": 6319 + }, + { + "epoch": 0.5699598683320557, + "grad_norm": 1.4104919456703624, + "learning_rate": 1.6458017705218848e-06, + "loss": 0.9837, + "step": 6320 + }, + { + "epoch": 0.570050051855526, + "grad_norm": 1.6791603858142443, + "learning_rate": 1.645226809617696e-06, + "loss": 0.9741, + "step": 6321 + }, + { + "epoch": 0.5701402353789963, + "grad_norm": 2.186958572958163, + "learning_rate": 1.6446518789847112e-06, + "loss": 0.9851, + "step": 6322 + }, + { + "epoch": 0.5702304189024665, + "grad_norm": 1.6472366804475134, + "learning_rate": 1.6440769786719883e-06, + "loss": 0.9732, + "step": 6323 + }, + { + "epoch": 0.5703206024259367, + "grad_norm": 1.267902256414381, + "learning_rate": 1.6435021087285803e-06, + "loss": 0.9903, + "step": 6324 + }, + { + "epoch": 0.5704107859494071, + "grad_norm": 1.797223198366498, + "learning_rate": 1.642927269203537e-06, + "loss": 0.9866, + "step": 6325 + }, + { + "epoch": 0.5705009694728773, + "grad_norm": 1.4709912508011296, + "learning_rate": 1.642352460145909e-06, + "loss": 1.0005, + "step": 6326 + }, + { + "epoch": 0.5705911529963476, + "grad_norm": 1.4459086245557065, + "learning_rate": 1.6417776816047402e-06, + "loss": 0.9554, + "step": 6327 + }, + { + "epoch": 0.5706813365198178, + "grad_norm": 1.4028126987143563, + "learning_rate": 1.6412029336290755e-06, + "loss": 0.9359, + "step": 6328 + }, + { + "epoch": 0.5707715200432881, + "grad_norm": 1.8525938871532999, + "learning_rate": 1.6406282162679551e-06, + "loss": 0.9502, + "step": 6329 + }, + { + "epoch": 0.5708617035667584, + "grad_norm": 1.410887594597216, + "learning_rate": 1.6400535295704162e-06, + "loss": 0.9218, + "step": 6330 + }, + { + "epoch": 0.5709518870902286, + "grad_norm": 1.3870428861658728, + "learning_rate": 1.6394788735854955e-06, + "loss": 0.8894, + "step": 6331 + }, + { + "epoch": 0.5710420706136988, + "grad_norm": 1.4890693720431833, + "learning_rate": 1.6389042483622246e-06, + "loss": 1.0218, + "step": 6332 + }, + { + "epoch": 0.5711322541371692, + "grad_norm": 1.6200031334646556, + "learning_rate": 1.638329653949635e-06, + "loss": 0.9466, + "step": 6333 + }, + { + "epoch": 0.5712224376606394, + "grad_norm": 1.4627616224464508, + "learning_rate": 1.637755090396753e-06, + "loss": 0.914, + "step": 6334 + }, + { + "epoch": 0.5713126211841096, + "grad_norm": 0.6849586375828515, + "learning_rate": 1.6371805577526039e-06, + "loss": 0.8116, + "step": 6335 + }, + { + "epoch": 0.5714028047075799, + "grad_norm": 1.5763144351818332, + "learning_rate": 1.636606056066211e-06, + "loss": 0.9021, + "step": 6336 + }, + { + "epoch": 0.5714929882310502, + "grad_norm": 1.8300469883373773, + "learning_rate": 1.636031585386592e-06, + "loss": 0.8825, + "step": 6337 + }, + { + "epoch": 0.5715831717545204, + "grad_norm": 1.351296768084904, + "learning_rate": 1.635457145762766e-06, + "loss": 1.0108, + "step": 6338 + }, + { + "epoch": 0.5716733552779907, + "grad_norm": 1.8412933146657906, + "learning_rate": 1.6348827372437456e-06, + "loss": 0.9766, + "step": 6339 + }, + { + "epoch": 0.571763538801461, + "grad_norm": 1.3911455873036664, + "learning_rate": 1.634308359878544e-06, + "loss": 0.9528, + "step": 6340 + }, + { + "epoch": 0.5718537223249313, + "grad_norm": 6.570316273873025, + "learning_rate": 1.6337340137161695e-06, + "loss": 1.0304, + "step": 6341 + }, + { + "epoch": 0.5719439058484015, + "grad_norm": 2.1108844019086965, + "learning_rate": 1.6331596988056277e-06, + "loss": 0.953, + "step": 6342 + }, + { + "epoch": 0.5720340893718717, + "grad_norm": 1.4511208608704491, + "learning_rate": 1.632585415195924e-06, + "loss": 0.9769, + "step": 6343 + }, + { + "epoch": 0.5721242728953421, + "grad_norm": 1.410530569002889, + "learning_rate": 1.6320111629360583e-06, + "loss": 0.981, + "step": 6344 + }, + { + "epoch": 0.5722144564188123, + "grad_norm": 2.0485727971116665, + "learning_rate": 1.631436942075029e-06, + "loss": 0.9282, + "step": 6345 + }, + { + "epoch": 0.5723046399422825, + "grad_norm": 1.647627862185623, + "learning_rate": 1.630862752661833e-06, + "loss": 0.9259, + "step": 6346 + }, + { + "epoch": 0.5723948234657528, + "grad_norm": 1.5938226271892075, + "learning_rate": 1.6302885947454612e-06, + "loss": 0.8787, + "step": 6347 + }, + { + "epoch": 0.5724850069892231, + "grad_norm": 1.65691218544161, + "learning_rate": 1.6297144683749057e-06, + "loss": 0.9689, + "step": 6348 + }, + { + "epoch": 0.5725751905126933, + "grad_norm": 1.6772409763841147, + "learning_rate": 1.629140373599153e-06, + "loss": 0.9561, + "step": 6349 + }, + { + "epoch": 0.5726653740361636, + "grad_norm": 1.9256763236224268, + "learning_rate": 1.628566310467189e-06, + "loss": 0.9666, + "step": 6350 + }, + { + "epoch": 0.5727555575596338, + "grad_norm": 1.7115311659556052, + "learning_rate": 1.6279922790279957e-06, + "loss": 0.9174, + "step": 6351 + }, + { + "epoch": 0.5728457410831042, + "grad_norm": 1.6105934641499993, + "learning_rate": 1.6274182793305512e-06, + "loss": 0.9701, + "step": 6352 + }, + { + "epoch": 0.5729359246065744, + "grad_norm": 2.1539632862330644, + "learning_rate": 1.626844311423835e-06, + "loss": 0.9807, + "step": 6353 + }, + { + "epoch": 0.5730261081300446, + "grad_norm": 2.143158464590452, + "learning_rate": 1.6262703753568181e-06, + "loss": 1.0352, + "step": 6354 + }, + { + "epoch": 0.5731162916535149, + "grad_norm": 1.636724576734855, + "learning_rate": 1.6256964711784747e-06, + "loss": 1.0282, + "step": 6355 + }, + { + "epoch": 0.5732064751769852, + "grad_norm": 1.57786905460466, + "learning_rate": 1.6251225989377723e-06, + "loss": 0.9674, + "step": 6356 + }, + { + "epoch": 0.5732966587004554, + "grad_norm": 1.5539389490337832, + "learning_rate": 1.624548758683676e-06, + "loss": 0.9567, + "step": 6357 + }, + { + "epoch": 0.5733868422239257, + "grad_norm": 4.130268547064952, + "learning_rate": 1.6239749504651505e-06, + "loss": 1.0468, + "step": 6358 + }, + { + "epoch": 0.5734770257473959, + "grad_norm": 1.466640499632945, + "learning_rate": 1.6234011743311552e-06, + "loss": 0.9294, + "step": 6359 + }, + { + "epoch": 0.5735672092708662, + "grad_norm": 1.4957875867275254, + "learning_rate": 1.6228274303306483e-06, + "loss": 0.8864, + "step": 6360 + }, + { + "epoch": 0.5736573927943365, + "grad_norm": 1.3739169797344957, + "learning_rate": 1.6222537185125847e-06, + "loss": 1.0374, + "step": 6361 + }, + { + "epoch": 0.5737475763178067, + "grad_norm": 1.43958090392302, + "learning_rate": 1.6216800389259172e-06, + "loss": 0.973, + "step": 6362 + }, + { + "epoch": 0.573837759841277, + "grad_norm": 1.2741042187999259, + "learning_rate": 1.6211063916195949e-06, + "loss": 0.9594, + "step": 6363 + }, + { + "epoch": 0.5739279433647473, + "grad_norm": 1.758727043452096, + "learning_rate": 1.6205327766425633e-06, + "loss": 0.9605, + "step": 6364 + }, + { + "epoch": 0.5740181268882175, + "grad_norm": 1.2916218175587664, + "learning_rate": 1.6199591940437689e-06, + "loss": 0.9673, + "step": 6365 + }, + { + "epoch": 0.5741083104116878, + "grad_norm": 1.4420208538150634, + "learning_rate": 1.6193856438721505e-06, + "loss": 0.9147, + "step": 6366 + }, + { + "epoch": 0.5741984939351581, + "grad_norm": 1.6050822960025812, + "learning_rate": 1.6188121261766483e-06, + "loss": 0.9629, + "step": 6367 + }, + { + "epoch": 0.5742886774586283, + "grad_norm": 1.3168655635602462, + "learning_rate": 1.6182386410061976e-06, + "loss": 0.8882, + "step": 6368 + }, + { + "epoch": 0.5743788609820986, + "grad_norm": 1.5833535025801062, + "learning_rate": 1.61766518840973e-06, + "loss": 0.867, + "step": 6369 + }, + { + "epoch": 0.5744690445055688, + "grad_norm": 1.4331958020933713, + "learning_rate": 1.6170917684361779e-06, + "loss": 0.8271, + "step": 6370 + }, + { + "epoch": 0.5745592280290391, + "grad_norm": 1.432948328400307, + "learning_rate": 1.6165183811344662e-06, + "loss": 1.0327, + "step": 6371 + }, + { + "epoch": 0.5746494115525094, + "grad_norm": 1.5366340493624877, + "learning_rate": 1.6159450265535218e-06, + "loss": 0.9488, + "step": 6372 + }, + { + "epoch": 0.5747395950759796, + "grad_norm": 1.3905937534207888, + "learning_rate": 1.6153717047422652e-06, + "loss": 1.0391, + "step": 6373 + }, + { + "epoch": 0.5748297785994498, + "grad_norm": 1.5928758205904745, + "learning_rate": 1.6147984157496155e-06, + "loss": 0.8419, + "step": 6374 + }, + { + "epoch": 0.5749199621229202, + "grad_norm": 1.429711346870365, + "learning_rate": 1.6142251596244886e-06, + "loss": 0.8814, + "step": 6375 + }, + { + "epoch": 0.5750101456463904, + "grad_norm": 1.4604857863296454, + "learning_rate": 1.6136519364157983e-06, + "loss": 1.0122, + "step": 6376 + }, + { + "epoch": 0.5751003291698606, + "grad_norm": 2.0293087412805733, + "learning_rate": 1.6130787461724555e-06, + "loss": 0.8957, + "step": 6377 + }, + { + "epoch": 0.5751905126933309, + "grad_norm": 1.635815500922357, + "learning_rate": 1.6125055889433679e-06, + "loss": 1.0298, + "step": 6378 + }, + { + "epoch": 0.5752806962168012, + "grad_norm": 1.4273159180982893, + "learning_rate": 1.6119324647774386e-06, + "loss": 0.9812, + "step": 6379 + }, + { + "epoch": 0.5753708797402715, + "grad_norm": 1.72775696738175, + "learning_rate": 1.6113593737235724e-06, + "loss": 0.965, + "step": 6380 + }, + { + "epoch": 0.5754610632637417, + "grad_norm": 1.494801493578076, + "learning_rate": 1.6107863158306665e-06, + "loss": 0.9049, + "step": 6381 + }, + { + "epoch": 0.5755512467872119, + "grad_norm": 2.185715628782941, + "learning_rate": 1.610213291147619e-06, + "loss": 0.8549, + "step": 6382 + }, + { + "epoch": 0.5756414303106823, + "grad_norm": 1.4332810562418108, + "learning_rate": 1.609640299723322e-06, + "loss": 0.9805, + "step": 6383 + }, + { + "epoch": 0.5757316138341525, + "grad_norm": 1.3615467656705855, + "learning_rate": 1.609067341606668e-06, + "loss": 0.9414, + "step": 6384 + }, + { + "epoch": 0.5758217973576227, + "grad_norm": 1.7583221374926659, + "learning_rate": 1.6084944168465438e-06, + "loss": 0.9463, + "step": 6385 + }, + { + "epoch": 0.5759119808810931, + "grad_norm": 1.4099579986106214, + "learning_rate": 1.6079215254918339e-06, + "loss": 0.979, + "step": 6386 + }, + { + "epoch": 0.5760021644045633, + "grad_norm": 1.8986000022882559, + "learning_rate": 1.6073486675914222e-06, + "loss": 0.8626, + "step": 6387 + }, + { + "epoch": 0.5760923479280335, + "grad_norm": 1.7519241381642552, + "learning_rate": 1.606775843194187e-06, + "loss": 0.9867, + "step": 6388 + }, + { + "epoch": 0.5761825314515038, + "grad_norm": 1.8971209572420153, + "learning_rate": 1.6062030523490053e-06, + "loss": 0.9452, + "step": 6389 + }, + { + "epoch": 0.5762727149749741, + "grad_norm": 1.5551221230440961, + "learning_rate": 1.60563029510475e-06, + "loss": 0.8748, + "step": 6390 + }, + { + "epoch": 0.5763628984984444, + "grad_norm": 0.6974171492547998, + "learning_rate": 1.6050575715102927e-06, + "loss": 0.7708, + "step": 6391 + }, + { + "epoch": 0.5764530820219146, + "grad_norm": 1.7122149292847413, + "learning_rate": 1.6044848816145014e-06, + "loss": 1.0153, + "step": 6392 + }, + { + "epoch": 0.5765432655453848, + "grad_norm": 1.763107057282008, + "learning_rate": 1.60391222546624e-06, + "loss": 0.9813, + "step": 6393 + }, + { + "epoch": 0.5766334490688552, + "grad_norm": 1.4153779376352467, + "learning_rate": 1.6033396031143725e-06, + "loss": 0.9473, + "step": 6394 + }, + { + "epoch": 0.5767236325923254, + "grad_norm": 1.634538313399408, + "learning_rate": 1.602767014607757e-06, + "loss": 0.9614, + "step": 6395 + }, + { + "epoch": 0.5768138161157956, + "grad_norm": 1.5070122685480438, + "learning_rate": 1.6021944599952493e-06, + "loss": 0.9702, + "step": 6396 + }, + { + "epoch": 0.5769039996392659, + "grad_norm": 1.811385963344111, + "learning_rate": 1.6016219393257048e-06, + "loss": 0.9711, + "step": 6397 + }, + { + "epoch": 0.5769941831627362, + "grad_norm": 1.480661505478187, + "learning_rate": 1.6010494526479722e-06, + "loss": 0.9072, + "step": 6398 + }, + { + "epoch": 0.5770843666862064, + "grad_norm": 1.8954942281197928, + "learning_rate": 1.6004770000109006e-06, + "loss": 0.895, + "step": 6399 + }, + { + "epoch": 0.5771745502096767, + "grad_norm": 1.5291040915315894, + "learning_rate": 1.5999045814633348e-06, + "loss": 0.9238, + "step": 6400 + }, + { + "epoch": 0.5772647337331469, + "grad_norm": 2.1050963912967005, + "learning_rate": 1.5993321970541151e-06, + "loss": 0.9699, + "step": 6401 + }, + { + "epoch": 0.5773549172566173, + "grad_norm": 1.710235939591504, + "learning_rate": 1.5987598468320825e-06, + "loss": 0.96, + "step": 6402 + }, + { + "epoch": 0.5774451007800875, + "grad_norm": 1.42388991233606, + "learning_rate": 1.5981875308460717e-06, + "loss": 0.9407, + "step": 6403 + }, + { + "epoch": 0.5775352843035577, + "grad_norm": 1.8394021964258018, + "learning_rate": 1.5976152491449169e-06, + "loss": 0.9396, + "step": 6404 + }, + { + "epoch": 0.577625467827028, + "grad_norm": 1.61337494289451, + "learning_rate": 1.5970430017774468e-06, + "loss": 1.0184, + "step": 6405 + }, + { + "epoch": 0.5777156513504983, + "grad_norm": 1.5621712147977416, + "learning_rate": 1.5964707887924904e-06, + "loss": 0.9695, + "step": 6406 + }, + { + "epoch": 0.5778058348739685, + "grad_norm": 1.2671252648137954, + "learning_rate": 1.5958986102388714e-06, + "loss": 0.843, + "step": 6407 + }, + { + "epoch": 0.5778960183974388, + "grad_norm": 1.5050201650390178, + "learning_rate": 1.5953264661654104e-06, + "loss": 0.9097, + "step": 6408 + }, + { + "epoch": 0.5779862019209091, + "grad_norm": 1.4898593175641828, + "learning_rate": 1.5947543566209276e-06, + "loss": 0.9894, + "step": 6409 + }, + { + "epoch": 0.5780763854443793, + "grad_norm": 1.4586844793542015, + "learning_rate": 1.5941822816542367e-06, + "loss": 0.9363, + "step": 6410 + }, + { + "epoch": 0.5781665689678496, + "grad_norm": 1.4225076797894831, + "learning_rate": 1.5936102413141519e-06, + "loss": 0.9138, + "step": 6411 + }, + { + "epoch": 0.5782567524913198, + "grad_norm": 1.8381950620848684, + "learning_rate": 1.5930382356494823e-06, + "loss": 0.872, + "step": 6412 + }, + { + "epoch": 0.5783469360147901, + "grad_norm": 1.4541953462033337, + "learning_rate": 1.5924662647090335e-06, + "loss": 1.0039, + "step": 6413 + }, + { + "epoch": 0.5784371195382604, + "grad_norm": 1.5114498244225067, + "learning_rate": 1.5918943285416108e-06, + "loss": 0.9746, + "step": 6414 + }, + { + "epoch": 0.5785273030617306, + "grad_norm": 4.0132218231559875, + "learning_rate": 1.5913224271960139e-06, + "loss": 0.8924, + "step": 6415 + }, + { + "epoch": 0.5786174865852008, + "grad_norm": 2.067177756496312, + "learning_rate": 1.590750560721041e-06, + "loss": 0.9693, + "step": 6416 + }, + { + "epoch": 0.5787076701086712, + "grad_norm": 1.2938035649807025, + "learning_rate": 1.5901787291654874e-06, + "loss": 0.9224, + "step": 6417 + }, + { + "epoch": 0.5787978536321414, + "grad_norm": 1.47653666478721, + "learning_rate": 1.5896069325781435e-06, + "loss": 0.832, + "step": 6418 + }, + { + "epoch": 0.5788880371556117, + "grad_norm": 1.5158143859204782, + "learning_rate": 1.5890351710077998e-06, + "loss": 0.9925, + "step": 6419 + }, + { + "epoch": 0.5789782206790819, + "grad_norm": 1.3629219434249797, + "learning_rate": 1.5884634445032406e-06, + "loss": 1.0151, + "step": 6420 + }, + { + "epoch": 0.5790684042025522, + "grad_norm": 0.6465752759568199, + "learning_rate": 1.5878917531132501e-06, + "loss": 0.799, + "step": 6421 + }, + { + "epoch": 0.5791585877260225, + "grad_norm": 1.7948041177835448, + "learning_rate": 1.5873200968866077e-06, + "loss": 0.8302, + "step": 6422 + }, + { + "epoch": 0.5792487712494927, + "grad_norm": 3.020460767900693, + "learning_rate": 1.586748475872089e-06, + "loss": 0.9287, + "step": 6423 + }, + { + "epoch": 0.5793389547729629, + "grad_norm": 1.536953483150644, + "learning_rate": 1.58617689011847e-06, + "loss": 0.823, + "step": 6424 + }, + { + "epoch": 0.5794291382964333, + "grad_norm": 1.5182884645305348, + "learning_rate": 1.5856053396745198e-06, + "loss": 0.9931, + "step": 6425 + }, + { + "epoch": 0.5795193218199035, + "grad_norm": 1.545571596894401, + "learning_rate": 1.5850338245890078e-06, + "loss": 1.0192, + "step": 6426 + }, + { + "epoch": 0.5796095053433737, + "grad_norm": 1.584695564660231, + "learning_rate": 1.5844623449106974e-06, + "loss": 0.9391, + "step": 6427 + }, + { + "epoch": 0.579699688866844, + "grad_norm": 1.5189572699126315, + "learning_rate": 1.583890900688351e-06, + "loss": 0.9965, + "step": 6428 + }, + { + "epoch": 0.5797898723903143, + "grad_norm": 1.7846020395440205, + "learning_rate": 1.583319491970728e-06, + "loss": 1.0332, + "step": 6429 + }, + { + "epoch": 0.5798800559137846, + "grad_norm": 1.4260120779812708, + "learning_rate": 1.5827481188065828e-06, + "loss": 0.9686, + "step": 6430 + }, + { + "epoch": 0.5799702394372548, + "grad_norm": 1.3946542552051708, + "learning_rate": 1.5821767812446689e-06, + "loss": 0.9271, + "step": 6431 + }, + { + "epoch": 0.5800604229607251, + "grad_norm": 1.559009697272854, + "learning_rate": 1.581605479333736e-06, + "loss": 1.0705, + "step": 6432 + }, + { + "epoch": 0.5801506064841954, + "grad_norm": 1.7042367788627097, + "learning_rate": 1.5810342131225308e-06, + "loss": 0.961, + "step": 6433 + }, + { + "epoch": 0.5802407900076656, + "grad_norm": 1.5480631879187174, + "learning_rate": 1.580462982659797e-06, + "loss": 0.9968, + "step": 6434 + }, + { + "epoch": 0.5803309735311358, + "grad_norm": 1.71587894971305, + "learning_rate": 1.5798917879942736e-06, + "loss": 1.032, + "step": 6435 + }, + { + "epoch": 0.5804211570546062, + "grad_norm": 2.043677591796868, + "learning_rate": 1.5793206291747006e-06, + "loss": 0.8987, + "step": 6436 + }, + { + "epoch": 0.5805113405780764, + "grad_norm": 1.4170928856201417, + "learning_rate": 1.57874950624981e-06, + "loss": 0.8909, + "step": 6437 + }, + { + "epoch": 0.5806015241015466, + "grad_norm": 4.826567265837624, + "learning_rate": 1.5781784192683351e-06, + "loss": 0.9529, + "step": 6438 + }, + { + "epoch": 0.5806917076250169, + "grad_norm": 1.1781338698966446, + "learning_rate": 1.5776073682790033e-06, + "loss": 0.9166, + "step": 6439 + }, + { + "epoch": 0.5807818911484872, + "grad_norm": 1.64220200086701, + "learning_rate": 1.5770363533305393e-06, + "loss": 0.9755, + "step": 6440 + }, + { + "epoch": 0.5808720746719575, + "grad_norm": 1.6165173012597844, + "learning_rate": 1.5764653744716665e-06, + "loss": 0.9552, + "step": 6441 + }, + { + "epoch": 0.5809622581954277, + "grad_norm": 1.6476732263101999, + "learning_rate": 1.575894431751103e-06, + "loss": 0.9818, + "step": 6442 + }, + { + "epoch": 0.5810524417188979, + "grad_norm": 1.6087107352301642, + "learning_rate": 1.575323525217565e-06, + "loss": 0.9814, + "step": 6443 + }, + { + "epoch": 0.5811426252423683, + "grad_norm": 2.3884284206538737, + "learning_rate": 1.574752654919766e-06, + "loss": 1.0347, + "step": 6444 + }, + { + "epoch": 0.5812328087658385, + "grad_norm": 1.5015661487673946, + "learning_rate": 1.5741818209064146e-06, + "loss": 0.9275, + "step": 6445 + }, + { + "epoch": 0.5813229922893087, + "grad_norm": 1.5310901636275718, + "learning_rate": 1.5736110232262183e-06, + "loss": 0.964, + "step": 6446 + }, + { + "epoch": 0.581413175812779, + "grad_norm": 1.4743566435430808, + "learning_rate": 1.5730402619278804e-06, + "loss": 1.002, + "step": 6447 + }, + { + "epoch": 0.5815033593362493, + "grad_norm": 0.7339205857798133, + "learning_rate": 1.5724695370601024e-06, + "loss": 0.7571, + "step": 6448 + }, + { + "epoch": 0.5815935428597195, + "grad_norm": 1.450564271378059, + "learning_rate": 1.5718988486715798e-06, + "loss": 0.8772, + "step": 6449 + }, + { + "epoch": 0.5816837263831898, + "grad_norm": 1.4785463438282136, + "learning_rate": 1.5713281968110087e-06, + "loss": 0.936, + "step": 6450 + }, + { + "epoch": 0.58177390990666, + "grad_norm": 1.5806529881966025, + "learning_rate": 1.5707575815270796e-06, + "loss": 0.9359, + "step": 6451 + }, + { + "epoch": 0.5818640934301303, + "grad_norm": 1.6543622324890068, + "learning_rate": 1.57018700286848e-06, + "loss": 0.9534, + "step": 6452 + }, + { + "epoch": 0.5819542769536006, + "grad_norm": 1.644606563628593, + "learning_rate": 1.5696164608838956e-06, + "loss": 0.9676, + "step": 6453 + }, + { + "epoch": 0.5820444604770708, + "grad_norm": 1.5241197672212623, + "learning_rate": 1.5690459556220073e-06, + "loss": 0.9666, + "step": 6454 + }, + { + "epoch": 0.582134644000541, + "grad_norm": 1.8849220329572676, + "learning_rate": 1.5684754871314949e-06, + "loss": 1.0169, + "step": 6455 + }, + { + "epoch": 0.5822248275240114, + "grad_norm": 2.1834809576120966, + "learning_rate": 1.5679050554610335e-06, + "loss": 0.9521, + "step": 6456 + }, + { + "epoch": 0.5823150110474816, + "grad_norm": 1.3468976083083843, + "learning_rate": 1.567334660659295e-06, + "loss": 0.9574, + "step": 6457 + }, + { + "epoch": 0.5824051945709519, + "grad_norm": 1.4318110636469503, + "learning_rate": 1.5667643027749488e-06, + "loss": 0.9721, + "step": 6458 + }, + { + "epoch": 0.5824953780944222, + "grad_norm": 1.7456239755356375, + "learning_rate": 1.5661939818566614e-06, + "loss": 1.0245, + "step": 6459 + }, + { + "epoch": 0.5825855616178924, + "grad_norm": 1.5216741284413442, + "learning_rate": 1.5656236979530956e-06, + "loss": 0.8224, + "step": 6460 + }, + { + "epoch": 0.5826757451413627, + "grad_norm": 1.5920923625760224, + "learning_rate": 1.5650534511129106e-06, + "loss": 0.8984, + "step": 6461 + }, + { + "epoch": 0.5827659286648329, + "grad_norm": 1.8515306944871786, + "learning_rate": 1.5644832413847635e-06, + "loss": 0.9368, + "step": 6462 + }, + { + "epoch": 0.5828561121883032, + "grad_norm": 1.2957318794710109, + "learning_rate": 1.5639130688173082e-06, + "loss": 0.9975, + "step": 6463 + }, + { + "epoch": 0.5829462957117735, + "grad_norm": 1.4602796736100558, + "learning_rate": 1.5633429334591932e-06, + "loss": 0.9327, + "step": 6464 + }, + { + "epoch": 0.5830364792352437, + "grad_norm": 1.6431607223580176, + "learning_rate": 1.562772835359068e-06, + "loss": 0.9918, + "step": 6465 + }, + { + "epoch": 0.5831266627587139, + "grad_norm": 1.7666882039802632, + "learning_rate": 1.5622027745655753e-06, + "loss": 0.9246, + "step": 6466 + }, + { + "epoch": 0.5832168462821843, + "grad_norm": 1.4280062102281612, + "learning_rate": 1.561632751127355e-06, + "loss": 0.9119, + "step": 6467 + }, + { + "epoch": 0.5833070298056545, + "grad_norm": 1.6285693508627312, + "learning_rate": 1.561062765093046e-06, + "loss": 1.0001, + "step": 6468 + }, + { + "epoch": 0.5833972133291248, + "grad_norm": 1.9217344403233128, + "learning_rate": 1.5604928165112817e-06, + "loss": 0.9549, + "step": 6469 + }, + { + "epoch": 0.583487396852595, + "grad_norm": 1.498155333809075, + "learning_rate": 1.5599229054306945e-06, + "loss": 0.9493, + "step": 6470 + }, + { + "epoch": 0.5835775803760653, + "grad_norm": 1.6329630367115961, + "learning_rate": 1.5593530318999111e-06, + "loss": 0.9496, + "step": 6471 + }, + { + "epoch": 0.5836677638995356, + "grad_norm": 1.6609153537942511, + "learning_rate": 1.5587831959675572e-06, + "loss": 0.9862, + "step": 6472 + }, + { + "epoch": 0.5837579474230058, + "grad_norm": 1.4135175887971136, + "learning_rate": 1.5582133976822534e-06, + "loss": 0.9377, + "step": 6473 + }, + { + "epoch": 0.583848130946476, + "grad_norm": 1.4768518235062549, + "learning_rate": 1.5576436370926185e-06, + "loss": 0.9092, + "step": 6474 + }, + { + "epoch": 0.5839383144699464, + "grad_norm": 1.3074390839924954, + "learning_rate": 1.5570739142472679e-06, + "loss": 0.9343, + "step": 6475 + }, + { + "epoch": 0.5840284979934166, + "grad_norm": 1.3339281195958785, + "learning_rate": 1.5565042291948127e-06, + "loss": 1.0172, + "step": 6476 + }, + { + "epoch": 0.5841186815168868, + "grad_norm": 1.4380986998671865, + "learning_rate": 1.5559345819838624e-06, + "loss": 0.9891, + "step": 6477 + }, + { + "epoch": 0.5842088650403571, + "grad_norm": 1.3084601035123389, + "learning_rate": 1.5553649726630226e-06, + "loss": 0.9375, + "step": 6478 + }, + { + "epoch": 0.5842990485638274, + "grad_norm": 1.4731727726359487, + "learning_rate": 1.5547954012808942e-06, + "loss": 0.8851, + "step": 6479 + }, + { + "epoch": 0.5843892320872976, + "grad_norm": 1.8543163696318632, + "learning_rate": 1.5542258678860776e-06, + "loss": 0.9333, + "step": 6480 + }, + { + "epoch": 0.5844794156107679, + "grad_norm": 1.5527181756476125, + "learning_rate": 1.553656372527167e-06, + "loss": 1.0799, + "step": 6481 + }, + { + "epoch": 0.5845695991342382, + "grad_norm": 1.9841109573231657, + "learning_rate": 1.5530869152527568e-06, + "loss": 1.058, + "step": 6482 + }, + { + "epoch": 0.5846597826577085, + "grad_norm": 1.4397123975445347, + "learning_rate": 1.5525174961114353e-06, + "loss": 0.9725, + "step": 6483 + }, + { + "epoch": 0.5847499661811787, + "grad_norm": 1.5150493039710407, + "learning_rate": 1.5519481151517875e-06, + "loss": 0.9906, + "step": 6484 + }, + { + "epoch": 0.5848401497046489, + "grad_norm": 1.6319349933467833, + "learning_rate": 1.551378772422398e-06, + "loss": 0.9589, + "step": 6485 + }, + { + "epoch": 0.5849303332281193, + "grad_norm": 1.3931575816175006, + "learning_rate": 1.5508094679718447e-06, + "loss": 0.9686, + "step": 6486 + }, + { + "epoch": 0.5850205167515895, + "grad_norm": 1.5005562068958667, + "learning_rate": 1.5502402018487048e-06, + "loss": 0.8603, + "step": 6487 + }, + { + "epoch": 0.5851107002750597, + "grad_norm": 1.5178965427189672, + "learning_rate": 1.54967097410155e-06, + "loss": 0.9634, + "step": 6488 + }, + { + "epoch": 0.58520088379853, + "grad_norm": 1.3641378777335773, + "learning_rate": 1.5491017847789519e-06, + "loss": 0.8774, + "step": 6489 + }, + { + "epoch": 0.5852910673220003, + "grad_norm": 1.4844924077231403, + "learning_rate": 1.5485326339294755e-06, + "loss": 0.9027, + "step": 6490 + }, + { + "epoch": 0.5853812508454705, + "grad_norm": 1.354879509541144, + "learning_rate": 1.5479635216016832e-06, + "loss": 0.9244, + "step": 6491 + }, + { + "epoch": 0.5854714343689408, + "grad_norm": 1.5386684319873338, + "learning_rate": 1.547394447844137e-06, + "loss": 0.8688, + "step": 6492 + }, + { + "epoch": 0.585561617892411, + "grad_norm": 1.418164620650885, + "learning_rate": 1.546825412705391e-06, + "loss": 0.8699, + "step": 6493 + }, + { + "epoch": 0.5856518014158814, + "grad_norm": 1.5115848452383505, + "learning_rate": 1.5462564162340007e-06, + "loss": 0.9847, + "step": 6494 + }, + { + "epoch": 0.5857419849393516, + "grad_norm": 1.9656272221053865, + "learning_rate": 1.5456874584785144e-06, + "loss": 0.9615, + "step": 6495 + }, + { + "epoch": 0.5858321684628218, + "grad_norm": 1.8070635986071466, + "learning_rate": 1.5451185394874785e-06, + "loss": 0.9326, + "step": 6496 + }, + { + "epoch": 0.5859223519862921, + "grad_norm": 1.5049956896393857, + "learning_rate": 1.5445496593094381e-06, + "loss": 0.9661, + "step": 6497 + }, + { + "epoch": 0.5860125355097624, + "grad_norm": 1.6422615970986638, + "learning_rate": 1.5439808179929316e-06, + "loss": 0.9687, + "step": 6498 + }, + { + "epoch": 0.5861027190332326, + "grad_norm": 1.7749787745751386, + "learning_rate": 1.543412015586496e-06, + "loss": 0.9224, + "step": 6499 + }, + { + "epoch": 0.5861929025567029, + "grad_norm": 1.8160178805029752, + "learning_rate": 1.5428432521386655e-06, + "loss": 0.9211, + "step": 6500 + }, + { + "epoch": 0.5862830860801731, + "grad_norm": 1.5009075439310302, + "learning_rate": 1.5422745276979688e-06, + "loss": 0.9903, + "step": 6501 + }, + { + "epoch": 0.5863732696036434, + "grad_norm": 1.520375230677044, + "learning_rate": 1.5417058423129336e-06, + "loss": 0.9464, + "step": 6502 + }, + { + "epoch": 0.5864634531271137, + "grad_norm": 1.3755776752406323, + "learning_rate": 1.5411371960320822e-06, + "loss": 0.991, + "step": 6503 + }, + { + "epoch": 0.5865536366505839, + "grad_norm": 1.4584867759996973, + "learning_rate": 1.5405685889039363e-06, + "loss": 0.99, + "step": 6504 + }, + { + "epoch": 0.5866438201740543, + "grad_norm": 1.667139852427468, + "learning_rate": 1.5400000209770118e-06, + "loss": 0.9519, + "step": 6505 + }, + { + "epoch": 0.5867340036975245, + "grad_norm": 1.694215762578928, + "learning_rate": 1.5394314922998208e-06, + "loss": 0.9689, + "step": 6506 + }, + { + "epoch": 0.5868241872209947, + "grad_norm": 2.06991262352858, + "learning_rate": 1.5388630029208756e-06, + "loss": 1.0096, + "step": 6507 + }, + { + "epoch": 0.586914370744465, + "grad_norm": 1.5136431291373287, + "learning_rate": 1.5382945528886806e-06, + "loss": 0.9834, + "step": 6508 + }, + { + "epoch": 0.5870045542679353, + "grad_norm": 2.015925421985356, + "learning_rate": 1.5377261422517412e-06, + "loss": 0.9378, + "step": 6509 + }, + { + "epoch": 0.5870947377914055, + "grad_norm": 1.6360346928301364, + "learning_rate": 1.5371577710585553e-06, + "loss": 0.9147, + "step": 6510 + }, + { + "epoch": 0.5871849213148758, + "grad_norm": 0.7469888879621042, + "learning_rate": 1.536589439357621e-06, + "loss": 0.767, + "step": 6511 + }, + { + "epoch": 0.587275104838346, + "grad_norm": 1.4375988470175847, + "learning_rate": 1.5360211471974315e-06, + "loss": 0.959, + "step": 6512 + }, + { + "epoch": 0.5873652883618163, + "grad_norm": 1.419332305790171, + "learning_rate": 1.5354528946264753e-06, + "loss": 1.012, + "step": 6513 + }, + { + "epoch": 0.5874554718852866, + "grad_norm": 1.4029340856271553, + "learning_rate": 1.5348846816932399e-06, + "loss": 0.8886, + "step": 6514 + }, + { + "epoch": 0.5875456554087568, + "grad_norm": 0.6847240367910067, + "learning_rate": 1.5343165084462077e-06, + "loss": 0.7762, + "step": 6515 + }, + { + "epoch": 0.587635838932227, + "grad_norm": 1.6252714077170851, + "learning_rate": 1.5337483749338595e-06, + "loss": 0.9097, + "step": 6516 + }, + { + "epoch": 0.5877260224556974, + "grad_norm": 1.6118733541325674, + "learning_rate": 1.5331802812046708e-06, + "loss": 0.9301, + "step": 6517 + }, + { + "epoch": 0.5878162059791676, + "grad_norm": 1.5602818575569486, + "learning_rate": 1.5326122273071133e-06, + "loss": 0.8057, + "step": 6518 + }, + { + "epoch": 0.5879063895026378, + "grad_norm": 1.7314029006367917, + "learning_rate": 1.532044213289659e-06, + "loss": 0.9929, + "step": 6519 + }, + { + "epoch": 0.5879965730261081, + "grad_norm": 1.4851759506435986, + "learning_rate": 1.5314762392007718e-06, + "loss": 0.8436, + "step": 6520 + }, + { + "epoch": 0.5880867565495784, + "grad_norm": 1.370119928206342, + "learning_rate": 1.530908305088916e-06, + "loss": 1.0065, + "step": 6521 + }, + { + "epoch": 0.5881769400730487, + "grad_norm": 1.5383629927801636, + "learning_rate": 1.5303404110025501e-06, + "loss": 0.9534, + "step": 6522 + }, + { + "epoch": 0.5882671235965189, + "grad_norm": 1.4144963394554564, + "learning_rate": 1.5297725569901293e-06, + "loss": 0.9107, + "step": 6523 + }, + { + "epoch": 0.5883573071199891, + "grad_norm": 1.6068056948778588, + "learning_rate": 1.5292047431001077e-06, + "loss": 0.8518, + "step": 6524 + }, + { + "epoch": 0.5884474906434595, + "grad_norm": 0.733187750612983, + "learning_rate": 1.5286369693809321e-06, + "loss": 0.7895, + "step": 6525 + }, + { + "epoch": 0.5885376741669297, + "grad_norm": 1.54476437929335, + "learning_rate": 1.5280692358810506e-06, + "loss": 0.9699, + "step": 6526 + }, + { + "epoch": 0.5886278576903999, + "grad_norm": 1.5581306400495376, + "learning_rate": 1.527501542648904e-06, + "loss": 0.9065, + "step": 6527 + }, + { + "epoch": 0.5887180412138703, + "grad_norm": 1.6626600346227762, + "learning_rate": 1.5269338897329308e-06, + "loss": 0.8631, + "step": 6528 + }, + { + "epoch": 0.5888082247373405, + "grad_norm": 1.5310388536263462, + "learning_rate": 1.5263662771815662e-06, + "loss": 0.98, + "step": 6529 + }, + { + "epoch": 0.5888984082608107, + "grad_norm": 1.6808293919035595, + "learning_rate": 1.5257987050432429e-06, + "loss": 1.0293, + "step": 6530 + }, + { + "epoch": 0.588988591784281, + "grad_norm": 1.8165850807675137, + "learning_rate": 1.5252311733663887e-06, + "loss": 0.8645, + "step": 6531 + }, + { + "epoch": 0.5890787753077513, + "grad_norm": 1.6800699197887654, + "learning_rate": 1.5246636821994281e-06, + "loss": 0.8993, + "step": 6532 + }, + { + "epoch": 0.5891689588312216, + "grad_norm": 1.4145603041349717, + "learning_rate": 1.524096231590784e-06, + "loss": 1.0033, + "step": 6533 + }, + { + "epoch": 0.5892591423546918, + "grad_norm": 1.7159552306713053, + "learning_rate": 1.5235288215888736e-06, + "loss": 1.0041, + "step": 6534 + }, + { + "epoch": 0.589349325878162, + "grad_norm": 2.2629313273813345, + "learning_rate": 1.5229614522421102e-06, + "loss": 0.9589, + "step": 6535 + }, + { + "epoch": 0.5894395094016324, + "grad_norm": 1.6126171158243843, + "learning_rate": 1.5223941235989071e-06, + "loss": 0.7941, + "step": 6536 + }, + { + "epoch": 0.5895296929251026, + "grad_norm": 1.3983703682412638, + "learning_rate": 1.52182683570767e-06, + "loss": 0.9237, + "step": 6537 + }, + { + "epoch": 0.5896198764485728, + "grad_norm": 1.6012729289954117, + "learning_rate": 1.5212595886168046e-06, + "loss": 0.975, + "step": 6538 + }, + { + "epoch": 0.5897100599720431, + "grad_norm": 1.518954601556197, + "learning_rate": 1.520692382374711e-06, + "loss": 0.9749, + "step": 6539 + }, + { + "epoch": 0.5898002434955134, + "grad_norm": 1.8994728059682227, + "learning_rate": 1.5201252170297854e-06, + "loss": 0.9792, + "step": 6540 + }, + { + "epoch": 0.5898904270189836, + "grad_norm": 1.5860249161705327, + "learning_rate": 1.5195580926304232e-06, + "loss": 0.8408, + "step": 6541 + }, + { + "epoch": 0.5899806105424539, + "grad_norm": 1.668990240055938, + "learning_rate": 1.5189910092250131e-06, + "loss": 0.9759, + "step": 6542 + }, + { + "epoch": 0.5900707940659241, + "grad_norm": 1.7359652402843393, + "learning_rate": 1.5184239668619427e-06, + "loss": 0.9605, + "step": 6543 + }, + { + "epoch": 0.5901609775893945, + "grad_norm": 1.6641272590465424, + "learning_rate": 1.5178569655895946e-06, + "loss": 0.9526, + "step": 6544 + }, + { + "epoch": 0.5902511611128647, + "grad_norm": 1.3348085864853176, + "learning_rate": 1.5172900054563487e-06, + "loss": 0.9842, + "step": 6545 + }, + { + "epoch": 0.5903413446363349, + "grad_norm": 1.6631670051210183, + "learning_rate": 1.5167230865105814e-06, + "loss": 0.9735, + "step": 6546 + }, + { + "epoch": 0.5904315281598052, + "grad_norm": 1.7088838558607942, + "learning_rate": 1.5161562088006644e-06, + "loss": 0.8664, + "step": 6547 + }, + { + "epoch": 0.5905217116832755, + "grad_norm": 1.7513931723426406, + "learning_rate": 1.5155893723749685e-06, + "loss": 0.9835, + "step": 6548 + }, + { + "epoch": 0.5906118952067457, + "grad_norm": 1.7013553209518297, + "learning_rate": 1.5150225772818582e-06, + "loss": 0.9579, + "step": 6549 + }, + { + "epoch": 0.590702078730216, + "grad_norm": 1.4560754008262193, + "learning_rate": 1.5144558235696949e-06, + "loss": 1.0055, + "step": 6550 + }, + { + "epoch": 0.5907922622536863, + "grad_norm": 3.232309170051773, + "learning_rate": 1.5138891112868388e-06, + "loss": 0.8723, + "step": 6551 + }, + { + "epoch": 0.5908824457771565, + "grad_norm": 0.7803850344798394, + "learning_rate": 1.5133224404816433e-06, + "loss": 0.7803, + "step": 6552 + }, + { + "epoch": 0.5909726293006268, + "grad_norm": 1.8241098114094447, + "learning_rate": 1.5127558112024617e-06, + "loss": 0.8798, + "step": 6553 + }, + { + "epoch": 0.591062812824097, + "grad_norm": 0.653183288346016, + "learning_rate": 1.5121892234976404e-06, + "loss": 0.7945, + "step": 6554 + }, + { + "epoch": 0.5911529963475673, + "grad_norm": 1.5090977226502527, + "learning_rate": 1.5116226774155243e-06, + "loss": 0.9409, + "step": 6555 + }, + { + "epoch": 0.5912431798710376, + "grad_norm": 1.624525294291601, + "learning_rate": 1.5110561730044547e-06, + "loss": 0.9285, + "step": 6556 + }, + { + "epoch": 0.5913333633945078, + "grad_norm": 1.4636214789985869, + "learning_rate": 1.510489710312768e-06, + "loss": 0.9511, + "step": 6557 + }, + { + "epoch": 0.591423546917978, + "grad_norm": 1.6567037428705336, + "learning_rate": 1.5099232893887987e-06, + "loss": 0.947, + "step": 6558 + }, + { + "epoch": 0.5915137304414484, + "grad_norm": 2.0003492527240176, + "learning_rate": 1.5093569102808758e-06, + "loss": 0.9279, + "step": 6559 + }, + { + "epoch": 0.5916039139649186, + "grad_norm": 1.7058210120092339, + "learning_rate": 1.5087905730373275e-06, + "loss": 0.9578, + "step": 6560 + }, + { + "epoch": 0.5916940974883889, + "grad_norm": 0.6849763892970505, + "learning_rate": 1.508224277706476e-06, + "loss": 0.7911, + "step": 6561 + }, + { + "epoch": 0.5917842810118591, + "grad_norm": 1.6835384492576295, + "learning_rate": 1.5076580243366399e-06, + "loss": 0.9087, + "step": 6562 + }, + { + "epoch": 0.5918744645353294, + "grad_norm": 1.6880103152036745, + "learning_rate": 1.507091812976137e-06, + "loss": 0.8628, + "step": 6563 + }, + { + "epoch": 0.5919646480587997, + "grad_norm": 1.5697348371959021, + "learning_rate": 1.5065256436732773e-06, + "loss": 0.9042, + "step": 6564 + }, + { + "epoch": 0.5920548315822699, + "grad_norm": 1.5868746990310245, + "learning_rate": 1.5059595164763717e-06, + "loss": 0.8846, + "step": 6565 + }, + { + "epoch": 0.5921450151057401, + "grad_norm": 0.7212893368523967, + "learning_rate": 1.5053934314337243e-06, + "loss": 0.7872, + "step": 6566 + }, + { + "epoch": 0.5922351986292105, + "grad_norm": 1.44933530436478, + "learning_rate": 1.5048273885936356e-06, + "loss": 0.9439, + "step": 6567 + }, + { + "epoch": 0.5923253821526807, + "grad_norm": 1.3179342757206536, + "learning_rate": 1.5042613880044053e-06, + "loss": 0.871, + "step": 6568 + }, + { + "epoch": 0.592415565676151, + "grad_norm": 1.6273392199720165, + "learning_rate": 1.5036954297143265e-06, + "loss": 0.9811, + "step": 6569 + }, + { + "epoch": 0.5925057491996212, + "grad_norm": 1.1921690453217157, + "learning_rate": 1.50312951377169e-06, + "loss": 0.9051, + "step": 6570 + }, + { + "epoch": 0.5925959327230915, + "grad_norm": 0.7757710789439187, + "learning_rate": 1.502563640224784e-06, + "loss": 0.7894, + "step": 6571 + }, + { + "epoch": 0.5926861162465618, + "grad_norm": 1.6231404448528295, + "learning_rate": 1.5019978091218903e-06, + "loss": 0.9497, + "step": 6572 + }, + { + "epoch": 0.592776299770032, + "grad_norm": 1.401930341644399, + "learning_rate": 1.50143202051129e-06, + "loss": 1.0145, + "step": 6573 + }, + { + "epoch": 0.5928664832935022, + "grad_norm": 1.59809925307747, + "learning_rate": 1.500866274441258e-06, + "loss": 1.0092, + "step": 6574 + }, + { + "epoch": 0.5929566668169726, + "grad_norm": 1.932612684648462, + "learning_rate": 1.5003005709600682e-06, + "loss": 0.936, + "step": 6575 + }, + { + "epoch": 0.5930468503404428, + "grad_norm": 1.493685624399376, + "learning_rate": 1.4997349101159885e-06, + "loss": 1.0333, + "step": 6576 + }, + { + "epoch": 0.593137033863913, + "grad_norm": 1.654470099125072, + "learning_rate": 1.4991692919572854e-06, + "loss": 0.909, + "step": 6577 + }, + { + "epoch": 0.5932272173873834, + "grad_norm": 2.958751666798221, + "learning_rate": 1.4986037165322199e-06, + "loss": 0.9612, + "step": 6578 + }, + { + "epoch": 0.5933174009108536, + "grad_norm": 0.7534140526798947, + "learning_rate": 1.498038183889049e-06, + "loss": 0.8168, + "step": 6579 + }, + { + "epoch": 0.5934075844343238, + "grad_norm": 1.5517883105728978, + "learning_rate": 1.4974726940760292e-06, + "loss": 0.8721, + "step": 6580 + }, + { + "epoch": 0.5934977679577941, + "grad_norm": 1.763064730918041, + "learning_rate": 1.496907247141409e-06, + "loss": 1.0243, + "step": 6581 + }, + { + "epoch": 0.5935879514812644, + "grad_norm": 0.8377949764549139, + "learning_rate": 1.4963418431334372e-06, + "loss": 0.8583, + "step": 6582 + }, + { + "epoch": 0.5936781350047347, + "grad_norm": 1.5391985063438545, + "learning_rate": 1.4957764821003566e-06, + "loss": 0.9356, + "step": 6583 + }, + { + "epoch": 0.5937683185282049, + "grad_norm": 1.5047924254724154, + "learning_rate": 1.4952111640904063e-06, + "loss": 0.9563, + "step": 6584 + }, + { + "epoch": 0.5938585020516751, + "grad_norm": 1.8337629710689767, + "learning_rate": 1.494645889151823e-06, + "loss": 0.9356, + "step": 6585 + }, + { + "epoch": 0.5939486855751455, + "grad_norm": 0.7030007782452916, + "learning_rate": 1.494080657332839e-06, + "loss": 0.7632, + "step": 6586 + }, + { + "epoch": 0.5940388690986157, + "grad_norm": 1.6891655649646742, + "learning_rate": 1.4935154686816832e-06, + "loss": 0.9099, + "step": 6587 + }, + { + "epoch": 0.5941290526220859, + "grad_norm": 1.6527089105500323, + "learning_rate": 1.4929503232465802e-06, + "loss": 1.0279, + "step": 6588 + }, + { + "epoch": 0.5942192361455562, + "grad_norm": 1.4611192070060228, + "learning_rate": 1.492385221075751e-06, + "loss": 0.8611, + "step": 6589 + }, + { + "epoch": 0.5943094196690265, + "grad_norm": 1.499752739554229, + "learning_rate": 1.4918201622174142e-06, + "loss": 0.9816, + "step": 6590 + }, + { + "epoch": 0.5943996031924967, + "grad_norm": 1.5331452765758742, + "learning_rate": 1.4912551467197827e-06, + "loss": 0.9754, + "step": 6591 + }, + { + "epoch": 0.594489786715967, + "grad_norm": 2.0754330056887635, + "learning_rate": 1.4906901746310678e-06, + "loss": 1.0352, + "step": 6592 + }, + { + "epoch": 0.5945799702394372, + "grad_norm": 1.340097564590071, + "learning_rate": 1.4901252459994757e-06, + "loss": 0.9121, + "step": 6593 + }, + { + "epoch": 0.5946701537629075, + "grad_norm": 3.3402018037237826, + "learning_rate": 1.489560360873208e-06, + "loss": 0.9393, + "step": 6594 + }, + { + "epoch": 0.5947603372863778, + "grad_norm": 1.560380795537198, + "learning_rate": 1.4889955193004659e-06, + "loss": 0.9277, + "step": 6595 + }, + { + "epoch": 0.594850520809848, + "grad_norm": 0.6913593513402034, + "learning_rate": 1.4884307213294428e-06, + "loss": 0.7759, + "step": 6596 + }, + { + "epoch": 0.5949407043333182, + "grad_norm": 1.708439017337553, + "learning_rate": 1.4878659670083321e-06, + "loss": 0.9356, + "step": 6597 + }, + { + "epoch": 0.5950308878567886, + "grad_norm": 1.469869227788209, + "learning_rate": 1.4873012563853208e-06, + "loss": 0.875, + "step": 6598 + }, + { + "epoch": 0.5951210713802588, + "grad_norm": 1.8736461201909673, + "learning_rate": 1.4867365895085935e-06, + "loss": 0.9583, + "step": 6599 + }, + { + "epoch": 0.5952112549037291, + "grad_norm": 1.7032485357155351, + "learning_rate": 1.4861719664263301e-06, + "loss": 0.9297, + "step": 6600 + }, + { + "epoch": 0.5953014384271994, + "grad_norm": 1.5321130461306232, + "learning_rate": 1.485607387186708e-06, + "loss": 1.0526, + "step": 6601 + }, + { + "epoch": 0.5953916219506696, + "grad_norm": 1.6585819454195294, + "learning_rate": 1.4850428518379001e-06, + "loss": 0.9467, + "step": 6602 + }, + { + "epoch": 0.5954818054741399, + "grad_norm": 1.300338257284913, + "learning_rate": 1.4844783604280746e-06, + "loss": 0.9439, + "step": 6603 + }, + { + "epoch": 0.5955719889976101, + "grad_norm": 1.860140090096753, + "learning_rate": 1.483913913005399e-06, + "loss": 0.8906, + "step": 6604 + }, + { + "epoch": 0.5956621725210804, + "grad_norm": 1.695670094549515, + "learning_rate": 1.483349509618034e-06, + "loss": 1.0055, + "step": 6605 + }, + { + "epoch": 0.5957523560445507, + "grad_norm": 1.5118956942835264, + "learning_rate": 1.4827851503141367e-06, + "loss": 0.8794, + "step": 6606 + }, + { + "epoch": 0.5958425395680209, + "grad_norm": 1.4236826043927548, + "learning_rate": 1.482220835141863e-06, + "loss": 0.8737, + "step": 6607 + }, + { + "epoch": 0.5959327230914911, + "grad_norm": 1.59831377223631, + "learning_rate": 1.481656564149362e-06, + "loss": 1.0316, + "step": 6608 + }, + { + "epoch": 0.5960229066149615, + "grad_norm": 1.9579650113191316, + "learning_rate": 1.4810923373847818e-06, + "loss": 0.9499, + "step": 6609 + }, + { + "epoch": 0.5961130901384317, + "grad_norm": 1.6702616384986033, + "learning_rate": 1.4805281548962647e-06, + "loss": 0.9556, + "step": 6610 + }, + { + "epoch": 0.596203273661902, + "grad_norm": 1.547657393060808, + "learning_rate": 1.4799640167319488e-06, + "loss": 0.8978, + "step": 6611 + }, + { + "epoch": 0.5962934571853722, + "grad_norm": 1.4517594350669494, + "learning_rate": 1.4793999229399714e-06, + "loss": 0.9391, + "step": 6612 + }, + { + "epoch": 0.5963836407088425, + "grad_norm": 1.457702645663315, + "learning_rate": 1.4788358735684626e-06, + "loss": 1.0153, + "step": 6613 + }, + { + "epoch": 0.5964738242323128, + "grad_norm": 2.4117586568157243, + "learning_rate": 1.4782718686655514e-06, + "loss": 1.0222, + "step": 6614 + }, + { + "epoch": 0.596564007755783, + "grad_norm": 1.6173745752020494, + "learning_rate": 1.4777079082793605e-06, + "loss": 1.0031, + "step": 6615 + }, + { + "epoch": 0.5966541912792532, + "grad_norm": 1.5213445126615617, + "learning_rate": 1.4771439924580108e-06, + "loss": 0.9248, + "step": 6616 + }, + { + "epoch": 0.5967443748027236, + "grad_norm": 1.470523088639067, + "learning_rate": 1.4765801212496189e-06, + "loss": 0.9932, + "step": 6617 + }, + { + "epoch": 0.5968345583261938, + "grad_norm": 1.4978387680631433, + "learning_rate": 1.4760162947022968e-06, + "loss": 0.8668, + "step": 6618 + }, + { + "epoch": 0.596924741849664, + "grad_norm": 2.667153095544161, + "learning_rate": 1.475452512864154e-06, + "loss": 0.9419, + "step": 6619 + }, + { + "epoch": 0.5970149253731343, + "grad_norm": 1.324217804764942, + "learning_rate": 1.4748887757832945e-06, + "loss": 0.9591, + "step": 6620 + }, + { + "epoch": 0.5971051088966046, + "grad_norm": 1.4756315818811097, + "learning_rate": 1.4743250835078209e-06, + "loss": 0.9844, + "step": 6621 + }, + { + "epoch": 0.5971952924200749, + "grad_norm": 1.3738645286823923, + "learning_rate": 1.4737614360858297e-06, + "loss": 0.9126, + "step": 6622 + }, + { + "epoch": 0.5972854759435451, + "grad_norm": 2.302822379904953, + "learning_rate": 1.4731978335654133e-06, + "loss": 0.9043, + "step": 6623 + }, + { + "epoch": 0.5973756594670154, + "grad_norm": 1.5436637761673746, + "learning_rate": 1.4726342759946638e-06, + "loss": 0.8929, + "step": 6624 + }, + { + "epoch": 0.5974658429904857, + "grad_norm": 1.4313215079622306, + "learning_rate": 1.4720707634216653e-06, + "loss": 0.9497, + "step": 6625 + }, + { + "epoch": 0.5975560265139559, + "grad_norm": 1.810897842512784, + "learning_rate": 1.4715072958945e-06, + "loss": 0.964, + "step": 6626 + }, + { + "epoch": 0.5976462100374261, + "grad_norm": 1.700779691177656, + "learning_rate": 1.470943873461247e-06, + "loss": 0.9527, + "step": 6627 + }, + { + "epoch": 0.5977363935608965, + "grad_norm": 1.9839811159434986, + "learning_rate": 1.470380496169979e-06, + "loss": 0.9103, + "step": 6628 + }, + { + "epoch": 0.5978265770843667, + "grad_norm": 1.4729697301102025, + "learning_rate": 1.4698171640687682e-06, + "loss": 0.9424, + "step": 6629 + }, + { + "epoch": 0.5979167606078369, + "grad_norm": 1.508391748354487, + "learning_rate": 1.4692538772056792e-06, + "loss": 0.8823, + "step": 6630 + }, + { + "epoch": 0.5980069441313072, + "grad_norm": 1.5635373296133228, + "learning_rate": 1.4686906356287772e-06, + "loss": 0.8625, + "step": 6631 + }, + { + "epoch": 0.5980971276547775, + "grad_norm": 1.4118761896311631, + "learning_rate": 1.4681274393861194e-06, + "loss": 0.9811, + "step": 6632 + }, + { + "epoch": 0.5981873111782477, + "grad_norm": 1.4169247990469271, + "learning_rate": 1.4675642885257603e-06, + "loss": 1.0455, + "step": 6633 + }, + { + "epoch": 0.598277494701718, + "grad_norm": 5.700511092980296, + "learning_rate": 1.4670011830957529e-06, + "loss": 1.0371, + "step": 6634 + }, + { + "epoch": 0.5983676782251882, + "grad_norm": 1.591731420002015, + "learning_rate": 1.4664381231441427e-06, + "loss": 0.9232, + "step": 6635 + }, + { + "epoch": 0.5984578617486586, + "grad_norm": 1.6518097864701948, + "learning_rate": 1.4658751087189746e-06, + "loss": 0.9583, + "step": 6636 + }, + { + "epoch": 0.5985480452721288, + "grad_norm": 1.8132377964452207, + "learning_rate": 1.4653121398682874e-06, + "loss": 1.0035, + "step": 6637 + }, + { + "epoch": 0.598638228795599, + "grad_norm": 1.3480454477210324, + "learning_rate": 1.4647492166401159e-06, + "loss": 0.8909, + "step": 6638 + }, + { + "epoch": 0.5987284123190693, + "grad_norm": 1.5571700168755815, + "learning_rate": 1.4641863390824934e-06, + "loss": 0.9369, + "step": 6639 + }, + { + "epoch": 0.5988185958425396, + "grad_norm": 1.4999062190939338, + "learning_rate": 1.4636235072434465e-06, + "loss": 0.9469, + "step": 6640 + }, + { + "epoch": 0.5989087793660098, + "grad_norm": 1.3753544177268116, + "learning_rate": 1.4630607211709994e-06, + "loss": 0.9798, + "step": 6641 + }, + { + "epoch": 0.5989989628894801, + "grad_norm": 1.415839411853647, + "learning_rate": 1.4624979809131723e-06, + "loss": 0.9562, + "step": 6642 + }, + { + "epoch": 0.5990891464129503, + "grad_norm": 1.3999135416445885, + "learning_rate": 1.4619352865179814e-06, + "loss": 0.8608, + "step": 6643 + }, + { + "epoch": 0.5991793299364206, + "grad_norm": 1.5684904848928287, + "learning_rate": 1.4613726380334391e-06, + "loss": 0.9225, + "step": 6644 + }, + { + "epoch": 0.5992695134598909, + "grad_norm": 1.6884142200616736, + "learning_rate": 1.4608100355075522e-06, + "loss": 0.8074, + "step": 6645 + }, + { + "epoch": 0.5993596969833611, + "grad_norm": 0.7028122524262734, + "learning_rate": 1.460247478988327e-06, + "loss": 0.7571, + "step": 6646 + }, + { + "epoch": 0.5994498805068315, + "grad_norm": 1.4078549021983484, + "learning_rate": 1.4596849685237623e-06, + "loss": 0.966, + "step": 6647 + }, + { + "epoch": 0.5995400640303017, + "grad_norm": 1.3754391835813817, + "learning_rate": 1.459122504161856e-06, + "loss": 0.8569, + "step": 6648 + }, + { + "epoch": 0.5996302475537719, + "grad_norm": 1.4359876721471936, + "learning_rate": 1.4585600859506001e-06, + "loss": 0.9231, + "step": 6649 + }, + { + "epoch": 0.5997204310772422, + "grad_norm": 1.4602664487513493, + "learning_rate": 1.4579977139379826e-06, + "loss": 1.0077, + "step": 6650 + }, + { + "epoch": 0.5998106146007125, + "grad_norm": 1.6727438122040954, + "learning_rate": 1.4574353881719895e-06, + "loss": 1.0118, + "step": 6651 + }, + { + "epoch": 0.5999007981241827, + "grad_norm": 1.5862007110609502, + "learning_rate": 1.4568731087005998e-06, + "loss": 0.9506, + "step": 6652 + }, + { + "epoch": 0.599990981647653, + "grad_norm": 1.3319599658316716, + "learning_rate": 1.4563108755717916e-06, + "loss": 0.9734, + "step": 6653 + }, + { + "epoch": 0.6000811651711232, + "grad_norm": 1.4791854893019585, + "learning_rate": 1.455748688833538e-06, + "loss": 0.9773, + "step": 6654 + }, + { + "epoch": 0.6001713486945935, + "grad_norm": 1.7592253707759802, + "learning_rate": 1.4551865485338065e-06, + "loss": 0.8606, + "step": 6655 + }, + { + "epoch": 0.6002615322180638, + "grad_norm": 1.4963380300428737, + "learning_rate": 1.4546244547205629e-06, + "loss": 1.0148, + "step": 6656 + }, + { + "epoch": 0.600351715741534, + "grad_norm": 1.5467443700021741, + "learning_rate": 1.4540624074417678e-06, + "loss": 0.9749, + "step": 6657 + }, + { + "epoch": 0.6004418992650042, + "grad_norm": 1.3590064151968988, + "learning_rate": 1.453500406745379e-06, + "loss": 1.0103, + "step": 6658 + }, + { + "epoch": 0.6005320827884746, + "grad_norm": 2.8974270357386396, + "learning_rate": 1.4529384526793486e-06, + "loss": 1.0249, + "step": 6659 + }, + { + "epoch": 0.6006222663119448, + "grad_norm": 1.5023121497708527, + "learning_rate": 1.4523765452916252e-06, + "loss": 1.0049, + "step": 6660 + }, + { + "epoch": 0.600712449835415, + "grad_norm": 1.5027531153891405, + "learning_rate": 1.4518146846301554e-06, + "loss": 0.9773, + "step": 6661 + }, + { + "epoch": 0.6008026333588853, + "grad_norm": 1.3201465389523472, + "learning_rate": 1.4512528707428787e-06, + "loss": 0.9288, + "step": 6662 + }, + { + "epoch": 0.6008928168823556, + "grad_norm": 1.4290630039770595, + "learning_rate": 1.4506911036777335e-06, + "loss": 0.9348, + "step": 6663 + }, + { + "epoch": 0.6009830004058259, + "grad_norm": 0.7210429974484629, + "learning_rate": 1.450129383482651e-06, + "loss": 0.7495, + "step": 6664 + }, + { + "epoch": 0.6010731839292961, + "grad_norm": 1.344463336477315, + "learning_rate": 1.4495677102055629e-06, + "loss": 0.9079, + "step": 6665 + }, + { + "epoch": 0.6011633674527663, + "grad_norm": 1.5670593118014045, + "learning_rate": 1.4490060838943924e-06, + "loss": 0.9878, + "step": 6666 + }, + { + "epoch": 0.6012535509762367, + "grad_norm": 1.4391041802449573, + "learning_rate": 1.4484445045970609e-06, + "loss": 0.9501, + "step": 6667 + }, + { + "epoch": 0.6013437344997069, + "grad_norm": 1.4677312036505301, + "learning_rate": 1.447882972361485e-06, + "loss": 0.9435, + "step": 6668 + }, + { + "epoch": 0.6014339180231771, + "grad_norm": 1.592951331387149, + "learning_rate": 1.4473214872355785e-06, + "loss": 0.9329, + "step": 6669 + }, + { + "epoch": 0.6015241015466475, + "grad_norm": 1.7472232860379333, + "learning_rate": 1.4467600492672508e-06, + "loss": 0.9885, + "step": 6670 + }, + { + "epoch": 0.6016142850701177, + "grad_norm": 1.5383782584317893, + "learning_rate": 1.4461986585044054e-06, + "loss": 0.9305, + "step": 6671 + }, + { + "epoch": 0.601704468593588, + "grad_norm": 1.6353036229938558, + "learning_rate": 1.4456373149949446e-06, + "loss": 0.9306, + "step": 6672 + }, + { + "epoch": 0.6017946521170582, + "grad_norm": 1.4429642855374307, + "learning_rate": 1.4450760187867648e-06, + "loss": 0.9601, + "step": 6673 + }, + { + "epoch": 0.6018848356405285, + "grad_norm": 1.4822906315420812, + "learning_rate": 1.4445147699277581e-06, + "loss": 0.9427, + "step": 6674 + }, + { + "epoch": 0.6019750191639988, + "grad_norm": 1.7223644052627165, + "learning_rate": 1.4439535684658154e-06, + "loss": 0.9958, + "step": 6675 + }, + { + "epoch": 0.602065202687469, + "grad_norm": 0.8270120519546353, + "learning_rate": 1.44339241444882e-06, + "loss": 0.8195, + "step": 6676 + }, + { + "epoch": 0.6021553862109392, + "grad_norm": 1.8310985671825635, + "learning_rate": 1.4428313079246518e-06, + "loss": 0.8957, + "step": 6677 + }, + { + "epoch": 0.6022455697344096, + "grad_norm": 0.725617522300548, + "learning_rate": 1.4422702489411896e-06, + "loss": 0.7969, + "step": 6678 + }, + { + "epoch": 0.6023357532578798, + "grad_norm": 1.5324527921141955, + "learning_rate": 1.4417092375463043e-06, + "loss": 0.8994, + "step": 6679 + }, + { + "epoch": 0.60242593678135, + "grad_norm": 1.568588829120206, + "learning_rate": 1.441148273787866e-06, + "loss": 0.9892, + "step": 6680 + }, + { + "epoch": 0.6025161203048203, + "grad_norm": 0.8041975511210511, + "learning_rate": 1.4405873577137383e-06, + "loss": 0.8028, + "step": 6681 + }, + { + "epoch": 0.6026063038282906, + "grad_norm": 1.6614253679630504, + "learning_rate": 1.4400264893717816e-06, + "loss": 0.9616, + "step": 6682 + }, + { + "epoch": 0.6026964873517608, + "grad_norm": 1.5422948983468747, + "learning_rate": 1.4394656688098526e-06, + "loss": 1.0015, + "step": 6683 + }, + { + "epoch": 0.6027866708752311, + "grad_norm": 1.4255636806922167, + "learning_rate": 1.4389048960758032e-06, + "loss": 0.9771, + "step": 6684 + }, + { + "epoch": 0.6028768543987013, + "grad_norm": 1.785315375049479, + "learning_rate": 1.4383441712174826e-06, + "loss": 0.9836, + "step": 6685 + }, + { + "epoch": 0.6029670379221717, + "grad_norm": 1.5808167116494176, + "learning_rate": 1.4377834942827333e-06, + "loss": 1.0168, + "step": 6686 + }, + { + "epoch": 0.6030572214456419, + "grad_norm": 1.5651046405945968, + "learning_rate": 1.437222865319397e-06, + "loss": 0.8923, + "step": 6687 + }, + { + "epoch": 0.6031474049691121, + "grad_norm": 1.4953376313492301, + "learning_rate": 1.4366622843753092e-06, + "loss": 0.8786, + "step": 6688 + }, + { + "epoch": 0.6032375884925824, + "grad_norm": 1.3489866462397222, + "learning_rate": 1.4361017514983006e-06, + "loss": 0.9704, + "step": 6689 + }, + { + "epoch": 0.6033277720160527, + "grad_norm": 1.6401625935272932, + "learning_rate": 1.4355412667362006e-06, + "loss": 1.0087, + "step": 6690 + }, + { + "epoch": 0.6034179555395229, + "grad_norm": 1.6642839884325127, + "learning_rate": 1.4349808301368311e-06, + "loss": 0.9054, + "step": 6691 + }, + { + "epoch": 0.6035081390629932, + "grad_norm": 1.747846231719927, + "learning_rate": 1.4344204417480139e-06, + "loss": 0.9944, + "step": 6692 + }, + { + "epoch": 0.6035983225864634, + "grad_norm": 1.6380371028251874, + "learning_rate": 1.4338601016175628e-06, + "loss": 0.9066, + "step": 6693 + }, + { + "epoch": 0.6036885061099337, + "grad_norm": 1.845717205014086, + "learning_rate": 1.433299809793289e-06, + "loss": 0.9495, + "step": 6694 + }, + { + "epoch": 0.603778689633404, + "grad_norm": 2.122849666697447, + "learning_rate": 1.432739566323001e-06, + "loss": 0.8214, + "step": 6695 + }, + { + "epoch": 0.6038688731568742, + "grad_norm": 1.6663594042152863, + "learning_rate": 1.4321793712545004e-06, + "loss": 0.9187, + "step": 6696 + }, + { + "epoch": 0.6039590566803446, + "grad_norm": 1.7609907113457943, + "learning_rate": 1.4316192246355873e-06, + "loss": 1.0004, + "step": 6697 + }, + { + "epoch": 0.6040492402038148, + "grad_norm": 0.6422060785135402, + "learning_rate": 1.4310591265140555e-06, + "loss": 0.7799, + "step": 6698 + }, + { + "epoch": 0.604139423727285, + "grad_norm": 1.3511759034201791, + "learning_rate": 1.4304990769376963e-06, + "loss": 0.961, + "step": 6699 + }, + { + "epoch": 0.6042296072507553, + "grad_norm": 1.397828806102537, + "learning_rate": 1.4299390759542962e-06, + "loss": 0.9391, + "step": 6700 + }, + { + "epoch": 0.6043197907742256, + "grad_norm": 1.4951952434872346, + "learning_rate": 1.4293791236116368e-06, + "loss": 0.9329, + "step": 6701 + }, + { + "epoch": 0.6044099742976958, + "grad_norm": 1.595955762159242, + "learning_rate": 1.4288192199574978e-06, + "loss": 0.9165, + "step": 6702 + }, + { + "epoch": 0.6045001578211661, + "grad_norm": 2.3991138610469953, + "learning_rate": 1.4282593650396524e-06, + "loss": 0.7535, + "step": 6703 + }, + { + "epoch": 0.6045903413446363, + "grad_norm": 1.4432985863383128, + "learning_rate": 1.4276995589058695e-06, + "loss": 1.0502, + "step": 6704 + }, + { + "epoch": 0.6046805248681066, + "grad_norm": 0.6392726813713299, + "learning_rate": 1.4271398016039168e-06, + "loss": 0.7495, + "step": 6705 + }, + { + "epoch": 0.6047707083915769, + "grad_norm": 1.743770957465346, + "learning_rate": 1.4265800931815542e-06, + "loss": 0.9936, + "step": 6706 + }, + { + "epoch": 0.6048608919150471, + "grad_norm": 1.3748625339803375, + "learning_rate": 1.4260204336865406e-06, + "loss": 0.9642, + "step": 6707 + }, + { + "epoch": 0.6049510754385173, + "grad_norm": 1.6928403206254434, + "learning_rate": 1.4254608231666281e-06, + "loss": 0.9522, + "step": 6708 + }, + { + "epoch": 0.6050412589619877, + "grad_norm": 1.380596044054571, + "learning_rate": 1.4249012616695661e-06, + "loss": 0.9593, + "step": 6709 + }, + { + "epoch": 0.6051314424854579, + "grad_norm": 1.8842806968056134, + "learning_rate": 1.4243417492431e-06, + "loss": 0.9552, + "step": 6710 + }, + { + "epoch": 0.6052216260089281, + "grad_norm": 1.4785205432805004, + "learning_rate": 1.4237822859349696e-06, + "loss": 0.9353, + "step": 6711 + }, + { + "epoch": 0.6053118095323984, + "grad_norm": 1.3849507411225854, + "learning_rate": 1.423222871792912e-06, + "loss": 0.9267, + "step": 6712 + }, + { + "epoch": 0.6054019930558687, + "grad_norm": 1.4860999274303432, + "learning_rate": 1.4226635068646586e-06, + "loss": 1.0118, + "step": 6713 + }, + { + "epoch": 0.605492176579339, + "grad_norm": 0.6779270173165575, + "learning_rate": 1.4221041911979393e-06, + "loss": 0.7733, + "step": 6714 + }, + { + "epoch": 0.6055823601028092, + "grad_norm": 1.708379845645056, + "learning_rate": 1.4215449248404765e-06, + "loss": 0.9681, + "step": 6715 + }, + { + "epoch": 0.6056725436262794, + "grad_norm": 1.4185188004246554, + "learning_rate": 1.4209857078399896e-06, + "loss": 0.9697, + "step": 6716 + }, + { + "epoch": 0.6057627271497498, + "grad_norm": 2.3385043633309976, + "learning_rate": 1.4204265402441955e-06, + "loss": 0.9254, + "step": 6717 + }, + { + "epoch": 0.60585291067322, + "grad_norm": 1.3281489650583702, + "learning_rate": 1.419867422100804e-06, + "loss": 0.8322, + "step": 6718 + }, + { + "epoch": 0.6059430941966902, + "grad_norm": 1.4489198772374152, + "learning_rate": 1.4193083534575236e-06, + "loss": 0.9334, + "step": 6719 + }, + { + "epoch": 0.6060332777201606, + "grad_norm": 1.6620474063066153, + "learning_rate": 1.4187493343620567e-06, + "loss": 0.9749, + "step": 6720 + }, + { + "epoch": 0.6061234612436308, + "grad_norm": 1.4503326659227707, + "learning_rate": 1.4181903648621006e-06, + "loss": 0.982, + "step": 6721 + }, + { + "epoch": 0.606213644767101, + "grad_norm": 1.320189250181873, + "learning_rate": 1.4176314450053512e-06, + "loss": 0.9839, + "step": 6722 + }, + { + "epoch": 0.6063038282905713, + "grad_norm": 1.5673933842858037, + "learning_rate": 1.4170725748394977e-06, + "loss": 0.9371, + "step": 6723 + }, + { + "epoch": 0.6063940118140416, + "grad_norm": 1.4103153804696384, + "learning_rate": 1.4165137544122266e-06, + "loss": 0.8607, + "step": 6724 + }, + { + "epoch": 0.6064841953375119, + "grad_norm": 1.4821991080871706, + "learning_rate": 1.4159549837712194e-06, + "loss": 1.0193, + "step": 6725 + }, + { + "epoch": 0.6065743788609821, + "grad_norm": 1.6492633649103248, + "learning_rate": 1.415396262964153e-06, + "loss": 0.8699, + "step": 6726 + }, + { + "epoch": 0.6066645623844523, + "grad_norm": 1.7469160290098373, + "learning_rate": 1.4148375920387016e-06, + "loss": 1.0237, + "step": 6727 + }, + { + "epoch": 0.6067547459079227, + "grad_norm": 0.673263910671857, + "learning_rate": 1.4142789710425325e-06, + "loss": 0.7767, + "step": 6728 + }, + { + "epoch": 0.6068449294313929, + "grad_norm": 1.8227056399315433, + "learning_rate": 1.4137204000233118e-06, + "loss": 0.8899, + "step": 6729 + }, + { + "epoch": 0.6069351129548631, + "grad_norm": 1.6835955201602135, + "learning_rate": 1.4131618790286987e-06, + "loss": 0.9448, + "step": 6730 + }, + { + "epoch": 0.6070252964783334, + "grad_norm": 1.5061792412437307, + "learning_rate": 1.4126034081063506e-06, + "loss": 0.9402, + "step": 6731 + }, + { + "epoch": 0.6071154800018037, + "grad_norm": 1.6671467884244784, + "learning_rate": 1.4120449873039186e-06, + "loss": 0.9378, + "step": 6732 + }, + { + "epoch": 0.6072056635252739, + "grad_norm": 1.6302055464488168, + "learning_rate": 1.4114866166690494e-06, + "loss": 0.9902, + "step": 6733 + }, + { + "epoch": 0.6072958470487442, + "grad_norm": 1.1851697947382578, + "learning_rate": 1.4109282962493877e-06, + "loss": 0.9105, + "step": 6734 + }, + { + "epoch": 0.6073860305722144, + "grad_norm": 1.8334332857917401, + "learning_rate": 1.4103700260925716e-06, + "loss": 0.9454, + "step": 6735 + }, + { + "epoch": 0.6074762140956848, + "grad_norm": 1.8460837642804808, + "learning_rate": 1.4098118062462364e-06, + "loss": 0.989, + "step": 6736 + }, + { + "epoch": 0.607566397619155, + "grad_norm": 1.317838891036253, + "learning_rate": 1.4092536367580123e-06, + "loss": 0.9491, + "step": 6737 + }, + { + "epoch": 0.6076565811426252, + "grad_norm": 1.8697043339161004, + "learning_rate": 1.4086955176755248e-06, + "loss": 0.8659, + "step": 6738 + }, + { + "epoch": 0.6077467646660955, + "grad_norm": 1.6914686231147085, + "learning_rate": 1.4081374490463964e-06, + "loss": 0.8891, + "step": 6739 + }, + { + "epoch": 0.6078369481895658, + "grad_norm": 1.691631205270095, + "learning_rate": 1.4075794309182443e-06, + "loss": 0.9645, + "step": 6740 + }, + { + "epoch": 0.607927131713036, + "grad_norm": 1.8979046234630899, + "learning_rate": 1.407021463338682e-06, + "loss": 0.9653, + "step": 6741 + }, + { + "epoch": 0.6080173152365063, + "grad_norm": 0.6461858838222846, + "learning_rate": 1.4064635463553177e-06, + "loss": 0.7222, + "step": 6742 + }, + { + "epoch": 0.6081074987599766, + "grad_norm": 1.7838378979760166, + "learning_rate": 1.4059056800157563e-06, + "loss": 0.9785, + "step": 6743 + }, + { + "epoch": 0.6081976822834468, + "grad_norm": 1.6050259241642735, + "learning_rate": 1.4053478643675982e-06, + "loss": 0.8841, + "step": 6744 + }, + { + "epoch": 0.6082878658069171, + "grad_norm": 1.5493837300462816, + "learning_rate": 1.4047900994584389e-06, + "loss": 0.8501, + "step": 6745 + }, + { + "epoch": 0.6083780493303873, + "grad_norm": 1.582373838653729, + "learning_rate": 1.404232385335871e-06, + "loss": 0.8621, + "step": 6746 + }, + { + "epoch": 0.6084682328538576, + "grad_norm": 1.573983451732036, + "learning_rate": 1.4036747220474806e-06, + "loss": 0.952, + "step": 6747 + }, + { + "epoch": 0.6085584163773279, + "grad_norm": 2.034986492552672, + "learning_rate": 1.4031171096408506e-06, + "loss": 0.903, + "step": 6748 + }, + { + "epoch": 0.6086485999007981, + "grad_norm": 1.5794808634642987, + "learning_rate": 1.4025595481635607e-06, + "loss": 0.9751, + "step": 6749 + }, + { + "epoch": 0.6087387834242683, + "grad_norm": 1.6533655925392827, + "learning_rate": 1.4020020376631836e-06, + "loss": 0.972, + "step": 6750 + }, + { + "epoch": 0.6088289669477387, + "grad_norm": 1.787590387366572, + "learning_rate": 1.4014445781872908e-06, + "loss": 0.9615, + "step": 6751 + }, + { + "epoch": 0.6089191504712089, + "grad_norm": 1.4939852764675694, + "learning_rate": 1.4008871697834465e-06, + "loss": 1.0136, + "step": 6752 + }, + { + "epoch": 0.6090093339946792, + "grad_norm": 1.8232419259674424, + "learning_rate": 1.400329812499213e-06, + "loss": 0.9829, + "step": 6753 + }, + { + "epoch": 0.6090995175181494, + "grad_norm": 1.3262924117085158, + "learning_rate": 1.3997725063821458e-06, + "loss": 0.9659, + "step": 6754 + }, + { + "epoch": 0.6091897010416197, + "grad_norm": 1.653500343791048, + "learning_rate": 1.3992152514797978e-06, + "loss": 0.9157, + "step": 6755 + }, + { + "epoch": 0.60927988456509, + "grad_norm": 0.6966953242680396, + "learning_rate": 1.398658047839718e-06, + "loss": 0.7973, + "step": 6756 + }, + { + "epoch": 0.6093700680885602, + "grad_norm": 0.6444335632277965, + "learning_rate": 1.3981008955094481e-06, + "loss": 0.7671, + "step": 6757 + }, + { + "epoch": 0.6094602516120304, + "grad_norm": 1.297264086714662, + "learning_rate": 1.39754379453653e-06, + "loss": 0.9264, + "step": 6758 + }, + { + "epoch": 0.6095504351355008, + "grad_norm": 1.5642962240150986, + "learning_rate": 1.3969867449684972e-06, + "loss": 0.9624, + "step": 6759 + }, + { + "epoch": 0.609640618658971, + "grad_norm": 1.8881757019367338, + "learning_rate": 1.396429746852879e-06, + "loss": 0.9449, + "step": 6760 + }, + { + "epoch": 0.6097308021824412, + "grad_norm": 2.0499372798350013, + "learning_rate": 1.395872800237204e-06, + "loss": 0.9306, + "step": 6761 + }, + { + "epoch": 0.6098209857059115, + "grad_norm": 0.8110206781848539, + "learning_rate": 1.3953159051689918e-06, + "loss": 0.7981, + "step": 6762 + }, + { + "epoch": 0.6099111692293818, + "grad_norm": 1.5449650079659518, + "learning_rate": 1.3947590616957618e-06, + "loss": 0.996, + "step": 6763 + }, + { + "epoch": 0.610001352752852, + "grad_norm": 1.8165886243919909, + "learning_rate": 1.3942022698650258e-06, + "loss": 0.979, + "step": 6764 + }, + { + "epoch": 0.6100915362763223, + "grad_norm": 1.5480892925433527, + "learning_rate": 1.3936455297242917e-06, + "loss": 0.9836, + "step": 6765 + }, + { + "epoch": 0.6101817197997926, + "grad_norm": 1.6237852251021783, + "learning_rate": 1.3930888413210652e-06, + "loss": 0.9706, + "step": 6766 + }, + { + "epoch": 0.6102719033232629, + "grad_norm": 1.610961354446285, + "learning_rate": 1.392532204702845e-06, + "loss": 0.916, + "step": 6767 + }, + { + "epoch": 0.6103620868467331, + "grad_norm": 1.8995591605890132, + "learning_rate": 1.3919756199171266e-06, + "loss": 0.8765, + "step": 6768 + }, + { + "epoch": 0.6104522703702033, + "grad_norm": 1.3334699898465698, + "learning_rate": 1.3914190870114009e-06, + "loss": 1.0241, + "step": 6769 + }, + { + "epoch": 0.6105424538936737, + "grad_norm": 1.450162677847237, + "learning_rate": 1.3908626060331541e-06, + "loss": 0.9978, + "step": 6770 + }, + { + "epoch": 0.6106326374171439, + "grad_norm": 1.5416938246663252, + "learning_rate": 1.3903061770298693e-06, + "loss": 0.9858, + "step": 6771 + }, + { + "epoch": 0.6107228209406141, + "grad_norm": 1.6087442291559368, + "learning_rate": 1.3897498000490223e-06, + "loss": 0.9286, + "step": 6772 + }, + { + "epoch": 0.6108130044640844, + "grad_norm": 1.4820563427334668, + "learning_rate": 1.3891934751380879e-06, + "loss": 1.0179, + "step": 6773 + }, + { + "epoch": 0.6109031879875547, + "grad_norm": 1.4508452221103765, + "learning_rate": 1.3886372023445334e-06, + "loss": 0.9737, + "step": 6774 + }, + { + "epoch": 0.610993371511025, + "grad_norm": 1.431016481497891, + "learning_rate": 1.3880809817158246e-06, + "loss": 1.0189, + "step": 6775 + }, + { + "epoch": 0.6110835550344952, + "grad_norm": 1.6245975729680344, + "learning_rate": 1.3875248132994206e-06, + "loss": 0.9456, + "step": 6776 + }, + { + "epoch": 0.6111737385579654, + "grad_norm": 1.4993682961673827, + "learning_rate": 1.386968697142776e-06, + "loss": 0.8732, + "step": 6777 + }, + { + "epoch": 0.6112639220814358, + "grad_norm": 1.5469883578601549, + "learning_rate": 1.386412633293343e-06, + "loss": 0.9997, + "step": 6778 + }, + { + "epoch": 0.611354105604906, + "grad_norm": 1.7648297434703155, + "learning_rate": 1.3858566217985672e-06, + "loss": 0.924, + "step": 6779 + }, + { + "epoch": 0.6114442891283762, + "grad_norm": 1.7180272056415, + "learning_rate": 1.3853006627058905e-06, + "loss": 1.0353, + "step": 6780 + }, + { + "epoch": 0.6115344726518465, + "grad_norm": 1.8267421872243508, + "learning_rate": 1.3847447560627512e-06, + "loss": 0.8398, + "step": 6781 + }, + { + "epoch": 0.6116246561753168, + "grad_norm": 1.686708193909838, + "learning_rate": 1.3841889019165812e-06, + "loss": 0.9996, + "step": 6782 + }, + { + "epoch": 0.611714839698787, + "grad_norm": 2.435391712523165, + "learning_rate": 1.3836331003148106e-06, + "loss": 0.9098, + "step": 6783 + }, + { + "epoch": 0.6118050232222573, + "grad_norm": 1.3137403712477838, + "learning_rate": 1.3830773513048612e-06, + "loss": 0.9918, + "step": 6784 + }, + { + "epoch": 0.6118952067457275, + "grad_norm": 1.5373145697199364, + "learning_rate": 1.382521654934155e-06, + "loss": 0.9708, + "step": 6785 + }, + { + "epoch": 0.6119853902691978, + "grad_norm": 1.5039162375378587, + "learning_rate": 1.3819660112501057e-06, + "loss": 0.8972, + "step": 6786 + }, + { + "epoch": 0.6120755737926681, + "grad_norm": 1.3986002225345722, + "learning_rate": 1.3814104203001234e-06, + "loss": 0.9126, + "step": 6787 + }, + { + "epoch": 0.6121657573161383, + "grad_norm": 1.5577008359445559, + "learning_rate": 1.3808548821316156e-06, + "loss": 0.9899, + "step": 6788 + }, + { + "epoch": 0.6122559408396087, + "grad_norm": 1.3895267049907485, + "learning_rate": 1.3802993967919824e-06, + "loss": 0.886, + "step": 6789 + }, + { + "epoch": 0.6123461243630789, + "grad_norm": 1.9401991101956468, + "learning_rate": 1.3797439643286227e-06, + "loss": 0.9074, + "step": 6790 + }, + { + "epoch": 0.6124363078865491, + "grad_norm": 1.5360169460981092, + "learning_rate": 1.3791885847889277e-06, + "loss": 0.9358, + "step": 6791 + }, + { + "epoch": 0.6125264914100194, + "grad_norm": 1.5772729216303916, + "learning_rate": 1.3786332582202853e-06, + "loss": 0.9874, + "step": 6792 + }, + { + "epoch": 0.6126166749334897, + "grad_norm": 1.8901350789517557, + "learning_rate": 1.3780779846700799e-06, + "loss": 0.954, + "step": 6793 + }, + { + "epoch": 0.6127068584569599, + "grad_norm": 1.3786140542954683, + "learning_rate": 1.3775227641856899e-06, + "loss": 0.964, + "step": 6794 + }, + { + "epoch": 0.6127970419804302, + "grad_norm": 1.4223225486572388, + "learning_rate": 1.37696759681449e-06, + "loss": 0.9502, + "step": 6795 + }, + { + "epoch": 0.6128872255039004, + "grad_norm": 0.724534269897933, + "learning_rate": 1.37641248260385e-06, + "loss": 0.8074, + "step": 6796 + }, + { + "epoch": 0.6129774090273707, + "grad_norm": 1.5492141454246373, + "learning_rate": 1.375857421601136e-06, + "loss": 1.0035, + "step": 6797 + }, + { + "epoch": 0.613067592550841, + "grad_norm": 1.5360961832091844, + "learning_rate": 1.3753024138537082e-06, + "loss": 0.9598, + "step": 6798 + }, + { + "epoch": 0.6131577760743112, + "grad_norm": 1.467755569481228, + "learning_rate": 1.3747474594089221e-06, + "loss": 1.0179, + "step": 6799 + }, + { + "epoch": 0.6132479595977814, + "grad_norm": 1.732030779228095, + "learning_rate": 1.374192558314131e-06, + "loss": 0.9361, + "step": 6800 + }, + { + "epoch": 0.6133381431212518, + "grad_norm": 1.6553020283528228, + "learning_rate": 1.373637710616681e-06, + "loss": 1.0194, + "step": 6801 + }, + { + "epoch": 0.613428326644722, + "grad_norm": 1.9929167964657977, + "learning_rate": 1.373082916363916e-06, + "loss": 1.0636, + "step": 6802 + }, + { + "epoch": 0.6135185101681923, + "grad_norm": 1.4107227402790707, + "learning_rate": 1.3725281756031732e-06, + "loss": 0.9946, + "step": 6803 + }, + { + "epoch": 0.6136086936916625, + "grad_norm": 1.4686020309907983, + "learning_rate": 1.3719734883817858e-06, + "loss": 0.9511, + "step": 6804 + }, + { + "epoch": 0.6136988772151328, + "grad_norm": 1.5220457321995888, + "learning_rate": 1.371418854747084e-06, + "loss": 1.0704, + "step": 6805 + }, + { + "epoch": 0.6137890607386031, + "grad_norm": 1.747999887527532, + "learning_rate": 1.3708642747463905e-06, + "loss": 0.9861, + "step": 6806 + }, + { + "epoch": 0.6138792442620733, + "grad_norm": 1.3377415037931522, + "learning_rate": 1.370309748427027e-06, + "loss": 0.9858, + "step": 6807 + }, + { + "epoch": 0.6139694277855435, + "grad_norm": 2.8819046038180742, + "learning_rate": 1.3697552758363079e-06, + "loss": 1.0305, + "step": 6808 + }, + { + "epoch": 0.6140596113090139, + "grad_norm": 1.8866467195586425, + "learning_rate": 1.3692008570215432e-06, + "loss": 0.9603, + "step": 6809 + }, + { + "epoch": 0.6141497948324841, + "grad_norm": 1.5214100184163286, + "learning_rate": 1.3686464920300398e-06, + "loss": 0.9041, + "step": 6810 + }, + { + "epoch": 0.6142399783559543, + "grad_norm": 1.9272945872493519, + "learning_rate": 1.3680921809090985e-06, + "loss": 0.9337, + "step": 6811 + }, + { + "epoch": 0.6143301618794246, + "grad_norm": 1.7260883077130091, + "learning_rate": 1.3675379237060175e-06, + "loss": 0.9923, + "step": 6812 + }, + { + "epoch": 0.6144203454028949, + "grad_norm": 1.3440674695098473, + "learning_rate": 1.366983720468088e-06, + "loss": 0.9479, + "step": 6813 + }, + { + "epoch": 0.6145105289263652, + "grad_norm": 1.5554493326673682, + "learning_rate": 1.3664295712425972e-06, + "loss": 0.9625, + "step": 6814 + }, + { + "epoch": 0.6146007124498354, + "grad_norm": 1.5861228496312365, + "learning_rate": 1.3658754760768296e-06, + "loss": 0.886, + "step": 6815 + }, + { + "epoch": 0.6146908959733057, + "grad_norm": 1.586396850810578, + "learning_rate": 1.3653214350180621e-06, + "loss": 0.9916, + "step": 6816 + }, + { + "epoch": 0.614781079496776, + "grad_norm": 1.4981519918353607, + "learning_rate": 1.3647674481135703e-06, + "loss": 1.0233, + "step": 6817 + }, + { + "epoch": 0.6148712630202462, + "grad_norm": 1.287015179210387, + "learning_rate": 1.3642135154106217e-06, + "loss": 0.8534, + "step": 6818 + }, + { + "epoch": 0.6149614465437164, + "grad_norm": 1.8787453437793162, + "learning_rate": 1.363659636956482e-06, + "loss": 0.9023, + "step": 6819 + }, + { + "epoch": 0.6150516300671868, + "grad_norm": 1.4467528718030294, + "learning_rate": 1.3631058127984112e-06, + "loss": 0.9233, + "step": 6820 + }, + { + "epoch": 0.615141813590657, + "grad_norm": 1.5833263229750318, + "learning_rate": 1.3625520429836632e-06, + "loss": 0.9329, + "step": 6821 + }, + { + "epoch": 0.6152319971141272, + "grad_norm": 1.4393688161686498, + "learning_rate": 1.361998327559491e-06, + "loss": 0.8631, + "step": 6822 + }, + { + "epoch": 0.6153221806375975, + "grad_norm": 0.7116829035050997, + "learning_rate": 1.3614446665731385e-06, + "loss": 0.7821, + "step": 6823 + }, + { + "epoch": 0.6154123641610678, + "grad_norm": 1.601119859515493, + "learning_rate": 1.3608910600718484e-06, + "loss": 0.9658, + "step": 6824 + }, + { + "epoch": 0.615502547684538, + "grad_norm": 1.5324373118536865, + "learning_rate": 1.360337508102857e-06, + "loss": 0.9581, + "step": 6825 + }, + { + "epoch": 0.6155927312080083, + "grad_norm": 1.8553179549126912, + "learning_rate": 1.3597840107133962e-06, + "loss": 1.051, + "step": 6826 + }, + { + "epoch": 0.6156829147314785, + "grad_norm": 1.7081617021019861, + "learning_rate": 1.3592305679506944e-06, + "loss": 0.9655, + "step": 6827 + }, + { + "epoch": 0.6157730982549489, + "grad_norm": 1.6927168704086601, + "learning_rate": 1.3586771798619726e-06, + "loss": 0.8632, + "step": 6828 + }, + { + "epoch": 0.6158632817784191, + "grad_norm": 1.3869527699900839, + "learning_rate": 1.358123846494451e-06, + "loss": 1.0147, + "step": 6829 + }, + { + "epoch": 0.6159534653018893, + "grad_norm": 2.510861168812707, + "learning_rate": 1.3575705678953422e-06, + "loss": 0.9657, + "step": 6830 + }, + { + "epoch": 0.6160436488253596, + "grad_norm": 1.659750783436012, + "learning_rate": 1.357017344111854e-06, + "loss": 0.9955, + "step": 6831 + }, + { + "epoch": 0.6161338323488299, + "grad_norm": 1.3502267399864327, + "learning_rate": 1.356464175191192e-06, + "loss": 0.9662, + "step": 6832 + }, + { + "epoch": 0.6162240158723001, + "grad_norm": 1.7845908173089395, + "learning_rate": 1.3559110611805542e-06, + "loss": 0.9817, + "step": 6833 + }, + { + "epoch": 0.6163141993957704, + "grad_norm": 1.507405675844022, + "learning_rate": 1.3553580021271372e-06, + "loss": 0.9933, + "step": 6834 + }, + { + "epoch": 0.6164043829192406, + "grad_norm": 1.3864799616287673, + "learning_rate": 1.3548049980781297e-06, + "loss": 0.8971, + "step": 6835 + }, + { + "epoch": 0.616494566442711, + "grad_norm": 1.6396545446394604, + "learning_rate": 1.3542520490807166e-06, + "loss": 0.9489, + "step": 6836 + }, + { + "epoch": 0.6165847499661812, + "grad_norm": 1.6492793387844609, + "learning_rate": 1.3536991551820802e-06, + "loss": 0.9443, + "step": 6837 + }, + { + "epoch": 0.6166749334896514, + "grad_norm": 1.2940051486017548, + "learning_rate": 1.3531463164293952e-06, + "loss": 0.9288, + "step": 6838 + }, + { + "epoch": 0.6167651170131218, + "grad_norm": 1.8881810683733435, + "learning_rate": 1.3525935328698332e-06, + "loss": 0.8592, + "step": 6839 + }, + { + "epoch": 0.616855300536592, + "grad_norm": 1.7798140911999794, + "learning_rate": 1.3520408045505605e-06, + "loss": 0.895, + "step": 6840 + }, + { + "epoch": 0.6169454840600622, + "grad_norm": 1.5807802882299526, + "learning_rate": 1.3514881315187396e-06, + "loss": 0.9192, + "step": 6841 + }, + { + "epoch": 0.6170356675835325, + "grad_norm": 1.4796634788940035, + "learning_rate": 1.3509355138215273e-06, + "loss": 1.0113, + "step": 6842 + }, + { + "epoch": 0.6171258511070028, + "grad_norm": 1.5013407437281105, + "learning_rate": 1.350382951506075e-06, + "loss": 1.0365, + "step": 6843 + }, + { + "epoch": 0.617216034630473, + "grad_norm": 0.7005153972504624, + "learning_rate": 1.349830444619532e-06, + "loss": 0.7692, + "step": 6844 + }, + { + "epoch": 0.6173062181539433, + "grad_norm": 1.5341197001520224, + "learning_rate": 1.3492779932090397e-06, + "loss": 0.9705, + "step": 6845 + }, + { + "epoch": 0.6173964016774135, + "grad_norm": 1.452192930503232, + "learning_rate": 1.3487255973217377e-06, + "loss": 0.9823, + "step": 6846 + }, + { + "epoch": 0.6174865852008838, + "grad_norm": 1.7933456922244713, + "learning_rate": 1.3481732570047592e-06, + "loss": 1.0237, + "step": 6847 + }, + { + "epoch": 0.6175767687243541, + "grad_norm": 2.1028724189982366, + "learning_rate": 1.3476209723052314e-06, + "loss": 0.888, + "step": 6848 + }, + { + "epoch": 0.6176669522478243, + "grad_norm": 2.083224242851018, + "learning_rate": 1.3470687432702806e-06, + "loss": 0.9756, + "step": 6849 + }, + { + "epoch": 0.6177571357712945, + "grad_norm": 1.6414594435572083, + "learning_rate": 1.346516569947024e-06, + "loss": 0.886, + "step": 6850 + }, + { + "epoch": 0.6178473192947649, + "grad_norm": 1.7410775656120376, + "learning_rate": 1.3459644523825774e-06, + "loss": 0.9395, + "step": 6851 + }, + { + "epoch": 0.6179375028182351, + "grad_norm": 1.796772033394953, + "learning_rate": 1.34541239062405e-06, + "loss": 0.9964, + "step": 6852 + }, + { + "epoch": 0.6180276863417054, + "grad_norm": 1.8400098073739444, + "learning_rate": 1.3448603847185464e-06, + "loss": 0.9574, + "step": 6853 + }, + { + "epoch": 0.6181178698651756, + "grad_norm": 1.4136264191840562, + "learning_rate": 1.344308434713168e-06, + "loss": 0.897, + "step": 6854 + }, + { + "epoch": 0.6182080533886459, + "grad_norm": 2.0090064866230417, + "learning_rate": 1.3437565406550083e-06, + "loss": 0.8909, + "step": 6855 + }, + { + "epoch": 0.6182982369121162, + "grad_norm": 1.8769982815807282, + "learning_rate": 1.34320470259116e-06, + "loss": 0.9445, + "step": 6856 + }, + { + "epoch": 0.6183884204355864, + "grad_norm": 1.6204640831667247, + "learning_rate": 1.3426529205687078e-06, + "loss": 0.8598, + "step": 6857 + }, + { + "epoch": 0.6184786039590566, + "grad_norm": 1.5720988878251891, + "learning_rate": 1.3421011946347323e-06, + "loss": 1.0103, + "step": 6858 + }, + { + "epoch": 0.618568787482527, + "grad_norm": 1.4634052181328643, + "learning_rate": 1.3415495248363113e-06, + "loss": 0.9154, + "step": 6859 + }, + { + "epoch": 0.6186589710059972, + "grad_norm": 1.7081014041548033, + "learning_rate": 1.3409979112205148e-06, + "loss": 0.9758, + "step": 6860 + }, + { + "epoch": 0.6187491545294674, + "grad_norm": 1.7913364135598344, + "learning_rate": 1.3404463538344107e-06, + "loss": 0.9806, + "step": 6861 + }, + { + "epoch": 0.6188393380529378, + "grad_norm": 1.9070434794957818, + "learning_rate": 1.3398948527250602e-06, + "loss": 1.0428, + "step": 6862 + }, + { + "epoch": 0.618929521576408, + "grad_norm": 0.7943376273331162, + "learning_rate": 1.3393434079395207e-06, + "loss": 0.7925, + "step": 6863 + }, + { + "epoch": 0.6190197050998782, + "grad_norm": 1.4408754441369898, + "learning_rate": 1.3387920195248449e-06, + "loss": 0.9403, + "step": 6864 + }, + { + "epoch": 0.6191098886233485, + "grad_norm": 1.41387141501507, + "learning_rate": 1.3382406875280791e-06, + "loss": 0.9698, + "step": 6865 + }, + { + "epoch": 0.6192000721468188, + "grad_norm": 1.5491473529193822, + "learning_rate": 1.3376894119962672e-06, + "loss": 0.9778, + "step": 6866 + }, + { + "epoch": 0.6192902556702891, + "grad_norm": 2.890824099720511, + "learning_rate": 1.3371381929764464e-06, + "loss": 0.9855, + "step": 6867 + }, + { + "epoch": 0.6193804391937593, + "grad_norm": 0.6569485579445438, + "learning_rate": 1.3365870305156506e-06, + "loss": 0.7713, + "step": 6868 + }, + { + "epoch": 0.6194706227172295, + "grad_norm": 2.229280225291557, + "learning_rate": 1.3360359246609073e-06, + "loss": 0.9606, + "step": 6869 + }, + { + "epoch": 0.6195608062406999, + "grad_norm": 1.5665210862528876, + "learning_rate": 1.3354848754592387e-06, + "loss": 0.9519, + "step": 6870 + }, + { + "epoch": 0.6196509897641701, + "grad_norm": 1.664731961117198, + "learning_rate": 1.334933882957666e-06, + "loss": 1.0421, + "step": 6871 + }, + { + "epoch": 0.6197411732876403, + "grad_norm": 1.4328489109683575, + "learning_rate": 1.3343829472032004e-06, + "loss": 0.9181, + "step": 6872 + }, + { + "epoch": 0.6198313568111106, + "grad_norm": 1.497295882717722, + "learning_rate": 1.3338320682428527e-06, + "loss": 0.9453, + "step": 6873 + }, + { + "epoch": 0.6199215403345809, + "grad_norm": 1.4926837675101208, + "learning_rate": 1.3332812461236263e-06, + "loss": 0.9158, + "step": 6874 + }, + { + "epoch": 0.6200117238580511, + "grad_norm": 1.40468586718981, + "learning_rate": 1.3327304808925192e-06, + "loss": 0.9178, + "step": 6875 + }, + { + "epoch": 0.6201019073815214, + "grad_norm": 1.825973418360888, + "learning_rate": 1.332179772596528e-06, + "loss": 0.8133, + "step": 6876 + }, + { + "epoch": 0.6201920909049916, + "grad_norm": 1.5810881641182581, + "learning_rate": 1.3316291212826402e-06, + "loss": 0.8787, + "step": 6877 + }, + { + "epoch": 0.620282274428462, + "grad_norm": 1.7212809567794651, + "learning_rate": 1.3310785269978413e-06, + "loss": 1.0044, + "step": 6878 + }, + { + "epoch": 0.6203724579519322, + "grad_norm": 1.6446263519116868, + "learning_rate": 1.3305279897891111e-06, + "loss": 1.0026, + "step": 6879 + }, + { + "epoch": 0.6204626414754024, + "grad_norm": 1.835786074623017, + "learning_rate": 1.329977509703424e-06, + "loss": 0.8196, + "step": 6880 + }, + { + "epoch": 0.6205528249988727, + "grad_norm": 1.7132072547557702, + "learning_rate": 1.32942708678775e-06, + "loss": 1.043, + "step": 6881 + }, + { + "epoch": 0.620643008522343, + "grad_norm": 1.523555027267496, + "learning_rate": 1.3288767210890548e-06, + "loss": 0.9701, + "step": 6882 + }, + { + "epoch": 0.6207331920458132, + "grad_norm": 1.607305436766555, + "learning_rate": 1.3283264126542986e-06, + "loss": 0.9159, + "step": 6883 + }, + { + "epoch": 0.6208233755692835, + "grad_norm": 1.6536050946681764, + "learning_rate": 1.3277761615304356e-06, + "loss": 0.9587, + "step": 6884 + }, + { + "epoch": 0.6209135590927538, + "grad_norm": 1.3390089275811756, + "learning_rate": 1.3272259677644178e-06, + "loss": 1.0528, + "step": 6885 + }, + { + "epoch": 0.621003742616224, + "grad_norm": 1.6972996891584895, + "learning_rate": 1.32667583140319e-06, + "loss": 0.8272, + "step": 6886 + }, + { + "epoch": 0.6210939261396943, + "grad_norm": 1.992713589486028, + "learning_rate": 1.3261257524936924e-06, + "loss": 0.9972, + "step": 6887 + }, + { + "epoch": 0.6211841096631645, + "grad_norm": 1.425807921277721, + "learning_rate": 1.3255757310828619e-06, + "loss": 0.9845, + "step": 6888 + }, + { + "epoch": 0.6212742931866349, + "grad_norm": 1.8173635933755066, + "learning_rate": 1.3250257672176282e-06, + "loss": 0.9714, + "step": 6889 + }, + { + "epoch": 0.6213644767101051, + "grad_norm": 1.4965160759834422, + "learning_rate": 1.3244758609449183e-06, + "loss": 0.8579, + "step": 6890 + }, + { + "epoch": 0.6214546602335753, + "grad_norm": 1.6790550993148003, + "learning_rate": 1.323926012311653e-06, + "loss": 0.9722, + "step": 6891 + }, + { + "epoch": 0.6215448437570456, + "grad_norm": 1.512949203747878, + "learning_rate": 1.3233762213647476e-06, + "loss": 0.9799, + "step": 6892 + }, + { + "epoch": 0.6216350272805159, + "grad_norm": 1.8438321111466356, + "learning_rate": 1.3228264881511137e-06, + "loss": 0.9277, + "step": 6893 + }, + { + "epoch": 0.6217252108039861, + "grad_norm": 2.011120042318578, + "learning_rate": 1.322276812717658e-06, + "loss": 1.0385, + "step": 6894 + }, + { + "epoch": 0.6218153943274564, + "grad_norm": 1.4025262485434609, + "learning_rate": 1.3217271951112818e-06, + "loss": 0.8637, + "step": 6895 + }, + { + "epoch": 0.6219055778509266, + "grad_norm": 1.6381348376686418, + "learning_rate": 1.321177635378881e-06, + "loss": 0.9378, + "step": 6896 + }, + { + "epoch": 0.6219957613743969, + "grad_norm": 1.7193535871985437, + "learning_rate": 1.3206281335673475e-06, + "loss": 0.8495, + "step": 6897 + }, + { + "epoch": 0.6220859448978672, + "grad_norm": 1.8526622226538705, + "learning_rate": 1.3200786897235677e-06, + "loss": 0.9297, + "step": 6898 + }, + { + "epoch": 0.6221761284213374, + "grad_norm": 1.6025760327897682, + "learning_rate": 1.3195293038944227e-06, + "loss": 0.9761, + "step": 6899 + }, + { + "epoch": 0.6222663119448076, + "grad_norm": 1.2960193063098562, + "learning_rate": 1.3189799761267902e-06, + "loss": 0.9477, + "step": 6900 + }, + { + "epoch": 0.622356495468278, + "grad_norm": 1.69392563306931, + "learning_rate": 1.3184307064675412e-06, + "loss": 0.947, + "step": 6901 + }, + { + "epoch": 0.6224466789917482, + "grad_norm": 0.7534333559394577, + "learning_rate": 1.3178814949635416e-06, + "loss": 0.8256, + "step": 6902 + }, + { + "epoch": 0.6225368625152184, + "grad_norm": 1.571976268987159, + "learning_rate": 1.3173323416616549e-06, + "loss": 0.9033, + "step": 6903 + }, + { + "epoch": 0.6226270460386887, + "grad_norm": 1.6482173537648075, + "learning_rate": 1.3167832466087361e-06, + "loss": 1.0058, + "step": 6904 + }, + { + "epoch": 0.622717229562159, + "grad_norm": 1.4292646939808613, + "learning_rate": 1.3162342098516388e-06, + "loss": 0.9028, + "step": 6905 + }, + { + "epoch": 0.6228074130856293, + "grad_norm": 1.3990831216737032, + "learning_rate": 1.3156852314372086e-06, + "loss": 0.8738, + "step": 6906 + }, + { + "epoch": 0.6228975966090995, + "grad_norm": 0.71363490718313, + "learning_rate": 1.3151363114122882e-06, + "loss": 0.7609, + "step": 6907 + }, + { + "epoch": 0.6229877801325698, + "grad_norm": 1.5561500455936905, + "learning_rate": 1.3145874498237133e-06, + "loss": 0.9668, + "step": 6908 + }, + { + "epoch": 0.6230779636560401, + "grad_norm": 1.5235545578024392, + "learning_rate": 1.3140386467183166e-06, + "loss": 1.0023, + "step": 6909 + }, + { + "epoch": 0.6231681471795103, + "grad_norm": 1.4055051738436155, + "learning_rate": 1.3134899021429258e-06, + "loss": 0.9945, + "step": 6910 + }, + { + "epoch": 0.6232583307029805, + "grad_norm": 1.7355270683399888, + "learning_rate": 1.3129412161443604e-06, + "loss": 0.9384, + "step": 6911 + }, + { + "epoch": 0.6233485142264509, + "grad_norm": 1.5811003783835669, + "learning_rate": 1.3123925887694402e-06, + "loss": 0.9633, + "step": 6912 + }, + { + "epoch": 0.6234386977499211, + "grad_norm": 1.5483078148788334, + "learning_rate": 1.3118440200649756e-06, + "loss": 0.962, + "step": 6913 + }, + { + "epoch": 0.6235288812733913, + "grad_norm": 1.5041512426006454, + "learning_rate": 1.3112955100777727e-06, + "loss": 0.918, + "step": 6914 + }, + { + "epoch": 0.6236190647968616, + "grad_norm": 1.3746183906209657, + "learning_rate": 1.3107470588546353e-06, + "loss": 0.9886, + "step": 6915 + }, + { + "epoch": 0.6237092483203319, + "grad_norm": 1.5316284646191425, + "learning_rate": 1.3101986664423583e-06, + "loss": 0.9277, + "step": 6916 + }, + { + "epoch": 0.6237994318438022, + "grad_norm": 1.46178994833658, + "learning_rate": 1.3096503328877358e-06, + "loss": 0.9286, + "step": 6917 + }, + { + "epoch": 0.6238896153672724, + "grad_norm": 1.6900345171275768, + "learning_rate": 1.309102058237553e-06, + "loss": 0.9752, + "step": 6918 + }, + { + "epoch": 0.6239797988907426, + "grad_norm": 1.2525790787667452, + "learning_rate": 1.3085538425385917e-06, + "loss": 1.005, + "step": 6919 + }, + { + "epoch": 0.624069982414213, + "grad_norm": 2.0450015047995853, + "learning_rate": 1.3080056858376298e-06, + "loss": 0.9873, + "step": 6920 + }, + { + "epoch": 0.6241601659376832, + "grad_norm": 1.7213108751756037, + "learning_rate": 1.3074575881814383e-06, + "loss": 0.9414, + "step": 6921 + }, + { + "epoch": 0.6242503494611534, + "grad_norm": 1.6076115696752924, + "learning_rate": 1.3069095496167841e-06, + "loss": 0.8819, + "step": 6922 + }, + { + "epoch": 0.6243405329846237, + "grad_norm": 1.5213192812236929, + "learning_rate": 1.3063615701904285e-06, + "loss": 0.8652, + "step": 6923 + }, + { + "epoch": 0.624430716508094, + "grad_norm": 1.718022973010261, + "learning_rate": 1.3058136499491283e-06, + "loss": 0.9843, + "step": 6924 + }, + { + "epoch": 0.6245209000315642, + "grad_norm": 2.0148382499652295, + "learning_rate": 1.3052657889396357e-06, + "loss": 0.9326, + "step": 6925 + }, + { + "epoch": 0.6246110835550345, + "grad_norm": 1.4264369345281394, + "learning_rate": 1.304717987208696e-06, + "loss": 0.9389, + "step": 6926 + }, + { + "epoch": 0.6247012670785047, + "grad_norm": 1.3823663023994037, + "learning_rate": 1.304170244803052e-06, + "loss": 0.9427, + "step": 6927 + }, + { + "epoch": 0.624791450601975, + "grad_norm": 2.0554592293287204, + "learning_rate": 1.3036225617694387e-06, + "loss": 0.924, + "step": 6928 + }, + { + "epoch": 0.6248816341254453, + "grad_norm": 1.490638848006504, + "learning_rate": 1.3030749381545892e-06, + "loss": 0.9831, + "step": 6929 + }, + { + "epoch": 0.6249718176489155, + "grad_norm": 1.709813205823011, + "learning_rate": 1.3025273740052285e-06, + "loss": 0.8903, + "step": 6930 + }, + { + "epoch": 0.6250620011723858, + "grad_norm": 1.3543231531349416, + "learning_rate": 1.3019798693680774e-06, + "loss": 1.0028, + "step": 6931 + }, + { + "epoch": 0.6251521846958561, + "grad_norm": 1.429561171656852, + "learning_rate": 1.3014324242898536e-06, + "loss": 1.069, + "step": 6932 + }, + { + "epoch": 0.6252423682193263, + "grad_norm": 2.2205490555407112, + "learning_rate": 1.3008850388172668e-06, + "loss": 1.0157, + "step": 6933 + }, + { + "epoch": 0.6253325517427966, + "grad_norm": 1.4475226746728413, + "learning_rate": 1.3003377129970233e-06, + "loss": 0.9134, + "step": 6934 + }, + { + "epoch": 0.6254227352662669, + "grad_norm": 1.4373396908478513, + "learning_rate": 1.2997904468758243e-06, + "loss": 0.8754, + "step": 6935 + }, + { + "epoch": 0.6255129187897371, + "grad_norm": 1.5117718985428579, + "learning_rate": 1.299243240500365e-06, + "loss": 0.9724, + "step": 6936 + }, + { + "epoch": 0.6256031023132074, + "grad_norm": 1.6898516763685534, + "learning_rate": 1.2986960939173368e-06, + "loss": 0.9687, + "step": 6937 + }, + { + "epoch": 0.6256932858366776, + "grad_norm": 1.6861847591009536, + "learning_rate": 1.298149007173424e-06, + "loss": 0.9987, + "step": 6938 + }, + { + "epoch": 0.625783469360148, + "grad_norm": 1.9821802452420128, + "learning_rate": 1.2976019803153087e-06, + "loss": 0.9868, + "step": 6939 + }, + { + "epoch": 0.6258736528836182, + "grad_norm": 1.5858672929830444, + "learning_rate": 1.2970550133896652e-06, + "loss": 0.9007, + "step": 6940 + }, + { + "epoch": 0.6259638364070884, + "grad_norm": 1.5768660254115339, + "learning_rate": 1.2965081064431634e-06, + "loss": 1.0344, + "step": 6941 + }, + { + "epoch": 0.6260540199305586, + "grad_norm": 1.6476259086785237, + "learning_rate": 1.2959612595224698e-06, + "loss": 0.9764, + "step": 6942 + }, + { + "epoch": 0.626144203454029, + "grad_norm": 1.9270484577494165, + "learning_rate": 1.2954144726742424e-06, + "loss": 0.9654, + "step": 6943 + }, + { + "epoch": 0.6262343869774992, + "grad_norm": 1.4421031063428615, + "learning_rate": 1.2948677459451385e-06, + "loss": 0.9524, + "step": 6944 + }, + { + "epoch": 0.6263245705009695, + "grad_norm": 1.3852566600364769, + "learning_rate": 1.2943210793818064e-06, + "loss": 1.0006, + "step": 6945 + }, + { + "epoch": 0.6264147540244397, + "grad_norm": 1.6238809547153634, + "learning_rate": 1.2937744730308899e-06, + "loss": 0.9319, + "step": 6946 + }, + { + "epoch": 0.62650493754791, + "grad_norm": 2.1161889681096016, + "learning_rate": 1.2932279269390305e-06, + "loss": 0.9558, + "step": 6947 + }, + { + "epoch": 0.6265951210713803, + "grad_norm": 1.4394074928369591, + "learning_rate": 1.292681441152861e-06, + "loss": 0.8602, + "step": 6948 + }, + { + "epoch": 0.6266853045948505, + "grad_norm": 1.5611135053258043, + "learning_rate": 1.292135015719011e-06, + "loss": 0.9636, + "step": 6949 + }, + { + "epoch": 0.6267754881183207, + "grad_norm": 1.5688453766906876, + "learning_rate": 1.2915886506841046e-06, + "loss": 0.9103, + "step": 6950 + }, + { + "epoch": 0.6268656716417911, + "grad_norm": 1.4863885974988702, + "learning_rate": 1.2910423460947613e-06, + "loss": 0.8867, + "step": 6951 + }, + { + "epoch": 0.6269558551652613, + "grad_norm": 1.4450580269801354, + "learning_rate": 1.290496101997594e-06, + "loss": 0.923, + "step": 6952 + }, + { + "epoch": 0.6270460386887315, + "grad_norm": 0.7684258979582375, + "learning_rate": 1.2899499184392105e-06, + "loss": 0.7833, + "step": 6953 + }, + { + "epoch": 0.6271362222122018, + "grad_norm": 1.67538454566541, + "learning_rate": 1.289403795466216e-06, + "loss": 0.9336, + "step": 6954 + }, + { + "epoch": 0.6272264057356721, + "grad_norm": 1.6804894240777761, + "learning_rate": 1.288857733125207e-06, + "loss": 0.9709, + "step": 6955 + }, + { + "epoch": 0.6273165892591424, + "grad_norm": 1.742598070425145, + "learning_rate": 1.2883117314627785e-06, + "loss": 0.9491, + "step": 6956 + }, + { + "epoch": 0.6274067727826126, + "grad_norm": 1.902717256008913, + "learning_rate": 1.2877657905255168e-06, + "loss": 0.8909, + "step": 6957 + }, + { + "epoch": 0.6274969563060829, + "grad_norm": 1.4250467158908557, + "learning_rate": 1.2872199103600046e-06, + "loss": 1.011, + "step": 6958 + }, + { + "epoch": 0.6275871398295532, + "grad_norm": 1.5115448607747781, + "learning_rate": 1.286674091012821e-06, + "loss": 1.0214, + "step": 6959 + }, + { + "epoch": 0.6276773233530234, + "grad_norm": 1.5114487202306002, + "learning_rate": 1.2861283325305356e-06, + "loss": 1.0109, + "step": 6960 + }, + { + "epoch": 0.6277675068764936, + "grad_norm": 1.4175450556354354, + "learning_rate": 1.2855826349597185e-06, + "loss": 0.9516, + "step": 6961 + }, + { + "epoch": 0.627857690399964, + "grad_norm": 1.721345917820944, + "learning_rate": 1.2850369983469302e-06, + "loss": 0.8099, + "step": 6962 + }, + { + "epoch": 0.6279478739234342, + "grad_norm": 1.4873366674941266, + "learning_rate": 1.2844914227387266e-06, + "loss": 0.9807, + "step": 6963 + }, + { + "epoch": 0.6280380574469044, + "grad_norm": 1.3963227291853455, + "learning_rate": 1.2839459081816606e-06, + "loss": 0.9751, + "step": 6964 + }, + { + "epoch": 0.6281282409703747, + "grad_norm": 1.5571199490423229, + "learning_rate": 1.283400454722278e-06, + "loss": 1.0198, + "step": 6965 + }, + { + "epoch": 0.628218424493845, + "grad_norm": 1.4538043495311528, + "learning_rate": 1.28285506240712e-06, + "loss": 1.0173, + "step": 6966 + }, + { + "epoch": 0.6283086080173153, + "grad_norm": 1.3722529279608475, + "learning_rate": 1.2823097312827225e-06, + "loss": 0.9787, + "step": 6967 + }, + { + "epoch": 0.6283987915407855, + "grad_norm": 1.4963049677262523, + "learning_rate": 1.2817644613956153e-06, + "loss": 0.9645, + "step": 6968 + }, + { + "epoch": 0.6284889750642557, + "grad_norm": 1.7908991419062976, + "learning_rate": 1.2812192527923253e-06, + "loss": 1.0162, + "step": 6969 + }, + { + "epoch": 0.6285791585877261, + "grad_norm": 1.524892571839687, + "learning_rate": 1.2806741055193712e-06, + "loss": 0.9861, + "step": 6970 + }, + { + "epoch": 0.6286693421111963, + "grad_norm": 1.827333001622575, + "learning_rate": 1.2801290196232695e-06, + "loss": 0.9138, + "step": 6971 + }, + { + "epoch": 0.6287595256346665, + "grad_norm": 1.8235031129200134, + "learning_rate": 1.2795839951505282e-06, + "loss": 0.9444, + "step": 6972 + }, + { + "epoch": 0.6288497091581368, + "grad_norm": 4.982434316490249, + "learning_rate": 1.2790390321476538e-06, + "loss": 0.9729, + "step": 6973 + }, + { + "epoch": 0.6289398926816071, + "grad_norm": 1.6763312586490042, + "learning_rate": 1.2784941306611446e-06, + "loss": 0.9256, + "step": 6974 + }, + { + "epoch": 0.6290300762050773, + "grad_norm": 1.7279548381890786, + "learning_rate": 1.2779492907374935e-06, + "loss": 1.0134, + "step": 6975 + }, + { + "epoch": 0.6291202597285476, + "grad_norm": 1.5779612994506826, + "learning_rate": 1.2774045124231911e-06, + "loss": 0.9443, + "step": 6976 + }, + { + "epoch": 0.6292104432520178, + "grad_norm": 1.4028854386506284, + "learning_rate": 1.2768597957647197e-06, + "loss": 0.8353, + "step": 6977 + }, + { + "epoch": 0.6293006267754881, + "grad_norm": 1.7057015763786239, + "learning_rate": 1.2763151408085582e-06, + "loss": 0.9644, + "step": 6978 + }, + { + "epoch": 0.6293908102989584, + "grad_norm": 1.6286302511089712, + "learning_rate": 1.2757705476011788e-06, + "loss": 0.9186, + "step": 6979 + }, + { + "epoch": 0.6294809938224286, + "grad_norm": 1.781005842641578, + "learning_rate": 1.27522601618905e-06, + "loss": 0.9019, + "step": 6980 + }, + { + "epoch": 0.629571177345899, + "grad_norm": 0.602803226476741, + "learning_rate": 1.2746815466186337e-06, + "loss": 0.8061, + "step": 6981 + }, + { + "epoch": 0.6296613608693692, + "grad_norm": 1.6436872695150437, + "learning_rate": 1.274137138936387e-06, + "loss": 0.9595, + "step": 6982 + }, + { + "epoch": 0.6297515443928394, + "grad_norm": 1.486063990138054, + "learning_rate": 1.2735927931887625e-06, + "loss": 0.9565, + "step": 6983 + }, + { + "epoch": 0.6298417279163097, + "grad_norm": 0.7131343477189805, + "learning_rate": 1.2730485094222061e-06, + "loss": 0.788, + "step": 6984 + }, + { + "epoch": 0.62993191143978, + "grad_norm": 1.5320395166464886, + "learning_rate": 1.2725042876831586e-06, + "loss": 0.9921, + "step": 6985 + }, + { + "epoch": 0.6300220949632502, + "grad_norm": 1.8446477870165734, + "learning_rate": 1.2719601280180573e-06, + "loss": 0.8465, + "step": 6986 + }, + { + "epoch": 0.6301122784867205, + "grad_norm": 1.5682070457429629, + "learning_rate": 1.2714160304733317e-06, + "loss": 1.0204, + "step": 6987 + }, + { + "epoch": 0.6302024620101907, + "grad_norm": 1.6272759271883754, + "learning_rate": 1.2708719950954082e-06, + "loss": 0.9652, + "step": 6988 + }, + { + "epoch": 0.630292645533661, + "grad_norm": 1.8790489348799466, + "learning_rate": 1.2703280219307065e-06, + "loss": 0.9709, + "step": 6989 + }, + { + "epoch": 0.6303828290571313, + "grad_norm": 1.5536842367873858, + "learning_rate": 1.2697841110256411e-06, + "loss": 0.8848, + "step": 6990 + }, + { + "epoch": 0.6304730125806015, + "grad_norm": 1.4506509700818842, + "learning_rate": 1.2692402624266221e-06, + "loss": 0.9982, + "step": 6991 + }, + { + "epoch": 0.6305631961040717, + "grad_norm": 1.762940856231482, + "learning_rate": 1.2686964761800529e-06, + "loss": 0.867, + "step": 6992 + }, + { + "epoch": 0.6306533796275421, + "grad_norm": 0.696390987175065, + "learning_rate": 1.268152752332333e-06, + "loss": 0.7854, + "step": 6993 + }, + { + "epoch": 0.6307435631510123, + "grad_norm": 1.7684080757191414, + "learning_rate": 1.2676090909298549e-06, + "loss": 0.8652, + "step": 6994 + }, + { + "epoch": 0.6308337466744826, + "grad_norm": 1.393417127057805, + "learning_rate": 1.2670654920190086e-06, + "loss": 1.0044, + "step": 6995 + }, + { + "epoch": 0.6309239301979528, + "grad_norm": 1.4793681782032144, + "learning_rate": 1.2665219556461754e-06, + "loss": 1.0027, + "step": 6996 + }, + { + "epoch": 0.6310141137214231, + "grad_norm": 1.7206057153986642, + "learning_rate": 1.2659784818577329e-06, + "loss": 1.0534, + "step": 6997 + }, + { + "epoch": 0.6311042972448934, + "grad_norm": 1.4996347777472447, + "learning_rate": 1.2654350707000542e-06, + "loss": 0.8372, + "step": 6998 + }, + { + "epoch": 0.6311944807683636, + "grad_norm": 1.5493879617300437, + "learning_rate": 1.264891722219505e-06, + "loss": 0.941, + "step": 6999 + }, + { + "epoch": 0.6312846642918338, + "grad_norm": 1.6519028817233028, + "learning_rate": 1.2643484364624483e-06, + "loss": 0.9435, + "step": 7000 + }, + { + "epoch": 0.6313748478153042, + "grad_norm": 1.6457670737662544, + "learning_rate": 1.2638052134752393e-06, + "loss": 1.0044, + "step": 7001 + }, + { + "epoch": 0.6314650313387744, + "grad_norm": 1.774042505256095, + "learning_rate": 1.2632620533042277e-06, + "loss": 1.0262, + "step": 7002 + }, + { + "epoch": 0.6315552148622446, + "grad_norm": 1.9131620819376272, + "learning_rate": 1.2627189559957612e-06, + "loss": 0.9225, + "step": 7003 + }, + { + "epoch": 0.631645398385715, + "grad_norm": 2.0128744120641695, + "learning_rate": 1.262175921596178e-06, + "loss": 1.0145, + "step": 7004 + }, + { + "epoch": 0.6317355819091852, + "grad_norm": 1.7566838010402321, + "learning_rate": 1.2616329501518137e-06, + "loss": 0.8269, + "step": 7005 + }, + { + "epoch": 0.6318257654326555, + "grad_norm": 1.7582291172956528, + "learning_rate": 1.2610900417089978e-06, + "loss": 0.877, + "step": 7006 + }, + { + "epoch": 0.6319159489561257, + "grad_norm": 1.5581263556017375, + "learning_rate": 1.2605471963140535e-06, + "loss": 0.9605, + "step": 7007 + }, + { + "epoch": 0.632006132479596, + "grad_norm": 1.5089958962700054, + "learning_rate": 1.2600044140133e-06, + "loss": 1.0265, + "step": 7008 + }, + { + "epoch": 0.6320963160030663, + "grad_norm": 1.6200645028918674, + "learning_rate": 1.2594616948530493e-06, + "loss": 0.9054, + "step": 7009 + }, + { + "epoch": 0.6321864995265365, + "grad_norm": 1.6495003550639815, + "learning_rate": 1.258919038879611e-06, + "loss": 0.9798, + "step": 7010 + }, + { + "epoch": 0.6322766830500067, + "grad_norm": 1.9862323989232349, + "learning_rate": 1.2583764461392859e-06, + "loss": 0.9531, + "step": 7011 + }, + { + "epoch": 0.6323668665734771, + "grad_norm": 1.5586854530069256, + "learning_rate": 1.2578339166783724e-06, + "loss": 0.919, + "step": 7012 + }, + { + "epoch": 0.6324570500969473, + "grad_norm": 1.3371510487201907, + "learning_rate": 1.2572914505431613e-06, + "loss": 0.888, + "step": 7013 + }, + { + "epoch": 0.6325472336204175, + "grad_norm": 1.356839544956721, + "learning_rate": 1.2567490477799383e-06, + "loss": 0.8548, + "step": 7014 + }, + { + "epoch": 0.6326374171438878, + "grad_norm": 1.5645409796307572, + "learning_rate": 1.2562067084349852e-06, + "loss": 1.046, + "step": 7015 + }, + { + "epoch": 0.6327276006673581, + "grad_norm": 1.4930883753912612, + "learning_rate": 1.2556644325545764e-06, + "loss": 0.9365, + "step": 7016 + }, + { + "epoch": 0.6328177841908283, + "grad_norm": 1.6629040047929688, + "learning_rate": 1.255122220184983e-06, + "loss": 0.9315, + "step": 7017 + }, + { + "epoch": 0.6329079677142986, + "grad_norm": 1.7545082741273474, + "learning_rate": 1.2545800713724694e-06, + "loss": 0.8674, + "step": 7018 + }, + { + "epoch": 0.6329981512377688, + "grad_norm": 1.5704990864447141, + "learning_rate": 1.254037986163294e-06, + "loss": 0.8221, + "step": 7019 + }, + { + "epoch": 0.6330883347612392, + "grad_norm": 1.4826528095402831, + "learning_rate": 1.2534959646037104e-06, + "loss": 0.9149, + "step": 7020 + }, + { + "epoch": 0.6331785182847094, + "grad_norm": 1.4683985999962916, + "learning_rate": 1.2529540067399675e-06, + "loss": 0.9566, + "step": 7021 + }, + { + "epoch": 0.6332687018081796, + "grad_norm": 1.5849964375260326, + "learning_rate": 1.252412112618308e-06, + "loss": 0.8896, + "step": 7022 + }, + { + "epoch": 0.6333588853316499, + "grad_norm": 1.6534824641070425, + "learning_rate": 1.2518702822849696e-06, + "loss": 0.8336, + "step": 7023 + }, + { + "epoch": 0.6334490688551202, + "grad_norm": 1.8985261621689444, + "learning_rate": 1.2513285157861831e-06, + "loss": 0.901, + "step": 7024 + }, + { + "epoch": 0.6335392523785904, + "grad_norm": 1.4924509504938939, + "learning_rate": 1.2507868131681764e-06, + "loss": 0.9495, + "step": 7025 + }, + { + "epoch": 0.6336294359020607, + "grad_norm": 1.4732160642538854, + "learning_rate": 1.250245174477169e-06, + "loss": 0.9333, + "step": 7026 + }, + { + "epoch": 0.6337196194255309, + "grad_norm": 0.7095859802011555, + "learning_rate": 1.2497035997593783e-06, + "loss": 0.7215, + "step": 7027 + }, + { + "epoch": 0.6338098029490012, + "grad_norm": 1.4806652894824506, + "learning_rate": 1.2491620890610135e-06, + "loss": 0.9462, + "step": 7028 + }, + { + "epoch": 0.6338999864724715, + "grad_norm": 1.6848002843478, + "learning_rate": 1.2486206424282788e-06, + "loss": 0.9392, + "step": 7029 + }, + { + "epoch": 0.6339901699959417, + "grad_norm": 1.658240292046108, + "learning_rate": 1.2480792599073743e-06, + "loss": 1.0047, + "step": 7030 + }, + { + "epoch": 0.634080353519412, + "grad_norm": 1.5679181572127052, + "learning_rate": 1.247537941544493e-06, + "loss": 0.9953, + "step": 7031 + }, + { + "epoch": 0.6341705370428823, + "grad_norm": 1.910955281459521, + "learning_rate": 1.2469966873858242e-06, + "loss": 0.9858, + "step": 7032 + }, + { + "epoch": 0.6342607205663525, + "grad_norm": 1.6645900982403783, + "learning_rate": 1.2464554974775496e-06, + "loss": 0.8771, + "step": 7033 + }, + { + "epoch": 0.6343509040898228, + "grad_norm": 1.5439687844120327, + "learning_rate": 1.2459143718658474e-06, + "loss": 0.9519, + "step": 7034 + }, + { + "epoch": 0.6344410876132931, + "grad_norm": 1.7527836730839081, + "learning_rate": 1.2453733105968886e-06, + "loss": 0.8833, + "step": 7035 + }, + { + "epoch": 0.6345312711367633, + "grad_norm": 0.6747903780931271, + "learning_rate": 1.2448323137168399e-06, + "loss": 0.786, + "step": 7036 + }, + { + "epoch": 0.6346214546602336, + "grad_norm": 1.375376476419834, + "learning_rate": 1.2442913812718625e-06, + "loss": 0.8724, + "step": 7037 + }, + { + "epoch": 0.6347116381837038, + "grad_norm": 1.4371210510990662, + "learning_rate": 1.2437505133081108e-06, + "loss": 0.9339, + "step": 7038 + }, + { + "epoch": 0.6348018217071741, + "grad_norm": 1.5996425378363726, + "learning_rate": 1.2432097098717358e-06, + "loss": 0.9677, + "step": 7039 + }, + { + "epoch": 0.6348920052306444, + "grad_norm": 1.8477053484532535, + "learning_rate": 1.2426689710088813e-06, + "loss": 0.9568, + "step": 7040 + }, + { + "epoch": 0.6349821887541146, + "grad_norm": 1.5487001229075852, + "learning_rate": 1.2421282967656855e-06, + "loss": 1.0065, + "step": 7041 + }, + { + "epoch": 0.6350723722775848, + "grad_norm": 1.620254851063378, + "learning_rate": 1.2415876871882827e-06, + "loss": 0.9745, + "step": 7042 + }, + { + "epoch": 0.6351625558010552, + "grad_norm": 0.7319476410661063, + "learning_rate": 1.2410471423227998e-06, + "loss": 0.8506, + "step": 7043 + }, + { + "epoch": 0.6352527393245254, + "grad_norm": 1.5168868642122193, + "learning_rate": 1.24050666221536e-06, + "loss": 0.9416, + "step": 7044 + }, + { + "epoch": 0.6353429228479957, + "grad_norm": 1.437010557360217, + "learning_rate": 1.23996624691208e-06, + "loss": 1.0166, + "step": 7045 + }, + { + "epoch": 0.6354331063714659, + "grad_norm": 1.3689721802447699, + "learning_rate": 1.2394258964590693e-06, + "loss": 0.9371, + "step": 7046 + }, + { + "epoch": 0.6355232898949362, + "grad_norm": 1.8347006019796306, + "learning_rate": 1.238885610902436e-06, + "loss": 0.9017, + "step": 7047 + }, + { + "epoch": 0.6356134734184065, + "grad_norm": 1.9448232380191477, + "learning_rate": 1.2383453902882787e-06, + "loss": 0.9606, + "step": 7048 + }, + { + "epoch": 0.6357036569418767, + "grad_norm": 1.7072170688918125, + "learning_rate": 1.2378052346626927e-06, + "loss": 1.0163, + "step": 7049 + }, + { + "epoch": 0.6357938404653469, + "grad_norm": 2.2997353318562843, + "learning_rate": 1.2372651440717665e-06, + "loss": 0.8766, + "step": 7050 + }, + { + "epoch": 0.6358840239888173, + "grad_norm": 1.6943369225051172, + "learning_rate": 1.236725118561584e-06, + "loss": 0.9848, + "step": 7051 + }, + { + "epoch": 0.6359742075122875, + "grad_norm": 1.3577418108099464, + "learning_rate": 1.2361851581782232e-06, + "loss": 0.9506, + "step": 7052 + }, + { + "epoch": 0.6360643910357577, + "grad_norm": 1.4941249073672882, + "learning_rate": 1.2356452629677554e-06, + "loss": 1.0133, + "step": 7053 + }, + { + "epoch": 0.6361545745592281, + "grad_norm": 1.7427292054319787, + "learning_rate": 1.2351054329762494e-06, + "loss": 0.8863, + "step": 7054 + }, + { + "epoch": 0.6362447580826983, + "grad_norm": 1.6108246727697613, + "learning_rate": 1.2345656682497648e-06, + "loss": 0.9795, + "step": 7055 + }, + { + "epoch": 0.6363349416061685, + "grad_norm": 1.2499605649449672, + "learning_rate": 1.2340259688343583e-06, + "loss": 0.9257, + "step": 7056 + }, + { + "epoch": 0.6364251251296388, + "grad_norm": 1.4258543230953495, + "learning_rate": 1.2334863347760803e-06, + "loss": 0.8516, + "step": 7057 + }, + { + "epoch": 0.6365153086531091, + "grad_norm": 0.6954374951084659, + "learning_rate": 1.2329467661209734e-06, + "loss": 0.7462, + "step": 7058 + }, + { + "epoch": 0.6366054921765794, + "grad_norm": 1.6296592088183948, + "learning_rate": 1.2324072629150788e-06, + "loss": 0.9911, + "step": 7059 + }, + { + "epoch": 0.6366956757000496, + "grad_norm": 2.694241783761746, + "learning_rate": 1.2318678252044287e-06, + "loss": 0.9988, + "step": 7060 + }, + { + "epoch": 0.6367858592235198, + "grad_norm": 1.8997445336308518, + "learning_rate": 1.2313284530350512e-06, + "loss": 1.009, + "step": 7061 + }, + { + "epoch": 0.6368760427469902, + "grad_norm": 1.5024324403829408, + "learning_rate": 1.230789146452969e-06, + "loss": 0.8855, + "step": 7062 + }, + { + "epoch": 0.6369662262704604, + "grad_norm": 1.56167496834051, + "learning_rate": 1.2302499055041974e-06, + "loss": 0.9484, + "step": 7063 + }, + { + "epoch": 0.6370564097939306, + "grad_norm": 0.7059950216363658, + "learning_rate": 1.2297107302347488e-06, + "loss": 0.783, + "step": 7064 + }, + { + "epoch": 0.6371465933174009, + "grad_norm": 1.5102044142745337, + "learning_rate": 1.2291716206906275e-06, + "loss": 1.0214, + "step": 7065 + }, + { + "epoch": 0.6372367768408712, + "grad_norm": 2.011231476370681, + "learning_rate": 1.2286325769178345e-06, + "loss": 1.0186, + "step": 7066 + }, + { + "epoch": 0.6373269603643414, + "grad_norm": 1.5066186002783557, + "learning_rate": 1.2280935989623633e-06, + "loss": 0.9779, + "step": 7067 + }, + { + "epoch": 0.6374171438878117, + "grad_norm": 1.6429579360323054, + "learning_rate": 1.2275546868702017e-06, + "loss": 0.8821, + "step": 7068 + }, + { + "epoch": 0.6375073274112819, + "grad_norm": 1.606217629666833, + "learning_rate": 1.2270158406873341e-06, + "loss": 0.8909, + "step": 7069 + }, + { + "epoch": 0.6375975109347523, + "grad_norm": 1.3958258510028463, + "learning_rate": 1.2264770604597363e-06, + "loss": 0.9934, + "step": 7070 + }, + { + "epoch": 0.6376876944582225, + "grad_norm": 1.513132779504378, + "learning_rate": 1.2259383462333819e-06, + "loss": 1.0539, + "step": 7071 + }, + { + "epoch": 0.6377778779816927, + "grad_norm": 1.7097721398665005, + "learning_rate": 1.2253996980542359e-06, + "loss": 0.9072, + "step": 7072 + }, + { + "epoch": 0.637868061505163, + "grad_norm": 1.624141906711352, + "learning_rate": 1.2248611159682578e-06, + "loss": 0.9342, + "step": 7073 + }, + { + "epoch": 0.6379582450286333, + "grad_norm": 0.7635234483641841, + "learning_rate": 1.2243226000214044e-06, + "loss": 0.7659, + "step": 7074 + }, + { + "epoch": 0.6380484285521035, + "grad_norm": 2.7268600520114, + "learning_rate": 1.2237841502596232e-06, + "loss": 0.9501, + "step": 7075 + }, + { + "epoch": 0.6381386120755738, + "grad_norm": 1.6694319513875733, + "learning_rate": 1.2232457667288583e-06, + "loss": 0.9568, + "step": 7076 + }, + { + "epoch": 0.6382287955990441, + "grad_norm": 1.7273673688704942, + "learning_rate": 1.2227074494750476e-06, + "loss": 0.9581, + "step": 7077 + }, + { + "epoch": 0.6383189791225143, + "grad_norm": 1.5954329636662063, + "learning_rate": 1.2221691985441238e-06, + "loss": 0.9868, + "step": 7078 + }, + { + "epoch": 0.6384091626459846, + "grad_norm": 1.5322234504429837, + "learning_rate": 1.2216310139820128e-06, + "loss": 0.9011, + "step": 7079 + }, + { + "epoch": 0.6384993461694548, + "grad_norm": 1.469708718226392, + "learning_rate": 1.2210928958346347e-06, + "loss": 0.994, + "step": 7080 + }, + { + "epoch": 0.6385895296929252, + "grad_norm": 1.4298034787689462, + "learning_rate": 1.2205548441479065e-06, + "loss": 0.8788, + "step": 7081 + }, + { + "epoch": 0.6386797132163954, + "grad_norm": 1.5671353058571598, + "learning_rate": 1.2200168589677357e-06, + "loss": 0.8827, + "step": 7082 + }, + { + "epoch": 0.6387698967398656, + "grad_norm": 3.9425134379678566, + "learning_rate": 1.2194789403400284e-06, + "loss": 0.9172, + "step": 7083 + }, + { + "epoch": 0.6388600802633358, + "grad_norm": 1.8219315691846794, + "learning_rate": 1.2189410883106816e-06, + "loss": 0.9318, + "step": 7084 + }, + { + "epoch": 0.6389502637868062, + "grad_norm": 1.3861988651210997, + "learning_rate": 1.2184033029255872e-06, + "loss": 1.0185, + "step": 7085 + }, + { + "epoch": 0.6390404473102764, + "grad_norm": 0.753274603582724, + "learning_rate": 1.2178655842306334e-06, + "loss": 0.782, + "step": 7086 + }, + { + "epoch": 0.6391306308337467, + "grad_norm": 2.017221454361186, + "learning_rate": 1.2173279322716999e-06, + "loss": 0.9492, + "step": 7087 + }, + { + "epoch": 0.6392208143572169, + "grad_norm": 1.7990067152075042, + "learning_rate": 1.216790347094663e-06, + "loss": 0.9494, + "step": 7088 + }, + { + "epoch": 0.6393109978806872, + "grad_norm": 1.6073883535196969, + "learning_rate": 1.2162528287453927e-06, + "loss": 0.9918, + "step": 7089 + }, + { + "epoch": 0.6394011814041575, + "grad_norm": 1.7436510447655131, + "learning_rate": 1.215715377269752e-06, + "loss": 0.9591, + "step": 7090 + }, + { + "epoch": 0.6394913649276277, + "grad_norm": 1.5031628641605694, + "learning_rate": 1.2151779927136003e-06, + "loss": 0.9333, + "step": 7091 + }, + { + "epoch": 0.6395815484510979, + "grad_norm": 1.7140134569661378, + "learning_rate": 1.2146406751227893e-06, + "loss": 1.0243, + "step": 7092 + }, + { + "epoch": 0.6396717319745683, + "grad_norm": 1.5085720066615438, + "learning_rate": 1.214103424543167e-06, + "loss": 0.985, + "step": 7093 + }, + { + "epoch": 0.6397619154980385, + "grad_norm": 1.4548703448544527, + "learning_rate": 1.2135662410205735e-06, + "loss": 0.8641, + "step": 7094 + }, + { + "epoch": 0.6398520990215087, + "grad_norm": 1.6223945270694313, + "learning_rate": 1.2130291246008444e-06, + "loss": 0.8683, + "step": 7095 + }, + { + "epoch": 0.639942282544979, + "grad_norm": 1.352821821822603, + "learning_rate": 1.21249207532981e-06, + "loss": 0.8704, + "step": 7096 + }, + { + "epoch": 0.6400324660684493, + "grad_norm": 1.481951451704593, + "learning_rate": 1.2119550932532936e-06, + "loss": 0.8719, + "step": 7097 + }, + { + "epoch": 0.6401226495919196, + "grad_norm": 1.6281777635104355, + "learning_rate": 1.2114181784171144e-06, + "loss": 0.9253, + "step": 7098 + }, + { + "epoch": 0.6402128331153898, + "grad_norm": 1.7225726533358492, + "learning_rate": 1.2108813308670837e-06, + "loss": 0.8529, + "step": 7099 + }, + { + "epoch": 0.6403030166388601, + "grad_norm": 1.7264097672653724, + "learning_rate": 1.2103445506490099e-06, + "loss": 0.9519, + "step": 7100 + }, + { + "epoch": 0.6403932001623304, + "grad_norm": 0.7518738543849726, + "learning_rate": 1.209807837808693e-06, + "loss": 0.7934, + "step": 7101 + }, + { + "epoch": 0.6404833836858006, + "grad_norm": 1.7005916968200254, + "learning_rate": 1.2092711923919282e-06, + "loss": 0.9878, + "step": 7102 + }, + { + "epoch": 0.6405735672092708, + "grad_norm": 1.8749715167107373, + "learning_rate": 1.2087346144445053e-06, + "loss": 0.9133, + "step": 7103 + }, + { + "epoch": 0.6406637507327412, + "grad_norm": 2.0018377205685196, + "learning_rate": 1.2081981040122081e-06, + "loss": 0.9589, + "step": 7104 + }, + { + "epoch": 0.6407539342562114, + "grad_norm": 1.4764114383344336, + "learning_rate": 1.2076616611408148e-06, + "loss": 0.9009, + "step": 7105 + }, + { + "epoch": 0.6408441177796816, + "grad_norm": 2.0813843829785577, + "learning_rate": 1.2071252858760972e-06, + "loss": 0.9853, + "step": 7106 + }, + { + "epoch": 0.6409343013031519, + "grad_norm": 1.999029639401713, + "learning_rate": 1.2065889782638218e-06, + "loss": 0.9336, + "step": 7107 + }, + { + "epoch": 0.6410244848266222, + "grad_norm": 1.5768704101398934, + "learning_rate": 1.2060527383497506e-06, + "loss": 0.9332, + "step": 7108 + }, + { + "epoch": 0.6411146683500925, + "grad_norm": 1.410533653752987, + "learning_rate": 1.2055165661796363e-06, + "loss": 1.0638, + "step": 7109 + }, + { + "epoch": 0.6412048518735627, + "grad_norm": 1.577283124811982, + "learning_rate": 1.2049804617992303e-06, + "loss": 1.0225, + "step": 7110 + }, + { + "epoch": 0.6412950353970329, + "grad_norm": 0.6599367216871361, + "learning_rate": 1.204444425254275e-06, + "loss": 0.7936, + "step": 7111 + }, + { + "epoch": 0.6413852189205033, + "grad_norm": 1.6716116581946832, + "learning_rate": 1.203908456590507e-06, + "loss": 0.9274, + "step": 7112 + }, + { + "epoch": 0.6414754024439735, + "grad_norm": 4.050111867984448, + "learning_rate": 1.20337255585366e-06, + "loss": 0.9267, + "step": 7113 + }, + { + "epoch": 0.6415655859674437, + "grad_norm": 0.6565092800745392, + "learning_rate": 1.2028367230894582e-06, + "loss": 0.7401, + "step": 7114 + }, + { + "epoch": 0.641655769490914, + "grad_norm": 1.6650822340215143, + "learning_rate": 1.2023009583436237e-06, + "loss": 0.8732, + "step": 7115 + }, + { + "epoch": 0.6417459530143843, + "grad_norm": 1.8013982481244804, + "learning_rate": 1.2017652616618698e-06, + "loss": 1.0164, + "step": 7116 + }, + { + "epoch": 0.6418361365378545, + "grad_norm": 1.5811669520104463, + "learning_rate": 1.2012296330899048e-06, + "loss": 0.9474, + "step": 7117 + }, + { + "epoch": 0.6419263200613248, + "grad_norm": 1.4801575829407199, + "learning_rate": 1.200694072673432e-06, + "loss": 0.9829, + "step": 7118 + }, + { + "epoch": 0.642016503584795, + "grad_norm": 1.459555395441386, + "learning_rate": 1.200158580458148e-06, + "loss": 0.9805, + "step": 7119 + }, + { + "epoch": 0.6421066871082654, + "grad_norm": 1.5025424074100784, + "learning_rate": 1.1996231564897448e-06, + "loss": 0.9253, + "step": 7120 + }, + { + "epoch": 0.6421968706317356, + "grad_norm": 1.5591200320053744, + "learning_rate": 1.1990878008139062e-06, + "loss": 0.9078, + "step": 7121 + }, + { + "epoch": 0.6422870541552058, + "grad_norm": 1.749243232041193, + "learning_rate": 1.1985525134763132e-06, + "loss": 0.98, + "step": 7122 + }, + { + "epoch": 0.6423772376786762, + "grad_norm": 1.4845390630704551, + "learning_rate": 1.1980172945226393e-06, + "loss": 0.995, + "step": 7123 + }, + { + "epoch": 0.6424674212021464, + "grad_norm": 1.7986442148835353, + "learning_rate": 1.197482143998551e-06, + "loss": 0.9585, + "step": 7124 + }, + { + "epoch": 0.6425576047256166, + "grad_norm": 1.7312089357429297, + "learning_rate": 1.196947061949712e-06, + "loss": 0.9362, + "step": 7125 + }, + { + "epoch": 0.6426477882490869, + "grad_norm": 3.03912259314583, + "learning_rate": 1.1964120484217768e-06, + "loss": 1.0076, + "step": 7126 + }, + { + "epoch": 0.6427379717725572, + "grad_norm": 0.7502625721021412, + "learning_rate": 1.1958771034603975e-06, + "loss": 0.7641, + "step": 7127 + }, + { + "epoch": 0.6428281552960274, + "grad_norm": 1.4602257938060417, + "learning_rate": 1.1953422271112175e-06, + "loss": 0.8901, + "step": 7128 + }, + { + "epoch": 0.6429183388194977, + "grad_norm": 1.7911125331606925, + "learning_rate": 1.1948074194198748e-06, + "loss": 0.9382, + "step": 7129 + }, + { + "epoch": 0.6430085223429679, + "grad_norm": 1.8786596187899751, + "learning_rate": 1.1942726804320033e-06, + "loss": 0.8628, + "step": 7130 + }, + { + "epoch": 0.6430987058664382, + "grad_norm": 1.888450948896783, + "learning_rate": 1.1937380101932295e-06, + "loss": 0.8921, + "step": 7131 + }, + { + "epoch": 0.6431888893899085, + "grad_norm": 1.3407945734652822, + "learning_rate": 1.1932034087491745e-06, + "loss": 0.9453, + "step": 7132 + }, + { + "epoch": 0.6432790729133787, + "grad_norm": 1.640771986416919, + "learning_rate": 1.1926688761454531e-06, + "loss": 0.8667, + "step": 7133 + }, + { + "epoch": 0.643369256436849, + "grad_norm": 1.1562235545665296, + "learning_rate": 1.1921344124276746e-06, + "loss": 0.9662, + "step": 7134 + }, + { + "epoch": 0.6434594399603193, + "grad_norm": 1.438575591014891, + "learning_rate": 1.1916000176414431e-06, + "loss": 0.9717, + "step": 7135 + }, + { + "epoch": 0.6435496234837895, + "grad_norm": 1.5855298951361219, + "learning_rate": 1.1910656918323546e-06, + "loss": 0.9867, + "step": 7136 + }, + { + "epoch": 0.6436398070072598, + "grad_norm": 1.4566872525597316, + "learning_rate": 1.1905314350460024e-06, + "loss": 0.9806, + "step": 7137 + }, + { + "epoch": 0.64372999053073, + "grad_norm": 1.396425728800246, + "learning_rate": 1.1899972473279717e-06, + "loss": 1.0044, + "step": 7138 + }, + { + "epoch": 0.6438201740542003, + "grad_norm": 1.754331133721797, + "learning_rate": 1.1894631287238414e-06, + "loss": 0.9426, + "step": 7139 + }, + { + "epoch": 0.6439103575776706, + "grad_norm": 1.5105886251846838, + "learning_rate": 1.188929079279187e-06, + "loss": 0.9584, + "step": 7140 + }, + { + "epoch": 0.6440005411011408, + "grad_norm": 1.9296850907160124, + "learning_rate": 1.1883950990395751e-06, + "loss": 0.8865, + "step": 7141 + }, + { + "epoch": 0.644090724624611, + "grad_norm": 1.63132726844395, + "learning_rate": 1.187861188050569e-06, + "loss": 0.889, + "step": 7142 + }, + { + "epoch": 0.6441809081480814, + "grad_norm": 1.292278877911263, + "learning_rate": 1.187327346357724e-06, + "loss": 0.9333, + "step": 7143 + }, + { + "epoch": 0.6442710916715516, + "grad_norm": 1.7976856559195156, + "learning_rate": 1.1867935740065912e-06, + "loss": 0.9712, + "step": 7144 + }, + { + "epoch": 0.6443612751950218, + "grad_norm": 1.5099042702954857, + "learning_rate": 1.1862598710427148e-06, + "loss": 0.8414, + "step": 7145 + }, + { + "epoch": 0.6444514587184921, + "grad_norm": 2.333728779480237, + "learning_rate": 1.1857262375116328e-06, + "loss": 0.9135, + "step": 7146 + }, + { + "epoch": 0.6445416422419624, + "grad_norm": 1.5453657243316505, + "learning_rate": 1.1851926734588783e-06, + "loss": 0.9554, + "step": 7147 + }, + { + "epoch": 0.6446318257654327, + "grad_norm": 1.5224077706048957, + "learning_rate": 1.184659178929977e-06, + "loss": 0.9755, + "step": 7148 + }, + { + "epoch": 0.6447220092889029, + "grad_norm": 1.6226977398581648, + "learning_rate": 1.1841257539704513e-06, + "loss": 0.9461, + "step": 7149 + }, + { + "epoch": 0.6448121928123732, + "grad_norm": 1.5804331301693113, + "learning_rate": 1.1835923986258146e-06, + "loss": 0.9647, + "step": 7150 + }, + { + "epoch": 0.6449023763358435, + "grad_norm": 1.5241056884252997, + "learning_rate": 1.1830591129415754e-06, + "loss": 0.9077, + "step": 7151 + }, + { + "epoch": 0.6449925598593137, + "grad_norm": 1.7839202942854608, + "learning_rate": 1.182525896963238e-06, + "loss": 0.8928, + "step": 7152 + }, + { + "epoch": 0.6450827433827839, + "grad_norm": 2.1254108256225055, + "learning_rate": 1.181992750736298e-06, + "loss": 0.8844, + "step": 7153 + }, + { + "epoch": 0.6451729269062543, + "grad_norm": 1.7326004297561755, + "learning_rate": 1.1814596743062474e-06, + "loss": 0.9474, + "step": 7154 + }, + { + "epoch": 0.6452631104297245, + "grad_norm": 1.4416964272457395, + "learning_rate": 1.1809266677185711e-06, + "loss": 0.9765, + "step": 7155 + }, + { + "epoch": 0.6453532939531947, + "grad_norm": 0.6985860551390269, + "learning_rate": 1.180393731018747e-06, + "loss": 0.8134, + "step": 7156 + }, + { + "epoch": 0.645443477476665, + "grad_norm": 0.5872506495164029, + "learning_rate": 1.1798608642522498e-06, + "loss": 0.7523, + "step": 7157 + }, + { + "epoch": 0.6455336610001353, + "grad_norm": 3.01784485067152, + "learning_rate": 1.1793280674645454e-06, + "loss": 0.9348, + "step": 7158 + }, + { + "epoch": 0.6456238445236056, + "grad_norm": 1.6906154272686327, + "learning_rate": 1.1787953407010954e-06, + "loss": 1.0223, + "step": 7159 + }, + { + "epoch": 0.6457140280470758, + "grad_norm": 1.61713383534903, + "learning_rate": 1.1782626840073554e-06, + "loss": 0.9548, + "step": 7160 + }, + { + "epoch": 0.645804211570546, + "grad_norm": 1.5541153818350015, + "learning_rate": 1.1777300974287738e-06, + "loss": 0.8528, + "step": 7161 + }, + { + "epoch": 0.6458943950940164, + "grad_norm": 1.4609862946710188, + "learning_rate": 1.1771975810107947e-06, + "loss": 0.9797, + "step": 7162 + }, + { + "epoch": 0.6459845786174866, + "grad_norm": 1.7298053649607439, + "learning_rate": 1.1766651347988542e-06, + "loss": 1.0207, + "step": 7163 + }, + { + "epoch": 0.6460747621409568, + "grad_norm": 1.9274341848098684, + "learning_rate": 1.1761327588383848e-06, + "loss": 1.0637, + "step": 7164 + }, + { + "epoch": 0.6461649456644271, + "grad_norm": 1.516190256323916, + "learning_rate": 1.1756004531748105e-06, + "loss": 0.9602, + "step": 7165 + }, + { + "epoch": 0.6462551291878974, + "grad_norm": 1.5846874403008517, + "learning_rate": 1.1750682178535521e-06, + "loss": 0.9887, + "step": 7166 + }, + { + "epoch": 0.6463453127113676, + "grad_norm": 1.5700610514633708, + "learning_rate": 1.1745360529200218e-06, + "loss": 0.9, + "step": 7167 + }, + { + "epoch": 0.6464354962348379, + "grad_norm": 1.4361198268029884, + "learning_rate": 1.1740039584196261e-06, + "loss": 0.9741, + "step": 7168 + }, + { + "epoch": 0.6465256797583081, + "grad_norm": 1.45669600897304, + "learning_rate": 1.1734719343977683e-06, + "loss": 0.9784, + "step": 7169 + }, + { + "epoch": 0.6466158632817784, + "grad_norm": 1.515007050375066, + "learning_rate": 1.1729399808998416e-06, + "loss": 0.8922, + "step": 7170 + }, + { + "epoch": 0.6467060468052487, + "grad_norm": 1.4962872014085802, + "learning_rate": 1.1724080979712368e-06, + "loss": 0.8451, + "step": 7171 + }, + { + "epoch": 0.6467962303287189, + "grad_norm": 1.404100895179869, + "learning_rate": 1.1718762856573365e-06, + "loss": 1.021, + "step": 7172 + }, + { + "epoch": 0.6468864138521893, + "grad_norm": 1.3802546319453715, + "learning_rate": 1.1713445440035172e-06, + "loss": 0.9376, + "step": 7173 + }, + { + "epoch": 0.6469765973756595, + "grad_norm": 1.8187413454669181, + "learning_rate": 1.1708128730551506e-06, + "loss": 0.9897, + "step": 7174 + }, + { + "epoch": 0.6470667808991297, + "grad_norm": 1.664104550685498, + "learning_rate": 1.1702812728576019e-06, + "loss": 0.9348, + "step": 7175 + }, + { + "epoch": 0.6471569644226, + "grad_norm": 1.5698539861582845, + "learning_rate": 1.1697497434562303e-06, + "loss": 0.978, + "step": 7176 + }, + { + "epoch": 0.6472471479460703, + "grad_norm": 1.597279399272275, + "learning_rate": 1.1692182848963885e-06, + "loss": 0.9846, + "step": 7177 + }, + { + "epoch": 0.6473373314695405, + "grad_norm": 1.5403640587989342, + "learning_rate": 1.1686868972234227e-06, + "loss": 0.9109, + "step": 7178 + }, + { + "epoch": 0.6474275149930108, + "grad_norm": 1.3042324894553514, + "learning_rate": 1.1681555804826755e-06, + "loss": 0.9579, + "step": 7179 + }, + { + "epoch": 0.647517698516481, + "grad_norm": 1.458799623692984, + "learning_rate": 1.1676243347194806e-06, + "loss": 0.8902, + "step": 7180 + }, + { + "epoch": 0.6476078820399513, + "grad_norm": 1.286816715663975, + "learning_rate": 1.167093159979167e-06, + "loss": 0.9657, + "step": 7181 + }, + { + "epoch": 0.6476980655634216, + "grad_norm": 1.4842971781357837, + "learning_rate": 1.1665620563070575e-06, + "loss": 0.9778, + "step": 7182 + }, + { + "epoch": 0.6477882490868918, + "grad_norm": 2.9300336709545243, + "learning_rate": 1.1660310237484691e-06, + "loss": 1.0622, + "step": 7183 + }, + { + "epoch": 0.647878432610362, + "grad_norm": 1.3799187938321247, + "learning_rate": 1.165500062348713e-06, + "loss": 0.8946, + "step": 7184 + }, + { + "epoch": 0.6479686161338324, + "grad_norm": 1.7482533594004004, + "learning_rate": 1.164969172153091e-06, + "loss": 0.9957, + "step": 7185 + }, + { + "epoch": 0.6480587996573026, + "grad_norm": 1.6936118741741355, + "learning_rate": 1.1644383532069055e-06, + "loss": 0.9718, + "step": 7186 + }, + { + "epoch": 0.6481489831807729, + "grad_norm": 1.3554664633783846, + "learning_rate": 1.1639076055554454e-06, + "loss": 0.9919, + "step": 7187 + }, + { + "epoch": 0.6482391667042431, + "grad_norm": 1.512157760866901, + "learning_rate": 1.163376929244e-06, + "loss": 0.9367, + "step": 7188 + }, + { + "epoch": 0.6483293502277134, + "grad_norm": 1.3417885237299245, + "learning_rate": 1.1628463243178472e-06, + "loss": 0.9139, + "step": 7189 + }, + { + "epoch": 0.6484195337511837, + "grad_norm": 1.6475969675612692, + "learning_rate": 1.1623157908222623e-06, + "loss": 1.0077, + "step": 7190 + }, + { + "epoch": 0.6485097172746539, + "grad_norm": 1.415772810539181, + "learning_rate": 1.1617853288025129e-06, + "loss": 0.989, + "step": 7191 + }, + { + "epoch": 0.6485999007981241, + "grad_norm": 1.3614861767389685, + "learning_rate": 1.1612549383038612e-06, + "loss": 0.9218, + "step": 7192 + }, + { + "epoch": 0.6486900843215945, + "grad_norm": 1.2488638005606412, + "learning_rate": 1.1607246193715629e-06, + "loss": 0.9299, + "step": 7193 + }, + { + "epoch": 0.6487802678450647, + "grad_norm": 1.8657885941387327, + "learning_rate": 1.1601943720508684e-06, + "loss": 0.9683, + "step": 7194 + }, + { + "epoch": 0.6488704513685349, + "grad_norm": 1.4119550058663355, + "learning_rate": 1.1596641963870202e-06, + "loss": 0.9988, + "step": 7195 + }, + { + "epoch": 0.6489606348920053, + "grad_norm": 1.3967423634208547, + "learning_rate": 1.1591340924252561e-06, + "loss": 0.9618, + "step": 7196 + }, + { + "epoch": 0.6490508184154755, + "grad_norm": 0.8116212641284715, + "learning_rate": 1.158604060210808e-06, + "loss": 0.825, + "step": 7197 + }, + { + "epoch": 0.6491410019389457, + "grad_norm": 1.7178855109074689, + "learning_rate": 1.1580740997889008e-06, + "loss": 1.0126, + "step": 7198 + }, + { + "epoch": 0.649231185462416, + "grad_norm": 1.5795532410992943, + "learning_rate": 1.1575442112047544e-06, + "loss": 0.9466, + "step": 7199 + }, + { + "epoch": 0.6493213689858863, + "grad_norm": 1.6145478603609111, + "learning_rate": 1.1570143945035797e-06, + "loss": 0.8877, + "step": 7200 + }, + { + "epoch": 0.6494115525093566, + "grad_norm": 1.8268674780227163, + "learning_rate": 1.1564846497305864e-06, + "loss": 0.9044, + "step": 7201 + }, + { + "epoch": 0.6495017360328268, + "grad_norm": 1.5008280375950458, + "learning_rate": 1.1559549769309726e-06, + "loss": 0.9238, + "step": 7202 + }, + { + "epoch": 0.649591919556297, + "grad_norm": 1.361919039486438, + "learning_rate": 1.1554253761499358e-06, + "loss": 0.9434, + "step": 7203 + }, + { + "epoch": 0.6496821030797674, + "grad_norm": 1.8205060078333977, + "learning_rate": 1.1548958474326617e-06, + "loss": 0.9928, + "step": 7204 + }, + { + "epoch": 0.6497722866032376, + "grad_norm": 1.4280168120893189, + "learning_rate": 1.154366390824334e-06, + "loss": 0.8837, + "step": 7205 + }, + { + "epoch": 0.6498624701267078, + "grad_norm": 0.7629324609054587, + "learning_rate": 1.1538370063701287e-06, + "loss": 0.7596, + "step": 7206 + }, + { + "epoch": 0.6499526536501781, + "grad_norm": 1.344842555321559, + "learning_rate": 1.1533076941152153e-06, + "loss": 0.9084, + "step": 7207 + }, + { + "epoch": 0.6500428371736484, + "grad_norm": 1.7651004518447495, + "learning_rate": 1.1527784541047583e-06, + "loss": 0.8748, + "step": 7208 + }, + { + "epoch": 0.6501330206971186, + "grad_norm": 1.8211222799808304, + "learning_rate": 1.1522492863839152e-06, + "loss": 0.9071, + "step": 7209 + }, + { + "epoch": 0.6502232042205889, + "grad_norm": 1.9372298759924553, + "learning_rate": 1.1517201909978382e-06, + "loss": 0.9156, + "step": 7210 + }, + { + "epoch": 0.6503133877440591, + "grad_norm": 0.6667611656169462, + "learning_rate": 1.151191167991671e-06, + "loss": 0.7806, + "step": 7211 + }, + { + "epoch": 0.6504035712675295, + "grad_norm": 1.8925745346932856, + "learning_rate": 1.1506622174105536e-06, + "loss": 0.9586, + "step": 7212 + }, + { + "epoch": 0.6504937547909997, + "grad_norm": 1.7808571767155426, + "learning_rate": 1.1501333392996194e-06, + "loss": 0.9813, + "step": 7213 + }, + { + "epoch": 0.6505839383144699, + "grad_norm": 0.6917251885792833, + "learning_rate": 1.1496045337039943e-06, + "loss": 0.7863, + "step": 7214 + }, + { + "epoch": 0.6506741218379402, + "grad_norm": 1.512970556407584, + "learning_rate": 1.1490758006687995e-06, + "loss": 0.9092, + "step": 7215 + }, + { + "epoch": 0.6507643053614105, + "grad_norm": 1.5967336663037415, + "learning_rate": 1.1485471402391502e-06, + "loss": 0.9302, + "step": 7216 + }, + { + "epoch": 0.6508544888848807, + "grad_norm": 1.5178029251032243, + "learning_rate": 1.1480185524601522e-06, + "loss": 0.9213, + "step": 7217 + }, + { + "epoch": 0.650944672408351, + "grad_norm": 1.7038272494787388, + "learning_rate": 1.1474900373769108e-06, + "loss": 0.9564, + "step": 7218 + }, + { + "epoch": 0.6510348559318213, + "grad_norm": 0.7723296737545781, + "learning_rate": 1.1469615950345184e-06, + "loss": 0.7257, + "step": 7219 + }, + { + "epoch": 0.6511250394552915, + "grad_norm": 0.7281319679287035, + "learning_rate": 1.1464332254780678e-06, + "loss": 0.7646, + "step": 7220 + }, + { + "epoch": 0.6512152229787618, + "grad_norm": 1.4528868285210201, + "learning_rate": 1.1459049287526404e-06, + "loss": 0.938, + "step": 7221 + }, + { + "epoch": 0.651305406502232, + "grad_norm": 2.2702460156529587, + "learning_rate": 1.1453767049033137e-06, + "loss": 0.9135, + "step": 7222 + }, + { + "epoch": 0.6513955900257024, + "grad_norm": 1.5641205585385887, + "learning_rate": 1.1448485539751586e-06, + "loss": 0.9933, + "step": 7223 + }, + { + "epoch": 0.6514857735491726, + "grad_norm": 1.4133226767099234, + "learning_rate": 1.1443204760132408e-06, + "loss": 0.9212, + "step": 7224 + }, + { + "epoch": 0.6515759570726428, + "grad_norm": 1.4194597124413713, + "learning_rate": 1.1437924710626185e-06, + "loss": 0.9296, + "step": 7225 + }, + { + "epoch": 0.651666140596113, + "grad_norm": 0.7685385172789658, + "learning_rate": 1.1432645391683429e-06, + "loss": 0.7482, + "step": 7226 + }, + { + "epoch": 0.6517563241195834, + "grad_norm": 1.8286604789886944, + "learning_rate": 1.1427366803754609e-06, + "loss": 0.9854, + "step": 7227 + }, + { + "epoch": 0.6518465076430536, + "grad_norm": 1.4092073497317432, + "learning_rate": 1.142208894729012e-06, + "loss": 0.9641, + "step": 7228 + }, + { + "epoch": 0.6519366911665239, + "grad_norm": 1.2802827487027921, + "learning_rate": 1.1416811822740301e-06, + "loss": 0.9911, + "step": 7229 + }, + { + "epoch": 0.6520268746899941, + "grad_norm": 1.5095021954444974, + "learning_rate": 1.1411535430555428e-06, + "loss": 1.0016, + "step": 7230 + }, + { + "epoch": 0.6521170582134644, + "grad_norm": 1.5275502565298351, + "learning_rate": 1.1406259771185705e-06, + "loss": 1.0056, + "step": 7231 + }, + { + "epoch": 0.6522072417369347, + "grad_norm": 1.3615791603804386, + "learning_rate": 1.1400984845081282e-06, + "loss": 0.9096, + "step": 7232 + }, + { + "epoch": 0.6522974252604049, + "grad_norm": 1.8275773621973666, + "learning_rate": 1.139571065269226e-06, + "loss": 1.0288, + "step": 7233 + }, + { + "epoch": 0.6523876087838751, + "grad_norm": 1.4808283950273946, + "learning_rate": 1.139043719446863e-06, + "loss": 0.9972, + "step": 7234 + }, + { + "epoch": 0.6524777923073455, + "grad_norm": 1.4771156687710116, + "learning_rate": 1.1385164470860385e-06, + "loss": 1.0218, + "step": 7235 + }, + { + "epoch": 0.6525679758308157, + "grad_norm": 1.7229979991358797, + "learning_rate": 1.1379892482317403e-06, + "loss": 0.9508, + "step": 7236 + }, + { + "epoch": 0.652658159354286, + "grad_norm": 1.6244093481749704, + "learning_rate": 1.1374621229289524e-06, + "loss": 0.938, + "step": 7237 + }, + { + "epoch": 0.6527483428777562, + "grad_norm": 1.5589018782445105, + "learning_rate": 1.1369350712226525e-06, + "loss": 0.9376, + "step": 7238 + }, + { + "epoch": 0.6528385264012265, + "grad_norm": 1.3946818209016887, + "learning_rate": 1.136408093157811e-06, + "loss": 0.8962, + "step": 7239 + }, + { + "epoch": 0.6529287099246968, + "grad_norm": 1.6126647214516405, + "learning_rate": 1.1358811887793935e-06, + "loss": 0.9267, + "step": 7240 + }, + { + "epoch": 0.653018893448167, + "grad_norm": 1.4544509252804207, + "learning_rate": 1.135354358132356e-06, + "loss": 0.9845, + "step": 7241 + }, + { + "epoch": 0.6531090769716373, + "grad_norm": 1.7833048027941212, + "learning_rate": 1.1348276012616542e-06, + "loss": 0.9923, + "step": 7242 + }, + { + "epoch": 0.6531992604951076, + "grad_norm": 1.8951163420782924, + "learning_rate": 1.134300918212231e-06, + "loss": 0.9074, + "step": 7243 + }, + { + "epoch": 0.6532894440185778, + "grad_norm": 2.343339706111977, + "learning_rate": 1.133774309029027e-06, + "loss": 0.9617, + "step": 7244 + }, + { + "epoch": 0.653379627542048, + "grad_norm": 1.4921203917746733, + "learning_rate": 1.133247773756975e-06, + "loss": 0.9903, + "step": 7245 + }, + { + "epoch": 0.6534698110655184, + "grad_norm": 0.6984298773910717, + "learning_rate": 1.1327213124410024e-06, + "loss": 0.819, + "step": 7246 + }, + { + "epoch": 0.6535599945889886, + "grad_norm": 1.446448791873887, + "learning_rate": 1.1321949251260292e-06, + "loss": 0.9938, + "step": 7247 + }, + { + "epoch": 0.6536501781124588, + "grad_norm": 1.7270173729363438, + "learning_rate": 1.1316686118569712e-06, + "loss": 0.8369, + "step": 7248 + }, + { + "epoch": 0.6537403616359291, + "grad_norm": 1.5603146338453013, + "learning_rate": 1.1311423726787335e-06, + "loss": 0.8622, + "step": 7249 + }, + { + "epoch": 0.6538305451593994, + "grad_norm": 1.5842402353851155, + "learning_rate": 1.130616207636221e-06, + "loss": 0.9242, + "step": 7250 + }, + { + "epoch": 0.6539207286828697, + "grad_norm": 1.5623265742378263, + "learning_rate": 1.1300901167743263e-06, + "loss": 0.9564, + "step": 7251 + }, + { + "epoch": 0.6540109122063399, + "grad_norm": 1.2998403176024944, + "learning_rate": 1.12956410013794e-06, + "loss": 0.9496, + "step": 7252 + }, + { + "epoch": 0.6541010957298101, + "grad_norm": 1.7460443886557901, + "learning_rate": 1.1290381577719436e-06, + "loss": 0.9383, + "step": 7253 + }, + { + "epoch": 0.6541912792532805, + "grad_norm": 1.7203522583181026, + "learning_rate": 1.1285122897212143e-06, + "loss": 0.8544, + "step": 7254 + }, + { + "epoch": 0.6542814627767507, + "grad_norm": 0.8053775532226257, + "learning_rate": 1.1279864960306228e-06, + "loss": 0.8192, + "step": 7255 + }, + { + "epoch": 0.6543716463002209, + "grad_norm": 1.6219455916784362, + "learning_rate": 1.1274607767450297e-06, + "loss": 1.0238, + "step": 7256 + }, + { + "epoch": 0.6544618298236912, + "grad_norm": 1.453542095743956, + "learning_rate": 1.126935131909296e-06, + "loss": 0.9256, + "step": 7257 + }, + { + "epoch": 0.6545520133471615, + "grad_norm": 1.3552398488365263, + "learning_rate": 1.1264095615682693e-06, + "loss": 0.9004, + "step": 7258 + }, + { + "epoch": 0.6546421968706317, + "grad_norm": 1.612507706261488, + "learning_rate": 1.1258840657667973e-06, + "loss": 0.8704, + "step": 7259 + }, + { + "epoch": 0.654732380394102, + "grad_norm": 1.5436452420787408, + "learning_rate": 1.125358644549716e-06, + "loss": 0.8802, + "step": 7260 + }, + { + "epoch": 0.6548225639175722, + "grad_norm": 1.4226054736088534, + "learning_rate": 1.1248332979618578e-06, + "loss": 0.8136, + "step": 7261 + }, + { + "epoch": 0.6549127474410426, + "grad_norm": 1.3158370239048218, + "learning_rate": 1.1243080260480482e-06, + "loss": 0.8417, + "step": 7262 + }, + { + "epoch": 0.6550029309645128, + "grad_norm": 1.8631221935272835, + "learning_rate": 1.1237828288531063e-06, + "loss": 1.0053, + "step": 7263 + }, + { + "epoch": 0.655093114487983, + "grad_norm": 1.5316335236730858, + "learning_rate": 1.1232577064218449e-06, + "loss": 0.869, + "step": 7264 + }, + { + "epoch": 0.6551832980114533, + "grad_norm": 1.6100217390073632, + "learning_rate": 1.1227326587990711e-06, + "loss": 1.0244, + "step": 7265 + }, + { + "epoch": 0.6552734815349236, + "grad_norm": 1.3611209224598164, + "learning_rate": 1.1222076860295832e-06, + "loss": 0.9338, + "step": 7266 + }, + { + "epoch": 0.6553636650583938, + "grad_norm": 2.1140318152885604, + "learning_rate": 1.1216827881581756e-06, + "loss": 0.97, + "step": 7267 + }, + { + "epoch": 0.6554538485818641, + "grad_norm": 1.7684980664597225, + "learning_rate": 1.1211579652296355e-06, + "loss": 0.9457, + "step": 7268 + }, + { + "epoch": 0.6555440321053344, + "grad_norm": 1.699460768729147, + "learning_rate": 1.1206332172887438e-06, + "loss": 1.0293, + "step": 7269 + }, + { + "epoch": 0.6556342156288046, + "grad_norm": 1.5644055762905302, + "learning_rate": 1.1201085443802756e-06, + "loss": 0.9599, + "step": 7270 + }, + { + "epoch": 0.6557243991522749, + "grad_norm": 1.379972353741244, + "learning_rate": 1.1195839465489964e-06, + "loss": 0.9155, + "step": 7271 + }, + { + "epoch": 0.6558145826757451, + "grad_norm": 1.8264651465924333, + "learning_rate": 1.1190594238396708e-06, + "loss": 0.9934, + "step": 7272 + }, + { + "epoch": 0.6559047661992154, + "grad_norm": 1.2256242115381983, + "learning_rate": 1.1185349762970515e-06, + "loss": 0.9193, + "step": 7273 + }, + { + "epoch": 0.6559949497226857, + "grad_norm": 1.508747818762744, + "learning_rate": 1.1180106039658896e-06, + "loss": 0.9476, + "step": 7274 + }, + { + "epoch": 0.6560851332461559, + "grad_norm": 0.6701442027059739, + "learning_rate": 1.117486306890925e-06, + "loss": 0.7594, + "step": 7275 + }, + { + "epoch": 0.6561753167696261, + "grad_norm": 1.830273140498255, + "learning_rate": 1.116962085116896e-06, + "loss": 0.8989, + "step": 7276 + }, + { + "epoch": 0.6562655002930965, + "grad_norm": 0.70793831319721, + "learning_rate": 1.1164379386885302e-06, + "loss": 0.719, + "step": 7277 + }, + { + "epoch": 0.6563556838165667, + "grad_norm": 1.6322510406495498, + "learning_rate": 1.1159138676505516e-06, + "loss": 0.9963, + "step": 7278 + }, + { + "epoch": 0.656445867340037, + "grad_norm": 1.54588363168052, + "learning_rate": 1.1153898720476761e-06, + "loss": 1.0068, + "step": 7279 + }, + { + "epoch": 0.6565360508635072, + "grad_norm": 2.3875494227861207, + "learning_rate": 1.114865951924615e-06, + "loss": 0.941, + "step": 7280 + }, + { + "epoch": 0.6566262343869775, + "grad_norm": 1.4956840571704646, + "learning_rate": 1.1143421073260721e-06, + "loss": 0.893, + "step": 7281 + }, + { + "epoch": 0.6567164179104478, + "grad_norm": 1.572272752070415, + "learning_rate": 1.1138183382967432e-06, + "loss": 0.8854, + "step": 7282 + }, + { + "epoch": 0.656806601433918, + "grad_norm": 1.6393878450228583, + "learning_rate": 1.11329464488132e-06, + "loss": 0.8607, + "step": 7283 + }, + { + "epoch": 0.6568967849573882, + "grad_norm": 1.5819296721826586, + "learning_rate": 1.112771027124487e-06, + "loss": 0.9229, + "step": 7284 + }, + { + "epoch": 0.6569869684808586, + "grad_norm": 1.5579725667279118, + "learning_rate": 1.112247485070922e-06, + "loss": 0.9613, + "step": 7285 + }, + { + "epoch": 0.6570771520043288, + "grad_norm": 1.830382689188082, + "learning_rate": 1.1117240187652968e-06, + "loss": 0.9557, + "step": 7286 + }, + { + "epoch": 0.657167335527799, + "grad_norm": 1.7020742692398656, + "learning_rate": 1.1112006282522767e-06, + "loss": 1.0304, + "step": 7287 + }, + { + "epoch": 0.6572575190512693, + "grad_norm": 1.5388171779864674, + "learning_rate": 1.1106773135765183e-06, + "loss": 0.9327, + "step": 7288 + }, + { + "epoch": 0.6573477025747396, + "grad_norm": 1.5254007844672082, + "learning_rate": 1.110154074782677e-06, + "loss": 0.9275, + "step": 7289 + }, + { + "epoch": 0.6574378860982099, + "grad_norm": 1.6317408937527713, + "learning_rate": 1.1096309119153948e-06, + "loss": 0.9455, + "step": 7290 + }, + { + "epoch": 0.6575280696216801, + "grad_norm": 0.7016231391714876, + "learning_rate": 1.1091078250193145e-06, + "loss": 0.8214, + "step": 7291 + }, + { + "epoch": 0.6576182531451504, + "grad_norm": 1.8181403957330113, + "learning_rate": 1.108584814139066e-06, + "loss": 0.9513, + "step": 7292 + }, + { + "epoch": 0.6577084366686207, + "grad_norm": 1.584428416899741, + "learning_rate": 1.108061879319276e-06, + "loss": 0.9466, + "step": 7293 + }, + { + "epoch": 0.6577986201920909, + "grad_norm": 3.405564965474866, + "learning_rate": 1.1075390206045648e-06, + "loss": 1.0056, + "step": 7294 + }, + { + "epoch": 0.6578888037155611, + "grad_norm": 1.485574819626481, + "learning_rate": 1.1070162380395454e-06, + "loss": 0.9905, + "step": 7295 + }, + { + "epoch": 0.6579789872390315, + "grad_norm": 1.709458848755292, + "learning_rate": 1.1064935316688253e-06, + "loss": 0.8922, + "step": 7296 + }, + { + "epoch": 0.6580691707625017, + "grad_norm": 1.5529130940779619, + "learning_rate": 1.105970901537002e-06, + "loss": 1.0325, + "step": 7297 + }, + { + "epoch": 0.6581593542859719, + "grad_norm": 1.493828794563909, + "learning_rate": 1.1054483476886727e-06, + "loss": 0.9929, + "step": 7298 + }, + { + "epoch": 0.6582495378094422, + "grad_norm": 1.4151822709408677, + "learning_rate": 1.1049258701684222e-06, + "loss": 0.8752, + "step": 7299 + }, + { + "epoch": 0.6583397213329125, + "grad_norm": 1.5451116819046142, + "learning_rate": 1.1044034690208315e-06, + "loss": 0.897, + "step": 7300 + }, + { + "epoch": 0.6584299048563828, + "grad_norm": 1.8016812605477042, + "learning_rate": 1.1038811442904755e-06, + "loss": 0.9843, + "step": 7301 + }, + { + "epoch": 0.658520088379853, + "grad_norm": 1.529995462747466, + "learning_rate": 1.103358896021921e-06, + "loss": 0.8808, + "step": 7302 + }, + { + "epoch": 0.6586102719033232, + "grad_norm": 1.6359726108692665, + "learning_rate": 1.1028367242597298e-06, + "loss": 0.9304, + "step": 7303 + }, + { + "epoch": 0.6587004554267936, + "grad_norm": 1.7978390305231002, + "learning_rate": 1.102314629048457e-06, + "loss": 0.9865, + "step": 7304 + }, + { + "epoch": 0.6587906389502638, + "grad_norm": 1.5876455645768999, + "learning_rate": 1.1017926104326484e-06, + "loss": 1.001, + "step": 7305 + }, + { + "epoch": 0.658880822473734, + "grad_norm": 1.4273431453275522, + "learning_rate": 1.1012706684568483e-06, + "loss": 1.0048, + "step": 7306 + }, + { + "epoch": 0.6589710059972043, + "grad_norm": 1.482763198338427, + "learning_rate": 1.1007488031655894e-06, + "loss": 0.9691, + "step": 7307 + }, + { + "epoch": 0.6590611895206746, + "grad_norm": 1.6476353144321754, + "learning_rate": 1.1002270146034013e-06, + "loss": 0.9055, + "step": 7308 + }, + { + "epoch": 0.6591513730441448, + "grad_norm": 1.2487581760264848, + "learning_rate": 1.0997053028148052e-06, + "loss": 0.9803, + "step": 7309 + }, + { + "epoch": 0.6592415565676151, + "grad_norm": 1.3799726992822956, + "learning_rate": 1.0991836678443173e-06, + "loss": 0.9523, + "step": 7310 + }, + { + "epoch": 0.6593317400910853, + "grad_norm": 1.809877010403095, + "learning_rate": 1.0986621097364465e-06, + "loss": 0.8957, + "step": 7311 + }, + { + "epoch": 0.6594219236145556, + "grad_norm": 1.5287505010571287, + "learning_rate": 1.0981406285356932e-06, + "loss": 0.865, + "step": 7312 + }, + { + "epoch": 0.6595121071380259, + "grad_norm": 1.7088997607401348, + "learning_rate": 1.0976192242865554e-06, + "loss": 0.8748, + "step": 7313 + }, + { + "epoch": 0.6596022906614961, + "grad_norm": 1.6882372058240696, + "learning_rate": 1.0970978970335202e-06, + "loss": 1.0182, + "step": 7314 + }, + { + "epoch": 0.6596924741849665, + "grad_norm": 2.0935964955097583, + "learning_rate": 1.0965766468210714e-06, + "loss": 0.9733, + "step": 7315 + }, + { + "epoch": 0.6597826577084367, + "grad_norm": 8.439008563779916, + "learning_rate": 1.0960554736936843e-06, + "loss": 0.9609, + "step": 7316 + }, + { + "epoch": 0.6598728412319069, + "grad_norm": 1.7883182626796112, + "learning_rate": 1.0955343776958283e-06, + "loss": 0.8673, + "step": 7317 + }, + { + "epoch": 0.6599630247553772, + "grad_norm": 1.5733864526887202, + "learning_rate": 1.0950133588719665e-06, + "loss": 0.9673, + "step": 7318 + }, + { + "epoch": 0.6600532082788475, + "grad_norm": 1.5380420690668648, + "learning_rate": 1.0944924172665551e-06, + "loss": 0.9738, + "step": 7319 + }, + { + "epoch": 0.6601433918023177, + "grad_norm": 1.6578229955975565, + "learning_rate": 1.0939715529240437e-06, + "loss": 0.987, + "step": 7320 + }, + { + "epoch": 0.660233575325788, + "grad_norm": 1.791210767125987, + "learning_rate": 1.0934507658888755e-06, + "loss": 0.9841, + "step": 7321 + }, + { + "epoch": 0.6603237588492582, + "grad_norm": 1.3859423545915805, + "learning_rate": 1.092930056205486e-06, + "loss": 0.9221, + "step": 7322 + }, + { + "epoch": 0.6604139423727285, + "grad_norm": 1.9940804258520355, + "learning_rate": 1.092409423918306e-06, + "loss": 0.9171, + "step": 7323 + }, + { + "epoch": 0.6605041258961988, + "grad_norm": 1.2598089166344155, + "learning_rate": 1.0918888690717581e-06, + "loss": 0.8732, + "step": 7324 + }, + { + "epoch": 0.660594309419669, + "grad_norm": 1.6689360268447797, + "learning_rate": 1.091368391710259e-06, + "loss": 1.0133, + "step": 7325 + }, + { + "epoch": 0.6606844929431392, + "grad_norm": 1.6136635241988442, + "learning_rate": 1.0908479918782198e-06, + "loss": 0.8985, + "step": 7326 + }, + { + "epoch": 0.6607746764666096, + "grad_norm": 1.4443047386653218, + "learning_rate": 1.0903276696200413e-06, + "loss": 0.9759, + "step": 7327 + }, + { + "epoch": 0.6608648599900798, + "grad_norm": 1.669372539625565, + "learning_rate": 1.0898074249801234e-06, + "loss": 0.9577, + "step": 7328 + }, + { + "epoch": 0.6609550435135501, + "grad_norm": 1.508095828302615, + "learning_rate": 1.0892872580028533e-06, + "loss": 0.9677, + "step": 7329 + }, + { + "epoch": 0.6610452270370203, + "grad_norm": 1.4769838730971845, + "learning_rate": 1.0887671687326178e-06, + "loss": 0.981, + "step": 7330 + }, + { + "epoch": 0.6611354105604906, + "grad_norm": 1.5144765376341864, + "learning_rate": 1.0882471572137908e-06, + "loss": 0.9122, + "step": 7331 + }, + { + "epoch": 0.6612255940839609, + "grad_norm": 1.642429485694444, + "learning_rate": 1.087727223490744e-06, + "loss": 0.9957, + "step": 7332 + }, + { + "epoch": 0.6613157776074311, + "grad_norm": 1.611167280087654, + "learning_rate": 1.0872073676078405e-06, + "loss": 1.0115, + "step": 7333 + }, + { + "epoch": 0.6614059611309013, + "grad_norm": 2.246994978840968, + "learning_rate": 1.0866875896094375e-06, + "loss": 0.9082, + "step": 7334 + }, + { + "epoch": 0.6614961446543717, + "grad_norm": 1.5541031088970565, + "learning_rate": 1.0861678895398854e-06, + "loss": 0.9677, + "step": 7335 + }, + { + "epoch": 0.6615863281778419, + "grad_norm": 1.7388934139779186, + "learning_rate": 1.0856482674435286e-06, + "loss": 0.9146, + "step": 7336 + }, + { + "epoch": 0.6616765117013121, + "grad_norm": 0.7107257318020007, + "learning_rate": 1.0851287233647024e-06, + "loss": 0.7772, + "step": 7337 + }, + { + "epoch": 0.6617666952247825, + "grad_norm": 1.5852866505042449, + "learning_rate": 1.084609257347738e-06, + "loss": 1.0097, + "step": 7338 + }, + { + "epoch": 0.6618568787482527, + "grad_norm": 1.7933835815045693, + "learning_rate": 1.0840898694369594e-06, + "loss": 0.9268, + "step": 7339 + }, + { + "epoch": 0.661947062271723, + "grad_norm": 1.4547500547732566, + "learning_rate": 1.083570559676683e-06, + "loss": 0.9411, + "step": 7340 + }, + { + "epoch": 0.6620372457951932, + "grad_norm": 1.4971088521417315, + "learning_rate": 1.08305132811122e-06, + "loss": 0.9309, + "step": 7341 + }, + { + "epoch": 0.6621274293186635, + "grad_norm": 1.5505457009792376, + "learning_rate": 1.0825321747848735e-06, + "loss": 0.9508, + "step": 7342 + }, + { + "epoch": 0.6622176128421338, + "grad_norm": 0.7687307541461755, + "learning_rate": 1.0820130997419417e-06, + "loss": 0.8461, + "step": 7343 + }, + { + "epoch": 0.662307796365604, + "grad_norm": 1.9646295085616665, + "learning_rate": 1.0814941030267123e-06, + "loss": 0.948, + "step": 7344 + }, + { + "epoch": 0.6623979798890742, + "grad_norm": 0.6548839838715456, + "learning_rate": 1.080975184683472e-06, + "loss": 0.7638, + "step": 7345 + }, + { + "epoch": 0.6624881634125446, + "grad_norm": 1.5504791964919338, + "learning_rate": 1.0804563447564948e-06, + "loss": 1.0084, + "step": 7346 + }, + { + "epoch": 0.6625783469360148, + "grad_norm": 1.597584468656029, + "learning_rate": 1.0799375832900545e-06, + "loss": 0.8974, + "step": 7347 + }, + { + "epoch": 0.662668530459485, + "grad_norm": 1.5709892826261096, + "learning_rate": 1.0794189003284118e-06, + "loss": 0.9867, + "step": 7348 + }, + { + "epoch": 0.6627587139829553, + "grad_norm": 1.507248768155355, + "learning_rate": 1.0789002959158242e-06, + "loss": 0.9731, + "step": 7349 + }, + { + "epoch": 0.6628488975064256, + "grad_norm": 1.9679482053477104, + "learning_rate": 1.0783817700965428e-06, + "loss": 0.9036, + "step": 7350 + }, + { + "epoch": 0.6629390810298958, + "grad_norm": 1.4904014723702415, + "learning_rate": 1.0778633229148102e-06, + "loss": 0.9983, + "step": 7351 + }, + { + "epoch": 0.6630292645533661, + "grad_norm": 1.8330576284769737, + "learning_rate": 1.0773449544148645e-06, + "loss": 0.9558, + "step": 7352 + }, + { + "epoch": 0.6631194480768363, + "grad_norm": 3.959596666545565, + "learning_rate": 1.076826664640934e-06, + "loss": 0.9935, + "step": 7353 + }, + { + "epoch": 0.6632096316003067, + "grad_norm": 0.781240234313964, + "learning_rate": 1.0763084536372424e-06, + "loss": 0.8054, + "step": 7354 + }, + { + "epoch": 0.6632998151237769, + "grad_norm": 1.5440393524195128, + "learning_rate": 1.0757903214480068e-06, + "loss": 0.9659, + "step": 7355 + }, + { + "epoch": 0.6633899986472471, + "grad_norm": 1.6570442742353395, + "learning_rate": 1.0752722681174376e-06, + "loss": 0.8958, + "step": 7356 + }, + { + "epoch": 0.6634801821707174, + "grad_norm": 2.0655462714585053, + "learning_rate": 1.074754293689737e-06, + "loss": 0.8701, + "step": 7357 + }, + { + "epoch": 0.6635703656941877, + "grad_norm": 1.6018255994593713, + "learning_rate": 1.0742363982091023e-06, + "loss": 0.9592, + "step": 7358 + }, + { + "epoch": 0.6636605492176579, + "grad_norm": 1.7695541548772835, + "learning_rate": 1.0737185817197215e-06, + "loss": 0.9103, + "step": 7359 + }, + { + "epoch": 0.6637507327411282, + "grad_norm": 1.6714149626502737, + "learning_rate": 1.0732008442657803e-06, + "loss": 0.9133, + "step": 7360 + }, + { + "epoch": 0.6638409162645985, + "grad_norm": 1.4242531308258073, + "learning_rate": 1.0726831858914516e-06, + "loss": 0.9318, + "step": 7361 + }, + { + "epoch": 0.6639310997880687, + "grad_norm": 1.9059676914979267, + "learning_rate": 1.0721656066409084e-06, + "loss": 0.9725, + "step": 7362 + }, + { + "epoch": 0.664021283311539, + "grad_norm": 1.781909820419799, + "learning_rate": 1.0716481065583108e-06, + "loss": 0.9617, + "step": 7363 + }, + { + "epoch": 0.6641114668350092, + "grad_norm": 0.668284068499676, + "learning_rate": 1.071130685687816e-06, + "loss": 0.7755, + "step": 7364 + }, + { + "epoch": 0.6642016503584796, + "grad_norm": 1.421459514571397, + "learning_rate": 1.0706133440735723e-06, + "loss": 0.888, + "step": 7365 + }, + { + "epoch": 0.6642918338819498, + "grad_norm": 1.7460808465468065, + "learning_rate": 1.070096081759723e-06, + "loss": 0.9031, + "step": 7366 + }, + { + "epoch": 0.66438201740542, + "grad_norm": 1.715336340017361, + "learning_rate": 1.069578898790404e-06, + "loss": 1.0501, + "step": 7367 + }, + { + "epoch": 0.6644722009288903, + "grad_norm": 1.711556170362881, + "learning_rate": 1.0690617952097424e-06, + "loss": 0.933, + "step": 7368 + }, + { + "epoch": 0.6645623844523606, + "grad_norm": 2.833139599927222, + "learning_rate": 1.068544771061863e-06, + "loss": 0.9996, + "step": 7369 + }, + { + "epoch": 0.6646525679758308, + "grad_norm": 1.6973222343389591, + "learning_rate": 1.0680278263908787e-06, + "loss": 0.996, + "step": 7370 + }, + { + "epoch": 0.6647427514993011, + "grad_norm": 0.7805645033375169, + "learning_rate": 1.0675109612408991e-06, + "loss": 0.7866, + "step": 7371 + }, + { + "epoch": 0.6648329350227713, + "grad_norm": 1.706318232810367, + "learning_rate": 1.0669941756560264e-06, + "loss": 0.9191, + "step": 7372 + }, + { + "epoch": 0.6649231185462416, + "grad_norm": 1.9157374935506055, + "learning_rate": 1.0664774696803548e-06, + "loss": 0.8984, + "step": 7373 + }, + { + "epoch": 0.6650133020697119, + "grad_norm": 1.5603275455274335, + "learning_rate": 1.065960843357973e-06, + "loss": 0.922, + "step": 7374 + }, + { + "epoch": 0.6651034855931821, + "grad_norm": 1.676085971256974, + "learning_rate": 1.065444296732963e-06, + "loss": 0.9644, + "step": 7375 + }, + { + "epoch": 0.6651936691166523, + "grad_norm": 0.7126760917699906, + "learning_rate": 1.064927829849397e-06, + "loss": 0.7883, + "step": 7376 + }, + { + "epoch": 0.6652838526401227, + "grad_norm": 1.8003801076941464, + "learning_rate": 1.0644114427513465e-06, + "loss": 0.9453, + "step": 7377 + }, + { + "epoch": 0.6653740361635929, + "grad_norm": 1.3812205747344053, + "learning_rate": 1.0638951354828693e-06, + "loss": 0.878, + "step": 7378 + }, + { + "epoch": 0.6654642196870632, + "grad_norm": 1.739446493770519, + "learning_rate": 1.063378908088021e-06, + "loss": 1.0079, + "step": 7379 + }, + { + "epoch": 0.6655544032105334, + "grad_norm": 1.5661837070132512, + "learning_rate": 1.0628627606108486e-06, + "loss": 0.9439, + "step": 7380 + }, + { + "epoch": 0.6656445867340037, + "grad_norm": 1.7278026425016102, + "learning_rate": 1.062346693095393e-06, + "loss": 1.0147, + "step": 7381 + }, + { + "epoch": 0.665734770257474, + "grad_norm": 1.7195342615457356, + "learning_rate": 1.0618307055856882e-06, + "loss": 0.8886, + "step": 7382 + }, + { + "epoch": 0.6658249537809442, + "grad_norm": 1.495223068385101, + "learning_rate": 1.061314798125759e-06, + "loss": 0.9996, + "step": 7383 + }, + { + "epoch": 0.6659151373044144, + "grad_norm": 1.641806839833956, + "learning_rate": 1.0607989707596293e-06, + "loss": 0.9082, + "step": 7384 + }, + { + "epoch": 0.6660053208278848, + "grad_norm": 1.3668269962765018, + "learning_rate": 1.0602832235313078e-06, + "loss": 1.0745, + "step": 7385 + }, + { + "epoch": 0.666095504351355, + "grad_norm": 1.3985348886005216, + "learning_rate": 1.0597675564848053e-06, + "loss": 0.9412, + "step": 7386 + }, + { + "epoch": 0.6661856878748252, + "grad_norm": 1.5412715800832726, + "learning_rate": 1.059251969664118e-06, + "loss": 1.0214, + "step": 7387 + }, + { + "epoch": 0.6662758713982956, + "grad_norm": 1.4416539255949992, + "learning_rate": 1.0587364631132402e-06, + "loss": 0.9593, + "step": 7388 + }, + { + "epoch": 0.6663660549217658, + "grad_norm": 1.4685413942928913, + "learning_rate": 1.0582210368761573e-06, + "loss": 0.9885, + "step": 7389 + }, + { + "epoch": 0.666456238445236, + "grad_norm": 1.9538831536332355, + "learning_rate": 1.0577056909968485e-06, + "loss": 0.9516, + "step": 7390 + }, + { + "epoch": 0.6665464219687063, + "grad_norm": 1.7098974262950153, + "learning_rate": 1.0571904255192857e-06, + "loss": 0.9682, + "step": 7391 + }, + { + "epoch": 0.6666366054921766, + "grad_norm": 2.0720884371408976, + "learning_rate": 1.0566752404874354e-06, + "loss": 0.983, + "step": 7392 + }, + { + "epoch": 0.6667267890156469, + "grad_norm": 1.9830232957596006, + "learning_rate": 1.0561601359452543e-06, + "loss": 0.9427, + "step": 7393 + }, + { + "epoch": 0.6668169725391171, + "grad_norm": 1.5550059083538104, + "learning_rate": 1.0556451119366947e-06, + "loss": 0.8858, + "step": 7394 + }, + { + "epoch": 0.6669071560625873, + "grad_norm": 1.4595463294804247, + "learning_rate": 1.0551301685057011e-06, + "loss": 0.8905, + "step": 7395 + }, + { + "epoch": 0.6669973395860577, + "grad_norm": 1.9076899108741507, + "learning_rate": 1.0546153056962117e-06, + "loss": 1.0326, + "step": 7396 + }, + { + "epoch": 0.6670875231095279, + "grad_norm": 1.7708824001321082, + "learning_rate": 1.0541005235521578e-06, + "loss": 0.9472, + "step": 7397 + }, + { + "epoch": 0.6671777066329981, + "grad_norm": 1.5506018239341586, + "learning_rate": 1.0535858221174614e-06, + "loss": 0.9611, + "step": 7398 + }, + { + "epoch": 0.6672678901564684, + "grad_norm": 1.2582435580095703, + "learning_rate": 1.0530712014360426e-06, + "loss": 0.9872, + "step": 7399 + }, + { + "epoch": 0.6673580736799387, + "grad_norm": 1.5320594360933444, + "learning_rate": 1.0525566615518088e-06, + "loss": 1.0008, + "step": 7400 + }, + { + "epoch": 0.667448257203409, + "grad_norm": 1.3747524125264794, + "learning_rate": 1.0520422025086662e-06, + "loss": 0.9656, + "step": 7401 + }, + { + "epoch": 0.6675384407268792, + "grad_norm": 1.9931890027262817, + "learning_rate": 1.0515278243505092e-06, + "loss": 1.0004, + "step": 7402 + }, + { + "epoch": 0.6676286242503494, + "grad_norm": 1.340487711886039, + "learning_rate": 1.0510135271212278e-06, + "loss": 0.983, + "step": 7403 + }, + { + "epoch": 0.6677188077738198, + "grad_norm": 1.606704347822855, + "learning_rate": 1.0504993108647052e-06, + "loss": 1.0172, + "step": 7404 + }, + { + "epoch": 0.66780899129729, + "grad_norm": 1.4556069456416438, + "learning_rate": 1.0499851756248168e-06, + "loss": 0.9238, + "step": 7405 + }, + { + "epoch": 0.6678991748207602, + "grad_norm": 1.5247714731305892, + "learning_rate": 1.0494711214454316e-06, + "loss": 0.931, + "step": 7406 + }, + { + "epoch": 0.6679893583442305, + "grad_norm": 1.5416430222356317, + "learning_rate": 1.0489571483704111e-06, + "loss": 0.867, + "step": 7407 + }, + { + "epoch": 0.6680795418677008, + "grad_norm": 1.7262262215381927, + "learning_rate": 1.048443256443612e-06, + "loss": 0.8697, + "step": 7408 + }, + { + "epoch": 0.668169725391171, + "grad_norm": 1.5727484488903298, + "learning_rate": 1.0479294457088801e-06, + "loss": 0.9704, + "step": 7409 + }, + { + "epoch": 0.6682599089146413, + "grad_norm": 1.4203932031043978, + "learning_rate": 1.0474157162100574e-06, + "loss": 1.0724, + "step": 7410 + }, + { + "epoch": 0.6683500924381116, + "grad_norm": 1.5958881621076084, + "learning_rate": 1.0469020679909786e-06, + "loss": 0.8879, + "step": 7411 + }, + { + "epoch": 0.6684402759615818, + "grad_norm": 1.532928558839895, + "learning_rate": 1.0463885010954705e-06, + "loss": 0.8737, + "step": 7412 + }, + { + "epoch": 0.6685304594850521, + "grad_norm": 1.7016432029284612, + "learning_rate": 1.0458750155673536e-06, + "loss": 1.0196, + "step": 7413 + }, + { + "epoch": 0.6686206430085223, + "grad_norm": 1.875954893464617, + "learning_rate": 1.0453616114504421e-06, + "loss": 0.9049, + "step": 7414 + }, + { + "epoch": 0.6687108265319927, + "grad_norm": 1.4650711086578945, + "learning_rate": 1.0448482887885406e-06, + "loss": 0.9836, + "step": 7415 + }, + { + "epoch": 0.6688010100554629, + "grad_norm": 1.74013278377796, + "learning_rate": 1.044335047625451e-06, + "loss": 0.9558, + "step": 7416 + }, + { + "epoch": 0.6688911935789331, + "grad_norm": 1.692833419481817, + "learning_rate": 1.0438218880049637e-06, + "loss": 0.9599, + "step": 7417 + }, + { + "epoch": 0.6689813771024034, + "grad_norm": 1.7416138666840653, + "learning_rate": 1.0433088099708653e-06, + "loss": 1.0085, + "step": 7418 + }, + { + "epoch": 0.6690715606258737, + "grad_norm": 0.6321370558476359, + "learning_rate": 1.0427958135669346e-06, + "loss": 0.7139, + "step": 7419 + }, + { + "epoch": 0.6691617441493439, + "grad_norm": 1.6087553442426834, + "learning_rate": 1.0422828988369428e-06, + "loss": 0.8873, + "step": 7420 + }, + { + "epoch": 0.6692519276728142, + "grad_norm": 1.4305693194684994, + "learning_rate": 1.041770065824655e-06, + "loss": 1.02, + "step": 7421 + }, + { + "epoch": 0.6693421111962844, + "grad_norm": 1.3798622945136132, + "learning_rate": 1.0412573145738287e-06, + "loss": 0.8077, + "step": 7422 + }, + { + "epoch": 0.6694322947197547, + "grad_norm": 1.7283968609887972, + "learning_rate": 1.040744645128216e-06, + "loss": 0.9205, + "step": 7423 + }, + { + "epoch": 0.669522478243225, + "grad_norm": 1.6364312477063767, + "learning_rate": 1.040232057531558e-06, + "loss": 1.008, + "step": 7424 + }, + { + "epoch": 0.6696126617666952, + "grad_norm": 2.0435001418876926, + "learning_rate": 1.0397195518275932e-06, + "loss": 0.9227, + "step": 7425 + }, + { + "epoch": 0.6697028452901654, + "grad_norm": 1.4414466914264443, + "learning_rate": 1.0392071280600512e-06, + "loss": 0.9229, + "step": 7426 + }, + { + "epoch": 0.6697930288136358, + "grad_norm": 1.4856250518683272, + "learning_rate": 1.0386947862726549e-06, + "loss": 0.951, + "step": 7427 + }, + { + "epoch": 0.669883212337106, + "grad_norm": 1.791369886732134, + "learning_rate": 1.0381825265091197e-06, + "loss": 0.8737, + "step": 7428 + }, + { + "epoch": 0.6699733958605762, + "grad_norm": 3.6766138542534725, + "learning_rate": 1.037670348813155e-06, + "loss": 0.8673, + "step": 7429 + }, + { + "epoch": 0.6700635793840465, + "grad_norm": 1.5259317178360796, + "learning_rate": 1.0371582532284624e-06, + "loss": 0.9787, + "step": 7430 + }, + { + "epoch": 0.6701537629075168, + "grad_norm": 1.4961142277125405, + "learning_rate": 1.0366462397987375e-06, + "loss": 0.8834, + "step": 7431 + }, + { + "epoch": 0.6702439464309871, + "grad_norm": 2.981714310866891, + "learning_rate": 1.0361343085676665e-06, + "loss": 0.9605, + "step": 7432 + }, + { + "epoch": 0.6703341299544573, + "grad_norm": 1.8329979271843195, + "learning_rate": 1.0356224595789309e-06, + "loss": 1.0662, + "step": 7433 + }, + { + "epoch": 0.6704243134779276, + "grad_norm": 1.4399081252512012, + "learning_rate": 1.0351106928762046e-06, + "loss": 1.0106, + "step": 7434 + }, + { + "epoch": 0.6705144970013979, + "grad_norm": 1.6019481659788979, + "learning_rate": 1.034599008503154e-06, + "loss": 1.0276, + "step": 7435 + }, + { + "epoch": 0.6706046805248681, + "grad_norm": 1.4669538838029605, + "learning_rate": 1.0340874065034406e-06, + "loss": 0.9416, + "step": 7436 + }, + { + "epoch": 0.6706948640483383, + "grad_norm": 1.5532474461879393, + "learning_rate": 1.0335758869207137e-06, + "loss": 0.8791, + "step": 7437 + }, + { + "epoch": 0.6707850475718087, + "grad_norm": 1.962896941065709, + "learning_rate": 1.0330644497986227e-06, + "loss": 0.9676, + "step": 7438 + }, + { + "epoch": 0.6708752310952789, + "grad_norm": 1.6491480737646513, + "learning_rate": 1.0325530951808029e-06, + "loss": 0.9946, + "step": 7439 + }, + { + "epoch": 0.6709654146187491, + "grad_norm": 1.599331823543606, + "learning_rate": 1.0320418231108887e-06, + "loss": 0.9597, + "step": 7440 + }, + { + "epoch": 0.6710555981422194, + "grad_norm": 1.6438070773233993, + "learning_rate": 1.0315306336325028e-06, + "loss": 0.9944, + "step": 7441 + }, + { + "epoch": 0.6711457816656897, + "grad_norm": 1.8798511056579346, + "learning_rate": 1.0310195267892635e-06, + "loss": 0.9134, + "step": 7442 + }, + { + "epoch": 0.67123596518916, + "grad_norm": 1.58241369075349, + "learning_rate": 1.030508502624781e-06, + "loss": 0.8916, + "step": 7443 + }, + { + "epoch": 0.6713261487126302, + "grad_norm": 1.5898511731773493, + "learning_rate": 1.0299975611826587e-06, + "loss": 0.9006, + "step": 7444 + }, + { + "epoch": 0.6714163322361004, + "grad_norm": 1.4206598140480278, + "learning_rate": 1.0294867025064928e-06, + "loss": 0.9581, + "step": 7445 + }, + { + "epoch": 0.6715065157595708, + "grad_norm": 1.3563355414490637, + "learning_rate": 1.028975926639874e-06, + "loss": 0.9595, + "step": 7446 + }, + { + "epoch": 0.671596699283041, + "grad_norm": 2.434871919619924, + "learning_rate": 1.0284652336263823e-06, + "loss": 0.8745, + "step": 7447 + }, + { + "epoch": 0.6716868828065112, + "grad_norm": 1.5576768821829223, + "learning_rate": 1.0279546235095938e-06, + "loss": 0.8818, + "step": 7448 + }, + { + "epoch": 0.6717770663299815, + "grad_norm": 0.7580886023228833, + "learning_rate": 1.0274440963330768e-06, + "loss": 0.8099, + "step": 7449 + }, + { + "epoch": 0.6718672498534518, + "grad_norm": 1.453494445433237, + "learning_rate": 1.0269336521403919e-06, + "loss": 0.9506, + "step": 7450 + }, + { + "epoch": 0.671957433376922, + "grad_norm": 1.556967745192659, + "learning_rate": 1.0264232909750936e-06, + "loss": 1.0796, + "step": 7451 + }, + { + "epoch": 0.6720476169003923, + "grad_norm": 1.5099162708841065, + "learning_rate": 1.025913012880728e-06, + "loss": 1.0125, + "step": 7452 + }, + { + "epoch": 0.6721378004238625, + "grad_norm": 1.4358358286996251, + "learning_rate": 1.0254028179008362e-06, + "loss": 0.8606, + "step": 7453 + }, + { + "epoch": 0.6722279839473329, + "grad_norm": 1.6339197374472836, + "learning_rate": 1.0248927060789483e-06, + "loss": 0.8437, + "step": 7454 + }, + { + "epoch": 0.6723181674708031, + "grad_norm": 1.5040967151864415, + "learning_rate": 1.0243826774585928e-06, + "loss": 0.909, + "step": 7455 + }, + { + "epoch": 0.6724083509942733, + "grad_norm": 1.8589801529373282, + "learning_rate": 1.0238727320832854e-06, + "loss": 0.9398, + "step": 7456 + }, + { + "epoch": 0.6724985345177437, + "grad_norm": 1.7820028086210953, + "learning_rate": 1.0233628699965403e-06, + "loss": 0.9555, + "step": 7457 + }, + { + "epoch": 0.6725887180412139, + "grad_norm": 1.5763005200608378, + "learning_rate": 1.0228530912418594e-06, + "loss": 0.9739, + "step": 7458 + }, + { + "epoch": 0.6726789015646841, + "grad_norm": 1.650020255340169, + "learning_rate": 1.0223433958627404e-06, + "loss": 0.9251, + "step": 7459 + }, + { + "epoch": 0.6727690850881544, + "grad_norm": 2.0395884803756386, + "learning_rate": 1.021833783902674e-06, + "loss": 0.9752, + "step": 7460 + }, + { + "epoch": 0.6728592686116247, + "grad_norm": 1.4736044181378514, + "learning_rate": 1.0213242554051427e-06, + "loss": 0.9529, + "step": 7461 + }, + { + "epoch": 0.6729494521350949, + "grad_norm": 1.569796197383074, + "learning_rate": 1.0208148104136229e-06, + "loss": 0.8937, + "step": 7462 + }, + { + "epoch": 0.6730396356585652, + "grad_norm": 2.124583932871889, + "learning_rate": 1.020305448971582e-06, + "loss": 0.9258, + "step": 7463 + }, + { + "epoch": 0.6731298191820354, + "grad_norm": 1.543248868214192, + "learning_rate": 1.0197961711224824e-06, + "loss": 0.9175, + "step": 7464 + }, + { + "epoch": 0.6732200027055057, + "grad_norm": 1.6827256515888187, + "learning_rate": 1.0192869769097777e-06, + "loss": 0.9611, + "step": 7465 + }, + { + "epoch": 0.673310186228976, + "grad_norm": 1.5528421616480212, + "learning_rate": 1.018777866376916e-06, + "loss": 0.9614, + "step": 7466 + }, + { + "epoch": 0.6734003697524462, + "grad_norm": 1.7386478337562898, + "learning_rate": 1.0182688395673374e-06, + "loss": 0.8221, + "step": 7467 + }, + { + "epoch": 0.6734905532759164, + "grad_norm": 1.482222190545873, + "learning_rate": 1.017759896524475e-06, + "loss": 0.9172, + "step": 7468 + }, + { + "epoch": 0.6735807367993868, + "grad_norm": 1.9418238108272161, + "learning_rate": 1.0172510372917528e-06, + "loss": 0.9885, + "step": 7469 + }, + { + "epoch": 0.673670920322857, + "grad_norm": 1.9539553898821267, + "learning_rate": 1.0167422619125925e-06, + "loss": 0.8574, + "step": 7470 + }, + { + "epoch": 0.6737611038463273, + "grad_norm": 1.7622769776053167, + "learning_rate": 1.0162335704304026e-06, + "loss": 0.9585, + "step": 7471 + }, + { + "epoch": 0.6738512873697975, + "grad_norm": 0.7630612793358922, + "learning_rate": 1.0157249628885903e-06, + "loss": 0.7935, + "step": 7472 + }, + { + "epoch": 0.6739414708932678, + "grad_norm": 1.3247714169382339, + "learning_rate": 1.0152164393305506e-06, + "loss": 0.8612, + "step": 7473 + }, + { + "epoch": 0.6740316544167381, + "grad_norm": 1.6589948833034864, + "learning_rate": 1.0147079997996746e-06, + "loss": 0.9471, + "step": 7474 + }, + { + "epoch": 0.6741218379402083, + "grad_norm": 1.6063204916837655, + "learning_rate": 1.0141996443393446e-06, + "loss": 0.9807, + "step": 7475 + }, + { + "epoch": 0.6742120214636785, + "grad_norm": 1.5651674009255119, + "learning_rate": 1.0136913729929369e-06, + "loss": 0.9386, + "step": 7476 + }, + { + "epoch": 0.6743022049871489, + "grad_norm": 1.4899374089759663, + "learning_rate": 1.0131831858038203e-06, + "loss": 0.9077, + "step": 7477 + }, + { + "epoch": 0.6743923885106191, + "grad_norm": 1.3880913187515511, + "learning_rate": 1.0126750828153538e-06, + "loss": 1.0122, + "step": 7478 + }, + { + "epoch": 0.6744825720340893, + "grad_norm": 1.65058621483402, + "learning_rate": 1.012167064070895e-06, + "loss": 0.9927, + "step": 7479 + }, + { + "epoch": 0.6745727555575597, + "grad_norm": 2.0950351016579543, + "learning_rate": 1.0116591296137885e-06, + "loss": 0.9774, + "step": 7480 + }, + { + "epoch": 0.6746629390810299, + "grad_norm": 1.4886395681374531, + "learning_rate": 1.0111512794873746e-06, + "loss": 0.9811, + "step": 7481 + }, + { + "epoch": 0.6747531226045002, + "grad_norm": 1.512944712559692, + "learning_rate": 1.010643513734986e-06, + "loss": 0.9518, + "step": 7482 + }, + { + "epoch": 0.6748433061279704, + "grad_norm": 1.546674850307935, + "learning_rate": 1.010135832399948e-06, + "loss": 0.9925, + "step": 7483 + }, + { + "epoch": 0.6749334896514407, + "grad_norm": 1.3027841995982916, + "learning_rate": 1.0096282355255792e-06, + "loss": 0.961, + "step": 7484 + }, + { + "epoch": 0.675023673174911, + "grad_norm": 1.8603203558119148, + "learning_rate": 1.0091207231551905e-06, + "loss": 1.0038, + "step": 7485 + }, + { + "epoch": 0.6751138566983812, + "grad_norm": 1.4630447937997493, + "learning_rate": 1.0086132953320842e-06, + "loss": 0.9178, + "step": 7486 + }, + { + "epoch": 0.6752040402218514, + "grad_norm": 1.6748334431751504, + "learning_rate": 1.0081059520995591e-06, + "loss": 0.9802, + "step": 7487 + }, + { + "epoch": 0.6752942237453218, + "grad_norm": 1.6843258003027657, + "learning_rate": 1.0075986935009028e-06, + "loss": 1.0201, + "step": 7488 + }, + { + "epoch": 0.675384407268792, + "grad_norm": 1.388317121859112, + "learning_rate": 1.0070915195793982e-06, + "loss": 0.9626, + "step": 7489 + }, + { + "epoch": 0.6754745907922622, + "grad_norm": 1.6647699132949383, + "learning_rate": 1.0065844303783197e-06, + "loss": 0.9282, + "step": 7490 + }, + { + "epoch": 0.6755647743157325, + "grad_norm": 1.5060544215446854, + "learning_rate": 1.0060774259409356e-06, + "loss": 0.8705, + "step": 7491 + }, + { + "epoch": 0.6756549578392028, + "grad_norm": 1.3631615791152794, + "learning_rate": 1.0055705063105065e-06, + "loss": 0.987, + "step": 7492 + }, + { + "epoch": 0.675745141362673, + "grad_norm": 1.4562077790389814, + "learning_rate": 1.0050636715302837e-06, + "loss": 0.9462, + "step": 7493 + }, + { + "epoch": 0.6758353248861433, + "grad_norm": 1.7963388057622316, + "learning_rate": 1.0045569216435157e-06, + "loss": 0.9628, + "step": 7494 + }, + { + "epoch": 0.6759255084096135, + "grad_norm": 1.4967569420116558, + "learning_rate": 1.0040502566934384e-06, + "loss": 1.0344, + "step": 7495 + }, + { + "epoch": 0.6760156919330839, + "grad_norm": 1.7408423597954885, + "learning_rate": 1.0035436767232866e-06, + "loss": 0.8887, + "step": 7496 + }, + { + "epoch": 0.6761058754565541, + "grad_norm": 1.5451326672527577, + "learning_rate": 1.0030371817762816e-06, + "loss": 0.9669, + "step": 7497 + }, + { + "epoch": 0.6761960589800243, + "grad_norm": 1.499795820326983, + "learning_rate": 1.0025307718956417e-06, + "loss": 0.9535, + "step": 7498 + }, + { + "epoch": 0.6762862425034946, + "grad_norm": 1.5234340569872953, + "learning_rate": 1.0020244471245765e-06, + "loss": 0.9713, + "step": 7499 + }, + { + "epoch": 0.6763764260269649, + "grad_norm": 1.501931456966386, + "learning_rate": 1.001518207506288e-06, + "loss": 0.9049, + "step": 7500 + }, + { + "epoch": 0.6764666095504351, + "grad_norm": 1.417577600255655, + "learning_rate": 1.0010120530839717e-06, + "loss": 1.0153, + "step": 7501 + }, + { + "epoch": 0.6765567930739054, + "grad_norm": 1.2272188896514633, + "learning_rate": 1.0005059839008161e-06, + "loss": 0.974, + "step": 7502 + }, + { + "epoch": 0.6766469765973756, + "grad_norm": 1.677409912282762, + "learning_rate": 1.0000000000000004e-06, + "loss": 0.9079, + "step": 7503 + }, + { + "epoch": 0.676737160120846, + "grad_norm": 1.5872764760445544, + "learning_rate": 9.994941014246985e-07, + "loss": 0.9986, + "step": 7504 + }, + { + "epoch": 0.6768273436443162, + "grad_norm": 1.465242702833697, + "learning_rate": 9.989882882180766e-07, + "loss": 0.8439, + "step": 7505 + }, + { + "epoch": 0.6769175271677864, + "grad_norm": 1.5614267859687414, + "learning_rate": 9.984825604232938e-07, + "loss": 0.9695, + "step": 7506 + }, + { + "epoch": 0.6770077106912568, + "grad_norm": 1.657931356097339, + "learning_rate": 9.97976918083502e-07, + "loss": 0.8915, + "step": 7507 + }, + { + "epoch": 0.677097894214727, + "grad_norm": 1.2859674461836283, + "learning_rate": 9.974713612418427e-07, + "loss": 1.0241, + "step": 7508 + }, + { + "epoch": 0.6771880777381972, + "grad_norm": 1.6783752283261082, + "learning_rate": 9.969658899414563e-07, + "loss": 0.9691, + "step": 7509 + }, + { + "epoch": 0.6772782612616675, + "grad_norm": 1.6767775057312155, + "learning_rate": 9.964605042254696e-07, + "loss": 0.8758, + "step": 7510 + }, + { + "epoch": 0.6773684447851378, + "grad_norm": 1.6786002965375315, + "learning_rate": 9.959552041370076e-07, + "loss": 0.8926, + "step": 7511 + }, + { + "epoch": 0.677458628308608, + "grad_norm": 1.4249231552017043, + "learning_rate": 9.954499897191824e-07, + "loss": 0.8713, + "step": 7512 + }, + { + "epoch": 0.6775488118320783, + "grad_norm": 2.0832560588492663, + "learning_rate": 9.949448610151043e-07, + "loss": 0.9083, + "step": 7513 + }, + { + "epoch": 0.6776389953555485, + "grad_norm": 1.5070019851144982, + "learning_rate": 9.944398180678719e-07, + "loss": 0.9325, + "step": 7514 + }, + { + "epoch": 0.6777291788790188, + "grad_norm": 1.3171770685943567, + "learning_rate": 9.939348609205789e-07, + "loss": 0.9021, + "step": 7515 + }, + { + "epoch": 0.6778193624024891, + "grad_norm": 1.4354708074021598, + "learning_rate": 9.93429989616311e-07, + "loss": 0.9617, + "step": 7516 + }, + { + "epoch": 0.6779095459259593, + "grad_norm": 1.617155212964607, + "learning_rate": 9.929252041981464e-07, + "loss": 0.8683, + "step": 7517 + }, + { + "epoch": 0.6779997294494295, + "grad_norm": 1.4791879876261256, + "learning_rate": 9.924205047091572e-07, + "loss": 0.9231, + "step": 7518 + }, + { + "epoch": 0.6780899129728999, + "grad_norm": 1.6357937841675794, + "learning_rate": 9.919158911924056e-07, + "loss": 1.0725, + "step": 7519 + }, + { + "epoch": 0.6781800964963701, + "grad_norm": 1.4715308789387629, + "learning_rate": 9.914113636909483e-07, + "loss": 0.9853, + "step": 7520 + }, + { + "epoch": 0.6782702800198404, + "grad_norm": 0.7759091735659396, + "learning_rate": 9.90906922247835e-07, + "loss": 0.8286, + "step": 7521 + }, + { + "epoch": 0.6783604635433106, + "grad_norm": 1.371508934868035, + "learning_rate": 9.904025669061072e-07, + "loss": 0.979, + "step": 7522 + }, + { + "epoch": 0.6784506470667809, + "grad_norm": 1.3220386394821266, + "learning_rate": 9.89898297708799e-07, + "loss": 0.9482, + "step": 7523 + }, + { + "epoch": 0.6785408305902512, + "grad_norm": 1.9767759908943323, + "learning_rate": 9.893941146989388e-07, + "loss": 0.9036, + "step": 7524 + }, + { + "epoch": 0.6786310141137214, + "grad_norm": 1.778704514287156, + "learning_rate": 9.888900179195437e-07, + "loss": 0.9119, + "step": 7525 + }, + { + "epoch": 0.6787211976371916, + "grad_norm": 1.6165069032492758, + "learning_rate": 9.883860074136285e-07, + "loss": 1.0247, + "step": 7526 + }, + { + "epoch": 0.678811381160662, + "grad_norm": 1.8381309230570226, + "learning_rate": 9.87882083224196e-07, + "loss": 0.9471, + "step": 7527 + }, + { + "epoch": 0.6789015646841322, + "grad_norm": 0.7287979601590726, + "learning_rate": 9.873782453942462e-07, + "loss": 0.8063, + "step": 7528 + }, + { + "epoch": 0.6789917482076024, + "grad_norm": 1.7768552003898532, + "learning_rate": 9.868744939667676e-07, + "loss": 0.8828, + "step": 7529 + }, + { + "epoch": 0.6790819317310728, + "grad_norm": 1.6172994676243189, + "learning_rate": 9.863708289847432e-07, + "loss": 0.8335, + "step": 7530 + }, + { + "epoch": 0.679172115254543, + "grad_norm": 1.9500141241100064, + "learning_rate": 9.85867250491149e-07, + "loss": 0.8626, + "step": 7531 + }, + { + "epoch": 0.6792622987780133, + "grad_norm": 1.6849830365105303, + "learning_rate": 9.853637585289528e-07, + "loss": 1.0601, + "step": 7532 + }, + { + "epoch": 0.6793524823014835, + "grad_norm": 1.5363769337272066, + "learning_rate": 9.848603531411159e-07, + "loss": 0.9449, + "step": 7533 + }, + { + "epoch": 0.6794426658249538, + "grad_norm": 1.9634334895305363, + "learning_rate": 9.843570343705899e-07, + "loss": 0.96, + "step": 7534 + }, + { + "epoch": 0.6795328493484241, + "grad_norm": 1.2857987673191196, + "learning_rate": 9.83853802260323e-07, + "loss": 1.0239, + "step": 7535 + }, + { + "epoch": 0.6796230328718943, + "grad_norm": 1.7133832895044032, + "learning_rate": 9.833506568532524e-07, + "loss": 0.8961, + "step": 7536 + }, + { + "epoch": 0.6797132163953645, + "grad_norm": 1.3479566598626163, + "learning_rate": 9.828475981923093e-07, + "loss": 0.7957, + "step": 7537 + }, + { + "epoch": 0.6798033999188349, + "grad_norm": 1.9512563692877245, + "learning_rate": 9.823446263204175e-07, + "loss": 0.9282, + "step": 7538 + }, + { + "epoch": 0.6798935834423051, + "grad_norm": 1.4409304611507352, + "learning_rate": 9.818417412804937e-07, + "loss": 0.9658, + "step": 7539 + }, + { + "epoch": 0.6799837669657753, + "grad_norm": 1.574289097882453, + "learning_rate": 9.813389431154463e-07, + "loss": 1.0353, + "step": 7540 + }, + { + "epoch": 0.6800739504892456, + "grad_norm": 1.760651935754578, + "learning_rate": 9.808362318681783e-07, + "loss": 0.9788, + "step": 7541 + }, + { + "epoch": 0.6801641340127159, + "grad_norm": 1.8163171889918406, + "learning_rate": 9.803336075815807e-07, + "loss": 0.9707, + "step": 7542 + }, + { + "epoch": 0.6802543175361861, + "grad_norm": 1.411591283113325, + "learning_rate": 9.79831070298544e-07, + "loss": 0.8807, + "step": 7543 + }, + { + "epoch": 0.6803445010596564, + "grad_norm": 1.9283477209417725, + "learning_rate": 9.793286200619443e-07, + "loss": 0.923, + "step": 7544 + }, + { + "epoch": 0.6804346845831266, + "grad_norm": 1.3924933751494852, + "learning_rate": 9.78826256914655e-07, + "loss": 1.0118, + "step": 7545 + }, + { + "epoch": 0.680524868106597, + "grad_norm": 1.9666960159094593, + "learning_rate": 9.7832398089954e-07, + "loss": 1.0047, + "step": 7546 + }, + { + "epoch": 0.6806150516300672, + "grad_norm": 1.5667653421345018, + "learning_rate": 9.778217920594565e-07, + "loss": 1.0548, + "step": 7547 + }, + { + "epoch": 0.6807052351535374, + "grad_norm": 1.5780854172983003, + "learning_rate": 9.773196904372547e-07, + "loss": 0.9959, + "step": 7548 + }, + { + "epoch": 0.6807954186770077, + "grad_norm": 1.7288422873930245, + "learning_rate": 9.768176760757742e-07, + "loss": 1.0401, + "step": 7549 + }, + { + "epoch": 0.680885602200478, + "grad_norm": 1.510469299802466, + "learning_rate": 9.76315749017853e-07, + "loss": 0.9292, + "step": 7550 + }, + { + "epoch": 0.6809757857239482, + "grad_norm": 1.7611582471287024, + "learning_rate": 9.758139093063161e-07, + "loss": 0.9562, + "step": 7551 + }, + { + "epoch": 0.6810659692474185, + "grad_norm": 1.584889558981733, + "learning_rate": 9.753121569839834e-07, + "loss": 1.0198, + "step": 7552 + }, + { + "epoch": 0.6811561527708888, + "grad_norm": 1.0391088489721407, + "learning_rate": 9.748104920936678e-07, + "loss": 0.7492, + "step": 7553 + }, + { + "epoch": 0.681246336294359, + "grad_norm": 1.5074349360556296, + "learning_rate": 9.743089146781738e-07, + "loss": 0.9279, + "step": 7554 + }, + { + "epoch": 0.6813365198178293, + "grad_norm": 1.5161759466467557, + "learning_rate": 9.738074247802988e-07, + "loss": 0.8762, + "step": 7555 + }, + { + "epoch": 0.6814267033412995, + "grad_norm": 1.68854250318572, + "learning_rate": 9.733060224428325e-07, + "loss": 0.9721, + "step": 7556 + }, + { + "epoch": 0.6815168868647699, + "grad_norm": 1.6928033498848074, + "learning_rate": 9.728047077085577e-07, + "loss": 0.963, + "step": 7557 + }, + { + "epoch": 0.6816070703882401, + "grad_norm": 1.8517181717341646, + "learning_rate": 9.723034806202497e-07, + "loss": 0.855, + "step": 7558 + }, + { + "epoch": 0.6816972539117103, + "grad_norm": 1.9497387637702241, + "learning_rate": 9.718023412206748e-07, + "loss": 0.9727, + "step": 7559 + }, + { + "epoch": 0.6817874374351806, + "grad_norm": 1.479329498712957, + "learning_rate": 9.713012895525935e-07, + "loss": 0.9748, + "step": 7560 + }, + { + "epoch": 0.6818776209586509, + "grad_norm": 1.621026536398391, + "learning_rate": 9.708003256587584e-07, + "loss": 0.9433, + "step": 7561 + }, + { + "epoch": 0.6819678044821211, + "grad_norm": 1.473691379199841, + "learning_rate": 9.702994495819147e-07, + "loss": 1.0024, + "step": 7562 + }, + { + "epoch": 0.6820579880055914, + "grad_norm": 2.096969406799775, + "learning_rate": 9.697986613647999e-07, + "loss": 0.8605, + "step": 7563 + }, + { + "epoch": 0.6821481715290616, + "grad_norm": 0.72279463809019, + "learning_rate": 9.692979610501425e-07, + "loss": 0.8144, + "step": 7564 + }, + { + "epoch": 0.6822383550525319, + "grad_norm": 1.474869822560539, + "learning_rate": 9.68797348680668e-07, + "loss": 1.0228, + "step": 7565 + }, + { + "epoch": 0.6823285385760022, + "grad_norm": 1.3975104130611185, + "learning_rate": 9.682968242990878e-07, + "loss": 0.9238, + "step": 7566 + }, + { + "epoch": 0.6824187220994724, + "grad_norm": 1.4356008506540674, + "learning_rate": 9.677963879481132e-07, + "loss": 1.0024, + "step": 7567 + }, + { + "epoch": 0.6825089056229426, + "grad_norm": 1.8468987684447051, + "learning_rate": 9.672960396704416e-07, + "loss": 0.9488, + "step": 7568 + }, + { + "epoch": 0.682599089146413, + "grad_norm": 1.38423669610736, + "learning_rate": 9.667957795087657e-07, + "loss": 0.9516, + "step": 7569 + }, + { + "epoch": 0.6826892726698832, + "grad_norm": 1.753777989920702, + "learning_rate": 9.662956075057712e-07, + "loss": 1.0156, + "step": 7570 + }, + { + "epoch": 0.6827794561933535, + "grad_norm": 2.097935947404115, + "learning_rate": 9.657955237041354e-07, + "loss": 1.0204, + "step": 7571 + }, + { + "epoch": 0.6828696397168237, + "grad_norm": 1.4610887744938377, + "learning_rate": 9.652955281465278e-07, + "loss": 0.9328, + "step": 7572 + }, + { + "epoch": 0.682959823240294, + "grad_norm": 1.500566693388252, + "learning_rate": 9.64795620875612e-07, + "loss": 0.9531, + "step": 7573 + }, + { + "epoch": 0.6830500067637643, + "grad_norm": 1.4410777971231985, + "learning_rate": 9.64295801934041e-07, + "loss": 0.8944, + "step": 7574 + }, + { + "epoch": 0.6831401902872345, + "grad_norm": 0.7291248627441319, + "learning_rate": 9.63796071364463e-07, + "loss": 0.7885, + "step": 7575 + }, + { + "epoch": 0.6832303738107048, + "grad_norm": 1.5895706429564478, + "learning_rate": 9.632964292095179e-07, + "loss": 1.0398, + "step": 7576 + }, + { + "epoch": 0.6833205573341751, + "grad_norm": 1.344051948156492, + "learning_rate": 9.627968755118374e-07, + "loss": 0.9432, + "step": 7577 + }, + { + "epoch": 0.6834107408576453, + "grad_norm": 0.7188405518918569, + "learning_rate": 9.622974103140468e-07, + "loss": 0.7694, + "step": 7578 + }, + { + "epoch": 0.6835009243811155, + "grad_norm": 1.4789635245653048, + "learning_rate": 9.617980336587632e-07, + "loss": 1.0111, + "step": 7579 + }, + { + "epoch": 0.6835911079045859, + "grad_norm": 1.5612936321969422, + "learning_rate": 9.612987455885964e-07, + "loss": 0.8368, + "step": 7580 + }, + { + "epoch": 0.6836812914280561, + "grad_norm": 1.4902164563142906, + "learning_rate": 9.607995461461467e-07, + "loss": 0.9562, + "step": 7581 + }, + { + "epoch": 0.6837714749515263, + "grad_norm": 1.5876174072343563, + "learning_rate": 9.603004353740111e-07, + "loss": 0.8513, + "step": 7582 + }, + { + "epoch": 0.6838616584749966, + "grad_norm": 0.731798033474628, + "learning_rate": 9.598014133147738e-07, + "loss": 0.8242, + "step": 7583 + }, + { + "epoch": 0.6839518419984669, + "grad_norm": 1.4251426625364714, + "learning_rate": 9.59302480011017e-07, + "loss": 0.9143, + "step": 7584 + }, + { + "epoch": 0.6840420255219372, + "grad_norm": 0.7162925754828567, + "learning_rate": 9.588036355053102e-07, + "loss": 0.7993, + "step": 7585 + }, + { + "epoch": 0.6841322090454074, + "grad_norm": 1.7285729350271495, + "learning_rate": 9.583048798402182e-07, + "loss": 0.8185, + "step": 7586 + }, + { + "epoch": 0.6842223925688776, + "grad_norm": 2.481240940798117, + "learning_rate": 9.57806213058298e-07, + "loss": 0.9473, + "step": 7587 + }, + { + "epoch": 0.684312576092348, + "grad_norm": 0.7592173659992498, + "learning_rate": 9.57307635202098e-07, + "loss": 0.8175, + "step": 7588 + }, + { + "epoch": 0.6844027596158182, + "grad_norm": 1.5365452973491098, + "learning_rate": 9.568091463141607e-07, + "loss": 1.0571, + "step": 7589 + }, + { + "epoch": 0.6844929431392884, + "grad_norm": 1.520588484959501, + "learning_rate": 9.563107464370187e-07, + "loss": 0.9217, + "step": 7590 + }, + { + "epoch": 0.6845831266627587, + "grad_norm": 1.541850551601588, + "learning_rate": 9.558124356131982e-07, + "loss": 0.9703, + "step": 7591 + }, + { + "epoch": 0.684673310186229, + "grad_norm": 1.7321384202341326, + "learning_rate": 9.553142138852187e-07, + "loss": 0.8853, + "step": 7592 + }, + { + "epoch": 0.6847634937096992, + "grad_norm": 1.8653583585266122, + "learning_rate": 9.548160812955905e-07, + "loss": 0.8619, + "step": 7593 + }, + { + "epoch": 0.6848536772331695, + "grad_norm": 1.371650474092769, + "learning_rate": 9.543180378868175e-07, + "loss": 0.774, + "step": 7594 + }, + { + "epoch": 0.6849438607566397, + "grad_norm": 1.6252008460709033, + "learning_rate": 9.538200837013962e-07, + "loss": 0.9837, + "step": 7595 + }, + { + "epoch": 0.68503404428011, + "grad_norm": 0.7204235330596052, + "learning_rate": 9.533222187818122e-07, + "loss": 0.8161, + "step": 7596 + }, + { + "epoch": 0.6851242278035803, + "grad_norm": 1.7147401678479457, + "learning_rate": 9.528244431705492e-07, + "loss": 0.9624, + "step": 7597 + }, + { + "epoch": 0.6852144113270505, + "grad_norm": 1.381479126926147, + "learning_rate": 9.523267569100774e-07, + "loss": 0.96, + "step": 7598 + }, + { + "epoch": 0.6853045948505209, + "grad_norm": 2.2834860933906507, + "learning_rate": 9.518291600428652e-07, + "loss": 0.912, + "step": 7599 + }, + { + "epoch": 0.6853947783739911, + "grad_norm": 1.840074528760404, + "learning_rate": 9.513316526113677e-07, + "loss": 0.9769, + "step": 7600 + }, + { + "epoch": 0.6854849618974613, + "grad_norm": 1.5061263702434469, + "learning_rate": 9.50834234658036e-07, + "loss": 0.976, + "step": 7601 + }, + { + "epoch": 0.6855751454209316, + "grad_norm": 1.4669038248501036, + "learning_rate": 9.503369062253123e-07, + "loss": 0.9688, + "step": 7602 + }, + { + "epoch": 0.6856653289444019, + "grad_norm": 0.7298285795157006, + "learning_rate": 9.498396673556317e-07, + "loss": 0.791, + "step": 7603 + }, + { + "epoch": 0.6857555124678721, + "grad_norm": 1.617162142200045, + "learning_rate": 9.493425180914219e-07, + "loss": 0.9872, + "step": 7604 + }, + { + "epoch": 0.6858456959913424, + "grad_norm": 3.4185385964705857, + "learning_rate": 9.488454584751e-07, + "loss": 0.885, + "step": 7605 + }, + { + "epoch": 0.6859358795148126, + "grad_norm": 1.4966984017511489, + "learning_rate": 9.483484885490813e-07, + "loss": 0.9543, + "step": 7606 + }, + { + "epoch": 0.686026063038283, + "grad_norm": 1.750485625324868, + "learning_rate": 9.478516083557675e-07, + "loss": 0.9659, + "step": 7607 + }, + { + "epoch": 0.6861162465617532, + "grad_norm": 1.6915388165473333, + "learning_rate": 9.473548179375561e-07, + "loss": 0.9741, + "step": 7608 + }, + { + "epoch": 0.6862064300852234, + "grad_norm": 0.7377846071689691, + "learning_rate": 9.468581173368358e-07, + "loss": 0.8189, + "step": 7609 + }, + { + "epoch": 0.6862966136086937, + "grad_norm": 1.5097791268952776, + "learning_rate": 9.463615065959878e-07, + "loss": 0.9593, + "step": 7610 + }, + { + "epoch": 0.686386797132164, + "grad_norm": 1.5246106448059502, + "learning_rate": 9.458649857573857e-07, + "loss": 1.0037, + "step": 7611 + }, + { + "epoch": 0.6864769806556342, + "grad_norm": 1.6138963611989345, + "learning_rate": 9.453685548633963e-07, + "loss": 0.9613, + "step": 7612 + }, + { + "epoch": 0.6865671641791045, + "grad_norm": 1.8694921978045207, + "learning_rate": 9.448722139563756e-07, + "loss": 0.9017, + "step": 7613 + }, + { + "epoch": 0.6866573477025747, + "grad_norm": 1.5425467771241457, + "learning_rate": 9.443759630786769e-07, + "loss": 0.8479, + "step": 7614 + }, + { + "epoch": 0.686747531226045, + "grad_norm": 1.5111626284752802, + "learning_rate": 9.438798022726408e-07, + "loss": 1.0275, + "step": 7615 + }, + { + "epoch": 0.6868377147495153, + "grad_norm": 1.726946964869873, + "learning_rate": 9.433837315806037e-07, + "loss": 0.9087, + "step": 7616 + }, + { + "epoch": 0.6869278982729855, + "grad_norm": 1.765232346346374, + "learning_rate": 9.428877510448925e-07, + "loss": 0.9062, + "step": 7617 + }, + { + "epoch": 0.6870180817964557, + "grad_norm": 1.3863130647642588, + "learning_rate": 9.423918607078272e-07, + "loss": 0.8851, + "step": 7618 + }, + { + "epoch": 0.6871082653199261, + "grad_norm": 1.676664604213551, + "learning_rate": 9.418960606117208e-07, + "loss": 0.9634, + "step": 7619 + }, + { + "epoch": 0.6871984488433963, + "grad_norm": 1.4602571422974995, + "learning_rate": 9.414003507988752e-07, + "loss": 0.893, + "step": 7620 + }, + { + "epoch": 0.6872886323668665, + "grad_norm": 1.5499963575751052, + "learning_rate": 9.409047313115904e-07, + "loss": 0.8864, + "step": 7621 + }, + { + "epoch": 0.6873788158903368, + "grad_norm": 1.6242817978872244, + "learning_rate": 9.404092021921521e-07, + "loss": 0.894, + "step": 7622 + }, + { + "epoch": 0.6874689994138071, + "grad_norm": 1.921662543734351, + "learning_rate": 9.399137634828447e-07, + "loss": 0.9349, + "step": 7623 + }, + { + "epoch": 0.6875591829372774, + "grad_norm": 1.4161139513275516, + "learning_rate": 9.394184152259396e-07, + "loss": 0.9735, + "step": 7624 + }, + { + "epoch": 0.6876493664607476, + "grad_norm": 1.6175811583309294, + "learning_rate": 9.389231574637033e-07, + "loss": 1.0567, + "step": 7625 + }, + { + "epoch": 0.6877395499842179, + "grad_norm": 1.5229033487401233, + "learning_rate": 9.384279902383938e-07, + "loss": 1.0373, + "step": 7626 + }, + { + "epoch": 0.6878297335076882, + "grad_norm": 1.4791141643967074, + "learning_rate": 9.379329135922615e-07, + "loss": 0.9712, + "step": 7627 + }, + { + "epoch": 0.6879199170311584, + "grad_norm": 1.693973275983808, + "learning_rate": 9.374379275675495e-07, + "loss": 0.9095, + "step": 7628 + }, + { + "epoch": 0.6880101005546286, + "grad_norm": 1.7589578923471052, + "learning_rate": 9.369430322064931e-07, + "loss": 0.9496, + "step": 7629 + }, + { + "epoch": 0.688100284078099, + "grad_norm": 1.673214456064152, + "learning_rate": 9.364482275513179e-07, + "loss": 1.0302, + "step": 7630 + }, + { + "epoch": 0.6881904676015692, + "grad_norm": 1.5366724500871394, + "learning_rate": 9.359535136442444e-07, + "loss": 1.018, + "step": 7631 + }, + { + "epoch": 0.6882806511250394, + "grad_norm": 1.5341580861086668, + "learning_rate": 9.354588905274843e-07, + "loss": 0.8849, + "step": 7632 + }, + { + "epoch": 0.6883708346485097, + "grad_norm": 1.5825816762857703, + "learning_rate": 9.349643582432414e-07, + "loss": 0.983, + "step": 7633 + }, + { + "epoch": 0.68846101817198, + "grad_norm": 1.3621396856980479, + "learning_rate": 9.344699168337127e-07, + "loss": 0.9446, + "step": 7634 + }, + { + "epoch": 0.6885512016954503, + "grad_norm": 1.6732502210096267, + "learning_rate": 9.339755663410845e-07, + "loss": 0.9213, + "step": 7635 + }, + { + "epoch": 0.6886413852189205, + "grad_norm": 2.4822800158313827, + "learning_rate": 9.334813068075405e-07, + "loss": 0.9826, + "step": 7636 + }, + { + "epoch": 0.6887315687423907, + "grad_norm": 1.4743045719235486, + "learning_rate": 9.329871382752506e-07, + "loss": 0.96, + "step": 7637 + }, + { + "epoch": 0.6888217522658611, + "grad_norm": 1.458497156296692, + "learning_rate": 9.32493060786383e-07, + "loss": 0.9438, + "step": 7638 + }, + { + "epoch": 0.6889119357893313, + "grad_norm": 1.2830913152421448, + "learning_rate": 9.31999074383093e-07, + "loss": 1.0104, + "step": 7639 + }, + { + "epoch": 0.6890021193128015, + "grad_norm": 1.6415653258799876, + "learning_rate": 9.315051791075308e-07, + "loss": 0.9473, + "step": 7640 + }, + { + "epoch": 0.6890923028362718, + "grad_norm": 1.2955747773685864, + "learning_rate": 9.310113750018382e-07, + "loss": 0.9445, + "step": 7641 + }, + { + "epoch": 0.6891824863597421, + "grad_norm": 1.8379969307683943, + "learning_rate": 9.305176621081496e-07, + "loss": 1.0266, + "step": 7642 + }, + { + "epoch": 0.6892726698832123, + "grad_norm": 1.3997044711181716, + "learning_rate": 9.300240404685911e-07, + "loss": 0.9876, + "step": 7643 + }, + { + "epoch": 0.6893628534066826, + "grad_norm": 1.5408426436697777, + "learning_rate": 9.295305101252812e-07, + "loss": 0.9555, + "step": 7644 + }, + { + "epoch": 0.6894530369301528, + "grad_norm": 1.7281066065769983, + "learning_rate": 9.290370711203314e-07, + "loss": 0.9374, + "step": 7645 + }, + { + "epoch": 0.6895432204536232, + "grad_norm": 1.3984249689830062, + "learning_rate": 9.285437234958433e-07, + "loss": 0.9462, + "step": 7646 + }, + { + "epoch": 0.6896334039770934, + "grad_norm": 1.620825320161103, + "learning_rate": 9.280504672939124e-07, + "loss": 0.9469, + "step": 7647 + }, + { + "epoch": 0.6897235875005636, + "grad_norm": 1.390523799685705, + "learning_rate": 9.275573025566266e-07, + "loss": 0.9929, + "step": 7648 + }, + { + "epoch": 0.689813771024034, + "grad_norm": 1.2816350404849663, + "learning_rate": 9.27064229326065e-07, + "loss": 0.9862, + "step": 7649 + }, + { + "epoch": 0.6899039545475042, + "grad_norm": 1.503508359118948, + "learning_rate": 9.265712476442995e-07, + "loss": 0.9186, + "step": 7650 + }, + { + "epoch": 0.6899941380709744, + "grad_norm": 1.3683016351860817, + "learning_rate": 9.260783575533949e-07, + "loss": 0.8819, + "step": 7651 + }, + { + "epoch": 0.6900843215944447, + "grad_norm": 1.5151107544907114, + "learning_rate": 9.255855590954045e-07, + "loss": 0.9028, + "step": 7652 + }, + { + "epoch": 0.690174505117915, + "grad_norm": 1.4364792890267282, + "learning_rate": 9.250928523123802e-07, + "loss": 1.0172, + "step": 7653 + }, + { + "epoch": 0.6902646886413852, + "grad_norm": 1.3602296181687015, + "learning_rate": 9.24600237246359e-07, + "loss": 0.9621, + "step": 7654 + }, + { + "epoch": 0.6903548721648555, + "grad_norm": 1.5050257095726336, + "learning_rate": 9.241077139393769e-07, + "loss": 0.9594, + "step": 7655 + }, + { + "epoch": 0.6904450556883257, + "grad_norm": 1.6759681864156735, + "learning_rate": 9.236152824334564e-07, + "loss": 0.9869, + "step": 7656 + }, + { + "epoch": 0.690535239211796, + "grad_norm": 1.5486089063019541, + "learning_rate": 9.231229427706151e-07, + "loss": 0.8504, + "step": 7657 + }, + { + "epoch": 0.6906254227352663, + "grad_norm": 1.6412874610036559, + "learning_rate": 9.226306949928622e-07, + "loss": 0.9608, + "step": 7658 + }, + { + "epoch": 0.6907156062587365, + "grad_norm": 1.5417760947629349, + "learning_rate": 9.221385391421988e-07, + "loss": 1.0299, + "step": 7659 + }, + { + "epoch": 0.6908057897822067, + "grad_norm": 1.4447095283348819, + "learning_rate": 9.216464752606192e-07, + "loss": 0.9697, + "step": 7660 + }, + { + "epoch": 0.6908959733056771, + "grad_norm": 1.4355689637437508, + "learning_rate": 9.211545033901078e-07, + "loss": 1.0317, + "step": 7661 + }, + { + "epoch": 0.6909861568291473, + "grad_norm": 1.9885886803397015, + "learning_rate": 9.206626235726426e-07, + "loss": 0.9847, + "step": 7662 + }, + { + "epoch": 0.6910763403526176, + "grad_norm": 1.8360812516444855, + "learning_rate": 9.20170835850194e-07, + "loss": 0.8872, + "step": 7663 + }, + { + "epoch": 0.6911665238760878, + "grad_norm": 0.6762699499025667, + "learning_rate": 9.196791402647237e-07, + "loss": 0.774, + "step": 7664 + }, + { + "epoch": 0.6912567073995581, + "grad_norm": 1.4880934068577716, + "learning_rate": 9.191875368581861e-07, + "loss": 0.8778, + "step": 7665 + }, + { + "epoch": 0.6913468909230284, + "grad_norm": 1.6799049414126435, + "learning_rate": 9.186960256725271e-07, + "loss": 0.9651, + "step": 7666 + }, + { + "epoch": 0.6914370744464986, + "grad_norm": 1.5161032954330966, + "learning_rate": 9.182046067496856e-07, + "loss": 1.0179, + "step": 7667 + }, + { + "epoch": 0.6915272579699688, + "grad_norm": 1.857652908541658, + "learning_rate": 9.177132801315927e-07, + "loss": 1.0391, + "step": 7668 + }, + { + "epoch": 0.6916174414934392, + "grad_norm": 1.4162702660318485, + "learning_rate": 9.172220458601692e-07, + "loss": 0.9286, + "step": 7669 + }, + { + "epoch": 0.6917076250169094, + "grad_norm": 1.8420674680475169, + "learning_rate": 9.167309039773324e-07, + "loss": 0.9189, + "step": 7670 + }, + { + "epoch": 0.6917978085403796, + "grad_norm": 1.2793764977756188, + "learning_rate": 9.162398545249872e-07, + "loss": 0.8538, + "step": 7671 + }, + { + "epoch": 0.69188799206385, + "grad_norm": 1.350405467919832, + "learning_rate": 9.157488975450334e-07, + "loss": 1.0085, + "step": 7672 + }, + { + "epoch": 0.6919781755873202, + "grad_norm": 1.486194579532509, + "learning_rate": 9.15258033079362e-07, + "loss": 0.7831, + "step": 7673 + }, + { + "epoch": 0.6920683591107905, + "grad_norm": 1.4366967818223022, + "learning_rate": 9.147672611698567e-07, + "loss": 0.9421, + "step": 7674 + }, + { + "epoch": 0.6921585426342607, + "grad_norm": 2.2862393934067455, + "learning_rate": 9.142765818583933e-07, + "loss": 0.944, + "step": 7675 + }, + { + "epoch": 0.692248726157731, + "grad_norm": 1.4890421526740556, + "learning_rate": 9.13785995186837e-07, + "loss": 0.9328, + "step": 7676 + }, + { + "epoch": 0.6923389096812013, + "grad_norm": 1.4760012897289398, + "learning_rate": 9.132955011970502e-07, + "loss": 0.918, + "step": 7677 + }, + { + "epoch": 0.6924290932046715, + "grad_norm": 1.2978089887469102, + "learning_rate": 9.128050999308827e-07, + "loss": 0.9384, + "step": 7678 + }, + { + "epoch": 0.6925192767281417, + "grad_norm": 1.7613101324641687, + "learning_rate": 9.123147914301789e-07, + "loss": 1.0403, + "step": 7679 + }, + { + "epoch": 0.6926094602516121, + "grad_norm": 1.793190936115817, + "learning_rate": 9.118245757367745e-07, + "loss": 0.9993, + "step": 7680 + }, + { + "epoch": 0.6926996437750823, + "grad_norm": 0.8330957471400829, + "learning_rate": 9.113344528924973e-07, + "loss": 0.761, + "step": 7681 + }, + { + "epoch": 0.6927898272985525, + "grad_norm": 1.382408126068515, + "learning_rate": 9.108444229391676e-07, + "loss": 0.9339, + "step": 7682 + }, + { + "epoch": 0.6928800108220228, + "grad_norm": 1.6287502881955223, + "learning_rate": 9.103544859185979e-07, + "loss": 0.9254, + "step": 7683 + }, + { + "epoch": 0.6929701943454931, + "grad_norm": 2.0483176701906216, + "learning_rate": 9.098646418725902e-07, + "loss": 1.0405, + "step": 7684 + }, + { + "epoch": 0.6930603778689634, + "grad_norm": 0.6981455702726589, + "learning_rate": 9.093748908429437e-07, + "loss": 0.8038, + "step": 7685 + }, + { + "epoch": 0.6931505613924336, + "grad_norm": 1.4772604448301394, + "learning_rate": 9.088852328714444e-07, + "loss": 0.9453, + "step": 7686 + }, + { + "epoch": 0.6932407449159038, + "grad_norm": 1.3947331899398143, + "learning_rate": 9.083956679998735e-07, + "loss": 0.8619, + "step": 7687 + }, + { + "epoch": 0.6933309284393742, + "grad_norm": 1.2900857736474791, + "learning_rate": 9.079061962700032e-07, + "loss": 0.9203, + "step": 7688 + }, + { + "epoch": 0.6934211119628444, + "grad_norm": 1.3248223023647427, + "learning_rate": 9.074168177235979e-07, + "loss": 0.9756, + "step": 7689 + }, + { + "epoch": 0.6935112954863146, + "grad_norm": 1.7054052224867648, + "learning_rate": 9.069275324024151e-07, + "loss": 0.9694, + "step": 7690 + }, + { + "epoch": 0.6936014790097849, + "grad_norm": 1.6308094496708878, + "learning_rate": 9.064383403482005e-07, + "loss": 0.8863, + "step": 7691 + }, + { + "epoch": 0.6936916625332552, + "grad_norm": 1.3983906572903846, + "learning_rate": 9.059492416026983e-07, + "loss": 0.9191, + "step": 7692 + }, + { + "epoch": 0.6937818460567254, + "grad_norm": 1.5642062217386736, + "learning_rate": 9.054602362076378e-07, + "loss": 0.9484, + "step": 7693 + }, + { + "epoch": 0.6938720295801957, + "grad_norm": 2.5952270163020033, + "learning_rate": 9.049713242047468e-07, + "loss": 0.9418, + "step": 7694 + }, + { + "epoch": 0.693962213103666, + "grad_norm": 1.672223063807031, + "learning_rate": 9.044825056357395e-07, + "loss": 0.922, + "step": 7695 + }, + { + "epoch": 0.6940523966271362, + "grad_norm": 1.5533198182632733, + "learning_rate": 9.039937805423255e-07, + "loss": 0.9066, + "step": 7696 + }, + { + "epoch": 0.6941425801506065, + "grad_norm": 0.6567899662370028, + "learning_rate": 9.035051489662051e-07, + "loss": 0.7181, + "step": 7697 + }, + { + "epoch": 0.6942327636740767, + "grad_norm": 1.4942965158157049, + "learning_rate": 9.030166109490718e-07, + "loss": 0.9022, + "step": 7698 + }, + { + "epoch": 0.6943229471975471, + "grad_norm": 2.1044168480931553, + "learning_rate": 9.025281665326099e-07, + "loss": 0.9662, + "step": 7699 + }, + { + "epoch": 0.6944131307210173, + "grad_norm": 1.7208903770444175, + "learning_rate": 9.020398157584967e-07, + "loss": 0.9569, + "step": 7700 + }, + { + "epoch": 0.6945033142444875, + "grad_norm": 1.6174155549023232, + "learning_rate": 9.015515586684002e-07, + "loss": 0.855, + "step": 7701 + }, + { + "epoch": 0.6945934977679578, + "grad_norm": 0.7282011360670876, + "learning_rate": 9.010633953039812e-07, + "loss": 0.7941, + "step": 7702 + }, + { + "epoch": 0.6946836812914281, + "grad_norm": 1.5448196319429028, + "learning_rate": 9.005753257068929e-07, + "loss": 0.9649, + "step": 7703 + }, + { + "epoch": 0.6947738648148983, + "grad_norm": 1.514399318019635, + "learning_rate": 9.000873499187797e-07, + "loss": 0.9646, + "step": 7704 + }, + { + "epoch": 0.6948640483383686, + "grad_norm": 1.778969023698058, + "learning_rate": 8.995994679812797e-07, + "loss": 0.9828, + "step": 7705 + }, + { + "epoch": 0.6949542318618388, + "grad_norm": 1.5162143937700123, + "learning_rate": 8.991116799360192e-07, + "loss": 0.9751, + "step": 7706 + }, + { + "epoch": 0.6950444153853091, + "grad_norm": 1.365719339355107, + "learning_rate": 8.986239858246217e-07, + "loss": 0.9446, + "step": 7707 + }, + { + "epoch": 0.6951345989087794, + "grad_norm": 1.8528727470713147, + "learning_rate": 8.981363856886972e-07, + "loss": 1.0324, + "step": 7708 + }, + { + "epoch": 0.6952247824322496, + "grad_norm": 1.4875820650175045, + "learning_rate": 8.976488795698533e-07, + "loss": 0.9359, + "step": 7709 + }, + { + "epoch": 0.6953149659557198, + "grad_norm": 1.358703973827586, + "learning_rate": 8.971614675096841e-07, + "loss": 0.9658, + "step": 7710 + }, + { + "epoch": 0.6954051494791902, + "grad_norm": 1.8648497179860386, + "learning_rate": 8.966741495497807e-07, + "loss": 1.0196, + "step": 7711 + }, + { + "epoch": 0.6954953330026604, + "grad_norm": 1.4013640825722353, + "learning_rate": 8.961869257317218e-07, + "loss": 0.9171, + "step": 7712 + }, + { + "epoch": 0.6955855165261307, + "grad_norm": 1.4921012174416386, + "learning_rate": 8.956997960970809e-07, + "loss": 0.9329, + "step": 7713 + }, + { + "epoch": 0.6956757000496009, + "grad_norm": 1.597113855343805, + "learning_rate": 8.952127606874224e-07, + "loss": 0.8637, + "step": 7714 + }, + { + "epoch": 0.6957658835730712, + "grad_norm": 1.3925403306281088, + "learning_rate": 8.947258195443028e-07, + "loss": 0.9556, + "step": 7715 + }, + { + "epoch": 0.6958560670965415, + "grad_norm": 1.5636407121482934, + "learning_rate": 8.942389727092716e-07, + "loss": 0.9308, + "step": 7716 + }, + { + "epoch": 0.6959462506200117, + "grad_norm": 0.7318964993961917, + "learning_rate": 8.937522202238677e-07, + "loss": 0.7658, + "step": 7717 + }, + { + "epoch": 0.696036434143482, + "grad_norm": 1.5921108568067546, + "learning_rate": 8.932655621296239e-07, + "loss": 0.9653, + "step": 7718 + }, + { + "epoch": 0.6961266176669523, + "grad_norm": 1.342988485975751, + "learning_rate": 8.927789984680649e-07, + "loss": 0.9486, + "step": 7719 + }, + { + "epoch": 0.6962168011904225, + "grad_norm": 1.5132733220260515, + "learning_rate": 8.922925292807068e-07, + "loss": 0.8199, + "step": 7720 + }, + { + "epoch": 0.6963069847138927, + "grad_norm": 1.9221661354927941, + "learning_rate": 8.91806154609058e-07, + "loss": 1.037, + "step": 7721 + }, + { + "epoch": 0.6963971682373631, + "grad_norm": 1.4715641738321503, + "learning_rate": 8.913198744946195e-07, + "loss": 0.9236, + "step": 7722 + }, + { + "epoch": 0.6964873517608333, + "grad_norm": 1.5408697216357894, + "learning_rate": 8.908336889788807e-07, + "loss": 0.9749, + "step": 7723 + }, + { + "epoch": 0.6965775352843036, + "grad_norm": 1.9343810049034673, + "learning_rate": 8.903475981033293e-07, + "loss": 0.8352, + "step": 7724 + }, + { + "epoch": 0.6966677188077738, + "grad_norm": 0.7904357291580553, + "learning_rate": 8.898616019094376e-07, + "loss": 0.8337, + "step": 7725 + }, + { + "epoch": 0.6967579023312441, + "grad_norm": 1.570820664507357, + "learning_rate": 8.89375700438677e-07, + "loss": 0.9887, + "step": 7726 + }, + { + "epoch": 0.6968480858547144, + "grad_norm": 1.5959713732558987, + "learning_rate": 8.888898937325047e-07, + "loss": 0.9135, + "step": 7727 + }, + { + "epoch": 0.6969382693781846, + "grad_norm": 2.2237725161634225, + "learning_rate": 8.884041818323733e-07, + "loss": 0.9197, + "step": 7728 + }, + { + "epoch": 0.6970284529016548, + "grad_norm": 1.586900465686943, + "learning_rate": 8.879185647797262e-07, + "loss": 0.9844, + "step": 7729 + }, + { + "epoch": 0.6971186364251252, + "grad_norm": 1.6707328388817426, + "learning_rate": 8.874330426159993e-07, + "loss": 0.9328, + "step": 7730 + }, + { + "epoch": 0.6972088199485954, + "grad_norm": 1.3744190896297237, + "learning_rate": 8.869476153826205e-07, + "loss": 0.9759, + "step": 7731 + }, + { + "epoch": 0.6972990034720656, + "grad_norm": 1.6745430166392248, + "learning_rate": 8.864622831210071e-07, + "loss": 0.9508, + "step": 7732 + }, + { + "epoch": 0.6973891869955359, + "grad_norm": 1.6512849630484931, + "learning_rate": 8.85977045872573e-07, + "loss": 0.9621, + "step": 7733 + }, + { + "epoch": 0.6974793705190062, + "grad_norm": 1.9445916490994803, + "learning_rate": 8.854919036787194e-07, + "loss": 0.9304, + "step": 7734 + }, + { + "epoch": 0.6975695540424764, + "grad_norm": 1.487530456359399, + "learning_rate": 8.850068565808417e-07, + "loss": 0.9157, + "step": 7735 + }, + { + "epoch": 0.6976597375659467, + "grad_norm": 1.9787054093253271, + "learning_rate": 8.845219046203271e-07, + "loss": 1.0017, + "step": 7736 + }, + { + "epoch": 0.6977499210894169, + "grad_norm": 1.845066053453823, + "learning_rate": 8.840370478385544e-07, + "loss": 0.9794, + "step": 7737 + }, + { + "epoch": 0.6978401046128873, + "grad_norm": 1.5538036958925225, + "learning_rate": 8.83552286276894e-07, + "loss": 0.9716, + "step": 7738 + }, + { + "epoch": 0.6979302881363575, + "grad_norm": 1.8224001852155998, + "learning_rate": 8.830676199767095e-07, + "loss": 0.9701, + "step": 7739 + }, + { + "epoch": 0.6980204716598277, + "grad_norm": 1.611270048077423, + "learning_rate": 8.825830489793527e-07, + "loss": 0.9338, + "step": 7740 + }, + { + "epoch": 0.698110655183298, + "grad_norm": 1.2973383799536766, + "learning_rate": 8.820985733261732e-07, + "loss": 0.9007, + "step": 7741 + }, + { + "epoch": 0.6982008387067683, + "grad_norm": 1.9231432939960411, + "learning_rate": 8.816141930585066e-07, + "loss": 0.8625, + "step": 7742 + }, + { + "epoch": 0.6982910222302385, + "grad_norm": 0.7522901696192269, + "learning_rate": 8.811299082176837e-07, + "loss": 0.7907, + "step": 7743 + }, + { + "epoch": 0.6983812057537088, + "grad_norm": 1.41341077422962, + "learning_rate": 8.806457188450265e-07, + "loss": 0.913, + "step": 7744 + }, + { + "epoch": 0.6984713892771791, + "grad_norm": 1.467616435148218, + "learning_rate": 8.801616249818487e-07, + "loss": 0.9892, + "step": 7745 + }, + { + "epoch": 0.6985615728006493, + "grad_norm": 1.7517080146175452, + "learning_rate": 8.796776266694564e-07, + "loss": 0.9315, + "step": 7746 + }, + { + "epoch": 0.6986517563241196, + "grad_norm": 1.2944823564908772, + "learning_rate": 8.79193723949145e-07, + "loss": 0.8934, + "step": 7747 + }, + { + "epoch": 0.6987419398475898, + "grad_norm": 1.4945091042520486, + "learning_rate": 8.787099168622063e-07, + "loss": 0.8453, + "step": 7748 + }, + { + "epoch": 0.6988321233710602, + "grad_norm": 1.6002623700553409, + "learning_rate": 8.782262054499199e-07, + "loss": 0.889, + "step": 7749 + }, + { + "epoch": 0.6989223068945304, + "grad_norm": 1.2886840264706343, + "learning_rate": 8.777425897535588e-07, + "loss": 0.9803, + "step": 7750 + }, + { + "epoch": 0.6990124904180006, + "grad_norm": 1.6313787855268516, + "learning_rate": 8.77259069814388e-07, + "loss": 0.9521, + "step": 7751 + }, + { + "epoch": 0.6991026739414709, + "grad_norm": 1.3557913448550984, + "learning_rate": 8.767756456736641e-07, + "loss": 0.9527, + "step": 7752 + }, + { + "epoch": 0.6991928574649412, + "grad_norm": 1.5173357717268736, + "learning_rate": 8.762923173726358e-07, + "loss": 0.8758, + "step": 7753 + }, + { + "epoch": 0.6992830409884114, + "grad_norm": 1.5738783248200927, + "learning_rate": 8.758090849525428e-07, + "loss": 1.0513, + "step": 7754 + }, + { + "epoch": 0.6993732245118817, + "grad_norm": 1.7669584926961042, + "learning_rate": 8.753259484546174e-07, + "loss": 0.8915, + "step": 7755 + }, + { + "epoch": 0.6994634080353519, + "grad_norm": 1.5292154926262629, + "learning_rate": 8.748429079200841e-07, + "loss": 0.9056, + "step": 7756 + }, + { + "epoch": 0.6995535915588222, + "grad_norm": 1.6629077325343722, + "learning_rate": 8.743599633901575e-07, + "loss": 1.0092, + "step": 7757 + }, + { + "epoch": 0.6996437750822925, + "grad_norm": 1.62249460016654, + "learning_rate": 8.738771149060453e-07, + "loss": 0.9662, + "step": 7758 + }, + { + "epoch": 0.6997339586057627, + "grad_norm": 1.5173534487199365, + "learning_rate": 8.73394362508947e-07, + "loss": 0.9174, + "step": 7759 + }, + { + "epoch": 0.6998241421292329, + "grad_norm": 1.3839428263690394, + "learning_rate": 8.72911706240054e-07, + "loss": 0.9172, + "step": 7760 + }, + { + "epoch": 0.6999143256527033, + "grad_norm": 1.5124163927662355, + "learning_rate": 8.724291461405493e-07, + "loss": 0.8285, + "step": 7761 + }, + { + "epoch": 0.7000045091761735, + "grad_norm": 1.539278663907391, + "learning_rate": 8.71946682251606e-07, + "loss": 0.9165, + "step": 7762 + }, + { + "epoch": 0.7000946926996438, + "grad_norm": 1.7323468696330118, + "learning_rate": 8.714643146143932e-07, + "loss": 0.9506, + "step": 7763 + }, + { + "epoch": 0.700184876223114, + "grad_norm": 1.510604645004414, + "learning_rate": 8.709820432700663e-07, + "loss": 0.9098, + "step": 7764 + }, + { + "epoch": 0.7002750597465843, + "grad_norm": 1.4199047481128368, + "learning_rate": 8.704998682597784e-07, + "loss": 0.8681, + "step": 7765 + }, + { + "epoch": 0.7003652432700546, + "grad_norm": 1.500630881835392, + "learning_rate": 8.700177896246688e-07, + "loss": 0.9693, + "step": 7766 + }, + { + "epoch": 0.7004554267935248, + "grad_norm": 1.7378010633065202, + "learning_rate": 8.695358074058721e-07, + "loss": 0.8776, + "step": 7767 + }, + { + "epoch": 0.7005456103169951, + "grad_norm": 1.659570424557009, + "learning_rate": 8.690539216445136e-07, + "loss": 0.9872, + "step": 7768 + }, + { + "epoch": 0.7006357938404654, + "grad_norm": 1.687134950500932, + "learning_rate": 8.685721323817106e-07, + "loss": 0.8531, + "step": 7769 + }, + { + "epoch": 0.7007259773639356, + "grad_norm": 1.4531723342897571, + "learning_rate": 8.680904396585718e-07, + "loss": 0.986, + "step": 7770 + }, + { + "epoch": 0.7008161608874058, + "grad_norm": 1.5200277659741204, + "learning_rate": 8.676088435161988e-07, + "loss": 0.8919, + "step": 7771 + }, + { + "epoch": 0.7009063444108762, + "grad_norm": 1.6140936406280593, + "learning_rate": 8.671273439956824e-07, + "loss": 0.9682, + "step": 7772 + }, + { + "epoch": 0.7009965279343464, + "grad_norm": 1.527074219319554, + "learning_rate": 8.666459411381075e-07, + "loss": 0.9548, + "step": 7773 + }, + { + "epoch": 0.7010867114578166, + "grad_norm": 1.332139131087166, + "learning_rate": 8.661646349845501e-07, + "loss": 0.976, + "step": 7774 + }, + { + "epoch": 0.7011768949812869, + "grad_norm": 1.8644955435034927, + "learning_rate": 8.656834255760783e-07, + "loss": 1.0057, + "step": 7775 + }, + { + "epoch": 0.7012670785047572, + "grad_norm": 1.6205236696458483, + "learning_rate": 8.652023129537509e-07, + "loss": 1.0026, + "step": 7776 + }, + { + "epoch": 0.7013572620282275, + "grad_norm": 1.4443594599853071, + "learning_rate": 8.647212971586195e-07, + "loss": 0.9777, + "step": 7777 + }, + { + "epoch": 0.7014474455516977, + "grad_norm": 1.596027168606759, + "learning_rate": 8.642403782317275e-07, + "loss": 0.9044, + "step": 7778 + }, + { + "epoch": 0.7015376290751679, + "grad_norm": 1.5769928752131064, + "learning_rate": 8.637595562141075e-07, + "loss": 0.9563, + "step": 7779 + }, + { + "epoch": 0.7016278125986383, + "grad_norm": 1.4413472812241201, + "learning_rate": 8.632788311467889e-07, + "loss": 0.9194, + "step": 7780 + }, + { + "epoch": 0.7017179961221085, + "grad_norm": 1.4778549754184747, + "learning_rate": 8.627982030707867e-07, + "loss": 0.9455, + "step": 7781 + }, + { + "epoch": 0.7018081796455787, + "grad_norm": 1.7234224099184254, + "learning_rate": 8.623176720271139e-07, + "loss": 1.0159, + "step": 7782 + }, + { + "epoch": 0.701898363169049, + "grad_norm": 1.864725125283064, + "learning_rate": 8.618372380567696e-07, + "loss": 1.021, + "step": 7783 + }, + { + "epoch": 0.7019885466925193, + "grad_norm": 1.6849931534534668, + "learning_rate": 8.613569012007478e-07, + "loss": 0.9131, + "step": 7784 + }, + { + "epoch": 0.7020787302159895, + "grad_norm": 1.9457596720680932, + "learning_rate": 8.608766615000338e-07, + "loss": 0.9887, + "step": 7785 + }, + { + "epoch": 0.7021689137394598, + "grad_norm": 1.8117572314809078, + "learning_rate": 8.603965189956039e-07, + "loss": 0.9702, + "step": 7786 + }, + { + "epoch": 0.70225909726293, + "grad_norm": 1.724049668519293, + "learning_rate": 8.599164737284276e-07, + "loss": 0.9591, + "step": 7787 + }, + { + "epoch": 0.7023492807864004, + "grad_norm": 1.5233168334122995, + "learning_rate": 8.594365257394634e-07, + "loss": 0.9069, + "step": 7788 + }, + { + "epoch": 0.7024394643098706, + "grad_norm": 1.702104245092021, + "learning_rate": 8.589566750696637e-07, + "loss": 0.893, + "step": 7789 + }, + { + "epoch": 0.7025296478333408, + "grad_norm": 1.7364071462219226, + "learning_rate": 8.584769217599721e-07, + "loss": 0.8819, + "step": 7790 + }, + { + "epoch": 0.7026198313568112, + "grad_norm": 1.7764241616799512, + "learning_rate": 8.579972658513239e-07, + "loss": 0.8918, + "step": 7791 + }, + { + "epoch": 0.7027100148802814, + "grad_norm": 1.8579230529898967, + "learning_rate": 8.57517707384646e-07, + "loss": 0.9553, + "step": 7792 + }, + { + "epoch": 0.7028001984037516, + "grad_norm": 1.580075710752173, + "learning_rate": 8.570382464008574e-07, + "loss": 0.9863, + "step": 7793 + }, + { + "epoch": 0.7028903819272219, + "grad_norm": 1.4879981545159986, + "learning_rate": 8.565588829408665e-07, + "loss": 0.9147, + "step": 7794 + }, + { + "epoch": 0.7029805654506922, + "grad_norm": 1.6539045316734993, + "learning_rate": 8.560796170455782e-07, + "loss": 0.9267, + "step": 7795 + }, + { + "epoch": 0.7030707489741624, + "grad_norm": 1.581726647711239, + "learning_rate": 8.556004487558828e-07, + "loss": 0.9868, + "step": 7796 + }, + { + "epoch": 0.7031609324976327, + "grad_norm": 1.4896663118027538, + "learning_rate": 8.55121378112669e-07, + "loss": 0.8608, + "step": 7797 + }, + { + "epoch": 0.7032511160211029, + "grad_norm": 1.693278344843471, + "learning_rate": 8.546424051568111e-07, + "loss": 0.9305, + "step": 7798 + }, + { + "epoch": 0.7033412995445733, + "grad_norm": 1.5086995416150648, + "learning_rate": 8.541635299291785e-07, + "loss": 1.0027, + "step": 7799 + }, + { + "epoch": 0.7034314830680435, + "grad_norm": 0.6180279236830489, + "learning_rate": 8.536847524706317e-07, + "loss": 0.8013, + "step": 7800 + }, + { + "epoch": 0.7035216665915137, + "grad_norm": 1.503168812071631, + "learning_rate": 8.532060728220225e-07, + "loss": 0.8807, + "step": 7801 + }, + { + "epoch": 0.703611850114984, + "grad_norm": 1.9304664059126488, + "learning_rate": 8.527274910241955e-07, + "loss": 0.9621, + "step": 7802 + }, + { + "epoch": 0.7037020336384543, + "grad_norm": 1.6531839601786658, + "learning_rate": 8.522490071179833e-07, + "loss": 0.947, + "step": 7803 + }, + { + "epoch": 0.7037922171619245, + "grad_norm": 1.3885531475320338, + "learning_rate": 8.517706211442159e-07, + "loss": 0.8303, + "step": 7804 + }, + { + "epoch": 0.7038824006853948, + "grad_norm": 1.6196910994569034, + "learning_rate": 8.512923331437097e-07, + "loss": 1.0277, + "step": 7805 + }, + { + "epoch": 0.703972584208865, + "grad_norm": 1.4603638364167146, + "learning_rate": 8.508141431572755e-07, + "loss": 0.9138, + "step": 7806 + }, + { + "epoch": 0.7040627677323353, + "grad_norm": 1.809352246077176, + "learning_rate": 8.503360512257152e-07, + "loss": 0.9368, + "step": 7807 + }, + { + "epoch": 0.7041529512558056, + "grad_norm": 1.689245098735253, + "learning_rate": 8.498580573898219e-07, + "loss": 0.9115, + "step": 7808 + }, + { + "epoch": 0.7042431347792758, + "grad_norm": 2.8811463707680733, + "learning_rate": 8.493801616903813e-07, + "loss": 0.957, + "step": 7809 + }, + { + "epoch": 0.704333318302746, + "grad_norm": 1.555528422906182, + "learning_rate": 8.489023641681705e-07, + "loss": 0.967, + "step": 7810 + }, + { + "epoch": 0.7044235018262164, + "grad_norm": 1.4225741333910462, + "learning_rate": 8.484246648639555e-07, + "loss": 0.9955, + "step": 7811 + }, + { + "epoch": 0.7045136853496866, + "grad_norm": 1.4726280796584503, + "learning_rate": 8.479470638184994e-07, + "loss": 0.8264, + "step": 7812 + }, + { + "epoch": 0.7046038688731568, + "grad_norm": 2.587075988379007, + "learning_rate": 8.474695610725513e-07, + "loss": 0.8707, + "step": 7813 + }, + { + "epoch": 0.7046940523966272, + "grad_norm": 1.4075984847201284, + "learning_rate": 8.469921566668552e-07, + "loss": 1.0215, + "step": 7814 + }, + { + "epoch": 0.7047842359200974, + "grad_norm": 1.714343647602366, + "learning_rate": 8.46514850642146e-07, + "loss": 0.9994, + "step": 7815 + }, + { + "epoch": 0.7048744194435677, + "grad_norm": 1.572925954963931, + "learning_rate": 8.460376430391499e-07, + "loss": 0.9357, + "step": 7816 + }, + { + "epoch": 0.7049646029670379, + "grad_norm": 1.1925682029973605, + "learning_rate": 8.455605338985858e-07, + "loss": 0.9791, + "step": 7817 + }, + { + "epoch": 0.7050547864905082, + "grad_norm": 1.4607320222833728, + "learning_rate": 8.45083523261161e-07, + "loss": 0.8036, + "step": 7818 + }, + { + "epoch": 0.7051449700139785, + "grad_norm": 1.289832607616262, + "learning_rate": 8.446066111675796e-07, + "loss": 0.9537, + "step": 7819 + }, + { + "epoch": 0.7052351535374487, + "grad_norm": 1.667242316334291, + "learning_rate": 8.441297976585314e-07, + "loss": 0.843, + "step": 7820 + }, + { + "epoch": 0.7053253370609189, + "grad_norm": 1.8537221940180828, + "learning_rate": 8.436530827747037e-07, + "loss": 1.0349, + "step": 7821 + }, + { + "epoch": 0.7054155205843893, + "grad_norm": 1.7042972135401409, + "learning_rate": 8.431764665567704e-07, + "loss": 1.0844, + "step": 7822 + }, + { + "epoch": 0.7055057041078595, + "grad_norm": 1.671842556932772, + "learning_rate": 8.426999490453996e-07, + "loss": 0.8904, + "step": 7823 + }, + { + "epoch": 0.7055958876313297, + "grad_norm": 1.7332608880284694, + "learning_rate": 8.422235302812504e-07, + "loss": 0.9518, + "step": 7824 + }, + { + "epoch": 0.7056860711548, + "grad_norm": 1.5728740391628662, + "learning_rate": 8.417472103049734e-07, + "loss": 1.0175, + "step": 7825 + }, + { + "epoch": 0.7057762546782703, + "grad_norm": 1.5026224577091836, + "learning_rate": 8.412709891572112e-07, + "loss": 0.9842, + "step": 7826 + }, + { + "epoch": 0.7058664382017406, + "grad_norm": 1.7069776164343258, + "learning_rate": 8.407948668785978e-07, + "loss": 0.9809, + "step": 7827 + }, + { + "epoch": 0.7059566217252108, + "grad_norm": 2.9420084343224633, + "learning_rate": 8.403188435097576e-07, + "loss": 0.9453, + "step": 7828 + }, + { + "epoch": 0.706046805248681, + "grad_norm": 1.526533453361428, + "learning_rate": 8.398429190913081e-07, + "loss": 1.0183, + "step": 7829 + }, + { + "epoch": 0.7061369887721514, + "grad_norm": 1.499832779782323, + "learning_rate": 8.393670936638578e-07, + "loss": 0.9491, + "step": 7830 + }, + { + "epoch": 0.7062271722956216, + "grad_norm": 1.6679520259974663, + "learning_rate": 8.388913672680067e-07, + "loss": 0.9452, + "step": 7831 + }, + { + "epoch": 0.7063173558190918, + "grad_norm": 1.5434140878536398, + "learning_rate": 8.384157399443472e-07, + "loss": 0.8955, + "step": 7832 + }, + { + "epoch": 0.7064075393425621, + "grad_norm": 1.6169215181705177, + "learning_rate": 8.379402117334601e-07, + "loss": 0.9745, + "step": 7833 + }, + { + "epoch": 0.7064977228660324, + "grad_norm": 1.545267676464194, + "learning_rate": 8.374647826759232e-07, + "loss": 0.9503, + "step": 7834 + }, + { + "epoch": 0.7065879063895026, + "grad_norm": 1.5862129611116715, + "learning_rate": 8.369894528122998e-07, + "loss": 0.9452, + "step": 7835 + }, + { + "epoch": 0.7066780899129729, + "grad_norm": 1.6242643304973263, + "learning_rate": 8.365142221831505e-07, + "loss": 0.8148, + "step": 7836 + }, + { + "epoch": 0.7067682734364432, + "grad_norm": 2.017539482817569, + "learning_rate": 8.360390908290222e-07, + "loss": 0.91, + "step": 7837 + }, + { + "epoch": 0.7068584569599135, + "grad_norm": 1.7630192931791608, + "learning_rate": 8.355640587904569e-07, + "loss": 0.9295, + "step": 7838 + }, + { + "epoch": 0.7069486404833837, + "grad_norm": 1.5704716084660832, + "learning_rate": 8.350891261079866e-07, + "loss": 0.9569, + "step": 7839 + }, + { + "epoch": 0.7070388240068539, + "grad_norm": 1.985992313231837, + "learning_rate": 8.346142928221356e-07, + "loss": 0.9922, + "step": 7840 + }, + { + "epoch": 0.7071290075303243, + "grad_norm": 1.6944296511367378, + "learning_rate": 8.341395589734189e-07, + "loss": 0.7922, + "step": 7841 + }, + { + "epoch": 0.7072191910537945, + "grad_norm": 2.9985327311426646, + "learning_rate": 8.336649246023433e-07, + "loss": 0.9038, + "step": 7842 + }, + { + "epoch": 0.7073093745772647, + "grad_norm": 1.4638461772487408, + "learning_rate": 8.331903897494084e-07, + "loss": 0.8503, + "step": 7843 + }, + { + "epoch": 0.707399558100735, + "grad_norm": 1.8697564875304895, + "learning_rate": 8.327159544551024e-07, + "loss": 0.9629, + "step": 7844 + }, + { + "epoch": 0.7074897416242053, + "grad_norm": 1.8195096320526039, + "learning_rate": 8.322416187599073e-07, + "loss": 0.9083, + "step": 7845 + }, + { + "epoch": 0.7075799251476755, + "grad_norm": 1.5985702474441947, + "learning_rate": 8.317673827042963e-07, + "loss": 0.975, + "step": 7846 + }, + { + "epoch": 0.7076701086711458, + "grad_norm": 1.7482614737120126, + "learning_rate": 8.312932463287339e-07, + "loss": 0.938, + "step": 7847 + }, + { + "epoch": 0.707760292194616, + "grad_norm": 1.714757408801262, + "learning_rate": 8.308192096736759e-07, + "loss": 0.9176, + "step": 7848 + }, + { + "epoch": 0.7078504757180863, + "grad_norm": 1.5533172856813053, + "learning_rate": 8.303452727795703e-07, + "loss": 0.9255, + "step": 7849 + }, + { + "epoch": 0.7079406592415566, + "grad_norm": 2.019688376949308, + "learning_rate": 8.298714356868542e-07, + "loss": 1.026, + "step": 7850 + }, + { + "epoch": 0.7080308427650268, + "grad_norm": 1.550769719386053, + "learning_rate": 8.293976984359605e-07, + "loss": 0.9594, + "step": 7851 + }, + { + "epoch": 0.708121026288497, + "grad_norm": 1.6523815538487203, + "learning_rate": 8.289240610673092e-07, + "loss": 0.9503, + "step": 7852 + }, + { + "epoch": 0.7082112098119674, + "grad_norm": 1.459775248172472, + "learning_rate": 8.284505236213144e-07, + "loss": 0.9841, + "step": 7853 + }, + { + "epoch": 0.7083013933354376, + "grad_norm": 1.5589316248064036, + "learning_rate": 8.279770861383806e-07, + "loss": 0.9393, + "step": 7854 + }, + { + "epoch": 0.7083915768589079, + "grad_norm": 1.4858709727576334, + "learning_rate": 8.275037486589042e-07, + "loss": 0.9464, + "step": 7855 + }, + { + "epoch": 0.7084817603823781, + "grad_norm": 0.6840931839071257, + "learning_rate": 8.270305112232739e-07, + "loss": 0.7602, + "step": 7856 + }, + { + "epoch": 0.7085719439058484, + "grad_norm": 1.7311934424020525, + "learning_rate": 8.265573738718665e-07, + "loss": 0.9456, + "step": 7857 + }, + { + "epoch": 0.7086621274293187, + "grad_norm": 1.6604929862995654, + "learning_rate": 8.260843366450559e-07, + "loss": 0.9112, + "step": 7858 + }, + { + "epoch": 0.7087523109527889, + "grad_norm": 1.6062178523190047, + "learning_rate": 8.256113995832017e-07, + "loss": 0.9496, + "step": 7859 + }, + { + "epoch": 0.7088424944762591, + "grad_norm": 1.3579533037547822, + "learning_rate": 8.251385627266583e-07, + "loss": 0.9924, + "step": 7860 + }, + { + "epoch": 0.7089326779997295, + "grad_norm": 1.7619470418793668, + "learning_rate": 8.24665826115771e-07, + "loss": 0.9567, + "step": 7861 + }, + { + "epoch": 0.7090228615231997, + "grad_norm": 1.4281065490999876, + "learning_rate": 8.241931897908763e-07, + "loss": 0.9657, + "step": 7862 + }, + { + "epoch": 0.7091130450466699, + "grad_norm": 1.5070934261075122, + "learning_rate": 8.237206537923016e-07, + "loss": 0.9727, + "step": 7863 + }, + { + "epoch": 0.7092032285701403, + "grad_norm": 1.9360414983542087, + "learning_rate": 8.232482181603671e-07, + "loss": 0.9346, + "step": 7864 + }, + { + "epoch": 0.7092934120936105, + "grad_norm": 1.5879338670311347, + "learning_rate": 8.227758829353828e-07, + "loss": 0.9291, + "step": 7865 + }, + { + "epoch": 0.7093835956170808, + "grad_norm": 1.5118917519002846, + "learning_rate": 8.223036481576522e-07, + "loss": 0.9173, + "step": 7866 + }, + { + "epoch": 0.709473779140551, + "grad_norm": 1.8506448137484548, + "learning_rate": 8.218315138674672e-07, + "loss": 0.9561, + "step": 7867 + }, + { + "epoch": 0.7095639626640213, + "grad_norm": 0.6412998924402801, + "learning_rate": 8.21359480105114e-07, + "loss": 0.7775, + "step": 7868 + }, + { + "epoch": 0.7096541461874916, + "grad_norm": 1.7237329556602436, + "learning_rate": 8.208875469108689e-07, + "loss": 0.9792, + "step": 7869 + }, + { + "epoch": 0.7097443297109618, + "grad_norm": 2.092031770227886, + "learning_rate": 8.204157143249997e-07, + "loss": 0.936, + "step": 7870 + }, + { + "epoch": 0.709834513234432, + "grad_norm": 1.4881898546565917, + "learning_rate": 8.199439823877668e-07, + "loss": 0.9185, + "step": 7871 + }, + { + "epoch": 0.7099246967579024, + "grad_norm": 1.4291284020345942, + "learning_rate": 8.194723511394186e-07, + "loss": 0.9303, + "step": 7872 + }, + { + "epoch": 0.7100148802813726, + "grad_norm": 1.6834699757829936, + "learning_rate": 8.190008206202002e-07, + "loss": 0.8991, + "step": 7873 + }, + { + "epoch": 0.7101050638048428, + "grad_norm": 1.6884017230558517, + "learning_rate": 8.185293908703423e-07, + "loss": 0.9982, + "step": 7874 + }, + { + "epoch": 0.7101952473283131, + "grad_norm": 1.457384981881108, + "learning_rate": 8.180580619300727e-07, + "loss": 0.9099, + "step": 7875 + }, + { + "epoch": 0.7102854308517834, + "grad_norm": 1.2407295740288256, + "learning_rate": 8.175868338396057e-07, + "loss": 0.9005, + "step": 7876 + }, + { + "epoch": 0.7103756143752537, + "grad_norm": 2.136692528154059, + "learning_rate": 8.171157066391499e-07, + "loss": 0.9145, + "step": 7877 + }, + { + "epoch": 0.7104657978987239, + "grad_norm": 1.8065419103098757, + "learning_rate": 8.166446803689045e-07, + "loss": 0.9702, + "step": 7878 + }, + { + "epoch": 0.7105559814221941, + "grad_norm": 0.6990237896586083, + "learning_rate": 8.161737550690595e-07, + "loss": 0.7844, + "step": 7879 + }, + { + "epoch": 0.7106461649456645, + "grad_norm": 1.5252208737373358, + "learning_rate": 8.157029307797976e-07, + "loss": 0.9724, + "step": 7880 + }, + { + "epoch": 0.7107363484691347, + "grad_norm": 1.802516209057268, + "learning_rate": 8.152322075412925e-07, + "loss": 1.0004, + "step": 7881 + }, + { + "epoch": 0.7108265319926049, + "grad_norm": 1.4721976863967798, + "learning_rate": 8.147615853937073e-07, + "loss": 0.8966, + "step": 7882 + }, + { + "epoch": 0.7109167155160752, + "grad_norm": 1.4846552534060102, + "learning_rate": 8.142910643771992e-07, + "loss": 0.8692, + "step": 7883 + }, + { + "epoch": 0.7110068990395455, + "grad_norm": 1.8560501633269535, + "learning_rate": 8.138206445319152e-07, + "loss": 0.9115, + "step": 7884 + }, + { + "epoch": 0.7110970825630157, + "grad_norm": 1.5130326428721248, + "learning_rate": 8.133503258979944e-07, + "loss": 0.851, + "step": 7885 + }, + { + "epoch": 0.711187266086486, + "grad_norm": 1.7055200657308596, + "learning_rate": 8.12880108515567e-07, + "loss": 0.8723, + "step": 7886 + }, + { + "epoch": 0.7112774496099563, + "grad_norm": 1.5289423155197883, + "learning_rate": 8.124099924247543e-07, + "loss": 0.9186, + "step": 7887 + }, + { + "epoch": 0.7113676331334265, + "grad_norm": 1.578040696715621, + "learning_rate": 8.119399776656701e-07, + "loss": 0.8539, + "step": 7888 + }, + { + "epoch": 0.7114578166568968, + "grad_norm": 1.4183386958491926, + "learning_rate": 8.114700642784167e-07, + "loss": 0.9436, + "step": 7889 + }, + { + "epoch": 0.711548000180367, + "grad_norm": 1.4412468716139972, + "learning_rate": 8.110002523030921e-07, + "loss": 1.0013, + "step": 7890 + }, + { + "epoch": 0.7116381837038374, + "grad_norm": 1.459673411146775, + "learning_rate": 8.105305417797808e-07, + "loss": 0.9686, + "step": 7891 + }, + { + "epoch": 0.7117283672273076, + "grad_norm": 1.5273623258041586, + "learning_rate": 8.100609327485635e-07, + "loss": 0.9867, + "step": 7892 + }, + { + "epoch": 0.7118185507507778, + "grad_norm": 1.7608313514107368, + "learning_rate": 8.095914252495082e-07, + "loss": 1.003, + "step": 7893 + }, + { + "epoch": 0.7119087342742481, + "grad_norm": 2.223300404721825, + "learning_rate": 8.091220193226762e-07, + "loss": 0.9135, + "step": 7894 + }, + { + "epoch": 0.7119989177977184, + "grad_norm": 1.4352526304788675, + "learning_rate": 8.0865271500812e-07, + "loss": 0.9343, + "step": 7895 + }, + { + "epoch": 0.7120891013211886, + "grad_norm": 1.6085475535195077, + "learning_rate": 8.081835123458831e-07, + "loss": 0.9855, + "step": 7896 + }, + { + "epoch": 0.7121792848446589, + "grad_norm": 2.809368339759816, + "learning_rate": 8.077144113760013e-07, + "loss": 0.9309, + "step": 7897 + }, + { + "epoch": 0.7122694683681291, + "grad_norm": 1.3841804160978355, + "learning_rate": 8.072454121384995e-07, + "loss": 0.9248, + "step": 7898 + }, + { + "epoch": 0.7123596518915994, + "grad_norm": 1.3396814987656844, + "learning_rate": 8.067765146733958e-07, + "loss": 0.9867, + "step": 7899 + }, + { + "epoch": 0.7124498354150697, + "grad_norm": 1.654544383948034, + "learning_rate": 8.063077190206993e-07, + "loss": 1.0231, + "step": 7900 + }, + { + "epoch": 0.7125400189385399, + "grad_norm": 0.6347772450483891, + "learning_rate": 8.058390252204101e-07, + "loss": 0.695, + "step": 7901 + }, + { + "epoch": 0.7126302024620101, + "grad_norm": 1.4654644680967288, + "learning_rate": 8.0537043331252e-07, + "loss": 0.9433, + "step": 7902 + }, + { + "epoch": 0.7127203859854805, + "grad_norm": 1.7243828457318215, + "learning_rate": 8.049019433370121e-07, + "loss": 0.8864, + "step": 7903 + }, + { + "epoch": 0.7128105695089507, + "grad_norm": 1.5254879968021549, + "learning_rate": 8.044335553338588e-07, + "loss": 0.9484, + "step": 7904 + }, + { + "epoch": 0.712900753032421, + "grad_norm": 1.4381846580415953, + "learning_rate": 8.039652693430281e-07, + "loss": 0.957, + "step": 7905 + }, + { + "epoch": 0.7129909365558912, + "grad_norm": 1.4505494786646616, + "learning_rate": 8.034970854044742e-07, + "loss": 0.9279, + "step": 7906 + }, + { + "epoch": 0.7130811200793615, + "grad_norm": 1.9399199601012205, + "learning_rate": 8.03029003558148e-07, + "loss": 1.039, + "step": 7907 + }, + { + "epoch": 0.7131713036028318, + "grad_norm": 2.020127346821355, + "learning_rate": 8.025610238439864e-07, + "loss": 1.0109, + "step": 7908 + }, + { + "epoch": 0.713261487126302, + "grad_norm": 1.995584979230971, + "learning_rate": 8.020931463019207e-07, + "loss": 1.0194, + "step": 7909 + }, + { + "epoch": 0.7133516706497723, + "grad_norm": 1.6046219794724943, + "learning_rate": 8.016253709718732e-07, + "loss": 0.8901, + "step": 7910 + }, + { + "epoch": 0.7134418541732426, + "grad_norm": 1.6871963510353565, + "learning_rate": 8.011576978937567e-07, + "loss": 0.8493, + "step": 7911 + }, + { + "epoch": 0.7135320376967128, + "grad_norm": 1.7432367560779256, + "learning_rate": 8.006901271074764e-07, + "loss": 0.9446, + "step": 7912 + }, + { + "epoch": 0.713622221220183, + "grad_norm": 1.6688378100205468, + "learning_rate": 8.002226586529261e-07, + "loss": 0.9884, + "step": 7913 + }, + { + "epoch": 0.7137124047436534, + "grad_norm": 1.3725280649758762, + "learning_rate": 7.997552925699956e-07, + "loss": 1.023, + "step": 7914 + }, + { + "epoch": 0.7138025882671236, + "grad_norm": 1.7060525213969149, + "learning_rate": 7.992880288985606e-07, + "loss": 1.0401, + "step": 7915 + }, + { + "epoch": 0.7138927717905938, + "grad_norm": 1.6405339533656065, + "learning_rate": 7.988208676784918e-07, + "loss": 0.9509, + "step": 7916 + }, + { + "epoch": 0.7139829553140641, + "grad_norm": 1.596027990210472, + "learning_rate": 7.983538089496497e-07, + "loss": 0.9441, + "step": 7917 + }, + { + "epoch": 0.7140731388375344, + "grad_norm": 1.4713213711574815, + "learning_rate": 7.978868527518864e-07, + "loss": 0.9039, + "step": 7918 + }, + { + "epoch": 0.7141633223610047, + "grad_norm": 1.4599075358649076, + "learning_rate": 7.974199991250455e-07, + "loss": 0.9229, + "step": 7919 + }, + { + "epoch": 0.7142535058844749, + "grad_norm": 1.8642193174309363, + "learning_rate": 7.969532481089616e-07, + "loss": 1.0442, + "step": 7920 + }, + { + "epoch": 0.7143436894079451, + "grad_norm": 1.6591371527434884, + "learning_rate": 7.964865997434589e-07, + "loss": 0.9606, + "step": 7921 + }, + { + "epoch": 0.7144338729314155, + "grad_norm": 1.7658207202561123, + "learning_rate": 7.96020054068357e-07, + "loss": 1.0198, + "step": 7922 + }, + { + "epoch": 0.7145240564548857, + "grad_norm": 1.2665626914031032, + "learning_rate": 7.95553611123462e-07, + "loss": 0.9038, + "step": 7923 + }, + { + "epoch": 0.7146142399783559, + "grad_norm": 1.967760776072014, + "learning_rate": 7.950872709485741e-07, + "loss": 1.1277, + "step": 7924 + }, + { + "epoch": 0.7147044235018262, + "grad_norm": 1.8375252209444124, + "learning_rate": 7.946210335834842e-07, + "loss": 1.0027, + "step": 7925 + }, + { + "epoch": 0.7147946070252965, + "grad_norm": 1.641746500981905, + "learning_rate": 7.94154899067974e-07, + "loss": 0.8693, + "step": 7926 + }, + { + "epoch": 0.7148847905487667, + "grad_norm": 1.5781289280002166, + "learning_rate": 7.936888674418177e-07, + "loss": 0.8717, + "step": 7927 + }, + { + "epoch": 0.714974974072237, + "grad_norm": 2.2964737729851272, + "learning_rate": 7.932229387447771e-07, + "loss": 0.9453, + "step": 7928 + }, + { + "epoch": 0.7150651575957072, + "grad_norm": 1.8114234094688029, + "learning_rate": 7.927571130166109e-07, + "loss": 0.9409, + "step": 7929 + }, + { + "epoch": 0.7151553411191776, + "grad_norm": 0.6385722184663986, + "learning_rate": 7.922913902970632e-07, + "loss": 0.7688, + "step": 7930 + }, + { + "epoch": 0.7152455246426478, + "grad_norm": 1.6181800486961173, + "learning_rate": 7.918257706258744e-07, + "loss": 0.938, + "step": 7931 + }, + { + "epoch": 0.715335708166118, + "grad_norm": 1.7086976601008246, + "learning_rate": 7.913602540427724e-07, + "loss": 0.9474, + "step": 7932 + }, + { + "epoch": 0.7154258916895884, + "grad_norm": 1.761917001603177, + "learning_rate": 7.908948405874775e-07, + "loss": 0.943, + "step": 7933 + }, + { + "epoch": 0.7155160752130586, + "grad_norm": 1.8868379732929712, + "learning_rate": 7.904295302997019e-07, + "loss": 0.8746, + "step": 7934 + }, + { + "epoch": 0.7156062587365288, + "grad_norm": 1.4385311740471405, + "learning_rate": 7.899643232191484e-07, + "loss": 0.9607, + "step": 7935 + }, + { + "epoch": 0.7156964422599991, + "grad_norm": 1.3261520991023668, + "learning_rate": 7.894992193855108e-07, + "loss": 0.9308, + "step": 7936 + }, + { + "epoch": 0.7157866257834694, + "grad_norm": 1.4880703353614668, + "learning_rate": 7.890342188384751e-07, + "loss": 0.8913, + "step": 7937 + }, + { + "epoch": 0.7158768093069396, + "grad_norm": 1.6848627605071502, + "learning_rate": 7.885693216177165e-07, + "loss": 0.8883, + "step": 7938 + }, + { + "epoch": 0.7159669928304099, + "grad_norm": 1.3897895875340638, + "learning_rate": 7.88104527762903e-07, + "loss": 0.9401, + "step": 7939 + }, + { + "epoch": 0.7160571763538801, + "grad_norm": 1.658899527208529, + "learning_rate": 7.876398373136936e-07, + "loss": 0.9018, + "step": 7940 + }, + { + "epoch": 0.7161473598773505, + "grad_norm": 1.6463733785011765, + "learning_rate": 7.87175250309738e-07, + "loss": 0.947, + "step": 7941 + }, + { + "epoch": 0.7162375434008207, + "grad_norm": 1.7017515051252114, + "learning_rate": 7.867107667906785e-07, + "loss": 1.0237, + "step": 7942 + }, + { + "epoch": 0.7163277269242909, + "grad_norm": 1.7977520336510429, + "learning_rate": 7.862463867961446e-07, + "loss": 0.9429, + "step": 7943 + }, + { + "epoch": 0.7164179104477612, + "grad_norm": 1.4082646244074595, + "learning_rate": 7.857821103657632e-07, + "loss": 1.0083, + "step": 7944 + }, + { + "epoch": 0.7165080939712315, + "grad_norm": 1.4628754679934919, + "learning_rate": 7.853179375391459e-07, + "loss": 0.9767, + "step": 7945 + }, + { + "epoch": 0.7165982774947017, + "grad_norm": 1.84099754876632, + "learning_rate": 7.848538683559012e-07, + "loss": 0.9365, + "step": 7946 + }, + { + "epoch": 0.716688461018172, + "grad_norm": 1.4668048396192717, + "learning_rate": 7.843899028556238e-07, + "loss": 0.9851, + "step": 7947 + }, + { + "epoch": 0.7167786445416422, + "grad_norm": 1.4496996305385366, + "learning_rate": 7.839260410779029e-07, + "loss": 0.9388, + "step": 7948 + }, + { + "epoch": 0.7168688280651125, + "grad_norm": 1.4215440469761345, + "learning_rate": 7.834622830623175e-07, + "loss": 0.9644, + "step": 7949 + }, + { + "epoch": 0.7169590115885828, + "grad_norm": 1.6596851353600648, + "learning_rate": 7.82998628848438e-07, + "loss": 0.9186, + "step": 7950 + }, + { + "epoch": 0.717049195112053, + "grad_norm": 1.4214919695712895, + "learning_rate": 7.825350784758261e-07, + "loss": 0.9387, + "step": 7951 + }, + { + "epoch": 0.7171393786355232, + "grad_norm": 1.3106326033954248, + "learning_rate": 7.820716319840342e-07, + "loss": 0.9078, + "step": 7952 + }, + { + "epoch": 0.7172295621589936, + "grad_norm": 1.7894485923245025, + "learning_rate": 7.816082894126069e-07, + "loss": 1.0296, + "step": 7953 + }, + { + "epoch": 0.7173197456824638, + "grad_norm": 1.5126725374284127, + "learning_rate": 7.811450508010778e-07, + "loss": 0.9129, + "step": 7954 + }, + { + "epoch": 0.717409929205934, + "grad_norm": 1.4624288362322801, + "learning_rate": 7.806819161889737e-07, + "loss": 0.9162, + "step": 7955 + }, + { + "epoch": 0.7175001127294043, + "grad_norm": 1.6585570767517352, + "learning_rate": 7.802188856158119e-07, + "loss": 0.9779, + "step": 7956 + }, + { + "epoch": 0.7175902962528746, + "grad_norm": 1.6263520411526757, + "learning_rate": 7.797559591211002e-07, + "loss": 0.9343, + "step": 7957 + }, + { + "epoch": 0.7176804797763449, + "grad_norm": 1.3710898374844016, + "learning_rate": 7.79293136744339e-07, + "loss": 0.908, + "step": 7958 + }, + { + "epoch": 0.7177706632998151, + "grad_norm": 1.525844062102159, + "learning_rate": 7.788304185250185e-07, + "loss": 0.942, + "step": 7959 + }, + { + "epoch": 0.7178608468232854, + "grad_norm": 1.599832856865559, + "learning_rate": 7.78367804502619e-07, + "loss": 0.9716, + "step": 7960 + }, + { + "epoch": 0.7179510303467557, + "grad_norm": 1.7178995282433702, + "learning_rate": 7.779052947166156e-07, + "loss": 0.9531, + "step": 7961 + }, + { + "epoch": 0.7180412138702259, + "grad_norm": 2.265456199934848, + "learning_rate": 7.774428892064697e-07, + "loss": 0.9014, + "step": 7962 + }, + { + "epoch": 0.7181313973936961, + "grad_norm": 1.595726732058394, + "learning_rate": 7.769805880116391e-07, + "loss": 0.9363, + "step": 7963 + }, + { + "epoch": 0.7182215809171665, + "grad_norm": 2.039388462481663, + "learning_rate": 7.765183911715678e-07, + "loss": 0.9272, + "step": 7964 + }, + { + "epoch": 0.7183117644406367, + "grad_norm": 1.6160238013592028, + "learning_rate": 7.760562987256933e-07, + "loss": 0.8461, + "step": 7965 + }, + { + "epoch": 0.718401947964107, + "grad_norm": 1.669009953514301, + "learning_rate": 7.755943107134444e-07, + "loss": 0.9491, + "step": 7966 + }, + { + "epoch": 0.7184921314875772, + "grad_norm": 0.6702302717800712, + "learning_rate": 7.751324271742401e-07, + "loss": 0.7691, + "step": 7967 + }, + { + "epoch": 0.7185823150110475, + "grad_norm": 1.5075244529600782, + "learning_rate": 7.746706481474916e-07, + "loss": 0.9571, + "step": 7968 + }, + { + "epoch": 0.7186724985345178, + "grad_norm": 0.628732550665194, + "learning_rate": 7.742089736725992e-07, + "loss": 0.7381, + "step": 7969 + }, + { + "epoch": 0.718762682057988, + "grad_norm": 1.6852848207311115, + "learning_rate": 7.737474037889559e-07, + "loss": 0.8612, + "step": 7970 + }, + { + "epoch": 0.7188528655814582, + "grad_norm": 1.6520807584507597, + "learning_rate": 7.732859385359458e-07, + "loss": 0.9159, + "step": 7971 + }, + { + "epoch": 0.7189430491049286, + "grad_norm": 1.8358749054322836, + "learning_rate": 7.728245779529434e-07, + "loss": 0.8649, + "step": 7972 + }, + { + "epoch": 0.7190332326283988, + "grad_norm": 1.6434116493944038, + "learning_rate": 7.723633220793146e-07, + "loss": 0.9745, + "step": 7973 + }, + { + "epoch": 0.719123416151869, + "grad_norm": 1.5738568138372804, + "learning_rate": 7.719021709544162e-07, + "loss": 0.8391, + "step": 7974 + }, + { + "epoch": 0.7192135996753393, + "grad_norm": 1.602903967300428, + "learning_rate": 7.714411246175964e-07, + "loss": 1.0339, + "step": 7975 + }, + { + "epoch": 0.7193037831988096, + "grad_norm": 1.5176870731635457, + "learning_rate": 7.709801831081946e-07, + "loss": 0.8629, + "step": 7976 + }, + { + "epoch": 0.7193939667222798, + "grad_norm": 1.5527198646875071, + "learning_rate": 7.705193464655391e-07, + "loss": 0.9028, + "step": 7977 + }, + { + "epoch": 0.7194841502457501, + "grad_norm": 1.3808184972429165, + "learning_rate": 7.700586147289534e-07, + "loss": 0.9224, + "step": 7978 + }, + { + "epoch": 0.7195743337692203, + "grad_norm": 1.496233263806506, + "learning_rate": 7.695979879377481e-07, + "loss": 0.902, + "step": 7979 + }, + { + "epoch": 0.7196645172926907, + "grad_norm": 1.5772963510602545, + "learning_rate": 7.691374661312266e-07, + "loss": 0.9644, + "step": 7980 + }, + { + "epoch": 0.7197547008161609, + "grad_norm": 1.4070241280840934, + "learning_rate": 7.686770493486834e-07, + "loss": 0.9171, + "step": 7981 + }, + { + "epoch": 0.7198448843396311, + "grad_norm": 1.3795455876913516, + "learning_rate": 7.68216737629404e-07, + "loss": 0.8863, + "step": 7982 + }, + { + "epoch": 0.7199350678631015, + "grad_norm": 1.3662307443335058, + "learning_rate": 7.67756531012665e-07, + "loss": 1.0114, + "step": 7983 + }, + { + "epoch": 0.7200252513865717, + "grad_norm": 1.7156942079130908, + "learning_rate": 7.67296429537732e-07, + "loss": 0.9991, + "step": 7984 + }, + { + "epoch": 0.7201154349100419, + "grad_norm": 1.3941516947393642, + "learning_rate": 7.668364332438661e-07, + "loss": 1.0108, + "step": 7985 + }, + { + "epoch": 0.7202056184335122, + "grad_norm": 1.451006093060394, + "learning_rate": 7.663765421703145e-07, + "loss": 0.9555, + "step": 7986 + }, + { + "epoch": 0.7202958019569825, + "grad_norm": 1.5436394501304058, + "learning_rate": 7.659167563563187e-07, + "loss": 0.8818, + "step": 7987 + }, + { + "epoch": 0.7203859854804527, + "grad_norm": 1.4134845711804171, + "learning_rate": 7.654570758411096e-07, + "loss": 0.9647, + "step": 7988 + }, + { + "epoch": 0.720476169003923, + "grad_norm": 1.5902123176321223, + "learning_rate": 7.649975006639103e-07, + "loss": 1.0034, + "step": 7989 + }, + { + "epoch": 0.7205663525273932, + "grad_norm": 0.6623689035947493, + "learning_rate": 7.645380308639337e-07, + "loss": 0.7527, + "step": 7990 + }, + { + "epoch": 0.7206565360508635, + "grad_norm": 1.9967740984095685, + "learning_rate": 7.640786664803853e-07, + "loss": 0.99, + "step": 7991 + }, + { + "epoch": 0.7207467195743338, + "grad_norm": 1.3480654772412077, + "learning_rate": 7.636194075524587e-07, + "loss": 0.8679, + "step": 7992 + }, + { + "epoch": 0.720836903097804, + "grad_norm": 1.4126759681131407, + "learning_rate": 7.631602541193429e-07, + "loss": 0.8996, + "step": 7993 + }, + { + "epoch": 0.7209270866212742, + "grad_norm": 1.8230912769796301, + "learning_rate": 7.627012062202132e-07, + "loss": 0.9983, + "step": 7994 + }, + { + "epoch": 0.7210172701447446, + "grad_norm": 2.069904099974987, + "learning_rate": 7.622422638942391e-07, + "loss": 0.9456, + "step": 7995 + }, + { + "epoch": 0.7211074536682148, + "grad_norm": 1.3454399128300956, + "learning_rate": 7.617834271805801e-07, + "loss": 1.0114, + "step": 7996 + }, + { + "epoch": 0.7211976371916851, + "grad_norm": 0.6949091609415763, + "learning_rate": 7.613246961183863e-07, + "loss": 0.7661, + "step": 7997 + }, + { + "epoch": 0.7212878207151553, + "grad_norm": 1.5568712702598904, + "learning_rate": 7.608660707468002e-07, + "loss": 0.9028, + "step": 7998 + }, + { + "epoch": 0.7213780042386256, + "grad_norm": 2.037749470609246, + "learning_rate": 7.604075511049522e-07, + "loss": 0.8384, + "step": 7999 + }, + { + "epoch": 0.7214681877620959, + "grad_norm": 1.587706833228745, + "learning_rate": 7.599491372319682e-07, + "loss": 0.8892, + "step": 8000 + }, + { + "epoch": 0.7215583712855661, + "grad_norm": 1.303155193390725, + "learning_rate": 7.594908291669601e-07, + "loss": 0.9416, + "step": 8001 + }, + { + "epoch": 0.7216485548090363, + "grad_norm": 1.544670615037381, + "learning_rate": 7.590326269490359e-07, + "loss": 0.978, + "step": 8002 + }, + { + "epoch": 0.7217387383325067, + "grad_norm": 1.7218526406056627, + "learning_rate": 7.585745306172899e-07, + "loss": 0.9011, + "step": 8003 + }, + { + "epoch": 0.7218289218559769, + "grad_norm": 1.5051818986131202, + "learning_rate": 7.5811654021081e-07, + "loss": 0.9232, + "step": 8004 + }, + { + "epoch": 0.7219191053794471, + "grad_norm": 1.4228847397684803, + "learning_rate": 7.576586557686748e-07, + "loss": 0.9733, + "step": 8005 + }, + { + "epoch": 0.7220092889029175, + "grad_norm": 1.625107908334046, + "learning_rate": 7.572008773299531e-07, + "loss": 0.9131, + "step": 8006 + }, + { + "epoch": 0.7220994724263877, + "grad_norm": 1.6041431755558775, + "learning_rate": 7.567432049337055e-07, + "loss": 0.9821, + "step": 8007 + }, + { + "epoch": 0.722189655949858, + "grad_norm": 1.600398166108934, + "learning_rate": 7.562856386189834e-07, + "loss": 0.8766, + "step": 8008 + }, + { + "epoch": 0.7222798394733282, + "grad_norm": 1.9902895394748457, + "learning_rate": 7.558281784248275e-07, + "loss": 0.9609, + "step": 8009 + }, + { + "epoch": 0.7223700229967985, + "grad_norm": 1.8075290784171996, + "learning_rate": 7.553708243902721e-07, + "loss": 0.9328, + "step": 8010 + }, + { + "epoch": 0.7224602065202688, + "grad_norm": 1.7679208328422147, + "learning_rate": 7.549135765543404e-07, + "loss": 0.838, + "step": 8011 + }, + { + "epoch": 0.722550390043739, + "grad_norm": 1.6207895016402714, + "learning_rate": 7.544564349560481e-07, + "loss": 0.9696, + "step": 8012 + }, + { + "epoch": 0.7226405735672092, + "grad_norm": 1.382890214191029, + "learning_rate": 7.539993996344009e-07, + "loss": 0.9267, + "step": 8013 + }, + { + "epoch": 0.7227307570906796, + "grad_norm": 1.5669464169307463, + "learning_rate": 7.535424706283941e-07, + "loss": 0.9248, + "step": 8014 + }, + { + "epoch": 0.7228209406141498, + "grad_norm": 1.765237006033321, + "learning_rate": 7.530856479770181e-07, + "loss": 0.9046, + "step": 8015 + }, + { + "epoch": 0.72291112413762, + "grad_norm": 1.601605075177241, + "learning_rate": 7.526289317192484e-07, + "loss": 1.0255, + "step": 8016 + }, + { + "epoch": 0.7230013076610903, + "grad_norm": 2.0737252375691906, + "learning_rate": 7.521723218940579e-07, + "loss": 0.9433, + "step": 8017 + }, + { + "epoch": 0.7230914911845606, + "grad_norm": 1.944481239273466, + "learning_rate": 7.517158185404038e-07, + "loss": 1.0036, + "step": 8018 + }, + { + "epoch": 0.7231816747080309, + "grad_norm": 1.5185426766757726, + "learning_rate": 7.512594216972403e-07, + "loss": 0.9671, + "step": 8019 + }, + { + "epoch": 0.7232718582315011, + "grad_norm": 1.393925682251416, + "learning_rate": 7.508031314035078e-07, + "loss": 0.9792, + "step": 8020 + }, + { + "epoch": 0.7233620417549713, + "grad_norm": 1.7511995155523754, + "learning_rate": 7.503469476981401e-07, + "loss": 0.9579, + "step": 8021 + }, + { + "epoch": 0.7234522252784417, + "grad_norm": 1.6669770269704576, + "learning_rate": 7.498908706200613e-07, + "loss": 0.8164, + "step": 8022 + }, + { + "epoch": 0.7235424088019119, + "grad_norm": 1.478714439499799, + "learning_rate": 7.494349002081866e-07, + "loss": 0.9871, + "step": 8023 + }, + { + "epoch": 0.7236325923253821, + "grad_norm": 2.001485392196756, + "learning_rate": 7.489790365014224e-07, + "loss": 0.9715, + "step": 8024 + }, + { + "epoch": 0.7237227758488524, + "grad_norm": 1.7845936896734478, + "learning_rate": 7.485232795386642e-07, + "loss": 0.9623, + "step": 8025 + }, + { + "epoch": 0.7238129593723227, + "grad_norm": 1.7545054883982496, + "learning_rate": 7.480676293588002e-07, + "loss": 0.966, + "step": 8026 + }, + { + "epoch": 0.7239031428957929, + "grad_norm": 0.8196379658445938, + "learning_rate": 7.476120860007093e-07, + "loss": 0.7979, + "step": 8027 + }, + { + "epoch": 0.7239933264192632, + "grad_norm": 1.8050428428743226, + "learning_rate": 7.471566495032608e-07, + "loss": 0.9845, + "step": 8028 + }, + { + "epoch": 0.7240835099427335, + "grad_norm": 1.5125792269803444, + "learning_rate": 7.467013199053152e-07, + "loss": 0.8785, + "step": 8029 + }, + { + "epoch": 0.7241736934662037, + "grad_norm": 1.563448579860651, + "learning_rate": 7.46246097245724e-07, + "loss": 0.9448, + "step": 8030 + }, + { + "epoch": 0.724263876989674, + "grad_norm": 1.8750757202117758, + "learning_rate": 7.457909815633276e-07, + "loss": 0.9501, + "step": 8031 + }, + { + "epoch": 0.7243540605131442, + "grad_norm": 1.8205486357283867, + "learning_rate": 7.453359728969618e-07, + "loss": 0.9374, + "step": 8032 + }, + { + "epoch": 0.7244442440366146, + "grad_norm": 1.504470679644314, + "learning_rate": 7.448810712854475e-07, + "loss": 0.8671, + "step": 8033 + }, + { + "epoch": 0.7245344275600848, + "grad_norm": 1.53261676436849, + "learning_rate": 7.444262767676022e-07, + "loss": 0.9836, + "step": 8034 + }, + { + "epoch": 0.724624611083555, + "grad_norm": 1.4775335761836346, + "learning_rate": 7.439715893822296e-07, + "loss": 1.0175, + "step": 8035 + }, + { + "epoch": 0.7247147946070253, + "grad_norm": 1.4267212995573866, + "learning_rate": 7.435170091681264e-07, + "loss": 0.9635, + "step": 8036 + }, + { + "epoch": 0.7248049781304956, + "grad_norm": 1.603946455996463, + "learning_rate": 7.430625361640803e-07, + "loss": 0.9044, + "step": 8037 + }, + { + "epoch": 0.7248951616539658, + "grad_norm": 1.7241349221071756, + "learning_rate": 7.426081704088694e-07, + "loss": 0.9878, + "step": 8038 + }, + { + "epoch": 0.7249853451774361, + "grad_norm": 1.594305427854709, + "learning_rate": 7.42153911941263e-07, + "loss": 1.0171, + "step": 8039 + }, + { + "epoch": 0.7250755287009063, + "grad_norm": 1.568730254637937, + "learning_rate": 7.416997608000192e-07, + "loss": 0.9849, + "step": 8040 + }, + { + "epoch": 0.7251657122243766, + "grad_norm": 1.7425575034506124, + "learning_rate": 7.412457170238918e-07, + "loss": 0.9258, + "step": 8041 + }, + { + "epoch": 0.7252558957478469, + "grad_norm": 1.634360059155649, + "learning_rate": 7.407917806516193e-07, + "loss": 0.8895, + "step": 8042 + }, + { + "epoch": 0.7253460792713171, + "grad_norm": 1.6201612617149541, + "learning_rate": 7.403379517219354e-07, + "loss": 0.9363, + "step": 8043 + }, + { + "epoch": 0.7254362627947873, + "grad_norm": 1.5065781675076773, + "learning_rate": 7.398842302735636e-07, + "loss": 0.9718, + "step": 8044 + }, + { + "epoch": 0.7255264463182577, + "grad_norm": 0.7311857244168566, + "learning_rate": 7.394306163452171e-07, + "loss": 0.7479, + "step": 8045 + }, + { + "epoch": 0.7256166298417279, + "grad_norm": 1.2645710925159719, + "learning_rate": 7.38977109975601e-07, + "loss": 0.9227, + "step": 8046 + }, + { + "epoch": 0.7257068133651982, + "grad_norm": 1.9853474194304717, + "learning_rate": 7.385237112034119e-07, + "loss": 0.8659, + "step": 8047 + }, + { + "epoch": 0.7257969968886684, + "grad_norm": 1.5436948202673082, + "learning_rate": 7.380704200673342e-07, + "loss": 1.0062, + "step": 8048 + }, + { + "epoch": 0.7258871804121387, + "grad_norm": 1.480181099920551, + "learning_rate": 7.376172366060478e-07, + "loss": 0.9698, + "step": 8049 + }, + { + "epoch": 0.725977363935609, + "grad_norm": 0.6894893690746143, + "learning_rate": 7.371641608582187e-07, + "loss": 0.7135, + "step": 8050 + }, + { + "epoch": 0.7260675474590792, + "grad_norm": 2.043645626284386, + "learning_rate": 7.367111928625067e-07, + "loss": 0.8047, + "step": 8051 + }, + { + "epoch": 0.7261577309825495, + "grad_norm": 1.6548124588565365, + "learning_rate": 7.362583326575613e-07, + "loss": 0.8566, + "step": 8052 + }, + { + "epoch": 0.7262479145060198, + "grad_norm": 1.8040705034853877, + "learning_rate": 7.358055802820234e-07, + "loss": 0.9306, + "step": 8053 + }, + { + "epoch": 0.72633809802949, + "grad_norm": 2.2251481767669623, + "learning_rate": 7.353529357745245e-07, + "loss": 0.9138, + "step": 8054 + }, + { + "epoch": 0.7264282815529602, + "grad_norm": 2.444991118058062, + "learning_rate": 7.349003991736851e-07, + "loss": 0.9826, + "step": 8055 + }, + { + "epoch": 0.7265184650764306, + "grad_norm": 1.5172103770109606, + "learning_rate": 7.344479705181206e-07, + "loss": 0.7903, + "step": 8056 + }, + { + "epoch": 0.7266086485999008, + "grad_norm": 1.709204436166489, + "learning_rate": 7.339956498464322e-07, + "loss": 0.8954, + "step": 8057 + }, + { + "epoch": 0.726698832123371, + "grad_norm": 1.5951521446346335, + "learning_rate": 7.335434371972169e-07, + "loss": 0.9927, + "step": 8058 + }, + { + "epoch": 0.7267890156468413, + "grad_norm": 1.433871125456868, + "learning_rate": 7.33091332609058e-07, + "loss": 0.9278, + "step": 8059 + }, + { + "epoch": 0.7268791991703116, + "grad_norm": 1.6476361826528825, + "learning_rate": 7.326393361205323e-07, + "loss": 0.9626, + "step": 8060 + }, + { + "epoch": 0.7269693826937819, + "grad_norm": 1.734415277236785, + "learning_rate": 7.321874477702068e-07, + "loss": 1.022, + "step": 8061 + }, + { + "epoch": 0.7270595662172521, + "grad_norm": 0.8251938317641115, + "learning_rate": 7.317356675966386e-07, + "loss": 0.7928, + "step": 8062 + }, + { + "epoch": 0.7271497497407223, + "grad_norm": 1.4648588866405445, + "learning_rate": 7.312839956383765e-07, + "loss": 0.9789, + "step": 8063 + }, + { + "epoch": 0.7272399332641927, + "grad_norm": 1.5709414764469314, + "learning_rate": 7.308324319339603e-07, + "loss": 0.9251, + "step": 8064 + }, + { + "epoch": 0.7273301167876629, + "grad_norm": 1.5854774897868236, + "learning_rate": 7.303809765219182e-07, + "loss": 0.9456, + "step": 8065 + }, + { + "epoch": 0.7274203003111331, + "grad_norm": 1.605096481489361, + "learning_rate": 7.299296294407719e-07, + "loss": 0.9877, + "step": 8066 + }, + { + "epoch": 0.7275104838346034, + "grad_norm": 1.470004418554607, + "learning_rate": 7.294783907290327e-07, + "loss": 0.9821, + "step": 8067 + }, + { + "epoch": 0.7276006673580737, + "grad_norm": 1.5766651464190153, + "learning_rate": 7.290272604252028e-07, + "loss": 0.8103, + "step": 8068 + }, + { + "epoch": 0.727690850881544, + "grad_norm": 1.7802826111641386, + "learning_rate": 7.285762385677758e-07, + "loss": 0.8514, + "step": 8069 + }, + { + "epoch": 0.7277810344050142, + "grad_norm": 1.494304652965393, + "learning_rate": 7.281253251952335e-07, + "loss": 0.9773, + "step": 8070 + }, + { + "epoch": 0.7278712179284844, + "grad_norm": 1.3252508591751857, + "learning_rate": 7.276745203460526e-07, + "loss": 0.9241, + "step": 8071 + }, + { + "epoch": 0.7279614014519548, + "grad_norm": 11.720570822084897, + "learning_rate": 7.272238240586959e-07, + "loss": 0.8333, + "step": 8072 + }, + { + "epoch": 0.728051584975425, + "grad_norm": 1.2220244313839854, + "learning_rate": 7.267732363716219e-07, + "loss": 0.8801, + "step": 8073 + }, + { + "epoch": 0.7281417684988952, + "grad_norm": 1.6163556451928598, + "learning_rate": 7.263227573232753e-07, + "loss": 0.8992, + "step": 8074 + }, + { + "epoch": 0.7282319520223655, + "grad_norm": 1.4636049449268929, + "learning_rate": 7.258723869520937e-07, + "loss": 1.0311, + "step": 8075 + }, + { + "epoch": 0.7283221355458358, + "grad_norm": 2.627713481241029, + "learning_rate": 7.254221252965059e-07, + "loss": 0.9904, + "step": 8076 + }, + { + "epoch": 0.728412319069306, + "grad_norm": 1.8789158937837227, + "learning_rate": 7.249719723949301e-07, + "loss": 0.8593, + "step": 8077 + }, + { + "epoch": 0.7285025025927763, + "grad_norm": 1.7253146257820984, + "learning_rate": 7.245219282857761e-07, + "loss": 0.9638, + "step": 8078 + }, + { + "epoch": 0.7285926861162466, + "grad_norm": 1.4529511542567024, + "learning_rate": 7.240719930074442e-07, + "loss": 0.9833, + "step": 8079 + }, + { + "epoch": 0.7286828696397168, + "grad_norm": 1.6377975077125777, + "learning_rate": 7.236221665983257e-07, + "loss": 0.9736, + "step": 8080 + }, + { + "epoch": 0.7287730531631871, + "grad_norm": 0.7418716963426905, + "learning_rate": 7.231724490968012e-07, + "loss": 0.7586, + "step": 8081 + }, + { + "epoch": 0.7288632366866573, + "grad_norm": 1.474093679228149, + "learning_rate": 7.227228405412438e-07, + "loss": 0.9134, + "step": 8082 + }, + { + "epoch": 0.7289534202101277, + "grad_norm": 1.6201566262529399, + "learning_rate": 7.222733409700165e-07, + "loss": 0.9075, + "step": 8083 + }, + { + "epoch": 0.7290436037335979, + "grad_norm": 1.4622079965288506, + "learning_rate": 7.21823950421473e-07, + "loss": 0.9556, + "step": 8084 + }, + { + "epoch": 0.7291337872570681, + "grad_norm": 1.9508055685816825, + "learning_rate": 7.213746689339577e-07, + "loss": 0.9154, + "step": 8085 + }, + { + "epoch": 0.7292239707805384, + "grad_norm": 1.633377050606204, + "learning_rate": 7.20925496545807e-07, + "loss": 0.9448, + "step": 8086 + }, + { + "epoch": 0.7293141543040087, + "grad_norm": 1.441147778501348, + "learning_rate": 7.20476433295344e-07, + "loss": 0.8529, + "step": 8087 + }, + { + "epoch": 0.7294043378274789, + "grad_norm": 1.6309755934918715, + "learning_rate": 7.200274792208882e-07, + "loss": 0.9825, + "step": 8088 + }, + { + "epoch": 0.7294945213509492, + "grad_norm": 1.6767620071065399, + "learning_rate": 7.195786343607444e-07, + "loss": 0.8347, + "step": 8089 + }, + { + "epoch": 0.7295847048744194, + "grad_norm": 1.8357910097807422, + "learning_rate": 7.191298987532131e-07, + "loss": 0.9332, + "step": 8090 + }, + { + "epoch": 0.7296748883978897, + "grad_norm": 1.4973513743116798, + "learning_rate": 7.186812724365805e-07, + "loss": 0.9515, + "step": 8091 + }, + { + "epoch": 0.72976507192136, + "grad_norm": 1.6529717300406928, + "learning_rate": 7.182327554491272e-07, + "loss": 0.9519, + "step": 8092 + }, + { + "epoch": 0.7298552554448302, + "grad_norm": 0.7372381521355039, + "learning_rate": 7.177843478291225e-07, + "loss": 0.7873, + "step": 8093 + }, + { + "epoch": 0.7299454389683004, + "grad_norm": 1.6723278537533999, + "learning_rate": 7.173360496148276e-07, + "loss": 0.9774, + "step": 8094 + }, + { + "epoch": 0.7300356224917708, + "grad_norm": 2.3388046988977473, + "learning_rate": 7.168878608444939e-07, + "loss": 0.9596, + "step": 8095 + }, + { + "epoch": 0.730125806015241, + "grad_norm": 1.768109354521003, + "learning_rate": 7.164397815563623e-07, + "loss": 0.9474, + "step": 8096 + }, + { + "epoch": 0.7302159895387113, + "grad_norm": 1.7880328017901992, + "learning_rate": 7.159918117886661e-07, + "loss": 0.9544, + "step": 8097 + }, + { + "epoch": 0.7303061730621815, + "grad_norm": 1.8228980435827928, + "learning_rate": 7.155439515796284e-07, + "loss": 0.9613, + "step": 8098 + }, + { + "epoch": 0.7303963565856518, + "grad_norm": 1.8946478424376487, + "learning_rate": 7.150962009674633e-07, + "loss": 0.938, + "step": 8099 + }, + { + "epoch": 0.7304865401091221, + "grad_norm": 1.756775885208184, + "learning_rate": 7.146485599903751e-07, + "loss": 0.9914, + "step": 8100 + }, + { + "epoch": 0.7305767236325923, + "grad_norm": 1.518637033664628, + "learning_rate": 7.142010286865592e-07, + "loss": 1.0133, + "step": 8101 + }, + { + "epoch": 0.7306669071560626, + "grad_norm": 0.5803883041148522, + "learning_rate": 7.137536070942012e-07, + "loss": 0.7491, + "step": 8102 + }, + { + "epoch": 0.7307570906795329, + "grad_norm": 1.5802445483180803, + "learning_rate": 7.133062952514786e-07, + "loss": 0.9453, + "step": 8103 + }, + { + "epoch": 0.7308472742030031, + "grad_norm": 0.6697282205395938, + "learning_rate": 7.128590931965562e-07, + "loss": 0.8054, + "step": 8104 + }, + { + "epoch": 0.7309374577264733, + "grad_norm": 1.5254229006118465, + "learning_rate": 7.124120009675945e-07, + "loss": 1.0037, + "step": 8105 + }, + { + "epoch": 0.7310276412499437, + "grad_norm": 1.6429687470977101, + "learning_rate": 7.119650186027399e-07, + "loss": 0.9707, + "step": 8106 + }, + { + "epoch": 0.7311178247734139, + "grad_norm": 1.7333337218333078, + "learning_rate": 7.11518146140132e-07, + "loss": 1.0636, + "step": 8107 + }, + { + "epoch": 0.7312080082968841, + "grad_norm": 1.635909579238254, + "learning_rate": 7.110713836179007e-07, + "loss": 0.9037, + "step": 8108 + }, + { + "epoch": 0.7312981918203544, + "grad_norm": 1.5882354774765872, + "learning_rate": 7.106247310741659e-07, + "loss": 0.8748, + "step": 8109 + }, + { + "epoch": 0.7313883753438247, + "grad_norm": 1.598346216125334, + "learning_rate": 7.101781885470393e-07, + "loss": 1.0075, + "step": 8110 + }, + { + "epoch": 0.731478558867295, + "grad_norm": 2.0727350995723666, + "learning_rate": 7.097317560746203e-07, + "loss": 0.9835, + "step": 8111 + }, + { + "epoch": 0.7315687423907652, + "grad_norm": 1.4547790629644746, + "learning_rate": 7.092854336950036e-07, + "loss": 1.0063, + "step": 8112 + }, + { + "epoch": 0.7316589259142354, + "grad_norm": 1.4765608045780196, + "learning_rate": 7.0883922144627e-07, + "loss": 1.0004, + "step": 8113 + }, + { + "epoch": 0.7317491094377058, + "grad_norm": 1.474482206204254, + "learning_rate": 7.083931193664934e-07, + "loss": 0.983, + "step": 8114 + }, + { + "epoch": 0.731839292961176, + "grad_norm": 1.5851107542409928, + "learning_rate": 7.079471274937378e-07, + "loss": 0.9931, + "step": 8115 + }, + { + "epoch": 0.7319294764846462, + "grad_norm": 1.5170121285061198, + "learning_rate": 7.075012458660574e-07, + "loss": 0.9323, + "step": 8116 + }, + { + "epoch": 0.7320196600081165, + "grad_norm": 0.6958839982446133, + "learning_rate": 7.070554745214976e-07, + "loss": 0.7773, + "step": 8117 + }, + { + "epoch": 0.7321098435315868, + "grad_norm": 1.6677320969092, + "learning_rate": 7.066098134980947e-07, + "loss": 0.9035, + "step": 8118 + }, + { + "epoch": 0.732200027055057, + "grad_norm": 1.5826834382199886, + "learning_rate": 7.061642628338727e-07, + "loss": 0.9517, + "step": 8119 + }, + { + "epoch": 0.7322902105785273, + "grad_norm": 1.3801078855802393, + "learning_rate": 7.057188225668513e-07, + "loss": 0.9493, + "step": 8120 + }, + { + "epoch": 0.7323803941019975, + "grad_norm": 1.5497783040887947, + "learning_rate": 7.052734927350358e-07, + "loss": 1.0925, + "step": 8121 + }, + { + "epoch": 0.7324705776254679, + "grad_norm": 1.7279461455960015, + "learning_rate": 7.048282733764252e-07, + "loss": 0.972, + "step": 8122 + }, + { + "epoch": 0.7325607611489381, + "grad_norm": 1.586502801276522, + "learning_rate": 7.043831645290077e-07, + "loss": 1.0219, + "step": 8123 + }, + { + "epoch": 0.7326509446724083, + "grad_norm": 1.721724000599609, + "learning_rate": 7.039381662307624e-07, + "loss": 0.9182, + "step": 8124 + }, + { + "epoch": 0.7327411281958787, + "grad_norm": 5.162254767331511, + "learning_rate": 7.034932785196601e-07, + "loss": 0.9307, + "step": 8125 + }, + { + "epoch": 0.7328313117193489, + "grad_norm": 1.663153314933596, + "learning_rate": 7.030485014336585e-07, + "loss": 0.9321, + "step": 8126 + }, + { + "epoch": 0.7329214952428191, + "grad_norm": 1.6359345735744182, + "learning_rate": 7.026038350107118e-07, + "loss": 0.9081, + "step": 8127 + }, + { + "epoch": 0.7330116787662894, + "grad_norm": 1.4632527981510908, + "learning_rate": 7.021592792887579e-07, + "loss": 1.0786, + "step": 8128 + }, + { + "epoch": 0.7331018622897597, + "grad_norm": 1.5830465190845446, + "learning_rate": 7.01714834305732e-07, + "loss": 0.9971, + "step": 8129 + }, + { + "epoch": 0.7331920458132299, + "grad_norm": 1.6792287510579125, + "learning_rate": 7.012705000995544e-07, + "loss": 0.9093, + "step": 8130 + }, + { + "epoch": 0.7332822293367002, + "grad_norm": 1.480285311081404, + "learning_rate": 7.008262767081392e-07, + "loss": 0.9422, + "step": 8131 + }, + { + "epoch": 0.7333724128601704, + "grad_norm": 1.836673475427117, + "learning_rate": 7.003821641693892e-07, + "loss": 1.0007, + "step": 8132 + }, + { + "epoch": 0.7334625963836408, + "grad_norm": 1.4597977052613333, + "learning_rate": 6.999381625211993e-07, + "loss": 0.9376, + "step": 8133 + }, + { + "epoch": 0.733552779907111, + "grad_norm": 1.54301371448526, + "learning_rate": 6.994942718014536e-07, + "loss": 0.866, + "step": 8134 + }, + { + "epoch": 0.7336429634305812, + "grad_norm": 1.8723679824759736, + "learning_rate": 6.990504920480282e-07, + "loss": 0.8562, + "step": 8135 + }, + { + "epoch": 0.7337331469540515, + "grad_norm": 1.5613354730230362, + "learning_rate": 6.986068232987879e-07, + "loss": 0.9429, + "step": 8136 + }, + { + "epoch": 0.7338233304775218, + "grad_norm": 1.3139787699437362, + "learning_rate": 6.981632655915888e-07, + "loss": 0.9001, + "step": 8137 + }, + { + "epoch": 0.733913514000992, + "grad_norm": 1.630834668366372, + "learning_rate": 6.977198189642783e-07, + "loss": 1.0107, + "step": 8138 + }, + { + "epoch": 0.7340036975244623, + "grad_norm": 1.6592471517101888, + "learning_rate": 6.972764834546935e-07, + "loss": 0.9754, + "step": 8139 + }, + { + "epoch": 0.7340938810479325, + "grad_norm": 1.5127694670181597, + "learning_rate": 6.96833259100663e-07, + "loss": 0.9627, + "step": 8140 + }, + { + "epoch": 0.7341840645714028, + "grad_norm": 1.6269932040147046, + "learning_rate": 6.96390145940003e-07, + "loss": 0.9219, + "step": 8141 + }, + { + "epoch": 0.7342742480948731, + "grad_norm": 1.8423714737597323, + "learning_rate": 6.959471440105253e-07, + "loss": 0.8992, + "step": 8142 + }, + { + "epoch": 0.7343644316183433, + "grad_norm": 1.681779135782685, + "learning_rate": 6.955042533500261e-07, + "loss": 0.9802, + "step": 8143 + }, + { + "epoch": 0.7344546151418135, + "grad_norm": 1.5939490998013608, + "learning_rate": 6.950614739962986e-07, + "loss": 0.9099, + "step": 8144 + }, + { + "epoch": 0.7345447986652839, + "grad_norm": 1.4339608286371688, + "learning_rate": 6.946188059871198e-07, + "loss": 0.8382, + "step": 8145 + }, + { + "epoch": 0.7346349821887541, + "grad_norm": 1.4968007461511332, + "learning_rate": 6.941762493602638e-07, + "loss": 1.0672, + "step": 8146 + }, + { + "epoch": 0.7347251657122243, + "grad_norm": 1.5580621639821968, + "learning_rate": 6.937338041534899e-07, + "loss": 0.8842, + "step": 8147 + }, + { + "epoch": 0.7348153492356947, + "grad_norm": 1.4075022737745548, + "learning_rate": 6.932914704045505e-07, + "loss": 0.9447, + "step": 8148 + }, + { + "epoch": 0.7349055327591649, + "grad_norm": 1.607623063355841, + "learning_rate": 6.928492481511878e-07, + "loss": 0.975, + "step": 8149 + }, + { + "epoch": 0.7349957162826352, + "grad_norm": 1.7121357662700447, + "learning_rate": 6.924071374311349e-07, + "loss": 0.8956, + "step": 8150 + }, + { + "epoch": 0.7350858998061054, + "grad_norm": 0.6475220848426431, + "learning_rate": 6.919651382821157e-07, + "loss": 0.7595, + "step": 8151 + }, + { + "epoch": 0.7351760833295757, + "grad_norm": 1.569637400188164, + "learning_rate": 6.915232507418425e-07, + "loss": 0.919, + "step": 8152 + }, + { + "epoch": 0.735266266853046, + "grad_norm": 1.5165045634124648, + "learning_rate": 6.910814748480204e-07, + "loss": 0.8326, + "step": 8153 + }, + { + "epoch": 0.7353564503765162, + "grad_norm": 2.0148965864882356, + "learning_rate": 6.906398106383445e-07, + "loss": 0.962, + "step": 8154 + }, + { + "epoch": 0.7354466338999864, + "grad_norm": 1.4014363447928948, + "learning_rate": 6.901982581504994e-07, + "loss": 0.9416, + "step": 8155 + }, + { + "epoch": 0.7355368174234568, + "grad_norm": 1.7581102585267754, + "learning_rate": 6.897568174221611e-07, + "loss": 0.822, + "step": 8156 + }, + { + "epoch": 0.735627000946927, + "grad_norm": 1.5157191453775942, + "learning_rate": 6.893154884909966e-07, + "loss": 1.0154, + "step": 8157 + }, + { + "epoch": 0.7357171844703972, + "grad_norm": 0.6451624005302368, + "learning_rate": 6.888742713946602e-07, + "loss": 0.7843, + "step": 8158 + }, + { + "epoch": 0.7358073679938675, + "grad_norm": 1.9083870337425375, + "learning_rate": 6.884331661708018e-07, + "loss": 0.9471, + "step": 8159 + }, + { + "epoch": 0.7358975515173378, + "grad_norm": 1.7209320089329843, + "learning_rate": 6.879921728570561e-07, + "loss": 0.9418, + "step": 8160 + }, + { + "epoch": 0.735987735040808, + "grad_norm": 1.4680152841026999, + "learning_rate": 6.875512914910539e-07, + "loss": 1.0238, + "step": 8161 + }, + { + "epoch": 0.7360779185642783, + "grad_norm": 1.5282147745858805, + "learning_rate": 6.871105221104119e-07, + "loss": 0.8975, + "step": 8162 + }, + { + "epoch": 0.7361681020877485, + "grad_norm": 1.7929291897400736, + "learning_rate": 6.866698647527391e-07, + "loss": 0.9265, + "step": 8163 + }, + { + "epoch": 0.7362582856112189, + "grad_norm": 1.7419052237471062, + "learning_rate": 6.862293194556353e-07, + "loss": 0.9698, + "step": 8164 + }, + { + "epoch": 0.7363484691346891, + "grad_norm": 1.5755452135790131, + "learning_rate": 6.857888862566896e-07, + "loss": 0.8911, + "step": 8165 + }, + { + "epoch": 0.7364386526581593, + "grad_norm": 1.468601949818847, + "learning_rate": 6.853485651934836e-07, + "loss": 0.9122, + "step": 8166 + }, + { + "epoch": 0.7365288361816296, + "grad_norm": 1.4957797764334837, + "learning_rate": 6.849083563035855e-07, + "loss": 0.8008, + "step": 8167 + }, + { + "epoch": 0.7366190197050999, + "grad_norm": 0.7683372335341865, + "learning_rate": 6.844682596245592e-07, + "loss": 0.8623, + "step": 8168 + }, + { + "epoch": 0.7367092032285701, + "grad_norm": 1.620609293786914, + "learning_rate": 6.840282751939539e-07, + "loss": 1.0067, + "step": 8169 + }, + { + "epoch": 0.7367993867520404, + "grad_norm": 1.5275306684846006, + "learning_rate": 6.835884030493126e-07, + "loss": 0.9913, + "step": 8170 + }, + { + "epoch": 0.7368895702755107, + "grad_norm": 1.8372025696729632, + "learning_rate": 6.831486432281672e-07, + "loss": 0.9694, + "step": 8171 + }, + { + "epoch": 0.736979753798981, + "grad_norm": 1.7687156431278368, + "learning_rate": 6.827089957680407e-07, + "loss": 0.9595, + "step": 8172 + }, + { + "epoch": 0.7370699373224512, + "grad_norm": 1.4943919250984865, + "learning_rate": 6.822694607064461e-07, + "loss": 1.0627, + "step": 8173 + }, + { + "epoch": 0.7371601208459214, + "grad_norm": 1.644253378835776, + "learning_rate": 6.818300380808877e-07, + "loss": 0.9693, + "step": 8174 + }, + { + "epoch": 0.7372503043693918, + "grad_norm": 1.647794625271971, + "learning_rate": 6.813907279288574e-07, + "loss": 0.9392, + "step": 8175 + }, + { + "epoch": 0.737340487892862, + "grad_norm": 1.501485089421431, + "learning_rate": 6.809515302878422e-07, + "loss": 0.9702, + "step": 8176 + }, + { + "epoch": 0.7374306714163322, + "grad_norm": 1.5150623653656194, + "learning_rate": 6.80512445195315e-07, + "loss": 0.9385, + "step": 8177 + }, + { + "epoch": 0.7375208549398025, + "grad_norm": 1.4938670188365668, + "learning_rate": 6.800734726887416e-07, + "loss": 0.9644, + "step": 8178 + }, + { + "epoch": 0.7376110384632728, + "grad_norm": 1.5540114423264295, + "learning_rate": 6.796346128055775e-07, + "loss": 0.9465, + "step": 8179 + }, + { + "epoch": 0.737701221986743, + "grad_norm": 1.635806683080183, + "learning_rate": 6.791958655832684e-07, + "loss": 0.921, + "step": 8180 + }, + { + "epoch": 0.7377914055102133, + "grad_norm": 1.7525182044461858, + "learning_rate": 6.787572310592518e-07, + "loss": 0.9092, + "step": 8181 + }, + { + "epoch": 0.7378815890336835, + "grad_norm": 1.6901408124417763, + "learning_rate": 6.783187092709521e-07, + "loss": 1.0355, + "step": 8182 + }, + { + "epoch": 0.7379717725571538, + "grad_norm": 1.4853716265184167, + "learning_rate": 6.778803002557891e-07, + "loss": 0.9522, + "step": 8183 + }, + { + "epoch": 0.7380619560806241, + "grad_norm": 1.343050841373944, + "learning_rate": 6.774420040511686e-07, + "loss": 0.9822, + "step": 8184 + }, + { + "epoch": 0.7381521396040943, + "grad_norm": 2.060042911749241, + "learning_rate": 6.770038206944886e-07, + "loss": 0.9461, + "step": 8185 + }, + { + "epoch": 0.7382423231275645, + "grad_norm": 1.3236122205172989, + "learning_rate": 6.765657502231375e-07, + "loss": 0.9634, + "step": 8186 + }, + { + "epoch": 0.7383325066510349, + "grad_norm": 1.7413202701802026, + "learning_rate": 6.761277926744939e-07, + "loss": 1.0161, + "step": 8187 + }, + { + "epoch": 0.7384226901745051, + "grad_norm": 1.5155601880875007, + "learning_rate": 6.756899480859268e-07, + "loss": 0.9403, + "step": 8188 + }, + { + "epoch": 0.7385128736979754, + "grad_norm": 1.6156811909154913, + "learning_rate": 6.752522164947956e-07, + "loss": 0.9955, + "step": 8189 + }, + { + "epoch": 0.7386030572214456, + "grad_norm": 1.2166783748133039, + "learning_rate": 6.748145979384498e-07, + "loss": 1.0022, + "step": 8190 + }, + { + "epoch": 0.7386932407449159, + "grad_norm": 1.6117142648121365, + "learning_rate": 6.743770924542303e-07, + "loss": 0.8881, + "step": 8191 + }, + { + "epoch": 0.7387834242683862, + "grad_norm": 0.6685030400985261, + "learning_rate": 6.739397000794658e-07, + "loss": 0.7716, + "step": 8192 + }, + { + "epoch": 0.7388736077918564, + "grad_norm": 2.166770700255699, + "learning_rate": 6.735024208514782e-07, + "loss": 0.9226, + "step": 8193 + }, + { + "epoch": 0.7389637913153266, + "grad_norm": 1.2537952504687364, + "learning_rate": 6.73065254807578e-07, + "loss": 0.8624, + "step": 8194 + }, + { + "epoch": 0.739053974838797, + "grad_norm": 1.435649842183097, + "learning_rate": 6.726282019850669e-07, + "loss": 0.9983, + "step": 8195 + }, + { + "epoch": 0.7391441583622672, + "grad_norm": 1.2469272516531276, + "learning_rate": 6.721912624212376e-07, + "loss": 1.0177, + "step": 8196 + }, + { + "epoch": 0.7392343418857374, + "grad_norm": 1.769352783904477, + "learning_rate": 6.717544361533696e-07, + "loss": 0.8439, + "step": 8197 + }, + { + "epoch": 0.7393245254092078, + "grad_norm": 1.3221119463148958, + "learning_rate": 6.713177232187386e-07, + "loss": 0.9292, + "step": 8198 + }, + { + "epoch": 0.739414708932678, + "grad_norm": 1.8184109034136273, + "learning_rate": 6.708811236546041e-07, + "loss": 0.9407, + "step": 8199 + }, + { + "epoch": 0.7395048924561483, + "grad_norm": 2.9744194105359365, + "learning_rate": 6.704446374982224e-07, + "loss": 0.9124, + "step": 8200 + }, + { + "epoch": 0.7395950759796185, + "grad_norm": 1.6459968179878264, + "learning_rate": 6.700082647868346e-07, + "loss": 0.9857, + "step": 8201 + }, + { + "epoch": 0.7396852595030888, + "grad_norm": 1.4298946043001695, + "learning_rate": 6.695720055576751e-07, + "loss": 1.0004, + "step": 8202 + }, + { + "epoch": 0.7397754430265591, + "grad_norm": 1.9601780001914857, + "learning_rate": 6.691358598479679e-07, + "loss": 0.9603, + "step": 8203 + }, + { + "epoch": 0.7398656265500293, + "grad_norm": 1.5363262658552903, + "learning_rate": 6.686998276949276e-07, + "loss": 0.9391, + "step": 8204 + }, + { + "epoch": 0.7399558100734995, + "grad_norm": 0.6889282476383023, + "learning_rate": 6.682639091357587e-07, + "loss": 0.7873, + "step": 8205 + }, + { + "epoch": 0.7400459935969699, + "grad_norm": 1.6070772081923852, + "learning_rate": 6.678281042076568e-07, + "loss": 0.936, + "step": 8206 + }, + { + "epoch": 0.7401361771204401, + "grad_norm": 1.502102411636622, + "learning_rate": 6.673924129478059e-07, + "loss": 0.9486, + "step": 8207 + }, + { + "epoch": 0.7402263606439103, + "grad_norm": 1.4262530656626164, + "learning_rate": 6.669568353933824e-07, + "loss": 0.8695, + "step": 8208 + }, + { + "epoch": 0.7403165441673806, + "grad_norm": 1.3541358748627599, + "learning_rate": 6.665213715815519e-07, + "loss": 0.8837, + "step": 8209 + }, + { + "epoch": 0.7404067276908509, + "grad_norm": 2.1454949220140755, + "learning_rate": 6.660860215494706e-07, + "loss": 0.9633, + "step": 8210 + }, + { + "epoch": 0.7404969112143212, + "grad_norm": 1.6953742372021292, + "learning_rate": 6.656507853342852e-07, + "loss": 0.9022, + "step": 8211 + }, + { + "epoch": 0.7405870947377914, + "grad_norm": 1.381906029099672, + "learning_rate": 6.652156629731323e-07, + "loss": 0.9596, + "step": 8212 + }, + { + "epoch": 0.7406772782612616, + "grad_norm": 1.8848229908489778, + "learning_rate": 6.647806545031396e-07, + "loss": 1.0137, + "step": 8213 + }, + { + "epoch": 0.740767461784732, + "grad_norm": 1.5739923887105134, + "learning_rate": 6.643457599614224e-07, + "loss": 0.8682, + "step": 8214 + }, + { + "epoch": 0.7408576453082022, + "grad_norm": 1.9200830556866189, + "learning_rate": 6.63910979385091e-07, + "loss": 0.9983, + "step": 8215 + }, + { + "epoch": 0.7409478288316724, + "grad_norm": 1.6901600676129591, + "learning_rate": 6.634763128112409e-07, + "loss": 1.0342, + "step": 8216 + }, + { + "epoch": 0.7410380123551427, + "grad_norm": 1.769203206153809, + "learning_rate": 6.630417602769622e-07, + "loss": 0.9341, + "step": 8217 + }, + { + "epoch": 0.741128195878613, + "grad_norm": 1.7680045103602464, + "learning_rate": 6.62607321819332e-07, + "loss": 0.9528, + "step": 8218 + }, + { + "epoch": 0.7412183794020832, + "grad_norm": 1.299770019435774, + "learning_rate": 6.621729974754196e-07, + "loss": 0.9486, + "step": 8219 + }, + { + "epoch": 0.7413085629255535, + "grad_norm": 1.6321243223072184, + "learning_rate": 6.617387872822835e-07, + "loss": 0.895, + "step": 8220 + }, + { + "epoch": 0.7413987464490238, + "grad_norm": 1.6303365086225776, + "learning_rate": 6.613046912769731e-07, + "loss": 0.9198, + "step": 8221 + }, + { + "epoch": 0.741488929972494, + "grad_norm": 2.0661030134016323, + "learning_rate": 6.608707094965289e-07, + "loss": 1.0144, + "step": 8222 + }, + { + "epoch": 0.7415791134959643, + "grad_norm": 2.3312253853569724, + "learning_rate": 6.604368419779787e-07, + "loss": 0.8799, + "step": 8223 + }, + { + "epoch": 0.7416692970194345, + "grad_norm": 1.6116667050766307, + "learning_rate": 6.600030887583434e-07, + "loss": 0.9563, + "step": 8224 + }, + { + "epoch": 0.7417594805429049, + "grad_norm": 1.9021900906082192, + "learning_rate": 6.595694498746336e-07, + "loss": 1.0601, + "step": 8225 + }, + { + "epoch": 0.7418496640663751, + "grad_norm": 1.9642847680423985, + "learning_rate": 6.591359253638491e-07, + "loss": 0.9183, + "step": 8226 + }, + { + "epoch": 0.7419398475898453, + "grad_norm": 1.7063540025251482, + "learning_rate": 6.587025152629808e-07, + "loss": 0.8462, + "step": 8227 + }, + { + "epoch": 0.7420300311133156, + "grad_norm": 1.9173042715710373, + "learning_rate": 6.582692196090107e-07, + "loss": 0.9921, + "step": 8228 + }, + { + "epoch": 0.7421202146367859, + "grad_norm": 1.5325147893792535, + "learning_rate": 6.578360384389074e-07, + "loss": 0.9636, + "step": 8229 + }, + { + "epoch": 0.7422103981602561, + "grad_norm": 1.3661949696674531, + "learning_rate": 6.574029717896355e-07, + "loss": 0.9273, + "step": 8230 + }, + { + "epoch": 0.7423005816837264, + "grad_norm": 1.6065989875916584, + "learning_rate": 6.569700196981436e-07, + "loss": 0.9395, + "step": 8231 + }, + { + "epoch": 0.7423907652071966, + "grad_norm": 1.5970355555972056, + "learning_rate": 6.565371822013763e-07, + "loss": 0.8533, + "step": 8232 + }, + { + "epoch": 0.742480948730667, + "grad_norm": 2.0222008656269495, + "learning_rate": 6.561044593362636e-07, + "loss": 1.005, + "step": 8233 + }, + { + "epoch": 0.7425711322541372, + "grad_norm": 2.269854576729424, + "learning_rate": 6.556718511397288e-07, + "loss": 0.9327, + "step": 8234 + }, + { + "epoch": 0.7426613157776074, + "grad_norm": 1.439879562586621, + "learning_rate": 6.552393576486843e-07, + "loss": 1.0642, + "step": 8235 + }, + { + "epoch": 0.7427514993010776, + "grad_norm": 1.7972259842120653, + "learning_rate": 6.548069789000325e-07, + "loss": 0.9669, + "step": 8236 + }, + { + "epoch": 0.742841682824548, + "grad_norm": 3.553750788755326, + "learning_rate": 6.543747149306673e-07, + "loss": 0.9358, + "step": 8237 + }, + { + "epoch": 0.7429318663480182, + "grad_norm": 1.5839739892057019, + "learning_rate": 6.5394256577747e-07, + "loss": 0.9978, + "step": 8238 + }, + { + "epoch": 0.7430220498714885, + "grad_norm": 1.5892917133909985, + "learning_rate": 6.535105314773161e-07, + "loss": 0.975, + "step": 8239 + }, + { + "epoch": 0.7431122333949587, + "grad_norm": 1.8516733643381769, + "learning_rate": 6.530786120670677e-07, + "loss": 0.8818, + "step": 8240 + }, + { + "epoch": 0.743202416918429, + "grad_norm": 1.6126534115463542, + "learning_rate": 6.526468075835787e-07, + "loss": 1.0052, + "step": 8241 + }, + { + "epoch": 0.7432926004418993, + "grad_norm": 1.525335762707792, + "learning_rate": 6.522151180636937e-07, + "loss": 0.9924, + "step": 8242 + }, + { + "epoch": 0.7433827839653695, + "grad_norm": 1.5326642103724455, + "learning_rate": 6.517835435442461e-07, + "loss": 0.9137, + "step": 8243 + }, + { + "epoch": 0.7434729674888398, + "grad_norm": 2.143787136812764, + "learning_rate": 6.513520840620606e-07, + "loss": 0.9701, + "step": 8244 + }, + { + "epoch": 0.7435631510123101, + "grad_norm": 1.6855895212730363, + "learning_rate": 6.509207396539525e-07, + "loss": 0.9137, + "step": 8245 + }, + { + "epoch": 0.7436533345357803, + "grad_norm": 0.6544766990311124, + "learning_rate": 6.50489510356724e-07, + "loss": 0.7536, + "step": 8246 + }, + { + "epoch": 0.7437435180592505, + "grad_norm": 1.7918869444998187, + "learning_rate": 6.500583962071732e-07, + "loss": 0.8783, + "step": 8247 + }, + { + "epoch": 0.7438337015827209, + "grad_norm": 1.505101113202005, + "learning_rate": 6.496273972420827e-07, + "loss": 0.9735, + "step": 8248 + }, + { + "epoch": 0.7439238851061911, + "grad_norm": 1.776810987502283, + "learning_rate": 6.491965134982287e-07, + "loss": 1.0029, + "step": 8249 + }, + { + "epoch": 0.7440140686296614, + "grad_norm": 1.6970607339784283, + "learning_rate": 6.487657450123765e-07, + "loss": 0.9157, + "step": 8250 + }, + { + "epoch": 0.7441042521531316, + "grad_norm": 1.482617110311551, + "learning_rate": 6.483350918212814e-07, + "loss": 0.9622, + "step": 8251 + }, + { + "epoch": 0.7441944356766019, + "grad_norm": 0.7136341137179639, + "learning_rate": 6.479045539616898e-07, + "loss": 0.7855, + "step": 8252 + }, + { + "epoch": 0.7442846192000722, + "grad_norm": 1.84937069731557, + "learning_rate": 6.474741314703358e-07, + "loss": 0.9523, + "step": 8253 + }, + { + "epoch": 0.7443748027235424, + "grad_norm": 2.0251036158296176, + "learning_rate": 6.47043824383948e-07, + "loss": 1.0068, + "step": 8254 + }, + { + "epoch": 0.7444649862470126, + "grad_norm": 1.891106347214889, + "learning_rate": 6.466136327392399e-07, + "loss": 0.9871, + "step": 8255 + }, + { + "epoch": 0.744555169770483, + "grad_norm": 1.6249835673748463, + "learning_rate": 6.461835565729206e-07, + "loss": 0.9858, + "step": 8256 + }, + { + "epoch": 0.7446453532939532, + "grad_norm": 1.4794415052942378, + "learning_rate": 6.457535959216844e-07, + "loss": 0.9813, + "step": 8257 + }, + { + "epoch": 0.7447355368174234, + "grad_norm": 1.4874570471708906, + "learning_rate": 6.453237508222186e-07, + "loss": 0.9108, + "step": 8258 + }, + { + "epoch": 0.7448257203408937, + "grad_norm": 1.454812741240537, + "learning_rate": 6.448940213112e-07, + "loss": 0.9683, + "step": 8259 + }, + { + "epoch": 0.744915903864364, + "grad_norm": 1.3880170735015427, + "learning_rate": 6.444644074252954e-07, + "loss": 1.0342, + "step": 8260 + }, + { + "epoch": 0.7450060873878342, + "grad_norm": 1.575120237090332, + "learning_rate": 6.440349092011628e-07, + "loss": 1.0204, + "step": 8261 + }, + { + "epoch": 0.7450962709113045, + "grad_norm": 1.4910973366345759, + "learning_rate": 6.436055266754475e-07, + "loss": 0.9445, + "step": 8262 + }, + { + "epoch": 0.7451864544347747, + "grad_norm": 1.6801372546939717, + "learning_rate": 6.431762598847879e-07, + "loss": 0.9289, + "step": 8263 + }, + { + "epoch": 0.7452766379582451, + "grad_norm": 1.476766259399199, + "learning_rate": 6.427471088658111e-07, + "loss": 0.9242, + "step": 8264 + }, + { + "epoch": 0.7453668214817153, + "grad_norm": 1.9822018355600377, + "learning_rate": 6.42318073655135e-07, + "loss": 0.953, + "step": 8265 + }, + { + "epoch": 0.7454570050051855, + "grad_norm": 1.4834645741870758, + "learning_rate": 6.41889154289367e-07, + "loss": 0.9149, + "step": 8266 + }, + { + "epoch": 0.7455471885286559, + "grad_norm": 1.5725293050925175, + "learning_rate": 6.414603508051055e-07, + "loss": 1.0112, + "step": 8267 + }, + { + "epoch": 0.7456373720521261, + "grad_norm": 0.6966064500185366, + "learning_rate": 6.410316632389365e-07, + "loss": 0.8066, + "step": 8268 + }, + { + "epoch": 0.7457275555755963, + "grad_norm": 1.6146600110294587, + "learning_rate": 6.406030916274406e-07, + "loss": 0.9609, + "step": 8269 + }, + { + "epoch": 0.7458177390990666, + "grad_norm": 2.8916413041320834, + "learning_rate": 6.401746360071831e-07, + "loss": 0.9094, + "step": 8270 + }, + { + "epoch": 0.7459079226225369, + "grad_norm": 2.8660986389308123, + "learning_rate": 6.397462964147251e-07, + "loss": 0.9105, + "step": 8271 + }, + { + "epoch": 0.7459981061460071, + "grad_norm": 1.5320188188218724, + "learning_rate": 6.393180728866128e-07, + "loss": 0.9418, + "step": 8272 + }, + { + "epoch": 0.7460882896694774, + "grad_norm": 1.444955730222084, + "learning_rate": 6.388899654593853e-07, + "loss": 0.9799, + "step": 8273 + }, + { + "epoch": 0.7461784731929476, + "grad_norm": 1.8959150471386625, + "learning_rate": 6.384619741695709e-07, + "loss": 0.9779, + "step": 8274 + }, + { + "epoch": 0.746268656716418, + "grad_norm": 1.2575986218315076, + "learning_rate": 6.380340990536883e-07, + "loss": 0.9501, + "step": 8275 + }, + { + "epoch": 0.7463588402398882, + "grad_norm": 2.2306015055309745, + "learning_rate": 6.37606340148247e-07, + "loss": 0.9544, + "step": 8276 + }, + { + "epoch": 0.7464490237633584, + "grad_norm": 2.198534247486681, + "learning_rate": 6.371786974897433e-07, + "loss": 0.8574, + "step": 8277 + }, + { + "epoch": 0.7465392072868287, + "grad_norm": 1.469023983338273, + "learning_rate": 6.367511711146691e-07, + "loss": 0.9822, + "step": 8278 + }, + { + "epoch": 0.746629390810299, + "grad_norm": 1.5329189158605714, + "learning_rate": 6.363237610595014e-07, + "loss": 0.895, + "step": 8279 + }, + { + "epoch": 0.7467195743337692, + "grad_norm": 1.7349734261975935, + "learning_rate": 6.358964673607094e-07, + "loss": 0.9299, + "step": 8280 + }, + { + "epoch": 0.7468097578572395, + "grad_norm": 1.3494101790011874, + "learning_rate": 6.354692900547525e-07, + "loss": 0.9846, + "step": 8281 + }, + { + "epoch": 0.7468999413807097, + "grad_norm": 1.6444373024679435, + "learning_rate": 6.350422291780797e-07, + "loss": 0.92, + "step": 8282 + }, + { + "epoch": 0.74699012490418, + "grad_norm": 1.9857925398149456, + "learning_rate": 6.346152847671302e-07, + "loss": 1.0545, + "step": 8283 + }, + { + "epoch": 0.7470803084276503, + "grad_norm": 1.6636807715775275, + "learning_rate": 6.34188456858334e-07, + "loss": 0.9442, + "step": 8284 + }, + { + "epoch": 0.7471704919511205, + "grad_norm": 1.7578896060562759, + "learning_rate": 6.337617454881081e-07, + "loss": 0.9031, + "step": 8285 + }, + { + "epoch": 0.7472606754745907, + "grad_norm": 1.4570393498172565, + "learning_rate": 6.333351506928651e-07, + "loss": 0.8597, + "step": 8286 + }, + { + "epoch": 0.7473508589980611, + "grad_norm": 1.7777586065026896, + "learning_rate": 6.329086725090018e-07, + "loss": 0.9251, + "step": 8287 + }, + { + "epoch": 0.7474410425215313, + "grad_norm": 1.527620880663279, + "learning_rate": 6.324823109729087e-07, + "loss": 0.9136, + "step": 8288 + }, + { + "epoch": 0.7475312260450016, + "grad_norm": 1.6325807543843525, + "learning_rate": 6.320560661209653e-07, + "loss": 0.8885, + "step": 8289 + }, + { + "epoch": 0.7476214095684719, + "grad_norm": 1.47328969754176, + "learning_rate": 6.316299379895411e-07, + "loss": 0.8329, + "step": 8290 + }, + { + "epoch": 0.7477115930919421, + "grad_norm": 1.6045582363583306, + "learning_rate": 6.312039266149965e-07, + "loss": 0.987, + "step": 8291 + }, + { + "epoch": 0.7478017766154124, + "grad_norm": 1.3292764496782412, + "learning_rate": 6.307780320336789e-07, + "loss": 0.9816, + "step": 8292 + }, + { + "epoch": 0.7478919601388826, + "grad_norm": 1.8121501321849023, + "learning_rate": 6.303522542819306e-07, + "loss": 1.0125, + "step": 8293 + }, + { + "epoch": 0.7479821436623529, + "grad_norm": 1.894982667607178, + "learning_rate": 6.299265933960796e-07, + "loss": 0.9245, + "step": 8294 + }, + { + "epoch": 0.7480723271858232, + "grad_norm": 1.842075492682859, + "learning_rate": 6.295010494124462e-07, + "loss": 0.9049, + "step": 8295 + }, + { + "epoch": 0.7481625107092934, + "grad_norm": 1.5313099440694258, + "learning_rate": 6.290756223673399e-07, + "loss": 0.9284, + "step": 8296 + }, + { + "epoch": 0.7482526942327636, + "grad_norm": 1.29041944811214, + "learning_rate": 6.28650312297061e-07, + "loss": 0.9822, + "step": 8297 + }, + { + "epoch": 0.748342877756234, + "grad_norm": 1.959734665563586, + "learning_rate": 6.282251192378987e-07, + "loss": 0.866, + "step": 8298 + }, + { + "epoch": 0.7484330612797042, + "grad_norm": 1.3490251306125813, + "learning_rate": 6.278000432261334e-07, + "loss": 0.8801, + "step": 8299 + }, + { + "epoch": 0.7485232448031744, + "grad_norm": 1.7544912199603533, + "learning_rate": 6.273750842980345e-07, + "loss": 0.9731, + "step": 8300 + }, + { + "epoch": 0.7486134283266447, + "grad_norm": 1.9456352977192197, + "learning_rate": 6.269502424898625e-07, + "loss": 0.9775, + "step": 8301 + }, + { + "epoch": 0.748703611850115, + "grad_norm": 1.873501687624568, + "learning_rate": 6.265255178378663e-07, + "loss": 0.9928, + "step": 8302 + }, + { + "epoch": 0.7487937953735853, + "grad_norm": 2.020465332292683, + "learning_rate": 6.261009103782861e-07, + "loss": 0.8476, + "step": 8303 + }, + { + "epoch": 0.7488839788970555, + "grad_norm": 1.6712009818193356, + "learning_rate": 6.256764201473519e-07, + "loss": 0.9962, + "step": 8304 + }, + { + "epoch": 0.7489741624205257, + "grad_norm": 0.6530961363405803, + "learning_rate": 6.252520471812835e-07, + "loss": 0.7747, + "step": 8305 + }, + { + "epoch": 0.7490643459439961, + "grad_norm": 1.8331060124118514, + "learning_rate": 6.248277915162912e-07, + "loss": 0.9673, + "step": 8306 + }, + { + "epoch": 0.7491545294674663, + "grad_norm": 1.448034085056984, + "learning_rate": 6.244036531885731e-07, + "loss": 0.8052, + "step": 8307 + }, + { + "epoch": 0.7492447129909365, + "grad_norm": 1.530064493702444, + "learning_rate": 6.239796322343216e-07, + "loss": 0.8834, + "step": 8308 + }, + { + "epoch": 0.7493348965144068, + "grad_norm": 1.5088590636572765, + "learning_rate": 6.235557286897137e-07, + "loss": 0.8863, + "step": 8309 + }, + { + "epoch": 0.7494250800378771, + "grad_norm": 1.584499816191528, + "learning_rate": 6.231319425909223e-07, + "loss": 0.9444, + "step": 8310 + }, + { + "epoch": 0.7495152635613473, + "grad_norm": 1.6523442550189462, + "learning_rate": 6.227082739741045e-07, + "loss": 0.9805, + "step": 8311 + }, + { + "epoch": 0.7496054470848176, + "grad_norm": 1.4088949125911918, + "learning_rate": 6.222847228754113e-07, + "loss": 0.9791, + "step": 8312 + }, + { + "epoch": 0.7496956306082878, + "grad_norm": 1.8517364549003679, + "learning_rate": 6.218612893309823e-07, + "loss": 1.0439, + "step": 8313 + }, + { + "epoch": 0.7497858141317582, + "grad_norm": 1.472525350485232, + "learning_rate": 6.214379733769468e-07, + "loss": 0.8875, + "step": 8314 + }, + { + "epoch": 0.7498759976552284, + "grad_norm": 1.574910430783362, + "learning_rate": 6.21014775049425e-07, + "loss": 0.9812, + "step": 8315 + }, + { + "epoch": 0.7499661811786986, + "grad_norm": 0.5996340022792044, + "learning_rate": 6.205916943845267e-07, + "loss": 0.766, + "step": 8316 + }, + { + "epoch": 0.750056364702169, + "grad_norm": 1.8228520045846526, + "learning_rate": 6.201687314183504e-07, + "loss": 0.9943, + "step": 8317 + }, + { + "epoch": 0.7501465482256392, + "grad_norm": 0.6758388208110842, + "learning_rate": 6.197458861869862e-07, + "loss": 0.8128, + "step": 8318 + }, + { + "epoch": 0.7502367317491094, + "grad_norm": 1.4920961042445482, + "learning_rate": 6.193231587265138e-07, + "loss": 0.9924, + "step": 8319 + }, + { + "epoch": 0.7503269152725797, + "grad_norm": 1.721965902320979, + "learning_rate": 6.189005490730024e-07, + "loss": 0.9351, + "step": 8320 + }, + { + "epoch": 0.75041709879605, + "grad_norm": 1.5678488160448203, + "learning_rate": 6.184780572625115e-07, + "loss": 0.9721, + "step": 8321 + }, + { + "epoch": 0.7505072823195202, + "grad_norm": 1.8295724469374066, + "learning_rate": 6.180556833310902e-07, + "loss": 0.8512, + "step": 8322 + }, + { + "epoch": 0.7505974658429905, + "grad_norm": 1.6749866086509935, + "learning_rate": 6.176334273147788e-07, + "loss": 0.9902, + "step": 8323 + }, + { + "epoch": 0.7506876493664607, + "grad_norm": 3.510090859044124, + "learning_rate": 6.172112892496042e-07, + "loss": 0.8708, + "step": 8324 + }, + { + "epoch": 0.750777832889931, + "grad_norm": 2.768939515940757, + "learning_rate": 6.167892691715883e-07, + "loss": 0.9936, + "step": 8325 + }, + { + "epoch": 0.7508680164134013, + "grad_norm": 1.3574505919568056, + "learning_rate": 6.163673671167378e-07, + "loss": 0.9327, + "step": 8326 + }, + { + "epoch": 0.7509581999368715, + "grad_norm": 1.6245128928553767, + "learning_rate": 6.15945583121054e-07, + "loss": 0.9893, + "step": 8327 + }, + { + "epoch": 0.7510483834603418, + "grad_norm": 1.7230220760895005, + "learning_rate": 6.15523917220524e-07, + "loss": 0.8995, + "step": 8328 + }, + { + "epoch": 0.7511385669838121, + "grad_norm": 1.5047064695816357, + "learning_rate": 6.151023694511273e-07, + "loss": 1.0176, + "step": 8329 + }, + { + "epoch": 0.7512287505072823, + "grad_norm": 1.5982870707945258, + "learning_rate": 6.146809398488328e-07, + "loss": 0.8933, + "step": 8330 + }, + { + "epoch": 0.7513189340307526, + "grad_norm": 1.5532571932068757, + "learning_rate": 6.142596284495989e-07, + "loss": 0.9833, + "step": 8331 + }, + { + "epoch": 0.7514091175542228, + "grad_norm": 1.572275405758592, + "learning_rate": 6.138384352893751e-07, + "loss": 0.9035, + "step": 8332 + }, + { + "epoch": 0.7514993010776931, + "grad_norm": 1.8210310929695543, + "learning_rate": 6.134173604040987e-07, + "loss": 0.9369, + "step": 8333 + }, + { + "epoch": 0.7515894846011634, + "grad_norm": 1.8705112769126775, + "learning_rate": 6.129964038296984e-07, + "loss": 0.9543, + "step": 8334 + }, + { + "epoch": 0.7516796681246336, + "grad_norm": 1.5935973299990351, + "learning_rate": 6.12575565602093e-07, + "loss": 1.0194, + "step": 8335 + }, + { + "epoch": 0.7517698516481038, + "grad_norm": 1.2250182520226776, + "learning_rate": 6.121548457571905e-07, + "loss": 1.0206, + "step": 8336 + }, + { + "epoch": 0.7518600351715742, + "grad_norm": 1.722379770387632, + "learning_rate": 6.11734244330889e-07, + "loss": 0.8991, + "step": 8337 + }, + { + "epoch": 0.7519502186950444, + "grad_norm": 1.5967288135122215, + "learning_rate": 6.113137613590773e-07, + "loss": 0.9573, + "step": 8338 + }, + { + "epoch": 0.7520404022185146, + "grad_norm": 1.5952523574177992, + "learning_rate": 6.108933968776313e-07, + "loss": 1.0089, + "step": 8339 + }, + { + "epoch": 0.752130585741985, + "grad_norm": 1.451935804508525, + "learning_rate": 6.104731509224212e-07, + "loss": 0.9486, + "step": 8340 + }, + { + "epoch": 0.7522207692654552, + "grad_norm": 1.3977967397959912, + "learning_rate": 6.100530235293027e-07, + "loss": 0.8821, + "step": 8341 + }, + { + "epoch": 0.7523109527889255, + "grad_norm": 1.6668574541882721, + "learning_rate": 6.096330147341253e-07, + "loss": 0.9691, + "step": 8342 + }, + { + "epoch": 0.7524011363123957, + "grad_norm": 1.8922135796234143, + "learning_rate": 6.09213124572725e-07, + "loss": 1.011, + "step": 8343 + }, + { + "epoch": 0.752491319835866, + "grad_norm": 1.3962364136693113, + "learning_rate": 6.087933530809297e-07, + "loss": 1.0256, + "step": 8344 + }, + { + "epoch": 0.7525815033593363, + "grad_norm": 1.4915168413470397, + "learning_rate": 6.083737002945566e-07, + "loss": 0.9389, + "step": 8345 + }, + { + "epoch": 0.7526716868828065, + "grad_norm": 1.5416298767373982, + "learning_rate": 6.079541662494126e-07, + "loss": 0.9024, + "step": 8346 + }, + { + "epoch": 0.7527618704062767, + "grad_norm": 1.7819512977709473, + "learning_rate": 6.075347509812954e-07, + "loss": 0.9612, + "step": 8347 + }, + { + "epoch": 0.7528520539297471, + "grad_norm": 1.6616952741118745, + "learning_rate": 6.0711545452599e-07, + "loss": 0.9061, + "step": 8348 + }, + { + "epoch": 0.7529422374532173, + "grad_norm": 1.4744780020898416, + "learning_rate": 6.066962769192756e-07, + "loss": 0.8715, + "step": 8349 + }, + { + "epoch": 0.7530324209766875, + "grad_norm": 1.8955642243706725, + "learning_rate": 6.062772181969167e-07, + "loss": 0.884, + "step": 8350 + }, + { + "epoch": 0.7531226045001578, + "grad_norm": 1.4662628217495894, + "learning_rate": 6.058582783946706e-07, + "loss": 1.0161, + "step": 8351 + }, + { + "epoch": 0.7532127880236281, + "grad_norm": 3.4850554892283263, + "learning_rate": 6.054394575482833e-07, + "loss": 0.9399, + "step": 8352 + }, + { + "epoch": 0.7533029715470984, + "grad_norm": 1.5158149364264526, + "learning_rate": 6.05020755693491e-07, + "loss": 0.865, + "step": 8353 + }, + { + "epoch": 0.7533931550705686, + "grad_norm": 1.5950554381873732, + "learning_rate": 6.046021728660198e-07, + "loss": 0.8622, + "step": 8354 + }, + { + "epoch": 0.7534833385940388, + "grad_norm": 1.6082543480540221, + "learning_rate": 6.041837091015858e-07, + "loss": 0.9896, + "step": 8355 + }, + { + "epoch": 0.7535735221175092, + "grad_norm": 1.4180407124903973, + "learning_rate": 6.037653644358931e-07, + "loss": 1.002, + "step": 8356 + }, + { + "epoch": 0.7536637056409794, + "grad_norm": 1.7605322925885427, + "learning_rate": 6.033471389046393e-07, + "loss": 0.9629, + "step": 8357 + }, + { + "epoch": 0.7537538891644496, + "grad_norm": 1.4895922876191159, + "learning_rate": 6.029290325435084e-07, + "loss": 0.9056, + "step": 8358 + }, + { + "epoch": 0.7538440726879199, + "grad_norm": 1.93447184035254, + "learning_rate": 6.025110453881756e-07, + "loss": 1.018, + "step": 8359 + }, + { + "epoch": 0.7539342562113902, + "grad_norm": 1.6887850636781727, + "learning_rate": 6.020931774743061e-07, + "loss": 0.8853, + "step": 8360 + }, + { + "epoch": 0.7540244397348604, + "grad_norm": 1.5691907681427204, + "learning_rate": 6.016754288375546e-07, + "loss": 0.8753, + "step": 8361 + }, + { + "epoch": 0.7541146232583307, + "grad_norm": 1.5099255081143312, + "learning_rate": 6.012577995135665e-07, + "loss": 0.9453, + "step": 8362 + }, + { + "epoch": 0.754204806781801, + "grad_norm": 1.7988257392594114, + "learning_rate": 6.008402895379743e-07, + "loss": 0.8944, + "step": 8363 + }, + { + "epoch": 0.7542949903052713, + "grad_norm": 1.8543449773309153, + "learning_rate": 6.004228989464047e-07, + "loss": 1.0331, + "step": 8364 + }, + { + "epoch": 0.7543851738287415, + "grad_norm": 1.8272875310331804, + "learning_rate": 6.000056277744692e-07, + "loss": 0.9665, + "step": 8365 + }, + { + "epoch": 0.7544753573522117, + "grad_norm": 1.461119043830704, + "learning_rate": 5.995884760577745e-07, + "loss": 0.9233, + "step": 8366 + }, + { + "epoch": 0.7545655408756821, + "grad_norm": 0.6693714204794011, + "learning_rate": 5.99171443831912e-07, + "loss": 0.7655, + "step": 8367 + }, + { + "epoch": 0.7546557243991523, + "grad_norm": 1.3834738577663785, + "learning_rate": 5.98754531132466e-07, + "loss": 0.9485, + "step": 8368 + }, + { + "epoch": 0.7547459079226225, + "grad_norm": 2.5975365632671594, + "learning_rate": 5.983377379950099e-07, + "loss": 0.9513, + "step": 8369 + }, + { + "epoch": 0.7548360914460928, + "grad_norm": 1.4699920110420552, + "learning_rate": 5.979210644551067e-07, + "loss": 0.8835, + "step": 8370 + }, + { + "epoch": 0.7549262749695631, + "grad_norm": 1.9435406930960346, + "learning_rate": 5.975045105483091e-07, + "loss": 0.9415, + "step": 8371 + }, + { + "epoch": 0.7550164584930333, + "grad_norm": 0.6543219718105429, + "learning_rate": 5.970880763101607e-07, + "loss": 0.7888, + "step": 8372 + }, + { + "epoch": 0.7551066420165036, + "grad_norm": 1.4588667938365083, + "learning_rate": 5.966717617761925e-07, + "loss": 0.9695, + "step": 8373 + }, + { + "epoch": 0.7551968255399738, + "grad_norm": 1.4338935725622397, + "learning_rate": 5.962555669819276e-07, + "loss": 0.9302, + "step": 8374 + }, + { + "epoch": 0.7552870090634441, + "grad_norm": 1.4132666693041265, + "learning_rate": 5.958394919628777e-07, + "loss": 0.9663, + "step": 8375 + }, + { + "epoch": 0.7553771925869144, + "grad_norm": 1.4651313194292883, + "learning_rate": 5.954235367545451e-07, + "loss": 0.8866, + "step": 8376 + }, + { + "epoch": 0.7554673761103846, + "grad_norm": 1.3365821844421417, + "learning_rate": 5.950077013924213e-07, + "loss": 0.944, + "step": 8377 + }, + { + "epoch": 0.7555575596338548, + "grad_norm": 1.7630046879429624, + "learning_rate": 5.945919859119865e-07, + "loss": 1.0333, + "step": 8378 + }, + { + "epoch": 0.7556477431573252, + "grad_norm": 1.7436597274329104, + "learning_rate": 5.94176390348714e-07, + "loss": 0.8398, + "step": 8379 + }, + { + "epoch": 0.7557379266807954, + "grad_norm": 1.2226182243113395, + "learning_rate": 5.937609147380622e-07, + "loss": 1.0571, + "step": 8380 + }, + { + "epoch": 0.7558281102042657, + "grad_norm": 1.5096955079529775, + "learning_rate": 5.933455591154844e-07, + "loss": 0.8927, + "step": 8381 + }, + { + "epoch": 0.7559182937277359, + "grad_norm": 1.8046365045385688, + "learning_rate": 5.929303235164191e-07, + "loss": 0.9225, + "step": 8382 + }, + { + "epoch": 0.7560084772512062, + "grad_norm": 1.6605043293185198, + "learning_rate": 5.92515207976297e-07, + "loss": 0.9233, + "step": 8383 + }, + { + "epoch": 0.7560986607746765, + "grad_norm": 1.5783688196357466, + "learning_rate": 5.921002125305383e-07, + "loss": 0.9406, + "step": 8384 + }, + { + "epoch": 0.7561888442981467, + "grad_norm": 1.6335079774454515, + "learning_rate": 5.916853372145525e-07, + "loss": 0.8708, + "step": 8385 + }, + { + "epoch": 0.756279027821617, + "grad_norm": 1.7227884185674534, + "learning_rate": 5.912705820637389e-07, + "loss": 0.8862, + "step": 8386 + }, + { + "epoch": 0.7563692113450873, + "grad_norm": 1.3867030183141547, + "learning_rate": 5.908559471134871e-07, + "loss": 1.016, + "step": 8387 + }, + { + "epoch": 0.7564593948685575, + "grad_norm": 1.7713087976962867, + "learning_rate": 5.904414323991764e-07, + "loss": 0.8883, + "step": 8388 + }, + { + "epoch": 0.7565495783920277, + "grad_norm": 1.626229994443366, + "learning_rate": 5.900270379561743e-07, + "loss": 0.9581, + "step": 8389 + }, + { + "epoch": 0.7566397619154981, + "grad_norm": 1.5776918780012181, + "learning_rate": 5.896127638198399e-07, + "loss": 0.9603, + "step": 8390 + }, + { + "epoch": 0.7567299454389683, + "grad_norm": 1.7129753900865698, + "learning_rate": 5.89198610025521e-07, + "loss": 0.9547, + "step": 8391 + }, + { + "epoch": 0.7568201289624386, + "grad_norm": 2.307166780029741, + "learning_rate": 5.887845766085559e-07, + "loss": 1.0182, + "step": 8392 + }, + { + "epoch": 0.7569103124859088, + "grad_norm": 1.8463728401161634, + "learning_rate": 5.883706636042722e-07, + "loss": 0.91, + "step": 8393 + }, + { + "epoch": 0.7570004960093791, + "grad_norm": 2.2434766300039275, + "learning_rate": 5.879568710479879e-07, + "loss": 0.8918, + "step": 8394 + }, + { + "epoch": 0.7570906795328494, + "grad_norm": 1.71209837666068, + "learning_rate": 5.875431989750078e-07, + "loss": 1.043, + "step": 8395 + }, + { + "epoch": 0.7571808630563196, + "grad_norm": 1.4034584736621196, + "learning_rate": 5.871296474206313e-07, + "loss": 0.8811, + "step": 8396 + }, + { + "epoch": 0.7572710465797898, + "grad_norm": 1.4574853429826164, + "learning_rate": 5.867162164201427e-07, + "loss": 0.9556, + "step": 8397 + }, + { + "epoch": 0.7573612301032602, + "grad_norm": 1.5953892524032998, + "learning_rate": 5.863029060088205e-07, + "loss": 0.9811, + "step": 8398 + }, + { + "epoch": 0.7574514136267304, + "grad_norm": 1.8811770732343158, + "learning_rate": 5.858897162219289e-07, + "loss": 0.8463, + "step": 8399 + }, + { + "epoch": 0.7575415971502006, + "grad_norm": 1.5574824833994736, + "learning_rate": 5.854766470947238e-07, + "loss": 0.9992, + "step": 8400 + }, + { + "epoch": 0.7576317806736709, + "grad_norm": 1.4939079552297492, + "learning_rate": 5.850636986624511e-07, + "loss": 0.9367, + "step": 8401 + }, + { + "epoch": 0.7577219641971412, + "grad_norm": 1.6638740110430938, + "learning_rate": 5.846508709603453e-07, + "loss": 0.9567, + "step": 8402 + }, + { + "epoch": 0.7578121477206115, + "grad_norm": 1.4035203508424385, + "learning_rate": 5.842381640236318e-07, + "loss": 0.9271, + "step": 8403 + }, + { + "epoch": 0.7579023312440817, + "grad_norm": 1.5005056005956203, + "learning_rate": 5.838255778875242e-07, + "loss": 0.9479, + "step": 8404 + }, + { + "epoch": 0.7579925147675519, + "grad_norm": 1.6097810103475674, + "learning_rate": 5.83413112587227e-07, + "loss": 0.9723, + "step": 8405 + }, + { + "epoch": 0.7580826982910223, + "grad_norm": 1.8213905122130554, + "learning_rate": 5.830007681579338e-07, + "loss": 0.9364, + "step": 8406 + }, + { + "epoch": 0.7581728818144925, + "grad_norm": 1.6284421996182605, + "learning_rate": 5.825885446348284e-07, + "loss": 0.9638, + "step": 8407 + }, + { + "epoch": 0.7582630653379627, + "grad_norm": 1.4146280580322241, + "learning_rate": 5.821764420530842e-07, + "loss": 0.887, + "step": 8408 + }, + { + "epoch": 0.7583532488614331, + "grad_norm": 1.3909698658770488, + "learning_rate": 5.817644604478633e-07, + "loss": 0.9676, + "step": 8409 + }, + { + "epoch": 0.7584434323849033, + "grad_norm": 1.5161601429452582, + "learning_rate": 5.81352599854319e-07, + "loss": 0.8947, + "step": 8410 + }, + { + "epoch": 0.7585336159083735, + "grad_norm": 1.4697141526417523, + "learning_rate": 5.809408603075938e-07, + "loss": 0.9862, + "step": 8411 + }, + { + "epoch": 0.7586237994318438, + "grad_norm": 1.509787891218812, + "learning_rate": 5.805292418428176e-07, + "loss": 0.9247, + "step": 8412 + }, + { + "epoch": 0.7587139829553141, + "grad_norm": 2.2841386664965033, + "learning_rate": 5.801177444951148e-07, + "loss": 0.9374, + "step": 8413 + }, + { + "epoch": 0.7588041664787843, + "grad_norm": 1.20772076020234, + "learning_rate": 5.797063682995944e-07, + "loss": 0.9761, + "step": 8414 + }, + { + "epoch": 0.7588943500022546, + "grad_norm": 4.080535527032932, + "learning_rate": 5.792951132913584e-07, + "loss": 1.0157, + "step": 8415 + }, + { + "epoch": 0.7589845335257248, + "grad_norm": 1.7753230083082727, + "learning_rate": 5.788839795054968e-07, + "loss": 0.9992, + "step": 8416 + }, + { + "epoch": 0.7590747170491952, + "grad_norm": 1.3438344086151446, + "learning_rate": 5.784729669770898e-07, + "loss": 0.8653, + "step": 8417 + }, + { + "epoch": 0.7591649005726654, + "grad_norm": 1.4144705505056232, + "learning_rate": 5.780620757412084e-07, + "loss": 0.9462, + "step": 8418 + }, + { + "epoch": 0.7592550840961356, + "grad_norm": 2.4501180503586677, + "learning_rate": 5.776513058329098e-07, + "loss": 0.9717, + "step": 8419 + }, + { + "epoch": 0.7593452676196059, + "grad_norm": 1.5538580133729294, + "learning_rate": 5.772406572872459e-07, + "loss": 0.9728, + "step": 8420 + }, + { + "epoch": 0.7594354511430762, + "grad_norm": 1.6437158443252855, + "learning_rate": 5.768301301392535e-07, + "loss": 0.9786, + "step": 8421 + }, + { + "epoch": 0.7595256346665464, + "grad_norm": 1.4722498325341975, + "learning_rate": 5.764197244239615e-07, + "loss": 0.9613, + "step": 8422 + }, + { + "epoch": 0.7596158181900167, + "grad_norm": 1.5507349732842628, + "learning_rate": 5.760094401763884e-07, + "loss": 0.9867, + "step": 8423 + }, + { + "epoch": 0.7597060017134869, + "grad_norm": 1.3829653380678046, + "learning_rate": 5.755992774315414e-07, + "loss": 0.9372, + "step": 8424 + }, + { + "epoch": 0.7597961852369572, + "grad_norm": 2.0305307141808244, + "learning_rate": 5.751892362244183e-07, + "loss": 0.9386, + "step": 8425 + }, + { + "epoch": 0.7598863687604275, + "grad_norm": 1.7949233361239292, + "learning_rate": 5.747793165900065e-07, + "loss": 0.9995, + "step": 8426 + }, + { + "epoch": 0.7599765522838977, + "grad_norm": 1.4908545967037072, + "learning_rate": 5.743695185632806e-07, + "loss": 1.0057, + "step": 8427 + }, + { + "epoch": 0.7600667358073679, + "grad_norm": 1.6889495981261182, + "learning_rate": 5.739598421792098e-07, + "loss": 0.9926, + "step": 8428 + }, + { + "epoch": 0.7601569193308383, + "grad_norm": 1.502135663789792, + "learning_rate": 5.735502874727474e-07, + "loss": 0.9279, + "step": 8429 + }, + { + "epoch": 0.7602471028543085, + "grad_norm": 1.490829968705422, + "learning_rate": 5.731408544788398e-07, + "loss": 1.082, + "step": 8430 + }, + { + "epoch": 0.7603372863777788, + "grad_norm": 1.4790633080418014, + "learning_rate": 5.727315432324225e-07, + "loss": 0.8649, + "step": 8431 + }, + { + "epoch": 0.760427469901249, + "grad_norm": 1.528085201813919, + "learning_rate": 5.723223537684196e-07, + "loss": 0.9419, + "step": 8432 + }, + { + "epoch": 0.7605176534247193, + "grad_norm": 1.5221082316751018, + "learning_rate": 5.719132861217462e-07, + "loss": 1.0289, + "step": 8433 + }, + { + "epoch": 0.7606078369481896, + "grad_norm": 1.3814549652198151, + "learning_rate": 5.715043403273044e-07, + "loss": 0.9181, + "step": 8434 + }, + { + "epoch": 0.7606980204716598, + "grad_norm": 2.1916435235642457, + "learning_rate": 5.710955164199902e-07, + "loss": 0.909, + "step": 8435 + }, + { + "epoch": 0.7607882039951301, + "grad_norm": 1.571090126128264, + "learning_rate": 5.706868144346841e-07, + "loss": 0.9695, + "step": 8436 + }, + { + "epoch": 0.7608783875186004, + "grad_norm": 1.7253659620329522, + "learning_rate": 5.702782344062613e-07, + "loss": 0.9436, + "step": 8437 + }, + { + "epoch": 0.7609685710420706, + "grad_norm": 1.4971052689524118, + "learning_rate": 5.698697763695826e-07, + "loss": 0.9251, + "step": 8438 + }, + { + "epoch": 0.7610587545655408, + "grad_norm": 1.5637266303373711, + "learning_rate": 5.694614403595002e-07, + "loss": 0.9733, + "step": 8439 + }, + { + "epoch": 0.7611489380890112, + "grad_norm": 1.527627825843714, + "learning_rate": 5.690532264108554e-07, + "loss": 0.8559, + "step": 8440 + }, + { + "epoch": 0.7612391216124814, + "grad_norm": 1.542280057933581, + "learning_rate": 5.686451345584795e-07, + "loss": 0.987, + "step": 8441 + }, + { + "epoch": 0.7613293051359517, + "grad_norm": 1.6467430481718361, + "learning_rate": 5.682371648371933e-07, + "loss": 1.0472, + "step": 8442 + }, + { + "epoch": 0.7614194886594219, + "grad_norm": 1.7139720047853366, + "learning_rate": 5.678293172818074e-07, + "loss": 0.9129, + "step": 8443 + }, + { + "epoch": 0.7615096721828922, + "grad_norm": 1.6116959215656372, + "learning_rate": 5.674215919271204e-07, + "loss": 0.9626, + "step": 8444 + }, + { + "epoch": 0.7615998557063625, + "grad_norm": 1.6255361699492363, + "learning_rate": 5.670139888079224e-07, + "loss": 0.8771, + "step": 8445 + }, + { + "epoch": 0.7616900392298327, + "grad_norm": 1.5016637158691226, + "learning_rate": 5.666065079589924e-07, + "loss": 1.0091, + "step": 8446 + }, + { + "epoch": 0.7617802227533029, + "grad_norm": 1.632970117873726, + "learning_rate": 5.661991494150986e-07, + "loss": 0.8967, + "step": 8447 + }, + { + "epoch": 0.7618704062767733, + "grad_norm": 1.3968065863802508, + "learning_rate": 5.657919132109999e-07, + "loss": 0.9952, + "step": 8448 + }, + { + "epoch": 0.7619605898002435, + "grad_norm": 1.476162164902984, + "learning_rate": 5.653847993814421e-07, + "loss": 0.9771, + "step": 8449 + }, + { + "epoch": 0.7620507733237137, + "grad_norm": 2.0209609971244356, + "learning_rate": 5.649778079611647e-07, + "loss": 0.962, + "step": 8450 + }, + { + "epoch": 0.762140956847184, + "grad_norm": 1.710114581636043, + "learning_rate": 5.645709389848923e-07, + "loss": 0.9304, + "step": 8451 + }, + { + "epoch": 0.7622311403706543, + "grad_norm": 1.6675851278008342, + "learning_rate": 5.641641924873435e-07, + "loss": 0.9114, + "step": 8452 + }, + { + "epoch": 0.7623213238941245, + "grad_norm": 1.5853470327931314, + "learning_rate": 5.637575685032217e-07, + "loss": 0.9479, + "step": 8453 + }, + { + "epoch": 0.7624115074175948, + "grad_norm": 1.5425034992543103, + "learning_rate": 5.633510670672246e-07, + "loss": 1.0293, + "step": 8454 + }, + { + "epoch": 0.762501690941065, + "grad_norm": 0.68476671130654, + "learning_rate": 5.629446882140354e-07, + "loss": 0.8065, + "step": 8455 + }, + { + "epoch": 0.7625918744645354, + "grad_norm": 1.8354031048842712, + "learning_rate": 5.625384319783295e-07, + "loss": 0.9594, + "step": 8456 + }, + { + "epoch": 0.7626820579880056, + "grad_norm": 1.9816448624451108, + "learning_rate": 5.621322983947705e-07, + "loss": 0.8911, + "step": 8457 + }, + { + "epoch": 0.7627722415114758, + "grad_norm": 1.6892618236390478, + "learning_rate": 5.617262874980122e-07, + "loss": 0.9871, + "step": 8458 + }, + { + "epoch": 0.7628624250349462, + "grad_norm": 1.3655047278613972, + "learning_rate": 5.613203993226981e-07, + "loss": 0.9968, + "step": 8459 + }, + { + "epoch": 0.7629526085584164, + "grad_norm": 1.8298189840789851, + "learning_rate": 5.609146339034599e-07, + "loss": 1.0039, + "step": 8460 + }, + { + "epoch": 0.7630427920818866, + "grad_norm": 3.416017998006966, + "learning_rate": 5.605089912749199e-07, + "loss": 0.8876, + "step": 8461 + }, + { + "epoch": 0.7631329756053569, + "grad_norm": 1.887749558176063, + "learning_rate": 5.601034714716901e-07, + "loss": 0.8924, + "step": 8462 + }, + { + "epoch": 0.7632231591288272, + "grad_norm": 1.4117476763820154, + "learning_rate": 5.59698074528372e-07, + "loss": 0.922, + "step": 8463 + }, + { + "epoch": 0.7633133426522974, + "grad_norm": 1.5724759358079436, + "learning_rate": 5.592928004795555e-07, + "loss": 0.9218, + "step": 8464 + }, + { + "epoch": 0.7634035261757677, + "grad_norm": 1.6387460030304026, + "learning_rate": 5.58887649359822e-07, + "loss": 0.9535, + "step": 8465 + }, + { + "epoch": 0.7634937096992379, + "grad_norm": 1.5584658615686147, + "learning_rate": 5.584826212037393e-07, + "loss": 0.8065, + "step": 8466 + }, + { + "epoch": 0.7635838932227083, + "grad_norm": 1.2864003290646877, + "learning_rate": 5.580777160458689e-07, + "loss": 0.8742, + "step": 8467 + }, + { + "epoch": 0.7636740767461785, + "grad_norm": 1.4960872958768667, + "learning_rate": 5.576729339207574e-07, + "loss": 0.9709, + "step": 8468 + }, + { + "epoch": 0.7637642602696487, + "grad_norm": 1.3215618204442963, + "learning_rate": 5.572682748629449e-07, + "loss": 1.0068, + "step": 8469 + }, + { + "epoch": 0.763854443793119, + "grad_norm": 1.3906781475755068, + "learning_rate": 5.568637389069582e-07, + "loss": 0.9325, + "step": 8470 + }, + { + "epoch": 0.7639446273165893, + "grad_norm": 1.6362551667646952, + "learning_rate": 5.564593260873145e-07, + "loss": 0.9631, + "step": 8471 + }, + { + "epoch": 0.7640348108400595, + "grad_norm": 1.4468511690686021, + "learning_rate": 5.560550364385206e-07, + "loss": 0.9853, + "step": 8472 + }, + { + "epoch": 0.7641249943635298, + "grad_norm": 1.5029940922888148, + "learning_rate": 5.556508699950728e-07, + "loss": 1.0175, + "step": 8473 + }, + { + "epoch": 0.764215177887, + "grad_norm": 1.6392699685811265, + "learning_rate": 5.552468267914577e-07, + "loss": 0.9017, + "step": 8474 + }, + { + "epoch": 0.7643053614104703, + "grad_norm": 1.9790041471980484, + "learning_rate": 5.548429068621481e-07, + "loss": 0.8661, + "step": 8475 + }, + { + "epoch": 0.7643955449339406, + "grad_norm": 0.6839310604037363, + "learning_rate": 5.544391102416115e-07, + "loss": 0.7758, + "step": 8476 + }, + { + "epoch": 0.7644857284574108, + "grad_norm": 1.8781693058214843, + "learning_rate": 5.540354369643003e-07, + "loss": 1.0054, + "step": 8477 + }, + { + "epoch": 0.764575911980881, + "grad_norm": 1.7644169143092598, + "learning_rate": 5.536318870646586e-07, + "loss": 0.8008, + "step": 8478 + }, + { + "epoch": 0.7646660955043514, + "grad_norm": 1.7856953347426678, + "learning_rate": 5.532284605771194e-07, + "loss": 0.8568, + "step": 8479 + }, + { + "epoch": 0.7647562790278216, + "grad_norm": 1.3698781594164198, + "learning_rate": 5.528251575361052e-07, + "loss": 0.9125, + "step": 8480 + }, + { + "epoch": 0.7648464625512919, + "grad_norm": 1.4943419875547375, + "learning_rate": 5.524219779760284e-07, + "loss": 0.9457, + "step": 8481 + }, + { + "epoch": 0.7649366460747622, + "grad_norm": 1.5324438462540795, + "learning_rate": 5.520189219312907e-07, + "loss": 0.9703, + "step": 8482 + }, + { + "epoch": 0.7650268295982324, + "grad_norm": 1.4901900578579963, + "learning_rate": 5.516159894362817e-07, + "loss": 0.8375, + "step": 8483 + }, + { + "epoch": 0.7651170131217027, + "grad_norm": 1.434397791596227, + "learning_rate": 5.512131805253839e-07, + "loss": 0.8858, + "step": 8484 + }, + { + "epoch": 0.7652071966451729, + "grad_norm": 1.4427655021642392, + "learning_rate": 5.508104952329653e-07, + "loss": 0.9615, + "step": 8485 + }, + { + "epoch": 0.7652973801686432, + "grad_norm": 1.98201834043299, + "learning_rate": 5.504079335933862e-07, + "loss": 0.9737, + "step": 8486 + }, + { + "epoch": 0.7653875636921135, + "grad_norm": 1.67549767573556, + "learning_rate": 5.500054956409952e-07, + "loss": 0.9761, + "step": 8487 + }, + { + "epoch": 0.7654777472155837, + "grad_norm": 1.7482672696258206, + "learning_rate": 5.496031814101303e-07, + "loss": 0.9099, + "step": 8488 + }, + { + "epoch": 0.7655679307390539, + "grad_norm": 1.5541359387889309, + "learning_rate": 5.492009909351203e-07, + "loss": 0.9559, + "step": 8489 + }, + { + "epoch": 0.7656581142625243, + "grad_norm": 3.8229586870065173, + "learning_rate": 5.4879892425028e-07, + "loss": 0.8739, + "step": 8490 + }, + { + "epoch": 0.7657482977859945, + "grad_norm": 1.545868902833984, + "learning_rate": 5.483969813899184e-07, + "loss": 1.0854, + "step": 8491 + }, + { + "epoch": 0.7658384813094647, + "grad_norm": 1.3999175864894342, + "learning_rate": 5.479951623883299e-07, + "loss": 1.0302, + "step": 8492 + }, + { + "epoch": 0.765928664832935, + "grad_norm": 1.6910639674237475, + "learning_rate": 5.475934672798004e-07, + "loss": 0.8775, + "step": 8493 + }, + { + "epoch": 0.7660188483564053, + "grad_norm": 1.6917121733885296, + "learning_rate": 5.471918960986047e-07, + "loss": 0.9133, + "step": 8494 + }, + { + "epoch": 0.7661090318798756, + "grad_norm": 1.594138229057981, + "learning_rate": 5.467904488790071e-07, + "loss": 1.0476, + "step": 8495 + }, + { + "epoch": 0.7661992154033458, + "grad_norm": 1.9038051498550217, + "learning_rate": 5.463891256552615e-07, + "loss": 0.956, + "step": 8496 + }, + { + "epoch": 0.766289398926816, + "grad_norm": 1.5868253430728712, + "learning_rate": 5.459879264616107e-07, + "loss": 0.8566, + "step": 8497 + }, + { + "epoch": 0.7663795824502864, + "grad_norm": 1.9099665822605831, + "learning_rate": 5.455868513322874e-07, + "loss": 0.8286, + "step": 8498 + }, + { + "epoch": 0.7664697659737566, + "grad_norm": 1.6740235400441865, + "learning_rate": 5.451859003015143e-07, + "loss": 0.8405, + "step": 8499 + }, + { + "epoch": 0.7665599494972268, + "grad_norm": 1.5203969494848, + "learning_rate": 5.447850734035009e-07, + "loss": 0.9838, + "step": 8500 + }, + { + "epoch": 0.7666501330206971, + "grad_norm": 1.6053197192890598, + "learning_rate": 5.443843706724494e-07, + "loss": 0.9254, + "step": 8501 + }, + { + "epoch": 0.7667403165441674, + "grad_norm": 1.8102188224662734, + "learning_rate": 5.439837921425494e-07, + "loss": 0.8829, + "step": 8502 + }, + { + "epoch": 0.7668305000676376, + "grad_norm": 1.283061491467157, + "learning_rate": 5.435833378479807e-07, + "loss": 0.9353, + "step": 8503 + }, + { + "epoch": 0.7669206835911079, + "grad_norm": 1.4212168752925118, + "learning_rate": 5.431830078229128e-07, + "loss": 0.8057, + "step": 8504 + }, + { + "epoch": 0.7670108671145782, + "grad_norm": 1.7880980044863513, + "learning_rate": 5.427828021015022e-07, + "loss": 0.9628, + "step": 8505 + }, + { + "epoch": 0.7671010506380485, + "grad_norm": 1.4901101396803849, + "learning_rate": 5.42382720717899e-07, + "loss": 0.9734, + "step": 8506 + }, + { + "epoch": 0.7671912341615187, + "grad_norm": 1.3455094196666035, + "learning_rate": 5.419827637062384e-07, + "loss": 0.9785, + "step": 8507 + }, + { + "epoch": 0.7672814176849889, + "grad_norm": 1.713629707772396, + "learning_rate": 5.415829311006487e-07, + "loss": 1.0944, + "step": 8508 + }, + { + "epoch": 0.7673716012084593, + "grad_norm": 1.9713791745278133, + "learning_rate": 5.411832229352447e-07, + "loss": 1.0253, + "step": 8509 + }, + { + "epoch": 0.7674617847319295, + "grad_norm": 1.5416034307484245, + "learning_rate": 5.407836392441319e-07, + "loss": 1.0403, + "step": 8510 + }, + { + "epoch": 0.7675519682553997, + "grad_norm": 1.5089851523566953, + "learning_rate": 5.403841800614049e-07, + "loss": 0.9657, + "step": 8511 + }, + { + "epoch": 0.76764215177887, + "grad_norm": 1.451641350812645, + "learning_rate": 5.39984845421148e-07, + "loss": 0.9047, + "step": 8512 + }, + { + "epoch": 0.7677323353023403, + "grad_norm": 1.531076693945447, + "learning_rate": 5.395856353574344e-07, + "loss": 0.9309, + "step": 8513 + }, + { + "epoch": 0.7678225188258105, + "grad_norm": 0.6320165408465757, + "learning_rate": 5.391865499043275e-07, + "loss": 0.7679, + "step": 8514 + }, + { + "epoch": 0.7679127023492808, + "grad_norm": 1.5610741021885552, + "learning_rate": 5.387875890958788e-07, + "loss": 0.9526, + "step": 8515 + }, + { + "epoch": 0.768002885872751, + "grad_norm": 1.429641722899945, + "learning_rate": 5.383887529661298e-07, + "loss": 0.9767, + "step": 8516 + }, + { + "epoch": 0.7680930693962214, + "grad_norm": 1.8679421148949955, + "learning_rate": 5.379900415491116e-07, + "loss": 0.8987, + "step": 8517 + }, + { + "epoch": 0.7681832529196916, + "grad_norm": 1.4168418420681896, + "learning_rate": 5.375914548788447e-07, + "loss": 0.9004, + "step": 8518 + }, + { + "epoch": 0.7682734364431618, + "grad_norm": 1.543706172058415, + "learning_rate": 5.371929929893384e-07, + "loss": 0.8272, + "step": 8519 + }, + { + "epoch": 0.768363619966632, + "grad_norm": 2.092903193706952, + "learning_rate": 5.367946559145917e-07, + "loss": 0.9142, + "step": 8520 + }, + { + "epoch": 0.7684538034901024, + "grad_norm": 1.436676867788848, + "learning_rate": 5.363964436885935e-07, + "loss": 0.9686, + "step": 8521 + }, + { + "epoch": 0.7685439870135726, + "grad_norm": 1.2660886127321094, + "learning_rate": 5.359983563453199e-07, + "loss": 1.0305, + "step": 8522 + }, + { + "epoch": 0.7686341705370429, + "grad_norm": 1.6153037484157573, + "learning_rate": 5.356003939187402e-07, + "loss": 1.064, + "step": 8523 + }, + { + "epoch": 0.7687243540605131, + "grad_norm": 1.8972364480357342, + "learning_rate": 5.352025564428082e-07, + "loss": 0.9241, + "step": 8524 + }, + { + "epoch": 0.7688145375839834, + "grad_norm": 1.961295349783255, + "learning_rate": 5.348048439514723e-07, + "loss": 0.9904, + "step": 8525 + }, + { + "epoch": 0.7689047211074537, + "grad_norm": 1.6559823737657657, + "learning_rate": 5.344072564786653e-07, + "loss": 0.8292, + "step": 8526 + }, + { + "epoch": 0.7689949046309239, + "grad_norm": 1.3929169016723861, + "learning_rate": 5.340097940583123e-07, + "loss": 0.9602, + "step": 8527 + }, + { + "epoch": 0.7690850881543942, + "grad_norm": 2.131083923312713, + "learning_rate": 5.336124567243275e-07, + "loss": 0.9096, + "step": 8528 + }, + { + "epoch": 0.7691752716778645, + "grad_norm": 2.010104878284738, + "learning_rate": 5.33215244510613e-07, + "loss": 0.9539, + "step": 8529 + }, + { + "epoch": 0.7692654552013347, + "grad_norm": 1.700127580006736, + "learning_rate": 5.328181574510624e-07, + "loss": 0.9423, + "step": 8530 + }, + { + "epoch": 0.769355638724805, + "grad_norm": 1.6325708967998918, + "learning_rate": 5.324211955795559e-07, + "loss": 0.9998, + "step": 8531 + }, + { + "epoch": 0.7694458222482753, + "grad_norm": 1.6808976892656389, + "learning_rate": 5.320243589299651e-07, + "loss": 0.874, + "step": 8532 + }, + { + "epoch": 0.7695360057717455, + "grad_norm": 1.5754177932042444, + "learning_rate": 5.316276475361505e-07, + "loss": 0.9056, + "step": 8533 + }, + { + "epoch": 0.7696261892952158, + "grad_norm": 1.8588778047420562, + "learning_rate": 5.312310614319613e-07, + "loss": 0.9894, + "step": 8534 + }, + { + "epoch": 0.769716372818686, + "grad_norm": 7.8028458416631405, + "learning_rate": 5.308346006512367e-07, + "loss": 0.9171, + "step": 8535 + }, + { + "epoch": 0.7698065563421563, + "grad_norm": 1.5076668624512328, + "learning_rate": 5.30438265227805e-07, + "loss": 0.8807, + "step": 8536 + }, + { + "epoch": 0.7698967398656266, + "grad_norm": 0.6878576865425389, + "learning_rate": 5.300420551954837e-07, + "loss": 0.7611, + "step": 8537 + }, + { + "epoch": 0.7699869233890968, + "grad_norm": 1.7951658662121266, + "learning_rate": 5.296459705880798e-07, + "loss": 0.8746, + "step": 8538 + }, + { + "epoch": 0.770077106912567, + "grad_norm": 1.4052890461259637, + "learning_rate": 5.292500114393881e-07, + "loss": 1.0012, + "step": 8539 + }, + { + "epoch": 0.7701672904360374, + "grad_norm": 1.640959278747479, + "learning_rate": 5.288541777831963e-07, + "loss": 0.9721, + "step": 8540 + }, + { + "epoch": 0.7702574739595076, + "grad_norm": 1.3399853867830023, + "learning_rate": 5.284584696532772e-07, + "loss": 1.0025, + "step": 8541 + }, + { + "epoch": 0.7703476574829778, + "grad_norm": 1.423511108773141, + "learning_rate": 5.280628870833954e-07, + "loss": 0.9504, + "step": 8542 + }, + { + "epoch": 0.7704378410064481, + "grad_norm": 1.634441530467246, + "learning_rate": 5.276674301073045e-07, + "loss": 0.9793, + "step": 8543 + }, + { + "epoch": 0.7705280245299184, + "grad_norm": 1.3043351297175256, + "learning_rate": 5.272720987587467e-07, + "loss": 0.9606, + "step": 8544 + }, + { + "epoch": 0.7706182080533887, + "grad_norm": 1.9646245329802254, + "learning_rate": 5.268768930714545e-07, + "loss": 0.9047, + "step": 8545 + }, + { + "epoch": 0.7707083915768589, + "grad_norm": 1.8738347565336975, + "learning_rate": 5.264818130791473e-07, + "loss": 0.9734, + "step": 8546 + }, + { + "epoch": 0.7707985751003291, + "grad_norm": 1.982696063242982, + "learning_rate": 5.260868588155378e-07, + "loss": 0.9525, + "step": 8547 + }, + { + "epoch": 0.7708887586237995, + "grad_norm": 1.5435266186414907, + "learning_rate": 5.256920303143242e-07, + "loss": 0.9749, + "step": 8548 + }, + { + "epoch": 0.7709789421472697, + "grad_norm": 1.506974852333776, + "learning_rate": 5.252973276091956e-07, + "loss": 1.0589, + "step": 8549 + }, + { + "epoch": 0.7710691256707399, + "grad_norm": 1.536395089954574, + "learning_rate": 5.249027507338307e-07, + "loss": 1.0183, + "step": 8550 + }, + { + "epoch": 0.7711593091942102, + "grad_norm": 1.3570200373351626, + "learning_rate": 5.245082997218966e-07, + "loss": 0.922, + "step": 8551 + }, + { + "epoch": 0.7712494927176805, + "grad_norm": 1.8428574680885712, + "learning_rate": 5.241139746070499e-07, + "loss": 0.9753, + "step": 8552 + }, + { + "epoch": 0.7713396762411507, + "grad_norm": 1.7891280649489811, + "learning_rate": 5.237197754229376e-07, + "loss": 0.8329, + "step": 8553 + }, + { + "epoch": 0.771429859764621, + "grad_norm": 1.4800609338949502, + "learning_rate": 5.233257022031931e-07, + "loss": 0.8568, + "step": 8554 + }, + { + "epoch": 0.7715200432880913, + "grad_norm": 1.4890926682257215, + "learning_rate": 5.229317549814432e-07, + "loss": 0.9668, + "step": 8555 + }, + { + "epoch": 0.7716102268115616, + "grad_norm": 1.822334639866562, + "learning_rate": 5.225379337912998e-07, + "loss": 1.0272, + "step": 8556 + }, + { + "epoch": 0.7717004103350318, + "grad_norm": 1.4777413965270667, + "learning_rate": 5.221442386663663e-07, + "loss": 0.9641, + "step": 8557 + }, + { + "epoch": 0.771790593858502, + "grad_norm": 1.3888124376024626, + "learning_rate": 5.217506696402354e-07, + "loss": 0.9935, + "step": 8558 + }, + { + "epoch": 0.7718807773819724, + "grad_norm": 1.7682752720286077, + "learning_rate": 5.213572267464883e-07, + "loss": 0.9685, + "step": 8559 + }, + { + "epoch": 0.7719709609054426, + "grad_norm": 1.5483075069057024, + "learning_rate": 5.209639100186965e-07, + "loss": 0.9292, + "step": 8560 + }, + { + "epoch": 0.7720611444289128, + "grad_norm": 0.7106081860112393, + "learning_rate": 5.205707194904179e-07, + "loss": 0.8421, + "step": 8561 + }, + { + "epoch": 0.7721513279523831, + "grad_norm": 1.6835660644520707, + "learning_rate": 5.201776551952042e-07, + "loss": 0.8766, + "step": 8562 + }, + { + "epoch": 0.7722415114758534, + "grad_norm": 1.884739060288829, + "learning_rate": 5.197847171665914e-07, + "loss": 0.9166, + "step": 8563 + }, + { + "epoch": 0.7723316949993236, + "grad_norm": 1.4797128644982764, + "learning_rate": 5.193919054381095e-07, + "loss": 0.9361, + "step": 8564 + }, + { + "epoch": 0.7724218785227939, + "grad_norm": 1.858612801901945, + "learning_rate": 5.189992200432738e-07, + "loss": 0.9144, + "step": 8565 + }, + { + "epoch": 0.7725120620462641, + "grad_norm": 1.8858939632772855, + "learning_rate": 5.186066610155906e-07, + "loss": 0.96, + "step": 8566 + }, + { + "epoch": 0.7726022455697344, + "grad_norm": 1.527701411602439, + "learning_rate": 5.182142283885555e-07, + "loss": 1.0018, + "step": 8567 + }, + { + "epoch": 0.7726924290932047, + "grad_norm": 0.6812647572940116, + "learning_rate": 5.178219221956528e-07, + "loss": 0.7323, + "step": 8568 + }, + { + "epoch": 0.7727826126166749, + "grad_norm": 1.6336286047597186, + "learning_rate": 5.174297424703565e-07, + "loss": 0.9658, + "step": 8569 + }, + { + "epoch": 0.7728727961401451, + "grad_norm": 1.8948102293386266, + "learning_rate": 5.170376892461299e-07, + "loss": 0.8805, + "step": 8570 + }, + { + "epoch": 0.7729629796636155, + "grad_norm": 0.7425982894447284, + "learning_rate": 5.16645762556424e-07, + "loss": 0.8319, + "step": 8571 + }, + { + "epoch": 0.7730531631870857, + "grad_norm": 2.053761204545609, + "learning_rate": 5.162539624346809e-07, + "loss": 0.9952, + "step": 8572 + }, + { + "epoch": 0.773143346710556, + "grad_norm": 1.6948977899257784, + "learning_rate": 5.158622889143309e-07, + "loss": 1.0136, + "step": 8573 + }, + { + "epoch": 0.7732335302340262, + "grad_norm": 1.6599839064955475, + "learning_rate": 5.154707420287939e-07, + "loss": 0.9088, + "step": 8574 + }, + { + "epoch": 0.7733237137574965, + "grad_norm": 1.4811478720437954, + "learning_rate": 5.150793218114793e-07, + "loss": 0.8766, + "step": 8575 + }, + { + "epoch": 0.7734138972809668, + "grad_norm": 1.501509463259778, + "learning_rate": 5.146880282957837e-07, + "loss": 0.9347, + "step": 8576 + }, + { + "epoch": 0.773504080804437, + "grad_norm": 1.5587343998597165, + "learning_rate": 5.142968615150964e-07, + "loss": 0.9401, + "step": 8577 + }, + { + "epoch": 0.7735942643279073, + "grad_norm": 1.7262637885424859, + "learning_rate": 5.139058215027921e-07, + "loss": 0.8807, + "step": 8578 + }, + { + "epoch": 0.7736844478513776, + "grad_norm": 1.302027703686326, + "learning_rate": 5.135149082922383e-07, + "loss": 0.9363, + "step": 8579 + }, + { + "epoch": 0.7737746313748478, + "grad_norm": 1.789787312237651, + "learning_rate": 5.131241219167879e-07, + "loss": 0.8786, + "step": 8580 + }, + { + "epoch": 0.773864814898318, + "grad_norm": 0.8003124074105444, + "learning_rate": 5.127334624097869e-07, + "loss": 0.8239, + "step": 8581 + }, + { + "epoch": 0.7739549984217884, + "grad_norm": 1.3324257274919797, + "learning_rate": 5.123429298045672e-07, + "loss": 1.0025, + "step": 8582 + }, + { + "epoch": 0.7740451819452586, + "grad_norm": 2.1350696064592998, + "learning_rate": 5.119525241344515e-07, + "loss": 0.8334, + "step": 8583 + }, + { + "epoch": 0.7741353654687289, + "grad_norm": 1.5179314907682406, + "learning_rate": 5.115622454327515e-07, + "loss": 0.9557, + "step": 8584 + }, + { + "epoch": 0.7742255489921991, + "grad_norm": 1.7636234787968148, + "learning_rate": 5.11172093732768e-07, + "loss": 0.9351, + "step": 8585 + }, + { + "epoch": 0.7743157325156694, + "grad_norm": 1.725740884539356, + "learning_rate": 5.107820690677911e-07, + "loss": 0.9786, + "step": 8586 + }, + { + "epoch": 0.7744059160391397, + "grad_norm": 1.6098657989250862, + "learning_rate": 5.103921714710991e-07, + "loss": 0.8951, + "step": 8587 + }, + { + "epoch": 0.7744960995626099, + "grad_norm": 1.5171598547824086, + "learning_rate": 5.100024009759605e-07, + "loss": 0.9632, + "step": 8588 + }, + { + "epoch": 0.7745862830860801, + "grad_norm": 2.121974698900111, + "learning_rate": 5.09612757615633e-07, + "loss": 0.9825, + "step": 8589 + }, + { + "epoch": 0.7746764666095505, + "grad_norm": 1.4119829936475778, + "learning_rate": 5.092232414233628e-07, + "loss": 0.9017, + "step": 8590 + }, + { + "epoch": 0.7747666501330207, + "grad_norm": 1.8351419876213488, + "learning_rate": 5.088338524323858e-07, + "loss": 0.9209, + "step": 8591 + }, + { + "epoch": 0.7748568336564909, + "grad_norm": 1.7244508007239383, + "learning_rate": 5.084445906759271e-07, + "loss": 1.009, + "step": 8592 + }, + { + "epoch": 0.7749470171799612, + "grad_norm": 2.0056395888518637, + "learning_rate": 5.080554561871995e-07, + "loss": 1.0085, + "step": 8593 + }, + { + "epoch": 0.7750372007034315, + "grad_norm": 1.4579554840324893, + "learning_rate": 5.076664489994078e-07, + "loss": 0.9743, + "step": 8594 + }, + { + "epoch": 0.7751273842269017, + "grad_norm": 1.7751404236600996, + "learning_rate": 5.07277569145742e-07, + "loss": 0.9152, + "step": 8595 + }, + { + "epoch": 0.775217567750372, + "grad_norm": 1.5939716110238218, + "learning_rate": 5.068888166593861e-07, + "loss": 0.8644, + "step": 8596 + }, + { + "epoch": 0.7753077512738422, + "grad_norm": 1.761362314643252, + "learning_rate": 5.065001915735087e-07, + "loss": 0.9357, + "step": 8597 + }, + { + "epoch": 0.7753979347973126, + "grad_norm": 1.3612955929605735, + "learning_rate": 5.061116939212702e-07, + "loss": 0.9137, + "step": 8598 + }, + { + "epoch": 0.7754881183207828, + "grad_norm": 3.5810347141180108, + "learning_rate": 5.05723323735819e-07, + "loss": 0.947, + "step": 8599 + }, + { + "epoch": 0.775578301844253, + "grad_norm": 0.6548151269663228, + "learning_rate": 5.053350810502932e-07, + "loss": 0.784, + "step": 8600 + }, + { + "epoch": 0.7756684853677234, + "grad_norm": 1.4799194121973367, + "learning_rate": 5.049469658978202e-07, + "loss": 0.9457, + "step": 8601 + }, + { + "epoch": 0.7757586688911936, + "grad_norm": 0.6385698849498456, + "learning_rate": 5.045589783115147e-07, + "loss": 0.7326, + "step": 8602 + }, + { + "epoch": 0.7758488524146638, + "grad_norm": 1.4947347419701393, + "learning_rate": 5.041711183244842e-07, + "loss": 0.9844, + "step": 8603 + }, + { + "epoch": 0.7759390359381341, + "grad_norm": 1.7950197676736352, + "learning_rate": 5.037833859698211e-07, + "loss": 0.9662, + "step": 8604 + }, + { + "epoch": 0.7760292194616044, + "grad_norm": 1.8103928649648802, + "learning_rate": 5.033957812806096e-07, + "loss": 1.0189, + "step": 8605 + }, + { + "epoch": 0.7761194029850746, + "grad_norm": 1.8032404394897215, + "learning_rate": 5.030083042899223e-07, + "loss": 1.0037, + "step": 8606 + }, + { + "epoch": 0.7762095865085449, + "grad_norm": 1.6688930978748757, + "learning_rate": 5.026209550308207e-07, + "loss": 0.9485, + "step": 8607 + }, + { + "epoch": 0.7762997700320151, + "grad_norm": 1.7742574535534206, + "learning_rate": 5.022337335363558e-07, + "loss": 1.002, + "step": 8608 + }, + { + "epoch": 0.7763899535554855, + "grad_norm": 1.539534598040737, + "learning_rate": 5.018466398395677e-07, + "loss": 0.9745, + "step": 8609 + }, + { + "epoch": 0.7764801370789557, + "grad_norm": 1.4116667374284735, + "learning_rate": 5.01459673973484e-07, + "loss": 0.8818, + "step": 8610 + }, + { + "epoch": 0.7765703206024259, + "grad_norm": 2.0239906291443184, + "learning_rate": 5.01072835971125e-07, + "loss": 1.0126, + "step": 8611 + }, + { + "epoch": 0.7766605041258962, + "grad_norm": 1.4428620055762853, + "learning_rate": 5.006861258654959e-07, + "loss": 0.8551, + "step": 8612 + }, + { + "epoch": 0.7767506876493665, + "grad_norm": 1.4941188436678308, + "learning_rate": 5.002995436895938e-07, + "loss": 0.9802, + "step": 8613 + }, + { + "epoch": 0.7768408711728367, + "grad_norm": 1.8271978912919231, + "learning_rate": 4.999130894764039e-07, + "loss": 0.9084, + "step": 8614 + }, + { + "epoch": 0.776931054696307, + "grad_norm": 1.5094448602593373, + "learning_rate": 4.995267632589006e-07, + "loss": 0.8885, + "step": 8615 + }, + { + "epoch": 0.7770212382197772, + "grad_norm": 0.7261529403934835, + "learning_rate": 4.99140565070048e-07, + "loss": 0.7824, + "step": 8616 + }, + { + "epoch": 0.7771114217432475, + "grad_norm": 1.3917188735504455, + "learning_rate": 4.987544949427969e-07, + "loss": 0.9752, + "step": 8617 + }, + { + "epoch": 0.7772016052667178, + "grad_norm": 0.6751403194738356, + "learning_rate": 4.98368552910091e-07, + "loss": 0.7893, + "step": 8618 + }, + { + "epoch": 0.777291788790188, + "grad_norm": 1.840886948085214, + "learning_rate": 4.979827390048596e-07, + "loss": 1.0643, + "step": 8619 + }, + { + "epoch": 0.7773819723136582, + "grad_norm": 1.2744040536304764, + "learning_rate": 4.975970532600231e-07, + "loss": 0.9125, + "step": 8620 + }, + { + "epoch": 0.7774721558371286, + "grad_norm": 2.330166529651519, + "learning_rate": 4.972114957084901e-07, + "loss": 0.9658, + "step": 8621 + }, + { + "epoch": 0.7775623393605988, + "grad_norm": 1.522672333884396, + "learning_rate": 4.968260663831585e-07, + "loss": 0.9263, + "step": 8622 + }, + { + "epoch": 0.777652522884069, + "grad_norm": 1.3415796806537998, + "learning_rate": 4.964407653169154e-07, + "loss": 1.0601, + "step": 8623 + }, + { + "epoch": 0.7777427064075394, + "grad_norm": 1.4521063238628955, + "learning_rate": 4.960555925426366e-07, + "loss": 0.9175, + "step": 8624 + }, + { + "epoch": 0.7778328899310096, + "grad_norm": 1.8845444939902856, + "learning_rate": 4.956705480931876e-07, + "loss": 1.0088, + "step": 8625 + }, + { + "epoch": 0.7779230734544799, + "grad_norm": 1.3483991703672524, + "learning_rate": 4.952856320014225e-07, + "loss": 0.9266, + "step": 8626 + }, + { + "epoch": 0.7780132569779501, + "grad_norm": 1.3738812316747648, + "learning_rate": 4.949008443001838e-07, + "loss": 0.8888, + "step": 8627 + }, + { + "epoch": 0.7781034405014204, + "grad_norm": 1.4674209100313602, + "learning_rate": 4.945161850223041e-07, + "loss": 0.9619, + "step": 8628 + }, + { + "epoch": 0.7781936240248907, + "grad_norm": 1.6353615753548867, + "learning_rate": 4.941316542006044e-07, + "loss": 0.8833, + "step": 8629 + }, + { + "epoch": 0.7782838075483609, + "grad_norm": 1.4119373601927145, + "learning_rate": 4.937472518678956e-07, + "loss": 0.976, + "step": 8630 + }, + { + "epoch": 0.7783739910718311, + "grad_norm": 2.042063291781227, + "learning_rate": 4.93362978056977e-07, + "loss": 0.9475, + "step": 8631 + }, + { + "epoch": 0.7784641745953015, + "grad_norm": 1.3924653380786716, + "learning_rate": 4.929788328006355e-07, + "loss": 0.8995, + "step": 8632 + }, + { + "epoch": 0.7785543581187717, + "grad_norm": 1.767825215809624, + "learning_rate": 4.925948161316506e-07, + "loss": 1.0208, + "step": 8633 + }, + { + "epoch": 0.778644541642242, + "grad_norm": 1.609974647582994, + "learning_rate": 4.922109280827868e-07, + "loss": 0.9376, + "step": 8634 + }, + { + "epoch": 0.7787347251657122, + "grad_norm": 1.4461749828404764, + "learning_rate": 4.918271686868016e-07, + "loss": 0.9397, + "step": 8635 + }, + { + "epoch": 0.7788249086891825, + "grad_norm": 1.7331921091991322, + "learning_rate": 4.914435379764379e-07, + "loss": 0.9906, + "step": 8636 + }, + { + "epoch": 0.7789150922126528, + "grad_norm": 1.5158611782145626, + "learning_rate": 4.910600359844294e-07, + "loss": 0.9373, + "step": 8637 + }, + { + "epoch": 0.779005275736123, + "grad_norm": 1.5062852942419427, + "learning_rate": 4.90676662743499e-07, + "loss": 0.8553, + "step": 8638 + }, + { + "epoch": 0.7790954592595932, + "grad_norm": 1.5744204711405048, + "learning_rate": 4.902934182863581e-07, + "loss": 0.9395, + "step": 8639 + }, + { + "epoch": 0.7791856427830636, + "grad_norm": 1.7906985069752253, + "learning_rate": 4.899103026457069e-07, + "loss": 0.8251, + "step": 8640 + }, + { + "epoch": 0.7792758263065338, + "grad_norm": 0.6191435371595181, + "learning_rate": 4.895273158542361e-07, + "loss": 0.7322, + "step": 8641 + }, + { + "epoch": 0.779366009830004, + "grad_norm": 2.131891296199179, + "learning_rate": 4.891444579446227e-07, + "loss": 0.9636, + "step": 8642 + }, + { + "epoch": 0.7794561933534743, + "grad_norm": 1.4961232314429618, + "learning_rate": 4.887617289495349e-07, + "loss": 0.9013, + "step": 8643 + }, + { + "epoch": 0.7795463768769446, + "grad_norm": 0.6124756983393002, + "learning_rate": 4.883791289016292e-07, + "loss": 0.7122, + "step": 8644 + }, + { + "epoch": 0.7796365604004148, + "grad_norm": 1.679973347316222, + "learning_rate": 4.879966578335514e-07, + "loss": 0.9563, + "step": 8645 + }, + { + "epoch": 0.7797267439238851, + "grad_norm": 1.8399760528747617, + "learning_rate": 4.876143157779358e-07, + "loss": 0.8498, + "step": 8646 + }, + { + "epoch": 0.7798169274473554, + "grad_norm": 1.8142019371046765, + "learning_rate": 4.872321027674058e-07, + "loss": 0.838, + "step": 8647 + }, + { + "epoch": 0.7799071109708257, + "grad_norm": 1.5167916124180372, + "learning_rate": 4.868500188345748e-07, + "loss": 1.019, + "step": 8648 + }, + { + "epoch": 0.7799972944942959, + "grad_norm": 1.444141057545538, + "learning_rate": 4.864680640120425e-07, + "loss": 1.0118, + "step": 8649 + }, + { + "epoch": 0.7800874780177661, + "grad_norm": 1.7657474036534533, + "learning_rate": 4.860862383324016e-07, + "loss": 0.9544, + "step": 8650 + }, + { + "epoch": 0.7801776615412365, + "grad_norm": 1.936313388838419, + "learning_rate": 4.857045418282295e-07, + "loss": 0.9452, + "step": 8651 + }, + { + "epoch": 0.7802678450647067, + "grad_norm": 1.7016909099848303, + "learning_rate": 4.853229745320966e-07, + "loss": 0.9006, + "step": 8652 + }, + { + "epoch": 0.7803580285881769, + "grad_norm": 1.5086445464516989, + "learning_rate": 4.849415364765587e-07, + "loss": 0.9374, + "step": 8653 + }, + { + "epoch": 0.7804482121116472, + "grad_norm": 1.9594654774195683, + "learning_rate": 4.845602276941631e-07, + "loss": 1.01, + "step": 8654 + }, + { + "epoch": 0.7805383956351175, + "grad_norm": 1.4541527999756008, + "learning_rate": 4.841790482174449e-07, + "loss": 0.9829, + "step": 8655 + }, + { + "epoch": 0.7806285791585877, + "grad_norm": 1.6803573337367765, + "learning_rate": 4.837979980789282e-07, + "loss": 0.9376, + "step": 8656 + }, + { + "epoch": 0.780718762682058, + "grad_norm": 1.5200295697648762, + "learning_rate": 4.834170773111273e-07, + "loss": 0.8895, + "step": 8657 + }, + { + "epoch": 0.7808089462055282, + "grad_norm": 1.6537538828555518, + "learning_rate": 4.830362859465431e-07, + "loss": 0.9681, + "step": 8658 + }, + { + "epoch": 0.7808991297289986, + "grad_norm": 1.5205751574430588, + "learning_rate": 4.826556240176675e-07, + "loss": 1.0143, + "step": 8659 + }, + { + "epoch": 0.7809893132524688, + "grad_norm": 1.7599008319833045, + "learning_rate": 4.822750915569807e-07, + "loss": 0.9623, + "step": 8660 + }, + { + "epoch": 0.781079496775939, + "grad_norm": 2.23721026883646, + "learning_rate": 4.818946885969514e-07, + "loss": 1.0187, + "step": 8661 + }, + { + "epoch": 0.7811696802994093, + "grad_norm": 1.7121270629888208, + "learning_rate": 4.815144151700383e-07, + "loss": 0.9967, + "step": 8662 + }, + { + "epoch": 0.7812598638228796, + "grad_norm": 1.5083718324809612, + "learning_rate": 4.811342713086885e-07, + "loss": 0.9514, + "step": 8663 + }, + { + "epoch": 0.7813500473463498, + "grad_norm": 2.0915557978859103, + "learning_rate": 4.807542570453367e-07, + "loss": 0.9854, + "step": 8664 + }, + { + "epoch": 0.7814402308698201, + "grad_norm": 1.5720442906369538, + "learning_rate": 4.803743724124098e-07, + "loss": 0.9126, + "step": 8665 + }, + { + "epoch": 0.7815304143932903, + "grad_norm": 1.7000092169567897, + "learning_rate": 4.799946174423192e-07, + "loss": 0.9531, + "step": 8666 + }, + { + "epoch": 0.7816205979167606, + "grad_norm": 0.6319065673594911, + "learning_rate": 4.796149921674706e-07, + "loss": 0.7339, + "step": 8667 + }, + { + "epoch": 0.7817107814402309, + "grad_norm": 1.5782277290976034, + "learning_rate": 4.792354966202534e-07, + "loss": 0.9073, + "step": 8668 + }, + { + "epoch": 0.7818009649637011, + "grad_norm": 1.5723340890324706, + "learning_rate": 4.788561308330489e-07, + "loss": 0.771, + "step": 8669 + }, + { + "epoch": 0.7818911484871713, + "grad_norm": 2.274434293900197, + "learning_rate": 4.784768948382272e-07, + "loss": 0.8971, + "step": 8670 + }, + { + "epoch": 0.7819813320106417, + "grad_norm": 1.5595238379503065, + "learning_rate": 4.780977886681461e-07, + "loss": 0.9573, + "step": 8671 + }, + { + "epoch": 0.7820715155341119, + "grad_norm": 1.6853181367734258, + "learning_rate": 4.777188123551541e-07, + "loss": 0.8955, + "step": 8672 + }, + { + "epoch": 0.7821616990575821, + "grad_norm": 1.4768593303092907, + "learning_rate": 4.773399659315856e-07, + "loss": 0.9811, + "step": 8673 + }, + { + "epoch": 0.7822518825810525, + "grad_norm": 1.793123990653955, + "learning_rate": 4.769612494297681e-07, + "loss": 0.9332, + "step": 8674 + }, + { + "epoch": 0.7823420661045227, + "grad_norm": 0.6561411131488035, + "learning_rate": 4.765826628820142e-07, + "loss": 0.7563, + "step": 8675 + }, + { + "epoch": 0.782432249627993, + "grad_norm": 1.9623090954884177, + "learning_rate": 4.7620420632062775e-07, + "loss": 0.8638, + "step": 8676 + }, + { + "epoch": 0.7825224331514632, + "grad_norm": 1.6269555429575744, + "learning_rate": 4.758258797779002e-07, + "loss": 0.9062, + "step": 8677 + }, + { + "epoch": 0.7826126166749335, + "grad_norm": 0.645722392510724, + "learning_rate": 4.7544768328611317e-07, + "loss": 0.7658, + "step": 8678 + }, + { + "epoch": 0.7827028001984038, + "grad_norm": 1.4321918337159383, + "learning_rate": 4.750696168775359e-07, + "loss": 0.9297, + "step": 8679 + }, + { + "epoch": 0.782792983721874, + "grad_norm": 0.6186013264684703, + "learning_rate": 4.746916805844279e-07, + "loss": 0.7535, + "step": 8680 + }, + { + "epoch": 0.7828831672453442, + "grad_norm": 1.6346809611392994, + "learning_rate": 4.743138744390356e-07, + "loss": 0.9743, + "step": 8681 + }, + { + "epoch": 0.7829733507688146, + "grad_norm": 1.5670141243994915, + "learning_rate": 4.739361984735959e-07, + "loss": 1.022, + "step": 8682 + }, + { + "epoch": 0.7830635342922848, + "grad_norm": 2.019608221302763, + "learning_rate": 4.7355865272033455e-07, + "loss": 0.904, + "step": 8683 + }, + { + "epoch": 0.783153717815755, + "grad_norm": 2.0981585648857237, + "learning_rate": 4.7318123721146563e-07, + "loss": 0.9412, + "step": 8684 + }, + { + "epoch": 0.7832439013392253, + "grad_norm": 1.454437811347194, + "learning_rate": 4.728039519791924e-07, + "loss": 0.9363, + "step": 8685 + }, + { + "epoch": 0.7833340848626956, + "grad_norm": 0.7129880655662212, + "learning_rate": 4.72426797055707e-07, + "loss": 0.8003, + "step": 8686 + }, + { + "epoch": 0.7834242683861659, + "grad_norm": 1.5988376418631294, + "learning_rate": 4.720497724731904e-07, + "loss": 0.9179, + "step": 8687 + }, + { + "epoch": 0.7835144519096361, + "grad_norm": 1.6638899879402405, + "learning_rate": 4.7167287826381153e-07, + "loss": 0.9871, + "step": 8688 + }, + { + "epoch": 0.7836046354331063, + "grad_norm": 1.6619796971298642, + "learning_rate": 4.712961144597307e-07, + "loss": 0.9581, + "step": 8689 + }, + { + "epoch": 0.7836948189565767, + "grad_norm": 1.8759527329019123, + "learning_rate": 4.7091948109309343e-07, + "loss": 0.9931, + "step": 8690 + }, + { + "epoch": 0.7837850024800469, + "grad_norm": 1.7285027282578551, + "learning_rate": 4.705429781960384e-07, + "loss": 0.9572, + "step": 8691 + }, + { + "epoch": 0.7838751860035171, + "grad_norm": 1.5653999882651393, + "learning_rate": 4.7016660580068923e-07, + "loss": 0.9458, + "step": 8692 + }, + { + "epoch": 0.7839653695269874, + "grad_norm": 1.6620940264879611, + "learning_rate": 4.6979036393916093e-07, + "loss": 0.9804, + "step": 8693 + }, + { + "epoch": 0.7840555530504577, + "grad_norm": 1.4535197880476949, + "learning_rate": 4.6941425264355603e-07, + "loss": 1.0248, + "step": 8694 + }, + { + "epoch": 0.7841457365739279, + "grad_norm": 1.4296408056753114, + "learning_rate": 4.6903827194596666e-07, + "loss": 0.9613, + "step": 8695 + }, + { + "epoch": 0.7842359200973982, + "grad_norm": 1.507564702203307, + "learning_rate": 4.686624218784743e-07, + "loss": 1.0056, + "step": 8696 + }, + { + "epoch": 0.7843261036208685, + "grad_norm": 1.9345457257437173, + "learning_rate": 4.6828670247314696e-07, + "loss": 0.9409, + "step": 8697 + }, + { + "epoch": 0.7844162871443388, + "grad_norm": 1.5995145776626336, + "learning_rate": 4.679111137620442e-07, + "loss": 0.9641, + "step": 8698 + }, + { + "epoch": 0.784506470667809, + "grad_norm": 1.4084680231382833, + "learning_rate": 4.67535655777213e-07, + "loss": 0.9125, + "step": 8699 + }, + { + "epoch": 0.7845966541912792, + "grad_norm": 1.485926650335612, + "learning_rate": 4.6716032855068956e-07, + "loss": 0.9534, + "step": 8700 + }, + { + "epoch": 0.7846868377147496, + "grad_norm": 1.6592819963674765, + "learning_rate": 4.6678513211449867e-07, + "loss": 0.9917, + "step": 8701 + }, + { + "epoch": 0.7847770212382198, + "grad_norm": 1.7880424023364492, + "learning_rate": 4.6641006650065516e-07, + "loss": 0.9151, + "step": 8702 + }, + { + "epoch": 0.78486720476169, + "grad_norm": 1.491534264866917, + "learning_rate": 4.6603513174115973e-07, + "loss": 0.9365, + "step": 8703 + }, + { + "epoch": 0.7849573882851603, + "grad_norm": 1.804474723129411, + "learning_rate": 4.6566032786800625e-07, + "loss": 0.9684, + "step": 8704 + }, + { + "epoch": 0.7850475718086306, + "grad_norm": 1.5770451845554834, + "learning_rate": 4.6528565491317274e-07, + "loss": 0.9317, + "step": 8705 + }, + { + "epoch": 0.7851377553321008, + "grad_norm": 1.8342203104279156, + "learning_rate": 4.649111129086305e-07, + "loss": 1.016, + "step": 8706 + }, + { + "epoch": 0.7852279388555711, + "grad_norm": 1.5893628942078846, + "learning_rate": 4.6453670188633596e-07, + "loss": 0.973, + "step": 8707 + }, + { + "epoch": 0.7853181223790413, + "grad_norm": 1.9747825720168115, + "learning_rate": 4.641624218782365e-07, + "loss": 0.9663, + "step": 8708 + }, + { + "epoch": 0.7854083059025116, + "grad_norm": 1.5531392271437943, + "learning_rate": 4.6378827291626765e-07, + "loss": 0.9133, + "step": 8709 + }, + { + "epoch": 0.7854984894259819, + "grad_norm": 1.3483713657609195, + "learning_rate": 4.634142550323541e-07, + "loss": 1.003, + "step": 8710 + }, + { + "epoch": 0.7855886729494521, + "grad_norm": 2.0129299628575934, + "learning_rate": 4.6304036825840943e-07, + "loss": 0.9999, + "step": 8711 + }, + { + "epoch": 0.7856788564729223, + "grad_norm": 1.449708100238533, + "learning_rate": 4.626666126263341e-07, + "loss": 0.8562, + "step": 8712 + }, + { + "epoch": 0.7857690399963927, + "grad_norm": 7.814656684739375, + "learning_rate": 4.622929881680213e-07, + "loss": 0.9758, + "step": 8713 + }, + { + "epoch": 0.7858592235198629, + "grad_norm": 1.6570525474204854, + "learning_rate": 4.6191949491534887e-07, + "loss": 0.8628, + "step": 8714 + }, + { + "epoch": 0.7859494070433332, + "grad_norm": 1.6568624695570642, + "learning_rate": 4.6154613290018617e-07, + "loss": 0.9624, + "step": 8715 + }, + { + "epoch": 0.7860395905668034, + "grad_norm": 1.6008523995043327, + "learning_rate": 4.6117290215439043e-07, + "loss": 0.8983, + "step": 8716 + }, + { + "epoch": 0.7861297740902737, + "grad_norm": 1.7339305823795566, + "learning_rate": 4.6079980270980744e-07, + "loss": 0.8179, + "step": 8717 + }, + { + "epoch": 0.786219957613744, + "grad_norm": 1.7234296727629588, + "learning_rate": 4.6042683459827245e-07, + "loss": 0.9235, + "step": 8718 + }, + { + "epoch": 0.7863101411372142, + "grad_norm": 0.722855741441877, + "learning_rate": 4.600539978516098e-07, + "loss": 0.7645, + "step": 8719 + }, + { + "epoch": 0.7864003246606845, + "grad_norm": 1.6620608904319414, + "learning_rate": 4.5968129250163004e-07, + "loss": 0.9105, + "step": 8720 + }, + { + "epoch": 0.7864905081841548, + "grad_norm": 1.7496081322207575, + "learning_rate": 4.5930871858013653e-07, + "loss": 0.8825, + "step": 8721 + }, + { + "epoch": 0.786580691707625, + "grad_norm": 0.6733779287874705, + "learning_rate": 4.589362761189182e-07, + "loss": 0.8229, + "step": 8722 + }, + { + "epoch": 0.7866708752310952, + "grad_norm": 1.6156119074927382, + "learning_rate": 4.585639651497539e-07, + "loss": 1.0126, + "step": 8723 + }, + { + "epoch": 0.7867610587545656, + "grad_norm": 1.4549982445827525, + "learning_rate": 4.581917857044115e-07, + "loss": 0.9034, + "step": 8724 + }, + { + "epoch": 0.7868512422780358, + "grad_norm": 1.8729515329863107, + "learning_rate": 4.5781973781464734e-07, + "loss": 0.9886, + "step": 8725 + }, + { + "epoch": 0.7869414258015061, + "grad_norm": 1.9463119696892666, + "learning_rate": 4.574478215122073e-07, + "loss": 0.9864, + "step": 8726 + }, + { + "epoch": 0.7870316093249763, + "grad_norm": 1.8161122730225867, + "learning_rate": 4.5707603682882357e-07, + "loss": 0.9256, + "step": 8727 + }, + { + "epoch": 0.7871217928484466, + "grad_norm": 2.2496612081834066, + "learning_rate": 4.56704383796221e-07, + "loss": 0.8165, + "step": 8728 + }, + { + "epoch": 0.7872119763719169, + "grad_norm": 1.5999240589001207, + "learning_rate": 4.5633286244610956e-07, + "loss": 0.9893, + "step": 8729 + }, + { + "epoch": 0.7873021598953871, + "grad_norm": 1.333752635749422, + "learning_rate": 4.5596147281018993e-07, + "loss": 1.0162, + "step": 8730 + }, + { + "epoch": 0.7873923434188573, + "grad_norm": 1.5779050919308282, + "learning_rate": 4.5559021492015137e-07, + "loss": 1.0516, + "step": 8731 + }, + { + "epoch": 0.7874825269423277, + "grad_norm": 1.4367269012734374, + "learning_rate": 4.552190888076712e-07, + "loss": 0.8575, + "step": 8732 + }, + { + "epoch": 0.7875727104657979, + "grad_norm": 1.5194158436704233, + "learning_rate": 4.548480945044164e-07, + "loss": 0.8979, + "step": 8733 + }, + { + "epoch": 0.7876628939892681, + "grad_norm": 1.4502826086414946, + "learning_rate": 4.54477232042042e-07, + "loss": 0.9446, + "step": 8734 + }, + { + "epoch": 0.7877530775127384, + "grad_norm": 1.4050654190418002, + "learning_rate": 4.541065014521921e-07, + "loss": 0.9434, + "step": 8735 + }, + { + "epoch": 0.7878432610362087, + "grad_norm": 2.0345599862417343, + "learning_rate": 4.5373590276649996e-07, + "loss": 0.9027, + "step": 8736 + }, + { + "epoch": 0.787933444559679, + "grad_norm": 1.313540681883493, + "learning_rate": 4.533654360165862e-07, + "loss": 0.9131, + "step": 8737 + }, + { + "epoch": 0.7880236280831492, + "grad_norm": 1.7724220628915277, + "learning_rate": 4.5299510123406115e-07, + "loss": 1.0058, + "step": 8738 + }, + { + "epoch": 0.7881138116066194, + "grad_norm": 1.451118314224392, + "learning_rate": 4.5262489845052456e-07, + "loss": 0.9151, + "step": 8739 + }, + { + "epoch": 0.7882039951300898, + "grad_norm": 1.6653317191485444, + "learning_rate": 4.5225482769756353e-07, + "loss": 0.9128, + "step": 8740 + }, + { + "epoch": 0.78829417865356, + "grad_norm": 2.2416899125644068, + "learning_rate": 4.5188488900675545e-07, + "loss": 0.8814, + "step": 8741 + }, + { + "epoch": 0.7883843621770302, + "grad_norm": 1.6190731830608935, + "learning_rate": 4.5151508240966363e-07, + "loss": 0.9789, + "step": 8742 + }, + { + "epoch": 0.7884745457005006, + "grad_norm": 1.4651436867408707, + "learning_rate": 4.511454079378445e-07, + "loss": 0.9957, + "step": 8743 + }, + { + "epoch": 0.7885647292239708, + "grad_norm": 1.6791969470124937, + "learning_rate": 4.507758656228382e-07, + "loss": 0.8509, + "step": 8744 + }, + { + "epoch": 0.788654912747441, + "grad_norm": 1.4735948723319654, + "learning_rate": 4.5040645549617864e-07, + "loss": 1.0101, + "step": 8745 + }, + { + "epoch": 0.7887450962709113, + "grad_norm": 1.5530405954187665, + "learning_rate": 4.5003717758938384e-07, + "loss": 0.8642, + "step": 8746 + }, + { + "epoch": 0.7888352797943816, + "grad_norm": 1.4697183703831076, + "learning_rate": 4.4966803193396365e-07, + "loss": 0.8885, + "step": 8747 + }, + { + "epoch": 0.7889254633178518, + "grad_norm": 1.5629204756506352, + "learning_rate": 4.492990185614154e-07, + "loss": 0.9558, + "step": 8748 + }, + { + "epoch": 0.7890156468413221, + "grad_norm": 1.4668783071291311, + "learning_rate": 4.489301375032255e-07, + "loss": 0.9387, + "step": 8749 + }, + { + "epoch": 0.7891058303647923, + "grad_norm": 1.6518680979093412, + "learning_rate": 4.4856138879086857e-07, + "loss": 0.9524, + "step": 8750 + }, + { + "epoch": 0.7891960138882627, + "grad_norm": 1.2831932311351597, + "learning_rate": 4.481927724558092e-07, + "loss": 0.8981, + "step": 8751 + }, + { + "epoch": 0.7892861974117329, + "grad_norm": 1.4913153371383048, + "learning_rate": 4.478242885294985e-07, + "loss": 0.9185, + "step": 8752 + }, + { + "epoch": 0.7893763809352031, + "grad_norm": 1.4662329838148296, + "learning_rate": 4.474559370433779e-07, + "loss": 0.9088, + "step": 8753 + }, + { + "epoch": 0.7894665644586734, + "grad_norm": 2.0407245045071365, + "learning_rate": 4.470877180288777e-07, + "loss": 0.8905, + "step": 8754 + }, + { + "epoch": 0.7895567479821437, + "grad_norm": 1.523540473783466, + "learning_rate": 4.4671963151741574e-07, + "loss": 0.9142, + "step": 8755 + }, + { + "epoch": 0.7896469315056139, + "grad_norm": 1.3332915398882288, + "learning_rate": 4.4635167754039973e-07, + "loss": 0.9247, + "step": 8756 + }, + { + "epoch": 0.7897371150290842, + "grad_norm": 1.3223515853099788, + "learning_rate": 4.459838561292253e-07, + "loss": 0.9744, + "step": 8757 + }, + { + "epoch": 0.7898272985525544, + "grad_norm": 2.1130721857871806, + "learning_rate": 4.456161673152774e-07, + "loss": 0.8684, + "step": 8758 + }, + { + "epoch": 0.7899174820760247, + "grad_norm": 1.603694111302889, + "learning_rate": 4.4524861112992806e-07, + "loss": 0.9388, + "step": 8759 + }, + { + "epoch": 0.790007665599495, + "grad_norm": 1.8571792588491591, + "learning_rate": 4.448811876045411e-07, + "loss": 0.9458, + "step": 8760 + }, + { + "epoch": 0.7900978491229652, + "grad_norm": 1.7315458299749709, + "learning_rate": 4.445138967704647e-07, + "loss": 1.0213, + "step": 8761 + }, + { + "epoch": 0.7901880326464354, + "grad_norm": 1.5661058399442065, + "learning_rate": 4.4414673865904075e-07, + "loss": 0.9207, + "step": 8762 + }, + { + "epoch": 0.7902782161699058, + "grad_norm": 1.4735203643831019, + "learning_rate": 4.437797133015955e-07, + "loss": 0.9631, + "step": 8763 + }, + { + "epoch": 0.790368399693376, + "grad_norm": 1.7709001042362498, + "learning_rate": 4.4341282072944586e-07, + "loss": 0.8672, + "step": 8764 + }, + { + "epoch": 0.7904585832168463, + "grad_norm": 2.0784676635003376, + "learning_rate": 4.430460609738973e-07, + "loss": 0.9641, + "step": 8765 + }, + { + "epoch": 0.7905487667403166, + "grad_norm": 0.7322909225380501, + "learning_rate": 4.4267943406624386e-07, + "loss": 0.8115, + "step": 8766 + }, + { + "epoch": 0.7906389502637868, + "grad_norm": 1.598611709185121, + "learning_rate": 4.4231294003776853e-07, + "loss": 1.0616, + "step": 8767 + }, + { + "epoch": 0.7907291337872571, + "grad_norm": 1.7448346336729867, + "learning_rate": 4.419465789197416e-07, + "loss": 0.9846, + "step": 8768 + }, + { + "epoch": 0.7908193173107273, + "grad_norm": 1.475711476346131, + "learning_rate": 4.415803507434237e-07, + "loss": 0.939, + "step": 8769 + }, + { + "epoch": 0.7909095008341976, + "grad_norm": 1.6452320626241377, + "learning_rate": 4.4121425554006307e-07, + "loss": 0.9711, + "step": 8770 + }, + { + "epoch": 0.7909996843576679, + "grad_norm": 1.5835614374719311, + "learning_rate": 4.4084829334089744e-07, + "loss": 0.9305, + "step": 8771 + }, + { + "epoch": 0.7910898678811381, + "grad_norm": 1.6220084310046738, + "learning_rate": 4.404824641771525e-07, + "loss": 0.8921, + "step": 8772 + }, + { + "epoch": 0.7911800514046083, + "grad_norm": 1.3681440222591006, + "learning_rate": 4.4011676808004327e-07, + "loss": 1.013, + "step": 8773 + }, + { + "epoch": 0.7912702349280787, + "grad_norm": 1.4002420727029161, + "learning_rate": 4.3975120508077145e-07, + "loss": 0.9583, + "step": 8774 + }, + { + "epoch": 0.7913604184515489, + "grad_norm": 1.3971526567640329, + "learning_rate": 4.39385775210531e-07, + "loss": 0.9949, + "step": 8775 + }, + { + "epoch": 0.7914506019750192, + "grad_norm": 1.6418372625538151, + "learning_rate": 4.390204785005003e-07, + "loss": 0.9643, + "step": 8776 + }, + { + "epoch": 0.7915407854984894, + "grad_norm": 2.547878231612292, + "learning_rate": 4.386553149818504e-07, + "loss": 0.952, + "step": 8777 + }, + { + "epoch": 0.7916309690219597, + "grad_norm": 1.4844569936492813, + "learning_rate": 4.3829028468573793e-07, + "loss": 0.8165, + "step": 8778 + }, + { + "epoch": 0.79172115254543, + "grad_norm": 1.59217726954055, + "learning_rate": 4.3792538764330935e-07, + "loss": 0.8859, + "step": 8779 + }, + { + "epoch": 0.7918113360689002, + "grad_norm": 1.9644028034239138, + "learning_rate": 4.3756062388569994e-07, + "loss": 0.9176, + "step": 8780 + }, + { + "epoch": 0.7919015195923704, + "grad_norm": 0.7321332846158637, + "learning_rate": 4.3719599344403346e-07, + "loss": 0.7812, + "step": 8781 + }, + { + "epoch": 0.7919917031158408, + "grad_norm": 1.3662031717468266, + "learning_rate": 4.3683149634942243e-07, + "loss": 0.9279, + "step": 8782 + }, + { + "epoch": 0.792081886639311, + "grad_norm": 1.8379945958694195, + "learning_rate": 4.364671326329663e-07, + "loss": 0.8325, + "step": 8783 + }, + { + "epoch": 0.7921720701627812, + "grad_norm": 1.845789541007751, + "learning_rate": 4.3610290232575673e-07, + "loss": 0.9647, + "step": 8784 + }, + { + "epoch": 0.7922622536862515, + "grad_norm": 1.7450889068514996, + "learning_rate": 4.357388054588702e-07, + "loss": 0.9517, + "step": 8785 + }, + { + "epoch": 0.7923524372097218, + "grad_norm": 1.4118937094209945, + "learning_rate": 4.3537484206337405e-07, + "loss": 0.8915, + "step": 8786 + }, + { + "epoch": 0.792442620733192, + "grad_norm": 1.611958624278017, + "learning_rate": 4.3501101217032366e-07, + "loss": 0.9785, + "step": 8787 + }, + { + "epoch": 0.7925328042566623, + "grad_norm": 1.7222457015922066, + "learning_rate": 4.346473158107629e-07, + "loss": 0.875, + "step": 8788 + }, + { + "epoch": 0.7926229877801325, + "grad_norm": 1.7282177339750917, + "learning_rate": 4.342837530157244e-07, + "loss": 0.9465, + "step": 8789 + }, + { + "epoch": 0.7927131713036029, + "grad_norm": 1.6060646603300288, + "learning_rate": 4.3392032381622987e-07, + "loss": 0.8467, + "step": 8790 + }, + { + "epoch": 0.7928033548270731, + "grad_norm": 1.309826125406694, + "learning_rate": 4.3355702824328765e-07, + "loss": 0.9081, + "step": 8791 + }, + { + "epoch": 0.7928935383505433, + "grad_norm": 1.4223240572925873, + "learning_rate": 4.3319386632789823e-07, + "loss": 0.9879, + "step": 8792 + }, + { + "epoch": 0.7929837218740137, + "grad_norm": 1.7105325716112776, + "learning_rate": 4.328308381010466e-07, + "loss": 0.937, + "step": 8793 + }, + { + "epoch": 0.7930739053974839, + "grad_norm": 1.5160434577822628, + "learning_rate": 4.3246794359370933e-07, + "loss": 0.9535, + "step": 8794 + }, + { + "epoch": 0.7931640889209541, + "grad_norm": 1.8216240214352606, + "learning_rate": 4.3210518283685025e-07, + "loss": 0.9216, + "step": 8795 + }, + { + "epoch": 0.7932542724444244, + "grad_norm": 1.6559225514448637, + "learning_rate": 4.317425558614225e-07, + "loss": 0.9418, + "step": 8796 + }, + { + "epoch": 0.7933444559678947, + "grad_norm": 1.3718411834650384, + "learning_rate": 4.3138006269836744e-07, + "loss": 0.9111, + "step": 8797 + }, + { + "epoch": 0.793434639491365, + "grad_norm": 1.3601361478648966, + "learning_rate": 4.3101770337861376e-07, + "loss": 0.9377, + "step": 8798 + }, + { + "epoch": 0.7935248230148352, + "grad_norm": 0.690964099435863, + "learning_rate": 4.30655477933082e-07, + "loss": 0.7946, + "step": 8799 + }, + { + "epoch": 0.7936150065383054, + "grad_norm": 1.5906486254663146, + "learning_rate": 4.30293386392677e-07, + "loss": 0.9707, + "step": 8800 + }, + { + "epoch": 0.7937051900617758, + "grad_norm": 1.3005595433538273, + "learning_rate": 4.299314287882967e-07, + "loss": 0.9668, + "step": 8801 + }, + { + "epoch": 0.793795373585246, + "grad_norm": 2.0884385563106966, + "learning_rate": 4.2956960515082353e-07, + "loss": 0.9788, + "step": 8802 + }, + { + "epoch": 0.7938855571087162, + "grad_norm": 1.5678847795552642, + "learning_rate": 4.29207915511131e-07, + "loss": 0.9054, + "step": 8803 + }, + { + "epoch": 0.7939757406321865, + "grad_norm": 1.5420537239344159, + "learning_rate": 4.2884635990008024e-07, + "loss": 1.0032, + "step": 8804 + }, + { + "epoch": 0.7940659241556568, + "grad_norm": 1.5933909292199175, + "learning_rate": 4.284849383485214e-07, + "loss": 0.9087, + "step": 8805 + }, + { + "epoch": 0.794156107679127, + "grad_norm": 1.8857276477381595, + "learning_rate": 4.2812365088729296e-07, + "loss": 0.9781, + "step": 8806 + }, + { + "epoch": 0.7942462912025973, + "grad_norm": 1.5357615124610782, + "learning_rate": 4.2776249754722227e-07, + "loss": 0.9701, + "step": 8807 + }, + { + "epoch": 0.7943364747260675, + "grad_norm": 2.264270035292745, + "learning_rate": 4.27401478359124e-07, + "loss": 1.0117, + "step": 8808 + }, + { + "epoch": 0.7944266582495378, + "grad_norm": 1.4959115099158296, + "learning_rate": 4.2704059335380283e-07, + "loss": 1.0463, + "step": 8809 + }, + { + "epoch": 0.7945168417730081, + "grad_norm": 1.3444403272404606, + "learning_rate": 4.266798425620515e-07, + "loss": 1.0469, + "step": 8810 + }, + { + "epoch": 0.7946070252964783, + "grad_norm": 1.547581520953052, + "learning_rate": 4.263192260146511e-07, + "loss": 0.9323, + "step": 8811 + }, + { + "epoch": 0.7946972088199485, + "grad_norm": 1.6409900077594641, + "learning_rate": 4.2595874374237216e-07, + "loss": 0.9927, + "step": 8812 + }, + { + "epoch": 0.7947873923434189, + "grad_norm": 1.2224879042294217, + "learning_rate": 4.255983957759712e-07, + "loss": 0.9458, + "step": 8813 + }, + { + "epoch": 0.7948775758668891, + "grad_norm": 1.291512628825105, + "learning_rate": 4.2523818214619745e-07, + "loss": 0.8859, + "step": 8814 + }, + { + "epoch": 0.7949677593903594, + "grad_norm": 1.4292465113513346, + "learning_rate": 4.24878102883784e-07, + "loss": 0.9921, + "step": 8815 + }, + { + "epoch": 0.7950579429138297, + "grad_norm": 1.3947119502258463, + "learning_rate": 4.24518158019457e-07, + "loss": 0.9504, + "step": 8816 + }, + { + "epoch": 0.7951481264372999, + "grad_norm": 1.8690518444824589, + "learning_rate": 4.241583475839274e-07, + "loss": 0.8917, + "step": 8817 + }, + { + "epoch": 0.7952383099607702, + "grad_norm": 1.8772796124611877, + "learning_rate": 4.237986716078965e-07, + "loss": 0.9906, + "step": 8818 + }, + { + "epoch": 0.7953284934842404, + "grad_norm": 1.7363319697299953, + "learning_rate": 4.2343913012205433e-07, + "loss": 0.9362, + "step": 8819 + }, + { + "epoch": 0.7954186770077107, + "grad_norm": 1.3867518891149224, + "learning_rate": 4.230797231570784e-07, + "loss": 0.9219, + "step": 8820 + }, + { + "epoch": 0.795508860531181, + "grad_norm": 2.0136497340941477, + "learning_rate": 4.227204507436357e-07, + "loss": 0.9521, + "step": 8821 + }, + { + "epoch": 0.7955990440546512, + "grad_norm": 1.4260205211794643, + "learning_rate": 4.223613129123811e-07, + "loss": 0.8663, + "step": 8822 + }, + { + "epoch": 0.7956892275781214, + "grad_norm": 1.3546201044585375, + "learning_rate": 4.220023096939589e-07, + "loss": 0.8526, + "step": 8823 + }, + { + "epoch": 0.7957794111015918, + "grad_norm": 1.6837508621723372, + "learning_rate": 4.21643441119e-07, + "loss": 0.9384, + "step": 8824 + }, + { + "epoch": 0.795869594625062, + "grad_norm": 0.6593560555073328, + "learning_rate": 4.212847072181256e-07, + "loss": 0.831, + "step": 8825 + }, + { + "epoch": 0.7959597781485322, + "grad_norm": 1.6838713591862413, + "learning_rate": 4.2092610802194505e-07, + "loss": 0.9212, + "step": 8826 + }, + { + "epoch": 0.7960499616720025, + "grad_norm": 1.3786015592680536, + "learning_rate": 4.2056764356105587e-07, + "loss": 0.8554, + "step": 8827 + }, + { + "epoch": 0.7961401451954728, + "grad_norm": 1.5812743083306722, + "learning_rate": 4.202093138660443e-07, + "loss": 0.9918, + "step": 8828 + }, + { + "epoch": 0.7962303287189431, + "grad_norm": 1.3630885996411979, + "learning_rate": 4.198511189674854e-07, + "loss": 0.9805, + "step": 8829 + }, + { + "epoch": 0.7963205122424133, + "grad_norm": 1.6266995858681388, + "learning_rate": 4.1949305889594066e-07, + "loss": 0.9627, + "step": 8830 + }, + { + "epoch": 0.7964106957658835, + "grad_norm": 1.594340944124581, + "learning_rate": 4.191351336819642e-07, + "loss": 0.9382, + "step": 8831 + }, + { + "epoch": 0.7965008792893539, + "grad_norm": 1.6418410381265045, + "learning_rate": 4.187773433560939e-07, + "loss": 0.8872, + "step": 8832 + }, + { + "epoch": 0.7965910628128241, + "grad_norm": 1.477780924262296, + "learning_rate": 4.184196879488604e-07, + "loss": 0.9401, + "step": 8833 + }, + { + "epoch": 0.7966812463362943, + "grad_norm": 1.5807936360400978, + "learning_rate": 4.1806216749077936e-07, + "loss": 1.0103, + "step": 8834 + }, + { + "epoch": 0.7967714298597646, + "grad_norm": 1.4053217049183349, + "learning_rate": 4.177047820123569e-07, + "loss": 0.9916, + "step": 8835 + }, + { + "epoch": 0.7968616133832349, + "grad_norm": 1.5179047889604271, + "learning_rate": 4.1734753154408733e-07, + "loss": 0.8207, + "step": 8836 + }, + { + "epoch": 0.7969517969067051, + "grad_norm": 1.6177994312084951, + "learning_rate": 4.169904161164528e-07, + "loss": 0.9256, + "step": 8837 + }, + { + "epoch": 0.7970419804301754, + "grad_norm": 1.8481362104543535, + "learning_rate": 4.1663343575992526e-07, + "loss": 1.0622, + "step": 8838 + }, + { + "epoch": 0.7971321639536457, + "grad_norm": 1.837775491316876, + "learning_rate": 4.1627659050496275e-07, + "loss": 0.9644, + "step": 8839 + }, + { + "epoch": 0.797222347477116, + "grad_norm": 1.4420480514012177, + "learning_rate": 4.1591988038201453e-07, + "loss": 0.8611, + "step": 8840 + }, + { + "epoch": 0.7973125310005862, + "grad_norm": 1.5720144888717331, + "learning_rate": 4.155633054215164e-07, + "loss": 0.9874, + "step": 8841 + }, + { + "epoch": 0.7974027145240564, + "grad_norm": 1.7243893440869467, + "learning_rate": 4.152068656538934e-07, + "loss": 0.8547, + "step": 8842 + }, + { + "epoch": 0.7974928980475268, + "grad_norm": 1.421559812384325, + "learning_rate": 4.148505611095594e-07, + "loss": 1.0057, + "step": 8843 + }, + { + "epoch": 0.797583081570997, + "grad_norm": 1.769386538215619, + "learning_rate": 4.1449439181891563e-07, + "loss": 0.8229, + "step": 8844 + }, + { + "epoch": 0.7976732650944672, + "grad_norm": 1.6452637986730643, + "learning_rate": 4.14138357812353e-07, + "loss": 0.9636, + "step": 8845 + }, + { + "epoch": 0.7977634486179375, + "grad_norm": 1.3799893715352418, + "learning_rate": 4.137824591202506e-07, + "loss": 1.0, + "step": 8846 + }, + { + "epoch": 0.7978536321414078, + "grad_norm": 1.3131574618917508, + "learning_rate": 4.134266957729737e-07, + "loss": 0.9324, + "step": 8847 + }, + { + "epoch": 0.797943815664878, + "grad_norm": 1.535471177740207, + "learning_rate": 4.1307106780088065e-07, + "loss": 1.0377, + "step": 8848 + }, + { + "epoch": 0.7980339991883483, + "grad_norm": 1.5269977929000043, + "learning_rate": 4.1271557523431387e-07, + "loss": 0.9556, + "step": 8849 + }, + { + "epoch": 0.7981241827118185, + "grad_norm": 1.7547449090657976, + "learning_rate": 4.1236021810360634e-07, + "loss": 0.9305, + "step": 8850 + }, + { + "epoch": 0.7982143662352889, + "grad_norm": 1.7021793224795898, + "learning_rate": 4.120049964390793e-07, + "loss": 0.8944, + "step": 8851 + }, + { + "epoch": 0.7983045497587591, + "grad_norm": 1.682765394027905, + "learning_rate": 4.116499102710418e-07, + "loss": 0.8952, + "step": 8852 + }, + { + "epoch": 0.7983947332822293, + "grad_norm": 1.7714233387922536, + "learning_rate": 4.112949596297928e-07, + "loss": 0.9245, + "step": 8853 + }, + { + "epoch": 0.7984849168056996, + "grad_norm": 1.7207439993637739, + "learning_rate": 4.1094014454561664e-07, + "loss": 0.8872, + "step": 8854 + }, + { + "epoch": 0.7985751003291699, + "grad_norm": 1.6415538519938881, + "learning_rate": 4.1058546504879057e-07, + "loss": 0.8518, + "step": 8855 + }, + { + "epoch": 0.7986652838526401, + "grad_norm": 1.6479054537012672, + "learning_rate": 4.1023092116957583e-07, + "loss": 1.0035, + "step": 8856 + }, + { + "epoch": 0.7987554673761104, + "grad_norm": 1.4843913027219466, + "learning_rate": 4.098765129382249e-07, + "loss": 0.9869, + "step": 8857 + }, + { + "epoch": 0.7988456508995806, + "grad_norm": 1.6967174140720465, + "learning_rate": 4.0952224038497764e-07, + "loss": 1.0191, + "step": 8858 + }, + { + "epoch": 0.7989358344230509, + "grad_norm": 2.0106322678567135, + "learning_rate": 4.091681035400627e-07, + "loss": 0.9399, + "step": 8859 + }, + { + "epoch": 0.7990260179465212, + "grad_norm": 1.8464042825358293, + "learning_rate": 4.088141024336971e-07, + "loss": 0.9773, + "step": 8860 + }, + { + "epoch": 0.7991162014699914, + "grad_norm": 1.3898203375539988, + "learning_rate": 4.0846023709608636e-07, + "loss": 1.0272, + "step": 8861 + }, + { + "epoch": 0.7992063849934617, + "grad_norm": 1.7223938895707591, + "learning_rate": 4.081065075574226e-07, + "loss": 1.0198, + "step": 8862 + }, + { + "epoch": 0.799296568516932, + "grad_norm": 1.5096424442391125, + "learning_rate": 4.077529138478906e-07, + "loss": 0.8784, + "step": 8863 + }, + { + "epoch": 0.7993867520404022, + "grad_norm": 0.7083755274902448, + "learning_rate": 4.073994559976588e-07, + "loss": 0.743, + "step": 8864 + }, + { + "epoch": 0.7994769355638724, + "grad_norm": 2.6146812167342355, + "learning_rate": 4.0704613403688716e-07, + "loss": 0.912, + "step": 8865 + }, + { + "epoch": 0.7995671190873428, + "grad_norm": 1.403571269001704, + "learning_rate": 4.0669294799572264e-07, + "loss": 1.0012, + "step": 8866 + }, + { + "epoch": 0.799657302610813, + "grad_norm": 1.6245880338345302, + "learning_rate": 4.0633989790430113e-07, + "loss": 1.0484, + "step": 8867 + }, + { + "epoch": 0.7997474861342833, + "grad_norm": 1.5984284043635932, + "learning_rate": 4.059869837927477e-07, + "loss": 0.9526, + "step": 8868 + }, + { + "epoch": 0.7998376696577535, + "grad_norm": 1.6857760064784866, + "learning_rate": 4.056342056911728e-07, + "loss": 0.93, + "step": 8869 + }, + { + "epoch": 0.7999278531812238, + "grad_norm": 2.388345466826081, + "learning_rate": 4.052815636296798e-07, + "loss": 0.9625, + "step": 8870 + }, + { + "epoch": 0.8000180367046941, + "grad_norm": 1.3328934728749533, + "learning_rate": 4.0492905763835593e-07, + "loss": 0.9789, + "step": 8871 + }, + { + "epoch": 0.8001082202281643, + "grad_norm": 1.3030083180979586, + "learning_rate": 4.0457668774728115e-07, + "loss": 0.9778, + "step": 8872 + }, + { + "epoch": 0.8001984037516345, + "grad_norm": 1.5534274875500071, + "learning_rate": 4.0422445398651985e-07, + "loss": 0.9628, + "step": 8873 + }, + { + "epoch": 0.8002885872751049, + "grad_norm": 1.5092116273403287, + "learning_rate": 4.0387235638612706e-07, + "loss": 0.8672, + "step": 8874 + }, + { + "epoch": 0.8003787707985751, + "grad_norm": 1.7146294183976585, + "learning_rate": 4.0352039497614586e-07, + "loss": 0.9493, + "step": 8875 + }, + { + "epoch": 0.8004689543220453, + "grad_norm": 2.0796468798126173, + "learning_rate": 4.031685697866074e-07, + "loss": 0.9354, + "step": 8876 + }, + { + "epoch": 0.8005591378455156, + "grad_norm": 1.4065880686977075, + "learning_rate": 4.0281688084753165e-07, + "loss": 0.9832, + "step": 8877 + }, + { + "epoch": 0.8006493213689859, + "grad_norm": 1.571351651687135, + "learning_rate": 4.0246532818892675e-07, + "loss": 1.0008, + "step": 8878 + }, + { + "epoch": 0.8007395048924562, + "grad_norm": 1.6999761187054074, + "learning_rate": 4.0211391184078814e-07, + "loss": 0.9864, + "step": 8879 + }, + { + "epoch": 0.8008296884159264, + "grad_norm": 1.5197514470064177, + "learning_rate": 4.0176263183310135e-07, + "loss": 0.9769, + "step": 8880 + }, + { + "epoch": 0.8009198719393966, + "grad_norm": 1.3228340223234565, + "learning_rate": 4.0141148819583925e-07, + "loss": 1.0021, + "step": 8881 + }, + { + "epoch": 0.801010055462867, + "grad_norm": 0.65095609992709, + "learning_rate": 4.010604809589637e-07, + "loss": 0.7719, + "step": 8882 + }, + { + "epoch": 0.8011002389863372, + "grad_norm": 1.6176213221619087, + "learning_rate": 4.0070961015242475e-07, + "loss": 0.9516, + "step": 8883 + }, + { + "epoch": 0.8011904225098074, + "grad_norm": 1.3631898253644323, + "learning_rate": 4.0035887580615933e-07, + "loss": 1.0174, + "step": 8884 + }, + { + "epoch": 0.8012806060332777, + "grad_norm": 1.560781983961482, + "learning_rate": 4.0000827795009594e-07, + "loss": 0.9999, + "step": 8885 + }, + { + "epoch": 0.801370789556748, + "grad_norm": 2.0807182938757562, + "learning_rate": 3.996578166141475e-07, + "loss": 0.9136, + "step": 8886 + }, + { + "epoch": 0.8014609730802182, + "grad_norm": 1.73433664640342, + "learning_rate": 3.9930749182821955e-07, + "loss": 0.9551, + "step": 8887 + }, + { + "epoch": 0.8015511566036885, + "grad_norm": 0.6293973959657095, + "learning_rate": 3.9895730362220116e-07, + "loss": 0.796, + "step": 8888 + }, + { + "epoch": 0.8016413401271588, + "grad_norm": 1.6564308643554315, + "learning_rate": 3.986072520259749e-07, + "loss": 0.9863, + "step": 8889 + }, + { + "epoch": 0.801731523650629, + "grad_norm": 2.140618094551476, + "learning_rate": 3.9825733706940736e-07, + "loss": 0.9374, + "step": 8890 + }, + { + "epoch": 0.8018217071740993, + "grad_norm": 1.404819142046857, + "learning_rate": 3.979075587823557e-07, + "loss": 0.9655, + "step": 8891 + }, + { + "epoch": 0.8019118906975695, + "grad_norm": 1.9301088930946049, + "learning_rate": 3.9755791719466504e-07, + "loss": 0.8214, + "step": 8892 + }, + { + "epoch": 0.8020020742210399, + "grad_norm": 1.68919527588856, + "learning_rate": 3.9720841233616875e-07, + "loss": 0.9066, + "step": 8893 + }, + { + "epoch": 0.8020922577445101, + "grad_norm": 1.6673114562497684, + "learning_rate": 3.968590442366888e-07, + "loss": 0.9229, + "step": 8894 + }, + { + "epoch": 0.8021824412679803, + "grad_norm": 1.5886212271689766, + "learning_rate": 3.9650981292603423e-07, + "loss": 0.9251, + "step": 8895 + }, + { + "epoch": 0.8022726247914506, + "grad_norm": 1.56026955672093, + "learning_rate": 3.961607184340041e-07, + "loss": 0.9259, + "step": 8896 + }, + { + "epoch": 0.8023628083149209, + "grad_norm": 1.3098013245279252, + "learning_rate": 3.9581176079038505e-07, + "loss": 0.8435, + "step": 8897 + }, + { + "epoch": 0.8024529918383911, + "grad_norm": 1.4215193922106832, + "learning_rate": 3.954629400249516e-07, + "loss": 0.8986, + "step": 8898 + }, + { + "epoch": 0.8025431753618614, + "grad_norm": 1.374949107528736, + "learning_rate": 3.9511425616746787e-07, + "loss": 1.0134, + "step": 8899 + }, + { + "epoch": 0.8026333588853316, + "grad_norm": 1.5600054091579851, + "learning_rate": 3.947657092476853e-07, + "loss": 1.0126, + "step": 8900 + }, + { + "epoch": 0.802723542408802, + "grad_norm": 1.4431765057214103, + "learning_rate": 3.944172992953425e-07, + "loss": 0.9593, + "step": 8901 + }, + { + "epoch": 0.8028137259322722, + "grad_norm": 1.43701370970173, + "learning_rate": 3.9406902634017e-07, + "loss": 0.9063, + "step": 8902 + }, + { + "epoch": 0.8029039094557424, + "grad_norm": 2.022206760645482, + "learning_rate": 3.9372089041188227e-07, + "loss": 0.9818, + "step": 8903 + }, + { + "epoch": 0.8029940929792126, + "grad_norm": 1.7674072857487182, + "learning_rate": 3.9337289154018593e-07, + "loss": 0.8648, + "step": 8904 + }, + { + "epoch": 0.803084276502683, + "grad_norm": 1.5749496391358706, + "learning_rate": 3.930250297547728e-07, + "loss": 0.9445, + "step": 8905 + }, + { + "epoch": 0.8031744600261532, + "grad_norm": 1.5607476135167224, + "learning_rate": 3.9267730508532513e-07, + "loss": 0.8314, + "step": 8906 + }, + { + "epoch": 0.8032646435496235, + "grad_norm": 1.4704511003506164, + "learning_rate": 3.923297175615121e-07, + "loss": 0.9389, + "step": 8907 + }, + { + "epoch": 0.8033548270730937, + "grad_norm": 1.4131756947133, + "learning_rate": 3.9198226721299243e-07, + "loss": 0.9916, + "step": 8908 + }, + { + "epoch": 0.803445010596564, + "grad_norm": 1.8338141316626049, + "learning_rate": 3.916349540694128e-07, + "loss": 0.9279, + "step": 8909 + }, + { + "epoch": 0.8035351941200343, + "grad_norm": 1.8580366174315048, + "learning_rate": 3.912877781604063e-07, + "loss": 0.9517, + "step": 8910 + }, + { + "epoch": 0.8036253776435045, + "grad_norm": 1.5425894355504166, + "learning_rate": 3.909407395155977e-07, + "loss": 0.9744, + "step": 8911 + }, + { + "epoch": 0.8037155611669748, + "grad_norm": 1.8020335578291766, + "learning_rate": 3.9059383816459725e-07, + "loss": 0.8941, + "step": 8912 + }, + { + "epoch": 0.8038057446904451, + "grad_norm": 1.6823210895748602, + "learning_rate": 3.902470741370045e-07, + "loss": 1.0123, + "step": 8913 + }, + { + "epoch": 0.8038959282139153, + "grad_norm": 1.4942271886944487, + "learning_rate": 3.8990044746240746e-07, + "loss": 0.9375, + "step": 8914 + }, + { + "epoch": 0.8039861117373855, + "grad_norm": 1.8448342190413343, + "learning_rate": 3.8955395817038237e-07, + "loss": 1.044, + "step": 8915 + }, + { + "epoch": 0.8040762952608559, + "grad_norm": 1.474124571071793, + "learning_rate": 3.892076062904934e-07, + "loss": 0.917, + "step": 8916 + }, + { + "epoch": 0.8041664787843261, + "grad_norm": 2.0893621442366777, + "learning_rate": 3.8886139185229384e-07, + "loss": 0.9091, + "step": 8917 + }, + { + "epoch": 0.8042566623077964, + "grad_norm": 1.5121162148323168, + "learning_rate": 3.8851531488532284e-07, + "loss": 0.8649, + "step": 8918 + }, + { + "epoch": 0.8043468458312666, + "grad_norm": 1.4030237276024429, + "learning_rate": 3.88169375419112e-07, + "loss": 0.9683, + "step": 8919 + }, + { + "epoch": 0.8044370293547369, + "grad_norm": 1.5794456923212223, + "learning_rate": 3.8782357348317717e-07, + "loss": 0.9241, + "step": 8920 + }, + { + "epoch": 0.8045272128782072, + "grad_norm": 1.8877953405662833, + "learning_rate": 3.8747790910702437e-07, + "loss": 0.995, + "step": 8921 + }, + { + "epoch": 0.8046173964016774, + "grad_norm": 1.8354430486520061, + "learning_rate": 3.8713238232014776e-07, + "loss": 0.9487, + "step": 8922 + }, + { + "epoch": 0.8047075799251476, + "grad_norm": 1.5117436685405892, + "learning_rate": 3.867869931520296e-07, + "loss": 1.0092, + "step": 8923 + }, + { + "epoch": 0.804797763448618, + "grad_norm": 1.4185213215583248, + "learning_rate": 3.864417416321406e-07, + "loss": 0.9631, + "step": 8924 + }, + { + "epoch": 0.8048879469720882, + "grad_norm": 1.6298385623252856, + "learning_rate": 3.8609662778993847e-07, + "loss": 0.9303, + "step": 8925 + }, + { + "epoch": 0.8049781304955584, + "grad_norm": 0.6062170265511877, + "learning_rate": 3.85751651654872e-07, + "loss": 0.7587, + "step": 8926 + }, + { + "epoch": 0.8050683140190287, + "grad_norm": 1.6663973193280308, + "learning_rate": 3.8540681325637505e-07, + "loss": 1.0636, + "step": 8927 + }, + { + "epoch": 0.805158497542499, + "grad_norm": 1.419076534118678, + "learning_rate": 3.8506211262387155e-07, + "loss": 0.9618, + "step": 8928 + }, + { + "epoch": 0.8052486810659693, + "grad_norm": 1.712956530565589, + "learning_rate": 3.847175497867732e-07, + "loss": 0.9595, + "step": 8929 + }, + { + "epoch": 0.8053388645894395, + "grad_norm": 1.6769726481746297, + "learning_rate": 3.843731247744801e-07, + "loss": 0.9016, + "step": 8930 + }, + { + "epoch": 0.8054290481129097, + "grad_norm": 1.5934738219079587, + "learning_rate": 3.8402883761638047e-07, + "loss": 0.9884, + "step": 8931 + }, + { + "epoch": 0.8055192316363801, + "grad_norm": 0.6962667196790007, + "learning_rate": 3.8368468834185076e-07, + "loss": 0.8157, + "step": 8932 + }, + { + "epoch": 0.8056094151598503, + "grad_norm": 1.6246490833211205, + "learning_rate": 3.8334067698025583e-07, + "loss": 0.9926, + "step": 8933 + }, + { + "epoch": 0.8056995986833205, + "grad_norm": 1.5231187535886916, + "learning_rate": 3.8299680356094897e-07, + "loss": 0.8873, + "step": 8934 + }, + { + "epoch": 0.8057897822067909, + "grad_norm": 1.808749890436197, + "learning_rate": 3.8265306811327024e-07, + "loss": 0.8396, + "step": 8935 + }, + { + "epoch": 0.8058799657302611, + "grad_norm": 1.7194371323780109, + "learning_rate": 3.8230947066654994e-07, + "loss": 0.92, + "step": 8936 + }, + { + "epoch": 0.8059701492537313, + "grad_norm": 1.4721005957324005, + "learning_rate": 3.819660112501053e-07, + "loss": 1.0108, + "step": 8937 + }, + { + "epoch": 0.8060603327772016, + "grad_norm": 0.6356359061896829, + "learning_rate": 3.816226898932422e-07, + "loss": 0.799, + "step": 8938 + }, + { + "epoch": 0.8061505163006719, + "grad_norm": 1.6163683304687801, + "learning_rate": 3.812795066252557e-07, + "loss": 0.9459, + "step": 8939 + }, + { + "epoch": 0.8062406998241421, + "grad_norm": 1.9074463764017762, + "learning_rate": 3.8093646147542577e-07, + "loss": 0.9637, + "step": 8940 + }, + { + "epoch": 0.8063308833476124, + "grad_norm": 1.4205359556970822, + "learning_rate": 3.805935544730259e-07, + "loss": 0.919, + "step": 8941 + }, + { + "epoch": 0.8064210668710826, + "grad_norm": 1.4417306591818313, + "learning_rate": 3.802507856473118e-07, + "loss": 0.8855, + "step": 8942 + }, + { + "epoch": 0.806511250394553, + "grad_norm": 1.9652033148843397, + "learning_rate": 3.7990815502753317e-07, + "loss": 0.957, + "step": 8943 + }, + { + "epoch": 0.8066014339180232, + "grad_norm": 1.4135307028599897, + "learning_rate": 3.795656626429231e-07, + "loss": 0.9576, + "step": 8944 + }, + { + "epoch": 0.8066916174414934, + "grad_norm": 1.9179813672056671, + "learning_rate": 3.792233085227059e-07, + "loss": 0.8694, + "step": 8945 + }, + { + "epoch": 0.8067818009649637, + "grad_norm": 1.6639238756492702, + "learning_rate": 3.788810926960928e-07, + "loss": 0.9626, + "step": 8946 + }, + { + "epoch": 0.806871984488434, + "grad_norm": 2.0029327824055807, + "learning_rate": 3.785390151922836e-07, + "loss": 0.9359, + "step": 8947 + }, + { + "epoch": 0.8069621680119042, + "grad_norm": 1.6583819073922603, + "learning_rate": 3.781970760404665e-07, + "loss": 0.9322, + "step": 8948 + }, + { + "epoch": 0.8070523515353745, + "grad_norm": 1.5971775970122124, + "learning_rate": 3.778552752698176e-07, + "loss": 1.0264, + "step": 8949 + }, + { + "epoch": 0.8071425350588447, + "grad_norm": 1.4181644525331718, + "learning_rate": 3.775136129095007e-07, + "loss": 0.86, + "step": 8950 + }, + { + "epoch": 0.807232718582315, + "grad_norm": 1.3960205161382413, + "learning_rate": 3.771720889886685e-07, + "loss": 0.9216, + "step": 8951 + }, + { + "epoch": 0.8073229021057853, + "grad_norm": 0.7170836163582502, + "learning_rate": 3.7683070353646194e-07, + "loss": 0.7965, + "step": 8952 + }, + { + "epoch": 0.8074130856292555, + "grad_norm": 1.9864954995124278, + "learning_rate": 3.7648945658200983e-07, + "loss": 0.9117, + "step": 8953 + }, + { + "epoch": 0.8075032691527257, + "grad_norm": 1.6519979200696437, + "learning_rate": 3.761483481544292e-07, + "loss": 0.8368, + "step": 8954 + }, + { + "epoch": 0.8075934526761961, + "grad_norm": 1.5124468959597044, + "learning_rate": 3.7580737828282525e-07, + "loss": 0.9166, + "step": 8955 + }, + { + "epoch": 0.8076836361996663, + "grad_norm": 1.4341621623318543, + "learning_rate": 3.754665469962921e-07, + "loss": 0.9238, + "step": 8956 + }, + { + "epoch": 0.8077738197231366, + "grad_norm": 1.6173253392249718, + "learning_rate": 3.7512585432390973e-07, + "loss": 0.9059, + "step": 8957 + }, + { + "epoch": 0.8078640032466069, + "grad_norm": 1.6321232267166932, + "learning_rate": 3.7478530029474987e-07, + "loss": 0.8863, + "step": 8958 + }, + { + "epoch": 0.8079541867700771, + "grad_norm": 0.651704783650101, + "learning_rate": 3.7444488493786854e-07, + "loss": 0.829, + "step": 8959 + }, + { + "epoch": 0.8080443702935474, + "grad_norm": 1.7955538993272813, + "learning_rate": 3.7410460828231405e-07, + "loss": 1.0008, + "step": 8960 + }, + { + "epoch": 0.8081345538170176, + "grad_norm": 1.3060032041022587, + "learning_rate": 3.737644703571188e-07, + "loss": 0.8834, + "step": 8961 + }, + { + "epoch": 0.8082247373404879, + "grad_norm": 1.6939674350486342, + "learning_rate": 3.734244711913059e-07, + "loss": 1.0172, + "step": 8962 + }, + { + "epoch": 0.8083149208639582, + "grad_norm": 1.4290378949991591, + "learning_rate": 3.7308461081388584e-07, + "loss": 0.9275, + "step": 8963 + }, + { + "epoch": 0.8084051043874284, + "grad_norm": 2.2944927707208267, + "learning_rate": 3.727448892538576e-07, + "loss": 0.8857, + "step": 8964 + }, + { + "epoch": 0.8084952879108986, + "grad_norm": 1.4432797545486935, + "learning_rate": 3.724053065402086e-07, + "loss": 1.0322, + "step": 8965 + }, + { + "epoch": 0.808585471434369, + "grad_norm": 1.2960318156799386, + "learning_rate": 3.7206586270191285e-07, + "loss": 0.9134, + "step": 8966 + }, + { + "epoch": 0.8086756549578392, + "grad_norm": 1.6018368369746085, + "learning_rate": 3.7172655776793385e-07, + "loss": 0.9975, + "step": 8967 + }, + { + "epoch": 0.8087658384813095, + "grad_norm": 1.6413014788373261, + "learning_rate": 3.7138739176722323e-07, + "loss": 0.8381, + "step": 8968 + }, + { + "epoch": 0.8088560220047797, + "grad_norm": 1.5893250915155765, + "learning_rate": 3.710483647287206e-07, + "loss": 0.913, + "step": 8969 + }, + { + "epoch": 0.80894620552825, + "grad_norm": 1.5579454034340152, + "learning_rate": 3.707094766813532e-07, + "loss": 0.8482, + "step": 8970 + }, + { + "epoch": 0.8090363890517203, + "grad_norm": 1.5936480470846393, + "learning_rate": 3.7037072765403754e-07, + "loss": 0.9146, + "step": 8971 + }, + { + "epoch": 0.8091265725751905, + "grad_norm": 1.8098656814222538, + "learning_rate": 3.700321176756762e-07, + "loss": 0.9054, + "step": 8972 + }, + { + "epoch": 0.8092167560986607, + "grad_norm": 1.6401258890169774, + "learning_rate": 3.69693646775163e-07, + "loss": 0.9965, + "step": 8973 + }, + { + "epoch": 0.8093069396221311, + "grad_norm": 1.7982876076363432, + "learning_rate": 3.693553149813764e-07, + "loss": 0.9304, + "step": 8974 + }, + { + "epoch": 0.8093971231456013, + "grad_norm": 0.6073500415975654, + "learning_rate": 3.690171223231866e-07, + "loss": 0.7531, + "step": 8975 + }, + { + "epoch": 0.8094873066690715, + "grad_norm": 1.3602552523617613, + "learning_rate": 3.6867906882944854e-07, + "loss": 1.0433, + "step": 8976 + }, + { + "epoch": 0.8095774901925418, + "grad_norm": 1.3314406800158014, + "learning_rate": 3.6834115452900737e-07, + "loss": 0.9482, + "step": 8977 + }, + { + "epoch": 0.8096676737160121, + "grad_norm": 1.387865178277287, + "learning_rate": 3.680033794506958e-07, + "loss": 0.8464, + "step": 8978 + }, + { + "epoch": 0.8097578572394823, + "grad_norm": 1.8286184965296124, + "learning_rate": 3.676657436233346e-07, + "loss": 0.7832, + "step": 8979 + }, + { + "epoch": 0.8098480407629526, + "grad_norm": 1.5504442132178704, + "learning_rate": 3.6732824707573305e-07, + "loss": 0.9421, + "step": 8980 + }, + { + "epoch": 0.8099382242864229, + "grad_norm": 1.5540989667586247, + "learning_rate": 3.6699088983668716e-07, + "loss": 0.9835, + "step": 8981 + }, + { + "epoch": 0.8100284078098932, + "grad_norm": 2.1961629177452426, + "learning_rate": 3.6665367193498376e-07, + "loss": 0.9035, + "step": 8982 + }, + { + "epoch": 0.8101185913333634, + "grad_norm": 3.2825157993747007, + "learning_rate": 3.663165933993948e-07, + "loss": 0.9492, + "step": 8983 + }, + { + "epoch": 0.8102087748568336, + "grad_norm": 1.6191829588432476, + "learning_rate": 3.659796542586822e-07, + "loss": 0.9713, + "step": 8984 + }, + { + "epoch": 0.810298958380304, + "grad_norm": 1.317931290813832, + "learning_rate": 3.6564285454159526e-07, + "loss": 0.9029, + "step": 8985 + }, + { + "epoch": 0.8103891419037742, + "grad_norm": 1.3910881299965057, + "learning_rate": 3.653061942768718e-07, + "loss": 0.9916, + "step": 8986 + }, + { + "epoch": 0.8104793254272444, + "grad_norm": 1.786762808459606, + "learning_rate": 3.649696734932375e-07, + "loss": 0.9062, + "step": 8987 + }, + { + "epoch": 0.8105695089507147, + "grad_norm": 1.6840384265611041, + "learning_rate": 3.646332922194064e-07, + "loss": 0.8778, + "step": 8988 + }, + { + "epoch": 0.810659692474185, + "grad_norm": 1.6421379016532882, + "learning_rate": 3.6429705048407943e-07, + "loss": 0.9325, + "step": 8989 + }, + { + "epoch": 0.8107498759976552, + "grad_norm": 1.7303669298928046, + "learning_rate": 3.6396094831594804e-07, + "loss": 0.94, + "step": 8990 + }, + { + "epoch": 0.8108400595211255, + "grad_norm": 1.6917600900318968, + "learning_rate": 3.6362498574368926e-07, + "loss": 0.9006, + "step": 8991 + }, + { + "epoch": 0.8109302430445957, + "grad_norm": 0.7178326433118071, + "learning_rate": 3.6328916279596935e-07, + "loss": 0.8254, + "step": 8992 + }, + { + "epoch": 0.811020426568066, + "grad_norm": 1.6741431706502765, + "learning_rate": 3.6295347950144305e-07, + "loss": 0.9921, + "step": 8993 + }, + { + "epoch": 0.8111106100915363, + "grad_norm": 1.6750147206812547, + "learning_rate": 3.626179358887522e-07, + "loss": 0.7984, + "step": 8994 + }, + { + "epoch": 0.8112007936150065, + "grad_norm": 1.49465379068022, + "learning_rate": 3.6228253198652816e-07, + "loss": 1.005, + "step": 8995 + }, + { + "epoch": 0.8112909771384768, + "grad_norm": 4.040257526159136, + "learning_rate": 3.6194726782338767e-07, + "loss": 0.9003, + "step": 8996 + }, + { + "epoch": 0.8113811606619471, + "grad_norm": 1.8268686525832951, + "learning_rate": 3.6161214342793953e-07, + "loss": 0.9178, + "step": 8997 + }, + { + "epoch": 0.8114713441854173, + "grad_norm": 1.6455685205325916, + "learning_rate": 3.612771588287764e-07, + "loss": 0.9725, + "step": 8998 + }, + { + "epoch": 0.8115615277088876, + "grad_norm": 2.198274182595623, + "learning_rate": 3.609423140544827e-07, + "loss": 1.0102, + "step": 8999 + }, + { + "epoch": 0.8116517112323578, + "grad_norm": 1.911517136281698, + "learning_rate": 3.6060760913362787e-07, + "loss": 0.9685, + "step": 9000 + }, + { + "epoch": 0.8117418947558281, + "grad_norm": 1.4218272106817031, + "learning_rate": 3.6027304409477146e-07, + "loss": 0.9521, + "step": 9001 + }, + { + "epoch": 0.8118320782792984, + "grad_norm": 1.4514915555733967, + "learning_rate": 3.599386189664604e-07, + "loss": 0.9169, + "step": 9002 + }, + { + "epoch": 0.8119222618027686, + "grad_norm": 1.7877373597801527, + "learning_rate": 3.5960433377722945e-07, + "loss": 0.9386, + "step": 9003 + }, + { + "epoch": 0.8120124453262388, + "grad_norm": 1.610918138537528, + "learning_rate": 3.5927018855560174e-07, + "loss": 1.003, + "step": 9004 + }, + { + "epoch": 0.8121026288497092, + "grad_norm": 1.5330186088960427, + "learning_rate": 3.5893618333008904e-07, + "loss": 0.9379, + "step": 9005 + }, + { + "epoch": 0.8121928123731794, + "grad_norm": 1.6835392989626745, + "learning_rate": 3.586023181291893e-07, + "loss": 1.039, + "step": 9006 + }, + { + "epoch": 0.8122829958966497, + "grad_norm": 0.7707111201618407, + "learning_rate": 3.5826859298139044e-07, + "loss": 0.8122, + "step": 9007 + }, + { + "epoch": 0.81237317942012, + "grad_norm": 1.6069707596687688, + "learning_rate": 3.5793500791516773e-07, + "loss": 0.9296, + "step": 9008 + }, + { + "epoch": 0.8124633629435902, + "grad_norm": 1.9475829574682315, + "learning_rate": 3.5760156295898415e-07, + "loss": 0.9158, + "step": 9009 + }, + { + "epoch": 0.8125535464670605, + "grad_norm": 1.4483341283031488, + "learning_rate": 3.5726825814129203e-07, + "loss": 0.9451, + "step": 9010 + }, + { + "epoch": 0.8126437299905307, + "grad_norm": 0.7210793689025269, + "learning_rate": 3.5693509349052886e-07, + "loss": 0.8148, + "step": 9011 + }, + { + "epoch": 0.812733913514001, + "grad_norm": 1.4722808440351212, + "learning_rate": 3.5660206903512433e-07, + "loss": 1.0164, + "step": 9012 + }, + { + "epoch": 0.8128240970374713, + "grad_norm": 1.5855238802629348, + "learning_rate": 3.56269184803492e-07, + "loss": 0.998, + "step": 9013 + }, + { + "epoch": 0.8129142805609415, + "grad_norm": 1.373060722878137, + "learning_rate": 3.5593644082403727e-07, + "loss": 0.9693, + "step": 9014 + }, + { + "epoch": 0.8130044640844117, + "grad_norm": 1.7488721891828716, + "learning_rate": 3.5560383712514994e-07, + "loss": 1.0467, + "step": 9015 + }, + { + "epoch": 0.8130946476078821, + "grad_norm": 1.4778569113466429, + "learning_rate": 3.5527137373521066e-07, + "loss": 0.9537, + "step": 9016 + }, + { + "epoch": 0.8131848311313523, + "grad_norm": 1.5026503830737978, + "learning_rate": 3.5493905068258645e-07, + "loss": 0.9024, + "step": 9017 + }, + { + "epoch": 0.8132750146548225, + "grad_norm": 1.5034553149033505, + "learning_rate": 3.546068679956333e-07, + "loss": 0.9626, + "step": 9018 + }, + { + "epoch": 0.8133651981782928, + "grad_norm": 1.3753584481234646, + "learning_rate": 3.5427482570269487e-07, + "loss": 1.01, + "step": 9019 + }, + { + "epoch": 0.8134553817017631, + "grad_norm": 2.043810581627612, + "learning_rate": 3.539429238321026e-07, + "loss": 0.7968, + "step": 9020 + }, + { + "epoch": 0.8135455652252334, + "grad_norm": 1.4288287475991728, + "learning_rate": 3.536111624121769e-07, + "loss": 0.8, + "step": 9021 + }, + { + "epoch": 0.8136357487487036, + "grad_norm": 1.5890371915139068, + "learning_rate": 3.532795414712244e-07, + "loss": 0.9079, + "step": 9022 + }, + { + "epoch": 0.8137259322721738, + "grad_norm": 1.432270655622404, + "learning_rate": 3.5294806103754124e-07, + "loss": 0.9645, + "step": 9023 + }, + { + "epoch": 0.8138161157956442, + "grad_norm": 1.5041986354625403, + "learning_rate": 3.526167211394115e-07, + "loss": 0.9594, + "step": 9024 + }, + { + "epoch": 0.8139062993191144, + "grad_norm": 1.2676638440506878, + "learning_rate": 3.522855218051066e-07, + "loss": 0.8901, + "step": 9025 + }, + { + "epoch": 0.8139964828425846, + "grad_norm": 1.7035200728209892, + "learning_rate": 3.5195446306288633e-07, + "loss": 0.9485, + "step": 9026 + }, + { + "epoch": 0.8140866663660549, + "grad_norm": 1.3135577435559591, + "learning_rate": 3.51623544940999e-07, + "loss": 0.8893, + "step": 9027 + }, + { + "epoch": 0.8141768498895252, + "grad_norm": 1.4097968406952477, + "learning_rate": 3.5129276746767886e-07, + "loss": 0.9325, + "step": 9028 + }, + { + "epoch": 0.8142670334129954, + "grad_norm": 2.475732610145873, + "learning_rate": 3.5096213067115165e-07, + "loss": 0.9212, + "step": 9029 + }, + { + "epoch": 0.8143572169364657, + "grad_norm": 1.6434956458962202, + "learning_rate": 3.506316345796272e-07, + "loss": 0.9677, + "step": 9030 + }, + { + "epoch": 0.814447400459936, + "grad_norm": 0.6236257464889368, + "learning_rate": 3.5030127922130714e-07, + "loss": 0.7365, + "step": 9031 + }, + { + "epoch": 0.8145375839834063, + "grad_norm": 1.439047436630157, + "learning_rate": 3.4997106462437784e-07, + "loss": 0.8363, + "step": 9032 + }, + { + "epoch": 0.8146277675068765, + "grad_norm": 0.6836566351167456, + "learning_rate": 3.496409908170157e-07, + "loss": 0.763, + "step": 9033 + }, + { + "epoch": 0.8147179510303467, + "grad_norm": 1.7801191521529722, + "learning_rate": 3.493110578273839e-07, + "loss": 0.9115, + "step": 9034 + }, + { + "epoch": 0.8148081345538171, + "grad_norm": 1.5337114596945902, + "learning_rate": 3.489812656836346e-07, + "loss": 1.0043, + "step": 9035 + }, + { + "epoch": 0.8148983180772873, + "grad_norm": 1.305326339651965, + "learning_rate": 3.486516144139078e-07, + "loss": 0.8609, + "step": 9036 + }, + { + "epoch": 0.8149885016007575, + "grad_norm": 1.6489338782837177, + "learning_rate": 3.4832210404632957e-07, + "loss": 0.9169, + "step": 9037 + }, + { + "epoch": 0.8150786851242278, + "grad_norm": 1.3874011013078875, + "learning_rate": 3.479927346090179e-07, + "loss": 0.9543, + "step": 9038 + }, + { + "epoch": 0.8151688686476981, + "grad_norm": 1.8077069409148063, + "learning_rate": 3.4766350613007455e-07, + "loss": 0.9379, + "step": 9039 + }, + { + "epoch": 0.8152590521711683, + "grad_norm": 1.4467124139053946, + "learning_rate": 3.4733441863759173e-07, + "loss": 0.9359, + "step": 9040 + }, + { + "epoch": 0.8153492356946386, + "grad_norm": 1.5545456404637699, + "learning_rate": 3.4700547215964916e-07, + "loss": 0.8416, + "step": 9041 + }, + { + "epoch": 0.8154394192181088, + "grad_norm": 1.4179414268362267, + "learning_rate": 3.46676666724314e-07, + "loss": 0.927, + "step": 9042 + }, + { + "epoch": 0.8155296027415792, + "grad_norm": 2.597915429974353, + "learning_rate": 3.463480023596421e-07, + "loss": 0.9379, + "step": 9043 + }, + { + "epoch": 0.8156197862650494, + "grad_norm": 1.6457114436086266, + "learning_rate": 3.460194790936772e-07, + "loss": 1.0173, + "step": 9044 + }, + { + "epoch": 0.8157099697885196, + "grad_norm": 1.3595300783293252, + "learning_rate": 3.456910969544495e-07, + "loss": 0.9354, + "step": 9045 + }, + { + "epoch": 0.8158001533119899, + "grad_norm": 1.8649994227840563, + "learning_rate": 3.4536285596997994e-07, + "loss": 0.9798, + "step": 9046 + }, + { + "epoch": 0.8158903368354602, + "grad_norm": 1.6738861679286867, + "learning_rate": 3.450347561682747e-07, + "loss": 0.9063, + "step": 9047 + }, + { + "epoch": 0.8159805203589304, + "grad_norm": 1.8514442245113516, + "learning_rate": 3.4470679757732945e-07, + "loss": 0.8876, + "step": 9048 + }, + { + "epoch": 0.8160707038824007, + "grad_norm": 1.5879118707978652, + "learning_rate": 3.4437898022512735e-07, + "loss": 1.03, + "step": 9049 + }, + { + "epoch": 0.8161608874058709, + "grad_norm": 1.5359717767874266, + "learning_rate": 3.4405130413963977e-07, + "loss": 0.9503, + "step": 9050 + }, + { + "epoch": 0.8162510709293412, + "grad_norm": 1.4262596686367321, + "learning_rate": 3.437237693488262e-07, + "loss": 1.0167, + "step": 9051 + }, + { + "epoch": 0.8163412544528115, + "grad_norm": 1.4028549325244106, + "learning_rate": 3.433963758806322e-07, + "loss": 0.9646, + "step": 9052 + }, + { + "epoch": 0.8164314379762817, + "grad_norm": 1.6148917447115305, + "learning_rate": 3.430691237629948e-07, + "loss": 0.8114, + "step": 9053 + }, + { + "epoch": 0.816521621499752, + "grad_norm": 1.7570088223543598, + "learning_rate": 3.427420130238354e-07, + "loss": 0.8951, + "step": 9054 + }, + { + "epoch": 0.8166118050232223, + "grad_norm": 1.295998104491437, + "learning_rate": 3.424150436910658e-07, + "loss": 1.0332, + "step": 9055 + }, + { + "epoch": 0.8167019885466925, + "grad_norm": 1.8798795784266011, + "learning_rate": 3.420882157925842e-07, + "loss": 1.0003, + "step": 9056 + }, + { + "epoch": 0.8167921720701627, + "grad_norm": 1.5982575346018617, + "learning_rate": 3.417615293562777e-07, + "loss": 0.9255, + "step": 9057 + }, + { + "epoch": 0.8168823555936331, + "grad_norm": 1.4812603044755057, + "learning_rate": 3.4143498441002105e-07, + "loss": 0.8602, + "step": 9058 + }, + { + "epoch": 0.8169725391171033, + "grad_norm": 1.4442302879294402, + "learning_rate": 3.411085809816767e-07, + "loss": 0.9546, + "step": 9059 + }, + { + "epoch": 0.8170627226405736, + "grad_norm": 1.6826718809124601, + "learning_rate": 3.407823190990953e-07, + "loss": 1.0602, + "step": 9060 + }, + { + "epoch": 0.8171529061640438, + "grad_norm": 1.7427461010871934, + "learning_rate": 3.4045619879011577e-07, + "loss": 1.0245, + "step": 9061 + }, + { + "epoch": 0.8172430896875141, + "grad_norm": 1.5512733735591382, + "learning_rate": 3.4013022008256334e-07, + "loss": 0.8744, + "step": 9062 + }, + { + "epoch": 0.8173332732109844, + "grad_norm": 1.4991015286560794, + "learning_rate": 3.398043830042532e-07, + "loss": 0.8855, + "step": 9063 + }, + { + "epoch": 0.8174234567344546, + "grad_norm": 1.5571148196779732, + "learning_rate": 3.394786875829871e-07, + "loss": 0.9448, + "step": 9064 + }, + { + "epoch": 0.8175136402579248, + "grad_norm": 1.4561058562490095, + "learning_rate": 3.3915313384655564e-07, + "loss": 0.8287, + "step": 9065 + }, + { + "epoch": 0.8176038237813952, + "grad_norm": 1.8516827636806505, + "learning_rate": 3.388277218227369e-07, + "loss": 0.9656, + "step": 9066 + }, + { + "epoch": 0.8176940073048654, + "grad_norm": 1.9161261197114936, + "learning_rate": 3.3850245153929557e-07, + "loss": 0.9692, + "step": 9067 + }, + { + "epoch": 0.8177841908283356, + "grad_norm": 1.6986664674724858, + "learning_rate": 3.381773230239875e-07, + "loss": 0.9059, + "step": 9068 + }, + { + "epoch": 0.8178743743518059, + "grad_norm": 1.4180265892990105, + "learning_rate": 3.3785233630455247e-07, + "loss": 0.9741, + "step": 9069 + }, + { + "epoch": 0.8179645578752762, + "grad_norm": 1.6871163673900738, + "learning_rate": 3.375274914087221e-07, + "loss": 0.96, + "step": 9070 + }, + { + "epoch": 0.8180547413987465, + "grad_norm": 1.7356789301075213, + "learning_rate": 3.3720278836421234e-07, + "loss": 0.9192, + "step": 9071 + }, + { + "epoch": 0.8181449249222167, + "grad_norm": 1.605064991024347, + "learning_rate": 3.368782271987294e-07, + "loss": 0.9703, + "step": 9072 + }, + { + "epoch": 0.8182351084456869, + "grad_norm": 1.6525858528123516, + "learning_rate": 3.3655380793996636e-07, + "loss": 0.9546, + "step": 9073 + }, + { + "epoch": 0.8183252919691573, + "grad_norm": 1.3548594927695288, + "learning_rate": 3.362295306156047e-07, + "loss": 0.8663, + "step": 9074 + }, + { + "epoch": 0.8184154754926275, + "grad_norm": 1.467474363190824, + "learning_rate": 3.3590539525331327e-07, + "loss": 0.96, + "step": 9075 + }, + { + "epoch": 0.8185056590160977, + "grad_norm": 1.5793104346195892, + "learning_rate": 3.3558140188074967e-07, + "loss": 0.8935, + "step": 9076 + }, + { + "epoch": 0.8185958425395681, + "grad_norm": 1.724796410308457, + "learning_rate": 3.3525755052555817e-07, + "loss": 0.979, + "step": 9077 + }, + { + "epoch": 0.8186860260630383, + "grad_norm": 0.753899292592573, + "learning_rate": 3.3493384121537147e-07, + "loss": 0.85, + "step": 9078 + }, + { + "epoch": 0.8187762095865085, + "grad_norm": 1.6545260111936315, + "learning_rate": 3.3461027397781075e-07, + "loss": 0.9456, + "step": 9079 + }, + { + "epoch": 0.8188663931099788, + "grad_norm": 1.6321346938610886, + "learning_rate": 3.3428684884048397e-07, + "loss": 0.931, + "step": 9080 + }, + { + "epoch": 0.8189565766334491, + "grad_norm": 1.374543070732025, + "learning_rate": 3.3396356583098826e-07, + "loss": 0.947, + "step": 9081 + }, + { + "epoch": 0.8190467601569194, + "grad_norm": 1.547382001923083, + "learning_rate": 3.3364042497690736e-07, + "loss": 0.9799, + "step": 9082 + }, + { + "epoch": 0.8191369436803896, + "grad_norm": 2.0980260654426712, + "learning_rate": 3.3331742630581405e-07, + "loss": 1.0003, + "step": 9083 + }, + { + "epoch": 0.8192271272038598, + "grad_norm": 1.3461783003419887, + "learning_rate": 3.3299456984526717e-07, + "loss": 0.8947, + "step": 9084 + }, + { + "epoch": 0.8193173107273302, + "grad_norm": 1.4075549958158056, + "learning_rate": 3.3267185562281605e-07, + "loss": 0.9823, + "step": 9085 + }, + { + "epoch": 0.8194074942508004, + "grad_norm": 1.6596086384290765, + "learning_rate": 3.3234928366599514e-07, + "loss": 0.9446, + "step": 9086 + }, + { + "epoch": 0.8194976777742706, + "grad_norm": 1.7441997453979614, + "learning_rate": 3.3202685400232946e-07, + "loss": 0.981, + "step": 9087 + }, + { + "epoch": 0.8195878612977409, + "grad_norm": 1.984114381992343, + "learning_rate": 3.317045666593292e-07, + "loss": 0.9783, + "step": 9088 + }, + { + "epoch": 0.8196780448212112, + "grad_norm": 1.759718408490974, + "learning_rate": 3.3138242166449426e-07, + "loss": 0.9493, + "step": 9089 + }, + { + "epoch": 0.8197682283446814, + "grad_norm": 1.502377295156767, + "learning_rate": 3.310604190453117e-07, + "loss": 1.0104, + "step": 9090 + }, + { + "epoch": 0.8198584118681517, + "grad_norm": 1.34424949943387, + "learning_rate": 3.307385588292566e-07, + "loss": 0.8985, + "step": 9091 + }, + { + "epoch": 0.8199485953916219, + "grad_norm": 1.6646208923202082, + "learning_rate": 3.304168410437924e-07, + "loss": 0.9371, + "step": 9092 + }, + { + "epoch": 0.8200387789150922, + "grad_norm": 0.663258615743364, + "learning_rate": 3.300952657163687e-07, + "loss": 0.8047, + "step": 9093 + }, + { + "epoch": 0.8201289624385625, + "grad_norm": 1.7080435468317356, + "learning_rate": 3.297738328744248e-07, + "loss": 0.93, + "step": 9094 + }, + { + "epoch": 0.8202191459620327, + "grad_norm": 1.671701440091342, + "learning_rate": 3.2945254254538714e-07, + "loss": 0.9386, + "step": 9095 + }, + { + "epoch": 0.820309329485503, + "grad_norm": 1.866412396937803, + "learning_rate": 3.2913139475666963e-07, + "loss": 0.9156, + "step": 9096 + }, + { + "epoch": 0.8203995130089733, + "grad_norm": 1.7193395817172132, + "learning_rate": 3.288103895356749e-07, + "loss": 0.8958, + "step": 9097 + }, + { + "epoch": 0.8204896965324435, + "grad_norm": 1.5182759805030217, + "learning_rate": 3.284895269097927e-07, + "loss": 0.8788, + "step": 9098 + }, + { + "epoch": 0.8205798800559138, + "grad_norm": 0.7061447993558532, + "learning_rate": 3.281688069063999e-07, + "loss": 0.783, + "step": 9099 + }, + { + "epoch": 0.8206700635793841, + "grad_norm": 2.3128004136714875, + "learning_rate": 3.2784822955286396e-07, + "loss": 0.7708, + "step": 9100 + }, + { + "epoch": 0.8207602471028543, + "grad_norm": 1.460071408994575, + "learning_rate": 3.275277948765365e-07, + "loss": 1.0205, + "step": 9101 + }, + { + "epoch": 0.8208504306263246, + "grad_norm": 1.4552657243265432, + "learning_rate": 3.2720750290475964e-07, + "loss": 0.9911, + "step": 9102 + }, + { + "epoch": 0.8209406141497948, + "grad_norm": 1.7632955527750183, + "learning_rate": 3.268873536648622e-07, + "loss": 0.9491, + "step": 9103 + }, + { + "epoch": 0.8210307976732651, + "grad_norm": 1.6125189432613525, + "learning_rate": 3.265673471841612e-07, + "loss": 0.9174, + "step": 9104 + }, + { + "epoch": 0.8211209811967354, + "grad_norm": 0.6609397617598989, + "learning_rate": 3.262474834899616e-07, + "loss": 0.7421, + "step": 9105 + }, + { + "epoch": 0.8212111647202056, + "grad_norm": 0.682822353545089, + "learning_rate": 3.2592776260955534e-07, + "loss": 0.8183, + "step": 9106 + }, + { + "epoch": 0.8213013482436758, + "grad_norm": 1.6531194214465332, + "learning_rate": 3.256081845702239e-07, + "loss": 0.986, + "step": 9107 + }, + { + "epoch": 0.8213915317671462, + "grad_norm": 1.5400044613315316, + "learning_rate": 3.2528874939923335e-07, + "loss": 0.9948, + "step": 9108 + }, + { + "epoch": 0.8214817152906164, + "grad_norm": 1.5825638992883702, + "learning_rate": 3.2496945712384217e-07, + "loss": 0.9003, + "step": 9109 + }, + { + "epoch": 0.8215718988140867, + "grad_norm": 1.652165757280249, + "learning_rate": 3.246503077712923e-07, + "loss": 0.9294, + "step": 9110 + }, + { + "epoch": 0.8216620823375569, + "grad_norm": 1.5057077690230636, + "learning_rate": 3.2433130136881625e-07, + "loss": 1.0364, + "step": 9111 + }, + { + "epoch": 0.8217522658610272, + "grad_norm": 1.854067560469027, + "learning_rate": 3.2401243794363287e-07, + "loss": 0.9432, + "step": 9112 + }, + { + "epoch": 0.8218424493844975, + "grad_norm": 1.446332417005073, + "learning_rate": 3.236937175229495e-07, + "loss": 1.0048, + "step": 9113 + }, + { + "epoch": 0.8219326329079677, + "grad_norm": 1.799239882934333, + "learning_rate": 3.233751401339615e-07, + "loss": 0.9168, + "step": 9114 + }, + { + "epoch": 0.8220228164314379, + "grad_norm": 1.8772004884435936, + "learning_rate": 3.2305670580385157e-07, + "loss": 1.0225, + "step": 9115 + }, + { + "epoch": 0.8221129999549083, + "grad_norm": 1.408970849733977, + "learning_rate": 3.227384145597898e-07, + "loss": 0.9452, + "step": 9116 + }, + { + "epoch": 0.8222031834783785, + "grad_norm": 1.8048935149211622, + "learning_rate": 3.224202664289346e-07, + "loss": 0.9306, + "step": 9117 + }, + { + "epoch": 0.8222933670018487, + "grad_norm": 1.5710247948460643, + "learning_rate": 3.2210226143843257e-07, + "loss": 0.8881, + "step": 9118 + }, + { + "epoch": 0.822383550525319, + "grad_norm": 0.6832222283180623, + "learning_rate": 3.217843996154173e-07, + "loss": 0.8153, + "step": 9119 + }, + { + "epoch": 0.8224737340487893, + "grad_norm": 0.653534060239279, + "learning_rate": 3.2146668098701055e-07, + "loss": 0.7294, + "step": 9120 + }, + { + "epoch": 0.8225639175722596, + "grad_norm": 2.0575432603443544, + "learning_rate": 3.2114910558032215e-07, + "loss": 1.077, + "step": 9121 + }, + { + "epoch": 0.8226541010957298, + "grad_norm": 1.642096014405407, + "learning_rate": 3.2083167342244945e-07, + "loss": 0.9747, + "step": 9122 + }, + { + "epoch": 0.8227442846192, + "grad_norm": 1.882841814374849, + "learning_rate": 3.205143845404763e-07, + "loss": 1.0461, + "step": 9123 + }, + { + "epoch": 0.8228344681426704, + "grad_norm": 1.574766153807754, + "learning_rate": 3.201972389614773e-07, + "loss": 0.9173, + "step": 9124 + }, + { + "epoch": 0.8229246516661406, + "grad_norm": 1.7782759349298782, + "learning_rate": 3.198802367125115e-07, + "loss": 0.938, + "step": 9125 + }, + { + "epoch": 0.8230148351896108, + "grad_norm": 1.4604245677667151, + "learning_rate": 3.195633778206288e-07, + "loss": 0.9552, + "step": 9126 + }, + { + "epoch": 0.8231050187130812, + "grad_norm": 1.7385357959694667, + "learning_rate": 3.19246662312864e-07, + "loss": 1.0208, + "step": 9127 + }, + { + "epoch": 0.8231952022365514, + "grad_norm": 0.6671189596807944, + "learning_rate": 3.189300902162417e-07, + "loss": 0.7793, + "step": 9128 + }, + { + "epoch": 0.8232853857600216, + "grad_norm": 1.6576938453659014, + "learning_rate": 3.1861366155777327e-07, + "loss": 0.96, + "step": 9129 + }, + { + "epoch": 0.8233755692834919, + "grad_norm": 1.3019856327277923, + "learning_rate": 3.182973763644583e-07, + "loss": 0.9539, + "step": 9130 + }, + { + "epoch": 0.8234657528069622, + "grad_norm": 1.4316491182707651, + "learning_rate": 3.1798123466328463e-07, + "loss": 0.8716, + "step": 9131 + }, + { + "epoch": 0.8235559363304324, + "grad_norm": 1.72835954726574, + "learning_rate": 3.17665236481226e-07, + "loss": 0.9491, + "step": 9132 + }, + { + "epoch": 0.8236461198539027, + "grad_norm": 1.8543112266279262, + "learning_rate": 3.1734938184524576e-07, + "loss": 0.9109, + "step": 9133 + }, + { + "epoch": 0.8237363033773729, + "grad_norm": 1.5881378994792774, + "learning_rate": 3.1703367078229427e-07, + "loss": 0.9596, + "step": 9134 + }, + { + "epoch": 0.8238264869008433, + "grad_norm": 1.74989059651224, + "learning_rate": 3.167181033193096e-07, + "loss": 0.9079, + "step": 9135 + }, + { + "epoch": 0.8239166704243135, + "grad_norm": 1.6872972437238136, + "learning_rate": 3.16402679483218e-07, + "loss": 0.9585, + "step": 9136 + }, + { + "epoch": 0.8240068539477837, + "grad_norm": 1.313015473276789, + "learning_rate": 3.1608739930093366e-07, + "loss": 0.879, + "step": 9137 + }, + { + "epoch": 0.824097037471254, + "grad_norm": 1.6395760952483984, + "learning_rate": 3.157722627993562e-07, + "loss": 0.863, + "step": 9138 + }, + { + "epoch": 0.8241872209947243, + "grad_norm": 1.490275411178323, + "learning_rate": 3.1545727000537727e-07, + "loss": 0.9104, + "step": 9139 + }, + { + "epoch": 0.8242774045181945, + "grad_norm": 1.4818935332311824, + "learning_rate": 3.151424209458713e-07, + "loss": 0.9486, + "step": 9140 + }, + { + "epoch": 0.8243675880416648, + "grad_norm": 1.5601543367285309, + "learning_rate": 3.148277156477053e-07, + "loss": 0.9659, + "step": 9141 + }, + { + "epoch": 0.824457771565135, + "grad_norm": 1.7113200417584364, + "learning_rate": 3.145131541377299e-07, + "loss": 0.9532, + "step": 9142 + }, + { + "epoch": 0.8245479550886053, + "grad_norm": 1.6487331034942003, + "learning_rate": 3.1419873644278606e-07, + "loss": 0.9134, + "step": 9143 + }, + { + "epoch": 0.8246381386120756, + "grad_norm": 1.4391243091940433, + "learning_rate": 3.1388446258970147e-07, + "loss": 0.968, + "step": 9144 + }, + { + "epoch": 0.8247283221355458, + "grad_norm": 1.7293002831490303, + "learning_rate": 3.1357033260529145e-07, + "loss": 0.9344, + "step": 9145 + }, + { + "epoch": 0.824818505659016, + "grad_norm": 1.5892484332958, + "learning_rate": 3.1325634651636025e-07, + "loss": 1.0222, + "step": 9146 + }, + { + "epoch": 0.8249086891824864, + "grad_norm": 1.4803640685904702, + "learning_rate": 3.1294250434969694e-07, + "loss": 0.9856, + "step": 9147 + }, + { + "epoch": 0.8249988727059566, + "grad_norm": 1.6509704654715955, + "learning_rate": 3.1262880613208274e-07, + "loss": 0.9906, + "step": 9148 + }, + { + "epoch": 0.8250890562294269, + "grad_norm": 1.470155327693566, + "learning_rate": 3.123152518902823e-07, + "loss": 0.9877, + "step": 9149 + }, + { + "epoch": 0.8251792397528972, + "grad_norm": 1.3208520452736332, + "learning_rate": 3.1200184165105017e-07, + "loss": 0.9011, + "step": 9150 + }, + { + "epoch": 0.8252694232763674, + "grad_norm": 1.647240733120746, + "learning_rate": 3.116885754411287e-07, + "loss": 0.9614, + "step": 9151 + }, + { + "epoch": 0.8253596067998377, + "grad_norm": 1.6185934248797895, + "learning_rate": 3.1137545328724703e-07, + "loss": 0.8839, + "step": 9152 + }, + { + "epoch": 0.8254497903233079, + "grad_norm": 2.0087177299181556, + "learning_rate": 3.1106247521612285e-07, + "loss": 0.9533, + "step": 9153 + }, + { + "epoch": 0.8255399738467782, + "grad_norm": 2.3795588057165724, + "learning_rate": 3.107496412544612e-07, + "loss": 0.9641, + "step": 9154 + }, + { + "epoch": 0.8256301573702485, + "grad_norm": 1.4023152723050774, + "learning_rate": 3.1043695142895397e-07, + "loss": 0.9776, + "step": 9155 + }, + { + "epoch": 0.8257203408937187, + "grad_norm": 1.4066308671379257, + "learning_rate": 3.101244057662828e-07, + "loss": 0.8682, + "step": 9156 + }, + { + "epoch": 0.8258105244171889, + "grad_norm": 1.738724555591096, + "learning_rate": 3.098120042931152e-07, + "loss": 0.9311, + "step": 9157 + }, + { + "epoch": 0.8259007079406593, + "grad_norm": 1.4752881237457263, + "learning_rate": 3.0949974703610647e-07, + "loss": 0.99, + "step": 9158 + }, + { + "epoch": 0.8259908914641295, + "grad_norm": 1.505683938183609, + "learning_rate": 3.0918763402190107e-07, + "loss": 0.9366, + "step": 9159 + }, + { + "epoch": 0.8260810749875998, + "grad_norm": 1.56847718438873, + "learning_rate": 3.088756652771296e-07, + "loss": 0.9022, + "step": 9160 + }, + { + "epoch": 0.82617125851107, + "grad_norm": 1.7028377185675583, + "learning_rate": 3.0856384082841147e-07, + "loss": 1.0427, + "step": 9161 + }, + { + "epoch": 0.8262614420345403, + "grad_norm": 1.3930606728270991, + "learning_rate": 3.0825216070235207e-07, + "loss": 0.8903, + "step": 9162 + }, + { + "epoch": 0.8263516255580106, + "grad_norm": 2.067904933035548, + "learning_rate": 3.0794062492554764e-07, + "loss": 1.0607, + "step": 9163 + }, + { + "epoch": 0.8264418090814808, + "grad_norm": 1.6839042784336056, + "learning_rate": 3.076292335245783e-07, + "loss": 0.9562, + "step": 9164 + }, + { + "epoch": 0.826531992604951, + "grad_norm": 1.8962817098522353, + "learning_rate": 3.073179865260145e-07, + "loss": 1.0161, + "step": 9165 + }, + { + "epoch": 0.8266221761284214, + "grad_norm": 1.4060545043690844, + "learning_rate": 3.070068839564135e-07, + "loss": 0.9504, + "step": 9166 + }, + { + "epoch": 0.8267123596518916, + "grad_norm": 1.9254869464440711, + "learning_rate": 3.0669592584232006e-07, + "loss": 0.9858, + "step": 9167 + }, + { + "epoch": 0.8268025431753618, + "grad_norm": 1.3986498282582356, + "learning_rate": 3.063851122102672e-07, + "loss": 0.8612, + "step": 9168 + }, + { + "epoch": 0.8268927266988321, + "grad_norm": 1.7404701424273006, + "learning_rate": 3.06074443086775e-07, + "loss": 0.8689, + "step": 9169 + }, + { + "epoch": 0.8269829102223024, + "grad_norm": 0.6144010925573207, + "learning_rate": 3.057639184983514e-07, + "loss": 0.7324, + "step": 9170 + }, + { + "epoch": 0.8270730937457726, + "grad_norm": 1.2472186138541725, + "learning_rate": 3.054535384714927e-07, + "loss": 0.9509, + "step": 9171 + }, + { + "epoch": 0.8271632772692429, + "grad_norm": 1.7273245118103755, + "learning_rate": 3.0514330303268135e-07, + "loss": 1.1166, + "step": 9172 + }, + { + "epoch": 0.8272534607927132, + "grad_norm": 1.5932486998587008, + "learning_rate": 3.0483321220838876e-07, + "loss": 0.9015, + "step": 9173 + }, + { + "epoch": 0.8273436443161835, + "grad_norm": 1.402004615346193, + "learning_rate": 3.045232660250734e-07, + "loss": 0.8079, + "step": 9174 + }, + { + "epoch": 0.8274338278396537, + "grad_norm": 1.547236236624462, + "learning_rate": 3.0421346450918185e-07, + "loss": 0.9768, + "step": 9175 + }, + { + "epoch": 0.8275240113631239, + "grad_norm": 1.3360292726639098, + "learning_rate": 3.039038076871485e-07, + "loss": 0.8981, + "step": 9176 + }, + { + "epoch": 0.8276141948865943, + "grad_norm": 1.3796532096360112, + "learning_rate": 3.035942955853934e-07, + "loss": 0.9642, + "step": 9177 + }, + { + "epoch": 0.8277043784100645, + "grad_norm": 1.6365982047690109, + "learning_rate": 3.0328492823032804e-07, + "loss": 0.9235, + "step": 9178 + }, + { + "epoch": 0.8277945619335347, + "grad_norm": 1.4903175660757362, + "learning_rate": 3.029757056483471e-07, + "loss": 0.8872, + "step": 9179 + }, + { + "epoch": 0.827884745457005, + "grad_norm": 1.3953980388671863, + "learning_rate": 3.026666278658372e-07, + "loss": 0.8348, + "step": 9180 + }, + { + "epoch": 0.8279749289804753, + "grad_norm": 1.650282347247833, + "learning_rate": 3.023576949091691e-07, + "loss": 0.9452, + "step": 9181 + }, + { + "epoch": 0.8280651125039455, + "grad_norm": 1.5812934567959664, + "learning_rate": 3.020489068047032e-07, + "loss": 0.8734, + "step": 9182 + }, + { + "epoch": 0.8281552960274158, + "grad_norm": 1.5002026421042955, + "learning_rate": 3.017402635787869e-07, + "loss": 0.9102, + "step": 9183 + }, + { + "epoch": 0.828245479550886, + "grad_norm": 1.5592802825477645, + "learning_rate": 3.0143176525775537e-07, + "loss": 0.9169, + "step": 9184 + }, + { + "epoch": 0.8283356630743564, + "grad_norm": 1.5491398886010765, + "learning_rate": 3.0112341186793155e-07, + "loss": 0.8929, + "step": 9185 + }, + { + "epoch": 0.8284258465978266, + "grad_norm": 1.4559028080632612, + "learning_rate": 3.008152034356264e-07, + "loss": 0.8819, + "step": 9186 + }, + { + "epoch": 0.8285160301212968, + "grad_norm": 0.6527028037927861, + "learning_rate": 3.005071399871366e-07, + "loss": 0.7653, + "step": 9187 + }, + { + "epoch": 0.828606213644767, + "grad_norm": 1.6531789125474228, + "learning_rate": 3.0019922154874853e-07, + "loss": 1.0095, + "step": 9188 + }, + { + "epoch": 0.8286963971682374, + "grad_norm": 1.7324573121860039, + "learning_rate": 2.998914481467356e-07, + "loss": 0.9056, + "step": 9189 + }, + { + "epoch": 0.8287865806917076, + "grad_norm": 2.0973992501069905, + "learning_rate": 2.9958381980735837e-07, + "loss": 0.897, + "step": 9190 + }, + { + "epoch": 0.8288767642151779, + "grad_norm": 1.569646513804605, + "learning_rate": 2.992763365568658e-07, + "loss": 0.8767, + "step": 9191 + }, + { + "epoch": 0.8289669477386481, + "grad_norm": 1.5825373840737433, + "learning_rate": 2.98968998421494e-07, + "loss": 0.9456, + "step": 9192 + }, + { + "epoch": 0.8290571312621184, + "grad_norm": 1.4786517988861485, + "learning_rate": 2.98661805427467e-07, + "loss": 0.9389, + "step": 9193 + }, + { + "epoch": 0.8291473147855887, + "grad_norm": 1.3482549249809141, + "learning_rate": 2.9835475760099483e-07, + "loss": 0.9608, + "step": 9194 + }, + { + "epoch": 0.8292374983090589, + "grad_norm": 1.392639158818757, + "learning_rate": 2.9804785496827856e-07, + "loss": 0.9613, + "step": 9195 + }, + { + "epoch": 0.8293276818325293, + "grad_norm": 1.836482904464205, + "learning_rate": 2.977410975555028e-07, + "loss": 0.9397, + "step": 9196 + }, + { + "epoch": 0.8294178653559995, + "grad_norm": 1.6202684622714667, + "learning_rate": 2.9743448538884376e-07, + "loss": 1.0094, + "step": 9197 + }, + { + "epoch": 0.8295080488794697, + "grad_norm": 1.4630700524664129, + "learning_rate": 2.9712801849446154e-07, + "loss": 0.9368, + "step": 9198 + }, + { + "epoch": 0.82959823240294, + "grad_norm": 1.5856399632861697, + "learning_rate": 2.9682169689850665e-07, + "loss": 0.8955, + "step": 9199 + }, + { + "epoch": 0.8296884159264103, + "grad_norm": 0.6799501810336873, + "learning_rate": 2.9651552062711573e-07, + "loss": 0.7285, + "step": 9200 + }, + { + "epoch": 0.8297785994498805, + "grad_norm": 1.6183756273886738, + "learning_rate": 2.9620948970641333e-07, + "loss": 1.0285, + "step": 9201 + }, + { + "epoch": 0.8298687829733508, + "grad_norm": 1.466443868895338, + "learning_rate": 2.959036041625125e-07, + "loss": 0.9627, + "step": 9202 + }, + { + "epoch": 0.829958966496821, + "grad_norm": 1.4869794784340695, + "learning_rate": 2.95597864021512e-07, + "loss": 0.8528, + "step": 9203 + }, + { + "epoch": 0.8300491500202913, + "grad_norm": 1.9173749637744268, + "learning_rate": 2.9529226930949966e-07, + "loss": 0.9937, + "step": 9204 + }, + { + "epoch": 0.8301393335437616, + "grad_norm": 1.5900156738900224, + "learning_rate": 2.949868200525505e-07, + "loss": 0.9629, + "step": 9205 + }, + { + "epoch": 0.8302295170672318, + "grad_norm": 1.6885759667457048, + "learning_rate": 2.9468151627672734e-07, + "loss": 0.9436, + "step": 9206 + }, + { + "epoch": 0.830319700590702, + "grad_norm": 1.5285053177760706, + "learning_rate": 2.9437635800808026e-07, + "loss": 0.9009, + "step": 9207 + }, + { + "epoch": 0.8304098841141724, + "grad_norm": 1.7112938496758394, + "learning_rate": 2.940713452726473e-07, + "loss": 0.9365, + "step": 9208 + }, + { + "epoch": 0.8305000676376426, + "grad_norm": 1.582413088082707, + "learning_rate": 2.937664780964526e-07, + "loss": 0.9411, + "step": 9209 + }, + { + "epoch": 0.8305902511611128, + "grad_norm": 1.7567195631142567, + "learning_rate": 2.9346175650551133e-07, + "loss": 0.9607, + "step": 9210 + }, + { + "epoch": 0.8306804346845831, + "grad_norm": 1.4728546413687402, + "learning_rate": 2.931571805258215e-07, + "loss": 0.9653, + "step": 9211 + }, + { + "epoch": 0.8307706182080534, + "grad_norm": 1.5863790414205778, + "learning_rate": 2.9285275018337353e-07, + "loss": 0.9949, + "step": 9212 + }, + { + "epoch": 0.8308608017315237, + "grad_norm": 1.7237304659836907, + "learning_rate": 2.9254846550414146e-07, + "loss": 1.0509, + "step": 9213 + }, + { + "epoch": 0.8309509852549939, + "grad_norm": 1.711313981391993, + "learning_rate": 2.922443265140893e-07, + "loss": 0.9961, + "step": 9214 + }, + { + "epoch": 0.8310411687784641, + "grad_norm": 1.7921052070754464, + "learning_rate": 2.919403332391674e-07, + "loss": 0.9388, + "step": 9215 + }, + { + "epoch": 0.8311313523019345, + "grad_norm": 1.4533156403214107, + "learning_rate": 2.9163648570531464e-07, + "loss": 0.9695, + "step": 9216 + }, + { + "epoch": 0.8312215358254047, + "grad_norm": 1.7138294878777431, + "learning_rate": 2.9133278393845717e-07, + "loss": 0.8898, + "step": 9217 + }, + { + "epoch": 0.8313117193488749, + "grad_norm": 1.5333363080341398, + "learning_rate": 2.9102922796450703e-07, + "loss": 1.0192, + "step": 9218 + }, + { + "epoch": 0.8314019028723453, + "grad_norm": 1.3515530624501155, + "learning_rate": 2.907258178093672e-07, + "loss": 0.9061, + "step": 9219 + }, + { + "epoch": 0.8314920863958155, + "grad_norm": 1.3903803663780403, + "learning_rate": 2.904225534989251e-07, + "loss": 0.8669, + "step": 9220 + }, + { + "epoch": 0.8315822699192857, + "grad_norm": 1.5536873058581528, + "learning_rate": 2.901194350590572e-07, + "loss": 0.8518, + "step": 9221 + }, + { + "epoch": 0.831672453442756, + "grad_norm": 1.632684145906424, + "learning_rate": 2.898164625156274e-07, + "loss": 0.8278, + "step": 9222 + }, + { + "epoch": 0.8317626369662263, + "grad_norm": 1.4222478692073488, + "learning_rate": 2.8951363589448676e-07, + "loss": 0.9022, + "step": 9223 + }, + { + "epoch": 0.8318528204896966, + "grad_norm": 1.4964230326937933, + "learning_rate": 2.8921095522147434e-07, + "loss": 0.9414, + "step": 9224 + }, + { + "epoch": 0.8319430040131668, + "grad_norm": 1.618653743047095, + "learning_rate": 2.8890842052241683e-07, + "loss": 0.9764, + "step": 9225 + }, + { + "epoch": 0.832033187536637, + "grad_norm": 0.6317458169569946, + "learning_rate": 2.886060318231267e-07, + "loss": 0.7277, + "step": 9226 + }, + { + "epoch": 0.8321233710601074, + "grad_norm": 1.6947903861908704, + "learning_rate": 2.8830378914940755e-07, + "loss": 0.9716, + "step": 9227 + }, + { + "epoch": 0.8322135545835776, + "grad_norm": 1.7554191062844497, + "learning_rate": 2.8800169252704675e-07, + "loss": 0.9325, + "step": 9228 + }, + { + "epoch": 0.8323037381070478, + "grad_norm": 1.4864913309449495, + "learning_rate": 2.8769974198182143e-07, + "loss": 0.9698, + "step": 9229 + }, + { + "epoch": 0.8323939216305181, + "grad_norm": 1.365994439684832, + "learning_rate": 2.873979375394955e-07, + "loss": 0.9432, + "step": 9230 + }, + { + "epoch": 0.8324841051539884, + "grad_norm": 1.6701743487186593, + "learning_rate": 2.870962792258209e-07, + "loss": 0.977, + "step": 9231 + }, + { + "epoch": 0.8325742886774586, + "grad_norm": 1.3673776548844399, + "learning_rate": 2.8679476706653716e-07, + "loss": 0.9587, + "step": 9232 + }, + { + "epoch": 0.8326644722009289, + "grad_norm": 1.8387538005859598, + "learning_rate": 2.864934010873692e-07, + "loss": 0.9674, + "step": 9233 + }, + { + "epoch": 0.8327546557243991, + "grad_norm": 1.3844868926905012, + "learning_rate": 2.8619218131403357e-07, + "loss": 0.9217, + "step": 9234 + }, + { + "epoch": 0.8328448392478695, + "grad_norm": 1.8137866090514723, + "learning_rate": 2.858911077722299e-07, + "loss": 0.9749, + "step": 9235 + }, + { + "epoch": 0.8329350227713397, + "grad_norm": 1.7065890020838912, + "learning_rate": 2.855901804876493e-07, + "loss": 0.7954, + "step": 9236 + }, + { + "epoch": 0.8330252062948099, + "grad_norm": 1.3816724916449614, + "learning_rate": 2.852893994859673e-07, + "loss": 0.9837, + "step": 9237 + }, + { + "epoch": 0.8331153898182801, + "grad_norm": 1.3230801979375855, + "learning_rate": 2.849887647928484e-07, + "loss": 0.8917, + "step": 9238 + }, + { + "epoch": 0.8332055733417505, + "grad_norm": 1.952498312545223, + "learning_rate": 2.8468827643394465e-07, + "loss": 0.8459, + "step": 9239 + }, + { + "epoch": 0.8332957568652207, + "grad_norm": 1.4285472220686062, + "learning_rate": 2.843879344348954e-07, + "loss": 0.9599, + "step": 9240 + }, + { + "epoch": 0.833385940388691, + "grad_norm": 1.4377126951012833, + "learning_rate": 2.840877388213272e-07, + "loss": 0.9166, + "step": 9241 + }, + { + "epoch": 0.8334761239121612, + "grad_norm": 1.6770760751255211, + "learning_rate": 2.8378768961885515e-07, + "loss": 0.9493, + "step": 9242 + }, + { + "epoch": 0.8335663074356315, + "grad_norm": 1.6647660465088674, + "learning_rate": 2.8348778685307983e-07, + "loss": 0.9812, + "step": 9243 + }, + { + "epoch": 0.8336564909591018, + "grad_norm": 1.5181315037857772, + "learning_rate": 2.831880305495915e-07, + "loss": 1.0203, + "step": 9244 + }, + { + "epoch": 0.833746674482572, + "grad_norm": 0.6734281154253058, + "learning_rate": 2.828884207339668e-07, + "loss": 0.8133, + "step": 9245 + }, + { + "epoch": 0.8338368580060423, + "grad_norm": 1.888717635033322, + "learning_rate": 2.8258895743177014e-07, + "loss": 1.0231, + "step": 9246 + }, + { + "epoch": 0.8339270415295126, + "grad_norm": 1.5618271714207246, + "learning_rate": 2.8228964066855356e-07, + "loss": 1.0217, + "step": 9247 + }, + { + "epoch": 0.8340172250529828, + "grad_norm": 1.436034077083282, + "learning_rate": 2.819904704698555e-07, + "loss": 0.9235, + "step": 9248 + }, + { + "epoch": 0.834107408576453, + "grad_norm": 0.6535953233714982, + "learning_rate": 2.8169144686120437e-07, + "loss": 0.7288, + "step": 9249 + }, + { + "epoch": 0.8341975920999234, + "grad_norm": 1.5932657590739454, + "learning_rate": 2.8139256986811254e-07, + "loss": 0.8571, + "step": 9250 + }, + { + "epoch": 0.8342877756233936, + "grad_norm": 1.6394213666332573, + "learning_rate": 2.8109383951608424e-07, + "loss": 0.9999, + "step": 9251 + }, + { + "epoch": 0.8343779591468639, + "grad_norm": 1.6063425326875855, + "learning_rate": 2.8079525583060683e-07, + "loss": 1.0133, + "step": 9252 + }, + { + "epoch": 0.8344681426703341, + "grad_norm": 1.3832113692892503, + "learning_rate": 2.804968188371577e-07, + "loss": 0.9542, + "step": 9253 + }, + { + "epoch": 0.8345583261938044, + "grad_norm": 1.6487589879691136, + "learning_rate": 2.801985285612014e-07, + "loss": 0.9484, + "step": 9254 + }, + { + "epoch": 0.8346485097172747, + "grad_norm": 1.5696970174699487, + "learning_rate": 2.7990038502818934e-07, + "loss": 0.919, + "step": 9255 + }, + { + "epoch": 0.8347386932407449, + "grad_norm": 1.561050505647675, + "learning_rate": 2.796023882635612e-07, + "loss": 0.9238, + "step": 9256 + }, + { + "epoch": 0.8348288767642151, + "grad_norm": 1.4416876624164283, + "learning_rate": 2.7930453829274323e-07, + "loss": 0.9973, + "step": 9257 + }, + { + "epoch": 0.8349190602876855, + "grad_norm": 1.740771278248012, + "learning_rate": 2.7900683514115054e-07, + "loss": 0.9897, + "step": 9258 + }, + { + "epoch": 0.8350092438111557, + "grad_norm": 1.6647756418503858, + "learning_rate": 2.787092788341836e-07, + "loss": 0.8829, + "step": 9259 + }, + { + "epoch": 0.8350994273346259, + "grad_norm": 1.503086887439521, + "learning_rate": 2.7841186939723195e-07, + "loss": 0.9432, + "step": 9260 + }, + { + "epoch": 0.8351896108580962, + "grad_norm": 1.529259224497619, + "learning_rate": 2.7811460685567255e-07, + "loss": 0.9231, + "step": 9261 + }, + { + "epoch": 0.8352797943815665, + "grad_norm": 1.5927213453314775, + "learning_rate": 2.778174912348692e-07, + "loss": 0.9403, + "step": 9262 + }, + { + "epoch": 0.8353699779050368, + "grad_norm": 5.612767630740166, + "learning_rate": 2.7752052256017354e-07, + "loss": 0.9099, + "step": 9263 + }, + { + "epoch": 0.835460161428507, + "grad_norm": 1.9480172469611268, + "learning_rate": 2.7722370085692493e-07, + "loss": 0.9131, + "step": 9264 + }, + { + "epoch": 0.8355503449519772, + "grad_norm": 1.645131198787612, + "learning_rate": 2.769270261504486e-07, + "loss": 0.9166, + "step": 9265 + }, + { + "epoch": 0.8356405284754476, + "grad_norm": 1.371794822974196, + "learning_rate": 2.7663049846606015e-07, + "loss": 0.9073, + "step": 9266 + }, + { + "epoch": 0.8357307119989178, + "grad_norm": 1.7521325469239923, + "learning_rate": 2.763341178290592e-07, + "loss": 0.9602, + "step": 9267 + }, + { + "epoch": 0.835820895522388, + "grad_norm": 1.3712502849373396, + "learning_rate": 2.7603788426473663e-07, + "loss": 0.9711, + "step": 9268 + }, + { + "epoch": 0.8359110790458584, + "grad_norm": 1.9321735136885383, + "learning_rate": 2.7574179779836695e-07, + "loss": 0.8841, + "step": 9269 + }, + { + "epoch": 0.8360012625693286, + "grad_norm": 1.5532733101906158, + "learning_rate": 2.754458584552146e-07, + "loss": 0.9038, + "step": 9270 + }, + { + "epoch": 0.8360914460927988, + "grad_norm": 1.9131385285809848, + "learning_rate": 2.751500662605308e-07, + "loss": 0.9399, + "step": 9271 + }, + { + "epoch": 0.8361816296162691, + "grad_norm": 1.6664035430471305, + "learning_rate": 2.7485442123955383e-07, + "loss": 0.9055, + "step": 9272 + }, + { + "epoch": 0.8362718131397394, + "grad_norm": 1.6466512538919618, + "learning_rate": 2.7455892341751075e-07, + "loss": 0.8308, + "step": 9273 + }, + { + "epoch": 0.8363619966632097, + "grad_norm": 1.9719717518909183, + "learning_rate": 2.7426357281961365e-07, + "loss": 0.822, + "step": 9274 + }, + { + "epoch": 0.8364521801866799, + "grad_norm": 1.6947240555337932, + "learning_rate": 2.7396836947106416e-07, + "loss": 0.9513, + "step": 9275 + }, + { + "epoch": 0.8365423637101501, + "grad_norm": 1.633956435455085, + "learning_rate": 2.736733133970506e-07, + "loss": 1.0462, + "step": 9276 + }, + { + "epoch": 0.8366325472336205, + "grad_norm": 1.2731816789597843, + "learning_rate": 2.7337840462274896e-07, + "loss": 0.9844, + "step": 9277 + }, + { + "epoch": 0.8367227307570907, + "grad_norm": 1.4689547619225969, + "learning_rate": 2.730836431733221e-07, + "loss": 0.9155, + "step": 9278 + }, + { + "epoch": 0.8368129142805609, + "grad_norm": 1.9158651776990658, + "learning_rate": 2.727890290739212e-07, + "loss": 0.9972, + "step": 9279 + }, + { + "epoch": 0.8369030978040312, + "grad_norm": 1.557472609742941, + "learning_rate": 2.7249456234968395e-07, + "loss": 0.982, + "step": 9280 + }, + { + "epoch": 0.8369932813275015, + "grad_norm": 1.544417770454233, + "learning_rate": 2.722002430257364e-07, + "loss": 1.0111, + "step": 9281 + }, + { + "epoch": 0.8370834648509717, + "grad_norm": 1.365093131119611, + "learning_rate": 2.7190607112719035e-07, + "loss": 0.9081, + "step": 9282 + }, + { + "epoch": 0.837173648374442, + "grad_norm": 1.8850205006989764, + "learning_rate": 2.716120466791476e-07, + "loss": 0.8658, + "step": 9283 + }, + { + "epoch": 0.8372638318979122, + "grad_norm": 1.7071869035021854, + "learning_rate": 2.7131816970669483e-07, + "loss": 0.9268, + "step": 9284 + }, + { + "epoch": 0.8373540154213825, + "grad_norm": 1.7925558175689653, + "learning_rate": 2.7102444023490777e-07, + "loss": 0.9585, + "step": 9285 + }, + { + "epoch": 0.8374441989448528, + "grad_norm": 1.3839341264604454, + "learning_rate": 2.70730858288849e-07, + "loss": 1.0014, + "step": 9286 + }, + { + "epoch": 0.837534382468323, + "grad_norm": 1.5713771418416604, + "learning_rate": 2.704374238935685e-07, + "loss": 0.8798, + "step": 9287 + }, + { + "epoch": 0.8376245659917932, + "grad_norm": 0.8976394093978086, + "learning_rate": 2.70144137074104e-07, + "loss": 0.7761, + "step": 9288 + }, + { + "epoch": 0.8377147495152636, + "grad_norm": 1.6550883411905164, + "learning_rate": 2.6985099785547926e-07, + "loss": 0.9655, + "step": 9289 + }, + { + "epoch": 0.8378049330387338, + "grad_norm": 1.512884828776537, + "learning_rate": 2.695580062627083e-07, + "loss": 0.9701, + "step": 9290 + }, + { + "epoch": 0.8378951165622041, + "grad_norm": 1.7957613603842513, + "learning_rate": 2.692651623207891e-07, + "loss": 0.9403, + "step": 9291 + }, + { + "epoch": 0.8379853000856744, + "grad_norm": 1.3029240551625094, + "learning_rate": 2.689724660547097e-07, + "loss": 0.9313, + "step": 9292 + }, + { + "epoch": 0.8380754836091446, + "grad_norm": 0.6336761044791561, + "learning_rate": 2.686799174894441e-07, + "loss": 0.7728, + "step": 9293 + }, + { + "epoch": 0.8381656671326149, + "grad_norm": 0.6733368561632076, + "learning_rate": 2.683875166499545e-07, + "loss": 0.7728, + "step": 9294 + }, + { + "epoch": 0.8382558506560851, + "grad_norm": 1.59948990856635, + "learning_rate": 2.680952635611899e-07, + "loss": 0.9529, + "step": 9295 + }, + { + "epoch": 0.8383460341795554, + "grad_norm": 1.6639661447629752, + "learning_rate": 2.678031582480875e-07, + "loss": 0.8691, + "step": 9296 + }, + { + "epoch": 0.8384362177030257, + "grad_norm": 1.460560714358201, + "learning_rate": 2.6751120073557e-07, + "loss": 0.9827, + "step": 9297 + }, + { + "epoch": 0.8385264012264959, + "grad_norm": 1.901146545304123, + "learning_rate": 2.672193910485505e-07, + "loss": 0.912, + "step": 9298 + }, + { + "epoch": 0.8386165847499661, + "grad_norm": 1.600511755163944, + "learning_rate": 2.669277292119265e-07, + "loss": 0.9462, + "step": 9299 + }, + { + "epoch": 0.8387067682734365, + "grad_norm": 1.6538542208193332, + "learning_rate": 2.666362152505848e-07, + "loss": 0.9476, + "step": 9300 + }, + { + "epoch": 0.8387969517969067, + "grad_norm": 1.6753132484518327, + "learning_rate": 2.663448491893989e-07, + "loss": 0.8799, + "step": 9301 + }, + { + "epoch": 0.838887135320377, + "grad_norm": 1.5188210207812882, + "learning_rate": 2.6605363105322974e-07, + "loss": 0.9219, + "step": 9302 + }, + { + "epoch": 0.8389773188438472, + "grad_norm": 1.605619771495249, + "learning_rate": 2.657625608669263e-07, + "loss": 0.9431, + "step": 9303 + }, + { + "epoch": 0.8390675023673175, + "grad_norm": 1.442866219191216, + "learning_rate": 2.654716386553224e-07, + "loss": 0.856, + "step": 9304 + }, + { + "epoch": 0.8391576858907878, + "grad_norm": 1.4146085074894996, + "learning_rate": 2.651808644432436e-07, + "loss": 0.9432, + "step": 9305 + }, + { + "epoch": 0.839247869414258, + "grad_norm": 1.709250467565656, + "learning_rate": 2.6489023825549807e-07, + "loss": 1.08, + "step": 9306 + }, + { + "epoch": 0.8393380529377282, + "grad_norm": 1.8330521656944143, + "learning_rate": 2.6459976011688547e-07, + "loss": 0.8652, + "step": 9307 + }, + { + "epoch": 0.8394282364611986, + "grad_norm": 1.6480594581183308, + "learning_rate": 2.6430943005219e-07, + "loss": 0.9404, + "step": 9308 + }, + { + "epoch": 0.8395184199846688, + "grad_norm": 1.3822448341161326, + "learning_rate": 2.6401924808618447e-07, + "loss": 0.9721, + "step": 9309 + }, + { + "epoch": 0.839608603508139, + "grad_norm": 1.621477270145239, + "learning_rate": 2.637292142436287e-07, + "loss": 0.9417, + "step": 9310 + }, + { + "epoch": 0.8396987870316093, + "grad_norm": 1.6342391940525982, + "learning_rate": 2.6343932854927e-07, + "loss": 0.9137, + "step": 9311 + }, + { + "epoch": 0.8397889705550796, + "grad_norm": 0.7025725207449399, + "learning_rate": 2.6314959102784316e-07, + "loss": 0.7705, + "step": 9312 + }, + { + "epoch": 0.8398791540785498, + "grad_norm": 1.714484244442488, + "learning_rate": 2.6286000170407074e-07, + "loss": 0.9408, + "step": 9313 + }, + { + "epoch": 0.8399693376020201, + "grad_norm": 1.7931269158306238, + "learning_rate": 2.625705606026607e-07, + "loss": 0.9225, + "step": 9314 + }, + { + "epoch": 0.8400595211254904, + "grad_norm": 1.473389137097327, + "learning_rate": 2.622812677483106e-07, + "loss": 1.0223, + "step": 9315 + }, + { + "epoch": 0.8401497046489607, + "grad_norm": 2.0569364474451137, + "learning_rate": 2.6199212316570453e-07, + "loss": 1.0032, + "step": 9316 + }, + { + "epoch": 0.8402398881724309, + "grad_norm": 1.5765235253347276, + "learning_rate": 2.617031268795138e-07, + "loss": 0.9184, + "step": 9317 + }, + { + "epoch": 0.8403300716959011, + "grad_norm": 1.3515951781098536, + "learning_rate": 2.614142789143976e-07, + "loss": 0.9742, + "step": 9318 + }, + { + "epoch": 0.8404202552193715, + "grad_norm": 1.4704604233568939, + "learning_rate": 2.6112557929500047e-07, + "loss": 0.9267, + "step": 9319 + }, + { + "epoch": 0.8405104387428417, + "grad_norm": 1.5228707848911462, + "learning_rate": 2.6083702804595817e-07, + "loss": 0.9705, + "step": 9320 + }, + { + "epoch": 0.8406006222663119, + "grad_norm": 1.477558990566, + "learning_rate": 2.6054862519188915e-07, + "loss": 0.9298, + "step": 9321 + }, + { + "epoch": 0.8406908057897822, + "grad_norm": 1.656196377443982, + "learning_rate": 2.6026037075740357e-07, + "loss": 0.8794, + "step": 9322 + }, + { + "epoch": 0.8407809893132525, + "grad_norm": 1.7518737162579539, + "learning_rate": 2.5997226476709524e-07, + "loss": 0.9099, + "step": 9323 + }, + { + "epoch": 0.8408711728367227, + "grad_norm": 1.608056278116941, + "learning_rate": 2.5968430724554856e-07, + "loss": 0.945, + "step": 9324 + }, + { + "epoch": 0.840961356360193, + "grad_norm": 1.7762624951969406, + "learning_rate": 2.5939649821733225e-07, + "loss": 0.879, + "step": 9325 + }, + { + "epoch": 0.8410515398836632, + "grad_norm": 1.8148760513143924, + "learning_rate": 2.5910883770700433e-07, + "loss": 0.9112, + "step": 9326 + }, + { + "epoch": 0.8411417234071336, + "grad_norm": 1.3083317280818219, + "learning_rate": 2.5882132573910965e-07, + "loss": 0.7834, + "step": 9327 + }, + { + "epoch": 0.8412319069306038, + "grad_norm": 1.2589210222089418, + "learning_rate": 2.585339623381801e-07, + "loss": 0.9822, + "step": 9328 + }, + { + "epoch": 0.841322090454074, + "grad_norm": 1.6715676568865145, + "learning_rate": 2.582467475287358e-07, + "loss": 0.9187, + "step": 9329 + }, + { + "epoch": 0.8414122739775443, + "grad_norm": 1.6066324513434045, + "learning_rate": 2.5795968133528224e-07, + "loss": 0.956, + "step": 9330 + }, + { + "epoch": 0.8415024575010146, + "grad_norm": 1.7531293409977171, + "learning_rate": 2.576727637823144e-07, + "loss": 0.9464, + "step": 9331 + }, + { + "epoch": 0.8415926410244848, + "grad_norm": 2.202646731144806, + "learning_rate": 2.5738599489431335e-07, + "loss": 0.9892, + "step": 9332 + }, + { + "epoch": 0.8416828245479551, + "grad_norm": 1.353234287984712, + "learning_rate": 2.5709937469574794e-07, + "loss": 1.0035, + "step": 9333 + }, + { + "epoch": 0.8417730080714253, + "grad_norm": 1.322605151447259, + "learning_rate": 2.568129032110742e-07, + "loss": 0.9847, + "step": 9334 + }, + { + "epoch": 0.8418631915948956, + "grad_norm": 2.104331535703964, + "learning_rate": 2.5652658046473565e-07, + "loss": 1.0182, + "step": 9335 + }, + { + "epoch": 0.8419533751183659, + "grad_norm": 1.313683385058435, + "learning_rate": 2.5624040648116184e-07, + "loss": 0.9422, + "step": 9336 + }, + { + "epoch": 0.8420435586418361, + "grad_norm": 1.5392080325337383, + "learning_rate": 2.5595438128477245e-07, + "loss": 0.8847, + "step": 9337 + }, + { + "epoch": 0.8421337421653065, + "grad_norm": 1.6775182156206274, + "learning_rate": 2.5566850489997096e-07, + "loss": 0.9517, + "step": 9338 + }, + { + "epoch": 0.8422239256887767, + "grad_norm": 2.274029212321159, + "learning_rate": 2.5538277735115166e-07, + "loss": 0.974, + "step": 9339 + }, + { + "epoch": 0.8423141092122469, + "grad_norm": 1.3631893443960479, + "learning_rate": 2.5509719866269306e-07, + "loss": 1.0003, + "step": 9340 + }, + { + "epoch": 0.8424042927357172, + "grad_norm": 1.5558444371752624, + "learning_rate": 2.548117688589628e-07, + "loss": 1.0064, + "step": 9341 + }, + { + "epoch": 0.8424944762591875, + "grad_norm": 1.461608324717057, + "learning_rate": 2.545264879643152e-07, + "loss": 0.9304, + "step": 9342 + }, + { + "epoch": 0.8425846597826577, + "grad_norm": 1.4954118494335498, + "learning_rate": 2.542413560030923e-07, + "loss": 0.9004, + "step": 9343 + }, + { + "epoch": 0.842674843306128, + "grad_norm": 1.7450669787868518, + "learning_rate": 2.53956372999623e-07, + "loss": 0.837, + "step": 9344 + }, + { + "epoch": 0.8427650268295982, + "grad_norm": 0.6809699576609047, + "learning_rate": 2.5367153897822293e-07, + "loss": 0.8106, + "step": 9345 + }, + { + "epoch": 0.8428552103530685, + "grad_norm": 1.3548505181209611, + "learning_rate": 2.5338685396319715e-07, + "loss": 0.9362, + "step": 9346 + }, + { + "epoch": 0.8429453938765388, + "grad_norm": 1.6237852251021783, + "learning_rate": 2.531023179788352e-07, + "loss": 0.8806, + "step": 9347 + }, + { + "epoch": 0.843035577400009, + "grad_norm": 1.5252159497309221, + "learning_rate": 2.528179310494158e-07, + "loss": 0.9151, + "step": 9348 + }, + { + "epoch": 0.8431257609234792, + "grad_norm": 1.5848134384107364, + "learning_rate": 2.5253369319920436e-07, + "loss": 0.9625, + "step": 9349 + }, + { + "epoch": 0.8432159444469496, + "grad_norm": 1.8798940365939276, + "learning_rate": 2.522496044524538e-07, + "loss": 0.9465, + "step": 9350 + }, + { + "epoch": 0.8433061279704198, + "grad_norm": 1.6832459128061221, + "learning_rate": 2.5196566483340386e-07, + "loss": 1.0347, + "step": 9351 + }, + { + "epoch": 0.84339631149389, + "grad_norm": 1.6140471850072065, + "learning_rate": 2.516818743662825e-07, + "loss": 0.9603, + "step": 9352 + }, + { + "epoch": 0.8434864950173603, + "grad_norm": 1.6433010256894822, + "learning_rate": 2.5139823307530285e-07, + "loss": 0.9468, + "step": 9353 + }, + { + "epoch": 0.8435766785408306, + "grad_norm": 1.708946079370035, + "learning_rate": 2.5111474098466836e-07, + "loss": 0.9791, + "step": 9354 + }, + { + "epoch": 0.8436668620643009, + "grad_norm": 1.3605362327963368, + "learning_rate": 2.50831398118567e-07, + "loss": 0.991, + "step": 9355 + }, + { + "epoch": 0.8437570455877711, + "grad_norm": 1.7182509651508837, + "learning_rate": 2.5054820450117576e-07, + "loss": 0.8911, + "step": 9356 + }, + { + "epoch": 0.8438472291112413, + "grad_norm": 1.4249967742348015, + "learning_rate": 2.502651601566579e-07, + "loss": 0.9072, + "step": 9357 + }, + { + "epoch": 0.8439374126347117, + "grad_norm": 1.8315499981120107, + "learning_rate": 2.499822651091645e-07, + "loss": 0.8956, + "step": 9358 + }, + { + "epoch": 0.8440275961581819, + "grad_norm": 1.7499478877346568, + "learning_rate": 2.496995193828344e-07, + "loss": 0.965, + "step": 9359 + }, + { + "epoch": 0.8441177796816521, + "grad_norm": 1.726103294375868, + "learning_rate": 2.494169230017913e-07, + "loss": 0.9423, + "step": 9360 + }, + { + "epoch": 0.8442079632051224, + "grad_norm": 1.479605229154052, + "learning_rate": 2.491344759901499e-07, + "loss": 0.9047, + "step": 9361 + }, + { + "epoch": 0.8442981467285927, + "grad_norm": 1.8076892016145334, + "learning_rate": 2.488521783720088e-07, + "loss": 0.9383, + "step": 9362 + }, + { + "epoch": 0.844388330252063, + "grad_norm": 1.6071800893679105, + "learning_rate": 2.4857003017145526e-07, + "loss": 0.977, + "step": 9363 + }, + { + "epoch": 0.8444785137755332, + "grad_norm": 1.6856808924080688, + "learning_rate": 2.482880314125644e-07, + "loss": 0.9817, + "step": 9364 + }, + { + "epoch": 0.8445686972990035, + "grad_norm": 1.7133841244087324, + "learning_rate": 2.4800618211939726e-07, + "loss": 0.9364, + "step": 9365 + }, + { + "epoch": 0.8446588808224738, + "grad_norm": 1.6344203789503402, + "learning_rate": 2.477244823160034e-07, + "loss": 0.9486, + "step": 9366 + }, + { + "epoch": 0.844749064345944, + "grad_norm": 1.6951507631843006, + "learning_rate": 2.474429320264184e-07, + "loss": 0.965, + "step": 9367 + }, + { + "epoch": 0.8448392478694142, + "grad_norm": 1.4814343687796179, + "learning_rate": 2.47161531274666e-07, + "loss": 0.9021, + "step": 9368 + }, + { + "epoch": 0.8449294313928846, + "grad_norm": 1.4211407117334958, + "learning_rate": 2.4688028008475714e-07, + "loss": 0.8762, + "step": 9369 + }, + { + "epoch": 0.8450196149163548, + "grad_norm": 2.1596591922548045, + "learning_rate": 2.465991784806891e-07, + "loss": 0.9343, + "step": 9370 + }, + { + "epoch": 0.845109798439825, + "grad_norm": 1.5751954638580987, + "learning_rate": 2.463182264864472e-07, + "loss": 1.022, + "step": 9371 + }, + { + "epoch": 0.8451999819632953, + "grad_norm": 1.6705491700321902, + "learning_rate": 2.460374241260039e-07, + "loss": 0.8685, + "step": 9372 + }, + { + "epoch": 0.8452901654867656, + "grad_norm": 1.5833173633907769, + "learning_rate": 2.4575677142331884e-07, + "loss": 1.01, + "step": 9373 + }, + { + "epoch": 0.8453803490102358, + "grad_norm": 1.5023785647099166, + "learning_rate": 2.454762684023395e-07, + "loss": 0.8869, + "step": 9374 + }, + { + "epoch": 0.8454705325337061, + "grad_norm": 1.5905694828200196, + "learning_rate": 2.4519591508699823e-07, + "loss": 0.8915, + "step": 9375 + }, + { + "epoch": 0.8455607160571763, + "grad_norm": 1.5373313191001734, + "learning_rate": 2.4491571150121815e-07, + "loss": 0.9912, + "step": 9376 + }, + { + "epoch": 0.8456508995806467, + "grad_norm": 1.782081744437334, + "learning_rate": 2.446356576689062e-07, + "loss": 0.9387, + "step": 9377 + }, + { + "epoch": 0.8457410831041169, + "grad_norm": 1.5682192843187162, + "learning_rate": 2.4435575361395976e-07, + "loss": 0.9541, + "step": 9378 + }, + { + "epoch": 0.8458312666275871, + "grad_norm": 1.4980610395455674, + "learning_rate": 2.440759993602606e-07, + "loss": 0.9238, + "step": 9379 + }, + { + "epoch": 0.8459214501510574, + "grad_norm": 1.655521736539212, + "learning_rate": 2.437963949316793e-07, + "loss": 0.8862, + "step": 9380 + }, + { + "epoch": 0.8460116336745277, + "grad_norm": 1.6439798114678648, + "learning_rate": 2.435169403520729e-07, + "loss": 0.9366, + "step": 9381 + }, + { + "epoch": 0.8461018171979979, + "grad_norm": 1.6109527705566768, + "learning_rate": 2.4323763564528653e-07, + "loss": 0.9899, + "step": 9382 + }, + { + "epoch": 0.8461920007214682, + "grad_norm": 1.5606723777903322, + "learning_rate": 2.429584808351517e-07, + "loss": 0.9596, + "step": 9383 + }, + { + "epoch": 0.8462821842449384, + "grad_norm": 2.5152361077570617, + "learning_rate": 2.42679475945488e-07, + "loss": 1.0666, + "step": 9384 + }, + { + "epoch": 0.8463723677684087, + "grad_norm": 1.4341265860141232, + "learning_rate": 2.424006210001008e-07, + "loss": 0.9433, + "step": 9385 + }, + { + "epoch": 0.846462551291879, + "grad_norm": 1.5249802040941476, + "learning_rate": 2.421219160227839e-07, + "loss": 0.9725, + "step": 9386 + }, + { + "epoch": 0.8465527348153492, + "grad_norm": 1.4377073884717755, + "learning_rate": 2.4184336103731785e-07, + "loss": 0.9376, + "step": 9387 + }, + { + "epoch": 0.8466429183388196, + "grad_norm": 1.4923049321326822, + "learning_rate": 2.4156495606747065e-07, + "loss": 0.9637, + "step": 9388 + }, + { + "epoch": 0.8467331018622898, + "grad_norm": 1.550477351240054, + "learning_rate": 2.412867011369972e-07, + "loss": 1.0567, + "step": 9389 + }, + { + "epoch": 0.84682328538576, + "grad_norm": 1.6019032929861765, + "learning_rate": 2.4100859626963997e-07, + "loss": 0.955, + "step": 9390 + }, + { + "epoch": 0.8469134689092302, + "grad_norm": 1.4722916128891685, + "learning_rate": 2.407306414891288e-07, + "loss": 0.9395, + "step": 9391 + }, + { + "epoch": 0.8470036524327006, + "grad_norm": 1.2552182472861073, + "learning_rate": 2.4045283681917893e-07, + "loss": 1.0452, + "step": 9392 + }, + { + "epoch": 0.8470938359561708, + "grad_norm": 1.6431450517489428, + "learning_rate": 2.4017518228349586e-07, + "loss": 1.0561, + "step": 9393 + }, + { + "epoch": 0.8471840194796411, + "grad_norm": 0.6427735144662362, + "learning_rate": 2.3989767790576887e-07, + "loss": 0.7926, + "step": 9394 + }, + { + "epoch": 0.8472742030031113, + "grad_norm": 1.6321599651178185, + "learning_rate": 2.396203237096781e-07, + "loss": 0.9078, + "step": 9395 + }, + { + "epoch": 0.8473643865265816, + "grad_norm": 1.6641208119092912, + "learning_rate": 2.393431197188873e-07, + "loss": 0.9355, + "step": 9396 + }, + { + "epoch": 0.8474545700500519, + "grad_norm": 1.9789649325170167, + "learning_rate": 2.3906606595705004e-07, + "loss": 1.0167, + "step": 9397 + }, + { + "epoch": 0.8475447535735221, + "grad_norm": 1.486337027604539, + "learning_rate": 2.387891624478056e-07, + "loss": 0.9979, + "step": 9398 + }, + { + "epoch": 0.8476349370969923, + "grad_norm": 1.4603083270597192, + "learning_rate": 2.3851240921478075e-07, + "loss": 0.963, + "step": 9399 + }, + { + "epoch": 0.8477251206204627, + "grad_norm": 1.5104470436286224, + "learning_rate": 2.3823580628159057e-07, + "loss": 0.9244, + "step": 9400 + }, + { + "epoch": 0.8478153041439329, + "grad_norm": 1.7810395769008909, + "learning_rate": 2.3795935367183517e-07, + "loss": 0.9318, + "step": 9401 + }, + { + "epoch": 0.8479054876674031, + "grad_norm": 1.3897365346449408, + "learning_rate": 2.376830514091035e-07, + "loss": 0.9628, + "step": 9402 + }, + { + "epoch": 0.8479956711908734, + "grad_norm": 1.536724813165228, + "learning_rate": 2.3740689951697135e-07, + "loss": 1.0236, + "step": 9403 + }, + { + "epoch": 0.8480858547143437, + "grad_norm": 1.5552736643793788, + "learning_rate": 2.371308980190012e-07, + "loss": 0.9445, + "step": 9404 + }, + { + "epoch": 0.848176038237814, + "grad_norm": 0.6753261113946467, + "learning_rate": 2.3685504693874337e-07, + "loss": 0.7985, + "step": 9405 + }, + { + "epoch": 0.8482662217612842, + "grad_norm": 1.830865159462039, + "learning_rate": 2.3657934629973497e-07, + "loss": 0.9797, + "step": 9406 + }, + { + "epoch": 0.8483564052847544, + "grad_norm": 1.6362902824741359, + "learning_rate": 2.3630379612549944e-07, + "loss": 0.9525, + "step": 9407 + }, + { + "epoch": 0.8484465888082248, + "grad_norm": 0.75982324533724, + "learning_rate": 2.3602839643954997e-07, + "loss": 0.8224, + "step": 9408 + }, + { + "epoch": 0.848536772331695, + "grad_norm": 1.3968106402221299, + "learning_rate": 2.3575314726538308e-07, + "loss": 0.9516, + "step": 9409 + }, + { + "epoch": 0.8486269558551652, + "grad_norm": 1.4496075296351685, + "learning_rate": 2.3547804862648645e-07, + "loss": 0.9769, + "step": 9410 + }, + { + "epoch": 0.8487171393786356, + "grad_norm": 1.3713665117615326, + "learning_rate": 2.3520310054633174e-07, + "loss": 0.9908, + "step": 9411 + }, + { + "epoch": 0.8488073229021058, + "grad_norm": 1.6381996756615618, + "learning_rate": 2.3492830304837973e-07, + "loss": 0.9069, + "step": 9412 + }, + { + "epoch": 0.848897506425576, + "grad_norm": 1.481514835423718, + "learning_rate": 2.3465365615607723e-07, + "loss": 0.8884, + "step": 9413 + }, + { + "epoch": 0.8489876899490463, + "grad_norm": 1.5544053018068216, + "learning_rate": 2.3437915989285884e-07, + "loss": 0.9371, + "step": 9414 + }, + { + "epoch": 0.8490778734725166, + "grad_norm": 1.662291394676833, + "learning_rate": 2.3410481428214647e-07, + "loss": 1.0078, + "step": 9415 + }, + { + "epoch": 0.8491680569959869, + "grad_norm": 1.5484033605851975, + "learning_rate": 2.338306193473476e-07, + "loss": 0.9687, + "step": 9416 + }, + { + "epoch": 0.8492582405194571, + "grad_norm": 1.4054592876529288, + "learning_rate": 2.3355657511185957e-07, + "loss": 0.9452, + "step": 9417 + }, + { + "epoch": 0.8493484240429273, + "grad_norm": 0.6657308825599723, + "learning_rate": 2.3328268159906428e-07, + "loss": 0.7638, + "step": 9418 + }, + { + "epoch": 0.8494386075663977, + "grad_norm": 1.7255699791680374, + "learning_rate": 2.330089388323322e-07, + "loss": 0.9928, + "step": 9419 + }, + { + "epoch": 0.8495287910898679, + "grad_norm": 2.030654232980622, + "learning_rate": 2.327353468350204e-07, + "loss": 0.9836, + "step": 9420 + }, + { + "epoch": 0.8496189746133381, + "grad_norm": 1.516602191323115, + "learning_rate": 2.3246190563047352e-07, + "loss": 0.9373, + "step": 9421 + }, + { + "epoch": 0.8497091581368084, + "grad_norm": 1.5791552362892265, + "learning_rate": 2.3218861524202293e-07, + "loss": 0.9265, + "step": 9422 + }, + { + "epoch": 0.8497993416602787, + "grad_norm": 1.9710714791786457, + "learning_rate": 2.3191547569298775e-07, + "loss": 0.9538, + "step": 9423 + }, + { + "epoch": 0.8498895251837489, + "grad_norm": 2.0754057797683325, + "learning_rate": 2.3164248700667245e-07, + "loss": 0.9311, + "step": 9424 + }, + { + "epoch": 0.8499797087072192, + "grad_norm": 1.4758047753271908, + "learning_rate": 2.313696492063717e-07, + "loss": 0.9764, + "step": 9425 + }, + { + "epoch": 0.8500698922306894, + "grad_norm": 1.841379031278369, + "learning_rate": 2.3109696231536401e-07, + "loss": 0.9311, + "step": 9426 + }, + { + "epoch": 0.8501600757541597, + "grad_norm": 1.6041108489130624, + "learning_rate": 2.3082442635691722e-07, + "loss": 0.9039, + "step": 9427 + }, + { + "epoch": 0.85025025927763, + "grad_norm": 1.5979555777628545, + "learning_rate": 2.305520413542854e-07, + "loss": 0.958, + "step": 9428 + }, + { + "epoch": 0.8503404428011002, + "grad_norm": 1.3677350065134317, + "learning_rate": 2.3027980733071018e-07, + "loss": 0.8663, + "step": 9429 + }, + { + "epoch": 0.8504306263245704, + "grad_norm": 0.6163674713430024, + "learning_rate": 2.3000772430942027e-07, + "loss": 0.7461, + "step": 9430 + }, + { + "epoch": 0.8505208098480408, + "grad_norm": 1.484126421543219, + "learning_rate": 2.2973579231363028e-07, + "loss": 0.9597, + "step": 9431 + }, + { + "epoch": 0.850610993371511, + "grad_norm": 1.5780916116034065, + "learning_rate": 2.2946401136654446e-07, + "loss": 0.9717, + "step": 9432 + }, + { + "epoch": 0.8507011768949813, + "grad_norm": 1.3846750162500716, + "learning_rate": 2.2919238149135077e-07, + "loss": 0.9728, + "step": 9433 + }, + { + "epoch": 0.8507913604184516, + "grad_norm": 1.447936032758045, + "learning_rate": 2.289209027112282e-07, + "loss": 0.9518, + "step": 9434 + }, + { + "epoch": 0.8508815439419218, + "grad_norm": 1.8266907633052927, + "learning_rate": 2.2864957504933934e-07, + "loss": 1.0052, + "step": 9435 + }, + { + "epoch": 0.8509717274653921, + "grad_norm": 1.593778871760927, + "learning_rate": 2.2837839852883589e-07, + "loss": 0.93, + "step": 9436 + }, + { + "epoch": 0.8510619109888623, + "grad_norm": 1.3888365142271903, + "learning_rate": 2.2810737317285623e-07, + "loss": 0.9436, + "step": 9437 + }, + { + "epoch": 0.8511520945123326, + "grad_norm": 1.3661069687281555, + "learning_rate": 2.278364990045254e-07, + "loss": 0.8836, + "step": 9438 + }, + { + "epoch": 0.8512422780358029, + "grad_norm": 1.6127995471548777, + "learning_rate": 2.2756577604695625e-07, + "loss": 0.9405, + "step": 9439 + }, + { + "epoch": 0.8513324615592731, + "grad_norm": 1.447320975662205, + "learning_rate": 2.2729520432324855e-07, + "loss": 0.9654, + "step": 9440 + }, + { + "epoch": 0.8514226450827433, + "grad_norm": 1.4924524681151925, + "learning_rate": 2.2702478385648826e-07, + "loss": 0.8579, + "step": 9441 + }, + { + "epoch": 0.8515128286062137, + "grad_norm": 1.7105177969855565, + "learning_rate": 2.2675451466974938e-07, + "loss": 0.9387, + "step": 9442 + }, + { + "epoch": 0.8516030121296839, + "grad_norm": 1.719036841298977, + "learning_rate": 2.26484396786093e-07, + "loss": 0.8775, + "step": 9443 + }, + { + "epoch": 0.8516931956531542, + "grad_norm": 0.6702989234527543, + "learning_rate": 2.2621443022856667e-07, + "loss": 0.803, + "step": 9444 + }, + { + "epoch": 0.8517833791766244, + "grad_norm": 1.518713017394093, + "learning_rate": 2.2594461502020646e-07, + "loss": 0.9814, + "step": 9445 + }, + { + "epoch": 0.8518735627000947, + "grad_norm": 1.812205652473946, + "learning_rate": 2.2567495118403278e-07, + "loss": 0.824, + "step": 9446 + }, + { + "epoch": 0.851963746223565, + "grad_norm": 1.4298368282190177, + "learning_rate": 2.254054387430566e-07, + "loss": 0.9589, + "step": 9447 + }, + { + "epoch": 0.8520539297470352, + "grad_norm": 1.6359970213109813, + "learning_rate": 2.2513607772027243e-07, + "loss": 0.9268, + "step": 9448 + }, + { + "epoch": 0.8521441132705054, + "grad_norm": 1.558550917722859, + "learning_rate": 2.2486686813866562e-07, + "loss": 0.9078, + "step": 9449 + }, + { + "epoch": 0.8522342967939758, + "grad_norm": 1.6095009911857936, + "learning_rate": 2.2459781002120514e-07, + "loss": 0.9946, + "step": 9450 + }, + { + "epoch": 0.852324480317446, + "grad_norm": 1.5945999272764528, + "learning_rate": 2.243289033908491e-07, + "loss": 1.0133, + "step": 9451 + }, + { + "epoch": 0.8524146638409162, + "grad_norm": 1.6702008287660033, + "learning_rate": 2.2406014827054176e-07, + "loss": 0.9536, + "step": 9452 + }, + { + "epoch": 0.8525048473643865, + "grad_norm": 2.1595642492722043, + "learning_rate": 2.2379154468321525e-07, + "loss": 1.0075, + "step": 9453 + }, + { + "epoch": 0.8525950308878568, + "grad_norm": 1.6053961298870136, + "learning_rate": 2.2352309265178793e-07, + "loss": 0.9794, + "step": 9454 + }, + { + "epoch": 0.852685214411327, + "grad_norm": 1.8979358402739506, + "learning_rate": 2.2325479219916565e-07, + "loss": 1.0146, + "step": 9455 + }, + { + "epoch": 0.8527753979347973, + "grad_norm": 1.517504205462239, + "learning_rate": 2.229866433482419e-07, + "loss": 0.9757, + "step": 9456 + }, + { + "epoch": 0.8528655814582676, + "grad_norm": 1.5458544823090088, + "learning_rate": 2.2271864612189552e-07, + "loss": 0.9933, + "step": 9457 + }, + { + "epoch": 0.8529557649817379, + "grad_norm": 1.71234964420708, + "learning_rate": 2.2245080054299415e-07, + "loss": 0.9772, + "step": 9458 + }, + { + "epoch": 0.8530459485052081, + "grad_norm": 1.5960362809151798, + "learning_rate": 2.2218310663439198e-07, + "loss": 0.8983, + "step": 9459 + }, + { + "epoch": 0.8531361320286783, + "grad_norm": 1.5787104663142622, + "learning_rate": 2.2191556441892968e-07, + "loss": 0.9685, + "step": 9460 + }, + { + "epoch": 0.8532263155521487, + "grad_norm": 1.4742264610456786, + "learning_rate": 2.216481739194358e-07, + "loss": 0.9571, + "step": 9461 + }, + { + "epoch": 0.8533164990756189, + "grad_norm": 1.6046412950677769, + "learning_rate": 2.2138093515872592e-07, + "loss": 0.8985, + "step": 9462 + }, + { + "epoch": 0.8534066825990891, + "grad_norm": 1.3356491134530368, + "learning_rate": 2.2111384815960087e-07, + "loss": 1.0238, + "step": 9463 + }, + { + "epoch": 0.8534968661225594, + "grad_norm": 1.4426622165328074, + "learning_rate": 2.208469129448518e-07, + "loss": 0.8674, + "step": 9464 + }, + { + "epoch": 0.8535870496460297, + "grad_norm": 0.641584700968148, + "learning_rate": 2.2058012953725357e-07, + "loss": 0.7701, + "step": 9465 + }, + { + "epoch": 0.8536772331695, + "grad_norm": 1.58315276906583, + "learning_rate": 2.203134979595711e-07, + "loss": 0.9151, + "step": 9466 + }, + { + "epoch": 0.8537674166929702, + "grad_norm": 1.7269984596788952, + "learning_rate": 2.2004701823455374e-07, + "loss": 0.898, + "step": 9467 + }, + { + "epoch": 0.8538576002164404, + "grad_norm": 1.6113154740270415, + "learning_rate": 2.1978069038493906e-07, + "loss": 0.9747, + "step": 9468 + }, + { + "epoch": 0.8539477837399108, + "grad_norm": 1.3179074565426132, + "learning_rate": 2.1951451443345225e-07, + "loss": 0.9468, + "step": 9469 + }, + { + "epoch": 0.854037967263381, + "grad_norm": 1.553688303304847, + "learning_rate": 2.1924849040280425e-07, + "loss": 0.932, + "step": 9470 + }, + { + "epoch": 0.8541281507868512, + "grad_norm": 1.7828785328250008, + "learning_rate": 2.1898261831569465e-07, + "loss": 1.0292, + "step": 9471 + }, + { + "epoch": 0.8542183343103215, + "grad_norm": 1.5666635354328653, + "learning_rate": 2.1871689819480798e-07, + "loss": 0.9555, + "step": 9472 + }, + { + "epoch": 0.8543085178337918, + "grad_norm": 1.4558648971399666, + "learning_rate": 2.1845133006281745e-07, + "loss": 0.9299, + "step": 9473 + }, + { + "epoch": 0.854398701357262, + "grad_norm": 1.412307493206788, + "learning_rate": 2.1818591394238294e-07, + "loss": 0.8037, + "step": 9474 + }, + { + "epoch": 0.8544888848807323, + "grad_norm": 1.8861893252655868, + "learning_rate": 2.1792064985615076e-07, + "loss": 0.9265, + "step": 9475 + }, + { + "epoch": 0.8545790684042025, + "grad_norm": 1.4195137119416998, + "learning_rate": 2.1765553782675528e-07, + "loss": 0.9193, + "step": 9476 + }, + { + "epoch": 0.8546692519276728, + "grad_norm": 1.429759539680919, + "learning_rate": 2.1739057787681703e-07, + "loss": 1.018, + "step": 9477 + }, + { + "epoch": 0.8547594354511431, + "grad_norm": 1.6654691526636334, + "learning_rate": 2.1712577002894372e-07, + "loss": 0.9733, + "step": 9478 + }, + { + "epoch": 0.8548496189746133, + "grad_norm": 1.507086544495736, + "learning_rate": 2.1686111430573105e-07, + "loss": 0.9203, + "step": 9479 + }, + { + "epoch": 0.8549398024980835, + "grad_norm": 1.5052037574496118, + "learning_rate": 2.165966107297592e-07, + "loss": 0.9992, + "step": 9480 + }, + { + "epoch": 0.8550299860215539, + "grad_norm": 1.6710858487962972, + "learning_rate": 2.16332259323599e-07, + "loss": 0.9769, + "step": 9481 + }, + { + "epoch": 0.8551201695450241, + "grad_norm": 1.6371707701770288, + "learning_rate": 2.1606806010980504e-07, + "loss": 0.8955, + "step": 9482 + }, + { + "epoch": 0.8552103530684944, + "grad_norm": 1.697780728505407, + "learning_rate": 2.1580401311092067e-07, + "loss": 0.943, + "step": 9483 + }, + { + "epoch": 0.8553005365919647, + "grad_norm": 1.5177744931070194, + "learning_rate": 2.1554011834947604e-07, + "loss": 0.9347, + "step": 9484 + }, + { + "epoch": 0.8553907201154349, + "grad_norm": 1.5576697648600788, + "learning_rate": 2.1527637584798764e-07, + "loss": 0.9717, + "step": 9485 + }, + { + "epoch": 0.8554809036389052, + "grad_norm": 1.8128224447786758, + "learning_rate": 2.150127856289603e-07, + "loss": 1.0105, + "step": 9486 + }, + { + "epoch": 0.8555710871623754, + "grad_norm": 0.6367398592018221, + "learning_rate": 2.1474934771488363e-07, + "loss": 0.7582, + "step": 9487 + }, + { + "epoch": 0.8556612706858457, + "grad_norm": 1.50580514693924, + "learning_rate": 2.1448606212823715e-07, + "loss": 0.987, + "step": 9488 + }, + { + "epoch": 0.855751454209316, + "grad_norm": 1.579940884244765, + "learning_rate": 2.1422292889148452e-07, + "loss": 0.927, + "step": 9489 + }, + { + "epoch": 0.8558416377327862, + "grad_norm": 1.4388580541526672, + "learning_rate": 2.139599480270784e-07, + "loss": 0.9631, + "step": 9490 + }, + { + "epoch": 0.8559318212562564, + "grad_norm": 1.780126116710286, + "learning_rate": 2.1369711955745773e-07, + "loss": 0.933, + "step": 9491 + }, + { + "epoch": 0.8560220047797268, + "grad_norm": 1.7659964381751219, + "learning_rate": 2.1343444350504813e-07, + "loss": 0.9014, + "step": 9492 + }, + { + "epoch": 0.856112188303197, + "grad_norm": 1.5220509797479789, + "learning_rate": 2.1317191989226302e-07, + "loss": 0.9507, + "step": 9493 + }, + { + "epoch": 0.8562023718266673, + "grad_norm": 1.7442990494871422, + "learning_rate": 2.129095487415027e-07, + "loss": 0.9298, + "step": 9494 + }, + { + "epoch": 0.8562925553501375, + "grad_norm": 1.8060652192845097, + "learning_rate": 2.1264733007515257e-07, + "loss": 0.9565, + "step": 9495 + }, + { + "epoch": 0.8563827388736078, + "grad_norm": 1.6186074919549505, + "learning_rate": 2.1238526391558852e-07, + "loss": 0.9308, + "step": 9496 + }, + { + "epoch": 0.8564729223970781, + "grad_norm": 1.7602695501428627, + "learning_rate": 2.1212335028517003e-07, + "loss": 1.0829, + "step": 9497 + }, + { + "epoch": 0.8565631059205483, + "grad_norm": 1.4882600237116164, + "learning_rate": 2.1186158920624563e-07, + "loss": 0.955, + "step": 9498 + }, + { + "epoch": 0.8566532894440185, + "grad_norm": 1.4977938800229829, + "learning_rate": 2.1159998070115015e-07, + "loss": 0.9472, + "step": 9499 + }, + { + "epoch": 0.8567434729674889, + "grad_norm": 1.499371158392987, + "learning_rate": 2.113385247922055e-07, + "loss": 0.9261, + "step": 9500 + }, + { + "epoch": 0.8568336564909591, + "grad_norm": 1.5830502842659953, + "learning_rate": 2.1107722150172068e-07, + "loss": 0.9527, + "step": 9501 + }, + { + "epoch": 0.8569238400144293, + "grad_norm": 1.4443841375493567, + "learning_rate": 2.108160708519906e-07, + "loss": 0.985, + "step": 9502 + }, + { + "epoch": 0.8570140235378996, + "grad_norm": 1.6646827652142222, + "learning_rate": 2.1055507286529984e-07, + "loss": 0.9251, + "step": 9503 + }, + { + "epoch": 0.8571042070613699, + "grad_norm": 1.532518523136975, + "learning_rate": 2.1029422756391612e-07, + "loss": 0.9666, + "step": 9504 + }, + { + "epoch": 0.8571943905848401, + "grad_norm": 1.4748998898946266, + "learning_rate": 2.1003353497009812e-07, + "loss": 0.8602, + "step": 9505 + }, + { + "epoch": 0.8572845741083104, + "grad_norm": 1.5015532875373858, + "learning_rate": 2.0977299510608825e-07, + "loss": 0.8726, + "step": 9506 + }, + { + "epoch": 0.8573747576317807, + "grad_norm": 1.4612074005992162, + "learning_rate": 2.0951260799411784e-07, + "loss": 0.9828, + "step": 9507 + }, + { + "epoch": 0.857464941155251, + "grad_norm": 1.6046702680247558, + "learning_rate": 2.0925237365640424e-07, + "loss": 0.9329, + "step": 9508 + }, + { + "epoch": 0.8575551246787212, + "grad_norm": 0.6725263100125324, + "learning_rate": 2.0899229211515211e-07, + "loss": 0.7774, + "step": 9509 + }, + { + "epoch": 0.8576453082021914, + "grad_norm": 1.9821688185161364, + "learning_rate": 2.0873236339255306e-07, + "loss": 0.8746, + "step": 9510 + }, + { + "epoch": 0.8577354917256618, + "grad_norm": 1.4600873299177444, + "learning_rate": 2.0847258751078644e-07, + "loss": 0.8916, + "step": 9511 + }, + { + "epoch": 0.857825675249132, + "grad_norm": 1.8523783757365455, + "learning_rate": 2.082129644920163e-07, + "loss": 0.9272, + "step": 9512 + }, + { + "epoch": 0.8579158587726022, + "grad_norm": 1.513198640741781, + "learning_rate": 2.0795349435839605e-07, + "loss": 0.9593, + "step": 9513 + }, + { + "epoch": 0.8580060422960725, + "grad_norm": 1.6416572593122514, + "learning_rate": 2.0769417713206484e-07, + "loss": 0.9139, + "step": 9514 + }, + { + "epoch": 0.8580962258195428, + "grad_norm": 1.2878468185254583, + "learning_rate": 2.074350128351492e-07, + "loss": 1.0094, + "step": 9515 + }, + { + "epoch": 0.858186409343013, + "grad_norm": 1.6531287238326073, + "learning_rate": 2.0717600148976256e-07, + "loss": 0.8625, + "step": 9516 + }, + { + "epoch": 0.8582765928664833, + "grad_norm": 1.4726407887585669, + "learning_rate": 2.0691714311800436e-07, + "loss": 0.8811, + "step": 9517 + }, + { + "epoch": 0.8583667763899535, + "grad_norm": 1.4368052254851544, + "learning_rate": 2.066584377419631e-07, + "loss": 1.0048, + "step": 9518 + }, + { + "epoch": 0.8584569599134239, + "grad_norm": 1.497873388239081, + "learning_rate": 2.0639988538371167e-07, + "loss": 0.9667, + "step": 9519 + }, + { + "epoch": 0.8585471434368941, + "grad_norm": 1.447222216079982, + "learning_rate": 2.0614148606531258e-07, + "loss": 0.9127, + "step": 9520 + }, + { + "epoch": 0.8586373269603643, + "grad_norm": 3.3249495136759073, + "learning_rate": 2.0588323980881285e-07, + "loss": 0.9755, + "step": 9521 + }, + { + "epoch": 0.8587275104838346, + "grad_norm": 1.570228308702918, + "learning_rate": 2.0562514663624752e-07, + "loss": 0.9343, + "step": 9522 + }, + { + "epoch": 0.8588176940073049, + "grad_norm": 1.3250596518854114, + "learning_rate": 2.0536720656963902e-07, + "loss": 1.0117, + "step": 9523 + }, + { + "epoch": 0.8589078775307751, + "grad_norm": 1.790995723568035, + "learning_rate": 2.051094196309957e-07, + "loss": 0.9732, + "step": 9524 + }, + { + "epoch": 0.8589980610542454, + "grad_norm": 1.4199992792369465, + "learning_rate": 2.0485178584231378e-07, + "loss": 0.9032, + "step": 9525 + }, + { + "epoch": 0.8590882445777156, + "grad_norm": 0.6909347262557595, + "learning_rate": 2.0459430522557587e-07, + "loss": 0.7787, + "step": 9526 + }, + { + "epoch": 0.8591784281011859, + "grad_norm": 1.5592506955432548, + "learning_rate": 2.0433697780275195e-07, + "loss": 0.95, + "step": 9527 + }, + { + "epoch": 0.8592686116246562, + "grad_norm": 1.2942730182754627, + "learning_rate": 2.040798035957978e-07, + "loss": 0.9642, + "step": 9528 + }, + { + "epoch": 0.8593587951481264, + "grad_norm": 0.7355016628638167, + "learning_rate": 2.038227826266574e-07, + "loss": 0.7949, + "step": 9529 + }, + { + "epoch": 0.8594489786715968, + "grad_norm": 1.6070204612507017, + "learning_rate": 2.0356591491726126e-07, + "loss": 0.9524, + "step": 9530 + }, + { + "epoch": 0.859539162195067, + "grad_norm": 1.9580313131772353, + "learning_rate": 2.033092004895267e-07, + "loss": 0.989, + "step": 9531 + }, + { + "epoch": 0.8596293457185372, + "grad_norm": 1.6393387611679893, + "learning_rate": 2.03052639365358e-07, + "loss": 0.9651, + "step": 9532 + }, + { + "epoch": 0.8597195292420075, + "grad_norm": 1.470124920059045, + "learning_rate": 2.0279623156664694e-07, + "loss": 0.965, + "step": 9533 + }, + { + "epoch": 0.8598097127654778, + "grad_norm": 1.522243168532839, + "learning_rate": 2.0253997711527005e-07, + "loss": 1.0087, + "step": 9534 + }, + { + "epoch": 0.859899896288948, + "grad_norm": 1.899763358588125, + "learning_rate": 2.0228387603309428e-07, + "loss": 0.9945, + "step": 9535 + }, + { + "epoch": 0.8599900798124183, + "grad_norm": 1.4707981195458038, + "learning_rate": 2.0202792834197035e-07, + "loss": 0.9684, + "step": 9536 + }, + { + "epoch": 0.8600802633358885, + "grad_norm": 1.5846228952991928, + "learning_rate": 2.017721340637375e-07, + "loss": 0.8778, + "step": 9537 + }, + { + "epoch": 0.8601704468593588, + "grad_norm": 1.5982515676279792, + "learning_rate": 2.0151649322022134e-07, + "loss": 0.9288, + "step": 9538 + }, + { + "epoch": 0.8602606303828291, + "grad_norm": 2.1112971837319985, + "learning_rate": 2.012610058332349e-07, + "loss": 0.7593, + "step": 9539 + }, + { + "epoch": 0.8603508139062993, + "grad_norm": 0.6884412174756515, + "learning_rate": 2.010056719245774e-07, + "loss": 0.7755, + "step": 9540 + }, + { + "epoch": 0.8604409974297695, + "grad_norm": 2.083629345527231, + "learning_rate": 2.0075049151603563e-07, + "loss": 0.9699, + "step": 9541 + }, + { + "epoch": 0.8605311809532399, + "grad_norm": 1.5123947958179962, + "learning_rate": 2.0049546462938326e-07, + "loss": 0.9524, + "step": 9542 + }, + { + "epoch": 0.8606213644767101, + "grad_norm": 1.368560319760714, + "learning_rate": 2.0024059128637961e-07, + "loss": 0.9698, + "step": 9543 + }, + { + "epoch": 0.8607115480001803, + "grad_norm": 1.4655444287763115, + "learning_rate": 1.99985871508773e-07, + "loss": 0.812, + "step": 9544 + }, + { + "epoch": 0.8608017315236506, + "grad_norm": 1.7505008117142173, + "learning_rate": 1.9973130531829674e-07, + "loss": 0.9738, + "step": 9545 + }, + { + "epoch": 0.8608919150471209, + "grad_norm": 1.6339455647593655, + "learning_rate": 1.994768927366721e-07, + "loss": 0.8992, + "step": 9546 + }, + { + "epoch": 0.8609820985705912, + "grad_norm": 1.528095031319601, + "learning_rate": 1.992226337856069e-07, + "loss": 0.9177, + "step": 9547 + }, + { + "epoch": 0.8610722820940614, + "grad_norm": 1.643334612558936, + "learning_rate": 1.9896852848679592e-07, + "loss": 0.8991, + "step": 9548 + }, + { + "epoch": 0.8611624656175316, + "grad_norm": 1.515378794385223, + "learning_rate": 1.9871457686192094e-07, + "loss": 0.9555, + "step": 9549 + }, + { + "epoch": 0.861252649141002, + "grad_norm": 1.6671989147363964, + "learning_rate": 1.984607789326509e-07, + "loss": 0.9415, + "step": 9550 + }, + { + "epoch": 0.8613428326644722, + "grad_norm": 1.45902808760968, + "learning_rate": 1.982071347206402e-07, + "loss": 0.8975, + "step": 9551 + }, + { + "epoch": 0.8614330161879424, + "grad_norm": 1.346794007381978, + "learning_rate": 1.9795364424753202e-07, + "loss": 0.9263, + "step": 9552 + }, + { + "epoch": 0.8615231997114128, + "grad_norm": 1.6731137116499901, + "learning_rate": 1.9770030753495505e-07, + "loss": 0.9141, + "step": 9553 + }, + { + "epoch": 0.861613383234883, + "grad_norm": 1.751460896113477, + "learning_rate": 1.9744712460452573e-07, + "loss": 0.9056, + "step": 9554 + }, + { + "epoch": 0.8617035667583532, + "grad_norm": 2.0566554635029757, + "learning_rate": 1.9719409547784703e-07, + "loss": 0.927, + "step": 9555 + }, + { + "epoch": 0.8617937502818235, + "grad_norm": 1.763550408828557, + "learning_rate": 1.9694122017650837e-07, + "loss": 0.9728, + "step": 9556 + }, + { + "epoch": 0.8618839338052938, + "grad_norm": 1.4153968037643447, + "learning_rate": 1.9668849872208738e-07, + "loss": 0.9609, + "step": 9557 + }, + { + "epoch": 0.8619741173287641, + "grad_norm": 1.417468947082222, + "learning_rate": 1.9643593113614632e-07, + "loss": 0.7932, + "step": 9558 + }, + { + "epoch": 0.8620643008522343, + "grad_norm": 1.60870206522868, + "learning_rate": 1.961835174402371e-07, + "loss": 0.9877, + "step": 9559 + }, + { + "epoch": 0.8621544843757045, + "grad_norm": 1.7997341463005667, + "learning_rate": 1.9593125765589535e-07, + "loss": 0.9151, + "step": 9560 + }, + { + "epoch": 0.8622446678991749, + "grad_norm": 1.5864215732188727, + "learning_rate": 1.9567915180464721e-07, + "loss": 0.8921, + "step": 9561 + }, + { + "epoch": 0.8623348514226451, + "grad_norm": 1.8745902567431287, + "learning_rate": 1.9542719990800217e-07, + "loss": 0.9323, + "step": 9562 + }, + { + "epoch": 0.8624250349461153, + "grad_norm": 1.5565085169543533, + "learning_rate": 1.9517540198745896e-07, + "loss": 1.0178, + "step": 9563 + }, + { + "epoch": 0.8625152184695856, + "grad_norm": 1.7096842874794176, + "learning_rate": 1.94923758064502e-07, + "loss": 0.9399, + "step": 9564 + }, + { + "epoch": 0.8626054019930559, + "grad_norm": 1.3832249430433399, + "learning_rate": 1.9467226816060322e-07, + "loss": 0.9176, + "step": 9565 + }, + { + "epoch": 0.8626955855165261, + "grad_norm": 1.8500170474942352, + "learning_rate": 1.9442093229722122e-07, + "loss": 0.8887, + "step": 9566 + }, + { + "epoch": 0.8627857690399964, + "grad_norm": 2.114522234655685, + "learning_rate": 1.9416975049580085e-07, + "loss": 1.0241, + "step": 9567 + }, + { + "epoch": 0.8628759525634666, + "grad_norm": 2.658208596183982, + "learning_rate": 1.9391872277777456e-07, + "loss": 0.9275, + "step": 9568 + }, + { + "epoch": 0.862966136086937, + "grad_norm": 1.8505807197114765, + "learning_rate": 1.9366784916456158e-07, + "loss": 0.9363, + "step": 9569 + }, + { + "epoch": 0.8630563196104072, + "grad_norm": 1.5255657491152772, + "learning_rate": 1.9341712967756774e-07, + "loss": 0.9038, + "step": 9570 + }, + { + "epoch": 0.8631465031338774, + "grad_norm": 1.439144272194694, + "learning_rate": 1.9316656433818566e-07, + "loss": 0.8583, + "step": 9571 + }, + { + "epoch": 0.8632366866573477, + "grad_norm": 1.5174161414616651, + "learning_rate": 1.929161531677954e-07, + "loss": 0.8447, + "step": 9572 + }, + { + "epoch": 0.863326870180818, + "grad_norm": 1.5273208811577172, + "learning_rate": 1.9266589618776251e-07, + "loss": 0.8994, + "step": 9573 + }, + { + "epoch": 0.8634170537042882, + "grad_norm": 1.6081037222399348, + "learning_rate": 1.924157934194417e-07, + "loss": 0.9949, + "step": 9574 + }, + { + "epoch": 0.8635072372277585, + "grad_norm": 1.707057856651379, + "learning_rate": 1.9216584488417142e-07, + "loss": 0.8381, + "step": 9575 + }, + { + "epoch": 0.8635974207512288, + "grad_norm": 1.9174621285847824, + "learning_rate": 1.919160506032802e-07, + "loss": 1.0006, + "step": 9576 + }, + { + "epoch": 0.863687604274699, + "grad_norm": 1.4688428484199705, + "learning_rate": 1.916664105980812e-07, + "loss": 1.0072, + "step": 9577 + }, + { + "epoch": 0.8637777877981693, + "grad_norm": 1.52982684583016, + "learning_rate": 1.914169248898747e-07, + "loss": 0.9802, + "step": 9578 + }, + { + "epoch": 0.8638679713216395, + "grad_norm": 1.2731133733039648, + "learning_rate": 1.9116759349994882e-07, + "loss": 0.8575, + "step": 9579 + }, + { + "epoch": 0.8639581548451098, + "grad_norm": 1.5266619084768538, + "learning_rate": 1.9091841644957763e-07, + "loss": 0.9803, + "step": 9580 + }, + { + "epoch": 0.8640483383685801, + "grad_norm": 1.2824866676752988, + "learning_rate": 1.9066939376002278e-07, + "loss": 1.0306, + "step": 9581 + }, + { + "epoch": 0.8641385218920503, + "grad_norm": 1.94929495071965, + "learning_rate": 1.9042052545253085e-07, + "loss": 0.8612, + "step": 9582 + }, + { + "epoch": 0.8642287054155205, + "grad_norm": 1.458534753604426, + "learning_rate": 1.901718115483384e-07, + "loss": 0.9246, + "step": 9583 + }, + { + "epoch": 0.8643188889389909, + "grad_norm": 1.3865551919252337, + "learning_rate": 1.8992325206866598e-07, + "loss": 0.9277, + "step": 9584 + }, + { + "epoch": 0.8644090724624611, + "grad_norm": 1.3538536981684215, + "learning_rate": 1.8967484703472225e-07, + "loss": 0.9191, + "step": 9585 + }, + { + "epoch": 0.8644992559859314, + "grad_norm": 1.272215367918323, + "learning_rate": 1.8942659646770288e-07, + "loss": 0.9886, + "step": 9586 + }, + { + "epoch": 0.8645894395094016, + "grad_norm": 1.7633354398405958, + "learning_rate": 1.8917850038878936e-07, + "loss": 1.0104, + "step": 9587 + }, + { + "epoch": 0.8646796230328719, + "grad_norm": 1.451087096909549, + "learning_rate": 1.8893055881915121e-07, + "loss": 0.8551, + "step": 9588 + }, + { + "epoch": 0.8647698065563422, + "grad_norm": 1.819540752485988, + "learning_rate": 1.886827717799442e-07, + "loss": 0.8942, + "step": 9589 + }, + { + "epoch": 0.8648599900798124, + "grad_norm": 2.437808970896359, + "learning_rate": 1.884351392923096e-07, + "loss": 0.9664, + "step": 9590 + }, + { + "epoch": 0.8649501736032826, + "grad_norm": 2.0475971607821246, + "learning_rate": 1.8818766137737896e-07, + "loss": 0.9874, + "step": 9591 + }, + { + "epoch": 0.865040357126753, + "grad_norm": 1.771378855732762, + "learning_rate": 1.8794033805626653e-07, + "loss": 0.9239, + "step": 9592 + }, + { + "epoch": 0.8651305406502232, + "grad_norm": 1.5503956197560638, + "learning_rate": 1.876931693500763e-07, + "loss": 0.9371, + "step": 9593 + }, + { + "epoch": 0.8652207241736934, + "grad_norm": 1.736187717866865, + "learning_rate": 1.8744615527989783e-07, + "loss": 0.7975, + "step": 9594 + }, + { + "epoch": 0.8653109076971637, + "grad_norm": 1.5678185543546364, + "learning_rate": 1.871992958668076e-07, + "loss": 1.0347, + "step": 9595 + }, + { + "epoch": 0.865401091220634, + "grad_norm": 1.9882122395048816, + "learning_rate": 1.8695259113186944e-07, + "loss": 0.9866, + "step": 9596 + }, + { + "epoch": 0.8654912747441043, + "grad_norm": 1.5090928250297508, + "learning_rate": 1.8670604109613252e-07, + "loss": 0.9047, + "step": 9597 + }, + { + "epoch": 0.8655814582675745, + "grad_norm": 1.6095479483405277, + "learning_rate": 1.8645964578063533e-07, + "loss": 0.9924, + "step": 9598 + }, + { + "epoch": 0.8656716417910447, + "grad_norm": 1.4757458078259262, + "learning_rate": 1.862134052064006e-07, + "loss": 1.0023, + "step": 9599 + }, + { + "epoch": 0.8657618253145151, + "grad_norm": 1.7438850514607118, + "learning_rate": 1.8596731939443932e-07, + "loss": 0.9893, + "step": 9600 + }, + { + "epoch": 0.8658520088379853, + "grad_norm": 1.4390263953634534, + "learning_rate": 1.857213883657487e-07, + "loss": 0.9738, + "step": 9601 + }, + { + "epoch": 0.8659421923614555, + "grad_norm": 0.6543536716690758, + "learning_rate": 1.8547561214131303e-07, + "loss": 0.7517, + "step": 9602 + }, + { + "epoch": 0.8660323758849259, + "grad_norm": 1.635312224857738, + "learning_rate": 1.8522999074210355e-07, + "loss": 1.0262, + "step": 9603 + }, + { + "epoch": 0.8661225594083961, + "grad_norm": 1.6930983187071806, + "learning_rate": 1.849845241890775e-07, + "loss": 0.9842, + "step": 9604 + }, + { + "epoch": 0.8662127429318663, + "grad_norm": 1.5478730498064952, + "learning_rate": 1.8473921250317992e-07, + "loss": 0.9463, + "step": 9605 + }, + { + "epoch": 0.8663029264553366, + "grad_norm": 1.543302861991507, + "learning_rate": 1.8449405570534225e-07, + "loss": 0.8766, + "step": 9606 + }, + { + "epoch": 0.8663931099788069, + "grad_norm": 1.483283273873736, + "learning_rate": 1.8424905381648204e-07, + "loss": 0.9889, + "step": 9607 + }, + { + "epoch": 0.8664832935022772, + "grad_norm": 1.430994988923859, + "learning_rate": 1.8400420685750452e-07, + "loss": 0.9027, + "step": 9608 + }, + { + "epoch": 0.8665734770257474, + "grad_norm": 1.5081406468489935, + "learning_rate": 1.8375951484930142e-07, + "loss": 0.8803, + "step": 9609 + }, + { + "epoch": 0.8666636605492176, + "grad_norm": 1.5081513177351533, + "learning_rate": 1.8351497781275094e-07, + "loss": 0.9499, + "step": 9610 + }, + { + "epoch": 0.866753844072688, + "grad_norm": 1.6852341026888455, + "learning_rate": 1.8327059576871907e-07, + "loss": 0.9362, + "step": 9611 + }, + { + "epoch": 0.8668440275961582, + "grad_norm": 2.3441250310301913, + "learning_rate": 1.8302636873805665e-07, + "loss": 0.9938, + "step": 9612 + }, + { + "epoch": 0.8669342111196284, + "grad_norm": 1.5171450042273247, + "learning_rate": 1.8278229674160373e-07, + "loss": 1.0255, + "step": 9613 + }, + { + "epoch": 0.8670243946430987, + "grad_norm": 1.491301028584865, + "learning_rate": 1.825383798001845e-07, + "loss": 0.878, + "step": 9614 + }, + { + "epoch": 0.867114578166569, + "grad_norm": 1.3877972344613518, + "learning_rate": 1.8229461793461297e-07, + "loss": 0.8963, + "step": 9615 + }, + { + "epoch": 0.8672047616900392, + "grad_norm": 1.7336644272151995, + "learning_rate": 1.8205101116568698e-07, + "loss": 0.9596, + "step": 9616 + }, + { + "epoch": 0.8672949452135095, + "grad_norm": 1.9477550691155028, + "learning_rate": 1.818075595141928e-07, + "loss": 0.9978, + "step": 9617 + }, + { + "epoch": 0.8673851287369797, + "grad_norm": 1.6864986274277105, + "learning_rate": 1.8156426300090288e-07, + "loss": 0.942, + "step": 9618 + }, + { + "epoch": 0.86747531226045, + "grad_norm": 1.4829771185644745, + "learning_rate": 1.8132112164657686e-07, + "loss": 0.9353, + "step": 9619 + }, + { + "epoch": 0.8675654957839203, + "grad_norm": 1.5793045470235136, + "learning_rate": 1.8107813547196106e-07, + "loss": 0.8425, + "step": 9620 + }, + { + "epoch": 0.8676556793073905, + "grad_norm": 1.5436511884564221, + "learning_rate": 1.8083530449778817e-07, + "loss": 0.9848, + "step": 9621 + }, + { + "epoch": 0.8677458628308607, + "grad_norm": 2.0392176543506055, + "learning_rate": 1.8059262874477787e-07, + "loss": 0.867, + "step": 9622 + }, + { + "epoch": 0.8678360463543311, + "grad_norm": 1.4740115134774265, + "learning_rate": 1.8035010823363627e-07, + "loss": 0.9339, + "step": 9623 + }, + { + "epoch": 0.8679262298778013, + "grad_norm": 1.5371267470428747, + "learning_rate": 1.8010774298505705e-07, + "loss": 0.9843, + "step": 9624 + }, + { + "epoch": 0.8680164134012716, + "grad_norm": 1.6429019205139626, + "learning_rate": 1.7986553301972007e-07, + "loss": 1.0332, + "step": 9625 + }, + { + "epoch": 0.8681065969247419, + "grad_norm": 1.968418820091622, + "learning_rate": 1.7962347835829171e-07, + "loss": 0.9367, + "step": 9626 + }, + { + "epoch": 0.8681967804482121, + "grad_norm": 1.4351257952271015, + "learning_rate": 1.793815790214257e-07, + "loss": 0.9526, + "step": 9627 + }, + { + "epoch": 0.8682869639716824, + "grad_norm": 1.8753752650796152, + "learning_rate": 1.791398350297626e-07, + "loss": 0.9029, + "step": 9628 + }, + { + "epoch": 0.8683771474951526, + "grad_norm": 1.4287375540871257, + "learning_rate": 1.7889824640392813e-07, + "loss": 0.9537, + "step": 9629 + }, + { + "epoch": 0.868467331018623, + "grad_norm": 1.7189457175054204, + "learning_rate": 1.7865681316453741e-07, + "loss": 0.9302, + "step": 9630 + }, + { + "epoch": 0.8685575145420932, + "grad_norm": 1.528486521887636, + "learning_rate": 1.7841553533218968e-07, + "loss": 0.9436, + "step": 9631 + }, + { + "epoch": 0.8686476980655634, + "grad_norm": 2.059187685793475, + "learning_rate": 1.7817441292747292e-07, + "loss": 1.0484, + "step": 9632 + }, + { + "epoch": 0.8687378815890336, + "grad_norm": 1.806601628063072, + "learning_rate": 1.779334459709607e-07, + "loss": 0.9435, + "step": 9633 + }, + { + "epoch": 0.868828065112504, + "grad_norm": 2.228229313745286, + "learning_rate": 1.7769263448321347e-07, + "loss": 0.9548, + "step": 9634 + }, + { + "epoch": 0.8689182486359742, + "grad_norm": 1.4217990393836726, + "learning_rate": 1.7745197848477879e-07, + "loss": 0.9353, + "step": 9635 + }, + { + "epoch": 0.8690084321594445, + "grad_norm": 1.7776022251527253, + "learning_rate": 1.7721147799619063e-07, + "loss": 1.0466, + "step": 9636 + }, + { + "epoch": 0.8690986156829147, + "grad_norm": 0.6033750203608574, + "learning_rate": 1.769711330379704e-07, + "loss": 0.7491, + "step": 9637 + }, + { + "epoch": 0.869188799206385, + "grad_norm": 1.892041897073953, + "learning_rate": 1.767309436306248e-07, + "loss": 0.9961, + "step": 9638 + }, + { + "epoch": 0.8692789827298553, + "grad_norm": 1.8991120949113816, + "learning_rate": 1.764909097946483e-07, + "loss": 1.0109, + "step": 9639 + }, + { + "epoch": 0.8693691662533255, + "grad_norm": 1.9930111248396838, + "learning_rate": 1.7625103155052236e-07, + "loss": 1.0123, + "step": 9640 + }, + { + "epoch": 0.8694593497767957, + "grad_norm": 1.6827150251091685, + "learning_rate": 1.760113089187143e-07, + "loss": 0.9587, + "step": 9641 + }, + { + "epoch": 0.8695495333002661, + "grad_norm": 1.5570916224908857, + "learning_rate": 1.7577174191967868e-07, + "loss": 0.9712, + "step": 9642 + }, + { + "epoch": 0.8696397168237363, + "grad_norm": 1.7026745961830347, + "learning_rate": 1.755323305738574e-07, + "loss": 0.9553, + "step": 9643 + }, + { + "epoch": 0.8697299003472065, + "grad_norm": 1.2949968549329354, + "learning_rate": 1.7529307490167677e-07, + "loss": 0.9772, + "step": 9644 + }, + { + "epoch": 0.8698200838706768, + "grad_norm": 2.2544615274760464, + "learning_rate": 1.7505397492355288e-07, + "loss": 0.9272, + "step": 9645 + }, + { + "epoch": 0.8699102673941471, + "grad_norm": 1.500891658885896, + "learning_rate": 1.7481503065988589e-07, + "loss": 0.982, + "step": 9646 + }, + { + "epoch": 0.8700004509176174, + "grad_norm": 1.3861869544508847, + "learning_rate": 1.7457624213106526e-07, + "loss": 0.9107, + "step": 9647 + }, + { + "epoch": 0.8700906344410876, + "grad_norm": 1.8900330302465547, + "learning_rate": 1.7433760935746465e-07, + "loss": 0.9463, + "step": 9648 + }, + { + "epoch": 0.8701808179645579, + "grad_norm": 1.5878918261834325, + "learning_rate": 1.740991323594456e-07, + "loss": 0.8904, + "step": 9649 + }, + { + "epoch": 0.8702710014880282, + "grad_norm": 1.8485753553744246, + "learning_rate": 1.7386081115735651e-07, + "loss": 0.9208, + "step": 9650 + }, + { + "epoch": 0.8703611850114984, + "grad_norm": 1.6898865954028588, + "learning_rate": 1.736226457715324e-07, + "loss": 0.9091, + "step": 9651 + }, + { + "epoch": 0.8704513685349686, + "grad_norm": 1.5083300240973883, + "learning_rate": 1.7338463622229505e-07, + "loss": 0.8983, + "step": 9652 + }, + { + "epoch": 0.870541552058439, + "grad_norm": 1.5398435208471892, + "learning_rate": 1.7314678252995152e-07, + "loss": 0.8937, + "step": 9653 + }, + { + "epoch": 0.8706317355819092, + "grad_norm": 1.7919741595865992, + "learning_rate": 1.7290908471479847e-07, + "loss": 0.8791, + "step": 9654 + }, + { + "epoch": 0.8707219191053794, + "grad_norm": 1.5479410524702184, + "learning_rate": 1.7267154279711637e-07, + "loss": 0.9453, + "step": 9655 + }, + { + "epoch": 0.8708121026288497, + "grad_norm": 1.4449166245479845, + "learning_rate": 1.724341567971741e-07, + "loss": 0.9936, + "step": 9656 + }, + { + "epoch": 0.87090228615232, + "grad_norm": 1.5928388496780523, + "learning_rate": 1.7219692673522657e-07, + "loss": 1.042, + "step": 9657 + }, + { + "epoch": 0.8709924696757902, + "grad_norm": 1.7023139212219667, + "learning_rate": 1.7195985263151558e-07, + "loss": 0.9657, + "step": 9658 + }, + { + "epoch": 0.8710826531992605, + "grad_norm": 1.558488349888141, + "learning_rate": 1.7172293450626985e-07, + "loss": 0.9902, + "step": 9659 + }, + { + "epoch": 0.8711728367227307, + "grad_norm": 1.53961860950132, + "learning_rate": 1.7148617237970475e-07, + "loss": 0.9479, + "step": 9660 + }, + { + "epoch": 0.8712630202462011, + "grad_norm": 1.3443231469249328, + "learning_rate": 1.7124956627202102e-07, + "loss": 0.9114, + "step": 9661 + }, + { + "epoch": 0.8713532037696713, + "grad_norm": 2.1601810833295105, + "learning_rate": 1.7101311620340852e-07, + "loss": 0.9597, + "step": 9662 + }, + { + "epoch": 0.8714433872931415, + "grad_norm": 1.641950598560195, + "learning_rate": 1.7077682219404155e-07, + "loss": 0.8869, + "step": 9663 + }, + { + "epoch": 0.8715335708166118, + "grad_norm": 1.7258954032420761, + "learning_rate": 1.705406842640824e-07, + "loss": 1.0754, + "step": 9664 + }, + { + "epoch": 0.8716237543400821, + "grad_norm": 1.5200156099371198, + "learning_rate": 1.7030470243367946e-07, + "loss": 0.9374, + "step": 9665 + }, + { + "epoch": 0.8717139378635523, + "grad_norm": 1.261917004200279, + "learning_rate": 1.7006887672296834e-07, + "loss": 0.8696, + "step": 9666 + }, + { + "epoch": 0.8718041213870226, + "grad_norm": 1.6569680151053838, + "learning_rate": 1.6983320715207094e-07, + "loss": 0.8902, + "step": 9667 + }, + { + "epoch": 0.8718943049104928, + "grad_norm": 1.4478241412927564, + "learning_rate": 1.6959769374109523e-07, + "loss": 0.9658, + "step": 9668 + }, + { + "epoch": 0.8719844884339631, + "grad_norm": 2.653069241031051, + "learning_rate": 1.6936233651013754e-07, + "loss": 0.8795, + "step": 9669 + }, + { + "epoch": 0.8720746719574334, + "grad_norm": 1.6970997895060214, + "learning_rate": 1.691271354792787e-07, + "loss": 0.9767, + "step": 9670 + }, + { + "epoch": 0.8721648554809036, + "grad_norm": 1.425954228189574, + "learning_rate": 1.6889209066858866e-07, + "loss": 0.8277, + "step": 9671 + }, + { + "epoch": 0.872255039004374, + "grad_norm": 1.5861037594694813, + "learning_rate": 1.6865720209812185e-07, + "loss": 0.9859, + "step": 9672 + }, + { + "epoch": 0.8723452225278442, + "grad_norm": 1.8051061763866343, + "learning_rate": 1.684224697879204e-07, + "loss": 0.9103, + "step": 9673 + }, + { + "epoch": 0.8724354060513144, + "grad_norm": 1.7198446775877336, + "learning_rate": 1.6818789375801302e-07, + "loss": 1.0189, + "step": 9674 + }, + { + "epoch": 0.8725255895747847, + "grad_norm": 1.3002096300567403, + "learning_rate": 1.679534740284152e-07, + "loss": 0.9031, + "step": 9675 + }, + { + "epoch": 0.872615773098255, + "grad_norm": 1.6120862622241425, + "learning_rate": 1.6771921061912853e-07, + "loss": 0.8942, + "step": 9676 + }, + { + "epoch": 0.8727059566217252, + "grad_norm": 1.6405756625068502, + "learning_rate": 1.6748510355014234e-07, + "loss": 0.882, + "step": 9677 + }, + { + "epoch": 0.8727961401451955, + "grad_norm": 1.4413462060347038, + "learning_rate": 1.6725115284143132e-07, + "loss": 0.9617, + "step": 9678 + }, + { + "epoch": 0.8728863236686657, + "grad_norm": 1.8410575086530694, + "learning_rate": 1.670173585129575e-07, + "loss": 0.9203, + "step": 9679 + }, + { + "epoch": 0.872976507192136, + "grad_norm": 1.421104390008572, + "learning_rate": 1.667837205846696e-07, + "loss": 1.0074, + "step": 9680 + }, + { + "epoch": 0.8730666907156063, + "grad_norm": 1.5003726814148497, + "learning_rate": 1.6655023907650278e-07, + "loss": 0.953, + "step": 9681 + }, + { + "epoch": 0.8731568742390765, + "grad_norm": 1.73844912286302, + "learning_rate": 1.6631691400837954e-07, + "loss": 0.9891, + "step": 9682 + }, + { + "epoch": 0.8732470577625467, + "grad_norm": 1.4821198850680026, + "learning_rate": 1.6608374540020752e-07, + "loss": 0.994, + "step": 9683 + }, + { + "epoch": 0.8733372412860171, + "grad_norm": 0.6860167368523724, + "learning_rate": 1.658507332718828e-07, + "loss": 0.7282, + "step": 9684 + }, + { + "epoch": 0.8734274248094873, + "grad_norm": 1.4556860556272453, + "learning_rate": 1.656178776432864e-07, + "loss": 0.8826, + "step": 9685 + }, + { + "epoch": 0.8735176083329576, + "grad_norm": 1.4276265783358828, + "learning_rate": 1.6538517853428814e-07, + "loss": 0.9249, + "step": 9686 + }, + { + "epoch": 0.8736077918564278, + "grad_norm": 1.4032454288176461, + "learning_rate": 1.6515263596474194e-07, + "loss": 0.9253, + "step": 9687 + }, + { + "epoch": 0.8736979753798981, + "grad_norm": 1.76743055539035, + "learning_rate": 1.6492024995449017e-07, + "loss": 0.9319, + "step": 9688 + }, + { + "epoch": 0.8737881589033684, + "grad_norm": 1.7237763169501708, + "learning_rate": 1.6468802052336116e-07, + "loss": 0.9137, + "step": 9689 + }, + { + "epoch": 0.8738783424268386, + "grad_norm": 1.432518995335514, + "learning_rate": 1.6445594769116998e-07, + "loss": 0.9656, + "step": 9690 + }, + { + "epoch": 0.8739685259503088, + "grad_norm": 1.781211116433294, + "learning_rate": 1.6422403147771836e-07, + "loss": 1.017, + "step": 9691 + }, + { + "epoch": 0.8740587094737792, + "grad_norm": 1.4973048793425074, + "learning_rate": 1.6399227190279485e-07, + "loss": 0.9072, + "step": 9692 + }, + { + "epoch": 0.8741488929972494, + "grad_norm": 1.508949997105689, + "learning_rate": 1.637606689861748e-07, + "loss": 0.8757, + "step": 9693 + }, + { + "epoch": 0.8742390765207196, + "grad_norm": 1.8638224905283607, + "learning_rate": 1.6352922274761883e-07, + "loss": 0.9071, + "step": 9694 + }, + { + "epoch": 0.87432926004419, + "grad_norm": 1.4573142260136194, + "learning_rate": 1.6329793320687602e-07, + "loss": 0.8124, + "step": 9695 + }, + { + "epoch": 0.8744194435676602, + "grad_norm": 1.4324242085483874, + "learning_rate": 1.630668003836808e-07, + "loss": 0.9799, + "step": 9696 + }, + { + "epoch": 0.8745096270911304, + "grad_norm": 1.4636003023214514, + "learning_rate": 1.62835824297755e-07, + "loss": 0.9502, + "step": 9697 + }, + { + "epoch": 0.8745998106146007, + "grad_norm": 1.5729439166742218, + "learning_rate": 1.626050049688066e-07, + "loss": 0.9532, + "step": 9698 + }, + { + "epoch": 0.874689994138071, + "grad_norm": 1.6770274545905077, + "learning_rate": 1.623743424165309e-07, + "loss": 0.9185, + "step": 9699 + }, + { + "epoch": 0.8747801776615413, + "grad_norm": 1.6864906400618158, + "learning_rate": 1.6214383666060826e-07, + "loss": 0.8938, + "step": 9700 + }, + { + "epoch": 0.8748703611850115, + "grad_norm": 1.596841693090379, + "learning_rate": 1.619134877207078e-07, + "loss": 0.9631, + "step": 9701 + }, + { + "epoch": 0.8749605447084817, + "grad_norm": 1.621599454019433, + "learning_rate": 1.616832956164831e-07, + "loss": 0.9485, + "step": 9702 + }, + { + "epoch": 0.8750507282319521, + "grad_norm": 2.098006746643099, + "learning_rate": 1.6145326036757667e-07, + "loss": 0.9158, + "step": 9703 + }, + { + "epoch": 0.8751409117554223, + "grad_norm": 1.5441313021721237, + "learning_rate": 1.612233819936155e-07, + "loss": 0.9092, + "step": 9704 + }, + { + "epoch": 0.8752310952788925, + "grad_norm": 1.6421206967951587, + "learning_rate": 1.6099366051421414e-07, + "loss": 0.94, + "step": 9705 + }, + { + "epoch": 0.8753212788023628, + "grad_norm": 1.8292207612863784, + "learning_rate": 1.6076409594897378e-07, + "loss": 1.006, + "step": 9706 + }, + { + "epoch": 0.8754114623258331, + "grad_norm": 0.6865752893783759, + "learning_rate": 1.605346883174823e-07, + "loss": 0.7287, + "step": 9707 + }, + { + "epoch": 0.8755016458493033, + "grad_norm": 1.680134558511487, + "learning_rate": 1.6030543763931427e-07, + "loss": 0.9182, + "step": 9708 + }, + { + "epoch": 0.8755918293727736, + "grad_norm": 2.218447919886399, + "learning_rate": 1.600763439340298e-07, + "loss": 0.9645, + "step": 9709 + }, + { + "epoch": 0.8756820128962438, + "grad_norm": 8.295629375510929, + "learning_rate": 1.5984740722117707e-07, + "loss": 0.9969, + "step": 9710 + }, + { + "epoch": 0.8757721964197142, + "grad_norm": 1.2980010418028096, + "learning_rate": 1.5961862752028998e-07, + "loss": 0.9582, + "step": 9711 + }, + { + "epoch": 0.8758623799431844, + "grad_norm": 1.3631363056593797, + "learning_rate": 1.5939000485088937e-07, + "loss": 0.8528, + "step": 9712 + }, + { + "epoch": 0.8759525634666546, + "grad_norm": 1.9155281527807047, + "learning_rate": 1.5916153923248254e-07, + "loss": 1.0377, + "step": 9713 + }, + { + "epoch": 0.8760427469901249, + "grad_norm": 1.4196338807513709, + "learning_rate": 1.5893323068456342e-07, + "loss": 0.9877, + "step": 9714 + }, + { + "epoch": 0.8761329305135952, + "grad_norm": 1.5384053797742359, + "learning_rate": 1.5870507922661248e-07, + "loss": 0.9923, + "step": 9715 + }, + { + "epoch": 0.8762231140370654, + "grad_norm": 0.6467682828733619, + "learning_rate": 1.5847708487809763e-07, + "loss": 0.7693, + "step": 9716 + }, + { + "epoch": 0.8763132975605357, + "grad_norm": 1.6402554777077467, + "learning_rate": 1.5824924765847113e-07, + "loss": 1.0259, + "step": 9717 + }, + { + "epoch": 0.8764034810840059, + "grad_norm": 1.2837936923606468, + "learning_rate": 1.5802156758717478e-07, + "loss": 0.9229, + "step": 9718 + }, + { + "epoch": 0.8764936646074762, + "grad_norm": 1.4611254892419405, + "learning_rate": 1.5779404468363433e-07, + "loss": 0.9448, + "step": 9719 + }, + { + "epoch": 0.8765838481309465, + "grad_norm": 1.6011493429072243, + "learning_rate": 1.5756667896726405e-07, + "loss": 1.0052, + "step": 9720 + }, + { + "epoch": 0.8766740316544167, + "grad_norm": 2.2155871346362734, + "learning_rate": 1.5733947045746377e-07, + "loss": 0.964, + "step": 9721 + }, + { + "epoch": 0.876764215177887, + "grad_norm": 1.3995311445836884, + "learning_rate": 1.5711241917362018e-07, + "loss": 1.0134, + "step": 9722 + }, + { + "epoch": 0.8768543987013573, + "grad_norm": 1.966465487400635, + "learning_rate": 1.5688552513510688e-07, + "loss": 0.8817, + "step": 9723 + }, + { + "epoch": 0.8769445822248275, + "grad_norm": 1.4648594562960056, + "learning_rate": 1.5665878836128266e-07, + "loss": 0.969, + "step": 9724 + }, + { + "epoch": 0.8770347657482978, + "grad_norm": 1.3245598007820478, + "learning_rate": 1.5643220887149554e-07, + "loss": 0.99, + "step": 9725 + }, + { + "epoch": 0.8771249492717681, + "grad_norm": 0.7754893204072342, + "learning_rate": 1.562057866850772e-07, + "loss": 0.789, + "step": 9726 + }, + { + "epoch": 0.8772151327952383, + "grad_norm": 1.6642602794043813, + "learning_rate": 1.5597952182134777e-07, + "loss": 0.8202, + "step": 9727 + }, + { + "epoch": 0.8773053163187086, + "grad_norm": 1.4431834442741174, + "learning_rate": 1.557534142996133e-07, + "loss": 1.0026, + "step": 9728 + }, + { + "epoch": 0.8773954998421788, + "grad_norm": 1.4686549541481388, + "learning_rate": 1.5552746413916662e-07, + "loss": 0.9746, + "step": 9729 + }, + { + "epoch": 0.8774856833656491, + "grad_norm": 2.0096509064759625, + "learning_rate": 1.5530167135928697e-07, + "loss": 0.9405, + "step": 9730 + }, + { + "epoch": 0.8775758668891194, + "grad_norm": 1.61972466071613, + "learning_rate": 1.5507603597924068e-07, + "loss": 0.889, + "step": 9731 + }, + { + "epoch": 0.8776660504125896, + "grad_norm": 1.7005405239894957, + "learning_rate": 1.548505580182793e-07, + "loss": 0.9082, + "step": 9732 + }, + { + "epoch": 0.8777562339360598, + "grad_norm": 1.5865874062837888, + "learning_rate": 1.5462523749564271e-07, + "loss": 0.944, + "step": 9733 + }, + { + "epoch": 0.8778464174595302, + "grad_norm": 1.5009141361753984, + "learning_rate": 1.5440007443055602e-07, + "loss": 0.9446, + "step": 9734 + }, + { + "epoch": 0.8779366009830004, + "grad_norm": 1.690089323177507, + "learning_rate": 1.541750688422314e-07, + "loss": 0.8827, + "step": 9735 + }, + { + "epoch": 0.8780267845064706, + "grad_norm": 1.5651358687437247, + "learning_rate": 1.5395022074986797e-07, + "loss": 0.9499, + "step": 9736 + }, + { + "epoch": 0.8781169680299409, + "grad_norm": 1.5621173390068384, + "learning_rate": 1.5372553017265033e-07, + "loss": 0.8898, + "step": 9737 + }, + { + "epoch": 0.8782071515534112, + "grad_norm": 2.006717011949154, + "learning_rate": 1.5350099712975116e-07, + "loss": 1.0051, + "step": 9738 + }, + { + "epoch": 0.8782973350768815, + "grad_norm": 1.4939079552297492, + "learning_rate": 1.5327662164032785e-07, + "loss": 0.9378, + "step": 9739 + }, + { + "epoch": 0.8783875186003517, + "grad_norm": 1.5394875186465005, + "learning_rate": 1.5305240372352656e-07, + "loss": 0.7965, + "step": 9740 + }, + { + "epoch": 0.8784777021238219, + "grad_norm": 1.4615135485151296, + "learning_rate": 1.5282834339847738e-07, + "loss": 0.9591, + "step": 9741 + }, + { + "epoch": 0.8785678856472923, + "grad_norm": 1.491593007803313, + "learning_rate": 1.526044406842999e-07, + "loss": 0.9293, + "step": 9742 + }, + { + "epoch": 0.8786580691707625, + "grad_norm": 2.2622046599282033, + "learning_rate": 1.523806956000977e-07, + "loss": 0.8745, + "step": 9743 + }, + { + "epoch": 0.8787482526942327, + "grad_norm": 1.4754097288640649, + "learning_rate": 1.5215710816496197e-07, + "loss": 0.9964, + "step": 9744 + }, + { + "epoch": 0.8788384362177031, + "grad_norm": 1.6190447623631041, + "learning_rate": 1.5193367839797077e-07, + "loss": 0.8904, + "step": 9745 + }, + { + "epoch": 0.8789286197411733, + "grad_norm": 1.4312427170747257, + "learning_rate": 1.5171040631818842e-07, + "loss": 0.944, + "step": 9746 + }, + { + "epoch": 0.8790188032646435, + "grad_norm": 1.6309305689456666, + "learning_rate": 1.5148729194466547e-07, + "loss": 0.8727, + "step": 9747 + }, + { + "epoch": 0.8791089867881138, + "grad_norm": 1.4601738713866153, + "learning_rate": 1.5126433529643956e-07, + "loss": 1.018, + "step": 9748 + }, + { + "epoch": 0.8791991703115841, + "grad_norm": 1.426254653722474, + "learning_rate": 1.5104153639253436e-07, + "loss": 0.9445, + "step": 9749 + }, + { + "epoch": 0.8792893538350544, + "grad_norm": 0.649418651593278, + "learning_rate": 1.5081889525196002e-07, + "loss": 0.8305, + "step": 9750 + }, + { + "epoch": 0.8793795373585246, + "grad_norm": 1.6349232004076153, + "learning_rate": 1.5059641189371398e-07, + "loss": 1.0037, + "step": 9751 + }, + { + "epoch": 0.8794697208819948, + "grad_norm": 1.550239910509373, + "learning_rate": 1.503740863367795e-07, + "loss": 0.9636, + "step": 9752 + }, + { + "epoch": 0.8795599044054652, + "grad_norm": 2.020972912349686, + "learning_rate": 1.50151918600127e-07, + "loss": 0.9731, + "step": 9753 + }, + { + "epoch": 0.8796500879289354, + "grad_norm": 1.5658621663593653, + "learning_rate": 1.4992990870271217e-07, + "loss": 0.9983, + "step": 9754 + }, + { + "epoch": 0.8797402714524056, + "grad_norm": 2.8328662562122373, + "learning_rate": 1.497080566634794e-07, + "loss": 0.9495, + "step": 9755 + }, + { + "epoch": 0.8798304549758759, + "grad_norm": 1.6835536022665605, + "learning_rate": 1.4948636250135693e-07, + "loss": 0.9575, + "step": 9756 + }, + { + "epoch": 0.8799206384993462, + "grad_norm": 1.8498269489903474, + "learning_rate": 1.4926482623526249e-07, + "loss": 1.0098, + "step": 9757 + }, + { + "epoch": 0.8800108220228164, + "grad_norm": 1.8290413410234476, + "learning_rate": 1.4904344788409694e-07, + "loss": 0.9489, + "step": 9758 + }, + { + "epoch": 0.8801010055462867, + "grad_norm": 1.7172033807299605, + "learning_rate": 1.4882222746675143e-07, + "loss": 0.9098, + "step": 9759 + }, + { + "epoch": 0.8801911890697569, + "grad_norm": 1.5306765981367507, + "learning_rate": 1.4860116500210018e-07, + "loss": 0.9914, + "step": 9760 + }, + { + "epoch": 0.8802813725932273, + "grad_norm": 2.4766361454263475, + "learning_rate": 1.4838026050900632e-07, + "loss": 1.0145, + "step": 9761 + }, + { + "epoch": 0.8803715561166975, + "grad_norm": 1.4459056564989734, + "learning_rate": 1.481595140063181e-07, + "loss": 0.9932, + "step": 9762 + }, + { + "epoch": 0.8804617396401677, + "grad_norm": 1.6161595270099407, + "learning_rate": 1.4793892551287136e-07, + "loss": 1.0373, + "step": 9763 + }, + { + "epoch": 0.880551923163638, + "grad_norm": 1.8271998485384744, + "learning_rate": 1.4771849504748768e-07, + "loss": 0.8778, + "step": 9764 + }, + { + "epoch": 0.8806421066871083, + "grad_norm": 1.7469413458111138, + "learning_rate": 1.4749822262897517e-07, + "loss": 1.0266, + "step": 9765 + }, + { + "epoch": 0.8807322902105785, + "grad_norm": 1.6213576504181482, + "learning_rate": 1.4727810827612895e-07, + "loss": 0.8899, + "step": 9766 + }, + { + "epoch": 0.8808224737340488, + "grad_norm": 1.8178563390189895, + "learning_rate": 1.470581520077303e-07, + "loss": 0.9585, + "step": 9767 + }, + { + "epoch": 0.8809126572575191, + "grad_norm": 2.0370864354832574, + "learning_rate": 1.4683835384254705e-07, + "loss": 0.9464, + "step": 9768 + }, + { + "epoch": 0.8810028407809893, + "grad_norm": 1.5073073731900037, + "learning_rate": 1.4661871379933376e-07, + "loss": 1.0283, + "step": 9769 + }, + { + "epoch": 0.8810930243044596, + "grad_norm": 1.5358952498033616, + "learning_rate": 1.4639923189683169e-07, + "loss": 0.9939, + "step": 9770 + }, + { + "epoch": 0.8811832078279298, + "grad_norm": 1.6147997639298473, + "learning_rate": 1.461799081537669e-07, + "loss": 0.997, + "step": 9771 + }, + { + "epoch": 0.8812733913514001, + "grad_norm": 2.4192855281540955, + "learning_rate": 1.4596074258885514e-07, + "loss": 1.0129, + "step": 9772 + }, + { + "epoch": 0.8813635748748704, + "grad_norm": 1.6855366199309632, + "learning_rate": 1.4574173522079502e-07, + "loss": 0.9164, + "step": 9773 + }, + { + "epoch": 0.8814537583983406, + "grad_norm": 1.4247729773126991, + "learning_rate": 1.4552288606827513e-07, + "loss": 0.9204, + "step": 9774 + }, + { + "epoch": 0.8815439419218108, + "grad_norm": 1.455023970652281, + "learning_rate": 1.4530419514996761e-07, + "loss": 1.0255, + "step": 9775 + }, + { + "epoch": 0.8816341254452812, + "grad_norm": 0.7132396522918576, + "learning_rate": 1.4508566248453291e-07, + "loss": 0.7902, + "step": 9776 + }, + { + "epoch": 0.8817243089687514, + "grad_norm": 1.6616164304638135, + "learning_rate": 1.448672880906172e-07, + "loss": 0.9397, + "step": 9777 + }, + { + "epoch": 0.8818144924922217, + "grad_norm": 1.4258360969758948, + "learning_rate": 1.4464907198685382e-07, + "loss": 0.9047, + "step": 9778 + }, + { + "epoch": 0.8819046760156919, + "grad_norm": 1.3408354711821784, + "learning_rate": 1.444310141918621e-07, + "loss": 0.9517, + "step": 9779 + }, + { + "epoch": 0.8819948595391622, + "grad_norm": 1.9922442259846653, + "learning_rate": 1.4421311472424735e-07, + "loss": 0.9777, + "step": 9780 + }, + { + "epoch": 0.8820850430626325, + "grad_norm": 1.841512583275051, + "learning_rate": 1.4399537360260273e-07, + "loss": 0.9821, + "step": 9781 + }, + { + "epoch": 0.8821752265861027, + "grad_norm": 1.5833556859630387, + "learning_rate": 1.4377779084550645e-07, + "loss": 0.9702, + "step": 9782 + }, + { + "epoch": 0.8822654101095729, + "grad_norm": 1.3627669886508278, + "learning_rate": 1.4356036647152413e-07, + "loss": 0.9413, + "step": 9783 + }, + { + "epoch": 0.8823555936330433, + "grad_norm": 1.5967921226181023, + "learning_rate": 1.4334310049920785e-07, + "loss": 0.9275, + "step": 9784 + }, + { + "epoch": 0.8824457771565135, + "grad_norm": 0.6135227189649102, + "learning_rate": 1.431259929470956e-07, + "loss": 0.7306, + "step": 9785 + }, + { + "epoch": 0.8825359606799837, + "grad_norm": 1.8112301817889795, + "learning_rate": 1.4290904383371237e-07, + "loss": 1.0238, + "step": 9786 + }, + { + "epoch": 0.882626144203454, + "grad_norm": 2.154329522750973, + "learning_rate": 1.4269225317756961e-07, + "loss": 0.9323, + "step": 9787 + }, + { + "epoch": 0.8827163277269243, + "grad_norm": 1.700606137093627, + "learning_rate": 1.424756209971645e-07, + "loss": 0.9059, + "step": 9788 + }, + { + "epoch": 0.8828065112503946, + "grad_norm": 1.882869355542302, + "learning_rate": 1.4225914731098199e-07, + "loss": 0.9448, + "step": 9789 + }, + { + "epoch": 0.8828966947738648, + "grad_norm": 1.6201667801048214, + "learning_rate": 1.4204283213749248e-07, + "loss": 0.8827, + "step": 9790 + }, + { + "epoch": 0.8829868782973351, + "grad_norm": 1.4395101010603182, + "learning_rate": 1.4182667549515315e-07, + "loss": 1.0056, + "step": 9791 + }, + { + "epoch": 0.8830770618208054, + "grad_norm": 1.5467916909335242, + "learning_rate": 1.4161067740240752e-07, + "loss": 0.9294, + "step": 9792 + }, + { + "epoch": 0.8831672453442756, + "grad_norm": 1.4599378296927554, + "learning_rate": 1.4139483787768614e-07, + "loss": 0.9439, + "step": 9793 + }, + { + "epoch": 0.8832574288677458, + "grad_norm": 1.4672093528932009, + "learning_rate": 1.4117915693940584e-07, + "loss": 0.86, + "step": 9794 + }, + { + "epoch": 0.8833476123912162, + "grad_norm": 1.4211449897508421, + "learning_rate": 1.409636346059684e-07, + "loss": 0.9419, + "step": 9795 + }, + { + "epoch": 0.8834377959146864, + "grad_norm": 1.8397080681209788, + "learning_rate": 1.4074827089576501e-07, + "loss": 0.9782, + "step": 9796 + }, + { + "epoch": 0.8835279794381566, + "grad_norm": 1.7934501184610172, + "learning_rate": 1.4053306582717085e-07, + "loss": 0.9623, + "step": 9797 + }, + { + "epoch": 0.8836181629616269, + "grad_norm": 1.4807903977099774, + "learning_rate": 1.4031801941854827e-07, + "loss": 0.9729, + "step": 9798 + }, + { + "epoch": 0.8837083464850972, + "grad_norm": 1.3901397201243064, + "learning_rate": 1.401031316882466e-07, + "loss": 0.9132, + "step": 9799 + }, + { + "epoch": 0.8837985300085675, + "grad_norm": 1.6728285449972575, + "learning_rate": 1.39888402654601e-07, + "loss": 0.9676, + "step": 9800 + }, + { + "epoch": 0.8838887135320377, + "grad_norm": 0.6435174308956214, + "learning_rate": 1.3967383233593344e-07, + "loss": 0.7885, + "step": 9801 + }, + { + "epoch": 0.8839788970555079, + "grad_norm": 1.8596732918105694, + "learning_rate": 1.3945942075055218e-07, + "loss": 0.9106, + "step": 9802 + }, + { + "epoch": 0.8840690805789783, + "grad_norm": 1.6729565981053685, + "learning_rate": 1.3924516791675212e-07, + "loss": 0.9717, + "step": 9803 + }, + { + "epoch": 0.8841592641024485, + "grad_norm": 1.4143381429700297, + "learning_rate": 1.3903107385281487e-07, + "loss": 0.8615, + "step": 9804 + }, + { + "epoch": 0.8842494476259187, + "grad_norm": 1.5249662114136724, + "learning_rate": 1.3881713857700717e-07, + "loss": 0.9586, + "step": 9805 + }, + { + "epoch": 0.884339631149389, + "grad_norm": 1.6338869055290222, + "learning_rate": 1.3860336210758372e-07, + "loss": 0.9281, + "step": 9806 + }, + { + "epoch": 0.8844298146728593, + "grad_norm": 1.8183302011413358, + "learning_rate": 1.3838974446278506e-07, + "loss": 0.9044, + "step": 9807 + }, + { + "epoch": 0.8845199981963295, + "grad_norm": 1.6071548703668075, + "learning_rate": 1.3817628566083817e-07, + "loss": 0.9261, + "step": 9808 + }, + { + "epoch": 0.8846101817197998, + "grad_norm": 1.3884621849501555, + "learning_rate": 1.3796298571995712e-07, + "loss": 0.9593, + "step": 9809 + }, + { + "epoch": 0.88470036524327, + "grad_norm": 1.6168296528618256, + "learning_rate": 1.377498446583405e-07, + "loss": 0.8414, + "step": 9810 + }, + { + "epoch": 0.8847905487667403, + "grad_norm": 1.4022461158534063, + "learning_rate": 1.3753686249417596e-07, + "loss": 0.9173, + "step": 9811 + }, + { + "epoch": 0.8848807322902106, + "grad_norm": 1.6331507136309336, + "learning_rate": 1.373240392456354e-07, + "loss": 0.8994, + "step": 9812 + }, + { + "epoch": 0.8849709158136808, + "grad_norm": 1.5937907643808238, + "learning_rate": 1.37111374930879e-07, + "loss": 0.8866, + "step": 9813 + }, + { + "epoch": 0.885061099337151, + "grad_norm": 1.4982682403976686, + "learning_rate": 1.3689886956805176e-07, + "loss": 0.8775, + "step": 9814 + }, + { + "epoch": 0.8851512828606214, + "grad_norm": 1.4267589822355746, + "learning_rate": 1.3668652317528585e-07, + "loss": 0.9346, + "step": 9815 + }, + { + "epoch": 0.8852414663840916, + "grad_norm": 1.5662252649788149, + "learning_rate": 1.3647433577070012e-07, + "loss": 0.9665, + "step": 9816 + }, + { + "epoch": 0.8853316499075619, + "grad_norm": 0.6537089880921926, + "learning_rate": 1.3626230737239942e-07, + "loss": 0.7507, + "step": 9817 + }, + { + "epoch": 0.8854218334310322, + "grad_norm": 0.7189388441703174, + "learning_rate": 1.3605043799847527e-07, + "loss": 0.7584, + "step": 9818 + }, + { + "epoch": 0.8855120169545024, + "grad_norm": 1.384030210041249, + "learning_rate": 1.3583872766700567e-07, + "loss": 0.8836, + "step": 9819 + }, + { + "epoch": 0.8856022004779727, + "grad_norm": 8.390152163120314, + "learning_rate": 1.3562717639605437e-07, + "loss": 0.927, + "step": 9820 + }, + { + "epoch": 0.8856923840014429, + "grad_norm": 1.6850196128631982, + "learning_rate": 1.3541578420367229e-07, + "loss": 0.9163, + "step": 9821 + }, + { + "epoch": 0.8857825675249132, + "grad_norm": 1.7793557652464544, + "learning_rate": 1.3520455110789697e-07, + "loss": 0.9626, + "step": 9822 + }, + { + "epoch": 0.8858727510483835, + "grad_norm": 1.4697219392317233, + "learning_rate": 1.3499347712675158e-07, + "loss": 0.9224, + "step": 9823 + }, + { + "epoch": 0.8859629345718537, + "grad_norm": 1.4630126087378077, + "learning_rate": 1.3478256227824635e-07, + "loss": 0.9078, + "step": 9824 + }, + { + "epoch": 0.8860531180953239, + "grad_norm": 1.4961274544128995, + "learning_rate": 1.3457180658037759e-07, + "loss": 0.8552, + "step": 9825 + }, + { + "epoch": 0.8861433016187943, + "grad_norm": 1.4167772231252276, + "learning_rate": 1.3436121005112843e-07, + "loss": 0.9772, + "step": 9826 + }, + { + "epoch": 0.8862334851422645, + "grad_norm": 1.565480479500977, + "learning_rate": 1.3415077270846719e-07, + "loss": 0.9139, + "step": 9827 + }, + { + "epoch": 0.8863236686657348, + "grad_norm": 1.8752134837368846, + "learning_rate": 1.3394049457035105e-07, + "loss": 1.0591, + "step": 9828 + }, + { + "epoch": 0.886413852189205, + "grad_norm": 1.5237701052934496, + "learning_rate": 1.3373037565472034e-07, + "loss": 0.9533, + "step": 9829 + }, + { + "epoch": 0.8865040357126753, + "grad_norm": 1.497411879301481, + "learning_rate": 1.3352041597950537e-07, + "loss": 0.924, + "step": 9830 + }, + { + "epoch": 0.8865942192361456, + "grad_norm": 1.5321366993097267, + "learning_rate": 1.333106155626196e-07, + "loss": 0.9327, + "step": 9831 + }, + { + "epoch": 0.8866844027596158, + "grad_norm": 1.4251706840956075, + "learning_rate": 1.331009744219651e-07, + "loss": 0.8865, + "step": 9832 + }, + { + "epoch": 0.886774586283086, + "grad_norm": 0.6922304015286983, + "learning_rate": 1.3289149257542943e-07, + "loss": 0.8143, + "step": 9833 + }, + { + "epoch": 0.8868647698065564, + "grad_norm": 2.0279669419402624, + "learning_rate": 1.3268217004088666e-07, + "loss": 0.848, + "step": 9834 + }, + { + "epoch": 0.8869549533300266, + "grad_norm": 1.607900592255378, + "learning_rate": 1.3247300683619788e-07, + "loss": 0.8349, + "step": 9835 + }, + { + "epoch": 0.8870451368534968, + "grad_norm": 1.7404868545364567, + "learning_rate": 1.3226400297920903e-07, + "loss": 0.9226, + "step": 9836 + }, + { + "epoch": 0.8871353203769671, + "grad_norm": 1.5805017683772928, + "learning_rate": 1.3205515848775428e-07, + "loss": 0.9225, + "step": 9837 + }, + { + "epoch": 0.8872255039004374, + "grad_norm": 1.5608851671276482, + "learning_rate": 1.3184647337965316e-07, + "loss": 0.8829, + "step": 9838 + }, + { + "epoch": 0.8873156874239077, + "grad_norm": 1.36199711434651, + "learning_rate": 1.3163794767271163e-07, + "loss": 0.8941, + "step": 9839 + }, + { + "epoch": 0.8874058709473779, + "grad_norm": 1.4372602345740375, + "learning_rate": 1.314295813847226e-07, + "loss": 0.9188, + "step": 9840 + }, + { + "epoch": 0.8874960544708482, + "grad_norm": 1.755803702450968, + "learning_rate": 1.3122137453346515e-07, + "loss": 1.0452, + "step": 9841 + }, + { + "epoch": 0.8875862379943185, + "grad_norm": 1.623255159713838, + "learning_rate": 1.3101332713670376e-07, + "loss": 1.0018, + "step": 9842 + }, + { + "epoch": 0.8876764215177887, + "grad_norm": 0.6903374163792226, + "learning_rate": 1.3080543921219133e-07, + "loss": 0.808, + "step": 9843 + }, + { + "epoch": 0.8877666050412589, + "grad_norm": 1.612972497836859, + "learning_rate": 1.3059771077766478e-07, + "loss": 0.9399, + "step": 9844 + }, + { + "epoch": 0.8878567885647293, + "grad_norm": 1.7722667577217355, + "learning_rate": 1.3039014185085018e-07, + "loss": 0.9832, + "step": 9845 + }, + { + "epoch": 0.8879469720881995, + "grad_norm": 1.805482763469591, + "learning_rate": 1.301827324494571e-07, + "loss": 0.952, + "step": 9846 + }, + { + "epoch": 0.8880371556116697, + "grad_norm": 1.9311361087931669, + "learning_rate": 1.2997548259118342e-07, + "loss": 0.9876, + "step": 9847 + }, + { + "epoch": 0.88812733913514, + "grad_norm": 1.607653020691833, + "learning_rate": 1.2976839229371272e-07, + "loss": 0.9768, + "step": 9848 + }, + { + "epoch": 0.8882175226586103, + "grad_norm": 0.6552229519311467, + "learning_rate": 1.2956146157471515e-07, + "loss": 0.7968, + "step": 9849 + }, + { + "epoch": 0.8883077061820805, + "grad_norm": 1.2869401973977508, + "learning_rate": 1.2935469045184745e-07, + "loss": 0.9659, + "step": 9850 + }, + { + "epoch": 0.8883978897055508, + "grad_norm": 1.7016047420810465, + "learning_rate": 1.291480789427517e-07, + "loss": 0.9186, + "step": 9851 + }, + { + "epoch": 0.888488073229021, + "grad_norm": 1.5078746407309276, + "learning_rate": 1.2894162706505807e-07, + "loss": 0.9872, + "step": 9852 + }, + { + "epoch": 0.8885782567524914, + "grad_norm": 1.7517807619495087, + "learning_rate": 1.2873533483638155e-07, + "loss": 0.9398, + "step": 9853 + }, + { + "epoch": 0.8886684402759616, + "grad_norm": 1.5581659097882188, + "learning_rate": 1.285292022743243e-07, + "loss": 0.8934, + "step": 9854 + }, + { + "epoch": 0.8887586237994318, + "grad_norm": 1.5763696407014969, + "learning_rate": 1.2832322939647467e-07, + "loss": 0.8912, + "step": 9855 + }, + { + "epoch": 0.8888488073229021, + "grad_norm": 1.544006462261532, + "learning_rate": 1.281174162204075e-07, + "loss": 0.9268, + "step": 9856 + }, + { + "epoch": 0.8889389908463724, + "grad_norm": 1.3373469006625738, + "learning_rate": 1.2791176276368366e-07, + "loss": 0.8259, + "step": 9857 + }, + { + "epoch": 0.8890291743698426, + "grad_norm": 1.3647498934288562, + "learning_rate": 1.2770626904385128e-07, + "loss": 0.9587, + "step": 9858 + }, + { + "epoch": 0.8891193578933129, + "grad_norm": 1.553554869940948, + "learning_rate": 1.2750093507844306e-07, + "loss": 0.8087, + "step": 9859 + }, + { + "epoch": 0.8892095414167831, + "grad_norm": 1.2377939318290139, + "learning_rate": 1.272957608849805e-07, + "loss": 0.9846, + "step": 9860 + }, + { + "epoch": 0.8892997249402534, + "grad_norm": 1.3396596976781794, + "learning_rate": 1.270907464809694e-07, + "loss": 0.9233, + "step": 9861 + }, + { + "epoch": 0.8893899084637237, + "grad_norm": 1.36705182083459, + "learning_rate": 1.2688589188390285e-07, + "loss": 0.9478, + "step": 9862 + }, + { + "epoch": 0.8894800919871939, + "grad_norm": 1.3931356761141653, + "learning_rate": 1.2668119711126023e-07, + "loss": 0.9552, + "step": 9863 + }, + { + "epoch": 0.8895702755106643, + "grad_norm": 1.432336489814143, + "learning_rate": 1.2647666218050735e-07, + "loss": 0.9487, + "step": 9864 + }, + { + "epoch": 0.8896604590341345, + "grad_norm": 1.4865046432684352, + "learning_rate": 1.2627228710909643e-07, + "loss": 0.9074, + "step": 9865 + }, + { + "epoch": 0.8897506425576047, + "grad_norm": 1.4755558841454728, + "learning_rate": 1.260680719144649e-07, + "loss": 0.9115, + "step": 9866 + }, + { + "epoch": 0.889840826081075, + "grad_norm": 2.546713326594256, + "learning_rate": 1.2586401661403877e-07, + "loss": 0.872, + "step": 9867 + }, + { + "epoch": 0.8899310096045453, + "grad_norm": 3.062131119913782, + "learning_rate": 1.2566012122522817e-07, + "loss": 0.9166, + "step": 9868 + }, + { + "epoch": 0.8900211931280155, + "grad_norm": 1.7799308309081516, + "learning_rate": 1.254563857654316e-07, + "loss": 0.9198, + "step": 9869 + }, + { + "epoch": 0.8901113766514858, + "grad_norm": 1.7798416861307071, + "learning_rate": 1.2525281025203205e-07, + "loss": 0.8326, + "step": 9870 + }, + { + "epoch": 0.890201560174956, + "grad_norm": 1.7583140018117598, + "learning_rate": 1.2504939470240006e-07, + "loss": 0.9284, + "step": 9871 + }, + { + "epoch": 0.8902917436984263, + "grad_norm": 1.6017015350282997, + "learning_rate": 1.2484613913389196e-07, + "loss": 1.0133, + "step": 9872 + }, + { + "epoch": 0.8903819272218966, + "grad_norm": 1.5075368678707337, + "learning_rate": 1.2464304356385057e-07, + "loss": 0.8609, + "step": 9873 + }, + { + "epoch": 0.8904721107453668, + "grad_norm": 2.2073871696901843, + "learning_rate": 1.2444010800960558e-07, + "loss": 1.0029, + "step": 9874 + }, + { + "epoch": 0.890562294268837, + "grad_norm": 1.583028295479759, + "learning_rate": 1.2423733248847267e-07, + "loss": 0.9309, + "step": 9875 + }, + { + "epoch": 0.8906524777923074, + "grad_norm": 1.5706758600661084, + "learning_rate": 1.2403471701775293e-07, + "loss": 0.8881, + "step": 9876 + }, + { + "epoch": 0.8907426613157776, + "grad_norm": 1.6089841497670696, + "learning_rate": 1.2383226161473515e-07, + "loss": 0.9858, + "step": 9877 + }, + { + "epoch": 0.8908328448392479, + "grad_norm": 1.6709972466081435, + "learning_rate": 1.2362996629669376e-07, + "loss": 1.0221, + "step": 9878 + }, + { + "epoch": 0.8909230283627181, + "grad_norm": 1.4492153774335337, + "learning_rate": 1.2342783108089007e-07, + "loss": 0.9573, + "step": 9879 + }, + { + "epoch": 0.8910132118861884, + "grad_norm": 1.6853711864369159, + "learning_rate": 1.2322585598457135e-07, + "loss": 1.0084, + "step": 9880 + }, + { + "epoch": 0.8911033954096587, + "grad_norm": 1.2410134099040617, + "learning_rate": 1.2302404102497034e-07, + "loss": 0.92, + "step": 9881 + }, + { + "epoch": 0.8911935789331289, + "grad_norm": 1.4562054050149822, + "learning_rate": 1.228223862193083e-07, + "loss": 0.8771, + "step": 9882 + }, + { + "epoch": 0.8912837624565991, + "grad_norm": 1.9473433713400674, + "learning_rate": 1.2262089158479038e-07, + "loss": 0.9578, + "step": 9883 + }, + { + "epoch": 0.8913739459800695, + "grad_norm": 1.6066438036330017, + "learning_rate": 1.2241955713861042e-07, + "loss": 0.9892, + "step": 9884 + }, + { + "epoch": 0.8914641295035397, + "grad_norm": 1.5545899633741473, + "learning_rate": 1.222183828979464e-07, + "loss": 0.8852, + "step": 9885 + }, + { + "epoch": 0.8915543130270099, + "grad_norm": 1.3277401983768065, + "learning_rate": 1.2201736887996372e-07, + "loss": 0.9465, + "step": 9886 + }, + { + "epoch": 0.8916444965504803, + "grad_norm": 1.6576379682331204, + "learning_rate": 1.2181651510181444e-07, + "loss": 1.024, + "step": 9887 + }, + { + "epoch": 0.8917346800739505, + "grad_norm": 1.5474658954911187, + "learning_rate": 1.2161582158063622e-07, + "loss": 0.9736, + "step": 9888 + }, + { + "epoch": 0.8918248635974207, + "grad_norm": 1.5249562053958285, + "learning_rate": 1.214152883335533e-07, + "loss": 0.8547, + "step": 9889 + }, + { + "epoch": 0.891915047120891, + "grad_norm": 1.779269941609751, + "learning_rate": 1.2121491537767648e-07, + "loss": 0.8972, + "step": 9890 + }, + { + "epoch": 0.8920052306443613, + "grad_norm": 1.4211853366927443, + "learning_rate": 1.2101470273010294e-07, + "loss": 0.9819, + "step": 9891 + }, + { + "epoch": 0.8920954141678316, + "grad_norm": 1.5842795138205181, + "learning_rate": 1.2081465040791528e-07, + "loss": 0.9214, + "step": 9892 + }, + { + "epoch": 0.8921855976913018, + "grad_norm": 1.6047920972987024, + "learning_rate": 1.2061475842818335e-07, + "loss": 0.909, + "step": 9893 + }, + { + "epoch": 0.892275781214772, + "grad_norm": 1.4507549195167582, + "learning_rate": 1.2041502680796313e-07, + "loss": 0.9217, + "step": 9894 + }, + { + "epoch": 0.8923659647382424, + "grad_norm": 1.3659515461824587, + "learning_rate": 1.2021545556429648e-07, + "loss": 0.9506, + "step": 9895 + }, + { + "epoch": 0.8924561482617126, + "grad_norm": 1.29896368022064, + "learning_rate": 1.2001604471421245e-07, + "loss": 0.9431, + "step": 9896 + }, + { + "epoch": 0.8925463317851828, + "grad_norm": 1.4032371034422044, + "learning_rate": 1.1981679427472567e-07, + "loss": 0.9948, + "step": 9897 + }, + { + "epoch": 0.8926365153086531, + "grad_norm": 1.5887316814239913, + "learning_rate": 1.196177042628368e-07, + "loss": 0.9947, + "step": 9898 + }, + { + "epoch": 0.8927266988321234, + "grad_norm": 1.8844747842714509, + "learning_rate": 1.194187746955344e-07, + "loss": 0.9526, + "step": 9899 + }, + { + "epoch": 0.8928168823555936, + "grad_norm": 1.4754774355428213, + "learning_rate": 1.1922000558979094e-07, + "loss": 0.9203, + "step": 9900 + }, + { + "epoch": 0.8929070658790639, + "grad_norm": 1.3704273745539006, + "learning_rate": 1.1902139696256752e-07, + "loss": 0.9743, + "step": 9901 + }, + { + "epoch": 0.8929972494025341, + "grad_norm": 1.6374455479311596, + "learning_rate": 1.188229488308099e-07, + "loss": 0.8834, + "step": 9902 + }, + { + "epoch": 0.8930874329260045, + "grad_norm": 0.5880381503622952, + "learning_rate": 1.1862466121145098e-07, + "loss": 0.7581, + "step": 9903 + }, + { + "epoch": 0.8931776164494747, + "grad_norm": 1.6833593643492277, + "learning_rate": 1.184265341214099e-07, + "loss": 0.8886, + "step": 9904 + }, + { + "epoch": 0.8932677999729449, + "grad_norm": 1.499089123408163, + "learning_rate": 1.182285675775918e-07, + "loss": 0.9703, + "step": 9905 + }, + { + "epoch": 0.8933579834964152, + "grad_norm": 1.798658597001019, + "learning_rate": 1.1803076159688851e-07, + "loss": 0.9305, + "step": 9906 + }, + { + "epoch": 0.8934481670198855, + "grad_norm": 1.6299563894698184, + "learning_rate": 1.1783311619617741e-07, + "loss": 0.9814, + "step": 9907 + }, + { + "epoch": 0.8935383505433557, + "grad_norm": 1.5997191450860702, + "learning_rate": 1.1763563139232302e-07, + "loss": 0.9164, + "step": 9908 + }, + { + "epoch": 0.893628534066826, + "grad_norm": 1.6564205009942046, + "learning_rate": 1.1743830720217562e-07, + "loss": 0.8835, + "step": 9909 + }, + { + "epoch": 0.8937187175902963, + "grad_norm": 1.9154784901087603, + "learning_rate": 1.1724114364257243e-07, + "loss": 0.9221, + "step": 9910 + }, + { + "epoch": 0.8938089011137665, + "grad_norm": 1.5069613253130885, + "learning_rate": 1.1704414073033619e-07, + "loss": 0.9515, + "step": 9911 + }, + { + "epoch": 0.8938990846372368, + "grad_norm": 1.6852795155680442, + "learning_rate": 1.1684729848227636e-07, + "loss": 0.9861, + "step": 9912 + }, + { + "epoch": 0.893989268160707, + "grad_norm": 1.4732507775481207, + "learning_rate": 1.1665061691518884e-07, + "loss": 1.0011, + "step": 9913 + }, + { + "epoch": 0.8940794516841774, + "grad_norm": 1.5093690417944157, + "learning_rate": 1.1645409604585532e-07, + "loss": 0.9435, + "step": 9914 + }, + { + "epoch": 0.8941696352076476, + "grad_norm": 1.5496509958942135, + "learning_rate": 1.162577358910437e-07, + "loss": 0.9506, + "step": 9915 + }, + { + "epoch": 0.8942598187311178, + "grad_norm": 1.5433524511785894, + "learning_rate": 1.160615364675095e-07, + "loss": 0.9231, + "step": 9916 + }, + { + "epoch": 0.894350002254588, + "grad_norm": 0.6312560496654124, + "learning_rate": 1.1586549779199262e-07, + "loss": 0.7605, + "step": 9917 + }, + { + "epoch": 0.8944401857780584, + "grad_norm": 1.4022402924439505, + "learning_rate": 1.1566961988122037e-07, + "loss": 1.007, + "step": 9918 + }, + { + "epoch": 0.8945303693015286, + "grad_norm": 1.4833273955379898, + "learning_rate": 1.1547390275190627e-07, + "loss": 0.9092, + "step": 9919 + }, + { + "epoch": 0.8946205528249989, + "grad_norm": 1.6282372141294628, + "learning_rate": 1.1527834642075007e-07, + "loss": 0.8837, + "step": 9920 + }, + { + "epoch": 0.8947107363484691, + "grad_norm": 1.8912254830111987, + "learning_rate": 1.1508295090443797e-07, + "loss": 0.9294, + "step": 9921 + }, + { + "epoch": 0.8948009198719394, + "grad_norm": 1.372014749834167, + "learning_rate": 1.148877162196411e-07, + "loss": 0.8932, + "step": 9922 + }, + { + "epoch": 0.8948911033954097, + "grad_norm": 1.9995848105535703, + "learning_rate": 1.1469264238301924e-07, + "loss": 0.88, + "step": 9923 + }, + { + "epoch": 0.8949812869188799, + "grad_norm": 2.0353630092684427, + "learning_rate": 1.1449772941121638e-07, + "loss": 0.974, + "step": 9924 + }, + { + "epoch": 0.8950714704423501, + "grad_norm": 1.648681188713519, + "learning_rate": 1.1430297732086369e-07, + "loss": 0.9156, + "step": 9925 + }, + { + "epoch": 0.8951616539658205, + "grad_norm": 1.5425000215156344, + "learning_rate": 1.1410838612857876e-07, + "loss": 0.9336, + "step": 9926 + }, + { + "epoch": 0.8952518374892907, + "grad_norm": 1.7295252033789288, + "learning_rate": 1.1391395585096497e-07, + "loss": 1.043, + "step": 9927 + }, + { + "epoch": 0.895342021012761, + "grad_norm": 2.0845606685455733, + "learning_rate": 1.1371968650461216e-07, + "loss": 0.8763, + "step": 9928 + }, + { + "epoch": 0.8954322045362312, + "grad_norm": 0.6944474614925543, + "learning_rate": 1.1352557810609687e-07, + "loss": 0.7737, + "step": 9929 + }, + { + "epoch": 0.8955223880597015, + "grad_norm": 2.2036232892076915, + "learning_rate": 1.1333163067198048e-07, + "loss": 0.93, + "step": 9930 + }, + { + "epoch": 0.8956125715831718, + "grad_norm": 1.4502490717624008, + "learning_rate": 1.1313784421881311e-07, + "loss": 1.0118, + "step": 9931 + }, + { + "epoch": 0.895702755106642, + "grad_norm": 1.6036392517384284, + "learning_rate": 1.1294421876312865e-07, + "loss": 0.9853, + "step": 9932 + }, + { + "epoch": 0.8957929386301122, + "grad_norm": 1.9111867048162783, + "learning_rate": 1.1275075432144831e-07, + "loss": 0.9218, + "step": 9933 + }, + { + "epoch": 0.8958831221535826, + "grad_norm": 1.7457584340758874, + "learning_rate": 1.1255745091028002e-07, + "loss": 0.9578, + "step": 9934 + }, + { + "epoch": 0.8959733056770528, + "grad_norm": 1.551192068267776, + "learning_rate": 1.1236430854611723e-07, + "loss": 0.9553, + "step": 9935 + }, + { + "epoch": 0.896063489200523, + "grad_norm": 1.5958460320082846, + "learning_rate": 1.1217132724544032e-07, + "loss": 0.9069, + "step": 9936 + }, + { + "epoch": 0.8961536727239934, + "grad_norm": 1.7275985440137893, + "learning_rate": 1.1197850702471434e-07, + "loss": 0.9984, + "step": 9937 + }, + { + "epoch": 0.8962438562474636, + "grad_norm": 1.4610913037616289, + "learning_rate": 1.1178584790039348e-07, + "loss": 1.0102, + "step": 9938 + }, + { + "epoch": 0.8963340397709338, + "grad_norm": 1.446895248238448, + "learning_rate": 1.1159334988891478e-07, + "loss": 0.8864, + "step": 9939 + }, + { + "epoch": 0.8964242232944041, + "grad_norm": 1.7148941484487932, + "learning_rate": 1.1140101300670446e-07, + "loss": 0.9199, + "step": 9940 + }, + { + "epoch": 0.8965144068178744, + "grad_norm": 1.4796778194187359, + "learning_rate": 1.1120883727017338e-07, + "loss": 0.811, + "step": 9941 + }, + { + "epoch": 0.8966045903413447, + "grad_norm": 2.591066834029841, + "learning_rate": 1.1101682269571889e-07, + "loss": 0.9961, + "step": 9942 + }, + { + "epoch": 0.8966947738648149, + "grad_norm": 1.607257598232407, + "learning_rate": 1.1082496929972473e-07, + "loss": 0.9228, + "step": 9943 + }, + { + "epoch": 0.8967849573882851, + "grad_norm": 1.5279124730391316, + "learning_rate": 1.1063327709856096e-07, + "loss": 0.9158, + "step": 9944 + }, + { + "epoch": 0.8968751409117555, + "grad_norm": 1.7572818887736648, + "learning_rate": 1.1044174610858403e-07, + "loss": 0.8243, + "step": 9945 + }, + { + "epoch": 0.8969653244352257, + "grad_norm": 1.5688491759543681, + "learning_rate": 1.1025037634613643e-07, + "loss": 0.9029, + "step": 9946 + }, + { + "epoch": 0.8970555079586959, + "grad_norm": 1.4286517784819175, + "learning_rate": 1.1005916782754643e-07, + "loss": 0.9694, + "step": 9947 + }, + { + "epoch": 0.8971456914821662, + "grad_norm": 1.757358678939357, + "learning_rate": 1.0986812056912919e-07, + "loss": 0.9307, + "step": 9948 + }, + { + "epoch": 0.8972358750056365, + "grad_norm": 1.447206235997865, + "learning_rate": 1.0967723458718613e-07, + "loss": 0.8199, + "step": 9949 + }, + { + "epoch": 0.8973260585291067, + "grad_norm": 2.2010402690872786, + "learning_rate": 1.0948650989800445e-07, + "loss": 0.9121, + "step": 9950 + }, + { + "epoch": 0.897416242052577, + "grad_norm": 1.6210240360574775, + "learning_rate": 1.0929594651785823e-07, + "loss": 0.9225, + "step": 9951 + }, + { + "epoch": 0.8975064255760472, + "grad_norm": 1.543559519023977, + "learning_rate": 1.0910554446300646e-07, + "loss": 0.9956, + "step": 9952 + }, + { + "epoch": 0.8975966090995176, + "grad_norm": 1.5057091941111453, + "learning_rate": 1.089153037496966e-07, + "loss": 0.8461, + "step": 9953 + }, + { + "epoch": 0.8976867926229878, + "grad_norm": 2.598262767685868, + "learning_rate": 1.0872522439415966e-07, + "loss": 0.9635, + "step": 9954 + }, + { + "epoch": 0.897776976146458, + "grad_norm": 1.75764749388303, + "learning_rate": 1.0853530641261554e-07, + "loss": 0.889, + "step": 9955 + }, + { + "epoch": 0.8978671596699282, + "grad_norm": 1.4326281712896936, + "learning_rate": 1.083455498212682e-07, + "loss": 0.9591, + "step": 9956 + }, + { + "epoch": 0.8979573431933986, + "grad_norm": 1.783085930974893, + "learning_rate": 1.0815595463630911e-07, + "loss": 0.9828, + "step": 9957 + }, + { + "epoch": 0.8980475267168688, + "grad_norm": 1.6952821931019795, + "learning_rate": 1.0796652087391556e-07, + "loss": 0.8126, + "step": 9958 + }, + { + "epoch": 0.8981377102403391, + "grad_norm": 1.527169220067374, + "learning_rate": 1.0777724855025083e-07, + "loss": 1.0385, + "step": 9959 + }, + { + "epoch": 0.8982278937638094, + "grad_norm": 1.5253751512199314, + "learning_rate": 1.075881376814649e-07, + "loss": 0.9417, + "step": 9960 + }, + { + "epoch": 0.8983180772872796, + "grad_norm": 1.6096666682855498, + "learning_rate": 1.0739918828369377e-07, + "loss": 0.9982, + "step": 9961 + }, + { + "epoch": 0.8984082608107499, + "grad_norm": 1.5730750233680169, + "learning_rate": 1.0721040037305983e-07, + "loss": 0.9502, + "step": 9962 + }, + { + "epoch": 0.8984984443342201, + "grad_norm": 1.5466420932714056, + "learning_rate": 1.0702177396567114e-07, + "loss": 0.8971, + "step": 9963 + }, + { + "epoch": 0.8985886278576904, + "grad_norm": 0.6307091785122916, + "learning_rate": 1.0683330907762233e-07, + "loss": 0.7644, + "step": 9964 + }, + { + "epoch": 0.8986788113811607, + "grad_norm": 1.4752554785389642, + "learning_rate": 1.0664500572499435e-07, + "loss": 0.875, + "step": 9965 + }, + { + "epoch": 0.8987689949046309, + "grad_norm": 2.2836196297618914, + "learning_rate": 1.0645686392385455e-07, + "loss": 0.8859, + "step": 9966 + }, + { + "epoch": 0.8988591784281011, + "grad_norm": 1.3412641108926915, + "learning_rate": 1.0626888369025588e-07, + "loss": 0.8656, + "step": 9967 + }, + { + "epoch": 0.8989493619515715, + "grad_norm": 1.391704568898806, + "learning_rate": 1.0608106504023817e-07, + "loss": 0.9721, + "step": 9968 + }, + { + "epoch": 0.8990395454750417, + "grad_norm": 1.2903884541612853, + "learning_rate": 1.0589340798982637e-07, + "loss": 0.9326, + "step": 9969 + }, + { + "epoch": 0.899129728998512, + "grad_norm": 1.6614146052458705, + "learning_rate": 1.057059125550337e-07, + "loss": 0.887, + "step": 9970 + }, + { + "epoch": 0.8992199125219822, + "grad_norm": 1.397349611082585, + "learning_rate": 1.0551857875185732e-07, + "loss": 0.9535, + "step": 9971 + }, + { + "epoch": 0.8993100960454525, + "grad_norm": 1.3031693723154252, + "learning_rate": 1.0533140659628181e-07, + "loss": 0.8292, + "step": 9972 + }, + { + "epoch": 0.8994002795689228, + "grad_norm": 1.4894981715615925, + "learning_rate": 1.0514439610427772e-07, + "loss": 1.0019, + "step": 9973 + }, + { + "epoch": 0.899490463092393, + "grad_norm": 1.4822285441909273, + "learning_rate": 1.0495754729180206e-07, + "loss": 0.9732, + "step": 9974 + }, + { + "epoch": 0.8995806466158632, + "grad_norm": 1.4045101422968644, + "learning_rate": 1.0477086017479741e-07, + "loss": 1.0443, + "step": 9975 + }, + { + "epoch": 0.8996708301393336, + "grad_norm": 2.401509755832951, + "learning_rate": 1.0458433476919327e-07, + "loss": 0.9043, + "step": 9976 + }, + { + "epoch": 0.8997610136628038, + "grad_norm": 1.4632904362021033, + "learning_rate": 1.0439797109090509e-07, + "loss": 0.9717, + "step": 9977 + }, + { + "epoch": 0.899851197186274, + "grad_norm": 1.653070817430822, + "learning_rate": 1.0421176915583396e-07, + "loss": 0.9156, + "step": 9978 + }, + { + "epoch": 0.8999413807097443, + "grad_norm": 1.6538997024745294, + "learning_rate": 1.0402572897986828e-07, + "loss": 0.9745, + "step": 9979 + }, + { + "epoch": 0.9000315642332146, + "grad_norm": 0.6073576227777028, + "learning_rate": 1.0383985057888134e-07, + "loss": 0.7695, + "step": 9980 + }, + { + "epoch": 0.9001217477566849, + "grad_norm": 1.784378383118851, + "learning_rate": 1.036541339687338e-07, + "loss": 0.8499, + "step": 9981 + }, + { + "epoch": 0.9002119312801551, + "grad_norm": 0.6609594887035976, + "learning_rate": 1.0346857916527186e-07, + "loss": 0.7393, + "step": 9982 + }, + { + "epoch": 0.9003021148036254, + "grad_norm": 1.5659891462668527, + "learning_rate": 1.0328318618432819e-07, + "loss": 0.9421, + "step": 9983 + }, + { + "epoch": 0.9003922983270957, + "grad_norm": 2.088833060037077, + "learning_rate": 1.0309795504172148e-07, + "loss": 1.0034, + "step": 9984 + }, + { + "epoch": 0.9004824818505659, + "grad_norm": 1.5053031477935606, + "learning_rate": 1.0291288575325685e-07, + "loss": 0.9194, + "step": 9985 + }, + { + "epoch": 0.9005726653740361, + "grad_norm": 1.61202510671729, + "learning_rate": 1.0272797833472502e-07, + "loss": 0.9048, + "step": 9986 + }, + { + "epoch": 0.9006628488975065, + "grad_norm": 2.374966872134808, + "learning_rate": 1.0254323280190335e-07, + "loss": 0.7492, + "step": 9987 + }, + { + "epoch": 0.9007530324209767, + "grad_norm": 2.394441637163909, + "learning_rate": 1.023586491705557e-07, + "loss": 0.9208, + "step": 9988 + }, + { + "epoch": 0.9008432159444469, + "grad_norm": 1.6051964447683644, + "learning_rate": 1.0217422745643145e-07, + "loss": 0.8797, + "step": 9989 + }, + { + "epoch": 0.9009333994679172, + "grad_norm": 1.5341564543365342, + "learning_rate": 1.0198996767526691e-07, + "loss": 0.8679, + "step": 9990 + }, + { + "epoch": 0.9010235829913875, + "grad_norm": 1.6344324134676618, + "learning_rate": 1.018058698427835e-07, + "loss": 0.924, + "step": 9991 + }, + { + "epoch": 0.9011137665148578, + "grad_norm": 1.3095250246323071, + "learning_rate": 1.0162193397469021e-07, + "loss": 0.9902, + "step": 9992 + }, + { + "epoch": 0.901203950038328, + "grad_norm": 1.7070978705982343, + "learning_rate": 1.0143816008668049e-07, + "loss": 0.8786, + "step": 9993 + }, + { + "epoch": 0.9012941335617982, + "grad_norm": 0.658970529787326, + "learning_rate": 1.0125454819443624e-07, + "loss": 0.7875, + "step": 9994 + }, + { + "epoch": 0.9013843170852686, + "grad_norm": 1.266525983945226, + "learning_rate": 1.0107109831362315e-07, + "loss": 0.8491, + "step": 9995 + }, + { + "epoch": 0.9014745006087388, + "grad_norm": 1.5513235532427474, + "learning_rate": 1.0088781045989447e-07, + "loss": 0.9515, + "step": 9996 + }, + { + "epoch": 0.901564684132209, + "grad_norm": 1.4685739452277944, + "learning_rate": 1.0070468464888926e-07, + "loss": 0.9884, + "step": 9997 + }, + { + "epoch": 0.9016548676556793, + "grad_norm": 1.571004155362933, + "learning_rate": 1.0052172089623324e-07, + "loss": 0.9609, + "step": 9998 + }, + { + "epoch": 0.9017450511791496, + "grad_norm": 1.535412793648475, + "learning_rate": 1.0033891921753746e-07, + "loss": 0.9811, + "step": 9999 + }, + { + "epoch": 0.9018352347026198, + "grad_norm": 1.563796078058719, + "learning_rate": 1.0015627962839968e-07, + "loss": 0.8558, + "step": 10000 + }, + { + "epoch": 0.9019254182260901, + "grad_norm": 1.4669178837869719, + "learning_rate": 9.99738021444041e-08, + "loss": 0.9161, + "step": 10001 + }, + { + "epoch": 0.9020156017495603, + "grad_norm": 1.4127894619770105, + "learning_rate": 9.979148678112003e-08, + "loss": 0.8752, + "step": 10002 + }, + { + "epoch": 0.9021057852730306, + "grad_norm": 1.9745643376648634, + "learning_rate": 9.960933355410417e-08, + "loss": 1.0477, + "step": 10003 + }, + { + "epoch": 0.9021959687965009, + "grad_norm": 1.453162900368897, + "learning_rate": 9.942734247889828e-08, + "loss": 0.8573, + "step": 10004 + }, + { + "epoch": 0.9022861523199711, + "grad_norm": 1.4519239815287948, + "learning_rate": 9.92455135710315e-08, + "loss": 0.9487, + "step": 10005 + }, + { + "epoch": 0.9023763358434415, + "grad_norm": 1.6788087897328066, + "learning_rate": 9.906384684601787e-08, + "loss": 0.8824, + "step": 10006 + }, + { + "epoch": 0.9024665193669117, + "grad_norm": 1.4914763188875129, + "learning_rate": 9.8882342319359e-08, + "loss": 0.9878, + "step": 10007 + }, + { + "epoch": 0.9025567028903819, + "grad_norm": 1.4045448137401388, + "learning_rate": 9.870100000654048e-08, + "loss": 0.8895, + "step": 10008 + }, + { + "epoch": 0.9026468864138522, + "grad_norm": 1.2870546366928801, + "learning_rate": 9.851981992303704e-08, + "loss": 0.913, + "step": 10009 + }, + { + "epoch": 0.9027370699373225, + "grad_norm": 1.3962210453620523, + "learning_rate": 9.833880208430678e-08, + "loss": 0.9464, + "step": 10010 + }, + { + "epoch": 0.9028272534607927, + "grad_norm": 1.7074034261132236, + "learning_rate": 9.815794650579601e-08, + "loss": 0.9332, + "step": 10011 + }, + { + "epoch": 0.902917436984263, + "grad_norm": 1.9405558083895045, + "learning_rate": 9.797725320293548e-08, + "loss": 0.9203, + "step": 10012 + }, + { + "epoch": 0.9030076205077332, + "grad_norm": 1.6971854837370857, + "learning_rate": 9.779672219114354e-08, + "loss": 1.0168, + "step": 10013 + }, + { + "epoch": 0.9030978040312035, + "grad_norm": 1.534508333026314, + "learning_rate": 9.761635348582386e-08, + "loss": 0.918, + "step": 10014 + }, + { + "epoch": 0.9031879875546738, + "grad_norm": 1.8839343515771048, + "learning_rate": 9.743614710236658e-08, + "loss": 0.9744, + "step": 10015 + }, + { + "epoch": 0.903278171078144, + "grad_norm": 1.493737419426584, + "learning_rate": 9.725610305614806e-08, + "loss": 0.9031, + "step": 10016 + }, + { + "epoch": 0.9033683546016142, + "grad_norm": 1.497584066308062, + "learning_rate": 9.707622136253002e-08, + "loss": 0.9361, + "step": 10017 + }, + { + "epoch": 0.9034585381250846, + "grad_norm": 1.4527101642409825, + "learning_rate": 9.689650203686173e-08, + "loss": 0.9638, + "step": 10018 + }, + { + "epoch": 0.9035487216485548, + "grad_norm": 1.8280495522083036, + "learning_rate": 9.671694509447715e-08, + "loss": 0.9196, + "step": 10019 + }, + { + "epoch": 0.903638905172025, + "grad_norm": 1.5859419348142658, + "learning_rate": 9.653755055069757e-08, + "loss": 0.9602, + "step": 10020 + }, + { + "epoch": 0.9037290886954953, + "grad_norm": 1.247122026415275, + "learning_rate": 9.635831842082987e-08, + "loss": 0.9192, + "step": 10021 + }, + { + "epoch": 0.9038192722189656, + "grad_norm": 1.6175792422333364, + "learning_rate": 9.617924872016691e-08, + "loss": 0.9586, + "step": 10022 + }, + { + "epoch": 0.9039094557424359, + "grad_norm": 1.6127515020072256, + "learning_rate": 9.600034146398806e-08, + "loss": 0.9798, + "step": 10023 + }, + { + "epoch": 0.9039996392659061, + "grad_norm": 0.6621893100254728, + "learning_rate": 9.582159666755863e-08, + "loss": 0.7828, + "step": 10024 + }, + { + "epoch": 0.9040898227893763, + "grad_norm": 1.3256281551339724, + "learning_rate": 9.564301434612976e-08, + "loss": 0.9078, + "step": 10025 + }, + { + "epoch": 0.9041800063128467, + "grad_norm": 1.5325662055346658, + "learning_rate": 9.546459451494015e-08, + "loss": 0.9895, + "step": 10026 + }, + { + "epoch": 0.9042701898363169, + "grad_norm": 0.6608968338957343, + "learning_rate": 9.528633718921231e-08, + "loss": 0.8066, + "step": 10027 + }, + { + "epoch": 0.9043603733597871, + "grad_norm": 2.825939619727301, + "learning_rate": 9.510824238415672e-08, + "loss": 0.9937, + "step": 10028 + }, + { + "epoch": 0.9044505568832575, + "grad_norm": 1.317332635126109, + "learning_rate": 9.493031011496944e-08, + "loss": 0.9752, + "step": 10029 + }, + { + "epoch": 0.9045407404067277, + "grad_norm": 1.5688019124594728, + "learning_rate": 9.475254039683234e-08, + "loss": 0.7943, + "step": 10030 + }, + { + "epoch": 0.904630923930198, + "grad_norm": 1.4541864107790003, + "learning_rate": 9.45749332449144e-08, + "loss": 0.9034, + "step": 10031 + }, + { + "epoch": 0.9047211074536682, + "grad_norm": 0.6625436579516404, + "learning_rate": 9.439748867436903e-08, + "loss": 0.7904, + "step": 10032 + }, + { + "epoch": 0.9048112909771385, + "grad_norm": 1.9741572042517281, + "learning_rate": 9.42202067003377e-08, + "loss": 0.9025, + "step": 10033 + }, + { + "epoch": 0.9049014745006088, + "grad_norm": 1.6183489623069707, + "learning_rate": 9.404308733794652e-08, + "loss": 1.0764, + "step": 10034 + }, + { + "epoch": 0.904991658024079, + "grad_norm": 1.4260000400676962, + "learning_rate": 9.38661306023083e-08, + "loss": 0.9599, + "step": 10035 + }, + { + "epoch": 0.9050818415475492, + "grad_norm": 1.727942213218101, + "learning_rate": 9.368933650852229e-08, + "loss": 0.9333, + "step": 10036 + }, + { + "epoch": 0.9051720250710196, + "grad_norm": 0.7090364127069727, + "learning_rate": 9.351270507167352e-08, + "loss": 0.7948, + "step": 10037 + }, + { + "epoch": 0.9052622085944898, + "grad_norm": 1.5972007344592685, + "learning_rate": 9.333623630683285e-08, + "loss": 0.8794, + "step": 10038 + }, + { + "epoch": 0.90535239211796, + "grad_norm": 1.5633762191605445, + "learning_rate": 9.315993022905799e-08, + "loss": 0.7866, + "step": 10039 + }, + { + "epoch": 0.9054425756414303, + "grad_norm": 1.591373466861913, + "learning_rate": 9.298378685339158e-08, + "loss": 0.917, + "step": 10040 + }, + { + "epoch": 0.9055327591649006, + "grad_norm": 1.7587948915002658, + "learning_rate": 9.280780619486406e-08, + "loss": 1.0084, + "step": 10041 + }, + { + "epoch": 0.9056229426883708, + "grad_norm": 1.561467325372573, + "learning_rate": 9.26319882684905e-08, + "loss": 1.0226, + "step": 10042 + }, + { + "epoch": 0.9057131262118411, + "grad_norm": 1.3551504767361016, + "learning_rate": 9.245633308927293e-08, + "loss": 0.91, + "step": 10043 + }, + { + "epoch": 0.9058033097353113, + "grad_norm": 1.626179926987707, + "learning_rate": 9.228084067219888e-08, + "loss": 0.928, + "step": 10044 + }, + { + "epoch": 0.9058934932587817, + "grad_norm": 1.6446927460230378, + "learning_rate": 9.210551103224284e-08, + "loss": 0.8877, + "step": 10045 + }, + { + "epoch": 0.9059836767822519, + "grad_norm": 1.4006345315699975, + "learning_rate": 9.193034418436463e-08, + "loss": 0.9927, + "step": 10046 + }, + { + "epoch": 0.9060738603057221, + "grad_norm": 0.6699203847084485, + "learning_rate": 9.175534014351005e-08, + "loss": 0.8144, + "step": 10047 + }, + { + "epoch": 0.9061640438291924, + "grad_norm": 1.9137421631855036, + "learning_rate": 9.158049892461228e-08, + "loss": 0.8831, + "step": 10048 + }, + { + "epoch": 0.9062542273526627, + "grad_norm": 1.3033714280979978, + "learning_rate": 9.140582054258871e-08, + "loss": 0.8805, + "step": 10049 + }, + { + "epoch": 0.9063444108761329, + "grad_norm": 1.681555060182513, + "learning_rate": 9.123130501234499e-08, + "loss": 0.9498, + "step": 10050 + }, + { + "epoch": 0.9064345943996032, + "grad_norm": 1.7696744006192646, + "learning_rate": 9.105695234877098e-08, + "loss": 0.9955, + "step": 10051 + }, + { + "epoch": 0.9065247779230734, + "grad_norm": 1.3153750855031576, + "learning_rate": 9.088276256674344e-08, + "loss": 0.9703, + "step": 10052 + }, + { + "epoch": 0.9066149614465437, + "grad_norm": 1.4756313395252925, + "learning_rate": 9.070873568112536e-08, + "loss": 0.9281, + "step": 10053 + }, + { + "epoch": 0.906705144970014, + "grad_norm": 1.6933486744543687, + "learning_rate": 9.053487170676577e-08, + "loss": 0.9968, + "step": 10054 + }, + { + "epoch": 0.9067953284934842, + "grad_norm": 1.671451907233705, + "learning_rate": 9.036117065849968e-08, + "loss": 1.0024, + "step": 10055 + }, + { + "epoch": 0.9068855120169546, + "grad_norm": 1.567832620790815, + "learning_rate": 9.018763255114837e-08, + "loss": 0.9254, + "step": 10056 + }, + { + "epoch": 0.9069756955404248, + "grad_norm": 1.4694069549707482, + "learning_rate": 9.00142573995184e-08, + "loss": 0.9076, + "step": 10057 + }, + { + "epoch": 0.907065879063895, + "grad_norm": 1.5459449361048212, + "learning_rate": 8.984104521840375e-08, + "loss": 1.0204, + "step": 10058 + }, + { + "epoch": 0.9071560625873653, + "grad_norm": 1.8125060508890107, + "learning_rate": 8.966799602258346e-08, + "loss": 0.965, + "step": 10059 + }, + { + "epoch": 0.9072462461108356, + "grad_norm": 1.5914045540787018, + "learning_rate": 8.949510982682329e-08, + "loss": 0.9863, + "step": 10060 + }, + { + "epoch": 0.9073364296343058, + "grad_norm": 5.287083220232223, + "learning_rate": 8.932238664587499e-08, + "loss": 1.0108, + "step": 10061 + }, + { + "epoch": 0.9074266131577761, + "grad_norm": 1.6794989191695173, + "learning_rate": 8.914982649447567e-08, + "loss": 1.038, + "step": 10062 + }, + { + "epoch": 0.9075167966812463, + "grad_norm": 1.4567539472786704, + "learning_rate": 8.897742938734975e-08, + "loss": 1.0063, + "step": 10063 + }, + { + "epoch": 0.9076069802047166, + "grad_norm": 1.4985673738747112, + "learning_rate": 8.880519533920661e-08, + "loss": 0.9661, + "step": 10064 + }, + { + "epoch": 0.9076971637281869, + "grad_norm": 1.6054311780408206, + "learning_rate": 8.863312436474268e-08, + "loss": 0.9452, + "step": 10065 + }, + { + "epoch": 0.9077873472516571, + "grad_norm": 1.9155030104360322, + "learning_rate": 8.846121647863936e-08, + "loss": 0.9442, + "step": 10066 + }, + { + "epoch": 0.9078775307751273, + "grad_norm": 1.3683575228614044, + "learning_rate": 8.828947169556555e-08, + "loss": 0.9438, + "step": 10067 + }, + { + "epoch": 0.9079677142985977, + "grad_norm": 1.4446894772050065, + "learning_rate": 8.81178900301749e-08, + "loss": 0.9076, + "step": 10068 + }, + { + "epoch": 0.9080578978220679, + "grad_norm": 1.6786473802004496, + "learning_rate": 8.794647149710787e-08, + "loss": 0.8589, + "step": 10069 + }, + { + "epoch": 0.9081480813455381, + "grad_norm": 1.616761819588933, + "learning_rate": 8.777521611099081e-08, + "loss": 0.9413, + "step": 10070 + }, + { + "epoch": 0.9082382648690084, + "grad_norm": 1.50875350762112, + "learning_rate": 8.760412388643624e-08, + "loss": 0.8795, + "step": 10071 + }, + { + "epoch": 0.9083284483924787, + "grad_norm": 1.573837423403633, + "learning_rate": 8.74331948380429e-08, + "loss": 0.9589, + "step": 10072 + }, + { + "epoch": 0.908418631915949, + "grad_norm": 1.5637397425472932, + "learning_rate": 8.726242898039516e-08, + "loss": 0.9422, + "step": 10073 + }, + { + "epoch": 0.9085088154394192, + "grad_norm": 1.6861790325820782, + "learning_rate": 8.709182632806334e-08, + "loss": 0.9527, + "step": 10074 + }, + { + "epoch": 0.9085989989628894, + "grad_norm": 1.6743513644872314, + "learning_rate": 8.692138689560469e-08, + "loss": 0.9471, + "step": 10075 + }, + { + "epoch": 0.9086891824863598, + "grad_norm": 1.5024171268756465, + "learning_rate": 8.675111069756203e-08, + "loss": 0.9291, + "step": 10076 + }, + { + "epoch": 0.90877936600983, + "grad_norm": 1.4912668153977078, + "learning_rate": 8.658099774846395e-08, + "loss": 0.9149, + "step": 10077 + }, + { + "epoch": 0.9088695495333002, + "grad_norm": 1.710844899637201, + "learning_rate": 8.641104806282595e-08, + "loss": 0.8854, + "step": 10078 + }, + { + "epoch": 0.9089597330567706, + "grad_norm": 1.64316543802131, + "learning_rate": 8.624126165514845e-08, + "loss": 0.9519, + "step": 10079 + }, + { + "epoch": 0.9090499165802408, + "grad_norm": 1.7050346377705394, + "learning_rate": 8.607163853991917e-08, + "loss": 0.9032, + "step": 10080 + }, + { + "epoch": 0.909140100103711, + "grad_norm": 1.5599425175664559, + "learning_rate": 8.590217873161054e-08, + "loss": 0.7902, + "step": 10081 + }, + { + "epoch": 0.9092302836271813, + "grad_norm": 1.487667567968612, + "learning_rate": 8.573288224468255e-08, + "loss": 0.9577, + "step": 10082 + }, + { + "epoch": 0.9093204671506516, + "grad_norm": 1.7139611547201197, + "learning_rate": 8.556374909358011e-08, + "loss": 0.8689, + "step": 10083 + }, + { + "epoch": 0.9094106506741219, + "grad_norm": 2.93408410333645, + "learning_rate": 8.539477929273476e-08, + "loss": 0.9147, + "step": 10084 + }, + { + "epoch": 0.9095008341975921, + "grad_norm": 1.4835238777247297, + "learning_rate": 8.522597285656386e-08, + "loss": 0.9752, + "step": 10085 + }, + { + "epoch": 0.9095910177210623, + "grad_norm": 0.6784237041151332, + "learning_rate": 8.505732979947078e-08, + "loss": 0.7917, + "step": 10086 + }, + { + "epoch": 0.9096812012445327, + "grad_norm": 1.2334770123242067, + "learning_rate": 8.488885013584557e-08, + "loss": 0.9581, + "step": 10087 + }, + { + "epoch": 0.9097713847680029, + "grad_norm": 1.5585392916159484, + "learning_rate": 8.472053388006295e-08, + "loss": 0.8812, + "step": 10088 + }, + { + "epoch": 0.9098615682914731, + "grad_norm": 1.7471704768834004, + "learning_rate": 8.455238104648565e-08, + "loss": 1.0283, + "step": 10089 + }, + { + "epoch": 0.9099517518149434, + "grad_norm": 1.4989034300242468, + "learning_rate": 8.438439164946043e-08, + "loss": 0.8142, + "step": 10090 + }, + { + "epoch": 0.9100419353384137, + "grad_norm": 1.528503290014997, + "learning_rate": 8.42165657033218e-08, + "loss": 0.9555, + "step": 10091 + }, + { + "epoch": 0.9101321188618839, + "grad_norm": 1.5114838173643868, + "learning_rate": 8.4048903222389e-08, + "loss": 0.9777, + "step": 10092 + }, + { + "epoch": 0.9102223023853542, + "grad_norm": 1.4537303853044239, + "learning_rate": 8.388140422096856e-08, + "loss": 0.991, + "step": 10093 + }, + { + "epoch": 0.9103124859088244, + "grad_norm": 1.6215461559284319, + "learning_rate": 8.371406871335173e-08, + "loss": 0.955, + "step": 10094 + }, + { + "epoch": 0.9104026694322948, + "grad_norm": 1.5143516933585732, + "learning_rate": 8.354689671381732e-08, + "loss": 0.9431, + "step": 10095 + }, + { + "epoch": 0.910492852955765, + "grad_norm": 1.4060747673210685, + "learning_rate": 8.337988823662834e-08, + "loss": 0.9367, + "step": 10096 + }, + { + "epoch": 0.9105830364792352, + "grad_norm": 1.7440633895535465, + "learning_rate": 8.321304329603607e-08, + "loss": 0.8945, + "step": 10097 + }, + { + "epoch": 0.9106732200027055, + "grad_norm": 0.6454924842757426, + "learning_rate": 8.304636190627557e-08, + "loss": 0.7875, + "step": 10098 + }, + { + "epoch": 0.9107634035261758, + "grad_norm": 1.8551999188474544, + "learning_rate": 8.287984408156945e-08, + "loss": 0.9558, + "step": 10099 + }, + { + "epoch": 0.910853587049646, + "grad_norm": 1.89154043182553, + "learning_rate": 8.271348983612591e-08, + "loss": 0.8471, + "step": 10100 + }, + { + "epoch": 0.9109437705731163, + "grad_norm": 1.380720031700388, + "learning_rate": 8.254729918413938e-08, + "loss": 0.9053, + "step": 10101 + }, + { + "epoch": 0.9110339540965866, + "grad_norm": 1.334061493327015, + "learning_rate": 8.238127213979006e-08, + "loss": 0.9115, + "step": 10102 + }, + { + "epoch": 0.9111241376200568, + "grad_norm": 1.7942408617798689, + "learning_rate": 8.221540871724398e-08, + "loss": 1.0157, + "step": 10103 + }, + { + "epoch": 0.9112143211435271, + "grad_norm": 1.4958406637679817, + "learning_rate": 8.2049708930654e-08, + "loss": 0.8314, + "step": 10104 + }, + { + "epoch": 0.9113045046669973, + "grad_norm": 1.4657789169860627, + "learning_rate": 8.188417279415793e-08, + "loss": 0.9075, + "step": 10105 + }, + { + "epoch": 0.9113946881904677, + "grad_norm": 1.6620080291655093, + "learning_rate": 8.171880032188117e-08, + "loss": 1.0333, + "step": 10106 + }, + { + "epoch": 0.9114848717139379, + "grad_norm": 1.3860486627873583, + "learning_rate": 8.155359152793351e-08, + "loss": 0.9223, + "step": 10107 + }, + { + "epoch": 0.9115750552374081, + "grad_norm": 1.8085005470467, + "learning_rate": 8.138854642641147e-08, + "loss": 0.9314, + "step": 10108 + }, + { + "epoch": 0.9116652387608783, + "grad_norm": 1.5457379564472857, + "learning_rate": 8.122366503139777e-08, + "loss": 0.8665, + "step": 10109 + }, + { + "epoch": 0.9117554222843487, + "grad_norm": 1.416124137140739, + "learning_rate": 8.105894735696117e-08, + "loss": 0.8816, + "step": 10110 + }, + { + "epoch": 0.9118456058078189, + "grad_norm": 1.7305162134830068, + "learning_rate": 8.089439341715576e-08, + "loss": 0.9188, + "step": 10111 + }, + { + "epoch": 0.9119357893312892, + "grad_norm": 1.668543124859343, + "learning_rate": 8.073000322602319e-08, + "loss": 0.9724, + "step": 10112 + }, + { + "epoch": 0.9120259728547594, + "grad_norm": 1.4691069960453758, + "learning_rate": 8.056577679758891e-08, + "loss": 0.9581, + "step": 10113 + }, + { + "epoch": 0.9121161563782297, + "grad_norm": 1.8858440891194577, + "learning_rate": 8.040171414586638e-08, + "loss": 0.8959, + "step": 10114 + }, + { + "epoch": 0.9122063399017, + "grad_norm": 1.3983779979812458, + "learning_rate": 8.023781528485419e-08, + "loss": 0.9057, + "step": 10115 + }, + { + "epoch": 0.9122965234251702, + "grad_norm": 1.7037304843112324, + "learning_rate": 8.00740802285369e-08, + "loss": 0.9685, + "step": 10116 + }, + { + "epoch": 0.9123867069486404, + "grad_norm": 1.5243048912902326, + "learning_rate": 7.99105089908858e-08, + "loss": 0.8044, + "step": 10117 + }, + { + "epoch": 0.9124768904721108, + "grad_norm": 1.398630608379927, + "learning_rate": 7.974710158585685e-08, + "loss": 0.9135, + "step": 10118 + }, + { + "epoch": 0.912567073995581, + "grad_norm": 1.5407778093488957, + "learning_rate": 7.958385802739375e-08, + "loss": 0.9678, + "step": 10119 + }, + { + "epoch": 0.9126572575190512, + "grad_norm": 1.5253651478846315, + "learning_rate": 7.942077832942452e-08, + "loss": 0.9977, + "step": 10120 + }, + { + "epoch": 0.9127474410425215, + "grad_norm": 0.7747609538916174, + "learning_rate": 7.925786250586508e-08, + "loss": 0.7823, + "step": 10121 + }, + { + "epoch": 0.9128376245659918, + "grad_norm": 0.671933060732251, + "learning_rate": 7.909511057061524e-08, + "loss": 0.7812, + "step": 10122 + }, + { + "epoch": 0.9129278080894621, + "grad_norm": 1.5711477155631366, + "learning_rate": 7.893252253756234e-08, + "loss": 0.928, + "step": 10123 + }, + { + "epoch": 0.9130179916129323, + "grad_norm": 1.8686399676229828, + "learning_rate": 7.877009842057925e-08, + "loss": 0.8717, + "step": 10124 + }, + { + "epoch": 0.9131081751364026, + "grad_norm": 1.4481405272657744, + "learning_rate": 7.860783823352512e-08, + "loss": 0.9657, + "step": 10125 + }, + { + "epoch": 0.9131983586598729, + "grad_norm": 1.7881151381336997, + "learning_rate": 7.844574199024445e-08, + "loss": 0.9488, + "step": 10126 + }, + { + "epoch": 0.9132885421833431, + "grad_norm": 1.7424609735133005, + "learning_rate": 7.82838097045686e-08, + "loss": 0.9828, + "step": 10127 + }, + { + "epoch": 0.9133787257068133, + "grad_norm": 1.602306807086091, + "learning_rate": 7.812204139031454e-08, + "loss": 0.9342, + "step": 10128 + }, + { + "epoch": 0.9134689092302837, + "grad_norm": 1.7720130880057632, + "learning_rate": 7.796043706128474e-08, + "loss": 0.9593, + "step": 10129 + }, + { + "epoch": 0.9135590927537539, + "grad_norm": 1.6495489197478677, + "learning_rate": 7.779899673126844e-08, + "loss": 0.9742, + "step": 10130 + }, + { + "epoch": 0.9136492762772241, + "grad_norm": 1.4598399236531554, + "learning_rate": 7.76377204140406e-08, + "loss": 0.985, + "step": 10131 + }, + { + "epoch": 0.9137394598006944, + "grad_norm": 1.4859690890479373, + "learning_rate": 7.74766081233622e-08, + "loss": 0.9343, + "step": 10132 + }, + { + "epoch": 0.9138296433241647, + "grad_norm": 1.4681700515168188, + "learning_rate": 7.73156598729805e-08, + "loss": 1.0053, + "step": 10133 + }, + { + "epoch": 0.913919826847635, + "grad_norm": 1.4690532777030596, + "learning_rate": 7.715487567662849e-08, + "loss": 0.9962, + "step": 10134 + }, + { + "epoch": 0.9140100103711052, + "grad_norm": 1.9908826556455845, + "learning_rate": 7.69942555480243e-08, + "loss": 0.9126, + "step": 10135 + }, + { + "epoch": 0.9141001938945754, + "grad_norm": 1.7180766780825705, + "learning_rate": 7.68337995008741e-08, + "loss": 1.0453, + "step": 10136 + }, + { + "epoch": 0.9141903774180458, + "grad_norm": 3.329378133908031, + "learning_rate": 7.667350754886803e-08, + "loss": 0.7954, + "step": 10137 + }, + { + "epoch": 0.914280560941516, + "grad_norm": 1.7102088944517109, + "learning_rate": 7.651337970568361e-08, + "loss": 0.9379, + "step": 10138 + }, + { + "epoch": 0.9143707444649862, + "grad_norm": 1.5518994659818168, + "learning_rate": 7.635341598498368e-08, + "loss": 0.905, + "step": 10139 + }, + { + "epoch": 0.9144609279884565, + "grad_norm": 1.526886698443412, + "learning_rate": 7.61936164004171e-08, + "loss": 0.9371, + "step": 10140 + }, + { + "epoch": 0.9145511115119268, + "grad_norm": 1.4506209753819705, + "learning_rate": 7.603398096561875e-08, + "loss": 0.985, + "step": 10141 + }, + { + "epoch": 0.914641295035397, + "grad_norm": 1.4900574985393171, + "learning_rate": 7.587450969420994e-08, + "loss": 0.9454, + "step": 10142 + }, + { + "epoch": 0.9147314785588673, + "grad_norm": 0.7225594533066291, + "learning_rate": 7.571520259979757e-08, + "loss": 0.7617, + "step": 10143 + }, + { + "epoch": 0.9148216620823375, + "grad_norm": 2.0937790370464935, + "learning_rate": 7.555605969597455e-08, + "loss": 0.7809, + "step": 10144 + }, + { + "epoch": 0.9149118456058078, + "grad_norm": 1.5840831119041456, + "learning_rate": 7.539708099631959e-08, + "loss": 1.0241, + "step": 10145 + }, + { + "epoch": 0.9150020291292781, + "grad_norm": 1.4026496574805303, + "learning_rate": 7.52382665143978e-08, + "loss": 0.949, + "step": 10146 + }, + { + "epoch": 0.9150922126527483, + "grad_norm": 1.6569892385175924, + "learning_rate": 7.507961626376014e-08, + "loss": 0.9978, + "step": 10147 + }, + { + "epoch": 0.9151823961762187, + "grad_norm": 1.7260352663126004, + "learning_rate": 7.492113025794378e-08, + "loss": 0.9777, + "step": 10148 + }, + { + "epoch": 0.9152725796996889, + "grad_norm": 1.5264242775381893, + "learning_rate": 7.476280851047101e-08, + "loss": 0.8903, + "step": 10149 + }, + { + "epoch": 0.9153627632231591, + "grad_norm": 1.2561511802712435, + "learning_rate": 7.460465103485125e-08, + "loss": 0.9644, + "step": 10150 + }, + { + "epoch": 0.9154529467466294, + "grad_norm": 1.6276663399583462, + "learning_rate": 7.444665784457948e-08, + "loss": 0.9336, + "step": 10151 + }, + { + "epoch": 0.9155431302700997, + "grad_norm": 1.8970055850986918, + "learning_rate": 7.42888289531356e-08, + "loss": 0.9019, + "step": 10152 + }, + { + "epoch": 0.9156333137935699, + "grad_norm": 1.7854380312175135, + "learning_rate": 7.41311643739877e-08, + "loss": 0.8638, + "step": 10153 + }, + { + "epoch": 0.9157234973170402, + "grad_norm": 1.436829369058123, + "learning_rate": 7.39736641205877e-08, + "loss": 0.9815, + "step": 10154 + }, + { + "epoch": 0.9158136808405104, + "grad_norm": 1.5116192294538782, + "learning_rate": 7.381632820637462e-08, + "loss": 0.9661, + "step": 10155 + }, + { + "epoch": 0.9159038643639807, + "grad_norm": 1.4347821125399067, + "learning_rate": 7.365915664477352e-08, + "loss": 0.928, + "step": 10156 + }, + { + "epoch": 0.915994047887451, + "grad_norm": 1.6533903229925488, + "learning_rate": 7.350214944919474e-08, + "loss": 0.9483, + "step": 10157 + }, + { + "epoch": 0.9160842314109212, + "grad_norm": 1.5596467478921145, + "learning_rate": 7.334530663303539e-08, + "loss": 0.9712, + "step": 10158 + }, + { + "epoch": 0.9161744149343914, + "grad_norm": 1.522240192689344, + "learning_rate": 7.318862820967742e-08, + "loss": 0.9494, + "step": 10159 + }, + { + "epoch": 0.9162645984578618, + "grad_norm": 1.5947834478955951, + "learning_rate": 7.303211419249056e-08, + "loss": 0.9699, + "step": 10160 + }, + { + "epoch": 0.916354781981332, + "grad_norm": 1.2379983765401965, + "learning_rate": 7.287576459482858e-08, + "loss": 0.9616, + "step": 10161 + }, + { + "epoch": 0.9164449655048023, + "grad_norm": 1.544050006827494, + "learning_rate": 7.271957943003259e-08, + "loss": 0.9834, + "step": 10162 + }, + { + "epoch": 0.9165351490282725, + "grad_norm": 1.893084416022147, + "learning_rate": 7.256355871142883e-08, + "loss": 0.9418, + "step": 10163 + }, + { + "epoch": 0.9166253325517428, + "grad_norm": 1.787402386434259, + "learning_rate": 7.240770245233019e-08, + "loss": 0.9894, + "step": 10164 + }, + { + "epoch": 0.9167155160752131, + "grad_norm": 1.5805912952743804, + "learning_rate": 7.225201066603492e-08, + "loss": 0.862, + "step": 10165 + }, + { + "epoch": 0.9168056995986833, + "grad_norm": 1.4874429419337143, + "learning_rate": 7.209648336582774e-08, + "loss": 1.0496, + "step": 10166 + }, + { + "epoch": 0.9168958831221535, + "grad_norm": 1.472846871340618, + "learning_rate": 7.19411205649787e-08, + "loss": 0.9277, + "step": 10167 + }, + { + "epoch": 0.9169860666456239, + "grad_norm": 1.6117956974302037, + "learning_rate": 7.178592227674474e-08, + "loss": 0.9687, + "step": 10168 + }, + { + "epoch": 0.9170762501690941, + "grad_norm": 1.7072918516216755, + "learning_rate": 7.163088851436771e-08, + "loss": 0.8575, + "step": 10169 + }, + { + "epoch": 0.9171664336925643, + "grad_norm": 1.5423780638539664, + "learning_rate": 7.147601929107639e-08, + "loss": 0.905, + "step": 10170 + }, + { + "epoch": 0.9172566172160346, + "grad_norm": 1.4755006231541288, + "learning_rate": 7.132131462008461e-08, + "loss": 0.985, + "step": 10171 + }, + { + "epoch": 0.9173468007395049, + "grad_norm": 1.6382076801740821, + "learning_rate": 7.116677451459297e-08, + "loss": 0.9484, + "step": 10172 + }, + { + "epoch": 0.9174369842629752, + "grad_norm": 1.2199857266951297, + "learning_rate": 7.101239898778799e-08, + "loss": 0.9659, + "step": 10173 + }, + { + "epoch": 0.9175271677864454, + "grad_norm": 1.4700914303897714, + "learning_rate": 7.085818805284094e-08, + "loss": 0.979, + "step": 10174 + }, + { + "epoch": 0.9176173513099157, + "grad_norm": 1.6568878672937273, + "learning_rate": 7.070414172291083e-08, + "loss": 1.0237, + "step": 10175 + }, + { + "epoch": 0.917707534833386, + "grad_norm": 1.5262172281181945, + "learning_rate": 7.055026001114095e-08, + "loss": 0.8852, + "step": 10176 + }, + { + "epoch": 0.9177977183568562, + "grad_norm": 0.6761333893170663, + "learning_rate": 7.039654293066211e-08, + "loss": 0.7671, + "step": 10177 + }, + { + "epoch": 0.9178879018803264, + "grad_norm": 1.6094254420995613, + "learning_rate": 7.024299049459003e-08, + "loss": 0.8673, + "step": 10178 + }, + { + "epoch": 0.9179780854037968, + "grad_norm": 1.711022988887878, + "learning_rate": 7.008960271602627e-08, + "loss": 0.8577, + "step": 10179 + }, + { + "epoch": 0.918068268927267, + "grad_norm": 1.4417265249331868, + "learning_rate": 6.993637960805921e-08, + "loss": 0.9048, + "step": 10180 + }, + { + "epoch": 0.9181584524507372, + "grad_norm": 1.410750499005308, + "learning_rate": 6.97833211837624e-08, + "loss": 0.9693, + "step": 10181 + }, + { + "epoch": 0.9182486359742075, + "grad_norm": 1.932024386309299, + "learning_rate": 6.963042745619562e-08, + "loss": 0.8902, + "step": 10182 + }, + { + "epoch": 0.9183388194976778, + "grad_norm": 2.3911216070440355, + "learning_rate": 6.947769843840511e-08, + "loss": 0.9117, + "step": 10183 + }, + { + "epoch": 0.918429003021148, + "grad_norm": 1.6137289021927874, + "learning_rate": 6.9325134143422e-08, + "loss": 0.9741, + "step": 10184 + }, + { + "epoch": 0.9185191865446183, + "grad_norm": 1.7565370806985945, + "learning_rate": 6.917273458426387e-08, + "loss": 0.924, + "step": 10185 + }, + { + "epoch": 0.9186093700680885, + "grad_norm": 0.7092658710983112, + "learning_rate": 6.902049977393476e-08, + "loss": 0.7549, + "step": 10186 + }, + { + "epoch": 0.9186995535915589, + "grad_norm": 1.3525642254964134, + "learning_rate": 6.886842972542362e-08, + "loss": 0.9462, + "step": 10187 + }, + { + "epoch": 0.9187897371150291, + "grad_norm": 1.9645150670002685, + "learning_rate": 6.871652445170672e-08, + "loss": 0.8304, + "step": 10188 + }, + { + "epoch": 0.9188799206384993, + "grad_norm": 1.4781739660302122, + "learning_rate": 6.856478396574416e-08, + "loss": 1.0245, + "step": 10189 + }, + { + "epoch": 0.9189701041619696, + "grad_norm": 1.631298549704257, + "learning_rate": 6.841320828048491e-08, + "loss": 1.003, + "step": 10190 + }, + { + "epoch": 0.9190602876854399, + "grad_norm": 1.5472815393613213, + "learning_rate": 6.826179740886062e-08, + "loss": 0.9404, + "step": 10191 + }, + { + "epoch": 0.9191504712089101, + "grad_norm": 1.5821889539693488, + "learning_rate": 6.811055136379184e-08, + "loss": 1.0069, + "step": 10192 + }, + { + "epoch": 0.9192406547323804, + "grad_norm": 1.6208435600893007, + "learning_rate": 6.79594701581827e-08, + "loss": 0.9229, + "step": 10193 + }, + { + "epoch": 0.9193308382558506, + "grad_norm": 4.527915411627465, + "learning_rate": 6.780855380492511e-08, + "loss": 0.9437, + "step": 10194 + }, + { + "epoch": 0.919421021779321, + "grad_norm": 1.3560556246086575, + "learning_rate": 6.765780231689544e-08, + "loss": 0.9574, + "step": 10195 + }, + { + "epoch": 0.9195112053027912, + "grad_norm": 1.607170075988371, + "learning_rate": 6.750721570695695e-08, + "loss": 1.0, + "step": 10196 + }, + { + "epoch": 0.9196013888262614, + "grad_norm": 2.1001972741609456, + "learning_rate": 6.735679398795868e-08, + "loss": 0.924, + "step": 10197 + }, + { + "epoch": 0.9196915723497318, + "grad_norm": 2.095103040177834, + "learning_rate": 6.720653717273506e-08, + "loss": 0.8772, + "step": 10198 + }, + { + "epoch": 0.919781755873202, + "grad_norm": 1.5319420943708582, + "learning_rate": 6.705644527410714e-08, + "loss": 0.945, + "step": 10199 + }, + { + "epoch": 0.9198719393966722, + "grad_norm": 1.6320258623018717, + "learning_rate": 6.690651830488136e-08, + "loss": 0.9095, + "step": 10200 + }, + { + "epoch": 0.9199621229201425, + "grad_norm": 1.9164320000443709, + "learning_rate": 6.675675627785037e-08, + "loss": 0.939, + "step": 10201 + }, + { + "epoch": 0.9200523064436128, + "grad_norm": 1.6674151250007962, + "learning_rate": 6.660715920579263e-08, + "loss": 0.9633, + "step": 10202 + }, + { + "epoch": 0.920142489967083, + "grad_norm": 0.625555459196235, + "learning_rate": 6.645772710147279e-08, + "loss": 0.7783, + "step": 10203 + }, + { + "epoch": 0.9202326734905533, + "grad_norm": 1.201607825090324, + "learning_rate": 6.630845997764112e-08, + "loss": 0.9141, + "step": 10204 + }, + { + "epoch": 0.9203228570140235, + "grad_norm": 1.4266617238027444, + "learning_rate": 6.615935784703409e-08, + "loss": 0.8857, + "step": 10205 + }, + { + "epoch": 0.9204130405374938, + "grad_norm": 1.8974466131889833, + "learning_rate": 6.601042072237328e-08, + "loss": 0.8702, + "step": 10206 + }, + { + "epoch": 0.9205032240609641, + "grad_norm": 1.4158313289281628, + "learning_rate": 6.586164861636767e-08, + "loss": 0.9706, + "step": 10207 + }, + { + "epoch": 0.9205934075844343, + "grad_norm": 1.972202664339521, + "learning_rate": 6.571304154171065e-08, + "loss": 0.8528, + "step": 10208 + }, + { + "epoch": 0.9206835911079045, + "grad_norm": 1.617296592976477, + "learning_rate": 6.556459951108273e-08, + "loss": 0.967, + "step": 10209 + }, + { + "epoch": 0.9207737746313749, + "grad_norm": 1.632426019952226, + "learning_rate": 6.541632253714957e-08, + "loss": 0.9262, + "step": 10210 + }, + { + "epoch": 0.9208639581548451, + "grad_norm": 1.5317783611916385, + "learning_rate": 6.526821063256261e-08, + "loss": 0.8681, + "step": 10211 + }, + { + "epoch": 0.9209541416783154, + "grad_norm": 1.7697969957091457, + "learning_rate": 6.512026380996016e-08, + "loss": 0.8517, + "step": 10212 + }, + { + "epoch": 0.9210443252017856, + "grad_norm": 1.440193803514273, + "learning_rate": 6.49724820819657e-08, + "loss": 0.9173, + "step": 10213 + }, + { + "epoch": 0.9211345087252559, + "grad_norm": 1.5608878401760247, + "learning_rate": 6.48248654611887e-08, + "loss": 0.9754, + "step": 10214 + }, + { + "epoch": 0.9212246922487262, + "grad_norm": 1.3682868679132119, + "learning_rate": 6.467741396022419e-08, + "loss": 0.9907, + "step": 10215 + }, + { + "epoch": 0.9213148757721964, + "grad_norm": 1.6374170821068, + "learning_rate": 6.453012759165455e-08, + "loss": 0.9469, + "step": 10216 + }, + { + "epoch": 0.9214050592956666, + "grad_norm": 1.5778599129298305, + "learning_rate": 6.438300636804639e-08, + "loss": 0.9067, + "step": 10217 + }, + { + "epoch": 0.921495242819137, + "grad_norm": 1.4711851669974318, + "learning_rate": 6.423605030195278e-08, + "loss": 0.9755, + "step": 10218 + }, + { + "epoch": 0.9215854263426072, + "grad_norm": 1.855885155783164, + "learning_rate": 6.408925940591304e-08, + "loss": 1.0164, + "step": 10219 + }, + { + "epoch": 0.9216756098660774, + "grad_norm": 1.4102089573528216, + "learning_rate": 6.394263369245222e-08, + "loss": 0.8754, + "step": 10220 + }, + { + "epoch": 0.9217657933895478, + "grad_norm": 1.5454894517077642, + "learning_rate": 6.379617317408126e-08, + "loss": 0.9452, + "step": 10221 + }, + { + "epoch": 0.921855976913018, + "grad_norm": 1.8552303120851374, + "learning_rate": 6.364987786329723e-08, + "loss": 0.9342, + "step": 10222 + }, + { + "epoch": 0.9219461604364882, + "grad_norm": 1.5010198464044127, + "learning_rate": 6.350374777258193e-08, + "loss": 0.9364, + "step": 10223 + }, + { + "epoch": 0.9220363439599585, + "grad_norm": 1.5526982909034397, + "learning_rate": 6.335778291440519e-08, + "loss": 0.921, + "step": 10224 + }, + { + "epoch": 0.9221265274834288, + "grad_norm": 1.4353018830594197, + "learning_rate": 6.321198330122057e-08, + "loss": 0.8469, + "step": 10225 + }, + { + "epoch": 0.9222167110068991, + "grad_norm": 9.309428073184984, + "learning_rate": 6.306634894546902e-08, + "loss": 0.8773, + "step": 10226 + }, + { + "epoch": 0.9223068945303693, + "grad_norm": 1.574063276843935, + "learning_rate": 6.292087985957661e-08, + "loss": 0.8848, + "step": 10227 + }, + { + "epoch": 0.9223970780538395, + "grad_norm": 1.622374026752547, + "learning_rate": 6.277557605595585e-08, + "loss": 0.979, + "step": 10228 + }, + { + "epoch": 0.9224872615773099, + "grad_norm": 1.4027090207907524, + "learning_rate": 6.263043754700481e-08, + "loss": 0.8694, + "step": 10229 + }, + { + "epoch": 0.9225774451007801, + "grad_norm": 1.5087527175031838, + "learning_rate": 6.248546434510671e-08, + "loss": 0.9424, + "step": 10230 + }, + { + "epoch": 0.9226676286242503, + "grad_norm": 1.5542588146119913, + "learning_rate": 6.234065646263298e-08, + "loss": 0.8737, + "step": 10231 + }, + { + "epoch": 0.9227578121477206, + "grad_norm": 1.752348754168171, + "learning_rate": 6.219601391193796e-08, + "loss": 0.8582, + "step": 10232 + }, + { + "epoch": 0.9228479956711909, + "grad_norm": 1.4242255934207158, + "learning_rate": 6.205153670536423e-08, + "loss": 0.8674, + "step": 10233 + }, + { + "epoch": 0.9229381791946611, + "grad_norm": 1.4765532962572778, + "learning_rate": 6.190722485523902e-08, + "loss": 0.9767, + "step": 10234 + }, + { + "epoch": 0.9230283627181314, + "grad_norm": 1.645622344516534, + "learning_rate": 6.176307837387607e-08, + "loss": 0.9225, + "step": 10235 + }, + { + "epoch": 0.9231185462416016, + "grad_norm": 1.9431014153345854, + "learning_rate": 6.16190972735744e-08, + "loss": 0.8937, + "step": 10236 + }, + { + "epoch": 0.923208729765072, + "grad_norm": 1.7262316080338684, + "learning_rate": 6.147528156661974e-08, + "loss": 0.9674, + "step": 10237 + }, + { + "epoch": 0.9232989132885422, + "grad_norm": 1.419812057844999, + "learning_rate": 6.133163126528273e-08, + "loss": 0.8634, + "step": 10238 + }, + { + "epoch": 0.9233890968120124, + "grad_norm": 3.957717218468225, + "learning_rate": 6.11881463818209e-08, + "loss": 0.9492, + "step": 10239 + }, + { + "epoch": 0.9234792803354827, + "grad_norm": 1.4740786374154766, + "learning_rate": 6.104482692847668e-08, + "loss": 0.9333, + "step": 10240 + }, + { + "epoch": 0.923569463858953, + "grad_norm": 1.5479376639597369, + "learning_rate": 6.090167291747917e-08, + "loss": 0.9092, + "step": 10241 + }, + { + "epoch": 0.9236596473824232, + "grad_norm": 1.576118855961398, + "learning_rate": 6.075868436104303e-08, + "loss": 0.8405, + "step": 10242 + }, + { + "epoch": 0.9237498309058935, + "grad_norm": 1.4892297160515635, + "learning_rate": 6.061586127136875e-08, + "loss": 0.9309, + "step": 10243 + }, + { + "epoch": 0.9238400144293638, + "grad_norm": 1.7172276777681785, + "learning_rate": 6.047320366064324e-08, + "loss": 0.9109, + "step": 10244 + }, + { + "epoch": 0.923930197952834, + "grad_norm": 1.4458837256687298, + "learning_rate": 6.033071154103786e-08, + "loss": 0.9495, + "step": 10245 + }, + { + "epoch": 0.9240203814763043, + "grad_norm": 1.6727624124531957, + "learning_rate": 6.018838492471178e-08, + "loss": 0.9661, + "step": 10246 + }, + { + "epoch": 0.9241105649997745, + "grad_norm": 0.6391676632037429, + "learning_rate": 6.00462238238082e-08, + "loss": 0.7807, + "step": 10247 + }, + { + "epoch": 0.9242007485232449, + "grad_norm": 1.4546284435179337, + "learning_rate": 5.990422825045827e-08, + "loss": 0.8998, + "step": 10248 + }, + { + "epoch": 0.9242909320467151, + "grad_norm": 1.544601773787471, + "learning_rate": 5.976239821677675e-08, + "loss": 0.9228, + "step": 10249 + }, + { + "epoch": 0.9243811155701853, + "grad_norm": 1.5953956784076364, + "learning_rate": 5.962073373486598e-08, + "loss": 0.8813, + "step": 10250 + }, + { + "epoch": 0.9244712990936556, + "grad_norm": 1.4373285771865099, + "learning_rate": 5.947923481681316e-08, + "loss": 0.9035, + "step": 10251 + }, + { + "epoch": 0.9245614826171259, + "grad_norm": 1.4504847176768147, + "learning_rate": 5.933790147469198e-08, + "loss": 0.9144, + "step": 10252 + }, + { + "epoch": 0.9246516661405961, + "grad_norm": 1.7290297611073449, + "learning_rate": 5.9196733720561665e-08, + "loss": 0.8946, + "step": 10253 + }, + { + "epoch": 0.9247418496640664, + "grad_norm": 2.1083278741376357, + "learning_rate": 5.905573156646793e-08, + "loss": 1.0259, + "step": 10254 + }, + { + "epoch": 0.9248320331875366, + "grad_norm": 1.9384804982642665, + "learning_rate": 5.8914895024441134e-08, + "loss": 1.0189, + "step": 10255 + }, + { + "epoch": 0.9249222167110069, + "grad_norm": 1.619157411421578, + "learning_rate": 5.877422410649857e-08, + "loss": 0.9745, + "step": 10256 + }, + { + "epoch": 0.9250124002344772, + "grad_norm": 1.6683620016103695, + "learning_rate": 5.863371882464285e-08, + "loss": 0.959, + "step": 10257 + }, + { + "epoch": 0.9251025837579474, + "grad_norm": 1.4977621233097198, + "learning_rate": 5.849337919086283e-08, + "loss": 0.9251, + "step": 10258 + }, + { + "epoch": 0.9251927672814176, + "grad_norm": 1.6531636974514319, + "learning_rate": 5.835320521713316e-08, + "loss": 0.9219, + "step": 10259 + }, + { + "epoch": 0.925282950804888, + "grad_norm": 1.4634650087021406, + "learning_rate": 5.8213196915414264e-08, + "loss": 0.926, + "step": 10260 + }, + { + "epoch": 0.9253731343283582, + "grad_norm": 1.8029515888783563, + "learning_rate": 5.807335429765237e-08, + "loss": 0.9963, + "step": 10261 + }, + { + "epoch": 0.9254633178518284, + "grad_norm": 1.6347379875809338, + "learning_rate": 5.7933677375779034e-08, + "loss": 0.9902, + "step": 10262 + }, + { + "epoch": 0.9255535013752987, + "grad_norm": 1.4841763012055889, + "learning_rate": 5.77941661617134e-08, + "loss": 0.9913, + "step": 10263 + }, + { + "epoch": 0.925643684898769, + "grad_norm": 1.6108740334352867, + "learning_rate": 5.765482066735816e-08, + "loss": 0.8731, + "step": 10264 + }, + { + "epoch": 0.9257338684222393, + "grad_norm": 1.7187201063850994, + "learning_rate": 5.7515640904604256e-08, + "loss": 0.9434, + "step": 10265 + }, + { + "epoch": 0.9258240519457095, + "grad_norm": 1.418425431285276, + "learning_rate": 5.7376626885326187e-08, + "loss": 0.9257, + "step": 10266 + }, + { + "epoch": 0.9259142354691798, + "grad_norm": 1.4949589102976688, + "learning_rate": 5.723777862138601e-08, + "loss": 1.0168, + "step": 10267 + }, + { + "epoch": 0.9260044189926501, + "grad_norm": 1.5768560463291272, + "learning_rate": 5.7099096124630705e-08, + "loss": 1.0202, + "step": 10268 + }, + { + "epoch": 0.9260946025161203, + "grad_norm": 1.358079183045289, + "learning_rate": 5.696057940689347e-08, + "loss": 0.95, + "step": 10269 + }, + { + "epoch": 0.9261847860395905, + "grad_norm": 1.715656756887109, + "learning_rate": 5.6822228479993736e-08, + "loss": 0.967, + "step": 10270 + }, + { + "epoch": 0.9262749695630609, + "grad_norm": 1.804339288583443, + "learning_rate": 5.668404335573584e-08, + "loss": 0.9876, + "step": 10271 + }, + { + "epoch": 0.9263651530865311, + "grad_norm": 1.5172525692470318, + "learning_rate": 5.654602404591058e-08, + "loss": 0.8852, + "step": 10272 + }, + { + "epoch": 0.9264553366100013, + "grad_norm": 1.5990792724731693, + "learning_rate": 5.640817056229474e-08, + "loss": 0.9989, + "step": 10273 + }, + { + "epoch": 0.9265455201334716, + "grad_norm": 1.3904794338204667, + "learning_rate": 5.6270482916650706e-08, + "loss": 0.987, + "step": 10274 + }, + { + "epoch": 0.9266357036569419, + "grad_norm": 1.6331588158827384, + "learning_rate": 5.613296112072663e-08, + "loss": 1.0247, + "step": 10275 + }, + { + "epoch": 0.9267258871804122, + "grad_norm": 2.0675690521554313, + "learning_rate": 5.59956051862569e-08, + "loss": 0.9477, + "step": 10276 + }, + { + "epoch": 0.9268160707038824, + "grad_norm": 1.479923520303007, + "learning_rate": 5.585841512496081e-08, + "loss": 1.0083, + "step": 10277 + }, + { + "epoch": 0.9269062542273526, + "grad_norm": 1.6648727378666592, + "learning_rate": 5.5721390948545e-08, + "loss": 1.0546, + "step": 10278 + }, + { + "epoch": 0.926996437750823, + "grad_norm": 1.554830573318152, + "learning_rate": 5.558453266870056e-08, + "loss": 0.9856, + "step": 10279 + }, + { + "epoch": 0.9270866212742932, + "grad_norm": 1.5205506972266014, + "learning_rate": 5.544784029710525e-08, + "loss": 0.8817, + "step": 10280 + }, + { + "epoch": 0.9271768047977634, + "grad_norm": 1.851608983997529, + "learning_rate": 5.531131384542242e-08, + "loss": 1.0146, + "step": 10281 + }, + { + "epoch": 0.9272669883212337, + "grad_norm": 1.6768251382222876, + "learning_rate": 5.51749533253012e-08, + "loss": 0.8961, + "step": 10282 + }, + { + "epoch": 0.927357171844704, + "grad_norm": 1.3873387629671425, + "learning_rate": 5.503875874837649e-08, + "loss": 0.8865, + "step": 10283 + }, + { + "epoch": 0.9274473553681742, + "grad_norm": 1.8039594891053208, + "learning_rate": 5.4902730126269225e-08, + "loss": 0.9128, + "step": 10284 + }, + { + "epoch": 0.9275375388916445, + "grad_norm": 1.5549468011127492, + "learning_rate": 5.476686747058656e-08, + "loss": 0.9773, + "step": 10285 + }, + { + "epoch": 0.9276277224151147, + "grad_norm": 3.005299496824367, + "learning_rate": 5.4631170792920124e-08, + "loss": 0.9203, + "step": 10286 + }, + { + "epoch": 0.927717905938585, + "grad_norm": 1.7773960650782823, + "learning_rate": 5.449564010484953e-08, + "loss": 0.9229, + "step": 10287 + }, + { + "epoch": 0.9278080894620553, + "grad_norm": 1.7136744377233588, + "learning_rate": 5.436027541793775e-08, + "loss": 0.9682, + "step": 10288 + }, + { + "epoch": 0.9278982729855255, + "grad_norm": 1.463809856490751, + "learning_rate": 5.4225076743735554e-08, + "loss": 0.9188, + "step": 10289 + }, + { + "epoch": 0.9279884565089958, + "grad_norm": 1.569977074473949, + "learning_rate": 5.409004409377882e-08, + "loss": 0.862, + "step": 10290 + }, + { + "epoch": 0.9280786400324661, + "grad_norm": 1.324064352661306, + "learning_rate": 5.3955177479589e-08, + "loss": 0.8665, + "step": 10291 + }, + { + "epoch": 0.9281688235559363, + "grad_norm": 1.4491537650208388, + "learning_rate": 5.3820476912674e-08, + "loss": 0.9534, + "step": 10292 + }, + { + "epoch": 0.9282590070794066, + "grad_norm": 1.3757805776079952, + "learning_rate": 5.3685942404527063e-08, + "loss": 0.969, + "step": 10293 + }, + { + "epoch": 0.9283491906028769, + "grad_norm": 1.4849385848013088, + "learning_rate": 5.355157396662702e-08, + "loss": 0.9274, + "step": 10294 + }, + { + "epoch": 0.9284393741263471, + "grad_norm": 1.3619207025626214, + "learning_rate": 5.34173716104398e-08, + "loss": 0.9268, + "step": 10295 + }, + { + "epoch": 0.9285295576498174, + "grad_norm": 1.9514224757346859, + "learning_rate": 5.328333534741536e-08, + "loss": 0.9537, + "step": 10296 + }, + { + "epoch": 0.9286197411732876, + "grad_norm": 1.659874099989747, + "learning_rate": 5.314946518899099e-08, + "loss": 0.9113, + "step": 10297 + }, + { + "epoch": 0.928709924696758, + "grad_norm": 1.5556990816493572, + "learning_rate": 5.301576114658912e-08, + "loss": 1.0114, + "step": 10298 + }, + { + "epoch": 0.9288001082202282, + "grad_norm": 1.91465463626601, + "learning_rate": 5.288222323161795e-08, + "loss": 0.9679, + "step": 10299 + }, + { + "epoch": 0.9288902917436984, + "grad_norm": 1.6437693663642, + "learning_rate": 5.274885145547214e-08, + "loss": 0.9867, + "step": 10300 + }, + { + "epoch": 0.9289804752671686, + "grad_norm": 1.4300867575890517, + "learning_rate": 5.261564582953082e-08, + "loss": 0.9362, + "step": 10301 + }, + { + "epoch": 0.929070658790639, + "grad_norm": 2.0094156359038666, + "learning_rate": 5.248260636516066e-08, + "loss": 0.9189, + "step": 10302 + }, + { + "epoch": 0.9291608423141092, + "grad_norm": 1.756572031383108, + "learning_rate": 5.2349733073712824e-08, + "loss": 0.8898, + "step": 10303 + }, + { + "epoch": 0.9292510258375795, + "grad_norm": 1.525948670280175, + "learning_rate": 5.221702596652533e-08, + "loss": 1.0356, + "step": 10304 + }, + { + "epoch": 0.9293412093610497, + "grad_norm": 1.9400883889463487, + "learning_rate": 5.208448505492091e-08, + "loss": 0.8876, + "step": 10305 + }, + { + "epoch": 0.92943139288452, + "grad_norm": 1.446996419273433, + "learning_rate": 5.1952110350208965e-08, + "loss": 0.925, + "step": 10306 + }, + { + "epoch": 0.9295215764079903, + "grad_norm": 1.5437000714463598, + "learning_rate": 5.181990186368446e-08, + "loss": 0.8981, + "step": 10307 + }, + { + "epoch": 0.9296117599314605, + "grad_norm": 1.46258207563624, + "learning_rate": 5.1687859606627915e-08, + "loss": 1.0394, + "step": 10308 + }, + { + "epoch": 0.9297019434549307, + "grad_norm": 1.699303004488918, + "learning_rate": 5.1555983590306327e-08, + "loss": 0.8593, + "step": 10309 + }, + { + "epoch": 0.9297921269784011, + "grad_norm": 0.6973718300511625, + "learning_rate": 5.1424273825971806e-08, + "loss": 0.7789, + "step": 10310 + }, + { + "epoch": 0.9298823105018713, + "grad_norm": 2.1075012927430374, + "learning_rate": 5.1292730324862475e-08, + "loss": 0.9046, + "step": 10311 + }, + { + "epoch": 0.9299724940253415, + "grad_norm": 1.4516898010080632, + "learning_rate": 5.116135309820224e-08, + "loss": 0.9582, + "step": 10312 + }, + { + "epoch": 0.9300626775488118, + "grad_norm": 1.6559532907413008, + "learning_rate": 5.103014215720147e-08, + "loss": 0.9771, + "step": 10313 + }, + { + "epoch": 0.9301528610722821, + "grad_norm": 1.59885575983363, + "learning_rate": 5.0899097513055214e-08, + "loss": 1.0006, + "step": 10314 + }, + { + "epoch": 0.9302430445957524, + "grad_norm": 1.8433767684500708, + "learning_rate": 5.076821917694563e-08, + "loss": 0.9328, + "step": 10315 + }, + { + "epoch": 0.9303332281192226, + "grad_norm": 1.7769315472171356, + "learning_rate": 5.063750716003889e-08, + "loss": 0.9956, + "step": 10316 + }, + { + "epoch": 0.9304234116426929, + "grad_norm": 1.6783801291544993, + "learning_rate": 5.050696147348921e-08, + "loss": 0.9303, + "step": 10317 + }, + { + "epoch": 0.9305135951661632, + "grad_norm": 2.2838363617768827, + "learning_rate": 5.037658212843454e-08, + "loss": 0.9172, + "step": 10318 + }, + { + "epoch": 0.9306037786896334, + "grad_norm": 1.4156453246958618, + "learning_rate": 5.0246369136000444e-08, + "loss": 0.837, + "step": 10319 + }, + { + "epoch": 0.9306939622131036, + "grad_norm": 1.494977489799537, + "learning_rate": 5.011632250729691e-08, + "loss": 1.0147, + "step": 10320 + }, + { + "epoch": 0.930784145736574, + "grad_norm": 1.5979452081723338, + "learning_rate": 4.998644225342019e-08, + "loss": 1.033, + "step": 10321 + }, + { + "epoch": 0.9308743292600442, + "grad_norm": 1.8199100295035662, + "learning_rate": 4.9856728385452296e-08, + "loss": 0.9448, + "step": 10322 + }, + { + "epoch": 0.9309645127835144, + "grad_norm": 1.5636495557645895, + "learning_rate": 4.9727180914461485e-08, + "loss": 1.0288, + "step": 10323 + }, + { + "epoch": 0.9310546963069847, + "grad_norm": 1.7206479084278752, + "learning_rate": 4.959779985150137e-08, + "loss": 0.9568, + "step": 10324 + }, + { + "epoch": 0.931144879830455, + "grad_norm": 2.3851716916646195, + "learning_rate": 4.9468585207611105e-08, + "loss": 0.892, + "step": 10325 + }, + { + "epoch": 0.9312350633539253, + "grad_norm": 1.7080054394905613, + "learning_rate": 4.9339536993816764e-08, + "loss": 0.9229, + "step": 10326 + }, + { + "epoch": 0.9313252468773955, + "grad_norm": 1.6247409467349534, + "learning_rate": 4.921065522112844e-08, + "loss": 0.8923, + "step": 10327 + }, + { + "epoch": 0.9314154304008657, + "grad_norm": 1.6184540732169161, + "learning_rate": 4.908193990054377e-08, + "loss": 0.9937, + "step": 10328 + }, + { + "epoch": 0.9315056139243361, + "grad_norm": 1.3950394417714147, + "learning_rate": 4.89533910430453e-08, + "loss": 0.9582, + "step": 10329 + }, + { + "epoch": 0.9315957974478063, + "grad_norm": 1.4405433946115727, + "learning_rate": 4.8825008659601376e-08, + "loss": 0.9351, + "step": 10330 + }, + { + "epoch": 0.9316859809712765, + "grad_norm": 2.079610537416851, + "learning_rate": 4.869679276116634e-08, + "loss": 0.9587, + "step": 10331 + }, + { + "epoch": 0.9317761644947468, + "grad_norm": 1.428805219729587, + "learning_rate": 4.856874335868055e-08, + "loss": 0.9617, + "step": 10332 + }, + { + "epoch": 0.9318663480182171, + "grad_norm": 1.5515626721031501, + "learning_rate": 4.844086046306928e-08, + "loss": 1.0511, + "step": 10333 + }, + { + "epoch": 0.9319565315416873, + "grad_norm": 1.537417389425878, + "learning_rate": 4.8313144085244896e-08, + "loss": 0.9445, + "step": 10334 + }, + { + "epoch": 0.9320467150651576, + "grad_norm": 0.6638066191510205, + "learning_rate": 4.818559423610424e-08, + "loss": 0.7757, + "step": 10335 + }, + { + "epoch": 0.9321368985886278, + "grad_norm": 1.4147227732156331, + "learning_rate": 4.8058210926531284e-08, + "loss": 0.8872, + "step": 10336 + }, + { + "epoch": 0.9322270821120981, + "grad_norm": 1.3947150699626503, + "learning_rate": 4.7930994167394435e-08, + "loss": 0.9783, + "step": 10337 + }, + { + "epoch": 0.9323172656355684, + "grad_norm": 4.224773486684694, + "learning_rate": 4.7803943969548786e-08, + "loss": 0.9875, + "step": 10338 + }, + { + "epoch": 0.9324074491590386, + "grad_norm": 1.7751023464515525, + "learning_rate": 4.7677060343834784e-08, + "loss": 0.8928, + "step": 10339 + }, + { + "epoch": 0.932497632682509, + "grad_norm": 1.9444442711179142, + "learning_rate": 4.75503433010791e-08, + "loss": 0.9377, + "step": 10340 + }, + { + "epoch": 0.9325878162059792, + "grad_norm": 2.2564301268301765, + "learning_rate": 4.742379285209419e-08, + "loss": 0.9278, + "step": 10341 + }, + { + "epoch": 0.9326779997294494, + "grad_norm": 1.4699551122649883, + "learning_rate": 4.72974090076772e-08, + "loss": 1.0416, + "step": 10342 + }, + { + "epoch": 0.9327681832529197, + "grad_norm": 1.30700587255418, + "learning_rate": 4.717119177861262e-08, + "loss": 0.9024, + "step": 10343 + }, + { + "epoch": 0.93285836677639, + "grad_norm": 1.6425179596358879, + "learning_rate": 4.70451411756696e-08, + "loss": 0.957, + "step": 10344 + }, + { + "epoch": 0.9329485502998602, + "grad_norm": 1.4082066804415068, + "learning_rate": 4.691925720960355e-08, + "loss": 0.9831, + "step": 10345 + }, + { + "epoch": 0.9330387338233305, + "grad_norm": 1.7114358811889923, + "learning_rate": 4.6793539891155645e-08, + "loss": 0.9692, + "step": 10346 + }, + { + "epoch": 0.9331289173468007, + "grad_norm": 1.5766042804002272, + "learning_rate": 4.6667989231052864e-08, + "loss": 0.9226, + "step": 10347 + }, + { + "epoch": 0.933219100870271, + "grad_norm": 1.7149615062697647, + "learning_rate": 4.654260524000797e-08, + "loss": 0.9506, + "step": 10348 + }, + { + "epoch": 0.9333092843937413, + "grad_norm": 1.442968994537151, + "learning_rate": 4.6417387928719076e-08, + "loss": 0.9308, + "step": 10349 + }, + { + "epoch": 0.9333994679172115, + "grad_norm": 1.367848045997578, + "learning_rate": 4.629233730787052e-08, + "loss": 0.8173, + "step": 10350 + }, + { + "epoch": 0.9334896514406817, + "grad_norm": 1.6848415344387864, + "learning_rate": 4.616745338813266e-08, + "loss": 0.9073, + "step": 10351 + }, + { + "epoch": 0.9335798349641521, + "grad_norm": 1.574288946437213, + "learning_rate": 4.6042736180160744e-08, + "loss": 0.9633, + "step": 10352 + }, + { + "epoch": 0.9336700184876223, + "grad_norm": 1.6452030069375192, + "learning_rate": 4.591818569459671e-08, + "loss": 1.0655, + "step": 10353 + }, + { + "epoch": 0.9337602020110926, + "grad_norm": 1.733372716578537, + "learning_rate": 4.5793801942067614e-08, + "loss": 0.9681, + "step": 10354 + }, + { + "epoch": 0.9338503855345628, + "grad_norm": 2.0553822086316624, + "learning_rate": 4.566958493318673e-08, + "loss": 0.8111, + "step": 10355 + }, + { + "epoch": 0.9339405690580331, + "grad_norm": 0.6272020410316573, + "learning_rate": 4.554553467855316e-08, + "loss": 0.8001, + "step": 10356 + }, + { + "epoch": 0.9340307525815034, + "grad_norm": 1.5809228098711443, + "learning_rate": 4.5421651188751074e-08, + "loss": 0.9611, + "step": 10357 + }, + { + "epoch": 0.9341209361049736, + "grad_norm": 1.887963178784751, + "learning_rate": 4.529793447435137e-08, + "loss": 0.96, + "step": 10358 + }, + { + "epoch": 0.9342111196284438, + "grad_norm": 1.4667896417878827, + "learning_rate": 4.5174384545909824e-08, + "loss": 0.9456, + "step": 10359 + }, + { + "epoch": 0.9343013031519142, + "grad_norm": 1.9792930830352096, + "learning_rate": 4.505100141396867e-08, + "loss": 0.9867, + "step": 10360 + }, + { + "epoch": 0.9343914866753844, + "grad_norm": 1.7735623219124952, + "learning_rate": 4.492778508905548e-08, + "loss": 0.98, + "step": 10361 + }, + { + "epoch": 0.9344816701988546, + "grad_norm": 1.6383350194312443, + "learning_rate": 4.480473558168385e-08, + "loss": 0.8855, + "step": 10362 + }, + { + "epoch": 0.934571853722325, + "grad_norm": 4.13834234692722, + "learning_rate": 4.4681852902352936e-08, + "loss": 0.8487, + "step": 10363 + }, + { + "epoch": 0.9346620372457952, + "grad_norm": 1.619747255295025, + "learning_rate": 4.455913706154812e-08, + "loss": 0.8626, + "step": 10364 + }, + { + "epoch": 0.9347522207692655, + "grad_norm": 1.716787223224611, + "learning_rate": 4.443658806973949e-08, + "loss": 0.974, + "step": 10365 + }, + { + "epoch": 0.9348424042927357, + "grad_norm": 0.7058039387486296, + "learning_rate": 4.431420593738444e-08, + "loss": 0.8023, + "step": 10366 + }, + { + "epoch": 0.934932587816206, + "grad_norm": 1.5843371505279324, + "learning_rate": 4.419199067492485e-08, + "loss": 0.8581, + "step": 10367 + }, + { + "epoch": 0.9350227713396763, + "grad_norm": 1.5532961039249746, + "learning_rate": 4.4069942292788596e-08, + "loss": 0.9535, + "step": 10368 + }, + { + "epoch": 0.9351129548631465, + "grad_norm": 1.5479334283112036, + "learning_rate": 4.39480608013898e-08, + "loss": 0.9489, + "step": 10369 + }, + { + "epoch": 0.9352031383866167, + "grad_norm": 1.4688676827542013, + "learning_rate": 4.3826346211128126e-08, + "loss": 1.0309, + "step": 10370 + }, + { + "epoch": 0.9352933219100871, + "grad_norm": 1.388453599217003, + "learning_rate": 4.370479853238884e-08, + "loss": 0.947, + "step": 10371 + }, + { + "epoch": 0.9353835054335573, + "grad_norm": 1.4713448673280751, + "learning_rate": 4.3583417775542756e-08, + "loss": 0.91, + "step": 10372 + }, + { + "epoch": 0.9354736889570275, + "grad_norm": 1.451577459777679, + "learning_rate": 4.3462203950947575e-08, + "loss": 0.8977, + "step": 10373 + }, + { + "epoch": 0.9355638724804978, + "grad_norm": 1.5089748823680518, + "learning_rate": 4.3341157068944814e-08, + "loss": 0.9306, + "step": 10374 + }, + { + "epoch": 0.9356540560039681, + "grad_norm": 1.7883647907520535, + "learning_rate": 4.322027713986376e-08, + "loss": 0.9593, + "step": 10375 + }, + { + "epoch": 0.9357442395274383, + "grad_norm": 1.8497649533218652, + "learning_rate": 4.309956417401816e-08, + "loss": 1.0181, + "step": 10376 + }, + { + "epoch": 0.9358344230509086, + "grad_norm": 1.424708048711423, + "learning_rate": 4.297901818170801e-08, + "loss": 0.9155, + "step": 10377 + }, + { + "epoch": 0.9359246065743788, + "grad_norm": 1.4213020931148235, + "learning_rate": 4.285863917321886e-08, + "loss": 0.8754, + "step": 10378 + }, + { + "epoch": 0.9360147900978492, + "grad_norm": 1.5890944306082817, + "learning_rate": 4.2738427158822253e-08, + "loss": 0.882, + "step": 10379 + }, + { + "epoch": 0.9361049736213194, + "grad_norm": 1.611552570779188, + "learning_rate": 4.261838214877511e-08, + "loss": 0.9503, + "step": 10380 + }, + { + "epoch": 0.9361951571447896, + "grad_norm": 1.3786470855872455, + "learning_rate": 4.249850415332079e-08, + "loss": 0.9509, + "step": 10381 + }, + { + "epoch": 0.9362853406682599, + "grad_norm": 1.4722109660232137, + "learning_rate": 4.237879318268756e-08, + "loss": 1.0048, + "step": 10382 + }, + { + "epoch": 0.9363755241917302, + "grad_norm": 1.9954635909431024, + "learning_rate": 4.225924924708968e-08, + "loss": 0.8753, + "step": 10383 + }, + { + "epoch": 0.9364657077152004, + "grad_norm": 1.8271784491956657, + "learning_rate": 4.2139872356727665e-08, + "loss": 0.9531, + "step": 10384 + }, + { + "epoch": 0.9365558912386707, + "grad_norm": 1.6139069976199225, + "learning_rate": 4.202066252178738e-08, + "loss": 0.9124, + "step": 10385 + }, + { + "epoch": 0.936646074762141, + "grad_norm": 1.7353117191718934, + "learning_rate": 4.1901619752440445e-08, + "loss": 1.0361, + "step": 10386 + }, + { + "epoch": 0.9367362582856112, + "grad_norm": 1.5966932756971497, + "learning_rate": 4.178274405884363e-08, + "loss": 0.9426, + "step": 10387 + }, + { + "epoch": 0.9368264418090815, + "grad_norm": 1.442587598392009, + "learning_rate": 4.166403545114105e-08, + "loss": 0.9567, + "step": 10388 + }, + { + "epoch": 0.9369166253325517, + "grad_norm": 1.7095590551618676, + "learning_rate": 4.154549393946083e-08, + "loss": 0.9994, + "step": 10389 + }, + { + "epoch": 0.937006808856022, + "grad_norm": 1.6935438079889273, + "learning_rate": 4.14271195339182e-08, + "loss": 0.9719, + "step": 10390 + }, + { + "epoch": 0.9370969923794923, + "grad_norm": 1.4923355268783642, + "learning_rate": 4.1308912244613084e-08, + "loss": 0.911, + "step": 10391 + }, + { + "epoch": 0.9371871759029625, + "grad_norm": 1.4543506006834295, + "learning_rate": 4.1190872081631636e-08, + "loss": 0.9021, + "step": 10392 + }, + { + "epoch": 0.9372773594264328, + "grad_norm": 1.6593628906267932, + "learning_rate": 4.107299905504558e-08, + "loss": 1.0113, + "step": 10393 + }, + { + "epoch": 0.9373675429499031, + "grad_norm": 1.4379324055358147, + "learning_rate": 4.095529317491286e-08, + "loss": 0.9356, + "step": 10394 + }, + { + "epoch": 0.9374577264733733, + "grad_norm": 1.6596709136622518, + "learning_rate": 4.0837754451276575e-08, + "loss": 0.9501, + "step": 10395 + }, + { + "epoch": 0.9375479099968436, + "grad_norm": 1.4130908304273444, + "learning_rate": 4.072038289416557e-08, + "loss": 0.9326, + "step": 10396 + }, + { + "epoch": 0.9376380935203138, + "grad_norm": 1.4228549975561229, + "learning_rate": 4.0603178513595185e-08, + "loss": 0.947, + "step": 10397 + }, + { + "epoch": 0.9377282770437841, + "grad_norm": 1.5044281289355255, + "learning_rate": 4.0486141319565624e-08, + "loss": 0.9472, + "step": 10398 + }, + { + "epoch": 0.9378184605672544, + "grad_norm": 1.7286429320614949, + "learning_rate": 4.0369271322062916e-08, + "loss": 0.8885, + "step": 10399 + }, + { + "epoch": 0.9379086440907246, + "grad_norm": 1.5910888595078836, + "learning_rate": 4.0252568531059295e-08, + "loss": 0.9633, + "step": 10400 + }, + { + "epoch": 0.9379988276141948, + "grad_norm": 1.6633611642857373, + "learning_rate": 4.013603295651235e-08, + "loss": 0.9466, + "step": 10401 + }, + { + "epoch": 0.9380890111376652, + "grad_norm": 2.1199006998450116, + "learning_rate": 4.001966460836592e-08, + "loss": 0.9711, + "step": 10402 + }, + { + "epoch": 0.9381791946611354, + "grad_norm": 1.530465216873554, + "learning_rate": 3.990346349654894e-08, + "loss": 1.0203, + "step": 10403 + }, + { + "epoch": 0.9382693781846057, + "grad_norm": 1.4811576911073694, + "learning_rate": 3.9787429630975924e-08, + "loss": 0.9209, + "step": 10404 + }, + { + "epoch": 0.9383595617080759, + "grad_norm": 1.451359240012598, + "learning_rate": 3.967156302154828e-08, + "loss": 0.9322, + "step": 10405 + }, + { + "epoch": 0.9384497452315462, + "grad_norm": 1.5338517487468564, + "learning_rate": 3.955586367815189e-08, + "loss": 0.9013, + "step": 10406 + }, + { + "epoch": 0.9385399287550165, + "grad_norm": 1.6541526037942824, + "learning_rate": 3.944033161065907e-08, + "loss": 1.015, + "step": 10407 + }, + { + "epoch": 0.9386301122784867, + "grad_norm": 1.462557460601642, + "learning_rate": 3.93249668289275e-08, + "loss": 0.8287, + "step": 10408 + }, + { + "epoch": 0.9387202958019569, + "grad_norm": 1.3305402000716298, + "learning_rate": 3.920976934280063e-08, + "loss": 0.8711, + "step": 10409 + }, + { + "epoch": 0.9388104793254273, + "grad_norm": 1.9164245355777971, + "learning_rate": 3.909473916210815e-08, + "loss": 1.0314, + "step": 10410 + }, + { + "epoch": 0.9389006628488975, + "grad_norm": 1.8647957011354532, + "learning_rate": 3.897987629666488e-08, + "loss": 0.9053, + "step": 10411 + }, + { + "epoch": 0.9389908463723677, + "grad_norm": 1.245646333640967, + "learning_rate": 3.886518075627143e-08, + "loss": 1.019, + "step": 10412 + }, + { + "epoch": 0.9390810298958381, + "grad_norm": 2.070372843312805, + "learning_rate": 3.875065255071419e-08, + "loss": 0.942, + "step": 10413 + }, + { + "epoch": 0.9391712134193083, + "grad_norm": 1.5746665298812854, + "learning_rate": 3.863629168976579e-08, + "loss": 0.9139, + "step": 10414 + }, + { + "epoch": 0.9392613969427785, + "grad_norm": 1.6512718240824757, + "learning_rate": 3.852209818318375e-08, + "loss": 1.0004, + "step": 10415 + }, + { + "epoch": 0.9393515804662488, + "grad_norm": 1.456586345419312, + "learning_rate": 3.840807204071161e-08, + "loss": 0.8512, + "step": 10416 + }, + { + "epoch": 0.9394417639897191, + "grad_norm": 1.616186154455003, + "learning_rate": 3.829421327207894e-08, + "loss": 0.929, + "step": 10417 + }, + { + "epoch": 0.9395319475131894, + "grad_norm": 1.5181266353091105, + "learning_rate": 3.8180521887000825e-08, + "loss": 0.9022, + "step": 10418 + }, + { + "epoch": 0.9396221310366596, + "grad_norm": 0.6231066636244832, + "learning_rate": 3.806699789517775e-08, + "loss": 0.8112, + "step": 10419 + }, + { + "epoch": 0.9397123145601298, + "grad_norm": 1.482825182909735, + "learning_rate": 3.7953641306296635e-08, + "loss": 0.9334, + "step": 10420 + }, + { + "epoch": 0.9398024980836002, + "grad_norm": 1.834434965800928, + "learning_rate": 3.784045213002951e-08, + "loss": 0.7956, + "step": 10421 + }, + { + "epoch": 0.9398926816070704, + "grad_norm": 1.593588727383404, + "learning_rate": 3.7727430376033986e-08, + "loss": 0.9581, + "step": 10422 + }, + { + "epoch": 0.9399828651305406, + "grad_norm": 1.875392427703811, + "learning_rate": 3.7614576053954126e-08, + "loss": 0.9504, + "step": 10423 + }, + { + "epoch": 0.9400730486540109, + "grad_norm": 1.8839640914264857, + "learning_rate": 3.75018891734189e-08, + "loss": 0.9413, + "step": 10424 + }, + { + "epoch": 0.9401632321774812, + "grad_norm": 1.735621305248159, + "learning_rate": 3.738936974404372e-08, + "loss": 0.992, + "step": 10425 + }, + { + "epoch": 0.9402534157009514, + "grad_norm": 1.4355778489696303, + "learning_rate": 3.7277017775429354e-08, + "loss": 1.0097, + "step": 10426 + }, + { + "epoch": 0.9403435992244217, + "grad_norm": 1.3901503106361572, + "learning_rate": 3.7164833277162136e-08, + "loss": 0.97, + "step": 10427 + }, + { + "epoch": 0.9404337827478919, + "grad_norm": 1.2023656356039178, + "learning_rate": 3.705281625881418e-08, + "loss": 0.9258, + "step": 10428 + }, + { + "epoch": 0.9405239662713623, + "grad_norm": 1.7933994681750527, + "learning_rate": 3.694096672994362e-08, + "loss": 0.8546, + "step": 10429 + }, + { + "epoch": 0.9406141497948325, + "grad_norm": 1.4716135072161345, + "learning_rate": 3.682928470009394e-08, + "loss": 0.9971, + "step": 10430 + }, + { + "epoch": 0.9407043333183027, + "grad_norm": 1.487252267676449, + "learning_rate": 3.6717770178794406e-08, + "loss": 0.9015, + "step": 10431 + }, + { + "epoch": 0.940794516841773, + "grad_norm": 1.3978444125760277, + "learning_rate": 3.6606423175560287e-08, + "loss": 1.0508, + "step": 10432 + }, + { + "epoch": 0.9408847003652433, + "grad_norm": 0.640230406212972, + "learning_rate": 3.649524369989221e-08, + "loss": 0.7463, + "step": 10433 + }, + { + "epoch": 0.9409748838887135, + "grad_norm": 2.504303661117677, + "learning_rate": 3.638423176127636e-08, + "loss": 0.9784, + "step": 10434 + }, + { + "epoch": 0.9410650674121838, + "grad_norm": 1.4779430575829096, + "learning_rate": 3.6273387369185396e-08, + "loss": 0.9794, + "step": 10435 + }, + { + "epoch": 0.9411552509356541, + "grad_norm": 1.5117474535994282, + "learning_rate": 3.616271053307685e-08, + "loss": 0.8928, + "step": 10436 + }, + { + "epoch": 0.9412454344591243, + "grad_norm": 1.7628933190380756, + "learning_rate": 3.6052201262394275e-08, + "loss": 0.9678, + "step": 10437 + }, + { + "epoch": 0.9413356179825946, + "grad_norm": 1.522784675204595, + "learning_rate": 3.5941859566566816e-08, + "loss": 1.0019, + "step": 10438 + }, + { + "epoch": 0.9414258015060648, + "grad_norm": 0.6973413164530966, + "learning_rate": 3.583168545500981e-08, + "loss": 0.7686, + "step": 10439 + }, + { + "epoch": 0.9415159850295352, + "grad_norm": 1.4807238195508157, + "learning_rate": 3.5721678937123746e-08, + "loss": 0.8886, + "step": 10440 + }, + { + "epoch": 0.9416061685530054, + "grad_norm": 1.3663305158003636, + "learning_rate": 3.561184002229467e-08, + "loss": 1.0275, + "step": 10441 + }, + { + "epoch": 0.9416963520764756, + "grad_norm": 1.4701571115885936, + "learning_rate": 3.550216871989531e-08, + "loss": 0.8964, + "step": 10442 + }, + { + "epoch": 0.9417865355999459, + "grad_norm": 0.6594514187828129, + "learning_rate": 3.539266503928262e-08, + "loss": 0.8482, + "step": 10443 + }, + { + "epoch": 0.9418767191234162, + "grad_norm": 1.3646954303543355, + "learning_rate": 3.528332898980091e-08, + "loss": 0.9575, + "step": 10444 + }, + { + "epoch": 0.9419669026468864, + "grad_norm": 1.4361213209443935, + "learning_rate": 3.517416058077849e-08, + "loss": 0.9147, + "step": 10445 + }, + { + "epoch": 0.9420570861703567, + "grad_norm": 1.764168467688178, + "learning_rate": 3.506515982153102e-08, + "loss": 0.8447, + "step": 10446 + }, + { + "epoch": 0.9421472696938269, + "grad_norm": 1.6136160957929402, + "learning_rate": 3.495632672135862e-08, + "loss": 0.9546, + "step": 10447 + }, + { + "epoch": 0.9422374532172972, + "grad_norm": 1.3314567960490413, + "learning_rate": 3.4847661289547417e-08, + "loss": 0.9791, + "step": 10448 + }, + { + "epoch": 0.9423276367407675, + "grad_norm": 5.229858818701496, + "learning_rate": 3.473916353536932e-08, + "loss": 0.9073, + "step": 10449 + }, + { + "epoch": 0.9424178202642377, + "grad_norm": 1.532997146668488, + "learning_rate": 3.463083346808249e-08, + "loss": 0.9144, + "step": 10450 + }, + { + "epoch": 0.9425080037877079, + "grad_norm": 1.6155733170349942, + "learning_rate": 3.452267109692975e-08, + "loss": 0.8423, + "step": 10451 + }, + { + "epoch": 0.9425981873111783, + "grad_norm": 1.4751452552265265, + "learning_rate": 3.441467643114016e-08, + "loss": 0.9477, + "step": 10452 + }, + { + "epoch": 0.9426883708346485, + "grad_norm": 0.6098136790057842, + "learning_rate": 3.430684947992857e-08, + "loss": 0.7589, + "step": 10453 + }, + { + "epoch": 0.9427785543581187, + "grad_norm": 1.5628924067801016, + "learning_rate": 3.419919025249518e-08, + "loss": 0.9567, + "step": 10454 + }, + { + "epoch": 0.942868737881589, + "grad_norm": 1.4427205531662042, + "learning_rate": 3.40916987580262e-08, + "loss": 0.968, + "step": 10455 + }, + { + "epoch": 0.9429589214050593, + "grad_norm": 1.4361665594897688, + "learning_rate": 3.398437500569362e-08, + "loss": 0.9212, + "step": 10456 + }, + { + "epoch": 0.9430491049285296, + "grad_norm": 1.5800858958351482, + "learning_rate": 3.3877219004654347e-08, + "loss": 0.828, + "step": 10457 + }, + { + "epoch": 0.9431392884519998, + "grad_norm": 1.9250060217626936, + "learning_rate": 3.3770230764051946e-08, + "loss": 0.9795, + "step": 10458 + }, + { + "epoch": 0.9432294719754701, + "grad_norm": 1.6860996369891972, + "learning_rate": 3.366341029301534e-08, + "loss": 0.9656, + "step": 10459 + }, + { + "epoch": 0.9433196554989404, + "grad_norm": 1.7676635721534337, + "learning_rate": 3.355675760065857e-08, + "loss": 0.9563, + "step": 10460 + }, + { + "epoch": 0.9434098390224106, + "grad_norm": 0.596986582447116, + "learning_rate": 3.345027269608236e-08, + "loss": 0.7268, + "step": 10461 + }, + { + "epoch": 0.9435000225458808, + "grad_norm": 0.683769574033361, + "learning_rate": 3.334395558837211e-08, + "loss": 0.7932, + "step": 10462 + }, + { + "epoch": 0.9435902060693512, + "grad_norm": 1.4302447957806188, + "learning_rate": 3.3237806286599667e-08, + "loss": 0.8569, + "step": 10463 + }, + { + "epoch": 0.9436803895928214, + "grad_norm": 1.460406773089728, + "learning_rate": 3.313182479982224e-08, + "loss": 0.9346, + "step": 10464 + }, + { + "epoch": 0.9437705731162916, + "grad_norm": 1.7076735346622076, + "learning_rate": 3.302601113708259e-08, + "loss": 1.0149, + "step": 10465 + }, + { + "epoch": 0.9438607566397619, + "grad_norm": 1.4157802201955545, + "learning_rate": 3.292036530740972e-08, + "loss": 0.9947, + "step": 10466 + }, + { + "epoch": 0.9439509401632322, + "grad_norm": 1.4668939916407702, + "learning_rate": 3.2814887319817294e-08, + "loss": 0.9372, + "step": 10467 + }, + { + "epoch": 0.9440411236867025, + "grad_norm": 1.5766258295078646, + "learning_rate": 3.270957718330591e-08, + "loss": 0.9894, + "step": 10468 + }, + { + "epoch": 0.9441313072101727, + "grad_norm": 1.417053684028011, + "learning_rate": 3.260443490686082e-08, + "loss": 0.9884, + "step": 10469 + }, + { + "epoch": 0.9442214907336429, + "grad_norm": 1.6421262139844883, + "learning_rate": 3.249946049945351e-08, + "loss": 0.9606, + "step": 10470 + }, + { + "epoch": 0.9443116742571133, + "grad_norm": 1.7214977841236967, + "learning_rate": 3.239465397004082e-08, + "loss": 0.9615, + "step": 10471 + }, + { + "epoch": 0.9444018577805835, + "grad_norm": 1.5256484982291714, + "learning_rate": 3.229001532756559e-08, + "loss": 0.9082, + "step": 10472 + }, + { + "epoch": 0.9444920413040537, + "grad_norm": 1.9741940386983876, + "learning_rate": 3.218554458095602e-08, + "loss": 0.946, + "step": 10473 + }, + { + "epoch": 0.944582224827524, + "grad_norm": 1.756264712994942, + "learning_rate": 3.20812417391263e-08, + "loss": 0.9538, + "step": 10474 + }, + { + "epoch": 0.9446724083509943, + "grad_norm": 1.2729802158972927, + "learning_rate": 3.1977106810975764e-08, + "loss": 0.9364, + "step": 10475 + }, + { + "epoch": 0.9447625918744645, + "grad_norm": 0.7957506662810014, + "learning_rate": 3.187313980539042e-08, + "loss": 0.8231, + "step": 10476 + }, + { + "epoch": 0.9448527753979348, + "grad_norm": 1.2171908455681517, + "learning_rate": 3.176934073124071e-08, + "loss": 1.0306, + "step": 10477 + }, + { + "epoch": 0.944942958921405, + "grad_norm": 1.7659204961565307, + "learning_rate": 3.166570959738357e-08, + "loss": 0.9452, + "step": 10478 + }, + { + "epoch": 0.9450331424448754, + "grad_norm": 1.5705791645519285, + "learning_rate": 3.1562246412661476e-08, + "loss": 0.8844, + "step": 10479 + }, + { + "epoch": 0.9451233259683456, + "grad_norm": 1.527588339341122, + "learning_rate": 3.145895118590225e-08, + "loss": 0.9775, + "step": 10480 + }, + { + "epoch": 0.9452135094918158, + "grad_norm": 1.6745531254676358, + "learning_rate": 3.135582392591996e-08, + "loss": 0.8568, + "step": 10481 + }, + { + "epoch": 0.9453036930152862, + "grad_norm": 0.6215717227917059, + "learning_rate": 3.125286464151333e-08, + "loss": 0.7376, + "step": 10482 + }, + { + "epoch": 0.9453938765387564, + "grad_norm": 1.6887688987685308, + "learning_rate": 3.115007334146824e-08, + "loss": 0.9205, + "step": 10483 + }, + { + "epoch": 0.9454840600622266, + "grad_norm": 0.6531604519946111, + "learning_rate": 3.104745003455478e-08, + "loss": 0.7585, + "step": 10484 + }, + { + "epoch": 0.9455742435856969, + "grad_norm": 1.525485886885164, + "learning_rate": 3.094499472952972e-08, + "loss": 0.9684, + "step": 10485 + }, + { + "epoch": 0.9456644271091672, + "grad_norm": 1.5832584848614784, + "learning_rate": 3.084270743513495e-08, + "loss": 0.9162, + "step": 10486 + }, + { + "epoch": 0.9457546106326374, + "grad_norm": 1.7525056884063825, + "learning_rate": 3.074058816009817e-08, + "loss": 0.8386, + "step": 10487 + }, + { + "epoch": 0.9458447941561077, + "grad_norm": 1.592435930840672, + "learning_rate": 3.063863691313284e-08, + "loss": 0.8939, + "step": 10488 + }, + { + "epoch": 0.9459349776795779, + "grad_norm": 1.771740408062824, + "learning_rate": 3.0536853702937794e-08, + "loss": 0.8872, + "step": 10489 + }, + { + "epoch": 0.9460251612030482, + "grad_norm": 2.123549695731583, + "learning_rate": 3.043523853819807e-08, + "loss": 0.9478, + "step": 10490 + }, + { + "epoch": 0.9461153447265185, + "grad_norm": 1.6013490115903486, + "learning_rate": 3.0333791427583855e-08, + "loss": 0.9672, + "step": 10491 + }, + { + "epoch": 0.9462055282499887, + "grad_norm": 1.509508276321249, + "learning_rate": 3.023251237975111e-08, + "loss": 1.0684, + "step": 10492 + }, + { + "epoch": 0.946295711773459, + "grad_norm": 0.669311935310578, + "learning_rate": 3.0131401403341584e-08, + "loss": 0.7948, + "step": 10493 + }, + { + "epoch": 0.9463858952969293, + "grad_norm": 1.3681969539982592, + "learning_rate": 3.00304585069826e-08, + "loss": 0.926, + "step": 10494 + }, + { + "epoch": 0.9464760788203995, + "grad_norm": 1.8827078382598281, + "learning_rate": 2.992968369928728e-08, + "loss": 0.9355, + "step": 10495 + }, + { + "epoch": 0.9465662623438698, + "grad_norm": 1.5704152846977621, + "learning_rate": 2.982907698885429e-08, + "loss": 0.9638, + "step": 10496 + }, + { + "epoch": 0.94665644586734, + "grad_norm": 1.8251556016613704, + "learning_rate": 2.9728638384267645e-08, + "loss": 0.9601, + "step": 10497 + }, + { + "epoch": 0.9467466293908103, + "grad_norm": 1.7602034520551746, + "learning_rate": 2.962836789409784e-08, + "loss": 0.957, + "step": 10498 + }, + { + "epoch": 0.9468368129142806, + "grad_norm": 1.5171556117815033, + "learning_rate": 2.95282655268998e-08, + "loss": 1.0417, + "step": 10499 + }, + { + "epoch": 0.9469269964377508, + "grad_norm": 1.6593554192086397, + "learning_rate": 2.942833129121558e-08, + "loss": 0.9494, + "step": 10500 + }, + { + "epoch": 0.947017179961221, + "grad_norm": 1.46196926626838, + "learning_rate": 2.9328565195571475e-08, + "loss": 0.8755, + "step": 10501 + }, + { + "epoch": 0.9471073634846914, + "grad_norm": 1.607786857912097, + "learning_rate": 2.9228967248480675e-08, + "loss": 1.0148, + "step": 10502 + }, + { + "epoch": 0.9471975470081616, + "grad_norm": 1.679184097439367, + "learning_rate": 2.912953745844082e-08, + "loss": 0.9894, + "step": 10503 + }, + { + "epoch": 0.9472877305316318, + "grad_norm": 1.6250030077392912, + "learning_rate": 2.9030275833936247e-08, + "loss": 0.9386, + "step": 10504 + }, + { + "epoch": 0.9473779140551022, + "grad_norm": 2.0110313648526574, + "learning_rate": 2.893118238343617e-08, + "loss": 0.8476, + "step": 10505 + }, + { + "epoch": 0.9474680975785724, + "grad_norm": 1.6186515336300764, + "learning_rate": 2.8832257115396052e-08, + "loss": 0.8562, + "step": 10506 + }, + { + "epoch": 0.9475582811020427, + "grad_norm": 1.5554123918563316, + "learning_rate": 2.873350003825692e-08, + "loss": 0.93, + "step": 10507 + }, + { + "epoch": 0.9476484646255129, + "grad_norm": 0.5869089338995715, + "learning_rate": 2.8634911160444696e-08, + "loss": 0.6737, + "step": 10508 + }, + { + "epoch": 0.9477386481489832, + "grad_norm": 1.4850061781269859, + "learning_rate": 2.853649049037199e-08, + "loss": 1.0079, + "step": 10509 + }, + { + "epoch": 0.9478288316724535, + "grad_norm": 1.649133833483983, + "learning_rate": 2.8438238036436525e-08, + "loss": 0.996, + "step": 10510 + }, + { + "epoch": 0.9479190151959237, + "grad_norm": 1.6334635335578376, + "learning_rate": 2.834015380702137e-08, + "loss": 0.9414, + "step": 10511 + }, + { + "epoch": 0.9480091987193939, + "grad_norm": 0.61081200893784, + "learning_rate": 2.824223781049606e-08, + "loss": 0.7142, + "step": 10512 + }, + { + "epoch": 0.9480993822428643, + "grad_norm": 2.0524928915479186, + "learning_rate": 2.8144490055215465e-08, + "loss": 1.0109, + "step": 10513 + }, + { + "epoch": 0.9481895657663345, + "grad_norm": 1.4022231196212382, + "learning_rate": 2.8046910549519355e-08, + "loss": 0.9827, + "step": 10514 + }, + { + "epoch": 0.9482797492898047, + "grad_norm": 1.3786105954756858, + "learning_rate": 2.794949930173418e-08, + "loss": 0.9521, + "step": 10515 + }, + { + "epoch": 0.948369932813275, + "grad_norm": 1.6846506293126857, + "learning_rate": 2.7852256320171296e-08, + "loss": 0.9785, + "step": 10516 + }, + { + "epoch": 0.9484601163367453, + "grad_norm": 1.7151643978987723, + "learning_rate": 2.775518161312851e-08, + "loss": 0.8104, + "step": 10517 + }, + { + "epoch": 0.9485502998602156, + "grad_norm": 1.4235439357053736, + "learning_rate": 2.76582751888883e-08, + "loss": 0.9723, + "step": 10518 + }, + { + "epoch": 0.9486404833836858, + "grad_norm": 1.8115041562105618, + "learning_rate": 2.756153705571962e-08, + "loss": 0.966, + "step": 10519 + }, + { + "epoch": 0.948730666907156, + "grad_norm": 1.4400993563230813, + "learning_rate": 2.74649672218763e-08, + "loss": 0.9305, + "step": 10520 + }, + { + "epoch": 0.9488208504306264, + "grad_norm": 1.487874613801191, + "learning_rate": 2.7368565695598424e-08, + "loss": 0.9327, + "step": 10521 + }, + { + "epoch": 0.9489110339540966, + "grad_norm": 1.5651095152735495, + "learning_rate": 2.727233248511185e-08, + "loss": 0.8746, + "step": 10522 + }, + { + "epoch": 0.9490012174775668, + "grad_norm": 2.085822398692645, + "learning_rate": 2.71762675986269e-08, + "loss": 0.9052, + "step": 10523 + }, + { + "epoch": 0.9490914010010371, + "grad_norm": 1.3554998828489675, + "learning_rate": 2.7080371044341242e-08, + "loss": 0.9872, + "step": 10524 + }, + { + "epoch": 0.9491815845245074, + "grad_norm": 1.4471675206343357, + "learning_rate": 2.6984642830436556e-08, + "loss": 0.9793, + "step": 10525 + }, + { + "epoch": 0.9492717680479776, + "grad_norm": 1.4500895965781728, + "learning_rate": 2.688908296508141e-08, + "loss": 0.9312, + "step": 10526 + }, + { + "epoch": 0.9493619515714479, + "grad_norm": 1.8649962907401418, + "learning_rate": 2.679369145642929e-08, + "loss": 1.0023, + "step": 10527 + }, + { + "epoch": 0.9494521350949181, + "grad_norm": 1.4986196365483198, + "learning_rate": 2.669846831261946e-08, + "loss": 0.9686, + "step": 10528 + }, + { + "epoch": 0.9495423186183884, + "grad_norm": 1.6617893220149598, + "learning_rate": 2.6603413541776976e-08, + "loss": 0.9372, + "step": 10529 + }, + { + "epoch": 0.9496325021418587, + "grad_norm": 1.3434419500645942, + "learning_rate": 2.6508527152012683e-08, + "loss": 0.9004, + "step": 10530 + }, + { + "epoch": 0.9497226856653289, + "grad_norm": 1.3976171207654526, + "learning_rate": 2.641380915142233e-08, + "loss": 0.8512, + "step": 10531 + }, + { + "epoch": 0.9498128691887993, + "grad_norm": 1.650028852730688, + "learning_rate": 2.6319259548088334e-08, + "loss": 0.9693, + "step": 10532 + }, + { + "epoch": 0.9499030527122695, + "grad_norm": 1.6426173147636154, + "learning_rate": 2.6224878350077585e-08, + "loss": 0.9455, + "step": 10533 + }, + { + "epoch": 0.9499932362357397, + "grad_norm": 1.8757371724902827, + "learning_rate": 2.6130665565443633e-08, + "loss": 0.8615, + "step": 10534 + }, + { + "epoch": 0.95008341975921, + "grad_norm": 1.4139650005638467, + "learning_rate": 2.603662120222494e-08, + "loss": 0.951, + "step": 10535 + }, + { + "epoch": 0.9501736032826803, + "grad_norm": 2.1826319341993434, + "learning_rate": 2.59427452684462e-08, + "loss": 0.7956, + "step": 10536 + }, + { + "epoch": 0.9502637868061505, + "grad_norm": 1.3999537341102495, + "learning_rate": 2.5849037772117443e-08, + "loss": 0.9298, + "step": 10537 + }, + { + "epoch": 0.9503539703296208, + "grad_norm": 1.554027474777876, + "learning_rate": 2.575549872123384e-08, + "loss": 0.9716, + "step": 10538 + }, + { + "epoch": 0.950444153853091, + "grad_norm": 1.3941987653747354, + "learning_rate": 2.5662128123776994e-08, + "loss": 0.9541, + "step": 10539 + }, + { + "epoch": 0.9505343373765613, + "grad_norm": 1.836878758465899, + "learning_rate": 2.5568925987713875e-08, + "loss": 0.9454, + "step": 10540 + }, + { + "epoch": 0.9506245209000316, + "grad_norm": 1.6648972257907328, + "learning_rate": 2.5475892320996785e-08, + "loss": 0.9211, + "step": 10541 + }, + { + "epoch": 0.9507147044235018, + "grad_norm": 0.684209056227182, + "learning_rate": 2.5383027131564038e-08, + "loss": 0.7629, + "step": 10542 + }, + { + "epoch": 0.950804887946972, + "grad_norm": 1.7820265566284508, + "learning_rate": 2.52903304273393e-08, + "loss": 0.9194, + "step": 10543 + }, + { + "epoch": 0.9508950714704424, + "grad_norm": 1.4984315778597372, + "learning_rate": 2.519780221623202e-08, + "loss": 0.9329, + "step": 10544 + }, + { + "epoch": 0.9509852549939126, + "grad_norm": 1.9834389468798266, + "learning_rate": 2.510544250613722e-08, + "loss": 0.9033, + "step": 10545 + }, + { + "epoch": 0.9510754385173829, + "grad_norm": 1.6177323754909363, + "learning_rate": 2.501325130493548e-08, + "loss": 1.0488, + "step": 10546 + }, + { + "epoch": 0.9511656220408531, + "grad_norm": 0.7242919967136517, + "learning_rate": 2.4921228620493395e-08, + "loss": 0.7795, + "step": 10547 + }, + { + "epoch": 0.9512558055643234, + "grad_norm": 0.8231984553368018, + "learning_rate": 2.4829374460662244e-08, + "loss": 0.8167, + "step": 10548 + }, + { + "epoch": 0.9513459890877937, + "grad_norm": 1.3886495982769516, + "learning_rate": 2.473768883327976e-08, + "loss": 0.961, + "step": 10549 + }, + { + "epoch": 0.9514361726112639, + "grad_norm": 1.5529806457610469, + "learning_rate": 2.464617174616923e-08, + "loss": 0.9106, + "step": 10550 + }, + { + "epoch": 0.9515263561347341, + "grad_norm": 1.7881840710714576, + "learning_rate": 2.455482320713953e-08, + "loss": 0.9479, + "step": 10551 + }, + { + "epoch": 0.9516165396582045, + "grad_norm": 1.6735663726792604, + "learning_rate": 2.4463643223984643e-08, + "loss": 1.0053, + "step": 10552 + }, + { + "epoch": 0.9517067231816747, + "grad_norm": 1.8811315099971222, + "learning_rate": 2.4372631804484567e-08, + "loss": 1.022, + "step": 10553 + }, + { + "epoch": 0.9517969067051449, + "grad_norm": 1.6076433810159179, + "learning_rate": 2.4281788956405313e-08, + "loss": 1.0243, + "step": 10554 + }, + { + "epoch": 0.9518870902286153, + "grad_norm": 2.178566921506414, + "learning_rate": 2.4191114687497572e-08, + "loss": 0.8776, + "step": 10555 + }, + { + "epoch": 0.9519772737520855, + "grad_norm": 1.8552973296637711, + "learning_rate": 2.4100609005498706e-08, + "loss": 0.9083, + "step": 10556 + }, + { + "epoch": 0.9520674572755558, + "grad_norm": 1.6861894251389913, + "learning_rate": 2.4010271918130764e-08, + "loss": 0.8773, + "step": 10557 + }, + { + "epoch": 0.952157640799026, + "grad_norm": 1.5157863098506785, + "learning_rate": 2.39201034331018e-08, + "loss": 0.9083, + "step": 10558 + }, + { + "epoch": 0.9522478243224963, + "grad_norm": 1.6911524346538718, + "learning_rate": 2.3830103558105663e-08, + "loss": 0.9407, + "step": 10559 + }, + { + "epoch": 0.9523380078459666, + "grad_norm": 2.132593388184014, + "learning_rate": 2.374027230082154e-08, + "loss": 0.954, + "step": 10560 + }, + { + "epoch": 0.9524281913694368, + "grad_norm": 0.684818870271795, + "learning_rate": 2.365060966891441e-08, + "loss": 0.7963, + "step": 10561 + }, + { + "epoch": 0.952518374892907, + "grad_norm": 1.5970896716563925, + "learning_rate": 2.3561115670034827e-08, + "loss": 0.9428, + "step": 10562 + }, + { + "epoch": 0.9526085584163774, + "grad_norm": 1.4060863399638888, + "learning_rate": 2.3471790311818675e-08, + "loss": 0.9649, + "step": 10563 + }, + { + "epoch": 0.9526987419398476, + "grad_norm": 1.445692270233542, + "learning_rate": 2.338263360188808e-08, + "loss": 0.9349, + "step": 10564 + }, + { + "epoch": 0.9527889254633178, + "grad_norm": 2.0108632463026446, + "learning_rate": 2.329364554784985e-08, + "loss": 0.8845, + "step": 10565 + }, + { + "epoch": 0.9528791089867881, + "grad_norm": 1.656089271088025, + "learning_rate": 2.3204826157297465e-08, + "loss": 1.0051, + "step": 10566 + }, + { + "epoch": 0.9529692925102584, + "grad_norm": 1.557669841390605, + "learning_rate": 2.3116175437809082e-08, + "loss": 0.8836, + "step": 10567 + }, + { + "epoch": 0.9530594760337286, + "grad_norm": 0.6553488402325162, + "learning_rate": 2.30276933969491e-08, + "loss": 0.7972, + "step": 10568 + }, + { + "epoch": 0.9531496595571989, + "grad_norm": 1.494123231873846, + "learning_rate": 2.2939380042267255e-08, + "loss": 1.0019, + "step": 10569 + }, + { + "epoch": 0.9532398430806691, + "grad_norm": 2.027768364395286, + "learning_rate": 2.2851235381298627e-08, + "loss": 0.8533, + "step": 10570 + }, + { + "epoch": 0.9533300266041395, + "grad_norm": 1.6376711457481965, + "learning_rate": 2.2763259421564986e-08, + "loss": 0.8766, + "step": 10571 + }, + { + "epoch": 0.9534202101276097, + "grad_norm": 1.832111348757885, + "learning_rate": 2.2675452170571873e-08, + "loss": 0.9183, + "step": 10572 + }, + { + "epoch": 0.9535103936510799, + "grad_norm": 1.588663098639546, + "learning_rate": 2.2587813635812414e-08, + "loss": 0.9412, + "step": 10573 + }, + { + "epoch": 0.9536005771745502, + "grad_norm": 1.8643761702663983, + "learning_rate": 2.2500343824763958e-08, + "loss": 1.0327, + "step": 10574 + }, + { + "epoch": 0.9536907606980205, + "grad_norm": 1.786948208238219, + "learning_rate": 2.2413042744890088e-08, + "loss": 0.8922, + "step": 10575 + }, + { + "epoch": 0.9537809442214907, + "grad_norm": 1.517963139655661, + "learning_rate": 2.2325910403639514e-08, + "loss": 0.9343, + "step": 10576 + }, + { + "epoch": 0.953871127744961, + "grad_norm": 1.4366416026773892, + "learning_rate": 2.223894680844718e-08, + "loss": 0.9026, + "step": 10577 + }, + { + "epoch": 0.9539613112684313, + "grad_norm": 1.3441134560102688, + "learning_rate": 2.2152151966733146e-08, + "loss": 0.7825, + "step": 10578 + }, + { + "epoch": 0.9540514947919015, + "grad_norm": 2.667042517003826, + "learning_rate": 2.2065525885903267e-08, + "loss": 0.9816, + "step": 10579 + }, + { + "epoch": 0.9541416783153718, + "grad_norm": 1.7360472993779548, + "learning_rate": 2.1979068573348747e-08, + "loss": 0.9886, + "step": 10580 + }, + { + "epoch": 0.954231861838842, + "grad_norm": 1.453317855014679, + "learning_rate": 2.1892780036447013e-08, + "loss": 0.9131, + "step": 10581 + }, + { + "epoch": 0.9543220453623124, + "grad_norm": 1.6234510815871972, + "learning_rate": 2.1806660282560175e-08, + "loss": 1.0935, + "step": 10582 + }, + { + "epoch": 0.9544122288857826, + "grad_norm": 1.3686583100704999, + "learning_rate": 2.1720709319037024e-08, + "loss": 0.9683, + "step": 10583 + }, + { + "epoch": 0.9545024124092528, + "grad_norm": 1.4529557488431344, + "learning_rate": 2.1634927153211023e-08, + "loss": 0.9094, + "step": 10584 + }, + { + "epoch": 0.954592595932723, + "grad_norm": 1.7724890504399295, + "learning_rate": 2.1549313792401437e-08, + "loss": 1.0105, + "step": 10585 + }, + { + "epoch": 0.9546827794561934, + "grad_norm": 1.4261784248546443, + "learning_rate": 2.1463869243913746e-08, + "loss": 1.0393, + "step": 10586 + }, + { + "epoch": 0.9547729629796636, + "grad_norm": 1.638511895428462, + "learning_rate": 2.1378593515037902e-08, + "loss": 0.9269, + "step": 10587 + }, + { + "epoch": 0.9548631465031339, + "grad_norm": 1.4589576565544233, + "learning_rate": 2.129348661305075e-08, + "loss": 1.0045, + "step": 10588 + }, + { + "epoch": 0.9549533300266041, + "grad_norm": 1.5050322837788812, + "learning_rate": 2.1208548545213813e-08, + "loss": 0.9737, + "step": 10589 + }, + { + "epoch": 0.9550435135500744, + "grad_norm": 1.6350481707508373, + "learning_rate": 2.1123779318774404e-08, + "loss": 0.9003, + "step": 10590 + }, + { + "epoch": 0.9551336970735447, + "grad_norm": 1.3375109502754092, + "learning_rate": 2.1039178940965408e-08, + "loss": 0.8885, + "step": 10591 + }, + { + "epoch": 0.9552238805970149, + "grad_norm": 1.4055558504741763, + "learning_rate": 2.0954747419005712e-08, + "loss": 1.0081, + "step": 10592 + }, + { + "epoch": 0.9553140641204851, + "grad_norm": 1.58582219047594, + "learning_rate": 2.087048476009934e-08, + "loss": 0.926, + "step": 10593 + }, + { + "epoch": 0.9554042476439555, + "grad_norm": 1.3913320394149586, + "learning_rate": 2.0786390971435862e-08, + "loss": 0.9199, + "step": 10594 + }, + { + "epoch": 0.9554944311674257, + "grad_norm": 1.7477817781920881, + "learning_rate": 2.070246606019088e-08, + "loss": 0.9112, + "step": 10595 + }, + { + "epoch": 0.955584614690896, + "grad_norm": 3.6320620182052723, + "learning_rate": 2.0618710033525112e-08, + "loss": 0.8598, + "step": 10596 + }, + { + "epoch": 0.9556747982143662, + "grad_norm": 1.61420545360394, + "learning_rate": 2.053512289858528e-08, + "loss": 0.9131, + "step": 10597 + }, + { + "epoch": 0.9557649817378365, + "grad_norm": 1.9828852670963217, + "learning_rate": 2.0451704662503456e-08, + "loss": 0.985, + "step": 10598 + }, + { + "epoch": 0.9558551652613068, + "grad_norm": 1.6075391946763395, + "learning_rate": 2.0368455332397282e-08, + "loss": 1.0215, + "step": 10599 + }, + { + "epoch": 0.955945348784777, + "grad_norm": 1.7112195204924625, + "learning_rate": 2.0285374915369967e-08, + "loss": 0.9608, + "step": 10600 + }, + { + "epoch": 0.9560355323082473, + "grad_norm": 1.532193340964935, + "learning_rate": 2.020246341851073e-08, + "loss": 0.9877, + "step": 10601 + }, + { + "epoch": 0.9561257158317176, + "grad_norm": 1.5223272728618025, + "learning_rate": 2.0119720848893463e-08, + "loss": 0.9907, + "step": 10602 + }, + { + "epoch": 0.9562158993551878, + "grad_norm": 1.5349586107222073, + "learning_rate": 2.0037147213578964e-08, + "loss": 0.896, + "step": 10603 + }, + { + "epoch": 0.956306082878658, + "grad_norm": 1.4564575211170703, + "learning_rate": 1.9954742519612265e-08, + "loss": 0.9762, + "step": 10604 + }, + { + "epoch": 0.9563962664021284, + "grad_norm": 1.7727458798381206, + "learning_rate": 1.9872506774024633e-08, + "loss": 0.9239, + "step": 10605 + }, + { + "epoch": 0.9564864499255986, + "grad_norm": 1.7158769349605365, + "learning_rate": 1.979043998383334e-08, + "loss": 0.9571, + "step": 10606 + }, + { + "epoch": 0.9565766334490688, + "grad_norm": 1.2615990832951371, + "learning_rate": 1.970854215604034e-08, + "loss": 0.8892, + "step": 10607 + }, + { + "epoch": 0.9566668169725391, + "grad_norm": 1.6894551712641148, + "learning_rate": 1.9626813297633826e-08, + "loss": 0.9194, + "step": 10608 + }, + { + "epoch": 0.9567570004960094, + "grad_norm": 1.7712915014817396, + "learning_rate": 1.954525341558688e-08, + "loss": 0.8527, + "step": 10609 + }, + { + "epoch": 0.9568471840194797, + "grad_norm": 1.721132603716284, + "learning_rate": 1.9463862516859498e-08, + "loss": 0.9154, + "step": 10610 + }, + { + "epoch": 0.9569373675429499, + "grad_norm": 1.543698526983787, + "learning_rate": 1.938264060839545e-08, + "loss": 1.0324, + "step": 10611 + }, + { + "epoch": 0.9570275510664201, + "grad_norm": 1.6848135863742608, + "learning_rate": 1.9301587697126086e-08, + "loss": 0.9208, + "step": 10612 + }, + { + "epoch": 0.9571177345898905, + "grad_norm": 1.716709104340564, + "learning_rate": 1.9220703789966318e-08, + "loss": 0.9014, + "step": 10613 + }, + { + "epoch": 0.9572079181133607, + "grad_norm": 1.4846583848787065, + "learning_rate": 1.913998889381818e-08, + "loss": 0.9039, + "step": 10614 + }, + { + "epoch": 0.9572981016368309, + "grad_norm": 1.5708156557715849, + "learning_rate": 1.9059443015568387e-08, + "loss": 0.9561, + "step": 10615 + }, + { + "epoch": 0.9573882851603012, + "grad_norm": 1.4746380733703683, + "learning_rate": 1.8979066162089884e-08, + "loss": 0.9558, + "step": 10616 + }, + { + "epoch": 0.9574784686837715, + "grad_norm": 1.5298686901167806, + "learning_rate": 1.889885834024052e-08, + "loss": 0.951, + "step": 10617 + }, + { + "epoch": 0.9575686522072417, + "grad_norm": 1.5796736115732743, + "learning_rate": 1.8818819556864374e-08, + "loss": 0.8982, + "step": 10618 + }, + { + "epoch": 0.957658835730712, + "grad_norm": 1.6402554777077467, + "learning_rate": 1.873894981879065e-08, + "loss": 0.9319, + "step": 10619 + }, + { + "epoch": 0.9577490192541822, + "grad_norm": 0.6535681466877094, + "learning_rate": 1.8659249132834342e-08, + "loss": 0.7916, + "step": 10620 + }, + { + "epoch": 0.9578392027776526, + "grad_norm": 1.4411812790221452, + "learning_rate": 1.857971750579579e-08, + "loss": 0.9251, + "step": 10621 + }, + { + "epoch": 0.9579293863011228, + "grad_norm": 1.443612083023844, + "learning_rate": 1.8500354944461116e-08, + "loss": 0.8271, + "step": 10622 + }, + { + "epoch": 0.958019569824593, + "grad_norm": 1.3716176654101928, + "learning_rate": 1.8421161455602242e-08, + "loss": 0.953, + "step": 10623 + }, + { + "epoch": 0.9581097533480634, + "grad_norm": 1.4762805487874002, + "learning_rate": 1.834213704597598e-08, + "loss": 0.9721, + "step": 10624 + }, + { + "epoch": 0.9581999368715336, + "grad_norm": 1.418337014877496, + "learning_rate": 1.8263281722325385e-08, + "loss": 0.958, + "step": 10625 + }, + { + "epoch": 0.9582901203950038, + "grad_norm": 1.742641714783307, + "learning_rate": 1.818459549137885e-08, + "loss": 0.957, + "step": 10626 + }, + { + "epoch": 0.9583803039184741, + "grad_norm": 1.5720826605256062, + "learning_rate": 1.8106078359850117e-08, + "loss": 0.9336, + "step": 10627 + }, + { + "epoch": 0.9584704874419444, + "grad_norm": 1.5585875546898722, + "learning_rate": 1.802773033443894e-08, + "loss": 0.964, + "step": 10628 + }, + { + "epoch": 0.9585606709654146, + "grad_norm": 1.5289670313242816, + "learning_rate": 1.7949551421830413e-08, + "loss": 0.9145, + "step": 10629 + }, + { + "epoch": 0.9586508544888849, + "grad_norm": 1.6112255826746857, + "learning_rate": 1.7871541628694752e-08, + "loss": 0.9173, + "step": 10630 + }, + { + "epoch": 0.9587410380123551, + "grad_norm": 1.4707938238498217, + "learning_rate": 1.779370096168864e-08, + "loss": 0.8552, + "step": 10631 + }, + { + "epoch": 0.9588312215358255, + "grad_norm": 1.5820841462511062, + "learning_rate": 1.771602942745387e-08, + "loss": 0.9091, + "step": 10632 + }, + { + "epoch": 0.9589214050592957, + "grad_norm": 1.8714977616651656, + "learning_rate": 1.763852703261759e-08, + "loss": 0.9297, + "step": 10633 + }, + { + "epoch": 0.9590115885827659, + "grad_norm": 1.465738902922317, + "learning_rate": 1.756119378379295e-08, + "loss": 1.0103, + "step": 10634 + }, + { + "epoch": 0.9591017721062362, + "grad_norm": 1.5027201148859561, + "learning_rate": 1.7484029687578005e-08, + "loss": 1.0369, + "step": 10635 + }, + { + "epoch": 0.9591919556297065, + "grad_norm": 1.4380923170398698, + "learning_rate": 1.740703475055727e-08, + "loss": 0.9232, + "step": 10636 + }, + { + "epoch": 0.9592821391531767, + "grad_norm": 1.6348914094613718, + "learning_rate": 1.7330208979300153e-08, + "loss": 0.9818, + "step": 10637 + }, + { + "epoch": 0.959372322676647, + "grad_norm": 1.3883949572398784, + "learning_rate": 1.725355238036208e-08, + "loss": 0.9888, + "step": 10638 + }, + { + "epoch": 0.9594625062001172, + "grad_norm": 3.6205627644670533, + "learning_rate": 1.7177064960283594e-08, + "loss": 0.9873, + "step": 10639 + }, + { + "epoch": 0.9595526897235875, + "grad_norm": 1.850958681437815, + "learning_rate": 1.7100746725591253e-08, + "loss": 0.9433, + "step": 10640 + }, + { + "epoch": 0.9596428732470578, + "grad_norm": 2.036709300883696, + "learning_rate": 1.7024597682796517e-08, + "loss": 0.9354, + "step": 10641 + }, + { + "epoch": 0.959733056770528, + "grad_norm": 1.546274087729056, + "learning_rate": 1.6948617838397293e-08, + "loss": 1.0177, + "step": 10642 + }, + { + "epoch": 0.9598232402939982, + "grad_norm": 1.296166651666213, + "learning_rate": 1.6872807198876404e-08, + "loss": 0.9739, + "step": 10643 + }, + { + "epoch": 0.9599134238174686, + "grad_norm": 1.7117541728670145, + "learning_rate": 1.679716577070245e-08, + "loss": 1.0503, + "step": 10644 + }, + { + "epoch": 0.9600036073409388, + "grad_norm": 1.5125337518126047, + "learning_rate": 1.6721693560329596e-08, + "loss": 1.0322, + "step": 10645 + }, + { + "epoch": 0.960093790864409, + "grad_norm": 1.528808749700567, + "learning_rate": 1.6646390574197366e-08, + "loss": 0.8577, + "step": 10646 + }, + { + "epoch": 0.9601839743878793, + "grad_norm": 0.7067347162886048, + "learning_rate": 1.6571256818731504e-08, + "loss": 0.7971, + "step": 10647 + }, + { + "epoch": 0.9602741579113496, + "grad_norm": 1.5208869123383468, + "learning_rate": 1.6496292300342218e-08, + "loss": 0.8822, + "step": 10648 + }, + { + "epoch": 0.9603643414348199, + "grad_norm": 1.6772004633530992, + "learning_rate": 1.642149702542639e-08, + "loss": 0.8856, + "step": 10649 + }, + { + "epoch": 0.9604545249582901, + "grad_norm": 1.3356499613463044, + "learning_rate": 1.634687100036558e-08, + "loss": 0.9083, + "step": 10650 + }, + { + "epoch": 0.9605447084817604, + "grad_norm": 1.621853128921544, + "learning_rate": 1.627241423152781e-08, + "loss": 0.8808, + "step": 10651 + }, + { + "epoch": 0.9606348920052307, + "grad_norm": 1.6559767587661762, + "learning_rate": 1.619812672526555e-08, + "loss": 0.9987, + "step": 10652 + }, + { + "epoch": 0.9607250755287009, + "grad_norm": 1.581848133933642, + "learning_rate": 1.6124008487917727e-08, + "loss": 0.9296, + "step": 10653 + }, + { + "epoch": 0.9608152590521711, + "grad_norm": 0.6570504393253087, + "learning_rate": 1.6050059525808623e-08, + "loss": 0.796, + "step": 10654 + }, + { + "epoch": 0.9609054425756415, + "grad_norm": 0.6385328275640237, + "learning_rate": 1.597627984524763e-08, + "loss": 0.7917, + "step": 10655 + }, + { + "epoch": 0.9609956260991117, + "grad_norm": 1.415241993002587, + "learning_rate": 1.590266945253038e-08, + "loss": 0.9862, + "step": 10656 + }, + { + "epoch": 0.9610858096225819, + "grad_norm": 1.4335679719774508, + "learning_rate": 1.582922835393763e-08, + "loss": 0.9621, + "step": 10657 + }, + { + "epoch": 0.9611759931460522, + "grad_norm": 1.462099806325973, + "learning_rate": 1.5755956555735473e-08, + "loss": 0.923, + "step": 10658 + }, + { + "epoch": 0.9612661766695225, + "grad_norm": 0.756297573453306, + "learning_rate": 1.5682854064176244e-08, + "loss": 0.8658, + "step": 10659 + }, + { + "epoch": 0.9613563601929928, + "grad_norm": 1.928113225718551, + "learning_rate": 1.5609920885497395e-08, + "loss": 0.879, + "step": 10660 + }, + { + "epoch": 0.961446543716463, + "grad_norm": 1.5719695955786894, + "learning_rate": 1.5537157025921732e-08, + "loss": 0.9714, + "step": 10661 + }, + { + "epoch": 0.9615367272399332, + "grad_norm": 1.7630561438160608, + "learning_rate": 1.5464562491658285e-08, + "loss": 0.9227, + "step": 10662 + }, + { + "epoch": 0.9616269107634036, + "grad_norm": 1.688965337818842, + "learning_rate": 1.5392137288900764e-08, + "loss": 0.9866, + "step": 10663 + }, + { + "epoch": 0.9617170942868738, + "grad_norm": 0.6139753993337312, + "learning_rate": 1.531988142382934e-08, + "loss": 0.7809, + "step": 10664 + }, + { + "epoch": 0.961807277810344, + "grad_norm": 1.546607948139545, + "learning_rate": 1.5247794902608634e-08, + "loss": 0.8795, + "step": 10665 + }, + { + "epoch": 0.9618974613338143, + "grad_norm": 1.9399846665277833, + "learning_rate": 1.5175877731390398e-08, + "loss": 0.903, + "step": 10666 + }, + { + "epoch": 0.9619876448572846, + "grad_norm": 1.831281691657799, + "learning_rate": 1.510412991631016e-08, + "loss": 0.9752, + "step": 10667 + }, + { + "epoch": 0.9620778283807548, + "grad_norm": 1.3256872805830409, + "learning_rate": 1.503255146349014e-08, + "loss": 0.9764, + "step": 10668 + }, + { + "epoch": 0.9621680119042251, + "grad_norm": 1.6335128669290393, + "learning_rate": 1.4961142379037893e-08, + "loss": 0.9086, + "step": 10669 + }, + { + "epoch": 0.9622581954276953, + "grad_norm": 1.4536670781214942, + "learning_rate": 1.4889902669046327e-08, + "loss": 0.9851, + "step": 10670 + }, + { + "epoch": 0.9623483789511657, + "grad_norm": 1.9426613018762104, + "learning_rate": 1.4818832339594135e-08, + "loss": 0.9471, + "step": 10671 + }, + { + "epoch": 0.9624385624746359, + "grad_norm": 1.3438278441836073, + "learning_rate": 1.474793139674535e-08, + "loss": 0.8941, + "step": 10672 + }, + { + "epoch": 0.9625287459981061, + "grad_norm": 1.5383186672885094, + "learning_rate": 1.4677199846549581e-08, + "loss": 0.9455, + "step": 10673 + }, + { + "epoch": 0.9626189295215765, + "grad_norm": 1.6128645906127979, + "learning_rate": 1.4606637695042224e-08, + "loss": 0.9462, + "step": 10674 + }, + { + "epoch": 0.9627091130450467, + "grad_norm": 1.7526075826973426, + "learning_rate": 1.4536244948243793e-08, + "loss": 1.0368, + "step": 10675 + }, + { + "epoch": 0.9627992965685169, + "grad_norm": 1.4999270421405118, + "learning_rate": 1.4466021612160595e-08, + "loss": 0.9122, + "step": 10676 + }, + { + "epoch": 0.9628894800919872, + "grad_norm": 1.5669957143436488, + "learning_rate": 1.4395967692784505e-08, + "loss": 0.9134, + "step": 10677 + }, + { + "epoch": 0.9629796636154575, + "grad_norm": 1.436499703714969, + "learning_rate": 1.4326083196092963e-08, + "loss": 0.8738, + "step": 10678 + }, + { + "epoch": 0.9630698471389277, + "grad_norm": 1.5009630608432272, + "learning_rate": 1.42563681280492e-08, + "loss": 0.9575, + "step": 10679 + }, + { + "epoch": 0.963160030662398, + "grad_norm": 1.54201159184177, + "learning_rate": 1.4186822494600902e-08, + "loss": 0.9219, + "step": 10680 + }, + { + "epoch": 0.9632502141858682, + "grad_norm": 1.3431858830345715, + "learning_rate": 1.4117446301682877e-08, + "loss": 0.9346, + "step": 10681 + }, + { + "epoch": 0.9633403977093385, + "grad_norm": 1.4554352802025923, + "learning_rate": 1.4048239555214392e-08, + "loss": 0.9011, + "step": 10682 + }, + { + "epoch": 0.9634305812328088, + "grad_norm": 1.4992406035436217, + "learning_rate": 1.3979202261100497e-08, + "loss": 0.9749, + "step": 10683 + }, + { + "epoch": 0.963520764756279, + "grad_norm": 1.6211069864858707, + "learning_rate": 1.3910334425231817e-08, + "loss": 1.0042, + "step": 10684 + }, + { + "epoch": 0.9636109482797492, + "grad_norm": 1.4629118119502136, + "learning_rate": 1.384163605348454e-08, + "loss": 0.9436, + "step": 10685 + }, + { + "epoch": 0.9637011318032196, + "grad_norm": 1.7627699056888342, + "learning_rate": 1.3773107151720642e-08, + "loss": 1.0018, + "step": 10686 + }, + { + "epoch": 0.9637913153266898, + "grad_norm": 1.671484857119183, + "learning_rate": 1.3704747725787003e-08, + "loss": 1.0039, + "step": 10687 + }, + { + "epoch": 0.9638814988501601, + "grad_norm": 0.7392491852875478, + "learning_rate": 1.3636557781516512e-08, + "loss": 0.7894, + "step": 10688 + }, + { + "epoch": 0.9639716823736303, + "grad_norm": 1.66750813540829, + "learning_rate": 1.3568537324727847e-08, + "loss": 0.9513, + "step": 10689 + }, + { + "epoch": 0.9640618658971006, + "grad_norm": 1.3739401026806406, + "learning_rate": 1.3500686361224589e-08, + "loss": 0.9407, + "step": 10690 + }, + { + "epoch": 0.9641520494205709, + "grad_norm": 1.8070229615492672, + "learning_rate": 1.3433004896796108e-08, + "loss": 0.9418, + "step": 10691 + }, + { + "epoch": 0.9642422329440411, + "grad_norm": 1.3896694114748234, + "learning_rate": 1.336549293721756e-08, + "loss": 0.8353, + "step": 10692 + }, + { + "epoch": 0.9643324164675113, + "grad_norm": 0.6837143927792353, + "learning_rate": 1.3298150488249227e-08, + "loss": 0.7737, + "step": 10693 + }, + { + "epoch": 0.9644225999909817, + "grad_norm": 1.8940914941458875, + "learning_rate": 1.3230977555637401e-08, + "loss": 1.0273, + "step": 10694 + }, + { + "epoch": 0.9645127835144519, + "grad_norm": 1.2733770718893285, + "learning_rate": 1.3163974145113499e-08, + "loss": 0.8995, + "step": 10695 + }, + { + "epoch": 0.9646029670379221, + "grad_norm": 2.1973898619391106, + "learning_rate": 1.3097140262394723e-08, + "loss": 0.8454, + "step": 10696 + }, + { + "epoch": 0.9646931505613925, + "grad_norm": 1.8552838363862032, + "learning_rate": 1.303047591318318e-08, + "loss": 0.9753, + "step": 10697 + }, + { + "epoch": 0.9647833340848627, + "grad_norm": 1.5623800613146765, + "learning_rate": 1.2963981103167875e-08, + "loss": 0.8782, + "step": 10698 + }, + { + "epoch": 0.964873517608333, + "grad_norm": 1.3980345758247241, + "learning_rate": 1.2897655838021825e-08, + "loss": 0.9294, + "step": 10699 + }, + { + "epoch": 0.9649637011318032, + "grad_norm": 1.5676806205589746, + "learning_rate": 1.2831500123404726e-08, + "loss": 0.951, + "step": 10700 + }, + { + "epoch": 0.9650538846552735, + "grad_norm": 1.7533214248257654, + "learning_rate": 1.2765513964961172e-08, + "loss": 1.0954, + "step": 10701 + }, + { + "epoch": 0.9651440681787438, + "grad_norm": 1.508639252271282, + "learning_rate": 1.2699697368321549e-08, + "loss": 0.941, + "step": 10702 + }, + { + "epoch": 0.965234251702214, + "grad_norm": 1.474791660789335, + "learning_rate": 1.2634050339101366e-08, + "loss": 0.9325, + "step": 10703 + }, + { + "epoch": 0.9653244352256842, + "grad_norm": 1.2915421651498824, + "learning_rate": 1.2568572882902361e-08, + "loss": 0.922, + "step": 10704 + }, + { + "epoch": 0.9654146187491546, + "grad_norm": 1.5939535871136006, + "learning_rate": 1.2503265005311402e-08, + "loss": 0.917, + "step": 10705 + }, + { + "epoch": 0.9655048022726248, + "grad_norm": 1.6148935901792194, + "learning_rate": 1.2438126711900698e-08, + "loss": 0.9075, + "step": 10706 + }, + { + "epoch": 0.965594985796095, + "grad_norm": 1.4205213537977734, + "learning_rate": 1.2373158008228247e-08, + "loss": 0.7954, + "step": 10707 + }, + { + "epoch": 0.9656851693195653, + "grad_norm": 1.3450231509006603, + "learning_rate": 1.2308358899837833e-08, + "loss": 0.9221, + "step": 10708 + }, + { + "epoch": 0.9657753528430356, + "grad_norm": 1.8644865923742413, + "learning_rate": 1.224372939225815e-08, + "loss": 0.9248, + "step": 10709 + }, + { + "epoch": 0.9658655363665059, + "grad_norm": 1.4802814455743112, + "learning_rate": 1.2179269491003674e-08, + "loss": 1.0278, + "step": 10710 + }, + { + "epoch": 0.9659557198899761, + "grad_norm": 1.4764986377610312, + "learning_rate": 1.2114979201574894e-08, + "loss": 0.8197, + "step": 10711 + }, + { + "epoch": 0.9660459034134463, + "grad_norm": 1.317851419432732, + "learning_rate": 1.2050858529456975e-08, + "loss": 0.9663, + "step": 10712 + }, + { + "epoch": 0.9661360869369167, + "grad_norm": 1.528230999661709, + "learning_rate": 1.1986907480121545e-08, + "loss": 0.945, + "step": 10713 + }, + { + "epoch": 0.9662262704603869, + "grad_norm": 1.5435647706633209, + "learning_rate": 1.192312605902468e-08, + "loss": 0.9652, + "step": 10714 + }, + { + "epoch": 0.9663164539838571, + "grad_norm": 0.8085405267715446, + "learning_rate": 1.1859514271608917e-08, + "loss": 0.8031, + "step": 10715 + }, + { + "epoch": 0.9664066375073274, + "grad_norm": 2.5786580777299735, + "learning_rate": 1.1796072123301914e-08, + "loss": 0.9828, + "step": 10716 + }, + { + "epoch": 0.9664968210307977, + "grad_norm": 0.689092049862532, + "learning_rate": 1.1732799619516897e-08, + "loss": 0.7832, + "step": 10717 + }, + { + "epoch": 0.9665870045542679, + "grad_norm": 2.0138559784520393, + "learning_rate": 1.1669696765652659e-08, + "loss": 0.9321, + "step": 10718 + }, + { + "epoch": 0.9666771880777382, + "grad_norm": 1.5191642102200555, + "learning_rate": 1.1606763567093336e-08, + "loss": 0.9765, + "step": 10719 + }, + { + "epoch": 0.9667673716012085, + "grad_norm": 0.6917942704033783, + "learning_rate": 1.1544000029208857e-08, + "loss": 0.7736, + "step": 10720 + }, + { + "epoch": 0.9668575551246787, + "grad_norm": 1.7062812746416545, + "learning_rate": 1.148140615735449e-08, + "loss": 0.8063, + "step": 10721 + }, + { + "epoch": 0.966947738648149, + "grad_norm": 2.0090076733713684, + "learning_rate": 1.1418981956871076e-08, + "loss": 0.9184, + "step": 10722 + }, + { + "epoch": 0.9670379221716192, + "grad_norm": 1.9360145288741928, + "learning_rate": 1.1356727433085245e-08, + "loss": 0.9346, + "step": 10723 + }, + { + "epoch": 0.9671281056950896, + "grad_norm": 1.3800612845150515, + "learning_rate": 1.1294642591308524e-08, + "loss": 0.9694, + "step": 10724 + }, + { + "epoch": 0.9672182892185598, + "grad_norm": 1.3923465059972522, + "learning_rate": 1.1232727436838452e-08, + "loss": 0.9353, + "step": 10725 + }, + { + "epoch": 0.96730847274203, + "grad_norm": 1.599468145784871, + "learning_rate": 1.1170981974958138e-08, + "loss": 0.9905, + "step": 10726 + }, + { + "epoch": 0.9673986562655003, + "grad_norm": 1.5255291786716627, + "learning_rate": 1.1109406210936035e-08, + "loss": 1.0067, + "step": 10727 + }, + { + "epoch": 0.9674888397889706, + "grad_norm": 1.772444594126614, + "learning_rate": 1.1048000150025939e-08, + "loss": 1.0605, + "step": 10728 + }, + { + "epoch": 0.9675790233124408, + "grad_norm": 1.6270425137696216, + "learning_rate": 1.0986763797467213e-08, + "loss": 0.8386, + "step": 10729 + }, + { + "epoch": 0.9676692068359111, + "grad_norm": 0.6004164293709189, + "learning_rate": 1.0925697158485459e-08, + "loss": 0.7445, + "step": 10730 + }, + { + "epoch": 0.9677593903593813, + "grad_norm": 1.4775439840311706, + "learning_rate": 1.0864800238290727e-08, + "loss": 0.9497, + "step": 10731 + }, + { + "epoch": 0.9678495738828516, + "grad_norm": 1.5442042558887827, + "learning_rate": 1.0804073042079309e-08, + "loss": 0.9084, + "step": 10732 + }, + { + "epoch": 0.9679397574063219, + "grad_norm": 0.6525314426519573, + "learning_rate": 1.0743515575032392e-08, + "loss": 0.7902, + "step": 10733 + }, + { + "epoch": 0.9680299409297921, + "grad_norm": 1.4927196252914425, + "learning_rate": 1.0683127842317619e-08, + "loss": 0.9561, + "step": 10734 + }, + { + "epoch": 0.9681201244532623, + "grad_norm": 1.5734138797627373, + "learning_rate": 1.0622909849087314e-08, + "loss": 0.9746, + "step": 10735 + }, + { + "epoch": 0.9682103079767327, + "grad_norm": 1.4080985740758973, + "learning_rate": 1.0562861600479588e-08, + "loss": 1.0051, + "step": 10736 + }, + { + "epoch": 0.9683004915002029, + "grad_norm": 1.6175514585631745, + "learning_rate": 1.0502983101618345e-08, + "loss": 1.0367, + "step": 10737 + }, + { + "epoch": 0.9683906750236732, + "grad_norm": 1.5379673503391509, + "learning_rate": 1.0443274357612386e-08, + "loss": 0.8262, + "step": 10738 + }, + { + "epoch": 0.9684808585471434, + "grad_norm": 1.8705064970921814, + "learning_rate": 1.0383735373556524e-08, + "loss": 0.9617, + "step": 10739 + }, + { + "epoch": 0.9685710420706137, + "grad_norm": 1.2733225858361277, + "learning_rate": 1.0324366154531139e-08, + "loss": 0.9735, + "step": 10740 + }, + { + "epoch": 0.968661225594084, + "grad_norm": 2.116044077084249, + "learning_rate": 1.0265166705601735e-08, + "loss": 0.9148, + "step": 10741 + }, + { + "epoch": 0.9687514091175542, + "grad_norm": 0.6918486995229638, + "learning_rate": 1.0206137031819606e-08, + "loss": 0.8117, + "step": 10742 + }, + { + "epoch": 0.9688415926410244, + "grad_norm": 1.4228968878172392, + "learning_rate": 1.0147277138221388e-08, + "loss": 0.9565, + "step": 10743 + }, + { + "epoch": 0.9689317761644948, + "grad_norm": 1.656346588197396, + "learning_rate": 1.0088587029829287e-08, + "loss": 0.9826, + "step": 10744 + }, + { + "epoch": 0.969021959687965, + "grad_norm": 1.5087798973224866, + "learning_rate": 1.003006671165152e-08, + "loss": 1.0488, + "step": 10745 + }, + { + "epoch": 0.9691121432114352, + "grad_norm": 1.5023856265797304, + "learning_rate": 9.971716188680978e-09, + "loss": 0.9388, + "step": 10746 + }, + { + "epoch": 0.9692023267349056, + "grad_norm": 2.0863468754277017, + "learning_rate": 9.91353546589635e-09, + "loss": 0.9139, + "step": 10747 + }, + { + "epoch": 0.9692925102583758, + "grad_norm": 1.4502106020342995, + "learning_rate": 9.855524548262106e-09, + "loss": 0.8807, + "step": 10748 + }, + { + "epoch": 0.969382693781846, + "grad_norm": 1.610615520693647, + "learning_rate": 9.797683440728288e-09, + "loss": 0.8917, + "step": 10749 + }, + { + "epoch": 0.9694728773053163, + "grad_norm": 1.507968084184782, + "learning_rate": 9.740012148229836e-09, + "loss": 0.8598, + "step": 10750 + }, + { + "epoch": 0.9695630608287866, + "grad_norm": 1.391621007950763, + "learning_rate": 9.682510675687705e-09, + "loss": 0.8975, + "step": 10751 + }, + { + "epoch": 0.9696532443522569, + "grad_norm": 1.7661526322034735, + "learning_rate": 9.625179028008191e-09, + "loss": 0.8693, + "step": 10752 + }, + { + "epoch": 0.9697434278757271, + "grad_norm": 1.6053529126778212, + "learning_rate": 9.568017210083379e-09, + "loss": 0.7926, + "step": 10753 + }, + { + "epoch": 0.9698336113991973, + "grad_norm": 1.5028308540947937, + "learning_rate": 9.511025226790259e-09, + "loss": 0.9203, + "step": 10754 + }, + { + "epoch": 0.9699237949226677, + "grad_norm": 1.4978513428402744, + "learning_rate": 9.454203082992052e-09, + "loss": 0.9208, + "step": 10755 + }, + { + "epoch": 0.9700139784461379, + "grad_norm": 1.7709338963695593, + "learning_rate": 9.3975507835371e-09, + "loss": 0.8532, + "step": 10756 + }, + { + "epoch": 0.9701041619696081, + "grad_norm": 0.642463441951609, + "learning_rate": 9.341068333259094e-09, + "loss": 0.766, + "step": 10757 + }, + { + "epoch": 0.9701943454930784, + "grad_norm": 1.5839129524732642, + "learning_rate": 9.28475573697729e-09, + "loss": 0.9214, + "step": 10758 + }, + { + "epoch": 0.9702845290165487, + "grad_norm": 1.543977509077839, + "learning_rate": 9.228612999497177e-09, + "loss": 0.942, + "step": 10759 + }, + { + "epoch": 0.970374712540019, + "grad_norm": 2.3655853389101194, + "learning_rate": 9.172640125608478e-09, + "loss": 0.8875, + "step": 10760 + }, + { + "epoch": 0.9704648960634892, + "grad_norm": 1.2666967116523644, + "learning_rate": 9.116837120087817e-09, + "loss": 0.957, + "step": 10761 + }, + { + "epoch": 0.9705550795869594, + "grad_norm": 2.6591589762609256, + "learning_rate": 9.061203987695832e-09, + "loss": 1.0119, + "step": 10762 + }, + { + "epoch": 0.9706452631104298, + "grad_norm": 1.9779600849077827, + "learning_rate": 9.005740733180055e-09, + "loss": 0.8903, + "step": 10763 + }, + { + "epoch": 0.9707354466339, + "grad_norm": 1.6498715639700703, + "learning_rate": 8.950447361272483e-09, + "loss": 1.0304, + "step": 10764 + }, + { + "epoch": 0.9708256301573702, + "grad_norm": 1.7699371613377206, + "learning_rate": 8.895323876691784e-09, + "loss": 0.9508, + "step": 10765 + }, + { + "epoch": 0.9709158136808405, + "grad_norm": 1.893220932079879, + "learning_rate": 8.840370284140419e-09, + "loss": 0.9957, + "step": 10766 + }, + { + "epoch": 0.9710059972043108, + "grad_norm": 1.6396053960625292, + "learning_rate": 8.78558658830797e-09, + "loss": 0.8599, + "step": 10767 + }, + { + "epoch": 0.971096180727781, + "grad_norm": 0.7201593927276363, + "learning_rate": 8.730972793868696e-09, + "loss": 0.7785, + "step": 10768 + }, + { + "epoch": 0.9711863642512513, + "grad_norm": 1.550682391419609, + "learning_rate": 8.67652890548265e-09, + "loss": 0.9099, + "step": 10769 + }, + { + "epoch": 0.9712765477747216, + "grad_norm": 1.4197974485007667, + "learning_rate": 8.622254927795004e-09, + "loss": 0.9243, + "step": 10770 + }, + { + "epoch": 0.9713667312981918, + "grad_norm": 1.4443300774528265, + "learning_rate": 8.568150865436941e-09, + "loss": 0.8974, + "step": 10771 + }, + { + "epoch": 0.9714569148216621, + "grad_norm": 1.3277258778160175, + "learning_rate": 8.514216723024991e-09, + "loss": 0.9546, + "step": 10772 + }, + { + "epoch": 0.9715470983451323, + "grad_norm": 2.7036395162088374, + "learning_rate": 8.460452505161031e-09, + "loss": 0.8682, + "step": 10773 + }, + { + "epoch": 0.9716372818686027, + "grad_norm": 1.4723621347909577, + "learning_rate": 8.4068582164325e-09, + "loss": 0.9709, + "step": 10774 + }, + { + "epoch": 0.9717274653920729, + "grad_norm": 1.8220945484448083, + "learning_rate": 8.353433861412406e-09, + "loss": 0.9245, + "step": 10775 + }, + { + "epoch": 0.9718176489155431, + "grad_norm": 1.461401717742708, + "learning_rate": 8.300179444658883e-09, + "loss": 0.9, + "step": 10776 + }, + { + "epoch": 0.9719078324390134, + "grad_norm": 3.103043133134986, + "learning_rate": 8.247094970716296e-09, + "loss": 1.0374, + "step": 10777 + }, + { + "epoch": 0.9719980159624837, + "grad_norm": 1.397048132149854, + "learning_rate": 8.19418044411413e-09, + "loss": 0.9767, + "step": 10778 + }, + { + "epoch": 0.9720881994859539, + "grad_norm": 1.47700752131898, + "learning_rate": 8.141435869367219e-09, + "loss": 0.9431, + "step": 10779 + }, + { + "epoch": 0.9721783830094242, + "grad_norm": 1.5591668244196482, + "learning_rate": 8.088861250975742e-09, + "loss": 0.9801, + "step": 10780 + }, + { + "epoch": 0.9722685665328944, + "grad_norm": 1.5623851733930567, + "learning_rate": 8.036456593426111e-09, + "loss": 0.9109, + "step": 10781 + }, + { + "epoch": 0.9723587500563647, + "grad_norm": 1.7231089025388224, + "learning_rate": 7.984221901189415e-09, + "loss": 0.9791, + "step": 10782 + }, + { + "epoch": 0.972448933579835, + "grad_norm": 1.4884828445988758, + "learning_rate": 7.932157178722976e-09, + "loss": 0.8687, + "step": 10783 + }, + { + "epoch": 0.9725391171033052, + "grad_norm": 3.0108899511156983, + "learning_rate": 7.880262430468799e-09, + "loss": 0.9124, + "step": 10784 + }, + { + "epoch": 0.9726293006267754, + "grad_norm": 1.5973782845515676, + "learning_rate": 7.828537660855339e-09, + "loss": 0.8758, + "step": 10785 + }, + { + "epoch": 0.9727194841502458, + "grad_norm": 1.691011448830999, + "learning_rate": 7.776982874295512e-09, + "loss": 0.96, + "step": 10786 + }, + { + "epoch": 0.972809667673716, + "grad_norm": 2.0875105623446295, + "learning_rate": 7.725598075188688e-09, + "loss": 0.8557, + "step": 10787 + }, + { + "epoch": 0.9728998511971862, + "grad_norm": 1.3985348886005216, + "learning_rate": 7.674383267918916e-09, + "loss": 0.891, + "step": 10788 + }, + { + "epoch": 0.9729900347206565, + "grad_norm": 1.5769777565828504, + "learning_rate": 7.623338456856476e-09, + "loss": 0.9296, + "step": 10789 + }, + { + "epoch": 0.9730802182441268, + "grad_norm": 1.426135460736888, + "learning_rate": 7.572463646356554e-09, + "loss": 0.837, + "step": 10790 + }, + { + "epoch": 0.9731704017675971, + "grad_norm": 1.5914902467803325, + "learning_rate": 7.521758840760339e-09, + "loss": 0.9112, + "step": 10791 + }, + { + "epoch": 0.9732605852910673, + "grad_norm": 1.8964161728009006, + "learning_rate": 7.471224044393931e-09, + "loss": 0.8998, + "step": 10792 + }, + { + "epoch": 0.9733507688145376, + "grad_norm": 1.4282234909185596, + "learning_rate": 7.420859261569434e-09, + "loss": 0.9997, + "step": 10793 + }, + { + "epoch": 0.9734409523380079, + "grad_norm": 1.6476844405344198, + "learning_rate": 7.370664496584078e-09, + "loss": 0.9987, + "step": 10794 + }, + { + "epoch": 0.9735311358614781, + "grad_norm": 1.6309428484906283, + "learning_rate": 7.3206397537211026e-09, + "loss": 0.8651, + "step": 10795 + }, + { + "epoch": 0.9736213193849483, + "grad_norm": 1.6668709709196023, + "learning_rate": 7.270785037248428e-09, + "loss": 0.9056, + "step": 10796 + }, + { + "epoch": 0.9737115029084187, + "grad_norm": 1.7698403734761108, + "learning_rate": 7.221100351420428e-09, + "loss": 0.9085, + "step": 10797 + }, + { + "epoch": 0.9738016864318889, + "grad_norm": 1.6024942060535448, + "learning_rate": 7.171585700475935e-09, + "loss": 0.8708, + "step": 10798 + }, + { + "epoch": 0.9738918699553591, + "grad_norm": 1.964893924383446, + "learning_rate": 7.122241088640235e-09, + "loss": 0.999, + "step": 10799 + }, + { + "epoch": 0.9739820534788294, + "grad_norm": 1.6270433929790313, + "learning_rate": 7.073066520123516e-09, + "loss": 1.0682, + "step": 10800 + }, + { + "epoch": 0.9740722370022997, + "grad_norm": 1.7077961829296842, + "learning_rate": 7.0240619991217555e-09, + "loss": 0.8338, + "step": 10801 + }, + { + "epoch": 0.97416242052577, + "grad_norm": 1.7391718608130975, + "learning_rate": 6.975227529816052e-09, + "loss": 1.0154, + "step": 10802 + }, + { + "epoch": 0.9742526040492402, + "grad_norm": 1.6518802939772885, + "learning_rate": 6.926563116373296e-09, + "loss": 0.9851, + "step": 10803 + }, + { + "epoch": 0.9743427875727104, + "grad_norm": 1.43115950718951, + "learning_rate": 6.878068762945943e-09, + "loss": 0.8823, + "step": 10804 + }, + { + "epoch": 0.9744329710961808, + "grad_norm": 1.4838867990858402, + "learning_rate": 6.829744473671794e-09, + "loss": 0.979, + "step": 10805 + }, + { + "epoch": 0.974523154619651, + "grad_norm": 0.7181843314631683, + "learning_rate": 6.781590252674219e-09, + "loss": 0.6807, + "step": 10806 + }, + { + "epoch": 0.9746133381431212, + "grad_norm": 1.258664618497195, + "learning_rate": 6.733606104061484e-09, + "loss": 0.9517, + "step": 10807 + }, + { + "epoch": 0.9747035216665915, + "grad_norm": 1.4365785382022591, + "learning_rate": 6.6857920319283165e-09, + "loss": 0.8848, + "step": 10808 + }, + { + "epoch": 0.9747937051900618, + "grad_norm": 1.5759392132920924, + "learning_rate": 6.638148040354563e-09, + "loss": 0.915, + "step": 10809 + }, + { + "epoch": 0.974883888713532, + "grad_norm": 1.5874106044079126, + "learning_rate": 6.590674133405194e-09, + "loss": 0.9005, + "step": 10810 + }, + { + "epoch": 0.9749740722370023, + "grad_norm": 1.4471385246199089, + "learning_rate": 6.5433703151311914e-09, + "loss": 0.9547, + "step": 10811 + }, + { + "epoch": 0.9750642557604725, + "grad_norm": 1.6208046528976323, + "learning_rate": 6.49623658956866e-09, + "loss": 0.9448, + "step": 10812 + }, + { + "epoch": 0.9751544392839429, + "grad_norm": 1.6720296485179535, + "learning_rate": 6.44927296073905e-09, + "loss": 0.8183, + "step": 10813 + }, + { + "epoch": 0.9752446228074131, + "grad_norm": 1.5314826010158769, + "learning_rate": 6.402479432649821e-09, + "loss": 1.0269, + "step": 10814 + }, + { + "epoch": 0.9753348063308833, + "grad_norm": 2.5805218276808293, + "learning_rate": 6.355856009293781e-09, + "loss": 0.9367, + "step": 10815 + }, + { + "epoch": 0.9754249898543537, + "grad_norm": 1.9855038893581578, + "learning_rate": 6.3094026946488575e-09, + "loss": 0.9118, + "step": 10816 + }, + { + "epoch": 0.9755151733778239, + "grad_norm": 1.4264738725956385, + "learning_rate": 6.2631194926787704e-09, + "loss": 0.9442, + "step": 10817 + }, + { + "epoch": 0.9756053569012941, + "grad_norm": 3.6793899577818308, + "learning_rate": 6.217006407332581e-09, + "loss": 0.8442, + "step": 10818 + }, + { + "epoch": 0.9756955404247644, + "grad_norm": 1.370753188916229, + "learning_rate": 6.1710634425453654e-09, + "loss": 0.8893, + "step": 10819 + }, + { + "epoch": 0.9757857239482347, + "grad_norm": 1.7298050893012935, + "learning_rate": 6.1252906022366544e-09, + "loss": 0.9615, + "step": 10820 + }, + { + "epoch": 0.9758759074717049, + "grad_norm": 1.3955515486935748, + "learning_rate": 6.079687890312213e-09, + "loss": 0.9234, + "step": 10821 + }, + { + "epoch": 0.9759660909951752, + "grad_norm": 1.7064736020639217, + "learning_rate": 6.034255310663372e-09, + "loss": 0.9189, + "step": 10822 + }, + { + "epoch": 0.9760562745186454, + "grad_norm": 2.113204982883255, + "learning_rate": 5.988992867166143e-09, + "loss": 1.0319, + "step": 10823 + }, + { + "epoch": 0.9761464580421157, + "grad_norm": 1.7836366692557704, + "learning_rate": 5.943900563682991e-09, + "loss": 0.9768, + "step": 10824 + }, + { + "epoch": 0.976236641565586, + "grad_norm": 0.7513028987216767, + "learning_rate": 5.898978404061506e-09, + "loss": 0.8571, + "step": 10825 + }, + { + "epoch": 0.9763268250890562, + "grad_norm": 2.0653536158050736, + "learning_rate": 5.85422639213462e-09, + "loss": 0.9393, + "step": 10826 + }, + { + "epoch": 0.9764170086125264, + "grad_norm": 1.5706343439615857, + "learning_rate": 5.809644531720614e-09, + "loss": 0.9523, + "step": 10827 + }, + { + "epoch": 0.9765071921359968, + "grad_norm": 1.4837290914551902, + "learning_rate": 5.765232826623556e-09, + "loss": 0.9673, + "step": 10828 + }, + { + "epoch": 0.976597375659467, + "grad_norm": 1.6666488964405033, + "learning_rate": 5.720991280633081e-09, + "loss": 0.923, + "step": 10829 + }, + { + "epoch": 0.9766875591829373, + "grad_norm": 1.457775099950233, + "learning_rate": 5.676919897523724e-09, + "loss": 0.9255, + "step": 10830 + }, + { + "epoch": 0.9767777427064075, + "grad_norm": 1.3362891643412285, + "learning_rate": 5.633018681056256e-09, + "loss": 0.9474, + "step": 10831 + }, + { + "epoch": 0.9768679262298778, + "grad_norm": 1.9998356036331681, + "learning_rate": 5.589287634976569e-09, + "loss": 0.9441, + "step": 10832 + }, + { + "epoch": 0.9769581097533481, + "grad_norm": 1.315486054320816, + "learning_rate": 5.5457267630159014e-09, + "loss": 0.9119, + "step": 10833 + }, + { + "epoch": 0.9770482932768183, + "grad_norm": 1.437376265798201, + "learning_rate": 5.5023360688910555e-09, + "loss": 0.9754, + "step": 10834 + }, + { + "epoch": 0.9771384768002885, + "grad_norm": 1.4086428734734322, + "learning_rate": 5.459115556304183e-09, + "loss": 0.9886, + "step": 10835 + }, + { + "epoch": 0.9772286603237589, + "grad_norm": 2.02046757432595, + "learning_rate": 5.416065228943889e-09, + "loss": 0.8774, + "step": 10836 + }, + { + "epoch": 0.9773188438472291, + "grad_norm": 1.636884658161581, + "learning_rate": 5.373185090482568e-09, + "loss": 0.8941, + "step": 10837 + }, + { + "epoch": 0.9774090273706993, + "grad_norm": 1.5886082452882084, + "learning_rate": 5.330475144579516e-09, + "loss": 0.9193, + "step": 10838 + }, + { + "epoch": 0.9774992108941697, + "grad_norm": 1.4001681737708147, + "learning_rate": 5.2879353948787065e-09, + "loss": 1.0227, + "step": 10839 + }, + { + "epoch": 0.9775893944176399, + "grad_norm": 1.820645805263582, + "learning_rate": 5.245565845010125e-09, + "loss": 1.0016, + "step": 10840 + }, + { + "epoch": 0.9776795779411102, + "grad_norm": 1.3614041760079794, + "learning_rate": 5.2033664985886575e-09, + "loss": 0.8756, + "step": 10841 + }, + { + "epoch": 0.9777697614645804, + "grad_norm": 1.6110082690348178, + "learning_rate": 5.161337359215201e-09, + "loss": 1.0876, + "step": 10842 + }, + { + "epoch": 0.9778599449880507, + "grad_norm": 1.5189365507866193, + "learning_rate": 5.119478430475999e-09, + "loss": 0.9517, + "step": 10843 + }, + { + "epoch": 0.977950128511521, + "grad_norm": 2.1346885618381153, + "learning_rate": 5.077789715942416e-09, + "loss": 0.8874, + "step": 10844 + }, + { + "epoch": 0.9780403120349912, + "grad_norm": 1.6276467849237986, + "learning_rate": 5.036271219171606e-09, + "loss": 0.9651, + "step": 10845 + }, + { + "epoch": 0.9781304955584614, + "grad_norm": 1.8274864972785971, + "learning_rate": 4.994922943706514e-09, + "loss": 0.9748, + "step": 10846 + }, + { + "epoch": 0.9782206790819318, + "grad_norm": 1.781232331840829, + "learning_rate": 4.953744893074763e-09, + "loss": 0.9262, + "step": 10847 + }, + { + "epoch": 0.978310862605402, + "grad_norm": 1.7809815455470681, + "learning_rate": 4.912737070789985e-09, + "loss": 0.8723, + "step": 10848 + }, + { + "epoch": 0.9784010461288722, + "grad_norm": 2.0308508847677884, + "learning_rate": 4.871899480351604e-09, + "loss": 1.0147, + "step": 10849 + }, + { + "epoch": 0.9784912296523425, + "grad_norm": 2.376131992610548, + "learning_rate": 4.831232125243501e-09, + "loss": 0.9588, + "step": 10850 + }, + { + "epoch": 0.9785814131758128, + "grad_norm": 1.4826987187714038, + "learning_rate": 4.7907350089360086e-09, + "loss": 1.0426, + "step": 10851 + }, + { + "epoch": 0.978671596699283, + "grad_norm": 1.5943112600543714, + "learning_rate": 4.750408134884365e-09, + "loss": 0.8402, + "step": 10852 + }, + { + "epoch": 0.9787617802227533, + "grad_norm": 1.7512674510230632, + "learning_rate": 4.710251506529816e-09, + "loss": 0.8511, + "step": 10853 + }, + { + "epoch": 0.9788519637462235, + "grad_norm": 1.4799275478465153, + "learning_rate": 4.6702651272982894e-09, + "loss": 0.9577, + "step": 10854 + }, + { + "epoch": 0.9789421472696939, + "grad_norm": 1.753796478438651, + "learning_rate": 4.630449000602166e-09, + "loss": 0.9157, + "step": 10855 + }, + { + "epoch": 0.9790323307931641, + "grad_norm": 1.9863922798951559, + "learning_rate": 4.590803129838283e-09, + "loss": 1.0078, + "step": 10856 + }, + { + "epoch": 0.9791225143166343, + "grad_norm": 1.6309759589457649, + "learning_rate": 4.551327518389714e-09, + "loss": 0.9451, + "step": 10857 + }, + { + "epoch": 0.9792126978401046, + "grad_norm": 1.7761973278335248, + "learning_rate": 4.512022169624652e-09, + "loss": 0.9299, + "step": 10858 + }, + { + "epoch": 0.9793028813635749, + "grad_norm": 1.4827903721789328, + "learning_rate": 4.472887086896637e-09, + "loss": 0.9162, + "step": 10859 + }, + { + "epoch": 0.9793930648870451, + "grad_norm": 2.181069103708121, + "learning_rate": 4.433922273545443e-09, + "loss": 0.8919, + "step": 10860 + }, + { + "epoch": 0.9794832484105154, + "grad_norm": 0.6370185661035622, + "learning_rate": 4.395127732895299e-09, + "loss": 0.7824, + "step": 10861 + }, + { + "epoch": 0.9795734319339856, + "grad_norm": 2.5685431701109547, + "learning_rate": 4.356503468256445e-09, + "loss": 1.0211, + "step": 10862 + }, + { + "epoch": 0.979663615457456, + "grad_norm": 1.8103069323377237, + "learning_rate": 4.318049482924913e-09, + "loss": 0.9547, + "step": 10863 + }, + { + "epoch": 0.9797537989809262, + "grad_norm": 1.6267494174879433, + "learning_rate": 4.279765780181188e-09, + "loss": 0.9082, + "step": 10864 + }, + { + "epoch": 0.9798439825043964, + "grad_norm": 1.9514219870273721, + "learning_rate": 4.241652363291992e-09, + "loss": 0.9547, + "step": 10865 + }, + { + "epoch": 0.9799341660278668, + "grad_norm": 2.020175026968555, + "learning_rate": 4.203709235509834e-09, + "loss": 0.9138, + "step": 10866 + }, + { + "epoch": 0.980024349551337, + "grad_norm": 1.5307096189548934, + "learning_rate": 4.165936400071679e-09, + "loss": 0.9606, + "step": 10867 + }, + { + "epoch": 0.9801145330748072, + "grad_norm": 1.3702753999357233, + "learning_rate": 4.12833386020095e-09, + "loss": 0.8908, + "step": 10868 + }, + { + "epoch": 0.9802047165982775, + "grad_norm": 1.7340268997165038, + "learning_rate": 4.090901619105746e-09, + "loss": 0.9827, + "step": 10869 + }, + { + "epoch": 0.9802949001217478, + "grad_norm": 1.5928503002750458, + "learning_rate": 4.053639679980181e-09, + "loss": 0.833, + "step": 10870 + }, + { + "epoch": 0.980385083645218, + "grad_norm": 1.6027236079040983, + "learning_rate": 4.01654804600371e-09, + "loss": 0.9487, + "step": 10871 + }, + { + "epoch": 0.9804752671686883, + "grad_norm": 1.5027877016772049, + "learning_rate": 3.9796267203409114e-09, + "loss": 0.953, + "step": 10872 + }, + { + "epoch": 0.9805654506921585, + "grad_norm": 1.379152529933119, + "learning_rate": 3.942875706142379e-09, + "loss": 0.922, + "step": 10873 + }, + { + "epoch": 0.9806556342156288, + "grad_norm": 1.8165131568211716, + "learning_rate": 3.906295006543825e-09, + "loss": 1.0035, + "step": 10874 + }, + { + "epoch": 0.9807458177390991, + "grad_norm": 1.5061095112857337, + "learning_rate": 3.8698846246665305e-09, + "loss": 0.9207, + "step": 10875 + }, + { + "epoch": 0.9808360012625693, + "grad_norm": 1.6425610699034683, + "learning_rate": 3.833644563617344e-09, + "loss": 0.9493, + "step": 10876 + }, + { + "epoch": 0.9809261847860395, + "grad_norm": 1.7594718052739797, + "learning_rate": 3.797574826488237e-09, + "loss": 0.9555, + "step": 10877 + }, + { + "epoch": 0.9810163683095099, + "grad_norm": 1.4821660520934872, + "learning_rate": 3.761675416356969e-09, + "loss": 0.8711, + "step": 10878 + }, + { + "epoch": 0.9811065518329801, + "grad_norm": 1.362607729441306, + "learning_rate": 3.725946336286867e-09, + "loss": 0.85, + "step": 10879 + }, + { + "epoch": 0.9811967353564504, + "grad_norm": 1.3518676854260667, + "learning_rate": 3.6903875893261604e-09, + "loss": 0.918, + "step": 10880 + }, + { + "epoch": 0.9812869188799206, + "grad_norm": 1.6421852322316868, + "learning_rate": 3.6549991785093105e-09, + "loss": 0.8777, + "step": 10881 + }, + { + "epoch": 0.9813771024033909, + "grad_norm": 1.5223643900542823, + "learning_rate": 3.6197811068554575e-09, + "loss": 0.9308, + "step": 10882 + }, + { + "epoch": 0.9814672859268612, + "grad_norm": 1.3378449592314878, + "learning_rate": 3.584733377369975e-09, + "loss": 1.0267, + "step": 10883 + }, + { + "epoch": 0.9815574694503314, + "grad_norm": 1.8181864229057263, + "learning_rate": 3.549855993043138e-09, + "loss": 0.9173, + "step": 10884 + }, + { + "epoch": 0.9816476529738016, + "grad_norm": 1.4349553349153081, + "learning_rate": 3.5151489568507887e-09, + "loss": 0.9274, + "step": 10885 + }, + { + "epoch": 0.981737836497272, + "grad_norm": 1.7948739229091843, + "learning_rate": 3.4806122717545572e-09, + "loss": 0.9583, + "step": 10886 + }, + { + "epoch": 0.9818280200207422, + "grad_norm": 1.6462405783369896, + "learning_rate": 3.446245940701198e-09, + "loss": 0.9046, + "step": 10887 + }, + { + "epoch": 0.9819182035442124, + "grad_norm": 1.4347884270081388, + "learning_rate": 3.41204996662281e-09, + "loss": 0.8805, + "step": 10888 + }, + { + "epoch": 0.9820083870676828, + "grad_norm": 1.2921666080211276, + "learning_rate": 3.3780243524375028e-09, + "loss": 0.9376, + "step": 10889 + }, + { + "epoch": 0.982098570591153, + "grad_norm": 1.5510279845929533, + "learning_rate": 3.3441691010485107e-09, + "loss": 0.9663, + "step": 10890 + }, + { + "epoch": 0.9821887541146233, + "grad_norm": 1.5931654213304092, + "learning_rate": 3.3104842153444113e-09, + "loss": 1.0238, + "step": 10891 + }, + { + "epoch": 0.9822789376380935, + "grad_norm": 1.6079401823750894, + "learning_rate": 3.27696969819935e-09, + "loss": 0.9337, + "step": 10892 + }, + { + "epoch": 0.9823691211615638, + "grad_norm": 1.3935489560860153, + "learning_rate": 3.2436255524732615e-09, + "loss": 0.8959, + "step": 10893 + }, + { + "epoch": 0.9824593046850341, + "grad_norm": 1.3185586962316165, + "learning_rate": 3.210451781010759e-09, + "loss": 0.9203, + "step": 10894 + }, + { + "epoch": 0.9825494882085043, + "grad_norm": 1.4773998005773012, + "learning_rate": 3.1774483866426895e-09, + "loss": 0.8012, + "step": 10895 + }, + { + "epoch": 0.9826396717319745, + "grad_norm": 1.395040851733187, + "learning_rate": 3.144615372185244e-09, + "loss": 0.9077, + "step": 10896 + }, + { + "epoch": 0.9827298552554449, + "grad_norm": 1.2562401462756816, + "learning_rate": 3.1119527404399604e-09, + "loss": 1.0315, + "step": 10897 + }, + { + "epoch": 0.9828200387789151, + "grad_norm": 1.8326416950686986, + "learning_rate": 3.0794604941932754e-09, + "loss": 0.9391, + "step": 10898 + }, + { + "epoch": 0.9829102223023853, + "grad_norm": 1.4569378124886554, + "learning_rate": 3.0471386362180827e-09, + "loss": 0.8399, + "step": 10899 + }, + { + "epoch": 0.9830004058258556, + "grad_norm": 1.4794288546159837, + "learning_rate": 3.0149871692719542e-09, + "loss": 0.9905, + "step": 10900 + }, + { + "epoch": 0.9830905893493259, + "grad_norm": 1.4207426320022039, + "learning_rate": 2.9830060960984728e-09, + "loss": 0.9298, + "step": 10901 + }, + { + "epoch": 0.9831807728727961, + "grad_norm": 0.6685063836414203, + "learning_rate": 2.9511954194263442e-09, + "loss": 0.7852, + "step": 10902 + }, + { + "epoch": 0.9832709563962664, + "grad_norm": 1.5654130864586842, + "learning_rate": 2.9195551419698426e-09, + "loss": 0.9409, + "step": 10903 + }, + { + "epoch": 0.9833611399197366, + "grad_norm": 1.8087833049794988, + "learning_rate": 2.888085266428808e-09, + "loss": 0.8994, + "step": 10904 + }, + { + "epoch": 0.983451323443207, + "grad_norm": 1.6355816298300505, + "learning_rate": 2.8567857954882037e-09, + "loss": 0.8973, + "step": 10905 + }, + { + "epoch": 0.9835415069666772, + "grad_norm": 1.7083891960252573, + "learning_rate": 2.82565673181856e-09, + "loss": 0.997, + "step": 10906 + }, + { + "epoch": 0.9836316904901474, + "grad_norm": 1.4498161462441217, + "learning_rate": 2.7946980780764184e-09, + "loss": 0.9606, + "step": 10907 + }, + { + "epoch": 0.9837218740136177, + "grad_norm": 1.8073220447402092, + "learning_rate": 2.763909836903e-09, + "loss": 0.9981, + "step": 10908 + }, + { + "epoch": 0.983812057537088, + "grad_norm": 1.3378894221168178, + "learning_rate": 2.7332920109255364e-09, + "loss": 0.8559, + "step": 10909 + }, + { + "epoch": 0.9839022410605582, + "grad_norm": 1.4504116526342334, + "learning_rate": 2.702844602756382e-09, + "loss": 0.9203, + "step": 10910 + }, + { + "epoch": 0.9839924245840285, + "grad_norm": 1.9413737058310383, + "learning_rate": 2.6725676149936814e-09, + "loss": 0.9213, + "step": 10911 + }, + { + "epoch": 0.9840826081074988, + "grad_norm": 1.576637851513606, + "learning_rate": 2.642461050220479e-09, + "loss": 1.0072, + "step": 10912 + }, + { + "epoch": 0.984172791630969, + "grad_norm": 1.4284490004940076, + "learning_rate": 2.612524911005609e-09, + "loss": 0.9182, + "step": 10913 + }, + { + "epoch": 0.9842629751544393, + "grad_norm": 1.4284856361785914, + "learning_rate": 2.582759199903917e-09, + "loss": 0.8225, + "step": 10914 + }, + { + "epoch": 0.9843531586779095, + "grad_norm": 1.4628308109538266, + "learning_rate": 2.553163919454704e-09, + "loss": 0.8994, + "step": 10915 + }, + { + "epoch": 0.9844433422013799, + "grad_norm": 1.453162818334534, + "learning_rate": 2.523739072183506e-09, + "loss": 0.9234, + "step": 10916 + }, + { + "epoch": 0.9845335257248501, + "grad_norm": 1.374428023234482, + "learning_rate": 2.4944846606007597e-09, + "loss": 0.9007, + "step": 10917 + }, + { + "epoch": 0.9846237092483203, + "grad_norm": 0.654389150005622, + "learning_rate": 2.46540068720269e-09, + "loss": 0.8157, + "step": 10918 + }, + { + "epoch": 0.9847138927717906, + "grad_norm": 2.2895870714327766, + "learning_rate": 2.4364871544708674e-09, + "loss": 0.9771, + "step": 10919 + }, + { + "epoch": 0.9848040762952609, + "grad_norm": 1.6229635830105875, + "learning_rate": 2.4077440648726523e-09, + "loss": 0.8885, + "step": 10920 + }, + { + "epoch": 0.9848942598187311, + "grad_norm": 1.598229265865865, + "learning_rate": 2.379171420860082e-09, + "loss": 1.0059, + "step": 10921 + }, + { + "epoch": 0.9849844433422014, + "grad_norm": 1.6982486454916024, + "learning_rate": 2.3507692248714296e-09, + "loss": 0.8627, + "step": 10922 + }, + { + "epoch": 0.9850746268656716, + "grad_norm": 1.671971541139712, + "learning_rate": 2.322537479330089e-09, + "loss": 0.9564, + "step": 10923 + }, + { + "epoch": 0.9851648103891419, + "grad_norm": 1.6001571071500655, + "learning_rate": 2.2944761866450223e-09, + "loss": 1.0158, + "step": 10924 + }, + { + "epoch": 0.9852549939126122, + "grad_norm": 1.9391536423798097, + "learning_rate": 2.266585349210315e-09, + "loss": 0.9246, + "step": 10925 + }, + { + "epoch": 0.9853451774360824, + "grad_norm": 2.09050396624492, + "learning_rate": 2.2388649694060623e-09, + "loss": 0.9017, + "step": 10926 + }, + { + "epoch": 0.9854353609595526, + "grad_norm": 1.441871133905791, + "learning_rate": 2.211315049597262e-09, + "loss": 1.001, + "step": 10927 + }, + { + "epoch": 0.985525544483023, + "grad_norm": 1.604119097842447, + "learning_rate": 2.1839355921349224e-09, + "loss": 0.9749, + "step": 10928 + }, + { + "epoch": 0.9856157280064932, + "grad_norm": 2.0735573728933616, + "learning_rate": 2.156726599354952e-09, + "loss": 0.9347, + "step": 10929 + }, + { + "epoch": 0.9857059115299635, + "grad_norm": 1.5620163741285347, + "learning_rate": 2.129688073578828e-09, + "loss": 0.9651, + "step": 10930 + }, + { + "epoch": 0.9857960950534337, + "grad_norm": 1.6701475113689974, + "learning_rate": 2.1028200171142597e-09, + "loss": 1.0702, + "step": 10931 + }, + { + "epoch": 0.985886278576904, + "grad_norm": 1.492357493998197, + "learning_rate": 2.076122432253191e-09, + "loss": 0.9222, + "step": 10932 + }, + { + "epoch": 0.9859764621003743, + "grad_norm": 1.5580476267761703, + "learning_rate": 2.0495953212738005e-09, + "loss": 0.9802, + "step": 10933 + }, + { + "epoch": 0.9860666456238445, + "grad_norm": 1.3422574914306482, + "learning_rate": 2.0232386864396102e-09, + "loss": 1.0247, + "step": 10934 + }, + { + "epoch": 0.9861568291473148, + "grad_norm": 3.933646481112239, + "learning_rate": 1.9970525299992656e-09, + "loss": 0.9625, + "step": 10935 + }, + { + "epoch": 0.9862470126707851, + "grad_norm": 2.2403173855315766, + "learning_rate": 1.9710368541874245e-09, + "loss": 0.9031, + "step": 10936 + }, + { + "epoch": 0.9863371961942553, + "grad_norm": 0.5929836547747477, + "learning_rate": 1.945191661223644e-09, + "loss": 0.7872, + "step": 10937 + }, + { + "epoch": 0.9864273797177255, + "grad_norm": 1.5681200046695405, + "learning_rate": 1.9195169533132714e-09, + "loss": 0.9855, + "step": 10938 + }, + { + "epoch": 0.9865175632411959, + "grad_norm": 0.598682519737689, + "learning_rate": 1.894012732646999e-09, + "loss": 0.7949, + "step": 10939 + }, + { + "epoch": 0.9866077467646661, + "grad_norm": 1.4504760881390577, + "learning_rate": 1.8686790014010854e-09, + "loss": 0.8917, + "step": 10940 + }, + { + "epoch": 0.9866979302881363, + "grad_norm": 0.618854589003572, + "learning_rate": 1.8435157617369134e-09, + "loss": 0.7602, + "step": 10941 + }, + { + "epoch": 0.9867881138116066, + "grad_norm": 1.722398042249652, + "learning_rate": 1.818523015801876e-09, + "loss": 1.0807, + "step": 10942 + }, + { + "epoch": 0.9868782973350769, + "grad_norm": 1.6592134558714184, + "learning_rate": 1.7937007657282677e-09, + "loss": 0.896, + "step": 10943 + }, + { + "epoch": 0.9869684808585472, + "grad_norm": 1.4381670855249926, + "learning_rate": 1.7690490136341718e-09, + "loss": 0.8853, + "step": 10944 + }, + { + "epoch": 0.9870586643820174, + "grad_norm": 1.8112682234727304, + "learning_rate": 1.744567761622795e-09, + "loss": 0.9367, + "step": 10945 + }, + { + "epoch": 0.9871488479054876, + "grad_norm": 1.3869186472202555, + "learning_rate": 1.7202570117831327e-09, + "loss": 0.9317, + "step": 10946 + }, + { + "epoch": 0.987239031428958, + "grad_norm": 2.332918391344328, + "learning_rate": 1.696116766189526e-09, + "loss": 0.9177, + "step": 10947 + }, + { + "epoch": 0.9873292149524282, + "grad_norm": 1.3972704402338314, + "learning_rate": 1.6721470269021042e-09, + "loss": 0.9796, + "step": 10948 + }, + { + "epoch": 0.9874193984758984, + "grad_norm": 1.7210234431565803, + "learning_rate": 1.6483477959654546e-09, + "loss": 0.9121, + "step": 10949 + }, + { + "epoch": 0.9875095819993687, + "grad_norm": 1.393431799107636, + "learning_rate": 1.6247190754106187e-09, + "loss": 0.9227, + "step": 10950 + }, + { + "epoch": 0.987599765522839, + "grad_norm": 1.3763901012597946, + "learning_rate": 1.6012608672537619e-09, + "loss": 0.9051, + "step": 10951 + }, + { + "epoch": 0.9876899490463092, + "grad_norm": 1.6684080166381063, + "learning_rate": 1.5779731734963942e-09, + "loss": 0.9479, + "step": 10952 + }, + { + "epoch": 0.9877801325697795, + "grad_norm": 2.767913130444445, + "learning_rate": 1.5548559961253705e-09, + "loss": 1.0031, + "step": 10953 + }, + { + "epoch": 0.9878703160932497, + "grad_norm": 1.683014736933794, + "learning_rate": 1.5319093371135573e-09, + "loss": 0.8731, + "step": 10954 + }, + { + "epoch": 0.9879604996167201, + "grad_norm": 1.671127580083584, + "learning_rate": 1.5091331984184997e-09, + "loss": 1.0321, + "step": 10955 + }, + { + "epoch": 0.9880506831401903, + "grad_norm": 1.5621902921819182, + "learning_rate": 1.486527581983754e-09, + "loss": 0.9395, + "step": 10956 + }, + { + "epoch": 0.9881408666636605, + "grad_norm": 1.5969053709721568, + "learning_rate": 1.4640924897382223e-09, + "loss": 1.0015, + "step": 10957 + }, + { + "epoch": 0.9882310501871309, + "grad_norm": 1.3615363904451503, + "learning_rate": 1.4418279235961506e-09, + "loss": 0.9826, + "step": 10958 + }, + { + "epoch": 0.9883212337106011, + "grad_norm": 1.5324933200958188, + "learning_rate": 1.4197338854573526e-09, + "loss": 0.8406, + "step": 10959 + }, + { + "epoch": 0.9884114172340713, + "grad_norm": 1.7855285656945208, + "learning_rate": 1.3978103772067651e-09, + "loss": 0.899, + "step": 10960 + }, + { + "epoch": 0.9885016007575416, + "grad_norm": 1.9562166692557281, + "learning_rate": 1.3760574007153358e-09, + "loss": 0.8604, + "step": 10961 + }, + { + "epoch": 0.9885917842810119, + "grad_norm": 1.715921189496899, + "learning_rate": 1.3544749578389137e-09, + "loss": 1.0224, + "step": 10962 + }, + { + "epoch": 0.9886819678044821, + "grad_norm": 1.5463458610800715, + "learning_rate": 1.3330630504189143e-09, + "loss": 0.9898, + "step": 10963 + }, + { + "epoch": 0.9887721513279524, + "grad_norm": 1.4794215220077283, + "learning_rate": 1.3118216802827652e-09, + "loss": 0.9096, + "step": 10964 + }, + { + "epoch": 0.9888623348514226, + "grad_norm": 1.615120419007669, + "learning_rate": 1.2907508492425722e-09, + "loss": 0.9108, + "step": 10965 + }, + { + "epoch": 0.988952518374893, + "grad_norm": 1.5735323713504417, + "learning_rate": 1.2698505590962305e-09, + "loss": 0.9043, + "step": 10966 + }, + { + "epoch": 0.9890427018983632, + "grad_norm": 1.5557731020274344, + "learning_rate": 1.2491208116272022e-09, + "loss": 0.9338, + "step": 10967 + }, + { + "epoch": 0.9891328854218334, + "grad_norm": 1.679110973687203, + "learning_rate": 1.2285616086040728e-09, + "loss": 0.9771, + "step": 10968 + }, + { + "epoch": 0.9892230689453037, + "grad_norm": 1.3573908740558458, + "learning_rate": 1.2081729517812167e-09, + "loss": 1.0052, + "step": 10969 + }, + { + "epoch": 0.989313252468774, + "grad_norm": 1.511305247291669, + "learning_rate": 1.1879548428983533e-09, + "loss": 0.9368, + "step": 10970 + }, + { + "epoch": 0.9894034359922442, + "grad_norm": 1.6901154206718756, + "learning_rate": 1.167907283680547e-09, + "loss": 0.8931, + "step": 10971 + }, + { + "epoch": 0.9894936195157145, + "grad_norm": 1.6671646645724076, + "learning_rate": 1.1480302758382077e-09, + "loss": 0.8841, + "step": 10972 + }, + { + "epoch": 0.9895838030391847, + "grad_norm": 1.4125287500341919, + "learning_rate": 1.1283238210675338e-09, + "loss": 0.9315, + "step": 10973 + }, + { + "epoch": 0.989673986562655, + "grad_norm": 1.9415661997309344, + "learning_rate": 1.1087879210498474e-09, + "loss": 0.8692, + "step": 10974 + }, + { + "epoch": 0.9897641700861253, + "grad_norm": 1.459274161079478, + "learning_rate": 1.0894225774522592e-09, + "loss": 0.9841, + "step": 10975 + }, + { + "epoch": 0.9898543536095955, + "grad_norm": 0.6623351575794001, + "learning_rate": 1.070227791927003e-09, + "loss": 0.7974, + "step": 10976 + }, + { + "epoch": 0.9899445371330657, + "grad_norm": 1.7240425465770377, + "learning_rate": 1.0512035661118802e-09, + "loss": 1.0439, + "step": 10977 + }, + { + "epoch": 0.9900347206565361, + "grad_norm": 1.5797446975389617, + "learning_rate": 1.0323499016300364e-09, + "loss": 0.9636, + "step": 10978 + }, + { + "epoch": 0.9901249041800063, + "grad_norm": 1.8273746874385164, + "learning_rate": 1.013666800090407e-09, + "loss": 0.8949, + "step": 10979 + }, + { + "epoch": 0.9902150877034765, + "grad_norm": 1.7521254710945018, + "learning_rate": 9.951542630870502e-10, + "loss": 0.9265, + "step": 10980 + }, + { + "epoch": 0.9903052712269468, + "grad_norm": 1.9020168642405537, + "learning_rate": 9.768122921995915e-10, + "loss": 0.9761, + "step": 10981 + }, + { + "epoch": 0.9903954547504171, + "grad_norm": 1.87301148827691, + "learning_rate": 9.58640888992779e-10, + "loss": 0.883, + "step": 10982 + }, + { + "epoch": 0.9904856382738874, + "grad_norm": 1.4395144901109471, + "learning_rate": 9.40640055017594e-10, + "loss": 0.8554, + "step": 10983 + }, + { + "epoch": 0.9905758217973576, + "grad_norm": 2.261343758549291, + "learning_rate": 9.228097918094757e-10, + "loss": 0.9413, + "step": 10984 + }, + { + "epoch": 0.9906660053208279, + "grad_norm": 1.7461295241546406, + "learning_rate": 9.051501008900952e-10, + "loss": 0.908, + "step": 10985 + }, + { + "epoch": 0.9907561888442982, + "grad_norm": 1.7091518472630112, + "learning_rate": 8.876609837662475e-10, + "loss": 0.9599, + "step": 10986 + }, + { + "epoch": 0.9908463723677684, + "grad_norm": 1.4630738819681288, + "learning_rate": 8.70342441930294e-10, + "loss": 1.0281, + "step": 10987 + }, + { + "epoch": 0.9909365558912386, + "grad_norm": 1.5534261829775122, + "learning_rate": 8.531944768594979e-10, + "loss": 0.9114, + "step": 10988 + }, + { + "epoch": 0.991026739414709, + "grad_norm": 1.381805915641186, + "learning_rate": 8.362170900175769e-10, + "loss": 1.0223, + "step": 10989 + }, + { + "epoch": 0.9911169229381792, + "grad_norm": 1.6477146100361886, + "learning_rate": 8.194102828527061e-10, + "loss": 0.8984, + "step": 10990 + }, + { + "epoch": 0.9912071064616494, + "grad_norm": 1.6936826826280666, + "learning_rate": 8.027740567992936e-10, + "loss": 0.9843, + "step": 10991 + }, + { + "epoch": 0.9912972899851197, + "grad_norm": 1.4130833223077477, + "learning_rate": 7.863084132766484e-10, + "loss": 0.9359, + "step": 10992 + }, + { + "epoch": 0.99138747350859, + "grad_norm": 1.7828989260400752, + "learning_rate": 7.700133536896469e-10, + "loss": 0.9537, + "step": 10993 + }, + { + "epoch": 0.9914776570320603, + "grad_norm": 1.4660789063413813, + "learning_rate": 7.538888794287324e-10, + "loss": 0.9072, + "step": 10994 + }, + { + "epoch": 0.9915678405555305, + "grad_norm": 1.5701371280045975, + "learning_rate": 7.379349918696931e-10, + "loss": 0.9628, + "step": 10995 + }, + { + "epoch": 0.9916580240790007, + "grad_norm": 1.5625731641805956, + "learning_rate": 7.221516923738846e-10, + "loss": 0.9156, + "step": 10996 + }, + { + "epoch": 0.9917482076024711, + "grad_norm": 2.071533649109594, + "learning_rate": 7.065389822880075e-10, + "loss": 1.0463, + "step": 10997 + }, + { + "epoch": 0.9918383911259413, + "grad_norm": 1.681841936810662, + "learning_rate": 6.910968629443292e-10, + "loss": 1.0201, + "step": 10998 + }, + { + "epoch": 0.9919285746494115, + "grad_norm": 1.5965713510885824, + "learning_rate": 6.758253356602406e-10, + "loss": 0.9363, + "step": 10999 + }, + { + "epoch": 0.9920187581728818, + "grad_norm": 1.6283462989054065, + "learning_rate": 6.607244017389213e-10, + "loss": 1.0352, + "step": 11000 + }, + { + "epoch": 0.9921089416963521, + "grad_norm": 1.978587986186776, + "learning_rate": 6.457940624686742e-10, + "loss": 0.9607, + "step": 11001 + }, + { + "epoch": 0.9921991252198223, + "grad_norm": 1.7579201220309426, + "learning_rate": 6.310343191238132e-10, + "loss": 1.0281, + "step": 11002 + }, + { + "epoch": 0.9922893087432926, + "grad_norm": 1.8337880639908348, + "learning_rate": 6.164451729635534e-10, + "loss": 0.8353, + "step": 11003 + }, + { + "epoch": 0.9923794922667628, + "grad_norm": 2.770236464918423, + "learning_rate": 6.020266252324546e-10, + "loss": 0.8139, + "step": 11004 + }, + { + "epoch": 0.9924696757902332, + "grad_norm": 1.5783400436280206, + "learning_rate": 5.877786771610882e-10, + "loss": 0.9154, + "step": 11005 + }, + { + "epoch": 0.9925598593137034, + "grad_norm": 0.7308072616453353, + "learning_rate": 5.737013299651483e-10, + "loss": 0.8477, + "step": 11006 + }, + { + "epoch": 0.9926500428371736, + "grad_norm": 1.643419410909217, + "learning_rate": 5.597945848458963e-10, + "loss": 0.9468, + "step": 11007 + }, + { + "epoch": 0.992740226360644, + "grad_norm": 1.595920954092472, + "learning_rate": 5.460584429894944e-10, + "loss": 0.9154, + "step": 11008 + }, + { + "epoch": 0.9928304098841142, + "grad_norm": 1.5847772572922394, + "learning_rate": 5.32492905568338e-10, + "loss": 0.9438, + "step": 11009 + }, + { + "epoch": 0.9929205934075844, + "grad_norm": 1.66603229688775, + "learning_rate": 5.190979737399459e-10, + "loss": 0.9759, + "step": 11010 + }, + { + "epoch": 0.9930107769310547, + "grad_norm": 1.6365623672540657, + "learning_rate": 5.058736486469594e-10, + "loss": 0.8319, + "step": 11011 + }, + { + "epoch": 0.993100960454525, + "grad_norm": 1.5465204622510278, + "learning_rate": 4.928199314180314e-10, + "loss": 0.9716, + "step": 11012 + }, + { + "epoch": 0.9931911439779952, + "grad_norm": 1.5841509147046857, + "learning_rate": 4.799368231669376e-10, + "loss": 0.8559, + "step": 11013 + }, + { + "epoch": 0.9932813275014655, + "grad_norm": 1.687712055587227, + "learning_rate": 4.672243249927988e-10, + "loss": 0.9307, + "step": 11014 + }, + { + "epoch": 0.9933715110249357, + "grad_norm": 1.4724109556673994, + "learning_rate": 4.546824379803027e-10, + "loss": 0.9721, + "step": 11015 + }, + { + "epoch": 0.993461694548406, + "grad_norm": 4.21580504307954, + "learning_rate": 4.4231116319970454e-10, + "loss": 0.97, + "step": 11016 + }, + { + "epoch": 0.9935518780718763, + "grad_norm": 1.575901769399046, + "learning_rate": 4.3011050170660423e-10, + "loss": 0.9447, + "step": 11017 + }, + { + "epoch": 0.9936420615953465, + "grad_norm": 1.5007927706989648, + "learning_rate": 4.18080454542169e-10, + "loss": 0.9337, + "step": 11018 + }, + { + "epoch": 0.9937322451188167, + "grad_norm": 0.6755993328106433, + "learning_rate": 4.0622102273246694e-10, + "loss": 0.7779, + "step": 11019 + }, + { + "epoch": 0.9938224286422871, + "grad_norm": 1.4915333057793734, + "learning_rate": 3.945322072897994e-10, + "loss": 1.0041, + "step": 11020 + }, + { + "epoch": 0.9939126121657573, + "grad_norm": 1.844555048645643, + "learning_rate": 3.830140092111467e-10, + "loss": 0.9993, + "step": 11021 + }, + { + "epoch": 0.9940027956892276, + "grad_norm": 2.6854667956810725, + "learning_rate": 3.7166642947972225e-10, + "loss": 0.9406, + "step": 11022 + }, + { + "epoch": 0.9940929792126978, + "grad_norm": 1.5278949962307082, + "learning_rate": 3.604894690634186e-10, + "loss": 0.918, + "step": 11023 + }, + { + "epoch": 0.9941831627361681, + "grad_norm": 3.489761636458409, + "learning_rate": 3.494831289161393e-10, + "loss": 1.0305, + "step": 11024 + }, + { + "epoch": 0.9942733462596384, + "grad_norm": 1.400646319382417, + "learning_rate": 3.3864740997668897e-10, + "loss": 0.9763, + "step": 11025 + }, + { + "epoch": 0.9943635297831086, + "grad_norm": 1.5291339500038716, + "learning_rate": 3.279823131701054e-10, + "loss": 0.9676, + "step": 11026 + }, + { + "epoch": 0.9944537133065788, + "grad_norm": 1.459040016430841, + "learning_rate": 3.1748783940610536e-10, + "loss": 0.9, + "step": 11027 + }, + { + "epoch": 0.9945438968300492, + "grad_norm": 1.5541994488542479, + "learning_rate": 3.071639895801947e-10, + "loss": 0.9014, + "step": 11028 + }, + { + "epoch": 0.9946340803535194, + "grad_norm": 1.9031578374325415, + "learning_rate": 2.9701076457322447e-10, + "loss": 1.0053, + "step": 11029 + }, + { + "epoch": 0.9947242638769896, + "grad_norm": 1.689981684387998, + "learning_rate": 2.870281652513906e-10, + "loss": 0.9853, + "step": 11030 + }, + { + "epoch": 0.99481444740046, + "grad_norm": 1.6623163509038157, + "learning_rate": 2.772161924669003e-10, + "loss": 0.926, + "step": 11031 + }, + { + "epoch": 0.9949046309239302, + "grad_norm": 1.695033529506999, + "learning_rate": 2.6757484705641765e-10, + "loss": 0.927, + "step": 11032 + }, + { + "epoch": 0.9949948144474005, + "grad_norm": 1.764924644720317, + "learning_rate": 2.58104129843062e-10, + "loss": 0.9402, + "step": 11033 + }, + { + "epoch": 0.9950849979708707, + "grad_norm": 1.8221696540116945, + "learning_rate": 2.4880404163463154e-10, + "loss": 0.8964, + "step": 11034 + }, + { + "epoch": 0.995175181494341, + "grad_norm": 1.9937462785739906, + "learning_rate": 2.3967458322471377e-10, + "loss": 0.9318, + "step": 11035 + }, + { + "epoch": 0.9952653650178113, + "grad_norm": 1.5781014317933815, + "learning_rate": 2.307157553922412e-10, + "loss": 0.9535, + "step": 11036 + }, + { + "epoch": 0.9953555485412815, + "grad_norm": 1.6232928331846588, + "learning_rate": 2.2192755890193538e-10, + "loss": 0.9848, + "step": 11037 + }, + { + "epoch": 0.9954457320647517, + "grad_norm": 0.6466263445182611, + "learning_rate": 2.133099945034189e-10, + "loss": 0.7461, + "step": 11038 + }, + { + "epoch": 0.9955359155882221, + "grad_norm": 1.870924717031315, + "learning_rate": 2.048630629318815e-10, + "loss": 0.9657, + "step": 11039 + }, + { + "epoch": 0.9956260991116923, + "grad_norm": 1.450639711869965, + "learning_rate": 1.965867649080799e-10, + "loss": 0.9402, + "step": 11040 + }, + { + "epoch": 0.9957162826351625, + "grad_norm": 0.6627457073105746, + "learning_rate": 1.8848110113856008e-10, + "loss": 0.7644, + "step": 11041 + }, + { + "epoch": 0.9958064661586328, + "grad_norm": 1.894676910778506, + "learning_rate": 1.8054607231454687e-10, + "loss": 0.9799, + "step": 11042 + }, + { + "epoch": 0.9958966496821031, + "grad_norm": 2.0927378926927775, + "learning_rate": 1.7278167911327635e-10, + "loss": 0.8988, + "step": 11043 + }, + { + "epoch": 0.9959868332055734, + "grad_norm": 1.685676719995117, + "learning_rate": 1.6518792219710753e-10, + "loss": 0.9739, + "step": 11044 + }, + { + "epoch": 0.9960770167290436, + "grad_norm": 1.4170145654784485, + "learning_rate": 1.5776480221418865e-10, + "loss": 0.9478, + "step": 11045 + }, + { + "epoch": 0.9961672002525138, + "grad_norm": 1.8957283563858627, + "learning_rate": 1.505123197977909e-10, + "loss": 0.947, + "step": 11046 + }, + { + "epoch": 0.9962573837759842, + "grad_norm": 1.5439026143239993, + "learning_rate": 1.4343047556675258e-10, + "loss": 0.8792, + "step": 11047 + }, + { + "epoch": 0.9963475672994544, + "grad_norm": 1.7320793699222699, + "learning_rate": 1.3651927012503506e-10, + "loss": 0.9469, + "step": 11048 + }, + { + "epoch": 0.9964377508229246, + "grad_norm": 1.579607125943905, + "learning_rate": 1.297787040630549e-10, + "loss": 0.9498, + "step": 11049 + }, + { + "epoch": 0.9965279343463949, + "grad_norm": 2.0158038392681474, + "learning_rate": 1.2320877795524153e-10, + "loss": 0.9659, + "step": 11050 + }, + { + "epoch": 0.9966181178698652, + "grad_norm": 1.749638928902407, + "learning_rate": 1.1680949236247962e-10, + "loss": 0.9282, + "step": 11051 + }, + { + "epoch": 0.9967083013933354, + "grad_norm": 1.4293541024452274, + "learning_rate": 1.1058084783099886e-10, + "loss": 0.9453, + "step": 11052 + }, + { + "epoch": 0.9967984849168057, + "grad_norm": 1.6339764256731986, + "learning_rate": 1.0452284489170793e-10, + "loss": 0.9527, + "step": 11053 + }, + { + "epoch": 0.996888668440276, + "grad_norm": 1.5365029366263752, + "learning_rate": 9.86354840621928e-11, + "loss": 0.9304, + "step": 11054 + }, + { + "epoch": 0.9969788519637462, + "grad_norm": 1.718336020811972, + "learning_rate": 9.291876584427427e-11, + "loss": 0.8161, + "step": 11055 + }, + { + "epoch": 0.9970690354872165, + "grad_norm": 1.6939347113646284, + "learning_rate": 8.737269072578435e-11, + "loss": 0.9141, + "step": 11056 + }, + { + "epoch": 0.9971592190106867, + "grad_norm": 1.5319410827652096, + "learning_rate": 8.199725918012212e-11, + "loss": 0.9401, + "step": 11057 + }, + { + "epoch": 0.9972494025341571, + "grad_norm": 1.517567363293366, + "learning_rate": 7.679247166603175e-11, + "loss": 0.949, + "step": 11058 + }, + { + "epoch": 0.9973395860576273, + "grad_norm": 1.476568312860582, + "learning_rate": 7.17583286273804e-11, + "loss": 0.9474, + "step": 11059 + }, + { + "epoch": 0.9974297695810975, + "grad_norm": 1.888457892681979, + "learning_rate": 6.689483049360233e-11, + "loss": 0.9673, + "step": 11060 + }, + { + "epoch": 0.9975199531045678, + "grad_norm": 1.7811239097538871, + "learning_rate": 6.220197768014302e-11, + "loss": 0.9219, + "step": 11061 + }, + { + "epoch": 0.9976101366280381, + "grad_norm": 1.5362324521682018, + "learning_rate": 5.7679770587126806e-11, + "loss": 0.964, + "step": 11062 + }, + { + "epoch": 0.9977003201515083, + "grad_norm": 0.6470388702481832, + "learning_rate": 5.33282096002452e-11, + "loss": 0.8676, + "step": 11063 + }, + { + "epoch": 0.9977905036749786, + "grad_norm": 1.6383173380575735, + "learning_rate": 4.914729509120086e-11, + "loss": 0.9088, + "step": 11064 + }, + { + "epoch": 0.9978806871984488, + "grad_norm": 1.3371779722626285, + "learning_rate": 4.513702741637537e-11, + "loss": 0.8665, + "step": 11065 + }, + { + "epoch": 0.9979708707219191, + "grad_norm": 1.521769309633143, + "learning_rate": 4.129740691816153e-11, + "loss": 1.0035, + "step": 11066 + }, + { + "epoch": 0.9980610542453894, + "grad_norm": 1.380617500930666, + "learning_rate": 3.762843392429715e-11, + "loss": 0.9622, + "step": 11067 + }, + { + "epoch": 0.9981512377688596, + "grad_norm": 1.8085518949538975, + "learning_rate": 3.413010874742106e-11, + "loss": 0.9229, + "step": 11068 + }, + { + "epoch": 0.9982414212923298, + "grad_norm": 2.3949015145967736, + "learning_rate": 3.080243168618324e-11, + "loss": 0.9415, + "step": 11069 + }, + { + "epoch": 0.9983316048158002, + "grad_norm": 1.6550946074352237, + "learning_rate": 2.7645403024800783e-11, + "loss": 0.8888, + "step": 11070 + }, + { + "epoch": 0.9984217883392704, + "grad_norm": 1.4770437596682278, + "learning_rate": 2.4659023032391756e-11, + "loss": 1.0134, + "step": 11071 + }, + { + "epoch": 0.9985119718627407, + "grad_norm": 1.6084238593496036, + "learning_rate": 2.1843291963863364e-11, + "loss": 0.9897, + "step": 11072 + }, + { + "epoch": 0.9986021553862109, + "grad_norm": 1.5149156149097023, + "learning_rate": 1.9198210059245822e-11, + "loss": 0.9599, + "step": 11073 + }, + { + "epoch": 0.9986923389096812, + "grad_norm": 1.8872324242308465, + "learning_rate": 1.672377754458054e-11, + "loss": 0.9467, + "step": 11074 + }, + { + "epoch": 0.9987825224331515, + "grad_norm": 1.4134838543142845, + "learning_rate": 1.4419994630809895e-11, + "loss": 0.9376, + "step": 11075 + }, + { + "epoch": 0.9988727059566217, + "grad_norm": 1.528457430700037, + "learning_rate": 1.2286861514443358e-11, + "loss": 0.8922, + "step": 11076 + }, + { + "epoch": 0.998962889480092, + "grad_norm": 1.336850038699352, + "learning_rate": 1.0324378377779553e-11, + "loss": 0.9818, + "step": 11077 + }, + { + "epoch": 0.9990530730035623, + "grad_norm": 1.455898140897048, + "learning_rate": 8.532545388018064e-12, + "loss": 0.9432, + "step": 11078 + }, + { + "epoch": 0.9991432565270325, + "grad_norm": 1.4273403057021337, + "learning_rate": 6.911362697925582e-12, + "loss": 0.7805, + "step": 11079 + }, + { + "epoch": 0.9992334400505027, + "grad_norm": 1.6103697730076314, + "learning_rate": 5.46083044605794e-12, + "loss": 0.9125, + "step": 11080 + }, + { + "epoch": 0.9993236235739731, + "grad_norm": 1.6136740144056563, + "learning_rate": 4.1809487563160276e-12, + "loss": 0.9953, + "step": 11081 + }, + { + "epoch": 0.9994138070974433, + "grad_norm": 1.3542097769501917, + "learning_rate": 3.0717177375017e-12, + "loss": 0.9449, + "step": 11082 + }, + { + "epoch": 0.9995039906209136, + "grad_norm": 1.9317351685501754, + "learning_rate": 2.1331374846500495e-12, + "loss": 0.9402, + "step": 11083 + }, + { + "epoch": 0.9995941741443838, + "grad_norm": 1.6507864031731565, + "learning_rate": 1.3652080774750885e-12, + "loss": 1.0426, + "step": 11084 + }, + { + "epoch": 0.9996843576678541, + "grad_norm": 1.2510942438000052, + "learning_rate": 7.679295817020204e-13, + "loss": 0.8579, + "step": 11085 + }, + { + "epoch": 0.9997745411913244, + "grad_norm": 1.8797173603124577, + "learning_rate": 3.413020484011042e-13, + "loss": 0.9235, + "step": 11086 + }, + { + "epoch": 0.9998647247147946, + "grad_norm": 1.6165984919353995, + "learning_rate": 8.53255139876552e-14, + "loss": 0.9832, + "step": 11087 + }, + { + "epoch": 0.9999549082382648, + "grad_norm": 1.6171132683417988, + "learning_rate": 0.0, + "loss": 0.8854, + "step": 11088 + }, + { + "epoch": 0.9999549082382648, + "step": 11088, + "total_flos": 1.3538718201382175e+18, + "train_loss": 0.9588839820764884, + "train_runtime": 155250.6091, + "train_samples_per_second": 4.285, + "train_steps_per_second": 0.071 + } + ], + "logging_steps": 1.0, + "max_steps": 11088, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.3538718201382175e+18, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +} diff --git a/sft_full/smoe/training_args.bin b/sft_full/smoe/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..ceb6de0a0a358e9ad3be6b0573f4b1895f599713 --- /dev/null +++ b/sft_full/smoe/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:894656220e9537a9337b5f57a61a0ede5e77ae8c5e2dd56e2d353113ca978d56 +size 8120 diff --git a/sft_full/smoe_cosinegating/added_tokens.json b/sft_full/smoe_cosinegating/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d3d3a1b74d87e381e471f7b33784015d2dc0ea --- /dev/null +++ b/sft_full/smoe_cosinegating/added_tokens.json @@ -0,0 +1,13 @@ +{ + "<|assistant|>": 32001, + "<|endoftext|>": 32000, + "<|end|>": 32007, + "<|placeholder1|>": 32002, + "<|placeholder2|>": 32003, + "<|placeholder3|>": 32004, + "<|placeholder4|>": 32005, + "<|placeholder5|>": 32008, + "<|placeholder6|>": 32009, + "<|system|>": 32006, + "<|user|>": 32010 +} diff --git a/sft_full/smoe_cosinegating/config.json b/sft_full/smoe_cosinegating/config.json new file mode 100644 index 0000000000000000000000000000000000000000..376002536c85d287bd80db8ed52e35be6445c60d --- /dev/null +++ b/sft_full/smoe_cosinegating/config.json @@ -0,0 +1,168 @@ +{ + "_name_or_path": "/cm/archive/namnv78/checkpoints/phi35-siglip224/pft", + "architectures": [ + "LlavaPhiForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "auto_map": { + "AutoConfig": "configuration_phi3.Phi3Config", + "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM" + }, + "balance_loss_coef": 0.1, + "bos_token_id": 1, + "clip_smoe": true, + "dropout": false, + "embd_pdrop": 0.0, + "eos_token_id": 32000, + "freeze_mm_mlp_adapter": false, + "hidden_act": "silu", + "hidden_size": 3072, + "image_aspect_ratio": "pad", + "initializer_range": 0.02, + "intermediate_size": 8192, + "local_rank": 0, + "max_position_embeddings": 131072, + "mlp_smoe": true, + "mm_hidden_size": 1152, + "mm_patch_merge_type": "flat", + "mm_projector_lr": null, + "mm_projector_type": "moe", + "mm_use_im_patch_token": false, + "mm_use_im_start_end": false, + "mm_vision_select_feature": "patch", + "mm_vision_select_layer": -2, + "mm_vision_tower": "google/siglip-so400m-patch14-224", + "model_type": "llava_phi", + "moe_name": "smoe_cosinegating", + "num_attention_heads": 32, + "num_experts": 4, + "num_hidden_layers": 32, + "num_key_value_heads": 32, + "num_layers": 3, + "num_selected": 2, + "original_max_position_embeddings": 4096, + "pad_token_id": 32000, + "resid_pdrop": 0.0, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "long_factor": [ + 1.0800000429153442, + 1.1100000143051147, + 1.1399999856948853, + 1.340000033378601, + 1.5899999141693115, + 1.600000023841858, + 1.6200000047683716, + 2.620000123977661, + 3.2300000190734863, + 3.2300000190734863, + 4.789999961853027, + 7.400000095367432, + 7.700000286102295, + 9.09000015258789, + 12.199999809265137, + 17.670000076293945, + 24.46000099182129, + 28.57000160217285, + 30.420001983642578, + 30.840002059936523, + 32.590003967285156, + 32.93000411987305, + 42.320003509521484, + 44.96000289916992, + 50.340003967285156, + 50.45000457763672, + 57.55000305175781, + 57.93000411987305, + 58.21000289916992, + 60.1400032043457, + 62.61000442504883, + 62.62000274658203, + 62.71000289916992, + 63.1400032043457, + 63.1400032043457, + 63.77000427246094, + 63.93000411987305, + 63.96000289916992, + 63.970001220703125, + 64.02999877929688, + 64.06999969482422, + 64.08000183105469, + 64.12000274658203, + 64.41000366210938, + 64.4800033569336, + 64.51000213623047, + 64.52999877929688, + 64.83999633789062 + ], + "short_factor": [ + 1.0, + 1.0199999809265137, + 1.0299999713897705, + 1.0299999713897705, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0699999332427979, + 1.0999999046325684, + 1.1099998950958252, + 1.1599998474121094, + 1.1599998474121094, + 1.1699998378753662, + 1.2899998426437378, + 1.339999794960022, + 1.679999828338623, + 1.7899998426437378, + 1.8199998140335083, + 1.8499997854232788, + 1.8799997568130493, + 1.9099997282028198, + 1.9399996995925903, + 1.9899996519088745, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0799996852874756, + 2.0899996757507324, + 2.189999580383301, + 2.2199995517730713, + 2.5899994373321533, + 2.729999542236328, + 2.749999523162842, + 2.8399994373321533 + ], + "type": "longrope" + }, + "rope_theta": 10000.0, + "router_z_loss_coef": 0.01, + "scales": [ + 1, + 3 + ], + "sliding_window": 262144, + "tie_word_embeddings": false, + "tokenizer_model_max_length": 2048, + "tokenizer_padding_side": "right", + "torch_dtype": "bfloat16", + "training": true, + "transformers_version": "4.43.0", + "tune_mm_mlp_adapter": false, + "use_cache": true, + "use_mm_proj": true, + "vocab_size": 32064 +} diff --git a/sft_full/smoe_cosinegating/generation_config.json b/sft_full/smoe_cosinegating/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..dad5c4578f0dc5969b38755d095fc30c368bb54a --- /dev/null +++ b/sft_full/smoe_cosinegating/generation_config.json @@ -0,0 +1,12 @@ +{ + "_from_model_config": true, + "bos_token_id": 1, + "do_sample": true, + "eos_token_id": [ + 32007, + 32001, + 32000 + ], + "pad_token_id": 32000, + "transformers_version": "4.43.0" +} diff --git a/sft_full/smoe_cosinegating/model-00001-of-00003.safetensors b/sft_full/smoe_cosinegating/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..decd07a52b3ef6767d5fee92e44db96bd2b7d118 --- /dev/null +++ b/sft_full/smoe_cosinegating/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e70af001e069480c661b44bda40ec265f8151b204f2ce7e1bd4a18adc30f49d +size 4972489328 diff --git a/sft_full/smoe_cosinegating/model-00002-of-00003.safetensors b/sft_full/smoe_cosinegating/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7059f71aee059d84aac0992f484c7b1ffd647e31 --- /dev/null +++ b/sft_full/smoe_cosinegating/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d9697408bce9e61d9db9e6a0303f7884a18627150cd410e22f1db4e78fb7d1c +size 4985533608 diff --git a/sft_full/smoe_cosinegating/model-00003-of-00003.safetensors b/sft_full/smoe_cosinegating/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a5440d6a7c5adcc6fbdfcbcc8700822879e3c7cc --- /dev/null +++ b/sft_full/smoe_cosinegating/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d8b17f0f381e5754ad51c721f0eddfd4aae81785deb6a72a998a93423e0f769f +size 248943664 diff --git a/sft_full/smoe_cosinegating/model.safetensors.index.json b/sft_full/smoe_cosinegating/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..f5e0d563e520320e7e1cb47747945b2591e60790 --- /dev/null +++ b/sft_full/smoe_cosinegating/model.safetensors.index.json @@ -0,0 +1,1033 @@ +{ + "metadata": { + "total_size": 10206819680 + }, + "weight_map": { + "lm_head.weight": "model-00003-of-00003.safetensors", + "model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.gate.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.gate.weight": "model-00003-of-00003.safetensors", + "model.norm.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.position_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors" + } +} diff --git a/sft_full/smoe_cosinegating/special_tokens_map.json b/sft_full/smoe_cosinegating/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..3e4d5a5bc1cb51753cc9ae0305ece0da60052b10 --- /dev/null +++ b/sft_full/smoe_cosinegating/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/sft_full/smoe_cosinegating/tokenizer.model b/sft_full/smoe_cosinegating/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/sft_full/smoe_cosinegating/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/sft_full/smoe_cosinegating/tokenizer_config.json b/sft_full/smoe_cosinegating/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d579bb0b91b24b214ea3c2e487e27a65017cdc4a --- /dev/null +++ b/sft_full/smoe_cosinegating/tokenizer_config.json @@ -0,0 +1,132 @@ +{ + "add_bos_token": false, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": false + }, + "32000": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "32001": { + "content": "<|assistant|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32002": { + "content": "<|placeholder1|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32003": { + "content": "<|placeholder2|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32004": { + "content": "<|placeholder3|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32005": { + "content": "<|placeholder4|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32006": { + "content": "<|system|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32007": { + "content": "<|end|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32008": { + "content": "<|placeholder5|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32009": { + "content": "<|placeholder6|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32010": { + "content": "<|user|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "chat_template": "{% for message in messages %}{% if message['role'] == 'system' and message['content'] %}{{'<|system|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'user' %}{{'<|user|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>\n' + message['content'] + '<|end|>\n'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}", + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "legacy": false, + "model_max_length": 2048, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +} diff --git a/sft_full/smoe_cosinegating/trainer_state.json b/sft_full/smoe_cosinegating/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..6bd76f57db4a90cf4f8288f71cdda88133f48f1a --- /dev/null +++ b/sft_full/smoe_cosinegating/trainer_state.json @@ -0,0 +1,77658 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999549082382648, + "eval_steps": 500, + "global_step": 11088, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 9.018352347026198e-05, + "grad_norm": 56.357384477601954, + "learning_rate": 0.0, + "loss": 2.2381, + "step": 1 + }, + { + "epoch": 0.00018036704694052397, + "grad_norm": 54.789146607129226, + "learning_rate": 4.773623799730706e-07, + "loss": 2.1713, + "step": 2 + }, + { + "epoch": 0.0002705505704107859, + "grad_norm": 34.718605844660225, + "learning_rate": 7.566014715123208e-07, + "loss": 1.8971, + "step": 3 + }, + { + "epoch": 0.00036073409388104793, + "grad_norm": 60.26651015359816, + "learning_rate": 9.547247599461412e-07, + "loss": 2.2097, + "step": 4 + }, + { + "epoch": 0.0004509176173513099, + "grad_norm": 40.082340177946506, + "learning_rate": 1.108401121501769e-06, + "loss": 1.8093, + "step": 5 + }, + { + "epoch": 0.0005411011408215718, + "grad_norm": 43.285880844320936, + "learning_rate": 1.2339638514853914e-06, + "loss": 1.9879, + "step": 6 + }, + { + "epoch": 0.0006312846642918339, + "grad_norm": 35.06857992350934, + "learning_rate": 1.3401256270225321e-06, + "loss": 1.7182, + "step": 7 + }, + { + "epoch": 0.0007214681877620959, + "grad_norm": 26.399799276802632, + "learning_rate": 1.4320871399192119e-06, + "loss": 1.652, + "step": 8 + }, + { + "epoch": 0.0008116517112323579, + "grad_norm": 16.053184806499942, + "learning_rate": 1.5132029430246416e-06, + "loss": 1.5273, + "step": 9 + }, + { + "epoch": 0.0009018352347026198, + "grad_norm": 16.29918281987484, + "learning_rate": 1.5857635014748399e-06, + "loss": 1.634, + "step": 10 + }, + { + "epoch": 0.0009920187581728818, + "grad_norm": 18.04247866499925, + "learning_rate": 1.6514025108267924e-06, + "loss": 1.711, + "step": 11 + }, + { + "epoch": 0.0010822022816431437, + "grad_norm": 11.256024252382137, + "learning_rate": 1.711326231458462e-06, + "loss": 1.5917, + "step": 12 + }, + { + "epoch": 0.0011723858051134058, + "grad_norm": 7.23744166322695, + "learning_rate": 1.7664507107987104e-06, + "loss": 1.4299, + "step": 13 + }, + { + "epoch": 0.0012625693285836677, + "grad_norm": 5.31765694240007, + "learning_rate": 1.8174880069956024e-06, + "loss": 1.3981, + "step": 14 + }, + { + "epoch": 0.0013527528520539298, + "grad_norm": 4.547303694084053, + "learning_rate": 1.8650025930140899e-06, + "loss": 1.3521, + "step": 15 + }, + { + "epoch": 0.0014429363755241917, + "grad_norm": 5.421374383070774, + "learning_rate": 1.9094495198922823e-06, + "loss": 1.4675, + "step": 16 + }, + { + "epoch": 0.0015331198989944536, + "grad_norm": 7.069416001951121, + "learning_rate": 1.9512009899507514e-06, + "loss": 1.3118, + "step": 17 + }, + { + "epoch": 0.0016233034224647158, + "grad_norm": 49.67418994291074, + "learning_rate": 1.990565322997712e-06, + "loss": 1.3672, + "step": 18 + }, + { + "epoch": 0.0017134869459349777, + "grad_norm": 3.404798565926644, + "learning_rate": 2.027800787770518e-06, + "loss": 1.2785, + "step": 19 + }, + { + "epoch": 0.0018036704694052396, + "grad_norm": 4.070504621485542, + "learning_rate": 2.06312588144791e-06, + "loss": 1.3743, + "step": 20 + }, + { + "epoch": 0.0018938539928755017, + "grad_norm": 3.2305984932285665, + "learning_rate": 2.0967270985348526e-06, + "loss": 1.2844, + "step": 21 + }, + { + "epoch": 0.0019840375163457636, + "grad_norm": 5.013963275520822, + "learning_rate": 2.128764890799863e-06, + "loss": 1.1969, + "step": 22 + }, + { + "epoch": 0.0020742210398160257, + "grad_norm": 3.182191654679494, + "learning_rate": 2.1593783012990145e-06, + "loss": 1.1988, + "step": 23 + }, + { + "epoch": 0.0021644045632862874, + "grad_norm": 2.647736042944366, + "learning_rate": 2.188688611431533e-06, + "loss": 1.3476, + "step": 24 + }, + { + "epoch": 0.0022545880867565495, + "grad_norm": 3.036547878159905, + "learning_rate": 2.216802243003538e-06, + "loss": 1.3253, + "step": 25 + }, + { + "epoch": 0.0023447716102268116, + "grad_norm": 2.799884562156334, + "learning_rate": 2.243813090771781e-06, + "loss": 1.1673, + "step": 26 + }, + { + "epoch": 0.0024349551336970737, + "grad_norm": 3.938269888073151, + "learning_rate": 2.269804414536962e-06, + "loss": 1.3327, + "step": 27 + }, + { + "epoch": 0.0025251386571673354, + "grad_norm": 3.312104255500451, + "learning_rate": 2.2948503869686733e-06, + "loss": 1.2438, + "step": 28 + }, + { + "epoch": 0.0026153221806375975, + "grad_norm": 2.5449633766941715, + "learning_rate": 2.3190173696980436e-06, + "loss": 0.9367, + "step": 29 + }, + { + "epoch": 0.0027055057041078597, + "grad_norm": 2.4929376029397283, + "learning_rate": 2.3423649729871604e-06, + "loss": 1.2456, + "step": 30 + }, + { + "epoch": 0.0027956892275781214, + "grad_norm": 2.2587511691927986, + "learning_rate": 2.364946941580084e-06, + "loss": 1.2432, + "step": 31 + }, + { + "epoch": 0.0028858727510483835, + "grad_norm": 2.3540358605517615, + "learning_rate": 2.3868118998653532e-06, + "loss": 1.3041, + "step": 32 + }, + { + "epoch": 0.0029760562745186456, + "grad_norm": 2.328779704569445, + "learning_rate": 2.408003982339113e-06, + "loss": 0.883, + "step": 33 + }, + { + "epoch": 0.0030662397979889073, + "grad_norm": 2.393391124626197, + "learning_rate": 2.4285633699238223e-06, + "loss": 1.1527, + "step": 34 + }, + { + "epoch": 0.0031564233214591694, + "grad_norm": 2.945408989011344, + "learning_rate": 2.4485267485243007e-06, + "loss": 1.1855, + "step": 35 + }, + { + "epoch": 0.0032466068449294315, + "grad_norm": 2.074095296688089, + "learning_rate": 2.467927702970783e-06, + "loss": 1.2228, + "step": 36 + }, + { + "epoch": 0.003336790368399693, + "grad_norm": 1.9202041806492065, + "learning_rate": 2.4867970569753584e-06, + "loss": 1.2412, + "step": 37 + }, + { + "epoch": 0.0034269738918699553, + "grad_norm": 2.2474987114042295, + "learning_rate": 2.5051631677435883e-06, + "loss": 1.0818, + "step": 38 + }, + { + "epoch": 0.0035171574153402174, + "grad_norm": 5.266609620537828, + "learning_rate": 2.523052182311031e-06, + "loss": 1.0497, + "step": 39 + }, + { + "epoch": 0.003607340938810479, + "grad_norm": 2.1925727880872716, + "learning_rate": 2.540488261420981e-06, + "loss": 1.2538, + "step": 40 + }, + { + "epoch": 0.0036975244622807412, + "grad_norm": 2.297630341561149, + "learning_rate": 2.557493775753984e-06, + "loss": 1.0535, + "step": 41 + }, + { + "epoch": 0.0037877079857510034, + "grad_norm": 2.0949647710200727, + "learning_rate": 2.5740894785079235e-06, + "loss": 1.168, + "step": 42 + }, + { + "epoch": 0.0038778915092212655, + "grad_norm": 3.186272927250764, + "learning_rate": 2.5902946576685834e-06, + "loss": 1.1711, + "step": 43 + }, + { + "epoch": 0.003968075032691527, + "grad_norm": 2.4208490967579293, + "learning_rate": 2.606127270772933e-06, + "loss": 1.1013, + "step": 44 + }, + { + "epoch": 0.004058258556161789, + "grad_norm": 2.57635853678969, + "learning_rate": 2.62160406452641e-06, + "loss": 1.075, + "step": 45 + }, + { + "epoch": 0.004148442079632051, + "grad_norm": 2.011233135980114, + "learning_rate": 2.636740681272085e-06, + "loss": 1.2358, + "step": 46 + }, + { + "epoch": 0.004238625603102313, + "grad_norm": 2.3779755072908437, + "learning_rate": 2.651551754008722e-06, + "loss": 1.2708, + "step": 47 + }, + { + "epoch": 0.004328809126572575, + "grad_norm": 1.682387980094668, + "learning_rate": 2.6660509914046035e-06, + "loss": 1.2067, + "step": 48 + }, + { + "epoch": 0.004418992650042837, + "grad_norm": 2.0954062899564403, + "learning_rate": 2.6802512540450642e-06, + "loss": 1.1665, + "step": 49 + }, + { + "epoch": 0.004509176173513099, + "grad_norm": 2.540772692527128, + "learning_rate": 2.694164622976609e-06, + "loss": 1.1536, + "step": 50 + }, + { + "epoch": 0.0045993596969833616, + "grad_norm": 3.830294869926907, + "learning_rate": 2.707802461463072e-06, + "loss": 1.0903, + "step": 51 + }, + { + "epoch": 0.004689543220453623, + "grad_norm": 1.7898138875480702, + "learning_rate": 2.7211754707448516e-06, + "loss": 1.1317, + "step": 52 + }, + { + "epoch": 0.004779726743923885, + "grad_norm": 5.36317438184451, + "learning_rate": 2.734293740486721e-06, + "loss": 1.1145, + "step": 53 + }, + { + "epoch": 0.0048699102673941475, + "grad_norm": 4.040206068467745, + "learning_rate": 2.747166794510033e-06, + "loss": 1.1588, + "step": 54 + }, + { + "epoch": 0.004960093790864409, + "grad_norm": 2.105835948057354, + "learning_rate": 2.759803632328562e-06, + "loss": 1.1275, + "step": 55 + }, + { + "epoch": 0.005050277314334671, + "grad_norm": 2.154524292273093, + "learning_rate": 2.772212766941744e-06, + "loss": 1.2468, + "step": 56 + }, + { + "epoch": 0.005140460837804933, + "grad_norm": 2.304880631970041, + "learning_rate": 2.7844022592828385e-06, + "loss": 1.2347, + "step": 57 + }, + { + "epoch": 0.005230644361275195, + "grad_norm": 1.8518791088994637, + "learning_rate": 2.7963797496711145e-06, + "loss": 1.0884, + "step": 58 + }, + { + "epoch": 0.005320827884745457, + "grad_norm": 1.7474469908395125, + "learning_rate": 2.80815248657541e-06, + "loss": 1.1773, + "step": 59 + }, + { + "epoch": 0.005411011408215719, + "grad_norm": 1.3743157418153948, + "learning_rate": 2.819727352960231e-06, + "loss": 0.8557, + "step": 60 + }, + { + "epoch": 0.005501194931685981, + "grad_norm": 3.9474262882475033, + "learning_rate": 2.8311108904541717e-06, + "loss": 1.1356, + "step": 61 + }, + { + "epoch": 0.005591378455156243, + "grad_norm": 4.1370553230051765, + "learning_rate": 2.842309321553155e-06, + "loss": 1.086, + "step": 62 + }, + { + "epoch": 0.005681561978626505, + "grad_norm": 1.5606927719952923, + "learning_rate": 2.8533285700471737e-06, + "loss": 1.0708, + "step": 63 + }, + { + "epoch": 0.005771745502096767, + "grad_norm": 1.9555728678888253, + "learning_rate": 2.8641742798384237e-06, + "loss": 1.2087, + "step": 64 + }, + { + "epoch": 0.005861929025567029, + "grad_norm": 1.8811451347247399, + "learning_rate": 2.874851832300479e-06, + "loss": 1.1648, + "step": 65 + }, + { + "epoch": 0.005952112549037291, + "grad_norm": 2.335828593595525, + "learning_rate": 2.8853663623121834e-06, + "loss": 1.1681, + "step": 66 + }, + { + "epoch": 0.006042296072507553, + "grad_norm": 1.9078656213411518, + "learning_rate": 2.895722773085839e-06, + "loss": 1.1952, + "step": 67 + }, + { + "epoch": 0.0061324795959778146, + "grad_norm": 1.5844716781824792, + "learning_rate": 2.905925749896893e-06, + "loss": 1.1614, + "step": 68 + }, + { + "epoch": 0.006222663119448077, + "grad_norm": 2.0228335388683023, + "learning_rate": 2.915979772811335e-06, + "loss": 1.1499, + "step": 69 + }, + { + "epoch": 0.006312846642918339, + "grad_norm": 1.7920189962018855, + "learning_rate": 2.925889128497372e-06, + "loss": 1.1114, + "step": 70 + }, + { + "epoch": 0.0064030301663886005, + "grad_norm": 2.117126914420847, + "learning_rate": 2.9356579211992906e-06, + "loss": 1.2083, + "step": 71 + }, + { + "epoch": 0.006493213689858863, + "grad_norm": 2.5581152082601144, + "learning_rate": 2.9452900829438533e-06, + "loss": 1.0819, + "step": 72 + }, + { + "epoch": 0.006583397213329125, + "grad_norm": 2.3438475524945552, + "learning_rate": 2.954789383042727e-06, + "loss": 1.1126, + "step": 73 + }, + { + "epoch": 0.006673580736799386, + "grad_norm": 1.2275778097442298, + "learning_rate": 2.9641594369484293e-06, + "loss": 0.9246, + "step": 74 + }, + { + "epoch": 0.006763764260269649, + "grad_norm": 2.3733319900094565, + "learning_rate": 2.9734037145158586e-06, + "loss": 1.1374, + "step": 75 + }, + { + "epoch": 0.006853947783739911, + "grad_norm": 2.2156520223113034, + "learning_rate": 2.982525547716659e-06, + "loss": 1.1382, + "step": 76 + }, + { + "epoch": 0.006944131307210172, + "grad_norm": 3.318976207430647, + "learning_rate": 2.9915281378493246e-06, + "loss": 1.1075, + "step": 77 + }, + { + "epoch": 0.007034314830680435, + "grad_norm": 1.9209619849999482, + "learning_rate": 3.000414562284102e-06, + "loss": 1.1315, + "step": 78 + }, + { + "epoch": 0.0071244983541506966, + "grad_norm": 2.1274823667913916, + "learning_rate": 3.009187780778246e-06, + "loss": 1.1229, + "step": 79 + }, + { + "epoch": 0.007214681877620958, + "grad_norm": 6.310643593540778, + "learning_rate": 3.017850641394051e-06, + "loss": 1.1184, + "step": 80 + }, + { + "epoch": 0.007304865401091221, + "grad_norm": 2.2742688733379004, + "learning_rate": 3.0264058860492832e-06, + "loss": 0.9864, + "step": 81 + }, + { + "epoch": 0.0073950489245614825, + "grad_norm": 1.7424557056023915, + "learning_rate": 3.0348561557270548e-06, + "loss": 1.1737, + "step": 82 + }, + { + "epoch": 0.007485232448031745, + "grad_norm": 0.8992145660859848, + "learning_rate": 3.043203995369939e-06, + "loss": 0.8846, + "step": 83 + }, + { + "epoch": 0.007575415971502007, + "grad_norm": 2.0432519661704776, + "learning_rate": 3.051451858480994e-06, + "loss": 1.2181, + "step": 84 + }, + { + "epoch": 0.007665599494972268, + "grad_norm": 2.030041496529225, + "learning_rate": 3.05960211145252e-06, + "loss": 1.0481, + "step": 85 + }, + { + "epoch": 0.007755783018442531, + "grad_norm": 1.6961345876787797, + "learning_rate": 3.0676570376416543e-06, + "loss": 1.0506, + "step": 86 + }, + { + "epoch": 0.007845966541912792, + "grad_norm": 3.0307848071986876, + "learning_rate": 3.0756188412103647e-06, + "loss": 1.148, + "step": 87 + }, + { + "epoch": 0.007936150065383054, + "grad_norm": 1.822686084514584, + "learning_rate": 3.083489650746004e-06, + "loss": 1.1248, + "step": 88 + }, + { + "epoch": 0.008026333588853317, + "grad_norm": 1.9974007405432577, + "learning_rate": 3.0912715226772975e-06, + "loss": 1.1523, + "step": 89 + }, + { + "epoch": 0.008116517112323578, + "grad_norm": 1.8335082519778307, + "learning_rate": 3.098966444499481e-06, + "loss": 1.065, + "step": 90 + }, + { + "epoch": 0.00820670063579384, + "grad_norm": 1.7282309777321978, + "learning_rate": 3.1065763378212426e-06, + "loss": 1.1334, + "step": 91 + }, + { + "epoch": 0.008296884159264103, + "grad_norm": 1.4233849022670682, + "learning_rate": 3.1141030612451554e-06, + "loss": 1.1343, + "step": 92 + }, + { + "epoch": 0.008387067682734364, + "grad_norm": 0.8814369896078021, + "learning_rate": 3.1215484130924052e-06, + "loss": 0.9152, + "step": 93 + }, + { + "epoch": 0.008477251206204626, + "grad_norm": 1.8190409267072274, + "learning_rate": 3.128914133981793e-06, + "loss": 1.139, + "step": 94 + }, + { + "epoch": 0.008567434729674889, + "grad_norm": 2.2049378856465154, + "learning_rate": 3.136201909272287e-06, + "loss": 1.1332, + "step": 95 + }, + { + "epoch": 0.00865761825314515, + "grad_norm": 1.8452190430414972, + "learning_rate": 3.1434133713776735e-06, + "loss": 1.0939, + "step": 96 + }, + { + "epoch": 0.008747801776615412, + "grad_norm": 1.7473421349117604, + "learning_rate": 3.15055010196128e-06, + "loss": 1.123, + "step": 97 + }, + { + "epoch": 0.008837985300085675, + "grad_norm": 5.997968329732382, + "learning_rate": 3.157613634018135e-06, + "loss": 1.1134, + "step": 98 + }, + { + "epoch": 0.008928168823555935, + "grad_norm": 1.5531254114039161, + "learning_rate": 3.1646054538514336e-06, + "loss": 1.141, + "step": 99 + }, + { + "epoch": 0.009018352347026198, + "grad_norm": 1.8079327225996025, + "learning_rate": 3.1715270029496797e-06, + "loss": 1.1727, + "step": 100 + }, + { + "epoch": 0.00910853587049646, + "grad_norm": 1.5010194493100453, + "learning_rate": 3.1783796797704243e-06, + "loss": 1.1517, + "step": 101 + }, + { + "epoch": 0.009198719393966723, + "grad_norm": 2.4630336977054768, + "learning_rate": 3.185164841436142e-06, + "loss": 1.0794, + "step": 102 + }, + { + "epoch": 0.009288902917436984, + "grad_norm": 1.7499072867766003, + "learning_rate": 3.1918838053473723e-06, + "loss": 1.1147, + "step": 103 + }, + { + "epoch": 0.009379086440907246, + "grad_norm": 1.5952538519679644, + "learning_rate": 3.198537850717922e-06, + "loss": 1.0238, + "step": 104 + }, + { + "epoch": 0.009469269964377509, + "grad_norm": 1.9698214567188819, + "learning_rate": 3.205128220036622e-06, + "loss": 1.0771, + "step": 105 + }, + { + "epoch": 0.00955945348784777, + "grad_norm": 1.6451033007295959, + "learning_rate": 3.2116561204597917e-06, + "loss": 1.0749, + "step": 106 + }, + { + "epoch": 0.009649637011318032, + "grad_norm": 0.8894267805819417, + "learning_rate": 3.218122725138335e-06, + "loss": 0.9067, + "step": 107 + }, + { + "epoch": 0.009739820534788295, + "grad_norm": 2.6839936026541, + "learning_rate": 3.224529174483104e-06, + "loss": 1.108, + "step": 108 + }, + { + "epoch": 0.009830004058258556, + "grad_norm": 2.8410957441567892, + "learning_rate": 3.2308765773719435e-06, + "loss": 1.0525, + "step": 109 + }, + { + "epoch": 0.009920187581728818, + "grad_norm": 1.8391951209308475, + "learning_rate": 3.2371660123016323e-06, + "loss": 1.1066, + "step": 110 + }, + { + "epoch": 0.010010371105199081, + "grad_norm": 1.582366380398802, + "learning_rate": 3.2433985284876787e-06, + "loss": 1.0317, + "step": 111 + }, + { + "epoch": 0.010100554628669342, + "grad_norm": 1.693003052705918, + "learning_rate": 3.2495751469148143e-06, + "loss": 1.1022, + "step": 112 + }, + { + "epoch": 0.010190738152139604, + "grad_norm": 3.0946567632255655, + "learning_rate": 3.2556968613407816e-06, + "loss": 1.0775, + "step": 113 + }, + { + "epoch": 0.010280921675609867, + "grad_norm": 2.0844714552767973, + "learning_rate": 3.2617646392559094e-06, + "loss": 1.144, + "step": 114 + }, + { + "epoch": 0.010371105199080128, + "grad_norm": 1.8432496895302848, + "learning_rate": 3.2677794228007836e-06, + "loss": 1.1578, + "step": 115 + }, + { + "epoch": 0.01046128872255039, + "grad_norm": 2.0419951064436934, + "learning_rate": 3.273742129644185e-06, + "loss": 1.0893, + "step": 116 + }, + { + "epoch": 0.010551472246020653, + "grad_norm": 1.729248029640733, + "learning_rate": 3.279653653823352e-06, + "loss": 1.1098, + "step": 117 + }, + { + "epoch": 0.010641655769490914, + "grad_norm": 1.8296981954547027, + "learning_rate": 3.285514866548481e-06, + "loss": 1.1054, + "step": 118 + }, + { + "epoch": 0.010731839292961176, + "grad_norm": 12.210838156385716, + "learning_rate": 3.2913266169732838e-06, + "loss": 1.1671, + "step": 119 + }, + { + "epoch": 0.010822022816431439, + "grad_norm": 2.0367577634293306, + "learning_rate": 3.2970897329333017e-06, + "loss": 1.084, + "step": 120 + }, + { + "epoch": 0.0109122063399017, + "grad_norm": 2.063750725878305, + "learning_rate": 3.302805021653585e-06, + "loss": 1.081, + "step": 121 + }, + { + "epoch": 0.011002389863371962, + "grad_norm": 1.8234815394453205, + "learning_rate": 3.3084732704272426e-06, + "loss": 1.0945, + "step": 122 + }, + { + "epoch": 0.011092573386842225, + "grad_norm": 1.9658036361570665, + "learning_rate": 3.314095247266304e-06, + "loss": 1.092, + "step": 123 + }, + { + "epoch": 0.011182756910312485, + "grad_norm": 3.4036498338500225, + "learning_rate": 3.3196717015262255e-06, + "loss": 1.0591, + "step": 124 + }, + { + "epoch": 0.011272940433782748, + "grad_norm": 1.6483227296622516, + "learning_rate": 3.325203364505307e-06, + "loss": 1.0848, + "step": 125 + }, + { + "epoch": 0.01136312395725301, + "grad_norm": 1.910328613492857, + "learning_rate": 3.3306909500202442e-06, + "loss": 1.1336, + "step": 126 + }, + { + "epoch": 0.011453307480723271, + "grad_norm": 1.8326860571235926, + "learning_rate": 3.3361351549589145e-06, + "loss": 1.1615, + "step": 127 + }, + { + "epoch": 0.011543491004193534, + "grad_norm": 2.145328783388847, + "learning_rate": 3.341536659811494e-06, + "loss": 1.036, + "step": 128 + }, + { + "epoch": 0.011633674527663796, + "grad_norm": 2.007294940665372, + "learning_rate": 3.346896129180904e-06, + "loss": 1.1229, + "step": 129 + }, + { + "epoch": 0.011723858051134057, + "grad_norm": 1.9036357651499491, + "learning_rate": 3.35221421227355e-06, + "loss": 1.104, + "step": 130 + }, + { + "epoch": 0.01181404157460432, + "grad_norm": 2.2892150828052436, + "learning_rate": 3.357491543371255e-06, + "loss": 1.123, + "step": 131 + }, + { + "epoch": 0.011904225098074582, + "grad_norm": 1.6072417259080471, + "learning_rate": 3.3627287422852543e-06, + "loss": 1.0525, + "step": 132 + }, + { + "epoch": 0.011994408621544843, + "grad_norm": 1.8998165067107224, + "learning_rate": 3.3679264147930497e-06, + "loss": 1.1272, + "step": 133 + }, + { + "epoch": 0.012084592145015106, + "grad_norm": 1.9886839333331447, + "learning_rate": 3.37308515305891e-06, + "loss": 1.0473, + "step": 134 + }, + { + "epoch": 0.012174775668485368, + "grad_norm": 1.499905980660429, + "learning_rate": 3.3782055360387313e-06, + "loss": 1.1648, + "step": 135 + }, + { + "epoch": 0.012264959191955629, + "grad_norm": 1.8976503474967064, + "learning_rate": 3.3832881298699633e-06, + "loss": 1.0587, + "step": 136 + }, + { + "epoch": 0.012355142715425892, + "grad_norm": 1.5880223743918767, + "learning_rate": 3.388333488247249e-06, + "loss": 1.0738, + "step": 137 + }, + { + "epoch": 0.012445326238896154, + "grad_norm": 1.5485452439347123, + "learning_rate": 3.393342152784406e-06, + "loss": 1.15, + "step": 138 + }, + { + "epoch": 0.012535509762366415, + "grad_norm": 1.6767239708837387, + "learning_rate": 3.3983146533633376e-06, + "loss": 1.1066, + "step": 139 + }, + { + "epoch": 0.012625693285836678, + "grad_norm": 2.022964009740309, + "learning_rate": 3.403251508470442e-06, + "loss": 1.1631, + "step": 140 + }, + { + "epoch": 0.01271587680930694, + "grad_norm": 1.968925710436164, + "learning_rate": 3.408153225521043e-06, + "loss": 1.0629, + "step": 141 + }, + { + "epoch": 0.012806060332777201, + "grad_norm": 2.0159418611090514, + "learning_rate": 3.413020301172361e-06, + "loss": 0.9891, + "step": 142 + }, + { + "epoch": 0.012896243856247463, + "grad_norm": 2.1825738207031176, + "learning_rate": 3.4178532216255024e-06, + "loss": 1.0834, + "step": 143 + }, + { + "epoch": 0.012986427379717726, + "grad_norm": 2.10770367001379, + "learning_rate": 3.422652462916924e-06, + "loss": 1.1393, + "step": 144 + }, + { + "epoch": 0.013076610903187987, + "grad_norm": 1.8159208572089827, + "learning_rate": 3.4274184911998124e-06, + "loss": 1.0021, + "step": 145 + }, + { + "epoch": 0.01316679442665825, + "grad_norm": 2.229142685045201, + "learning_rate": 3.4321517630157976e-06, + "loss": 1.1903, + "step": 146 + }, + { + "epoch": 0.013256977950128512, + "grad_norm": 1.8395859199525901, + "learning_rate": 3.4368527255573845e-06, + "loss": 1.1573, + "step": 147 + }, + { + "epoch": 0.013347161473598773, + "grad_norm": 1.9167364632986899, + "learning_rate": 3.4415218169214994e-06, + "loss": 1.0673, + "step": 148 + }, + { + "epoch": 0.013437344997069035, + "grad_norm": 1.752207997982711, + "learning_rate": 3.4461594663544882e-06, + "loss": 1.0772, + "step": 149 + }, + { + "epoch": 0.013527528520539298, + "grad_norm": 1.8022458291705983, + "learning_rate": 3.450766094488929e-06, + "loss": 1.0404, + "step": 150 + }, + { + "epoch": 0.013617712044009559, + "grad_norm": 2.0789959452022115, + "learning_rate": 3.4553421135725735e-06, + "loss": 1.1415, + "step": 151 + }, + { + "epoch": 0.013707895567479821, + "grad_norm": 1.8850449113023682, + "learning_rate": 3.45988792768973e-06, + "loss": 1.1082, + "step": 152 + }, + { + "epoch": 0.013798079090950084, + "grad_norm": 2.9264075063547645, + "learning_rate": 3.464403932975393e-06, + "loss": 1.1572, + "step": 153 + }, + { + "epoch": 0.013888262614420345, + "grad_norm": 3.940764874994167, + "learning_rate": 3.468890517822395e-06, + "loss": 1.1494, + "step": 154 + }, + { + "epoch": 0.013978446137890607, + "grad_norm": 2.754363326430334, + "learning_rate": 3.473348063081853e-06, + "loss": 1.0873, + "step": 155 + }, + { + "epoch": 0.01406862966136087, + "grad_norm": 1.6891463690304969, + "learning_rate": 3.4777769422571727e-06, + "loss": 1.047, + "step": 156 + }, + { + "epoch": 0.01415881318483113, + "grad_norm": 11.883200935671292, + "learning_rate": 3.4821775216918497e-06, + "loss": 1.09, + "step": 157 + }, + { + "epoch": 0.014248996708301393, + "grad_norm": 1.6863266433027138, + "learning_rate": 3.4865501607513164e-06, + "loss": 1.0433, + "step": 158 + }, + { + "epoch": 0.014339180231771656, + "grad_norm": 1.8484379282263215, + "learning_rate": 3.4908952119990423e-06, + "loss": 1.1626, + "step": 159 + }, + { + "epoch": 0.014429363755241916, + "grad_norm": 1.0103913661240564, + "learning_rate": 3.495213021367122e-06, + "loss": 0.8649, + "step": 160 + }, + { + "epoch": 0.014519547278712179, + "grad_norm": 2.2034295054274335, + "learning_rate": 3.4995039283215464e-06, + "loss": 1.1157, + "step": 161 + }, + { + "epoch": 0.014609730802182442, + "grad_norm": 2.317318201269532, + "learning_rate": 3.5037682660223533e-06, + "loss": 1.1067, + "step": 162 + }, + { + "epoch": 0.014699914325652702, + "grad_norm": 2.862304187723856, + "learning_rate": 3.508006361478857e-06, + "loss": 1.1222, + "step": 163 + }, + { + "epoch": 0.014790097849122965, + "grad_norm": 2.210049400295298, + "learning_rate": 3.5122185357001253e-06, + "loss": 1.0761, + "step": 164 + }, + { + "epoch": 0.014880281372593228, + "grad_norm": 2.123429784136207, + "learning_rate": 3.5164051038408817e-06, + "loss": 1.1519, + "step": 165 + }, + { + "epoch": 0.01497046489606349, + "grad_norm": 1.7904077661103808, + "learning_rate": 3.5205663753430093e-06, + "loss": 1.1356, + "step": 166 + }, + { + "epoch": 0.015060648419533751, + "grad_norm": 1.506723672933348, + "learning_rate": 3.5247026540727915e-06, + "loss": 1.1877, + "step": 167 + }, + { + "epoch": 0.015150831943004013, + "grad_norm": 1.5221643067145354, + "learning_rate": 3.5288142384540645e-06, + "loss": 1.0375, + "step": 168 + }, + { + "epoch": 0.015241015466474276, + "grad_norm": 3.4815068093709494, + "learning_rate": 3.532901421597421e-06, + "loss": 1.1239, + "step": 169 + }, + { + "epoch": 0.015331198989944537, + "grad_norm": 2.701410797431274, + "learning_rate": 3.5369644914255915e-06, + "loss": 1.0375, + "step": 170 + }, + { + "epoch": 0.0154213825134148, + "grad_norm": 2.1134491182171784, + "learning_rate": 3.5410037307951596e-06, + "loss": 1.0899, + "step": 171 + }, + { + "epoch": 0.015511566036885062, + "grad_norm": 1.631489633081241, + "learning_rate": 3.545019417614725e-06, + "loss": 1.1375, + "step": 172 + }, + { + "epoch": 0.015601749560355323, + "grad_norm": 1.7936923825251259, + "learning_rate": 3.5490118249596387e-06, + "loss": 1.1344, + "step": 173 + }, + { + "epoch": 0.015691933083825584, + "grad_norm": 2.038096699779922, + "learning_rate": 3.5529812211834352e-06, + "loss": 1.1411, + "step": 174 + }, + { + "epoch": 0.015782116607295848, + "grad_norm": 1.5405603855658196, + "learning_rate": 3.5569278700260707e-06, + "loss": 1.0941, + "step": 175 + }, + { + "epoch": 0.01587230013076611, + "grad_norm": 2.0856048661661393, + "learning_rate": 3.5608520307190746e-06, + "loss": 1.0776, + "step": 176 + }, + { + "epoch": 0.01596248365423637, + "grad_norm": 1.5910676561328911, + "learning_rate": 3.564753958087731e-06, + "loss": 1.0314, + "step": 177 + }, + { + "epoch": 0.016052667177706634, + "grad_norm": 1.6756186239654107, + "learning_rate": 3.5686339026503684e-06, + "loss": 1.1017, + "step": 178 + }, + { + "epoch": 0.016142850701176895, + "grad_norm": 1.9738520557576749, + "learning_rate": 3.5724921107148806e-06, + "loss": 1.1773, + "step": 179 + }, + { + "epoch": 0.016233034224647155, + "grad_norm": 2.5908736860785697, + "learning_rate": 3.576328824472552e-06, + "loss": 1.1506, + "step": 180 + }, + { + "epoch": 0.01632321774811742, + "grad_norm": 1.67300156044987, + "learning_rate": 3.5801442820892838e-06, + "loss": 1.1867, + "step": 181 + }, + { + "epoch": 0.01641340127158768, + "grad_norm": 3.486838799695718, + "learning_rate": 3.583938717794313e-06, + "loss": 1.1175, + "step": 182 + }, + { + "epoch": 0.01650358479505794, + "grad_norm": 2.274692988377933, + "learning_rate": 3.5877123619664928e-06, + "loss": 1.1361, + "step": 183 + }, + { + "epoch": 0.016593768318528206, + "grad_norm": 1.6368626642800144, + "learning_rate": 3.5914654412182268e-06, + "loss": 1.1133, + "step": 184 + }, + { + "epoch": 0.016683951841998466, + "grad_norm": 3.422504663234345, + "learning_rate": 3.595198178477127e-06, + "loss": 1.1298, + "step": 185 + }, + { + "epoch": 0.016774135365468727, + "grad_norm": 1.9316766039329687, + "learning_rate": 3.5989107930654757e-06, + "loss": 1.0272, + "step": 186 + }, + { + "epoch": 0.01686431888893899, + "grad_norm": 2.061963329429411, + "learning_rate": 3.6026035007775437e-06, + "loss": 1.1593, + "step": 187 + }, + { + "epoch": 0.016954502412409252, + "grad_norm": 2.075982617197046, + "learning_rate": 3.6062765139548636e-06, + "loss": 1.066, + "step": 188 + }, + { + "epoch": 0.017044685935879513, + "grad_norm": 2.431767006159783, + "learning_rate": 3.6099300415594945e-06, + "loss": 0.9605, + "step": 189 + }, + { + "epoch": 0.017134869459349777, + "grad_norm": 1.7490180530390667, + "learning_rate": 3.6135642892453575e-06, + "loss": 0.9777, + "step": 190 + }, + { + "epoch": 0.01722505298282004, + "grad_norm": 1.5130113698780485, + "learning_rate": 3.6171794594277004e-06, + "loss": 1.1101, + "step": 191 + }, + { + "epoch": 0.0173152365062903, + "grad_norm": 1.6210435974452122, + "learning_rate": 3.620775751350745e-06, + "loss": 1.0467, + "step": 192 + }, + { + "epoch": 0.017405420029760563, + "grad_norm": 2.5538763689245934, + "learning_rate": 3.6243533611535794e-06, + "loss": 1.1367, + "step": 193 + }, + { + "epoch": 0.017495603553230824, + "grad_norm": 2.107337249951744, + "learning_rate": 3.627912481934351e-06, + "loss": 1.086, + "step": 194 + }, + { + "epoch": 0.017585787076701085, + "grad_norm": 1.9970832059485988, + "learning_rate": 3.6314533038128e-06, + "loss": 1.0699, + "step": 195 + }, + { + "epoch": 0.01767597060017135, + "grad_norm": 0.8350605509269654, + "learning_rate": 3.6349760139912048e-06, + "loss": 0.8505, + "step": 196 + }, + { + "epoch": 0.01776615412364161, + "grad_norm": 2.2289634207885465, + "learning_rate": 3.638480796813769e-06, + "loss": 1.1529, + "step": 197 + }, + { + "epoch": 0.01785633764711187, + "grad_norm": 2.510164482988901, + "learning_rate": 3.641967833824504e-06, + "loss": 1.0911, + "step": 198 + }, + { + "epoch": 0.017946521170582135, + "grad_norm": 2.2124240582035513, + "learning_rate": 3.645437303823663e-06, + "loss": 1.0866, + "step": 199 + }, + { + "epoch": 0.018036704694052396, + "grad_norm": 2.1579301342123767, + "learning_rate": 3.64888938292275e-06, + "loss": 1.0748, + "step": 200 + }, + { + "epoch": 0.01812688821752266, + "grad_norm": 2.6642055081553035, + "learning_rate": 3.6523242445981603e-06, + "loss": 1.0971, + "step": 201 + }, + { + "epoch": 0.01821707174099292, + "grad_norm": 1.653844201712915, + "learning_rate": 3.655742059743495e-06, + "loss": 1.0603, + "step": 202 + }, + { + "epoch": 0.018307255264463182, + "grad_norm": 2.0578443983845442, + "learning_rate": 3.659142996720576e-06, + "loss": 1.1482, + "step": 203 + }, + { + "epoch": 0.018397438787933446, + "grad_norm": 2.9607684004112893, + "learning_rate": 3.6625272214092135e-06, + "loss": 1.0607, + "step": 204 + }, + { + "epoch": 0.018487622311403707, + "grad_norm": 0.6910675210479686, + "learning_rate": 3.6658948972557535e-06, + "loss": 0.826, + "step": 205 + }, + { + "epoch": 0.018577805834873968, + "grad_norm": 3.122405233785283, + "learning_rate": 3.6692461853204432e-06, + "loss": 1.0143, + "step": 206 + }, + { + "epoch": 0.018667989358344232, + "grad_norm": 2.189649561186827, + "learning_rate": 3.672581244323656e-06, + "loss": 1.0325, + "step": 207 + }, + { + "epoch": 0.018758172881814493, + "grad_norm": 1.7810233791815204, + "learning_rate": 3.6759002306909926e-06, + "loss": 0.9659, + "step": 208 + }, + { + "epoch": 0.018848356405284754, + "grad_norm": 2.571511310805446, + "learning_rate": 3.67920329859731e-06, + "loss": 1.1023, + "step": 209 + }, + { + "epoch": 0.018938539928755018, + "grad_norm": 1.7933643046024466, + "learning_rate": 3.6824906000096923e-06, + "loss": 1.0862, + "step": 210 + }, + { + "epoch": 0.01902872345222528, + "grad_norm": 2.154231024640959, + "learning_rate": 3.6857622847294067e-06, + "loss": 1.0989, + "step": 211 + }, + { + "epoch": 0.01911890697569554, + "grad_norm": 2.1293597656994896, + "learning_rate": 3.6890185004328626e-06, + "loss": 1.0853, + "step": 212 + }, + { + "epoch": 0.019209090499165804, + "grad_norm": 2.07761623493501, + "learning_rate": 3.6922593927116113e-06, + "loss": 1.0269, + "step": 213 + }, + { + "epoch": 0.019299274022636065, + "grad_norm": 2.981979767032911, + "learning_rate": 3.695485105111406e-06, + "loss": 1.0773, + "step": 214 + }, + { + "epoch": 0.019389457546106326, + "grad_norm": 1.4906995454777918, + "learning_rate": 3.698695779170352e-06, + "loss": 1.0558, + "step": 215 + }, + { + "epoch": 0.01947964106957659, + "grad_norm": 2.3284744922042733, + "learning_rate": 3.7018915544561744e-06, + "loss": 1.0805, + "step": 216 + }, + { + "epoch": 0.01956982459304685, + "grad_norm": 2.0805625673772083, + "learning_rate": 3.7050725686026164e-06, + "loss": 1.0367, + "step": 217 + }, + { + "epoch": 0.01966000811651711, + "grad_norm": 2.176105468262668, + "learning_rate": 3.708238957345014e-06, + "loss": 1.1461, + "step": 218 + }, + { + "epoch": 0.019750191639987376, + "grad_norm": 1.669903028950673, + "learning_rate": 3.7113908545550482e-06, + "loss": 1.1267, + "step": 219 + }, + { + "epoch": 0.019840375163457637, + "grad_norm": 2.175538739535672, + "learning_rate": 3.7145283922747028e-06, + "loss": 1.0814, + "step": 220 + }, + { + "epoch": 0.019930558686927898, + "grad_norm": 1.8840899427411253, + "learning_rate": 3.7176517007494612e-06, + "loss": 1.1717, + "step": 221 + }, + { + "epoch": 0.020020742210398162, + "grad_norm": 2.057967668040238, + "learning_rate": 3.7207609084607496e-06, + "loss": 1.0271, + "step": 222 + }, + { + "epoch": 0.020110925733868423, + "grad_norm": 2.741818916785466, + "learning_rate": 3.723856142157645e-06, + "loss": 1.134, + "step": 223 + }, + { + "epoch": 0.020201109257338683, + "grad_norm": 1.5657239840771984, + "learning_rate": 3.726937526887885e-06, + "loss": 1.0912, + "step": 224 + }, + { + "epoch": 0.020291292780808948, + "grad_norm": 2.077280087910427, + "learning_rate": 3.7300051860281798e-06, + "loss": 1.0459, + "step": 225 + }, + { + "epoch": 0.02038147630427921, + "grad_norm": 2.7236897984749753, + "learning_rate": 3.733059241313852e-06, + "loss": 1.1592, + "step": 226 + }, + { + "epoch": 0.02047165982774947, + "grad_norm": 1.7549065199297702, + "learning_rate": 3.736099812867827e-06, + "loss": 1.0724, + "step": 227 + }, + { + "epoch": 0.020561843351219734, + "grad_norm": 2.707317827036448, + "learning_rate": 3.73912701922898e-06, + "loss": 1.1463, + "step": 228 + }, + { + "epoch": 0.020652026874689994, + "grad_norm": 0.9049208200180647, + "learning_rate": 3.742140977379868e-06, + "loss": 0.8563, + "step": 229 + }, + { + "epoch": 0.020742210398160255, + "grad_norm": 1.7531802707602082, + "learning_rate": 3.745141802773854e-06, + "loss": 1.1167, + "step": 230 + }, + { + "epoch": 0.02083239392163052, + "grad_norm": 1.978971739422147, + "learning_rate": 3.748129609361645e-06, + "loss": 1.0753, + "step": 231 + }, + { + "epoch": 0.02092257744510078, + "grad_norm": 1.8877268875883564, + "learning_rate": 3.7511045096172555e-06, + "loss": 1.058, + "step": 232 + }, + { + "epoch": 0.02101276096857104, + "grad_norm": 1.6721527724490906, + "learning_rate": 3.7540666145634137e-06, + "loss": 1.0786, + "step": 233 + }, + { + "epoch": 0.021102944492041305, + "grad_norm": 1.547614797299875, + "learning_rate": 3.7570160337964225e-06, + "loss": 1.0604, + "step": 234 + }, + { + "epoch": 0.021193128015511566, + "grad_norm": 3.303165166316513, + "learning_rate": 3.7599528755104913e-06, + "loss": 1.113, + "step": 235 + }, + { + "epoch": 0.021283311538981827, + "grad_norm": 1.5877683998354777, + "learning_rate": 3.7628772465215515e-06, + "loss": 1.0797, + "step": 236 + }, + { + "epoch": 0.02137349506245209, + "grad_norm": 2.111076629368223, + "learning_rate": 3.7657892522905666e-06, + "loss": 1.0729, + "step": 237 + }, + { + "epoch": 0.021463678585922352, + "grad_norm": 1.9837950211856397, + "learning_rate": 3.7686889969463542e-06, + "loss": 1.0002, + "step": 238 + }, + { + "epoch": 0.021553862109392613, + "grad_norm": 2.6339004034794518, + "learning_rate": 3.771576583307928e-06, + "loss": 1.0562, + "step": 239 + }, + { + "epoch": 0.021644045632862877, + "grad_norm": 1.7919367727084132, + "learning_rate": 3.7744521129063722e-06, + "loss": 1.1159, + "step": 240 + }, + { + "epoch": 0.021734229156333138, + "grad_norm": 1.8981993098027714, + "learning_rate": 3.7773156860062653e-06, + "loss": 1.1073, + "step": 241 + }, + { + "epoch": 0.0218244126798034, + "grad_norm": 2.7339038551966723, + "learning_rate": 3.7801674016266554e-06, + "loss": 1.1797, + "step": 242 + }, + { + "epoch": 0.021914596203273663, + "grad_norm": 0.6245080203617156, + "learning_rate": 3.7830073575616035e-06, + "loss": 0.8019, + "step": 243 + }, + { + "epoch": 0.022004779726743924, + "grad_norm": 1.9233497605815044, + "learning_rate": 3.785835650400313e-06, + "loss": 1.1185, + "step": 244 + }, + { + "epoch": 0.022094963250214185, + "grad_norm": 3.7356314999270297, + "learning_rate": 3.7886523755468334e-06, + "loss": 0.9905, + "step": 245 + }, + { + "epoch": 0.02218514677368445, + "grad_norm": 1.6702258809310522, + "learning_rate": 3.7914576272393746e-06, + "loss": 1.0728, + "step": 246 + }, + { + "epoch": 0.02227533029715471, + "grad_norm": 2.286294246222489, + "learning_rate": 3.7942514985692284e-06, + "loss": 1.1862, + "step": 247 + }, + { + "epoch": 0.02236551382062497, + "grad_norm": 4.068132926217025, + "learning_rate": 3.797034081499296e-06, + "loss": 1.0742, + "step": 248 + }, + { + "epoch": 0.022455697344095235, + "grad_norm": 2.7466294700764307, + "learning_rate": 3.7998054668822595e-06, + "loss": 1.0597, + "step": 249 + }, + { + "epoch": 0.022545880867565496, + "grad_norm": 1.7599203399490297, + "learning_rate": 3.8025657444783776e-06, + "loss": 1.0571, + "step": 250 + }, + { + "epoch": 0.022636064391035757, + "grad_norm": 2.254615288712834, + "learning_rate": 3.80531500297293e-06, + "loss": 0.9929, + "step": 251 + }, + { + "epoch": 0.02272624791450602, + "grad_norm": 3.1780635801548773, + "learning_rate": 3.8080533299933147e-06, + "loss": 1.0546, + "step": 252 + }, + { + "epoch": 0.022816431437976282, + "grad_norm": 1.7841546318412074, + "learning_rate": 3.8107808121258067e-06, + "loss": 1.0449, + "step": 253 + }, + { + "epoch": 0.022906614961446543, + "grad_norm": 2.5141351683399678, + "learning_rate": 3.813497534931985e-06, + "loss": 1.1458, + "step": 254 + }, + { + "epoch": 0.022996798484916807, + "grad_norm": 1.6970650188913583, + "learning_rate": 3.816203582964841e-06, + "loss": 1.1089, + "step": 255 + }, + { + "epoch": 0.023086982008387068, + "grad_norm": 1.8321688016808886, + "learning_rate": 3.818899039784565e-06, + "loss": 1.1365, + "step": 256 + }, + { + "epoch": 0.02317716553185733, + "grad_norm": 1.959939345472277, + "learning_rate": 3.821583987974031e-06, + "loss": 1.1402, + "step": 257 + }, + { + "epoch": 0.023267349055327593, + "grad_norm": 2.3911113368987142, + "learning_rate": 3.8242585091539755e-06, + "loss": 1.0217, + "step": 258 + }, + { + "epoch": 0.023357532578797854, + "grad_norm": 1.5519858805466427, + "learning_rate": 3.8269226839978895e-06, + "loss": 1.1652, + "step": 259 + }, + { + "epoch": 0.023447716102268115, + "grad_norm": 2.225366103666792, + "learning_rate": 3.82957659224662e-06, + "loss": 1.0389, + "step": 260 + }, + { + "epoch": 0.02353789962573838, + "grad_norm": 9.274147064392103, + "learning_rate": 3.8322203127226855e-06, + "loss": 0.9684, + "step": 261 + }, + { + "epoch": 0.02362808314920864, + "grad_norm": 1.8694349354197766, + "learning_rate": 3.834853923344326e-06, + "loss": 1.1923, + "step": 262 + }, + { + "epoch": 0.0237182666726789, + "grad_norm": 1.514999417849388, + "learning_rate": 3.837477501139285e-06, + "loss": 1.0834, + "step": 263 + }, + { + "epoch": 0.023808450196149165, + "grad_norm": 2.434977081629625, + "learning_rate": 3.840091122258324e-06, + "loss": 1.096, + "step": 264 + }, + { + "epoch": 0.023898633719619426, + "grad_norm": 2.0274024322778166, + "learning_rate": 3.84269486198849e-06, + "loss": 1.0067, + "step": 265 + }, + { + "epoch": 0.023988817243089686, + "grad_norm": 1.606770899281637, + "learning_rate": 3.845288794766121e-06, + "loss": 1.101, + "step": 266 + }, + { + "epoch": 0.02407900076655995, + "grad_norm": 1.8446692421985582, + "learning_rate": 3.847872994189619e-06, + "loss": 1.0811, + "step": 267 + }, + { + "epoch": 0.02416918429003021, + "grad_norm": 1.7966411272520917, + "learning_rate": 3.8504475330319805e-06, + "loss": 1.1444, + "step": 268 + }, + { + "epoch": 0.024259367813500472, + "grad_norm": 1.6324763339840693, + "learning_rate": 3.853012483253093e-06, + "loss": 1.0626, + "step": 269 + }, + { + "epoch": 0.024349551336970737, + "grad_norm": 2.5192966558965177, + "learning_rate": 3.855567916011802e-06, + "loss": 1.0945, + "step": 270 + }, + { + "epoch": 0.024439734860440997, + "grad_norm": 2.090293878582706, + "learning_rate": 3.858113901677755e-06, + "loss": 1.0871, + "step": 271 + }, + { + "epoch": 0.024529918383911258, + "grad_norm": 1.8253382121193449, + "learning_rate": 3.860650509843034e-06, + "loss": 1.0436, + "step": 272 + }, + { + "epoch": 0.024620101907381522, + "grad_norm": 1.9046993514983903, + "learning_rate": 3.863177809333563e-06, + "loss": 1.0697, + "step": 273 + }, + { + "epoch": 0.024710285430851783, + "grad_norm": 1.628363136843907, + "learning_rate": 3.86569586822032e-06, + "loss": 1.0587, + "step": 274 + }, + { + "epoch": 0.024800468954322044, + "grad_norm": 2.0898661104236007, + "learning_rate": 3.868204753830331e-06, + "loss": 1.0505, + "step": 275 + }, + { + "epoch": 0.02489065247779231, + "grad_norm": 3.542510048172941, + "learning_rate": 3.870704532757476e-06, + "loss": 1.0021, + "step": 276 + }, + { + "epoch": 0.02498083600126257, + "grad_norm": 1.8136532501769576, + "learning_rate": 3.8731952708730974e-06, + "loss": 1.1103, + "step": 277 + }, + { + "epoch": 0.02507101952473283, + "grad_norm": 1.9229976197543914, + "learning_rate": 3.8756770333364085e-06, + "loss": 1.1073, + "step": 278 + }, + { + "epoch": 0.025161203048203094, + "grad_norm": 1.8454172387304888, + "learning_rate": 3.878149884604725e-06, + "loss": 1.0023, + "step": 279 + }, + { + "epoch": 0.025251386571673355, + "grad_norm": 1.9152362572172323, + "learning_rate": 3.8806138884435125e-06, + "loss": 1.0553, + "step": 280 + }, + { + "epoch": 0.025341570095143616, + "grad_norm": 4.418041279245497, + "learning_rate": 3.883069107936248e-06, + "loss": 1.0129, + "step": 281 + }, + { + "epoch": 0.02543175361861388, + "grad_norm": 2.0976331769928094, + "learning_rate": 3.885515605494114e-06, + "loss": 1.1415, + "step": 282 + }, + { + "epoch": 0.02552193714208414, + "grad_norm": 2.278088089757735, + "learning_rate": 3.8879534428655145e-06, + "loss": 1.1378, + "step": 283 + }, + { + "epoch": 0.025612120665554402, + "grad_norm": 1.721400765746744, + "learning_rate": 3.890382681145432e-06, + "loss": 1.1215, + "step": 284 + }, + { + "epoch": 0.025702304189024666, + "grad_norm": 2.3825695476808506, + "learning_rate": 3.892803380784608e-06, + "loss": 1.1128, + "step": 285 + }, + { + "epoch": 0.025792487712494927, + "grad_norm": 0.9575911771823572, + "learning_rate": 3.8952156015985725e-06, + "loss": 0.8799, + "step": 286 + }, + { + "epoch": 0.025882671235965188, + "grad_norm": 2.7022078176356388, + "learning_rate": 3.897619402776516e-06, + "loss": 1.0856, + "step": 287 + }, + { + "epoch": 0.025972854759435452, + "grad_norm": 2.033299160758774, + "learning_rate": 3.900014842889995e-06, + "loss": 1.1602, + "step": 288 + }, + { + "epoch": 0.026063038282905713, + "grad_norm": 1.5312623782533408, + "learning_rate": 3.902401979901503e-06, + "loss": 1.0038, + "step": 289 + }, + { + "epoch": 0.026153221806375974, + "grad_norm": 1.9435722809528984, + "learning_rate": 3.904780871172884e-06, + "loss": 1.1267, + "step": 290 + }, + { + "epoch": 0.026243405329846238, + "grad_norm": 2.5402879779996272, + "learning_rate": 3.907151573473601e-06, + "loss": 0.9554, + "step": 291 + }, + { + "epoch": 0.0263335888533165, + "grad_norm": 1.9513006005577875, + "learning_rate": 3.909514142988868e-06, + "loss": 1.09, + "step": 292 + }, + { + "epoch": 0.02642377237678676, + "grad_norm": 1.8878321551174473, + "learning_rate": 3.911868635327639e-06, + "loss": 1.0456, + "step": 293 + }, + { + "epoch": 0.026513955900257024, + "grad_norm": 2.6927718862760273, + "learning_rate": 3.914215105530455e-06, + "loss": 1.0568, + "step": 294 + }, + { + "epoch": 0.026604139423727285, + "grad_norm": 1.864985360504865, + "learning_rate": 3.916553608077179e-06, + "loss": 1.0718, + "step": 295 + }, + { + "epoch": 0.026694322947197546, + "grad_norm": 1.7101222495388801, + "learning_rate": 3.91888419689457e-06, + "loss": 1.0651, + "step": 296 + }, + { + "epoch": 0.02678450647066781, + "grad_norm": 5.753755089664135, + "learning_rate": 3.921206925363754e-06, + "loss": 1.018, + "step": 297 + }, + { + "epoch": 0.02687468999413807, + "grad_norm": 2.1851932897555146, + "learning_rate": 3.923521846327559e-06, + "loss": 0.9947, + "step": 298 + }, + { + "epoch": 0.02696487351760833, + "grad_norm": 2.3183224553781545, + "learning_rate": 3.925829012097725e-06, + "loss": 1.0371, + "step": 299 + }, + { + "epoch": 0.027055057041078596, + "grad_norm": 2.2262954284042564, + "learning_rate": 3.928128474462e-06, + "loss": 0.9968, + "step": 300 + }, + { + "epoch": 0.027145240564548857, + "grad_norm": 1.9877264601568585, + "learning_rate": 3.930420284691115e-06, + "loss": 1.0862, + "step": 301 + }, + { + "epoch": 0.027235424088019117, + "grad_norm": 1.748500726696951, + "learning_rate": 3.932704493545644e-06, + "loss": 1.1092, + "step": 302 + }, + { + "epoch": 0.02732560761148938, + "grad_norm": 2.040440586981514, + "learning_rate": 3.934981151282745e-06, + "loss": 1.1617, + "step": 303 + }, + { + "epoch": 0.027415791134959643, + "grad_norm": 1.789082156410679, + "learning_rate": 3.9372503076628006e-06, + "loss": 1.0725, + "step": 304 + }, + { + "epoch": 0.027505974658429903, + "grad_norm": 1.7847466531047012, + "learning_rate": 3.939512011955941e-06, + "loss": 1.0936, + "step": 305 + }, + { + "epoch": 0.027596158181900168, + "grad_norm": 2.1955039595879033, + "learning_rate": 3.941766312948463e-06, + "loss": 1.0687, + "step": 306 + }, + { + "epoch": 0.02768634170537043, + "grad_norm": 1.690461139177008, + "learning_rate": 3.944013258949147e-06, + "loss": 1.0426, + "step": 307 + }, + { + "epoch": 0.02777652522884069, + "grad_norm": 1.9241036000027911, + "learning_rate": 3.946252897795465e-06, + "loss": 0.8981, + "step": 308 + }, + { + "epoch": 0.027866708752310954, + "grad_norm": 2.3624095475577147, + "learning_rate": 3.9484852768596935e-06, + "loss": 1.0832, + "step": 309 + }, + { + "epoch": 0.027956892275781214, + "grad_norm": 2.414874289105716, + "learning_rate": 3.950710443054923e-06, + "loss": 0.9761, + "step": 310 + }, + { + "epoch": 0.028047075799251475, + "grad_norm": 2.4258208586427643, + "learning_rate": 3.952928442840981e-06, + "loss": 0.9861, + "step": 311 + }, + { + "epoch": 0.02813725932272174, + "grad_norm": 1.802943323989027, + "learning_rate": 3.955139322230243e-06, + "loss": 1.0949, + "step": 312 + }, + { + "epoch": 0.028227442846192, + "grad_norm": 1.8131744839826764, + "learning_rate": 3.957343126793365e-06, + "loss": 1.0067, + "step": 313 + }, + { + "epoch": 0.02831762636966226, + "grad_norm": 1.7885726857888258, + "learning_rate": 3.959539901664921e-06, + "loss": 1.0532, + "step": 314 + }, + { + "epoch": 0.028407809893132525, + "grad_norm": 2.273292982614042, + "learning_rate": 3.9617296915489425e-06, + "loss": 1.1659, + "step": 315 + }, + { + "epoch": 0.028497993416602786, + "grad_norm": 2.2149507473190724, + "learning_rate": 3.963912540724387e-06, + "loss": 1.013, + "step": 316 + }, + { + "epoch": 0.028588176940073047, + "grad_norm": 2.9007404565960746, + "learning_rate": 3.966088493050501e-06, + "loss": 1.0711, + "step": 317 + }, + { + "epoch": 0.02867836046354331, + "grad_norm": 6.8024548306940495, + "learning_rate": 3.968257591972113e-06, + "loss": 1.0537, + "step": 318 + }, + { + "epoch": 0.028768543987013572, + "grad_norm": 1.5764730135092697, + "learning_rate": 3.970419880524835e-06, + "loss": 1.1476, + "step": 319 + }, + { + "epoch": 0.028858727510483833, + "grad_norm": 1.6468138449966443, + "learning_rate": 3.972575401340192e-06, + "loss": 1.0317, + "step": 320 + }, + { + "epoch": 0.028948911033954097, + "grad_norm": 1.5440649846390306, + "learning_rate": 3.974724196650656e-06, + "loss": 1.1059, + "step": 321 + }, + { + "epoch": 0.029039094557424358, + "grad_norm": 1.7921459828417767, + "learning_rate": 3.976866308294617e-06, + "loss": 1.0538, + "step": 322 + }, + { + "epoch": 0.02912927808089462, + "grad_norm": 1.6774683997962714, + "learning_rate": 3.979001777721269e-06, + "loss": 1.0258, + "step": 323 + }, + { + "epoch": 0.029219461604364883, + "grad_norm": 2.0675851959793716, + "learning_rate": 3.981130645995424e-06, + "loss": 1.0355, + "step": 324 + }, + { + "epoch": 0.029309645127835144, + "grad_norm": 1.7791870619284578, + "learning_rate": 3.983252953802248e-06, + "loss": 1.1362, + "step": 325 + }, + { + "epoch": 0.029399828651305405, + "grad_norm": 0.7190734301436381, + "learning_rate": 3.9853687414519285e-06, + "loss": 0.8728, + "step": 326 + }, + { + "epoch": 0.02949001217477567, + "grad_norm": 2.3473696414718352, + "learning_rate": 3.987478048884265e-06, + "loss": 1.1109, + "step": 327 + }, + { + "epoch": 0.02958019569824593, + "grad_norm": 1.925074325711966, + "learning_rate": 3.989580915673196e-06, + "loss": 1.0408, + "step": 328 + }, + { + "epoch": 0.02967037922171619, + "grad_norm": 1.8353543268442523, + "learning_rate": 3.991677381031255e-06, + "loss": 1.04, + "step": 329 + }, + { + "epoch": 0.029760562745186455, + "grad_norm": 2.043847560595493, + "learning_rate": 3.993767483813953e-06, + "loss": 1.1372, + "step": 330 + }, + { + "epoch": 0.029850746268656716, + "grad_norm": 6.913148774369209, + "learning_rate": 3.995851262524104e-06, + "loss": 0.9774, + "step": 331 + }, + { + "epoch": 0.02994092979212698, + "grad_norm": 1.752347257545148, + "learning_rate": 3.997928755316079e-06, + "loss": 1.0594, + "step": 332 + }, + { + "epoch": 0.03003111331559724, + "grad_norm": 0.842671800410149, + "learning_rate": 4e-06, + "loss": 0.8582, + "step": 333 + }, + { + "epoch": 0.030121296839067502, + "grad_norm": 1.6255948371831506, + "learning_rate": 3.999999914674486e-06, + "loss": 1.0272, + "step": 334 + }, + { + "epoch": 0.030211480362537766, + "grad_norm": 1.6273084529541129, + "learning_rate": 3.999999658697952e-06, + "loss": 1.0458, + "step": 335 + }, + { + "epoch": 0.030301663886008027, + "grad_norm": 0.6866167203166834, + "learning_rate": 3.9999992320704185e-06, + "loss": 0.7962, + "step": 336 + }, + { + "epoch": 0.030391847409478288, + "grad_norm": 2.1284994594151225, + "learning_rate": 3.999998634791922e-06, + "loss": 1.0734, + "step": 337 + }, + { + "epoch": 0.030482030932948552, + "grad_norm": 1.4818520235993813, + "learning_rate": 3.999997866862515e-06, + "loss": 1.0439, + "step": 338 + }, + { + "epoch": 0.030572214456418813, + "grad_norm": 1.5759212100767834, + "learning_rate": 3.999996928282262e-06, + "loss": 1.1493, + "step": 339 + }, + { + "epoch": 0.030662397979889074, + "grad_norm": 2.747206656404717, + "learning_rate": 3.999995819051244e-06, + "loss": 1.056, + "step": 340 + }, + { + "epoch": 0.030752581503359338, + "grad_norm": 1.5312634681576467, + "learning_rate": 3.9999945391695536e-06, + "loss": 1.0201, + "step": 341 + }, + { + "epoch": 0.0308427650268296, + "grad_norm": 2.192330830049103, + "learning_rate": 3.999993088637302e-06, + "loss": 1.1437, + "step": 342 + }, + { + "epoch": 0.03093294855029986, + "grad_norm": 1.637095478556842, + "learning_rate": 3.999991467454612e-06, + "loss": 1.1272, + "step": 343 + }, + { + "epoch": 0.031023132073770124, + "grad_norm": 1.7652700962099943, + "learning_rate": 3.999989675621622e-06, + "loss": 0.9759, + "step": 344 + }, + { + "epoch": 0.031113315597240385, + "grad_norm": 1.6415960435473855, + "learning_rate": 3.999987713138485e-06, + "loss": 1.0644, + "step": 345 + }, + { + "epoch": 0.031203499120710645, + "grad_norm": 0.7612702443104135, + "learning_rate": 3.999985580005369e-06, + "loss": 0.9047, + "step": 346 + }, + { + "epoch": 0.031293682644180906, + "grad_norm": 3.01461095017387, + "learning_rate": 3.999983276222455e-06, + "loss": 1.1304, + "step": 347 + }, + { + "epoch": 0.03138386616765117, + "grad_norm": 1.7739355253439717, + "learning_rate": 3.999980801789941e-06, + "loss": 1.0715, + "step": 348 + }, + { + "epoch": 0.031474049691121435, + "grad_norm": 1.6106883495515523, + "learning_rate": 3.999978156708036e-06, + "loss": 1.064, + "step": 349 + }, + { + "epoch": 0.031564233214591696, + "grad_norm": 2.3763015593785095, + "learning_rate": 3.9999753409769675e-06, + "loss": 1.1687, + "step": 350 + }, + { + "epoch": 0.031654416738061956, + "grad_norm": 0.6922629485333106, + "learning_rate": 3.999972354596975e-06, + "loss": 0.8473, + "step": 351 + }, + { + "epoch": 0.03174460026153222, + "grad_norm": 1.6297792432642226, + "learning_rate": 3.999969197568314e-06, + "loss": 1.0124, + "step": 352 + }, + { + "epoch": 0.03183478378500248, + "grad_norm": 1.6254930481822187, + "learning_rate": 3.999965869891253e-06, + "loss": 1.1099, + "step": 353 + }, + { + "epoch": 0.03192496730847274, + "grad_norm": 1.5617545829352755, + "learning_rate": 3.999962371566075e-06, + "loss": 1.0863, + "step": 354 + }, + { + "epoch": 0.03201515083194301, + "grad_norm": 4.14658971177436, + "learning_rate": 3.999958702593082e-06, + "loss": 1.1289, + "step": 355 + }, + { + "epoch": 0.03210533435541327, + "grad_norm": 1.465847637778284, + "learning_rate": 3.999954862972583e-06, + "loss": 1.0853, + "step": 356 + }, + { + "epoch": 0.03219551787888353, + "grad_norm": 1.735351699889204, + "learning_rate": 3.999950852704908e-06, + "loss": 1.0082, + "step": 357 + }, + { + "epoch": 0.03228570140235379, + "grad_norm": 1.8887913536932168, + "learning_rate": 3.9999466717903995e-06, + "loss": 1.1191, + "step": 358 + }, + { + "epoch": 0.03237588492582405, + "grad_norm": 0.7145707072782044, + "learning_rate": 3.999942320229413e-06, + "loss": 0.8325, + "step": 359 + }, + { + "epoch": 0.03246606844929431, + "grad_norm": 2.43976233716336, + "learning_rate": 3.99993779802232e-06, + "loss": 1.1922, + "step": 360 + }, + { + "epoch": 0.03255625197276458, + "grad_norm": 1.878287548577912, + "learning_rate": 3.999933105169506e-06, + "loss": 1.0487, + "step": 361 + }, + { + "epoch": 0.03264643549623484, + "grad_norm": 1.5243590868580295, + "learning_rate": 3.999928241671373e-06, + "loss": 1.0607, + "step": 362 + }, + { + "epoch": 0.0327366190197051, + "grad_norm": 1.7816325663606185, + "learning_rate": 3.999923207528334e-06, + "loss": 0.9561, + "step": 363 + }, + { + "epoch": 0.03282680254317536, + "grad_norm": 1.4115618518813502, + "learning_rate": 3.9999180027408196e-06, + "loss": 1.0573, + "step": 364 + }, + { + "epoch": 0.03291698606664562, + "grad_norm": 1.6516661641222474, + "learning_rate": 3.9999126273092735e-06, + "loss": 1.0368, + "step": 365 + }, + { + "epoch": 0.03300716959011588, + "grad_norm": 1.7953329352401894, + "learning_rate": 3.999907081234156e-06, + "loss": 1.1479, + "step": 366 + }, + { + "epoch": 0.03309735311358615, + "grad_norm": 2.039390099177579, + "learning_rate": 3.999901364515938e-06, + "loss": 1.0283, + "step": 367 + }, + { + "epoch": 0.03318753663705641, + "grad_norm": 7.119495340099962, + "learning_rate": 3.999895477155108e-06, + "loss": 1.1063, + "step": 368 + }, + { + "epoch": 0.03327772016052667, + "grad_norm": 0.7318549645187418, + "learning_rate": 3.999889419152169e-06, + "loss": 0.8142, + "step": 369 + }, + { + "epoch": 0.03336790368399693, + "grad_norm": 1.783427129340679, + "learning_rate": 3.999883190507638e-06, + "loss": 1.0572, + "step": 370 + }, + { + "epoch": 0.033458087207467194, + "grad_norm": 1.9182809859136822, + "learning_rate": 3.999876791222044e-06, + "loss": 1.0456, + "step": 371 + }, + { + "epoch": 0.033548270730937454, + "grad_norm": 2.0219034985946767, + "learning_rate": 3.999870221295936e-06, + "loss": 1.0129, + "step": 372 + }, + { + "epoch": 0.03363845425440772, + "grad_norm": 1.6156465126833663, + "learning_rate": 3.999863480729875e-06, + "loss": 1.0252, + "step": 373 + }, + { + "epoch": 0.03372863777787798, + "grad_norm": 1.8557244375637745, + "learning_rate": 3.999856569524433e-06, + "loss": 0.9984, + "step": 374 + }, + { + "epoch": 0.033818821301348244, + "grad_norm": 1.603194730852038, + "learning_rate": 3.999849487680202e-06, + "loss": 0.9974, + "step": 375 + }, + { + "epoch": 0.033909004824818505, + "grad_norm": 1.5660765341305625, + "learning_rate": 3.999842235197786e-06, + "loss": 0.9918, + "step": 376 + }, + { + "epoch": 0.033999188348288766, + "grad_norm": 1.7445479341384793, + "learning_rate": 3.999834812077803e-06, + "loss": 0.9903, + "step": 377 + }, + { + "epoch": 0.034089371871759026, + "grad_norm": 1.567328583441746, + "learning_rate": 3.999827218320886e-06, + "loss": 1.1081, + "step": 378 + }, + { + "epoch": 0.034179555395229294, + "grad_norm": 2.43030218953372, + "learning_rate": 3.999819453927685e-06, + "loss": 0.9718, + "step": 379 + }, + { + "epoch": 0.034269738918699555, + "grad_norm": 1.999395636797547, + "learning_rate": 3.999811518898861e-06, + "loss": 1.0588, + "step": 380 + }, + { + "epoch": 0.034359922442169816, + "grad_norm": 2.1415710586205385, + "learning_rate": 3.999803413235092e-06, + "loss": 1.0726, + "step": 381 + }, + { + "epoch": 0.03445010596564008, + "grad_norm": 3.255506692056013, + "learning_rate": 3.999795136937068e-06, + "loss": 1.1639, + "step": 382 + }, + { + "epoch": 0.03454028948911034, + "grad_norm": 1.7404810327133662, + "learning_rate": 3.999786690005496e-06, + "loss": 1.048, + "step": 383 + }, + { + "epoch": 0.0346304730125806, + "grad_norm": 2.219961883114593, + "learning_rate": 3.999778072441098e-06, + "loss": 1.0389, + "step": 384 + }, + { + "epoch": 0.034720656536050866, + "grad_norm": 2.142518813808456, + "learning_rate": 3.999769284244608e-06, + "loss": 1.0885, + "step": 385 + }, + { + "epoch": 0.03481084005952113, + "grad_norm": 1.344857712987083, + "learning_rate": 3.999760325416775e-06, + "loss": 1.073, + "step": 386 + }, + { + "epoch": 0.03490102358299139, + "grad_norm": 1.8412690361991566, + "learning_rate": 3.999751195958366e-06, + "loss": 1.1461, + "step": 387 + }, + { + "epoch": 0.03499120710646165, + "grad_norm": 0.7865107605156209, + "learning_rate": 3.999741895870157e-06, + "loss": 0.838, + "step": 388 + }, + { + "epoch": 0.03508139062993191, + "grad_norm": 2.5046592211063947, + "learning_rate": 3.999732425152944e-06, + "loss": 1.1034, + "step": 389 + }, + { + "epoch": 0.03517157415340217, + "grad_norm": 1.410180089014651, + "learning_rate": 3.999722783807533e-06, + "loss": 1.1155, + "step": 390 + }, + { + "epoch": 0.03526175767687244, + "grad_norm": 3.1608920566914613, + "learning_rate": 3.999712971834748e-06, + "loss": 0.9903, + "step": 391 + }, + { + "epoch": 0.0353519412003427, + "grad_norm": 5.228130384808382, + "learning_rate": 3.999702989235427e-06, + "loss": 1.0699, + "step": 392 + }, + { + "epoch": 0.03544212472381296, + "grad_norm": 7.250916521742161, + "learning_rate": 3.999692836010419e-06, + "loss": 1.1494, + "step": 393 + }, + { + "epoch": 0.03553230824728322, + "grad_norm": 2.1863205045887946, + "learning_rate": 3.999682512160593e-06, + "loss": 1.2004, + "step": 394 + }, + { + "epoch": 0.03562249177075348, + "grad_norm": 0.7833011784616131, + "learning_rate": 3.99967201768683e-06, + "loss": 0.8162, + "step": 395 + }, + { + "epoch": 0.03571267529422374, + "grad_norm": 1.6501658269609532, + "learning_rate": 3.999661352590023e-06, + "loss": 1.0482, + "step": 396 + }, + { + "epoch": 0.03580285881769401, + "grad_norm": 1.6599898670071105, + "learning_rate": 3.999650516871083e-06, + "loss": 1.0638, + "step": 397 + }, + { + "epoch": 0.03589304234116427, + "grad_norm": 1.5079875310792132, + "learning_rate": 3.9996395105309365e-06, + "loss": 1.0869, + "step": 398 + }, + { + "epoch": 0.03598322586463453, + "grad_norm": 1.5151749562126349, + "learning_rate": 3.99962833357052e-06, + "loss": 1.0826, + "step": 399 + }, + { + "epoch": 0.03607340938810479, + "grad_norm": 2.1385250511459204, + "learning_rate": 3.999616985990789e-06, + "loss": 1.0433, + "step": 400 + }, + { + "epoch": 0.03616359291157505, + "grad_norm": 1.5125078216854912, + "learning_rate": 3.9996054677927104e-06, + "loss": 0.9991, + "step": 401 + }, + { + "epoch": 0.03625377643504532, + "grad_norm": 2.116389951365063, + "learning_rate": 3.9995937789772675e-06, + "loss": 0.9999, + "step": 402 + }, + { + "epoch": 0.03634395995851558, + "grad_norm": 0.7156598378326349, + "learning_rate": 3.999581919545458e-06, + "loss": 0.8411, + "step": 403 + }, + { + "epoch": 0.03643414348198584, + "grad_norm": 2.2834754435549245, + "learning_rate": 3.9995698894982935e-06, + "loss": 1.0853, + "step": 404 + }, + { + "epoch": 0.0365243270054561, + "grad_norm": 1.8611936250427847, + "learning_rate": 3.9995576888368e-06, + "loss": 1.1106, + "step": 405 + }, + { + "epoch": 0.036614510528926364, + "grad_norm": 1.7166397317915192, + "learning_rate": 3.9995453175620194e-06, + "loss": 1.0507, + "step": 406 + }, + { + "epoch": 0.036704694052396625, + "grad_norm": 2.0565704884393528, + "learning_rate": 3.999532775675007e-06, + "loss": 0.9186, + "step": 407 + }, + { + "epoch": 0.03679487757586689, + "grad_norm": 2.188709033907299, + "learning_rate": 3.9995200631768326e-06, + "loss": 1.0328, + "step": 408 + }, + { + "epoch": 0.03688506109933715, + "grad_norm": 2.066880746746351, + "learning_rate": 3.9995071800685815e-06, + "loss": 1.1884, + "step": 409 + }, + { + "epoch": 0.036975244622807414, + "grad_norm": 2.9387715610896064, + "learning_rate": 3.999494126351352e-06, + "loss": 1.0312, + "step": 410 + }, + { + "epoch": 0.037065428146277675, + "grad_norm": 3.7485918580139863, + "learning_rate": 3.99948090202626e-06, + "loss": 1.0632, + "step": 411 + }, + { + "epoch": 0.037155611669747936, + "grad_norm": 2.117746314593042, + "learning_rate": 3.999467507094431e-06, + "loss": 1.0647, + "step": 412 + }, + { + "epoch": 0.0372457951932182, + "grad_norm": 1.5829340113310206, + "learning_rate": 3.999453941557011e-06, + "loss": 0.9648, + "step": 413 + }, + { + "epoch": 0.037335978716688464, + "grad_norm": 0.73028300978699, + "learning_rate": 3.999440205415154e-06, + "loss": 0.7941, + "step": 414 + }, + { + "epoch": 0.037426162240158725, + "grad_norm": 2.2767362487719494, + "learning_rate": 3.999426298670035e-06, + "loss": 1.1088, + "step": 415 + }, + { + "epoch": 0.037516345763628986, + "grad_norm": 1.9072611268989388, + "learning_rate": 3.9994122213228385e-06, + "loss": 1.1047, + "step": 416 + }, + { + "epoch": 0.03760652928709925, + "grad_norm": 1.7736657621719147, + "learning_rate": 3.9993979733747675e-06, + "loss": 1.0924, + "step": 417 + }, + { + "epoch": 0.03769671281056951, + "grad_norm": 1.519063372522483, + "learning_rate": 3.999383554827037e-06, + "loss": 1.0621, + "step": 418 + }, + { + "epoch": 0.03778689633403977, + "grad_norm": 1.7139996165004734, + "learning_rate": 3.999368965680876e-06, + "loss": 1.0198, + "step": 419 + }, + { + "epoch": 0.037877079857510036, + "grad_norm": 5.815252165242822, + "learning_rate": 3.999354205937531e-06, + "loss": 1.1041, + "step": 420 + }, + { + "epoch": 0.0379672633809803, + "grad_norm": 2.0770702697077814, + "learning_rate": 3.999339275598261e-06, + "loss": 1.1037, + "step": 421 + }, + { + "epoch": 0.03805744690445056, + "grad_norm": 1.7251610915664828, + "learning_rate": 3.99932417466434e-06, + "loss": 1.0424, + "step": 422 + }, + { + "epoch": 0.03814763042792082, + "grad_norm": 2.6711034720490208, + "learning_rate": 3.999308903137056e-06, + "loss": 0.9719, + "step": 423 + }, + { + "epoch": 0.03823781395139108, + "grad_norm": 1.7255166454899573, + "learning_rate": 3.999293461017711e-06, + "loss": 1.1087, + "step": 424 + }, + { + "epoch": 0.03832799747486134, + "grad_norm": 2.1021947654121096, + "learning_rate": 3.9992778483076255e-06, + "loss": 1.0531, + "step": 425 + }, + { + "epoch": 0.03841818099833161, + "grad_norm": 1.5373572957872474, + "learning_rate": 3.99926206500813e-06, + "loss": 1.1166, + "step": 426 + }, + { + "epoch": 0.03850836452180187, + "grad_norm": 2.276528371163675, + "learning_rate": 3.999246111120571e-06, + "loss": 0.987, + "step": 427 + }, + { + "epoch": 0.03859854804527213, + "grad_norm": 1.5988566545429028, + "learning_rate": 3.999229986646311e-06, + "loss": 1.1338, + "step": 428 + }, + { + "epoch": 0.03868873156874239, + "grad_norm": 1.6902949887209817, + "learning_rate": 3.999213691586723e-06, + "loss": 1.0294, + "step": 429 + }, + { + "epoch": 0.03877891509221265, + "grad_norm": 3.7712781584045323, + "learning_rate": 3.9991972259432e-06, + "loss": 1.1061, + "step": 430 + }, + { + "epoch": 0.03886909861568291, + "grad_norm": 1.6345665376364136, + "learning_rate": 3.999180589717147e-06, + "loss": 1.1436, + "step": 431 + }, + { + "epoch": 0.03895928213915318, + "grad_norm": 1.786552300683468, + "learning_rate": 3.999163782909983e-06, + "loss": 0.9711, + "step": 432 + }, + { + "epoch": 0.03904946566262344, + "grad_norm": 2.44289085720256, + "learning_rate": 3.99914680552314e-06, + "loss": 1.0947, + "step": 433 + }, + { + "epoch": 0.0391396491860937, + "grad_norm": 2.2423732560998024, + "learning_rate": 3.999129657558069e-06, + "loss": 0.9925, + "step": 434 + }, + { + "epoch": 0.03922983270956396, + "grad_norm": 0.7278849963406284, + "learning_rate": 3.999112339016234e-06, + "loss": 0.8075, + "step": 435 + }, + { + "epoch": 0.03932001623303422, + "grad_norm": 1.5522934005550821, + "learning_rate": 3.999094849899109e-06, + "loss": 1.0728, + "step": 436 + }, + { + "epoch": 0.039410199756504484, + "grad_norm": 1.443218466935773, + "learning_rate": 3.99907719020819e-06, + "loss": 0.9789, + "step": 437 + }, + { + "epoch": 0.03950038327997475, + "grad_norm": 1.921250963080518, + "learning_rate": 3.999059359944982e-06, + "loss": 1.0853, + "step": 438 + }, + { + "epoch": 0.03959056680344501, + "grad_norm": 1.6813603786836426, + "learning_rate": 3.999041359111007e-06, + "loss": 1.1157, + "step": 439 + }, + { + "epoch": 0.03968075032691527, + "grad_norm": 1.6467846724405877, + "learning_rate": 3.999023187707801e-06, + "loss": 1.1354, + "step": 440 + }, + { + "epoch": 0.039770933850385534, + "grad_norm": 2.0610442370350475, + "learning_rate": 3.999004845736913e-06, + "loss": 1.1269, + "step": 441 + }, + { + "epoch": 0.039861117373855795, + "grad_norm": 1.823253629512671, + "learning_rate": 3.9989863331999096e-06, + "loss": 1.139, + "step": 442 + }, + { + "epoch": 0.039951300897326056, + "grad_norm": 2.3052615727687558, + "learning_rate": 3.99896765009837e-06, + "loss": 0.9966, + "step": 443 + }, + { + "epoch": 0.040041484420796324, + "grad_norm": 1.5059427319918857, + "learning_rate": 3.998948796433888e-06, + "loss": 1.0281, + "step": 444 + }, + { + "epoch": 0.040131667944266584, + "grad_norm": 1.775114031607708, + "learning_rate": 3.998929772208073e-06, + "loss": 1.102, + "step": 445 + }, + { + "epoch": 0.040221851467736845, + "grad_norm": 0.6874153561938198, + "learning_rate": 3.998910577422547e-06, + "loss": 0.8342, + "step": 446 + }, + { + "epoch": 0.040312034991207106, + "grad_norm": 1.541564920865456, + "learning_rate": 3.99889121207895e-06, + "loss": 1.1029, + "step": 447 + }, + { + "epoch": 0.04040221851467737, + "grad_norm": 1.6653871233987823, + "learning_rate": 3.9988716761789324e-06, + "loss": 1.0864, + "step": 448 + }, + { + "epoch": 0.04049240203814763, + "grad_norm": 1.892121408506872, + "learning_rate": 3.998851969724161e-06, + "loss": 1.0016, + "step": 449 + }, + { + "epoch": 0.040582585561617895, + "grad_norm": 2.0257713967880675, + "learning_rate": 3.998832092716319e-06, + "loss": 1.1114, + "step": 450 + }, + { + "epoch": 0.040672769085088156, + "grad_norm": 1.5396334755681191, + "learning_rate": 3.998812045157102e-06, + "loss": 1.0432, + "step": 451 + }, + { + "epoch": 0.04076295260855842, + "grad_norm": 1.806810591667028, + "learning_rate": 3.998791827048219e-06, + "loss": 1.111, + "step": 452 + }, + { + "epoch": 0.04085313613202868, + "grad_norm": 3.4045233593712343, + "learning_rate": 3.998771438391396e-06, + "loss": 1.0066, + "step": 453 + }, + { + "epoch": 0.04094331965549894, + "grad_norm": 0.6521369240730712, + "learning_rate": 3.9987508791883725e-06, + "loss": 0.812, + "step": 454 + }, + { + "epoch": 0.0410335031789692, + "grad_norm": 1.7772428714503523, + "learning_rate": 3.998730149440904e-06, + "loss": 1.1755, + "step": 455 + }, + { + "epoch": 0.04112368670243947, + "grad_norm": 1.7885005684783177, + "learning_rate": 3.998709249150758e-06, + "loss": 1.1118, + "step": 456 + }, + { + "epoch": 0.04121387022590973, + "grad_norm": 1.589875317009439, + "learning_rate": 3.998688178319717e-06, + "loss": 1.0861, + "step": 457 + }, + { + "epoch": 0.04130405374937999, + "grad_norm": 2.132386888229941, + "learning_rate": 3.9986669369495805e-06, + "loss": 1.0513, + "step": 458 + }, + { + "epoch": 0.04139423727285025, + "grad_norm": 1.837386707783756, + "learning_rate": 3.998645525042161e-06, + "loss": 1.1122, + "step": 459 + }, + { + "epoch": 0.04148442079632051, + "grad_norm": 2.1763600755753516, + "learning_rate": 3.998623942599284e-06, + "loss": 1.1335, + "step": 460 + }, + { + "epoch": 0.04157460431979077, + "grad_norm": 2.5096206563798087, + "learning_rate": 3.998602189622793e-06, + "loss": 1.0562, + "step": 461 + }, + { + "epoch": 0.04166478784326104, + "grad_norm": 2.09399241851775, + "learning_rate": 3.998580266114542e-06, + "loss": 1.096, + "step": 462 + }, + { + "epoch": 0.0417549713667313, + "grad_norm": 1.6767195628918476, + "learning_rate": 3.998558172076404e-06, + "loss": 0.9663, + "step": 463 + }, + { + "epoch": 0.04184515489020156, + "grad_norm": 1.8444186145947583, + "learning_rate": 3.998535907510262e-06, + "loss": 0.9949, + "step": 464 + }, + { + "epoch": 0.04193533841367182, + "grad_norm": 2.2205455123542333, + "learning_rate": 3.998513472418016e-06, + "loss": 1.0748, + "step": 465 + }, + { + "epoch": 0.04202552193714208, + "grad_norm": 1.912902605152844, + "learning_rate": 3.998490866801582e-06, + "loss": 1.0935, + "step": 466 + }, + { + "epoch": 0.04211570546061234, + "grad_norm": 2.4707400835612714, + "learning_rate": 3.998468090662886e-06, + "loss": 1.0525, + "step": 467 + }, + { + "epoch": 0.04220588898408261, + "grad_norm": 1.5755362854086872, + "learning_rate": 3.998445144003874e-06, + "loss": 0.9555, + "step": 468 + }, + { + "epoch": 0.04229607250755287, + "grad_norm": 2.0366208010855997, + "learning_rate": 3.998422026826504e-06, + "loss": 1.0877, + "step": 469 + }, + { + "epoch": 0.04238625603102313, + "grad_norm": 2.1810002356122014, + "learning_rate": 3.998398739132746e-06, + "loss": 1.0924, + "step": 470 + }, + { + "epoch": 0.04247643955449339, + "grad_norm": 1.7087176131086308, + "learning_rate": 3.99837528092459e-06, + "loss": 1.0184, + "step": 471 + }, + { + "epoch": 0.042566623077963654, + "grad_norm": 2.2600142886967607, + "learning_rate": 3.998351652204034e-06, + "loss": 1.1092, + "step": 472 + }, + { + "epoch": 0.042656806601433915, + "grad_norm": 1.5570757747089332, + "learning_rate": 3.998327852973098e-06, + "loss": 1.1012, + "step": 473 + }, + { + "epoch": 0.04274699012490418, + "grad_norm": 3.3878863124936522, + "learning_rate": 3.99830388323381e-06, + "loss": 1.0114, + "step": 474 + }, + { + "epoch": 0.042837173648374444, + "grad_norm": 1.6147189257712236, + "learning_rate": 3.998279742988216e-06, + "loss": 1.0951, + "step": 475 + }, + { + "epoch": 0.042927357171844704, + "grad_norm": 1.6372393597222001, + "learning_rate": 3.998255432238377e-06, + "loss": 1.0304, + "step": 476 + }, + { + "epoch": 0.043017540695314965, + "grad_norm": 1.7405637692651597, + "learning_rate": 3.9982309509863656e-06, + "loss": 1.1333, + "step": 477 + }, + { + "epoch": 0.043107724218785226, + "grad_norm": 1.6749864663103993, + "learning_rate": 3.998206299234272e-06, + "loss": 1.01, + "step": 478 + }, + { + "epoch": 0.04319790774225549, + "grad_norm": 1.5416179683657634, + "learning_rate": 3.998181476984198e-06, + "loss": 1.0775, + "step": 479 + }, + { + "epoch": 0.043288091265725755, + "grad_norm": 2.0187971599727432, + "learning_rate": 3.998156484238263e-06, + "loss": 0.9861, + "step": 480 + }, + { + "epoch": 0.043378274789196015, + "grad_norm": 1.9266171759328543, + "learning_rate": 3.998131320998599e-06, + "loss": 1.0006, + "step": 481 + }, + { + "epoch": 0.043468458312666276, + "grad_norm": 2.0553014729954975, + "learning_rate": 3.998105987267353e-06, + "loss": 1.1098, + "step": 482 + }, + { + "epoch": 0.04355864183613654, + "grad_norm": 1.7641874554610066, + "learning_rate": 3.998080483046687e-06, + "loss": 0.9877, + "step": 483 + }, + { + "epoch": 0.0436488253596068, + "grad_norm": 1.5480220664666968, + "learning_rate": 3.998054808338776e-06, + "loss": 1.0105, + "step": 484 + }, + { + "epoch": 0.04373900888307706, + "grad_norm": 3.2327272207554842, + "learning_rate": 3.998028963145812e-06, + "loss": 1.0251, + "step": 485 + }, + { + "epoch": 0.043829192406547327, + "grad_norm": 1.6553776440972696, + "learning_rate": 3.99800294747e-06, + "loss": 1.0721, + "step": 486 + }, + { + "epoch": 0.04391937593001759, + "grad_norm": 1.6878356599763766, + "learning_rate": 3.99797676131356e-06, + "loss": 1.0981, + "step": 487 + }, + { + "epoch": 0.04400955945348785, + "grad_norm": 2.002340258871749, + "learning_rate": 3.997950404678726e-06, + "loss": 0.9207, + "step": 488 + }, + { + "epoch": 0.04409974297695811, + "grad_norm": 2.13708280987162, + "learning_rate": 3.997923877567746e-06, + "loss": 1.1784, + "step": 489 + }, + { + "epoch": 0.04418992650042837, + "grad_norm": 1.5945535018093455, + "learning_rate": 3.9978971799828855e-06, + "loss": 0.891, + "step": 490 + }, + { + "epoch": 0.04428011002389863, + "grad_norm": 2.1500091907393495, + "learning_rate": 3.997870311926421e-06, + "loss": 1.05, + "step": 491 + }, + { + "epoch": 0.0443702935473689, + "grad_norm": 2.1605246369454214, + "learning_rate": 3.997843273400645e-06, + "loss": 1.0182, + "step": 492 + }, + { + "epoch": 0.04446047707083916, + "grad_norm": 1.5581171746024942, + "learning_rate": 3.997816064407865e-06, + "loss": 1.0381, + "step": 493 + }, + { + "epoch": 0.04455066059430942, + "grad_norm": 1.7244499711776164, + "learning_rate": 3.997788684950402e-06, + "loss": 1.0833, + "step": 494 + }, + { + "epoch": 0.04464084411777968, + "grad_norm": 3.041687447115489, + "learning_rate": 3.997761135030593e-06, + "loss": 1.1047, + "step": 495 + }, + { + "epoch": 0.04473102764124994, + "grad_norm": 1.738015555317651, + "learning_rate": 3.997733414650789e-06, + "loss": 0.9624, + "step": 496 + }, + { + "epoch": 0.0448212111647202, + "grad_norm": 2.781731810251711, + "learning_rate": 3.9977055238133554e-06, + "loss": 1.0191, + "step": 497 + }, + { + "epoch": 0.04491139468819047, + "grad_norm": 1.70027468089086, + "learning_rate": 3.99767746252067e-06, + "loss": 0.9411, + "step": 498 + }, + { + "epoch": 0.04500157821166073, + "grad_norm": 1.9045367437069527, + "learning_rate": 3.997649230775129e-06, + "loss": 1.0601, + "step": 499 + }, + { + "epoch": 0.04509176173513099, + "grad_norm": 2.269433549626301, + "learning_rate": 3.9976208285791395e-06, + "loss": 0.9987, + "step": 500 + }, + { + "epoch": 0.04518194525860125, + "grad_norm": 0.7777576590577504, + "learning_rate": 3.997592255935127e-06, + "loss": 0.9092, + "step": 501 + }, + { + "epoch": 0.045272128782071513, + "grad_norm": 2.8491218887236465, + "learning_rate": 3.997563512845529e-06, + "loss": 1.0871, + "step": 502 + }, + { + "epoch": 0.045362312305541774, + "grad_norm": 1.6794451161644486, + "learning_rate": 3.9975345993127975e-06, + "loss": 1.0182, + "step": 503 + }, + { + "epoch": 0.04545249582901204, + "grad_norm": 1.7134988503097448, + "learning_rate": 3.9975055153393985e-06, + "loss": 1.0743, + "step": 504 + }, + { + "epoch": 0.0455426793524823, + "grad_norm": 1.7781920703648064, + "learning_rate": 3.997476260927816e-06, + "loss": 1.0794, + "step": 505 + }, + { + "epoch": 0.045632862875952564, + "grad_norm": 1.839087847305675, + "learning_rate": 3.997446836080545e-06, + "loss": 1.0344, + "step": 506 + }, + { + "epoch": 0.045723046399422825, + "grad_norm": 2.1997282207116218, + "learning_rate": 3.997417240800095e-06, + "loss": 1.1108, + "step": 507 + }, + { + "epoch": 0.045813229922893085, + "grad_norm": 3.5934214566193843, + "learning_rate": 3.997387475088994e-06, + "loss": 1.1589, + "step": 508 + }, + { + "epoch": 0.045903413446363346, + "grad_norm": 2.480819749403072, + "learning_rate": 3.99735753894978e-06, + "loss": 1.1585, + "step": 509 + }, + { + "epoch": 0.045993596969833614, + "grad_norm": 1.931907087717976, + "learning_rate": 3.997327432385006e-06, + "loss": 1.0601, + "step": 510 + }, + { + "epoch": 0.046083780493303875, + "grad_norm": 1.9220865024706189, + "learning_rate": 3.997297155397244e-06, + "loss": 0.996, + "step": 511 + }, + { + "epoch": 0.046173964016774136, + "grad_norm": 1.6454230493283442, + "learning_rate": 3.997266707989074e-06, + "loss": 0.9788, + "step": 512 + }, + { + "epoch": 0.046264147540244396, + "grad_norm": 1.9950326387226054, + "learning_rate": 3.997236090163097e-06, + "loss": 0.983, + "step": 513 + }, + { + "epoch": 0.04635433106371466, + "grad_norm": 2.816883062668953, + "learning_rate": 3.9972053019219235e-06, + "loss": 1.1902, + "step": 514 + }, + { + "epoch": 0.04644451458718492, + "grad_norm": 3.1076035557353743, + "learning_rate": 3.997174343268181e-06, + "loss": 1.0765, + "step": 515 + }, + { + "epoch": 0.046534698110655186, + "grad_norm": 1.8831709524499158, + "learning_rate": 3.9971432142045115e-06, + "loss": 1.066, + "step": 516 + }, + { + "epoch": 0.04662488163412545, + "grad_norm": 0.8532832198093201, + "learning_rate": 3.99711191473357e-06, + "loss": 0.9009, + "step": 517 + }, + { + "epoch": 0.04671506515759571, + "grad_norm": 1.946299536141381, + "learning_rate": 3.99708044485803e-06, + "loss": 1.0783, + "step": 518 + }, + { + "epoch": 0.04680524868106597, + "grad_norm": 1.710856884315643, + "learning_rate": 3.997048804580574e-06, + "loss": 1.0664, + "step": 519 + }, + { + "epoch": 0.04689543220453623, + "grad_norm": 2.4804587536708023, + "learning_rate": 3.997016993903901e-06, + "loss": 1.0129, + "step": 520 + }, + { + "epoch": 0.04698561572800649, + "grad_norm": 2.5150230113923913, + "learning_rate": 3.996985012830728e-06, + "loss": 1.1375, + "step": 521 + }, + { + "epoch": 0.04707579925147676, + "grad_norm": 1.5332226405754386, + "learning_rate": 3.996952861363782e-06, + "loss": 1.0049, + "step": 522 + }, + { + "epoch": 0.04716598277494702, + "grad_norm": 1.4547121138083194, + "learning_rate": 3.9969205395058064e-06, + "loss": 1.0662, + "step": 523 + }, + { + "epoch": 0.04725616629841728, + "grad_norm": 1.7369873717732398, + "learning_rate": 3.99688804725956e-06, + "loss": 1.0582, + "step": 524 + }, + { + "epoch": 0.04734634982188754, + "grad_norm": 1.2629407509915644, + "learning_rate": 3.996855384627815e-06, + "loss": 1.013, + "step": 525 + }, + { + "epoch": 0.0474365333453578, + "grad_norm": 1.7405914385588874, + "learning_rate": 3.996822551613357e-06, + "loss": 1.0968, + "step": 526 + }, + { + "epoch": 0.04752671686882806, + "grad_norm": 1.8263691351933926, + "learning_rate": 3.996789548218989e-06, + "loss": 0.9893, + "step": 527 + }, + { + "epoch": 0.04761690039229833, + "grad_norm": 1.89820131944171, + "learning_rate": 3.996756374447526e-06, + "loss": 0.9467, + "step": 528 + }, + { + "epoch": 0.04770708391576859, + "grad_norm": 3.3251554467140347, + "learning_rate": 3.9967230303018005e-06, + "loss": 1.0635, + "step": 529 + }, + { + "epoch": 0.04779726743923885, + "grad_norm": 1.7275934377925006, + "learning_rate": 3.996689515784655e-06, + "loss": 1.0462, + "step": 530 + }, + { + "epoch": 0.04788745096270911, + "grad_norm": 1.9304221912977126, + "learning_rate": 3.996655830898951e-06, + "loss": 1.0333, + "step": 531 + }, + { + "epoch": 0.04797763448617937, + "grad_norm": 3.119850340701692, + "learning_rate": 3.996621975647562e-06, + "loss": 1.0584, + "step": 532 + }, + { + "epoch": 0.04806781800964964, + "grad_norm": 1.3316956933637318, + "learning_rate": 3.996587950033377e-06, + "loss": 0.9942, + "step": 533 + }, + { + "epoch": 0.0481580015331199, + "grad_norm": 2.5310044346125835, + "learning_rate": 3.996553754059299e-06, + "loss": 0.9898, + "step": 534 + }, + { + "epoch": 0.04824818505659016, + "grad_norm": 1.854950906869932, + "learning_rate": 3.996519387728245e-06, + "loss": 1.0561, + "step": 535 + }, + { + "epoch": 0.04833836858006042, + "grad_norm": 1.3277720262801038, + "learning_rate": 3.9964848510431495e-06, + "loss": 1.0395, + "step": 536 + }, + { + "epoch": 0.048428552103530684, + "grad_norm": 2.2344393420792152, + "learning_rate": 3.996450144006957e-06, + "loss": 1.1174, + "step": 537 + }, + { + "epoch": 0.048518735627000945, + "grad_norm": 4.428417743705821, + "learning_rate": 3.99641526662263e-06, + "loss": 1.0373, + "step": 538 + }, + { + "epoch": 0.04860891915047121, + "grad_norm": 1.8874030284097751, + "learning_rate": 3.996380218893145e-06, + "loss": 1.1399, + "step": 539 + }, + { + "epoch": 0.04869910267394147, + "grad_norm": 1.8858609036216956, + "learning_rate": 3.996345000821491e-06, + "loss": 1.0866, + "step": 540 + }, + { + "epoch": 0.048789286197411734, + "grad_norm": 3.253575632213158, + "learning_rate": 3.996309612410674e-06, + "loss": 1.1684, + "step": 541 + }, + { + "epoch": 0.048879469720881995, + "grad_norm": 1.9105852577535989, + "learning_rate": 3.996274053663713e-06, + "loss": 0.9946, + "step": 542 + }, + { + "epoch": 0.048969653244352256, + "grad_norm": 0.7896271375585227, + "learning_rate": 3.996238324583643e-06, + "loss": 0.8563, + "step": 543 + }, + { + "epoch": 0.049059836767822516, + "grad_norm": 2.0858009093636425, + "learning_rate": 3.996202425173512e-06, + "loss": 1.1242, + "step": 544 + }, + { + "epoch": 0.049150020291292784, + "grad_norm": 1.8457392290408967, + "learning_rate": 3.996166355436383e-06, + "loss": 1.0128, + "step": 545 + }, + { + "epoch": 0.049240203814763045, + "grad_norm": 1.5778005285294332, + "learning_rate": 3.996130115375333e-06, + "loss": 1.0703, + "step": 546 + }, + { + "epoch": 0.049330387338233306, + "grad_norm": 0.8836359218041692, + "learning_rate": 3.996093704993456e-06, + "loss": 0.8717, + "step": 547 + }, + { + "epoch": 0.04942057086170357, + "grad_norm": 1.731432781830362, + "learning_rate": 3.996057124293857e-06, + "loss": 1.1426, + "step": 548 + }, + { + "epoch": 0.04951075438517383, + "grad_norm": 2.040987472705066, + "learning_rate": 3.996020373279659e-06, + "loss": 1.1084, + "step": 549 + }, + { + "epoch": 0.04960093790864409, + "grad_norm": 1.8471958481986075, + "learning_rate": 3.995983451953996e-06, + "loss": 1.0663, + "step": 550 + }, + { + "epoch": 0.049691121432114356, + "grad_norm": 1.7175629417957734, + "learning_rate": 3.99594636032002e-06, + "loss": 1.0802, + "step": 551 + }, + { + "epoch": 0.04978130495558462, + "grad_norm": 2.480618690104015, + "learning_rate": 3.995909098380894e-06, + "loss": 0.9436, + "step": 552 + }, + { + "epoch": 0.04987148847905488, + "grad_norm": 1.6961495578744263, + "learning_rate": 3.995871666139799e-06, + "loss": 0.9439, + "step": 553 + }, + { + "epoch": 0.04996167200252514, + "grad_norm": 4.578307060293147, + "learning_rate": 3.995834063599928e-06, + "loss": 1.0357, + "step": 554 + }, + { + "epoch": 0.0500518555259954, + "grad_norm": 1.6269875622387049, + "learning_rate": 3.99579629076449e-06, + "loss": 1.0835, + "step": 555 + }, + { + "epoch": 0.05014203904946566, + "grad_norm": 1.995465383146727, + "learning_rate": 3.9957583476367084e-06, + "loss": 1.0325, + "step": 556 + }, + { + "epoch": 0.05023222257293593, + "grad_norm": 1.8776511205409672, + "learning_rate": 3.995720234219819e-06, + "loss": 1.0612, + "step": 557 + }, + { + "epoch": 0.05032240609640619, + "grad_norm": 1.6654099574306633, + "learning_rate": 3.995681950517075e-06, + "loss": 0.9641, + "step": 558 + }, + { + "epoch": 0.05041258961987645, + "grad_norm": 2.0369569862255092, + "learning_rate": 3.995643496531743e-06, + "loss": 1.0721, + "step": 559 + }, + { + "epoch": 0.05050277314334671, + "grad_norm": 2.194705539842694, + "learning_rate": 3.9956048722671044e-06, + "loss": 1.0666, + "step": 560 + }, + { + "epoch": 0.05059295666681697, + "grad_norm": 1.8329975369733962, + "learning_rate": 3.995566077726454e-06, + "loss": 1.0359, + "step": 561 + }, + { + "epoch": 0.05068314019028723, + "grad_norm": 3.57103042563085, + "learning_rate": 3.995527112913103e-06, + "loss": 1.0168, + "step": 562 + }, + { + "epoch": 0.0507733237137575, + "grad_norm": 4.331564958138371, + "learning_rate": 3.995487977830375e-06, + "loss": 1.0127, + "step": 563 + }, + { + "epoch": 0.05086350723722776, + "grad_norm": 1.7267384849375367, + "learning_rate": 3.9954486724816105e-06, + "loss": 1.1543, + "step": 564 + }, + { + "epoch": 0.05095369076069802, + "grad_norm": 2.4981791541037515, + "learning_rate": 3.995409196870161e-06, + "loss": 1.0954, + "step": 565 + }, + { + "epoch": 0.05104387428416828, + "grad_norm": 0.7807642761913829, + "learning_rate": 3.995369550999398e-06, + "loss": 0.8165, + "step": 566 + }, + { + "epoch": 0.05113405780763854, + "grad_norm": 2.1262748484474288, + "learning_rate": 3.995329734872702e-06, + "loss": 1.0823, + "step": 567 + }, + { + "epoch": 0.051224241331108804, + "grad_norm": 1.5561323496764254, + "learning_rate": 3.9952897484934706e-06, + "loss": 1.1732, + "step": 568 + }, + { + "epoch": 0.05131442485457907, + "grad_norm": 2.3042750102595164, + "learning_rate": 3.995249591865115e-06, + "loss": 1.0629, + "step": 569 + }, + { + "epoch": 0.05140460837804933, + "grad_norm": 1.8472549615634086, + "learning_rate": 3.995209264991063e-06, + "loss": 1.0157, + "step": 570 + }, + { + "epoch": 0.05149479190151959, + "grad_norm": 2.2556655338119715, + "learning_rate": 3.995168767874756e-06, + "loss": 1.1068, + "step": 571 + }, + { + "epoch": 0.051584975424989854, + "grad_norm": 1.6188198089275547, + "learning_rate": 3.995128100519648e-06, + "loss": 1.0437, + "step": 572 + }, + { + "epoch": 0.051675158948460115, + "grad_norm": 3.0011998002613667, + "learning_rate": 3.995087262929209e-06, + "loss": 1.0719, + "step": 573 + }, + { + "epoch": 0.051765342471930376, + "grad_norm": 1.7665045243957609, + "learning_rate": 3.995046255106925e-06, + "loss": 1.068, + "step": 574 + }, + { + "epoch": 0.05185552599540064, + "grad_norm": 2.240449238245324, + "learning_rate": 3.995005077056293e-06, + "loss": 1.1221, + "step": 575 + }, + { + "epoch": 0.051945709518870904, + "grad_norm": 2.336739290862002, + "learning_rate": 3.9949637287808284e-06, + "loss": 0.9507, + "step": 576 + }, + { + "epoch": 0.052035893042341165, + "grad_norm": 1.606048924666444, + "learning_rate": 3.994922210284057e-06, + "loss": 1.0057, + "step": 577 + }, + { + "epoch": 0.052126076565811426, + "grad_norm": 2.439994122857692, + "learning_rate": 3.994880521569524e-06, + "loss": 1.087, + "step": 578 + }, + { + "epoch": 0.05221626008928169, + "grad_norm": 1.8988796394295988, + "learning_rate": 3.994838662640785e-06, + "loss": 1.0664, + "step": 579 + }, + { + "epoch": 0.05230644361275195, + "grad_norm": 2.4063992144405826, + "learning_rate": 3.9947966335014116e-06, + "loss": 1.0941, + "step": 580 + }, + { + "epoch": 0.052396627136222215, + "grad_norm": 2.504688539463144, + "learning_rate": 3.99475443415499e-06, + "loss": 1.0331, + "step": 581 + }, + { + "epoch": 0.052486810659692476, + "grad_norm": 1.7305208288756202, + "learning_rate": 3.994712064605121e-06, + "loss": 1.0539, + "step": 582 + }, + { + "epoch": 0.05257699418316274, + "grad_norm": 2.1805495878274277, + "learning_rate": 3.99466952485542e-06, + "loss": 1.2049, + "step": 583 + }, + { + "epoch": 0.052667177706633, + "grad_norm": 1.9256149746374782, + "learning_rate": 3.994626814909518e-06, + "loss": 1.0757, + "step": 584 + }, + { + "epoch": 0.05275736123010326, + "grad_norm": 2.3636675053992517, + "learning_rate": 3.994583934771056e-06, + "loss": 1.0408, + "step": 585 + }, + { + "epoch": 0.05284754475357352, + "grad_norm": 1.5242675867278346, + "learning_rate": 3.9945408844436955e-06, + "loss": 1.1178, + "step": 586 + }, + { + "epoch": 0.05293772827704379, + "grad_norm": 2.043662309077731, + "learning_rate": 3.994497663931109e-06, + "loss": 1.0831, + "step": 587 + }, + { + "epoch": 0.05302791180051405, + "grad_norm": 1.8994399198914778, + "learning_rate": 3.994454273236984e-06, + "loss": 1.0763, + "step": 588 + }, + { + "epoch": 0.05311809532398431, + "grad_norm": 1.5296607041471686, + "learning_rate": 3.994410712365023e-06, + "loss": 1.063, + "step": 589 + }, + { + "epoch": 0.05320827884745457, + "grad_norm": 2.2563445391093757, + "learning_rate": 3.994366981318943e-06, + "loss": 1.1643, + "step": 590 + }, + { + "epoch": 0.05329846237092483, + "grad_norm": 1.7458462465062092, + "learning_rate": 3.9943230801024765e-06, + "loss": 1.0673, + "step": 591 + }, + { + "epoch": 0.05338864589439509, + "grad_norm": 1.9183922822491597, + "learning_rate": 3.9942790087193666e-06, + "loss": 1.1036, + "step": 592 + }, + { + "epoch": 0.05347882941786536, + "grad_norm": 2.1055817582583796, + "learning_rate": 3.994234767173376e-06, + "loss": 1.0518, + "step": 593 + }, + { + "epoch": 0.05356901294133562, + "grad_norm": 1.8040340941647792, + "learning_rate": 3.994190355468279e-06, + "loss": 0.9442, + "step": 594 + }, + { + "epoch": 0.05365919646480588, + "grad_norm": 1.5312557609605437, + "learning_rate": 3.994145773607865e-06, + "loss": 1.0598, + "step": 595 + }, + { + "epoch": 0.05374937998827614, + "grad_norm": 1.5096519990038477, + "learning_rate": 3.994101021595938e-06, + "loss": 1.0303, + "step": 596 + }, + { + "epoch": 0.0538395635117464, + "grad_norm": 1.8070095695913528, + "learning_rate": 3.9940560994363165e-06, + "loss": 1.0738, + "step": 597 + }, + { + "epoch": 0.05392974703521666, + "grad_norm": 2.3874514589817615, + "learning_rate": 3.994011007132833e-06, + "loss": 0.9323, + "step": 598 + }, + { + "epoch": 0.05401993055868693, + "grad_norm": 0.7028602313519063, + "learning_rate": 3.993965744689337e-06, + "loss": 0.8867, + "step": 599 + }, + { + "epoch": 0.05411011408215719, + "grad_norm": 2.2327387226961655, + "learning_rate": 3.993920312109687e-06, + "loss": 1.0432, + "step": 600 + }, + { + "epoch": 0.05420029760562745, + "grad_norm": 1.8977067584952116, + "learning_rate": 3.993874709397764e-06, + "loss": 1.0525, + "step": 601 + }, + { + "epoch": 0.05429048112909771, + "grad_norm": 2.4481830749557427, + "learning_rate": 3.993828936557454e-06, + "loss": 1.041, + "step": 602 + }, + { + "epoch": 0.054380664652567974, + "grad_norm": 2.3607847536714828, + "learning_rate": 3.993782993592667e-06, + "loss": 1.0192, + "step": 603 + }, + { + "epoch": 0.054470848176038235, + "grad_norm": 2.7541029406709727, + "learning_rate": 3.993736880507321e-06, + "loss": 1.0959, + "step": 604 + }, + { + "epoch": 0.0545610316995085, + "grad_norm": 2.40289111840897, + "learning_rate": 3.99369059730535e-06, + "loss": 1.0082, + "step": 605 + }, + { + "epoch": 0.05465121522297876, + "grad_norm": 1.386197317169746, + "learning_rate": 3.993644143990706e-06, + "loss": 1.0611, + "step": 606 + }, + { + "epoch": 0.054741398746449024, + "grad_norm": 2.134549394355483, + "learning_rate": 3.99359752056735e-06, + "loss": 0.9781, + "step": 607 + }, + { + "epoch": 0.054831582269919285, + "grad_norm": 1.7604075625332802, + "learning_rate": 3.993550727039261e-06, + "loss": 0.9097, + "step": 608 + }, + { + "epoch": 0.054921765793389546, + "grad_norm": 1.8918846914990206, + "learning_rate": 3.993503763410431e-06, + "loss": 1.0792, + "step": 609 + }, + { + "epoch": 0.05501194931685981, + "grad_norm": 1.88847436828833, + "learning_rate": 3.9934566296848686e-06, + "loss": 0.9756, + "step": 610 + }, + { + "epoch": 0.055102132840330074, + "grad_norm": 2.862635853304693, + "learning_rate": 3.993409325866595e-06, + "loss": 1.1391, + "step": 611 + }, + { + "epoch": 0.055192316363800335, + "grad_norm": 0.634405086536749, + "learning_rate": 3.993361851959645e-06, + "loss": 0.8352, + "step": 612 + }, + { + "epoch": 0.055282499887270596, + "grad_norm": 1.765615007490391, + "learning_rate": 3.993314207968071e-06, + "loss": 1.0995, + "step": 613 + }, + { + "epoch": 0.05537268341074086, + "grad_norm": 1.767571313328553, + "learning_rate": 3.993266393895938e-06, + "loss": 1.0429, + "step": 614 + }, + { + "epoch": 0.05546286693421112, + "grad_norm": 2.182904020232647, + "learning_rate": 3.993218409747326e-06, + "loss": 1.059, + "step": 615 + }, + { + "epoch": 0.05555305045768138, + "grad_norm": 1.4795060462428469, + "learning_rate": 3.993170255526328e-06, + "loss": 1.0538, + "step": 616 + }, + { + "epoch": 0.055643233981151646, + "grad_norm": 3.440197250566126, + "learning_rate": 3.993121931237054e-06, + "loss": 1.0855, + "step": 617 + }, + { + "epoch": 0.05573341750462191, + "grad_norm": 1.7794632901904028, + "learning_rate": 3.993073436883627e-06, + "loss": 1.0765, + "step": 618 + }, + { + "epoch": 0.05582360102809217, + "grad_norm": 1.976570822958329, + "learning_rate": 3.993024772470184e-06, + "loss": 1.0378, + "step": 619 + }, + { + "epoch": 0.05591378455156243, + "grad_norm": 1.937436318120064, + "learning_rate": 3.992975938000878e-06, + "loss": 1.0401, + "step": 620 + }, + { + "epoch": 0.05600396807503269, + "grad_norm": 0.6561437248280483, + "learning_rate": 3.992926933479876e-06, + "loss": 0.8244, + "step": 621 + }, + { + "epoch": 0.05609415159850295, + "grad_norm": 1.8240115846550455, + "learning_rate": 3.9928777589113595e-06, + "loss": 1.0825, + "step": 622 + }, + { + "epoch": 0.05618433512197322, + "grad_norm": 1.6394264566226184, + "learning_rate": 3.992828414299524e-06, + "loss": 1.0171, + "step": 623 + }, + { + "epoch": 0.05627451864544348, + "grad_norm": 1.638992660096394, + "learning_rate": 3.992778899648579e-06, + "loss": 1.0563, + "step": 624 + }, + { + "epoch": 0.05636470216891374, + "grad_norm": 2.003068239357061, + "learning_rate": 3.992729214962751e-06, + "loss": 1.0768, + "step": 625 + }, + { + "epoch": 0.056454885692384, + "grad_norm": 1.9035514742056652, + "learning_rate": 3.992679360246279e-06, + "loss": 1.0901, + "step": 626 + }, + { + "epoch": 0.05654506921585426, + "grad_norm": 1.8074751952224972, + "learning_rate": 3.992629335503416e-06, + "loss": 1.0892, + "step": 627 + }, + { + "epoch": 0.05663525273932452, + "grad_norm": 1.4386080741977123, + "learning_rate": 3.9925791407384304e-06, + "loss": 1.0389, + "step": 628 + }, + { + "epoch": 0.05672543626279479, + "grad_norm": 2.0225808233796916, + "learning_rate": 3.992528775955606e-06, + "loss": 1.1018, + "step": 629 + }, + { + "epoch": 0.05681561978626505, + "grad_norm": 2.3830120065685163, + "learning_rate": 3.992478241159239e-06, + "loss": 1.0265, + "step": 630 + }, + { + "epoch": 0.05690580330973531, + "grad_norm": 1.9208761581424358, + "learning_rate": 3.992427536353643e-06, + "loss": 1.0772, + "step": 631 + }, + { + "epoch": 0.05699598683320557, + "grad_norm": 2.0752857287288125, + "learning_rate": 3.992376661543143e-06, + "loss": 1.0457, + "step": 632 + }, + { + "epoch": 0.05708617035667583, + "grad_norm": 2.4157781831206075, + "learning_rate": 3.992325616732081e-06, + "loss": 1.0331, + "step": 633 + }, + { + "epoch": 0.057176353880146094, + "grad_norm": 3.6958911918460657, + "learning_rate": 3.992274401924811e-06, + "loss": 1.1607, + "step": 634 + }, + { + "epoch": 0.05726653740361636, + "grad_norm": 1.6239731552038144, + "learning_rate": 3.992223017125704e-06, + "loss": 0.9768, + "step": 635 + }, + { + "epoch": 0.05735672092708662, + "grad_norm": 1.5769752619949273, + "learning_rate": 3.992171462339145e-06, + "loss": 1.083, + "step": 636 + }, + { + "epoch": 0.057446904450556883, + "grad_norm": 5.390143950739385, + "learning_rate": 3.992119737569532e-06, + "loss": 1.1244, + "step": 637 + }, + { + "epoch": 0.057537087974027144, + "grad_norm": 1.8302061836115704, + "learning_rate": 3.992067842821277e-06, + "loss": 1.0539, + "step": 638 + }, + { + "epoch": 0.057627271497497405, + "grad_norm": 13.590478121871014, + "learning_rate": 3.99201577809881e-06, + "loss": 1.1888, + "step": 639 + }, + { + "epoch": 0.057717455020967666, + "grad_norm": 2.0339205279658326, + "learning_rate": 3.991963543406574e-06, + "loss": 1.0745, + "step": 640 + }, + { + "epoch": 0.057807638544437934, + "grad_norm": 2.1608710046385795, + "learning_rate": 3.991911138749024e-06, + "loss": 1.0536, + "step": 641 + }, + { + "epoch": 0.057897822067908195, + "grad_norm": 2.1132476296280136, + "learning_rate": 3.991858564130633e-06, + "loss": 1.0808, + "step": 642 + }, + { + "epoch": 0.057988005591378455, + "grad_norm": 5.929850901487084, + "learning_rate": 3.991805819555885e-06, + "loss": 1.0385, + "step": 643 + }, + { + "epoch": 0.058078189114848716, + "grad_norm": 1.7719545592085317, + "learning_rate": 3.991752905029283e-06, + "loss": 1.0598, + "step": 644 + }, + { + "epoch": 0.05816837263831898, + "grad_norm": 1.4985538505116527, + "learning_rate": 3.991699820555341e-06, + "loss": 0.9323, + "step": 645 + }, + { + "epoch": 0.05825855616178924, + "grad_norm": 3.3774786430491206, + "learning_rate": 3.991646566138588e-06, + "loss": 1.0574, + "step": 646 + }, + { + "epoch": 0.058348739685259506, + "grad_norm": 2.5225317306715636, + "learning_rate": 3.991593141783567e-06, + "loss": 0.9826, + "step": 647 + }, + { + "epoch": 0.058438923208729766, + "grad_norm": 2.776886935967193, + "learning_rate": 3.991539547494839e-06, + "loss": 1.0785, + "step": 648 + }, + { + "epoch": 0.05852910673220003, + "grad_norm": 3.262084794205976, + "learning_rate": 3.991485783276974e-06, + "loss": 1.0526, + "step": 649 + }, + { + "epoch": 0.05861929025567029, + "grad_norm": 1.671391283453086, + "learning_rate": 3.991431849134563e-06, + "loss": 1.1127, + "step": 650 + }, + { + "epoch": 0.05870947377914055, + "grad_norm": 2.7581467898658008, + "learning_rate": 3.991377745072205e-06, + "loss": 0.9482, + "step": 651 + }, + { + "epoch": 0.05879965730261081, + "grad_norm": 1.9480694457605923, + "learning_rate": 3.991323471094517e-06, + "loss": 1.0868, + "step": 652 + }, + { + "epoch": 0.05888984082608108, + "grad_norm": 2.0194533312761025, + "learning_rate": 3.991269027206131e-06, + "loss": 1.0278, + "step": 653 + }, + { + "epoch": 0.05898002434955134, + "grad_norm": 0.7883753286578448, + "learning_rate": 3.9912144134116916e-06, + "loss": 0.8771, + "step": 654 + }, + { + "epoch": 0.0590702078730216, + "grad_norm": 1.8636064225686537, + "learning_rate": 3.99115962971586e-06, + "loss": 0.9786, + "step": 655 + }, + { + "epoch": 0.05916039139649186, + "grad_norm": 2.630836265819302, + "learning_rate": 3.991104676123308e-06, + "loss": 1.0894, + "step": 656 + }, + { + "epoch": 0.05925057491996212, + "grad_norm": 0.652139300442026, + "learning_rate": 3.991049552638727e-06, + "loss": 0.8116, + "step": 657 + }, + { + "epoch": 0.05934075844343238, + "grad_norm": 2.3181962660476962, + "learning_rate": 3.99099425926682e-06, + "loss": 1.0185, + "step": 658 + }, + { + "epoch": 0.05943094196690265, + "grad_norm": 1.9743128100119656, + "learning_rate": 3.990938796012304e-06, + "loss": 1.0239, + "step": 659 + }, + { + "epoch": 0.05952112549037291, + "grad_norm": 2.3308386636654195, + "learning_rate": 3.990883162879912e-06, + "loss": 1.2115, + "step": 660 + }, + { + "epoch": 0.05961130901384317, + "grad_norm": 1.9235289984972133, + "learning_rate": 3.990827359874391e-06, + "loss": 1.1239, + "step": 661 + }, + { + "epoch": 0.05970149253731343, + "grad_norm": 2.000165694048846, + "learning_rate": 3.990771387000503e-06, + "loss": 1.0577, + "step": 662 + }, + { + "epoch": 0.05979167606078369, + "grad_norm": 2.122584821715829, + "learning_rate": 3.990715244263023e-06, + "loss": 1.0821, + "step": 663 + }, + { + "epoch": 0.05988185958425396, + "grad_norm": 0.7085633932758941, + "learning_rate": 3.990658931666741e-06, + "loss": 0.8438, + "step": 664 + }, + { + "epoch": 0.05997204310772422, + "grad_norm": 1.7495528739839092, + "learning_rate": 3.990602449216463e-06, + "loss": 0.9744, + "step": 665 + }, + { + "epoch": 0.06006222663119448, + "grad_norm": 1.542509681881485, + "learning_rate": 3.990545796917008e-06, + "loss": 0.9564, + "step": 666 + }, + { + "epoch": 0.06015241015466474, + "grad_norm": 1.7280941896641022, + "learning_rate": 3.99048897477321e-06, + "loss": 1.1038, + "step": 667 + }, + { + "epoch": 0.060242593678135004, + "grad_norm": 3.7468924043965437, + "learning_rate": 3.990431982789917e-06, + "loss": 1.0718, + "step": 668 + }, + { + "epoch": 0.060332777201605264, + "grad_norm": 1.6602546931980078, + "learning_rate": 3.990374820971992e-06, + "loss": 0.9897, + "step": 669 + }, + { + "epoch": 0.06042296072507553, + "grad_norm": 1.9292674978228181, + "learning_rate": 3.990317489324312e-06, + "loss": 0.9448, + "step": 670 + }, + { + "epoch": 0.06051314424854579, + "grad_norm": 1.8852315218173819, + "learning_rate": 3.99025998785177e-06, + "loss": 1.039, + "step": 671 + }, + { + "epoch": 0.060603327772016054, + "grad_norm": 1.5995212852061058, + "learning_rate": 3.990202316559271e-06, + "loss": 0.9967, + "step": 672 + }, + { + "epoch": 0.060693511295486315, + "grad_norm": 2.11317248954766, + "learning_rate": 3.990144475451738e-06, + "loss": 1.107, + "step": 673 + }, + { + "epoch": 0.060783694818956575, + "grad_norm": 1.5992735345995253, + "learning_rate": 3.9900864645341036e-06, + "loss": 1.0534, + "step": 674 + }, + { + "epoch": 0.060873878342426836, + "grad_norm": 2.0081854447909353, + "learning_rate": 3.990028283811319e-06, + "loss": 1.0626, + "step": 675 + }, + { + "epoch": 0.060964061865897104, + "grad_norm": 1.6782511402650753, + "learning_rate": 3.989969933288348e-06, + "loss": 1.0482, + "step": 676 + }, + { + "epoch": 0.061054245389367365, + "grad_norm": 0.670092947334414, + "learning_rate": 3.98991141297017e-06, + "loss": 0.7888, + "step": 677 + }, + { + "epoch": 0.061144428912837626, + "grad_norm": 2.2434728042113825, + "learning_rate": 3.989852722861778e-06, + "loss": 1.0014, + "step": 678 + }, + { + "epoch": 0.061234612436307886, + "grad_norm": 0.7233747055451258, + "learning_rate": 3.98979386296818e-06, + "loss": 0.8169, + "step": 679 + }, + { + "epoch": 0.06132479595977815, + "grad_norm": 1.6503891803802848, + "learning_rate": 3.989734833294398e-06, + "loss": 1.0347, + "step": 680 + }, + { + "epoch": 0.06141497948324841, + "grad_norm": 1.8832061482697808, + "learning_rate": 3.989675633845469e-06, + "loss": 1.1219, + "step": 681 + }, + { + "epoch": 0.061505163006718676, + "grad_norm": 1.9089747482316224, + "learning_rate": 3.989616264626443e-06, + "loss": 1.0304, + "step": 682 + }, + { + "epoch": 0.06159534653018894, + "grad_norm": 2.0118886931404596, + "learning_rate": 3.989556725642388e-06, + "loss": 1.0889, + "step": 683 + }, + { + "epoch": 0.0616855300536592, + "grad_norm": 1.9208011883785603, + "learning_rate": 3.989497016898382e-06, + "loss": 1.1081, + "step": 684 + }, + { + "epoch": 0.06177571357712946, + "grad_norm": 2.0813464672196376, + "learning_rate": 3.98943713839952e-06, + "loss": 1.0499, + "step": 685 + }, + { + "epoch": 0.06186589710059972, + "grad_norm": 0.6770635137347035, + "learning_rate": 3.9893770901509125e-06, + "loss": 0.8452, + "step": 686 + }, + { + "epoch": 0.06195608062406998, + "grad_norm": 1.6083894693463925, + "learning_rate": 3.989316872157682e-06, + "loss": 1.0442, + "step": 687 + }, + { + "epoch": 0.06204626414754025, + "grad_norm": 1.7640879194186907, + "learning_rate": 3.989256484424968e-06, + "loss": 1.1279, + "step": 688 + }, + { + "epoch": 0.06213644767101051, + "grad_norm": 1.8100315253531405, + "learning_rate": 3.98919592695792e-06, + "loss": 1.0195, + "step": 689 + }, + { + "epoch": 0.06222663119448077, + "grad_norm": 2.192044034411172, + "learning_rate": 3.9891351997617096e-06, + "loss": 1.08, + "step": 690 + }, + { + "epoch": 0.06231681471795103, + "grad_norm": 1.89441905525151, + "learning_rate": 3.989074302841514e-06, + "loss": 0.9968, + "step": 691 + }, + { + "epoch": 0.06240699824142129, + "grad_norm": 2.2387061123616996, + "learning_rate": 3.989013236202533e-06, + "loss": 1.1687, + "step": 692 + }, + { + "epoch": 0.06249718176489155, + "grad_norm": 1.5728212879222658, + "learning_rate": 3.988951999849974e-06, + "loss": 1.0914, + "step": 693 + }, + { + "epoch": 0.06258736528836181, + "grad_norm": 2.0463905052272, + "learning_rate": 3.988890593789064e-06, + "loss": 0.9688, + "step": 694 + }, + { + "epoch": 0.06267754881183207, + "grad_norm": 1.902470703901987, + "learning_rate": 3.9888290180250415e-06, + "loss": 1.0419, + "step": 695 + }, + { + "epoch": 0.06276773233530233, + "grad_norm": 3.0997324428151884, + "learning_rate": 3.988767272563161e-06, + "loss": 1.0204, + "step": 696 + }, + { + "epoch": 0.06285791585877261, + "grad_norm": 2.5855517142838833, + "learning_rate": 3.988705357408691e-06, + "loss": 0.9975, + "step": 697 + }, + { + "epoch": 0.06294809938224287, + "grad_norm": 1.6719194923256202, + "learning_rate": 3.9886432725669146e-06, + "loss": 1.0155, + "step": 698 + }, + { + "epoch": 0.06303828290571313, + "grad_norm": 1.9568216969270484, + "learning_rate": 3.988581018043128e-06, + "loss": 1.0175, + "step": 699 + }, + { + "epoch": 0.06312846642918339, + "grad_norm": 1.9643378089833514, + "learning_rate": 3.988518593842645e-06, + "loss": 1.0664, + "step": 700 + }, + { + "epoch": 0.06321864995265365, + "grad_norm": 1.8803279201134582, + "learning_rate": 3.9884559999707906e-06, + "loss": 1.0518, + "step": 701 + }, + { + "epoch": 0.06330883347612391, + "grad_norm": 1.8726677694909304, + "learning_rate": 3.988393236432906e-06, + "loss": 1.065, + "step": 702 + }, + { + "epoch": 0.06339901699959417, + "grad_norm": 1.9496044931597114, + "learning_rate": 3.988330303234347e-06, + "loss": 1.134, + "step": 703 + }, + { + "epoch": 0.06348920052306443, + "grad_norm": 1.8917869591241394, + "learning_rate": 3.988267200380483e-06, + "loss": 1.0069, + "step": 704 + }, + { + "epoch": 0.0635793840465347, + "grad_norm": 1.4456247353295835, + "learning_rate": 3.988203927876698e-06, + "loss": 1.1022, + "step": 705 + }, + { + "epoch": 0.06366956757000496, + "grad_norm": 1.8636952705969685, + "learning_rate": 3.988140485728391e-06, + "loss": 0.9399, + "step": 706 + }, + { + "epoch": 0.06375975109347522, + "grad_norm": 2.4704923637872773, + "learning_rate": 3.988076873940975e-06, + "loss": 1.0031, + "step": 707 + }, + { + "epoch": 0.06384993461694548, + "grad_norm": 2.1008994038396294, + "learning_rate": 3.9880130925198786e-06, + "loss": 1.0479, + "step": 708 + }, + { + "epoch": 0.06394011814041575, + "grad_norm": 3.081397007033273, + "learning_rate": 3.987949141470543e-06, + "loss": 1.0654, + "step": 709 + }, + { + "epoch": 0.06403030166388601, + "grad_norm": 1.7029169638029653, + "learning_rate": 3.987885020798425e-06, + "loss": 1.0537, + "step": 710 + }, + { + "epoch": 0.06412048518735627, + "grad_norm": 1.6404096189539195, + "learning_rate": 3.987820730508996e-06, + "loss": 1.0212, + "step": 711 + }, + { + "epoch": 0.06421066871082654, + "grad_norm": 1.9169406971402478, + "learning_rate": 3.987756270607742e-06, + "loss": 1.146, + "step": 712 + }, + { + "epoch": 0.0643008522342968, + "grad_norm": 0.659154956494985, + "learning_rate": 3.987691641100162e-06, + "loss": 0.8036, + "step": 713 + }, + { + "epoch": 0.06439103575776706, + "grad_norm": 2.713416557332238, + "learning_rate": 3.987626841991771e-06, + "loss": 1.1394, + "step": 714 + }, + { + "epoch": 0.06448121928123732, + "grad_norm": 2.6345932368954537, + "learning_rate": 3.987561873288099e-06, + "loss": 1.0915, + "step": 715 + }, + { + "epoch": 0.06457140280470758, + "grad_norm": 1.639605032532063, + "learning_rate": 3.987496734994688e-06, + "loss": 0.9992, + "step": 716 + }, + { + "epoch": 0.06466158632817784, + "grad_norm": 1.953943249010107, + "learning_rate": 3.987431427117097e-06, + "loss": 1.086, + "step": 717 + }, + { + "epoch": 0.0647517698516481, + "grad_norm": 1.9858971712473097, + "learning_rate": 3.9873659496608985e-06, + "loss": 1.0665, + "step": 718 + }, + { + "epoch": 0.06484195337511836, + "grad_norm": 2.5129931407186676, + "learning_rate": 3.987300302631678e-06, + "loss": 1.0044, + "step": 719 + }, + { + "epoch": 0.06493213689858862, + "grad_norm": 2.20213392976097, + "learning_rate": 3.987234486035039e-06, + "loss": 1.0915, + "step": 720 + }, + { + "epoch": 0.0650223204220589, + "grad_norm": 3.4140996516320983, + "learning_rate": 3.987168499876595e-06, + "loss": 1.103, + "step": 721 + }, + { + "epoch": 0.06511250394552916, + "grad_norm": 1.8071595142723358, + "learning_rate": 3.987102344161978e-06, + "loss": 1.0636, + "step": 722 + }, + { + "epoch": 0.06520268746899942, + "grad_norm": 1.9993399484560368, + "learning_rate": 3.987036018896832e-06, + "loss": 1.087, + "step": 723 + }, + { + "epoch": 0.06529287099246968, + "grad_norm": 1.5701417592917688, + "learning_rate": 3.986969524086817e-06, + "loss": 1.1176, + "step": 724 + }, + { + "epoch": 0.06538305451593994, + "grad_norm": 1.9686685575563776, + "learning_rate": 3.986902859737605e-06, + "loss": 1.0315, + "step": 725 + }, + { + "epoch": 0.0654732380394102, + "grad_norm": 1.685847285475849, + "learning_rate": 3.986836025854886e-06, + "loss": 0.9879, + "step": 726 + }, + { + "epoch": 0.06556342156288046, + "grad_norm": 2.001262862136176, + "learning_rate": 3.986769022444362e-06, + "loss": 1.0092, + "step": 727 + }, + { + "epoch": 0.06565360508635072, + "grad_norm": 1.882106537854971, + "learning_rate": 3.986701849511751e-06, + "loss": 1.0459, + "step": 728 + }, + { + "epoch": 0.06574378860982098, + "grad_norm": 2.123441910335629, + "learning_rate": 3.986634507062782e-06, + "loss": 1.0342, + "step": 729 + }, + { + "epoch": 0.06583397213329124, + "grad_norm": 3.561858002219981, + "learning_rate": 3.986566995103204e-06, + "loss": 1.0691, + "step": 730 + }, + { + "epoch": 0.0659241556567615, + "grad_norm": 1.8792259911803408, + "learning_rate": 3.986499313638776e-06, + "loss": 0.9844, + "step": 731 + }, + { + "epoch": 0.06601433918023177, + "grad_norm": 1.7255357822454236, + "learning_rate": 3.986431462675272e-06, + "loss": 0.9822, + "step": 732 + }, + { + "epoch": 0.06610452270370204, + "grad_norm": 2.2710909449425833, + "learning_rate": 3.9863634422184835e-06, + "loss": 1.0732, + "step": 733 + }, + { + "epoch": 0.0661947062271723, + "grad_norm": 1.9077444003461266, + "learning_rate": 3.986295252274213e-06, + "loss": 1.0861, + "step": 734 + }, + { + "epoch": 0.06628488975064256, + "grad_norm": 1.4492445786634498, + "learning_rate": 3.9862268928482796e-06, + "loss": 1.1017, + "step": 735 + }, + { + "epoch": 0.06637507327411282, + "grad_norm": 3.0565120779679034, + "learning_rate": 3.986158363946515e-06, + "loss": 0.9645, + "step": 736 + }, + { + "epoch": 0.06646525679758308, + "grad_norm": 1.8192621565236098, + "learning_rate": 3.9860896655747685e-06, + "loss": 1.0577, + "step": 737 + }, + { + "epoch": 0.06655544032105334, + "grad_norm": 1.671210611545442, + "learning_rate": 3.9860207977388994e-06, + "loss": 1.1293, + "step": 738 + }, + { + "epoch": 0.0666456238445236, + "grad_norm": 1.6289904090125973, + "learning_rate": 3.9859517604447854e-06, + "loss": 0.9845, + "step": 739 + }, + { + "epoch": 0.06673580736799387, + "grad_norm": 1.6105170040612684, + "learning_rate": 3.985882553698317e-06, + "loss": 1.0203, + "step": 740 + }, + { + "epoch": 0.06682599089146413, + "grad_norm": 1.8695331986349801, + "learning_rate": 3.985813177505399e-06, + "loss": 1.11, + "step": 741 + }, + { + "epoch": 0.06691617441493439, + "grad_norm": 1.860018706894191, + "learning_rate": 3.985743631871951e-06, + "loss": 1.0032, + "step": 742 + }, + { + "epoch": 0.06700635793840465, + "grad_norm": 1.8365305490588342, + "learning_rate": 3.985673916803907e-06, + "loss": 1.0501, + "step": 743 + }, + { + "epoch": 0.06709654146187491, + "grad_norm": 2.478896620545678, + "learning_rate": 3.985604032307215e-06, + "loss": 1.0723, + "step": 744 + }, + { + "epoch": 0.06718672498534518, + "grad_norm": 3.6915843779360356, + "learning_rate": 3.985533978387839e-06, + "loss": 1.0602, + "step": 745 + }, + { + "epoch": 0.06727690850881544, + "grad_norm": 1.5422550144123315, + "learning_rate": 3.985463755051756e-06, + "loss": 1.0906, + "step": 746 + }, + { + "epoch": 0.0673670920322857, + "grad_norm": 2.298804789435245, + "learning_rate": 3.9853933623049575e-06, + "loss": 1.0485, + "step": 747 + }, + { + "epoch": 0.06745727555575597, + "grad_norm": 1.695616945275011, + "learning_rate": 3.98532280015345e-06, + "loss": 1.0494, + "step": 748 + }, + { + "epoch": 0.06754745907922623, + "grad_norm": 1.7135737067321997, + "learning_rate": 3.985252068603254e-06, + "loss": 0.976, + "step": 749 + }, + { + "epoch": 0.06763764260269649, + "grad_norm": 1.8569199839676778, + "learning_rate": 3.985181167660406e-06, + "loss": 1.082, + "step": 750 + }, + { + "epoch": 0.06772782612616675, + "grad_norm": 1.7263317385470536, + "learning_rate": 3.985110097330953e-06, + "loss": 1.0783, + "step": 751 + }, + { + "epoch": 0.06781800964963701, + "grad_norm": 1.6107278711509103, + "learning_rate": 3.985038857620962e-06, + "loss": 1.046, + "step": 752 + }, + { + "epoch": 0.06790819317310727, + "grad_norm": 0.6989998288256718, + "learning_rate": 3.9849674485365094e-06, + "loss": 0.8274, + "step": 753 + }, + { + "epoch": 0.06799837669657753, + "grad_norm": 2.604074451085386, + "learning_rate": 3.98489587008369e-06, + "loss": 0.9785, + "step": 754 + }, + { + "epoch": 0.06808856022004779, + "grad_norm": 3.560353753590787, + "learning_rate": 3.98482412226861e-06, + "loss": 0.8914, + "step": 755 + }, + { + "epoch": 0.06817874374351805, + "grad_norm": 1.5192803418430043, + "learning_rate": 3.984752205097391e-06, + "loss": 1.0889, + "step": 756 + }, + { + "epoch": 0.06826892726698833, + "grad_norm": 1.8435800118398826, + "learning_rate": 3.984680118576171e-06, + "loss": 1.0574, + "step": 757 + }, + { + "epoch": 0.06835911079045859, + "grad_norm": 1.6702966815995228, + "learning_rate": 3.984607862711099e-06, + "loss": 1.0083, + "step": 758 + }, + { + "epoch": 0.06844929431392885, + "grad_norm": 2.1562440291612304, + "learning_rate": 3.984535437508341e-06, + "loss": 1.0871, + "step": 759 + }, + { + "epoch": 0.06853947783739911, + "grad_norm": 0.6644507675532749, + "learning_rate": 3.984462842974078e-06, + "loss": 0.8324, + "step": 760 + }, + { + "epoch": 0.06862966136086937, + "grad_norm": 2.551899172827004, + "learning_rate": 3.984390079114502e-06, + "loss": 1.0647, + "step": 761 + }, + { + "epoch": 0.06871984488433963, + "grad_norm": 1.7918506867881838, + "learning_rate": 3.984317145935824e-06, + "loss": 1.1678, + "step": 762 + }, + { + "epoch": 0.06881002840780989, + "grad_norm": 2.2784667084542676, + "learning_rate": 3.984244043444264e-06, + "loss": 1.0277, + "step": 763 + }, + { + "epoch": 0.06890021193128015, + "grad_norm": 2.0009115049361723, + "learning_rate": 3.984170771646062e-06, + "loss": 1.0028, + "step": 764 + }, + { + "epoch": 0.06899039545475041, + "grad_norm": 2.182229040395924, + "learning_rate": 3.9840973305474695e-06, + "loss": 1.0949, + "step": 765 + }, + { + "epoch": 0.06908057897822067, + "grad_norm": 1.9870773297847968, + "learning_rate": 3.984023720154752e-06, + "loss": 1.1012, + "step": 766 + }, + { + "epoch": 0.06917076250169094, + "grad_norm": 1.50579343025496, + "learning_rate": 3.9839499404741915e-06, + "loss": 1.0642, + "step": 767 + }, + { + "epoch": 0.0692609460251612, + "grad_norm": 0.741519828433524, + "learning_rate": 3.983875991512082e-06, + "loss": 0.8475, + "step": 768 + }, + { + "epoch": 0.06935112954863147, + "grad_norm": 1.9602888637913531, + "learning_rate": 3.9838018732747345e-06, + "loss": 1.0678, + "step": 769 + }, + { + "epoch": 0.06944131307210173, + "grad_norm": 2.8084106385621066, + "learning_rate": 3.9837275857684716e-06, + "loss": 1.042, + "step": 770 + }, + { + "epoch": 0.06953149659557199, + "grad_norm": 3.819525202658398, + "learning_rate": 3.983653128999634e-06, + "loss": 1.1881, + "step": 771 + }, + { + "epoch": 0.06962168011904225, + "grad_norm": 1.6843914080724762, + "learning_rate": 3.983578502974574e-06, + "loss": 1.0826, + "step": 772 + }, + { + "epoch": 0.06971186364251251, + "grad_norm": 1.7837015648235905, + "learning_rate": 3.983503707699658e-06, + "loss": 1.0419, + "step": 773 + }, + { + "epoch": 0.06980204716598278, + "grad_norm": 1.6310385234453442, + "learning_rate": 3.983428743181268e-06, + "loss": 1.082, + "step": 774 + }, + { + "epoch": 0.06989223068945304, + "grad_norm": 1.7580194648125111, + "learning_rate": 3.983353609425802e-06, + "loss": 1.1183, + "step": 775 + }, + { + "epoch": 0.0699824142129233, + "grad_norm": 2.1188888189249884, + "learning_rate": 3.983278306439671e-06, + "loss": 1.0488, + "step": 776 + }, + { + "epoch": 0.07007259773639356, + "grad_norm": 4.324294174901146, + "learning_rate": 3.983202834229297e-06, + "loss": 1.0851, + "step": 777 + }, + { + "epoch": 0.07016278125986382, + "grad_norm": 1.5743781451609937, + "learning_rate": 3.983127192801123e-06, + "loss": 1.1338, + "step": 778 + }, + { + "epoch": 0.07025296478333408, + "grad_norm": 2.070951197728426, + "learning_rate": 3.983051382161602e-06, + "loss": 1.0874, + "step": 779 + }, + { + "epoch": 0.07034314830680434, + "grad_norm": 2.0572573759544985, + "learning_rate": 3.982975402317203e-06, + "loss": 1.0764, + "step": 780 + }, + { + "epoch": 0.07043333183027461, + "grad_norm": 3.453986841420796, + "learning_rate": 3.982899253274409e-06, + "loss": 1.0774, + "step": 781 + }, + { + "epoch": 0.07052351535374488, + "grad_norm": 1.9831090177847177, + "learning_rate": 3.982822935039717e-06, + "loss": 1.0346, + "step": 782 + }, + { + "epoch": 0.07061369887721514, + "grad_norm": 2.479094068815981, + "learning_rate": 3.982746447619638e-06, + "loss": 0.9835, + "step": 783 + }, + { + "epoch": 0.0707038824006854, + "grad_norm": 0.7092209938687842, + "learning_rate": 3.9826697910207e-06, + "loss": 0.8212, + "step": 784 + }, + { + "epoch": 0.07079406592415566, + "grad_norm": 3.39871881131856, + "learning_rate": 3.982592965249442e-06, + "loss": 1.0347, + "step": 785 + }, + { + "epoch": 0.07088424944762592, + "grad_norm": 2.1908884235631843, + "learning_rate": 3.982515970312422e-06, + "loss": 1.0969, + "step": 786 + }, + { + "epoch": 0.07097443297109618, + "grad_norm": 3.2801818562312564, + "learning_rate": 3.982438806216207e-06, + "loss": 1.0233, + "step": 787 + }, + { + "epoch": 0.07106461649456644, + "grad_norm": 1.7296696662844966, + "learning_rate": 3.982361472967382e-06, + "loss": 1.0373, + "step": 788 + }, + { + "epoch": 0.0711548000180367, + "grad_norm": 1.726793644825311, + "learning_rate": 3.982283970572546e-06, + "loss": 1.1262, + "step": 789 + }, + { + "epoch": 0.07124498354150696, + "grad_norm": 2.1235320687024863, + "learning_rate": 3.982206299038311e-06, + "loss": 1.029, + "step": 790 + }, + { + "epoch": 0.07133516706497722, + "grad_norm": 1.6496043482276375, + "learning_rate": 3.9821284583713054e-06, + "loss": 1.1258, + "step": 791 + }, + { + "epoch": 0.07142535058844748, + "grad_norm": 2.283326131804681, + "learning_rate": 3.98205044857817e-06, + "loss": 1.0298, + "step": 792 + }, + { + "epoch": 0.07151553411191776, + "grad_norm": 0.7433115428102727, + "learning_rate": 3.981972269665561e-06, + "loss": 0.8506, + "step": 793 + }, + { + "epoch": 0.07160571763538802, + "grad_norm": 1.5694366593669502, + "learning_rate": 3.98189392164015e-06, + "loss": 1.0402, + "step": 794 + }, + { + "epoch": 0.07169590115885828, + "grad_norm": 1.814210282121096, + "learning_rate": 3.981815404508621e-06, + "loss": 1.096, + "step": 795 + }, + { + "epoch": 0.07178608468232854, + "grad_norm": 1.8632980761777935, + "learning_rate": 3.981736718277674e-06, + "loss": 1.0563, + "step": 796 + }, + { + "epoch": 0.0718762682057988, + "grad_norm": 2.2716838981924585, + "learning_rate": 3.9816578629540235e-06, + "loss": 1.0827, + "step": 797 + }, + { + "epoch": 0.07196645172926906, + "grad_norm": 1.7632807470075018, + "learning_rate": 3.981578838544398e-06, + "loss": 1.0611, + "step": 798 + }, + { + "epoch": 0.07205663525273932, + "grad_norm": 0.676036411690138, + "learning_rate": 3.981499645055539e-06, + "loss": 0.8044, + "step": 799 + }, + { + "epoch": 0.07214681877620958, + "grad_norm": 0.6581981217281906, + "learning_rate": 3.981420282494204e-06, + "loss": 0.8386, + "step": 800 + }, + { + "epoch": 0.07223700229967984, + "grad_norm": 0.7474592804423141, + "learning_rate": 3.981340750867166e-06, + "loss": 0.8366, + "step": 801 + }, + { + "epoch": 0.0723271858231501, + "grad_norm": 1.512726046482086, + "learning_rate": 3.981261050181209e-06, + "loss": 1.0374, + "step": 802 + }, + { + "epoch": 0.07241736934662037, + "grad_norm": 1.8049254157383274, + "learning_rate": 3.9811811804431355e-06, + "loss": 1.003, + "step": 803 + }, + { + "epoch": 0.07250755287009064, + "grad_norm": 2.1125619055350824, + "learning_rate": 3.981101141659759e-06, + "loss": 1.0936, + "step": 804 + }, + { + "epoch": 0.0725977363935609, + "grad_norm": 1.7263047384081596, + "learning_rate": 3.98102093383791e-06, + "loss": 1.0793, + "step": 805 + }, + { + "epoch": 0.07268791991703116, + "grad_norm": 1.7863230825520278, + "learning_rate": 3.9809405569844315e-06, + "loss": 0.9719, + "step": 806 + }, + { + "epoch": 0.07277810344050142, + "grad_norm": 1.7248329026380222, + "learning_rate": 3.980860011106182e-06, + "loss": 1.0134, + "step": 807 + }, + { + "epoch": 0.07286828696397168, + "grad_norm": 1.3690219016848437, + "learning_rate": 3.980779296210033e-06, + "loss": 0.9868, + "step": 808 + }, + { + "epoch": 0.07295847048744195, + "grad_norm": 1.7940554847174803, + "learning_rate": 3.980698412302874e-06, + "loss": 1.1288, + "step": 809 + }, + { + "epoch": 0.0730486540109122, + "grad_norm": 1.8741234319824467, + "learning_rate": 3.980617359391604e-06, + "loss": 1.0651, + "step": 810 + }, + { + "epoch": 0.07313883753438247, + "grad_norm": 2.0938143364430344, + "learning_rate": 3.98053613748314e-06, + "loss": 1.0074, + "step": 811 + }, + { + "epoch": 0.07322902105785273, + "grad_norm": 2.0104000769475885, + "learning_rate": 3.980454746584413e-06, + "loss": 0.9142, + "step": 812 + }, + { + "epoch": 0.07331920458132299, + "grad_norm": 2.0698301509152133, + "learning_rate": 3.9803731867023665e-06, + "loss": 1.1331, + "step": 813 + }, + { + "epoch": 0.07340938810479325, + "grad_norm": 1.9144695861146075, + "learning_rate": 3.9802914578439596e-06, + "loss": 1.0357, + "step": 814 + }, + { + "epoch": 0.07349957162826351, + "grad_norm": 2.211678963602813, + "learning_rate": 3.980209560016167e-06, + "loss": 1.152, + "step": 815 + }, + { + "epoch": 0.07358975515173379, + "grad_norm": 1.9085275145884169, + "learning_rate": 3.980127493225975e-06, + "loss": 1.1465, + "step": 816 + }, + { + "epoch": 0.07367993867520405, + "grad_norm": 2.656591685141185, + "learning_rate": 3.980045257480387e-06, + "loss": 1.0785, + "step": 817 + }, + { + "epoch": 0.0737701221986743, + "grad_norm": 4.87375742408596, + "learning_rate": 3.9799628527864205e-06, + "loss": 1.0072, + "step": 818 + }, + { + "epoch": 0.07386030572214457, + "grad_norm": 1.680877193256308, + "learning_rate": 3.979880279151106e-06, + "loss": 1.0379, + "step": 819 + }, + { + "epoch": 0.07395048924561483, + "grad_norm": 1.8161137170981356, + "learning_rate": 3.979797536581489e-06, + "loss": 0.9928, + "step": 820 + }, + { + "epoch": 0.07404067276908509, + "grad_norm": 2.5621900603635996, + "learning_rate": 3.97971462508463e-06, + "loss": 0.8933, + "step": 821 + }, + { + "epoch": 0.07413085629255535, + "grad_norm": 1.692412255631433, + "learning_rate": 3.979631544667603e-06, + "loss": 1.0332, + "step": 822 + }, + { + "epoch": 0.07422103981602561, + "grad_norm": 1.818677634670842, + "learning_rate": 3.979548295337496e-06, + "loss": 1.062, + "step": 823 + }, + { + "epoch": 0.07431122333949587, + "grad_norm": 1.7029572850513524, + "learning_rate": 3.9794648771014146e-06, + "loss": 1.0152, + "step": 824 + }, + { + "epoch": 0.07440140686296613, + "grad_norm": 1.6720882529325696, + "learning_rate": 3.9793812899664745e-06, + "loss": 1.0521, + "step": 825 + }, + { + "epoch": 0.0744915903864364, + "grad_norm": 2.118461085063165, + "learning_rate": 3.979297533939809e-06, + "loss": 1.0073, + "step": 826 + }, + { + "epoch": 0.07458177390990665, + "grad_norm": 1.7872191682035228, + "learning_rate": 3.979213609028564e-06, + "loss": 1.025, + "step": 827 + }, + { + "epoch": 0.07467195743337693, + "grad_norm": 1.7760302038932196, + "learning_rate": 3.979129515239901e-06, + "loss": 1.0602, + "step": 828 + }, + { + "epoch": 0.07476214095684719, + "grad_norm": 3.9387299569412035, + "learning_rate": 3.979045252580994e-06, + "loss": 1.0571, + "step": 829 + }, + { + "epoch": 0.07485232448031745, + "grad_norm": 1.9149867103653369, + "learning_rate": 3.978960821059034e-06, + "loss": 0.9917, + "step": 830 + }, + { + "epoch": 0.07494250800378771, + "grad_norm": 1.8695268859684817, + "learning_rate": 3.978876220681225e-06, + "loss": 1.0885, + "step": 831 + }, + { + "epoch": 0.07503269152725797, + "grad_norm": 3.368391560751655, + "learning_rate": 3.978791451454786e-06, + "loss": 0.9754, + "step": 832 + }, + { + "epoch": 0.07512287505072823, + "grad_norm": 2.6261437058941706, + "learning_rate": 3.978706513386949e-06, + "loss": 1.0061, + "step": 833 + }, + { + "epoch": 0.0752130585741985, + "grad_norm": 2.4616155749860034, + "learning_rate": 3.978621406484962e-06, + "loss": 1.0601, + "step": 834 + }, + { + "epoch": 0.07530324209766875, + "grad_norm": 2.327618127728226, + "learning_rate": 3.978536130756086e-06, + "loss": 1.0396, + "step": 835 + }, + { + "epoch": 0.07539342562113902, + "grad_norm": 1.7210373656690114, + "learning_rate": 3.978450686207599e-06, + "loss": 1.1112, + "step": 836 + }, + { + "epoch": 0.07548360914460928, + "grad_norm": 2.3890569194082345, + "learning_rate": 3.978365072846789e-06, + "loss": 1.0274, + "step": 837 + }, + { + "epoch": 0.07557379266807954, + "grad_norm": 1.6257619538598773, + "learning_rate": 3.9782792906809625e-06, + "loss": 0.9968, + "step": 838 + }, + { + "epoch": 0.0756639761915498, + "grad_norm": 1.6638929970247773, + "learning_rate": 3.97819333971744e-06, + "loss": 1.0336, + "step": 839 + }, + { + "epoch": 0.07575415971502007, + "grad_norm": 3.1918808993824137, + "learning_rate": 3.978107219963553e-06, + "loss": 1.0641, + "step": 840 + }, + { + "epoch": 0.07584434323849033, + "grad_norm": 1.7093573811684835, + "learning_rate": 3.978020931426651e-06, + "loss": 0.9539, + "step": 841 + }, + { + "epoch": 0.0759345267619606, + "grad_norm": 1.8958292182067253, + "learning_rate": 3.977934474114096e-06, + "loss": 0.9931, + "step": 842 + }, + { + "epoch": 0.07602471028543085, + "grad_norm": 0.6579321011799829, + "learning_rate": 3.977847848033267e-06, + "loss": 0.8164, + "step": 843 + }, + { + "epoch": 0.07611489380890112, + "grad_norm": 1.9886989791722482, + "learning_rate": 3.977761053191553e-06, + "loss": 1.0902, + "step": 844 + }, + { + "epoch": 0.07620507733237138, + "grad_norm": 2.2914247789749056, + "learning_rate": 3.977674089596361e-06, + "loss": 1.0304, + "step": 845 + }, + { + "epoch": 0.07629526085584164, + "grad_norm": 0.6903580948160075, + "learning_rate": 3.97758695725511e-06, + "loss": 0.8129, + "step": 846 + }, + { + "epoch": 0.0763854443793119, + "grad_norm": 1.55844008374196, + "learning_rate": 3.977499656175236e-06, + "loss": 1.042, + "step": 847 + }, + { + "epoch": 0.07647562790278216, + "grad_norm": 3.620708292801968, + "learning_rate": 3.977412186364187e-06, + "loss": 0.957, + "step": 848 + }, + { + "epoch": 0.07656581142625242, + "grad_norm": 2.958707830517773, + "learning_rate": 3.977324547829428e-06, + "loss": 1.0711, + "step": 849 + }, + { + "epoch": 0.07665599494972268, + "grad_norm": 2.596066922311212, + "learning_rate": 3.977236740578435e-06, + "loss": 1.0687, + "step": 850 + }, + { + "epoch": 0.07674617847319294, + "grad_norm": 2.1782472288844583, + "learning_rate": 3.9771487646187015e-06, + "loss": 1.0311, + "step": 851 + }, + { + "epoch": 0.07683636199666322, + "grad_norm": 2.2354493659318844, + "learning_rate": 3.9770606199577325e-06, + "loss": 1.0844, + "step": 852 + }, + { + "epoch": 0.07692654552013348, + "grad_norm": 2.020097487146218, + "learning_rate": 3.9769723066030505e-06, + "loss": 0.9799, + "step": 853 + }, + { + "epoch": 0.07701672904360374, + "grad_norm": 1.8767938298201214, + "learning_rate": 3.976883824562191e-06, + "loss": 1.1054, + "step": 854 + }, + { + "epoch": 0.077106912567074, + "grad_norm": 1.7255898752475232, + "learning_rate": 3.976795173842703e-06, + "loss": 1.0447, + "step": 855 + }, + { + "epoch": 0.07719709609054426, + "grad_norm": 7.006486884271793, + "learning_rate": 3.97670635445215e-06, + "loss": 1.0253, + "step": 856 + }, + { + "epoch": 0.07728727961401452, + "grad_norm": 1.7820292993356748, + "learning_rate": 3.976617366398112e-06, + "loss": 1.0204, + "step": 857 + }, + { + "epoch": 0.07737746313748478, + "grad_norm": 1.4177958905914243, + "learning_rate": 3.976528209688181e-06, + "loss": 1.0765, + "step": 858 + }, + { + "epoch": 0.07746764666095504, + "grad_norm": 3.695903835606162, + "learning_rate": 3.976438884329965e-06, + "loss": 1.049, + "step": 859 + }, + { + "epoch": 0.0775578301844253, + "grad_norm": 2.135751006461845, + "learning_rate": 3.976349390331085e-06, + "loss": 1.1752, + "step": 860 + }, + { + "epoch": 0.07764801370789556, + "grad_norm": 1.760147104184517, + "learning_rate": 3.976259727699178e-06, + "loss": 1.0122, + "step": 861 + }, + { + "epoch": 0.07773819723136582, + "grad_norm": 2.9847624362488694, + "learning_rate": 3.976169896441895e-06, + "loss": 1.0005, + "step": 862 + }, + { + "epoch": 0.07782838075483609, + "grad_norm": 2.009908213251491, + "learning_rate": 3.976079896566898e-06, + "loss": 1.0811, + "step": 863 + }, + { + "epoch": 0.07791856427830636, + "grad_norm": 3.425389962376276, + "learning_rate": 3.97598972808187e-06, + "loss": 1.1123, + "step": 864 + }, + { + "epoch": 0.07800874780177662, + "grad_norm": 2.46775488776665, + "learning_rate": 3.975899390994501e-06, + "loss": 1.0473, + "step": 865 + }, + { + "epoch": 0.07809893132524688, + "grad_norm": 1.644175873911245, + "learning_rate": 3.975808885312502e-06, + "loss": 1.1081, + "step": 866 + }, + { + "epoch": 0.07818911484871714, + "grad_norm": 1.528243402367929, + "learning_rate": 3.975718211043594e-06, + "loss": 1.093, + "step": 867 + }, + { + "epoch": 0.0782792983721874, + "grad_norm": 2.219593263946063, + "learning_rate": 3.975627368195515e-06, + "loss": 0.9596, + "step": 868 + }, + { + "epoch": 0.07836948189565766, + "grad_norm": 1.513091417812095, + "learning_rate": 3.975536356776015e-06, + "loss": 1.0126, + "step": 869 + }, + { + "epoch": 0.07845966541912792, + "grad_norm": 0.7464267726454112, + "learning_rate": 3.975445176792861e-06, + "loss": 0.8905, + "step": 870 + }, + { + "epoch": 0.07854984894259819, + "grad_norm": 1.6520580288206648, + "learning_rate": 3.975353828253831e-06, + "loss": 1.052, + "step": 871 + }, + { + "epoch": 0.07864003246606845, + "grad_norm": 2.036180703311193, + "learning_rate": 3.97526231116672e-06, + "loss": 1.0477, + "step": 872 + }, + { + "epoch": 0.07873021598953871, + "grad_norm": 3.056305673790347, + "learning_rate": 3.975170625539338e-06, + "loss": 0.9861, + "step": 873 + }, + { + "epoch": 0.07882039951300897, + "grad_norm": 1.7846406488375743, + "learning_rate": 3.975078771379507e-06, + "loss": 1.0476, + "step": 874 + }, + { + "epoch": 0.07891058303647923, + "grad_norm": 2.309149299298379, + "learning_rate": 3.974986748695064e-06, + "loss": 0.9574, + "step": 875 + }, + { + "epoch": 0.0790007665599495, + "grad_norm": 0.7244268223391356, + "learning_rate": 3.974894557493862e-06, + "loss": 0.8281, + "step": 876 + }, + { + "epoch": 0.07909095008341976, + "grad_norm": 1.8673103962440258, + "learning_rate": 3.974802197783768e-06, + "loss": 1.0935, + "step": 877 + }, + { + "epoch": 0.07918113360689003, + "grad_norm": 1.9473419021466676, + "learning_rate": 3.974709669572661e-06, + "loss": 1.1036, + "step": 878 + }, + { + "epoch": 0.07927131713036029, + "grad_norm": 1.6984714311651317, + "learning_rate": 3.974616972868436e-06, + "loss": 1.0349, + "step": 879 + }, + { + "epoch": 0.07936150065383055, + "grad_norm": 1.6141167571558142, + "learning_rate": 3.974524107679003e-06, + "loss": 1.0483, + "step": 880 + }, + { + "epoch": 0.07945168417730081, + "grad_norm": 1.5669769997676748, + "learning_rate": 3.974431074012286e-06, + "loss": 1.0393, + "step": 881 + }, + { + "epoch": 0.07954186770077107, + "grad_norm": 2.0177641170774963, + "learning_rate": 3.974337871876223e-06, + "loss": 1.0681, + "step": 882 + }, + { + "epoch": 0.07963205122424133, + "grad_norm": 1.8586229999512747, + "learning_rate": 3.974244501278766e-06, + "loss": 1.0124, + "step": 883 + }, + { + "epoch": 0.07972223474771159, + "grad_norm": 1.6453068369942756, + "learning_rate": 3.974150962227883e-06, + "loss": 0.9978, + "step": 884 + }, + { + "epoch": 0.07981241827118185, + "grad_norm": 2.670797833600036, + "learning_rate": 3.9740572547315535e-06, + "loss": 0.9741, + "step": 885 + }, + { + "epoch": 0.07990260179465211, + "grad_norm": 2.465463889772564, + "learning_rate": 3.973963378797775e-06, + "loss": 1.0819, + "step": 886 + }, + { + "epoch": 0.07999278531812237, + "grad_norm": 2.614296480732888, + "learning_rate": 3.973869334434556e-06, + "loss": 1.0606, + "step": 887 + }, + { + "epoch": 0.08008296884159265, + "grad_norm": 1.8624997209382328, + "learning_rate": 3.973775121649922e-06, + "loss": 1.0616, + "step": 888 + }, + { + "epoch": 0.08017315236506291, + "grad_norm": 10.445245666974857, + "learning_rate": 3.973680740451911e-06, + "loss": 0.9611, + "step": 889 + }, + { + "epoch": 0.08026333588853317, + "grad_norm": 1.6334925791253962, + "learning_rate": 3.9735861908485776e-06, + "loss": 1.0112, + "step": 890 + }, + { + "epoch": 0.08035351941200343, + "grad_norm": 1.7289422667105, + "learning_rate": 3.973491472847987e-06, + "loss": 1.0635, + "step": 891 + }, + { + "epoch": 0.08044370293547369, + "grad_norm": 3.6935742021736737, + "learning_rate": 3.973396586458222e-06, + "loss": 0.9753, + "step": 892 + }, + { + "epoch": 0.08053388645894395, + "grad_norm": 1.8207406780607158, + "learning_rate": 3.97330153168738e-06, + "loss": 1.0302, + "step": 893 + }, + { + "epoch": 0.08062406998241421, + "grad_norm": 1.7919655779700108, + "learning_rate": 3.973206308543571e-06, + "loss": 0.9895, + "step": 894 + }, + { + "epoch": 0.08071425350588447, + "grad_norm": 1.8750048955217666, + "learning_rate": 3.973110917034918e-06, + "loss": 1.0176, + "step": 895 + }, + { + "epoch": 0.08080443702935473, + "grad_norm": 3.268365905251857, + "learning_rate": 3.973015357169563e-06, + "loss": 0.9548, + "step": 896 + }, + { + "epoch": 0.080894620552825, + "grad_norm": 1.7666784199137766, + "learning_rate": 3.972919628955659e-06, + "loss": 0.9346, + "step": 897 + }, + { + "epoch": 0.08098480407629526, + "grad_norm": 2.6373508176150096, + "learning_rate": 3.972823732401373e-06, + "loss": 1.0257, + "step": 898 + }, + { + "epoch": 0.08107498759976552, + "grad_norm": 6.673708089899456, + "learning_rate": 3.972727667514888e-06, + "loss": 1.0373, + "step": 899 + }, + { + "epoch": 0.08116517112323579, + "grad_norm": 1.7275762559370613, + "learning_rate": 3.972631434304402e-06, + "loss": 1.0036, + "step": 900 + }, + { + "epoch": 0.08125535464670605, + "grad_norm": 2.7089521630317206, + "learning_rate": 3.972535032778124e-06, + "loss": 1.1585, + "step": 901 + }, + { + "epoch": 0.08134553817017631, + "grad_norm": 1.7566299190453725, + "learning_rate": 3.97243846294428e-06, + "loss": 1.1293, + "step": 902 + }, + { + "epoch": 0.08143572169364657, + "grad_norm": 2.693906998097409, + "learning_rate": 3.972341724811111e-06, + "loss": 1.1771, + "step": 903 + }, + { + "epoch": 0.08152590521711683, + "grad_norm": 2.209974315028004, + "learning_rate": 3.972244818386872e-06, + "loss": 0.987, + "step": 904 + }, + { + "epoch": 0.0816160887405871, + "grad_norm": 3.4596662748504228, + "learning_rate": 3.972147743679828e-06, + "loss": 0.9852, + "step": 905 + }, + { + "epoch": 0.08170627226405736, + "grad_norm": 1.7296423047233838, + "learning_rate": 3.972050500698265e-06, + "loss": 1.006, + "step": 906 + }, + { + "epoch": 0.08179645578752762, + "grad_norm": 4.0202692983584125, + "learning_rate": 3.971953089450481e-06, + "loss": 1.0354, + "step": 907 + }, + { + "epoch": 0.08188663931099788, + "grad_norm": 1.6015087583875833, + "learning_rate": 3.971855509944784e-06, + "loss": 1.1095, + "step": 908 + }, + { + "epoch": 0.08197682283446814, + "grad_norm": 1.7454290639668195, + "learning_rate": 3.971757762189504e-06, + "loss": 1.0274, + "step": 909 + }, + { + "epoch": 0.0820670063579384, + "grad_norm": 2.6367719178667284, + "learning_rate": 3.9716598461929785e-06, + "loss": 1.0259, + "step": 910 + }, + { + "epoch": 0.08215718988140866, + "grad_norm": 1.6746339938981554, + "learning_rate": 3.971561761963563e-06, + "loss": 1.0128, + "step": 911 + }, + { + "epoch": 0.08224737340487893, + "grad_norm": 1.9068545023998957, + "learning_rate": 3.971463509509628e-06, + "loss": 1.1155, + "step": 912 + }, + { + "epoch": 0.0823375569283492, + "grad_norm": 2.224964411590221, + "learning_rate": 3.9713650888395555e-06, + "loss": 1.0314, + "step": 913 + }, + { + "epoch": 0.08242774045181946, + "grad_norm": 2.0814313471603847, + "learning_rate": 3.9712664999617425e-06, + "loss": 1.034, + "step": 914 + }, + { + "epoch": 0.08251792397528972, + "grad_norm": 2.826324479860762, + "learning_rate": 3.971167742884603e-06, + "loss": 1.0154, + "step": 915 + }, + { + "epoch": 0.08260810749875998, + "grad_norm": 1.9165679104559286, + "learning_rate": 3.971068817616564e-06, + "loss": 1.0868, + "step": 916 + }, + { + "epoch": 0.08269829102223024, + "grad_norm": 1.7134466019283798, + "learning_rate": 3.970969724166064e-06, + "loss": 1.0645, + "step": 917 + }, + { + "epoch": 0.0827884745457005, + "grad_norm": 0.8242465913960697, + "learning_rate": 3.970870462541559e-06, + "loss": 0.8157, + "step": 918 + }, + { + "epoch": 0.08287865806917076, + "grad_norm": 2.7399788245893433, + "learning_rate": 3.97077103275152e-06, + "loss": 1.0595, + "step": 919 + }, + { + "epoch": 0.08296884159264102, + "grad_norm": 2.006763228609864, + "learning_rate": 3.970671434804428e-06, + "loss": 1.0279, + "step": 920 + }, + { + "epoch": 0.08305902511611128, + "grad_norm": 3.657484579725215, + "learning_rate": 3.970571668708784e-06, + "loss": 1.0355, + "step": 921 + }, + { + "epoch": 0.08314920863958154, + "grad_norm": 0.6768028167248904, + "learning_rate": 3.9704717344731e-06, + "loss": 0.8261, + "step": 922 + }, + { + "epoch": 0.0832393921630518, + "grad_norm": 0.9214157965134658, + "learning_rate": 3.9703716321059026e-06, + "loss": 0.842, + "step": 923 + }, + { + "epoch": 0.08332957568652208, + "grad_norm": 1.739948972852682, + "learning_rate": 3.9702713616157325e-06, + "loss": 1.1271, + "step": 924 + }, + { + "epoch": 0.08341975920999234, + "grad_norm": 2.9268920582533453, + "learning_rate": 3.9701709230111455e-06, + "loss": 1.0519, + "step": 925 + }, + { + "epoch": 0.0835099427334626, + "grad_norm": 1.963789488300085, + "learning_rate": 3.970070316300713e-06, + "loss": 1.0059, + "step": 926 + }, + { + "epoch": 0.08360012625693286, + "grad_norm": 1.9113439444287412, + "learning_rate": 3.969969541493017e-06, + "loss": 1.0588, + "step": 927 + }, + { + "epoch": 0.08369030978040312, + "grad_norm": 2.0918464547914564, + "learning_rate": 3.969868598596658e-06, + "loss": 1.0911, + "step": 928 + }, + { + "epoch": 0.08378049330387338, + "grad_norm": 1.6859070358847281, + "learning_rate": 3.969767487620249e-06, + "loss": 0.9829, + "step": 929 + }, + { + "epoch": 0.08387067682734364, + "grad_norm": 1.3697343918269127, + "learning_rate": 3.969666208572416e-06, + "loss": 1.0901, + "step": 930 + }, + { + "epoch": 0.0839608603508139, + "grad_norm": 2.229049204170751, + "learning_rate": 3.969564761461802e-06, + "loss": 1.1076, + "step": 931 + }, + { + "epoch": 0.08405104387428416, + "grad_norm": 1.6404232082699985, + "learning_rate": 3.969463146297062e-06, + "loss": 1.0206, + "step": 932 + }, + { + "epoch": 0.08414122739775443, + "grad_norm": 2.2506940618892184, + "learning_rate": 3.969361363086867e-06, + "loss": 1.0601, + "step": 933 + }, + { + "epoch": 0.08423141092122469, + "grad_norm": 1.9135075601588394, + "learning_rate": 3.9692594118399014e-06, + "loss": 1.0111, + "step": 934 + }, + { + "epoch": 0.08432159444469496, + "grad_norm": 3.671892401471399, + "learning_rate": 3.969157292564865e-06, + "loss": 1.0753, + "step": 935 + }, + { + "epoch": 0.08441177796816522, + "grad_norm": 1.7435117059680605, + "learning_rate": 3.96905500527047e-06, + "loss": 1.0288, + "step": 936 + }, + { + "epoch": 0.08450196149163548, + "grad_norm": 1.7680965443138699, + "learning_rate": 3.968952549965445e-06, + "loss": 1.0272, + "step": 937 + }, + { + "epoch": 0.08459214501510574, + "grad_norm": 18.141460231902588, + "learning_rate": 3.968849926658532e-06, + "loss": 1.0737, + "step": 938 + }, + { + "epoch": 0.084682328538576, + "grad_norm": 11.066343167950281, + "learning_rate": 3.9687471353584866e-06, + "loss": 0.9515, + "step": 939 + }, + { + "epoch": 0.08477251206204627, + "grad_norm": 1.4842666586438544, + "learning_rate": 3.9686441760740795e-06, + "loss": 1.0578, + "step": 940 + }, + { + "epoch": 0.08486269558551653, + "grad_norm": 5.879185261236246, + "learning_rate": 3.968541048814098e-06, + "loss": 0.9831, + "step": 941 + }, + { + "epoch": 0.08495287910898679, + "grad_norm": 3.813095546546759, + "learning_rate": 3.968437753587339e-06, + "loss": 0.9708, + "step": 942 + }, + { + "epoch": 0.08504306263245705, + "grad_norm": 1.67431640625, + "learning_rate": 3.968334290402616e-06, + "loss": 1.0875, + "step": 943 + }, + { + "epoch": 0.08513324615592731, + "grad_norm": 1.5518369372428469, + "learning_rate": 3.968230659268759e-06, + "loss": 0.9397, + "step": 944 + }, + { + "epoch": 0.08522342967939757, + "grad_norm": 2.451509952407481, + "learning_rate": 3.968126860194609e-06, + "loss": 1.114, + "step": 945 + }, + { + "epoch": 0.08531361320286783, + "grad_norm": 2.0207271613581668, + "learning_rate": 3.968022893189025e-06, + "loss": 1.0944, + "step": 946 + }, + { + "epoch": 0.0854037967263381, + "grad_norm": 1.809049149393309, + "learning_rate": 3.967918758260874e-06, + "loss": 1.0811, + "step": 947 + }, + { + "epoch": 0.08549398024980837, + "grad_norm": 1.7160122654927776, + "learning_rate": 3.967814455419044e-06, + "loss": 1.096, + "step": 948 + }, + { + "epoch": 0.08558416377327863, + "grad_norm": 2.313346604722328, + "learning_rate": 3.967709984672434e-06, + "loss": 1.0905, + "step": 949 + }, + { + "epoch": 0.08567434729674889, + "grad_norm": 1.7237810887001441, + "learning_rate": 3.967605346029959e-06, + "loss": 1.074, + "step": 950 + }, + { + "epoch": 0.08576453082021915, + "grad_norm": 1.8348472442242845, + "learning_rate": 3.9675005395005466e-06, + "loss": 1.1221, + "step": 951 + }, + { + "epoch": 0.08585471434368941, + "grad_norm": 1.9473588590197268, + "learning_rate": 3.967395565093139e-06, + "loss": 1.0693, + "step": 952 + }, + { + "epoch": 0.08594489786715967, + "grad_norm": 1.745455290196361, + "learning_rate": 3.967290422816693e-06, + "loss": 1.0495, + "step": 953 + }, + { + "epoch": 0.08603508139062993, + "grad_norm": 3.5931017747155947, + "learning_rate": 3.967185112680183e-06, + "loss": 1.011, + "step": 954 + }, + { + "epoch": 0.08612526491410019, + "grad_norm": 1.8490111724037332, + "learning_rate": 3.96707963469259e-06, + "loss": 1.1272, + "step": 955 + }, + { + "epoch": 0.08621544843757045, + "grad_norm": 2.7901992831960065, + "learning_rate": 3.966973988862917e-06, + "loss": 1.0388, + "step": 956 + }, + { + "epoch": 0.08630563196104071, + "grad_norm": 2.064736396461141, + "learning_rate": 3.966868175200178e-06, + "loss": 1.1025, + "step": 957 + }, + { + "epoch": 0.08639581548451097, + "grad_norm": 1.9522851587901549, + "learning_rate": 3.9667621937134e-06, + "loss": 1.1048, + "step": 958 + }, + { + "epoch": 0.08648599900798125, + "grad_norm": 2.5062010153837417, + "learning_rate": 3.966656044411627e-06, + "loss": 0.9591, + "step": 959 + }, + { + "epoch": 0.08657618253145151, + "grad_norm": 1.4759794830095494, + "learning_rate": 3.966549727303918e-06, + "loss": 1.0213, + "step": 960 + }, + { + "epoch": 0.08666636605492177, + "grad_norm": 2.122786322558216, + "learning_rate": 3.966443242399341e-06, + "loss": 1.0803, + "step": 961 + }, + { + "epoch": 0.08675654957839203, + "grad_norm": 0.7047257956441497, + "learning_rate": 3.966336589706985e-06, + "loss": 0.8371, + "step": 962 + }, + { + "epoch": 0.08684673310186229, + "grad_norm": 1.7625879144047991, + "learning_rate": 3.966229769235948e-06, + "loss": 1.0592, + "step": 963 + }, + { + "epoch": 0.08693691662533255, + "grad_norm": 2.4430808709805083, + "learning_rate": 3.966122780995345e-06, + "loss": 1.126, + "step": 964 + }, + { + "epoch": 0.08702710014880281, + "grad_norm": 2.1560313901792414, + "learning_rate": 3.966015624994306e-06, + "loss": 1.0811, + "step": 965 + }, + { + "epoch": 0.08711728367227307, + "grad_norm": 1.7571699536346566, + "learning_rate": 3.9659083012419735e-06, + "loss": 1.0189, + "step": 966 + }, + { + "epoch": 0.08720746719574334, + "grad_norm": 2.5867385775936143, + "learning_rate": 3.965800809747505e-06, + "loss": 0.9814, + "step": 967 + }, + { + "epoch": 0.0872976507192136, + "grad_norm": 1.941700105283677, + "learning_rate": 3.965693150520071e-06, + "loss": 1.0097, + "step": 968 + }, + { + "epoch": 0.08738783424268386, + "grad_norm": 2.3329613139267984, + "learning_rate": 3.96558532356886e-06, + "loss": 0.982, + "step": 969 + }, + { + "epoch": 0.08747801776615412, + "grad_norm": 2.0027525557036348, + "learning_rate": 3.9654773289030704e-06, + "loss": 1.095, + "step": 970 + }, + { + "epoch": 0.08756820128962439, + "grad_norm": 2.684170989876216, + "learning_rate": 3.9653691665319176e-06, + "loss": 1.0628, + "step": 971 + }, + { + "epoch": 0.08765838481309465, + "grad_norm": 2.0640205789101995, + "learning_rate": 3.96526083646463e-06, + "loss": 0.9769, + "step": 972 + }, + { + "epoch": 0.08774856833656491, + "grad_norm": 2.041468344677266, + "learning_rate": 3.9651523387104526e-06, + "loss": 0.9982, + "step": 973 + }, + { + "epoch": 0.08783875186003517, + "grad_norm": 2.6141021305963226, + "learning_rate": 3.965043673278641e-06, + "loss": 1.0558, + "step": 974 + }, + { + "epoch": 0.08792893538350544, + "grad_norm": 1.553791497214213, + "learning_rate": 3.964934840178469e-06, + "loss": 1.0542, + "step": 975 + }, + { + "epoch": 0.0880191189069757, + "grad_norm": 2.717617018074381, + "learning_rate": 3.964825839419221e-06, + "loss": 1.0954, + "step": 976 + }, + { + "epoch": 0.08810930243044596, + "grad_norm": 2.372095540565463, + "learning_rate": 3.964716671010199e-06, + "loss": 1.0351, + "step": 977 + }, + { + "epoch": 0.08819948595391622, + "grad_norm": 1.8549580403232477, + "learning_rate": 3.9646073349607165e-06, + "loss": 1.0105, + "step": 978 + }, + { + "epoch": 0.08828966947738648, + "grad_norm": 2.200021275504148, + "learning_rate": 3.964497831280105e-06, + "loss": 1.1818, + "step": 979 + }, + { + "epoch": 0.08837985300085674, + "grad_norm": 2.0653249871881805, + "learning_rate": 3.964388159977705e-06, + "loss": 1.0458, + "step": 980 + }, + { + "epoch": 0.088470036524327, + "grad_norm": 1.9655105945667466, + "learning_rate": 3.964278321062876e-06, + "loss": 0.954, + "step": 981 + }, + { + "epoch": 0.08856022004779726, + "grad_norm": 2.673414266586792, + "learning_rate": 3.96416831454499e-06, + "loss": 1.0717, + "step": 982 + }, + { + "epoch": 0.08865040357126754, + "grad_norm": 0.7021583906590387, + "learning_rate": 3.964058140433434e-06, + "loss": 0.8676, + "step": 983 + }, + { + "epoch": 0.0887405870947378, + "grad_norm": 1.7864652880059675, + "learning_rate": 3.963947798737606e-06, + "loss": 1.1086, + "step": 984 + }, + { + "epoch": 0.08883077061820806, + "grad_norm": 1.741246332738428, + "learning_rate": 3.963837289466923e-06, + "loss": 1.0671, + "step": 985 + }, + { + "epoch": 0.08892095414167832, + "grad_norm": 2.1377399243003707, + "learning_rate": 3.9637266126308145e-06, + "loss": 1.0181, + "step": 986 + }, + { + "epoch": 0.08901113766514858, + "grad_norm": 0.7099646315025976, + "learning_rate": 3.963615768238724e-06, + "loss": 0.8576, + "step": 987 + }, + { + "epoch": 0.08910132118861884, + "grad_norm": 1.9451708761237267, + "learning_rate": 3.963504756300107e-06, + "loss": 0.9354, + "step": 988 + }, + { + "epoch": 0.0891915047120891, + "grad_norm": 1.6232934941149393, + "learning_rate": 3.96339357682444e-06, + "loss": 1.0874, + "step": 989 + }, + { + "epoch": 0.08928168823555936, + "grad_norm": 1.6782799079192843, + "learning_rate": 3.963282229821206e-06, + "loss": 1.0666, + "step": 990 + }, + { + "epoch": 0.08937187175902962, + "grad_norm": 2.8007800582393734, + "learning_rate": 3.963170715299906e-06, + "loss": 1.0587, + "step": 991 + }, + { + "epoch": 0.08946205528249988, + "grad_norm": 1.6228124860135111, + "learning_rate": 3.963059033270056e-06, + "loss": 0.9732, + "step": 992 + }, + { + "epoch": 0.08955223880597014, + "grad_norm": 1.5845490941211944, + "learning_rate": 3.9629471837411855e-06, + "loss": 0.9977, + "step": 993 + }, + { + "epoch": 0.0896424223294404, + "grad_norm": 2.008235664155999, + "learning_rate": 3.962835166722838e-06, + "loss": 0.9962, + "step": 994 + }, + { + "epoch": 0.08973260585291068, + "grad_norm": 1.8834843366846072, + "learning_rate": 3.96272298222457e-06, + "loss": 1.0555, + "step": 995 + }, + { + "epoch": 0.08982278937638094, + "grad_norm": 1.6452898101836098, + "learning_rate": 3.962610630255956e-06, + "loss": 1.1738, + "step": 996 + }, + { + "epoch": 0.0899129728998512, + "grad_norm": 4.605746221232183, + "learning_rate": 3.96249811082658e-06, + "loss": 1.019, + "step": 997 + }, + { + "epoch": 0.09000315642332146, + "grad_norm": 0.8043301724122667, + "learning_rate": 3.962385423946046e-06, + "loss": 0.9114, + "step": 998 + }, + { + "epoch": 0.09009333994679172, + "grad_norm": 1.5027919852446403, + "learning_rate": 3.962272569623966e-06, + "loss": 1.0236, + "step": 999 + }, + { + "epoch": 0.09018352347026198, + "grad_norm": 2.09435912424216, + "learning_rate": 3.9621595478699704e-06, + "loss": 0.9429, + "step": 1000 + }, + { + "epoch": 0.09027370699373224, + "grad_norm": 2.1787491285942817, + "learning_rate": 3.962046358693703e-06, + "loss": 1.0408, + "step": 1001 + }, + { + "epoch": 0.0903638905172025, + "grad_norm": 1.825207264770757, + "learning_rate": 3.961933002104822e-06, + "loss": 1.1277, + "step": 1002 + }, + { + "epoch": 0.09045407404067277, + "grad_norm": 1.806687934874644, + "learning_rate": 3.961819478112999e-06, + "loss": 0.9754, + "step": 1003 + }, + { + "epoch": 0.09054425756414303, + "grad_norm": 1.8714806907201915, + "learning_rate": 3.961705786727921e-06, + "loss": 1.0716, + "step": 1004 + }, + { + "epoch": 0.09063444108761329, + "grad_norm": 1.6878983769131757, + "learning_rate": 3.961591927959288e-06, + "loss": 0.9719, + "step": 1005 + }, + { + "epoch": 0.09072462461108355, + "grad_norm": 0.7137456078068531, + "learning_rate": 3.961477901816816e-06, + "loss": 0.8318, + "step": 1006 + }, + { + "epoch": 0.09081480813455382, + "grad_norm": 1.702492622774024, + "learning_rate": 3.961363708310233e-06, + "loss": 1.1095, + "step": 1007 + }, + { + "epoch": 0.09090499165802408, + "grad_norm": 3.5466537070455857, + "learning_rate": 3.961249347449286e-06, + "loss": 1.0705, + "step": 1008 + }, + { + "epoch": 0.09099517518149434, + "grad_norm": 1.7417119493332052, + "learning_rate": 3.961134819243728e-06, + "loss": 1.0329, + "step": 1009 + }, + { + "epoch": 0.0910853587049646, + "grad_norm": 2.8435073581343446, + "learning_rate": 3.961020123703335e-06, + "loss": 1.055, + "step": 1010 + }, + { + "epoch": 0.09117554222843487, + "grad_norm": 1.950396653939022, + "learning_rate": 3.960905260837892e-06, + "loss": 1.0584, + "step": 1011 + }, + { + "epoch": 0.09126572575190513, + "grad_norm": 1.8391409987660776, + "learning_rate": 3.960790230657199e-06, + "loss": 1.0175, + "step": 1012 + }, + { + "epoch": 0.09135590927537539, + "grad_norm": 1.319747081426019, + "learning_rate": 3.960675033171072e-06, + "loss": 0.9963, + "step": 1013 + }, + { + "epoch": 0.09144609279884565, + "grad_norm": 3.1662463528280416, + "learning_rate": 3.960559668389341e-06, + "loss": 0.9922, + "step": 1014 + }, + { + "epoch": 0.09153627632231591, + "grad_norm": 9.34848137420759, + "learning_rate": 3.960444136321847e-06, + "loss": 1.0533, + "step": 1015 + }, + { + "epoch": 0.09162645984578617, + "grad_norm": 2.2389832040038353, + "learning_rate": 3.960328436978451e-06, + "loss": 1.1316, + "step": 1016 + }, + { + "epoch": 0.09171664336925643, + "grad_norm": 0.6556968628488775, + "learning_rate": 3.960212570369024e-06, + "loss": 0.8328, + "step": 1017 + }, + { + "epoch": 0.09180682689272669, + "grad_norm": 2.345871232462352, + "learning_rate": 3.9600965365034515e-06, + "loss": 0.9819, + "step": 1018 + }, + { + "epoch": 0.09189701041619697, + "grad_norm": 2.608438929247439, + "learning_rate": 3.959980335391634e-06, + "loss": 1.0987, + "step": 1019 + }, + { + "epoch": 0.09198719393966723, + "grad_norm": 1.881271808282414, + "learning_rate": 3.959863967043487e-06, + "loss": 0.9869, + "step": 1020 + }, + { + "epoch": 0.09207737746313749, + "grad_norm": 1.735982201770658, + "learning_rate": 3.9597474314689405e-06, + "loss": 1.0769, + "step": 1021 + }, + { + "epoch": 0.09216756098660775, + "grad_norm": 2.5259859898254935, + "learning_rate": 3.959630728677937e-06, + "loss": 1.0665, + "step": 1022 + }, + { + "epoch": 0.09225774451007801, + "grad_norm": 1.8333620878334373, + "learning_rate": 3.959513858680434e-06, + "loss": 1.0266, + "step": 1023 + }, + { + "epoch": 0.09234792803354827, + "grad_norm": 1.8351506921286966, + "learning_rate": 3.959396821486405e-06, + "loss": 1.1372, + "step": 1024 + }, + { + "epoch": 0.09243811155701853, + "grad_norm": 0.8091378707816768, + "learning_rate": 3.959279617105835e-06, + "loss": 0.8633, + "step": 1025 + }, + { + "epoch": 0.09252829508048879, + "grad_norm": 1.948131311459639, + "learning_rate": 3.9591622455487235e-06, + "loss": 1.0557, + "step": 1026 + }, + { + "epoch": 0.09261847860395905, + "grad_norm": 4.837896735903993, + "learning_rate": 3.959044706825087e-06, + "loss": 1.1193, + "step": 1027 + }, + { + "epoch": 0.09270866212742931, + "grad_norm": 2.271099343307741, + "learning_rate": 3.958927000944954e-06, + "loss": 0.9533, + "step": 1028 + }, + { + "epoch": 0.09279884565089958, + "grad_norm": 5.596163777668525, + "learning_rate": 3.958809127918368e-06, + "loss": 1.1293, + "step": 1029 + }, + { + "epoch": 0.09288902917436984, + "grad_norm": 9.923009900644121, + "learning_rate": 3.958691087755387e-06, + "loss": 0.9824, + "step": 1030 + }, + { + "epoch": 0.09297921269784011, + "grad_norm": 1.7208725048098912, + "learning_rate": 3.958572880466081e-06, + "loss": 0.9204, + "step": 1031 + }, + { + "epoch": 0.09306939622131037, + "grad_norm": 1.564647419612155, + "learning_rate": 3.9584545060605385e-06, + "loss": 1.031, + "step": 1032 + }, + { + "epoch": 0.09315957974478063, + "grad_norm": 1.59287312638293, + "learning_rate": 3.958335964548859e-06, + "loss": 1.0129, + "step": 1033 + }, + { + "epoch": 0.0932497632682509, + "grad_norm": 1.9595437130852227, + "learning_rate": 3.958217255941156e-06, + "loss": 1.092, + "step": 1034 + }, + { + "epoch": 0.09333994679172115, + "grad_norm": 1.4565167784374304, + "learning_rate": 3.95809838024756e-06, + "loss": 0.9948, + "step": 1035 + }, + { + "epoch": 0.09343013031519141, + "grad_norm": 1.862083898057537, + "learning_rate": 3.957979337478212e-06, + "loss": 1.0118, + "step": 1036 + }, + { + "epoch": 0.09352031383866168, + "grad_norm": 2.2139911389746465, + "learning_rate": 3.957860127643272e-06, + "loss": 1.0704, + "step": 1037 + }, + { + "epoch": 0.09361049736213194, + "grad_norm": 2.3593637832476024, + "learning_rate": 3.95774075075291e-06, + "loss": 1.0523, + "step": 1038 + }, + { + "epoch": 0.0937006808856022, + "grad_norm": 1.8109988211254577, + "learning_rate": 3.957621206817312e-06, + "loss": 1.1038, + "step": 1039 + }, + { + "epoch": 0.09379086440907246, + "grad_norm": 1.7881334716212192, + "learning_rate": 3.957501495846679e-06, + "loss": 1.0284, + "step": 1040 + }, + { + "epoch": 0.09388104793254272, + "grad_norm": 0.6770369929570362, + "learning_rate": 3.957381617851225e-06, + "loss": 0.838, + "step": 1041 + }, + { + "epoch": 0.09397123145601298, + "grad_norm": 1.6345385321938173, + "learning_rate": 3.9572615728411776e-06, + "loss": 0.9671, + "step": 1042 + }, + { + "epoch": 0.09406141497948325, + "grad_norm": 3.058984100649649, + "learning_rate": 3.957141360826781e-06, + "loss": 0.9282, + "step": 1043 + }, + { + "epoch": 0.09415159850295352, + "grad_norm": 1.8065099720493571, + "learning_rate": 3.957020981818292e-06, + "loss": 1.0316, + "step": 1044 + }, + { + "epoch": 0.09424178202642378, + "grad_norm": 2.7209205294117407, + "learning_rate": 3.956900435825982e-06, + "loss": 1.0405, + "step": 1045 + }, + { + "epoch": 0.09433196554989404, + "grad_norm": 1.754387600442057, + "learning_rate": 3.9567797228601364e-06, + "loss": 0.9932, + "step": 1046 + }, + { + "epoch": 0.0944221490733643, + "grad_norm": 1.6041928161207228, + "learning_rate": 3.956658842931055e-06, + "loss": 1.1194, + "step": 1047 + }, + { + "epoch": 0.09451233259683456, + "grad_norm": 1.6789929750532968, + "learning_rate": 3.956537796049052e-06, + "loss": 1.0591, + "step": 1048 + }, + { + "epoch": 0.09460251612030482, + "grad_norm": 0.7285311699811857, + "learning_rate": 3.956416582224457e-06, + "loss": 0.8571, + "step": 1049 + }, + { + "epoch": 0.09469269964377508, + "grad_norm": 3.508849672451749, + "learning_rate": 3.956295201467611e-06, + "loss": 1.0827, + "step": 1050 + }, + { + "epoch": 0.09478288316724534, + "grad_norm": 1.4136108182869862, + "learning_rate": 3.956173653788872e-06, + "loss": 1.0754, + "step": 1051 + }, + { + "epoch": 0.0948730666907156, + "grad_norm": 2.3521797662726693, + "learning_rate": 3.95605193919861e-06, + "loss": 1.0813, + "step": 1052 + }, + { + "epoch": 0.09496325021418586, + "grad_norm": 1.8531307870089406, + "learning_rate": 3.955930057707211e-06, + "loss": 1.0336, + "step": 1053 + }, + { + "epoch": 0.09505343373765612, + "grad_norm": 1.70092652541066, + "learning_rate": 3.955808009325075e-06, + "loss": 1.0563, + "step": 1054 + }, + { + "epoch": 0.0951436172611264, + "grad_norm": 2.2983197709149747, + "learning_rate": 3.955685794062615e-06, + "loss": 1.0182, + "step": 1055 + }, + { + "epoch": 0.09523380078459666, + "grad_norm": 2.173172926551191, + "learning_rate": 3.95556341193026e-06, + "loss": 1.148, + "step": 1056 + }, + { + "epoch": 0.09532398430806692, + "grad_norm": 1.6008238787208864, + "learning_rate": 3.955440862938452e-06, + "loss": 1.142, + "step": 1057 + }, + { + "epoch": 0.09541416783153718, + "grad_norm": 2.4615738303366985, + "learning_rate": 3.955318147097647e-06, + "loss": 1.0484, + "step": 1058 + }, + { + "epoch": 0.09550435135500744, + "grad_norm": 1.8932368625097913, + "learning_rate": 3.955195264418316e-06, + "loss": 1.0096, + "step": 1059 + }, + { + "epoch": 0.0955945348784777, + "grad_norm": 1.515185026824145, + "learning_rate": 3.955072214910944e-06, + "loss": 0.9024, + "step": 1060 + }, + { + "epoch": 0.09568471840194796, + "grad_norm": 1.9981197220843272, + "learning_rate": 3.954948998586032e-06, + "loss": 0.9669, + "step": 1061 + }, + { + "epoch": 0.09577490192541822, + "grad_norm": 1.9926067434744907, + "learning_rate": 3.954825615454089e-06, + "loss": 1.0545, + "step": 1062 + }, + { + "epoch": 0.09586508544888848, + "grad_norm": 1.9724815971183252, + "learning_rate": 3.954702065525649e-06, + "loss": 1.0439, + "step": 1063 + }, + { + "epoch": 0.09595526897235875, + "grad_norm": 2.1818714207598293, + "learning_rate": 3.954578348811248e-06, + "loss": 1.0752, + "step": 1064 + }, + { + "epoch": 0.096045452495829, + "grad_norm": 3.0054462587701463, + "learning_rate": 3.954454465321447e-06, + "loss": 1.0276, + "step": 1065 + }, + { + "epoch": 0.09613563601929928, + "grad_norm": 2.201923123693802, + "learning_rate": 3.954330415066813e-06, + "loss": 1.1274, + "step": 1066 + }, + { + "epoch": 0.09622581954276954, + "grad_norm": 1.8853888395444285, + "learning_rate": 3.954206198057932e-06, + "loss": 1.0313, + "step": 1067 + }, + { + "epoch": 0.0963160030662398, + "grad_norm": 1.7974069222663007, + "learning_rate": 3.954081814305403e-06, + "loss": 1.1041, + "step": 1068 + }, + { + "epoch": 0.09640618658971006, + "grad_norm": 2.234778841243312, + "learning_rate": 3.953957263819839e-06, + "loss": 0.984, + "step": 1069 + }, + { + "epoch": 0.09649637011318032, + "grad_norm": 1.7896422400529084, + "learning_rate": 3.953832546611867e-06, + "loss": 0.9798, + "step": 1070 + }, + { + "epoch": 0.09658655363665059, + "grad_norm": 2.0842326194255385, + "learning_rate": 3.953707662692129e-06, + "loss": 1.0733, + "step": 1071 + }, + { + "epoch": 0.09667673716012085, + "grad_norm": 2.880641378139716, + "learning_rate": 3.95358261207128e-06, + "loss": 0.9773, + "step": 1072 + }, + { + "epoch": 0.0967669206835911, + "grad_norm": 1.9837862478020114, + "learning_rate": 3.953457394759992e-06, + "loss": 1.0394, + "step": 1073 + }, + { + "epoch": 0.09685710420706137, + "grad_norm": 1.9096741487251383, + "learning_rate": 3.953332010768947e-06, + "loss": 1.0644, + "step": 1074 + }, + { + "epoch": 0.09694728773053163, + "grad_norm": 2.000290611134217, + "learning_rate": 3.9532064601088436e-06, + "loss": 1.0111, + "step": 1075 + }, + { + "epoch": 0.09703747125400189, + "grad_norm": 2.7006072738973264, + "learning_rate": 3.953080742790396e-06, + "loss": 1.1947, + "step": 1076 + }, + { + "epoch": 0.09712765477747215, + "grad_norm": 0.7659378191209157, + "learning_rate": 3.95295485882433e-06, + "loss": 0.8984, + "step": 1077 + }, + { + "epoch": 0.09721783830094242, + "grad_norm": 0.8193155679455442, + "learning_rate": 3.952828808221387e-06, + "loss": 0.8663, + "step": 1078 + }, + { + "epoch": 0.09730802182441269, + "grad_norm": 6.559749935315155, + "learning_rate": 3.9527025909923225e-06, + "loss": 1.025, + "step": 1079 + }, + { + "epoch": 0.09739820534788295, + "grad_norm": 13.387704679917304, + "learning_rate": 3.952576207147906e-06, + "loss": 1.1341, + "step": 1080 + }, + { + "epoch": 0.09748838887135321, + "grad_norm": 3.476030287563472, + "learning_rate": 3.95244965669892e-06, + "loss": 1.0815, + "step": 1081 + }, + { + "epoch": 0.09757857239482347, + "grad_norm": 2.461966260200071, + "learning_rate": 3.952322939656165e-06, + "loss": 1.0037, + "step": 1082 + }, + { + "epoch": 0.09766875591829373, + "grad_norm": 1.5357455998139868, + "learning_rate": 3.952196056030451e-06, + "loss": 1.0399, + "step": 1083 + }, + { + "epoch": 0.09775893944176399, + "grad_norm": 1.8717048300195156, + "learning_rate": 3.952069005832605e-06, + "loss": 1.0128, + "step": 1084 + }, + { + "epoch": 0.09784912296523425, + "grad_norm": 1.6481300537355932, + "learning_rate": 3.951941789073468e-06, + "loss": 1.0718, + "step": 1085 + }, + { + "epoch": 0.09793930648870451, + "grad_norm": 2.259671933829147, + "learning_rate": 3.9518144057638955e-06, + "loss": 1.0693, + "step": 1086 + }, + { + "epoch": 0.09802949001217477, + "grad_norm": 1.696482663260239, + "learning_rate": 3.951686855914755e-06, + "loss": 1.041, + "step": 1087 + }, + { + "epoch": 0.09811967353564503, + "grad_norm": 2.169410240087135, + "learning_rate": 3.9515591395369305e-06, + "loss": 1.0739, + "step": 1088 + }, + { + "epoch": 0.0982098570591153, + "grad_norm": 1.729288357355456, + "learning_rate": 3.95143125664132e-06, + "loss": 1.0312, + "step": 1089 + }, + { + "epoch": 0.09830004058258557, + "grad_norm": 1.828153039440083, + "learning_rate": 3.951303207238833e-06, + "loss": 1.0702, + "step": 1090 + }, + { + "epoch": 0.09839022410605583, + "grad_norm": 1.6014537495232832, + "learning_rate": 3.951174991340399e-06, + "loss": 1.0161, + "step": 1091 + }, + { + "epoch": 0.09848040762952609, + "grad_norm": 1.6401201470489841, + "learning_rate": 3.9510466089569546e-06, + "loss": 1.0098, + "step": 1092 + }, + { + "epoch": 0.09857059115299635, + "grad_norm": 1.8015426383201556, + "learning_rate": 3.950918060099456e-06, + "loss": 1.0081, + "step": 1093 + }, + { + "epoch": 0.09866077467646661, + "grad_norm": 1.7846825302928515, + "learning_rate": 3.950789344778871e-06, + "loss": 1.0993, + "step": 1094 + }, + { + "epoch": 0.09875095819993687, + "grad_norm": 1.7475196426881188, + "learning_rate": 3.950660463006184e-06, + "loss": 0.9869, + "step": 1095 + }, + { + "epoch": 0.09884114172340713, + "grad_norm": 1.6157044322927747, + "learning_rate": 3.950531414792389e-06, + "loss": 1.1116, + "step": 1096 + }, + { + "epoch": 0.0989313252468774, + "grad_norm": 1.8192445954240832, + "learning_rate": 3.950402200148498e-06, + "loss": 1.0498, + "step": 1097 + }, + { + "epoch": 0.09902150877034765, + "grad_norm": 1.5011245168163707, + "learning_rate": 3.950272819085538e-06, + "loss": 1.0207, + "step": 1098 + }, + { + "epoch": 0.09911169229381792, + "grad_norm": 3.2975998212789297, + "learning_rate": 3.9501432716145474e-06, + "loss": 1.1281, + "step": 1099 + }, + { + "epoch": 0.09920187581728818, + "grad_norm": 1.8484444418975932, + "learning_rate": 3.950013557746579e-06, + "loss": 0.9403, + "step": 1100 + }, + { + "epoch": 0.09929205934075844, + "grad_norm": 1.919616550759571, + "learning_rate": 3.949883677492703e-06, + "loss": 1.0813, + "step": 1101 + }, + { + "epoch": 0.09938224286422871, + "grad_norm": 1.9878893992592077, + "learning_rate": 3.9497536308639994e-06, + "loss": 1.0615, + "step": 1102 + }, + { + "epoch": 0.09947242638769897, + "grad_norm": 2.4338128984706207, + "learning_rate": 3.949623417871565e-06, + "loss": 1.0966, + "step": 1103 + }, + { + "epoch": 0.09956260991116923, + "grad_norm": 1.6471587369170133, + "learning_rate": 3.949493038526511e-06, + "loss": 1.0246, + "step": 1104 + }, + { + "epoch": 0.0996527934346395, + "grad_norm": 1.8337075834042695, + "learning_rate": 3.949362492839961e-06, + "loss": 0.9962, + "step": 1105 + }, + { + "epoch": 0.09974297695810976, + "grad_norm": 1.546339616699566, + "learning_rate": 3.949231780823054e-06, + "loss": 1.0054, + "step": 1106 + }, + { + "epoch": 0.09983316048158002, + "grad_norm": 1.8581909127963812, + "learning_rate": 3.949100902486945e-06, + "loss": 0.9875, + "step": 1107 + }, + { + "epoch": 0.09992334400505028, + "grad_norm": 1.831685893808562, + "learning_rate": 3.948969857842799e-06, + "loss": 1.0092, + "step": 1108 + }, + { + "epoch": 0.10001352752852054, + "grad_norm": 1.9232425941117424, + "learning_rate": 3.948838646901798e-06, + "loss": 1.0337, + "step": 1109 + }, + { + "epoch": 0.1001037110519908, + "grad_norm": 2.0513052715919056, + "learning_rate": 3.948707269675138e-06, + "loss": 1.0335, + "step": 1110 + }, + { + "epoch": 0.10019389457546106, + "grad_norm": 3.991788422407322, + "learning_rate": 3.948575726174028e-06, + "loss": 1.0301, + "step": 1111 + }, + { + "epoch": 0.10028407809893132, + "grad_norm": 2.3909087168564356, + "learning_rate": 3.9484440164096935e-06, + "loss": 1.1238, + "step": 1112 + }, + { + "epoch": 0.10037426162240158, + "grad_norm": 2.23267123480886, + "learning_rate": 3.948312140393372e-06, + "loss": 1.0702, + "step": 1113 + }, + { + "epoch": 0.10046444514587186, + "grad_norm": 1.9713335191526422, + "learning_rate": 3.948180098136316e-06, + "loss": 1.0138, + "step": 1114 + }, + { + "epoch": 0.10055462866934212, + "grad_norm": 2.3231809206336766, + "learning_rate": 3.948047889649791e-06, + "loss": 1.0132, + "step": 1115 + }, + { + "epoch": 0.10064481219281238, + "grad_norm": 1.7568030277411977, + "learning_rate": 3.947915514945079e-06, + "loss": 1.0235, + "step": 1116 + }, + { + "epoch": 0.10073499571628264, + "grad_norm": 1.756019932892268, + "learning_rate": 3.947782974033474e-06, + "loss": 1.1046, + "step": 1117 + }, + { + "epoch": 0.1008251792397529, + "grad_norm": 4.483364509748288, + "learning_rate": 3.9476502669262866e-06, + "loss": 1.0024, + "step": 1118 + }, + { + "epoch": 0.10091536276322316, + "grad_norm": 1.6243452807265693, + "learning_rate": 3.947517393634839e-06, + "loss": 1.013, + "step": 1119 + }, + { + "epoch": 0.10100554628669342, + "grad_norm": 1.5984304179969733, + "learning_rate": 3.947384354170469e-06, + "loss": 1.079, + "step": 1120 + }, + { + "epoch": 0.10109572981016368, + "grad_norm": 2.030480811394053, + "learning_rate": 3.947251148544528e-06, + "loss": 1.0762, + "step": 1121 + }, + { + "epoch": 0.10118591333363394, + "grad_norm": 1.7537920602443642, + "learning_rate": 3.947117776768382e-06, + "loss": 1.0688, + "step": 1122 + }, + { + "epoch": 0.1012760968571042, + "grad_norm": 1.5009205695425891, + "learning_rate": 3.9469842388534105e-06, + "loss": 1.0213, + "step": 1123 + }, + { + "epoch": 0.10136628038057446, + "grad_norm": 2.1743662426041106, + "learning_rate": 3.946850534811009e-06, + "loss": 1.054, + "step": 1124 + }, + { + "epoch": 0.10145646390404472, + "grad_norm": 1.6119655018988879, + "learning_rate": 3.946716664652585e-06, + "loss": 1.0281, + "step": 1125 + }, + { + "epoch": 0.101546647427515, + "grad_norm": 3.206304247064469, + "learning_rate": 3.94658262838956e-06, + "loss": 0.951, + "step": 1126 + }, + { + "epoch": 0.10163683095098526, + "grad_norm": 1.6050817761132496, + "learning_rate": 3.946448426033373e-06, + "loss": 0.9623, + "step": 1127 + }, + { + "epoch": 0.10172701447445552, + "grad_norm": 3.176850724173492, + "learning_rate": 3.946314057595473e-06, + "loss": 1.0786, + "step": 1128 + }, + { + "epoch": 0.10181719799792578, + "grad_norm": 2.8925374071344976, + "learning_rate": 3.946179523087326e-06, + "loss": 1.1074, + "step": 1129 + }, + { + "epoch": 0.10190738152139604, + "grad_norm": 1.6600813545835817, + "learning_rate": 3.9460448225204104e-06, + "loss": 0.9596, + "step": 1130 + }, + { + "epoch": 0.1019975650448663, + "grad_norm": 1.9966604842895337, + "learning_rate": 3.945909955906221e-06, + "loss": 1.0366, + "step": 1131 + }, + { + "epoch": 0.10208774856833656, + "grad_norm": 2.0643439864053095, + "learning_rate": 3.945774923256264e-06, + "loss": 1.0418, + "step": 1132 + }, + { + "epoch": 0.10217793209180683, + "grad_norm": 2.4068322406181033, + "learning_rate": 3.945639724582062e-06, + "loss": 1.1072, + "step": 1133 + }, + { + "epoch": 0.10226811561527709, + "grad_norm": 1.9106961288913653, + "learning_rate": 3.94550435989515e-06, + "loss": 1.0047, + "step": 1134 + }, + { + "epoch": 0.10235829913874735, + "grad_norm": 1.6860731945281968, + "learning_rate": 3.945368829207079e-06, + "loss": 1.0613, + "step": 1135 + }, + { + "epoch": 0.10244848266221761, + "grad_norm": 1.755529659149988, + "learning_rate": 3.945233132529414e-06, + "loss": 0.9986, + "step": 1136 + }, + { + "epoch": 0.10253866618568787, + "grad_norm": 1.6116489530271718, + "learning_rate": 3.9450972698737304e-06, + "loss": 1.0408, + "step": 1137 + }, + { + "epoch": 0.10262884970915814, + "grad_norm": 2.5178902419986366, + "learning_rate": 3.944961241251623e-06, + "loss": 1.0216, + "step": 1138 + }, + { + "epoch": 0.1027190332326284, + "grad_norm": 1.9594212479434048, + "learning_rate": 3.9448250466746985e-06, + "loss": 1.0114, + "step": 1139 + }, + { + "epoch": 0.10280921675609866, + "grad_norm": 1.887341382770268, + "learning_rate": 3.944688686154578e-06, + "loss": 1.0692, + "step": 1140 + }, + { + "epoch": 0.10289940027956893, + "grad_norm": 2.553545028519568, + "learning_rate": 3.944552159702894e-06, + "loss": 0.9994, + "step": 1141 + }, + { + "epoch": 0.10298958380303919, + "grad_norm": 1.7831773870977647, + "learning_rate": 3.944415467331299e-06, + "loss": 1.0334, + "step": 1142 + }, + { + "epoch": 0.10307976732650945, + "grad_norm": 1.5961481638710902, + "learning_rate": 3.944278609051455e-06, + "loss": 0.9969, + "step": 1143 + }, + { + "epoch": 0.10316995084997971, + "grad_norm": 1.428710854061643, + "learning_rate": 3.944141584875039e-06, + "loss": 0.977, + "step": 1144 + }, + { + "epoch": 0.10326013437344997, + "grad_norm": 1.8788889925792707, + "learning_rate": 3.944004394813743e-06, + "loss": 1.0452, + "step": 1145 + }, + { + "epoch": 0.10335031789692023, + "grad_norm": 7.366347846587468, + "learning_rate": 3.943867038879273e-06, + "loss": 1.0656, + "step": 1146 + }, + { + "epoch": 0.10344050142039049, + "grad_norm": 3.6462529113569593, + "learning_rate": 3.943729517083349e-06, + "loss": 1.0765, + "step": 1147 + }, + { + "epoch": 0.10353068494386075, + "grad_norm": 1.880341550663791, + "learning_rate": 3.943591829437705e-06, + "loss": 1.0712, + "step": 1148 + }, + { + "epoch": 0.10362086846733101, + "grad_norm": 2.193471224642845, + "learning_rate": 3.9434539759540895e-06, + "loss": 0.9853, + "step": 1149 + }, + { + "epoch": 0.10371105199080129, + "grad_norm": 1.7782135229081992, + "learning_rate": 3.943315956644264e-06, + "loss": 0.9798, + "step": 1150 + }, + { + "epoch": 0.10380123551427155, + "grad_norm": 2.2471804436424465, + "learning_rate": 3.943177771520006e-06, + "loss": 1.1087, + "step": 1151 + }, + { + "epoch": 0.10389141903774181, + "grad_norm": 2.0094291620385567, + "learning_rate": 3.9430394205931065e-06, + "loss": 1.0877, + "step": 1152 + }, + { + "epoch": 0.10398160256121207, + "grad_norm": 1.9110113620909346, + "learning_rate": 3.942900903875369e-06, + "loss": 0.9152, + "step": 1153 + }, + { + "epoch": 0.10407178608468233, + "grad_norm": 1.6390667008890696, + "learning_rate": 3.942762221378614e-06, + "loss": 1.0453, + "step": 1154 + }, + { + "epoch": 0.10416196960815259, + "grad_norm": 1.7995034274624815, + "learning_rate": 3.942623373114673e-06, + "loss": 1.0782, + "step": 1155 + }, + { + "epoch": 0.10425215313162285, + "grad_norm": 1.835851529320255, + "learning_rate": 3.942484359095396e-06, + "loss": 1.0277, + "step": 1156 + }, + { + "epoch": 0.10434233665509311, + "grad_norm": 1.8314492413451313, + "learning_rate": 3.942345179332642e-06, + "loss": 0.9741, + "step": 1157 + }, + { + "epoch": 0.10443252017856337, + "grad_norm": 1.4795626079053463, + "learning_rate": 3.942205833838287e-06, + "loss": 0.9927, + "step": 1158 + }, + { + "epoch": 0.10452270370203363, + "grad_norm": 2.2932469785681424, + "learning_rate": 3.9420663226242204e-06, + "loss": 1.1356, + "step": 1159 + }, + { + "epoch": 0.1046128872255039, + "grad_norm": 2.1135007846488785, + "learning_rate": 3.941926645702348e-06, + "loss": 1.0903, + "step": 1160 + }, + { + "epoch": 0.10470307074897416, + "grad_norm": 2.656022095720463, + "learning_rate": 3.941786803084586e-06, + "loss": 1.0254, + "step": 1161 + }, + { + "epoch": 0.10479325427244443, + "grad_norm": 1.9671614156604083, + "learning_rate": 3.941646794782867e-06, + "loss": 0.9903, + "step": 1162 + }, + { + "epoch": 0.10488343779591469, + "grad_norm": 3.264017539949599, + "learning_rate": 3.941506620809137e-06, + "loss": 1.0521, + "step": 1163 + }, + { + "epoch": 0.10497362131938495, + "grad_norm": 2.4628646812917863, + "learning_rate": 3.941366281175357e-06, + "loss": 1.0313, + "step": 1164 + }, + { + "epoch": 0.10506380484285521, + "grad_norm": 1.5454590607388472, + "learning_rate": 3.941225775893502e-06, + "loss": 0.9791, + "step": 1165 + }, + { + "epoch": 0.10515398836632547, + "grad_norm": 1.854935675891306, + "learning_rate": 3.941085104975559e-06, + "loss": 1.0132, + "step": 1166 + }, + { + "epoch": 0.10524417188979573, + "grad_norm": 0.7533363599702356, + "learning_rate": 3.9409442684335325e-06, + "loss": 0.8613, + "step": 1167 + }, + { + "epoch": 0.105334355413266, + "grad_norm": 1.9989831246698517, + "learning_rate": 3.940803266279438e-06, + "loss": 1.033, + "step": 1168 + }, + { + "epoch": 0.10542453893673626, + "grad_norm": 1.777535698555011, + "learning_rate": 3.9406620985253076e-06, + "loss": 0.9774, + "step": 1169 + }, + { + "epoch": 0.10551472246020652, + "grad_norm": 1.6734776168844496, + "learning_rate": 3.940520765183187e-06, + "loss": 1.0107, + "step": 1170 + }, + { + "epoch": 0.10560490598367678, + "grad_norm": 2.7019844074321733, + "learning_rate": 3.940379266265134e-06, + "loss": 1.0797, + "step": 1171 + }, + { + "epoch": 0.10569508950714704, + "grad_norm": 3.7668566450523437, + "learning_rate": 3.940237601783223e-06, + "loss": 1.0053, + "step": 1172 + }, + { + "epoch": 0.1057852730306173, + "grad_norm": 1.5773894607095125, + "learning_rate": 3.940095771749542e-06, + "loss": 1.0387, + "step": 1173 + }, + { + "epoch": 0.10587545655408757, + "grad_norm": 2.0966170310051684, + "learning_rate": 3.939953776176192e-06, + "loss": 1.0961, + "step": 1174 + }, + { + "epoch": 0.10596564007755783, + "grad_norm": 2.081679323380382, + "learning_rate": 3.939811615075288e-06, + "loss": 1.0653, + "step": 1175 + }, + { + "epoch": 0.1060558236010281, + "grad_norm": 2.1839862395680982, + "learning_rate": 3.9396692884589616e-06, + "loss": 1.1019, + "step": 1176 + }, + { + "epoch": 0.10614600712449836, + "grad_norm": 3.106018094591637, + "learning_rate": 3.9395267963393565e-06, + "loss": 1.0259, + "step": 1177 + }, + { + "epoch": 0.10623619064796862, + "grad_norm": 2.201301090269287, + "learning_rate": 3.939384138728631e-06, + "loss": 1.0753, + "step": 1178 + }, + { + "epoch": 0.10632637417143888, + "grad_norm": 2.312891952702393, + "learning_rate": 3.939241315638956e-06, + "loss": 1.026, + "step": 1179 + }, + { + "epoch": 0.10641655769490914, + "grad_norm": 1.7586265882885777, + "learning_rate": 3.93909832708252e-06, + "loss": 1.0275, + "step": 1180 + }, + { + "epoch": 0.1065067412183794, + "grad_norm": 1.780653000220142, + "learning_rate": 3.938955173071523e-06, + "loss": 1.0477, + "step": 1181 + }, + { + "epoch": 0.10659692474184966, + "grad_norm": 0.6020303368481053, + "learning_rate": 3.938811853618179e-06, + "loss": 0.8474, + "step": 1182 + }, + { + "epoch": 0.10668710826531992, + "grad_norm": 3.619131533945341, + "learning_rate": 3.938668368734717e-06, + "loss": 1.0764, + "step": 1183 + }, + { + "epoch": 0.10677729178879018, + "grad_norm": 4.349406339979504, + "learning_rate": 3.93852471843338e-06, + "loss": 0.9174, + "step": 1184 + }, + { + "epoch": 0.10686747531226044, + "grad_norm": 2.180670075807596, + "learning_rate": 3.9383809027264254e-06, + "loss": 1.0263, + "step": 1185 + }, + { + "epoch": 0.10695765883573072, + "grad_norm": 1.4635843381622873, + "learning_rate": 3.938236921626124e-06, + "loss": 1.0804, + "step": 1186 + }, + { + "epoch": 0.10704784235920098, + "grad_norm": 2.4904360940052195, + "learning_rate": 3.938092775144761e-06, + "loss": 1.0399, + "step": 1187 + }, + { + "epoch": 0.10713802588267124, + "grad_norm": 1.6743544259681566, + "learning_rate": 3.9379484632946355e-06, + "loss": 0.9636, + "step": 1188 + }, + { + "epoch": 0.1072282094061415, + "grad_norm": 2.0134792291229675, + "learning_rate": 3.937803986088062e-06, + "loss": 1.084, + "step": 1189 + }, + { + "epoch": 0.10731839292961176, + "grad_norm": 3.6700549220827083, + "learning_rate": 3.937659343537367e-06, + "loss": 0.9995, + "step": 1190 + }, + { + "epoch": 0.10740857645308202, + "grad_norm": 1.7901122499146, + "learning_rate": 3.937514535654893e-06, + "loss": 1.1091, + "step": 1191 + }, + { + "epoch": 0.10749875997655228, + "grad_norm": 4.748726322670329, + "learning_rate": 3.937369562452996e-06, + "loss": 1.0466, + "step": 1192 + }, + { + "epoch": 0.10758894350002254, + "grad_norm": 1.571808135959641, + "learning_rate": 3.937224423944044e-06, + "loss": 1.0964, + "step": 1193 + }, + { + "epoch": 0.1076791270234928, + "grad_norm": 1.785795201784366, + "learning_rate": 3.937079120140423e-06, + "loss": 0.9744, + "step": 1194 + }, + { + "epoch": 0.10776931054696307, + "grad_norm": 1.4737586794946254, + "learning_rate": 3.936933651054531e-06, + "loss": 1.0772, + "step": 1195 + }, + { + "epoch": 0.10785949407043333, + "grad_norm": 3.5048837648455797, + "learning_rate": 3.936788016698779e-06, + "loss": 1.1583, + "step": 1196 + }, + { + "epoch": 0.1079496775939036, + "grad_norm": 1.9208140352263057, + "learning_rate": 3.936642217085594e-06, + "loss": 1.0723, + "step": 1197 + }, + { + "epoch": 0.10803986111737386, + "grad_norm": 10.779172835454563, + "learning_rate": 3.936496252227417e-06, + "loss": 0.9478, + "step": 1198 + }, + { + "epoch": 0.10813004464084412, + "grad_norm": 1.790175112784496, + "learning_rate": 3.936350122136703e-06, + "loss": 1.0229, + "step": 1199 + }, + { + "epoch": 0.10822022816431438, + "grad_norm": 0.691414116421363, + "learning_rate": 3.936203826825919e-06, + "loss": 0.8686, + "step": 1200 + }, + { + "epoch": 0.10831041168778464, + "grad_norm": 1.6471949952298242, + "learning_rate": 3.9360573663075475e-06, + "loss": 0.9982, + "step": 1201 + }, + { + "epoch": 0.1084005952112549, + "grad_norm": 2.1621396993907207, + "learning_rate": 3.935910740594087e-06, + "loss": 1.0711, + "step": 1202 + }, + { + "epoch": 0.10849077873472517, + "grad_norm": 2.3158810616821213, + "learning_rate": 3.935763949698047e-06, + "loss": 1.0472, + "step": 1203 + }, + { + "epoch": 0.10858096225819543, + "grad_norm": 2.3002074065193754, + "learning_rate": 3.935616993631954e-06, + "loss": 1.1041, + "step": 1204 + }, + { + "epoch": 0.10867114578166569, + "grad_norm": 1.751223068629249, + "learning_rate": 3.935469872408345e-06, + "loss": 1.0336, + "step": 1205 + }, + { + "epoch": 0.10876132930513595, + "grad_norm": 2.660744219764281, + "learning_rate": 3.935322586039776e-06, + "loss": 1.067, + "step": 1206 + }, + { + "epoch": 0.10885151282860621, + "grad_norm": 2.206739234172887, + "learning_rate": 3.935175134538811e-06, + "loss": 0.9747, + "step": 1207 + }, + { + "epoch": 0.10894169635207647, + "grad_norm": 2.481415046948694, + "learning_rate": 3.935027517918034e-06, + "loss": 0.9773, + "step": 1208 + }, + { + "epoch": 0.10903187987554674, + "grad_norm": 2.487772699485019, + "learning_rate": 3.93487973619004e-06, + "loss": 1.0729, + "step": 1209 + }, + { + "epoch": 0.109122063399017, + "grad_norm": 2.228717604234289, + "learning_rate": 3.934731789367438e-06, + "loss": 1.0456, + "step": 1210 + }, + { + "epoch": 0.10921224692248727, + "grad_norm": 0.7734287916521667, + "learning_rate": 3.9345836774628505e-06, + "loss": 0.8464, + "step": 1211 + }, + { + "epoch": 0.10930243044595753, + "grad_norm": 1.9112536938408933, + "learning_rate": 3.934435400488917e-06, + "loss": 1.0299, + "step": 1212 + }, + { + "epoch": 0.10939261396942779, + "grad_norm": 2.177324771421986, + "learning_rate": 3.934286958458289e-06, + "loss": 1.0712, + "step": 1213 + }, + { + "epoch": 0.10948279749289805, + "grad_norm": 1.87424078511383, + "learning_rate": 3.934138351383632e-06, + "loss": 1.1026, + "step": 1214 + }, + { + "epoch": 0.10957298101636831, + "grad_norm": 1.677896227595686, + "learning_rate": 3.933989579277626e-06, + "loss": 0.9721, + "step": 1215 + }, + { + "epoch": 0.10966316453983857, + "grad_norm": 1.937874419427743, + "learning_rate": 3.933840642152966e-06, + "loss": 1.0325, + "step": 1216 + }, + { + "epoch": 0.10975334806330883, + "grad_norm": 2.041605448940384, + "learning_rate": 3.933691540022359e-06, + "loss": 1.0654, + "step": 1217 + }, + { + "epoch": 0.10984353158677909, + "grad_norm": 2.1025390625, + "learning_rate": 3.933542272898527e-06, + "loss": 1.0071, + "step": 1218 + }, + { + "epoch": 0.10993371511024935, + "grad_norm": 0.7562267583633214, + "learning_rate": 3.933392840794207e-06, + "loss": 0.8043, + "step": 1219 + }, + { + "epoch": 0.11002389863371961, + "grad_norm": 2.5774917287166574, + "learning_rate": 3.93324324372215e-06, + "loss": 1.1472, + "step": 1220 + }, + { + "epoch": 0.11011408215718989, + "grad_norm": 1.715204222728385, + "learning_rate": 3.9330934816951185e-06, + "loss": 1.0624, + "step": 1221 + }, + { + "epoch": 0.11020426568066015, + "grad_norm": 3.81207635745801, + "learning_rate": 3.932943554725893e-06, + "loss": 1.0693, + "step": 1222 + }, + { + "epoch": 0.11029444920413041, + "grad_norm": 0.6986371823782863, + "learning_rate": 3.932793462827265e-06, + "loss": 0.8785, + "step": 1223 + }, + { + "epoch": 0.11038463272760067, + "grad_norm": 1.9012735164474581, + "learning_rate": 3.932643206012041e-06, + "loss": 1.0977, + "step": 1224 + }, + { + "epoch": 0.11047481625107093, + "grad_norm": 2.2650900373605136, + "learning_rate": 3.932492784293043e-06, + "loss": 0.9643, + "step": 1225 + }, + { + "epoch": 0.11056499977454119, + "grad_norm": 1.6915617908734932, + "learning_rate": 3.932342197683104e-06, + "loss": 1.0297, + "step": 1226 + }, + { + "epoch": 0.11065518329801145, + "grad_norm": 2.141520737402649, + "learning_rate": 3.932191446195075e-06, + "loss": 0.913, + "step": 1227 + }, + { + "epoch": 0.11074536682148171, + "grad_norm": 1.874739183723572, + "learning_rate": 3.9320405298418175e-06, + "loss": 1.0426, + "step": 1228 + }, + { + "epoch": 0.11083555034495197, + "grad_norm": 1.9278428376671126, + "learning_rate": 3.9318894486362076e-06, + "loss": 0.9635, + "step": 1229 + }, + { + "epoch": 0.11092573386842224, + "grad_norm": 1.699106006609048, + "learning_rate": 3.9317382025911395e-06, + "loss": 1.1137, + "step": 1230 + }, + { + "epoch": 0.1110159173918925, + "grad_norm": 1.7345071003411987, + "learning_rate": 3.9315867917195145e-06, + "loss": 1.1407, + "step": 1231 + }, + { + "epoch": 0.11110610091536276, + "grad_norm": 1.5409969043440672, + "learning_rate": 3.931435216034256e-06, + "loss": 1.0442, + "step": 1232 + }, + { + "epoch": 0.11119628443883303, + "grad_norm": 1.8114474956706677, + "learning_rate": 3.931283475548293e-06, + "loss": 1.0414, + "step": 1233 + }, + { + "epoch": 0.11128646796230329, + "grad_norm": 1.3984615814010932, + "learning_rate": 3.931131570274576e-06, + "loss": 1.0418, + "step": 1234 + }, + { + "epoch": 0.11137665148577355, + "grad_norm": 2.2004890635226664, + "learning_rate": 3.930979500226065e-06, + "loss": 1.0759, + "step": 1235 + }, + { + "epoch": 0.11146683500924381, + "grad_norm": 3.083689969719022, + "learning_rate": 3.930827265415736e-06, + "loss": 0.8751, + "step": 1236 + }, + { + "epoch": 0.11155701853271408, + "grad_norm": 1.6391791375827933, + "learning_rate": 3.930674865856578e-06, + "loss": 1.0488, + "step": 1237 + }, + { + "epoch": 0.11164720205618434, + "grad_norm": 2.3907301137915478, + "learning_rate": 3.930522301561595e-06, + "loss": 1.0777, + "step": 1238 + }, + { + "epoch": 0.1117373855796546, + "grad_norm": 2.105681626239823, + "learning_rate": 3.930369572543804e-06, + "loss": 1.0851, + "step": 1239 + }, + { + "epoch": 0.11182756910312486, + "grad_norm": 1.5087729443920619, + "learning_rate": 3.930216678816237e-06, + "loss": 0.9836, + "step": 1240 + }, + { + "epoch": 0.11191775262659512, + "grad_norm": 0.6671993218833007, + "learning_rate": 3.930063620391941e-06, + "loss": 0.8518, + "step": 1241 + }, + { + "epoch": 0.11200793615006538, + "grad_norm": 2.154034900386087, + "learning_rate": 3.9299103972839735e-06, + "loss": 1.0441, + "step": 1242 + }, + { + "epoch": 0.11209811967353564, + "grad_norm": 0.6499859579110083, + "learning_rate": 3.92975700950541e-06, + "loss": 0.8163, + "step": 1243 + }, + { + "epoch": 0.1121883031970059, + "grad_norm": 1.8626466066149083, + "learning_rate": 3.929603457069338e-06, + "loss": 1.075, + "step": 1244 + }, + { + "epoch": 0.11227848672047618, + "grad_norm": 1.6394380181111672, + "learning_rate": 3.929449739988859e-06, + "loss": 1.008, + "step": 1245 + }, + { + "epoch": 0.11236867024394644, + "grad_norm": 1.8063199805196575, + "learning_rate": 3.929295858277089e-06, + "loss": 1.0385, + "step": 1246 + }, + { + "epoch": 0.1124588537674167, + "grad_norm": 2.2909192571510775, + "learning_rate": 3.9291418119471585e-06, + "loss": 1.0409, + "step": 1247 + }, + { + "epoch": 0.11254903729088696, + "grad_norm": 2.0599017104986546, + "learning_rate": 3.928987601012212e-06, + "loss": 1.069, + "step": 1248 + }, + { + "epoch": 0.11263922081435722, + "grad_norm": 2.2706312970452913, + "learning_rate": 3.928833225485407e-06, + "loss": 1.0344, + "step": 1249 + }, + { + "epoch": 0.11272940433782748, + "grad_norm": 2.027393377207498, + "learning_rate": 3.928678685379915e-06, + "loss": 1.0293, + "step": 1250 + }, + { + "epoch": 0.11281958786129774, + "grad_norm": 1.634863701251738, + "learning_rate": 3.928523980708924e-06, + "loss": 1.0898, + "step": 1251 + }, + { + "epoch": 0.112909771384768, + "grad_norm": 4.879798874697103, + "learning_rate": 3.928369111485632e-06, + "loss": 0.9965, + "step": 1252 + }, + { + "epoch": 0.11299995490823826, + "grad_norm": 2.4106951193072823, + "learning_rate": 3.928214077723255e-06, + "loss": 1.0316, + "step": 1253 + }, + { + "epoch": 0.11309013843170852, + "grad_norm": 2.200863157352164, + "learning_rate": 3.928058879435021e-06, + "loss": 1.0317, + "step": 1254 + }, + { + "epoch": 0.11318032195517878, + "grad_norm": 3.69045992230933, + "learning_rate": 3.9279035166341725e-06, + "loss": 1.027, + "step": 1255 + }, + { + "epoch": 0.11327050547864904, + "grad_norm": 2.157971123701042, + "learning_rate": 3.927747989333965e-06, + "loss": 1.0219, + "step": 1256 + }, + { + "epoch": 0.11336068900211932, + "grad_norm": 3.436913457593926, + "learning_rate": 3.927592297547669e-06, + "loss": 1.0025, + "step": 1257 + }, + { + "epoch": 0.11345087252558958, + "grad_norm": 11.751981527564565, + "learning_rate": 3.927436441288571e-06, + "loss": 1.0681, + "step": 1258 + }, + { + "epoch": 0.11354105604905984, + "grad_norm": 2.877209850699698, + "learning_rate": 3.927280420569968e-06, + "loss": 1.0572, + "step": 1259 + }, + { + "epoch": 0.1136312395725301, + "grad_norm": 1.7341069194736822, + "learning_rate": 3.927124235405171e-06, + "loss": 1.0651, + "step": 1260 + }, + { + "epoch": 0.11372142309600036, + "grad_norm": 1.900688438172197, + "learning_rate": 3.92696788580751e-06, + "loss": 0.8995, + "step": 1261 + }, + { + "epoch": 0.11381160661947062, + "grad_norm": 1.5871528516255824, + "learning_rate": 3.9268113717903225e-06, + "loss": 0.8795, + "step": 1262 + }, + { + "epoch": 0.11390179014294088, + "grad_norm": 1.931403875547773, + "learning_rate": 3.926654693366965e-06, + "loss": 1.1074, + "step": 1263 + }, + { + "epoch": 0.11399197366641114, + "grad_norm": 1.9809049167411623, + "learning_rate": 3.926497850550805e-06, + "loss": 1.0507, + "step": 1264 + }, + { + "epoch": 0.1140821571898814, + "grad_norm": 1.9774526167448228, + "learning_rate": 3.926340843355226e-06, + "loss": 1.0343, + "step": 1265 + }, + { + "epoch": 0.11417234071335167, + "grad_norm": 1.924526587398004, + "learning_rate": 3.926183671793625e-06, + "loss": 1.0551, + "step": 1266 + }, + { + "epoch": 0.11426252423682193, + "grad_norm": 2.246576353553872, + "learning_rate": 3.926026335879412e-06, + "loss": 1.0042, + "step": 1267 + }, + { + "epoch": 0.11435270776029219, + "grad_norm": 0.6370882939465111, + "learning_rate": 3.925868835626012e-06, + "loss": 0.7889, + "step": 1268 + }, + { + "epoch": 0.11444289128376246, + "grad_norm": 5.049375121481037, + "learning_rate": 3.925711171046864e-06, + "loss": 1.0467, + "step": 1269 + }, + { + "epoch": 0.11453307480723272, + "grad_norm": 2.1206491471547975, + "learning_rate": 3.925553342155421e-06, + "loss": 1.0008, + "step": 1270 + }, + { + "epoch": 0.11462325833070298, + "grad_norm": 0.6900324778238605, + "learning_rate": 3.9253953489651485e-06, + "loss": 0.8632, + "step": 1271 + }, + { + "epoch": 0.11471344185417325, + "grad_norm": 0.7128093098289349, + "learning_rate": 3.925237191489529e-06, + "loss": 0.8382, + "step": 1272 + }, + { + "epoch": 0.1148036253776435, + "grad_norm": 1.6297953349615113, + "learning_rate": 3.925078869742056e-06, + "loss": 1.012, + "step": 1273 + }, + { + "epoch": 0.11489380890111377, + "grad_norm": 1.7284450710574286, + "learning_rate": 3.92492038373624e-06, + "loss": 0.9725, + "step": 1274 + }, + { + "epoch": 0.11498399242458403, + "grad_norm": 6.927295775251155, + "learning_rate": 3.924761733485602e-06, + "loss": 1.0942, + "step": 1275 + }, + { + "epoch": 0.11507417594805429, + "grad_norm": 3.3824999799256137, + "learning_rate": 3.92460291900368e-06, + "loss": 1.0346, + "step": 1276 + }, + { + "epoch": 0.11516435947152455, + "grad_norm": 1.497482889934223, + "learning_rate": 3.924443940304025e-06, + "loss": 1.0585, + "step": 1277 + }, + { + "epoch": 0.11525454299499481, + "grad_norm": 1.8879388690564287, + "learning_rate": 3.924284797400202e-06, + "loss": 1.0928, + "step": 1278 + }, + { + "epoch": 0.11534472651846507, + "grad_norm": 2.865021513746617, + "learning_rate": 3.924125490305789e-06, + "loss": 1.0615, + "step": 1279 + }, + { + "epoch": 0.11543491004193533, + "grad_norm": 1.6224045934384494, + "learning_rate": 3.923966019034381e-06, + "loss": 1.0174, + "step": 1280 + }, + { + "epoch": 0.1155250935654056, + "grad_norm": 1.7900558445919537, + "learning_rate": 3.923806383599583e-06, + "loss": 1.0464, + "step": 1281 + }, + { + "epoch": 0.11561527708887587, + "grad_norm": 1.9593649101910144, + "learning_rate": 3.923646584015017e-06, + "loss": 1.0053, + "step": 1282 + }, + { + "epoch": 0.11570546061234613, + "grad_norm": 6.271483462608934, + "learning_rate": 3.923486620294316e-06, + "loss": 1.037, + "step": 1283 + }, + { + "epoch": 0.11579564413581639, + "grad_norm": 1.7357361409775338, + "learning_rate": 3.923326492451132e-06, + "loss": 1.0495, + "step": 1284 + }, + { + "epoch": 0.11588582765928665, + "grad_norm": 15.55203127825696, + "learning_rate": 3.923166200499125e-06, + "loss": 1.0142, + "step": 1285 + }, + { + "epoch": 0.11597601118275691, + "grad_norm": 1.962786771389757, + "learning_rate": 3.923005744451975e-06, + "loss": 1.0104, + "step": 1286 + }, + { + "epoch": 0.11606619470622717, + "grad_norm": 1.9582456068585998, + "learning_rate": 3.9228451243233715e-06, + "loss": 0.9909, + "step": 1287 + }, + { + "epoch": 0.11615637822969743, + "grad_norm": 2.1999189708699625, + "learning_rate": 3.9226843401270195e-06, + "loss": 1.067, + "step": 1288 + }, + { + "epoch": 0.1162465617531677, + "grad_norm": 3.3584539525889223, + "learning_rate": 3.9225233918766376e-06, + "loss": 1.0387, + "step": 1289 + }, + { + "epoch": 0.11633674527663795, + "grad_norm": 1.8712360750035724, + "learning_rate": 3.92236227958596e-06, + "loss": 1.0542, + "step": 1290 + }, + { + "epoch": 0.11642692880010821, + "grad_norm": 1.9711846329537457, + "learning_rate": 3.922201003268731e-06, + "loss": 1.0623, + "step": 1291 + }, + { + "epoch": 0.11651711232357848, + "grad_norm": 1.1407221400035483, + "learning_rate": 3.922039562938715e-06, + "loss": 0.8652, + "step": 1292 + }, + { + "epoch": 0.11660729584704875, + "grad_norm": 1.578523491129714, + "learning_rate": 3.921877958609685e-06, + "loss": 1.0858, + "step": 1293 + }, + { + "epoch": 0.11669747937051901, + "grad_norm": 0.7932196159230324, + "learning_rate": 3.921716190295431e-06, + "loss": 0.9288, + "step": 1294 + }, + { + "epoch": 0.11678766289398927, + "grad_norm": 5.005403459961752, + "learning_rate": 3.921554258009755e-06, + "loss": 0.9943, + "step": 1295 + }, + { + "epoch": 0.11687784641745953, + "grad_norm": 6.488452925182141, + "learning_rate": 3.921392161766474e-06, + "loss": 1.0222, + "step": 1296 + }, + { + "epoch": 0.1169680299409298, + "grad_norm": 4.3488245899703815, + "learning_rate": 3.92122990157942e-06, + "loss": 1.0123, + "step": 1297 + }, + { + "epoch": 0.11705821346440005, + "grad_norm": 1.7849112912315608, + "learning_rate": 3.921067477462437e-06, + "loss": 1.082, + "step": 1298 + }, + { + "epoch": 0.11714839698787032, + "grad_norm": 2.3858897865785735, + "learning_rate": 3.920904889429385e-06, + "loss": 1.0942, + "step": 1299 + }, + { + "epoch": 0.11723858051134058, + "grad_norm": 2.0427629939781684, + "learning_rate": 3.920742137494135e-06, + "loss": 1.0768, + "step": 1300 + }, + { + "epoch": 0.11732876403481084, + "grad_norm": 1.4519322740454932, + "learning_rate": 3.920579221670575e-06, + "loss": 1.0931, + "step": 1301 + }, + { + "epoch": 0.1174189475582811, + "grad_norm": 1.8664364760814756, + "learning_rate": 3.920416141972606e-06, + "loss": 1.0385, + "step": 1302 + }, + { + "epoch": 0.11750913108175136, + "grad_norm": 1.9743157686367496, + "learning_rate": 3.920252898414143e-06, + "loss": 1.0371, + "step": 1303 + }, + { + "epoch": 0.11759931460522162, + "grad_norm": 2.095968750458644, + "learning_rate": 3.920089491009114e-06, + "loss": 1.0505, + "step": 1304 + }, + { + "epoch": 0.1176894981286919, + "grad_norm": 1.578318442395602, + "learning_rate": 3.919925919771463e-06, + "loss": 1.1211, + "step": 1305 + }, + { + "epoch": 0.11777968165216215, + "grad_norm": 1.8774844557808217, + "learning_rate": 3.919762184715146e-06, + "loss": 0.9528, + "step": 1306 + }, + { + "epoch": 0.11786986517563242, + "grad_norm": 1.859354163301546, + "learning_rate": 3.919598285854134e-06, + "loss": 1.0362, + "step": 1307 + }, + { + "epoch": 0.11796004869910268, + "grad_norm": 2.4860799445312667, + "learning_rate": 3.919434223202411e-06, + "loss": 1.0929, + "step": 1308 + }, + { + "epoch": 0.11805023222257294, + "grad_norm": 1.5608953246871256, + "learning_rate": 3.919269996773977e-06, + "loss": 1.096, + "step": 1309 + }, + { + "epoch": 0.1181404157460432, + "grad_norm": 1.5081023891254253, + "learning_rate": 3.919105606582844e-06, + "loss": 1.0504, + "step": 1310 + }, + { + "epoch": 0.11823059926951346, + "grad_norm": 1.6591675450878158, + "learning_rate": 3.918941052643039e-06, + "loss": 1.0051, + "step": 1311 + }, + { + "epoch": 0.11832078279298372, + "grad_norm": 1.7577699105083617, + "learning_rate": 3.918776334968602e-06, + "loss": 1.0505, + "step": 1312 + }, + { + "epoch": 0.11841096631645398, + "grad_norm": 1.9927512175403495, + "learning_rate": 3.918611453573589e-06, + "loss": 1.1484, + "step": 1313 + }, + { + "epoch": 0.11850114983992424, + "grad_norm": 2.7140509675988254, + "learning_rate": 3.918446408472066e-06, + "loss": 0.9737, + "step": 1314 + }, + { + "epoch": 0.1185913333633945, + "grad_norm": 2.141220789508029, + "learning_rate": 3.918281199678119e-06, + "loss": 1.0769, + "step": 1315 + }, + { + "epoch": 0.11868151688686476, + "grad_norm": 0.7355328218694742, + "learning_rate": 3.9181158272058414e-06, + "loss": 0.8461, + "step": 1316 + }, + { + "epoch": 0.11877170041033504, + "grad_norm": 1.5865196324722777, + "learning_rate": 3.9179502910693455e-06, + "loss": 1.0672, + "step": 1317 + }, + { + "epoch": 0.1188618839338053, + "grad_norm": 1.9052611741180163, + "learning_rate": 3.917784591282756e-06, + "loss": 0.9958, + "step": 1318 + }, + { + "epoch": 0.11895206745727556, + "grad_norm": 1.591197494275866, + "learning_rate": 3.9176187278602105e-06, + "loss": 1.0645, + "step": 1319 + }, + { + "epoch": 0.11904225098074582, + "grad_norm": 1.6661781151297361, + "learning_rate": 3.9174527008158606e-06, + "loss": 0.9792, + "step": 1320 + }, + { + "epoch": 0.11913243450421608, + "grad_norm": 1.8774185476908183, + "learning_rate": 3.917286510163874e-06, + "loss": 0.9594, + "step": 1321 + }, + { + "epoch": 0.11922261802768634, + "grad_norm": 2.572324759369154, + "learning_rate": 3.917120155918431e-06, + "loss": 1.0124, + "step": 1322 + }, + { + "epoch": 0.1193128015511566, + "grad_norm": 0.7017398966777791, + "learning_rate": 3.916953638093725e-06, + "loss": 0.8775, + "step": 1323 + }, + { + "epoch": 0.11940298507462686, + "grad_norm": 2.340020938844817, + "learning_rate": 3.916786956703964e-06, + "loss": 1.0291, + "step": 1324 + }, + { + "epoch": 0.11949316859809712, + "grad_norm": 1.9940663771299456, + "learning_rate": 3.916620111763372e-06, + "loss": 1.0261, + "step": 1325 + }, + { + "epoch": 0.11958335212156739, + "grad_norm": 1.7061071622537747, + "learning_rate": 3.916453103286183e-06, + "loss": 1.0981, + "step": 1326 + }, + { + "epoch": 0.11967353564503765, + "grad_norm": 1.9518126083892409, + "learning_rate": 3.916285931286648e-06, + "loss": 1.0837, + "step": 1327 + }, + { + "epoch": 0.11976371916850792, + "grad_norm": 1.7644802196990559, + "learning_rate": 3.916118595779031e-06, + "loss": 0.9819, + "step": 1328 + }, + { + "epoch": 0.11985390269197818, + "grad_norm": 1.769950833796178, + "learning_rate": 3.915951096777611e-06, + "loss": 1.0091, + "step": 1329 + }, + { + "epoch": 0.11994408621544844, + "grad_norm": 1.8682674333495857, + "learning_rate": 3.915783434296678e-06, + "loss": 0.9853, + "step": 1330 + }, + { + "epoch": 0.1200342697389187, + "grad_norm": 1.8658268651584664, + "learning_rate": 3.91561560835054e-06, + "loss": 1.0193, + "step": 1331 + }, + { + "epoch": 0.12012445326238896, + "grad_norm": 1.7276588514856455, + "learning_rate": 3.915447618953515e-06, + "loss": 1.0553, + "step": 1332 + }, + { + "epoch": 0.12021463678585922, + "grad_norm": 2.355972767515451, + "learning_rate": 3.915279466119937e-06, + "loss": 0.9896, + "step": 1333 + }, + { + "epoch": 0.12030482030932949, + "grad_norm": 0.7171310352711804, + "learning_rate": 3.9151111498641546e-06, + "loss": 0.8623, + "step": 1334 + }, + { + "epoch": 0.12039500383279975, + "grad_norm": 2.1448147278914282, + "learning_rate": 3.914942670200529e-06, + "loss": 0.9121, + "step": 1335 + }, + { + "epoch": 0.12048518735627001, + "grad_norm": 1.793463678137356, + "learning_rate": 3.914774027143436e-06, + "loss": 1.0972, + "step": 1336 + }, + { + "epoch": 0.12057537087974027, + "grad_norm": 1.799842374575752, + "learning_rate": 3.914605220707265e-06, + "loss": 1.0265, + "step": 1337 + }, + { + "epoch": 0.12066555440321053, + "grad_norm": 1.7683882569182654, + "learning_rate": 3.9144362509064194e-06, + "loss": 1.1853, + "step": 1338 + }, + { + "epoch": 0.12075573792668079, + "grad_norm": 1.9203010875341053, + "learning_rate": 3.914267117755317e-06, + "loss": 1.1037, + "step": 1339 + }, + { + "epoch": 0.12084592145015106, + "grad_norm": 2.223775196500931, + "learning_rate": 3.914097821268389e-06, + "loss": 1.0, + "step": 1340 + }, + { + "epoch": 0.12093610497362133, + "grad_norm": 2.294931640604222, + "learning_rate": 3.913928361460081e-06, + "loss": 1.0495, + "step": 1341 + }, + { + "epoch": 0.12102628849709159, + "grad_norm": 1.6848176901697367, + "learning_rate": 3.913758738344851e-06, + "loss": 1.0251, + "step": 1342 + }, + { + "epoch": 0.12111647202056185, + "grad_norm": 0.7246986190252147, + "learning_rate": 3.913588951937174e-06, + "loss": 0.8387, + "step": 1343 + }, + { + "epoch": 0.12120665554403211, + "grad_norm": 0.7172695337389923, + "learning_rate": 3.9134190022515355e-06, + "loss": 0.839, + "step": 1344 + }, + { + "epoch": 0.12129683906750237, + "grad_norm": 2.0328950529467, + "learning_rate": 3.913248889302438e-06, + "loss": 0.9982, + "step": 1345 + }, + { + "epoch": 0.12138702259097263, + "grad_norm": 1.5884386456545658, + "learning_rate": 3.913078613104395e-06, + "loss": 0.9719, + "step": 1346 + }, + { + "epoch": 0.12147720611444289, + "grad_norm": 3.526554598142385, + "learning_rate": 3.912908173671936e-06, + "loss": 0.9223, + "step": 1347 + }, + { + "epoch": 0.12156738963791315, + "grad_norm": 3.2021820913525825, + "learning_rate": 3.9127375710196044e-06, + "loss": 0.9659, + "step": 1348 + }, + { + "epoch": 0.12165757316138341, + "grad_norm": 1.7370861960345012, + "learning_rate": 3.912566805161957e-06, + "loss": 0.9787, + "step": 1349 + }, + { + "epoch": 0.12174775668485367, + "grad_norm": 1.8656765238996937, + "learning_rate": 3.912395876113564e-06, + "loss": 1.0546, + "step": 1350 + }, + { + "epoch": 0.12183794020832393, + "grad_norm": 1.7857329858073243, + "learning_rate": 3.912224783889009e-06, + "loss": 1.0384, + "step": 1351 + }, + { + "epoch": 0.12192812373179421, + "grad_norm": 2.150705811618284, + "learning_rate": 3.912053528502892e-06, + "loss": 1.0481, + "step": 1352 + }, + { + "epoch": 0.12201830725526447, + "grad_norm": 1.7480416239228043, + "learning_rate": 3.911882109969825e-06, + "loss": 0.963, + "step": 1353 + }, + { + "epoch": 0.12210849077873473, + "grad_norm": 1.8229858675946193, + "learning_rate": 3.911710528304435e-06, + "loss": 1.0425, + "step": 1354 + }, + { + "epoch": 0.12219867430220499, + "grad_norm": 6.753825763600092, + "learning_rate": 3.911538783521361e-06, + "loss": 0.9875, + "step": 1355 + }, + { + "epoch": 0.12228885782567525, + "grad_norm": 3.4877412555336953, + "learning_rate": 3.9113668756352575e-06, + "loss": 1.0477, + "step": 1356 + }, + { + "epoch": 0.12237904134914551, + "grad_norm": 1.5552239187506807, + "learning_rate": 3.911194804660793e-06, + "loss": 0.9615, + "step": 1357 + }, + { + "epoch": 0.12246922487261577, + "grad_norm": 1.805144545243852, + "learning_rate": 3.91102257061265e-06, + "loss": 1.116, + "step": 1358 + }, + { + "epoch": 0.12255940839608603, + "grad_norm": 1.4735833040267015, + "learning_rate": 3.910850173505524e-06, + "loss": 1.0764, + "step": 1359 + }, + { + "epoch": 0.1226495919195563, + "grad_norm": 1.9397150114482842, + "learning_rate": 3.9106776133541255e-06, + "loss": 1.1097, + "step": 1360 + }, + { + "epoch": 0.12273977544302656, + "grad_norm": 1.5900243708104345, + "learning_rate": 3.9105048901731766e-06, + "loss": 0.9958, + "step": 1361 + }, + { + "epoch": 0.12282995896649682, + "grad_norm": 1.9451399883700091, + "learning_rate": 3.9103320039774165e-06, + "loss": 1.0266, + "step": 1362 + }, + { + "epoch": 0.12292014248996708, + "grad_norm": 1.950549632642047, + "learning_rate": 3.9101589547815965e-06, + "loss": 1.0925, + "step": 1363 + }, + { + "epoch": 0.12301032601343735, + "grad_norm": 1.6826933468825935, + "learning_rate": 3.909985742600482e-06, + "loss": 1.025, + "step": 1364 + }, + { + "epoch": 0.12310050953690761, + "grad_norm": 1.746408592018454, + "learning_rate": 3.909812367448852e-06, + "loss": 0.9978, + "step": 1365 + }, + { + "epoch": 0.12319069306037787, + "grad_norm": 2.3117437414821174, + "learning_rate": 3.909638829341501e-06, + "loss": 1.0118, + "step": 1366 + }, + { + "epoch": 0.12328087658384813, + "grad_norm": 1.8090660846125781, + "learning_rate": 3.909465128293234e-06, + "loss": 1.0885, + "step": 1367 + }, + { + "epoch": 0.1233710601073184, + "grad_norm": 1.8781668939218235, + "learning_rate": 3.9092912643188745e-06, + "loss": 1.092, + "step": 1368 + }, + { + "epoch": 0.12346124363078866, + "grad_norm": 1.7883466596127608, + "learning_rate": 3.909117237433256e-06, + "loss": 1.0292, + "step": 1369 + }, + { + "epoch": 0.12355142715425892, + "grad_norm": 6.692240238271358, + "learning_rate": 3.908943047651229e-06, + "loss": 1.0774, + "step": 1370 + }, + { + "epoch": 0.12364161067772918, + "grad_norm": 3.5542466781095174, + "learning_rate": 3.908768694987655e-06, + "loss": 1.0225, + "step": 1371 + }, + { + "epoch": 0.12373179420119944, + "grad_norm": 3.6158402572557335, + "learning_rate": 3.908594179457411e-06, + "loss": 0.9552, + "step": 1372 + }, + { + "epoch": 0.1238219777246697, + "grad_norm": 2.151086124825888, + "learning_rate": 3.908419501075388e-06, + "loss": 1.0417, + "step": 1373 + }, + { + "epoch": 0.12391216124813996, + "grad_norm": 1.5438729642629618, + "learning_rate": 3.90824465985649e-06, + "loss": 1.1117, + "step": 1374 + }, + { + "epoch": 0.12400234477161022, + "grad_norm": 1.7765942680585063, + "learning_rate": 3.908069655815636e-06, + "loss": 1.0777, + "step": 1375 + }, + { + "epoch": 0.1240925282950805, + "grad_norm": 2.0734262910265384, + "learning_rate": 3.907894488967758e-06, + "loss": 0.9768, + "step": 1376 + }, + { + "epoch": 0.12418271181855076, + "grad_norm": 1.788258800973971, + "learning_rate": 3.9077191593278005e-06, + "loss": 1.0956, + "step": 1377 + }, + { + "epoch": 0.12427289534202102, + "grad_norm": 0.762674758770946, + "learning_rate": 3.9075436669107265e-06, + "loss": 0.8156, + "step": 1378 + }, + { + "epoch": 0.12436307886549128, + "grad_norm": 1.9061050359918896, + "learning_rate": 3.90736801173151e-06, + "loss": 1.0485, + "step": 1379 + }, + { + "epoch": 0.12445326238896154, + "grad_norm": 1.524587031246405, + "learning_rate": 3.907192193805136e-06, + "loss": 1.0441, + "step": 1380 + }, + { + "epoch": 0.1245434459124318, + "grad_norm": 4.045969504603036, + "learning_rate": 3.907016213146608e-06, + "loss": 1.085, + "step": 1381 + }, + { + "epoch": 0.12463362943590206, + "grad_norm": 2.5184867172269607, + "learning_rate": 3.906840069770942e-06, + "loss": 1.1486, + "step": 1382 + }, + { + "epoch": 0.12472381295937232, + "grad_norm": 1.3861315276392743, + "learning_rate": 3.906663763693167e-06, + "loss": 1.0657, + "step": 1383 + }, + { + "epoch": 0.12481399648284258, + "grad_norm": 1.7888889145242335, + "learning_rate": 3.906487294928327e-06, + "loss": 1.0154, + "step": 1384 + }, + { + "epoch": 0.12490418000631284, + "grad_norm": 1.7733949496504378, + "learning_rate": 3.906310663491478e-06, + "loss": 1.0383, + "step": 1385 + }, + { + "epoch": 0.1249943635297831, + "grad_norm": 2.100771616909743, + "learning_rate": 3.906133869397692e-06, + "loss": 1.0908, + "step": 1386 + }, + { + "epoch": 0.12508454705325336, + "grad_norm": 1.83591808552318, + "learning_rate": 3.905956912662054e-06, + "loss": 1.0028, + "step": 1387 + }, + { + "epoch": 0.12517473057672363, + "grad_norm": 2.8159789714949244, + "learning_rate": 3.905779793299662e-06, + "loss": 1.0957, + "step": 1388 + }, + { + "epoch": 0.12526491410019389, + "grad_norm": 1.6504681992216166, + "learning_rate": 3.905602511325631e-06, + "loss": 1.0705, + "step": 1389 + }, + { + "epoch": 0.12535509762366415, + "grad_norm": 1.8504196696209465, + "learning_rate": 3.905425066755086e-06, + "loss": 1.0005, + "step": 1390 + }, + { + "epoch": 0.1254452811471344, + "grad_norm": 1.579714060030913, + "learning_rate": 3.905247459603168e-06, + "loss": 0.9947, + "step": 1391 + }, + { + "epoch": 0.12553546467060467, + "grad_norm": 2.018062566628485, + "learning_rate": 3.905069689885031e-06, + "loss": 0.9154, + "step": 1392 + }, + { + "epoch": 0.12562564819407493, + "grad_norm": 1.8127737167463072, + "learning_rate": 3.904891757615843e-06, + "loss": 0.9183, + "step": 1393 + }, + { + "epoch": 0.12571583171754522, + "grad_norm": 0.6215568831061975, + "learning_rate": 3.9047136628107874e-06, + "loss": 0.8011, + "step": 1394 + }, + { + "epoch": 0.12580601524101548, + "grad_norm": 1.8339096232852299, + "learning_rate": 3.904535405485059e-06, + "loss": 1.1664, + "step": 1395 + }, + { + "epoch": 0.12589619876448574, + "grad_norm": 3.727146830621149, + "learning_rate": 3.90435698565387e-06, + "loss": 0.9853, + "step": 1396 + }, + { + "epoch": 0.125986382287956, + "grad_norm": 1.7027026711488065, + "learning_rate": 3.904178403332441e-06, + "loss": 1.0014, + "step": 1397 + }, + { + "epoch": 0.12607656581142626, + "grad_norm": 1.910923840652518, + "learning_rate": 3.903999658536012e-06, + "loss": 1.1221, + "step": 1398 + }, + { + "epoch": 0.12616674933489652, + "grad_norm": 1.7113570304263221, + "learning_rate": 3.903820751279833e-06, + "loss": 1.0332, + "step": 1399 + }, + { + "epoch": 0.12625693285836678, + "grad_norm": 3.0716216036188233, + "learning_rate": 3.90364168157917e-06, + "loss": 1.0628, + "step": 1400 + }, + { + "epoch": 0.12634711638183704, + "grad_norm": 1.7702792778163745, + "learning_rate": 3.903462449449302e-06, + "loss": 1.0681, + "step": 1401 + }, + { + "epoch": 0.1264372999053073, + "grad_norm": 5.9556341215110455, + "learning_rate": 3.903283054905522e-06, + "loss": 1.0088, + "step": 1402 + }, + { + "epoch": 0.12652748342877757, + "grad_norm": 1.9172816395672685, + "learning_rate": 3.9031034979631385e-06, + "loss": 0.9468, + "step": 1403 + }, + { + "epoch": 0.12661766695224783, + "grad_norm": 1.4934916763712267, + "learning_rate": 3.902923778637469e-06, + "loss": 1.0716, + "step": 1404 + }, + { + "epoch": 0.1267078504757181, + "grad_norm": 1.6468503280615694, + "learning_rate": 3.902743896943852e-06, + "loss": 0.969, + "step": 1405 + }, + { + "epoch": 0.12679803399918835, + "grad_norm": 0.7357121732102115, + "learning_rate": 3.902563852897633e-06, + "loss": 0.8705, + "step": 1406 + }, + { + "epoch": 0.1268882175226586, + "grad_norm": 0.6962875644405563, + "learning_rate": 3.9023836465141755e-06, + "loss": 0.8443, + "step": 1407 + }, + { + "epoch": 0.12697840104612887, + "grad_norm": 1.8756123496534367, + "learning_rate": 3.902203277808856e-06, + "loss": 1.0783, + "step": 1408 + }, + { + "epoch": 0.12706858456959913, + "grad_norm": 1.8346296481350235, + "learning_rate": 3.902022746797064e-06, + "loss": 1.0951, + "step": 1409 + }, + { + "epoch": 0.1271587680930694, + "grad_norm": 2.3111627165930098, + "learning_rate": 3.9018420534942035e-06, + "loss": 1.1459, + "step": 1410 + }, + { + "epoch": 0.12724895161653965, + "grad_norm": 2.2935268366299524, + "learning_rate": 3.9016611979156935e-06, + "loss": 1.0258, + "step": 1411 + }, + { + "epoch": 0.1273391351400099, + "grad_norm": 2.357049465393548, + "learning_rate": 3.9014801800769635e-06, + "loss": 0.9384, + "step": 1412 + }, + { + "epoch": 0.12742931866348017, + "grad_norm": 1.8319788032621465, + "learning_rate": 3.901298999993459e-06, + "loss": 1.0622, + "step": 1413 + }, + { + "epoch": 0.12751950218695043, + "grad_norm": 1.913825506991891, + "learning_rate": 3.901117657680642e-06, + "loss": 0.9643, + "step": 1414 + }, + { + "epoch": 0.1276096857104207, + "grad_norm": 1.6213744874100098, + "learning_rate": 3.900936153153982e-06, + "loss": 0.9636, + "step": 1415 + }, + { + "epoch": 0.12769986923389096, + "grad_norm": 0.7161277953679451, + "learning_rate": 3.900754486428968e-06, + "loss": 0.7973, + "step": 1416 + }, + { + "epoch": 0.12779005275736122, + "grad_norm": 19.45398310105398, + "learning_rate": 3.900572657521102e-06, + "loss": 1.124, + "step": 1417 + }, + { + "epoch": 0.1278802362808315, + "grad_norm": 1.5261370875326459, + "learning_rate": 3.900390666445896e-06, + "loss": 1.0187, + "step": 1418 + }, + { + "epoch": 0.12797041980430177, + "grad_norm": 2.168567911468675, + "learning_rate": 3.9002085132188795e-06, + "loss": 0.9987, + "step": 1419 + }, + { + "epoch": 0.12806060332777203, + "grad_norm": 5.5933957120912705, + "learning_rate": 3.9000261978555964e-06, + "loss": 1.0804, + "step": 1420 + }, + { + "epoch": 0.1281507868512423, + "grad_norm": 2.428851115526442, + "learning_rate": 3.8998437203716e-06, + "loss": 1.104, + "step": 1421 + }, + { + "epoch": 0.12824097037471255, + "grad_norm": 1.5393979244407012, + "learning_rate": 3.899661080782462e-06, + "loss": 1.0595, + "step": 1422 + }, + { + "epoch": 0.1283311538981828, + "grad_norm": 1.9048365371138225, + "learning_rate": 3.899478279103767e-06, + "loss": 1.0652, + "step": 1423 + }, + { + "epoch": 0.12842133742165307, + "grad_norm": 1.888825561203609, + "learning_rate": 3.8992953153511105e-06, + "loss": 1.0665, + "step": 1424 + }, + { + "epoch": 0.12851152094512333, + "grad_norm": 1.9635331805490182, + "learning_rate": 3.899112189540106e-06, + "loss": 1.0958, + "step": 1425 + }, + { + "epoch": 0.1286017044685936, + "grad_norm": 2.0694480386769647, + "learning_rate": 3.898928901686377e-06, + "loss": 1.0687, + "step": 1426 + }, + { + "epoch": 0.12869188799206385, + "grad_norm": 0.7107895498069999, + "learning_rate": 3.898745451805564e-06, + "loss": 0.8026, + "step": 1427 + }, + { + "epoch": 0.1287820715155341, + "grad_norm": 2.0502562804640485, + "learning_rate": 3.898561839913319e-06, + "loss": 0.9703, + "step": 1428 + }, + { + "epoch": 0.12887225503900437, + "grad_norm": 6.866258647363934, + "learning_rate": 3.89837806602531e-06, + "loss": 0.8856, + "step": 1429 + }, + { + "epoch": 0.12896243856247463, + "grad_norm": 2.2617102113868937, + "learning_rate": 3.898194130157217e-06, + "loss": 1.0422, + "step": 1430 + }, + { + "epoch": 0.1290526220859449, + "grad_norm": 2.415544764677805, + "learning_rate": 3.8980100323247335e-06, + "loss": 0.9773, + "step": 1431 + }, + { + "epoch": 0.12914280560941516, + "grad_norm": 11.71502219353925, + "learning_rate": 3.897825772543568e-06, + "loss": 0.9747, + "step": 1432 + }, + { + "epoch": 0.12923298913288542, + "grad_norm": 1.989596368503976, + "learning_rate": 3.897641350829444e-06, + "loss": 1.0963, + "step": 1433 + }, + { + "epoch": 0.12932317265635568, + "grad_norm": 1.55078125, + "learning_rate": 3.897456767198096e-06, + "loss": 1.1197, + "step": 1434 + }, + { + "epoch": 0.12941335617982594, + "grad_norm": 3.800877349360004, + "learning_rate": 3.897272021665275e-06, + "loss": 0.9761, + "step": 1435 + }, + { + "epoch": 0.1295035397032962, + "grad_norm": 1.5108095732284852, + "learning_rate": 3.897087114246743e-06, + "loss": 1.1463, + "step": 1436 + }, + { + "epoch": 0.12959372322676646, + "grad_norm": 2.229754492241827, + "learning_rate": 3.896902044958279e-06, + "loss": 1.1504, + "step": 1437 + }, + { + "epoch": 0.12968390675023672, + "grad_norm": 1.8970121205273984, + "learning_rate": 3.896716813815672e-06, + "loss": 1.0906, + "step": 1438 + }, + { + "epoch": 0.12977409027370698, + "grad_norm": 1.5779841898427103, + "learning_rate": 3.896531420834728e-06, + "loss": 0.9837, + "step": 1439 + }, + { + "epoch": 0.12986427379717724, + "grad_norm": 1.7555410671575196, + "learning_rate": 3.896345866031266e-06, + "loss": 0.9456, + "step": 1440 + }, + { + "epoch": 0.1299544573206475, + "grad_norm": 2.286887219175611, + "learning_rate": 3.896160149421119e-06, + "loss": 1.0954, + "step": 1441 + }, + { + "epoch": 0.1300446408441178, + "grad_norm": 2.0749147328701585, + "learning_rate": 3.8959742710201314e-06, + "loss": 1.1021, + "step": 1442 + }, + { + "epoch": 0.13013482436758805, + "grad_norm": 2.632827985135053, + "learning_rate": 3.895788230844166e-06, + "loss": 0.9291, + "step": 1443 + }, + { + "epoch": 0.13022500789105831, + "grad_norm": 6.692549180885727, + "learning_rate": 3.895602028909095e-06, + "loss": 1.1943, + "step": 1444 + }, + { + "epoch": 0.13031519141452858, + "grad_norm": 0.6357026681175245, + "learning_rate": 3.895415665230807e-06, + "loss": 0.8357, + "step": 1445 + }, + { + "epoch": 0.13040537493799884, + "grad_norm": 1.3575827962871065, + "learning_rate": 3.895229139825203e-06, + "loss": 1.0624, + "step": 1446 + }, + { + "epoch": 0.1304955584614691, + "grad_norm": 1.891082267004591, + "learning_rate": 3.895042452708198e-06, + "loss": 0.9518, + "step": 1447 + }, + { + "epoch": 0.13058574198493936, + "grad_norm": 1.8517482358590107, + "learning_rate": 3.894855603895723e-06, + "loss": 0.9405, + "step": 1448 + }, + { + "epoch": 0.13067592550840962, + "grad_norm": 1.6072294877939262, + "learning_rate": 3.894668593403718e-06, + "loss": 1.1152, + "step": 1449 + }, + { + "epoch": 0.13076610903187988, + "grad_norm": 1.776462210556326, + "learning_rate": 3.8944814212481425e-06, + "loss": 1.0442, + "step": 1450 + }, + { + "epoch": 0.13085629255535014, + "grad_norm": 1.776540855738864, + "learning_rate": 3.894294087444966e-06, + "loss": 1.0162, + "step": 1451 + }, + { + "epoch": 0.1309464760788204, + "grad_norm": 4.023650110053045, + "learning_rate": 3.894106592010173e-06, + "loss": 0.9264, + "step": 1452 + }, + { + "epoch": 0.13103665960229066, + "grad_norm": 1.5863101314753794, + "learning_rate": 3.893918934959762e-06, + "loss": 0.9117, + "step": 1453 + }, + { + "epoch": 0.13112684312576092, + "grad_norm": 2.008205152776585, + "learning_rate": 3.893731116309743e-06, + "loss": 0.9744, + "step": 1454 + }, + { + "epoch": 0.13121702664923118, + "grad_norm": 1.816244138641246, + "learning_rate": 3.893543136076145e-06, + "loss": 1.0464, + "step": 1455 + }, + { + "epoch": 0.13130721017270144, + "grad_norm": 2.453428990669153, + "learning_rate": 3.893354994275006e-06, + "loss": 1.0676, + "step": 1456 + }, + { + "epoch": 0.1313973936961717, + "grad_norm": 2.6828940199751115, + "learning_rate": 3.893166690922378e-06, + "loss": 1.0079, + "step": 1457 + }, + { + "epoch": 0.13148757721964197, + "grad_norm": 2.6600103352281375, + "learning_rate": 3.892978226034329e-06, + "loss": 1.0692, + "step": 1458 + }, + { + "epoch": 0.13157776074311223, + "grad_norm": 3.1666124406153746, + "learning_rate": 3.89278959962694e-06, + "loss": 1.0909, + "step": 1459 + }, + { + "epoch": 0.1316679442665825, + "grad_norm": 2.685218818741353, + "learning_rate": 3.8926008117163056e-06, + "loss": 1.0125, + "step": 1460 + }, + { + "epoch": 0.13175812779005275, + "grad_norm": 1.9065276397134165, + "learning_rate": 3.892411862318535e-06, + "loss": 1.0264, + "step": 1461 + }, + { + "epoch": 0.131848311313523, + "grad_norm": 2.3932056335908287, + "learning_rate": 3.892222751449749e-06, + "loss": 1.0598, + "step": 1462 + }, + { + "epoch": 0.13193849483699327, + "grad_norm": 1.71986068904392, + "learning_rate": 3.892033479126084e-06, + "loss": 0.9888, + "step": 1463 + }, + { + "epoch": 0.13202867836046353, + "grad_norm": 2.0209004761333587, + "learning_rate": 3.891844045363691e-06, + "loss": 1.0188, + "step": 1464 + }, + { + "epoch": 0.13211886188393382, + "grad_norm": 1.798069697733916, + "learning_rate": 3.891654450178732e-06, + "loss": 0.9603, + "step": 1465 + }, + { + "epoch": 0.13220904540740408, + "grad_norm": 1.6150556878329725, + "learning_rate": 3.891464693587385e-06, + "loss": 1.0798, + "step": 1466 + }, + { + "epoch": 0.13229922893087434, + "grad_norm": 3.7029640388112726, + "learning_rate": 3.89127477560584e-06, + "loss": 1.078, + "step": 1467 + }, + { + "epoch": 0.1323894124543446, + "grad_norm": 1.5604677335415547, + "learning_rate": 3.891084696250304e-06, + "loss": 1.0725, + "step": 1468 + }, + { + "epoch": 0.13247959597781486, + "grad_norm": 1.8518381034376044, + "learning_rate": 3.890894455536993e-06, + "loss": 1.0764, + "step": 1469 + }, + { + "epoch": 0.13256977950128512, + "grad_norm": 3.452320402143697, + "learning_rate": 3.890704053482142e-06, + "loss": 1.088, + "step": 1470 + }, + { + "epoch": 0.13265996302475538, + "grad_norm": 1.7656550447051373, + "learning_rate": 3.890513490101995e-06, + "loss": 0.9822, + "step": 1471 + }, + { + "epoch": 0.13275014654822564, + "grad_norm": 0.6153354137953073, + "learning_rate": 3.890322765412814e-06, + "loss": 0.8444, + "step": 1472 + }, + { + "epoch": 0.1328403300716959, + "grad_norm": 1.5096212814043934, + "learning_rate": 3.890131879430871e-06, + "loss": 1.0638, + "step": 1473 + }, + { + "epoch": 0.13293051359516617, + "grad_norm": 1.6346786275341012, + "learning_rate": 3.889940832172454e-06, + "loss": 1.0631, + "step": 1474 + }, + { + "epoch": 0.13302069711863643, + "grad_norm": 1.5704442059146804, + "learning_rate": 3.889749623653864e-06, + "loss": 1.0812, + "step": 1475 + }, + { + "epoch": 0.1331108806421067, + "grad_norm": 2.184078265160653, + "learning_rate": 3.889558253891416e-06, + "loss": 0.9809, + "step": 1476 + }, + { + "epoch": 0.13320106416557695, + "grad_norm": 0.6337046515587976, + "learning_rate": 3.8893667229014385e-06, + "loss": 0.8193, + "step": 1477 + }, + { + "epoch": 0.1332912476890472, + "grad_norm": 2.2148313707009093, + "learning_rate": 3.8891750307002746e-06, + "loss": 0.9393, + "step": 1478 + }, + { + "epoch": 0.13338143121251747, + "grad_norm": 1.5809504078151297, + "learning_rate": 3.888983177304281e-06, + "loss": 0.9711, + "step": 1479 + }, + { + "epoch": 0.13347161473598773, + "grad_norm": 2.775889170014556, + "learning_rate": 3.888791162729826e-06, + "loss": 0.8185, + "step": 1480 + }, + { + "epoch": 0.133561798259458, + "grad_norm": 2.213876987565153, + "learning_rate": 3.888598986993295e-06, + "loss": 1.0606, + "step": 1481 + }, + { + "epoch": 0.13365198178292825, + "grad_norm": 1.639623063546035, + "learning_rate": 3.888406650111085e-06, + "loss": 1.0666, + "step": 1482 + }, + { + "epoch": 0.1337421653063985, + "grad_norm": 1.792139996245773, + "learning_rate": 3.888214152099607e-06, + "loss": 1.0386, + "step": 1483 + }, + { + "epoch": 0.13383234882986877, + "grad_norm": 1.958121416786926, + "learning_rate": 3.888021492975285e-06, + "loss": 1.1078, + "step": 1484 + }, + { + "epoch": 0.13392253235333904, + "grad_norm": 1.8737421266926058, + "learning_rate": 3.88782867275456e-06, + "loss": 1.0859, + "step": 1485 + }, + { + "epoch": 0.1340127158768093, + "grad_norm": 2.577688191718852, + "learning_rate": 3.8876356914538824e-06, + "loss": 0.9982, + "step": 1486 + }, + { + "epoch": 0.13410289940027956, + "grad_norm": 4.689518404962646, + "learning_rate": 3.88744254908972e-06, + "loss": 0.9912, + "step": 1487 + }, + { + "epoch": 0.13419308292374982, + "grad_norm": 3.065225439446323, + "learning_rate": 3.887249245678552e-06, + "loss": 1.0099, + "step": 1488 + }, + { + "epoch": 0.1342832664472201, + "grad_norm": 1.8029609777466682, + "learning_rate": 3.887055781236872e-06, + "loss": 1.1086, + "step": 1489 + }, + { + "epoch": 0.13437344997069037, + "grad_norm": 1.7396893573666787, + "learning_rate": 3.886862155781186e-06, + "loss": 0.9081, + "step": 1490 + }, + { + "epoch": 0.13446363349416063, + "grad_norm": 0.6864355256184999, + "learning_rate": 3.886668369328019e-06, + "loss": 0.841, + "step": 1491 + }, + { + "epoch": 0.1345538170176309, + "grad_norm": 1.9271106889862295, + "learning_rate": 3.886474421893904e-06, + "loss": 1.063, + "step": 1492 + }, + { + "epoch": 0.13464400054110115, + "grad_norm": 1.2797495266853318, + "learning_rate": 3.886280313495388e-06, + "loss": 0.959, + "step": 1493 + }, + { + "epoch": 0.1347341840645714, + "grad_norm": 1.8366582226890869, + "learning_rate": 3.886086044149035e-06, + "loss": 1.0061, + "step": 1494 + }, + { + "epoch": 0.13482436758804167, + "grad_norm": 9.598539368456667, + "learning_rate": 3.885891613871421e-06, + "loss": 1.0458, + "step": 1495 + }, + { + "epoch": 0.13491455111151193, + "grad_norm": 2.935742928156716, + "learning_rate": 3.885697022679136e-06, + "loss": 1.0797, + "step": 1496 + }, + { + "epoch": 0.1350047346349822, + "grad_norm": 2.017372615720854, + "learning_rate": 3.885502270588784e-06, + "loss": 1.0666, + "step": 1497 + }, + { + "epoch": 0.13509491815845245, + "grad_norm": 2.3536133805229795, + "learning_rate": 3.885307357616981e-06, + "loss": 1.0274, + "step": 1498 + }, + { + "epoch": 0.13518510168192271, + "grad_norm": 1.8521578189845975, + "learning_rate": 3.885112283780359e-06, + "loss": 1.0318, + "step": 1499 + }, + { + "epoch": 0.13527528520539298, + "grad_norm": 1.6154365828123303, + "learning_rate": 3.8849170490955624e-06, + "loss": 0.9274, + "step": 1500 + }, + { + "epoch": 0.13536546872886324, + "grad_norm": 2.2684266755880675, + "learning_rate": 3.88472165357925e-06, + "loss": 1.026, + "step": 1501 + }, + { + "epoch": 0.1354556522523335, + "grad_norm": 1.7941594045409086, + "learning_rate": 3.884526097248093e-06, + "loss": 0.9932, + "step": 1502 + }, + { + "epoch": 0.13554583577580376, + "grad_norm": 2.113177341014913, + "learning_rate": 3.884330380118779e-06, + "loss": 1.0539, + "step": 1503 + }, + { + "epoch": 0.13563601929927402, + "grad_norm": 1.8096195220432212, + "learning_rate": 3.884134502208007e-06, + "loss": 0.9899, + "step": 1504 + }, + { + "epoch": 0.13572620282274428, + "grad_norm": 2.14312052924579, + "learning_rate": 3.88393846353249e-06, + "loss": 1.1441, + "step": 1505 + }, + { + "epoch": 0.13581638634621454, + "grad_norm": 1.4671866029974294, + "learning_rate": 3.883742264108955e-06, + "loss": 1.0861, + "step": 1506 + }, + { + "epoch": 0.1359065698696848, + "grad_norm": 1.6797513727862947, + "learning_rate": 3.883545903954145e-06, + "loss": 1.041, + "step": 1507 + }, + { + "epoch": 0.13599675339315506, + "grad_norm": 1.541481402151025, + "learning_rate": 3.883349383084811e-06, + "loss": 1.0262, + "step": 1508 + }, + { + "epoch": 0.13608693691662532, + "grad_norm": 3.029097746668884, + "learning_rate": 3.883152701517723e-06, + "loss": 0.9568, + "step": 1509 + }, + { + "epoch": 0.13617712044009558, + "grad_norm": 2.2083184463670964, + "learning_rate": 3.882955859269664e-06, + "loss": 1.0257, + "step": 1510 + }, + { + "epoch": 0.13626730396356584, + "grad_norm": 2.5730209226502603, + "learning_rate": 3.882758856357428e-06, + "loss": 1.0412, + "step": 1511 + }, + { + "epoch": 0.1363574874870361, + "grad_norm": 1.9581553872067399, + "learning_rate": 3.882561692797824e-06, + "loss": 1.0273, + "step": 1512 + }, + { + "epoch": 0.1364476710105064, + "grad_norm": 1.8006865834407295, + "learning_rate": 3.882364368607677e-06, + "loss": 1.0371, + "step": 1513 + }, + { + "epoch": 0.13653785453397665, + "grad_norm": 2.5953591192698737, + "learning_rate": 3.8821668838038225e-06, + "loss": 1.0473, + "step": 1514 + }, + { + "epoch": 0.13662803805744692, + "grad_norm": 2.1138269985438614, + "learning_rate": 3.881969238403111e-06, + "loss": 0.9488, + "step": 1515 + }, + { + "epoch": 0.13671822158091718, + "grad_norm": 28.467205252752574, + "learning_rate": 3.881771432422408e-06, + "loss": 1.1617, + "step": 1516 + }, + { + "epoch": 0.13680840510438744, + "grad_norm": 2.0784551602010346, + "learning_rate": 3.88157346587859e-06, + "loss": 1.0873, + "step": 1517 + }, + { + "epoch": 0.1368985886278577, + "grad_norm": 7.169568058554171, + "learning_rate": 3.881375338788549e-06, + "loss": 1.0112, + "step": 1518 + }, + { + "epoch": 0.13698877215132796, + "grad_norm": 2.1477870997049915, + "learning_rate": 3.88117705116919e-06, + "loss": 1.047, + "step": 1519 + }, + { + "epoch": 0.13707895567479822, + "grad_norm": 0.794991192618953, + "learning_rate": 3.880978603037432e-06, + "loss": 0.8935, + "step": 1520 + }, + { + "epoch": 0.13716913919826848, + "grad_norm": 1.5239594959072946, + "learning_rate": 3.880779994410209e-06, + "loss": 1.0356, + "step": 1521 + }, + { + "epoch": 0.13725932272173874, + "grad_norm": 2.131510242671101, + "learning_rate": 3.880581225304466e-06, + "loss": 0.9093, + "step": 1522 + }, + { + "epoch": 0.137349506245209, + "grad_norm": 3.900587815222246, + "learning_rate": 3.880382295737163e-06, + "loss": 1.0643, + "step": 1523 + }, + { + "epoch": 0.13743968976867926, + "grad_norm": 3.1234371091795654, + "learning_rate": 3.880183205725274e-06, + "loss": 1.008, + "step": 1524 + }, + { + "epoch": 0.13752987329214952, + "grad_norm": 1.9261367191223324, + "learning_rate": 3.879983955285788e-06, + "loss": 0.9723, + "step": 1525 + }, + { + "epoch": 0.13762005681561978, + "grad_norm": 1.9522723969132145, + "learning_rate": 3.879784544435703e-06, + "loss": 1.0088, + "step": 1526 + }, + { + "epoch": 0.13771024033909005, + "grad_norm": 2.9000300636048606, + "learning_rate": 3.879584973192037e-06, + "loss": 0.9751, + "step": 1527 + }, + { + "epoch": 0.1378004238625603, + "grad_norm": 2.6373043512031833, + "learning_rate": 3.8793852415718165e-06, + "loss": 0.9642, + "step": 1528 + }, + { + "epoch": 0.13789060738603057, + "grad_norm": 2.1130208473747794, + "learning_rate": 3.879185349592085e-06, + "loss": 1.0403, + "step": 1529 + }, + { + "epoch": 0.13798079090950083, + "grad_norm": 1.8326132039158252, + "learning_rate": 3.878985297269897e-06, + "loss": 1.0543, + "step": 1530 + }, + { + "epoch": 0.1380709744329711, + "grad_norm": 2.163093763212424, + "learning_rate": 3.878785084622323e-06, + "loss": 1.0708, + "step": 1531 + }, + { + "epoch": 0.13816115795644135, + "grad_norm": 2.275942336882239, + "learning_rate": 3.878584711666447e-06, + "loss": 1.2208, + "step": 1532 + }, + { + "epoch": 0.1382513414799116, + "grad_norm": 1.7909360845077957, + "learning_rate": 3.8783841784193635e-06, + "loss": 1.0504, + "step": 1533 + }, + { + "epoch": 0.13834152500338187, + "grad_norm": 1.908926476275171, + "learning_rate": 3.8781834848981855e-06, + "loss": 1.0148, + "step": 1534 + }, + { + "epoch": 0.13843170852685213, + "grad_norm": 1.9206224407125614, + "learning_rate": 3.877982631120037e-06, + "loss": 0.9682, + "step": 1535 + }, + { + "epoch": 0.1385218920503224, + "grad_norm": 3.0492450592759113, + "learning_rate": 3.877781617102053e-06, + "loss": 1.0958, + "step": 1536 + }, + { + "epoch": 0.13861207557379268, + "grad_norm": 1.630917339094357, + "learning_rate": 3.877580442861389e-06, + "loss": 1.0703, + "step": 1537 + }, + { + "epoch": 0.13870225909726294, + "grad_norm": 2.789698851408351, + "learning_rate": 3.877379108415209e-06, + "loss": 1.0609, + "step": 1538 + }, + { + "epoch": 0.1387924426207332, + "grad_norm": 1.8503484810138293, + "learning_rate": 3.8771776137806915e-06, + "loss": 1.0615, + "step": 1539 + }, + { + "epoch": 0.13888262614420346, + "grad_norm": 2.2843560798581644, + "learning_rate": 3.8769759589750295e-06, + "loss": 0.9783, + "step": 1540 + }, + { + "epoch": 0.13897280966767372, + "grad_norm": 1.5422547052305153, + "learning_rate": 3.876774144015429e-06, + "loss": 1.0442, + "step": 1541 + }, + { + "epoch": 0.13906299319114399, + "grad_norm": 5.394499251749324, + "learning_rate": 3.87657216891911e-06, + "loss": 1.0726, + "step": 1542 + }, + { + "epoch": 0.13915317671461425, + "grad_norm": 1.9403901232912948, + "learning_rate": 3.876370033703307e-06, + "loss": 1.0322, + "step": 1543 + }, + { + "epoch": 0.1392433602380845, + "grad_norm": 1.3869297350637368, + "learning_rate": 3.876167738385265e-06, + "loss": 0.9347, + "step": 1544 + }, + { + "epoch": 0.13933354376155477, + "grad_norm": 2.322930089344423, + "learning_rate": 3.875965282982247e-06, + "loss": 1.0824, + "step": 1545 + }, + { + "epoch": 0.13942372728502503, + "grad_norm": 1.760213407561163, + "learning_rate": 3.875762667511528e-06, + "loss": 1.0512, + "step": 1546 + }, + { + "epoch": 0.1395139108084953, + "grad_norm": 1.4681594148242336, + "learning_rate": 3.875559891990394e-06, + "loss": 1.0426, + "step": 1547 + }, + { + "epoch": 0.13960409433196555, + "grad_norm": 2.162884002284086, + "learning_rate": 3.875356956436149e-06, + "loss": 1.0052, + "step": 1548 + }, + { + "epoch": 0.1396942778554358, + "grad_norm": 1.8610185683982892, + "learning_rate": 3.875153860866108e-06, + "loss": 1.0189, + "step": 1549 + }, + { + "epoch": 0.13978446137890607, + "grad_norm": 3.501196929082719, + "learning_rate": 3.8749506052976e-06, + "loss": 1.0837, + "step": 1550 + }, + { + "epoch": 0.13987464490237633, + "grad_norm": 1.5405763258605534, + "learning_rate": 3.874747189747968e-06, + "loss": 1.0636, + "step": 1551 + }, + { + "epoch": 0.1399648284258466, + "grad_norm": 1.7243486943991633, + "learning_rate": 3.874543614234568e-06, + "loss": 1.0441, + "step": 1552 + }, + { + "epoch": 0.14005501194931685, + "grad_norm": 1.6652638571677636, + "learning_rate": 3.874339878774771e-06, + "loss": 1.0159, + "step": 1553 + }, + { + "epoch": 0.14014519547278712, + "grad_norm": 2.8610469745084997, + "learning_rate": 3.874135983385961e-06, + "loss": 1.0404, + "step": 1554 + }, + { + "epoch": 0.14023537899625738, + "grad_norm": 1.9068898237234229, + "learning_rate": 3.873931928085535e-06, + "loss": 1.037, + "step": 1555 + }, + { + "epoch": 0.14032556251972764, + "grad_norm": 2.2515127606692125, + "learning_rate": 3.873727712890904e-06, + "loss": 1.1227, + "step": 1556 + }, + { + "epoch": 0.1404157460431979, + "grad_norm": 1.5554398293060239, + "learning_rate": 3.873523337819493e-06, + "loss": 1.0092, + "step": 1557 + }, + { + "epoch": 0.14050592956666816, + "grad_norm": 1.5533947961678096, + "learning_rate": 3.873318802888739e-06, + "loss": 0.9653, + "step": 1558 + }, + { + "epoch": 0.14059611309013842, + "grad_norm": 0.5680949064453146, + "learning_rate": 3.873114108116097e-06, + "loss": 0.7688, + "step": 1559 + }, + { + "epoch": 0.14068629661360868, + "grad_norm": 1.9306744356701606, + "learning_rate": 3.872909253519031e-06, + "loss": 1.1297, + "step": 1560 + }, + { + "epoch": 0.14077648013707897, + "grad_norm": 2.22548320109657, + "learning_rate": 3.8727042391150195e-06, + "loss": 1.1108, + "step": 1561 + }, + { + "epoch": 0.14086666366054923, + "grad_norm": 1.8642039063914393, + "learning_rate": 3.872499064921556e-06, + "loss": 0.9735, + "step": 1562 + }, + { + "epoch": 0.1409568471840195, + "grad_norm": 2.127388901639369, + "learning_rate": 3.872293730956149e-06, + "loss": 1.0532, + "step": 1563 + }, + { + "epoch": 0.14104703070748975, + "grad_norm": 2.799907468220499, + "learning_rate": 3.872088237236316e-06, + "loss": 1.0334, + "step": 1564 + }, + { + "epoch": 0.14113721423096, + "grad_norm": 1.9231980894504521, + "learning_rate": 3.871882583779592e-06, + "loss": 0.9393, + "step": 1565 + }, + { + "epoch": 0.14122739775443027, + "grad_norm": 1.5639377845331255, + "learning_rate": 3.871676770603525e-06, + "loss": 1.0399, + "step": 1566 + }, + { + "epoch": 0.14131758127790053, + "grad_norm": 3.3730046060340118, + "learning_rate": 3.871470797725676e-06, + "loss": 1.032, + "step": 1567 + }, + { + "epoch": 0.1414077648013708, + "grad_norm": 1.8623103205211085, + "learning_rate": 3.8712646651636185e-06, + "loss": 1.1125, + "step": 1568 + }, + { + "epoch": 0.14149794832484106, + "grad_norm": 1.5515175712050233, + "learning_rate": 3.871058372934942e-06, + "loss": 1.035, + "step": 1569 + }, + { + "epoch": 0.14158813184831132, + "grad_norm": 1.7056461537280512, + "learning_rate": 3.8708519210572485e-06, + "loss": 0.8713, + "step": 1570 + }, + { + "epoch": 0.14167831537178158, + "grad_norm": 1.9880760701552336, + "learning_rate": 3.870645309548153e-06, + "loss": 0.977, + "step": 1571 + }, + { + "epoch": 0.14176849889525184, + "grad_norm": 0.6028954983413314, + "learning_rate": 3.870438538425284e-06, + "loss": 0.7821, + "step": 1572 + }, + { + "epoch": 0.1418586824187221, + "grad_norm": 2.78405740041862, + "learning_rate": 3.870231607706287e-06, + "loss": 0.9869, + "step": 1573 + }, + { + "epoch": 0.14194886594219236, + "grad_norm": 3.470446884665528, + "learning_rate": 3.870024517408817e-06, + "loss": 1.045, + "step": 1574 + }, + { + "epoch": 0.14203904946566262, + "grad_norm": 1.8823559472642064, + "learning_rate": 3.8698172675505425e-06, + "loss": 0.9612, + "step": 1575 + }, + { + "epoch": 0.14212923298913288, + "grad_norm": 0.7469346665047829, + "learning_rate": 3.86960985814915e-06, + "loss": 0.7965, + "step": 1576 + }, + { + "epoch": 0.14221941651260314, + "grad_norm": 2.642146558122927, + "learning_rate": 3.869402289222335e-06, + "loss": 0.9672, + "step": 1577 + }, + { + "epoch": 0.1423096000360734, + "grad_norm": 2.3606552793516693, + "learning_rate": 3.869194560787808e-06, + "loss": 1.0189, + "step": 1578 + }, + { + "epoch": 0.14239978355954366, + "grad_norm": 1.8176201814483077, + "learning_rate": 3.868986672863296e-06, + "loss": 0.9689, + "step": 1579 + }, + { + "epoch": 0.14248996708301392, + "grad_norm": 1.5837017517448209, + "learning_rate": 3.868778625466535e-06, + "loss": 1.1369, + "step": 1580 + }, + { + "epoch": 0.14258015060648419, + "grad_norm": 2.004781015285831, + "learning_rate": 3.868570418615278e-06, + "loss": 0.9581, + "step": 1581 + }, + { + "epoch": 0.14267033412995445, + "grad_norm": 0.6920551121779986, + "learning_rate": 3.8683620523272885e-06, + "loss": 0.8048, + "step": 1582 + }, + { + "epoch": 0.1427605176534247, + "grad_norm": 2.2077896840508755, + "learning_rate": 3.8681535266203464e-06, + "loss": 1.1434, + "step": 1583 + }, + { + "epoch": 0.14285070117689497, + "grad_norm": 2.831344879110083, + "learning_rate": 3.867944841512246e-06, + "loss": 1.064, + "step": 1584 + }, + { + "epoch": 0.14294088470036526, + "grad_norm": 2.144038465126362, + "learning_rate": 3.867735997020791e-06, + "loss": 1.0442, + "step": 1585 + }, + { + "epoch": 0.14303106822383552, + "grad_norm": 1.5987868657160291, + "learning_rate": 3.867526993163802e-06, + "loss": 0.9441, + "step": 1586 + }, + { + "epoch": 0.14312125174730578, + "grad_norm": 1.6703592720298979, + "learning_rate": 3.867317829959113e-06, + "loss": 0.91, + "step": 1587 + }, + { + "epoch": 0.14321143527077604, + "grad_norm": 1.897907638388752, + "learning_rate": 3.8671085074245704e-06, + "loss": 0.8308, + "step": 1588 + }, + { + "epoch": 0.1433016187942463, + "grad_norm": 1.8074234869500254, + "learning_rate": 3.866899025578035e-06, + "loss": 1.1031, + "step": 1589 + }, + { + "epoch": 0.14339180231771656, + "grad_norm": 4.116417481658145, + "learning_rate": 3.86668938443738e-06, + "loss": 1.0037, + "step": 1590 + }, + { + "epoch": 0.14348198584118682, + "grad_norm": 3.291075464264959, + "learning_rate": 3.866479584020495e-06, + "loss": 1.0506, + "step": 1591 + }, + { + "epoch": 0.14357216936465708, + "grad_norm": 1.907718905465826, + "learning_rate": 3.866269624345279e-06, + "loss": 1.0696, + "step": 1592 + }, + { + "epoch": 0.14366235288812734, + "grad_norm": 1.9777582946572545, + "learning_rate": 3.866059505429649e-06, + "loss": 1.1159, + "step": 1593 + }, + { + "epoch": 0.1437525364115976, + "grad_norm": 0.6586490331866968, + "learning_rate": 3.865849227291532e-06, + "loss": 0.8323, + "step": 1594 + }, + { + "epoch": 0.14384271993506786, + "grad_norm": 2.634195570078676, + "learning_rate": 3.865638789948872e-06, + "loss": 1.032, + "step": 1595 + }, + { + "epoch": 0.14393290345853813, + "grad_norm": 3.0878327163391215, + "learning_rate": 3.865428193419622e-06, + "loss": 0.99, + "step": 1596 + }, + { + "epoch": 0.14402308698200839, + "grad_norm": 2.2209229909847763, + "learning_rate": 3.865217437721753e-06, + "loss": 1.0225, + "step": 1597 + }, + { + "epoch": 0.14411327050547865, + "grad_norm": 3.160401889981901, + "learning_rate": 3.865006522873249e-06, + "loss": 0.9676, + "step": 1598 + }, + { + "epoch": 0.1442034540289489, + "grad_norm": 1.9147005224493268, + "learning_rate": 3.864795448892103e-06, + "loss": 1.03, + "step": 1599 + }, + { + "epoch": 0.14429363755241917, + "grad_norm": 1.958982853684448, + "learning_rate": 3.864584215796327e-06, + "loss": 1.0825, + "step": 1600 + }, + { + "epoch": 0.14438382107588943, + "grad_norm": 2.555822091324042, + "learning_rate": 3.8643728236039455e-06, + "loss": 1.059, + "step": 1601 + }, + { + "epoch": 0.1444740045993597, + "grad_norm": 5.2375922948311855, + "learning_rate": 3.864161272332994e-06, + "loss": 1.0192, + "step": 1602 + }, + { + "epoch": 0.14456418812282995, + "grad_norm": 1.5847434072650743, + "learning_rate": 3.863949562001524e-06, + "loss": 1.078, + "step": 1603 + }, + { + "epoch": 0.1446543716463002, + "grad_norm": 1.7160845809756873, + "learning_rate": 3.8637376926276005e-06, + "loss": 1.0098, + "step": 1604 + }, + { + "epoch": 0.14474455516977047, + "grad_norm": 2.0236692553617637, + "learning_rate": 3.8635256642293e-06, + "loss": 1.0429, + "step": 1605 + }, + { + "epoch": 0.14483473869324073, + "grad_norm": 1.7988983703880725, + "learning_rate": 3.863313476824714e-06, + "loss": 1.016, + "step": 1606 + }, + { + "epoch": 0.144924922216711, + "grad_norm": 2.3972862040130742, + "learning_rate": 3.863101130431948e-06, + "loss": 1.0196, + "step": 1607 + }, + { + "epoch": 0.14501510574018128, + "grad_norm": 1.7448150253630166, + "learning_rate": 3.862888625069121e-06, + "loss": 1.0772, + "step": 1608 + }, + { + "epoch": 0.14510528926365154, + "grad_norm": 1.422381132821163, + "learning_rate": 3.8626759607543645e-06, + "loss": 0.9983, + "step": 1609 + }, + { + "epoch": 0.1451954727871218, + "grad_norm": 2.0675606342543764, + "learning_rate": 3.862463137505825e-06, + "loss": 0.9911, + "step": 1610 + }, + { + "epoch": 0.14528565631059207, + "grad_norm": 1.7644004964253124, + "learning_rate": 3.862250155341659e-06, + "loss": 0.9971, + "step": 1611 + }, + { + "epoch": 0.14537583983406233, + "grad_norm": 1.748617920758346, + "learning_rate": 3.862037014280043e-06, + "loss": 1.1619, + "step": 1612 + }, + { + "epoch": 0.1454660233575326, + "grad_norm": 3.80069091940449, + "learning_rate": 3.861823714339162e-06, + "loss": 1.024, + "step": 1613 + }, + { + "epoch": 0.14555620688100285, + "grad_norm": 2.3274238093737765, + "learning_rate": 3.861610255537215e-06, + "loss": 1.0438, + "step": 1614 + }, + { + "epoch": 0.1456463904044731, + "grad_norm": 2.2038905119173426, + "learning_rate": 3.8613966378924165e-06, + "loss": 1.0611, + "step": 1615 + }, + { + "epoch": 0.14573657392794337, + "grad_norm": 1.846851455874047, + "learning_rate": 3.861182861422993e-06, + "loss": 0.9805, + "step": 1616 + }, + { + "epoch": 0.14582675745141363, + "grad_norm": 1.5286010874276073, + "learning_rate": 3.860968926147185e-06, + "loss": 1.1048, + "step": 1617 + }, + { + "epoch": 0.1459169409748839, + "grad_norm": 1.79635320633748, + "learning_rate": 3.860754832083247e-06, + "loss": 0.8673, + "step": 1618 + }, + { + "epoch": 0.14600712449835415, + "grad_norm": 1.8598024534101596, + "learning_rate": 3.8605405792494475e-06, + "loss": 1.109, + "step": 1619 + }, + { + "epoch": 0.1460973080218244, + "grad_norm": 4.832004013539851, + "learning_rate": 3.860326167664066e-06, + "loss": 1.0489, + "step": 1620 + }, + { + "epoch": 0.14618749154529467, + "grad_norm": 1.588647340688736, + "learning_rate": 3.860111597345399e-06, + "loss": 1.0175, + "step": 1621 + }, + { + "epoch": 0.14627767506876493, + "grad_norm": 2.187622284877237, + "learning_rate": 3.859896868311753e-06, + "loss": 0.8864, + "step": 1622 + }, + { + "epoch": 0.1463678585922352, + "grad_norm": 1.6262752957322923, + "learning_rate": 3.859681980581452e-06, + "loss": 1.1215, + "step": 1623 + }, + { + "epoch": 0.14645804211570546, + "grad_norm": 1.751728634607772, + "learning_rate": 3.859466934172829e-06, + "loss": 1.0086, + "step": 1624 + }, + { + "epoch": 0.14654822563917572, + "grad_norm": 1.5022318448718461, + "learning_rate": 3.859251729104235e-06, + "loss": 0.9656, + "step": 1625 + }, + { + "epoch": 0.14663840916264598, + "grad_norm": 1.4785791582493892, + "learning_rate": 3.859036365394031e-06, + "loss": 0.9808, + "step": 1626 + }, + { + "epoch": 0.14672859268611624, + "grad_norm": 5.482576863401949, + "learning_rate": 3.858820843060594e-06, + "loss": 1.0171, + "step": 1627 + }, + { + "epoch": 0.1468187762095865, + "grad_norm": 1.6228589110897302, + "learning_rate": 3.858605162122314e-06, + "loss": 1.0068, + "step": 1628 + }, + { + "epoch": 0.14690895973305676, + "grad_norm": 1.7910802532087857, + "learning_rate": 3.858389322597592e-06, + "loss": 0.9835, + "step": 1629 + }, + { + "epoch": 0.14699914325652702, + "grad_norm": 1.6678568961095759, + "learning_rate": 3.858173324504847e-06, + "loss": 1.0433, + "step": 1630 + }, + { + "epoch": 0.14708932677999728, + "grad_norm": 1.6224981268591274, + "learning_rate": 3.857957167862508e-06, + "loss": 0.9946, + "step": 1631 + }, + { + "epoch": 0.14717951030346757, + "grad_norm": 0.670286896579002, + "learning_rate": 3.857740852689018e-06, + "loss": 0.8172, + "step": 1632 + }, + { + "epoch": 0.14726969382693783, + "grad_norm": 1.9783602895509749, + "learning_rate": 3.857524379002835e-06, + "loss": 1.0362, + "step": 1633 + }, + { + "epoch": 0.1473598773504081, + "grad_norm": 2.078253491093468, + "learning_rate": 3.85730774682243e-06, + "loss": 1.0508, + "step": 1634 + }, + { + "epoch": 0.14745006087387835, + "grad_norm": 1.5328441029439535, + "learning_rate": 3.8570909561662875e-06, + "loss": 1.0917, + "step": 1635 + }, + { + "epoch": 0.1475402443973486, + "grad_norm": 1.884013066240425, + "learning_rate": 3.8568740070529045e-06, + "loss": 1.0964, + "step": 1636 + }, + { + "epoch": 0.14763042792081887, + "grad_norm": 1.8352987923443926, + "learning_rate": 3.856656899500792e-06, + "loss": 0.9971, + "step": 1637 + }, + { + "epoch": 0.14772061144428913, + "grad_norm": 1.6692872029264112, + "learning_rate": 3.856439633528476e-06, + "loss": 0.9598, + "step": 1638 + }, + { + "epoch": 0.1478107949677594, + "grad_norm": 2.072669073554787, + "learning_rate": 3.856222209154494e-06, + "loss": 1.0818, + "step": 1639 + }, + { + "epoch": 0.14790097849122966, + "grad_norm": 1.945147894210819, + "learning_rate": 3.856004626397397e-06, + "loss": 1.0513, + "step": 1640 + }, + { + "epoch": 0.14799116201469992, + "grad_norm": 1.914909455671257, + "learning_rate": 3.855786885275753e-06, + "loss": 1.0033, + "step": 1641 + }, + { + "epoch": 0.14808134553817018, + "grad_norm": 2.014072382083677, + "learning_rate": 3.855568985808138e-06, + "loss": 0.9452, + "step": 1642 + }, + { + "epoch": 0.14817152906164044, + "grad_norm": 2.0545375754772293, + "learning_rate": 3.855350928013145e-06, + "loss": 0.997, + "step": 1643 + }, + { + "epoch": 0.1482617125851107, + "grad_norm": 2.018010583341311, + "learning_rate": 3.8551327119093825e-06, + "loss": 1.0318, + "step": 1644 + }, + { + "epoch": 0.14835189610858096, + "grad_norm": 2.206341066723357, + "learning_rate": 3.854914337515467e-06, + "loss": 0.9607, + "step": 1645 + }, + { + "epoch": 0.14844207963205122, + "grad_norm": 1.6238350727445634, + "learning_rate": 3.8546958048500324e-06, + "loss": 0.921, + "step": 1646 + }, + { + "epoch": 0.14853226315552148, + "grad_norm": 0.6827530186843767, + "learning_rate": 3.854477113931725e-06, + "loss": 0.8695, + "step": 1647 + }, + { + "epoch": 0.14862244667899174, + "grad_norm": 1.883289197669037, + "learning_rate": 3.854258264779205e-06, + "loss": 1.0474, + "step": 1648 + }, + { + "epoch": 0.148712630202462, + "grad_norm": 1.747674964255489, + "learning_rate": 3.854039257411145e-06, + "loss": 1.0058, + "step": 1649 + }, + { + "epoch": 0.14880281372593226, + "grad_norm": 1.7786972760812798, + "learning_rate": 3.853820091846232e-06, + "loss": 0.9818, + "step": 1650 + }, + { + "epoch": 0.14889299724940253, + "grad_norm": 2.118800600961143, + "learning_rate": 3.853600768103169e-06, + "loss": 1.0291, + "step": 1651 + }, + { + "epoch": 0.1489831807728728, + "grad_norm": 2.4064347827855315, + "learning_rate": 3.853381286200667e-06, + "loss": 1.0148, + "step": 1652 + }, + { + "epoch": 0.14907336429634305, + "grad_norm": 1.778672210214794, + "learning_rate": 3.853161646157453e-06, + "loss": 1.1189, + "step": 1653 + }, + { + "epoch": 0.1491635478198133, + "grad_norm": 2.294896629681623, + "learning_rate": 3.852941847992269e-06, + "loss": 0.9932, + "step": 1654 + }, + { + "epoch": 0.14925373134328357, + "grad_norm": 2.3270330748656227, + "learning_rate": 3.852721891723871e-06, + "loss": 1.042, + "step": 1655 + }, + { + "epoch": 0.14934391486675386, + "grad_norm": 1.9419058266345492, + "learning_rate": 3.852501777371025e-06, + "loss": 1.0117, + "step": 1656 + }, + { + "epoch": 0.14943409839022412, + "grad_norm": 1.8002984647198792, + "learning_rate": 3.8522815049525125e-06, + "loss": 1.0642, + "step": 1657 + }, + { + "epoch": 0.14952428191369438, + "grad_norm": 1.9354669918201062, + "learning_rate": 3.852061074487129e-06, + "loss": 1.0813, + "step": 1658 + }, + { + "epoch": 0.14961446543716464, + "grad_norm": 1.7095651914802483, + "learning_rate": 3.851840485993682e-06, + "loss": 0.9337, + "step": 1659 + }, + { + "epoch": 0.1497046489606349, + "grad_norm": 1.6597319655440546, + "learning_rate": 3.851619739490994e-06, + "loss": 1.0394, + "step": 1660 + }, + { + "epoch": 0.14979483248410516, + "grad_norm": 4.97073036890864, + "learning_rate": 3.8513988349978996e-06, + "loss": 1.0614, + "step": 1661 + }, + { + "epoch": 0.14988501600757542, + "grad_norm": 4.508891117096859, + "learning_rate": 3.851177772533249e-06, + "loss": 1.0519, + "step": 1662 + }, + { + "epoch": 0.14997519953104568, + "grad_norm": 1.948144345228762, + "learning_rate": 3.850956552115903e-06, + "loss": 1.2033, + "step": 1663 + }, + { + "epoch": 0.15006538305451594, + "grad_norm": 0.6945857418211531, + "learning_rate": 3.850735173764738e-06, + "loss": 0.8204, + "step": 1664 + }, + { + "epoch": 0.1501555665779862, + "grad_norm": 1.9340973785800761, + "learning_rate": 3.850513637498642e-06, + "loss": 1.0511, + "step": 1665 + }, + { + "epoch": 0.15024575010145647, + "grad_norm": 2.4086866055882883, + "learning_rate": 3.850291943336521e-06, + "loss": 1.0455, + "step": 1666 + }, + { + "epoch": 0.15033593362492673, + "grad_norm": 1.7384283453607559, + "learning_rate": 3.850070091297287e-06, + "loss": 1.0834, + "step": 1667 + }, + { + "epoch": 0.150426117148397, + "grad_norm": 0.7143698200028015, + "learning_rate": 3.8498480813998735e-06, + "loss": 0.9117, + "step": 1668 + }, + { + "epoch": 0.15051630067186725, + "grad_norm": 1.6114163091399767, + "learning_rate": 3.84962591366322e-06, + "loss": 1.0266, + "step": 1669 + }, + { + "epoch": 0.1506064841953375, + "grad_norm": 2.031437204977632, + "learning_rate": 3.8494035881062855e-06, + "loss": 1.0785, + "step": 1670 + }, + { + "epoch": 0.15069666771880777, + "grad_norm": 1.9108978267366061, + "learning_rate": 3.84918110474804e-06, + "loss": 1.0088, + "step": 1671 + }, + { + "epoch": 0.15078685124227803, + "grad_norm": 1.5275981720439225, + "learning_rate": 3.8489584636074655e-06, + "loss": 1.0134, + "step": 1672 + }, + { + "epoch": 0.1508770347657483, + "grad_norm": 0.7092769639083641, + "learning_rate": 3.848735664703561e-06, + "loss": 0.803, + "step": 1673 + }, + { + "epoch": 0.15096721828921855, + "grad_norm": 1.3656131071712014, + "learning_rate": 3.8485127080553346e-06, + "loss": 1.0241, + "step": 1674 + }, + { + "epoch": 0.1510574018126888, + "grad_norm": 2.034368966792429, + "learning_rate": 3.8482895936818115e-06, + "loss": 1.0283, + "step": 1675 + }, + { + "epoch": 0.15114758533615907, + "grad_norm": 1.727488273878917, + "learning_rate": 3.848066321602029e-06, + "loss": 1.046, + "step": 1676 + }, + { + "epoch": 0.15123776885962933, + "grad_norm": 1.9007830787960338, + "learning_rate": 3.847842891835038e-06, + "loss": 0.9947, + "step": 1677 + }, + { + "epoch": 0.1513279523830996, + "grad_norm": 0.6292980939970517, + "learning_rate": 3.847619304399902e-06, + "loss": 0.8187, + "step": 1678 + }, + { + "epoch": 0.15141813590656986, + "grad_norm": 2.048982190793389, + "learning_rate": 3.8473955593157e-06, + "loss": 1.0613, + "step": 1679 + }, + { + "epoch": 0.15150831943004014, + "grad_norm": 1.899436844637633, + "learning_rate": 3.847171656601522e-06, + "loss": 0.995, + "step": 1680 + }, + { + "epoch": 0.1515985029535104, + "grad_norm": 1.6977169722938292, + "learning_rate": 3.846947596276473e-06, + "loss": 1.0575, + "step": 1681 + }, + { + "epoch": 0.15168868647698067, + "grad_norm": 3.9778182589450073, + "learning_rate": 3.846723378359672e-06, + "loss": 1.0785, + "step": 1682 + }, + { + "epoch": 0.15177887000045093, + "grad_norm": 1.9746820000669854, + "learning_rate": 3.846499002870249e-06, + "loss": 1.0326, + "step": 1683 + }, + { + "epoch": 0.1518690535239212, + "grad_norm": 1.6664563364191114, + "learning_rate": 3.846274469827349e-06, + "loss": 1.0457, + "step": 1684 + }, + { + "epoch": 0.15195923704739145, + "grad_norm": 2.6201777897273066, + "learning_rate": 3.846049779250132e-06, + "loss": 1.0371, + "step": 1685 + }, + { + "epoch": 0.1520494205708617, + "grad_norm": 0.7802295505370324, + "learning_rate": 3.845824931157769e-06, + "loss": 0.8616, + "step": 1686 + }, + { + "epoch": 0.15213960409433197, + "grad_norm": 3.158788887391404, + "learning_rate": 3.845599925569444e-06, + "loss": 1.0158, + "step": 1687 + }, + { + "epoch": 0.15222978761780223, + "grad_norm": 2.1513781588391945, + "learning_rate": 3.845374762504357e-06, + "loss": 0.9922, + "step": 1688 + }, + { + "epoch": 0.1523199711412725, + "grad_norm": 1.794765461446072, + "learning_rate": 3.8451494419817204e-06, + "loss": 1.0765, + "step": 1689 + }, + { + "epoch": 0.15241015466474275, + "grad_norm": 2.385423874045954, + "learning_rate": 3.8449239640207594e-06, + "loss": 0.9609, + "step": 1690 + }, + { + "epoch": 0.152500338188213, + "grad_norm": 2.521291948535948, + "learning_rate": 3.844698328640713e-06, + "loss": 1.0728, + "step": 1691 + }, + { + "epoch": 0.15259052171168327, + "grad_norm": 2.1583423138091655, + "learning_rate": 3.844472535860833e-06, + "loss": 1.0496, + "step": 1692 + }, + { + "epoch": 0.15268070523515354, + "grad_norm": 2.0602955452470035, + "learning_rate": 3.8442465857003864e-06, + "loss": 1.0807, + "step": 1693 + }, + { + "epoch": 0.1527708887586238, + "grad_norm": 1.4624637240614062, + "learning_rate": 3.844020478178653e-06, + "loss": 0.9741, + "step": 1694 + }, + { + "epoch": 0.15286107228209406, + "grad_norm": 1.9520256305364225, + "learning_rate": 3.843794213314923e-06, + "loss": 1.0566, + "step": 1695 + }, + { + "epoch": 0.15295125580556432, + "grad_norm": 2.6540880268449922, + "learning_rate": 3.843567791128505e-06, + "loss": 1.0639, + "step": 1696 + }, + { + "epoch": 0.15304143932903458, + "grad_norm": 5.104190354065199, + "learning_rate": 3.843341211638717e-06, + "loss": 1.0181, + "step": 1697 + }, + { + "epoch": 0.15313162285250484, + "grad_norm": 2.351362454342825, + "learning_rate": 3.843114474864894e-06, + "loss": 0.9914, + "step": 1698 + }, + { + "epoch": 0.1532218063759751, + "grad_norm": 1.4865036007413805, + "learning_rate": 3.84288758082638e-06, + "loss": 1.0314, + "step": 1699 + }, + { + "epoch": 0.15331198989944536, + "grad_norm": 1.4055782409055242, + "learning_rate": 3.842660529542536e-06, + "loss": 1.1209, + "step": 1700 + }, + { + "epoch": 0.15340217342291562, + "grad_norm": 1.5207011370054273, + "learning_rate": 3.842433321032736e-06, + "loss": 1.0461, + "step": 1701 + }, + { + "epoch": 0.15349235694638588, + "grad_norm": 1.4678960712138343, + "learning_rate": 3.842205955316365e-06, + "loss": 0.9557, + "step": 1702 + }, + { + "epoch": 0.15358254046985614, + "grad_norm": 2.0040605571147925, + "learning_rate": 3.8419784324128256e-06, + "loss": 0.9361, + "step": 1703 + }, + { + "epoch": 0.15367272399332643, + "grad_norm": 2.2781600928514414, + "learning_rate": 3.841750752341529e-06, + "loss": 0.9931, + "step": 1704 + }, + { + "epoch": 0.1537629075167967, + "grad_norm": 1.6828742027482002, + "learning_rate": 3.841522915121902e-06, + "loss": 1.034, + "step": 1705 + }, + { + "epoch": 0.15385309104026695, + "grad_norm": 1.5975725296581407, + "learning_rate": 3.841294920773387e-06, + "loss": 0.9543, + "step": 1706 + }, + { + "epoch": 0.15394327456373721, + "grad_norm": 2.5527567902611894, + "learning_rate": 3.841066769315436e-06, + "loss": 1.0427, + "step": 1707 + }, + { + "epoch": 0.15403345808720748, + "grad_norm": 1.766368540614841, + "learning_rate": 3.840838460767518e-06, + "loss": 1.0401, + "step": 1708 + }, + { + "epoch": 0.15412364161067774, + "grad_norm": 1.8574124405197106, + "learning_rate": 3.840609995149111e-06, + "loss": 1.0014, + "step": 1709 + }, + { + "epoch": 0.154213825134148, + "grad_norm": 2.2269878851120386, + "learning_rate": 3.84038137247971e-06, + "loss": 1.0333, + "step": 1710 + }, + { + "epoch": 0.15430400865761826, + "grad_norm": 1.930720435068662, + "learning_rate": 3.840152592778823e-06, + "loss": 1.0128, + "step": 1711 + }, + { + "epoch": 0.15439419218108852, + "grad_norm": 1.456461040610899, + "learning_rate": 3.83992365606597e-06, + "loss": 1.0746, + "step": 1712 + }, + { + "epoch": 0.15448437570455878, + "grad_norm": 1.7187748300319383, + "learning_rate": 3.8396945623606855e-06, + "loss": 0.9887, + "step": 1713 + }, + { + "epoch": 0.15457455922802904, + "grad_norm": 1.658651230723893, + "learning_rate": 3.8394653116825174e-06, + "loss": 0.986, + "step": 1714 + }, + { + "epoch": 0.1546647427514993, + "grad_norm": 1.631907819332434, + "learning_rate": 3.839235904051026e-06, + "loss": 1.0207, + "step": 1715 + }, + { + "epoch": 0.15475492627496956, + "grad_norm": 1.9282708160808029, + "learning_rate": 3.8390063394857855e-06, + "loss": 1.0039, + "step": 1716 + }, + { + "epoch": 0.15484510979843982, + "grad_norm": 1.554875348114777, + "learning_rate": 3.838776618006385e-06, + "loss": 1.1132, + "step": 1717 + }, + { + "epoch": 0.15493529332191008, + "grad_norm": 1.6688879548959252, + "learning_rate": 3.838546739632423e-06, + "loss": 1.0379, + "step": 1718 + }, + { + "epoch": 0.15502547684538034, + "grad_norm": 1.6388525698385759, + "learning_rate": 3.838316704383517e-06, + "loss": 1.0785, + "step": 1719 + }, + { + "epoch": 0.1551156603688506, + "grad_norm": 1.6635962894320353, + "learning_rate": 3.838086512279292e-06, + "loss": 1.0303, + "step": 1720 + }, + { + "epoch": 0.15520584389232087, + "grad_norm": 2.9758577088107523, + "learning_rate": 3.837856163339391e-06, + "loss": 1.0213, + "step": 1721 + }, + { + "epoch": 0.15529602741579113, + "grad_norm": 1.6154362876368917, + "learning_rate": 3.837625657583469e-06, + "loss": 0.9568, + "step": 1722 + }, + { + "epoch": 0.1553862109392614, + "grad_norm": 2.9869830334656933, + "learning_rate": 3.837394995031193e-06, + "loss": 0.9877, + "step": 1723 + }, + { + "epoch": 0.15547639446273165, + "grad_norm": 3.054370444153693, + "learning_rate": 3.837164175702245e-06, + "loss": 1.0254, + "step": 1724 + }, + { + "epoch": 0.1555665779862019, + "grad_norm": 2.3614442353458083, + "learning_rate": 3.836933199616319e-06, + "loss": 0.9426, + "step": 1725 + }, + { + "epoch": 0.15565676150967217, + "grad_norm": 1.4499745004976832, + "learning_rate": 3.836702066793124e-06, + "loss": 1.0986, + "step": 1726 + }, + { + "epoch": 0.15574694503314246, + "grad_norm": 1.635639790988116, + "learning_rate": 3.836470777252381e-06, + "loss": 1.0897, + "step": 1727 + }, + { + "epoch": 0.15583712855661272, + "grad_norm": 2.495340869529705, + "learning_rate": 3.836239331013825e-06, + "loss": 1.0458, + "step": 1728 + }, + { + "epoch": 0.15592731208008298, + "grad_norm": 2.0384168516169137, + "learning_rate": 3.836007728097205e-06, + "loss": 0.9878, + "step": 1729 + }, + { + "epoch": 0.15601749560355324, + "grad_norm": 2.397843874553683, + "learning_rate": 3.835775968522282e-06, + "loss": 0.9809, + "step": 1730 + }, + { + "epoch": 0.1561076791270235, + "grad_norm": 1.9898447061453275, + "learning_rate": 3.83554405230883e-06, + "loss": 1.0512, + "step": 1731 + }, + { + "epoch": 0.15619786265049376, + "grad_norm": 7.188541502750581, + "learning_rate": 3.835311979476639e-06, + "loss": 1.0106, + "step": 1732 + }, + { + "epoch": 0.15628804617396402, + "grad_norm": 2.178243179070686, + "learning_rate": 3.83507975004551e-06, + "loss": 0.923, + "step": 1733 + }, + { + "epoch": 0.15637822969743428, + "grad_norm": 1.5983395782338097, + "learning_rate": 3.834847364035258e-06, + "loss": 1.0057, + "step": 1734 + }, + { + "epoch": 0.15646841322090455, + "grad_norm": 2.49272986940889, + "learning_rate": 3.834614821465712e-06, + "loss": 0.9661, + "step": 1735 + }, + { + "epoch": 0.1565585967443748, + "grad_norm": 4.474242136624207, + "learning_rate": 3.834382122356713e-06, + "loss": 1.0637, + "step": 1736 + }, + { + "epoch": 0.15664878026784507, + "grad_norm": 1.4400990252089738, + "learning_rate": 3.834149266728117e-06, + "loss": 0.982, + "step": 1737 + }, + { + "epoch": 0.15673896379131533, + "grad_norm": 0.6904950792135903, + "learning_rate": 3.833916254599792e-06, + "loss": 0.8582, + "step": 1738 + }, + { + "epoch": 0.1568291473147856, + "grad_norm": 0.6204575212971285, + "learning_rate": 3.83368308599162e-06, + "loss": 0.8039, + "step": 1739 + }, + { + "epoch": 0.15691933083825585, + "grad_norm": 0.5757942902163647, + "learning_rate": 3.833449760923498e-06, + "loss": 0.7759, + "step": 1740 + }, + { + "epoch": 0.1570095143617261, + "grad_norm": 2.9810083383304717, + "learning_rate": 3.83321627941533e-06, + "loss": 1.0646, + "step": 1741 + }, + { + "epoch": 0.15709969788519637, + "grad_norm": 2.104150249555644, + "learning_rate": 3.832982641487043e-06, + "loss": 1.0842, + "step": 1742 + }, + { + "epoch": 0.15718988140866663, + "grad_norm": 1.57931564285936, + "learning_rate": 3.832748847158568e-06, + "loss": 1.1174, + "step": 1743 + }, + { + "epoch": 0.1572800649321369, + "grad_norm": 2.247979846687528, + "learning_rate": 3.832514896449858e-06, + "loss": 1.0516, + "step": 1744 + }, + { + "epoch": 0.15737024845560715, + "grad_norm": 1.6041257118182497, + "learning_rate": 3.832280789380871e-06, + "loss": 0.9616, + "step": 1745 + }, + { + "epoch": 0.15746043197907741, + "grad_norm": 1.665693889288335, + "learning_rate": 3.832046525971584e-06, + "loss": 0.9293, + "step": 1746 + }, + { + "epoch": 0.15755061550254768, + "grad_norm": 2.1473956547034727, + "learning_rate": 3.831812106241987e-06, + "loss": 1.0839, + "step": 1747 + }, + { + "epoch": 0.15764079902601794, + "grad_norm": 1.628740553810206, + "learning_rate": 3.8315775302120796e-06, + "loss": 0.9317, + "step": 1748 + }, + { + "epoch": 0.1577309825494882, + "grad_norm": 1.567327290439896, + "learning_rate": 3.831342797901878e-06, + "loss": 1.0415, + "step": 1749 + }, + { + "epoch": 0.15782116607295846, + "grad_norm": 0.8017802264371358, + "learning_rate": 3.831107909331411e-06, + "loss": 0.8169, + "step": 1750 + }, + { + "epoch": 0.15791134959642875, + "grad_norm": 2.1792314387287326, + "learning_rate": 3.830872864520721e-06, + "loss": 1.0717, + "step": 1751 + }, + { + "epoch": 0.158001533119899, + "grad_norm": 0.7153114081524823, + "learning_rate": 3.830637663489862e-06, + "loss": 0.896, + "step": 1752 + }, + { + "epoch": 0.15809171664336927, + "grad_norm": 0.7793598197431441, + "learning_rate": 3.830402306258904e-06, + "loss": 0.9262, + "step": 1753 + }, + { + "epoch": 0.15818190016683953, + "grad_norm": 1.7368297903485217, + "learning_rate": 3.830166792847929e-06, + "loss": 1.0136, + "step": 1754 + }, + { + "epoch": 0.1582720836903098, + "grad_norm": 2.4697859316892954, + "learning_rate": 3.829931123277031e-06, + "loss": 0.9679, + "step": 1755 + }, + { + "epoch": 0.15836226721378005, + "grad_norm": 1.925590149719788, + "learning_rate": 3.8296952975663204e-06, + "loss": 1.0127, + "step": 1756 + }, + { + "epoch": 0.1584524507372503, + "grad_norm": 1.8616574807596415, + "learning_rate": 3.829459315735918e-06, + "loss": 1.0913, + "step": 1757 + }, + { + "epoch": 0.15854263426072057, + "grad_norm": 1.8343043429074297, + "learning_rate": 3.829223177805959e-06, + "loss": 0.9505, + "step": 1758 + }, + { + "epoch": 0.15863281778419083, + "grad_norm": 1.7237474787017995, + "learning_rate": 3.828986883796591e-06, + "loss": 1.0665, + "step": 1759 + }, + { + "epoch": 0.1587230013076611, + "grad_norm": 1.7162362182463924, + "learning_rate": 3.828750433727979e-06, + "loss": 1.0142, + "step": 1760 + }, + { + "epoch": 0.15881318483113135, + "grad_norm": 1.8945488683136973, + "learning_rate": 3.828513827620296e-06, + "loss": 1.0724, + "step": 1761 + }, + { + "epoch": 0.15890336835460162, + "grad_norm": 2.3033943298293904, + "learning_rate": 3.82827706549373e-06, + "loss": 1.0771, + "step": 1762 + }, + { + "epoch": 0.15899355187807188, + "grad_norm": 1.8671897345493624, + "learning_rate": 3.828040147368484e-06, + "loss": 1.0116, + "step": 1763 + }, + { + "epoch": 0.15908373540154214, + "grad_norm": 1.5140194598399883, + "learning_rate": 3.827803073264774e-06, + "loss": 1.02, + "step": 1764 + }, + { + "epoch": 0.1591739189250124, + "grad_norm": 1.6824004509143886, + "learning_rate": 3.827565843202826e-06, + "loss": 1.0247, + "step": 1765 + }, + { + "epoch": 0.15926410244848266, + "grad_norm": 2.144971903014508, + "learning_rate": 3.827328457202884e-06, + "loss": 0.8999, + "step": 1766 + }, + { + "epoch": 0.15935428597195292, + "grad_norm": 3.6063075386627146, + "learning_rate": 3.8270909152852014e-06, + "loss": 0.9529, + "step": 1767 + }, + { + "epoch": 0.15944446949542318, + "grad_norm": 1.8973052490570317, + "learning_rate": 3.826853217470048e-06, + "loss": 1.014, + "step": 1768 + }, + { + "epoch": 0.15953465301889344, + "grad_norm": 1.9956411307639033, + "learning_rate": 3.826615363777705e-06, + "loss": 1.1049, + "step": 1769 + }, + { + "epoch": 0.1596248365423637, + "grad_norm": 2.0455687336886004, + "learning_rate": 3.826377354228468e-06, + "loss": 1.0034, + "step": 1770 + }, + { + "epoch": 0.15971502006583396, + "grad_norm": 0.6540557189167044, + "learning_rate": 3.826139188842643e-06, + "loss": 0.8294, + "step": 1771 + }, + { + "epoch": 0.15980520358930422, + "grad_norm": 1.6893163724909455, + "learning_rate": 3.825900867640554e-06, + "loss": 1.0794, + "step": 1772 + }, + { + "epoch": 0.15989538711277448, + "grad_norm": 3.77132531990799, + "learning_rate": 3.825662390642535e-06, + "loss": 1.0604, + "step": 1773 + }, + { + "epoch": 0.15998557063624474, + "grad_norm": 1.733209029044274, + "learning_rate": 3.825423757868934e-06, + "loss": 1.1313, + "step": 1774 + }, + { + "epoch": 0.16007575415971503, + "grad_norm": 1.5230497991744658, + "learning_rate": 3.825184969340114e-06, + "loss": 1.0509, + "step": 1775 + }, + { + "epoch": 0.1601659376831853, + "grad_norm": 2.395582954481639, + "learning_rate": 3.824946025076447e-06, + "loss": 1.0547, + "step": 1776 + }, + { + "epoch": 0.16025612120665556, + "grad_norm": 1.9668349988740845, + "learning_rate": 3.824706925098323e-06, + "loss": 0.9698, + "step": 1777 + }, + { + "epoch": 0.16034630473012582, + "grad_norm": 5.585032057571957, + "learning_rate": 3.824467669426143e-06, + "loss": 1.0821, + "step": 1778 + }, + { + "epoch": 0.16043648825359608, + "grad_norm": 1.9340600886709713, + "learning_rate": 3.824228258080321e-06, + "loss": 0.9905, + "step": 1779 + }, + { + "epoch": 0.16052667177706634, + "grad_norm": 1.8173387328413089, + "learning_rate": 3.823988691081285e-06, + "loss": 0.9355, + "step": 1780 + }, + { + "epoch": 0.1606168553005366, + "grad_norm": 1.5086270834876558, + "learning_rate": 3.823748968449478e-06, + "loss": 1.0574, + "step": 1781 + }, + { + "epoch": 0.16070703882400686, + "grad_norm": 3.239714117977564, + "learning_rate": 3.823509090205352e-06, + "loss": 0.9143, + "step": 1782 + }, + { + "epoch": 0.16079722234747712, + "grad_norm": 1.794222723579342, + "learning_rate": 3.823269056369376e-06, + "loss": 0.9796, + "step": 1783 + }, + { + "epoch": 0.16088740587094738, + "grad_norm": 1.6074358173567573, + "learning_rate": 3.8230288669620295e-06, + "loss": 1.0543, + "step": 1784 + }, + { + "epoch": 0.16097758939441764, + "grad_norm": 1.5488295873801026, + "learning_rate": 3.822788522003809e-06, + "loss": 1.0262, + "step": 1785 + }, + { + "epoch": 0.1610677729178879, + "grad_norm": 2.7164745235557954, + "learning_rate": 3.822548021515221e-06, + "loss": 1.0974, + "step": 1786 + }, + { + "epoch": 0.16115795644135816, + "grad_norm": 2.423992879894866, + "learning_rate": 3.822307365516787e-06, + "loss": 0.9968, + "step": 1787 + }, + { + "epoch": 0.16124813996482842, + "grad_norm": 1.8450108758869455, + "learning_rate": 3.8220665540290395e-06, + "loss": 1.043, + "step": 1788 + }, + { + "epoch": 0.16133832348829868, + "grad_norm": 1.587515435932237, + "learning_rate": 3.8218255870725265e-06, + "loss": 1.0883, + "step": 1789 + }, + { + "epoch": 0.16142850701176895, + "grad_norm": 1.7791642140618005, + "learning_rate": 3.82158446466781e-06, + "loss": 0.9754, + "step": 1790 + }, + { + "epoch": 0.1615186905352392, + "grad_norm": 1.9136433667973445, + "learning_rate": 3.821343186835462e-06, + "loss": 1.035, + "step": 1791 + }, + { + "epoch": 0.16160887405870947, + "grad_norm": 1.8289319075292387, + "learning_rate": 3.821101753596072e-06, + "loss": 1.0962, + "step": 1792 + }, + { + "epoch": 0.16169905758217973, + "grad_norm": 1.9182468686584744, + "learning_rate": 3.820860164970237e-06, + "loss": 1.0112, + "step": 1793 + }, + { + "epoch": 0.16178924110565, + "grad_norm": 1.5876189840587385, + "learning_rate": 3.820618420978574e-06, + "loss": 1.0928, + "step": 1794 + }, + { + "epoch": 0.16187942462912025, + "grad_norm": 1.7325186890534, + "learning_rate": 3.820376521641708e-06, + "loss": 1.0218, + "step": 1795 + }, + { + "epoch": 0.1619696081525905, + "grad_norm": 1.6183594221427702, + "learning_rate": 3.82013446698028e-06, + "loss": 1.0314, + "step": 1796 + }, + { + "epoch": 0.16205979167606077, + "grad_norm": 1.668236168322567, + "learning_rate": 3.819892257014943e-06, + "loss": 0.9885, + "step": 1797 + }, + { + "epoch": 0.16214997519953103, + "grad_norm": 2.464114412233809, + "learning_rate": 3.819649891766364e-06, + "loss": 1.0511, + "step": 1798 + }, + { + "epoch": 0.16224015872300132, + "grad_norm": 2.246275468192605, + "learning_rate": 3.819407371255222e-06, + "loss": 1.0195, + "step": 1799 + }, + { + "epoch": 0.16233034224647158, + "grad_norm": 1.7023816367524938, + "learning_rate": 3.819164695502212e-06, + "loss": 1.034, + "step": 1800 + }, + { + "epoch": 0.16242052576994184, + "grad_norm": 2.0999235548001836, + "learning_rate": 3.818921864528039e-06, + "loss": 0.9693, + "step": 1801 + }, + { + "epoch": 0.1625107092934121, + "grad_norm": 2.1298036696165354, + "learning_rate": 3.818678878353423e-06, + "loss": 1.0792, + "step": 1802 + }, + { + "epoch": 0.16260089281688236, + "grad_norm": 1.8602467905413411, + "learning_rate": 3.818435736999097e-06, + "loss": 0.968, + "step": 1803 + }, + { + "epoch": 0.16269107634035262, + "grad_norm": 1.9592865456992141, + "learning_rate": 3.818192440485807e-06, + "loss": 1.0149, + "step": 1804 + }, + { + "epoch": 0.16278125986382289, + "grad_norm": 1.7447046134829574, + "learning_rate": 3.817948988834314e-06, + "loss": 0.9604, + "step": 1805 + }, + { + "epoch": 0.16287144338729315, + "grad_norm": 1.5940184273653788, + "learning_rate": 3.817705382065388e-06, + "loss": 1.0198, + "step": 1806 + }, + { + "epoch": 0.1629616269107634, + "grad_norm": 1.3437194820753284, + "learning_rate": 3.8174616201998155e-06, + "loss": 0.9912, + "step": 1807 + }, + { + "epoch": 0.16305181043423367, + "grad_norm": 1.7563703258197823, + "learning_rate": 3.817217703258397e-06, + "loss": 1.0691, + "step": 1808 + }, + { + "epoch": 0.16314199395770393, + "grad_norm": 1.5555013702053018, + "learning_rate": 3.816973631261943e-06, + "loss": 1.0141, + "step": 1809 + }, + { + "epoch": 0.1632321774811742, + "grad_norm": 1.919976141503235, + "learning_rate": 3.816729404231281e-06, + "loss": 1.0542, + "step": 1810 + }, + { + "epoch": 0.16332236100464445, + "grad_norm": 1.9899473273432606, + "learning_rate": 3.816485022187249e-06, + "loss": 1.0875, + "step": 1811 + }, + { + "epoch": 0.1634125445281147, + "grad_norm": 2.3469999033761706, + "learning_rate": 3.816240485150698e-06, + "loss": 0.9765, + "step": 1812 + }, + { + "epoch": 0.16350272805158497, + "grad_norm": 1.5593714864993218, + "learning_rate": 3.815995793142495e-06, + "loss": 1.0215, + "step": 1813 + }, + { + "epoch": 0.16359291157505523, + "grad_norm": 2.5235362314095218, + "learning_rate": 3.815750946183518e-06, + "loss": 0.9641, + "step": 1814 + }, + { + "epoch": 0.1636830950985255, + "grad_norm": 1.8538882943114545, + "learning_rate": 3.815505944294658e-06, + "loss": 1.0785, + "step": 1815 + }, + { + "epoch": 0.16377327862199575, + "grad_norm": 1.4641428987458833, + "learning_rate": 3.81526078749682e-06, + "loss": 0.9821, + "step": 1816 + }, + { + "epoch": 0.16386346214546602, + "grad_norm": 2.2178590086970127, + "learning_rate": 3.8150154758109225e-06, + "loss": 0.9621, + "step": 1817 + }, + { + "epoch": 0.16395364566893628, + "grad_norm": 1.6958886499154322, + "learning_rate": 3.814770009257896e-06, + "loss": 0.9975, + "step": 1818 + }, + { + "epoch": 0.16404382919240654, + "grad_norm": 1.5041200956424585, + "learning_rate": 3.814524387858687e-06, + "loss": 1.0364, + "step": 1819 + }, + { + "epoch": 0.1641340127158768, + "grad_norm": 1.9276195982801763, + "learning_rate": 3.814278611634251e-06, + "loss": 1.0447, + "step": 1820 + }, + { + "epoch": 0.16422419623934706, + "grad_norm": 1.9611598037003575, + "learning_rate": 3.8140326806055606e-06, + "loss": 0.9937, + "step": 1821 + }, + { + "epoch": 0.16431437976281732, + "grad_norm": 4.499796332942622, + "learning_rate": 3.8137865947935992e-06, + "loss": 1.0902, + "step": 1822 + }, + { + "epoch": 0.1644045632862876, + "grad_norm": 1.8433581436906514, + "learning_rate": 3.8135403542193646e-06, + "loss": 1.0426, + "step": 1823 + }, + { + "epoch": 0.16449474680975787, + "grad_norm": 2.029823741495627, + "learning_rate": 3.813293958903867e-06, + "loss": 0.9565, + "step": 1824 + }, + { + "epoch": 0.16458493033322813, + "grad_norm": 1.6865459853687803, + "learning_rate": 3.8130474088681306e-06, + "loss": 0.9887, + "step": 1825 + }, + { + "epoch": 0.1646751138566984, + "grad_norm": 1.7517242111979752, + "learning_rate": 3.8128007041331927e-06, + "loss": 0.9501, + "step": 1826 + }, + { + "epoch": 0.16476529738016865, + "grad_norm": 1.8026405384443387, + "learning_rate": 3.812553844720102e-06, + "loss": 0.9372, + "step": 1827 + }, + { + "epoch": 0.1648554809036389, + "grad_norm": 1.8235987522506714, + "learning_rate": 3.8123068306499236e-06, + "loss": 1.0381, + "step": 1828 + }, + { + "epoch": 0.16494566442710917, + "grad_norm": 1.6488926697519444, + "learning_rate": 3.812059661943733e-06, + "loss": 1.0199, + "step": 1829 + }, + { + "epoch": 0.16503584795057943, + "grad_norm": 1.6962154111097536, + "learning_rate": 3.811812338622621e-06, + "loss": 0.9903, + "step": 1830 + }, + { + "epoch": 0.1651260314740497, + "grad_norm": 1.5297109693649753, + "learning_rate": 3.81156486070769e-06, + "loss": 1.0152, + "step": 1831 + }, + { + "epoch": 0.16521621499751996, + "grad_norm": 1.9841471864186504, + "learning_rate": 3.811317228220056e-06, + "loss": 1.067, + "step": 1832 + }, + { + "epoch": 0.16530639852099022, + "grad_norm": 2.0973756058990625, + "learning_rate": 3.811069441180849e-06, + "loss": 1.0664, + "step": 1833 + }, + { + "epoch": 0.16539658204446048, + "grad_norm": 2.2039368128018095, + "learning_rate": 3.8108214996112107e-06, + "loss": 1.0205, + "step": 1834 + }, + { + "epoch": 0.16548676556793074, + "grad_norm": 1.6154341476133494, + "learning_rate": 3.810573403532297e-06, + "loss": 1.0599, + "step": 1835 + }, + { + "epoch": 0.165576949091401, + "grad_norm": 1.9149743224332314, + "learning_rate": 3.8103251529652774e-06, + "loss": 1.0015, + "step": 1836 + }, + { + "epoch": 0.16566713261487126, + "grad_norm": 2.105739823763444, + "learning_rate": 3.810076747931334e-06, + "loss": 0.9976, + "step": 1837 + }, + { + "epoch": 0.16575731613834152, + "grad_norm": 1.9639976302596827, + "learning_rate": 3.809828188451662e-06, + "loss": 0.9789, + "step": 1838 + }, + { + "epoch": 0.16584749966181178, + "grad_norm": 2.5548917822865915, + "learning_rate": 3.809579474547469e-06, + "loss": 1.0237, + "step": 1839 + }, + { + "epoch": 0.16593768318528204, + "grad_norm": 1.8256739299607168, + "learning_rate": 3.809330606239977e-06, + "loss": 1.0689, + "step": 1840 + }, + { + "epoch": 0.1660278667087523, + "grad_norm": 2.0110178494944067, + "learning_rate": 3.809081583550422e-06, + "loss": 1.0869, + "step": 1841 + }, + { + "epoch": 0.16611805023222256, + "grad_norm": 1.8730933349702599, + "learning_rate": 3.808832406500051e-06, + "loss": 1.102, + "step": 1842 + }, + { + "epoch": 0.16620823375569282, + "grad_norm": 1.7552236524266194, + "learning_rate": 3.8085830751101253e-06, + "loss": 1.094, + "step": 1843 + }, + { + "epoch": 0.16629841727916309, + "grad_norm": 1.652641900761169, + "learning_rate": 3.808333589401919e-06, + "loss": 1.0538, + "step": 1844 + }, + { + "epoch": 0.16638860080263335, + "grad_norm": 1.78594218843367, + "learning_rate": 3.8080839493967194e-06, + "loss": 0.9547, + "step": 1845 + }, + { + "epoch": 0.1664787843261036, + "grad_norm": 0.6677020695035959, + "learning_rate": 3.807834155115828e-06, + "loss": 0.7836, + "step": 1846 + }, + { + "epoch": 0.1665689678495739, + "grad_norm": 0.7473962409431507, + "learning_rate": 3.8075842065805584e-06, + "loss": 0.8879, + "step": 1847 + }, + { + "epoch": 0.16665915137304416, + "grad_norm": 1.5763168551903368, + "learning_rate": 3.8073341038122374e-06, + "loss": 1.0093, + "step": 1848 + }, + { + "epoch": 0.16674933489651442, + "grad_norm": 1.760794183404476, + "learning_rate": 3.8070838468322048e-06, + "loss": 1.0885, + "step": 1849 + }, + { + "epoch": 0.16683951841998468, + "grad_norm": 2.3610865971907073, + "learning_rate": 3.8068334356618143e-06, + "loss": 0.9589, + "step": 1850 + }, + { + "epoch": 0.16692970194345494, + "grad_norm": 2.2473266402312473, + "learning_rate": 3.8065828703224324e-06, + "loss": 1.0563, + "step": 1851 + }, + { + "epoch": 0.1670198854669252, + "grad_norm": 1.8337820833255998, + "learning_rate": 3.8063321508354386e-06, + "loss": 1.0235, + "step": 1852 + }, + { + "epoch": 0.16711006899039546, + "grad_norm": 2.007549342859017, + "learning_rate": 3.8060812772222255e-06, + "loss": 1.0021, + "step": 1853 + }, + { + "epoch": 0.16720025251386572, + "grad_norm": 1.4605002335362443, + "learning_rate": 3.8058302495041993e-06, + "loss": 1.0884, + "step": 1854 + }, + { + "epoch": 0.16729043603733598, + "grad_norm": 2.0183196279289857, + "learning_rate": 3.805579067702779e-06, + "loss": 1.0319, + "step": 1855 + }, + { + "epoch": 0.16738061956080624, + "grad_norm": 1.8401802548973065, + "learning_rate": 3.8053277318393967e-06, + "loss": 1.024, + "step": 1856 + }, + { + "epoch": 0.1674708030842765, + "grad_norm": 1.6364073536793653, + "learning_rate": 3.805076241935498e-06, + "loss": 1.0342, + "step": 1857 + }, + { + "epoch": 0.16756098660774676, + "grad_norm": 1.786314874195742, + "learning_rate": 3.804824598012541e-06, + "loss": 0.8952, + "step": 1858 + }, + { + "epoch": 0.16765117013121703, + "grad_norm": 1.5824297367786764, + "learning_rate": 3.8045728000919975e-06, + "loss": 1.0631, + "step": 1859 + }, + { + "epoch": 0.1677413536546873, + "grad_norm": 1.6794168653476527, + "learning_rate": 3.8043208481953524e-06, + "loss": 1.059, + "step": 1860 + }, + { + "epoch": 0.16783153717815755, + "grad_norm": 1.6951199611197638, + "learning_rate": 3.804068742344104e-06, + "loss": 0.9474, + "step": 1861 + }, + { + "epoch": 0.1679217207016278, + "grad_norm": 1.7996186912167964, + "learning_rate": 3.8038164825597628e-06, + "loss": 0.992, + "step": 1862 + }, + { + "epoch": 0.16801190422509807, + "grad_norm": 2.800778355723313, + "learning_rate": 3.8035640688638537e-06, + "loss": 1.1029, + "step": 1863 + }, + { + "epoch": 0.16810208774856833, + "grad_norm": 1.7087082645255909, + "learning_rate": 3.8033115012779125e-06, + "loss": 1.0995, + "step": 1864 + }, + { + "epoch": 0.1681922712720386, + "grad_norm": 1.4354690634470837, + "learning_rate": 3.8030587798234915e-06, + "loss": 1.0631, + "step": 1865 + }, + { + "epoch": 0.16828245479550885, + "grad_norm": 1.819589954443083, + "learning_rate": 3.802805904522153e-06, + "loss": 0.9564, + "step": 1866 + }, + { + "epoch": 0.1683726383189791, + "grad_norm": 1.707515272572077, + "learning_rate": 3.8025528753954742e-06, + "loss": 1.0147, + "step": 1867 + }, + { + "epoch": 0.16846282184244937, + "grad_norm": 1.8783851582988726, + "learning_rate": 3.802299692465045e-06, + "loss": 0.9664, + "step": 1868 + }, + { + "epoch": 0.16855300536591963, + "grad_norm": 4.028993196642167, + "learning_rate": 3.802046355752468e-06, + "loss": 1.0115, + "step": 1869 + }, + { + "epoch": 0.16864318888938992, + "grad_norm": 3.5338249608264083, + "learning_rate": 3.80179286527936e-06, + "loss": 1.079, + "step": 1870 + }, + { + "epoch": 0.16873337241286018, + "grad_norm": 1.566675405613219, + "learning_rate": 3.801539221067349e-06, + "loss": 0.9369, + "step": 1871 + }, + { + "epoch": 0.16882355593633044, + "grad_norm": 1.5898084331966873, + "learning_rate": 3.801285423138079e-06, + "loss": 0.9796, + "step": 1872 + }, + { + "epoch": 0.1689137394598007, + "grad_norm": 5.449057562300177, + "learning_rate": 3.8010314715132037e-06, + "loss": 0.9773, + "step": 1873 + }, + { + "epoch": 0.16900392298327097, + "grad_norm": 1.726807106608033, + "learning_rate": 3.800777366214393e-06, + "loss": 0.9763, + "step": 1874 + }, + { + "epoch": 0.16909410650674123, + "grad_norm": 2.2188851758585932, + "learning_rate": 3.800523107263328e-06, + "loss": 0.9943, + "step": 1875 + }, + { + "epoch": 0.1691842900302115, + "grad_norm": 1.6135394834138774, + "learning_rate": 3.800268694681703e-06, + "loss": 1.0782, + "step": 1876 + }, + { + "epoch": 0.16927447355368175, + "grad_norm": 0.663662262521951, + "learning_rate": 3.800014128491227e-06, + "loss": 0.8366, + "step": 1877 + }, + { + "epoch": 0.169364657077152, + "grad_norm": 2.2808563728920346, + "learning_rate": 3.79975940871362e-06, + "loss": 1.0733, + "step": 1878 + }, + { + "epoch": 0.16945484060062227, + "grad_norm": 2.7477252395120964, + "learning_rate": 3.799504535370617e-06, + "loss": 1.0255, + "step": 1879 + }, + { + "epoch": 0.16954502412409253, + "grad_norm": 1.9177243451871635, + "learning_rate": 3.799249508483964e-06, + "loss": 0.9599, + "step": 1880 + }, + { + "epoch": 0.1696352076475628, + "grad_norm": 0.734713618350916, + "learning_rate": 3.798994328075422e-06, + "loss": 0.8966, + "step": 1881 + }, + { + "epoch": 0.16972539117103305, + "grad_norm": 1.8994946460251791, + "learning_rate": 3.798738994166765e-06, + "loss": 0.9483, + "step": 1882 + }, + { + "epoch": 0.1698155746945033, + "grad_norm": 1.6432813665337924, + "learning_rate": 3.7984835067797788e-06, + "loss": 1.0224, + "step": 1883 + }, + { + "epoch": 0.16990575821797357, + "grad_norm": 1.5126485010812942, + "learning_rate": 3.798227865936263e-06, + "loss": 1.1156, + "step": 1884 + }, + { + "epoch": 0.16999594174144383, + "grad_norm": 1.9725453562773687, + "learning_rate": 3.7979720716580297e-06, + "loss": 0.8869, + "step": 1885 + }, + { + "epoch": 0.1700861252649141, + "grad_norm": 1.795226692825057, + "learning_rate": 3.7977161239669057e-06, + "loss": 1.009, + "step": 1886 + }, + { + "epoch": 0.17017630878838436, + "grad_norm": 2.6043810540008305, + "learning_rate": 3.7974600228847294e-06, + "loss": 1.0019, + "step": 1887 + }, + { + "epoch": 0.17026649231185462, + "grad_norm": 2.6375882897808673, + "learning_rate": 3.7972037684333534e-06, + "loss": 0.9609, + "step": 1888 + }, + { + "epoch": 0.17035667583532488, + "grad_norm": 1.7196555093170514, + "learning_rate": 3.796947360634642e-06, + "loss": 0.9494, + "step": 1889 + }, + { + "epoch": 0.17044685935879514, + "grad_norm": 1.8738458260023594, + "learning_rate": 3.796690799510473e-06, + "loss": 1.0694, + "step": 1890 + }, + { + "epoch": 0.1705370428822654, + "grad_norm": 4.225451085314993, + "learning_rate": 3.7964340850827387e-06, + "loss": 0.985, + "step": 1891 + }, + { + "epoch": 0.17062722640573566, + "grad_norm": 0.6775592060894017, + "learning_rate": 3.7961772173733425e-06, + "loss": 0.8046, + "step": 1892 + }, + { + "epoch": 0.17071740992920592, + "grad_norm": 2.302357675997606, + "learning_rate": 3.7959201964042024e-06, + "loss": 1.0117, + "step": 1893 + }, + { + "epoch": 0.1708075934526762, + "grad_norm": 1.5508458967210723, + "learning_rate": 3.795663022197248e-06, + "loss": 1.052, + "step": 1894 + }, + { + "epoch": 0.17089777697614647, + "grad_norm": 1.9866522390473234, + "learning_rate": 3.7954056947744242e-06, + "loss": 1.0174, + "step": 1895 + }, + { + "epoch": 0.17098796049961673, + "grad_norm": 1.6277498307214087, + "learning_rate": 3.7951482141576863e-06, + "loss": 0.9984, + "step": 1896 + }, + { + "epoch": 0.171078144023087, + "grad_norm": 1.3816044158548006, + "learning_rate": 3.794890580369004e-06, + "loss": 1.0762, + "step": 1897 + }, + { + "epoch": 0.17116832754655725, + "grad_norm": 1.9668549393660288, + "learning_rate": 3.7946327934303612e-06, + "loss": 1.0525, + "step": 1898 + }, + { + "epoch": 0.1712585110700275, + "grad_norm": 1.7454297469465447, + "learning_rate": 3.794374853363752e-06, + "loss": 1.1204, + "step": 1899 + }, + { + "epoch": 0.17134869459349777, + "grad_norm": 1.5552469138461436, + "learning_rate": 3.794116760191187e-06, + "loss": 1.043, + "step": 1900 + }, + { + "epoch": 0.17143887811696804, + "grad_norm": 2.1908674206650756, + "learning_rate": 3.7938585139346877e-06, + "loss": 1.1145, + "step": 1901 + }, + { + "epoch": 0.1715290616404383, + "grad_norm": 2.0131314252616437, + "learning_rate": 3.793600114616288e-06, + "loss": 0.9993, + "step": 1902 + }, + { + "epoch": 0.17161924516390856, + "grad_norm": 0.7380982025734454, + "learning_rate": 3.793341562258037e-06, + "loss": 0.7947, + "step": 1903 + }, + { + "epoch": 0.17170942868737882, + "grad_norm": 1.465303638746299, + "learning_rate": 3.7930828568819953e-06, + "loss": 1.0314, + "step": 1904 + }, + { + "epoch": 0.17179961221084908, + "grad_norm": 1.7349956192488978, + "learning_rate": 3.7928239985102378e-06, + "loss": 1.0196, + "step": 1905 + }, + { + "epoch": 0.17188979573431934, + "grad_norm": 1.714232107584532, + "learning_rate": 3.7925649871648505e-06, + "loss": 1.0219, + "step": 1906 + }, + { + "epoch": 0.1719799792577896, + "grad_norm": 1.7496956833138595, + "learning_rate": 3.792305822867935e-06, + "loss": 0.942, + "step": 1907 + }, + { + "epoch": 0.17207016278125986, + "grad_norm": 1.9750950609338636, + "learning_rate": 3.792046505641604e-06, + "loss": 0.9828, + "step": 1908 + }, + { + "epoch": 0.17216034630473012, + "grad_norm": 1.8365677421408033, + "learning_rate": 3.791787035507984e-06, + "loss": 0.9796, + "step": 1909 + }, + { + "epoch": 0.17225052982820038, + "grad_norm": 1.9054521235404946, + "learning_rate": 3.7915274124892136e-06, + "loss": 0.9825, + "step": 1910 + }, + { + "epoch": 0.17234071335167064, + "grad_norm": 1.5773828857766476, + "learning_rate": 3.7912676366074466e-06, + "loss": 1.0038, + "step": 1911 + }, + { + "epoch": 0.1724308968751409, + "grad_norm": 2.801298170922426, + "learning_rate": 3.7910077078848478e-06, + "loss": 0.9835, + "step": 1912 + }, + { + "epoch": 0.17252108039861117, + "grad_norm": 1.9720332908876723, + "learning_rate": 3.7907476263435957e-06, + "loss": 0.983, + "step": 1913 + }, + { + "epoch": 0.17261126392208143, + "grad_norm": 1.7672512700643872, + "learning_rate": 3.7904873920058826e-06, + "loss": 0.9895, + "step": 1914 + }, + { + "epoch": 0.1727014474455517, + "grad_norm": 1.8782397573134157, + "learning_rate": 3.7902270048939114e-06, + "loss": 0.8941, + "step": 1915 + }, + { + "epoch": 0.17279163096902195, + "grad_norm": 2.159624527510048, + "learning_rate": 3.7899664650299023e-06, + "loss": 0.9781, + "step": 1916 + }, + { + "epoch": 0.1728818144924922, + "grad_norm": 1.4980792622751842, + "learning_rate": 3.7897057724360836e-06, + "loss": 1.0289, + "step": 1917 + }, + { + "epoch": 0.1729719980159625, + "grad_norm": 1.8135750311934464, + "learning_rate": 3.7894449271347004e-06, + "loss": 1.0387, + "step": 1918 + }, + { + "epoch": 0.17306218153943276, + "grad_norm": 9.572007981191417, + "learning_rate": 3.789183929148009e-06, + "loss": 1.0779, + "step": 1919 + }, + { + "epoch": 0.17315236506290302, + "grad_norm": 1.87998458784969, + "learning_rate": 3.7889227784982795e-06, + "loss": 1.0469, + "step": 1920 + }, + { + "epoch": 0.17324254858637328, + "grad_norm": 1.4477003837071114, + "learning_rate": 3.7886614752077945e-06, + "loss": 1.0131, + "step": 1921 + }, + { + "epoch": 0.17333273210984354, + "grad_norm": 1.701789052013886, + "learning_rate": 3.7884000192988495e-06, + "loss": 1.0157, + "step": 1922 + }, + { + "epoch": 0.1734229156333138, + "grad_norm": 1.907294753145241, + "learning_rate": 3.7881384107937546e-06, + "loss": 1.0683, + "step": 1923 + }, + { + "epoch": 0.17351309915678406, + "grad_norm": 3.0804316983569344, + "learning_rate": 3.78787664971483e-06, + "loss": 1.0133, + "step": 1924 + }, + { + "epoch": 0.17360328268025432, + "grad_norm": 1.4703284365107538, + "learning_rate": 3.7876147360844115e-06, + "loss": 0.9892, + "step": 1925 + }, + { + "epoch": 0.17369346620372458, + "grad_norm": 2.001344705565027, + "learning_rate": 3.7873526699248474e-06, + "loss": 1.0887, + "step": 1926 + }, + { + "epoch": 0.17378364972719484, + "grad_norm": 2.9990937135611766, + "learning_rate": 3.7870904512584974e-06, + "loss": 1.0286, + "step": 1927 + }, + { + "epoch": 0.1738738332506651, + "grad_norm": 1.8287679079734314, + "learning_rate": 3.7868280801077368e-06, + "loss": 1.0191, + "step": 1928 + }, + { + "epoch": 0.17396401677413537, + "grad_norm": 1.3483976232261612, + "learning_rate": 3.7865655564949517e-06, + "loss": 1.0117, + "step": 1929 + }, + { + "epoch": 0.17405420029760563, + "grad_norm": 1.438381588164339, + "learning_rate": 3.786302880442542e-06, + "loss": 0.9883, + "step": 1930 + }, + { + "epoch": 0.1741443838210759, + "grad_norm": 1.4416554966899944, + "learning_rate": 3.7860400519729215e-06, + "loss": 1.0188, + "step": 1931 + }, + { + "epoch": 0.17423456734454615, + "grad_norm": 2.7885115049828344, + "learning_rate": 3.7857770711085157e-06, + "loss": 1.0412, + "step": 1932 + }, + { + "epoch": 0.1743247508680164, + "grad_norm": 1.5108334021227245, + "learning_rate": 3.785513937871763e-06, + "loss": 1.0647, + "step": 1933 + }, + { + "epoch": 0.17441493439148667, + "grad_norm": 2.588992616197202, + "learning_rate": 3.785250652285116e-06, + "loss": 0.9888, + "step": 1934 + }, + { + "epoch": 0.17450511791495693, + "grad_norm": 1.6225658672581857, + "learning_rate": 3.78498721437104e-06, + "loss": 1.0958, + "step": 1935 + }, + { + "epoch": 0.1745953014384272, + "grad_norm": 2.5530788945082077, + "learning_rate": 3.784723624152012e-06, + "loss": 1.0113, + "step": 1936 + }, + { + "epoch": 0.17468548496189745, + "grad_norm": 1.9842075066954223, + "learning_rate": 3.784459881650524e-06, + "loss": 0.9616, + "step": 1937 + }, + { + "epoch": 0.1747756684853677, + "grad_norm": 1.621260521824443, + "learning_rate": 3.784195986889079e-06, + "loss": 0.9934, + "step": 1938 + }, + { + "epoch": 0.17486585200883797, + "grad_norm": 3.428804670642728, + "learning_rate": 3.7839319398901946e-06, + "loss": 0.9414, + "step": 1939 + }, + { + "epoch": 0.17495603553230824, + "grad_norm": 1.5466966622557117, + "learning_rate": 3.7836677406764013e-06, + "loss": 1.0923, + "step": 1940 + }, + { + "epoch": 0.1750462190557785, + "grad_norm": 1.4672226776683415, + "learning_rate": 3.7834033892702407e-06, + "loss": 1.0628, + "step": 1941 + }, + { + "epoch": 0.17513640257924878, + "grad_norm": 1.693297282733786, + "learning_rate": 3.783138885694269e-06, + "loss": 0.9809, + "step": 1942 + }, + { + "epoch": 0.17522658610271905, + "grad_norm": 1.7802743749568979, + "learning_rate": 3.7828742299710558e-06, + "loss": 1.0942, + "step": 1943 + }, + { + "epoch": 0.1753167696261893, + "grad_norm": 1.5821967144373261, + "learning_rate": 3.782609422123183e-06, + "loss": 0.9972, + "step": 1944 + }, + { + "epoch": 0.17540695314965957, + "grad_norm": 2.566206414790472, + "learning_rate": 3.7823444621732444e-06, + "loss": 1.0214, + "step": 1945 + }, + { + "epoch": 0.17549713667312983, + "grad_norm": 1.915457516889419, + "learning_rate": 3.782079350143849e-06, + "loss": 1.0373, + "step": 1946 + }, + { + "epoch": 0.1755873201966001, + "grad_norm": 2.1886859403829235, + "learning_rate": 3.781814086057617e-06, + "loss": 1.118, + "step": 1947 + }, + { + "epoch": 0.17567750372007035, + "grad_norm": 2.757341901270887, + "learning_rate": 3.7815486699371826e-06, + "loss": 1.0094, + "step": 1948 + }, + { + "epoch": 0.1757676872435406, + "grad_norm": 1.709943857330994, + "learning_rate": 3.7812831018051918e-06, + "loss": 0.9958, + "step": 1949 + }, + { + "epoch": 0.17585787076701087, + "grad_norm": 1.5457287789877363, + "learning_rate": 3.7810173816843058e-06, + "loss": 1.0271, + "step": 1950 + }, + { + "epoch": 0.17594805429048113, + "grad_norm": 1.718395959629871, + "learning_rate": 3.7807515095971955e-06, + "loss": 1.0934, + "step": 1951 + }, + { + "epoch": 0.1760382378139514, + "grad_norm": 1.8464793031484448, + "learning_rate": 3.7804854855665475e-06, + "loss": 1.0142, + "step": 1952 + }, + { + "epoch": 0.17612842133742165, + "grad_norm": 3.0372973025157792, + "learning_rate": 3.7802193096150606e-06, + "loss": 1.0259, + "step": 1953 + }, + { + "epoch": 0.17621860486089191, + "grad_norm": 1.773430105857727, + "learning_rate": 3.779952981765446e-06, + "loss": 0.9701, + "step": 1954 + }, + { + "epoch": 0.17630878838436218, + "grad_norm": 1.9999221548189552, + "learning_rate": 3.779686502040429e-06, + "loss": 1.0358, + "step": 1955 + }, + { + "epoch": 0.17639897190783244, + "grad_norm": 0.6998455498807008, + "learning_rate": 3.779419870462746e-06, + "loss": 0.8126, + "step": 1956 + }, + { + "epoch": 0.1764891554313027, + "grad_norm": 2.045496002894571, + "learning_rate": 3.779153087055148e-06, + "loss": 1.0003, + "step": 1957 + }, + { + "epoch": 0.17657933895477296, + "grad_norm": 2.576301808583601, + "learning_rate": 3.7788861518403988e-06, + "loss": 1.0186, + "step": 1958 + }, + { + "epoch": 0.17666952247824322, + "grad_norm": 1.80198678730091, + "learning_rate": 3.7786190648412742e-06, + "loss": 0.9127, + "step": 1959 + }, + { + "epoch": 0.17675970600171348, + "grad_norm": 0.6504864092337315, + "learning_rate": 3.778351826080564e-06, + "loss": 0.8254, + "step": 1960 + }, + { + "epoch": 0.17684988952518374, + "grad_norm": 1.6358948593630651, + "learning_rate": 3.7780844355810704e-06, + "loss": 1.1215, + "step": 1961 + }, + { + "epoch": 0.176940073048654, + "grad_norm": 1.5390688513610222, + "learning_rate": 3.777816893365608e-06, + "loss": 1.032, + "step": 1962 + }, + { + "epoch": 0.17703025657212426, + "grad_norm": 1.5152380537837609, + "learning_rate": 3.7775491994570057e-06, + "loss": 1.0966, + "step": 1963 + }, + { + "epoch": 0.17712044009559452, + "grad_norm": 1.6897155205146055, + "learning_rate": 3.777281353878105e-06, + "loss": 0.9928, + "step": 1964 + }, + { + "epoch": 0.17721062361906478, + "grad_norm": 0.6213609612459262, + "learning_rate": 3.777013356651758e-06, + "loss": 0.8302, + "step": 1965 + }, + { + "epoch": 0.17730080714253507, + "grad_norm": 1.7762681326319814, + "learning_rate": 3.776745207800834e-06, + "loss": 1.0355, + "step": 1966 + }, + { + "epoch": 0.17739099066600533, + "grad_norm": 1.610969790293234, + "learning_rate": 3.7764769073482122e-06, + "loss": 1.0482, + "step": 1967 + }, + { + "epoch": 0.1774811741894756, + "grad_norm": 1.4424757052645012, + "learning_rate": 3.7762084553167846e-06, + "loss": 1.0871, + "step": 1968 + }, + { + "epoch": 0.17757135771294585, + "grad_norm": 1.8702709961431478, + "learning_rate": 3.775939851729458e-06, + "loss": 1.113, + "step": 1969 + }, + { + "epoch": 0.17766154123641612, + "grad_norm": 2.2757053665097184, + "learning_rate": 3.775671096609151e-06, + "loss": 0.9811, + "step": 1970 + }, + { + "epoch": 0.17775172475988638, + "grad_norm": 1.6960611404383694, + "learning_rate": 3.775402189978795e-06, + "loss": 1.0037, + "step": 1971 + }, + { + "epoch": 0.17784190828335664, + "grad_norm": 1.431855356055874, + "learning_rate": 3.7751331318613343e-06, + "loss": 1.0014, + "step": 1972 + }, + { + "epoch": 0.1779320918068269, + "grad_norm": 1.9564044732965538, + "learning_rate": 3.774863922279727e-06, + "loss": 1.0423, + "step": 1973 + }, + { + "epoch": 0.17802227533029716, + "grad_norm": 1.3833656282370306, + "learning_rate": 3.7745945612569435e-06, + "loss": 1.0794, + "step": 1974 + }, + { + "epoch": 0.17811245885376742, + "grad_norm": 1.2907916407114395, + "learning_rate": 3.7743250488159674e-06, + "loss": 0.8966, + "step": 1975 + }, + { + "epoch": 0.17820264237723768, + "grad_norm": 1.6625699078769232, + "learning_rate": 3.774055384979794e-06, + "loss": 0.9175, + "step": 1976 + }, + { + "epoch": 0.17829282590070794, + "grad_norm": 1.9905657581275253, + "learning_rate": 3.773785569771433e-06, + "loss": 1.0015, + "step": 1977 + }, + { + "epoch": 0.1783830094241782, + "grad_norm": 1.8899178561772292, + "learning_rate": 3.7735156032139066e-06, + "loss": 1.007, + "step": 1978 + }, + { + "epoch": 0.17847319294764846, + "grad_norm": 1.8611712714889503, + "learning_rate": 3.773245485330251e-06, + "loss": 1.1221, + "step": 1979 + }, + { + "epoch": 0.17856337647111872, + "grad_norm": 1.7388036050219007, + "learning_rate": 3.7729752161435115e-06, + "loss": 0.9893, + "step": 1980 + }, + { + "epoch": 0.17865355999458898, + "grad_norm": 1.6376478521559201, + "learning_rate": 3.7727047956767514e-06, + "loss": 1.084, + "step": 1981 + }, + { + "epoch": 0.17874374351805924, + "grad_norm": 1.7563897372547788, + "learning_rate": 3.7724342239530436e-06, + "loss": 1.1495, + "step": 1982 + }, + { + "epoch": 0.1788339270415295, + "grad_norm": 1.7843453800662499, + "learning_rate": 3.772163500995474e-06, + "loss": 0.9672, + "step": 1983 + }, + { + "epoch": 0.17892411056499977, + "grad_norm": 3.1722021075173474, + "learning_rate": 3.7718926268271437e-06, + "loss": 1.1124, + "step": 1984 + }, + { + "epoch": 0.17901429408847003, + "grad_norm": 2.0546059247782855, + "learning_rate": 3.771621601471164e-06, + "loss": 1.0449, + "step": 1985 + }, + { + "epoch": 0.1791044776119403, + "grad_norm": 1.7352461817767673, + "learning_rate": 3.771350424950661e-06, + "loss": 1.0698, + "step": 1986 + }, + { + "epoch": 0.17919466113541055, + "grad_norm": 1.6914123112171275, + "learning_rate": 3.771079097288772e-06, + "loss": 0.9665, + "step": 1987 + }, + { + "epoch": 0.1792848446588808, + "grad_norm": 1.3841490670971766, + "learning_rate": 3.770807618508649e-06, + "loss": 1.037, + "step": 1988 + }, + { + "epoch": 0.17937502818235107, + "grad_norm": 1.6689480267586911, + "learning_rate": 3.7705359886334555e-06, + "loss": 1.0074, + "step": 1989 + }, + { + "epoch": 0.17946521170582136, + "grad_norm": 1.7301383295046646, + "learning_rate": 3.7702642076863694e-06, + "loss": 0.9995, + "step": 1990 + }, + { + "epoch": 0.17955539522929162, + "grad_norm": 1.7510865789126888, + "learning_rate": 3.7699922756905795e-06, + "loss": 1.0793, + "step": 1991 + }, + { + "epoch": 0.17964557875276188, + "grad_norm": 2.063867289157871, + "learning_rate": 3.7697201926692895e-06, + "loss": 0.9762, + "step": 1992 + }, + { + "epoch": 0.17973576227623214, + "grad_norm": 1.714634076553797, + "learning_rate": 3.7694479586457144e-06, + "loss": 1.0374, + "step": 1993 + }, + { + "epoch": 0.1798259457997024, + "grad_norm": 5.011333400215967, + "learning_rate": 3.7691755736430827e-06, + "loss": 1.0312, + "step": 1994 + }, + { + "epoch": 0.17991612932317266, + "grad_norm": 1.5247076754946438, + "learning_rate": 3.768903037684636e-06, + "loss": 1.0641, + "step": 1995 + }, + { + "epoch": 0.18000631284664292, + "grad_norm": 1.469680572499928, + "learning_rate": 3.7686303507936284e-06, + "loss": 1.0687, + "step": 1996 + }, + { + "epoch": 0.18009649637011318, + "grad_norm": 2.624244217835632, + "learning_rate": 3.7683575129933272e-06, + "loss": 1.0603, + "step": 1997 + }, + { + "epoch": 0.18018667989358345, + "grad_norm": 1.5121607565520294, + "learning_rate": 3.7680845243070128e-06, + "loss": 1.0265, + "step": 1998 + }, + { + "epoch": 0.1802768634170537, + "grad_norm": 0.7093196948166627, + "learning_rate": 3.7678113847579767e-06, + "loss": 0.7925, + "step": 1999 + }, + { + "epoch": 0.18036704694052397, + "grad_norm": 3.074480096548798, + "learning_rate": 3.7675380943695264e-06, + "loss": 1.0923, + "step": 2000 + }, + { + "epoch": 0.18045723046399423, + "grad_norm": 1.846413127637471, + "learning_rate": 3.7672646531649795e-06, + "loss": 0.9876, + "step": 2001 + }, + { + "epoch": 0.1805474139874645, + "grad_norm": 0.5788745402970833, + "learning_rate": 3.7669910611676682e-06, + "loss": 0.8144, + "step": 2002 + }, + { + "epoch": 0.18063759751093475, + "grad_norm": 2.6201794276058883, + "learning_rate": 3.7667173184009356e-06, + "loss": 1.0073, + "step": 2003 + }, + { + "epoch": 0.180727781034405, + "grad_norm": 1.5530845007422882, + "learning_rate": 3.7664434248881403e-06, + "loss": 1.0832, + "step": 2004 + }, + { + "epoch": 0.18081796455787527, + "grad_norm": 1.660220299775491, + "learning_rate": 3.766169380652652e-06, + "loss": 1.0339, + "step": 2005 + }, + { + "epoch": 0.18090814808134553, + "grad_norm": 1.583422164767448, + "learning_rate": 3.7658951857178537e-06, + "loss": 1.0631, + "step": 2006 + }, + { + "epoch": 0.1809983316048158, + "grad_norm": 1.7491850999233745, + "learning_rate": 3.7656208401071414e-06, + "loss": 1.0336, + "step": 2007 + }, + { + "epoch": 0.18108851512828605, + "grad_norm": 2.2124118808935633, + "learning_rate": 3.7653463438439225e-06, + "loss": 0.9595, + "step": 2008 + }, + { + "epoch": 0.18117869865175631, + "grad_norm": 2.3357989930391203, + "learning_rate": 3.7650716969516203e-06, + "loss": 1.039, + "step": 2009 + }, + { + "epoch": 0.18126888217522658, + "grad_norm": 3.610131783021735, + "learning_rate": 3.764796899453668e-06, + "loss": 0.9207, + "step": 2010 + }, + { + "epoch": 0.18135906569869684, + "grad_norm": 1.7195459776840825, + "learning_rate": 3.7645219513735134e-06, + "loss": 1.0655, + "step": 2011 + }, + { + "epoch": 0.1814492492221671, + "grad_norm": 1.5331123859743467, + "learning_rate": 3.764246852734617e-06, + "loss": 1.0642, + "step": 2012 + }, + { + "epoch": 0.18153943274563739, + "grad_norm": 1.9638711330804783, + "learning_rate": 3.7639716035604502e-06, + "loss": 0.9924, + "step": 2013 + }, + { + "epoch": 0.18162961626910765, + "grad_norm": 0.6407352678255194, + "learning_rate": 3.7636962038745e-06, + "loss": 0.7862, + "step": 2014 + }, + { + "epoch": 0.1817197997925779, + "grad_norm": 1.639441508355839, + "learning_rate": 3.763420653700265e-06, + "loss": 0.9807, + "step": 2015 + }, + { + "epoch": 0.18180998331604817, + "grad_norm": 1.4640859042223673, + "learning_rate": 3.7631449530612565e-06, + "loss": 0.9815, + "step": 2016 + }, + { + "epoch": 0.18190016683951843, + "grad_norm": 1.6313692129720982, + "learning_rate": 3.762869101980999e-06, + "loss": 1.1128, + "step": 2017 + }, + { + "epoch": 0.1819903503629887, + "grad_norm": 2.535995274024916, + "learning_rate": 3.7625931004830287e-06, + "loss": 1.0208, + "step": 2018 + }, + { + "epoch": 0.18208053388645895, + "grad_norm": 1.5956695067507463, + "learning_rate": 3.7623169485908966e-06, + "loss": 0.9815, + "step": 2019 + }, + { + "epoch": 0.1821707174099292, + "grad_norm": 0.7215593415757446, + "learning_rate": 3.7620406463281647e-06, + "loss": 0.8322, + "step": 2020 + }, + { + "epoch": 0.18226090093339947, + "grad_norm": 2.0204510540224527, + "learning_rate": 3.7617641937184095e-06, + "loss": 1.047, + "step": 2021 + }, + { + "epoch": 0.18235108445686973, + "grad_norm": 1.9635450799914036, + "learning_rate": 3.761487590785219e-06, + "loss": 1.0555, + "step": 2022 + }, + { + "epoch": 0.18244126798034, + "grad_norm": 1.986238220640848, + "learning_rate": 3.7612108375521942e-06, + "loss": 1.0005, + "step": 2023 + }, + { + "epoch": 0.18253145150381025, + "grad_norm": 1.5187504583915363, + "learning_rate": 3.76093393404295e-06, + "loss": 1.0748, + "step": 2024 + }, + { + "epoch": 0.18262163502728052, + "grad_norm": 1.717252113299889, + "learning_rate": 3.7606568802811126e-06, + "loss": 1.0147, + "step": 2025 + }, + { + "epoch": 0.18271181855075078, + "grad_norm": 2.6174006588801886, + "learning_rate": 3.760379676290322e-06, + "loss": 1.0377, + "step": 2026 + }, + { + "epoch": 0.18280200207422104, + "grad_norm": 3.0735445749610433, + "learning_rate": 3.760102322094231e-06, + "loss": 0.9512, + "step": 2027 + }, + { + "epoch": 0.1828921855976913, + "grad_norm": 1.408048877918889, + "learning_rate": 3.759824817716504e-06, + "loss": 0.9858, + "step": 2028 + }, + { + "epoch": 0.18298236912116156, + "grad_norm": 1.2457297819006174, + "learning_rate": 3.759547163180821e-06, + "loss": 1.0726, + "step": 2029 + }, + { + "epoch": 0.18307255264463182, + "grad_norm": 1.6070659331425725, + "learning_rate": 3.759269358510871e-06, + "loss": 0.8263, + "step": 2030 + }, + { + "epoch": 0.18316273616810208, + "grad_norm": 1.7755335717729468, + "learning_rate": 3.75899140373036e-06, + "loss": 0.9296, + "step": 2031 + }, + { + "epoch": 0.18325291969157234, + "grad_norm": 3.15794281839274, + "learning_rate": 3.7587132988630028e-06, + "loss": 1.0873, + "step": 2032 + }, + { + "epoch": 0.1833431032150426, + "grad_norm": 2.8616454068969572, + "learning_rate": 3.7584350439325295e-06, + "loss": 1.0568, + "step": 2033 + }, + { + "epoch": 0.18343328673851286, + "grad_norm": 2.1147027443930635, + "learning_rate": 3.758156638962682e-06, + "loss": 1.0127, + "step": 2034 + }, + { + "epoch": 0.18352347026198312, + "grad_norm": 1.9859171004189293, + "learning_rate": 3.757878083977216e-06, + "loss": 0.9877, + "step": 2035 + }, + { + "epoch": 0.18361365378545338, + "grad_norm": 2.9643543025369348, + "learning_rate": 3.7575993789999e-06, + "loss": 1.0098, + "step": 2036 + }, + { + "epoch": 0.18370383730892367, + "grad_norm": 1.5375820727268816, + "learning_rate": 3.757320524054512e-06, + "loss": 1.0584, + "step": 2037 + }, + { + "epoch": 0.18379402083239393, + "grad_norm": 1.5536514740791532, + "learning_rate": 3.757041519164848e-06, + "loss": 0.9525, + "step": 2038 + }, + { + "epoch": 0.1838842043558642, + "grad_norm": 2.34558460911311, + "learning_rate": 3.7567623643547133e-06, + "loss": 1.0314, + "step": 2039 + }, + { + "epoch": 0.18397438787933446, + "grad_norm": 1.6677144412854736, + "learning_rate": 3.756483059647927e-06, + "loss": 1.0227, + "step": 2040 + }, + { + "epoch": 0.18406457140280472, + "grad_norm": 1.6650359838910322, + "learning_rate": 3.756203605068321e-06, + "loss": 0.9931, + "step": 2041 + }, + { + "epoch": 0.18415475492627498, + "grad_norm": 1.631140112555546, + "learning_rate": 3.7559240006397396e-06, + "loss": 1.0114, + "step": 2042 + }, + { + "epoch": 0.18424493844974524, + "grad_norm": 1.8810823967878836, + "learning_rate": 3.7556442463860406e-06, + "loss": 1.1419, + "step": 2043 + }, + { + "epoch": 0.1843351219732155, + "grad_norm": 1.5940689814872717, + "learning_rate": 3.7553643423310934e-06, + "loss": 1.1341, + "step": 2044 + }, + { + "epoch": 0.18442530549668576, + "grad_norm": 1.7752151654350798, + "learning_rate": 3.755084288498782e-06, + "loss": 0.8783, + "step": 2045 + }, + { + "epoch": 0.18451548902015602, + "grad_norm": 1.7144291057107905, + "learning_rate": 3.754804084913002e-06, + "loss": 1.0247, + "step": 2046 + }, + { + "epoch": 0.18460567254362628, + "grad_norm": 1.4469375959330997, + "learning_rate": 3.754523731597661e-06, + "loss": 1.0912, + "step": 2047 + }, + { + "epoch": 0.18469585606709654, + "grad_norm": 2.646666258019993, + "learning_rate": 3.754243228576681e-06, + "loss": 0.9201, + "step": 2048 + }, + { + "epoch": 0.1847860395905668, + "grad_norm": 1.6782213776544446, + "learning_rate": 3.753962575873996e-06, + "loss": 1.0732, + "step": 2049 + }, + { + "epoch": 0.18487622311403706, + "grad_norm": 1.636144132052433, + "learning_rate": 3.7536817735135527e-06, + "loss": 0.9844, + "step": 2050 + }, + { + "epoch": 0.18496640663750732, + "grad_norm": 1.3836284751562407, + "learning_rate": 3.753400821519311e-06, + "loss": 1.0664, + "step": 2051 + }, + { + "epoch": 0.18505659016097759, + "grad_norm": 1.4827611080223764, + "learning_rate": 3.7531197199152426e-06, + "loss": 0.9505, + "step": 2052 + }, + { + "epoch": 0.18514677368444785, + "grad_norm": 0.7158414205272693, + "learning_rate": 3.7528384687253335e-06, + "loss": 0.855, + "step": 2053 + }, + { + "epoch": 0.1852369572079181, + "grad_norm": 1.6127574153334068, + "learning_rate": 3.7525570679735815e-06, + "loss": 1.1296, + "step": 2054 + }, + { + "epoch": 0.18532714073138837, + "grad_norm": 1.6423233687419385, + "learning_rate": 3.7522755176839965e-06, + "loss": 1.0382, + "step": 2055 + }, + { + "epoch": 0.18541732425485863, + "grad_norm": 1.8180712216509118, + "learning_rate": 3.7519938178806027e-06, + "loss": 1.032, + "step": 2056 + }, + { + "epoch": 0.1855075077783289, + "grad_norm": 3.8032024391333397, + "learning_rate": 3.7517119685874358e-06, + "loss": 0.9778, + "step": 2057 + }, + { + "epoch": 0.18559769130179915, + "grad_norm": 1.4460276715861275, + "learning_rate": 3.7514299698285447e-06, + "loss": 1.0138, + "step": 2058 + }, + { + "epoch": 0.1856878748252694, + "grad_norm": 1.8805523995216942, + "learning_rate": 3.751147821627991e-06, + "loss": 1.0017, + "step": 2059 + }, + { + "epoch": 0.18577805834873967, + "grad_norm": 2.537959589937844, + "learning_rate": 3.75086552400985e-06, + "loss": 1.04, + "step": 2060 + }, + { + "epoch": 0.18586824187220996, + "grad_norm": 1.9605351981457346, + "learning_rate": 3.750583076998208e-06, + "loss": 0.9989, + "step": 2061 + }, + { + "epoch": 0.18595842539568022, + "grad_norm": 2.0933043731621903, + "learning_rate": 3.7503004806171655e-06, + "loss": 1.0006, + "step": 2062 + }, + { + "epoch": 0.18604860891915048, + "grad_norm": 1.6099792383170317, + "learning_rate": 3.7500177348908354e-06, + "loss": 0.9501, + "step": 2063 + }, + { + "epoch": 0.18613879244262074, + "grad_norm": 1.5666027373939773, + "learning_rate": 3.749734839843342e-06, + "loss": 1.0418, + "step": 2064 + }, + { + "epoch": 0.186228975966091, + "grad_norm": 1.5473857768461827, + "learning_rate": 3.7494517954988245e-06, + "loss": 1.0237, + "step": 2065 + }, + { + "epoch": 0.18631915948956126, + "grad_norm": 2.444073445410608, + "learning_rate": 3.749168601881433e-06, + "loss": 1.0157, + "step": 2066 + }, + { + "epoch": 0.18640934301303153, + "grad_norm": 2.0654394992748624, + "learning_rate": 3.7488852590153315e-06, + "loss": 1.0052, + "step": 2067 + }, + { + "epoch": 0.1864995265365018, + "grad_norm": 1.634134296029839, + "learning_rate": 3.748601766924697e-06, + "loss": 1.0186, + "step": 2068 + }, + { + "epoch": 0.18658971005997205, + "grad_norm": 1.8092837243125401, + "learning_rate": 3.7483181256337176e-06, + "loss": 0.9869, + "step": 2069 + }, + { + "epoch": 0.1866798935834423, + "grad_norm": 1.5845496959795646, + "learning_rate": 3.7480343351665962e-06, + "loss": 1.0253, + "step": 2070 + }, + { + "epoch": 0.18677007710691257, + "grad_norm": 2.438987473696941, + "learning_rate": 3.747750395547546e-06, + "loss": 1.0084, + "step": 2071 + }, + { + "epoch": 0.18686026063038283, + "grad_norm": 1.2900312078382659, + "learning_rate": 3.7474663068007956e-06, + "loss": 1.0075, + "step": 2072 + }, + { + "epoch": 0.1869504441538531, + "grad_norm": 1.2188353386369573, + "learning_rate": 3.747182068950584e-06, + "loss": 1.0045, + "step": 2073 + }, + { + "epoch": 0.18704062767732335, + "grad_norm": 1.7583020694116707, + "learning_rate": 3.7468976820211643e-06, + "loss": 1.0234, + "step": 2074 + }, + { + "epoch": 0.1871308112007936, + "grad_norm": 1.675094357010648, + "learning_rate": 3.746613146036803e-06, + "loss": 1.0244, + "step": 2075 + }, + { + "epoch": 0.18722099472426387, + "grad_norm": 2.197259657109951, + "learning_rate": 3.7463284610217766e-06, + "loss": 1.0621, + "step": 2076 + }, + { + "epoch": 0.18731117824773413, + "grad_norm": 1.6623464699891537, + "learning_rate": 3.746043627000377e-06, + "loss": 0.9553, + "step": 2077 + }, + { + "epoch": 0.1874013617712044, + "grad_norm": 1.542046070667368, + "learning_rate": 3.7457586439969076e-06, + "loss": 0.9666, + "step": 2078 + }, + { + "epoch": 0.18749154529467466, + "grad_norm": 1.3535144360029274, + "learning_rate": 3.7454735120356842e-06, + "loss": 1.0117, + "step": 2079 + }, + { + "epoch": 0.18758172881814492, + "grad_norm": 1.457319297646737, + "learning_rate": 3.7451882311410373e-06, + "loss": 1.087, + "step": 2080 + }, + { + "epoch": 0.18767191234161518, + "grad_norm": 1.5032138568207016, + "learning_rate": 3.7449028013373074e-06, + "loss": 1.0592, + "step": 2081 + }, + { + "epoch": 0.18776209586508544, + "grad_norm": 1.568247941976468, + "learning_rate": 3.7446172226488485e-06, + "loss": 1.0188, + "step": 2082 + }, + { + "epoch": 0.1878522793885557, + "grad_norm": 1.937037443505131, + "learning_rate": 3.7443314951000285e-06, + "loss": 1.0862, + "step": 2083 + }, + { + "epoch": 0.18794246291202596, + "grad_norm": 2.08778042765642, + "learning_rate": 3.7440456187152276e-06, + "loss": 1.0053, + "step": 2084 + }, + { + "epoch": 0.18803264643549625, + "grad_norm": 2.083245758759139, + "learning_rate": 3.7437595935188377e-06, + "loss": 1.1059, + "step": 2085 + }, + { + "epoch": 0.1881228299589665, + "grad_norm": 1.9404416057525908, + "learning_rate": 3.7434734195352647e-06, + "loss": 1.0207, + "step": 2086 + }, + { + "epoch": 0.18821301348243677, + "grad_norm": 0.6648206085947123, + "learning_rate": 3.743187096788926e-06, + "loss": 0.8186, + "step": 2087 + }, + { + "epoch": 0.18830319700590703, + "grad_norm": 1.7551969608953448, + "learning_rate": 3.7429006253042524e-06, + "loss": 1.0206, + "step": 2088 + }, + { + "epoch": 0.1883933805293773, + "grad_norm": 1.9422082606807245, + "learning_rate": 3.7426140051056867e-06, + "loss": 0.9673, + "step": 2089 + }, + { + "epoch": 0.18848356405284755, + "grad_norm": 1.6900500351084413, + "learning_rate": 3.7423272362176856e-06, + "loss": 0.9605, + "step": 2090 + }, + { + "epoch": 0.1885737475763178, + "grad_norm": 1.565284765130715, + "learning_rate": 3.742040318664718e-06, + "loss": 1.0647, + "step": 2091 + }, + { + "epoch": 0.18866393109978807, + "grad_norm": 1.6712977045479278, + "learning_rate": 3.7417532524712643e-06, + "loss": 1.0743, + "step": 2092 + }, + { + "epoch": 0.18875411462325833, + "grad_norm": 1.7399891895747617, + "learning_rate": 3.7414660376618195e-06, + "loss": 1.0451, + "step": 2093 + }, + { + "epoch": 0.1888442981467286, + "grad_norm": 0.5999113931397164, + "learning_rate": 3.74117867426089e-06, + "loss": 0.828, + "step": 2094 + }, + { + "epoch": 0.18893448167019886, + "grad_norm": 1.5897659920087404, + "learning_rate": 3.7408911622929954e-06, + "loss": 0.9961, + "step": 2095 + }, + { + "epoch": 0.18902466519366912, + "grad_norm": 1.4771811990460288, + "learning_rate": 3.740603501782668e-06, + "loss": 1.0523, + "step": 2096 + }, + { + "epoch": 0.18911484871713938, + "grad_norm": 2.2217782000823862, + "learning_rate": 3.7403156927544516e-06, + "loss": 0.9569, + "step": 2097 + }, + { + "epoch": 0.18920503224060964, + "grad_norm": 1.7902208601166578, + "learning_rate": 3.740027735232904e-06, + "loss": 0.9478, + "step": 2098 + }, + { + "epoch": 0.1892952157640799, + "grad_norm": 1.6395338517138118, + "learning_rate": 3.7397396292425966e-06, + "loss": 1.0042, + "step": 2099 + }, + { + "epoch": 0.18938539928755016, + "grad_norm": 2.327235211840799, + "learning_rate": 3.7394513748081105e-06, + "loss": 0.9977, + "step": 2100 + }, + { + "epoch": 0.18947558281102042, + "grad_norm": 1.4578757612724706, + "learning_rate": 3.7391629719540418e-06, + "loss": 1.05, + "step": 2101 + }, + { + "epoch": 0.18956576633449068, + "grad_norm": 1.5987946947425393, + "learning_rate": 3.7388744207049998e-06, + "loss": 1.0074, + "step": 2102 + }, + { + "epoch": 0.18965594985796094, + "grad_norm": 2.019631595442851, + "learning_rate": 3.7385857210856027e-06, + "loss": 0.9898, + "step": 2103 + }, + { + "epoch": 0.1897461333814312, + "grad_norm": 1.540554195115634, + "learning_rate": 3.738296873120486e-06, + "loss": 1.0049, + "step": 2104 + }, + { + "epoch": 0.18983631690490146, + "grad_norm": 4.345227326481186, + "learning_rate": 3.7380078768342955e-06, + "loss": 0.9897, + "step": 2105 + }, + { + "epoch": 0.18992650042837173, + "grad_norm": 2.220177849183897, + "learning_rate": 3.7377187322516895e-06, + "loss": 1.0357, + "step": 2106 + }, + { + "epoch": 0.19001668395184199, + "grad_norm": 1.6089841497670696, + "learning_rate": 3.7374294393973395e-06, + "loss": 0.9696, + "step": 2107 + }, + { + "epoch": 0.19010686747531225, + "grad_norm": 1.627820721270664, + "learning_rate": 3.7371399982959294e-06, + "loss": 1.0725, + "step": 2108 + }, + { + "epoch": 0.19019705099878254, + "grad_norm": 2.0067670304471683, + "learning_rate": 3.7368504089721565e-06, + "loss": 0.932, + "step": 2109 + }, + { + "epoch": 0.1902872345222528, + "grad_norm": 1.4880216276945353, + "learning_rate": 3.73656067145073e-06, + "loss": 0.9672, + "step": 2110 + }, + { + "epoch": 0.19037741804572306, + "grad_norm": 1.4498299597647462, + "learning_rate": 3.736270785756371e-06, + "loss": 1.0659, + "step": 2111 + }, + { + "epoch": 0.19046760156919332, + "grad_norm": 2.0751891084466085, + "learning_rate": 3.7359807519138156e-06, + "loss": 1.021, + "step": 2112 + }, + { + "epoch": 0.19055778509266358, + "grad_norm": 1.6891870189859006, + "learning_rate": 3.73569056994781e-06, + "loss": 1.0158, + "step": 2113 + }, + { + "epoch": 0.19064796861613384, + "grad_norm": 1.930437567724183, + "learning_rate": 3.7354002398831144e-06, + "loss": 1.0549, + "step": 2114 + }, + { + "epoch": 0.1907381521396041, + "grad_norm": 2.205125590270533, + "learning_rate": 3.7351097617445015e-06, + "loss": 1.0177, + "step": 2115 + }, + { + "epoch": 0.19082833566307436, + "grad_norm": 0.6913533163551603, + "learning_rate": 3.7348191355567567e-06, + "loss": 0.8495, + "step": 2116 + }, + { + "epoch": 0.19091851918654462, + "grad_norm": 1.9079261039883146, + "learning_rate": 3.734528361344677e-06, + "loss": 1.0477, + "step": 2117 + }, + { + "epoch": 0.19100870271001488, + "grad_norm": 1.7717888516805762, + "learning_rate": 3.734237439133074e-06, + "loss": 1.0151, + "step": 2118 + }, + { + "epoch": 0.19109888623348514, + "grad_norm": 1.7007501125332853, + "learning_rate": 3.7339463689467702e-06, + "loss": 1.0671, + "step": 2119 + }, + { + "epoch": 0.1911890697569554, + "grad_norm": 1.426792402846866, + "learning_rate": 3.733655150810601e-06, + "loss": 0.9891, + "step": 2120 + }, + { + "epoch": 0.19127925328042567, + "grad_norm": 2.185188270855226, + "learning_rate": 3.7333637847494154e-06, + "loss": 1.1494, + "step": 2121 + }, + { + "epoch": 0.19136943680389593, + "grad_norm": 2.0421111602641138, + "learning_rate": 3.7330722707880734e-06, + "loss": 1.1731, + "step": 2122 + }, + { + "epoch": 0.1914596203273662, + "grad_norm": 7.446446015251614, + "learning_rate": 3.7327806089514497e-06, + "loss": 0.9271, + "step": 2123 + }, + { + "epoch": 0.19154980385083645, + "grad_norm": 1.5296590675774293, + "learning_rate": 3.7324887992644297e-06, + "loss": 1.1092, + "step": 2124 + }, + { + "epoch": 0.1916399873743067, + "grad_norm": 0.7587745694576926, + "learning_rate": 3.7321968417519123e-06, + "loss": 0.8231, + "step": 2125 + }, + { + "epoch": 0.19173017089777697, + "grad_norm": 1.7078347137486296, + "learning_rate": 3.7319047364388097e-06, + "loss": 1.007, + "step": 2126 + }, + { + "epoch": 0.19182035442124723, + "grad_norm": 0.6071887534889873, + "learning_rate": 3.7316124833500453e-06, + "loss": 0.8091, + "step": 2127 + }, + { + "epoch": 0.1919105379447175, + "grad_norm": 1.4982021206583218, + "learning_rate": 3.731320082510556e-06, + "loss": 1.0003, + "step": 2128 + }, + { + "epoch": 0.19200072146818775, + "grad_norm": 3.5581886571794596, + "learning_rate": 3.7310275339452906e-06, + "loss": 1.0328, + "step": 2129 + }, + { + "epoch": 0.192090904991658, + "grad_norm": 1.9064408503815213, + "learning_rate": 3.7307348376792113e-06, + "loss": 1.0056, + "step": 2130 + }, + { + "epoch": 0.19218108851512827, + "grad_norm": 1.5157454924518772, + "learning_rate": 3.730441993737292e-06, + "loss": 1.0089, + "step": 2131 + }, + { + "epoch": 0.19227127203859856, + "grad_norm": 2.6999423303449874, + "learning_rate": 3.7301490021445205e-06, + "loss": 1.0013, + "step": 2132 + }, + { + "epoch": 0.19236145556206882, + "grad_norm": 1.941474652380551, + "learning_rate": 3.7298558629258966e-06, + "loss": 0.9584, + "step": 2133 + }, + { + "epoch": 0.19245163908553908, + "grad_norm": 1.5979533397274293, + "learning_rate": 3.7295625761064314e-06, + "loss": 1.0548, + "step": 2134 + }, + { + "epoch": 0.19254182260900934, + "grad_norm": 2.5216780158991634, + "learning_rate": 3.7292691417111504e-06, + "loss": 0.9793, + "step": 2135 + }, + { + "epoch": 0.1926320061324796, + "grad_norm": 2.267684421462885, + "learning_rate": 3.728975559765092e-06, + "loss": 0.9758, + "step": 2136 + }, + { + "epoch": 0.19272218965594987, + "grad_norm": 2.006180036519211, + "learning_rate": 3.728681830293305e-06, + "loss": 1.0499, + "step": 2137 + }, + { + "epoch": 0.19281237317942013, + "grad_norm": 1.8464694899541512, + "learning_rate": 3.7283879533208523e-06, + "loss": 0.9756, + "step": 2138 + }, + { + "epoch": 0.1929025567028904, + "grad_norm": 1.4188409650099896, + "learning_rate": 3.7280939288728094e-06, + "loss": 1.0029, + "step": 2139 + }, + { + "epoch": 0.19299274022636065, + "grad_norm": 1.7063254287312082, + "learning_rate": 3.7277997569742637e-06, + "loss": 0.981, + "step": 2140 + }, + { + "epoch": 0.1930829237498309, + "grad_norm": 2.3956422701958795, + "learning_rate": 3.7275054376503155e-06, + "loss": 0.9797, + "step": 2141 + }, + { + "epoch": 0.19317310727330117, + "grad_norm": 2.180997174757466, + "learning_rate": 3.7272109709260783e-06, + "loss": 0.9555, + "step": 2142 + }, + { + "epoch": 0.19326329079677143, + "grad_norm": 0.7560207381986663, + "learning_rate": 3.7269163568266774e-06, + "loss": 0.8257, + "step": 2143 + }, + { + "epoch": 0.1933534743202417, + "grad_norm": 1.8869795530148423, + "learning_rate": 3.7266215953772512e-06, + "loss": 1.0003, + "step": 2144 + }, + { + "epoch": 0.19344365784371195, + "grad_norm": 1.711934535094, + "learning_rate": 3.7263266866029492e-06, + "loss": 1.0505, + "step": 2145 + }, + { + "epoch": 0.1935338413671822, + "grad_norm": 1.4723435937686462, + "learning_rate": 3.726031630528936e-06, + "loss": 0.9812, + "step": 2146 + }, + { + "epoch": 0.19362402489065247, + "grad_norm": 1.393415330469658, + "learning_rate": 3.7257364271803865e-06, + "loss": 0.9078, + "step": 2147 + }, + { + "epoch": 0.19371420841412273, + "grad_norm": 2.169811009030715, + "learning_rate": 3.7254410765824896e-06, + "loss": 0.937, + "step": 2148 + }, + { + "epoch": 0.193804391937593, + "grad_norm": 2.067120088038918, + "learning_rate": 3.725145578760446e-06, + "loss": 0.9315, + "step": 2149 + }, + { + "epoch": 0.19389457546106326, + "grad_norm": 4.54260199945087, + "learning_rate": 3.7248499337394696e-06, + "loss": 0.929, + "step": 2150 + }, + { + "epoch": 0.19398475898453352, + "grad_norm": 1.8868329821138325, + "learning_rate": 3.7245541415447848e-06, + "loss": 1.103, + "step": 2151 + }, + { + "epoch": 0.19407494250800378, + "grad_norm": 1.8511437051098365, + "learning_rate": 3.724258202201633e-06, + "loss": 1.0401, + "step": 2152 + }, + { + "epoch": 0.19416512603147404, + "grad_norm": 1.8488340595729613, + "learning_rate": 3.7239621157352633e-06, + "loss": 1.03, + "step": 2153 + }, + { + "epoch": 0.1942553095549443, + "grad_norm": 1.9060005900331047, + "learning_rate": 3.7236658821709403e-06, + "loss": 1.0059, + "step": 2154 + }, + { + "epoch": 0.19434549307841456, + "grad_norm": 2.0197992202384087, + "learning_rate": 3.7233695015339404e-06, + "loss": 1.0528, + "step": 2155 + }, + { + "epoch": 0.19443567660188485, + "grad_norm": 1.8228015827000477, + "learning_rate": 3.7230729738495513e-06, + "loss": 1.0051, + "step": 2156 + }, + { + "epoch": 0.1945258601253551, + "grad_norm": 1.7042963741842965, + "learning_rate": 3.722776299143075e-06, + "loss": 1.0064, + "step": 2157 + }, + { + "epoch": 0.19461604364882537, + "grad_norm": 0.6680409799529171, + "learning_rate": 3.722479477439826e-06, + "loss": 0.8239, + "step": 2158 + }, + { + "epoch": 0.19470622717229563, + "grad_norm": 1.296805506303547, + "learning_rate": 3.7221825087651306e-06, + "loss": 0.99, + "step": 2159 + }, + { + "epoch": 0.1947964106957659, + "grad_norm": 1.7857924648609986, + "learning_rate": 3.7218853931443274e-06, + "loss": 0.9514, + "step": 2160 + }, + { + "epoch": 0.19488659421923615, + "grad_norm": 1.3585109814904475, + "learning_rate": 3.721588130602768e-06, + "loss": 0.9514, + "step": 2161 + }, + { + "epoch": 0.19497677774270641, + "grad_norm": 1.926986904996554, + "learning_rate": 3.7212907211658164e-06, + "loss": 0.9917, + "step": 2162 + }, + { + "epoch": 0.19506696126617667, + "grad_norm": 1.577755119865475, + "learning_rate": 3.72099316485885e-06, + "loss": 0.9785, + "step": 2163 + }, + { + "epoch": 0.19515714478964694, + "grad_norm": 1.8537012294699564, + "learning_rate": 3.720695461707256e-06, + "loss": 1.0461, + "step": 2164 + }, + { + "epoch": 0.1952473283131172, + "grad_norm": 1.664549205686406, + "learning_rate": 3.7203976117364383e-06, + "loss": 1.0282, + "step": 2165 + }, + { + "epoch": 0.19533751183658746, + "grad_norm": 1.9551101971567033, + "learning_rate": 3.7200996149718105e-06, + "loss": 0.9941, + "step": 2166 + }, + { + "epoch": 0.19542769536005772, + "grad_norm": 1.4923250624275741, + "learning_rate": 3.7198014714387985e-06, + "loss": 1.0005, + "step": 2167 + }, + { + "epoch": 0.19551787888352798, + "grad_norm": 1.82777062266198, + "learning_rate": 3.7195031811628422e-06, + "loss": 0.9446, + "step": 2168 + }, + { + "epoch": 0.19560806240699824, + "grad_norm": 3.691885457899924, + "learning_rate": 3.719204744169393e-06, + "loss": 1.0413, + "step": 2169 + }, + { + "epoch": 0.1956982459304685, + "grad_norm": 1.866111285604701, + "learning_rate": 3.718906160483916e-06, + "loss": 0.9878, + "step": 2170 + }, + { + "epoch": 0.19578842945393876, + "grad_norm": 1.712972884774709, + "learning_rate": 3.7186074301318868e-06, + "loss": 0.9885, + "step": 2171 + }, + { + "epoch": 0.19587861297740902, + "grad_norm": 1.5451418482529473, + "learning_rate": 3.7183085531387957e-06, + "loss": 1.0869, + "step": 2172 + }, + { + "epoch": 0.19596879650087928, + "grad_norm": 1.9093708069500872, + "learning_rate": 3.7180095295301443e-06, + "loss": 0.9837, + "step": 2173 + }, + { + "epoch": 0.19605898002434954, + "grad_norm": 1.3212473341843662, + "learning_rate": 3.7177103593314465e-06, + "loss": 1.073, + "step": 2174 + }, + { + "epoch": 0.1961491635478198, + "grad_norm": 1.7182842664006186, + "learning_rate": 3.7174110425682297e-06, + "loss": 1.0465, + "step": 2175 + }, + { + "epoch": 0.19623934707129007, + "grad_norm": 1.769655471719404, + "learning_rate": 3.7171115792660333e-06, + "loss": 1.138, + "step": 2176 + }, + { + "epoch": 0.19632953059476033, + "grad_norm": 2.458388591942726, + "learning_rate": 3.7168119694504083e-06, + "loss": 1.0792, + "step": 2177 + }, + { + "epoch": 0.1964197141182306, + "grad_norm": 3.8724518057654476, + "learning_rate": 3.71651221314692e-06, + "loss": 1.0067, + "step": 2178 + }, + { + "epoch": 0.19650989764170085, + "grad_norm": 1.6823286007191727, + "learning_rate": 3.716212310381145e-06, + "loss": 1.0089, + "step": 2179 + }, + { + "epoch": 0.19660008116517114, + "grad_norm": 1.6273424431497543, + "learning_rate": 3.7159122611786725e-06, + "loss": 1.0566, + "step": 2180 + }, + { + "epoch": 0.1966902646886414, + "grad_norm": 1.8031303656909778, + "learning_rate": 3.7156120655651045e-06, + "loss": 0.9956, + "step": 2181 + }, + { + "epoch": 0.19678044821211166, + "grad_norm": 1.4523917368265864, + "learning_rate": 3.7153117235660553e-06, + "loss": 0.9917, + "step": 2182 + }, + { + "epoch": 0.19687063173558192, + "grad_norm": 0.6958349386832474, + "learning_rate": 3.7150112352071514e-06, + "loss": 0.8229, + "step": 2183 + }, + { + "epoch": 0.19696081525905218, + "grad_norm": 1.752814754022845, + "learning_rate": 3.7147106005140326e-06, + "loss": 0.9793, + "step": 2184 + }, + { + "epoch": 0.19705099878252244, + "grad_norm": 0.6908488855770256, + "learning_rate": 3.714409819512351e-06, + "loss": 0.8321, + "step": 2185 + }, + { + "epoch": 0.1971411823059927, + "grad_norm": 2.555432785466841, + "learning_rate": 3.7141088922277695e-06, + "loss": 1.0395, + "step": 2186 + }, + { + "epoch": 0.19723136582946296, + "grad_norm": 2.1914811011058037, + "learning_rate": 3.7138078186859664e-06, + "loss": 1.0527, + "step": 2187 + }, + { + "epoch": 0.19732154935293322, + "grad_norm": 6.609087854374891, + "learning_rate": 3.7135065989126303e-06, + "loss": 1.0241, + "step": 2188 + }, + { + "epoch": 0.19741173287640348, + "grad_norm": 0.7268815519344279, + "learning_rate": 3.713205232933463e-06, + "loss": 0.8865, + "step": 2189 + }, + { + "epoch": 0.19750191639987374, + "grad_norm": 1.7937494709931234, + "learning_rate": 3.7129037207741792e-06, + "loss": 1.0197, + "step": 2190 + }, + { + "epoch": 0.197592099923344, + "grad_norm": 2.211919559699295, + "learning_rate": 3.7126020624605046e-06, + "loss": 1.0234, + "step": 2191 + }, + { + "epoch": 0.19768228344681427, + "grad_norm": 1.5161262548508925, + "learning_rate": 3.7123002580181785e-06, + "loss": 0.9663, + "step": 2192 + }, + { + "epoch": 0.19777246697028453, + "grad_norm": 1.3665089694586128, + "learning_rate": 3.7119983074729532e-06, + "loss": 1.0828, + "step": 2193 + }, + { + "epoch": 0.1978626504937548, + "grad_norm": 2.170947309004837, + "learning_rate": 3.7116962108505926e-06, + "loss": 1.068, + "step": 2194 + }, + { + "epoch": 0.19795283401722505, + "grad_norm": 1.480878466274532, + "learning_rate": 3.711393968176873e-06, + "loss": 1.1123, + "step": 2195 + }, + { + "epoch": 0.1980430175406953, + "grad_norm": 1.5147254083138477, + "learning_rate": 3.711091579477584e-06, + "loss": 1.0198, + "step": 2196 + }, + { + "epoch": 0.19813320106416557, + "grad_norm": 2.2338738579718593, + "learning_rate": 3.7107890447785255e-06, + "loss": 1.0439, + "step": 2197 + }, + { + "epoch": 0.19822338458763583, + "grad_norm": 1.827502478873141, + "learning_rate": 3.710486364105513e-06, + "loss": 1.065, + "step": 2198 + }, + { + "epoch": 0.1983135681111061, + "grad_norm": 3.465242416687279, + "learning_rate": 3.7101835374843728e-06, + "loss": 1.0103, + "step": 2199 + }, + { + "epoch": 0.19840375163457635, + "grad_norm": 1.8256710569324166, + "learning_rate": 3.7098805649409427e-06, + "loss": 1.0313, + "step": 2200 + }, + { + "epoch": 0.1984939351580466, + "grad_norm": 1.9817724507776813, + "learning_rate": 3.7095774465010748e-06, + "loss": 1.0828, + "step": 2201 + }, + { + "epoch": 0.19858411868151687, + "grad_norm": 2.1955008103593525, + "learning_rate": 3.7092741821906328e-06, + "loss": 1.013, + "step": 2202 + }, + { + "epoch": 0.19867430220498714, + "grad_norm": 3.1091747123582123, + "learning_rate": 3.708970772035493e-06, + "loss": 0.9948, + "step": 2203 + }, + { + "epoch": 0.19876448572845742, + "grad_norm": 1.6310967004294177, + "learning_rate": 3.7086672160615427e-06, + "loss": 0.9809, + "step": 2204 + }, + { + "epoch": 0.19885466925192768, + "grad_norm": 1.8598052096145312, + "learning_rate": 3.7083635142946852e-06, + "loss": 0.9989, + "step": 2205 + }, + { + "epoch": 0.19894485277539795, + "grad_norm": 1.8375532466932716, + "learning_rate": 3.7080596667608327e-06, + "loss": 1.051, + "step": 2206 + }, + { + "epoch": 0.1990350362988682, + "grad_norm": 1.5093037242922565, + "learning_rate": 3.707755673485911e-06, + "loss": 1.0579, + "step": 2207 + }, + { + "epoch": 0.19912521982233847, + "grad_norm": 2.395376233710561, + "learning_rate": 3.7074515344958584e-06, + "loss": 0.9272, + "step": 2208 + }, + { + "epoch": 0.19921540334580873, + "grad_norm": 0.6643875224242272, + "learning_rate": 3.707147249816627e-06, + "loss": 0.8515, + "step": 2209 + }, + { + "epoch": 0.199305586869279, + "grad_norm": 1.7205768932929568, + "learning_rate": 3.706842819474178e-06, + "loss": 1.0386, + "step": 2210 + }, + { + "epoch": 0.19939577039274925, + "grad_norm": 2.32105045592801, + "learning_rate": 3.706538243494489e-06, + "loss": 0.9541, + "step": 2211 + }, + { + "epoch": 0.1994859539162195, + "grad_norm": 2.2910464285546395, + "learning_rate": 3.706233521903547e-06, + "loss": 0.9863, + "step": 2212 + }, + { + "epoch": 0.19957613743968977, + "grad_norm": 2.436003690337845, + "learning_rate": 3.705928654727353e-06, + "loss": 1.0977, + "step": 2213 + }, + { + "epoch": 0.19966632096316003, + "grad_norm": 1.3612215501178524, + "learning_rate": 3.7056236419919195e-06, + "loss": 0.9883, + "step": 2214 + }, + { + "epoch": 0.1997565044866303, + "grad_norm": 1.450215451906665, + "learning_rate": 3.705318483723273e-06, + "loss": 1.0663, + "step": 2215 + }, + { + "epoch": 0.19984668801010055, + "grad_norm": 1.67355540311686, + "learning_rate": 3.7050131799474493e-06, + "loss": 0.9878, + "step": 2216 + }, + { + "epoch": 0.19993687153357081, + "grad_norm": 1.608465289433002, + "learning_rate": 3.7047077306905e-06, + "loss": 1.0155, + "step": 2217 + }, + { + "epoch": 0.20002705505704108, + "grad_norm": 1.4906511636353217, + "learning_rate": 3.704402135978488e-06, + "loss": 1.0491, + "step": 2218 + }, + { + "epoch": 0.20011723858051134, + "grad_norm": 2.2791757428850383, + "learning_rate": 3.7040963958374877e-06, + "loss": 1.0421, + "step": 2219 + }, + { + "epoch": 0.2002074221039816, + "grad_norm": 2.291570233714726, + "learning_rate": 3.7037905102935864e-06, + "loss": 1.0473, + "step": 2220 + }, + { + "epoch": 0.20029760562745186, + "grad_norm": 2.9112789164737194, + "learning_rate": 3.7034844793728837e-06, + "loss": 1.0242, + "step": 2221 + }, + { + "epoch": 0.20038778915092212, + "grad_norm": 1.4553878556938349, + "learning_rate": 3.7031783031014933e-06, + "loss": 0.9056, + "step": 2222 + }, + { + "epoch": 0.20047797267439238, + "grad_norm": 1.683044343950091, + "learning_rate": 3.702871981505538e-06, + "loss": 0.9961, + "step": 2223 + }, + { + "epoch": 0.20056815619786264, + "grad_norm": 1.7058458205851394, + "learning_rate": 3.7025655146111563e-06, + "loss": 0.9576, + "step": 2224 + }, + { + "epoch": 0.2006583397213329, + "grad_norm": 0.6814947764713551, + "learning_rate": 3.702258902444497e-06, + "loss": 0.9019, + "step": 2225 + }, + { + "epoch": 0.20074852324480316, + "grad_norm": 1.4437026672531001, + "learning_rate": 3.701952145031722e-06, + "loss": 1.0667, + "step": 2226 + }, + { + "epoch": 0.20083870676827342, + "grad_norm": 1.6215557864717873, + "learning_rate": 3.701645242399005e-06, + "loss": 1.0985, + "step": 2227 + }, + { + "epoch": 0.2009288902917437, + "grad_norm": 4.280413246525989, + "learning_rate": 3.701338194572533e-06, + "loss": 1.0841, + "step": 2228 + }, + { + "epoch": 0.20101907381521397, + "grad_norm": 1.6815700183892448, + "learning_rate": 3.7010310015785056e-06, + "loss": 1.0, + "step": 2229 + }, + { + "epoch": 0.20110925733868423, + "grad_norm": 1.4969266560029415, + "learning_rate": 3.700723663443134e-06, + "loss": 0.9623, + "step": 2230 + }, + { + "epoch": 0.2011994408621545, + "grad_norm": 0.7718272931881207, + "learning_rate": 3.7004161801926416e-06, + "loss": 0.8294, + "step": 2231 + }, + { + "epoch": 0.20128962438562475, + "grad_norm": 1.8867620301268115, + "learning_rate": 3.7001085518532643e-06, + "loss": 1.0405, + "step": 2232 + }, + { + "epoch": 0.20137980790909502, + "grad_norm": 1.5572843859788348, + "learning_rate": 3.6998007784512515e-06, + "loss": 1.0113, + "step": 2233 + }, + { + "epoch": 0.20146999143256528, + "grad_norm": 1.3710516681353688, + "learning_rate": 3.6994928600128637e-06, + "loss": 0.8891, + "step": 2234 + }, + { + "epoch": 0.20156017495603554, + "grad_norm": 1.8445954404953053, + "learning_rate": 3.6991847965643742e-06, + "loss": 0.9821, + "step": 2235 + }, + { + "epoch": 0.2016503584795058, + "grad_norm": 2.35475616051216, + "learning_rate": 3.698876588132068e-06, + "loss": 0.9709, + "step": 2236 + }, + { + "epoch": 0.20174054200297606, + "grad_norm": 1.7408353065539193, + "learning_rate": 3.6985682347422446e-06, + "loss": 0.9755, + "step": 2237 + }, + { + "epoch": 0.20183072552644632, + "grad_norm": 2.0396568630577434, + "learning_rate": 3.698259736421213e-06, + "loss": 1.0274, + "step": 2238 + }, + { + "epoch": 0.20192090904991658, + "grad_norm": 1.671793071129801, + "learning_rate": 3.697951093195297e-06, + "loss": 0.9944, + "step": 2239 + }, + { + "epoch": 0.20201109257338684, + "grad_norm": 1.733758077788289, + "learning_rate": 3.6976423050908307e-06, + "loss": 0.9877, + "step": 2240 + }, + { + "epoch": 0.2021012760968571, + "grad_norm": 1.7089850027900633, + "learning_rate": 3.697333372134163e-06, + "loss": 1.0009, + "step": 2241 + }, + { + "epoch": 0.20219145962032736, + "grad_norm": 1.8420671444727779, + "learning_rate": 3.697024294351653e-06, + "loss": 1.1059, + "step": 2242 + }, + { + "epoch": 0.20228164314379762, + "grad_norm": 1.5565594468912978, + "learning_rate": 3.696715071769672e-06, + "loss": 0.9756, + "step": 2243 + }, + { + "epoch": 0.20237182666726788, + "grad_norm": 2.3079831111296647, + "learning_rate": 3.696405704414606e-06, + "loss": 1.0152, + "step": 2244 + }, + { + "epoch": 0.20246201019073815, + "grad_norm": 1.4270689283697404, + "learning_rate": 3.6960961923128514e-06, + "loss": 1.1182, + "step": 2245 + }, + { + "epoch": 0.2025521937142084, + "grad_norm": 2.7927873459172257, + "learning_rate": 3.6957865354908177e-06, + "loss": 0.9443, + "step": 2246 + }, + { + "epoch": 0.20264237723767867, + "grad_norm": 1.765201686650172, + "learning_rate": 3.6954767339749262e-06, + "loss": 1.0313, + "step": 2247 + }, + { + "epoch": 0.20273256076114893, + "grad_norm": 2.911337388767463, + "learning_rate": 3.6951667877916113e-06, + "loss": 0.9928, + "step": 2248 + }, + { + "epoch": 0.2028227442846192, + "grad_norm": 2.01068041031708, + "learning_rate": 3.694856696967319e-06, + "loss": 1.0194, + "step": 2249 + }, + { + "epoch": 0.20291292780808945, + "grad_norm": 1.7264026550592506, + "learning_rate": 3.6945464615285077e-06, + "loss": 1.0016, + "step": 2250 + }, + { + "epoch": 0.2030031113315597, + "grad_norm": 0.705149998264168, + "learning_rate": 3.694236081501648e-06, + "loss": 0.9301, + "step": 2251 + }, + { + "epoch": 0.20309329485503, + "grad_norm": 1.5408285629393923, + "learning_rate": 3.6939255569132246e-06, + "loss": 0.9763, + "step": 2252 + }, + { + "epoch": 0.20318347837850026, + "grad_norm": 0.9005925744365818, + "learning_rate": 3.693614887789733e-06, + "loss": 0.8569, + "step": 2253 + }, + { + "epoch": 0.20327366190197052, + "grad_norm": 0.7865614202444238, + "learning_rate": 3.69330407415768e-06, + "loss": 0.8586, + "step": 2254 + }, + { + "epoch": 0.20336384542544078, + "grad_norm": 1.912557703282123, + "learning_rate": 3.6929931160435867e-06, + "loss": 1.0214, + "step": 2255 + }, + { + "epoch": 0.20345402894891104, + "grad_norm": 1.8519134832652178, + "learning_rate": 3.6926820134739858e-06, + "loss": 1.0699, + "step": 2256 + }, + { + "epoch": 0.2035442124723813, + "grad_norm": 1.6862511428817824, + "learning_rate": 3.692370766475422e-06, + "loss": 0.9459, + "step": 2257 + }, + { + "epoch": 0.20363439599585156, + "grad_norm": 1.7186836229858595, + "learning_rate": 3.692059375074453e-06, + "loss": 1.102, + "step": 2258 + }, + { + "epoch": 0.20372457951932182, + "grad_norm": 1.6131673777296536, + "learning_rate": 3.6917478392976475e-06, + "loss": 1.0811, + "step": 2259 + }, + { + "epoch": 0.20381476304279209, + "grad_norm": 1.6220513514375676, + "learning_rate": 3.691436159171589e-06, + "loss": 1.0764, + "step": 2260 + }, + { + "epoch": 0.20390494656626235, + "grad_norm": 4.08431478629642, + "learning_rate": 3.6911243347228703e-06, + "loss": 0.9973, + "step": 2261 + }, + { + "epoch": 0.2039951300897326, + "grad_norm": 0.63789243119703, + "learning_rate": 3.690812365978099e-06, + "loss": 0.7706, + "step": 2262 + }, + { + "epoch": 0.20408531361320287, + "grad_norm": 1.6488872475000411, + "learning_rate": 3.690500252963893e-06, + "loss": 0.9701, + "step": 2263 + }, + { + "epoch": 0.20417549713667313, + "grad_norm": 1.5383760112133935, + "learning_rate": 3.6901879957068846e-06, + "loss": 1.0061, + "step": 2264 + }, + { + "epoch": 0.2042656806601434, + "grad_norm": 13.838634226006949, + "learning_rate": 3.689875594233717e-06, + "loss": 1.103, + "step": 2265 + }, + { + "epoch": 0.20435586418361365, + "grad_norm": 1.691452131463985, + "learning_rate": 3.689563048571046e-06, + "loss": 0.9985, + "step": 2266 + }, + { + "epoch": 0.2044460477070839, + "grad_norm": 1.7214548501845046, + "learning_rate": 3.6892503587455395e-06, + "loss": 1.0073, + "step": 2267 + }, + { + "epoch": 0.20453623123055417, + "grad_norm": 1.9902442579987996, + "learning_rate": 3.6889375247838766e-06, + "loss": 1.0231, + "step": 2268 + }, + { + "epoch": 0.20462641475402443, + "grad_norm": 1.4186308184739769, + "learning_rate": 3.688624546712753e-06, + "loss": 1.0119, + "step": 2269 + }, + { + "epoch": 0.2047165982774947, + "grad_norm": 3.065760686673319, + "learning_rate": 3.688311424558871e-06, + "loss": 0.9695, + "step": 2270 + }, + { + "epoch": 0.20480678180096495, + "grad_norm": 2.070839638447773, + "learning_rate": 3.6879981583489496e-06, + "loss": 0.9162, + "step": 2271 + }, + { + "epoch": 0.20489696532443522, + "grad_norm": 1.7051046922149895, + "learning_rate": 3.687684748109718e-06, + "loss": 1.0301, + "step": 2272 + }, + { + "epoch": 0.20498714884790548, + "grad_norm": 2.079544729695997, + "learning_rate": 3.6873711938679174e-06, + "loss": 1.0755, + "step": 2273 + }, + { + "epoch": 0.20507733237137574, + "grad_norm": 1.442608257178365, + "learning_rate": 3.6870574956503027e-06, + "loss": 1.0013, + "step": 2274 + }, + { + "epoch": 0.20516751589484603, + "grad_norm": 2.3030447584897114, + "learning_rate": 3.68674365348364e-06, + "loss": 1.0048, + "step": 2275 + }, + { + "epoch": 0.2052576994183163, + "grad_norm": 1.635027172901074, + "learning_rate": 3.6864296673947086e-06, + "loss": 1.0228, + "step": 2276 + }, + { + "epoch": 0.20534788294178655, + "grad_norm": 1.540111203823153, + "learning_rate": 3.686115537410298e-06, + "loss": 0.9955, + "step": 2277 + }, + { + "epoch": 0.2054380664652568, + "grad_norm": 1.596527820329231, + "learning_rate": 3.685801263557214e-06, + "loss": 0.9598, + "step": 2278 + }, + { + "epoch": 0.20552824998872707, + "grad_norm": 1.915067447839096, + "learning_rate": 3.68548684586227e-06, + "loss": 1.1031, + "step": 2279 + }, + { + "epoch": 0.20561843351219733, + "grad_norm": 7.4517841532515146, + "learning_rate": 3.685172284352295e-06, + "loss": 1.0137, + "step": 2280 + }, + { + "epoch": 0.2057086170356676, + "grad_norm": 1.294378060006172, + "learning_rate": 3.684857579054128e-06, + "loss": 0.9468, + "step": 2281 + }, + { + "epoch": 0.20579880055913785, + "grad_norm": 1.8331347560265288, + "learning_rate": 3.6845427299946233e-06, + "loss": 1.0758, + "step": 2282 + }, + { + "epoch": 0.2058889840826081, + "grad_norm": 1.594798173482101, + "learning_rate": 3.6842277372006434e-06, + "loss": 1.1131, + "step": 2283 + }, + { + "epoch": 0.20597916760607837, + "grad_norm": 1.7010496601478342, + "learning_rate": 3.6839126006990664e-06, + "loss": 0.9857, + "step": 2284 + }, + { + "epoch": 0.20606935112954863, + "grad_norm": 1.4338427750919804, + "learning_rate": 3.6835973205167818e-06, + "loss": 1.0705, + "step": 2285 + }, + { + "epoch": 0.2061595346530189, + "grad_norm": 1.7749657452998677, + "learning_rate": 3.6832818966806904e-06, + "loss": 1.0525, + "step": 2286 + }, + { + "epoch": 0.20624971817648916, + "grad_norm": 1.8207017212301497, + "learning_rate": 3.682966329217706e-06, + "loss": 1.0945, + "step": 2287 + }, + { + "epoch": 0.20633990169995942, + "grad_norm": 1.7010680209670592, + "learning_rate": 3.6826506181547543e-06, + "loss": 1.0516, + "step": 2288 + }, + { + "epoch": 0.20643008522342968, + "grad_norm": 1.7881834710863647, + "learning_rate": 3.682334763518774e-06, + "loss": 1.0373, + "step": 2289 + }, + { + "epoch": 0.20652026874689994, + "grad_norm": 2.104358953874193, + "learning_rate": 3.6820187653367158e-06, + "loss": 0.9825, + "step": 2290 + }, + { + "epoch": 0.2066104522703702, + "grad_norm": 1.5710860287757806, + "learning_rate": 3.6817026236355412e-06, + "loss": 0.906, + "step": 2291 + }, + { + "epoch": 0.20670063579384046, + "grad_norm": 2.3361334574998063, + "learning_rate": 3.681386338442227e-06, + "loss": 0.9961, + "step": 2292 + }, + { + "epoch": 0.20679081931731072, + "grad_norm": 1.7305702196945316, + "learning_rate": 3.681069909783758e-06, + "loss": 0.9757, + "step": 2293 + }, + { + "epoch": 0.20688100284078098, + "grad_norm": 2.351293808297256, + "learning_rate": 3.680753337687136e-06, + "loss": 1.0223, + "step": 2294 + }, + { + "epoch": 0.20697118636425124, + "grad_norm": 1.909155273281398, + "learning_rate": 3.680436622179371e-06, + "loss": 0.9457, + "step": 2295 + }, + { + "epoch": 0.2070613698877215, + "grad_norm": 1.6541373976545037, + "learning_rate": 3.680119763287488e-06, + "loss": 0.9945, + "step": 2296 + }, + { + "epoch": 0.20715155341119176, + "grad_norm": 2.112356607660564, + "learning_rate": 3.6798027610385227e-06, + "loss": 0.9521, + "step": 2297 + }, + { + "epoch": 0.20724173693466202, + "grad_norm": 2.040856635367077, + "learning_rate": 3.6794856154595235e-06, + "loss": 1.0595, + "step": 2298 + }, + { + "epoch": 0.2073319204581323, + "grad_norm": 2.0177967289152767, + "learning_rate": 3.6791683265775506e-06, + "loss": 1.0193, + "step": 2299 + }, + { + "epoch": 0.20742210398160257, + "grad_norm": 1.8127428089483746, + "learning_rate": 3.6788508944196773e-06, + "loss": 0.975, + "step": 2300 + }, + { + "epoch": 0.20751228750507283, + "grad_norm": 1.523527485072789, + "learning_rate": 3.678533319012989e-06, + "loss": 1.0894, + "step": 2301 + }, + { + "epoch": 0.2076024710285431, + "grad_norm": 1.7811364254798858, + "learning_rate": 3.6782156003845826e-06, + "loss": 0.9667, + "step": 2302 + }, + { + "epoch": 0.20769265455201336, + "grad_norm": 0.6508253780941965, + "learning_rate": 3.6778977385615676e-06, + "loss": 0.8325, + "step": 2303 + }, + { + "epoch": 0.20778283807548362, + "grad_norm": 1.7559081209776284, + "learning_rate": 3.6775797335710656e-06, + "loss": 1.0633, + "step": 2304 + }, + { + "epoch": 0.20787302159895388, + "grad_norm": 1.8190128778774584, + "learning_rate": 3.6772615854402105e-06, + "loss": 0.9608, + "step": 2305 + }, + { + "epoch": 0.20796320512242414, + "grad_norm": 1.3661475013473552, + "learning_rate": 3.6769432941961487e-06, + "loss": 0.8939, + "step": 2306 + }, + { + "epoch": 0.2080533886458944, + "grad_norm": 1.97610106378093, + "learning_rate": 3.676624859866038e-06, + "loss": 1.0141, + "step": 2307 + }, + { + "epoch": 0.20814357216936466, + "grad_norm": 2.6732273360315246, + "learning_rate": 3.67630628247705e-06, + "loss": 0.9384, + "step": 2308 + }, + { + "epoch": 0.20823375569283492, + "grad_norm": 1.7164668784808739, + "learning_rate": 3.675987562056367e-06, + "loss": 1.0737, + "step": 2309 + }, + { + "epoch": 0.20832393921630518, + "grad_norm": 1.4244219427336697, + "learning_rate": 3.675668698631184e-06, + "loss": 1.0509, + "step": 2310 + }, + { + "epoch": 0.20841412273977544, + "grad_norm": 1.555247067145639, + "learning_rate": 3.675349692228708e-06, + "loss": 1.0781, + "step": 2311 + }, + { + "epoch": 0.2085043062632457, + "grad_norm": 1.7288963458936342, + "learning_rate": 3.6750305428761578e-06, + "loss": 0.9871, + "step": 2312 + }, + { + "epoch": 0.20859448978671596, + "grad_norm": 2.2253032134828854, + "learning_rate": 3.674711250600766e-06, + "loss": 1.0202, + "step": 2313 + }, + { + "epoch": 0.20868467331018623, + "grad_norm": 0.6634438044805058, + "learning_rate": 3.6743918154297765e-06, + "loss": 0.8137, + "step": 2314 + }, + { + "epoch": 0.20877485683365649, + "grad_norm": 1.751538485958787, + "learning_rate": 3.6740722373904446e-06, + "loss": 0.9988, + "step": 2315 + }, + { + "epoch": 0.20886504035712675, + "grad_norm": 1.5598942199936443, + "learning_rate": 3.6737525165100383e-06, + "loss": 0.9899, + "step": 2316 + }, + { + "epoch": 0.208955223880597, + "grad_norm": 1.735595960693512, + "learning_rate": 3.6734326528158385e-06, + "loss": 1.0333, + "step": 2317 + }, + { + "epoch": 0.20904540740406727, + "grad_norm": 1.3594490118815643, + "learning_rate": 3.673112646335138e-06, + "loss": 1.0014, + "step": 2318 + }, + { + "epoch": 0.20913559092753753, + "grad_norm": 1.6378929278273149, + "learning_rate": 3.672792497095241e-06, + "loss": 0.9598, + "step": 2319 + }, + { + "epoch": 0.2092257744510078, + "grad_norm": 2.3147175440660948, + "learning_rate": 3.672472205123464e-06, + "loss": 1.0119, + "step": 2320 + }, + { + "epoch": 0.20931595797447805, + "grad_norm": 5.1308727576433775, + "learning_rate": 3.6721517704471363e-06, + "loss": 0.9417, + "step": 2321 + }, + { + "epoch": 0.2094061414979483, + "grad_norm": 1.5672449924804603, + "learning_rate": 3.6718311930936e-06, + "loss": 1.0469, + "step": 2322 + }, + { + "epoch": 0.2094963250214186, + "grad_norm": 1.9685954078650985, + "learning_rate": 3.6715104730902074e-06, + "loss": 0.9684, + "step": 2323 + }, + { + "epoch": 0.20958650854488886, + "grad_norm": 1.7770624294445343, + "learning_rate": 3.671189610464325e-06, + "loss": 1.0479, + "step": 2324 + }, + { + "epoch": 0.20967669206835912, + "grad_norm": 1.5646397244805226, + "learning_rate": 3.6708686052433303e-06, + "loss": 0.9635, + "step": 2325 + }, + { + "epoch": 0.20976687559182938, + "grad_norm": 1.6987958009003725, + "learning_rate": 3.6705474574546127e-06, + "loss": 1.0277, + "step": 2326 + }, + { + "epoch": 0.20985705911529964, + "grad_norm": 1.5716140439449502, + "learning_rate": 3.670226167125575e-06, + "loss": 1.0287, + "step": 2327 + }, + { + "epoch": 0.2099472426387699, + "grad_norm": 1.7430374740669194, + "learning_rate": 3.6699047342836313e-06, + "loss": 1.0168, + "step": 2328 + }, + { + "epoch": 0.21003742616224017, + "grad_norm": 1.4352167489512992, + "learning_rate": 3.669583158956208e-06, + "loss": 1.0483, + "step": 2329 + }, + { + "epoch": 0.21012760968571043, + "grad_norm": 0.6979197791490376, + "learning_rate": 3.669261441170743e-06, + "loss": 0.8457, + "step": 2330 + }, + { + "epoch": 0.2102177932091807, + "grad_norm": 2.4674334335237598, + "learning_rate": 3.668939580954688e-06, + "loss": 1.0236, + "step": 2331 + }, + { + "epoch": 0.21030797673265095, + "grad_norm": 1.4996452706828605, + "learning_rate": 3.668617578335506e-06, + "loss": 0.9879, + "step": 2332 + }, + { + "epoch": 0.2103981602561212, + "grad_norm": 1.894729132002207, + "learning_rate": 3.6682954333406707e-06, + "loss": 1.0426, + "step": 2333 + }, + { + "epoch": 0.21048834377959147, + "grad_norm": 2.283946285905572, + "learning_rate": 3.6679731459976707e-06, + "loss": 1.146, + "step": 2334 + }, + { + "epoch": 0.21057852730306173, + "grad_norm": 2.3101882720781677, + "learning_rate": 3.6676507163340046e-06, + "loss": 1.0112, + "step": 2335 + }, + { + "epoch": 0.210668710826532, + "grad_norm": 1.9462008611788308, + "learning_rate": 3.6673281443771842e-06, + "loss": 1.0439, + "step": 2336 + }, + { + "epoch": 0.21075889435000225, + "grad_norm": 1.5254655690166359, + "learning_rate": 3.667005430154733e-06, + "loss": 0.9053, + "step": 2337 + }, + { + "epoch": 0.2108490778734725, + "grad_norm": 0.6625619427495376, + "learning_rate": 3.666682573694186e-06, + "loss": 0.9184, + "step": 2338 + }, + { + "epoch": 0.21093926139694277, + "grad_norm": 0.6513803643719662, + "learning_rate": 3.6663595750230924e-06, + "loss": 0.7878, + "step": 2339 + }, + { + "epoch": 0.21102944492041303, + "grad_norm": 3.1106123259510174, + "learning_rate": 3.666036434169012e-06, + "loss": 1.0663, + "step": 2340 + }, + { + "epoch": 0.2111196284438833, + "grad_norm": 1.5233440663956395, + "learning_rate": 3.665713151159516e-06, + "loss": 0.9964, + "step": 2341 + }, + { + "epoch": 0.21120981196735356, + "grad_norm": 1.5625527182268633, + "learning_rate": 3.665389726022189e-06, + "loss": 1.1136, + "step": 2342 + }, + { + "epoch": 0.21129999549082382, + "grad_norm": 1.6537526574261248, + "learning_rate": 3.6650661587846283e-06, + "loss": 0.9895, + "step": 2343 + }, + { + "epoch": 0.21139017901429408, + "grad_norm": 2.1468999736282863, + "learning_rate": 3.6647424494744418e-06, + "loss": 0.9803, + "step": 2344 + }, + { + "epoch": 0.21148036253776434, + "grad_norm": 1.5502898929595472, + "learning_rate": 3.6644185981192503e-06, + "loss": 1.0643, + "step": 2345 + }, + { + "epoch": 0.2115705460612346, + "grad_norm": 1.9234124213783175, + "learning_rate": 3.6640946047466868e-06, + "loss": 0.9437, + "step": 2346 + }, + { + "epoch": 0.2116607295847049, + "grad_norm": 1.654435657503221, + "learning_rate": 3.6637704693843953e-06, + "loss": 0.9811, + "step": 2347 + }, + { + "epoch": 0.21175091310817515, + "grad_norm": 1.633602480543497, + "learning_rate": 3.6634461920600337e-06, + "loss": 0.9498, + "step": 2348 + }, + { + "epoch": 0.2118410966316454, + "grad_norm": 2.0683913043820343, + "learning_rate": 3.66312177280127e-06, + "loss": 1.0372, + "step": 2349 + }, + { + "epoch": 0.21193128015511567, + "grad_norm": 1.7156091602476677, + "learning_rate": 3.6627972116357872e-06, + "loss": 1.0549, + "step": 2350 + }, + { + "epoch": 0.21202146367858593, + "grad_norm": 1.6949955516520505, + "learning_rate": 3.662472508591278e-06, + "loss": 1.0757, + "step": 2351 + }, + { + "epoch": 0.2121116472020562, + "grad_norm": 1.6404370154811991, + "learning_rate": 3.662147663695447e-06, + "loss": 1.0025, + "step": 2352 + }, + { + "epoch": 0.21220183072552645, + "grad_norm": 1.5450444034257167, + "learning_rate": 3.6618226769760127e-06, + "loss": 1.0075, + "step": 2353 + }, + { + "epoch": 0.2122920142489967, + "grad_norm": 2.3717055309200443, + "learning_rate": 3.661497548460704e-06, + "loss": 0.8945, + "step": 2354 + }, + { + "epoch": 0.21238219777246697, + "grad_norm": 1.8594709740240407, + "learning_rate": 3.6611722781772635e-06, + "loss": 1.0323, + "step": 2355 + }, + { + "epoch": 0.21247238129593723, + "grad_norm": 0.8260340654585461, + "learning_rate": 3.6608468661534444e-06, + "loss": 0.882, + "step": 2356 + }, + { + "epoch": 0.2125625648194075, + "grad_norm": 2.0245504374741947, + "learning_rate": 3.660521312417013e-06, + "loss": 0.9394, + "step": 2357 + }, + { + "epoch": 0.21265274834287776, + "grad_norm": 1.7218164312852353, + "learning_rate": 3.660195616995747e-06, + "loss": 0.9842, + "step": 2358 + }, + { + "epoch": 0.21274293186634802, + "grad_norm": 1.8163696285370783, + "learning_rate": 3.6598697799174367e-06, + "loss": 1.0434, + "step": 2359 + }, + { + "epoch": 0.21283311538981828, + "grad_norm": 2.099805132589363, + "learning_rate": 3.6595438012098844e-06, + "loss": 1.0735, + "step": 2360 + }, + { + "epoch": 0.21292329891328854, + "grad_norm": 1.5816069610924899, + "learning_rate": 3.6592176809009045e-06, + "loss": 0.9836, + "step": 2361 + }, + { + "epoch": 0.2130134824367588, + "grad_norm": 1.6147524427226256, + "learning_rate": 3.6588914190183227e-06, + "loss": 1.0518, + "step": 2362 + }, + { + "epoch": 0.21310366596022906, + "grad_norm": 1.8244375877713586, + "learning_rate": 3.658565015589978e-06, + "loss": 0.8819, + "step": 2363 + }, + { + "epoch": 0.21319384948369932, + "grad_norm": 5.622580113703118, + "learning_rate": 3.6582384706437217e-06, + "loss": 1.0215, + "step": 2364 + }, + { + "epoch": 0.21328403300716958, + "grad_norm": 2.178529274418756, + "learning_rate": 3.6579117842074156e-06, + "loss": 1.0299, + "step": 2365 + }, + { + "epoch": 0.21337421653063984, + "grad_norm": 1.5120734062440095, + "learning_rate": 3.657584956308934e-06, + "loss": 0.9512, + "step": 2366 + }, + { + "epoch": 0.2134644000541101, + "grad_norm": 1.7895273995530834, + "learning_rate": 3.6572579869761648e-06, + "loss": 1.0819, + "step": 2367 + }, + { + "epoch": 0.21355458357758036, + "grad_norm": 1.769770052461471, + "learning_rate": 3.6569308762370056e-06, + "loss": 1.0642, + "step": 2368 + }, + { + "epoch": 0.21364476710105063, + "grad_norm": 1.785980435019676, + "learning_rate": 3.6566036241193676e-06, + "loss": 1.053, + "step": 2369 + }, + { + "epoch": 0.2137349506245209, + "grad_norm": 0.6162373435866056, + "learning_rate": 3.656276230651174e-06, + "loss": 0.8462, + "step": 2370 + }, + { + "epoch": 0.21382513414799117, + "grad_norm": 0.7073879290671713, + "learning_rate": 3.65594869586036e-06, + "loss": 0.8795, + "step": 2371 + }, + { + "epoch": 0.21391531767146144, + "grad_norm": 1.4590937766201175, + "learning_rate": 3.6556210197748724e-06, + "loss": 0.9324, + "step": 2372 + }, + { + "epoch": 0.2140055011949317, + "grad_norm": 1.674310781536739, + "learning_rate": 3.655293202422671e-06, + "loss": 0.9811, + "step": 2373 + }, + { + "epoch": 0.21409568471840196, + "grad_norm": 1.3370831133988008, + "learning_rate": 3.654965243831725e-06, + "loss": 1.0256, + "step": 2374 + }, + { + "epoch": 0.21418586824187222, + "grad_norm": 1.8207862466736862, + "learning_rate": 3.65463714403002e-06, + "loss": 1.0547, + "step": 2375 + }, + { + "epoch": 0.21427605176534248, + "grad_norm": 1.640955355853466, + "learning_rate": 3.65430890304555e-06, + "loss": 0.9685, + "step": 2376 + }, + { + "epoch": 0.21436623528881274, + "grad_norm": 1.4854738685469167, + "learning_rate": 3.653980520906323e-06, + "loss": 1.0419, + "step": 2377 + }, + { + "epoch": 0.214456418812283, + "grad_norm": 2.3147735760158996, + "learning_rate": 3.653651997640358e-06, + "loss": 0.9914, + "step": 2378 + }, + { + "epoch": 0.21454660233575326, + "grad_norm": 2.00593948575724, + "learning_rate": 3.653323333275686e-06, + "loss": 1.0171, + "step": 2379 + }, + { + "epoch": 0.21463678585922352, + "grad_norm": 5.979537402986199, + "learning_rate": 3.652994527840351e-06, + "loss": 0.9969, + "step": 2380 + }, + { + "epoch": 0.21472696938269378, + "grad_norm": 2.507049539558169, + "learning_rate": 3.6526655813624087e-06, + "loss": 0.979, + "step": 2381 + }, + { + "epoch": 0.21481715290616404, + "grad_norm": 0.5765638925178893, + "learning_rate": 3.652336493869925e-06, + "loss": 0.8391, + "step": 2382 + }, + { + "epoch": 0.2149073364296343, + "grad_norm": 1.8687537738672635, + "learning_rate": 3.6520072653909823e-06, + "loss": 1.0223, + "step": 2383 + }, + { + "epoch": 0.21499751995310457, + "grad_norm": 2.171482078083029, + "learning_rate": 3.6516778959536702e-06, + "loss": 1.0096, + "step": 2384 + }, + { + "epoch": 0.21508770347657483, + "grad_norm": 1.8970179646897687, + "learning_rate": 3.6513483855860923e-06, + "loss": 1.0605, + "step": 2385 + }, + { + "epoch": 0.2151778870000451, + "grad_norm": 1.7198285966678493, + "learning_rate": 3.6510187343163654e-06, + "loss": 0.9554, + "step": 2386 + }, + { + "epoch": 0.21526807052351535, + "grad_norm": 1.9253685685092274, + "learning_rate": 3.650688942172616e-06, + "loss": 1.0437, + "step": 2387 + }, + { + "epoch": 0.2153582540469856, + "grad_norm": 2.3742205696061327, + "learning_rate": 3.650359009182984e-06, + "loss": 0.9518, + "step": 2388 + }, + { + "epoch": 0.21544843757045587, + "grad_norm": 2.0387761291288973, + "learning_rate": 3.650028935375622e-06, + "loss": 1.0939, + "step": 2389 + }, + { + "epoch": 0.21553862109392613, + "grad_norm": 1.4221372257769043, + "learning_rate": 3.6496987207786926e-06, + "loss": 0.9783, + "step": 2390 + }, + { + "epoch": 0.2156288046173964, + "grad_norm": 2.104100506437995, + "learning_rate": 3.6493683654203724e-06, + "loss": 0.9603, + "step": 2391 + }, + { + "epoch": 0.21571898814086665, + "grad_norm": 1.7715962804674734, + "learning_rate": 3.6490378693288484e-06, + "loss": 0.9959, + "step": 2392 + }, + { + "epoch": 0.2158091716643369, + "grad_norm": 1.8229838404313283, + "learning_rate": 3.648707232532321e-06, + "loss": 1.051, + "step": 2393 + }, + { + "epoch": 0.2158993551878072, + "grad_norm": 2.073117755995691, + "learning_rate": 3.6483764550590017e-06, + "loss": 1.0111, + "step": 2394 + }, + { + "epoch": 0.21598953871127746, + "grad_norm": 1.5768765336194408, + "learning_rate": 3.6480455369371133e-06, + "loss": 1.0324, + "step": 2395 + }, + { + "epoch": 0.21607972223474772, + "grad_norm": 1.9600329497545532, + "learning_rate": 3.647714478194893e-06, + "loss": 1.0179, + "step": 2396 + }, + { + "epoch": 0.21616990575821798, + "grad_norm": 3.3971511686444087, + "learning_rate": 3.647383278860588e-06, + "loss": 0.9585, + "step": 2397 + }, + { + "epoch": 0.21626008928168824, + "grad_norm": 1.8565430432340144, + "learning_rate": 3.6470519389624587e-06, + "loss": 1.0029, + "step": 2398 + }, + { + "epoch": 0.2163502728051585, + "grad_norm": 1.5619955392937457, + "learning_rate": 3.646720458528776e-06, + "loss": 1.0732, + "step": 2399 + }, + { + "epoch": 0.21644045632862877, + "grad_norm": 2.0354247401170755, + "learning_rate": 3.6463888375878235e-06, + "loss": 0.9846, + "step": 2400 + }, + { + "epoch": 0.21653063985209903, + "grad_norm": 4.026992797360114, + "learning_rate": 3.646057076167897e-06, + "loss": 1.0286, + "step": 2401 + }, + { + "epoch": 0.2166208233755693, + "grad_norm": 1.5588608132919646, + "learning_rate": 3.645725174297305e-06, + "loss": 1.0195, + "step": 2402 + }, + { + "epoch": 0.21671100689903955, + "grad_norm": 1.7241028401487, + "learning_rate": 3.645393132004367e-06, + "loss": 0.9919, + "step": 2403 + }, + { + "epoch": 0.2168011904225098, + "grad_norm": 1.7619160543789316, + "learning_rate": 3.6450609493174135e-06, + "loss": 0.9714, + "step": 2404 + }, + { + "epoch": 0.21689137394598007, + "grad_norm": 1.796600586213262, + "learning_rate": 3.6447286262647896e-06, + "loss": 0.9767, + "step": 2405 + }, + { + "epoch": 0.21698155746945033, + "grad_norm": 2.286171711978164, + "learning_rate": 3.64439616287485e-06, + "loss": 0.9783, + "step": 2406 + }, + { + "epoch": 0.2170717409929206, + "grad_norm": 1.4504313780927909, + "learning_rate": 3.644063559175963e-06, + "loss": 0.9799, + "step": 2407 + }, + { + "epoch": 0.21716192451639085, + "grad_norm": 1.7937777154888028, + "learning_rate": 3.6437308151965074e-06, + "loss": 0.9933, + "step": 2408 + }, + { + "epoch": 0.2172521080398611, + "grad_norm": 1.4820559405581166, + "learning_rate": 3.643397930964876e-06, + "loss": 1.0268, + "step": 2409 + }, + { + "epoch": 0.21734229156333137, + "grad_norm": 1.7626936217746394, + "learning_rate": 3.6430649065094707e-06, + "loss": 1.0763, + "step": 2410 + }, + { + "epoch": 0.21743247508680164, + "grad_norm": 2.032979023920424, + "learning_rate": 3.6427317418587086e-06, + "loss": 1.0782, + "step": 2411 + }, + { + "epoch": 0.2175226586102719, + "grad_norm": 1.8774772809275668, + "learning_rate": 3.6423984370410157e-06, + "loss": 0.9543, + "step": 2412 + }, + { + "epoch": 0.21761284213374216, + "grad_norm": 1.5619569979327579, + "learning_rate": 3.6420649920848324e-06, + "loss": 0.9793, + "step": 2413 + }, + { + "epoch": 0.21770302565721242, + "grad_norm": 1.5237874729471204, + "learning_rate": 3.6417314070186096e-06, + "loss": 1.0203, + "step": 2414 + }, + { + "epoch": 0.21779320918068268, + "grad_norm": 2.929221317076274, + "learning_rate": 3.641397681870811e-06, + "loss": 1.02, + "step": 2415 + }, + { + "epoch": 0.21788339270415294, + "grad_norm": 1.5806614349618668, + "learning_rate": 3.641063816669911e-06, + "loss": 1.028, + "step": 2416 + }, + { + "epoch": 0.2179735762276232, + "grad_norm": 1.6928227156115547, + "learning_rate": 3.640729811444398e-06, + "loss": 1.0705, + "step": 2417 + }, + { + "epoch": 0.2180637597510935, + "grad_norm": 1.558397323736129, + "learning_rate": 3.6403956662227706e-06, + "loss": 1.0131, + "step": 2418 + }, + { + "epoch": 0.21815394327456375, + "grad_norm": 1.6092121819661118, + "learning_rate": 3.6400613810335396e-06, + "loss": 1.0619, + "step": 2419 + }, + { + "epoch": 0.218244126798034, + "grad_norm": 1.7185328259677883, + "learning_rate": 3.639726955905228e-06, + "loss": 0.9277, + "step": 2420 + }, + { + "epoch": 0.21833431032150427, + "grad_norm": 1.668628070963069, + "learning_rate": 3.639392390866372e-06, + "loss": 1.0407, + "step": 2421 + }, + { + "epoch": 0.21842449384497453, + "grad_norm": 2.73863646724187, + "learning_rate": 3.639057685945517e-06, + "loss": 0.9923, + "step": 2422 + }, + { + "epoch": 0.2185146773684448, + "grad_norm": 1.4959738262508187, + "learning_rate": 3.638722841171223e-06, + "loss": 0.9041, + "step": 2423 + }, + { + "epoch": 0.21860486089191505, + "grad_norm": 1.5569242556455285, + "learning_rate": 3.638387856572061e-06, + "loss": 1.0511, + "step": 2424 + }, + { + "epoch": 0.21869504441538531, + "grad_norm": 0.6545999081449855, + "learning_rate": 3.638052732176612e-06, + "loss": 0.8456, + "step": 2425 + }, + { + "epoch": 0.21878522793885558, + "grad_norm": 2.199857520778301, + "learning_rate": 3.637717468013472e-06, + "loss": 1.0256, + "step": 2426 + }, + { + "epoch": 0.21887541146232584, + "grad_norm": 0.59742116356128, + "learning_rate": 3.6373820641112475e-06, + "loss": 0.7969, + "step": 2427 + }, + { + "epoch": 0.2189655949857961, + "grad_norm": 1.8015307275595687, + "learning_rate": 3.6370465204985567e-06, + "loss": 1.0301, + "step": 2428 + }, + { + "epoch": 0.21905577850926636, + "grad_norm": 0.6325216920246524, + "learning_rate": 3.6367108372040304e-06, + "loss": 0.858, + "step": 2429 + }, + { + "epoch": 0.21914596203273662, + "grad_norm": 1.7528131217771812, + "learning_rate": 3.6363750142563107e-06, + "loss": 1.0171, + "step": 2430 + }, + { + "epoch": 0.21923614555620688, + "grad_norm": 0.715207158669332, + "learning_rate": 3.636039051684052e-06, + "loss": 0.894, + "step": 2431 + }, + { + "epoch": 0.21932632907967714, + "grad_norm": 1.6571335505120848, + "learning_rate": 3.6357029495159203e-06, + "loss": 0.9771, + "step": 2432 + }, + { + "epoch": 0.2194165126031474, + "grad_norm": 1.6629457263451297, + "learning_rate": 3.6353667077805934e-06, + "loss": 0.9873, + "step": 2433 + }, + { + "epoch": 0.21950669612661766, + "grad_norm": 1.9700797903293872, + "learning_rate": 3.6350303265067625e-06, + "loss": 1.0538, + "step": 2434 + }, + { + "epoch": 0.21959687965008792, + "grad_norm": 1.6096351191107892, + "learning_rate": 3.6346938057231285e-06, + "loss": 1.0437, + "step": 2435 + }, + { + "epoch": 0.21968706317355818, + "grad_norm": 1.6218046169407694, + "learning_rate": 3.6343571454584047e-06, + "loss": 0.9561, + "step": 2436 + }, + { + "epoch": 0.21977724669702844, + "grad_norm": 1.5963303863430216, + "learning_rate": 3.6340203457413176e-06, + "loss": 1.0488, + "step": 2437 + }, + { + "epoch": 0.2198674302204987, + "grad_norm": 1.6148679010793425, + "learning_rate": 3.633683406600605e-06, + "loss": 1.0377, + "step": 2438 + }, + { + "epoch": 0.21995761374396897, + "grad_norm": 1.5954915423526224, + "learning_rate": 3.6333463280650165e-06, + "loss": 1.0757, + "step": 2439 + }, + { + "epoch": 0.22004779726743923, + "grad_norm": 0.7892535516576208, + "learning_rate": 3.6330091101633126e-06, + "loss": 0.8067, + "step": 2440 + }, + { + "epoch": 0.2201379807909095, + "grad_norm": 1.8238072717470832, + "learning_rate": 3.632671752924267e-06, + "loss": 1.0526, + "step": 2441 + }, + { + "epoch": 0.22022816431437978, + "grad_norm": 2.0376200634738777, + "learning_rate": 3.632334256376665e-06, + "loss": 1.0155, + "step": 2442 + }, + { + "epoch": 0.22031834783785004, + "grad_norm": 1.5422133516040655, + "learning_rate": 3.6319966205493044e-06, + "loss": 1.0718, + "step": 2443 + }, + { + "epoch": 0.2204085313613203, + "grad_norm": 0.7234963148032438, + "learning_rate": 3.6316588454709922e-06, + "loss": 0.8907, + "step": 2444 + }, + { + "epoch": 0.22049871488479056, + "grad_norm": 1.5010772651441422, + "learning_rate": 3.6313209311705514e-06, + "loss": 0.9654, + "step": 2445 + }, + { + "epoch": 0.22058889840826082, + "grad_norm": 2.976065847058057, + "learning_rate": 3.6309828776768133e-06, + "loss": 1.0531, + "step": 2446 + }, + { + "epoch": 0.22067908193173108, + "grad_norm": 2.0043611660781178, + "learning_rate": 3.630644685018623e-06, + "loss": 0.9764, + "step": 2447 + }, + { + "epoch": 0.22076926545520134, + "grad_norm": 2.34434959371413, + "learning_rate": 3.6303063532248367e-06, + "loss": 1.0978, + "step": 2448 + }, + { + "epoch": 0.2208594489786716, + "grad_norm": 1.5844005033407975, + "learning_rate": 3.6299678823243236e-06, + "loss": 1.097, + "step": 2449 + }, + { + "epoch": 0.22094963250214186, + "grad_norm": 1.5937413982084432, + "learning_rate": 3.629629272345963e-06, + "loss": 0.9207, + "step": 2450 + }, + { + "epoch": 0.22103981602561212, + "grad_norm": 1.7113818980643505, + "learning_rate": 3.6292905233186468e-06, + "loss": 1.1345, + "step": 2451 + }, + { + "epoch": 0.22112999954908238, + "grad_norm": 1.6634060276686873, + "learning_rate": 3.6289516352712796e-06, + "loss": 0.9727, + "step": 2452 + }, + { + "epoch": 0.22122018307255265, + "grad_norm": 0.7068203774958642, + "learning_rate": 3.6286126082327764e-06, + "loss": 0.8409, + "step": 2453 + }, + { + "epoch": 0.2213103665960229, + "grad_norm": 1.5192444047622382, + "learning_rate": 3.628273442232066e-06, + "loss": 1.0453, + "step": 2454 + }, + { + "epoch": 0.22140055011949317, + "grad_norm": 1.463085614847156, + "learning_rate": 3.627934137298087e-06, + "loss": 1.01, + "step": 2455 + }, + { + "epoch": 0.22149073364296343, + "grad_norm": 1.9127650008764487, + "learning_rate": 3.627594693459792e-06, + "loss": 0.9939, + "step": 2456 + }, + { + "epoch": 0.2215809171664337, + "grad_norm": 1.647362959955175, + "learning_rate": 3.6272551107461424e-06, + "loss": 0.9782, + "step": 2457 + }, + { + "epoch": 0.22167110068990395, + "grad_norm": 1.571476367434118, + "learning_rate": 3.6269153891861137e-06, + "loss": 1.0118, + "step": 2458 + }, + { + "epoch": 0.2217612842133742, + "grad_norm": 1.335822563358011, + "learning_rate": 3.6265755288086944e-06, + "loss": 0.984, + "step": 2459 + }, + { + "epoch": 0.22185146773684447, + "grad_norm": 1.8719146138427212, + "learning_rate": 3.626235529642881e-06, + "loss": 0.9994, + "step": 2460 + }, + { + "epoch": 0.22194165126031473, + "grad_norm": 2.005835721508845, + "learning_rate": 3.625895391717686e-06, + "loss": 0.9524, + "step": 2461 + }, + { + "epoch": 0.222031834783785, + "grad_norm": 1.5668049825659716, + "learning_rate": 3.625555115062131e-06, + "loss": 0.9858, + "step": 2462 + }, + { + "epoch": 0.22212201830725525, + "grad_norm": 1.8969270953545716, + "learning_rate": 3.6252146997052507e-06, + "loss": 1.083, + "step": 2463 + }, + { + "epoch": 0.22221220183072551, + "grad_norm": 0.6615759802495884, + "learning_rate": 3.6248741456760898e-06, + "loss": 0.7981, + "step": 2464 + }, + { + "epoch": 0.22230238535419578, + "grad_norm": 1.8578441312389486, + "learning_rate": 3.624533453003708e-06, + "loss": 0.9976, + "step": 2465 + }, + { + "epoch": 0.22239256887766606, + "grad_norm": 1.7207976887900678, + "learning_rate": 3.6241926217171745e-06, + "loss": 1.0239, + "step": 2466 + }, + { + "epoch": 0.22248275240113632, + "grad_norm": 1.721890441492978, + "learning_rate": 3.6238516518455703e-06, + "loss": 0.9339, + "step": 2467 + }, + { + "epoch": 0.22257293592460659, + "grad_norm": 2.13650628799738, + "learning_rate": 3.62351054341799e-06, + "loss": 1.1379, + "step": 2468 + }, + { + "epoch": 0.22266311944807685, + "grad_norm": 1.7929031925225545, + "learning_rate": 3.623169296463538e-06, + "loss": 1.0425, + "step": 2469 + }, + { + "epoch": 0.2227533029715471, + "grad_norm": 4.388867878494728, + "learning_rate": 3.6228279110113316e-06, + "loss": 1.0156, + "step": 2470 + }, + { + "epoch": 0.22284348649501737, + "grad_norm": 1.882532692532908, + "learning_rate": 3.6224863870904994e-06, + "loss": 1.0252, + "step": 2471 + }, + { + "epoch": 0.22293367001848763, + "grad_norm": 2.052624497163322, + "learning_rate": 3.6221447247301827e-06, + "loss": 1.0279, + "step": 2472 + }, + { + "epoch": 0.2230238535419579, + "grad_norm": 1.4564621046422428, + "learning_rate": 3.6218029239595332e-06, + "loss": 0.9646, + "step": 2473 + }, + { + "epoch": 0.22311403706542815, + "grad_norm": 1.7999200644227975, + "learning_rate": 3.621460984807716e-06, + "loss": 1.0417, + "step": 2474 + }, + { + "epoch": 0.2232042205888984, + "grad_norm": 1.611082929971289, + "learning_rate": 3.621118907303907e-06, + "loss": 0.9858, + "step": 2475 + }, + { + "epoch": 0.22329440411236867, + "grad_norm": 5.121457061566725, + "learning_rate": 3.620776691477294e-06, + "loss": 1.0159, + "step": 2476 + }, + { + "epoch": 0.22338458763583893, + "grad_norm": 0.6832745706095237, + "learning_rate": 3.6204343373570765e-06, + "loss": 0.7888, + "step": 2477 + }, + { + "epoch": 0.2234747711593092, + "grad_norm": 1.8302730102343168, + "learning_rate": 3.620091844972467e-06, + "loss": 1.041, + "step": 2478 + }, + { + "epoch": 0.22356495468277945, + "grad_norm": 1.5586163129583552, + "learning_rate": 3.619749214352688e-06, + "loss": 1.037, + "step": 2479 + }, + { + "epoch": 0.22365513820624972, + "grad_norm": 1.7395769069651914, + "learning_rate": 3.6194064455269744e-06, + "loss": 1.1192, + "step": 2480 + }, + { + "epoch": 0.22374532172971998, + "grad_norm": 2.257043621785926, + "learning_rate": 3.6190635385245737e-06, + "loss": 0.935, + "step": 2481 + }, + { + "epoch": 0.22383550525319024, + "grad_norm": 1.6611959634894884, + "learning_rate": 3.618720493374745e-06, + "loss": 0.9484, + "step": 2482 + }, + { + "epoch": 0.2239256887766605, + "grad_norm": 1.7341952531478828, + "learning_rate": 3.6183773101067575e-06, + "loss": 0.8726, + "step": 2483 + }, + { + "epoch": 0.22401587230013076, + "grad_norm": 1.5108417658176885, + "learning_rate": 3.6180339887498948e-06, + "loss": 0.9369, + "step": 2484 + }, + { + "epoch": 0.22410605582360102, + "grad_norm": 1.6406167892977206, + "learning_rate": 3.61769052933345e-06, + "loss": 1.0171, + "step": 2485 + }, + { + "epoch": 0.22419623934707128, + "grad_norm": 1.970416075535045, + "learning_rate": 3.6173469318867297e-06, + "loss": 1.1706, + "step": 2486 + }, + { + "epoch": 0.22428642287054154, + "grad_norm": 1.569382045521905, + "learning_rate": 3.617003196439051e-06, + "loss": 1.0339, + "step": 2487 + }, + { + "epoch": 0.2243766063940118, + "grad_norm": 1.3459147382431917, + "learning_rate": 3.616659323019744e-06, + "loss": 1.0281, + "step": 2488 + }, + { + "epoch": 0.22446678991748206, + "grad_norm": 1.5415711072568945, + "learning_rate": 3.616315311658149e-06, + "loss": 1.0455, + "step": 2489 + }, + { + "epoch": 0.22455697344095235, + "grad_norm": 1.681914091304477, + "learning_rate": 3.6159711623836195e-06, + "loss": 1.0749, + "step": 2490 + }, + { + "epoch": 0.2246471569644226, + "grad_norm": 0.6454720537723974, + "learning_rate": 3.6156268752255203e-06, + "loss": 0.8023, + "step": 2491 + }, + { + "epoch": 0.22473734048789287, + "grad_norm": 1.5566414673921418, + "learning_rate": 3.615282450213227e-06, + "loss": 0.9824, + "step": 2492 + }, + { + "epoch": 0.22482752401136313, + "grad_norm": 1.6323513128032963, + "learning_rate": 3.614937887376128e-06, + "loss": 1.0328, + "step": 2493 + }, + { + "epoch": 0.2249177075348334, + "grad_norm": 1.8654239258643746, + "learning_rate": 3.614593186743625e-06, + "loss": 1.0293, + "step": 2494 + }, + { + "epoch": 0.22500789105830366, + "grad_norm": 2.282308150185054, + "learning_rate": 3.614248348345128e-06, + "loss": 0.9489, + "step": 2495 + }, + { + "epoch": 0.22509807458177392, + "grad_norm": 1.7030665536353682, + "learning_rate": 3.6139033722100614e-06, + "loss": 1.0691, + "step": 2496 + }, + { + "epoch": 0.22518825810524418, + "grad_norm": 1.962453491728937, + "learning_rate": 3.6135582583678596e-06, + "loss": 1.0928, + "step": 2497 + }, + { + "epoch": 0.22527844162871444, + "grad_norm": 2.3498287463183196, + "learning_rate": 3.61321300684797e-06, + "loss": 1.0316, + "step": 2498 + }, + { + "epoch": 0.2253686251521847, + "grad_norm": 2.9114680872954533, + "learning_rate": 3.6128676176798527e-06, + "loss": 1.0436, + "step": 2499 + }, + { + "epoch": 0.22545880867565496, + "grad_norm": 1.4021764883956942, + "learning_rate": 3.612522090892976e-06, + "loss": 1.0376, + "step": 2500 + }, + { + "epoch": 0.22554899219912522, + "grad_norm": 1.6658034552903005, + "learning_rate": 3.6121764265168232e-06, + "loss": 0.9842, + "step": 2501 + }, + { + "epoch": 0.22563917572259548, + "grad_norm": 1.5930876758104728, + "learning_rate": 3.611830624580888e-06, + "loss": 1.0412, + "step": 2502 + }, + { + "epoch": 0.22572935924606574, + "grad_norm": 1.7481574849345147, + "learning_rate": 3.6114846851146767e-06, + "loss": 1.0289, + "step": 2503 + }, + { + "epoch": 0.225819542769536, + "grad_norm": 1.6119169142067045, + "learning_rate": 3.6111386081477068e-06, + "loss": 0.9209, + "step": 2504 + }, + { + "epoch": 0.22590972629300626, + "grad_norm": 1.7003175186506627, + "learning_rate": 3.6107923937095066e-06, + "loss": 1.0381, + "step": 2505 + }, + { + "epoch": 0.22599990981647652, + "grad_norm": 2.096901415048784, + "learning_rate": 3.6104460418296173e-06, + "loss": 0.9398, + "step": 2506 + }, + { + "epoch": 0.22609009333994678, + "grad_norm": 1.7796777177362642, + "learning_rate": 3.6100995525375924e-06, + "loss": 1.0678, + "step": 2507 + }, + { + "epoch": 0.22618027686341705, + "grad_norm": 1.513003727284925, + "learning_rate": 3.6097529258629952e-06, + "loss": 0.963, + "step": 2508 + }, + { + "epoch": 0.2262704603868873, + "grad_norm": 1.9708504749329556, + "learning_rate": 3.6094061618354027e-06, + "loss": 1.0472, + "step": 2509 + }, + { + "epoch": 0.22636064391035757, + "grad_norm": 1.7468419189596391, + "learning_rate": 3.609059260484402e-06, + "loss": 1.0153, + "step": 2510 + }, + { + "epoch": 0.22645082743382783, + "grad_norm": 3.543365759092612, + "learning_rate": 3.6087122218395935e-06, + "loss": 0.9632, + "step": 2511 + }, + { + "epoch": 0.2265410109572981, + "grad_norm": 2.5411861048751616, + "learning_rate": 3.608365045930587e-06, + "loss": 0.9862, + "step": 2512 + }, + { + "epoch": 0.22663119448076835, + "grad_norm": 2.2421779300073426, + "learning_rate": 3.608017732787007e-06, + "loss": 1.0011, + "step": 2513 + }, + { + "epoch": 0.22672137800423864, + "grad_norm": 1.6019086510215403, + "learning_rate": 3.6076702824384875e-06, + "loss": 0.9478, + "step": 2514 + }, + { + "epoch": 0.2268115615277089, + "grad_norm": 1.6127427798115324, + "learning_rate": 3.607322694914675e-06, + "loss": 1.033, + "step": 2515 + }, + { + "epoch": 0.22690174505117916, + "grad_norm": 1.539916600285139, + "learning_rate": 3.606974970245227e-06, + "loss": 0.9879, + "step": 2516 + }, + { + "epoch": 0.22699192857464942, + "grad_norm": 1.600891568211722, + "learning_rate": 3.606627108459814e-06, + "loss": 0.9654, + "step": 2517 + }, + { + "epoch": 0.22708211209811968, + "grad_norm": 13.723374545319851, + "learning_rate": 3.6062791095881174e-06, + "loss": 0.944, + "step": 2518 + }, + { + "epoch": 0.22717229562158994, + "grad_norm": 2.1722724605755266, + "learning_rate": 3.6059309736598303e-06, + "loss": 1.0886, + "step": 2519 + }, + { + "epoch": 0.2272624791450602, + "grad_norm": 2.519438039548846, + "learning_rate": 3.605582700704657e-06, + "loss": 0.9918, + "step": 2520 + }, + { + "epoch": 0.22735266266853046, + "grad_norm": 1.8409361623235636, + "learning_rate": 3.6052342907523146e-06, + "loss": 1.0186, + "step": 2521 + }, + { + "epoch": 0.22744284619200072, + "grad_norm": 1.5819216089756059, + "learning_rate": 3.604885743832532e-06, + "loss": 0.9486, + "step": 2522 + }, + { + "epoch": 0.22753302971547099, + "grad_norm": 0.6883390032193749, + "learning_rate": 3.6045370599750482e-06, + "loss": 0.8183, + "step": 2523 + }, + { + "epoch": 0.22762321323894125, + "grad_norm": 1.6374058703675045, + "learning_rate": 3.604188239209615e-06, + "loss": 0.9996, + "step": 2524 + }, + { + "epoch": 0.2277133967624115, + "grad_norm": 3.0964514757441597, + "learning_rate": 3.603839281565996e-06, + "loss": 0.9403, + "step": 2525 + }, + { + "epoch": 0.22780358028588177, + "grad_norm": 2.037157828028154, + "learning_rate": 3.603490187073966e-06, + "loss": 1.0361, + "step": 2526 + }, + { + "epoch": 0.22789376380935203, + "grad_norm": 2.105993088807796, + "learning_rate": 3.6031409557633117e-06, + "loss": 1.0725, + "step": 2527 + }, + { + "epoch": 0.2279839473328223, + "grad_norm": 1.6513329699201085, + "learning_rate": 3.602791587663831e-06, + "loss": 1.0593, + "step": 2528 + }, + { + "epoch": 0.22807413085629255, + "grad_norm": 1.5441179462507368, + "learning_rate": 3.6024420828053348e-06, + "loss": 1.0554, + "step": 2529 + }, + { + "epoch": 0.2281643143797628, + "grad_norm": 1.5978979846536985, + "learning_rate": 3.6020924412176445e-06, + "loss": 1.0894, + "step": 2530 + }, + { + "epoch": 0.22825449790323307, + "grad_norm": 1.56081917956782, + "learning_rate": 3.601742662930593e-06, + "loss": 0.9902, + "step": 2531 + }, + { + "epoch": 0.22834468142670333, + "grad_norm": 1.6920554520784257, + "learning_rate": 3.6013927479740248e-06, + "loss": 0.9797, + "step": 2532 + }, + { + "epoch": 0.2284348649501736, + "grad_norm": 2.057759704144291, + "learning_rate": 3.6010426963777985e-06, + "loss": 1.0969, + "step": 2533 + }, + { + "epoch": 0.22852504847364385, + "grad_norm": 3.014326062897701, + "learning_rate": 3.6006925081717804e-06, + "loss": 1.0612, + "step": 2534 + }, + { + "epoch": 0.22861523199711412, + "grad_norm": 2.2037940125552047, + "learning_rate": 3.600342183385852e-06, + "loss": 0.9854, + "step": 2535 + }, + { + "epoch": 0.22870541552058438, + "grad_norm": 1.5572391445867355, + "learning_rate": 3.5999917220499043e-06, + "loss": 0.9553, + "step": 2536 + }, + { + "epoch": 0.22879559904405466, + "grad_norm": 2.3957698537184418, + "learning_rate": 3.5996411241938404e-06, + "loss": 1.0189, + "step": 2537 + }, + { + "epoch": 0.22888578256752493, + "grad_norm": 1.6267485381196487, + "learning_rate": 3.5992903898475752e-06, + "loss": 0.9764, + "step": 2538 + }, + { + "epoch": 0.2289759660909952, + "grad_norm": 1.7582028782015282, + "learning_rate": 3.5989395190410365e-06, + "loss": 1.1153, + "step": 2539 + }, + { + "epoch": 0.22906614961446545, + "grad_norm": 1.5835621149848782, + "learning_rate": 3.598588511804161e-06, + "loss": 1.0218, + "step": 2540 + }, + { + "epoch": 0.2291563331379357, + "grad_norm": 5.998454530676364, + "learning_rate": 3.5982373681668987e-06, + "loss": 1.019, + "step": 2541 + }, + { + "epoch": 0.22924651666140597, + "grad_norm": 1.747885243705889, + "learning_rate": 3.597886088159212e-06, + "loss": 1.0271, + "step": 2542 + }, + { + "epoch": 0.22933670018487623, + "grad_norm": 1.3881604504341698, + "learning_rate": 3.597534671811074e-06, + "loss": 0.9426, + "step": 2543 + }, + { + "epoch": 0.2294268837083465, + "grad_norm": 1.7937981177066757, + "learning_rate": 3.5971831191524684e-06, + "loss": 0.9952, + "step": 2544 + }, + { + "epoch": 0.22951706723181675, + "grad_norm": 1.4956740943639806, + "learning_rate": 3.5968314302133925e-06, + "loss": 1.0651, + "step": 2545 + }, + { + "epoch": 0.229607250755287, + "grad_norm": 1.9123729489080659, + "learning_rate": 3.596479605023854e-06, + "loss": 1.0162, + "step": 2546 + }, + { + "epoch": 0.22969743427875727, + "grad_norm": 1.91372814761492, + "learning_rate": 3.596127643613873e-06, + "loss": 1.0311, + "step": 2547 + }, + { + "epoch": 0.22978761780222753, + "grad_norm": 1.4664093196391685, + "learning_rate": 3.59577554601348e-06, + "loss": 1.0395, + "step": 2548 + }, + { + "epoch": 0.2298778013256978, + "grad_norm": 2.1862153777166817, + "learning_rate": 3.595423312252719e-06, + "loss": 1.0649, + "step": 2549 + }, + { + "epoch": 0.22996798484916806, + "grad_norm": 2.0772962710380187, + "learning_rate": 3.5950709423616436e-06, + "loss": 0.9989, + "step": 2550 + }, + { + "epoch": 0.23005816837263832, + "grad_norm": 1.8616011300272477, + "learning_rate": 3.5947184363703203e-06, + "loss": 0.9918, + "step": 2551 + }, + { + "epoch": 0.23014835189610858, + "grad_norm": 3.247884942455054, + "learning_rate": 3.5943657943088274e-06, + "loss": 0.9647, + "step": 2552 + }, + { + "epoch": 0.23023853541957884, + "grad_norm": 1.3594888661368152, + "learning_rate": 3.5940130162072525e-06, + "loss": 0.943, + "step": 2553 + }, + { + "epoch": 0.2303287189430491, + "grad_norm": 1.9189022612656448, + "learning_rate": 3.5936601020956985e-06, + "loss": 1.0148, + "step": 2554 + }, + { + "epoch": 0.23041890246651936, + "grad_norm": 2.7934157647357245, + "learning_rate": 3.5933070520042772e-06, + "loss": 1.049, + "step": 2555 + }, + { + "epoch": 0.23050908598998962, + "grad_norm": 0.790243039953421, + "learning_rate": 3.5929538659631133e-06, + "loss": 0.8686, + "step": 2556 + }, + { + "epoch": 0.23059926951345988, + "grad_norm": 1.5553421865691734, + "learning_rate": 3.592600544002341e-06, + "loss": 1.0495, + "step": 2557 + }, + { + "epoch": 0.23068945303693014, + "grad_norm": 3.8733869086897554, + "learning_rate": 3.5922470861521098e-06, + "loss": 1.018, + "step": 2558 + }, + { + "epoch": 0.2307796365604004, + "grad_norm": 4.737173531186629, + "learning_rate": 3.591893492442577e-06, + "loss": 0.9851, + "step": 2559 + }, + { + "epoch": 0.23086982008387066, + "grad_norm": 1.9167108392462802, + "learning_rate": 3.591539762903914e-06, + "loss": 1.0784, + "step": 2560 + }, + { + "epoch": 0.23096000360734095, + "grad_norm": 1.7204897311989784, + "learning_rate": 3.591185897566303e-06, + "loss": 0.947, + "step": 2561 + }, + { + "epoch": 0.2310501871308112, + "grad_norm": 1.701997856443604, + "learning_rate": 3.590831896459937e-06, + "loss": 1.0172, + "step": 2562 + }, + { + "epoch": 0.23114037065428147, + "grad_norm": 1.8310276039663673, + "learning_rate": 3.5904777596150222e-06, + "loss": 1.0206, + "step": 2563 + }, + { + "epoch": 0.23123055417775173, + "grad_norm": 1.8362743961383352, + "learning_rate": 3.590123487061775e-06, + "loss": 1.0146, + "step": 2564 + }, + { + "epoch": 0.231320737701222, + "grad_norm": 2.2613718033560772, + "learning_rate": 3.589769078830424e-06, + "loss": 1.1111, + "step": 2565 + }, + { + "epoch": 0.23141092122469226, + "grad_norm": 2.444148557508953, + "learning_rate": 3.58941453495121e-06, + "loss": 0.9967, + "step": 2566 + }, + { + "epoch": 0.23150110474816252, + "grad_norm": 2.075058014759568, + "learning_rate": 3.5890598554543834e-06, + "loss": 1.0276, + "step": 2567 + }, + { + "epoch": 0.23159128827163278, + "grad_norm": 1.69197668451338, + "learning_rate": 3.5887050403702073e-06, + "loss": 1.0687, + "step": 2568 + }, + { + "epoch": 0.23168147179510304, + "grad_norm": 1.7724905973112417, + "learning_rate": 3.588350089728958e-06, + "loss": 0.9953, + "step": 2569 + }, + { + "epoch": 0.2317716553185733, + "grad_norm": 1.603035226431274, + "learning_rate": 3.5879950035609204e-06, + "loss": 1.0456, + "step": 2570 + }, + { + "epoch": 0.23186183884204356, + "grad_norm": 1.9301839954250652, + "learning_rate": 3.5876397818963933e-06, + "loss": 1.0161, + "step": 2571 + }, + { + "epoch": 0.23195202236551382, + "grad_norm": 1.6559167202727194, + "learning_rate": 3.5872844247656858e-06, + "loss": 1.0719, + "step": 2572 + }, + { + "epoch": 0.23204220588898408, + "grad_norm": 1.3573007653088296, + "learning_rate": 3.5869289321991195e-06, + "loss": 1.0227, + "step": 2573 + }, + { + "epoch": 0.23213238941245434, + "grad_norm": 1.4402655659943375, + "learning_rate": 3.5865733042270263e-06, + "loss": 1.0048, + "step": 2574 + }, + { + "epoch": 0.2322225729359246, + "grad_norm": 1.475246993899131, + "learning_rate": 3.5862175408797498e-06, + "loss": 1.0476, + "step": 2575 + }, + { + "epoch": 0.23231275645939486, + "grad_norm": 1.361470941525573, + "learning_rate": 3.585861642187647e-06, + "loss": 0.9565, + "step": 2576 + }, + { + "epoch": 0.23240293998286513, + "grad_norm": 1.4659873465685413, + "learning_rate": 3.5855056081810845e-06, + "loss": 1.0146, + "step": 2577 + }, + { + "epoch": 0.2324931235063354, + "grad_norm": 1.5213021204725083, + "learning_rate": 3.5851494388904406e-06, + "loss": 1.0163, + "step": 2578 + }, + { + "epoch": 0.23258330702980565, + "grad_norm": 1.5624841307787887, + "learning_rate": 3.5847931343461064e-06, + "loss": 0.951, + "step": 2579 + }, + { + "epoch": 0.2326734905532759, + "grad_norm": 1.983083950787657, + "learning_rate": 3.5844366945784835e-06, + "loss": 0.943, + "step": 2580 + }, + { + "epoch": 0.23276367407674617, + "grad_norm": 1.6922532002805795, + "learning_rate": 3.5840801196179856e-06, + "loss": 1.0712, + "step": 2581 + }, + { + "epoch": 0.23285385760021643, + "grad_norm": 1.8189217161935538, + "learning_rate": 3.583723409495037e-06, + "loss": 1.0753, + "step": 2582 + }, + { + "epoch": 0.2329440411236867, + "grad_norm": 2.352473491231521, + "learning_rate": 3.5833665642400747e-06, + "loss": 1.0698, + "step": 2583 + }, + { + "epoch": 0.23303422464715695, + "grad_norm": 1.7250271007233424, + "learning_rate": 3.5830095838835472e-06, + "loss": 0.9415, + "step": 2584 + }, + { + "epoch": 0.23312440817062724, + "grad_norm": 4.176712488976968, + "learning_rate": 3.5826524684559125e-06, + "loss": 0.9873, + "step": 2585 + }, + { + "epoch": 0.2332145916940975, + "grad_norm": 1.8984513144912756, + "learning_rate": 3.5822952179876433e-06, + "loss": 1.0944, + "step": 2586 + }, + { + "epoch": 0.23330477521756776, + "grad_norm": 1.3626089105036738, + "learning_rate": 3.5819378325092205e-06, + "loss": 0.9886, + "step": 2587 + }, + { + "epoch": 0.23339495874103802, + "grad_norm": 1.4805260800657591, + "learning_rate": 3.581580312051139e-06, + "loss": 1.0408, + "step": 2588 + }, + { + "epoch": 0.23348514226450828, + "grad_norm": 1.5316152331682864, + "learning_rate": 3.5812226566439057e-06, + "loss": 0.9429, + "step": 2589 + }, + { + "epoch": 0.23357532578797854, + "grad_norm": 2.385262052586524, + "learning_rate": 3.580864866318036e-06, + "loss": 0.9953, + "step": 2590 + }, + { + "epoch": 0.2336655093114488, + "grad_norm": 1.6056312051892978, + "learning_rate": 3.580506941104059e-06, + "loss": 1.0256, + "step": 2591 + }, + { + "epoch": 0.23375569283491907, + "grad_norm": 1.9259122916123466, + "learning_rate": 3.580148881032515e-06, + "loss": 1.0043, + "step": 2592 + }, + { + "epoch": 0.23384587635838933, + "grad_norm": 1.5837464630159717, + "learning_rate": 3.5797906861339556e-06, + "loss": 1.0227, + "step": 2593 + }, + { + "epoch": 0.2339360598818596, + "grad_norm": 1.7308894694369468, + "learning_rate": 3.5794323564389435e-06, + "loss": 1.048, + "step": 2594 + }, + { + "epoch": 0.23402624340532985, + "grad_norm": 1.8183813370342288, + "learning_rate": 3.579073891978055e-06, + "loss": 1.0237, + "step": 2595 + }, + { + "epoch": 0.2341164269288001, + "grad_norm": 2.5351228638318304, + "learning_rate": 3.5787152927818746e-06, + "loss": 0.9832, + "step": 2596 + }, + { + "epoch": 0.23420661045227037, + "grad_norm": 1.459707466500171, + "learning_rate": 3.5783565588810003e-06, + "loss": 1.0801, + "step": 2597 + }, + { + "epoch": 0.23429679397574063, + "grad_norm": 1.542972535720237, + "learning_rate": 3.5779976903060412e-06, + "loss": 0.9917, + "step": 2598 + }, + { + "epoch": 0.2343869774992109, + "grad_norm": 1.6981281156655246, + "learning_rate": 3.577638687087619e-06, + "loss": 1.0498, + "step": 2599 + }, + { + "epoch": 0.23447716102268115, + "grad_norm": 1.8994741239098136, + "learning_rate": 3.577279549256364e-06, + "loss": 1.0274, + "step": 2600 + }, + { + "epoch": 0.2345673445461514, + "grad_norm": 1.5082024578971345, + "learning_rate": 3.5769202768429213e-06, + "loss": 1.0651, + "step": 2601 + }, + { + "epoch": 0.23465752806962167, + "grad_norm": 1.6310940693543354, + "learning_rate": 3.5765608698779454e-06, + "loss": 0.9987, + "step": 2602 + }, + { + "epoch": 0.23474771159309193, + "grad_norm": 1.8598248875131054, + "learning_rate": 3.5762013283921033e-06, + "loss": 1.0285, + "step": 2603 + }, + { + "epoch": 0.2348378951165622, + "grad_norm": 2.032298360396767, + "learning_rate": 3.5758416524160728e-06, + "loss": 1.0237, + "step": 2604 + }, + { + "epoch": 0.23492807864003246, + "grad_norm": 1.5795052412863264, + "learning_rate": 3.5754818419805427e-06, + "loss": 1.0087, + "step": 2605 + }, + { + "epoch": 0.23501826216350272, + "grad_norm": 1.3029207156396665, + "learning_rate": 3.575121897116216e-06, + "loss": 0.9951, + "step": 2606 + }, + { + "epoch": 0.23510844568697298, + "grad_norm": 1.4317327992741717, + "learning_rate": 3.574761817853803e-06, + "loss": 1.0565, + "step": 2607 + }, + { + "epoch": 0.23519862921044324, + "grad_norm": 1.4727805006799157, + "learning_rate": 3.5744016042240287e-06, + "loss": 0.9531, + "step": 2608 + }, + { + "epoch": 0.23528881273391353, + "grad_norm": 2.261372119648907, + "learning_rate": 3.5740412562576286e-06, + "loss": 0.9518, + "step": 2609 + }, + { + "epoch": 0.2353789962573838, + "grad_norm": 1.8282160940053025, + "learning_rate": 3.573680773985349e-06, + "loss": 1.0134, + "step": 2610 + }, + { + "epoch": 0.23546917978085405, + "grad_norm": 1.9551531218055047, + "learning_rate": 3.5733201574379486e-06, + "loss": 1.0123, + "step": 2611 + }, + { + "epoch": 0.2355593633043243, + "grad_norm": 1.8319357906108555, + "learning_rate": 3.5729594066461975e-06, + "loss": 0.9608, + "step": 2612 + }, + { + "epoch": 0.23564954682779457, + "grad_norm": 1.374149319454619, + "learning_rate": 3.572598521640876e-06, + "loss": 1.0319, + "step": 2613 + }, + { + "epoch": 0.23573973035126483, + "grad_norm": 0.8139014262281464, + "learning_rate": 3.5722375024527782e-06, + "loss": 0.8733, + "step": 2614 + }, + { + "epoch": 0.2358299138747351, + "grad_norm": 1.4030566090523813, + "learning_rate": 3.571876349112707e-06, + "loss": 1.0395, + "step": 2615 + }, + { + "epoch": 0.23592009739820535, + "grad_norm": 1.5213776576722837, + "learning_rate": 3.5715150616514784e-06, + "loss": 1.0155, + "step": 2616 + }, + { + "epoch": 0.2360102809216756, + "grad_norm": 1.6481542841161712, + "learning_rate": 3.5711536400999196e-06, + "loss": 1.0669, + "step": 2617 + }, + { + "epoch": 0.23610046444514587, + "grad_norm": 0.7186277949062692, + "learning_rate": 3.570792084488869e-06, + "loss": 0.8752, + "step": 2618 + }, + { + "epoch": 0.23619064796861614, + "grad_norm": 0.7148936520027708, + "learning_rate": 3.5704303948491764e-06, + "loss": 0.887, + "step": 2619 + }, + { + "epoch": 0.2362808314920864, + "grad_norm": 1.6610990831470538, + "learning_rate": 3.5700685712117035e-06, + "loss": 1.0395, + "step": 2620 + }, + { + "epoch": 0.23637101501555666, + "grad_norm": 1.848508029601467, + "learning_rate": 3.5697066136073227e-06, + "loss": 1.0041, + "step": 2621 + }, + { + "epoch": 0.23646119853902692, + "grad_norm": 1.6075939953080847, + "learning_rate": 3.5693445220669184e-06, + "loss": 1.0267, + "step": 2622 + }, + { + "epoch": 0.23655138206249718, + "grad_norm": 1.457519204093347, + "learning_rate": 3.568982296621386e-06, + "loss": 1.0103, + "step": 2623 + }, + { + "epoch": 0.23664156558596744, + "grad_norm": 1.9008106108693614, + "learning_rate": 3.5686199373016325e-06, + "loss": 1.0919, + "step": 2624 + }, + { + "epoch": 0.2367317491094377, + "grad_norm": 1.6856684458575302, + "learning_rate": 3.568257444138577e-06, + "loss": 0.9658, + "step": 2625 + }, + { + "epoch": 0.23682193263290796, + "grad_norm": 1.6434368923176703, + "learning_rate": 3.5678948171631495e-06, + "loss": 1.0171, + "step": 2626 + }, + { + "epoch": 0.23691211615637822, + "grad_norm": 1.5073279358200442, + "learning_rate": 3.5675320564062908e-06, + "loss": 1.021, + "step": 2627 + }, + { + "epoch": 0.23700229967984848, + "grad_norm": 1.9483847506188534, + "learning_rate": 3.5671691618989533e-06, + "loss": 0.9268, + "step": 2628 + }, + { + "epoch": 0.23709248320331874, + "grad_norm": 2.6190013464435906, + "learning_rate": 3.5668061336721024e-06, + "loss": 1.0224, + "step": 2629 + }, + { + "epoch": 0.237182666726789, + "grad_norm": 3.9146330944807963, + "learning_rate": 3.5664429717567117e-06, + "loss": 1.009, + "step": 2630 + }, + { + "epoch": 0.23727285025025927, + "grad_norm": 1.6820581075222398, + "learning_rate": 3.56607967618377e-06, + "loss": 1.0479, + "step": 2631 + }, + { + "epoch": 0.23736303377372953, + "grad_norm": 1.6040227090555081, + "learning_rate": 3.5657162469842754e-06, + "loss": 1.034, + "step": 2632 + }, + { + "epoch": 0.23745321729719981, + "grad_norm": 1.5565350160467992, + "learning_rate": 3.5653526841892374e-06, + "loss": 1.0223, + "step": 2633 + }, + { + "epoch": 0.23754340082067008, + "grad_norm": 2.846870470933211, + "learning_rate": 3.564988987829676e-06, + "loss": 1.0531, + "step": 2634 + }, + { + "epoch": 0.23763358434414034, + "grad_norm": 2.025072769929699, + "learning_rate": 3.564625157936626e-06, + "loss": 0.9611, + "step": 2635 + }, + { + "epoch": 0.2377237678676106, + "grad_norm": 1.5073462837761824, + "learning_rate": 3.56426119454113e-06, + "loss": 1.0559, + "step": 2636 + }, + { + "epoch": 0.23781395139108086, + "grad_norm": 1.8483813679573118, + "learning_rate": 3.5638970976742436e-06, + "loss": 1.0947, + "step": 2637 + }, + { + "epoch": 0.23790413491455112, + "grad_norm": 2.016927017661763, + "learning_rate": 3.5635328673670335e-06, + "loss": 1.0439, + "step": 2638 + }, + { + "epoch": 0.23799431843802138, + "grad_norm": 1.9409588733897243, + "learning_rate": 3.5631685036505783e-06, + "loss": 1.1077, + "step": 2639 + }, + { + "epoch": 0.23808450196149164, + "grad_norm": 1.83117688394341, + "learning_rate": 3.562804006555966e-06, + "loss": 1.033, + "step": 2640 + }, + { + "epoch": 0.2381746854849619, + "grad_norm": 1.6526064832951461, + "learning_rate": 3.5624393761143e-06, + "loss": 0.9586, + "step": 2641 + }, + { + "epoch": 0.23826486900843216, + "grad_norm": 1.811934218397883, + "learning_rate": 3.5620746123566906e-06, + "loss": 0.9319, + "step": 2642 + }, + { + "epoch": 0.23835505253190242, + "grad_norm": 2.0875379730242893, + "learning_rate": 3.5617097153142623e-06, + "loss": 0.9724, + "step": 2643 + }, + { + "epoch": 0.23844523605537268, + "grad_norm": 1.5801229388055544, + "learning_rate": 3.5613446850181497e-06, + "loss": 1.0577, + "step": 2644 + }, + { + "epoch": 0.23853541957884294, + "grad_norm": 0.6087612582659825, + "learning_rate": 3.5609795214994996e-06, + "loss": 0.7756, + "step": 2645 + }, + { + "epoch": 0.2386256031023132, + "grad_norm": 2.73563220958531, + "learning_rate": 3.560614224789469e-06, + "loss": 1.075, + "step": 2646 + }, + { + "epoch": 0.23871578662578347, + "grad_norm": 2.121873350522014, + "learning_rate": 3.5602487949192285e-06, + "loss": 0.9665, + "step": 2647 + }, + { + "epoch": 0.23880597014925373, + "grad_norm": 2.033774582351458, + "learning_rate": 3.559883231919957e-06, + "loss": 0.96, + "step": 2648 + }, + { + "epoch": 0.238896153672724, + "grad_norm": 1.4125826347282884, + "learning_rate": 3.5595175358228473e-06, + "loss": 0.9945, + "step": 2649 + }, + { + "epoch": 0.23898633719619425, + "grad_norm": 1.7049325572795617, + "learning_rate": 3.5591517066591027e-06, + "loss": 1.0185, + "step": 2650 + }, + { + "epoch": 0.2390765207196645, + "grad_norm": 2.4932684870904422, + "learning_rate": 3.5587857444599364e-06, + "loss": 1.0071, + "step": 2651 + }, + { + "epoch": 0.23916670424313477, + "grad_norm": 1.417930329277822, + "learning_rate": 3.5584196492565766e-06, + "loss": 0.9836, + "step": 2652 + }, + { + "epoch": 0.23925688776660503, + "grad_norm": 2.294193285274847, + "learning_rate": 3.5580534210802587e-06, + "loss": 1.0822, + "step": 2653 + }, + { + "epoch": 0.2393470712900753, + "grad_norm": 1.409940116571122, + "learning_rate": 3.557687059962232e-06, + "loss": 1.0602, + "step": 2654 + }, + { + "epoch": 0.23943725481354555, + "grad_norm": 1.4185946845858164, + "learning_rate": 3.5573205659337558e-06, + "loss": 1.0115, + "step": 2655 + }, + { + "epoch": 0.23952743833701584, + "grad_norm": 1.482333415612528, + "learning_rate": 3.5569539390261025e-06, + "loss": 1.0064, + "step": 2656 + }, + { + "epoch": 0.2396176218604861, + "grad_norm": 1.3048939798485621, + "learning_rate": 3.5565871792705543e-06, + "loss": 1.0302, + "step": 2657 + }, + { + "epoch": 0.23970780538395636, + "grad_norm": 1.4254295437894717, + "learning_rate": 3.5562202866984045e-06, + "loss": 1.0116, + "step": 2658 + }, + { + "epoch": 0.23979798890742662, + "grad_norm": 2.2330463433793057, + "learning_rate": 3.5558532613409594e-06, + "loss": 1.0323, + "step": 2659 + }, + { + "epoch": 0.23988817243089688, + "grad_norm": 1.5123575127692008, + "learning_rate": 3.555486103229535e-06, + "loss": 0.9578, + "step": 2660 + }, + { + "epoch": 0.23997835595436715, + "grad_norm": 2.13716101384446, + "learning_rate": 3.5551188123954595e-06, + "loss": 1.0398, + "step": 2661 + }, + { + "epoch": 0.2400685394778374, + "grad_norm": 1.6210281542639855, + "learning_rate": 3.5547513888700715e-06, + "loss": 0.9113, + "step": 2662 + }, + { + "epoch": 0.24015872300130767, + "grad_norm": 1.5303529719928435, + "learning_rate": 3.5543838326847224e-06, + "loss": 1.0522, + "step": 2663 + }, + { + "epoch": 0.24024890652477793, + "grad_norm": 2.6546120474040835, + "learning_rate": 3.5540161438707744e-06, + "loss": 0.9964, + "step": 2664 + }, + { + "epoch": 0.2403390900482482, + "grad_norm": 1.51766005286313, + "learning_rate": 3.5536483224596e-06, + "loss": 1.1743, + "step": 2665 + }, + { + "epoch": 0.24042927357171845, + "grad_norm": 1.7652055360245222, + "learning_rate": 3.553280368482584e-06, + "loss": 0.9954, + "step": 2666 + }, + { + "epoch": 0.2405194570951887, + "grad_norm": 1.871925503617858, + "learning_rate": 3.5529122819711227e-06, + "loss": 0.8413, + "step": 2667 + }, + { + "epoch": 0.24060964061865897, + "grad_norm": 1.4964355397054054, + "learning_rate": 3.5525440629566223e-06, + "loss": 0.9608, + "step": 2668 + }, + { + "epoch": 0.24069982414212923, + "grad_norm": 1.727742202054535, + "learning_rate": 3.552175711470502e-06, + "loss": 1.0367, + "step": 2669 + }, + { + "epoch": 0.2407900076655995, + "grad_norm": 1.7257955237126872, + "learning_rate": 3.5518072275441912e-06, + "loss": 1.065, + "step": 2670 + }, + { + "epoch": 0.24088019118906975, + "grad_norm": 6.994875803479538, + "learning_rate": 3.551438611209131e-06, + "loss": 1.0456, + "step": 2671 + }, + { + "epoch": 0.24097037471254001, + "grad_norm": 1.7431014875191937, + "learning_rate": 3.551069862496774e-06, + "loss": 1.1162, + "step": 2672 + }, + { + "epoch": 0.24106055823601027, + "grad_norm": 0.6587851693032407, + "learning_rate": 3.5507009814385846e-06, + "loss": 0.837, + "step": 2673 + }, + { + "epoch": 0.24115074175948054, + "grad_norm": 0.6546318449340311, + "learning_rate": 3.550331968066036e-06, + "loss": 0.8277, + "step": 2674 + }, + { + "epoch": 0.2412409252829508, + "grad_norm": 1.6711821501945945, + "learning_rate": 3.549962822410616e-06, + "loss": 1.0354, + "step": 2675 + }, + { + "epoch": 0.24133110880642106, + "grad_norm": 0.6969569401828494, + "learning_rate": 3.5495935445038217e-06, + "loss": 0.8247, + "step": 2676 + }, + { + "epoch": 0.24142129232989132, + "grad_norm": 1.5706037564402622, + "learning_rate": 3.5492241343771612e-06, + "loss": 1.057, + "step": 2677 + }, + { + "epoch": 0.24151147585336158, + "grad_norm": 1.5690137512256017, + "learning_rate": 3.548854592062156e-06, + "loss": 1.0157, + "step": 2678 + }, + { + "epoch": 0.24160165937683184, + "grad_norm": 1.580019201210301, + "learning_rate": 3.548484917590336e-06, + "loss": 0.9988, + "step": 2679 + }, + { + "epoch": 0.24169184290030213, + "grad_norm": 2.1773927702927294, + "learning_rate": 3.5481151109932447e-06, + "loss": 0.935, + "step": 2680 + }, + { + "epoch": 0.2417820264237724, + "grad_norm": 1.6153742258031227, + "learning_rate": 3.5477451723024364e-06, + "loss": 1.0434, + "step": 2681 + }, + { + "epoch": 0.24187220994724265, + "grad_norm": 0.5813964095355494, + "learning_rate": 3.5473751015494757e-06, + "loss": 0.8087, + "step": 2682 + }, + { + "epoch": 0.2419623934707129, + "grad_norm": 1.4689249786685987, + "learning_rate": 3.547004898765939e-06, + "loss": 1.0308, + "step": 2683 + }, + { + "epoch": 0.24205257699418317, + "grad_norm": 1.4120276969619292, + "learning_rate": 3.546634563983414e-06, + "loss": 1.0358, + "step": 2684 + }, + { + "epoch": 0.24214276051765343, + "grad_norm": 1.4202581581992044, + "learning_rate": 3.5462640972335002e-06, + "loss": 0.9837, + "step": 2685 + }, + { + "epoch": 0.2422329440411237, + "grad_norm": 1.7445503257736947, + "learning_rate": 3.5458934985478077e-06, + "loss": 1.0363, + "step": 2686 + }, + { + "epoch": 0.24232312756459395, + "grad_norm": 1.8644718868543224, + "learning_rate": 3.5455227679579577e-06, + "loss": 0.9964, + "step": 2687 + }, + { + "epoch": 0.24241331108806422, + "grad_norm": 1.8454506353216436, + "learning_rate": 3.545151905495584e-06, + "loss": 1.1266, + "step": 2688 + }, + { + "epoch": 0.24250349461153448, + "grad_norm": 1.8818188969900174, + "learning_rate": 3.544780911192329e-06, + "loss": 0.9985, + "step": 2689 + }, + { + "epoch": 0.24259367813500474, + "grad_norm": 1.6958675618509416, + "learning_rate": 3.544409785079849e-06, + "loss": 0.9662, + "step": 2690 + }, + { + "epoch": 0.242683861658475, + "grad_norm": 1.5384090217466642, + "learning_rate": 3.5440385271898103e-06, + "loss": 0.9619, + "step": 2691 + }, + { + "epoch": 0.24277404518194526, + "grad_norm": 1.6759513288500338, + "learning_rate": 3.5436671375538903e-06, + "loss": 0.9484, + "step": 2692 + }, + { + "epoch": 0.24286422870541552, + "grad_norm": 4.690906558395681, + "learning_rate": 3.543295616203779e-06, + "loss": 1.0219, + "step": 2693 + }, + { + "epoch": 0.24295441222888578, + "grad_norm": 1.4527230475923418, + "learning_rate": 3.542923963171176e-06, + "loss": 1.1134, + "step": 2694 + }, + { + "epoch": 0.24304459575235604, + "grad_norm": 1.412228570025611, + "learning_rate": 3.542552178487793e-06, + "loss": 0.9362, + "step": 2695 + }, + { + "epoch": 0.2431347792758263, + "grad_norm": 3.2102425744450818, + "learning_rate": 3.5421802621853523e-06, + "loss": 1.0097, + "step": 2696 + }, + { + "epoch": 0.24322496279929656, + "grad_norm": 2.8610118079210154, + "learning_rate": 3.5418082142955887e-06, + "loss": 1.0144, + "step": 2697 + }, + { + "epoch": 0.24331514632276682, + "grad_norm": 0.667730813285279, + "learning_rate": 3.5414360348502463e-06, + "loss": 0.8531, + "step": 2698 + }, + { + "epoch": 0.24340532984623708, + "grad_norm": 2.8738716233158024, + "learning_rate": 3.5410637238810825e-06, + "loss": 0.9193, + "step": 2699 + }, + { + "epoch": 0.24349551336970734, + "grad_norm": 1.7930326429630552, + "learning_rate": 3.5406912814198635e-06, + "loss": 0.9718, + "step": 2700 + }, + { + "epoch": 0.2435856968931776, + "grad_norm": 1.7543370454852703, + "learning_rate": 3.54031870749837e-06, + "loss": 0.9861, + "step": 2701 + }, + { + "epoch": 0.24367588041664787, + "grad_norm": 1.5657130203326486, + "learning_rate": 3.539946002148391e-06, + "loss": 1.0308, + "step": 2702 + }, + { + "epoch": 0.24376606394011813, + "grad_norm": 2.247682331399284, + "learning_rate": 3.5395731654017277e-06, + "loss": 1.018, + "step": 2703 + }, + { + "epoch": 0.24385624746358842, + "grad_norm": 4.707194345268312, + "learning_rate": 3.5392001972901923e-06, + "loss": 1.0589, + "step": 2704 + }, + { + "epoch": 0.24394643098705868, + "grad_norm": 1.6057630580113995, + "learning_rate": 3.5388270978456098e-06, + "loss": 0.9782, + "step": 2705 + }, + { + "epoch": 0.24403661451052894, + "grad_norm": 1.7309634360556754, + "learning_rate": 3.5384538670998137e-06, + "loss": 0.9856, + "step": 2706 + }, + { + "epoch": 0.2441267980339992, + "grad_norm": 1.6828965870030121, + "learning_rate": 3.538080505084651e-06, + "loss": 1.0387, + "step": 2707 + }, + { + "epoch": 0.24421698155746946, + "grad_norm": 1.5585149682959698, + "learning_rate": 3.5377070118319788e-06, + "loss": 1.0321, + "step": 2708 + }, + { + "epoch": 0.24430716508093972, + "grad_norm": 2.305624457814501, + "learning_rate": 3.5373333873736657e-06, + "loss": 0.9751, + "step": 2709 + }, + { + "epoch": 0.24439734860440998, + "grad_norm": 1.4971750997859086, + "learning_rate": 3.536959631741591e-06, + "loss": 1.0263, + "step": 2710 + }, + { + "epoch": 0.24448753212788024, + "grad_norm": 1.2207825169495077, + "learning_rate": 3.536585744967646e-06, + "loss": 1.0437, + "step": 2711 + }, + { + "epoch": 0.2445777156513505, + "grad_norm": 2.370635992417419, + "learning_rate": 3.5362117270837326e-06, + "loss": 1.0038, + "step": 2712 + }, + { + "epoch": 0.24466789917482076, + "grad_norm": 1.339636783907889, + "learning_rate": 3.5358375781217634e-06, + "loss": 0.9194, + "step": 2713 + }, + { + "epoch": 0.24475808269829102, + "grad_norm": 1.7891571819454972, + "learning_rate": 3.535463298113664e-06, + "loss": 1.0484, + "step": 2714 + }, + { + "epoch": 0.24484826622176128, + "grad_norm": 2.1788363418498027, + "learning_rate": 3.5350888870913697e-06, + "loss": 0.98, + "step": 2715 + }, + { + "epoch": 0.24493844974523155, + "grad_norm": 1.8725328744566865, + "learning_rate": 3.5347143450868273e-06, + "loss": 1.0226, + "step": 2716 + }, + { + "epoch": 0.2450286332687018, + "grad_norm": 2.0319237398645758, + "learning_rate": 3.534339672131994e-06, + "loss": 1.0387, + "step": 2717 + }, + { + "epoch": 0.24511881679217207, + "grad_norm": 1.3906638750793172, + "learning_rate": 3.5339648682588397e-06, + "loss": 0.9513, + "step": 2718 + }, + { + "epoch": 0.24520900031564233, + "grad_norm": 1.6117124896683737, + "learning_rate": 3.533589933499345e-06, + "loss": 1.0614, + "step": 2719 + }, + { + "epoch": 0.2452991838391126, + "grad_norm": 1.8251588020734675, + "learning_rate": 3.533214867885501e-06, + "loss": 0.9548, + "step": 2720 + }, + { + "epoch": 0.24538936736258285, + "grad_norm": 2.274051229478618, + "learning_rate": 3.53283967144931e-06, + "loss": 0.9596, + "step": 2721 + }, + { + "epoch": 0.2454795508860531, + "grad_norm": 1.7824941606680096, + "learning_rate": 3.532464344222787e-06, + "loss": 1.028, + "step": 2722 + }, + { + "epoch": 0.24556973440952337, + "grad_norm": 1.6020617195460551, + "learning_rate": 3.532088886237956e-06, + "loss": 1.0582, + "step": 2723 + }, + { + "epoch": 0.24565991793299363, + "grad_norm": 0.6412695689124808, + "learning_rate": 3.5317132975268535e-06, + "loss": 0.7648, + "step": 2724 + }, + { + "epoch": 0.2457501014564639, + "grad_norm": 1.5051049149656348, + "learning_rate": 3.531337578121526e-06, + "loss": 1.0436, + "step": 2725 + }, + { + "epoch": 0.24584028497993415, + "grad_norm": 2.8501302388197742, + "learning_rate": 3.530961728054033e-06, + "loss": 1.0504, + "step": 2726 + }, + { + "epoch": 0.24593046850340441, + "grad_norm": 1.427177268322143, + "learning_rate": 3.5305857473564435e-06, + "loss": 0.9787, + "step": 2727 + }, + { + "epoch": 0.2460206520268747, + "grad_norm": 4.39779848162525, + "learning_rate": 3.5302096360608385e-06, + "loss": 0.951, + "step": 2728 + }, + { + "epoch": 0.24611083555034496, + "grad_norm": 1.5646189245788862, + "learning_rate": 3.5298333941993105e-06, + "loss": 1.0696, + "step": 2729 + }, + { + "epoch": 0.24620101907381522, + "grad_norm": 1.985087529613967, + "learning_rate": 3.529457021803962e-06, + "loss": 1.0103, + "step": 2730 + }, + { + "epoch": 0.24629120259728549, + "grad_norm": 1.935661551179997, + "learning_rate": 3.529080518906906e-06, + "loss": 1.0323, + "step": 2731 + }, + { + "epoch": 0.24638138612075575, + "grad_norm": 1.3325448039039123, + "learning_rate": 3.5287038855402696e-06, + "loss": 1.0006, + "step": 2732 + }, + { + "epoch": 0.246471569644226, + "grad_norm": 4.190602306330121, + "learning_rate": 3.528327121736188e-06, + "loss": 1.0536, + "step": 2733 + }, + { + "epoch": 0.24656175316769627, + "grad_norm": 1.5761832954153714, + "learning_rate": 3.52795022752681e-06, + "loss": 1.1002, + "step": 2734 + }, + { + "epoch": 0.24665193669116653, + "grad_norm": 1.9777189347208612, + "learning_rate": 3.5275732029442925e-06, + "loss": 1.087, + "step": 2735 + }, + { + "epoch": 0.2467421202146368, + "grad_norm": 2.140078064038371, + "learning_rate": 3.5271960480208077e-06, + "loss": 0.9556, + "step": 2736 + }, + { + "epoch": 0.24683230373810705, + "grad_norm": 0.6801167809610772, + "learning_rate": 3.526818762788534e-06, + "loss": 0.8656, + "step": 2737 + }, + { + "epoch": 0.2469224872615773, + "grad_norm": 1.6894966604984962, + "learning_rate": 3.5264413472796653e-06, + "loss": 1.0563, + "step": 2738 + }, + { + "epoch": 0.24701267078504757, + "grad_norm": 2.265947305839343, + "learning_rate": 3.5260638015264037e-06, + "loss": 0.9922, + "step": 2739 + }, + { + "epoch": 0.24710285430851783, + "grad_norm": 1.837540401611437, + "learning_rate": 3.5256861255609644e-06, + "loss": 0.9882, + "step": 2740 + }, + { + "epoch": 0.2471930378319881, + "grad_norm": 2.00779623180879, + "learning_rate": 3.5253083194155723e-06, + "loss": 1.0287, + "step": 2741 + }, + { + "epoch": 0.24728322135545835, + "grad_norm": 4.41316319309591, + "learning_rate": 3.5249303831224637e-06, + "loss": 1.0386, + "step": 2742 + }, + { + "epoch": 0.24737340487892862, + "grad_norm": 1.524578430190138, + "learning_rate": 3.524552316713887e-06, + "loss": 1.0074, + "step": 2743 + }, + { + "epoch": 0.24746358840239888, + "grad_norm": 1.8088894760884346, + "learning_rate": 3.5241741202220995e-06, + "loss": 1.0775, + "step": 2744 + }, + { + "epoch": 0.24755377192586914, + "grad_norm": 1.4344091772857108, + "learning_rate": 3.5237957936793724e-06, + "loss": 1.0015, + "step": 2745 + }, + { + "epoch": 0.2476439554493394, + "grad_norm": 1.6614405791510953, + "learning_rate": 3.523417337117986e-06, + "loss": 0.982, + "step": 2746 + }, + { + "epoch": 0.24773413897280966, + "grad_norm": 1.6798055208128397, + "learning_rate": 3.523038750570232e-06, + "loss": 0.9668, + "step": 2747 + }, + { + "epoch": 0.24782432249627992, + "grad_norm": 1.7713570512230108, + "learning_rate": 3.522660034068414e-06, + "loss": 1.0426, + "step": 2748 + }, + { + "epoch": 0.24791450601975018, + "grad_norm": 1.765128479406773, + "learning_rate": 3.5222811876448464e-06, + "loss": 1.0304, + "step": 2749 + }, + { + "epoch": 0.24800468954322044, + "grad_norm": 1.4131890650133143, + "learning_rate": 3.521902211331854e-06, + "loss": 1.0349, + "step": 2750 + }, + { + "epoch": 0.2480948730666907, + "grad_norm": 1.4307088070274752, + "learning_rate": 3.5215231051617726e-06, + "loss": 0.9632, + "step": 2751 + }, + { + "epoch": 0.248185056590161, + "grad_norm": 1.8476739926927899, + "learning_rate": 3.521143869166951e-06, + "loss": 0.922, + "step": 2752 + }, + { + "epoch": 0.24827524011363125, + "grad_norm": 2.0225103308716785, + "learning_rate": 3.5207645033797464e-06, + "loss": 1.0463, + "step": 2753 + }, + { + "epoch": 0.2483654236371015, + "grad_norm": 1.6708161753521054, + "learning_rate": 3.5203850078325293e-06, + "loss": 0.9412, + "step": 2754 + }, + { + "epoch": 0.24845560716057177, + "grad_norm": 2.7519040018737146, + "learning_rate": 3.5200053825576797e-06, + "loss": 0.9233, + "step": 2755 + }, + { + "epoch": 0.24854579068404203, + "grad_norm": 2.0462854133202764, + "learning_rate": 3.51962562758759e-06, + "loss": 1.0213, + "step": 2756 + }, + { + "epoch": 0.2486359742075123, + "grad_norm": 1.7389682740821237, + "learning_rate": 3.5192457429546627e-06, + "loss": 0.9902, + "step": 2757 + }, + { + "epoch": 0.24872615773098256, + "grad_norm": 1.6213597090990244, + "learning_rate": 3.5188657286913115e-06, + "loss": 0.9833, + "step": 2758 + }, + { + "epoch": 0.24881634125445282, + "grad_norm": 1.846079954401711, + "learning_rate": 3.518485584829961e-06, + "loss": 0.9808, + "step": 2759 + }, + { + "epoch": 0.24890652477792308, + "grad_norm": 1.697626459638888, + "learning_rate": 3.5181053114030485e-06, + "loss": 1.0255, + "step": 2760 + }, + { + "epoch": 0.24899670830139334, + "grad_norm": 1.9783940329662626, + "learning_rate": 3.5177249084430198e-06, + "loss": 1.0119, + "step": 2761 + }, + { + "epoch": 0.2490868918248636, + "grad_norm": 2.5919928976067617, + "learning_rate": 3.517344375982333e-06, + "loss": 1.0493, + "step": 2762 + }, + { + "epoch": 0.24917707534833386, + "grad_norm": 4.097468892806222, + "learning_rate": 3.5169637140534565e-06, + "loss": 1.0477, + "step": 2763 + }, + { + "epoch": 0.24926725887180412, + "grad_norm": 2.0426175634385233, + "learning_rate": 3.5165829226888733e-06, + "loss": 0.9854, + "step": 2764 + }, + { + "epoch": 0.24935744239527438, + "grad_norm": 1.5793310410330097, + "learning_rate": 3.516202001921072e-06, + "loss": 1.0416, + "step": 2765 + }, + { + "epoch": 0.24944762591874464, + "grad_norm": 1.5472873176762525, + "learning_rate": 3.515820951782555e-06, + "loss": 1.0419, + "step": 2766 + }, + { + "epoch": 0.2495378094422149, + "grad_norm": 2.0629013133480916, + "learning_rate": 3.5154397723058366e-06, + "loss": 0.946, + "step": 2767 + }, + { + "epoch": 0.24962799296568516, + "grad_norm": 1.6052752375868207, + "learning_rate": 3.5150584635234416e-06, + "loss": 1.0697, + "step": 2768 + }, + { + "epoch": 0.24971817648915542, + "grad_norm": 1.26905931357308, + "learning_rate": 3.5146770254679035e-06, + "loss": 1.0587, + "step": 2769 + }, + { + "epoch": 0.24980836001262569, + "grad_norm": 2.116294193786521, + "learning_rate": 3.51429545817177e-06, + "loss": 0.9401, + "step": 2770 + }, + { + "epoch": 0.24989854353609595, + "grad_norm": 1.9973923253339843, + "learning_rate": 3.5139137616675985e-06, + "loss": 1.0174, + "step": 2771 + }, + { + "epoch": 0.2499887270595662, + "grad_norm": 1.5814889991080958, + "learning_rate": 3.513531935987957e-06, + "loss": 0.9572, + "step": 2772 + }, + { + "epoch": 0.2500789105830365, + "grad_norm": 1.419248481597814, + "learning_rate": 3.5131499811654253e-06, + "loss": 1.0394, + "step": 2773 + }, + { + "epoch": 0.25016909410650673, + "grad_norm": 1.7292048744974784, + "learning_rate": 3.512767897232594e-06, + "loss": 0.9832, + "step": 2774 + }, + { + "epoch": 0.250259277629977, + "grad_norm": 1.7463321395047635, + "learning_rate": 3.512385684222064e-06, + "loss": 1.0115, + "step": 2775 + }, + { + "epoch": 0.25034946115344725, + "grad_norm": 1.9963994995776928, + "learning_rate": 3.512003342166449e-06, + "loss": 1.1138, + "step": 2776 + }, + { + "epoch": 0.25043964467691754, + "grad_norm": 1.3601498642562846, + "learning_rate": 3.511620871098371e-06, + "loss": 1.0528, + "step": 2777 + }, + { + "epoch": 0.25052982820038777, + "grad_norm": 1.7189532766644202, + "learning_rate": 3.511238271050465e-06, + "loss": 1.0446, + "step": 2778 + }, + { + "epoch": 0.25062001172385806, + "grad_norm": 1.527278889013015, + "learning_rate": 3.5108555420553778e-06, + "loss": 0.9317, + "step": 2779 + }, + { + "epoch": 0.2507101952473283, + "grad_norm": 1.7596532381435017, + "learning_rate": 3.510472684145764e-06, + "loss": 0.8304, + "step": 2780 + }, + { + "epoch": 0.2508003787707986, + "grad_norm": 1.5908131935892356, + "learning_rate": 3.5100896973542926e-06, + "loss": 0.904, + "step": 2781 + }, + { + "epoch": 0.2508905622942688, + "grad_norm": 1.820909394073011, + "learning_rate": 3.509706581713642e-06, + "loss": 1.0555, + "step": 2782 + }, + { + "epoch": 0.2509807458177391, + "grad_norm": 1.5274513772571106, + "learning_rate": 3.509323337256501e-06, + "loss": 1.0377, + "step": 2783 + }, + { + "epoch": 0.25107092934120934, + "grad_norm": 2.054998686110176, + "learning_rate": 3.5089399640155703e-06, + "loss": 0.959, + "step": 2784 + }, + { + "epoch": 0.2511611128646796, + "grad_norm": 1.329717870201836, + "learning_rate": 3.508556462023562e-06, + "loss": 1.0054, + "step": 2785 + }, + { + "epoch": 0.25125129638814986, + "grad_norm": 2.243960858906347, + "learning_rate": 3.5081728313131984e-06, + "loss": 0.9996, + "step": 2786 + }, + { + "epoch": 0.25134147991162015, + "grad_norm": 2.5211883063799116, + "learning_rate": 3.5077890719172125e-06, + "loss": 1.0062, + "step": 2787 + }, + { + "epoch": 0.25143166343509044, + "grad_norm": 2.1621024278978838, + "learning_rate": 3.5074051838683497e-06, + "loss": 1.0174, + "step": 2788 + }, + { + "epoch": 0.25152184695856067, + "grad_norm": 1.6511099607701243, + "learning_rate": 3.5070211671993643e-06, + "loss": 0.94, + "step": 2789 + }, + { + "epoch": 0.25161203048203096, + "grad_norm": 1.6555981342905153, + "learning_rate": 3.5066370219430238e-06, + "loss": 1.0365, + "step": 2790 + }, + { + "epoch": 0.2517022140055012, + "grad_norm": 1.576098056279435, + "learning_rate": 3.5062527481321044e-06, + "loss": 1.0129, + "step": 2791 + }, + { + "epoch": 0.2517923975289715, + "grad_norm": 2.1508324054103634, + "learning_rate": 3.5058683457993954e-06, + "loss": 1.0261, + "step": 2792 + }, + { + "epoch": 0.2518825810524417, + "grad_norm": 0.7634351124298699, + "learning_rate": 3.5054838149776963e-06, + "loss": 0.7737, + "step": 2793 + }, + { + "epoch": 0.251972764575912, + "grad_norm": 1.5997463442484496, + "learning_rate": 3.505099155699816e-06, + "loss": 1.045, + "step": 2794 + }, + { + "epoch": 0.25206294809938223, + "grad_norm": 1.7526639689142633, + "learning_rate": 3.5047143679985775e-06, + "loss": 0.894, + "step": 2795 + }, + { + "epoch": 0.2521531316228525, + "grad_norm": 1.3902986122032779, + "learning_rate": 3.5043294519068126e-06, + "loss": 1.0282, + "step": 2796 + }, + { + "epoch": 0.25224331514632276, + "grad_norm": 2.5311310351637646, + "learning_rate": 3.503944407457363e-06, + "loss": 0.9355, + "step": 2797 + }, + { + "epoch": 0.25233349866979304, + "grad_norm": 1.5284565727745711, + "learning_rate": 3.5035592346830846e-06, + "loss": 1.0254, + "step": 2798 + }, + { + "epoch": 0.2524236821932633, + "grad_norm": 1.7009443969698868, + "learning_rate": 3.503173933616841e-06, + "loss": 1.0842, + "step": 2799 + }, + { + "epoch": 0.25251386571673357, + "grad_norm": 1.8819761199287455, + "learning_rate": 3.50278850429151e-06, + "loss": 1.0844, + "step": 2800 + }, + { + "epoch": 0.2526040492402038, + "grad_norm": 1.824554673822279, + "learning_rate": 3.502402946739977e-06, + "loss": 1.0374, + "step": 2801 + }, + { + "epoch": 0.2526942327636741, + "grad_norm": 1.738215824567565, + "learning_rate": 3.5020172609951405e-06, + "loss": 1.0339, + "step": 2802 + }, + { + "epoch": 0.2527844162871443, + "grad_norm": 1.2849912682470377, + "learning_rate": 3.501631447089909e-06, + "loss": 1.0296, + "step": 2803 + }, + { + "epoch": 0.2528745998106146, + "grad_norm": 3.567313706600364, + "learning_rate": 3.501245505057203e-06, + "loss": 0.9852, + "step": 2804 + }, + { + "epoch": 0.25296478333408484, + "grad_norm": 1.6167662435847874, + "learning_rate": 3.5008594349299526e-06, + "loss": 1.089, + "step": 2805 + }, + { + "epoch": 0.25305496685755513, + "grad_norm": 1.8525639654485595, + "learning_rate": 3.500473236741099e-06, + "loss": 0.9984, + "step": 2806 + }, + { + "epoch": 0.25314515038102536, + "grad_norm": 2.045663839286768, + "learning_rate": 3.500086910523596e-06, + "loss": 0.9733, + "step": 2807 + }, + { + "epoch": 0.25323533390449565, + "grad_norm": 1.4102971648505236, + "learning_rate": 3.499700456310406e-06, + "loss": 0.9621, + "step": 2808 + }, + { + "epoch": 0.2533255174279659, + "grad_norm": 1.3496621945502487, + "learning_rate": 3.499313874134504e-06, + "loss": 1.0662, + "step": 2809 + }, + { + "epoch": 0.2534157009514362, + "grad_norm": 2.020036704230393, + "learning_rate": 3.498927164028875e-06, + "loss": 0.8837, + "step": 2810 + }, + { + "epoch": 0.25350588447490646, + "grad_norm": 1.281971100506749, + "learning_rate": 3.498540326026515e-06, + "loss": 1.0503, + "step": 2811 + }, + { + "epoch": 0.2535960679983767, + "grad_norm": 1.629475226617501, + "learning_rate": 3.4981533601604323e-06, + "loss": 1.0837, + "step": 2812 + }, + { + "epoch": 0.253686251521847, + "grad_norm": 1.6708351537959327, + "learning_rate": 3.4977662664636443e-06, + "loss": 1.0228, + "step": 2813 + }, + { + "epoch": 0.2537764350453172, + "grad_norm": 2.14360407118513, + "learning_rate": 3.497379044969179e-06, + "loss": 0.9733, + "step": 2814 + }, + { + "epoch": 0.2538666185687875, + "grad_norm": 1.9281495796009467, + "learning_rate": 3.4969916957100777e-06, + "loss": 1.0098, + "step": 2815 + }, + { + "epoch": 0.25395680209225774, + "grad_norm": 1.5129570829931853, + "learning_rate": 3.4966042187193905e-06, + "loss": 1.0076, + "step": 2816 + }, + { + "epoch": 0.254046985615728, + "grad_norm": 1.5693490788040718, + "learning_rate": 3.496216614030179e-06, + "loss": 1.009, + "step": 2817 + }, + { + "epoch": 0.25413716913919826, + "grad_norm": 1.6731635146326846, + "learning_rate": 3.495828881675516e-06, + "loss": 0.9108, + "step": 2818 + }, + { + "epoch": 0.25422735266266855, + "grad_norm": 1.9471802841050372, + "learning_rate": 3.4954410216884845e-06, + "loss": 1.0838, + "step": 2819 + }, + { + "epoch": 0.2543175361861388, + "grad_norm": 1.7364238287829625, + "learning_rate": 3.49505303410218e-06, + "loss": 1.0173, + "step": 2820 + }, + { + "epoch": 0.25440771970960907, + "grad_norm": 1.8337777928363501, + "learning_rate": 3.4946649189497067e-06, + "loss": 1.0423, + "step": 2821 + }, + { + "epoch": 0.2544979032330793, + "grad_norm": 4.284249250596346, + "learning_rate": 3.4942766762641805e-06, + "loss": 1.0015, + "step": 2822 + }, + { + "epoch": 0.2545880867565496, + "grad_norm": 1.8131998618291856, + "learning_rate": 3.49388830607873e-06, + "loss": 1.0367, + "step": 2823 + }, + { + "epoch": 0.2546782702800198, + "grad_norm": 1.3773044435601554, + "learning_rate": 3.493499808426491e-06, + "loss": 0.9792, + "step": 2824 + }, + { + "epoch": 0.2547684538034901, + "grad_norm": 1.9484640430150635, + "learning_rate": 3.493111183340614e-06, + "loss": 1.0628, + "step": 2825 + }, + { + "epoch": 0.25485863732696035, + "grad_norm": 1.8844966716296692, + "learning_rate": 3.4927224308542576e-06, + "loss": 0.9153, + "step": 2826 + }, + { + "epoch": 0.25494882085043064, + "grad_norm": 1.46760197676689, + "learning_rate": 3.4923335510005923e-06, + "loss": 1.0956, + "step": 2827 + }, + { + "epoch": 0.25503900437390087, + "grad_norm": 1.8034302927394403, + "learning_rate": 3.4919445438128e-06, + "loss": 1.0326, + "step": 2828 + }, + { + "epoch": 0.25512918789737116, + "grad_norm": 1.6218314457155427, + "learning_rate": 3.491555409324073e-06, + "loss": 1.0925, + "step": 2829 + }, + { + "epoch": 0.2552193714208414, + "grad_norm": 0.6645908946222006, + "learning_rate": 3.4911661475676136e-06, + "loss": 0.7976, + "step": 2830 + }, + { + "epoch": 0.2553095549443117, + "grad_norm": 1.5817054695672979, + "learning_rate": 3.490776758576637e-06, + "loss": 0.9904, + "step": 2831 + }, + { + "epoch": 0.2553997384677819, + "grad_norm": 1.632725690514392, + "learning_rate": 3.4903872423843668e-06, + "loss": 1.0628, + "step": 2832 + }, + { + "epoch": 0.2554899219912522, + "grad_norm": 1.7975572037381897, + "learning_rate": 3.4899975990240396e-06, + "loss": 0.9726, + "step": 2833 + }, + { + "epoch": 0.25558010551472243, + "grad_norm": 3.5112588176025836, + "learning_rate": 3.489607828528901e-06, + "loss": 1.0369, + "step": 2834 + }, + { + "epoch": 0.2556702890381927, + "grad_norm": 0.6616945796006624, + "learning_rate": 3.4892179309322093e-06, + "loss": 0.9072, + "step": 2835 + }, + { + "epoch": 0.255760472561663, + "grad_norm": 1.683584828364195, + "learning_rate": 3.488827906267232e-06, + "loss": 0.93, + "step": 2836 + }, + { + "epoch": 0.25585065608513324, + "grad_norm": 1.7448228823777905, + "learning_rate": 3.4884377545672485e-06, + "loss": 1.0504, + "step": 2837 + }, + { + "epoch": 0.25594083960860353, + "grad_norm": 1.7126635201899065, + "learning_rate": 3.4880474758655485e-06, + "loss": 0.9851, + "step": 2838 + }, + { + "epoch": 0.25603102313207377, + "grad_norm": 1.6752065104247835, + "learning_rate": 3.487657070195433e-06, + "loss": 1.082, + "step": 2839 + }, + { + "epoch": 0.25612120665554405, + "grad_norm": 1.5874709059084242, + "learning_rate": 3.487266537590213e-06, + "loss": 1.0874, + "step": 2840 + }, + { + "epoch": 0.2562113901790143, + "grad_norm": 3.624724147429497, + "learning_rate": 3.4868758780832116e-06, + "loss": 0.9318, + "step": 2841 + }, + { + "epoch": 0.2563015737024846, + "grad_norm": 1.6571669289644575, + "learning_rate": 3.486485091707762e-06, + "loss": 1.0238, + "step": 2842 + }, + { + "epoch": 0.2563917572259548, + "grad_norm": 1.724141628820466, + "learning_rate": 3.4860941784972077e-06, + "loss": 0.8905, + "step": 2843 + }, + { + "epoch": 0.2564819407494251, + "grad_norm": 2.704678304161866, + "learning_rate": 3.485703138484904e-06, + "loss": 1.1076, + "step": 2844 + }, + { + "epoch": 0.25657212427289533, + "grad_norm": 1.6071944046836435, + "learning_rate": 3.485311971704216e-06, + "loss": 1.0639, + "step": 2845 + }, + { + "epoch": 0.2566623077963656, + "grad_norm": 1.4633385007038568, + "learning_rate": 3.484920678188521e-06, + "loss": 1.042, + "step": 2846 + }, + { + "epoch": 0.25675249131983585, + "grad_norm": 2.0338013105244372, + "learning_rate": 3.4845292579712063e-06, + "loss": 1.0384, + "step": 2847 + }, + { + "epoch": 0.25684267484330614, + "grad_norm": 1.829034693076592, + "learning_rate": 3.484137711085669e-06, + "loss": 0.8943, + "step": 2848 + }, + { + "epoch": 0.2569328583667764, + "grad_norm": 1.4249509301018903, + "learning_rate": 3.4837460375653198e-06, + "loss": 0.9997, + "step": 2849 + }, + { + "epoch": 0.25702304189024666, + "grad_norm": 1.479436912385341, + "learning_rate": 3.483354237443576e-06, + "loss": 1.0477, + "step": 2850 + }, + { + "epoch": 0.2571132254137169, + "grad_norm": 1.4011200477579964, + "learning_rate": 3.48296231075387e-06, + "loss": 0.9554, + "step": 2851 + }, + { + "epoch": 0.2572034089371872, + "grad_norm": 1.5112688839271347, + "learning_rate": 3.4825702575296433e-06, + "loss": 1.0285, + "step": 2852 + }, + { + "epoch": 0.2572935924606574, + "grad_norm": 0.7086576018748026, + "learning_rate": 3.482178077804347e-06, + "loss": 0.8513, + "step": 2853 + }, + { + "epoch": 0.2573837759841277, + "grad_norm": 2.1936037194151616, + "learning_rate": 3.4817857716114443e-06, + "loss": 0.975, + "step": 2854 + }, + { + "epoch": 0.25747395950759794, + "grad_norm": 1.682224150287985, + "learning_rate": 3.4813933389844094e-06, + "loss": 0.966, + "step": 2855 + }, + { + "epoch": 0.2575641430310682, + "grad_norm": 1.679675363896019, + "learning_rate": 3.4810007799567264e-06, + "loss": 1.0173, + "step": 2856 + }, + { + "epoch": 0.25765432655453846, + "grad_norm": 2.363536070649903, + "learning_rate": 3.480608094561891e-06, + "loss": 1.0294, + "step": 2857 + }, + { + "epoch": 0.25774451007800875, + "grad_norm": 1.6980706205565954, + "learning_rate": 3.4802152828334083e-06, + "loss": 1.0345, + "step": 2858 + }, + { + "epoch": 0.25783469360147904, + "grad_norm": 1.7407767566978674, + "learning_rate": 3.479822344804796e-06, + "loss": 0.9506, + "step": 2859 + }, + { + "epoch": 0.25792487712494927, + "grad_norm": 1.9674887477257317, + "learning_rate": 3.479429280509582e-06, + "loss": 1.0702, + "step": 2860 + }, + { + "epoch": 0.25801506064841956, + "grad_norm": 1.5591517623088416, + "learning_rate": 3.4790360899813038e-06, + "loss": 1.0549, + "step": 2861 + }, + { + "epoch": 0.2581052441718898, + "grad_norm": 1.966663950848186, + "learning_rate": 3.4786427732535115e-06, + "loss": 0.9488, + "step": 2862 + }, + { + "epoch": 0.2581954276953601, + "grad_norm": 2.553391434064068, + "learning_rate": 3.478249330359764e-06, + "loss": 1.0574, + "step": 2863 + }, + { + "epoch": 0.2582856112188303, + "grad_norm": 1.6060749032734827, + "learning_rate": 3.4778557613336333e-06, + "loss": 0.9469, + "step": 2864 + }, + { + "epoch": 0.2583757947423006, + "grad_norm": 1.691563482222597, + "learning_rate": 3.4774620662087004e-06, + "loss": 1.0291, + "step": 2865 + }, + { + "epoch": 0.25846597826577083, + "grad_norm": 1.9086314475185262, + "learning_rate": 3.477068245018557e-06, + "loss": 1.0153, + "step": 2866 + }, + { + "epoch": 0.2585561617892411, + "grad_norm": 1.802313229517811, + "learning_rate": 3.476674297796807e-06, + "loss": 1.019, + "step": 2867 + }, + { + "epoch": 0.25864634531271136, + "grad_norm": 1.7499147803128996, + "learning_rate": 3.4762802245770627e-06, + "loss": 1.0542, + "step": 2868 + }, + { + "epoch": 0.25873652883618165, + "grad_norm": 1.3060911930778878, + "learning_rate": 3.4758860253929497e-06, + "loss": 0.966, + "step": 2869 + }, + { + "epoch": 0.2588267123596519, + "grad_norm": 1.6907024372861454, + "learning_rate": 3.4754917002781038e-06, + "loss": 1.0316, + "step": 2870 + }, + { + "epoch": 0.25891689588312217, + "grad_norm": 2.6297109202783027, + "learning_rate": 3.475097249266169e-06, + "loss": 1.0503, + "step": 2871 + }, + { + "epoch": 0.2590070794065924, + "grad_norm": 2.085020056429115, + "learning_rate": 3.4747026723908044e-06, + "loss": 0.9956, + "step": 2872 + }, + { + "epoch": 0.2590972629300627, + "grad_norm": 1.7964402709510625, + "learning_rate": 3.474307969685676e-06, + "loss": 0.9935, + "step": 2873 + }, + { + "epoch": 0.2591874464535329, + "grad_norm": 1.5186481802761074, + "learning_rate": 3.473913141184462e-06, + "loss": 0.9628, + "step": 2874 + }, + { + "epoch": 0.2592776299770032, + "grad_norm": 1.7397024452714596, + "learning_rate": 3.4735181869208523e-06, + "loss": 0.8876, + "step": 2875 + }, + { + "epoch": 0.25936781350047344, + "grad_norm": 1.7763386661699796, + "learning_rate": 3.473123106928546e-06, + "loss": 0.8969, + "step": 2876 + }, + { + "epoch": 0.25945799702394373, + "grad_norm": 2.570024282882501, + "learning_rate": 3.4727279012412533e-06, + "loss": 0.9706, + "step": 2877 + }, + { + "epoch": 0.25954818054741396, + "grad_norm": 0.7432748158120074, + "learning_rate": 3.4723325698926953e-06, + "loss": 0.8134, + "step": 2878 + }, + { + "epoch": 0.25963836407088425, + "grad_norm": 2.649101241668736, + "learning_rate": 3.4719371129166045e-06, + "loss": 1.0195, + "step": 2879 + }, + { + "epoch": 0.2597285475943545, + "grad_norm": 5.078462654038798, + "learning_rate": 3.471541530346723e-06, + "loss": 0.9345, + "step": 2880 + }, + { + "epoch": 0.2598187311178248, + "grad_norm": 1.9961806425826434, + "learning_rate": 3.4711458222168037e-06, + "loss": 1.0787, + "step": 2881 + }, + { + "epoch": 0.259908914641295, + "grad_norm": 1.621609231272178, + "learning_rate": 3.4707499885606114e-06, + "loss": 1.0562, + "step": 2882 + }, + { + "epoch": 0.2599990981647653, + "grad_norm": 1.6322898941014186, + "learning_rate": 3.4703540294119204e-06, + "loss": 1.0521, + "step": 2883 + }, + { + "epoch": 0.2600892816882356, + "grad_norm": 1.492480903260504, + "learning_rate": 3.4699579448045163e-06, + "loss": 1.0669, + "step": 2884 + }, + { + "epoch": 0.2601794652117058, + "grad_norm": 1.3128573975562061, + "learning_rate": 3.4695617347721947e-06, + "loss": 0.9016, + "step": 2885 + }, + { + "epoch": 0.2602696487351761, + "grad_norm": 1.7600835752150927, + "learning_rate": 3.469165399348763e-06, + "loss": 0.9807, + "step": 2886 + }, + { + "epoch": 0.26035983225864634, + "grad_norm": 1.5999121373609178, + "learning_rate": 3.4687689385680384e-06, + "loss": 0.9946, + "step": 2887 + }, + { + "epoch": 0.26045001578211663, + "grad_norm": 1.5553362082432876, + "learning_rate": 3.4683723524638494e-06, + "loss": 0.9924, + "step": 2888 + }, + { + "epoch": 0.26054019930558686, + "grad_norm": 1.8825129353761478, + "learning_rate": 3.4679756410700354e-06, + "loss": 1.0847, + "step": 2889 + }, + { + "epoch": 0.26063038282905715, + "grad_norm": 2.1396375174441786, + "learning_rate": 3.4675788044204445e-06, + "loss": 1.0074, + "step": 2890 + }, + { + "epoch": 0.2607205663525274, + "grad_norm": 2.066074164374976, + "learning_rate": 3.467181842548938e-06, + "loss": 0.9454, + "step": 2891 + }, + { + "epoch": 0.26081074987599767, + "grad_norm": 1.7341641134634047, + "learning_rate": 3.466784755489387e-06, + "loss": 1.0092, + "step": 2892 + }, + { + "epoch": 0.2609009333994679, + "grad_norm": 1.565940045545188, + "learning_rate": 3.4663875432756726e-06, + "loss": 1.1309, + "step": 2893 + }, + { + "epoch": 0.2609911169229382, + "grad_norm": 1.711066602600064, + "learning_rate": 3.465990205941687e-06, + "loss": 1.1806, + "step": 2894 + }, + { + "epoch": 0.2610813004464084, + "grad_norm": 2.4777164117582036, + "learning_rate": 3.465592743521335e-06, + "loss": 0.9994, + "step": 2895 + }, + { + "epoch": 0.2611714839698787, + "grad_norm": 1.8587267090108441, + "learning_rate": 3.465195156048528e-06, + "loss": 1.0628, + "step": 2896 + }, + { + "epoch": 0.26126166749334895, + "grad_norm": 0.5617163285374344, + "learning_rate": 3.464797443557191e-06, + "loss": 0.8107, + "step": 2897 + }, + { + "epoch": 0.26135185101681924, + "grad_norm": 1.5224451207384786, + "learning_rate": 3.46439960608126e-06, + "loss": 0.9884, + "step": 2898 + }, + { + "epoch": 0.26144203454028947, + "grad_norm": 2.081893715958997, + "learning_rate": 3.4640016436546797e-06, + "loss": 0.9506, + "step": 2899 + }, + { + "epoch": 0.26153221806375976, + "grad_norm": 2.0185783100637593, + "learning_rate": 3.4636035563114065e-06, + "loss": 0.99, + "step": 2900 + }, + { + "epoch": 0.26162240158723, + "grad_norm": 1.3722037145307755, + "learning_rate": 3.4632053440854085e-06, + "loss": 0.9413, + "step": 2901 + }, + { + "epoch": 0.2617125851107003, + "grad_norm": 1.6480923693612495, + "learning_rate": 3.462807007010662e-06, + "loss": 1.0675, + "step": 2902 + }, + { + "epoch": 0.2618027686341705, + "grad_norm": 7.360750559807807, + "learning_rate": 3.462408545121155e-06, + "loss": 1.0416, + "step": 2903 + }, + { + "epoch": 0.2618929521576408, + "grad_norm": 1.8913608726784141, + "learning_rate": 3.4620099584508883e-06, + "loss": 1.0725, + "step": 2904 + }, + { + "epoch": 0.26198313568111103, + "grad_norm": 3.70840335451233, + "learning_rate": 3.46161124703387e-06, + "loss": 0.9736, + "step": 2905 + }, + { + "epoch": 0.2620733192045813, + "grad_norm": 1.7618948093584885, + "learning_rate": 3.461212410904122e-06, + "loss": 1.0057, + "step": 2906 + }, + { + "epoch": 0.2621635027280516, + "grad_norm": 1.5947199095419629, + "learning_rate": 3.4608134500956726e-06, + "loss": 1.0663, + "step": 2907 + }, + { + "epoch": 0.26225368625152184, + "grad_norm": 1.3924502277896105, + "learning_rate": 3.4604143646425655e-06, + "loss": 0.9991, + "step": 2908 + }, + { + "epoch": 0.26234386977499213, + "grad_norm": 1.5303213456474474, + "learning_rate": 3.460015154578852e-06, + "loss": 1.0813, + "step": 2909 + }, + { + "epoch": 0.26243405329846237, + "grad_norm": 5.4384408772565545, + "learning_rate": 3.459615819938595e-06, + "loss": 0.9868, + "step": 2910 + }, + { + "epoch": 0.26252423682193265, + "grad_norm": 2.3166103433948013, + "learning_rate": 3.4592163607558684e-06, + "loss": 0.9765, + "step": 2911 + }, + { + "epoch": 0.2626144203454029, + "grad_norm": 1.3218458853326749, + "learning_rate": 3.4588167770647553e-06, + "loss": 1.0171, + "step": 2912 + }, + { + "epoch": 0.2627046038688732, + "grad_norm": 2.0381815092364826, + "learning_rate": 3.458417068899351e-06, + "loss": 1.1017, + "step": 2913 + }, + { + "epoch": 0.2627947873923434, + "grad_norm": 1.545222314685022, + "learning_rate": 3.4580172362937612e-06, + "loss": 0.9841, + "step": 2914 + }, + { + "epoch": 0.2628849709158137, + "grad_norm": 1.3127640503985238, + "learning_rate": 3.457617279282101e-06, + "loss": 0.9772, + "step": 2915 + }, + { + "epoch": 0.26297515443928393, + "grad_norm": 2.152798640925993, + "learning_rate": 3.4572171978984975e-06, + "loss": 0.9931, + "step": 2916 + }, + { + "epoch": 0.2630653379627542, + "grad_norm": 1.4847524062774835, + "learning_rate": 3.456816992177088e-06, + "loss": 1.0137, + "step": 2917 + }, + { + "epoch": 0.26315552148622445, + "grad_norm": 1.5663609676522514, + "learning_rate": 3.4564166621520193e-06, + "loss": 0.9885, + "step": 2918 + }, + { + "epoch": 0.26324570500969474, + "grad_norm": 2.6110577668045907, + "learning_rate": 3.4560162078574507e-06, + "loss": 1.0029, + "step": 2919 + }, + { + "epoch": 0.263335888533165, + "grad_norm": 1.6302279226407854, + "learning_rate": 3.455615629327551e-06, + "loss": 1.018, + "step": 2920 + }, + { + "epoch": 0.26342607205663526, + "grad_norm": 2.2584526448093696, + "learning_rate": 3.4552149265964994e-06, + "loss": 0.9912, + "step": 2921 + }, + { + "epoch": 0.2635162555801055, + "grad_norm": 1.5454796556992978, + "learning_rate": 3.4548140996984866e-06, + "loss": 0.9596, + "step": 2922 + }, + { + "epoch": 0.2636064391035758, + "grad_norm": 1.4376200750207055, + "learning_rate": 3.4544131486677124e-06, + "loss": 1.0347, + "step": 2923 + }, + { + "epoch": 0.263696622627046, + "grad_norm": 1.4516911148891474, + "learning_rate": 3.454012073538389e-06, + "loss": 0.9512, + "step": 2924 + }, + { + "epoch": 0.2637868061505163, + "grad_norm": 1.5788836780050184, + "learning_rate": 3.453610874344738e-06, + "loss": 1.0182, + "step": 2925 + }, + { + "epoch": 0.26387698967398654, + "grad_norm": 1.7476875831001828, + "learning_rate": 3.453209551120993e-06, + "loss": 1.0502, + "step": 2926 + }, + { + "epoch": 0.26396717319745683, + "grad_norm": 1.7779389778538992, + "learning_rate": 3.452808103901395e-06, + "loss": 0.9857, + "step": 2927 + }, + { + "epoch": 0.26405735672092706, + "grad_norm": 8.5913628609264, + "learning_rate": 3.4524065327202e-06, + "loss": 1.0873, + "step": 2928 + }, + { + "epoch": 0.26414754024439735, + "grad_norm": 0.6482922954450443, + "learning_rate": 3.4520048376116702e-06, + "loss": 0.763, + "step": 2929 + }, + { + "epoch": 0.26423772376786764, + "grad_norm": 1.550407922042275, + "learning_rate": 3.4516030186100817e-06, + "loss": 1.0507, + "step": 2930 + }, + { + "epoch": 0.26432790729133787, + "grad_norm": 1.4630105716847464, + "learning_rate": 3.4512010757497197e-06, + "loss": 0.93, + "step": 2931 + }, + { + "epoch": 0.26441809081480816, + "grad_norm": 1.5026660908633338, + "learning_rate": 3.4507990090648804e-06, + "loss": 0.9289, + "step": 2932 + }, + { + "epoch": 0.2645082743382784, + "grad_norm": 1.8838399402635635, + "learning_rate": 3.4503968185898696e-06, + "loss": 1.035, + "step": 2933 + }, + { + "epoch": 0.2645984578617487, + "grad_norm": 1.6640364953816476, + "learning_rate": 3.4499945043590047e-06, + "loss": 1.0176, + "step": 2934 + }, + { + "epoch": 0.2646886413852189, + "grad_norm": 1.8945693808582198, + "learning_rate": 3.4495920664066137e-06, + "loss": 1.0268, + "step": 2935 + }, + { + "epoch": 0.2647788249086892, + "grad_norm": 1.8853783436739044, + "learning_rate": 3.449189504767035e-06, + "loss": 1.0016, + "step": 2936 + }, + { + "epoch": 0.26486900843215944, + "grad_norm": 1.7043013403670302, + "learning_rate": 3.4487868194746163e-06, + "loss": 1.0372, + "step": 2937 + }, + { + "epoch": 0.2649591919556297, + "grad_norm": 1.70684862360205, + "learning_rate": 3.4483840105637175e-06, + "loss": 1.0536, + "step": 2938 + }, + { + "epoch": 0.26504937547909996, + "grad_norm": 1.815457791690075, + "learning_rate": 3.4479810780687097e-06, + "loss": 0.9986, + "step": 2939 + }, + { + "epoch": 0.26513955900257025, + "grad_norm": 1.4908523578115838, + "learning_rate": 3.4475780220239714e-06, + "loss": 1.0038, + "step": 2940 + }, + { + "epoch": 0.2652297425260405, + "grad_norm": 1.558724993005758, + "learning_rate": 3.4471748424638948e-06, + "loss": 1.0003, + "step": 2941 + }, + { + "epoch": 0.26531992604951077, + "grad_norm": 0.5974164494152989, + "learning_rate": 3.4467715394228803e-06, + "loss": 0.8307, + "step": 2942 + }, + { + "epoch": 0.265410109572981, + "grad_norm": 1.2862321240338648, + "learning_rate": 3.4463681129353413e-06, + "loss": 1.0125, + "step": 2943 + }, + { + "epoch": 0.2655002930964513, + "grad_norm": 1.47883721385458, + "learning_rate": 3.4459645630357e-06, + "loss": 1.0028, + "step": 2944 + }, + { + "epoch": 0.2655904766199215, + "grad_norm": 1.7281753118824066, + "learning_rate": 3.4455608897583884e-06, + "loss": 1.0499, + "step": 2945 + }, + { + "epoch": 0.2656806601433918, + "grad_norm": 1.8231150782553502, + "learning_rate": 3.4451570931378514e-06, + "loss": 1.0155, + "step": 2946 + }, + { + "epoch": 0.26577084366686204, + "grad_norm": 1.5824309421073313, + "learning_rate": 3.444753173208543e-06, + "loss": 1.02, + "step": 2947 + }, + { + "epoch": 0.26586102719033233, + "grad_norm": 1.6436283777300724, + "learning_rate": 3.444349130004927e-06, + "loss": 1.0107, + "step": 2948 + }, + { + "epoch": 0.26595121071380257, + "grad_norm": 1.804918546872946, + "learning_rate": 3.4439449635614794e-06, + "loss": 0.9344, + "step": 2949 + }, + { + "epoch": 0.26604139423727285, + "grad_norm": 1.5532270309736804, + "learning_rate": 3.4435406739126854e-06, + "loss": 1.0384, + "step": 2950 + }, + { + "epoch": 0.2661315777607431, + "grad_norm": 2.1872429832967706, + "learning_rate": 3.443136261093042e-06, + "loss": 0.8218, + "step": 2951 + }, + { + "epoch": 0.2662217612842134, + "grad_norm": 1.9798746214771805, + "learning_rate": 3.4427317251370553e-06, + "loss": 0.9579, + "step": 2952 + }, + { + "epoch": 0.2663119448076836, + "grad_norm": 2.1065719630084447, + "learning_rate": 3.4423270660792422e-06, + "loss": 0.9227, + "step": 2953 + }, + { + "epoch": 0.2664021283311539, + "grad_norm": 1.8226996883051136, + "learning_rate": 3.4419222839541314e-06, + "loss": 0.9027, + "step": 2954 + }, + { + "epoch": 0.2664923118546242, + "grad_norm": 1.3996322046537202, + "learning_rate": 3.4415173787962607e-06, + "loss": 0.983, + "step": 2955 + }, + { + "epoch": 0.2665824953780944, + "grad_norm": 2.47774980165267, + "learning_rate": 3.4411123506401783e-06, + "loss": 0.8736, + "step": 2956 + }, + { + "epoch": 0.2666726789015647, + "grad_norm": 2.006609961045395, + "learning_rate": 3.440707199520444e-06, + "loss": 0.8613, + "step": 2957 + }, + { + "epoch": 0.26676286242503494, + "grad_norm": 1.4653827946992588, + "learning_rate": 3.440301925471628e-06, + "loss": 0.9713, + "step": 2958 + }, + { + "epoch": 0.26685304594850523, + "grad_norm": 1.8350320084610012, + "learning_rate": 3.43989652852831e-06, + "loss": 1.0477, + "step": 2959 + }, + { + "epoch": 0.26694322947197546, + "grad_norm": 1.9394108363340254, + "learning_rate": 3.4394910087250804e-06, + "loss": 1.0736, + "step": 2960 + }, + { + "epoch": 0.26703341299544575, + "grad_norm": 1.8917007538445259, + "learning_rate": 3.4390853660965405e-06, + "loss": 1.0259, + "step": 2961 + }, + { + "epoch": 0.267123596518916, + "grad_norm": 2.007419770420246, + "learning_rate": 3.438679600677302e-06, + "loss": 1.089, + "step": 2962 + }, + { + "epoch": 0.2672137800423863, + "grad_norm": 2.3137252242432678, + "learning_rate": 3.4382737125019874e-06, + "loss": 0.9633, + "step": 2963 + }, + { + "epoch": 0.2673039635658565, + "grad_norm": 1.441615557269957, + "learning_rate": 3.4378677016052294e-06, + "loss": 0.9769, + "step": 2964 + }, + { + "epoch": 0.2673941470893268, + "grad_norm": 1.7580534367213636, + "learning_rate": 3.43746156802167e-06, + "loss": 1.0296, + "step": 2965 + }, + { + "epoch": 0.267484330612797, + "grad_norm": 1.5907276892349702, + "learning_rate": 3.4370553117859643e-06, + "loss": 0.9434, + "step": 2966 + }, + { + "epoch": 0.2675745141362673, + "grad_norm": 1.8498948711733083, + "learning_rate": 3.4366489329327754e-06, + "loss": 1.0371, + "step": 2967 + }, + { + "epoch": 0.26766469765973755, + "grad_norm": 2.9068358456376386, + "learning_rate": 3.4362424314967777e-06, + "loss": 1.032, + "step": 2968 + }, + { + "epoch": 0.26775488118320784, + "grad_norm": 3.8431789315125027, + "learning_rate": 3.4358358075126567e-06, + "loss": 1.0413, + "step": 2969 + }, + { + "epoch": 0.26784506470667807, + "grad_norm": 1.7720434953074584, + "learning_rate": 3.4354290610151077e-06, + "loss": 1.0288, + "step": 2970 + }, + { + "epoch": 0.26793524823014836, + "grad_norm": 1.3584223950136187, + "learning_rate": 3.4350221920388354e-06, + "loss": 1.0179, + "step": 2971 + }, + { + "epoch": 0.2680254317536186, + "grad_norm": 1.786223378500293, + "learning_rate": 3.4346152006185574e-06, + "loss": 1.1074, + "step": 2972 + }, + { + "epoch": 0.2681156152770889, + "grad_norm": 1.8402870761788375, + "learning_rate": 3.4342080867890006e-06, + "loss": 1.1057, + "step": 2973 + }, + { + "epoch": 0.2682057988005591, + "grad_norm": 4.089012141604553, + "learning_rate": 3.4338008505849016e-06, + "loss": 0.9303, + "step": 2974 + }, + { + "epoch": 0.2682959823240294, + "grad_norm": 2.374282829019945, + "learning_rate": 3.433393492041008e-06, + "loss": 1.0186, + "step": 2975 + }, + { + "epoch": 0.26838616584749964, + "grad_norm": 1.614153166862023, + "learning_rate": 3.432986011192078e-06, + "loss": 1.0608, + "step": 2976 + }, + { + "epoch": 0.2684763493709699, + "grad_norm": 1.9571212778303941, + "learning_rate": 3.4325784080728796e-06, + "loss": 1.078, + "step": 2977 + }, + { + "epoch": 0.2685665328944402, + "grad_norm": 2.961721351761255, + "learning_rate": 3.4321706827181926e-06, + "loss": 1.0383, + "step": 2978 + }, + { + "epoch": 0.26865671641791045, + "grad_norm": 1.3980532922887914, + "learning_rate": 3.4317628351628064e-06, + "loss": 0.9618, + "step": 2979 + }, + { + "epoch": 0.26874689994138073, + "grad_norm": 1.4776866206319104, + "learning_rate": 3.43135486544152e-06, + "loss": 0.9936, + "step": 2980 + }, + { + "epoch": 0.26883708346485097, + "grad_norm": 1.547085680245365, + "learning_rate": 3.4309467735891442e-06, + "loss": 1.0486, + "step": 2981 + }, + { + "epoch": 0.26892726698832126, + "grad_norm": 1.9969131967535925, + "learning_rate": 3.4305385596405e-06, + "loss": 1.0999, + "step": 2982 + }, + { + "epoch": 0.2690174505117915, + "grad_norm": 1.65581629130798, + "learning_rate": 3.4301302236304174e-06, + "loss": 0.9738, + "step": 2983 + }, + { + "epoch": 0.2691076340352618, + "grad_norm": 0.5997827633347861, + "learning_rate": 3.429721765593739e-06, + "loss": 0.7857, + "step": 2984 + }, + { + "epoch": 0.269197817558732, + "grad_norm": 1.45113490841838, + "learning_rate": 3.4293131855653155e-06, + "loss": 1.0364, + "step": 2985 + }, + { + "epoch": 0.2692880010822023, + "grad_norm": 1.7814955960330923, + "learning_rate": 3.4289044835800102e-06, + "loss": 0.9469, + "step": 2986 + }, + { + "epoch": 0.26937818460567253, + "grad_norm": 2.165565712297497, + "learning_rate": 3.4284956596726953e-06, + "loss": 1.0096, + "step": 2987 + }, + { + "epoch": 0.2694683681291428, + "grad_norm": 1.437769988869414, + "learning_rate": 3.4280867138782544e-06, + "loss": 1.0076, + "step": 2988 + }, + { + "epoch": 0.26955855165261305, + "grad_norm": 2.1068232042163535, + "learning_rate": 3.4276776462315803e-06, + "loss": 1.0129, + "step": 2989 + }, + { + "epoch": 0.26964873517608334, + "grad_norm": 1.7646362777147457, + "learning_rate": 3.427268456767578e-06, + "loss": 1.0133, + "step": 2990 + }, + { + "epoch": 0.2697389186995536, + "grad_norm": 6.126829535831139, + "learning_rate": 3.42685914552116e-06, + "loss": 1.0465, + "step": 2991 + }, + { + "epoch": 0.26982910222302386, + "grad_norm": 1.440961402127293, + "learning_rate": 3.426449712527253e-06, + "loss": 0.9807, + "step": 2992 + }, + { + "epoch": 0.2699192857464941, + "grad_norm": 2.304622843207186, + "learning_rate": 3.4260401578207904e-06, + "loss": 0.925, + "step": 2993 + }, + { + "epoch": 0.2700094692699644, + "grad_norm": 2.2245094873042115, + "learning_rate": 3.4256304814367185e-06, + "loss": 0.9714, + "step": 2994 + }, + { + "epoch": 0.2700996527934346, + "grad_norm": 2.8241357646057046, + "learning_rate": 3.4252206834099936e-06, + "loss": 0.9714, + "step": 2995 + }, + { + "epoch": 0.2701898363169049, + "grad_norm": 1.5350967667487514, + "learning_rate": 3.424810763775581e-06, + "loss": 1.0068, + "step": 2996 + }, + { + "epoch": 0.27028001984037514, + "grad_norm": 1.527084211459391, + "learning_rate": 3.4244007225684587e-06, + "loss": 0.9498, + "step": 2997 + }, + { + "epoch": 0.27037020336384543, + "grad_norm": 1.9847775749292762, + "learning_rate": 3.4239905598236115e-06, + "loss": 1.0162, + "step": 2998 + }, + { + "epoch": 0.27046038688731566, + "grad_norm": 1.4409858069743398, + "learning_rate": 3.4235802755760386e-06, + "loss": 1.0396, + "step": 2999 + }, + { + "epoch": 0.27055057041078595, + "grad_norm": 0.8394979785351843, + "learning_rate": 3.4231698698607464e-06, + "loss": 0.9316, + "step": 3000 + }, + { + "epoch": 0.2706407539342562, + "grad_norm": 1.6216528238696934, + "learning_rate": 3.4227593427127543e-06, + "loss": 1.0608, + "step": 3001 + }, + { + "epoch": 0.2707309374577265, + "grad_norm": 3.0671510217161866, + "learning_rate": 3.42234869416709e-06, + "loss": 0.9822, + "step": 3002 + }, + { + "epoch": 0.27082112098119676, + "grad_norm": 3.3827690844139506, + "learning_rate": 3.421937924258792e-06, + "loss": 1.0078, + "step": 3003 + }, + { + "epoch": 0.270911304504667, + "grad_norm": 2.012100687158188, + "learning_rate": 3.4215270330229096e-06, + "loss": 1.0577, + "step": 3004 + }, + { + "epoch": 0.2710014880281373, + "grad_norm": 1.618188004900763, + "learning_rate": 3.421116020494503e-06, + "loss": 0.9756, + "step": 3005 + }, + { + "epoch": 0.2710916715516075, + "grad_norm": 1.3421392323315875, + "learning_rate": 3.420704886708642e-06, + "loss": 0.9519, + "step": 3006 + }, + { + "epoch": 0.2711818550750778, + "grad_norm": 1.1594266813126217, + "learning_rate": 3.4202936317004056e-06, + "loss": 0.9619, + "step": 3007 + }, + { + "epoch": 0.27127203859854804, + "grad_norm": 1.5680626841539713, + "learning_rate": 3.4198822555048856e-06, + "loss": 1.0243, + "step": 3008 + }, + { + "epoch": 0.2713622221220183, + "grad_norm": 2.8458941414528174, + "learning_rate": 3.419470758157182e-06, + "loss": 1.0947, + "step": 3009 + }, + { + "epoch": 0.27145240564548856, + "grad_norm": 1.6614308210464765, + "learning_rate": 3.4190591396924068e-06, + "loss": 1.0088, + "step": 3010 + }, + { + "epoch": 0.27154258916895885, + "grad_norm": 0.6761672400498053, + "learning_rate": 3.418647400145681e-06, + "loss": 0.8535, + "step": 3011 + }, + { + "epoch": 0.2716327726924291, + "grad_norm": 0.7098596807628115, + "learning_rate": 3.4182355395521367e-06, + "loss": 0.8134, + "step": 3012 + }, + { + "epoch": 0.27172295621589937, + "grad_norm": 1.7232395144859836, + "learning_rate": 3.417823557946916e-06, + "loss": 1.0085, + "step": 3013 + }, + { + "epoch": 0.2718131397393696, + "grad_norm": 1.5844863491318968, + "learning_rate": 3.417411455365172e-06, + "loss": 0.9044, + "step": 3014 + }, + { + "epoch": 0.2719033232628399, + "grad_norm": 1.7402597890040692, + "learning_rate": 3.416999231842066e-06, + "loss": 0.9196, + "step": 3015 + }, + { + "epoch": 0.2719935067863101, + "grad_norm": 2.015346300832756, + "learning_rate": 3.416586887412773e-06, + "loss": 1.0597, + "step": 3016 + }, + { + "epoch": 0.2720836903097804, + "grad_norm": 1.5556338703029828, + "learning_rate": 3.416174422112476e-06, + "loss": 0.9727, + "step": 3017 + }, + { + "epoch": 0.27217387383325065, + "grad_norm": 0.5924524635447873, + "learning_rate": 3.4157618359763687e-06, + "loss": 0.7771, + "step": 3018 + }, + { + "epoch": 0.27226405735672093, + "grad_norm": 1.4584164822806838, + "learning_rate": 3.4153491290396542e-06, + "loss": 0.9284, + "step": 3019 + }, + { + "epoch": 0.27235424088019117, + "grad_norm": 1.6148619954813703, + "learning_rate": 3.4149363013375485e-06, + "loss": 1.0278, + "step": 3020 + }, + { + "epoch": 0.27244442440366146, + "grad_norm": 0.6931742461298352, + "learning_rate": 3.414523352905276e-06, + "loss": 0.7934, + "step": 3021 + }, + { + "epoch": 0.2725346079271317, + "grad_norm": 1.9121204722689338, + "learning_rate": 3.414110283778071e-06, + "loss": 0.9804, + "step": 3022 + }, + { + "epoch": 0.272624791450602, + "grad_norm": 1.9099694533177256, + "learning_rate": 3.4136970939911797e-06, + "loss": 0.9525, + "step": 3023 + }, + { + "epoch": 0.2727149749740722, + "grad_norm": 1.644857849889769, + "learning_rate": 3.413283783579857e-06, + "loss": 0.9792, + "step": 3024 + }, + { + "epoch": 0.2728051584975425, + "grad_norm": 2.1714641813714093, + "learning_rate": 3.412870352579369e-06, + "loss": 0.9397, + "step": 3025 + }, + { + "epoch": 0.2728953420210128, + "grad_norm": 0.6528124695773777, + "learning_rate": 3.4124568010249915e-06, + "loss": 0.8345, + "step": 3026 + }, + { + "epoch": 0.272985525544483, + "grad_norm": 2.9129756097109696, + "learning_rate": 3.4120431289520124e-06, + "loss": 0.99, + "step": 3027 + }, + { + "epoch": 0.2730757090679533, + "grad_norm": 1.8728813917708729, + "learning_rate": 3.4116293363957276e-06, + "loss": 1.0156, + "step": 3028 + }, + { + "epoch": 0.27316589259142354, + "grad_norm": 2.0089939070476763, + "learning_rate": 3.4112154233914438e-06, + "loss": 1.0216, + "step": 3029 + }, + { + "epoch": 0.27325607611489383, + "grad_norm": 1.7407631975030045, + "learning_rate": 3.410801389974479e-06, + "loss": 0.9138, + "step": 3030 + }, + { + "epoch": 0.27334625963836406, + "grad_norm": 2.331417750646019, + "learning_rate": 3.410387236180161e-06, + "loss": 1.0404, + "step": 3031 + }, + { + "epoch": 0.27343644316183435, + "grad_norm": 4.2202297300027425, + "learning_rate": 3.409972962043826e-06, + "loss": 0.8913, + "step": 3032 + }, + { + "epoch": 0.2735266266853046, + "grad_norm": 1.9628697332532488, + "learning_rate": 3.4095585676008234e-06, + "loss": 1.063, + "step": 3033 + }, + { + "epoch": 0.2736168102087749, + "grad_norm": 1.479971527908287, + "learning_rate": 3.4091440528865125e-06, + "loss": 1.0031, + "step": 3034 + }, + { + "epoch": 0.2737069937322451, + "grad_norm": 1.8101430234726303, + "learning_rate": 3.4087294179362606e-06, + "loss": 1.0015, + "step": 3035 + }, + { + "epoch": 0.2737971772557154, + "grad_norm": 1.77581372480995, + "learning_rate": 3.4083146627854474e-06, + "loss": 1.0552, + "step": 3036 + }, + { + "epoch": 0.27388736077918563, + "grad_norm": 2.9127703299742365, + "learning_rate": 3.4078997874694614e-06, + "loss": 0.9343, + "step": 3037 + }, + { + "epoch": 0.2739775443026559, + "grad_norm": 2.095015300066667, + "learning_rate": 3.407484792023703e-06, + "loss": 0.8999, + "step": 3038 + }, + { + "epoch": 0.27406772782612615, + "grad_norm": 1.98328664174403, + "learning_rate": 3.407069676483581e-06, + "loss": 0.9786, + "step": 3039 + }, + { + "epoch": 0.27415791134959644, + "grad_norm": 1.4988808430499028, + "learning_rate": 3.406654440884516e-06, + "loss": 1.0434, + "step": 3040 + }, + { + "epoch": 0.2742480948730667, + "grad_norm": 1.9884931829852188, + "learning_rate": 3.4062390852619372e-06, + "loss": 1.0318, + "step": 3041 + }, + { + "epoch": 0.27433827839653696, + "grad_norm": 3.138696030700062, + "learning_rate": 3.4058236096512867e-06, + "loss": 1.0016, + "step": 3042 + }, + { + "epoch": 0.2744284619200072, + "grad_norm": 2.471092271879745, + "learning_rate": 3.405408014088013e-06, + "loss": 0.91, + "step": 3043 + }, + { + "epoch": 0.2745186454434775, + "grad_norm": 2.3243973350836544, + "learning_rate": 3.404992298607579e-06, + "loss": 0.8934, + "step": 3044 + }, + { + "epoch": 0.2746088289669477, + "grad_norm": 2.1366625121141216, + "learning_rate": 3.4045764632454547e-06, + "loss": 0.9895, + "step": 3045 + }, + { + "epoch": 0.274699012490418, + "grad_norm": 8.909085434750944, + "learning_rate": 3.4041605080371223e-06, + "loss": 0.9714, + "step": 3046 + }, + { + "epoch": 0.27478919601388824, + "grad_norm": 3.783759553290444, + "learning_rate": 3.4037444330180726e-06, + "loss": 1.0213, + "step": 3047 + }, + { + "epoch": 0.2748793795373585, + "grad_norm": 2.074132195514894, + "learning_rate": 3.403328238223808e-06, + "loss": 1.018, + "step": 3048 + }, + { + "epoch": 0.27496956306082876, + "grad_norm": 1.982477557523102, + "learning_rate": 3.4029119236898395e-06, + "loss": 1.0325, + "step": 3049 + }, + { + "epoch": 0.27505974658429905, + "grad_norm": 1.0314891422308627, + "learning_rate": 3.4024954894516906e-06, + "loss": 0.8462, + "step": 3050 + }, + { + "epoch": 0.27514993010776934, + "grad_norm": 2.4925947185490704, + "learning_rate": 3.4020789355448933e-06, + "loss": 1.0106, + "step": 3051 + }, + { + "epoch": 0.27524011363123957, + "grad_norm": 1.9382293312902443, + "learning_rate": 3.40166226200499e-06, + "loss": 1.079, + "step": 3052 + }, + { + "epoch": 0.27533029715470986, + "grad_norm": 2.471372346372368, + "learning_rate": 3.401245468867534e-06, + "loss": 0.9915, + "step": 3053 + }, + { + "epoch": 0.2754204806781801, + "grad_norm": 0.7064347886242134, + "learning_rate": 3.400828556168088e-06, + "loss": 0.8702, + "step": 3054 + }, + { + "epoch": 0.2755106642016504, + "grad_norm": 2.974594306531926, + "learning_rate": 3.4004115239422255e-06, + "loss": 1.0279, + "step": 3055 + }, + { + "epoch": 0.2756008477251206, + "grad_norm": 1.8190549509599254, + "learning_rate": 3.3999943722255305e-06, + "loss": 0.9927, + "step": 3056 + }, + { + "epoch": 0.2756910312485909, + "grad_norm": 1.545113302107087, + "learning_rate": 3.3995771010535955e-06, + "loss": 0.9049, + "step": 3057 + }, + { + "epoch": 0.27578121477206113, + "grad_norm": 0.6942145218635639, + "learning_rate": 3.3991597104620253e-06, + "loss": 0.7773, + "step": 3058 + }, + { + "epoch": 0.2758713982955314, + "grad_norm": 1.6720385605139192, + "learning_rate": 3.398742200486434e-06, + "loss": 0.9888, + "step": 3059 + }, + { + "epoch": 0.27596158181900166, + "grad_norm": 1.9231546995134057, + "learning_rate": 3.3983245711624453e-06, + "loss": 1.0259, + "step": 3060 + }, + { + "epoch": 0.27605176534247194, + "grad_norm": 0.771574724918637, + "learning_rate": 3.3979068225256946e-06, + "loss": 0.8924, + "step": 3061 + }, + { + "epoch": 0.2761419488659422, + "grad_norm": 1.9088520988532953, + "learning_rate": 3.3974889546118246e-06, + "loss": 0.9753, + "step": 3062 + }, + { + "epoch": 0.27623213238941247, + "grad_norm": 2.195200201128514, + "learning_rate": 3.3970709674564918e-06, + "loss": 0.9503, + "step": 3063 + }, + { + "epoch": 0.2763223159128827, + "grad_norm": 1.5103769582154658, + "learning_rate": 3.3966528610953607e-06, + "loss": 1.0039, + "step": 3064 + }, + { + "epoch": 0.276412499436353, + "grad_norm": 1.7950394252602158, + "learning_rate": 3.3962346355641067e-06, + "loss": 1.0144, + "step": 3065 + }, + { + "epoch": 0.2765026829598232, + "grad_norm": 1.5269603979874837, + "learning_rate": 3.3958162908984146e-06, + "loss": 1.04, + "step": 3066 + }, + { + "epoch": 0.2765928664832935, + "grad_norm": 1.6832822436268953, + "learning_rate": 3.39539782713398e-06, + "loss": 1.0001, + "step": 3067 + }, + { + "epoch": 0.27668305000676374, + "grad_norm": 1.9447087955917974, + "learning_rate": 3.394979244306509e-06, + "loss": 1.036, + "step": 3068 + }, + { + "epoch": 0.27677323353023403, + "grad_norm": 1.6194982559369016, + "learning_rate": 3.3945605424517166e-06, + "loss": 1.017, + "step": 3069 + }, + { + "epoch": 0.27686341705370426, + "grad_norm": 1.8603622636462236, + "learning_rate": 3.3941417216053294e-06, + "loss": 0.9994, + "step": 3070 + }, + { + "epoch": 0.27695360057717455, + "grad_norm": 1.7711988931953948, + "learning_rate": 3.3937227818030835e-06, + "loss": 0.9669, + "step": 3071 + }, + { + "epoch": 0.2770437841006448, + "grad_norm": 1.7804437787061125, + "learning_rate": 3.393303723080725e-06, + "loss": 0.9703, + "step": 3072 + }, + { + "epoch": 0.2771339676241151, + "grad_norm": 2.091140943420951, + "learning_rate": 3.3928845454740097e-06, + "loss": 1.0161, + "step": 3073 + }, + { + "epoch": 0.27722415114758536, + "grad_norm": 1.7931874792117546, + "learning_rate": 3.392465249018705e-06, + "loss": 0.9896, + "step": 3074 + }, + { + "epoch": 0.2773143346710556, + "grad_norm": 1.8074455159102332, + "learning_rate": 3.3920458337505872e-06, + "loss": 1.0001, + "step": 3075 + }, + { + "epoch": 0.2774045181945259, + "grad_norm": 2.356385010106993, + "learning_rate": 3.391626299705443e-06, + "loss": 0.8443, + "step": 3076 + }, + { + "epoch": 0.2774947017179961, + "grad_norm": 10.918694421881124, + "learning_rate": 3.39120664691907e-06, + "loss": 1.0164, + "step": 3077 + }, + { + "epoch": 0.2775848852414664, + "grad_norm": 1.733100560343038, + "learning_rate": 3.390786875427275e-06, + "loss": 0.9818, + "step": 3078 + }, + { + "epoch": 0.27767506876493664, + "grad_norm": 1.7044657757949624, + "learning_rate": 3.390366985265875e-06, + "loss": 1.0039, + "step": 3079 + }, + { + "epoch": 0.2777652522884069, + "grad_norm": 2.1323508856494584, + "learning_rate": 3.389946976470697e-06, + "loss": 0.9577, + "step": 3080 + }, + { + "epoch": 0.27785543581187716, + "grad_norm": 1.6452227880467587, + "learning_rate": 3.3895268490775787e-06, + "loss": 1.0968, + "step": 3081 + }, + { + "epoch": 0.27794561933534745, + "grad_norm": 1.7757642497613784, + "learning_rate": 3.3891066031223685e-06, + "loss": 0.9577, + "step": 3082 + }, + { + "epoch": 0.2780358028588177, + "grad_norm": 1.8576036879709057, + "learning_rate": 3.3886862386409237e-06, + "loss": 1.0348, + "step": 3083 + }, + { + "epoch": 0.27812598638228797, + "grad_norm": 4.565705727784546, + "learning_rate": 3.388265755669111e-06, + "loss": 0.9632, + "step": 3084 + }, + { + "epoch": 0.2782161699057582, + "grad_norm": 2.24458360408453, + "learning_rate": 3.3878451542428093e-06, + "loss": 1.0683, + "step": 3085 + }, + { + "epoch": 0.2783063534292285, + "grad_norm": 1.6865651402179775, + "learning_rate": 3.387424434397907e-06, + "loss": 0.9894, + "step": 3086 + }, + { + "epoch": 0.2783965369526987, + "grad_norm": 1.8813499373445164, + "learning_rate": 3.3870035961703013e-06, + "loss": 0.9159, + "step": 3087 + }, + { + "epoch": 0.278486720476169, + "grad_norm": 2.0579007048012627, + "learning_rate": 3.3865826395959014e-06, + "loss": 1.1294, + "step": 3088 + }, + { + "epoch": 0.27857690399963925, + "grad_norm": 1.5770420097604618, + "learning_rate": 3.3861615647106253e-06, + "loss": 1.09, + "step": 3089 + }, + { + "epoch": 0.27866708752310954, + "grad_norm": 3.3687253480479904, + "learning_rate": 3.3857403715504012e-06, + "loss": 1.0187, + "step": 3090 + }, + { + "epoch": 0.27875727104657977, + "grad_norm": 1.9283541501481398, + "learning_rate": 3.385319060151167e-06, + "loss": 0.9664, + "step": 3091 + }, + { + "epoch": 0.27884745457005006, + "grad_norm": 1.8149118490329308, + "learning_rate": 3.3848976305488728e-06, + "loss": 0.9913, + "step": 3092 + }, + { + "epoch": 0.2789376380935203, + "grad_norm": 4.257815859732046, + "learning_rate": 3.384476082779476e-06, + "loss": 1.0451, + "step": 3093 + }, + { + "epoch": 0.2790278216169906, + "grad_norm": 2.28537221034814, + "learning_rate": 3.3840544168789463e-06, + "loss": 0.9808, + "step": 3094 + }, + { + "epoch": 0.2791180051404608, + "grad_norm": 1.5692701533641369, + "learning_rate": 3.3836326328832617e-06, + "loss": 1.1106, + "step": 3095 + }, + { + "epoch": 0.2792081886639311, + "grad_norm": 1.9051426657373312, + "learning_rate": 3.383210730828412e-06, + "loss": 0.9546, + "step": 3096 + }, + { + "epoch": 0.2792983721874014, + "grad_norm": 1.4264400266654964, + "learning_rate": 3.3827887107503953e-06, + "loss": 1.0736, + "step": 3097 + }, + { + "epoch": 0.2793885557108716, + "grad_norm": 1.7203399240246433, + "learning_rate": 3.3823665726852216e-06, + "loss": 0.9523, + "step": 3098 + }, + { + "epoch": 0.2794787392343419, + "grad_norm": 1.4948071557637004, + "learning_rate": 3.3819443166689095e-06, + "loss": 1.1259, + "step": 3099 + }, + { + "epoch": 0.27956892275781214, + "grad_norm": 1.6188138441066464, + "learning_rate": 3.3815219427374886e-06, + "loss": 0.9988, + "step": 3100 + }, + { + "epoch": 0.27965910628128243, + "grad_norm": 1.655287768996771, + "learning_rate": 3.3810994509269975e-06, + "loss": 1.0206, + "step": 3101 + }, + { + "epoch": 0.27974928980475267, + "grad_norm": 1.6134374510583864, + "learning_rate": 3.3806768412734864e-06, + "loss": 1.0392, + "step": 3102 + }, + { + "epoch": 0.27983947332822295, + "grad_norm": 1.741528647336942, + "learning_rate": 3.380254113813014e-06, + "loss": 1.0822, + "step": 3103 + }, + { + "epoch": 0.2799296568516932, + "grad_norm": 2.546452773966465, + "learning_rate": 3.3798312685816496e-06, + "loss": 1.0396, + "step": 3104 + }, + { + "epoch": 0.2800198403751635, + "grad_norm": 1.7221123838093384, + "learning_rate": 3.3794083056154738e-06, + "loss": 1.0046, + "step": 3105 + }, + { + "epoch": 0.2801100238986337, + "grad_norm": 1.875778227473584, + "learning_rate": 3.3789852249505746e-06, + "loss": 1.0165, + "step": 3106 + }, + { + "epoch": 0.280200207422104, + "grad_norm": 1.7236938811552378, + "learning_rate": 3.378562026623053e-06, + "loss": 1.0599, + "step": 3107 + }, + { + "epoch": 0.28029039094557423, + "grad_norm": 1.9518003320352244, + "learning_rate": 3.3781387106690175e-06, + "loss": 1.0788, + "step": 3108 + }, + { + "epoch": 0.2803805744690445, + "grad_norm": 1.6142456275770838, + "learning_rate": 3.3777152771245885e-06, + "loss": 1.0352, + "step": 3109 + }, + { + "epoch": 0.28047075799251475, + "grad_norm": 2.0859658457226677, + "learning_rate": 3.377291726025895e-06, + "loss": 1.0374, + "step": 3110 + }, + { + "epoch": 0.28056094151598504, + "grad_norm": 2.0898124906394626, + "learning_rate": 3.3768680574090782e-06, + "loss": 1.119, + "step": 3111 + }, + { + "epoch": 0.2806511250394553, + "grad_norm": 2.453295659245906, + "learning_rate": 3.3764442713102857e-06, + "loss": 0.9797, + "step": 3112 + }, + { + "epoch": 0.28074130856292556, + "grad_norm": 1.8046678153942008, + "learning_rate": 3.3760203677656786e-06, + "loss": 1.002, + "step": 3113 + }, + { + "epoch": 0.2808314920863958, + "grad_norm": 2.0088564759896332, + "learning_rate": 3.3755963468114262e-06, + "loss": 1.0596, + "step": 3114 + }, + { + "epoch": 0.2809216756098661, + "grad_norm": 1.58706144147366, + "learning_rate": 3.3751722084837095e-06, + "loss": 0.912, + "step": 3115 + }, + { + "epoch": 0.2810118591333363, + "grad_norm": 1.5666957216747497, + "learning_rate": 3.3747479528187166e-06, + "loss": 0.9841, + "step": 3116 + }, + { + "epoch": 0.2811020426568066, + "grad_norm": 1.7562064060217288, + "learning_rate": 3.3743235798526485e-06, + "loss": 1.0552, + "step": 3117 + }, + { + "epoch": 0.28119222618027684, + "grad_norm": 1.399738920936132, + "learning_rate": 3.373899089621714e-06, + "loss": 0.9396, + "step": 3118 + }, + { + "epoch": 0.2812824097037471, + "grad_norm": 1.7510664279086419, + "learning_rate": 3.373474482162134e-06, + "loss": 0.9846, + "step": 3119 + }, + { + "epoch": 0.28137259322721736, + "grad_norm": 3.8066477650731523, + "learning_rate": 3.3730497575101376e-06, + "loss": 1.0083, + "step": 3120 + }, + { + "epoch": 0.28146277675068765, + "grad_norm": 1.6504542592438238, + "learning_rate": 3.3726249157019654e-06, + "loss": 1.0003, + "step": 3121 + }, + { + "epoch": 0.28155296027415794, + "grad_norm": 4.674508975756567, + "learning_rate": 3.372199956773866e-06, + "loss": 1.0926, + "step": 3122 + }, + { + "epoch": 0.28164314379762817, + "grad_norm": 1.551501051808676, + "learning_rate": 3.371774880762101e-06, + "loss": 0.9575, + "step": 3123 + }, + { + "epoch": 0.28173332732109846, + "grad_norm": 1.8063764719261173, + "learning_rate": 3.3713496877029392e-06, + "loss": 0.9959, + "step": 3124 + }, + { + "epoch": 0.2818235108445687, + "grad_norm": 2.030144023578063, + "learning_rate": 3.37092437763266e-06, + "loss": 1.072, + "step": 3125 + }, + { + "epoch": 0.281913694368039, + "grad_norm": 1.9705215839287022, + "learning_rate": 3.3704989505875537e-06, + "loss": 1.0375, + "step": 3126 + }, + { + "epoch": 0.2820038778915092, + "grad_norm": 1.8042202299182506, + "learning_rate": 3.3700734066039205e-06, + "loss": 0.9769, + "step": 3127 + }, + { + "epoch": 0.2820940614149795, + "grad_norm": 1.693012628836039, + "learning_rate": 3.36964774571807e-06, + "loss": 1.0338, + "step": 3128 + }, + { + "epoch": 0.28218424493844974, + "grad_norm": 2.5917174866856603, + "learning_rate": 3.3692219679663206e-06, + "loss": 1.004, + "step": 3129 + }, + { + "epoch": 0.28227442846192, + "grad_norm": 2.424410569435774, + "learning_rate": 3.3687960733850043e-06, + "loss": 1.0221, + "step": 3130 + }, + { + "epoch": 0.28236461198539026, + "grad_norm": 1.7148920630251159, + "learning_rate": 3.3683700620104586e-06, + "loss": 0.9404, + "step": 3131 + }, + { + "epoch": 0.28245479550886055, + "grad_norm": 1.7977079368299393, + "learning_rate": 3.3679439338790347e-06, + "loss": 1.0249, + "step": 3132 + }, + { + "epoch": 0.2825449790323308, + "grad_norm": 1.831163017628782, + "learning_rate": 3.3675176890270916e-06, + "loss": 1.1275, + "step": 3133 + }, + { + "epoch": 0.28263516255580107, + "grad_norm": 1.837738063208368, + "learning_rate": 3.367091327490998e-06, + "loss": 1.0295, + "step": 3134 + }, + { + "epoch": 0.2827253460792713, + "grad_norm": 2.3735084869638046, + "learning_rate": 3.3666648493071347e-06, + "loss": 0.9895, + "step": 3135 + }, + { + "epoch": 0.2828155296027416, + "grad_norm": 2.019998345138325, + "learning_rate": 3.3662382545118914e-06, + "loss": 0.9845, + "step": 3136 + }, + { + "epoch": 0.2829057131262118, + "grad_norm": 1.650138375230898, + "learning_rate": 3.3658115431416663e-06, + "loss": 1.0894, + "step": 3137 + }, + { + "epoch": 0.2829958966496821, + "grad_norm": 0.7054971522362303, + "learning_rate": 3.36538471523287e-06, + "loss": 0.8462, + "step": 3138 + }, + { + "epoch": 0.28308608017315234, + "grad_norm": 1.7130158921198815, + "learning_rate": 3.3649577708219204e-06, + "loss": 1.0655, + "step": 3139 + }, + { + "epoch": 0.28317626369662263, + "grad_norm": 2.7619123809142483, + "learning_rate": 3.3645307099452477e-06, + "loss": 0.9498, + "step": 3140 + }, + { + "epoch": 0.28326644722009287, + "grad_norm": 1.9811811200512153, + "learning_rate": 3.3641035326392907e-06, + "loss": 1.0077, + "step": 3141 + }, + { + "epoch": 0.28335663074356315, + "grad_norm": 2.0070023739149256, + "learning_rate": 3.363676238940499e-06, + "loss": 1.0152, + "step": 3142 + }, + { + "epoch": 0.2834468142670334, + "grad_norm": 1.9153668999847018, + "learning_rate": 3.363248828885331e-06, + "loss": 1.0401, + "step": 3143 + }, + { + "epoch": 0.2835369977905037, + "grad_norm": 1.6169379590033723, + "learning_rate": 3.3628213025102562e-06, + "loss": 1.032, + "step": 3144 + }, + { + "epoch": 0.28362718131397396, + "grad_norm": 1.660675183142028, + "learning_rate": 3.3623936598517536e-06, + "loss": 1.0419, + "step": 3145 + }, + { + "epoch": 0.2837173648374442, + "grad_norm": 2.0798116163995943, + "learning_rate": 3.3619659009463117e-06, + "loss": 0.8545, + "step": 3146 + }, + { + "epoch": 0.2838075483609145, + "grad_norm": 2.2199268712207876, + "learning_rate": 3.3615380258304287e-06, + "loss": 1.0612, + "step": 3147 + }, + { + "epoch": 0.2838977318843847, + "grad_norm": 1.8028026828243995, + "learning_rate": 3.3611100345406146e-06, + "loss": 0.9984, + "step": 3148 + }, + { + "epoch": 0.283987915407855, + "grad_norm": 1.698919861913619, + "learning_rate": 3.3606819271133873e-06, + "loss": 0.9946, + "step": 3149 + }, + { + "epoch": 0.28407809893132524, + "grad_norm": 0.7308679397709132, + "learning_rate": 3.360253703585275e-06, + "loss": 0.8309, + "step": 3150 + }, + { + "epoch": 0.28416828245479553, + "grad_norm": 2.5603449063116717, + "learning_rate": 3.3598253639928164e-06, + "loss": 0.9457, + "step": 3151 + }, + { + "epoch": 0.28425846597826576, + "grad_norm": 2.7730722106214696, + "learning_rate": 3.3593969083725596e-06, + "loss": 0.9576, + "step": 3152 + }, + { + "epoch": 0.28434864950173605, + "grad_norm": 1.657288567591882, + "learning_rate": 3.358968336761063e-06, + "loss": 0.962, + "step": 3153 + }, + { + "epoch": 0.2844388330252063, + "grad_norm": 2.0998606544766236, + "learning_rate": 3.3585396491948945e-06, + "loss": 0.9714, + "step": 3154 + }, + { + "epoch": 0.28452901654867657, + "grad_norm": 2.2115721314077534, + "learning_rate": 3.358110845710633e-06, + "loss": 1.0639, + "step": 3155 + }, + { + "epoch": 0.2846192000721468, + "grad_norm": 1.7118708185630223, + "learning_rate": 3.357681926344865e-06, + "loss": 0.9933, + "step": 3156 + }, + { + "epoch": 0.2847093835956171, + "grad_norm": 1.7286290018431323, + "learning_rate": 3.357252891134189e-06, + "loss": 1.0047, + "step": 3157 + }, + { + "epoch": 0.2847995671190873, + "grad_norm": 2.1034322029651213, + "learning_rate": 3.356823740115212e-06, + "loss": 1.0527, + "step": 3158 + }, + { + "epoch": 0.2848897506425576, + "grad_norm": 1.3246848487332525, + "learning_rate": 3.3563944733245525e-06, + "loss": 0.9517, + "step": 3159 + }, + { + "epoch": 0.28497993416602785, + "grad_norm": 2.2538950372744275, + "learning_rate": 3.3559650907988375e-06, + "loss": 0.9973, + "step": 3160 + }, + { + "epoch": 0.28507011768949814, + "grad_norm": 1.809110168155854, + "learning_rate": 3.3555355925747045e-06, + "loss": 0.8572, + "step": 3161 + }, + { + "epoch": 0.28516030121296837, + "grad_norm": 1.6993021626662264, + "learning_rate": 3.3551059786888e-06, + "loss": 0.9903, + "step": 3162 + }, + { + "epoch": 0.28525048473643866, + "grad_norm": 1.8511202642197455, + "learning_rate": 3.3546762491777807e-06, + "loss": 1.0567, + "step": 3163 + }, + { + "epoch": 0.2853406682599089, + "grad_norm": 1.8929586588944427, + "learning_rate": 3.3542464040783156e-06, + "loss": 0.9931, + "step": 3164 + }, + { + "epoch": 0.2854308517833792, + "grad_norm": 2.1739022652732025, + "learning_rate": 3.353816443427079e-06, + "loss": 1.0169, + "step": 3165 + }, + { + "epoch": 0.2855210353068494, + "grad_norm": 1.983663416356893, + "learning_rate": 3.3533863672607597e-06, + "loss": 0.9864, + "step": 3166 + }, + { + "epoch": 0.2856112188303197, + "grad_norm": 1.528276319754214, + "learning_rate": 3.352956175616052e-06, + "loss": 1.0351, + "step": 3167 + }, + { + "epoch": 0.28570140235378993, + "grad_norm": 1.7017082831474428, + "learning_rate": 3.352525868529664e-06, + "loss": 0.9741, + "step": 3168 + }, + { + "epoch": 0.2857915858772602, + "grad_norm": 1.8055992300506019, + "learning_rate": 3.3520954460383103e-06, + "loss": 0.9421, + "step": 3169 + }, + { + "epoch": 0.2858817694007305, + "grad_norm": 2.0817539968057113, + "learning_rate": 3.3516649081787182e-06, + "loss": 0.9442, + "step": 3170 + }, + { + "epoch": 0.28597195292420075, + "grad_norm": 1.5513687366226685, + "learning_rate": 3.3512342549876236e-06, + "loss": 1.1025, + "step": 3171 + }, + { + "epoch": 0.28606213644767103, + "grad_norm": 1.7734640514385949, + "learning_rate": 3.350803486501771e-06, + "loss": 0.9622, + "step": 3172 + }, + { + "epoch": 0.28615231997114127, + "grad_norm": 1.7570148608035718, + "learning_rate": 3.3503726027579175e-06, + "loss": 1.0282, + "step": 3173 + }, + { + "epoch": 0.28624250349461156, + "grad_norm": 1.8253965312814049, + "learning_rate": 3.349941603792827e-06, + "loss": 1.0638, + "step": 3174 + }, + { + "epoch": 0.2863326870180818, + "grad_norm": 1.6762647664829966, + "learning_rate": 3.3495104896432755e-06, + "loss": 1.0021, + "step": 3175 + }, + { + "epoch": 0.2864228705415521, + "grad_norm": 1.5207491899319194, + "learning_rate": 3.3490792603460477e-06, + "loss": 0.9675, + "step": 3176 + }, + { + "epoch": 0.2865130540650223, + "grad_norm": 1.5585552774905913, + "learning_rate": 3.3486479159379393e-06, + "loss": 0.9362, + "step": 3177 + }, + { + "epoch": 0.2866032375884926, + "grad_norm": 0.6735404685973692, + "learning_rate": 3.3482164564557537e-06, + "loss": 0.8133, + "step": 3178 + }, + { + "epoch": 0.28669342111196283, + "grad_norm": 2.4594655312303733, + "learning_rate": 3.3477848819363065e-06, + "loss": 1.0645, + "step": 3179 + }, + { + "epoch": 0.2867836046354331, + "grad_norm": 0.6614800672389736, + "learning_rate": 3.3473531924164213e-06, + "loss": 0.8428, + "step": 3180 + }, + { + "epoch": 0.28687378815890335, + "grad_norm": 2.7121033374093373, + "learning_rate": 3.3469213879329325e-06, + "loss": 0.9607, + "step": 3181 + }, + { + "epoch": 0.28696397168237364, + "grad_norm": 2.371743931623134, + "learning_rate": 3.3464894685226837e-06, + "loss": 0.9657, + "step": 3182 + }, + { + "epoch": 0.2870541552058439, + "grad_norm": 1.7367775574619202, + "learning_rate": 3.34605743422253e-06, + "loss": 1.0815, + "step": 3183 + }, + { + "epoch": 0.28714433872931416, + "grad_norm": 2.030475879757947, + "learning_rate": 3.345625285069333e-06, + "loss": 0.9742, + "step": 3184 + }, + { + "epoch": 0.2872345222527844, + "grad_norm": 1.6345225601249727, + "learning_rate": 3.345193021099967e-06, + "loss": 0.9844, + "step": 3185 + }, + { + "epoch": 0.2873247057762547, + "grad_norm": 1.4070302282284124, + "learning_rate": 3.3447606423513157e-06, + "loss": 0.9738, + "step": 3186 + }, + { + "epoch": 0.2874148892997249, + "grad_norm": 9.164952435742046, + "learning_rate": 3.344328148860271e-06, + "loss": 0.9731, + "step": 3187 + }, + { + "epoch": 0.2875050728231952, + "grad_norm": 1.4756533129575946, + "learning_rate": 3.3438955406637365e-06, + "loss": 1.0145, + "step": 3188 + }, + { + "epoch": 0.28759525634666544, + "grad_norm": 1.61711238373347, + "learning_rate": 3.343462817798624e-06, + "loss": 1.0239, + "step": 3189 + }, + { + "epoch": 0.28768543987013573, + "grad_norm": 3.606108405177945, + "learning_rate": 3.343029980301856e-06, + "loss": 0.9703, + "step": 3190 + }, + { + "epoch": 0.28777562339360596, + "grad_norm": 2.3090555468222385, + "learning_rate": 3.342597028210365e-06, + "loss": 1.0305, + "step": 3191 + }, + { + "epoch": 0.28786580691707625, + "grad_norm": 1.9607471510460197, + "learning_rate": 3.342163961561092e-06, + "loss": 0.9537, + "step": 3192 + }, + { + "epoch": 0.28795599044054654, + "grad_norm": 1.557493658160039, + "learning_rate": 3.34173078039099e-06, + "loss": 0.9738, + "step": 3193 + }, + { + "epoch": 0.28804617396401677, + "grad_norm": 2.286410100221048, + "learning_rate": 3.3412974847370193e-06, + "loss": 1.1722, + "step": 3194 + }, + { + "epoch": 0.28813635748748706, + "grad_norm": 2.763804898245165, + "learning_rate": 3.3408640746361514e-06, + "loss": 1.0286, + "step": 3195 + }, + { + "epoch": 0.2882265410109573, + "grad_norm": 0.7446663268645469, + "learning_rate": 3.3404305501253663e-06, + "loss": 0.8363, + "step": 3196 + }, + { + "epoch": 0.2883167245344276, + "grad_norm": 1.633012531202069, + "learning_rate": 3.3399969112416565e-06, + "loss": 1.0722, + "step": 3197 + }, + { + "epoch": 0.2884069080578978, + "grad_norm": 1.724885497232034, + "learning_rate": 3.3395631580220213e-06, + "loss": 0.9423, + "step": 3198 + }, + { + "epoch": 0.2884970915813681, + "grad_norm": 1.7335068875625168, + "learning_rate": 3.3391292905034714e-06, + "loss": 0.9837, + "step": 3199 + }, + { + "epoch": 0.28858727510483834, + "grad_norm": 1.7400855140938485, + "learning_rate": 3.338695308723027e-06, + "loss": 1.0671, + "step": 3200 + }, + { + "epoch": 0.2886774586283086, + "grad_norm": 1.7775309369792514, + "learning_rate": 3.338261212717716e-06, + "loss": 1.0597, + "step": 3201 + }, + { + "epoch": 0.28876764215177886, + "grad_norm": 1.6627685100584484, + "learning_rate": 3.33782700252458e-06, + "loss": 1.0384, + "step": 3202 + }, + { + "epoch": 0.28885782567524915, + "grad_norm": 2.1075956398473394, + "learning_rate": 3.337392678180668e-06, + "loss": 1.0472, + "step": 3203 + }, + { + "epoch": 0.2889480091987194, + "grad_norm": 1.81229425775302, + "learning_rate": 3.3369582397230377e-06, + "loss": 1.0073, + "step": 3204 + }, + { + "epoch": 0.28903819272218967, + "grad_norm": 1.7853669223806987, + "learning_rate": 3.336523687188759e-06, + "loss": 1.0461, + "step": 3205 + }, + { + "epoch": 0.2891283762456599, + "grad_norm": 0.7092794009418741, + "learning_rate": 3.336089020614909e-06, + "loss": 0.8502, + "step": 3206 + }, + { + "epoch": 0.2892185597691302, + "grad_norm": 1.6366590247717143, + "learning_rate": 3.3356542400385774e-06, + "loss": 0.8829, + "step": 3207 + }, + { + "epoch": 0.2893087432926004, + "grad_norm": 1.687724204534043, + "learning_rate": 3.3352193454968607e-06, + "loss": 0.9396, + "step": 3208 + }, + { + "epoch": 0.2893989268160707, + "grad_norm": 2.5572635832964785, + "learning_rate": 3.3347843370268675e-06, + "loss": 1.0731, + "step": 3209 + }, + { + "epoch": 0.28948911033954094, + "grad_norm": 3.2190172399300376, + "learning_rate": 3.334349214665715e-06, + "loss": 1.0232, + "step": 3210 + }, + { + "epoch": 0.28957929386301123, + "grad_norm": 1.649407268756361, + "learning_rate": 3.3339139784505293e-06, + "loss": 1.0884, + "step": 3211 + }, + { + "epoch": 0.28966947738648147, + "grad_norm": 1.6261852050352197, + "learning_rate": 3.333478628418448e-06, + "loss": 1.0764, + "step": 3212 + }, + { + "epoch": 0.28975966090995176, + "grad_norm": 1.4497817763181091, + "learning_rate": 3.333043164606618e-06, + "loss": 0.8968, + "step": 3213 + }, + { + "epoch": 0.289849844433422, + "grad_norm": 1.8711719217874183, + "learning_rate": 3.3326075870521948e-06, + "loss": 1.088, + "step": 3214 + }, + { + "epoch": 0.2899400279568923, + "grad_norm": 1.745996323156691, + "learning_rate": 3.3321718957923437e-06, + "loss": 1.008, + "step": 3215 + }, + { + "epoch": 0.29003021148036257, + "grad_norm": 1.6957582511807967, + "learning_rate": 3.3317360908642413e-06, + "loss": 1.0038, + "step": 3216 + }, + { + "epoch": 0.2901203950038328, + "grad_norm": 2.81077531279882, + "learning_rate": 3.331300172305072e-06, + "loss": 0.9579, + "step": 3217 + }, + { + "epoch": 0.2902105785273031, + "grad_norm": 1.5058911192797473, + "learning_rate": 3.330864140152032e-06, + "loss": 0.9992, + "step": 3218 + }, + { + "epoch": 0.2903007620507733, + "grad_norm": 1.6871121278684273, + "learning_rate": 3.330427994442325e-06, + "loss": 1.0217, + "step": 3219 + }, + { + "epoch": 0.2903909455742436, + "grad_norm": 1.657382146290397, + "learning_rate": 3.3299917352131657e-06, + "loss": 1.042, + "step": 3220 + }, + { + "epoch": 0.29048112909771384, + "grad_norm": 2.1757727028697698, + "learning_rate": 3.329555362501778e-06, + "loss": 0.9751, + "step": 3221 + }, + { + "epoch": 0.29057131262118413, + "grad_norm": 1.5780316315781602, + "learning_rate": 3.3291188763453954e-06, + "loss": 0.9689, + "step": 3222 + }, + { + "epoch": 0.29066149614465436, + "grad_norm": 1.8090057233400414, + "learning_rate": 3.3286822767812618e-06, + "loss": 0.9964, + "step": 3223 + }, + { + "epoch": 0.29075167966812465, + "grad_norm": 1.8671807324913314, + "learning_rate": 3.32824556384663e-06, + "loss": 0.9897, + "step": 3224 + }, + { + "epoch": 0.2908418631915949, + "grad_norm": 1.4597491158523404, + "learning_rate": 3.3278087375787628e-06, + "loss": 1.006, + "step": 3225 + }, + { + "epoch": 0.2909320467150652, + "grad_norm": 5.03300720236157, + "learning_rate": 3.327371798014933e-06, + "loss": 1.009, + "step": 3226 + }, + { + "epoch": 0.2910222302385354, + "grad_norm": 1.8258557051044386, + "learning_rate": 3.3269347451924218e-06, + "loss": 1.0056, + "step": 3227 + }, + { + "epoch": 0.2911124137620057, + "grad_norm": 1.5448492638346891, + "learning_rate": 3.326497579148522e-06, + "loss": 1.0713, + "step": 3228 + }, + { + "epoch": 0.29120259728547593, + "grad_norm": 1.8188057094575807, + "learning_rate": 3.3260602999205345e-06, + "loss": 1.0094, + "step": 3229 + }, + { + "epoch": 0.2912927808089462, + "grad_norm": 1.9912762521281242, + "learning_rate": 3.32562290754577e-06, + "loss": 1.0205, + "step": 3230 + }, + { + "epoch": 0.29138296433241645, + "grad_norm": 6.914038775559579, + "learning_rate": 3.3251854020615494e-06, + "loss": 0.99, + "step": 3231 + }, + { + "epoch": 0.29147314785588674, + "grad_norm": 1.641547534060342, + "learning_rate": 3.324747783505204e-06, + "loss": 1.0119, + "step": 3232 + }, + { + "epoch": 0.29156333137935697, + "grad_norm": 5.157565602168704, + "learning_rate": 3.324310051914073e-06, + "loss": 0.9246, + "step": 3233 + }, + { + "epoch": 0.29165351490282726, + "grad_norm": 4.902081510665345, + "learning_rate": 3.3238722073255056e-06, + "loss": 0.9716, + "step": 3234 + }, + { + "epoch": 0.2917436984262975, + "grad_norm": 2.0914812464300754, + "learning_rate": 3.323434249776863e-06, + "loss": 0.9687, + "step": 3235 + }, + { + "epoch": 0.2918338819497678, + "grad_norm": 0.7201000885140029, + "learning_rate": 3.3229961793055117e-06, + "loss": 0.8263, + "step": 3236 + }, + { + "epoch": 0.291924065473238, + "grad_norm": 1.4708193546888064, + "learning_rate": 3.3225579959488314e-06, + "loss": 0.9334, + "step": 3237 + }, + { + "epoch": 0.2920142489967083, + "grad_norm": 1.9621227679364173, + "learning_rate": 3.322119699744211e-06, + "loss": 1.03, + "step": 3238 + }, + { + "epoch": 0.29210443252017854, + "grad_norm": 2.1194303112411514, + "learning_rate": 3.3216812907290476e-06, + "loss": 1.0205, + "step": 3239 + }, + { + "epoch": 0.2921946160436488, + "grad_norm": 1.4328846853994295, + "learning_rate": 3.3212427689407484e-06, + "loss": 1.0269, + "step": 3240 + }, + { + "epoch": 0.2922847995671191, + "grad_norm": 0.6487802266188679, + "learning_rate": 3.3208041344167317e-06, + "loss": 0.8053, + "step": 3241 + }, + { + "epoch": 0.29237498309058935, + "grad_norm": 1.7190717222642145, + "learning_rate": 3.3203653871944224e-06, + "loss": 0.9126, + "step": 3242 + }, + { + "epoch": 0.29246516661405964, + "grad_norm": 1.5390905387376939, + "learning_rate": 3.3199265273112587e-06, + "loss": 0.9646, + "step": 3243 + }, + { + "epoch": 0.29255535013752987, + "grad_norm": 1.8800369635458516, + "learning_rate": 3.3194875548046852e-06, + "loss": 0.9482, + "step": 3244 + }, + { + "epoch": 0.29264553366100016, + "grad_norm": 1.503440408269733, + "learning_rate": 3.319048469712158e-06, + "loss": 0.9844, + "step": 3245 + }, + { + "epoch": 0.2927357171844704, + "grad_norm": 1.8357002916698553, + "learning_rate": 3.3186092720711423e-06, + "loss": 1.1012, + "step": 3246 + }, + { + "epoch": 0.2928259007079407, + "grad_norm": 1.6828111568859978, + "learning_rate": 3.3181699619191125e-06, + "loss": 0.9312, + "step": 3247 + }, + { + "epoch": 0.2929160842314109, + "grad_norm": 1.625647929204778, + "learning_rate": 3.3177305392935536e-06, + "loss": 1.0495, + "step": 3248 + }, + { + "epoch": 0.2930062677548812, + "grad_norm": 0.7325050408512125, + "learning_rate": 3.3172910042319595e-06, + "loss": 0.8926, + "step": 3249 + }, + { + "epoch": 0.29309645127835143, + "grad_norm": 1.70546400809579, + "learning_rate": 3.316851356771833e-06, + "loss": 1.0398, + "step": 3250 + }, + { + "epoch": 0.2931866348018217, + "grad_norm": 1.8637699150182996, + "learning_rate": 3.3164115969506876e-06, + "loss": 1.0365, + "step": 3251 + }, + { + "epoch": 0.29327681832529195, + "grad_norm": 2.4149627488384966, + "learning_rate": 3.315971724806046e-06, + "loss": 1.0945, + "step": 3252 + }, + { + "epoch": 0.29336700184876224, + "grad_norm": 1.5437636247412025, + "learning_rate": 3.315531740375441e-06, + "loss": 0.9461, + "step": 3253 + }, + { + "epoch": 0.2934571853722325, + "grad_norm": 1.6971177013628291, + "learning_rate": 3.315091643696414e-06, + "loss": 0.9702, + "step": 3254 + }, + { + "epoch": 0.29354736889570276, + "grad_norm": 1.4870012040072442, + "learning_rate": 3.3146514348065164e-06, + "loss": 0.9308, + "step": 3255 + }, + { + "epoch": 0.293637552419173, + "grad_norm": 1.9398423464121295, + "learning_rate": 3.31421111374331e-06, + "loss": 1.0091, + "step": 3256 + }, + { + "epoch": 0.2937277359426433, + "grad_norm": 3.1286678913969443, + "learning_rate": 3.3137706805443647e-06, + "loss": 1.0161, + "step": 3257 + }, + { + "epoch": 0.2938179194661135, + "grad_norm": 1.569797488351349, + "learning_rate": 3.313330135247261e-06, + "loss": 1.0557, + "step": 3258 + }, + { + "epoch": 0.2939081029895838, + "grad_norm": 1.3971313258452192, + "learning_rate": 3.312889477889588e-06, + "loss": 0.9752, + "step": 3259 + }, + { + "epoch": 0.29399828651305404, + "grad_norm": 1.6339803653223723, + "learning_rate": 3.3124487085089464e-06, + "loss": 1.0154, + "step": 3260 + }, + { + "epoch": 0.29408847003652433, + "grad_norm": 1.8048174262238668, + "learning_rate": 3.312007827142943e-06, + "loss": 1.0408, + "step": 3261 + }, + { + "epoch": 0.29417865355999456, + "grad_norm": 1.414865760905842, + "learning_rate": 3.3115668338291983e-06, + "loss": 0.9216, + "step": 3262 + }, + { + "epoch": 0.29426883708346485, + "grad_norm": 1.704264058669957, + "learning_rate": 3.3111257286053394e-06, + "loss": 1.0867, + "step": 3263 + }, + { + "epoch": 0.29435902060693514, + "grad_norm": 1.6344246822121664, + "learning_rate": 3.3106845115090043e-06, + "loss": 1.0068, + "step": 3264 + }, + { + "epoch": 0.2944492041304054, + "grad_norm": 1.6131933894951995, + "learning_rate": 3.310243182577839e-06, + "loss": 0.955, + "step": 3265 + }, + { + "epoch": 0.29453938765387566, + "grad_norm": 1.616333224461707, + "learning_rate": 3.3098017418495007e-06, + "loss": 0.9001, + "step": 3266 + }, + { + "epoch": 0.2946295711773459, + "grad_norm": 1.728734716980875, + "learning_rate": 3.309360189361656e-06, + "loss": 0.9701, + "step": 3267 + }, + { + "epoch": 0.2947197547008162, + "grad_norm": 1.5389899222209455, + "learning_rate": 3.3089185251519797e-06, + "loss": 1.0849, + "step": 3268 + }, + { + "epoch": 0.2948099382242864, + "grad_norm": 1.9151915042629806, + "learning_rate": 3.3084767492581574e-06, + "loss": 0.9829, + "step": 3269 + }, + { + "epoch": 0.2949001217477567, + "grad_norm": 3.942632932290519, + "learning_rate": 3.3080348617178846e-06, + "loss": 1.0205, + "step": 3270 + }, + { + "epoch": 0.29499030527122694, + "grad_norm": 1.6519169538050866, + "learning_rate": 3.307592862568865e-06, + "loss": 0.9798, + "step": 3271 + }, + { + "epoch": 0.2950804887946972, + "grad_norm": 1.6514761159603624, + "learning_rate": 3.307150751848812e-06, + "loss": 1.0698, + "step": 3272 + }, + { + "epoch": 0.29517067231816746, + "grad_norm": 1.6921150537201202, + "learning_rate": 3.3067085295954497e-06, + "loss": 0.9093, + "step": 3273 + }, + { + "epoch": 0.29526085584163775, + "grad_norm": 1.8206700313615711, + "learning_rate": 3.3062661958465098e-06, + "loss": 0.8703, + "step": 3274 + }, + { + "epoch": 0.295351039365108, + "grad_norm": 1.9118824903098883, + "learning_rate": 3.305823750639736e-06, + "loss": 0.8962, + "step": 3275 + }, + { + "epoch": 0.29544122288857827, + "grad_norm": 1.5363244036045778, + "learning_rate": 3.3053811940128795e-06, + "loss": 0.9974, + "step": 3276 + }, + { + "epoch": 0.2955314064120485, + "grad_norm": 1.9478131503038745, + "learning_rate": 3.3049385260037016e-06, + "loss": 0.9812, + "step": 3277 + }, + { + "epoch": 0.2956215899355188, + "grad_norm": 1.5662444452063964, + "learning_rate": 3.3044957466499736e-06, + "loss": 0.892, + "step": 3278 + }, + { + "epoch": 0.295711773458989, + "grad_norm": 2.9402566514907744, + "learning_rate": 3.304052855989475e-06, + "loss": 0.9208, + "step": 3279 + }, + { + "epoch": 0.2958019569824593, + "grad_norm": 1.5749357119564134, + "learning_rate": 3.3036098540599966e-06, + "loss": 1.0631, + "step": 3280 + }, + { + "epoch": 0.29589214050592955, + "grad_norm": 2.000812008050441, + "learning_rate": 3.3031667408993373e-06, + "loss": 0.9598, + "step": 3281 + }, + { + "epoch": 0.29598232402939983, + "grad_norm": 1.8571978733800691, + "learning_rate": 3.302723516545306e-06, + "loss": 0.9993, + "step": 3282 + }, + { + "epoch": 0.29607250755287007, + "grad_norm": 1.5983690383285638, + "learning_rate": 3.302280181035722e-06, + "loss": 1.0252, + "step": 3283 + }, + { + "epoch": 0.29616269107634036, + "grad_norm": 0.6178641412789088, + "learning_rate": 3.3018367344084117e-06, + "loss": 0.8557, + "step": 3284 + }, + { + "epoch": 0.2962528745998106, + "grad_norm": 1.3806283371739532, + "learning_rate": 3.3013931767012125e-06, + "loss": 0.9921, + "step": 3285 + }, + { + "epoch": 0.2963430581232809, + "grad_norm": 1.492996236327551, + "learning_rate": 3.300949507951972e-06, + "loss": 0.9968, + "step": 3286 + }, + { + "epoch": 0.2964332416467511, + "grad_norm": 2.605995410953208, + "learning_rate": 3.300505728198546e-06, + "loss": 1.0408, + "step": 3287 + }, + { + "epoch": 0.2965234251702214, + "grad_norm": 1.8995117162396855, + "learning_rate": 3.3000618374788e-06, + "loss": 0.9843, + "step": 3288 + }, + { + "epoch": 0.2966136086936917, + "grad_norm": 1.4038592467084392, + "learning_rate": 3.2996178358306104e-06, + "loss": 1.0351, + "step": 3289 + }, + { + "epoch": 0.2967037922171619, + "grad_norm": 1.8471937185333493, + "learning_rate": 3.2991737232918606e-06, + "loss": 0.9648, + "step": 3290 + }, + { + "epoch": 0.2967939757406322, + "grad_norm": 1.8666582195340626, + "learning_rate": 3.298729499900445e-06, + "loss": 0.9496, + "step": 3291 + }, + { + "epoch": 0.29688415926410244, + "grad_norm": 4.089518915810093, + "learning_rate": 3.2982851656942677e-06, + "loss": 0.9487, + "step": 3292 + }, + { + "epoch": 0.29697434278757273, + "grad_norm": 2.569900062191306, + "learning_rate": 3.2978407207112416e-06, + "loss": 1.0245, + "step": 3293 + }, + { + "epoch": 0.29706452631104296, + "grad_norm": 2.5266606676874264, + "learning_rate": 3.2973961649892888e-06, + "loss": 1.0054, + "step": 3294 + }, + { + "epoch": 0.29715470983451325, + "grad_norm": 1.5080827856182824, + "learning_rate": 3.296951498566341e-06, + "loss": 1.0645, + "step": 3295 + }, + { + "epoch": 0.2972448933579835, + "grad_norm": 2.698450092226654, + "learning_rate": 3.2965067214803404e-06, + "loss": 0.9971, + "step": 3296 + }, + { + "epoch": 0.2973350768814538, + "grad_norm": 1.3195913930120238, + "learning_rate": 3.2960618337692372e-06, + "loss": 1.0371, + "step": 3297 + }, + { + "epoch": 0.297425260404924, + "grad_norm": 1.5163228895965815, + "learning_rate": 3.2956168354709927e-06, + "loss": 0.9591, + "step": 3298 + }, + { + "epoch": 0.2975154439283943, + "grad_norm": 1.6250660222653082, + "learning_rate": 3.2951717266235754e-06, + "loss": 1.0477, + "step": 3299 + }, + { + "epoch": 0.29760562745186453, + "grad_norm": 1.4529705991034818, + "learning_rate": 3.294726507264964e-06, + "loss": 0.8713, + "step": 3300 + }, + { + "epoch": 0.2976958109753348, + "grad_norm": 1.7588633469182091, + "learning_rate": 3.2942811774331487e-06, + "loss": 0.9711, + "step": 3301 + }, + { + "epoch": 0.29778599449880505, + "grad_norm": 1.8099985290226308, + "learning_rate": 3.293835737166127e-06, + "loss": 0.997, + "step": 3302 + }, + { + "epoch": 0.29787617802227534, + "grad_norm": 1.6362254416747977, + "learning_rate": 3.293390186501906e-06, + "loss": 0.9459, + "step": 3303 + }, + { + "epoch": 0.2979663615457456, + "grad_norm": 2.237353813467911, + "learning_rate": 3.2929445254785024e-06, + "loss": 1.0099, + "step": 3304 + }, + { + "epoch": 0.29805654506921586, + "grad_norm": 1.7863417681004738, + "learning_rate": 3.2924987541339423e-06, + "loss": 0.9592, + "step": 3305 + }, + { + "epoch": 0.2981467285926861, + "grad_norm": 1.9533824903510908, + "learning_rate": 3.292052872506262e-06, + "loss": 1.0193, + "step": 3306 + }, + { + "epoch": 0.2982369121161564, + "grad_norm": 2.483728864605792, + "learning_rate": 3.291606880633506e-06, + "loss": 1.0378, + "step": 3307 + }, + { + "epoch": 0.2983270956396266, + "grad_norm": 2.5724818573989454, + "learning_rate": 3.2911607785537297e-06, + "loss": 1.0531, + "step": 3308 + }, + { + "epoch": 0.2984172791630969, + "grad_norm": 1.9338332519167396, + "learning_rate": 3.290714566304997e-06, + "loss": 1.0868, + "step": 3309 + }, + { + "epoch": 0.29850746268656714, + "grad_norm": 2.659303861892565, + "learning_rate": 3.2902682439253794e-06, + "loss": 1.0668, + "step": 3310 + }, + { + "epoch": 0.2985976462100374, + "grad_norm": 2.0169205161657398, + "learning_rate": 3.289821811452961e-06, + "loss": 1.0081, + "step": 3311 + }, + { + "epoch": 0.2986878297335077, + "grad_norm": 1.9680687467422693, + "learning_rate": 3.289375268925834e-06, + "loss": 1.0562, + "step": 3312 + }, + { + "epoch": 0.29877801325697795, + "grad_norm": 1.826603443740878, + "learning_rate": 3.288928616382099e-06, + "loss": 0.9864, + "step": 3313 + }, + { + "epoch": 0.29886819678044824, + "grad_norm": 1.5245262754733893, + "learning_rate": 3.288481853859868e-06, + "loss": 1.0338, + "step": 3314 + }, + { + "epoch": 0.29895838030391847, + "grad_norm": 1.5026599029658452, + "learning_rate": 3.2880349813972604e-06, + "loss": 0.9768, + "step": 3315 + }, + { + "epoch": 0.29904856382738876, + "grad_norm": 1.7628362456644464, + "learning_rate": 3.2875879990324052e-06, + "loss": 1.0015, + "step": 3316 + }, + { + "epoch": 0.299138747350859, + "grad_norm": 1.8111175000026691, + "learning_rate": 3.287140906803443e-06, + "loss": 1.055, + "step": 3317 + }, + { + "epoch": 0.2992289308743293, + "grad_norm": 1.9465530913133697, + "learning_rate": 3.2866937047485216e-06, + "loss": 0.9213, + "step": 3318 + }, + { + "epoch": 0.2993191143977995, + "grad_norm": 1.8791256972736483, + "learning_rate": 3.2862463929057985e-06, + "loss": 1.0588, + "step": 3319 + }, + { + "epoch": 0.2994092979212698, + "grad_norm": 2.13687986781148, + "learning_rate": 3.285798971313441e-06, + "loss": 0.8438, + "step": 3320 + }, + { + "epoch": 0.29949948144474003, + "grad_norm": 1.4564768373040298, + "learning_rate": 3.2853514400096248e-06, + "loss": 1.0243, + "step": 3321 + }, + { + "epoch": 0.2995896649682103, + "grad_norm": 1.5566380212395883, + "learning_rate": 3.2849037990325367e-06, + "loss": 1.0269, + "step": 3322 + }, + { + "epoch": 0.29967984849168056, + "grad_norm": 1.5007455085815549, + "learning_rate": 3.2844560484203717e-06, + "loss": 0.9583, + "step": 3323 + }, + { + "epoch": 0.29977003201515084, + "grad_norm": 1.9045741735202086, + "learning_rate": 3.2840081882113333e-06, + "loss": 1.0809, + "step": 3324 + }, + { + "epoch": 0.2998602155386211, + "grad_norm": 1.6224594062934796, + "learning_rate": 3.283560218443638e-06, + "loss": 0.977, + "step": 3325 + }, + { + "epoch": 0.29995039906209137, + "grad_norm": 1.532391648063712, + "learning_rate": 3.2831121391555064e-06, + "loss": 1.0152, + "step": 3326 + }, + { + "epoch": 0.3000405825855616, + "grad_norm": 6.035127174282711, + "learning_rate": 3.2826639503851724e-06, + "loss": 0.9553, + "step": 3327 + }, + { + "epoch": 0.3001307661090319, + "grad_norm": 1.3931872732928876, + "learning_rate": 3.282215652170877e-06, + "loss": 1.0406, + "step": 3328 + }, + { + "epoch": 0.3002209496325021, + "grad_norm": 1.6669974078391092, + "learning_rate": 3.281767244550873e-06, + "loss": 1.0365, + "step": 3329 + }, + { + "epoch": 0.3003111331559724, + "grad_norm": 0.611003338517466, + "learning_rate": 3.2813187275634193e-06, + "loss": 0.7907, + "step": 3330 + }, + { + "epoch": 0.30040131667944264, + "grad_norm": 1.5243011374133169, + "learning_rate": 3.280870101246787e-06, + "loss": 1.0278, + "step": 3331 + }, + { + "epoch": 0.30049150020291293, + "grad_norm": 1.7721676754637232, + "learning_rate": 3.280421365639255e-06, + "loss": 0.9191, + "step": 3332 + }, + { + "epoch": 0.30058168372638316, + "grad_norm": 1.5980929125390502, + "learning_rate": 3.279972520779112e-06, + "loss": 0.9878, + "step": 3333 + }, + { + "epoch": 0.30067186724985345, + "grad_norm": 1.4510379693540345, + "learning_rate": 3.279523566704656e-06, + "loss": 1.0687, + "step": 3334 + }, + { + "epoch": 0.30076205077332374, + "grad_norm": 3.9610719676771433, + "learning_rate": 3.2790745034541935e-06, + "loss": 1.0513, + "step": 3335 + }, + { + "epoch": 0.300852234296794, + "grad_norm": 1.7287204427116236, + "learning_rate": 3.278625331066042e-06, + "loss": 1.0471, + "step": 3336 + }, + { + "epoch": 0.30094241782026426, + "grad_norm": 1.8761377061841484, + "learning_rate": 3.278176049578527e-06, + "loss": 1.0506, + "step": 3337 + }, + { + "epoch": 0.3010326013437345, + "grad_norm": 1.962140569148435, + "learning_rate": 3.2777266590299835e-06, + "loss": 0.9598, + "step": 3338 + }, + { + "epoch": 0.3011227848672048, + "grad_norm": 2.045850541093586, + "learning_rate": 3.2772771594587562e-06, + "loss": 0.9359, + "step": 3339 + }, + { + "epoch": 0.301212968390675, + "grad_norm": 1.6718697235879794, + "learning_rate": 3.2768275509031988e-06, + "loss": 1.1035, + "step": 3340 + }, + { + "epoch": 0.3013031519141453, + "grad_norm": 1.612836577899667, + "learning_rate": 3.276377833401675e-06, + "loss": 0.984, + "step": 3341 + }, + { + "epoch": 0.30139333543761554, + "grad_norm": 1.698711310725722, + "learning_rate": 3.2759280069925557e-06, + "loss": 1.0702, + "step": 3342 + }, + { + "epoch": 0.30148351896108583, + "grad_norm": 1.5948199999896362, + "learning_rate": 3.2754780717142233e-06, + "loss": 1.0396, + "step": 3343 + }, + { + "epoch": 0.30157370248455606, + "grad_norm": 1.7882378022535055, + "learning_rate": 3.27502802760507e-06, + "loss": 0.9783, + "step": 3344 + }, + { + "epoch": 0.30166388600802635, + "grad_norm": 1.8854061639257893, + "learning_rate": 3.2745778747034943e-06, + "loss": 0.9159, + "step": 3345 + }, + { + "epoch": 0.3017540695314966, + "grad_norm": 2.6009767348302395, + "learning_rate": 3.274127613047906e-06, + "loss": 0.9848, + "step": 3346 + }, + { + "epoch": 0.30184425305496687, + "grad_norm": 1.7400977769323516, + "learning_rate": 3.273677242676725e-06, + "loss": 1.0223, + "step": 3347 + }, + { + "epoch": 0.3019344365784371, + "grad_norm": 5.508614123291463, + "learning_rate": 3.2732267636283782e-06, + "loss": 0.9989, + "step": 3348 + }, + { + "epoch": 0.3020246201019074, + "grad_norm": 1.957450776047616, + "learning_rate": 3.2727761759413034e-06, + "loss": 0.9661, + "step": 3349 + }, + { + "epoch": 0.3021148036253776, + "grad_norm": 13.929033521874933, + "learning_rate": 3.2723254796539477e-06, + "loss": 1.013, + "step": 3350 + }, + { + "epoch": 0.3022049871488479, + "grad_norm": 1.8077651034961193, + "learning_rate": 3.271874674804766e-06, + "loss": 0.9848, + "step": 3351 + }, + { + "epoch": 0.30229517067231815, + "grad_norm": 1.727541063651505, + "learning_rate": 3.2714237614322242e-06, + "loss": 1.0295, + "step": 3352 + }, + { + "epoch": 0.30238535419578844, + "grad_norm": 1.4786962199810778, + "learning_rate": 3.2709727395747974e-06, + "loss": 0.8375, + "step": 3353 + }, + { + "epoch": 0.30247553771925867, + "grad_norm": 0.6760823675951445, + "learning_rate": 3.2705216092709673e-06, + "loss": 0.8102, + "step": 3354 + }, + { + "epoch": 0.30256572124272896, + "grad_norm": 2.0506235988511006, + "learning_rate": 3.2700703705592282e-06, + "loss": 1.091, + "step": 3355 + }, + { + "epoch": 0.3026559047661992, + "grad_norm": 1.6865248511683422, + "learning_rate": 3.269619023478082e-06, + "loss": 0.9074, + "step": 3356 + }, + { + "epoch": 0.3027460882896695, + "grad_norm": 1.7246023244918878, + "learning_rate": 3.26916756806604e-06, + "loss": 1.1056, + "step": 3357 + }, + { + "epoch": 0.3028362718131397, + "grad_norm": 1.875085828724058, + "learning_rate": 3.268716004361623e-06, + "loss": 1.0213, + "step": 3358 + }, + { + "epoch": 0.30292645533661, + "grad_norm": 0.6677816474724498, + "learning_rate": 3.268264332403361e-06, + "loss": 0.8252, + "step": 3359 + }, + { + "epoch": 0.3030166388600803, + "grad_norm": 1.5116778227714425, + "learning_rate": 3.2678125522297933e-06, + "loss": 1.0365, + "step": 3360 + }, + { + "epoch": 0.3031068223835505, + "grad_norm": 2.5392520775079936, + "learning_rate": 3.267360663879468e-06, + "loss": 0.9449, + "step": 3361 + }, + { + "epoch": 0.3031970059070208, + "grad_norm": 1.6122608421459181, + "learning_rate": 3.266908667390942e-06, + "loss": 0.9921, + "step": 3362 + }, + { + "epoch": 0.30328718943049104, + "grad_norm": 2.070052681220298, + "learning_rate": 3.2664565628027833e-06, + "loss": 1.0599, + "step": 3363 + }, + { + "epoch": 0.30337737295396133, + "grad_norm": 2.736570338026381, + "learning_rate": 3.2660043501535675e-06, + "loss": 1.0052, + "step": 3364 + }, + { + "epoch": 0.30346755647743157, + "grad_norm": 1.707809585052264, + "learning_rate": 3.2655520294818797e-06, + "loss": 1.0725, + "step": 3365 + }, + { + "epoch": 0.30355774000090185, + "grad_norm": 1.4837899109020873, + "learning_rate": 3.2650996008263146e-06, + "loss": 1.0735, + "step": 3366 + }, + { + "epoch": 0.3036479235243721, + "grad_norm": 2.6533198632165154, + "learning_rate": 3.2646470642254756e-06, + "loss": 0.9611, + "step": 3367 + }, + { + "epoch": 0.3037381070478424, + "grad_norm": 2.5019986746317207, + "learning_rate": 3.2641944197179767e-06, + "loss": 1.0267, + "step": 3368 + }, + { + "epoch": 0.3038282905713126, + "grad_norm": 0.6627740814858764, + "learning_rate": 3.2637416673424383e-06, + "loss": 0.8101, + "step": 3369 + }, + { + "epoch": 0.3039184740947829, + "grad_norm": 1.6446566499936721, + "learning_rate": 3.2632888071374937e-06, + "loss": 1.0175, + "step": 3370 + }, + { + "epoch": 0.30400865761825313, + "grad_norm": 1.7198534111277688, + "learning_rate": 3.2628358391417815e-06, + "loss": 1.0366, + "step": 3371 + }, + { + "epoch": 0.3040988411417234, + "grad_norm": 1.7107892255242672, + "learning_rate": 3.2623827633939526e-06, + "loss": 0.9945, + "step": 3372 + }, + { + "epoch": 0.30418902466519365, + "grad_norm": 1.9904021996295833, + "learning_rate": 3.2619295799326657e-06, + "loss": 0.9954, + "step": 3373 + }, + { + "epoch": 0.30427920818866394, + "grad_norm": 1.9449349768588626, + "learning_rate": 3.2614762887965883e-06, + "loss": 0.9013, + "step": 3374 + }, + { + "epoch": 0.3043693917121342, + "grad_norm": 2.517459177029056, + "learning_rate": 3.2610228900243984e-06, + "loss": 0.9939, + "step": 3375 + }, + { + "epoch": 0.30445957523560446, + "grad_norm": 0.6079939083638077, + "learning_rate": 3.260569383654783e-06, + "loss": 0.8615, + "step": 3376 + }, + { + "epoch": 0.3045497587590747, + "grad_norm": 0.6882298670109328, + "learning_rate": 3.2601157697264365e-06, + "loss": 0.8316, + "step": 3377 + }, + { + "epoch": 0.304639942282545, + "grad_norm": 1.841854802471699, + "learning_rate": 3.2596620482780647e-06, + "loss": 0.9814, + "step": 3378 + }, + { + "epoch": 0.3047301258060152, + "grad_norm": 0.5694355427684608, + "learning_rate": 3.2592082193483803e-06, + "loss": 0.8303, + "step": 3379 + }, + { + "epoch": 0.3048203093294855, + "grad_norm": 1.5239845271889503, + "learning_rate": 3.258754282976109e-06, + "loss": 0.9893, + "step": 3380 + }, + { + "epoch": 0.30491049285295574, + "grad_norm": 1.8669789090020072, + "learning_rate": 3.25830023919998e-06, + "loss": 1.0751, + "step": 3381 + }, + { + "epoch": 0.305000676376426, + "grad_norm": 1.594189078490348, + "learning_rate": 3.2578460880587374e-06, + "loss": 0.96, + "step": 3382 + }, + { + "epoch": 0.3050908598998963, + "grad_norm": 1.8641982151454717, + "learning_rate": 3.2573918295911306e-06, + "loss": 1.0648, + "step": 3383 + }, + { + "epoch": 0.30518104342336655, + "grad_norm": 1.5538929965075898, + "learning_rate": 3.2569374638359196e-06, + "loss": 0.9435, + "step": 3384 + }, + { + "epoch": 0.30527122694683684, + "grad_norm": 1.4319964665310703, + "learning_rate": 3.2564829908318736e-06, + "loss": 1.0197, + "step": 3385 + }, + { + "epoch": 0.30536141047030707, + "grad_norm": 1.3085232843813392, + "learning_rate": 3.2560284106177705e-06, + "loss": 1.0673, + "step": 3386 + }, + { + "epoch": 0.30545159399377736, + "grad_norm": 1.5553785925448667, + "learning_rate": 3.2555737232323978e-06, + "loss": 0.9784, + "step": 3387 + }, + { + "epoch": 0.3055417775172476, + "grad_norm": 1.360563876417919, + "learning_rate": 3.255118928714552e-06, + "loss": 1.0332, + "step": 3388 + }, + { + "epoch": 0.3056319610407179, + "grad_norm": 1.6346831488911473, + "learning_rate": 3.2546640271030386e-06, + "loss": 0.9357, + "step": 3389 + }, + { + "epoch": 0.3057221445641881, + "grad_norm": 0.6228771875516451, + "learning_rate": 3.2542090184366717e-06, + "loss": 0.8624, + "step": 3390 + }, + { + "epoch": 0.3058123280876584, + "grad_norm": 1.9121242752507734, + "learning_rate": 3.253753902754276e-06, + "loss": 1.0861, + "step": 3391 + }, + { + "epoch": 0.30590251161112864, + "grad_norm": 1.5059547641490458, + "learning_rate": 3.253298680094685e-06, + "loss": 0.9358, + "step": 3392 + }, + { + "epoch": 0.3059926951345989, + "grad_norm": 5.44261402492556, + "learning_rate": 3.2528433504967394e-06, + "loss": 0.9471, + "step": 3393 + }, + { + "epoch": 0.30608287865806916, + "grad_norm": 2.1054158672988295, + "learning_rate": 3.252387913999291e-06, + "loss": 1.0152, + "step": 3394 + }, + { + "epoch": 0.30617306218153945, + "grad_norm": 2.28113315230767, + "learning_rate": 3.2519323706411998e-06, + "loss": 1.0059, + "step": 3395 + }, + { + "epoch": 0.3062632457050097, + "grad_norm": 1.6760568814469219, + "learning_rate": 3.251476720461336e-06, + "loss": 0.8938, + "step": 3396 + }, + { + "epoch": 0.30635342922847997, + "grad_norm": 1.5502738987508211, + "learning_rate": 3.251020963498578e-06, + "loss": 0.929, + "step": 3397 + }, + { + "epoch": 0.3064436127519502, + "grad_norm": 1.4598995336167466, + "learning_rate": 3.250565099791813e-06, + "loss": 1.0447, + "step": 3398 + }, + { + "epoch": 0.3065337962754205, + "grad_norm": 1.4052866709114573, + "learning_rate": 3.2501091293799387e-06, + "loss": 1.0578, + "step": 3399 + }, + { + "epoch": 0.3066239797988907, + "grad_norm": 1.5029048607791475, + "learning_rate": 3.24965305230186e-06, + "loss": 1.0179, + "step": 3400 + }, + { + "epoch": 0.306714163322361, + "grad_norm": 2.1322100001432065, + "learning_rate": 3.249196868596492e-06, + "loss": 1.0769, + "step": 3401 + }, + { + "epoch": 0.30680434684583124, + "grad_norm": 1.9573808003814757, + "learning_rate": 3.24874057830276e-06, + "loss": 0.9308, + "step": 3402 + }, + { + "epoch": 0.30689453036930153, + "grad_norm": 1.5449332949364587, + "learning_rate": 3.2482841814595954e-06, + "loss": 0.9277, + "step": 3403 + }, + { + "epoch": 0.30698471389277177, + "grad_norm": 1.6326420161119362, + "learning_rate": 3.247827678105943e-06, + "loss": 0.9435, + "step": 3404 + }, + { + "epoch": 0.30707489741624205, + "grad_norm": 1.5031732531632103, + "learning_rate": 3.247371068280751e-06, + "loss": 1.0032, + "step": 3405 + }, + { + "epoch": 0.3071650809397123, + "grad_norm": 1.3920144843794873, + "learning_rate": 3.2469143520229823e-06, + "loss": 0.8648, + "step": 3406 + }, + { + "epoch": 0.3072552644631826, + "grad_norm": 1.4876636415115039, + "learning_rate": 3.2464575293716054e-06, + "loss": 0.9567, + "step": 3407 + }, + { + "epoch": 0.30734544798665286, + "grad_norm": 1.5539050409572188, + "learning_rate": 3.2460006003655997e-06, + "loss": 0.9459, + "step": 3408 + }, + { + "epoch": 0.3074356315101231, + "grad_norm": 3.84087965565648, + "learning_rate": 3.245543565043952e-06, + "loss": 1.0594, + "step": 3409 + }, + { + "epoch": 0.3075258150335934, + "grad_norm": 1.785890724457339, + "learning_rate": 3.2450864234456592e-06, + "loss": 0.9935, + "step": 3410 + }, + { + "epoch": 0.3076159985570636, + "grad_norm": 1.373500092591462, + "learning_rate": 3.244629175609728e-06, + "loss": 0.9976, + "step": 3411 + }, + { + "epoch": 0.3077061820805339, + "grad_norm": 1.508586309445255, + "learning_rate": 3.2441718215751726e-06, + "loss": 0.9909, + "step": 3412 + }, + { + "epoch": 0.30779636560400414, + "grad_norm": 1.6963076856751522, + "learning_rate": 3.2437143613810173e-06, + "loss": 0.9661, + "step": 3413 + }, + { + "epoch": 0.30788654912747443, + "grad_norm": 2.3901664942444367, + "learning_rate": 3.2432567950662947e-06, + "loss": 1.0489, + "step": 3414 + }, + { + "epoch": 0.30797673265094466, + "grad_norm": 1.481186665013897, + "learning_rate": 3.2427991226700468e-06, + "loss": 0.9885, + "step": 3415 + }, + { + "epoch": 0.30806691617441495, + "grad_norm": 1.9685039669284667, + "learning_rate": 3.2423413442313246e-06, + "loss": 0.9234, + "step": 3416 + }, + { + "epoch": 0.3081570996978852, + "grad_norm": 3.6210778162995583, + "learning_rate": 3.2418834597891904e-06, + "loss": 0.9215, + "step": 3417 + }, + { + "epoch": 0.3082472832213555, + "grad_norm": 2.073895272931683, + "learning_rate": 3.2414254693827098e-06, + "loss": 1.0878, + "step": 3418 + }, + { + "epoch": 0.3083374667448257, + "grad_norm": 1.2899309411896165, + "learning_rate": 3.2409673730509644e-06, + "loss": 0.9816, + "step": 3419 + }, + { + "epoch": 0.308427650268296, + "grad_norm": 1.9225042910087726, + "learning_rate": 3.2405091708330393e-06, + "loss": 0.978, + "step": 3420 + }, + { + "epoch": 0.3085178337917662, + "grad_norm": 1.7481092047634108, + "learning_rate": 3.2400508627680323e-06, + "loss": 0.9905, + "step": 3421 + }, + { + "epoch": 0.3086080173152365, + "grad_norm": 1.9087717230159627, + "learning_rate": 3.2395924488950474e-06, + "loss": 1.0971, + "step": 3422 + }, + { + "epoch": 0.30869820083870675, + "grad_norm": 2.038779052677059, + "learning_rate": 3.2391339292532004e-06, + "loss": 1.1344, + "step": 3423 + }, + { + "epoch": 0.30878838436217704, + "grad_norm": 1.6966491914285122, + "learning_rate": 3.238675303881614e-06, + "loss": 0.9539, + "step": 3424 + }, + { + "epoch": 0.30887856788564727, + "grad_norm": 1.4968017018635227, + "learning_rate": 3.2382165728194203e-06, + "loss": 0.9657, + "step": 3425 + }, + { + "epoch": 0.30896875140911756, + "grad_norm": 2.5439834562579433, + "learning_rate": 3.237757736105761e-06, + "loss": 0.9841, + "step": 3426 + }, + { + "epoch": 0.3090589349325878, + "grad_norm": 1.5434759537702372, + "learning_rate": 3.2372987937797867e-06, + "loss": 1.0818, + "step": 3427 + }, + { + "epoch": 0.3091491184560581, + "grad_norm": 1.3319835486498703, + "learning_rate": 3.2368397458806573e-06, + "loss": 0.9061, + "step": 3428 + }, + { + "epoch": 0.3092393019795283, + "grad_norm": 2.0986580330179523, + "learning_rate": 3.2363805924475412e-06, + "loss": 1.0306, + "step": 3429 + }, + { + "epoch": 0.3093294855029986, + "grad_norm": 1.7070908874289954, + "learning_rate": 3.2359213335196153e-06, + "loss": 0.9412, + "step": 3430 + }, + { + "epoch": 0.3094196690264689, + "grad_norm": 1.5117786011195888, + "learning_rate": 3.2354619691360663e-06, + "loss": 0.9483, + "step": 3431 + }, + { + "epoch": 0.3095098525499391, + "grad_norm": 1.4019287685469144, + "learning_rate": 3.2350024993360898e-06, + "loss": 1.035, + "step": 3432 + }, + { + "epoch": 0.3096000360734094, + "grad_norm": 1.7155332808305788, + "learning_rate": 3.2345429241588902e-06, + "loss": 0.998, + "step": 3433 + }, + { + "epoch": 0.30969021959687965, + "grad_norm": 2.0314982335925373, + "learning_rate": 3.234083243643681e-06, + "loss": 1.0273, + "step": 3434 + }, + { + "epoch": 0.30978040312034993, + "grad_norm": 1.5957989704758406, + "learning_rate": 3.233623457829686e-06, + "loss": 1.0708, + "step": 3435 + }, + { + "epoch": 0.30987058664382017, + "grad_norm": 2.6206474549604946, + "learning_rate": 3.2331635667561344e-06, + "loss": 0.9678, + "step": 3436 + }, + { + "epoch": 0.30996077016729046, + "grad_norm": 1.9320035927069767, + "learning_rate": 3.2327035704622674e-06, + "loss": 1.0277, + "step": 3437 + }, + { + "epoch": 0.3100509536907607, + "grad_norm": 1.3593568910291423, + "learning_rate": 3.2322434689873353e-06, + "loss": 0.894, + "step": 3438 + }, + { + "epoch": 0.310141137214231, + "grad_norm": 1.9765494123787002, + "learning_rate": 3.2317832623705957e-06, + "loss": 1.0854, + "step": 3439 + }, + { + "epoch": 0.3102313207377012, + "grad_norm": 1.5303341209001342, + "learning_rate": 3.231322950651316e-06, + "loss": 0.9459, + "step": 3440 + }, + { + "epoch": 0.3103215042611715, + "grad_norm": 1.5773764619651964, + "learning_rate": 3.2308625338687735e-06, + "loss": 1.0284, + "step": 3441 + }, + { + "epoch": 0.31041168778464173, + "grad_norm": 2.468346091921551, + "learning_rate": 3.230402012062252e-06, + "loss": 1.1043, + "step": 3442 + }, + { + "epoch": 0.310501871308112, + "grad_norm": 0.6105418160510229, + "learning_rate": 3.2299413852710466e-06, + "loss": 0.851, + "step": 3443 + }, + { + "epoch": 0.31059205483158225, + "grad_norm": 1.6537648395959559, + "learning_rate": 3.2294806535344606e-06, + "loss": 0.9592, + "step": 3444 + }, + { + "epoch": 0.31068223835505254, + "grad_norm": 1.8492876073074775, + "learning_rate": 3.2290198168918056e-06, + "loss": 1.0377, + "step": 3445 + }, + { + "epoch": 0.3107724218785228, + "grad_norm": 3.5597660214974267, + "learning_rate": 3.2285588753824035e-06, + "loss": 1.0363, + "step": 3446 + }, + { + "epoch": 0.31086260540199306, + "grad_norm": 1.6087161446945977, + "learning_rate": 3.228097829045584e-06, + "loss": 1.0104, + "step": 3447 + }, + { + "epoch": 0.3109527889254633, + "grad_norm": 1.3941134299727886, + "learning_rate": 3.227636677920685e-06, + "loss": 1.0428, + "step": 3448 + }, + { + "epoch": 0.3110429724489336, + "grad_norm": 1.7113741661483788, + "learning_rate": 3.2271754220470567e-06, + "loss": 1.0453, + "step": 3449 + }, + { + "epoch": 0.3111331559724038, + "grad_norm": 3.687213692622585, + "learning_rate": 3.2267140614640547e-06, + "loss": 1.08, + "step": 3450 + }, + { + "epoch": 0.3112233394958741, + "grad_norm": 4.227959646979765, + "learning_rate": 3.2262525962110445e-06, + "loss": 1.0469, + "step": 3451 + }, + { + "epoch": 0.31131352301934434, + "grad_norm": 3.7636608525477544, + "learning_rate": 3.2257910263274015e-06, + "loss": 0.9268, + "step": 3452 + }, + { + "epoch": 0.31140370654281463, + "grad_norm": 2.0373973843998203, + "learning_rate": 3.225329351852509e-06, + "loss": 0.9458, + "step": 3453 + }, + { + "epoch": 0.3114938900662849, + "grad_norm": 2.146026368084947, + "learning_rate": 3.2248675728257596e-06, + "loss": 1.006, + "step": 3454 + }, + { + "epoch": 0.31158407358975515, + "grad_norm": 2.4047129291977005, + "learning_rate": 3.2244056892865557e-06, + "loss": 1.0755, + "step": 3455 + }, + { + "epoch": 0.31167425711322544, + "grad_norm": 1.5510238342421456, + "learning_rate": 3.2239437012743063e-06, + "loss": 0.9901, + "step": 3456 + }, + { + "epoch": 0.31176444063669567, + "grad_norm": 1.4758412855304917, + "learning_rate": 3.223481608828432e-06, + "loss": 1.0155, + "step": 3457 + }, + { + "epoch": 0.31185462416016596, + "grad_norm": 2.3951435146457323, + "learning_rate": 3.223019411988361e-06, + "loss": 1.0149, + "step": 3458 + }, + { + "epoch": 0.3119448076836362, + "grad_norm": 1.7103107297230555, + "learning_rate": 3.22255711079353e-06, + "loss": 1.0027, + "step": 3459 + }, + { + "epoch": 0.3120349912071065, + "grad_norm": 1.3554674307956927, + "learning_rate": 3.222094705283385e-06, + "loss": 0.9637, + "step": 3460 + }, + { + "epoch": 0.3121251747305767, + "grad_norm": 1.8646912428107272, + "learning_rate": 3.2216321954973805e-06, + "loss": 0.9953, + "step": 3461 + }, + { + "epoch": 0.312215358254047, + "grad_norm": 1.4372111320501901, + "learning_rate": 3.2211695814749816e-06, + "loss": 0.9898, + "step": 3462 + }, + { + "epoch": 0.31230554177751724, + "grad_norm": 2.7872305017039594, + "learning_rate": 3.220706863255661e-06, + "loss": 1.103, + "step": 3463 + }, + { + "epoch": 0.3123957253009875, + "grad_norm": 1.6500743675812264, + "learning_rate": 3.2202440408788994e-06, + "loss": 1.0038, + "step": 3464 + }, + { + "epoch": 0.31248590882445776, + "grad_norm": 1.8840737451548988, + "learning_rate": 3.2197811143841883e-06, + "loss": 1.0694, + "step": 3465 + }, + { + "epoch": 0.31257609234792805, + "grad_norm": 1.3977241614758924, + "learning_rate": 3.2193180838110267e-06, + "loss": 0.9989, + "step": 3466 + }, + { + "epoch": 0.3126662758713983, + "grad_norm": 3.7825264785959263, + "learning_rate": 3.2188549491989225e-06, + "loss": 0.8951, + "step": 3467 + }, + { + "epoch": 0.31275645939486857, + "grad_norm": 1.6115615213305017, + "learning_rate": 3.2183917105873934e-06, + "loss": 1.0246, + "step": 3468 + }, + { + "epoch": 0.3128466429183388, + "grad_norm": 1.86905560753538, + "learning_rate": 3.217928368015966e-06, + "loss": 1.0618, + "step": 3469 + }, + { + "epoch": 0.3129368264418091, + "grad_norm": 1.4828708456316142, + "learning_rate": 3.217464921524174e-06, + "loss": 1.0298, + "step": 3470 + }, + { + "epoch": 0.3130270099652793, + "grad_norm": 1.503604610958473, + "learning_rate": 3.2170013711515616e-06, + "loss": 1.017, + "step": 3471 + }, + { + "epoch": 0.3131171934887496, + "grad_norm": 1.9869074843955974, + "learning_rate": 3.216537716937682e-06, + "loss": 0.9211, + "step": 3472 + }, + { + "epoch": 0.31320737701221985, + "grad_norm": 1.7688273193929045, + "learning_rate": 3.2160739589220968e-06, + "loss": 0.9783, + "step": 3473 + }, + { + "epoch": 0.31329756053569013, + "grad_norm": 1.5920561222458884, + "learning_rate": 3.215610097144376e-06, + "loss": 1.0603, + "step": 3474 + }, + { + "epoch": 0.31338774405916037, + "grad_norm": 1.9527735889921065, + "learning_rate": 3.215146131644099e-06, + "loss": 0.9532, + "step": 3475 + }, + { + "epoch": 0.31347792758263066, + "grad_norm": 1.5630467793303564, + "learning_rate": 3.214682062460854e-06, + "loss": 1.0744, + "step": 3476 + }, + { + "epoch": 0.3135681111061009, + "grad_norm": 1.5922242135109923, + "learning_rate": 3.2142178896342367e-06, + "loss": 0.9173, + "step": 3477 + }, + { + "epoch": 0.3136582946295712, + "grad_norm": 2.0146994424817755, + "learning_rate": 3.2137536132038552e-06, + "loss": 0.9921, + "step": 3478 + }, + { + "epoch": 0.31374847815304147, + "grad_norm": 1.6143510056564871, + "learning_rate": 3.2132892332093226e-06, + "loss": 1.0039, + "step": 3479 + }, + { + "epoch": 0.3138386616765117, + "grad_norm": 1.6305690118572715, + "learning_rate": 3.2128247496902623e-06, + "loss": 1.1172, + "step": 3480 + }, + { + "epoch": 0.313928845199982, + "grad_norm": 1.451613922358056, + "learning_rate": 3.2123601626863064e-06, + "loss": 1.016, + "step": 3481 + }, + { + "epoch": 0.3140190287234522, + "grad_norm": 0.6445236899192626, + "learning_rate": 3.2118954722370974e-06, + "loss": 0.8083, + "step": 3482 + }, + { + "epoch": 0.3141092122469225, + "grad_norm": 1.7567499636976749, + "learning_rate": 3.2114306783822837e-06, + "loss": 0.9456, + "step": 3483 + }, + { + "epoch": 0.31419939577039274, + "grad_norm": 1.5859533600413476, + "learning_rate": 3.210965781161525e-06, + "loss": 0.9783, + "step": 3484 + }, + { + "epoch": 0.31428957929386303, + "grad_norm": 1.5303027278626307, + "learning_rate": 3.2105007806144892e-06, + "loss": 1.1155, + "step": 3485 + }, + { + "epoch": 0.31437976281733326, + "grad_norm": 2.4036187982881416, + "learning_rate": 3.2100356767808513e-06, + "loss": 1.0021, + "step": 3486 + }, + { + "epoch": 0.31446994634080355, + "grad_norm": 2.0081174151375003, + "learning_rate": 3.2095704697002977e-06, + "loss": 1.0602, + "step": 3487 + }, + { + "epoch": 0.3145601298642738, + "grad_norm": 1.560189331498607, + "learning_rate": 3.209105159412522e-06, + "loss": 1.008, + "step": 3488 + }, + { + "epoch": 0.3146503133877441, + "grad_norm": 2.4254450910648355, + "learning_rate": 3.208639745957228e-06, + "loss": 0.9666, + "step": 3489 + }, + { + "epoch": 0.3147404969112143, + "grad_norm": 1.366443278888947, + "learning_rate": 3.2081742293741256e-06, + "loss": 1.0275, + "step": 3490 + }, + { + "epoch": 0.3148306804346846, + "grad_norm": 1.425104937119498, + "learning_rate": 3.2077086097029366e-06, + "loss": 1.0631, + "step": 3491 + }, + { + "epoch": 0.31492086395815483, + "grad_norm": 1.4146380017390865, + "learning_rate": 3.2072428869833895e-06, + "loss": 1.0018, + "step": 3492 + }, + { + "epoch": 0.3150110474816251, + "grad_norm": 1.4493139189449549, + "learning_rate": 3.206777061255223e-06, + "loss": 1.0168, + "step": 3493 + }, + { + "epoch": 0.31510123100509535, + "grad_norm": 1.7865606414000563, + "learning_rate": 3.206311132558183e-06, + "loss": 1.0432, + "step": 3494 + }, + { + "epoch": 0.31519141452856564, + "grad_norm": 3.0551787076585195, + "learning_rate": 3.205845100932026e-06, + "loss": 1.018, + "step": 3495 + }, + { + "epoch": 0.31528159805203587, + "grad_norm": 1.6967303416376707, + "learning_rate": 3.205378966416516e-06, + "loss": 1.032, + "step": 3496 + }, + { + "epoch": 0.31537178157550616, + "grad_norm": 2.2599630178953456, + "learning_rate": 3.204912729051426e-06, + "loss": 0.9982, + "step": 3497 + }, + { + "epoch": 0.3154619650989764, + "grad_norm": 1.6520150942216765, + "learning_rate": 3.2044463888765384e-06, + "loss": 0.9758, + "step": 3498 + }, + { + "epoch": 0.3155521486224467, + "grad_norm": 2.164124966487599, + "learning_rate": 3.2039799459316436e-06, + "loss": 0.9009, + "step": 3499 + }, + { + "epoch": 0.3156423321459169, + "grad_norm": 1.739515093837427, + "learning_rate": 3.2035134002565407e-06, + "loss": 1.0392, + "step": 3500 + }, + { + "epoch": 0.3157325156693872, + "grad_norm": 4.672049528708579, + "learning_rate": 3.203046751891039e-06, + "loss": 1.0163, + "step": 3501 + }, + { + "epoch": 0.3158226991928575, + "grad_norm": 1.7419786541897786, + "learning_rate": 3.2025800008749545e-06, + "loss": 1.0164, + "step": 3502 + }, + { + "epoch": 0.3159128827163277, + "grad_norm": 1.8024267924094002, + "learning_rate": 3.202113147248114e-06, + "loss": 0.9756, + "step": 3503 + }, + { + "epoch": 0.316003066239798, + "grad_norm": 1.7312934236554833, + "learning_rate": 3.20164619105035e-06, + "loss": 0.9768, + "step": 3504 + }, + { + "epoch": 0.31609324976326825, + "grad_norm": 1.3353321421629503, + "learning_rate": 3.201179132321508e-06, + "loss": 0.8554, + "step": 3505 + }, + { + "epoch": 0.31618343328673854, + "grad_norm": 1.7455395666100135, + "learning_rate": 3.200711971101439e-06, + "loss": 0.9434, + "step": 3506 + }, + { + "epoch": 0.31627361681020877, + "grad_norm": 1.8774892178364277, + "learning_rate": 3.2002447074300047e-06, + "loss": 1.0049, + "step": 3507 + }, + { + "epoch": 0.31636380033367906, + "grad_norm": 1.8956165469239294, + "learning_rate": 3.1997773413470736e-06, + "loss": 1.0077, + "step": 3508 + }, + { + "epoch": 0.3164539838571493, + "grad_norm": 2.613261178571089, + "learning_rate": 3.199309872892524e-06, + "loss": 0.9963, + "step": 3509 + }, + { + "epoch": 0.3165441673806196, + "grad_norm": 2.2499957614434902, + "learning_rate": 3.198842302106243e-06, + "loss": 1.0585, + "step": 3510 + }, + { + "epoch": 0.3166343509040898, + "grad_norm": 1.8405215559746106, + "learning_rate": 3.1983746290281265e-06, + "loss": 0.9878, + "step": 3511 + }, + { + "epoch": 0.3167245344275601, + "grad_norm": 1.3279881575154049, + "learning_rate": 3.197906853698079e-06, + "loss": 1.0134, + "step": 3512 + }, + { + "epoch": 0.31681471795103033, + "grad_norm": 1.5334586738413503, + "learning_rate": 3.1974389761560137e-06, + "loss": 0.9547, + "step": 3513 + }, + { + "epoch": 0.3169049014745006, + "grad_norm": 1.6851556529476008, + "learning_rate": 3.1969709964418525e-06, + "loss": 0.9569, + "step": 3514 + }, + { + "epoch": 0.31699508499797086, + "grad_norm": 1.760711923551785, + "learning_rate": 3.196502914595525e-06, + "loss": 1.0247, + "step": 3515 + }, + { + "epoch": 0.31708526852144114, + "grad_norm": 1.6315735857750793, + "learning_rate": 3.1960347306569723e-06, + "loss": 1.0233, + "step": 3516 + }, + { + "epoch": 0.3171754520449114, + "grad_norm": 1.7851993881121968, + "learning_rate": 3.195566444666141e-06, + "loss": 0.9522, + "step": 3517 + }, + { + "epoch": 0.31726563556838167, + "grad_norm": 1.5539565165019364, + "learning_rate": 3.1950980566629886e-06, + "loss": 1.1121, + "step": 3518 + }, + { + "epoch": 0.3173558190918519, + "grad_norm": 1.4844162784659356, + "learning_rate": 3.1946295666874797e-06, + "loss": 1.0189, + "step": 3519 + }, + { + "epoch": 0.3174460026153222, + "grad_norm": 1.4441616941096036, + "learning_rate": 3.19416097477959e-06, + "loss": 1.0239, + "step": 3520 + }, + { + "epoch": 0.3175361861387924, + "grad_norm": 1.5282679734698101, + "learning_rate": 3.1936922809793005e-06, + "loss": 1.0623, + "step": 3521 + }, + { + "epoch": 0.3176263696622627, + "grad_norm": 1.762402048838598, + "learning_rate": 3.193223485326604e-06, + "loss": 1.1256, + "step": 3522 + }, + { + "epoch": 0.31771655318573294, + "grad_norm": 1.9101051370533975, + "learning_rate": 3.1927545878615005e-06, + "loss": 1.0922, + "step": 3523 + }, + { + "epoch": 0.31780673670920323, + "grad_norm": 2.0437793181296753, + "learning_rate": 3.192285588623999e-06, + "loss": 0.9607, + "step": 3524 + }, + { + "epoch": 0.31789692023267346, + "grad_norm": 1.2107228242642984, + "learning_rate": 3.191816487654117e-06, + "loss": 0.9355, + "step": 3525 + }, + { + "epoch": 0.31798710375614375, + "grad_norm": 1.86969158219981, + "learning_rate": 3.19134728499188e-06, + "loss": 0.9689, + "step": 3526 + }, + { + "epoch": 0.31807728727961404, + "grad_norm": 1.9387165065211238, + "learning_rate": 3.1908779806773235e-06, + "loss": 0.8943, + "step": 3527 + }, + { + "epoch": 0.3181674708030843, + "grad_norm": 1.6996456618436604, + "learning_rate": 3.190408574750492e-06, + "loss": 1.0621, + "step": 3528 + }, + { + "epoch": 0.31825765432655456, + "grad_norm": 1.648508513308599, + "learning_rate": 3.1899390672514367e-06, + "loss": 1.0779, + "step": 3529 + }, + { + "epoch": 0.3183478378500248, + "grad_norm": 2.3459121903151425, + "learning_rate": 3.189469458220219e-06, + "loss": 0.9236, + "step": 3530 + }, + { + "epoch": 0.3184380213734951, + "grad_norm": 2.332214552737638, + "learning_rate": 3.1889997476969086e-06, + "loss": 1.006, + "step": 3531 + }, + { + "epoch": 0.3185282048969653, + "grad_norm": 2.2071991173023813, + "learning_rate": 3.188529935721583e-06, + "loss": 0.9994, + "step": 3532 + }, + { + "epoch": 0.3186183884204356, + "grad_norm": 1.4928549826496216, + "learning_rate": 3.18806002233433e-06, + "loss": 1.0656, + "step": 3533 + }, + { + "epoch": 0.31870857194390584, + "grad_norm": 1.4288632043827671, + "learning_rate": 3.187590007575245e-06, + "loss": 1.0689, + "step": 3534 + }, + { + "epoch": 0.3187987554673761, + "grad_norm": 1.9087062082719408, + "learning_rate": 3.1871198914844327e-06, + "loss": 0.9403, + "step": 3535 + }, + { + "epoch": 0.31888893899084636, + "grad_norm": 1.6555838774847904, + "learning_rate": 3.1866496741020057e-06, + "loss": 0.9404, + "step": 3536 + }, + { + "epoch": 0.31897912251431665, + "grad_norm": 1.2761352141467852, + "learning_rate": 3.186179355468085e-06, + "loss": 0.9719, + "step": 3537 + }, + { + "epoch": 0.3190693060377869, + "grad_norm": 1.6718753565137252, + "learning_rate": 3.1857089356228015e-06, + "loss": 1.0023, + "step": 3538 + }, + { + "epoch": 0.31915948956125717, + "grad_norm": 1.537364972395823, + "learning_rate": 3.1852384146062933e-06, + "loss": 0.9513, + "step": 3539 + }, + { + "epoch": 0.3192496730847274, + "grad_norm": 0.5886977981701811, + "learning_rate": 3.184767792458708e-06, + "loss": 0.7954, + "step": 3540 + }, + { + "epoch": 0.3193398566081977, + "grad_norm": 1.647115385112157, + "learning_rate": 3.1842970692202023e-06, + "loss": 1.1321, + "step": 3541 + }, + { + "epoch": 0.3194300401316679, + "grad_norm": 1.7951181862688985, + "learning_rate": 3.1838262449309403e-06, + "loss": 0.9855, + "step": 3542 + }, + { + "epoch": 0.3195202236551382, + "grad_norm": 1.9549844054358685, + "learning_rate": 3.1833553196310956e-06, + "loss": 0.9226, + "step": 3543 + }, + { + "epoch": 0.31961040717860845, + "grad_norm": 2.739294281035094, + "learning_rate": 3.18288429336085e-06, + "loss": 0.8671, + "step": 3544 + }, + { + "epoch": 0.31970059070207874, + "grad_norm": 1.9366935774508711, + "learning_rate": 3.182413166160394e-06, + "loss": 1.112, + "step": 3545 + }, + { + "epoch": 0.31979077422554897, + "grad_norm": 2.156850316611944, + "learning_rate": 3.1819419380699275e-06, + "loss": 1.0441, + "step": 3546 + }, + { + "epoch": 0.31988095774901926, + "grad_norm": 1.4049994948579687, + "learning_rate": 3.181470609129658e-06, + "loss": 1.0134, + "step": 3547 + }, + { + "epoch": 0.3199711412724895, + "grad_norm": 2.2821499865135078, + "learning_rate": 3.1809991793798e-06, + "loss": 1.0586, + "step": 3548 + }, + { + "epoch": 0.3200613247959598, + "grad_norm": 1.8528210832998908, + "learning_rate": 3.1805276488605806e-06, + "loss": 0.9621, + "step": 3549 + }, + { + "epoch": 0.32015150831943007, + "grad_norm": 2.079224145231115, + "learning_rate": 3.1800560176122336e-06, + "loss": 1.0001, + "step": 3550 + }, + { + "epoch": 0.3202416918429003, + "grad_norm": 1.5845417965702655, + "learning_rate": 3.179584285675e-06, + "loss": 1.0148, + "step": 3551 + }, + { + "epoch": 0.3203318753663706, + "grad_norm": 1.837388524417371, + "learning_rate": 3.1791124530891315e-06, + "loss": 0.9239, + "step": 3552 + }, + { + "epoch": 0.3204220588898408, + "grad_norm": 1.8252904713063407, + "learning_rate": 3.178640519894886e-06, + "loss": 1.0047, + "step": 3553 + }, + { + "epoch": 0.3205122424133111, + "grad_norm": 1.6729891620793413, + "learning_rate": 3.1781684861325324e-06, + "loss": 0.9643, + "step": 3554 + }, + { + "epoch": 0.32060242593678134, + "grad_norm": 1.76232730479346, + "learning_rate": 3.177696351842348e-06, + "loss": 1.0128, + "step": 3555 + }, + { + "epoch": 0.32069260946025163, + "grad_norm": 0.6993460725762525, + "learning_rate": 3.1772241170646167e-06, + "loss": 0.7979, + "step": 3556 + }, + { + "epoch": 0.32078279298372187, + "grad_norm": 1.8482101932688304, + "learning_rate": 3.1767517818396334e-06, + "loss": 1.109, + "step": 3557 + }, + { + "epoch": 0.32087297650719215, + "grad_norm": 1.6652888404338042, + "learning_rate": 3.1762793462076986e-06, + "loss": 1.1619, + "step": 3558 + }, + { + "epoch": 0.3209631600306624, + "grad_norm": 1.9736851190263764, + "learning_rate": 3.1758068102091236e-06, + "loss": 0.8897, + "step": 3559 + }, + { + "epoch": 0.3210533435541327, + "grad_norm": 1.9847917494605842, + "learning_rate": 3.175334173884229e-06, + "loss": 0.9723, + "step": 3560 + }, + { + "epoch": 0.3211435270776029, + "grad_norm": 1.4179280593119938, + "learning_rate": 3.174861437273342e-06, + "loss": 1.0014, + "step": 3561 + }, + { + "epoch": 0.3212337106010732, + "grad_norm": 1.8945880055528992, + "learning_rate": 3.174388600416799e-06, + "loss": 0.9641, + "step": 3562 + }, + { + "epoch": 0.32132389412454343, + "grad_norm": 0.666657571929249, + "learning_rate": 3.1739156633549445e-06, + "loss": 0.808, + "step": 3563 + }, + { + "epoch": 0.3214140776480137, + "grad_norm": 1.4196393389114148, + "learning_rate": 3.173442626128133e-06, + "loss": 1.0701, + "step": 3564 + }, + { + "epoch": 0.32150426117148395, + "grad_norm": 2.0986163394755413, + "learning_rate": 3.1729694887767265e-06, + "loss": 1.1114, + "step": 3565 + }, + { + "epoch": 0.32159444469495424, + "grad_norm": 2.836575691147925, + "learning_rate": 3.172496251341096e-06, + "loss": 1.0014, + "step": 3566 + }, + { + "epoch": 0.3216846282184245, + "grad_norm": 1.7401870395717471, + "learning_rate": 3.172022913861619e-06, + "loss": 0.8611, + "step": 3567 + }, + { + "epoch": 0.32177481174189476, + "grad_norm": 2.0205111164786906, + "learning_rate": 3.171549476378686e-06, + "loss": 0.9602, + "step": 3568 + }, + { + "epoch": 0.321864995265365, + "grad_norm": 2.2556073993472876, + "learning_rate": 3.1710759389326906e-06, + "loss": 0.971, + "step": 3569 + }, + { + "epoch": 0.3219551787888353, + "grad_norm": 2.1824784544135807, + "learning_rate": 3.1706023015640396e-06, + "loss": 0.9953, + "step": 3570 + }, + { + "epoch": 0.3220453623123055, + "grad_norm": 2.052075600649857, + "learning_rate": 3.1701285643131453e-06, + "loss": 1.0273, + "step": 3571 + }, + { + "epoch": 0.3221355458357758, + "grad_norm": 2.08807583478989, + "learning_rate": 3.16965472722043e-06, + "loss": 1.0093, + "step": 3572 + }, + { + "epoch": 0.32222572935924604, + "grad_norm": 2.3069307433911983, + "learning_rate": 3.169180790326324e-06, + "loss": 0.9293, + "step": 3573 + }, + { + "epoch": 0.3223159128827163, + "grad_norm": 1.408327560112009, + "learning_rate": 3.168706753671266e-06, + "loss": 1.0033, + "step": 3574 + }, + { + "epoch": 0.3224060964061866, + "grad_norm": 1.3208584982623786, + "learning_rate": 3.168232617295704e-06, + "loss": 0.9289, + "step": 3575 + }, + { + "epoch": 0.32249627992965685, + "grad_norm": 2.841644513742081, + "learning_rate": 3.167758381240093e-06, + "loss": 1.0004, + "step": 3576 + }, + { + "epoch": 0.32258646345312714, + "grad_norm": 1.4632299867399003, + "learning_rate": 3.1672840455448978e-06, + "loss": 1.0504, + "step": 3577 + }, + { + "epoch": 0.32267664697659737, + "grad_norm": 1.5650744781809645, + "learning_rate": 3.166809610250592e-06, + "loss": 0.9448, + "step": 3578 + }, + { + "epoch": 0.32276683050006766, + "grad_norm": 1.481005810274811, + "learning_rate": 3.166335075397656e-06, + "loss": 0.9286, + "step": 3579 + }, + { + "epoch": 0.3228570140235379, + "grad_norm": 2.2761845198319826, + "learning_rate": 3.1658604410265808e-06, + "loss": 0.9695, + "step": 3580 + }, + { + "epoch": 0.3229471975470082, + "grad_norm": 0.8480572784643174, + "learning_rate": 3.1653857071778644e-06, + "loss": 0.837, + "step": 3581 + }, + { + "epoch": 0.3230373810704784, + "grad_norm": 1.8510179322423617, + "learning_rate": 3.1649108738920133e-06, + "loss": 0.9926, + "step": 3582 + }, + { + "epoch": 0.3231275645939487, + "grad_norm": 1.5616738996606256, + "learning_rate": 3.1644359412095432e-06, + "loss": 1.0073, + "step": 3583 + }, + { + "epoch": 0.32321774811741893, + "grad_norm": 0.7651700205515108, + "learning_rate": 3.163960909170978e-06, + "loss": 0.8477, + "step": 3584 + }, + { + "epoch": 0.3233079316408892, + "grad_norm": 1.664312925882266, + "learning_rate": 3.1634857778168496e-06, + "loss": 0.9043, + "step": 3585 + }, + { + "epoch": 0.32339811516435946, + "grad_norm": 1.5960216414435158, + "learning_rate": 3.1630105471877002e-06, + "loss": 1.0391, + "step": 3586 + }, + { + "epoch": 0.32348829868782975, + "grad_norm": 2.119342340810009, + "learning_rate": 3.162535217324077e-06, + "loss": 0.9845, + "step": 3587 + }, + { + "epoch": 0.3235784822113, + "grad_norm": 3.2985186344144837, + "learning_rate": 3.1620597882665393e-06, + "loss": 1.0323, + "step": 3588 + }, + { + "epoch": 0.32366866573477027, + "grad_norm": 2.831166018152038, + "learning_rate": 3.1615842600556535e-06, + "loss": 1.0635, + "step": 3589 + }, + { + "epoch": 0.3237588492582405, + "grad_norm": 1.6479622395318934, + "learning_rate": 3.1611086327319932e-06, + "loss": 1.0003, + "step": 3590 + }, + { + "epoch": 0.3238490327817108, + "grad_norm": 1.587137304009122, + "learning_rate": 3.160632906336142e-06, + "loss": 1.0704, + "step": 3591 + }, + { + "epoch": 0.323939216305181, + "grad_norm": 1.667446224369269, + "learning_rate": 3.160157080908692e-06, + "loss": 1.0264, + "step": 3592 + }, + { + "epoch": 0.3240293998286513, + "grad_norm": 1.9053336270365255, + "learning_rate": 3.1596811564902426e-06, + "loss": 1.0598, + "step": 3593 + }, + { + "epoch": 0.32411958335212154, + "grad_norm": 2.7366499674105964, + "learning_rate": 3.1592051331214023e-06, + "loss": 0.9422, + "step": 3594 + }, + { + "epoch": 0.32420976687559183, + "grad_norm": 1.476112741263571, + "learning_rate": 3.158729010842789e-06, + "loss": 1.0264, + "step": 3595 + }, + { + "epoch": 0.32429995039906206, + "grad_norm": 1.4579567105024405, + "learning_rate": 3.1582527896950266e-06, + "loss": 1.055, + "step": 3596 + }, + { + "epoch": 0.32439013392253235, + "grad_norm": 1.5244469843748438, + "learning_rate": 3.157776469718749e-06, + "loss": 1.0164, + "step": 3597 + }, + { + "epoch": 0.32448031744600264, + "grad_norm": 1.9971370471352945, + "learning_rate": 3.1573000509546004e-06, + "loss": 0.992, + "step": 3598 + }, + { + "epoch": 0.3245705009694729, + "grad_norm": 2.047039229232812, + "learning_rate": 3.1568235334432296e-06, + "loss": 1.0581, + "step": 3599 + }, + { + "epoch": 0.32466068449294316, + "grad_norm": 1.9917017325768642, + "learning_rate": 3.1563469172252964e-06, + "loss": 1.0695, + "step": 3600 + }, + { + "epoch": 0.3247508680164134, + "grad_norm": 2.1729092776744974, + "learning_rate": 3.155870202341468e-06, + "loss": 1.0749, + "step": 3601 + }, + { + "epoch": 0.3248410515398837, + "grad_norm": 1.595716721512989, + "learning_rate": 3.155393388832421e-06, + "loss": 0.9985, + "step": 3602 + }, + { + "epoch": 0.3249312350633539, + "grad_norm": 1.7924830624513037, + "learning_rate": 3.1549164767388386e-06, + "loss": 1.0285, + "step": 3603 + }, + { + "epoch": 0.3250214185868242, + "grad_norm": 1.5656212720997909, + "learning_rate": 3.1544394661014145e-06, + "loss": 1.006, + "step": 3604 + }, + { + "epoch": 0.32511160211029444, + "grad_norm": 1.598592842734904, + "learning_rate": 3.15396235696085e-06, + "loss": 0.9217, + "step": 3605 + }, + { + "epoch": 0.32520178563376473, + "grad_norm": 1.518422484699504, + "learning_rate": 3.153485149357854e-06, + "loss": 0.9939, + "step": 3606 + }, + { + "epoch": 0.32529196915723496, + "grad_norm": 1.4681239315595835, + "learning_rate": 3.153007843333145e-06, + "loss": 0.9893, + "step": 3607 + }, + { + "epoch": 0.32538215268070525, + "grad_norm": 1.9338202449810793, + "learning_rate": 3.152530438927449e-06, + "loss": 1.0352, + "step": 3608 + }, + { + "epoch": 0.3254723362041755, + "grad_norm": 1.68229905196735, + "learning_rate": 3.1520529361815008e-06, + "loss": 1.0471, + "step": 3609 + }, + { + "epoch": 0.32556251972764577, + "grad_norm": 1.5069034187994523, + "learning_rate": 3.151575335136044e-06, + "loss": 1.046, + "step": 3610 + }, + { + "epoch": 0.325652703251116, + "grad_norm": 2.6310505843539382, + "learning_rate": 3.1510976358318298e-06, + "loss": 0.9873, + "step": 3611 + }, + { + "epoch": 0.3257428867745863, + "grad_norm": 1.5535580160003615, + "learning_rate": 3.1506198383096186e-06, + "loss": 0.9554, + "step": 3612 + }, + { + "epoch": 0.3258330702980565, + "grad_norm": 1.5111018062909694, + "learning_rate": 3.150141942610178e-06, + "loss": 1.0209, + "step": 3613 + }, + { + "epoch": 0.3259232538215268, + "grad_norm": 0.7046870576303113, + "learning_rate": 3.1496639487742853e-06, + "loss": 0.8231, + "step": 3614 + }, + { + "epoch": 0.32601343734499705, + "grad_norm": 1.6479070451779492, + "learning_rate": 3.1491858568427247e-06, + "loss": 1.0535, + "step": 3615 + }, + { + "epoch": 0.32610362086846734, + "grad_norm": 2.1030399843805445, + "learning_rate": 3.1487076668562903e-06, + "loss": 1.0558, + "step": 3616 + }, + { + "epoch": 0.32619380439193757, + "grad_norm": 2.0115797272579057, + "learning_rate": 3.1482293788557847e-06, + "loss": 1.0503, + "step": 3617 + }, + { + "epoch": 0.32628398791540786, + "grad_norm": 2.569295200675966, + "learning_rate": 3.1477509928820165e-06, + "loss": 0.9734, + "step": 3618 + }, + { + "epoch": 0.3263741714388781, + "grad_norm": 1.9324835778805667, + "learning_rate": 3.147272508975805e-06, + "loss": 1.0056, + "step": 3619 + }, + { + "epoch": 0.3264643549623484, + "grad_norm": 0.6745637057627697, + "learning_rate": 3.1467939271779775e-06, + "loss": 0.8434, + "step": 3620 + }, + { + "epoch": 0.32655453848581867, + "grad_norm": 1.2785184166140897, + "learning_rate": 3.146315247529368e-06, + "loss": 0.9718, + "step": 3621 + }, + { + "epoch": 0.3266447220092889, + "grad_norm": 1.861312625976591, + "learning_rate": 3.1458364700708212e-06, + "loss": 0.9653, + "step": 3622 + }, + { + "epoch": 0.3267349055327592, + "grad_norm": 2.2503623670602595, + "learning_rate": 3.1453575948431892e-06, + "loss": 1.0321, + "step": 3623 + }, + { + "epoch": 0.3268250890562294, + "grad_norm": 1.7119355099733218, + "learning_rate": 3.144878621887331e-06, + "loss": 1.0116, + "step": 3624 + }, + { + "epoch": 0.3269152725796997, + "grad_norm": 1.7230546624650447, + "learning_rate": 3.1443995512441167e-06, + "loss": 1.0478, + "step": 3625 + }, + { + "epoch": 0.32700545610316994, + "grad_norm": 1.5654973082712693, + "learning_rate": 3.1439203829544224e-06, + "loss": 1.0734, + "step": 3626 + }, + { + "epoch": 0.32709563962664023, + "grad_norm": 1.2607334884237065, + "learning_rate": 3.143441117059133e-06, + "loss": 1.0853, + "step": 3627 + }, + { + "epoch": 0.32718582315011047, + "grad_norm": 1.8986016347754526, + "learning_rate": 3.142961753599143e-06, + "loss": 0.9555, + "step": 3628 + }, + { + "epoch": 0.32727600667358075, + "grad_norm": 1.8700361030071835, + "learning_rate": 3.1424822926153543e-06, + "loss": 0.974, + "step": 3629 + }, + { + "epoch": 0.327366190197051, + "grad_norm": 1.691514643836618, + "learning_rate": 3.142002734148676e-06, + "loss": 1.0529, + "step": 3630 + }, + { + "epoch": 0.3274563737205213, + "grad_norm": 1.8378920519637791, + "learning_rate": 3.141523078240028e-06, + "loss": 1.0346, + "step": 3631 + }, + { + "epoch": 0.3275465572439915, + "grad_norm": 2.0139212097848294, + "learning_rate": 3.1410433249303366e-06, + "loss": 1.0593, + "step": 3632 + }, + { + "epoch": 0.3276367407674618, + "grad_norm": 2.282341682818089, + "learning_rate": 3.1405634742605366e-06, + "loss": 1.0948, + "step": 3633 + }, + { + "epoch": 0.32772692429093203, + "grad_norm": 2.483014099886095, + "learning_rate": 3.1400835262715727e-06, + "loss": 1.0419, + "step": 3634 + }, + { + "epoch": 0.3278171078144023, + "grad_norm": 1.6014094581985012, + "learning_rate": 3.139603481004396e-06, + "loss": 0.9953, + "step": 3635 + }, + { + "epoch": 0.32790729133787255, + "grad_norm": 1.619304211569943, + "learning_rate": 3.139123338499966e-06, + "loss": 1.0243, + "step": 3636 + }, + { + "epoch": 0.32799747486134284, + "grad_norm": 1.8131147854716811, + "learning_rate": 3.1386430987992524e-06, + "loss": 1.0428, + "step": 3637 + }, + { + "epoch": 0.3280876583848131, + "grad_norm": 1.813570167044401, + "learning_rate": 3.1381627619432307e-06, + "loss": 0.9523, + "step": 3638 + }, + { + "epoch": 0.32817784190828336, + "grad_norm": 2.0158012372232346, + "learning_rate": 3.1376823279728864e-06, + "loss": 1.0162, + "step": 3639 + }, + { + "epoch": 0.3282680254317536, + "grad_norm": 1.7656329669603856, + "learning_rate": 3.1372017969292125e-06, + "loss": 1.0942, + "step": 3640 + }, + { + "epoch": 0.3283582089552239, + "grad_norm": 1.5244117164960327, + "learning_rate": 3.136721168853211e-06, + "loss": 0.9924, + "step": 3641 + }, + { + "epoch": 0.3284483924786941, + "grad_norm": 1.7641657647857498, + "learning_rate": 3.1362404437858924e-06, + "loss": 0.9714, + "step": 3642 + }, + { + "epoch": 0.3285385760021644, + "grad_norm": 1.72219489530534, + "learning_rate": 3.135759621768273e-06, + "loss": 1.065, + "step": 3643 + }, + { + "epoch": 0.32862875952563464, + "grad_norm": 1.9048917340128668, + "learning_rate": 3.13527870284138e-06, + "loss": 1.0575, + "step": 3644 + }, + { + "epoch": 0.32871894304910493, + "grad_norm": 3.340449049681139, + "learning_rate": 3.134797687046249e-06, + "loss": 1.0659, + "step": 3645 + }, + { + "epoch": 0.3288091265725752, + "grad_norm": 2.4306774024091555, + "learning_rate": 3.1343165744239218e-06, + "loss": 0.9446, + "step": 3646 + }, + { + "epoch": 0.32889931009604545, + "grad_norm": 1.8634299292915188, + "learning_rate": 3.13383536501545e-06, + "loss": 1.0283, + "step": 3647 + }, + { + "epoch": 0.32898949361951574, + "grad_norm": 1.644261208877534, + "learning_rate": 3.133354058861893e-06, + "loss": 1.0313, + "step": 3648 + }, + { + "epoch": 0.32907967714298597, + "grad_norm": 2.5480165322155197, + "learning_rate": 3.132872656004318e-06, + "loss": 0.9784, + "step": 3649 + }, + { + "epoch": 0.32916986066645626, + "grad_norm": 1.5437047820476626, + "learning_rate": 3.132391156483802e-06, + "loss": 0.9114, + "step": 3650 + }, + { + "epoch": 0.3292600441899265, + "grad_norm": 1.8037481464209917, + "learning_rate": 3.131909560341428e-06, + "loss": 0.9665, + "step": 3651 + }, + { + "epoch": 0.3293502277133968, + "grad_norm": 1.7627967530541135, + "learning_rate": 3.1314278676182893e-06, + "loss": 0.9603, + "step": 3652 + }, + { + "epoch": 0.329440411236867, + "grad_norm": 1.4568301310067453, + "learning_rate": 3.130946078355486e-06, + "loss": 1.0792, + "step": 3653 + }, + { + "epoch": 0.3295305947603373, + "grad_norm": 1.881589500234788, + "learning_rate": 3.130464192594128e-06, + "loss": 1.0471, + "step": 3654 + }, + { + "epoch": 0.32962077828380754, + "grad_norm": 1.4494253664201775, + "learning_rate": 3.1299822103753315e-06, + "loss": 0.9359, + "step": 3655 + }, + { + "epoch": 0.3297109618072778, + "grad_norm": 2.1393461096526636, + "learning_rate": 3.1295001317402217e-06, + "loss": 1.0108, + "step": 3656 + }, + { + "epoch": 0.32980114533074806, + "grad_norm": 2.1625950654583432, + "learning_rate": 3.1290179567299335e-06, + "loss": 0.955, + "step": 3657 + }, + { + "epoch": 0.32989132885421835, + "grad_norm": 1.4087989171089503, + "learning_rate": 3.128535685385607e-06, + "loss": 1.0417, + "step": 3658 + }, + { + "epoch": 0.3299815123776886, + "grad_norm": 1.7105511094045485, + "learning_rate": 3.1280533177483935e-06, + "loss": 1.0914, + "step": 3659 + }, + { + "epoch": 0.33007169590115887, + "grad_norm": 1.6697531096190597, + "learning_rate": 3.127570853859451e-06, + "loss": 1.0235, + "step": 3660 + }, + { + "epoch": 0.3301618794246291, + "grad_norm": 1.7861101869679947, + "learning_rate": 3.1270882937599456e-06, + "loss": 0.8952, + "step": 3661 + }, + { + "epoch": 0.3302520629480994, + "grad_norm": 2.9601089808016625, + "learning_rate": 3.1266056374910532e-06, + "loss": 1.03, + "step": 3662 + }, + { + "epoch": 0.3303422464715696, + "grad_norm": 1.7505569253015778, + "learning_rate": 3.126122885093955e-06, + "loss": 0.9047, + "step": 3663 + }, + { + "epoch": 0.3304324299950399, + "grad_norm": 2.1125528769110993, + "learning_rate": 3.1256400366098427e-06, + "loss": 1.0035, + "step": 3664 + }, + { + "epoch": 0.33052261351851014, + "grad_norm": 2.0013809205119495, + "learning_rate": 3.125157092079916e-06, + "loss": 0.9627, + "step": 3665 + }, + { + "epoch": 0.33061279704198043, + "grad_norm": 2.890907232579484, + "learning_rate": 3.1246740515453824e-06, + "loss": 0.9767, + "step": 3666 + }, + { + "epoch": 0.33070298056545067, + "grad_norm": 1.4264980238918519, + "learning_rate": 3.124190915047457e-06, + "loss": 0.9486, + "step": 3667 + }, + { + "epoch": 0.33079316408892095, + "grad_norm": 2.0081465031422074, + "learning_rate": 3.123707682627364e-06, + "loss": 0.9837, + "step": 3668 + }, + { + "epoch": 0.33088334761239124, + "grad_norm": 5.150252304563261, + "learning_rate": 3.1232243543263356e-06, + "loss": 0.9915, + "step": 3669 + }, + { + "epoch": 0.3309735311358615, + "grad_norm": 1.869809978540659, + "learning_rate": 3.1227409301856122e-06, + "loss": 0.9103, + "step": 3670 + }, + { + "epoch": 0.33106371465933176, + "grad_norm": 1.5482573834623095, + "learning_rate": 3.1222574102464413e-06, + "loss": 0.9911, + "step": 3671 + }, + { + "epoch": 0.331153898182802, + "grad_norm": 1.5202586337952995, + "learning_rate": 3.12177379455008e-06, + "loss": 1.0417, + "step": 3672 + }, + { + "epoch": 0.3312440817062723, + "grad_norm": 1.5711498400341952, + "learning_rate": 3.121290083137794e-06, + "loss": 1.0013, + "step": 3673 + }, + { + "epoch": 0.3313342652297425, + "grad_norm": 1.7488838451064486, + "learning_rate": 3.1208062760508547e-06, + "loss": 0.9767, + "step": 3674 + }, + { + "epoch": 0.3314244487532128, + "grad_norm": 1.580939323452895, + "learning_rate": 3.1203223733305438e-06, + "loss": 0.9428, + "step": 3675 + }, + { + "epoch": 0.33151463227668304, + "grad_norm": 0.651321478159234, + "learning_rate": 3.1198383750181512e-06, + "loss": 0.8016, + "step": 3676 + }, + { + "epoch": 0.33160481580015333, + "grad_norm": 1.5525286076316431, + "learning_rate": 3.1193542811549734e-06, + "loss": 0.93, + "step": 3677 + }, + { + "epoch": 0.33169499932362356, + "grad_norm": 1.4978019185829583, + "learning_rate": 3.1188700917823166e-06, + "loss": 0.9353, + "step": 3678 + }, + { + "epoch": 0.33178518284709385, + "grad_norm": 0.7005450068466242, + "learning_rate": 3.1183858069414937e-06, + "loss": 0.8646, + "step": 3679 + }, + { + "epoch": 0.3318753663705641, + "grad_norm": 3.1748182049466154, + "learning_rate": 3.117901426673827e-06, + "loss": 1.054, + "step": 3680 + }, + { + "epoch": 0.3319655498940344, + "grad_norm": 1.9685474473207156, + "learning_rate": 3.1174169510206466e-06, + "loss": 0.9669, + "step": 3681 + }, + { + "epoch": 0.3320557334175046, + "grad_norm": 1.4426204870404782, + "learning_rate": 3.1169323800232908e-06, + "loss": 1.0035, + "step": 3682 + }, + { + "epoch": 0.3321459169409749, + "grad_norm": 5.927127491585926, + "learning_rate": 3.1164477137231054e-06, + "loss": 1.033, + "step": 3683 + }, + { + "epoch": 0.3322361004644451, + "grad_norm": 2.0501203362106177, + "learning_rate": 3.115962952161445e-06, + "loss": 1.0092, + "step": 3684 + }, + { + "epoch": 0.3323262839879154, + "grad_norm": 3.0399413419384853, + "learning_rate": 3.1154780953796727e-06, + "loss": 1.0402, + "step": 3685 + }, + { + "epoch": 0.33241646751138565, + "grad_norm": 1.7510458001203917, + "learning_rate": 3.114993143419158e-06, + "loss": 0.9104, + "step": 3686 + }, + { + "epoch": 0.33250665103485594, + "grad_norm": 1.3145469416655862, + "learning_rate": 3.1145080963212806e-06, + "loss": 0.9165, + "step": 3687 + }, + { + "epoch": 0.33259683455832617, + "grad_norm": 1.5508258342339547, + "learning_rate": 3.114022954127427e-06, + "loss": 1.0092, + "step": 3688 + }, + { + "epoch": 0.33268701808179646, + "grad_norm": 1.7789783381086226, + "learning_rate": 3.1135377168789923e-06, + "loss": 1.211, + "step": 3689 + }, + { + "epoch": 0.3327772016052667, + "grad_norm": 1.724420176374334, + "learning_rate": 3.1130523846173803e-06, + "loss": 1.0736, + "step": 3690 + }, + { + "epoch": 0.332867385128737, + "grad_norm": 2.1951032111161406, + "learning_rate": 3.1125669573840006e-06, + "loss": 0.9484, + "step": 3691 + }, + { + "epoch": 0.3329575686522072, + "grad_norm": 1.9738465598744246, + "learning_rate": 3.112081435220274e-06, + "loss": 0.9499, + "step": 3692 + }, + { + "epoch": 0.3330477521756775, + "grad_norm": 2.7999531946357386, + "learning_rate": 3.111595818167627e-06, + "loss": 0.9917, + "step": 3693 + }, + { + "epoch": 0.3331379356991478, + "grad_norm": 1.7336157434210737, + "learning_rate": 3.1111101062674953e-06, + "loss": 1.063, + "step": 3694 + }, + { + "epoch": 0.333228119222618, + "grad_norm": 1.4788702636467501, + "learning_rate": 3.1106242995613233e-06, + "loss": 0.9919, + "step": 3695 + }, + { + "epoch": 0.3333183027460883, + "grad_norm": 2.221335395712809, + "learning_rate": 3.1101383980905616e-06, + "loss": 1.0052, + "step": 3696 + }, + { + "epoch": 0.33340848626955855, + "grad_norm": 0.6721034659578542, + "learning_rate": 3.109652401896671e-06, + "loss": 0.7894, + "step": 3697 + }, + { + "epoch": 0.33349866979302883, + "grad_norm": 1.4674110802870457, + "learning_rate": 3.109166311021119e-06, + "loss": 1.0498, + "step": 3698 + }, + { + "epoch": 0.33358885331649907, + "grad_norm": 1.9648951377746757, + "learning_rate": 3.1086801255053807e-06, + "loss": 1.0396, + "step": 3699 + }, + { + "epoch": 0.33367903683996936, + "grad_norm": 1.595032194882916, + "learning_rate": 3.108193845390942e-06, + "loss": 1.045, + "step": 3700 + }, + { + "epoch": 0.3337692203634396, + "grad_norm": 1.207990148194668, + "learning_rate": 3.1077074707192933e-06, + "loss": 0.9808, + "step": 3701 + }, + { + "epoch": 0.3338594038869099, + "grad_norm": 1.7021983714392959, + "learning_rate": 3.1072210015319353e-06, + "loss": 1.0229, + "step": 3702 + }, + { + "epoch": 0.3339495874103801, + "grad_norm": 1.4688408194525417, + "learning_rate": 3.106734437870376e-06, + "loss": 0.9604, + "step": 3703 + }, + { + "epoch": 0.3340397709338504, + "grad_norm": 2.0983964974266893, + "learning_rate": 3.1062477797761327e-06, + "loss": 0.9514, + "step": 3704 + }, + { + "epoch": 0.33412995445732063, + "grad_norm": 3.935070317571954, + "learning_rate": 3.105761027290729e-06, + "loss": 1.0517, + "step": 3705 + }, + { + "epoch": 0.3342201379807909, + "grad_norm": 1.4318074835615682, + "learning_rate": 3.105274180455697e-06, + "loss": 0.9731, + "step": 3706 + }, + { + "epoch": 0.33431032150426115, + "grad_norm": 1.5625737745034605, + "learning_rate": 3.1047872393125775e-06, + "loss": 1.0271, + "step": 3707 + }, + { + "epoch": 0.33440050502773144, + "grad_norm": 0.6261192551918768, + "learning_rate": 3.1043002039029186e-06, + "loss": 0.8167, + "step": 3708 + }, + { + "epoch": 0.3344906885512017, + "grad_norm": 1.497614712405278, + "learning_rate": 3.1038130742682782e-06, + "loss": 0.9971, + "step": 3709 + }, + { + "epoch": 0.33458087207467196, + "grad_norm": 1.4407232884041266, + "learning_rate": 3.103325850450219e-06, + "loss": 0.9834, + "step": 3710 + }, + { + "epoch": 0.3346710555981422, + "grad_norm": 1.5137296800958473, + "learning_rate": 3.1028385324903154e-06, + "loss": 1.0403, + "step": 3711 + }, + { + "epoch": 0.3347612391216125, + "grad_norm": 1.630900162055461, + "learning_rate": 3.1023511204301465e-06, + "loss": 1.0121, + "step": 3712 + }, + { + "epoch": 0.3348514226450827, + "grad_norm": 1.4529637893344298, + "learning_rate": 3.1018636143113022e-06, + "loss": 0.9666, + "step": 3713 + }, + { + "epoch": 0.334941606168553, + "grad_norm": 1.568449290965273, + "learning_rate": 3.1013760141753787e-06, + "loss": 0.9099, + "step": 3714 + }, + { + "epoch": 0.33503178969202324, + "grad_norm": 1.6985813392295948, + "learning_rate": 3.100888320063981e-06, + "loss": 0.9741, + "step": 3715 + }, + { + "epoch": 0.33512197321549353, + "grad_norm": 1.374023220601773, + "learning_rate": 3.100400532018721e-06, + "loss": 1.0077, + "step": 3716 + }, + { + "epoch": 0.3352121567389638, + "grad_norm": 2.0000238417158034, + "learning_rate": 3.0999126500812204e-06, + "loss": 1.0689, + "step": 3717 + }, + { + "epoch": 0.33530234026243405, + "grad_norm": 0.7269905634140379, + "learning_rate": 3.0994246742931076e-06, + "loss": 0.8712, + "step": 3718 + }, + { + "epoch": 0.33539252378590434, + "grad_norm": 1.7046657907028444, + "learning_rate": 3.098936604696019e-06, + "loss": 0.9484, + "step": 3719 + }, + { + "epoch": 0.3354827073093746, + "grad_norm": 1.727915652185637, + "learning_rate": 3.0984484413316e-06, + "loss": 1.0306, + "step": 3720 + }, + { + "epoch": 0.33557289083284486, + "grad_norm": 3.8437182417387206, + "learning_rate": 3.0979601842415033e-06, + "loss": 0.9797, + "step": 3721 + }, + { + "epoch": 0.3356630743563151, + "grad_norm": 1.4236362155261837, + "learning_rate": 3.0974718334673896e-06, + "loss": 0.9225, + "step": 3722 + }, + { + "epoch": 0.3357532578797854, + "grad_norm": 1.66966093824372, + "learning_rate": 3.0969833890509282e-06, + "loss": 0.9971, + "step": 3723 + }, + { + "epoch": 0.3358434414032556, + "grad_norm": 1.6606435980453493, + "learning_rate": 3.096494851033795e-06, + "loss": 0.9971, + "step": 3724 + }, + { + "epoch": 0.3359336249267259, + "grad_norm": 1.8657011875973004, + "learning_rate": 3.0960062194576747e-06, + "loss": 0.9279, + "step": 3725 + }, + { + "epoch": 0.33602380845019614, + "grad_norm": 2.5842203852796346, + "learning_rate": 3.0955174943642606e-06, + "loss": 0.9552, + "step": 3726 + }, + { + "epoch": 0.3361139919736664, + "grad_norm": 1.709582903002949, + "learning_rate": 3.0950286757952534e-06, + "loss": 0.9227, + "step": 3727 + }, + { + "epoch": 0.33620417549713666, + "grad_norm": 1.6685229772095063, + "learning_rate": 3.0945397637923617e-06, + "loss": 1.0662, + "step": 3728 + }, + { + "epoch": 0.33629435902060695, + "grad_norm": 1.7496223723466844, + "learning_rate": 3.0940507583973025e-06, + "loss": 1.0117, + "step": 3729 + }, + { + "epoch": 0.3363845425440772, + "grad_norm": 1.3722646120905428, + "learning_rate": 3.093561659651799e-06, + "loss": 0.9743, + "step": 3730 + }, + { + "epoch": 0.33647472606754747, + "grad_norm": 2.334987383365241, + "learning_rate": 3.093072467597586e-06, + "loss": 0.9396, + "step": 3731 + }, + { + "epoch": 0.3365649095910177, + "grad_norm": 0.6415005957183688, + "learning_rate": 3.092583182276402e-06, + "loss": 0.8223, + "step": 3732 + }, + { + "epoch": 0.336655093114488, + "grad_norm": 0.6251891803530102, + "learning_rate": 3.092093803729997e-06, + "loss": 0.8273, + "step": 3733 + }, + { + "epoch": 0.3367452766379582, + "grad_norm": 15.707940998119092, + "learning_rate": 3.0916043320001264e-06, + "loss": 1.0135, + "step": 3734 + }, + { + "epoch": 0.3368354601614285, + "grad_norm": 3.5992385747788753, + "learning_rate": 3.0911147671285557e-06, + "loss": 1.0116, + "step": 3735 + }, + { + "epoch": 0.33692564368489875, + "grad_norm": 1.6931036697795838, + "learning_rate": 3.0906251091570565e-06, + "loss": 1.0295, + "step": 3736 + }, + { + "epoch": 0.33701582720836903, + "grad_norm": 4.599607923632998, + "learning_rate": 3.0901353581274094e-06, + "loss": 1.0162, + "step": 3737 + }, + { + "epoch": 0.33710601073183927, + "grad_norm": 1.9231194289718434, + "learning_rate": 3.089645514081402e-06, + "loss": 0.8837, + "step": 3738 + }, + { + "epoch": 0.33719619425530956, + "grad_norm": 6.994617572662066, + "learning_rate": 3.0891555770608323e-06, + "loss": 0.9274, + "step": 3739 + }, + { + "epoch": 0.33728637777877984, + "grad_norm": 3.7072012955763345, + "learning_rate": 3.088665547107503e-06, + "loss": 1.0307, + "step": 3740 + }, + { + "epoch": 0.3373765613022501, + "grad_norm": 1.4292928015355575, + "learning_rate": 3.0881754242632254e-06, + "loss": 1.0812, + "step": 3741 + }, + { + "epoch": 0.33746674482572037, + "grad_norm": 1.3275097487786736, + "learning_rate": 3.0876852085698213e-06, + "loss": 1.047, + "step": 3742 + }, + { + "epoch": 0.3375569283491906, + "grad_norm": 2.657207989443773, + "learning_rate": 3.087194900069117e-06, + "loss": 1.0291, + "step": 3743 + }, + { + "epoch": 0.3376471118726609, + "grad_norm": 1.608976963041885, + "learning_rate": 3.08670449880295e-06, + "loss": 1.0622, + "step": 3744 + }, + { + "epoch": 0.3377372953961311, + "grad_norm": 1.7899694684157934, + "learning_rate": 3.086214004813163e-06, + "loss": 1.0226, + "step": 3745 + }, + { + "epoch": 0.3378274789196014, + "grad_norm": 1.4883253836908943, + "learning_rate": 3.0857234181416074e-06, + "loss": 1.0648, + "step": 3746 + }, + { + "epoch": 0.33791766244307164, + "grad_norm": 1.780372872141685, + "learning_rate": 3.085232738830143e-06, + "loss": 1.0555, + "step": 3747 + }, + { + "epoch": 0.33800784596654193, + "grad_norm": 1.6200150543266927, + "learning_rate": 3.084741966920638e-06, + "loss": 1.0659, + "step": 3748 + }, + { + "epoch": 0.33809802949001216, + "grad_norm": 1.6164149406276915, + "learning_rate": 3.084251102454966e-06, + "loss": 0.9651, + "step": 3749 + }, + { + "epoch": 0.33818821301348245, + "grad_norm": 0.7292988838940471, + "learning_rate": 3.083760145475013e-06, + "loss": 0.8511, + "step": 3750 + }, + { + "epoch": 0.3382783965369527, + "grad_norm": 1.5732580239785263, + "learning_rate": 3.0832690960226678e-06, + "loss": 0.946, + "step": 3751 + }, + { + "epoch": 0.338368580060423, + "grad_norm": 1.9582481636302709, + "learning_rate": 3.08277795413983e-06, + "loss": 0.9668, + "step": 3752 + }, + { + "epoch": 0.3384587635838932, + "grad_norm": 1.474228240014417, + "learning_rate": 3.0822867198684073e-06, + "loss": 1.055, + "step": 3753 + }, + { + "epoch": 0.3385489471073635, + "grad_norm": 1.7429648405721159, + "learning_rate": 3.081795393250314e-06, + "loss": 1.0123, + "step": 3754 + }, + { + "epoch": 0.33863913063083373, + "grad_norm": 1.8515277971967135, + "learning_rate": 3.081303974327473e-06, + "loss": 0.9716, + "step": 3755 + }, + { + "epoch": 0.338729314154304, + "grad_norm": 1.4504621163976452, + "learning_rate": 3.080812463141814e-06, + "loss": 1.0014, + "step": 3756 + }, + { + "epoch": 0.33881949767777425, + "grad_norm": 1.3845267150274254, + "learning_rate": 3.080320859735276e-06, + "loss": 0.8777, + "step": 3757 + }, + { + "epoch": 0.33890968120124454, + "grad_norm": 1.76526287044884, + "learning_rate": 3.079829164149806e-06, + "loss": 1.0563, + "step": 3758 + }, + { + "epoch": 0.3389998647247148, + "grad_norm": 2.4113622087362137, + "learning_rate": 3.0793373764273573e-06, + "loss": 0.9959, + "step": 3759 + }, + { + "epoch": 0.33909004824818506, + "grad_norm": 1.9856459985186448, + "learning_rate": 3.078845496609892e-06, + "loss": 1.0283, + "step": 3760 + }, + { + "epoch": 0.3391802317716553, + "grad_norm": 1.7110024357147984, + "learning_rate": 3.078353524739381e-06, + "loss": 1.094, + "step": 3761 + }, + { + "epoch": 0.3392704152951256, + "grad_norm": 1.78513087415293, + "learning_rate": 3.077861460857801e-06, + "loss": 1.0887, + "step": 3762 + }, + { + "epoch": 0.3393605988185958, + "grad_norm": 1.571854323143081, + "learning_rate": 3.077369305007138e-06, + "loss": 1.0252, + "step": 3763 + }, + { + "epoch": 0.3394507823420661, + "grad_norm": 1.537562457421025, + "learning_rate": 3.0768770572293852e-06, + "loss": 1.0458, + "step": 3764 + }, + { + "epoch": 0.3395409658655364, + "grad_norm": 2.0342548153208355, + "learning_rate": 3.0763847175665437e-06, + "loss": 1.0835, + "step": 3765 + }, + { + "epoch": 0.3396311493890066, + "grad_norm": 1.8877389491327004, + "learning_rate": 3.0758922860606237e-06, + "loss": 0.9447, + "step": 3766 + }, + { + "epoch": 0.3397213329124769, + "grad_norm": 1.5767411311323403, + "learning_rate": 3.0753997627536404e-06, + "loss": 0.9402, + "step": 3767 + }, + { + "epoch": 0.33981151643594715, + "grad_norm": 1.873485462125189, + "learning_rate": 3.0749071476876203e-06, + "loss": 0.9546, + "step": 3768 + }, + { + "epoch": 0.33990169995941744, + "grad_norm": 1.4510864396957026, + "learning_rate": 3.0744144409045952e-06, + "loss": 0.9728, + "step": 3769 + }, + { + "epoch": 0.33999188348288767, + "grad_norm": 4.317138609745677, + "learning_rate": 3.0739216424466056e-06, + "loss": 1.0275, + "step": 3770 + }, + { + "epoch": 0.34008206700635796, + "grad_norm": 2.031322184527185, + "learning_rate": 3.0734287523557002e-06, + "loss": 0.9922, + "step": 3771 + }, + { + "epoch": 0.3401722505298282, + "grad_norm": 1.8136496350882083, + "learning_rate": 3.0729357706739348e-06, + "loss": 0.9624, + "step": 3772 + }, + { + "epoch": 0.3402624340532985, + "grad_norm": 1.5628569386004645, + "learning_rate": 3.0724426974433737e-06, + "loss": 0.9899, + "step": 3773 + }, + { + "epoch": 0.3403526175767687, + "grad_norm": 1.3262133584816853, + "learning_rate": 3.0719495327060874e-06, + "loss": 0.9803, + "step": 3774 + }, + { + "epoch": 0.340442801100239, + "grad_norm": 1.9351517998740844, + "learning_rate": 3.071456276504157e-06, + "loss": 1.0136, + "step": 3775 + }, + { + "epoch": 0.34053298462370923, + "grad_norm": 1.6657386182785217, + "learning_rate": 3.070962928879669e-06, + "loss": 1.0043, + "step": 3776 + }, + { + "epoch": 0.3406231681471795, + "grad_norm": 2.1630728211015984, + "learning_rate": 3.0704694898747185e-06, + "loss": 1.0189, + "step": 3777 + }, + { + "epoch": 0.34071335167064976, + "grad_norm": 1.620840103344479, + "learning_rate": 3.069975959531408e-06, + "loss": 1.1334, + "step": 3778 + }, + { + "epoch": 0.34080353519412004, + "grad_norm": 0.6777872949876147, + "learning_rate": 3.06948233789185e-06, + "loss": 0.8854, + "step": 3779 + }, + { + "epoch": 0.3408937187175903, + "grad_norm": 1.575457821261133, + "learning_rate": 3.0689886249981614e-06, + "loss": 1.0215, + "step": 3780 + }, + { + "epoch": 0.34098390224106057, + "grad_norm": 2.492507196567136, + "learning_rate": 3.0684948208924693e-06, + "loss": 1.038, + "step": 3781 + }, + { + "epoch": 0.3410740857645308, + "grad_norm": 1.6772702588779569, + "learning_rate": 3.068000925616907e-06, + "loss": 1.0277, + "step": 3782 + }, + { + "epoch": 0.3411642692880011, + "grad_norm": 1.9157658685412544, + "learning_rate": 3.067506939213617e-06, + "loss": 1.002, + "step": 3783 + }, + { + "epoch": 0.3412544528114713, + "grad_norm": 1.3610248419395536, + "learning_rate": 3.0670128617247493e-06, + "loss": 1.0126, + "step": 3784 + }, + { + "epoch": 0.3413446363349416, + "grad_norm": 1.6968419783103197, + "learning_rate": 3.06651869319246e-06, + "loss": 0.9154, + "step": 3785 + }, + { + "epoch": 0.34143481985841184, + "grad_norm": 1.6860608922793254, + "learning_rate": 3.0660244336589154e-06, + "loss": 1.0033, + "step": 3786 + }, + { + "epoch": 0.34152500338188213, + "grad_norm": 1.5154683926872274, + "learning_rate": 3.065530083166288e-06, + "loss": 1.022, + "step": 3787 + }, + { + "epoch": 0.3416151869053524, + "grad_norm": 1.4037855806537607, + "learning_rate": 3.0650356417567586e-06, + "loss": 0.9656, + "step": 3788 + }, + { + "epoch": 0.34170537042882265, + "grad_norm": 1.7417317294633619, + "learning_rate": 3.0645411094725156e-06, + "loss": 1.0559, + "step": 3789 + }, + { + "epoch": 0.34179555395229294, + "grad_norm": 1.5632208877798073, + "learning_rate": 3.0640464863557556e-06, + "loss": 1.0039, + "step": 3790 + }, + { + "epoch": 0.3418857374757632, + "grad_norm": 1.2993377374147401, + "learning_rate": 3.063551772448682e-06, + "loss": 1.0406, + "step": 3791 + }, + { + "epoch": 0.34197592099923346, + "grad_norm": 2.028257660134623, + "learning_rate": 3.0630569677935075e-06, + "loss": 0.9335, + "step": 3792 + }, + { + "epoch": 0.3420661045227037, + "grad_norm": 1.9715184924743194, + "learning_rate": 3.06256207243245e-06, + "loss": 1.0947, + "step": 3793 + }, + { + "epoch": 0.342156288046174, + "grad_norm": 0.6101668421647709, + "learning_rate": 3.0620670864077385e-06, + "loss": 0.8058, + "step": 3794 + }, + { + "epoch": 0.3422464715696442, + "grad_norm": 0.6200346403673818, + "learning_rate": 3.0615720097616063e-06, + "loss": 0.8085, + "step": 3795 + }, + { + "epoch": 0.3423366550931145, + "grad_norm": 1.6173702265757963, + "learning_rate": 3.0610768425362967e-06, + "loss": 0.9616, + "step": 3796 + }, + { + "epoch": 0.34242683861658474, + "grad_norm": 1.5253349811794032, + "learning_rate": 3.0605815847740603e-06, + "loss": 0.9871, + "step": 3797 + }, + { + "epoch": 0.342517022140055, + "grad_norm": 5.879745839133908, + "learning_rate": 3.0600862365171553e-06, + "loss": 0.9737, + "step": 3798 + }, + { + "epoch": 0.34260720566352526, + "grad_norm": 1.5288410311254597, + "learning_rate": 3.0595907978078474e-06, + "loss": 0.9877, + "step": 3799 + }, + { + "epoch": 0.34269738918699555, + "grad_norm": 1.5075776702963992, + "learning_rate": 3.05909526868841e-06, + "loss": 0.9757, + "step": 3800 + }, + { + "epoch": 0.3427875727104658, + "grad_norm": 1.64302490544728, + "learning_rate": 3.0585996492011243e-06, + "loss": 1.0143, + "step": 3801 + }, + { + "epoch": 0.34287775623393607, + "grad_norm": 1.7212108681302185, + "learning_rate": 3.05810393938828e-06, + "loss": 1.0506, + "step": 3802 + }, + { + "epoch": 0.3429679397574063, + "grad_norm": 2.0030825225087407, + "learning_rate": 3.0576081392921723e-06, + "loss": 0.9223, + "step": 3803 + }, + { + "epoch": 0.3430581232808766, + "grad_norm": 1.7527332078909852, + "learning_rate": 3.057112248955107e-06, + "loss": 1.0515, + "step": 3804 + }, + { + "epoch": 0.3431483068043468, + "grad_norm": 2.177140692969764, + "learning_rate": 3.0566162684193963e-06, + "loss": 1.0266, + "step": 3805 + }, + { + "epoch": 0.3432384903278171, + "grad_norm": 1.4579325079713679, + "learning_rate": 3.056120197727359e-06, + "loss": 0.9552, + "step": 3806 + }, + { + "epoch": 0.34332867385128735, + "grad_norm": 1.5192967407960922, + "learning_rate": 3.0556240369213236e-06, + "loss": 0.9977, + "step": 3807 + }, + { + "epoch": 0.34341885737475764, + "grad_norm": 4.143927642558771, + "learning_rate": 3.055127786043624e-06, + "loss": 0.9182, + "step": 3808 + }, + { + "epoch": 0.34350904089822787, + "grad_norm": 1.5844862738966052, + "learning_rate": 3.054631445136604e-06, + "loss": 1.092, + "step": 3809 + }, + { + "epoch": 0.34359922442169816, + "grad_norm": 1.941437135753911, + "learning_rate": 3.0541350142426147e-06, + "loss": 1.0142, + "step": 3810 + }, + { + "epoch": 0.3436894079451684, + "grad_norm": 1.5582086761444582, + "learning_rate": 3.053638493404012e-06, + "loss": 0.9646, + "step": 3811 + }, + { + "epoch": 0.3437795914686387, + "grad_norm": 1.49922732479472, + "learning_rate": 3.0531418826631643e-06, + "loss": 0.9428, + "step": 3812 + }, + { + "epoch": 0.34386977499210897, + "grad_norm": 1.5155258146256543, + "learning_rate": 3.052645182062444e-06, + "loss": 1.0876, + "step": 3813 + }, + { + "epoch": 0.3439599585155792, + "grad_norm": 1.8907471924695005, + "learning_rate": 3.0521483916442324e-06, + "loss": 1.0508, + "step": 3814 + }, + { + "epoch": 0.3440501420390495, + "grad_norm": 1.590541002495153, + "learning_rate": 3.0516515114509183e-06, + "loss": 0.9999, + "step": 3815 + }, + { + "epoch": 0.3441403255625197, + "grad_norm": 1.8146161846890916, + "learning_rate": 3.0511545415249e-06, + "loss": 0.9537, + "step": 3816 + }, + { + "epoch": 0.34423050908599, + "grad_norm": 1.7776979198334846, + "learning_rate": 3.050657481908579e-06, + "loss": 1.0423, + "step": 3817 + }, + { + "epoch": 0.34432069260946024, + "grad_norm": 1.7115797117829445, + "learning_rate": 3.0501603326443677e-06, + "loss": 1.1164, + "step": 3818 + }, + { + "epoch": 0.34441087613293053, + "grad_norm": 1.9347517765134115, + "learning_rate": 3.049663093774687e-06, + "loss": 0.9923, + "step": 3819 + }, + { + "epoch": 0.34450105965640077, + "grad_norm": 2.0411253101134417, + "learning_rate": 3.0491657653419643e-06, + "loss": 0.8989, + "step": 3820 + }, + { + "epoch": 0.34459124317987105, + "grad_norm": 2.0624775740820334, + "learning_rate": 3.0486683473886325e-06, + "loss": 1.0541, + "step": 3821 + }, + { + "epoch": 0.3446814267033413, + "grad_norm": 2.9929461202261325, + "learning_rate": 3.0481708399571355e-06, + "loss": 1.0622, + "step": 3822 + }, + { + "epoch": 0.3447716102268116, + "grad_norm": 1.880273713923872, + "learning_rate": 3.047673243089922e-06, + "loss": 1.0297, + "step": 3823 + }, + { + "epoch": 0.3448617937502818, + "grad_norm": 1.633615834594487, + "learning_rate": 3.047175556829451e-06, + "loss": 1.0371, + "step": 3824 + }, + { + "epoch": 0.3449519772737521, + "grad_norm": 2.043405752324788, + "learning_rate": 3.046677781218188e-06, + "loss": 1.0555, + "step": 3825 + }, + { + "epoch": 0.34504216079722233, + "grad_norm": 1.855388181844535, + "learning_rate": 3.0461799162986043e-06, + "loss": 0.9804, + "step": 3826 + }, + { + "epoch": 0.3451323443206926, + "grad_norm": 1.6882617785148692, + "learning_rate": 3.045681962113183e-06, + "loss": 0.9288, + "step": 3827 + }, + { + "epoch": 0.34522252784416285, + "grad_norm": 2.1355400111854097, + "learning_rate": 3.0451839187044095e-06, + "loss": 0.9565, + "step": 3828 + }, + { + "epoch": 0.34531271136763314, + "grad_norm": 1.6299980036015367, + "learning_rate": 3.0446857861147816e-06, + "loss": 1.0212, + "step": 3829 + }, + { + "epoch": 0.3454028948911034, + "grad_norm": 1.6430645924256333, + "learning_rate": 3.044187564386802e-06, + "loss": 0.9791, + "step": 3830 + }, + { + "epoch": 0.34549307841457366, + "grad_norm": 1.684776439494947, + "learning_rate": 3.0436892535629818e-06, + "loss": 1.0685, + "step": 3831 + }, + { + "epoch": 0.3455832619380439, + "grad_norm": 2.3339822298303425, + "learning_rate": 3.0431908536858393e-06, + "loss": 1.0863, + "step": 3832 + }, + { + "epoch": 0.3456734454615142, + "grad_norm": 3.8570328373835223, + "learning_rate": 3.0426923647979016e-06, + "loss": 0.9985, + "step": 3833 + }, + { + "epoch": 0.3457636289849844, + "grad_norm": 2.2534633049170485, + "learning_rate": 3.0421937869417016e-06, + "loss": 0.9445, + "step": 3834 + }, + { + "epoch": 0.3458538125084547, + "grad_norm": 1.841123229179849, + "learning_rate": 3.041695120159782e-06, + "loss": 1.0684, + "step": 3835 + }, + { + "epoch": 0.345943996031925, + "grad_norm": 1.5973188048568656, + "learning_rate": 3.04119636449469e-06, + "loss": 0.9392, + "step": 3836 + }, + { + "epoch": 0.3460341795553952, + "grad_norm": 1.3359992475907267, + "learning_rate": 3.040697519988983e-06, + "loss": 1.0362, + "step": 3837 + }, + { + "epoch": 0.3461243630788655, + "grad_norm": 0.6380827400120399, + "learning_rate": 3.040198586685226e-06, + "loss": 0.8292, + "step": 3838 + }, + { + "epoch": 0.34621454660233575, + "grad_norm": 1.7469457130928276, + "learning_rate": 3.039699564625989e-06, + "loss": 0.9546, + "step": 3839 + }, + { + "epoch": 0.34630473012580604, + "grad_norm": 0.6744754943354921, + "learning_rate": 3.039200453853853e-06, + "loss": 0.823, + "step": 3840 + }, + { + "epoch": 0.34639491364927627, + "grad_norm": 2.188942896579553, + "learning_rate": 3.038701254411404e-06, + "loss": 1.0526, + "step": 3841 + }, + { + "epoch": 0.34648509717274656, + "grad_norm": 1.7885174983114807, + "learning_rate": 3.0382019663412367e-06, + "loss": 1.0266, + "step": 3842 + }, + { + "epoch": 0.3465752806962168, + "grad_norm": 1.9761049849346983, + "learning_rate": 3.0377025896859532e-06, + "loss": 1.0433, + "step": 3843 + }, + { + "epoch": 0.3466654642196871, + "grad_norm": 1.5942266910710012, + "learning_rate": 3.0372031244881627e-06, + "loss": 1.1136, + "step": 3844 + }, + { + "epoch": 0.3467556477431573, + "grad_norm": 1.7821998740510294, + "learning_rate": 3.0367035707904826e-06, + "loss": 0.9619, + "step": 3845 + }, + { + "epoch": 0.3468458312666276, + "grad_norm": 1.6905023921711562, + "learning_rate": 3.036203928635537e-06, + "loss": 0.9185, + "step": 3846 + }, + { + "epoch": 0.34693601479009784, + "grad_norm": 2.0011561151673622, + "learning_rate": 3.035704198065959e-06, + "loss": 1.0718, + "step": 3847 + }, + { + "epoch": 0.3470261983135681, + "grad_norm": 3.7330741372276663, + "learning_rate": 3.0352043791243886e-06, + "loss": 0.9918, + "step": 3848 + }, + { + "epoch": 0.34711638183703836, + "grad_norm": 0.6844100873479103, + "learning_rate": 3.034704471853472e-06, + "loss": 0.8266, + "step": 3849 + }, + { + "epoch": 0.34720656536050865, + "grad_norm": 1.6112362367265614, + "learning_rate": 3.0342044762958646e-06, + "loss": 1.0547, + "step": 3850 + }, + { + "epoch": 0.3472967488839789, + "grad_norm": 1.636480710147559, + "learning_rate": 3.0337043924942286e-06, + "loss": 1.0438, + "step": 3851 + }, + { + "epoch": 0.34738693240744917, + "grad_norm": 1.6377968526353686, + "learning_rate": 3.0332042204912343e-06, + "loss": 0.9944, + "step": 3852 + }, + { + "epoch": 0.3474771159309194, + "grad_norm": 0.6665420465420784, + "learning_rate": 3.0327039603295587e-06, + "loss": 0.8307, + "step": 3853 + }, + { + "epoch": 0.3475672994543897, + "grad_norm": 2.2264936469959946, + "learning_rate": 3.032203612051887e-06, + "loss": 0.9659, + "step": 3854 + }, + { + "epoch": 0.3476574829778599, + "grad_norm": 1.8999720521176424, + "learning_rate": 3.0317031757009116e-06, + "loss": 0.9871, + "step": 3855 + }, + { + "epoch": 0.3477476665013302, + "grad_norm": 1.4719507768890747, + "learning_rate": 3.0312026513193326e-06, + "loss": 0.9658, + "step": 3856 + }, + { + "epoch": 0.34783785002480044, + "grad_norm": 1.8667489019513046, + "learning_rate": 3.0307020389498573e-06, + "loss": 0.9831, + "step": 3857 + }, + { + "epoch": 0.34792803354827073, + "grad_norm": 1.5666549371050302, + "learning_rate": 3.0302013386352004e-06, + "loss": 0.9336, + "step": 3858 + }, + { + "epoch": 0.348018217071741, + "grad_norm": 3.2788312443954224, + "learning_rate": 3.0297005504180854e-06, + "loss": 0.9968, + "step": 3859 + }, + { + "epoch": 0.34810840059521125, + "grad_norm": 0.6032789933074263, + "learning_rate": 3.0291996743412417e-06, + "loss": 0.7498, + "step": 3860 + }, + { + "epoch": 0.34819858411868154, + "grad_norm": 1.5311270683923, + "learning_rate": 3.0286987104474063e-06, + "loss": 1.0975, + "step": 3861 + }, + { + "epoch": 0.3482887676421518, + "grad_norm": 1.372887853147387, + "learning_rate": 3.028197658779325e-06, + "loss": 0.9812, + "step": 3862 + }, + { + "epoch": 0.34837895116562206, + "grad_norm": 1.634661198314919, + "learning_rate": 3.0276965193797503e-06, + "loss": 1.0237, + "step": 3863 + }, + { + "epoch": 0.3484691346890923, + "grad_norm": 1.3697236869778564, + "learning_rate": 3.0271952922914423e-06, + "loss": 0.9672, + "step": 3864 + }, + { + "epoch": 0.3485593182125626, + "grad_norm": 1.5587799802342777, + "learning_rate": 3.0266939775571675e-06, + "loss": 0.9843, + "step": 3865 + }, + { + "epoch": 0.3486495017360328, + "grad_norm": 1.8489675244026804, + "learning_rate": 3.026192575219701e-06, + "loss": 0.9915, + "step": 3866 + }, + { + "epoch": 0.3487396852595031, + "grad_norm": 1.4963479884288198, + "learning_rate": 3.025691085321826e-06, + "loss": 0.897, + "step": 3867 + }, + { + "epoch": 0.34882986878297334, + "grad_norm": 1.7054723259909759, + "learning_rate": 3.025189507906332e-06, + "loss": 0.9887, + "step": 3868 + }, + { + "epoch": 0.34892005230644363, + "grad_norm": 1.649045714494796, + "learning_rate": 3.0246878430160166e-06, + "loss": 1.0236, + "step": 3869 + }, + { + "epoch": 0.34901023582991386, + "grad_norm": 2.5102868158676013, + "learning_rate": 3.024186090693684e-06, + "loss": 0.9453, + "step": 3870 + }, + { + "epoch": 0.34910041935338415, + "grad_norm": 1.9125600094823734, + "learning_rate": 3.023684250982147e-06, + "loss": 1.0339, + "step": 3871 + }, + { + "epoch": 0.3491906028768544, + "grad_norm": 1.7340689036748975, + "learning_rate": 3.0231823239242252e-06, + "loss": 0.9222, + "step": 3872 + }, + { + "epoch": 0.34928078640032467, + "grad_norm": 1.5335221074180245, + "learning_rate": 3.0226803095627457e-06, + "loss": 0.9586, + "step": 3873 + }, + { + "epoch": 0.3493709699237949, + "grad_norm": 2.024225501310926, + "learning_rate": 3.022178207940543e-06, + "loss": 0.9862, + "step": 3874 + }, + { + "epoch": 0.3494611534472652, + "grad_norm": 2.2585460698713775, + "learning_rate": 3.02167601910046e-06, + "loss": 1.018, + "step": 3875 + }, + { + "epoch": 0.3495513369707354, + "grad_norm": 1.5550877808206645, + "learning_rate": 3.021173743085345e-06, + "loss": 1.0254, + "step": 3876 + }, + { + "epoch": 0.3496415204942057, + "grad_norm": 0.6423075143832729, + "learning_rate": 3.0206713799380557e-06, + "loss": 0.8166, + "step": 3877 + }, + { + "epoch": 0.34973170401767595, + "grad_norm": 1.6876913597962118, + "learning_rate": 3.0201689297014565e-06, + "loss": 1.013, + "step": 3878 + }, + { + "epoch": 0.34982188754114624, + "grad_norm": 1.2613096252956062, + "learning_rate": 3.0196663924184187e-06, + "loss": 1.0459, + "step": 3879 + }, + { + "epoch": 0.34991207106461647, + "grad_norm": 2.3749637600994857, + "learning_rate": 3.019163768131822e-06, + "loss": 1.0364, + "step": 3880 + }, + { + "epoch": 0.35000225458808676, + "grad_norm": 1.8706356116853229, + "learning_rate": 3.0186610568845533e-06, + "loss": 0.926, + "step": 3881 + }, + { + "epoch": 0.350092438111557, + "grad_norm": 1.7507811573986896, + "learning_rate": 3.018158258719507e-06, + "loss": 0.9383, + "step": 3882 + }, + { + "epoch": 0.3501826216350273, + "grad_norm": 1.3837473236444051, + "learning_rate": 3.0176553736795827e-06, + "loss": 1.0297, + "step": 3883 + }, + { + "epoch": 0.35027280515849757, + "grad_norm": 0.5562676234078813, + "learning_rate": 3.017152401807691e-06, + "loss": 0.7547, + "step": 3884 + }, + { + "epoch": 0.3503629886819678, + "grad_norm": 1.5816362806294915, + "learning_rate": 3.0166493431467476e-06, + "loss": 0.9861, + "step": 3885 + }, + { + "epoch": 0.3504531722054381, + "grad_norm": 1.3308415623607306, + "learning_rate": 3.016146197739677e-06, + "loss": 0.9753, + "step": 3886 + }, + { + "epoch": 0.3505433557289083, + "grad_norm": 2.7496859631261508, + "learning_rate": 3.0156429656294097e-06, + "loss": 0.9331, + "step": 3887 + }, + { + "epoch": 0.3506335392523786, + "grad_norm": 1.4283284048724674, + "learning_rate": 3.0151396468588844e-06, + "loss": 0.9399, + "step": 3888 + }, + { + "epoch": 0.35072372277584885, + "grad_norm": 4.370304094292171, + "learning_rate": 3.014636241471047e-06, + "loss": 0.9546, + "step": 3889 + }, + { + "epoch": 0.35081390629931913, + "grad_norm": 1.6676119349026282, + "learning_rate": 3.0141327495088514e-06, + "loss": 0.9938, + "step": 3890 + }, + { + "epoch": 0.35090408982278937, + "grad_norm": 2.1633772333135486, + "learning_rate": 3.0136291710152566e-06, + "loss": 0.9577, + "step": 3891 + }, + { + "epoch": 0.35099427334625966, + "grad_norm": 2.1801984728137427, + "learning_rate": 3.0131255060332325e-06, + "loss": 1.0423, + "step": 3892 + }, + { + "epoch": 0.3510844568697299, + "grad_norm": 1.6239636858088693, + "learning_rate": 3.012621754605754e-06, + "loss": 1.0199, + "step": 3893 + }, + { + "epoch": 0.3511746403932002, + "grad_norm": 1.6977622619195087, + "learning_rate": 3.0121179167758035e-06, + "loss": 1.0267, + "step": 3894 + }, + { + "epoch": 0.3512648239166704, + "grad_norm": 1.8709419841977042, + "learning_rate": 3.0116139925863717e-06, + "loss": 1.0778, + "step": 3895 + }, + { + "epoch": 0.3513550074401407, + "grad_norm": 1.7393160022594925, + "learning_rate": 3.011109982080456e-06, + "loss": 1.0401, + "step": 3896 + }, + { + "epoch": 0.35144519096361093, + "grad_norm": 1.9567307414730966, + "learning_rate": 3.0106058853010614e-06, + "loss": 0.9436, + "step": 3897 + }, + { + "epoch": 0.3515353744870812, + "grad_norm": 1.4707019091597315, + "learning_rate": 3.010101702291201e-06, + "loss": 0.9863, + "step": 3898 + }, + { + "epoch": 0.35162555801055145, + "grad_norm": 1.426037073046705, + "learning_rate": 3.009597433093893e-06, + "loss": 0.995, + "step": 3899 + }, + { + "epoch": 0.35171574153402174, + "grad_norm": 1.726596745630399, + "learning_rate": 3.009093077752165e-06, + "loss": 1.009, + "step": 3900 + }, + { + "epoch": 0.351805925057492, + "grad_norm": 1.6430333943331155, + "learning_rate": 3.008588636309052e-06, + "loss": 1.0291, + "step": 3901 + }, + { + "epoch": 0.35189610858096226, + "grad_norm": 1.6417743834707323, + "learning_rate": 3.0080841088075947e-06, + "loss": 0.9651, + "step": 3902 + }, + { + "epoch": 0.3519862921044325, + "grad_norm": 1.8703552572642062, + "learning_rate": 3.0075794952908436e-06, + "loss": 1.0437, + "step": 3903 + }, + { + "epoch": 0.3520764756279028, + "grad_norm": 1.414402367651153, + "learning_rate": 3.0070747958018528e-06, + "loss": 1.0145, + "step": 3904 + }, + { + "epoch": 0.352166659151373, + "grad_norm": 1.462366720794215, + "learning_rate": 3.0065700103836894e-06, + "loss": 1.0157, + "step": 3905 + }, + { + "epoch": 0.3522568426748433, + "grad_norm": 1.6559347176514947, + "learning_rate": 3.0060651390794214e-06, + "loss": 1.0205, + "step": 3906 + }, + { + "epoch": 0.3523470261983136, + "grad_norm": 1.3991025483459991, + "learning_rate": 3.005560181932128e-06, + "loss": 0.9811, + "step": 3907 + }, + { + "epoch": 0.35243720972178383, + "grad_norm": 1.6494737595121336, + "learning_rate": 3.005055138984896e-06, + "loss": 0.9553, + "step": 3908 + }, + { + "epoch": 0.3525273932452541, + "grad_norm": 3.456435881819154, + "learning_rate": 3.0045500102808174e-06, + "loss": 0.9745, + "step": 3909 + }, + { + "epoch": 0.35261757676872435, + "grad_norm": 2.326038116529813, + "learning_rate": 3.0040447958629927e-06, + "loss": 1.0806, + "step": 3910 + }, + { + "epoch": 0.35270776029219464, + "grad_norm": 1.659647138733656, + "learning_rate": 3.00353949577453e-06, + "loss": 1.0229, + "step": 3911 + }, + { + "epoch": 0.35279794381566487, + "grad_norm": 3.248922169296457, + "learning_rate": 3.003034110058544e-06, + "loss": 0.8393, + "step": 3912 + }, + { + "epoch": 0.35288812733913516, + "grad_norm": 1.5152226336723247, + "learning_rate": 3.002528638758157e-06, + "loss": 0.9781, + "step": 3913 + }, + { + "epoch": 0.3529783108626054, + "grad_norm": 2.225689204767427, + "learning_rate": 3.0020230819164985e-06, + "loss": 0.9978, + "step": 3914 + }, + { + "epoch": 0.3530684943860757, + "grad_norm": 1.4293457623399217, + "learning_rate": 3.0015174395767064e-06, + "loss": 1.0079, + "step": 3915 + }, + { + "epoch": 0.3531586779095459, + "grad_norm": 2.1718876241413123, + "learning_rate": 3.001011711781923e-06, + "loss": 1.0188, + "step": 3916 + }, + { + "epoch": 0.3532488614330162, + "grad_norm": 1.5592952669663755, + "learning_rate": 3.0005058985753017e-06, + "loss": 0.9566, + "step": 3917 + }, + { + "epoch": 0.35333904495648644, + "grad_norm": 1.4806404920019531, + "learning_rate": 3e-06, + "loss": 0.9907, + "step": 3918 + }, + { + "epoch": 0.3534292284799567, + "grad_norm": 0.6497030891357347, + "learning_rate": 2.9994940160991843e-06, + "loss": 0.8432, + "step": 3919 + }, + { + "epoch": 0.35351941200342696, + "grad_norm": 0.6257844531945246, + "learning_rate": 2.9989879469160285e-06, + "loss": 0.8007, + "step": 3920 + }, + { + "epoch": 0.35360959552689725, + "grad_norm": 1.5891347893070038, + "learning_rate": 2.9984817924937124e-06, + "loss": 1.0176, + "step": 3921 + }, + { + "epoch": 0.3536997790503675, + "grad_norm": 1.7087953299458585, + "learning_rate": 2.997975552875424e-06, + "loss": 0.9371, + "step": 3922 + }, + { + "epoch": 0.35378996257383777, + "grad_norm": 1.862981552237448, + "learning_rate": 2.997469228104358e-06, + "loss": 0.9873, + "step": 3923 + }, + { + "epoch": 0.353880146097308, + "grad_norm": 1.6618659337640387, + "learning_rate": 2.996962818223718e-06, + "loss": 0.8982, + "step": 3924 + }, + { + "epoch": 0.3539703296207783, + "grad_norm": 4.898102817326975, + "learning_rate": 2.9964563232767135e-06, + "loss": 0.9425, + "step": 3925 + }, + { + "epoch": 0.3540605131442485, + "grad_norm": 1.7258850425637462, + "learning_rate": 2.9959497433065617e-06, + "loss": 0.9958, + "step": 3926 + }, + { + "epoch": 0.3541506966677188, + "grad_norm": 1.83790041913819, + "learning_rate": 2.9954430783564848e-06, + "loss": 0.9208, + "step": 3927 + }, + { + "epoch": 0.35424088019118904, + "grad_norm": 1.5107417140533357, + "learning_rate": 2.994936328469716e-06, + "loss": 1.0248, + "step": 3928 + }, + { + "epoch": 0.35433106371465933, + "grad_norm": 1.5224386217244803, + "learning_rate": 2.994429493689494e-06, + "loss": 1.0004, + "step": 3929 + }, + { + "epoch": 0.35442124723812957, + "grad_norm": 1.372628334035108, + "learning_rate": 2.9939225740590642e-06, + "loss": 1.0205, + "step": 3930 + }, + { + "epoch": 0.35451143076159986, + "grad_norm": 1.5834741864378434, + "learning_rate": 2.99341556962168e-06, + "loss": 0.9541, + "step": 3931 + }, + { + "epoch": 0.35460161428507014, + "grad_norm": 1.515053159454648, + "learning_rate": 2.992908480420602e-06, + "loss": 1.0458, + "step": 3932 + }, + { + "epoch": 0.3546917978085404, + "grad_norm": 1.7890753599812437, + "learning_rate": 2.9924013064990974e-06, + "loss": 0.9884, + "step": 3933 + }, + { + "epoch": 0.35478198133201067, + "grad_norm": 0.6489300063937884, + "learning_rate": 2.991894047900441e-06, + "loss": 0.8482, + "step": 3934 + }, + { + "epoch": 0.3548721648554809, + "grad_norm": 1.4646309252947352, + "learning_rate": 2.991386704667916e-06, + "loss": 1.0111, + "step": 3935 + }, + { + "epoch": 0.3549623483789512, + "grad_norm": 1.47810968951629, + "learning_rate": 2.9908792768448097e-06, + "loss": 1.0621, + "step": 3936 + }, + { + "epoch": 0.3550525319024214, + "grad_norm": 1.8186882532918454, + "learning_rate": 2.990371764474421e-06, + "loss": 0.9952, + "step": 3937 + }, + { + "epoch": 0.3551427154258917, + "grad_norm": 1.9303132559978777, + "learning_rate": 2.9898641676000518e-06, + "loss": 0.967, + "step": 3938 + }, + { + "epoch": 0.35523289894936194, + "grad_norm": 1.5424932205817863, + "learning_rate": 2.9893564862650138e-06, + "loss": 0.9958, + "step": 3939 + }, + { + "epoch": 0.35532308247283223, + "grad_norm": 1.7510394006911516, + "learning_rate": 2.9888487205126254e-06, + "loss": 1.0496, + "step": 3940 + }, + { + "epoch": 0.35541326599630246, + "grad_norm": 1.5723017149390548, + "learning_rate": 2.9883408703862115e-06, + "loss": 0.9919, + "step": 3941 + }, + { + "epoch": 0.35550344951977275, + "grad_norm": 2.196730837697467, + "learning_rate": 2.987832935929105e-06, + "loss": 1.0714, + "step": 3942 + }, + { + "epoch": 0.355593633043243, + "grad_norm": 1.6793223849176047, + "learning_rate": 2.9873249171846454e-06, + "loss": 1.0589, + "step": 3943 + }, + { + "epoch": 0.3556838165667133, + "grad_norm": 5.876666218279243, + "learning_rate": 2.98681681419618e-06, + "loss": 0.9911, + "step": 3944 + }, + { + "epoch": 0.3557740000901835, + "grad_norm": 2.161760338209853, + "learning_rate": 2.9863086270070627e-06, + "loss": 1.0535, + "step": 3945 + }, + { + "epoch": 0.3558641836136538, + "grad_norm": 1.6685644153738726, + "learning_rate": 2.985800355660655e-06, + "loss": 1.002, + "step": 3946 + }, + { + "epoch": 0.35595436713712403, + "grad_norm": 0.6741874731772352, + "learning_rate": 2.9852920002003252e-06, + "loss": 0.8462, + "step": 3947 + }, + { + "epoch": 0.3560445506605943, + "grad_norm": 1.9376883261429403, + "learning_rate": 2.9847835606694494e-06, + "loss": 1.0327, + "step": 3948 + }, + { + "epoch": 0.35613473418406455, + "grad_norm": 1.7850001828276383, + "learning_rate": 2.98427503711141e-06, + "loss": 0.9261, + "step": 3949 + }, + { + "epoch": 0.35622491770753484, + "grad_norm": 2.552379627285053, + "learning_rate": 2.9837664295695973e-06, + "loss": 1.0311, + "step": 3950 + }, + { + "epoch": 0.35631510123100507, + "grad_norm": 1.3966528304073986, + "learning_rate": 2.983257738087408e-06, + "loss": 0.991, + "step": 3951 + }, + { + "epoch": 0.35640528475447536, + "grad_norm": 1.7598527385052323, + "learning_rate": 2.982748962708247e-06, + "loss": 1.0288, + "step": 3952 + }, + { + "epoch": 0.3564954682779456, + "grad_norm": 1.4902604527022878, + "learning_rate": 2.982240103475526e-06, + "loss": 1.0511, + "step": 3953 + }, + { + "epoch": 0.3565856518014159, + "grad_norm": 2.0638401417076087, + "learning_rate": 2.981731160432663e-06, + "loss": 1.0, + "step": 3954 + }, + { + "epoch": 0.35667583532488617, + "grad_norm": 1.7494164583854799, + "learning_rate": 2.981222133623084e-06, + "loss": 0.9001, + "step": 3955 + }, + { + "epoch": 0.3567660188483564, + "grad_norm": 1.7152823406507143, + "learning_rate": 2.980713023090222e-06, + "loss": 1.0089, + "step": 3956 + }, + { + "epoch": 0.3568562023718267, + "grad_norm": 1.6120199302075495, + "learning_rate": 2.980203828877518e-06, + "loss": 0.9716, + "step": 3957 + }, + { + "epoch": 0.3569463858952969, + "grad_norm": 2.8200654743250166, + "learning_rate": 2.9796945510284182e-06, + "loss": 1.0378, + "step": 3958 + }, + { + "epoch": 0.3570365694187672, + "grad_norm": 1.672802106546018, + "learning_rate": 2.9791851895863774e-06, + "loss": 1.0576, + "step": 3959 + }, + { + "epoch": 0.35712675294223745, + "grad_norm": 1.6553672741423988, + "learning_rate": 2.978675744594857e-06, + "loss": 1.029, + "step": 3960 + }, + { + "epoch": 0.35721693646570774, + "grad_norm": 7.3337664331798535, + "learning_rate": 2.978166216097326e-06, + "loss": 0.9572, + "step": 3961 + }, + { + "epoch": 0.35730711998917797, + "grad_norm": 2.0800421340049438, + "learning_rate": 2.9776566041372596e-06, + "loss": 0.9856, + "step": 3962 + }, + { + "epoch": 0.35739730351264826, + "grad_norm": 1.5375768781840777, + "learning_rate": 2.977146908758141e-06, + "loss": 1.056, + "step": 3963 + }, + { + "epoch": 0.3574874870361185, + "grad_norm": 1.6484390186465532, + "learning_rate": 2.9766371300034604e-06, + "loss": 1.0167, + "step": 3964 + }, + { + "epoch": 0.3575776705595888, + "grad_norm": 1.7047862778672127, + "learning_rate": 2.9761272679167142e-06, + "loss": 1.0492, + "step": 3965 + }, + { + "epoch": 0.357667854083059, + "grad_norm": 2.2778888135172064, + "learning_rate": 2.9756173225414072e-06, + "loss": 0.8498, + "step": 3966 + }, + { + "epoch": 0.3577580376065293, + "grad_norm": 1.6429503900116116, + "learning_rate": 2.975107293921051e-06, + "loss": 1.1748, + "step": 3967 + }, + { + "epoch": 0.35784822112999953, + "grad_norm": 1.8143704070093232, + "learning_rate": 2.9745971820991643e-06, + "loss": 1.0382, + "step": 3968 + }, + { + "epoch": 0.3579384046534698, + "grad_norm": 1.6767988338795823, + "learning_rate": 2.9740869871192715e-06, + "loss": 1.042, + "step": 3969 + }, + { + "epoch": 0.35802858817694005, + "grad_norm": 3.311898518781343, + "learning_rate": 2.9735767090249065e-06, + "loss": 1.0639, + "step": 3970 + }, + { + "epoch": 0.35811877170041034, + "grad_norm": 1.7136991326081865, + "learning_rate": 2.973066347859608e-06, + "loss": 0.9144, + "step": 3971 + }, + { + "epoch": 0.3582089552238806, + "grad_norm": 1.3318949878612354, + "learning_rate": 2.972555903666923e-06, + "loss": 1.0121, + "step": 3972 + }, + { + "epoch": 0.35829913874735086, + "grad_norm": 2.1219491937219668, + "learning_rate": 2.972045376490406e-06, + "loss": 0.913, + "step": 3973 + }, + { + "epoch": 0.3583893222708211, + "grad_norm": 1.4140665465239064, + "learning_rate": 2.9715347663736177e-06, + "loss": 0.9291, + "step": 3974 + }, + { + "epoch": 0.3584795057942914, + "grad_norm": 5.054729007509252, + "learning_rate": 2.9710240733601266e-06, + "loss": 0.9996, + "step": 3975 + }, + { + "epoch": 0.3585696893177616, + "grad_norm": 1.9649817719718212, + "learning_rate": 2.970513297493507e-06, + "loss": 1.0841, + "step": 3976 + }, + { + "epoch": 0.3586598728412319, + "grad_norm": 1.4563838553407855, + "learning_rate": 2.9700024388173416e-06, + "loss": 1.0528, + "step": 3977 + }, + { + "epoch": 0.35875005636470214, + "grad_norm": 1.6956082275101023, + "learning_rate": 2.969491497375219e-06, + "loss": 1.0511, + "step": 3978 + }, + { + "epoch": 0.35884023988817243, + "grad_norm": 1.5074395227400097, + "learning_rate": 2.9689804732107364e-06, + "loss": 1.0247, + "step": 3979 + }, + { + "epoch": 0.3589304234116427, + "grad_norm": 2.104129287337356, + "learning_rate": 2.9684693663674968e-06, + "loss": 0.9579, + "step": 3980 + }, + { + "epoch": 0.35902060693511295, + "grad_norm": 1.937849751560268, + "learning_rate": 2.9679581768891115e-06, + "loss": 1.0858, + "step": 3981 + }, + { + "epoch": 0.35911079045858324, + "grad_norm": 1.6929724228083718, + "learning_rate": 2.967446904819197e-06, + "loss": 0.9248, + "step": 3982 + }, + { + "epoch": 0.3592009739820535, + "grad_norm": 1.4217447074468315, + "learning_rate": 2.966935550201378e-06, + "loss": 1.0245, + "step": 3983 + }, + { + "epoch": 0.35929115750552376, + "grad_norm": 1.677757964673362, + "learning_rate": 2.966424113079286e-06, + "loss": 0.9969, + "step": 3984 + }, + { + "epoch": 0.359381341028994, + "grad_norm": 1.6904007034858146, + "learning_rate": 2.9659125934965596e-06, + "loss": 1.0527, + "step": 3985 + }, + { + "epoch": 0.3594715245524643, + "grad_norm": 1.764377457079526, + "learning_rate": 2.9654009914968457e-06, + "loss": 1.0598, + "step": 3986 + }, + { + "epoch": 0.3595617080759345, + "grad_norm": 3.8676917778062863, + "learning_rate": 2.9648893071237956e-06, + "loss": 0.8921, + "step": 3987 + }, + { + "epoch": 0.3596518915994048, + "grad_norm": 1.5502022303662024, + "learning_rate": 2.964377540421069e-06, + "loss": 0.9742, + "step": 3988 + }, + { + "epoch": 0.35974207512287504, + "grad_norm": 1.7596829105968375, + "learning_rate": 2.963865691432334e-06, + "loss": 1.0229, + "step": 3989 + }, + { + "epoch": 0.3598322586463453, + "grad_norm": 1.5654497151274618, + "learning_rate": 2.963353760201263e-06, + "loss": 0.9194, + "step": 3990 + }, + { + "epoch": 0.35992244216981556, + "grad_norm": 1.776275649184204, + "learning_rate": 2.962841746771537e-06, + "loss": 0.9896, + "step": 3991 + }, + { + "epoch": 0.36001262569328585, + "grad_norm": 1.5909571392680688, + "learning_rate": 2.9623296511868445e-06, + "loss": 1.0038, + "step": 3992 + }, + { + "epoch": 0.3601028092167561, + "grad_norm": 2.3559100242234057, + "learning_rate": 2.96181747349088e-06, + "loss": 0.9306, + "step": 3993 + }, + { + "epoch": 0.36019299274022637, + "grad_norm": 1.8505636490659771, + "learning_rate": 2.961305213727345e-06, + "loss": 1.0213, + "step": 3994 + }, + { + "epoch": 0.3602831762636966, + "grad_norm": 0.6512396829194252, + "learning_rate": 2.960792871939949e-06, + "loss": 0.8405, + "step": 3995 + }, + { + "epoch": 0.3603733597871669, + "grad_norm": 1.5440527090207292, + "learning_rate": 2.9602804481724064e-06, + "loss": 1.0938, + "step": 3996 + }, + { + "epoch": 0.3604635433106371, + "grad_norm": 1.8906371691603776, + "learning_rate": 2.9597679424684427e-06, + "loss": 0.9985, + "step": 3997 + }, + { + "epoch": 0.3605537268341074, + "grad_norm": 1.9129017326931432, + "learning_rate": 2.9592553548717848e-06, + "loss": 1.0155, + "step": 3998 + }, + { + "epoch": 0.36064391035757765, + "grad_norm": 1.5548817115514333, + "learning_rate": 2.958742685426171e-06, + "loss": 1.0837, + "step": 3999 + }, + { + "epoch": 0.36073409388104793, + "grad_norm": 1.4996050473642866, + "learning_rate": 2.9582299341753446e-06, + "loss": 1.0679, + "step": 4000 + }, + { + "epoch": 0.36082427740451817, + "grad_norm": 1.9880292391747156, + "learning_rate": 2.957717101163057e-06, + "loss": 0.9483, + "step": 4001 + }, + { + "epoch": 0.36091446092798846, + "grad_norm": 2.1183077955569933, + "learning_rate": 2.9572041864330655e-06, + "loss": 1.0003, + "step": 4002 + }, + { + "epoch": 0.36100464445145874, + "grad_norm": 1.7904110286352517, + "learning_rate": 2.9566911900291346e-06, + "loss": 1.0228, + "step": 4003 + }, + { + "epoch": 0.361094827974929, + "grad_norm": 2.128206302473388, + "learning_rate": 2.9561781119950368e-06, + "loss": 0.9207, + "step": 4004 + }, + { + "epoch": 0.36118501149839927, + "grad_norm": 1.6198835520630064, + "learning_rate": 2.9556649523745493e-06, + "loss": 0.9481, + "step": 4005 + }, + { + "epoch": 0.3612751950218695, + "grad_norm": 2.1935375273750584, + "learning_rate": 2.955151711211459e-06, + "loss": 1.061, + "step": 4006 + }, + { + "epoch": 0.3613653785453398, + "grad_norm": 1.8012204244966525, + "learning_rate": 2.9546383885495583e-06, + "loss": 0.9719, + "step": 4007 + }, + { + "epoch": 0.36145556206881, + "grad_norm": 2.6449470136848103, + "learning_rate": 2.9541249844326464e-06, + "loss": 0.9515, + "step": 4008 + }, + { + "epoch": 0.3615457455922803, + "grad_norm": 1.4287552425792291, + "learning_rate": 2.9536114989045295e-06, + "loss": 0.9859, + "step": 4009 + }, + { + "epoch": 0.36163592911575054, + "grad_norm": 1.446358709313312, + "learning_rate": 2.9530979320090216e-06, + "loss": 1.0599, + "step": 4010 + }, + { + "epoch": 0.36172611263922083, + "grad_norm": 1.8083495928946014, + "learning_rate": 2.9525842837899422e-06, + "loss": 1.0836, + "step": 4011 + }, + { + "epoch": 0.36181629616269106, + "grad_norm": 1.752202759375749, + "learning_rate": 2.95207055429112e-06, + "loss": 1.0833, + "step": 4012 + }, + { + "epoch": 0.36190647968616135, + "grad_norm": 1.806890357000748, + "learning_rate": 2.951556743556388e-06, + "loss": 0.9937, + "step": 4013 + }, + { + "epoch": 0.3619966632096316, + "grad_norm": 2.5435510018606196, + "learning_rate": 2.951042851629588e-06, + "loss": 1.0535, + "step": 4014 + }, + { + "epoch": 0.3620868467331019, + "grad_norm": 1.638366961840662, + "learning_rate": 2.950528878554568e-06, + "loss": 0.9576, + "step": 4015 + }, + { + "epoch": 0.3621770302565721, + "grad_norm": 1.3958009554775392, + "learning_rate": 2.950014824375183e-06, + "loss": 1.0417, + "step": 4016 + }, + { + "epoch": 0.3622672137800424, + "grad_norm": 0.6334859067151606, + "learning_rate": 2.949500689135295e-06, + "loss": 0.7651, + "step": 4017 + }, + { + "epoch": 0.36235739730351263, + "grad_norm": 1.3847926988631793, + "learning_rate": 2.9489864728787722e-06, + "loss": 1.0821, + "step": 4018 + }, + { + "epoch": 0.3624475808269829, + "grad_norm": 1.4926530201742312, + "learning_rate": 2.9484721756494915e-06, + "loss": 1.0635, + "step": 4019 + }, + { + "epoch": 0.36253776435045315, + "grad_norm": 1.988001475857469, + "learning_rate": 2.9479577974913343e-06, + "loss": 1.0051, + "step": 4020 + }, + { + "epoch": 0.36262794787392344, + "grad_norm": 1.6413346708469432, + "learning_rate": 2.9474433384481908e-06, + "loss": 1.0036, + "step": 4021 + }, + { + "epoch": 0.3627181313973937, + "grad_norm": 1.4561226392070747, + "learning_rate": 2.9469287985639577e-06, + "loss": 1.0008, + "step": 4022 + }, + { + "epoch": 0.36280831492086396, + "grad_norm": 1.4892166682328758, + "learning_rate": 2.9464141778825384e-06, + "loss": 1.0507, + "step": 4023 + }, + { + "epoch": 0.3628984984443342, + "grad_norm": 1.7084792780368878, + "learning_rate": 2.9458994764478427e-06, + "loss": 1.0391, + "step": 4024 + }, + { + "epoch": 0.3629886819678045, + "grad_norm": 2.073828479093549, + "learning_rate": 2.9453846943037883e-06, + "loss": 1.0378, + "step": 4025 + }, + { + "epoch": 0.36307886549127477, + "grad_norm": 2.6103589606902737, + "learning_rate": 2.9448698314942987e-06, + "loss": 0.9526, + "step": 4026 + }, + { + "epoch": 0.363169049014745, + "grad_norm": 0.7454284977564827, + "learning_rate": 2.944354888063305e-06, + "loss": 0.8614, + "step": 4027 + }, + { + "epoch": 0.3632592325382153, + "grad_norm": 1.5537698615869335, + "learning_rate": 2.9438398640547453e-06, + "loss": 0.9795, + "step": 4028 + }, + { + "epoch": 0.3633494160616855, + "grad_norm": 1.6405963713555, + "learning_rate": 2.943324759512564e-06, + "loss": 0.9425, + "step": 4029 + }, + { + "epoch": 0.3634395995851558, + "grad_norm": 2.5896877972582075, + "learning_rate": 2.9428095744807134e-06, + "loss": 1.0284, + "step": 4030 + }, + { + "epoch": 0.36352978310862605, + "grad_norm": 1.3714641277282191, + "learning_rate": 2.942294309003151e-06, + "loss": 0.9571, + "step": 4031 + }, + { + "epoch": 0.36361996663209634, + "grad_norm": 1.66543665638751, + "learning_rate": 2.941778963123843e-06, + "loss": 1.0235, + "step": 4032 + }, + { + "epoch": 0.36371015015556657, + "grad_norm": 2.2529151892887964, + "learning_rate": 2.94126353688676e-06, + "loss": 1.0078, + "step": 4033 + }, + { + "epoch": 0.36380033367903686, + "grad_norm": 2.049703143886417, + "learning_rate": 2.9407480303358825e-06, + "loss": 0.9563, + "step": 4034 + }, + { + "epoch": 0.3638905172025071, + "grad_norm": 1.5963415878654936, + "learning_rate": 2.940232443515195e-06, + "loss": 1.012, + "step": 4035 + }, + { + "epoch": 0.3639807007259774, + "grad_norm": 1.4543281414514326, + "learning_rate": 2.9397167764686916e-06, + "loss": 1.0269, + "step": 4036 + }, + { + "epoch": 0.3640708842494476, + "grad_norm": 1.499036320435276, + "learning_rate": 2.9392010292403714e-06, + "loss": 0.9814, + "step": 4037 + }, + { + "epoch": 0.3641610677729179, + "grad_norm": 1.7121840860827464, + "learning_rate": 2.9386852018742404e-06, + "loss": 0.9927, + "step": 4038 + }, + { + "epoch": 0.36425125129638813, + "grad_norm": 1.6064469450504621, + "learning_rate": 2.938169294414312e-06, + "loss": 0.9759, + "step": 4039 + }, + { + "epoch": 0.3643414348198584, + "grad_norm": 1.5890011815906355, + "learning_rate": 2.9376533069046067e-06, + "loss": 1.0336, + "step": 4040 + }, + { + "epoch": 0.36443161834332866, + "grad_norm": 0.7142927620744416, + "learning_rate": 2.9371372393891514e-06, + "loss": 0.873, + "step": 4041 + }, + { + "epoch": 0.36452180186679894, + "grad_norm": 1.3832004240064042, + "learning_rate": 2.936621091911979e-06, + "loss": 1.0784, + "step": 4042 + }, + { + "epoch": 0.3646119853902692, + "grad_norm": 2.3140609215750967, + "learning_rate": 2.936104864517131e-06, + "loss": 0.9347, + "step": 4043 + }, + { + "epoch": 0.36470216891373947, + "grad_norm": 2.01962734561932, + "learning_rate": 2.9355885572486535e-06, + "loss": 0.9626, + "step": 4044 + }, + { + "epoch": 0.3647923524372097, + "grad_norm": 2.49385115721856, + "learning_rate": 2.9350721701506026e-06, + "loss": 1.0348, + "step": 4045 + }, + { + "epoch": 0.36488253596068, + "grad_norm": 1.5512728356364318, + "learning_rate": 2.9345557032670375e-06, + "loss": 0.9953, + "step": 4046 + }, + { + "epoch": 0.3649727194841502, + "grad_norm": 1.8043859320673754, + "learning_rate": 2.934039156642027e-06, + "loss": 1.0509, + "step": 4047 + }, + { + "epoch": 0.3650629030076205, + "grad_norm": 1.7543116995083805, + "learning_rate": 2.9335225303196454e-06, + "loss": 0.8655, + "step": 4048 + }, + { + "epoch": 0.36515308653109074, + "grad_norm": 1.7504910052715905, + "learning_rate": 2.933005824343974e-06, + "loss": 0.9732, + "step": 4049 + }, + { + "epoch": 0.36524327005456103, + "grad_norm": 1.6513889882043384, + "learning_rate": 2.932489038759101e-06, + "loss": 0.9744, + "step": 4050 + }, + { + "epoch": 0.3653334535780313, + "grad_norm": 1.463274142792362, + "learning_rate": 2.9319721736091215e-06, + "loss": 0.8926, + "step": 4051 + }, + { + "epoch": 0.36542363710150155, + "grad_norm": 1.3031830937085718, + "learning_rate": 2.9314552289381377e-06, + "loss": 0.9079, + "step": 4052 + }, + { + "epoch": 0.36551382062497184, + "grad_norm": 1.786313739704508, + "learning_rate": 2.9309382047902574e-06, + "loss": 0.9719, + "step": 4053 + }, + { + "epoch": 0.3656040041484421, + "grad_norm": 1.6530032453787826, + "learning_rate": 2.9304211012095963e-06, + "loss": 1.0472, + "step": 4054 + }, + { + "epoch": 0.36569418767191236, + "grad_norm": 1.7683414054683968, + "learning_rate": 2.929903918240277e-06, + "loss": 1.0581, + "step": 4055 + }, + { + "epoch": 0.3657843711953826, + "grad_norm": 0.6400988674539381, + "learning_rate": 2.9293866559264273e-06, + "loss": 0.8162, + "step": 4056 + }, + { + "epoch": 0.3658745547188529, + "grad_norm": 1.800637725605034, + "learning_rate": 2.928869314312184e-06, + "loss": 1.0042, + "step": 4057 + }, + { + "epoch": 0.3659647382423231, + "grad_norm": 1.5169340163524059, + "learning_rate": 2.9283518934416892e-06, + "loss": 1.0054, + "step": 4058 + }, + { + "epoch": 0.3660549217657934, + "grad_norm": 1.812154079181046, + "learning_rate": 2.927834393359092e-06, + "loss": 1.0098, + "step": 4059 + }, + { + "epoch": 0.36614510528926364, + "grad_norm": 1.7717070350222657, + "learning_rate": 2.927316814108548e-06, + "loss": 1.0694, + "step": 4060 + }, + { + "epoch": 0.36623528881273393, + "grad_norm": 1.5181909450564, + "learning_rate": 2.92679915573422e-06, + "loss": 0.9077, + "step": 4061 + }, + { + "epoch": 0.36632547233620416, + "grad_norm": 1.6222171429913113, + "learning_rate": 2.926281418280278e-06, + "loss": 0.9436, + "step": 4062 + }, + { + "epoch": 0.36641565585967445, + "grad_norm": 1.4986007044661336, + "learning_rate": 2.925763601790899e-06, + "loss": 1.0058, + "step": 4063 + }, + { + "epoch": 0.3665058393831447, + "grad_norm": 1.6235791376796493, + "learning_rate": 2.9252457063102635e-06, + "loss": 0.9484, + "step": 4064 + }, + { + "epoch": 0.36659602290661497, + "grad_norm": 1.5556514186309387, + "learning_rate": 2.9247277318825626e-06, + "loss": 0.9648, + "step": 4065 + }, + { + "epoch": 0.3666862064300852, + "grad_norm": 1.429279206555867, + "learning_rate": 2.924209678551993e-06, + "loss": 1.0947, + "step": 4066 + }, + { + "epoch": 0.3667763899535555, + "grad_norm": 1.603220904420717, + "learning_rate": 2.923691546362757e-06, + "loss": 1.011, + "step": 4067 + }, + { + "epoch": 0.3668665734770257, + "grad_norm": 19.30962935658455, + "learning_rate": 2.9231733353590663e-06, + "loss": 0.9615, + "step": 4068 + }, + { + "epoch": 0.366956757000496, + "grad_norm": 2.810674456364202, + "learning_rate": 2.922655045585136e-06, + "loss": 1.0739, + "step": 4069 + }, + { + "epoch": 0.36704694052396625, + "grad_norm": 0.6919533025511572, + "learning_rate": 2.92213667708519e-06, + "loss": 0.8581, + "step": 4070 + }, + { + "epoch": 0.36713712404743654, + "grad_norm": 5.538546130561381, + "learning_rate": 2.921618229903457e-06, + "loss": 1.0105, + "step": 4071 + }, + { + "epoch": 0.36722730757090677, + "grad_norm": 12.20212964090439, + "learning_rate": 2.9210997040841752e-06, + "loss": 0.9239, + "step": 4072 + }, + { + "epoch": 0.36731749109437706, + "grad_norm": 1.9503233690391424, + "learning_rate": 2.9205810996715885e-06, + "loss": 0.9655, + "step": 4073 + }, + { + "epoch": 0.36740767461784735, + "grad_norm": 1.7385129624136677, + "learning_rate": 2.9200624167099456e-06, + "loss": 0.9697, + "step": 4074 + }, + { + "epoch": 0.3674978581413176, + "grad_norm": 1.409092890550669, + "learning_rate": 2.919543655243505e-06, + "loss": 0.9176, + "step": 4075 + }, + { + "epoch": 0.36758804166478787, + "grad_norm": 2.1417160038392287, + "learning_rate": 2.919024815316529e-06, + "loss": 1.0283, + "step": 4076 + }, + { + "epoch": 0.3676782251882581, + "grad_norm": 1.7480538991435342, + "learning_rate": 2.9185058969732877e-06, + "loss": 1.1277, + "step": 4077 + }, + { + "epoch": 0.3677684087117284, + "grad_norm": 1.8929262264322955, + "learning_rate": 2.917986900258059e-06, + "loss": 1.0327, + "step": 4078 + }, + { + "epoch": 0.3678585922351986, + "grad_norm": 1.803613648736528, + "learning_rate": 2.917467825215126e-06, + "loss": 0.9952, + "step": 4079 + }, + { + "epoch": 0.3679487757586689, + "grad_norm": 1.4246109849749695, + "learning_rate": 2.9169486718887803e-06, + "loss": 0.9523, + "step": 4080 + }, + { + "epoch": 0.36803895928213914, + "grad_norm": 1.4539809628516747, + "learning_rate": 2.9164294403233173e-06, + "loss": 1.0504, + "step": 4081 + }, + { + "epoch": 0.36812914280560943, + "grad_norm": 1.8378613071353171, + "learning_rate": 2.915910130563041e-06, + "loss": 0.9341, + "step": 4082 + }, + { + "epoch": 0.36821932632907967, + "grad_norm": 1.9890511152086976, + "learning_rate": 2.915390742652262e-06, + "loss": 0.9437, + "step": 4083 + }, + { + "epoch": 0.36830950985254995, + "grad_norm": 2.1133187056234766, + "learning_rate": 2.914871276635298e-06, + "loss": 0.938, + "step": 4084 + }, + { + "epoch": 0.3683996933760202, + "grad_norm": 2.4112534460000608, + "learning_rate": 2.914351732556472e-06, + "loss": 0.9787, + "step": 4085 + }, + { + "epoch": 0.3684898768994905, + "grad_norm": 1.6272625943668066, + "learning_rate": 2.9138321104601144e-06, + "loss": 0.9239, + "step": 4086 + }, + { + "epoch": 0.3685800604229607, + "grad_norm": 1.5650859034060345, + "learning_rate": 2.9133124103905623e-06, + "loss": 0.9307, + "step": 4087 + }, + { + "epoch": 0.368670243946431, + "grad_norm": 4.256402970532583, + "learning_rate": 2.9127926323921596e-06, + "loss": 0.9858, + "step": 4088 + }, + { + "epoch": 0.36876042746990123, + "grad_norm": 1.484370101117533, + "learning_rate": 2.912272776509256e-06, + "loss": 1.0161, + "step": 4089 + }, + { + "epoch": 0.3688506109933715, + "grad_norm": 1.7849934376400975, + "learning_rate": 2.911752842786209e-06, + "loss": 0.9673, + "step": 4090 + }, + { + "epoch": 0.36894079451684175, + "grad_norm": 2.097921400876151, + "learning_rate": 2.911232831267383e-06, + "loss": 0.9435, + "step": 4091 + }, + { + "epoch": 0.36903097804031204, + "grad_norm": 1.8341866443637367, + "learning_rate": 2.910712741997146e-06, + "loss": 1.0171, + "step": 4092 + }, + { + "epoch": 0.3691211615637823, + "grad_norm": 1.636129049984669, + "learning_rate": 2.910192575019877e-06, + "loss": 0.987, + "step": 4093 + }, + { + "epoch": 0.36921134508725256, + "grad_norm": 1.4567538654465304, + "learning_rate": 2.9096723303799583e-06, + "loss": 1.0301, + "step": 4094 + }, + { + "epoch": 0.3693015286107228, + "grad_norm": 2.1555671163404067, + "learning_rate": 2.9091520081217805e-06, + "loss": 1.0058, + "step": 4095 + }, + { + "epoch": 0.3693917121341931, + "grad_norm": 1.4390029513850375, + "learning_rate": 2.908631608289741e-06, + "loss": 0.9433, + "step": 4096 + }, + { + "epoch": 0.3694818956576633, + "grad_norm": 3.824938011913776, + "learning_rate": 2.9081111309282423e-06, + "loss": 0.9921, + "step": 4097 + }, + { + "epoch": 0.3695720791811336, + "grad_norm": 0.6787219588667863, + "learning_rate": 2.9075905760816942e-06, + "loss": 0.8693, + "step": 4098 + }, + { + "epoch": 0.3696622627046039, + "grad_norm": 2.954417758129473, + "learning_rate": 2.907069943794514e-06, + "loss": 1.0374, + "step": 4099 + }, + { + "epoch": 0.3697524462280741, + "grad_norm": 1.761213411989419, + "learning_rate": 2.906549234111125e-06, + "loss": 0.9731, + "step": 4100 + }, + { + "epoch": 0.3698426297515444, + "grad_norm": 1.453580231270023, + "learning_rate": 2.906028447075956e-06, + "loss": 1.0525, + "step": 4101 + }, + { + "epoch": 0.36993281327501465, + "grad_norm": 1.5743495233707383, + "learning_rate": 2.905507582733445e-06, + "loss": 1.0791, + "step": 4102 + }, + { + "epoch": 0.37002299679848494, + "grad_norm": 1.6944928275255107, + "learning_rate": 2.904986641128033e-06, + "loss": 1.1166, + "step": 4103 + }, + { + "epoch": 0.37011318032195517, + "grad_norm": 1.8153602131256645, + "learning_rate": 2.9044656223041716e-06, + "loss": 0.9868, + "step": 4104 + }, + { + "epoch": 0.37020336384542546, + "grad_norm": 1.5218838325259847, + "learning_rate": 2.9039445263063157e-06, + "loss": 0.9561, + "step": 4105 + }, + { + "epoch": 0.3702935473688957, + "grad_norm": 2.0077502764251713, + "learning_rate": 2.903423353178929e-06, + "loss": 0.8966, + "step": 4106 + }, + { + "epoch": 0.370383730892366, + "grad_norm": 1.52250079187833, + "learning_rate": 2.9029021029664802e-06, + "loss": 0.9926, + "step": 4107 + }, + { + "epoch": 0.3704739144158362, + "grad_norm": 1.4207799698251133, + "learning_rate": 2.9023807757134455e-06, + "loss": 1.0519, + "step": 4108 + }, + { + "epoch": 0.3705640979393065, + "grad_norm": 1.5620450692702168, + "learning_rate": 2.901859371464307e-06, + "loss": 1.0663, + "step": 4109 + }, + { + "epoch": 0.37065428146277674, + "grad_norm": 1.8732276168986817, + "learning_rate": 2.9013378902635535e-06, + "loss": 1.0344, + "step": 4110 + }, + { + "epoch": 0.370744464986247, + "grad_norm": 1.7599017802921944, + "learning_rate": 2.9008163321556823e-06, + "loss": 0.9695, + "step": 4111 + }, + { + "epoch": 0.37083464850971726, + "grad_norm": 1.9120155443932307, + "learning_rate": 2.900294697185194e-06, + "loss": 1.0097, + "step": 4112 + }, + { + "epoch": 0.37092483203318755, + "grad_norm": 1.567029034864705, + "learning_rate": 2.899772985396599e-06, + "loss": 1.0146, + "step": 4113 + }, + { + "epoch": 0.3710150155566578, + "grad_norm": 1.5929554473336476, + "learning_rate": 2.8992511968344104e-06, + "loss": 0.9882, + "step": 4114 + }, + { + "epoch": 0.37110519908012807, + "grad_norm": 1.6505920648376464, + "learning_rate": 2.8987293315431523e-06, + "loss": 0.9393, + "step": 4115 + }, + { + "epoch": 0.3711953826035983, + "grad_norm": 1.4015802201561707, + "learning_rate": 2.898207389567351e-06, + "loss": 0.9297, + "step": 4116 + }, + { + "epoch": 0.3712855661270686, + "grad_norm": 5.073756196139721, + "learning_rate": 2.897685370951543e-06, + "loss": 0.9899, + "step": 4117 + }, + { + "epoch": 0.3713757496505388, + "grad_norm": 3.676718516348772, + "learning_rate": 2.89716327574027e-06, + "loss": 0.8906, + "step": 4118 + }, + { + "epoch": 0.3714659331740091, + "grad_norm": 2.2941177322853727, + "learning_rate": 2.8966411039780787e-06, + "loss": 1.0883, + "step": 4119 + }, + { + "epoch": 0.37155611669747934, + "grad_norm": 1.5612826373412758, + "learning_rate": 2.8961188557095248e-06, + "loss": 1.0034, + "step": 4120 + }, + { + "epoch": 0.37164630022094963, + "grad_norm": 1.744083621470146, + "learning_rate": 2.895596530979168e-06, + "loss": 0.9455, + "step": 4121 + }, + { + "epoch": 0.3717364837444199, + "grad_norm": 1.5585717986050114, + "learning_rate": 2.895074129831578e-06, + "loss": 1.01, + "step": 4122 + }, + { + "epoch": 0.37182666726789015, + "grad_norm": 1.6362215074312572, + "learning_rate": 2.8945516523113275e-06, + "loss": 0.9131, + "step": 4123 + }, + { + "epoch": 0.37191685079136044, + "grad_norm": 0.678264509520404, + "learning_rate": 2.894029098462998e-06, + "loss": 0.7893, + "step": 4124 + }, + { + "epoch": 0.3720070343148307, + "grad_norm": 1.5582444796330532, + "learning_rate": 2.8935064683311756e-06, + "loss": 1.0194, + "step": 4125 + }, + { + "epoch": 0.37209721783830096, + "grad_norm": 0.679278261479177, + "learning_rate": 2.8929837619604544e-06, + "loss": 0.8931, + "step": 4126 + }, + { + "epoch": 0.3721874013617712, + "grad_norm": 1.8993627860005122, + "learning_rate": 2.8924609793954346e-06, + "loss": 1.0428, + "step": 4127 + }, + { + "epoch": 0.3722775848852415, + "grad_norm": 1.6852183281469315, + "learning_rate": 2.891938120680724e-06, + "loss": 0.9544, + "step": 4128 + }, + { + "epoch": 0.3723677684087117, + "grad_norm": 1.4930985151315574, + "learning_rate": 2.8914151858609343e-06, + "loss": 0.9938, + "step": 4129 + }, + { + "epoch": 0.372457951932182, + "grad_norm": 1.6100833408038562, + "learning_rate": 2.8908921749806858e-06, + "loss": 1.0149, + "step": 4130 + }, + { + "epoch": 0.37254813545565224, + "grad_norm": 1.5296994357865943, + "learning_rate": 2.890369088084605e-06, + "loss": 1.0111, + "step": 4131 + }, + { + "epoch": 0.37263831897912253, + "grad_norm": 1.8396638753963934, + "learning_rate": 2.889845925217323e-06, + "loss": 1.0217, + "step": 4132 + }, + { + "epoch": 0.37272850250259276, + "grad_norm": 1.620423327695745, + "learning_rate": 2.8893226864234813e-06, + "loss": 1.0071, + "step": 4133 + }, + { + "epoch": 0.37281868602606305, + "grad_norm": 1.5420869649565856, + "learning_rate": 2.8887993717477236e-06, + "loss": 1.0053, + "step": 4134 + }, + { + "epoch": 0.3729088695495333, + "grad_norm": 2.365499769799014, + "learning_rate": 2.8882759812347035e-06, + "loss": 0.8929, + "step": 4135 + }, + { + "epoch": 0.3729990530730036, + "grad_norm": 1.6746411124110137, + "learning_rate": 2.887752514929078e-06, + "loss": 0.9993, + "step": 4136 + }, + { + "epoch": 0.3730892365964738, + "grad_norm": 1.6451385898821513, + "learning_rate": 2.887228972875513e-06, + "loss": 0.9511, + "step": 4137 + }, + { + "epoch": 0.3731794201199441, + "grad_norm": 1.8181283314380305, + "learning_rate": 2.88670535511868e-06, + "loss": 1.0002, + "step": 4138 + }, + { + "epoch": 0.3732696036434143, + "grad_norm": 1.6734590245847405, + "learning_rate": 2.886181661703257e-06, + "loss": 0.9446, + "step": 4139 + }, + { + "epoch": 0.3733597871668846, + "grad_norm": 1.8859141906966554, + "learning_rate": 2.8856578926739285e-06, + "loss": 0.9657, + "step": 4140 + }, + { + "epoch": 0.37344997069035485, + "grad_norm": 1.934226007840063, + "learning_rate": 2.8851340480753846e-06, + "loss": 0.9586, + "step": 4141 + }, + { + "epoch": 0.37354015421382514, + "grad_norm": 2.051003638425638, + "learning_rate": 2.8846101279523232e-06, + "loss": 1.0, + "step": 4142 + }, + { + "epoch": 0.37363033773729537, + "grad_norm": 1.6179023673799156, + "learning_rate": 2.8840861323494487e-06, + "loss": 0.9717, + "step": 4143 + }, + { + "epoch": 0.37372052126076566, + "grad_norm": 2.965963401018158, + "learning_rate": 2.88356206131147e-06, + "loss": 1.0566, + "step": 4144 + }, + { + "epoch": 0.37381070478423595, + "grad_norm": 1.6547114585748526, + "learning_rate": 2.883037914883104e-06, + "loss": 0.9825, + "step": 4145 + }, + { + "epoch": 0.3739008883077062, + "grad_norm": 1.5588629545080286, + "learning_rate": 2.882513693109075e-06, + "loss": 0.9633, + "step": 4146 + }, + { + "epoch": 0.37399107183117647, + "grad_norm": 1.2904343674385632, + "learning_rate": 2.8819893960341106e-06, + "loss": 0.9759, + "step": 4147 + }, + { + "epoch": 0.3740812553546467, + "grad_norm": 1.8834446522494783, + "learning_rate": 2.881465023702948e-06, + "loss": 1.058, + "step": 4148 + }, + { + "epoch": 0.374171438878117, + "grad_norm": 1.603376895976282, + "learning_rate": 2.8809405761603294e-06, + "loss": 1.0779, + "step": 4149 + }, + { + "epoch": 0.3742616224015872, + "grad_norm": 2.248727332568681, + "learning_rate": 2.880416053451003e-06, + "loss": 0.9207, + "step": 4150 + }, + { + "epoch": 0.3743518059250575, + "grad_norm": 1.7557688043677246, + "learning_rate": 2.879891455619725e-06, + "loss": 1.0854, + "step": 4151 + }, + { + "epoch": 0.37444198944852775, + "grad_norm": 1.553358343697988, + "learning_rate": 2.879366782711256e-06, + "loss": 1.0097, + "step": 4152 + }, + { + "epoch": 0.37453217297199803, + "grad_norm": 1.4414060018895873, + "learning_rate": 2.8788420347703643e-06, + "loss": 1.0193, + "step": 4153 + }, + { + "epoch": 0.37462235649546827, + "grad_norm": 0.6258922168960858, + "learning_rate": 2.8783172118418244e-06, + "loss": 0.8144, + "step": 4154 + }, + { + "epoch": 0.37471254001893856, + "grad_norm": 1.5575141705255653, + "learning_rate": 2.877792313970417e-06, + "loss": 1.0097, + "step": 4155 + }, + { + "epoch": 0.3748027235424088, + "grad_norm": 1.4282281650534785, + "learning_rate": 2.8772673412009293e-06, + "loss": 1.0568, + "step": 4156 + }, + { + "epoch": 0.3748929070658791, + "grad_norm": 1.7187491677022133, + "learning_rate": 2.8767422935781545e-06, + "loss": 1.0452, + "step": 4157 + }, + { + "epoch": 0.3749830905893493, + "grad_norm": 1.9932434275561959, + "learning_rate": 2.8762171711468935e-06, + "loss": 1.036, + "step": 4158 + }, + { + "epoch": 0.3750732741128196, + "grad_norm": 1.4443478225930881, + "learning_rate": 2.875691973951952e-06, + "loss": 1.0017, + "step": 4159 + }, + { + "epoch": 0.37516345763628983, + "grad_norm": 0.6772225505804417, + "learning_rate": 2.8751667020381425e-06, + "loss": 0.8857, + "step": 4160 + }, + { + "epoch": 0.3752536411597601, + "grad_norm": 1.4956280254838745, + "learning_rate": 2.8746413554502837e-06, + "loss": 1.0659, + "step": 4161 + }, + { + "epoch": 0.37534382468323035, + "grad_norm": 1.502162566544485, + "learning_rate": 2.8741159342332027e-06, + "loss": 0.937, + "step": 4162 + }, + { + "epoch": 0.37543400820670064, + "grad_norm": 1.7810566445862903, + "learning_rate": 2.87359043843173e-06, + "loss": 1.0563, + "step": 4163 + }, + { + "epoch": 0.3755241917301709, + "grad_norm": 1.5965320763906417, + "learning_rate": 2.873064868090704e-06, + "loss": 1.037, + "step": 4164 + }, + { + "epoch": 0.37561437525364116, + "grad_norm": 2.032825504528205, + "learning_rate": 2.8725392232549697e-06, + "loss": 1.0903, + "step": 4165 + }, + { + "epoch": 0.3757045587771114, + "grad_norm": 1.9148823130727124, + "learning_rate": 2.872013503969378e-06, + "loss": 1.0734, + "step": 4166 + }, + { + "epoch": 0.3757947423005817, + "grad_norm": 1.5116312164277237, + "learning_rate": 2.8714877102787853e-06, + "loss": 1.0107, + "step": 4167 + }, + { + "epoch": 0.3758849258240519, + "grad_norm": 1.6661454180372826, + "learning_rate": 2.8709618422280564e-06, + "loss": 1.0092, + "step": 4168 + }, + { + "epoch": 0.3759751093475222, + "grad_norm": 1.5371713395576931, + "learning_rate": 2.8704358998620605e-06, + "loss": 0.9787, + "step": 4169 + }, + { + "epoch": 0.3760652928709925, + "grad_norm": 1.7232551485153804, + "learning_rate": 2.8699098832256735e-06, + "loss": 1.0307, + "step": 4170 + }, + { + "epoch": 0.37615547639446273, + "grad_norm": 1.6741503624559684, + "learning_rate": 2.86938379236378e-06, + "loss": 0.9158, + "step": 4171 + }, + { + "epoch": 0.376245659917933, + "grad_norm": 1.5152932030001156, + "learning_rate": 2.868857627321266e-06, + "loss": 0.9277, + "step": 4172 + }, + { + "epoch": 0.37633584344140325, + "grad_norm": 2.0594494541352995, + "learning_rate": 2.8683313881430296e-06, + "loss": 1.0062, + "step": 4173 + }, + { + "epoch": 0.37642602696487354, + "grad_norm": 1.6037245136950473, + "learning_rate": 2.8678050748739706e-06, + "loss": 0.9929, + "step": 4174 + }, + { + "epoch": 0.37651621048834377, + "grad_norm": 1.4998991455505, + "learning_rate": 2.8672786875589976e-06, + "loss": 0.9743, + "step": 4175 + }, + { + "epoch": 0.37660639401181406, + "grad_norm": 1.6708547027946363, + "learning_rate": 2.866752226243025e-06, + "loss": 0.9533, + "step": 4176 + }, + { + "epoch": 0.3766965775352843, + "grad_norm": 1.684976952165947, + "learning_rate": 2.8662256909709733e-06, + "loss": 1.044, + "step": 4177 + }, + { + "epoch": 0.3767867610587546, + "grad_norm": 1.6898310067820195, + "learning_rate": 2.865699081787769e-06, + "loss": 1.0405, + "step": 4178 + }, + { + "epoch": 0.3768769445822248, + "grad_norm": 3.869484698860663, + "learning_rate": 2.8651723987383465e-06, + "loss": 0.9887, + "step": 4179 + }, + { + "epoch": 0.3769671281056951, + "grad_norm": 1.7488840495956426, + "learning_rate": 2.8646456418676437e-06, + "loss": 0.9774, + "step": 4180 + }, + { + "epoch": 0.37705731162916534, + "grad_norm": 0.6955556337262917, + "learning_rate": 2.8641188112206067e-06, + "loss": 0.834, + "step": 4181 + }, + { + "epoch": 0.3771474951526356, + "grad_norm": 2.117024433007605, + "learning_rate": 2.863591906842189e-06, + "loss": 1.0335, + "step": 4182 + }, + { + "epoch": 0.37723767867610586, + "grad_norm": 1.8318530160709956, + "learning_rate": 2.863064928777347e-06, + "loss": 1.0331, + "step": 4183 + }, + { + "epoch": 0.37732786219957615, + "grad_norm": 1.6398491614453723, + "learning_rate": 2.862537877071047e-06, + "loss": 0.9788, + "step": 4184 + }, + { + "epoch": 0.3774180457230464, + "grad_norm": 0.6389282365329118, + "learning_rate": 2.8620107517682597e-06, + "loss": 0.7978, + "step": 4185 + }, + { + "epoch": 0.37750822924651667, + "grad_norm": 1.3895894170091285, + "learning_rate": 2.8614835529139618e-06, + "loss": 0.9349, + "step": 4186 + }, + { + "epoch": 0.3775984127699869, + "grad_norm": 1.6299296944322827, + "learning_rate": 2.8609562805531367e-06, + "loss": 0.9613, + "step": 4187 + }, + { + "epoch": 0.3776885962934572, + "grad_norm": 1.6015357782879385, + "learning_rate": 2.8604289347307746e-06, + "loss": 1.0376, + "step": 4188 + }, + { + "epoch": 0.3777787798169274, + "grad_norm": 1.6345537747989054, + "learning_rate": 2.859901515491871e-06, + "loss": 0.9959, + "step": 4189 + }, + { + "epoch": 0.3778689633403977, + "grad_norm": 1.5788505321639628, + "learning_rate": 2.8593740228814298e-06, + "loss": 0.8899, + "step": 4190 + }, + { + "epoch": 0.37795914686386795, + "grad_norm": 2.673846047608167, + "learning_rate": 2.8588464569444574e-06, + "loss": 0.9895, + "step": 4191 + }, + { + "epoch": 0.37804933038733823, + "grad_norm": 1.5902999485845102, + "learning_rate": 2.8583188177259697e-06, + "loss": 0.9904, + "step": 4192 + }, + { + "epoch": 0.3781395139108085, + "grad_norm": 2.0312349759059756, + "learning_rate": 2.857791105270988e-06, + "loss": 0.9885, + "step": 4193 + }, + { + "epoch": 0.37822969743427876, + "grad_norm": 1.634152898047979, + "learning_rate": 2.857263319624539e-06, + "loss": 1.0285, + "step": 4194 + }, + { + "epoch": 0.37831988095774904, + "grad_norm": 1.828742029116632, + "learning_rate": 2.856735460831657e-06, + "loss": 0.9847, + "step": 4195 + }, + { + "epoch": 0.3784100644812193, + "grad_norm": 1.6649961762484613, + "learning_rate": 2.856207528937382e-06, + "loss": 1.0098, + "step": 4196 + }, + { + "epoch": 0.37850024800468957, + "grad_norm": 1.6559157124137245, + "learning_rate": 2.855679523986759e-06, + "loss": 0.9935, + "step": 4197 + }, + { + "epoch": 0.3785904315281598, + "grad_norm": 2.809434873560314, + "learning_rate": 2.8551514460248406e-06, + "loss": 1.0471, + "step": 4198 + }, + { + "epoch": 0.3786806150516301, + "grad_norm": 6.355947409660123, + "learning_rate": 2.8546232950966868e-06, + "loss": 1.0059, + "step": 4199 + }, + { + "epoch": 0.3787707985751003, + "grad_norm": 1.8758807339039525, + "learning_rate": 2.85409507124736e-06, + "loss": 0.9493, + "step": 4200 + }, + { + "epoch": 0.3788609820985706, + "grad_norm": 1.5125271313968784, + "learning_rate": 2.8535667745219324e-06, + "loss": 0.9968, + "step": 4201 + }, + { + "epoch": 0.37895116562204084, + "grad_norm": 1.6171095087530591, + "learning_rate": 2.853038404965481e-06, + "loss": 0.9896, + "step": 4202 + }, + { + "epoch": 0.37904134914551113, + "grad_norm": 4.47477582657725, + "learning_rate": 2.8525099626230894e-06, + "loss": 1.0738, + "step": 4203 + }, + { + "epoch": 0.37913153266898136, + "grad_norm": 1.7479796327412604, + "learning_rate": 2.8519814475398472e-06, + "loss": 0.9347, + "step": 4204 + }, + { + "epoch": 0.37922171619245165, + "grad_norm": 1.3649319598947824, + "learning_rate": 2.8514528597608502e-06, + "loss": 0.9791, + "step": 4205 + }, + { + "epoch": 0.3793118997159219, + "grad_norm": 1.2838053922950898, + "learning_rate": 2.8509241993312004e-06, + "loss": 1.0311, + "step": 4206 + }, + { + "epoch": 0.3794020832393922, + "grad_norm": 1.5784363392319807, + "learning_rate": 2.850395466296006e-06, + "loss": 1.0528, + "step": 4207 + }, + { + "epoch": 0.3794922667628624, + "grad_norm": 2.080095088692565, + "learning_rate": 2.849866660700381e-06, + "loss": 1.0279, + "step": 4208 + }, + { + "epoch": 0.3795824502863327, + "grad_norm": 1.6713223123135055, + "learning_rate": 2.8493377825894464e-06, + "loss": 1.059, + "step": 4209 + }, + { + "epoch": 0.37967263380980293, + "grad_norm": 1.981025090974816, + "learning_rate": 2.848808832008329e-06, + "loss": 1.0639, + "step": 4210 + }, + { + "epoch": 0.3797628173332732, + "grad_norm": 2.1323559171025246, + "learning_rate": 2.848279809002162e-06, + "loss": 1.0324, + "step": 4211 + }, + { + "epoch": 0.37985300085674345, + "grad_norm": 1.7335471162029197, + "learning_rate": 2.8477507136160842e-06, + "loss": 1.0165, + "step": 4212 + }, + { + "epoch": 0.37994318438021374, + "grad_norm": 0.6425939865241729, + "learning_rate": 2.847221545895241e-06, + "loss": 0.8267, + "step": 4213 + }, + { + "epoch": 0.38003336790368397, + "grad_norm": 1.5970012935855336, + "learning_rate": 2.846692305884785e-06, + "loss": 0.9945, + "step": 4214 + }, + { + "epoch": 0.38012355142715426, + "grad_norm": 1.5787068418038854, + "learning_rate": 2.8461629936298718e-06, + "loss": 1.0477, + "step": 4215 + }, + { + "epoch": 0.3802137349506245, + "grad_norm": 1.9423481366115132, + "learning_rate": 2.845633609175666e-06, + "loss": 1.0608, + "step": 4216 + }, + { + "epoch": 0.3803039184740948, + "grad_norm": 2.3264276858711708, + "learning_rate": 2.8451041525673383e-06, + "loss": 1.0414, + "step": 4217 + }, + { + "epoch": 0.38039410199756507, + "grad_norm": 2.10905047145482, + "learning_rate": 2.8445746238500647e-06, + "loss": 1.0312, + "step": 4218 + }, + { + "epoch": 0.3804842855210353, + "grad_norm": 1.7459113859999544, + "learning_rate": 2.844045023069027e-06, + "loss": 0.9862, + "step": 4219 + }, + { + "epoch": 0.3805744690445056, + "grad_norm": 1.6465808120667822, + "learning_rate": 2.8435153502694136e-06, + "loss": 1.0416, + "step": 4220 + }, + { + "epoch": 0.3806646525679758, + "grad_norm": 1.8328335109589313, + "learning_rate": 2.84298560549642e-06, + "loss": 0.94, + "step": 4221 + }, + { + "epoch": 0.3807548360914461, + "grad_norm": 2.4346981083899917, + "learning_rate": 2.8424557887952462e-06, + "loss": 1.0283, + "step": 4222 + }, + { + "epoch": 0.38084501961491635, + "grad_norm": 2.1819059506015623, + "learning_rate": 2.841925900211099e-06, + "loss": 0.9331, + "step": 4223 + }, + { + "epoch": 0.38093520313838664, + "grad_norm": 1.5208880096779571, + "learning_rate": 2.841395939789192e-06, + "loss": 1.0565, + "step": 4224 + }, + { + "epoch": 0.38102538666185687, + "grad_norm": 1.7392040760578587, + "learning_rate": 2.8408659075747435e-06, + "loss": 1.0219, + "step": 4225 + }, + { + "epoch": 0.38111557018532716, + "grad_norm": 1.8756170529048137, + "learning_rate": 2.8403358036129796e-06, + "loss": 1.0032, + "step": 4226 + }, + { + "epoch": 0.3812057537087974, + "grad_norm": 1.7160101814279198, + "learning_rate": 2.839805627949132e-06, + "loss": 1.0597, + "step": 4227 + }, + { + "epoch": 0.3812959372322677, + "grad_norm": 1.9312401829936068, + "learning_rate": 2.8392753806284367e-06, + "loss": 1.063, + "step": 4228 + }, + { + "epoch": 0.3813861207557379, + "grad_norm": 1.8597419440803546, + "learning_rate": 2.838745061696139e-06, + "loss": 0.9563, + "step": 4229 + }, + { + "epoch": 0.3814763042792082, + "grad_norm": 1.5138168560860132, + "learning_rate": 2.838214671197487e-06, + "loss": 0.9439, + "step": 4230 + }, + { + "epoch": 0.38156648780267843, + "grad_norm": 1.7320170139111397, + "learning_rate": 2.8376842091777377e-06, + "loss": 1.1124, + "step": 4231 + }, + { + "epoch": 0.3816566713261487, + "grad_norm": 1.6679069990033073, + "learning_rate": 2.8371536756821524e-06, + "loss": 0.913, + "step": 4232 + }, + { + "epoch": 0.38174685484961896, + "grad_norm": 1.6796564143653319, + "learning_rate": 2.836623070756e-06, + "loss": 0.9883, + "step": 4233 + }, + { + "epoch": 0.38183703837308924, + "grad_norm": 1.9256599186867145, + "learning_rate": 2.8360923944445542e-06, + "loss": 1.0322, + "step": 4234 + }, + { + "epoch": 0.3819272218965595, + "grad_norm": 1.4126657152444235, + "learning_rate": 2.8355616467930947e-06, + "loss": 0.9893, + "step": 4235 + }, + { + "epoch": 0.38201740542002977, + "grad_norm": 2.286428661342178, + "learning_rate": 2.8350308278469085e-06, + "loss": 0.9842, + "step": 4236 + }, + { + "epoch": 0.3821075889435, + "grad_norm": 1.8046250105438582, + "learning_rate": 2.8344999376512877e-06, + "loss": 1.0408, + "step": 4237 + }, + { + "epoch": 0.3821977724669703, + "grad_norm": 1.60580084491536, + "learning_rate": 2.8339689762515307e-06, + "loss": 0.9974, + "step": 4238 + }, + { + "epoch": 0.3822879559904405, + "grad_norm": 1.9413143266679271, + "learning_rate": 2.8334379436929424e-06, + "loss": 0.939, + "step": 4239 + }, + { + "epoch": 0.3823781395139108, + "grad_norm": 1.8548154945846813, + "learning_rate": 2.832906840020833e-06, + "loss": 0.9753, + "step": 4240 + }, + { + "epoch": 0.3824683230373811, + "grad_norm": 1.639783952485094, + "learning_rate": 2.83237566528052e-06, + "loss": 1.0174, + "step": 4241 + }, + { + "epoch": 0.38255850656085133, + "grad_norm": 1.6442056727757077, + "learning_rate": 2.831844419517325e-06, + "loss": 0.9859, + "step": 4242 + }, + { + "epoch": 0.3826486900843216, + "grad_norm": 1.591530544724945, + "learning_rate": 2.8313131027765774e-06, + "loss": 1.0374, + "step": 4243 + }, + { + "epoch": 0.38273887360779185, + "grad_norm": 1.7967383954740406, + "learning_rate": 2.8307817151036124e-06, + "loss": 0.9799, + "step": 4244 + }, + { + "epoch": 0.38282905713126214, + "grad_norm": 1.7120093206526257, + "learning_rate": 2.8302502565437704e-06, + "loss": 0.9842, + "step": 4245 + }, + { + "epoch": 0.3829192406547324, + "grad_norm": 2.1540937838101346, + "learning_rate": 2.829718727142398e-06, + "loss": 0.9359, + "step": 4246 + }, + { + "epoch": 0.38300942417820266, + "grad_norm": 1.7789561576694446, + "learning_rate": 2.829187126944849e-06, + "loss": 1.0147, + "step": 4247 + }, + { + "epoch": 0.3830996077016729, + "grad_norm": 3.1412008336474013, + "learning_rate": 2.8286554559964826e-06, + "loss": 0.9465, + "step": 4248 + }, + { + "epoch": 0.3831897912251432, + "grad_norm": 1.725257276616148, + "learning_rate": 2.8281237143426637e-06, + "loss": 1.0253, + "step": 4249 + }, + { + "epoch": 0.3832799747486134, + "grad_norm": 3.411844681079081, + "learning_rate": 2.8275919020287626e-06, + "loss": 0.9093, + "step": 4250 + }, + { + "epoch": 0.3833701582720837, + "grad_norm": 1.7731606691717223, + "learning_rate": 2.827060019100158e-06, + "loss": 1.0326, + "step": 4251 + }, + { + "epoch": 0.38346034179555394, + "grad_norm": 1.5732617368140602, + "learning_rate": 2.8265280656022315e-06, + "loss": 1.0018, + "step": 4252 + }, + { + "epoch": 0.3835505253190242, + "grad_norm": 1.691376436953082, + "learning_rate": 2.825996041580373e-06, + "loss": 0.8922, + "step": 4253 + }, + { + "epoch": 0.38364070884249446, + "grad_norm": 1.5518664351437061, + "learning_rate": 2.825463947079978e-06, + "loss": 1.0279, + "step": 4254 + }, + { + "epoch": 0.38373089236596475, + "grad_norm": 1.6076427878032045, + "learning_rate": 2.8249317821464483e-06, + "loss": 1.0096, + "step": 4255 + }, + { + "epoch": 0.383821075889435, + "grad_norm": 1.5312342740243967, + "learning_rate": 2.824399546825189e-06, + "loss": 0.9381, + "step": 4256 + }, + { + "epoch": 0.38391125941290527, + "grad_norm": 2.5598502066120488, + "learning_rate": 2.823867241161616e-06, + "loss": 0.9652, + "step": 4257 + }, + { + "epoch": 0.3840014429363755, + "grad_norm": 1.6740363580180415, + "learning_rate": 2.8233348652011456e-06, + "loss": 1.0574, + "step": 4258 + }, + { + "epoch": 0.3840916264598458, + "grad_norm": 2.3777563262974013, + "learning_rate": 2.8228024189892057e-06, + "loss": 0.9214, + "step": 4259 + }, + { + "epoch": 0.384181809983316, + "grad_norm": 1.3459316552369394, + "learning_rate": 2.822269902571226e-06, + "loss": 1.0115, + "step": 4260 + }, + { + "epoch": 0.3842719935067863, + "grad_norm": 1.5141081937935668, + "learning_rate": 2.8217373159926446e-06, + "loss": 0.9863, + "step": 4261 + }, + { + "epoch": 0.38436217703025655, + "grad_norm": 1.3325023545178194, + "learning_rate": 2.8212046592989046e-06, + "loss": 0.9447, + "step": 4262 + }, + { + "epoch": 0.38445236055372684, + "grad_norm": 5.746561432042324, + "learning_rate": 2.820671932535455e-06, + "loss": 1.0138, + "step": 4263 + }, + { + "epoch": 0.3845425440771971, + "grad_norm": 2.078292839901887, + "learning_rate": 2.8201391357477506e-06, + "loss": 0.9883, + "step": 4264 + }, + { + "epoch": 0.38463272760066736, + "grad_norm": 1.3988391150346693, + "learning_rate": 2.8196062689812525e-06, + "loss": 0.9812, + "step": 4265 + }, + { + "epoch": 0.38472291112413765, + "grad_norm": 2.5010526348372926, + "learning_rate": 2.819073332281429e-06, + "loss": 0.9508, + "step": 4266 + }, + { + "epoch": 0.3848130946476079, + "grad_norm": 0.6158448354458217, + "learning_rate": 2.8185403256937524e-06, + "loss": 0.7875, + "step": 4267 + }, + { + "epoch": 0.38490327817107817, + "grad_norm": 1.7714987757390825, + "learning_rate": 2.8180072492637016e-06, + "loss": 0.9692, + "step": 4268 + }, + { + "epoch": 0.3849934616945484, + "grad_norm": 1.550589215666408, + "learning_rate": 2.817474103036762e-06, + "loss": 0.9988, + "step": 4269 + }, + { + "epoch": 0.3850836452180187, + "grad_norm": 1.815437698551784, + "learning_rate": 2.816940887058425e-06, + "loss": 1.0322, + "step": 4270 + }, + { + "epoch": 0.3851738287414889, + "grad_norm": 1.6048921538154575, + "learning_rate": 2.816407601374186e-06, + "loss": 1.0456, + "step": 4271 + }, + { + "epoch": 0.3852640122649592, + "grad_norm": 1.6694587288342542, + "learning_rate": 2.815874246029549e-06, + "loss": 0.9367, + "step": 4272 + }, + { + "epoch": 0.38535419578842944, + "grad_norm": 1.6262867308096365, + "learning_rate": 2.815340821070023e-06, + "loss": 0.9562, + "step": 4273 + }, + { + "epoch": 0.38544437931189973, + "grad_norm": 1.8574687257908973, + "learning_rate": 2.814807326541122e-06, + "loss": 0.9859, + "step": 4274 + }, + { + "epoch": 0.38553456283536996, + "grad_norm": 1.5250836990400753, + "learning_rate": 2.8142737624883676e-06, + "loss": 0.9391, + "step": 4275 + }, + { + "epoch": 0.38562474635884025, + "grad_norm": 1.8252936714819616, + "learning_rate": 2.8137401289572854e-06, + "loss": 1.0087, + "step": 4276 + }, + { + "epoch": 0.3857149298823105, + "grad_norm": 2.61656568741385, + "learning_rate": 2.8132064259934086e-06, + "loss": 1.05, + "step": 4277 + }, + { + "epoch": 0.3858051134057808, + "grad_norm": 1.6351724025112606, + "learning_rate": 2.812672653642276e-06, + "loss": 1.0031, + "step": 4278 + }, + { + "epoch": 0.385895296929251, + "grad_norm": 1.3987184807001938, + "learning_rate": 2.812138811949431e-06, + "loss": 1.0166, + "step": 4279 + }, + { + "epoch": 0.3859854804527213, + "grad_norm": 1.4288873987221418, + "learning_rate": 2.8116049009604247e-06, + "loss": 1.0706, + "step": 4280 + }, + { + "epoch": 0.38607566397619153, + "grad_norm": 1.5074792207010141, + "learning_rate": 2.8110709207208132e-06, + "loss": 0.8835, + "step": 4281 + }, + { + "epoch": 0.3861658474996618, + "grad_norm": 1.9232002589216073, + "learning_rate": 2.810536871276158e-06, + "loss": 1.0349, + "step": 4282 + }, + { + "epoch": 0.38625603102313205, + "grad_norm": 0.8302506929421685, + "learning_rate": 2.8100027526720283e-06, + "loss": 0.8673, + "step": 4283 + }, + { + "epoch": 0.38634621454660234, + "grad_norm": 1.6521525531649441, + "learning_rate": 2.8094685649539974e-06, + "loss": 0.9309, + "step": 4284 + }, + { + "epoch": 0.3864363980700726, + "grad_norm": 0.7164033646353853, + "learning_rate": 2.8089343081676455e-06, + "loss": 0.8391, + "step": 4285 + }, + { + "epoch": 0.38652658159354286, + "grad_norm": 1.8306791932695483, + "learning_rate": 2.8083999823585577e-06, + "loss": 0.9851, + "step": 4286 + }, + { + "epoch": 0.3866167651170131, + "grad_norm": 1.374503002511773, + "learning_rate": 2.8078655875723254e-06, + "loss": 1.0305, + "step": 4287 + }, + { + "epoch": 0.3867069486404834, + "grad_norm": 2.7686877344849936, + "learning_rate": 2.807331123854547e-06, + "loss": 1.0875, + "step": 4288 + }, + { + "epoch": 0.38679713216395367, + "grad_norm": 2.7205381103774062, + "learning_rate": 2.806796591250826e-06, + "loss": 0.8756, + "step": 4289 + }, + { + "epoch": 0.3868873156874239, + "grad_norm": 2.5579695311425072, + "learning_rate": 2.8062619898067707e-06, + "loss": 1.074, + "step": 4290 + }, + { + "epoch": 0.3869774992108942, + "grad_norm": 1.532836092612777, + "learning_rate": 2.8057273195679963e-06, + "loss": 1.0167, + "step": 4291 + }, + { + "epoch": 0.3870676827343644, + "grad_norm": 1.6106450523043023, + "learning_rate": 2.8051925805801253e-06, + "loss": 1.0216, + "step": 4292 + }, + { + "epoch": 0.3871578662578347, + "grad_norm": 1.5725161145571742, + "learning_rate": 2.804657772888783e-06, + "loss": 1.0348, + "step": 4293 + }, + { + "epoch": 0.38724804978130495, + "grad_norm": 1.833972219928603, + "learning_rate": 2.804122896539602e-06, + "loss": 0.9979, + "step": 4294 + }, + { + "epoch": 0.38733823330477524, + "grad_norm": 2.1565856672324206, + "learning_rate": 2.8035879515782225e-06, + "loss": 0.9515, + "step": 4295 + }, + { + "epoch": 0.38742841682824547, + "grad_norm": 1.8292910777588045, + "learning_rate": 2.803052938050288e-06, + "loss": 1.0539, + "step": 4296 + }, + { + "epoch": 0.38751860035171576, + "grad_norm": 2.588498601932601, + "learning_rate": 2.802517856001449e-06, + "loss": 1.0498, + "step": 4297 + }, + { + "epoch": 0.387608783875186, + "grad_norm": 1.9868668057397834, + "learning_rate": 2.801982705477361e-06, + "loss": 0.9692, + "step": 4298 + }, + { + "epoch": 0.3876989673986563, + "grad_norm": 1.3526093943243562, + "learning_rate": 2.8014474865236867e-06, + "loss": 0.9885, + "step": 4299 + }, + { + "epoch": 0.3877891509221265, + "grad_norm": 2.5250778299554972, + "learning_rate": 2.800912199186094e-06, + "loss": 0.996, + "step": 4300 + }, + { + "epoch": 0.3878793344455968, + "grad_norm": 1.5147589342552992, + "learning_rate": 2.800376843510256e-06, + "loss": 1.013, + "step": 4301 + }, + { + "epoch": 0.38796951796906703, + "grad_norm": 1.5106308445405094, + "learning_rate": 2.799841419541852e-06, + "loss": 1.0243, + "step": 4302 + }, + { + "epoch": 0.3880597014925373, + "grad_norm": 0.633410277419188, + "learning_rate": 2.799305927326568e-06, + "loss": 0.8215, + "step": 4303 + }, + { + "epoch": 0.38814988501600756, + "grad_norm": 0.7149400906706368, + "learning_rate": 2.7987703669100955e-06, + "loss": 0.8313, + "step": 4304 + }, + { + "epoch": 0.38824006853947785, + "grad_norm": 2.343519174017579, + "learning_rate": 2.79823473833813e-06, + "loss": 1.011, + "step": 4305 + }, + { + "epoch": 0.3883302520629481, + "grad_norm": 1.5890517452930064, + "learning_rate": 2.797699041656376e-06, + "loss": 0.9529, + "step": 4306 + }, + { + "epoch": 0.38842043558641837, + "grad_norm": 2.634580567508444, + "learning_rate": 2.7971632769105412e-06, + "loss": 1.013, + "step": 4307 + }, + { + "epoch": 0.3885106191098886, + "grad_norm": 2.1862266103933936, + "learning_rate": 2.79662744414634e-06, + "loss": 0.9666, + "step": 4308 + }, + { + "epoch": 0.3886008026333589, + "grad_norm": 1.5414228590212065, + "learning_rate": 2.7960915434094923e-06, + "loss": 0.9482, + "step": 4309 + }, + { + "epoch": 0.3886909861568291, + "grad_norm": 1.6626053998993677, + "learning_rate": 2.7955555747457256e-06, + "loss": 0.9774, + "step": 4310 + }, + { + "epoch": 0.3887811696802994, + "grad_norm": 1.982202857936954, + "learning_rate": 2.79501953820077e-06, + "loss": 0.9977, + "step": 4311 + }, + { + "epoch": 0.3888713532037697, + "grad_norm": 3.0862061975089223, + "learning_rate": 2.7944834338203637e-06, + "loss": 1.0187, + "step": 4312 + }, + { + "epoch": 0.38896153672723993, + "grad_norm": 1.6079766578023929, + "learning_rate": 2.79394726165025e-06, + "loss": 0.9614, + "step": 4313 + }, + { + "epoch": 0.3890517202507102, + "grad_norm": 2.1119906561962702, + "learning_rate": 2.793411021736178e-06, + "loss": 0.9808, + "step": 4314 + }, + { + "epoch": 0.38914190377418045, + "grad_norm": 1.7383561364662712, + "learning_rate": 2.7928747141239027e-06, + "loss": 1.0931, + "step": 4315 + }, + { + "epoch": 0.38923208729765074, + "grad_norm": 0.6614168309285262, + "learning_rate": 2.7923383388591856e-06, + "loss": 0.8254, + "step": 4316 + }, + { + "epoch": 0.389322270821121, + "grad_norm": 1.5358936974924235, + "learning_rate": 2.7918018959877923e-06, + "loss": 0.9709, + "step": 4317 + }, + { + "epoch": 0.38941245434459126, + "grad_norm": 1.4808251749535613, + "learning_rate": 2.791265385555495e-06, + "loss": 1.0267, + "step": 4318 + }, + { + "epoch": 0.3895026378680615, + "grad_norm": 1.4488110238652712, + "learning_rate": 2.790728807608072e-06, + "loss": 1.0594, + "step": 4319 + }, + { + "epoch": 0.3895928213915318, + "grad_norm": 2.1541041878568303, + "learning_rate": 2.790192162191307e-06, + "loss": 0.9918, + "step": 4320 + }, + { + "epoch": 0.389683004915002, + "grad_norm": 1.9480430099666202, + "learning_rate": 2.78965544935099e-06, + "loss": 1.0296, + "step": 4321 + }, + { + "epoch": 0.3897731884384723, + "grad_norm": 1.774805759493837, + "learning_rate": 2.789118669132916e-06, + "loss": 0.9786, + "step": 4322 + }, + { + "epoch": 0.38986337196194254, + "grad_norm": 1.9911576545025833, + "learning_rate": 2.7885818215828856e-06, + "loss": 1.067, + "step": 4323 + }, + { + "epoch": 0.38995355548541283, + "grad_norm": 2.0968101116208775, + "learning_rate": 2.7880449067467064e-06, + "loss": 0.9563, + "step": 4324 + }, + { + "epoch": 0.39004373900888306, + "grad_norm": 1.853389884454158, + "learning_rate": 2.78750792467019e-06, + "loss": 0.9535, + "step": 4325 + }, + { + "epoch": 0.39013392253235335, + "grad_norm": 1.9329310659124546, + "learning_rate": 2.786970875399156e-06, + "loss": 0.868, + "step": 4326 + }, + { + "epoch": 0.3902241060558236, + "grad_norm": 1.4980018340380943, + "learning_rate": 2.7864337589794267e-06, + "loss": 0.9342, + "step": 4327 + }, + { + "epoch": 0.39031428957929387, + "grad_norm": 1.7907557574558648, + "learning_rate": 2.7858965754568335e-06, + "loss": 1.0155, + "step": 4328 + }, + { + "epoch": 0.3904044731027641, + "grad_norm": 1.7185287333159915, + "learning_rate": 2.785359324877211e-06, + "loss": 0.9351, + "step": 4329 + }, + { + "epoch": 0.3904946566262344, + "grad_norm": 1.5565678712315258, + "learning_rate": 2.7848220072864e-06, + "loss": 1.0267, + "step": 4330 + }, + { + "epoch": 0.3905848401497046, + "grad_norm": 4.405501281758876, + "learning_rate": 2.784284622730248e-06, + "loss": 0.9906, + "step": 4331 + }, + { + "epoch": 0.3906750236731749, + "grad_norm": 1.7741377200155317, + "learning_rate": 2.7837471712546073e-06, + "loss": 1.0186, + "step": 4332 + }, + { + "epoch": 0.39076520719664515, + "grad_norm": 1.2911036197377903, + "learning_rate": 2.783209652905337e-06, + "loss": 0.9746, + "step": 4333 + }, + { + "epoch": 0.39085539072011544, + "grad_norm": 1.5596968874747235, + "learning_rate": 2.7826720677283e-06, + "loss": 0.9614, + "step": 4334 + }, + { + "epoch": 0.39094557424358567, + "grad_norm": 1.526422637498482, + "learning_rate": 2.782134415769367e-06, + "loss": 0.9673, + "step": 4335 + }, + { + "epoch": 0.39103575776705596, + "grad_norm": 1.8757008196497102, + "learning_rate": 2.7815966970744126e-06, + "loss": 1.0464, + "step": 4336 + }, + { + "epoch": 0.39112594129052625, + "grad_norm": 1.8305518840906696, + "learning_rate": 2.7810589116893184e-06, + "loss": 1.0651, + "step": 4337 + }, + { + "epoch": 0.3912161248139965, + "grad_norm": 5.701511436980165, + "learning_rate": 2.780521059659972e-06, + "loss": 1.0016, + "step": 4338 + }, + { + "epoch": 0.39130630833746677, + "grad_norm": 2.250375080634346, + "learning_rate": 2.7799831410322637e-06, + "loss": 0.9827, + "step": 4339 + }, + { + "epoch": 0.391396491860937, + "grad_norm": 1.7476164389088802, + "learning_rate": 2.779445155852094e-06, + "loss": 0.9531, + "step": 4340 + }, + { + "epoch": 0.3914866753844073, + "grad_norm": 0.6476254951712241, + "learning_rate": 2.7789071041653655e-06, + "loss": 0.8635, + "step": 4341 + }, + { + "epoch": 0.3915768589078775, + "grad_norm": 1.7228318728238754, + "learning_rate": 2.7783689860179875e-06, + "loss": 1.0547, + "step": 4342 + }, + { + "epoch": 0.3916670424313478, + "grad_norm": 1.6659710465579407, + "learning_rate": 2.7778308014558767e-06, + "loss": 0.9358, + "step": 4343 + }, + { + "epoch": 0.39175722595481804, + "grad_norm": 2.147569182661557, + "learning_rate": 2.7772925505249524e-06, + "loss": 1.0716, + "step": 4344 + }, + { + "epoch": 0.39184740947828833, + "grad_norm": 1.501660699243293, + "learning_rate": 2.7767542332711417e-06, + "loss": 0.9667, + "step": 4345 + }, + { + "epoch": 0.39193759300175857, + "grad_norm": 1.4435012603109452, + "learning_rate": 2.776215849740377e-06, + "loss": 0.9708, + "step": 4346 + }, + { + "epoch": 0.39202777652522885, + "grad_norm": 1.4736526316023528, + "learning_rate": 2.775677399978596e-06, + "loss": 1.0121, + "step": 4347 + }, + { + "epoch": 0.3921179600486991, + "grad_norm": 0.5837270112177293, + "learning_rate": 2.775138884031742e-06, + "loss": 0.8108, + "step": 4348 + }, + { + "epoch": 0.3922081435721694, + "grad_norm": 1.4364960523286212, + "learning_rate": 2.774600301945764e-06, + "loss": 0.9569, + "step": 4349 + }, + { + "epoch": 0.3922983270956396, + "grad_norm": 1.5258657030678775, + "learning_rate": 2.774061653766618e-06, + "loss": 1.0293, + "step": 4350 + }, + { + "epoch": 0.3923885106191099, + "grad_norm": 1.7055581587082296, + "learning_rate": 2.773522939540263e-06, + "loss": 0.9875, + "step": 4351 + }, + { + "epoch": 0.39247869414258013, + "grad_norm": 2.4058206348548734, + "learning_rate": 2.7729841593126663e-06, + "loss": 0.9213, + "step": 4352 + }, + { + "epoch": 0.3925688776660504, + "grad_norm": 1.4215996444186494, + "learning_rate": 2.7724453131297988e-06, + "loss": 1.0594, + "step": 4353 + }, + { + "epoch": 0.39265906118952065, + "grad_norm": 1.8075344864337213, + "learning_rate": 2.771906401037637e-06, + "loss": 1.0232, + "step": 4354 + }, + { + "epoch": 0.39274924471299094, + "grad_norm": 2.5406391589156123, + "learning_rate": 2.7713674230821664e-06, + "loss": 0.9563, + "step": 4355 + }, + { + "epoch": 0.3928394282364612, + "grad_norm": 1.424608223582274, + "learning_rate": 2.7708283793093724e-06, + "loss": 0.9524, + "step": 4356 + }, + { + "epoch": 0.39292961175993146, + "grad_norm": 1.7712400829438684, + "learning_rate": 2.7702892697652514e-06, + "loss": 0.9697, + "step": 4357 + }, + { + "epoch": 0.3930197952834017, + "grad_norm": 1.5235336664609567, + "learning_rate": 2.7697500944958024e-06, + "loss": 0.9785, + "step": 4358 + }, + { + "epoch": 0.393109978806872, + "grad_norm": 2.007201461731556, + "learning_rate": 2.7692108535470312e-06, + "loss": 1.1011, + "step": 4359 + }, + { + "epoch": 0.3932001623303423, + "grad_norm": 2.005516788158466, + "learning_rate": 2.768671546964948e-06, + "loss": 1.0347, + "step": 4360 + }, + { + "epoch": 0.3932903458538125, + "grad_norm": 1.7200460229305334, + "learning_rate": 2.7681321747955713e-06, + "loss": 1.025, + "step": 4361 + }, + { + "epoch": 0.3933805293772828, + "grad_norm": 1.61060604677902, + "learning_rate": 2.767592737084921e-06, + "loss": 1.0163, + "step": 4362 + }, + { + "epoch": 0.39347071290075303, + "grad_norm": 1.9591759296145796, + "learning_rate": 2.767053233879026e-06, + "loss": 1.0895, + "step": 4363 + }, + { + "epoch": 0.3935608964242233, + "grad_norm": 1.7241304970449465, + "learning_rate": 2.76651366522392e-06, + "loss": 1.0718, + "step": 4364 + }, + { + "epoch": 0.39365107994769355, + "grad_norm": 1.6903404066815078, + "learning_rate": 2.7659740311656413e-06, + "loss": 0.9657, + "step": 4365 + }, + { + "epoch": 0.39374126347116384, + "grad_norm": 1.8878071490854307, + "learning_rate": 2.7654343317502352e-06, + "loss": 1.0198, + "step": 4366 + }, + { + "epoch": 0.39383144699463407, + "grad_norm": 1.5456019856505345, + "learning_rate": 2.7648945670237502e-06, + "loss": 0.9797, + "step": 4367 + }, + { + "epoch": 0.39392163051810436, + "grad_norm": 1.8833083137043105, + "learning_rate": 2.7643547370322446e-06, + "loss": 1.051, + "step": 4368 + }, + { + "epoch": 0.3940118140415746, + "grad_norm": 1.7430833642910382, + "learning_rate": 2.7638148418217775e-06, + "loss": 1.0329, + "step": 4369 + }, + { + "epoch": 0.3941019975650449, + "grad_norm": 1.656750225605337, + "learning_rate": 2.7632748814384163e-06, + "loss": 1.0414, + "step": 4370 + }, + { + "epoch": 0.3941921810885151, + "grad_norm": 1.3484305106431649, + "learning_rate": 2.7627348559282335e-06, + "loss": 1.0254, + "step": 4371 + }, + { + "epoch": 0.3942823646119854, + "grad_norm": 1.509968298648748, + "learning_rate": 2.7621947653373075e-06, + "loss": 0.937, + "step": 4372 + }, + { + "epoch": 0.39437254813545564, + "grad_norm": 0.6609801619087752, + "learning_rate": 2.7616546097117213e-06, + "loss": 0.8019, + "step": 4373 + }, + { + "epoch": 0.3944627316589259, + "grad_norm": 1.5111500855792717, + "learning_rate": 2.761114389097564e-06, + "loss": 0.9904, + "step": 4374 + }, + { + "epoch": 0.39455291518239616, + "grad_norm": 1.6310352344835193, + "learning_rate": 2.7605741035409305e-06, + "loss": 1.033, + "step": 4375 + }, + { + "epoch": 0.39464309870586645, + "grad_norm": 1.4242669411780455, + "learning_rate": 2.76003375308792e-06, + "loss": 0.9783, + "step": 4376 + }, + { + "epoch": 0.3947332822293367, + "grad_norm": 1.3302441912034764, + "learning_rate": 2.75949333778464e-06, + "loss": 0.9636, + "step": 4377 + }, + { + "epoch": 0.39482346575280697, + "grad_norm": 1.5367174436547177, + "learning_rate": 2.7589528576772e-06, + "loss": 0.8795, + "step": 4378 + }, + { + "epoch": 0.3949136492762772, + "grad_norm": 3.3917653499808997, + "learning_rate": 2.758412312811717e-06, + "loss": 0.9853, + "step": 4379 + }, + { + "epoch": 0.3950038327997475, + "grad_norm": 0.7575809478943206, + "learning_rate": 2.7578717032343146e-06, + "loss": 0.8676, + "step": 4380 + }, + { + "epoch": 0.3950940163232177, + "grad_norm": 1.5623592313298365, + "learning_rate": 2.757331028991119e-06, + "loss": 1.0195, + "step": 4381 + }, + { + "epoch": 0.395184199846688, + "grad_norm": 1.8206606028644532, + "learning_rate": 2.7567902901282642e-06, + "loss": 1.0094, + "step": 4382 + }, + { + "epoch": 0.3952743833701583, + "grad_norm": 1.8055988339185012, + "learning_rate": 2.7562494866918892e-06, + "loss": 0.987, + "step": 4383 + }, + { + "epoch": 0.39536456689362853, + "grad_norm": 1.5775599459519436, + "learning_rate": 2.7557086187281378e-06, + "loss": 0.9684, + "step": 4384 + }, + { + "epoch": 0.3954547504170988, + "grad_norm": 1.8864778163443767, + "learning_rate": 2.75516768628316e-06, + "loss": 1.0958, + "step": 4385 + }, + { + "epoch": 0.39554493394056905, + "grad_norm": 1.5799104014475545, + "learning_rate": 2.7546266894031114e-06, + "loss": 1.0664, + "step": 4386 + }, + { + "epoch": 0.39563511746403934, + "grad_norm": 1.5016124324058329, + "learning_rate": 2.7540856281341526e-06, + "loss": 0.9445, + "step": 4387 + }, + { + "epoch": 0.3957253009875096, + "grad_norm": 1.624581943435381, + "learning_rate": 2.7535445025224506e-06, + "loss": 0.9581, + "step": 4388 + }, + { + "epoch": 0.39581548451097986, + "grad_norm": 1.8253693637960724, + "learning_rate": 2.753003312614176e-06, + "loss": 0.9961, + "step": 4389 + }, + { + "epoch": 0.3959056680344501, + "grad_norm": 2.4477841496953436, + "learning_rate": 2.7524620584555065e-06, + "loss": 1.0002, + "step": 4390 + }, + { + "epoch": 0.3959958515579204, + "grad_norm": 1.7431002565134348, + "learning_rate": 2.7519207400926253e-06, + "loss": 1.0352, + "step": 4391 + }, + { + "epoch": 0.3960860350813906, + "grad_norm": 1.4144209939488204, + "learning_rate": 2.751379357571721e-06, + "loss": 0.9765, + "step": 4392 + }, + { + "epoch": 0.3961762186048609, + "grad_norm": 1.4167426406833459, + "learning_rate": 2.7508379109389865e-06, + "loss": 1.0627, + "step": 4393 + }, + { + "epoch": 0.39626640212833114, + "grad_norm": 1.7585646990863892, + "learning_rate": 2.750296400240622e-06, + "loss": 1.0121, + "step": 4394 + }, + { + "epoch": 0.39635658565180143, + "grad_norm": 1.3374043492848462, + "learning_rate": 2.7497548255228305e-06, + "loss": 1.0302, + "step": 4395 + }, + { + "epoch": 0.39644676917527166, + "grad_norm": 1.73611198255729, + "learning_rate": 2.749213186831824e-06, + "loss": 0.943, + "step": 4396 + }, + { + "epoch": 0.39653695269874195, + "grad_norm": 1.5706137752473408, + "learning_rate": 2.7486714842138173e-06, + "loss": 1.0423, + "step": 4397 + }, + { + "epoch": 0.3966271362222122, + "grad_norm": 1.9838790874630177, + "learning_rate": 2.748129717715031e-06, + "loss": 0.9464, + "step": 4398 + }, + { + "epoch": 0.3967173197456825, + "grad_norm": 1.5158461576523634, + "learning_rate": 2.747587887381692e-06, + "loss": 1.016, + "step": 4399 + }, + { + "epoch": 0.3968075032691527, + "grad_norm": 1.5487360692417576, + "learning_rate": 2.7470459932600328e-06, + "loss": 0.9436, + "step": 4400 + }, + { + "epoch": 0.396897686792623, + "grad_norm": 2.523430791792796, + "learning_rate": 2.7465040353962897e-06, + "loss": 0.9913, + "step": 4401 + }, + { + "epoch": 0.3969878703160932, + "grad_norm": 1.7477723657319857, + "learning_rate": 2.745962013836706e-06, + "loss": 0.9871, + "step": 4402 + }, + { + "epoch": 0.3970780538395635, + "grad_norm": 2.234089866531398, + "learning_rate": 2.74541992862753e-06, + "loss": 0.9741, + "step": 4403 + }, + { + "epoch": 0.39716823736303375, + "grad_norm": 1.7982908558653181, + "learning_rate": 2.744877779815016e-06, + "loss": 0.9977, + "step": 4404 + }, + { + "epoch": 0.39725842088650404, + "grad_norm": 1.3779697425908783, + "learning_rate": 2.7443355674454234e-06, + "loss": 0.9892, + "step": 4405 + }, + { + "epoch": 0.39734860440997427, + "grad_norm": 1.8947485730298834, + "learning_rate": 2.743793291565015e-06, + "loss": 1.0347, + "step": 4406 + }, + { + "epoch": 0.39743878793344456, + "grad_norm": 1.7482483135069604, + "learning_rate": 2.7432509522200617e-06, + "loss": 0.9175, + "step": 4407 + }, + { + "epoch": 0.39752897145691485, + "grad_norm": 2.2142395594836226, + "learning_rate": 2.7427085494568383e-06, + "loss": 0.939, + "step": 4408 + }, + { + "epoch": 0.3976191549803851, + "grad_norm": 0.6621772033697378, + "learning_rate": 2.742166083321628e-06, + "loss": 0.8441, + "step": 4409 + }, + { + "epoch": 0.39770933850385537, + "grad_norm": 1.874076361453057, + "learning_rate": 2.7416235538607137e-06, + "loss": 1.0396, + "step": 4410 + }, + { + "epoch": 0.3977995220273256, + "grad_norm": 1.7740317537977346, + "learning_rate": 2.7410809611203894e-06, + "loss": 0.9737, + "step": 4411 + }, + { + "epoch": 0.3978897055507959, + "grad_norm": 1.6556179352059122, + "learning_rate": 2.7405383051469507e-06, + "loss": 0.9533, + "step": 4412 + }, + { + "epoch": 0.3979798890742661, + "grad_norm": 1.7420305258335145, + "learning_rate": 2.7399955859867e-06, + "loss": 0.8119, + "step": 4413 + }, + { + "epoch": 0.3980700725977364, + "grad_norm": 0.625425456195175, + "learning_rate": 2.7394528036859465e-06, + "loss": 0.7877, + "step": 4414 + }, + { + "epoch": 0.39816025612120665, + "grad_norm": 1.609723692220317, + "learning_rate": 2.738909958291002e-06, + "loss": 0.998, + "step": 4415 + }, + { + "epoch": 0.39825043964467693, + "grad_norm": 1.9806139288966327, + "learning_rate": 2.7383670498481863e-06, + "loss": 0.9092, + "step": 4416 + }, + { + "epoch": 0.39834062316814717, + "grad_norm": 0.5773437737450704, + "learning_rate": 2.737824078403822e-06, + "loss": 0.7468, + "step": 4417 + }, + { + "epoch": 0.39843080669161746, + "grad_norm": 1.6264090664454671, + "learning_rate": 2.737281044004239e-06, + "loss": 0.9733, + "step": 4418 + }, + { + "epoch": 0.3985209902150877, + "grad_norm": 1.6614468214379625, + "learning_rate": 2.736737946695772e-06, + "loss": 0.9809, + "step": 4419 + }, + { + "epoch": 0.398611173738558, + "grad_norm": 1.7360433166830551, + "learning_rate": 2.736194786524761e-06, + "loss": 0.9904, + "step": 4420 + }, + { + "epoch": 0.3987013572620282, + "grad_norm": 1.5325487817962615, + "learning_rate": 2.7356515635375517e-06, + "loss": 0.9922, + "step": 4421 + }, + { + "epoch": 0.3987915407854985, + "grad_norm": 1.968068262168529, + "learning_rate": 2.735108277780495e-06, + "loss": 0.9078, + "step": 4422 + }, + { + "epoch": 0.39888172430896873, + "grad_norm": 1.6347910014271452, + "learning_rate": 2.7345649292999456e-06, + "loss": 0.9823, + "step": 4423 + }, + { + "epoch": 0.398971907832439, + "grad_norm": 2.0839769386509195, + "learning_rate": 2.734021518142267e-06, + "loss": 1.0014, + "step": 4424 + }, + { + "epoch": 0.39906209135590925, + "grad_norm": 2.2138731106214875, + "learning_rate": 2.733478044353825e-06, + "loss": 0.8241, + "step": 4425 + }, + { + "epoch": 0.39915227487937954, + "grad_norm": 2.224131866998469, + "learning_rate": 2.7329345079809917e-06, + "loss": 0.9999, + "step": 4426 + }, + { + "epoch": 0.3992424584028498, + "grad_norm": 1.5500346579830033, + "learning_rate": 2.7323909090701447e-06, + "loss": 0.9307, + "step": 4427 + }, + { + "epoch": 0.39933264192632006, + "grad_norm": 1.8560758541120164, + "learning_rate": 2.731847247667667e-06, + "loss": 1.0476, + "step": 4428 + }, + { + "epoch": 0.3994228254497903, + "grad_norm": 1.677354549804943, + "learning_rate": 2.731303523819947e-06, + "loss": 0.9688, + "step": 4429 + }, + { + "epoch": 0.3995130089732606, + "grad_norm": 1.6544801144412165, + "learning_rate": 2.7307597375733783e-06, + "loss": 1.0562, + "step": 4430 + }, + { + "epoch": 0.3996031924967309, + "grad_norm": 1.8366878193758203, + "learning_rate": 2.7302158889743587e-06, + "loss": 0.9893, + "step": 4431 + }, + { + "epoch": 0.3996933760202011, + "grad_norm": 1.6182712480295436, + "learning_rate": 2.7296719780692937e-06, + "loss": 0.953, + "step": 4432 + }, + { + "epoch": 0.3997835595436714, + "grad_norm": 2.268798079388428, + "learning_rate": 2.7291280049045916e-06, + "loss": 0.9072, + "step": 4433 + }, + { + "epoch": 0.39987374306714163, + "grad_norm": 1.5863432717463202, + "learning_rate": 2.7285839695266683e-06, + "loss": 0.9379, + "step": 4434 + }, + { + "epoch": 0.3999639265906119, + "grad_norm": 2.112929112116829, + "learning_rate": 2.7280398719819423e-06, + "loss": 0.943, + "step": 4435 + }, + { + "epoch": 0.40005411011408215, + "grad_norm": 1.8216292567249515, + "learning_rate": 2.727495712316841e-06, + "loss": 1.061, + "step": 4436 + }, + { + "epoch": 0.40014429363755244, + "grad_norm": 3.2885699775515524, + "learning_rate": 2.7269514905777945e-06, + "loss": 0.9366, + "step": 4437 + }, + { + "epoch": 0.4002344771610227, + "grad_norm": 1.6739178592304593, + "learning_rate": 2.7264072068112377e-06, + "loss": 1.1166, + "step": 4438 + }, + { + "epoch": 0.40032466068449296, + "grad_norm": 1.7918990524162608, + "learning_rate": 2.7258628610636133e-06, + "loss": 0.9567, + "step": 4439 + }, + { + "epoch": 0.4004148442079632, + "grad_norm": 2.0686121453859663, + "learning_rate": 2.7253184533813667e-06, + "loss": 1.0709, + "step": 4440 + }, + { + "epoch": 0.4005050277314335, + "grad_norm": 1.6534711450135977, + "learning_rate": 2.72477398381095e-06, + "loss": 1.0051, + "step": 4441 + }, + { + "epoch": 0.4005952112549037, + "grad_norm": 1.7357186963262479, + "learning_rate": 2.724229452398821e-06, + "loss": 1.0, + "step": 4442 + }, + { + "epoch": 0.400685394778374, + "grad_norm": 1.6763881482482021, + "learning_rate": 2.7236848591914422e-06, + "loss": 1.0108, + "step": 4443 + }, + { + "epoch": 0.40077557830184424, + "grad_norm": 2.5309376465024576, + "learning_rate": 2.7231402042352803e-06, + "loss": 0.9775, + "step": 4444 + }, + { + "epoch": 0.4008657618253145, + "grad_norm": 2.068632084462266, + "learning_rate": 2.722595487576809e-06, + "loss": 1.0237, + "step": 4445 + }, + { + "epoch": 0.40095594534878476, + "grad_norm": 1.6258883615576616, + "learning_rate": 2.722050709262506e-06, + "loss": 1.0399, + "step": 4446 + }, + { + "epoch": 0.40104612887225505, + "grad_norm": 1.4012921994445635, + "learning_rate": 2.7215058693388557e-06, + "loss": 0.9768, + "step": 4447 + }, + { + "epoch": 0.4011363123957253, + "grad_norm": 1.8600881154712818, + "learning_rate": 2.720960967852346e-06, + "loss": 0.8936, + "step": 4448 + }, + { + "epoch": 0.40122649591919557, + "grad_norm": 1.5422026845074317, + "learning_rate": 2.720416004849471e-06, + "loss": 0.9275, + "step": 4449 + }, + { + "epoch": 0.4013166794426658, + "grad_norm": 1.7578379989469997, + "learning_rate": 2.7198709803767304e-06, + "loss": 0.9462, + "step": 4450 + }, + { + "epoch": 0.4014068629661361, + "grad_norm": 1.415943307016454, + "learning_rate": 2.7193258944806286e-06, + "loss": 0.9659, + "step": 4451 + }, + { + "epoch": 0.4014970464896063, + "grad_norm": 1.6991103565239223, + "learning_rate": 2.718780747207675e-06, + "loss": 0.9976, + "step": 4452 + }, + { + "epoch": 0.4015872300130766, + "grad_norm": 1.7646163489966078, + "learning_rate": 2.7182355386043847e-06, + "loss": 1.0566, + "step": 4453 + }, + { + "epoch": 0.40167741353654685, + "grad_norm": 1.5480879834723649, + "learning_rate": 2.717690268717278e-06, + "loss": 1.0584, + "step": 4454 + }, + { + "epoch": 0.40176759706001713, + "grad_norm": 1.4497946856920847, + "learning_rate": 2.7171449375928803e-06, + "loss": 1.0198, + "step": 4455 + }, + { + "epoch": 0.4018577805834874, + "grad_norm": 2.0763497478264763, + "learning_rate": 2.716599545277722e-06, + "loss": 1.0261, + "step": 4456 + }, + { + "epoch": 0.40194796410695766, + "grad_norm": 1.5365872690013276, + "learning_rate": 2.7160540918183394e-06, + "loss": 0.9529, + "step": 4457 + }, + { + "epoch": 0.40203814763042794, + "grad_norm": 2.4785086990151455, + "learning_rate": 2.715508577261273e-06, + "loss": 0.9123, + "step": 4458 + }, + { + "epoch": 0.4021283311538982, + "grad_norm": 1.7890498398245425, + "learning_rate": 2.7149630016530702e-06, + "loss": 0.9734, + "step": 4459 + }, + { + "epoch": 0.40221851467736847, + "grad_norm": 1.4308946026655416, + "learning_rate": 2.7144173650402815e-06, + "loss": 1.0097, + "step": 4460 + }, + { + "epoch": 0.4023086982008387, + "grad_norm": 1.4300740871044213, + "learning_rate": 2.7138716674694636e-06, + "loss": 1.0154, + "step": 4461 + }, + { + "epoch": 0.402398881724309, + "grad_norm": 1.6629457263451297, + "learning_rate": 2.7133259089871795e-06, + "loss": 0.9403, + "step": 4462 + }, + { + "epoch": 0.4024890652477792, + "grad_norm": 3.171945054356466, + "learning_rate": 2.712780089639995e-06, + "loss": 0.9898, + "step": 4463 + }, + { + "epoch": 0.4025792487712495, + "grad_norm": 1.448998776421514, + "learning_rate": 2.712234209474483e-06, + "loss": 0.8701, + "step": 4464 + }, + { + "epoch": 0.40266943229471974, + "grad_norm": 1.476825024715348, + "learning_rate": 2.7116882685372218e-06, + "loss": 1.0341, + "step": 4465 + }, + { + "epoch": 0.40275961581819003, + "grad_norm": 1.579048415347121, + "learning_rate": 2.7111422668747927e-06, + "loss": 0.9585, + "step": 4466 + }, + { + "epoch": 0.40284979934166026, + "grad_norm": 1.5703892475560706, + "learning_rate": 2.7105962045337846e-06, + "loss": 0.9883, + "step": 4467 + }, + { + "epoch": 0.40293998286513055, + "grad_norm": 1.7241244125658373, + "learning_rate": 2.7100500815607898e-06, + "loss": 0.9848, + "step": 4468 + }, + { + "epoch": 0.4030301663886008, + "grad_norm": 1.4730163457757302, + "learning_rate": 2.709503898002407e-06, + "loss": 1.0161, + "step": 4469 + }, + { + "epoch": 0.4031203499120711, + "grad_norm": 2.1972839625970915, + "learning_rate": 2.708957653905239e-06, + "loss": 0.9898, + "step": 4470 + }, + { + "epoch": 0.4032105334355413, + "grad_norm": 1.5146271871998185, + "learning_rate": 2.7084113493158956e-06, + "loss": 1.0024, + "step": 4471 + }, + { + "epoch": 0.4033007169590116, + "grad_norm": 1.445418317890443, + "learning_rate": 2.7078649842809888e-06, + "loss": 1.0036, + "step": 4472 + }, + { + "epoch": 0.40339090048248183, + "grad_norm": 1.5370318964004515, + "learning_rate": 2.707318558847139e-06, + "loss": 1.042, + "step": 4473 + }, + { + "epoch": 0.4034810840059521, + "grad_norm": 0.7240815000712436, + "learning_rate": 2.7067720730609697e-06, + "loss": 0.8173, + "step": 4474 + }, + { + "epoch": 0.40357126752942235, + "grad_norm": 1.4942123495601296, + "learning_rate": 2.70622552696911e-06, + "loss": 0.9642, + "step": 4475 + }, + { + "epoch": 0.40366145105289264, + "grad_norm": 1.4548724752430953, + "learning_rate": 2.7056789206181943e-06, + "loss": 0.9446, + "step": 4476 + }, + { + "epoch": 0.40375163457636287, + "grad_norm": 1.6186234001230124, + "learning_rate": 2.7051322540548615e-06, + "loss": 1.0477, + "step": 4477 + }, + { + "epoch": 0.40384181809983316, + "grad_norm": 3.3137576037146648, + "learning_rate": 2.704585527325757e-06, + "loss": 0.9674, + "step": 4478 + }, + { + "epoch": 0.40393200162330345, + "grad_norm": 1.5555174639367386, + "learning_rate": 2.7040387404775303e-06, + "loss": 0.9493, + "step": 4479 + }, + { + "epoch": 0.4040221851467737, + "grad_norm": 1.8269123718009854, + "learning_rate": 2.703491893556837e-06, + "loss": 1.022, + "step": 4480 + }, + { + "epoch": 0.40411236867024397, + "grad_norm": 1.5047418824189378, + "learning_rate": 2.702944986610335e-06, + "loss": 0.9684, + "step": 4481 + }, + { + "epoch": 0.4042025521937142, + "grad_norm": 1.9037286311632864, + "learning_rate": 2.7023980196846917e-06, + "loss": 1.0072, + "step": 4482 + }, + { + "epoch": 0.4042927357171845, + "grad_norm": 0.7107546225285768, + "learning_rate": 2.7018509928265763e-06, + "loss": 0.8357, + "step": 4483 + }, + { + "epoch": 0.4043829192406547, + "grad_norm": 2.059586750541917, + "learning_rate": 2.7013039060826635e-06, + "loss": 1.023, + "step": 4484 + }, + { + "epoch": 0.404473102764125, + "grad_norm": 2.0330395371645067, + "learning_rate": 2.7007567594996347e-06, + "loss": 1.0606, + "step": 4485 + }, + { + "epoch": 0.40456328628759525, + "grad_norm": 1.7579136798137358, + "learning_rate": 2.7002095531241757e-06, + "loss": 0.8965, + "step": 4486 + }, + { + "epoch": 0.40465346981106554, + "grad_norm": 1.8848569541505964, + "learning_rate": 2.6996622870029767e-06, + "loss": 0.9705, + "step": 4487 + }, + { + "epoch": 0.40474365333453577, + "grad_norm": 1.7051168570647404, + "learning_rate": 2.6991149611827335e-06, + "loss": 1.0998, + "step": 4488 + }, + { + "epoch": 0.40483383685800606, + "grad_norm": 1.5724031565915693, + "learning_rate": 2.6985675757101466e-06, + "loss": 1.0334, + "step": 4489 + }, + { + "epoch": 0.4049240203814763, + "grad_norm": 1.8383088076982435, + "learning_rate": 2.698020130631922e-06, + "loss": 0.9527, + "step": 4490 + }, + { + "epoch": 0.4050142039049466, + "grad_norm": 1.5405222365208526, + "learning_rate": 2.6974726259947713e-06, + "loss": 0.9277, + "step": 4491 + }, + { + "epoch": 0.4051043874284168, + "grad_norm": 2.458305574249613, + "learning_rate": 2.6969250618454106e-06, + "loss": 0.9704, + "step": 4492 + }, + { + "epoch": 0.4051945709518871, + "grad_norm": 1.39598678580427, + "learning_rate": 2.696377438230561e-06, + "loss": 1.0661, + "step": 4493 + }, + { + "epoch": 0.40528475447535733, + "grad_norm": 1.6982688616543742, + "learning_rate": 2.6958297551969484e-06, + "loss": 0.9287, + "step": 4494 + }, + { + "epoch": 0.4053749379988276, + "grad_norm": 1.6246575214711418, + "learning_rate": 2.695282012791304e-06, + "loss": 0.9614, + "step": 4495 + }, + { + "epoch": 0.40546512152229786, + "grad_norm": 1.532183848965469, + "learning_rate": 2.6947342110603646e-06, + "loss": 1.0458, + "step": 4496 + }, + { + "epoch": 0.40555530504576814, + "grad_norm": 1.4532434558783407, + "learning_rate": 2.6941863500508717e-06, + "loss": 1.0483, + "step": 4497 + }, + { + "epoch": 0.4056454885692384, + "grad_norm": 3.2510771800290086, + "learning_rate": 2.693638429809572e-06, + "loss": 1.0461, + "step": 4498 + }, + { + "epoch": 0.40573567209270867, + "grad_norm": 1.7068823568324556, + "learning_rate": 2.6930904503832167e-06, + "loss": 0.9781, + "step": 4499 + }, + { + "epoch": 0.4058258556161789, + "grad_norm": 1.9026943251910264, + "learning_rate": 2.692542411818562e-06, + "loss": 0.9772, + "step": 4500 + }, + { + "epoch": 0.4059160391396492, + "grad_norm": 1.93518414067239, + "learning_rate": 2.69199431416237e-06, + "loss": 1.0047, + "step": 4501 + }, + { + "epoch": 0.4060062226631194, + "grad_norm": 1.513821502181254, + "learning_rate": 2.691446157461408e-06, + "loss": 0.9469, + "step": 4502 + }, + { + "epoch": 0.4060964061865897, + "grad_norm": 1.5407163766995455, + "learning_rate": 2.690897941762447e-06, + "loss": 1.0576, + "step": 4503 + }, + { + "epoch": 0.40618658971006, + "grad_norm": 1.480188992521821, + "learning_rate": 2.6903496671122642e-06, + "loss": 0.9705, + "step": 4504 + }, + { + "epoch": 0.40627677323353023, + "grad_norm": 0.6428735164229875, + "learning_rate": 2.689801333557641e-06, + "loss": 0.7666, + "step": 4505 + }, + { + "epoch": 0.4063669567570005, + "grad_norm": 3.462810483156816, + "learning_rate": 2.689252941145365e-06, + "loss": 0.9959, + "step": 4506 + }, + { + "epoch": 0.40645714028047075, + "grad_norm": 1.4356237689333908, + "learning_rate": 2.6887044899222277e-06, + "loss": 0.9916, + "step": 4507 + }, + { + "epoch": 0.40654732380394104, + "grad_norm": 1.7158805476185335, + "learning_rate": 2.688155979935025e-06, + "loss": 1.039, + "step": 4508 + }, + { + "epoch": 0.4066375073274113, + "grad_norm": 1.6048921538154575, + "learning_rate": 2.68760741123056e-06, + "loss": 0.9516, + "step": 4509 + }, + { + "epoch": 0.40672769085088156, + "grad_norm": 1.7458917215032268, + "learning_rate": 2.6870587838556394e-06, + "loss": 1.0102, + "step": 4510 + }, + { + "epoch": 0.4068178743743518, + "grad_norm": 1.7875789171417764, + "learning_rate": 2.686510097857075e-06, + "loss": 1.0158, + "step": 4511 + }, + { + "epoch": 0.4069080578978221, + "grad_norm": 1.2852100024823383, + "learning_rate": 2.685961353281683e-06, + "loss": 0.9885, + "step": 4512 + }, + { + "epoch": 0.4069982414212923, + "grad_norm": 1.5218595500059977, + "learning_rate": 2.6854125501762863e-06, + "loss": 0.974, + "step": 4513 + }, + { + "epoch": 0.4070884249447626, + "grad_norm": 1.5959416448159491, + "learning_rate": 2.684863688587712e-06, + "loss": 0.9974, + "step": 4514 + }, + { + "epoch": 0.40717860846823284, + "grad_norm": 1.791125511534085, + "learning_rate": 2.6843147685627916e-06, + "loss": 0.9676, + "step": 4515 + }, + { + "epoch": 0.4072687919917031, + "grad_norm": 1.7335623822118487, + "learning_rate": 2.683765790148361e-06, + "loss": 0.9688, + "step": 4516 + }, + { + "epoch": 0.40735897551517336, + "grad_norm": 1.9007506544078667, + "learning_rate": 2.6832167533912637e-06, + "loss": 1.0469, + "step": 4517 + }, + { + "epoch": 0.40744915903864365, + "grad_norm": 2.2832739225951464, + "learning_rate": 2.682667658338345e-06, + "loss": 1.065, + "step": 4518 + }, + { + "epoch": 0.4075393425621139, + "grad_norm": 1.6879548060392005, + "learning_rate": 2.682118505036458e-06, + "loss": 0.9669, + "step": 4519 + }, + { + "epoch": 0.40762952608558417, + "grad_norm": 0.6139910775306733, + "learning_rate": 2.681569293532459e-06, + "loss": 0.7871, + "step": 4520 + }, + { + "epoch": 0.4077197096090544, + "grad_norm": 0.6508770518408856, + "learning_rate": 2.6810200238732102e-06, + "loss": 0.8153, + "step": 4521 + }, + { + "epoch": 0.4078098931325247, + "grad_norm": 1.3635518589712243, + "learning_rate": 2.6804706961055776e-06, + "loss": 1.0365, + "step": 4522 + }, + { + "epoch": 0.4079000766559949, + "grad_norm": 1.3841226265913724, + "learning_rate": 2.6799213102764326e-06, + "loss": 1.0268, + "step": 4523 + }, + { + "epoch": 0.4079902601794652, + "grad_norm": 1.6733093526091998, + "learning_rate": 2.679371866432653e-06, + "loss": 1.085, + "step": 4524 + }, + { + "epoch": 0.40808044370293545, + "grad_norm": 1.85742148992011, + "learning_rate": 2.6788223646211194e-06, + "loss": 0.9679, + "step": 4525 + }, + { + "epoch": 0.40817062722640574, + "grad_norm": 1.4987822994159226, + "learning_rate": 2.6782728048887183e-06, + "loss": 0.8831, + "step": 4526 + }, + { + "epoch": 0.408260810749876, + "grad_norm": 1.2819572915567383, + "learning_rate": 2.6777231872823416e-06, + "loss": 0.9796, + "step": 4527 + }, + { + "epoch": 0.40835099427334626, + "grad_norm": 1.6169322821427619, + "learning_rate": 2.6771735118488864e-06, + "loss": 0.9911, + "step": 4528 + }, + { + "epoch": 0.40844117779681655, + "grad_norm": 1.6723796047332917, + "learning_rate": 2.6766237786352523e-06, + "loss": 1.0314, + "step": 4529 + }, + { + "epoch": 0.4085313613202868, + "grad_norm": 2.224939872700786, + "learning_rate": 2.676073987688347e-06, + "loss": 0.9213, + "step": 4530 + }, + { + "epoch": 0.40862154484375707, + "grad_norm": 1.7613549374899746, + "learning_rate": 2.6755241390550818e-06, + "loss": 0.9924, + "step": 4531 + }, + { + "epoch": 0.4087117283672273, + "grad_norm": 1.4689112635611026, + "learning_rate": 2.6749742327823716e-06, + "loss": 0.9953, + "step": 4532 + }, + { + "epoch": 0.4088019118906976, + "grad_norm": 1.5271363568782552, + "learning_rate": 2.674424268917138e-06, + "loss": 1.0058, + "step": 4533 + }, + { + "epoch": 0.4088920954141678, + "grad_norm": 1.9725885662893488, + "learning_rate": 2.6738742475063074e-06, + "loss": 1.055, + "step": 4534 + }, + { + "epoch": 0.4089822789376381, + "grad_norm": 1.79898968529196, + "learning_rate": 2.6733241685968104e-06, + "loss": 1.0323, + "step": 4535 + }, + { + "epoch": 0.40907246246110834, + "grad_norm": 1.77382430509027, + "learning_rate": 2.6727740322355826e-06, + "loss": 1.0081, + "step": 4536 + }, + { + "epoch": 0.40916264598457863, + "grad_norm": 1.8777076244730302, + "learning_rate": 2.6722238384695644e-06, + "loss": 0.9754, + "step": 4537 + }, + { + "epoch": 0.40925282950804887, + "grad_norm": 2.1456503805542293, + "learning_rate": 2.671673587345702e-06, + "loss": 1.0021, + "step": 4538 + }, + { + "epoch": 0.40934301303151915, + "grad_norm": 0.6514339384222988, + "learning_rate": 2.6711232789109455e-06, + "loss": 0.8058, + "step": 4539 + }, + { + "epoch": 0.4094331965549894, + "grad_norm": 1.5425652471310554, + "learning_rate": 2.6705729132122497e-06, + "loss": 1.0017, + "step": 4540 + }, + { + "epoch": 0.4095233800784597, + "grad_norm": 5.007574638157066, + "learning_rate": 2.670022490296576e-06, + "loss": 1.0699, + "step": 4541 + }, + { + "epoch": 0.4096135636019299, + "grad_norm": 1.6836950709750733, + "learning_rate": 2.669472010210889e-06, + "loss": 1.0534, + "step": 4542 + }, + { + "epoch": 0.4097037471254002, + "grad_norm": 1.9379550030408434, + "learning_rate": 2.668921473002159e-06, + "loss": 0.9745, + "step": 4543 + }, + { + "epoch": 0.40979393064887043, + "grad_norm": 1.5011030750654084, + "learning_rate": 2.6683708787173596e-06, + "loss": 1.0484, + "step": 4544 + }, + { + "epoch": 0.4098841141723407, + "grad_norm": 1.4383082398177593, + "learning_rate": 2.6678202274034718e-06, + "loss": 1.0149, + "step": 4545 + }, + { + "epoch": 0.40997429769581095, + "grad_norm": 1.7370350689068377, + "learning_rate": 2.66726951910748e-06, + "loss": 1.0827, + "step": 4546 + }, + { + "epoch": 0.41006448121928124, + "grad_norm": 1.577693842541723, + "learning_rate": 2.6667187538763737e-06, + "loss": 0.9982, + "step": 4547 + }, + { + "epoch": 0.4101546647427515, + "grad_norm": 1.736134985017871, + "learning_rate": 2.6661679317571473e-06, + "loss": 0.9765, + "step": 4548 + }, + { + "epoch": 0.41024484826622176, + "grad_norm": 1.5458015030756733, + "learning_rate": 2.665617052796799e-06, + "loss": 0.9787, + "step": 4549 + }, + { + "epoch": 0.41033503178969205, + "grad_norm": 1.692316669243565, + "learning_rate": 2.6650661170423346e-06, + "loss": 0.8763, + "step": 4550 + }, + { + "epoch": 0.4104252153131623, + "grad_norm": 1.7658933587498788, + "learning_rate": 2.6645151245407614e-06, + "loss": 1.0754, + "step": 4551 + }, + { + "epoch": 0.4105153988366326, + "grad_norm": 1.4720224488868556, + "learning_rate": 2.6639640753390936e-06, + "loss": 0.9906, + "step": 4552 + }, + { + "epoch": 0.4106055823601028, + "grad_norm": 1.4238363299767876, + "learning_rate": 2.66341296948435e-06, + "loss": 0.8978, + "step": 4553 + }, + { + "epoch": 0.4106957658835731, + "grad_norm": 1.4387065136499602, + "learning_rate": 2.6628618070235534e-06, + "loss": 1.0007, + "step": 4554 + }, + { + "epoch": 0.4107859494070433, + "grad_norm": 1.6557179442108672, + "learning_rate": 2.662310588003733e-06, + "loss": 1.0013, + "step": 4555 + }, + { + "epoch": 0.4108761329305136, + "grad_norm": 2.2396052041575323, + "learning_rate": 2.6617593124719205e-06, + "loss": 0.9731, + "step": 4556 + }, + { + "epoch": 0.41096631645398385, + "grad_norm": 1.3376861638789355, + "learning_rate": 2.661207980475155e-06, + "loss": 1.0009, + "step": 4557 + }, + { + "epoch": 0.41105649997745414, + "grad_norm": 1.8462200107088793, + "learning_rate": 2.6606565920604793e-06, + "loss": 0.9563, + "step": 4558 + }, + { + "epoch": 0.41114668350092437, + "grad_norm": 1.4968712283029204, + "learning_rate": 2.66010514727494e-06, + "loss": 1.0732, + "step": 4559 + }, + { + "epoch": 0.41123686702439466, + "grad_norm": 1.6321976521611024, + "learning_rate": 2.659553646165589e-06, + "loss": 0.8912, + "step": 4560 + }, + { + "epoch": 0.4113270505478649, + "grad_norm": 1.616414203135642, + "learning_rate": 2.659002088779485e-06, + "loss": 0.9659, + "step": 4561 + }, + { + "epoch": 0.4114172340713352, + "grad_norm": 0.5983372965453855, + "learning_rate": 2.6584504751636888e-06, + "loss": 0.7654, + "step": 4562 + }, + { + "epoch": 0.4115074175948054, + "grad_norm": 1.6599092187532285, + "learning_rate": 2.657898805365268e-06, + "loss": 0.9612, + "step": 4563 + }, + { + "epoch": 0.4115976011182757, + "grad_norm": 1.5643539111506306, + "learning_rate": 2.657347079431293e-06, + "loss": 0.9624, + "step": 4564 + }, + { + "epoch": 0.41168778464174594, + "grad_norm": 1.3066792617810385, + "learning_rate": 2.6567952974088403e-06, + "loss": 0.9704, + "step": 4565 + }, + { + "epoch": 0.4117779681652162, + "grad_norm": 1.3500070607036043, + "learning_rate": 2.6562434593449917e-06, + "loss": 1.0422, + "step": 4566 + }, + { + "epoch": 0.41186815168868646, + "grad_norm": 4.083476446848662, + "learning_rate": 2.6556915652868325e-06, + "loss": 1.0151, + "step": 4567 + }, + { + "epoch": 0.41195833521215675, + "grad_norm": 1.6778204188636914, + "learning_rate": 2.6551396152814534e-06, + "loss": 1.0014, + "step": 4568 + }, + { + "epoch": 0.412048518735627, + "grad_norm": 1.6469914027848378, + "learning_rate": 2.65458760937595e-06, + "loss": 0.987, + "step": 4569 + }, + { + "epoch": 0.41213870225909727, + "grad_norm": 1.3057657685822228, + "learning_rate": 2.654035547617423e-06, + "loss": 0.9407, + "step": 4570 + }, + { + "epoch": 0.4122288857825675, + "grad_norm": 1.8660303464631836, + "learning_rate": 2.653483430052976e-06, + "loss": 0.8971, + "step": 4571 + }, + { + "epoch": 0.4123190693060378, + "grad_norm": 1.7503942318048926, + "learning_rate": 2.6529312567297197e-06, + "loss": 0.9817, + "step": 4572 + }, + { + "epoch": 0.412409252829508, + "grad_norm": 2.205553056035946, + "learning_rate": 2.652379027694768e-06, + "loss": 1.0706, + "step": 4573 + }, + { + "epoch": 0.4124994363529783, + "grad_norm": 2.2024588152599494, + "learning_rate": 2.651826742995241e-06, + "loss": 1.05, + "step": 4574 + }, + { + "epoch": 0.4125896198764486, + "grad_norm": 1.5446733161449477, + "learning_rate": 2.651274402678262e-06, + "loss": 0.9906, + "step": 4575 + }, + { + "epoch": 0.41267980339991883, + "grad_norm": 1.8109710426943377, + "learning_rate": 2.6507220067909597e-06, + "loss": 1.0238, + "step": 4576 + }, + { + "epoch": 0.4127699869233891, + "grad_norm": 1.7156221539138505, + "learning_rate": 2.650169555380468e-06, + "loss": 0.9525, + "step": 4577 + }, + { + "epoch": 0.41286017044685935, + "grad_norm": 1.8307548582644702, + "learning_rate": 2.6496170484939253e-06, + "loss": 0.9891, + "step": 4578 + }, + { + "epoch": 0.41295035397032964, + "grad_norm": 1.4092404675610484, + "learning_rate": 2.6490644861784735e-06, + "loss": 1.0077, + "step": 4579 + }, + { + "epoch": 0.4130405374937999, + "grad_norm": 1.6609266939287177, + "learning_rate": 2.648511868481261e-06, + "loss": 1.0146, + "step": 4580 + }, + { + "epoch": 0.41313072101727016, + "grad_norm": 1.7945303837629194, + "learning_rate": 2.6479591954494397e-06, + "loss": 0.9915, + "step": 4581 + }, + { + "epoch": 0.4132209045407404, + "grad_norm": 1.403024194915269, + "learning_rate": 2.647406467130167e-06, + "loss": 0.8908, + "step": 4582 + }, + { + "epoch": 0.4133110880642107, + "grad_norm": 1.6369193962400799, + "learning_rate": 2.646853683570605e-06, + "loss": 1.0122, + "step": 4583 + }, + { + "epoch": 0.4134012715876809, + "grad_norm": 1.5182465366353546, + "learning_rate": 2.6463008448179196e-06, + "loss": 0.9326, + "step": 4584 + }, + { + "epoch": 0.4134914551111512, + "grad_norm": 1.4021587196628804, + "learning_rate": 2.6457479509192828e-06, + "loss": 0.9447, + "step": 4585 + }, + { + "epoch": 0.41358163863462144, + "grad_norm": 1.7760914844972684, + "learning_rate": 2.645195001921871e-06, + "loss": 0.9888, + "step": 4586 + }, + { + "epoch": 0.41367182215809173, + "grad_norm": 1.6063570038892636, + "learning_rate": 2.644641997872863e-06, + "loss": 0.9698, + "step": 4587 + }, + { + "epoch": 0.41376200568156196, + "grad_norm": 1.9189412745464831, + "learning_rate": 2.644088938819445e-06, + "loss": 0.9934, + "step": 4588 + }, + { + "epoch": 0.41385218920503225, + "grad_norm": 1.7124394482676326, + "learning_rate": 2.6435358248088077e-06, + "loss": 1.0779, + "step": 4589 + }, + { + "epoch": 0.4139423727285025, + "grad_norm": 1.6586315378923708, + "learning_rate": 2.642982655888146e-06, + "loss": 1.0362, + "step": 4590 + }, + { + "epoch": 0.41403255625197277, + "grad_norm": 1.5638903726074216, + "learning_rate": 2.6424294321046585e-06, + "loss": 0.9549, + "step": 4591 + }, + { + "epoch": 0.414122739775443, + "grad_norm": 0.7404119837120342, + "learning_rate": 2.641876153505549e-06, + "loss": 0.8113, + "step": 4592 + }, + { + "epoch": 0.4142129232989133, + "grad_norm": 1.7951245613764337, + "learning_rate": 2.641322820138027e-06, + "loss": 0.9948, + "step": 4593 + }, + { + "epoch": 0.4143031068223835, + "grad_norm": 2.0245139303936415, + "learning_rate": 2.640769432049306e-06, + "loss": 0.9383, + "step": 4594 + }, + { + "epoch": 0.4143932903458538, + "grad_norm": 2.2304358271158207, + "learning_rate": 2.6402159892866038e-06, + "loss": 0.9175, + "step": 4595 + }, + { + "epoch": 0.41448347386932405, + "grad_norm": 1.8673767887716946, + "learning_rate": 2.639662491897143e-06, + "loss": 0.9798, + "step": 4596 + }, + { + "epoch": 0.41457365739279434, + "grad_norm": 2.820922362943136, + "learning_rate": 2.639108939928152e-06, + "loss": 1.0079, + "step": 4597 + }, + { + "epoch": 0.4146638409162646, + "grad_norm": 2.401380094452402, + "learning_rate": 2.638555333426862e-06, + "loss": 1.004, + "step": 4598 + }, + { + "epoch": 0.41475402443973486, + "grad_norm": 1.3384592448981387, + "learning_rate": 2.6380016724405093e-06, + "loss": 1.0247, + "step": 4599 + }, + { + "epoch": 0.41484420796320515, + "grad_norm": 1.7722440896965506, + "learning_rate": 2.637447957016336e-06, + "loss": 0.9481, + "step": 4600 + }, + { + "epoch": 0.4149343914866754, + "grad_norm": 2.008726631792991, + "learning_rate": 2.636894187201589e-06, + "loss": 1.0523, + "step": 4601 + }, + { + "epoch": 0.41502457501014567, + "grad_norm": 1.6315943358186236, + "learning_rate": 2.6363403630435176e-06, + "loss": 0.8988, + "step": 4602 + }, + { + "epoch": 0.4151147585336159, + "grad_norm": 1.5349756187962962, + "learning_rate": 2.635786484589378e-06, + "loss": 0.9842, + "step": 4603 + }, + { + "epoch": 0.4152049420570862, + "grad_norm": 1.8220849964696326, + "learning_rate": 2.63523255188643e-06, + "loss": 0.892, + "step": 4604 + }, + { + "epoch": 0.4152951255805564, + "grad_norm": 1.411505352330372, + "learning_rate": 2.6346785649819375e-06, + "loss": 1.0176, + "step": 4605 + }, + { + "epoch": 0.4153853091040267, + "grad_norm": 1.6471197275814273, + "learning_rate": 2.6341245239231706e-06, + "loss": 1.0163, + "step": 4606 + }, + { + "epoch": 0.41547549262749695, + "grad_norm": 1.1874294260034537, + "learning_rate": 2.6335704287574024e-06, + "loss": 0.9928, + "step": 4607 + }, + { + "epoch": 0.41556567615096723, + "grad_norm": 1.7653515359212846, + "learning_rate": 2.6330162795319124e-06, + "loss": 1.0649, + "step": 4608 + }, + { + "epoch": 0.41565585967443747, + "grad_norm": 1.8696410845491656, + "learning_rate": 2.632462076293983e-06, + "loss": 1.0466, + "step": 4609 + }, + { + "epoch": 0.41574604319790776, + "grad_norm": 1.3634111839830665, + "learning_rate": 2.6319078190909017e-06, + "loss": 0.9908, + "step": 4610 + }, + { + "epoch": 0.415836226721378, + "grad_norm": 3.7598320813931885, + "learning_rate": 2.6313535079699606e-06, + "loss": 0.9087, + "step": 4611 + }, + { + "epoch": 0.4159264102448483, + "grad_norm": 1.8464392107561907, + "learning_rate": 2.6307991429784572e-06, + "loss": 0.969, + "step": 4612 + }, + { + "epoch": 0.4160165937683185, + "grad_norm": 1.4324708121423524, + "learning_rate": 2.6302447241636924e-06, + "loss": 1.0537, + "step": 4613 + }, + { + "epoch": 0.4161067772917888, + "grad_norm": 1.6260997278802451, + "learning_rate": 2.629690251572973e-06, + "loss": 1.0672, + "step": 4614 + }, + { + "epoch": 0.41619696081525903, + "grad_norm": 1.613966973903984, + "learning_rate": 2.629135725253609e-06, + "loss": 0.9696, + "step": 4615 + }, + { + "epoch": 0.4162871443387293, + "grad_norm": 4.594343704990936, + "learning_rate": 2.6285811452529162e-06, + "loss": 0.9807, + "step": 4616 + }, + { + "epoch": 0.41637732786219955, + "grad_norm": 1.4539342288236408, + "learning_rate": 2.6280265116182136e-06, + "loss": 1.0095, + "step": 4617 + }, + { + "epoch": 0.41646751138566984, + "grad_norm": 1.479681525374438, + "learning_rate": 2.6274718243968266e-06, + "loss": 0.9983, + "step": 4618 + }, + { + "epoch": 0.4165576949091401, + "grad_norm": 1.7007829854498138, + "learning_rate": 2.626917083636084e-06, + "loss": 0.9116, + "step": 4619 + }, + { + "epoch": 0.41664787843261036, + "grad_norm": 1.3546686366966771, + "learning_rate": 2.6263622893833183e-06, + "loss": 0.9286, + "step": 4620 + }, + { + "epoch": 0.4167380619560806, + "grad_norm": 1.427960379229854, + "learning_rate": 2.625807441685869e-06, + "loss": 0.9374, + "step": 4621 + }, + { + "epoch": 0.4168282454795509, + "grad_norm": 0.6132655050601291, + "learning_rate": 2.625252540591078e-06, + "loss": 0.7992, + "step": 4622 + }, + { + "epoch": 0.4169184290030212, + "grad_norm": 1.6093970732425147, + "learning_rate": 2.6246975861462927e-06, + "loss": 1.0667, + "step": 4623 + }, + { + "epoch": 0.4170086125264914, + "grad_norm": 1.5680809296382447, + "learning_rate": 2.624142578398864e-06, + "loss": 1.0357, + "step": 4624 + }, + { + "epoch": 0.4170987960499617, + "grad_norm": 1.2759696265405396, + "learning_rate": 2.6235875173961498e-06, + "loss": 0.9707, + "step": 4625 + }, + { + "epoch": 0.41718897957343193, + "grad_norm": 1.4124367151177593, + "learning_rate": 2.62303240318551e-06, + "loss": 0.9228, + "step": 4626 + }, + { + "epoch": 0.4172791630969022, + "grad_norm": 2.149421494300816, + "learning_rate": 2.62247723581431e-06, + "loss": 1.0637, + "step": 4627 + }, + { + "epoch": 0.41736934662037245, + "grad_norm": 1.5677795478470764, + "learning_rate": 2.62192201532992e-06, + "loss": 0.9949, + "step": 4628 + }, + { + "epoch": 0.41745953014384274, + "grad_norm": 1.574916334806203, + "learning_rate": 2.6213667417797145e-06, + "loss": 1.0121, + "step": 4629 + }, + { + "epoch": 0.41754971366731297, + "grad_norm": 1.3651406797639078, + "learning_rate": 2.6208114152110725e-06, + "loss": 1.0191, + "step": 4630 + }, + { + "epoch": 0.41763989719078326, + "grad_norm": 1.6213900008154307, + "learning_rate": 2.6202560356713774e-06, + "loss": 1.0088, + "step": 4631 + }, + { + "epoch": 0.4177300807142535, + "grad_norm": 0.6430317167688353, + "learning_rate": 2.619700603208017e-06, + "loss": 0.8762, + "step": 4632 + }, + { + "epoch": 0.4178202642377238, + "grad_norm": 1.577349632822812, + "learning_rate": 2.6191451178683842e-06, + "loss": 0.9076, + "step": 4633 + }, + { + "epoch": 0.417910447761194, + "grad_norm": 1.6640499633889925, + "learning_rate": 2.6185895796998764e-06, + "loss": 1.0252, + "step": 4634 + }, + { + "epoch": 0.4180006312846643, + "grad_norm": 1.7401114783256653, + "learning_rate": 2.6180339887498946e-06, + "loss": 1.0637, + "step": 4635 + }, + { + "epoch": 0.41809081480813454, + "grad_norm": 1.9617273935897042, + "learning_rate": 2.617478345065846e-06, + "loss": 0.9336, + "step": 4636 + }, + { + "epoch": 0.4181809983316048, + "grad_norm": 1.7956019660383182, + "learning_rate": 2.616922648695139e-06, + "loss": 1.0116, + "step": 4637 + }, + { + "epoch": 0.41827118185507506, + "grad_norm": 2.1608311735241217, + "learning_rate": 2.61636689968519e-06, + "loss": 0.8881, + "step": 4638 + }, + { + "epoch": 0.41836136537854535, + "grad_norm": 1.775590639839328, + "learning_rate": 2.6158110980834186e-06, + "loss": 1.0559, + "step": 4639 + }, + { + "epoch": 0.4184515489020156, + "grad_norm": 1.7698721652015987, + "learning_rate": 2.615255243937249e-06, + "loss": 1.0463, + "step": 4640 + }, + { + "epoch": 0.41854173242548587, + "grad_norm": 1.8821212322660774, + "learning_rate": 2.61469933729411e-06, + "loss": 1.0181, + "step": 4641 + }, + { + "epoch": 0.4186319159489561, + "grad_norm": 1.8157793173153127, + "learning_rate": 2.614143378201433e-06, + "loss": 0.9714, + "step": 4642 + }, + { + "epoch": 0.4187220994724264, + "grad_norm": 1.5704838293943948, + "learning_rate": 2.6135873667066567e-06, + "loss": 0.9748, + "step": 4643 + }, + { + "epoch": 0.4188122829958966, + "grad_norm": 1.615408540904845, + "learning_rate": 2.613031302857224e-06, + "loss": 0.953, + "step": 4644 + }, + { + "epoch": 0.4189024665193669, + "grad_norm": 1.6723624258665695, + "learning_rate": 2.6124751867005792e-06, + "loss": 1.0205, + "step": 4645 + }, + { + "epoch": 0.4189926500428372, + "grad_norm": 1.3007632417549202, + "learning_rate": 2.611919018284175e-06, + "loss": 1.0461, + "step": 4646 + }, + { + "epoch": 0.41908283356630743, + "grad_norm": 3.1259156983589738, + "learning_rate": 2.611362797655466e-06, + "loss": 1.0417, + "step": 4647 + }, + { + "epoch": 0.4191730170897777, + "grad_norm": 1.3774344827125715, + "learning_rate": 2.6108065248619124e-06, + "loss": 1.0542, + "step": 4648 + }, + { + "epoch": 0.41926320061324795, + "grad_norm": 1.4297147654450577, + "learning_rate": 2.610250199950978e-06, + "loss": 0.9302, + "step": 4649 + }, + { + "epoch": 0.41935338413671824, + "grad_norm": 1.755327697981498, + "learning_rate": 2.609693822970131e-06, + "loss": 0.9677, + "step": 4650 + }, + { + "epoch": 0.4194435676601885, + "grad_norm": 2.044941228940768, + "learning_rate": 2.609137393966846e-06, + "loss": 0.9617, + "step": 4651 + }, + { + "epoch": 0.41953375118365877, + "grad_norm": 1.6488498696251477, + "learning_rate": 2.6085809129886e-06, + "loss": 0.947, + "step": 4652 + }, + { + "epoch": 0.419623934707129, + "grad_norm": 1.9883799350432914, + "learning_rate": 2.608024380082874e-06, + "loss": 1.0057, + "step": 4653 + }, + { + "epoch": 0.4197141182305993, + "grad_norm": 2.143886114753922, + "learning_rate": 2.6074677952971554e-06, + "loss": 0.9922, + "step": 4654 + }, + { + "epoch": 0.4198043017540695, + "grad_norm": 1.5905146952156781, + "learning_rate": 2.606911158678935e-06, + "loss": 0.915, + "step": 4655 + }, + { + "epoch": 0.4198944852775398, + "grad_norm": 2.9360916332736435, + "learning_rate": 2.606354470275708e-06, + "loss": 0.9793, + "step": 4656 + }, + { + "epoch": 0.41998466880101004, + "grad_norm": 1.6149739028900718, + "learning_rate": 2.6057977301349744e-06, + "loss": 0.9657, + "step": 4657 + }, + { + "epoch": 0.42007485232448033, + "grad_norm": 1.8206881025112391, + "learning_rate": 2.6052409383042383e-06, + "loss": 0.9566, + "step": 4658 + }, + { + "epoch": 0.42016503584795056, + "grad_norm": 1.5324629825520928, + "learning_rate": 2.6046840948310074e-06, + "loss": 0.9702, + "step": 4659 + }, + { + "epoch": 0.42025521937142085, + "grad_norm": 1.6533686928787767, + "learning_rate": 2.6041271997627962e-06, + "loss": 0.9937, + "step": 4660 + }, + { + "epoch": 0.4203454028948911, + "grad_norm": 1.8709839726938524, + "learning_rate": 2.6035702531471202e-06, + "loss": 0.9828, + "step": 4661 + }, + { + "epoch": 0.4204355864183614, + "grad_norm": 1.4468477085909417, + "learning_rate": 2.6030132550315035e-06, + "loss": 0.9893, + "step": 4662 + }, + { + "epoch": 0.4205257699418316, + "grad_norm": 0.6698172128847403, + "learning_rate": 2.60245620546347e-06, + "loss": 0.8507, + "step": 4663 + }, + { + "epoch": 0.4206159534653019, + "grad_norm": 1.6401713152703497, + "learning_rate": 2.6018991044905517e-06, + "loss": 0.9681, + "step": 4664 + }, + { + "epoch": 0.42070613698877213, + "grad_norm": 1.563854088555108, + "learning_rate": 2.6013419521602825e-06, + "loss": 1.0267, + "step": 4665 + }, + { + "epoch": 0.4207963205122424, + "grad_norm": 1.601000803630944, + "learning_rate": 2.600784748520202e-06, + "loss": 1.1218, + "step": 4666 + }, + { + "epoch": 0.42088650403571265, + "grad_norm": 1.5341457312193423, + "learning_rate": 2.6002274936178544e-06, + "loss": 0.9279, + "step": 4667 + }, + { + "epoch": 0.42097668755918294, + "grad_norm": 1.5555330210521858, + "learning_rate": 2.5996701875007873e-06, + "loss": 0.9986, + "step": 4668 + }, + { + "epoch": 0.4210668710826532, + "grad_norm": 1.471903560547611, + "learning_rate": 2.5991128302165533e-06, + "loss": 0.981, + "step": 4669 + }, + { + "epoch": 0.42115705460612346, + "grad_norm": 1.4642734892597342, + "learning_rate": 2.5985554218127094e-06, + "loss": 1.0528, + "step": 4670 + }, + { + "epoch": 0.42124723812959375, + "grad_norm": 2.112272293188089, + "learning_rate": 2.597997962336816e-06, + "loss": 0.9248, + "step": 4671 + }, + { + "epoch": 0.421337421653064, + "grad_norm": 1.8732491902067459, + "learning_rate": 2.5974404518364393e-06, + "loss": 0.9835, + "step": 4672 + }, + { + "epoch": 0.42142760517653427, + "grad_norm": 1.69883825492514, + "learning_rate": 2.596882890359149e-06, + "loss": 1.1163, + "step": 4673 + }, + { + "epoch": 0.4215177887000045, + "grad_norm": 2.280309404828759, + "learning_rate": 2.5963252779525196e-06, + "loss": 0.912, + "step": 4674 + }, + { + "epoch": 0.4216079722234748, + "grad_norm": 1.9109865346061878, + "learning_rate": 2.595767614664129e-06, + "loss": 0.9975, + "step": 4675 + }, + { + "epoch": 0.421698155746945, + "grad_norm": 1.545234658158169, + "learning_rate": 2.5952099005415607e-06, + "loss": 0.968, + "step": 4676 + }, + { + "epoch": 0.4217883392704153, + "grad_norm": 5.1470074671899955, + "learning_rate": 2.594652135632402e-06, + "loss": 0.9534, + "step": 4677 + }, + { + "epoch": 0.42187852279388555, + "grad_norm": 1.343205585644998, + "learning_rate": 2.594094319984244e-06, + "loss": 0.9502, + "step": 4678 + }, + { + "epoch": 0.42196870631735584, + "grad_norm": 2.1185252337669187, + "learning_rate": 2.5935364536446825e-06, + "loss": 1.0494, + "step": 4679 + }, + { + "epoch": 0.42205888984082607, + "grad_norm": 1.7131136637009556, + "learning_rate": 2.5929785366613185e-06, + "loss": 0.9907, + "step": 4680 + }, + { + "epoch": 0.42214907336429636, + "grad_norm": 2.0312543428814753, + "learning_rate": 2.592420569081756e-06, + "loss": 1.0247, + "step": 4681 + }, + { + "epoch": 0.4222392568877666, + "grad_norm": 2.213376482615731, + "learning_rate": 2.5918625509536037e-06, + "loss": 0.9563, + "step": 4682 + }, + { + "epoch": 0.4223294404112369, + "grad_norm": 0.6587169237899204, + "learning_rate": 2.591304482324475e-06, + "loss": 0.8159, + "step": 4683 + }, + { + "epoch": 0.4224196239347071, + "grad_norm": 2.238064048041518, + "learning_rate": 2.5907463632419878e-06, + "loss": 1.0025, + "step": 4684 + }, + { + "epoch": 0.4225098074581774, + "grad_norm": 2.6879191404453575, + "learning_rate": 2.5901881937537632e-06, + "loss": 0.9599, + "step": 4685 + }, + { + "epoch": 0.42259999098164763, + "grad_norm": 0.7341470161590911, + "learning_rate": 2.589629973907428e-06, + "loss": 0.8491, + "step": 4686 + }, + { + "epoch": 0.4226901745051179, + "grad_norm": 1.646488067514209, + "learning_rate": 2.589071703750612e-06, + "loss": 0.942, + "step": 4687 + }, + { + "epoch": 0.42278035802858815, + "grad_norm": 1.334749597146759, + "learning_rate": 2.5885133833309504e-06, + "loss": 1.0335, + "step": 4688 + }, + { + "epoch": 0.42287054155205844, + "grad_norm": 1.5722695676386902, + "learning_rate": 2.5879550126960814e-06, + "loss": 1.071, + "step": 4689 + }, + { + "epoch": 0.4229607250755287, + "grad_norm": 1.4760121122033836, + "learning_rate": 2.5873965918936494e-06, + "loss": 1.0504, + "step": 4690 + }, + { + "epoch": 0.42305090859899896, + "grad_norm": 1.7298720732561534, + "learning_rate": 2.586838120971301e-06, + "loss": 1.019, + "step": 4691 + }, + { + "epoch": 0.4231410921224692, + "grad_norm": 1.8651727636257684, + "learning_rate": 2.586279599976689e-06, + "loss": 1.054, + "step": 4692 + }, + { + "epoch": 0.4232312756459395, + "grad_norm": 1.4588537195984241, + "learning_rate": 2.585721028957468e-06, + "loss": 0.99, + "step": 4693 + }, + { + "epoch": 0.4233214591694098, + "grad_norm": 1.3901165664975466, + "learning_rate": 2.585162407961299e-06, + "loss": 1.0025, + "step": 4694 + }, + { + "epoch": 0.42341164269288, + "grad_norm": 1.905021521779637, + "learning_rate": 2.584603737035847e-06, + "loss": 0.9318, + "step": 4695 + }, + { + "epoch": 0.4235018262163503, + "grad_norm": 1.6804658018092529, + "learning_rate": 2.5840450162287806e-06, + "loss": 0.9861, + "step": 4696 + }, + { + "epoch": 0.42359200973982053, + "grad_norm": 2.212760470300792, + "learning_rate": 2.583486245587774e-06, + "loss": 0.9031, + "step": 4697 + }, + { + "epoch": 0.4236821932632908, + "grad_norm": 1.7788800987764055, + "learning_rate": 2.5829274251605023e-06, + "loss": 1.0969, + "step": 4698 + }, + { + "epoch": 0.42377237678676105, + "grad_norm": 1.4204010082906262, + "learning_rate": 2.582368554994649e-06, + "loss": 0.9621, + "step": 4699 + }, + { + "epoch": 0.42386256031023134, + "grad_norm": 0.7526048882468878, + "learning_rate": 2.5818096351378994e-06, + "loss": 0.8513, + "step": 4700 + }, + { + "epoch": 0.4239527438337016, + "grad_norm": 2.1265380005893237, + "learning_rate": 2.5812506656379435e-06, + "loss": 1.0482, + "step": 4701 + }, + { + "epoch": 0.42404292735717186, + "grad_norm": 1.7942273744174702, + "learning_rate": 2.580691646542476e-06, + "loss": 0.9411, + "step": 4702 + }, + { + "epoch": 0.4241331108806421, + "grad_norm": 1.6103323894259234, + "learning_rate": 2.5801325778991958e-06, + "loss": 1.0273, + "step": 4703 + }, + { + "epoch": 0.4242232944041124, + "grad_norm": 1.8763420388394831, + "learning_rate": 2.5795734597558043e-06, + "loss": 0.8487, + "step": 4704 + }, + { + "epoch": 0.4243134779275826, + "grad_norm": 2.0205622094613833, + "learning_rate": 2.579014292160011e-06, + "loss": 0.9911, + "step": 4705 + }, + { + "epoch": 0.4244036614510529, + "grad_norm": 1.2815605345503742, + "learning_rate": 2.5784550751595236e-06, + "loss": 0.9736, + "step": 4706 + }, + { + "epoch": 0.42449384497452314, + "grad_norm": 1.3449821589396291, + "learning_rate": 2.577895808802061e-06, + "loss": 1.051, + "step": 4707 + }, + { + "epoch": 0.4245840284979934, + "grad_norm": 22.537017888198267, + "learning_rate": 2.577336493135341e-06, + "loss": 0.9108, + "step": 4708 + }, + { + "epoch": 0.42467421202146366, + "grad_norm": 1.3327567622167584, + "learning_rate": 2.576777128207088e-06, + "loss": 1.0894, + "step": 4709 + }, + { + "epoch": 0.42476439554493395, + "grad_norm": 1.4161763277613582, + "learning_rate": 2.5762177140650306e-06, + "loss": 0.9535, + "step": 4710 + }, + { + "epoch": 0.4248545790684042, + "grad_norm": 1.8325932988749634, + "learning_rate": 2.5756582507569003e-06, + "loss": 1.0022, + "step": 4711 + }, + { + "epoch": 0.42494476259187447, + "grad_norm": 1.5112069616507768, + "learning_rate": 2.5750987383304335e-06, + "loss": 1.0162, + "step": 4712 + }, + { + "epoch": 0.4250349461153447, + "grad_norm": 2.4090516265493114, + "learning_rate": 2.574539176833372e-06, + "loss": 0.9771, + "step": 4713 + }, + { + "epoch": 0.425125129638815, + "grad_norm": 1.9601254548580171, + "learning_rate": 2.5739795663134594e-06, + "loss": 0.9352, + "step": 4714 + }, + { + "epoch": 0.4252153131622852, + "grad_norm": 2.257363561856337, + "learning_rate": 2.5734199068184454e-06, + "loss": 1.0795, + "step": 4715 + }, + { + "epoch": 0.4253054966857555, + "grad_norm": 1.5825119230234035, + "learning_rate": 2.572860198396083e-06, + "loss": 0.9162, + "step": 4716 + }, + { + "epoch": 0.4253956802092258, + "grad_norm": 1.4110569917049476, + "learning_rate": 2.57230044109413e-06, + "loss": 1.0155, + "step": 4717 + }, + { + "epoch": 0.42548586373269603, + "grad_norm": 1.9878732078960109, + "learning_rate": 2.5717406349603483e-06, + "loss": 0.9738, + "step": 4718 + }, + { + "epoch": 0.4255760472561663, + "grad_norm": 1.464076133509844, + "learning_rate": 2.5711807800425026e-06, + "loss": 1.0126, + "step": 4719 + }, + { + "epoch": 0.42566623077963656, + "grad_norm": 2.721947904852059, + "learning_rate": 2.5706208763883633e-06, + "loss": 0.8292, + "step": 4720 + }, + { + "epoch": 0.42575641430310684, + "grad_norm": 1.6697529668323279, + "learning_rate": 2.570060924045704e-06, + "loss": 0.9515, + "step": 4721 + }, + { + "epoch": 0.4258465978265771, + "grad_norm": 1.6618000105702326, + "learning_rate": 2.569500923062304e-06, + "loss": 1.0251, + "step": 4722 + }, + { + "epoch": 0.42593678135004737, + "grad_norm": 1.8458511535381357, + "learning_rate": 2.5689408734859445e-06, + "loss": 0.9148, + "step": 4723 + }, + { + "epoch": 0.4260269648735176, + "grad_norm": 1.8020738443407787, + "learning_rate": 2.5683807753644127e-06, + "loss": 0.9405, + "step": 4724 + }, + { + "epoch": 0.4261171483969879, + "grad_norm": 1.5163739113992838, + "learning_rate": 2.5678206287454996e-06, + "loss": 0.969, + "step": 4725 + }, + { + "epoch": 0.4262073319204581, + "grad_norm": 2.371437110737775, + "learning_rate": 2.567260433676999e-06, + "loss": 1.029, + "step": 4726 + }, + { + "epoch": 0.4262975154439284, + "grad_norm": 1.6909600565753615, + "learning_rate": 2.5667001902067107e-06, + "loss": 0.9679, + "step": 4727 + }, + { + "epoch": 0.42638769896739864, + "grad_norm": 2.1567060568787735, + "learning_rate": 2.566139898382437e-06, + "loss": 1.0429, + "step": 4728 + }, + { + "epoch": 0.42647788249086893, + "grad_norm": 1.5495462184043403, + "learning_rate": 2.5655795582519853e-06, + "loss": 1.0059, + "step": 4729 + }, + { + "epoch": 0.42656806601433916, + "grad_norm": 1.5708800851085902, + "learning_rate": 2.565019169863168e-06, + "loss": 1.0606, + "step": 4730 + }, + { + "epoch": 0.42665824953780945, + "grad_norm": 0.5872315168707797, + "learning_rate": 2.5644587332637994e-06, + "loss": 0.785, + "step": 4731 + }, + { + "epoch": 0.4267484330612797, + "grad_norm": 1.4714174600226633, + "learning_rate": 2.5638982485016994e-06, + "loss": 1.02, + "step": 4732 + }, + { + "epoch": 0.42683861658475, + "grad_norm": 0.720525042440832, + "learning_rate": 2.5633377156246917e-06, + "loss": 0.7901, + "step": 4733 + }, + { + "epoch": 0.4269288001082202, + "grad_norm": 1.7038830811098533, + "learning_rate": 2.562777134680603e-06, + "loss": 0.9764, + "step": 4734 + }, + { + "epoch": 0.4270189836316905, + "grad_norm": 1.861262221235396, + "learning_rate": 2.562216505717267e-06, + "loss": 0.9958, + "step": 4735 + }, + { + "epoch": 0.42710916715516073, + "grad_norm": 1.5155062284829306, + "learning_rate": 2.561655828782518e-06, + "loss": 1.0173, + "step": 4736 + }, + { + "epoch": 0.427199350678631, + "grad_norm": 2.4447696558564638, + "learning_rate": 2.561095103924197e-06, + "loss": 0.9745, + "step": 4737 + }, + { + "epoch": 0.42728953420210125, + "grad_norm": 1.8827640004856312, + "learning_rate": 2.560534331190148e-06, + "loss": 0.9407, + "step": 4738 + }, + { + "epoch": 0.42737971772557154, + "grad_norm": 0.5774501521820573, + "learning_rate": 2.559973510628218e-06, + "loss": 0.7863, + "step": 4739 + }, + { + "epoch": 0.4274699012490418, + "grad_norm": 1.6253583219582417, + "learning_rate": 2.5594126422862615e-06, + "loss": 0.9382, + "step": 4740 + }, + { + "epoch": 0.42756008477251206, + "grad_norm": 2.532343122136576, + "learning_rate": 2.558851726212134e-06, + "loss": 0.9407, + "step": 4741 + }, + { + "epoch": 0.42765026829598235, + "grad_norm": 1.5716729036101573, + "learning_rate": 2.5582907624536953e-06, + "loss": 0.9665, + "step": 4742 + }, + { + "epoch": 0.4277404518194526, + "grad_norm": 1.8777197503761733, + "learning_rate": 2.557729751058811e-06, + "loss": 1.0084, + "step": 4743 + }, + { + "epoch": 0.42783063534292287, + "grad_norm": 1.6677324543085599, + "learning_rate": 2.557168692075348e-06, + "loss": 0.9686, + "step": 4744 + }, + { + "epoch": 0.4279208188663931, + "grad_norm": 1.443808685695332, + "learning_rate": 2.556607585551181e-06, + "loss": 0.9329, + "step": 4745 + }, + { + "epoch": 0.4280110023898634, + "grad_norm": 0.7520743532428383, + "learning_rate": 2.5560464315341844e-06, + "loss": 0.9148, + "step": 4746 + }, + { + "epoch": 0.4281011859133336, + "grad_norm": 1.838031305326697, + "learning_rate": 2.555485230072242e-06, + "loss": 0.9863, + "step": 4747 + }, + { + "epoch": 0.4281913694368039, + "grad_norm": 1.9250722202773827, + "learning_rate": 2.5549239812132354e-06, + "loss": 1.038, + "step": 4748 + }, + { + "epoch": 0.42828155296027415, + "grad_norm": 1.5672413414542385, + "learning_rate": 2.5543626850050556e-06, + "loss": 1.0585, + "step": 4749 + }, + { + "epoch": 0.42837173648374444, + "grad_norm": 1.4996512325453883, + "learning_rate": 2.5538013414955944e-06, + "loss": 1.045, + "step": 4750 + }, + { + "epoch": 0.42846192000721467, + "grad_norm": 1.526596784255495, + "learning_rate": 2.5532399507327494e-06, + "loss": 0.9462, + "step": 4751 + }, + { + "epoch": 0.42855210353068496, + "grad_norm": 0.6026628017885349, + "learning_rate": 2.552678512764421e-06, + "loss": 0.8232, + "step": 4752 + }, + { + "epoch": 0.4286422870541552, + "grad_norm": 3.873123945353329, + "learning_rate": 2.5521170276385147e-06, + "loss": 1.0407, + "step": 4753 + }, + { + "epoch": 0.4287324705776255, + "grad_norm": 2.4251093770952363, + "learning_rate": 2.5515554954029394e-06, + "loss": 1.0401, + "step": 4754 + }, + { + "epoch": 0.4288226541010957, + "grad_norm": 1.5383423800355696, + "learning_rate": 2.550993916105608e-06, + "loss": 1.0165, + "step": 4755 + }, + { + "epoch": 0.428912837624566, + "grad_norm": 1.3710802735103813, + "learning_rate": 2.550432289794437e-06, + "loss": 0.9474, + "step": 4756 + }, + { + "epoch": 0.42900302114803623, + "grad_norm": 1.5769246890433768, + "learning_rate": 2.5498706165173483e-06, + "loss": 1.0013, + "step": 4757 + }, + { + "epoch": 0.4290932046715065, + "grad_norm": 1.3144134470150208, + "learning_rate": 2.5493088963222668e-06, + "loss": 0.9666, + "step": 4758 + }, + { + "epoch": 0.42918338819497676, + "grad_norm": 1.485271062933077, + "learning_rate": 2.548747129257121e-06, + "loss": 1.0398, + "step": 4759 + }, + { + "epoch": 0.42927357171844704, + "grad_norm": 1.5298664303986933, + "learning_rate": 2.548185315369845e-06, + "loss": 0.9138, + "step": 4760 + }, + { + "epoch": 0.4293637552419173, + "grad_norm": 1.4228369843646629, + "learning_rate": 2.5476234547083746e-06, + "loss": 1.0775, + "step": 4761 + }, + { + "epoch": 0.42945393876538757, + "grad_norm": 1.6441617355749651, + "learning_rate": 2.547061547320652e-06, + "loss": 1.0627, + "step": 4762 + }, + { + "epoch": 0.4295441222888578, + "grad_norm": 1.5854631287646266, + "learning_rate": 2.5464995932546217e-06, + "loss": 0.9731, + "step": 4763 + }, + { + "epoch": 0.4296343058123281, + "grad_norm": 1.7004576627984511, + "learning_rate": 2.545937592558232e-06, + "loss": 0.9426, + "step": 4764 + }, + { + "epoch": 0.4297244893357984, + "grad_norm": 1.5025955631884689, + "learning_rate": 2.5453755452794374e-06, + "loss": 0.9394, + "step": 4765 + }, + { + "epoch": 0.4298146728592686, + "grad_norm": 1.5811439569069539, + "learning_rate": 2.5448134514661938e-06, + "loss": 1.0355, + "step": 4766 + }, + { + "epoch": 0.4299048563827389, + "grad_norm": 1.3836308875503605, + "learning_rate": 2.5442513111664623e-06, + "loss": 0.9879, + "step": 4767 + }, + { + "epoch": 0.42999503990620913, + "grad_norm": 1.6809810182179359, + "learning_rate": 2.5436891244282084e-06, + "loss": 0.9853, + "step": 4768 + }, + { + "epoch": 0.4300852234296794, + "grad_norm": 1.719445313347491, + "learning_rate": 2.5431268912994004e-06, + "loss": 1.0919, + "step": 4769 + }, + { + "epoch": 0.43017540695314965, + "grad_norm": 1.6792832708772711, + "learning_rate": 2.5425646118280108e-06, + "loss": 0.9583, + "step": 4770 + }, + { + "epoch": 0.43026559047661994, + "grad_norm": 0.5521658679759871, + "learning_rate": 2.5420022860620172e-06, + "loss": 0.7997, + "step": 4771 + }, + { + "epoch": 0.4303557740000902, + "grad_norm": 1.9485939874649667, + "learning_rate": 2.5414399140493995e-06, + "loss": 0.9231, + "step": 4772 + }, + { + "epoch": 0.43044595752356046, + "grad_norm": 1.7970358071726735, + "learning_rate": 2.5408774958381436e-06, + "loss": 0.9267, + "step": 4773 + }, + { + "epoch": 0.4305361410470307, + "grad_norm": 1.5389627337191656, + "learning_rate": 2.540315031476237e-06, + "loss": 1.0229, + "step": 4774 + }, + { + "epoch": 0.430626324570501, + "grad_norm": 1.6796661375594677, + "learning_rate": 2.5397525210116737e-06, + "loss": 0.9965, + "step": 4775 + }, + { + "epoch": 0.4307165080939712, + "grad_norm": 1.5226677930853352, + "learning_rate": 2.539189964492448e-06, + "loss": 0.949, + "step": 4776 + }, + { + "epoch": 0.4308066916174415, + "grad_norm": 1.471090197444927, + "learning_rate": 2.5386273619665613e-06, + "loss": 1.0319, + "step": 4777 + }, + { + "epoch": 0.43089687514091174, + "grad_norm": 0.6839860716776389, + "learning_rate": 2.5380647134820186e-06, + "loss": 0.8216, + "step": 4778 + }, + { + "epoch": 0.43098705866438203, + "grad_norm": 1.4231830490931925, + "learning_rate": 2.5375020190868277e-06, + "loss": 1.0456, + "step": 4779 + }, + { + "epoch": 0.43107724218785226, + "grad_norm": 1.5384673696187359, + "learning_rate": 2.536939278829001e-06, + "loss": 0.9819, + "step": 4780 + }, + { + "epoch": 0.43116742571132255, + "grad_norm": 1.709773673756972, + "learning_rate": 2.5363764927565536e-06, + "loss": 1.0492, + "step": 4781 + }, + { + "epoch": 0.4312576092347928, + "grad_norm": 1.9047279535117445, + "learning_rate": 2.5358136609175064e-06, + "loss": 0.9474, + "step": 4782 + }, + { + "epoch": 0.43134779275826307, + "grad_norm": 1.5274217980777818, + "learning_rate": 2.535250783359884e-06, + "loss": 0.9921, + "step": 4783 + }, + { + "epoch": 0.4314379762817333, + "grad_norm": 1.374169358867154, + "learning_rate": 2.5346878601317124e-06, + "loss": 0.8848, + "step": 4784 + }, + { + "epoch": 0.4315281598052036, + "grad_norm": 0.5864056306627857, + "learning_rate": 2.534124891281025e-06, + "loss": 0.7846, + "step": 4785 + }, + { + "epoch": 0.4316183433286738, + "grad_norm": 1.645066706455688, + "learning_rate": 2.533561876855857e-06, + "loss": 1.0001, + "step": 4786 + }, + { + "epoch": 0.4317085268521441, + "grad_norm": 1.4703864862391776, + "learning_rate": 2.532998816904247e-06, + "loss": 0.9413, + "step": 4787 + }, + { + "epoch": 0.4317987103756144, + "grad_norm": 1.5703950167558904, + "learning_rate": 2.53243571147424e-06, + "loss": 0.8824, + "step": 4788 + }, + { + "epoch": 0.43188889389908464, + "grad_norm": 1.8553851620769073, + "learning_rate": 2.5318725606138815e-06, + "loss": 0.9001, + "step": 4789 + }, + { + "epoch": 0.4319790774225549, + "grad_norm": 1.9901186504168078, + "learning_rate": 2.5313093643712235e-06, + "loss": 0.9958, + "step": 4790 + }, + { + "epoch": 0.43206926094602516, + "grad_norm": 1.3714268380321606, + "learning_rate": 2.530746122794321e-06, + "loss": 1.0357, + "step": 4791 + }, + { + "epoch": 0.43215944446949545, + "grad_norm": 1.4366445898733287, + "learning_rate": 2.5301828359312323e-06, + "loss": 0.9226, + "step": 4792 + }, + { + "epoch": 0.4322496279929657, + "grad_norm": 1.5325695502475003, + "learning_rate": 2.529619503830021e-06, + "loss": 0.95, + "step": 4793 + }, + { + "epoch": 0.43233981151643597, + "grad_norm": 2.026530725705032, + "learning_rate": 2.529056126538753e-06, + "loss": 0.9213, + "step": 4794 + }, + { + "epoch": 0.4324299950399062, + "grad_norm": 1.5767147448160639, + "learning_rate": 2.5284927041054995e-06, + "loss": 1.0202, + "step": 4795 + }, + { + "epoch": 0.4325201785633765, + "grad_norm": 0.7113382656122452, + "learning_rate": 2.5279292365783348e-06, + "loss": 0.7624, + "step": 4796 + }, + { + "epoch": 0.4326103620868467, + "grad_norm": 1.5024491819143933, + "learning_rate": 2.527365724005336e-06, + "loss": 1.0771, + "step": 4797 + }, + { + "epoch": 0.432700545610317, + "grad_norm": 1.59945927662119, + "learning_rate": 2.526802166434586e-06, + "loss": 1.0441, + "step": 4798 + }, + { + "epoch": 0.43279072913378724, + "grad_norm": 1.3784322815603665, + "learning_rate": 2.5262385639141708e-06, + "loss": 0.9554, + "step": 4799 + }, + { + "epoch": 0.43288091265725753, + "grad_norm": 1.5384793023771146, + "learning_rate": 2.525674916492179e-06, + "loss": 1.0599, + "step": 4800 + }, + { + "epoch": 0.43297109618072777, + "grad_norm": 1.3256720386092862, + "learning_rate": 2.5251112242167056e-06, + "loss": 0.9963, + "step": 4801 + }, + { + "epoch": 0.43306127970419805, + "grad_norm": 1.7137083148407992, + "learning_rate": 2.5245474871358464e-06, + "loss": 0.95, + "step": 4802 + }, + { + "epoch": 0.4331514632276683, + "grad_norm": 1.7711446451043562, + "learning_rate": 2.5239837052977037e-06, + "loss": 0.9083, + "step": 4803 + }, + { + "epoch": 0.4332416467511386, + "grad_norm": 1.7520374970974084, + "learning_rate": 2.523419878750381e-06, + "loss": 1.0997, + "step": 4804 + }, + { + "epoch": 0.4333318302746088, + "grad_norm": 1.553121573676584, + "learning_rate": 2.522856007541989e-06, + "loss": 0.9405, + "step": 4805 + }, + { + "epoch": 0.4334220137980791, + "grad_norm": 1.6287682929954943, + "learning_rate": 2.5222920917206397e-06, + "loss": 0.8999, + "step": 4806 + }, + { + "epoch": 0.43351219732154933, + "grad_norm": 0.613784146495266, + "learning_rate": 2.5217281313344493e-06, + "loss": 0.8001, + "step": 4807 + }, + { + "epoch": 0.4336023808450196, + "grad_norm": 1.812247883592723, + "learning_rate": 2.5211641264315372e-06, + "loss": 1.0367, + "step": 4808 + }, + { + "epoch": 0.43369256436848985, + "grad_norm": 1.5170345241060639, + "learning_rate": 2.5206000770600286e-06, + "loss": 0.9688, + "step": 4809 + }, + { + "epoch": 0.43378274789196014, + "grad_norm": 1.6747298065439564, + "learning_rate": 2.520035983268051e-06, + "loss": 0.9473, + "step": 4810 + }, + { + "epoch": 0.4338729314154304, + "grad_norm": 1.5044216313284517, + "learning_rate": 2.5194718451037357e-06, + "loss": 0.9914, + "step": 4811 + }, + { + "epoch": 0.43396311493890066, + "grad_norm": 4.263923277631681, + "learning_rate": 2.518907662615218e-06, + "loss": 1.0183, + "step": 4812 + }, + { + "epoch": 0.43405329846237095, + "grad_norm": 1.5976149126968113, + "learning_rate": 2.5183434358506373e-06, + "loss": 0.9507, + "step": 4813 + }, + { + "epoch": 0.4341434819858412, + "grad_norm": 1.4548848478231584, + "learning_rate": 2.5177791648581368e-06, + "loss": 0.9733, + "step": 4814 + }, + { + "epoch": 0.4342336655093115, + "grad_norm": 1.6190086835565893, + "learning_rate": 2.517214849685863e-06, + "loss": 0.9894, + "step": 4815 + }, + { + "epoch": 0.4343238490327817, + "grad_norm": 1.649828644726797, + "learning_rate": 2.5166504903819663e-06, + "loss": 1.0064, + "step": 4816 + }, + { + "epoch": 0.434414032556252, + "grad_norm": 2.1799020969210163, + "learning_rate": 2.5160860869946014e-06, + "loss": 1.0617, + "step": 4817 + }, + { + "epoch": 0.4345042160797222, + "grad_norm": 1.4561122419826815, + "learning_rate": 2.5155216395719253e-06, + "loss": 1.0716, + "step": 4818 + }, + { + "epoch": 0.4345943996031925, + "grad_norm": 3.3031492869369363, + "learning_rate": 2.5149571481621e-06, + "loss": 0.9502, + "step": 4819 + }, + { + "epoch": 0.43468458312666275, + "grad_norm": 0.5937451061247868, + "learning_rate": 2.514392612813292e-06, + "loss": 0.7817, + "step": 4820 + }, + { + "epoch": 0.43477476665013304, + "grad_norm": 1.5132057308954696, + "learning_rate": 2.5138280335736695e-06, + "loss": 1.0465, + "step": 4821 + }, + { + "epoch": 0.43486495017360327, + "grad_norm": 1.4933956508037238, + "learning_rate": 2.5132634104914064e-06, + "loss": 1.0724, + "step": 4822 + }, + { + "epoch": 0.43495513369707356, + "grad_norm": 1.7682487100936002, + "learning_rate": 2.5126987436146794e-06, + "loss": 1.0024, + "step": 4823 + }, + { + "epoch": 0.4350453172205438, + "grad_norm": 1.5045953930385836, + "learning_rate": 2.5121340329916675e-06, + "loss": 0.9639, + "step": 4824 + }, + { + "epoch": 0.4351355007440141, + "grad_norm": 1.4996525044063267, + "learning_rate": 2.5115692786705566e-06, + "loss": 0.9104, + "step": 4825 + }, + { + "epoch": 0.4352256842674843, + "grad_norm": 2.498237179090633, + "learning_rate": 2.511004480699534e-06, + "loss": 1.0231, + "step": 4826 + }, + { + "epoch": 0.4353158677909546, + "grad_norm": 0.7574639846247081, + "learning_rate": 2.510439639126791e-06, + "loss": 0.8567, + "step": 4827 + }, + { + "epoch": 0.43540605131442484, + "grad_norm": 1.5078060960386592, + "learning_rate": 2.509874754000524e-06, + "loss": 0.9872, + "step": 4828 + }, + { + "epoch": 0.4354962348378951, + "grad_norm": 1.764973545578632, + "learning_rate": 2.509309825368932e-06, + "loss": 0.9785, + "step": 4829 + }, + { + "epoch": 0.43558641836136536, + "grad_norm": 1.5054783914333523, + "learning_rate": 2.5087448532802173e-06, + "loss": 0.9909, + "step": 4830 + }, + { + "epoch": 0.43567660188483565, + "grad_norm": 1.6866609818786193, + "learning_rate": 2.508179837782586e-06, + "loss": 0.9678, + "step": 4831 + }, + { + "epoch": 0.4357667854083059, + "grad_norm": 1.545476416060788, + "learning_rate": 2.5076147789242493e-06, + "loss": 0.9613, + "step": 4832 + }, + { + "epoch": 0.43585696893177617, + "grad_norm": 2.181659421780036, + "learning_rate": 2.5070496767534202e-06, + "loss": 1.0063, + "step": 4833 + }, + { + "epoch": 0.4359471524552464, + "grad_norm": 1.6941173929209612, + "learning_rate": 2.506484531318317e-06, + "loss": 1.0173, + "step": 4834 + }, + { + "epoch": 0.4360373359787167, + "grad_norm": 0.7670888724706415, + "learning_rate": 2.5059193426671613e-06, + "loss": 0.8287, + "step": 4835 + }, + { + "epoch": 0.436127519502187, + "grad_norm": 1.8111658118606861, + "learning_rate": 2.5053541108481772e-06, + "loss": 1.1238, + "step": 4836 + }, + { + "epoch": 0.4362177030256572, + "grad_norm": 1.912901670374578, + "learning_rate": 2.5047888359095935e-06, + "loss": 0.9728, + "step": 4837 + }, + { + "epoch": 0.4363078865491275, + "grad_norm": 1.568850543687041, + "learning_rate": 2.5042235178996436e-06, + "loss": 0.9733, + "step": 4838 + }, + { + "epoch": 0.43639807007259773, + "grad_norm": 1.678639781568955, + "learning_rate": 2.5036581568665627e-06, + "loss": 1.0551, + "step": 4839 + }, + { + "epoch": 0.436488253596068, + "grad_norm": 1.5342190043597501, + "learning_rate": 2.503092752858591e-06, + "loss": 1.0528, + "step": 4840 + }, + { + "epoch": 0.43657843711953825, + "grad_norm": 0.7736587978468543, + "learning_rate": 2.502527305923971e-06, + "loss": 0.8454, + "step": 4841 + }, + { + "epoch": 0.43666862064300854, + "grad_norm": 2.059447486076422, + "learning_rate": 2.5019618161109506e-06, + "loss": 0.9737, + "step": 4842 + }, + { + "epoch": 0.4367588041664788, + "grad_norm": 1.788059603186233, + "learning_rate": 2.5013962834677804e-06, + "loss": 1.0077, + "step": 4843 + }, + { + "epoch": 0.43684898768994906, + "grad_norm": 1.8000546394108434, + "learning_rate": 2.500830708042715e-06, + "loss": 0.9779, + "step": 4844 + }, + { + "epoch": 0.4369391712134193, + "grad_norm": 1.5870011995307414, + "learning_rate": 2.500265089884011e-06, + "loss": 1.0524, + "step": 4845 + }, + { + "epoch": 0.4370293547368896, + "grad_norm": 1.5330268515557521, + "learning_rate": 2.499699429039932e-06, + "loss": 0.9907, + "step": 4846 + }, + { + "epoch": 0.4371195382603598, + "grad_norm": 1.6561911230598645, + "learning_rate": 2.4991337255587425e-06, + "loss": 0.9412, + "step": 4847 + }, + { + "epoch": 0.4372097217838301, + "grad_norm": 1.7481894664040758, + "learning_rate": 2.4985679794887106e-06, + "loss": 0.8581, + "step": 4848 + }, + { + "epoch": 0.43729990530730034, + "grad_norm": 1.8854401167224426, + "learning_rate": 2.49800219087811e-06, + "loss": 1.0211, + "step": 4849 + }, + { + "epoch": 0.43739008883077063, + "grad_norm": 1.8846888393118402, + "learning_rate": 2.4974363597752163e-06, + "loss": 1.027, + "step": 4850 + }, + { + "epoch": 0.43748027235424086, + "grad_norm": 1.7689417516068762, + "learning_rate": 2.4968704862283097e-06, + "loss": 1.037, + "step": 4851 + }, + { + "epoch": 0.43757045587771115, + "grad_norm": 1.361838684228072, + "learning_rate": 2.4963045702856737e-06, + "loss": 1.0597, + "step": 4852 + }, + { + "epoch": 0.4376606394011814, + "grad_norm": 1.6708232387830761, + "learning_rate": 2.4957386119955946e-06, + "loss": 0.9332, + "step": 4853 + }, + { + "epoch": 0.4377508229246517, + "grad_norm": 0.6629525950862005, + "learning_rate": 2.495172611406364e-06, + "loss": 0.8061, + "step": 4854 + }, + { + "epoch": 0.4378410064481219, + "grad_norm": 1.8989244629278337, + "learning_rate": 2.4946065685662757e-06, + "loss": 1.021, + "step": 4855 + }, + { + "epoch": 0.4379311899715922, + "grad_norm": 2.111978802914163, + "learning_rate": 2.4940404835236283e-06, + "loss": 0.9889, + "step": 4856 + }, + { + "epoch": 0.4380213734950624, + "grad_norm": 1.3484005848946967, + "learning_rate": 2.4934743563267223e-06, + "loss": 0.9778, + "step": 4857 + }, + { + "epoch": 0.4381115570185327, + "grad_norm": 2.2566802145674765, + "learning_rate": 2.4929081870238635e-06, + "loss": 0.953, + "step": 4858 + }, + { + "epoch": 0.43820174054200295, + "grad_norm": 1.5258204676309493, + "learning_rate": 2.49234197566336e-06, + "loss": 1.0273, + "step": 4859 + }, + { + "epoch": 0.43829192406547324, + "grad_norm": 1.556568407324361, + "learning_rate": 2.4917757222935247e-06, + "loss": 1.0102, + "step": 4860 + }, + { + "epoch": 0.4383821075889435, + "grad_norm": 1.8292430490984983, + "learning_rate": 2.4912094269626725e-06, + "loss": 1.0093, + "step": 4861 + }, + { + "epoch": 0.43847229111241376, + "grad_norm": 1.7758245326086202, + "learning_rate": 2.4906430897191245e-06, + "loss": 1.0302, + "step": 4862 + }, + { + "epoch": 0.43856247463588405, + "grad_norm": 1.9477989514799396, + "learning_rate": 2.490076710611202e-06, + "loss": 1.0271, + "step": 4863 + }, + { + "epoch": 0.4386526581593543, + "grad_norm": 1.8206808347878263, + "learning_rate": 2.4895102896872326e-06, + "loss": 1.0361, + "step": 4864 + }, + { + "epoch": 0.43874284168282457, + "grad_norm": 1.7473177790560162, + "learning_rate": 2.4889438269955457e-06, + "loss": 0.9218, + "step": 4865 + }, + { + "epoch": 0.4388330252062948, + "grad_norm": 1.5266857241950011, + "learning_rate": 2.4883773225844755e-06, + "loss": 0.9728, + "step": 4866 + }, + { + "epoch": 0.4389232087297651, + "grad_norm": 1.5708620239120128, + "learning_rate": 2.48781077650236e-06, + "loss": 0.983, + "step": 4867 + }, + { + "epoch": 0.4390133922532353, + "grad_norm": 1.9814269627623413, + "learning_rate": 2.4872441887975386e-06, + "loss": 0.9544, + "step": 4868 + }, + { + "epoch": 0.4391035757767056, + "grad_norm": 1.5585390621524393, + "learning_rate": 2.486677559518356e-06, + "loss": 0.9321, + "step": 4869 + }, + { + "epoch": 0.43919375930017585, + "grad_norm": 1.2505427612686963, + "learning_rate": 2.4861108887131614e-06, + "loss": 0.98, + "step": 4870 + }, + { + "epoch": 0.43928394282364613, + "grad_norm": 1.3460112773938635, + "learning_rate": 2.485544176430305e-06, + "loss": 0.9542, + "step": 4871 + }, + { + "epoch": 0.43937412634711637, + "grad_norm": 1.8031110607646392, + "learning_rate": 2.4849774227181425e-06, + "loss": 1.0664, + "step": 4872 + }, + { + "epoch": 0.43946430987058666, + "grad_norm": 1.3208644548393929, + "learning_rate": 2.484410627625032e-06, + "loss": 0.9888, + "step": 4873 + }, + { + "epoch": 0.4395544933940569, + "grad_norm": 1.88603864776236, + "learning_rate": 2.4838437911993356e-06, + "loss": 1.0309, + "step": 4874 + }, + { + "epoch": 0.4396446769175272, + "grad_norm": 3.0074567311916947, + "learning_rate": 2.483276913489419e-06, + "loss": 1.0294, + "step": 4875 + }, + { + "epoch": 0.4397348604409974, + "grad_norm": 3.0456969502430353, + "learning_rate": 2.4827099945436516e-06, + "loss": 0.894, + "step": 4876 + }, + { + "epoch": 0.4398250439644677, + "grad_norm": 2.7127334845822992, + "learning_rate": 2.482143034410405e-06, + "loss": 1.0681, + "step": 4877 + }, + { + "epoch": 0.43991522748793793, + "grad_norm": 1.5243953725682025, + "learning_rate": 2.4815760331380573e-06, + "loss": 1.017, + "step": 4878 + }, + { + "epoch": 0.4400054110114082, + "grad_norm": 1.5343774271089272, + "learning_rate": 2.481008990774987e-06, + "loss": 0.949, + "step": 4879 + }, + { + "epoch": 0.44009559453487845, + "grad_norm": 1.5214941687997126, + "learning_rate": 2.480441907369577e-06, + "loss": 0.987, + "step": 4880 + }, + { + "epoch": 0.44018577805834874, + "grad_norm": 1.576177017978753, + "learning_rate": 2.479874782970214e-06, + "loss": 0.9318, + "step": 4881 + }, + { + "epoch": 0.440275961581819, + "grad_norm": 2.760106070456735, + "learning_rate": 2.4793076176252887e-06, + "loss": 1.0123, + "step": 4882 + }, + { + "epoch": 0.44036614510528926, + "grad_norm": 1.5790827649188452, + "learning_rate": 2.478740411383195e-06, + "loss": 0.8715, + "step": 4883 + }, + { + "epoch": 0.44045632862875955, + "grad_norm": 1.3646663417272238, + "learning_rate": 2.4781731642923296e-06, + "loss": 0.98, + "step": 4884 + }, + { + "epoch": 0.4405465121522298, + "grad_norm": 1.447159200814631, + "learning_rate": 2.477605876401093e-06, + "loss": 0.9742, + "step": 4885 + }, + { + "epoch": 0.4406366956757001, + "grad_norm": 1.377398739397013, + "learning_rate": 2.4770385477578894e-06, + "loss": 0.9411, + "step": 4886 + }, + { + "epoch": 0.4407268791991703, + "grad_norm": 1.2556140238774425, + "learning_rate": 2.476471178411127e-06, + "loss": 0.9444, + "step": 4887 + }, + { + "epoch": 0.4408170627226406, + "grad_norm": 1.4475270394285245, + "learning_rate": 2.475903768409216e-06, + "loss": 1.0105, + "step": 4888 + }, + { + "epoch": 0.44090724624611083, + "grad_norm": 1.9092866443225853, + "learning_rate": 2.475336317800572e-06, + "loss": 0.9629, + "step": 4889 + }, + { + "epoch": 0.4409974297695811, + "grad_norm": 1.5406249969049108, + "learning_rate": 2.4747688266336118e-06, + "loss": 0.9763, + "step": 4890 + }, + { + "epoch": 0.44108761329305135, + "grad_norm": 1.6212593453637203, + "learning_rate": 2.4742012949567574e-06, + "loss": 1.0904, + "step": 4891 + }, + { + "epoch": 0.44117779681652164, + "grad_norm": 1.753125873096269, + "learning_rate": 2.4736337228184338e-06, + "loss": 0.9717, + "step": 4892 + }, + { + "epoch": 0.44126798033999187, + "grad_norm": 1.6967068049636294, + "learning_rate": 2.4730661102670692e-06, + "loss": 1.0875, + "step": 4893 + }, + { + "epoch": 0.44135816386346216, + "grad_norm": 1.7297277650910574, + "learning_rate": 2.472498457351096e-06, + "loss": 0.9906, + "step": 4894 + }, + { + "epoch": 0.4414483473869324, + "grad_norm": 1.9833036519348717, + "learning_rate": 2.4719307641189495e-06, + "loss": 0.9734, + "step": 4895 + }, + { + "epoch": 0.4415385309104027, + "grad_norm": 1.913356292611226, + "learning_rate": 2.4713630306190673e-06, + "loss": 0.9556, + "step": 4896 + }, + { + "epoch": 0.4416287144338729, + "grad_norm": 1.901145228521914, + "learning_rate": 2.4707952568998923e-06, + "loss": 1.0686, + "step": 4897 + }, + { + "epoch": 0.4417188979573432, + "grad_norm": 1.916539422939214, + "learning_rate": 2.4702274430098703e-06, + "loss": 0.9632, + "step": 4898 + }, + { + "epoch": 0.44180908148081344, + "grad_norm": 1.809141072121934, + "learning_rate": 2.4696595889974497e-06, + "loss": 0.9441, + "step": 4899 + }, + { + "epoch": 0.4418992650042837, + "grad_norm": 1.5341930522373002, + "learning_rate": 2.469091694911084e-06, + "loss": 1.046, + "step": 4900 + }, + { + "epoch": 0.44198944852775396, + "grad_norm": 1.8691239151960113, + "learning_rate": 2.4685237607992276e-06, + "loss": 0.9869, + "step": 4901 + }, + { + "epoch": 0.44207963205122425, + "grad_norm": 1.6440300619481487, + "learning_rate": 2.4679557867103416e-06, + "loss": 1.0516, + "step": 4902 + }, + { + "epoch": 0.4421698155746945, + "grad_norm": 1.3919528664989438, + "learning_rate": 2.4673877726928865e-06, + "loss": 0.9921, + "step": 4903 + }, + { + "epoch": 0.44225999909816477, + "grad_norm": 1.9079405995445406, + "learning_rate": 2.46681971879533e-06, + "loss": 0.8799, + "step": 4904 + }, + { + "epoch": 0.442350182621635, + "grad_norm": 2.2991197477251775, + "learning_rate": 2.4662516250661407e-06, + "loss": 0.9085, + "step": 4905 + }, + { + "epoch": 0.4424403661451053, + "grad_norm": 2.027289182189521, + "learning_rate": 2.465683491553792e-06, + "loss": 0.9628, + "step": 4906 + }, + { + "epoch": 0.4425305496685755, + "grad_norm": 1.619723777534007, + "learning_rate": 2.4651153183067604e-06, + "loss": 0.9935, + "step": 4907 + }, + { + "epoch": 0.4426207331920458, + "grad_norm": 1.6379913262730297, + "learning_rate": 2.4645471053735245e-06, + "loss": 0.9305, + "step": 4908 + }, + { + "epoch": 0.4427109167155161, + "grad_norm": 1.5090183318423211, + "learning_rate": 2.4639788528025684e-06, + "loss": 0.9261, + "step": 4909 + }, + { + "epoch": 0.44280110023898633, + "grad_norm": 1.3116655421868189, + "learning_rate": 2.463410560642378e-06, + "loss": 0.9936, + "step": 4910 + }, + { + "epoch": 0.4428912837624566, + "grad_norm": 0.6346784795949344, + "learning_rate": 2.4628422289414448e-06, + "loss": 0.8258, + "step": 4911 + }, + { + "epoch": 0.44298146728592686, + "grad_norm": 0.6594475999972866, + "learning_rate": 2.4622738577482592e-06, + "loss": 0.8009, + "step": 4912 + }, + { + "epoch": 0.44307165080939714, + "grad_norm": 1.5076668624512328, + "learning_rate": 2.461705447111319e-06, + "loss": 0.9767, + "step": 4913 + }, + { + "epoch": 0.4431618343328674, + "grad_norm": 1.513771418097858, + "learning_rate": 2.4611369970791246e-06, + "loss": 0.962, + "step": 4914 + }, + { + "epoch": 0.44325201785633767, + "grad_norm": 1.6745553323164273, + "learning_rate": 2.460568507700179e-06, + "loss": 0.9905, + "step": 4915 + }, + { + "epoch": 0.4433422013798079, + "grad_norm": 2.3272600039153284, + "learning_rate": 2.4599999790229887e-06, + "loss": 1.027, + "step": 4916 + }, + { + "epoch": 0.4434323849032782, + "grad_norm": 1.6541853938095619, + "learning_rate": 2.459431411096064e-06, + "loss": 0.9496, + "step": 4917 + }, + { + "epoch": 0.4435225684267484, + "grad_norm": 1.6884732442175678, + "learning_rate": 2.458862803967918e-06, + "loss": 0.9585, + "step": 4918 + }, + { + "epoch": 0.4436127519502187, + "grad_norm": 1.6204667315012122, + "learning_rate": 2.4582941576870667e-06, + "loss": 1.0056, + "step": 4919 + }, + { + "epoch": 0.44370293547368894, + "grad_norm": 1.7052048748573958, + "learning_rate": 2.4577254723020315e-06, + "loss": 0.9683, + "step": 4920 + }, + { + "epoch": 0.44379311899715923, + "grad_norm": 1.8795915967618273, + "learning_rate": 2.457156747861335e-06, + "loss": 1.0084, + "step": 4921 + }, + { + "epoch": 0.44388330252062946, + "grad_norm": 1.6958064049809574, + "learning_rate": 2.456587984413504e-06, + "loss": 1.0507, + "step": 4922 + }, + { + "epoch": 0.44397348604409975, + "grad_norm": 1.3247438813036014, + "learning_rate": 2.4560191820070683e-06, + "loss": 0.9215, + "step": 4923 + }, + { + "epoch": 0.44406366956757, + "grad_norm": 2.1337263783341287, + "learning_rate": 2.4554503406905617e-06, + "loss": 0.9689, + "step": 4924 + }, + { + "epoch": 0.4441538530910403, + "grad_norm": 1.4922702626549251, + "learning_rate": 2.454881460512521e-06, + "loss": 1.0281, + "step": 4925 + }, + { + "epoch": 0.4442440366145105, + "grad_norm": 1.3763229767925635, + "learning_rate": 2.4543125415214856e-06, + "loss": 0.9571, + "step": 4926 + }, + { + "epoch": 0.4443342201379808, + "grad_norm": 1.6546720509442792, + "learning_rate": 2.4537435837659996e-06, + "loss": 1.0138, + "step": 4927 + }, + { + "epoch": 0.44442440366145103, + "grad_norm": 2.7713437195908193, + "learning_rate": 2.4531745872946085e-06, + "loss": 0.9428, + "step": 4928 + }, + { + "epoch": 0.4445145871849213, + "grad_norm": 1.7761200769143983, + "learning_rate": 2.4526055521558632e-06, + "loss": 1.0512, + "step": 4929 + }, + { + "epoch": 0.44460477070839155, + "grad_norm": 1.5547246113857383, + "learning_rate": 2.4520364783983164e-06, + "loss": 1.0024, + "step": 4930 + }, + { + "epoch": 0.44469495423186184, + "grad_norm": 1.419512956130831, + "learning_rate": 2.451467366070525e-06, + "loss": 0.9738, + "step": 4931 + }, + { + "epoch": 0.4447851377553321, + "grad_norm": 2.6369643259481617, + "learning_rate": 2.450898215221048e-06, + "loss": 1.0614, + "step": 4932 + }, + { + "epoch": 0.44487532127880236, + "grad_norm": 1.7309845786330547, + "learning_rate": 2.4503290258984498e-06, + "loss": 1.1339, + "step": 4933 + }, + { + "epoch": 0.44496550480227265, + "grad_norm": 2.3087850062221156, + "learning_rate": 2.4497597981512952e-06, + "loss": 0.8529, + "step": 4934 + }, + { + "epoch": 0.4450556883257429, + "grad_norm": 0.6445438268185814, + "learning_rate": 2.4491905320281555e-06, + "loss": 0.8292, + "step": 4935 + }, + { + "epoch": 0.44514587184921317, + "grad_norm": 3.5654564352084894, + "learning_rate": 2.448621227577602e-06, + "loss": 0.9651, + "step": 4936 + }, + { + "epoch": 0.4452360553726834, + "grad_norm": 1.5465392702076575, + "learning_rate": 2.4480518848482123e-06, + "loss": 0.9304, + "step": 4937 + }, + { + "epoch": 0.4453262388961537, + "grad_norm": 1.5439595192262254, + "learning_rate": 2.447482503888565e-06, + "loss": 1.0038, + "step": 4938 + }, + { + "epoch": 0.4454164224196239, + "grad_norm": 1.9118801209405192, + "learning_rate": 2.4469130847472434e-06, + "loss": 0.9603, + "step": 4939 + }, + { + "epoch": 0.4455066059430942, + "grad_norm": 1.6461824296366083, + "learning_rate": 2.4463436274728326e-06, + "loss": 1.0285, + "step": 4940 + }, + { + "epoch": 0.44559678946656445, + "grad_norm": 2.3635878182807892, + "learning_rate": 2.4457741321139227e-06, + "loss": 0.8381, + "step": 4941 + }, + { + "epoch": 0.44568697299003474, + "grad_norm": 1.7358620257241284, + "learning_rate": 2.4452045987191063e-06, + "loss": 1.0086, + "step": 4942 + }, + { + "epoch": 0.44577715651350497, + "grad_norm": 1.603934192739235, + "learning_rate": 2.4446350273369776e-06, + "loss": 0.9628, + "step": 4943 + }, + { + "epoch": 0.44586734003697526, + "grad_norm": 1.4625739250084087, + "learning_rate": 2.4440654180161374e-06, + "loss": 1.0252, + "step": 4944 + }, + { + "epoch": 0.4459575235604455, + "grad_norm": 1.527764772109344, + "learning_rate": 2.4434957708051875e-06, + "loss": 0.9431, + "step": 4945 + }, + { + "epoch": 0.4460477070839158, + "grad_norm": 1.8333652088977737, + "learning_rate": 2.4429260857527324e-06, + "loss": 0.9347, + "step": 4946 + }, + { + "epoch": 0.446137890607386, + "grad_norm": 2.707207479974545, + "learning_rate": 2.4423563629073815e-06, + "loss": 0.8983, + "step": 4947 + }, + { + "epoch": 0.4462280741308563, + "grad_norm": 1.4548884530584796, + "learning_rate": 2.4417866023177466e-06, + "loss": 0.986, + "step": 4948 + }, + { + "epoch": 0.44631825765432653, + "grad_norm": 2.0890038059066645, + "learning_rate": 2.441216804032443e-06, + "loss": 1.021, + "step": 4949 + }, + { + "epoch": 0.4464084411777968, + "grad_norm": 1.8709034992248705, + "learning_rate": 2.440646968100089e-06, + "loss": 1.003, + "step": 4950 + }, + { + "epoch": 0.44649862470126706, + "grad_norm": 0.6757311278702316, + "learning_rate": 2.4400770945693055e-06, + "loss": 0.8119, + "step": 4951 + }, + { + "epoch": 0.44658880822473734, + "grad_norm": 1.695335493503164, + "learning_rate": 2.4395071834887177e-06, + "loss": 1.0202, + "step": 4952 + }, + { + "epoch": 0.4466789917482076, + "grad_norm": 0.699767403616843, + "learning_rate": 2.438937234906954e-06, + "loss": 0.8727, + "step": 4953 + }, + { + "epoch": 0.44676917527167787, + "grad_norm": 1.6043256788173554, + "learning_rate": 2.4383672488726447e-06, + "loss": 0.9797, + "step": 4954 + }, + { + "epoch": 0.44685935879514815, + "grad_norm": 1.648899465615818, + "learning_rate": 2.4377972254344256e-06, + "loss": 0.9104, + "step": 4955 + }, + { + "epoch": 0.4469495423186184, + "grad_norm": 1.698331825036055, + "learning_rate": 2.437227164640932e-06, + "loss": 0.8764, + "step": 4956 + }, + { + "epoch": 0.4470397258420887, + "grad_norm": 1.8049578443555165, + "learning_rate": 2.436657066540807e-06, + "loss": 0.9319, + "step": 4957 + }, + { + "epoch": 0.4471299093655589, + "grad_norm": 2.6666524608551567, + "learning_rate": 2.4360869311826927e-06, + "loss": 0.8663, + "step": 4958 + }, + { + "epoch": 0.4472200928890292, + "grad_norm": 1.722148448464648, + "learning_rate": 2.4355167586152367e-06, + "loss": 1.0812, + "step": 4959 + }, + { + "epoch": 0.44731027641249943, + "grad_norm": 2.033892394686317, + "learning_rate": 2.4349465488870896e-06, + "loss": 0.9571, + "step": 4960 + }, + { + "epoch": 0.4474004599359697, + "grad_norm": 1.479893313377313, + "learning_rate": 2.434376302046905e-06, + "loss": 1.0731, + "step": 4961 + }, + { + "epoch": 0.44749064345943995, + "grad_norm": 1.5956236353747715, + "learning_rate": 2.433806018143339e-06, + "loss": 1.018, + "step": 4962 + }, + { + "epoch": 0.44758082698291024, + "grad_norm": 0.6744770408410309, + "learning_rate": 2.433235697225051e-06, + "loss": 0.8538, + "step": 4963 + }, + { + "epoch": 0.4476710105063805, + "grad_norm": 1.642379912075458, + "learning_rate": 2.4326653393407048e-06, + "loss": 1.0, + "step": 4964 + }, + { + "epoch": 0.44776119402985076, + "grad_norm": 1.425283350296211, + "learning_rate": 2.432094944538966e-06, + "loss": 0.9554, + "step": 4965 + }, + { + "epoch": 0.447851377553321, + "grad_norm": 1.391041853929802, + "learning_rate": 2.4315245128685047e-06, + "loss": 0.9551, + "step": 4966 + }, + { + "epoch": 0.4479415610767913, + "grad_norm": 1.5969535196189155, + "learning_rate": 2.4309540443779925e-06, + "loss": 0.9901, + "step": 4967 + }, + { + "epoch": 0.4480317446002615, + "grad_norm": 1.4526553471757238, + "learning_rate": 2.4303835391161047e-06, + "loss": 0.9222, + "step": 4968 + }, + { + "epoch": 0.4481219281237318, + "grad_norm": 1.545952107406952, + "learning_rate": 2.42981299713152e-06, + "loss": 0.9358, + "step": 4969 + }, + { + "epoch": 0.44821211164720204, + "grad_norm": 1.7119398969233997, + "learning_rate": 2.4292424184729204e-06, + "loss": 1.0255, + "step": 4970 + }, + { + "epoch": 0.4483022951706723, + "grad_norm": 1.545048106904418, + "learning_rate": 2.4286718031889913e-06, + "loss": 0.9793, + "step": 4971 + }, + { + "epoch": 0.44839247869414256, + "grad_norm": 1.3825707520631445, + "learning_rate": 2.4281011513284202e-06, + "loss": 0.9608, + "step": 4972 + }, + { + "epoch": 0.44848266221761285, + "grad_norm": 3.6914688993112916, + "learning_rate": 2.4275304629398985e-06, + "loss": 0.9257, + "step": 4973 + }, + { + "epoch": 0.4485728457410831, + "grad_norm": 1.3806529018886207, + "learning_rate": 2.4269597380721194e-06, + "loss": 1.0632, + "step": 4974 + }, + { + "epoch": 0.44866302926455337, + "grad_norm": 1.475272770891931, + "learning_rate": 2.426388976773782e-06, + "loss": 0.9824, + "step": 4975 + }, + { + "epoch": 0.4487532127880236, + "grad_norm": 1.701700927596213, + "learning_rate": 2.425818179093586e-06, + "loss": 1.0936, + "step": 4976 + }, + { + "epoch": 0.4488433963114939, + "grad_norm": 1.5641339723032002, + "learning_rate": 2.4252473450802346e-06, + "loss": 0.8996, + "step": 4977 + }, + { + "epoch": 0.4489335798349641, + "grad_norm": 2.457990450034279, + "learning_rate": 2.4246764747824355e-06, + "loss": 1.0212, + "step": 4978 + }, + { + "epoch": 0.4490237633584344, + "grad_norm": 1.3136347452089527, + "learning_rate": 2.424105568248897e-06, + "loss": 0.997, + "step": 4979 + }, + { + "epoch": 0.4491139468819047, + "grad_norm": 1.9584119118150582, + "learning_rate": 2.4235346255283337e-06, + "loss": 0.9249, + "step": 4980 + }, + { + "epoch": 0.44920413040537494, + "grad_norm": 1.550976411852837, + "learning_rate": 2.42296364666946e-06, + "loss": 1.0005, + "step": 4981 + }, + { + "epoch": 0.4492943139288452, + "grad_norm": 1.674402981731131, + "learning_rate": 2.4223926317209965e-06, + "loss": 0.9341, + "step": 4982 + }, + { + "epoch": 0.44938449745231546, + "grad_norm": 2.1606783520481527, + "learning_rate": 2.4218215807316647e-06, + "loss": 0.9891, + "step": 4983 + }, + { + "epoch": 0.44947468097578575, + "grad_norm": 1.7268450061403697, + "learning_rate": 2.4212504937501894e-06, + "loss": 1.0756, + "step": 4984 + }, + { + "epoch": 0.449564864499256, + "grad_norm": 1.3809657291963635, + "learning_rate": 2.4206793708253e-06, + "loss": 1.0428, + "step": 4985 + }, + { + "epoch": 0.44965504802272627, + "grad_norm": 2.3036926191182987, + "learning_rate": 2.420108212005726e-06, + "loss": 0.9397, + "step": 4986 + }, + { + "epoch": 0.4497452315461965, + "grad_norm": 1.7820636162667918, + "learning_rate": 2.4195370173402034e-06, + "loss": 0.9885, + "step": 4987 + }, + { + "epoch": 0.4498354150696668, + "grad_norm": 1.6528497015995982, + "learning_rate": 2.4189657868774696e-06, + "loss": 1.1083, + "step": 4988 + }, + { + "epoch": 0.449925598593137, + "grad_norm": 2.0815662773772683, + "learning_rate": 2.418394520666264e-06, + "loss": 0.9249, + "step": 4989 + }, + { + "epoch": 0.4500157821166073, + "grad_norm": 2.9060091308534313, + "learning_rate": 2.4178232187553307e-06, + "loss": 0.9727, + "step": 4990 + }, + { + "epoch": 0.45010596564007754, + "grad_norm": 1.4431603156353614, + "learning_rate": 2.417251881193417e-06, + "loss": 1.0156, + "step": 4991 + }, + { + "epoch": 0.45019614916354783, + "grad_norm": 1.5584061206131494, + "learning_rate": 2.4166805080292723e-06, + "loss": 1.0482, + "step": 4992 + }, + { + "epoch": 0.45028633268701806, + "grad_norm": 1.7304178406953064, + "learning_rate": 2.4161090993116485e-06, + "loss": 0.9385, + "step": 4993 + }, + { + "epoch": 0.45037651621048835, + "grad_norm": 1.9642841611583564, + "learning_rate": 2.4155376550893026e-06, + "loss": 0.8697, + "step": 4994 + }, + { + "epoch": 0.4504666997339586, + "grad_norm": 2.3629550677211557, + "learning_rate": 2.4149661754109926e-06, + "loss": 0.9988, + "step": 4995 + }, + { + "epoch": 0.4505568832574289, + "grad_norm": 2.0150980889084193, + "learning_rate": 2.41439466032548e-06, + "loss": 1.0446, + "step": 4996 + }, + { + "epoch": 0.4506470667808991, + "grad_norm": 2.106414299508023, + "learning_rate": 2.41382310988153e-06, + "loss": 0.9353, + "step": 4997 + }, + { + "epoch": 0.4507372503043694, + "grad_norm": 1.75318849825993, + "learning_rate": 2.413251524127911e-06, + "loss": 1.0467, + "step": 4998 + }, + { + "epoch": 0.45082743382783963, + "grad_norm": 1.8168418444011958, + "learning_rate": 2.412679903113393e-06, + "loss": 0.9582, + "step": 4999 + }, + { + "epoch": 0.4509176173513099, + "grad_norm": 2.0645117918284295, + "learning_rate": 2.4121082468867505e-06, + "loss": 0.9958, + "step": 5000 + }, + { + "epoch": 0.45100780087478015, + "grad_norm": 1.9498801170079982, + "learning_rate": 2.4115365554967597e-06, + "loss": 0.9281, + "step": 5001 + }, + { + "epoch": 0.45109798439825044, + "grad_norm": 1.8956039066317105, + "learning_rate": 2.4109648289922006e-06, + "loss": 0.9871, + "step": 5002 + }, + { + "epoch": 0.45118816792172073, + "grad_norm": 1.6399150220440561, + "learning_rate": 2.4103930674218565e-06, + "loss": 0.9959, + "step": 5003 + }, + { + "epoch": 0.45127835144519096, + "grad_norm": 1.5936548541222415, + "learning_rate": 2.409821270834513e-06, + "loss": 0.9628, + "step": 5004 + }, + { + "epoch": 0.45136853496866125, + "grad_norm": 2.267654877644304, + "learning_rate": 2.409249439278959e-06, + "loss": 1.0424, + "step": 5005 + }, + { + "epoch": 0.4514587184921315, + "grad_norm": 0.6188070800731914, + "learning_rate": 2.408677572803986e-06, + "loss": 0.7895, + "step": 5006 + }, + { + "epoch": 0.45154890201560177, + "grad_norm": 1.501351859951274, + "learning_rate": 2.408105671458389e-06, + "loss": 0.9924, + "step": 5007 + }, + { + "epoch": 0.451639085539072, + "grad_norm": 1.60759785130384, + "learning_rate": 2.4075337352909663e-06, + "loss": 0.9325, + "step": 5008 + }, + { + "epoch": 0.4517292690625423, + "grad_norm": 1.4704150238128821, + "learning_rate": 2.4069617643505177e-06, + "loss": 1.1159, + "step": 5009 + }, + { + "epoch": 0.4518194525860125, + "grad_norm": 1.3530033788138391, + "learning_rate": 2.406389758685848e-06, + "loss": 1.0768, + "step": 5010 + }, + { + "epoch": 0.4519096361094828, + "grad_norm": 1.4742533879337387, + "learning_rate": 2.405817718345763e-06, + "loss": 1.0281, + "step": 5011 + }, + { + "epoch": 0.45199981963295305, + "grad_norm": 1.6826259017249316, + "learning_rate": 2.4052456433790726e-06, + "loss": 0.942, + "step": 5012 + }, + { + "epoch": 0.45209000315642334, + "grad_norm": 1.5765907458978248, + "learning_rate": 2.4046735338345897e-06, + "loss": 1.0409, + "step": 5013 + }, + { + "epoch": 0.45218018667989357, + "grad_norm": 1.6153386554013716, + "learning_rate": 2.404101389761129e-06, + "loss": 1.003, + "step": 5014 + }, + { + "epoch": 0.45227037020336386, + "grad_norm": 0.6622289363971771, + "learning_rate": 2.4035292112075097e-06, + "loss": 0.8004, + "step": 5015 + }, + { + "epoch": 0.4523605537268341, + "grad_norm": 1.4790559736213023, + "learning_rate": 2.4029569982225534e-06, + "loss": 1.0496, + "step": 5016 + }, + { + "epoch": 0.4524507372503044, + "grad_norm": 0.7034743394868986, + "learning_rate": 2.402384750855084e-06, + "loss": 0.8729, + "step": 5017 + }, + { + "epoch": 0.4525409207737746, + "grad_norm": 1.8298211339660935, + "learning_rate": 2.4018124691539286e-06, + "loss": 0.993, + "step": 5018 + }, + { + "epoch": 0.4526311042972449, + "grad_norm": 1.8630210326886694, + "learning_rate": 2.4012401531679178e-06, + "loss": 0.9808, + "step": 5019 + }, + { + "epoch": 0.45272128782071513, + "grad_norm": 1.8617453332716756, + "learning_rate": 2.4006678029458847e-06, + "loss": 0.9915, + "step": 5020 + }, + { + "epoch": 0.4528114713441854, + "grad_norm": 1.4175104919260595, + "learning_rate": 2.400095418536666e-06, + "loss": 0.8931, + "step": 5021 + }, + { + "epoch": 0.45290165486765566, + "grad_norm": 1.5677942989601439, + "learning_rate": 2.3995229999890996e-06, + "loss": 0.9674, + "step": 5022 + }, + { + "epoch": 0.45299183839112594, + "grad_norm": 1.852729783523986, + "learning_rate": 2.398950547352028e-06, + "loss": 0.9329, + "step": 5023 + }, + { + "epoch": 0.4530820219145962, + "grad_norm": 2.374158609826702, + "learning_rate": 2.398378060674295e-06, + "loss": 0.9048, + "step": 5024 + }, + { + "epoch": 0.45317220543806647, + "grad_norm": 1.4477108413441075, + "learning_rate": 2.39780554000475e-06, + "loss": 0.9889, + "step": 5025 + }, + { + "epoch": 0.4532623889615367, + "grad_norm": 1.6102878981313138, + "learning_rate": 2.3972329853922434e-06, + "loss": 1.019, + "step": 5026 + }, + { + "epoch": 0.453352572485007, + "grad_norm": 0.5989957521332557, + "learning_rate": 2.3966603968856278e-06, + "loss": 0.8077, + "step": 5027 + }, + { + "epoch": 0.4534427560084773, + "grad_norm": 1.6258923941240724, + "learning_rate": 2.39608777453376e-06, + "loss": 1.0619, + "step": 5028 + }, + { + "epoch": 0.4535329395319475, + "grad_norm": 1.8127192002966552, + "learning_rate": 2.3955151183854993e-06, + "loss": 1.0081, + "step": 5029 + }, + { + "epoch": 0.4536231230554178, + "grad_norm": 1.6500237954475725, + "learning_rate": 2.3949424284897073e-06, + "loss": 0.9555, + "step": 5030 + }, + { + "epoch": 0.45371330657888803, + "grad_norm": 1.4753346661577325, + "learning_rate": 2.39436970489525e-06, + "loss": 0.8588, + "step": 5031 + }, + { + "epoch": 0.4538034901023583, + "grad_norm": 1.6047908344817163, + "learning_rate": 2.3937969476509955e-06, + "loss": 0.9887, + "step": 5032 + }, + { + "epoch": 0.45389367362582855, + "grad_norm": 1.5686164920730277, + "learning_rate": 2.393224156805813e-06, + "loss": 1.0638, + "step": 5033 + }, + { + "epoch": 0.45398385714929884, + "grad_norm": 1.753349572687888, + "learning_rate": 2.392651332408578e-06, + "loss": 0.964, + "step": 5034 + }, + { + "epoch": 0.4540740406727691, + "grad_norm": 2.550000385209597, + "learning_rate": 2.3920784745081655e-06, + "loss": 0.9738, + "step": 5035 + }, + { + "epoch": 0.45416422419623936, + "grad_norm": 4.411085140772124, + "learning_rate": 2.391505583153456e-06, + "loss": 1.0243, + "step": 5036 + }, + { + "epoch": 0.4542544077197096, + "grad_norm": 1.4395371804611556, + "learning_rate": 2.3909326583933315e-06, + "loss": 0.9936, + "step": 5037 + }, + { + "epoch": 0.4543445912431799, + "grad_norm": 1.4790058406742879, + "learning_rate": 2.3903597002766777e-06, + "loss": 0.9558, + "step": 5038 + }, + { + "epoch": 0.4544347747666501, + "grad_norm": 1.5043150507191199, + "learning_rate": 2.389786708852381e-06, + "loss": 0.9534, + "step": 5039 + }, + { + "epoch": 0.4545249582901204, + "grad_norm": 0.9445771123386498, + "learning_rate": 2.389213684169333e-06, + "loss": 0.9321, + "step": 5040 + }, + { + "epoch": 0.45461514181359064, + "grad_norm": 1.3916555722112296, + "learning_rate": 2.388640626276428e-06, + "loss": 0.9492, + "step": 5041 + }, + { + "epoch": 0.45470532533706093, + "grad_norm": 0.682860913513516, + "learning_rate": 2.388067535222561e-06, + "loss": 0.7727, + "step": 5042 + }, + { + "epoch": 0.45479550886053116, + "grad_norm": 2.0428850729383017, + "learning_rate": 2.3874944110566332e-06, + "loss": 1.0048, + "step": 5043 + }, + { + "epoch": 0.45488569238400145, + "grad_norm": 0.7936642607991959, + "learning_rate": 2.3869212538275447e-06, + "loss": 0.9152, + "step": 5044 + }, + { + "epoch": 0.4549758759074717, + "grad_norm": 1.5211099695282368, + "learning_rate": 2.386348063584202e-06, + "loss": 1.0069, + "step": 5045 + }, + { + "epoch": 0.45506605943094197, + "grad_norm": 1.5060177730864177, + "learning_rate": 2.385774840375511e-06, + "loss": 1.0404, + "step": 5046 + }, + { + "epoch": 0.4551562429544122, + "grad_norm": 1.7334816496312138, + "learning_rate": 2.385201584250385e-06, + "loss": 0.9436, + "step": 5047 + }, + { + "epoch": 0.4552464264778825, + "grad_norm": 0.6350972644401023, + "learning_rate": 2.3846282952577346e-06, + "loss": 0.85, + "step": 5048 + }, + { + "epoch": 0.4553366100013527, + "grad_norm": 1.442858370290771, + "learning_rate": 2.3840549734464785e-06, + "loss": 1.0794, + "step": 5049 + }, + { + "epoch": 0.455426793524823, + "grad_norm": 1.6553905344367188, + "learning_rate": 2.3834816188655336e-06, + "loss": 0.9262, + "step": 5050 + }, + { + "epoch": 0.4555169770482933, + "grad_norm": 1.763151140529704, + "learning_rate": 2.3829082315638224e-06, + "loss": 0.9361, + "step": 5051 + }, + { + "epoch": 0.45560716057176354, + "grad_norm": 1.8734206859861744, + "learning_rate": 2.3823348115902695e-06, + "loss": 0.8523, + "step": 5052 + }, + { + "epoch": 0.4556973440952338, + "grad_norm": 1.5015810740073332, + "learning_rate": 2.3817613589938026e-06, + "loss": 1.0069, + "step": 5053 + }, + { + "epoch": 0.45578752761870406, + "grad_norm": 1.5893124154114973, + "learning_rate": 2.3811878738233517e-06, + "loss": 1.0199, + "step": 5054 + }, + { + "epoch": 0.45587771114217435, + "grad_norm": 2.0460917597728603, + "learning_rate": 2.380614356127849e-06, + "loss": 0.9351, + "step": 5055 + }, + { + "epoch": 0.4559678946656446, + "grad_norm": 2.1279058223119627, + "learning_rate": 2.3800408059562318e-06, + "loss": 0.9883, + "step": 5056 + }, + { + "epoch": 0.45605807818911487, + "grad_norm": 1.6301272273581187, + "learning_rate": 2.3794672233574365e-06, + "loss": 1.0065, + "step": 5057 + }, + { + "epoch": 0.4561482617125851, + "grad_norm": 1.658477077612085, + "learning_rate": 2.3788936083804058e-06, + "loss": 1.0013, + "step": 5058 + }, + { + "epoch": 0.4562384452360554, + "grad_norm": 1.6015779075811842, + "learning_rate": 2.378319961074083e-06, + "loss": 0.9945, + "step": 5059 + }, + { + "epoch": 0.4563286287595256, + "grad_norm": 1.4903534807876697, + "learning_rate": 2.377746281487415e-06, + "loss": 1.027, + "step": 5060 + }, + { + "epoch": 0.4564188122829959, + "grad_norm": 1.5663090625778648, + "learning_rate": 2.377172569669352e-06, + "loss": 1.0767, + "step": 5061 + }, + { + "epoch": 0.45650899580646614, + "grad_norm": 1.7110328821916427, + "learning_rate": 2.376598825668845e-06, + "loss": 0.8443, + "step": 5062 + }, + { + "epoch": 0.45659917932993643, + "grad_norm": 1.6309374396548357, + "learning_rate": 2.3760250495348495e-06, + "loss": 0.9777, + "step": 5063 + }, + { + "epoch": 0.45668936285340667, + "grad_norm": 1.660055790273899, + "learning_rate": 2.3754512413163236e-06, + "loss": 1.0364, + "step": 5064 + }, + { + "epoch": 0.45677954637687695, + "grad_norm": 1.6111299150107674, + "learning_rate": 2.3748774010622285e-06, + "loss": 0.9727, + "step": 5065 + }, + { + "epoch": 0.4568697299003472, + "grad_norm": 1.4342141121451562, + "learning_rate": 2.3743035288215254e-06, + "loss": 0.9986, + "step": 5066 + }, + { + "epoch": 0.4569599134238175, + "grad_norm": 1.5166620855297166, + "learning_rate": 2.3737296246431815e-06, + "loss": 0.9973, + "step": 5067 + }, + { + "epoch": 0.4570500969472877, + "grad_norm": 1.4874120862610565, + "learning_rate": 2.3731556885761656e-06, + "loss": 0.9606, + "step": 5068 + }, + { + "epoch": 0.457140280470758, + "grad_norm": 0.6384519379150725, + "learning_rate": 2.372581720669449e-06, + "loss": 0.8175, + "step": 5069 + }, + { + "epoch": 0.45723046399422823, + "grad_norm": 2.086785076405484, + "learning_rate": 2.3720077209720046e-06, + "loss": 1.0191, + "step": 5070 + }, + { + "epoch": 0.4573206475176985, + "grad_norm": 0.6856576770238908, + "learning_rate": 2.3714336895328112e-06, + "loss": 0.8105, + "step": 5071 + }, + { + "epoch": 0.45741083104116875, + "grad_norm": 1.7673667486267843, + "learning_rate": 2.370859626400847e-06, + "loss": 0.9455, + "step": 5072 + }, + { + "epoch": 0.45750101456463904, + "grad_norm": 1.4564029269492083, + "learning_rate": 2.3702855316250943e-06, + "loss": 0.9633, + "step": 5073 + }, + { + "epoch": 0.45759119808810933, + "grad_norm": 1.6565093071397163, + "learning_rate": 2.369711405254539e-06, + "loss": 0.9913, + "step": 5074 + }, + { + "epoch": 0.45768138161157956, + "grad_norm": 1.6088759750447428, + "learning_rate": 2.3691372473381673e-06, + "loss": 1.0443, + "step": 5075 + }, + { + "epoch": 0.45777156513504985, + "grad_norm": 1.5568352054761248, + "learning_rate": 2.3685630579249708e-06, + "loss": 1.0507, + "step": 5076 + }, + { + "epoch": 0.4578617486585201, + "grad_norm": 1.8187011659758252, + "learning_rate": 2.367988837063942e-06, + "loss": 0.9594, + "step": 5077 + }, + { + "epoch": 0.4579519321819904, + "grad_norm": 1.911377187006029, + "learning_rate": 2.367414584804076e-06, + "loss": 1.0651, + "step": 5078 + }, + { + "epoch": 0.4580421157054606, + "grad_norm": 1.4229133922414492, + "learning_rate": 2.366840301194372e-06, + "loss": 1.0194, + "step": 5079 + }, + { + "epoch": 0.4581322992289309, + "grad_norm": 1.627559992890137, + "learning_rate": 2.3662659862838308e-06, + "loss": 0.977, + "step": 5080 + }, + { + "epoch": 0.45822248275240113, + "grad_norm": 1.4974519227068632, + "learning_rate": 2.365691640121456e-06, + "loss": 1.0476, + "step": 5081 + }, + { + "epoch": 0.4583126662758714, + "grad_norm": 1.6360982296747268, + "learning_rate": 2.365117262756254e-06, + "loss": 0.9183, + "step": 5082 + }, + { + "epoch": 0.45840284979934165, + "grad_norm": 2.581402261967716, + "learning_rate": 2.3645428542372342e-06, + "loss": 0.9854, + "step": 5083 + }, + { + "epoch": 0.45849303332281194, + "grad_norm": 0.6131330177335075, + "learning_rate": 2.3639684146134083e-06, + "loss": 0.7693, + "step": 5084 + }, + { + "epoch": 0.45858321684628217, + "grad_norm": 1.65932474293726, + "learning_rate": 2.3633939439337897e-06, + "loss": 0.9488, + "step": 5085 + }, + { + "epoch": 0.45867340036975246, + "grad_norm": 1.5246690516089727, + "learning_rate": 2.362819442247396e-06, + "loss": 0.9608, + "step": 5086 + }, + { + "epoch": 0.4587635838932227, + "grad_norm": 0.6324541466568494, + "learning_rate": 2.3622449096032477e-06, + "loss": 0.78, + "step": 5087 + }, + { + "epoch": 0.458853767416693, + "grad_norm": 2.351228912112932, + "learning_rate": 2.361670346050366e-06, + "loss": 1.0261, + "step": 5088 + }, + { + "epoch": 0.4589439509401632, + "grad_norm": 2.076034756687402, + "learning_rate": 2.3610957516377757e-06, + "loss": 0.9233, + "step": 5089 + }, + { + "epoch": 0.4590341344636335, + "grad_norm": 1.6057016616760424, + "learning_rate": 2.3605211264145048e-06, + "loss": 0.9961, + "step": 5090 + }, + { + "epoch": 0.45912431798710374, + "grad_norm": 1.2695034552245805, + "learning_rate": 2.3599464704295836e-06, + "loss": 0.9531, + "step": 5091 + }, + { + "epoch": 0.459214501510574, + "grad_norm": 2.5266520808091633, + "learning_rate": 2.359371783732045e-06, + "loss": 0.8855, + "step": 5092 + }, + { + "epoch": 0.45930468503404426, + "grad_norm": 1.5016328348603032, + "learning_rate": 2.358797066370924e-06, + "loss": 0.972, + "step": 5093 + }, + { + "epoch": 0.45939486855751455, + "grad_norm": 1.6745274261433798, + "learning_rate": 2.3582223183952594e-06, + "loss": 0.9876, + "step": 5094 + }, + { + "epoch": 0.4594850520809848, + "grad_norm": 3.718594587909324, + "learning_rate": 2.357647539854091e-06, + "loss": 1.0252, + "step": 5095 + }, + { + "epoch": 0.45957523560445507, + "grad_norm": 2.254134935119257, + "learning_rate": 2.3570727307964624e-06, + "loss": 0.9984, + "step": 5096 + }, + { + "epoch": 0.4596654191279253, + "grad_norm": 1.5126399897586151, + "learning_rate": 2.35649789127142e-06, + "loss": 1.0415, + "step": 5097 + }, + { + "epoch": 0.4597556026513956, + "grad_norm": 1.6482234291539815, + "learning_rate": 2.3559230213280115e-06, + "loss": 0.9547, + "step": 5098 + }, + { + "epoch": 0.4598457861748659, + "grad_norm": 5.829025894366679, + "learning_rate": 2.3553481210152886e-06, + "loss": 0.9032, + "step": 5099 + }, + { + "epoch": 0.4599359696983361, + "grad_norm": 0.8286228842507589, + "learning_rate": 2.3547731903823043e-06, + "loss": 0.8341, + "step": 5100 + }, + { + "epoch": 0.4600261532218064, + "grad_norm": 1.4234856506310452, + "learning_rate": 2.3541982294781155e-06, + "loss": 0.9645, + "step": 5101 + }, + { + "epoch": 0.46011633674527663, + "grad_norm": 1.8869444275775886, + "learning_rate": 2.3536232383517804e-06, + "loss": 0.9867, + "step": 5102 + }, + { + "epoch": 0.4602065202687469, + "grad_norm": 1.4668557959048374, + "learning_rate": 2.3530482170523602e-06, + "loss": 0.9824, + "step": 5103 + }, + { + "epoch": 0.46029670379221715, + "grad_norm": 1.4796446265320586, + "learning_rate": 2.3524731656289206e-06, + "loss": 1.0834, + "step": 5104 + }, + { + "epoch": 0.46038688731568744, + "grad_norm": 1.6431047863041586, + "learning_rate": 2.351898084130526e-06, + "loss": 1.08, + "step": 5105 + }, + { + "epoch": 0.4604770708391577, + "grad_norm": 1.6222153058550566, + "learning_rate": 2.351322972606247e-06, + "loss": 1.0551, + "step": 5106 + }, + { + "epoch": 0.46056725436262796, + "grad_norm": 1.6152184334536783, + "learning_rate": 2.350747831105155e-06, + "loss": 1.1357, + "step": 5107 + }, + { + "epoch": 0.4606574378860982, + "grad_norm": 1.832993179612444, + "learning_rate": 2.350172659676323e-06, + "loss": 1.0678, + "step": 5108 + }, + { + "epoch": 0.4607476214095685, + "grad_norm": 1.5401413133115, + "learning_rate": 2.3495974583688306e-06, + "loss": 1.0263, + "step": 5109 + }, + { + "epoch": 0.4608378049330387, + "grad_norm": 1.4172945127281507, + "learning_rate": 2.3490222272317543e-06, + "loss": 1.0597, + "step": 5110 + }, + { + "epoch": 0.460927988456509, + "grad_norm": 1.7549322648847436, + "learning_rate": 2.348446966314177e-06, + "loss": 1.0342, + "step": 5111 + }, + { + "epoch": 0.46101817197997924, + "grad_norm": 1.4426301551689058, + "learning_rate": 2.3478716756651837e-06, + "loss": 0.963, + "step": 5112 + }, + { + "epoch": 0.46110835550344953, + "grad_norm": 1.6928471513215322, + "learning_rate": 2.347296355333861e-06, + "loss": 0.9832, + "step": 5113 + }, + { + "epoch": 0.46119853902691976, + "grad_norm": 1.6891741748348739, + "learning_rate": 2.3467210053692972e-06, + "loss": 0.9947, + "step": 5114 + }, + { + "epoch": 0.46128872255039005, + "grad_norm": 1.739169735956827, + "learning_rate": 2.3461456258205866e-06, + "loss": 0.9926, + "step": 5115 + }, + { + "epoch": 0.4613789060738603, + "grad_norm": 1.5411692495609417, + "learning_rate": 2.345570216736822e-06, + "loss": 0.9295, + "step": 5116 + }, + { + "epoch": 0.4614690895973306, + "grad_norm": 1.837752723182235, + "learning_rate": 2.3449947781671013e-06, + "loss": 0.9992, + "step": 5117 + }, + { + "epoch": 0.4615592731208008, + "grad_norm": 1.8084894731092365, + "learning_rate": 2.3444193101605237e-06, + "loss": 1.0168, + "step": 5118 + }, + { + "epoch": 0.4616494566442711, + "grad_norm": 1.5087348607274789, + "learning_rate": 2.3438438127661913e-06, + "loss": 1.0085, + "step": 5119 + }, + { + "epoch": 0.4617396401677413, + "grad_norm": 1.6744901222673494, + "learning_rate": 2.3432682860332096e-06, + "loss": 0.8191, + "step": 5120 + }, + { + "epoch": 0.4618298236912116, + "grad_norm": 1.6216108485564236, + "learning_rate": 2.342692730010684e-06, + "loss": 0.9382, + "step": 5121 + }, + { + "epoch": 0.4619200072146819, + "grad_norm": 2.0491709172392336, + "learning_rate": 2.342117144747726e-06, + "loss": 0.9965, + "step": 5122 + }, + { + "epoch": 0.46201019073815214, + "grad_norm": 1.6254211760282937, + "learning_rate": 2.3415415302934457e-06, + "loss": 0.9908, + "step": 5123 + }, + { + "epoch": 0.4621003742616224, + "grad_norm": 1.528728822949784, + "learning_rate": 2.340965886696959e-06, + "loss": 1.0204, + "step": 5124 + }, + { + "epoch": 0.46219055778509266, + "grad_norm": 1.6346362574353703, + "learning_rate": 2.340390214007384e-06, + "loss": 0.9397, + "step": 5125 + }, + { + "epoch": 0.46228074130856295, + "grad_norm": 1.9570884467806606, + "learning_rate": 2.339814512273838e-06, + "loss": 0.8733, + "step": 5126 + }, + { + "epoch": 0.4623709248320332, + "grad_norm": 1.7129485273850769, + "learning_rate": 2.3392387815454447e-06, + "loss": 1.0975, + "step": 5127 + }, + { + "epoch": 0.46246110835550347, + "grad_norm": 1.8035223696563767, + "learning_rate": 2.3386630218713273e-06, + "loss": 0.9887, + "step": 5128 + }, + { + "epoch": 0.4625512918789737, + "grad_norm": 1.6956466134617025, + "learning_rate": 2.3380872333006135e-06, + "loss": 1.0198, + "step": 5129 + }, + { + "epoch": 0.462641475402444, + "grad_norm": 2.113520864274963, + "learning_rate": 2.3375114158824335e-06, + "loss": 1.0383, + "step": 5130 + }, + { + "epoch": 0.4627316589259142, + "grad_norm": 1.5911389822638111, + "learning_rate": 2.3369355696659184e-06, + "loss": 1.0276, + "step": 5131 + }, + { + "epoch": 0.4628218424493845, + "grad_norm": 1.533315316730192, + "learning_rate": 2.336359694700202e-06, + "loss": 0.9772, + "step": 5132 + }, + { + "epoch": 0.46291202597285475, + "grad_norm": 1.83356202028612, + "learning_rate": 2.335783791034422e-06, + "loss": 0.9926, + "step": 5133 + }, + { + "epoch": 0.46300220949632503, + "grad_norm": 1.607970133802986, + "learning_rate": 2.3352078587177173e-06, + "loss": 0.9763, + "step": 5134 + }, + { + "epoch": 0.46309239301979527, + "grad_norm": 1.6486106890055026, + "learning_rate": 2.33463189779923e-06, + "loss": 0.9994, + "step": 5135 + }, + { + "epoch": 0.46318257654326556, + "grad_norm": 1.410347584770456, + "learning_rate": 2.334055908328104e-06, + "loss": 1.064, + "step": 5136 + }, + { + "epoch": 0.4632727600667358, + "grad_norm": 1.4089504592151287, + "learning_rate": 2.3334798903534866e-06, + "loss": 0.901, + "step": 5137 + }, + { + "epoch": 0.4633629435902061, + "grad_norm": 1.3309602878874147, + "learning_rate": 2.3329038439245252e-06, + "loss": 0.8861, + "step": 5138 + }, + { + "epoch": 0.4634531271136763, + "grad_norm": 2.3485117034458067, + "learning_rate": 2.3323277690903724e-06, + "loss": 0.9881, + "step": 5139 + }, + { + "epoch": 0.4635433106371466, + "grad_norm": 1.5460805683239083, + "learning_rate": 2.3317516659001827e-06, + "loss": 0.9654, + "step": 5140 + }, + { + "epoch": 0.46363349416061683, + "grad_norm": 1.5489128638004808, + "learning_rate": 2.331175534403111e-06, + "loss": 1.0038, + "step": 5141 + }, + { + "epoch": 0.4637236776840871, + "grad_norm": 1.4438376660173908, + "learning_rate": 2.3305993746483167e-06, + "loss": 1.07, + "step": 5142 + }, + { + "epoch": 0.46381386120755735, + "grad_norm": 1.9765549610582203, + "learning_rate": 2.3300231866849606e-06, + "loss": 0.9708, + "step": 5143 + }, + { + "epoch": 0.46390404473102764, + "grad_norm": 1.695049072050404, + "learning_rate": 2.3294469705622067e-06, + "loss": 1.0324, + "step": 5144 + }, + { + "epoch": 0.4639942282544979, + "grad_norm": 1.6399765186257018, + "learning_rate": 2.3288707263292203e-06, + "loss": 0.9672, + "step": 5145 + }, + { + "epoch": 0.46408441177796816, + "grad_norm": 1.480113769488644, + "learning_rate": 2.3282944540351707e-06, + "loss": 1.0094, + "step": 5146 + }, + { + "epoch": 0.46417459530143845, + "grad_norm": 1.6845151145848019, + "learning_rate": 2.327718153729228e-06, + "loss": 0.9603, + "step": 5147 + }, + { + "epoch": 0.4642647788249087, + "grad_norm": 1.6848745054833922, + "learning_rate": 2.327141825460566e-06, + "loss": 1.0314, + "step": 5148 + }, + { + "epoch": 0.464354962348379, + "grad_norm": 0.6755925394636112, + "learning_rate": 2.326565469278358e-06, + "loss": 0.8227, + "step": 5149 + }, + { + "epoch": 0.4644451458718492, + "grad_norm": 0.7198551223615027, + "learning_rate": 2.3259890852317846e-06, + "loss": 0.8314, + "step": 5150 + }, + { + "epoch": 0.4645353293953195, + "grad_norm": 2.219344153569893, + "learning_rate": 2.3254126733700246e-06, + "loss": 0.9894, + "step": 5151 + }, + { + "epoch": 0.46462551291878973, + "grad_norm": 1.6287968367826926, + "learning_rate": 2.324836233742262e-06, + "loss": 1.0216, + "step": 5152 + }, + { + "epoch": 0.46471569644226, + "grad_norm": 1.7305598181190198, + "learning_rate": 2.3242597663976793e-06, + "loss": 1.027, + "step": 5153 + }, + { + "epoch": 0.46480587996573025, + "grad_norm": 1.9093556354745718, + "learning_rate": 2.3236832713854663e-06, + "loss": 0.9709, + "step": 5154 + }, + { + "epoch": 0.46489606348920054, + "grad_norm": 1.5732746180122656, + "learning_rate": 2.323106748754812e-06, + "loss": 0.9864, + "step": 5155 + }, + { + "epoch": 0.4649862470126708, + "grad_norm": 1.5171862554103095, + "learning_rate": 2.3225301985549077e-06, + "loss": 1.0823, + "step": 5156 + }, + { + "epoch": 0.46507643053614106, + "grad_norm": 1.9038823542232157, + "learning_rate": 2.321953620834948e-06, + "loss": 0.9203, + "step": 5157 + }, + { + "epoch": 0.4651666140596113, + "grad_norm": 1.6013853393821202, + "learning_rate": 2.3213770156441314e-06, + "loss": 0.9561, + "step": 5158 + }, + { + "epoch": 0.4652567975830816, + "grad_norm": 3.850597082630333, + "learning_rate": 2.3208003830316554e-06, + "loss": 1.0347, + "step": 5159 + }, + { + "epoch": 0.4653469811065518, + "grad_norm": 1.8492471245545685, + "learning_rate": 2.3202237230467215e-06, + "loss": 1.082, + "step": 5160 + }, + { + "epoch": 0.4654371646300221, + "grad_norm": 1.9803233196726875, + "learning_rate": 2.3196470357385338e-06, + "loss": 0.9526, + "step": 5161 + }, + { + "epoch": 0.46552734815349234, + "grad_norm": 1.4336374884229521, + "learning_rate": 2.319070321156299e-06, + "loss": 0.9824, + "step": 5162 + }, + { + "epoch": 0.4656175316769626, + "grad_norm": 1.496775021330039, + "learning_rate": 2.318493579349224e-06, + "loss": 1.1216, + "step": 5163 + }, + { + "epoch": 0.46570771520043286, + "grad_norm": 1.6712272316088619, + "learning_rate": 2.317916810366522e-06, + "loss": 0.9763, + "step": 5164 + }, + { + "epoch": 0.46579789872390315, + "grad_norm": 1.5165636754262297, + "learning_rate": 2.317340014257404e-06, + "loss": 0.9358, + "step": 5165 + }, + { + "epoch": 0.4658880822473734, + "grad_norm": 1.5167703921442268, + "learning_rate": 2.316763191071086e-06, + "loss": 0.919, + "step": 5166 + }, + { + "epoch": 0.46597826577084367, + "grad_norm": 2.189166279088356, + "learning_rate": 2.316186340856787e-06, + "loss": 0.9459, + "step": 5167 + }, + { + "epoch": 0.4660684492943139, + "grad_norm": 1.8502071908967967, + "learning_rate": 2.315609463663725e-06, + "loss": 0.9391, + "step": 5168 + }, + { + "epoch": 0.4661586328177842, + "grad_norm": 1.746499989336758, + "learning_rate": 2.315032559541123e-06, + "loss": 0.963, + "step": 5169 + }, + { + "epoch": 0.4662488163412545, + "grad_norm": 2.285692670413401, + "learning_rate": 2.314455628538207e-06, + "loss": 1.0078, + "step": 5170 + }, + { + "epoch": 0.4663389998647247, + "grad_norm": 1.3276094221288202, + "learning_rate": 2.3138786707042023e-06, + "loss": 0.9613, + "step": 5171 + }, + { + "epoch": 0.466429183388195, + "grad_norm": 1.4432159889451996, + "learning_rate": 2.3133016860883387e-06, + "loss": 1.0008, + "step": 5172 + }, + { + "epoch": 0.46651936691166523, + "grad_norm": 2.5212061792819935, + "learning_rate": 2.3127246747398475e-06, + "loss": 0.969, + "step": 5173 + }, + { + "epoch": 0.4666095504351355, + "grad_norm": 1.2663756605121552, + "learning_rate": 2.312147636707963e-06, + "loss": 0.9864, + "step": 5174 + }, + { + "epoch": 0.46669973395860576, + "grad_norm": 1.983524110044126, + "learning_rate": 2.3115705720419214e-06, + "loss": 0.9678, + "step": 5175 + }, + { + "epoch": 0.46678991748207604, + "grad_norm": 1.8420980132469404, + "learning_rate": 2.31099348079096e-06, + "loss": 1.0094, + "step": 5176 + }, + { + "epoch": 0.4668801010055463, + "grad_norm": 1.6183024815147, + "learning_rate": 2.31041636300432e-06, + "loss": 1.0323, + "step": 5177 + }, + { + "epoch": 0.46697028452901657, + "grad_norm": 1.5963910976533509, + "learning_rate": 2.3098392187312445e-06, + "loss": 1.0716, + "step": 5178 + }, + { + "epoch": 0.4670604680524868, + "grad_norm": 1.7871001029686042, + "learning_rate": 2.309262048020978e-06, + "loss": 0.9679, + "step": 5179 + }, + { + "epoch": 0.4671506515759571, + "grad_norm": 1.8031564799028466, + "learning_rate": 2.308684850922769e-06, + "loss": 1.0223, + "step": 5180 + }, + { + "epoch": 0.4672408350994273, + "grad_norm": 1.663927529462513, + "learning_rate": 2.3081076274858664e-06, + "loss": 0.9807, + "step": 5181 + }, + { + "epoch": 0.4673310186228976, + "grad_norm": 1.4313517405234422, + "learning_rate": 2.307530377759522e-06, + "loss": 1.0176, + "step": 5182 + }, + { + "epoch": 0.46742120214636784, + "grad_norm": 1.9720867883568836, + "learning_rate": 2.30695310179299e-06, + "loss": 0.9172, + "step": 5183 + }, + { + "epoch": 0.46751138566983813, + "grad_norm": 1.4296062842475914, + "learning_rate": 2.3063757996355267e-06, + "loss": 0.9866, + "step": 5184 + }, + { + "epoch": 0.46760156919330836, + "grad_norm": 1.8709037540948443, + "learning_rate": 2.3057984713363903e-06, + "loss": 1.0253, + "step": 5185 + }, + { + "epoch": 0.46769175271677865, + "grad_norm": 0.6933498972930029, + "learning_rate": 2.3052211169448436e-06, + "loss": 0.794, + "step": 5186 + }, + { + "epoch": 0.4677819362402489, + "grad_norm": 1.6442005975814753, + "learning_rate": 2.3046437365101474e-06, + "loss": 0.9344, + "step": 5187 + }, + { + "epoch": 0.4678721197637192, + "grad_norm": 1.7987873681635826, + "learning_rate": 2.3040663300815673e-06, + "loss": 0.9657, + "step": 5188 + }, + { + "epoch": 0.4679623032871894, + "grad_norm": 0.7768955951639945, + "learning_rate": 2.3034888977083723e-06, + "loss": 0.8029, + "step": 5189 + }, + { + "epoch": 0.4680524868106597, + "grad_norm": 1.9332603090021998, + "learning_rate": 2.30291143943983e-06, + "loss": 0.9052, + "step": 5190 + }, + { + "epoch": 0.46814267033412993, + "grad_norm": 1.6496271839305159, + "learning_rate": 2.3023339553252145e-06, + "loss": 0.9541, + "step": 5191 + }, + { + "epoch": 0.4682328538576002, + "grad_norm": 1.5218174071645065, + "learning_rate": 2.301756445413799e-06, + "loss": 1.0224, + "step": 5192 + }, + { + "epoch": 0.4683230373810705, + "grad_norm": 1.6349312209615365, + "learning_rate": 2.3011789097548585e-06, + "loss": 0.9937, + "step": 5193 + }, + { + "epoch": 0.46841322090454074, + "grad_norm": 1.4841844135203657, + "learning_rate": 2.3006013483976738e-06, + "loss": 0.9883, + "step": 5194 + }, + { + "epoch": 0.468503404428011, + "grad_norm": 2.109952946359717, + "learning_rate": 2.300023761391524e-06, + "loss": 0.9861, + "step": 5195 + }, + { + "epoch": 0.46859358795148126, + "grad_norm": 1.8106973167886213, + "learning_rate": 2.299446148785693e-06, + "loss": 1.0388, + "step": 5196 + }, + { + "epoch": 0.46868377147495155, + "grad_norm": 2.4571819410827622, + "learning_rate": 2.2988685106294654e-06, + "loss": 0.9692, + "step": 5197 + }, + { + "epoch": 0.4687739549984218, + "grad_norm": 1.637553436921618, + "learning_rate": 2.2982908469721284e-06, + "loss": 0.8881, + "step": 5198 + }, + { + "epoch": 0.46886413852189207, + "grad_norm": 1.876619593175027, + "learning_rate": 2.2977131578629714e-06, + "loss": 1.0595, + "step": 5199 + }, + { + "epoch": 0.4689543220453623, + "grad_norm": 1.605635734097545, + "learning_rate": 2.297135443351286e-06, + "loss": 1.0101, + "step": 5200 + }, + { + "epoch": 0.4690445055688326, + "grad_norm": 1.2894764091194526, + "learning_rate": 2.296557703486367e-06, + "loss": 0.9977, + "step": 5201 + }, + { + "epoch": 0.4691346890923028, + "grad_norm": 8.047153268095174, + "learning_rate": 2.295979938317509e-06, + "loss": 1.0147, + "step": 5202 + }, + { + "epoch": 0.4692248726157731, + "grad_norm": 1.8145832717258985, + "learning_rate": 2.295402147894011e-06, + "loss": 0.9936, + "step": 5203 + }, + { + "epoch": 0.46931505613924335, + "grad_norm": 1.4600696944232636, + "learning_rate": 2.2948243322651723e-06, + "loss": 0.9174, + "step": 5204 + }, + { + "epoch": 0.46940523966271364, + "grad_norm": 1.8542974047341645, + "learning_rate": 2.2942464914802962e-06, + "loss": 0.9951, + "step": 5205 + }, + { + "epoch": 0.46949542318618387, + "grad_norm": 1.8857048896249953, + "learning_rate": 2.293668625588687e-06, + "loss": 0.9675, + "step": 5206 + }, + { + "epoch": 0.46958560670965416, + "grad_norm": 2.3249626443794766, + "learning_rate": 2.293090734639651e-06, + "loss": 1.032, + "step": 5207 + }, + { + "epoch": 0.4696757902331244, + "grad_norm": 1.7255844867486474, + "learning_rate": 2.2925128186824983e-06, + "loss": 1.109, + "step": 5208 + }, + { + "epoch": 0.4697659737565947, + "grad_norm": 1.8243291196578717, + "learning_rate": 2.2919348777665384e-06, + "loss": 1.0919, + "step": 5209 + }, + { + "epoch": 0.4698561572800649, + "grad_norm": 1.681784381105024, + "learning_rate": 2.2913569119410856e-06, + "loss": 0.9925, + "step": 5210 + }, + { + "epoch": 0.4699463408035352, + "grad_norm": 1.6312637650066157, + "learning_rate": 2.290778921255454e-06, + "loss": 1.0179, + "step": 5211 + }, + { + "epoch": 0.47003652432700543, + "grad_norm": 1.5241352537466144, + "learning_rate": 2.2902009057589613e-06, + "loss": 1.0259, + "step": 5212 + }, + { + "epoch": 0.4701267078504757, + "grad_norm": 1.2632574849438263, + "learning_rate": 2.2896228655009276e-06, + "loss": 0.9437, + "step": 5213 + }, + { + "epoch": 0.47021689137394596, + "grad_norm": 1.5173524273881651, + "learning_rate": 2.289044800530674e-06, + "loss": 0.9573, + "step": 5214 + }, + { + "epoch": 0.47030707489741624, + "grad_norm": 2.9665843895064703, + "learning_rate": 2.2884667108975245e-06, + "loss": 0.9536, + "step": 5215 + }, + { + "epoch": 0.4703972584208865, + "grad_norm": 1.7486127395696365, + "learning_rate": 2.287888596650804e-06, + "loss": 1.0303, + "step": 5216 + }, + { + "epoch": 0.47048744194435677, + "grad_norm": 4.608781782867868, + "learning_rate": 2.287310457839841e-06, + "loss": 0.9904, + "step": 5217 + }, + { + "epoch": 0.47057762546782705, + "grad_norm": 1.799926621212959, + "learning_rate": 2.286732294513966e-06, + "loss": 1.0644, + "step": 5218 + }, + { + "epoch": 0.4706678089912973, + "grad_norm": 2.180568503350155, + "learning_rate": 2.2861541067225106e-06, + "loss": 0.9979, + "step": 5219 + }, + { + "epoch": 0.4707579925147676, + "grad_norm": 2.694943167771957, + "learning_rate": 2.2855758945148095e-06, + "loss": 0.976, + "step": 5220 + }, + { + "epoch": 0.4708481760382378, + "grad_norm": 1.6598443669588496, + "learning_rate": 2.2849976579401977e-06, + "loss": 1.0504, + "step": 5221 + }, + { + "epoch": 0.4709383595617081, + "grad_norm": 2.3711485502895604, + "learning_rate": 2.284419397048014e-06, + "loss": 0.9854, + "step": 5222 + }, + { + "epoch": 0.47102854308517833, + "grad_norm": 1.4903458019910443, + "learning_rate": 2.2838411118875997e-06, + "loss": 1.0008, + "step": 5223 + }, + { + "epoch": 0.4711187266086486, + "grad_norm": 1.3721501988949065, + "learning_rate": 2.283262802508296e-06, + "loss": 0.9162, + "step": 5224 + }, + { + "epoch": 0.47120891013211885, + "grad_norm": 1.2388032601124845, + "learning_rate": 2.2826844689594492e-06, + "loss": 0.9697, + "step": 5225 + }, + { + "epoch": 0.47129909365558914, + "grad_norm": 1.716617093095741, + "learning_rate": 2.282106111290404e-06, + "loss": 1.0712, + "step": 5226 + }, + { + "epoch": 0.4713892771790594, + "grad_norm": 0.6906485635526407, + "learning_rate": 2.2815277295505098e-06, + "loss": 0.838, + "step": 5227 + }, + { + "epoch": 0.47147946070252966, + "grad_norm": 2.2373471000044756, + "learning_rate": 2.2809493237891174e-06, + "loss": 0.941, + "step": 5228 + }, + { + "epoch": 0.4715696442259999, + "grad_norm": 1.6296768378314028, + "learning_rate": 2.2803708940555796e-06, + "loss": 1.0239, + "step": 5229 + }, + { + "epoch": 0.4716598277494702, + "grad_norm": 1.4430939839475936, + "learning_rate": 2.2797924403992514e-06, + "loss": 1.0623, + "step": 5230 + }, + { + "epoch": 0.4717500112729404, + "grad_norm": 8.92700859997039, + "learning_rate": 2.2792139628694892e-06, + "loss": 1.0096, + "step": 5231 + }, + { + "epoch": 0.4718401947964107, + "grad_norm": 1.4643003549408777, + "learning_rate": 2.2786354615156524e-06, + "loss": 0.9546, + "step": 5232 + }, + { + "epoch": 0.47193037831988094, + "grad_norm": 1.6493031908877192, + "learning_rate": 2.2780569363871016e-06, + "loss": 1.0543, + "step": 5233 + }, + { + "epoch": 0.4720205618433512, + "grad_norm": 0.6208078938164207, + "learning_rate": 2.277478387533199e-06, + "loss": 0.8103, + "step": 5234 + }, + { + "epoch": 0.47211074536682146, + "grad_norm": 1.8196632635463617, + "learning_rate": 2.276899815003311e-06, + "loss": 1.0265, + "step": 5235 + }, + { + "epoch": 0.47220092889029175, + "grad_norm": 1.5930361925696452, + "learning_rate": 2.2763212188468045e-06, + "loss": 0.9842, + "step": 5236 + }, + { + "epoch": 0.472291112413762, + "grad_norm": 1.4781745305543392, + "learning_rate": 2.2757425991130473e-06, + "loss": 1.0063, + "step": 5237 + }, + { + "epoch": 0.47238129593723227, + "grad_norm": 0.7618893554561764, + "learning_rate": 2.2751639558514117e-06, + "loss": 0.8232, + "step": 5238 + }, + { + "epoch": 0.4724714794607025, + "grad_norm": 1.4464237374522657, + "learning_rate": 2.2745852891112697e-06, + "loss": 1.0348, + "step": 5239 + }, + { + "epoch": 0.4725616629841728, + "grad_norm": 2.2356485858597694, + "learning_rate": 2.274006598941997e-06, + "loss": 0.9974, + "step": 5240 + }, + { + "epoch": 0.4726518465076431, + "grad_norm": 1.7691309730904945, + "learning_rate": 2.27342788539297e-06, + "loss": 0.974, + "step": 5241 + }, + { + "epoch": 0.4727420300311133, + "grad_norm": 2.320010387134319, + "learning_rate": 2.2728491485135684e-06, + "loss": 0.951, + "step": 5242 + }, + { + "epoch": 0.4728322135545836, + "grad_norm": 1.449012021848777, + "learning_rate": 2.272270388353173e-06, + "loss": 1.0473, + "step": 5243 + }, + { + "epoch": 0.47292239707805384, + "grad_norm": 1.934901494616462, + "learning_rate": 2.2716916049611666e-06, + "loss": 0.9668, + "step": 5244 + }, + { + "epoch": 0.4730125806015241, + "grad_norm": 1.553837682899924, + "learning_rate": 2.2711127983869346e-06, + "loss": 0.9754, + "step": 5245 + }, + { + "epoch": 0.47310276412499436, + "grad_norm": 1.3015230830018143, + "learning_rate": 2.270533968679864e-06, + "loss": 1.0031, + "step": 5246 + }, + { + "epoch": 0.47319294764846465, + "grad_norm": 1.3972312370592592, + "learning_rate": 2.269955115889343e-06, + "loss": 1.0001, + "step": 5247 + }, + { + "epoch": 0.4732831311719349, + "grad_norm": 2.0439861382837443, + "learning_rate": 2.269376240064763e-06, + "loss": 0.951, + "step": 5248 + }, + { + "epoch": 0.47337331469540517, + "grad_norm": 1.7812759665219455, + "learning_rate": 2.268797341255517e-06, + "loss": 1.0082, + "step": 5249 + }, + { + "epoch": 0.4734634982188754, + "grad_norm": 1.7332179703591464, + "learning_rate": 2.268218419511e-06, + "loss": 0.9108, + "step": 5250 + }, + { + "epoch": 0.4735536817423457, + "grad_norm": 5.886385011483531, + "learning_rate": 2.267639474880608e-06, + "loss": 0.9129, + "step": 5251 + }, + { + "epoch": 0.4736438652658159, + "grad_norm": 1.848480492395801, + "learning_rate": 2.2670605074137407e-06, + "loss": 0.9703, + "step": 5252 + }, + { + "epoch": 0.4737340487892862, + "grad_norm": 1.8502475237861375, + "learning_rate": 2.2664815171597983e-06, + "loss": 0.9732, + "step": 5253 + }, + { + "epoch": 0.47382423231275644, + "grad_norm": 1.824851863544835, + "learning_rate": 2.265902504168183e-06, + "loss": 0.9363, + "step": 5254 + }, + { + "epoch": 0.47391441583622673, + "grad_norm": 1.698791730872019, + "learning_rate": 2.2653234684883007e-06, + "loss": 0.9493, + "step": 5255 + }, + { + "epoch": 0.47400459935969697, + "grad_norm": 3.7613725512827703, + "learning_rate": 2.264744410169556e-06, + "loss": 0.9098, + "step": 5256 + }, + { + "epoch": 0.47409478288316725, + "grad_norm": 1.5422532366160433, + "learning_rate": 2.264165329261359e-06, + "loss": 0.9736, + "step": 5257 + }, + { + "epoch": 0.4741849664066375, + "grad_norm": 0.5875178577874436, + "learning_rate": 2.26358622581312e-06, + "loss": 0.8129, + "step": 5258 + }, + { + "epoch": 0.4742751499301078, + "grad_norm": 1.8468760482346196, + "learning_rate": 2.2630070998742504e-06, + "loss": 1.0549, + "step": 5259 + }, + { + "epoch": 0.474365333453578, + "grad_norm": 1.4975392502367861, + "learning_rate": 2.262427951494165e-06, + "loss": 1.0114, + "step": 5260 + }, + { + "epoch": 0.4744555169770483, + "grad_norm": 0.6532100932344205, + "learning_rate": 2.2618487807222794e-06, + "loss": 0.8525, + "step": 5261 + }, + { + "epoch": 0.47454570050051853, + "grad_norm": 1.6262094691721365, + "learning_rate": 2.261269587608012e-06, + "loss": 1.0747, + "step": 5262 + }, + { + "epoch": 0.4746358840239888, + "grad_norm": 1.429398554385607, + "learning_rate": 2.260690372200783e-06, + "loss": 0.997, + "step": 5263 + }, + { + "epoch": 0.47472606754745905, + "grad_norm": 1.3589468095263428, + "learning_rate": 2.2601111345500138e-06, + "loss": 1.0758, + "step": 5264 + }, + { + "epoch": 0.47481625107092934, + "grad_norm": 1.542716244667197, + "learning_rate": 2.2595318747051286e-06, + "loss": 0.9834, + "step": 5265 + }, + { + "epoch": 0.47490643459439963, + "grad_norm": 1.8913684990891975, + "learning_rate": 2.258952592715553e-06, + "loss": 1.0157, + "step": 5266 + }, + { + "epoch": 0.47499661811786986, + "grad_norm": 0.7042692833868209, + "learning_rate": 2.2583732886307142e-06, + "loss": 0.8546, + "step": 5267 + }, + { + "epoch": 0.47508680164134015, + "grad_norm": 1.5279773851488665, + "learning_rate": 2.2577939625000414e-06, + "loss": 0.9434, + "step": 5268 + }, + { + "epoch": 0.4751769851648104, + "grad_norm": 0.6789428534259381, + "learning_rate": 2.257214614372967e-06, + "loss": 0.8141, + "step": 5269 + }, + { + "epoch": 0.4752671686882807, + "grad_norm": 1.6598331630699166, + "learning_rate": 2.2566352442989227e-06, + "loss": 1.0175, + "step": 5270 + }, + { + "epoch": 0.4753573522117509, + "grad_norm": 1.4026435807794664, + "learning_rate": 2.256055852327344e-06, + "loss": 0.9966, + "step": 5271 + }, + { + "epoch": 0.4754475357352212, + "grad_norm": 1.913830490064127, + "learning_rate": 2.2554764385076685e-06, + "loss": 0.9374, + "step": 5272 + }, + { + "epoch": 0.4755377192586914, + "grad_norm": 1.7012825895911579, + "learning_rate": 2.2548970028893348e-06, + "loss": 1.0062, + "step": 5273 + }, + { + "epoch": 0.4756279027821617, + "grad_norm": 1.917345742159407, + "learning_rate": 2.254317545521783e-06, + "loss": 1.0042, + "step": 5274 + }, + { + "epoch": 0.47571808630563195, + "grad_norm": 2.2511728197049794, + "learning_rate": 2.253738066454457e-06, + "loss": 1.0245, + "step": 5275 + }, + { + "epoch": 0.47580826982910224, + "grad_norm": 1.9221036201275852, + "learning_rate": 2.2531585657367986e-06, + "loss": 0.9693, + "step": 5276 + }, + { + "epoch": 0.47589845335257247, + "grad_norm": 1.6152878813547416, + "learning_rate": 2.252579043418256e-06, + "loss": 0.9455, + "step": 5277 + }, + { + "epoch": 0.47598863687604276, + "grad_norm": 1.477803269065249, + "learning_rate": 2.251999499548277e-06, + "loss": 0.9671, + "step": 5278 + }, + { + "epoch": 0.476078820399513, + "grad_norm": 1.9293278656328219, + "learning_rate": 2.251419934176311e-06, + "loss": 0.9498, + "step": 5279 + }, + { + "epoch": 0.4761690039229833, + "grad_norm": 2.455837523792611, + "learning_rate": 2.25084034735181e-06, + "loss": 1.0423, + "step": 5280 + }, + { + "epoch": 0.4762591874464535, + "grad_norm": 1.6984661671904737, + "learning_rate": 2.2502607391242274e-06, + "loss": 1.0465, + "step": 5281 + }, + { + "epoch": 0.4763493709699238, + "grad_norm": 0.7394513339239113, + "learning_rate": 2.2496811095430182e-06, + "loss": 0.8193, + "step": 5282 + }, + { + "epoch": 0.47643955449339404, + "grad_norm": 1.7473496394487538, + "learning_rate": 2.249101458657641e-06, + "loss": 0.9872, + "step": 5283 + }, + { + "epoch": 0.4765297380168643, + "grad_norm": 1.8925521108691687, + "learning_rate": 2.2485217865175526e-06, + "loss": 1.061, + "step": 5284 + }, + { + "epoch": 0.47661992154033456, + "grad_norm": 1.6129544645292302, + "learning_rate": 2.2479420931722156e-06, + "loss": 1.1277, + "step": 5285 + }, + { + "epoch": 0.47671010506380485, + "grad_norm": 0.5839025382683682, + "learning_rate": 2.2473623786710923e-06, + "loss": 0.8561, + "step": 5286 + }, + { + "epoch": 0.4768002885872751, + "grad_norm": 1.8175221942664141, + "learning_rate": 2.2467826430636465e-06, + "loss": 0.9046, + "step": 5287 + }, + { + "epoch": 0.47689047211074537, + "grad_norm": 1.3589451428107109, + "learning_rate": 2.246202886399345e-06, + "loss": 1.016, + "step": 5288 + }, + { + "epoch": 0.47698065563421566, + "grad_norm": 1.4252988234292223, + "learning_rate": 2.2456231087276556e-06, + "loss": 0.9891, + "step": 5289 + }, + { + "epoch": 0.4770708391576859, + "grad_norm": 1.4262785579888095, + "learning_rate": 2.245043310098048e-06, + "loss": 0.9181, + "step": 5290 + }, + { + "epoch": 0.4771610226811562, + "grad_norm": 1.7801829705869283, + "learning_rate": 2.244463490559995e-06, + "loss": 1.0357, + "step": 5291 + }, + { + "epoch": 0.4772512062046264, + "grad_norm": 1.5816233167881302, + "learning_rate": 2.2438836501629683e-06, + "loss": 1.0041, + "step": 5292 + }, + { + "epoch": 0.4773413897280967, + "grad_norm": 1.4396304224522771, + "learning_rate": 2.2433037889564437e-06, + "loss": 0.9972, + "step": 5293 + }, + { + "epoch": 0.47743157325156693, + "grad_norm": 1.8570711625062473, + "learning_rate": 2.242723906989899e-06, + "loss": 0.9257, + "step": 5294 + }, + { + "epoch": 0.4775217567750372, + "grad_norm": 1.2888592646517596, + "learning_rate": 2.2421440043128114e-06, + "loss": 0.9238, + "step": 5295 + }, + { + "epoch": 0.47761194029850745, + "grad_norm": 1.7851538459878507, + "learning_rate": 2.241564080974662e-06, + "loss": 0.9061, + "step": 5296 + }, + { + "epoch": 0.47770212382197774, + "grad_norm": 1.7691567131956787, + "learning_rate": 2.2409841370249343e-06, + "loss": 0.9439, + "step": 5297 + }, + { + "epoch": 0.477792307345448, + "grad_norm": 1.9523341904887386, + "learning_rate": 2.2404041725131106e-06, + "loss": 1.0761, + "step": 5298 + }, + { + "epoch": 0.47788249086891826, + "grad_norm": 1.9653467099607116, + "learning_rate": 2.239824187488677e-06, + "loss": 0.9698, + "step": 5299 + }, + { + "epoch": 0.4779726743923885, + "grad_norm": 0.8010349151035779, + "learning_rate": 2.239244182001122e-06, + "loss": 0.8555, + "step": 5300 + }, + { + "epoch": 0.4780628579158588, + "grad_norm": 1.7402819146407105, + "learning_rate": 2.2386641560999336e-06, + "loss": 1.0083, + "step": 5301 + }, + { + "epoch": 0.478153041439329, + "grad_norm": 1.565839022628954, + "learning_rate": 2.238084109834604e-06, + "loss": 1.0255, + "step": 5302 + }, + { + "epoch": 0.4782432249627993, + "grad_norm": 2.1144811921905773, + "learning_rate": 2.237504043254625e-06, + "loss": 0.9946, + "step": 5303 + }, + { + "epoch": 0.47833340848626954, + "grad_norm": 1.8892570780335916, + "learning_rate": 2.2369239564094915e-06, + "loss": 0.9632, + "step": 5304 + }, + { + "epoch": 0.47842359200973983, + "grad_norm": 2.895323198425323, + "learning_rate": 2.2363438493486995e-06, + "loss": 0.9228, + "step": 5305 + }, + { + "epoch": 0.47851377553321006, + "grad_norm": 1.8051955923902756, + "learning_rate": 2.235763722121747e-06, + "loss": 0.8808, + "step": 5306 + }, + { + "epoch": 0.47860395905668035, + "grad_norm": 1.4137941390347357, + "learning_rate": 2.2351835747781346e-06, + "loss": 1.0104, + "step": 5307 + }, + { + "epoch": 0.4786941425801506, + "grad_norm": 1.5508514311546036, + "learning_rate": 2.234603407367362e-06, + "loss": 1.0335, + "step": 5308 + }, + { + "epoch": 0.47878432610362087, + "grad_norm": 1.6702406551086533, + "learning_rate": 2.2340232199389337e-06, + "loss": 1.0043, + "step": 5309 + }, + { + "epoch": 0.4788745096270911, + "grad_norm": 1.6548324852841771, + "learning_rate": 2.2334430125423538e-06, + "loss": 0.8735, + "step": 5310 + }, + { + "epoch": 0.4789646931505614, + "grad_norm": 1.701936569641157, + "learning_rate": 2.232862785227128e-06, + "loss": 1.0375, + "step": 5311 + }, + { + "epoch": 0.4790548766740317, + "grad_norm": 1.642322933227322, + "learning_rate": 2.232282538042766e-06, + "loss": 1.0246, + "step": 5312 + }, + { + "epoch": 0.4791450601975019, + "grad_norm": 1.8375730331335431, + "learning_rate": 2.231702271038777e-06, + "loss": 1.0412, + "step": 5313 + }, + { + "epoch": 0.4792352437209722, + "grad_norm": 2.2232117185068985, + "learning_rate": 2.231121984264673e-06, + "loss": 0.9713, + "step": 5314 + }, + { + "epoch": 0.47932542724444244, + "grad_norm": 1.4224262217075345, + "learning_rate": 2.2305416777699665e-06, + "loss": 1.0223, + "step": 5315 + }, + { + "epoch": 0.4794156107679127, + "grad_norm": 1.7403273295399004, + "learning_rate": 2.229961351604173e-06, + "loss": 0.9932, + "step": 5316 + }, + { + "epoch": 0.47950579429138296, + "grad_norm": 1.5804136695635607, + "learning_rate": 2.2293810058168085e-06, + "loss": 0.9499, + "step": 5317 + }, + { + "epoch": 0.47959597781485325, + "grad_norm": 1.4007618517152165, + "learning_rate": 2.2288006404573922e-06, + "loss": 1.0143, + "step": 5318 + }, + { + "epoch": 0.4796861613383235, + "grad_norm": 1.7275146344637364, + "learning_rate": 2.228220255575444e-06, + "loss": 0.9743, + "step": 5319 + }, + { + "epoch": 0.47977634486179377, + "grad_norm": 1.8029410759325752, + "learning_rate": 2.2276398512204847e-06, + "loss": 0.9391, + "step": 5320 + }, + { + "epoch": 0.479866528385264, + "grad_norm": 1.6922943391473182, + "learning_rate": 2.2270594274420382e-06, + "loss": 0.9984, + "step": 5321 + }, + { + "epoch": 0.4799567119087343, + "grad_norm": 0.7046049860584859, + "learning_rate": 2.22647898428963e-06, + "loss": 0.8454, + "step": 5322 + }, + { + "epoch": 0.4800468954322045, + "grad_norm": 1.364735611822448, + "learning_rate": 2.225898521812785e-06, + "loss": 1.0364, + "step": 5323 + }, + { + "epoch": 0.4801370789556748, + "grad_norm": 1.581443696352762, + "learning_rate": 2.2253180400610337e-06, + "loss": 1.0433, + "step": 5324 + }, + { + "epoch": 0.48022726247914505, + "grad_norm": 1.8075289465141553, + "learning_rate": 2.2247375390839037e-06, + "loss": 0.9744, + "step": 5325 + }, + { + "epoch": 0.48031744600261533, + "grad_norm": 1.8118915852119546, + "learning_rate": 2.224157018930928e-06, + "loss": 0.9623, + "step": 5326 + }, + { + "epoch": 0.48040762952608557, + "grad_norm": 1.440056807536597, + "learning_rate": 2.2235764796516395e-06, + "loss": 0.9896, + "step": 5327 + }, + { + "epoch": 0.48049781304955586, + "grad_norm": 2.1236121189383104, + "learning_rate": 2.222995921295573e-06, + "loss": 0.992, + "step": 5328 + }, + { + "epoch": 0.4805879965730261, + "grad_norm": 5.6052722431257225, + "learning_rate": 2.222415343912265e-06, + "loss": 0.9702, + "step": 5329 + }, + { + "epoch": 0.4806781800964964, + "grad_norm": 2.0405455125454885, + "learning_rate": 2.221834747551254e-06, + "loss": 1.0593, + "step": 5330 + }, + { + "epoch": 0.4807683636199666, + "grad_norm": 1.7651382720644964, + "learning_rate": 2.221254132262078e-06, + "loss": 0.9627, + "step": 5331 + }, + { + "epoch": 0.4808585471434369, + "grad_norm": 8.070209454509124, + "learning_rate": 2.2206734980942802e-06, + "loss": 1.0348, + "step": 5332 + }, + { + "epoch": 0.48094873066690713, + "grad_norm": 1.8406194197824342, + "learning_rate": 2.2200928450974024e-06, + "loss": 0.9701, + "step": 5333 + }, + { + "epoch": 0.4810389141903774, + "grad_norm": 1.6962038149410814, + "learning_rate": 2.21951217332099e-06, + "loss": 1.0122, + "step": 5334 + }, + { + "epoch": 0.48112909771384765, + "grad_norm": 2.4896476024381786, + "learning_rate": 2.2189314828145883e-06, + "loss": 0.907, + "step": 5335 + }, + { + "epoch": 0.48121928123731794, + "grad_norm": 1.4642098237684447, + "learning_rate": 2.2183507736277453e-06, + "loss": 0.9785, + "step": 5336 + }, + { + "epoch": 0.48130946476078823, + "grad_norm": 1.5113276485818017, + "learning_rate": 2.2177700458100107e-06, + "loss": 1.0067, + "step": 5337 + }, + { + "epoch": 0.48139964828425846, + "grad_norm": 1.5952897954776668, + "learning_rate": 2.2171892994109346e-06, + "loss": 0.969, + "step": 5338 + }, + { + "epoch": 0.48148983180772875, + "grad_norm": 1.5244334559945119, + "learning_rate": 2.21660853448007e-06, + "loss": 1.0558, + "step": 5339 + }, + { + "epoch": 0.481580015331199, + "grad_norm": 2.474209988828179, + "learning_rate": 2.2160277510669703e-06, + "loss": 0.9448, + "step": 5340 + }, + { + "epoch": 0.4816701988546693, + "grad_norm": 1.5762356316246624, + "learning_rate": 2.215446949221193e-06, + "loss": 0.9862, + "step": 5341 + }, + { + "epoch": 0.4817603823781395, + "grad_norm": 1.8745028790451574, + "learning_rate": 2.2148661289922924e-06, + "loss": 1.0023, + "step": 5342 + }, + { + "epoch": 0.4818505659016098, + "grad_norm": 2.3896503331559193, + "learning_rate": 2.21428529042983e-06, + "loss": 1.0572, + "step": 5343 + }, + { + "epoch": 0.48194074942508003, + "grad_norm": 5.191645608841923, + "learning_rate": 2.2137044335833647e-06, + "loss": 1.0409, + "step": 5344 + }, + { + "epoch": 0.4820309329485503, + "grad_norm": 1.7547537406449991, + "learning_rate": 2.213123558502459e-06, + "loss": 0.9833, + "step": 5345 + }, + { + "epoch": 0.48212111647202055, + "grad_norm": 1.7778378648670083, + "learning_rate": 2.2125426652366763e-06, + "loss": 1.0139, + "step": 5346 + }, + { + "epoch": 0.48221129999549084, + "grad_norm": 1.6272177601215458, + "learning_rate": 2.211961753835581e-06, + "loss": 1.0854, + "step": 5347 + }, + { + "epoch": 0.48230148351896107, + "grad_norm": 1.4370851332705847, + "learning_rate": 2.21138082434874e-06, + "loss": 0.9846, + "step": 5348 + }, + { + "epoch": 0.48239166704243136, + "grad_norm": 1.6939120507226744, + "learning_rate": 2.210799876825722e-06, + "loss": 1.0193, + "step": 5349 + }, + { + "epoch": 0.4824818505659016, + "grad_norm": 1.6349921029743018, + "learning_rate": 2.210218911316096e-06, + "loss": 0.9497, + "step": 5350 + }, + { + "epoch": 0.4825720340893719, + "grad_norm": 1.8471329252350923, + "learning_rate": 2.2096379278694336e-06, + "loss": 0.9417, + "step": 5351 + }, + { + "epoch": 0.4826622176128421, + "grad_norm": 2.123789835394052, + "learning_rate": 2.2090569265353074e-06, + "loss": 0.9368, + "step": 5352 + }, + { + "epoch": 0.4827524011363124, + "grad_norm": 1.9499957842659008, + "learning_rate": 2.2084759073632912e-06, + "loss": 0.9898, + "step": 5353 + }, + { + "epoch": 0.48284258465978264, + "grad_norm": 1.5750060278156524, + "learning_rate": 2.2078948704029606e-06, + "loss": 0.9553, + "step": 5354 + }, + { + "epoch": 0.4829327681832529, + "grad_norm": 1.7038033910433477, + "learning_rate": 2.2073138157038935e-06, + "loss": 0.9186, + "step": 5355 + }, + { + "epoch": 0.48302295170672316, + "grad_norm": 1.4921470756447652, + "learning_rate": 2.2067327433156687e-06, + "loss": 0.9794, + "step": 5356 + }, + { + "epoch": 0.48311313523019345, + "grad_norm": 1.4836554138321365, + "learning_rate": 2.2061516532878667e-06, + "loss": 1.0078, + "step": 5357 + }, + { + "epoch": 0.4832033187536637, + "grad_norm": 1.809277794422442, + "learning_rate": 2.2055705456700686e-06, + "loss": 1.038, + "step": 5358 + }, + { + "epoch": 0.48329350227713397, + "grad_norm": 1.7044466822356834, + "learning_rate": 2.204989420511858e-06, + "loss": 0.8921, + "step": 5359 + }, + { + "epoch": 0.48338368580060426, + "grad_norm": 2.2216262442061026, + "learning_rate": 2.20440827786282e-06, + "loss": 0.9377, + "step": 5360 + }, + { + "epoch": 0.4834738693240745, + "grad_norm": 1.8565038745037306, + "learning_rate": 2.20382711777254e-06, + "loss": 0.9944, + "step": 5361 + }, + { + "epoch": 0.4835640528475448, + "grad_norm": 1.5819823457670839, + "learning_rate": 2.203245940290607e-06, + "loss": 1.005, + "step": 5362 + }, + { + "epoch": 0.483654236371015, + "grad_norm": 1.669880970236409, + "learning_rate": 2.2026647454666097e-06, + "loss": 0.9916, + "step": 5363 + }, + { + "epoch": 0.4837444198944853, + "grad_norm": 2.0013027716498764, + "learning_rate": 2.2020835333501384e-06, + "loss": 1.0284, + "step": 5364 + }, + { + "epoch": 0.48383460341795553, + "grad_norm": 1.601959849122341, + "learning_rate": 2.2015023039907863e-06, + "loss": 1.05, + "step": 5365 + }, + { + "epoch": 0.4839247869414258, + "grad_norm": 1.9310592532477695, + "learning_rate": 2.2009210574381464e-06, + "loss": 0.9712, + "step": 5366 + }, + { + "epoch": 0.48401497046489605, + "grad_norm": 1.6430014700903326, + "learning_rate": 2.2003397937418134e-06, + "loss": 0.9336, + "step": 5367 + }, + { + "epoch": 0.48410515398836634, + "grad_norm": 0.6713950638685909, + "learning_rate": 2.1997585129513852e-06, + "loss": 0.8401, + "step": 5368 + }, + { + "epoch": 0.4841953375118366, + "grad_norm": 1.5593303574648791, + "learning_rate": 2.1991772151164595e-06, + "loss": 1.015, + "step": 5369 + }, + { + "epoch": 0.48428552103530687, + "grad_norm": 1.2973837716261547, + "learning_rate": 2.1985959002866346e-06, + "loss": 1.0345, + "step": 5370 + }, + { + "epoch": 0.4843757045587771, + "grad_norm": 2.0940738256751907, + "learning_rate": 2.198014568511513e-06, + "loss": 1.0563, + "step": 5371 + }, + { + "epoch": 0.4844658880822474, + "grad_norm": 0.6593310825131411, + "learning_rate": 2.1974332198406965e-06, + "loss": 0.8125, + "step": 5372 + }, + { + "epoch": 0.4845560716057176, + "grad_norm": 1.8496911615901641, + "learning_rate": 2.196851854323789e-06, + "loss": 0.9757, + "step": 5373 + }, + { + "epoch": 0.4846462551291879, + "grad_norm": 1.661929630731597, + "learning_rate": 2.196270472010396e-06, + "loss": 0.9757, + "step": 5374 + }, + { + "epoch": 0.48473643865265814, + "grad_norm": 1.6150344300512156, + "learning_rate": 2.195689072950124e-06, + "loss": 0.9899, + "step": 5375 + }, + { + "epoch": 0.48482662217612843, + "grad_norm": 0.5851415950636006, + "learning_rate": 2.195107657192581e-06, + "loss": 0.7998, + "step": 5376 + }, + { + "epoch": 0.48491680569959866, + "grad_norm": 0.5905350934581978, + "learning_rate": 2.194526224787378e-06, + "loss": 0.7539, + "step": 5377 + }, + { + "epoch": 0.48500698922306895, + "grad_norm": 1.8969262783905978, + "learning_rate": 2.1939447757841236e-06, + "loss": 0.9599, + "step": 5378 + }, + { + "epoch": 0.4850971727465392, + "grad_norm": 2.2335049729318555, + "learning_rate": 2.193363310232432e-06, + "loss": 1.0518, + "step": 5379 + }, + { + "epoch": 0.4851873562700095, + "grad_norm": 1.370242123335325, + "learning_rate": 2.192781828181917e-06, + "loss": 1.0907, + "step": 5380 + }, + { + "epoch": 0.4852775397934797, + "grad_norm": 1.6030544867818912, + "learning_rate": 2.192200329682193e-06, + "loss": 1.0176, + "step": 5381 + }, + { + "epoch": 0.48536772331695, + "grad_norm": 1.2806725480061931, + "learning_rate": 2.1916188147828767e-06, + "loss": 0.9894, + "step": 5382 + }, + { + "epoch": 0.48545790684042023, + "grad_norm": 1.5291693428424975, + "learning_rate": 2.191037283533587e-06, + "loss": 1.0048, + "step": 5383 + }, + { + "epoch": 0.4855480903638905, + "grad_norm": 2.410500277753623, + "learning_rate": 2.1904557359839428e-06, + "loss": 0.986, + "step": 5384 + }, + { + "epoch": 0.4856382738873608, + "grad_norm": 2.429645501169366, + "learning_rate": 2.189874172183565e-06, + "loss": 0.9716, + "step": 5385 + }, + { + "epoch": 0.48572845741083104, + "grad_norm": 1.452321558511412, + "learning_rate": 2.1892925921820763e-06, + "loss": 0.9612, + "step": 5386 + }, + { + "epoch": 0.4858186409343013, + "grad_norm": 1.6512412864114399, + "learning_rate": 2.1887109960290994e-06, + "loss": 0.9858, + "step": 5387 + }, + { + "epoch": 0.48590882445777156, + "grad_norm": 3.355439475734495, + "learning_rate": 2.18812938377426e-06, + "loss": 1.084, + "step": 5388 + }, + { + "epoch": 0.48599900798124185, + "grad_norm": 1.6317415512611992, + "learning_rate": 2.187547755467184e-06, + "loss": 1.0262, + "step": 5389 + }, + { + "epoch": 0.4860891915047121, + "grad_norm": 1.5171694407778022, + "learning_rate": 2.1869661111574994e-06, + "loss": 0.8828, + "step": 5390 + }, + { + "epoch": 0.48617937502818237, + "grad_norm": 1.959943542251219, + "learning_rate": 2.1863844508948353e-06, + "loss": 1.0444, + "step": 5391 + }, + { + "epoch": 0.4862695585516526, + "grad_norm": 1.4487368047078757, + "learning_rate": 2.185802774728823e-06, + "loss": 1.0062, + "step": 5392 + }, + { + "epoch": 0.4863597420751229, + "grad_norm": 1.3328044358934128, + "learning_rate": 2.1852210827090927e-06, + "loss": 0.9435, + "step": 5393 + }, + { + "epoch": 0.4864499255985931, + "grad_norm": 1.831945876875401, + "learning_rate": 2.184639374885278e-06, + "loss": 0.9318, + "step": 5394 + }, + { + "epoch": 0.4865401091220634, + "grad_norm": 2.393438839920133, + "learning_rate": 2.184057651307014e-06, + "loss": 1.0869, + "step": 5395 + }, + { + "epoch": 0.48663029264553365, + "grad_norm": 2.070858980425534, + "learning_rate": 2.183475912023937e-06, + "loss": 1.0611, + "step": 5396 + }, + { + "epoch": 0.48672047616900393, + "grad_norm": 4.964700260084727, + "learning_rate": 2.1828941570856826e-06, + "loss": 1.0041, + "step": 5397 + }, + { + "epoch": 0.48681065969247417, + "grad_norm": 0.6313499692563247, + "learning_rate": 2.1823123865418903e-06, + "loss": 0.8091, + "step": 5398 + }, + { + "epoch": 0.48690084321594446, + "grad_norm": 1.7070577169850005, + "learning_rate": 2.1817306004422e-06, + "loss": 0.9628, + "step": 5399 + }, + { + "epoch": 0.4869910267394147, + "grad_norm": 2.240223413122974, + "learning_rate": 2.1811487988362527e-06, + "loss": 1.0129, + "step": 5400 + }, + { + "epoch": 0.487081210262885, + "grad_norm": 2.0025685029808242, + "learning_rate": 2.1805669817736917e-06, + "loss": 1.0643, + "step": 5401 + }, + { + "epoch": 0.4871713937863552, + "grad_norm": 1.7536696378495038, + "learning_rate": 2.17998514930416e-06, + "loss": 0.8144, + "step": 5402 + }, + { + "epoch": 0.4872615773098255, + "grad_norm": 2.0095934854380593, + "learning_rate": 2.1794033014773025e-06, + "loss": 0.9915, + "step": 5403 + }, + { + "epoch": 0.48735176083329573, + "grad_norm": 1.487536466776608, + "learning_rate": 2.178821438342766e-06, + "loss": 1.055, + "step": 5404 + }, + { + "epoch": 0.487441944356766, + "grad_norm": 1.6312998650774282, + "learning_rate": 2.1782395599501996e-06, + "loss": 0.9836, + "step": 5405 + }, + { + "epoch": 0.48753212788023625, + "grad_norm": 0.6510933639345794, + "learning_rate": 2.1776576663492498e-06, + "loss": 0.8215, + "step": 5406 + }, + { + "epoch": 0.48762231140370654, + "grad_norm": 2.2117212207258916, + "learning_rate": 2.177075757589569e-06, + "loss": 1.0771, + "step": 5407 + }, + { + "epoch": 0.48771249492717683, + "grad_norm": 1.4092140325747942, + "learning_rate": 2.176493833720808e-06, + "loss": 0.9164, + "step": 5408 + }, + { + "epoch": 0.48780267845064706, + "grad_norm": 1.7857314504070936, + "learning_rate": 2.1759118947926195e-06, + "loss": 0.9566, + "step": 5409 + }, + { + "epoch": 0.48789286197411735, + "grad_norm": 1.4955055933279866, + "learning_rate": 2.1753299408546587e-06, + "loss": 1.0013, + "step": 5410 + }, + { + "epoch": 0.4879830454975876, + "grad_norm": 1.3211625652561567, + "learning_rate": 2.1747479719565803e-06, + "loss": 0.9957, + "step": 5411 + }, + { + "epoch": 0.4880732290210579, + "grad_norm": 1.3537423520563423, + "learning_rate": 2.174165988148042e-06, + "loss": 1.0525, + "step": 5412 + }, + { + "epoch": 0.4881634125445281, + "grad_norm": 1.7280209970478957, + "learning_rate": 2.1735839894787003e-06, + "loss": 0.9576, + "step": 5413 + }, + { + "epoch": 0.4882535960679984, + "grad_norm": 0.7626207928268715, + "learning_rate": 2.1730019759982163e-06, + "loss": 0.866, + "step": 5414 + }, + { + "epoch": 0.48834377959146863, + "grad_norm": 1.5195689061850834, + "learning_rate": 2.172419947756249e-06, + "loss": 1.0437, + "step": 5415 + }, + { + "epoch": 0.4884339631149389, + "grad_norm": 1.7582211167864854, + "learning_rate": 2.171837904802461e-06, + "loss": 1.0367, + "step": 5416 + }, + { + "epoch": 0.48852414663840915, + "grad_norm": 1.4568160565471282, + "learning_rate": 2.171255847186516e-06, + "loss": 0.9883, + "step": 5417 + }, + { + "epoch": 0.48861433016187944, + "grad_norm": 1.5809046371914122, + "learning_rate": 2.1706737749580783e-06, + "loss": 0.9815, + "step": 5418 + }, + { + "epoch": 0.4887045136853497, + "grad_norm": 1.6979218543349823, + "learning_rate": 2.1700916881668127e-06, + "loss": 1.0662, + "step": 5419 + }, + { + "epoch": 0.48879469720881996, + "grad_norm": 1.6183736385717755, + "learning_rate": 2.1695095868623862e-06, + "loss": 1.079, + "step": 5420 + }, + { + "epoch": 0.4888848807322902, + "grad_norm": 1.8130567945073461, + "learning_rate": 2.168927471094467e-06, + "loss": 0.9233, + "step": 5421 + }, + { + "epoch": 0.4889750642557605, + "grad_norm": 2.184679992950991, + "learning_rate": 2.168345340912725e-06, + "loss": 1.0078, + "step": 5422 + }, + { + "epoch": 0.4890652477792307, + "grad_norm": 1.7791146311718111, + "learning_rate": 2.1677631963668298e-06, + "loss": 1.011, + "step": 5423 + }, + { + "epoch": 0.489155431302701, + "grad_norm": 1.7537980417970402, + "learning_rate": 2.167181037506453e-06, + "loss": 0.9498, + "step": 5424 + }, + { + "epoch": 0.48924561482617124, + "grad_norm": 1.5541491319146168, + "learning_rate": 2.1665988643812693e-06, + "loss": 0.9278, + "step": 5425 + }, + { + "epoch": 0.4893357983496415, + "grad_norm": 1.7405533589219597, + "learning_rate": 2.166016677040951e-06, + "loss": 0.9849, + "step": 5426 + }, + { + "epoch": 0.48942598187311176, + "grad_norm": 1.558232698234029, + "learning_rate": 2.165434475535175e-06, + "loss": 0.9804, + "step": 5427 + }, + { + "epoch": 0.48951616539658205, + "grad_norm": 1.653679995046271, + "learning_rate": 2.1648522599136173e-06, + "loss": 1.0351, + "step": 5428 + }, + { + "epoch": 0.4896063489200523, + "grad_norm": 1.469257023730779, + "learning_rate": 2.164270030225956e-06, + "loss": 0.9632, + "step": 5429 + }, + { + "epoch": 0.48969653244352257, + "grad_norm": 1.366714700963464, + "learning_rate": 2.16368778652187e-06, + "loss": 0.9029, + "step": 5430 + }, + { + "epoch": 0.4897867159669928, + "grad_norm": 1.5651607746501763, + "learning_rate": 2.163105528851039e-06, + "loss": 1.0301, + "step": 5431 + }, + { + "epoch": 0.4898768994904631, + "grad_norm": 1.5076918479598347, + "learning_rate": 2.1625232572631448e-06, + "loss": 0.9969, + "step": 5432 + }, + { + "epoch": 0.4899670830139334, + "grad_norm": 0.745924246342268, + "learning_rate": 2.161940971807871e-06, + "loss": 0.8284, + "step": 5433 + }, + { + "epoch": 0.4900572665374036, + "grad_norm": 2.554146525649479, + "learning_rate": 2.1613586725348994e-06, + "loss": 0.9784, + "step": 5434 + }, + { + "epoch": 0.4901474500608739, + "grad_norm": 1.555233270130521, + "learning_rate": 2.1607763594939176e-06, + "loss": 1.0692, + "step": 5435 + }, + { + "epoch": 0.49023763358434413, + "grad_norm": 2.405850860387393, + "learning_rate": 2.1601940327346093e-06, + "loss": 0.9021, + "step": 5436 + }, + { + "epoch": 0.4903278171078144, + "grad_norm": 1.5260371792914995, + "learning_rate": 2.159611692306663e-06, + "loss": 1.0633, + "step": 5437 + }, + { + "epoch": 0.49041800063128466, + "grad_norm": 1.6994674325254007, + "learning_rate": 2.1590293382597667e-06, + "loss": 0.9499, + "step": 5438 + }, + { + "epoch": 0.49050818415475494, + "grad_norm": 1.6419359328424492, + "learning_rate": 2.1584469706436102e-06, + "loss": 0.9301, + "step": 5439 + }, + { + "epoch": 0.4905983676782252, + "grad_norm": 1.5379383609757071, + "learning_rate": 2.1578645895078855e-06, + "loss": 0.9043, + "step": 5440 + }, + { + "epoch": 0.49068855120169547, + "grad_norm": 1.6620465456140503, + "learning_rate": 2.157282194902283e-06, + "loss": 1.0523, + "step": 5441 + }, + { + "epoch": 0.4907787347251657, + "grad_norm": 2.1592798369398736, + "learning_rate": 2.1566997868764965e-06, + "loss": 0.9308, + "step": 5442 + }, + { + "epoch": 0.490868918248636, + "grad_norm": 1.7414108392681509, + "learning_rate": 2.15611736548022e-06, + "loss": 0.9508, + "step": 5443 + }, + { + "epoch": 0.4909591017721062, + "grad_norm": 1.6471749483171998, + "learning_rate": 2.155534930763149e-06, + "loss": 1.0, + "step": 5444 + }, + { + "epoch": 0.4910492852955765, + "grad_norm": 1.7380889443310321, + "learning_rate": 2.1549524827749804e-06, + "loss": 0.8749, + "step": 5445 + }, + { + "epoch": 0.49113946881904674, + "grad_norm": 1.4759382108920467, + "learning_rate": 2.1543700215654115e-06, + "loss": 0.8421, + "step": 5446 + }, + { + "epoch": 0.49122965234251703, + "grad_norm": 1.675790924542295, + "learning_rate": 2.153787547184141e-06, + "loss": 0.9802, + "step": 5447 + }, + { + "epoch": 0.49131983586598726, + "grad_norm": 1.5936387715140705, + "learning_rate": 2.1532050596808695e-06, + "loss": 0.9206, + "step": 5448 + }, + { + "epoch": 0.49141001938945755, + "grad_norm": 1.7879833314916604, + "learning_rate": 2.152622559105297e-06, + "loss": 1.0366, + "step": 5449 + }, + { + "epoch": 0.4915002029129278, + "grad_norm": 1.8746520037373808, + "learning_rate": 2.152040045507126e-06, + "loss": 1.0359, + "step": 5450 + }, + { + "epoch": 0.4915903864363981, + "grad_norm": 1.89241479064525, + "learning_rate": 2.1514575189360607e-06, + "loss": 1.0048, + "step": 5451 + }, + { + "epoch": 0.4916805699598683, + "grad_norm": 1.5756781465372995, + "learning_rate": 2.1508749794418043e-06, + "loss": 0.9463, + "step": 5452 + }, + { + "epoch": 0.4917707534833386, + "grad_norm": 1.56172595878796, + "learning_rate": 2.1502924270740626e-06, + "loss": 0.9615, + "step": 5453 + }, + { + "epoch": 0.49186093700680883, + "grad_norm": 1.81858337530485, + "learning_rate": 2.1497098618825427e-06, + "loss": 0.9923, + "step": 5454 + }, + { + "epoch": 0.4919511205302791, + "grad_norm": 1.7155833115923815, + "learning_rate": 2.1491272839169516e-06, + "loss": 1.0327, + "step": 5455 + }, + { + "epoch": 0.4920413040537494, + "grad_norm": 1.5612112452489797, + "learning_rate": 2.1485446932269986e-06, + "loss": 1.1022, + "step": 5456 + }, + { + "epoch": 0.49213148757721964, + "grad_norm": 2.1162852937444523, + "learning_rate": 2.147962089862393e-06, + "loss": 1.0275, + "step": 5457 + }, + { + "epoch": 0.49222167110068993, + "grad_norm": 2.2554154392729817, + "learning_rate": 2.1473794738728462e-06, + "loss": 0.8951, + "step": 5458 + }, + { + "epoch": 0.49231185462416016, + "grad_norm": 1.576107283808581, + "learning_rate": 2.14679684530807e-06, + "loss": 1.0238, + "step": 5459 + }, + { + "epoch": 0.49240203814763045, + "grad_norm": 1.656148223628897, + "learning_rate": 2.1462142042177774e-06, + "loss": 0.9736, + "step": 5460 + }, + { + "epoch": 0.4924922216711007, + "grad_norm": 2.6565816335658803, + "learning_rate": 2.145631550651683e-06, + "loss": 0.9868, + "step": 5461 + }, + { + "epoch": 0.49258240519457097, + "grad_norm": 1.7538652648894852, + "learning_rate": 2.1450488846595016e-06, + "loss": 1.0881, + "step": 5462 + }, + { + "epoch": 0.4926725887180412, + "grad_norm": 1.7718095743792386, + "learning_rate": 2.14446620629095e-06, + "loss": 0.9677, + "step": 5463 + }, + { + "epoch": 0.4927627722415115, + "grad_norm": 1.368237511829977, + "learning_rate": 2.1438835155957445e-06, + "loss": 1.0573, + "step": 5464 + }, + { + "epoch": 0.4928529557649817, + "grad_norm": 1.6203173145307268, + "learning_rate": 2.143300812623604e-06, + "loss": 1.0287, + "step": 5465 + }, + { + "epoch": 0.492943139288452, + "grad_norm": 2.188398340232469, + "learning_rate": 2.1427180974242485e-06, + "loss": 0.971, + "step": 5466 + }, + { + "epoch": 0.49303332281192225, + "grad_norm": 1.5693983007453673, + "learning_rate": 2.142135370047398e-06, + "loss": 1.0168, + "step": 5467 + }, + { + "epoch": 0.49312350633539254, + "grad_norm": 2.002342044920387, + "learning_rate": 2.1415526305427735e-06, + "loss": 1.0205, + "step": 5468 + }, + { + "epoch": 0.49321368985886277, + "grad_norm": 1.3442588995790934, + "learning_rate": 2.140969878960098e-06, + "loss": 0.9131, + "step": 5469 + }, + { + "epoch": 0.49330387338233306, + "grad_norm": 1.459174903368546, + "learning_rate": 2.1403871153490956e-06, + "loss": 1.0492, + "step": 5470 + }, + { + "epoch": 0.4933940569058033, + "grad_norm": 1.5546076001669444, + "learning_rate": 2.13980433975949e-06, + "loss": 1.1152, + "step": 5471 + }, + { + "epoch": 0.4934842404292736, + "grad_norm": 1.5071886580603364, + "learning_rate": 2.1392215522410076e-06, + "loss": 1.0347, + "step": 5472 + }, + { + "epoch": 0.4935744239527438, + "grad_norm": 2.6621281825385705, + "learning_rate": 2.1386387528433743e-06, + "loss": 1.1037, + "step": 5473 + }, + { + "epoch": 0.4936646074762141, + "grad_norm": 2.0135225671716523, + "learning_rate": 2.1380559416163186e-06, + "loss": 1.0154, + "step": 5474 + }, + { + "epoch": 0.49375479099968433, + "grad_norm": 2.40651800471406, + "learning_rate": 2.1374731186095685e-06, + "loss": 1.0357, + "step": 5475 + }, + { + "epoch": 0.4938449745231546, + "grad_norm": 1.47460266515961, + "learning_rate": 2.136890283872854e-06, + "loss": 0.9412, + "step": 5476 + }, + { + "epoch": 0.49393515804662486, + "grad_norm": 1.6201352146610448, + "learning_rate": 2.136307437455906e-06, + "loss": 0.9704, + "step": 5477 + }, + { + "epoch": 0.49402534157009514, + "grad_norm": 1.3632024616548313, + "learning_rate": 2.135724579408456e-06, + "loss": 0.94, + "step": 5478 + }, + { + "epoch": 0.49411552509356543, + "grad_norm": 1.7951331278915326, + "learning_rate": 2.1351417097802356e-06, + "loss": 1.087, + "step": 5479 + }, + { + "epoch": 0.49420570861703567, + "grad_norm": 1.9329862622889367, + "learning_rate": 2.1345588286209798e-06, + "loss": 0.8716, + "step": 5480 + }, + { + "epoch": 0.49429589214050595, + "grad_norm": 2.2840504636250363, + "learning_rate": 2.1339759359804227e-06, + "loss": 0.834, + "step": 5481 + }, + { + "epoch": 0.4943860756639762, + "grad_norm": 1.4685503235488426, + "learning_rate": 2.1333930319082997e-06, + "loss": 0.967, + "step": 5482 + }, + { + "epoch": 0.4944762591874465, + "grad_norm": 1.3647419009886401, + "learning_rate": 2.132810116454348e-06, + "loss": 0.9976, + "step": 5483 + }, + { + "epoch": 0.4945664427109167, + "grad_norm": 1.6120414495889586, + "learning_rate": 2.132227189668305e-06, + "loss": 0.9838, + "step": 5484 + }, + { + "epoch": 0.494656626234387, + "grad_norm": 1.731992580204107, + "learning_rate": 2.1316442515999096e-06, + "loss": 0.9373, + "step": 5485 + }, + { + "epoch": 0.49474680975785723, + "grad_norm": 2.5315020400417123, + "learning_rate": 2.1310613022989e-06, + "loss": 1.0205, + "step": 5486 + }, + { + "epoch": 0.4948369932813275, + "grad_norm": 1.9529435340504167, + "learning_rate": 2.130478341815017e-06, + "loss": 0.8377, + "step": 5487 + }, + { + "epoch": 0.49492717680479775, + "grad_norm": 1.599934564682909, + "learning_rate": 2.1298953701980033e-06, + "loss": 0.8903, + "step": 5488 + }, + { + "epoch": 0.49501736032826804, + "grad_norm": 2.1945431267269955, + "learning_rate": 2.1293123874976003e-06, + "loss": 0.9944, + "step": 5489 + }, + { + "epoch": 0.4951075438517383, + "grad_norm": 1.7242176825066784, + "learning_rate": 2.1287293937635513e-06, + "loss": 0.9387, + "step": 5490 + }, + { + "epoch": 0.49519772737520856, + "grad_norm": 1.5817950037002517, + "learning_rate": 2.1281463890456005e-06, + "loss": 1.045, + "step": 5491 + }, + { + "epoch": 0.4952879108986788, + "grad_norm": 1.6207277919007415, + "learning_rate": 2.127563373393493e-06, + "loss": 0.9862, + "step": 5492 + }, + { + "epoch": 0.4953780944221491, + "grad_norm": 1.587085402408826, + "learning_rate": 2.1269803468569756e-06, + "loss": 0.9825, + "step": 5493 + }, + { + "epoch": 0.4954682779456193, + "grad_norm": 1.530595444891978, + "learning_rate": 2.126397309485794e-06, + "loss": 1.0287, + "step": 5494 + }, + { + "epoch": 0.4955584614690896, + "grad_norm": 1.8002296645620157, + "learning_rate": 2.1258142613296983e-06, + "loss": 0.9463, + "step": 5495 + }, + { + "epoch": 0.49564864499255984, + "grad_norm": 1.3742203669622557, + "learning_rate": 2.125231202438435e-06, + "loss": 0.9658, + "step": 5496 + }, + { + "epoch": 0.49573882851603013, + "grad_norm": 1.410864612475202, + "learning_rate": 2.1246481328617553e-06, + "loss": 1.0219, + "step": 5497 + }, + { + "epoch": 0.49582901203950036, + "grad_norm": 5.628988759099919, + "learning_rate": 2.1240650526494096e-06, + "loss": 0.9746, + "step": 5498 + }, + { + "epoch": 0.49591919556297065, + "grad_norm": 5.0546659914293155, + "learning_rate": 2.1234819618511493e-06, + "loss": 0.9602, + "step": 5499 + }, + { + "epoch": 0.4960093790864409, + "grad_norm": 0.7390539172845618, + "learning_rate": 2.122898860516728e-06, + "loss": 0.8566, + "step": 5500 + }, + { + "epoch": 0.49609956260991117, + "grad_norm": 1.6272785644368437, + "learning_rate": 2.1223157486958976e-06, + "loss": 0.9779, + "step": 5501 + }, + { + "epoch": 0.4961897461333814, + "grad_norm": 1.3439872443553202, + "learning_rate": 2.1217326264384127e-06, + "loss": 0.8829, + "step": 5502 + }, + { + "epoch": 0.4962799296568517, + "grad_norm": 2.081309581739998, + "learning_rate": 2.1211494937940296e-06, + "loss": 1.0117, + "step": 5503 + }, + { + "epoch": 0.496370113180322, + "grad_norm": 1.4035688484137672, + "learning_rate": 2.1205663508125034e-06, + "loss": 1.0032, + "step": 5504 + }, + { + "epoch": 0.4964602967037922, + "grad_norm": 2.787164464351319, + "learning_rate": 2.1199831975435914e-06, + "loss": 0.9545, + "step": 5505 + }, + { + "epoch": 0.4965504802272625, + "grad_norm": 1.6203822769252363, + "learning_rate": 2.1194000340370517e-06, + "loss": 1.0362, + "step": 5506 + }, + { + "epoch": 0.49664066375073274, + "grad_norm": 1.7781381697074834, + "learning_rate": 2.1188168603426423e-06, + "loss": 0.9396, + "step": 5507 + }, + { + "epoch": 0.496730847274203, + "grad_norm": 2.5707960050851884, + "learning_rate": 2.118233676510123e-06, + "loss": 0.9532, + "step": 5508 + }, + { + "epoch": 0.49682103079767326, + "grad_norm": 1.683367224940689, + "learning_rate": 2.117650482589255e-06, + "loss": 1.0029, + "step": 5509 + }, + { + "epoch": 0.49691121432114355, + "grad_norm": 1.9954446532264345, + "learning_rate": 2.1170672786297988e-06, + "loss": 1.0274, + "step": 5510 + }, + { + "epoch": 0.4970013978446138, + "grad_norm": 1.3399976636382274, + "learning_rate": 2.1164840646815174e-06, + "loss": 0.9557, + "step": 5511 + }, + { + "epoch": 0.49709158136808407, + "grad_norm": 1.6774312323902263, + "learning_rate": 2.1159008407941726e-06, + "loss": 1.0725, + "step": 5512 + }, + { + "epoch": 0.4971817648915543, + "grad_norm": 1.55971049214443, + "learning_rate": 2.1153176070175293e-06, + "loss": 0.9809, + "step": 5513 + }, + { + "epoch": 0.4972719484150246, + "grad_norm": 1.8597818778949744, + "learning_rate": 2.114734363401352e-06, + "loss": 0.8831, + "step": 5514 + }, + { + "epoch": 0.4973621319384948, + "grad_norm": 1.909681327459029, + "learning_rate": 2.1141511099954056e-06, + "loss": 0.9867, + "step": 5515 + }, + { + "epoch": 0.4974523154619651, + "grad_norm": 1.433760963271645, + "learning_rate": 2.1135678468494576e-06, + "loss": 0.9685, + "step": 5516 + }, + { + "epoch": 0.49754249898543534, + "grad_norm": 2.1741199554576087, + "learning_rate": 2.112984574013275e-06, + "loss": 1.028, + "step": 5517 + }, + { + "epoch": 0.49763268250890563, + "grad_norm": 1.552481615138884, + "learning_rate": 2.112401291536625e-06, + "loss": 0.9451, + "step": 5518 + }, + { + "epoch": 0.49772286603237587, + "grad_norm": 1.4890698523805237, + "learning_rate": 2.111817999469278e-06, + "loss": 1.0241, + "step": 5519 + }, + { + "epoch": 0.49781304955584615, + "grad_norm": 1.8553812427966998, + "learning_rate": 2.1112346978610016e-06, + "loss": 0.94, + "step": 5520 + }, + { + "epoch": 0.4979032330793164, + "grad_norm": 1.6052219915912591, + "learning_rate": 2.1106513867615678e-06, + "loss": 0.9429, + "step": 5521 + }, + { + "epoch": 0.4979934166027867, + "grad_norm": 1.3005370864596042, + "learning_rate": 2.110068066220748e-06, + "loss": 0.9907, + "step": 5522 + }, + { + "epoch": 0.4980836001262569, + "grad_norm": 1.2553979194947185, + "learning_rate": 2.109484736288313e-06, + "loss": 1.0567, + "step": 5523 + }, + { + "epoch": 0.4981737836497272, + "grad_norm": 1.7149094414782704, + "learning_rate": 2.108901397014037e-06, + "loss": 1.0033, + "step": 5524 + }, + { + "epoch": 0.49826396717319743, + "grad_norm": 1.4492740260684722, + "learning_rate": 2.1083180484476934e-06, + "loss": 0.9932, + "step": 5525 + }, + { + "epoch": 0.4983541506966677, + "grad_norm": 1.578550375839333, + "learning_rate": 2.1077346906390567e-06, + "loss": 1.0084, + "step": 5526 + }, + { + "epoch": 0.498444334220138, + "grad_norm": 1.699083414936558, + "learning_rate": 2.107151323637902e-06, + "loss": 1.0097, + "step": 5527 + }, + { + "epoch": 0.49853451774360824, + "grad_norm": 1.7521003652766132, + "learning_rate": 2.106567947494006e-06, + "loss": 0.9762, + "step": 5528 + }, + { + "epoch": 0.49862470126707853, + "grad_norm": 1.8554417015915, + "learning_rate": 2.1059845622571447e-06, + "loss": 1.0545, + "step": 5529 + }, + { + "epoch": 0.49871488479054876, + "grad_norm": 1.4667343756189928, + "learning_rate": 2.1054011679770956e-06, + "loss": 1.0294, + "step": 5530 + }, + { + "epoch": 0.49880506831401905, + "grad_norm": 1.5829943327473968, + "learning_rate": 2.104817764703638e-06, + "loss": 1.0003, + "step": 5531 + }, + { + "epoch": 0.4988952518374893, + "grad_norm": 3.0170452351072883, + "learning_rate": 2.1042343524865516e-06, + "loss": 0.9941, + "step": 5532 + }, + { + "epoch": 0.4989854353609596, + "grad_norm": 1.5736961538542826, + "learning_rate": 2.103650931375615e-06, + "loss": 1.1002, + "step": 5533 + }, + { + "epoch": 0.4990756188844298, + "grad_norm": 2.054531309043819, + "learning_rate": 2.1030675014206094e-06, + "loss": 0.9279, + "step": 5534 + }, + { + "epoch": 0.4991658024079001, + "grad_norm": 1.952749353524207, + "learning_rate": 2.1024840626713166e-06, + "loss": 0.9234, + "step": 5535 + }, + { + "epoch": 0.4992559859313703, + "grad_norm": 1.6516504298597978, + "learning_rate": 2.1019006151775177e-06, + "loss": 1.0911, + "step": 5536 + }, + { + "epoch": 0.4993461694548406, + "grad_norm": 1.6556493281718736, + "learning_rate": 2.101317158988997e-06, + "loss": 0.9783, + "step": 5537 + }, + { + "epoch": 0.49943635297831085, + "grad_norm": 2.631512419963957, + "learning_rate": 2.1007336941555374e-06, + "loss": 0.965, + "step": 5538 + }, + { + "epoch": 0.49952653650178114, + "grad_norm": 1.6258422429505275, + "learning_rate": 2.1001502207269238e-06, + "loss": 0.9528, + "step": 5539 + }, + { + "epoch": 0.49961672002525137, + "grad_norm": 1.8461902439827074, + "learning_rate": 2.0995667387529407e-06, + "loss": 0.947, + "step": 5540 + }, + { + "epoch": 0.49970690354872166, + "grad_norm": 3.6146385489139066, + "learning_rate": 2.098983248283375e-06, + "loss": 0.9386, + "step": 5541 + }, + { + "epoch": 0.4997970870721919, + "grad_norm": 1.944730618020465, + "learning_rate": 2.098399749368012e-06, + "loss": 0.9901, + "step": 5542 + }, + { + "epoch": 0.4998872705956622, + "grad_norm": 0.628320219533793, + "learning_rate": 2.09781624205664e-06, + "loss": 0.8074, + "step": 5543 + }, + { + "epoch": 0.4999774541191324, + "grad_norm": 1.4469651130314367, + "learning_rate": 2.0972327263990477e-06, + "loss": 1.0395, + "step": 5544 + }, + { + "epoch": 0.5000676376426026, + "grad_norm": 2.076562394371059, + "learning_rate": 2.0966492024450226e-06, + "loss": 0.9232, + "step": 5545 + }, + { + "epoch": 0.500157821166073, + "grad_norm": 1.6638693540711038, + "learning_rate": 2.0960656702443545e-06, + "loss": 1.1378, + "step": 5546 + }, + { + "epoch": 0.5002480046895432, + "grad_norm": 1.4907264946826606, + "learning_rate": 2.0954821298468343e-06, + "loss": 0.9885, + "step": 5547 + }, + { + "epoch": 0.5003381882130135, + "grad_norm": 1.5409070110844074, + "learning_rate": 2.0948985813022513e-06, + "loss": 1.053, + "step": 5548 + }, + { + "epoch": 0.5004283717364837, + "grad_norm": 1.8091007453063777, + "learning_rate": 2.094315024660399e-06, + "loss": 1.0287, + "step": 5549 + }, + { + "epoch": 0.500518555259954, + "grad_norm": 3.999468529679595, + "learning_rate": 2.0937314599710676e-06, + "loss": 1.0118, + "step": 5550 + }, + { + "epoch": 0.5006087387834243, + "grad_norm": 1.598129239753646, + "learning_rate": 2.0931478872840526e-06, + "loss": 0.981, + "step": 5551 + }, + { + "epoch": 0.5006989223068945, + "grad_norm": 1.6115847481491374, + "learning_rate": 2.092564306649145e-06, + "loss": 0.9696, + "step": 5552 + }, + { + "epoch": 0.5007891058303648, + "grad_norm": 1.5179126424820728, + "learning_rate": 2.091980718116141e-06, + "loss": 1.023, + "step": 5553 + }, + { + "epoch": 0.5008792893538351, + "grad_norm": 1.5214835131426072, + "learning_rate": 2.091397121734835e-06, + "loss": 1.0284, + "step": 5554 + }, + { + "epoch": 0.5009694728773053, + "grad_norm": 1.9367670087814697, + "learning_rate": 2.090813517555022e-06, + "loss": 1.0151, + "step": 5555 + }, + { + "epoch": 0.5010596564007755, + "grad_norm": 1.7042517478366488, + "learning_rate": 2.0902299056265e-06, + "loss": 0.9337, + "step": 5556 + }, + { + "epoch": 0.5011498399242459, + "grad_norm": 1.878400706843287, + "learning_rate": 2.0896462859990643e-06, + "loss": 0.9708, + "step": 5557 + }, + { + "epoch": 0.5012400234477161, + "grad_norm": 1.5282512027605324, + "learning_rate": 2.089062658722513e-06, + "loss": 1.0506, + "step": 5558 + }, + { + "epoch": 0.5013302069711864, + "grad_norm": 9.418162359322677, + "learning_rate": 2.0884790238466452e-06, + "loss": 1.0384, + "step": 5559 + }, + { + "epoch": 0.5014203904946566, + "grad_norm": 2.109995884823515, + "learning_rate": 2.087895381421259e-06, + "loss": 1.0223, + "step": 5560 + }, + { + "epoch": 0.5015105740181269, + "grad_norm": 1.8545483042827682, + "learning_rate": 2.087311731496154e-06, + "loss": 0.9737, + "step": 5561 + }, + { + "epoch": 0.5016007575415972, + "grad_norm": 1.3236426166353141, + "learning_rate": 2.08672807412113e-06, + "loss": 0.9808, + "step": 5562 + }, + { + "epoch": 0.5016909410650674, + "grad_norm": 1.967311394072209, + "learning_rate": 2.08614440934599e-06, + "loss": 1.0807, + "step": 5563 + }, + { + "epoch": 0.5017811245885376, + "grad_norm": 5.410190094659481, + "learning_rate": 2.0855607372205337e-06, + "loss": 0.9051, + "step": 5564 + }, + { + "epoch": 0.501871308112008, + "grad_norm": 1.6046181906169117, + "learning_rate": 2.0849770577945623e-06, + "loss": 1.0014, + "step": 5565 + }, + { + "epoch": 0.5019614916354782, + "grad_norm": 1.402637971493579, + "learning_rate": 2.084393371117881e-06, + "loss": 1.024, + "step": 5566 + }, + { + "epoch": 0.5020516751589484, + "grad_norm": 1.3699170141076746, + "learning_rate": 2.0838096772402902e-06, + "loss": 0.9234, + "step": 5567 + }, + { + "epoch": 0.5021418586824187, + "grad_norm": 7.856341828968095, + "learning_rate": 2.0832259762115973e-06, + "loss": 0.8775, + "step": 5568 + }, + { + "epoch": 0.502232042205889, + "grad_norm": 1.5020668572074674, + "learning_rate": 2.082642268081605e-06, + "loss": 1.0162, + "step": 5569 + }, + { + "epoch": 0.5023222257293593, + "grad_norm": 1.733852892121106, + "learning_rate": 2.082058552900118e-06, + "loss": 0.9421, + "step": 5570 + }, + { + "epoch": 0.5024124092528295, + "grad_norm": 3.230801879976346, + "learning_rate": 2.081474830716944e-06, + "loss": 0.9656, + "step": 5571 + }, + { + "epoch": 0.5025025927762997, + "grad_norm": 1.9861909261937003, + "learning_rate": 2.080891101581887e-06, + "loss": 1.019, + "step": 5572 + }, + { + "epoch": 0.5025927762997701, + "grad_norm": 1.8018501733610404, + "learning_rate": 2.080307365544755e-06, + "loss": 1.0347, + "step": 5573 + }, + { + "epoch": 0.5026829598232403, + "grad_norm": 2.118939677612565, + "learning_rate": 2.0797236226553567e-06, + "loss": 1.0314, + "step": 5574 + }, + { + "epoch": 0.5027731433467105, + "grad_norm": 1.6589981168353412, + "learning_rate": 2.079139872963499e-06, + "loss": 0.928, + "step": 5575 + }, + { + "epoch": 0.5028633268701809, + "grad_norm": 2.038449250005659, + "learning_rate": 2.078556116518991e-06, + "loss": 0.8876, + "step": 5576 + }, + { + "epoch": 0.5029535103936511, + "grad_norm": 3.2527902802892084, + "learning_rate": 2.077972353371642e-06, + "loss": 0.9671, + "step": 5577 + }, + { + "epoch": 0.5030436939171213, + "grad_norm": 1.5201766107250203, + "learning_rate": 2.077388583571262e-06, + "loss": 0.9374, + "step": 5578 + }, + { + "epoch": 0.5031338774405916, + "grad_norm": 0.655439489391251, + "learning_rate": 2.0768048071676608e-06, + "loss": 0.7509, + "step": 5579 + }, + { + "epoch": 0.5032240609640619, + "grad_norm": 1.5611545110230434, + "learning_rate": 2.0762210242106505e-06, + "loss": 0.9527, + "step": 5580 + }, + { + "epoch": 0.5033142444875321, + "grad_norm": 2.3245093412418787, + "learning_rate": 2.0756372347500424e-06, + "loss": 1.0012, + "step": 5581 + }, + { + "epoch": 0.5034044280110024, + "grad_norm": 2.1043775345798656, + "learning_rate": 2.0750534388356473e-06, + "loss": 1.0946, + "step": 5582 + }, + { + "epoch": 0.5034946115344726, + "grad_norm": 1.716244970148972, + "learning_rate": 2.07446963651728e-06, + "loss": 0.9375, + "step": 5583 + }, + { + "epoch": 0.503584795057943, + "grad_norm": 1.6354870953145517, + "learning_rate": 2.0738858278447516e-06, + "loss": 0.9268, + "step": 5584 + }, + { + "epoch": 0.5036749785814132, + "grad_norm": 1.5514537976805112, + "learning_rate": 2.073302012867878e-06, + "loss": 1.0811, + "step": 5585 + }, + { + "epoch": 0.5037651621048834, + "grad_norm": 1.533945164034839, + "learning_rate": 2.0727181916364725e-06, + "loss": 1.0252, + "step": 5586 + }, + { + "epoch": 0.5038553456283537, + "grad_norm": 1.473164114159697, + "learning_rate": 2.0721343642003493e-06, + "loss": 1.0287, + "step": 5587 + }, + { + "epoch": 0.503945529151824, + "grad_norm": 1.8245218094671918, + "learning_rate": 2.0715505306093247e-06, + "loss": 1.0197, + "step": 5588 + }, + { + "epoch": 0.5040357126752942, + "grad_norm": 1.8156310037063899, + "learning_rate": 2.070966690913214e-06, + "loss": 1.0229, + "step": 5589 + }, + { + "epoch": 0.5041258961987645, + "grad_norm": 2.140585905957168, + "learning_rate": 2.0703828451618346e-06, + "loss": 0.8808, + "step": 5590 + }, + { + "epoch": 0.5042160797222347, + "grad_norm": 1.346877472912848, + "learning_rate": 2.069798993405002e-06, + "loss": 0.9885, + "step": 5591 + }, + { + "epoch": 0.504306263245705, + "grad_norm": 1.5614051797924804, + "learning_rate": 2.0692151356925345e-06, + "loss": 1.032, + "step": 5592 + }, + { + "epoch": 0.5043964467691753, + "grad_norm": 2.862097106299196, + "learning_rate": 2.068631272074251e-06, + "loss": 1.0137, + "step": 5593 + }, + { + "epoch": 0.5044866302926455, + "grad_norm": 2.560568849424206, + "learning_rate": 2.0680474025999676e-06, + "loss": 1.1205, + "step": 5594 + }, + { + "epoch": 0.5045768138161157, + "grad_norm": 1.3713657728786552, + "learning_rate": 2.0674635273195055e-06, + "loss": 0.8687, + "step": 5595 + }, + { + "epoch": 0.5046669973395861, + "grad_norm": 2.174113046733777, + "learning_rate": 2.066879646282682e-06, + "loss": 0.9862, + "step": 5596 + }, + { + "epoch": 0.5047571808630563, + "grad_norm": 1.982337686789961, + "learning_rate": 2.0662957595393194e-06, + "loss": 0.9653, + "step": 5597 + }, + { + "epoch": 0.5048473643865266, + "grad_norm": 1.8890162788081455, + "learning_rate": 2.0657118671392373e-06, + "loss": 1.0148, + "step": 5598 + }, + { + "epoch": 0.5049375479099969, + "grad_norm": 1.7430668138576966, + "learning_rate": 2.0651279691322558e-06, + "loss": 1.0467, + "step": 5599 + }, + { + "epoch": 0.5050277314334671, + "grad_norm": 1.5688642209481871, + "learning_rate": 2.0645440655681973e-06, + "loss": 0.9454, + "step": 5600 + }, + { + "epoch": 0.5051179149569374, + "grad_norm": 1.83537062963568, + "learning_rate": 2.0639601564968826e-06, + "loss": 1.056, + "step": 5601 + }, + { + "epoch": 0.5052080984804076, + "grad_norm": 0.6872546668575911, + "learning_rate": 2.0633762419681355e-06, + "loss": 0.9028, + "step": 5602 + }, + { + "epoch": 0.5052982820038779, + "grad_norm": 1.3134523751112888, + "learning_rate": 2.062792322031777e-06, + "loss": 1.0109, + "step": 5603 + }, + { + "epoch": 0.5053884655273482, + "grad_norm": 2.199072174011956, + "learning_rate": 2.062208396737632e-06, + "loss": 0.9478, + "step": 5604 + }, + { + "epoch": 0.5054786490508184, + "grad_norm": 1.4145742933895535, + "learning_rate": 2.0616244661355235e-06, + "loss": 1.0138, + "step": 5605 + }, + { + "epoch": 0.5055688325742886, + "grad_norm": 9.440134249233678, + "learning_rate": 2.0610405302752752e-06, + "loss": 1.0267, + "step": 5606 + }, + { + "epoch": 0.505659016097759, + "grad_norm": 1.5947464464563348, + "learning_rate": 2.060456589206713e-06, + "loss": 0.9822, + "step": 5607 + }, + { + "epoch": 0.5057491996212292, + "grad_norm": 1.6724540919687578, + "learning_rate": 2.0598726429796614e-06, + "loss": 0.9392, + "step": 5608 + }, + { + "epoch": 0.5058393831446995, + "grad_norm": 1.5786169814902118, + "learning_rate": 2.059288691643945e-06, + "loss": 0.9501, + "step": 5609 + }, + { + "epoch": 0.5059295666681697, + "grad_norm": 1.6601586195984928, + "learning_rate": 2.0587047352493913e-06, + "loss": 1.0745, + "step": 5610 + }, + { + "epoch": 0.50601975019164, + "grad_norm": 1.8701829704487587, + "learning_rate": 2.0581207738458248e-06, + "loss": 0.9699, + "step": 5611 + }, + { + "epoch": 0.5061099337151103, + "grad_norm": 1.8514084892764857, + "learning_rate": 2.0575368074830743e-06, + "loss": 0.9469, + "step": 5612 + }, + { + "epoch": 0.5062001172385805, + "grad_norm": 1.992980619639689, + "learning_rate": 2.0569528362109667e-06, + "loss": 0.9047, + "step": 5613 + }, + { + "epoch": 0.5062903007620507, + "grad_norm": 1.5110551032894208, + "learning_rate": 2.056368860079327e-06, + "loss": 0.987, + "step": 5614 + }, + { + "epoch": 0.5063804842855211, + "grad_norm": 2.0051421579778204, + "learning_rate": 2.0557848791379874e-06, + "loss": 0.9342, + "step": 5615 + }, + { + "epoch": 0.5064706678089913, + "grad_norm": 1.8006614264551204, + "learning_rate": 2.0552008934367734e-06, + "loss": 0.8908, + "step": 5616 + }, + { + "epoch": 0.5065608513324615, + "grad_norm": 1.8590858298726425, + "learning_rate": 2.0546169030255154e-06, + "loss": 1.063, + "step": 5617 + }, + { + "epoch": 0.5066510348559318, + "grad_norm": 1.974451317201263, + "learning_rate": 2.054032907954041e-06, + "loss": 1.0665, + "step": 5618 + }, + { + "epoch": 0.5067412183794021, + "grad_norm": 1.6693326211603816, + "learning_rate": 2.053448908272182e-06, + "loss": 0.9572, + "step": 5619 + }, + { + "epoch": 0.5068314019028723, + "grad_norm": 1.5051606730623142, + "learning_rate": 2.0528649040297673e-06, + "loss": 1.0263, + "step": 5620 + }, + { + "epoch": 0.5069215854263426, + "grad_norm": 1.418054078952686, + "learning_rate": 2.0522808952766266e-06, + "loss": 0.9947, + "step": 5621 + }, + { + "epoch": 0.5070117689498129, + "grad_norm": 1.5933231455318932, + "learning_rate": 2.0516968820625925e-06, + "loss": 1.0369, + "step": 5622 + }, + { + "epoch": 0.5071019524732832, + "grad_norm": 1.731719518680431, + "learning_rate": 2.051112864437495e-06, + "loss": 0.9795, + "step": 5623 + }, + { + "epoch": 0.5071921359967534, + "grad_norm": 2.1150007039603285, + "learning_rate": 2.050528842451166e-06, + "loss": 0.9591, + "step": 5624 + }, + { + "epoch": 0.5072823195202236, + "grad_norm": 1.5732831801621343, + "learning_rate": 2.049944816153438e-06, + "loss": 0.873, + "step": 5625 + }, + { + "epoch": 0.507372503043694, + "grad_norm": 1.4781923532766372, + "learning_rate": 2.049360785594142e-06, + "loss": 1.033, + "step": 5626 + }, + { + "epoch": 0.5074626865671642, + "grad_norm": 0.61871975815205, + "learning_rate": 2.048776750823113e-06, + "loss": 0.7476, + "step": 5627 + }, + { + "epoch": 0.5075528700906344, + "grad_norm": 0.6312609832110334, + "learning_rate": 2.0481927118901817e-06, + "loss": 0.8005, + "step": 5628 + }, + { + "epoch": 0.5076430536141047, + "grad_norm": 1.6770329280252396, + "learning_rate": 2.0476086688451824e-06, + "loss": 0.9344, + "step": 5629 + }, + { + "epoch": 0.507733237137575, + "grad_norm": 1.788329861481429, + "learning_rate": 2.04702462173795e-06, + "loss": 0.9819, + "step": 5630 + }, + { + "epoch": 0.5078234206610452, + "grad_norm": 3.5506281552185746, + "learning_rate": 2.0464405706183167e-06, + "loss": 0.9747, + "step": 5631 + }, + { + "epoch": 0.5079136041845155, + "grad_norm": 1.7287494048740248, + "learning_rate": 2.045856515536118e-06, + "loss": 1.0075, + "step": 5632 + }, + { + "epoch": 0.5080037877079857, + "grad_norm": 2.0844373702250265, + "learning_rate": 2.045272456541188e-06, + "loss": 1.0065, + "step": 5633 + }, + { + "epoch": 0.508093971231456, + "grad_norm": 1.388549327137624, + "learning_rate": 2.0446883936833635e-06, + "loss": 1.0595, + "step": 5634 + }, + { + "epoch": 0.5081841547549263, + "grad_norm": 1.7496293901651536, + "learning_rate": 2.0441043270124782e-06, + "loss": 0.9636, + "step": 5635 + }, + { + "epoch": 0.5082743382783965, + "grad_norm": 1.9107602652503966, + "learning_rate": 2.0435202565783683e-06, + "loss": 1.0172, + "step": 5636 + }, + { + "epoch": 0.5083645218018668, + "grad_norm": 1.5273144028910093, + "learning_rate": 2.042936182430871e-06, + "loss": 0.9523, + "step": 5637 + }, + { + "epoch": 0.5084547053253371, + "grad_norm": 1.7950222248837306, + "learning_rate": 2.0423521046198206e-06, + "loss": 0.9961, + "step": 5638 + }, + { + "epoch": 0.5085448888488073, + "grad_norm": 1.5974342546381166, + "learning_rate": 2.041768023195056e-06, + "loss": 0.9086, + "step": 5639 + }, + { + "epoch": 0.5086350723722776, + "grad_norm": 1.4763176124587267, + "learning_rate": 2.0411839382064126e-06, + "loss": 1.0586, + "step": 5640 + }, + { + "epoch": 0.5087252558957478, + "grad_norm": 1.5547373394669166, + "learning_rate": 2.040599849703729e-06, + "loss": 0.884, + "step": 5641 + }, + { + "epoch": 0.5088154394192181, + "grad_norm": 1.9984127541233556, + "learning_rate": 2.040015757736843e-06, + "loss": 0.9255, + "step": 5642 + }, + { + "epoch": 0.5089056229426884, + "grad_norm": 1.672566279878483, + "learning_rate": 2.039431662355591e-06, + "loss": 0.9169, + "step": 5643 + }, + { + "epoch": 0.5089958064661586, + "grad_norm": 1.6669862520260836, + "learning_rate": 2.0388475636098126e-06, + "loss": 1.0056, + "step": 5644 + }, + { + "epoch": 0.5090859899896288, + "grad_norm": 1.3901561418147184, + "learning_rate": 2.038263461549346e-06, + "loss": 1.0239, + "step": 5645 + }, + { + "epoch": 0.5091761735130992, + "grad_norm": 1.968413490721889, + "learning_rate": 2.0376793562240297e-06, + "loss": 0.9962, + "step": 5646 + }, + { + "epoch": 0.5092663570365694, + "grad_norm": 1.514885790961075, + "learning_rate": 2.037095247683703e-06, + "loss": 1.0018, + "step": 5647 + }, + { + "epoch": 0.5093565405600397, + "grad_norm": 2.1048638601972813, + "learning_rate": 2.0365111359782046e-06, + "loss": 0.9444, + "step": 5648 + }, + { + "epoch": 0.50944672408351, + "grad_norm": 1.433516747176263, + "learning_rate": 2.0359270211573757e-06, + "loss": 1.0212, + "step": 5649 + }, + { + "epoch": 0.5095369076069802, + "grad_norm": 1.9324560035543417, + "learning_rate": 2.0353429032710545e-06, + "loss": 1.0517, + "step": 5650 + }, + { + "epoch": 0.5096270911304505, + "grad_norm": 1.5077887814884814, + "learning_rate": 2.0347587823690825e-06, + "loss": 0.9694, + "step": 5651 + }, + { + "epoch": 0.5097172746539207, + "grad_norm": 1.7580438758313583, + "learning_rate": 2.034174658501299e-06, + "loss": 0.9895, + "step": 5652 + }, + { + "epoch": 0.509807458177391, + "grad_norm": 1.4500230886002314, + "learning_rate": 2.0335905317175453e-06, + "loss": 1.0251, + "step": 5653 + }, + { + "epoch": 0.5098976417008613, + "grad_norm": 2.3374173889249708, + "learning_rate": 2.033006402067663e-06, + "loss": 0.9935, + "step": 5654 + }, + { + "epoch": 0.5099878252243315, + "grad_norm": 1.5242896411077533, + "learning_rate": 2.0324222696014912e-06, + "loss": 0.8937, + "step": 5655 + }, + { + "epoch": 0.5100780087478017, + "grad_norm": 1.6368245748390378, + "learning_rate": 2.0318381343688733e-06, + "loss": 1.0257, + "step": 5656 + }, + { + "epoch": 0.5101681922712721, + "grad_norm": 1.573726984269335, + "learning_rate": 2.0312539964196505e-06, + "loss": 0.972, + "step": 5657 + }, + { + "epoch": 0.5102583757947423, + "grad_norm": 1.5252337698692728, + "learning_rate": 2.030669855803664e-06, + "loss": 0.9701, + "step": 5658 + }, + { + "epoch": 0.5103485593182125, + "grad_norm": 1.6329377778831586, + "learning_rate": 2.0300857125707563e-06, + "loss": 1.041, + "step": 5659 + }, + { + "epoch": 0.5104387428416828, + "grad_norm": 2.063741021601776, + "learning_rate": 2.0295015667707697e-06, + "loss": 1.0192, + "step": 5660 + }, + { + "epoch": 0.5105289263651531, + "grad_norm": 1.7754239959434208, + "learning_rate": 2.0289174184535472e-06, + "loss": 1.0009, + "step": 5661 + }, + { + "epoch": 0.5106191098886234, + "grad_norm": 2.332609223223865, + "learning_rate": 2.02833326766893e-06, + "loss": 0.9675, + "step": 5662 + }, + { + "epoch": 0.5107092934120936, + "grad_norm": 1.6943284092276916, + "learning_rate": 2.027749114466763e-06, + "loss": 1.0567, + "step": 5663 + }, + { + "epoch": 0.5107994769355638, + "grad_norm": 1.6766383684524504, + "learning_rate": 2.027164958896889e-06, + "loss": 1.0378, + "step": 5664 + }, + { + "epoch": 0.5108896604590342, + "grad_norm": 1.9795736061820184, + "learning_rate": 2.02658080100915e-06, + "loss": 1.1088, + "step": 5665 + }, + { + "epoch": 0.5109798439825044, + "grad_norm": 1.8701519915576648, + "learning_rate": 2.0259966408533915e-06, + "loss": 1.0661, + "step": 5666 + }, + { + "epoch": 0.5110700275059746, + "grad_norm": 1.775116583528129, + "learning_rate": 2.025412478479455e-06, + "loss": 1.0785, + "step": 5667 + }, + { + "epoch": 0.5111602110294449, + "grad_norm": 1.7217881833325346, + "learning_rate": 2.0248283139371862e-06, + "loss": 1.0336, + "step": 5668 + }, + { + "epoch": 0.5112503945529152, + "grad_norm": 1.7391071543482937, + "learning_rate": 2.024244147276429e-06, + "loss": 0.9642, + "step": 5669 + }, + { + "epoch": 0.5113405780763854, + "grad_norm": 1.306034101321398, + "learning_rate": 2.023659978547027e-06, + "loss": 1.0589, + "step": 5670 + }, + { + "epoch": 0.5114307615998557, + "grad_norm": 1.4787524095452766, + "learning_rate": 2.023075807798826e-06, + "loss": 0.9917, + "step": 5671 + }, + { + "epoch": 0.511520945123326, + "grad_norm": 2.165097756766701, + "learning_rate": 2.0224916350816696e-06, + "loss": 0.9164, + "step": 5672 + }, + { + "epoch": 0.5116111286467963, + "grad_norm": 1.588367948778682, + "learning_rate": 2.0219074604454026e-06, + "loss": 1.0148, + "step": 5673 + }, + { + "epoch": 0.5117013121702665, + "grad_norm": 21.04942318812405, + "learning_rate": 2.02132328393987e-06, + "loss": 0.9931, + "step": 5674 + }, + { + "epoch": 0.5117914956937367, + "grad_norm": 1.6743542835738192, + "learning_rate": 2.0207391056149174e-06, + "loss": 0.9448, + "step": 5675 + }, + { + "epoch": 0.5118816792172071, + "grad_norm": 1.78524920265384, + "learning_rate": 2.020154925520391e-06, + "loss": 0.958, + "step": 5676 + }, + { + "epoch": 0.5119718627406773, + "grad_norm": 1.5762567320036314, + "learning_rate": 2.0195707437061332e-06, + "loss": 1.0885, + "step": 5677 + }, + { + "epoch": 0.5120620462641475, + "grad_norm": 2.0674013795591026, + "learning_rate": 2.0189865602219934e-06, + "loss": 0.987, + "step": 5678 + }, + { + "epoch": 0.5121522297876178, + "grad_norm": 1.610397014358186, + "learning_rate": 2.0184023751178154e-06, + "loss": 1.0634, + "step": 5679 + }, + { + "epoch": 0.5122424133110881, + "grad_norm": 1.340977536601467, + "learning_rate": 2.017818188443444e-06, + "loss": 0.9406, + "step": 5680 + }, + { + "epoch": 0.5123325968345583, + "grad_norm": 1.7570823678959393, + "learning_rate": 2.017234000248728e-06, + "loss": 1.0049, + "step": 5681 + }, + { + "epoch": 0.5124227803580286, + "grad_norm": 1.6952156004043404, + "learning_rate": 2.0166498105835108e-06, + "loss": 0.9301, + "step": 5682 + }, + { + "epoch": 0.5125129638814988, + "grad_norm": 1.7550101088099597, + "learning_rate": 2.0160656194976407e-06, + "loss": 0.9291, + "step": 5683 + }, + { + "epoch": 0.5126031474049692, + "grad_norm": 1.6225009922912104, + "learning_rate": 2.0154814270409634e-06, + "loss": 0.9442, + "step": 5684 + }, + { + "epoch": 0.5126933309284394, + "grad_norm": 1.8783972163649278, + "learning_rate": 2.0148972332633247e-06, + "loss": 0.9568, + "step": 5685 + }, + { + "epoch": 0.5127835144519096, + "grad_norm": 1.8352680040730542, + "learning_rate": 2.0143130382145733e-06, + "loss": 1.0518, + "step": 5686 + }, + { + "epoch": 0.5128736979753798, + "grad_norm": 1.8522972870159624, + "learning_rate": 2.0137288419445533e-06, + "loss": 0.9642, + "step": 5687 + }, + { + "epoch": 0.5129638814988502, + "grad_norm": 1.6626162266079267, + "learning_rate": 2.0131446445031134e-06, + "loss": 1.0367, + "step": 5688 + }, + { + "epoch": 0.5130540650223204, + "grad_norm": 1.8667628870869553, + "learning_rate": 2.0125604459400994e-06, + "loss": 1.0749, + "step": 5689 + }, + { + "epoch": 0.5131442485457907, + "grad_norm": 2.2580058401148833, + "learning_rate": 2.0119762463053596e-06, + "loss": 1.0501, + "step": 5690 + }, + { + "epoch": 0.5132344320692609, + "grad_norm": 2.4672935148996302, + "learning_rate": 2.0113920456487406e-06, + "loss": 0.9819, + "step": 5691 + }, + { + "epoch": 0.5133246155927312, + "grad_norm": 1.9100016587934232, + "learning_rate": 2.010807844020088e-06, + "loss": 0.988, + "step": 5692 + }, + { + "epoch": 0.5134147991162015, + "grad_norm": 2.190882438301853, + "learning_rate": 2.0102236414692524e-06, + "loss": 1.0469, + "step": 5693 + }, + { + "epoch": 0.5135049826396717, + "grad_norm": 1.4445028150616208, + "learning_rate": 2.0096394380460777e-06, + "loss": 0.9828, + "step": 5694 + }, + { + "epoch": 0.513595166163142, + "grad_norm": 1.5620940634799108, + "learning_rate": 2.0090552338004136e-06, + "loss": 1.0244, + "step": 5695 + }, + { + "epoch": 0.5136853496866123, + "grad_norm": 1.510053955023634, + "learning_rate": 2.0084710287821077e-06, + "loss": 0.9898, + "step": 5696 + }, + { + "epoch": 0.5137755332100825, + "grad_norm": 1.6414685215274545, + "learning_rate": 2.007886823041006e-06, + "loss": 1.0309, + "step": 5697 + }, + { + "epoch": 0.5138657167335527, + "grad_norm": 1.95406343421932, + "learning_rate": 2.0073026166269577e-06, + "loss": 1.0111, + "step": 5698 + }, + { + "epoch": 0.5139559002570231, + "grad_norm": 1.3469573046679875, + "learning_rate": 2.0067184095898093e-06, + "loss": 1.0523, + "step": 5699 + }, + { + "epoch": 0.5140460837804933, + "grad_norm": 1.597518056848049, + "learning_rate": 2.0061342019794094e-06, + "loss": 1.0046, + "step": 5700 + }, + { + "epoch": 0.5141362673039636, + "grad_norm": 1.9455063198617342, + "learning_rate": 2.0055499938456058e-06, + "loss": 1.0722, + "step": 5701 + }, + { + "epoch": 0.5142264508274338, + "grad_norm": 1.8218809255304045, + "learning_rate": 2.0049657852382464e-06, + "loss": 0.9582, + "step": 5702 + }, + { + "epoch": 0.5143166343509041, + "grad_norm": 2.079818953007405, + "learning_rate": 2.0043815762071782e-06, + "loss": 0.9233, + "step": 5703 + }, + { + "epoch": 0.5144068178743744, + "grad_norm": 2.0111396029979076, + "learning_rate": 2.0037973668022492e-06, + "loss": 0.9609, + "step": 5704 + }, + { + "epoch": 0.5144970013978446, + "grad_norm": 1.6209379190209434, + "learning_rate": 2.003213157073309e-06, + "loss": 0.8738, + "step": 5705 + }, + { + "epoch": 0.5145871849213148, + "grad_norm": 1.7811988689378753, + "learning_rate": 2.002628947070204e-06, + "loss": 0.9889, + "step": 5706 + }, + { + "epoch": 0.5146773684447852, + "grad_norm": 1.9677650773318065, + "learning_rate": 2.002044736842783e-06, + "loss": 1.0552, + "step": 5707 + }, + { + "epoch": 0.5147675519682554, + "grad_norm": 1.4976524420636579, + "learning_rate": 2.001460526440894e-06, + "loss": 0.9749, + "step": 5708 + }, + { + "epoch": 0.5148577354917256, + "grad_norm": 0.6282203677832657, + "learning_rate": 2.0008763159143843e-06, + "loss": 0.8181, + "step": 5709 + }, + { + "epoch": 0.5149479190151959, + "grad_norm": 2.2201986821938995, + "learning_rate": 2.000292105313103e-06, + "loss": 1.0198, + "step": 5710 + }, + { + "epoch": 0.5150381025386662, + "grad_norm": 1.6912164383382577, + "learning_rate": 1.999707894686897e-06, + "loss": 1.0308, + "step": 5711 + }, + { + "epoch": 0.5151282860621365, + "grad_norm": 1.537796739304849, + "learning_rate": 1.9991236840856155e-06, + "loss": 0.982, + "step": 5712 + }, + { + "epoch": 0.5152184695856067, + "grad_norm": 1.5755651882853408, + "learning_rate": 1.9985394735591065e-06, + "loss": 0.9669, + "step": 5713 + }, + { + "epoch": 0.5153086531090769, + "grad_norm": 1.6549960733926066, + "learning_rate": 1.997955263157217e-06, + "loss": 1.0348, + "step": 5714 + }, + { + "epoch": 0.5153988366325473, + "grad_norm": 1.604052064773621, + "learning_rate": 1.997371052929796e-06, + "loss": 1.0278, + "step": 5715 + }, + { + "epoch": 0.5154890201560175, + "grad_norm": 1.7157449982581479, + "learning_rate": 1.996786842926691e-06, + "loss": 0.9505, + "step": 5716 + }, + { + "epoch": 0.5155792036794877, + "grad_norm": 1.5227881979691147, + "learning_rate": 1.9962026331977506e-06, + "loss": 0.999, + "step": 5717 + }, + { + "epoch": 0.5156693872029581, + "grad_norm": 1.7551598773308559, + "learning_rate": 1.9956184237928224e-06, + "loss": 1.0076, + "step": 5718 + }, + { + "epoch": 0.5157595707264283, + "grad_norm": 1.8146903516334134, + "learning_rate": 1.995034214761754e-06, + "loss": 0.9824, + "step": 5719 + }, + { + "epoch": 0.5158497542498985, + "grad_norm": 1.4995804835818256, + "learning_rate": 1.9944500061543945e-06, + "loss": 0.9685, + "step": 5720 + }, + { + "epoch": 0.5159399377733688, + "grad_norm": 1.7934819569505644, + "learning_rate": 1.99386579802059e-06, + "loss": 1.0166, + "step": 5721 + }, + { + "epoch": 0.5160301212968391, + "grad_norm": 1.604511356065124, + "learning_rate": 1.993281590410191e-06, + "loss": 0.9714, + "step": 5722 + }, + { + "epoch": 0.5161203048203094, + "grad_norm": 1.3875272301416144, + "learning_rate": 1.992697383373043e-06, + "loss": 0.9419, + "step": 5723 + }, + { + "epoch": 0.5162104883437796, + "grad_norm": 0.7777264673663516, + "learning_rate": 1.9921131769589937e-06, + "loss": 0.8589, + "step": 5724 + }, + { + "epoch": 0.5163006718672498, + "grad_norm": 1.491691946458337, + "learning_rate": 1.991528971217893e-06, + "loss": 0.991, + "step": 5725 + }, + { + "epoch": 0.5163908553907202, + "grad_norm": 1.7739539381754157, + "learning_rate": 1.9909447661995858e-06, + "loss": 0.9872, + "step": 5726 + }, + { + "epoch": 0.5164810389141904, + "grad_norm": 1.804286697600146, + "learning_rate": 1.990360561953922e-06, + "loss": 1.0061, + "step": 5727 + }, + { + "epoch": 0.5165712224376606, + "grad_norm": 1.4451948788182345, + "learning_rate": 1.9897763585307483e-06, + "loss": 1.0177, + "step": 5728 + }, + { + "epoch": 0.5166614059611309, + "grad_norm": 1.641351666058116, + "learning_rate": 1.989192155979912e-06, + "loss": 1.0366, + "step": 5729 + }, + { + "epoch": 0.5167515894846012, + "grad_norm": 1.3920689918778768, + "learning_rate": 1.98860795435126e-06, + "loss": 1.0302, + "step": 5730 + }, + { + "epoch": 0.5168417730080714, + "grad_norm": 1.7032959178590097, + "learning_rate": 1.9880237536946406e-06, + "loss": 1.0022, + "step": 5731 + }, + { + "epoch": 0.5169319565315417, + "grad_norm": 3.055756754945974, + "learning_rate": 1.987439554059901e-06, + "loss": 0.9612, + "step": 5732 + }, + { + "epoch": 0.5170221400550119, + "grad_norm": 1.5831920744808876, + "learning_rate": 1.9868553554968864e-06, + "loss": 0.9703, + "step": 5733 + }, + { + "epoch": 0.5171123235784822, + "grad_norm": 1.9251681390628508, + "learning_rate": 1.986271158055447e-06, + "loss": 0.9411, + "step": 5734 + }, + { + "epoch": 0.5172025071019525, + "grad_norm": 1.64063539047584, + "learning_rate": 1.9856869617854273e-06, + "loss": 0.9883, + "step": 5735 + }, + { + "epoch": 0.5172926906254227, + "grad_norm": 1.5257957789856846, + "learning_rate": 1.9851027667366746e-06, + "loss": 0.9709, + "step": 5736 + }, + { + "epoch": 0.517382874148893, + "grad_norm": 1.5923819560761372, + "learning_rate": 1.984518572959037e-06, + "loss": 1.0902, + "step": 5737 + }, + { + "epoch": 0.5174730576723633, + "grad_norm": 1.8908962969372154, + "learning_rate": 1.9839343805023587e-06, + "loss": 0.9147, + "step": 5738 + }, + { + "epoch": 0.5175632411958335, + "grad_norm": 2.0117633110962747, + "learning_rate": 1.9833501894164886e-06, + "loss": 1.085, + "step": 5739 + }, + { + "epoch": 0.5176534247193038, + "grad_norm": 1.5819159571704615, + "learning_rate": 1.982765999751273e-06, + "loss": 1.0271, + "step": 5740 + }, + { + "epoch": 0.5177436082427741, + "grad_norm": 2.1556766134637906, + "learning_rate": 1.9821818115565553e-06, + "loss": 1.0248, + "step": 5741 + }, + { + "epoch": 0.5178337917662443, + "grad_norm": 1.6014384151864498, + "learning_rate": 1.9815976248821853e-06, + "loss": 0.9005, + "step": 5742 + }, + { + "epoch": 0.5179239752897146, + "grad_norm": 4.187861868182562, + "learning_rate": 1.981013439778007e-06, + "loss": 1.0045, + "step": 5743 + }, + { + "epoch": 0.5180141588131848, + "grad_norm": 1.6859109248911939, + "learning_rate": 1.9804292562938666e-06, + "loss": 0.999, + "step": 5744 + }, + { + "epoch": 0.5181043423366551, + "grad_norm": 1.696820831839287, + "learning_rate": 1.97984507447961e-06, + "loss": 1.0251, + "step": 5745 + }, + { + "epoch": 0.5181945258601254, + "grad_norm": 1.6272932157393931, + "learning_rate": 1.9792608943850824e-06, + "loss": 0.9331, + "step": 5746 + }, + { + "epoch": 0.5182847093835956, + "grad_norm": 1.568534489756805, + "learning_rate": 1.9786767160601305e-06, + "loss": 0.9697, + "step": 5747 + }, + { + "epoch": 0.5183748929070658, + "grad_norm": 1.6932786968507991, + "learning_rate": 1.9780925395545977e-06, + "loss": 0.9261, + "step": 5748 + }, + { + "epoch": 0.5184650764305362, + "grad_norm": 0.6224734738071711, + "learning_rate": 1.9775083649183306e-06, + "loss": 0.8038, + "step": 5749 + }, + { + "epoch": 0.5185552599540064, + "grad_norm": 1.867100478183468, + "learning_rate": 1.976924192201174e-06, + "loss": 1.0328, + "step": 5750 + }, + { + "epoch": 0.5186454434774767, + "grad_norm": 1.529833780995895, + "learning_rate": 1.9763400214529723e-06, + "loss": 0.9256, + "step": 5751 + }, + { + "epoch": 0.5187356270009469, + "grad_norm": 1.606786108536863, + "learning_rate": 1.9757558527235713e-06, + "loss": 0.8639, + "step": 5752 + }, + { + "epoch": 0.5188258105244172, + "grad_norm": 1.5951114897791439, + "learning_rate": 1.9751716860628136e-06, + "loss": 0.9829, + "step": 5753 + }, + { + "epoch": 0.5189159940478875, + "grad_norm": 1.7033715025973035, + "learning_rate": 1.974587521520545e-06, + "loss": 1.0187, + "step": 5754 + }, + { + "epoch": 0.5190061775713577, + "grad_norm": 1.5472417069233055, + "learning_rate": 1.9740033591466088e-06, + "loss": 1.0278, + "step": 5755 + }, + { + "epoch": 0.5190963610948279, + "grad_norm": 1.486500713739567, + "learning_rate": 1.97341919899085e-06, + "loss": 1.1063, + "step": 5756 + }, + { + "epoch": 0.5191865446182983, + "grad_norm": 1.6133065210168882, + "learning_rate": 1.9728350411031114e-06, + "loss": 1.0196, + "step": 5757 + }, + { + "epoch": 0.5192767281417685, + "grad_norm": 2.1550059392364864, + "learning_rate": 1.9722508855332367e-06, + "loss": 1.0661, + "step": 5758 + }, + { + "epoch": 0.5193669116652387, + "grad_norm": 1.3384076309879238, + "learning_rate": 1.97166673233107e-06, + "loss": 0.9256, + "step": 5759 + }, + { + "epoch": 0.519457095188709, + "grad_norm": 1.5899037342799545, + "learning_rate": 1.971082581546453e-06, + "loss": 1.037, + "step": 5760 + }, + { + "epoch": 0.5195472787121793, + "grad_norm": 3.117436372769947, + "learning_rate": 1.9704984332292306e-06, + "loss": 1.0205, + "step": 5761 + }, + { + "epoch": 0.5196374622356495, + "grad_norm": 1.599489535918474, + "learning_rate": 1.9699142874292444e-06, + "loss": 1.0826, + "step": 5762 + }, + { + "epoch": 0.5197276457591198, + "grad_norm": 1.9091106899896324, + "learning_rate": 1.969330144196336e-06, + "loss": 1.0807, + "step": 5763 + }, + { + "epoch": 0.51981782928259, + "grad_norm": 2.0074133569017443, + "learning_rate": 1.9687460035803497e-06, + "loss": 0.9347, + "step": 5764 + }, + { + "epoch": 0.5199080128060604, + "grad_norm": 1.792915227086988, + "learning_rate": 1.9681618656311265e-06, + "loss": 0.9752, + "step": 5765 + }, + { + "epoch": 0.5199981963295306, + "grad_norm": 1.3100187826630085, + "learning_rate": 1.9675777303985086e-06, + "loss": 0.9037, + "step": 5766 + }, + { + "epoch": 0.5200883798530008, + "grad_norm": 1.5855557588308185, + "learning_rate": 1.9669935979323376e-06, + "loss": 1.0231, + "step": 5767 + }, + { + "epoch": 0.5201785633764712, + "grad_norm": 2.2601642965838766, + "learning_rate": 1.9664094682824545e-06, + "loss": 1.052, + "step": 5768 + }, + { + "epoch": 0.5202687468999414, + "grad_norm": 1.4421544394624843, + "learning_rate": 1.965825341498701e-06, + "loss": 0.9558, + "step": 5769 + }, + { + "epoch": 0.5203589304234116, + "grad_norm": 1.7915944003531008, + "learning_rate": 1.9652412176309177e-06, + "loss": 1.0684, + "step": 5770 + }, + { + "epoch": 0.5204491139468819, + "grad_norm": 1.6562745434364055, + "learning_rate": 1.9646570967289453e-06, + "loss": 0.8593, + "step": 5771 + }, + { + "epoch": 0.5205392974703522, + "grad_norm": 1.6151072072682509, + "learning_rate": 1.9640729788426246e-06, + "loss": 1.0323, + "step": 5772 + }, + { + "epoch": 0.5206294809938224, + "grad_norm": 1.4959373292749554, + "learning_rate": 1.963488864021795e-06, + "loss": 1.026, + "step": 5773 + }, + { + "epoch": 0.5207196645172927, + "grad_norm": 1.4787494267956742, + "learning_rate": 1.962904752316298e-06, + "loss": 0.8868, + "step": 5774 + }, + { + "epoch": 0.5208098480407629, + "grad_norm": 5.035075184810969, + "learning_rate": 1.9623206437759706e-06, + "loss": 0.9208, + "step": 5775 + }, + { + "epoch": 0.5209000315642333, + "grad_norm": 1.6173190740605103, + "learning_rate": 1.9617365384506545e-06, + "loss": 0.9534, + "step": 5776 + }, + { + "epoch": 0.5209902150877035, + "grad_norm": 1.2997173277102905, + "learning_rate": 1.9611524363901872e-06, + "loss": 0.9809, + "step": 5777 + }, + { + "epoch": 0.5210803986111737, + "grad_norm": 1.555810417383843, + "learning_rate": 1.960568337644409e-06, + "loss": 0.9855, + "step": 5778 + }, + { + "epoch": 0.521170582134644, + "grad_norm": 1.5084740170930193, + "learning_rate": 1.9599842422631576e-06, + "loss": 0.9744, + "step": 5779 + }, + { + "epoch": 0.5212607656581143, + "grad_norm": 2.5446769731189796, + "learning_rate": 1.9594001502962703e-06, + "loss": 0.9835, + "step": 5780 + }, + { + "epoch": 0.5213509491815845, + "grad_norm": 1.6591686228203015, + "learning_rate": 1.9588160617935868e-06, + "loss": 1.0533, + "step": 5781 + }, + { + "epoch": 0.5214411327050548, + "grad_norm": 1.79819930659287, + "learning_rate": 1.958231976804944e-06, + "loss": 1.0791, + "step": 5782 + }, + { + "epoch": 0.521531316228525, + "grad_norm": 3.677864287153848, + "learning_rate": 1.957647895380179e-06, + "loss": 0.9156, + "step": 5783 + }, + { + "epoch": 0.5216214997519953, + "grad_norm": 1.476430089755869, + "learning_rate": 1.9570638175691297e-06, + "loss": 0.9229, + "step": 5784 + }, + { + "epoch": 0.5217116832754656, + "grad_norm": 1.943163377855851, + "learning_rate": 1.956479743421632e-06, + "loss": 1.0343, + "step": 5785 + }, + { + "epoch": 0.5218018667989358, + "grad_norm": 1.777847855729834, + "learning_rate": 1.955895672987522e-06, + "loss": 0.9303, + "step": 5786 + }, + { + "epoch": 0.521892050322406, + "grad_norm": 1.4054572520016322, + "learning_rate": 1.9553116063166367e-06, + "loss": 0.8982, + "step": 5787 + }, + { + "epoch": 0.5219822338458764, + "grad_norm": 1.4663265606989064, + "learning_rate": 1.954727543458812e-06, + "loss": 0.8951, + "step": 5788 + }, + { + "epoch": 0.5220724173693466, + "grad_norm": 1.6190602980913313, + "learning_rate": 1.954143484463883e-06, + "loss": 1.0244, + "step": 5789 + }, + { + "epoch": 0.5221626008928169, + "grad_norm": 2.212416406981208, + "learning_rate": 1.9535594293816836e-06, + "loss": 0.9773, + "step": 5790 + }, + { + "epoch": 0.5222527844162872, + "grad_norm": 1.936025119673701, + "learning_rate": 1.952975378262051e-06, + "loss": 0.9382, + "step": 5791 + }, + { + "epoch": 0.5223429679397574, + "grad_norm": 1.9952044213522078, + "learning_rate": 1.952391331154817e-06, + "loss": 0.9428, + "step": 5792 + }, + { + "epoch": 0.5224331514632277, + "grad_norm": 1.6036423738778258, + "learning_rate": 1.9518072881098185e-06, + "loss": 1.0458, + "step": 5793 + }, + { + "epoch": 0.5225233349866979, + "grad_norm": 3.7984741359812824, + "learning_rate": 1.9512232491768867e-06, + "loss": 0.9811, + "step": 5794 + }, + { + "epoch": 0.5226135185101682, + "grad_norm": 1.518775261562914, + "learning_rate": 1.9506392144058573e-06, + "loss": 0.9812, + "step": 5795 + }, + { + "epoch": 0.5227037020336385, + "grad_norm": 1.7083446765926245, + "learning_rate": 1.9500551838465623e-06, + "loss": 0.9864, + "step": 5796 + }, + { + "epoch": 0.5227938855571087, + "grad_norm": 1.3351218873719768, + "learning_rate": 1.9494711575488337e-06, + "loss": 1.004, + "step": 5797 + }, + { + "epoch": 0.5228840690805789, + "grad_norm": 2.3252627808748993, + "learning_rate": 1.948887135562505e-06, + "loss": 0.9367, + "step": 5798 + }, + { + "epoch": 0.5229742526040493, + "grad_norm": 1.5292501820028939, + "learning_rate": 1.9483031179374074e-06, + "loss": 1.04, + "step": 5799 + }, + { + "epoch": 0.5230644361275195, + "grad_norm": 1.4641230323352943, + "learning_rate": 1.9477191047233736e-06, + "loss": 1.0475, + "step": 5800 + }, + { + "epoch": 0.5231546196509897, + "grad_norm": 1.6484826972152873, + "learning_rate": 1.9471350959702334e-06, + "loss": 0.8694, + "step": 5801 + }, + { + "epoch": 0.52324480317446, + "grad_norm": 1.904397971575273, + "learning_rate": 1.9465510917278184e-06, + "loss": 0.9933, + "step": 5802 + }, + { + "epoch": 0.5233349866979303, + "grad_norm": 1.9174493836209, + "learning_rate": 1.9459670920459593e-06, + "loss": 0.9905, + "step": 5803 + }, + { + "epoch": 0.5234251702214006, + "grad_norm": 1.6735390198702993, + "learning_rate": 1.945383096974485e-06, + "loss": 0.9071, + "step": 5804 + }, + { + "epoch": 0.5235153537448708, + "grad_norm": 4.809673717955653, + "learning_rate": 1.944799106563227e-06, + "loss": 0.9141, + "step": 5805 + }, + { + "epoch": 0.523605537268341, + "grad_norm": 1.6921908560849035, + "learning_rate": 1.9442151208620133e-06, + "loss": 1.0289, + "step": 5806 + }, + { + "epoch": 0.5236957207918114, + "grad_norm": 1.9386502204960026, + "learning_rate": 1.943631139920672e-06, + "loss": 1.0352, + "step": 5807 + }, + { + "epoch": 0.5237859043152816, + "grad_norm": 1.4574333229320806, + "learning_rate": 1.943047163789034e-06, + "loss": 0.9435, + "step": 5808 + }, + { + "epoch": 0.5238760878387518, + "grad_norm": 1.4014491891295646, + "learning_rate": 1.942463192516925e-06, + "loss": 0.9486, + "step": 5809 + }, + { + "epoch": 0.5239662713622221, + "grad_norm": 1.4193899215525334, + "learning_rate": 1.9418792261541746e-06, + "loss": 1.0063, + "step": 5810 + }, + { + "epoch": 0.5240564548856924, + "grad_norm": 1.6813521542056638, + "learning_rate": 1.9412952647506094e-06, + "loss": 0.9493, + "step": 5811 + }, + { + "epoch": 0.5241466384091626, + "grad_norm": 1.4871608090956356, + "learning_rate": 1.9407113083560552e-06, + "loss": 0.9629, + "step": 5812 + }, + { + "epoch": 0.5242368219326329, + "grad_norm": 1.789568433895711, + "learning_rate": 1.940127357020339e-06, + "loss": 0.9661, + "step": 5813 + }, + { + "epoch": 0.5243270054561032, + "grad_norm": 1.5704418527647048, + "learning_rate": 1.939543410793287e-06, + "loss": 0.9981, + "step": 5814 + }, + { + "epoch": 0.5244171889795735, + "grad_norm": 1.9393457419487055, + "learning_rate": 1.9389594697247246e-06, + "loss": 1.0245, + "step": 5815 + }, + { + "epoch": 0.5245073725030437, + "grad_norm": 1.9545345255685314, + "learning_rate": 1.9383755338644763e-06, + "loss": 1.064, + "step": 5816 + }, + { + "epoch": 0.5245975560265139, + "grad_norm": 2.6677253131755556, + "learning_rate": 1.937791603262368e-06, + "loss": 1.0179, + "step": 5817 + }, + { + "epoch": 0.5246877395499843, + "grad_norm": 1.8826339447070382, + "learning_rate": 1.9372076779682235e-06, + "loss": 0.9422, + "step": 5818 + }, + { + "epoch": 0.5247779230734545, + "grad_norm": 1.9480649174324576, + "learning_rate": 1.9366237580318648e-06, + "loss": 0.9396, + "step": 5819 + }, + { + "epoch": 0.5248681065969247, + "grad_norm": 1.4455178604632366, + "learning_rate": 1.9360398435031176e-06, + "loss": 0.9951, + "step": 5820 + }, + { + "epoch": 0.524958290120395, + "grad_norm": 1.5779876649226994, + "learning_rate": 1.9354559344318025e-06, + "loss": 0.9939, + "step": 5821 + }, + { + "epoch": 0.5250484736438653, + "grad_norm": 1.7437483291891427, + "learning_rate": 1.934872030867744e-06, + "loss": 1.0415, + "step": 5822 + }, + { + "epoch": 0.5251386571673355, + "grad_norm": 1.8135428223961383, + "learning_rate": 1.934288132860763e-06, + "loss": 1.0008, + "step": 5823 + }, + { + "epoch": 0.5252288406908058, + "grad_norm": 1.8208583291765563, + "learning_rate": 1.93370424046068e-06, + "loss": 0.9829, + "step": 5824 + }, + { + "epoch": 0.525319024214276, + "grad_norm": 1.6445249435050173, + "learning_rate": 1.9331203537173177e-06, + "loss": 1.0524, + "step": 5825 + }, + { + "epoch": 0.5254092077377464, + "grad_norm": 1.2831964361983204, + "learning_rate": 1.9325364726804947e-06, + "loss": 1.1042, + "step": 5826 + }, + { + "epoch": 0.5254993912612166, + "grad_norm": 1.5404553766866687, + "learning_rate": 1.9319525974000327e-06, + "loss": 0.9177, + "step": 5827 + }, + { + "epoch": 0.5255895747846868, + "grad_norm": 2.4270453606992692, + "learning_rate": 1.93136872792575e-06, + "loss": 1.0471, + "step": 5828 + }, + { + "epoch": 0.525679758308157, + "grad_norm": 1.4405769919383513, + "learning_rate": 1.9307848643074653e-06, + "loss": 0.9462, + "step": 5829 + }, + { + "epoch": 0.5257699418316274, + "grad_norm": 1.4714505954527997, + "learning_rate": 1.9302010065949985e-06, + "loss": 1.0216, + "step": 5830 + }, + { + "epoch": 0.5258601253550976, + "grad_norm": 1.570389019824064, + "learning_rate": 1.9296171548381657e-06, + "loss": 0.949, + "step": 5831 + }, + { + "epoch": 0.5259503088785679, + "grad_norm": 1.4147168747683738, + "learning_rate": 1.9290333090867862e-06, + "loss": 0.976, + "step": 5832 + }, + { + "epoch": 0.5260404924020381, + "grad_norm": 1.6361195052226085, + "learning_rate": 1.928449469390676e-06, + "loss": 0.992, + "step": 5833 + }, + { + "epoch": 0.5261306759255084, + "grad_norm": 1.5344766369505023, + "learning_rate": 1.927865635799651e-06, + "loss": 0.9598, + "step": 5834 + }, + { + "epoch": 0.5262208594489787, + "grad_norm": 1.3792004582086808, + "learning_rate": 1.927281808363528e-06, + "loss": 0.9183, + "step": 5835 + }, + { + "epoch": 0.5263110429724489, + "grad_norm": 1.5577273147540096, + "learning_rate": 1.9266979871321216e-06, + "loss": 1.0644, + "step": 5836 + }, + { + "epoch": 0.5264012264959192, + "grad_norm": 1.4770208384103416, + "learning_rate": 1.9261141721552482e-06, + "loss": 0.9411, + "step": 5837 + }, + { + "epoch": 0.5264914100193895, + "grad_norm": 1.8153893690469765, + "learning_rate": 1.9255303634827204e-06, + "loss": 1.0521, + "step": 5838 + }, + { + "epoch": 0.5265815935428597, + "grad_norm": 2.1058454583496635, + "learning_rate": 1.924946561164352e-06, + "loss": 1.0529, + "step": 5839 + }, + { + "epoch": 0.52667177706633, + "grad_norm": 1.6105493501565082, + "learning_rate": 1.9243627652499582e-06, + "loss": 0.9402, + "step": 5840 + }, + { + "epoch": 0.5267619605898003, + "grad_norm": 1.6554258203273395, + "learning_rate": 1.9237789757893493e-06, + "loss": 1.0516, + "step": 5841 + }, + { + "epoch": 0.5268521441132705, + "grad_norm": 1.6755428544782116, + "learning_rate": 1.9231951928323395e-06, + "loss": 1.0318, + "step": 5842 + }, + { + "epoch": 0.5269423276367408, + "grad_norm": 1.7943428442188765, + "learning_rate": 1.922611416428738e-06, + "loss": 0.9405, + "step": 5843 + }, + { + "epoch": 0.527032511160211, + "grad_norm": 1.590589793477024, + "learning_rate": 1.922027646628358e-06, + "loss": 1.0116, + "step": 5844 + }, + { + "epoch": 0.5271226946836813, + "grad_norm": 0.6618769865213324, + "learning_rate": 1.9214438834810092e-06, + "loss": 0.8718, + "step": 5845 + }, + { + "epoch": 0.5272128782071516, + "grad_norm": 2.062556294915619, + "learning_rate": 1.9208601270365008e-06, + "loss": 0.9188, + "step": 5846 + }, + { + "epoch": 0.5273030617306218, + "grad_norm": 3.239061139737187, + "learning_rate": 1.9202763773446435e-06, + "loss": 1.0693, + "step": 5847 + }, + { + "epoch": 0.527393245254092, + "grad_norm": 2.3717837390255063, + "learning_rate": 1.9196926344552444e-06, + "loss": 0.8906, + "step": 5848 + }, + { + "epoch": 0.5274834287775624, + "grad_norm": 1.9863080799212884, + "learning_rate": 1.919108898418113e-06, + "loss": 1.015, + "step": 5849 + }, + { + "epoch": 0.5275736123010326, + "grad_norm": 1.4367626621158995, + "learning_rate": 1.918525169283057e-06, + "loss": 0.8839, + "step": 5850 + }, + { + "epoch": 0.5276637958245028, + "grad_norm": 1.6446864401473063, + "learning_rate": 1.9179414470998817e-06, + "loss": 1.0156, + "step": 5851 + }, + { + "epoch": 0.5277539793479731, + "grad_norm": 3.1986749766810183, + "learning_rate": 1.917357731918395e-06, + "loss": 0.9445, + "step": 5852 + }, + { + "epoch": 0.5278441628714434, + "grad_norm": 2.088081886373923, + "learning_rate": 1.9167740237884025e-06, + "loss": 1.1186, + "step": 5853 + }, + { + "epoch": 0.5279343463949137, + "grad_norm": 1.5767743213157417, + "learning_rate": 1.916190322759709e-06, + "loss": 1.0288, + "step": 5854 + }, + { + "epoch": 0.5280245299183839, + "grad_norm": 1.975716571948393, + "learning_rate": 1.91560662888212e-06, + "loss": 1.0884, + "step": 5855 + }, + { + "epoch": 0.5281147134418541, + "grad_norm": 1.4844825304133482, + "learning_rate": 1.915022942205438e-06, + "loss": 1.0361, + "step": 5856 + }, + { + "epoch": 0.5282048969653245, + "grad_norm": 3.1479765978262013, + "learning_rate": 1.914439262779468e-06, + "loss": 0.9806, + "step": 5857 + }, + { + "epoch": 0.5282950804887947, + "grad_norm": 2.128730304691866, + "learning_rate": 1.9138555906540103e-06, + "loss": 0.9275, + "step": 5858 + }, + { + "epoch": 0.5283852640122649, + "grad_norm": 1.8979358402739506, + "learning_rate": 1.91327192587887e-06, + "loss": 0.8975, + "step": 5859 + }, + { + "epoch": 0.5284754475357353, + "grad_norm": 2.259017992107582, + "learning_rate": 1.912688268503846e-06, + "loss": 0.9649, + "step": 5860 + }, + { + "epoch": 0.5285656310592055, + "grad_norm": 1.6987327143650721, + "learning_rate": 1.912104618578741e-06, + "loss": 0.9557, + "step": 5861 + }, + { + "epoch": 0.5286558145826757, + "grad_norm": 2.385421075497569, + "learning_rate": 1.9115209761533554e-06, + "loss": 0.9683, + "step": 5862 + }, + { + "epoch": 0.528745998106146, + "grad_norm": 2.0204909385245737, + "learning_rate": 1.9109373412774863e-06, + "loss": 0.9059, + "step": 5863 + }, + { + "epoch": 0.5288361816296163, + "grad_norm": 1.4374702284673941, + "learning_rate": 1.910353714000936e-06, + "loss": 0.9915, + "step": 5864 + }, + { + "epoch": 0.5289263651530866, + "grad_norm": 1.773509356013469, + "learning_rate": 1.9097700943734997e-06, + "loss": 0.9496, + "step": 5865 + }, + { + "epoch": 0.5290165486765568, + "grad_norm": 1.6637797227321098, + "learning_rate": 1.909186482444977e-06, + "loss": 0.8882, + "step": 5866 + }, + { + "epoch": 0.529106732200027, + "grad_norm": 1.6792670145017594, + "learning_rate": 1.9086028782651652e-06, + "loss": 0.8952, + "step": 5867 + }, + { + "epoch": 0.5291969157234974, + "grad_norm": 1.5114271883271864, + "learning_rate": 1.908019281883859e-06, + "loss": 0.9863, + "step": 5868 + }, + { + "epoch": 0.5292870992469676, + "grad_norm": 1.5235607390807677, + "learning_rate": 1.9074356933508545e-06, + "loss": 0.9853, + "step": 5869 + }, + { + "epoch": 0.5293772827704378, + "grad_norm": 2.922952708219013, + "learning_rate": 1.9068521127159477e-06, + "loss": 1.0146, + "step": 5870 + }, + { + "epoch": 0.5294674662939081, + "grad_norm": 1.527684712706178, + "learning_rate": 1.9062685400289322e-06, + "loss": 0.938, + "step": 5871 + }, + { + "epoch": 0.5295576498173784, + "grad_norm": 1.6523251363368694, + "learning_rate": 1.9056849753396018e-06, + "loss": 0.9608, + "step": 5872 + }, + { + "epoch": 0.5296478333408486, + "grad_norm": 2.242453104164598, + "learning_rate": 1.9051014186977485e-06, + "loss": 1.0255, + "step": 5873 + }, + { + "epoch": 0.5297380168643189, + "grad_norm": 1.5567735641555938, + "learning_rate": 1.9045178701531664e-06, + "loss": 0.99, + "step": 5874 + }, + { + "epoch": 0.5298282003877891, + "grad_norm": 1.317028181728926, + "learning_rate": 1.903934329755645e-06, + "loss": 0.982, + "step": 5875 + }, + { + "epoch": 0.5299183839112594, + "grad_norm": 1.5362162340021324, + "learning_rate": 1.9033507975549775e-06, + "loss": 1.0224, + "step": 5876 + }, + { + "epoch": 0.5300085674347297, + "grad_norm": 1.5329839270309649, + "learning_rate": 1.9027672736009525e-06, + "loss": 0.9798, + "step": 5877 + }, + { + "epoch": 0.5300987509581999, + "grad_norm": 1.541560667706941, + "learning_rate": 1.9021837579433593e-06, + "loss": 0.9369, + "step": 5878 + }, + { + "epoch": 0.5301889344816701, + "grad_norm": 1.898084631256014, + "learning_rate": 1.901600250631988e-06, + "loss": 0.9638, + "step": 5879 + }, + { + "epoch": 0.5302791180051405, + "grad_norm": 1.937120892751445, + "learning_rate": 1.901016751716625e-06, + "loss": 0.9744, + "step": 5880 + }, + { + "epoch": 0.5303693015286107, + "grad_norm": 1.7155835895369829, + "learning_rate": 1.9004332612470593e-06, + "loss": 0.9514, + "step": 5881 + }, + { + "epoch": 0.530459485052081, + "grad_norm": 1.1975934573917904, + "learning_rate": 1.8998497792730763e-06, + "loss": 1.0129, + "step": 5882 + }, + { + "epoch": 0.5305496685755512, + "grad_norm": 1.4895044941701505, + "learning_rate": 1.8992663058444629e-06, + "loss": 1.0395, + "step": 5883 + }, + { + "epoch": 0.5306398520990215, + "grad_norm": 1.4454291219313038, + "learning_rate": 1.8986828410110032e-06, + "loss": 1.0231, + "step": 5884 + }, + { + "epoch": 0.5307300356224918, + "grad_norm": 1.9132129262903443, + "learning_rate": 1.8980993848224823e-06, + "loss": 0.98, + "step": 5885 + }, + { + "epoch": 0.530820219145962, + "grad_norm": 1.4587228073292187, + "learning_rate": 1.8975159373286843e-06, + "loss": 0.943, + "step": 5886 + }, + { + "epoch": 0.5309104026694323, + "grad_norm": 2.306393370268387, + "learning_rate": 1.8969324985793904e-06, + "loss": 1.0251, + "step": 5887 + }, + { + "epoch": 0.5310005861929026, + "grad_norm": 1.7415116029671014, + "learning_rate": 1.8963490686243851e-06, + "loss": 0.9812, + "step": 5888 + }, + { + "epoch": 0.5310907697163728, + "grad_norm": 1.7304433299795963, + "learning_rate": 1.8957656475134486e-06, + "loss": 0.9267, + "step": 5889 + }, + { + "epoch": 0.531180953239843, + "grad_norm": 1.6721252539321467, + "learning_rate": 1.895182235296361e-06, + "loss": 0.9674, + "step": 5890 + }, + { + "epoch": 0.5312711367633134, + "grad_norm": 1.4275386483059667, + "learning_rate": 1.8945988320229042e-06, + "loss": 0.91, + "step": 5891 + }, + { + "epoch": 0.5313613202867836, + "grad_norm": 1.5661859143321917, + "learning_rate": 1.8940154377428553e-06, + "loss": 1.069, + "step": 5892 + }, + { + "epoch": 0.5314515038102539, + "grad_norm": 1.4503461457492222, + "learning_rate": 1.8934320525059944e-06, + "loss": 1.0169, + "step": 5893 + }, + { + "epoch": 0.5315416873337241, + "grad_norm": 2.1818554669142083, + "learning_rate": 1.8928486763620984e-06, + "loss": 0.8728, + "step": 5894 + }, + { + "epoch": 0.5316318708571944, + "grad_norm": 1.6227040577192746, + "learning_rate": 1.892265309360943e-06, + "loss": 0.9977, + "step": 5895 + }, + { + "epoch": 0.5317220543806647, + "grad_norm": 1.4438460049766513, + "learning_rate": 1.8916819515523067e-06, + "loss": 0.9739, + "step": 5896 + }, + { + "epoch": 0.5318122379041349, + "grad_norm": 1.376635315925272, + "learning_rate": 1.891098602985963e-06, + "loss": 0.9034, + "step": 5897 + }, + { + "epoch": 0.5319024214276051, + "grad_norm": 1.7138018734215994, + "learning_rate": 1.8905152637116872e-06, + "loss": 1.0118, + "step": 5898 + }, + { + "epoch": 0.5319926049510755, + "grad_norm": 1.8340548338269187, + "learning_rate": 1.8899319337792527e-06, + "loss": 1.0268, + "step": 5899 + }, + { + "epoch": 0.5320827884745457, + "grad_norm": 1.3121539522567132, + "learning_rate": 1.8893486132384325e-06, + "loss": 1.0274, + "step": 5900 + }, + { + "epoch": 0.5321729719980159, + "grad_norm": 2.555407501391695, + "learning_rate": 1.888765302138999e-06, + "loss": 0.9555, + "step": 5901 + }, + { + "epoch": 0.5322631555214862, + "grad_norm": 1.773204099109747, + "learning_rate": 1.8881820005307224e-06, + "loss": 0.9301, + "step": 5902 + }, + { + "epoch": 0.5323533390449565, + "grad_norm": 1.4007048739766466, + "learning_rate": 1.8875987084633748e-06, + "loss": 0.9186, + "step": 5903 + }, + { + "epoch": 0.5324435225684268, + "grad_norm": 2.986136031724506, + "learning_rate": 1.8870154259867246e-06, + "loss": 1.0177, + "step": 5904 + }, + { + "epoch": 0.532533706091897, + "grad_norm": 1.839418268916475, + "learning_rate": 1.886432153150542e-06, + "loss": 0.9251, + "step": 5905 + }, + { + "epoch": 0.5326238896153672, + "grad_norm": 1.7174865934245716, + "learning_rate": 1.8858488900045944e-06, + "loss": 0.9624, + "step": 5906 + }, + { + "epoch": 0.5327140731388376, + "grad_norm": 1.9429076009897863, + "learning_rate": 1.885265636598648e-06, + "loss": 0.9398, + "step": 5907 + }, + { + "epoch": 0.5328042566623078, + "grad_norm": 1.3531635922044907, + "learning_rate": 1.884682392982471e-06, + "loss": 1.0196, + "step": 5908 + }, + { + "epoch": 0.532894440185778, + "grad_norm": 2.520611955574423, + "learning_rate": 1.8840991592058274e-06, + "loss": 1.0411, + "step": 5909 + }, + { + "epoch": 0.5329846237092484, + "grad_norm": 1.4102134798647028, + "learning_rate": 1.8835159353184828e-06, + "loss": 1.0023, + "step": 5910 + }, + { + "epoch": 0.5330748072327186, + "grad_norm": 1.4159966829282626, + "learning_rate": 1.8829327213702013e-06, + "loss": 1.0616, + "step": 5911 + }, + { + "epoch": 0.5331649907561888, + "grad_norm": 1.7542973615415922, + "learning_rate": 1.8823495174107452e-06, + "loss": 1.0603, + "step": 5912 + }, + { + "epoch": 0.5332551742796591, + "grad_norm": 1.9729349373841194, + "learning_rate": 1.8817663234898773e-06, + "loss": 0.9556, + "step": 5913 + }, + { + "epoch": 0.5333453578031294, + "grad_norm": 1.717176098217308, + "learning_rate": 1.881183139657358e-06, + "loss": 0.9946, + "step": 5914 + }, + { + "epoch": 0.5334355413265996, + "grad_norm": 1.4925831375097804, + "learning_rate": 1.8805999659629488e-06, + "loss": 1.0007, + "step": 5915 + }, + { + "epoch": 0.5335257248500699, + "grad_norm": 0.6383461311076681, + "learning_rate": 1.880016802456409e-06, + "loss": 0.7729, + "step": 5916 + }, + { + "epoch": 0.5336159083735401, + "grad_norm": 1.6701485820162056, + "learning_rate": 1.8794336491874964e-06, + "loss": 0.9277, + "step": 5917 + }, + { + "epoch": 0.5337060918970105, + "grad_norm": 1.8255647517067013, + "learning_rate": 1.8788505062059708e-06, + "loss": 0.9464, + "step": 5918 + }, + { + "epoch": 0.5337962754204807, + "grad_norm": 1.7163770073269211, + "learning_rate": 1.8782673735615869e-06, + "loss": 0.9698, + "step": 5919 + }, + { + "epoch": 0.5338864589439509, + "grad_norm": 1.4518013945439745, + "learning_rate": 1.8776842513041026e-06, + "loss": 0.9439, + "step": 5920 + }, + { + "epoch": 0.5339766424674212, + "grad_norm": 1.8873867960954802, + "learning_rate": 1.8771011394832727e-06, + "loss": 0.9526, + "step": 5921 + }, + { + "epoch": 0.5340668259908915, + "grad_norm": 1.7088071197138206, + "learning_rate": 1.8765180381488501e-06, + "loss": 0.9659, + "step": 5922 + }, + { + "epoch": 0.5341570095143617, + "grad_norm": 0.7271136172818359, + "learning_rate": 1.8759349473505905e-06, + "loss": 0.9035, + "step": 5923 + }, + { + "epoch": 0.534247193037832, + "grad_norm": 1.5100267981368107, + "learning_rate": 1.8753518671382447e-06, + "loss": 0.9085, + "step": 5924 + }, + { + "epoch": 0.5343373765613022, + "grad_norm": 1.454164768789778, + "learning_rate": 1.8747687975615649e-06, + "loss": 0.9639, + "step": 5925 + }, + { + "epoch": 0.5344275600847725, + "grad_norm": 1.5160902430063208, + "learning_rate": 1.874185738670302e-06, + "loss": 0.8719, + "step": 5926 + }, + { + "epoch": 0.5345177436082428, + "grad_norm": 1.6359482729319361, + "learning_rate": 1.8736026905142057e-06, + "loss": 0.9723, + "step": 5927 + }, + { + "epoch": 0.534607927131713, + "grad_norm": 1.714523946012477, + "learning_rate": 1.873019653143025e-06, + "loss": 0.9629, + "step": 5928 + }, + { + "epoch": 0.5346981106551832, + "grad_norm": 1.8283358721279344, + "learning_rate": 1.8724366266065069e-06, + "loss": 0.9977, + "step": 5929 + }, + { + "epoch": 0.5347882941786536, + "grad_norm": 1.6724118948226283, + "learning_rate": 1.8718536109543998e-06, + "loss": 1.0578, + "step": 5930 + }, + { + "epoch": 0.5348784777021238, + "grad_norm": 1.9011903748201004, + "learning_rate": 1.8712706062364485e-06, + "loss": 1.0005, + "step": 5931 + }, + { + "epoch": 0.534968661225594, + "grad_norm": 1.4940274862699898, + "learning_rate": 1.8706876125024e-06, + "loss": 0.8747, + "step": 5932 + }, + { + "epoch": 0.5350588447490644, + "grad_norm": 1.6749880320562707, + "learning_rate": 1.870104629801997e-06, + "loss": 1.0216, + "step": 5933 + }, + { + "epoch": 0.5351490282725346, + "grad_norm": 1.8324864194192427, + "learning_rate": 1.8695216581849823e-06, + "loss": 0.9961, + "step": 5934 + }, + { + "epoch": 0.5352392117960049, + "grad_norm": 1.7516946761447436, + "learning_rate": 1.8689386977011003e-06, + "loss": 0.9516, + "step": 5935 + }, + { + "epoch": 0.5353293953194751, + "grad_norm": 0.6822733306362259, + "learning_rate": 1.8683557484000903e-06, + "loss": 0.8122, + "step": 5936 + }, + { + "epoch": 0.5354195788429454, + "grad_norm": 4.160941093807122, + "learning_rate": 1.8677728103316947e-06, + "loss": 1.0128, + "step": 5937 + }, + { + "epoch": 0.5355097623664157, + "grad_norm": 1.4532539556565287, + "learning_rate": 1.8671898835456518e-06, + "loss": 0.9383, + "step": 5938 + }, + { + "epoch": 0.5355999458898859, + "grad_norm": 1.9701102870723124, + "learning_rate": 1.8666069680917003e-06, + "loss": 1.0856, + "step": 5939 + }, + { + "epoch": 0.5356901294133561, + "grad_norm": 2.0660550083987337, + "learning_rate": 1.8660240640195775e-06, + "loss": 0.9423, + "step": 5940 + }, + { + "epoch": 0.5357803129368265, + "grad_norm": 1.3866210043400915, + "learning_rate": 1.8654411713790203e-06, + "loss": 1.0219, + "step": 5941 + }, + { + "epoch": 0.5358704964602967, + "grad_norm": 1.7211301102782286, + "learning_rate": 1.8648582902197648e-06, + "loss": 1.0054, + "step": 5942 + }, + { + "epoch": 0.535960679983767, + "grad_norm": 1.7406188334658848, + "learning_rate": 1.8642754205915452e-06, + "loss": 0.9894, + "step": 5943 + }, + { + "epoch": 0.5360508635072372, + "grad_norm": 1.7284877623990236, + "learning_rate": 1.8636925625440943e-06, + "loss": 0.9335, + "step": 5944 + }, + { + "epoch": 0.5361410470307075, + "grad_norm": 3.014609526595336, + "learning_rate": 1.863109716127146e-06, + "loss": 0.8906, + "step": 5945 + }, + { + "epoch": 0.5362312305541778, + "grad_norm": 1.460168564746048, + "learning_rate": 1.8625268813904311e-06, + "loss": 1.0287, + "step": 5946 + }, + { + "epoch": 0.536321414077648, + "grad_norm": 1.6083633057695477, + "learning_rate": 1.8619440583836814e-06, + "loss": 1.0075, + "step": 5947 + }, + { + "epoch": 0.5364115976011182, + "grad_norm": 1.6474590561544533, + "learning_rate": 1.8613612471566249e-06, + "loss": 0.99, + "step": 5948 + }, + { + "epoch": 0.5365017811245886, + "grad_norm": 1.603006744484804, + "learning_rate": 1.8607784477589922e-06, + "loss": 0.9318, + "step": 5949 + }, + { + "epoch": 0.5365919646480588, + "grad_norm": 1.5839655601051954, + "learning_rate": 1.8601956602405103e-06, + "loss": 1.0025, + "step": 5950 + }, + { + "epoch": 0.536682148171529, + "grad_norm": 1.4521868558931081, + "learning_rate": 1.8596128846509043e-06, + "loss": 1.0271, + "step": 5951 + }, + { + "epoch": 0.5367723316949993, + "grad_norm": 1.8631267363582737, + "learning_rate": 1.859030121039902e-06, + "loss": 0.9416, + "step": 5952 + }, + { + "epoch": 0.5368625152184696, + "grad_norm": 1.4578827934040475, + "learning_rate": 1.8584473694572268e-06, + "loss": 0.9384, + "step": 5953 + }, + { + "epoch": 0.5369526987419398, + "grad_norm": 1.622037314229186, + "learning_rate": 1.8578646299526026e-06, + "loss": 1.0057, + "step": 5954 + }, + { + "epoch": 0.5370428822654101, + "grad_norm": 3.358633553942795, + "learning_rate": 1.8572819025757518e-06, + "loss": 0.9569, + "step": 5955 + }, + { + "epoch": 0.5371330657888804, + "grad_norm": 1.601742245891885, + "learning_rate": 1.8566991873763959e-06, + "loss": 0.9995, + "step": 5956 + }, + { + "epoch": 0.5372232493123507, + "grad_norm": 1.5924682699063644, + "learning_rate": 1.856116484404256e-06, + "loss": 0.9279, + "step": 5957 + }, + { + "epoch": 0.5373134328358209, + "grad_norm": 1.77346989942355, + "learning_rate": 1.8555337937090502e-06, + "loss": 0.948, + "step": 5958 + }, + { + "epoch": 0.5374036163592911, + "grad_norm": 1.6625240897821105, + "learning_rate": 1.8549511153404984e-06, + "loss": 0.964, + "step": 5959 + }, + { + "epoch": 0.5374937998827615, + "grad_norm": 1.5295088074284968, + "learning_rate": 1.854368449348317e-06, + "loss": 1.0966, + "step": 5960 + }, + { + "epoch": 0.5375839834062317, + "grad_norm": 1.6498712027016382, + "learning_rate": 1.853785795782222e-06, + "loss": 0.942, + "step": 5961 + }, + { + "epoch": 0.5376741669297019, + "grad_norm": 1.5925622890153845, + "learning_rate": 1.85320315469193e-06, + "loss": 0.851, + "step": 5962 + }, + { + "epoch": 0.5377643504531722, + "grad_norm": 0.6534953203712333, + "learning_rate": 1.8526205261271534e-06, + "loss": 0.8189, + "step": 5963 + }, + { + "epoch": 0.5378545339766425, + "grad_norm": 1.6014198053571438, + "learning_rate": 1.852037910137607e-06, + "loss": 0.8947, + "step": 5964 + }, + { + "epoch": 0.5379447175001127, + "grad_norm": 1.3746695554914439, + "learning_rate": 1.851455306773002e-06, + "loss": 1.0704, + "step": 5965 + }, + { + "epoch": 0.538034901023583, + "grad_norm": 1.3027308062228573, + "learning_rate": 1.8508727160830483e-06, + "loss": 0.9583, + "step": 5966 + }, + { + "epoch": 0.5381250845470532, + "grad_norm": 1.4878206115898949, + "learning_rate": 1.8502901381174575e-06, + "loss": 1.0511, + "step": 5967 + }, + { + "epoch": 0.5382152680705236, + "grad_norm": 1.8131141937365138, + "learning_rate": 1.8497075729259372e-06, + "loss": 0.8895, + "step": 5968 + }, + { + "epoch": 0.5383054515939938, + "grad_norm": 0.7640262961830275, + "learning_rate": 1.8491250205581963e-06, + "loss": 0.8795, + "step": 5969 + }, + { + "epoch": 0.538395635117464, + "grad_norm": 1.9189219543743774, + "learning_rate": 1.8485424810639393e-06, + "loss": 0.9883, + "step": 5970 + }, + { + "epoch": 0.5384858186409343, + "grad_norm": 2.5847432585669936, + "learning_rate": 1.847959954492874e-06, + "loss": 1.046, + "step": 5971 + }, + { + "epoch": 0.5385760021644046, + "grad_norm": 1.8806935650239411, + "learning_rate": 1.8473774408947035e-06, + "loss": 1.0775, + "step": 5972 + }, + { + "epoch": 0.5386661856878748, + "grad_norm": 0.6573922117163036, + "learning_rate": 1.8467949403191308e-06, + "loss": 0.7902, + "step": 5973 + }, + { + "epoch": 0.5387563692113451, + "grad_norm": 1.5923282789411388, + "learning_rate": 1.8462124528158592e-06, + "loss": 1.0429, + "step": 5974 + }, + { + "epoch": 0.5388465527348153, + "grad_norm": 1.4847712740731707, + "learning_rate": 1.8456299784345881e-06, + "loss": 0.9725, + "step": 5975 + }, + { + "epoch": 0.5389367362582856, + "grad_norm": 1.5422943572926484, + "learning_rate": 1.8450475172250194e-06, + "loss": 1.0058, + "step": 5976 + }, + { + "epoch": 0.5390269197817559, + "grad_norm": 1.393356212872253, + "learning_rate": 1.844465069236851e-06, + "loss": 0.9628, + "step": 5977 + }, + { + "epoch": 0.5391171033052261, + "grad_norm": 1.7089579378535509, + "learning_rate": 1.8438826345197796e-06, + "loss": 0.8915, + "step": 5978 + }, + { + "epoch": 0.5392072868286965, + "grad_norm": 1.7455544545648864, + "learning_rate": 1.8433002131235036e-06, + "loss": 0.9731, + "step": 5979 + }, + { + "epoch": 0.5392974703521667, + "grad_norm": 1.3639360839014132, + "learning_rate": 1.8427178050977167e-06, + "loss": 1.0046, + "step": 5980 + }, + { + "epoch": 0.5393876538756369, + "grad_norm": 1.471031527183318, + "learning_rate": 1.8421354104921143e-06, + "loss": 1.0643, + "step": 5981 + }, + { + "epoch": 0.5394778373991072, + "grad_norm": 0.6025530104050644, + "learning_rate": 1.8415530293563894e-06, + "loss": 0.7861, + "step": 5982 + }, + { + "epoch": 0.5395680209225775, + "grad_norm": 1.5167245710588773, + "learning_rate": 1.8409706617402333e-06, + "loss": 0.9664, + "step": 5983 + }, + { + "epoch": 0.5396582044460477, + "grad_norm": 1.5389518891684992, + "learning_rate": 1.8403883076933378e-06, + "loss": 1.0049, + "step": 5984 + }, + { + "epoch": 0.539748387969518, + "grad_norm": 1.5062127991193937, + "learning_rate": 1.839805967265391e-06, + "loss": 0.9772, + "step": 5985 + }, + { + "epoch": 0.5398385714929882, + "grad_norm": 1.7700015489388279, + "learning_rate": 1.839223640506083e-06, + "loss": 1.0399, + "step": 5986 + }, + { + "epoch": 0.5399287550164585, + "grad_norm": 1.7218988877316617, + "learning_rate": 1.8386413274650998e-06, + "loss": 0.9798, + "step": 5987 + }, + { + "epoch": 0.5400189385399288, + "grad_norm": 1.4415487410250318, + "learning_rate": 1.8380590281921294e-06, + "loss": 1.0324, + "step": 5988 + }, + { + "epoch": 0.540109122063399, + "grad_norm": 1.551527482758407, + "learning_rate": 1.8374767427368552e-06, + "loss": 0.8895, + "step": 5989 + }, + { + "epoch": 0.5401993055868692, + "grad_norm": 1.7311078478021078, + "learning_rate": 1.8368944711489608e-06, + "loss": 1.0118, + "step": 5990 + }, + { + "epoch": 0.5402894891103396, + "grad_norm": 1.8144140331709575, + "learning_rate": 1.8363122134781304e-06, + "loss": 1.0108, + "step": 5991 + }, + { + "epoch": 0.5403796726338098, + "grad_norm": 1.5827318688780507, + "learning_rate": 1.835729969774044e-06, + "loss": 1.0091, + "step": 5992 + }, + { + "epoch": 0.54046985615728, + "grad_norm": 6.3573483698707625, + "learning_rate": 1.8351477400863823e-06, + "loss": 0.9923, + "step": 5993 + }, + { + "epoch": 0.5405600396807503, + "grad_norm": 1.6330334089591099, + "learning_rate": 1.8345655244648249e-06, + "loss": 0.9688, + "step": 5994 + }, + { + "epoch": 0.5406502232042206, + "grad_norm": 1.3593136126809107, + "learning_rate": 1.8339833229590486e-06, + "loss": 0.9775, + "step": 5995 + }, + { + "epoch": 0.5407404067276909, + "grad_norm": 1.6725198090524889, + "learning_rate": 1.833401135618731e-06, + "loss": 1.0528, + "step": 5996 + }, + { + "epoch": 0.5408305902511611, + "grad_norm": 1.3596959228535297, + "learning_rate": 1.8328189624935466e-06, + "loss": 0.9096, + "step": 5997 + }, + { + "epoch": 0.5409207737746313, + "grad_norm": 1.8420085765088086, + "learning_rate": 1.832236803633171e-06, + "loss": 1.0196, + "step": 5998 + }, + { + "epoch": 0.5410109572981017, + "grad_norm": 1.9179957867676538, + "learning_rate": 1.831654659087276e-06, + "loss": 0.9518, + "step": 5999 + }, + { + "epoch": 0.5411011408215719, + "grad_norm": 2.416334151269402, + "learning_rate": 1.831072528905533e-06, + "loss": 1.0027, + "step": 6000 + }, + { + "epoch": 0.5411913243450421, + "grad_norm": 1.430808456477172, + "learning_rate": 1.8304904131376142e-06, + "loss": 1.0376, + "step": 6001 + }, + { + "epoch": 0.5412815078685124, + "grad_norm": 1.7033309811963429, + "learning_rate": 1.8299083118331874e-06, + "loss": 1.0007, + "step": 6002 + }, + { + "epoch": 0.5413716913919827, + "grad_norm": 1.5404603293664945, + "learning_rate": 1.8293262250419217e-06, + "loss": 1.1195, + "step": 6003 + }, + { + "epoch": 0.541461874915453, + "grad_norm": 1.7231719268065233, + "learning_rate": 1.828744152813484e-06, + "loss": 1.0989, + "step": 6004 + }, + { + "epoch": 0.5415520584389232, + "grad_norm": 1.6671481470184513, + "learning_rate": 1.8281620951975382e-06, + "loss": 0.9369, + "step": 6005 + }, + { + "epoch": 0.5416422419623935, + "grad_norm": 1.5367750800416315, + "learning_rate": 1.827580052243751e-06, + "loss": 0.9448, + "step": 6006 + }, + { + "epoch": 0.5417324254858638, + "grad_norm": 2.0380237023320142, + "learning_rate": 1.826998024001784e-06, + "loss": 1.0177, + "step": 6007 + }, + { + "epoch": 0.541822609009334, + "grad_norm": 2.111700739697291, + "learning_rate": 1.8264160105212995e-06, + "loss": 1.0135, + "step": 6008 + }, + { + "epoch": 0.5419127925328042, + "grad_norm": 1.5037886140071934, + "learning_rate": 1.8258340118519582e-06, + "loss": 0.9512, + "step": 6009 + }, + { + "epoch": 0.5420029760562746, + "grad_norm": 3.1173325124869504, + "learning_rate": 1.82525202804342e-06, + "loss": 0.9693, + "step": 6010 + }, + { + "epoch": 0.5420931595797448, + "grad_norm": 1.8238336127961117, + "learning_rate": 1.8246700591453415e-06, + "loss": 0.9671, + "step": 6011 + }, + { + "epoch": 0.542183343103215, + "grad_norm": 1.6793211781459516, + "learning_rate": 1.8240881052073801e-06, + "loss": 0.9848, + "step": 6012 + }, + { + "epoch": 0.5422735266266853, + "grad_norm": 1.4866581273510795, + "learning_rate": 1.8235061662791923e-06, + "loss": 0.9555, + "step": 6013 + }, + { + "epoch": 0.5423637101501556, + "grad_norm": 1.8225686821992795, + "learning_rate": 1.8229242424104309e-06, + "loss": 1.0574, + "step": 6014 + }, + { + "epoch": 0.5424538936736258, + "grad_norm": 0.615506889902779, + "learning_rate": 1.8223423336507503e-06, + "loss": 0.7968, + "step": 6015 + }, + { + "epoch": 0.5425440771970961, + "grad_norm": 1.6884155615426768, + "learning_rate": 1.8217604400498012e-06, + "loss": 0.9671, + "step": 6016 + }, + { + "epoch": 0.5426342607205663, + "grad_norm": 1.3383859427544416, + "learning_rate": 1.8211785616572333e-06, + "loss": 1.033, + "step": 6017 + }, + { + "epoch": 0.5427244442440367, + "grad_norm": 1.6650818760533013, + "learning_rate": 1.8205966985226975e-06, + "loss": 1.0532, + "step": 6018 + }, + { + "epoch": 0.5428146277675069, + "grad_norm": 1.2662529976782002, + "learning_rate": 1.8200148506958397e-06, + "loss": 0.9917, + "step": 6019 + }, + { + "epoch": 0.5429048112909771, + "grad_norm": 0.6505900584135896, + "learning_rate": 1.819433018226308e-06, + "loss": 0.8428, + "step": 6020 + }, + { + "epoch": 0.5429949948144474, + "grad_norm": 0.6006016178720325, + "learning_rate": 1.8188512011637471e-06, + "loss": 0.7873, + "step": 6021 + }, + { + "epoch": 0.5430851783379177, + "grad_norm": 1.625874284156518, + "learning_rate": 1.8182693995578e-06, + "loss": 0.9634, + "step": 6022 + }, + { + "epoch": 0.5431753618613879, + "grad_norm": 1.2261724641573082, + "learning_rate": 1.8176876134581098e-06, + "loss": 0.8635, + "step": 6023 + }, + { + "epoch": 0.5432655453848582, + "grad_norm": 1.6769932629899769, + "learning_rate": 1.8171058429143176e-06, + "loss": 0.9313, + "step": 6024 + }, + { + "epoch": 0.5433557289083284, + "grad_norm": 1.3710494075027788, + "learning_rate": 1.8165240879760637e-06, + "loss": 1.0251, + "step": 6025 + }, + { + "epoch": 0.5434459124317987, + "grad_norm": 1.950896494756963, + "learning_rate": 1.8159423486929862e-06, + "loss": 1.0162, + "step": 6026 + }, + { + "epoch": 0.543536095955269, + "grad_norm": 1.7175515591629786, + "learning_rate": 1.815360625114722e-06, + "loss": 0.9867, + "step": 6027 + }, + { + "epoch": 0.5436262794787392, + "grad_norm": 1.6296785202583606, + "learning_rate": 1.814778917290908e-06, + "loss": 0.9728, + "step": 6028 + }, + { + "epoch": 0.5437164630022095, + "grad_norm": 2.8603874040196406, + "learning_rate": 1.8141972252711773e-06, + "loss": 1.003, + "step": 6029 + }, + { + "epoch": 0.5438066465256798, + "grad_norm": 2.2234675798774015, + "learning_rate": 1.8136155491051645e-06, + "loss": 0.9839, + "step": 6030 + }, + { + "epoch": 0.54389683004915, + "grad_norm": 0.6690106882045537, + "learning_rate": 1.8130338888424998e-06, + "loss": 0.832, + "step": 6031 + }, + { + "epoch": 0.5439870135726202, + "grad_norm": 1.6421419669098938, + "learning_rate": 1.812452244532816e-06, + "loss": 1.0062, + "step": 6032 + }, + { + "epoch": 0.5440771970960906, + "grad_norm": 1.6420848346126766, + "learning_rate": 1.8118706162257405e-06, + "loss": 0.9245, + "step": 6033 + }, + { + "epoch": 0.5441673806195608, + "grad_norm": 1.4671069755841315, + "learning_rate": 1.8112890039709002e-06, + "loss": 1.0684, + "step": 6034 + }, + { + "epoch": 0.5442575641430311, + "grad_norm": 1.701883896248632, + "learning_rate": 1.8107074078179238e-06, + "loss": 0.9957, + "step": 6035 + }, + { + "epoch": 0.5443477476665013, + "grad_norm": 1.4449125819179394, + "learning_rate": 1.8101258278164348e-06, + "loss": 0.938, + "step": 6036 + }, + { + "epoch": 0.5444379311899716, + "grad_norm": 1.7463119336383952, + "learning_rate": 1.8095442640160575e-06, + "loss": 1.092, + "step": 6037 + }, + { + "epoch": 0.5445281147134419, + "grad_norm": 2.555181519691268, + "learning_rate": 1.8089627164664132e-06, + "loss": 0.9511, + "step": 6038 + }, + { + "epoch": 0.5446182982369121, + "grad_norm": 1.3642675990611757, + "learning_rate": 1.8083811852171233e-06, + "loss": 0.9069, + "step": 6039 + }, + { + "epoch": 0.5447084817603823, + "grad_norm": 1.6861820018905895, + "learning_rate": 1.8077996703178078e-06, + "loss": 0.9856, + "step": 6040 + }, + { + "epoch": 0.5447986652838527, + "grad_norm": 1.5893863704273175, + "learning_rate": 1.8072181718180833e-06, + "loss": 0.9329, + "step": 6041 + }, + { + "epoch": 0.5448888488073229, + "grad_norm": 1.5457384962961507, + "learning_rate": 1.806636689767568e-06, + "loss": 1.0736, + "step": 6042 + }, + { + "epoch": 0.5449790323307931, + "grad_norm": 1.518133152782399, + "learning_rate": 1.8060552242158769e-06, + "loss": 0.9272, + "step": 6043 + }, + { + "epoch": 0.5450692158542634, + "grad_norm": 1.4129941497919203, + "learning_rate": 1.8054737752126224e-06, + "loss": 0.9576, + "step": 6044 + }, + { + "epoch": 0.5451593993777337, + "grad_norm": 2.3381099741436855, + "learning_rate": 1.804892342807419e-06, + "loss": 0.9743, + "step": 6045 + }, + { + "epoch": 0.545249582901204, + "grad_norm": 2.7154629008175664, + "learning_rate": 1.8043109270498756e-06, + "loss": 1.0579, + "step": 6046 + }, + { + "epoch": 0.5453397664246742, + "grad_norm": 1.7309502132050136, + "learning_rate": 1.803729527989604e-06, + "loss": 0.9956, + "step": 6047 + }, + { + "epoch": 0.5454299499481444, + "grad_norm": 1.4624379658380593, + "learning_rate": 1.8031481456762112e-06, + "loss": 0.9788, + "step": 6048 + }, + { + "epoch": 0.5455201334716148, + "grad_norm": 1.6483868053288644, + "learning_rate": 1.8025667801593033e-06, + "loss": 0.9128, + "step": 6049 + }, + { + "epoch": 0.545610316995085, + "grad_norm": 1.6194747745661038, + "learning_rate": 1.8019854314884871e-06, + "loss": 0.9244, + "step": 6050 + }, + { + "epoch": 0.5457005005185552, + "grad_norm": 1.3884756644440979, + "learning_rate": 1.8014040997133652e-06, + "loss": 1.0367, + "step": 6051 + }, + { + "epoch": 0.5457906840420256, + "grad_norm": 0.7605401727021547, + "learning_rate": 1.8008227848835414e-06, + "loss": 0.8513, + "step": 6052 + }, + { + "epoch": 0.5458808675654958, + "grad_norm": 3.2605947785076497, + "learning_rate": 1.8002414870486144e-06, + "loss": 0.9366, + "step": 6053 + }, + { + "epoch": 0.545971051088966, + "grad_norm": 1.4444359673145346, + "learning_rate": 1.7996602062581864e-06, + "loss": 0.9397, + "step": 6054 + }, + { + "epoch": 0.5460612346124363, + "grad_norm": 2.0864750887318158, + "learning_rate": 1.7990789425618544e-06, + "loss": 1.0349, + "step": 6055 + }, + { + "epoch": 0.5461514181359066, + "grad_norm": 1.3087384499784382, + "learning_rate": 1.7984976960092137e-06, + "loss": 0.9692, + "step": 6056 + }, + { + "epoch": 0.5462416016593769, + "grad_norm": 1.826028322503671, + "learning_rate": 1.7979164666498617e-06, + "loss": 0.949, + "step": 6057 + }, + { + "epoch": 0.5463317851828471, + "grad_norm": 0.6229544783335954, + "learning_rate": 1.7973352545333901e-06, + "loss": 0.7907, + "step": 6058 + }, + { + "epoch": 0.5464219687063173, + "grad_norm": 1.6119802923819988, + "learning_rate": 1.796754059709393e-06, + "loss": 0.9048, + "step": 6059 + }, + { + "epoch": 0.5465121522297877, + "grad_norm": 1.6627171051988299, + "learning_rate": 1.7961728822274603e-06, + "loss": 1.0466, + "step": 6060 + }, + { + "epoch": 0.5466023357532579, + "grad_norm": 1.5644362849973419, + "learning_rate": 1.7955917221371802e-06, + "loss": 0.9403, + "step": 6061 + }, + { + "epoch": 0.5466925192767281, + "grad_norm": 1.893959195004615, + "learning_rate": 1.7950105794881422e-06, + "loss": 0.9031, + "step": 6062 + }, + { + "epoch": 0.5467827028001984, + "grad_norm": 1.5091538071712045, + "learning_rate": 1.7944294543299317e-06, + "loss": 0.9712, + "step": 6063 + }, + { + "epoch": 0.5468728863236687, + "grad_norm": 1.618376879605467, + "learning_rate": 1.7938483467121333e-06, + "loss": 1.039, + "step": 6064 + }, + { + "epoch": 0.5469630698471389, + "grad_norm": 1.6638711452157195, + "learning_rate": 1.7932672566843313e-06, + "loss": 0.9662, + "step": 6065 + }, + { + "epoch": 0.5470532533706092, + "grad_norm": 1.6761920132090697, + "learning_rate": 1.7926861842961065e-06, + "loss": 0.9082, + "step": 6066 + }, + { + "epoch": 0.5471434368940794, + "grad_norm": 1.6368610620651733, + "learning_rate": 1.7921051295970399e-06, + "loss": 0.907, + "step": 6067 + }, + { + "epoch": 0.5472336204175497, + "grad_norm": 2.310299831929944, + "learning_rate": 1.7915240926367092e-06, + "loss": 0.9461, + "step": 6068 + }, + { + "epoch": 0.54732380394102, + "grad_norm": 1.375427613090092, + "learning_rate": 1.7909430734646932e-06, + "loss": 1.0284, + "step": 6069 + }, + { + "epoch": 0.5474139874644902, + "grad_norm": 0.674300782660073, + "learning_rate": 1.790362072130567e-06, + "loss": 0.8236, + "step": 6070 + }, + { + "epoch": 0.5475041709879604, + "grad_norm": 1.6015611602033268, + "learning_rate": 1.7897810886839037e-06, + "loss": 0.9723, + "step": 6071 + }, + { + "epoch": 0.5475943545114308, + "grad_norm": 1.4986695110988661, + "learning_rate": 1.7892001231742782e-06, + "loss": 0.995, + "step": 6072 + }, + { + "epoch": 0.547684538034901, + "grad_norm": 1.7170171845853073, + "learning_rate": 1.7886191756512594e-06, + "loss": 0.9934, + "step": 6073 + }, + { + "epoch": 0.5477747215583713, + "grad_norm": 1.4620601807822426, + "learning_rate": 1.7880382461644192e-06, + "loss": 1.0665, + "step": 6074 + }, + { + "epoch": 0.5478649050818416, + "grad_norm": 1.535127906440597, + "learning_rate": 1.7874573347633235e-06, + "loss": 0.9594, + "step": 6075 + }, + { + "epoch": 0.5479550886053118, + "grad_norm": 1.929086815628116, + "learning_rate": 1.7868764414975408e-06, + "loss": 1.0578, + "step": 6076 + }, + { + "epoch": 0.5480452721287821, + "grad_norm": 1.878883853390262, + "learning_rate": 1.7862955664166353e-06, + "loss": 0.9328, + "step": 6077 + }, + { + "epoch": 0.5481354556522523, + "grad_norm": 1.2846359080740237, + "learning_rate": 1.78571470957017e-06, + "loss": 1.0679, + "step": 6078 + }, + { + "epoch": 0.5482256391757226, + "grad_norm": 1.902796634602292, + "learning_rate": 1.7851338710077074e-06, + "loss": 1.0488, + "step": 6079 + }, + { + "epoch": 0.5483158226991929, + "grad_norm": 1.744701880427012, + "learning_rate": 1.7845530507788076e-06, + "loss": 1.0923, + "step": 6080 + }, + { + "epoch": 0.5484060062226631, + "grad_norm": 2.753980876371409, + "learning_rate": 1.7839722489330298e-06, + "loss": 0.9552, + "step": 6081 + }, + { + "epoch": 0.5484961897461333, + "grad_norm": 1.6594349449810153, + "learning_rate": 1.7833914655199308e-06, + "loss": 0.9543, + "step": 6082 + }, + { + "epoch": 0.5485863732696037, + "grad_norm": 1.7131350961559553, + "learning_rate": 1.7828107005890658e-06, + "loss": 1.0292, + "step": 6083 + }, + { + "epoch": 0.5486765567930739, + "grad_norm": 9.182532388290918, + "learning_rate": 1.7822299541899898e-06, + "loss": 0.9028, + "step": 6084 + }, + { + "epoch": 0.5487667403165442, + "grad_norm": 1.3715288824836156, + "learning_rate": 1.7816492263722545e-06, + "loss": 0.9848, + "step": 6085 + }, + { + "epoch": 0.5488569238400144, + "grad_norm": 2.2992684486298565, + "learning_rate": 1.781068517185412e-06, + "loss": 1.0093, + "step": 6086 + }, + { + "epoch": 0.5489471073634847, + "grad_norm": 3.034419496879829, + "learning_rate": 1.7804878266790104e-06, + "loss": 1.0518, + "step": 6087 + }, + { + "epoch": 0.549037290886955, + "grad_norm": 1.6205271270656545, + "learning_rate": 1.779907154902597e-06, + "loss": 0.921, + "step": 6088 + }, + { + "epoch": 0.5491274744104252, + "grad_norm": 1.8491704756411151, + "learning_rate": 1.7793265019057198e-06, + "loss": 0.9141, + "step": 6089 + }, + { + "epoch": 0.5492176579338954, + "grad_norm": 0.7212758256992458, + "learning_rate": 1.7787458677379212e-06, + "loss": 0.832, + "step": 6090 + }, + { + "epoch": 0.5493078414573658, + "grad_norm": 1.7117296588503084, + "learning_rate": 1.7781652524487463e-06, + "loss": 1.0425, + "step": 6091 + }, + { + "epoch": 0.549398024980836, + "grad_norm": 2.1564025824920354, + "learning_rate": 1.777584656087735e-06, + "loss": 0.9303, + "step": 6092 + }, + { + "epoch": 0.5494882085043062, + "grad_norm": 1.8620118109413493, + "learning_rate": 1.777004078704427e-06, + "loss": 0.9078, + "step": 6093 + }, + { + "epoch": 0.5495783920277765, + "grad_norm": 1.7769544237911554, + "learning_rate": 1.7764235203483603e-06, + "loss": 1.0068, + "step": 6094 + }, + { + "epoch": 0.5496685755512468, + "grad_norm": 1.7175741161362985, + "learning_rate": 1.775842981069072e-06, + "loss": 0.9584, + "step": 6095 + }, + { + "epoch": 0.549758759074717, + "grad_norm": 1.4352752220915765, + "learning_rate": 1.7752624609160966e-06, + "loss": 0.9972, + "step": 6096 + }, + { + "epoch": 0.5498489425981873, + "grad_norm": 2.462013227500058, + "learning_rate": 1.7746819599389665e-06, + "loss": 1.0215, + "step": 6097 + }, + { + "epoch": 0.5499391261216575, + "grad_norm": 1.5556204598752534, + "learning_rate": 1.774101478187215e-06, + "loss": 0.9908, + "step": 6098 + }, + { + "epoch": 0.5500293096451279, + "grad_norm": 2.1500658557475663, + "learning_rate": 1.773521015710371e-06, + "loss": 1.0545, + "step": 6099 + }, + { + "epoch": 0.5501194931685981, + "grad_norm": 1.8326153505249638, + "learning_rate": 1.7729405725579614e-06, + "loss": 1.0638, + "step": 6100 + }, + { + "epoch": 0.5502096766920683, + "grad_norm": 1.5823357936523117, + "learning_rate": 1.7723601487795151e-06, + "loss": 0.963, + "step": 6101 + }, + { + "epoch": 0.5502998602155387, + "grad_norm": 1.360552792751203, + "learning_rate": 1.7717797444245557e-06, + "loss": 0.9344, + "step": 6102 + }, + { + "epoch": 0.5503900437390089, + "grad_norm": 1.2474657117454535, + "learning_rate": 1.7711993595426076e-06, + "loss": 0.9194, + "step": 6103 + }, + { + "epoch": 0.5504802272624791, + "grad_norm": 4.311418923032511, + "learning_rate": 1.7706189941831915e-06, + "loss": 1.0956, + "step": 6104 + }, + { + "epoch": 0.5505704107859494, + "grad_norm": 1.5480043546433029, + "learning_rate": 1.770038648395827e-06, + "loss": 0.9464, + "step": 6105 + }, + { + "epoch": 0.5506605943094197, + "grad_norm": 0.6124461860760969, + "learning_rate": 1.7694583222300336e-06, + "loss": 0.8317, + "step": 6106 + }, + { + "epoch": 0.55075077783289, + "grad_norm": 1.7784013556878076, + "learning_rate": 1.7688780157353272e-06, + "loss": 0.9667, + "step": 6107 + }, + { + "epoch": 0.5508409613563602, + "grad_norm": 1.6036700268471487, + "learning_rate": 1.768297728961223e-06, + "loss": 0.9307, + "step": 6108 + }, + { + "epoch": 0.5509311448798304, + "grad_norm": 1.719998131906249, + "learning_rate": 1.7677174619572342e-06, + "loss": 0.9884, + "step": 6109 + }, + { + "epoch": 0.5510213284033008, + "grad_norm": 1.6378775707318733, + "learning_rate": 1.7671372147728717e-06, + "loss": 1.0067, + "step": 6110 + }, + { + "epoch": 0.551111511926771, + "grad_norm": 4.275078773888175, + "learning_rate": 1.7665569874576471e-06, + "loss": 1.0552, + "step": 6111 + }, + { + "epoch": 0.5512016954502412, + "grad_norm": 1.8277865365516464, + "learning_rate": 1.7659767800610664e-06, + "loss": 0.9923, + "step": 6112 + }, + { + "epoch": 0.5512918789737115, + "grad_norm": 1.8927723063559139, + "learning_rate": 1.7653965926326379e-06, + "loss": 0.8942, + "step": 6113 + }, + { + "epoch": 0.5513820624971818, + "grad_norm": 1.1993980646425364, + "learning_rate": 1.764816425221866e-06, + "loss": 0.9416, + "step": 6114 + }, + { + "epoch": 0.551472246020652, + "grad_norm": 1.3599959522776113, + "learning_rate": 1.7642362778782524e-06, + "loss": 0.9038, + "step": 6115 + }, + { + "epoch": 0.5515624295441223, + "grad_norm": 1.8030331117167597, + "learning_rate": 1.7636561506513005e-06, + "loss": 0.9575, + "step": 6116 + }, + { + "epoch": 0.5516526130675925, + "grad_norm": 1.431312596314065, + "learning_rate": 1.7630760435905083e-06, + "loss": 0.9627, + "step": 6117 + }, + { + "epoch": 0.5517427965910628, + "grad_norm": 1.7675769110406743, + "learning_rate": 1.762495956745375e-06, + "loss": 0.9432, + "step": 6118 + }, + { + "epoch": 0.5518329801145331, + "grad_norm": 1.6993173154108683, + "learning_rate": 1.7619158901653962e-06, + "loss": 1.0276, + "step": 6119 + }, + { + "epoch": 0.5519231636380033, + "grad_norm": 1.6922937756079546, + "learning_rate": 1.761335843900066e-06, + "loss": 0.9829, + "step": 6120 + }, + { + "epoch": 0.5520133471614735, + "grad_norm": 2.3976516678841846, + "learning_rate": 1.7607558179988785e-06, + "loss": 0.9728, + "step": 6121 + }, + { + "epoch": 0.5521035306849439, + "grad_norm": 1.9264633487330718, + "learning_rate": 1.760175812511323e-06, + "loss": 0.9176, + "step": 6122 + }, + { + "epoch": 0.5521937142084141, + "grad_norm": 1.7637746111411883, + "learning_rate": 1.75959582748689e-06, + "loss": 0.945, + "step": 6123 + }, + { + "epoch": 0.5522838977318844, + "grad_norm": 2.105396729548601, + "learning_rate": 1.7590158629750657e-06, + "loss": 1.0164, + "step": 6124 + }, + { + "epoch": 0.5523740812553547, + "grad_norm": 2.808345778425433, + "learning_rate": 1.7584359190253376e-06, + "loss": 1.0365, + "step": 6125 + }, + { + "epoch": 0.5524642647788249, + "grad_norm": 1.8747155927488517, + "learning_rate": 1.7578559956871892e-06, + "loss": 1.0496, + "step": 6126 + }, + { + "epoch": 0.5525544483022952, + "grad_norm": 1.4609051053132598, + "learning_rate": 1.7572760930101012e-06, + "loss": 0.9834, + "step": 6127 + }, + { + "epoch": 0.5526446318257654, + "grad_norm": 0.6896622856519385, + "learning_rate": 1.7566962110435563e-06, + "loss": 0.8707, + "step": 6128 + }, + { + "epoch": 0.5527348153492357, + "grad_norm": 1.5537626496442252, + "learning_rate": 1.7561163498370313e-06, + "loss": 1.004, + "step": 6129 + }, + { + "epoch": 0.552824998872706, + "grad_norm": 1.721531022567054, + "learning_rate": 1.755536509440005e-06, + "loss": 0.9499, + "step": 6130 + }, + { + "epoch": 0.5529151823961762, + "grad_norm": 2.0101218394459885, + "learning_rate": 1.7549566899019519e-06, + "loss": 0.9831, + "step": 6131 + }, + { + "epoch": 0.5530053659196464, + "grad_norm": 1.5505829114940872, + "learning_rate": 1.754376891272344e-06, + "loss": 1.0032, + "step": 6132 + }, + { + "epoch": 0.5530955494431168, + "grad_norm": 1.3493503261216766, + "learning_rate": 1.753797113600655e-06, + "loss": 0.9348, + "step": 6133 + }, + { + "epoch": 0.553185732966587, + "grad_norm": 1.5563652149125993, + "learning_rate": 1.7532173569363535e-06, + "loss": 1.1221, + "step": 6134 + }, + { + "epoch": 0.5532759164900573, + "grad_norm": 1.6719279058053589, + "learning_rate": 1.7526376213289077e-06, + "loss": 0.8061, + "step": 6135 + }, + { + "epoch": 0.5533661000135275, + "grad_norm": 1.6615461208704936, + "learning_rate": 1.7520579068277844e-06, + "loss": 0.9731, + "step": 6136 + }, + { + "epoch": 0.5534562835369978, + "grad_norm": 1.4583691547172355, + "learning_rate": 1.7514782134824472e-06, + "loss": 1.0138, + "step": 6137 + }, + { + "epoch": 0.5535464670604681, + "grad_norm": 1.5880331090459119, + "learning_rate": 1.7508985413423599e-06, + "loss": 0.9831, + "step": 6138 + }, + { + "epoch": 0.5536366505839383, + "grad_norm": 1.7275592119196599, + "learning_rate": 1.7503188904569814e-06, + "loss": 1.0565, + "step": 6139 + }, + { + "epoch": 0.5537268341074085, + "grad_norm": 1.727896472796656, + "learning_rate": 1.7497392608757728e-06, + "loss": 0.9513, + "step": 6140 + }, + { + "epoch": 0.5538170176308789, + "grad_norm": 1.6544847257894173, + "learning_rate": 1.7491596526481897e-06, + "loss": 1.0038, + "step": 6141 + }, + { + "epoch": 0.5539072011543491, + "grad_norm": 1.6777398461284847, + "learning_rate": 1.7485800658236888e-06, + "loss": 0.9121, + "step": 6142 + }, + { + "epoch": 0.5539973846778193, + "grad_norm": 1.78511130779297, + "learning_rate": 1.7480005004517232e-06, + "loss": 1.0606, + "step": 6143 + }, + { + "epoch": 0.5540875682012896, + "grad_norm": 1.5942903986264823, + "learning_rate": 1.7474209565817435e-06, + "loss": 1.0487, + "step": 6144 + }, + { + "epoch": 0.5541777517247599, + "grad_norm": 1.5588296888549407, + "learning_rate": 1.7468414342632014e-06, + "loss": 0.9806, + "step": 6145 + }, + { + "epoch": 0.5542679352482301, + "grad_norm": 1.5518941657318936, + "learning_rate": 1.746261933545543e-06, + "loss": 1.0419, + "step": 6146 + }, + { + "epoch": 0.5543581187717004, + "grad_norm": 1.4033528472165842, + "learning_rate": 1.7456824544782165e-06, + "loss": 0.9823, + "step": 6147 + }, + { + "epoch": 0.5544483022951707, + "grad_norm": 1.6865389171037946, + "learning_rate": 1.7451029971106653e-06, + "loss": 0.9934, + "step": 6148 + }, + { + "epoch": 0.554538485818641, + "grad_norm": 1.7130156833489694, + "learning_rate": 1.7445235614923313e-06, + "loss": 1.0149, + "step": 6149 + }, + { + "epoch": 0.5546286693421112, + "grad_norm": 1.4421655985945343, + "learning_rate": 1.7439441476726556e-06, + "loss": 0.974, + "step": 6150 + }, + { + "epoch": 0.5547188528655814, + "grad_norm": 2.3296787274590134, + "learning_rate": 1.7433647557010776e-06, + "loss": 0.9954, + "step": 6151 + }, + { + "epoch": 0.5548090363890518, + "grad_norm": 1.4707960122248143, + "learning_rate": 1.7427853856270338e-06, + "loss": 0.9498, + "step": 6152 + }, + { + "epoch": 0.554899219912522, + "grad_norm": 1.4802394880666643, + "learning_rate": 1.7422060374999593e-06, + "loss": 0.9509, + "step": 6153 + }, + { + "epoch": 0.5549894034359922, + "grad_norm": 1.5679768515065367, + "learning_rate": 1.7416267113692862e-06, + "loss": 0.9456, + "step": 6154 + }, + { + "epoch": 0.5550795869594625, + "grad_norm": 1.4510290966415997, + "learning_rate": 1.7410474072844475e-06, + "loss": 1.0179, + "step": 6155 + }, + { + "epoch": 0.5551697704829328, + "grad_norm": 1.5304919332137776, + "learning_rate": 1.740468125294871e-06, + "loss": 0.9735, + "step": 6156 + }, + { + "epoch": 0.555259954006403, + "grad_norm": 1.77074958004826, + "learning_rate": 1.739888865449986e-06, + "loss": 0.8984, + "step": 6157 + }, + { + "epoch": 0.5553501375298733, + "grad_norm": 1.9577514176152289, + "learning_rate": 1.7393096277992174e-06, + "loss": 1.0099, + "step": 6158 + }, + { + "epoch": 0.5554403210533435, + "grad_norm": 1.6360671173309178, + "learning_rate": 1.738730412391988e-06, + "loss": 0.9991, + "step": 6159 + }, + { + "epoch": 0.5555305045768139, + "grad_norm": 1.5278700290147469, + "learning_rate": 1.738151219277721e-06, + "loss": 1.0408, + "step": 6160 + }, + { + "epoch": 0.5556206881002841, + "grad_norm": 1.8282810372819573, + "learning_rate": 1.7375720485058349e-06, + "loss": 0.9434, + "step": 6161 + }, + { + "epoch": 0.5557108716237543, + "grad_norm": 1.8027105030202673, + "learning_rate": 1.7369929001257498e-06, + "loss": 1.0028, + "step": 6162 + }, + { + "epoch": 0.5558010551472246, + "grad_norm": 2.0549481012700963, + "learning_rate": 1.73641377418688e-06, + "loss": 0.9008, + "step": 6163 + }, + { + "epoch": 0.5558912386706949, + "grad_norm": 1.4815084787178294, + "learning_rate": 1.7358346707386408e-06, + "loss": 0.9209, + "step": 6164 + }, + { + "epoch": 0.5559814221941651, + "grad_norm": 1.7181355852041764, + "learning_rate": 1.7352555898304439e-06, + "loss": 1.0342, + "step": 6165 + }, + { + "epoch": 0.5560716057176354, + "grad_norm": 3.7322814691425386, + "learning_rate": 1.7346765315116996e-06, + "loss": 0.9051, + "step": 6166 + }, + { + "epoch": 0.5561617892411056, + "grad_norm": 1.5191221495505642, + "learning_rate": 1.734097495831817e-06, + "loss": 0.9458, + "step": 6167 + }, + { + "epoch": 0.5562519727645759, + "grad_norm": 2.613898736050316, + "learning_rate": 1.7335184828402015e-06, + "loss": 1.002, + "step": 6168 + }, + { + "epoch": 0.5563421562880462, + "grad_norm": 1.4848642446457743, + "learning_rate": 1.7329394925862595e-06, + "loss": 0.967, + "step": 6169 + }, + { + "epoch": 0.5564323398115164, + "grad_norm": 1.7106129238272447, + "learning_rate": 1.7323605251193922e-06, + "loss": 1.0353, + "step": 6170 + }, + { + "epoch": 0.5565225233349868, + "grad_norm": 2.0016946765338526, + "learning_rate": 1.7317815804890001e-06, + "loss": 1.0058, + "step": 6171 + }, + { + "epoch": 0.556612706858457, + "grad_norm": 1.6389958603569852, + "learning_rate": 1.731202658744483e-06, + "loss": 0.9396, + "step": 6172 + }, + { + "epoch": 0.5567028903819272, + "grad_norm": 1.5799592943621779, + "learning_rate": 1.7306237599352365e-06, + "loss": 1.0483, + "step": 6173 + }, + { + "epoch": 0.5567930739053975, + "grad_norm": 1.3311236917097224, + "learning_rate": 1.730044884110657e-06, + "loss": 0.9706, + "step": 6174 + }, + { + "epoch": 0.5568832574288678, + "grad_norm": 1.6325913420939582, + "learning_rate": 1.7294660313201366e-06, + "loss": 0.9431, + "step": 6175 + }, + { + "epoch": 0.556973440952338, + "grad_norm": 2.4683260976241868, + "learning_rate": 1.7288872016130652e-06, + "loss": 0.9686, + "step": 6176 + }, + { + "epoch": 0.5570636244758083, + "grad_norm": 2.420580117863201, + "learning_rate": 1.7283083950388334e-06, + "loss": 0.8997, + "step": 6177 + }, + { + "epoch": 0.5571538079992785, + "grad_norm": 1.3766269595136476, + "learning_rate": 1.727729611646827e-06, + "loss": 0.8894, + "step": 6178 + }, + { + "epoch": 0.5572439915227488, + "grad_norm": 1.7095763483663298, + "learning_rate": 1.7271508514864318e-06, + "loss": 1.0402, + "step": 6179 + }, + { + "epoch": 0.5573341750462191, + "grad_norm": 1.9521982664663542, + "learning_rate": 1.7265721146070302e-06, + "loss": 0.9857, + "step": 6180 + }, + { + "epoch": 0.5574243585696893, + "grad_norm": 1.6335064449177323, + "learning_rate": 1.7259934010580035e-06, + "loss": 0.9841, + "step": 6181 + }, + { + "epoch": 0.5575145420931595, + "grad_norm": 1.5129832417933484, + "learning_rate": 1.725414710888731e-06, + "loss": 0.9347, + "step": 6182 + }, + { + "epoch": 0.5576047256166299, + "grad_norm": 1.3769687516428226, + "learning_rate": 1.7248360441485885e-06, + "loss": 0.9627, + "step": 6183 + }, + { + "epoch": 0.5576949091401001, + "grad_norm": 1.727437483764591, + "learning_rate": 1.7242574008869528e-06, + "loss": 1.059, + "step": 6184 + }, + { + "epoch": 0.5577850926635703, + "grad_norm": 0.6891125494325073, + "learning_rate": 1.7236787811531951e-06, + "loss": 0.866, + "step": 6185 + }, + { + "epoch": 0.5578752761870406, + "grad_norm": 1.987059092049512, + "learning_rate": 1.7231001849966887e-06, + "loss": 0.9522, + "step": 6186 + }, + { + "epoch": 0.5579654597105109, + "grad_norm": 0.6499884109197537, + "learning_rate": 1.722521612466801e-06, + "loss": 0.8444, + "step": 6187 + }, + { + "epoch": 0.5580556432339812, + "grad_norm": 1.6405691954794988, + "learning_rate": 1.7219430636128984e-06, + "loss": 0.9311, + "step": 6188 + }, + { + "epoch": 0.5581458267574514, + "grad_norm": 1.4730267855403, + "learning_rate": 1.7213645384843479e-06, + "loss": 0.9262, + "step": 6189 + }, + { + "epoch": 0.5582360102809216, + "grad_norm": 4.30097017211079, + "learning_rate": 1.7207860371305108e-06, + "loss": 0.9782, + "step": 6190 + }, + { + "epoch": 0.558326193804392, + "grad_norm": 4.337625480837612, + "learning_rate": 1.7202075596007487e-06, + "loss": 0.953, + "step": 6191 + }, + { + "epoch": 0.5584163773278622, + "grad_norm": 1.7093992242020695, + "learning_rate": 1.7196291059444206e-06, + "loss": 0.9461, + "step": 6192 + }, + { + "epoch": 0.5585065608513324, + "grad_norm": 1.7510496125352053, + "learning_rate": 1.7190506762108828e-06, + "loss": 0.9697, + "step": 6193 + }, + { + "epoch": 0.5585967443748028, + "grad_norm": 1.8504899536041508, + "learning_rate": 1.7184722704494907e-06, + "loss": 0.9839, + "step": 6194 + }, + { + "epoch": 0.558686927898273, + "grad_norm": 1.587931990231312, + "learning_rate": 1.717893888709596e-06, + "loss": 0.9967, + "step": 6195 + }, + { + "epoch": 0.5587771114217432, + "grad_norm": 1.6015111403254958, + "learning_rate": 1.7173155310405515e-06, + "loss": 1.0084, + "step": 6196 + }, + { + "epoch": 0.5588672949452135, + "grad_norm": 1.3339244108560138, + "learning_rate": 1.7167371974917043e-06, + "loss": 1.0578, + "step": 6197 + }, + { + "epoch": 0.5589574784686838, + "grad_norm": 2.5470411562400814, + "learning_rate": 1.7161588881124003e-06, + "loss": 1.002, + "step": 6198 + }, + { + "epoch": 0.559047661992154, + "grad_norm": 5.086189322382732, + "learning_rate": 1.7155806029519861e-06, + "loss": 0.971, + "step": 6199 + }, + { + "epoch": 0.5591378455156243, + "grad_norm": 2.0200105020986565, + "learning_rate": 1.7150023420598023e-06, + "loss": 1.0283, + "step": 6200 + }, + { + "epoch": 0.5592280290390945, + "grad_norm": 1.4420087015362968, + "learning_rate": 1.714424105485191e-06, + "loss": 1.038, + "step": 6201 + }, + { + "epoch": 0.5593182125625649, + "grad_norm": 2.2315120895711273, + "learning_rate": 1.7138458932774896e-06, + "loss": 0.9483, + "step": 6202 + }, + { + "epoch": 0.5594083960860351, + "grad_norm": 0.659398744707009, + "learning_rate": 1.7132677054860335e-06, + "loss": 0.8444, + "step": 6203 + }, + { + "epoch": 0.5594985796095053, + "grad_norm": 1.5713834384530503, + "learning_rate": 1.7126895421601586e-06, + "loss": 1.01, + "step": 6204 + }, + { + "epoch": 0.5595887631329756, + "grad_norm": 1.4811829628240907, + "learning_rate": 1.712111403349196e-06, + "loss": 0.9258, + "step": 6205 + }, + { + "epoch": 0.5596789466564459, + "grad_norm": 1.4630141568962376, + "learning_rate": 1.7115332891024757e-06, + "loss": 1.0202, + "step": 6206 + }, + { + "epoch": 0.5597691301799161, + "grad_norm": 0.7191972584883277, + "learning_rate": 1.7109551994693257e-06, + "loss": 0.8938, + "step": 6207 + }, + { + "epoch": 0.5598593137033864, + "grad_norm": 1.3625774589008957, + "learning_rate": 1.7103771344990725e-06, + "loss": 0.9242, + "step": 6208 + }, + { + "epoch": 0.5599494972268566, + "grad_norm": 1.6362883882849717, + "learning_rate": 1.709799094241039e-06, + "loss": 0.8986, + "step": 6209 + }, + { + "epoch": 0.560039680750327, + "grad_norm": 5.452798303474914, + "learning_rate": 1.709221078744546e-06, + "loss": 1.0158, + "step": 6210 + }, + { + "epoch": 0.5601298642737972, + "grad_norm": 1.296520092671293, + "learning_rate": 1.7086430880589148e-06, + "loss": 0.9943, + "step": 6211 + }, + { + "epoch": 0.5602200477972674, + "grad_norm": 1.551555373068861, + "learning_rate": 1.7080651222334612e-06, + "loss": 0.9915, + "step": 6212 + }, + { + "epoch": 0.5603102313207377, + "grad_norm": 1.7771203204558421, + "learning_rate": 1.7074871813175018e-06, + "loss": 1.0448, + "step": 6213 + }, + { + "epoch": 0.560400414844208, + "grad_norm": 1.9779606875957354, + "learning_rate": 1.706909265360349e-06, + "loss": 0.8972, + "step": 6214 + }, + { + "epoch": 0.5604905983676782, + "grad_norm": 0.7020229179657104, + "learning_rate": 1.7063313744113128e-06, + "loss": 0.8331, + "step": 6215 + }, + { + "epoch": 0.5605807818911485, + "grad_norm": 0.6491845205673286, + "learning_rate": 1.7057535085197042e-06, + "loss": 0.807, + "step": 6216 + }, + { + "epoch": 0.5606709654146187, + "grad_norm": 2.0844452624510295, + "learning_rate": 1.705175667734828e-06, + "loss": 0.9036, + "step": 6217 + }, + { + "epoch": 0.560761148938089, + "grad_norm": 2.2855937832334874, + "learning_rate": 1.7045978521059894e-06, + "loss": 0.9627, + "step": 6218 + }, + { + "epoch": 0.5608513324615593, + "grad_norm": 1.2818436526627677, + "learning_rate": 1.7040200616824914e-06, + "loss": 0.9824, + "step": 6219 + }, + { + "epoch": 0.5609415159850295, + "grad_norm": 4.196072131385149, + "learning_rate": 1.7034422965136333e-06, + "loss": 0.9283, + "step": 6220 + }, + { + "epoch": 0.5610316995084998, + "grad_norm": 1.5242953501682925, + "learning_rate": 1.7028645566487137e-06, + "loss": 1.0244, + "step": 6221 + }, + { + "epoch": 0.5611218830319701, + "grad_norm": 1.5149064867978042, + "learning_rate": 1.7022868421370284e-06, + "loss": 0.9594, + "step": 6222 + }, + { + "epoch": 0.5612120665554403, + "grad_norm": 1.4758798141239433, + "learning_rate": 1.701709153027872e-06, + "loss": 1.03, + "step": 6223 + }, + { + "epoch": 0.5613022500789105, + "grad_norm": 1.5425745206912975, + "learning_rate": 1.7011314893705353e-06, + "loss": 0.9493, + "step": 6224 + }, + { + "epoch": 0.5613924336023809, + "grad_norm": 1.468703005424412, + "learning_rate": 1.700553851214307e-06, + "loss": 0.987, + "step": 6225 + }, + { + "epoch": 0.5614826171258511, + "grad_norm": 1.6118221011287988, + "learning_rate": 1.699976238608476e-06, + "loss": 0.85, + "step": 6226 + }, + { + "epoch": 0.5615728006493214, + "grad_norm": 1.78453256743115, + "learning_rate": 1.699398651602326e-06, + "loss": 0.931, + "step": 6227 + }, + { + "epoch": 0.5616629841727916, + "grad_norm": 3.3659644505600594, + "learning_rate": 1.6988210902451413e-06, + "loss": 1.0276, + "step": 6228 + }, + { + "epoch": 0.5617531676962619, + "grad_norm": 1.6984112103207696, + "learning_rate": 1.6982435545862011e-06, + "loss": 0.9601, + "step": 6229 + }, + { + "epoch": 0.5618433512197322, + "grad_norm": 1.6403065690325627, + "learning_rate": 1.6976660446747853e-06, + "loss": 0.9873, + "step": 6230 + }, + { + "epoch": 0.5619335347432024, + "grad_norm": 2.087783625174231, + "learning_rate": 1.6970885605601696e-06, + "loss": 0.9114, + "step": 6231 + }, + { + "epoch": 0.5620237182666726, + "grad_norm": 1.759485830065122, + "learning_rate": 1.6965111022916282e-06, + "loss": 1.0933, + "step": 6232 + }, + { + "epoch": 0.562113901790143, + "grad_norm": 1.6465042855177734, + "learning_rate": 1.6959336699184323e-06, + "loss": 0.9556, + "step": 6233 + }, + { + "epoch": 0.5622040853136132, + "grad_norm": 1.753926980080895, + "learning_rate": 1.6953562634898529e-06, + "loss": 0.9798, + "step": 6234 + }, + { + "epoch": 0.5622942688370834, + "grad_norm": 2.0244571664640514, + "learning_rate": 1.6947788830551569e-06, + "loss": 0.9524, + "step": 6235 + }, + { + "epoch": 0.5623844523605537, + "grad_norm": 0.7652600255369906, + "learning_rate": 1.6942015286636093e-06, + "loss": 0.8675, + "step": 6236 + }, + { + "epoch": 0.562474635884024, + "grad_norm": 1.654332184217706, + "learning_rate": 1.6936242003644735e-06, + "loss": 0.9672, + "step": 6237 + }, + { + "epoch": 0.5625648194074943, + "grad_norm": 1.7198143871091984, + "learning_rate": 1.6930468982070106e-06, + "loss": 1.0508, + "step": 6238 + }, + { + "epoch": 0.5626550029309645, + "grad_norm": 0.6406045538267904, + "learning_rate": 1.692469622240478e-06, + "loss": 0.7559, + "step": 6239 + }, + { + "epoch": 0.5627451864544347, + "grad_norm": 1.735735316824726, + "learning_rate": 1.6918923725141339e-06, + "loss": 0.9387, + "step": 6240 + }, + { + "epoch": 0.5628353699779051, + "grad_norm": 1.3576772326133126, + "learning_rate": 1.6913151490772312e-06, + "loss": 1.036, + "step": 6241 + }, + { + "epoch": 0.5629255535013753, + "grad_norm": 1.355541348230887, + "learning_rate": 1.6907379519790215e-06, + "loss": 1.0234, + "step": 6242 + }, + { + "epoch": 0.5630157370248455, + "grad_norm": 1.5239757662871378, + "learning_rate": 1.6901607812687558e-06, + "loss": 0.8897, + "step": 6243 + }, + { + "epoch": 0.5631059205483159, + "grad_norm": 1.5620834558296444, + "learning_rate": 1.6895836369956794e-06, + "loss": 0.937, + "step": 6244 + }, + { + "epoch": 0.5631961040717861, + "grad_norm": 0.665498683925276, + "learning_rate": 1.6890065192090402e-06, + "loss": 0.7797, + "step": 6245 + }, + { + "epoch": 0.5632862875952563, + "grad_norm": 1.6802059349363, + "learning_rate": 1.6884294279580793e-06, + "loss": 0.8807, + "step": 6246 + }, + { + "epoch": 0.5633764711187266, + "grad_norm": 2.1257942061921415, + "learning_rate": 1.6878523632920371e-06, + "loss": 1.0058, + "step": 6247 + }, + { + "epoch": 0.5634666546421969, + "grad_norm": 1.6021600864446834, + "learning_rate": 1.6872753252601525e-06, + "loss": 1.0456, + "step": 6248 + }, + { + "epoch": 0.5635568381656672, + "grad_norm": 1.4886938609469356, + "learning_rate": 1.6866983139116616e-06, + "loss": 0.9226, + "step": 6249 + }, + { + "epoch": 0.5636470216891374, + "grad_norm": 1.3697371332991903, + "learning_rate": 1.6861213292957981e-06, + "loss": 0.952, + "step": 6250 + }, + { + "epoch": 0.5637372052126076, + "grad_norm": 1.6033630670234438, + "learning_rate": 1.685544371461793e-06, + "loss": 1.0139, + "step": 6251 + }, + { + "epoch": 0.563827388736078, + "grad_norm": 1.6931822441058644, + "learning_rate": 1.6849674404588767e-06, + "loss": 1.01, + "step": 6252 + }, + { + "epoch": 0.5639175722595482, + "grad_norm": 1.828919197437551, + "learning_rate": 1.6843905363362758e-06, + "loss": 1.0162, + "step": 6253 + }, + { + "epoch": 0.5640077557830184, + "grad_norm": 1.4850648581986716, + "learning_rate": 1.6838136591432136e-06, + "loss": 0.9238, + "step": 6254 + }, + { + "epoch": 0.5640979393064887, + "grad_norm": 3.602051277672546, + "learning_rate": 1.6832368089289139e-06, + "loss": 1.0058, + "step": 6255 + }, + { + "epoch": 0.564188122829959, + "grad_norm": 1.509382705194934, + "learning_rate": 1.682659985742596e-06, + "loss": 0.9992, + "step": 6256 + }, + { + "epoch": 0.5642783063534292, + "grad_norm": 2.170397798140436, + "learning_rate": 1.6820831896334782e-06, + "loss": 1.0239, + "step": 6257 + }, + { + "epoch": 0.5643684898768995, + "grad_norm": 1.8567958863981568, + "learning_rate": 1.681506420650776e-06, + "loss": 1.0035, + "step": 6258 + }, + { + "epoch": 0.5644586734003697, + "grad_norm": 1.3058006882781268, + "learning_rate": 1.680929678843701e-06, + "loss": 0.9671, + "step": 6259 + }, + { + "epoch": 0.56454885692384, + "grad_norm": 1.5794028217143135, + "learning_rate": 1.6803529642614662e-06, + "loss": 1.0286, + "step": 6260 + }, + { + "epoch": 0.5646390404473103, + "grad_norm": 2.0572490317518195, + "learning_rate": 1.6797762769532785e-06, + "loss": 0.9864, + "step": 6261 + }, + { + "epoch": 0.5647292239707805, + "grad_norm": 1.3934799634097281, + "learning_rate": 1.679199616968345e-06, + "loss": 1.0158, + "step": 6262 + }, + { + "epoch": 0.5648194074942507, + "grad_norm": 1.50053594869168, + "learning_rate": 1.6786229843558689e-06, + "loss": 1.0106, + "step": 6263 + }, + { + "epoch": 0.5649095910177211, + "grad_norm": 2.5281603760006575, + "learning_rate": 1.6780463791650514e-06, + "loss": 0.9646, + "step": 6264 + }, + { + "epoch": 0.5649997745411913, + "grad_norm": 1.8123654282696848, + "learning_rate": 1.6774698014450928e-06, + "loss": 0.9247, + "step": 6265 + }, + { + "epoch": 0.5650899580646616, + "grad_norm": 0.6527475717558561, + "learning_rate": 1.6768932512451883e-06, + "loss": 0.8155, + "step": 6266 + }, + { + "epoch": 0.5651801415881319, + "grad_norm": 1.6535392746922672, + "learning_rate": 1.676316728614534e-06, + "loss": 0.9337, + "step": 6267 + }, + { + "epoch": 0.5652703251116021, + "grad_norm": 0.5808862522169058, + "learning_rate": 1.675740233602321e-06, + "loss": 0.826, + "step": 6268 + }, + { + "epoch": 0.5653605086350724, + "grad_norm": 1.4883445265727464, + "learning_rate": 1.6751637662577385e-06, + "loss": 1.028, + "step": 6269 + }, + { + "epoch": 0.5654506921585426, + "grad_norm": 0.8122436412446142, + "learning_rate": 1.6745873266299753e-06, + "loss": 0.7692, + "step": 6270 + }, + { + "epoch": 0.565540875682013, + "grad_norm": 1.3705047472565746, + "learning_rate": 1.6740109147682148e-06, + "loss": 0.9427, + "step": 6271 + }, + { + "epoch": 0.5656310592054832, + "grad_norm": 1.5782809038863348, + "learning_rate": 1.6734345307216418e-06, + "loss": 0.9973, + "step": 6272 + }, + { + "epoch": 0.5657212427289534, + "grad_norm": 1.8691557402283625, + "learning_rate": 1.6728581745394346e-06, + "loss": 0.9552, + "step": 6273 + }, + { + "epoch": 0.5658114262524236, + "grad_norm": 1.519781018595252, + "learning_rate": 1.672281846270772e-06, + "loss": 1.0547, + "step": 6274 + }, + { + "epoch": 0.565901609775894, + "grad_norm": 1.721120621328154, + "learning_rate": 1.6717055459648295e-06, + "loss": 0.9779, + "step": 6275 + }, + { + "epoch": 0.5659917932993642, + "grad_norm": 1.6519377369681516, + "learning_rate": 1.6711292736707793e-06, + "loss": 0.9338, + "step": 6276 + }, + { + "epoch": 0.5660819768228345, + "grad_norm": 7.6496877768635825, + "learning_rate": 1.6705530294377938e-06, + "loss": 1.0501, + "step": 6277 + }, + { + "epoch": 0.5661721603463047, + "grad_norm": 1.6002319763818191, + "learning_rate": 1.6699768133150395e-06, + "loss": 0.9733, + "step": 6278 + }, + { + "epoch": 0.566262343869775, + "grad_norm": 1.936690561351781, + "learning_rate": 1.6694006253516837e-06, + "loss": 0.972, + "step": 6279 + }, + { + "epoch": 0.5663525273932453, + "grad_norm": 1.786197550639595, + "learning_rate": 1.6688244655968896e-06, + "loss": 0.9772, + "step": 6280 + }, + { + "epoch": 0.5664427109167155, + "grad_norm": 1.9461606792516728, + "learning_rate": 1.6682483340998175e-06, + "loss": 0.9029, + "step": 6281 + }, + { + "epoch": 0.5665328944401857, + "grad_norm": 1.696310848080814, + "learning_rate": 1.6676722309096276e-06, + "loss": 0.9418, + "step": 6282 + }, + { + "epoch": 0.5666230779636561, + "grad_norm": 0.848365311335306, + "learning_rate": 1.6670961560754744e-06, + "loss": 0.8842, + "step": 6283 + }, + { + "epoch": 0.5667132614871263, + "grad_norm": 2.0363056356511584, + "learning_rate": 1.6665201096465138e-06, + "loss": 0.9788, + "step": 6284 + }, + { + "epoch": 0.5668034450105965, + "grad_norm": 1.980653893358423, + "learning_rate": 1.6659440916718961e-06, + "loss": 1.0256, + "step": 6285 + }, + { + "epoch": 0.5668936285340668, + "grad_norm": 1.6343051348426931, + "learning_rate": 1.6653681022007696e-06, + "loss": 0.9965, + "step": 6286 + }, + { + "epoch": 0.5669838120575371, + "grad_norm": 4.398743640822431, + "learning_rate": 1.6647921412822825e-06, + "loss": 0.9132, + "step": 6287 + }, + { + "epoch": 0.5670739955810074, + "grad_norm": 1.5407054671190474, + "learning_rate": 1.6642162089655782e-06, + "loss": 0.979, + "step": 6288 + }, + { + "epoch": 0.5671641791044776, + "grad_norm": 1.8055349894925494, + "learning_rate": 1.663640305299798e-06, + "loss": 0.9478, + "step": 6289 + }, + { + "epoch": 0.5672543626279479, + "grad_norm": 0.6775367294715616, + "learning_rate": 1.6630644303340824e-06, + "loss": 0.8467, + "step": 6290 + }, + { + "epoch": 0.5673445461514182, + "grad_norm": 2.786973015617969, + "learning_rate": 1.662488584117567e-06, + "loss": 0.9909, + "step": 6291 + }, + { + "epoch": 0.5674347296748884, + "grad_norm": 1.395057899340019, + "learning_rate": 1.6619127666993867e-06, + "loss": 1.0365, + "step": 6292 + }, + { + "epoch": 0.5675249131983586, + "grad_norm": 1.672939638935745, + "learning_rate": 1.6613369781286727e-06, + "loss": 0.9841, + "step": 6293 + }, + { + "epoch": 0.567615096721829, + "grad_norm": 1.711647130389777, + "learning_rate": 1.6607612184545562e-06, + "loss": 0.9215, + "step": 6294 + }, + { + "epoch": 0.5677052802452992, + "grad_norm": 1.701020927170395, + "learning_rate": 1.6601854877261617e-06, + "loss": 1.0511, + "step": 6295 + }, + { + "epoch": 0.5677954637687694, + "grad_norm": 1.8077227017084154, + "learning_rate": 1.6596097859926163e-06, + "loss": 0.9774, + "step": 6296 + }, + { + "epoch": 0.5678856472922397, + "grad_norm": 1.7983342092168382, + "learning_rate": 1.6590341133030407e-06, + "loss": 1.0712, + "step": 6297 + }, + { + "epoch": 0.56797583081571, + "grad_norm": 1.5593523746293938, + "learning_rate": 1.658458469706554e-06, + "loss": 1.0546, + "step": 6298 + }, + { + "epoch": 0.5680660143391802, + "grad_norm": 1.3714420930303701, + "learning_rate": 1.6578828552522746e-06, + "loss": 0.9358, + "step": 6299 + }, + { + "epoch": 0.5681561978626505, + "grad_norm": 1.6288710482885116, + "learning_rate": 1.6573072699893156e-06, + "loss": 1.0283, + "step": 6300 + }, + { + "epoch": 0.5682463813861207, + "grad_norm": 1.5689976439740962, + "learning_rate": 1.6567317139667906e-06, + "loss": 0.9536, + "step": 6301 + }, + { + "epoch": 0.5683365649095911, + "grad_norm": 1.825112558839112, + "learning_rate": 1.6561561872338087e-06, + "loss": 1.0, + "step": 6302 + }, + { + "epoch": 0.5684267484330613, + "grad_norm": 1.6523404313002281, + "learning_rate": 1.6555806898394764e-06, + "loss": 0.8926, + "step": 6303 + }, + { + "epoch": 0.5685169319565315, + "grad_norm": 1.667138350816068, + "learning_rate": 1.6550052218328987e-06, + "loss": 0.9583, + "step": 6304 + }, + { + "epoch": 0.5686071154800018, + "grad_norm": 1.6125993742373743, + "learning_rate": 1.6544297832631777e-06, + "loss": 0.9542, + "step": 6305 + }, + { + "epoch": 0.5686972990034721, + "grad_norm": 1.9142470290315532, + "learning_rate": 1.6538543741794135e-06, + "loss": 0.9885, + "step": 6306 + }, + { + "epoch": 0.5687874825269423, + "grad_norm": 1.789180768364157, + "learning_rate": 1.6532789946307028e-06, + "loss": 0.9517, + "step": 6307 + }, + { + "epoch": 0.5688776660504126, + "grad_norm": 0.6445419773018282, + "learning_rate": 1.6527036446661393e-06, + "loss": 0.8502, + "step": 6308 + }, + { + "epoch": 0.5689678495738828, + "grad_norm": 1.9826240323726168, + "learning_rate": 1.6521283243348165e-06, + "loss": 1.012, + "step": 6309 + }, + { + "epoch": 0.5690580330973531, + "grad_norm": 1.484690823335487, + "learning_rate": 1.6515530336858227e-06, + "loss": 0.9686, + "step": 6310 + }, + { + "epoch": 0.5691482166208234, + "grad_norm": 1.8145974618191638, + "learning_rate": 1.6509777727682457e-06, + "loss": 1.0271, + "step": 6311 + }, + { + "epoch": 0.5692384001442936, + "grad_norm": 1.5167739288771451, + "learning_rate": 1.65040254163117e-06, + "loss": 1.0203, + "step": 6312 + }, + { + "epoch": 0.569328583667764, + "grad_norm": 2.0255248154980086, + "learning_rate": 1.649827340323676e-06, + "loss": 0.8883, + "step": 6313 + }, + { + "epoch": 0.5694187671912342, + "grad_norm": 2.4293944743293245, + "learning_rate": 1.6492521688948454e-06, + "loss": 0.9439, + "step": 6314 + }, + { + "epoch": 0.5695089507147044, + "grad_norm": 1.9751246956199198, + "learning_rate": 1.6486770273937526e-06, + "loss": 1.0021, + "step": 6315 + }, + { + "epoch": 0.5695991342381747, + "grad_norm": 1.3931236108265552, + "learning_rate": 1.6481019158694738e-06, + "loss": 0.9613, + "step": 6316 + }, + { + "epoch": 0.569689317761645, + "grad_norm": 0.7621396198068604, + "learning_rate": 1.6475268343710792e-06, + "loss": 0.8891, + "step": 6317 + }, + { + "epoch": 0.5697795012851152, + "grad_norm": 1.621006460022806, + "learning_rate": 1.6469517829476396e-06, + "loss": 0.9134, + "step": 6318 + }, + { + "epoch": 0.5698696848085855, + "grad_norm": 1.9724637079224157, + "learning_rate": 1.64637676164822e-06, + "loss": 0.9652, + "step": 6319 + }, + { + "epoch": 0.5699598683320557, + "grad_norm": 1.4291145552466473, + "learning_rate": 1.6458017705218848e-06, + "loss": 1.0159, + "step": 6320 + }, + { + "epoch": 0.570050051855526, + "grad_norm": 2.042007949746375, + "learning_rate": 1.645226809617696e-06, + "loss": 1.0079, + "step": 6321 + }, + { + "epoch": 0.5701402353789963, + "grad_norm": 1.6641336345243725, + "learning_rate": 1.6446518789847112e-06, + "loss": 1.0168, + "step": 6322 + }, + { + "epoch": 0.5702304189024665, + "grad_norm": 1.9263426173357672, + "learning_rate": 1.6440769786719883e-06, + "loss": 1.0042, + "step": 6323 + }, + { + "epoch": 0.5703206024259367, + "grad_norm": 1.2756876811318534, + "learning_rate": 1.6435021087285803e-06, + "loss": 1.019, + "step": 6324 + }, + { + "epoch": 0.5704107859494071, + "grad_norm": 1.6205549332801292, + "learning_rate": 1.642927269203537e-06, + "loss": 1.0174, + "step": 6325 + }, + { + "epoch": 0.5705009694728773, + "grad_norm": 1.5217143167818339, + "learning_rate": 1.642352460145909e-06, + "loss": 1.0351, + "step": 6326 + }, + { + "epoch": 0.5705911529963476, + "grad_norm": 1.9730244812178859, + "learning_rate": 1.6417776816047402e-06, + "loss": 0.987, + "step": 6327 + }, + { + "epoch": 0.5706813365198178, + "grad_norm": 1.5300263166969241, + "learning_rate": 1.6412029336290755e-06, + "loss": 0.9684, + "step": 6328 + }, + { + "epoch": 0.5707715200432881, + "grad_norm": 1.9376343711280408, + "learning_rate": 1.6406282162679551e-06, + "loss": 0.9805, + "step": 6329 + }, + { + "epoch": 0.5708617035667584, + "grad_norm": 1.4839656064502706, + "learning_rate": 1.6400535295704162e-06, + "loss": 0.9517, + "step": 6330 + }, + { + "epoch": 0.5709518870902286, + "grad_norm": 1.5305301764927033, + "learning_rate": 1.6394788735854955e-06, + "loss": 0.9222, + "step": 6331 + }, + { + "epoch": 0.5710420706136988, + "grad_norm": 1.5327177214800558, + "learning_rate": 1.6389042483622246e-06, + "loss": 1.0524, + "step": 6332 + }, + { + "epoch": 0.5711322541371692, + "grad_norm": 1.8050226998420797, + "learning_rate": 1.638329653949635e-06, + "loss": 0.9786, + "step": 6333 + }, + { + "epoch": 0.5712224376606394, + "grad_norm": 1.4925896067852835, + "learning_rate": 1.637755090396753e-06, + "loss": 0.9449, + "step": 6334 + }, + { + "epoch": 0.5713126211841096, + "grad_norm": 0.8986291432196944, + "learning_rate": 1.6371805577526039e-06, + "loss": 0.8477, + "step": 6335 + }, + { + "epoch": 0.5714028047075799, + "grad_norm": 1.4913087823990205, + "learning_rate": 1.636606056066211e-06, + "loss": 0.9323, + "step": 6336 + }, + { + "epoch": 0.5714929882310502, + "grad_norm": 1.5535791941857395, + "learning_rate": 1.636031585386592e-06, + "loss": 0.9109, + "step": 6337 + }, + { + "epoch": 0.5715831717545204, + "grad_norm": 1.73235884318465, + "learning_rate": 1.635457145762766e-06, + "loss": 1.0451, + "step": 6338 + }, + { + "epoch": 0.5716733552779907, + "grad_norm": 2.217223918634514, + "learning_rate": 1.6348827372437456e-06, + "loss": 1.0131, + "step": 6339 + }, + { + "epoch": 0.571763538801461, + "grad_norm": 1.5352381715926082, + "learning_rate": 1.634308359878544e-06, + "loss": 0.9871, + "step": 6340 + }, + { + "epoch": 0.5718537223249313, + "grad_norm": 1.4986001476365662, + "learning_rate": 1.6337340137161695e-06, + "loss": 1.0607, + "step": 6341 + }, + { + "epoch": 0.5719439058484015, + "grad_norm": 1.557399129402463, + "learning_rate": 1.6331596988056277e-06, + "loss": 0.9852, + "step": 6342 + }, + { + "epoch": 0.5720340893718717, + "grad_norm": 1.503076656460316, + "learning_rate": 1.632585415195924e-06, + "loss": 1.0116, + "step": 6343 + }, + { + "epoch": 0.5721242728953421, + "grad_norm": 1.3855644472318354, + "learning_rate": 1.6320111629360583e-06, + "loss": 1.0097, + "step": 6344 + }, + { + "epoch": 0.5722144564188123, + "grad_norm": 1.8201854481378519, + "learning_rate": 1.631436942075029e-06, + "loss": 0.961, + "step": 6345 + }, + { + "epoch": 0.5723046399422825, + "grad_norm": 1.6167843081088904, + "learning_rate": 1.630862752661833e-06, + "loss": 0.9578, + "step": 6346 + }, + { + "epoch": 0.5723948234657528, + "grad_norm": 1.5243847371909072, + "learning_rate": 1.6302885947454612e-06, + "loss": 0.9123, + "step": 6347 + }, + { + "epoch": 0.5724850069892231, + "grad_norm": 1.8822643700967052, + "learning_rate": 1.6297144683749057e-06, + "loss": 0.9998, + "step": 6348 + }, + { + "epoch": 0.5725751905126933, + "grad_norm": 1.5876458649191951, + "learning_rate": 1.629140373599153e-06, + "loss": 0.9897, + "step": 6349 + }, + { + "epoch": 0.5726653740361636, + "grad_norm": 1.7381034845901815, + "learning_rate": 1.628566310467189e-06, + "loss": 0.9985, + "step": 6350 + }, + { + "epoch": 0.5727555575596338, + "grad_norm": 1.7027845129834398, + "learning_rate": 1.6279922790279957e-06, + "loss": 0.9504, + "step": 6351 + }, + { + "epoch": 0.5728457410831042, + "grad_norm": 1.5981817523783877, + "learning_rate": 1.6274182793305512e-06, + "loss": 0.9997, + "step": 6352 + }, + { + "epoch": 0.5729359246065744, + "grad_norm": 13.681211655703747, + "learning_rate": 1.626844311423835e-06, + "loss": 1.0145, + "step": 6353 + }, + { + "epoch": 0.5730261081300446, + "grad_norm": 1.8945204901635495, + "learning_rate": 1.6262703753568181e-06, + "loss": 1.0615, + "step": 6354 + }, + { + "epoch": 0.5731162916535149, + "grad_norm": 1.730707156462015, + "learning_rate": 1.6256964711784747e-06, + "loss": 1.0574, + "step": 6355 + }, + { + "epoch": 0.5732064751769852, + "grad_norm": 1.703319223510968, + "learning_rate": 1.6251225989377723e-06, + "loss": 0.9999, + "step": 6356 + }, + { + "epoch": 0.5732966587004554, + "grad_norm": 5.454155504852989, + "learning_rate": 1.624548758683676e-06, + "loss": 0.9881, + "step": 6357 + }, + { + "epoch": 0.5733868422239257, + "grad_norm": 1.5685263576901336, + "learning_rate": 1.6239749504651505e-06, + "loss": 1.0795, + "step": 6358 + }, + { + "epoch": 0.5734770257473959, + "grad_norm": 1.6131883645277236, + "learning_rate": 1.6234011743311552e-06, + "loss": 0.9615, + "step": 6359 + }, + { + "epoch": 0.5735672092708662, + "grad_norm": 1.460719862449945, + "learning_rate": 1.6228274303306483e-06, + "loss": 0.9165, + "step": 6360 + }, + { + "epoch": 0.5736573927943365, + "grad_norm": 1.4669654231244929, + "learning_rate": 1.6222537185125847e-06, + "loss": 1.0694, + "step": 6361 + }, + { + "epoch": 0.5737475763178067, + "grad_norm": 2.435621466943666, + "learning_rate": 1.6216800389259172e-06, + "loss": 1.0043, + "step": 6362 + }, + { + "epoch": 0.573837759841277, + "grad_norm": 1.495096058263297, + "learning_rate": 1.6211063916195949e-06, + "loss": 0.9926, + "step": 6363 + }, + { + "epoch": 0.5739279433647473, + "grad_norm": 1.5918132008844537, + "learning_rate": 1.6205327766425633e-06, + "loss": 0.9937, + "step": 6364 + }, + { + "epoch": 0.5740181268882175, + "grad_norm": 2.252510048254296, + "learning_rate": 1.6199591940437689e-06, + "loss": 0.9978, + "step": 6365 + }, + { + "epoch": 0.5741083104116878, + "grad_norm": 1.4670197054755816, + "learning_rate": 1.6193856438721505e-06, + "loss": 0.947, + "step": 6366 + }, + { + "epoch": 0.5741984939351581, + "grad_norm": 1.6660918277606682, + "learning_rate": 1.6188121261766483e-06, + "loss": 0.9915, + "step": 6367 + }, + { + "epoch": 0.5742886774586283, + "grad_norm": 1.388122922233665, + "learning_rate": 1.6182386410061976e-06, + "loss": 0.9178, + "step": 6368 + }, + { + "epoch": 0.5743788609820986, + "grad_norm": 1.8575762214270657, + "learning_rate": 1.61766518840973e-06, + "loss": 0.8998, + "step": 6369 + }, + { + "epoch": 0.5744690445055688, + "grad_norm": 1.514945989092936, + "learning_rate": 1.6170917684361779e-06, + "loss": 0.8595, + "step": 6370 + }, + { + "epoch": 0.5745592280290391, + "grad_norm": 1.5199447892853168, + "learning_rate": 1.6165183811344662e-06, + "loss": 1.0647, + "step": 6371 + }, + { + "epoch": 0.5746494115525094, + "grad_norm": 3.846739064223444, + "learning_rate": 1.6159450265535218e-06, + "loss": 0.9786, + "step": 6372 + }, + { + "epoch": 0.5747395950759796, + "grad_norm": 2.6967096660409515, + "learning_rate": 1.6153717047422652e-06, + "loss": 1.0733, + "step": 6373 + }, + { + "epoch": 0.5748297785994498, + "grad_norm": 1.9706525540928408, + "learning_rate": 1.6147984157496155e-06, + "loss": 0.8805, + "step": 6374 + }, + { + "epoch": 0.5749199621229202, + "grad_norm": 1.6052031285694395, + "learning_rate": 1.6142251596244886e-06, + "loss": 0.914, + "step": 6375 + }, + { + "epoch": 0.5750101456463904, + "grad_norm": 1.5605721596971889, + "learning_rate": 1.6136519364157983e-06, + "loss": 1.0406, + "step": 6376 + }, + { + "epoch": 0.5751003291698606, + "grad_norm": 1.9404215781316456, + "learning_rate": 1.6130787461724555e-06, + "loss": 0.928, + "step": 6377 + }, + { + "epoch": 0.5751905126933309, + "grad_norm": 1.3953298210957068, + "learning_rate": 1.6125055889433679e-06, + "loss": 1.0615, + "step": 6378 + }, + { + "epoch": 0.5752806962168012, + "grad_norm": 1.7038769942915555, + "learning_rate": 1.6119324647774386e-06, + "loss": 1.0134, + "step": 6379 + }, + { + "epoch": 0.5753708797402715, + "grad_norm": 1.7066126120222562, + "learning_rate": 1.6113593737235724e-06, + "loss": 0.995, + "step": 6380 + }, + { + "epoch": 0.5754610632637417, + "grad_norm": 1.5917533634534302, + "learning_rate": 1.6107863158306665e-06, + "loss": 0.9374, + "step": 6381 + }, + { + "epoch": 0.5755512467872119, + "grad_norm": 1.9889534823739865, + "learning_rate": 1.610213291147619e-06, + "loss": 0.8864, + "step": 6382 + }, + { + "epoch": 0.5756414303106823, + "grad_norm": 1.5053422685672777, + "learning_rate": 1.609640299723322e-06, + "loss": 1.01, + "step": 6383 + }, + { + "epoch": 0.5757316138341525, + "grad_norm": 1.3044688789717251, + "learning_rate": 1.609067341606668e-06, + "loss": 0.9712, + "step": 6384 + }, + { + "epoch": 0.5758217973576227, + "grad_norm": 3.165768161259173, + "learning_rate": 1.6084944168465438e-06, + "loss": 0.9793, + "step": 6385 + }, + { + "epoch": 0.5759119808810931, + "grad_norm": 1.5162837376500922, + "learning_rate": 1.6079215254918339e-06, + "loss": 1.0138, + "step": 6386 + }, + { + "epoch": 0.5760021644045633, + "grad_norm": 2.0687968914330206, + "learning_rate": 1.6073486675914222e-06, + "loss": 0.8926, + "step": 6387 + }, + { + "epoch": 0.5760923479280335, + "grad_norm": 1.8861259334707772, + "learning_rate": 1.606775843194187e-06, + "loss": 1.0156, + "step": 6388 + }, + { + "epoch": 0.5761825314515038, + "grad_norm": 1.8728473386001814, + "learning_rate": 1.6062030523490053e-06, + "loss": 0.9769, + "step": 6389 + }, + { + "epoch": 0.5762727149749741, + "grad_norm": 1.7821513120885883, + "learning_rate": 1.60563029510475e-06, + "loss": 0.9108, + "step": 6390 + }, + { + "epoch": 0.5763628984984444, + "grad_norm": 0.6441885297091038, + "learning_rate": 1.6050575715102927e-06, + "loss": 0.8063, + "step": 6391 + }, + { + "epoch": 0.5764530820219146, + "grad_norm": 1.7282218036820876, + "learning_rate": 1.6044848816145014e-06, + "loss": 1.0424, + "step": 6392 + }, + { + "epoch": 0.5765432655453848, + "grad_norm": 1.7114827580565566, + "learning_rate": 1.60391222546624e-06, + "loss": 1.0173, + "step": 6393 + }, + { + "epoch": 0.5766334490688552, + "grad_norm": 1.6630016401859713, + "learning_rate": 1.6033396031143725e-06, + "loss": 0.9815, + "step": 6394 + }, + { + "epoch": 0.5767236325923254, + "grad_norm": 1.5245525485372233, + "learning_rate": 1.602767014607757e-06, + "loss": 0.9883, + "step": 6395 + }, + { + "epoch": 0.5768138161157956, + "grad_norm": 2.4451107636620324, + "learning_rate": 1.6021944599952493e-06, + "loss": 0.9973, + "step": 6396 + }, + { + "epoch": 0.5769039996392659, + "grad_norm": 1.8721524550396234, + "learning_rate": 1.6016219393257048e-06, + "loss": 1.0012, + "step": 6397 + }, + { + "epoch": 0.5769941831627362, + "grad_norm": 1.6228775688949664, + "learning_rate": 1.6010494526479722e-06, + "loss": 0.9419, + "step": 6398 + }, + { + "epoch": 0.5770843666862064, + "grad_norm": 2.313277036694597, + "learning_rate": 1.6004770000109006e-06, + "loss": 0.9277, + "step": 6399 + }, + { + "epoch": 0.5771745502096767, + "grad_norm": 1.5505945972566473, + "learning_rate": 1.5999045814633348e-06, + "loss": 0.9543, + "step": 6400 + }, + { + "epoch": 0.5772647337331469, + "grad_norm": 1.5783917796435671, + "learning_rate": 1.5993321970541151e-06, + "loss": 1.0011, + "step": 6401 + }, + { + "epoch": 0.5773549172566173, + "grad_norm": 1.9032578670930886, + "learning_rate": 1.5987598468320825e-06, + "loss": 0.9884, + "step": 6402 + }, + { + "epoch": 0.5774451007800875, + "grad_norm": 2.245392107630758, + "learning_rate": 1.5981875308460717e-06, + "loss": 0.9759, + "step": 6403 + }, + { + "epoch": 0.5775352843035577, + "grad_norm": 1.7459179407829786, + "learning_rate": 1.5976152491449169e-06, + "loss": 0.9748, + "step": 6404 + }, + { + "epoch": 0.577625467827028, + "grad_norm": 2.078000057738141, + "learning_rate": 1.5970430017774468e-06, + "loss": 1.0485, + "step": 6405 + }, + { + "epoch": 0.5777156513504983, + "grad_norm": 1.5229558721349936, + "learning_rate": 1.5964707887924904e-06, + "loss": 1.0024, + "step": 6406 + }, + { + "epoch": 0.5778058348739685, + "grad_norm": 1.3084278058407461, + "learning_rate": 1.5958986102388714e-06, + "loss": 0.8731, + "step": 6407 + }, + { + "epoch": 0.5778960183974388, + "grad_norm": 1.5211525238077361, + "learning_rate": 1.5953264661654104e-06, + "loss": 0.9446, + "step": 6408 + }, + { + "epoch": 0.5779862019209091, + "grad_norm": 1.6344874064373724, + "learning_rate": 1.5947543566209276e-06, + "loss": 1.0218, + "step": 6409 + }, + { + "epoch": 0.5780763854443793, + "grad_norm": 1.6132162972368373, + "learning_rate": 1.5941822816542367e-06, + "loss": 0.9669, + "step": 6410 + }, + { + "epoch": 0.5781665689678496, + "grad_norm": 1.599667651706675, + "learning_rate": 1.5936102413141519e-06, + "loss": 0.9444, + "step": 6411 + }, + { + "epoch": 0.5782567524913198, + "grad_norm": 1.8585253785919855, + "learning_rate": 1.5930382356494823e-06, + "loss": 0.9081, + "step": 6412 + }, + { + "epoch": 0.5783469360147901, + "grad_norm": 1.4824601521532266, + "learning_rate": 1.5924662647090335e-06, + "loss": 1.0379, + "step": 6413 + }, + { + "epoch": 0.5784371195382604, + "grad_norm": 1.5016651447950322, + "learning_rate": 1.5918943285416108e-06, + "loss": 1.0037, + "step": 6414 + }, + { + "epoch": 0.5785273030617306, + "grad_norm": 1.333711610221911, + "learning_rate": 1.5913224271960139e-06, + "loss": 0.9224, + "step": 6415 + }, + { + "epoch": 0.5786174865852008, + "grad_norm": 1.6403336765857677, + "learning_rate": 1.590750560721041e-06, + "loss": 0.9961, + "step": 6416 + }, + { + "epoch": 0.5787076701086712, + "grad_norm": 1.5251652237235682, + "learning_rate": 1.5901787291654874e-06, + "loss": 0.9575, + "step": 6417 + }, + { + "epoch": 0.5787978536321414, + "grad_norm": 1.7158248284703306, + "learning_rate": 1.5896069325781435e-06, + "loss": 0.8599, + "step": 6418 + }, + { + "epoch": 0.5788880371556117, + "grad_norm": 1.788890580491766, + "learning_rate": 1.5890351710077998e-06, + "loss": 1.0221, + "step": 6419 + }, + { + "epoch": 0.5789782206790819, + "grad_norm": 2.4709267978670435, + "learning_rate": 1.5884634445032406e-06, + "loss": 1.042, + "step": 6420 + }, + { + "epoch": 0.5790684042025522, + "grad_norm": 0.6033471869314833, + "learning_rate": 1.5878917531132501e-06, + "loss": 0.8347, + "step": 6421 + }, + { + "epoch": 0.5791585877260225, + "grad_norm": 1.7795177536175535, + "learning_rate": 1.5873200968866077e-06, + "loss": 0.8582, + "step": 6422 + }, + { + "epoch": 0.5792487712494927, + "grad_norm": 4.883695623262554, + "learning_rate": 1.586748475872089e-06, + "loss": 0.957, + "step": 6423 + }, + { + "epoch": 0.5793389547729629, + "grad_norm": 1.8048393548927297, + "learning_rate": 1.58617689011847e-06, + "loss": 0.8559, + "step": 6424 + }, + { + "epoch": 0.5794291382964333, + "grad_norm": 1.5888392768945305, + "learning_rate": 1.5856053396745198e-06, + "loss": 1.0211, + "step": 6425 + }, + { + "epoch": 0.5795193218199035, + "grad_norm": 1.4009944194026351, + "learning_rate": 1.5850338245890078e-06, + "loss": 1.0516, + "step": 6426 + }, + { + "epoch": 0.5796095053433737, + "grad_norm": 2.6214280121425153, + "learning_rate": 1.5844623449106974e-06, + "loss": 0.9711, + "step": 6427 + }, + { + "epoch": 0.579699688866844, + "grad_norm": 1.4759147070764955, + "learning_rate": 1.583890900688351e-06, + "loss": 1.0291, + "step": 6428 + }, + { + "epoch": 0.5797898723903143, + "grad_norm": 2.1418655031081064, + "learning_rate": 1.583319491970728e-06, + "loss": 1.0623, + "step": 6429 + }, + { + "epoch": 0.5798800559137846, + "grad_norm": 1.4488216380459809, + "learning_rate": 1.5827481188065828e-06, + "loss": 1.0004, + "step": 6430 + }, + { + "epoch": 0.5799702394372548, + "grad_norm": 1.9084934103624385, + "learning_rate": 1.5821767812446689e-06, + "loss": 0.9552, + "step": 6431 + }, + { + "epoch": 0.5800604229607251, + "grad_norm": 1.948258524661295, + "learning_rate": 1.581605479333736e-06, + "loss": 1.1025, + "step": 6432 + }, + { + "epoch": 0.5801506064841954, + "grad_norm": 2.4271364219873384, + "learning_rate": 1.5810342131225308e-06, + "loss": 0.9888, + "step": 6433 + }, + { + "epoch": 0.5802407900076656, + "grad_norm": 1.5199636123718785, + "learning_rate": 1.580462982659797e-06, + "loss": 1.0258, + "step": 6434 + }, + { + "epoch": 0.5803309735311358, + "grad_norm": 1.7201411772608235, + "learning_rate": 1.5798917879942736e-06, + "loss": 1.057, + "step": 6435 + }, + { + "epoch": 0.5804211570546062, + "grad_norm": 1.371441223804262, + "learning_rate": 1.5793206291747006e-06, + "loss": 0.93, + "step": 6436 + }, + { + "epoch": 0.5805113405780764, + "grad_norm": 1.326931001592764, + "learning_rate": 1.57874950624981e-06, + "loss": 0.9229, + "step": 6437 + }, + { + "epoch": 0.5806015241015466, + "grad_norm": 1.6162943562073069, + "learning_rate": 1.5781784192683351e-06, + "loss": 0.9826, + "step": 6438 + }, + { + "epoch": 0.5806917076250169, + "grad_norm": 1.2317752743964216, + "learning_rate": 1.5776073682790033e-06, + "loss": 0.9479, + "step": 6439 + }, + { + "epoch": 0.5807818911484872, + "grad_norm": 1.7217099451618338, + "learning_rate": 1.5770363533305393e-06, + "loss": 1.0061, + "step": 6440 + }, + { + "epoch": 0.5808720746719575, + "grad_norm": 2.515295729797473, + "learning_rate": 1.5764653744716665e-06, + "loss": 0.9888, + "step": 6441 + }, + { + "epoch": 0.5809622581954277, + "grad_norm": 8.434362209518136, + "learning_rate": 1.575894431751103e-06, + "loss": 1.011, + "step": 6442 + }, + { + "epoch": 0.5810524417188979, + "grad_norm": 1.7007299959575537, + "learning_rate": 1.575323525217565e-06, + "loss": 1.0155, + "step": 6443 + }, + { + "epoch": 0.5811426252423683, + "grad_norm": 1.691458615380032, + "learning_rate": 1.574752654919766e-06, + "loss": 1.0709, + "step": 6444 + }, + { + "epoch": 0.5812328087658385, + "grad_norm": 1.481451267137423, + "learning_rate": 1.5741818209064146e-06, + "loss": 0.9588, + "step": 6445 + }, + { + "epoch": 0.5813229922893087, + "grad_norm": 1.8394217685530834, + "learning_rate": 1.5736110232262183e-06, + "loss": 0.9969, + "step": 6446 + }, + { + "epoch": 0.581413175812779, + "grad_norm": 1.8077819188596853, + "learning_rate": 1.5730402619278804e-06, + "loss": 1.0326, + "step": 6447 + }, + { + "epoch": 0.5815033593362493, + "grad_norm": 0.6544052262052056, + "learning_rate": 1.5724695370601024e-06, + "loss": 0.7934, + "step": 6448 + }, + { + "epoch": 0.5815935428597195, + "grad_norm": 1.456599358168218, + "learning_rate": 1.5718988486715798e-06, + "loss": 0.9087, + "step": 6449 + }, + { + "epoch": 0.5816837263831898, + "grad_norm": 1.7083880795646402, + "learning_rate": 1.5713281968110087e-06, + "loss": 0.9661, + "step": 6450 + }, + { + "epoch": 0.58177390990666, + "grad_norm": 1.6200288883035352, + "learning_rate": 1.5707575815270796e-06, + "loss": 0.9676, + "step": 6451 + }, + { + "epoch": 0.5818640934301303, + "grad_norm": 2.044725177271417, + "learning_rate": 1.57018700286848e-06, + "loss": 0.9838, + "step": 6452 + }, + { + "epoch": 0.5819542769536006, + "grad_norm": 1.6365842923309941, + "learning_rate": 1.5696164608838956e-06, + "loss": 1.0011, + "step": 6453 + }, + { + "epoch": 0.5820444604770708, + "grad_norm": 2.58301593512428, + "learning_rate": 1.5690459556220073e-06, + "loss": 1.0039, + "step": 6454 + }, + { + "epoch": 0.582134644000541, + "grad_norm": 1.8337011474051585, + "learning_rate": 1.5684754871314949e-06, + "loss": 1.0486, + "step": 6455 + }, + { + "epoch": 0.5822248275240114, + "grad_norm": 1.8587318397973958, + "learning_rate": 1.5679050554610335e-06, + "loss": 0.9868, + "step": 6456 + }, + { + "epoch": 0.5823150110474816, + "grad_norm": 1.3692552896717498, + "learning_rate": 1.567334660659295e-06, + "loss": 0.9904, + "step": 6457 + }, + { + "epoch": 0.5824051945709519, + "grad_norm": 1.3547344142441766, + "learning_rate": 1.5667643027749488e-06, + "loss": 1.0002, + "step": 6458 + }, + { + "epoch": 0.5824953780944222, + "grad_norm": 1.7793112125091253, + "learning_rate": 1.5661939818566614e-06, + "loss": 1.0575, + "step": 6459 + }, + { + "epoch": 0.5825855616178924, + "grad_norm": 1.748955142222938, + "learning_rate": 1.5656236979530956e-06, + "loss": 0.8535, + "step": 6460 + }, + { + "epoch": 0.5826757451413627, + "grad_norm": 1.755383113984401, + "learning_rate": 1.5650534511129106e-06, + "loss": 0.9306, + "step": 6461 + }, + { + "epoch": 0.5827659286648329, + "grad_norm": 1.7694171927705866, + "learning_rate": 1.5644832413847635e-06, + "loss": 0.9706, + "step": 6462 + }, + { + "epoch": 0.5828561121883032, + "grad_norm": 1.7210435995932962, + "learning_rate": 1.5639130688173082e-06, + "loss": 1.0304, + "step": 6463 + }, + { + "epoch": 0.5829462957117735, + "grad_norm": 1.6269259410478458, + "learning_rate": 1.5633429334591932e-06, + "loss": 0.9627, + "step": 6464 + }, + { + "epoch": 0.5830364792352437, + "grad_norm": 2.155973444355782, + "learning_rate": 1.562772835359068e-06, + "loss": 1.0218, + "step": 6465 + }, + { + "epoch": 0.5831266627587139, + "grad_norm": 3.0230949266920177, + "learning_rate": 1.5622027745655753e-06, + "loss": 0.9536, + "step": 6466 + }, + { + "epoch": 0.5832168462821843, + "grad_norm": 1.8017058078114783, + "learning_rate": 1.561632751127355e-06, + "loss": 0.9435, + "step": 6467 + }, + { + "epoch": 0.5833070298056545, + "grad_norm": 1.6813777492161248, + "learning_rate": 1.561062765093046e-06, + "loss": 1.031, + "step": 6468 + }, + { + "epoch": 0.5833972133291248, + "grad_norm": 2.4473374242822006, + "learning_rate": 1.5604928165112817e-06, + "loss": 0.9832, + "step": 6469 + }, + { + "epoch": 0.583487396852595, + "grad_norm": 1.5423196320517059, + "learning_rate": 1.5599229054306945e-06, + "loss": 0.9801, + "step": 6470 + }, + { + "epoch": 0.5835775803760653, + "grad_norm": 1.7393279963731065, + "learning_rate": 1.5593530318999111e-06, + "loss": 0.9812, + "step": 6471 + }, + { + "epoch": 0.5836677638995356, + "grad_norm": 1.6661396226432297, + "learning_rate": 1.5587831959675572e-06, + "loss": 1.0148, + "step": 6472 + }, + { + "epoch": 0.5837579474230058, + "grad_norm": 1.4874259513567905, + "learning_rate": 1.5582133976822534e-06, + "loss": 0.9737, + "step": 6473 + }, + { + "epoch": 0.583848130946476, + "grad_norm": 4.513842168060826, + "learning_rate": 1.5576436370926185e-06, + "loss": 0.9389, + "step": 6474 + }, + { + "epoch": 0.5839383144699464, + "grad_norm": 4.517900467788198, + "learning_rate": 1.5570739142472679e-06, + "loss": 0.9662, + "step": 6475 + }, + { + "epoch": 0.5840284979934166, + "grad_norm": 1.4834707617943244, + "learning_rate": 1.5565042291948127e-06, + "loss": 1.0521, + "step": 6476 + }, + { + "epoch": 0.5841186815168868, + "grad_norm": 1.4311920753984595, + "learning_rate": 1.5559345819838624e-06, + "loss": 1.022, + "step": 6477 + }, + { + "epoch": 0.5842088650403571, + "grad_norm": 1.3729084753661396, + "learning_rate": 1.5553649726630226e-06, + "loss": 0.9671, + "step": 6478 + }, + { + "epoch": 0.5842990485638274, + "grad_norm": 1.5085943695177715, + "learning_rate": 1.5547954012808942e-06, + "loss": 0.9168, + "step": 6479 + }, + { + "epoch": 0.5843892320872976, + "grad_norm": 1.8413362382115004, + "learning_rate": 1.5542258678860776e-06, + "loss": 0.9656, + "step": 6480 + }, + { + "epoch": 0.5844794156107679, + "grad_norm": 1.6065553575292792, + "learning_rate": 1.553656372527167e-06, + "loss": 1.1067, + "step": 6481 + }, + { + "epoch": 0.5845695991342382, + "grad_norm": 2.045458820608276, + "learning_rate": 1.5530869152527568e-06, + "loss": 1.0837, + "step": 6482 + }, + { + "epoch": 0.5846597826577085, + "grad_norm": 1.5361105401327595, + "learning_rate": 1.5525174961114353e-06, + "loss": 1.0069, + "step": 6483 + }, + { + "epoch": 0.5847499661811787, + "grad_norm": 1.554969110335408, + "learning_rate": 1.5519481151517875e-06, + "loss": 1.0253, + "step": 6484 + }, + { + "epoch": 0.5848401497046489, + "grad_norm": 1.8643871680133415, + "learning_rate": 1.551378772422398e-06, + "loss": 0.9902, + "step": 6485 + }, + { + "epoch": 0.5849303332281193, + "grad_norm": 1.5315605160136625, + "learning_rate": 1.5508094679718447e-06, + "loss": 0.9989, + "step": 6486 + }, + { + "epoch": 0.5850205167515895, + "grad_norm": 1.5928926592843835, + "learning_rate": 1.5502402018487048e-06, + "loss": 0.8921, + "step": 6487 + }, + { + "epoch": 0.5851107002750597, + "grad_norm": 2.61899588438849, + "learning_rate": 1.54967097410155e-06, + "loss": 0.9907, + "step": 6488 + }, + { + "epoch": 0.58520088379853, + "grad_norm": 1.4886189876021445, + "learning_rate": 1.5491017847789519e-06, + "loss": 0.9113, + "step": 6489 + }, + { + "epoch": 0.5852910673220003, + "grad_norm": 1.5028652799511395, + "learning_rate": 1.5485326339294755e-06, + "loss": 0.9305, + "step": 6490 + }, + { + "epoch": 0.5853812508454705, + "grad_norm": 2.0746747968525843, + "learning_rate": 1.5479635216016832e-06, + "loss": 0.9537, + "step": 6491 + }, + { + "epoch": 0.5854714343689408, + "grad_norm": 1.5860096581490881, + "learning_rate": 1.547394447844137e-06, + "loss": 0.9033, + "step": 6492 + }, + { + "epoch": 0.585561617892411, + "grad_norm": 1.5248299519520185, + "learning_rate": 1.546825412705391e-06, + "loss": 0.9017, + "step": 6493 + }, + { + "epoch": 0.5856518014158814, + "grad_norm": 1.830786959593731, + "learning_rate": 1.5462564162340007e-06, + "loss": 1.0173, + "step": 6494 + }, + { + "epoch": 0.5857419849393516, + "grad_norm": 1.805339546813208, + "learning_rate": 1.5456874584785144e-06, + "loss": 0.9917, + "step": 6495 + }, + { + "epoch": 0.5858321684628218, + "grad_norm": 1.9015412253425499, + "learning_rate": 1.5451185394874785e-06, + "loss": 0.9586, + "step": 6496 + }, + { + "epoch": 0.5859223519862921, + "grad_norm": 1.9529764347793515, + "learning_rate": 1.5445496593094381e-06, + "loss": 0.9954, + "step": 6497 + }, + { + "epoch": 0.5860125355097624, + "grad_norm": 1.6859232282343832, + "learning_rate": 1.5439808179929316e-06, + "loss": 0.998, + "step": 6498 + }, + { + "epoch": 0.5861027190332326, + "grad_norm": 2.5487625522875046, + "learning_rate": 1.543412015586496e-06, + "loss": 0.9489, + "step": 6499 + }, + { + "epoch": 0.5861929025567029, + "grad_norm": 1.798402286453253, + "learning_rate": 1.5428432521386655e-06, + "loss": 0.952, + "step": 6500 + }, + { + "epoch": 0.5862830860801731, + "grad_norm": 1.5821588158686792, + "learning_rate": 1.5422745276979688e-06, + "loss": 1.0192, + "step": 6501 + }, + { + "epoch": 0.5863732696036434, + "grad_norm": 1.4574036314105412, + "learning_rate": 1.5417058423129336e-06, + "loss": 0.9781, + "step": 6502 + }, + { + "epoch": 0.5864634531271137, + "grad_norm": 1.576842816387665, + "learning_rate": 1.5411371960320822e-06, + "loss": 1.0223, + "step": 6503 + }, + { + "epoch": 0.5865536366505839, + "grad_norm": 1.4738326092087395, + "learning_rate": 1.5405685889039363e-06, + "loss": 1.0181, + "step": 6504 + }, + { + "epoch": 0.5866438201740543, + "grad_norm": 1.8477905095619398, + "learning_rate": 1.5400000209770118e-06, + "loss": 0.9819, + "step": 6505 + }, + { + "epoch": 0.5867340036975245, + "grad_norm": 1.6718764260544443, + "learning_rate": 1.5394314922998208e-06, + "loss": 0.9969, + "step": 6506 + }, + { + "epoch": 0.5868241872209947, + "grad_norm": 2.2613568321115203, + "learning_rate": 1.5388630029208756e-06, + "loss": 1.0394, + "step": 6507 + }, + { + "epoch": 0.586914370744465, + "grad_norm": 1.5869101562451924, + "learning_rate": 1.5382945528886806e-06, + "loss": 1.0124, + "step": 6508 + }, + { + "epoch": 0.5870045542679353, + "grad_norm": 3.395392235189645, + "learning_rate": 1.5377261422517412e-06, + "loss": 0.9718, + "step": 6509 + }, + { + "epoch": 0.5870947377914055, + "grad_norm": 1.401258468398164, + "learning_rate": 1.5371577710585553e-06, + "loss": 0.9441, + "step": 6510 + }, + { + "epoch": 0.5871849213148758, + "grad_norm": 0.7339609074355706, + "learning_rate": 1.536589439357621e-06, + "loss": 0.8028, + "step": 6511 + }, + { + "epoch": 0.587275104838346, + "grad_norm": 1.4882832524644092, + "learning_rate": 1.5360211471974315e-06, + "loss": 0.9903, + "step": 6512 + }, + { + "epoch": 0.5873652883618163, + "grad_norm": 1.5512796749255229, + "learning_rate": 1.5354528946264753e-06, + "loss": 1.043, + "step": 6513 + }, + { + "epoch": 0.5874554718852866, + "grad_norm": 3.6011405250781126, + "learning_rate": 1.5348846816932399e-06, + "loss": 0.9204, + "step": 6514 + }, + { + "epoch": 0.5875456554087568, + "grad_norm": 0.6236993129982711, + "learning_rate": 1.5343165084462077e-06, + "loss": 0.8116, + "step": 6515 + }, + { + "epoch": 0.587635838932227, + "grad_norm": 1.7839043231976093, + "learning_rate": 1.5337483749338595e-06, + "loss": 0.9421, + "step": 6516 + }, + { + "epoch": 0.5877260224556974, + "grad_norm": 1.8446505012401302, + "learning_rate": 1.5331802812046708e-06, + "loss": 0.9598, + "step": 6517 + }, + { + "epoch": 0.5878162059791676, + "grad_norm": 1.7746236590231041, + "learning_rate": 1.5326122273071133e-06, + "loss": 0.8424, + "step": 6518 + }, + { + "epoch": 0.5879063895026378, + "grad_norm": 1.8379944661527228, + "learning_rate": 1.532044213289659e-06, + "loss": 1.0234, + "step": 6519 + }, + { + "epoch": 0.5879965730261081, + "grad_norm": 1.4928973042606533, + "learning_rate": 1.5314762392007718e-06, + "loss": 0.8756, + "step": 6520 + }, + { + "epoch": 0.5880867565495784, + "grad_norm": 1.827910647092875, + "learning_rate": 1.530908305088916e-06, + "loss": 1.0383, + "step": 6521 + }, + { + "epoch": 0.5881769400730487, + "grad_norm": 1.674253322940358, + "learning_rate": 1.5303404110025501e-06, + "loss": 0.9807, + "step": 6522 + }, + { + "epoch": 0.5882671235965189, + "grad_norm": 1.7194669441979193, + "learning_rate": 1.5297725569901293e-06, + "loss": 0.9421, + "step": 6523 + }, + { + "epoch": 0.5883573071199891, + "grad_norm": 1.7536351731462292, + "learning_rate": 1.5292047431001077e-06, + "loss": 0.8844, + "step": 6524 + }, + { + "epoch": 0.5884474906434595, + "grad_norm": 0.6743414430162069, + "learning_rate": 1.5286369693809321e-06, + "loss": 0.8262, + "step": 6525 + }, + { + "epoch": 0.5885376741669297, + "grad_norm": 1.6266610385974598, + "learning_rate": 1.5280692358810506e-06, + "loss": 1.0002, + "step": 6526 + }, + { + "epoch": 0.5886278576903999, + "grad_norm": 1.6643372787916768, + "learning_rate": 1.527501542648904e-06, + "loss": 0.9335, + "step": 6527 + }, + { + "epoch": 0.5887180412138703, + "grad_norm": 1.4422365190600328, + "learning_rate": 1.5269338897329308e-06, + "loss": 0.8901, + "step": 6528 + }, + { + "epoch": 0.5888082247373405, + "grad_norm": 1.5711671391915987, + "learning_rate": 1.5263662771815662e-06, + "loss": 1.0101, + "step": 6529 + }, + { + "epoch": 0.5888984082608107, + "grad_norm": 1.9561258684939253, + "learning_rate": 1.5257987050432429e-06, + "loss": 1.0581, + "step": 6530 + }, + { + "epoch": 0.588988591784281, + "grad_norm": 1.622187748561569, + "learning_rate": 1.5252311733663887e-06, + "loss": 0.8939, + "step": 6531 + }, + { + "epoch": 0.5890787753077513, + "grad_norm": 1.6622394013338757, + "learning_rate": 1.5246636821994281e-06, + "loss": 0.9272, + "step": 6532 + }, + { + "epoch": 0.5891689588312216, + "grad_norm": 2.2650007769181224, + "learning_rate": 1.524096231590784e-06, + "loss": 1.0294, + "step": 6533 + }, + { + "epoch": 0.5892591423546918, + "grad_norm": 1.765212559422759, + "learning_rate": 1.5235288215888736e-06, + "loss": 1.0337, + "step": 6534 + }, + { + "epoch": 0.589349325878162, + "grad_norm": 1.7312985878196046, + "learning_rate": 1.5229614522421102e-06, + "loss": 0.9899, + "step": 6535 + }, + { + "epoch": 0.5894395094016324, + "grad_norm": 1.6484496491332512, + "learning_rate": 1.5223941235989071e-06, + "loss": 0.8229, + "step": 6536 + }, + { + "epoch": 0.5895296929251026, + "grad_norm": 1.7222800330573615, + "learning_rate": 1.52182683570767e-06, + "loss": 0.9562, + "step": 6537 + }, + { + "epoch": 0.5896198764485728, + "grad_norm": 2.017063898839889, + "learning_rate": 1.5212595886168046e-06, + "loss": 1.0081, + "step": 6538 + }, + { + "epoch": 0.5897100599720431, + "grad_norm": 1.6131067066123446, + "learning_rate": 1.520692382374711e-06, + "loss": 0.9997, + "step": 6539 + }, + { + "epoch": 0.5898002434955134, + "grad_norm": 3.371990416127406, + "learning_rate": 1.5201252170297854e-06, + "loss": 1.0112, + "step": 6540 + }, + { + "epoch": 0.5898904270189836, + "grad_norm": 2.2153784696787175, + "learning_rate": 1.5195580926304232e-06, + "loss": 0.8711, + "step": 6541 + }, + { + "epoch": 0.5899806105424539, + "grad_norm": 7.503225777238013, + "learning_rate": 1.5189910092250131e-06, + "loss": 1.0057, + "step": 6542 + }, + { + "epoch": 0.5900707940659241, + "grad_norm": 2.4716338685068124, + "learning_rate": 1.5184239668619427e-06, + "loss": 0.9936, + "step": 6543 + }, + { + "epoch": 0.5901609775893945, + "grad_norm": 1.6296357275546904, + "learning_rate": 1.5178569655895946e-06, + "loss": 0.9871, + "step": 6544 + }, + { + "epoch": 0.5902511611128647, + "grad_norm": 1.5270622755041205, + "learning_rate": 1.5172900054563487e-06, + "loss": 1.0132, + "step": 6545 + }, + { + "epoch": 0.5903413446363349, + "grad_norm": 1.4718578005553193, + "learning_rate": 1.5167230865105814e-06, + "loss": 1.0051, + "step": 6546 + }, + { + "epoch": 0.5904315281598052, + "grad_norm": 1.767534826601935, + "learning_rate": 1.5161562088006644e-06, + "loss": 0.9008, + "step": 6547 + }, + { + "epoch": 0.5905217116832755, + "grad_norm": 2.0254664319528413, + "learning_rate": 1.5155893723749685e-06, + "loss": 1.0145, + "step": 6548 + }, + { + "epoch": 0.5906118952067457, + "grad_norm": 1.6374975277248676, + "learning_rate": 1.5150225772818582e-06, + "loss": 0.9917, + "step": 6549 + }, + { + "epoch": 0.590702078730216, + "grad_norm": 1.5527590191882745, + "learning_rate": 1.5144558235696949e-06, + "loss": 1.0358, + "step": 6550 + }, + { + "epoch": 0.5907922622536863, + "grad_norm": 2.4135680448440406, + "learning_rate": 1.5138891112868388e-06, + "loss": 0.9061, + "step": 6551 + }, + { + "epoch": 0.5908824457771565, + "grad_norm": 0.7006096884001233, + "learning_rate": 1.5133224404816433e-06, + "loss": 0.8159, + "step": 6552 + }, + { + "epoch": 0.5909726293006268, + "grad_norm": 1.7439236735606642, + "learning_rate": 1.5127558112024617e-06, + "loss": 0.9109, + "step": 6553 + }, + { + "epoch": 0.591062812824097, + "grad_norm": 0.6402276132446495, + "learning_rate": 1.5121892234976404e-06, + "loss": 0.8296, + "step": 6554 + }, + { + "epoch": 0.5911529963475673, + "grad_norm": 1.8857248029890343, + "learning_rate": 1.5116226774155243e-06, + "loss": 0.9696, + "step": 6555 + }, + { + "epoch": 0.5912431798710376, + "grad_norm": 2.1251841072661475, + "learning_rate": 1.5110561730044547e-06, + "loss": 0.9563, + "step": 6556 + }, + { + "epoch": 0.5913333633945078, + "grad_norm": 1.5068925808577533, + "learning_rate": 1.510489710312768e-06, + "loss": 0.9845, + "step": 6557 + }, + { + "epoch": 0.591423546917978, + "grad_norm": 1.7794149885115955, + "learning_rate": 1.5099232893887987e-06, + "loss": 0.9783, + "step": 6558 + }, + { + "epoch": 0.5915137304414484, + "grad_norm": 1.388961654452369, + "learning_rate": 1.5093569102808758e-06, + "loss": 0.9556, + "step": 6559 + }, + { + "epoch": 0.5916039139649186, + "grad_norm": 1.6287741481721105, + "learning_rate": 1.5087905730373275e-06, + "loss": 0.987, + "step": 6560 + }, + { + "epoch": 0.5916940974883889, + "grad_norm": 0.6223598028205791, + "learning_rate": 1.508224277706476e-06, + "loss": 0.8277, + "step": 6561 + }, + { + "epoch": 0.5917842810118591, + "grad_norm": 1.7186031625619806, + "learning_rate": 1.5076580243366399e-06, + "loss": 0.9399, + "step": 6562 + }, + { + "epoch": 0.5918744645353294, + "grad_norm": 1.9223529252443623, + "learning_rate": 1.507091812976137e-06, + "loss": 0.8936, + "step": 6563 + }, + { + "epoch": 0.5919646480587997, + "grad_norm": 1.7384656486858654, + "learning_rate": 1.5065256436732773e-06, + "loss": 0.9318, + "step": 6564 + }, + { + "epoch": 0.5920548315822699, + "grad_norm": 1.8771920424244946, + "learning_rate": 1.5059595164763717e-06, + "loss": 0.9144, + "step": 6565 + }, + { + "epoch": 0.5921450151057401, + "grad_norm": 0.571847026953774, + "learning_rate": 1.5053934314337243e-06, + "loss": 0.8226, + "step": 6566 + }, + { + "epoch": 0.5922351986292105, + "grad_norm": 1.6880913863777687, + "learning_rate": 1.5048273885936356e-06, + "loss": 0.9726, + "step": 6567 + }, + { + "epoch": 0.5923253821526807, + "grad_norm": 1.393151891353851, + "learning_rate": 1.5042613880044053e-06, + "loss": 0.9017, + "step": 6568 + }, + { + "epoch": 0.592415565676151, + "grad_norm": 1.6987729244074827, + "learning_rate": 1.5036954297143265e-06, + "loss": 1.0095, + "step": 6569 + }, + { + "epoch": 0.5925057491996212, + "grad_norm": 1.1985114680295776, + "learning_rate": 1.50312951377169e-06, + "loss": 0.9403, + "step": 6570 + }, + { + "epoch": 0.5925959327230915, + "grad_norm": 0.5840706139618831, + "learning_rate": 1.502563640224784e-06, + "loss": 0.8249, + "step": 6571 + }, + { + "epoch": 0.5926861162465618, + "grad_norm": 1.6755017312000224, + "learning_rate": 1.5019978091218903e-06, + "loss": 0.9814, + "step": 6572 + }, + { + "epoch": 0.592776299770032, + "grad_norm": 1.3717453323988413, + "learning_rate": 1.50143202051129e-06, + "loss": 1.0458, + "step": 6573 + }, + { + "epoch": 0.5928664832935022, + "grad_norm": 1.646833679144104, + "learning_rate": 1.500866274441258e-06, + "loss": 1.0457, + "step": 6574 + }, + { + "epoch": 0.5929566668169726, + "grad_norm": 1.7601419569333403, + "learning_rate": 1.5003005709600682e-06, + "loss": 0.9663, + "step": 6575 + }, + { + "epoch": 0.5930468503404428, + "grad_norm": 1.750141955476426, + "learning_rate": 1.4997349101159885e-06, + "loss": 1.0627, + "step": 6576 + }, + { + "epoch": 0.593137033863913, + "grad_norm": 1.870437283922441, + "learning_rate": 1.4991692919572854e-06, + "loss": 0.9397, + "step": 6577 + }, + { + "epoch": 0.5932272173873834, + "grad_norm": 1.8541529437068065, + "learning_rate": 1.4986037165322199e-06, + "loss": 0.9969, + "step": 6578 + }, + { + "epoch": 0.5933174009108536, + "grad_norm": 0.6687892091255301, + "learning_rate": 1.498038183889049e-06, + "loss": 0.8535, + "step": 6579 + }, + { + "epoch": 0.5934075844343238, + "grad_norm": 1.7351814663225675, + "learning_rate": 1.4974726940760292e-06, + "loss": 0.9022, + "step": 6580 + }, + { + "epoch": 0.5934977679577941, + "grad_norm": 1.557893753611036, + "learning_rate": 1.496907247141409e-06, + "loss": 1.0583, + "step": 6581 + }, + { + "epoch": 0.5935879514812644, + "grad_norm": 0.7471053769727569, + "learning_rate": 1.4963418431334372e-06, + "loss": 0.8937, + "step": 6582 + }, + { + "epoch": 0.5936781350047347, + "grad_norm": 1.816885804891198, + "learning_rate": 1.4957764821003566e-06, + "loss": 0.9677, + "step": 6583 + }, + { + "epoch": 0.5937683185282049, + "grad_norm": 4.849962741669368, + "learning_rate": 1.4952111640904063e-06, + "loss": 0.9873, + "step": 6584 + }, + { + "epoch": 0.5938585020516751, + "grad_norm": 1.8318160526470646, + "learning_rate": 1.494645889151823e-06, + "loss": 0.9701, + "step": 6585 + }, + { + "epoch": 0.5939486855751455, + "grad_norm": 0.6574317872449903, + "learning_rate": 1.494080657332839e-06, + "loss": 0.7989, + "step": 6586 + }, + { + "epoch": 0.5940388690986157, + "grad_norm": 1.7663569325813506, + "learning_rate": 1.4935154686816832e-06, + "loss": 0.9396, + "step": 6587 + }, + { + "epoch": 0.5941290526220859, + "grad_norm": 1.8696849512729683, + "learning_rate": 1.4929503232465802e-06, + "loss": 1.0604, + "step": 6588 + }, + { + "epoch": 0.5942192361455562, + "grad_norm": 1.406782134358014, + "learning_rate": 1.492385221075751e-06, + "loss": 0.893, + "step": 6589 + }, + { + "epoch": 0.5943094196690265, + "grad_norm": 1.7784972084544712, + "learning_rate": 1.4918201622174142e-06, + "loss": 1.0165, + "step": 6590 + }, + { + "epoch": 0.5943996031924967, + "grad_norm": 2.161188082671293, + "learning_rate": 1.4912551467197827e-06, + "loss": 1.0041, + "step": 6591 + }, + { + "epoch": 0.594489786715967, + "grad_norm": 1.8502939763919315, + "learning_rate": 1.4906901746310678e-06, + "loss": 1.0668, + "step": 6592 + }, + { + "epoch": 0.5945799702394372, + "grad_norm": 1.264370331691962, + "learning_rate": 1.4901252459994757e-06, + "loss": 0.943, + "step": 6593 + }, + { + "epoch": 0.5946701537629075, + "grad_norm": 2.2744052570675404, + "learning_rate": 1.489560360873208e-06, + "loss": 0.9696, + "step": 6594 + }, + { + "epoch": 0.5947603372863778, + "grad_norm": 1.6527817598752617, + "learning_rate": 1.4889955193004659e-06, + "loss": 0.9607, + "step": 6595 + }, + { + "epoch": 0.594850520809848, + "grad_norm": 0.6327242848639411, + "learning_rate": 1.4884307213294428e-06, + "loss": 0.8123, + "step": 6596 + }, + { + "epoch": 0.5949407043333182, + "grad_norm": 1.9910953295468763, + "learning_rate": 1.4878659670083321e-06, + "loss": 0.9708, + "step": 6597 + }, + { + "epoch": 0.5950308878567886, + "grad_norm": 1.5742933383432636, + "learning_rate": 1.4873012563853208e-06, + "loss": 0.9045, + "step": 6598 + }, + { + "epoch": 0.5951210713802588, + "grad_norm": 2.553769848000002, + "learning_rate": 1.4867365895085935e-06, + "loss": 0.9911, + "step": 6599 + }, + { + "epoch": 0.5952112549037291, + "grad_norm": 1.398256897585856, + "learning_rate": 1.4861719664263301e-06, + "loss": 0.9631, + "step": 6600 + }, + { + "epoch": 0.5953014384271994, + "grad_norm": 1.802830719362301, + "learning_rate": 1.485607387186708e-06, + "loss": 1.085, + "step": 6601 + }, + { + "epoch": 0.5953916219506696, + "grad_norm": 1.8000753016085693, + "learning_rate": 1.4850428518379001e-06, + "loss": 0.9808, + "step": 6602 + }, + { + "epoch": 0.5954818054741399, + "grad_norm": 1.4009435352243944, + "learning_rate": 1.4844783604280746e-06, + "loss": 0.9757, + "step": 6603 + }, + { + "epoch": 0.5955719889976101, + "grad_norm": 2.0816680992290535, + "learning_rate": 1.483913913005399e-06, + "loss": 0.9232, + "step": 6604 + }, + { + "epoch": 0.5956621725210804, + "grad_norm": 2.1411328235460796, + "learning_rate": 1.483349509618034e-06, + "loss": 1.0433, + "step": 6605 + }, + { + "epoch": 0.5957523560445507, + "grad_norm": 1.5489538076813938, + "learning_rate": 1.4827851503141367e-06, + "loss": 0.915, + "step": 6606 + }, + { + "epoch": 0.5958425395680209, + "grad_norm": 1.5321145244650154, + "learning_rate": 1.482220835141863e-06, + "loss": 0.9085, + "step": 6607 + }, + { + "epoch": 0.5959327230914911, + "grad_norm": 1.9419581897057376, + "learning_rate": 1.481656564149362e-06, + "loss": 1.0599, + "step": 6608 + }, + { + "epoch": 0.5960229066149615, + "grad_norm": 3.9813692611006712, + "learning_rate": 1.4810923373847818e-06, + "loss": 0.9771, + "step": 6609 + }, + { + "epoch": 0.5961130901384317, + "grad_norm": 1.6811446855311338, + "learning_rate": 1.4805281548962647e-06, + "loss": 0.9887, + "step": 6610 + }, + { + "epoch": 0.596203273661902, + "grad_norm": 2.204047692975394, + "learning_rate": 1.4799640167319488e-06, + "loss": 0.9272, + "step": 6611 + }, + { + "epoch": 0.5962934571853722, + "grad_norm": 1.5060554505382016, + "learning_rate": 1.4793999229399714e-06, + "loss": 0.9665, + "step": 6612 + }, + { + "epoch": 0.5963836407088425, + "grad_norm": 1.8516371184817582, + "learning_rate": 1.4788358735684626e-06, + "loss": 1.0461, + "step": 6613 + }, + { + "epoch": 0.5964738242323128, + "grad_norm": 1.4972914241933708, + "learning_rate": 1.4782718686655514e-06, + "loss": 1.0512, + "step": 6614 + }, + { + "epoch": 0.596564007755783, + "grad_norm": 2.02612538561081, + "learning_rate": 1.4777079082793605e-06, + "loss": 1.0353, + "step": 6615 + }, + { + "epoch": 0.5966541912792532, + "grad_norm": 1.8050017640588651, + "learning_rate": 1.4771439924580108e-06, + "loss": 0.9503, + "step": 6616 + }, + { + "epoch": 0.5967443748027236, + "grad_norm": 1.4315173836364405, + "learning_rate": 1.4765801212496189e-06, + "loss": 1.0258, + "step": 6617 + }, + { + "epoch": 0.5968345583261938, + "grad_norm": 1.9524401265047948, + "learning_rate": 1.4760162947022968e-06, + "loss": 0.8985, + "step": 6618 + }, + { + "epoch": 0.596924741849664, + "grad_norm": 1.5930997232472452, + "learning_rate": 1.475452512864154e-06, + "loss": 0.9732, + "step": 6619 + }, + { + "epoch": 0.5970149253731343, + "grad_norm": 1.3542759288821766, + "learning_rate": 1.4748887757832945e-06, + "loss": 0.9928, + "step": 6620 + }, + { + "epoch": 0.5971051088966046, + "grad_norm": 4.116821272774393, + "learning_rate": 1.4743250835078209e-06, + "loss": 1.0159, + "step": 6621 + }, + { + "epoch": 0.5971952924200749, + "grad_norm": 1.3675771975303017, + "learning_rate": 1.4737614360858297e-06, + "loss": 0.9436, + "step": 6622 + }, + { + "epoch": 0.5972854759435451, + "grad_norm": 2.2191429327529915, + "learning_rate": 1.4731978335654133e-06, + "loss": 0.9404, + "step": 6623 + }, + { + "epoch": 0.5973756594670154, + "grad_norm": 2.009431890984518, + "learning_rate": 1.4726342759946638e-06, + "loss": 0.9219, + "step": 6624 + }, + { + "epoch": 0.5974658429904857, + "grad_norm": 1.8661751656370127, + "learning_rate": 1.4720707634216653e-06, + "loss": 0.98, + "step": 6625 + }, + { + "epoch": 0.5975560265139559, + "grad_norm": 2.578982863607048, + "learning_rate": 1.4715072958945e-06, + "loss": 0.9941, + "step": 6626 + }, + { + "epoch": 0.5976462100374261, + "grad_norm": 2.1257882619655923, + "learning_rate": 1.470943873461247e-06, + "loss": 0.9802, + "step": 6627 + }, + { + "epoch": 0.5977363935608965, + "grad_norm": 2.069120935105485, + "learning_rate": 1.470380496169979e-06, + "loss": 0.9422, + "step": 6628 + }, + { + "epoch": 0.5978265770843667, + "grad_norm": 1.716332764543528, + "learning_rate": 1.4698171640687682e-06, + "loss": 0.9738, + "step": 6629 + }, + { + "epoch": 0.5979167606078369, + "grad_norm": 1.4917476465134094, + "learning_rate": 1.4692538772056792e-06, + "loss": 0.9173, + "step": 6630 + }, + { + "epoch": 0.5980069441313072, + "grad_norm": 2.279959640313025, + "learning_rate": 1.4686906356287772e-06, + "loss": 0.896, + "step": 6631 + }, + { + "epoch": 0.5980971276547775, + "grad_norm": 2.070107734198625, + "learning_rate": 1.4681274393861194e-06, + "loss": 1.0118, + "step": 6632 + }, + { + "epoch": 0.5981873111782477, + "grad_norm": 1.5575685882161596, + "learning_rate": 1.4675642885257603e-06, + "loss": 1.0782, + "step": 6633 + }, + { + "epoch": 0.598277494701718, + "grad_norm": 1.6667066887182824, + "learning_rate": 1.4670011830957529e-06, + "loss": 1.0687, + "step": 6634 + }, + { + "epoch": 0.5983676782251882, + "grad_norm": 1.6875201330043574, + "learning_rate": 1.4664381231441427e-06, + "loss": 0.9474, + "step": 6635 + }, + { + "epoch": 0.5984578617486586, + "grad_norm": 2.977083415909471, + "learning_rate": 1.4658751087189746e-06, + "loss": 0.9892, + "step": 6636 + }, + { + "epoch": 0.5985480452721288, + "grad_norm": 1.730579381295351, + "learning_rate": 1.4653121398682874e-06, + "loss": 1.0322, + "step": 6637 + }, + { + "epoch": 0.598638228795599, + "grad_norm": 1.2885273602903657, + "learning_rate": 1.4647492166401159e-06, + "loss": 0.9207, + "step": 6638 + }, + { + "epoch": 0.5987284123190693, + "grad_norm": 2.014646070648149, + "learning_rate": 1.4641863390824934e-06, + "loss": 0.9689, + "step": 6639 + }, + { + "epoch": 0.5988185958425396, + "grad_norm": 1.7321779924517513, + "learning_rate": 1.4636235072434465e-06, + "loss": 0.9733, + "step": 6640 + }, + { + "epoch": 0.5989087793660098, + "grad_norm": 1.411094797011726, + "learning_rate": 1.4630607211709994e-06, + "loss": 1.0073, + "step": 6641 + }, + { + "epoch": 0.5989989628894801, + "grad_norm": 1.5854565121242392, + "learning_rate": 1.4624979809131723e-06, + "loss": 0.9901, + "step": 6642 + }, + { + "epoch": 0.5990891464129503, + "grad_norm": 1.5589688645606607, + "learning_rate": 1.4619352865179814e-06, + "loss": 0.8916, + "step": 6643 + }, + { + "epoch": 0.5991793299364206, + "grad_norm": 1.647671345207049, + "learning_rate": 1.4613726380334391e-06, + "loss": 0.9559, + "step": 6644 + }, + { + "epoch": 0.5992695134598909, + "grad_norm": 1.5977159409692774, + "learning_rate": 1.4608100355075522e-06, + "loss": 0.8329, + "step": 6645 + }, + { + "epoch": 0.5993596969833611, + "grad_norm": 0.6269307830727577, + "learning_rate": 1.460247478988327e-06, + "loss": 0.7936, + "step": 6646 + }, + { + "epoch": 0.5994498805068315, + "grad_norm": 1.5377458081263484, + "learning_rate": 1.4596849685237623e-06, + "loss": 1.0003, + "step": 6647 + }, + { + "epoch": 0.5995400640303017, + "grad_norm": 1.6937062613269762, + "learning_rate": 1.459122504161856e-06, + "loss": 0.8889, + "step": 6648 + }, + { + "epoch": 0.5996302475537719, + "grad_norm": 1.6285063254936567, + "learning_rate": 1.4585600859506001e-06, + "loss": 0.9552, + "step": 6649 + }, + { + "epoch": 0.5997204310772422, + "grad_norm": 1.5169856462648257, + "learning_rate": 1.4579977139379826e-06, + "loss": 1.0383, + "step": 6650 + }, + { + "epoch": 0.5998106146007125, + "grad_norm": 10.255229081339051, + "learning_rate": 1.4574353881719895e-06, + "loss": 1.0481, + "step": 6651 + }, + { + "epoch": 0.5999007981241827, + "grad_norm": 1.6208152439837198, + "learning_rate": 1.4568731087005998e-06, + "loss": 0.9857, + "step": 6652 + }, + { + "epoch": 0.599990981647653, + "grad_norm": 2.6475894436662823, + "learning_rate": 1.4563108755717916e-06, + "loss": 1.0085, + "step": 6653 + }, + { + "epoch": 0.6000811651711232, + "grad_norm": 1.6960837723672653, + "learning_rate": 1.455748688833538e-06, + "loss": 1.009, + "step": 6654 + }, + { + "epoch": 0.6001713486945935, + "grad_norm": 1.6623597365560001, + "learning_rate": 1.4551865485338065e-06, + "loss": 0.8918, + "step": 6655 + }, + { + "epoch": 0.6002615322180638, + "grad_norm": 2.101989929523536, + "learning_rate": 1.4546244547205629e-06, + "loss": 1.0461, + "step": 6656 + }, + { + "epoch": 0.600351715741534, + "grad_norm": 1.6991790416067964, + "learning_rate": 1.4540624074417678e-06, + "loss": 1.0061, + "step": 6657 + }, + { + "epoch": 0.6004418992650042, + "grad_norm": 1.4045272022816637, + "learning_rate": 1.453500406745379e-06, + "loss": 1.0402, + "step": 6658 + }, + { + "epoch": 0.6005320827884746, + "grad_norm": 1.9111473461099462, + "learning_rate": 1.4529384526793486e-06, + "loss": 1.0566, + "step": 6659 + }, + { + "epoch": 0.6006222663119448, + "grad_norm": 1.6653689418454336, + "learning_rate": 1.4523765452916252e-06, + "loss": 1.039, + "step": 6660 + }, + { + "epoch": 0.600712449835415, + "grad_norm": 1.7174481403234036, + "learning_rate": 1.4518146846301554e-06, + "loss": 1.0129, + "step": 6661 + }, + { + "epoch": 0.6008026333588853, + "grad_norm": 1.4212469034152826, + "learning_rate": 1.4512528707428787e-06, + "loss": 0.9586, + "step": 6662 + }, + { + "epoch": 0.6008928168823556, + "grad_norm": 1.4587650568298625, + "learning_rate": 1.4506911036777335e-06, + "loss": 0.9634, + "step": 6663 + }, + { + "epoch": 0.6009830004058259, + "grad_norm": 0.6788078623111775, + "learning_rate": 1.450129383482651e-06, + "loss": 0.7852, + "step": 6664 + }, + { + "epoch": 0.6010731839292961, + "grad_norm": 1.8544467400179252, + "learning_rate": 1.4495677102055629e-06, + "loss": 0.9365, + "step": 6665 + }, + { + "epoch": 0.6011633674527663, + "grad_norm": 2.0803403590570184, + "learning_rate": 1.4490060838943924e-06, + "loss": 1.0198, + "step": 6666 + }, + { + "epoch": 0.6012535509762367, + "grad_norm": 1.44536957499215, + "learning_rate": 1.4484445045970609e-06, + "loss": 0.9853, + "step": 6667 + }, + { + "epoch": 0.6013437344997069, + "grad_norm": 1.698667660500811, + "learning_rate": 1.447882972361485e-06, + "loss": 0.9811, + "step": 6668 + }, + { + "epoch": 0.6014339180231771, + "grad_norm": 3.393369627373139, + "learning_rate": 1.4473214872355785e-06, + "loss": 0.9666, + "step": 6669 + }, + { + "epoch": 0.6015241015466475, + "grad_norm": 1.904214992229948, + "learning_rate": 1.4467600492672508e-06, + "loss": 1.0208, + "step": 6670 + }, + { + "epoch": 0.6016142850701177, + "grad_norm": 1.4953846657936531, + "learning_rate": 1.4461986585044054e-06, + "loss": 0.9625, + "step": 6671 + }, + { + "epoch": 0.601704468593588, + "grad_norm": 1.5543772325549565, + "learning_rate": 1.4456373149949446e-06, + "loss": 0.9568, + "step": 6672 + }, + { + "epoch": 0.6017946521170582, + "grad_norm": 1.6278074661443616, + "learning_rate": 1.4450760187867648e-06, + "loss": 0.9908, + "step": 6673 + }, + { + "epoch": 0.6018848356405285, + "grad_norm": 1.5450788917275349, + "learning_rate": 1.4445147699277581e-06, + "loss": 0.9741, + "step": 6674 + }, + { + "epoch": 0.6019750191639988, + "grad_norm": 1.6248107946725665, + "learning_rate": 1.4439535684658154e-06, + "loss": 1.0299, + "step": 6675 + }, + { + "epoch": 0.602065202687469, + "grad_norm": 0.7770271998594654, + "learning_rate": 1.44339241444882e-06, + "loss": 0.8548, + "step": 6676 + }, + { + "epoch": 0.6021553862109392, + "grad_norm": 2.047523686943863, + "learning_rate": 1.4428313079246518e-06, + "loss": 0.9325, + "step": 6677 + }, + { + "epoch": 0.6022455697344096, + "grad_norm": 0.6665803858435958, + "learning_rate": 1.4422702489411896e-06, + "loss": 0.8327, + "step": 6678 + }, + { + "epoch": 0.6023357532578798, + "grad_norm": 1.7370045291663698, + "learning_rate": 1.4417092375463043e-06, + "loss": 0.928, + "step": 6679 + }, + { + "epoch": 0.60242593678135, + "grad_norm": 1.6101498266366827, + "learning_rate": 1.441148273787866e-06, + "loss": 1.02, + "step": 6680 + }, + { + "epoch": 0.6025161203048203, + "grad_norm": 0.7694439136497175, + "learning_rate": 1.4405873577137383e-06, + "loss": 0.8391, + "step": 6681 + }, + { + "epoch": 0.6026063038282906, + "grad_norm": 2.3630864338929864, + "learning_rate": 1.4400264893717816e-06, + "loss": 0.9951, + "step": 6682 + }, + { + "epoch": 0.6026964873517608, + "grad_norm": 1.4681185724565124, + "learning_rate": 1.4394656688098526e-06, + "loss": 1.0342, + "step": 6683 + }, + { + "epoch": 0.6027866708752311, + "grad_norm": 1.6807695319127272, + "learning_rate": 1.4389048960758032e-06, + "loss": 1.0081, + "step": 6684 + }, + { + "epoch": 0.6028768543987013, + "grad_norm": 1.9525253596105956, + "learning_rate": 1.4383441712174826e-06, + "loss": 1.0138, + "step": 6685 + }, + { + "epoch": 0.6029670379221717, + "grad_norm": 1.9983441531620445, + "learning_rate": 1.4377834942827333e-06, + "loss": 1.0497, + "step": 6686 + }, + { + "epoch": 0.6030572214456419, + "grad_norm": 1.6166107328642658, + "learning_rate": 1.437222865319397e-06, + "loss": 0.9259, + "step": 6687 + }, + { + "epoch": 0.6031474049691121, + "grad_norm": 1.4877722162515172, + "learning_rate": 1.4366622843753092e-06, + "loss": 0.9113, + "step": 6688 + }, + { + "epoch": 0.6032375884925824, + "grad_norm": 1.8298774861667333, + "learning_rate": 1.4361017514983006e-06, + "loss": 0.9989, + "step": 6689 + }, + { + "epoch": 0.6033277720160527, + "grad_norm": 1.4822651374392661, + "learning_rate": 1.4355412667362006e-06, + "loss": 1.0388, + "step": 6690 + }, + { + "epoch": 0.6034179555395229, + "grad_norm": 1.8272181813127164, + "learning_rate": 1.4349808301368311e-06, + "loss": 0.9382, + "step": 6691 + }, + { + "epoch": 0.6035081390629932, + "grad_norm": 1.7379955957135231, + "learning_rate": 1.4344204417480139e-06, + "loss": 1.0246, + "step": 6692 + }, + { + "epoch": 0.6035983225864634, + "grad_norm": 1.899848947041824, + "learning_rate": 1.4338601016175628e-06, + "loss": 0.9334, + "step": 6693 + }, + { + "epoch": 0.6036885061099337, + "grad_norm": 1.5584021429009949, + "learning_rate": 1.433299809793289e-06, + "loss": 0.9804, + "step": 6694 + }, + { + "epoch": 0.603778689633404, + "grad_norm": 2.3287436668158357, + "learning_rate": 1.432739566323001e-06, + "loss": 0.8516, + "step": 6695 + }, + { + "epoch": 0.6038688731568742, + "grad_norm": 1.6973413378089017, + "learning_rate": 1.4321793712545004e-06, + "loss": 0.9536, + "step": 6696 + }, + { + "epoch": 0.6039590566803446, + "grad_norm": 1.893437272238824, + "learning_rate": 1.4316192246355873e-06, + "loss": 1.0277, + "step": 6697 + }, + { + "epoch": 0.6040492402038148, + "grad_norm": 0.6045910041168223, + "learning_rate": 1.4310591265140555e-06, + "loss": 0.8162, + "step": 6698 + }, + { + "epoch": 0.604139423727285, + "grad_norm": 1.4454955114019816, + "learning_rate": 1.4304990769376963e-06, + "loss": 0.9898, + "step": 6699 + }, + { + "epoch": 0.6042296072507553, + "grad_norm": 1.408628105975965, + "learning_rate": 1.4299390759542962e-06, + "loss": 0.9677, + "step": 6700 + }, + { + "epoch": 0.6043197907742256, + "grad_norm": 1.6678503919145657, + "learning_rate": 1.4293791236116368e-06, + "loss": 0.9646, + "step": 6701 + }, + { + "epoch": 0.6044099742976958, + "grad_norm": 5.196813656186262, + "learning_rate": 1.4288192199574978e-06, + "loss": 0.95, + "step": 6702 + }, + { + "epoch": 0.6045001578211661, + "grad_norm": 0.6242533276736196, + "learning_rate": 1.4282593650396524e-06, + "loss": 0.7891, + "step": 6703 + }, + { + "epoch": 0.6045903413446363, + "grad_norm": 1.3112356817520145, + "learning_rate": 1.4276995589058695e-06, + "loss": 1.0801, + "step": 6704 + }, + { + "epoch": 0.6046805248681066, + "grad_norm": 0.5915238912228359, + "learning_rate": 1.4271398016039168e-06, + "loss": 0.7855, + "step": 6705 + }, + { + "epoch": 0.6047707083915769, + "grad_norm": 1.8345451107139963, + "learning_rate": 1.4265800931815542e-06, + "loss": 1.0209, + "step": 6706 + }, + { + "epoch": 0.6048608919150471, + "grad_norm": 1.5797252284524221, + "learning_rate": 1.4260204336865406e-06, + "loss": 0.9948, + "step": 6707 + }, + { + "epoch": 0.6049510754385173, + "grad_norm": 2.6301913018509273, + "learning_rate": 1.4254608231666281e-06, + "loss": 0.9819, + "step": 6708 + }, + { + "epoch": 0.6050412589619877, + "grad_norm": 1.4461164558231436, + "learning_rate": 1.4249012616695661e-06, + "loss": 0.9914, + "step": 6709 + }, + { + "epoch": 0.6051314424854579, + "grad_norm": 1.6074787560893882, + "learning_rate": 1.4243417492431e-06, + "loss": 0.9889, + "step": 6710 + }, + { + "epoch": 0.6052216260089281, + "grad_norm": 1.5076582439360795, + "learning_rate": 1.4237822859349696e-06, + "loss": 0.9611, + "step": 6711 + }, + { + "epoch": 0.6053118095323984, + "grad_norm": 1.3599714088940762, + "learning_rate": 1.423222871792912e-06, + "loss": 0.9555, + "step": 6712 + }, + { + "epoch": 0.6054019930558687, + "grad_norm": 1.5274350658543923, + "learning_rate": 1.4226635068646586e-06, + "loss": 1.0392, + "step": 6713 + }, + { + "epoch": 0.605492176579339, + "grad_norm": 0.626199120341097, + "learning_rate": 1.4221041911979393e-06, + "loss": 0.8094, + "step": 6714 + }, + { + "epoch": 0.6055823601028092, + "grad_norm": 1.7014047176463836, + "learning_rate": 1.4215449248404765e-06, + "loss": 1.001, + "step": 6715 + }, + { + "epoch": 0.6056725436262794, + "grad_norm": 1.5094626296611744, + "learning_rate": 1.4209857078399896e-06, + "loss": 1.0031, + "step": 6716 + }, + { + "epoch": 0.6057627271497498, + "grad_norm": 1.791437363551935, + "learning_rate": 1.4204265402441955e-06, + "loss": 0.9552, + "step": 6717 + }, + { + "epoch": 0.60585291067322, + "grad_norm": 1.4586354442290839, + "learning_rate": 1.419867422100804e-06, + "loss": 0.8604, + "step": 6718 + }, + { + "epoch": 0.6059430941966902, + "grad_norm": 1.67924522077327, + "learning_rate": 1.4193083534575236e-06, + "loss": 0.9696, + "step": 6719 + }, + { + "epoch": 0.6060332777201606, + "grad_norm": 1.6900664699055403, + "learning_rate": 1.4187493343620567e-06, + "loss": 1.007, + "step": 6720 + }, + { + "epoch": 0.6061234612436308, + "grad_norm": 1.7223579684804593, + "learning_rate": 1.4181903648621006e-06, + "loss": 1.0163, + "step": 6721 + }, + { + "epoch": 0.606213644767101, + "grad_norm": 5.948012512111779, + "learning_rate": 1.4176314450053512e-06, + "loss": 1.0151, + "step": 6722 + }, + { + "epoch": 0.6063038282905713, + "grad_norm": 1.6265401877156396, + "learning_rate": 1.4170725748394977e-06, + "loss": 0.9734, + "step": 6723 + }, + { + "epoch": 0.6063940118140416, + "grad_norm": 1.4604085688926551, + "learning_rate": 1.4165137544122266e-06, + "loss": 0.8946, + "step": 6724 + }, + { + "epoch": 0.6064841953375119, + "grad_norm": 1.7682960359029847, + "learning_rate": 1.4159549837712194e-06, + "loss": 1.051, + "step": 6725 + }, + { + "epoch": 0.6065743788609821, + "grad_norm": 1.378066544796, + "learning_rate": 1.415396262964153e-06, + "loss": 0.9009, + "step": 6726 + }, + { + "epoch": 0.6066645623844523, + "grad_norm": 1.691598084035146, + "learning_rate": 1.4148375920387016e-06, + "loss": 1.0552, + "step": 6727 + }, + { + "epoch": 0.6067547459079227, + "grad_norm": 0.6368873700558744, + "learning_rate": 1.4142789710425325e-06, + "loss": 0.8124, + "step": 6728 + }, + { + "epoch": 0.6068449294313929, + "grad_norm": 2.3022809411300935, + "learning_rate": 1.4137204000233118e-06, + "loss": 0.9164, + "step": 6729 + }, + { + "epoch": 0.6069351129548631, + "grad_norm": 1.9654402993275952, + "learning_rate": 1.4131618790286987e-06, + "loss": 0.9792, + "step": 6730 + }, + { + "epoch": 0.6070252964783334, + "grad_norm": 1.71273876112888, + "learning_rate": 1.4126034081063506e-06, + "loss": 0.9716, + "step": 6731 + }, + { + "epoch": 0.6071154800018037, + "grad_norm": 1.6866773083442845, + "learning_rate": 1.4120449873039186e-06, + "loss": 0.9674, + "step": 6732 + }, + { + "epoch": 0.6072056635252739, + "grad_norm": 3.9329955963567915, + "learning_rate": 1.4114866166690494e-06, + "loss": 1.02, + "step": 6733 + }, + { + "epoch": 0.6072958470487442, + "grad_norm": 1.2375938842301766, + "learning_rate": 1.4109282962493877e-06, + "loss": 0.9394, + "step": 6734 + }, + { + "epoch": 0.6073860305722144, + "grad_norm": 2.258285525708607, + "learning_rate": 1.4103700260925716e-06, + "loss": 0.9754, + "step": 6735 + }, + { + "epoch": 0.6074762140956848, + "grad_norm": 2.632834414608814, + "learning_rate": 1.4098118062462364e-06, + "loss": 1.0227, + "step": 6736 + }, + { + "epoch": 0.607566397619155, + "grad_norm": 1.3810396198221666, + "learning_rate": 1.4092536367580123e-06, + "loss": 0.9792, + "step": 6737 + }, + { + "epoch": 0.6076565811426252, + "grad_norm": 1.5613024127552348, + "learning_rate": 1.4086955176755248e-06, + "loss": 0.8984, + "step": 6738 + }, + { + "epoch": 0.6077467646660955, + "grad_norm": 1.5741099279762496, + "learning_rate": 1.4081374490463964e-06, + "loss": 0.9246, + "step": 6739 + }, + { + "epoch": 0.6078369481895658, + "grad_norm": 5.681700715671403, + "learning_rate": 1.4075794309182443e-06, + "loss": 0.995, + "step": 6740 + }, + { + "epoch": 0.607927131713036, + "grad_norm": 1.881233914835654, + "learning_rate": 1.407021463338682e-06, + "loss": 0.9897, + "step": 6741 + }, + { + "epoch": 0.6080173152365063, + "grad_norm": 0.591693403307046, + "learning_rate": 1.4064635463553177e-06, + "loss": 0.7581, + "step": 6742 + }, + { + "epoch": 0.6081074987599766, + "grad_norm": 1.9210702094970122, + "learning_rate": 1.4059056800157563e-06, + "loss": 1.0147, + "step": 6743 + }, + { + "epoch": 0.6081976822834468, + "grad_norm": 2.125876189960912, + "learning_rate": 1.4053478643675982e-06, + "loss": 0.9157, + "step": 6744 + }, + { + "epoch": 0.6082878658069171, + "grad_norm": 1.5790167073869203, + "learning_rate": 1.4047900994584389e-06, + "loss": 0.8808, + "step": 6745 + }, + { + "epoch": 0.6083780493303873, + "grad_norm": 1.6282712581493775, + "learning_rate": 1.404232385335871e-06, + "loss": 0.8886, + "step": 6746 + }, + { + "epoch": 0.6084682328538576, + "grad_norm": 10.65828419172716, + "learning_rate": 1.4036747220474806e-06, + "loss": 0.9817, + "step": 6747 + }, + { + "epoch": 0.6085584163773279, + "grad_norm": 1.5897980104657354, + "learning_rate": 1.4031171096408506e-06, + "loss": 0.9344, + "step": 6748 + }, + { + "epoch": 0.6086485999007981, + "grad_norm": 1.6669060137866856, + "learning_rate": 1.4025595481635607e-06, + "loss": 1.0063, + "step": 6749 + }, + { + "epoch": 0.6087387834242683, + "grad_norm": 2.0230434431690334, + "learning_rate": 1.4020020376631836e-06, + "loss": 1.0062, + "step": 6750 + }, + { + "epoch": 0.6088289669477387, + "grad_norm": 2.2386774641135956, + "learning_rate": 1.4014445781872908e-06, + "loss": 0.9939, + "step": 6751 + }, + { + "epoch": 0.6089191504712089, + "grad_norm": 1.6354502129809407, + "learning_rate": 1.4008871697834465e-06, + "loss": 1.0465, + "step": 6752 + }, + { + "epoch": 0.6090093339946792, + "grad_norm": 1.4734814504822111, + "learning_rate": 1.400329812499213e-06, + "loss": 1.0171, + "step": 6753 + }, + { + "epoch": 0.6090995175181494, + "grad_norm": 1.402421231905348, + "learning_rate": 1.3997725063821458e-06, + "loss": 0.9932, + "step": 6754 + }, + { + "epoch": 0.6091897010416197, + "grad_norm": 1.6058802761758184, + "learning_rate": 1.3992152514797978e-06, + "loss": 0.9485, + "step": 6755 + }, + { + "epoch": 0.60927988456509, + "grad_norm": 0.6327569489762538, + "learning_rate": 1.398658047839718e-06, + "loss": 0.8331, + "step": 6756 + }, + { + "epoch": 0.6093700680885602, + "grad_norm": 1.013816573265401, + "learning_rate": 1.3981008955094481e-06, + "loss": 0.8032, + "step": 6757 + }, + { + "epoch": 0.6094602516120304, + "grad_norm": 1.3086225819614103, + "learning_rate": 1.39754379453653e-06, + "loss": 0.9581, + "step": 6758 + }, + { + "epoch": 0.6095504351355008, + "grad_norm": 2.3360188447701655, + "learning_rate": 1.3969867449684972e-06, + "loss": 0.9914, + "step": 6759 + }, + { + "epoch": 0.609640618658971, + "grad_norm": 2.0806560734278827, + "learning_rate": 1.396429746852879e-06, + "loss": 0.9746, + "step": 6760 + }, + { + "epoch": 0.6097308021824412, + "grad_norm": 1.7166413984321702, + "learning_rate": 1.395872800237204e-06, + "loss": 0.959, + "step": 6761 + }, + { + "epoch": 0.6098209857059115, + "grad_norm": 0.6897019107628419, + "learning_rate": 1.3953159051689918e-06, + "loss": 0.8345, + "step": 6762 + }, + { + "epoch": 0.6099111692293818, + "grad_norm": 1.8765032146659086, + "learning_rate": 1.3947590616957618e-06, + "loss": 1.0272, + "step": 6763 + }, + { + "epoch": 0.610001352752852, + "grad_norm": 1.6949861273858968, + "learning_rate": 1.3942022698650258e-06, + "loss": 1.0133, + "step": 6764 + }, + { + "epoch": 0.6100915362763223, + "grad_norm": 1.59214791949699, + "learning_rate": 1.3936455297242917e-06, + "loss": 1.018, + "step": 6765 + }, + { + "epoch": 0.6101817197997926, + "grad_norm": 2.1068697145045503, + "learning_rate": 1.3930888413210652e-06, + "loss": 1.0004, + "step": 6766 + }, + { + "epoch": 0.6102719033232629, + "grad_norm": 2.193500789387415, + "learning_rate": 1.392532204702845e-06, + "loss": 0.9464, + "step": 6767 + }, + { + "epoch": 0.6103620868467331, + "grad_norm": 2.1547371906307404, + "learning_rate": 1.3919756199171266e-06, + "loss": 0.9071, + "step": 6768 + }, + { + "epoch": 0.6104522703702033, + "grad_norm": 1.478616567413667, + "learning_rate": 1.3914190870114009e-06, + "loss": 1.0539, + "step": 6769 + }, + { + "epoch": 0.6105424538936737, + "grad_norm": 1.4574040403887758, + "learning_rate": 1.3908626060331541e-06, + "loss": 1.0316, + "step": 6770 + }, + { + "epoch": 0.6106326374171439, + "grad_norm": 1.8097058825447683, + "learning_rate": 1.3903061770298693e-06, + "loss": 1.0156, + "step": 6771 + }, + { + "epoch": 0.6107228209406141, + "grad_norm": 1.6556673284572454, + "learning_rate": 1.3897498000490223e-06, + "loss": 0.9602, + "step": 6772 + }, + { + "epoch": 0.6108130044640844, + "grad_norm": 1.4645754962709652, + "learning_rate": 1.3891934751380879e-06, + "loss": 1.0539, + "step": 6773 + }, + { + "epoch": 0.6109031879875547, + "grad_norm": 1.5917795004803317, + "learning_rate": 1.3886372023445334e-06, + "loss": 1.002, + "step": 6774 + }, + { + "epoch": 0.610993371511025, + "grad_norm": 1.5135535015458244, + "learning_rate": 1.3880809817158246e-06, + "loss": 1.0529, + "step": 6775 + }, + { + "epoch": 0.6110835550344952, + "grad_norm": 1.6604428032796723, + "learning_rate": 1.3875248132994206e-06, + "loss": 0.9757, + "step": 6776 + }, + { + "epoch": 0.6111737385579654, + "grad_norm": 1.755515942280871, + "learning_rate": 1.386968697142776e-06, + "loss": 0.9068, + "step": 6777 + }, + { + "epoch": 0.6112639220814358, + "grad_norm": 1.5447234015408928, + "learning_rate": 1.386412633293343e-06, + "loss": 1.0312, + "step": 6778 + }, + { + "epoch": 0.611354105604906, + "grad_norm": 2.396838422394214, + "learning_rate": 1.3858566217985672e-06, + "loss": 0.9571, + "step": 6779 + }, + { + "epoch": 0.6114442891283762, + "grad_norm": 2.186463355618531, + "learning_rate": 1.3853006627058905e-06, + "loss": 1.0665, + "step": 6780 + }, + { + "epoch": 0.6115344726518465, + "grad_norm": 5.427145031988936, + "learning_rate": 1.3847447560627512e-06, + "loss": 0.8709, + "step": 6781 + }, + { + "epoch": 0.6116246561753168, + "grad_norm": 1.6926050320917978, + "learning_rate": 1.3841889019165812e-06, + "loss": 1.0271, + "step": 6782 + }, + { + "epoch": 0.611714839698787, + "grad_norm": 1.8682523747422297, + "learning_rate": 1.3836331003148106e-06, + "loss": 0.9393, + "step": 6783 + }, + { + "epoch": 0.6118050232222573, + "grad_norm": 1.3105933554944442, + "learning_rate": 1.3830773513048612e-06, + "loss": 1.0211, + "step": 6784 + }, + { + "epoch": 0.6118952067457275, + "grad_norm": 1.6126361138907268, + "learning_rate": 1.382521654934155e-06, + "loss": 0.9988, + "step": 6785 + }, + { + "epoch": 0.6119853902691978, + "grad_norm": 1.8618931748244871, + "learning_rate": 1.3819660112501057e-06, + "loss": 0.9293, + "step": 6786 + }, + { + "epoch": 0.6120755737926681, + "grad_norm": 1.5696289700458441, + "learning_rate": 1.3814104203001234e-06, + "loss": 0.9477, + "step": 6787 + }, + { + "epoch": 0.6121657573161383, + "grad_norm": 1.4398473564678196, + "learning_rate": 1.3808548821316156e-06, + "loss": 1.0252, + "step": 6788 + }, + { + "epoch": 0.6122559408396087, + "grad_norm": 1.2743357199489294, + "learning_rate": 1.3802993967919824e-06, + "loss": 0.9173, + "step": 6789 + }, + { + "epoch": 0.6123461243630789, + "grad_norm": 1.761258625580653, + "learning_rate": 1.3797439643286227e-06, + "loss": 0.9392, + "step": 6790 + }, + { + "epoch": 0.6124363078865491, + "grad_norm": 1.6876924899483472, + "learning_rate": 1.3791885847889277e-06, + "loss": 0.9695, + "step": 6791 + }, + { + "epoch": 0.6125264914100194, + "grad_norm": 1.5448180114324293, + "learning_rate": 1.3786332582202853e-06, + "loss": 1.0173, + "step": 6792 + }, + { + "epoch": 0.6126166749334897, + "grad_norm": 1.4429047194950246, + "learning_rate": 1.3780779846700799e-06, + "loss": 0.9841, + "step": 6793 + }, + { + "epoch": 0.6127068584569599, + "grad_norm": 1.4491730140399324, + "learning_rate": 1.3775227641856899e-06, + "loss": 0.9957, + "step": 6794 + }, + { + "epoch": 0.6127970419804302, + "grad_norm": 1.463153240024207, + "learning_rate": 1.37696759681449e-06, + "loss": 0.9836, + "step": 6795 + }, + { + "epoch": 0.6128872255039004, + "grad_norm": 0.699606058271545, + "learning_rate": 1.37641248260385e-06, + "loss": 0.8442, + "step": 6796 + }, + { + "epoch": 0.6129774090273707, + "grad_norm": 2.0375108917506886, + "learning_rate": 1.375857421601136e-06, + "loss": 1.0346, + "step": 6797 + }, + { + "epoch": 0.613067592550841, + "grad_norm": 1.6704868008030187, + "learning_rate": 1.3753024138537082e-06, + "loss": 0.9932, + "step": 6798 + }, + { + "epoch": 0.6131577760743112, + "grad_norm": 1.6134259987982165, + "learning_rate": 1.3747474594089221e-06, + "loss": 1.0505, + "step": 6799 + }, + { + "epoch": 0.6132479595977814, + "grad_norm": 1.6313548174977508, + "learning_rate": 1.374192558314131e-06, + "loss": 0.9671, + "step": 6800 + }, + { + "epoch": 0.6133381431212518, + "grad_norm": 2.4693209796655546, + "learning_rate": 1.373637710616681e-06, + "loss": 1.0476, + "step": 6801 + }, + { + "epoch": 0.613428326644722, + "grad_norm": 1.760384538647017, + "learning_rate": 1.373082916363916e-06, + "loss": 1.0965, + "step": 6802 + }, + { + "epoch": 0.6135185101681923, + "grad_norm": 1.4840840907884885, + "learning_rate": 1.3725281756031732e-06, + "loss": 1.0275, + "step": 6803 + }, + { + "epoch": 0.6136086936916625, + "grad_norm": 1.325335276668155, + "learning_rate": 1.3719734883817858e-06, + "loss": 0.9846, + "step": 6804 + }, + { + "epoch": 0.6136988772151328, + "grad_norm": 1.715882562366805, + "learning_rate": 1.371418854747084e-06, + "loss": 1.0997, + "step": 6805 + }, + { + "epoch": 0.6137890607386031, + "grad_norm": 2.036777780231241, + "learning_rate": 1.3708642747463905e-06, + "loss": 1.0212, + "step": 6806 + }, + { + "epoch": 0.6138792442620733, + "grad_norm": 2.267060029266991, + "learning_rate": 1.370309748427027e-06, + "loss": 1.0159, + "step": 6807 + }, + { + "epoch": 0.6139694277855435, + "grad_norm": 1.941549069468514, + "learning_rate": 1.3697552758363079e-06, + "loss": 1.0592, + "step": 6808 + }, + { + "epoch": 0.6140596113090139, + "grad_norm": 1.9882128390850935, + "learning_rate": 1.3692008570215432e-06, + "loss": 0.9905, + "step": 6809 + }, + { + "epoch": 0.6141497948324841, + "grad_norm": 1.605941665682611, + "learning_rate": 1.3686464920300398e-06, + "loss": 0.9351, + "step": 6810 + }, + { + "epoch": 0.6142399783559543, + "grad_norm": 1.8190697615019313, + "learning_rate": 1.3680921809090985e-06, + "loss": 0.9635, + "step": 6811 + }, + { + "epoch": 0.6143301618794246, + "grad_norm": 1.6745318399057918, + "learning_rate": 1.3675379237060175e-06, + "loss": 1.0303, + "step": 6812 + }, + { + "epoch": 0.6144203454028949, + "grad_norm": 1.4781502558221986, + "learning_rate": 1.366983720468088e-06, + "loss": 0.9756, + "step": 6813 + }, + { + "epoch": 0.6145105289263652, + "grad_norm": 1.3042805202540675, + "learning_rate": 1.3664295712425972e-06, + "loss": 0.9933, + "step": 6814 + }, + { + "epoch": 0.6146007124498354, + "grad_norm": 1.7125763729913615, + "learning_rate": 1.3658754760768296e-06, + "loss": 0.9201, + "step": 6815 + }, + { + "epoch": 0.6146908959733057, + "grad_norm": 1.7923337522528335, + "learning_rate": 1.3653214350180621e-06, + "loss": 1.0254, + "step": 6816 + }, + { + "epoch": 0.614781079496776, + "grad_norm": 1.6733721153101624, + "learning_rate": 1.3647674481135703e-06, + "loss": 1.0543, + "step": 6817 + }, + { + "epoch": 0.6148712630202462, + "grad_norm": 1.3668170099991208, + "learning_rate": 1.3642135154106217e-06, + "loss": 0.8862, + "step": 6818 + }, + { + "epoch": 0.6149614465437164, + "grad_norm": 1.7068928328552977, + "learning_rate": 1.363659636956482e-06, + "loss": 0.9331, + "step": 6819 + }, + { + "epoch": 0.6150516300671868, + "grad_norm": 1.6526734945200567, + "learning_rate": 1.3631058127984112e-06, + "loss": 0.9555, + "step": 6820 + }, + { + "epoch": 0.615141813590657, + "grad_norm": 1.6212989033872673, + "learning_rate": 1.3625520429836632e-06, + "loss": 0.9631, + "step": 6821 + }, + { + "epoch": 0.6152319971141272, + "grad_norm": 1.7188921089377829, + "learning_rate": 1.361998327559491e-06, + "loss": 0.8937, + "step": 6822 + }, + { + "epoch": 0.6153221806375975, + "grad_norm": 0.6569335420329784, + "learning_rate": 1.3614446665731385e-06, + "loss": 0.8187, + "step": 6823 + }, + { + "epoch": 0.6154123641610678, + "grad_norm": 1.6345537747989054, + "learning_rate": 1.3608910600718484e-06, + "loss": 0.9948, + "step": 6824 + }, + { + "epoch": 0.615502547684538, + "grad_norm": 1.6144700369513283, + "learning_rate": 1.360337508102857e-06, + "loss": 0.9913, + "step": 6825 + }, + { + "epoch": 0.6155927312080083, + "grad_norm": 1.7291586159514871, + "learning_rate": 1.3597840107133962e-06, + "loss": 1.0811, + "step": 6826 + }, + { + "epoch": 0.6156829147314785, + "grad_norm": 1.5908164158370017, + "learning_rate": 1.3592305679506944e-06, + "loss": 0.9934, + "step": 6827 + }, + { + "epoch": 0.6157730982549489, + "grad_norm": 1.6827571055752923, + "learning_rate": 1.3586771798619726e-06, + "loss": 0.8909, + "step": 6828 + }, + { + "epoch": 0.6158632817784191, + "grad_norm": 1.6896583035632657, + "learning_rate": 1.358123846494451e-06, + "loss": 1.0487, + "step": 6829 + }, + { + "epoch": 0.6159534653018893, + "grad_norm": 1.934374102713973, + "learning_rate": 1.3575705678953422e-06, + "loss": 0.9907, + "step": 6830 + }, + { + "epoch": 0.6160436488253596, + "grad_norm": 1.7189800455289992, + "learning_rate": 1.357017344111854e-06, + "loss": 1.0258, + "step": 6831 + }, + { + "epoch": 0.6161338323488299, + "grad_norm": 1.5283318564859547, + "learning_rate": 1.356464175191192e-06, + "loss": 0.9985, + "step": 6832 + }, + { + "epoch": 0.6162240158723001, + "grad_norm": 2.929721353971067, + "learning_rate": 1.3559110611805542e-06, + "loss": 1.0127, + "step": 6833 + }, + { + "epoch": 0.6163141993957704, + "grad_norm": 1.8999994704597136, + "learning_rate": 1.3553580021271372e-06, + "loss": 1.0295, + "step": 6834 + }, + { + "epoch": 0.6164043829192406, + "grad_norm": 1.429390881729132, + "learning_rate": 1.3548049980781297e-06, + "loss": 0.9311, + "step": 6835 + }, + { + "epoch": 0.616494566442711, + "grad_norm": 2.2573104352982005, + "learning_rate": 1.3542520490807166e-06, + "loss": 0.9795, + "step": 6836 + }, + { + "epoch": 0.6165847499661812, + "grad_norm": 1.8268108372052154, + "learning_rate": 1.3536991551820802e-06, + "loss": 0.9745, + "step": 6837 + }, + { + "epoch": 0.6166749334896514, + "grad_norm": 1.767810717712119, + "learning_rate": 1.3531463164293952e-06, + "loss": 0.9562, + "step": 6838 + }, + { + "epoch": 0.6167651170131218, + "grad_norm": 2.0485801292131303, + "learning_rate": 1.3525935328698332e-06, + "loss": 0.8959, + "step": 6839 + }, + { + "epoch": 0.616855300536592, + "grad_norm": 1.8858984512546864, + "learning_rate": 1.3520408045505605e-06, + "loss": 0.923, + "step": 6840 + }, + { + "epoch": 0.6169454840600622, + "grad_norm": 1.6377049936589596, + "learning_rate": 1.3514881315187396e-06, + "loss": 0.9551, + "step": 6841 + }, + { + "epoch": 0.6170356675835325, + "grad_norm": 1.8691996183423158, + "learning_rate": 1.3509355138215273e-06, + "loss": 1.0447, + "step": 6842 + }, + { + "epoch": 0.6171258511070028, + "grad_norm": 1.5858506803754846, + "learning_rate": 1.350382951506075e-06, + "loss": 1.0689, + "step": 6843 + }, + { + "epoch": 0.617216034630473, + "grad_norm": 0.6439061225533678, + "learning_rate": 1.349830444619532e-06, + "loss": 0.8045, + "step": 6844 + }, + { + "epoch": 0.6173062181539433, + "grad_norm": 1.553840905105837, + "learning_rate": 1.3492779932090397e-06, + "loss": 1.0042, + "step": 6845 + }, + { + "epoch": 0.6173964016774135, + "grad_norm": 1.6637148785054636, + "learning_rate": 1.3487255973217377e-06, + "loss": 1.0138, + "step": 6846 + }, + { + "epoch": 0.6174865852008838, + "grad_norm": 2.025105970459165, + "learning_rate": 1.3481732570047592e-06, + "loss": 1.0559, + "step": 6847 + }, + { + "epoch": 0.6175767687243541, + "grad_norm": 1.983267888290342, + "learning_rate": 1.3476209723052314e-06, + "loss": 0.9203, + "step": 6848 + }, + { + "epoch": 0.6176669522478243, + "grad_norm": 1.6277209756261692, + "learning_rate": 1.3470687432702806e-06, + "loss": 1.0111, + "step": 6849 + }, + { + "epoch": 0.6177571357712945, + "grad_norm": 1.981690280261354, + "learning_rate": 1.346516569947024e-06, + "loss": 0.9175, + "step": 6850 + }, + { + "epoch": 0.6178473192947649, + "grad_norm": 1.7015213021719695, + "learning_rate": 1.3459644523825774e-06, + "loss": 0.9705, + "step": 6851 + }, + { + "epoch": 0.6179375028182351, + "grad_norm": 1.6911066860425692, + "learning_rate": 1.34541239062405e-06, + "loss": 1.0291, + "step": 6852 + }, + { + "epoch": 0.6180276863417054, + "grad_norm": 2.7689023184803556, + "learning_rate": 1.3448603847185464e-06, + "loss": 0.9883, + "step": 6853 + }, + { + "epoch": 0.6181178698651756, + "grad_norm": 1.779532089347082, + "learning_rate": 1.344308434713168e-06, + "loss": 0.9295, + "step": 6854 + }, + { + "epoch": 0.6182080533886459, + "grad_norm": 1.865460287263914, + "learning_rate": 1.3437565406550083e-06, + "loss": 0.9186, + "step": 6855 + }, + { + "epoch": 0.6182982369121162, + "grad_norm": 1.9755277676958884, + "learning_rate": 1.34320470259116e-06, + "loss": 0.9718, + "step": 6856 + }, + { + "epoch": 0.6183884204355864, + "grad_norm": 5.580687033357293, + "learning_rate": 1.3426529205687078e-06, + "loss": 0.8925, + "step": 6857 + }, + { + "epoch": 0.6184786039590566, + "grad_norm": 1.984816794889214, + "learning_rate": 1.3421011946347323e-06, + "loss": 1.0431, + "step": 6858 + }, + { + "epoch": 0.618568787482527, + "grad_norm": 1.7064704584917243, + "learning_rate": 1.3415495248363113e-06, + "loss": 0.9487, + "step": 6859 + }, + { + "epoch": 0.6186589710059972, + "grad_norm": 7.235140864985397, + "learning_rate": 1.3409979112205148e-06, + "loss": 1.0113, + "step": 6860 + }, + { + "epoch": 0.6187491545294674, + "grad_norm": 1.8843428854158961, + "learning_rate": 1.3404463538344107e-06, + "loss": 1.0104, + "step": 6861 + }, + { + "epoch": 0.6188393380529378, + "grad_norm": 1.8761988939495708, + "learning_rate": 1.3398948527250602e-06, + "loss": 1.0745, + "step": 6862 + }, + { + "epoch": 0.618929521576408, + "grad_norm": 0.7881684478714901, + "learning_rate": 1.3393434079395207e-06, + "loss": 0.8288, + "step": 6863 + }, + { + "epoch": 0.6190197050998782, + "grad_norm": 1.5265122904178108, + "learning_rate": 1.3387920195248449e-06, + "loss": 0.9736, + "step": 6864 + }, + { + "epoch": 0.6191098886233485, + "grad_norm": 1.5253711655238846, + "learning_rate": 1.3382406875280791e-06, + "loss": 0.9995, + "step": 6865 + }, + { + "epoch": 0.6192000721468188, + "grad_norm": 1.562667227379326, + "learning_rate": 1.3376894119962672e-06, + "loss": 1.0081, + "step": 6866 + }, + { + "epoch": 0.6192902556702891, + "grad_norm": 2.0198864504559646, + "learning_rate": 1.3371381929764464e-06, + "loss": 1.0128, + "step": 6867 + }, + { + "epoch": 0.6193804391937593, + "grad_norm": 0.6512488124638894, + "learning_rate": 1.3365870305156506e-06, + "loss": 0.8072, + "step": 6868 + }, + { + "epoch": 0.6194706227172295, + "grad_norm": 2.163678186768642, + "learning_rate": 1.3360359246609073e-06, + "loss": 0.9905, + "step": 6869 + }, + { + "epoch": 0.6195608062406999, + "grad_norm": 1.682600042307817, + "learning_rate": 1.3354848754592387e-06, + "loss": 0.9827, + "step": 6870 + }, + { + "epoch": 0.6196509897641701, + "grad_norm": 1.5998824821468287, + "learning_rate": 1.334933882957666e-06, + "loss": 1.0734, + "step": 6871 + }, + { + "epoch": 0.6197411732876403, + "grad_norm": 1.3527090688789907, + "learning_rate": 1.3343829472032004e-06, + "loss": 0.9492, + "step": 6872 + }, + { + "epoch": 0.6198313568111106, + "grad_norm": 1.5983440532198876, + "learning_rate": 1.3338320682428527e-06, + "loss": 0.9689, + "step": 6873 + }, + { + "epoch": 0.6199215403345809, + "grad_norm": 6.088895301784532, + "learning_rate": 1.3332812461236263e-06, + "loss": 0.9495, + "step": 6874 + }, + { + "epoch": 0.6200117238580511, + "grad_norm": 1.6012891582671098, + "learning_rate": 1.3327304808925192e-06, + "loss": 0.9528, + "step": 6875 + }, + { + "epoch": 0.6201019073815214, + "grad_norm": 1.7985554673839201, + "learning_rate": 1.332179772596528e-06, + "loss": 0.8481, + "step": 6876 + }, + { + "epoch": 0.6201920909049916, + "grad_norm": 3.1128079426723176, + "learning_rate": 1.3316291212826402e-06, + "loss": 0.9085, + "step": 6877 + }, + { + "epoch": 0.620282274428462, + "grad_norm": 2.4427137124030147, + "learning_rate": 1.3310785269978413e-06, + "loss": 1.0372, + "step": 6878 + }, + { + "epoch": 0.6203724579519322, + "grad_norm": 1.7011569490007872, + "learning_rate": 1.3305279897891111e-06, + "loss": 1.0267, + "step": 6879 + }, + { + "epoch": 0.6204626414754024, + "grad_norm": 1.9033088507583633, + "learning_rate": 1.329977509703424e-06, + "loss": 0.8507, + "step": 6880 + }, + { + "epoch": 0.6205528249988727, + "grad_norm": 1.9101619915398118, + "learning_rate": 1.32942708678775e-06, + "loss": 1.0712, + "step": 6881 + }, + { + "epoch": 0.620643008522343, + "grad_norm": 1.5056387298071987, + "learning_rate": 1.3288767210890548e-06, + "loss": 0.9976, + "step": 6882 + }, + { + "epoch": 0.6207331920458132, + "grad_norm": 2.045207735063954, + "learning_rate": 1.3283264126542986e-06, + "loss": 0.9469, + "step": 6883 + }, + { + "epoch": 0.6208233755692835, + "grad_norm": 1.5280781026871588, + "learning_rate": 1.3277761615304356e-06, + "loss": 0.9851, + "step": 6884 + }, + { + "epoch": 0.6209135590927538, + "grad_norm": 1.782323346580111, + "learning_rate": 1.3272259677644178e-06, + "loss": 1.0882, + "step": 6885 + }, + { + "epoch": 0.621003742616224, + "grad_norm": 2.8197898486135875, + "learning_rate": 1.32667583140319e-06, + "loss": 0.8601, + "step": 6886 + }, + { + "epoch": 0.6210939261396943, + "grad_norm": 2.584642437493744, + "learning_rate": 1.3261257524936924e-06, + "loss": 1.0282, + "step": 6887 + }, + { + "epoch": 0.6211841096631645, + "grad_norm": 1.5057704716227702, + "learning_rate": 1.3255757310828619e-06, + "loss": 1.0157, + "step": 6888 + }, + { + "epoch": 0.6212742931866349, + "grad_norm": 1.8804531272679808, + "learning_rate": 1.3250257672176282e-06, + "loss": 1.0061, + "step": 6889 + }, + { + "epoch": 0.6213644767101051, + "grad_norm": 1.6008064532402015, + "learning_rate": 1.3244758609449183e-06, + "loss": 0.8934, + "step": 6890 + }, + { + "epoch": 0.6214546602335753, + "grad_norm": 1.7590834028716964, + "learning_rate": 1.323926012311653e-06, + "loss": 1.0058, + "step": 6891 + }, + { + "epoch": 0.6215448437570456, + "grad_norm": 1.6081230701445333, + "learning_rate": 1.3233762213647476e-06, + "loss": 1.0106, + "step": 6892 + }, + { + "epoch": 0.6216350272805159, + "grad_norm": 5.27969355474183, + "learning_rate": 1.3228264881511137e-06, + "loss": 0.9604, + "step": 6893 + }, + { + "epoch": 0.6217252108039861, + "grad_norm": 1.7631355222045253, + "learning_rate": 1.322276812717658e-06, + "loss": 1.0728, + "step": 6894 + }, + { + "epoch": 0.6218153943274564, + "grad_norm": 1.8256561693497197, + "learning_rate": 1.3217271951112818e-06, + "loss": 0.8951, + "step": 6895 + }, + { + "epoch": 0.6219055778509266, + "grad_norm": 1.5742945499013974, + "learning_rate": 1.321177635378881e-06, + "loss": 0.967, + "step": 6896 + }, + { + "epoch": 0.6219957613743969, + "grad_norm": 1.8288626853450574, + "learning_rate": 1.3206281335673475e-06, + "loss": 0.8806, + "step": 6897 + }, + { + "epoch": 0.6220859448978672, + "grad_norm": 2.3490702961179166, + "learning_rate": 1.3200786897235677e-06, + "loss": 0.9638, + "step": 6898 + }, + { + "epoch": 0.6221761284213374, + "grad_norm": 1.6141749532103182, + "learning_rate": 1.3195293038944227e-06, + "loss": 1.0025, + "step": 6899 + }, + { + "epoch": 0.6222663119448076, + "grad_norm": 1.382111323114128, + "learning_rate": 1.3189799761267902e-06, + "loss": 0.9806, + "step": 6900 + }, + { + "epoch": 0.622356495468278, + "grad_norm": 2.4551672725834273, + "learning_rate": 1.3184307064675412e-06, + "loss": 0.9787, + "step": 6901 + }, + { + "epoch": 0.6224466789917482, + "grad_norm": 0.7958231697298985, + "learning_rate": 1.3178814949635416e-06, + "loss": 0.8613, + "step": 6902 + }, + { + "epoch": 0.6225368625152184, + "grad_norm": 1.8935464403386997, + "learning_rate": 1.3173323416616549e-06, + "loss": 0.9319, + "step": 6903 + }, + { + "epoch": 0.6226270460386887, + "grad_norm": 1.8263699837186849, + "learning_rate": 1.3167832466087361e-06, + "loss": 1.0406, + "step": 6904 + }, + { + "epoch": 0.622717229562159, + "grad_norm": 1.3955177216014532, + "learning_rate": 1.3162342098516388e-06, + "loss": 0.9317, + "step": 6905 + }, + { + "epoch": 0.6228074130856293, + "grad_norm": 1.506616938679023, + "learning_rate": 1.3156852314372086e-06, + "loss": 0.9076, + "step": 6906 + }, + { + "epoch": 0.6228975966090995, + "grad_norm": 0.6768945993349186, + "learning_rate": 1.3151363114122882e-06, + "loss": 0.7974, + "step": 6907 + }, + { + "epoch": 0.6229877801325698, + "grad_norm": 1.6865666952146383, + "learning_rate": 1.3145874498237133e-06, + "loss": 1.0017, + "step": 6908 + }, + { + "epoch": 0.6230779636560401, + "grad_norm": 1.5273504622914367, + "learning_rate": 1.3140386467183166e-06, + "loss": 1.0321, + "step": 6909 + }, + { + "epoch": 0.6231681471795103, + "grad_norm": 1.6210315370686477, + "learning_rate": 1.3134899021429258e-06, + "loss": 1.0285, + "step": 6910 + }, + { + "epoch": 0.6232583307029805, + "grad_norm": 1.664531086597884, + "learning_rate": 1.3129412161443604e-06, + "loss": 0.9699, + "step": 6911 + }, + { + "epoch": 0.6233485142264509, + "grad_norm": 1.8041024188553505, + "learning_rate": 1.3123925887694402e-06, + "loss": 0.9935, + "step": 6912 + }, + { + "epoch": 0.6234386977499211, + "grad_norm": 1.4607960841639285, + "learning_rate": 1.3118440200649756e-06, + "loss": 0.9969, + "step": 6913 + }, + { + "epoch": 0.6235288812733913, + "grad_norm": 1.6630009233537544, + "learning_rate": 1.3112955100777727e-06, + "loss": 0.9487, + "step": 6914 + }, + { + "epoch": 0.6236190647968616, + "grad_norm": 1.74053171616708, + "learning_rate": 1.3107470588546353e-06, + "loss": 1.0223, + "step": 6915 + }, + { + "epoch": 0.6237092483203319, + "grad_norm": 1.639578494590331, + "learning_rate": 1.3101986664423583e-06, + "loss": 0.9545, + "step": 6916 + }, + { + "epoch": 0.6237994318438022, + "grad_norm": 1.7355384704537455, + "learning_rate": 1.3096503328877358e-06, + "loss": 0.9596, + "step": 6917 + }, + { + "epoch": 0.6238896153672724, + "grad_norm": 1.546563473552984, + "learning_rate": 1.309102058237553e-06, + "loss": 1.006, + "step": 6918 + }, + { + "epoch": 0.6239797988907426, + "grad_norm": 1.3503446439286453, + "learning_rate": 1.3085538425385917e-06, + "loss": 1.0356, + "step": 6919 + }, + { + "epoch": 0.624069982414213, + "grad_norm": 2.112259877118028, + "learning_rate": 1.3080056858376298e-06, + "loss": 1.0154, + "step": 6920 + }, + { + "epoch": 0.6241601659376832, + "grad_norm": 1.7844655643602578, + "learning_rate": 1.3074575881814383e-06, + "loss": 0.9684, + "step": 6921 + }, + { + "epoch": 0.6242503494611534, + "grad_norm": 1.933532343958662, + "learning_rate": 1.3069095496167841e-06, + "loss": 0.911, + "step": 6922 + }, + { + "epoch": 0.6243405329846237, + "grad_norm": 1.4210171469688078, + "learning_rate": 1.3063615701904285e-06, + "loss": 0.8955, + "step": 6923 + }, + { + "epoch": 0.624430716508094, + "grad_norm": 1.9500553881286062, + "learning_rate": 1.3058136499491283e-06, + "loss": 1.0174, + "step": 6924 + }, + { + "epoch": 0.6245209000315642, + "grad_norm": 1.8871587710113331, + "learning_rate": 1.3052657889396357e-06, + "loss": 0.967, + "step": 6925 + }, + { + "epoch": 0.6246110835550345, + "grad_norm": 1.5576906575541074, + "learning_rate": 1.304717987208696e-06, + "loss": 0.9726, + "step": 6926 + }, + { + "epoch": 0.6247012670785047, + "grad_norm": 1.3935730364205674, + "learning_rate": 1.304170244803052e-06, + "loss": 0.9732, + "step": 6927 + }, + { + "epoch": 0.624791450601975, + "grad_norm": 1.6451356189559845, + "learning_rate": 1.3036225617694387e-06, + "loss": 0.9541, + "step": 6928 + }, + { + "epoch": 0.6248816341254453, + "grad_norm": 1.4572652265846437, + "learning_rate": 1.3030749381545892e-06, + "loss": 1.0151, + "step": 6929 + }, + { + "epoch": 0.6249718176489155, + "grad_norm": 1.770043642077163, + "learning_rate": 1.3025273740052285e-06, + "loss": 0.9236, + "step": 6930 + }, + { + "epoch": 0.6250620011723858, + "grad_norm": 1.5476087120954651, + "learning_rate": 1.3019798693680774e-06, + "loss": 1.0321, + "step": 6931 + }, + { + "epoch": 0.6251521846958561, + "grad_norm": 1.4241206283375003, + "learning_rate": 1.3014324242898536e-06, + "loss": 1.1002, + "step": 6932 + }, + { + "epoch": 0.6252423682193263, + "grad_norm": 1.9319058536076155, + "learning_rate": 1.3008850388172668e-06, + "loss": 1.043, + "step": 6933 + }, + { + "epoch": 0.6253325517427966, + "grad_norm": 1.6505064071581759, + "learning_rate": 1.3003377129970233e-06, + "loss": 0.9432, + "step": 6934 + }, + { + "epoch": 0.6254227352662669, + "grad_norm": 1.4126762212694517, + "learning_rate": 1.2997904468758243e-06, + "loss": 0.9069, + "step": 6935 + }, + { + "epoch": 0.6255129187897371, + "grad_norm": 1.6782454577071548, + "learning_rate": 1.299243240500365e-06, + "loss": 1.0035, + "step": 6936 + }, + { + "epoch": 0.6256031023132074, + "grad_norm": 1.9002563805520067, + "learning_rate": 1.2986960939173368e-06, + "loss": 1.0019, + "step": 6937 + }, + { + "epoch": 0.6256932858366776, + "grad_norm": 1.664332623115957, + "learning_rate": 1.298149007173424e-06, + "loss": 1.0299, + "step": 6938 + }, + { + "epoch": 0.625783469360148, + "grad_norm": 1.9010635862436718, + "learning_rate": 1.2976019803153087e-06, + "loss": 1.0185, + "step": 6939 + }, + { + "epoch": 0.6258736528836182, + "grad_norm": 1.66247346613081, + "learning_rate": 1.2970550133896652e-06, + "loss": 0.9318, + "step": 6940 + }, + { + "epoch": 0.6259638364070884, + "grad_norm": 1.892430412897796, + "learning_rate": 1.2965081064431634e-06, + "loss": 1.0653, + "step": 6941 + }, + { + "epoch": 0.6260540199305586, + "grad_norm": 1.789443995684186, + "learning_rate": 1.2959612595224698e-06, + "loss": 1.0124, + "step": 6942 + }, + { + "epoch": 0.626144203454029, + "grad_norm": 1.9400750552730617, + "learning_rate": 1.2954144726742424e-06, + "loss": 0.9967, + "step": 6943 + }, + { + "epoch": 0.6262343869774992, + "grad_norm": 1.4710573780427527, + "learning_rate": 1.2948677459451385e-06, + "loss": 0.9818, + "step": 6944 + }, + { + "epoch": 0.6263245705009695, + "grad_norm": 1.8069467646024944, + "learning_rate": 1.2943210793818064e-06, + "loss": 1.0319, + "step": 6945 + }, + { + "epoch": 0.6264147540244397, + "grad_norm": 1.7261471486095714, + "learning_rate": 1.2937744730308899e-06, + "loss": 0.9648, + "step": 6946 + }, + { + "epoch": 0.62650493754791, + "grad_norm": 2.0280721601706215, + "learning_rate": 1.2932279269390305e-06, + "loss": 0.9807, + "step": 6947 + }, + { + "epoch": 0.6265951210713803, + "grad_norm": 1.6715551765639474, + "learning_rate": 1.292681441152861e-06, + "loss": 0.8896, + "step": 6948 + }, + { + "epoch": 0.6266853045948505, + "grad_norm": 2.6531233393256177, + "learning_rate": 1.292135015719011e-06, + "loss": 0.9982, + "step": 6949 + }, + { + "epoch": 0.6267754881183207, + "grad_norm": 1.618989613005486, + "learning_rate": 1.2915886506841046e-06, + "loss": 0.9423, + "step": 6950 + }, + { + "epoch": 0.6268656716417911, + "grad_norm": 1.4766161066244583, + "learning_rate": 1.2910423460947613e-06, + "loss": 0.9198, + "step": 6951 + }, + { + "epoch": 0.6269558551652613, + "grad_norm": 1.5281890325255805, + "learning_rate": 1.290496101997594e-06, + "loss": 0.951, + "step": 6952 + }, + { + "epoch": 0.6270460386887315, + "grad_norm": 0.7365853748160174, + "learning_rate": 1.2899499184392105e-06, + "loss": 0.8189, + "step": 6953 + }, + { + "epoch": 0.6271362222122018, + "grad_norm": 1.758587340076751, + "learning_rate": 1.289403795466216e-06, + "loss": 0.9633, + "step": 6954 + }, + { + "epoch": 0.6272264057356721, + "grad_norm": 1.671942379743468, + "learning_rate": 1.288857733125207e-06, + "loss": 1.0021, + "step": 6955 + }, + { + "epoch": 0.6273165892591424, + "grad_norm": 2.281059048120404, + "learning_rate": 1.2883117314627785e-06, + "loss": 0.982, + "step": 6956 + }, + { + "epoch": 0.6274067727826126, + "grad_norm": 1.6211902267978437, + "learning_rate": 1.2877657905255168e-06, + "loss": 0.9201, + "step": 6957 + }, + { + "epoch": 0.6274969563060829, + "grad_norm": 1.447590368082089, + "learning_rate": 1.2872199103600046e-06, + "loss": 1.0443, + "step": 6958 + }, + { + "epoch": 0.6275871398295532, + "grad_norm": 1.4591240056437373, + "learning_rate": 1.286674091012821e-06, + "loss": 1.0509, + "step": 6959 + }, + { + "epoch": 0.6276773233530234, + "grad_norm": 1.5999100510824242, + "learning_rate": 1.2861283325305356e-06, + "loss": 1.0446, + "step": 6960 + }, + { + "epoch": 0.6277675068764936, + "grad_norm": 1.4430927448459614, + "learning_rate": 1.2855826349597185e-06, + "loss": 0.9853, + "step": 6961 + }, + { + "epoch": 0.627857690399964, + "grad_norm": 1.522850119010434, + "learning_rate": 1.2850369983469302e-06, + "loss": 0.8401, + "step": 6962 + }, + { + "epoch": 0.6279478739234342, + "grad_norm": 2.3285717631514475, + "learning_rate": 1.2844914227387266e-06, + "loss": 1.0113, + "step": 6963 + }, + { + "epoch": 0.6280380574469044, + "grad_norm": 1.3308838407900543, + "learning_rate": 1.2839459081816606e-06, + "loss": 1.0056, + "step": 6964 + }, + { + "epoch": 0.6281282409703747, + "grad_norm": 1.6859922384262174, + "learning_rate": 1.283400454722278e-06, + "loss": 1.0524, + "step": 6965 + }, + { + "epoch": 0.628218424493845, + "grad_norm": 1.577230596616796, + "learning_rate": 1.28285506240712e-06, + "loss": 1.0455, + "step": 6966 + }, + { + "epoch": 0.6283086080173153, + "grad_norm": 1.787947194640307, + "learning_rate": 1.2823097312827225e-06, + "loss": 1.0118, + "step": 6967 + }, + { + "epoch": 0.6283987915407855, + "grad_norm": 1.7391147629664065, + "learning_rate": 1.2817644613956153e-06, + "loss": 0.9934, + "step": 6968 + }, + { + "epoch": 0.6284889750642557, + "grad_norm": 1.703472976833003, + "learning_rate": 1.2812192527923253e-06, + "loss": 1.0494, + "step": 6969 + }, + { + "epoch": 0.6285791585877261, + "grad_norm": 1.7539434280363295, + "learning_rate": 1.2806741055193712e-06, + "loss": 1.0203, + "step": 6970 + }, + { + "epoch": 0.6286693421111963, + "grad_norm": 1.6729142711174116, + "learning_rate": 1.2801290196232695e-06, + "loss": 0.9454, + "step": 6971 + }, + { + "epoch": 0.6287595256346665, + "grad_norm": 2.275772835388237, + "learning_rate": 1.2795839951505282e-06, + "loss": 0.9756, + "step": 6972 + }, + { + "epoch": 0.6288497091581368, + "grad_norm": 1.841700691956003, + "learning_rate": 1.2790390321476538e-06, + "loss": 1.0013, + "step": 6973 + }, + { + "epoch": 0.6289398926816071, + "grad_norm": 1.4460366574253642, + "learning_rate": 1.2784941306611446e-06, + "loss": 0.9582, + "step": 6974 + }, + { + "epoch": 0.6290300762050773, + "grad_norm": 1.9917893554162516, + "learning_rate": 1.2779492907374935e-06, + "loss": 1.0447, + "step": 6975 + }, + { + "epoch": 0.6291202597285476, + "grad_norm": 1.6972298744571632, + "learning_rate": 1.2774045124231911e-06, + "loss": 0.9734, + "step": 6976 + }, + { + "epoch": 0.6292104432520178, + "grad_norm": 1.4370730222068693, + "learning_rate": 1.2768597957647197e-06, + "loss": 0.8661, + "step": 6977 + }, + { + "epoch": 0.6293006267754881, + "grad_norm": 1.9242129488016173, + "learning_rate": 1.2763151408085582e-06, + "loss": 0.9967, + "step": 6978 + }, + { + "epoch": 0.6293908102989584, + "grad_norm": 1.6163229727737844, + "learning_rate": 1.2757705476011788e-06, + "loss": 0.9536, + "step": 6979 + }, + { + "epoch": 0.6294809938224286, + "grad_norm": 1.5401375206320038, + "learning_rate": 1.27522601618905e-06, + "loss": 0.9297, + "step": 6980 + }, + { + "epoch": 0.629571177345899, + "grad_norm": 0.6011965988855483, + "learning_rate": 1.2746815466186337e-06, + "loss": 0.841, + "step": 6981 + }, + { + "epoch": 0.6296613608693692, + "grad_norm": 1.629394897058507, + "learning_rate": 1.274137138936387e-06, + "loss": 0.9889, + "step": 6982 + }, + { + "epoch": 0.6297515443928394, + "grad_norm": 1.5336219944347318, + "learning_rate": 1.2735927931887625e-06, + "loss": 0.9878, + "step": 6983 + }, + { + "epoch": 0.6298417279163097, + "grad_norm": 0.6944848825019138, + "learning_rate": 1.2730485094222061e-06, + "loss": 0.8235, + "step": 6984 + }, + { + "epoch": 0.62993191143978, + "grad_norm": 1.6133162007459907, + "learning_rate": 1.2725042876831586e-06, + "loss": 1.0241, + "step": 6985 + }, + { + "epoch": 0.6300220949632502, + "grad_norm": 3.14790601006923, + "learning_rate": 1.2719601280180573e-06, + "loss": 0.884, + "step": 6986 + }, + { + "epoch": 0.6301122784867205, + "grad_norm": 2.9902964061679107, + "learning_rate": 1.2714160304733317e-06, + "loss": 1.049, + "step": 6987 + }, + { + "epoch": 0.6302024620101907, + "grad_norm": 1.4266175208287166, + "learning_rate": 1.2708719950954082e-06, + "loss": 0.9971, + "step": 6988 + }, + { + "epoch": 0.630292645533661, + "grad_norm": 2.8161928305006447, + "learning_rate": 1.2703280219307065e-06, + "loss": 1.0043, + "step": 6989 + }, + { + "epoch": 0.6303828290571313, + "grad_norm": 1.5540520217514646, + "learning_rate": 1.2697841110256411e-06, + "loss": 0.9169, + "step": 6990 + }, + { + "epoch": 0.6304730125806015, + "grad_norm": 1.650975375444127, + "learning_rate": 1.2692402624266221e-06, + "loss": 1.0313, + "step": 6991 + }, + { + "epoch": 0.6305631961040717, + "grad_norm": 1.7583950857494712, + "learning_rate": 1.2686964761800529e-06, + "loss": 0.8968, + "step": 6992 + }, + { + "epoch": 0.6306533796275421, + "grad_norm": 0.6428040915881806, + "learning_rate": 1.268152752332333e-06, + "loss": 0.8212, + "step": 6993 + }, + { + "epoch": 0.6307435631510123, + "grad_norm": 1.7122156951359608, + "learning_rate": 1.2676090909298549e-06, + "loss": 0.8968, + "step": 6994 + }, + { + "epoch": 0.6308337466744826, + "grad_norm": 1.5958436416146964, + "learning_rate": 1.2670654920190086e-06, + "loss": 1.0306, + "step": 6995 + }, + { + "epoch": 0.6309239301979528, + "grad_norm": 1.4303585624653763, + "learning_rate": 1.2665219556461754e-06, + "loss": 1.0346, + "step": 6996 + }, + { + "epoch": 0.6310141137214231, + "grad_norm": 1.6760135658552733, + "learning_rate": 1.2659784818577329e-06, + "loss": 1.085, + "step": 6997 + }, + { + "epoch": 0.6311042972448934, + "grad_norm": 1.5320901706692727, + "learning_rate": 1.2654350707000542e-06, + "loss": 0.8723, + "step": 6998 + }, + { + "epoch": 0.6311944807683636, + "grad_norm": 1.7467620048760126, + "learning_rate": 1.264891722219505e-06, + "loss": 0.967, + "step": 6999 + }, + { + "epoch": 0.6312846642918338, + "grad_norm": 1.5625908634468941, + "learning_rate": 1.2643484364624483e-06, + "loss": 0.9823, + "step": 7000 + }, + { + "epoch": 0.6313748478153042, + "grad_norm": 1.6532734449112407, + "learning_rate": 1.2638052134752393e-06, + "loss": 1.037, + "step": 7001 + }, + { + "epoch": 0.6314650313387744, + "grad_norm": 1.6560550790924173, + "learning_rate": 1.2632620533042277e-06, + "loss": 1.0612, + "step": 7002 + }, + { + "epoch": 0.6315552148622446, + "grad_norm": 3.4695785495185576, + "learning_rate": 1.2627189559957612e-06, + "loss": 0.9503, + "step": 7003 + }, + { + "epoch": 0.631645398385715, + "grad_norm": 1.754639402224568, + "learning_rate": 1.262175921596178e-06, + "loss": 1.048, + "step": 7004 + }, + { + "epoch": 0.6317355819091852, + "grad_norm": 1.5939218765029783, + "learning_rate": 1.2616329501518137e-06, + "loss": 0.8593, + "step": 7005 + }, + { + "epoch": 0.6318257654326555, + "grad_norm": 1.6235785502894404, + "learning_rate": 1.2610900417089978e-06, + "loss": 0.9056, + "step": 7006 + }, + { + "epoch": 0.6319159489561257, + "grad_norm": 3.324085058085709, + "learning_rate": 1.2605471963140535e-06, + "loss": 0.9928, + "step": 7007 + }, + { + "epoch": 0.632006132479596, + "grad_norm": 1.7746681279224903, + "learning_rate": 1.2600044140133e-06, + "loss": 1.0598, + "step": 7008 + }, + { + "epoch": 0.6320963160030663, + "grad_norm": 1.585549744055747, + "learning_rate": 1.2594616948530493e-06, + "loss": 0.937, + "step": 7009 + }, + { + "epoch": 0.6321864995265365, + "grad_norm": 1.7893925658019034, + "learning_rate": 1.258919038879611e-06, + "loss": 1.0067, + "step": 7010 + }, + { + "epoch": 0.6322766830500067, + "grad_norm": 1.8353006759982355, + "learning_rate": 1.2583764461392859e-06, + "loss": 0.9851, + "step": 7011 + }, + { + "epoch": 0.6323668665734771, + "grad_norm": 1.6669723071548144, + "learning_rate": 1.2578339166783724e-06, + "loss": 0.9475, + "step": 7012 + }, + { + "epoch": 0.6324570500969473, + "grad_norm": 1.601216571944334, + "learning_rate": 1.2572914505431613e-06, + "loss": 0.9194, + "step": 7013 + }, + { + "epoch": 0.6325472336204175, + "grad_norm": 1.4004952831876643, + "learning_rate": 1.2567490477799383e-06, + "loss": 0.8857, + "step": 7014 + }, + { + "epoch": 0.6326374171438878, + "grad_norm": 1.5025828694421535, + "learning_rate": 1.2562067084349852e-06, + "loss": 1.0784, + "step": 7015 + }, + { + "epoch": 0.6327276006673581, + "grad_norm": 1.8625941259166394, + "learning_rate": 1.2556644325545764e-06, + "loss": 0.9664, + "step": 7016 + }, + { + "epoch": 0.6328177841908283, + "grad_norm": 1.6442290184674055, + "learning_rate": 1.255122220184983e-06, + "loss": 0.9651, + "step": 7017 + }, + { + "epoch": 0.6329079677142986, + "grad_norm": 1.8907396896506896, + "learning_rate": 1.2545800713724694e-06, + "loss": 0.8971, + "step": 7018 + }, + { + "epoch": 0.6329981512377688, + "grad_norm": 1.5282984722880375, + "learning_rate": 1.254037986163294e-06, + "loss": 0.8503, + "step": 7019 + }, + { + "epoch": 0.6330883347612392, + "grad_norm": 1.4267797866580845, + "learning_rate": 1.2534959646037104e-06, + "loss": 0.9442, + "step": 7020 + }, + { + "epoch": 0.6331785182847094, + "grad_norm": 1.7170810572754067, + "learning_rate": 1.2529540067399675e-06, + "loss": 0.9906, + "step": 7021 + }, + { + "epoch": 0.6332687018081796, + "grad_norm": 1.6786561860481264, + "learning_rate": 1.252412112618308e-06, + "loss": 0.9171, + "step": 7022 + }, + { + "epoch": 0.6333588853316499, + "grad_norm": 6.825806108315729, + "learning_rate": 1.2518702822849696e-06, + "loss": 0.8713, + "step": 7023 + }, + { + "epoch": 0.6334490688551202, + "grad_norm": 2.287537571332718, + "learning_rate": 1.2513285157861831e-06, + "loss": 0.9273, + "step": 7024 + }, + { + "epoch": 0.6335392523785904, + "grad_norm": 2.651617790729194, + "learning_rate": 1.2507868131681764e-06, + "loss": 0.9812, + "step": 7025 + }, + { + "epoch": 0.6336294359020607, + "grad_norm": 2.0819940331086797, + "learning_rate": 1.250245174477169e-06, + "loss": 0.9616, + "step": 7026 + }, + { + "epoch": 0.6337196194255309, + "grad_norm": 0.6537600692279922, + "learning_rate": 1.2497035997593783e-06, + "loss": 0.7572, + "step": 7027 + }, + { + "epoch": 0.6338098029490012, + "grad_norm": 1.4951950043024866, + "learning_rate": 1.2491620890610135e-06, + "loss": 0.9822, + "step": 7028 + }, + { + "epoch": 0.6338999864724715, + "grad_norm": 1.4567919168957146, + "learning_rate": 1.2486206424282788e-06, + "loss": 0.9666, + "step": 7029 + }, + { + "epoch": 0.6339901699959417, + "grad_norm": 1.6669035822645917, + "learning_rate": 1.2480792599073743e-06, + "loss": 1.0376, + "step": 7030 + }, + { + "epoch": 0.634080353519412, + "grad_norm": 1.526328527506449, + "learning_rate": 1.247537941544493e-06, + "loss": 1.022, + "step": 7031 + }, + { + "epoch": 0.6341705370428823, + "grad_norm": 3.377698455333687, + "learning_rate": 1.2469966873858242e-06, + "loss": 1.0156, + "step": 7032 + }, + { + "epoch": 0.6342607205663525, + "grad_norm": 1.9018149140212366, + "learning_rate": 1.2464554974775496e-06, + "loss": 0.9093, + "step": 7033 + }, + { + "epoch": 0.6343509040898228, + "grad_norm": 2.6777927829151422, + "learning_rate": 1.2459143718658474e-06, + "loss": 0.979, + "step": 7034 + }, + { + "epoch": 0.6344410876132931, + "grad_norm": 1.7401878616162194, + "learning_rate": 1.2453733105968886e-06, + "loss": 0.9168, + "step": 7035 + }, + { + "epoch": 0.6345312711367633, + "grad_norm": 0.6368333678260094, + "learning_rate": 1.2448323137168399e-06, + "loss": 0.8227, + "step": 7036 + }, + { + "epoch": 0.6346214546602336, + "grad_norm": 1.889289510384706, + "learning_rate": 1.2442913812718625e-06, + "loss": 0.9011, + "step": 7037 + }, + { + "epoch": 0.6347116381837038, + "grad_norm": 1.5709050517148693, + "learning_rate": 1.2437505133081108e-06, + "loss": 0.9644, + "step": 7038 + }, + { + "epoch": 0.6348018217071741, + "grad_norm": 1.6346254641882365, + "learning_rate": 1.2432097098717358e-06, + "loss": 0.9979, + "step": 7039 + }, + { + "epoch": 0.6348920052306444, + "grad_norm": 1.8167227520582412, + "learning_rate": 1.2426689710088813e-06, + "loss": 0.986, + "step": 7040 + }, + { + "epoch": 0.6349821887541146, + "grad_norm": 1.5971509512143696, + "learning_rate": 1.2421282967656855e-06, + "loss": 1.0403, + "step": 7041 + }, + { + "epoch": 0.6350723722775848, + "grad_norm": 1.956490264860559, + "learning_rate": 1.2415876871882827e-06, + "loss": 1.0076, + "step": 7042 + }, + { + "epoch": 0.6351625558010552, + "grad_norm": 0.6682168823167562, + "learning_rate": 1.2410471423227998e-06, + "loss": 0.8867, + "step": 7043 + }, + { + "epoch": 0.6352527393245254, + "grad_norm": 1.4782205788677845, + "learning_rate": 1.24050666221536e-06, + "loss": 0.9755, + "step": 7044 + }, + { + "epoch": 0.6353429228479957, + "grad_norm": 1.484200798654592, + "learning_rate": 1.23996624691208e-06, + "loss": 1.0463, + "step": 7045 + }, + { + "epoch": 0.6354331063714659, + "grad_norm": 1.3829798624085514, + "learning_rate": 1.2394258964590693e-06, + "loss": 0.9663, + "step": 7046 + }, + { + "epoch": 0.6355232898949362, + "grad_norm": 1.8475914070735622, + "learning_rate": 1.238885610902436e-06, + "loss": 0.9302, + "step": 7047 + }, + { + "epoch": 0.6356134734184065, + "grad_norm": 1.5965626151651762, + "learning_rate": 1.2383453902882787e-06, + "loss": 0.9926, + "step": 7048 + }, + { + "epoch": 0.6357036569418767, + "grad_norm": 1.4145921589834025, + "learning_rate": 1.2378052346626927e-06, + "loss": 1.0504, + "step": 7049 + }, + { + "epoch": 0.6357938404653469, + "grad_norm": 1.8015112731479237, + "learning_rate": 1.2372651440717665e-06, + "loss": 0.9063, + "step": 7050 + }, + { + "epoch": 0.6358840239888173, + "grad_norm": 1.7304228696919377, + "learning_rate": 1.236725118561584e-06, + "loss": 1.0177, + "step": 7051 + }, + { + "epoch": 0.6359742075122875, + "grad_norm": 2.238898972624522, + "learning_rate": 1.2361851581782232e-06, + "loss": 0.9836, + "step": 7052 + }, + { + "epoch": 0.6360643910357577, + "grad_norm": 1.4773368621455807, + "learning_rate": 1.2356452629677554e-06, + "loss": 1.0465, + "step": 7053 + }, + { + "epoch": 0.6361545745592281, + "grad_norm": 1.625124192994006, + "learning_rate": 1.2351054329762494e-06, + "loss": 0.9162, + "step": 7054 + }, + { + "epoch": 0.6362447580826983, + "grad_norm": 1.6013747686752642, + "learning_rate": 1.2345656682497648e-06, + "loss": 1.0107, + "step": 7055 + }, + { + "epoch": 0.6363349416061685, + "grad_norm": 1.3060091827940175, + "learning_rate": 1.2340259688343583e-06, + "loss": 0.959, + "step": 7056 + }, + { + "epoch": 0.6364251251296388, + "grad_norm": 1.8133464020837815, + "learning_rate": 1.2334863347760803e-06, + "loss": 0.8805, + "step": 7057 + }, + { + "epoch": 0.6365153086531091, + "grad_norm": 0.6395941325424317, + "learning_rate": 1.2329467661209734e-06, + "loss": 0.7824, + "step": 7058 + }, + { + "epoch": 0.6366054921765794, + "grad_norm": 1.696449496204711, + "learning_rate": 1.2324072629150788e-06, + "loss": 1.0262, + "step": 7059 + }, + { + "epoch": 0.6366956757000496, + "grad_norm": 1.7916198842646234, + "learning_rate": 1.2318678252044287e-06, + "loss": 1.03, + "step": 7060 + }, + { + "epoch": 0.6367858592235198, + "grad_norm": 1.8705398918492788, + "learning_rate": 1.2313284530350512e-06, + "loss": 1.0403, + "step": 7061 + }, + { + "epoch": 0.6368760427469902, + "grad_norm": 1.6797378177646751, + "learning_rate": 1.230789146452969e-06, + "loss": 0.9208, + "step": 7062 + }, + { + "epoch": 0.6369662262704604, + "grad_norm": 3.2692815578192214, + "learning_rate": 1.2302499055041974e-06, + "loss": 0.9823, + "step": 7063 + }, + { + "epoch": 0.6370564097939306, + "grad_norm": 0.6683347714219194, + "learning_rate": 1.2297107302347488e-06, + "loss": 0.8187, + "step": 7064 + }, + { + "epoch": 0.6371465933174009, + "grad_norm": 1.7016797013986926, + "learning_rate": 1.2291716206906275e-06, + "loss": 1.0533, + "step": 7065 + }, + { + "epoch": 0.6372367768408712, + "grad_norm": 1.8960094230393552, + "learning_rate": 1.2286325769178345e-06, + "loss": 1.048, + "step": 7066 + }, + { + "epoch": 0.6373269603643414, + "grad_norm": 1.5148652522276087, + "learning_rate": 1.2280935989623633e-06, + "loss": 1.0103, + "step": 7067 + }, + { + "epoch": 0.6374171438878117, + "grad_norm": 1.8391677035542122, + "learning_rate": 1.2275546868702017e-06, + "loss": 0.9168, + "step": 7068 + }, + { + "epoch": 0.6375073274112819, + "grad_norm": 2.086361616930378, + "learning_rate": 1.2270158406873341e-06, + "loss": 0.9201, + "step": 7069 + }, + { + "epoch": 0.6375975109347523, + "grad_norm": 1.386462937521031, + "learning_rate": 1.2264770604597363e-06, + "loss": 1.022, + "step": 7070 + }, + { + "epoch": 0.6376876944582225, + "grad_norm": 1.4758392661844366, + "learning_rate": 1.2259383462333819e-06, + "loss": 1.0828, + "step": 7071 + }, + { + "epoch": 0.6377778779816927, + "grad_norm": 1.769005905887628, + "learning_rate": 1.2253996980542359e-06, + "loss": 0.9385, + "step": 7072 + }, + { + "epoch": 0.637868061505163, + "grad_norm": 1.592442069324936, + "learning_rate": 1.2248611159682578e-06, + "loss": 0.9634, + "step": 7073 + }, + { + "epoch": 0.6379582450286333, + "grad_norm": 0.7147403913307715, + "learning_rate": 1.2243226000214044e-06, + "loss": 0.8006, + "step": 7074 + }, + { + "epoch": 0.6380484285521035, + "grad_norm": 1.5428616643516448, + "learning_rate": 1.2237841502596232e-06, + "loss": 0.9786, + "step": 7075 + }, + { + "epoch": 0.6381386120755738, + "grad_norm": 1.5606174572390747, + "learning_rate": 1.2232457667288583e-06, + "loss": 0.9872, + "step": 7076 + }, + { + "epoch": 0.6382287955990441, + "grad_norm": 1.9540223158476693, + "learning_rate": 1.2227074494750476e-06, + "loss": 0.987, + "step": 7077 + }, + { + "epoch": 0.6383189791225143, + "grad_norm": 1.7710775393975475, + "learning_rate": 1.2221691985441238e-06, + "loss": 1.0173, + "step": 7078 + }, + { + "epoch": 0.6384091626459846, + "grad_norm": 1.6264622051821138, + "learning_rate": 1.2216310139820128e-06, + "loss": 0.9304, + "step": 7079 + }, + { + "epoch": 0.6384993461694548, + "grad_norm": 1.6344307359341987, + "learning_rate": 1.2210928958346347e-06, + "loss": 1.0257, + "step": 7080 + }, + { + "epoch": 0.6385895296929252, + "grad_norm": 1.6741374741485935, + "learning_rate": 1.2205548441479065e-06, + "loss": 0.9103, + "step": 7081 + }, + { + "epoch": 0.6386797132163954, + "grad_norm": 1.5961930492879277, + "learning_rate": 1.2200168589677357e-06, + "loss": 0.9161, + "step": 7082 + }, + { + "epoch": 0.6387698967398656, + "grad_norm": 1.529499376709017, + "learning_rate": 1.2194789403400284e-06, + "loss": 0.9489, + "step": 7083 + }, + { + "epoch": 0.6388600802633358, + "grad_norm": 1.9636733586867468, + "learning_rate": 1.2189410883106816e-06, + "loss": 0.961, + "step": 7084 + }, + { + "epoch": 0.6389502637868062, + "grad_norm": 1.532151482361236, + "learning_rate": 1.2184033029255872e-06, + "loss": 1.051, + "step": 7085 + }, + { + "epoch": 0.6390404473102764, + "grad_norm": 0.6510909150887848, + "learning_rate": 1.2178655842306334e-06, + "loss": 0.8177, + "step": 7086 + }, + { + "epoch": 0.6391306308337467, + "grad_norm": 1.7058310752497146, + "learning_rate": 1.2173279322716999e-06, + "loss": 0.9788, + "step": 7087 + }, + { + "epoch": 0.6392208143572169, + "grad_norm": 1.6034111704474006, + "learning_rate": 1.216790347094663e-06, + "loss": 0.985, + "step": 7088 + }, + { + "epoch": 0.6393109978806872, + "grad_norm": 2.1773472189490666, + "learning_rate": 1.2162528287453927e-06, + "loss": 1.0224, + "step": 7089 + }, + { + "epoch": 0.6394011814041575, + "grad_norm": 1.734591152708982, + "learning_rate": 1.215715377269752e-06, + "loss": 0.9882, + "step": 7090 + }, + { + "epoch": 0.6394913649276277, + "grad_norm": 1.8366588068388334, + "learning_rate": 1.2151779927136003e-06, + "loss": 0.9654, + "step": 7091 + }, + { + "epoch": 0.6395815484510979, + "grad_norm": 3.1608762168506304, + "learning_rate": 1.2146406751227893e-06, + "loss": 1.0594, + "step": 7092 + }, + { + "epoch": 0.6396717319745683, + "grad_norm": 1.8533471757499924, + "learning_rate": 1.214103424543167e-06, + "loss": 1.0139, + "step": 7093 + }, + { + "epoch": 0.6397619154980385, + "grad_norm": 1.5422250234875872, + "learning_rate": 1.2135662410205735e-06, + "loss": 0.8961, + "step": 7094 + }, + { + "epoch": 0.6398520990215087, + "grad_norm": 1.6920197323579653, + "learning_rate": 1.2130291246008444e-06, + "loss": 0.8989, + "step": 7095 + }, + { + "epoch": 0.639942282544979, + "grad_norm": 1.5360914492752373, + "learning_rate": 1.21249207532981e-06, + "loss": 0.9001, + "step": 7096 + }, + { + "epoch": 0.6400324660684493, + "grad_norm": 2.420250132220769, + "learning_rate": 1.2119550932532936e-06, + "loss": 0.9011, + "step": 7097 + }, + { + "epoch": 0.6401226495919196, + "grad_norm": 1.536374838764088, + "learning_rate": 1.2114181784171144e-06, + "loss": 0.9543, + "step": 7098 + }, + { + "epoch": 0.6402128331153898, + "grad_norm": 2.252779727040723, + "learning_rate": 1.2108813308670837e-06, + "loss": 0.8837, + "step": 7099 + }, + { + "epoch": 0.6403030166388601, + "grad_norm": 1.9108696289985172, + "learning_rate": 1.2103445506490099e-06, + "loss": 0.9834, + "step": 7100 + }, + { + "epoch": 0.6403932001623304, + "grad_norm": 0.7125786503665553, + "learning_rate": 1.209807837808693e-06, + "loss": 0.8293, + "step": 7101 + }, + { + "epoch": 0.6404833836858006, + "grad_norm": 1.918000075323364, + "learning_rate": 1.2092711923919282e-06, + "loss": 1.0213, + "step": 7102 + }, + { + "epoch": 0.6405735672092708, + "grad_norm": 1.384882310005299, + "learning_rate": 1.2087346144445053e-06, + "loss": 0.9422, + "step": 7103 + }, + { + "epoch": 0.6406637507327412, + "grad_norm": 2.253670982630166, + "learning_rate": 1.2081981040122081e-06, + "loss": 0.9948, + "step": 7104 + }, + { + "epoch": 0.6407539342562114, + "grad_norm": 1.5667433530557844, + "learning_rate": 1.2076616611408148e-06, + "loss": 0.9363, + "step": 7105 + }, + { + "epoch": 0.6408441177796816, + "grad_norm": 1.7160310914306278, + "learning_rate": 1.2071252858760972e-06, + "loss": 1.0171, + "step": 7106 + }, + { + "epoch": 0.6409343013031519, + "grad_norm": 1.8128190253103649, + "learning_rate": 1.2065889782638218e-06, + "loss": 0.9651, + "step": 7107 + }, + { + "epoch": 0.6410244848266222, + "grad_norm": 1.7549379029115422, + "learning_rate": 1.2060527383497506e-06, + "loss": 0.9625, + "step": 7108 + }, + { + "epoch": 0.6411146683500925, + "grad_norm": 1.514778608758067, + "learning_rate": 1.2055165661796363e-06, + "loss": 1.0956, + "step": 7109 + }, + { + "epoch": 0.6412048518735627, + "grad_norm": 1.539186269363021, + "learning_rate": 1.2049804617992303e-06, + "loss": 1.0531, + "step": 7110 + }, + { + "epoch": 0.6412950353970329, + "grad_norm": 0.6242270458372007, + "learning_rate": 1.204444425254275e-06, + "loss": 0.8296, + "step": 7111 + }, + { + "epoch": 0.6413852189205033, + "grad_norm": 1.649763830218012, + "learning_rate": 1.203908456590507e-06, + "loss": 0.9629, + "step": 7112 + }, + { + "epoch": 0.6414754024439735, + "grad_norm": 2.027080305619313, + "learning_rate": 1.20337255585366e-06, + "loss": 0.9587, + "step": 7113 + }, + { + "epoch": 0.6415655859674437, + "grad_norm": 0.610143665765189, + "learning_rate": 1.2028367230894582e-06, + "loss": 0.7755, + "step": 7114 + }, + { + "epoch": 0.641655769490914, + "grad_norm": 1.6859317132463392, + "learning_rate": 1.2023009583436237e-06, + "loss": 0.9019, + "step": 7115 + }, + { + "epoch": 0.6417459530143843, + "grad_norm": 1.976164706161173, + "learning_rate": 1.2017652616618698e-06, + "loss": 1.049, + "step": 7116 + }, + { + "epoch": 0.6418361365378545, + "grad_norm": 1.803606047830871, + "learning_rate": 1.2012296330899048e-06, + "loss": 0.9775, + "step": 7117 + }, + { + "epoch": 0.6419263200613248, + "grad_norm": 1.8124896082086932, + "learning_rate": 1.200694072673432e-06, + "loss": 1.0123, + "step": 7118 + }, + { + "epoch": 0.642016503584795, + "grad_norm": 1.5767222298045134, + "learning_rate": 1.200158580458148e-06, + "loss": 1.011, + "step": 7119 + }, + { + "epoch": 0.6421066871082654, + "grad_norm": 1.6241231900315676, + "learning_rate": 1.1996231564897448e-06, + "loss": 0.9557, + "step": 7120 + }, + { + "epoch": 0.6421968706317356, + "grad_norm": 11.209256121194098, + "learning_rate": 1.1990878008139062e-06, + "loss": 0.9376, + "step": 7121 + }, + { + "epoch": 0.6422870541552058, + "grad_norm": 1.4968706708299953, + "learning_rate": 1.1985525134763132e-06, + "loss": 1.0104, + "step": 7122 + }, + { + "epoch": 0.6423772376786762, + "grad_norm": 1.5562037449592503, + "learning_rate": 1.1980172945226393e-06, + "loss": 1.0261, + "step": 7123 + }, + { + "epoch": 0.6424674212021464, + "grad_norm": 2.5107267090696066, + "learning_rate": 1.197482143998551e-06, + "loss": 0.9853, + "step": 7124 + }, + { + "epoch": 0.6425576047256166, + "grad_norm": 1.6396995477395546, + "learning_rate": 1.196947061949712e-06, + "loss": 0.9689, + "step": 7125 + }, + { + "epoch": 0.6426477882490869, + "grad_norm": 1.6701958325627888, + "learning_rate": 1.1964120484217768e-06, + "loss": 1.0412, + "step": 7126 + }, + { + "epoch": 0.6427379717725572, + "grad_norm": 0.6996448007343953, + "learning_rate": 1.1958771034603975e-06, + "loss": 0.8001, + "step": 7127 + }, + { + "epoch": 0.6428281552960274, + "grad_norm": 2.2850542094605415, + "learning_rate": 1.1953422271112175e-06, + "loss": 0.9225, + "step": 7128 + }, + { + "epoch": 0.6429183388194977, + "grad_norm": 2.2593014569270533, + "learning_rate": 1.1948074194198748e-06, + "loss": 0.9686, + "step": 7129 + }, + { + "epoch": 0.6430085223429679, + "grad_norm": 1.8157204266418907, + "learning_rate": 1.1942726804320033e-06, + "loss": 0.8965, + "step": 7130 + }, + { + "epoch": 0.6430987058664382, + "grad_norm": 2.1422861155843282, + "learning_rate": 1.1937380101932295e-06, + "loss": 0.9234, + "step": 7131 + }, + { + "epoch": 0.6431888893899085, + "grad_norm": 1.4489898912223307, + "learning_rate": 1.1932034087491745e-06, + "loss": 0.9786, + "step": 7132 + }, + { + "epoch": 0.6432790729133787, + "grad_norm": 2.0274720491300644, + "learning_rate": 1.1926688761454531e-06, + "loss": 0.8958, + "step": 7133 + }, + { + "epoch": 0.643369256436849, + "grad_norm": 1.2025566244771397, + "learning_rate": 1.1921344124276746e-06, + "loss": 0.9994, + "step": 7134 + }, + { + "epoch": 0.6434594399603193, + "grad_norm": 1.4850471179564166, + "learning_rate": 1.1916000176414431e-06, + "loss": 1.0052, + "step": 7135 + }, + { + "epoch": 0.6435496234837895, + "grad_norm": 1.7167481295074232, + "learning_rate": 1.1910656918323546e-06, + "loss": 1.0171, + "step": 7136 + }, + { + "epoch": 0.6436398070072598, + "grad_norm": 1.5994147064658613, + "learning_rate": 1.1905314350460024e-06, + "loss": 1.0106, + "step": 7137 + }, + { + "epoch": 0.64372999053073, + "grad_norm": 1.7415633516225713, + "learning_rate": 1.1899972473279717e-06, + "loss": 1.0363, + "step": 7138 + }, + { + "epoch": 0.6438201740542003, + "grad_norm": 1.7125376704172866, + "learning_rate": 1.1894631287238414e-06, + "loss": 0.9729, + "step": 7139 + }, + { + "epoch": 0.6439103575776706, + "grad_norm": 1.3785073889477948, + "learning_rate": 1.188929079279187e-06, + "loss": 0.9928, + "step": 7140 + }, + { + "epoch": 0.6440005411011408, + "grad_norm": 1.5184815220517174, + "learning_rate": 1.1883950990395751e-06, + "loss": 0.9175, + "step": 7141 + }, + { + "epoch": 0.644090724624611, + "grad_norm": 1.8163436386638403, + "learning_rate": 1.187861188050569e-06, + "loss": 0.9193, + "step": 7142 + }, + { + "epoch": 0.6441809081480814, + "grad_norm": 1.37358666522408, + "learning_rate": 1.187327346357724e-06, + "loss": 0.9708, + "step": 7143 + }, + { + "epoch": 0.6442710916715516, + "grad_norm": 1.8833677494089012, + "learning_rate": 1.1867935740065912e-06, + "loss": 1.0065, + "step": 7144 + }, + { + "epoch": 0.6443612751950218, + "grad_norm": 1.7967542524816895, + "learning_rate": 1.1862598710427148e-06, + "loss": 0.8742, + "step": 7145 + }, + { + "epoch": 0.6444514587184921, + "grad_norm": 2.066251983364697, + "learning_rate": 1.1857262375116328e-06, + "loss": 0.943, + "step": 7146 + }, + { + "epoch": 0.6445416422419624, + "grad_norm": 1.8183902528888423, + "learning_rate": 1.1851926734588783e-06, + "loss": 0.988, + "step": 7147 + }, + { + "epoch": 0.6446318257654327, + "grad_norm": 3.609855586832815, + "learning_rate": 1.184659178929977e-06, + "loss": 1.0038, + "step": 7148 + }, + { + "epoch": 0.6447220092889029, + "grad_norm": 1.4766349976518514, + "learning_rate": 1.1841257539704513e-06, + "loss": 0.9753, + "step": 7149 + }, + { + "epoch": 0.6448121928123732, + "grad_norm": 1.5360522579502471, + "learning_rate": 1.1835923986258146e-06, + "loss": 1.0013, + "step": 7150 + }, + { + "epoch": 0.6449023763358435, + "grad_norm": 1.3950202576044881, + "learning_rate": 1.1830591129415754e-06, + "loss": 0.9401, + "step": 7151 + }, + { + "epoch": 0.6449925598593137, + "grad_norm": 1.7348580589827818, + "learning_rate": 1.182525896963238e-06, + "loss": 0.928, + "step": 7152 + }, + { + "epoch": 0.6450827433827839, + "grad_norm": 2.02832242830595, + "learning_rate": 1.181992750736298e-06, + "loss": 0.9128, + "step": 7153 + }, + { + "epoch": 0.6451729269062543, + "grad_norm": 2.0153875876487097, + "learning_rate": 1.1814596743062474e-06, + "loss": 0.9779, + "step": 7154 + }, + { + "epoch": 0.6452631104297245, + "grad_norm": 1.6942377155427684, + "learning_rate": 1.1809266677185711e-06, + "loss": 1.0053, + "step": 7155 + }, + { + "epoch": 0.6453532939531947, + "grad_norm": 0.6541170470944132, + "learning_rate": 1.180393731018747e-06, + "loss": 0.8492, + "step": 7156 + }, + { + "epoch": 0.645443477476665, + "grad_norm": 0.5626729858163746, + "learning_rate": 1.1798608642522498e-06, + "loss": 0.7882, + "step": 7157 + }, + { + "epoch": 0.6455336610001353, + "grad_norm": 4.795232734004326, + "learning_rate": 1.1793280674645454e-06, + "loss": 0.9639, + "step": 7158 + }, + { + "epoch": 0.6456238445236056, + "grad_norm": 1.5412051394973154, + "learning_rate": 1.1787953407010954e-06, + "loss": 1.056, + "step": 7159 + }, + { + "epoch": 0.6457140280470758, + "grad_norm": 1.5228903545954158, + "learning_rate": 1.1782626840073554e-06, + "loss": 0.9831, + "step": 7160 + }, + { + "epoch": 0.645804211570546, + "grad_norm": 1.7527673502694339, + "learning_rate": 1.1777300974287738e-06, + "loss": 0.8855, + "step": 7161 + }, + { + "epoch": 0.6458943950940164, + "grad_norm": 1.9846423712398618, + "learning_rate": 1.1771975810107947e-06, + "loss": 1.0112, + "step": 7162 + }, + { + "epoch": 0.6459845786174866, + "grad_norm": 2.1299958686609455, + "learning_rate": 1.1766651347988542e-06, + "loss": 1.0554, + "step": 7163 + }, + { + "epoch": 0.6460747621409568, + "grad_norm": 2.02759269700006, + "learning_rate": 1.1761327588383848e-06, + "loss": 1.0895, + "step": 7164 + }, + { + "epoch": 0.6461649456644271, + "grad_norm": 1.427212600236704, + "learning_rate": 1.1756004531748105e-06, + "loss": 0.9946, + "step": 7165 + }, + { + "epoch": 0.6462551291878974, + "grad_norm": 1.5929606109605785, + "learning_rate": 1.1750682178535521e-06, + "loss": 1.0164, + "step": 7166 + }, + { + "epoch": 0.6463453127113676, + "grad_norm": 1.8643062179619703, + "learning_rate": 1.1745360529200218e-06, + "loss": 0.9286, + "step": 7167 + }, + { + "epoch": 0.6464354962348379, + "grad_norm": 1.6161300961655078, + "learning_rate": 1.1740039584196261e-06, + "loss": 1.0103, + "step": 7168 + }, + { + "epoch": 0.6465256797583081, + "grad_norm": 1.5051545746210944, + "learning_rate": 1.1734719343977683e-06, + "loss": 1.0128, + "step": 7169 + }, + { + "epoch": 0.6466158632817784, + "grad_norm": 1.6695189229664533, + "learning_rate": 1.1729399808998416e-06, + "loss": 0.9237, + "step": 7170 + }, + { + "epoch": 0.6467060468052487, + "grad_norm": 4.3467372803910305, + "learning_rate": 1.1724080979712368e-06, + "loss": 0.875, + "step": 7171 + }, + { + "epoch": 0.6467962303287189, + "grad_norm": 1.96241965645744, + "learning_rate": 1.1718762856573365e-06, + "loss": 1.0543, + "step": 7172 + }, + { + "epoch": 0.6468864138521893, + "grad_norm": 1.8178508305553784, + "learning_rate": 1.1713445440035172e-06, + "loss": 0.9673, + "step": 7173 + }, + { + "epoch": 0.6469765973756595, + "grad_norm": 1.849522106202515, + "learning_rate": 1.1708128730551506e-06, + "loss": 1.0173, + "step": 7174 + }, + { + "epoch": 0.6470667808991297, + "grad_norm": 5.316066666632121, + "learning_rate": 1.1702812728576019e-06, + "loss": 0.9687, + "step": 7175 + }, + { + "epoch": 0.6471569644226, + "grad_norm": 2.4005491423198118, + "learning_rate": 1.1697497434562303e-06, + "loss": 1.0103, + "step": 7176 + }, + { + "epoch": 0.6472471479460703, + "grad_norm": 1.738933586559054, + "learning_rate": 1.1692182848963885e-06, + "loss": 1.0151, + "step": 7177 + }, + { + "epoch": 0.6473373314695405, + "grad_norm": 1.9600508307658224, + "learning_rate": 1.1686868972234227e-06, + "loss": 0.9473, + "step": 7178 + }, + { + "epoch": 0.6474275149930108, + "grad_norm": 1.5425191103402156, + "learning_rate": 1.1681555804826755e-06, + "loss": 0.9919, + "step": 7179 + }, + { + "epoch": 0.647517698516481, + "grad_norm": 1.581065921849563, + "learning_rate": 1.1676243347194806e-06, + "loss": 0.9189, + "step": 7180 + }, + { + "epoch": 0.6476078820399513, + "grad_norm": 1.8856290272640304, + "learning_rate": 1.167093159979167e-06, + "loss": 0.9976, + "step": 7181 + }, + { + "epoch": 0.6476980655634216, + "grad_norm": 1.784138462402594, + "learning_rate": 1.1665620563070575e-06, + "loss": 1.011, + "step": 7182 + }, + { + "epoch": 0.6477882490868918, + "grad_norm": 1.7734387099476185, + "learning_rate": 1.1660310237484691e-06, + "loss": 1.092, + "step": 7183 + }, + { + "epoch": 0.647878432610362, + "grad_norm": 1.630544154603703, + "learning_rate": 1.165500062348713e-06, + "loss": 0.9256, + "step": 7184 + }, + { + "epoch": 0.6479686161338324, + "grad_norm": 1.7107242120329098, + "learning_rate": 1.164969172153091e-06, + "loss": 1.0245, + "step": 7185 + }, + { + "epoch": 0.6480587996573026, + "grad_norm": 1.526170050349242, + "learning_rate": 1.1644383532069055e-06, + "loss": 0.9992, + "step": 7186 + }, + { + "epoch": 0.6481489831807729, + "grad_norm": 1.4700898085946525, + "learning_rate": 1.1639076055554454e-06, + "loss": 1.0221, + "step": 7187 + }, + { + "epoch": 0.6482391667042431, + "grad_norm": 1.9667136545009951, + "learning_rate": 1.163376929244e-06, + "loss": 0.9747, + "step": 7188 + }, + { + "epoch": 0.6483293502277134, + "grad_norm": 1.7930473360279868, + "learning_rate": 1.1628463243178472e-06, + "loss": 0.946, + "step": 7189 + }, + { + "epoch": 0.6484195337511837, + "grad_norm": 1.8043022239921254, + "learning_rate": 1.1623157908222623e-06, + "loss": 1.0417, + "step": 7190 + }, + { + "epoch": 0.6485097172746539, + "grad_norm": 1.6862066752281175, + "learning_rate": 1.1617853288025129e-06, + "loss": 1.0198, + "step": 7191 + }, + { + "epoch": 0.6485999007981241, + "grad_norm": 1.5032676232575488, + "learning_rate": 1.1612549383038612e-06, + "loss": 0.9559, + "step": 7192 + }, + { + "epoch": 0.6486900843215945, + "grad_norm": 1.480115621921883, + "learning_rate": 1.1607246193715629e-06, + "loss": 0.9619, + "step": 7193 + }, + { + "epoch": 0.6487802678450647, + "grad_norm": 2.890384148538572, + "learning_rate": 1.1601943720508684e-06, + "loss": 0.9976, + "step": 7194 + }, + { + "epoch": 0.6488704513685349, + "grad_norm": 1.4587228073292187, + "learning_rate": 1.1596641963870202e-06, + "loss": 1.0272, + "step": 7195 + }, + { + "epoch": 0.6489606348920053, + "grad_norm": 1.6002229624607915, + "learning_rate": 1.1591340924252561e-06, + "loss": 0.9948, + "step": 7196 + }, + { + "epoch": 0.6490508184154755, + "grad_norm": 0.7562163936619738, + "learning_rate": 1.158604060210808e-06, + "loss": 0.8611, + "step": 7197 + }, + { + "epoch": 0.6491410019389457, + "grad_norm": 1.5879058649474307, + "learning_rate": 1.1580740997889008e-06, + "loss": 1.0425, + "step": 7198 + }, + { + "epoch": 0.649231185462416, + "grad_norm": 2.2585411084115803, + "learning_rate": 1.1575442112047544e-06, + "loss": 0.9787, + "step": 7199 + }, + { + "epoch": 0.6493213689858863, + "grad_norm": 2.0186267354552516, + "learning_rate": 1.1570143945035797e-06, + "loss": 0.9194, + "step": 7200 + }, + { + "epoch": 0.6494115525093566, + "grad_norm": 2.0248296359889277, + "learning_rate": 1.1564846497305864e-06, + "loss": 0.9412, + "step": 7201 + }, + { + "epoch": 0.6495017360328268, + "grad_norm": 1.7572321633163492, + "learning_rate": 1.1559549769309726e-06, + "loss": 0.9591, + "step": 7202 + }, + { + "epoch": 0.649591919556297, + "grad_norm": 1.6458337115838626, + "learning_rate": 1.1554253761499358e-06, + "loss": 0.9768, + "step": 7203 + }, + { + "epoch": 0.6496821030797674, + "grad_norm": 2.155089909388378, + "learning_rate": 1.1548958474326617e-06, + "loss": 1.0199, + "step": 7204 + }, + { + "epoch": 0.6497722866032376, + "grad_norm": 2.9739289736036167, + "learning_rate": 1.154366390824334e-06, + "loss": 0.915, + "step": 7205 + }, + { + "epoch": 0.6498624701267078, + "grad_norm": 0.7681142085212025, + "learning_rate": 1.1538370063701287e-06, + "loss": 0.7951, + "step": 7206 + }, + { + "epoch": 0.6499526536501781, + "grad_norm": 1.5349646684140383, + "learning_rate": 1.1533076941152153e-06, + "loss": 0.942, + "step": 7207 + }, + { + "epoch": 0.6500428371736484, + "grad_norm": 1.906026232984209, + "learning_rate": 1.1527784541047583e-06, + "loss": 0.9119, + "step": 7208 + }, + { + "epoch": 0.6501330206971186, + "grad_norm": 2.9116652704623855, + "learning_rate": 1.1522492863839152e-06, + "loss": 0.9356, + "step": 7209 + }, + { + "epoch": 0.6502232042205889, + "grad_norm": 1.8857712666873658, + "learning_rate": 1.1517201909978382e-06, + "loss": 0.9455, + "step": 7210 + }, + { + "epoch": 0.6503133877440591, + "grad_norm": 0.5990696071149443, + "learning_rate": 1.151191167991671e-06, + "loss": 0.816, + "step": 7211 + }, + { + "epoch": 0.6504035712675295, + "grad_norm": 1.574997550720248, + "learning_rate": 1.1506622174105536e-06, + "loss": 0.9901, + "step": 7212 + }, + { + "epoch": 0.6504937547909997, + "grad_norm": 1.7716490344117828, + "learning_rate": 1.1501333392996194e-06, + "loss": 1.0157, + "step": 7213 + }, + { + "epoch": 0.6505839383144699, + "grad_norm": 0.6561415219340672, + "learning_rate": 1.1496045337039943e-06, + "loss": 0.8224, + "step": 7214 + }, + { + "epoch": 0.6506741218379402, + "grad_norm": 1.511728291641601, + "learning_rate": 1.1490758006687995e-06, + "loss": 0.9382, + "step": 7215 + }, + { + "epoch": 0.6507643053614105, + "grad_norm": 1.879018165602078, + "learning_rate": 1.1485471402391502e-06, + "loss": 0.9608, + "step": 7216 + }, + { + "epoch": 0.6508544888848807, + "grad_norm": 1.6907964933219592, + "learning_rate": 1.1480185524601522e-06, + "loss": 0.9546, + "step": 7217 + }, + { + "epoch": 0.650944672408351, + "grad_norm": 1.5222982988969622, + "learning_rate": 1.1474900373769108e-06, + "loss": 0.9871, + "step": 7218 + }, + { + "epoch": 0.6510348559318213, + "grad_norm": 0.5670375331468532, + "learning_rate": 1.1469615950345184e-06, + "loss": 0.7618, + "step": 7219 + }, + { + "epoch": 0.6511250394552915, + "grad_norm": 0.7276356063381688, + "learning_rate": 1.1464332254780678e-06, + "loss": 0.7996, + "step": 7220 + }, + { + "epoch": 0.6512152229787618, + "grad_norm": 1.6427975756348434, + "learning_rate": 1.1459049287526404e-06, + "loss": 0.9656, + "step": 7221 + }, + { + "epoch": 0.651305406502232, + "grad_norm": 1.5286882731467193, + "learning_rate": 1.1453767049033137e-06, + "loss": 0.9401, + "step": 7222 + }, + { + "epoch": 0.6513955900257024, + "grad_norm": 1.4421362540252396, + "learning_rate": 1.1448485539751586e-06, + "loss": 1.0234, + "step": 7223 + }, + { + "epoch": 0.6514857735491726, + "grad_norm": 1.5123175487343388, + "learning_rate": 1.1443204760132408e-06, + "loss": 0.9518, + "step": 7224 + }, + { + "epoch": 0.6515759570726428, + "grad_norm": 2.0018720210324457, + "learning_rate": 1.1437924710626185e-06, + "loss": 0.9627, + "step": 7225 + }, + { + "epoch": 0.651666140596113, + "grad_norm": 0.651704783650101, + "learning_rate": 1.1432645391683429e-06, + "loss": 0.784, + "step": 7226 + }, + { + "epoch": 0.6517563241195834, + "grad_norm": 2.55890039834041, + "learning_rate": 1.1427366803754609e-06, + "loss": 1.0168, + "step": 7227 + }, + { + "epoch": 0.6518465076430536, + "grad_norm": 1.508857009493496, + "learning_rate": 1.142208894729012e-06, + "loss": 0.9952, + "step": 7228 + }, + { + "epoch": 0.6519366911665239, + "grad_norm": 1.527189046922575, + "learning_rate": 1.1416811822740301e-06, + "loss": 1.021, + "step": 7229 + }, + { + "epoch": 0.6520268746899941, + "grad_norm": 1.8709456797329074, + "learning_rate": 1.1411535430555428e-06, + "loss": 1.0314, + "step": 7230 + }, + { + "epoch": 0.6521170582134644, + "grad_norm": 1.5862306220612266, + "learning_rate": 1.1406259771185705e-06, + "loss": 1.0355, + "step": 7231 + }, + { + "epoch": 0.6522072417369347, + "grad_norm": 1.3387080236384692, + "learning_rate": 1.1400984845081282e-06, + "loss": 0.9412, + "step": 7232 + }, + { + "epoch": 0.6522974252604049, + "grad_norm": 1.6713678892179928, + "learning_rate": 1.139571065269226e-06, + "loss": 1.0626, + "step": 7233 + }, + { + "epoch": 0.6523876087838751, + "grad_norm": 1.609891197603533, + "learning_rate": 1.139043719446863e-06, + "loss": 1.0287, + "step": 7234 + }, + { + "epoch": 0.6524777923073455, + "grad_norm": 1.512706896920418, + "learning_rate": 1.1385164470860385e-06, + "loss": 1.0559, + "step": 7235 + }, + { + "epoch": 0.6525679758308157, + "grad_norm": 1.8031640165924014, + "learning_rate": 1.1379892482317403e-06, + "loss": 0.9829, + "step": 7236 + }, + { + "epoch": 0.652658159354286, + "grad_norm": 1.6519316752392676, + "learning_rate": 1.1374621229289524e-06, + "loss": 0.9692, + "step": 7237 + }, + { + "epoch": 0.6527483428777562, + "grad_norm": 1.7835424963078015, + "learning_rate": 1.1369350712226525e-06, + "loss": 0.9646, + "step": 7238 + }, + { + "epoch": 0.6528385264012265, + "grad_norm": 1.9952434363544593, + "learning_rate": 1.136408093157811e-06, + "loss": 0.9334, + "step": 7239 + }, + { + "epoch": 0.6529287099246968, + "grad_norm": 2.1626414787450043, + "learning_rate": 1.1358811887793935e-06, + "loss": 0.9577, + "step": 7240 + }, + { + "epoch": 0.653018893448167, + "grad_norm": 1.492686962002205, + "learning_rate": 1.135354358132356e-06, + "loss": 1.0156, + "step": 7241 + }, + { + "epoch": 0.6531090769716373, + "grad_norm": 2.002856003527942, + "learning_rate": 1.1348276012616542e-06, + "loss": 1.0209, + "step": 7242 + }, + { + "epoch": 0.6531992604951076, + "grad_norm": 1.77472851517254, + "learning_rate": 1.134300918212231e-06, + "loss": 0.936, + "step": 7243 + }, + { + "epoch": 0.6532894440185778, + "grad_norm": 15.541465669343005, + "learning_rate": 1.133774309029027e-06, + "loss": 0.9914, + "step": 7244 + }, + { + "epoch": 0.653379627542048, + "grad_norm": 1.7683427537289051, + "learning_rate": 1.133247773756975e-06, + "loss": 1.025, + "step": 7245 + }, + { + "epoch": 0.6534698110655184, + "grad_norm": 0.6562029730886131, + "learning_rate": 1.1327213124410024e-06, + "loss": 0.8549, + "step": 7246 + }, + { + "epoch": 0.6535599945889886, + "grad_norm": 5.151969473825376, + "learning_rate": 1.1321949251260292e-06, + "loss": 1.0221, + "step": 7247 + }, + { + "epoch": 0.6536501781124588, + "grad_norm": 1.9541000373830568, + "learning_rate": 1.1316686118569712e-06, + "loss": 0.8727, + "step": 7248 + }, + { + "epoch": 0.6537403616359291, + "grad_norm": 1.5834747887047722, + "learning_rate": 1.1311423726787335e-06, + "loss": 0.8957, + "step": 7249 + }, + { + "epoch": 0.6538305451593994, + "grad_norm": 1.7577741152493054, + "learning_rate": 1.130616207636221e-06, + "loss": 0.9521, + "step": 7250 + }, + { + "epoch": 0.6539207286828697, + "grad_norm": 2.019025314973456, + "learning_rate": 1.1300901167743263e-06, + "loss": 0.9877, + "step": 7251 + }, + { + "epoch": 0.6540109122063399, + "grad_norm": 2.037780475619291, + "learning_rate": 1.12956410013794e-06, + "loss": 0.9811, + "step": 7252 + }, + { + "epoch": 0.6541010957298101, + "grad_norm": 1.7837982020451493, + "learning_rate": 1.1290381577719436e-06, + "loss": 0.9725, + "step": 7253 + }, + { + "epoch": 0.6541912792532805, + "grad_norm": 2.072468797106009, + "learning_rate": 1.1285122897212143e-06, + "loss": 0.8841, + "step": 7254 + }, + { + "epoch": 0.6542814627767507, + "grad_norm": 0.7441635729037313, + "learning_rate": 1.1279864960306228e-06, + "loss": 0.8554, + "step": 7255 + }, + { + "epoch": 0.6543716463002209, + "grad_norm": 1.9410513662884954, + "learning_rate": 1.1274607767450297e-06, + "loss": 1.0572, + "step": 7256 + }, + { + "epoch": 0.6544618298236912, + "grad_norm": 1.463878262549833, + "learning_rate": 1.126935131909296e-06, + "loss": 0.9567, + "step": 7257 + }, + { + "epoch": 0.6545520133471615, + "grad_norm": 1.3650608198009626, + "learning_rate": 1.1264095615682693e-06, + "loss": 0.9274, + "step": 7258 + }, + { + "epoch": 0.6546421968706317, + "grad_norm": 2.6751770673955293, + "learning_rate": 1.1258840657667973e-06, + "loss": 0.9002, + "step": 7259 + }, + { + "epoch": 0.654732380394102, + "grad_norm": 2.811819799610163, + "learning_rate": 1.125358644549716e-06, + "loss": 0.9105, + "step": 7260 + }, + { + "epoch": 0.6548225639175722, + "grad_norm": 1.2605422826966435, + "learning_rate": 1.1248332979618578e-06, + "loss": 0.8409, + "step": 7261 + }, + { + "epoch": 0.6549127474410426, + "grad_norm": 1.3016639898877496, + "learning_rate": 1.1243080260480482e-06, + "loss": 0.8751, + "step": 7262 + }, + { + "epoch": 0.6550029309645128, + "grad_norm": 2.79846214568991, + "learning_rate": 1.1237828288531063e-06, + "loss": 1.0371, + "step": 7263 + }, + { + "epoch": 0.655093114487983, + "grad_norm": 1.6536618289524319, + "learning_rate": 1.1232577064218449e-06, + "loss": 0.8995, + "step": 7264 + }, + { + "epoch": 0.6551832980114533, + "grad_norm": 1.7855502639100758, + "learning_rate": 1.1227326587990711e-06, + "loss": 1.0582, + "step": 7265 + }, + { + "epoch": 0.6552734815349236, + "grad_norm": 1.3632990007966412, + "learning_rate": 1.1222076860295832e-06, + "loss": 0.9635, + "step": 7266 + }, + { + "epoch": 0.6553636650583938, + "grad_norm": 2.030812025536274, + "learning_rate": 1.1216827881581756e-06, + "loss": 1.0026, + "step": 7267 + }, + { + "epoch": 0.6554538485818641, + "grad_norm": 1.858410945976526, + "learning_rate": 1.1211579652296355e-06, + "loss": 0.9738, + "step": 7268 + }, + { + "epoch": 0.6555440321053344, + "grad_norm": 2.1041290607176393, + "learning_rate": 1.1206332172887438e-06, + "loss": 1.0657, + "step": 7269 + }, + { + "epoch": 0.6556342156288046, + "grad_norm": 3.732050854793336, + "learning_rate": 1.1201085443802756e-06, + "loss": 0.9921, + "step": 7270 + }, + { + "epoch": 0.6557243991522749, + "grad_norm": 1.6209721898841472, + "learning_rate": 1.1195839465489964e-06, + "loss": 0.9489, + "step": 7271 + }, + { + "epoch": 0.6558145826757451, + "grad_norm": 2.4038744012200013, + "learning_rate": 1.1190594238396708e-06, + "loss": 1.0232, + "step": 7272 + }, + { + "epoch": 0.6559047661992154, + "grad_norm": 1.349185595521958, + "learning_rate": 1.1185349762970515e-06, + "loss": 0.9532, + "step": 7273 + }, + { + "epoch": 0.6559949497226857, + "grad_norm": 1.4151771325357212, + "learning_rate": 1.1180106039658896e-06, + "loss": 0.986, + "step": 7274 + }, + { + "epoch": 0.6560851332461559, + "grad_norm": 0.6137840736626806, + "learning_rate": 1.117486306890925e-06, + "loss": 0.7951, + "step": 7275 + }, + { + "epoch": 0.6561753167696261, + "grad_norm": 1.8659768109899495, + "learning_rate": 1.116962085116896e-06, + "loss": 0.9308, + "step": 7276 + }, + { + "epoch": 0.6562655002930965, + "grad_norm": 0.664256864040632, + "learning_rate": 1.1164379386885302e-06, + "loss": 0.755, + "step": 7277 + }, + { + "epoch": 0.6563556838165667, + "grad_norm": 1.882125285876538, + "learning_rate": 1.1159138676505516e-06, + "loss": 1.026, + "step": 7278 + }, + { + "epoch": 0.656445867340037, + "grad_norm": 1.963375202546401, + "learning_rate": 1.1153898720476761e-06, + "loss": 1.0394, + "step": 7279 + }, + { + "epoch": 0.6565360508635072, + "grad_norm": 1.7485126578069454, + "learning_rate": 1.114865951924615e-06, + "loss": 0.971, + "step": 7280 + }, + { + "epoch": 0.6566262343869775, + "grad_norm": 1.4932607416123937, + "learning_rate": 1.1143421073260721e-06, + "loss": 0.924, + "step": 7281 + }, + { + "epoch": 0.6567164179104478, + "grad_norm": 1.9165341981085366, + "learning_rate": 1.1138183382967432e-06, + "loss": 0.9169, + "step": 7282 + }, + { + "epoch": 0.656806601433918, + "grad_norm": 1.7024661548404054, + "learning_rate": 1.11329464488132e-06, + "loss": 0.889, + "step": 7283 + }, + { + "epoch": 0.6568967849573882, + "grad_norm": 1.6898785535073957, + "learning_rate": 1.112771027124487e-06, + "loss": 0.9562, + "step": 7284 + }, + { + "epoch": 0.6569869684808586, + "grad_norm": 1.711447234842353, + "learning_rate": 1.112247485070922e-06, + "loss": 0.9921, + "step": 7285 + }, + { + "epoch": 0.6570771520043288, + "grad_norm": 1.6089116142326525, + "learning_rate": 1.1117240187652968e-06, + "loss": 0.9848, + "step": 7286 + }, + { + "epoch": 0.657167335527799, + "grad_norm": 1.671202337117476, + "learning_rate": 1.1112006282522767e-06, + "loss": 1.0584, + "step": 7287 + }, + { + "epoch": 0.6572575190512693, + "grad_norm": 2.028616386333703, + "learning_rate": 1.1106773135765183e-06, + "loss": 0.9635, + "step": 7288 + }, + { + "epoch": 0.6573477025747396, + "grad_norm": 1.4583080107897288, + "learning_rate": 1.110154074782677e-06, + "loss": 0.9545, + "step": 7289 + }, + { + "epoch": 0.6574378860982099, + "grad_norm": 1.774010653871286, + "learning_rate": 1.1096309119153948e-06, + "loss": 0.9797, + "step": 7290 + }, + { + "epoch": 0.6575280696216801, + "grad_norm": 0.6321163115535114, + "learning_rate": 1.1091078250193145e-06, + "loss": 0.8565, + "step": 7291 + }, + { + "epoch": 0.6576182531451504, + "grad_norm": 1.8389993661627368, + "learning_rate": 1.108584814139066e-06, + "loss": 0.984, + "step": 7292 + }, + { + "epoch": 0.6577084366686207, + "grad_norm": 1.6165875045166969, + "learning_rate": 1.108061879319276e-06, + "loss": 0.9789, + "step": 7293 + }, + { + "epoch": 0.6577986201920909, + "grad_norm": 1.6418762521844277, + "learning_rate": 1.1075390206045648e-06, + "loss": 1.0376, + "step": 7294 + }, + { + "epoch": 0.6578888037155611, + "grad_norm": 2.1861069739675685, + "learning_rate": 1.1070162380395454e-06, + "loss": 1.026, + "step": 7295 + }, + { + "epoch": 0.6579789872390315, + "grad_norm": 1.6360854059231913, + "learning_rate": 1.1064935316688253e-06, + "loss": 0.9265, + "step": 7296 + }, + { + "epoch": 0.6580691707625017, + "grad_norm": 1.9187941629999734, + "learning_rate": 1.105970901537002e-06, + "loss": 1.0664, + "step": 7297 + }, + { + "epoch": 0.6581593542859719, + "grad_norm": 1.545164684789925, + "learning_rate": 1.1054483476886727e-06, + "loss": 1.0233, + "step": 7298 + }, + { + "epoch": 0.6582495378094422, + "grad_norm": 1.5941561761016705, + "learning_rate": 1.1049258701684222e-06, + "loss": 0.9051, + "step": 7299 + }, + { + "epoch": 0.6583397213329125, + "grad_norm": 1.567083730713641, + "learning_rate": 1.1044034690208315e-06, + "loss": 0.9294, + "step": 7300 + }, + { + "epoch": 0.6584299048563828, + "grad_norm": 1.9747767165275065, + "learning_rate": 1.1038811442904755e-06, + "loss": 1.012, + "step": 7301 + }, + { + "epoch": 0.658520088379853, + "grad_norm": 1.7634461048219807, + "learning_rate": 1.103358896021921e-06, + "loss": 0.9135, + "step": 7302 + }, + { + "epoch": 0.6586102719033232, + "grad_norm": 2.8693832596095183, + "learning_rate": 1.1028367242597298e-06, + "loss": 0.9604, + "step": 7303 + }, + { + "epoch": 0.6587004554267936, + "grad_norm": 1.7200582207207602, + "learning_rate": 1.102314629048457e-06, + "loss": 1.0229, + "step": 7304 + }, + { + "epoch": 0.6587906389502638, + "grad_norm": 15.707714173166586, + "learning_rate": 1.1017926104326484e-06, + "loss": 1.031, + "step": 7305 + }, + { + "epoch": 0.658880822473734, + "grad_norm": 1.8307749135219553, + "learning_rate": 1.1012706684568483e-06, + "loss": 1.0343, + "step": 7306 + }, + { + "epoch": 0.6589710059972043, + "grad_norm": 1.4603337963207712, + "learning_rate": 1.1007488031655894e-06, + "loss": 0.9959, + "step": 7307 + }, + { + "epoch": 0.6590611895206746, + "grad_norm": 2.096729492960499, + "learning_rate": 1.1002270146034013e-06, + "loss": 0.9382, + "step": 7308 + }, + { + "epoch": 0.6591513730441448, + "grad_norm": 1.2564144064898597, + "learning_rate": 1.0997053028148052e-06, + "loss": 1.0115, + "step": 7309 + }, + { + "epoch": 0.6592415565676151, + "grad_norm": 1.4744765468166752, + "learning_rate": 1.0991836678443173e-06, + "loss": 0.9888, + "step": 7310 + }, + { + "epoch": 0.6593317400910853, + "grad_norm": 1.6670021990948631, + "learning_rate": 1.0986621097364465e-06, + "loss": 0.9261, + "step": 7311 + }, + { + "epoch": 0.6594219236145556, + "grad_norm": 2.696267575209791, + "learning_rate": 1.0981406285356932e-06, + "loss": 0.8956, + "step": 7312 + }, + { + "epoch": 0.6595121071380259, + "grad_norm": 1.7255859375, + "learning_rate": 1.0976192242865554e-06, + "loss": 0.9046, + "step": 7313 + }, + { + "epoch": 0.6596022906614961, + "grad_norm": 1.7699675369518513, + "learning_rate": 1.0970978970335202e-06, + "loss": 1.0462, + "step": 7314 + }, + { + "epoch": 0.6596924741849665, + "grad_norm": 1.5869026441896903, + "learning_rate": 1.0965766468210714e-06, + "loss": 1.0031, + "step": 7315 + }, + { + "epoch": 0.6597826577084367, + "grad_norm": 1.5628582353007776, + "learning_rate": 1.0960554736936843e-06, + "loss": 0.9919, + "step": 7316 + }, + { + "epoch": 0.6598728412319069, + "grad_norm": 2.9249522278625886, + "learning_rate": 1.0955343776958283e-06, + "loss": 0.8953, + "step": 7317 + }, + { + "epoch": 0.6599630247553772, + "grad_norm": 1.7291958434925692, + "learning_rate": 1.0950133588719665e-06, + "loss": 1.0002, + "step": 7318 + }, + { + "epoch": 0.6600532082788475, + "grad_norm": 1.6488300597308967, + "learning_rate": 1.0944924172665551e-06, + "loss": 1.002, + "step": 7319 + }, + { + "epoch": 0.6601433918023177, + "grad_norm": 1.6846269238676048, + "learning_rate": 1.0939715529240437e-06, + "loss": 1.0167, + "step": 7320 + }, + { + "epoch": 0.660233575325788, + "grad_norm": 1.6361009984260695, + "learning_rate": 1.0934507658888755e-06, + "loss": 1.0169, + "step": 7321 + }, + { + "epoch": 0.6603237588492582, + "grad_norm": 1.6533947210810642, + "learning_rate": 1.092930056205486e-06, + "loss": 0.9561, + "step": 7322 + }, + { + "epoch": 0.6604139423727285, + "grad_norm": 2.410289000281993, + "learning_rate": 1.092409423918306e-06, + "loss": 0.9496, + "step": 7323 + }, + { + "epoch": 0.6605041258961988, + "grad_norm": 1.768897206142069, + "learning_rate": 1.0918888690717581e-06, + "loss": 0.9055, + "step": 7324 + }, + { + "epoch": 0.660594309419669, + "grad_norm": 1.72713567963508, + "learning_rate": 1.091368391710259e-06, + "loss": 1.0465, + "step": 7325 + }, + { + "epoch": 0.6606844929431392, + "grad_norm": 2.376061653968489, + "learning_rate": 1.0908479918782198e-06, + "loss": 0.93, + "step": 7326 + }, + { + "epoch": 0.6607746764666096, + "grad_norm": 1.4784821641328019, + "learning_rate": 1.0903276696200413e-06, + "loss": 1.0055, + "step": 7327 + }, + { + "epoch": 0.6608648599900798, + "grad_norm": 1.8611450105239957, + "learning_rate": 1.0898074249801234e-06, + "loss": 0.9851, + "step": 7328 + }, + { + "epoch": 0.6609550435135501, + "grad_norm": 1.425578565430829, + "learning_rate": 1.0892872580028533e-06, + "loss": 0.9994, + "step": 7329 + }, + { + "epoch": 0.6610452270370203, + "grad_norm": 1.8467916195503107, + "learning_rate": 1.0887671687326178e-06, + "loss": 1.0141, + "step": 7330 + }, + { + "epoch": 0.6611354105604906, + "grad_norm": 1.616682554278354, + "learning_rate": 1.0882471572137908e-06, + "loss": 0.9491, + "step": 7331 + }, + { + "epoch": 0.6612255940839609, + "grad_norm": 1.6259682045111765, + "learning_rate": 1.087727223490744e-06, + "loss": 1.0292, + "step": 7332 + }, + { + "epoch": 0.6613157776074311, + "grad_norm": 1.735017399769363, + "learning_rate": 1.0872073676078405e-06, + "loss": 1.0436, + "step": 7333 + }, + { + "epoch": 0.6614059611309013, + "grad_norm": 2.530708066594971, + "learning_rate": 1.0866875896094375e-06, + "loss": 0.9339, + "step": 7334 + }, + { + "epoch": 0.6614961446543717, + "grad_norm": 1.6021182700896275, + "learning_rate": 1.0861678895398854e-06, + "loss": 1.0004, + "step": 7335 + }, + { + "epoch": 0.6615863281778419, + "grad_norm": 1.9387210566846187, + "learning_rate": 1.0856482674435286e-06, + "loss": 0.9483, + "step": 7336 + }, + { + "epoch": 0.6616765117013121, + "grad_norm": 0.6746097840803239, + "learning_rate": 1.0851287233647024e-06, + "loss": 0.8125, + "step": 7337 + }, + { + "epoch": 0.6617666952247825, + "grad_norm": 2.10476044186486, + "learning_rate": 1.084609257347738e-06, + "loss": 1.0387, + "step": 7338 + }, + { + "epoch": 0.6618568787482527, + "grad_norm": 1.9835685232570865, + "learning_rate": 1.0840898694369594e-06, + "loss": 0.9623, + "step": 7339 + }, + { + "epoch": 0.661947062271723, + "grad_norm": 1.4445991198954673, + "learning_rate": 1.083570559676683e-06, + "loss": 0.9702, + "step": 7340 + }, + { + "epoch": 0.6620372457951932, + "grad_norm": 1.536159895884301, + "learning_rate": 1.08305132811122e-06, + "loss": 0.9599, + "step": 7341 + }, + { + "epoch": 0.6621274293186635, + "grad_norm": 1.4286496924351568, + "learning_rate": 1.0825321747848735e-06, + "loss": 0.9841, + "step": 7342 + }, + { + "epoch": 0.6622176128421338, + "grad_norm": 0.711390089240653, + "learning_rate": 1.0820130997419417e-06, + "loss": 0.8822, + "step": 7343 + }, + { + "epoch": 0.662307796365604, + "grad_norm": 1.8962504658104382, + "learning_rate": 1.0814941030267123e-06, + "loss": 0.9744, + "step": 7344 + }, + { + "epoch": 0.6623979798890742, + "grad_norm": 0.5846226244052974, + "learning_rate": 1.080975184683472e-06, + "loss": 0.8, + "step": 7345 + }, + { + "epoch": 0.6624881634125446, + "grad_norm": 1.5409785702581171, + "learning_rate": 1.0804563447564948e-06, + "loss": 1.0433, + "step": 7346 + }, + { + "epoch": 0.6625783469360148, + "grad_norm": 1.6959382058348276, + "learning_rate": 1.0799375832900545e-06, + "loss": 0.9275, + "step": 7347 + }, + { + "epoch": 0.662668530459485, + "grad_norm": 1.9764632249108631, + "learning_rate": 1.0794189003284118e-06, + "loss": 1.0188, + "step": 7348 + }, + { + "epoch": 0.6627587139829553, + "grad_norm": 1.743581650240125, + "learning_rate": 1.0789002959158242e-06, + "loss": 1.0004, + "step": 7349 + }, + { + "epoch": 0.6628488975064256, + "grad_norm": 2.1283433082404573, + "learning_rate": 1.0783817700965428e-06, + "loss": 0.9343, + "step": 7350 + }, + { + "epoch": 0.6629390810298958, + "grad_norm": 1.4216219498703837, + "learning_rate": 1.0778633229148102e-06, + "loss": 1.032, + "step": 7351 + }, + { + "epoch": 0.6630292645533661, + "grad_norm": 1.6871024475874, + "learning_rate": 1.0773449544148645e-06, + "loss": 0.9853, + "step": 7352 + }, + { + "epoch": 0.6631194480768363, + "grad_norm": 1.4671064880560452, + "learning_rate": 1.076826664640934e-06, + "loss": 1.0238, + "step": 7353 + }, + { + "epoch": 0.6632096316003067, + "grad_norm": 0.7513959529597142, + "learning_rate": 1.0763084536372424e-06, + "loss": 0.841, + "step": 7354 + }, + { + "epoch": 0.6632998151237769, + "grad_norm": 3.411877664123714, + "learning_rate": 1.0757903214480068e-06, + "loss": 0.9926, + "step": 7355 + }, + { + "epoch": 0.6633899986472471, + "grad_norm": 1.546550061554914, + "learning_rate": 1.0752722681174376e-06, + "loss": 0.9255, + "step": 7356 + }, + { + "epoch": 0.6634801821707174, + "grad_norm": 1.750406354319739, + "learning_rate": 1.074754293689737e-06, + "loss": 0.9046, + "step": 7357 + }, + { + "epoch": 0.6635703656941877, + "grad_norm": 1.7217906758175912, + "learning_rate": 1.0742363982091023e-06, + "loss": 0.9862, + "step": 7358 + }, + { + "epoch": 0.6636605492176579, + "grad_norm": 2.0455529988802605, + "learning_rate": 1.0737185817197215e-06, + "loss": 0.942, + "step": 7359 + }, + { + "epoch": 0.6637507327411282, + "grad_norm": 2.084015378569461, + "learning_rate": 1.0732008442657803e-06, + "loss": 0.9461, + "step": 7360 + }, + { + "epoch": 0.6638409162645985, + "grad_norm": 1.5922838086470839, + "learning_rate": 1.0726831858914516e-06, + "loss": 0.9672, + "step": 7361 + }, + { + "epoch": 0.6639310997880687, + "grad_norm": 1.7815023544642854, + "learning_rate": 1.0721656066409084e-06, + "loss": 1.0045, + "step": 7362 + }, + { + "epoch": 0.664021283311539, + "grad_norm": 1.9844091217231212, + "learning_rate": 1.0716481065583108e-06, + "loss": 0.9957, + "step": 7363 + }, + { + "epoch": 0.6641114668350092, + "grad_norm": 0.6360551148921376, + "learning_rate": 1.071130685687816e-06, + "loss": 0.8114, + "step": 7364 + }, + { + "epoch": 0.6642016503584796, + "grad_norm": 1.4842908333457097, + "learning_rate": 1.0706133440735723e-06, + "loss": 0.9228, + "step": 7365 + }, + { + "epoch": 0.6642918338819498, + "grad_norm": 5.858212124188761, + "learning_rate": 1.070096081759723e-06, + "loss": 0.9319, + "step": 7366 + }, + { + "epoch": 0.66438201740542, + "grad_norm": 1.9483753283140572, + "learning_rate": 1.069578898790404e-06, + "loss": 1.0844, + "step": 7367 + }, + { + "epoch": 0.6644722009288903, + "grad_norm": 6.411819344548699, + "learning_rate": 1.0690617952097424e-06, + "loss": 0.9668, + "step": 7368 + }, + { + "epoch": 0.6645623844523606, + "grad_norm": 1.703372622345992, + "learning_rate": 1.068544771061863e-06, + "loss": 1.0302, + "step": 7369 + }, + { + "epoch": 0.6646525679758308, + "grad_norm": 1.9658511785214974, + "learning_rate": 1.0680278263908787e-06, + "loss": 1.0221, + "step": 7370 + }, + { + "epoch": 0.6647427514993011, + "grad_norm": 0.762483848338175, + "learning_rate": 1.0675109612408991e-06, + "loss": 0.822, + "step": 7371 + }, + { + "epoch": 0.6648329350227713, + "grad_norm": 1.8911610623643615, + "learning_rate": 1.0669941756560264e-06, + "loss": 0.9509, + "step": 7372 + }, + { + "epoch": 0.6649231185462416, + "grad_norm": 1.9924825412459202, + "learning_rate": 1.0664774696803548e-06, + "loss": 0.928, + "step": 7373 + }, + { + "epoch": 0.6650133020697119, + "grad_norm": 1.614738563540803, + "learning_rate": 1.065960843357973e-06, + "loss": 0.9555, + "step": 7374 + }, + { + "epoch": 0.6651034855931821, + "grad_norm": 1.5982843857086566, + "learning_rate": 1.065444296732963e-06, + "loss": 0.9998, + "step": 7375 + }, + { + "epoch": 0.6651936691166523, + "grad_norm": 0.6723516680698691, + "learning_rate": 1.064927829849397e-06, + "loss": 0.824, + "step": 7376 + }, + { + "epoch": 0.6652838526401227, + "grad_norm": 1.8070573975567503, + "learning_rate": 1.0644114427513465e-06, + "loss": 0.9776, + "step": 7377 + }, + { + "epoch": 0.6653740361635929, + "grad_norm": 1.4512580445532397, + "learning_rate": 1.0638951354828693e-06, + "loss": 0.9105, + "step": 7378 + }, + { + "epoch": 0.6654642196870632, + "grad_norm": 1.8711478398733543, + "learning_rate": 1.063378908088021e-06, + "loss": 1.0358, + "step": 7379 + }, + { + "epoch": 0.6655544032105334, + "grad_norm": 1.8631820812266189, + "learning_rate": 1.0628627606108486e-06, + "loss": 0.976, + "step": 7380 + }, + { + "epoch": 0.6656445867340037, + "grad_norm": 2.1669433368198145, + "learning_rate": 1.062346693095393e-06, + "loss": 1.0495, + "step": 7381 + }, + { + "epoch": 0.665734770257474, + "grad_norm": 1.5925304009905819, + "learning_rate": 1.0618307055856882e-06, + "loss": 0.9198, + "step": 7382 + }, + { + "epoch": 0.6658249537809442, + "grad_norm": 1.659200523384843, + "learning_rate": 1.061314798125759e-06, + "loss": 1.0344, + "step": 7383 + }, + { + "epoch": 0.6659151373044144, + "grad_norm": 1.4714154346054238, + "learning_rate": 1.0607989707596293e-06, + "loss": 0.9408, + "step": 7384 + }, + { + "epoch": 0.6660053208278848, + "grad_norm": 1.3665630985575354, + "learning_rate": 1.0602832235313078e-06, + "loss": 1.1085, + "step": 7385 + }, + { + "epoch": 0.666095504351355, + "grad_norm": 1.7766344604412079, + "learning_rate": 1.0597675564848053e-06, + "loss": 0.9719, + "step": 7386 + }, + { + "epoch": 0.6661856878748252, + "grad_norm": 1.7299881175079643, + "learning_rate": 1.059251969664118e-06, + "loss": 1.0477, + "step": 7387 + }, + { + "epoch": 0.6662758713982956, + "grad_norm": 2.777619810910925, + "learning_rate": 1.0587364631132402e-06, + "loss": 0.9898, + "step": 7388 + }, + { + "epoch": 0.6663660549217658, + "grad_norm": 1.6386064009379508, + "learning_rate": 1.0582210368761573e-06, + "loss": 1.0186, + "step": 7389 + }, + { + "epoch": 0.666456238445236, + "grad_norm": 2.2765026076796433, + "learning_rate": 1.0577056909968485e-06, + "loss": 0.9825, + "step": 7390 + }, + { + "epoch": 0.6665464219687063, + "grad_norm": 1.6516674632748332, + "learning_rate": 1.0571904255192857e-06, + "loss": 1.0001, + "step": 7391 + }, + { + "epoch": 0.6666366054921766, + "grad_norm": 1.9519131372190754, + "learning_rate": 1.0566752404874354e-06, + "loss": 1.0149, + "step": 7392 + }, + { + "epoch": 0.6667267890156469, + "grad_norm": 1.5702912438315708, + "learning_rate": 1.0561601359452543e-06, + "loss": 0.9768, + "step": 7393 + }, + { + "epoch": 0.6668169725391171, + "grad_norm": 1.6288402370185704, + "learning_rate": 1.0556451119366947e-06, + "loss": 0.9192, + "step": 7394 + }, + { + "epoch": 0.6669071560625873, + "grad_norm": 1.413246382941667, + "learning_rate": 1.0551301685057011e-06, + "loss": 0.9171, + "step": 7395 + }, + { + "epoch": 0.6669973395860577, + "grad_norm": 1.8706417294310838, + "learning_rate": 1.0546153056962117e-06, + "loss": 1.0647, + "step": 7396 + }, + { + "epoch": 0.6670875231095279, + "grad_norm": 2.769639716559881, + "learning_rate": 1.0541005235521578e-06, + "loss": 0.9777, + "step": 7397 + }, + { + "epoch": 0.6671777066329981, + "grad_norm": 1.9499075672054667, + "learning_rate": 1.0535858221174614e-06, + "loss": 0.9969, + "step": 7398 + }, + { + "epoch": 0.6672678901564684, + "grad_norm": 1.2657426379127386, + "learning_rate": 1.0530712014360426e-06, + "loss": 1.0211, + "step": 7399 + }, + { + "epoch": 0.6673580736799387, + "grad_norm": 1.7550478068521518, + "learning_rate": 1.0525566615518088e-06, + "loss": 1.035, + "step": 7400 + }, + { + "epoch": 0.667448257203409, + "grad_norm": 1.4396507924564836, + "learning_rate": 1.0520422025086662e-06, + "loss": 0.9958, + "step": 7401 + }, + { + "epoch": 0.6675384407268792, + "grad_norm": 1.6489900503604453, + "learning_rate": 1.0515278243505092e-06, + "loss": 1.0327, + "step": 7402 + }, + { + "epoch": 0.6676286242503494, + "grad_norm": 1.3740794828544003, + "learning_rate": 1.0510135271212278e-06, + "loss": 1.0099, + "step": 7403 + }, + { + "epoch": 0.6677188077738198, + "grad_norm": 1.4242809187921377, + "learning_rate": 1.0504993108647052e-06, + "loss": 1.0484, + "step": 7404 + }, + { + "epoch": 0.66780899129729, + "grad_norm": 4.147261228304865, + "learning_rate": 1.0499851756248168e-06, + "loss": 0.956, + "step": 7405 + }, + { + "epoch": 0.6678991748207602, + "grad_norm": 1.7143694454135117, + "learning_rate": 1.0494711214454316e-06, + "loss": 0.9665, + "step": 7406 + }, + { + "epoch": 0.6679893583442305, + "grad_norm": 3.3261592593712606, + "learning_rate": 1.0489571483704111e-06, + "loss": 0.8985, + "step": 7407 + }, + { + "epoch": 0.6680795418677008, + "grad_norm": 1.6491603622328872, + "learning_rate": 1.048443256443612e-06, + "loss": 0.8974, + "step": 7408 + }, + { + "epoch": 0.668169725391171, + "grad_norm": 1.9346787000137742, + "learning_rate": 1.0479294457088801e-06, + "loss": 0.9969, + "step": 7409 + }, + { + "epoch": 0.6682599089146413, + "grad_norm": 1.4253991019869026, + "learning_rate": 1.0474157162100574e-06, + "loss": 1.1045, + "step": 7410 + }, + { + "epoch": 0.6683500924381116, + "grad_norm": 1.6821601587944013, + "learning_rate": 1.0469020679909786e-06, + "loss": 0.9226, + "step": 7411 + }, + { + "epoch": 0.6684402759615818, + "grad_norm": 1.610662667393193, + "learning_rate": 1.0463885010954705e-06, + "loss": 0.9088, + "step": 7412 + }, + { + "epoch": 0.6685304594850521, + "grad_norm": 1.8228544896738994, + "learning_rate": 1.0458750155673536e-06, + "loss": 1.0521, + "step": 7413 + }, + { + "epoch": 0.6686206430085223, + "grad_norm": 1.603444477691367, + "learning_rate": 1.0453616114504421e-06, + "loss": 0.9383, + "step": 7414 + }, + { + "epoch": 0.6687108265319927, + "grad_norm": 1.5734605501488608, + "learning_rate": 1.0448482887885406e-06, + "loss": 1.016, + "step": 7415 + }, + { + "epoch": 0.6688010100554629, + "grad_norm": 1.6612482047663304, + "learning_rate": 1.044335047625451e-06, + "loss": 0.9904, + "step": 7416 + }, + { + "epoch": 0.6688911935789331, + "grad_norm": 1.7366939541191615, + "learning_rate": 1.0438218880049637e-06, + "loss": 0.9947, + "step": 7417 + }, + { + "epoch": 0.6689813771024034, + "grad_norm": 1.783088939474602, + "learning_rate": 1.0433088099708653e-06, + "loss": 1.0416, + "step": 7418 + }, + { + "epoch": 0.6690715606258737, + "grad_norm": 0.6129006097489615, + "learning_rate": 1.0427958135669346e-06, + "loss": 0.7497, + "step": 7419 + }, + { + "epoch": 0.6691617441493439, + "grad_norm": 2.1354379668375247, + "learning_rate": 1.0422828988369428e-06, + "loss": 0.9184, + "step": 7420 + }, + { + "epoch": 0.6692519276728142, + "grad_norm": 1.7255218269428954, + "learning_rate": 1.041770065824655e-06, + "loss": 1.0494, + "step": 7421 + }, + { + "epoch": 0.6693421111962844, + "grad_norm": 1.703551912610097, + "learning_rate": 1.0412573145738287e-06, + "loss": 0.8381, + "step": 7422 + }, + { + "epoch": 0.6694322947197547, + "grad_norm": 2.3270972114588573, + "learning_rate": 1.040744645128216e-06, + "loss": 0.9553, + "step": 7423 + }, + { + "epoch": 0.669522478243225, + "grad_norm": 1.6264884440835055, + "learning_rate": 1.040232057531558e-06, + "loss": 1.04, + "step": 7424 + }, + { + "epoch": 0.6696126617666952, + "grad_norm": 3.191430155969215, + "learning_rate": 1.0397195518275932e-06, + "loss": 0.9542, + "step": 7425 + }, + { + "epoch": 0.6697028452901654, + "grad_norm": 1.4886117002626762, + "learning_rate": 1.0392071280600512e-06, + "loss": 0.9548, + "step": 7426 + }, + { + "epoch": 0.6697930288136358, + "grad_norm": 3.456706817403578, + "learning_rate": 1.0386947862726549e-06, + "loss": 0.9842, + "step": 7427 + }, + { + "epoch": 0.669883212337106, + "grad_norm": 1.9596911721308445, + "learning_rate": 1.0381825265091197e-06, + "loss": 0.9034, + "step": 7428 + }, + { + "epoch": 0.6699733958605762, + "grad_norm": 1.8239054441454332, + "learning_rate": 1.037670348813155e-06, + "loss": 0.9021, + "step": 7429 + }, + { + "epoch": 0.6700635793840465, + "grad_norm": 1.6365988603259998, + "learning_rate": 1.0371582532284624e-06, + "loss": 1.0129, + "step": 7430 + }, + { + "epoch": 0.6701537629075168, + "grad_norm": 1.5238828350494982, + "learning_rate": 1.0366462397987375e-06, + "loss": 0.9182, + "step": 7431 + }, + { + "epoch": 0.6702439464309871, + "grad_norm": 2.5290305685999215, + "learning_rate": 1.0361343085676665e-06, + "loss": 0.9907, + "step": 7432 + }, + { + "epoch": 0.6703341299544573, + "grad_norm": 2.0894348314889033, + "learning_rate": 1.0356224595789309e-06, + "loss": 1.0993, + "step": 7433 + }, + { + "epoch": 0.6704243134779276, + "grad_norm": 4.017975234460911, + "learning_rate": 1.0351106928762046e-06, + "loss": 1.0391, + "step": 7434 + }, + { + "epoch": 0.6705144970013979, + "grad_norm": 1.8341937935796222, + "learning_rate": 1.034599008503154e-06, + "loss": 1.0576, + "step": 7435 + }, + { + "epoch": 0.6706046805248681, + "grad_norm": 1.6715416977109123, + "learning_rate": 1.0340874065034406e-06, + "loss": 0.972, + "step": 7436 + }, + { + "epoch": 0.6706948640483383, + "grad_norm": 1.6357142021859779, + "learning_rate": 1.0335758869207137e-06, + "loss": 0.9097, + "step": 7437 + }, + { + "epoch": 0.6707850475718087, + "grad_norm": 2.2882097237642354, + "learning_rate": 1.0330644497986227e-06, + "loss": 1.0011, + "step": 7438 + }, + { + "epoch": 0.6708752310952789, + "grad_norm": 2.3330392879312516, + "learning_rate": 1.0325530951808029e-06, + "loss": 1.0235, + "step": 7439 + }, + { + "epoch": 0.6709654146187491, + "grad_norm": 1.696117016858061, + "learning_rate": 1.0320418231108887e-06, + "loss": 0.9915, + "step": 7440 + }, + { + "epoch": 0.6710555981422194, + "grad_norm": 1.5819540876244047, + "learning_rate": 1.0315306336325028e-06, + "loss": 1.0267, + "step": 7441 + }, + { + "epoch": 0.6711457816656897, + "grad_norm": 1.815925255535427, + "learning_rate": 1.0310195267892635e-06, + "loss": 0.9425, + "step": 7442 + }, + { + "epoch": 0.67123596518916, + "grad_norm": 1.5862591046251635, + "learning_rate": 1.030508502624781e-06, + "loss": 0.9227, + "step": 7443 + }, + { + "epoch": 0.6713261487126302, + "grad_norm": 2.4923779645026216, + "learning_rate": 1.0299975611826587e-06, + "loss": 0.9321, + "step": 7444 + }, + { + "epoch": 0.6714163322361004, + "grad_norm": 2.686339637961324, + "learning_rate": 1.0294867025064928e-06, + "loss": 0.9917, + "step": 7445 + }, + { + "epoch": 0.6715065157595708, + "grad_norm": 1.4039028925529162, + "learning_rate": 1.028975926639874e-06, + "loss": 0.991, + "step": 7446 + }, + { + "epoch": 0.671596699283041, + "grad_norm": 5.798940614436042, + "learning_rate": 1.0284652336263823e-06, + "loss": 0.9126, + "step": 7447 + }, + { + "epoch": 0.6716868828065112, + "grad_norm": 1.7408423597954885, + "learning_rate": 1.0279546235095938e-06, + "loss": 0.9134, + "step": 7448 + }, + { + "epoch": 0.6717770663299815, + "grad_norm": 0.7369184075820664, + "learning_rate": 1.0274440963330768e-06, + "loss": 0.8463, + "step": 7449 + }, + { + "epoch": 0.6718672498534518, + "grad_norm": 1.728638380616357, + "learning_rate": 1.0269336521403919e-06, + "loss": 0.984, + "step": 7450 + }, + { + "epoch": 0.671957433376922, + "grad_norm": 1.7122311513328108, + "learning_rate": 1.0264232909750936e-06, + "loss": 1.112, + "step": 7451 + }, + { + "epoch": 0.6720476169003923, + "grad_norm": 1.4310759592642461, + "learning_rate": 1.025913012880728e-06, + "loss": 1.0448, + "step": 7452 + }, + { + "epoch": 0.6721378004238625, + "grad_norm": 1.4162075570355117, + "learning_rate": 1.0254028179008362e-06, + "loss": 0.8921, + "step": 7453 + }, + { + "epoch": 0.6722279839473329, + "grad_norm": 1.4324272877611974, + "learning_rate": 1.0248927060789483e-06, + "loss": 0.8709, + "step": 7454 + }, + { + "epoch": 0.6723181674708031, + "grad_norm": 1.553034684991963, + "learning_rate": 1.0243826774585928e-06, + "loss": 0.9421, + "step": 7455 + }, + { + "epoch": 0.6724083509942733, + "grad_norm": 1.8848009175371225, + "learning_rate": 1.0238727320832854e-06, + "loss": 0.972, + "step": 7456 + }, + { + "epoch": 0.6724985345177437, + "grad_norm": 1.7618107739962146, + "learning_rate": 1.0233628699965403e-06, + "loss": 0.9853, + "step": 7457 + }, + { + "epoch": 0.6725887180412139, + "grad_norm": 1.930768717822148, + "learning_rate": 1.0228530912418594e-06, + "loss": 1.008, + "step": 7458 + }, + { + "epoch": 0.6726789015646841, + "grad_norm": 1.7641090028781228, + "learning_rate": 1.0223433958627404e-06, + "loss": 0.9536, + "step": 7459 + }, + { + "epoch": 0.6727690850881544, + "grad_norm": 2.623588818473968, + "learning_rate": 1.021833783902674e-06, + "loss": 1.0074, + "step": 7460 + }, + { + "epoch": 0.6728592686116247, + "grad_norm": 2.242548471562733, + "learning_rate": 1.0213242554051427e-06, + "loss": 0.987, + "step": 7461 + }, + { + "epoch": 0.6729494521350949, + "grad_norm": 1.5587581079389938, + "learning_rate": 1.0208148104136229e-06, + "loss": 0.9214, + "step": 7462 + }, + { + "epoch": 0.6730396356585652, + "grad_norm": 1.6390282262777034, + "learning_rate": 1.020305448971582e-06, + "loss": 0.9562, + "step": 7463 + }, + { + "epoch": 0.6731298191820354, + "grad_norm": 1.494989849426104, + "learning_rate": 1.0197961711224824e-06, + "loss": 0.9499, + "step": 7464 + }, + { + "epoch": 0.6732200027055057, + "grad_norm": 1.658765214564237, + "learning_rate": 1.0192869769097777e-06, + "loss": 0.9902, + "step": 7465 + }, + { + "epoch": 0.673310186228976, + "grad_norm": 1.632516350423301, + "learning_rate": 1.018777866376916e-06, + "loss": 0.9899, + "step": 7466 + }, + { + "epoch": 0.6734003697524462, + "grad_norm": 1.824431184411246, + "learning_rate": 1.0182688395673374e-06, + "loss": 0.8505, + "step": 7467 + }, + { + "epoch": 0.6734905532759164, + "grad_norm": 1.6173455350016717, + "learning_rate": 1.017759896524475e-06, + "loss": 0.9479, + "step": 7468 + }, + { + "epoch": 0.6735807367993868, + "grad_norm": 2.6737966486859412, + "learning_rate": 1.0172510372917528e-06, + "loss": 1.0239, + "step": 7469 + }, + { + "epoch": 0.673670920322857, + "grad_norm": 13.854327223380427, + "learning_rate": 1.0167422619125925e-06, + "loss": 0.8865, + "step": 7470 + }, + { + "epoch": 0.6737611038463273, + "grad_norm": 1.6722597053920278, + "learning_rate": 1.0162335704304026e-06, + "loss": 0.9863, + "step": 7471 + }, + { + "epoch": 0.6738512873697975, + "grad_norm": 0.6928176267827382, + "learning_rate": 1.0157249628885903e-06, + "loss": 0.83, + "step": 7472 + }, + { + "epoch": 0.6739414708932678, + "grad_norm": 1.229129801482916, + "learning_rate": 1.0152164393305506e-06, + "loss": 0.8933, + "step": 7473 + }, + { + "epoch": 0.6740316544167381, + "grad_norm": 1.7808111888393474, + "learning_rate": 1.0147079997996746e-06, + "loss": 0.9779, + "step": 7474 + }, + { + "epoch": 0.6741218379402083, + "grad_norm": 1.6715531083872075, + "learning_rate": 1.0141996443393446e-06, + "loss": 1.0146, + "step": 7475 + }, + { + "epoch": 0.6742120214636785, + "grad_norm": 1.6707055823874624, + "learning_rate": 1.0136913729929369e-06, + "loss": 0.9701, + "step": 7476 + }, + { + "epoch": 0.6743022049871489, + "grad_norm": 1.5507954709724607, + "learning_rate": 1.0131831858038203e-06, + "loss": 0.942, + "step": 7477 + }, + { + "epoch": 0.6743923885106191, + "grad_norm": 1.4103549806674667, + "learning_rate": 1.0126750828153538e-06, + "loss": 1.0477, + "step": 7478 + }, + { + "epoch": 0.6744825720340893, + "grad_norm": 2.0861629978550056, + "learning_rate": 1.012167064070895e-06, + "loss": 1.0264, + "step": 7479 + }, + { + "epoch": 0.6745727555575597, + "grad_norm": 1.6468003808042184, + "learning_rate": 1.0116591296137885e-06, + "loss": 1.0098, + "step": 7480 + }, + { + "epoch": 0.6746629390810299, + "grad_norm": 1.438073251283087, + "learning_rate": 1.0111512794873746e-06, + "loss": 1.0102, + "step": 7481 + }, + { + "epoch": 0.6747531226045002, + "grad_norm": 1.414551961079936, + "learning_rate": 1.010643513734986e-06, + "loss": 0.9827, + "step": 7482 + }, + { + "epoch": 0.6748433061279704, + "grad_norm": 1.7486443718509297, + "learning_rate": 1.010135832399948e-06, + "loss": 1.0224, + "step": 7483 + }, + { + "epoch": 0.6749334896514407, + "grad_norm": 1.440211682390472, + "learning_rate": 1.0096282355255792e-06, + "loss": 0.9915, + "step": 7484 + }, + { + "epoch": 0.675023673174911, + "grad_norm": 3.8481032368458283, + "learning_rate": 1.0091207231551905e-06, + "loss": 1.0371, + "step": 7485 + }, + { + "epoch": 0.6751138566983812, + "grad_norm": 1.4269298368820225, + "learning_rate": 1.0086132953320842e-06, + "loss": 0.9479, + "step": 7486 + }, + { + "epoch": 0.6752040402218514, + "grad_norm": 1.6595278998153369, + "learning_rate": 1.0081059520995591e-06, + "loss": 1.0096, + "step": 7487 + }, + { + "epoch": 0.6752942237453218, + "grad_norm": 1.72156550677375, + "learning_rate": 1.0075986935009028e-06, + "loss": 1.0491, + "step": 7488 + }, + { + "epoch": 0.675384407268792, + "grad_norm": 1.6305171036324755, + "learning_rate": 1.0070915195793982e-06, + "loss": 0.9945, + "step": 7489 + }, + { + "epoch": 0.6754745907922622, + "grad_norm": 1.8060505661206288, + "learning_rate": 1.0065844303783197e-06, + "loss": 0.9618, + "step": 7490 + }, + { + "epoch": 0.6755647743157325, + "grad_norm": 1.5301373390190751, + "learning_rate": 1.0060774259409356e-06, + "loss": 0.9001, + "step": 7491 + }, + { + "epoch": 0.6756549578392028, + "grad_norm": 1.335803376542319, + "learning_rate": 1.0055705063105065e-06, + "loss": 1.0164, + "step": 7492 + }, + { + "epoch": 0.675745141362673, + "grad_norm": 1.6143259725240735, + "learning_rate": 1.0050636715302837e-06, + "loss": 0.9746, + "step": 7493 + }, + { + "epoch": 0.6758353248861433, + "grad_norm": 1.8234923915887356, + "learning_rate": 1.0045569216435157e-06, + "loss": 0.9918, + "step": 7494 + }, + { + "epoch": 0.6759255084096135, + "grad_norm": 2.4044771475496245, + "learning_rate": 1.0040502566934384e-06, + "loss": 1.0668, + "step": 7495 + }, + { + "epoch": 0.6760156919330839, + "grad_norm": 1.8957526919921692, + "learning_rate": 1.0035436767232866e-06, + "loss": 0.9249, + "step": 7496 + }, + { + "epoch": 0.6761058754565541, + "grad_norm": 1.7688163340500604, + "learning_rate": 1.0030371817762816e-06, + "loss": 0.9942, + "step": 7497 + }, + { + "epoch": 0.6761960589800243, + "grad_norm": 1.4827682633227908, + "learning_rate": 1.0025307718956417e-06, + "loss": 0.9923, + "step": 7498 + }, + { + "epoch": 0.6762862425034946, + "grad_norm": 1.5296158926766643, + "learning_rate": 1.0020244471245765e-06, + "loss": 1.0019, + "step": 7499 + }, + { + "epoch": 0.6763764260269649, + "grad_norm": 1.9630423871485978, + "learning_rate": 1.001518207506288e-06, + "loss": 0.9419, + "step": 7500 + }, + { + "epoch": 0.6764666095504351, + "grad_norm": 1.3925509456855254, + "learning_rate": 1.0010120530839717e-06, + "loss": 1.0482, + "step": 7501 + }, + { + "epoch": 0.6765567930739054, + "grad_norm": 1.2715966897278808, + "learning_rate": 1.0005059839008161e-06, + "loss": 1.0092, + "step": 7502 + }, + { + "epoch": 0.6766469765973756, + "grad_norm": 1.7582358295591582, + "learning_rate": 1.0000000000000004e-06, + "loss": 0.9374, + "step": 7503 + }, + { + "epoch": 0.676737160120846, + "grad_norm": 1.5213561879368456, + "learning_rate": 9.994941014246985e-07, + "loss": 1.0285, + "step": 7504 + }, + { + "epoch": 0.6768273436443162, + "grad_norm": 2.2033274672522007, + "learning_rate": 9.989882882180766e-07, + "loss": 0.875, + "step": 7505 + }, + { + "epoch": 0.6769175271677864, + "grad_norm": 1.6595403269217919, + "learning_rate": 9.984825604232938e-07, + "loss": 1.0021, + "step": 7506 + }, + { + "epoch": 0.6770077106912568, + "grad_norm": 1.7417437753841278, + "learning_rate": 9.97976918083502e-07, + "loss": 0.9242, + "step": 7507 + }, + { + "epoch": 0.677097894214727, + "grad_norm": 1.380073075294704, + "learning_rate": 9.974713612418427e-07, + "loss": 1.0557, + "step": 7508 + }, + { + "epoch": 0.6771880777381972, + "grad_norm": 1.703749865771909, + "learning_rate": 9.969658899414563e-07, + "loss": 0.9976, + "step": 7509 + }, + { + "epoch": 0.6772782612616675, + "grad_norm": 1.6476562239537027, + "learning_rate": 9.964605042254696e-07, + "loss": 0.9023, + "step": 7510 + }, + { + "epoch": 0.6773684447851378, + "grad_norm": 1.7712395445219624, + "learning_rate": 9.959552041370076e-07, + "loss": 0.9267, + "step": 7511 + }, + { + "epoch": 0.677458628308608, + "grad_norm": 1.4592566791412285, + "learning_rate": 9.954499897191824e-07, + "loss": 0.9052, + "step": 7512 + }, + { + "epoch": 0.6775488118320783, + "grad_norm": 4.146893517482323, + "learning_rate": 9.949448610151043e-07, + "loss": 0.9346, + "step": 7513 + }, + { + "epoch": 0.6776389953555485, + "grad_norm": 1.463070296902992, + "learning_rate": 9.944398180678719e-07, + "loss": 0.9625, + "step": 7514 + }, + { + "epoch": 0.6777291788790188, + "grad_norm": 1.590845490707748, + "learning_rate": 9.939348609205789e-07, + "loss": 0.9362, + "step": 7515 + }, + { + "epoch": 0.6778193624024891, + "grad_norm": 1.4067961162209828, + "learning_rate": 9.93429989616311e-07, + "loss": 0.9909, + "step": 7516 + }, + { + "epoch": 0.6779095459259593, + "grad_norm": 1.5545663450947564, + "learning_rate": 9.929252041981464e-07, + "loss": 0.9004, + "step": 7517 + }, + { + "epoch": 0.6779997294494295, + "grad_norm": 1.9869682608355552, + "learning_rate": 9.924205047091572e-07, + "loss": 0.9565, + "step": 7518 + }, + { + "epoch": 0.6780899129728999, + "grad_norm": 1.734426411734487, + "learning_rate": 9.919158911924056e-07, + "loss": 1.1051, + "step": 7519 + }, + { + "epoch": 0.6781800964963701, + "grad_norm": 1.6720434086197755, + "learning_rate": 9.914113636909483e-07, + "loss": 1.0215, + "step": 7520 + }, + { + "epoch": 0.6782702800198404, + "grad_norm": 0.7438439606370304, + "learning_rate": 9.90906922247835e-07, + "loss": 0.8648, + "step": 7521 + }, + { + "epoch": 0.6783604635433106, + "grad_norm": 1.533135868225267, + "learning_rate": 9.904025669061072e-07, + "loss": 1.0148, + "step": 7522 + }, + { + "epoch": 0.6784506470667809, + "grad_norm": 1.6248777783720278, + "learning_rate": 9.89898297708799e-07, + "loss": 0.981, + "step": 7523 + }, + { + "epoch": 0.6785408305902512, + "grad_norm": 1.7314922672796347, + "learning_rate": 9.893941146989388e-07, + "loss": 0.9346, + "step": 7524 + }, + { + "epoch": 0.6786310141137214, + "grad_norm": 1.870628091928748, + "learning_rate": 9.888900179195437e-07, + "loss": 0.9452, + "step": 7525 + }, + { + "epoch": 0.6787211976371916, + "grad_norm": 1.651311024094831, + "learning_rate": 9.883860074136285e-07, + "loss": 1.0513, + "step": 7526 + }, + { + "epoch": 0.678811381160662, + "grad_norm": 1.9090611725922935, + "learning_rate": 9.87882083224196e-07, + "loss": 0.978, + "step": 7527 + }, + { + "epoch": 0.6789015646841322, + "grad_norm": 0.8472734632960497, + "learning_rate": 9.873782453942462e-07, + "loss": 0.842, + "step": 7528 + }, + { + "epoch": 0.6789917482076024, + "grad_norm": 1.8067718623152795, + "learning_rate": 9.868744939667676e-07, + "loss": 0.9098, + "step": 7529 + }, + { + "epoch": 0.6790819317310728, + "grad_norm": 1.850661883645285, + "learning_rate": 9.863708289847432e-07, + "loss": 0.8682, + "step": 7530 + }, + { + "epoch": 0.679172115254543, + "grad_norm": 1.9210132434632716, + "learning_rate": 9.85867250491149e-07, + "loss": 0.8926, + "step": 7531 + }, + { + "epoch": 0.6792622987780133, + "grad_norm": 1.7770064819939009, + "learning_rate": 9.853637585289528e-07, + "loss": 1.0906, + "step": 7532 + }, + { + "epoch": 0.6793524823014835, + "grad_norm": 1.75549482352532, + "learning_rate": 9.848603531411159e-07, + "loss": 0.9765, + "step": 7533 + }, + { + "epoch": 0.6794426658249538, + "grad_norm": 2.6677102987234274, + "learning_rate": 9.843570343705899e-07, + "loss": 0.9897, + "step": 7534 + }, + { + "epoch": 0.6795328493484241, + "grad_norm": 1.2770999594140817, + "learning_rate": 9.83853802260323e-07, + "loss": 1.0558, + "step": 7535 + }, + { + "epoch": 0.6796230328718943, + "grad_norm": 1.7853858182446207, + "learning_rate": 9.833506568532524e-07, + "loss": 0.9267, + "step": 7536 + }, + { + "epoch": 0.6797132163953645, + "grad_norm": 1.2684558715392102, + "learning_rate": 9.828475981923093e-07, + "loss": 0.829, + "step": 7537 + }, + { + "epoch": 0.6798033999188349, + "grad_norm": 2.064574498819439, + "learning_rate": 9.823446263204175e-07, + "loss": 0.9602, + "step": 7538 + }, + { + "epoch": 0.6798935834423051, + "grad_norm": 1.8048962228799226, + "learning_rate": 9.818417412804937e-07, + "loss": 0.9965, + "step": 7539 + }, + { + "epoch": 0.6799837669657753, + "grad_norm": 1.6309516195375682, + "learning_rate": 9.813389431154463e-07, + "loss": 1.0647, + "step": 7540 + }, + { + "epoch": 0.6800739504892456, + "grad_norm": 1.8316143675891052, + "learning_rate": 9.808362318681783e-07, + "loss": 1.0139, + "step": 7541 + }, + { + "epoch": 0.6801641340127159, + "grad_norm": 1.8805357277277392, + "learning_rate": 9.803336075815807e-07, + "loss": 1.0051, + "step": 7542 + }, + { + "epoch": 0.6802543175361861, + "grad_norm": 1.530108123399524, + "learning_rate": 9.79831070298544e-07, + "loss": 0.9134, + "step": 7543 + }, + { + "epoch": 0.6803445010596564, + "grad_norm": 1.724091500693, + "learning_rate": 9.793286200619443e-07, + "loss": 0.9537, + "step": 7544 + }, + { + "epoch": 0.6804346845831266, + "grad_norm": 2.3002250271057147, + "learning_rate": 9.78826256914655e-07, + "loss": 1.0461, + "step": 7545 + }, + { + "epoch": 0.680524868106597, + "grad_norm": 1.72606088938011, + "learning_rate": 9.7832398089954e-07, + "loss": 1.0355, + "step": 7546 + }, + { + "epoch": 0.6806150516300672, + "grad_norm": 1.5576357849350186, + "learning_rate": 9.778217920594565e-07, + "loss": 1.0895, + "step": 7547 + }, + { + "epoch": 0.6807052351535374, + "grad_norm": 2.3713872435673298, + "learning_rate": 9.773196904372547e-07, + "loss": 1.0246, + "step": 7548 + }, + { + "epoch": 0.6807954186770077, + "grad_norm": 1.788052069501073, + "learning_rate": 9.768176760757742e-07, + "loss": 1.0698, + "step": 7549 + }, + { + "epoch": 0.680885602200478, + "grad_norm": 1.5736170676689993, + "learning_rate": 9.76315749017853e-07, + "loss": 0.9643, + "step": 7550 + }, + { + "epoch": 0.6809757857239482, + "grad_norm": 2.7796330412858032, + "learning_rate": 9.758139093063161e-07, + "loss": 0.9874, + "step": 7551 + }, + { + "epoch": 0.6810659692474185, + "grad_norm": 1.3688539212982878, + "learning_rate": 9.753121569839834e-07, + "loss": 1.052, + "step": 7552 + }, + { + "epoch": 0.6811561527708888, + "grad_norm": 0.728287034567265, + "learning_rate": 9.748104920936678e-07, + "loss": 0.7855, + "step": 7553 + }, + { + "epoch": 0.681246336294359, + "grad_norm": 1.5852830410293206, + "learning_rate": 9.743089146781738e-07, + "loss": 0.9606, + "step": 7554 + }, + { + "epoch": 0.6813365198178293, + "grad_norm": 1.3678845181939687, + "learning_rate": 9.738074247802988e-07, + "loss": 0.9094, + "step": 7555 + }, + { + "epoch": 0.6814267033412995, + "grad_norm": 1.6364863920406514, + "learning_rate": 9.733060224428325e-07, + "loss": 0.9978, + "step": 7556 + }, + { + "epoch": 0.6815168868647699, + "grad_norm": 1.79007455777086, + "learning_rate": 9.728047077085577e-07, + "loss": 0.9943, + "step": 7557 + }, + { + "epoch": 0.6816070703882401, + "grad_norm": 2.954082515434463, + "learning_rate": 9.723034806202497e-07, + "loss": 0.8866, + "step": 7558 + }, + { + "epoch": 0.6816972539117103, + "grad_norm": 6.6352439710180695, + "learning_rate": 9.718023412206748e-07, + "loss": 1.0018, + "step": 7559 + }, + { + "epoch": 0.6817874374351806, + "grad_norm": 1.449383255909135, + "learning_rate": 9.713012895525935e-07, + "loss": 1.0079, + "step": 7560 + }, + { + "epoch": 0.6818776209586509, + "grad_norm": 1.6676656192962043, + "learning_rate": 9.708003256587584e-07, + "loss": 0.9749, + "step": 7561 + }, + { + "epoch": 0.6819678044821211, + "grad_norm": 1.7567750030688252, + "learning_rate": 9.702994495819147e-07, + "loss": 1.0352, + "step": 7562 + }, + { + "epoch": 0.6820579880055914, + "grad_norm": 1.8287387046021866, + "learning_rate": 9.697986613647999e-07, + "loss": 0.8881, + "step": 7563 + }, + { + "epoch": 0.6821481715290616, + "grad_norm": 0.711820703567552, + "learning_rate": 9.692979610501425e-07, + "loss": 0.8499, + "step": 7564 + }, + { + "epoch": 0.6822383550525319, + "grad_norm": 1.686338872921686, + "learning_rate": 9.68797348680668e-07, + "loss": 1.0586, + "step": 7565 + }, + { + "epoch": 0.6823285385760022, + "grad_norm": 1.450426446753301, + "learning_rate": 9.682968242990878e-07, + "loss": 0.9534, + "step": 7566 + }, + { + "epoch": 0.6824187220994724, + "grad_norm": 1.4744887548970562, + "learning_rate": 9.677963879481132e-07, + "loss": 1.0356, + "step": 7567 + }, + { + "epoch": 0.6825089056229426, + "grad_norm": 1.7443673903799193, + "learning_rate": 9.672960396704416e-07, + "loss": 0.9731, + "step": 7568 + }, + { + "epoch": 0.682599089146413, + "grad_norm": 3.9069785697511357, + "learning_rate": 9.667957795087657e-07, + "loss": 0.9876, + "step": 7569 + }, + { + "epoch": 0.6826892726698832, + "grad_norm": 1.8464359181055838, + "learning_rate": 9.662956075057712e-07, + "loss": 1.0437, + "step": 7570 + }, + { + "epoch": 0.6827794561933535, + "grad_norm": 4.467026171284584, + "learning_rate": 9.657955237041354e-07, + "loss": 1.0496, + "step": 7571 + }, + { + "epoch": 0.6828696397168237, + "grad_norm": 1.4597533623861954, + "learning_rate": 9.652955281465278e-07, + "loss": 0.9642, + "step": 7572 + }, + { + "epoch": 0.682959823240294, + "grad_norm": 1.486795640120786, + "learning_rate": 9.64795620875612e-07, + "loss": 0.9861, + "step": 7573 + }, + { + "epoch": 0.6830500067637643, + "grad_norm": 1.6086141769349809, + "learning_rate": 9.64295801934041e-07, + "loss": 0.9294, + "step": 7574 + }, + { + "epoch": 0.6831401902872345, + "grad_norm": 0.6941015220709489, + "learning_rate": 9.63796071364463e-07, + "loss": 0.8251, + "step": 7575 + }, + { + "epoch": 0.6832303738107048, + "grad_norm": 2.4970927500397386, + "learning_rate": 9.632964292095179e-07, + "loss": 1.0693, + "step": 7576 + }, + { + "epoch": 0.6833205573341751, + "grad_norm": 1.3499333700762142, + "learning_rate": 9.627968755118374e-07, + "loss": 0.9705, + "step": 7577 + }, + { + "epoch": 0.6834107408576453, + "grad_norm": 0.6654760910528531, + "learning_rate": 9.622974103140468e-07, + "loss": 0.8043, + "step": 7578 + }, + { + "epoch": 0.6835009243811155, + "grad_norm": 1.8530517898173378, + "learning_rate": 9.617980336587632e-07, + "loss": 1.0458, + "step": 7579 + }, + { + "epoch": 0.6835911079045859, + "grad_norm": 2.577827945126061, + "learning_rate": 9.612987455885964e-07, + "loss": 0.868, + "step": 7580 + }, + { + "epoch": 0.6836812914280561, + "grad_norm": 1.5690448255973717, + "learning_rate": 9.607995461461467e-07, + "loss": 0.9859, + "step": 7581 + }, + { + "epoch": 0.6837714749515263, + "grad_norm": 1.4734291052149344, + "learning_rate": 9.603004353740111e-07, + "loss": 0.8856, + "step": 7582 + }, + { + "epoch": 0.6838616584749966, + "grad_norm": 0.6891385838723041, + "learning_rate": 9.598014133147738e-07, + "loss": 0.8602, + "step": 7583 + }, + { + "epoch": 0.6839518419984669, + "grad_norm": 1.5282383320913862, + "learning_rate": 9.59302480011017e-07, + "loss": 0.9435, + "step": 7584 + }, + { + "epoch": 0.6840420255219372, + "grad_norm": 0.675561020086416, + "learning_rate": 9.588036355053102e-07, + "loss": 0.8358, + "step": 7585 + }, + { + "epoch": 0.6841322090454074, + "grad_norm": 1.9689720876254555, + "learning_rate": 9.583048798402182e-07, + "loss": 0.8538, + "step": 7586 + }, + { + "epoch": 0.6842223925688776, + "grad_norm": 1.5299176238828383, + "learning_rate": 9.57806213058298e-07, + "loss": 0.978, + "step": 7587 + }, + { + "epoch": 0.684312576092348, + "grad_norm": 0.6924804489529528, + "learning_rate": 9.57307635202098e-07, + "loss": 0.8539, + "step": 7588 + }, + { + "epoch": 0.6844027596158182, + "grad_norm": 1.8089455577529985, + "learning_rate": 9.568091463141607e-07, + "loss": 1.0878, + "step": 7589 + }, + { + "epoch": 0.6844929431392884, + "grad_norm": 1.3922503113625566, + "learning_rate": 9.563107464370187e-07, + "loss": 0.9515, + "step": 7590 + }, + { + "epoch": 0.6845831266627587, + "grad_norm": 1.5507150631095823, + "learning_rate": 9.558124356131982e-07, + "loss": 0.9998, + "step": 7591 + }, + { + "epoch": 0.684673310186229, + "grad_norm": 2.2096302494729714, + "learning_rate": 9.553142138852187e-07, + "loss": 0.9184, + "step": 7592 + }, + { + "epoch": 0.6847634937096992, + "grad_norm": 1.808743958795947, + "learning_rate": 9.548160812955905e-07, + "loss": 0.893, + "step": 7593 + }, + { + "epoch": 0.6848536772331695, + "grad_norm": 1.462441960022666, + "learning_rate": 9.543180378868175e-07, + "loss": 0.805, + "step": 7594 + }, + { + "epoch": 0.6849438607566397, + "grad_norm": 2.157480193010736, + "learning_rate": 9.538200837013962e-07, + "loss": 1.0198, + "step": 7595 + }, + { + "epoch": 0.68503404428011, + "grad_norm": 0.6688761654499518, + "learning_rate": 9.533222187818122e-07, + "loss": 0.8518, + "step": 7596 + }, + { + "epoch": 0.6851242278035803, + "grad_norm": 1.6690520938219633, + "learning_rate": 9.528244431705492e-07, + "loss": 0.9926, + "step": 7597 + }, + { + "epoch": 0.6852144113270505, + "grad_norm": 1.542961564831965, + "learning_rate": 9.523267569100774e-07, + "loss": 0.9891, + "step": 7598 + }, + { + "epoch": 0.6853045948505209, + "grad_norm": 1.5435009002485205, + "learning_rate": 9.518291600428652e-07, + "loss": 0.9399, + "step": 7599 + }, + { + "epoch": 0.6853947783739911, + "grad_norm": 1.8134614433352652, + "learning_rate": 9.513316526113677e-07, + "loss": 1.0051, + "step": 7600 + }, + { + "epoch": 0.6854849618974613, + "grad_norm": 1.5658912477987932, + "learning_rate": 9.50834234658036e-07, + "loss": 1.0074, + "step": 7601 + }, + { + "epoch": 0.6855751454209316, + "grad_norm": 1.5399337084237494, + "learning_rate": 9.503369062253123e-07, + "loss": 1.0015, + "step": 7602 + }, + { + "epoch": 0.6856653289444019, + "grad_norm": 0.6391446058510483, + "learning_rate": 9.498396673556317e-07, + "loss": 0.8267, + "step": 7603 + }, + { + "epoch": 0.6857555124678721, + "grad_norm": 1.5882313492981532, + "learning_rate": 9.493425180914219e-07, + "loss": 1.0199, + "step": 7604 + }, + { + "epoch": 0.6858456959913424, + "grad_norm": 1.825026404829918, + "learning_rate": 9.488454584751e-07, + "loss": 0.9126, + "step": 7605 + }, + { + "epoch": 0.6859358795148126, + "grad_norm": 1.5871658454104292, + "learning_rate": 9.483484885490813e-07, + "loss": 0.9898, + "step": 7606 + }, + { + "epoch": 0.686026063038283, + "grad_norm": 1.8090711585579142, + "learning_rate": 9.478516083557675e-07, + "loss": 0.9978, + "step": 7607 + }, + { + "epoch": 0.6861162465617532, + "grad_norm": 2.0618066489276474, + "learning_rate": 9.473548179375561e-07, + "loss": 1.0051, + "step": 7608 + }, + { + "epoch": 0.6862064300852234, + "grad_norm": 0.688982254338226, + "learning_rate": 9.468581173368358e-07, + "loss": 0.8542, + "step": 7609 + }, + { + "epoch": 0.6862966136086937, + "grad_norm": 1.5546799089256023, + "learning_rate": 9.463615065959878e-07, + "loss": 0.9926, + "step": 7610 + }, + { + "epoch": 0.686386797132164, + "grad_norm": 2.2740295268534805, + "learning_rate": 9.458649857573857e-07, + "loss": 1.0384, + "step": 7611 + }, + { + "epoch": 0.6864769806556342, + "grad_norm": 1.5723440968140563, + "learning_rate": 9.453685548633963e-07, + "loss": 0.9936, + "step": 7612 + }, + { + "epoch": 0.6865671641791045, + "grad_norm": 1.672712312459954, + "learning_rate": 9.448722139563756e-07, + "loss": 0.9321, + "step": 7613 + }, + { + "epoch": 0.6866573477025747, + "grad_norm": 1.8839633321170226, + "learning_rate": 9.443759630786769e-07, + "loss": 0.8807, + "step": 7614 + }, + { + "epoch": 0.686747531226045, + "grad_norm": 1.5296562620252312, + "learning_rate": 9.438798022726408e-07, + "loss": 1.0552, + "step": 7615 + }, + { + "epoch": 0.6868377147495153, + "grad_norm": 1.732936296779099, + "learning_rate": 9.433837315806037e-07, + "loss": 0.9362, + "step": 7616 + }, + { + "epoch": 0.6869278982729855, + "grad_norm": 1.91091691611956, + "learning_rate": 9.428877510448925e-07, + "loss": 0.9431, + "step": 7617 + }, + { + "epoch": 0.6870180817964557, + "grad_norm": 1.6848229260322107, + "learning_rate": 9.423918607078272e-07, + "loss": 0.9134, + "step": 7618 + }, + { + "epoch": 0.6871082653199261, + "grad_norm": 1.7062989503874348, + "learning_rate": 9.418960606117208e-07, + "loss": 0.9964, + "step": 7619 + }, + { + "epoch": 0.6871984488433963, + "grad_norm": 1.4892686987307036, + "learning_rate": 9.414003507988752e-07, + "loss": 0.9272, + "step": 7620 + }, + { + "epoch": 0.6872886323668665, + "grad_norm": 1.6245252209106, + "learning_rate": 9.409047313115904e-07, + "loss": 0.9203, + "step": 7621 + }, + { + "epoch": 0.6873788158903368, + "grad_norm": 1.8384292251164658, + "learning_rate": 9.404092021921521e-07, + "loss": 0.926, + "step": 7622 + }, + { + "epoch": 0.6874689994138071, + "grad_norm": 1.800300583643544, + "learning_rate": 9.399137634828447e-07, + "loss": 0.964, + "step": 7623 + }, + { + "epoch": 0.6875591829372774, + "grad_norm": 1.6182297743333802, + "learning_rate": 9.394184152259396e-07, + "loss": 1.0061, + "step": 7624 + }, + { + "epoch": 0.6876493664607476, + "grad_norm": 1.469876445845788, + "learning_rate": 9.389231574637033e-07, + "loss": 1.0902, + "step": 7625 + }, + { + "epoch": 0.6877395499842179, + "grad_norm": 1.7334356427565274, + "learning_rate": 9.384279902383938e-07, + "loss": 1.0724, + "step": 7626 + }, + { + "epoch": 0.6878297335076882, + "grad_norm": 2.0356421302493937, + "learning_rate": 9.379329135922615e-07, + "loss": 1.0025, + "step": 7627 + }, + { + "epoch": 0.6879199170311584, + "grad_norm": 1.891576417234942, + "learning_rate": 9.374379275675495e-07, + "loss": 0.9444, + "step": 7628 + }, + { + "epoch": 0.6880101005546286, + "grad_norm": 2.4752180138330075, + "learning_rate": 9.369430322064931e-07, + "loss": 0.9814, + "step": 7629 + }, + { + "epoch": 0.688100284078099, + "grad_norm": 5.985511452756081, + "learning_rate": 9.364482275513179e-07, + "loss": 1.0612, + "step": 7630 + }, + { + "epoch": 0.6881904676015692, + "grad_norm": 1.5701148824512328, + "learning_rate": 9.359535136442444e-07, + "loss": 1.0512, + "step": 7631 + }, + { + "epoch": 0.6882806511250394, + "grad_norm": 1.6421026206107892, + "learning_rate": 9.354588905274843e-07, + "loss": 0.9189, + "step": 7632 + }, + { + "epoch": 0.6883708346485097, + "grad_norm": 1.4817384292822704, + "learning_rate": 9.349643582432414e-07, + "loss": 1.0167, + "step": 7633 + }, + { + "epoch": 0.68846101817198, + "grad_norm": 1.5461152648199858, + "learning_rate": 9.344699168337127e-07, + "loss": 0.9771, + "step": 7634 + }, + { + "epoch": 0.6885512016954503, + "grad_norm": 2.1152315571238733, + "learning_rate": 9.339755663410845e-07, + "loss": 0.947, + "step": 7635 + }, + { + "epoch": 0.6886413852189205, + "grad_norm": 2.212197312162053, + "learning_rate": 9.334813068075405e-07, + "loss": 1.0188, + "step": 7636 + }, + { + "epoch": 0.6887315687423907, + "grad_norm": 1.484551991401224, + "learning_rate": 9.329871382752506e-07, + "loss": 0.9893, + "step": 7637 + }, + { + "epoch": 0.6888217522658611, + "grad_norm": 1.7750691037762873, + "learning_rate": 9.32493060786383e-07, + "loss": 0.9738, + "step": 7638 + }, + { + "epoch": 0.6889119357893313, + "grad_norm": 1.9073756912143194, + "learning_rate": 9.31999074383093e-07, + "loss": 1.0441, + "step": 7639 + }, + { + "epoch": 0.6890021193128015, + "grad_norm": 1.7549710514256818, + "learning_rate": 9.315051791075308e-07, + "loss": 0.9788, + "step": 7640 + }, + { + "epoch": 0.6890923028362718, + "grad_norm": 1.3980254093556965, + "learning_rate": 9.310113750018382e-07, + "loss": 0.9795, + "step": 7641 + }, + { + "epoch": 0.6891824863597421, + "grad_norm": 2.9657560008050283, + "learning_rate": 9.305176621081496e-07, + "loss": 1.0556, + "step": 7642 + }, + { + "epoch": 0.6892726698832123, + "grad_norm": 1.9269627163917233, + "learning_rate": 9.300240404685911e-07, + "loss": 1.0216, + "step": 7643 + }, + { + "epoch": 0.6893628534066826, + "grad_norm": 1.9899376824939394, + "learning_rate": 9.295305101252812e-07, + "loss": 0.9869, + "step": 7644 + }, + { + "epoch": 0.6894530369301528, + "grad_norm": 1.8832460909303692, + "learning_rate": 9.290370711203314e-07, + "loss": 0.9664, + "step": 7645 + }, + { + "epoch": 0.6895432204536232, + "grad_norm": 1.4262673581376804, + "learning_rate": 9.285437234958433e-07, + "loss": 0.9831, + "step": 7646 + }, + { + "epoch": 0.6896334039770934, + "grad_norm": 1.7492888231711017, + "learning_rate": 9.280504672939124e-07, + "loss": 0.9756, + "step": 7647 + }, + { + "epoch": 0.6897235875005636, + "grad_norm": 1.4461683883653813, + "learning_rate": 9.275573025566266e-07, + "loss": 1.0235, + "step": 7648 + }, + { + "epoch": 0.689813771024034, + "grad_norm": 1.2451955015199883, + "learning_rate": 9.27064229326065e-07, + "loss": 1.018, + "step": 7649 + }, + { + "epoch": 0.6899039545475042, + "grad_norm": 1.4174679378805835, + "learning_rate": 9.265712476442995e-07, + "loss": 0.9509, + "step": 7650 + }, + { + "epoch": 0.6899941380709744, + "grad_norm": 1.7236659406120578, + "learning_rate": 9.260783575533949e-07, + "loss": 0.9153, + "step": 7651 + }, + { + "epoch": 0.6900843215944447, + "grad_norm": 2.022264059330329, + "learning_rate": 9.255855590954045e-07, + "loss": 0.9337, + "step": 7652 + }, + { + "epoch": 0.690174505117915, + "grad_norm": 1.5456286717304528, + "learning_rate": 9.250928523123802e-07, + "loss": 1.0471, + "step": 7653 + }, + { + "epoch": 0.6902646886413852, + "grad_norm": 1.5818326849008495, + "learning_rate": 9.24600237246359e-07, + "loss": 0.9906, + "step": 7654 + }, + { + "epoch": 0.6903548721648555, + "grad_norm": 2.416270015217254, + "learning_rate": 9.241077139393769e-07, + "loss": 0.9903, + "step": 7655 + }, + { + "epoch": 0.6904450556883257, + "grad_norm": 2.2212596185977205, + "learning_rate": 9.236152824334564e-07, + "loss": 1.0154, + "step": 7656 + }, + { + "epoch": 0.690535239211796, + "grad_norm": 1.6979875686289296, + "learning_rate": 9.231229427706151e-07, + "loss": 0.8837, + "step": 7657 + }, + { + "epoch": 0.6906254227352663, + "grad_norm": 5.484194326005918, + "learning_rate": 9.226306949928622e-07, + "loss": 0.991, + "step": 7658 + }, + { + "epoch": 0.6907156062587365, + "grad_norm": 1.5517789384329321, + "learning_rate": 9.221385391421988e-07, + "loss": 1.0576, + "step": 7659 + }, + { + "epoch": 0.6908057897822067, + "grad_norm": 1.4763906873419115, + "learning_rate": 9.216464752606192e-07, + "loss": 1.0034, + "step": 7660 + }, + { + "epoch": 0.6908959733056771, + "grad_norm": 1.4480504676233, + "learning_rate": 9.211545033901078e-07, + "loss": 1.0614, + "step": 7661 + }, + { + "epoch": 0.6909861568291473, + "grad_norm": 1.5326124085961559, + "learning_rate": 9.206626235726426e-07, + "loss": 1.0161, + "step": 7662 + }, + { + "epoch": 0.6910763403526176, + "grad_norm": 2.527865183603471, + "learning_rate": 9.20170835850194e-07, + "loss": 0.9247, + "step": 7663 + }, + { + "epoch": 0.6911665238760878, + "grad_norm": 0.642461122568572, + "learning_rate": 9.196791402647237e-07, + "loss": 0.8099, + "step": 7664 + }, + { + "epoch": 0.6912567073995581, + "grad_norm": 1.406672223601431, + "learning_rate": 9.191875368581861e-07, + "loss": 0.9089, + "step": 7665 + }, + { + "epoch": 0.6913468909230284, + "grad_norm": 1.734869740348235, + "learning_rate": 9.186960256725271e-07, + "loss": 0.9956, + "step": 7666 + }, + { + "epoch": 0.6914370744464986, + "grad_norm": 1.6046922574070255, + "learning_rate": 9.182046067496856e-07, + "loss": 1.0475, + "step": 7667 + }, + { + "epoch": 0.6915272579699688, + "grad_norm": 1.6669647983296585, + "learning_rate": 9.177132801315927e-07, + "loss": 1.0702, + "step": 7668 + }, + { + "epoch": 0.6916174414934392, + "grad_norm": 1.5270410418237967, + "learning_rate": 9.172220458601692e-07, + "loss": 0.9573, + "step": 7669 + }, + { + "epoch": 0.6917076250169094, + "grad_norm": 1.8227096948746833, + "learning_rate": 9.167309039773324e-07, + "loss": 0.9546, + "step": 7670 + }, + { + "epoch": 0.6917978085403796, + "grad_norm": 1.321300385307522, + "learning_rate": 9.162398545249872e-07, + "loss": 0.8826, + "step": 7671 + }, + { + "epoch": 0.69188799206385, + "grad_norm": 1.3636477177320638, + "learning_rate": 9.157488975450334e-07, + "loss": 1.0401, + "step": 7672 + }, + { + "epoch": 0.6919781755873202, + "grad_norm": 1.7136465425099336, + "learning_rate": 9.15258033079362e-07, + "loss": 0.8147, + "step": 7673 + }, + { + "epoch": 0.6920683591107905, + "grad_norm": 1.522639060473898, + "learning_rate": 9.147672611698567e-07, + "loss": 0.9723, + "step": 7674 + }, + { + "epoch": 0.6921585426342607, + "grad_norm": 1.8370818771779909, + "learning_rate": 9.142765818583933e-07, + "loss": 0.9772, + "step": 7675 + }, + { + "epoch": 0.692248726157731, + "grad_norm": 1.4120610863328982, + "learning_rate": 9.13785995186837e-07, + "loss": 0.964, + "step": 7676 + }, + { + "epoch": 0.6923389096812013, + "grad_norm": 1.4669432382920369, + "learning_rate": 9.132955011970502e-07, + "loss": 0.9523, + "step": 7677 + }, + { + "epoch": 0.6924290932046715, + "grad_norm": 1.3395972289239262, + "learning_rate": 9.128050999308827e-07, + "loss": 0.9722, + "step": 7678 + }, + { + "epoch": 0.6925192767281417, + "grad_norm": 1.6822583772969715, + "learning_rate": 9.123147914301789e-07, + "loss": 1.0709, + "step": 7679 + }, + { + "epoch": 0.6926094602516121, + "grad_norm": 1.7990689359797947, + "learning_rate": 9.118245757367745e-07, + "loss": 1.0304, + "step": 7680 + }, + { + "epoch": 0.6926996437750823, + "grad_norm": 0.6215304153239608, + "learning_rate": 9.113344528924973e-07, + "loss": 0.7971, + "step": 7681 + }, + { + "epoch": 0.6927898272985525, + "grad_norm": 1.6573320848614221, + "learning_rate": 9.108444229391676e-07, + "loss": 0.9643, + "step": 7682 + }, + { + "epoch": 0.6928800108220228, + "grad_norm": 23.810627540800322, + "learning_rate": 9.103544859185979e-07, + "loss": 0.9575, + "step": 7683 + }, + { + "epoch": 0.6929701943454931, + "grad_norm": 1.5834015363091438, + "learning_rate": 9.098646418725902e-07, + "loss": 1.0724, + "step": 7684 + }, + { + "epoch": 0.6930603778689634, + "grad_norm": 0.6322835666152324, + "learning_rate": 9.093748908429437e-07, + "loss": 0.8402, + "step": 7685 + }, + { + "epoch": 0.6931505613924336, + "grad_norm": 2.278528445171963, + "learning_rate": 9.088852328714444e-07, + "loss": 0.9761, + "step": 7686 + }, + { + "epoch": 0.6932407449159038, + "grad_norm": 1.7596577093671177, + "learning_rate": 9.083956679998735e-07, + "loss": 0.8918, + "step": 7687 + }, + { + "epoch": 0.6933309284393742, + "grad_norm": 1.3262671097955998, + "learning_rate": 9.079061962700032e-07, + "loss": 0.9521, + "step": 7688 + }, + { + "epoch": 0.6934211119628444, + "grad_norm": 1.8907345826729565, + "learning_rate": 9.074168177235979e-07, + "loss": 1.0107, + "step": 7689 + }, + { + "epoch": 0.6935112954863146, + "grad_norm": 1.6158930071074327, + "learning_rate": 9.069275324024151e-07, + "loss": 0.9986, + "step": 7690 + }, + { + "epoch": 0.6936014790097849, + "grad_norm": 1.8787729449974444, + "learning_rate": 9.064383403482005e-07, + "loss": 0.9172, + "step": 7691 + }, + { + "epoch": 0.6936916625332552, + "grad_norm": 1.2923990706063522, + "learning_rate": 9.059492416026983e-07, + "loss": 0.9464, + "step": 7692 + }, + { + "epoch": 0.6937818460567254, + "grad_norm": 1.6036841504915997, + "learning_rate": 9.054602362076378e-07, + "loss": 0.9801, + "step": 7693 + }, + { + "epoch": 0.6938720295801957, + "grad_norm": 1.9854240347961356, + "learning_rate": 9.049713242047468e-07, + "loss": 0.9744, + "step": 7694 + }, + { + "epoch": 0.693962213103666, + "grad_norm": 1.6767000822727822, + "learning_rate": 9.044825056357395e-07, + "loss": 0.9514, + "step": 7695 + }, + { + "epoch": 0.6940523966271362, + "grad_norm": 1.6820385469879895, + "learning_rate": 9.039937805423255e-07, + "loss": 0.938, + "step": 7696 + }, + { + "epoch": 0.6941425801506065, + "grad_norm": 0.6029815038640501, + "learning_rate": 9.035051489662051e-07, + "loss": 0.7548, + "step": 7697 + }, + { + "epoch": 0.6942327636740767, + "grad_norm": 1.5749903602910609, + "learning_rate": 9.030166109490718e-07, + "loss": 0.9357, + "step": 7698 + }, + { + "epoch": 0.6943229471975471, + "grad_norm": 2.2100393674977585, + "learning_rate": 9.025281665326099e-07, + "loss": 0.9926, + "step": 7699 + }, + { + "epoch": 0.6944131307210173, + "grad_norm": 1.8912675254398048, + "learning_rate": 9.020398157584967e-07, + "loss": 0.9929, + "step": 7700 + }, + { + "epoch": 0.6945033142444875, + "grad_norm": 1.9837019372128224, + "learning_rate": 9.015515586684002e-07, + "loss": 0.8855, + "step": 7701 + }, + { + "epoch": 0.6945934977679578, + "grad_norm": 0.6513220043610874, + "learning_rate": 9.010633953039812e-07, + "loss": 0.8295, + "step": 7702 + }, + { + "epoch": 0.6946836812914281, + "grad_norm": 1.5491920611246324, + "learning_rate": 9.005753257068929e-07, + "loss": 0.9969, + "step": 7703 + }, + { + "epoch": 0.6947738648148983, + "grad_norm": 1.6819138786729917, + "learning_rate": 9.000873499187797e-07, + "loss": 0.996, + "step": 7704 + }, + { + "epoch": 0.6948640483383686, + "grad_norm": 1.6262869507141822, + "learning_rate": 8.995994679812797e-07, + "loss": 1.0147, + "step": 7705 + }, + { + "epoch": 0.6949542318618388, + "grad_norm": 2.130756099446478, + "learning_rate": 8.991116799360192e-07, + "loss": 1.0071, + "step": 7706 + }, + { + "epoch": 0.6950444153853091, + "grad_norm": 1.5510000847989225, + "learning_rate": 8.986239858246217e-07, + "loss": 0.9762, + "step": 7707 + }, + { + "epoch": 0.6951345989087794, + "grad_norm": 2.1379628578221377, + "learning_rate": 8.981363856886972e-07, + "loss": 1.0652, + "step": 7708 + }, + { + "epoch": 0.6952247824322496, + "grad_norm": 1.694207459781288, + "learning_rate": 8.976488795698533e-07, + "loss": 0.964, + "step": 7709 + }, + { + "epoch": 0.6953149659557198, + "grad_norm": 1.7863244839749413, + "learning_rate": 8.971614675096841e-07, + "loss": 0.9975, + "step": 7710 + }, + { + "epoch": 0.6954051494791902, + "grad_norm": 1.963284975739787, + "learning_rate": 8.966741495497807e-07, + "loss": 1.0506, + "step": 7711 + }, + { + "epoch": 0.6954953330026604, + "grad_norm": 1.3415288976854711, + "learning_rate": 8.961869257317218e-07, + "loss": 0.9512, + "step": 7712 + }, + { + "epoch": 0.6955855165261307, + "grad_norm": 1.5565706282784253, + "learning_rate": 8.956997960970809e-07, + "loss": 0.965, + "step": 7713 + }, + { + "epoch": 0.6956757000496009, + "grad_norm": 1.6770118161071241, + "learning_rate": 8.952127606874224e-07, + "loss": 0.8921, + "step": 7714 + }, + { + "epoch": 0.6957658835730712, + "grad_norm": 1.5547781299647703, + "learning_rate": 8.947258195443028e-07, + "loss": 0.985, + "step": 7715 + }, + { + "epoch": 0.6958560670965415, + "grad_norm": 2.354128125530723, + "learning_rate": 8.942389727092716e-07, + "loss": 0.9593, + "step": 7716 + }, + { + "epoch": 0.6959462506200117, + "grad_norm": 0.6787536168572883, + "learning_rate": 8.937522202238677e-07, + "loss": 0.8014, + "step": 7717 + }, + { + "epoch": 0.696036434143482, + "grad_norm": 1.8637994649400627, + "learning_rate": 8.932655621296239e-07, + "loss": 0.9981, + "step": 7718 + }, + { + "epoch": 0.6961266176669523, + "grad_norm": 1.3720456375672723, + "learning_rate": 8.927789984680649e-07, + "loss": 0.9794, + "step": 7719 + }, + { + "epoch": 0.6962168011904225, + "grad_norm": 1.8216367824275221, + "learning_rate": 8.922925292807068e-07, + "loss": 0.85, + "step": 7720 + }, + { + "epoch": 0.6963069847138927, + "grad_norm": 1.5489595027971332, + "learning_rate": 8.91806154609058e-07, + "loss": 1.0662, + "step": 7721 + }, + { + "epoch": 0.6963971682373631, + "grad_norm": 2.1025426911545555, + "learning_rate": 8.913198744946195e-07, + "loss": 0.9548, + "step": 7722 + }, + { + "epoch": 0.6964873517608333, + "grad_norm": 1.6256081103511175, + "learning_rate": 8.908336889788807e-07, + "loss": 1.0082, + "step": 7723 + }, + { + "epoch": 0.6965775352843036, + "grad_norm": 1.6642624282728138, + "learning_rate": 8.903475981033293e-07, + "loss": 0.8606, + "step": 7724 + }, + { + "epoch": 0.6966677188077738, + "grad_norm": 0.7390043159147193, + "learning_rate": 8.898616019094376e-07, + "loss": 0.8696, + "step": 7725 + }, + { + "epoch": 0.6967579023312441, + "grad_norm": 1.7137360699266249, + "learning_rate": 8.89375700438677e-07, + "loss": 1.0223, + "step": 7726 + }, + { + "epoch": 0.6968480858547144, + "grad_norm": 1.9088400458160093, + "learning_rate": 8.888898937325047e-07, + "loss": 0.9469, + "step": 7727 + }, + { + "epoch": 0.6969382693781846, + "grad_norm": 2.768000816097029, + "learning_rate": 8.884041818323733e-07, + "loss": 0.9467, + "step": 7728 + }, + { + "epoch": 0.6970284529016548, + "grad_norm": 1.6195367528708746, + "learning_rate": 8.879185647797262e-07, + "loss": 1.0164, + "step": 7729 + }, + { + "epoch": 0.6971186364251252, + "grad_norm": 1.6102331891804391, + "learning_rate": 8.874330426159993e-07, + "loss": 0.9611, + "step": 7730 + }, + { + "epoch": 0.6972088199485954, + "grad_norm": 1.6183468997961574, + "learning_rate": 8.869476153826205e-07, + "loss": 1.0094, + "step": 7731 + }, + { + "epoch": 0.6972990034720656, + "grad_norm": 1.5622176868981237, + "learning_rate": 8.864622831210071e-07, + "loss": 0.9843, + "step": 7732 + }, + { + "epoch": 0.6973891869955359, + "grad_norm": 1.892122857574885, + "learning_rate": 8.85977045872573e-07, + "loss": 0.9911, + "step": 7733 + }, + { + "epoch": 0.6974793705190062, + "grad_norm": 1.5433293561012207, + "learning_rate": 8.854919036787194e-07, + "loss": 0.9609, + "step": 7734 + }, + { + "epoch": 0.6975695540424764, + "grad_norm": 1.7229899042224508, + "learning_rate": 8.850068565808417e-07, + "loss": 0.9494, + "step": 7735 + }, + { + "epoch": 0.6976597375659467, + "grad_norm": 2.072849317315989, + "learning_rate": 8.845219046203271e-07, + "loss": 1.0304, + "step": 7736 + }, + { + "epoch": 0.6977499210894169, + "grad_norm": 1.9727007266025633, + "learning_rate": 8.840370478385544e-07, + "loss": 1.0078, + "step": 7737 + }, + { + "epoch": 0.6978401046128873, + "grad_norm": 1.725105533965172, + "learning_rate": 8.83552286276894e-07, + "loss": 1.0036, + "step": 7738 + }, + { + "epoch": 0.6979302881363575, + "grad_norm": 1.7762463210280544, + "learning_rate": 8.830676199767095e-07, + "loss": 1.002, + "step": 7739 + }, + { + "epoch": 0.6980204716598277, + "grad_norm": 1.5665707775287652, + "learning_rate": 8.825830489793527e-07, + "loss": 0.9624, + "step": 7740 + }, + { + "epoch": 0.698110655183298, + "grad_norm": 1.4680182886621533, + "learning_rate": 8.820985733261732e-07, + "loss": 0.9364, + "step": 7741 + }, + { + "epoch": 0.6982008387067683, + "grad_norm": 1.7592077524702188, + "learning_rate": 8.816141930585066e-07, + "loss": 0.8929, + "step": 7742 + }, + { + "epoch": 0.6982910222302385, + "grad_norm": 0.6994114615555258, + "learning_rate": 8.811299082176837e-07, + "loss": 0.8266, + "step": 7743 + }, + { + "epoch": 0.6983812057537088, + "grad_norm": 1.602361340289045, + "learning_rate": 8.806457188450265e-07, + "loss": 0.9494, + "step": 7744 + }, + { + "epoch": 0.6984713892771791, + "grad_norm": 1.5343816224866234, + "learning_rate": 8.801616249818487e-07, + "loss": 1.0226, + "step": 7745 + }, + { + "epoch": 0.6985615728006493, + "grad_norm": 2.4429106692932865, + "learning_rate": 8.796776266694564e-07, + "loss": 0.963, + "step": 7746 + }, + { + "epoch": 0.6986517563241196, + "grad_norm": 1.2805485549831024, + "learning_rate": 8.79193723949145e-07, + "loss": 0.9293, + "step": 7747 + }, + { + "epoch": 0.6987419398475898, + "grad_norm": 4.796920726446347, + "learning_rate": 8.787099168622063e-07, + "loss": 0.8797, + "step": 7748 + }, + { + "epoch": 0.6988321233710602, + "grad_norm": 1.7339606262629426, + "learning_rate": 8.782262054499199e-07, + "loss": 0.9232, + "step": 7749 + }, + { + "epoch": 0.6989223068945304, + "grad_norm": 1.3865163306131911, + "learning_rate": 8.777425897535588e-07, + "loss": 1.0129, + "step": 7750 + }, + { + "epoch": 0.6990124904180006, + "grad_norm": 1.6796522979596151, + "learning_rate": 8.77259069814388e-07, + "loss": 0.9844, + "step": 7751 + }, + { + "epoch": 0.6991026739414709, + "grad_norm": 1.5212024432407192, + "learning_rate": 8.767756456736641e-07, + "loss": 0.9803, + "step": 7752 + }, + { + "epoch": 0.6991928574649412, + "grad_norm": 1.6579962108186832, + "learning_rate": 8.762923173726358e-07, + "loss": 0.9081, + "step": 7753 + }, + { + "epoch": 0.6992830409884114, + "grad_norm": 1.8199173003051405, + "learning_rate": 8.758090849525428e-07, + "loss": 1.0797, + "step": 7754 + }, + { + "epoch": 0.6993732245118817, + "grad_norm": 1.7798678741280285, + "learning_rate": 8.753259484546174e-07, + "loss": 0.9219, + "step": 7755 + }, + { + "epoch": 0.6994634080353519, + "grad_norm": 2.470764593633829, + "learning_rate": 8.748429079200841e-07, + "loss": 0.9418, + "step": 7756 + }, + { + "epoch": 0.6995535915588222, + "grad_norm": 2.681141628396112, + "learning_rate": 8.743599633901575e-07, + "loss": 1.0372, + "step": 7757 + }, + { + "epoch": 0.6996437750822925, + "grad_norm": 1.6073099609573294, + "learning_rate": 8.738771149060453e-07, + "loss": 0.9987, + "step": 7758 + }, + { + "epoch": 0.6997339586057627, + "grad_norm": 1.6487029526297732, + "learning_rate": 8.73394362508947e-07, + "loss": 0.9501, + "step": 7759 + }, + { + "epoch": 0.6998241421292329, + "grad_norm": 1.624404578062556, + "learning_rate": 8.72911706240054e-07, + "loss": 0.9479, + "step": 7760 + }, + { + "epoch": 0.6999143256527033, + "grad_norm": 1.4358974314216293, + "learning_rate": 8.724291461405493e-07, + "loss": 0.8591, + "step": 7761 + }, + { + "epoch": 0.7000045091761735, + "grad_norm": 1.8034259300409436, + "learning_rate": 8.71946682251606e-07, + "loss": 0.9485, + "step": 7762 + }, + { + "epoch": 0.7000946926996438, + "grad_norm": 1.4750529651507434, + "learning_rate": 8.714643146143932e-07, + "loss": 0.9791, + "step": 7763 + }, + { + "epoch": 0.700184876223114, + "grad_norm": 1.424137871936454, + "learning_rate": 8.709820432700663e-07, + "loss": 0.9371, + "step": 7764 + }, + { + "epoch": 0.7002750597465843, + "grad_norm": 1.3092355549197368, + "learning_rate": 8.704998682597784e-07, + "loss": 0.8948, + "step": 7765 + }, + { + "epoch": 0.7003652432700546, + "grad_norm": 1.804420220205151, + "learning_rate": 8.700177896246688e-07, + "loss": 1.0016, + "step": 7766 + }, + { + "epoch": 0.7004554267935248, + "grad_norm": 1.9142256063915029, + "learning_rate": 8.695358074058721e-07, + "loss": 0.9083, + "step": 7767 + }, + { + "epoch": 0.7005456103169951, + "grad_norm": 4.412269971000989, + "learning_rate": 8.690539216445136e-07, + "loss": 1.0183, + "step": 7768 + }, + { + "epoch": 0.7006357938404654, + "grad_norm": 2.2388005744337662, + "learning_rate": 8.685721323817106e-07, + "loss": 0.8829, + "step": 7769 + }, + { + "epoch": 0.7007259773639356, + "grad_norm": 1.8573851636914918, + "learning_rate": 8.680904396585718e-07, + "loss": 1.0177, + "step": 7770 + }, + { + "epoch": 0.7008161608874058, + "grad_norm": 1.886228572756619, + "learning_rate": 8.676088435161988e-07, + "loss": 0.921, + "step": 7771 + }, + { + "epoch": 0.7009063444108762, + "grad_norm": 1.7988052615225545, + "learning_rate": 8.671273439956824e-07, + "loss": 0.9959, + "step": 7772 + }, + { + "epoch": 0.7009965279343464, + "grad_norm": 1.633799350521441, + "learning_rate": 8.666459411381075e-07, + "loss": 0.9858, + "step": 7773 + }, + { + "epoch": 0.7010867114578166, + "grad_norm": 1.3903366389958196, + "learning_rate": 8.661646349845501e-07, + "loss": 1.0071, + "step": 7774 + }, + { + "epoch": 0.7011768949812869, + "grad_norm": 1.7639298530220657, + "learning_rate": 8.656834255760783e-07, + "loss": 1.0375, + "step": 7775 + }, + { + "epoch": 0.7012670785047572, + "grad_norm": 1.554231969911952, + "learning_rate": 8.652023129537509e-07, + "loss": 1.0358, + "step": 7776 + }, + { + "epoch": 0.7013572620282275, + "grad_norm": 1.6929305962601249, + "learning_rate": 8.647212971586195e-07, + "loss": 1.0082, + "step": 7777 + }, + { + "epoch": 0.7014474455516977, + "grad_norm": 4.081943405155273, + "learning_rate": 8.642403782317275e-07, + "loss": 0.9318, + "step": 7778 + }, + { + "epoch": 0.7015376290751679, + "grad_norm": 2.750154057435715, + "learning_rate": 8.637595562141075e-07, + "loss": 0.9854, + "step": 7779 + }, + { + "epoch": 0.7016278125986383, + "grad_norm": 1.6376390441800133, + "learning_rate": 8.632788311467889e-07, + "loss": 0.9493, + "step": 7780 + }, + { + "epoch": 0.7017179961221085, + "grad_norm": 1.417173472611592, + "learning_rate": 8.627982030707867e-07, + "loss": 0.9758, + "step": 7781 + }, + { + "epoch": 0.7018081796455787, + "grad_norm": 1.7003290867640588, + "learning_rate": 8.623176720271139e-07, + "loss": 1.0484, + "step": 7782 + }, + { + "epoch": 0.701898363169049, + "grad_norm": 1.983948008336742, + "learning_rate": 8.618372380567696e-07, + "loss": 1.0519, + "step": 7783 + }, + { + "epoch": 0.7019885466925193, + "grad_norm": 1.8983171212919385, + "learning_rate": 8.613569012007478e-07, + "loss": 0.9424, + "step": 7784 + }, + { + "epoch": 0.7020787302159895, + "grad_norm": 2.448961259609493, + "learning_rate": 8.608766615000338e-07, + "loss": 1.0209, + "step": 7785 + }, + { + "epoch": 0.7021689137394598, + "grad_norm": 1.803328890321985, + "learning_rate": 8.603965189956039e-07, + "loss": 1.0032, + "step": 7786 + }, + { + "epoch": 0.70225909726293, + "grad_norm": 67.1086055689302, + "learning_rate": 8.599164737284276e-07, + "loss": 0.9869, + "step": 7787 + }, + { + "epoch": 0.7023492807864004, + "grad_norm": 1.6595351549581565, + "learning_rate": 8.594365257394634e-07, + "loss": 0.9405, + "step": 7788 + }, + { + "epoch": 0.7024394643098706, + "grad_norm": 1.7943800480956602, + "learning_rate": 8.589566750696637e-07, + "loss": 0.9237, + "step": 7789 + }, + { + "epoch": 0.7025296478333408, + "grad_norm": 3.174289629907929, + "learning_rate": 8.584769217599721e-07, + "loss": 0.9148, + "step": 7790 + }, + { + "epoch": 0.7026198313568112, + "grad_norm": 3.675871407095393, + "learning_rate": 8.579972658513239e-07, + "loss": 0.9297, + "step": 7791 + }, + { + "epoch": 0.7027100148802814, + "grad_norm": 1.9144072553038856, + "learning_rate": 8.57517707384646e-07, + "loss": 0.9858, + "step": 7792 + }, + { + "epoch": 0.7028001984037516, + "grad_norm": 1.5564378249717943, + "learning_rate": 8.570382464008574e-07, + "loss": 1.022, + "step": 7793 + }, + { + "epoch": 0.7028903819272219, + "grad_norm": 1.7124145263725825, + "learning_rate": 8.565588829408665e-07, + "loss": 0.9481, + "step": 7794 + }, + { + "epoch": 0.7029805654506922, + "grad_norm": 1.6938443484896717, + "learning_rate": 8.560796170455782e-07, + "loss": 0.9575, + "step": 7795 + }, + { + "epoch": 0.7030707489741624, + "grad_norm": 1.6605838718568813, + "learning_rate": 8.556004487558828e-07, + "loss": 1.0203, + "step": 7796 + }, + { + "epoch": 0.7031609324976327, + "grad_norm": 1.5203868349883793, + "learning_rate": 8.55121378112669e-07, + "loss": 0.8969, + "step": 7797 + }, + { + "epoch": 0.7032511160211029, + "grad_norm": 2.3013588871596156, + "learning_rate": 8.546424051568111e-07, + "loss": 0.9608, + "step": 7798 + }, + { + "epoch": 0.7033412995445733, + "grad_norm": 1.7781840255943095, + "learning_rate": 8.541635299291785e-07, + "loss": 1.0306, + "step": 7799 + }, + { + "epoch": 0.7034314830680435, + "grad_norm": 0.6081011246231537, + "learning_rate": 8.536847524706317e-07, + "loss": 0.8373, + "step": 7800 + }, + { + "epoch": 0.7035216665915137, + "grad_norm": 1.6012229745625586, + "learning_rate": 8.532060728220225e-07, + "loss": 0.9105, + "step": 7801 + }, + { + "epoch": 0.703611850114984, + "grad_norm": 1.7964837353185354, + "learning_rate": 8.527274910241955e-07, + "loss": 0.9904, + "step": 7802 + }, + { + "epoch": 0.7037020336384543, + "grad_norm": 1.5399523645935562, + "learning_rate": 8.522490071179833e-07, + "loss": 0.9776, + "step": 7803 + }, + { + "epoch": 0.7037922171619245, + "grad_norm": 1.932844290321562, + "learning_rate": 8.517706211442159e-07, + "loss": 0.8585, + "step": 7804 + }, + { + "epoch": 0.7038824006853948, + "grad_norm": 1.9322758662836528, + "learning_rate": 8.512923331437097e-07, + "loss": 1.0587, + "step": 7805 + }, + { + "epoch": 0.703972584208865, + "grad_norm": 1.478149529993505, + "learning_rate": 8.508141431572755e-07, + "loss": 0.9502, + "step": 7806 + }, + { + "epoch": 0.7040627677323353, + "grad_norm": 2.2991647530584975, + "learning_rate": 8.503360512257152e-07, + "loss": 0.9663, + "step": 7807 + }, + { + "epoch": 0.7041529512558056, + "grad_norm": 1.6189713521900368, + "learning_rate": 8.498580573898219e-07, + "loss": 0.9465, + "step": 7808 + }, + { + "epoch": 0.7042431347792758, + "grad_norm": 2.352621353376025, + "learning_rate": 8.493801616903813e-07, + "loss": 0.9881, + "step": 7809 + }, + { + "epoch": 0.704333318302746, + "grad_norm": 1.4995234845170935, + "learning_rate": 8.489023641681705e-07, + "loss": 0.9986, + "step": 7810 + }, + { + "epoch": 0.7044235018262164, + "grad_norm": 1.5129151649366088, + "learning_rate": 8.484246648639555e-07, + "loss": 1.0303, + "step": 7811 + }, + { + "epoch": 0.7045136853496866, + "grad_norm": 1.3730931498080992, + "learning_rate": 8.479470638184994e-07, + "loss": 0.8538, + "step": 7812 + }, + { + "epoch": 0.7046038688731568, + "grad_norm": 1.8764738012691158, + "learning_rate": 8.474695610725513e-07, + "loss": 0.8951, + "step": 7813 + }, + { + "epoch": 0.7046940523966272, + "grad_norm": 1.4932455735446755, + "learning_rate": 8.469921566668552e-07, + "loss": 1.0532, + "step": 7814 + }, + { + "epoch": 0.7047842359200974, + "grad_norm": 1.568104724471679, + "learning_rate": 8.46514850642146e-07, + "loss": 1.0283, + "step": 7815 + }, + { + "epoch": 0.7048744194435677, + "grad_norm": 1.6440294818647871, + "learning_rate": 8.460376430391499e-07, + "loss": 0.966, + "step": 7816 + }, + { + "epoch": 0.7049646029670379, + "grad_norm": 1.2001583630652861, + "learning_rate": 8.455605338985858e-07, + "loss": 1.0083, + "step": 7817 + }, + { + "epoch": 0.7050547864905082, + "grad_norm": 1.5367053420660548, + "learning_rate": 8.45083523261161e-07, + "loss": 0.8394, + "step": 7818 + }, + { + "epoch": 0.7051449700139785, + "grad_norm": 1.318315881082316, + "learning_rate": 8.446066111675796e-07, + "loss": 0.9861, + "step": 7819 + }, + { + "epoch": 0.7052351535374487, + "grad_norm": 2.125682833298981, + "learning_rate": 8.441297976585314e-07, + "loss": 0.8738, + "step": 7820 + }, + { + "epoch": 0.7053253370609189, + "grad_norm": 1.5891521927418302, + "learning_rate": 8.436530827747037e-07, + "loss": 1.0646, + "step": 7821 + }, + { + "epoch": 0.7054155205843893, + "grad_norm": 1.6833558235302724, + "learning_rate": 8.431764665567704e-07, + "loss": 1.1174, + "step": 7822 + }, + { + "epoch": 0.7055057041078595, + "grad_norm": 2.238834865203305, + "learning_rate": 8.426999490453996e-07, + "loss": 0.9226, + "step": 7823 + }, + { + "epoch": 0.7055958876313297, + "grad_norm": 1.9274382668194336, + "learning_rate": 8.422235302812504e-07, + "loss": 0.9874, + "step": 7824 + }, + { + "epoch": 0.7056860711548, + "grad_norm": 1.732921025254728, + "learning_rate": 8.417472103049734e-07, + "loss": 1.0506, + "step": 7825 + }, + { + "epoch": 0.7057762546782703, + "grad_norm": 1.7105177969855565, + "learning_rate": 8.412709891572112e-07, + "loss": 1.0179, + "step": 7826 + }, + { + "epoch": 0.7058664382017406, + "grad_norm": 1.88143883451275, + "learning_rate": 8.407948668785978e-07, + "loss": 1.0158, + "step": 7827 + }, + { + "epoch": 0.7059566217252108, + "grad_norm": 1.550539396629995, + "learning_rate": 8.403188435097576e-07, + "loss": 0.9719, + "step": 7828 + }, + { + "epoch": 0.706046805248681, + "grad_norm": 1.5330955127441632, + "learning_rate": 8.398429190913081e-07, + "loss": 1.0526, + "step": 7829 + }, + { + "epoch": 0.7061369887721514, + "grad_norm": 1.8771064370187474, + "learning_rate": 8.393670936638578e-07, + "loss": 0.981, + "step": 7830 + }, + { + "epoch": 0.7062271722956216, + "grad_norm": 2.2589559332613702, + "learning_rate": 8.388913672680067e-07, + "loss": 0.9784, + "step": 7831 + }, + { + "epoch": 0.7063173558190918, + "grad_norm": 1.681556123566532, + "learning_rate": 8.384157399443472e-07, + "loss": 0.9281, + "step": 7832 + }, + { + "epoch": 0.7064075393425621, + "grad_norm": 1.9484770133777245, + "learning_rate": 8.379402117334601e-07, + "loss": 1.0043, + "step": 7833 + }, + { + "epoch": 0.7064977228660324, + "grad_norm": 1.985543275087265, + "learning_rate": 8.374647826759232e-07, + "loss": 0.9804, + "step": 7834 + }, + { + "epoch": 0.7065879063895026, + "grad_norm": 2.068314534710775, + "learning_rate": 8.369894528122998e-07, + "loss": 0.9787, + "step": 7835 + }, + { + "epoch": 0.7066780899129729, + "grad_norm": 1.9380063041450628, + "learning_rate": 8.365142221831505e-07, + "loss": 0.8404, + "step": 7836 + }, + { + "epoch": 0.7067682734364432, + "grad_norm": 1.4866469814439887, + "learning_rate": 8.360390908290222e-07, + "loss": 0.9419, + "step": 7837 + }, + { + "epoch": 0.7068584569599135, + "grad_norm": 1.9777250226029672, + "learning_rate": 8.355640587904569e-07, + "loss": 0.9581, + "step": 7838 + }, + { + "epoch": 0.7069486404833837, + "grad_norm": 1.703304316328899, + "learning_rate": 8.350891261079866e-07, + "loss": 0.9915, + "step": 7839 + }, + { + "epoch": 0.7070388240068539, + "grad_norm": 1.6429598225320636, + "learning_rate": 8.346142928221356e-07, + "loss": 1.0247, + "step": 7840 + }, + { + "epoch": 0.7071290075303243, + "grad_norm": 2.0416603347783937, + "learning_rate": 8.341395589734189e-07, + "loss": 0.8248, + "step": 7841 + }, + { + "epoch": 0.7072191910537945, + "grad_norm": 2.1028201512857834, + "learning_rate": 8.336649246023433e-07, + "loss": 0.9327, + "step": 7842 + }, + { + "epoch": 0.7073093745772647, + "grad_norm": 1.5616252739522174, + "learning_rate": 8.331903897494084e-07, + "loss": 0.885, + "step": 7843 + }, + { + "epoch": 0.707399558100735, + "grad_norm": 1.949849976368343, + "learning_rate": 8.327159544551024e-07, + "loss": 0.9928, + "step": 7844 + }, + { + "epoch": 0.7074897416242053, + "grad_norm": 1.685775723619319, + "learning_rate": 8.322416187599073e-07, + "loss": 0.9414, + "step": 7845 + }, + { + "epoch": 0.7075799251476755, + "grad_norm": 1.8558696755372495, + "learning_rate": 8.317673827042963e-07, + "loss": 1.005, + "step": 7846 + }, + { + "epoch": 0.7076701086711458, + "grad_norm": 1.8008798170938127, + "learning_rate": 8.312932463287339e-07, + "loss": 0.9685, + "step": 7847 + }, + { + "epoch": 0.707760292194616, + "grad_norm": 3.1916444052242747, + "learning_rate": 8.308192096736759e-07, + "loss": 0.9478, + "step": 7848 + }, + { + "epoch": 0.7078504757180863, + "grad_norm": 1.4941268222146358, + "learning_rate": 8.303452727795703e-07, + "loss": 0.9569, + "step": 7849 + }, + { + "epoch": 0.7079406592415566, + "grad_norm": 2.6272542447654708, + "learning_rate": 8.298714356868542e-07, + "loss": 1.0583, + "step": 7850 + }, + { + "epoch": 0.7080308427650268, + "grad_norm": 1.9693253297150275, + "learning_rate": 8.293976984359605e-07, + "loss": 0.994, + "step": 7851 + }, + { + "epoch": 0.708121026288497, + "grad_norm": 1.7572571279783065, + "learning_rate": 8.289240610673092e-07, + "loss": 0.9825, + "step": 7852 + }, + { + "epoch": 0.7082112098119674, + "grad_norm": 1.4052257622537423, + "learning_rate": 8.284505236213144e-07, + "loss": 1.0139, + "step": 7853 + }, + { + "epoch": 0.7083013933354376, + "grad_norm": 1.8448068208536932, + "learning_rate": 8.279770861383806e-07, + "loss": 0.9736, + "step": 7854 + }, + { + "epoch": 0.7083915768589079, + "grad_norm": 1.948040133828508, + "learning_rate": 8.275037486589042e-07, + "loss": 0.9754, + "step": 7855 + }, + { + "epoch": 0.7084817603823781, + "grad_norm": 0.6712441476818963, + "learning_rate": 8.270305112232739e-07, + "loss": 0.7963, + "step": 7856 + }, + { + "epoch": 0.7085719439058484, + "grad_norm": 2.1698929778749245, + "learning_rate": 8.265573738718665e-07, + "loss": 0.9788, + "step": 7857 + }, + { + "epoch": 0.7086621274293187, + "grad_norm": 1.584326240346034, + "learning_rate": 8.260843366450559e-07, + "loss": 0.9424, + "step": 7858 + }, + { + "epoch": 0.7087523109527889, + "grad_norm": 1.50121059839282, + "learning_rate": 8.256113995832017e-07, + "loss": 0.98, + "step": 7859 + }, + { + "epoch": 0.7088424944762591, + "grad_norm": 1.5746960543464312, + "learning_rate": 8.251385627266583e-07, + "loss": 1.0226, + "step": 7860 + }, + { + "epoch": 0.7089326779997295, + "grad_norm": 2.00049763687326, + "learning_rate": 8.24665826115771e-07, + "loss": 0.992, + "step": 7861 + }, + { + "epoch": 0.7090228615231997, + "grad_norm": 1.4611653033038738, + "learning_rate": 8.241931897908763e-07, + "loss": 1.0006, + "step": 7862 + }, + { + "epoch": 0.7091130450466699, + "grad_norm": 1.5346981073539976, + "learning_rate": 8.237206537923016e-07, + "loss": 1.0028, + "step": 7863 + }, + { + "epoch": 0.7092032285701403, + "grad_norm": 1.95942629757645, + "learning_rate": 8.232482181603671e-07, + "loss": 0.9685, + "step": 7864 + }, + { + "epoch": 0.7092934120936105, + "grad_norm": 1.7268718597681776, + "learning_rate": 8.227758829353828e-07, + "loss": 0.9605, + "step": 7865 + }, + { + "epoch": 0.7093835956170808, + "grad_norm": 1.694037665681372, + "learning_rate": 8.223036481576522e-07, + "loss": 0.9465, + "step": 7866 + }, + { + "epoch": 0.709473779140551, + "grad_norm": 2.6125099400965484, + "learning_rate": 8.218315138674672e-07, + "loss": 0.9842, + "step": 7867 + }, + { + "epoch": 0.7095639626640213, + "grad_norm": 0.6159422420123692, + "learning_rate": 8.21359480105114e-07, + "loss": 0.8137, + "step": 7868 + }, + { + "epoch": 0.7096541461874916, + "grad_norm": 1.710397504481573, + "learning_rate": 8.208875469108689e-07, + "loss": 1.0122, + "step": 7869 + }, + { + "epoch": 0.7097443297109618, + "grad_norm": 1.9426069326928412, + "learning_rate": 8.204157143249997e-07, + "loss": 0.9681, + "step": 7870 + }, + { + "epoch": 0.709834513234432, + "grad_norm": 1.417194922447745, + "learning_rate": 8.199439823877668e-07, + "loss": 0.953, + "step": 7871 + }, + { + "epoch": 0.7099246967579024, + "grad_norm": 1.5212840191964083, + "learning_rate": 8.194723511394186e-07, + "loss": 0.9658, + "step": 7872 + }, + { + "epoch": 0.7100148802813726, + "grad_norm": 2.0159852644820813, + "learning_rate": 8.190008206202002e-07, + "loss": 0.9285, + "step": 7873 + }, + { + "epoch": 0.7101050638048428, + "grad_norm": 1.8059318183034596, + "learning_rate": 8.185293908703423e-07, + "loss": 1.0292, + "step": 7874 + }, + { + "epoch": 0.7101952473283131, + "grad_norm": 1.5876221377028048, + "learning_rate": 8.180580619300727e-07, + "loss": 0.9411, + "step": 7875 + }, + { + "epoch": 0.7102854308517834, + "grad_norm": 1.2782883723897185, + "learning_rate": 8.175868338396057e-07, + "loss": 0.9291, + "step": 7876 + }, + { + "epoch": 0.7103756143752537, + "grad_norm": 1.9612659316087495, + "learning_rate": 8.171157066391499e-07, + "loss": 0.948, + "step": 7877 + }, + { + "epoch": 0.7104657978987239, + "grad_norm": 1.7071347412582263, + "learning_rate": 8.166446803689045e-07, + "loss": 1.0028, + "step": 7878 + }, + { + "epoch": 0.7105559814221941, + "grad_norm": 0.6687245916727063, + "learning_rate": 8.161737550690595e-07, + "loss": 0.8203, + "step": 7879 + }, + { + "epoch": 0.7106461649456645, + "grad_norm": 1.5912900901952367, + "learning_rate": 8.157029307797976e-07, + "loss": 1.0032, + "step": 7880 + }, + { + "epoch": 0.7107363484691347, + "grad_norm": 1.373220809837404, + "learning_rate": 8.152322075412925e-07, + "loss": 1.0331, + "step": 7881 + }, + { + "epoch": 0.7108265319926049, + "grad_norm": 1.576073171935646, + "learning_rate": 8.147615853937073e-07, + "loss": 0.9279, + "step": 7882 + }, + { + "epoch": 0.7109167155160752, + "grad_norm": 1.594264601859678, + "learning_rate": 8.142910643771992e-07, + "loss": 0.9026, + "step": 7883 + }, + { + "epoch": 0.7110068990395455, + "grad_norm": 2.3969602726336507, + "learning_rate": 8.138206445319152e-07, + "loss": 0.9367, + "step": 7884 + }, + { + "epoch": 0.7110970825630157, + "grad_norm": 1.7560931125493424, + "learning_rate": 8.133503258979944e-07, + "loss": 0.8798, + "step": 7885 + }, + { + "epoch": 0.711187266086486, + "grad_norm": 1.6822145836081541, + "learning_rate": 8.12880108515567e-07, + "loss": 0.9023, + "step": 7886 + }, + { + "epoch": 0.7112774496099563, + "grad_norm": 1.4952519291934026, + "learning_rate": 8.124099924247543e-07, + "loss": 0.949, + "step": 7887 + }, + { + "epoch": 0.7113676331334265, + "grad_norm": 1.5270576696941263, + "learning_rate": 8.119399776656701e-07, + "loss": 0.8807, + "step": 7888 + }, + { + "epoch": 0.7114578166568968, + "grad_norm": 1.3469984577711038, + "learning_rate": 8.114700642784167e-07, + "loss": 0.9772, + "step": 7889 + }, + { + "epoch": 0.711548000180367, + "grad_norm": 2.169366169666726, + "learning_rate": 8.110002523030921e-07, + "loss": 1.0358, + "step": 7890 + }, + { + "epoch": 0.7116381837038374, + "grad_norm": 1.4832585201613457, + "learning_rate": 8.105305417797808e-07, + "loss": 0.9998, + "step": 7891 + }, + { + "epoch": 0.7117283672273076, + "grad_norm": 1.6170865087257642, + "learning_rate": 8.100609327485635e-07, + "loss": 1.0146, + "step": 7892 + }, + { + "epoch": 0.7118185507507778, + "grad_norm": 2.4545361778777557, + "learning_rate": 8.095914252495082e-07, + "loss": 1.0335, + "step": 7893 + }, + { + "epoch": 0.7119087342742481, + "grad_norm": 1.8868151022118442, + "learning_rate": 8.091220193226762e-07, + "loss": 0.9456, + "step": 7894 + }, + { + "epoch": 0.7119989177977184, + "grad_norm": 1.420285856460769, + "learning_rate": 8.0865271500812e-07, + "loss": 0.9646, + "step": 7895 + }, + { + "epoch": 0.7120891013211886, + "grad_norm": 1.56994366465774, + "learning_rate": 8.081835123458831e-07, + "loss": 1.016, + "step": 7896 + }, + { + "epoch": 0.7121792848446589, + "grad_norm": 2.7000980076839123, + "learning_rate": 8.077144113760013e-07, + "loss": 0.9644, + "step": 7897 + }, + { + "epoch": 0.7122694683681291, + "grad_norm": 6.090174168694135, + "learning_rate": 8.072454121384995e-07, + "loss": 0.9566, + "step": 7898 + }, + { + "epoch": 0.7123596518915994, + "grad_norm": 1.4259370901561568, + "learning_rate": 8.067765146733958e-07, + "loss": 1.0183, + "step": 7899 + }, + { + "epoch": 0.7124498354150697, + "grad_norm": 2.063484188269747, + "learning_rate": 8.063077190206993e-07, + "loss": 1.0551, + "step": 7900 + }, + { + "epoch": 0.7125400189385399, + "grad_norm": 0.6084241785081119, + "learning_rate": 8.058390252204101e-07, + "loss": 0.7309, + "step": 7901 + }, + { + "epoch": 0.7126302024620101, + "grad_norm": 1.644827265612901, + "learning_rate": 8.0537043331252e-07, + "loss": 0.9685, + "step": 7902 + }, + { + "epoch": 0.7127203859854805, + "grad_norm": 1.768623843597157, + "learning_rate": 8.049019433370121e-07, + "loss": 0.9137, + "step": 7903 + }, + { + "epoch": 0.7128105695089507, + "grad_norm": 1.4650076405713293, + "learning_rate": 8.044335553338588e-07, + "loss": 0.9792, + "step": 7904 + }, + { + "epoch": 0.712900753032421, + "grad_norm": 2.4477854159183994, + "learning_rate": 8.039652693430281e-07, + "loss": 0.9912, + "step": 7905 + }, + { + "epoch": 0.7129909365558912, + "grad_norm": 1.6229375809320463, + "learning_rate": 8.034970854044742e-07, + "loss": 0.9554, + "step": 7906 + }, + { + "epoch": 0.7130811200793615, + "grad_norm": 2.016375971038038, + "learning_rate": 8.03029003558148e-07, + "loss": 1.0741, + "step": 7907 + }, + { + "epoch": 0.7131713036028318, + "grad_norm": 2.003566066131526, + "learning_rate": 8.025610238439864e-07, + "loss": 1.0405, + "step": 7908 + }, + { + "epoch": 0.713261487126302, + "grad_norm": 1.718019989345908, + "learning_rate": 8.020931463019207e-07, + "loss": 1.0551, + "step": 7909 + }, + { + "epoch": 0.7133516706497723, + "grad_norm": 2.1523818550959968, + "learning_rate": 8.016253709718732e-07, + "loss": 0.9213, + "step": 7910 + }, + { + "epoch": 0.7134418541732426, + "grad_norm": 1.785798940014383, + "learning_rate": 8.011576978937567e-07, + "loss": 0.8834, + "step": 7911 + }, + { + "epoch": 0.7135320376967128, + "grad_norm": 1.9709751330838716, + "learning_rate": 8.006901271074764e-07, + "loss": 0.9749, + "step": 7912 + }, + { + "epoch": 0.713622221220183, + "grad_norm": 1.722241617763508, + "learning_rate": 8.002226586529261e-07, + "loss": 1.0154, + "step": 7913 + }, + { + "epoch": 0.7137124047436534, + "grad_norm": 1.436597208894559, + "learning_rate": 7.997552925699956e-07, + "loss": 1.0568, + "step": 7914 + }, + { + "epoch": 0.7138025882671236, + "grad_norm": 1.6653600657394094, + "learning_rate": 7.992880288985606e-07, + "loss": 1.0738, + "step": 7915 + }, + { + "epoch": 0.7138927717905938, + "grad_norm": 1.7906877889492812, + "learning_rate": 7.988208676784918e-07, + "loss": 0.985, + "step": 7916 + }, + { + "epoch": 0.7139829553140641, + "grad_norm": 1.7442788200692956, + "learning_rate": 7.983538089496497e-07, + "loss": 0.9755, + "step": 7917 + }, + { + "epoch": 0.7140731388375344, + "grad_norm": 2.1622515100144604, + "learning_rate": 7.978868527518864e-07, + "loss": 0.9402, + "step": 7918 + }, + { + "epoch": 0.7141633223610047, + "grad_norm": 1.5040700055427894, + "learning_rate": 7.974199991250455e-07, + "loss": 0.9565, + "step": 7919 + }, + { + "epoch": 0.7142535058844749, + "grad_norm": 1.6346613441668967, + "learning_rate": 7.969532481089616e-07, + "loss": 1.076, + "step": 7920 + }, + { + "epoch": 0.7143436894079451, + "grad_norm": 1.7404011690330918, + "learning_rate": 7.964865997434589e-07, + "loss": 0.9933, + "step": 7921 + }, + { + "epoch": 0.7144338729314155, + "grad_norm": 1.7764559697912052, + "learning_rate": 7.96020054068357e-07, + "loss": 1.0514, + "step": 7922 + }, + { + "epoch": 0.7145240564548857, + "grad_norm": 1.3059377561752896, + "learning_rate": 7.95553611123462e-07, + "loss": 0.9337, + "step": 7923 + }, + { + "epoch": 0.7146142399783559, + "grad_norm": 1.8810707361730696, + "learning_rate": 7.950872709485741e-07, + "loss": 1.1606, + "step": 7924 + }, + { + "epoch": 0.7147044235018262, + "grad_norm": 1.8827900865415794, + "learning_rate": 7.946210335834842e-07, + "loss": 1.0353, + "step": 7925 + }, + { + "epoch": 0.7147946070252965, + "grad_norm": 1.567093163462591, + "learning_rate": 7.94154899067974e-07, + "loss": 0.9038, + "step": 7926 + }, + { + "epoch": 0.7148847905487667, + "grad_norm": 1.9882975579512936, + "learning_rate": 7.936888674418177e-07, + "loss": 0.9043, + "step": 7927 + }, + { + "epoch": 0.714974974072237, + "grad_norm": 1.6172745538399376, + "learning_rate": 7.932229387447771e-07, + "loss": 0.9804, + "step": 7928 + }, + { + "epoch": 0.7150651575957072, + "grad_norm": 1.7928170200215785, + "learning_rate": 7.927571130166109e-07, + "loss": 0.9693, + "step": 7929 + }, + { + "epoch": 0.7151553411191776, + "grad_norm": 0.610500982686782, + "learning_rate": 7.922913902970632e-07, + "loss": 0.805, + "step": 7930 + }, + { + "epoch": 0.7152455246426478, + "grad_norm": 1.6046690051118944, + "learning_rate": 7.918257706258744e-07, + "loss": 0.9685, + "step": 7931 + }, + { + "epoch": 0.715335708166118, + "grad_norm": 1.66550923530265, + "learning_rate": 7.913602540427724e-07, + "loss": 0.9798, + "step": 7932 + }, + { + "epoch": 0.7154258916895884, + "grad_norm": 1.732641367463518, + "learning_rate": 7.908948405874775e-07, + "loss": 0.9739, + "step": 7933 + }, + { + "epoch": 0.7155160752130586, + "grad_norm": 2.1741688642080113, + "learning_rate": 7.904295302997019e-07, + "loss": 0.9037, + "step": 7934 + }, + { + "epoch": 0.7156062587365288, + "grad_norm": 1.698312873093159, + "learning_rate": 7.899643232191484e-07, + "loss": 0.9908, + "step": 7935 + }, + { + "epoch": 0.7156964422599991, + "grad_norm": 1.4183157504135298, + "learning_rate": 7.894992193855108e-07, + "loss": 0.9615, + "step": 7936 + }, + { + "epoch": 0.7157866257834694, + "grad_norm": 1.5708281775810702, + "learning_rate": 7.890342188384751e-07, + "loss": 0.9275, + "step": 7937 + }, + { + "epoch": 0.7158768093069396, + "grad_norm": 1.8305723974211774, + "learning_rate": 7.885693216177165e-07, + "loss": 0.9199, + "step": 7938 + }, + { + "epoch": 0.7159669928304099, + "grad_norm": 4.216413514952862, + "learning_rate": 7.88104527762903e-07, + "loss": 0.9702, + "step": 7939 + }, + { + "epoch": 0.7160571763538801, + "grad_norm": 1.9126879057482253, + "learning_rate": 7.876398373136936e-07, + "loss": 0.9358, + "step": 7940 + }, + { + "epoch": 0.7161473598773505, + "grad_norm": 1.4413211456234059, + "learning_rate": 7.87175250309738e-07, + "loss": 0.9736, + "step": 7941 + }, + { + "epoch": 0.7162375434008207, + "grad_norm": 1.9324337340685611, + "learning_rate": 7.867107667906785e-07, + "loss": 1.0522, + "step": 7942 + }, + { + "epoch": 0.7163277269242909, + "grad_norm": 1.578858007027155, + "learning_rate": 7.862463867961446e-07, + "loss": 0.9759, + "step": 7943 + }, + { + "epoch": 0.7164179104477612, + "grad_norm": 1.4708334572479513, + "learning_rate": 7.857821103657632e-07, + "loss": 1.0433, + "step": 7944 + }, + { + "epoch": 0.7165080939712315, + "grad_norm": 1.4092229570815997, + "learning_rate": 7.853179375391459e-07, + "loss": 1.0066, + "step": 7945 + }, + { + "epoch": 0.7165982774947017, + "grad_norm": 1.7395083778664249, + "learning_rate": 7.848538683559012e-07, + "loss": 0.9679, + "step": 7946 + }, + { + "epoch": 0.716688461018172, + "grad_norm": 1.542885152749217, + "learning_rate": 7.843899028556238e-07, + "loss": 1.0163, + "step": 7947 + }, + { + "epoch": 0.7167786445416422, + "grad_norm": 2.157580642342355, + "learning_rate": 7.839260410779029e-07, + "loss": 0.9713, + "step": 7948 + }, + { + "epoch": 0.7168688280651125, + "grad_norm": 1.4653228383219217, + "learning_rate": 7.834622830623175e-07, + "loss": 0.9991, + "step": 7949 + }, + { + "epoch": 0.7169590115885828, + "grad_norm": 5.823685578798286, + "learning_rate": 7.82998628848438e-07, + "loss": 0.9483, + "step": 7950 + }, + { + "epoch": 0.717049195112053, + "grad_norm": 1.6416760665443715, + "learning_rate": 7.825350784758261e-07, + "loss": 0.9647, + "step": 7951 + }, + { + "epoch": 0.7171393786355232, + "grad_norm": 1.4421359233788034, + "learning_rate": 7.820716319840342e-07, + "loss": 0.9426, + "step": 7952 + }, + { + "epoch": 0.7172295621589936, + "grad_norm": 1.814492084589975, + "learning_rate": 7.816082894126069e-07, + "loss": 1.0565, + "step": 7953 + }, + { + "epoch": 0.7173197456824638, + "grad_norm": 1.5571929066790184, + "learning_rate": 7.811450508010778e-07, + "loss": 0.9448, + "step": 7954 + }, + { + "epoch": 0.717409929205934, + "grad_norm": 1.412065011959284, + "learning_rate": 7.806819161889737e-07, + "loss": 0.9525, + "step": 7955 + }, + { + "epoch": 0.7175001127294043, + "grad_norm": 1.7087644250200884, + "learning_rate": 7.802188856158119e-07, + "loss": 1.008, + "step": 7956 + }, + { + "epoch": 0.7175902962528746, + "grad_norm": 1.6663051769510184, + "learning_rate": 7.797559591211002e-07, + "loss": 0.9659, + "step": 7957 + }, + { + "epoch": 0.7176804797763449, + "grad_norm": 1.7689749072609597, + "learning_rate": 7.79293136744339e-07, + "loss": 0.9407, + "step": 7958 + }, + { + "epoch": 0.7177706632998151, + "grad_norm": 1.6264951136809114, + "learning_rate": 7.788304185250185e-07, + "loss": 0.9785, + "step": 7959 + }, + { + "epoch": 0.7178608468232854, + "grad_norm": 1.4940542956575584, + "learning_rate": 7.78367804502619e-07, + "loss": 1.0044, + "step": 7960 + }, + { + "epoch": 0.7179510303467557, + "grad_norm": 1.7729694572266452, + "learning_rate": 7.779052947166156e-07, + "loss": 0.988, + "step": 7961 + }, + { + "epoch": 0.7180412138702259, + "grad_norm": 1.7592999755691112, + "learning_rate": 7.774428892064697e-07, + "loss": 0.9285, + "step": 7962 + }, + { + "epoch": 0.7181313973936961, + "grad_norm": 2.3596827830470954, + "learning_rate": 7.769805880116391e-07, + "loss": 0.9651, + "step": 7963 + }, + { + "epoch": 0.7182215809171665, + "grad_norm": 2.3606622481191213, + "learning_rate": 7.765183911715678e-07, + "loss": 0.9533, + "step": 7964 + }, + { + "epoch": 0.7183117644406367, + "grad_norm": 1.5535533352755078, + "learning_rate": 7.760562987256933e-07, + "loss": 0.8785, + "step": 7965 + }, + { + "epoch": 0.718401947964107, + "grad_norm": 1.6222829845812252, + "learning_rate": 7.755943107134444e-07, + "loss": 0.9829, + "step": 7966 + }, + { + "epoch": 0.7184921314875772, + "grad_norm": 0.6282240205972622, + "learning_rate": 7.751324271742401e-07, + "loss": 0.8056, + "step": 7967 + }, + { + "epoch": 0.7185823150110475, + "grad_norm": 1.4709253637179087, + "learning_rate": 7.746706481474916e-07, + "loss": 0.9934, + "step": 7968 + }, + { + "epoch": 0.7186724985345178, + "grad_norm": 1.1324832306182422, + "learning_rate": 7.742089736725992e-07, + "loss": 0.7739, + "step": 7969 + }, + { + "epoch": 0.718762682057988, + "grad_norm": 1.7604802889491038, + "learning_rate": 7.737474037889559e-07, + "loss": 0.8949, + "step": 7970 + }, + { + "epoch": 0.7188528655814582, + "grad_norm": 1.5249842689828448, + "learning_rate": 7.732859385359458e-07, + "loss": 0.9462, + "step": 7971 + }, + { + "epoch": 0.7189430491049286, + "grad_norm": 1.7124391698127084, + "learning_rate": 7.728245779529434e-07, + "loss": 0.8937, + "step": 7972 + }, + { + "epoch": 0.7190332326283988, + "grad_norm": 1.9464837037723983, + "learning_rate": 7.723633220793146e-07, + "loss": 1.0095, + "step": 7973 + }, + { + "epoch": 0.719123416151869, + "grad_norm": 2.158464925085582, + "learning_rate": 7.719021709544162e-07, + "loss": 0.866, + "step": 7974 + }, + { + "epoch": 0.7192135996753393, + "grad_norm": 1.7724990714518667, + "learning_rate": 7.714411246175964e-07, + "loss": 1.0679, + "step": 7975 + }, + { + "epoch": 0.7193037831988096, + "grad_norm": 2.1747880240437056, + "learning_rate": 7.709801831081946e-07, + "loss": 0.8957, + "step": 7976 + }, + { + "epoch": 0.7193939667222798, + "grad_norm": 1.714176682569846, + "learning_rate": 7.705193464655391e-07, + "loss": 0.934, + "step": 7977 + }, + { + "epoch": 0.7194841502457501, + "grad_norm": 1.4652672727588938, + "learning_rate": 7.700586147289534e-07, + "loss": 0.9509, + "step": 7978 + }, + { + "epoch": 0.7195743337692203, + "grad_norm": 2.1032684095894894, + "learning_rate": 7.695979879377481e-07, + "loss": 0.9341, + "step": 7979 + }, + { + "epoch": 0.7196645172926907, + "grad_norm": 1.4736594266617997, + "learning_rate": 7.691374661312266e-07, + "loss": 0.9966, + "step": 7980 + }, + { + "epoch": 0.7197547008161609, + "grad_norm": 1.413164854976589, + "learning_rate": 7.686770493486834e-07, + "loss": 0.9503, + "step": 7981 + }, + { + "epoch": 0.7198448843396311, + "grad_norm": 1.5083709631312858, + "learning_rate": 7.68216737629404e-07, + "loss": 0.921, + "step": 7982 + }, + { + "epoch": 0.7199350678631015, + "grad_norm": 1.4633177272569826, + "learning_rate": 7.67756531012665e-07, + "loss": 1.0472, + "step": 7983 + }, + { + "epoch": 0.7200252513865717, + "grad_norm": 2.2018413727158066, + "learning_rate": 7.67296429537732e-07, + "loss": 1.0335, + "step": 7984 + }, + { + "epoch": 0.7201154349100419, + "grad_norm": 1.4958471189552571, + "learning_rate": 7.668364332438661e-07, + "loss": 1.04, + "step": 7985 + }, + { + "epoch": 0.7202056184335122, + "grad_norm": 1.672451027009204, + "learning_rate": 7.663765421703145e-07, + "loss": 0.9869, + "step": 7986 + }, + { + "epoch": 0.7202958019569825, + "grad_norm": 1.5147679845583106, + "learning_rate": 7.659167563563187e-07, + "loss": 0.912, + "step": 7987 + }, + { + "epoch": 0.7203859854804527, + "grad_norm": 1.492055437725085, + "learning_rate": 7.654570758411096e-07, + "loss": 0.9998, + "step": 7988 + }, + { + "epoch": 0.720476169003923, + "grad_norm": 1.5793880280658177, + "learning_rate": 7.649975006639103e-07, + "loss": 1.0359, + "step": 7989 + }, + { + "epoch": 0.7205663525273932, + "grad_norm": 0.6216332353836111, + "learning_rate": 7.645380308639337e-07, + "loss": 0.7882, + "step": 7990 + }, + { + "epoch": 0.7206565360508635, + "grad_norm": 1.4476899259945728, + "learning_rate": 7.640786664803853e-07, + "loss": 1.0216, + "step": 7991 + }, + { + "epoch": 0.7207467195743338, + "grad_norm": 1.4152562282137475, + "learning_rate": 7.636194075524587e-07, + "loss": 0.8979, + "step": 7992 + }, + { + "epoch": 0.720836903097804, + "grad_norm": 1.4175929052192588, + "learning_rate": 7.631602541193429e-07, + "loss": 0.9267, + "step": 7993 + }, + { + "epoch": 0.7209270866212742, + "grad_norm": 1.916366809387798, + "learning_rate": 7.627012062202132e-07, + "loss": 1.0326, + "step": 7994 + }, + { + "epoch": 0.7210172701447446, + "grad_norm": 1.6832881216392102, + "learning_rate": 7.622422638942391e-07, + "loss": 0.9755, + "step": 7995 + }, + { + "epoch": 0.7211074536682148, + "grad_norm": 1.4591332376261132, + "learning_rate": 7.617834271805801e-07, + "loss": 1.0414, + "step": 7996 + }, + { + "epoch": 0.7211976371916851, + "grad_norm": 0.6470250522341563, + "learning_rate": 7.613246961183863e-07, + "loss": 0.8023, + "step": 7997 + }, + { + "epoch": 0.7212878207151553, + "grad_norm": 1.5662557096750198, + "learning_rate": 7.608660707468002e-07, + "loss": 0.9311, + "step": 7998 + }, + { + "epoch": 0.7213780042386256, + "grad_norm": 1.6430388359035313, + "learning_rate": 7.604075511049522e-07, + "loss": 0.869, + "step": 7999 + }, + { + "epoch": 0.7214681877620959, + "grad_norm": 1.593259548942171, + "learning_rate": 7.599491372319682e-07, + "loss": 0.9218, + "step": 8000 + }, + { + "epoch": 0.7215583712855661, + "grad_norm": 1.3421724063259972, + "learning_rate": 7.594908291669601e-07, + "loss": 0.9741, + "step": 8001 + }, + { + "epoch": 0.7216485548090363, + "grad_norm": 1.7455880544367381, + "learning_rate": 7.590326269490359e-07, + "loss": 1.0103, + "step": 8002 + }, + { + "epoch": 0.7217387383325067, + "grad_norm": 1.4320166953887767, + "learning_rate": 7.585745306172899e-07, + "loss": 0.9322, + "step": 8003 + }, + { + "epoch": 0.7218289218559769, + "grad_norm": 1.584814942805261, + "learning_rate": 7.5811654021081e-07, + "loss": 0.9539, + "step": 8004 + }, + { + "epoch": 0.7219191053794471, + "grad_norm": 1.384756327653027, + "learning_rate": 7.576586557686748e-07, + "loss": 1.0054, + "step": 8005 + }, + { + "epoch": 0.7220092889029175, + "grad_norm": 2.029305921576623, + "learning_rate": 7.572008773299531e-07, + "loss": 0.9465, + "step": 8006 + }, + { + "epoch": 0.7220994724263877, + "grad_norm": 1.663084647267189, + "learning_rate": 7.567432049337055e-07, + "loss": 1.0119, + "step": 8007 + }, + { + "epoch": 0.722189655949858, + "grad_norm": 1.7322982864421617, + "learning_rate": 7.562856386189834e-07, + "loss": 0.9123, + "step": 8008 + }, + { + "epoch": 0.7222798394733282, + "grad_norm": 2.0284023571364047, + "learning_rate": 7.558281784248275e-07, + "loss": 0.9868, + "step": 8009 + }, + { + "epoch": 0.7223700229967985, + "grad_norm": 2.334086932239452, + "learning_rate": 7.553708243902721e-07, + "loss": 0.9638, + "step": 8010 + }, + { + "epoch": 0.7224602065202688, + "grad_norm": 1.5796044845727981, + "learning_rate": 7.549135765543404e-07, + "loss": 0.8659, + "step": 8011 + }, + { + "epoch": 0.722550390043739, + "grad_norm": 1.7181873440940332, + "learning_rate": 7.544564349560481e-07, + "loss": 0.9977, + "step": 8012 + }, + { + "epoch": 0.7226405735672092, + "grad_norm": 1.8018560615362014, + "learning_rate": 7.539993996344009e-07, + "loss": 0.9609, + "step": 8013 + }, + { + "epoch": 0.7227307570906796, + "grad_norm": 2.593752114168707, + "learning_rate": 7.535424706283941e-07, + "loss": 0.9518, + "step": 8014 + }, + { + "epoch": 0.7228209406141498, + "grad_norm": 2.0407688995611473, + "learning_rate": 7.530856479770181e-07, + "loss": 0.9383, + "step": 8015 + }, + { + "epoch": 0.72291112413762, + "grad_norm": 1.6533604012601362, + "learning_rate": 7.526289317192484e-07, + "loss": 1.0561, + "step": 8016 + }, + { + "epoch": 0.7230013076610903, + "grad_norm": 1.7152593365431685, + "learning_rate": 7.521723218940579e-07, + "loss": 0.9741, + "step": 8017 + }, + { + "epoch": 0.7230914911845606, + "grad_norm": 1.7293596351528282, + "learning_rate": 7.517158185404038e-07, + "loss": 1.0372, + "step": 8018 + }, + { + "epoch": 0.7231816747080309, + "grad_norm": 1.636603376378121, + "learning_rate": 7.512594216972403e-07, + "loss": 0.9987, + "step": 8019 + }, + { + "epoch": 0.7232718582315011, + "grad_norm": 1.4142735782119102, + "learning_rate": 7.508031314035078e-07, + "loss": 1.0085, + "step": 8020 + }, + { + "epoch": 0.7233620417549713, + "grad_norm": 2.2084524254505724, + "learning_rate": 7.503469476981401e-07, + "loss": 0.9968, + "step": 8021 + }, + { + "epoch": 0.7234522252784417, + "grad_norm": 2.2244791557105392, + "learning_rate": 7.498908706200613e-07, + "loss": 0.8477, + "step": 8022 + }, + { + "epoch": 0.7235424088019119, + "grad_norm": 1.4430233534564176, + "learning_rate": 7.494349002081866e-07, + "loss": 1.0224, + "step": 8023 + }, + { + "epoch": 0.7236325923253821, + "grad_norm": 1.5053139971788407, + "learning_rate": 7.489790365014224e-07, + "loss": 1.0054, + "step": 8024 + }, + { + "epoch": 0.7237227758488524, + "grad_norm": 1.773605607589405, + "learning_rate": 7.485232795386642e-07, + "loss": 0.9938, + "step": 8025 + }, + { + "epoch": 0.7238129593723227, + "grad_norm": 1.5631104611922184, + "learning_rate": 7.480676293588002e-07, + "loss": 1.0001, + "step": 8026 + }, + { + "epoch": 0.7239031428957929, + "grad_norm": 0.7985908883524365, + "learning_rate": 7.476120860007093e-07, + "loss": 0.8339, + "step": 8027 + }, + { + "epoch": 0.7239933264192632, + "grad_norm": 1.8047731057212608, + "learning_rate": 7.471566495032608e-07, + "loss": 1.0171, + "step": 8028 + }, + { + "epoch": 0.7240835099427335, + "grad_norm": 1.5980336832356623, + "learning_rate": 7.467013199053152e-07, + "loss": 0.9127, + "step": 8029 + }, + { + "epoch": 0.7241736934662037, + "grad_norm": 1.7024364654516586, + "learning_rate": 7.46246097245724e-07, + "loss": 0.9727, + "step": 8030 + }, + { + "epoch": 0.724263876989674, + "grad_norm": 1.895880464519926, + "learning_rate": 7.457909815633276e-07, + "loss": 0.9866, + "step": 8031 + }, + { + "epoch": 0.7243540605131442, + "grad_norm": 1.9416799678537906, + "learning_rate": 7.453359728969618e-07, + "loss": 0.9665, + "step": 8032 + }, + { + "epoch": 0.7244442440366146, + "grad_norm": 1.509897954157901, + "learning_rate": 7.448810712854475e-07, + "loss": 0.9008, + "step": 8033 + }, + { + "epoch": 0.7245344275600848, + "grad_norm": 1.7323088840249106, + "learning_rate": 7.444262767676022e-07, + "loss": 1.0165, + "step": 8034 + }, + { + "epoch": 0.724624611083555, + "grad_norm": 2.1965947326474744, + "learning_rate": 7.439715893822296e-07, + "loss": 1.0496, + "step": 8035 + }, + { + "epoch": 0.7247147946070253, + "grad_norm": 1.3592101249241373, + "learning_rate": 7.435170091681264e-07, + "loss": 0.9966, + "step": 8036 + }, + { + "epoch": 0.7248049781304956, + "grad_norm": 10.589062759379146, + "learning_rate": 7.430625361640803e-07, + "loss": 0.9319, + "step": 8037 + }, + { + "epoch": 0.7248951616539658, + "grad_norm": 1.6596915997278758, + "learning_rate": 7.426081704088694e-07, + "loss": 1.0196, + "step": 8038 + }, + { + "epoch": 0.7249853451774361, + "grad_norm": 1.8299764403405685, + "learning_rate": 7.42153911941263e-07, + "loss": 1.0488, + "step": 8039 + }, + { + "epoch": 0.7250755287009063, + "grad_norm": 1.7979866238182423, + "learning_rate": 7.416997608000192e-07, + "loss": 1.0171, + "step": 8040 + }, + { + "epoch": 0.7251657122243766, + "grad_norm": 1.7623146554875746, + "learning_rate": 7.412457170238918e-07, + "loss": 0.9563, + "step": 8041 + }, + { + "epoch": 0.7252558957478469, + "grad_norm": 1.5976897518310866, + "learning_rate": 7.407917806516193e-07, + "loss": 0.9187, + "step": 8042 + }, + { + "epoch": 0.7253460792713171, + "grad_norm": 1.9642807019157367, + "learning_rate": 7.403379517219354e-07, + "loss": 0.9696, + "step": 8043 + }, + { + "epoch": 0.7254362627947873, + "grad_norm": 1.781124980621584, + "learning_rate": 7.398842302735636e-07, + "loss": 1.0057, + "step": 8044 + }, + { + "epoch": 0.7255264463182577, + "grad_norm": 0.6669078455221693, + "learning_rate": 7.394306163452171e-07, + "loss": 0.783, + "step": 8045 + }, + { + "epoch": 0.7256166298417279, + "grad_norm": 1.616452552276053, + "learning_rate": 7.38977109975601e-07, + "loss": 0.9526, + "step": 8046 + }, + { + "epoch": 0.7257068133651982, + "grad_norm": 2.7289981838568336, + "learning_rate": 7.385237112034119e-07, + "loss": 0.8905, + "step": 8047 + }, + { + "epoch": 0.7257969968886684, + "grad_norm": 1.6809269080579186, + "learning_rate": 7.380704200673342e-07, + "loss": 1.0408, + "step": 8048 + }, + { + "epoch": 0.7258871804121387, + "grad_norm": 1.4002282808019355, + "learning_rate": 7.376172366060478e-07, + "loss": 0.9996, + "step": 8049 + }, + { + "epoch": 0.725977363935609, + "grad_norm": 0.651663785604497, + "learning_rate": 7.371641608582187e-07, + "loss": 0.7492, + "step": 8050 + }, + { + "epoch": 0.7260675474590792, + "grad_norm": 4.008454210093829, + "learning_rate": 7.367111928625067e-07, + "loss": 0.8442, + "step": 8051 + }, + { + "epoch": 0.7261577309825495, + "grad_norm": 1.7487745081266017, + "learning_rate": 7.362583326575613e-07, + "loss": 0.8915, + "step": 8052 + }, + { + "epoch": 0.7262479145060198, + "grad_norm": 1.6925065688170582, + "learning_rate": 7.358055802820234e-07, + "loss": 0.9579, + "step": 8053 + }, + { + "epoch": 0.72633809802949, + "grad_norm": 2.0458152298861165, + "learning_rate": 7.353529357745245e-07, + "loss": 0.9479, + "step": 8054 + }, + { + "epoch": 0.7264282815529602, + "grad_norm": 1.8860721467378163, + "learning_rate": 7.349003991736851e-07, + "loss": 1.0139, + "step": 8055 + }, + { + "epoch": 0.7265184650764306, + "grad_norm": 1.9340741418120457, + "learning_rate": 7.344479705181206e-07, + "loss": 0.8241, + "step": 8056 + }, + { + "epoch": 0.7266086485999008, + "grad_norm": 1.4943606544996266, + "learning_rate": 7.339956498464322e-07, + "loss": 0.9283, + "step": 8057 + }, + { + "epoch": 0.726698832123371, + "grad_norm": 1.486816887315193, + "learning_rate": 7.335434371972169e-07, + "loss": 1.0236, + "step": 8058 + }, + { + "epoch": 0.7267890156468413, + "grad_norm": 1.6366647788824833, + "learning_rate": 7.33091332609058e-07, + "loss": 0.9622, + "step": 8059 + }, + { + "epoch": 0.7268791991703116, + "grad_norm": 1.6438389134020164, + "learning_rate": 7.326393361205323e-07, + "loss": 0.9914, + "step": 8060 + }, + { + "epoch": 0.7269693826937819, + "grad_norm": 1.6237022646611499, + "learning_rate": 7.321874477702068e-07, + "loss": 1.0551, + "step": 8061 + }, + { + "epoch": 0.7270595662172521, + "grad_norm": 0.7568977970561456, + "learning_rate": 7.317356675966386e-07, + "loss": 0.8285, + "step": 8062 + }, + { + "epoch": 0.7271497497407223, + "grad_norm": 1.545640857706195, + "learning_rate": 7.312839956383765e-07, + "loss": 1.011, + "step": 8063 + }, + { + "epoch": 0.7272399332641927, + "grad_norm": 1.4645782637022282, + "learning_rate": 7.308324319339603e-07, + "loss": 0.9555, + "step": 8064 + }, + { + "epoch": 0.7273301167876629, + "grad_norm": 1.5774796176566956, + "learning_rate": 7.303809765219182e-07, + "loss": 0.9785, + "step": 8065 + }, + { + "epoch": 0.7274203003111331, + "grad_norm": 1.6785664210491216, + "learning_rate": 7.299296294407719e-07, + "loss": 1.0223, + "step": 8066 + }, + { + "epoch": 0.7275104838346034, + "grad_norm": 1.5801608862026646, + "learning_rate": 7.294783907290327e-07, + "loss": 1.0152, + "step": 8067 + }, + { + "epoch": 0.7276006673580737, + "grad_norm": 1.568857686271626, + "learning_rate": 7.290272604252028e-07, + "loss": 0.8426, + "step": 8068 + }, + { + "epoch": 0.727690850881544, + "grad_norm": 2.341158438632002, + "learning_rate": 7.285762385677758e-07, + "loss": 0.877, + "step": 8069 + }, + { + "epoch": 0.7277810344050142, + "grad_norm": 1.5231110051779675, + "learning_rate": 7.281253251952335e-07, + "loss": 1.012, + "step": 8070 + }, + { + "epoch": 0.7278712179284844, + "grad_norm": 1.5048950907746597, + "learning_rate": 7.276745203460526e-07, + "loss": 0.9571, + "step": 8071 + }, + { + "epoch": 0.7279614014519548, + "grad_norm": 0.7218644624411619, + "learning_rate": 7.272238240586959e-07, + "loss": 0.8691, + "step": 8072 + }, + { + "epoch": 0.728051584975425, + "grad_norm": 1.3250781396101488, + "learning_rate": 7.267732363716219e-07, + "loss": 0.9102, + "step": 8073 + }, + { + "epoch": 0.7281417684988952, + "grad_norm": 1.4988044106578653, + "learning_rate": 7.263227573232753e-07, + "loss": 0.9311, + "step": 8074 + }, + { + "epoch": 0.7282319520223655, + "grad_norm": 1.506906187604174, + "learning_rate": 7.258723869520937e-07, + "loss": 1.0658, + "step": 8075 + }, + { + "epoch": 0.7283221355458358, + "grad_norm": 1.9506009691903852, + "learning_rate": 7.254221252965059e-07, + "loss": 1.0206, + "step": 8076 + }, + { + "epoch": 0.728412319069306, + "grad_norm": 1.8147872436502237, + "learning_rate": 7.249719723949301e-07, + "loss": 0.8901, + "step": 8077 + }, + { + "epoch": 0.7285025025927763, + "grad_norm": 1.515804634067223, + "learning_rate": 7.245219282857761e-07, + "loss": 0.9973, + "step": 8078 + }, + { + "epoch": 0.7285926861162466, + "grad_norm": 2.023422771385524, + "learning_rate": 7.240719930074442e-07, + "loss": 1.0146, + "step": 8079 + }, + { + "epoch": 0.7286828696397168, + "grad_norm": 1.6236873606991555, + "learning_rate": 7.236221665983257e-07, + "loss": 0.9985, + "step": 8080 + }, + { + "epoch": 0.7287730531631871, + "grad_norm": 0.7321848981862358, + "learning_rate": 7.231724490968012e-07, + "loss": 0.7943, + "step": 8081 + }, + { + "epoch": 0.7288632366866573, + "grad_norm": 1.600462447331584, + "learning_rate": 7.227228405412438e-07, + "loss": 0.9457, + "step": 8082 + }, + { + "epoch": 0.7289534202101277, + "grad_norm": 1.7866835456852692, + "learning_rate": 7.222733409700165e-07, + "loss": 0.9374, + "step": 8083 + }, + { + "epoch": 0.7290436037335979, + "grad_norm": 1.4465387040045723, + "learning_rate": 7.21823950421473e-07, + "loss": 0.9868, + "step": 8084 + }, + { + "epoch": 0.7291337872570681, + "grad_norm": 1.7936126284216836, + "learning_rate": 7.213746689339577e-07, + "loss": 0.9458, + "step": 8085 + }, + { + "epoch": 0.7292239707805384, + "grad_norm": 1.890261780709057, + "learning_rate": 7.20925496545807e-07, + "loss": 0.9736, + "step": 8086 + }, + { + "epoch": 0.7293141543040087, + "grad_norm": 1.4913068639332923, + "learning_rate": 7.20476433295344e-07, + "loss": 0.8849, + "step": 8087 + }, + { + "epoch": 0.7294043378274789, + "grad_norm": 1.9393405785611588, + "learning_rate": 7.200274792208882e-07, + "loss": 1.0165, + "step": 8088 + }, + { + "epoch": 0.7294945213509492, + "grad_norm": 1.6876722883647846, + "learning_rate": 7.195786343607444e-07, + "loss": 0.8681, + "step": 8089 + }, + { + "epoch": 0.7295847048744194, + "grad_norm": 1.644144986926977, + "learning_rate": 7.191298987532131e-07, + "loss": 0.9636, + "step": 8090 + }, + { + "epoch": 0.7296748883978897, + "grad_norm": 1.5000770072243457, + "learning_rate": 7.186812724365805e-07, + "loss": 0.9806, + "step": 8091 + }, + { + "epoch": 0.72976507192136, + "grad_norm": 1.6741343410644058, + "learning_rate": 7.182327554491272e-07, + "loss": 0.9807, + "step": 8092 + }, + { + "epoch": 0.7298552554448302, + "grad_norm": 0.6820276467838781, + "learning_rate": 7.177843478291225e-07, + "loss": 0.8241, + "step": 8093 + }, + { + "epoch": 0.7299454389683004, + "grad_norm": 1.8481778140579705, + "learning_rate": 7.173360496148276e-07, + "loss": 1.0061, + "step": 8094 + }, + { + "epoch": 0.7300356224917708, + "grad_norm": 1.7924407647000489, + "learning_rate": 7.168878608444939e-07, + "loss": 0.9919, + "step": 8095 + }, + { + "epoch": 0.730125806015241, + "grad_norm": 1.8455381612374588, + "learning_rate": 7.164397815563623e-07, + "loss": 0.9808, + "step": 8096 + }, + { + "epoch": 0.7302159895387113, + "grad_norm": 10.723348537533502, + "learning_rate": 7.159918117886661e-07, + "loss": 0.9859, + "step": 8097 + }, + { + "epoch": 0.7303061730621815, + "grad_norm": 1.7314559153034779, + "learning_rate": 7.155439515796284e-07, + "loss": 0.9913, + "step": 8098 + }, + { + "epoch": 0.7303963565856518, + "grad_norm": 2.80287456095405, + "learning_rate": 7.150962009674633e-07, + "loss": 0.9711, + "step": 8099 + }, + { + "epoch": 0.7304865401091221, + "grad_norm": 1.437986291647605, + "learning_rate": 7.146485599903751e-07, + "loss": 1.0231, + "step": 8100 + }, + { + "epoch": 0.7305767236325923, + "grad_norm": 1.6071111812194758, + "learning_rate": 7.142010286865592e-07, + "loss": 1.0476, + "step": 8101 + }, + { + "epoch": 0.7306669071560626, + "grad_norm": 0.550091142038531, + "learning_rate": 7.137536070942012e-07, + "loss": 0.7853, + "step": 8102 + }, + { + "epoch": 0.7307570906795329, + "grad_norm": 1.629591909587742, + "learning_rate": 7.133062952514786e-07, + "loss": 0.9739, + "step": 8103 + }, + { + "epoch": 0.7308472742030031, + "grad_norm": 0.6500322196384342, + "learning_rate": 7.128590931965562e-07, + "loss": 0.8404, + "step": 8104 + }, + { + "epoch": 0.7309374577264733, + "grad_norm": 1.4058396906427268, + "learning_rate": 7.124120009675945e-07, + "loss": 1.0397, + "step": 8105 + }, + { + "epoch": 0.7310276412499437, + "grad_norm": 1.5434440556734716, + "learning_rate": 7.119650186027399e-07, + "loss": 1.005, + "step": 8106 + }, + { + "epoch": 0.7311178247734139, + "grad_norm": 1.6338951500505297, + "learning_rate": 7.11518146140132e-07, + "loss": 1.0954, + "step": 8107 + }, + { + "epoch": 0.7312080082968841, + "grad_norm": 1.8028228506610067, + "learning_rate": 7.110713836179007e-07, + "loss": 0.9377, + "step": 8108 + }, + { + "epoch": 0.7312981918203544, + "grad_norm": 1.4627471160782006, + "learning_rate": 7.106247310741659e-07, + "loss": 0.9045, + "step": 8109 + }, + { + "epoch": 0.7313883753438247, + "grad_norm": 2.48751804977846, + "learning_rate": 7.101781885470393e-07, + "loss": 1.0427, + "step": 8110 + }, + { + "epoch": 0.731478558867295, + "grad_norm": 1.6910339368576397, + "learning_rate": 7.097317560746203e-07, + "loss": 1.0119, + "step": 8111 + }, + { + "epoch": 0.7315687423907652, + "grad_norm": 1.9822593885457531, + "learning_rate": 7.092854336950036e-07, + "loss": 1.0385, + "step": 8112 + }, + { + "epoch": 0.7316589259142354, + "grad_norm": 1.4409266555301945, + "learning_rate": 7.0883922144627e-07, + "loss": 1.0312, + "step": 8113 + }, + { + "epoch": 0.7317491094377058, + "grad_norm": 1.6819498838880458, + "learning_rate": 7.083931193664934e-07, + "loss": 1.0168, + "step": 8114 + }, + { + "epoch": 0.731839292961176, + "grad_norm": 1.620529407487447, + "learning_rate": 7.079471274937378e-07, + "loss": 1.0229, + "step": 8115 + }, + { + "epoch": 0.7319294764846462, + "grad_norm": 1.5562886183516347, + "learning_rate": 7.075012458660574e-07, + "loss": 0.9627, + "step": 8116 + }, + { + "epoch": 0.7320196600081165, + "grad_norm": 0.7182511796314234, + "learning_rate": 7.070554745214976e-07, + "loss": 0.8125, + "step": 8117 + }, + { + "epoch": 0.7321098435315868, + "grad_norm": 1.8510006724211607, + "learning_rate": 7.066098134980947e-07, + "loss": 0.9356, + "step": 8118 + }, + { + "epoch": 0.732200027055057, + "grad_norm": 1.5127916102236734, + "learning_rate": 7.061642628338727e-07, + "loss": 0.9801, + "step": 8119 + }, + { + "epoch": 0.7322902105785273, + "grad_norm": 1.4480813387897866, + "learning_rate": 7.057188225668513e-07, + "loss": 0.976, + "step": 8120 + }, + { + "epoch": 0.7323803941019975, + "grad_norm": 1.660613878792194, + "learning_rate": 7.052734927350358e-07, + "loss": 1.1288, + "step": 8121 + }, + { + "epoch": 0.7324705776254679, + "grad_norm": 1.6175967818112154, + "learning_rate": 7.048282733764252e-07, + "loss": 1.0012, + "step": 8122 + }, + { + "epoch": 0.7325607611489381, + "grad_norm": 1.5807559300712182, + "learning_rate": 7.043831645290077e-07, + "loss": 1.0602, + "step": 8123 + }, + { + "epoch": 0.7326509446724083, + "grad_norm": 1.7193056075508397, + "learning_rate": 7.039381662307624e-07, + "loss": 0.9502, + "step": 8124 + }, + { + "epoch": 0.7327411281958787, + "grad_norm": 1.5285762877189706, + "learning_rate": 7.034932785196601e-07, + "loss": 0.9594, + "step": 8125 + }, + { + "epoch": 0.7328313117193489, + "grad_norm": 1.787974264010883, + "learning_rate": 7.030485014336585e-07, + "loss": 0.9619, + "step": 8126 + }, + { + "epoch": 0.7329214952428191, + "grad_norm": 1.6335375330559336, + "learning_rate": 7.026038350107118e-07, + "loss": 0.9399, + "step": 8127 + }, + { + "epoch": 0.7330116787662894, + "grad_norm": 1.5788123267549576, + "learning_rate": 7.021592792887579e-07, + "loss": 1.1078, + "step": 8128 + }, + { + "epoch": 0.7331018622897597, + "grad_norm": 1.687370507252188, + "learning_rate": 7.01714834305732e-07, + "loss": 1.0318, + "step": 8129 + }, + { + "epoch": 0.7331920458132299, + "grad_norm": 1.5106196387704418, + "learning_rate": 7.012705000995544e-07, + "loss": 0.9382, + "step": 8130 + }, + { + "epoch": 0.7332822293367002, + "grad_norm": 1.4035980650559332, + "learning_rate": 7.008262767081392e-07, + "loss": 0.9717, + "step": 8131 + }, + { + "epoch": 0.7333724128601704, + "grad_norm": 2.062985680934012, + "learning_rate": 7.003821641693892e-07, + "loss": 1.0345, + "step": 8132 + }, + { + "epoch": 0.7334625963836408, + "grad_norm": 1.915383455316206, + "learning_rate": 6.999381625211993e-07, + "loss": 0.9693, + "step": 8133 + }, + { + "epoch": 0.733552779907111, + "grad_norm": 1.8114529577975793, + "learning_rate": 6.994942718014536e-07, + "loss": 0.9002, + "step": 8134 + }, + { + "epoch": 0.7336429634305812, + "grad_norm": 1.816985007342311, + "learning_rate": 6.990504920480282e-07, + "loss": 0.8861, + "step": 8135 + }, + { + "epoch": 0.7337331469540515, + "grad_norm": 2.0975079190965187, + "learning_rate": 6.986068232987879e-07, + "loss": 0.9756, + "step": 8136 + }, + { + "epoch": 0.7338233304775218, + "grad_norm": 1.584368451046623, + "learning_rate": 6.981632655915888e-07, + "loss": 0.9316, + "step": 8137 + }, + { + "epoch": 0.733913514000992, + "grad_norm": 1.481888384812883, + "learning_rate": 6.977198189642783e-07, + "loss": 1.0411, + "step": 8138 + }, + { + "epoch": 0.7340036975244623, + "grad_norm": 1.8199514268325028, + "learning_rate": 6.972764834546935e-07, + "loss": 1.0055, + "step": 8139 + }, + { + "epoch": 0.7340938810479325, + "grad_norm": 1.598531096431898, + "learning_rate": 6.96833259100663e-07, + "loss": 0.9952, + "step": 8140 + }, + { + "epoch": 0.7341840645714028, + "grad_norm": 2.0756509150604794, + "learning_rate": 6.96390145940003e-07, + "loss": 0.954, + "step": 8141 + }, + { + "epoch": 0.7342742480948731, + "grad_norm": 1.5888348501674323, + "learning_rate": 6.959471440105253e-07, + "loss": 0.9313, + "step": 8142 + }, + { + "epoch": 0.7343644316183433, + "grad_norm": 2.498063100086577, + "learning_rate": 6.955042533500261e-07, + "loss": 1.0115, + "step": 8143 + }, + { + "epoch": 0.7344546151418135, + "grad_norm": 1.745952352990122, + "learning_rate": 6.950614739962986e-07, + "loss": 0.9344, + "step": 8144 + }, + { + "epoch": 0.7345447986652839, + "grad_norm": 1.5721710745299002, + "learning_rate": 6.946188059871198e-07, + "loss": 0.8692, + "step": 8145 + }, + { + "epoch": 0.7346349821887541, + "grad_norm": 1.5107636502596298, + "learning_rate": 6.941762493602638e-07, + "loss": 1.0976, + "step": 8146 + }, + { + "epoch": 0.7347251657122243, + "grad_norm": 1.4452651557385052, + "learning_rate": 6.937338041534899e-07, + "loss": 0.9212, + "step": 8147 + }, + { + "epoch": 0.7348153492356947, + "grad_norm": 1.448115008191709, + "learning_rate": 6.932914704045505e-07, + "loss": 0.98, + "step": 8148 + }, + { + "epoch": 0.7349055327591649, + "grad_norm": 1.615193930399781, + "learning_rate": 6.928492481511878e-07, + "loss": 1.0072, + "step": 8149 + }, + { + "epoch": 0.7349957162826352, + "grad_norm": 4.946289930127022, + "learning_rate": 6.924071374311349e-07, + "loss": 0.9251, + "step": 8150 + }, + { + "epoch": 0.7350858998061054, + "grad_norm": 0.5969259749609912, + "learning_rate": 6.919651382821157e-07, + "loss": 0.7957, + "step": 8151 + }, + { + "epoch": 0.7351760833295757, + "grad_norm": 1.672994149940601, + "learning_rate": 6.915232507418425e-07, + "loss": 0.9474, + "step": 8152 + }, + { + "epoch": 0.735266266853046, + "grad_norm": 1.690184118635871, + "learning_rate": 6.910814748480204e-07, + "loss": 0.8657, + "step": 8153 + }, + { + "epoch": 0.7353564503765162, + "grad_norm": 2.093324646516663, + "learning_rate": 6.906398106383445e-07, + "loss": 0.9929, + "step": 8154 + }, + { + "epoch": 0.7354466338999864, + "grad_norm": 1.6610703767752713, + "learning_rate": 6.901982581504994e-07, + "loss": 0.9758, + "step": 8155 + }, + { + "epoch": 0.7355368174234568, + "grad_norm": 2.2195806963044227, + "learning_rate": 6.897568174221611e-07, + "loss": 0.8511, + "step": 8156 + }, + { + "epoch": 0.735627000946927, + "grad_norm": 1.6131451343815464, + "learning_rate": 6.893154884909966e-07, + "loss": 1.051, + "step": 8157 + }, + { + "epoch": 0.7357171844703972, + "grad_norm": 0.6008192400088485, + "learning_rate": 6.888742713946602e-07, + "loss": 0.8208, + "step": 8158 + }, + { + "epoch": 0.7358073679938675, + "grad_norm": 1.796219814102648, + "learning_rate": 6.884331661708018e-07, + "loss": 0.9806, + "step": 8159 + }, + { + "epoch": 0.7358975515173378, + "grad_norm": 1.5079317194002968, + "learning_rate": 6.879921728570561e-07, + "loss": 0.9769, + "step": 8160 + }, + { + "epoch": 0.735987735040808, + "grad_norm": 2.121083970864025, + "learning_rate": 6.875512914910539e-07, + "loss": 1.057, + "step": 8161 + }, + { + "epoch": 0.7360779185642783, + "grad_norm": 1.573259387878312, + "learning_rate": 6.871105221104119e-07, + "loss": 0.9261, + "step": 8162 + }, + { + "epoch": 0.7361681020877485, + "grad_norm": 2.0889188912561507, + "learning_rate": 6.866698647527391e-07, + "loss": 0.9557, + "step": 8163 + }, + { + "epoch": 0.7362582856112189, + "grad_norm": 1.6544940204990732, + "learning_rate": 6.862293194556353e-07, + "loss": 1.0026, + "step": 8164 + }, + { + "epoch": 0.7363484691346891, + "grad_norm": 1.4649989338372944, + "learning_rate": 6.857888862566896e-07, + "loss": 0.9227, + "step": 8165 + }, + { + "epoch": 0.7364386526581593, + "grad_norm": 1.5243296822869072, + "learning_rate": 6.853485651934836e-07, + "loss": 0.943, + "step": 8166 + }, + { + "epoch": 0.7365288361816296, + "grad_norm": 2.0029776579991108, + "learning_rate": 6.849083563035855e-07, + "loss": 0.8325, + "step": 8167 + }, + { + "epoch": 0.7366190197050999, + "grad_norm": 0.7047489274819594, + "learning_rate": 6.844682596245592e-07, + "loss": 0.8975, + "step": 8168 + }, + { + "epoch": 0.7367092032285701, + "grad_norm": 1.6525326884581526, + "learning_rate": 6.840282751939539e-07, + "loss": 1.0395, + "step": 8169 + }, + { + "epoch": 0.7367993867520404, + "grad_norm": 1.4885430693925674, + "learning_rate": 6.835884030493126e-07, + "loss": 1.0215, + "step": 8170 + }, + { + "epoch": 0.7368895702755107, + "grad_norm": 2.5296683373765116, + "learning_rate": 6.831486432281672e-07, + "loss": 1.0023, + "step": 8171 + }, + { + "epoch": 0.736979753798981, + "grad_norm": 3.5970256969391974, + "learning_rate": 6.827089957680407e-07, + "loss": 0.9886, + "step": 8172 + }, + { + "epoch": 0.7370699373224512, + "grad_norm": 1.4879906237932354, + "learning_rate": 6.822694607064461e-07, + "loss": 1.091, + "step": 8173 + }, + { + "epoch": 0.7371601208459214, + "grad_norm": 1.7025760150805853, + "learning_rate": 6.818300380808877e-07, + "loss": 1.0012, + "step": 8174 + }, + { + "epoch": 0.7372503043693918, + "grad_norm": 1.5297069170365345, + "learning_rate": 6.813907279288574e-07, + "loss": 0.9706, + "step": 8175 + }, + { + "epoch": 0.737340487892862, + "grad_norm": 2.188675591790582, + "learning_rate": 6.809515302878422e-07, + "loss": 1.0033, + "step": 8176 + }, + { + "epoch": 0.7374306714163322, + "grad_norm": 1.9470776129497815, + "learning_rate": 6.80512445195315e-07, + "loss": 0.9676, + "step": 8177 + }, + { + "epoch": 0.7375208549398025, + "grad_norm": 1.6752625843162354, + "learning_rate": 6.800734726887416e-07, + "loss": 0.9983, + "step": 8178 + }, + { + "epoch": 0.7376110384632728, + "grad_norm": 1.8972445534911817, + "learning_rate": 6.796346128055775e-07, + "loss": 0.983, + "step": 8179 + }, + { + "epoch": 0.737701221986743, + "grad_norm": 1.6956082275101023, + "learning_rate": 6.791958655832684e-07, + "loss": 0.9529, + "step": 8180 + }, + { + "epoch": 0.7377914055102133, + "grad_norm": 1.8627952085709012, + "learning_rate": 6.787572310592518e-07, + "loss": 0.941, + "step": 8181 + }, + { + "epoch": 0.7378815890336835, + "grad_norm": 1.9536414112219667, + "learning_rate": 6.783187092709521e-07, + "loss": 1.0636, + "step": 8182 + }, + { + "epoch": 0.7379717725571538, + "grad_norm": 1.6138072045526959, + "learning_rate": 6.778803002557891e-07, + "loss": 0.986, + "step": 8183 + }, + { + "epoch": 0.7380619560806241, + "grad_norm": 2.106855342821949, + "learning_rate": 6.774420040511686e-07, + "loss": 1.0153, + "step": 8184 + }, + { + "epoch": 0.7381521396040943, + "grad_norm": 1.8073433493957314, + "learning_rate": 6.770038206944886e-07, + "loss": 0.9808, + "step": 8185 + }, + { + "epoch": 0.7382423231275645, + "grad_norm": 1.318424929649102, + "learning_rate": 6.765657502231375e-07, + "loss": 0.9928, + "step": 8186 + }, + { + "epoch": 0.7383325066510349, + "grad_norm": 1.5885013095753056, + "learning_rate": 6.761277926744939e-07, + "loss": 1.0436, + "step": 8187 + }, + { + "epoch": 0.7384226901745051, + "grad_norm": 1.5071134379724016, + "learning_rate": 6.756899480859268e-07, + "loss": 0.9705, + "step": 8188 + }, + { + "epoch": 0.7385128736979754, + "grad_norm": 1.6185389966677142, + "learning_rate": 6.752522164947956e-07, + "loss": 1.0275, + "step": 8189 + }, + { + "epoch": 0.7386030572214456, + "grad_norm": 1.2466644128398277, + "learning_rate": 6.748145979384498e-07, + "loss": 1.0349, + "step": 8190 + }, + { + "epoch": 0.7386932407449159, + "grad_norm": 1.6203090009219316, + "learning_rate": 6.743770924542303e-07, + "loss": 0.9222, + "step": 8191 + }, + { + "epoch": 0.7387834242683862, + "grad_norm": 0.6366266055218884, + "learning_rate": 6.739397000794658e-07, + "loss": 0.8072, + "step": 8192 + }, + { + "epoch": 0.7388736077918564, + "grad_norm": 1.5856751473946944, + "learning_rate": 6.735024208514782e-07, + "loss": 0.9501, + "step": 8193 + }, + { + "epoch": 0.7389637913153266, + "grad_norm": 1.4125164706189792, + "learning_rate": 6.73065254807578e-07, + "loss": 0.8938, + "step": 8194 + }, + { + "epoch": 0.739053974838797, + "grad_norm": 1.634611243247055, + "learning_rate": 6.726282019850669e-07, + "loss": 1.0311, + "step": 8195 + }, + { + "epoch": 0.7391441583622672, + "grad_norm": 1.3444137708109047, + "learning_rate": 6.721912624212376e-07, + "loss": 1.0478, + "step": 8196 + }, + { + "epoch": 0.7392343418857374, + "grad_norm": 1.7948783728108721, + "learning_rate": 6.717544361533696e-07, + "loss": 0.8784, + "step": 8197 + }, + { + "epoch": 0.7393245254092078, + "grad_norm": 1.8206604719127606, + "learning_rate": 6.713177232187386e-07, + "loss": 0.9615, + "step": 8198 + }, + { + "epoch": 0.739414708932678, + "grad_norm": 1.4967272341111717, + "learning_rate": 6.708811236546041e-07, + "loss": 0.974, + "step": 8199 + }, + { + "epoch": 0.7395048924561483, + "grad_norm": 1.7382150701719075, + "learning_rate": 6.704446374982224e-07, + "loss": 0.943, + "step": 8200 + }, + { + "epoch": 0.7395950759796185, + "grad_norm": 2.037013518875199, + "learning_rate": 6.700082647868346e-07, + "loss": 1.0145, + "step": 8201 + }, + { + "epoch": 0.7396852595030888, + "grad_norm": 1.5240601660949344, + "learning_rate": 6.695720055576751e-07, + "loss": 1.0322, + "step": 8202 + }, + { + "epoch": 0.7397754430265591, + "grad_norm": 1.797513102137494, + "learning_rate": 6.691358598479679e-07, + "loss": 0.9898, + "step": 8203 + }, + { + "epoch": 0.7398656265500293, + "grad_norm": 22.874059438911225, + "learning_rate": 6.686998276949276e-07, + "loss": 0.9723, + "step": 8204 + }, + { + "epoch": 0.7399558100734995, + "grad_norm": 0.6350715252250267, + "learning_rate": 6.682639091357587e-07, + "loss": 0.8232, + "step": 8205 + }, + { + "epoch": 0.7400459935969699, + "grad_norm": 1.6741186043483212, + "learning_rate": 6.678281042076568e-07, + "loss": 0.9678, + "step": 8206 + }, + { + "epoch": 0.7401361771204401, + "grad_norm": 1.2854971853764752, + "learning_rate": 6.673924129478059e-07, + "loss": 0.9837, + "step": 8207 + }, + { + "epoch": 0.7402263606439103, + "grad_norm": 1.481106743947222, + "learning_rate": 6.669568353933824e-07, + "loss": 0.9001, + "step": 8208 + }, + { + "epoch": 0.7403165441673806, + "grad_norm": 1.3659864545265046, + "learning_rate": 6.665213715815519e-07, + "loss": 0.9165, + "step": 8209 + }, + { + "epoch": 0.7404067276908509, + "grad_norm": 2.8354377411323797, + "learning_rate": 6.660860215494706e-07, + "loss": 0.9948, + "step": 8210 + }, + { + "epoch": 0.7404969112143212, + "grad_norm": 1.6506246367009612, + "learning_rate": 6.656507853342852e-07, + "loss": 0.933, + "step": 8211 + }, + { + "epoch": 0.7405870947377914, + "grad_norm": 1.5071775057922223, + "learning_rate": 6.652156629731323e-07, + "loss": 0.9911, + "step": 8212 + }, + { + "epoch": 0.7406772782612616, + "grad_norm": 1.6922572860346772, + "learning_rate": 6.647806545031396e-07, + "loss": 1.0444, + "step": 8213 + }, + { + "epoch": 0.740767461784732, + "grad_norm": 1.5960421814805859, + "learning_rate": 6.643457599614224e-07, + "loss": 0.9032, + "step": 8214 + }, + { + "epoch": 0.7408576453082022, + "grad_norm": 1.7307472435459401, + "learning_rate": 6.63910979385091e-07, + "loss": 1.0355, + "step": 8215 + }, + { + "epoch": 0.7409478288316724, + "grad_norm": 1.7912188864945335, + "learning_rate": 6.634763128112409e-07, + "loss": 1.0673, + "step": 8216 + }, + { + "epoch": 0.7410380123551427, + "grad_norm": 2.2542583646709002, + "learning_rate": 6.630417602769622e-07, + "loss": 0.9646, + "step": 8217 + }, + { + "epoch": 0.741128195878613, + "grad_norm": 1.5794755048213982, + "learning_rate": 6.62607321819332e-07, + "loss": 0.9854, + "step": 8218 + }, + { + "epoch": 0.7412183794020832, + "grad_norm": 1.2577276260457446, + "learning_rate": 6.621729974754196e-07, + "loss": 0.9791, + "step": 8219 + }, + { + "epoch": 0.7413085629255535, + "grad_norm": 1.5962582466551734, + "learning_rate": 6.617387872822835e-07, + "loss": 0.9281, + "step": 8220 + }, + { + "epoch": 0.7413987464490238, + "grad_norm": 1.7745005912446568, + "learning_rate": 6.613046912769731e-07, + "loss": 0.9494, + "step": 8221 + }, + { + "epoch": 0.741488929972494, + "grad_norm": 1.9403447831906135, + "learning_rate": 6.608707094965289e-07, + "loss": 1.048, + "step": 8222 + }, + { + "epoch": 0.7415791134959643, + "grad_norm": 2.0690002889287276, + "learning_rate": 6.604368419779787e-07, + "loss": 0.914, + "step": 8223 + }, + { + "epoch": 0.7416692970194345, + "grad_norm": 1.4994334899510358, + "learning_rate": 6.600030887583434e-07, + "loss": 0.9923, + "step": 8224 + }, + { + "epoch": 0.7417594805429049, + "grad_norm": 1.7879573290339361, + "learning_rate": 6.595694498746336e-07, + "loss": 1.0847, + "step": 8225 + }, + { + "epoch": 0.7418496640663751, + "grad_norm": 1.656459363264492, + "learning_rate": 6.591359253638491e-07, + "loss": 0.9509, + "step": 8226 + }, + { + "epoch": 0.7419398475898453, + "grad_norm": 5.093937408182539, + "learning_rate": 6.587025152629808e-07, + "loss": 0.8739, + "step": 8227 + }, + { + "epoch": 0.7420300311133156, + "grad_norm": 1.8526382218659891, + "learning_rate": 6.582692196090107e-07, + "loss": 1.0195, + "step": 8228 + }, + { + "epoch": 0.7421202146367859, + "grad_norm": 1.4173035966284249, + "learning_rate": 6.578360384389074e-07, + "loss": 0.9963, + "step": 8229 + }, + { + "epoch": 0.7422103981602561, + "grad_norm": 1.4035322842881974, + "learning_rate": 6.574029717896355e-07, + "loss": 0.9598, + "step": 8230 + }, + { + "epoch": 0.7423005816837264, + "grad_norm": 1.534793878937486, + "learning_rate": 6.569700196981436e-07, + "loss": 0.9735, + "step": 8231 + }, + { + "epoch": 0.7423907652071966, + "grad_norm": 1.6294860539783282, + "learning_rate": 6.565371822013763e-07, + "loss": 0.8813, + "step": 8232 + }, + { + "epoch": 0.742480948730667, + "grad_norm": 1.7362248633523962, + "learning_rate": 6.561044593362636e-07, + "loss": 1.0328, + "step": 8233 + }, + { + "epoch": 0.7425711322541372, + "grad_norm": 1.488055915494626, + "learning_rate": 6.556718511397288e-07, + "loss": 0.9657, + "step": 8234 + }, + { + "epoch": 0.7426613157776074, + "grad_norm": 1.5602104960768857, + "learning_rate": 6.552393576486843e-07, + "loss": 1.0945, + "step": 8235 + }, + { + "epoch": 0.7427514993010776, + "grad_norm": 1.9887285310268568, + "learning_rate": 6.548069789000325e-07, + "loss": 0.9954, + "step": 8236 + }, + { + "epoch": 0.742841682824548, + "grad_norm": 1.8801813379042247, + "learning_rate": 6.543747149306673e-07, + "loss": 0.9677, + "step": 8237 + }, + { + "epoch": 0.7429318663480182, + "grad_norm": 1.5882200905751593, + "learning_rate": 6.5394256577747e-07, + "loss": 1.0289, + "step": 8238 + }, + { + "epoch": 0.7430220498714885, + "grad_norm": 1.630057241569776, + "learning_rate": 6.535105314773161e-07, + "loss": 1.0094, + "step": 8239 + }, + { + "epoch": 0.7431122333949587, + "grad_norm": 1.7161234814158324, + "learning_rate": 6.530786120670677e-07, + "loss": 0.909, + "step": 8240 + }, + { + "epoch": 0.743202416918429, + "grad_norm": 1.662743560617952, + "learning_rate": 6.526468075835787e-07, + "loss": 1.039, + "step": 8241 + }, + { + "epoch": 0.7432926004418993, + "grad_norm": 2.2491097278414616, + "learning_rate": 6.522151180636937e-07, + "loss": 1.0252, + "step": 8242 + }, + { + "epoch": 0.7433827839653695, + "grad_norm": 1.9169393912073427, + "learning_rate": 6.517835435442461e-07, + "loss": 0.9443, + "step": 8243 + }, + { + "epoch": 0.7434729674888398, + "grad_norm": 2.553259867697734, + "learning_rate": 6.513520840620606e-07, + "loss": 0.9992, + "step": 8244 + }, + { + "epoch": 0.7435631510123101, + "grad_norm": 5.502353684777788, + "learning_rate": 6.509207396539525e-07, + "loss": 0.941, + "step": 8245 + }, + { + "epoch": 0.7436533345357803, + "grad_norm": 0.6216263796439894, + "learning_rate": 6.50489510356724e-07, + "loss": 0.7896, + "step": 8246 + }, + { + "epoch": 0.7437435180592505, + "grad_norm": 1.9241216290427743, + "learning_rate": 6.500583962071732e-07, + "loss": 0.9108, + "step": 8247 + }, + { + "epoch": 0.7438337015827209, + "grad_norm": 2.7761905384523295, + "learning_rate": 6.496273972420827e-07, + "loss": 1.0056, + "step": 8248 + }, + { + "epoch": 0.7439238851061911, + "grad_norm": 2.17153906110871, + "learning_rate": 6.491965134982287e-07, + "loss": 1.0339, + "step": 8249 + }, + { + "epoch": 0.7440140686296614, + "grad_norm": 1.6861923944292014, + "learning_rate": 6.487657450123765e-07, + "loss": 0.9478, + "step": 8250 + }, + { + "epoch": 0.7441042521531316, + "grad_norm": 1.824390672836599, + "learning_rate": 6.483350918212814e-07, + "loss": 0.9911, + "step": 8251 + }, + { + "epoch": 0.7441944356766019, + "grad_norm": 0.7002505569265126, + "learning_rate": 6.479045539616898e-07, + "loss": 0.8217, + "step": 8252 + }, + { + "epoch": 0.7442846192000722, + "grad_norm": 2.028564673495031, + "learning_rate": 6.474741314703358e-07, + "loss": 0.9839, + "step": 8253 + }, + { + "epoch": 0.7443748027235424, + "grad_norm": 1.948872567972018, + "learning_rate": 6.47043824383948e-07, + "loss": 1.0398, + "step": 8254 + }, + { + "epoch": 0.7444649862470126, + "grad_norm": 1.7876969502392668, + "learning_rate": 6.466136327392399e-07, + "loss": 1.0208, + "step": 8255 + }, + { + "epoch": 0.744555169770483, + "grad_norm": 1.7521417318523451, + "learning_rate": 6.461835565729206e-07, + "loss": 1.0165, + "step": 8256 + }, + { + "epoch": 0.7446453532939532, + "grad_norm": 1.4715295827719612, + "learning_rate": 6.457535959216844e-07, + "loss": 1.0091, + "step": 8257 + }, + { + "epoch": 0.7447355368174234, + "grad_norm": 1.6227008987917944, + "learning_rate": 6.453237508222186e-07, + "loss": 0.9454, + "step": 8258 + }, + { + "epoch": 0.7448257203408937, + "grad_norm": 1.6402551869986426, + "learning_rate": 6.448940213112e-07, + "loss": 1.007, + "step": 8259 + }, + { + "epoch": 0.744915903864364, + "grad_norm": 1.417225035787538, + "learning_rate": 6.444644074252954e-07, + "loss": 1.0665, + "step": 8260 + }, + { + "epoch": 0.7450060873878342, + "grad_norm": 1.7433754330664142, + "learning_rate": 6.440349092011628e-07, + "loss": 1.0491, + "step": 8261 + }, + { + "epoch": 0.7450962709113045, + "grad_norm": 1.6704071589055247, + "learning_rate": 6.436055266754475e-07, + "loss": 0.9731, + "step": 8262 + }, + { + "epoch": 0.7451864544347747, + "grad_norm": 4.282291035183126, + "learning_rate": 6.431762598847879e-07, + "loss": 0.9565, + "step": 8263 + }, + { + "epoch": 0.7452766379582451, + "grad_norm": 1.7997732258042676, + "learning_rate": 6.427471088658111e-07, + "loss": 0.9582, + "step": 8264 + }, + { + "epoch": 0.7453668214817153, + "grad_norm": 3.6411876325573065, + "learning_rate": 6.42318073655135e-07, + "loss": 0.9787, + "step": 8265 + }, + { + "epoch": 0.7454570050051855, + "grad_norm": 1.7091592404957323, + "learning_rate": 6.41889154289367e-07, + "loss": 0.9477, + "step": 8266 + }, + { + "epoch": 0.7455471885286559, + "grad_norm": 1.784323199534072, + "learning_rate": 6.414603508051055e-07, + "loss": 1.0381, + "step": 8267 + }, + { + "epoch": 0.7456373720521261, + "grad_norm": 0.644065272340541, + "learning_rate": 6.410316632389365e-07, + "loss": 0.8419, + "step": 8268 + }, + { + "epoch": 0.7457275555755963, + "grad_norm": 1.7133276282981165, + "learning_rate": 6.406030916274406e-07, + "loss": 0.9892, + "step": 8269 + }, + { + "epoch": 0.7458177390990666, + "grad_norm": 1.6975692284334376, + "learning_rate": 6.401746360071831e-07, + "loss": 0.9437, + "step": 8270 + }, + { + "epoch": 0.7459079226225369, + "grad_norm": 2.6894376112989833, + "learning_rate": 6.397462964147251e-07, + "loss": 0.9446, + "step": 8271 + }, + { + "epoch": 0.7459981061460071, + "grad_norm": 1.589824779500417, + "learning_rate": 6.393180728866128e-07, + "loss": 0.9733, + "step": 8272 + }, + { + "epoch": 0.7460882896694774, + "grad_norm": 1.5378814658235478, + "learning_rate": 6.388899654593853e-07, + "loss": 1.0158, + "step": 8273 + }, + { + "epoch": 0.7461784731929476, + "grad_norm": 1.939996218825126, + "learning_rate": 6.384619741695709e-07, + "loss": 1.006, + "step": 8274 + }, + { + "epoch": 0.746268656716418, + "grad_norm": 1.3631481116573065, + "learning_rate": 6.380340990536883e-07, + "loss": 0.9835, + "step": 8275 + }, + { + "epoch": 0.7463588402398882, + "grad_norm": 2.079529481268384, + "learning_rate": 6.37606340148247e-07, + "loss": 0.9873, + "step": 8276 + }, + { + "epoch": 0.7464490237633584, + "grad_norm": 1.97522210670961, + "learning_rate": 6.371786974897433e-07, + "loss": 0.8919, + "step": 8277 + }, + { + "epoch": 0.7465392072868287, + "grad_norm": 1.3380988856925113, + "learning_rate": 6.367511711146691e-07, + "loss": 1.014, + "step": 8278 + }, + { + "epoch": 0.746629390810299, + "grad_norm": 2.2504456396650423, + "learning_rate": 6.363237610595014e-07, + "loss": 0.9269, + "step": 8279 + }, + { + "epoch": 0.7467195743337692, + "grad_norm": 1.865824437304074, + "learning_rate": 6.358964673607094e-07, + "loss": 0.9624, + "step": 8280 + }, + { + "epoch": 0.7468097578572395, + "grad_norm": 1.3210143983043272, + "learning_rate": 6.354692900547525e-07, + "loss": 1.0161, + "step": 8281 + }, + { + "epoch": 0.7468999413807097, + "grad_norm": 1.818335445914482, + "learning_rate": 6.350422291780797e-07, + "loss": 0.9527, + "step": 8282 + }, + { + "epoch": 0.74699012490418, + "grad_norm": 1.931641920993783, + "learning_rate": 6.346152847671302e-07, + "loss": 1.0836, + "step": 8283 + }, + { + "epoch": 0.7470803084276503, + "grad_norm": 1.7758564857156325, + "learning_rate": 6.34188456858334e-07, + "loss": 0.9747, + "step": 8284 + }, + { + "epoch": 0.7471704919511205, + "grad_norm": 1.813341405841212, + "learning_rate": 6.337617454881081e-07, + "loss": 0.9339, + "step": 8285 + }, + { + "epoch": 0.7472606754745907, + "grad_norm": 1.5321947414188948, + "learning_rate": 6.333351506928651e-07, + "loss": 0.8928, + "step": 8286 + }, + { + "epoch": 0.7473508589980611, + "grad_norm": 1.4832191384038522, + "learning_rate": 6.329086725090018e-07, + "loss": 0.9558, + "step": 8287 + }, + { + "epoch": 0.7474410425215313, + "grad_norm": 2.1252176509917553, + "learning_rate": 6.324823109729087e-07, + "loss": 0.9447, + "step": 8288 + }, + { + "epoch": 0.7475312260450016, + "grad_norm": 1.8037720707569451, + "learning_rate": 6.320560661209653e-07, + "loss": 0.9158, + "step": 8289 + }, + { + "epoch": 0.7476214095684719, + "grad_norm": 1.435183856762963, + "learning_rate": 6.316299379895411e-07, + "loss": 0.8632, + "step": 8290 + }, + { + "epoch": 0.7477115930919421, + "grad_norm": 1.6486231260960778, + "learning_rate": 6.312039266149965e-07, + "loss": 1.0158, + "step": 8291 + }, + { + "epoch": 0.7478017766154124, + "grad_norm": 1.4414193997907547, + "learning_rate": 6.307780320336789e-07, + "loss": 1.0113, + "step": 8292 + }, + { + "epoch": 0.7478919601388826, + "grad_norm": 2.2289686620125755, + "learning_rate": 6.303522542819306e-07, + "loss": 1.0405, + "step": 8293 + }, + { + "epoch": 0.7479821436623529, + "grad_norm": 1.9206945003667328, + "learning_rate": 6.299265933960796e-07, + "loss": 0.9512, + "step": 8294 + }, + { + "epoch": 0.7480723271858232, + "grad_norm": 1.9540195705249803, + "learning_rate": 6.295010494124462e-07, + "loss": 0.9331, + "step": 8295 + }, + { + "epoch": 0.7481625107092934, + "grad_norm": 1.530308803984666, + "learning_rate": 6.290756223673399e-07, + "loss": 0.9578, + "step": 8296 + }, + { + "epoch": 0.7482526942327636, + "grad_norm": 1.3572842974205341, + "learning_rate": 6.28650312297061e-07, + "loss": 1.0138, + "step": 8297 + }, + { + "epoch": 0.748342877756234, + "grad_norm": 2.4042114936796954, + "learning_rate": 6.282251192378987e-07, + "loss": 0.8977, + "step": 8298 + }, + { + "epoch": 0.7484330612797042, + "grad_norm": 1.335316653183813, + "learning_rate": 6.278000432261334e-07, + "loss": 0.9077, + "step": 8299 + }, + { + "epoch": 0.7485232448031744, + "grad_norm": 1.9086634881999875, + "learning_rate": 6.273750842980345e-07, + "loss": 1.0034, + "step": 8300 + }, + { + "epoch": 0.7486134283266447, + "grad_norm": 1.570401165484998, + "learning_rate": 6.269502424898625e-07, + "loss": 1.0082, + "step": 8301 + }, + { + "epoch": 0.748703611850115, + "grad_norm": 1.985009700165308, + "learning_rate": 6.265255178378663e-07, + "loss": 1.0239, + "step": 8302 + }, + { + "epoch": 0.7487937953735853, + "grad_norm": 1.6525963844895935, + "learning_rate": 6.261009103782861e-07, + "loss": 0.8764, + "step": 8303 + }, + { + "epoch": 0.7488839788970555, + "grad_norm": 1.5060006754444324, + "learning_rate": 6.256764201473519e-07, + "loss": 1.0292, + "step": 8304 + }, + { + "epoch": 0.7489741624205257, + "grad_norm": 0.6283558634766947, + "learning_rate": 6.252520471812835e-07, + "loss": 0.8099, + "step": 8305 + }, + { + "epoch": 0.7490643459439961, + "grad_norm": 2.122445871586298, + "learning_rate": 6.248277915162912e-07, + "loss": 0.9957, + "step": 8306 + }, + { + "epoch": 0.7491545294674663, + "grad_norm": 1.3309105776204306, + "learning_rate": 6.244036531885731e-07, + "loss": 0.8361, + "step": 8307 + }, + { + "epoch": 0.7492447129909365, + "grad_norm": 1.7996441277406015, + "learning_rate": 6.239796322343216e-07, + "loss": 0.9147, + "step": 8308 + }, + { + "epoch": 0.7493348965144068, + "grad_norm": 1.5204438359613908, + "learning_rate": 6.235557286897137e-07, + "loss": 0.9187, + "step": 8309 + }, + { + "epoch": 0.7494250800378771, + "grad_norm": 1.8281624292959868, + "learning_rate": 6.231319425909223e-07, + "loss": 0.9774, + "step": 8310 + }, + { + "epoch": 0.7495152635613473, + "grad_norm": 1.9026633744138923, + "learning_rate": 6.227082739741045e-07, + "loss": 1.0138, + "step": 8311 + }, + { + "epoch": 0.7496054470848176, + "grad_norm": 4.763223211689736, + "learning_rate": 6.222847228754113e-07, + "loss": 1.0102, + "step": 8312 + }, + { + "epoch": 0.7496956306082878, + "grad_norm": 1.885945922023675, + "learning_rate": 6.218612893309823e-07, + "loss": 1.0784, + "step": 8313 + }, + { + "epoch": 0.7497858141317582, + "grad_norm": 1.4495791580887143, + "learning_rate": 6.214379733769468e-07, + "loss": 0.9207, + "step": 8314 + }, + { + "epoch": 0.7498759976552284, + "grad_norm": 1.5581893970021525, + "learning_rate": 6.21014775049425e-07, + "loss": 1.0136, + "step": 8315 + }, + { + "epoch": 0.7499661811786986, + "grad_norm": 0.5738551487282118, + "learning_rate": 6.205916943845267e-07, + "loss": 0.8014, + "step": 8316 + }, + { + "epoch": 0.750056364702169, + "grad_norm": 1.843183527418545, + "learning_rate": 6.201687314183504e-07, + "loss": 1.0276, + "step": 8317 + }, + { + "epoch": 0.7501465482256392, + "grad_norm": 0.6583071889595175, + "learning_rate": 6.197458861869862e-07, + "loss": 0.8487, + "step": 8318 + }, + { + "epoch": 0.7502367317491094, + "grad_norm": 1.6750270328902415, + "learning_rate": 6.193231587265138e-07, + "loss": 1.0283, + "step": 8319 + }, + { + "epoch": 0.7503269152725797, + "grad_norm": 2.1190066246876085, + "learning_rate": 6.189005490730024e-07, + "loss": 0.9628, + "step": 8320 + }, + { + "epoch": 0.75041709879605, + "grad_norm": 2.5082995931018455, + "learning_rate": 6.184780572625115e-07, + "loss": 1.0039, + "step": 8321 + }, + { + "epoch": 0.7505072823195202, + "grad_norm": 1.8147983448459073, + "learning_rate": 6.180556833310902e-07, + "loss": 0.8823, + "step": 8322 + }, + { + "epoch": 0.7505974658429905, + "grad_norm": 1.4953361963768221, + "learning_rate": 6.176334273147788e-07, + "loss": 1.0194, + "step": 8323 + }, + { + "epoch": 0.7506876493664607, + "grad_norm": 2.244689183798138, + "learning_rate": 6.172112892496042e-07, + "loss": 0.8992, + "step": 8324 + }, + { + "epoch": 0.750777832889931, + "grad_norm": 1.472165701438967, + "learning_rate": 6.167892691715883e-07, + "loss": 1.0248, + "step": 8325 + }, + { + "epoch": 0.7508680164134013, + "grad_norm": 1.631722483409222, + "learning_rate": 6.163673671167378e-07, + "loss": 0.9601, + "step": 8326 + }, + { + "epoch": 0.7509581999368715, + "grad_norm": 1.5861208203729875, + "learning_rate": 6.15945583121054e-07, + "loss": 1.0219, + "step": 8327 + }, + { + "epoch": 0.7510483834603418, + "grad_norm": 1.6072077556403992, + "learning_rate": 6.15523917220524e-07, + "loss": 0.9345, + "step": 8328 + }, + { + "epoch": 0.7511385669838121, + "grad_norm": 1.6307121530199866, + "learning_rate": 6.151023694511273e-07, + "loss": 1.0488, + "step": 8329 + }, + { + "epoch": 0.7512287505072823, + "grad_norm": 1.494268434331366, + "learning_rate": 6.146809398488328e-07, + "loss": 0.9231, + "step": 8330 + }, + { + "epoch": 0.7513189340307526, + "grad_norm": 1.5145119581584787, + "learning_rate": 6.142596284495989e-07, + "loss": 1.0152, + "step": 8331 + }, + { + "epoch": 0.7514091175542228, + "grad_norm": 2.509514822098882, + "learning_rate": 6.138384352893751e-07, + "loss": 0.9374, + "step": 8332 + }, + { + "epoch": 0.7514993010776931, + "grad_norm": 1.8720046595531739, + "learning_rate": 6.134173604040987e-07, + "loss": 0.9719, + "step": 8333 + }, + { + "epoch": 0.7515894846011634, + "grad_norm": 1.5768581631094603, + "learning_rate": 6.129964038296984e-07, + "loss": 0.9834, + "step": 8334 + }, + { + "epoch": 0.7516796681246336, + "grad_norm": 1.6510301784111683, + "learning_rate": 6.12575565602093e-07, + "loss": 1.0492, + "step": 8335 + }, + { + "epoch": 0.7517698516481038, + "grad_norm": 1.308230267046951, + "learning_rate": 6.121548457571905e-07, + "loss": 1.0552, + "step": 8336 + }, + { + "epoch": 0.7518600351715742, + "grad_norm": 1.7410632555982872, + "learning_rate": 6.11734244330889e-07, + "loss": 0.9304, + "step": 8337 + }, + { + "epoch": 0.7519502186950444, + "grad_norm": 1.6649141236803617, + "learning_rate": 6.113137613590773e-07, + "loss": 0.9951, + "step": 8338 + }, + { + "epoch": 0.7520404022185146, + "grad_norm": 1.6676438884049993, + "learning_rate": 6.108933968776313e-07, + "loss": 1.0389, + "step": 8339 + }, + { + "epoch": 0.752130585741985, + "grad_norm": 1.5498852964228904, + "learning_rate": 6.104731509224212e-07, + "loss": 0.9824, + "step": 8340 + }, + { + "epoch": 0.7522207692654552, + "grad_norm": 1.4991116277244134, + "learning_rate": 6.100530235293027e-07, + "loss": 0.9146, + "step": 8341 + }, + { + "epoch": 0.7523109527889255, + "grad_norm": 1.9688318023641243, + "learning_rate": 6.096330147341253e-07, + "loss": 1.0013, + "step": 8342 + }, + { + "epoch": 0.7524011363123957, + "grad_norm": 3.2990644515981105, + "learning_rate": 6.09213124572725e-07, + "loss": 1.0386, + "step": 8343 + }, + { + "epoch": 0.752491319835866, + "grad_norm": 1.4969059504930768, + "learning_rate": 6.087933530809297e-07, + "loss": 1.0584, + "step": 8344 + }, + { + "epoch": 0.7525815033593363, + "grad_norm": 1.5633565462309966, + "learning_rate": 6.083737002945566e-07, + "loss": 0.9695, + "step": 8345 + }, + { + "epoch": 0.7526716868828065, + "grad_norm": 1.7171960221305604, + "learning_rate": 6.079541662494126e-07, + "loss": 0.9317, + "step": 8346 + }, + { + "epoch": 0.7527618704062767, + "grad_norm": 1.7514491212137564, + "learning_rate": 6.075347509812954e-07, + "loss": 0.9876, + "step": 8347 + }, + { + "epoch": 0.7528520539297471, + "grad_norm": 1.587901810985546, + "learning_rate": 6.0711545452599e-07, + "loss": 0.94, + "step": 8348 + }, + { + "epoch": 0.7529422374532173, + "grad_norm": 1.8213598815403607, + "learning_rate": 6.066962769192756e-07, + "loss": 0.9021, + "step": 8349 + }, + { + "epoch": 0.7530324209766875, + "grad_norm": 2.054535602713217, + "learning_rate": 6.062772181969167e-07, + "loss": 0.915, + "step": 8350 + }, + { + "epoch": 0.7531226045001578, + "grad_norm": 1.4441625195660326, + "learning_rate": 6.058582783946706e-07, + "loss": 1.0488, + "step": 8351 + }, + { + "epoch": 0.7532127880236281, + "grad_norm": 1.698328034664395, + "learning_rate": 6.054394575482833e-07, + "loss": 0.9697, + "step": 8352 + }, + { + "epoch": 0.7533029715470984, + "grad_norm": 1.5589920338151262, + "learning_rate": 6.05020755693491e-07, + "loss": 0.8992, + "step": 8353 + }, + { + "epoch": 0.7533931550705686, + "grad_norm": 5.191496814543474, + "learning_rate": 6.046021728660198e-07, + "loss": 0.8964, + "step": 8354 + }, + { + "epoch": 0.7534833385940388, + "grad_norm": 1.6973973827778324, + "learning_rate": 6.041837091015858e-07, + "loss": 1.0191, + "step": 8355 + }, + { + "epoch": 0.7535735221175092, + "grad_norm": 1.527708434448211, + "learning_rate": 6.037653644358931e-07, + "loss": 1.0314, + "step": 8356 + }, + { + "epoch": 0.7536637056409794, + "grad_norm": 1.9509023608162963, + "learning_rate": 6.033471389046393e-07, + "loss": 0.9922, + "step": 8357 + }, + { + "epoch": 0.7537538891644496, + "grad_norm": 1.495940835573308, + "learning_rate": 6.029290325435084e-07, + "loss": 0.9358, + "step": 8358 + }, + { + "epoch": 0.7538440726879199, + "grad_norm": 1.97000717713651, + "learning_rate": 6.025110453881756e-07, + "loss": 1.0465, + "step": 8359 + }, + { + "epoch": 0.7539342562113902, + "grad_norm": 2.5096201813705656, + "learning_rate": 6.020931774743061e-07, + "loss": 0.9229, + "step": 8360 + }, + { + "epoch": 0.7540244397348604, + "grad_norm": 1.6648429509003904, + "learning_rate": 6.016754288375546e-07, + "loss": 0.9054, + "step": 8361 + }, + { + "epoch": 0.7541146232583307, + "grad_norm": 1.5291975629770636, + "learning_rate": 6.012577995135665e-07, + "loss": 0.975, + "step": 8362 + }, + { + "epoch": 0.754204806781801, + "grad_norm": 1.7347623375369943, + "learning_rate": 6.008402895379743e-07, + "loss": 0.921, + "step": 8363 + }, + { + "epoch": 0.7542949903052713, + "grad_norm": 1.7696142450873342, + "learning_rate": 6.004228989464047e-07, + "loss": 1.0677, + "step": 8364 + }, + { + "epoch": 0.7543851738287415, + "grad_norm": 10.363239286807238, + "learning_rate": 6.000056277744692e-07, + "loss": 1.0017, + "step": 8365 + }, + { + "epoch": 0.7544753573522117, + "grad_norm": 1.4778602185513967, + "learning_rate": 5.995884760577745e-07, + "loss": 0.951, + "step": 8366 + }, + { + "epoch": 0.7545655408756821, + "grad_norm": 0.6318792596554583, + "learning_rate": 5.99171443831912e-07, + "loss": 0.8018, + "step": 8367 + }, + { + "epoch": 0.7546557243991523, + "grad_norm": 1.8014932743214827, + "learning_rate": 5.98754531132466e-07, + "loss": 0.9779, + "step": 8368 + }, + { + "epoch": 0.7547459079226225, + "grad_norm": 2.273830209484367, + "learning_rate": 5.983377379950099e-07, + "loss": 0.9834, + "step": 8369 + }, + { + "epoch": 0.7548360914460928, + "grad_norm": 1.7964883803032445, + "learning_rate": 5.979210644551067e-07, + "loss": 0.9197, + "step": 8370 + }, + { + "epoch": 0.7549262749695631, + "grad_norm": 2.309321203366609, + "learning_rate": 5.975045105483091e-07, + "loss": 0.9767, + "step": 8371 + }, + { + "epoch": 0.7550164584930333, + "grad_norm": 0.6019070121832006, + "learning_rate": 5.970880763101607e-07, + "loss": 0.8246, + "step": 8372 + }, + { + "epoch": 0.7551066420165036, + "grad_norm": 1.5753389447941377, + "learning_rate": 5.966717617761925e-07, + "loss": 1.0057, + "step": 8373 + }, + { + "epoch": 0.7551968255399738, + "grad_norm": 1.499020256511283, + "learning_rate": 5.962555669819276e-07, + "loss": 0.9642, + "step": 8374 + }, + { + "epoch": 0.7552870090634441, + "grad_norm": 1.391730308600352, + "learning_rate": 5.958394919628777e-07, + "loss": 1.0006, + "step": 8375 + }, + { + "epoch": 0.7553771925869144, + "grad_norm": 1.29219972726925, + "learning_rate": 5.954235367545451e-07, + "loss": 0.9225, + "step": 8376 + }, + { + "epoch": 0.7554673761103846, + "grad_norm": 1.7686662391267143, + "learning_rate": 5.950077013924213e-07, + "loss": 0.9737, + "step": 8377 + }, + { + "epoch": 0.7555575596338548, + "grad_norm": 1.7748520373071937, + "learning_rate": 5.945919859119865e-07, + "loss": 1.0672, + "step": 8378 + }, + { + "epoch": 0.7556477431573252, + "grad_norm": 2.748845811868013, + "learning_rate": 5.94176390348714e-07, + "loss": 0.8734, + "step": 8379 + }, + { + "epoch": 0.7557379266807954, + "grad_norm": 1.274892104950388, + "learning_rate": 5.937609147380622e-07, + "loss": 1.0884, + "step": 8380 + }, + { + "epoch": 0.7558281102042657, + "grad_norm": 1.7390852194070616, + "learning_rate": 5.933455591154844e-07, + "loss": 0.9217, + "step": 8381 + }, + { + "epoch": 0.7559182937277359, + "grad_norm": 2.0699433771396283, + "learning_rate": 5.929303235164191e-07, + "loss": 0.9535, + "step": 8382 + }, + { + "epoch": 0.7560084772512062, + "grad_norm": 1.594755640833522, + "learning_rate": 5.92515207976297e-07, + "loss": 0.9518, + "step": 8383 + }, + { + "epoch": 0.7560986607746765, + "grad_norm": 1.4368132733878882, + "learning_rate": 5.921002125305383e-07, + "loss": 0.9695, + "step": 8384 + }, + { + "epoch": 0.7561888442981467, + "grad_norm": 1.663682849540618, + "learning_rate": 5.916853372145525e-07, + "loss": 0.9, + "step": 8385 + }, + { + "epoch": 0.756279027821617, + "grad_norm": 1.705808991949744, + "learning_rate": 5.912705820637389e-07, + "loss": 0.9167, + "step": 8386 + }, + { + "epoch": 0.7563692113450873, + "grad_norm": 1.5249285321608492, + "learning_rate": 5.908559471134871e-07, + "loss": 1.0461, + "step": 8387 + }, + { + "epoch": 0.7564593948685575, + "grad_norm": 1.6075665581481666, + "learning_rate": 5.904414323991764e-07, + "loss": 0.9205, + "step": 8388 + }, + { + "epoch": 0.7565495783920277, + "grad_norm": 2.4014470109242168, + "learning_rate": 5.900270379561743e-07, + "loss": 0.9879, + "step": 8389 + }, + { + "epoch": 0.7566397619154981, + "grad_norm": 1.7485338608929508, + "learning_rate": 5.896127638198399e-07, + "loss": 0.9982, + "step": 8390 + }, + { + "epoch": 0.7567299454389683, + "grad_norm": 1.987168096023801, + "learning_rate": 5.89198610025521e-07, + "loss": 0.987, + "step": 8391 + }, + { + "epoch": 0.7568201289624386, + "grad_norm": 1.6956042201353896, + "learning_rate": 5.887845766085559e-07, + "loss": 1.0494, + "step": 8392 + }, + { + "epoch": 0.7569103124859088, + "grad_norm": 1.8450513869782081, + "learning_rate": 5.883706636042722e-07, + "loss": 0.9438, + "step": 8393 + }, + { + "epoch": 0.7570004960093791, + "grad_norm": 2.2845554181409407, + "learning_rate": 5.879568710479879e-07, + "loss": 0.9225, + "step": 8394 + }, + { + "epoch": 0.7570906795328494, + "grad_norm": 1.7186957610848823, + "learning_rate": 5.875431989750078e-07, + "loss": 1.0692, + "step": 8395 + }, + { + "epoch": 0.7571808630563196, + "grad_norm": 1.4246711485355934, + "learning_rate": 5.871296474206313e-07, + "loss": 0.9147, + "step": 8396 + }, + { + "epoch": 0.7572710465797898, + "grad_norm": 1.5057574879654714, + "learning_rate": 5.867162164201427e-07, + "loss": 0.9866, + "step": 8397 + }, + { + "epoch": 0.7573612301032602, + "grad_norm": 2.0449161620547587, + "learning_rate": 5.863029060088205e-07, + "loss": 1.0123, + "step": 8398 + }, + { + "epoch": 0.7574514136267304, + "grad_norm": 1.6891230089780993, + "learning_rate": 5.858897162219289e-07, + "loss": 0.8804, + "step": 8399 + }, + { + "epoch": 0.7575415971502006, + "grad_norm": 1.753689894893559, + "learning_rate": 5.854766470947238e-07, + "loss": 1.0288, + "step": 8400 + }, + { + "epoch": 0.7576317806736709, + "grad_norm": 1.4775088068252642, + "learning_rate": 5.850636986624511e-07, + "loss": 0.9704, + "step": 8401 + }, + { + "epoch": 0.7577219641971412, + "grad_norm": 1.598151542911597, + "learning_rate": 5.846508709603453e-07, + "loss": 0.9919, + "step": 8402 + }, + { + "epoch": 0.7578121477206115, + "grad_norm": 1.4790314715572523, + "learning_rate": 5.842381640236318e-07, + "loss": 0.9596, + "step": 8403 + }, + { + "epoch": 0.7579023312440817, + "grad_norm": 1.5622135662751278, + "learning_rate": 5.838255778875242e-07, + "loss": 0.9828, + "step": 8404 + }, + { + "epoch": 0.7579925147675519, + "grad_norm": 1.6339568731997067, + "learning_rate": 5.83413112587227e-07, + "loss": 1.0023, + "step": 8405 + }, + { + "epoch": 0.7580826982910223, + "grad_norm": 1.6155545748740159, + "learning_rate": 5.830007681579338e-07, + "loss": 0.9678, + "step": 8406 + }, + { + "epoch": 0.7581728818144925, + "grad_norm": 1.823167583816976, + "learning_rate": 5.825885446348284e-07, + "loss": 0.9939, + "step": 8407 + }, + { + "epoch": 0.7582630653379627, + "grad_norm": 1.6684665339162736, + "learning_rate": 5.821764420530842e-07, + "loss": 0.9192, + "step": 8408 + }, + { + "epoch": 0.7583532488614331, + "grad_norm": 2.4282603104642586, + "learning_rate": 5.817644604478633e-07, + "loss": 1.0011, + "step": 8409 + }, + { + "epoch": 0.7584434323849033, + "grad_norm": 1.5871281406082791, + "learning_rate": 5.81352599854319e-07, + "loss": 0.9258, + "step": 8410 + }, + { + "epoch": 0.7585336159083735, + "grad_norm": 1.5336610924057679, + "learning_rate": 5.809408603075938e-07, + "loss": 1.0147, + "step": 8411 + }, + { + "epoch": 0.7586237994318438, + "grad_norm": 4.987127709894725, + "learning_rate": 5.805292418428176e-07, + "loss": 0.9612, + "step": 8412 + }, + { + "epoch": 0.7587139829553141, + "grad_norm": 2.483317119963627, + "learning_rate": 5.801177444951148e-07, + "loss": 0.9678, + "step": 8413 + }, + { + "epoch": 0.7588041664787843, + "grad_norm": 1.4075559274325133, + "learning_rate": 5.797063682995944e-07, + "loss": 1.0098, + "step": 8414 + }, + { + "epoch": 0.7588943500022546, + "grad_norm": 1.5696090716925195, + "learning_rate": 5.792951132913584e-07, + "loss": 1.0474, + "step": 8415 + }, + { + "epoch": 0.7589845335257248, + "grad_norm": 2.0212266545784683, + "learning_rate": 5.788839795054968e-07, + "loss": 1.0331, + "step": 8416 + }, + { + "epoch": 0.7590747170491952, + "grad_norm": 1.383316637804019, + "learning_rate": 5.784729669770898e-07, + "loss": 0.8959, + "step": 8417 + }, + { + "epoch": 0.7591649005726654, + "grad_norm": 2.2509858832507676, + "learning_rate": 5.780620757412084e-07, + "loss": 0.9813, + "step": 8418 + }, + { + "epoch": 0.7592550840961356, + "grad_norm": 2.8777025208846823, + "learning_rate": 5.776513058329098e-07, + "loss": 1.0035, + "step": 8419 + }, + { + "epoch": 0.7593452676196059, + "grad_norm": 1.4028823370831316, + "learning_rate": 5.772406572872459e-07, + "loss": 1.0053, + "step": 8420 + }, + { + "epoch": 0.7594354511430762, + "grad_norm": 2.2871444008846593, + "learning_rate": 5.768301301392535e-07, + "loss": 1.0112, + "step": 8421 + }, + { + "epoch": 0.7595256346665464, + "grad_norm": 1.6954463782823859, + "learning_rate": 5.764197244239615e-07, + "loss": 0.9911, + "step": 8422 + }, + { + "epoch": 0.7596158181900167, + "grad_norm": 2.774344196668491, + "learning_rate": 5.760094401763884e-07, + "loss": 1.0178, + "step": 8423 + }, + { + "epoch": 0.7597060017134869, + "grad_norm": 1.7112661943024676, + "learning_rate": 5.755992774315414e-07, + "loss": 0.9654, + "step": 8424 + }, + { + "epoch": 0.7597961852369572, + "grad_norm": 1.6961102696145363, + "learning_rate": 5.751892362244183e-07, + "loss": 0.9712, + "step": 8425 + }, + { + "epoch": 0.7598863687604275, + "grad_norm": 1.7044370304649898, + "learning_rate": 5.747793165900065e-07, + "loss": 1.033, + "step": 8426 + }, + { + "epoch": 0.7599765522838977, + "grad_norm": 1.5664580757528848, + "learning_rate": 5.743695185632806e-07, + "loss": 1.0368, + "step": 8427 + }, + { + "epoch": 0.7600667358073679, + "grad_norm": 1.7594082519207643, + "learning_rate": 5.739598421792098e-07, + "loss": 1.0231, + "step": 8428 + }, + { + "epoch": 0.7601569193308383, + "grad_norm": 1.4794256314959524, + "learning_rate": 5.735502874727474e-07, + "loss": 0.9604, + "step": 8429 + }, + { + "epoch": 0.7602471028543085, + "grad_norm": 1.5143462616905092, + "learning_rate": 5.731408544788398e-07, + "loss": 1.111, + "step": 8430 + }, + { + "epoch": 0.7603372863777788, + "grad_norm": 1.6046473125711143, + "learning_rate": 5.727315432324225e-07, + "loss": 0.9003, + "step": 8431 + }, + { + "epoch": 0.760427469901249, + "grad_norm": 1.845147395358635, + "learning_rate": 5.723223537684196e-07, + "loss": 0.9686, + "step": 8432 + }, + { + "epoch": 0.7605176534247193, + "grad_norm": 1.7039748704919884, + "learning_rate": 5.719132861217462e-07, + "loss": 1.0576, + "step": 8433 + }, + { + "epoch": 0.7606078369481896, + "grad_norm": 1.3134249652243546, + "learning_rate": 5.715043403273044e-07, + "loss": 0.9481, + "step": 8434 + }, + { + "epoch": 0.7606980204716598, + "grad_norm": 2.964143089763346, + "learning_rate": 5.710955164199902e-07, + "loss": 0.9421, + "step": 8435 + }, + { + "epoch": 0.7607882039951301, + "grad_norm": 1.6903228461551387, + "learning_rate": 5.706868144346841e-07, + "loss": 1.0012, + "step": 8436 + }, + { + "epoch": 0.7608783875186004, + "grad_norm": 1.944723875161345, + "learning_rate": 5.702782344062613e-07, + "loss": 0.9742, + "step": 8437 + }, + { + "epoch": 0.7609685710420706, + "grad_norm": 1.8949818498049231, + "learning_rate": 5.698697763695826e-07, + "loss": 0.9558, + "step": 8438 + }, + { + "epoch": 0.7610587545655408, + "grad_norm": 1.5794191247582834, + "learning_rate": 5.694614403595002e-07, + "loss": 0.9999, + "step": 8439 + }, + { + "epoch": 0.7611489380890112, + "grad_norm": 1.5190252329404987, + "learning_rate": 5.690532264108554e-07, + "loss": 0.8907, + "step": 8440 + }, + { + "epoch": 0.7612391216124814, + "grad_norm": 1.496078690336044, + "learning_rate": 5.686451345584795e-07, + "loss": 1.0194, + "step": 8441 + }, + { + "epoch": 0.7613293051359517, + "grad_norm": 1.7351497260138073, + "learning_rate": 5.682371648371933e-07, + "loss": 1.0813, + "step": 8442 + }, + { + "epoch": 0.7614194886594219, + "grad_norm": 2.0710475552420853, + "learning_rate": 5.678293172818074e-07, + "loss": 0.9404, + "step": 8443 + }, + { + "epoch": 0.7615096721828922, + "grad_norm": 1.8044372649329246, + "learning_rate": 5.674215919271204e-07, + "loss": 0.9957, + "step": 8444 + }, + { + "epoch": 0.7615998557063625, + "grad_norm": 1.7602425965092634, + "learning_rate": 5.670139888079224e-07, + "loss": 0.9085, + "step": 8445 + }, + { + "epoch": 0.7616900392298327, + "grad_norm": 1.8811608505670578, + "learning_rate": 5.666065079589924e-07, + "loss": 1.0421, + "step": 8446 + }, + { + "epoch": 0.7617802227533029, + "grad_norm": 5.745712340226481, + "learning_rate": 5.661991494150986e-07, + "loss": 0.9318, + "step": 8447 + }, + { + "epoch": 0.7618704062767733, + "grad_norm": 1.4453678429807504, + "learning_rate": 5.657919132109999e-07, + "loss": 1.0293, + "step": 8448 + }, + { + "epoch": 0.7619605898002435, + "grad_norm": 1.498349792013836, + "learning_rate": 5.653847993814421e-07, + "loss": 1.0115, + "step": 8449 + }, + { + "epoch": 0.7620507733237137, + "grad_norm": 1.4817636910944092, + "learning_rate": 5.649778079611647e-07, + "loss": 0.9926, + "step": 8450 + }, + { + "epoch": 0.762140956847184, + "grad_norm": 1.5775877538536514, + "learning_rate": 5.645709389848923e-07, + "loss": 0.9636, + "step": 8451 + }, + { + "epoch": 0.7622311403706543, + "grad_norm": 1.8974630735722022, + "learning_rate": 5.641641924873435e-07, + "loss": 0.9415, + "step": 8452 + }, + { + "epoch": 0.7623213238941245, + "grad_norm": 1.5660549159840882, + "learning_rate": 5.637575685032217e-07, + "loss": 0.9777, + "step": 8453 + }, + { + "epoch": 0.7624115074175948, + "grad_norm": 2.18838450396934, + "learning_rate": 5.633510670672246e-07, + "loss": 1.0625, + "step": 8454 + }, + { + "epoch": 0.762501690941065, + "grad_norm": 0.6493212640758179, + "learning_rate": 5.629446882140354e-07, + "loss": 0.8421, + "step": 8455 + }, + { + "epoch": 0.7625918744645354, + "grad_norm": 1.4558142112316055, + "learning_rate": 5.625384319783295e-07, + "loss": 0.9884, + "step": 8456 + }, + { + "epoch": 0.7626820579880056, + "grad_norm": 1.728703134140075, + "learning_rate": 5.621322983947705e-07, + "loss": 0.9273, + "step": 8457 + }, + { + "epoch": 0.7627722415114758, + "grad_norm": 1.6485650614336655, + "learning_rate": 5.617262874980122e-07, + "loss": 1.0152, + "step": 8458 + }, + { + "epoch": 0.7628624250349462, + "grad_norm": 2.0863297340145452, + "learning_rate": 5.613203993226981e-07, + "loss": 1.0276, + "step": 8459 + }, + { + "epoch": 0.7629526085584164, + "grad_norm": 3.4620546901532774, + "learning_rate": 5.609146339034599e-07, + "loss": 1.0371, + "step": 8460 + }, + { + "epoch": 0.7630427920818866, + "grad_norm": 2.1405976008311516, + "learning_rate": 5.605089912749199e-07, + "loss": 0.9215, + "step": 8461 + }, + { + "epoch": 0.7631329756053569, + "grad_norm": 2.0696353596716435, + "learning_rate": 5.601034714716901e-07, + "loss": 0.9177, + "step": 8462 + }, + { + "epoch": 0.7632231591288272, + "grad_norm": 1.3311511401510405, + "learning_rate": 5.59698074528372e-07, + "loss": 0.9541, + "step": 8463 + }, + { + "epoch": 0.7633133426522974, + "grad_norm": 1.7293245480936077, + "learning_rate": 5.592928004795555e-07, + "loss": 0.9529, + "step": 8464 + }, + { + "epoch": 0.7634035261757677, + "grad_norm": 1.7843775814324996, + "learning_rate": 5.58887649359822e-07, + "loss": 0.9854, + "step": 8465 + }, + { + "epoch": 0.7634937096992379, + "grad_norm": 3.181878162413183, + "learning_rate": 5.584826212037393e-07, + "loss": 0.837, + "step": 8466 + }, + { + "epoch": 0.7635838932227083, + "grad_norm": 1.3338299909091649, + "learning_rate": 5.580777160458689e-07, + "loss": 0.899, + "step": 8467 + }, + { + "epoch": 0.7636740767461785, + "grad_norm": 1.5411165733778442, + "learning_rate": 5.576729339207574e-07, + "loss": 1.0031, + "step": 8468 + }, + { + "epoch": 0.7637642602696487, + "grad_norm": 1.286545996311086, + "learning_rate": 5.572682748629449e-07, + "loss": 1.0392, + "step": 8469 + }, + { + "epoch": 0.763854443793119, + "grad_norm": 1.5163765842969392, + "learning_rate": 5.568637389069582e-07, + "loss": 0.9658, + "step": 8470 + }, + { + "epoch": 0.7639446273165893, + "grad_norm": 1.7698984333574752, + "learning_rate": 5.564593260873145e-07, + "loss": 0.9969, + "step": 8471 + }, + { + "epoch": 0.7640348108400595, + "grad_norm": 1.3542394422880486, + "learning_rate": 5.560550364385206e-07, + "loss": 1.015, + "step": 8472 + }, + { + "epoch": 0.7641249943635298, + "grad_norm": 1.5543434874281812, + "learning_rate": 5.556508699950728e-07, + "loss": 1.056, + "step": 8473 + }, + { + "epoch": 0.764215177887, + "grad_norm": 1.8297593725667787, + "learning_rate": 5.552468267914577e-07, + "loss": 0.9332, + "step": 8474 + }, + { + "epoch": 0.7643053614104703, + "grad_norm": 1.8930202475551625, + "learning_rate": 5.548429068621481e-07, + "loss": 0.8943, + "step": 8475 + }, + { + "epoch": 0.7643955449339406, + "grad_norm": 0.6303042872077769, + "learning_rate": 5.544391102416115e-07, + "loss": 0.8121, + "step": 8476 + }, + { + "epoch": 0.7644857284574108, + "grad_norm": 1.796673307294493, + "learning_rate": 5.540354369643003e-07, + "loss": 1.0351, + "step": 8477 + }, + { + "epoch": 0.764575911980881, + "grad_norm": 2.930505582394414, + "learning_rate": 5.536318870646586e-07, + "loss": 0.833, + "step": 8478 + }, + { + "epoch": 0.7646660955043514, + "grad_norm": 1.775177224136651, + "learning_rate": 5.532284605771194e-07, + "loss": 0.8872, + "step": 8479 + }, + { + "epoch": 0.7647562790278216, + "grad_norm": 1.4847060788003246, + "learning_rate": 5.528251575361052e-07, + "loss": 0.9473, + "step": 8480 + }, + { + "epoch": 0.7648464625512919, + "grad_norm": 1.5456786490452994, + "learning_rate": 5.524219779760284e-07, + "loss": 0.9775, + "step": 8481 + }, + { + "epoch": 0.7649366460747622, + "grad_norm": 1.573575174621453, + "learning_rate": 5.520189219312907e-07, + "loss": 1.006, + "step": 8482 + }, + { + "epoch": 0.7650268295982324, + "grad_norm": 1.6207730998767778, + "learning_rate": 5.516159894362817e-07, + "loss": 0.8643, + "step": 8483 + }, + { + "epoch": 0.7651170131217027, + "grad_norm": 1.4254341434558497, + "learning_rate": 5.512131805253839e-07, + "loss": 0.9162, + "step": 8484 + }, + { + "epoch": 0.7652071966451729, + "grad_norm": 1.4025916940290577, + "learning_rate": 5.508104952329653e-07, + "loss": 0.9934, + "step": 8485 + }, + { + "epoch": 0.7652973801686432, + "grad_norm": 2.3261132477131294, + "learning_rate": 5.504079335933862e-07, + "loss": 1.0072, + "step": 8486 + }, + { + "epoch": 0.7653875636921135, + "grad_norm": 1.7005161988300848, + "learning_rate": 5.500054956409952e-07, + "loss": 1.0102, + "step": 8487 + }, + { + "epoch": 0.7654777472155837, + "grad_norm": 1.617344945347532, + "learning_rate": 5.496031814101303e-07, + "loss": 0.9374, + "step": 8488 + }, + { + "epoch": 0.7655679307390539, + "grad_norm": 1.6387373464461006, + "learning_rate": 5.492009909351203e-07, + "loss": 0.9835, + "step": 8489 + }, + { + "epoch": 0.7656581142625243, + "grad_norm": 1.7925270882957864, + "learning_rate": 5.4879892425028e-07, + "loss": 0.9046, + "step": 8490 + }, + { + "epoch": 0.7657482977859945, + "grad_norm": 1.5083953047328031, + "learning_rate": 5.483969813899184e-07, + "loss": 1.1151, + "step": 8491 + }, + { + "epoch": 0.7658384813094647, + "grad_norm": 1.3145014171599958, + "learning_rate": 5.479951623883299e-07, + "loss": 1.0615, + "step": 8492 + }, + { + "epoch": 0.765928664832935, + "grad_norm": 1.7864761648224243, + "learning_rate": 5.475934672798004e-07, + "loss": 0.914, + "step": 8493 + }, + { + "epoch": 0.7660188483564053, + "grad_norm": 1.9394804154518774, + "learning_rate": 5.471918960986047e-07, + "loss": 0.9434, + "step": 8494 + }, + { + "epoch": 0.7661090318798756, + "grad_norm": 2.312553198949947, + "learning_rate": 5.467904488790071e-07, + "loss": 1.0759, + "step": 8495 + }, + { + "epoch": 0.7661992154033458, + "grad_norm": 1.5445972974564648, + "learning_rate": 5.463891256552615e-07, + "loss": 0.9839, + "step": 8496 + }, + { + "epoch": 0.766289398926816, + "grad_norm": 2.126910977256872, + "learning_rate": 5.459879264616107e-07, + "loss": 0.8864, + "step": 8497 + }, + { + "epoch": 0.7663795824502864, + "grad_norm": 3.2316029044188186, + "learning_rate": 5.455868513322874e-07, + "loss": 0.864, + "step": 8498 + }, + { + "epoch": 0.7664697659737566, + "grad_norm": 1.7378679448519374, + "learning_rate": 5.451859003015143e-07, + "loss": 0.8709, + "step": 8499 + }, + { + "epoch": 0.7665599494972268, + "grad_norm": 1.7937656867108958, + "learning_rate": 5.447850734035009e-07, + "loss": 1.0173, + "step": 8500 + }, + { + "epoch": 0.7666501330206971, + "grad_norm": 2.0457601058917327, + "learning_rate": 5.443843706724494e-07, + "loss": 0.9596, + "step": 8501 + }, + { + "epoch": 0.7667403165441674, + "grad_norm": 1.7280507297663446, + "learning_rate": 5.439837921425494e-07, + "loss": 0.9129, + "step": 8502 + }, + { + "epoch": 0.7668305000676376, + "grad_norm": 1.2787890642152562, + "learning_rate": 5.435833378479807e-07, + "loss": 0.9694, + "step": 8503 + }, + { + "epoch": 0.7669206835911079, + "grad_norm": 1.3704490776246239, + "learning_rate": 5.431830078229128e-07, + "loss": 0.8414, + "step": 8504 + }, + { + "epoch": 0.7670108671145782, + "grad_norm": 1.484508066738127, + "learning_rate": 5.427828021015022e-07, + "loss": 0.9951, + "step": 8505 + }, + { + "epoch": 0.7671010506380485, + "grad_norm": 2.2109761049384695, + "learning_rate": 5.42382720717899e-07, + "loss": 1.0082, + "step": 8506 + }, + { + "epoch": 0.7671912341615187, + "grad_norm": 1.4067826427918293, + "learning_rate": 5.419827637062384e-07, + "loss": 1.0079, + "step": 8507 + }, + { + "epoch": 0.7672814176849889, + "grad_norm": 2.3337681115484084, + "learning_rate": 5.415829311006487e-07, + "loss": 1.122, + "step": 8508 + }, + { + "epoch": 0.7673716012084593, + "grad_norm": 1.5788567234673054, + "learning_rate": 5.411832229352447e-07, + "loss": 1.0586, + "step": 8509 + }, + { + "epoch": 0.7674617847319295, + "grad_norm": 1.5011651758882774, + "learning_rate": 5.407836392441319e-07, + "loss": 1.0741, + "step": 8510 + }, + { + "epoch": 0.7675519682553997, + "grad_norm": 1.4862366897590167, + "learning_rate": 5.403841800614049e-07, + "loss": 0.9952, + "step": 8511 + }, + { + "epoch": 0.76764215177887, + "grad_norm": 1.255692964862918, + "learning_rate": 5.39984845421148e-07, + "loss": 0.9354, + "step": 8512 + }, + { + "epoch": 0.7677323353023403, + "grad_norm": 2.3057413051904114, + "learning_rate": 5.395856353574344e-07, + "loss": 0.9669, + "step": 8513 + }, + { + "epoch": 0.7678225188258105, + "grad_norm": 0.586471136595941, + "learning_rate": 5.391865499043275e-07, + "loss": 0.8041, + "step": 8514 + }, + { + "epoch": 0.7679127023492808, + "grad_norm": 1.3813973637331622, + "learning_rate": 5.387875890958788e-07, + "loss": 0.9849, + "step": 8515 + }, + { + "epoch": 0.768002885872751, + "grad_norm": 1.3924188509432416, + "learning_rate": 5.383887529661298e-07, + "loss": 1.0079, + "step": 8516 + }, + { + "epoch": 0.7680930693962214, + "grad_norm": 1.7773424085674883, + "learning_rate": 5.379900415491116e-07, + "loss": 0.9321, + "step": 8517 + }, + { + "epoch": 0.7681832529196916, + "grad_norm": 1.3659062512763076, + "learning_rate": 5.375914548788447e-07, + "loss": 0.9331, + "step": 8518 + }, + { + "epoch": 0.7682734364431618, + "grad_norm": 2.5449876403508473, + "learning_rate": 5.371929929893384e-07, + "loss": 0.8622, + "step": 8519 + }, + { + "epoch": 0.768363619966632, + "grad_norm": 1.7763604095312897, + "learning_rate": 5.367946559145917e-07, + "loss": 0.9485, + "step": 8520 + }, + { + "epoch": 0.7684538034901024, + "grad_norm": 1.375579105150619, + "learning_rate": 5.363964436885935e-07, + "loss": 0.9998, + "step": 8521 + }, + { + "epoch": 0.7685439870135726, + "grad_norm": 1.2813147086573873, + "learning_rate": 5.359983563453199e-07, + "loss": 1.0644, + "step": 8522 + }, + { + "epoch": 0.7686341705370429, + "grad_norm": 1.5273413304458536, + "learning_rate": 5.356003939187402e-07, + "loss": 1.0951, + "step": 8523 + }, + { + "epoch": 0.7687243540605131, + "grad_norm": 1.6325390600465015, + "learning_rate": 5.352025564428082e-07, + "loss": 0.9541, + "step": 8524 + }, + { + "epoch": 0.7688145375839834, + "grad_norm": 2.4070122242305323, + "learning_rate": 5.348048439514723e-07, + "loss": 1.0218, + "step": 8525 + }, + { + "epoch": 0.7689047211074537, + "grad_norm": 4.338718491607177, + "learning_rate": 5.344072564786653e-07, + "loss": 0.858, + "step": 8526 + }, + { + "epoch": 0.7689949046309239, + "grad_norm": 3.3867998476449435, + "learning_rate": 5.340097940583123e-07, + "loss": 0.9922, + "step": 8527 + }, + { + "epoch": 0.7690850881543942, + "grad_norm": 1.9997714627346108, + "learning_rate": 5.336124567243275e-07, + "loss": 0.9411, + "step": 8528 + }, + { + "epoch": 0.7691752716778645, + "grad_norm": 1.8573634703044888, + "learning_rate": 5.33215244510613e-07, + "loss": 0.985, + "step": 8529 + }, + { + "epoch": 0.7692654552013347, + "grad_norm": 1.9573594235042173, + "learning_rate": 5.328181574510624e-07, + "loss": 0.9748, + "step": 8530 + }, + { + "epoch": 0.769355638724805, + "grad_norm": 1.6202282904848173, + "learning_rate": 5.324211955795559e-07, + "loss": 1.03, + "step": 8531 + }, + { + "epoch": 0.7694458222482753, + "grad_norm": 1.6891762214369084, + "learning_rate": 5.320243589299651e-07, + "loss": 0.9098, + "step": 8532 + }, + { + "epoch": 0.7695360057717455, + "grad_norm": 1.5508023123668073, + "learning_rate": 5.316276475361505e-07, + "loss": 0.9318, + "step": 8533 + }, + { + "epoch": 0.7696261892952158, + "grad_norm": 2.0993254713579748, + "learning_rate": 5.312310614319613e-07, + "loss": 1.0164, + "step": 8534 + }, + { + "epoch": 0.769716372818686, + "grad_norm": 2.152056167692044, + "learning_rate": 5.308346006512367e-07, + "loss": 0.9496, + "step": 8535 + }, + { + "epoch": 0.7698065563421563, + "grad_norm": 1.5491915994285899, + "learning_rate": 5.30438265227805e-07, + "loss": 0.9122, + "step": 8536 + }, + { + "epoch": 0.7698967398656266, + "grad_norm": 0.6019790247767222, + "learning_rate": 5.300420551954837e-07, + "loss": 0.7968, + "step": 8537 + }, + { + "epoch": 0.7699869233890968, + "grad_norm": 1.4809081701210574, + "learning_rate": 5.296459705880798e-07, + "loss": 0.9073, + "step": 8538 + }, + { + "epoch": 0.770077106912567, + "grad_norm": 1.57563207144325, + "learning_rate": 5.292500114393881e-07, + "loss": 1.0319, + "step": 8539 + }, + { + "epoch": 0.7701672904360374, + "grad_norm": 1.5914877749401073, + "learning_rate": 5.288541777831963e-07, + "loss": 1.0025, + "step": 8540 + }, + { + "epoch": 0.7702574739595076, + "grad_norm": 3.509237768762955, + "learning_rate": 5.284584696532772e-07, + "loss": 1.0348, + "step": 8541 + }, + { + "epoch": 0.7703476574829778, + "grad_norm": 1.5334908573809694, + "learning_rate": 5.280628870833954e-07, + "loss": 0.9844, + "step": 8542 + }, + { + "epoch": 0.7704378410064481, + "grad_norm": 1.948172064644993, + "learning_rate": 5.276674301073045e-07, + "loss": 1.009, + "step": 8543 + }, + { + "epoch": 0.7705280245299184, + "grad_norm": 1.4581436942186123, + "learning_rate": 5.272720987587467e-07, + "loss": 0.9947, + "step": 8544 + }, + { + "epoch": 0.7706182080533887, + "grad_norm": 2.5652530117063703, + "learning_rate": 5.268768930714545e-07, + "loss": 0.9364, + "step": 8545 + }, + { + "epoch": 0.7707083915768589, + "grad_norm": 1.5548041218478499, + "learning_rate": 5.264818130791473e-07, + "loss": 1.0085, + "step": 8546 + }, + { + "epoch": 0.7707985751003291, + "grad_norm": 1.9635393731249728, + "learning_rate": 5.260868588155378e-07, + "loss": 0.9846, + "step": 8547 + }, + { + "epoch": 0.7708887586237995, + "grad_norm": 1.2905014791794243, + "learning_rate": 5.256920303143242e-07, + "loss": 1.0067, + "step": 8548 + }, + { + "epoch": 0.7709789421472697, + "grad_norm": 2.1608817070566566, + "learning_rate": 5.252973276091956e-07, + "loss": 1.093, + "step": 8549 + }, + { + "epoch": 0.7710691256707399, + "grad_norm": 1.3985804479411985, + "learning_rate": 5.249027507338307e-07, + "loss": 1.0509, + "step": 8550 + }, + { + "epoch": 0.7711593091942102, + "grad_norm": 1.5951833824042096, + "learning_rate": 5.245082997218966e-07, + "loss": 0.9529, + "step": 8551 + }, + { + "epoch": 0.7712494927176805, + "grad_norm": 1.6755378030572383, + "learning_rate": 5.241139746070499e-07, + "loss": 1.0054, + "step": 8552 + }, + { + "epoch": 0.7713396762411507, + "grad_norm": 1.662270454105334, + "learning_rate": 5.237197754229376e-07, + "loss": 0.8613, + "step": 8553 + }, + { + "epoch": 0.771429859764621, + "grad_norm": 1.9579468677196923, + "learning_rate": 5.233257022031931e-07, + "loss": 0.8884, + "step": 8554 + }, + { + "epoch": 0.7715200432880913, + "grad_norm": 1.601175996617347, + "learning_rate": 5.229317549814432e-07, + "loss": 0.9939, + "step": 8555 + }, + { + "epoch": 0.7716102268115616, + "grad_norm": 10.800507526481793, + "learning_rate": 5.225379337912998e-07, + "loss": 1.0624, + "step": 8556 + }, + { + "epoch": 0.7717004103350318, + "grad_norm": 2.1908137698610184, + "learning_rate": 5.221442386663663e-07, + "loss": 0.9941, + "step": 8557 + }, + { + "epoch": 0.771790593858502, + "grad_norm": 1.7162693502136865, + "learning_rate": 5.217506696402354e-07, + "loss": 1.0228, + "step": 8558 + }, + { + "epoch": 0.7718807773819724, + "grad_norm": 2.0698867072094487, + "learning_rate": 5.213572267464883e-07, + "loss": 0.9976, + "step": 8559 + }, + { + "epoch": 0.7719709609054426, + "grad_norm": 3.213085634754332, + "learning_rate": 5.209639100186965e-07, + "loss": 0.9583, + "step": 8560 + }, + { + "epoch": 0.7720611444289128, + "grad_norm": 0.6638717601828511, + "learning_rate": 5.205707194904179e-07, + "loss": 0.8777, + "step": 8561 + }, + { + "epoch": 0.7721513279523831, + "grad_norm": 1.6298398057350294, + "learning_rate": 5.201776551952042e-07, + "loss": 0.9081, + "step": 8562 + }, + { + "epoch": 0.7722415114758534, + "grad_norm": 1.862368954138221, + "learning_rate": 5.197847171665914e-07, + "loss": 0.9442, + "step": 8563 + }, + { + "epoch": 0.7723316949993236, + "grad_norm": 1.5211436682270374, + "learning_rate": 5.193919054381095e-07, + "loss": 0.9686, + "step": 8564 + }, + { + "epoch": 0.7724218785227939, + "grad_norm": 1.6544431511388784, + "learning_rate": 5.189992200432738e-07, + "loss": 0.9457, + "step": 8565 + }, + { + "epoch": 0.7725120620462641, + "grad_norm": 3.290724092243154, + "learning_rate": 5.186066610155906e-07, + "loss": 0.9917, + "step": 8566 + }, + { + "epoch": 0.7726022455697344, + "grad_norm": 1.9347483876979847, + "learning_rate": 5.182142283885555e-07, + "loss": 1.0339, + "step": 8567 + }, + { + "epoch": 0.7726924290932047, + "grad_norm": 0.657935566388102, + "learning_rate": 5.178219221956528e-07, + "loss": 0.7682, + "step": 8568 + }, + { + "epoch": 0.7727826126166749, + "grad_norm": 1.5835816874559858, + "learning_rate": 5.174297424703565e-07, + "loss": 0.9977, + "step": 8569 + }, + { + "epoch": 0.7728727961401451, + "grad_norm": 2.09494428592426, + "learning_rate": 5.170376892461299e-07, + "loss": 0.911, + "step": 8570 + }, + { + "epoch": 0.7729629796636155, + "grad_norm": 0.7215505853580397, + "learning_rate": 5.16645762556424e-07, + "loss": 0.8677, + "step": 8571 + }, + { + "epoch": 0.7730531631870857, + "grad_norm": 1.5190326098067717, + "learning_rate": 5.162539624346809e-07, + "loss": 1.026, + "step": 8572 + }, + { + "epoch": 0.773143346710556, + "grad_norm": 1.510196125986119, + "learning_rate": 5.158622889143309e-07, + "loss": 1.0451, + "step": 8573 + }, + { + "epoch": 0.7732335302340262, + "grad_norm": 1.6973690093656941, + "learning_rate": 5.154707420287939e-07, + "loss": 0.9415, + "step": 8574 + }, + { + "epoch": 0.7733237137574965, + "grad_norm": 2.6035713939603204, + "learning_rate": 5.150793218114793e-07, + "loss": 0.9072, + "step": 8575 + }, + { + "epoch": 0.7734138972809668, + "grad_norm": 1.54317030737033, + "learning_rate": 5.146880282957837e-07, + "loss": 0.9693, + "step": 8576 + }, + { + "epoch": 0.773504080804437, + "grad_norm": 1.703995228595053, + "learning_rate": 5.142968615150964e-07, + "loss": 0.9707, + "step": 8577 + }, + { + "epoch": 0.7735942643279073, + "grad_norm": 1.6154981995043909, + "learning_rate": 5.139058215027921e-07, + "loss": 0.9082, + "step": 8578 + }, + { + "epoch": 0.7736844478513776, + "grad_norm": 1.2594939656069535, + "learning_rate": 5.135149082922383e-07, + "loss": 0.9685, + "step": 8579 + }, + { + "epoch": 0.7737746313748478, + "grad_norm": 1.6842686833674183, + "learning_rate": 5.131241219167879e-07, + "loss": 0.9101, + "step": 8580 + }, + { + "epoch": 0.773864814898318, + "grad_norm": 0.7830648323929839, + "learning_rate": 5.127334624097869e-07, + "loss": 0.8598, + "step": 8581 + }, + { + "epoch": 0.7739549984217884, + "grad_norm": 1.4061936261215584, + "learning_rate": 5.123429298045672e-07, + "loss": 1.034, + "step": 8582 + }, + { + "epoch": 0.7740451819452586, + "grad_norm": 1.5499709003547166, + "learning_rate": 5.119525241344515e-07, + "loss": 0.8687, + "step": 8583 + }, + { + "epoch": 0.7741353654687289, + "grad_norm": 1.561240107899733, + "learning_rate": 5.115622454327515e-07, + "loss": 0.9868, + "step": 8584 + }, + { + "epoch": 0.7742255489921991, + "grad_norm": 1.7681189958262111, + "learning_rate": 5.11172093732768e-07, + "loss": 0.9702, + "step": 8585 + }, + { + "epoch": 0.7743157325156694, + "grad_norm": 3.1738484073871156, + "learning_rate": 5.107820690677911e-07, + "loss": 1.0096, + "step": 8586 + }, + { + "epoch": 0.7744059160391397, + "grad_norm": 1.7211601698185086, + "learning_rate": 5.103921714710991e-07, + "loss": 0.9275, + "step": 8587 + }, + { + "epoch": 0.7744960995626099, + "grad_norm": 1.6750160728936019, + "learning_rate": 5.100024009759605e-07, + "loss": 0.9941, + "step": 8588 + }, + { + "epoch": 0.7745862830860801, + "grad_norm": 2.4225363535953317, + "learning_rate": 5.09612757615633e-07, + "loss": 1.0185, + "step": 8589 + }, + { + "epoch": 0.7746764666095505, + "grad_norm": 1.401607394482864, + "learning_rate": 5.092232414233628e-07, + "loss": 0.9321, + "step": 8590 + }, + { + "epoch": 0.7747666501330207, + "grad_norm": 2.0002798838281257, + "learning_rate": 5.088338524323858e-07, + "loss": 0.9482, + "step": 8591 + }, + { + "epoch": 0.7748568336564909, + "grad_norm": 2.1971051373681822, + "learning_rate": 5.084445906759271e-07, + "loss": 1.0395, + "step": 8592 + }, + { + "epoch": 0.7749470171799612, + "grad_norm": 1.6487082308784675, + "learning_rate": 5.080554561871995e-07, + "loss": 1.043, + "step": 8593 + }, + { + "epoch": 0.7750372007034315, + "grad_norm": 1.46884796140546, + "learning_rate": 5.076664489994078e-07, + "loss": 1.0074, + "step": 8594 + }, + { + "epoch": 0.7751273842269017, + "grad_norm": 2.1850717418424255, + "learning_rate": 5.07277569145742e-07, + "loss": 0.9499, + "step": 8595 + }, + { + "epoch": 0.775217567750372, + "grad_norm": 2.669680581006284, + "learning_rate": 5.068888166593861e-07, + "loss": 0.8974, + "step": 8596 + }, + { + "epoch": 0.7753077512738422, + "grad_norm": 1.6857720464458192, + "learning_rate": 5.065001915735087e-07, + "loss": 0.9682, + "step": 8597 + }, + { + "epoch": 0.7753979347973126, + "grad_norm": 1.406537174996095, + "learning_rate": 5.061116939212702e-07, + "loss": 0.9459, + "step": 8598 + }, + { + "epoch": 0.7754881183207828, + "grad_norm": 2.161324542091057, + "learning_rate": 5.05723323735819e-07, + "loss": 0.978, + "step": 8599 + }, + { + "epoch": 0.775578301844253, + "grad_norm": 0.6546192115366064, + "learning_rate": 5.053350810502932e-07, + "loss": 0.8204, + "step": 8600 + }, + { + "epoch": 0.7756684853677234, + "grad_norm": 1.6726887228965956, + "learning_rate": 5.049469658978202e-07, + "loss": 0.976, + "step": 8601 + }, + { + "epoch": 0.7757586688911936, + "grad_norm": 0.6201233871029911, + "learning_rate": 5.045589783115147e-07, + "loss": 0.7682, + "step": 8602 + }, + { + "epoch": 0.7758488524146638, + "grad_norm": 1.6736945830124443, + "learning_rate": 5.041711183244842e-07, + "loss": 1.0173, + "step": 8603 + }, + { + "epoch": 0.7759390359381341, + "grad_norm": 2.7622244235713933, + "learning_rate": 5.037833859698211e-07, + "loss": 0.9985, + "step": 8604 + }, + { + "epoch": 0.7760292194616044, + "grad_norm": 2.209947344122467, + "learning_rate": 5.033957812806096e-07, + "loss": 1.0515, + "step": 8605 + }, + { + "epoch": 0.7761194029850746, + "grad_norm": 1.9995616790156452, + "learning_rate": 5.030083042899223e-07, + "loss": 1.0317, + "step": 8606 + }, + { + "epoch": 0.7762095865085449, + "grad_norm": 1.82541083323262, + "learning_rate": 5.026209550308207e-07, + "loss": 0.9783, + "step": 8607 + }, + { + "epoch": 0.7762997700320151, + "grad_norm": 1.933689862569456, + "learning_rate": 5.022337335363558e-07, + "loss": 1.0355, + "step": 8608 + }, + { + "epoch": 0.7763899535554855, + "grad_norm": 2.72008017085237, + "learning_rate": 5.018466398395677e-07, + "loss": 1.0072, + "step": 8609 + }, + { + "epoch": 0.7764801370789557, + "grad_norm": 1.5172489550604673, + "learning_rate": 5.01459673973484e-07, + "loss": 0.9136, + "step": 8610 + }, + { + "epoch": 0.7765703206024259, + "grad_norm": 2.443542717535475, + "learning_rate": 5.01072835971125e-07, + "loss": 1.0441, + "step": 8611 + }, + { + "epoch": 0.7766605041258962, + "grad_norm": 1.4259982009330625, + "learning_rate": 5.006861258654959e-07, + "loss": 0.8852, + "step": 8612 + }, + { + "epoch": 0.7767506876493665, + "grad_norm": 1.83333542852571, + "learning_rate": 5.002995436895938e-07, + "loss": 1.0096, + "step": 8613 + }, + { + "epoch": 0.7768408711728367, + "grad_norm": 1.7471445493240993, + "learning_rate": 4.999130894764039e-07, + "loss": 0.9397, + "step": 8614 + }, + { + "epoch": 0.776931054696307, + "grad_norm": 2.0709271364311053, + "learning_rate": 4.995267632589006e-07, + "loss": 0.9202, + "step": 8615 + }, + { + "epoch": 0.7770212382197772, + "grad_norm": 0.7056711930865732, + "learning_rate": 4.99140565070048e-07, + "loss": 0.8186, + "step": 8616 + }, + { + "epoch": 0.7771114217432475, + "grad_norm": 2.3016490497877107, + "learning_rate": 4.987544949427969e-07, + "loss": 1.0046, + "step": 8617 + }, + { + "epoch": 0.7772016052667178, + "grad_norm": 1.0122169240200358, + "learning_rate": 4.98368552910091e-07, + "loss": 0.8247, + "step": 8618 + }, + { + "epoch": 0.777291788790188, + "grad_norm": 1.6646320639850916, + "learning_rate": 4.979827390048596e-07, + "loss": 1.0938, + "step": 8619 + }, + { + "epoch": 0.7773819723136582, + "grad_norm": 1.2986168711098796, + "learning_rate": 4.975970532600231e-07, + "loss": 0.9422, + "step": 8620 + }, + { + "epoch": 0.7774721558371286, + "grad_norm": 2.127862012749016, + "learning_rate": 4.972114957084901e-07, + "loss": 1.0009, + "step": 8621 + }, + { + "epoch": 0.7775623393605988, + "grad_norm": 1.4362049903822238, + "learning_rate": 4.968260663831585e-07, + "loss": 0.9577, + "step": 8622 + }, + { + "epoch": 0.777652522884069, + "grad_norm": 1.3407959960155063, + "learning_rate": 4.964407653169154e-07, + "loss": 1.0888, + "step": 8623 + }, + { + "epoch": 0.7777427064075394, + "grad_norm": 1.3705028336505063, + "learning_rate": 4.960555925426366e-07, + "loss": 0.9487, + "step": 8624 + }, + { + "epoch": 0.7778328899310096, + "grad_norm": 1.9036896192226451, + "learning_rate": 4.956705480931876e-07, + "loss": 1.0386, + "step": 8625 + }, + { + "epoch": 0.7779230734544799, + "grad_norm": 1.3416744437298356, + "learning_rate": 4.952856320014225e-07, + "loss": 0.962, + "step": 8626 + }, + { + "epoch": 0.7780132569779501, + "grad_norm": 1.9972902657182399, + "learning_rate": 4.949008443001838e-07, + "loss": 0.9202, + "step": 8627 + }, + { + "epoch": 0.7781034405014204, + "grad_norm": 1.5070306994495009, + "learning_rate": 4.945161850223041e-07, + "loss": 0.9942, + "step": 8628 + }, + { + "epoch": 0.7781936240248907, + "grad_norm": 1.6740101522778168, + "learning_rate": 4.941316542006044e-07, + "loss": 0.9116, + "step": 8629 + }, + { + "epoch": 0.7782838075483609, + "grad_norm": 1.4867672566024501, + "learning_rate": 4.937472518678956e-07, + "loss": 1.007, + "step": 8630 + }, + { + "epoch": 0.7783739910718311, + "grad_norm": 1.829709531942969, + "learning_rate": 4.93362978056977e-07, + "loss": 0.9809, + "step": 8631 + }, + { + "epoch": 0.7784641745953015, + "grad_norm": 1.4382400266426534, + "learning_rate": 4.929788328006355e-07, + "loss": 0.9324, + "step": 8632 + }, + { + "epoch": 0.7785543581187717, + "grad_norm": 2.5129816608966586, + "learning_rate": 4.925948161316506e-07, + "loss": 1.0512, + "step": 8633 + }, + { + "epoch": 0.778644541642242, + "grad_norm": 1.8224438153935263, + "learning_rate": 4.922109280827868e-07, + "loss": 0.9695, + "step": 8634 + }, + { + "epoch": 0.7787347251657122, + "grad_norm": 1.9745750235598498, + "learning_rate": 4.918271686868016e-07, + "loss": 0.9721, + "step": 8635 + }, + { + "epoch": 0.7788249086891825, + "grad_norm": 3.108781614185021, + "learning_rate": 4.914435379764379e-07, + "loss": 1.0236, + "step": 8636 + }, + { + "epoch": 0.7789150922126528, + "grad_norm": 2.086964329501073, + "learning_rate": 4.910600359844294e-07, + "loss": 0.9737, + "step": 8637 + }, + { + "epoch": 0.779005275736123, + "grad_norm": 1.5588897194606761, + "learning_rate": 4.90676662743499e-07, + "loss": 0.8863, + "step": 8638 + }, + { + "epoch": 0.7790954592595932, + "grad_norm": 1.7642618504240157, + "learning_rate": 4.902934182863581e-07, + "loss": 0.9749, + "step": 8639 + }, + { + "epoch": 0.7791856427830636, + "grad_norm": 1.7497995125504247, + "learning_rate": 4.899103026457069e-07, + "loss": 0.8534, + "step": 8640 + }, + { + "epoch": 0.7792758263065338, + "grad_norm": 0.5878198777526212, + "learning_rate": 4.895273158542361e-07, + "loss": 0.7675, + "step": 8641 + }, + { + "epoch": 0.779366009830004, + "grad_norm": 1.572891622516587, + "learning_rate": 4.891444579446227e-07, + "loss": 0.9934, + "step": 8642 + }, + { + "epoch": 0.7794561933534743, + "grad_norm": 1.8215378333159302, + "learning_rate": 4.887617289495349e-07, + "loss": 0.9325, + "step": 8643 + }, + { + "epoch": 0.7795463768769446, + "grad_norm": 0.5898449373549075, + "learning_rate": 4.883791289016292e-07, + "loss": 0.7489, + "step": 8644 + }, + { + "epoch": 0.7796365604004148, + "grad_norm": 1.6498487316496775, + "learning_rate": 4.879966578335514e-07, + "loss": 0.9881, + "step": 8645 + }, + { + "epoch": 0.7797267439238851, + "grad_norm": 2.041894458856681, + "learning_rate": 4.876143157779358e-07, + "loss": 0.878, + "step": 8646 + }, + { + "epoch": 0.7798169274473554, + "grad_norm": 2.394950195304337, + "learning_rate": 4.872321027674058e-07, + "loss": 0.8668, + "step": 8647 + }, + { + "epoch": 0.7799071109708257, + "grad_norm": 3.3280310102631914, + "learning_rate": 4.868500188345748e-07, + "loss": 1.0512, + "step": 8648 + }, + { + "epoch": 0.7799972944942959, + "grad_norm": 1.7253225025066008, + "learning_rate": 4.864680640120425e-07, + "loss": 1.045, + "step": 8649 + }, + { + "epoch": 0.7800874780177661, + "grad_norm": 1.8798930854021259, + "learning_rate": 4.860862383324016e-07, + "loss": 0.9883, + "step": 8650 + }, + { + "epoch": 0.7801776615412365, + "grad_norm": 1.540166468602365, + "learning_rate": 4.857045418282295e-07, + "loss": 0.9772, + "step": 8651 + }, + { + "epoch": 0.7802678450647067, + "grad_norm": 1.8034357130471133, + "learning_rate": 4.853229745320966e-07, + "loss": 0.927, + "step": 8652 + }, + { + "epoch": 0.7803580285881769, + "grad_norm": 1.8220883331241862, + "learning_rate": 4.849415364765587e-07, + "loss": 0.9643, + "step": 8653 + }, + { + "epoch": 0.7804482121116472, + "grad_norm": 1.7007529863103947, + "learning_rate": 4.845602276941631e-07, + "loss": 1.0459, + "step": 8654 + }, + { + "epoch": 0.7805383956351175, + "grad_norm": 1.578953138558402, + "learning_rate": 4.841790482174449e-07, + "loss": 1.0195, + "step": 8655 + }, + { + "epoch": 0.7806285791585877, + "grad_norm": 1.9938345888184832, + "learning_rate": 4.837979980789282e-07, + "loss": 0.9688, + "step": 8656 + }, + { + "epoch": 0.780718762682058, + "grad_norm": 1.7019951248502558, + "learning_rate": 4.834170773111273e-07, + "loss": 0.9207, + "step": 8657 + }, + { + "epoch": 0.7808089462055282, + "grad_norm": 1.8330141208643778, + "learning_rate": 4.830362859465431e-07, + "loss": 0.999, + "step": 8658 + }, + { + "epoch": 0.7808991297289986, + "grad_norm": 1.648214099082824, + "learning_rate": 4.826556240176675e-07, + "loss": 1.0459, + "step": 8659 + }, + { + "epoch": 0.7809893132524688, + "grad_norm": 1.8915870047557075, + "learning_rate": 4.822750915569807e-07, + "loss": 0.9936, + "step": 8660 + }, + { + "epoch": 0.781079496775939, + "grad_norm": 2.0141146420046936, + "learning_rate": 4.818946885969514e-07, + "loss": 1.0563, + "step": 8661 + }, + { + "epoch": 0.7811696802994093, + "grad_norm": 1.6863982524698096, + "learning_rate": 4.815144151700383e-07, + "loss": 1.0254, + "step": 8662 + }, + { + "epoch": 0.7812598638228796, + "grad_norm": 1.5852087441781977, + "learning_rate": 4.811342713086885e-07, + "loss": 0.982, + "step": 8663 + }, + { + "epoch": 0.7813500473463498, + "grad_norm": 1.8290668897708087, + "learning_rate": 4.807542570453367e-07, + "loss": 1.019, + "step": 8664 + }, + { + "epoch": 0.7814402308698201, + "grad_norm": 1.5436260899533658, + "learning_rate": 4.803743724124098e-07, + "loss": 0.946, + "step": 8665 + }, + { + "epoch": 0.7815304143932903, + "grad_norm": 2.0674319398833436, + "learning_rate": 4.799946174423192e-07, + "loss": 0.986, + "step": 8666 + }, + { + "epoch": 0.7816205979167606, + "grad_norm": 0.8437878988726041, + "learning_rate": 4.796149921674706e-07, + "loss": 0.7699, + "step": 8667 + }, + { + "epoch": 0.7817107814402309, + "grad_norm": 1.4706914528919297, + "learning_rate": 4.792354966202534e-07, + "loss": 0.9391, + "step": 8668 + }, + { + "epoch": 0.7818009649637011, + "grad_norm": 2.394258219102269, + "learning_rate": 4.788561308330489e-07, + "loss": 0.8058, + "step": 8669 + }, + { + "epoch": 0.7818911484871713, + "grad_norm": 2.405082716758549, + "learning_rate": 4.784768948382272e-07, + "loss": 0.9274, + "step": 8670 + }, + { + "epoch": 0.7819813320106417, + "grad_norm": 1.5616080981046556, + "learning_rate": 4.780977886681461e-07, + "loss": 0.9865, + "step": 8671 + }, + { + "epoch": 0.7820715155341119, + "grad_norm": 1.5955249402073315, + "learning_rate": 4.777188123551541e-07, + "loss": 0.9244, + "step": 8672 + }, + { + "epoch": 0.7821616990575821, + "grad_norm": 1.582310178680652, + "learning_rate": 4.773399659315856e-07, + "loss": 1.0103, + "step": 8673 + }, + { + "epoch": 0.7822518825810525, + "grad_norm": 2.0430572082157856, + "learning_rate": 4.769612494297681e-07, + "loss": 0.9631, + "step": 8674 + }, + { + "epoch": 0.7823420661045227, + "grad_norm": 0.648272872616915, + "learning_rate": 4.765826628820142e-07, + "loss": 0.7928, + "step": 8675 + }, + { + "epoch": 0.782432249627993, + "grad_norm": 2.269721701111038, + "learning_rate": 4.7620420632062775e-07, + "loss": 0.8894, + "step": 8676 + }, + { + "epoch": 0.7825224331514632, + "grad_norm": 2.480730274260794, + "learning_rate": 4.758258797779002e-07, + "loss": 0.9341, + "step": 8677 + }, + { + "epoch": 0.7826126166749335, + "grad_norm": 0.6171486999290957, + "learning_rate": 4.7544768328611317e-07, + "loss": 0.8018, + "step": 8678 + }, + { + "epoch": 0.7827028001984038, + "grad_norm": 1.614980324781814, + "learning_rate": 4.750696168775359e-07, + "loss": 0.9609, + "step": 8679 + }, + { + "epoch": 0.782792983721874, + "grad_norm": 0.6016380213111618, + "learning_rate": 4.746916805844279e-07, + "loss": 0.7896, + "step": 8680 + }, + { + "epoch": 0.7828831672453442, + "grad_norm": 1.7557040306003762, + "learning_rate": 4.743138744390356e-07, + "loss": 1.0062, + "step": 8681 + }, + { + "epoch": 0.7829733507688146, + "grad_norm": 1.5532679378812837, + "learning_rate": 4.739361984735959e-07, + "loss": 1.0523, + "step": 8682 + }, + { + "epoch": 0.7830635342922848, + "grad_norm": 1.8141891895251758, + "learning_rate": 4.7355865272033455e-07, + "loss": 0.934, + "step": 8683 + }, + { + "epoch": 0.783153717815755, + "grad_norm": 1.6217283913972362, + "learning_rate": 4.7318123721146563e-07, + "loss": 0.9724, + "step": 8684 + }, + { + "epoch": 0.7832439013392253, + "grad_norm": 1.5148849253495764, + "learning_rate": 4.728039519791924e-07, + "loss": 0.9669, + "step": 8685 + }, + { + "epoch": 0.7833340848626956, + "grad_norm": 0.6670750445038183, + "learning_rate": 4.72426797055707e-07, + "loss": 0.836, + "step": 8686 + }, + { + "epoch": 0.7834242683861659, + "grad_norm": 2.01706165302332, + "learning_rate": 4.720497724731904e-07, + "loss": 0.9458, + "step": 8687 + }, + { + "epoch": 0.7835144519096361, + "grad_norm": 1.7112628505494574, + "learning_rate": 4.7167287826381153e-07, + "loss": 1.0183, + "step": 8688 + }, + { + "epoch": 0.7836046354331063, + "grad_norm": 1.874191872967327, + "learning_rate": 4.712961144597307e-07, + "loss": 0.9882, + "step": 8689 + }, + { + "epoch": 0.7836948189565767, + "grad_norm": 2.0176954650731145, + "learning_rate": 4.7091948109309343e-07, + "loss": 1.0218, + "step": 8690 + }, + { + "epoch": 0.7837850024800469, + "grad_norm": 1.8283914225184958, + "learning_rate": 4.705429781960384e-07, + "loss": 0.9894, + "step": 8691 + }, + { + "epoch": 0.7838751860035171, + "grad_norm": 1.6512993291609201, + "learning_rate": 4.7016660580068923e-07, + "loss": 0.9799, + "step": 8692 + }, + { + "epoch": 0.7839653695269874, + "grad_norm": 1.8101964979838145, + "learning_rate": 4.6979036393916093e-07, + "loss": 1.0172, + "step": 8693 + }, + { + "epoch": 0.7840555530504577, + "grad_norm": 1.86328412901458, + "learning_rate": 4.6941425264355603e-07, + "loss": 1.0586, + "step": 8694 + }, + { + "epoch": 0.7841457365739279, + "grad_norm": 1.7566474953437126, + "learning_rate": 4.6903827194596666e-07, + "loss": 0.9923, + "step": 8695 + }, + { + "epoch": 0.7842359200973982, + "grad_norm": 1.505897135576538, + "learning_rate": 4.686624218784743e-07, + "loss": 1.0372, + "step": 8696 + }, + { + "epoch": 0.7843261036208685, + "grad_norm": 2.0037570473968835, + "learning_rate": 4.6828670247314696e-07, + "loss": 0.9743, + "step": 8697 + }, + { + "epoch": 0.7844162871443388, + "grad_norm": 1.7785919836632562, + "learning_rate": 4.679111137620442e-07, + "loss": 0.9983, + "step": 8698 + }, + { + "epoch": 0.784506470667809, + "grad_norm": 1.6057218551675505, + "learning_rate": 4.67535655777213e-07, + "loss": 0.9447, + "step": 8699 + }, + { + "epoch": 0.7845966541912792, + "grad_norm": 1.472433301001175, + "learning_rate": 4.6716032855068956e-07, + "loss": 0.9888, + "step": 8700 + }, + { + "epoch": 0.7846868377147496, + "grad_norm": 3.00541690695433, + "learning_rate": 4.6678513211449867e-07, + "loss": 1.0216, + "step": 8701 + }, + { + "epoch": 0.7847770212382198, + "grad_norm": 1.9616572666294108, + "learning_rate": 4.6641006650065516e-07, + "loss": 0.9472, + "step": 8702 + }, + { + "epoch": 0.78486720476169, + "grad_norm": 1.5049385788255938, + "learning_rate": 4.6603513174115973e-07, + "loss": 0.9704, + "step": 8703 + }, + { + "epoch": 0.7849573882851603, + "grad_norm": 1.9550761128509564, + "learning_rate": 4.6566032786800625e-07, + "loss": 0.9937, + "step": 8704 + }, + { + "epoch": 0.7850475718086306, + "grad_norm": 1.6395021501709106, + "learning_rate": 4.6528565491317274e-07, + "loss": 0.958, + "step": 8705 + }, + { + "epoch": 0.7851377553321008, + "grad_norm": 6.739402363709315, + "learning_rate": 4.649111129086305e-07, + "loss": 1.0459, + "step": 8706 + }, + { + "epoch": 0.7852279388555711, + "grad_norm": 1.8028206685780945, + "learning_rate": 4.6453670188633596e-07, + "loss": 1.0057, + "step": 8707 + }, + { + "epoch": 0.7853181223790413, + "grad_norm": 2.1722257042824666, + "learning_rate": 4.641624218782365e-07, + "loss": 0.9958, + "step": 8708 + }, + { + "epoch": 0.7854083059025116, + "grad_norm": 1.4882057951775516, + "learning_rate": 4.6378827291626765e-07, + "loss": 0.9481, + "step": 8709 + }, + { + "epoch": 0.7854984894259819, + "grad_norm": 1.303241133603557, + "learning_rate": 4.634142550323541e-07, + "loss": 1.0349, + "step": 8710 + }, + { + "epoch": 0.7855886729494521, + "grad_norm": 2.0445882822084003, + "learning_rate": 4.6304036825840943e-07, + "loss": 1.0317, + "step": 8711 + }, + { + "epoch": 0.7856788564729223, + "grad_norm": 3.4761486707265057, + "learning_rate": 4.626666126263341e-07, + "loss": 0.8853, + "step": 8712 + }, + { + "epoch": 0.7857690399963927, + "grad_norm": 1.7899821886849394, + "learning_rate": 4.622929881680213e-07, + "loss": 1.0041, + "step": 8713 + }, + { + "epoch": 0.7858592235198629, + "grad_norm": 1.719605181091305, + "learning_rate": 4.6191949491534887e-07, + "loss": 0.902, + "step": 8714 + }, + { + "epoch": 0.7859494070433332, + "grad_norm": 1.787682079831768, + "learning_rate": 4.6154613290018617e-07, + "loss": 0.9961, + "step": 8715 + }, + { + "epoch": 0.7860395905668034, + "grad_norm": 1.8290475327169238, + "learning_rate": 4.6117290215439043e-07, + "loss": 0.93, + "step": 8716 + }, + { + "epoch": 0.7861297740902737, + "grad_norm": 1.653511374306121, + "learning_rate": 4.6079980270980744e-07, + "loss": 0.8437, + "step": 8717 + }, + { + "epoch": 0.786219957613744, + "grad_norm": 1.745202163309544, + "learning_rate": 4.6042683459827245e-07, + "loss": 0.9514, + "step": 8718 + }, + { + "epoch": 0.7863101411372142, + "grad_norm": 0.7043062460175954, + "learning_rate": 4.600539978516098e-07, + "loss": 0.8004, + "step": 8719 + }, + { + "epoch": 0.7864003246606845, + "grad_norm": 1.5325767063527693, + "learning_rate": 4.5968129250163004e-07, + "loss": 0.9438, + "step": 8720 + }, + { + "epoch": 0.7864905081841548, + "grad_norm": 2.1612992806494637, + "learning_rate": 4.5930871858013653e-07, + "loss": 0.9135, + "step": 8721 + }, + { + "epoch": 0.786580691707625, + "grad_norm": 0.6571220779687773, + "learning_rate": 4.589362761189182e-07, + "loss": 0.859, + "step": 8722 + }, + { + "epoch": 0.7866708752310952, + "grad_norm": 1.617932797506344, + "learning_rate": 4.585639651497539e-07, + "loss": 1.0489, + "step": 8723 + }, + { + "epoch": 0.7867610587545656, + "grad_norm": 1.5227086597894084, + "learning_rate": 4.581917857044115e-07, + "loss": 0.936, + "step": 8724 + }, + { + "epoch": 0.7868512422780358, + "grad_norm": 1.6401072820603781, + "learning_rate": 4.5781973781464734e-07, + "loss": 1.0165, + "step": 8725 + }, + { + "epoch": 0.7869414258015061, + "grad_norm": 1.4908849813357943, + "learning_rate": 4.574478215122073e-07, + "loss": 1.0176, + "step": 8726 + }, + { + "epoch": 0.7870316093249763, + "grad_norm": 1.8892285572958796, + "learning_rate": 4.5707603682882357e-07, + "loss": 0.9613, + "step": 8727 + }, + { + "epoch": 0.7871217928484466, + "grad_norm": 1.8514383008804705, + "learning_rate": 4.56704383796221e-07, + "loss": 0.8503, + "step": 8728 + }, + { + "epoch": 0.7872119763719169, + "grad_norm": 1.7479015439106405, + "learning_rate": 4.5633286244610956e-07, + "loss": 1.0198, + "step": 8729 + }, + { + "epoch": 0.7873021598953871, + "grad_norm": 1.4853518032698276, + "learning_rate": 4.5596147281018993e-07, + "loss": 1.0458, + "step": 8730 + }, + { + "epoch": 0.7873923434188573, + "grad_norm": 1.636339749176343, + "learning_rate": 4.5559021492015137e-07, + "loss": 1.0809, + "step": 8731 + }, + { + "epoch": 0.7874825269423277, + "grad_norm": 2.612237239929513, + "learning_rate": 4.552190888076712e-07, + "loss": 0.8888, + "step": 8732 + }, + { + "epoch": 0.7875727104657979, + "grad_norm": 1.9931498876999334, + "learning_rate": 4.548480945044164e-07, + "loss": 0.9297, + "step": 8733 + }, + { + "epoch": 0.7876628939892681, + "grad_norm": 2.1039686078315136, + "learning_rate": 4.54477232042042e-07, + "loss": 0.9775, + "step": 8734 + }, + { + "epoch": 0.7877530775127384, + "grad_norm": 1.4206445421250102, + "learning_rate": 4.541065014521921e-07, + "loss": 0.9764, + "step": 8735 + }, + { + "epoch": 0.7878432610362087, + "grad_norm": 2.4118311177117167, + "learning_rate": 4.5373590276649996e-07, + "loss": 0.9299, + "step": 8736 + }, + { + "epoch": 0.787933444559679, + "grad_norm": 3.0874793727664485, + "learning_rate": 4.533654360165862e-07, + "loss": 0.9493, + "step": 8737 + }, + { + "epoch": 0.7880236280831492, + "grad_norm": 1.9120588130324965, + "learning_rate": 4.5299510123406115e-07, + "loss": 1.0373, + "step": 8738 + }, + { + "epoch": 0.7881138116066194, + "grad_norm": 2.4685576158722444, + "learning_rate": 4.5262489845052456e-07, + "loss": 0.9473, + "step": 8739 + }, + { + "epoch": 0.7882039951300898, + "grad_norm": 1.7735377884122654, + "learning_rate": 4.5225482769756353e-07, + "loss": 0.9459, + "step": 8740 + }, + { + "epoch": 0.78829417865356, + "grad_norm": 2.24069759831203, + "learning_rate": 4.5188488900675545e-07, + "loss": 0.9121, + "step": 8741 + }, + { + "epoch": 0.7883843621770302, + "grad_norm": 1.5001830942942431, + "learning_rate": 4.5151508240966363e-07, + "loss": 1.0083, + "step": 8742 + }, + { + "epoch": 0.7884745457005006, + "grad_norm": 1.8111975363852795, + "learning_rate": 4.511454079378445e-07, + "loss": 1.0248, + "step": 8743 + }, + { + "epoch": 0.7885647292239708, + "grad_norm": 3.2089714472597937, + "learning_rate": 4.507758656228382e-07, + "loss": 0.8809, + "step": 8744 + }, + { + "epoch": 0.788654912747441, + "grad_norm": 1.6006730750094795, + "learning_rate": 4.5040645549617864e-07, + "loss": 1.0441, + "step": 8745 + }, + { + "epoch": 0.7887450962709113, + "grad_norm": 1.53596261858653, + "learning_rate": 4.5003717758938384e-07, + "loss": 0.8987, + "step": 8746 + }, + { + "epoch": 0.7888352797943816, + "grad_norm": 1.6010360969176876, + "learning_rate": 4.4966803193396365e-07, + "loss": 0.92, + "step": 8747 + }, + { + "epoch": 0.7889254633178518, + "grad_norm": 1.5112535021972464, + "learning_rate": 4.492990185614154e-07, + "loss": 0.9861, + "step": 8748 + }, + { + "epoch": 0.7890156468413221, + "grad_norm": 1.5075576645938973, + "learning_rate": 4.489301375032255e-07, + "loss": 0.9682, + "step": 8749 + }, + { + "epoch": 0.7891058303647923, + "grad_norm": 2.2036587765183646, + "learning_rate": 4.4856138879086857e-07, + "loss": 0.9854, + "step": 8750 + }, + { + "epoch": 0.7891960138882627, + "grad_norm": 1.3649746234337483, + "learning_rate": 4.481927724558092e-07, + "loss": 0.9306, + "step": 8751 + }, + { + "epoch": 0.7892861974117329, + "grad_norm": 1.928197123014822, + "learning_rate": 4.478242885294985e-07, + "loss": 0.9499, + "step": 8752 + }, + { + "epoch": 0.7893763809352031, + "grad_norm": 1.5613021836978196, + "learning_rate": 4.474559370433779e-07, + "loss": 0.9431, + "step": 8753 + }, + { + "epoch": 0.7894665644586734, + "grad_norm": 1.7607780702545137, + "learning_rate": 4.470877180288777e-07, + "loss": 0.925, + "step": 8754 + }, + { + "epoch": 0.7895567479821437, + "grad_norm": 1.6863358331980853, + "learning_rate": 4.4671963151741574e-07, + "loss": 0.9449, + "step": 8755 + }, + { + "epoch": 0.7896469315056139, + "grad_norm": 1.459847844569366, + "learning_rate": 4.4635167754039973e-07, + "loss": 0.9572, + "step": 8756 + }, + { + "epoch": 0.7897371150290842, + "grad_norm": 1.3974689134200893, + "learning_rate": 4.459838561292253e-07, + "loss": 1.006, + "step": 8757 + }, + { + "epoch": 0.7898272985525544, + "grad_norm": 1.6658151199594444, + "learning_rate": 4.456161673152774e-07, + "loss": 0.8966, + "step": 8758 + }, + { + "epoch": 0.7899174820760247, + "grad_norm": 1.435537491333064, + "learning_rate": 4.4524861112992806e-07, + "loss": 0.971, + "step": 8759 + }, + { + "epoch": 0.790007665599495, + "grad_norm": 2.133788392090856, + "learning_rate": 4.448811876045411e-07, + "loss": 0.978, + "step": 8760 + }, + { + "epoch": 0.7900978491229652, + "grad_norm": 3.730747647869655, + "learning_rate": 4.445138967704647e-07, + "loss": 1.0506, + "step": 8761 + }, + { + "epoch": 0.7901880326464354, + "grad_norm": 1.7023911601270894, + "learning_rate": 4.4414673865904075e-07, + "loss": 0.9522, + "step": 8762 + }, + { + "epoch": 0.7902782161699058, + "grad_norm": 1.3573002822536173, + "learning_rate": 4.437797133015955e-07, + "loss": 0.9974, + "step": 8763 + }, + { + "epoch": 0.790368399693376, + "grad_norm": 2.303748608742437, + "learning_rate": 4.4341282072944586e-07, + "loss": 0.8949, + "step": 8764 + }, + { + "epoch": 0.7904585832168463, + "grad_norm": 1.486557009263611, + "learning_rate": 4.430460609738973e-07, + "loss": 1.0014, + "step": 8765 + }, + { + "epoch": 0.7905487667403166, + "grad_norm": 0.7037422649642802, + "learning_rate": 4.4267943406624386e-07, + "loss": 0.8472, + "step": 8766 + }, + { + "epoch": 0.7906389502637868, + "grad_norm": 2.3104664908684622, + "learning_rate": 4.4231294003776853e-07, + "loss": 1.0902, + "step": 8767 + }, + { + "epoch": 0.7907291337872571, + "grad_norm": 1.9444721045873063, + "learning_rate": 4.419465789197416e-07, + "loss": 1.0218, + "step": 8768 + }, + { + "epoch": 0.7908193173107273, + "grad_norm": 1.5239201490303642, + "learning_rate": 4.415803507434237e-07, + "loss": 0.9725, + "step": 8769 + }, + { + "epoch": 0.7909095008341976, + "grad_norm": 1.5036120634798116, + "learning_rate": 4.4121425554006307e-07, + "loss": 1.0033, + "step": 8770 + }, + { + "epoch": 0.7909996843576679, + "grad_norm": 3.9420671164161565, + "learning_rate": 4.4084829334089744e-07, + "loss": 0.963, + "step": 8771 + }, + { + "epoch": 0.7910898678811381, + "grad_norm": 2.3967058224015103, + "learning_rate": 4.404824641771525e-07, + "loss": 0.9219, + "step": 8772 + }, + { + "epoch": 0.7911800514046083, + "grad_norm": 1.5047463188677745, + "learning_rate": 4.4011676808004327e-07, + "loss": 1.046, + "step": 8773 + }, + { + "epoch": 0.7912702349280787, + "grad_norm": 1.497104393060385, + "learning_rate": 4.3975120508077145e-07, + "loss": 0.9899, + "step": 8774 + }, + { + "epoch": 0.7913604184515489, + "grad_norm": 1.5409643360494092, + "learning_rate": 4.39385775210531e-07, + "loss": 1.0274, + "step": 8775 + }, + { + "epoch": 0.7914506019750192, + "grad_norm": 2.083002776306617, + "learning_rate": 4.390204785005003e-07, + "loss": 0.9945, + "step": 8776 + }, + { + "epoch": 0.7915407854984894, + "grad_norm": 1.3753547644280633, + "learning_rate": 4.386553149818504e-07, + "loss": 0.9852, + "step": 8777 + }, + { + "epoch": 0.7916309690219597, + "grad_norm": 2.3144366428738015, + "learning_rate": 4.3829028468573793e-07, + "loss": 0.849, + "step": 8778 + }, + { + "epoch": 0.79172115254543, + "grad_norm": 5.715097396968484, + "learning_rate": 4.3792538764330935e-07, + "loss": 0.9172, + "step": 8779 + }, + { + "epoch": 0.7918113360689002, + "grad_norm": 1.8060095102264668, + "learning_rate": 4.3756062388569994e-07, + "loss": 0.9521, + "step": 8780 + }, + { + "epoch": 0.7919015195923704, + "grad_norm": 0.7420680702891669, + "learning_rate": 4.3719599344403346e-07, + "loss": 0.8169, + "step": 8781 + }, + { + "epoch": 0.7919917031158408, + "grad_norm": 1.4435750054328402, + "learning_rate": 4.3683149634942243e-07, + "loss": 0.9613, + "step": 8782 + }, + { + "epoch": 0.792081886639311, + "grad_norm": 1.8535438591723061, + "learning_rate": 4.364671326329663e-07, + "loss": 0.8607, + "step": 8783 + }, + { + "epoch": 0.7921720701627812, + "grad_norm": 2.7196228555463184, + "learning_rate": 4.3610290232575673e-07, + "loss": 0.9971, + "step": 8784 + }, + { + "epoch": 0.7922622536862515, + "grad_norm": 1.704482770980993, + "learning_rate": 4.357388054588702e-07, + "loss": 0.9796, + "step": 8785 + }, + { + "epoch": 0.7923524372097218, + "grad_norm": 1.495980041804263, + "learning_rate": 4.3537484206337405e-07, + "loss": 0.921, + "step": 8786 + }, + { + "epoch": 0.792442620733192, + "grad_norm": 2.726080209656121, + "learning_rate": 4.3501101217032366e-07, + "loss": 1.0068, + "step": 8787 + }, + { + "epoch": 0.7925328042566623, + "grad_norm": 1.6403300428914989, + "learning_rate": 4.346473158107629e-07, + "loss": 0.9022, + "step": 8788 + }, + { + "epoch": 0.7926229877801325, + "grad_norm": 1.7201196241919199, + "learning_rate": 4.342837530157244e-07, + "loss": 0.9793, + "step": 8789 + }, + { + "epoch": 0.7927131713036029, + "grad_norm": 1.5407385824864575, + "learning_rate": 4.3392032381622987e-07, + "loss": 0.8822, + "step": 8790 + }, + { + "epoch": 0.7928033548270731, + "grad_norm": 1.3580678596536535, + "learning_rate": 4.3355702824328765e-07, + "loss": 0.9392, + "step": 8791 + }, + { + "epoch": 0.7928935383505433, + "grad_norm": 1.500784430433455, + "learning_rate": 4.3319386632789823e-07, + "loss": 1.0211, + "step": 8792 + }, + { + "epoch": 0.7929837218740137, + "grad_norm": 1.7553897013095825, + "learning_rate": 4.328308381010466e-07, + "loss": 0.973, + "step": 8793 + }, + { + "epoch": 0.7930739053974839, + "grad_norm": 1.5385358654920975, + "learning_rate": 4.3246794359370933e-07, + "loss": 0.9877, + "step": 8794 + }, + { + "epoch": 0.7931640889209541, + "grad_norm": 1.8475109470647058, + "learning_rate": 4.3210518283685025e-07, + "loss": 0.9538, + "step": 8795 + }, + { + "epoch": 0.7932542724444244, + "grad_norm": 1.7658176148263944, + "learning_rate": 4.317425558614225e-07, + "loss": 0.977, + "step": 8796 + }, + { + "epoch": 0.7933444559678947, + "grad_norm": 1.3411445642360373, + "learning_rate": 4.3138006269836744e-07, + "loss": 0.9405, + "step": 8797 + }, + { + "epoch": 0.793434639491365, + "grad_norm": 1.7143920442552338, + "learning_rate": 4.3101770337861376e-07, + "loss": 0.9687, + "step": 8798 + }, + { + "epoch": 0.7935248230148352, + "grad_norm": 0.6698511158578452, + "learning_rate": 4.30655477933082e-07, + "loss": 0.8305, + "step": 8799 + }, + { + "epoch": 0.7936150065383054, + "grad_norm": 1.3209593503185826, + "learning_rate": 4.30293386392677e-07, + "loss": 1.0012, + "step": 8800 + }, + { + "epoch": 0.7937051900617758, + "grad_norm": 1.3272819536091516, + "learning_rate": 4.299314287882967e-07, + "loss": 0.9963, + "step": 8801 + }, + { + "epoch": 0.793795373585246, + "grad_norm": 1.599258179357906, + "learning_rate": 4.2956960515082353e-07, + "loss": 1.012, + "step": 8802 + }, + { + "epoch": 0.7938855571087162, + "grad_norm": 1.6547108822360126, + "learning_rate": 4.29207915511131e-07, + "loss": 0.9387, + "step": 8803 + }, + { + "epoch": 0.7939757406321865, + "grad_norm": 1.6329671978105886, + "learning_rate": 4.2884635990008024e-07, + "loss": 1.0371, + "step": 8804 + }, + { + "epoch": 0.7940659241556568, + "grad_norm": 1.8002727063546466, + "learning_rate": 4.284849383485214e-07, + "loss": 0.943, + "step": 8805 + }, + { + "epoch": 0.794156107679127, + "grad_norm": 3.0582003870313614, + "learning_rate": 4.2812365088729296e-07, + "loss": 1.006, + "step": 8806 + }, + { + "epoch": 0.7942462912025973, + "grad_norm": 1.6186144149732298, + "learning_rate": 4.2776249754722227e-07, + "loss": 0.9972, + "step": 8807 + }, + { + "epoch": 0.7943364747260675, + "grad_norm": 5.212011315287371, + "learning_rate": 4.27401478359124e-07, + "loss": 1.0475, + "step": 8808 + }, + { + "epoch": 0.7944266582495378, + "grad_norm": 1.5690422424212158, + "learning_rate": 4.2704059335380283e-07, + "loss": 1.0776, + "step": 8809 + }, + { + "epoch": 0.7945168417730081, + "grad_norm": 1.5625536337230503, + "learning_rate": 4.266798425620515e-07, + "loss": 1.0806, + "step": 8810 + }, + { + "epoch": 0.7946070252964783, + "grad_norm": 1.5960894599068935, + "learning_rate": 4.263192260146511e-07, + "loss": 0.9627, + "step": 8811 + }, + { + "epoch": 0.7946972088199485, + "grad_norm": 6.594119961018508, + "learning_rate": 4.2595874374237216e-07, + "loss": 1.0237, + "step": 8812 + }, + { + "epoch": 0.7947873923434189, + "grad_norm": 1.40464317922293, + "learning_rate": 4.255983957759712e-07, + "loss": 0.9758, + "step": 8813 + }, + { + "epoch": 0.7948775758668891, + "grad_norm": 1.4769394005009397, + "learning_rate": 4.2523818214619745e-07, + "loss": 0.9168, + "step": 8814 + }, + { + "epoch": 0.7949677593903594, + "grad_norm": 3.622721909912498, + "learning_rate": 4.24878102883784e-07, + "loss": 1.0221, + "step": 8815 + }, + { + "epoch": 0.7950579429138297, + "grad_norm": 1.4178072414422591, + "learning_rate": 4.24518158019457e-07, + "loss": 0.9811, + "step": 8816 + }, + { + "epoch": 0.7951481264372999, + "grad_norm": 1.827239514950111, + "learning_rate": 4.241583475839274e-07, + "loss": 0.9182, + "step": 8817 + }, + { + "epoch": 0.7952383099607702, + "grad_norm": 1.647262009491638, + "learning_rate": 4.237986716078965e-07, + "loss": 1.0233, + "step": 8818 + }, + { + "epoch": 0.7953284934842404, + "grad_norm": 1.5908105708246079, + "learning_rate": 4.2343913012205433e-07, + "loss": 0.9653, + "step": 8819 + }, + { + "epoch": 0.7954186770077107, + "grad_norm": 1.3988135913646735, + "learning_rate": 4.230797231570784e-07, + "loss": 0.9534, + "step": 8820 + }, + { + "epoch": 0.795508860531181, + "grad_norm": 1.87133468956123, + "learning_rate": 4.227204507436357e-07, + "loss": 0.9786, + "step": 8821 + }, + { + "epoch": 0.7955990440546512, + "grad_norm": 2.1888913770468954, + "learning_rate": 4.223613129123811e-07, + "loss": 0.899, + "step": 8822 + }, + { + "epoch": 0.7956892275781214, + "grad_norm": 1.352355592037826, + "learning_rate": 4.220023096939589e-07, + "loss": 0.8855, + "step": 8823 + }, + { + "epoch": 0.7957794111015918, + "grad_norm": 1.863947143566579, + "learning_rate": 4.21643441119e-07, + "loss": 0.9712, + "step": 8824 + }, + { + "epoch": 0.795869594625062, + "grad_norm": 0.6374351356268753, + "learning_rate": 4.212847072181256e-07, + "loss": 0.867, + "step": 8825 + }, + { + "epoch": 0.7959597781485322, + "grad_norm": 2.0576367696401485, + "learning_rate": 4.2092610802194505e-07, + "loss": 0.9497, + "step": 8826 + }, + { + "epoch": 0.7960499616720025, + "grad_norm": 3.9734025252374665, + "learning_rate": 4.2056764356105587e-07, + "loss": 0.8857, + "step": 8827 + }, + { + "epoch": 0.7961401451954728, + "grad_norm": 1.6101323540111452, + "learning_rate": 4.202093138660443e-07, + "loss": 1.0259, + "step": 8828 + }, + { + "epoch": 0.7962303287189431, + "grad_norm": 1.425309027241242, + "learning_rate": 4.198511189674854e-07, + "loss": 1.0111, + "step": 8829 + }, + { + "epoch": 0.7963205122424133, + "grad_norm": 1.5077131960094063, + "learning_rate": 4.1949305889594066e-07, + "loss": 0.997, + "step": 8830 + }, + { + "epoch": 0.7964106957658835, + "grad_norm": 1.9052079901704502, + "learning_rate": 4.191351336819642e-07, + "loss": 0.9747, + "step": 8831 + }, + { + "epoch": 0.7965008792893539, + "grad_norm": 1.750450144502631, + "learning_rate": 4.187773433560939e-07, + "loss": 0.9175, + "step": 8832 + }, + { + "epoch": 0.7965910628128241, + "grad_norm": 1.6603427198816483, + "learning_rate": 4.184196879488604e-07, + "loss": 0.9711, + "step": 8833 + }, + { + "epoch": 0.7966812463362943, + "grad_norm": 3.865560046600874, + "learning_rate": 4.1806216749077936e-07, + "loss": 1.0405, + "step": 8834 + }, + { + "epoch": 0.7967714298597646, + "grad_norm": 1.5182845387467838, + "learning_rate": 4.177047820123569e-07, + "loss": 1.0206, + "step": 8835 + }, + { + "epoch": 0.7968616133832349, + "grad_norm": 1.8154386178505473, + "learning_rate": 4.1734753154408733e-07, + "loss": 0.8489, + "step": 8836 + }, + { + "epoch": 0.7969517969067051, + "grad_norm": 1.6578337097249356, + "learning_rate": 4.169904161164528e-07, + "loss": 0.9599, + "step": 8837 + }, + { + "epoch": 0.7970419804301754, + "grad_norm": 2.4300153210631836, + "learning_rate": 4.1663343575992526e-07, + "loss": 1.0975, + "step": 8838 + }, + { + "epoch": 0.7971321639536457, + "grad_norm": 2.207978226042931, + "learning_rate": 4.1627659050496275e-07, + "loss": 0.9934, + "step": 8839 + }, + { + "epoch": 0.797222347477116, + "grad_norm": 1.5299309479333818, + "learning_rate": 4.1591988038201453e-07, + "loss": 0.8946, + "step": 8840 + }, + { + "epoch": 0.7973125310005862, + "grad_norm": 1.4238149801937388, + "learning_rate": 4.155633054215164e-07, + "loss": 1.0208, + "step": 8841 + }, + { + "epoch": 0.7974027145240564, + "grad_norm": 2.01160935780398, + "learning_rate": 4.152068656538934e-07, + "loss": 0.8852, + "step": 8842 + }, + { + "epoch": 0.7974928980475268, + "grad_norm": 1.3833302105251746, + "learning_rate": 4.148505611095594e-07, + "loss": 1.0341, + "step": 8843 + }, + { + "epoch": 0.797583081570997, + "grad_norm": 1.6882470914323342, + "learning_rate": 4.1449439181891563e-07, + "loss": 0.8548, + "step": 8844 + }, + { + "epoch": 0.7976732650944672, + "grad_norm": 1.8304962038919306, + "learning_rate": 4.14138357812353e-07, + "loss": 0.9935, + "step": 8845 + }, + { + "epoch": 0.7977634486179375, + "grad_norm": 1.9881145654744496, + "learning_rate": 4.137824591202506e-07, + "loss": 1.0329, + "step": 8846 + }, + { + "epoch": 0.7978536321414078, + "grad_norm": 1.2721768557600563, + "learning_rate": 4.134266957729737e-07, + "loss": 0.9665, + "step": 8847 + }, + { + "epoch": 0.797943815664878, + "grad_norm": 1.4661147643389054, + "learning_rate": 4.1307106780088065e-07, + "loss": 1.0705, + "step": 8848 + }, + { + "epoch": 0.7980339991883483, + "grad_norm": 1.493025220025585, + "learning_rate": 4.1271557523431387e-07, + "loss": 0.9874, + "step": 8849 + }, + { + "epoch": 0.7981241827118185, + "grad_norm": 1.8948229378385932, + "learning_rate": 4.1236021810360634e-07, + "loss": 0.9602, + "step": 8850 + }, + { + "epoch": 0.7982143662352889, + "grad_norm": 1.9014876865787729, + "learning_rate": 4.120049964390793e-07, + "loss": 0.9229, + "step": 8851 + }, + { + "epoch": 0.7983045497587591, + "grad_norm": 2.3817522535524653, + "learning_rate": 4.116499102710418e-07, + "loss": 0.9278, + "step": 8852 + }, + { + "epoch": 0.7983947332822293, + "grad_norm": 2.1360430163291757, + "learning_rate": 4.112949596297928e-07, + "loss": 0.9535, + "step": 8853 + }, + { + "epoch": 0.7984849168056996, + "grad_norm": 1.7673950100462135, + "learning_rate": 4.1094014454561664e-07, + "loss": 0.9189, + "step": 8854 + }, + { + "epoch": 0.7985751003291699, + "grad_norm": 1.9448504531167436, + "learning_rate": 4.1058546504879057e-07, + "loss": 0.8829, + "step": 8855 + }, + { + "epoch": 0.7986652838526401, + "grad_norm": 1.7617904073347745, + "learning_rate": 4.1023092116957583e-07, + "loss": 1.0344, + "step": 8856 + }, + { + "epoch": 0.7987554673761104, + "grad_norm": 1.4611905944600871, + "learning_rate": 4.098765129382249e-07, + "loss": 1.0177, + "step": 8857 + }, + { + "epoch": 0.7988456508995806, + "grad_norm": 1.7946556649364458, + "learning_rate": 4.0952224038497764e-07, + "loss": 1.0484, + "step": 8858 + }, + { + "epoch": 0.7989358344230509, + "grad_norm": 1.857331507396873, + "learning_rate": 4.091681035400627e-07, + "loss": 0.9717, + "step": 8859 + }, + { + "epoch": 0.7990260179465212, + "grad_norm": 1.7091245757223594, + "learning_rate": 4.088141024336971e-07, + "loss": 1.0113, + "step": 8860 + }, + { + "epoch": 0.7991162014699914, + "grad_norm": 1.4792584225129704, + "learning_rate": 4.0846023709608636e-07, + "loss": 1.0555, + "step": 8861 + }, + { + "epoch": 0.7992063849934617, + "grad_norm": 1.71740100980671, + "learning_rate": 4.081065075574226e-07, + "loss": 1.0548, + "step": 8862 + }, + { + "epoch": 0.799296568516932, + "grad_norm": 1.9015261167808564, + "learning_rate": 4.077529138478906e-07, + "loss": 0.9089, + "step": 8863 + }, + { + "epoch": 0.7993867520404022, + "grad_norm": 0.6757357808069653, + "learning_rate": 4.073994559976588e-07, + "loss": 0.7783, + "step": 8864 + }, + { + "epoch": 0.7994769355638724, + "grad_norm": 2.3311750670961904, + "learning_rate": 4.0704613403688716e-07, + "loss": 0.9385, + "step": 8865 + }, + { + "epoch": 0.7995671190873428, + "grad_norm": 1.3640090180319862, + "learning_rate": 4.0669294799572264e-07, + "loss": 1.0348, + "step": 8866 + }, + { + "epoch": 0.799657302610813, + "grad_norm": 1.8763891160639956, + "learning_rate": 4.0633989790430113e-07, + "loss": 1.0767, + "step": 8867 + }, + { + "epoch": 0.7997474861342833, + "grad_norm": 1.8382104317722823, + "learning_rate": 4.059869837927477e-07, + "loss": 0.9869, + "step": 8868 + }, + { + "epoch": 0.7998376696577535, + "grad_norm": 1.8504588382928668, + "learning_rate": 4.056342056911728e-07, + "loss": 0.9636, + "step": 8869 + }, + { + "epoch": 0.7999278531812238, + "grad_norm": 2.366510174730819, + "learning_rate": 4.052815636296798e-07, + "loss": 0.9934, + "step": 8870 + }, + { + "epoch": 0.8000180367046941, + "grad_norm": 1.4147127458406596, + "learning_rate": 4.0492905763835593e-07, + "loss": 1.012, + "step": 8871 + }, + { + "epoch": 0.8001082202281643, + "grad_norm": 2.0618250349315312, + "learning_rate": 4.0457668774728115e-07, + "loss": 1.0105, + "step": 8872 + }, + { + "epoch": 0.8001984037516345, + "grad_norm": 1.7163695062759576, + "learning_rate": 4.0422445398651985e-07, + "loss": 0.9955, + "step": 8873 + }, + { + "epoch": 0.8002885872751049, + "grad_norm": 1.9962858044226666, + "learning_rate": 4.0387235638612706e-07, + "loss": 0.8976, + "step": 8874 + }, + { + "epoch": 0.8003787707985751, + "grad_norm": 2.1807151204207957, + "learning_rate": 4.0352039497614586e-07, + "loss": 0.983, + "step": 8875 + }, + { + "epoch": 0.8004689543220453, + "grad_norm": 2.686159109900676, + "learning_rate": 4.031685697866074e-07, + "loss": 0.9687, + "step": 8876 + }, + { + "epoch": 0.8005591378455156, + "grad_norm": 1.4654628411810537, + "learning_rate": 4.0281688084753165e-07, + "loss": 1.0157, + "step": 8877 + }, + { + "epoch": 0.8006493213689859, + "grad_norm": 1.6135151026266903, + "learning_rate": 4.0246532818892675e-07, + "loss": 1.0282, + "step": 8878 + }, + { + "epoch": 0.8007395048924562, + "grad_norm": 1.8644679866752316, + "learning_rate": 4.0211391184078814e-07, + "loss": 1.0158, + "step": 8879 + }, + { + "epoch": 0.8008296884159264, + "grad_norm": 1.59979567415206, + "learning_rate": 4.0176263183310135e-07, + "loss": 1.0139, + "step": 8880 + }, + { + "epoch": 0.8009198719393966, + "grad_norm": 1.4191699444204138, + "learning_rate": 4.0141148819583925e-07, + "loss": 1.0367, + "step": 8881 + }, + { + "epoch": 0.801010055462867, + "grad_norm": 0.8332091040202251, + "learning_rate": 4.010604809589637e-07, + "loss": 0.8074, + "step": 8882 + }, + { + "epoch": 0.8011002389863372, + "grad_norm": 1.6350245481508925, + "learning_rate": 4.0070961015242475e-07, + "loss": 0.9822, + "step": 8883 + }, + { + "epoch": 0.8011904225098074, + "grad_norm": 1.4715618245821358, + "learning_rate": 4.0035887580615933e-07, + "loss": 1.0485, + "step": 8884 + }, + { + "epoch": 0.8012806060332777, + "grad_norm": 1.4558187967821858, + "learning_rate": 4.0000827795009594e-07, + "loss": 1.0295, + "step": 8885 + }, + { + "epoch": 0.801370789556748, + "grad_norm": 2.7075996187574716, + "learning_rate": 3.996578166141475e-07, + "loss": 0.9387, + "step": 8886 + }, + { + "epoch": 0.8014609730802182, + "grad_norm": 1.7185078537034262, + "learning_rate": 3.9930749182821955e-07, + "loss": 0.985, + "step": 8887 + }, + { + "epoch": 0.8015511566036885, + "grad_norm": 0.6004065765267197, + "learning_rate": 3.9895730362220116e-07, + "loss": 0.8318, + "step": 8888 + }, + { + "epoch": 0.8016413401271588, + "grad_norm": 2.627496259053305, + "learning_rate": 3.986072520259749e-07, + "loss": 1.0209, + "step": 8889 + }, + { + "epoch": 0.801731523650629, + "grad_norm": 1.545117622638709, + "learning_rate": 3.9825733706940736e-07, + "loss": 0.9661, + "step": 8890 + }, + { + "epoch": 0.8018217071740993, + "grad_norm": 1.376187332070845, + "learning_rate": 3.979075587823557e-07, + "loss": 0.9968, + "step": 8891 + }, + { + "epoch": 0.8019118906975695, + "grad_norm": 2.045867672053883, + "learning_rate": 3.9755791719466504e-07, + "loss": 0.8555, + "step": 8892 + }, + { + "epoch": 0.8020020742210399, + "grad_norm": 1.7657819695028634, + "learning_rate": 3.9720841233616875e-07, + "loss": 0.9377, + "step": 8893 + }, + { + "epoch": 0.8020922577445101, + "grad_norm": 1.5932768325018671, + "learning_rate": 3.968590442366888e-07, + "loss": 0.9586, + "step": 8894 + }, + { + "epoch": 0.8021824412679803, + "grad_norm": 1.4978448167032947, + "learning_rate": 3.9650981292603423e-07, + "loss": 0.9582, + "step": 8895 + }, + { + "epoch": 0.8022726247914506, + "grad_norm": 1.9902752843064617, + "learning_rate": 3.961607184340041e-07, + "loss": 0.9555, + "step": 8896 + }, + { + "epoch": 0.8023628083149209, + "grad_norm": 1.36175727373915, + "learning_rate": 3.9581176079038505e-07, + "loss": 0.8793, + "step": 8897 + }, + { + "epoch": 0.8024529918383911, + "grad_norm": 1.3276848905297352, + "learning_rate": 3.954629400249516e-07, + "loss": 0.9276, + "step": 8898 + }, + { + "epoch": 0.8025431753618614, + "grad_norm": 1.4467348265668762, + "learning_rate": 3.9511425616746787e-07, + "loss": 1.048, + "step": 8899 + }, + { + "epoch": 0.8026333588853316, + "grad_norm": 1.4582126930836474, + "learning_rate": 3.947657092476853e-07, + "loss": 1.0436, + "step": 8900 + }, + { + "epoch": 0.802723542408802, + "grad_norm": 2.38932056617448, + "learning_rate": 3.944172992953425e-07, + "loss": 0.992, + "step": 8901 + }, + { + "epoch": 0.8028137259322722, + "grad_norm": 1.3415423600244345, + "learning_rate": 3.9406902634017e-07, + "loss": 0.9379, + "step": 8902 + }, + { + "epoch": 0.8029039094557424, + "grad_norm": 1.50649658653969, + "learning_rate": 3.9372089041188227e-07, + "loss": 1.0146, + "step": 8903 + }, + { + "epoch": 0.8029940929792126, + "grad_norm": 1.7822032853788141, + "learning_rate": 3.9337289154018593e-07, + "loss": 0.8913, + "step": 8904 + }, + { + "epoch": 0.803084276502683, + "grad_norm": 1.5087860601203946, + "learning_rate": 3.930250297547728e-07, + "loss": 0.9805, + "step": 8905 + }, + { + "epoch": 0.8031744600261532, + "grad_norm": 2.2242417404331873, + "learning_rate": 3.9267730508532513e-07, + "loss": 0.859, + "step": 8906 + }, + { + "epoch": 0.8032646435496235, + "grad_norm": 1.686793073295029, + "learning_rate": 3.923297175615121e-07, + "loss": 0.9693, + "step": 8907 + }, + { + "epoch": 0.8033548270730937, + "grad_norm": 1.3286624045580389, + "learning_rate": 3.9198226721299243e-07, + "loss": 1.0215, + "step": 8908 + }, + { + "epoch": 0.803445010596564, + "grad_norm": 1.97407936657839, + "learning_rate": 3.916349540694128e-07, + "loss": 0.9598, + "step": 8909 + }, + { + "epoch": 0.8035351941200343, + "grad_norm": 1.93928789892879, + "learning_rate": 3.912877781604063e-07, + "loss": 0.9922, + "step": 8910 + }, + { + "epoch": 0.8036253776435045, + "grad_norm": 1.6797829534333124, + "learning_rate": 3.909407395155977e-07, + "loss": 1.0067, + "step": 8911 + }, + { + "epoch": 0.8037155611669748, + "grad_norm": 1.7145143509712246, + "learning_rate": 3.9059383816459725e-07, + "loss": 0.9242, + "step": 8912 + }, + { + "epoch": 0.8038057446904451, + "grad_norm": 1.6699773411155845, + "learning_rate": 3.902470741370045e-07, + "loss": 1.042, + "step": 8913 + }, + { + "epoch": 0.8038959282139153, + "grad_norm": 1.4518247960738488, + "learning_rate": 3.8990044746240746e-07, + "loss": 0.9689, + "step": 8914 + }, + { + "epoch": 0.8039861117373855, + "grad_norm": 2.0120098014577636, + "learning_rate": 3.8955395817038237e-07, + "loss": 1.0778, + "step": 8915 + }, + { + "epoch": 0.8040762952608559, + "grad_norm": 1.9722009718859004, + "learning_rate": 3.892076062904934e-07, + "loss": 0.9454, + "step": 8916 + }, + { + "epoch": 0.8041664787843261, + "grad_norm": 1.9839758884633725, + "learning_rate": 3.8886139185229384e-07, + "loss": 0.9406, + "step": 8917 + }, + { + "epoch": 0.8042566623077964, + "grad_norm": 1.6240301172207412, + "learning_rate": 3.8851531488532284e-07, + "loss": 0.8954, + "step": 8918 + }, + { + "epoch": 0.8043468458312666, + "grad_norm": 1.7055100006552895, + "learning_rate": 3.88169375419112e-07, + "loss": 1.0027, + "step": 8919 + }, + { + "epoch": 0.8044370293547369, + "grad_norm": 2.5044584572930413, + "learning_rate": 3.8782357348317717e-07, + "loss": 0.9509, + "step": 8920 + }, + { + "epoch": 0.8045272128782072, + "grad_norm": 1.7847759751645746, + "learning_rate": 3.8747790910702437e-07, + "loss": 1.0259, + "step": 8921 + }, + { + "epoch": 0.8046173964016774, + "grad_norm": 1.5480531771789299, + "learning_rate": 3.8713238232014776e-07, + "loss": 0.9811, + "step": 8922 + }, + { + "epoch": 0.8047075799251476, + "grad_norm": 1.5463992071053987, + "learning_rate": 3.867869931520296e-07, + "loss": 1.0432, + "step": 8923 + }, + { + "epoch": 0.804797763448618, + "grad_norm": 1.5808520031062672, + "learning_rate": 3.864417416321406e-07, + "loss": 0.9943, + "step": 8924 + }, + { + "epoch": 0.8048879469720882, + "grad_norm": 1.866385123898559, + "learning_rate": 3.8609662778993847e-07, + "loss": 0.9606, + "step": 8925 + }, + { + "epoch": 0.8049781304955584, + "grad_norm": 0.5768943524240109, + "learning_rate": 3.85751651654872e-07, + "loss": 0.7949, + "step": 8926 + }, + { + "epoch": 0.8050683140190287, + "grad_norm": 1.7126630329575134, + "learning_rate": 3.8540681325637505e-07, + "loss": 1.0906, + "step": 8927 + }, + { + "epoch": 0.805158497542499, + "grad_norm": 1.3991141360571793, + "learning_rate": 3.8506211262387155e-07, + "loss": 0.9901, + "step": 8928 + }, + { + "epoch": 0.8052486810659693, + "grad_norm": 1.785621498967024, + "learning_rate": 3.847175497867732e-07, + "loss": 0.992, + "step": 8929 + }, + { + "epoch": 0.8053388645894395, + "grad_norm": 1.9028508883103032, + "learning_rate": 3.843731247744801e-07, + "loss": 0.9327, + "step": 8930 + }, + { + "epoch": 0.8054290481129097, + "grad_norm": 1.6727314119231236, + "learning_rate": 3.8402883761638047e-07, + "loss": 1.0242, + "step": 8931 + }, + { + "epoch": 0.8055192316363801, + "grad_norm": 0.6642831547969081, + "learning_rate": 3.8368468834185076e-07, + "loss": 0.8507, + "step": 8932 + }, + { + "epoch": 0.8056094151598503, + "grad_norm": 2.5825212391721286, + "learning_rate": 3.8334067698025583e-07, + "loss": 1.0288, + "step": 8933 + }, + { + "epoch": 0.8056995986833205, + "grad_norm": 1.7663278446739736, + "learning_rate": 3.8299680356094897e-07, + "loss": 0.9181, + "step": 8934 + }, + { + "epoch": 0.8057897822067909, + "grad_norm": 2.0367502717101877, + "learning_rate": 3.8265306811327024e-07, + "loss": 0.8685, + "step": 8935 + }, + { + "epoch": 0.8058799657302611, + "grad_norm": 1.6698812557879696, + "learning_rate": 3.8230947066654994e-07, + "loss": 0.9549, + "step": 8936 + }, + { + "epoch": 0.8059701492537313, + "grad_norm": 1.5969656871866833, + "learning_rate": 3.819660112501053e-07, + "loss": 1.0415, + "step": 8937 + }, + { + "epoch": 0.8060603327772016, + "grad_norm": 0.6260728211566544, + "learning_rate": 3.816226898932422e-07, + "loss": 0.8349, + "step": 8938 + }, + { + "epoch": 0.8061505163006719, + "grad_norm": 2.4580624210200814, + "learning_rate": 3.812795066252557e-07, + "loss": 0.9711, + "step": 8939 + }, + { + "epoch": 0.8062406998241421, + "grad_norm": 1.9594035437054744, + "learning_rate": 3.8093646147542577e-07, + "loss": 0.9941, + "step": 8940 + }, + { + "epoch": 0.8063308833476124, + "grad_norm": 2.24182700206549, + "learning_rate": 3.805935544730259e-07, + "loss": 0.9495, + "step": 8941 + }, + { + "epoch": 0.8064210668710826, + "grad_norm": 1.4084028083924947, + "learning_rate": 3.802507856473118e-07, + "loss": 0.9181, + "step": 8942 + }, + { + "epoch": 0.806511250394553, + "grad_norm": 2.088811601681232, + "learning_rate": 3.7990815502753317e-07, + "loss": 0.988, + "step": 8943 + }, + { + "epoch": 0.8066014339180232, + "grad_norm": 1.5692012517801077, + "learning_rate": 3.795656626429231e-07, + "loss": 0.988, + "step": 8944 + }, + { + "epoch": 0.8066916174414934, + "grad_norm": 1.84175428581956, + "learning_rate": 3.792233085227059e-07, + "loss": 0.9006, + "step": 8945 + }, + { + "epoch": 0.8067818009649637, + "grad_norm": 1.6969573308315575, + "learning_rate": 3.788810926960928e-07, + "loss": 0.9924, + "step": 8946 + }, + { + "epoch": 0.806871984488434, + "grad_norm": 1.5777877598358692, + "learning_rate": 3.785390151922836e-07, + "loss": 0.9659, + "step": 8947 + }, + { + "epoch": 0.8069621680119042, + "grad_norm": 2.6808076080077705, + "learning_rate": 3.781970760404665e-07, + "loss": 0.9688, + "step": 8948 + }, + { + "epoch": 0.8070523515353745, + "grad_norm": 1.5193557441774828, + "learning_rate": 3.778552752698176e-07, + "loss": 1.0598, + "step": 8949 + }, + { + "epoch": 0.8071425350588447, + "grad_norm": 1.4635142892831328, + "learning_rate": 3.775136129095007e-07, + "loss": 0.893, + "step": 8950 + }, + { + "epoch": 0.807232718582315, + "grad_norm": 1.4722329095336582, + "learning_rate": 3.771720889886685e-07, + "loss": 0.9556, + "step": 8951 + }, + { + "epoch": 0.8073229021057853, + "grad_norm": 0.6962679609656434, + "learning_rate": 3.7683070353646194e-07, + "loss": 0.833, + "step": 8952 + }, + { + "epoch": 0.8074130856292555, + "grad_norm": 1.7375019512577214, + "learning_rate": 3.7648945658200983e-07, + "loss": 0.9471, + "step": 8953 + }, + { + "epoch": 0.8075032691527257, + "grad_norm": 1.713904051642356, + "learning_rate": 3.761483481544292e-07, + "loss": 0.8721, + "step": 8954 + }, + { + "epoch": 0.8075934526761961, + "grad_norm": 2.2772360198776656, + "learning_rate": 3.7580737828282525e-07, + "loss": 0.9522, + "step": 8955 + }, + { + "epoch": 0.8076836361996663, + "grad_norm": 1.4575925671347112, + "learning_rate": 3.754665469962921e-07, + "loss": 0.9598, + "step": 8956 + }, + { + "epoch": 0.8077738197231366, + "grad_norm": 1.6845693925495677, + "learning_rate": 3.7512585432390973e-07, + "loss": 0.9378, + "step": 8957 + }, + { + "epoch": 0.8078640032466069, + "grad_norm": 1.6570473676920114, + "learning_rate": 3.7478530029474987e-07, + "loss": 0.9184, + "step": 8958 + }, + { + "epoch": 0.8079541867700771, + "grad_norm": 0.6275147154021078, + "learning_rate": 3.7444488493786854e-07, + "loss": 0.8649, + "step": 8959 + }, + { + "epoch": 0.8080443702935474, + "grad_norm": 1.7733439282170114, + "learning_rate": 3.7410460828231405e-07, + "loss": 1.0337, + "step": 8960 + }, + { + "epoch": 0.8081345538170176, + "grad_norm": 1.3126324632194724, + "learning_rate": 3.737644703571188e-07, + "loss": 0.9172, + "step": 8961 + }, + { + "epoch": 0.8082247373404879, + "grad_norm": 1.7204059599696226, + "learning_rate": 3.734244711913059e-07, + "loss": 1.0502, + "step": 8962 + }, + { + "epoch": 0.8083149208639582, + "grad_norm": 1.4796443848335512, + "learning_rate": 3.7308461081388584e-07, + "loss": 0.9596, + "step": 8963 + }, + { + "epoch": 0.8084051043874284, + "grad_norm": 1.7833967156005541, + "learning_rate": 3.727448892538576e-07, + "loss": 0.9152, + "step": 8964 + }, + { + "epoch": 0.8084952879108986, + "grad_norm": 1.4957355438967148, + "learning_rate": 3.724053065402086e-07, + "loss": 1.0611, + "step": 8965 + }, + { + "epoch": 0.808585471434369, + "grad_norm": 5.297525821639062, + "learning_rate": 3.7206586270191285e-07, + "loss": 0.9477, + "step": 8966 + }, + { + "epoch": 0.8086756549578392, + "grad_norm": 1.8834685136448337, + "learning_rate": 3.7172655776793385e-07, + "loss": 1.028, + "step": 8967 + }, + { + "epoch": 0.8087658384813095, + "grad_norm": 1.5737279690146528, + "learning_rate": 3.7138739176722323e-07, + "loss": 0.8731, + "step": 8968 + }, + { + "epoch": 0.8088560220047797, + "grad_norm": 2.1503798703679182, + "learning_rate": 3.710483647287206e-07, + "loss": 0.9494, + "step": 8969 + }, + { + "epoch": 0.80894620552825, + "grad_norm": 1.6298930520386865, + "learning_rate": 3.707094766813532e-07, + "loss": 0.8806, + "step": 8970 + }, + { + "epoch": 0.8090363890517203, + "grad_norm": 1.8540166372686413, + "learning_rate": 3.7037072765403754e-07, + "loss": 0.9494, + "step": 8971 + }, + { + "epoch": 0.8091265725751905, + "grad_norm": 1.9807773928940615, + "learning_rate": 3.700321176756762e-07, + "loss": 0.9343, + "step": 8972 + }, + { + "epoch": 0.8092167560986607, + "grad_norm": 1.5958103998327482, + "learning_rate": 3.69693646775163e-07, + "loss": 1.0296, + "step": 8973 + }, + { + "epoch": 0.8093069396221311, + "grad_norm": 2.116026162185897, + "learning_rate": 3.693553149813764e-07, + "loss": 0.9625, + "step": 8974 + }, + { + "epoch": 0.8093971231456013, + "grad_norm": 0.5863991254073356, + "learning_rate": 3.690171223231866e-07, + "loss": 0.7888, + "step": 8975 + }, + { + "epoch": 0.8094873066690715, + "grad_norm": 1.3340452747211387, + "learning_rate": 3.6867906882944854e-07, + "loss": 1.0747, + "step": 8976 + }, + { + "epoch": 0.8095774901925418, + "grad_norm": 1.3339761535480548, + "learning_rate": 3.6834115452900737e-07, + "loss": 0.9819, + "step": 8977 + }, + { + "epoch": 0.8096676737160121, + "grad_norm": 1.4068408148809577, + "learning_rate": 3.680033794506958e-07, + "loss": 0.8783, + "step": 8978 + }, + { + "epoch": 0.8097578572394823, + "grad_norm": 2.431236533787166, + "learning_rate": 3.676657436233346e-07, + "loss": 0.8089, + "step": 8979 + }, + { + "epoch": 0.8098480407629526, + "grad_norm": 1.7629208407273358, + "learning_rate": 3.6732824707573305e-07, + "loss": 0.9773, + "step": 8980 + }, + { + "epoch": 0.8099382242864229, + "grad_norm": 1.7070332751927704, + "learning_rate": 3.6699088983668716e-07, + "loss": 1.0127, + "step": 8981 + }, + { + "epoch": 0.8100284078098932, + "grad_norm": 2.5154955338992147, + "learning_rate": 3.6665367193498376e-07, + "loss": 0.9362, + "step": 8982 + }, + { + "epoch": 0.8101185913333634, + "grad_norm": 2.3362886809141603, + "learning_rate": 3.663165933993948e-07, + "loss": 0.9766, + "step": 8983 + }, + { + "epoch": 0.8102087748568336, + "grad_norm": 1.8915471121809517, + "learning_rate": 3.659796542586822e-07, + "loss": 1.0018, + "step": 8984 + }, + { + "epoch": 0.810298958380304, + "grad_norm": 1.5678607532845714, + "learning_rate": 3.6564285454159526e-07, + "loss": 0.9365, + "step": 8985 + }, + { + "epoch": 0.8103891419037742, + "grad_norm": 1.5466215908673053, + "learning_rate": 3.653061942768718e-07, + "loss": 1.0231, + "step": 8986 + }, + { + "epoch": 0.8104793254272444, + "grad_norm": 2.029500471958539, + "learning_rate": 3.649696734932375e-07, + "loss": 0.9338, + "step": 8987 + }, + { + "epoch": 0.8105695089507147, + "grad_norm": 1.8197173754312874, + "learning_rate": 3.646332922194064e-07, + "loss": 0.9131, + "step": 8988 + }, + { + "epoch": 0.810659692474185, + "grad_norm": 1.8763138935918389, + "learning_rate": 3.6429705048407943e-07, + "loss": 0.964, + "step": 8989 + }, + { + "epoch": 0.8107498759976552, + "grad_norm": 2.328446436460285, + "learning_rate": 3.6396094831594804e-07, + "loss": 0.9689, + "step": 8990 + }, + { + "epoch": 0.8108400595211255, + "grad_norm": 1.961035129502857, + "learning_rate": 3.6362498574368926e-07, + "loss": 0.9296, + "step": 8991 + }, + { + "epoch": 0.8109302430445957, + "grad_norm": 0.6946845701759048, + "learning_rate": 3.6328916279596935e-07, + "loss": 0.8605, + "step": 8992 + }, + { + "epoch": 0.811020426568066, + "grad_norm": 1.7288753846348537, + "learning_rate": 3.6295347950144305e-07, + "loss": 1.0237, + "step": 8993 + }, + { + "epoch": 0.8111106100915363, + "grad_norm": 1.5242730612490774, + "learning_rate": 3.626179358887522e-07, + "loss": 0.8308, + "step": 8994 + }, + { + "epoch": 0.8112007936150065, + "grad_norm": 2.066617727725679, + "learning_rate": 3.6228253198652816e-07, + "loss": 1.0347, + "step": 8995 + }, + { + "epoch": 0.8112909771384768, + "grad_norm": 1.7634788906087537, + "learning_rate": 3.6194726782338767e-07, + "loss": 0.9291, + "step": 8996 + }, + { + "epoch": 0.8113811606619471, + "grad_norm": 2.0546684699392026, + "learning_rate": 3.6161214342793953e-07, + "loss": 0.9535, + "step": 8997 + }, + { + "epoch": 0.8114713441854173, + "grad_norm": 1.5971092276294843, + "learning_rate": 3.612771588287764e-07, + "loss": 1.0044, + "step": 8998 + }, + { + "epoch": 0.8115615277088876, + "grad_norm": 1.992474583902836, + "learning_rate": 3.609423140544827e-07, + "loss": 1.0439, + "step": 8999 + }, + { + "epoch": 0.8116517112323578, + "grad_norm": 2.257066966586613, + "learning_rate": 3.6060760913362787e-07, + "loss": 0.9916, + "step": 9000 + }, + { + "epoch": 0.8117418947558281, + "grad_norm": 1.4674824866008536, + "learning_rate": 3.6027304409477146e-07, + "loss": 0.9872, + "step": 9001 + }, + { + "epoch": 0.8118320782792984, + "grad_norm": 1.4373913599508579, + "learning_rate": 3.599386189664604e-07, + "loss": 0.9502, + "step": 9002 + }, + { + "epoch": 0.8119222618027686, + "grad_norm": 1.722013738619487, + "learning_rate": 3.5960433377722945e-07, + "loss": 0.9687, + "step": 9003 + }, + { + "epoch": 0.8120124453262388, + "grad_norm": 1.6135170235505654, + "learning_rate": 3.5927018855560174e-07, + "loss": 1.0326, + "step": 9004 + }, + { + "epoch": 0.8121026288497092, + "grad_norm": 1.4837218604436362, + "learning_rate": 3.5893618333008904e-07, + "loss": 0.9717, + "step": 9005 + }, + { + "epoch": 0.8121928123731794, + "grad_norm": 1.623571648438567, + "learning_rate": 3.586023181291893e-07, + "loss": 1.0712, + "step": 9006 + }, + { + "epoch": 0.8122829958966497, + "grad_norm": 0.7560549539670521, + "learning_rate": 3.5826859298139044e-07, + "loss": 0.8486, + "step": 9007 + }, + { + "epoch": 0.81237317942012, + "grad_norm": 1.7320863211707482, + "learning_rate": 3.5793500791516773e-07, + "loss": 0.9607, + "step": 9008 + }, + { + "epoch": 0.8124633629435902, + "grad_norm": 3.066852823771591, + "learning_rate": 3.5760156295898415e-07, + "loss": 0.9501, + "step": 9009 + }, + { + "epoch": 0.8125535464670605, + "grad_norm": 2.2930454845683537, + "learning_rate": 3.5726825814129203e-07, + "loss": 0.9758, + "step": 9010 + }, + { + "epoch": 0.8126437299905307, + "grad_norm": 0.7048750608741503, + "learning_rate": 3.5693509349052886e-07, + "loss": 0.8508, + "step": 9011 + }, + { + "epoch": 0.812733913514001, + "grad_norm": 1.5741474902208585, + "learning_rate": 3.5660206903512433e-07, + "loss": 1.0519, + "step": 9012 + }, + { + "epoch": 0.8128240970374713, + "grad_norm": 1.6471582303081866, + "learning_rate": 3.56269184803492e-07, + "loss": 1.0311, + "step": 9013 + }, + { + "epoch": 0.8129142805609415, + "grad_norm": 1.33369780069184, + "learning_rate": 3.5593644082403727e-07, + "loss": 1.0025, + "step": 9014 + }, + { + "epoch": 0.8130044640844117, + "grad_norm": 1.9883291542262977, + "learning_rate": 3.5560383712514994e-07, + "loss": 1.0784, + "step": 9015 + }, + { + "epoch": 0.8130946476078821, + "grad_norm": 2.2885722914602775, + "learning_rate": 3.5527137373521066e-07, + "loss": 0.9834, + "step": 9016 + }, + { + "epoch": 0.8131848311313523, + "grad_norm": 1.4142165969429321, + "learning_rate": 3.5493905068258645e-07, + "loss": 0.9354, + "step": 9017 + }, + { + "epoch": 0.8132750146548225, + "grad_norm": 1.6292692735886187, + "learning_rate": 3.546068679956333e-07, + "loss": 0.9953, + "step": 9018 + }, + { + "epoch": 0.8133651981782928, + "grad_norm": 1.3503800440013172, + "learning_rate": 3.5427482570269487e-07, + "loss": 1.043, + "step": 9019 + }, + { + "epoch": 0.8134553817017631, + "grad_norm": 2.2040286544349446, + "learning_rate": 3.539429238321026e-07, + "loss": 0.8285, + "step": 9020 + }, + { + "epoch": 0.8135455652252334, + "grad_norm": 1.4556333161145585, + "learning_rate": 3.536111624121769e-07, + "loss": 0.8318, + "step": 9021 + }, + { + "epoch": 0.8136357487487036, + "grad_norm": 1.7320947865148852, + "learning_rate": 3.532795414712244e-07, + "loss": 0.9423, + "step": 9022 + }, + { + "epoch": 0.8137259322721738, + "grad_norm": 1.4492240144783086, + "learning_rate": 3.5294806103754124e-07, + "loss": 0.9978, + "step": 9023 + }, + { + "epoch": 0.8138161157956442, + "grad_norm": 1.8351699198493443, + "learning_rate": 3.526167211394115e-07, + "loss": 0.99, + "step": 9024 + }, + { + "epoch": 0.8139062993191144, + "grad_norm": 1.5269742943249487, + "learning_rate": 3.522855218051066e-07, + "loss": 0.9206, + "step": 9025 + }, + { + "epoch": 0.8139964828425846, + "grad_norm": 2.2279577334245526, + "learning_rate": 3.5195446306288633e-07, + "loss": 0.9794, + "step": 9026 + }, + { + "epoch": 0.8140866663660549, + "grad_norm": 1.470652302019641, + "learning_rate": 3.51623544940999e-07, + "loss": 0.9205, + "step": 9027 + }, + { + "epoch": 0.8141768498895252, + "grad_norm": 1.3846968403722666, + "learning_rate": 3.5129276746767886e-07, + "loss": 0.966, + "step": 9028 + }, + { + "epoch": 0.8142670334129954, + "grad_norm": 1.4700894031455933, + "learning_rate": 3.5096213067115165e-07, + "loss": 0.9532, + "step": 9029 + }, + { + "epoch": 0.8143572169364657, + "grad_norm": 1.85496427403927, + "learning_rate": 3.506316345796272e-07, + "loss": 0.9965, + "step": 9030 + }, + { + "epoch": 0.814447400459936, + "grad_norm": 0.6863148835087564, + "learning_rate": 3.5030127922130714e-07, + "loss": 0.7728, + "step": 9031 + }, + { + "epoch": 0.8145375839834063, + "grad_norm": 1.7767110851358348, + "learning_rate": 3.4997106462437784e-07, + "loss": 0.8698, + "step": 9032 + }, + { + "epoch": 0.8146277675068765, + "grad_norm": 0.6641629311405879, + "learning_rate": 3.496409908170157e-07, + "loss": 0.7988, + "step": 9033 + }, + { + "epoch": 0.8147179510303467, + "grad_norm": 2.014829375092612, + "learning_rate": 3.493110578273839e-07, + "loss": 0.9417, + "step": 9034 + }, + { + "epoch": 0.8148081345538171, + "grad_norm": 1.6028356190555715, + "learning_rate": 3.489812656836346e-07, + "loss": 1.0345, + "step": 9035 + }, + { + "epoch": 0.8148983180772873, + "grad_norm": 1.4302578814935658, + "learning_rate": 3.486516144139078e-07, + "loss": 0.895, + "step": 9036 + }, + { + "epoch": 0.8149885016007575, + "grad_norm": 1.6796445619284734, + "learning_rate": 3.4832210404632957e-07, + "loss": 0.9482, + "step": 9037 + }, + { + "epoch": 0.8150786851242278, + "grad_norm": 1.3616263939883528, + "learning_rate": 3.479927346090179e-07, + "loss": 0.9877, + "step": 9038 + }, + { + "epoch": 0.8151688686476981, + "grad_norm": 2.0918361969999815, + "learning_rate": 3.4766350613007455e-07, + "loss": 0.9652, + "step": 9039 + }, + { + "epoch": 0.8152590521711683, + "grad_norm": 1.681767440072617, + "learning_rate": 3.4733441863759173e-07, + "loss": 0.9676, + "step": 9040 + }, + { + "epoch": 0.8153492356946386, + "grad_norm": 1.9198201677425277, + "learning_rate": 3.4700547215964916e-07, + "loss": 0.8693, + "step": 9041 + }, + { + "epoch": 0.8154394192181088, + "grad_norm": 1.5899219540568856, + "learning_rate": 3.46676666724314e-07, + "loss": 0.9588, + "step": 9042 + }, + { + "epoch": 0.8155296027415792, + "grad_norm": 1.430326725374591, + "learning_rate": 3.463480023596421e-07, + "loss": 0.9724, + "step": 9043 + }, + { + "epoch": 0.8156197862650494, + "grad_norm": 1.7642731343996263, + "learning_rate": 3.460194790936772e-07, + "loss": 1.0486, + "step": 9044 + }, + { + "epoch": 0.8157099697885196, + "grad_norm": 1.4837482132937079, + "learning_rate": 3.456910969544495e-07, + "loss": 0.9702, + "step": 9045 + }, + { + "epoch": 0.8158001533119899, + "grad_norm": 2.186257581746743, + "learning_rate": 3.4536285596997994e-07, + "loss": 1.0081, + "step": 9046 + }, + { + "epoch": 0.8158903368354602, + "grad_norm": 1.8251935490435902, + "learning_rate": 3.450347561682747e-07, + "loss": 0.9385, + "step": 9047 + }, + { + "epoch": 0.8159805203589304, + "grad_norm": 1.7127224046841922, + "learning_rate": 3.4470679757732945e-07, + "loss": 0.913, + "step": 9048 + }, + { + "epoch": 0.8160707038824007, + "grad_norm": 1.5472665156415086, + "learning_rate": 3.4437898022512735e-07, + "loss": 1.0599, + "step": 9049 + }, + { + "epoch": 0.8161608874058709, + "grad_norm": 1.6706325870203205, + "learning_rate": 3.4405130413963977e-07, + "loss": 0.9793, + "step": 9050 + }, + { + "epoch": 0.8162510709293412, + "grad_norm": 1.370893110500518, + "learning_rate": 3.437237693488262e-07, + "loss": 1.0507, + "step": 9051 + }, + { + "epoch": 0.8163412544528115, + "grad_norm": 1.5718930010210415, + "learning_rate": 3.433963758806322e-07, + "loss": 0.993, + "step": 9052 + }, + { + "epoch": 0.8164314379762817, + "grad_norm": 1.7376882162996474, + "learning_rate": 3.430691237629948e-07, + "loss": 0.8406, + "step": 9053 + }, + { + "epoch": 0.816521621499752, + "grad_norm": 1.7682487100936002, + "learning_rate": 3.427420130238354e-07, + "loss": 0.9288, + "step": 9054 + }, + { + "epoch": 0.8166118050232223, + "grad_norm": 1.5270322203918516, + "learning_rate": 3.424150436910658e-07, + "loss": 1.0644, + "step": 9055 + }, + { + "epoch": 0.8167019885466925, + "grad_norm": 2.1430296374330675, + "learning_rate": 3.420882157925842e-07, + "loss": 1.033, + "step": 9056 + }, + { + "epoch": 0.8167921720701627, + "grad_norm": 2.155960284686503, + "learning_rate": 3.417615293562777e-07, + "loss": 0.9559, + "step": 9057 + }, + { + "epoch": 0.8168823555936331, + "grad_norm": 1.5179062025973222, + "learning_rate": 3.4143498441002105e-07, + "loss": 0.889, + "step": 9058 + }, + { + "epoch": 0.8169725391171033, + "grad_norm": 1.376181485015033, + "learning_rate": 3.411085809816767e-07, + "loss": 0.9863, + "step": 9059 + }, + { + "epoch": 0.8170627226405736, + "grad_norm": 2.4129514637447516, + "learning_rate": 3.407823190990953e-07, + "loss": 1.0947, + "step": 9060 + }, + { + "epoch": 0.8171529061640438, + "grad_norm": 1.7409284344315408, + "learning_rate": 3.4045619879011577e-07, + "loss": 1.0604, + "step": 9061 + }, + { + "epoch": 0.8172430896875141, + "grad_norm": 1.7182481206392166, + "learning_rate": 3.4013022008256334e-07, + "loss": 0.9059, + "step": 9062 + }, + { + "epoch": 0.8173332732109844, + "grad_norm": 1.8764454038805678, + "learning_rate": 3.398043830042532e-07, + "loss": 0.9174, + "step": 9063 + }, + { + "epoch": 0.8174234567344546, + "grad_norm": 1.9570125495184691, + "learning_rate": 3.394786875829871e-07, + "loss": 0.9752, + "step": 9064 + }, + { + "epoch": 0.8175136402579248, + "grad_norm": 1.5229860859646722, + "learning_rate": 3.3915313384655564e-07, + "loss": 0.8608, + "step": 9065 + }, + { + "epoch": 0.8176038237813952, + "grad_norm": 2.768812164228778, + "learning_rate": 3.388277218227369e-07, + "loss": 0.996, + "step": 9066 + }, + { + "epoch": 0.8176940073048654, + "grad_norm": 2.1516585185248553, + "learning_rate": 3.3850245153929557e-07, + "loss": 0.9984, + "step": 9067 + }, + { + "epoch": 0.8177841908283356, + "grad_norm": 1.5712339060830893, + "learning_rate": 3.381773230239875e-07, + "loss": 0.936, + "step": 9068 + }, + { + "epoch": 0.8178743743518059, + "grad_norm": 1.4569949229930217, + "learning_rate": 3.3785233630455247e-07, + "loss": 1.0069, + "step": 9069 + }, + { + "epoch": 0.8179645578752762, + "grad_norm": 1.8671084590765803, + "learning_rate": 3.375274914087221e-07, + "loss": 0.9846, + "step": 9070 + }, + { + "epoch": 0.8180547413987465, + "grad_norm": 1.4101905712906952, + "learning_rate": 3.3720278836421234e-07, + "loss": 0.9521, + "step": 9071 + }, + { + "epoch": 0.8181449249222167, + "grad_norm": 2.1008827216176194, + "learning_rate": 3.368782271987294e-07, + "loss": 0.9996, + "step": 9072 + }, + { + "epoch": 0.8182351084456869, + "grad_norm": 5.801160683087144, + "learning_rate": 3.3655380793996636e-07, + "loss": 0.9855, + "step": 9073 + }, + { + "epoch": 0.8183252919691573, + "grad_norm": 1.50745913461236, + "learning_rate": 3.362295306156047e-07, + "loss": 0.8967, + "step": 9074 + }, + { + "epoch": 0.8184154754926275, + "grad_norm": 1.7346603571205026, + "learning_rate": 3.3590539525331327e-07, + "loss": 0.9921, + "step": 9075 + }, + { + "epoch": 0.8185056590160977, + "grad_norm": 1.429439418927372, + "learning_rate": 3.3558140188074967e-07, + "loss": 0.926, + "step": 9076 + }, + { + "epoch": 0.8185958425395681, + "grad_norm": 1.6710028824749963, + "learning_rate": 3.3525755052555817e-07, + "loss": 1.0058, + "step": 9077 + }, + { + "epoch": 0.8186860260630383, + "grad_norm": 0.7407968442474775, + "learning_rate": 3.3493384121537147e-07, + "loss": 0.8864, + "step": 9078 + }, + { + "epoch": 0.8187762095865085, + "grad_norm": 1.8245341581710346, + "learning_rate": 3.3461027397781075e-07, + "loss": 0.9789, + "step": 9079 + }, + { + "epoch": 0.8188663931099788, + "grad_norm": 2.0197512951325103, + "learning_rate": 3.3428684884048397e-07, + "loss": 0.9646, + "step": 9080 + }, + { + "epoch": 0.8189565766334491, + "grad_norm": 1.4111519043366796, + "learning_rate": 3.3396356583098826e-07, + "loss": 0.9764, + "step": 9081 + }, + { + "epoch": 0.8190467601569194, + "grad_norm": 1.8233784408615323, + "learning_rate": 3.3364042497690736e-07, + "loss": 1.0129, + "step": 9082 + }, + { + "epoch": 0.8191369436803896, + "grad_norm": 1.3711731281832291, + "learning_rate": 3.3331742630581405e-07, + "loss": 1.0358, + "step": 9083 + }, + { + "epoch": 0.8192271272038598, + "grad_norm": 3.5058965059402096, + "learning_rate": 3.3299456984526717e-07, + "loss": 0.9266, + "step": 9084 + }, + { + "epoch": 0.8193173107273302, + "grad_norm": 2.32804634934262, + "learning_rate": 3.3267185562281605e-07, + "loss": 1.0175, + "step": 9085 + }, + { + "epoch": 0.8194074942508004, + "grad_norm": 1.9776481089726183, + "learning_rate": 3.3234928366599514e-07, + "loss": 0.9731, + "step": 9086 + }, + { + "epoch": 0.8194976777742706, + "grad_norm": 1.8011600650357982, + "learning_rate": 3.3202685400232946e-07, + "loss": 1.0098, + "step": 9087 + }, + { + "epoch": 0.8195878612977409, + "grad_norm": 8.500605393336562, + "learning_rate": 3.317045666593292e-07, + "loss": 1.0036, + "step": 9088 + }, + { + "epoch": 0.8196780448212112, + "grad_norm": 1.6185429002404672, + "learning_rate": 3.3138242166449426e-07, + "loss": 0.9832, + "step": 9089 + }, + { + "epoch": 0.8197682283446814, + "grad_norm": 1.5708120130446284, + "learning_rate": 3.310604190453117e-07, + "loss": 1.0392, + "step": 9090 + }, + { + "epoch": 0.8198584118681517, + "grad_norm": 1.400803806955185, + "learning_rate": 3.307385588292566e-07, + "loss": 0.9258, + "step": 9091 + }, + { + "epoch": 0.8199485953916219, + "grad_norm": 1.8439906981517002, + "learning_rate": 3.304168410437924e-07, + "loss": 0.9689, + "step": 9092 + }, + { + "epoch": 0.8200387789150922, + "grad_norm": 0.6279022780349581, + "learning_rate": 3.300952657163687e-07, + "loss": 0.8408, + "step": 9093 + }, + { + "epoch": 0.8201289624385625, + "grad_norm": 1.9435287938478991, + "learning_rate": 3.297738328744248e-07, + "loss": 0.9596, + "step": 9094 + }, + { + "epoch": 0.8202191459620327, + "grad_norm": 1.8406802339151689, + "learning_rate": 3.2945254254538714e-07, + "loss": 0.967, + "step": 9095 + }, + { + "epoch": 0.820309329485503, + "grad_norm": 1.4618361865791296, + "learning_rate": 3.2913139475666963e-07, + "loss": 0.9468, + "step": 9096 + }, + { + "epoch": 0.8203995130089733, + "grad_norm": 1.966749052434208, + "learning_rate": 3.288103895356749e-07, + "loss": 0.9294, + "step": 9097 + }, + { + "epoch": 0.8204896965324435, + "grad_norm": 1.5515670515072417, + "learning_rate": 3.284895269097927e-07, + "loss": 0.9122, + "step": 9098 + }, + { + "epoch": 0.8205798800559138, + "grad_norm": 0.6644780597796154, + "learning_rate": 3.281688069063999e-07, + "loss": 0.819, + "step": 9099 + }, + { + "epoch": 0.8206700635793841, + "grad_norm": 2.00677439648645, + "learning_rate": 3.2784822955286396e-07, + "loss": 0.8043, + "step": 9100 + }, + { + "epoch": 0.8207602471028543, + "grad_norm": 1.5555655142298828, + "learning_rate": 3.275277948765365e-07, + "loss": 1.0478, + "step": 9101 + }, + { + "epoch": 0.8208504306263246, + "grad_norm": 1.6090724623670463, + "learning_rate": 3.2720750290475964e-07, + "loss": 1.0218, + "step": 9102 + }, + { + "epoch": 0.8209406141497948, + "grad_norm": 1.9950091556348544, + "learning_rate": 3.268873536648622e-07, + "loss": 0.9809, + "step": 9103 + }, + { + "epoch": 0.8210307976732651, + "grad_norm": 1.6475612245798374, + "learning_rate": 3.265673471841612e-07, + "loss": 0.9502, + "step": 9104 + }, + { + "epoch": 0.8211209811967354, + "grad_norm": 0.6279337454230152, + "learning_rate": 3.262474834899616e-07, + "loss": 0.7781, + "step": 9105 + }, + { + "epoch": 0.8212111647202056, + "grad_norm": 0.6772482499993837, + "learning_rate": 3.2592776260955534e-07, + "loss": 0.8542, + "step": 9106 + }, + { + "epoch": 0.8213013482436758, + "grad_norm": 4.687438557540023, + "learning_rate": 3.256081845702239e-07, + "loss": 1.0179, + "step": 9107 + }, + { + "epoch": 0.8213915317671462, + "grad_norm": 1.5629723407155345, + "learning_rate": 3.2528874939923335e-07, + "loss": 1.027, + "step": 9108 + }, + { + "epoch": 0.8214817152906164, + "grad_norm": 2.041367787912341, + "learning_rate": 3.2496945712384217e-07, + "loss": 0.9376, + "step": 9109 + }, + { + "epoch": 0.8215718988140867, + "grad_norm": 1.7717790284984825, + "learning_rate": 3.246503077712923e-07, + "loss": 0.9573, + "step": 9110 + }, + { + "epoch": 0.8216620823375569, + "grad_norm": 1.9142222435134446, + "learning_rate": 3.2433130136881625e-07, + "loss": 1.0693, + "step": 9111 + }, + { + "epoch": 0.8217522658610272, + "grad_norm": 2.3174302409350624, + "learning_rate": 3.2401243794363287e-07, + "loss": 0.9765, + "step": 9112 + }, + { + "epoch": 0.8218424493844975, + "grad_norm": 1.6262667193715181, + "learning_rate": 3.236937175229495e-07, + "loss": 1.0365, + "step": 9113 + }, + { + "epoch": 0.8219326329079677, + "grad_norm": 1.7749359925240609, + "learning_rate": 3.233751401339615e-07, + "loss": 0.946, + "step": 9114 + }, + { + "epoch": 0.8220228164314379, + "grad_norm": 3.031833238630587, + "learning_rate": 3.2305670580385157e-07, + "loss": 1.0591, + "step": 9115 + }, + { + "epoch": 0.8221129999549083, + "grad_norm": 1.4624141635321586, + "learning_rate": 3.227384145597898e-07, + "loss": 0.9773, + "step": 9116 + }, + { + "epoch": 0.8222031834783785, + "grad_norm": 1.7808027542528237, + "learning_rate": 3.224202664289346e-07, + "loss": 0.9614, + "step": 9117 + }, + { + "epoch": 0.8222933670018487, + "grad_norm": 1.474710099778072, + "learning_rate": 3.2210226143843257e-07, + "loss": 0.921, + "step": 9118 + }, + { + "epoch": 0.822383550525319, + "grad_norm": 0.6646417447632312, + "learning_rate": 3.217843996154173e-07, + "loss": 0.8513, + "step": 9119 + }, + { + "epoch": 0.8224737340487893, + "grad_norm": 0.6026734089504125, + "learning_rate": 3.2146668098701055e-07, + "loss": 0.7662, + "step": 9120 + }, + { + "epoch": 0.8225639175722596, + "grad_norm": 1.8047275291065923, + "learning_rate": 3.2114910558032215e-07, + "loss": 1.1076, + "step": 9121 + }, + { + "epoch": 0.8226541010957298, + "grad_norm": 2.472504091326932, + "learning_rate": 3.2083167342244945e-07, + "loss": 1.0098, + "step": 9122 + }, + { + "epoch": 0.8227442846192, + "grad_norm": 4.085795354467582, + "learning_rate": 3.205143845404763e-07, + "loss": 1.0786, + "step": 9123 + }, + { + "epoch": 0.8228344681426704, + "grad_norm": 1.900685427657525, + "learning_rate": 3.201972389614773e-07, + "loss": 0.9504, + "step": 9124 + }, + { + "epoch": 0.8229246516661406, + "grad_norm": 2.015321220749371, + "learning_rate": 3.198802367125115e-07, + "loss": 0.9707, + "step": 9125 + }, + { + "epoch": 0.8230148351896108, + "grad_norm": 3.369109912677046, + "learning_rate": 3.195633778206288e-07, + "loss": 0.9901, + "step": 9126 + }, + { + "epoch": 0.8231050187130812, + "grad_norm": 2.15499686716653, + "learning_rate": 3.19246662312864e-07, + "loss": 1.055, + "step": 9127 + }, + { + "epoch": 0.8231952022365514, + "grad_norm": 0.6407733605882123, + "learning_rate": 3.189300902162417e-07, + "loss": 0.8152, + "step": 9128 + }, + { + "epoch": 0.8232853857600216, + "grad_norm": 1.600503785571907, + "learning_rate": 3.1861366155777327e-07, + "loss": 0.989, + "step": 9129 + }, + { + "epoch": 0.8233755692834919, + "grad_norm": 1.3698440899505315, + "learning_rate": 3.182973763644583e-07, + "loss": 0.9866, + "step": 9130 + }, + { + "epoch": 0.8234657528069622, + "grad_norm": 1.4946490850024243, + "learning_rate": 3.1798123466328463e-07, + "loss": 0.9044, + "step": 9131 + }, + { + "epoch": 0.8235559363304324, + "grad_norm": 1.8915702411537956, + "learning_rate": 3.17665236481226e-07, + "loss": 0.9788, + "step": 9132 + }, + { + "epoch": 0.8236461198539027, + "grad_norm": 1.8614399447081624, + "learning_rate": 3.1734938184524576e-07, + "loss": 0.9349, + "step": 9133 + }, + { + "epoch": 0.8237363033773729, + "grad_norm": 1.7723689956568713, + "learning_rate": 3.1703367078229427e-07, + "loss": 0.9888, + "step": 9134 + }, + { + "epoch": 0.8238264869008433, + "grad_norm": 1.8617485988463631, + "learning_rate": 3.167181033193096e-07, + "loss": 0.9435, + "step": 9135 + }, + { + "epoch": 0.8239166704243135, + "grad_norm": 1.6566432989737985, + "learning_rate": 3.16402679483218e-07, + "loss": 0.9879, + "step": 9136 + }, + { + "epoch": 0.8240068539477837, + "grad_norm": 1.3288722012041558, + "learning_rate": 3.1608739930093366e-07, + "loss": 0.911, + "step": 9137 + }, + { + "epoch": 0.824097037471254, + "grad_norm": 1.852472202540349, + "learning_rate": 3.157722627993562e-07, + "loss": 0.8945, + "step": 9138 + }, + { + "epoch": 0.8241872209947243, + "grad_norm": 1.4608480660710594, + "learning_rate": 3.1545727000537727e-07, + "loss": 0.9416, + "step": 9139 + }, + { + "epoch": 0.8242774045181945, + "grad_norm": 1.67836223040759, + "learning_rate": 3.151424209458713e-07, + "loss": 0.9784, + "step": 9140 + }, + { + "epoch": 0.8243675880416648, + "grad_norm": 1.6500494429376333, + "learning_rate": 3.148277156477053e-07, + "loss": 0.9941, + "step": 9141 + }, + { + "epoch": 0.824457771565135, + "grad_norm": 1.9884855693766401, + "learning_rate": 3.145131541377299e-07, + "loss": 0.9777, + "step": 9142 + }, + { + "epoch": 0.8245479550886053, + "grad_norm": 1.6554653540338353, + "learning_rate": 3.1419873644278606e-07, + "loss": 0.9443, + "step": 9143 + }, + { + "epoch": 0.8246381386120756, + "grad_norm": 1.3813703096217111, + "learning_rate": 3.1388446258970147e-07, + "loss": 1.0003, + "step": 9144 + }, + { + "epoch": 0.8247283221355458, + "grad_norm": 1.7544864637886015, + "learning_rate": 3.1357033260529145e-07, + "loss": 0.9658, + "step": 9145 + }, + { + "epoch": 0.824818505659016, + "grad_norm": 1.7090692640747207, + "learning_rate": 3.1325634651636025e-07, + "loss": 1.0567, + "step": 9146 + }, + { + "epoch": 0.8249086891824864, + "grad_norm": 1.4451055430110578, + "learning_rate": 3.1294250434969694e-07, + "loss": 1.0191, + "step": 9147 + }, + { + "epoch": 0.8249988727059566, + "grad_norm": 2.065251797710283, + "learning_rate": 3.1262880613208274e-07, + "loss": 1.0212, + "step": 9148 + }, + { + "epoch": 0.8250890562294269, + "grad_norm": 1.3976024500054043, + "learning_rate": 3.123152518902823e-07, + "loss": 1.0174, + "step": 9149 + }, + { + "epoch": 0.8251792397528972, + "grad_norm": 1.4329133043012172, + "learning_rate": 3.1200184165105017e-07, + "loss": 0.9325, + "step": 9150 + }, + { + "epoch": 0.8252694232763674, + "grad_norm": 1.779218284638363, + "learning_rate": 3.116885754411287e-07, + "loss": 0.9874, + "step": 9151 + }, + { + "epoch": 0.8253596067998377, + "grad_norm": 1.7902042793409443, + "learning_rate": 3.1137545328724703e-07, + "loss": 0.9128, + "step": 9152 + }, + { + "epoch": 0.8254497903233079, + "grad_norm": 1.710885870167597, + "learning_rate": 3.1106247521612285e-07, + "loss": 0.9806, + "step": 9153 + }, + { + "epoch": 0.8255399738467782, + "grad_norm": 2.0929457984682567, + "learning_rate": 3.107496412544612e-07, + "loss": 0.9966, + "step": 9154 + }, + { + "epoch": 0.8256301573702485, + "grad_norm": 1.3576272713027808, + "learning_rate": 3.1043695142895397e-07, + "loss": 1.0067, + "step": 9155 + }, + { + "epoch": 0.8257203408937187, + "grad_norm": 1.4609990234476178, + "learning_rate": 3.101244057662828e-07, + "loss": 0.8999, + "step": 9156 + }, + { + "epoch": 0.8258105244171889, + "grad_norm": 1.9567637612894735, + "learning_rate": 3.098120042931152e-07, + "loss": 0.9635, + "step": 9157 + }, + { + "epoch": 0.8259007079406593, + "grad_norm": 1.744224896589721, + "learning_rate": 3.0949974703610647e-07, + "loss": 1.0217, + "step": 9158 + }, + { + "epoch": 0.8259908914641295, + "grad_norm": 2.745111194713976, + "learning_rate": 3.0918763402190107e-07, + "loss": 0.9693, + "step": 9159 + }, + { + "epoch": 0.8260810749875998, + "grad_norm": 2.152307970473083, + "learning_rate": 3.088756652771296e-07, + "loss": 0.9358, + "step": 9160 + }, + { + "epoch": 0.82617125851107, + "grad_norm": 2.4032206105830367, + "learning_rate": 3.0856384082841147e-07, + "loss": 1.0741, + "step": 9161 + }, + { + "epoch": 0.8262614420345403, + "grad_norm": 1.546666295007683, + "learning_rate": 3.0825216070235207e-07, + "loss": 0.9264, + "step": 9162 + }, + { + "epoch": 0.8263516255580106, + "grad_norm": 1.7390458729129898, + "learning_rate": 3.0794062492554764e-07, + "loss": 1.0911, + "step": 9163 + }, + { + "epoch": 0.8264418090814808, + "grad_norm": 1.6806102967957934, + "learning_rate": 3.076292335245783e-07, + "loss": 0.9915, + "step": 9164 + }, + { + "epoch": 0.826531992604951, + "grad_norm": 2.006623268460804, + "learning_rate": 3.073179865260145e-07, + "loss": 1.0481, + "step": 9165 + }, + { + "epoch": 0.8266221761284214, + "grad_norm": 1.6949025726450082, + "learning_rate": 3.070068839564135e-07, + "loss": 0.9841, + "step": 9166 + }, + { + "epoch": 0.8267123596518916, + "grad_norm": 2.1348547467530166, + "learning_rate": 3.0669592584232006e-07, + "loss": 1.0121, + "step": 9167 + }, + { + "epoch": 0.8268025431753618, + "grad_norm": 1.5151778672556522, + "learning_rate": 3.063851122102672e-07, + "loss": 0.8883, + "step": 9168 + }, + { + "epoch": 0.8268927266988321, + "grad_norm": 2.3375611037157373, + "learning_rate": 3.06074443086775e-07, + "loss": 0.8991, + "step": 9169 + }, + { + "epoch": 0.8269829102223024, + "grad_norm": 0.6028439385742728, + "learning_rate": 3.057639184983514e-07, + "loss": 0.7682, + "step": 9170 + }, + { + "epoch": 0.8270730937457726, + "grad_norm": 1.2931324374029742, + "learning_rate": 3.054535384714927e-07, + "loss": 0.9827, + "step": 9171 + }, + { + "epoch": 0.8271632772692429, + "grad_norm": 1.9039535447712406, + "learning_rate": 3.0514330303268135e-07, + "loss": 1.1436, + "step": 9172 + }, + { + "epoch": 0.8272534607927132, + "grad_norm": 1.8778228491327982, + "learning_rate": 3.0483321220838876e-07, + "loss": 0.9288, + "step": 9173 + }, + { + "epoch": 0.8273436443161835, + "grad_norm": 1.3915050165248395, + "learning_rate": 3.045232660250734e-07, + "loss": 0.8375, + "step": 9174 + }, + { + "epoch": 0.8274338278396537, + "grad_norm": 1.8929537468290865, + "learning_rate": 3.0421346450918185e-07, + "loss": 1.0078, + "step": 9175 + }, + { + "epoch": 0.8275240113631239, + "grad_norm": 1.3979777001115417, + "learning_rate": 3.039038076871485e-07, + "loss": 0.9268, + "step": 9176 + }, + { + "epoch": 0.8276141948865943, + "grad_norm": 1.435680149664863, + "learning_rate": 3.035942955853934e-07, + "loss": 0.9991, + "step": 9177 + }, + { + "epoch": 0.8277043784100645, + "grad_norm": 1.8131524588797947, + "learning_rate": 3.0328492823032804e-07, + "loss": 0.9553, + "step": 9178 + }, + { + "epoch": 0.8277945619335347, + "grad_norm": 1.7215781092632874, + "learning_rate": 3.029757056483471e-07, + "loss": 0.9165, + "step": 9179 + }, + { + "epoch": 0.827884745457005, + "grad_norm": 2.011809056301346, + "learning_rate": 3.026666278658372e-07, + "loss": 0.8705, + "step": 9180 + }, + { + "epoch": 0.8279749289804753, + "grad_norm": 1.2656292856402893, + "learning_rate": 3.023576949091691e-07, + "loss": 0.9782, + "step": 9181 + }, + { + "epoch": 0.8280651125039455, + "grad_norm": 1.845653327391219, + "learning_rate": 3.020489068047032e-07, + "loss": 0.901, + "step": 9182 + }, + { + "epoch": 0.8281552960274158, + "grad_norm": 1.8168039194223513, + "learning_rate": 3.017402635787869e-07, + "loss": 0.9391, + "step": 9183 + }, + { + "epoch": 0.828245479550886, + "grad_norm": 1.6838849517276728, + "learning_rate": 3.0143176525775537e-07, + "loss": 0.9519, + "step": 9184 + }, + { + "epoch": 0.8283356630743564, + "grad_norm": 1.5026660908633338, + "learning_rate": 3.0112341186793155e-07, + "loss": 0.9231, + "step": 9185 + }, + { + "epoch": 0.8284258465978266, + "grad_norm": 1.6937961389092004, + "learning_rate": 3.008152034356264e-07, + "loss": 0.9151, + "step": 9186 + }, + { + "epoch": 0.8285160301212968, + "grad_norm": 0.6457087309421486, + "learning_rate": 3.005071399871366e-07, + "loss": 0.8011, + "step": 9187 + }, + { + "epoch": 0.828606213644767, + "grad_norm": 1.6105962027305671, + "learning_rate": 3.0019922154874853e-07, + "loss": 1.0433, + "step": 9188 + }, + { + "epoch": 0.8286963971682374, + "grad_norm": 2.1489055331818454, + "learning_rate": 2.998914481467356e-07, + "loss": 0.9374, + "step": 9189 + }, + { + "epoch": 0.8287865806917076, + "grad_norm": 1.9109826045978764, + "learning_rate": 2.9958381980735837e-07, + "loss": 0.9269, + "step": 9190 + }, + { + "epoch": 0.8288767642151779, + "grad_norm": 1.737796398651434, + "learning_rate": 2.992763365568658e-07, + "loss": 0.907, + "step": 9191 + }, + { + "epoch": 0.8289669477386481, + "grad_norm": 1.6878425815545248, + "learning_rate": 2.98968998421494e-07, + "loss": 0.9789, + "step": 9192 + }, + { + "epoch": 0.8290571312621184, + "grad_norm": 1.7636198963435081, + "learning_rate": 2.98661805427467e-07, + "loss": 0.9694, + "step": 9193 + }, + { + "epoch": 0.8291473147855887, + "grad_norm": 1.4645868915426008, + "learning_rate": 2.9835475760099483e-07, + "loss": 0.9929, + "step": 9194 + }, + { + "epoch": 0.8292374983090589, + "grad_norm": 1.5711843622862878, + "learning_rate": 2.9804785496827856e-07, + "loss": 0.9946, + "step": 9195 + }, + { + "epoch": 0.8293276818325293, + "grad_norm": 2.9949855222017754, + "learning_rate": 2.977410975555028e-07, + "loss": 0.9727, + "step": 9196 + }, + { + "epoch": 0.8294178653559995, + "grad_norm": 1.7799098678871703, + "learning_rate": 2.9743448538884376e-07, + "loss": 1.0406, + "step": 9197 + }, + { + "epoch": 0.8295080488794697, + "grad_norm": 1.8904425793119608, + "learning_rate": 2.9712801849446154e-07, + "loss": 0.9672, + "step": 9198 + }, + { + "epoch": 0.82959823240294, + "grad_norm": 1.5778846935454054, + "learning_rate": 2.9682169689850665e-07, + "loss": 0.9269, + "step": 9199 + }, + { + "epoch": 0.8296884159264103, + "grad_norm": 0.649329778034724, + "learning_rate": 2.9651552062711573e-07, + "loss": 0.7633, + "step": 9200 + }, + { + "epoch": 0.8297785994498805, + "grad_norm": 1.466964610499187, + "learning_rate": 2.9620948970641333e-07, + "loss": 1.0547, + "step": 9201 + }, + { + "epoch": 0.8298687829733508, + "grad_norm": 2.2375862147594323, + "learning_rate": 2.959036041625125e-07, + "loss": 0.9944, + "step": 9202 + }, + { + "epoch": 0.829958966496821, + "grad_norm": 1.9714540349607166, + "learning_rate": 2.95597864021512e-07, + "loss": 0.8881, + "step": 9203 + }, + { + "epoch": 0.8300491500202913, + "grad_norm": 31.322499884642628, + "learning_rate": 2.9529226930949966e-07, + "loss": 1.0218, + "step": 9204 + }, + { + "epoch": 0.8301393335437616, + "grad_norm": 1.7536522356382294, + "learning_rate": 2.949868200525505e-07, + "loss": 0.99, + "step": 9205 + }, + { + "epoch": 0.8302295170672318, + "grad_norm": 1.626340020141699, + "learning_rate": 2.9468151627672734e-07, + "loss": 0.9714, + "step": 9206 + }, + { + "epoch": 0.830319700590702, + "grad_norm": 1.5213781278090857, + "learning_rate": 2.9437635800808026e-07, + "loss": 0.9339, + "step": 9207 + }, + { + "epoch": 0.8304098841141724, + "grad_norm": 1.7016296121272356, + "learning_rate": 2.940713452726473e-07, + "loss": 0.9709, + "step": 9208 + }, + { + "epoch": 0.8305000676376426, + "grad_norm": 1.7135167298826255, + "learning_rate": 2.937664780964526e-07, + "loss": 0.973, + "step": 9209 + }, + { + "epoch": 0.8305902511611128, + "grad_norm": 2.8177511572741953, + "learning_rate": 2.9346175650551133e-07, + "loss": 0.9869, + "step": 9210 + }, + { + "epoch": 0.8306804346845831, + "grad_norm": 1.522873994339886, + "learning_rate": 2.931571805258215e-07, + "loss": 1.0004, + "step": 9211 + }, + { + "epoch": 0.8307706182080534, + "grad_norm": 2.5603625989957326, + "learning_rate": 2.9285275018337353e-07, + "loss": 1.0259, + "step": 9212 + }, + { + "epoch": 0.8308608017315237, + "grad_norm": 3.4633595966105757, + "learning_rate": 2.9254846550414146e-07, + "loss": 1.0795, + "step": 9213 + }, + { + "epoch": 0.8309509852549939, + "grad_norm": 1.767248774241126, + "learning_rate": 2.922443265140893e-07, + "loss": 1.0289, + "step": 9214 + }, + { + "epoch": 0.8310411687784641, + "grad_norm": 1.6257140718115632, + "learning_rate": 2.919403332391674e-07, + "loss": 0.9779, + "step": 9215 + }, + { + "epoch": 0.8311313523019345, + "grad_norm": 1.961946508994701, + "learning_rate": 2.9163648570531464e-07, + "loss": 1.0018, + "step": 9216 + }, + { + "epoch": 0.8312215358254047, + "grad_norm": 1.876273168005278, + "learning_rate": 2.9133278393845717e-07, + "loss": 0.918, + "step": 9217 + }, + { + "epoch": 0.8313117193488749, + "grad_norm": 1.5770771589204502, + "learning_rate": 2.9102922796450703e-07, + "loss": 1.0503, + "step": 9218 + }, + { + "epoch": 0.8314019028723453, + "grad_norm": 1.707264550369784, + "learning_rate": 2.907258178093672e-07, + "loss": 0.9375, + "step": 9219 + }, + { + "epoch": 0.8314920863958155, + "grad_norm": 1.6020575525849357, + "learning_rate": 2.904225534989251e-07, + "loss": 0.8984, + "step": 9220 + }, + { + "epoch": 0.8315822699192857, + "grad_norm": 1.7659449329668697, + "learning_rate": 2.901194350590572e-07, + "loss": 0.8806, + "step": 9221 + }, + { + "epoch": 0.831672453442756, + "grad_norm": 1.6550158094820786, + "learning_rate": 2.898164625156274e-07, + "loss": 0.863, + "step": 9222 + }, + { + "epoch": 0.8317626369662263, + "grad_norm": 1.4414693513616295, + "learning_rate": 2.8951363589448676e-07, + "loss": 0.9368, + "step": 9223 + }, + { + "epoch": 0.8318528204896966, + "grad_norm": 1.6064591149198089, + "learning_rate": 2.8921095522147434e-07, + "loss": 0.975, + "step": 9224 + }, + { + "epoch": 0.8319430040131668, + "grad_norm": 2.3998868955345882, + "learning_rate": 2.8890842052241683e-07, + "loss": 1.011, + "step": 9225 + }, + { + "epoch": 0.832033187536637, + "grad_norm": 0.5991541722002542, + "learning_rate": 2.886060318231267e-07, + "loss": 0.7642, + "step": 9226 + }, + { + "epoch": 0.8321233710601074, + "grad_norm": 1.8170399862622797, + "learning_rate": 2.8830378914940755e-07, + "loss": 1.003, + "step": 9227 + }, + { + "epoch": 0.8322135545835776, + "grad_norm": 1.8145964107049497, + "learning_rate": 2.8800169252704675e-07, + "loss": 0.9634, + "step": 9228 + }, + { + "epoch": 0.8323037381070478, + "grad_norm": 1.4286859057743932, + "learning_rate": 2.8769974198182143e-07, + "loss": 1.0, + "step": 9229 + }, + { + "epoch": 0.8323939216305181, + "grad_norm": 1.38387490533238, + "learning_rate": 2.873979375394955e-07, + "loss": 0.9792, + "step": 9230 + }, + { + "epoch": 0.8324841051539884, + "grad_norm": 1.4552698201119314, + "learning_rate": 2.870962792258209e-07, + "loss": 1.0103, + "step": 9231 + }, + { + "epoch": 0.8325742886774586, + "grad_norm": 2.1223669007246375, + "learning_rate": 2.8679476706653716e-07, + "loss": 0.9899, + "step": 9232 + }, + { + "epoch": 0.8326644722009289, + "grad_norm": 1.831321725319502, + "learning_rate": 2.864934010873692e-07, + "loss": 0.9994, + "step": 9233 + }, + { + "epoch": 0.8327546557243991, + "grad_norm": 1.5322814114674996, + "learning_rate": 2.8619218131403357e-07, + "loss": 0.9535, + "step": 9234 + }, + { + "epoch": 0.8328448392478695, + "grad_norm": 1.777354146067638, + "learning_rate": 2.858911077722299e-07, + "loss": 1.0019, + "step": 9235 + }, + { + "epoch": 0.8329350227713397, + "grad_norm": 2.032538841411328, + "learning_rate": 2.855901804876493e-07, + "loss": 0.8295, + "step": 9236 + }, + { + "epoch": 0.8330252062948099, + "grad_norm": 1.6060543430963088, + "learning_rate": 2.852893994859673e-07, + "loss": 1.0164, + "step": 9237 + }, + { + "epoch": 0.8331153898182801, + "grad_norm": 1.3311629163887366, + "learning_rate": 2.849887647928484e-07, + "loss": 0.9234, + "step": 9238 + }, + { + "epoch": 0.8332055733417505, + "grad_norm": 2.1214616150917887, + "learning_rate": 2.8468827643394465e-07, + "loss": 0.8728, + "step": 9239 + }, + { + "epoch": 0.8332957568652207, + "grad_norm": 1.5838942872568724, + "learning_rate": 2.843879344348954e-07, + "loss": 0.9918, + "step": 9240 + }, + { + "epoch": 0.833385940388691, + "grad_norm": 2.083222869487636, + "learning_rate": 2.840877388213272e-07, + "loss": 0.9471, + "step": 9241 + }, + { + "epoch": 0.8334761239121612, + "grad_norm": 1.8392754910273559, + "learning_rate": 2.8378768961885515e-07, + "loss": 0.9806, + "step": 9242 + }, + { + "epoch": 0.8335663074356315, + "grad_norm": 1.6549677653789392, + "learning_rate": 2.8348778685307983e-07, + "loss": 1.0099, + "step": 9243 + }, + { + "epoch": 0.8336564909591018, + "grad_norm": 1.5483486977681646, + "learning_rate": 2.831880305495915e-07, + "loss": 1.0523, + "step": 9244 + }, + { + "epoch": 0.833746674482572, + "grad_norm": 0.712706576065206, + "learning_rate": 2.828884207339668e-07, + "loss": 0.8494, + "step": 9245 + }, + { + "epoch": 0.8338368580060423, + "grad_norm": 2.3884568698591244, + "learning_rate": 2.8258895743177014e-07, + "loss": 1.0547, + "step": 9246 + }, + { + "epoch": 0.8339270415295126, + "grad_norm": 1.7085068389737836, + "learning_rate": 2.8228964066855356e-07, + "loss": 1.0559, + "step": 9247 + }, + { + "epoch": 0.8340172250529828, + "grad_norm": 2.350166793747852, + "learning_rate": 2.819904704698555e-07, + "loss": 0.956, + "step": 9248 + }, + { + "epoch": 0.834107408576453, + "grad_norm": 0.7671152519811365, + "learning_rate": 2.8169144686120437e-07, + "loss": 0.7654, + "step": 9249 + }, + { + "epoch": 0.8341975920999234, + "grad_norm": 1.9942833142318837, + "learning_rate": 2.8139256986811254e-07, + "loss": 0.8939, + "step": 9250 + }, + { + "epoch": 0.8342877756233936, + "grad_norm": 2.1601552566608815, + "learning_rate": 2.8109383951608424e-07, + "loss": 1.0337, + "step": 9251 + }, + { + "epoch": 0.8343779591468639, + "grad_norm": 1.6197356268540457, + "learning_rate": 2.8079525583060683e-07, + "loss": 1.0437, + "step": 9252 + }, + { + "epoch": 0.8344681426703341, + "grad_norm": 1.3427781871364861, + "learning_rate": 2.804968188371577e-07, + "loss": 0.9849, + "step": 9253 + }, + { + "epoch": 0.8345583261938044, + "grad_norm": 1.5988603824928174, + "learning_rate": 2.801985285612014e-07, + "loss": 0.9784, + "step": 9254 + }, + { + "epoch": 0.8346485097172747, + "grad_norm": 1.8418570677548054, + "learning_rate": 2.7990038502818934e-07, + "loss": 0.9506, + "step": 9255 + }, + { + "epoch": 0.8347386932407449, + "grad_norm": 1.4866368778876557, + "learning_rate": 2.796023882635612e-07, + "loss": 0.9543, + "step": 9256 + }, + { + "epoch": 0.8348288767642151, + "grad_norm": 1.5734717629574617, + "learning_rate": 2.7930453829274323e-07, + "loss": 1.0262, + "step": 9257 + }, + { + "epoch": 0.8349190602876855, + "grad_norm": 1.9422976866912538, + "learning_rate": 2.7900683514115054e-07, + "loss": 1.0227, + "step": 9258 + }, + { + "epoch": 0.8350092438111557, + "grad_norm": 1.6989527000772002, + "learning_rate": 2.787092788341836e-07, + "loss": 0.917, + "step": 9259 + }, + { + "epoch": 0.8350994273346259, + "grad_norm": 1.554594410932067, + "learning_rate": 2.7841186939723195e-07, + "loss": 0.9763, + "step": 9260 + }, + { + "epoch": 0.8351896108580962, + "grad_norm": 2.101955674841071, + "learning_rate": 2.7811460685567255e-07, + "loss": 0.9545, + "step": 9261 + }, + { + "epoch": 0.8352797943815665, + "grad_norm": 2.593019957236486, + "learning_rate": 2.778174912348692e-07, + "loss": 0.9709, + "step": 9262 + }, + { + "epoch": 0.8353699779050368, + "grad_norm": 1.5277103072016311, + "learning_rate": 2.7752052256017354e-07, + "loss": 0.9441, + "step": 9263 + }, + { + "epoch": 0.835460161428507, + "grad_norm": 1.7498901196452048, + "learning_rate": 2.7722370085692493e-07, + "loss": 0.9456, + "step": 9264 + }, + { + "epoch": 0.8355503449519772, + "grad_norm": 1.9403245702568694, + "learning_rate": 2.769270261504486e-07, + "loss": 0.951, + "step": 9265 + }, + { + "epoch": 0.8356405284754476, + "grad_norm": 1.457818603502848, + "learning_rate": 2.7663049846606015e-07, + "loss": 0.9394, + "step": 9266 + }, + { + "epoch": 0.8357307119989178, + "grad_norm": 1.2718007621375325, + "learning_rate": 2.763341178290592e-07, + "loss": 0.9909, + "step": 9267 + }, + { + "epoch": 0.835820895522388, + "grad_norm": 1.3787634237955353, + "learning_rate": 2.7603788426473663e-07, + "loss": 1.0017, + "step": 9268 + }, + { + "epoch": 0.8359110790458584, + "grad_norm": 2.042702418631148, + "learning_rate": 2.7574179779836695e-07, + "loss": 0.9128, + "step": 9269 + }, + { + "epoch": 0.8360012625693286, + "grad_norm": 1.603902456480688, + "learning_rate": 2.754458584552146e-07, + "loss": 0.9359, + "step": 9270 + }, + { + "epoch": 0.8360914460927988, + "grad_norm": 2.844889821096284, + "learning_rate": 2.751500662605308e-07, + "loss": 0.9611, + "step": 9271 + }, + { + "epoch": 0.8361816296162691, + "grad_norm": 1.5344228764229846, + "learning_rate": 2.7485442123955383e-07, + "loss": 0.9349, + "step": 9272 + }, + { + "epoch": 0.8362718131397394, + "grad_norm": 1.6506614688540837, + "learning_rate": 2.7455892341751075e-07, + "loss": 0.8628, + "step": 9273 + }, + { + "epoch": 0.8363619966632097, + "grad_norm": 1.8299073879592078, + "learning_rate": 2.7426357281961365e-07, + "loss": 0.8561, + "step": 9274 + }, + { + "epoch": 0.8364521801866799, + "grad_norm": 2.448056664758479, + "learning_rate": 2.7396836947106416e-07, + "loss": 0.9852, + "step": 9275 + }, + { + "epoch": 0.8365423637101501, + "grad_norm": 1.6992607023826172, + "learning_rate": 2.736733133970506e-07, + "loss": 1.0742, + "step": 9276 + }, + { + "epoch": 0.8366325472336205, + "grad_norm": 1.3525972319395696, + "learning_rate": 2.7337840462274896e-07, + "loss": 1.0172, + "step": 9277 + }, + { + "epoch": 0.8367227307570907, + "grad_norm": 1.8429271995708596, + "learning_rate": 2.730836431733221e-07, + "loss": 0.944, + "step": 9278 + }, + { + "epoch": 0.8368129142805609, + "grad_norm": 1.737536324320306, + "learning_rate": 2.727890290739212e-07, + "loss": 1.03, + "step": 9279 + }, + { + "epoch": 0.8369030978040312, + "grad_norm": 1.75527295937876, + "learning_rate": 2.7249456234968395e-07, + "loss": 1.0091, + "step": 9280 + }, + { + "epoch": 0.8369932813275015, + "grad_norm": 1.6549831799807777, + "learning_rate": 2.722002430257364e-07, + "loss": 1.0429, + "step": 9281 + }, + { + "epoch": 0.8370834648509717, + "grad_norm": 1.6298292001508712, + "learning_rate": 2.7190607112719035e-07, + "loss": 0.9411, + "step": 9282 + }, + { + "epoch": 0.837173648374442, + "grad_norm": 1.9900718675029145, + "learning_rate": 2.716120466791476e-07, + "loss": 0.8982, + "step": 9283 + }, + { + "epoch": 0.8372638318979122, + "grad_norm": 1.9036973841028064, + "learning_rate": 2.7131816970669483e-07, + "loss": 0.957, + "step": 9284 + }, + { + "epoch": 0.8373540154213825, + "grad_norm": 1.9656397153367229, + "learning_rate": 2.7102444023490777e-07, + "loss": 0.9842, + "step": 9285 + }, + { + "epoch": 0.8374441989448528, + "grad_norm": 1.4573045735018266, + "learning_rate": 2.70730858288849e-07, + "loss": 1.0339, + "step": 9286 + }, + { + "epoch": 0.837534382468323, + "grad_norm": 1.8065819643137062, + "learning_rate": 2.704374238935685e-07, + "loss": 0.913, + "step": 9287 + }, + { + "epoch": 0.8376245659917932, + "grad_norm": 0.6716003965723686, + "learning_rate": 2.70144137074104e-07, + "loss": 0.8123, + "step": 9288 + }, + { + "epoch": 0.8377147495152636, + "grad_norm": 1.4934240679906152, + "learning_rate": 2.6985099785547926e-07, + "loss": 0.9976, + "step": 9289 + }, + { + "epoch": 0.8378049330387338, + "grad_norm": 1.4893837197032913, + "learning_rate": 2.695580062627083e-07, + "loss": 1.0006, + "step": 9290 + }, + { + "epoch": 0.8378951165622041, + "grad_norm": 1.9279671850181885, + "learning_rate": 2.692651623207891e-07, + "loss": 0.9747, + "step": 9291 + }, + { + "epoch": 0.8379853000856744, + "grad_norm": 1.4957071706259353, + "learning_rate": 2.689724660547097e-07, + "loss": 0.9653, + "step": 9292 + }, + { + "epoch": 0.8380754836091446, + "grad_norm": 0.6272933369638264, + "learning_rate": 2.686799174894441e-07, + "loss": 0.8082, + "step": 9293 + }, + { + "epoch": 0.8381656671326149, + "grad_norm": 0.6494761042967556, + "learning_rate": 2.683875166499545e-07, + "loss": 0.8091, + "step": 9294 + }, + { + "epoch": 0.8382558506560851, + "grad_norm": 1.6943172222939726, + "learning_rate": 2.680952635611899e-07, + "loss": 0.9836, + "step": 9295 + }, + { + "epoch": 0.8383460341795554, + "grad_norm": 1.5245556762527588, + "learning_rate": 2.678031582480875e-07, + "loss": 0.904, + "step": 9296 + }, + { + "epoch": 0.8384362177030257, + "grad_norm": 1.6031611209837844, + "learning_rate": 2.6751120073557e-07, + "loss": 1.0193, + "step": 9297 + }, + { + "epoch": 0.8385264012264959, + "grad_norm": 1.633083120306183, + "learning_rate": 2.672193910485505e-07, + "loss": 0.9456, + "step": 9298 + }, + { + "epoch": 0.8386165847499661, + "grad_norm": 1.5598612056024113, + "learning_rate": 2.669277292119265e-07, + "loss": 0.9766, + "step": 9299 + }, + { + "epoch": 0.8387067682734365, + "grad_norm": 1.7704348246642092, + "learning_rate": 2.666362152505848e-07, + "loss": 0.9809, + "step": 9300 + }, + { + "epoch": 0.8387969517969067, + "grad_norm": 1.8018019425278256, + "learning_rate": 2.663448491893989e-07, + "loss": 0.9104, + "step": 9301 + }, + { + "epoch": 0.838887135320377, + "grad_norm": 1.633295672170634, + "learning_rate": 2.6605363105322974e-07, + "loss": 0.9529, + "step": 9302 + }, + { + "epoch": 0.8389773188438472, + "grad_norm": 1.57242438421098, + "learning_rate": 2.657625608669263e-07, + "loss": 0.9759, + "step": 9303 + }, + { + "epoch": 0.8390675023673175, + "grad_norm": 3.0480768425433955, + "learning_rate": 2.654716386553224e-07, + "loss": 0.8911, + "step": 9304 + }, + { + "epoch": 0.8391576858907878, + "grad_norm": 2.0908288648164066, + "learning_rate": 2.651808644432436e-07, + "loss": 0.9747, + "step": 9305 + }, + { + "epoch": 0.839247869414258, + "grad_norm": 1.8914091517517948, + "learning_rate": 2.6489023825549807e-07, + "loss": 1.1093, + "step": 9306 + }, + { + "epoch": 0.8393380529377282, + "grad_norm": 2.3895015694804047, + "learning_rate": 2.6459976011688547e-07, + "loss": 0.8971, + "step": 9307 + }, + { + "epoch": 0.8394282364611986, + "grad_norm": 1.9089111139217252, + "learning_rate": 2.6430943005219e-07, + "loss": 0.9728, + "step": 9308 + }, + { + "epoch": 0.8395184199846688, + "grad_norm": 1.3892720960089, + "learning_rate": 2.6401924808618447e-07, + "loss": 1.0022, + "step": 9309 + }, + { + "epoch": 0.839608603508139, + "grad_norm": 1.6862844398324008, + "learning_rate": 2.637292142436287e-07, + "loss": 0.9709, + "step": 9310 + }, + { + "epoch": 0.8396987870316093, + "grad_norm": 7.128259549006198, + "learning_rate": 2.6343932854927e-07, + "loss": 0.9451, + "step": 9311 + }, + { + "epoch": 0.8397889705550796, + "grad_norm": 0.6841682853478238, + "learning_rate": 2.6314959102784316e-07, + "loss": 0.807, + "step": 9312 + }, + { + "epoch": 0.8398791540785498, + "grad_norm": 1.892005794504176, + "learning_rate": 2.6286000170407074e-07, + "loss": 0.9701, + "step": 9313 + }, + { + "epoch": 0.8399693376020201, + "grad_norm": 2.2695858763326906, + "learning_rate": 2.625705606026607e-07, + "loss": 0.955, + "step": 9314 + }, + { + "epoch": 0.8400595211254904, + "grad_norm": 1.4512112228378078, + "learning_rate": 2.622812677483106e-07, + "loss": 1.0553, + "step": 9315 + }, + { + "epoch": 0.8401497046489607, + "grad_norm": 1.9372293837047567, + "learning_rate": 2.6199212316570453e-07, + "loss": 1.0288, + "step": 9316 + }, + { + "epoch": 0.8402398881724309, + "grad_norm": 1.5418561956390082, + "learning_rate": 2.617031268795138e-07, + "loss": 0.9541, + "step": 9317 + }, + { + "epoch": 0.8403300716959011, + "grad_norm": 1.6023586620304637, + "learning_rate": 2.614142789143976e-07, + "loss": 1.0034, + "step": 9318 + }, + { + "epoch": 0.8404202552193715, + "grad_norm": 1.593499182618012, + "learning_rate": 2.6112557929500047e-07, + "loss": 0.9592, + "step": 9319 + }, + { + "epoch": 0.8405104387428417, + "grad_norm": 1.6140246583503801, + "learning_rate": 2.6083702804595817e-07, + "loss": 0.9985, + "step": 9320 + }, + { + "epoch": 0.8406006222663119, + "grad_norm": 1.4446865066435788, + "learning_rate": 2.6054862519188915e-07, + "loss": 0.959, + "step": 9321 + }, + { + "epoch": 0.8406908057897822, + "grad_norm": 2.661935533317593, + "learning_rate": 2.6026037075740357e-07, + "loss": 0.9115, + "step": 9322 + }, + { + "epoch": 0.8407809893132525, + "grad_norm": 1.8055209922742772, + "learning_rate": 2.5997226476709524e-07, + "loss": 0.9422, + "step": 9323 + }, + { + "epoch": 0.8408711728367227, + "grad_norm": 1.6797449856169604, + "learning_rate": 2.5968430724554856e-07, + "loss": 0.9767, + "step": 9324 + }, + { + "epoch": 0.840961356360193, + "grad_norm": 1.5015484447000056, + "learning_rate": 2.5939649821733225e-07, + "loss": 0.9054, + "step": 9325 + }, + { + "epoch": 0.8410515398836632, + "grad_norm": 2.0672921659538264, + "learning_rate": 2.5910883770700433e-07, + "loss": 0.9424, + "step": 9326 + }, + { + "epoch": 0.8411417234071336, + "grad_norm": 1.2838431377852804, + "learning_rate": 2.5882132573910965e-07, + "loss": 0.8181, + "step": 9327 + }, + { + "epoch": 0.8412319069306038, + "grad_norm": 1.2639299509460316, + "learning_rate": 2.585339623381801e-07, + "loss": 1.0144, + "step": 9328 + }, + { + "epoch": 0.841322090454074, + "grad_norm": 1.6588921255642373, + "learning_rate": 2.582467475287358e-07, + "loss": 0.9518, + "step": 9329 + }, + { + "epoch": 0.8414122739775443, + "grad_norm": 1.667310955764302, + "learning_rate": 2.5795968133528224e-07, + "loss": 0.9841, + "step": 9330 + }, + { + "epoch": 0.8415024575010146, + "grad_norm": 1.6613951604267287, + "learning_rate": 2.576727637823144e-07, + "loss": 0.9748, + "step": 9331 + }, + { + "epoch": 0.8415926410244848, + "grad_norm": 2.007252418386428, + "learning_rate": 2.5738599489431335e-07, + "loss": 1.0206, + "step": 9332 + }, + { + "epoch": 0.8416828245479551, + "grad_norm": 1.4765345656563142, + "learning_rate": 2.5709937469574794e-07, + "loss": 1.0373, + "step": 9333 + }, + { + "epoch": 0.8417730080714253, + "grad_norm": 1.956088206270164, + "learning_rate": 2.568129032110742e-07, + "loss": 1.0151, + "step": 9334 + }, + { + "epoch": 0.8418631915948956, + "grad_norm": 1.5453127283418215, + "learning_rate": 2.5652658046473565e-07, + "loss": 1.0508, + "step": 9335 + }, + { + "epoch": 0.8419533751183659, + "grad_norm": 1.3950259829736835, + "learning_rate": 2.5624040648116184e-07, + "loss": 0.9758, + "step": 9336 + }, + { + "epoch": 0.8420435586418361, + "grad_norm": 1.6636494585091324, + "learning_rate": 2.5595438128477245e-07, + "loss": 0.9177, + "step": 9337 + }, + { + "epoch": 0.8421337421653065, + "grad_norm": 1.6081928985767202, + "learning_rate": 2.5566850489997096e-07, + "loss": 0.9795, + "step": 9338 + }, + { + "epoch": 0.8422239256887767, + "grad_norm": 1.612374926307346, + "learning_rate": 2.5538277735115166e-07, + "loss": 1.0033, + "step": 9339 + }, + { + "epoch": 0.8423141092122469, + "grad_norm": 1.4476115319169254, + "learning_rate": 2.5509719866269306e-07, + "loss": 1.0307, + "step": 9340 + }, + { + "epoch": 0.8424042927357172, + "grad_norm": 3.841587481782497, + "learning_rate": 2.548117688589628e-07, + "loss": 1.0387, + "step": 9341 + }, + { + "epoch": 0.8424944762591875, + "grad_norm": 1.8833168588891123, + "learning_rate": 2.545264879643152e-07, + "loss": 0.96, + "step": 9342 + }, + { + "epoch": 0.8425846597826577, + "grad_norm": 1.5609985767287144, + "learning_rate": 2.542413560030923e-07, + "loss": 0.9349, + "step": 9343 + }, + { + "epoch": 0.842674843306128, + "grad_norm": 1.6088908680238605, + "learning_rate": 2.53956372999623e-07, + "loss": 0.8714, + "step": 9344 + }, + { + "epoch": 0.8427650268295982, + "grad_norm": 0.6570972241675253, + "learning_rate": 2.5367153897822293e-07, + "loss": 0.8463, + "step": 9345 + }, + { + "epoch": 0.8428552103530685, + "grad_norm": 1.4807135950701642, + "learning_rate": 2.5338685396319715e-07, + "loss": 0.9686, + "step": 9346 + }, + { + "epoch": 0.8429453938765388, + "grad_norm": 4.708923820444851, + "learning_rate": 2.531023179788352e-07, + "loss": 0.9103, + "step": 9347 + }, + { + "epoch": 0.843035577400009, + "grad_norm": 1.6668581693619884, + "learning_rate": 2.528179310494158e-07, + "loss": 0.9481, + "step": 9348 + }, + { + "epoch": 0.8431257609234792, + "grad_norm": 1.7250832136787522, + "learning_rate": 2.5253369319920436e-07, + "loss": 0.9962, + "step": 9349 + }, + { + "epoch": 0.8432159444469496, + "grad_norm": 2.7122395051958303, + "learning_rate": 2.522496044524538e-07, + "loss": 0.9758, + "step": 9350 + }, + { + "epoch": 0.8433061279704198, + "grad_norm": 1.8027851597667957, + "learning_rate": 2.5196566483340386e-07, + "loss": 1.0684, + "step": 9351 + }, + { + "epoch": 0.84339631149389, + "grad_norm": 1.521760849323832, + "learning_rate": 2.516818743662825e-07, + "loss": 0.9911, + "step": 9352 + }, + { + "epoch": 0.8434864950173603, + "grad_norm": 1.6315586806513622, + "learning_rate": 2.5139823307530285e-07, + "loss": 0.9759, + "step": 9353 + }, + { + "epoch": 0.8435766785408306, + "grad_norm": 2.03474712103436, + "learning_rate": 2.5111474098466836e-07, + "loss": 1.0114, + "step": 9354 + }, + { + "epoch": 0.8436668620643009, + "grad_norm": 2.0709445204586303, + "learning_rate": 2.50831398118567e-07, + "loss": 1.0232, + "step": 9355 + }, + { + "epoch": 0.8437570455877711, + "grad_norm": 2.2925231604617786, + "learning_rate": 2.5054820450117576e-07, + "loss": 0.9192, + "step": 9356 + }, + { + "epoch": 0.8438472291112413, + "grad_norm": 1.565590358206211, + "learning_rate": 2.502651601566579e-07, + "loss": 0.9374, + "step": 9357 + }, + { + "epoch": 0.8439374126347117, + "grad_norm": 2.6316256692151936, + "learning_rate": 2.499822651091645e-07, + "loss": 0.9257, + "step": 9358 + }, + { + "epoch": 0.8440275961581819, + "grad_norm": 2.1825131931321287, + "learning_rate": 2.496995193828344e-07, + "loss": 0.997, + "step": 9359 + }, + { + "epoch": 0.8441177796816521, + "grad_norm": 2.199768756504257, + "learning_rate": 2.494169230017913e-07, + "loss": 0.9694, + "step": 9360 + }, + { + "epoch": 0.8442079632051224, + "grad_norm": 1.5278261013287766, + "learning_rate": 2.491344759901499e-07, + "loss": 0.9326, + "step": 9361 + }, + { + "epoch": 0.8442981467285927, + "grad_norm": 1.687172752014921, + "learning_rate": 2.488521783720088e-07, + "loss": 0.9694, + "step": 9362 + }, + { + "epoch": 0.844388330252063, + "grad_norm": 1.653967237902237, + "learning_rate": 2.4857003017145526e-07, + "loss": 1.0119, + "step": 9363 + }, + { + "epoch": 0.8444785137755332, + "grad_norm": 1.6199921691375598, + "learning_rate": 2.482880314125644e-07, + "loss": 1.0161, + "step": 9364 + }, + { + "epoch": 0.8445686972990035, + "grad_norm": 1.6715234404668184, + "learning_rate": 2.4800618211939726e-07, + "loss": 0.968, + "step": 9365 + }, + { + "epoch": 0.8446588808224738, + "grad_norm": 1.7574251553483013, + "learning_rate": 2.477244823160034e-07, + "loss": 0.9828, + "step": 9366 + }, + { + "epoch": 0.844749064345944, + "grad_norm": 1.783795127912639, + "learning_rate": 2.474429320264184e-07, + "loss": 0.994, + "step": 9367 + }, + { + "epoch": 0.8448392478694142, + "grad_norm": 1.462153698603274, + "learning_rate": 2.47161531274666e-07, + "loss": 0.9339, + "step": 9368 + }, + { + "epoch": 0.8449294313928846, + "grad_norm": 1.3879028422845678, + "learning_rate": 2.4688028008475714e-07, + "loss": 0.9069, + "step": 9369 + }, + { + "epoch": 0.8450196149163548, + "grad_norm": 1.8346855927413412, + "learning_rate": 2.465991784806891e-07, + "loss": 0.9682, + "step": 9370 + }, + { + "epoch": 0.845109798439825, + "grad_norm": 1.6745748378857404, + "learning_rate": 2.463182264864472e-07, + "loss": 1.0554, + "step": 9371 + }, + { + "epoch": 0.8451999819632953, + "grad_norm": 2.044753278105394, + "learning_rate": 2.460374241260039e-07, + "loss": 0.8983, + "step": 9372 + }, + { + "epoch": 0.8452901654867656, + "grad_norm": 1.68607672964058, + "learning_rate": 2.4575677142331884e-07, + "loss": 1.0438, + "step": 9373 + }, + { + "epoch": 0.8453803490102358, + "grad_norm": 1.8846626530323307, + "learning_rate": 2.454762684023395e-07, + "loss": 0.9165, + "step": 9374 + }, + { + "epoch": 0.8454705325337061, + "grad_norm": 1.7611062619730513, + "learning_rate": 2.4519591508699823e-07, + "loss": 0.9251, + "step": 9375 + }, + { + "epoch": 0.8455607160571763, + "grad_norm": 1.5532809848860003, + "learning_rate": 2.4491571150121815e-07, + "loss": 1.0231, + "step": 9376 + }, + { + "epoch": 0.8456508995806467, + "grad_norm": 2.0216932400778114, + "learning_rate": 2.446356576689062e-07, + "loss": 0.9701, + "step": 9377 + }, + { + "epoch": 0.8457410831041169, + "grad_norm": 1.6851944892092372, + "learning_rate": 2.4435575361395976e-07, + "loss": 0.9825, + "step": 9378 + }, + { + "epoch": 0.8458312666275871, + "grad_norm": 1.4698968832997026, + "learning_rate": 2.440759993602606e-07, + "loss": 0.9524, + "step": 9379 + }, + { + "epoch": 0.8459214501510574, + "grad_norm": 1.5657062440910314, + "learning_rate": 2.437963949316793e-07, + "loss": 0.9171, + "step": 9380 + }, + { + "epoch": 0.8460116336745277, + "grad_norm": 3.360465081361405, + "learning_rate": 2.435169403520729e-07, + "loss": 0.9678, + "step": 9381 + }, + { + "epoch": 0.8461018171979979, + "grad_norm": 2.223106620225705, + "learning_rate": 2.4323763564528653e-07, + "loss": 1.0212, + "step": 9382 + }, + { + "epoch": 0.8461920007214682, + "grad_norm": 1.6878257013294462, + "learning_rate": 2.429584808351517e-07, + "loss": 0.9968, + "step": 9383 + }, + { + "epoch": 0.8462821842449384, + "grad_norm": 1.9328705022646449, + "learning_rate": 2.42679475945488e-07, + "loss": 1.0987, + "step": 9384 + }, + { + "epoch": 0.8463723677684087, + "grad_norm": 1.364400846592113, + "learning_rate": 2.424006210001008e-07, + "loss": 0.975, + "step": 9385 + }, + { + "epoch": 0.846462551291879, + "grad_norm": 4.61762949317401, + "learning_rate": 2.421219160227839e-07, + "loss": 1.0069, + "step": 9386 + }, + { + "epoch": 0.8465527348153492, + "grad_norm": 2.927396565213605, + "learning_rate": 2.4184336103731785e-07, + "loss": 0.969, + "step": 9387 + }, + { + "epoch": 0.8466429183388196, + "grad_norm": 3.0006355566098506, + "learning_rate": 2.4156495606747065e-07, + "loss": 0.9966, + "step": 9388 + }, + { + "epoch": 0.8467331018622898, + "grad_norm": 1.7186924317862535, + "learning_rate": 2.412867011369972e-07, + "loss": 1.0925, + "step": 9389 + }, + { + "epoch": 0.84682328538576, + "grad_norm": 1.6287807352206445, + "learning_rate": 2.4100859626963997e-07, + "loss": 0.985, + "step": 9390 + }, + { + "epoch": 0.8469134689092302, + "grad_norm": 1.5884554563382096, + "learning_rate": 2.407306414891288e-07, + "loss": 0.9757, + "step": 9391 + }, + { + "epoch": 0.8470036524327006, + "grad_norm": 1.2563339452073194, + "learning_rate": 2.4045283681917893e-07, + "loss": 1.0789, + "step": 9392 + }, + { + "epoch": 0.8470938359561708, + "grad_norm": 1.4528557312577146, + "learning_rate": 2.4017518228349586e-07, + "loss": 1.0876, + "step": 9393 + }, + { + "epoch": 0.8471840194796411, + "grad_norm": 1.570335653713204, + "learning_rate": 2.3989767790576887e-07, + "loss": 0.8286, + "step": 9394 + }, + { + "epoch": 0.8472742030031113, + "grad_norm": 1.8917757425776127, + "learning_rate": 2.396203237096781e-07, + "loss": 0.944, + "step": 9395 + }, + { + "epoch": 0.8473643865265816, + "grad_norm": 1.6954379408876536, + "learning_rate": 2.393431197188873e-07, + "loss": 0.9674, + "step": 9396 + }, + { + "epoch": 0.8474545700500519, + "grad_norm": 3.6425375437585306, + "learning_rate": 2.3906606595705004e-07, + "loss": 1.0487, + "step": 9397 + }, + { + "epoch": 0.8475447535735221, + "grad_norm": 1.9647314445297976, + "learning_rate": 2.387891624478056e-07, + "loss": 1.026, + "step": 9398 + }, + { + "epoch": 0.8476349370969923, + "grad_norm": 1.514080164819108, + "learning_rate": 2.3851240921478075e-07, + "loss": 0.9968, + "step": 9399 + }, + { + "epoch": 0.8477251206204627, + "grad_norm": 1.5967775647322642, + "learning_rate": 2.3823580628159057e-07, + "loss": 0.957, + "step": 9400 + }, + { + "epoch": 0.8478153041439329, + "grad_norm": 2.980086035798584, + "learning_rate": 2.3795935367183517e-07, + "loss": 0.9606, + "step": 9401 + }, + { + "epoch": 0.8479054876674031, + "grad_norm": 1.408708796402084, + "learning_rate": 2.376830514091035e-07, + "loss": 0.9967, + "step": 9402 + }, + { + "epoch": 0.8479956711908734, + "grad_norm": 1.9776056725857372, + "learning_rate": 2.3740689951697135e-07, + "loss": 1.0543, + "step": 9403 + }, + { + "epoch": 0.8480858547143437, + "grad_norm": 1.4666717922959207, + "learning_rate": 2.371308980190012e-07, + "loss": 0.9758, + "step": 9404 + }, + { + "epoch": 0.848176038237814, + "grad_norm": 0.6278142986205385, + "learning_rate": 2.3685504693874337e-07, + "loss": 0.8352, + "step": 9405 + }, + { + "epoch": 0.8482662217612842, + "grad_norm": 2.352470349440162, + "learning_rate": 2.3657934629973497e-07, + "loss": 1.0095, + "step": 9406 + }, + { + "epoch": 0.8483564052847544, + "grad_norm": 2.1684226720742488, + "learning_rate": 2.3630379612549944e-07, + "loss": 0.9804, + "step": 9407 + }, + { + "epoch": 0.8484465888082248, + "grad_norm": 0.7196945121181697, + "learning_rate": 2.3602839643954997e-07, + "loss": 0.8586, + "step": 9408 + }, + { + "epoch": 0.848536772331695, + "grad_norm": 1.7424995587648027, + "learning_rate": 2.3575314726538308e-07, + "loss": 0.9799, + "step": 9409 + }, + { + "epoch": 0.8486269558551652, + "grad_norm": 1.4919769777715464, + "learning_rate": 2.3547804862648645e-07, + "loss": 1.0098, + "step": 9410 + }, + { + "epoch": 0.8487171393786356, + "grad_norm": 1.7229442398958486, + "learning_rate": 2.3520310054633174e-07, + "loss": 1.0234, + "step": 9411 + }, + { + "epoch": 0.8488073229021058, + "grad_norm": 1.7112786636571236, + "learning_rate": 2.3492830304837973e-07, + "loss": 0.9364, + "step": 9412 + }, + { + "epoch": 0.848897506425576, + "grad_norm": 1.5100033512457234, + "learning_rate": 2.3465365615607723e-07, + "loss": 0.9218, + "step": 9413 + }, + { + "epoch": 0.8489876899490463, + "grad_norm": 1.487740726386276, + "learning_rate": 2.3437915989285884e-07, + "loss": 0.967, + "step": 9414 + }, + { + "epoch": 0.8490778734725166, + "grad_norm": 1.8336688370480592, + "learning_rate": 2.3410481428214647e-07, + "loss": 1.0387, + "step": 9415 + }, + { + "epoch": 0.8491680569959869, + "grad_norm": 1.5714966214295691, + "learning_rate": 2.338306193473476e-07, + "loss": 1.0019, + "step": 9416 + }, + { + "epoch": 0.8492582405194571, + "grad_norm": 1.8735303841441375, + "learning_rate": 2.3355657511185957e-07, + "loss": 0.9771, + "step": 9417 + }, + { + "epoch": 0.8493484240429273, + "grad_norm": 0.6575331178454341, + "learning_rate": 2.3328268159906428e-07, + "loss": 0.7997, + "step": 9418 + }, + { + "epoch": 0.8494386075663977, + "grad_norm": 5.729282706270405, + "learning_rate": 2.330089388323322e-07, + "loss": 1.027, + "step": 9419 + }, + { + "epoch": 0.8495287910898679, + "grad_norm": 1.9977179859753122, + "learning_rate": 2.327353468350204e-07, + "loss": 1.0165, + "step": 9420 + }, + { + "epoch": 0.8496189746133381, + "grad_norm": 1.7742293014447779, + "learning_rate": 2.3246190563047352e-07, + "loss": 0.968, + "step": 9421 + }, + { + "epoch": 0.8497091581368084, + "grad_norm": 1.6598042193401326, + "learning_rate": 2.3218861524202293e-07, + "loss": 0.96, + "step": 9422 + }, + { + "epoch": 0.8497993416602787, + "grad_norm": 1.9164882936266932, + "learning_rate": 2.3191547569298775e-07, + "loss": 0.985, + "step": 9423 + }, + { + "epoch": 0.8498895251837489, + "grad_norm": 1.6780132377247845, + "learning_rate": 2.3164248700667245e-07, + "loss": 0.9618, + "step": 9424 + }, + { + "epoch": 0.8499797087072192, + "grad_norm": 2.019290164451239, + "learning_rate": 2.313696492063717e-07, + "loss": 1.0078, + "step": 9425 + }, + { + "epoch": 0.8500698922306894, + "grad_norm": 2.092097754971651, + "learning_rate": 2.3109696231536401e-07, + "loss": 0.9649, + "step": 9426 + }, + { + "epoch": 0.8501600757541597, + "grad_norm": 1.9338157449308369, + "learning_rate": 2.3082442635691722e-07, + "loss": 0.9399, + "step": 9427 + }, + { + "epoch": 0.85025025927763, + "grad_norm": 1.6034997899787178, + "learning_rate": 2.305520413542854e-07, + "loss": 0.9916, + "step": 9428 + }, + { + "epoch": 0.8503404428011002, + "grad_norm": 1.6233301386060879, + "learning_rate": 2.3027980733071018e-07, + "loss": 0.9013, + "step": 9429 + }, + { + "epoch": 0.8504306263245704, + "grad_norm": 0.5974693255351645, + "learning_rate": 2.3000772430942027e-07, + "loss": 0.7819, + "step": 9430 + }, + { + "epoch": 0.8505208098480408, + "grad_norm": 1.5642594921280704, + "learning_rate": 2.2973579231363028e-07, + "loss": 0.9913, + "step": 9431 + }, + { + "epoch": 0.850610993371511, + "grad_norm": 1.5407117343343075, + "learning_rate": 2.2946401136654446e-07, + "loss": 1.0018, + "step": 9432 + }, + { + "epoch": 0.8507011768949813, + "grad_norm": 1.434181196961874, + "learning_rate": 2.2919238149135077e-07, + "loss": 1.0031, + "step": 9433 + }, + { + "epoch": 0.8507913604184516, + "grad_norm": 1.4217151090829179, + "learning_rate": 2.289209027112282e-07, + "loss": 0.9814, + "step": 9434 + }, + { + "epoch": 0.8508815439419218, + "grad_norm": 8.551119850393912, + "learning_rate": 2.2864957504933934e-07, + "loss": 1.036, + "step": 9435 + }, + { + "epoch": 0.8509717274653921, + "grad_norm": 1.7057351925187774, + "learning_rate": 2.2837839852883589e-07, + "loss": 0.96, + "step": 9436 + }, + { + "epoch": 0.8510619109888623, + "grad_norm": 1.4738431240747178, + "learning_rate": 2.2810737317285623e-07, + "loss": 0.9709, + "step": 9437 + }, + { + "epoch": 0.8511520945123326, + "grad_norm": 1.724018345582538, + "learning_rate": 2.278364990045254e-07, + "loss": 0.9167, + "step": 9438 + }, + { + "epoch": 0.8512422780358029, + "grad_norm": 2.0926857136449684, + "learning_rate": 2.2756577604695625e-07, + "loss": 0.9692, + "step": 9439 + }, + { + "epoch": 0.8513324615592731, + "grad_norm": 1.456181500833516, + "learning_rate": 2.2729520432324855e-07, + "loss": 0.9946, + "step": 9440 + }, + { + "epoch": 0.8514226450827433, + "grad_norm": 2.2146284479789506, + "learning_rate": 2.2702478385648826e-07, + "loss": 0.8935, + "step": 9441 + }, + { + "epoch": 0.8515128286062137, + "grad_norm": 1.5842935845939883, + "learning_rate": 2.2675451466974938e-07, + "loss": 0.9761, + "step": 9442 + }, + { + "epoch": 0.8516030121296839, + "grad_norm": 1.5826081154337428, + "learning_rate": 2.26484396786093e-07, + "loss": 0.9097, + "step": 9443 + }, + { + "epoch": 0.8516931956531542, + "grad_norm": 0.64281482453174, + "learning_rate": 2.2621443022856667e-07, + "loss": 0.8399, + "step": 9444 + }, + { + "epoch": 0.8517833791766244, + "grad_norm": 2.5587999566535413, + "learning_rate": 2.2594461502020646e-07, + "loss": 1.0172, + "step": 9445 + }, + { + "epoch": 0.8518735627000947, + "grad_norm": 1.7319765432456495, + "learning_rate": 2.2567495118403278e-07, + "loss": 0.8536, + "step": 9446 + }, + { + "epoch": 0.851963746223565, + "grad_norm": 1.8932594041539148, + "learning_rate": 2.254054387430566e-07, + "loss": 0.9921, + "step": 9447 + }, + { + "epoch": 0.8520539297470352, + "grad_norm": 1.6249712427969278, + "learning_rate": 2.2513607772027243e-07, + "loss": 0.9612, + "step": 9448 + }, + { + "epoch": 0.8521441132705054, + "grad_norm": 2.007102276654131, + "learning_rate": 2.2486686813866562e-07, + "loss": 0.9405, + "step": 9449 + }, + { + "epoch": 0.8522342967939758, + "grad_norm": 1.6939713759227688, + "learning_rate": 2.2459781002120514e-07, + "loss": 1.0283, + "step": 9450 + }, + { + "epoch": 0.852324480317446, + "grad_norm": 3.9671881298231457, + "learning_rate": 2.243289033908491e-07, + "loss": 1.0448, + "step": 9451 + }, + { + "epoch": 0.8524146638409162, + "grad_norm": 2.484879928236265, + "learning_rate": 2.2406014827054176e-07, + "loss": 0.9906, + "step": 9452 + }, + { + "epoch": 0.8525048473643865, + "grad_norm": 1.8411062003424277, + "learning_rate": 2.2379154468321525e-07, + "loss": 1.0373, + "step": 9453 + }, + { + "epoch": 0.8525950308878568, + "grad_norm": 1.4012578728868983, + "learning_rate": 2.2352309265178793e-07, + "loss": 1.0087, + "step": 9454 + }, + { + "epoch": 0.852685214411327, + "grad_norm": 1.9386204586862572, + "learning_rate": 2.2325479219916565e-07, + "loss": 1.0468, + "step": 9455 + }, + { + "epoch": 0.8527753979347973, + "grad_norm": 1.669088947862964, + "learning_rate": 2.229866433482419e-07, + "loss": 1.0098, + "step": 9456 + }, + { + "epoch": 0.8528655814582676, + "grad_norm": 1.6256147835603703, + "learning_rate": 2.2271864612189552e-07, + "loss": 1.0254, + "step": 9457 + }, + { + "epoch": 0.8529557649817379, + "grad_norm": 1.8059080546008879, + "learning_rate": 2.2245080054299415e-07, + "loss": 1.0084, + "step": 9458 + }, + { + "epoch": 0.8530459485052081, + "grad_norm": 2.0868522552777105, + "learning_rate": 2.2218310663439198e-07, + "loss": 0.9268, + "step": 9459 + }, + { + "epoch": 0.8531361320286783, + "grad_norm": 1.6114129061502722, + "learning_rate": 2.2191556441892968e-07, + "loss": 0.9986, + "step": 9460 + }, + { + "epoch": 0.8532263155521487, + "grad_norm": 1.6144023998976111, + "learning_rate": 2.216481739194358e-07, + "loss": 0.9823, + "step": 9461 + }, + { + "epoch": 0.8533164990756189, + "grad_norm": 3.7649207505273554, + "learning_rate": 2.2138093515872592e-07, + "loss": 0.9273, + "step": 9462 + }, + { + "epoch": 0.8534066825990891, + "grad_norm": 5.261234252645101, + "learning_rate": 2.2111384815960087e-07, + "loss": 1.056, + "step": 9463 + }, + { + "epoch": 0.8534968661225594, + "grad_norm": 1.573845528028005, + "learning_rate": 2.208469129448518e-07, + "loss": 0.9016, + "step": 9464 + }, + { + "epoch": 0.8535870496460297, + "grad_norm": 0.6242706811952355, + "learning_rate": 2.2058012953725357e-07, + "loss": 0.8059, + "step": 9465 + }, + { + "epoch": 0.8536772331695, + "grad_norm": 1.7486389862227663, + "learning_rate": 2.203134979595711e-07, + "loss": 0.9473, + "step": 9466 + }, + { + "epoch": 0.8537674166929702, + "grad_norm": 1.8209988196598021, + "learning_rate": 2.2004701823455374e-07, + "loss": 0.9289, + "step": 9467 + }, + { + "epoch": 0.8538576002164404, + "grad_norm": 2.1144769074917025, + "learning_rate": 2.1978069038493906e-07, + "loss": 1.0074, + "step": 9468 + }, + { + "epoch": 0.8539477837399108, + "grad_norm": 1.3107122325472789, + "learning_rate": 2.1951451443345225e-07, + "loss": 0.9797, + "step": 9469 + }, + { + "epoch": 0.854037967263381, + "grad_norm": 1.5309292204259808, + "learning_rate": 2.1924849040280425e-07, + "loss": 0.9626, + "step": 9470 + }, + { + "epoch": 0.8541281507868512, + "grad_norm": 2.033214147649636, + "learning_rate": 2.1898261831569465e-07, + "loss": 1.0609, + "step": 9471 + }, + { + "epoch": 0.8542183343103215, + "grad_norm": 1.4587807468550644, + "learning_rate": 2.1871689819480798e-07, + "loss": 0.9866, + "step": 9472 + }, + { + "epoch": 0.8543085178337918, + "grad_norm": 1.3864893763864616, + "learning_rate": 2.1845133006281745e-07, + "loss": 0.9644, + "step": 9473 + }, + { + "epoch": 0.854398701357262, + "grad_norm": 1.4028371723849624, + "learning_rate": 2.1818591394238294e-07, + "loss": 0.8338, + "step": 9474 + }, + { + "epoch": 0.8544888848807323, + "grad_norm": 1.8042424962037176, + "learning_rate": 2.1792064985615076e-07, + "loss": 0.9593, + "step": 9475 + }, + { + "epoch": 0.8545790684042025, + "grad_norm": 1.4368595686951307, + "learning_rate": 2.1765553782675528e-07, + "loss": 0.9493, + "step": 9476 + }, + { + "epoch": 0.8546692519276728, + "grad_norm": 1.4701763288659506, + "learning_rate": 2.1739057787681703e-07, + "loss": 1.052, + "step": 9477 + }, + { + "epoch": 0.8547594354511431, + "grad_norm": 1.5763740268105644, + "learning_rate": 2.1712577002894372e-07, + "loss": 0.9991, + "step": 9478 + }, + { + "epoch": 0.8548496189746133, + "grad_norm": 1.611462322773482, + "learning_rate": 2.1686111430573105e-07, + "loss": 0.9511, + "step": 9479 + }, + { + "epoch": 0.8549398024980835, + "grad_norm": 1.6317814396102481, + "learning_rate": 2.165966107297592e-07, + "loss": 1.0299, + "step": 9480 + }, + { + "epoch": 0.8550299860215539, + "grad_norm": 1.653728509072747, + "learning_rate": 2.16332259323599e-07, + "loss": 1.0044, + "step": 9481 + }, + { + "epoch": 0.8551201695450241, + "grad_norm": 2.1808342873621385, + "learning_rate": 2.1606806010980504e-07, + "loss": 0.9234, + "step": 9482 + }, + { + "epoch": 0.8552103530684944, + "grad_norm": 1.7478345688767352, + "learning_rate": 2.1580401311092067e-07, + "loss": 0.9737, + "step": 9483 + }, + { + "epoch": 0.8553005365919647, + "grad_norm": 23.44190583589035, + "learning_rate": 2.1554011834947604e-07, + "loss": 0.9685, + "step": 9484 + }, + { + "epoch": 0.8553907201154349, + "grad_norm": 2.3208248715842843, + "learning_rate": 2.1527637584798764e-07, + "loss": 1.0074, + "step": 9485 + }, + { + "epoch": 0.8554809036389052, + "grad_norm": 1.738153963036287, + "learning_rate": 2.150127856289603e-07, + "loss": 1.0445, + "step": 9486 + }, + { + "epoch": 0.8555710871623754, + "grad_norm": 0.6191606488284096, + "learning_rate": 2.1474934771488363e-07, + "loss": 0.7941, + "step": 9487 + }, + { + "epoch": 0.8556612706858457, + "grad_norm": 1.67779825108958, + "learning_rate": 2.1448606212823715e-07, + "loss": 1.022, + "step": 9488 + }, + { + "epoch": 0.855751454209316, + "grad_norm": 2.3146662489228524, + "learning_rate": 2.1422292889148452e-07, + "loss": 0.9598, + "step": 9489 + }, + { + "epoch": 0.8558416377327862, + "grad_norm": 1.4073572250369948, + "learning_rate": 2.139599480270784e-07, + "loss": 0.9959, + "step": 9490 + }, + { + "epoch": 0.8559318212562564, + "grad_norm": 1.9983873422157672, + "learning_rate": 2.1369711955745773e-07, + "loss": 0.9664, + "step": 9491 + }, + { + "epoch": 0.8560220047797268, + "grad_norm": 1.5889998312026379, + "learning_rate": 2.1343444350504813e-07, + "loss": 0.9332, + "step": 9492 + }, + { + "epoch": 0.856112188303197, + "grad_norm": 1.478929912621671, + "learning_rate": 2.1317191989226302e-07, + "loss": 0.979, + "step": 9493 + }, + { + "epoch": 0.8562023718266673, + "grad_norm": 1.812161183752441, + "learning_rate": 2.129095487415027e-07, + "loss": 0.9602, + "step": 9494 + }, + { + "epoch": 0.8562925553501375, + "grad_norm": 1.855383427314558, + "learning_rate": 2.1264733007515257e-07, + "loss": 0.9888, + "step": 9495 + }, + { + "epoch": 0.8563827388736078, + "grad_norm": 1.7083434903222812, + "learning_rate": 2.1238526391558852e-07, + "loss": 0.9575, + "step": 9496 + }, + { + "epoch": 0.8564729223970781, + "grad_norm": 2.1556939776481623, + "learning_rate": 2.1212335028517003e-07, + "loss": 1.1157, + "step": 9497 + }, + { + "epoch": 0.8565631059205483, + "grad_norm": 1.4897412925500313, + "learning_rate": 2.1186158920624563e-07, + "loss": 0.9841, + "step": 9498 + }, + { + "epoch": 0.8566532894440185, + "grad_norm": 1.6890034512653402, + "learning_rate": 2.1159998070115015e-07, + "loss": 0.9773, + "step": 9499 + }, + { + "epoch": 0.8567434729674889, + "grad_norm": 1.520862613901252, + "learning_rate": 2.113385247922055e-07, + "loss": 0.9594, + "step": 9500 + }, + { + "epoch": 0.8568336564909591, + "grad_norm": 1.493219627808501, + "learning_rate": 2.1107722150172068e-07, + "loss": 0.9864, + "step": 9501 + }, + { + "epoch": 0.8569238400144293, + "grad_norm": 1.4667625778800688, + "learning_rate": 2.108160708519906e-07, + "loss": 1.0158, + "step": 9502 + }, + { + "epoch": 0.8570140235378996, + "grad_norm": 1.8269231383134468, + "learning_rate": 2.1055507286529984e-07, + "loss": 0.959, + "step": 9503 + }, + { + "epoch": 0.8571042070613699, + "grad_norm": 1.6399464248419606, + "learning_rate": 2.1029422756391612e-07, + "loss": 1.0012, + "step": 9504 + }, + { + "epoch": 0.8571943905848401, + "grad_norm": 1.3972665157000672, + "learning_rate": 2.1003353497009812e-07, + "loss": 0.8915, + "step": 9505 + }, + { + "epoch": 0.8572845741083104, + "grad_norm": 2.005592633037024, + "learning_rate": 2.0977299510608825e-07, + "loss": 0.9014, + "step": 9506 + }, + { + "epoch": 0.8573747576317807, + "grad_norm": 1.535100183603647, + "learning_rate": 2.0951260799411784e-07, + "loss": 1.012, + "step": 9507 + }, + { + "epoch": 0.857464941155251, + "grad_norm": 1.521893232107125, + "learning_rate": 2.0925237365640424e-07, + "loss": 0.9668, + "step": 9508 + }, + { + "epoch": 0.8575551246787212, + "grad_norm": 0.6419566201203895, + "learning_rate": 2.0899229211515211e-07, + "loss": 0.813, + "step": 9509 + }, + { + "epoch": 0.8576453082021914, + "grad_norm": 1.7775335524942757, + "learning_rate": 2.0873236339255306e-07, + "loss": 0.9006, + "step": 9510 + }, + { + "epoch": 0.8577354917256618, + "grad_norm": 1.454440024331719, + "learning_rate": 2.0847258751078644e-07, + "loss": 0.9228, + "step": 9511 + }, + { + "epoch": 0.857825675249132, + "grad_norm": 1.9425926344494768, + "learning_rate": 2.082129644920163e-07, + "loss": 0.9565, + "step": 9512 + }, + { + "epoch": 0.8579158587726022, + "grad_norm": 1.624901915304349, + "learning_rate": 2.0795349435839605e-07, + "loss": 0.9875, + "step": 9513 + }, + { + "epoch": 0.8580060422960725, + "grad_norm": 1.8940401364900517, + "learning_rate": 2.0769417713206484e-07, + "loss": 0.9499, + "step": 9514 + }, + { + "epoch": 0.8580962258195428, + "grad_norm": 1.33386480150175, + "learning_rate": 2.074350128351492e-07, + "loss": 1.0408, + "step": 9515 + }, + { + "epoch": 0.858186409343013, + "grad_norm": 1.5394637461027054, + "learning_rate": 2.0717600148976256e-07, + "loss": 0.8961, + "step": 9516 + }, + { + "epoch": 0.8582765928664833, + "grad_norm": 1.6279689369893537, + "learning_rate": 2.0691714311800436e-07, + "loss": 0.9106, + "step": 9517 + }, + { + "epoch": 0.8583667763899535, + "grad_norm": 1.596082065743033, + "learning_rate": 2.066584377419631e-07, + "loss": 1.0355, + "step": 9518 + }, + { + "epoch": 0.8584569599134239, + "grad_norm": 1.484876126475597, + "learning_rate": 2.0639988538371167e-07, + "loss": 1.0005, + "step": 9519 + }, + { + "epoch": 0.8585471434368941, + "grad_norm": 3.5584107068216353, + "learning_rate": 2.0614148606531258e-07, + "loss": 0.9441, + "step": 9520 + }, + { + "epoch": 0.8586373269603643, + "grad_norm": 2.046834930726603, + "learning_rate": 2.0588323980881285e-07, + "loss": 1.005, + "step": 9521 + }, + { + "epoch": 0.8587275104838346, + "grad_norm": 1.4616147679706686, + "learning_rate": 2.0562514663624752e-07, + "loss": 0.9639, + "step": 9522 + }, + { + "epoch": 0.8588176940073049, + "grad_norm": 1.3862445289084202, + "learning_rate": 2.0536720656963902e-07, + "loss": 1.0414, + "step": 9523 + }, + { + "epoch": 0.8589078775307751, + "grad_norm": 1.8139745042012276, + "learning_rate": 2.051094196309957e-07, + "loss": 1.0029, + "step": 9524 + }, + { + "epoch": 0.8589980610542454, + "grad_norm": 1.5849684587583404, + "learning_rate": 2.0485178584231378e-07, + "loss": 0.9345, + "step": 9525 + }, + { + "epoch": 0.8590882445777156, + "grad_norm": 0.6714429797094794, + "learning_rate": 2.0459430522557587e-07, + "loss": 0.8144, + "step": 9526 + }, + { + "epoch": 0.8591784281011859, + "grad_norm": 1.8605328969612573, + "learning_rate": 2.0433697780275195e-07, + "loss": 0.9819, + "step": 9527 + }, + { + "epoch": 0.8592686116246562, + "grad_norm": 1.677184826482892, + "learning_rate": 2.040798035957978e-07, + "loss": 0.9945, + "step": 9528 + }, + { + "epoch": 0.8593587951481264, + "grad_norm": 0.7228208199833214, + "learning_rate": 2.038227826266574e-07, + "loss": 0.8306, + "step": 9529 + }, + { + "epoch": 0.8594489786715968, + "grad_norm": 1.7163390155554201, + "learning_rate": 2.0356591491726126e-07, + "loss": 0.9866, + "step": 9530 + }, + { + "epoch": 0.859539162195067, + "grad_norm": 1.4687850623308272, + "learning_rate": 2.033092004895267e-07, + "loss": 1.0203, + "step": 9531 + }, + { + "epoch": 0.8596293457185372, + "grad_norm": 1.5499908969981009, + "learning_rate": 2.03052639365358e-07, + "loss": 0.997, + "step": 9532 + }, + { + "epoch": 0.8597195292420075, + "grad_norm": 1.4420292032805764, + "learning_rate": 2.0279623156664694e-07, + "loss": 0.9972, + "step": 9533 + }, + { + "epoch": 0.8598097127654778, + "grad_norm": 1.6047256123475042, + "learning_rate": 2.0253997711527005e-07, + "loss": 1.037, + "step": 9534 + }, + { + "epoch": 0.859899896288948, + "grad_norm": 2.259608521271567, + "learning_rate": 2.0228387603309428e-07, + "loss": 1.0276, + "step": 9535 + }, + { + "epoch": 0.8599900798124183, + "grad_norm": 1.454875506944631, + "learning_rate": 2.0202792834197035e-07, + "loss": 1.0007, + "step": 9536 + }, + { + "epoch": 0.8600802633358885, + "grad_norm": 1.8447992604331116, + "learning_rate": 2.017721340637375e-07, + "loss": 0.9074, + "step": 9537 + }, + { + "epoch": 0.8601704468593588, + "grad_norm": 1.566761157385513, + "learning_rate": 2.0151649322022134e-07, + "loss": 0.9618, + "step": 9538 + }, + { + "epoch": 0.8602606303828291, + "grad_norm": 2.37651706228265, + "learning_rate": 2.012610058332349e-07, + "loss": 0.7901, + "step": 9539 + }, + { + "epoch": 0.8603508139062993, + "grad_norm": 0.6725800607312006, + "learning_rate": 2.010056719245774e-07, + "loss": 0.8111, + "step": 9540 + }, + { + "epoch": 0.8604409974297695, + "grad_norm": 1.6151291283633769, + "learning_rate": 2.0075049151603563e-07, + "loss": 1.0029, + "step": 9541 + }, + { + "epoch": 0.8605311809532399, + "grad_norm": 1.6383337097064052, + "learning_rate": 2.0049546462938326e-07, + "loss": 0.9839, + "step": 9542 + }, + { + "epoch": 0.8606213644767101, + "grad_norm": 1.8642083186937721, + "learning_rate": 2.0024059128637961e-07, + "loss": 1.0017, + "step": 9543 + }, + { + "epoch": 0.8607115480001803, + "grad_norm": 1.960770497296779, + "learning_rate": 1.99985871508773e-07, + "loss": 0.8418, + "step": 9544 + }, + { + "epoch": 0.8608017315236506, + "grad_norm": 1.8015308599017856, + "learning_rate": 1.9973130531829674e-07, + "loss": 1.0042, + "step": 9545 + }, + { + "epoch": 0.8608919150471209, + "grad_norm": 1.739950685679218, + "learning_rate": 1.994768927366721e-07, + "loss": 0.929, + "step": 9546 + }, + { + "epoch": 0.8609820985705912, + "grad_norm": 1.7028711812280979, + "learning_rate": 1.992226337856069e-07, + "loss": 0.9491, + "step": 9547 + }, + { + "epoch": 0.8610722820940614, + "grad_norm": 1.7274268562963122, + "learning_rate": 1.9896852848679592e-07, + "loss": 0.9334, + "step": 9548 + }, + { + "epoch": 0.8611624656175316, + "grad_norm": 1.6575556952553767, + "learning_rate": 1.9871457686192094e-07, + "loss": 0.9886, + "step": 9549 + }, + { + "epoch": 0.861252649141002, + "grad_norm": 2.770794191777158, + "learning_rate": 1.984607789326509e-07, + "loss": 0.9741, + "step": 9550 + }, + { + "epoch": 0.8613428326644722, + "grad_norm": 1.3951310008760496, + "learning_rate": 1.982071347206402e-07, + "loss": 0.9275, + "step": 9551 + }, + { + "epoch": 0.8614330161879424, + "grad_norm": 1.4052171516942968, + "learning_rate": 1.9795364424753202e-07, + "loss": 0.9537, + "step": 9552 + }, + { + "epoch": 0.8615231997114128, + "grad_norm": 1.7713970259525296, + "learning_rate": 1.9770030753495505e-07, + "loss": 0.9477, + "step": 9553 + }, + { + "epoch": 0.861613383234883, + "grad_norm": 1.8391220718367738, + "learning_rate": 1.9744712460452573e-07, + "loss": 0.9353, + "step": 9554 + }, + { + "epoch": 0.8617035667583532, + "grad_norm": 1.7926261757424584, + "learning_rate": 1.9719409547784703e-07, + "loss": 0.9581, + "step": 9555 + }, + { + "epoch": 0.8617937502818235, + "grad_norm": 2.221846340761842, + "learning_rate": 1.9694122017650837e-07, + "loss": 1.0011, + "step": 9556 + }, + { + "epoch": 0.8618839338052938, + "grad_norm": 6.682001152671023, + "learning_rate": 1.9668849872208738e-07, + "loss": 0.9905, + "step": 9557 + }, + { + "epoch": 0.8619741173287641, + "grad_norm": 1.4158630709849058, + "learning_rate": 1.9643593113614632e-07, + "loss": 0.8256, + "step": 9558 + }, + { + "epoch": 0.8620643008522343, + "grad_norm": 1.457964396357309, + "learning_rate": 1.961835174402371e-07, + "loss": 1.0165, + "step": 9559 + }, + { + "epoch": 0.8621544843757045, + "grad_norm": 1.803078863389996, + "learning_rate": 1.9593125765589535e-07, + "loss": 0.9418, + "step": 9560 + }, + { + "epoch": 0.8622446678991749, + "grad_norm": 1.5830742306104293, + "learning_rate": 1.9567915180464721e-07, + "loss": 0.9244, + "step": 9561 + }, + { + "epoch": 0.8623348514226451, + "grad_norm": 1.8845122962645637, + "learning_rate": 1.9542719990800217e-07, + "loss": 0.9641, + "step": 9562 + }, + { + "epoch": 0.8624250349461153, + "grad_norm": 1.7065741933168104, + "learning_rate": 1.9517540198745896e-07, + "loss": 1.046, + "step": 9563 + }, + { + "epoch": 0.8625152184695856, + "grad_norm": 1.8612486431047708, + "learning_rate": 1.94923758064502e-07, + "loss": 0.9762, + "step": 9564 + }, + { + "epoch": 0.8626054019930559, + "grad_norm": 2.3928775522138435, + "learning_rate": 1.9467226816060322e-07, + "loss": 0.944, + "step": 9565 + }, + { + "epoch": 0.8626955855165261, + "grad_norm": 1.8111330336561657, + "learning_rate": 1.9442093229722122e-07, + "loss": 0.9195, + "step": 9566 + }, + { + "epoch": 0.8627857690399964, + "grad_norm": 1.9847232182669834, + "learning_rate": 1.9416975049580085e-07, + "loss": 1.0483, + "step": 9567 + }, + { + "epoch": 0.8628759525634666, + "grad_norm": 1.7062649960260623, + "learning_rate": 1.9391872277777456e-07, + "loss": 0.9618, + "step": 9568 + }, + { + "epoch": 0.862966136086937, + "grad_norm": 2.0935823743533772, + "learning_rate": 1.9366784916456158e-07, + "loss": 0.9665, + "step": 9569 + }, + { + "epoch": 0.8630563196104072, + "grad_norm": 1.7071866940184492, + "learning_rate": 1.9341712967756774e-07, + "loss": 0.9325, + "step": 9570 + }, + { + "epoch": 0.8631465031338774, + "grad_norm": 1.6109196925543419, + "learning_rate": 1.9316656433818566e-07, + "loss": 0.8893, + "step": 9571 + }, + { + "epoch": 0.8632366866573477, + "grad_norm": 1.4842408772156561, + "learning_rate": 1.929161531677954e-07, + "loss": 0.877, + "step": 9572 + }, + { + "epoch": 0.863326870180818, + "grad_norm": 1.511273143455424, + "learning_rate": 1.9266589618776251e-07, + "loss": 0.9313, + "step": 9573 + }, + { + "epoch": 0.8634170537042882, + "grad_norm": 2.0835247587632932, + "learning_rate": 1.924157934194417e-07, + "loss": 1.0273, + "step": 9574 + }, + { + "epoch": 0.8635072372277585, + "grad_norm": 1.7134950239177722, + "learning_rate": 1.9216584488417142e-07, + "loss": 0.8697, + "step": 9575 + }, + { + "epoch": 0.8635974207512288, + "grad_norm": 2.6067516365576613, + "learning_rate": 1.919160506032802e-07, + "loss": 1.0344, + "step": 9576 + }, + { + "epoch": 0.863687604274699, + "grad_norm": 1.6486676674647107, + "learning_rate": 1.916664105980812e-07, + "loss": 1.0384, + "step": 9577 + }, + { + "epoch": 0.8637777877981693, + "grad_norm": 1.5268439136015428, + "learning_rate": 1.914169248898747e-07, + "loss": 1.0118, + "step": 9578 + }, + { + "epoch": 0.8638679713216395, + "grad_norm": 1.691078630041607, + "learning_rate": 1.9116759349994882e-07, + "loss": 0.8933, + "step": 9579 + }, + { + "epoch": 0.8639581548451098, + "grad_norm": 1.5436495667193262, + "learning_rate": 1.9091841644957763e-07, + "loss": 1.0133, + "step": 9580 + }, + { + "epoch": 0.8640483383685801, + "grad_norm": 1.3148382793092843, + "learning_rate": 1.9066939376002278e-07, + "loss": 1.059, + "step": 9581 + }, + { + "epoch": 0.8641385218920503, + "grad_norm": 1.9902707322143778, + "learning_rate": 1.9042052545253085e-07, + "loss": 0.8939, + "step": 9582 + }, + { + "epoch": 0.8642287054155205, + "grad_norm": 1.478579319497818, + "learning_rate": 1.901718115483384e-07, + "loss": 0.956, + "step": 9583 + }, + { + "epoch": 0.8643188889389909, + "grad_norm": 1.5185944088956893, + "learning_rate": 1.8992325206866598e-07, + "loss": 0.9566, + "step": 9584 + }, + { + "epoch": 0.8644090724624611, + "grad_norm": 1.3265911445176033, + "learning_rate": 1.8967484703472225e-07, + "loss": 0.9509, + "step": 9585 + }, + { + "epoch": 0.8644992559859314, + "grad_norm": 1.3060494812008352, + "learning_rate": 1.8942659646770288e-07, + "loss": 1.0235, + "step": 9586 + }, + { + "epoch": 0.8645894395094016, + "grad_norm": 2.1504919598200054, + "learning_rate": 1.8917850038878936e-07, + "loss": 1.0454, + "step": 9587 + }, + { + "epoch": 0.8646796230328719, + "grad_norm": 1.526573904223291, + "learning_rate": 1.8893055881915121e-07, + "loss": 0.8847, + "step": 9588 + }, + { + "epoch": 0.8647698065563422, + "grad_norm": 1.746363608310781, + "learning_rate": 1.886827717799442e-07, + "loss": 0.9227, + "step": 9589 + }, + { + "epoch": 0.8648599900798124, + "grad_norm": 1.7393951618809351, + "learning_rate": 1.884351392923096e-07, + "loss": 0.9935, + "step": 9590 + }, + { + "epoch": 0.8649501736032826, + "grad_norm": 2.2661030232173287, + "learning_rate": 1.8818766137737896e-07, + "loss": 1.0173, + "step": 9591 + }, + { + "epoch": 0.865040357126753, + "grad_norm": 1.8536185265166356, + "learning_rate": 1.8794033805626653e-07, + "loss": 0.9531, + "step": 9592 + }, + { + "epoch": 0.8651305406502232, + "grad_norm": 1.732013572564807, + "learning_rate": 1.876931693500763e-07, + "loss": 0.9707, + "step": 9593 + }, + { + "epoch": 0.8652207241736934, + "grad_norm": 2.0880942178493687, + "learning_rate": 1.8744615527989783e-07, + "loss": 0.8253, + "step": 9594 + }, + { + "epoch": 0.8653109076971637, + "grad_norm": 1.7244632438708776, + "learning_rate": 1.871992958668076e-07, + "loss": 1.0648, + "step": 9595 + }, + { + "epoch": 0.865401091220634, + "grad_norm": 1.8278722995961674, + "learning_rate": 1.8695259113186944e-07, + "loss": 1.0158, + "step": 9596 + }, + { + "epoch": 0.8654912747441043, + "grad_norm": 1.4398883384221235, + "learning_rate": 1.8670604109613252e-07, + "loss": 0.9406, + "step": 9597 + }, + { + "epoch": 0.8655814582675745, + "grad_norm": 1.7839321221537625, + "learning_rate": 1.8645964578063533e-07, + "loss": 1.0215, + "step": 9598 + }, + { + "epoch": 0.8656716417910447, + "grad_norm": 1.5681841646712167, + "learning_rate": 1.862134052064006e-07, + "loss": 1.0337, + "step": 9599 + }, + { + "epoch": 0.8657618253145151, + "grad_norm": 2.267515143834932, + "learning_rate": 1.8596731939443932e-07, + "loss": 1.0149, + "step": 9600 + }, + { + "epoch": 0.8658520088379853, + "grad_norm": 1.55500874483158, + "learning_rate": 1.857213883657487e-07, + "loss": 1.0048, + "step": 9601 + }, + { + "epoch": 0.8659421923614555, + "grad_norm": 0.6335717344398516, + "learning_rate": 1.8547561214131303e-07, + "loss": 0.7876, + "step": 9602 + }, + { + "epoch": 0.8660323758849259, + "grad_norm": 1.7687312121791965, + "learning_rate": 1.8522999074210355e-07, + "loss": 1.0551, + "step": 9603 + }, + { + "epoch": 0.8661225594083961, + "grad_norm": 1.7117777115640258, + "learning_rate": 1.849845241890775e-07, + "loss": 1.0137, + "step": 9604 + }, + { + "epoch": 0.8662127429318663, + "grad_norm": 1.60726056500462, + "learning_rate": 1.8473921250317992e-07, + "loss": 0.9795, + "step": 9605 + }, + { + "epoch": 0.8663029264553366, + "grad_norm": 1.639253242674473, + "learning_rate": 1.8449405570534225e-07, + "loss": 0.91, + "step": 9606 + }, + { + "epoch": 0.8663931099788069, + "grad_norm": 1.4763507994009573, + "learning_rate": 1.8424905381648204e-07, + "loss": 1.0201, + "step": 9607 + }, + { + "epoch": 0.8664832935022772, + "grad_norm": 1.6366950058409708, + "learning_rate": 1.8400420685750452e-07, + "loss": 0.9331, + "step": 9608 + }, + { + "epoch": 0.8665734770257474, + "grad_norm": 1.5989907061585278, + "learning_rate": 1.8375951484930142e-07, + "loss": 0.9152, + "step": 9609 + }, + { + "epoch": 0.8666636605492176, + "grad_norm": 1.4846448954108051, + "learning_rate": 1.8351497781275094e-07, + "loss": 0.9809, + "step": 9610 + }, + { + "epoch": 0.866753844072688, + "grad_norm": 1.686205473383365, + "learning_rate": 1.8327059576871907e-07, + "loss": 0.967, + "step": 9611 + }, + { + "epoch": 0.8668440275961582, + "grad_norm": 1.482244950945859, + "learning_rate": 1.8302636873805665e-07, + "loss": 1.0246, + "step": 9612 + }, + { + "epoch": 0.8669342111196284, + "grad_norm": 1.5287348273484263, + "learning_rate": 1.8278229674160373e-07, + "loss": 1.0558, + "step": 9613 + }, + { + "epoch": 0.8670243946430987, + "grad_norm": 1.4998221291939824, + "learning_rate": 1.825383798001845e-07, + "loss": 0.9103, + "step": 9614 + }, + { + "epoch": 0.867114578166569, + "grad_norm": 1.5936893376791943, + "learning_rate": 1.8229461793461297e-07, + "loss": 0.9212, + "step": 9615 + }, + { + "epoch": 0.8672047616900392, + "grad_norm": 1.5123525468815, + "learning_rate": 1.8205101116568698e-07, + "loss": 0.9908, + "step": 9616 + }, + { + "epoch": 0.8672949452135095, + "grad_norm": 2.618542949348318, + "learning_rate": 1.818075595141928e-07, + "loss": 1.0253, + "step": 9617 + }, + { + "epoch": 0.8673851287369797, + "grad_norm": 1.7534822151058809, + "learning_rate": 1.8156426300090288e-07, + "loss": 0.9734, + "step": 9618 + }, + { + "epoch": 0.86747531226045, + "grad_norm": 1.471902345698942, + "learning_rate": 1.8132112164657686e-07, + "loss": 0.9684, + "step": 9619 + }, + { + "epoch": 0.8675654957839203, + "grad_norm": 1.7671813856806664, + "learning_rate": 1.8107813547196106e-07, + "loss": 0.8741, + "step": 9620 + }, + { + "epoch": 0.8676556793073905, + "grad_norm": 1.547285930882637, + "learning_rate": 1.8083530449778817e-07, + "loss": 1.0152, + "step": 9621 + }, + { + "epoch": 0.8677458628308607, + "grad_norm": 1.5004160621908098, + "learning_rate": 1.8059262874477787e-07, + "loss": 0.8962, + "step": 9622 + }, + { + "epoch": 0.8678360463543311, + "grad_norm": 1.5960919993088019, + "learning_rate": 1.8035010823363627e-07, + "loss": 0.9596, + "step": 9623 + }, + { + "epoch": 0.8679262298778013, + "grad_norm": 2.0863461897738795, + "learning_rate": 1.8010774298505705e-07, + "loss": 1.0126, + "step": 9624 + }, + { + "epoch": 0.8680164134012716, + "grad_norm": 2.1195555109420567, + "learning_rate": 1.7986553301972007e-07, + "loss": 1.07, + "step": 9625 + }, + { + "epoch": 0.8681065969247419, + "grad_norm": 1.8055544005496251, + "learning_rate": 1.7962347835829171e-07, + "loss": 0.97, + "step": 9626 + }, + { + "epoch": 0.8681967804482121, + "grad_norm": 1.4227094612037046, + "learning_rate": 1.793815790214257e-07, + "loss": 0.9837, + "step": 9627 + }, + { + "epoch": 0.8682869639716824, + "grad_norm": 1.7309152961294432, + "learning_rate": 1.791398350297626e-07, + "loss": 0.9356, + "step": 9628 + }, + { + "epoch": 0.8683771474951526, + "grad_norm": 1.4585595182581312, + "learning_rate": 1.7889824640392813e-07, + "loss": 0.9878, + "step": 9629 + }, + { + "epoch": 0.868467331018623, + "grad_norm": 1.997805941645698, + "learning_rate": 1.7865681316453741e-07, + "loss": 0.9597, + "step": 9630 + }, + { + "epoch": 0.8685575145420932, + "grad_norm": 1.5337912822412467, + "learning_rate": 1.7841553533218968e-07, + "loss": 0.9773, + "step": 9631 + }, + { + "epoch": 0.8686476980655634, + "grad_norm": 1.6544298931450032, + "learning_rate": 1.7817441292747292e-07, + "loss": 1.0814, + "step": 9632 + }, + { + "epoch": 0.8687378815890336, + "grad_norm": 1.834747058367577, + "learning_rate": 1.779334459709607e-07, + "loss": 0.9688, + "step": 9633 + }, + { + "epoch": 0.868828065112504, + "grad_norm": 2.0433766995298837, + "learning_rate": 1.7769263448321347e-07, + "loss": 0.9874, + "step": 9634 + }, + { + "epoch": 0.8689182486359742, + "grad_norm": 6.582846064179801, + "learning_rate": 1.7745197848477879e-07, + "loss": 0.9662, + "step": 9635 + }, + { + "epoch": 0.8690084321594445, + "grad_norm": 2.302380354234014, + "learning_rate": 1.7721147799619063e-07, + "loss": 1.0789, + "step": 9636 + }, + { + "epoch": 0.8690986156829147, + "grad_norm": 0.5855507146239918, + "learning_rate": 1.769711330379704e-07, + "loss": 0.785, + "step": 9637 + }, + { + "epoch": 0.869188799206385, + "grad_norm": 5.548276912134576, + "learning_rate": 1.767309436306248e-07, + "loss": 1.0223, + "step": 9638 + }, + { + "epoch": 0.8692789827298553, + "grad_norm": 1.5325611495595477, + "learning_rate": 1.764909097946483e-07, + "loss": 1.0403, + "step": 9639 + }, + { + "epoch": 0.8693691662533255, + "grad_norm": 1.8123153724030654, + "learning_rate": 1.7625103155052236e-07, + "loss": 1.0427, + "step": 9640 + }, + { + "epoch": 0.8694593497767957, + "grad_norm": 1.7991076986318093, + "learning_rate": 1.760113089187143e-07, + "loss": 0.9867, + "step": 9641 + }, + { + "epoch": 0.8695495333002661, + "grad_norm": 1.6660222794644726, + "learning_rate": 1.7577174191967868e-07, + "loss": 1.0015, + "step": 9642 + }, + { + "epoch": 0.8696397168237363, + "grad_norm": 2.1706316565651265, + "learning_rate": 1.755323305738574e-07, + "loss": 0.9868, + "step": 9643 + }, + { + "epoch": 0.8697299003472065, + "grad_norm": 1.4840291473631468, + "learning_rate": 1.7529307490167677e-07, + "loss": 1.008, + "step": 9644 + }, + { + "epoch": 0.8698200838706768, + "grad_norm": 1.9290418896237629, + "learning_rate": 1.7505397492355288e-07, + "loss": 0.9598, + "step": 9645 + }, + { + "epoch": 0.8699102673941471, + "grad_norm": 2.369894764652246, + "learning_rate": 1.7481503065988589e-07, + "loss": 1.0147, + "step": 9646 + }, + { + "epoch": 0.8700004509176174, + "grad_norm": 1.416572492872665, + "learning_rate": 1.7457624213106526e-07, + "loss": 0.9431, + "step": 9647 + }, + { + "epoch": 0.8700906344410876, + "grad_norm": 1.808673370783851, + "learning_rate": 1.7433760935746465e-07, + "loss": 0.9808, + "step": 9648 + }, + { + "epoch": 0.8701808179645579, + "grad_norm": 1.5589464596405738, + "learning_rate": 1.740991323594456e-07, + "loss": 0.9235, + "step": 9649 + }, + { + "epoch": 0.8702710014880282, + "grad_norm": 1.6013052384841795, + "learning_rate": 1.7386081115735651e-07, + "loss": 0.9558, + "step": 9650 + }, + { + "epoch": 0.8703611850114984, + "grad_norm": 1.8993589574670457, + "learning_rate": 1.736226457715324e-07, + "loss": 0.9377, + "step": 9651 + }, + { + "epoch": 0.8704513685349686, + "grad_norm": 3.014173088991472, + "learning_rate": 1.7338463622229505e-07, + "loss": 0.9301, + "step": 9652 + }, + { + "epoch": 0.870541552058439, + "grad_norm": 4.063903800801989, + "learning_rate": 1.7314678252995152e-07, + "loss": 0.9274, + "step": 9653 + }, + { + "epoch": 0.8706317355819092, + "grad_norm": 1.8160289741759041, + "learning_rate": 1.7290908471479847e-07, + "loss": 0.9102, + "step": 9654 + }, + { + "epoch": 0.8707219191053794, + "grad_norm": 1.6442359786094822, + "learning_rate": 1.7267154279711637e-07, + "loss": 0.978, + "step": 9655 + }, + { + "epoch": 0.8708121026288497, + "grad_norm": 1.4156687344496388, + "learning_rate": 1.724341567971741e-07, + "loss": 1.0234, + "step": 9656 + }, + { + "epoch": 0.87090228615232, + "grad_norm": 1.626540774036317, + "learning_rate": 1.7219692673522657e-07, + "loss": 1.0757, + "step": 9657 + }, + { + "epoch": 0.8709924696757902, + "grad_norm": 1.541949667172223, + "learning_rate": 1.7195985263151558e-07, + "loss": 0.9955, + "step": 9658 + }, + { + "epoch": 0.8710826531992605, + "grad_norm": 2.124273737048055, + "learning_rate": 1.7172293450626985e-07, + "loss": 1.0267, + "step": 9659 + }, + { + "epoch": 0.8711728367227307, + "grad_norm": 1.5934759166131667, + "learning_rate": 1.7148617237970475e-07, + "loss": 0.9782, + "step": 9660 + }, + { + "epoch": 0.8712630202462011, + "grad_norm": 1.4474792734615685, + "learning_rate": 1.7124956627202102e-07, + "loss": 0.9424, + "step": 9661 + }, + { + "epoch": 0.8713532037696713, + "grad_norm": 2.112995911124408, + "learning_rate": 1.7101311620340852e-07, + "loss": 0.9942, + "step": 9662 + }, + { + "epoch": 0.8714433872931415, + "grad_norm": 1.767904986935503, + "learning_rate": 1.7077682219404155e-07, + "loss": 0.917, + "step": 9663 + }, + { + "epoch": 0.8715335708166118, + "grad_norm": 2.255634564183634, + "learning_rate": 1.705406842640824e-07, + "loss": 1.1083, + "step": 9664 + }, + { + "epoch": 0.8716237543400821, + "grad_norm": 1.565801717869743, + "learning_rate": 1.7030470243367946e-07, + "loss": 0.9694, + "step": 9665 + }, + { + "epoch": 0.8717139378635523, + "grad_norm": 1.3478286411761442, + "learning_rate": 1.7006887672296834e-07, + "loss": 0.8995, + "step": 9666 + }, + { + "epoch": 0.8718041213870226, + "grad_norm": 2.363043554032984, + "learning_rate": 1.6983320715207094e-07, + "loss": 0.9253, + "step": 9667 + }, + { + "epoch": 0.8718943049104928, + "grad_norm": 1.5517174035507166, + "learning_rate": 1.6959769374109523e-07, + "loss": 1.0003, + "step": 9668 + }, + { + "epoch": 0.8719844884339631, + "grad_norm": 1.6083706434728526, + "learning_rate": 1.6936233651013754e-07, + "loss": 0.9094, + "step": 9669 + }, + { + "epoch": 0.8720746719574334, + "grad_norm": 2.1764316100494994, + "learning_rate": 1.691271354792787e-07, + "loss": 1.0078, + "step": 9670 + }, + { + "epoch": 0.8721648554809036, + "grad_norm": 1.7943602504143954, + "learning_rate": 1.6889209066858866e-07, + "loss": 0.8603, + "step": 9671 + }, + { + "epoch": 0.872255039004374, + "grad_norm": 1.62665927976484, + "learning_rate": 1.6865720209812185e-07, + "loss": 1.0205, + "step": 9672 + }, + { + "epoch": 0.8723452225278442, + "grad_norm": 2.0245244115260723, + "learning_rate": 1.684224697879204e-07, + "loss": 0.9421, + "step": 9673 + }, + { + "epoch": 0.8724354060513144, + "grad_norm": 1.812850129082693, + "learning_rate": 1.6818789375801302e-07, + "loss": 1.0509, + "step": 9674 + }, + { + "epoch": 0.8725255895747847, + "grad_norm": 1.3385965754110798, + "learning_rate": 1.679534740284152e-07, + "loss": 0.9325, + "step": 9675 + }, + { + "epoch": 0.872615773098255, + "grad_norm": 2.0666479535434905, + "learning_rate": 1.6771921061912853e-07, + "loss": 0.9211, + "step": 9676 + }, + { + "epoch": 0.8727059566217252, + "grad_norm": 1.5857076995177695, + "learning_rate": 1.6748510355014234e-07, + "loss": 0.9133, + "step": 9677 + }, + { + "epoch": 0.8727961401451955, + "grad_norm": 1.4419614141436485, + "learning_rate": 1.6725115284143132e-07, + "loss": 0.9916, + "step": 9678 + }, + { + "epoch": 0.8728863236686657, + "grad_norm": 2.0661213611730216, + "learning_rate": 1.670173585129575e-07, + "loss": 0.9501, + "step": 9679 + }, + { + "epoch": 0.872976507192136, + "grad_norm": 1.4439368222065838, + "learning_rate": 1.667837205846696e-07, + "loss": 1.0386, + "step": 9680 + }, + { + "epoch": 0.8730666907156063, + "grad_norm": 1.64103402988035, + "learning_rate": 1.6655023907650278e-07, + "loss": 0.9835, + "step": 9681 + }, + { + "epoch": 0.8731568742390765, + "grad_norm": 2.115464188339575, + "learning_rate": 1.6631691400837954e-07, + "loss": 1.0226, + "step": 9682 + }, + { + "epoch": 0.8732470577625467, + "grad_norm": 1.6040065074417644, + "learning_rate": 1.6608374540020752e-07, + "loss": 1.0268, + "step": 9683 + }, + { + "epoch": 0.8733372412860171, + "grad_norm": 0.6604887221064417, + "learning_rate": 1.658507332718828e-07, + "loss": 0.764, + "step": 9684 + }, + { + "epoch": 0.8734274248094873, + "grad_norm": 1.4229882043120408, + "learning_rate": 1.656178776432864e-07, + "loss": 0.9107, + "step": 9685 + }, + { + "epoch": 0.8735176083329576, + "grad_norm": 1.5826780904203888, + "learning_rate": 1.6538517853428814e-07, + "loss": 0.9552, + "step": 9686 + }, + { + "epoch": 0.8736077918564278, + "grad_norm": 1.3562897496177626, + "learning_rate": 1.6515263596474194e-07, + "loss": 0.9594, + "step": 9687 + }, + { + "epoch": 0.8736979753798981, + "grad_norm": 1.6868489740377348, + "learning_rate": 1.6492024995449017e-07, + "loss": 0.9588, + "step": 9688 + }, + { + "epoch": 0.8737881589033684, + "grad_norm": 1.8059549216033952, + "learning_rate": 1.6468802052336116e-07, + "loss": 0.9411, + "step": 9689 + }, + { + "epoch": 0.8738783424268386, + "grad_norm": 1.5031205143534454, + "learning_rate": 1.6445594769116998e-07, + "loss": 0.9975, + "step": 9690 + }, + { + "epoch": 0.8739685259503088, + "grad_norm": 1.9006871210726146, + "learning_rate": 1.6422403147771836e-07, + "loss": 1.046, + "step": 9691 + }, + { + "epoch": 0.8740587094737792, + "grad_norm": 1.472825989186923, + "learning_rate": 1.6399227190279485e-07, + "loss": 0.9403, + "step": 9692 + }, + { + "epoch": 0.8741488929972494, + "grad_norm": 1.524530185184183, + "learning_rate": 1.637606689861748e-07, + "loss": 0.9078, + "step": 9693 + }, + { + "epoch": 0.8742390765207196, + "grad_norm": 2.0400716688619096, + "learning_rate": 1.6352922274761883e-07, + "loss": 0.935, + "step": 9694 + }, + { + "epoch": 0.87432926004419, + "grad_norm": 1.5004654797883146, + "learning_rate": 1.6329793320687602e-07, + "loss": 0.8404, + "step": 9695 + }, + { + "epoch": 0.8744194435676602, + "grad_norm": 1.533435740713786, + "learning_rate": 1.630668003836808e-07, + "loss": 1.0128, + "step": 9696 + }, + { + "epoch": 0.8745096270911304, + "grad_norm": 1.6493420763043574, + "learning_rate": 1.62835824297755e-07, + "loss": 0.9804, + "step": 9697 + }, + { + "epoch": 0.8745998106146007, + "grad_norm": 1.5467980105574475, + "learning_rate": 1.626050049688066e-07, + "loss": 0.9883, + "step": 9698 + }, + { + "epoch": 0.874689994138071, + "grad_norm": 1.575302394720265, + "learning_rate": 1.623743424165309e-07, + "loss": 0.9507, + "step": 9699 + }, + { + "epoch": 0.8747801776615413, + "grad_norm": 1.7721544237166278, + "learning_rate": 1.6214383666060826e-07, + "loss": 0.926, + "step": 9700 + }, + { + "epoch": 0.8748703611850115, + "grad_norm": 1.557829475884263, + "learning_rate": 1.619134877207078e-07, + "loss": 0.9987, + "step": 9701 + }, + { + "epoch": 0.8749605447084817, + "grad_norm": 1.6800451563261394, + "learning_rate": 1.616832956164831e-07, + "loss": 0.9781, + "step": 9702 + }, + { + "epoch": 0.8750507282319521, + "grad_norm": 1.4970896620722385, + "learning_rate": 1.6145326036757667e-07, + "loss": 0.9464, + "step": 9703 + }, + { + "epoch": 0.8751409117554223, + "grad_norm": 1.750458861546711, + "learning_rate": 1.612233819936155e-07, + "loss": 0.9425, + "step": 9704 + }, + { + "epoch": 0.8752310952788925, + "grad_norm": 1.6655148181651185, + "learning_rate": 1.6099366051421414e-07, + "loss": 0.9736, + "step": 9705 + }, + { + "epoch": 0.8753212788023628, + "grad_norm": 2.118872390857985, + "learning_rate": 1.6076409594897378e-07, + "loss": 1.0385, + "step": 9706 + }, + { + "epoch": 0.8754114623258331, + "grad_norm": 0.7152582853144719, + "learning_rate": 1.605346883174823e-07, + "loss": 0.7646, + "step": 9707 + }, + { + "epoch": 0.8755016458493033, + "grad_norm": 1.8300448387155968, + "learning_rate": 1.6030543763931427e-07, + "loss": 0.9543, + "step": 9708 + }, + { + "epoch": 0.8755918293727736, + "grad_norm": 1.7738871403503773, + "learning_rate": 1.600763439340298e-07, + "loss": 0.9963, + "step": 9709 + }, + { + "epoch": 0.8756820128962438, + "grad_norm": 1.5037704604636513, + "learning_rate": 1.5984740722117707e-07, + "loss": 1.031, + "step": 9710 + }, + { + "epoch": 0.8757721964197142, + "grad_norm": 1.3628916798634272, + "learning_rate": 1.5961862752028998e-07, + "loss": 0.9908, + "step": 9711 + }, + { + "epoch": 0.8758623799431844, + "grad_norm": 1.2851938166823704, + "learning_rate": 1.5939000485088937e-07, + "loss": 0.8828, + "step": 9712 + }, + { + "epoch": 0.8759525634666546, + "grad_norm": 1.4051562400231647, + "learning_rate": 1.5916153923248254e-07, + "loss": 1.0695, + "step": 9713 + }, + { + "epoch": 0.8760427469901249, + "grad_norm": 1.8301681447560587, + "learning_rate": 1.5893323068456342e-07, + "loss": 1.0156, + "step": 9714 + }, + { + "epoch": 0.8761329305135952, + "grad_norm": 1.5179483754925265, + "learning_rate": 1.5870507922661248e-07, + "loss": 1.0218, + "step": 9715 + }, + { + "epoch": 0.8762231140370654, + "grad_norm": 0.7238792174948775, + "learning_rate": 1.5847708487809763e-07, + "loss": 0.8042, + "step": 9716 + }, + { + "epoch": 0.8763132975605357, + "grad_norm": 2.1882042977184435, + "learning_rate": 1.5824924765847113e-07, + "loss": 1.0545, + "step": 9717 + }, + { + "epoch": 0.8764034810840059, + "grad_norm": 1.3656306094266095, + "learning_rate": 1.5802156758717478e-07, + "loss": 0.9559, + "step": 9718 + }, + { + "epoch": 0.8764936646074762, + "grad_norm": 1.7785212713797505, + "learning_rate": 1.5779404468363433e-07, + "loss": 0.9777, + "step": 9719 + }, + { + "epoch": 0.8765838481309465, + "grad_norm": 1.7017034495030734, + "learning_rate": 1.5756667896726405e-07, + "loss": 1.0332, + "step": 9720 + }, + { + "epoch": 0.8766740316544167, + "grad_norm": 2.0107700518270266, + "learning_rate": 1.5733947045746377e-07, + "loss": 0.9917, + "step": 9721 + }, + { + "epoch": 0.876764215177887, + "grad_norm": 1.8138888069738293, + "learning_rate": 1.5711241917362018e-07, + "loss": 1.0467, + "step": 9722 + }, + { + "epoch": 0.8768543987013573, + "grad_norm": 1.778023725954227, + "learning_rate": 1.5688552513510688e-07, + "loss": 0.9145, + "step": 9723 + }, + { + "epoch": 0.8769445822248275, + "grad_norm": 1.4002865549147214, + "learning_rate": 1.5665878836128266e-07, + "loss": 0.9988, + "step": 9724 + }, + { + "epoch": 0.8770347657482978, + "grad_norm": 2.483770812839645, + "learning_rate": 1.5643220887149554e-07, + "loss": 1.0253, + "step": 9725 + }, + { + "epoch": 0.8771249492717681, + "grad_norm": 0.6639412713190164, + "learning_rate": 1.562057866850772e-07, + "loss": 0.8255, + "step": 9726 + }, + { + "epoch": 0.8772151327952383, + "grad_norm": 1.6688002360922554, + "learning_rate": 1.5597952182134777e-07, + "loss": 0.8537, + "step": 9727 + }, + { + "epoch": 0.8773053163187086, + "grad_norm": 1.4594482338151906, + "learning_rate": 1.557534142996133e-07, + "loss": 1.032, + "step": 9728 + }, + { + "epoch": 0.8773954998421788, + "grad_norm": 1.6122808795467376, + "learning_rate": 1.5552746413916662e-07, + "loss": 1.008, + "step": 9729 + }, + { + "epoch": 0.8774856833656491, + "grad_norm": 2.04229947099132, + "learning_rate": 1.5530167135928697e-07, + "loss": 0.9715, + "step": 9730 + }, + { + "epoch": 0.8775758668891194, + "grad_norm": 1.9376456913469275, + "learning_rate": 1.5507603597924068e-07, + "loss": 0.921, + "step": 9731 + }, + { + "epoch": 0.8776660504125896, + "grad_norm": 1.7893162843036017, + "learning_rate": 1.548505580182793e-07, + "loss": 0.9365, + "step": 9732 + }, + { + "epoch": 0.8777562339360598, + "grad_norm": 1.5356849749990211, + "learning_rate": 1.5462523749564271e-07, + "loss": 0.9765, + "step": 9733 + }, + { + "epoch": 0.8778464174595302, + "grad_norm": 1.5259086715846837, + "learning_rate": 1.5440007443055602e-07, + "loss": 0.9783, + "step": 9734 + }, + { + "epoch": 0.8779366009830004, + "grad_norm": 1.6216534119574328, + "learning_rate": 1.541750688422314e-07, + "loss": 0.9083, + "step": 9735 + }, + { + "epoch": 0.8780267845064706, + "grad_norm": 1.778913404260187, + "learning_rate": 1.5395022074986797e-07, + "loss": 0.9777, + "step": 9736 + }, + { + "epoch": 0.8781169680299409, + "grad_norm": 1.5365698132778283, + "learning_rate": 1.5372553017265033e-07, + "loss": 0.9192, + "step": 9737 + }, + { + "epoch": 0.8782071515534112, + "grad_norm": 2.169868475448369, + "learning_rate": 1.5350099712975116e-07, + "loss": 1.0395, + "step": 9738 + }, + { + "epoch": 0.8782973350768815, + "grad_norm": 2.368925558141984, + "learning_rate": 1.5327662164032785e-07, + "loss": 0.9671, + "step": 9739 + }, + { + "epoch": 0.8783875186003517, + "grad_norm": 1.7107561964708553, + "learning_rate": 1.5305240372352656e-07, + "loss": 0.8278, + "step": 9740 + }, + { + "epoch": 0.8784777021238219, + "grad_norm": 1.5765803113976602, + "learning_rate": 1.5282834339847738e-07, + "loss": 0.986, + "step": 9741 + }, + { + "epoch": 0.8785678856472923, + "grad_norm": 1.5506200442375258, + "learning_rate": 1.526044406842999e-07, + "loss": 0.958, + "step": 9742 + }, + { + "epoch": 0.8786580691707625, + "grad_norm": 1.6211678729288947, + "learning_rate": 1.523806956000977e-07, + "loss": 0.9063, + "step": 9743 + }, + { + "epoch": 0.8787482526942327, + "grad_norm": 1.3847591685156693, + "learning_rate": 1.5215710816496197e-07, + "loss": 1.029, + "step": 9744 + }, + { + "epoch": 0.8788384362177031, + "grad_norm": 1.708320392776727, + "learning_rate": 1.5193367839797077e-07, + "loss": 0.9225, + "step": 9745 + }, + { + "epoch": 0.8789286197411733, + "grad_norm": 1.5356744178083408, + "learning_rate": 1.5171040631818842e-07, + "loss": 0.9759, + "step": 9746 + }, + { + "epoch": 0.8790188032646435, + "grad_norm": 1.6206903530232697, + "learning_rate": 1.5148729194466547e-07, + "loss": 0.9057, + "step": 9747 + }, + { + "epoch": 0.8791089867881138, + "grad_norm": 1.5728226522007747, + "learning_rate": 1.5126433529643956e-07, + "loss": 1.0473, + "step": 9748 + }, + { + "epoch": 0.8791991703115841, + "grad_norm": 1.4558058589416492, + "learning_rate": 1.5104153639253436e-07, + "loss": 0.9749, + "step": 9749 + }, + { + "epoch": 0.8792893538350544, + "grad_norm": 0.6742994788363447, + "learning_rate": 1.5081889525196002e-07, + "loss": 0.8667, + "step": 9750 + }, + { + "epoch": 0.8793795373585246, + "grad_norm": 2.5542382828181514, + "learning_rate": 1.5059641189371398e-07, + "loss": 1.0361, + "step": 9751 + }, + { + "epoch": 0.8794697208819948, + "grad_norm": 1.576335458881647, + "learning_rate": 1.503740863367795e-07, + "loss": 0.9922, + "step": 9752 + }, + { + "epoch": 0.8795599044054652, + "grad_norm": 2.372816788935669, + "learning_rate": 1.50151918600127e-07, + "loss": 1.0094, + "step": 9753 + }, + { + "epoch": 0.8796500879289354, + "grad_norm": 3.4023332387628247, + "learning_rate": 1.4992990870271217e-07, + "loss": 1.0283, + "step": 9754 + }, + { + "epoch": 0.8797402714524056, + "grad_norm": 1.9925075498315845, + "learning_rate": 1.497080566634794e-07, + "loss": 0.9846, + "step": 9755 + }, + { + "epoch": 0.8798304549758759, + "grad_norm": 1.5762566563755371, + "learning_rate": 1.4948636250135693e-07, + "loss": 0.9883, + "step": 9756 + }, + { + "epoch": 0.8799206384993462, + "grad_norm": 1.798508274995773, + "learning_rate": 1.4926482623526249e-07, + "loss": 1.0397, + "step": 9757 + }, + { + "epoch": 0.8800108220228164, + "grad_norm": 1.8400221817089348, + "learning_rate": 1.4904344788409694e-07, + "loss": 0.9817, + "step": 9758 + }, + { + "epoch": 0.8801010055462867, + "grad_norm": 1.546573725174229, + "learning_rate": 1.4882222746675143e-07, + "loss": 0.9374, + "step": 9759 + }, + { + "epoch": 0.8801911890697569, + "grad_norm": 1.540407164612034, + "learning_rate": 1.4860116500210018e-07, + "loss": 1.0195, + "step": 9760 + }, + { + "epoch": 0.8802813725932273, + "grad_norm": 1.9556586959340068, + "learning_rate": 1.4838026050900632e-07, + "loss": 1.052, + "step": 9761 + }, + { + "epoch": 0.8803715561166975, + "grad_norm": 1.5376857273225086, + "learning_rate": 1.481595140063181e-07, + "loss": 1.0266, + "step": 9762 + }, + { + "epoch": 0.8804617396401677, + "grad_norm": 1.768111714285878, + "learning_rate": 1.4793892551287136e-07, + "loss": 1.0683, + "step": 9763 + }, + { + "epoch": 0.880551923163638, + "grad_norm": 1.7186335437508744, + "learning_rate": 1.4771849504748768e-07, + "loss": 0.9072, + "step": 9764 + }, + { + "epoch": 0.8806421066871083, + "grad_norm": 1.7255069043163118, + "learning_rate": 1.4749822262897517e-07, + "loss": 1.0557, + "step": 9765 + }, + { + "epoch": 0.8807322902105785, + "grad_norm": 1.5802162590993205, + "learning_rate": 1.4727810827612895e-07, + "loss": 0.9195, + "step": 9766 + }, + { + "epoch": 0.8808224737340488, + "grad_norm": 2.552093308617436, + "learning_rate": 1.470581520077303e-07, + "loss": 0.9885, + "step": 9767 + }, + { + "epoch": 0.8809126572575191, + "grad_norm": 2.6448490284341806, + "learning_rate": 1.4683835384254705e-07, + "loss": 0.9802, + "step": 9768 + }, + { + "epoch": 0.8810028407809893, + "grad_norm": 1.6868869939648343, + "learning_rate": 1.4661871379933376e-07, + "loss": 1.0589, + "step": 9769 + }, + { + "epoch": 0.8810930243044596, + "grad_norm": 1.517853033236542, + "learning_rate": 1.4639923189683169e-07, + "loss": 1.0279, + "step": 9770 + }, + { + "epoch": 0.8811832078279298, + "grad_norm": 1.5715422867836455, + "learning_rate": 1.461799081537669e-07, + "loss": 1.0318, + "step": 9771 + }, + { + "epoch": 0.8812733913514001, + "grad_norm": 1.6282944662452585, + "learning_rate": 1.4596074258885514e-07, + "loss": 1.0447, + "step": 9772 + }, + { + "epoch": 0.8813635748748704, + "grad_norm": 1.4426179253887546, + "learning_rate": 1.4574173522079502e-07, + "loss": 0.9508, + "step": 9773 + }, + { + "epoch": 0.8814537583983406, + "grad_norm": 1.9807630692686011, + "learning_rate": 1.4552288606827513e-07, + "loss": 0.9528, + "step": 9774 + }, + { + "epoch": 0.8815439419218108, + "grad_norm": 1.8282425671455962, + "learning_rate": 1.4530419514996761e-07, + "loss": 1.0549, + "step": 9775 + }, + { + "epoch": 0.8816341254452812, + "grad_norm": 0.7045532978437923, + "learning_rate": 1.4508566248453291e-07, + "loss": 0.8262, + "step": 9776 + }, + { + "epoch": 0.8817243089687514, + "grad_norm": 1.8925243326736814, + "learning_rate": 1.448672880906172e-07, + "loss": 0.9716, + "step": 9777 + }, + { + "epoch": 0.8818144924922217, + "grad_norm": 1.7447486150943514, + "learning_rate": 1.4464907198685382e-07, + "loss": 0.9372, + "step": 9778 + }, + { + "epoch": 0.8819046760156919, + "grad_norm": 1.389367681733702, + "learning_rate": 1.444310141918621e-07, + "loss": 0.9822, + "step": 9779 + }, + { + "epoch": 0.8819948595391622, + "grad_norm": 2.231452898475891, + "learning_rate": 1.4421311472424735e-07, + "loss": 1.0084, + "step": 9780 + }, + { + "epoch": 0.8820850430626325, + "grad_norm": 1.867088347160601, + "learning_rate": 1.4399537360260273e-07, + "loss": 1.0177, + "step": 9781 + }, + { + "epoch": 0.8821752265861027, + "grad_norm": 1.7061819936313494, + "learning_rate": 1.4377779084550645e-07, + "loss": 1.0032, + "step": 9782 + }, + { + "epoch": 0.8822654101095729, + "grad_norm": 1.7655000895648978, + "learning_rate": 1.4356036647152413e-07, + "loss": 0.9725, + "step": 9783 + }, + { + "epoch": 0.8823555936330433, + "grad_norm": 1.4532308232321667, + "learning_rate": 1.4334310049920785e-07, + "loss": 0.9587, + "step": 9784 + }, + { + "epoch": 0.8824457771565135, + "grad_norm": 0.5983809272579331, + "learning_rate": 1.431259929470956e-07, + "loss": 0.7664, + "step": 9785 + }, + { + "epoch": 0.8825359606799837, + "grad_norm": 1.5879526349073603, + "learning_rate": 1.4290904383371237e-07, + "loss": 1.0562, + "step": 9786 + }, + { + "epoch": 0.882626144203454, + "grad_norm": 1.8863416970263005, + "learning_rate": 1.4269225317756961e-07, + "loss": 0.9681, + "step": 9787 + }, + { + "epoch": 0.8827163277269243, + "grad_norm": 2.042704869691622, + "learning_rate": 1.424756209971645e-07, + "loss": 0.9426, + "step": 9788 + }, + { + "epoch": 0.8828065112503946, + "grad_norm": 1.91534163104494, + "learning_rate": 1.4225914731098199e-07, + "loss": 0.9787, + "step": 9789 + }, + { + "epoch": 0.8828966947738648, + "grad_norm": 1.9142947309094593, + "learning_rate": 1.4204283213749248e-07, + "loss": 0.9112, + "step": 9790 + }, + { + "epoch": 0.8829868782973351, + "grad_norm": 1.5263542228436129, + "learning_rate": 1.4182667549515315e-07, + "loss": 1.0389, + "step": 9791 + }, + { + "epoch": 0.8830770618208054, + "grad_norm": 1.7219427798131632, + "learning_rate": 1.4161067740240752e-07, + "loss": 0.9621, + "step": 9792 + }, + { + "epoch": 0.8831672453442756, + "grad_norm": 1.416670191517353, + "learning_rate": 1.4139483787768614e-07, + "loss": 0.9725, + "step": 9793 + }, + { + "epoch": 0.8832574288677458, + "grad_norm": 2.306685173515358, + "learning_rate": 1.4117915693940584e-07, + "loss": 0.889, + "step": 9794 + }, + { + "epoch": 0.8833476123912162, + "grad_norm": 1.6228924068389177, + "learning_rate": 1.409636346059684e-07, + "loss": 0.9723, + "step": 9795 + }, + { + "epoch": 0.8834377959146864, + "grad_norm": 2.3649146118233997, + "learning_rate": 1.4074827089576501e-07, + "loss": 1.0158, + "step": 9796 + }, + { + "epoch": 0.8835279794381566, + "grad_norm": 2.1128072436216114, + "learning_rate": 1.4053306582717085e-07, + "loss": 0.9933, + "step": 9797 + }, + { + "epoch": 0.8836181629616269, + "grad_norm": 1.7617597554038116, + "learning_rate": 1.4031801941854827e-07, + "loss": 1.0064, + "step": 9798 + }, + { + "epoch": 0.8837083464850972, + "grad_norm": 1.735384192130453, + "learning_rate": 1.401031316882466e-07, + "loss": 0.9472, + "step": 9799 + }, + { + "epoch": 0.8837985300085675, + "grad_norm": 1.7460709470074927, + "learning_rate": 1.39888402654601e-07, + "loss": 0.998, + "step": 9800 + }, + { + "epoch": 0.8838887135320377, + "grad_norm": 0.602470159068652, + "learning_rate": 1.3967383233593344e-07, + "loss": 0.8244, + "step": 9801 + }, + { + "epoch": 0.8839788970555079, + "grad_norm": 9.00785823654911, + "learning_rate": 1.3945942075055218e-07, + "loss": 0.9402, + "step": 9802 + }, + { + "epoch": 0.8840690805789783, + "grad_norm": 1.6116154455009564, + "learning_rate": 1.3924516791675212e-07, + "loss": 1.0055, + "step": 9803 + }, + { + "epoch": 0.8841592641024485, + "grad_norm": 1.6571162855315265, + "learning_rate": 1.3903107385281487e-07, + "loss": 0.8909, + "step": 9804 + }, + { + "epoch": 0.8842494476259187, + "grad_norm": 1.7906108968821794, + "learning_rate": 1.3881713857700717e-07, + "loss": 0.9914, + "step": 9805 + }, + { + "epoch": 0.884339631149389, + "grad_norm": 1.7345423574721697, + "learning_rate": 1.3860336210758372e-07, + "loss": 0.958, + "step": 9806 + }, + { + "epoch": 0.8844298146728593, + "grad_norm": 2.038058329652495, + "learning_rate": 1.3838974446278506e-07, + "loss": 0.938, + "step": 9807 + }, + { + "epoch": 0.8845199981963295, + "grad_norm": 1.6240813519673907, + "learning_rate": 1.3817628566083817e-07, + "loss": 0.9555, + "step": 9808 + }, + { + "epoch": 0.8846101817197998, + "grad_norm": 1.3509412010013342, + "learning_rate": 1.3796298571995712e-07, + "loss": 0.9934, + "step": 9809 + }, + { + "epoch": 0.88470036524327, + "grad_norm": 1.6261743556967476, + "learning_rate": 1.377498446583405e-07, + "loss": 0.8759, + "step": 9810 + }, + { + "epoch": 0.8847905487667403, + "grad_norm": 1.4280305025652338, + "learning_rate": 1.3753686249417596e-07, + "loss": 0.9518, + "step": 9811 + }, + { + "epoch": 0.8848807322902106, + "grad_norm": 1.9916774321176025, + "learning_rate": 1.373240392456354e-07, + "loss": 0.9328, + "step": 9812 + }, + { + "epoch": 0.8849709158136808, + "grad_norm": 1.7349899851265367, + "learning_rate": 1.37111374930879e-07, + "loss": 0.9177, + "step": 9813 + }, + { + "epoch": 0.885061099337151, + "grad_norm": 1.9360328780133191, + "learning_rate": 1.3689886956805176e-07, + "loss": 0.9083, + "step": 9814 + }, + { + "epoch": 0.8851512828606214, + "grad_norm": 1.471113292140887, + "learning_rate": 1.3668652317528585e-07, + "loss": 0.9668, + "step": 9815 + }, + { + "epoch": 0.8852414663840916, + "grad_norm": 1.6781769813712866, + "learning_rate": 1.3647433577070012e-07, + "loss": 0.9985, + "step": 9816 + }, + { + "epoch": 0.8853316499075619, + "grad_norm": 0.6426990012712971, + "learning_rate": 1.3626230737239942e-07, + "loss": 0.7866, + "step": 9817 + }, + { + "epoch": 0.8854218334310322, + "grad_norm": 0.7030956686047891, + "learning_rate": 1.3605043799847527e-07, + "loss": 0.794, + "step": 9818 + }, + { + "epoch": 0.8855120169545024, + "grad_norm": 1.3782443870680625, + "learning_rate": 1.3583872766700567e-07, + "loss": 0.916, + "step": 9819 + }, + { + "epoch": 0.8856022004779727, + "grad_norm": 1.4734916442437425, + "learning_rate": 1.3562717639605437e-07, + "loss": 0.9597, + "step": 9820 + }, + { + "epoch": 0.8856923840014429, + "grad_norm": 1.6877136095271623, + "learning_rate": 1.3541578420367229e-07, + "loss": 0.9493, + "step": 9821 + }, + { + "epoch": 0.8857825675249132, + "grad_norm": 5.078104905308799, + "learning_rate": 1.3520455110789697e-07, + "loss": 0.9937, + "step": 9822 + }, + { + "epoch": 0.8858727510483835, + "grad_norm": 1.3505745071512198, + "learning_rate": 1.3499347712675158e-07, + "loss": 0.9553, + "step": 9823 + }, + { + "epoch": 0.8859629345718537, + "grad_norm": 1.5829117196890992, + "learning_rate": 1.3478256227824635e-07, + "loss": 0.9439, + "step": 9824 + }, + { + "epoch": 0.8860531180953239, + "grad_norm": 1.8493441398600985, + "learning_rate": 1.3457180658037759e-07, + "loss": 0.8875, + "step": 9825 + }, + { + "epoch": 0.8861433016187943, + "grad_norm": 1.3598871252958946, + "learning_rate": 1.3436121005112843e-07, + "loss": 1.0108, + "step": 9826 + }, + { + "epoch": 0.8862334851422645, + "grad_norm": 2.2652135014908477, + "learning_rate": 1.3415077270846719e-07, + "loss": 0.9461, + "step": 9827 + }, + { + "epoch": 0.8863236686657348, + "grad_norm": 1.9373955852306635, + "learning_rate": 1.3394049457035105e-07, + "loss": 1.0893, + "step": 9828 + }, + { + "epoch": 0.886413852189205, + "grad_norm": 1.617624269926736, + "learning_rate": 1.3373037565472034e-07, + "loss": 0.9869, + "step": 9829 + }, + { + "epoch": 0.8865040357126753, + "grad_norm": 1.9754764754725032, + "learning_rate": 1.3352041597950537e-07, + "loss": 0.9576, + "step": 9830 + }, + { + "epoch": 0.8865942192361456, + "grad_norm": 1.7497835025384183, + "learning_rate": 1.333106155626196e-07, + "loss": 0.9654, + "step": 9831 + }, + { + "epoch": 0.8866844027596158, + "grad_norm": 1.601651519534401, + "learning_rate": 1.331009744219651e-07, + "loss": 0.9183, + "step": 9832 + }, + { + "epoch": 0.886774586283086, + "grad_norm": 0.6691672743497238, + "learning_rate": 1.3289149257542943e-07, + "loss": 0.849, + "step": 9833 + }, + { + "epoch": 0.8868647698065564, + "grad_norm": 1.6045393655333808, + "learning_rate": 1.3268217004088666e-07, + "loss": 0.8786, + "step": 9834 + }, + { + "epoch": 0.8869549533300266, + "grad_norm": 1.6994620313451803, + "learning_rate": 1.3247300683619788e-07, + "loss": 0.8642, + "step": 9835 + }, + { + "epoch": 0.8870451368534968, + "grad_norm": 2.3440079610643823, + "learning_rate": 1.3226400297920903e-07, + "loss": 0.9544, + "step": 9836 + }, + { + "epoch": 0.8871353203769671, + "grad_norm": 1.601765615097483, + "learning_rate": 1.3205515848775428e-07, + "loss": 0.9603, + "step": 9837 + }, + { + "epoch": 0.8872255039004374, + "grad_norm": 1.673757758603336, + "learning_rate": 1.3184647337965316e-07, + "loss": 0.9159, + "step": 9838 + }, + { + "epoch": 0.8873156874239077, + "grad_norm": 1.3277835631455102, + "learning_rate": 1.3163794767271163e-07, + "loss": 0.9259, + "step": 9839 + }, + { + "epoch": 0.8874058709473779, + "grad_norm": 1.5320153172825348, + "learning_rate": 1.314295813847226e-07, + "loss": 0.9475, + "step": 9840 + }, + { + "epoch": 0.8874960544708482, + "grad_norm": 1.8426534980592288, + "learning_rate": 1.3122137453346515e-07, + "loss": 1.0765, + "step": 9841 + }, + { + "epoch": 0.8875862379943185, + "grad_norm": 1.7498699548630736, + "learning_rate": 1.3101332713670376e-07, + "loss": 1.035, + "step": 9842 + }, + { + "epoch": 0.8876764215177887, + "grad_norm": 0.6899751410636452, + "learning_rate": 1.3080543921219133e-07, + "loss": 0.8436, + "step": 9843 + }, + { + "epoch": 0.8877666050412589, + "grad_norm": 1.8209651710158363, + "learning_rate": 1.3059771077766478e-07, + "loss": 0.9702, + "step": 9844 + }, + { + "epoch": 0.8878567885647293, + "grad_norm": 1.966572783810156, + "learning_rate": 1.3039014185085018e-07, + "loss": 1.0176, + "step": 9845 + }, + { + "epoch": 0.8879469720881995, + "grad_norm": 6.162018372941353, + "learning_rate": 1.301827324494571e-07, + "loss": 0.9852, + "step": 9846 + }, + { + "epoch": 0.8880371556116697, + "grad_norm": 2.165966752294665, + "learning_rate": 1.2997548259118342e-07, + "loss": 1.0156, + "step": 9847 + }, + { + "epoch": 0.88812733913514, + "grad_norm": 1.680794710249205, + "learning_rate": 1.2976839229371272e-07, + "loss": 1.0056, + "step": 9848 + }, + { + "epoch": 0.8882175226586103, + "grad_norm": 0.6309252961382787, + "learning_rate": 1.2956146157471515e-07, + "loss": 0.8326, + "step": 9849 + }, + { + "epoch": 0.8883077061820805, + "grad_norm": 2.1175372408459854, + "learning_rate": 1.2935469045184745e-07, + "loss": 0.9992, + "step": 9850 + }, + { + "epoch": 0.8883978897055508, + "grad_norm": 1.7542543469381002, + "learning_rate": 1.291480789427517e-07, + "loss": 0.9539, + "step": 9851 + }, + { + "epoch": 0.888488073229021, + "grad_norm": 2.1299224388077875, + "learning_rate": 1.2894162706505807e-07, + "loss": 1.0188, + "step": 9852 + }, + { + "epoch": 0.8885782567524914, + "grad_norm": 1.7414383581939972, + "learning_rate": 1.2873533483638155e-07, + "loss": 0.9737, + "step": 9853 + }, + { + "epoch": 0.8886684402759616, + "grad_norm": 1.4432836365605184, + "learning_rate": 1.285292022743243e-07, + "loss": 0.9258, + "step": 9854 + }, + { + "epoch": 0.8887586237994318, + "grad_norm": 1.7732739478161, + "learning_rate": 1.2832322939647467e-07, + "loss": 0.9238, + "step": 9855 + }, + { + "epoch": 0.8888488073229021, + "grad_norm": 1.697710020755958, + "learning_rate": 1.281174162204075e-07, + "loss": 0.9652, + "step": 9856 + }, + { + "epoch": 0.8889389908463724, + "grad_norm": 1.4207297942768524, + "learning_rate": 1.2791176276368366e-07, + "loss": 0.8585, + "step": 9857 + }, + { + "epoch": 0.8890291743698426, + "grad_norm": 1.5362371080694128, + "learning_rate": 1.2770626904385128e-07, + "loss": 0.9885, + "step": 9858 + }, + { + "epoch": 0.8891193578933129, + "grad_norm": 1.9584631640096566, + "learning_rate": 1.2750093507844306e-07, + "loss": 0.841, + "step": 9859 + }, + { + "epoch": 0.8892095414167831, + "grad_norm": 1.8018249003008018, + "learning_rate": 1.272957608849805e-07, + "loss": 1.0178, + "step": 9860 + }, + { + "epoch": 0.8892997249402534, + "grad_norm": 1.3648797312700887, + "learning_rate": 1.270907464809694e-07, + "loss": 0.9547, + "step": 9861 + }, + { + "epoch": 0.8893899084637237, + "grad_norm": 1.49569585304754, + "learning_rate": 1.2688589188390285e-07, + "loss": 0.9801, + "step": 9862 + }, + { + "epoch": 0.8894800919871939, + "grad_norm": 1.3983950475290798, + "learning_rate": 1.2668119711126023e-07, + "loss": 0.99, + "step": 9863 + }, + { + "epoch": 0.8895702755106643, + "grad_norm": 2.201157144194276, + "learning_rate": 1.2647666218050735e-07, + "loss": 0.9801, + "step": 9864 + }, + { + "epoch": 0.8896604590341345, + "grad_norm": 1.8294543136434476, + "learning_rate": 1.2627228710909643e-07, + "loss": 0.9417, + "step": 9865 + }, + { + "epoch": 0.8897506425576047, + "grad_norm": 1.3945347120738931, + "learning_rate": 1.260680719144649e-07, + "loss": 0.9456, + "step": 9866 + }, + { + "epoch": 0.889840826081075, + "grad_norm": 3.2483166957042258, + "learning_rate": 1.2586401661403877e-07, + "loss": 0.9042, + "step": 9867 + }, + { + "epoch": 0.8899310096045453, + "grad_norm": 1.7289368886598702, + "learning_rate": 1.2566012122522817e-07, + "loss": 0.9496, + "step": 9868 + }, + { + "epoch": 0.8900211931280155, + "grad_norm": 1.8301395499588573, + "learning_rate": 1.254563857654316e-07, + "loss": 0.9476, + "step": 9869 + }, + { + "epoch": 0.8901113766514858, + "grad_norm": 2.495185412166572, + "learning_rate": 1.2525281025203205e-07, + "loss": 0.8642, + "step": 9870 + }, + { + "epoch": 0.890201560174956, + "grad_norm": 1.7953687906363052, + "learning_rate": 1.2504939470240006e-07, + "loss": 0.9595, + "step": 9871 + }, + { + "epoch": 0.8902917436984263, + "grad_norm": 2.3203468191774888, + "learning_rate": 1.2484613913389196e-07, + "loss": 1.0431, + "step": 9872 + }, + { + "epoch": 0.8903819272218966, + "grad_norm": 1.752947300899498, + "learning_rate": 1.2464304356385057e-07, + "loss": 0.8903, + "step": 9873 + }, + { + "epoch": 0.8904721107453668, + "grad_norm": 6.9390980022770306, + "learning_rate": 1.2444010800960558e-07, + "loss": 1.0331, + "step": 9874 + }, + { + "epoch": 0.890562294268837, + "grad_norm": 1.7069004453915597, + "learning_rate": 1.2423733248847267e-07, + "loss": 0.9619, + "step": 9875 + }, + { + "epoch": 0.8906524777923074, + "grad_norm": 1.6311803810200773, + "learning_rate": 1.2403471701775293e-07, + "loss": 0.9169, + "step": 9876 + }, + { + "epoch": 0.8907426613157776, + "grad_norm": 1.6972658357082364, + "learning_rate": 1.2383226161473515e-07, + "loss": 1.0148, + "step": 9877 + }, + { + "epoch": 0.8908328448392479, + "grad_norm": 1.7062736593429633, + "learning_rate": 1.2362996629669376e-07, + "loss": 1.0531, + "step": 9878 + }, + { + "epoch": 0.8909230283627181, + "grad_norm": 1.4553455900624712, + "learning_rate": 1.2342783108089007e-07, + "loss": 0.9882, + "step": 9879 + }, + { + "epoch": 0.8910132118861884, + "grad_norm": 2.5879140255427875, + "learning_rate": 1.2322585598457135e-07, + "loss": 1.0341, + "step": 9880 + }, + { + "epoch": 0.8911033954096587, + "grad_norm": 1.2789992124481997, + "learning_rate": 1.2302404102497034e-07, + "loss": 0.9514, + "step": 9881 + }, + { + "epoch": 0.8911935789331289, + "grad_norm": 1.7535697087381301, + "learning_rate": 1.228223862193083e-07, + "loss": 0.905, + "step": 9882 + }, + { + "epoch": 0.8912837624565991, + "grad_norm": 2.0824296008388736, + "learning_rate": 1.2262089158479038e-07, + "loss": 0.9887, + "step": 9883 + }, + { + "epoch": 0.8913739459800695, + "grad_norm": 1.8771286008106431, + "learning_rate": 1.2241955713861042e-07, + "loss": 1.0195, + "step": 9884 + }, + { + "epoch": 0.8914641295035397, + "grad_norm": 1.40937639900358, + "learning_rate": 1.222183828979464e-07, + "loss": 0.9223, + "step": 9885 + }, + { + "epoch": 0.8915543130270099, + "grad_norm": 1.3403710309245298, + "learning_rate": 1.2201736887996372e-07, + "loss": 0.9767, + "step": 9886 + }, + { + "epoch": 0.8916444965504803, + "grad_norm": 1.6590361283261894, + "learning_rate": 1.2181651510181444e-07, + "loss": 1.0543, + "step": 9887 + }, + { + "epoch": 0.8917346800739505, + "grad_norm": 1.87482763133896, + "learning_rate": 1.2161582158063622e-07, + "loss": 1.0058, + "step": 9888 + }, + { + "epoch": 0.8918248635974207, + "grad_norm": 2.9289193328866414, + "learning_rate": 1.214152883335533e-07, + "loss": 0.8838, + "step": 9889 + }, + { + "epoch": 0.891915047120891, + "grad_norm": 1.7976655630759888, + "learning_rate": 1.2121491537767648e-07, + "loss": 0.9294, + "step": 9890 + }, + { + "epoch": 0.8920052306443613, + "grad_norm": 1.5064039222929584, + "learning_rate": 1.2101470273010294e-07, + "loss": 1.0115, + "step": 9891 + }, + { + "epoch": 0.8920954141678316, + "grad_norm": 1.8068169255195032, + "learning_rate": 1.2081465040791528e-07, + "loss": 0.949, + "step": 9892 + }, + { + "epoch": 0.8921855976913018, + "grad_norm": 1.7215800480996435, + "learning_rate": 1.2061475842818335e-07, + "loss": 0.9443, + "step": 9893 + }, + { + "epoch": 0.892275781214772, + "grad_norm": 1.6860285100687136, + "learning_rate": 1.2041502680796313e-07, + "loss": 0.9538, + "step": 9894 + }, + { + "epoch": 0.8923659647382424, + "grad_norm": 1.697820750454664, + "learning_rate": 1.2021545556429648e-07, + "loss": 0.9825, + "step": 9895 + }, + { + "epoch": 0.8924561482617126, + "grad_norm": 1.3517942285062616, + "learning_rate": 1.2001604471421245e-07, + "loss": 0.9744, + "step": 9896 + }, + { + "epoch": 0.8925463317851828, + "grad_norm": 1.444183320912288, + "learning_rate": 1.1981679427472567e-07, + "loss": 1.0275, + "step": 9897 + }, + { + "epoch": 0.8926365153086531, + "grad_norm": 1.7553796505354726, + "learning_rate": 1.196177042628368e-07, + "loss": 1.0286, + "step": 9898 + }, + { + "epoch": 0.8927266988321234, + "grad_norm": 78.64594129407601, + "learning_rate": 1.194187746955344e-07, + "loss": 0.9798, + "step": 9899 + }, + { + "epoch": 0.8928168823555936, + "grad_norm": 1.3871069995942507, + "learning_rate": 1.1922000558979094e-07, + "loss": 0.9541, + "step": 9900 + }, + { + "epoch": 0.8929070658790639, + "grad_norm": 1.4928533057308881, + "learning_rate": 1.1902139696256752e-07, + "loss": 1.0067, + "step": 9901 + }, + { + "epoch": 0.8929972494025341, + "grad_norm": 1.5803488745792529, + "learning_rate": 1.188229488308099e-07, + "loss": 0.9147, + "step": 9902 + }, + { + "epoch": 0.8930874329260045, + "grad_norm": 0.5655359619622232, + "learning_rate": 1.1862466121145098e-07, + "loss": 0.7941, + "step": 9903 + }, + { + "epoch": 0.8931776164494747, + "grad_norm": 2.2071562334978925, + "learning_rate": 1.184265341214099e-07, + "loss": 0.9196, + "step": 9904 + }, + { + "epoch": 0.8932677999729449, + "grad_norm": 1.5008357421895206, + "learning_rate": 1.182285675775918e-07, + "loss": 1.0007, + "step": 9905 + }, + { + "epoch": 0.8933579834964152, + "grad_norm": 2.029049899627464, + "learning_rate": 1.1803076159688851e-07, + "loss": 0.9597, + "step": 9906 + }, + { + "epoch": 0.8934481670198855, + "grad_norm": 1.763258707141566, + "learning_rate": 1.1783311619617741e-07, + "loss": 1.0083, + "step": 9907 + }, + { + "epoch": 0.8935383505433557, + "grad_norm": 1.7502593120822525, + "learning_rate": 1.1763563139232302e-07, + "loss": 0.9432, + "step": 9908 + }, + { + "epoch": 0.893628534066826, + "grad_norm": 1.9057426402812967, + "learning_rate": 1.1743830720217562e-07, + "loss": 0.9155, + "step": 9909 + }, + { + "epoch": 0.8937187175902963, + "grad_norm": 2.0822815592528294, + "learning_rate": 1.1724114364257243e-07, + "loss": 0.9502, + "step": 9910 + }, + { + "epoch": 0.8938089011137665, + "grad_norm": 1.5336554959382362, + "learning_rate": 1.1704414073033619e-07, + "loss": 0.9846, + "step": 9911 + }, + { + "epoch": 0.8938990846372368, + "grad_norm": 1.7520572967348613, + "learning_rate": 1.1684729848227636e-07, + "loss": 1.0177, + "step": 9912 + }, + { + "epoch": 0.893989268160707, + "grad_norm": 1.5666709923515392, + "learning_rate": 1.1665061691518884e-07, + "loss": 1.034, + "step": 9913 + }, + { + "epoch": 0.8940794516841774, + "grad_norm": 2.0352813623671357, + "learning_rate": 1.1645409604585532e-07, + "loss": 0.979, + "step": 9914 + }, + { + "epoch": 0.8941696352076476, + "grad_norm": 1.5158861073008147, + "learning_rate": 1.162577358910437e-07, + "loss": 0.9817, + "step": 9915 + }, + { + "epoch": 0.8942598187311178, + "grad_norm": 1.5978846305145726, + "learning_rate": 1.160615364675095e-07, + "loss": 0.9598, + "step": 9916 + }, + { + "epoch": 0.894350002254588, + "grad_norm": 0.612250926503438, + "learning_rate": 1.1586549779199262e-07, + "loss": 0.7957, + "step": 9917 + }, + { + "epoch": 0.8944401857780584, + "grad_norm": 1.4818324750560905, + "learning_rate": 1.1566961988122037e-07, + "loss": 1.0334, + "step": 9918 + }, + { + "epoch": 0.8945303693015286, + "grad_norm": 1.8200847172145957, + "learning_rate": 1.1547390275190627e-07, + "loss": 0.9413, + "step": 9919 + }, + { + "epoch": 0.8946205528249989, + "grad_norm": 1.5521434883220218, + "learning_rate": 1.1527834642075007e-07, + "loss": 0.9183, + "step": 9920 + }, + { + "epoch": 0.8947107363484691, + "grad_norm": 1.7332963768369374, + "learning_rate": 1.1508295090443797e-07, + "loss": 0.96, + "step": 9921 + }, + { + "epoch": 0.8948009198719394, + "grad_norm": 1.5288216935421677, + "learning_rate": 1.148877162196411e-07, + "loss": 0.9272, + "step": 9922 + }, + { + "epoch": 0.8948911033954097, + "grad_norm": 2.1347584772295707, + "learning_rate": 1.1469264238301924e-07, + "loss": 0.9076, + "step": 9923 + }, + { + "epoch": 0.8949812869188799, + "grad_norm": 1.7174367569297406, + "learning_rate": 1.1449772941121638e-07, + "loss": 1.0036, + "step": 9924 + }, + { + "epoch": 0.8950714704423501, + "grad_norm": 1.9312674043235347, + "learning_rate": 1.1430297732086369e-07, + "loss": 0.9476, + "step": 9925 + }, + { + "epoch": 0.8951616539658205, + "grad_norm": 1.5527709956619076, + "learning_rate": 1.1410838612857876e-07, + "loss": 0.9649, + "step": 9926 + }, + { + "epoch": 0.8952518374892907, + "grad_norm": 2.065311711686801, + "learning_rate": 1.1391395585096497e-07, + "loss": 1.0786, + "step": 9927 + }, + { + "epoch": 0.895342021012761, + "grad_norm": 3.019242562808787, + "learning_rate": 1.1371968650461216e-07, + "loss": 0.9069, + "step": 9928 + }, + { + "epoch": 0.8954322045362312, + "grad_norm": 0.69684874737475, + "learning_rate": 1.1352557810609687e-07, + "loss": 0.8096, + "step": 9929 + }, + { + "epoch": 0.8955223880597015, + "grad_norm": 1.5227417751071965, + "learning_rate": 1.1333163067198048e-07, + "loss": 0.9616, + "step": 9930 + }, + { + "epoch": 0.8956125715831718, + "grad_norm": 1.4935738080550396, + "learning_rate": 1.1313784421881311e-07, + "loss": 1.0415, + "step": 9931 + }, + { + "epoch": 0.895702755106642, + "grad_norm": 1.6612212949731269, + "learning_rate": 1.1294421876312865e-07, + "loss": 1.0124, + "step": 9932 + }, + { + "epoch": 0.8957929386301122, + "grad_norm": 2.055425359136746, + "learning_rate": 1.1275075432144831e-07, + "loss": 0.9523, + "step": 9933 + }, + { + "epoch": 0.8958831221535826, + "grad_norm": 1.8868746801668586, + "learning_rate": 1.1255745091028002e-07, + "loss": 0.9947, + "step": 9934 + }, + { + "epoch": 0.8959733056770528, + "grad_norm": 1.5791972832576182, + "learning_rate": 1.1236430854611723e-07, + "loss": 0.9849, + "step": 9935 + }, + { + "epoch": 0.896063489200523, + "grad_norm": 2.413898252559987, + "learning_rate": 1.1217132724544032e-07, + "loss": 0.9403, + "step": 9936 + }, + { + "epoch": 0.8961536727239934, + "grad_norm": 1.7399750076340643, + "learning_rate": 1.1197850702471434e-07, + "loss": 1.0237, + "step": 9937 + }, + { + "epoch": 0.8962438562474636, + "grad_norm": 2.16514180395176, + "learning_rate": 1.1178584790039348e-07, + "loss": 1.043, + "step": 9938 + }, + { + "epoch": 0.8963340397709338, + "grad_norm": 7.969614498143774, + "learning_rate": 1.1159334988891478e-07, + "loss": 0.9222, + "step": 9939 + }, + { + "epoch": 0.8964242232944041, + "grad_norm": 2.1189217871957156, + "learning_rate": 1.1140101300670446e-07, + "loss": 0.9518, + "step": 9940 + }, + { + "epoch": 0.8965144068178744, + "grad_norm": 1.6147734827441682, + "learning_rate": 1.1120883727017338e-07, + "loss": 0.8429, + "step": 9941 + }, + { + "epoch": 0.8966045903413447, + "grad_norm": 1.6501362079696011, + "learning_rate": 1.1101682269571889e-07, + "loss": 1.0249, + "step": 9942 + }, + { + "epoch": 0.8966947738648149, + "grad_norm": 1.7771482255304372, + "learning_rate": 1.1082496929972473e-07, + "loss": 0.9523, + "step": 9943 + }, + { + "epoch": 0.8967849573882851, + "grad_norm": 1.6827240930426695, + "learning_rate": 1.1063327709856096e-07, + "loss": 0.943, + "step": 9944 + }, + { + "epoch": 0.8968751409117555, + "grad_norm": 1.7836031846253495, + "learning_rate": 1.1044174610858403e-07, + "loss": 0.8626, + "step": 9945 + }, + { + "epoch": 0.8969653244352257, + "grad_norm": 1.731293217088598, + "learning_rate": 1.1025037634613643e-07, + "loss": 0.9313, + "step": 9946 + }, + { + "epoch": 0.8970555079586959, + "grad_norm": 1.5570531128660996, + "learning_rate": 1.1005916782754643e-07, + "loss": 1.0025, + "step": 9947 + }, + { + "epoch": 0.8971456914821662, + "grad_norm": 1.8671154822343001, + "learning_rate": 1.0986812056912919e-07, + "loss": 0.9608, + "step": 9948 + }, + { + "epoch": 0.8972358750056365, + "grad_norm": 1.581395000096999, + "learning_rate": 1.0967723458718613e-07, + "loss": 0.8561, + "step": 9949 + }, + { + "epoch": 0.8973260585291067, + "grad_norm": 2.4305374278043455, + "learning_rate": 1.0948650989800445e-07, + "loss": 0.9444, + "step": 9950 + }, + { + "epoch": 0.897416242052577, + "grad_norm": 1.4970429200853383, + "learning_rate": 1.0929594651785823e-07, + "loss": 0.9592, + "step": 9951 + }, + { + "epoch": 0.8975064255760472, + "grad_norm": 2.0968831092006828, + "learning_rate": 1.0910554446300646e-07, + "loss": 1.0276, + "step": 9952 + }, + { + "epoch": 0.8975966090995176, + "grad_norm": 1.705274013527491, + "learning_rate": 1.089153037496966e-07, + "loss": 0.8752, + "step": 9953 + }, + { + "epoch": 0.8976867926229878, + "grad_norm": 1.7541029380161766, + "learning_rate": 1.0872522439415966e-07, + "loss": 0.9973, + "step": 9954 + }, + { + "epoch": 0.897776976146458, + "grad_norm": 2.06244971474123, + "learning_rate": 1.0853530641261554e-07, + "loss": 0.922, + "step": 9955 + }, + { + "epoch": 0.8978671596699282, + "grad_norm": 1.7612477283991073, + "learning_rate": 1.083455498212682e-07, + "loss": 0.9869, + "step": 9956 + }, + { + "epoch": 0.8979573431933986, + "grad_norm": 1.9897118838125074, + "learning_rate": 1.0815595463630911e-07, + "loss": 1.0132, + "step": 9957 + }, + { + "epoch": 0.8980475267168688, + "grad_norm": 1.767678004019444, + "learning_rate": 1.0796652087391556e-07, + "loss": 0.8471, + "step": 9958 + }, + { + "epoch": 0.8981377102403391, + "grad_norm": 1.4618126191214973, + "learning_rate": 1.0777724855025083e-07, + "loss": 1.0754, + "step": 9959 + }, + { + "epoch": 0.8982278937638094, + "grad_norm": 1.531008720851506, + "learning_rate": 1.075881376814649e-07, + "loss": 0.9747, + "step": 9960 + }, + { + "epoch": 0.8983180772872796, + "grad_norm": 2.4654634062558443, + "learning_rate": 1.0739918828369377e-07, + "loss": 1.0324, + "step": 9961 + }, + { + "epoch": 0.8984082608107499, + "grad_norm": 1.7499917575097115, + "learning_rate": 1.0721040037305983e-07, + "loss": 0.9834, + "step": 9962 + }, + { + "epoch": 0.8984984443342201, + "grad_norm": 1.6285845029396946, + "learning_rate": 1.0702177396567114e-07, + "loss": 0.9274, + "step": 9963 + }, + { + "epoch": 0.8985886278576904, + "grad_norm": 0.6229905967048679, + "learning_rate": 1.0683330907762233e-07, + "loss": 0.8, + "step": 9964 + }, + { + "epoch": 0.8986788113811607, + "grad_norm": 1.6597118545839762, + "learning_rate": 1.0664500572499435e-07, + "loss": 0.9061, + "step": 9965 + }, + { + "epoch": 0.8987689949046309, + "grad_norm": 1.8183473121578935, + "learning_rate": 1.0645686392385455e-07, + "loss": 0.9178, + "step": 9966 + }, + { + "epoch": 0.8988591784281011, + "grad_norm": 1.5390990586949191, + "learning_rate": 1.0626888369025588e-07, + "loss": 0.9024, + "step": 9967 + }, + { + "epoch": 0.8989493619515715, + "grad_norm": 1.4374351486836345, + "learning_rate": 1.0608106504023817e-07, + "loss": 1.0052, + "step": 9968 + }, + { + "epoch": 0.8990395454750417, + "grad_norm": 1.5523358679361576, + "learning_rate": 1.0589340798982637e-07, + "loss": 0.9649, + "step": 9969 + }, + { + "epoch": 0.899129728998512, + "grad_norm": 2.227898340881708, + "learning_rate": 1.057059125550337e-07, + "loss": 0.9192, + "step": 9970 + }, + { + "epoch": 0.8992199125219822, + "grad_norm": 1.3528642504498607, + "learning_rate": 1.0551857875185732e-07, + "loss": 0.9851, + "step": 9971 + }, + { + "epoch": 0.8993100960454525, + "grad_norm": 1.2685868726821845, + "learning_rate": 1.0533140659628181e-07, + "loss": 0.8644, + "step": 9972 + }, + { + "epoch": 0.8994002795689228, + "grad_norm": 1.565988308903108, + "learning_rate": 1.0514439610427772e-07, + "loss": 1.0293, + "step": 9973 + }, + { + "epoch": 0.899490463092393, + "grad_norm": 1.6135512303889206, + "learning_rate": 1.0495754729180206e-07, + "loss": 1.0011, + "step": 9974 + }, + { + "epoch": 0.8995806466158632, + "grad_norm": 1.3538691952010111, + "learning_rate": 1.0477086017479741e-07, + "loss": 1.0759, + "step": 9975 + }, + { + "epoch": 0.8996708301393336, + "grad_norm": 1.7122974303787841, + "learning_rate": 1.0458433476919327e-07, + "loss": 0.9341, + "step": 9976 + }, + { + "epoch": 0.8997610136628038, + "grad_norm": 1.3838654727878328, + "learning_rate": 1.0439797109090509e-07, + "loss": 1.0025, + "step": 9977 + }, + { + "epoch": 0.899851197186274, + "grad_norm": 1.7502989513633598, + "learning_rate": 1.0421176915583396e-07, + "loss": 0.9475, + "step": 9978 + }, + { + "epoch": 0.8999413807097443, + "grad_norm": 1.732472312562887, + "learning_rate": 1.0402572897986828e-07, + "loss": 1.0078, + "step": 9979 + }, + { + "epoch": 0.9000315642332146, + "grad_norm": 0.5870284541867905, + "learning_rate": 1.0383985057888134e-07, + "loss": 0.8053, + "step": 9980 + }, + { + "epoch": 0.9001217477566849, + "grad_norm": 1.716833467928952, + "learning_rate": 1.036541339687338e-07, + "loss": 0.8805, + "step": 9981 + }, + { + "epoch": 0.9002119312801551, + "grad_norm": 0.6409324512998914, + "learning_rate": 1.0346857916527186e-07, + "loss": 0.7756, + "step": 9982 + }, + { + "epoch": 0.9003021148036254, + "grad_norm": 1.54924561693188, + "learning_rate": 1.0328318618432819e-07, + "loss": 0.9741, + "step": 9983 + }, + { + "epoch": 0.9003922983270957, + "grad_norm": 1.8475299816254027, + "learning_rate": 1.0309795504172148e-07, + "loss": 1.0366, + "step": 9984 + }, + { + "epoch": 0.9004824818505659, + "grad_norm": 1.874200968560626, + "learning_rate": 1.0291288575325685e-07, + "loss": 0.9522, + "step": 9985 + }, + { + "epoch": 0.9005726653740361, + "grad_norm": 1.7449776925133984, + "learning_rate": 1.0272797833472502e-07, + "loss": 0.9401, + "step": 9986 + }, + { + "epoch": 0.9006628488975065, + "grad_norm": 0.6928616738838391, + "learning_rate": 1.0254323280190335e-07, + "loss": 0.7846, + "step": 9987 + }, + { + "epoch": 0.9007530324209767, + "grad_norm": 2.033978668843722, + "learning_rate": 1.023586491705557e-07, + "loss": 0.9545, + "step": 9988 + }, + { + "epoch": 0.9008432159444469, + "grad_norm": 2.361300257667447, + "learning_rate": 1.0217422745643145e-07, + "loss": 0.9136, + "step": 9989 + }, + { + "epoch": 0.9009333994679172, + "grad_norm": 1.4906911487417136, + "learning_rate": 1.0198996767526691e-07, + "loss": 0.8965, + "step": 9990 + }, + { + "epoch": 0.9010235829913875, + "grad_norm": 1.799046738187608, + "learning_rate": 1.018058698427835e-07, + "loss": 0.9546, + "step": 9991 + }, + { + "epoch": 0.9011137665148578, + "grad_norm": 1.3949392881137133, + "learning_rate": 1.0162193397469021e-07, + "loss": 1.0193, + "step": 9992 + }, + { + "epoch": 0.901203950038328, + "grad_norm": 2.0446591796033107, + "learning_rate": 1.0143816008668049e-07, + "loss": 0.9146, + "step": 9993 + }, + { + "epoch": 0.9012941335617982, + "grad_norm": 0.6381607576938352, + "learning_rate": 1.0125454819443624e-07, + "loss": 0.8233, + "step": 9994 + }, + { + "epoch": 0.9013843170852686, + "grad_norm": 1.3436169669368712, + "learning_rate": 1.0107109831362315e-07, + "loss": 0.8835, + "step": 9995 + }, + { + "epoch": 0.9014745006087388, + "grad_norm": 1.8186349629205234, + "learning_rate": 1.0088781045989447e-07, + "loss": 0.9813, + "step": 9996 + }, + { + "epoch": 0.901564684132209, + "grad_norm": 1.4466160849475593, + "learning_rate": 1.0070468464888926e-07, + "loss": 1.0253, + "step": 9997 + }, + { + "epoch": 0.9016548676556793, + "grad_norm": 1.8917918112051202, + "learning_rate": 1.0052172089623324e-07, + "loss": 0.9931, + "step": 9998 + }, + { + "epoch": 0.9017450511791496, + "grad_norm": 1.5380640034398334, + "learning_rate": 1.0033891921753746e-07, + "loss": 1.0129, + "step": 9999 + }, + { + "epoch": 0.9018352347026198, + "grad_norm": 1.6361879202633325, + "learning_rate": 1.0015627962839968e-07, + "loss": 0.8863, + "step": 10000 + }, + { + "epoch": 0.9019254182260901, + "grad_norm": 1.5978618759980683, + "learning_rate": 9.99738021444041e-08, + "loss": 0.9508, + "step": 10001 + }, + { + "epoch": 0.9020156017495603, + "grad_norm": 1.3667785032770539, + "learning_rate": 9.979148678112003e-08, + "loss": 0.9039, + "step": 10002 + }, + { + "epoch": 0.9021057852730306, + "grad_norm": 2.252923020450961, + "learning_rate": 9.960933355410417e-08, + "loss": 1.0752, + "step": 10003 + }, + { + "epoch": 0.9021959687965009, + "grad_norm": 1.5413925425504376, + "learning_rate": 9.942734247889828e-08, + "loss": 0.8889, + "step": 10004 + }, + { + "epoch": 0.9022861523199711, + "grad_norm": 1.684544058332551, + "learning_rate": 9.92455135710315e-08, + "loss": 0.9827, + "step": 10005 + }, + { + "epoch": 0.9023763358434415, + "grad_norm": 2.2469181571285555, + "learning_rate": 9.906384684601787e-08, + "loss": 0.9045, + "step": 10006 + }, + { + "epoch": 0.9024665193669117, + "grad_norm": 1.5572035476436423, + "learning_rate": 9.8882342319359e-08, + "loss": 1.0183, + "step": 10007 + }, + { + "epoch": 0.9025567028903819, + "grad_norm": 1.484213248022901, + "learning_rate": 9.870100000654048e-08, + "loss": 0.9242, + "step": 10008 + }, + { + "epoch": 0.9026468864138522, + "grad_norm": 1.2757822460724306, + "learning_rate": 9.851981992303704e-08, + "loss": 0.9451, + "step": 10009 + }, + { + "epoch": 0.9027370699373225, + "grad_norm": 3.768744984587057, + "learning_rate": 9.833880208430678e-08, + "loss": 0.9792, + "step": 10010 + }, + { + "epoch": 0.9028272534607927, + "grad_norm": 1.9038464135686286, + "learning_rate": 9.815794650579601e-08, + "loss": 0.9688, + "step": 10011 + }, + { + "epoch": 0.902917436984263, + "grad_norm": 1.492999190614536, + "learning_rate": 9.797725320293548e-08, + "loss": 0.9505, + "step": 10012 + }, + { + "epoch": 0.9030076205077332, + "grad_norm": 3.015058711351729, + "learning_rate": 9.779672219114354e-08, + "loss": 1.0501, + "step": 10013 + }, + { + "epoch": 0.9030978040312035, + "grad_norm": 1.9489645018107018, + "learning_rate": 9.761635348582386e-08, + "loss": 0.9527, + "step": 10014 + }, + { + "epoch": 0.9031879875546738, + "grad_norm": 1.8831117004471838, + "learning_rate": 9.743614710236658e-08, + "loss": 1.005, + "step": 10015 + }, + { + "epoch": 0.903278171078144, + "grad_norm": 1.5898924873998488, + "learning_rate": 9.725610305614806e-08, + "loss": 0.9341, + "step": 10016 + }, + { + "epoch": 0.9033683546016142, + "grad_norm": 1.471849944271959, + "learning_rate": 9.707622136253002e-08, + "loss": 0.9678, + "step": 10017 + }, + { + "epoch": 0.9034585381250846, + "grad_norm": 1.4492883382783022, + "learning_rate": 9.689650203686173e-08, + "loss": 0.9972, + "step": 10018 + }, + { + "epoch": 0.9035487216485548, + "grad_norm": 1.9539497160166526, + "learning_rate": 9.671694509447715e-08, + "loss": 0.9542, + "step": 10019 + }, + { + "epoch": 0.903638905172025, + "grad_norm": 7.40572693341805, + "learning_rate": 9.653755055069757e-08, + "loss": 0.9899, + "step": 10020 + }, + { + "epoch": 0.9037290886954953, + "grad_norm": 1.432510340788102, + "learning_rate": 9.635831842082987e-08, + "loss": 0.9509, + "step": 10021 + }, + { + "epoch": 0.9038192722189656, + "grad_norm": 2.157100454489326, + "learning_rate": 9.617924872016691e-08, + "loss": 0.9888, + "step": 10022 + }, + { + "epoch": 0.9039094557424359, + "grad_norm": 1.7755478725312215, + "learning_rate": 9.600034146398806e-08, + "loss": 1.0127, + "step": 10023 + }, + { + "epoch": 0.9039996392659061, + "grad_norm": 0.6572045695237141, + "learning_rate": 9.582159666755863e-08, + "loss": 0.819, + "step": 10024 + }, + { + "epoch": 0.9040898227893763, + "grad_norm": 2.966306946051497, + "learning_rate": 9.564301434612976e-08, + "loss": 0.9425, + "step": 10025 + }, + { + "epoch": 0.9041800063128467, + "grad_norm": 1.558501812128232, + "learning_rate": 9.546459451494015e-08, + "loss": 1.0147, + "step": 10026 + }, + { + "epoch": 0.9042701898363169, + "grad_norm": 0.6457363999337987, + "learning_rate": 9.528633718921231e-08, + "loss": 0.8424, + "step": 10027 + }, + { + "epoch": 0.9043603733597871, + "grad_norm": 1.3673624853141564, + "learning_rate": 9.510824238415672e-08, + "loss": 1.022, + "step": 10028 + }, + { + "epoch": 0.9044505568832575, + "grad_norm": 1.9604286050247202, + "learning_rate": 9.493031011496944e-08, + "loss": 1.005, + "step": 10029 + }, + { + "epoch": 0.9045407404067277, + "grad_norm": 1.5610366071945478, + "learning_rate": 9.475254039683234e-08, + "loss": 0.8297, + "step": 10030 + }, + { + "epoch": 0.904630923930198, + "grad_norm": 1.3463859431380136, + "learning_rate": 9.45749332449144e-08, + "loss": 0.9363, + "step": 10031 + }, + { + "epoch": 0.9047211074536682, + "grad_norm": 0.6437811214368566, + "learning_rate": 9.439748867436903e-08, + "loss": 0.826, + "step": 10032 + }, + { + "epoch": 0.9048112909771385, + "grad_norm": 1.739156095688624, + "learning_rate": 9.42202067003377e-08, + "loss": 0.9368, + "step": 10033 + }, + { + "epoch": 0.9049014745006088, + "grad_norm": 1.6943891269569116, + "learning_rate": 9.404308733794652e-08, + "loss": 1.1104, + "step": 10034 + }, + { + "epoch": 0.904991658024079, + "grad_norm": 1.558056731673712, + "learning_rate": 9.38661306023083e-08, + "loss": 0.9929, + "step": 10035 + }, + { + "epoch": 0.9050818415475492, + "grad_norm": 2.4025781532619783, + "learning_rate": 9.368933650852229e-08, + "loss": 0.9664, + "step": 10036 + }, + { + "epoch": 0.9051720250710196, + "grad_norm": 0.7766040752495533, + "learning_rate": 9.351270507167352e-08, + "loss": 0.8309, + "step": 10037 + }, + { + "epoch": 0.9052622085944898, + "grad_norm": 1.735192389801031, + "learning_rate": 9.333623630683285e-08, + "loss": 0.9127, + "step": 10038 + }, + { + "epoch": 0.90535239211796, + "grad_norm": 1.6663536811044632, + "learning_rate": 9.315993022905799e-08, + "loss": 0.8255, + "step": 10039 + }, + { + "epoch": 0.9054425756414303, + "grad_norm": 1.6192625434723003, + "learning_rate": 9.298378685339158e-08, + "loss": 0.9516, + "step": 10040 + }, + { + "epoch": 0.9055327591649006, + "grad_norm": 3.396627711583386, + "learning_rate": 9.280780619486406e-08, + "loss": 1.0376, + "step": 10041 + }, + { + "epoch": 0.9056229426883708, + "grad_norm": 1.5031839593121818, + "learning_rate": 9.26319882684905e-08, + "loss": 1.0529, + "step": 10042 + }, + { + "epoch": 0.9057131262118411, + "grad_norm": 1.397440080467062, + "learning_rate": 9.245633308927293e-08, + "loss": 0.9419, + "step": 10043 + }, + { + "epoch": 0.9058033097353113, + "grad_norm": 1.5852907863508254, + "learning_rate": 9.228084067219888e-08, + "loss": 0.9559, + "step": 10044 + }, + { + "epoch": 0.9058934932587817, + "grad_norm": 2.772396526280446, + "learning_rate": 9.210551103224284e-08, + "loss": 0.9213, + "step": 10045 + }, + { + "epoch": 0.9059836767822519, + "grad_norm": 1.5236649563521825, + "learning_rate": 9.193034418436463e-08, + "loss": 1.0275, + "step": 10046 + }, + { + "epoch": 0.9060738603057221, + "grad_norm": 0.6493980233642577, + "learning_rate": 9.175534014351005e-08, + "loss": 0.8505, + "step": 10047 + }, + { + "epoch": 0.9061640438291924, + "grad_norm": 2.4822657046052257, + "learning_rate": 9.158049892461228e-08, + "loss": 0.9227, + "step": 10048 + }, + { + "epoch": 0.9062542273526627, + "grad_norm": 1.3083680826581745, + "learning_rate": 9.140582054258871e-08, + "loss": 0.9108, + "step": 10049 + }, + { + "epoch": 0.9063444108761329, + "grad_norm": 1.6313976381803652, + "learning_rate": 9.123130501234499e-08, + "loss": 0.9819, + "step": 10050 + }, + { + "epoch": 0.9064345943996032, + "grad_norm": 3.3638911209358042, + "learning_rate": 9.105695234877098e-08, + "loss": 1.0244, + "step": 10051 + }, + { + "epoch": 0.9065247779230734, + "grad_norm": 1.3563901204799256, + "learning_rate": 9.088276256674344e-08, + "loss": 0.9981, + "step": 10052 + }, + { + "epoch": 0.9066149614465437, + "grad_norm": 1.6325562928616528, + "learning_rate": 9.070873568112536e-08, + "loss": 0.9596, + "step": 10053 + }, + { + "epoch": 0.906705144970014, + "grad_norm": 1.6824684010864206, + "learning_rate": 9.053487170676577e-08, + "loss": 1.0292, + "step": 10054 + }, + { + "epoch": 0.9067953284934842, + "grad_norm": 1.6080442686013012, + "learning_rate": 9.036117065849968e-08, + "loss": 1.0326, + "step": 10055 + }, + { + "epoch": 0.9068855120169546, + "grad_norm": 1.5906582182467737, + "learning_rate": 9.018763255114837e-08, + "loss": 0.9508, + "step": 10056 + }, + { + "epoch": 0.9069756955404248, + "grad_norm": 1.630071429109293, + "learning_rate": 9.00142573995184e-08, + "loss": 0.9428, + "step": 10057 + }, + { + "epoch": 0.907065879063895, + "grad_norm": 1.5758452614760292, + "learning_rate": 8.984104521840375e-08, + "loss": 1.0513, + "step": 10058 + }, + { + "epoch": 0.9071560625873653, + "grad_norm": 4.335630883725241, + "learning_rate": 8.966799602258346e-08, + "loss": 0.9942, + "step": 10059 + }, + { + "epoch": 0.9072462461108356, + "grad_norm": 1.6410796489180584, + "learning_rate": 8.949510982682329e-08, + "loss": 1.0185, + "step": 10060 + }, + { + "epoch": 0.9073364296343058, + "grad_norm": 2.7106380242265047, + "learning_rate": 8.932238664587499e-08, + "loss": 1.0451, + "step": 10061 + }, + { + "epoch": 0.9074266131577761, + "grad_norm": 1.8347778553805827, + "learning_rate": 8.914982649447567e-08, + "loss": 1.0661, + "step": 10062 + }, + { + "epoch": 0.9075167966812463, + "grad_norm": 1.539281994034694, + "learning_rate": 8.897742938734975e-08, + "loss": 1.0422, + "step": 10063 + }, + { + "epoch": 0.9076069802047166, + "grad_norm": 1.535360541109377, + "learning_rate": 8.880519533920661e-08, + "loss": 0.9977, + "step": 10064 + }, + { + "epoch": 0.9076971637281869, + "grad_norm": 1.2927048059374902, + "learning_rate": 8.863312436474268e-08, + "loss": 0.9748, + "step": 10065 + }, + { + "epoch": 0.9077873472516571, + "grad_norm": 1.9231785640999257, + "learning_rate": 8.846121647863936e-08, + "loss": 0.9795, + "step": 10066 + }, + { + "epoch": 0.9078775307751273, + "grad_norm": 1.4669878514052712, + "learning_rate": 8.828947169556555e-08, + "loss": 0.9792, + "step": 10067 + }, + { + "epoch": 0.9079677142985977, + "grad_norm": 1.2946500881350171, + "learning_rate": 8.81178900301749e-08, + "loss": 0.9394, + "step": 10068 + }, + { + "epoch": 0.9080578978220679, + "grad_norm": 2.571850533907872, + "learning_rate": 8.794647149710787e-08, + "loss": 0.8898, + "step": 10069 + }, + { + "epoch": 0.9081480813455381, + "grad_norm": 1.6321337443553763, + "learning_rate": 8.777521611099081e-08, + "loss": 0.9685, + "step": 10070 + }, + { + "epoch": 0.9082382648690084, + "grad_norm": 1.5292472977476392, + "learning_rate": 8.760412388643624e-08, + "loss": 0.9122, + "step": 10071 + }, + { + "epoch": 0.9083284483924787, + "grad_norm": 1.5693557633558266, + "learning_rate": 8.74331948380429e-08, + "loss": 0.9884, + "step": 10072 + }, + { + "epoch": 0.908418631915949, + "grad_norm": 21.960451124105322, + "learning_rate": 8.726242898039516e-08, + "loss": 0.973, + "step": 10073 + }, + { + "epoch": 0.9085088154394192, + "grad_norm": 1.7517565358582914, + "learning_rate": 8.709182632806334e-08, + "loss": 0.9824, + "step": 10074 + }, + { + "epoch": 0.9085989989628894, + "grad_norm": 2.2692660835034877, + "learning_rate": 8.692138689560469e-08, + "loss": 0.9817, + "step": 10075 + }, + { + "epoch": 0.9086891824863598, + "grad_norm": 1.4734931813921404, + "learning_rate": 8.675111069756203e-08, + "loss": 0.9599, + "step": 10076 + }, + { + "epoch": 0.90877936600983, + "grad_norm": 1.766384940206558, + "learning_rate": 8.658099774846395e-08, + "loss": 0.9459, + "step": 10077 + }, + { + "epoch": 0.9088695495333002, + "grad_norm": 1.6353017276618325, + "learning_rate": 8.641104806282595e-08, + "loss": 0.9196, + "step": 10078 + }, + { + "epoch": 0.9089597330567706, + "grad_norm": 1.9704352537996532, + "learning_rate": 8.624126165514845e-08, + "loss": 0.9819, + "step": 10079 + }, + { + "epoch": 0.9090499165802408, + "grad_norm": 1.8213474458638272, + "learning_rate": 8.607163853991917e-08, + "loss": 0.9387, + "step": 10080 + }, + { + "epoch": 0.909140100103711, + "grad_norm": 1.6150928882777928, + "learning_rate": 8.590217873161054e-08, + "loss": 0.8212, + "step": 10081 + }, + { + "epoch": 0.9092302836271813, + "grad_norm": 1.601623161841829, + "learning_rate": 8.573288224468255e-08, + "loss": 0.9905, + "step": 10082 + }, + { + "epoch": 0.9093204671506516, + "grad_norm": 1.7824002617623498, + "learning_rate": 8.556374909358011e-08, + "loss": 0.9012, + "step": 10083 + }, + { + "epoch": 0.9094106506741219, + "grad_norm": 1.69118605803677, + "learning_rate": 8.539477929273476e-08, + "loss": 0.9486, + "step": 10084 + }, + { + "epoch": 0.9095008341975921, + "grad_norm": 1.4555216068615304, + "learning_rate": 8.522597285656386e-08, + "loss": 1.0068, + "step": 10085 + }, + { + "epoch": 0.9095910177210623, + "grad_norm": 0.7152877012714604, + "learning_rate": 8.505732979947078e-08, + "loss": 0.8268, + "step": 10086 + }, + { + "epoch": 0.9096812012445327, + "grad_norm": 1.2912721595450645, + "learning_rate": 8.488885013584557e-08, + "loss": 0.9884, + "step": 10087 + }, + { + "epoch": 0.9097713847680029, + "grad_norm": 1.522137130940834, + "learning_rate": 8.472053388006295e-08, + "loss": 0.9162, + "step": 10088 + }, + { + "epoch": 0.9098615682914731, + "grad_norm": 1.9222390675158816, + "learning_rate": 8.455238104648565e-08, + "loss": 1.061, + "step": 10089 + }, + { + "epoch": 0.9099517518149434, + "grad_norm": 1.5876898644516104, + "learning_rate": 8.438439164946043e-08, + "loss": 0.8463, + "step": 10090 + }, + { + "epoch": 0.9100419353384137, + "grad_norm": 2.8055975242555964, + "learning_rate": 8.42165657033218e-08, + "loss": 0.9896, + "step": 10091 + }, + { + "epoch": 0.9101321188618839, + "grad_norm": 5.5844350648285035, + "learning_rate": 8.4048903222389e-08, + "loss": 1.0127, + "step": 10092 + }, + { + "epoch": 0.9102223023853542, + "grad_norm": 1.4988631072645853, + "learning_rate": 8.388140422096856e-08, + "loss": 1.028, + "step": 10093 + }, + { + "epoch": 0.9103124859088244, + "grad_norm": 2.2953903694005975, + "learning_rate": 8.371406871335173e-08, + "loss": 0.9857, + "step": 10094 + }, + { + "epoch": 0.9104026694322948, + "grad_norm": 1.6904217892290931, + "learning_rate": 8.354689671381732e-08, + "loss": 0.978, + "step": 10095 + }, + { + "epoch": 0.910492852955765, + "grad_norm": 1.403360832123132, + "learning_rate": 8.337988823662834e-08, + "loss": 0.9729, + "step": 10096 + }, + { + "epoch": 0.9105830364792352, + "grad_norm": 1.7955063624745995, + "learning_rate": 8.321304329603607e-08, + "loss": 0.9275, + "step": 10097 + }, + { + "epoch": 0.9106732200027055, + "grad_norm": 0.6245346005492801, + "learning_rate": 8.304636190627557e-08, + "loss": 0.8236, + "step": 10098 + }, + { + "epoch": 0.9107634035261758, + "grad_norm": 1.8775189962835634, + "learning_rate": 8.287984408156945e-08, + "loss": 0.987, + "step": 10099 + }, + { + "epoch": 0.910853587049646, + "grad_norm": 2.056614889217539, + "learning_rate": 8.271348983612591e-08, + "loss": 0.8753, + "step": 10100 + }, + { + "epoch": 0.9109437705731163, + "grad_norm": 1.377802982898772, + "learning_rate": 8.254729918413938e-08, + "loss": 0.9346, + "step": 10101 + }, + { + "epoch": 0.9110339540965866, + "grad_norm": 1.3038639964576615, + "learning_rate": 8.238127213979006e-08, + "loss": 0.9432, + "step": 10102 + }, + { + "epoch": 0.9111241376200568, + "grad_norm": 1.78461827158062, + "learning_rate": 8.221540871724398e-08, + "loss": 1.0443, + "step": 10103 + }, + { + "epoch": 0.9112143211435271, + "grad_norm": 1.5355948484113273, + "learning_rate": 8.2049708930654e-08, + "loss": 0.8607, + "step": 10104 + }, + { + "epoch": 0.9113045046669973, + "grad_norm": 1.8192131422848088, + "learning_rate": 8.188417279415793e-08, + "loss": 0.9408, + "step": 10105 + }, + { + "epoch": 0.9113946881904677, + "grad_norm": 1.879178667840747, + "learning_rate": 8.171880032188117e-08, + "loss": 1.0638, + "step": 10106 + }, + { + "epoch": 0.9114848717139379, + "grad_norm": 1.633951693214168, + "learning_rate": 8.155359152793351e-08, + "loss": 0.9537, + "step": 10107 + }, + { + "epoch": 0.9115750552374081, + "grad_norm": 3.0588924413142244, + "learning_rate": 8.138854642641147e-08, + "loss": 0.9583, + "step": 10108 + }, + { + "epoch": 0.9116652387608783, + "grad_norm": 1.5422650627876078, + "learning_rate": 8.122366503139777e-08, + "loss": 0.8987, + "step": 10109 + }, + { + "epoch": 0.9117554222843487, + "grad_norm": 1.479771835034576, + "learning_rate": 8.105894735696117e-08, + "loss": 0.9141, + "step": 10110 + }, + { + "epoch": 0.9118456058078189, + "grad_norm": 1.9191987547329372, + "learning_rate": 8.089439341715576e-08, + "loss": 0.9541, + "step": 10111 + }, + { + "epoch": 0.9119357893312892, + "grad_norm": 1.680668602005781, + "learning_rate": 8.073000322602319e-08, + "loss": 1.0022, + "step": 10112 + }, + { + "epoch": 0.9120259728547594, + "grad_norm": 1.4370133778766332, + "learning_rate": 8.056577679758891e-08, + "loss": 0.9881, + "step": 10113 + }, + { + "epoch": 0.9121161563782297, + "grad_norm": 1.952604300708483, + "learning_rate": 8.040171414586638e-08, + "loss": 0.9272, + "step": 10114 + }, + { + "epoch": 0.9122063399017, + "grad_norm": 1.5797598651494047, + "learning_rate": 8.023781528485419e-08, + "loss": 0.9374, + "step": 10115 + }, + { + "epoch": 0.9122965234251702, + "grad_norm": 2.290912180302616, + "learning_rate": 8.00740802285369e-08, + "loss": 0.9982, + "step": 10116 + }, + { + "epoch": 0.9123867069486404, + "grad_norm": 1.654243621509483, + "learning_rate": 7.99105089908858e-08, + "loss": 0.8411, + "step": 10117 + }, + { + "epoch": 0.9124768904721108, + "grad_norm": 1.2928391590463353, + "learning_rate": 7.974710158585685e-08, + "loss": 0.9495, + "step": 10118 + }, + { + "epoch": 0.912567073995581, + "grad_norm": 1.6395024410135397, + "learning_rate": 7.958385802739375e-08, + "loss": 1.0015, + "step": 10119 + }, + { + "epoch": 0.9126572575190512, + "grad_norm": 1.6072481045245257, + "learning_rate": 7.942077832942452e-08, + "loss": 1.035, + "step": 10120 + }, + { + "epoch": 0.9127474410425215, + "grad_norm": 0.7405000879080319, + "learning_rate": 7.925786250586508e-08, + "loss": 0.818, + "step": 10121 + }, + { + "epoch": 0.9128376245659918, + "grad_norm": 0.6571403775664457, + "learning_rate": 7.909511057061524e-08, + "loss": 0.8174, + "step": 10122 + }, + { + "epoch": 0.9129278080894621, + "grad_norm": 1.7546999352575838, + "learning_rate": 7.893252253756234e-08, + "loss": 0.9614, + "step": 10123 + }, + { + "epoch": 0.9130179916129323, + "grad_norm": 1.923026569514185, + "learning_rate": 7.877009842057925e-08, + "loss": 0.8991, + "step": 10124 + }, + { + "epoch": 0.9131081751364026, + "grad_norm": 1.6093849996607954, + "learning_rate": 7.860783823352512e-08, + "loss": 1.0027, + "step": 10125 + }, + { + "epoch": 0.9131983586598729, + "grad_norm": 1.616006170940892, + "learning_rate": 7.844574199024445e-08, + "loss": 0.9832, + "step": 10126 + }, + { + "epoch": 0.9132885421833431, + "grad_norm": 1.6107912961102795, + "learning_rate": 7.82838097045686e-08, + "loss": 1.0167, + "step": 10127 + }, + { + "epoch": 0.9133787257068133, + "grad_norm": 2.2459674302114894, + "learning_rate": 7.812204139031454e-08, + "loss": 0.9613, + "step": 10128 + }, + { + "epoch": 0.9134689092302837, + "grad_norm": 2.1137469161161713, + "learning_rate": 7.796043706128474e-08, + "loss": 0.9886, + "step": 10129 + }, + { + "epoch": 0.9135590927537539, + "grad_norm": 1.8304288644063658, + "learning_rate": 7.779899673126844e-08, + "loss": 1.0049, + "step": 10130 + }, + { + "epoch": 0.9136492762772241, + "grad_norm": 1.844164009810561, + "learning_rate": 7.76377204140406e-08, + "loss": 1.0156, + "step": 10131 + }, + { + "epoch": 0.9137394598006944, + "grad_norm": 8.073762818979073, + "learning_rate": 7.74766081233622e-08, + "loss": 0.9665, + "step": 10132 + }, + { + "epoch": 0.9138296433241647, + "grad_norm": 1.4406301171466365, + "learning_rate": 7.73156598729805e-08, + "loss": 1.0424, + "step": 10133 + }, + { + "epoch": 0.913919826847635, + "grad_norm": 1.4843216735898175, + "learning_rate": 7.715487567662849e-08, + "loss": 1.0306, + "step": 10134 + }, + { + "epoch": 0.9140100103711052, + "grad_norm": 1.9642418608786327, + "learning_rate": 7.69942555480243e-08, + "loss": 0.9461, + "step": 10135 + }, + { + "epoch": 0.9141001938945754, + "grad_norm": 1.8970902925639197, + "learning_rate": 7.68337995008741e-08, + "loss": 1.0778, + "step": 10136 + }, + { + "epoch": 0.9141903774180458, + "grad_norm": 5.631424011861226, + "learning_rate": 7.667350754886803e-08, + "loss": 0.8303, + "step": 10137 + }, + { + "epoch": 0.914280560941516, + "grad_norm": 1.683035348571992, + "learning_rate": 7.651337970568361e-08, + "loss": 0.9661, + "step": 10138 + }, + { + "epoch": 0.9143707444649862, + "grad_norm": 1.5815845753731375, + "learning_rate": 7.635341598498368e-08, + "loss": 0.9382, + "step": 10139 + }, + { + "epoch": 0.9144609279884565, + "grad_norm": 1.5845961888477504, + "learning_rate": 7.61936164004171e-08, + "loss": 0.9705, + "step": 10140 + }, + { + "epoch": 0.9145511115119268, + "grad_norm": 1.4493794724830096, + "learning_rate": 7.603398096561875e-08, + "loss": 1.0195, + "step": 10141 + }, + { + "epoch": 0.914641295035397, + "grad_norm": 1.5996496919760852, + "learning_rate": 7.587450969420994e-08, + "loss": 0.9772, + "step": 10142 + }, + { + "epoch": 0.9147314785588673, + "grad_norm": 0.6837350753745374, + "learning_rate": 7.571520259979757e-08, + "loss": 0.7982, + "step": 10143 + }, + { + "epoch": 0.9148216620823375, + "grad_norm": 1.6990095338604436, + "learning_rate": 7.555605969597455e-08, + "loss": 0.8132, + "step": 10144 + }, + { + "epoch": 0.9149118456058078, + "grad_norm": 1.4097244573783172, + "learning_rate": 7.539708099631959e-08, + "loss": 1.0576, + "step": 10145 + }, + { + "epoch": 0.9150020291292781, + "grad_norm": 1.8180369286893951, + "learning_rate": 7.52382665143978e-08, + "loss": 0.9771, + "step": 10146 + }, + { + "epoch": 0.9150922126527483, + "grad_norm": 1.9801022041799794, + "learning_rate": 7.507961626376014e-08, + "loss": 1.0267, + "step": 10147 + }, + { + "epoch": 0.9151823961762187, + "grad_norm": 1.730122895478663, + "learning_rate": 7.492113025794378e-08, + "loss": 1.0127, + "step": 10148 + }, + { + "epoch": 0.9152725796996889, + "grad_norm": 1.6805864634268202, + "learning_rate": 7.476280851047101e-08, + "loss": 0.9214, + "step": 10149 + }, + { + "epoch": 0.9153627632231591, + "grad_norm": 1.2771605380660984, + "learning_rate": 7.460465103485125e-08, + "loss": 0.9979, + "step": 10150 + }, + { + "epoch": 0.9154529467466294, + "grad_norm": 1.8165118443139414, + "learning_rate": 7.444665784457948e-08, + "loss": 0.9657, + "step": 10151 + }, + { + "epoch": 0.9155431302700997, + "grad_norm": 1.9510474179643573, + "learning_rate": 7.42888289531356e-08, + "loss": 0.9315, + "step": 10152 + }, + { + "epoch": 0.9156333137935699, + "grad_norm": 2.4180087824205256, + "learning_rate": 7.41311643739877e-08, + "loss": 0.898, + "step": 10153 + }, + { + "epoch": 0.9157234973170402, + "grad_norm": 1.5468496455416976, + "learning_rate": 7.39736641205877e-08, + "loss": 1.0129, + "step": 10154 + }, + { + "epoch": 0.9158136808405104, + "grad_norm": 1.6461896711807535, + "learning_rate": 7.381632820637462e-08, + "loss": 0.9959, + "step": 10155 + }, + { + "epoch": 0.9159038643639807, + "grad_norm": 1.5381270920833416, + "learning_rate": 7.365915664477352e-08, + "loss": 0.9603, + "step": 10156 + }, + { + "epoch": 0.915994047887451, + "grad_norm": 10.845377951016305, + "learning_rate": 7.350214944919474e-08, + "loss": 0.978, + "step": 10157 + }, + { + "epoch": 0.9160842314109212, + "grad_norm": 1.7395681353912118, + "learning_rate": 7.334530663303539e-08, + "loss": 0.9984, + "step": 10158 + }, + { + "epoch": 0.9161744149343914, + "grad_norm": 1.5158824112161204, + "learning_rate": 7.318862820967742e-08, + "loss": 0.9801, + "step": 10159 + }, + { + "epoch": 0.9162645984578618, + "grad_norm": 1.854788308107547, + "learning_rate": 7.303211419249056e-08, + "loss": 1.0065, + "step": 10160 + }, + { + "epoch": 0.916354781981332, + "grad_norm": 1.412686347484688, + "learning_rate": 7.287576459482858e-08, + "loss": 0.9903, + "step": 10161 + }, + { + "epoch": 0.9164449655048023, + "grad_norm": 1.5611420643423852, + "learning_rate": 7.271957943003259e-08, + "loss": 1.0123, + "step": 10162 + }, + { + "epoch": 0.9165351490282725, + "grad_norm": 1.9857469756953312, + "learning_rate": 7.256355871142883e-08, + "loss": 0.9727, + "step": 10163 + }, + { + "epoch": 0.9166253325517428, + "grad_norm": 1.5485397782406822, + "learning_rate": 7.240770245233019e-08, + "loss": 1.0251, + "step": 10164 + }, + { + "epoch": 0.9167155160752131, + "grad_norm": 1.7108576507747801, + "learning_rate": 7.225201066603492e-08, + "loss": 0.8889, + "step": 10165 + }, + { + "epoch": 0.9168056995986833, + "grad_norm": 3.15395596672675, + "learning_rate": 7.209648336582774e-08, + "loss": 1.0841, + "step": 10166 + }, + { + "epoch": 0.9168958831221535, + "grad_norm": 1.960335627741261, + "learning_rate": 7.19411205649787e-08, + "loss": 0.9586, + "step": 10167 + }, + { + "epoch": 0.9169860666456239, + "grad_norm": 1.4950747692586575, + "learning_rate": 7.178592227674474e-08, + "loss": 0.9967, + "step": 10168 + }, + { + "epoch": 0.9170762501690941, + "grad_norm": 1.6299645807361416, + "learning_rate": 7.163088851436771e-08, + "loss": 0.8934, + "step": 10169 + }, + { + "epoch": 0.9171664336925643, + "grad_norm": 1.450946528521482, + "learning_rate": 7.147601929107639e-08, + "loss": 0.9406, + "step": 10170 + }, + { + "epoch": 0.9172566172160346, + "grad_norm": 1.6087682376471006, + "learning_rate": 7.132131462008461e-08, + "loss": 1.0183, + "step": 10171 + }, + { + "epoch": 0.9173468007395049, + "grad_norm": 1.7099208511194006, + "learning_rate": 7.116677451459297e-08, + "loss": 0.9808, + "step": 10172 + }, + { + "epoch": 0.9174369842629752, + "grad_norm": 1.446647728358006, + "learning_rate": 7.101239898778799e-08, + "loss": 0.9998, + "step": 10173 + }, + { + "epoch": 0.9175271677864454, + "grad_norm": 1.4394856711943516, + "learning_rate": 7.085818805284094e-08, + "loss": 1.0095, + "step": 10174 + }, + { + "epoch": 0.9176173513099157, + "grad_norm": 2.174482358522795, + "learning_rate": 7.070414172291083e-08, + "loss": 1.053, + "step": 10175 + }, + { + "epoch": 0.917707534833386, + "grad_norm": 1.5336303893117493, + "learning_rate": 7.055026001114095e-08, + "loss": 0.919, + "step": 10176 + }, + { + "epoch": 0.9177977183568562, + "grad_norm": 0.6410469433411621, + "learning_rate": 7.039654293066211e-08, + "loss": 0.8027, + "step": 10177 + }, + { + "epoch": 0.9178879018803264, + "grad_norm": 1.7947937563727006, + "learning_rate": 7.024299049459003e-08, + "loss": 0.8986, + "step": 10178 + }, + { + "epoch": 0.9179780854037968, + "grad_norm": 1.8681743360317686, + "learning_rate": 7.008960271602627e-08, + "loss": 0.8848, + "step": 10179 + }, + { + "epoch": 0.918068268927267, + "grad_norm": 2.5874174095646723, + "learning_rate": 6.993637960805921e-08, + "loss": 0.9384, + "step": 10180 + }, + { + "epoch": 0.9181584524507372, + "grad_norm": 5.748957995662229, + "learning_rate": 6.97833211837624e-08, + "loss": 1.0017, + "step": 10181 + }, + { + "epoch": 0.9182486359742075, + "grad_norm": 2.1340000074506364, + "learning_rate": 6.963042745619562e-08, + "loss": 0.9266, + "step": 10182 + }, + { + "epoch": 0.9183388194976778, + "grad_norm": 1.5163854677170614, + "learning_rate": 6.947769843840511e-08, + "loss": 0.9425, + "step": 10183 + }, + { + "epoch": 0.918429003021148, + "grad_norm": 1.8479409512626852, + "learning_rate": 6.9325134143422e-08, + "loss": 1.004, + "step": 10184 + }, + { + "epoch": 0.9185191865446183, + "grad_norm": 1.8316245207075543, + "learning_rate": 6.917273458426387e-08, + "loss": 0.9522, + "step": 10185 + }, + { + "epoch": 0.9186093700680885, + "grad_norm": 0.6920881195714333, + "learning_rate": 6.902049977393476e-08, + "loss": 0.7911, + "step": 10186 + }, + { + "epoch": 0.9186995535915589, + "grad_norm": 1.9064211534052493, + "learning_rate": 6.886842972542362e-08, + "loss": 0.9839, + "step": 10187 + }, + { + "epoch": 0.9187897371150291, + "grad_norm": 1.4751815393229915, + "learning_rate": 6.871652445170672e-08, + "loss": 0.8644, + "step": 10188 + }, + { + "epoch": 0.9188799206384993, + "grad_norm": 3.4404931648628896, + "learning_rate": 6.856478396574416e-08, + "loss": 1.0581, + "step": 10189 + }, + { + "epoch": 0.9189701041619696, + "grad_norm": 1.8379769543146276, + "learning_rate": 6.841320828048491e-08, + "loss": 1.0368, + "step": 10190 + }, + { + "epoch": 0.9190602876854399, + "grad_norm": 1.5801132066224868, + "learning_rate": 6.826179740886062e-08, + "loss": 0.9714, + "step": 10191 + }, + { + "epoch": 0.9191504712089101, + "grad_norm": 1.7069306158435607, + "learning_rate": 6.811055136379184e-08, + "loss": 1.0368, + "step": 10192 + }, + { + "epoch": 0.9192406547323804, + "grad_norm": 1.9830804041126056, + "learning_rate": 6.79594701581827e-08, + "loss": 0.9511, + "step": 10193 + }, + { + "epoch": 0.9193308382558506, + "grad_norm": 1.7511767109709047, + "learning_rate": 6.780855380492511e-08, + "loss": 0.9787, + "step": 10194 + }, + { + "epoch": 0.919421021779321, + "grad_norm": 1.396379843047382, + "learning_rate": 6.765780231689544e-08, + "loss": 0.9926, + "step": 10195 + }, + { + "epoch": 0.9195112053027912, + "grad_norm": 2.5979981161614356, + "learning_rate": 6.750721570695695e-08, + "loss": 1.031, + "step": 10196 + }, + { + "epoch": 0.9196013888262614, + "grad_norm": 2.2095939948553514, + "learning_rate": 6.735679398795868e-08, + "loss": 0.9568, + "step": 10197 + }, + { + "epoch": 0.9196915723497318, + "grad_norm": 2.0726633220599404, + "learning_rate": 6.720653717273506e-08, + "loss": 0.9077, + "step": 10198 + }, + { + "epoch": 0.919781755873202, + "grad_norm": 1.5604422179686426, + "learning_rate": 6.705644527410714e-08, + "loss": 0.9746, + "step": 10199 + }, + { + "epoch": 0.9198719393966722, + "grad_norm": 1.71102751751994, + "learning_rate": 6.690651830488136e-08, + "loss": 0.9435, + "step": 10200 + }, + { + "epoch": 0.9199621229201425, + "grad_norm": 2.6506695639244584, + "learning_rate": 6.675675627785037e-08, + "loss": 0.9692, + "step": 10201 + }, + { + "epoch": 0.9200523064436128, + "grad_norm": 5.3235536357190885, + "learning_rate": 6.660715920579263e-08, + "loss": 0.9952, + "step": 10202 + }, + { + "epoch": 0.920142489967083, + "grad_norm": 0.5951646466012154, + "learning_rate": 6.645772710147279e-08, + "loss": 0.8147, + "step": 10203 + }, + { + "epoch": 0.9202326734905533, + "grad_norm": 1.4690904425661924, + "learning_rate": 6.630845997764112e-08, + "loss": 0.9457, + "step": 10204 + }, + { + "epoch": 0.9203228570140235, + "grad_norm": 1.5591497744051757, + "learning_rate": 6.615935784703409e-08, + "loss": 0.9183, + "step": 10205 + }, + { + "epoch": 0.9204130405374938, + "grad_norm": 1.6505787036878523, + "learning_rate": 6.601042072237328e-08, + "loss": 0.9035, + "step": 10206 + }, + { + "epoch": 0.9205032240609641, + "grad_norm": 3.109175632545333, + "learning_rate": 6.586164861636767e-08, + "loss": 1.0042, + "step": 10207 + }, + { + "epoch": 0.9205934075844343, + "grad_norm": 1.9594535532024209, + "learning_rate": 6.571304154171065e-08, + "loss": 0.8837, + "step": 10208 + }, + { + "epoch": 0.9206835911079045, + "grad_norm": 1.9260943237610817, + "learning_rate": 6.556459951108273e-08, + "loss": 0.9991, + "step": 10209 + }, + { + "epoch": 0.9207737746313749, + "grad_norm": 17.24387593681359, + "learning_rate": 6.541632253714957e-08, + "loss": 0.9557, + "step": 10210 + }, + { + "epoch": 0.9208639581548451, + "grad_norm": 1.744338960893766, + "learning_rate": 6.526821063256261e-08, + "loss": 0.9, + "step": 10211 + }, + { + "epoch": 0.9209541416783154, + "grad_norm": 2.6582111972347127, + "learning_rate": 6.512026380996016e-08, + "loss": 0.8823, + "step": 10212 + }, + { + "epoch": 0.9210443252017856, + "grad_norm": 1.4676347109882149, + "learning_rate": 6.49724820819657e-08, + "loss": 0.9439, + "step": 10213 + }, + { + "epoch": 0.9211345087252559, + "grad_norm": 1.721627964361496, + "learning_rate": 6.48248654611887e-08, + "loss": 1.0087, + "step": 10214 + }, + { + "epoch": 0.9212246922487262, + "grad_norm": 1.3751808394374923, + "learning_rate": 6.467741396022419e-08, + "loss": 1.0223, + "step": 10215 + }, + { + "epoch": 0.9213148757721964, + "grad_norm": 1.7205940757602474, + "learning_rate": 6.453012759165455e-08, + "loss": 0.9819, + "step": 10216 + }, + { + "epoch": 0.9214050592956666, + "grad_norm": 1.4640612330477025, + "learning_rate": 6.438300636804639e-08, + "loss": 0.9423, + "step": 10217 + }, + { + "epoch": 0.921495242819137, + "grad_norm": 1.4233615356676819, + "learning_rate": 6.423605030195278e-08, + "loss": 1.0079, + "step": 10218 + }, + { + "epoch": 0.9215854263426072, + "grad_norm": 1.9717007881989104, + "learning_rate": 6.408925940591304e-08, + "loss": 1.0463, + "step": 10219 + }, + { + "epoch": 0.9216756098660774, + "grad_norm": 1.4797683709833425, + "learning_rate": 6.394263369245222e-08, + "loss": 0.9085, + "step": 10220 + }, + { + "epoch": 0.9217657933895478, + "grad_norm": 1.6056646148676978, + "learning_rate": 6.379617317408126e-08, + "loss": 0.9788, + "step": 10221 + }, + { + "epoch": 0.921855976913018, + "grad_norm": 1.6949992088158485, + "learning_rate": 6.364987786329723e-08, + "loss": 0.9689, + "step": 10222 + }, + { + "epoch": 0.9219461604364882, + "grad_norm": 1.4912773672120363, + "learning_rate": 6.350374777258193e-08, + "loss": 0.9674, + "step": 10223 + }, + { + "epoch": 0.9220363439599585, + "grad_norm": 2.240989019173815, + "learning_rate": 6.335778291440519e-08, + "loss": 0.9541, + "step": 10224 + }, + { + "epoch": 0.9221265274834288, + "grad_norm": 1.457613667526401, + "learning_rate": 6.321198330122057e-08, + "loss": 0.8809, + "step": 10225 + }, + { + "epoch": 0.9222167110068991, + "grad_norm": 1.5834177981819084, + "learning_rate": 6.306634894546902e-08, + "loss": 0.9107, + "step": 10226 + }, + { + "epoch": 0.9223068945303693, + "grad_norm": 1.31052745457912, + "learning_rate": 6.292087985957661e-08, + "loss": 0.9164, + "step": 10227 + }, + { + "epoch": 0.9223970780538395, + "grad_norm": 4.848795963550392, + "learning_rate": 6.277557605595585e-08, + "loss": 1.0138, + "step": 10228 + }, + { + "epoch": 0.9224872615773099, + "grad_norm": 1.4179209130995472, + "learning_rate": 6.263043754700481e-08, + "loss": 0.8987, + "step": 10229 + }, + { + "epoch": 0.9225774451007801, + "grad_norm": 1.6571618215293205, + "learning_rate": 6.248546434510671e-08, + "loss": 0.9741, + "step": 10230 + }, + { + "epoch": 0.9226676286242503, + "grad_norm": 1.6516243019919572, + "learning_rate": 6.234065646263298e-08, + "loss": 0.9052, + "step": 10231 + }, + { + "epoch": 0.9227578121477206, + "grad_norm": 4.043969248940569, + "learning_rate": 6.219601391193796e-08, + "loss": 0.8843, + "step": 10232 + }, + { + "epoch": 0.9228479956711909, + "grad_norm": 1.4412380213376172, + "learning_rate": 6.205153670536423e-08, + "loss": 0.8964, + "step": 10233 + }, + { + "epoch": 0.9229381791946611, + "grad_norm": 1.5061939624793523, + "learning_rate": 6.190722485523902e-08, + "loss": 1.0062, + "step": 10234 + }, + { + "epoch": 0.9230283627181314, + "grad_norm": 1.5453775266845604, + "learning_rate": 6.176307837387607e-08, + "loss": 0.952, + "step": 10235 + }, + { + "epoch": 0.9231185462416016, + "grad_norm": 4.1550024928189915, + "learning_rate": 6.16190972735744e-08, + "loss": 0.923, + "step": 10236 + }, + { + "epoch": 0.923208729765072, + "grad_norm": 1.892247410273887, + "learning_rate": 6.147528156661974e-08, + "loss": 0.9985, + "step": 10237 + }, + { + "epoch": 0.9232989132885422, + "grad_norm": 1.444082035344491, + "learning_rate": 6.133163126528273e-08, + "loss": 0.8953, + "step": 10238 + }, + { + "epoch": 0.9233890968120124, + "grad_norm": 1.583371421288927, + "learning_rate": 6.11881463818209e-08, + "loss": 0.9792, + "step": 10239 + }, + { + "epoch": 0.9234792803354827, + "grad_norm": 1.519256253107688, + "learning_rate": 6.104482692847668e-08, + "loss": 0.9638, + "step": 10240 + }, + { + "epoch": 0.923569463858953, + "grad_norm": 1.8779037878479223, + "learning_rate": 6.090167291747917e-08, + "loss": 0.9428, + "step": 10241 + }, + { + "epoch": 0.9236596473824232, + "grad_norm": 1.5739722425686506, + "learning_rate": 6.075868436104303e-08, + "loss": 0.8707, + "step": 10242 + }, + { + "epoch": 0.9237498309058935, + "grad_norm": 2.6823463705959685, + "learning_rate": 6.061586127136875e-08, + "loss": 0.9623, + "step": 10243 + }, + { + "epoch": 0.9238400144293638, + "grad_norm": 3.856438388272462, + "learning_rate": 6.047320366064324e-08, + "loss": 0.9434, + "step": 10244 + }, + { + "epoch": 0.923930197952834, + "grad_norm": 1.5272704592239994, + "learning_rate": 6.033071154103786e-08, + "loss": 0.9819, + "step": 10245 + }, + { + "epoch": 0.9240203814763043, + "grad_norm": 1.6531052874884336, + "learning_rate": 6.018838492471178e-08, + "loss": 0.9916, + "step": 10246 + }, + { + "epoch": 0.9241105649997745, + "grad_norm": 0.62792437181848, + "learning_rate": 6.00462238238082e-08, + "loss": 0.8169, + "step": 10247 + }, + { + "epoch": 0.9242007485232449, + "grad_norm": 1.53718134361379, + "learning_rate": 5.990422825045827e-08, + "loss": 0.9348, + "step": 10248 + }, + { + "epoch": 0.9242909320467151, + "grad_norm": 1.5219849533059122, + "learning_rate": 5.976239821677675e-08, + "loss": 0.9509, + "step": 10249 + }, + { + "epoch": 0.9243811155701853, + "grad_norm": 2.0078611850402193, + "learning_rate": 5.962073373486598e-08, + "loss": 0.9144, + "step": 10250 + }, + { + "epoch": 0.9244712990936556, + "grad_norm": 1.4280038727993634, + "learning_rate": 5.947923481681316e-08, + "loss": 0.9367, + "step": 10251 + }, + { + "epoch": 0.9245614826171259, + "grad_norm": 1.605856521444307, + "learning_rate": 5.933790147469198e-08, + "loss": 0.9498, + "step": 10252 + }, + { + "epoch": 0.9246516661405961, + "grad_norm": 4.251893014618744, + "learning_rate": 5.9196733720561665e-08, + "loss": 0.9242, + "step": 10253 + }, + { + "epoch": 0.9247418496640664, + "grad_norm": 1.6688810261353357, + "learning_rate": 5.905573156646793e-08, + "loss": 1.0589, + "step": 10254 + }, + { + "epoch": 0.9248320331875366, + "grad_norm": 2.174542661794398, + "learning_rate": 5.8914895024441134e-08, + "loss": 1.051, + "step": 10255 + }, + { + "epoch": 0.9249222167110069, + "grad_norm": 1.7686768884005166, + "learning_rate": 5.877422410649857e-08, + "loss": 1.0079, + "step": 10256 + }, + { + "epoch": 0.9250124002344772, + "grad_norm": 1.7790224301241302, + "learning_rate": 5.863371882464285e-08, + "loss": 0.9872, + "step": 10257 + }, + { + "epoch": 0.9251025837579474, + "grad_norm": 1.8542026416274628, + "learning_rate": 5.849337919086283e-08, + "loss": 0.9537, + "step": 10258 + }, + { + "epoch": 0.9251927672814176, + "grad_norm": 3.5610754276761845, + "learning_rate": 5.835320521713316e-08, + "loss": 0.9553, + "step": 10259 + }, + { + "epoch": 0.925282950804888, + "grad_norm": 1.7227797691021796, + "learning_rate": 5.8213196915414264e-08, + "loss": 0.9584, + "step": 10260 + }, + { + "epoch": 0.9253731343283582, + "grad_norm": 1.9283284950140698, + "learning_rate": 5.807335429765237e-08, + "loss": 1.0268, + "step": 10261 + }, + { + "epoch": 0.9254633178518284, + "grad_norm": 1.731067769068851, + "learning_rate": 5.7933677375779034e-08, + "loss": 1.0242, + "step": 10262 + }, + { + "epoch": 0.9255535013752987, + "grad_norm": 1.8195494661119538, + "learning_rate": 5.77941661617134e-08, + "loss": 1.0253, + "step": 10263 + }, + { + "epoch": 0.925643684898769, + "grad_norm": 1.5978164406005446, + "learning_rate": 5.765482066735816e-08, + "loss": 0.9061, + "step": 10264 + }, + { + "epoch": 0.9257338684222393, + "grad_norm": 2.034522486905795, + "learning_rate": 5.7515640904604256e-08, + "loss": 0.9732, + "step": 10265 + }, + { + "epoch": 0.9258240519457095, + "grad_norm": 1.9846954087027882, + "learning_rate": 5.7376626885326187e-08, + "loss": 0.9576, + "step": 10266 + }, + { + "epoch": 0.9259142354691798, + "grad_norm": 1.5459595100065169, + "learning_rate": 5.723777862138601e-08, + "loss": 1.0461, + "step": 10267 + }, + { + "epoch": 0.9260044189926501, + "grad_norm": 1.7003104375242355, + "learning_rate": 5.7099096124630705e-08, + "loss": 1.0536, + "step": 10268 + }, + { + "epoch": 0.9260946025161203, + "grad_norm": 1.313177841989802, + "learning_rate": 5.696057940689347e-08, + "loss": 0.9839, + "step": 10269 + }, + { + "epoch": 0.9261847860395905, + "grad_norm": 1.9744905007762483, + "learning_rate": 5.6822228479993736e-08, + "loss": 0.9961, + "step": 10270 + }, + { + "epoch": 0.9262749695630609, + "grad_norm": 1.7123198477107768, + "learning_rate": 5.668404335573584e-08, + "loss": 1.0183, + "step": 10271 + }, + { + "epoch": 0.9263651530865311, + "grad_norm": 1.7049492681344602, + "learning_rate": 5.654602404591058e-08, + "loss": 0.9198, + "step": 10272 + }, + { + "epoch": 0.9264553366100013, + "grad_norm": 1.9718827043434204, + "learning_rate": 5.640817056229474e-08, + "loss": 1.0333, + "step": 10273 + }, + { + "epoch": 0.9265455201334716, + "grad_norm": 1.47419427751324, + "learning_rate": 5.6270482916650706e-08, + "loss": 1.0166, + "step": 10274 + }, + { + "epoch": 0.9266357036569419, + "grad_norm": 1.7690375104499356, + "learning_rate": 5.613296112072663e-08, + "loss": 1.0496, + "step": 10275 + }, + { + "epoch": 0.9267258871804122, + "grad_norm": 2.51938674867238, + "learning_rate": 5.59956051862569e-08, + "loss": 0.976, + "step": 10276 + }, + { + "epoch": 0.9268160707038824, + "grad_norm": 2.316484678333447, + "learning_rate": 5.585841512496081e-08, + "loss": 1.0413, + "step": 10277 + }, + { + "epoch": 0.9269062542273526, + "grad_norm": 2.0214612581709632, + "learning_rate": 5.5721390948545e-08, + "loss": 1.089, + "step": 10278 + }, + { + "epoch": 0.926996437750823, + "grad_norm": 1.428452004822415, + "learning_rate": 5.558453266870056e-08, + "loss": 1.0147, + "step": 10279 + }, + { + "epoch": 0.9270866212742932, + "grad_norm": 1.5230878379774504, + "learning_rate": 5.544784029710525e-08, + "loss": 0.9143, + "step": 10280 + }, + { + "epoch": 0.9271768047977634, + "grad_norm": 1.92416332442284, + "learning_rate": 5.531131384542242e-08, + "loss": 1.0452, + "step": 10281 + }, + { + "epoch": 0.9272669883212337, + "grad_norm": 1.7236698827458403, + "learning_rate": 5.51749533253012e-08, + "loss": 0.9274, + "step": 10282 + }, + { + "epoch": 0.927357171844704, + "grad_norm": 1.6542302177898627, + "learning_rate": 5.503875874837649e-08, + "loss": 0.9167, + "step": 10283 + }, + { + "epoch": 0.9274473553681742, + "grad_norm": 1.7394233980343585, + "learning_rate": 5.4902730126269225e-08, + "loss": 0.9446, + "step": 10284 + }, + { + "epoch": 0.9275375388916445, + "grad_norm": 1.6601967481273645, + "learning_rate": 5.476686747058656e-08, + "loss": 1.0078, + "step": 10285 + }, + { + "epoch": 0.9276277224151147, + "grad_norm": 2.1577062803030738, + "learning_rate": 5.4631170792920124e-08, + "loss": 0.9509, + "step": 10286 + }, + { + "epoch": 0.927717905938585, + "grad_norm": 4.359013586889079, + "learning_rate": 5.449564010484953e-08, + "loss": 0.9569, + "step": 10287 + }, + { + "epoch": 0.9278080894620553, + "grad_norm": 2.105690004970744, + "learning_rate": 5.436027541793775e-08, + "loss": 0.9982, + "step": 10288 + }, + { + "epoch": 0.9278982729855255, + "grad_norm": 1.4870058537196758, + "learning_rate": 5.4225076743735554e-08, + "loss": 0.949, + "step": 10289 + }, + { + "epoch": 0.9279884565089958, + "grad_norm": 1.531626907983798, + "learning_rate": 5.409004409377882e-08, + "loss": 0.894, + "step": 10290 + }, + { + "epoch": 0.9280786400324661, + "grad_norm": 1.4877539474032377, + "learning_rate": 5.3955177479589e-08, + "loss": 0.9006, + "step": 10291 + }, + { + "epoch": 0.9281688235559363, + "grad_norm": 1.9276511378227053, + "learning_rate": 5.3820476912674e-08, + "loss": 0.9842, + "step": 10292 + }, + { + "epoch": 0.9282590070794066, + "grad_norm": 1.3860327084770405, + "learning_rate": 5.3685942404527063e-08, + "loss": 1.0025, + "step": 10293 + }, + { + "epoch": 0.9283491906028769, + "grad_norm": 1.5957988210719403, + "learning_rate": 5.355157396662702e-08, + "loss": 0.958, + "step": 10294 + }, + { + "epoch": 0.9284393741263471, + "grad_norm": 1.4299192813953354, + "learning_rate": 5.34173716104398e-08, + "loss": 0.9628, + "step": 10295 + }, + { + "epoch": 0.9285295576498174, + "grad_norm": 1.6866958256162388, + "learning_rate": 5.328333534741536e-08, + "loss": 0.9844, + "step": 10296 + }, + { + "epoch": 0.9286197411732876, + "grad_norm": 1.7602689406431462, + "learning_rate": 5.314946518899099e-08, + "loss": 0.9472, + "step": 10297 + }, + { + "epoch": 0.928709924696758, + "grad_norm": 1.7172908484587208, + "learning_rate": 5.301576114658912e-08, + "loss": 1.0448, + "step": 10298 + }, + { + "epoch": 0.9288001082202282, + "grad_norm": 1.8821128716669295, + "learning_rate": 5.288222323161795e-08, + "loss": 1.0042, + "step": 10299 + }, + { + "epoch": 0.9288902917436984, + "grad_norm": 1.6724093287456172, + "learning_rate": 5.274885145547214e-08, + "loss": 1.0193, + "step": 10300 + }, + { + "epoch": 0.9289804752671686, + "grad_norm": 1.9100882863316608, + "learning_rate": 5.261564582953082e-08, + "loss": 0.9671, + "step": 10301 + }, + { + "epoch": 0.929070658790639, + "grad_norm": 2.244311559423722, + "learning_rate": 5.248260636516066e-08, + "loss": 0.9482, + "step": 10302 + }, + { + "epoch": 0.9291608423141092, + "grad_norm": 1.9131438873066549, + "learning_rate": 5.2349733073712824e-08, + "loss": 0.9206, + "step": 10303 + }, + { + "epoch": 0.9292510258375795, + "grad_norm": 2.3686990979042384, + "learning_rate": 5.221702596652533e-08, + "loss": 1.0666, + "step": 10304 + }, + { + "epoch": 0.9293412093610497, + "grad_norm": 2.1935345927063503, + "learning_rate": 5.208448505492091e-08, + "loss": 0.9164, + "step": 10305 + }, + { + "epoch": 0.92943139288452, + "grad_norm": 1.4229471545027406, + "learning_rate": 5.1952110350208965e-08, + "loss": 0.957, + "step": 10306 + }, + { + "epoch": 0.9295215764079903, + "grad_norm": 1.533373936170839, + "learning_rate": 5.181990186368446e-08, + "loss": 0.9308, + "step": 10307 + }, + { + "epoch": 0.9296117599314605, + "grad_norm": 1.5405236294032083, + "learning_rate": 5.1687859606627915e-08, + "loss": 1.0745, + "step": 10308 + }, + { + "epoch": 0.9297019434549307, + "grad_norm": 2.6879410492699773, + "learning_rate": 5.1555983590306327e-08, + "loss": 0.8916, + "step": 10309 + }, + { + "epoch": 0.9297921269784011, + "grad_norm": 0.6558875490808902, + "learning_rate": 5.1424273825971806e-08, + "loss": 0.8151, + "step": 10310 + }, + { + "epoch": 0.9298823105018713, + "grad_norm": 2.764363318980733, + "learning_rate": 5.1292730324862475e-08, + "loss": 0.9379, + "step": 10311 + }, + { + "epoch": 0.9299724940253415, + "grad_norm": 1.5036231629112031, + "learning_rate": 5.116135309820224e-08, + "loss": 0.9907, + "step": 10312 + }, + { + "epoch": 0.9300626775488118, + "grad_norm": 2.018209648448127, + "learning_rate": 5.103014215720147e-08, + "loss": 1.0131, + "step": 10313 + }, + { + "epoch": 0.9301528610722821, + "grad_norm": 3.40705106222563, + "learning_rate": 5.0899097513055214e-08, + "loss": 1.0334, + "step": 10314 + }, + { + "epoch": 0.9302430445957524, + "grad_norm": 1.6249172483061325, + "learning_rate": 5.076821917694563e-08, + "loss": 0.9629, + "step": 10315 + }, + { + "epoch": 0.9303332281192226, + "grad_norm": 5.803325015695138, + "learning_rate": 5.063750716003889e-08, + "loss": 1.0256, + "step": 10316 + }, + { + "epoch": 0.9304234116426929, + "grad_norm": 1.6607071265492628, + "learning_rate": 5.050696147348921e-08, + "loss": 0.9554, + "step": 10317 + }, + { + "epoch": 0.9305135951661632, + "grad_norm": 2.749468058376824, + "learning_rate": 5.037658212843454e-08, + "loss": 0.9461, + "step": 10318 + }, + { + "epoch": 0.9306037786896334, + "grad_norm": 1.454056472027973, + "learning_rate": 5.0246369136000444e-08, + "loss": 0.8695, + "step": 10319 + }, + { + "epoch": 0.9306939622131036, + "grad_norm": 1.537348145840084, + "learning_rate": 5.011632250729691e-08, + "loss": 1.05, + "step": 10320 + }, + { + "epoch": 0.930784145736574, + "grad_norm": 1.6146107661059959, + "learning_rate": 4.998644225342019e-08, + "loss": 1.0645, + "step": 10321 + }, + { + "epoch": 0.9308743292600442, + "grad_norm": 1.69462332363261, + "learning_rate": 4.9856728385452296e-08, + "loss": 0.9764, + "step": 10322 + }, + { + "epoch": 0.9309645127835144, + "grad_norm": 1.635680896123159, + "learning_rate": 4.9727180914461485e-08, + "loss": 1.06, + "step": 10323 + }, + { + "epoch": 0.9310546963069847, + "grad_norm": 1.778352287796861, + "learning_rate": 4.959779985150137e-08, + "loss": 0.9855, + "step": 10324 + }, + { + "epoch": 0.931144879830455, + "grad_norm": 2.026760833376254, + "learning_rate": 4.9468585207611105e-08, + "loss": 0.9206, + "step": 10325 + }, + { + "epoch": 0.9312350633539253, + "grad_norm": 1.5169524838745378, + "learning_rate": 4.9339536993816764e-08, + "loss": 0.9526, + "step": 10326 + }, + { + "epoch": 0.9313252468773955, + "grad_norm": 1.8243295117229068, + "learning_rate": 4.921065522112844e-08, + "loss": 0.9232, + "step": 10327 + }, + { + "epoch": 0.9314154304008657, + "grad_norm": 1.590935709346677, + "learning_rate": 4.908193990054377e-08, + "loss": 1.021, + "step": 10328 + }, + { + "epoch": 0.9315056139243361, + "grad_norm": 1.8458957793592317, + "learning_rate": 4.89533910430453e-08, + "loss": 0.9903, + "step": 10329 + }, + { + "epoch": 0.9315957974478063, + "grad_norm": 1.5264426302432963, + "learning_rate": 4.8825008659601376e-08, + "loss": 0.9625, + "step": 10330 + }, + { + "epoch": 0.9316859809712765, + "grad_norm": 2.078267372245046, + "learning_rate": 4.869679276116634e-08, + "loss": 0.9906, + "step": 10331 + }, + { + "epoch": 0.9317761644947468, + "grad_norm": 1.4845466916022274, + "learning_rate": 4.856874335868055e-08, + "loss": 0.9936, + "step": 10332 + }, + { + "epoch": 0.9318663480182171, + "grad_norm": 1.9104100470253156, + "learning_rate": 4.844086046306928e-08, + "loss": 1.0824, + "step": 10333 + }, + { + "epoch": 0.9319565315416873, + "grad_norm": 1.6220376816969224, + "learning_rate": 4.8313144085244896e-08, + "loss": 0.9718, + "step": 10334 + }, + { + "epoch": 0.9320467150651576, + "grad_norm": 0.651975562073539, + "learning_rate": 4.818559423610424e-08, + "loss": 0.8121, + "step": 10335 + }, + { + "epoch": 0.9321368985886278, + "grad_norm": 1.3552020687317174, + "learning_rate": 4.8058210926531284e-08, + "loss": 0.9175, + "step": 10336 + }, + { + "epoch": 0.9322270821120981, + "grad_norm": 1.437867490814984, + "learning_rate": 4.7930994167394435e-08, + "loss": 1.0099, + "step": 10337 + }, + { + "epoch": 0.9323172656355684, + "grad_norm": 1.3191524578509257, + "learning_rate": 4.7803943969548786e-08, + "loss": 1.0172, + "step": 10338 + }, + { + "epoch": 0.9324074491590386, + "grad_norm": 1.4565126861585505, + "learning_rate": 4.7677060343834784e-08, + "loss": 0.9221, + "step": 10339 + }, + { + "epoch": 0.932497632682509, + "grad_norm": 1.5741346161653966, + "learning_rate": 4.75503433010791e-08, + "loss": 0.9689, + "step": 10340 + }, + { + "epoch": 0.9325878162059792, + "grad_norm": 1.8839771894665545, + "learning_rate": 4.742379285209419e-08, + "loss": 0.956, + "step": 10341 + }, + { + "epoch": 0.9326779997294494, + "grad_norm": 1.5125385594772913, + "learning_rate": 4.72974090076772e-08, + "loss": 1.0752, + "step": 10342 + }, + { + "epoch": 0.9327681832529197, + "grad_norm": 2.1734254640620305, + "learning_rate": 4.717119177861262e-08, + "loss": 0.9318, + "step": 10343 + }, + { + "epoch": 0.93285836677639, + "grad_norm": 1.8338468294900645, + "learning_rate": 4.70451411756696e-08, + "loss": 0.9899, + "step": 10344 + }, + { + "epoch": 0.9329485502998602, + "grad_norm": 1.4524763566947319, + "learning_rate": 4.691925720960355e-08, + "loss": 1.0173, + "step": 10345 + }, + { + "epoch": 0.9330387338233305, + "grad_norm": 1.7983498532889028, + "learning_rate": 4.6793539891155645e-08, + "loss": 0.9999, + "step": 10346 + }, + { + "epoch": 0.9331289173468007, + "grad_norm": 2.182384722659539, + "learning_rate": 4.6667989231052864e-08, + "loss": 0.9575, + "step": 10347 + }, + { + "epoch": 0.933219100870271, + "grad_norm": 1.9498782217685218, + "learning_rate": 4.654260524000797e-08, + "loss": 0.9863, + "step": 10348 + }, + { + "epoch": 0.9333092843937413, + "grad_norm": 1.7220671115606239, + "learning_rate": 4.6417387928719076e-08, + "loss": 0.966, + "step": 10349 + }, + { + "epoch": 0.9333994679172115, + "grad_norm": 1.4504259536184299, + "learning_rate": 4.629233730787052e-08, + "loss": 0.8507, + "step": 10350 + }, + { + "epoch": 0.9334896514406817, + "grad_norm": 1.7397119013927025, + "learning_rate": 4.616745338813266e-08, + "loss": 0.9376, + "step": 10351 + }, + { + "epoch": 0.9335798349641521, + "grad_norm": 1.801737566578324, + "learning_rate": 4.6042736180160744e-08, + "loss": 0.9923, + "step": 10352 + }, + { + "epoch": 0.9336700184876223, + "grad_norm": 1.9051886558733189, + "learning_rate": 4.591818569459671e-08, + "loss": 1.0959, + "step": 10353 + }, + { + "epoch": 0.9337602020110926, + "grad_norm": 1.916467766876618, + "learning_rate": 4.5793801942067614e-08, + "loss": 0.997, + "step": 10354 + }, + { + "epoch": 0.9338503855345628, + "grad_norm": 1.6599862763400572, + "learning_rate": 4.566958493318673e-08, + "loss": 0.8448, + "step": 10355 + }, + { + "epoch": 0.9339405690580331, + "grad_norm": 0.6358143052128287, + "learning_rate": 4.554553467855316e-08, + "loss": 0.8364, + "step": 10356 + }, + { + "epoch": 0.9340307525815034, + "grad_norm": 3.917100895828519, + "learning_rate": 4.5421651188751074e-08, + "loss": 0.9946, + "step": 10357 + }, + { + "epoch": 0.9341209361049736, + "grad_norm": 1.6401673904917136, + "learning_rate": 4.529793447435137e-08, + "loss": 0.9882, + "step": 10358 + }, + { + "epoch": 0.9342111196284438, + "grad_norm": 2.0721323903475475, + "learning_rate": 4.5174384545909824e-08, + "loss": 0.978, + "step": 10359 + }, + { + "epoch": 0.9343013031519142, + "grad_norm": 2.30412601161453, + "learning_rate": 4.505100141396867e-08, + "loss": 1.0202, + "step": 10360 + }, + { + "epoch": 0.9343914866753844, + "grad_norm": 2.0852015193457296, + "learning_rate": 4.492778508905548e-08, + "loss": 1.0091, + "step": 10361 + }, + { + "epoch": 0.9344816701988546, + "grad_norm": 1.6175258116748326, + "learning_rate": 4.480473558168385e-08, + "loss": 0.9139, + "step": 10362 + }, + { + "epoch": 0.934571853722325, + "grad_norm": 1.790249093694608, + "learning_rate": 4.4681852902352936e-08, + "loss": 0.8835, + "step": 10363 + }, + { + "epoch": 0.9346620372457952, + "grad_norm": 1.7688207821112107, + "learning_rate": 4.455913706154812e-08, + "loss": 0.8898, + "step": 10364 + }, + { + "epoch": 0.9347522207692655, + "grad_norm": 4.63723805965355, + "learning_rate": 4.443658806973949e-08, + "loss": 1.0048, + "step": 10365 + }, + { + "epoch": 0.9348424042927357, + "grad_norm": 0.6670555430939216, + "learning_rate": 4.431420593738444e-08, + "loss": 0.8386, + "step": 10366 + }, + { + "epoch": 0.934932587816206, + "grad_norm": 1.6722693290214978, + "learning_rate": 4.419199067492485e-08, + "loss": 0.8897, + "step": 10367 + }, + { + "epoch": 0.9350227713396763, + "grad_norm": 16.17321005207908, + "learning_rate": 4.4069942292788596e-08, + "loss": 0.9865, + "step": 10368 + }, + { + "epoch": 0.9351129548631465, + "grad_norm": 1.6109163625164764, + "learning_rate": 4.39480608013898e-08, + "loss": 0.9809, + "step": 10369 + }, + { + "epoch": 0.9352031383866167, + "grad_norm": 1.516975509023401, + "learning_rate": 4.3826346211128126e-08, + "loss": 1.0655, + "step": 10370 + }, + { + "epoch": 0.9352933219100871, + "grad_norm": 1.473307822094354, + "learning_rate": 4.370479853238884e-08, + "loss": 0.9791, + "step": 10371 + }, + { + "epoch": 0.9353835054335573, + "grad_norm": 1.8598650759294963, + "learning_rate": 4.3583417775542756e-08, + "loss": 0.9397, + "step": 10372 + }, + { + "epoch": 0.9354736889570275, + "grad_norm": 1.5119444212916529, + "learning_rate": 4.3462203950947575e-08, + "loss": 0.9326, + "step": 10373 + }, + { + "epoch": 0.9355638724804978, + "grad_norm": 1.6194125730512876, + "learning_rate": 4.3341157068944814e-08, + "loss": 0.9614, + "step": 10374 + }, + { + "epoch": 0.9356540560039681, + "grad_norm": 1.660024552450954, + "learning_rate": 4.322027713986376e-08, + "loss": 0.9895, + "step": 10375 + }, + { + "epoch": 0.9357442395274383, + "grad_norm": 1.7400802389914847, + "learning_rate": 4.309956417401816e-08, + "loss": 1.044, + "step": 10376 + }, + { + "epoch": 0.9358344230509086, + "grad_norm": 1.4151329077856956, + "learning_rate": 4.297901818170801e-08, + "loss": 0.9479, + "step": 10377 + }, + { + "epoch": 0.9359246065743788, + "grad_norm": 1.4131621133957641, + "learning_rate": 4.285863917321886e-08, + "loss": 0.9088, + "step": 10378 + }, + { + "epoch": 0.9360147900978492, + "grad_norm": 1.832003725410405, + "learning_rate": 4.2738427158822253e-08, + "loss": 0.9098, + "step": 10379 + }, + { + "epoch": 0.9361049736213194, + "grad_norm": 2.2138516796159537, + "learning_rate": 4.261838214877511e-08, + "loss": 0.9815, + "step": 10380 + }, + { + "epoch": 0.9361951571447896, + "grad_norm": 1.468311406656338, + "learning_rate": 4.249850415332079e-08, + "loss": 0.9827, + "step": 10381 + }, + { + "epoch": 0.9362853406682599, + "grad_norm": 1.579486976823435, + "learning_rate": 4.237879318268756e-08, + "loss": 1.0346, + "step": 10382 + }, + { + "epoch": 0.9363755241917302, + "grad_norm": 2.1774764246053078, + "learning_rate": 4.225924924708968e-08, + "loss": 0.9097, + "step": 10383 + }, + { + "epoch": 0.9364657077152004, + "grad_norm": 2.285862687947247, + "learning_rate": 4.2139872356727665e-08, + "loss": 0.9848, + "step": 10384 + }, + { + "epoch": 0.9365558912386707, + "grad_norm": 1.8323164272605221, + "learning_rate": 4.202066252178738e-08, + "loss": 0.942, + "step": 10385 + }, + { + "epoch": 0.936646074762141, + "grad_norm": 1.4765066307841328, + "learning_rate": 4.1901619752440445e-08, + "loss": 1.0662, + "step": 10386 + }, + { + "epoch": 0.9367362582856112, + "grad_norm": 1.5325977855556359, + "learning_rate": 4.178274405884363e-08, + "loss": 0.9765, + "step": 10387 + }, + { + "epoch": 0.9368264418090815, + "grad_norm": 1.5453387251461737, + "learning_rate": 4.166403545114105e-08, + "loss": 0.9895, + "step": 10388 + }, + { + "epoch": 0.9369166253325517, + "grad_norm": 2.2386333727849634, + "learning_rate": 4.154549393946083e-08, + "loss": 1.0333, + "step": 10389 + }, + { + "epoch": 0.937006808856022, + "grad_norm": 1.5526815537399221, + "learning_rate": 4.14271195339182e-08, + "loss": 1.0048, + "step": 10390 + }, + { + "epoch": 0.9370969923794923, + "grad_norm": 1.388610838502377, + "learning_rate": 4.1308912244613084e-08, + "loss": 0.9439, + "step": 10391 + }, + { + "epoch": 0.9371871759029625, + "grad_norm": 1.4620839073170604, + "learning_rate": 4.1190872081631636e-08, + "loss": 0.9347, + "step": 10392 + }, + { + "epoch": 0.9372773594264328, + "grad_norm": 1.8150441959349148, + "learning_rate": 4.107299905504558e-08, + "loss": 1.0418, + "step": 10393 + }, + { + "epoch": 0.9373675429499031, + "grad_norm": 1.5562199080288703, + "learning_rate": 4.095529317491286e-08, + "loss": 0.9687, + "step": 10394 + }, + { + "epoch": 0.9374577264733733, + "grad_norm": 1.5647952197467985, + "learning_rate": 4.0837754451276575e-08, + "loss": 0.9832, + "step": 10395 + }, + { + "epoch": 0.9375479099968436, + "grad_norm": 1.8204582057808278, + "learning_rate": 4.072038289416557e-08, + "loss": 0.9633, + "step": 10396 + }, + { + "epoch": 0.9376380935203138, + "grad_norm": 1.4548915666636126, + "learning_rate": 4.0603178513595185e-08, + "loss": 0.9813, + "step": 10397 + }, + { + "epoch": 0.9377282770437841, + "grad_norm": 1.7150726513168602, + "learning_rate": 4.0486141319565624e-08, + "loss": 0.9831, + "step": 10398 + }, + { + "epoch": 0.9378184605672544, + "grad_norm": 1.6076777869787824, + "learning_rate": 4.0369271322062916e-08, + "loss": 0.9189, + "step": 10399 + }, + { + "epoch": 0.9379086440907246, + "grad_norm": 1.7444679833030148, + "learning_rate": 4.0252568531059295e-08, + "loss": 0.999, + "step": 10400 + }, + { + "epoch": 0.9379988276141948, + "grad_norm": 2.011095502288176, + "learning_rate": 4.013603295651235e-08, + "loss": 0.9776, + "step": 10401 + }, + { + "epoch": 0.9380890111376652, + "grad_norm": 1.5502626719480548, + "learning_rate": 4.001966460836592e-08, + "loss": 1.0037, + "step": 10402 + }, + { + "epoch": 0.9381791946611354, + "grad_norm": 1.5464183248861942, + "learning_rate": 3.990346349654894e-08, + "loss": 1.0533, + "step": 10403 + }, + { + "epoch": 0.9382693781846057, + "grad_norm": 1.469757221700566, + "learning_rate": 3.9787429630975924e-08, + "loss": 0.9526, + "step": 10404 + }, + { + "epoch": 0.9383595617080759, + "grad_norm": 1.632916898903421, + "learning_rate": 3.967156302154828e-08, + "loss": 0.9671, + "step": 10405 + }, + { + "epoch": 0.9384497452315462, + "grad_norm": 2.912094310616636, + "learning_rate": 3.955586367815189e-08, + "loss": 0.9336, + "step": 10406 + }, + { + "epoch": 0.9385399287550165, + "grad_norm": 1.6713027687843662, + "learning_rate": 3.944033161065907e-08, + "loss": 1.0467, + "step": 10407 + }, + { + "epoch": 0.9386301122784867, + "grad_norm": 1.5626906469385342, + "learning_rate": 3.93249668289275e-08, + "loss": 0.8654, + "step": 10408 + }, + { + "epoch": 0.9387202958019569, + "grad_norm": 1.5916371271211733, + "learning_rate": 3.920976934280063e-08, + "loss": 0.9037, + "step": 10409 + }, + { + "epoch": 0.9388104793254273, + "grad_norm": 1.7315167080466625, + "learning_rate": 3.909473916210815e-08, + "loss": 1.0577, + "step": 10410 + }, + { + "epoch": 0.9389006628488975, + "grad_norm": 2.6286669051115177, + "learning_rate": 3.897987629666488e-08, + "loss": 0.9412, + "step": 10411 + }, + { + "epoch": 0.9389908463723677, + "grad_norm": 1.5392283238961355, + "learning_rate": 3.886518075627143e-08, + "loss": 1.0501, + "step": 10412 + }, + { + "epoch": 0.9390810298958381, + "grad_norm": 2.5060379547447678, + "learning_rate": 3.875065255071419e-08, + "loss": 0.9736, + "step": 10413 + }, + { + "epoch": 0.9391712134193083, + "grad_norm": 1.5972741003411042, + "learning_rate": 3.863629168976579e-08, + "loss": 0.9464, + "step": 10414 + }, + { + "epoch": 0.9392613969427785, + "grad_norm": 1.4560326639169001, + "learning_rate": 3.852209818318375e-08, + "loss": 1.0342, + "step": 10415 + }, + { + "epoch": 0.9393515804662488, + "grad_norm": 1.6532475590118936, + "learning_rate": 3.840807204071161e-08, + "loss": 0.8828, + "step": 10416 + }, + { + "epoch": 0.9394417639897191, + "grad_norm": 1.5661990820667557, + "learning_rate": 3.829421327207894e-08, + "loss": 0.9637, + "step": 10417 + }, + { + "epoch": 0.9395319475131894, + "grad_norm": 2.20681216072409, + "learning_rate": 3.8180521887000825e-08, + "loss": 0.9348, + "step": 10418 + }, + { + "epoch": 0.9396221310366596, + "grad_norm": 0.5933305614072721, + "learning_rate": 3.806699789517775e-08, + "loss": 0.8474, + "step": 10419 + }, + { + "epoch": 0.9397123145601298, + "grad_norm": 1.5878180267892474, + "learning_rate": 3.7953641306296635e-08, + "loss": 0.9606, + "step": 10420 + }, + { + "epoch": 0.9398024980836002, + "grad_norm": 3.4492661677003253, + "learning_rate": 3.784045213002951e-08, + "loss": 0.824, + "step": 10421 + }, + { + "epoch": 0.9398926816070704, + "grad_norm": 1.6457966265226738, + "learning_rate": 3.7727430376033986e-08, + "loss": 0.9919, + "step": 10422 + }, + { + "epoch": 0.9399828651305406, + "grad_norm": 1.6262203182763488, + "learning_rate": 3.7614576053954126e-08, + "loss": 0.9826, + "step": 10423 + }, + { + "epoch": 0.9400730486540109, + "grad_norm": 1.672993864920358, + "learning_rate": 3.75018891734189e-08, + "loss": 0.9731, + "step": 10424 + }, + { + "epoch": 0.9401632321774812, + "grad_norm": 2.0063030342864367, + "learning_rate": 3.738936974404372e-08, + "loss": 1.0224, + "step": 10425 + }, + { + "epoch": 0.9402534157009514, + "grad_norm": 1.4961041084110243, + "learning_rate": 3.7277017775429354e-08, + "loss": 1.0425, + "step": 10426 + }, + { + "epoch": 0.9403435992244217, + "grad_norm": 1.4150650095749107, + "learning_rate": 3.7164833277162136e-08, + "loss": 1.0029, + "step": 10427 + }, + { + "epoch": 0.9404337827478919, + "grad_norm": 1.3186615774151598, + "learning_rate": 3.705281625881418e-08, + "loss": 0.9588, + "step": 10428 + }, + { + "epoch": 0.9405239662713623, + "grad_norm": 1.866150252684516, + "learning_rate": 3.694096672994362e-08, + "loss": 0.8839, + "step": 10429 + }, + { + "epoch": 0.9406141497948325, + "grad_norm": 1.5170644629706247, + "learning_rate": 3.682928470009394e-08, + "loss": 1.0297, + "step": 10430 + }, + { + "epoch": 0.9407043333183027, + "grad_norm": 1.5890853535106877, + "learning_rate": 3.6717770178794406e-08, + "loss": 0.9366, + "step": 10431 + }, + { + "epoch": 0.940794516841773, + "grad_norm": 1.7997176531706391, + "learning_rate": 3.6606423175560287e-08, + "loss": 1.0856, + "step": 10432 + }, + { + "epoch": 0.9408847003652433, + "grad_norm": 0.6322147230395577, + "learning_rate": 3.649524369989221e-08, + "loss": 0.7829, + "step": 10433 + }, + { + "epoch": 0.9409748838887135, + "grad_norm": 2.7154758074375858, + "learning_rate": 3.638423176127636e-08, + "loss": 1.0104, + "step": 10434 + }, + { + "epoch": 0.9410650674121838, + "grad_norm": 2.167112109049386, + "learning_rate": 3.6273387369185396e-08, + "loss": 1.0124, + "step": 10435 + }, + { + "epoch": 0.9411552509356541, + "grad_norm": 1.494635127396396, + "learning_rate": 3.616271053307685e-08, + "loss": 0.9244, + "step": 10436 + }, + { + "epoch": 0.9412454344591243, + "grad_norm": 1.706142097945883, + "learning_rate": 3.6052201262394275e-08, + "loss": 1.0031, + "step": 10437 + }, + { + "epoch": 0.9413356179825946, + "grad_norm": 1.5698237631220178, + "learning_rate": 3.5941859566566816e-08, + "loss": 1.0317, + "step": 10438 + }, + { + "epoch": 0.9414258015060648, + "grad_norm": 0.7289814214183599, + "learning_rate": 3.583168545500981e-08, + "loss": 0.8046, + "step": 10439 + }, + { + "epoch": 0.9415159850295352, + "grad_norm": 1.5956094403821295, + "learning_rate": 3.5721678937123746e-08, + "loss": 0.9225, + "step": 10440 + }, + { + "epoch": 0.9416061685530054, + "grad_norm": 1.5142046378453002, + "learning_rate": 3.561184002229467e-08, + "loss": 1.0599, + "step": 10441 + }, + { + "epoch": 0.9416963520764756, + "grad_norm": 1.78152142514673, + "learning_rate": 3.550216871989531e-08, + "loss": 0.9284, + "step": 10442 + }, + { + "epoch": 0.9417865355999459, + "grad_norm": 0.6375719216859111, + "learning_rate": 3.539266503928262e-08, + "loss": 0.8837, + "step": 10443 + }, + { + "epoch": 0.9418767191234162, + "grad_norm": 1.4718671956432474, + "learning_rate": 3.528332898980091e-08, + "loss": 0.9892, + "step": 10444 + }, + { + "epoch": 0.9419669026468864, + "grad_norm": 1.4991796952607952, + "learning_rate": 3.517416058077849e-08, + "loss": 0.9441, + "step": 10445 + }, + { + "epoch": 0.9420570861703567, + "grad_norm": 1.7806526654853545, + "learning_rate": 3.506515982153102e-08, + "loss": 0.8834, + "step": 10446 + }, + { + "epoch": 0.9421472696938269, + "grad_norm": 1.6584146137861553, + "learning_rate": 3.495632672135862e-08, + "loss": 0.9837, + "step": 10447 + }, + { + "epoch": 0.9422374532172972, + "grad_norm": 1.839190000458921, + "learning_rate": 3.4847661289547417e-08, + "loss": 1.0108, + "step": 10448 + }, + { + "epoch": 0.9423276367407675, + "grad_norm": 1.7686065211249087, + "learning_rate": 3.473916353536932e-08, + "loss": 0.9351, + "step": 10449 + }, + { + "epoch": 0.9424178202642377, + "grad_norm": 1.5151926584148492, + "learning_rate": 3.463083346808249e-08, + "loss": 0.9459, + "step": 10450 + }, + { + "epoch": 0.9425080037877079, + "grad_norm": 3.1073692408431266, + "learning_rate": 3.452267109692975e-08, + "loss": 0.8757, + "step": 10451 + }, + { + "epoch": 0.9425981873111783, + "grad_norm": 1.462587455025727, + "learning_rate": 3.441467643114016e-08, + "loss": 0.9795, + "step": 10452 + }, + { + "epoch": 0.9426883708346485, + "grad_norm": 0.5893823448443551, + "learning_rate": 3.430684947992857e-08, + "loss": 0.795, + "step": 10453 + }, + { + "epoch": 0.9427785543581187, + "grad_norm": 1.6403993723898123, + "learning_rate": 3.419919025249518e-08, + "loss": 0.9937, + "step": 10454 + }, + { + "epoch": 0.942868737881589, + "grad_norm": 1.7388020967402869, + "learning_rate": 3.40916987580262e-08, + "loss": 0.9994, + "step": 10455 + }, + { + "epoch": 0.9429589214050593, + "grad_norm": 1.5740254096484962, + "learning_rate": 3.398437500569362e-08, + "loss": 0.9528, + "step": 10456 + }, + { + "epoch": 0.9430491049285296, + "grad_norm": 1.6741698015385251, + "learning_rate": 3.3877219004654347e-08, + "loss": 0.8605, + "step": 10457 + }, + { + "epoch": 0.9431392884519998, + "grad_norm": 1.8826859301053445, + "learning_rate": 3.3770230764051946e-08, + "loss": 1.0128, + "step": 10458 + }, + { + "epoch": 0.9432294719754701, + "grad_norm": 2.043396418138531, + "learning_rate": 3.366341029301534e-08, + "loss": 0.9977, + "step": 10459 + }, + { + "epoch": 0.9433196554989404, + "grad_norm": 2.1202063627391, + "learning_rate": 3.355675760065857e-08, + "loss": 0.986, + "step": 10460 + }, + { + "epoch": 0.9434098390224106, + "grad_norm": 0.5762111585770642, + "learning_rate": 3.345027269608236e-08, + "loss": 0.7628, + "step": 10461 + }, + { + "epoch": 0.9435000225458808, + "grad_norm": 0.6582400257848634, + "learning_rate": 3.334395558837211e-08, + "loss": 0.8299, + "step": 10462 + }, + { + "epoch": 0.9435902060693512, + "grad_norm": 1.4146184513337876, + "learning_rate": 3.3237806286599667e-08, + "loss": 0.8871, + "step": 10463 + }, + { + "epoch": 0.9436803895928214, + "grad_norm": 1.4732396920395558, + "learning_rate": 3.313182479982224e-08, + "loss": 0.9672, + "step": 10464 + }, + { + "epoch": 0.9437705731162916, + "grad_norm": 1.5939527644406355, + "learning_rate": 3.302601113708259e-08, + "loss": 1.0453, + "step": 10465 + }, + { + "epoch": 0.9438607566397619, + "grad_norm": 1.4265610326151492, + "learning_rate": 3.292036530740972e-08, + "loss": 1.0254, + "step": 10466 + }, + { + "epoch": 0.9439509401632322, + "grad_norm": 5.202016901466166, + "learning_rate": 3.2814887319817294e-08, + "loss": 0.9721, + "step": 10467 + }, + { + "epoch": 0.9440411236867025, + "grad_norm": 1.6563527777110034, + "learning_rate": 3.270957718330591e-08, + "loss": 1.0204, + "step": 10468 + }, + { + "epoch": 0.9441313072101727, + "grad_norm": 1.3498515948211036, + "learning_rate": 3.260443490686082e-08, + "loss": 1.0243, + "step": 10469 + }, + { + "epoch": 0.9442214907336429, + "grad_norm": 2.2394086784451512, + "learning_rate": 3.249946049945351e-08, + "loss": 0.992, + "step": 10470 + }, + { + "epoch": 0.9443116742571133, + "grad_norm": 1.631054018020721, + "learning_rate": 3.239465397004082e-08, + "loss": 0.9941, + "step": 10471 + }, + { + "epoch": 0.9444018577805835, + "grad_norm": 1.624560736927653, + "learning_rate": 3.229001532756559e-08, + "loss": 0.9394, + "step": 10472 + }, + { + "epoch": 0.9444920413040537, + "grad_norm": 2.8003971363172315, + "learning_rate": 3.218554458095602e-08, + "loss": 0.9763, + "step": 10473 + }, + { + "epoch": 0.944582224827524, + "grad_norm": 2.37337689400819, + "learning_rate": 3.20812417391263e-08, + "loss": 0.9829, + "step": 10474 + }, + { + "epoch": 0.9446724083509943, + "grad_norm": 1.320940489087627, + "learning_rate": 3.1977106810975764e-08, + "loss": 0.9695, + "step": 10475 + }, + { + "epoch": 0.9447625918744645, + "grad_norm": 2.341386544339103, + "learning_rate": 3.187313980539042e-08, + "loss": 0.8593, + "step": 10476 + }, + { + "epoch": 0.9448527753979348, + "grad_norm": 1.4354590149027509, + "learning_rate": 3.176934073124071e-08, + "loss": 1.0599, + "step": 10477 + }, + { + "epoch": 0.944942958921405, + "grad_norm": 1.8894586669891071, + "learning_rate": 3.166570959738357e-08, + "loss": 0.9767, + "step": 10478 + }, + { + "epoch": 0.9450331424448754, + "grad_norm": 1.5458237129316499, + "learning_rate": 3.1562246412661476e-08, + "loss": 0.9136, + "step": 10479 + }, + { + "epoch": 0.9451233259683456, + "grad_norm": 1.5392168616326938, + "learning_rate": 3.145895118590225e-08, + "loss": 1.0114, + "step": 10480 + }, + { + "epoch": 0.9452135094918158, + "grad_norm": 1.5071354269878943, + "learning_rate": 3.135582392591996e-08, + "loss": 0.8864, + "step": 10481 + }, + { + "epoch": 0.9453036930152862, + "grad_norm": 0.5947558766924964, + "learning_rate": 3.125286464151333e-08, + "loss": 0.7732, + "step": 10482 + }, + { + "epoch": 0.9453938765387564, + "grad_norm": 2.1072502455236592, + "learning_rate": 3.115007334146824e-08, + "loss": 0.95, + "step": 10483 + }, + { + "epoch": 0.9454840600622266, + "grad_norm": 0.6304482930585897, + "learning_rate": 3.104745003455478e-08, + "loss": 0.7946, + "step": 10484 + }, + { + "epoch": 0.9455742435856969, + "grad_norm": 1.6404288038380166, + "learning_rate": 3.094499472952972e-08, + "loss": 1.0011, + "step": 10485 + }, + { + "epoch": 0.9456644271091672, + "grad_norm": 2.024581409094127, + "learning_rate": 3.084270743513495e-08, + "loss": 0.9437, + "step": 10486 + }, + { + "epoch": 0.9457546106326374, + "grad_norm": 1.5529923134814356, + "learning_rate": 3.074058816009817e-08, + "loss": 0.8645, + "step": 10487 + }, + { + "epoch": 0.9458447941561077, + "grad_norm": 1.9645439510773797, + "learning_rate": 3.063863691313284e-08, + "loss": 0.925, + "step": 10488 + }, + { + "epoch": 0.9459349776795779, + "grad_norm": 1.598560105559699, + "learning_rate": 3.0536853702937794e-08, + "loss": 0.9211, + "step": 10489 + }, + { + "epoch": 0.9460251612030482, + "grad_norm": 2.083111280687388, + "learning_rate": 3.043523853819807e-08, + "loss": 0.9816, + "step": 10490 + }, + { + "epoch": 0.9461153447265185, + "grad_norm": 2.0248150352492535, + "learning_rate": 3.0333791427583855e-08, + "loss": 0.9993, + "step": 10491 + }, + { + "epoch": 0.9462055282499887, + "grad_norm": 1.6141068606636928, + "learning_rate": 3.023251237975111e-08, + "loss": 1.1038, + "step": 10492 + }, + { + "epoch": 0.946295711773459, + "grad_norm": 0.6574053131574108, + "learning_rate": 3.0131401403341584e-08, + "loss": 0.8313, + "step": 10493 + }, + { + "epoch": 0.9463858952969293, + "grad_norm": 1.5045869153907214, + "learning_rate": 3.00304585069826e-08, + "loss": 0.9588, + "step": 10494 + }, + { + "epoch": 0.9464760788203995, + "grad_norm": 1.9061380572359936, + "learning_rate": 2.992968369928728e-08, + "loss": 0.9703, + "step": 10495 + }, + { + "epoch": 0.9465662623438698, + "grad_norm": 1.7713179504885683, + "learning_rate": 2.982907698885429e-08, + "loss": 0.9947, + "step": 10496 + }, + { + "epoch": 0.94665644586734, + "grad_norm": 1.9832612163457826, + "learning_rate": 2.9728638384267645e-08, + "loss": 0.9902, + "step": 10497 + }, + { + "epoch": 0.9467466293908103, + "grad_norm": 2.3670926216676067, + "learning_rate": 2.962836789409784e-08, + "loss": 0.9858, + "step": 10498 + }, + { + "epoch": 0.9468368129142806, + "grad_norm": 1.6436250414327933, + "learning_rate": 2.95282655268998e-08, + "loss": 1.0708, + "step": 10499 + }, + { + "epoch": 0.9469269964377508, + "grad_norm": 6.542864679676077, + "learning_rate": 2.942833129121558e-08, + "loss": 0.9804, + "step": 10500 + }, + { + "epoch": 0.947017179961221, + "grad_norm": 1.6811256107417425, + "learning_rate": 2.9328565195571475e-08, + "loss": 0.9054, + "step": 10501 + }, + { + "epoch": 0.9471073634846914, + "grad_norm": 1.6206945456321185, + "learning_rate": 2.9228967248480675e-08, + "loss": 1.0471, + "step": 10502 + }, + { + "epoch": 0.9471975470081616, + "grad_norm": 1.7604272680320774, + "learning_rate": 2.912953745844082e-08, + "loss": 1.0226, + "step": 10503 + }, + { + "epoch": 0.9472877305316318, + "grad_norm": 1.5678362704401225, + "learning_rate": 2.9030275833936247e-08, + "loss": 0.9718, + "step": 10504 + }, + { + "epoch": 0.9473779140551022, + "grad_norm": 1.881692196909612, + "learning_rate": 2.893118238343617e-08, + "loss": 0.881, + "step": 10505 + }, + { + "epoch": 0.9474680975785724, + "grad_norm": 2.399413863769324, + "learning_rate": 2.8832257115396052e-08, + "loss": 0.8906, + "step": 10506 + }, + { + "epoch": 0.9475582811020427, + "grad_norm": 1.5250400037677834, + "learning_rate": 2.873350003825692e-08, + "loss": 0.9634, + "step": 10507 + }, + { + "epoch": 0.9476484646255129, + "grad_norm": 0.5627013481899534, + "learning_rate": 2.8634911160444696e-08, + "loss": 0.7093, + "step": 10508 + }, + { + "epoch": 0.9477386481489832, + "grad_norm": 1.7445495057848475, + "learning_rate": 2.853649049037199e-08, + "loss": 1.0421, + "step": 10509 + }, + { + "epoch": 0.9478288316724535, + "grad_norm": 1.7888172764525605, + "learning_rate": 2.8438238036436525e-08, + "loss": 1.0268, + "step": 10510 + }, + { + "epoch": 0.9479190151959237, + "grad_norm": 1.7832352801428961, + "learning_rate": 2.834015380702137e-08, + "loss": 0.9746, + "step": 10511 + }, + { + "epoch": 0.9480091987193939, + "grad_norm": 0.5882635612918692, + "learning_rate": 2.824223781049606e-08, + "loss": 0.7505, + "step": 10512 + }, + { + "epoch": 0.9480993822428643, + "grad_norm": 2.531825341665468, + "learning_rate": 2.8144490055215465e-08, + "loss": 1.0409, + "step": 10513 + }, + { + "epoch": 0.9481895657663345, + "grad_norm": 1.5821654463000836, + "learning_rate": 2.8046910549519355e-08, + "loss": 1.0134, + "step": 10514 + }, + { + "epoch": 0.9482797492898047, + "grad_norm": 1.5287250019565506, + "learning_rate": 2.794949930173418e-08, + "loss": 0.983, + "step": 10515 + }, + { + "epoch": 0.948369932813275, + "grad_norm": 1.6718235186199795, + "learning_rate": 2.7852256320171296e-08, + "loss": 1.0105, + "step": 10516 + }, + { + "epoch": 0.9484601163367453, + "grad_norm": 1.932805989407897, + "learning_rate": 2.775518161312851e-08, + "loss": 0.843, + "step": 10517 + }, + { + "epoch": 0.9485502998602156, + "grad_norm": 2.6702890429685784, + "learning_rate": 2.76582751888883e-08, + "loss": 1.0087, + "step": 10518 + }, + { + "epoch": 0.9486404833836858, + "grad_norm": 1.9697264414585363, + "learning_rate": 2.756153705571962e-08, + "loss": 1.0014, + "step": 10519 + }, + { + "epoch": 0.948730666907156, + "grad_norm": 1.360615263147521, + "learning_rate": 2.74649672218763e-08, + "loss": 0.96, + "step": 10520 + }, + { + "epoch": 0.9488208504306264, + "grad_norm": 3.0726711541499867, + "learning_rate": 2.7368565695598424e-08, + "loss": 0.9703, + "step": 10521 + }, + { + "epoch": 0.9489110339540966, + "grad_norm": 1.6382846670337965, + "learning_rate": 2.727233248511185e-08, + "loss": 0.905, + "step": 10522 + }, + { + "epoch": 0.9490012174775668, + "grad_norm": 1.9817414116565664, + "learning_rate": 2.71762675986269e-08, + "loss": 0.9323, + "step": 10523 + }, + { + "epoch": 0.9490914010010371, + "grad_norm": 1.5229955570158646, + "learning_rate": 2.7080371044341242e-08, + "loss": 1.0203, + "step": 10524 + }, + { + "epoch": 0.9491815845245074, + "grad_norm": 1.4880950090317322, + "learning_rate": 2.6984642830436556e-08, + "loss": 1.0105, + "step": 10525 + }, + { + "epoch": 0.9492717680479776, + "grad_norm": 1.8399945175047756, + "learning_rate": 2.688908296508141e-08, + "loss": 0.9601, + "step": 10526 + }, + { + "epoch": 0.9493619515714479, + "grad_norm": 1.7558519747263062, + "learning_rate": 2.679369145642929e-08, + "loss": 1.0311, + "step": 10527 + }, + { + "epoch": 0.9494521350949181, + "grad_norm": 1.6454470297796489, + "learning_rate": 2.669846831261946e-08, + "loss": 1.0007, + "step": 10528 + }, + { + "epoch": 0.9495423186183884, + "grad_norm": 1.8192925604145251, + "learning_rate": 2.6603413541776976e-08, + "loss": 0.9699, + "step": 10529 + }, + { + "epoch": 0.9496325021418587, + "grad_norm": 1.3834793724201486, + "learning_rate": 2.6508527152012683e-08, + "loss": 0.9336, + "step": 10530 + }, + { + "epoch": 0.9497226856653289, + "grad_norm": 1.4550650986465996, + "learning_rate": 2.641380915142233e-08, + "loss": 0.8836, + "step": 10531 + }, + { + "epoch": 0.9498128691887993, + "grad_norm": 1.661575536470496, + "learning_rate": 2.6319259548088334e-08, + "loss": 1.0042, + "step": 10532 + }, + { + "epoch": 0.9499030527122695, + "grad_norm": 1.6552155341089174, + "learning_rate": 2.6224878350077585e-08, + "loss": 0.9816, + "step": 10533 + }, + { + "epoch": 0.9499932362357397, + "grad_norm": 1.5953616053452246, + "learning_rate": 2.6130665565443633e-08, + "loss": 0.8944, + "step": 10534 + }, + { + "epoch": 0.95008341975921, + "grad_norm": 1.442117572382735, + "learning_rate": 2.603662120222494e-08, + "loss": 0.9839, + "step": 10535 + }, + { + "epoch": 0.9501736032826803, + "grad_norm": 1.9429649067692312, + "learning_rate": 2.59427452684462e-08, + "loss": 0.8236, + "step": 10536 + }, + { + "epoch": 0.9502637868061505, + "grad_norm": 1.5595969124414408, + "learning_rate": 2.5849037772117443e-08, + "loss": 0.9585, + "step": 10537 + }, + { + "epoch": 0.9503539703296208, + "grad_norm": 1.7463443584794336, + "learning_rate": 2.575549872123384e-08, + "loss": 1.0054, + "step": 10538 + }, + { + "epoch": 0.950444153853091, + "grad_norm": 1.8102793408436655, + "learning_rate": 2.5662128123776994e-08, + "loss": 0.9857, + "step": 10539 + }, + { + "epoch": 0.9505343373765613, + "grad_norm": 1.5008094510909393, + "learning_rate": 2.5568925987713875e-08, + "loss": 0.9769, + "step": 10540 + }, + { + "epoch": 0.9506245209000316, + "grad_norm": 1.9846008052159616, + "learning_rate": 2.5475892320996785e-08, + "loss": 0.9541, + "step": 10541 + }, + { + "epoch": 0.9507147044235018, + "grad_norm": 0.6667734348711409, + "learning_rate": 2.5383027131564038e-08, + "loss": 0.7984, + "step": 10542 + }, + { + "epoch": 0.950804887946972, + "grad_norm": 1.650407526956721, + "learning_rate": 2.52903304273393e-08, + "loss": 0.9483, + "step": 10543 + }, + { + "epoch": 0.9508950714704424, + "grad_norm": 1.4973848115819457, + "learning_rate": 2.519780221623202e-08, + "loss": 0.9626, + "step": 10544 + }, + { + "epoch": 0.9509852549939126, + "grad_norm": 1.6098662432202808, + "learning_rate": 2.510544250613722e-08, + "loss": 0.9363, + "step": 10545 + }, + { + "epoch": 0.9510754385173829, + "grad_norm": 2.020517370423563, + "learning_rate": 2.501325130493548e-08, + "loss": 1.0786, + "step": 10546 + }, + { + "epoch": 0.9511656220408531, + "grad_norm": 0.747629631947843, + "learning_rate": 2.4921228620493395e-08, + "loss": 0.8157, + "step": 10547 + }, + { + "epoch": 0.9512558055643234, + "grad_norm": 0.7465932482219203, + "learning_rate": 2.4829374460662244e-08, + "loss": 0.8524, + "step": 10548 + }, + { + "epoch": 0.9513459890877937, + "grad_norm": 1.4251214159721448, + "learning_rate": 2.473768883327976e-08, + "loss": 0.9939, + "step": 10549 + }, + { + "epoch": 0.9514361726112639, + "grad_norm": 1.656650278917359, + "learning_rate": 2.464617174616923e-08, + "loss": 0.9426, + "step": 10550 + }, + { + "epoch": 0.9515263561347341, + "grad_norm": 1.7541013749295287, + "learning_rate": 2.455482320713953e-08, + "loss": 0.9774, + "step": 10551 + }, + { + "epoch": 0.9516165396582045, + "grad_norm": 1.7227900100643334, + "learning_rate": 2.4463643223984643e-08, + "loss": 1.0407, + "step": 10552 + }, + { + "epoch": 0.9517067231816747, + "grad_norm": 1.8106739448144527, + "learning_rate": 2.4372631804484567e-08, + "loss": 1.0537, + "step": 10553 + }, + { + "epoch": 0.9517969067051449, + "grad_norm": 1.6902098620201835, + "learning_rate": 2.4281788956405313e-08, + "loss": 1.0575, + "step": 10554 + }, + { + "epoch": 0.9518870902286153, + "grad_norm": 1.8328365028416151, + "learning_rate": 2.4191114687497572e-08, + "loss": 0.9078, + "step": 10555 + }, + { + "epoch": 0.9519772737520855, + "grad_norm": 2.120557292168763, + "learning_rate": 2.4100609005498706e-08, + "loss": 0.9386, + "step": 10556 + }, + { + "epoch": 0.9520674572755558, + "grad_norm": 2.0267406000070154, + "learning_rate": 2.4010271918130764e-08, + "loss": 0.9104, + "step": 10557 + }, + { + "epoch": 0.952157640799026, + "grad_norm": 1.6261826393198184, + "learning_rate": 2.39201034331018e-08, + "loss": 0.9415, + "step": 10558 + }, + { + "epoch": 0.9522478243224963, + "grad_norm": 1.958721900158998, + "learning_rate": 2.3830103558105663e-08, + "loss": 0.9679, + "step": 10559 + }, + { + "epoch": 0.9523380078459666, + "grad_norm": 1.750895952027607, + "learning_rate": 2.374027230082154e-08, + "loss": 0.9831, + "step": 10560 + }, + { + "epoch": 0.9524281913694368, + "grad_norm": 0.6795700234802127, + "learning_rate": 2.365060966891441e-08, + "loss": 0.8316, + "step": 10561 + }, + { + "epoch": 0.952518374892907, + "grad_norm": 1.7344633200212525, + "learning_rate": 2.3561115670034827e-08, + "loss": 0.9751, + "step": 10562 + }, + { + "epoch": 0.9526085584163774, + "grad_norm": 1.7895177403583489, + "learning_rate": 2.3471790311818675e-08, + "loss": 1.0001, + "step": 10563 + }, + { + "epoch": 0.9526987419398476, + "grad_norm": 1.753542992031374, + "learning_rate": 2.338263360188808e-08, + "loss": 0.9695, + "step": 10564 + }, + { + "epoch": 0.9527889254633178, + "grad_norm": 1.6421567033807254, + "learning_rate": 2.329364554784985e-08, + "loss": 0.9171, + "step": 10565 + }, + { + "epoch": 0.9528791089867881, + "grad_norm": 5.221896199720394, + "learning_rate": 2.3204826157297465e-08, + "loss": 1.0356, + "step": 10566 + }, + { + "epoch": 0.9529692925102584, + "grad_norm": 1.682477966323029, + "learning_rate": 2.3116175437809082e-08, + "loss": 0.9197, + "step": 10567 + }, + { + "epoch": 0.9530594760337286, + "grad_norm": 0.6447893955106491, + "learning_rate": 2.30276933969491e-08, + "loss": 0.8335, + "step": 10568 + }, + { + "epoch": 0.9531496595571989, + "grad_norm": 1.4378960105424137, + "learning_rate": 2.2939380042267255e-08, + "loss": 1.0345, + "step": 10569 + }, + { + "epoch": 0.9532398430806691, + "grad_norm": 1.7223167171327665, + "learning_rate": 2.2851235381298627e-08, + "loss": 0.8831, + "step": 10570 + }, + { + "epoch": 0.9533300266041395, + "grad_norm": 1.602762358995507, + "learning_rate": 2.2763259421564986e-08, + "loss": 0.9059, + "step": 10571 + }, + { + "epoch": 0.9534202101276097, + "grad_norm": 1.7269087914533945, + "learning_rate": 2.2675452170571873e-08, + "loss": 0.9491, + "step": 10572 + }, + { + "epoch": 0.9535103936510799, + "grad_norm": 1.598719609070876, + "learning_rate": 2.2587813635812414e-08, + "loss": 0.9717, + "step": 10573 + }, + { + "epoch": 0.9536005771745502, + "grad_norm": 1.5331930172337997, + "learning_rate": 2.2500343824763958e-08, + "loss": 1.0647, + "step": 10574 + }, + { + "epoch": 0.9536907606980205, + "grad_norm": 1.72167788725622, + "learning_rate": 2.2413042744890088e-08, + "loss": 0.9263, + "step": 10575 + }, + { + "epoch": 0.9537809442214907, + "grad_norm": 16.177999293262825, + "learning_rate": 2.2325910403639514e-08, + "loss": 0.9667, + "step": 10576 + }, + { + "epoch": 0.953871127744961, + "grad_norm": 1.4756496776696917, + "learning_rate": 2.223894680844718e-08, + "loss": 0.9343, + "step": 10577 + }, + { + "epoch": 0.9539613112684313, + "grad_norm": 1.4771680448028484, + "learning_rate": 2.2152151966733146e-08, + "loss": 0.813, + "step": 10578 + }, + { + "epoch": 0.9540514947919015, + "grad_norm": 1.3576463253059314, + "learning_rate": 2.2065525885903267e-08, + "loss": 1.0129, + "step": 10579 + }, + { + "epoch": 0.9541416783153718, + "grad_norm": 1.638713340552918, + "learning_rate": 2.1979068573348747e-08, + "loss": 1.0237, + "step": 10580 + }, + { + "epoch": 0.954231861838842, + "grad_norm": 1.4849678060464835, + "learning_rate": 2.1892780036447013e-08, + "loss": 0.9419, + "step": 10581 + }, + { + "epoch": 0.9543220453623124, + "grad_norm": 1.7944133316245492, + "learning_rate": 2.1806660282560175e-08, + "loss": 1.121, + "step": 10582 + }, + { + "epoch": 0.9544122288857826, + "grad_norm": 1.4115542089502966, + "learning_rate": 2.1720709319037024e-08, + "loss": 1.0033, + "step": 10583 + }, + { + "epoch": 0.9545024124092528, + "grad_norm": 1.3362122637966594, + "learning_rate": 2.1634927153211023e-08, + "loss": 0.9443, + "step": 10584 + }, + { + "epoch": 0.954592595932723, + "grad_norm": 2.954038609866422, + "learning_rate": 2.1549313792401437e-08, + "loss": 1.0433, + "step": 10585 + }, + { + "epoch": 0.9546827794561934, + "grad_norm": 1.4950171997203634, + "learning_rate": 2.1463869243913746e-08, + "loss": 1.0706, + "step": 10586 + }, + { + "epoch": 0.9547729629796636, + "grad_norm": 1.7866185582620708, + "learning_rate": 2.1378593515037902e-08, + "loss": 0.9621, + "step": 10587 + }, + { + "epoch": 0.9548631465031339, + "grad_norm": 1.5724830619251529, + "learning_rate": 2.129348661305075e-08, + "loss": 1.0333, + "step": 10588 + }, + { + "epoch": 0.9549533300266041, + "grad_norm": 1.5420347066553142, + "learning_rate": 2.1208548545213813e-08, + "loss": 1.0127, + "step": 10589 + }, + { + "epoch": 0.9550435135500744, + "grad_norm": 1.5787308540300822, + "learning_rate": 2.1123779318774404e-08, + "loss": 0.9302, + "step": 10590 + }, + { + "epoch": 0.9551336970735447, + "grad_norm": 1.4640840315075183, + "learning_rate": 2.1039178940965408e-08, + "loss": 0.924, + "step": 10591 + }, + { + "epoch": 0.9552238805970149, + "grad_norm": 1.5177353786065202, + "learning_rate": 2.0954747419005712e-08, + "loss": 1.0413, + "step": 10592 + }, + { + "epoch": 0.9553140641204851, + "grad_norm": 1.7425181669495007, + "learning_rate": 2.087048476009934e-08, + "loss": 0.9571, + "step": 10593 + }, + { + "epoch": 0.9554042476439555, + "grad_norm": 1.5697539745413542, + "learning_rate": 2.0786390971435862e-08, + "loss": 0.9512, + "step": 10594 + }, + { + "epoch": 0.9554944311674257, + "grad_norm": 2.499194396873084, + "learning_rate": 2.070246606019088e-08, + "loss": 0.9483, + "step": 10595 + }, + { + "epoch": 0.955584614690896, + "grad_norm": 1.949883601798357, + "learning_rate": 2.0618710033525112e-08, + "loss": 0.8929, + "step": 10596 + }, + { + "epoch": 0.9556747982143662, + "grad_norm": 1.7042921773988735, + "learning_rate": 2.053512289858528e-08, + "loss": 0.9458, + "step": 10597 + }, + { + "epoch": 0.9557649817378365, + "grad_norm": 2.6083462451781796, + "learning_rate": 2.0451704662503456e-08, + "loss": 1.02, + "step": 10598 + }, + { + "epoch": 0.9558551652613068, + "grad_norm": 1.8857086826629297, + "learning_rate": 2.0368455332397282e-08, + "loss": 1.059, + "step": 10599 + }, + { + "epoch": 0.955945348784777, + "grad_norm": 1.6065920869974075, + "learning_rate": 2.0285374915369967e-08, + "loss": 0.9956, + "step": 10600 + }, + { + "epoch": 0.9560355323082473, + "grad_norm": 2.6374642681576, + "learning_rate": 2.020246341851073e-08, + "loss": 1.0183, + "step": 10601 + }, + { + "epoch": 0.9561257158317176, + "grad_norm": 1.469052709673839, + "learning_rate": 2.0119720848893463e-08, + "loss": 1.0207, + "step": 10602 + }, + { + "epoch": 0.9562158993551878, + "grad_norm": 1.6241170244898213, + "learning_rate": 2.0037147213578964e-08, + "loss": 0.9266, + "step": 10603 + }, + { + "epoch": 0.956306082878658, + "grad_norm": 1.5973673141847435, + "learning_rate": 1.9954742519612265e-08, + "loss": 1.0048, + "step": 10604 + }, + { + "epoch": 0.9563962664021284, + "grad_norm": 1.6658259973687293, + "learning_rate": 1.9872506774024633e-08, + "loss": 0.9565, + "step": 10605 + }, + { + "epoch": 0.9564864499255986, + "grad_norm": 1.9656643983322095, + "learning_rate": 1.979043998383334e-08, + "loss": 0.9913, + "step": 10606 + }, + { + "epoch": 0.9565766334490688, + "grad_norm": 1.3365072184438034, + "learning_rate": 1.970854215604034e-08, + "loss": 0.9233, + "step": 10607 + }, + { + "epoch": 0.9566668169725391, + "grad_norm": 1.657469893997676, + "learning_rate": 1.9626813297633826e-08, + "loss": 0.9521, + "step": 10608 + }, + { + "epoch": 0.9567570004960094, + "grad_norm": 1.6031087714669572, + "learning_rate": 1.954525341558688e-08, + "loss": 0.8756, + "step": 10609 + }, + { + "epoch": 0.9568471840194797, + "grad_norm": 1.7293509496209716, + "learning_rate": 1.9463862516859498e-08, + "loss": 0.9461, + "step": 10610 + }, + { + "epoch": 0.9569373675429499, + "grad_norm": 3.705212354867873, + "learning_rate": 1.938264060839545e-08, + "loss": 1.0627, + "step": 10611 + }, + { + "epoch": 0.9570275510664201, + "grad_norm": 2.2190316249868207, + "learning_rate": 1.9301587697126086e-08, + "loss": 0.9532, + "step": 10612 + }, + { + "epoch": 0.9571177345898905, + "grad_norm": 2.377335755042987, + "learning_rate": 1.9220703789966318e-08, + "loss": 0.9328, + "step": 10613 + }, + { + "epoch": 0.9572079181133607, + "grad_norm": 1.6349541157808125, + "learning_rate": 1.913998889381818e-08, + "loss": 0.9363, + "step": 10614 + }, + { + "epoch": 0.9572981016368309, + "grad_norm": 1.5807538939233035, + "learning_rate": 1.9059443015568387e-08, + "loss": 0.9917, + "step": 10615 + }, + { + "epoch": 0.9573882851603012, + "grad_norm": 1.7868241212745737, + "learning_rate": 1.8979066162089884e-08, + "loss": 0.9928, + "step": 10616 + }, + { + "epoch": 0.9574784686837715, + "grad_norm": 1.6725166016625708, + "learning_rate": 1.889885834024052e-08, + "loss": 0.9872, + "step": 10617 + }, + { + "epoch": 0.9575686522072417, + "grad_norm": 2.020642563247692, + "learning_rate": 1.8818819556864374e-08, + "loss": 0.9298, + "step": 10618 + }, + { + "epoch": 0.957658835730712, + "grad_norm": 2.121646141778037, + "learning_rate": 1.873894981879065e-08, + "loss": 0.9624, + "step": 10619 + }, + { + "epoch": 0.9577490192541822, + "grad_norm": 0.6337161734919088, + "learning_rate": 1.8659249132834342e-08, + "loss": 0.8279, + "step": 10620 + }, + { + "epoch": 0.9578392027776526, + "grad_norm": 1.4019803822297165, + "learning_rate": 1.857971750579579e-08, + "loss": 0.9588, + "step": 10621 + }, + { + "epoch": 0.9579293863011228, + "grad_norm": 1.6973543308549472, + "learning_rate": 1.8500354944461116e-08, + "loss": 0.8619, + "step": 10622 + }, + { + "epoch": 0.958019569824593, + "grad_norm": 1.3955154578905118, + "learning_rate": 1.8421161455602242e-08, + "loss": 0.9882, + "step": 10623 + }, + { + "epoch": 0.9581097533480634, + "grad_norm": 1.485450756364869, + "learning_rate": 1.834213704597598e-08, + "loss": 1.005, + "step": 10624 + }, + { + "epoch": 0.9581999368715336, + "grad_norm": 1.5859769618945536, + "learning_rate": 1.8263281722325385e-08, + "loss": 0.9891, + "step": 10625 + }, + { + "epoch": 0.9582901203950038, + "grad_norm": 1.79855725695854, + "learning_rate": 1.818459549137885e-08, + "loss": 0.9887, + "step": 10626 + }, + { + "epoch": 0.9583803039184741, + "grad_norm": 2.205585269370161, + "learning_rate": 1.8106078359850117e-08, + "loss": 0.9692, + "step": 10627 + }, + { + "epoch": 0.9584704874419444, + "grad_norm": 2.108209132379805, + "learning_rate": 1.802773033443894e-08, + "loss": 0.9949, + "step": 10628 + }, + { + "epoch": 0.9585606709654146, + "grad_norm": 1.5839545721032584, + "learning_rate": 1.7949551421830413e-08, + "loss": 0.9439, + "step": 10629 + }, + { + "epoch": 0.9586508544888849, + "grad_norm": 1.478431689199165, + "learning_rate": 1.7871541628694752e-08, + "loss": 0.9467, + "step": 10630 + }, + { + "epoch": 0.9587410380123551, + "grad_norm": 1.4214945692937473, + "learning_rate": 1.779370096168864e-08, + "loss": 0.8875, + "step": 10631 + }, + { + "epoch": 0.9588312215358255, + "grad_norm": 1.6831264334076543, + "learning_rate": 1.771602942745387e-08, + "loss": 0.9438, + "step": 10632 + }, + { + "epoch": 0.9589214050592957, + "grad_norm": 3.921761317334236, + "learning_rate": 1.763852703261759e-08, + "loss": 0.9617, + "step": 10633 + }, + { + "epoch": 0.9590115885827659, + "grad_norm": 1.6078950317672365, + "learning_rate": 1.756119378379295e-08, + "loss": 1.0388, + "step": 10634 + }, + { + "epoch": 0.9591017721062362, + "grad_norm": 1.6667045430012148, + "learning_rate": 1.7484029687578005e-08, + "loss": 1.0686, + "step": 10635 + }, + { + "epoch": 0.9591919556297065, + "grad_norm": 1.813731334335219, + "learning_rate": 1.740703475055727e-08, + "loss": 0.9529, + "step": 10636 + }, + { + "epoch": 0.9592821391531767, + "grad_norm": 1.8374870740993798, + "learning_rate": 1.7330208979300153e-08, + "loss": 1.0149, + "step": 10637 + }, + { + "epoch": 0.959372322676647, + "grad_norm": 1.501505652392581, + "learning_rate": 1.725355238036208e-08, + "loss": 1.0206, + "step": 10638 + }, + { + "epoch": 0.9594625062001172, + "grad_norm": 1.6344405093658214, + "learning_rate": 1.7177064960283594e-08, + "loss": 1.0223, + "step": 10639 + }, + { + "epoch": 0.9595526897235875, + "grad_norm": 2.447507415542045, + "learning_rate": 1.7100746725591253e-08, + "loss": 0.9721, + "step": 10640 + }, + { + "epoch": 0.9596428732470578, + "grad_norm": 2.1572599395476995, + "learning_rate": 1.7024597682796517e-08, + "loss": 0.9689, + "step": 10641 + }, + { + "epoch": 0.959733056770528, + "grad_norm": 1.7513922194266796, + "learning_rate": 1.6948617838397293e-08, + "loss": 1.0525, + "step": 10642 + }, + { + "epoch": 0.9598232402939982, + "grad_norm": 1.404690025590402, + "learning_rate": 1.6872807198876404e-08, + "loss": 1.0096, + "step": 10643 + }, + { + "epoch": 0.9599134238174686, + "grad_norm": 1.9040644264180142, + "learning_rate": 1.679716577070245e-08, + "loss": 1.0799, + "step": 10644 + }, + { + "epoch": 0.9600036073409388, + "grad_norm": 1.6862363675911765, + "learning_rate": 1.6721693560329596e-08, + "loss": 1.0655, + "step": 10645 + }, + { + "epoch": 0.960093790864409, + "grad_norm": 2.283209808052666, + "learning_rate": 1.6646390574197366e-08, + "loss": 0.8865, + "step": 10646 + }, + { + "epoch": 0.9601839743878793, + "grad_norm": 0.7015700418453481, + "learning_rate": 1.6571256818731504e-08, + "loss": 0.8336, + "step": 10647 + }, + { + "epoch": 0.9602741579113496, + "grad_norm": 1.464083950085079, + "learning_rate": 1.6496292300342218e-08, + "loss": 0.9121, + "step": 10648 + }, + { + "epoch": 0.9603643414348199, + "grad_norm": 1.6151604964163655, + "learning_rate": 1.642149702542639e-08, + "loss": 0.9182, + "step": 10649 + }, + { + "epoch": 0.9604545249582901, + "grad_norm": 1.7956274594623043, + "learning_rate": 1.634687100036558e-08, + "loss": 0.9416, + "step": 10650 + }, + { + "epoch": 0.9605447084817604, + "grad_norm": 1.667115468951208, + "learning_rate": 1.627241423152781e-08, + "loss": 0.912, + "step": 10651 + }, + { + "epoch": 0.9606348920052307, + "grad_norm": 1.6304915875753332, + "learning_rate": 1.619812672526555e-08, + "loss": 1.0302, + "step": 10652 + }, + { + "epoch": 0.9607250755287009, + "grad_norm": 1.786991102963902, + "learning_rate": 1.6124008487917727e-08, + "loss": 0.962, + "step": 10653 + }, + { + "epoch": 0.9608152590521711, + "grad_norm": 0.6431533187029095, + "learning_rate": 1.6050059525808623e-08, + "loss": 0.8315, + "step": 10654 + }, + { + "epoch": 0.9609054425756415, + "grad_norm": 0.6324554660634196, + "learning_rate": 1.597627984524763e-08, + "loss": 0.8274, + "step": 10655 + }, + { + "epoch": 0.9609956260991117, + "grad_norm": 1.4552323842046548, + "learning_rate": 1.590266945253038e-08, + "loss": 1.0193, + "step": 10656 + }, + { + "epoch": 0.9610858096225819, + "grad_norm": 1.7370349316507785, + "learning_rate": 1.582922835393763e-08, + "loss": 0.994, + "step": 10657 + }, + { + "epoch": 0.9611759931460522, + "grad_norm": 1.5017989814940782, + "learning_rate": 1.5755956555735473e-08, + "loss": 0.9535, + "step": 10658 + }, + { + "epoch": 0.9612661766695225, + "grad_norm": 0.7662734574776988, + "learning_rate": 1.5682854064176244e-08, + "loss": 0.9025, + "step": 10659 + }, + { + "epoch": 0.9613563601929928, + "grad_norm": 8.527873841889525, + "learning_rate": 1.5609920885497395e-08, + "loss": 0.91, + "step": 10660 + }, + { + "epoch": 0.961446543716463, + "grad_norm": 1.6733524532416282, + "learning_rate": 1.5537157025921732e-08, + "loss": 1.0014, + "step": 10661 + }, + { + "epoch": 0.9615367272399332, + "grad_norm": 3.0396975758353104, + "learning_rate": 1.5464562491658285e-08, + "loss": 0.9587, + "step": 10662 + }, + { + "epoch": 0.9616269107634036, + "grad_norm": 1.4342774468120154, + "learning_rate": 1.5392137288900764e-08, + "loss": 1.0136, + "step": 10663 + }, + { + "epoch": 0.9617170942868738, + "grad_norm": 0.5939735694393096, + "learning_rate": 1.531988142382934e-08, + "loss": 0.8168, + "step": 10664 + }, + { + "epoch": 0.961807277810344, + "grad_norm": 1.7723547365093077, + "learning_rate": 1.5247794902608634e-08, + "loss": 0.9121, + "step": 10665 + }, + { + "epoch": 0.9618974613338143, + "grad_norm": 2.7618804409494047, + "learning_rate": 1.5175877731390398e-08, + "loss": 0.9372, + "step": 10666 + }, + { + "epoch": 0.9619876448572846, + "grad_norm": 1.966129072172624, + "learning_rate": 1.510412991631016e-08, + "loss": 1.0061, + "step": 10667 + }, + { + "epoch": 0.9620778283807548, + "grad_norm": 2.265537970614301, + "learning_rate": 1.503255146349014e-08, + "loss": 1.0046, + "step": 10668 + }, + { + "epoch": 0.9621680119042251, + "grad_norm": 1.8831774092862816, + "learning_rate": 1.4961142379037893e-08, + "loss": 0.9405, + "step": 10669 + }, + { + "epoch": 0.9622581954276953, + "grad_norm": 1.6273798022440715, + "learning_rate": 1.4889902669046327e-08, + "loss": 1.0192, + "step": 10670 + }, + { + "epoch": 0.9623483789511657, + "grad_norm": 3.812458726002891, + "learning_rate": 1.4818832339594135e-08, + "loss": 0.9813, + "step": 10671 + }, + { + "epoch": 0.9624385624746359, + "grad_norm": 1.5453148883303556, + "learning_rate": 1.474793139674535e-08, + "loss": 0.9322, + "step": 10672 + }, + { + "epoch": 0.9625287459981061, + "grad_norm": 1.870730179602796, + "learning_rate": 1.4677199846549581e-08, + "loss": 0.9769, + "step": 10673 + }, + { + "epoch": 0.9626189295215765, + "grad_norm": 1.573259009017379, + "learning_rate": 1.4606637695042224e-08, + "loss": 0.9754, + "step": 10674 + }, + { + "epoch": 0.9627091130450467, + "grad_norm": 1.7638407781586156, + "learning_rate": 1.4536244948243793e-08, + "loss": 1.0663, + "step": 10675 + }, + { + "epoch": 0.9627992965685169, + "grad_norm": 2.5304519902446008, + "learning_rate": 1.4466021612160595e-08, + "loss": 0.9433, + "step": 10676 + }, + { + "epoch": 0.9628894800919872, + "grad_norm": 2.4952745600749973, + "learning_rate": 1.4395967692784505e-08, + "loss": 0.9413, + "step": 10677 + }, + { + "epoch": 0.9629796636154575, + "grad_norm": 2.23202017092734, + "learning_rate": 1.4326083196092963e-08, + "loss": 0.9104, + "step": 10678 + }, + { + "epoch": 0.9630698471389277, + "grad_norm": 1.5296875124688776, + "learning_rate": 1.42563681280492e-08, + "loss": 0.99, + "step": 10679 + }, + { + "epoch": 0.963160030662398, + "grad_norm": 1.554329068832443, + "learning_rate": 1.4186822494600902e-08, + "loss": 0.953, + "step": 10680 + }, + { + "epoch": 0.9632502141858682, + "grad_norm": 1.3948527587927886, + "learning_rate": 1.4117446301682877e-08, + "loss": 0.9668, + "step": 10681 + }, + { + "epoch": 0.9633403977093385, + "grad_norm": 1.5013867961168286, + "learning_rate": 1.4048239555214392e-08, + "loss": 0.9311, + "step": 10682 + }, + { + "epoch": 0.9634305812328088, + "grad_norm": 1.5737921276316906, + "learning_rate": 1.3979202261100497e-08, + "loss": 1.0068, + "step": 10683 + }, + { + "epoch": 0.963520764756279, + "grad_norm": 1.6945693677006743, + "learning_rate": 1.3910334425231817e-08, + "loss": 1.0396, + "step": 10684 + }, + { + "epoch": 0.9636109482797492, + "grad_norm": 1.6624222672725215, + "learning_rate": 1.384163605348454e-08, + "loss": 0.9757, + "step": 10685 + }, + { + "epoch": 0.9637011318032196, + "grad_norm": 1.887768250155017, + "learning_rate": 1.3773107151720642e-08, + "loss": 1.035, + "step": 10686 + }, + { + "epoch": 0.9637913153266898, + "grad_norm": 1.8893787911213333, + "learning_rate": 1.3704747725787003e-08, + "loss": 1.0392, + "step": 10687 + }, + { + "epoch": 0.9638814988501601, + "grad_norm": 0.72874034044201, + "learning_rate": 1.3636557781516512e-08, + "loss": 0.8238, + "step": 10688 + }, + { + "epoch": 0.9639716823736303, + "grad_norm": 1.5385198265734645, + "learning_rate": 1.3568537324727847e-08, + "loss": 0.9786, + "step": 10689 + }, + { + "epoch": 0.9640618658971006, + "grad_norm": 1.5344903098478158, + "learning_rate": 1.3500686361224589e-08, + "loss": 0.9766, + "step": 10690 + }, + { + "epoch": 0.9641520494205709, + "grad_norm": 1.7519446875667786, + "learning_rate": 1.3433004896796108e-08, + "loss": 0.9741, + "step": 10691 + }, + { + "epoch": 0.9642422329440411, + "grad_norm": 1.8228521353789078, + "learning_rate": 1.336549293721756e-08, + "loss": 0.8702, + "step": 10692 + }, + { + "epoch": 0.9643324164675113, + "grad_norm": 0.661293089243326, + "learning_rate": 1.3298150488249227e-08, + "loss": 0.8101, + "step": 10693 + }, + { + "epoch": 0.9644225999909817, + "grad_norm": 2.21632198675109, + "learning_rate": 1.3230977555637401e-08, + "loss": 1.052, + "step": 10694 + }, + { + "epoch": 0.9645127835144519, + "grad_norm": 1.2907531286745133, + "learning_rate": 1.3163974145113499e-08, + "loss": 0.9365, + "step": 10695 + }, + { + "epoch": 0.9646029670379221, + "grad_norm": 1.8316884319943323, + "learning_rate": 1.3097140262394723e-08, + "loss": 0.8734, + "step": 10696 + }, + { + "epoch": 0.9646931505613925, + "grad_norm": 2.3339431056759343, + "learning_rate": 1.303047591318318e-08, + "loss": 1.0108, + "step": 10697 + }, + { + "epoch": 0.9647833340848627, + "grad_norm": 2.021762581814527, + "learning_rate": 1.2963981103167875e-08, + "loss": 0.9104, + "step": 10698 + }, + { + "epoch": 0.964873517608333, + "grad_norm": 1.4840413571890145, + "learning_rate": 1.2897655838021825e-08, + "loss": 0.9613, + "step": 10699 + }, + { + "epoch": 0.9649637011318032, + "grad_norm": 2.2297327861718075, + "learning_rate": 1.2831500123404726e-08, + "loss": 0.9824, + "step": 10700 + }, + { + "epoch": 0.9650538846552735, + "grad_norm": 2.6831042691506646, + "learning_rate": 1.2765513964961172e-08, + "loss": 1.1245, + "step": 10701 + }, + { + "epoch": 0.9651440681787438, + "grad_norm": 1.58142575580141, + "learning_rate": 1.2699697368321549e-08, + "loss": 0.9701, + "step": 10702 + }, + { + "epoch": 0.965234251702214, + "grad_norm": 2.489458557151711, + "learning_rate": 1.2634050339101366e-08, + "loss": 0.9635, + "step": 10703 + }, + { + "epoch": 0.9653244352256842, + "grad_norm": 1.4659197708127956, + "learning_rate": 1.2568572882902361e-08, + "loss": 0.9564, + "step": 10704 + }, + { + "epoch": 0.9654146187491546, + "grad_norm": 1.374723666773946, + "learning_rate": 1.2503265005311402e-08, + "loss": 0.9485, + "step": 10705 + }, + { + "epoch": 0.9655048022726248, + "grad_norm": 1.6053685809069997, + "learning_rate": 1.2438126711900698e-08, + "loss": 0.9385, + "step": 10706 + }, + { + "epoch": 0.965594985796095, + "grad_norm": 1.3845727353068757, + "learning_rate": 1.2373158008228247e-08, + "loss": 0.8267, + "step": 10707 + }, + { + "epoch": 0.9656851693195653, + "grad_norm": 2.4456162919070543, + "learning_rate": 1.2308358899837833e-08, + "loss": 0.9509, + "step": 10708 + }, + { + "epoch": 0.9657753528430356, + "grad_norm": 1.7735437033693955, + "learning_rate": 1.224372939225815e-08, + "loss": 0.9544, + "step": 10709 + }, + { + "epoch": 0.9658655363665059, + "grad_norm": 1.5173287008719858, + "learning_rate": 1.2179269491003674e-08, + "loss": 1.0596, + "step": 10710 + }, + { + "epoch": 0.9659557198899761, + "grad_norm": 1.9453077201324935, + "learning_rate": 1.2114979201574894e-08, + "loss": 0.8507, + "step": 10711 + }, + { + "epoch": 0.9660459034134463, + "grad_norm": 1.4102784418228032, + "learning_rate": 1.2050858529456975e-08, + "loss": 0.9945, + "step": 10712 + }, + { + "epoch": 0.9661360869369167, + "grad_norm": 1.7325027257724475, + "learning_rate": 1.1986907480121545e-08, + "loss": 0.9774, + "step": 10713 + }, + { + "epoch": 0.9662262704603869, + "grad_norm": 1.5688624733047964, + "learning_rate": 1.192312605902468e-08, + "loss": 0.9997, + "step": 10714 + }, + { + "epoch": 0.9663164539838571, + "grad_norm": 0.820730520641806, + "learning_rate": 1.1859514271608917e-08, + "loss": 0.8392, + "step": 10715 + }, + { + "epoch": 0.9664066375073274, + "grad_norm": 3.4842943943433706, + "learning_rate": 1.1796072123301914e-08, + "loss": 1.0066, + "step": 10716 + }, + { + "epoch": 0.9664968210307977, + "grad_norm": 0.645287341904407, + "learning_rate": 1.1732799619516897e-08, + "loss": 0.8191, + "step": 10717 + }, + { + "epoch": 0.9665870045542679, + "grad_norm": 1.7490352287299706, + "learning_rate": 1.1669696765652659e-08, + "loss": 0.9612, + "step": 10718 + }, + { + "epoch": 0.9666771880777382, + "grad_norm": 3.229283993138015, + "learning_rate": 1.1606763567093336e-08, + "loss": 1.0062, + "step": 10719 + }, + { + "epoch": 0.9667673716012085, + "grad_norm": 0.6806500514160532, + "learning_rate": 1.1544000029208857e-08, + "loss": 0.8101, + "step": 10720 + }, + { + "epoch": 0.9668575551246787, + "grad_norm": 1.8149486969561999, + "learning_rate": 1.148140615735449e-08, + "loss": 0.8395, + "step": 10721 + }, + { + "epoch": 0.966947738648149, + "grad_norm": 1.42824343935199, + "learning_rate": 1.1418981956871076e-08, + "loss": 0.9529, + "step": 10722 + }, + { + "epoch": 0.9670379221716192, + "grad_norm": 2.144638976012714, + "learning_rate": 1.1356727433085245e-08, + "loss": 0.965, + "step": 10723 + }, + { + "epoch": 0.9671281056950896, + "grad_norm": 1.4940204646842237, + "learning_rate": 1.1294642591308524e-08, + "loss": 1.0025, + "step": 10724 + }, + { + "epoch": 0.9672182892185598, + "grad_norm": 1.4175035959026603, + "learning_rate": 1.1232727436838452e-08, + "loss": 0.9647, + "step": 10725 + }, + { + "epoch": 0.96730847274203, + "grad_norm": 1.7092839441682537, + "learning_rate": 1.1170981974958138e-08, + "loss": 1.0199, + "step": 10726 + }, + { + "epoch": 0.9673986562655003, + "grad_norm": 1.6677355994196241, + "learning_rate": 1.1109406210936035e-08, + "loss": 1.037, + "step": 10727 + }, + { + "epoch": 0.9674888397889706, + "grad_norm": 2.3311555326819597, + "learning_rate": 1.1048000150025939e-08, + "loss": 1.0954, + "step": 10728 + }, + { + "epoch": 0.9675790233124408, + "grad_norm": 2.375052903238657, + "learning_rate": 1.0986763797467213e-08, + "loss": 0.8741, + "step": 10729 + }, + { + "epoch": 0.9676692068359111, + "grad_norm": 0.5813719580486062, + "learning_rate": 1.0925697158485459e-08, + "loss": 0.7811, + "step": 10730 + }, + { + "epoch": 0.9677593903593813, + "grad_norm": 1.5688687800087389, + "learning_rate": 1.0864800238290727e-08, + "loss": 0.9826, + "step": 10731 + }, + { + "epoch": 0.9678495738828516, + "grad_norm": 2.5425138051906346, + "learning_rate": 1.0804073042079309e-08, + "loss": 0.9441, + "step": 10732 + }, + { + "epoch": 0.9679397574063219, + "grad_norm": 0.6405062681750532, + "learning_rate": 1.0743515575032392e-08, + "loss": 0.8259, + "step": 10733 + }, + { + "epoch": 0.9680299409297921, + "grad_norm": 1.5554608285913942, + "learning_rate": 1.0683127842317619e-08, + "loss": 0.9878, + "step": 10734 + }, + { + "epoch": 0.9681201244532623, + "grad_norm": 1.8971325192050805, + "learning_rate": 1.0622909849087314e-08, + "loss": 1.0101, + "step": 10735 + }, + { + "epoch": 0.9682103079767327, + "grad_norm": 1.4990949284408004, + "learning_rate": 1.0562861600479588e-08, + "loss": 1.0374, + "step": 10736 + }, + { + "epoch": 0.9683004915002029, + "grad_norm": 4.004704093526635, + "learning_rate": 1.0502983101618345e-08, + "loss": 1.0731, + "step": 10737 + }, + { + "epoch": 0.9683906750236732, + "grad_norm": 1.7225915459831453, + "learning_rate": 1.0443274357612386e-08, + "loss": 0.8558, + "step": 10738 + }, + { + "epoch": 0.9684808585471434, + "grad_norm": 1.566213467499944, + "learning_rate": 1.0383735373556524e-08, + "loss": 0.9903, + "step": 10739 + }, + { + "epoch": 0.9685710420706137, + "grad_norm": 1.3506758322764085, + "learning_rate": 1.0324366154531139e-08, + "loss": 1.0095, + "step": 10740 + }, + { + "epoch": 0.968661225594084, + "grad_norm": 1.7647165307490396, + "learning_rate": 1.0265166705601735e-08, + "loss": 0.9406, + "step": 10741 + }, + { + "epoch": 0.9687514091175542, + "grad_norm": 0.6741163660066295, + "learning_rate": 1.0206137031819606e-08, + "loss": 0.8481, + "step": 10742 + }, + { + "epoch": 0.9688415926410244, + "grad_norm": 1.4758373276096244, + "learning_rate": 1.0147277138221388e-08, + "loss": 0.9896, + "step": 10743 + }, + { + "epoch": 0.9689317761644948, + "grad_norm": 1.6695861835929824, + "learning_rate": 1.0088587029829287e-08, + "loss": 1.0115, + "step": 10744 + }, + { + "epoch": 0.969021959687965, + "grad_norm": 1.444351949338647, + "learning_rate": 1.003006671165152e-08, + "loss": 1.078, + "step": 10745 + }, + { + "epoch": 0.9691121432114352, + "grad_norm": 1.8272641755416805, + "learning_rate": 9.971716188680978e-09, + "loss": 0.9701, + "step": 10746 + }, + { + "epoch": 0.9692023267349056, + "grad_norm": 1.6881680755599313, + "learning_rate": 9.91353546589635e-09, + "loss": 0.9382, + "step": 10747 + }, + { + "epoch": 0.9692925102583758, + "grad_norm": 1.4664510225219554, + "learning_rate": 9.855524548262106e-09, + "loss": 0.9143, + "step": 10748 + }, + { + "epoch": 0.969382693781846, + "grad_norm": 1.642237279774471, + "learning_rate": 9.797683440728288e-09, + "loss": 0.9199, + "step": 10749 + }, + { + "epoch": 0.9694728773053163, + "grad_norm": 1.4741654087720428, + "learning_rate": 9.740012148229836e-09, + "loss": 0.8912, + "step": 10750 + }, + { + "epoch": 0.9695630608287866, + "grad_norm": 1.4315358705173549, + "learning_rate": 9.682510675687705e-09, + "loss": 0.928, + "step": 10751 + }, + { + "epoch": 0.9696532443522569, + "grad_norm": 1.5897137263035443, + "learning_rate": 9.625179028008191e-09, + "loss": 0.9037, + "step": 10752 + }, + { + "epoch": 0.9697434278757271, + "grad_norm": 1.7175764759256213, + "learning_rate": 9.568017210083379e-09, + "loss": 0.8221, + "step": 10753 + }, + { + "epoch": 0.9698336113991973, + "grad_norm": 6.683239586886918, + "learning_rate": 9.511025226790259e-09, + "loss": 0.9531, + "step": 10754 + }, + { + "epoch": 0.9699237949226677, + "grad_norm": 1.4361666424949644, + "learning_rate": 9.454203082992052e-09, + "loss": 0.9503, + "step": 10755 + }, + { + "epoch": 0.9700139784461379, + "grad_norm": 2.0266231954245737, + "learning_rate": 9.3975507835371e-09, + "loss": 0.8869, + "step": 10756 + }, + { + "epoch": 0.9701041619696081, + "grad_norm": 0.6259783478471883, + "learning_rate": 9.341068333259094e-09, + "loss": 0.8022, + "step": 10757 + }, + { + "epoch": 0.9701943454930784, + "grad_norm": 1.5377721654156218, + "learning_rate": 9.28475573697729e-09, + "loss": 0.9533, + "step": 10758 + }, + { + "epoch": 0.9702845290165487, + "grad_norm": 1.5472504901795836, + "learning_rate": 9.228612999497177e-09, + "loss": 0.9794, + "step": 10759 + }, + { + "epoch": 0.970374712540019, + "grad_norm": 2.7027738486410944, + "learning_rate": 9.172640125608478e-09, + "loss": 0.9217, + "step": 10760 + }, + { + "epoch": 0.9704648960634892, + "grad_norm": 1.766210408344082, + "learning_rate": 9.116837120087817e-09, + "loss": 0.9936, + "step": 10761 + }, + { + "epoch": 0.9705550795869594, + "grad_norm": 2.8334083734186786, + "learning_rate": 9.061203987695832e-09, + "loss": 1.0429, + "step": 10762 + }, + { + "epoch": 0.9706452631104298, + "grad_norm": 1.7944411006433114, + "learning_rate": 9.005740733180055e-09, + "loss": 0.9219, + "step": 10763 + }, + { + "epoch": 0.9707354466339, + "grad_norm": 1.5763565579243743, + "learning_rate": 8.950447361272483e-09, + "loss": 1.0577, + "step": 10764 + }, + { + "epoch": 0.9708256301573702, + "grad_norm": 1.8606379090855363, + "learning_rate": 8.895323876691784e-09, + "loss": 0.9827, + "step": 10765 + }, + { + "epoch": 0.9709158136808405, + "grad_norm": 2.2404229534567746, + "learning_rate": 8.840370284140419e-09, + "loss": 1.0283, + "step": 10766 + }, + { + "epoch": 0.9710059972043108, + "grad_norm": 1.6006703194514715, + "learning_rate": 8.78558658830797e-09, + "loss": 0.896, + "step": 10767 + }, + { + "epoch": 0.971096180727781, + "grad_norm": 0.7008743292361568, + "learning_rate": 8.730972793868696e-09, + "loss": 0.8146, + "step": 10768 + }, + { + "epoch": 0.9711863642512513, + "grad_norm": 1.9539867483683384, + "learning_rate": 8.67652890548265e-09, + "loss": 0.9387, + "step": 10769 + }, + { + "epoch": 0.9712765477747216, + "grad_norm": 1.419249741517111, + "learning_rate": 8.622254927795004e-09, + "loss": 0.9594, + "step": 10770 + }, + { + "epoch": 0.9713667312981918, + "grad_norm": 1.5586716864276349, + "learning_rate": 8.568150865436941e-09, + "loss": 0.9261, + "step": 10771 + }, + { + "epoch": 0.9714569148216621, + "grad_norm": 1.3492468693491741, + "learning_rate": 8.514216723024991e-09, + "loss": 0.985, + "step": 10772 + }, + { + "epoch": 0.9715470983451323, + "grad_norm": 2.0934150769538573, + "learning_rate": 8.460452505161031e-09, + "loss": 0.8945, + "step": 10773 + }, + { + "epoch": 0.9716372818686027, + "grad_norm": 2.4869160642490766, + "learning_rate": 8.4068582164325e-09, + "loss": 1.0019, + "step": 10774 + }, + { + "epoch": 0.9717274653920729, + "grad_norm": 1.5587521427142872, + "learning_rate": 8.353433861412406e-09, + "loss": 0.9556, + "step": 10775 + }, + { + "epoch": 0.9718176489155431, + "grad_norm": 1.5929319488779499, + "learning_rate": 8.300179444658883e-09, + "loss": 0.935, + "step": 10776 + }, + { + "epoch": 0.9719078324390134, + "grad_norm": 3.987097554242115, + "learning_rate": 8.247094970716296e-09, + "loss": 1.0716, + "step": 10777 + }, + { + "epoch": 0.9719980159624837, + "grad_norm": 1.5125927037604943, + "learning_rate": 8.19418044411413e-09, + "loss": 1.0123, + "step": 10778 + }, + { + "epoch": 0.9720881994859539, + "grad_norm": 1.4852335805892325, + "learning_rate": 8.141435869367219e-09, + "loss": 0.977, + "step": 10779 + }, + { + "epoch": 0.9721783830094242, + "grad_norm": 1.8388846908603371, + "learning_rate": 8.088861250975742e-09, + "loss": 1.0115, + "step": 10780 + }, + { + "epoch": 0.9722685665328944, + "grad_norm": 1.6730956853123176, + "learning_rate": 8.036456593426111e-09, + "loss": 0.9416, + "step": 10781 + }, + { + "epoch": 0.9723587500563647, + "grad_norm": 1.6180718993679013, + "learning_rate": 7.984221901189415e-09, + "loss": 1.0113, + "step": 10782 + }, + { + "epoch": 0.972448933579835, + "grad_norm": 1.6225063557786874, + "learning_rate": 7.932157178722976e-09, + "loss": 0.8998, + "step": 10783 + }, + { + "epoch": 0.9725391171033052, + "grad_norm": 1.4427054321406787, + "learning_rate": 7.880262430468799e-09, + "loss": 0.9442, + "step": 10784 + }, + { + "epoch": 0.9726293006267754, + "grad_norm": 1.662691724807631, + "learning_rate": 7.828537660855339e-09, + "loss": 0.909, + "step": 10785 + }, + { + "epoch": 0.9727194841502458, + "grad_norm": 2.4639294072541698, + "learning_rate": 7.776982874295512e-09, + "loss": 0.9897, + "step": 10786 + }, + { + "epoch": 0.972809667673716, + "grad_norm": 2.0944471195090313, + "learning_rate": 7.725598075188688e-09, + "loss": 0.883, + "step": 10787 + }, + { + "epoch": 0.9728998511971862, + "grad_norm": 1.4680964050496321, + "learning_rate": 7.674383267918916e-09, + "loss": 0.9267, + "step": 10788 + }, + { + "epoch": 0.9729900347206565, + "grad_norm": 1.521438460826394, + "learning_rate": 7.623338456856476e-09, + "loss": 0.9612, + "step": 10789 + }, + { + "epoch": 0.9730802182441268, + "grad_norm": 1.457651124058495, + "learning_rate": 7.572463646356554e-09, + "loss": 0.8667, + "step": 10790 + }, + { + "epoch": 0.9731704017675971, + "grad_norm": 1.642402557896714, + "learning_rate": 7.521758840760339e-09, + "loss": 0.9423, + "step": 10791 + }, + { + "epoch": 0.9732605852910673, + "grad_norm": 2.1844894401031376, + "learning_rate": 7.471224044393931e-09, + "loss": 0.9302, + "step": 10792 + }, + { + "epoch": 0.9733507688145376, + "grad_norm": 1.4980306413111613, + "learning_rate": 7.420859261569434e-09, + "loss": 1.0325, + "step": 10793 + }, + { + "epoch": 0.9734409523380079, + "grad_norm": 2.4554757681590664, + "learning_rate": 7.370664496584078e-09, + "loss": 1.0324, + "step": 10794 + }, + { + "epoch": 0.9735311358614781, + "grad_norm": 1.569673626500646, + "learning_rate": 7.3206397537211026e-09, + "loss": 0.9023, + "step": 10795 + }, + { + "epoch": 0.9736213193849483, + "grad_norm": 2.0259080334470645, + "learning_rate": 7.270785037248428e-09, + "loss": 0.9409, + "step": 10796 + }, + { + "epoch": 0.9737115029084187, + "grad_norm": 1.7031978626589, + "learning_rate": 7.221100351420428e-09, + "loss": 0.9373, + "step": 10797 + }, + { + "epoch": 0.9738016864318889, + "grad_norm": 1.7598636443246394, + "learning_rate": 7.171585700475935e-09, + "loss": 0.9015, + "step": 10798 + }, + { + "epoch": 0.9738918699553591, + "grad_norm": 2.123185673504511, + "learning_rate": 7.122241088640235e-09, + "loss": 1.0303, + "step": 10799 + }, + { + "epoch": 0.9739820534788294, + "grad_norm": 1.6845350001960113, + "learning_rate": 7.073066520123516e-09, + "loss": 1.0987, + "step": 10800 + }, + { + "epoch": 0.9740722370022997, + "grad_norm": 1.9957388545578905, + "learning_rate": 7.0240619991217555e-09, + "loss": 0.8699, + "step": 10801 + }, + { + "epoch": 0.97416242052577, + "grad_norm": 1.6328371767068468, + "learning_rate": 6.975227529816052e-09, + "loss": 1.0498, + "step": 10802 + }, + { + "epoch": 0.9742526040492402, + "grad_norm": 1.5791541039496417, + "learning_rate": 6.926563116373296e-09, + "loss": 1.0192, + "step": 10803 + }, + { + "epoch": 0.9743427875727104, + "grad_norm": 1.5074422114795372, + "learning_rate": 6.878068762945943e-09, + "loss": 0.9119, + "step": 10804 + }, + { + "epoch": 0.9744329710961808, + "grad_norm": 1.70751485368515, + "learning_rate": 6.829744473671794e-09, + "loss": 1.0109, + "step": 10805 + }, + { + "epoch": 0.974523154619651, + "grad_norm": 0.588855395349255, + "learning_rate": 6.781590252674219e-09, + "loss": 0.7158, + "step": 10806 + }, + { + "epoch": 0.9746133381431212, + "grad_norm": 1.201451909366846, + "learning_rate": 6.733606104061484e-09, + "loss": 0.9829, + "step": 10807 + }, + { + "epoch": 0.9747035216665915, + "grad_norm": 1.3685139352341926, + "learning_rate": 6.6857920319283165e-09, + "loss": 0.9191, + "step": 10808 + }, + { + "epoch": 0.9747937051900618, + "grad_norm": 1.6654431700090842, + "learning_rate": 6.638148040354563e-09, + "loss": 0.947, + "step": 10809 + }, + { + "epoch": 0.974883888713532, + "grad_norm": 2.7015730972920586, + "learning_rate": 6.590674133405194e-09, + "loss": 0.9324, + "step": 10810 + }, + { + "epoch": 0.9749740722370023, + "grad_norm": 1.4317680187713426, + "learning_rate": 6.5433703151311914e-09, + "loss": 0.9885, + "step": 10811 + }, + { + "epoch": 0.9750642557604725, + "grad_norm": 2.2207195512873685, + "learning_rate": 6.49623658956866e-09, + "loss": 0.9721, + "step": 10812 + }, + { + "epoch": 0.9751544392839429, + "grad_norm": 1.9557718271478175, + "learning_rate": 6.44927296073905e-09, + "loss": 0.85, + "step": 10813 + }, + { + "epoch": 0.9752446228074131, + "grad_norm": 1.4847923896623292, + "learning_rate": 6.402479432649821e-09, + "loss": 1.0598, + "step": 10814 + }, + { + "epoch": 0.9753348063308833, + "grad_norm": 1.5081759790498452, + "learning_rate": 6.355856009293781e-09, + "loss": 0.9688, + "step": 10815 + }, + { + "epoch": 0.9754249898543537, + "grad_norm": 1.9469034820481612, + "learning_rate": 6.3094026946488575e-09, + "loss": 0.9424, + "step": 10816 + }, + { + "epoch": 0.9755151733778239, + "grad_norm": 1.6050677390376438, + "learning_rate": 6.2631194926787704e-09, + "loss": 0.974, + "step": 10817 + }, + { + "epoch": 0.9756053569012941, + "grad_norm": 1.7783433052767763, + "learning_rate": 6.217006407332581e-09, + "loss": 0.8723, + "step": 10818 + }, + { + "epoch": 0.9756955404247644, + "grad_norm": 2.6907601210706775, + "learning_rate": 6.1710634425453654e-09, + "loss": 0.9192, + "step": 10819 + }, + { + "epoch": 0.9757857239482347, + "grad_norm": 1.5574332675751565, + "learning_rate": 6.1252906022366544e-09, + "loss": 0.9954, + "step": 10820 + }, + { + "epoch": 0.9758759074717049, + "grad_norm": 1.5751679921249406, + "learning_rate": 6.079687890312213e-09, + "loss": 0.9526, + "step": 10821 + }, + { + "epoch": 0.9759660909951752, + "grad_norm": 2.48419025621941, + "learning_rate": 6.034255310663372e-09, + "loss": 0.9538, + "step": 10822 + }, + { + "epoch": 0.9760562745186454, + "grad_norm": 1.735523084451258, + "learning_rate": 5.988992867166143e-09, + "loss": 1.066, + "step": 10823 + }, + { + "epoch": 0.9761464580421157, + "grad_norm": 1.8890801415514968, + "learning_rate": 5.943900563682991e-09, + "loss": 1.0121, + "step": 10824 + }, + { + "epoch": 0.976236641565586, + "grad_norm": 0.7209343267998477, + "learning_rate": 5.898978404061506e-09, + "loss": 0.8933, + "step": 10825 + }, + { + "epoch": 0.9763268250890562, + "grad_norm": 4.087467643773049, + "learning_rate": 5.85422639213462e-09, + "loss": 0.9692, + "step": 10826 + }, + { + "epoch": 0.9764170086125264, + "grad_norm": 1.4871801272836955, + "learning_rate": 5.809644531720614e-09, + "loss": 0.9845, + "step": 10827 + }, + { + "epoch": 0.9765071921359968, + "grad_norm": 1.6598478861134838, + "learning_rate": 5.765232826623556e-09, + "loss": 0.9995, + "step": 10828 + }, + { + "epoch": 0.976597375659467, + "grad_norm": 2.5694088724325974, + "learning_rate": 5.720991280633081e-09, + "loss": 0.9529, + "step": 10829 + }, + { + "epoch": 0.9766875591829373, + "grad_norm": 1.3664575862674073, + "learning_rate": 5.676919897523724e-09, + "loss": 0.955, + "step": 10830 + }, + { + "epoch": 0.9767777427064075, + "grad_norm": 1.6505194799919836, + "learning_rate": 5.633018681056256e-09, + "loss": 0.9778, + "step": 10831 + }, + { + "epoch": 0.9768679262298778, + "grad_norm": 1.775565597254866, + "learning_rate": 5.589287634976569e-09, + "loss": 0.9739, + "step": 10832 + }, + { + "epoch": 0.9769581097533481, + "grad_norm": 1.3233587572616665, + "learning_rate": 5.5457267630159014e-09, + "loss": 0.9465, + "step": 10833 + }, + { + "epoch": 0.9770482932768183, + "grad_norm": 1.588920606431448, + "learning_rate": 5.5023360688910555e-09, + "loss": 1.008, + "step": 10834 + }, + { + "epoch": 0.9771384768002885, + "grad_norm": 1.4140545333722192, + "learning_rate": 5.459115556304183e-09, + "loss": 1.0181, + "step": 10835 + }, + { + "epoch": 0.9772286603237589, + "grad_norm": 2.3621263440650093, + "learning_rate": 5.416065228943889e-09, + "loss": 0.9029, + "step": 10836 + }, + { + "epoch": 0.9773188438472291, + "grad_norm": 1.641791737176524, + "learning_rate": 5.373185090482568e-09, + "loss": 0.9247, + "step": 10837 + }, + { + "epoch": 0.9774090273706993, + "grad_norm": 1.6104047129172177, + "learning_rate": 5.330475144579516e-09, + "loss": 0.9501, + "step": 10838 + }, + { + "epoch": 0.9774992108941697, + "grad_norm": 1.4647371380474516, + "learning_rate": 5.2879353948787065e-09, + "loss": 1.0579, + "step": 10839 + }, + { + "epoch": 0.9775893944176399, + "grad_norm": 2.27466437410699, + "learning_rate": 5.245565845010125e-09, + "loss": 1.0318, + "step": 10840 + }, + { + "epoch": 0.9776795779411102, + "grad_norm": 1.7525347336443877, + "learning_rate": 5.2033664985886575e-09, + "loss": 0.9043, + "step": 10841 + }, + { + "epoch": 0.9777697614645804, + "grad_norm": 1.7735688417171223, + "learning_rate": 5.161337359215201e-09, + "loss": 1.1164, + "step": 10842 + }, + { + "epoch": 0.9778599449880507, + "grad_norm": 1.5257349932134319, + "learning_rate": 5.119478430475999e-09, + "loss": 0.9839, + "step": 10843 + }, + { + "epoch": 0.977950128511521, + "grad_norm": 1.8148373627108934, + "learning_rate": 5.077789715942416e-09, + "loss": 0.9168, + "step": 10844 + }, + { + "epoch": 0.9780403120349912, + "grad_norm": 1.7283283714117754, + "learning_rate": 5.036271219171606e-09, + "loss": 0.9965, + "step": 10845 + }, + { + "epoch": 0.9781304955584614, + "grad_norm": 1.8721549383626008, + "learning_rate": 4.994922943706514e-09, + "loss": 1.0094, + "step": 10846 + }, + { + "epoch": 0.9782206790819318, + "grad_norm": 1.7775696327959916, + "learning_rate": 4.953744893074763e-09, + "loss": 0.9589, + "step": 10847 + }, + { + "epoch": 0.978310862605402, + "grad_norm": 2.0407824515431248, + "learning_rate": 4.912737070789985e-09, + "loss": 0.9001, + "step": 10848 + }, + { + "epoch": 0.9784010461288722, + "grad_norm": 1.7715817459414906, + "learning_rate": 4.871899480351604e-09, + "loss": 1.0434, + "step": 10849 + }, + { + "epoch": 0.9784912296523425, + "grad_norm": 1.781173570565953, + "learning_rate": 4.831232125243501e-09, + "loss": 0.9877, + "step": 10850 + }, + { + "epoch": 0.9785814131758128, + "grad_norm": 1.514905463816319, + "learning_rate": 4.7907350089360086e-09, + "loss": 1.0761, + "step": 10851 + }, + { + "epoch": 0.978671596699283, + "grad_norm": 2.4122304538500647, + "learning_rate": 4.750408134884365e-09, + "loss": 0.8702, + "step": 10852 + }, + { + "epoch": 0.9787617802227533, + "grad_norm": 1.4305243205803426, + "learning_rate": 4.710251506529816e-09, + "loss": 0.8821, + "step": 10853 + }, + { + "epoch": 0.9788519637462235, + "grad_norm": 2.42373566063057, + "learning_rate": 4.6702651272982894e-09, + "loss": 0.9867, + "step": 10854 + }, + { + "epoch": 0.9789421472696939, + "grad_norm": 1.6463402156687266, + "learning_rate": 4.630449000602166e-09, + "loss": 0.9456, + "step": 10855 + }, + { + "epoch": 0.9790323307931641, + "grad_norm": 1.8198774742970631, + "learning_rate": 4.590803129838283e-09, + "loss": 1.0325, + "step": 10856 + }, + { + "epoch": 0.9791225143166343, + "grad_norm": 2.4294767135003434, + "learning_rate": 4.551327518389714e-09, + "loss": 0.9806, + "step": 10857 + }, + { + "epoch": 0.9792126978401046, + "grad_norm": 2.188542798668558, + "learning_rate": 4.512022169624652e-09, + "loss": 0.9652, + "step": 10858 + }, + { + "epoch": 0.9793028813635749, + "grad_norm": 1.4626634164134809, + "learning_rate": 4.472887086896637e-09, + "loss": 0.9473, + "step": 10859 + }, + { + "epoch": 0.9793930648870451, + "grad_norm": 1.687529669606796, + "learning_rate": 4.433922273545443e-09, + "loss": 0.9207, + "step": 10860 + }, + { + "epoch": 0.9794832484105154, + "grad_norm": 0.6294630204277888, + "learning_rate": 4.395127732895299e-09, + "loss": 0.8182, + "step": 10861 + }, + { + "epoch": 0.9795734319339856, + "grad_norm": 2.3075675942882956, + "learning_rate": 4.356503468256445e-09, + "loss": 1.051, + "step": 10862 + }, + { + "epoch": 0.979663615457456, + "grad_norm": 1.907582177108146, + "learning_rate": 4.318049482924913e-09, + "loss": 0.9849, + "step": 10863 + }, + { + "epoch": 0.9797537989809262, + "grad_norm": 1.6112031645437868, + "learning_rate": 4.279765780181188e-09, + "loss": 0.9414, + "step": 10864 + }, + { + "epoch": 0.9798439825043964, + "grad_norm": 2.94072173361833, + "learning_rate": 4.241652363291992e-09, + "loss": 0.988, + "step": 10865 + }, + { + "epoch": 0.9799341660278668, + "grad_norm": 3.3606414536847886, + "learning_rate": 4.203709235509834e-09, + "loss": 0.9481, + "step": 10866 + }, + { + "epoch": 0.980024349551337, + "grad_norm": 2.0887329572054236, + "learning_rate": 4.165936400071679e-09, + "loss": 0.9921, + "step": 10867 + }, + { + "epoch": 0.9801145330748072, + "grad_norm": 1.4806907306217412, + "learning_rate": 4.12833386020095e-09, + "loss": 0.9226, + "step": 10868 + }, + { + "epoch": 0.9802047165982775, + "grad_norm": 2.033473045468749, + "learning_rate": 4.090901619105746e-09, + "loss": 1.0143, + "step": 10869 + }, + { + "epoch": 0.9802949001217478, + "grad_norm": 1.5269866291702336, + "learning_rate": 4.053639679980181e-09, + "loss": 0.8652, + "step": 10870 + }, + { + "epoch": 0.980385083645218, + "grad_norm": 1.7596373855318606, + "learning_rate": 4.01654804600371e-09, + "loss": 0.977, + "step": 10871 + }, + { + "epoch": 0.9804752671686883, + "grad_norm": 1.4922010810593516, + "learning_rate": 3.9796267203409114e-09, + "loss": 0.9851, + "step": 10872 + }, + { + "epoch": 0.9805654506921585, + "grad_norm": 1.5268076860827535, + "learning_rate": 3.942875706142379e-09, + "loss": 0.9523, + "step": 10873 + }, + { + "epoch": 0.9806556342156288, + "grad_norm": 5.128081721019194, + "learning_rate": 3.906295006543825e-09, + "loss": 1.0335, + "step": 10874 + }, + { + "epoch": 0.9807458177390991, + "grad_norm": 1.8863167976296817, + "learning_rate": 3.8698846246665305e-09, + "loss": 0.9522, + "step": 10875 + }, + { + "epoch": 0.9808360012625693, + "grad_norm": 1.6083745717245006, + "learning_rate": 3.833644563617344e-09, + "loss": 0.9819, + "step": 10876 + }, + { + "epoch": 0.9809261847860395, + "grad_norm": 2.0785669989672106, + "learning_rate": 3.797574826488237e-09, + "loss": 0.9873, + "step": 10877 + }, + { + "epoch": 0.9810163683095099, + "grad_norm": 1.5534739142696552, + "learning_rate": 3.761675416356969e-09, + "loss": 0.9027, + "step": 10878 + }, + { + "epoch": 0.9811065518329801, + "grad_norm": 1.5245776482735258, + "learning_rate": 3.725946336286867e-09, + "loss": 0.8803, + "step": 10879 + }, + { + "epoch": 0.9811967353564504, + "grad_norm": 1.3049464168729012, + "learning_rate": 3.6903875893261604e-09, + "loss": 0.9461, + "step": 10880 + }, + { + "epoch": 0.9812869188799206, + "grad_norm": 2.0565271300963937, + "learning_rate": 3.6549991785093105e-09, + "loss": 0.9064, + "step": 10881 + }, + { + "epoch": 0.9813771024033909, + "grad_norm": 1.8175223910331169, + "learning_rate": 3.6197811068554575e-09, + "loss": 0.9631, + "step": 10882 + }, + { + "epoch": 0.9814672859268612, + "grad_norm": 1.3082904976718752, + "learning_rate": 3.584733377369975e-09, + "loss": 1.0597, + "step": 10883 + }, + { + "epoch": 0.9815574694503314, + "grad_norm": 1.9556492477186544, + "learning_rate": 3.549855993043138e-09, + "loss": 0.9513, + "step": 10884 + }, + { + "epoch": 0.9816476529738016, + "grad_norm": 1.628167732834829, + "learning_rate": 3.5151489568507887e-09, + "loss": 0.956, + "step": 10885 + }, + { + "epoch": 0.981737836497272, + "grad_norm": 1.9332669685198225, + "learning_rate": 3.4806122717545572e-09, + "loss": 0.9856, + "step": 10886 + }, + { + "epoch": 0.9818280200207422, + "grad_norm": 1.657017583939071, + "learning_rate": 3.446245940701198e-09, + "loss": 0.9369, + "step": 10887 + }, + { + "epoch": 0.9819182035442124, + "grad_norm": 1.5359242777289233, + "learning_rate": 3.41204996662281e-09, + "loss": 0.9146, + "step": 10888 + }, + { + "epoch": 0.9820083870676828, + "grad_norm": 1.366870516406443, + "learning_rate": 3.3780243524375028e-09, + "loss": 0.9749, + "step": 10889 + }, + { + "epoch": 0.982098570591153, + "grad_norm": 1.8702780074250949, + "learning_rate": 3.3441691010485107e-09, + "loss": 0.9993, + "step": 10890 + }, + { + "epoch": 0.9821887541146233, + "grad_norm": 1.7927024493951194, + "learning_rate": 3.3104842153444113e-09, + "loss": 1.0552, + "step": 10891 + }, + { + "epoch": 0.9822789376380935, + "grad_norm": 2.1010415945461287, + "learning_rate": 3.27696969819935e-09, + "loss": 0.9677, + "step": 10892 + }, + { + "epoch": 0.9823691211615638, + "grad_norm": 1.5282375520473481, + "learning_rate": 3.2436255524732615e-09, + "loss": 0.9318, + "step": 10893 + }, + { + "epoch": 0.9824593046850341, + "grad_norm": 1.3833329681411322, + "learning_rate": 3.210451781010759e-09, + "loss": 0.9557, + "step": 10894 + }, + { + "epoch": 0.9825494882085043, + "grad_norm": 1.4940479125136765, + "learning_rate": 3.1774483866426895e-09, + "loss": 0.8374, + "step": 10895 + }, + { + "epoch": 0.9826396717319745, + "grad_norm": 1.4798533587364822, + "learning_rate": 3.144615372185244e-09, + "loss": 0.9387, + "step": 10896 + }, + { + "epoch": 0.9827298552554449, + "grad_norm": 1.5653470613087765, + "learning_rate": 3.1119527404399604e-09, + "loss": 1.065, + "step": 10897 + }, + { + "epoch": 0.9828200387789151, + "grad_norm": 2.329160933467608, + "learning_rate": 3.0794604941932754e-09, + "loss": 0.9728, + "step": 10898 + }, + { + "epoch": 0.9829102223023853, + "grad_norm": 1.379786612902546, + "learning_rate": 3.0471386362180827e-09, + "loss": 0.87, + "step": 10899 + }, + { + "epoch": 0.9830004058258556, + "grad_norm": 1.4614327962995683, + "learning_rate": 3.0149871692719542e-09, + "loss": 1.024, + "step": 10900 + }, + { + "epoch": 0.9830905893493259, + "grad_norm": 1.5163456098409382, + "learning_rate": 2.9830060960984728e-09, + "loss": 0.9602, + "step": 10901 + }, + { + "epoch": 0.9831807728727961, + "grad_norm": 0.6548075718353361, + "learning_rate": 2.9511954194263442e-09, + "loss": 0.8211, + "step": 10902 + }, + { + "epoch": 0.9832709563962664, + "grad_norm": 1.5990207506685852, + "learning_rate": 2.9195551419698426e-09, + "loss": 0.9708, + "step": 10903 + }, + { + "epoch": 0.9833611399197366, + "grad_norm": 1.9561126441387389, + "learning_rate": 2.888085266428808e-09, + "loss": 0.9346, + "step": 10904 + }, + { + "epoch": 0.983451323443207, + "grad_norm": 1.7476193720439976, + "learning_rate": 2.8567857954882037e-09, + "loss": 0.9323, + "step": 10905 + }, + { + "epoch": 0.9835415069666772, + "grad_norm": 1.7821310441013056, + "learning_rate": 2.82565673181856e-09, + "loss": 1.027, + "step": 10906 + }, + { + "epoch": 0.9836316904901474, + "grad_norm": 1.6129098238348596, + "learning_rate": 2.7946980780764184e-09, + "loss": 0.9939, + "step": 10907 + }, + { + "epoch": 0.9837218740136177, + "grad_norm": 1.820187674896323, + "learning_rate": 2.763909836903e-09, + "loss": 1.0292, + "step": 10908 + }, + { + "epoch": 0.983812057537088, + "grad_norm": 1.6873322862712055, + "learning_rate": 2.7332920109255364e-09, + "loss": 0.8901, + "step": 10909 + }, + { + "epoch": 0.9839022410605582, + "grad_norm": 1.4478919852735173, + "learning_rate": 2.702844602756382e-09, + "loss": 0.9563, + "step": 10910 + }, + { + "epoch": 0.9839924245840285, + "grad_norm": 2.2221684515593574, + "learning_rate": 2.6725676149936814e-09, + "loss": 0.9478, + "step": 10911 + }, + { + "epoch": 0.9840826081074988, + "grad_norm": 1.6004295249421765, + "learning_rate": 2.642461050220479e-09, + "loss": 1.0432, + "step": 10912 + }, + { + "epoch": 0.984172791630969, + "grad_norm": 1.3611492986360791, + "learning_rate": 2.612524911005609e-09, + "loss": 0.9537, + "step": 10913 + }, + { + "epoch": 0.9842629751544393, + "grad_norm": 1.4006749586722036, + "learning_rate": 2.582759199903917e-09, + "loss": 0.8491, + "step": 10914 + }, + { + "epoch": 0.9843531586779095, + "grad_norm": 5.666146235783471, + "learning_rate": 2.553163919454704e-09, + "loss": 0.9314, + "step": 10915 + }, + { + "epoch": 0.9844433422013799, + "grad_norm": 1.5785253035964417, + "learning_rate": 2.523739072183506e-09, + "loss": 0.9537, + "step": 10916 + }, + { + "epoch": 0.9845335257248501, + "grad_norm": 3.881078198449873, + "learning_rate": 2.4944846606007597e-09, + "loss": 0.9343, + "step": 10917 + }, + { + "epoch": 0.9846237092483203, + "grad_norm": 0.612202759066237, + "learning_rate": 2.46540068720269e-09, + "loss": 0.8525, + "step": 10918 + }, + { + "epoch": 0.9847138927717906, + "grad_norm": 1.4121525126050698, + "learning_rate": 2.4364871544708674e-09, + "loss": 1.0094, + "step": 10919 + }, + { + "epoch": 0.9848040762952609, + "grad_norm": 1.6940683466872524, + "learning_rate": 2.4077440648726523e-09, + "loss": 0.9234, + "step": 10920 + }, + { + "epoch": 0.9848942598187311, + "grad_norm": 1.50698078519928, + "learning_rate": 2.379171420860082e-09, + "loss": 1.036, + "step": 10921 + }, + { + "epoch": 0.9849844433422014, + "grad_norm": 1.6274154024653313, + "learning_rate": 2.3507692248714296e-09, + "loss": 0.8923, + "step": 10922 + }, + { + "epoch": 0.9850746268656716, + "grad_norm": 1.8179117506113613, + "learning_rate": 2.322537479330089e-09, + "loss": 0.9878, + "step": 10923 + }, + { + "epoch": 0.9851648103891419, + "grad_norm": 3.297242926004382, + "learning_rate": 2.2944761866450223e-09, + "loss": 1.0491, + "step": 10924 + }, + { + "epoch": 0.9852549939126122, + "grad_norm": 1.7683428211419037, + "learning_rate": 2.266585349210315e-09, + "loss": 0.9554, + "step": 10925 + }, + { + "epoch": 0.9853451774360824, + "grad_norm": 2.0848354900419195, + "learning_rate": 2.2388649694060623e-09, + "loss": 0.9274, + "step": 10926 + }, + { + "epoch": 0.9854353609595526, + "grad_norm": 1.4197947617084505, + "learning_rate": 2.211315049597262e-09, + "loss": 1.0295, + "step": 10927 + }, + { + "epoch": 0.985525544483023, + "grad_norm": 1.5254950298377148, + "learning_rate": 2.1839355921349224e-09, + "loss": 1.0088, + "step": 10928 + }, + { + "epoch": 0.9856157280064932, + "grad_norm": 2.3679070213274023, + "learning_rate": 2.156726599354952e-09, + "loss": 0.9647, + "step": 10929 + }, + { + "epoch": 0.9857059115299635, + "grad_norm": 2.1229748893542895, + "learning_rate": 2.129688073578828e-09, + "loss": 0.9943, + "step": 10930 + }, + { + "epoch": 0.9857960950534337, + "grad_norm": 1.7403218496752701, + "learning_rate": 2.1028200171142597e-09, + "loss": 1.1075, + "step": 10931 + }, + { + "epoch": 0.985886278576904, + "grad_norm": 1.547442399587624, + "learning_rate": 2.076122432253191e-09, + "loss": 0.9529, + "step": 10932 + }, + { + "epoch": 0.9859764621003743, + "grad_norm": 1.8752345892066105, + "learning_rate": 2.0495953212738005e-09, + "loss": 1.012, + "step": 10933 + }, + { + "epoch": 0.9860666456238445, + "grad_norm": 2.076915991599884, + "learning_rate": 2.0232386864396102e-09, + "loss": 1.0555, + "step": 10934 + }, + { + "epoch": 0.9861568291473148, + "grad_norm": 1.7189010553738349, + "learning_rate": 1.9970525299992656e-09, + "loss": 0.9976, + "step": 10935 + }, + { + "epoch": 0.9862470126707851, + "grad_norm": 1.6278438625584275, + "learning_rate": 1.9710368541874245e-09, + "loss": 0.9348, + "step": 10936 + }, + { + "epoch": 0.9863371961942553, + "grad_norm": 0.5908870690833892, + "learning_rate": 1.945191661223644e-09, + "loss": 0.8239, + "step": 10937 + }, + { + "epoch": 0.9864273797177255, + "grad_norm": 2.8102567415613846, + "learning_rate": 1.9195169533132714e-09, + "loss": 1.0204, + "step": 10938 + }, + { + "epoch": 0.9865175632411959, + "grad_norm": 0.5853217895389787, + "learning_rate": 1.894012732646999e-09, + "loss": 0.8306, + "step": 10939 + }, + { + "epoch": 0.9866077467646661, + "grad_norm": 1.6263597374850007, + "learning_rate": 1.8686790014010854e-09, + "loss": 0.9198, + "step": 10940 + }, + { + "epoch": 0.9866979302881363, + "grad_norm": 0.6057895917812537, + "learning_rate": 1.8435157617369134e-09, + "loss": 0.7962, + "step": 10941 + }, + { + "epoch": 0.9867881138116066, + "grad_norm": 1.6714126803326235, + "learning_rate": 1.818523015801876e-09, + "loss": 1.1123, + "step": 10942 + }, + { + "epoch": 0.9868782973350769, + "grad_norm": 1.687269548292439, + "learning_rate": 1.7937007657282677e-09, + "loss": 0.9327, + "step": 10943 + }, + { + "epoch": 0.9869684808585472, + "grad_norm": 1.5524512842616007, + "learning_rate": 1.7690490136341718e-09, + "loss": 0.9153, + "step": 10944 + }, + { + "epoch": 0.9870586643820174, + "grad_norm": 1.5175465466369449, + "learning_rate": 1.744567761622795e-09, + "loss": 0.9697, + "step": 10945 + }, + { + "epoch": 0.9871488479054876, + "grad_norm": 1.46674169037856, + "learning_rate": 1.7202570117831327e-09, + "loss": 0.9648, + "step": 10946 + }, + { + "epoch": 0.987239031428958, + "grad_norm": 2.1752326314759065, + "learning_rate": 1.696116766189526e-09, + "loss": 0.9525, + "step": 10947 + }, + { + "epoch": 0.9873292149524282, + "grad_norm": 1.6117563499395418, + "learning_rate": 1.6721470269021042e-09, + "loss": 1.0132, + "step": 10948 + }, + { + "epoch": 0.9874193984758984, + "grad_norm": 3.840206839236037, + "learning_rate": 1.6483477959654546e-09, + "loss": 0.9454, + "step": 10949 + }, + { + "epoch": 0.9875095819993687, + "grad_norm": 2.3708452726129434, + "learning_rate": 1.6247190754106187e-09, + "loss": 0.9533, + "step": 10950 + }, + { + "epoch": 0.987599765522839, + "grad_norm": 2.300153300110671, + "learning_rate": 1.6012608672537619e-09, + "loss": 0.9389, + "step": 10951 + }, + { + "epoch": 0.9876899490463092, + "grad_norm": 1.8186203454854986, + "learning_rate": 1.5779731734963942e-09, + "loss": 0.9778, + "step": 10952 + }, + { + "epoch": 0.9877801325697795, + "grad_norm": 2.034910454431878, + "learning_rate": 1.5548559961253705e-09, + "loss": 1.0312, + "step": 10953 + }, + { + "epoch": 0.9878703160932497, + "grad_norm": 1.51786826955099, + "learning_rate": 1.5319093371135573e-09, + "loss": 0.897, + "step": 10954 + }, + { + "epoch": 0.9879604996167201, + "grad_norm": 1.5905939155344244, + "learning_rate": 1.5091331984184997e-09, + "loss": 1.0638, + "step": 10955 + }, + { + "epoch": 0.9880506831401903, + "grad_norm": 1.653475182406137, + "learning_rate": 1.486527581983754e-09, + "loss": 0.9719, + "step": 10956 + }, + { + "epoch": 0.9881408666636605, + "grad_norm": 1.6738721381046104, + "learning_rate": 1.4640924897382223e-09, + "loss": 1.0295, + "step": 10957 + }, + { + "epoch": 0.9882310501871309, + "grad_norm": 1.4889231021154388, + "learning_rate": 1.4418279235961506e-09, + "loss": 1.0128, + "step": 10958 + }, + { + "epoch": 0.9883212337106011, + "grad_norm": 1.646618748223751, + "learning_rate": 1.4197338854573526e-09, + "loss": 0.8773, + "step": 10959 + }, + { + "epoch": 0.9884114172340713, + "grad_norm": 1.8213088292744362, + "learning_rate": 1.3978103772067651e-09, + "loss": 0.9301, + "step": 10960 + }, + { + "epoch": 0.9885016007575416, + "grad_norm": 1.8689733926706893, + "learning_rate": 1.3760574007153358e-09, + "loss": 0.8946, + "step": 10961 + }, + { + "epoch": 0.9885917842810119, + "grad_norm": 2.24802280018739, + "learning_rate": 1.3544749578389137e-09, + "loss": 1.0541, + "step": 10962 + }, + { + "epoch": 0.9886819678044821, + "grad_norm": 1.6958002188758785, + "learning_rate": 1.3330630504189143e-09, + "loss": 1.0218, + "step": 10963 + }, + { + "epoch": 0.9887721513279524, + "grad_norm": 1.8393937064647694, + "learning_rate": 1.3118216802827652e-09, + "loss": 0.9437, + "step": 10964 + }, + { + "epoch": 0.9888623348514226, + "grad_norm": 1.7195940199491344, + "learning_rate": 1.2907508492425722e-09, + "loss": 0.9416, + "step": 10965 + }, + { + "epoch": 0.988952518374893, + "grad_norm": 1.5336815348844612, + "learning_rate": 1.2698505590962305e-09, + "loss": 0.9362, + "step": 10966 + }, + { + "epoch": 0.9890427018983632, + "grad_norm": 1.4496513605269499, + "learning_rate": 1.2491208116272022e-09, + "loss": 0.9642, + "step": 10967 + }, + { + "epoch": 0.9891328854218334, + "grad_norm": 1.5079468187977132, + "learning_rate": 1.2285616086040728e-09, + "loss": 1.0048, + "step": 10968 + }, + { + "epoch": 0.9892230689453037, + "grad_norm": 1.2720087848006099, + "learning_rate": 1.2081729517812167e-09, + "loss": 1.0359, + "step": 10969 + }, + { + "epoch": 0.989313252468774, + "grad_norm": 1.4183954274757382, + "learning_rate": 1.1879548428983533e-09, + "loss": 0.9691, + "step": 10970 + }, + { + "epoch": 0.9894034359922442, + "grad_norm": 1.8006166064282916, + "learning_rate": 1.167907283680547e-09, + "loss": 0.9187, + "step": 10971 + }, + { + "epoch": 0.9894936195157145, + "grad_norm": 1.6559072895681382, + "learning_rate": 1.1480302758382077e-09, + "loss": 0.9181, + "step": 10972 + }, + { + "epoch": 0.9895838030391847, + "grad_norm": 1.5589374364077426, + "learning_rate": 1.1283238210675338e-09, + "loss": 0.9657, + "step": 10973 + }, + { + "epoch": 0.989673986562655, + "grad_norm": 2.829137294281601, + "learning_rate": 1.1087879210498474e-09, + "loss": 0.8987, + "step": 10974 + }, + { + "epoch": 0.9897641700861253, + "grad_norm": 1.5550922269548282, + "learning_rate": 1.0894225774522592e-09, + "loss": 1.0185, + "step": 10975 + }, + { + "epoch": 0.9898543536095955, + "grad_norm": 0.6454940540503779, + "learning_rate": 1.070227791927003e-09, + "loss": 0.8325, + "step": 10976 + }, + { + "epoch": 0.9899445371330657, + "grad_norm": 1.8258634745438498, + "learning_rate": 1.0512035661118802e-09, + "loss": 1.0712, + "step": 10977 + }, + { + "epoch": 0.9900347206565361, + "grad_norm": 2.0287929052304063, + "learning_rate": 1.0323499016300364e-09, + "loss": 0.9929, + "step": 10978 + }, + { + "epoch": 0.9901249041800063, + "grad_norm": 2.3625652647338558, + "learning_rate": 1.013666800090407e-09, + "loss": 0.9271, + "step": 10979 + }, + { + "epoch": 0.9902150877034765, + "grad_norm": 1.713453907057531, + "learning_rate": 9.951542630870502e-10, + "loss": 0.9589, + "step": 10980 + }, + { + "epoch": 0.9903052712269468, + "grad_norm": 1.7658424581112229, + "learning_rate": 9.768122921995915e-10, + "loss": 1.0109, + "step": 10981 + }, + { + "epoch": 0.9903954547504171, + "grad_norm": 2.0102797018512852, + "learning_rate": 9.58640888992779e-10, + "loss": 0.9171, + "step": 10982 + }, + { + "epoch": 0.9904856382738874, + "grad_norm": 1.5823689418236493, + "learning_rate": 9.40640055017594e-10, + "loss": 0.8888, + "step": 10983 + }, + { + "epoch": 0.9905758217973576, + "grad_norm": 1.6912365271009806, + "learning_rate": 9.228097918094757e-10, + "loss": 0.9726, + "step": 10984 + }, + { + "epoch": 0.9906660053208279, + "grad_norm": 2.1910838605674834, + "learning_rate": 9.051501008900952e-10, + "loss": 0.9401, + "step": 10985 + }, + { + "epoch": 0.9907561888442982, + "grad_norm": 1.5910353635242114, + "learning_rate": 8.876609837662475e-10, + "loss": 0.9893, + "step": 10986 + }, + { + "epoch": 0.9908463723677684, + "grad_norm": 1.599442581591431, + "learning_rate": 8.70342441930294e-10, + "loss": 1.056, + "step": 10987 + }, + { + "epoch": 0.9909365558912386, + "grad_norm": 1.6113903426671763, + "learning_rate": 8.531944768594979e-10, + "loss": 0.9437, + "step": 10988 + }, + { + "epoch": 0.991026739414709, + "grad_norm": 1.5115453339698808, + "learning_rate": 8.362170900175769e-10, + "loss": 1.0565, + "step": 10989 + }, + { + "epoch": 0.9911169229381792, + "grad_norm": 2.074982181150842, + "learning_rate": 8.194102828527061e-10, + "loss": 0.9292, + "step": 10990 + }, + { + "epoch": 0.9912071064616494, + "grad_norm": 1.7873528319911016, + "learning_rate": 8.027740567992936e-10, + "loss": 1.0173, + "step": 10991 + }, + { + "epoch": 0.9912972899851197, + "grad_norm": 1.4363923988765797, + "learning_rate": 7.863084132766484e-10, + "loss": 0.9665, + "step": 10992 + }, + { + "epoch": 0.99138747350859, + "grad_norm": 1.8930659024776864, + "learning_rate": 7.700133536896469e-10, + "loss": 0.9891, + "step": 10993 + }, + { + "epoch": 0.9914776570320603, + "grad_norm": 1.5470416817305204, + "learning_rate": 7.538888794287324e-10, + "loss": 0.9388, + "step": 10994 + }, + { + "epoch": 0.9915678405555305, + "grad_norm": 1.4942472133141391, + "learning_rate": 7.379349918696931e-10, + "loss": 0.9963, + "step": 10995 + }, + { + "epoch": 0.9916580240790007, + "grad_norm": 1.9056095861084923, + "learning_rate": 7.221516923738846e-10, + "loss": 0.9449, + "step": 10996 + }, + { + "epoch": 0.9917482076024711, + "grad_norm": 2.053495111883646, + "learning_rate": 7.065389822880075e-10, + "loss": 1.0751, + "step": 10997 + }, + { + "epoch": 0.9918383911259413, + "grad_norm": 1.7374165302233682, + "learning_rate": 6.910968629443292e-10, + "loss": 1.0497, + "step": 10998 + }, + { + "epoch": 0.9919285746494115, + "grad_norm": 1.7199047717635232, + "learning_rate": 6.758253356602406e-10, + "loss": 0.9671, + "step": 10999 + }, + { + "epoch": 0.9920187581728818, + "grad_norm": 1.5802910169755886, + "learning_rate": 6.607244017389213e-10, + "loss": 1.0683, + "step": 11000 + }, + { + "epoch": 0.9921089416963521, + "grad_norm": 1.8930548194528691, + "learning_rate": 6.457940624686742e-10, + "loss": 0.9919, + "step": 11001 + }, + { + "epoch": 0.9921991252198223, + "grad_norm": 1.7936739729343933, + "learning_rate": 6.310343191238132e-10, + "loss": 1.0551, + "step": 11002 + }, + { + "epoch": 0.9922893087432926, + "grad_norm": 1.8047956293913832, + "learning_rate": 6.164451729635534e-10, + "loss": 0.8654, + "step": 11003 + }, + { + "epoch": 0.9923794922667628, + "grad_norm": 1.5810710489087414, + "learning_rate": 6.020266252324546e-10, + "loss": 0.8431, + "step": 11004 + }, + { + "epoch": 0.9924696757902332, + "grad_norm": 1.6114584020488438, + "learning_rate": 5.877786771610882e-10, + "loss": 0.9448, + "step": 11005 + }, + { + "epoch": 0.9925598593137034, + "grad_norm": 0.7172368333923692, + "learning_rate": 5.737013299651483e-10, + "loss": 0.8838, + "step": 11006 + }, + { + "epoch": 0.9926500428371736, + "grad_norm": 1.8857878289806085, + "learning_rate": 5.597945848458963e-10, + "loss": 0.9765, + "step": 11007 + }, + { + "epoch": 0.992740226360644, + "grad_norm": 2.1910001817126004, + "learning_rate": 5.460584429894944e-10, + "loss": 0.9476, + "step": 11008 + }, + { + "epoch": 0.9928304098841142, + "grad_norm": 1.6976723836526284, + "learning_rate": 5.32492905568338e-10, + "loss": 0.9731, + "step": 11009 + }, + { + "epoch": 0.9929205934075844, + "grad_norm": 2.8304145776216916, + "learning_rate": 5.190979737399459e-10, + "loss": 1.0084, + "step": 11010 + }, + { + "epoch": 0.9930107769310547, + "grad_norm": 1.9887351846292958, + "learning_rate": 5.058736486469594e-10, + "loss": 0.865, + "step": 11011 + }, + { + "epoch": 0.993100960454525, + "grad_norm": 2.20495053674835, + "learning_rate": 4.928199314180314e-10, + "loss": 1.0058, + "step": 11012 + }, + { + "epoch": 0.9931911439779952, + "grad_norm": 1.5230750802131952, + "learning_rate": 4.799368231669376e-10, + "loss": 0.8883, + "step": 11013 + }, + { + "epoch": 0.9932813275014655, + "grad_norm": 2.2192714239652527, + "learning_rate": 4.672243249927988e-10, + "loss": 0.9591, + "step": 11014 + }, + { + "epoch": 0.9933715110249357, + "grad_norm": 1.629311417428095, + "learning_rate": 4.546824379803027e-10, + "loss": 1.001, + "step": 11015 + }, + { + "epoch": 0.993461694548406, + "grad_norm": 1.6395970348414168, + "learning_rate": 4.4231116319970454e-10, + "loss": 1.005, + "step": 11016 + }, + { + "epoch": 0.9935518780718763, + "grad_norm": 2.144379824102318, + "learning_rate": 4.3011050170660423e-10, + "loss": 0.9749, + "step": 11017 + }, + { + "epoch": 0.9936420615953465, + "grad_norm": 1.7558214907012193, + "learning_rate": 4.18080454542169e-10, + "loss": 0.9684, + "step": 11018 + }, + { + "epoch": 0.9937322451188167, + "grad_norm": 0.6672377574136265, + "learning_rate": 4.0622102273246694e-10, + "loss": 0.8137, + "step": 11019 + }, + { + "epoch": 0.9938224286422871, + "grad_norm": 1.440607278512121, + "learning_rate": 3.945322072897994e-10, + "loss": 1.034, + "step": 11020 + }, + { + "epoch": 0.9939126121657573, + "grad_norm": 1.9384412632570047, + "learning_rate": 3.830140092111467e-10, + "loss": 1.0322, + "step": 11021 + }, + { + "epoch": 0.9940027956892276, + "grad_norm": 1.7287706434847447, + "learning_rate": 3.7166642947972225e-10, + "loss": 0.9696, + "step": 11022 + }, + { + "epoch": 0.9940929792126978, + "grad_norm": 1.8156868114881788, + "learning_rate": 3.604894690634186e-10, + "loss": 0.9485, + "step": 11023 + }, + { + "epoch": 0.9941831627361681, + "grad_norm": 1.6097668661415, + "learning_rate": 3.494831289161393e-10, + "loss": 1.0558, + "step": 11024 + }, + { + "epoch": 0.9942733462596384, + "grad_norm": 1.3911427166369605, + "learning_rate": 3.3864740997668897e-10, + "loss": 1.0066, + "step": 11025 + }, + { + "epoch": 0.9943635297831086, + "grad_norm": 2.696546277713465, + "learning_rate": 3.279823131701054e-10, + "loss": 1.0012, + "step": 11026 + }, + { + "epoch": 0.9944537133065788, + "grad_norm": 1.6028950427766502, + "learning_rate": 3.1748783940610536e-10, + "loss": 0.9321, + "step": 11027 + }, + { + "epoch": 0.9945438968300492, + "grad_norm": 3.090220894926989, + "learning_rate": 3.071639895801947e-10, + "loss": 0.934, + "step": 11028 + }, + { + "epoch": 0.9946340803535194, + "grad_norm": 2.22888587062543, + "learning_rate": 2.9701076457322447e-10, + "loss": 1.0353, + "step": 11029 + }, + { + "epoch": 0.9947242638769896, + "grad_norm": 1.959126095575491, + "learning_rate": 2.870281652513906e-10, + "loss": 1.0146, + "step": 11030 + }, + { + "epoch": 0.99481444740046, + "grad_norm": 5.271648459550304, + "learning_rate": 2.772161924669003e-10, + "loss": 0.9607, + "step": 11031 + }, + { + "epoch": 0.9949046309239302, + "grad_norm": 1.5360365811402652, + "learning_rate": 2.6757484705641765e-10, + "loss": 0.9565, + "step": 11032 + }, + { + "epoch": 0.9949948144474005, + "grad_norm": 1.5857738541984938, + "learning_rate": 2.58104129843062e-10, + "loss": 0.9757, + "step": 11033 + }, + { + "epoch": 0.9950849979708707, + "grad_norm": 2.054996017674952, + "learning_rate": 2.4880404163463154e-10, + "loss": 0.9285, + "step": 11034 + }, + { + "epoch": 0.995175181494341, + "grad_norm": 1.8197874695486207, + "learning_rate": 2.3967458322471377e-10, + "loss": 0.9619, + "step": 11035 + }, + { + "epoch": 0.9952653650178113, + "grad_norm": 1.7273091216565182, + "learning_rate": 2.307157553922412e-10, + "loss": 0.9864, + "step": 11036 + }, + { + "epoch": 0.9953555485412815, + "grad_norm": 1.646141731565916, + "learning_rate": 2.2192755890193538e-10, + "loss": 1.0191, + "step": 11037 + }, + { + "epoch": 0.9954457320647517, + "grad_norm": 0.8272726602963965, + "learning_rate": 2.133099945034189e-10, + "loss": 0.7823, + "step": 11038 + }, + { + "epoch": 0.9955359155882221, + "grad_norm": 1.827276244757522, + "learning_rate": 2.048630629318815e-10, + "loss": 0.9934, + "step": 11039 + }, + { + "epoch": 0.9956260991116923, + "grad_norm": 1.4173224371248725, + "learning_rate": 1.965867649080799e-10, + "loss": 0.9707, + "step": 11040 + }, + { + "epoch": 0.9957162826351625, + "grad_norm": 0.611212381246747, + "learning_rate": 1.8848110113856008e-10, + "loss": 0.8009, + "step": 11041 + }, + { + "epoch": 0.9958064661586328, + "grad_norm": 2.381569159638219, + "learning_rate": 1.8054607231454687e-10, + "loss": 1.0121, + "step": 11042 + }, + { + "epoch": 0.9958966496821031, + "grad_norm": 1.4319461845110288, + "learning_rate": 1.7278167911327635e-10, + "loss": 0.9272, + "step": 11043 + }, + { + "epoch": 0.9959868332055734, + "grad_norm": 1.7223054351323974, + "learning_rate": 1.6518792219710753e-10, + "loss": 1.0079, + "step": 11044 + }, + { + "epoch": 0.9960770167290436, + "grad_norm": 1.4247735629953788, + "learning_rate": 1.5776480221418865e-10, + "loss": 0.981, + "step": 11045 + }, + { + "epoch": 0.9961672002525138, + "grad_norm": 2.8842017488532905, + "learning_rate": 1.505123197977909e-10, + "loss": 0.9736, + "step": 11046 + }, + { + "epoch": 0.9962573837759842, + "grad_norm": 1.4965484963871283, + "learning_rate": 1.4343047556675258e-10, + "loss": 0.9074, + "step": 11047 + }, + { + "epoch": 0.9963475672994544, + "grad_norm": 1.8497558664640554, + "learning_rate": 1.3651927012503506e-10, + "loss": 0.98, + "step": 11048 + }, + { + "epoch": 0.9964377508229246, + "grad_norm": 1.5693859954047735, + "learning_rate": 1.297787040630549e-10, + "loss": 0.9795, + "step": 11049 + }, + { + "epoch": 0.9965279343463949, + "grad_norm": 2.1246123240766557, + "learning_rate": 1.2320877795524153e-10, + "loss": 0.9987, + "step": 11050 + }, + { + "epoch": 0.9966181178698652, + "grad_norm": 1.8739323437344377, + "learning_rate": 1.1680949236247962e-10, + "loss": 0.9613, + "step": 11051 + }, + { + "epoch": 0.9967083013933354, + "grad_norm": 1.5799138722907689, + "learning_rate": 1.1058084783099886e-10, + "loss": 0.9784, + "step": 11052 + }, + { + "epoch": 0.9967984849168057, + "grad_norm": 1.8725631136883194, + "learning_rate": 1.0452284489170793e-10, + "loss": 0.9885, + "step": 11053 + }, + { + "epoch": 0.996888668440276, + "grad_norm": 1.6617277000770887, + "learning_rate": 9.86354840621928e-11, + "loss": 0.9593, + "step": 11054 + }, + { + "epoch": 0.9969788519637462, + "grad_norm": 1.7247582584110954, + "learning_rate": 9.291876584427427e-11, + "loss": 0.8518, + "step": 11055 + }, + { + "epoch": 0.9970690354872165, + "grad_norm": 1.9319673730492275, + "learning_rate": 8.737269072578435e-11, + "loss": 0.9442, + "step": 11056 + }, + { + "epoch": 0.9971592190106867, + "grad_norm": 1.4911578555549059, + "learning_rate": 8.199725918012212e-11, + "loss": 0.9726, + "step": 11057 + }, + { + "epoch": 0.9972494025341571, + "grad_norm": 2.096551870812949, + "learning_rate": 7.679247166603175e-11, + "loss": 0.9841, + "step": 11058 + }, + { + "epoch": 0.9973395860576273, + "grad_norm": 1.6504731106883337, + "learning_rate": 7.17583286273804e-11, + "loss": 0.9811, + "step": 11059 + }, + { + "epoch": 0.9974297695810975, + "grad_norm": 1.4203639122918692, + "learning_rate": 6.689483049360233e-11, + "loss": 0.9958, + "step": 11060 + }, + { + "epoch": 0.9975199531045678, + "grad_norm": 1.7927385568887932, + "learning_rate": 6.220197768014302e-11, + "loss": 0.9537, + "step": 11061 + }, + { + "epoch": 0.9976101366280381, + "grad_norm": 1.7814538404045004, + "learning_rate": 5.7679770587126806e-11, + "loss": 0.9967, + "step": 11062 + }, + { + "epoch": 0.9977003201515083, + "grad_norm": 0.6102560716316825, + "learning_rate": 5.33282096002452e-11, + "loss": 0.903, + "step": 11063 + }, + { + "epoch": 0.9977905036749786, + "grad_norm": 1.6679419486329252, + "learning_rate": 4.914729509120086e-11, + "loss": 0.9435, + "step": 11064 + }, + { + "epoch": 0.9978806871984488, + "grad_norm": 1.3706397802110015, + "learning_rate": 4.513702741637537e-11, + "loss": 0.8981, + "step": 11065 + }, + { + "epoch": 0.9979708707219191, + "grad_norm": 1.6311994552089424, + "learning_rate": 4.129740691816153e-11, + "loss": 1.0354, + "step": 11066 + }, + { + "epoch": 0.9980610542453894, + "grad_norm": 1.3466894690232372, + "learning_rate": 3.762843392429715e-11, + "loss": 0.9959, + "step": 11067 + }, + { + "epoch": 0.9981512377688596, + "grad_norm": 1.9757857172181297, + "learning_rate": 3.413010874742106e-11, + "loss": 0.956, + "step": 11068 + }, + { + "epoch": 0.9982414212923298, + "grad_norm": 2.8832157338880573, + "learning_rate": 3.080243168618324e-11, + "loss": 0.9677, + "step": 11069 + }, + { + "epoch": 0.9983316048158002, + "grad_norm": 1.4750384180270566, + "learning_rate": 2.7645403024800783e-11, + "loss": 0.9153, + "step": 11070 + }, + { + "epoch": 0.9984217883392704, + "grad_norm": 1.3963639640914964, + "learning_rate": 2.4659023032391756e-11, + "loss": 1.044, + "step": 11071 + }, + { + "epoch": 0.9985119718627407, + "grad_norm": 1.687214014756811, + "learning_rate": 2.1843291963863364e-11, + "loss": 1.0218, + "step": 11072 + }, + { + "epoch": 0.9986021553862109, + "grad_norm": 1.9282365046429144, + "learning_rate": 1.9198210059245822e-11, + "loss": 0.9938, + "step": 11073 + }, + { + "epoch": 0.9986923389096812, + "grad_norm": 1.7434949544795868, + "learning_rate": 1.672377754458054e-11, + "loss": 0.9786, + "step": 11074 + }, + { + "epoch": 0.9987825224331515, + "grad_norm": 1.4813286289640544, + "learning_rate": 1.4419994630809895e-11, + "loss": 0.968, + "step": 11075 + }, + { + "epoch": 0.9988727059566217, + "grad_norm": 4.390551149011914, + "learning_rate": 1.2286861514443358e-11, + "loss": 0.9262, + "step": 11076 + }, + { + "epoch": 0.998962889480092, + "grad_norm": 1.3105378243010177, + "learning_rate": 1.0324378377779553e-11, + "loss": 1.0119, + "step": 11077 + }, + { + "epoch": 0.9990530730035623, + "grad_norm": 1.6396525816328287, + "learning_rate": 8.532545388018064e-12, + "loss": 0.9737, + "step": 11078 + }, + { + "epoch": 0.9991432565270325, + "grad_norm": 1.435564313510153, + "learning_rate": 6.911362697925582e-12, + "loss": 0.8119, + "step": 11079 + }, + { + "epoch": 0.9992334400505027, + "grad_norm": 1.5377351875615057, + "learning_rate": 5.46083044605794e-12, + "loss": 0.9473, + "step": 11080 + }, + { + "epoch": 0.9993236235739731, + "grad_norm": 1.6689859544906995, + "learning_rate": 4.1809487563160276e-12, + "loss": 1.0268, + "step": 11081 + }, + { + "epoch": 0.9994138070974433, + "grad_norm": 1.4523459366263294, + "learning_rate": 3.0717177375017e-12, + "loss": 0.9773, + "step": 11082 + }, + { + "epoch": 0.9995039906209136, + "grad_norm": 1.9407673020334224, + "learning_rate": 2.1331374846500495e-12, + "loss": 0.9768, + "step": 11083 + }, + { + "epoch": 0.9995941741443838, + "grad_norm": 2.175990315499914, + "learning_rate": 1.3652080774750885e-12, + "loss": 1.0732, + "step": 11084 + }, + { + "epoch": 0.9996843576678541, + "grad_norm": 1.3054038268495938, + "learning_rate": 7.679295817020204e-13, + "loss": 0.891, + "step": 11085 + }, + { + "epoch": 0.9997745411913244, + "grad_norm": 2.366152496182964, + "learning_rate": 3.413020484011042e-13, + "loss": 0.9521, + "step": 11086 + }, + { + "epoch": 0.9998647247147946, + "grad_norm": 1.564118958024902, + "learning_rate": 8.53255139876552e-14, + "loss": 1.017, + "step": 11087 + }, + { + "epoch": 0.9999549082382648, + "grad_norm": 1.909743687672825, + "learning_rate": 0.0, + "loss": 0.9163, + "step": 11088 + }, + { + "epoch": 0.9999549082382648, + "step": 11088, + "total_flos": 1.3538720343203512e+18, + "train_loss": 0.7547961819008487, + "train_runtime": 106983.5577, + "train_samples_per_second": 6.219, + "train_steps_per_second": 0.104 + } + ], + "logging_steps": 1.0, + "max_steps": 11088, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.3538720343203512e+18, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +} diff --git a/sft_full/smoe_cosinegating/training_args.bin b/sft_full/smoe_cosinegating/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..c95ae66756f8ff8fcee0312bb8a4a2ea01839815 --- /dev/null +++ b/sft_full/smoe_cosinegating/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:74abf7f3a5b16db530b49de04d931d31874d9902af82c92d683f25980e00ed19 +size 8184 diff --git a/sft_full/smoe_perturbed/added_tokens.json b/sft_full/smoe_perturbed/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d3d3a1b74d87e381e471f7b33784015d2dc0ea --- /dev/null +++ b/sft_full/smoe_perturbed/added_tokens.json @@ -0,0 +1,13 @@ +{ + "<|assistant|>": 32001, + "<|endoftext|>": 32000, + "<|end|>": 32007, + "<|placeholder1|>": 32002, + "<|placeholder2|>": 32003, + "<|placeholder3|>": 32004, + "<|placeholder4|>": 32005, + "<|placeholder5|>": 32008, + "<|placeholder6|>": 32009, + "<|system|>": 32006, + "<|user|>": 32010 +} diff --git a/sft_full/smoe_perturbed/config.json b/sft_full/smoe_perturbed/config.json new file mode 100644 index 0000000000000000000000000000000000000000..fad7751e96bd5777e51b9533a82bb18f8af23148 --- /dev/null +++ b/sft_full/smoe_perturbed/config.json @@ -0,0 +1,168 @@ +{ + "_name_or_path": "/cm/archive/namnv78/checkpoints/phi35-siglip224/pft", + "architectures": [ + "LlavaPhiForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "auto_map": { + "AutoConfig": "configuration_phi3.Phi3Config", + "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM" + }, + "balance_loss_coef": 0.1, + "bos_token_id": 1, + "clip_smoe": true, + "dropout": false, + "embd_pdrop": 0.0, + "eos_token_id": 32000, + "freeze_mm_mlp_adapter": false, + "hidden_act": "silu", + "hidden_size": 3072, + "image_aspect_ratio": "pad", + "initializer_range": 0.02, + "intermediate_size": 8192, + "local_rank": 0, + "max_position_embeddings": 131072, + "mlp_smoe": true, + "mm_hidden_size": 1152, + "mm_patch_merge_type": "flat", + "mm_projector_lr": null, + "mm_projector_type": "moe", + "mm_use_im_patch_token": false, + "mm_use_im_start_end": false, + "mm_vision_select_feature": "patch", + "mm_vision_select_layer": -2, + "mm_vision_tower": "google/siglip-so400m-patch14-224", + "model_type": "llava_phi", + "moe_name": "smoe_perturbed", + "num_attention_heads": 32, + "num_experts": 4, + "num_hidden_layers": 32, + "num_key_value_heads": 32, + "num_layers": 3, + "num_selected": 2, + "original_max_position_embeddings": 4096, + "pad_token_id": 32000, + "resid_pdrop": 0.0, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "long_factor": [ + 1.0800000429153442, + 1.1100000143051147, + 1.1399999856948853, + 1.340000033378601, + 1.5899999141693115, + 1.600000023841858, + 1.6200000047683716, + 2.620000123977661, + 3.2300000190734863, + 3.2300000190734863, + 4.789999961853027, + 7.400000095367432, + 7.700000286102295, + 9.09000015258789, + 12.199999809265137, + 17.670000076293945, + 24.46000099182129, + 28.57000160217285, + 30.420001983642578, + 30.840002059936523, + 32.590003967285156, + 32.93000411987305, + 42.320003509521484, + 44.96000289916992, + 50.340003967285156, + 50.45000457763672, + 57.55000305175781, + 57.93000411987305, + 58.21000289916992, + 60.1400032043457, + 62.61000442504883, + 62.62000274658203, + 62.71000289916992, + 63.1400032043457, + 63.1400032043457, + 63.77000427246094, + 63.93000411987305, + 63.96000289916992, + 63.970001220703125, + 64.02999877929688, + 64.06999969482422, + 64.08000183105469, + 64.12000274658203, + 64.41000366210938, + 64.4800033569336, + 64.51000213623047, + 64.52999877929688, + 64.83999633789062 + ], + "short_factor": [ + 1.0, + 1.0199999809265137, + 1.0299999713897705, + 1.0299999713897705, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0699999332427979, + 1.0999999046325684, + 1.1099998950958252, + 1.1599998474121094, + 1.1599998474121094, + 1.1699998378753662, + 1.2899998426437378, + 1.339999794960022, + 1.679999828338623, + 1.7899998426437378, + 1.8199998140335083, + 1.8499997854232788, + 1.8799997568130493, + 1.9099997282028198, + 1.9399996995925903, + 1.9899996519088745, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0799996852874756, + 2.0899996757507324, + 2.189999580383301, + 2.2199995517730713, + 2.5899994373321533, + 2.729999542236328, + 2.749999523162842, + 2.8399994373321533 + ], + "type": "longrope" + }, + "rope_theta": 10000.0, + "router_z_loss_coef": 0.01, + "scales": [ + 1, + 3 + ], + "sliding_window": 262144, + "tie_word_embeddings": false, + "tokenizer_model_max_length": 2048, + "tokenizer_padding_side": "right", + "torch_dtype": "bfloat16", + "training": true, + "transformers_version": "4.43.0", + "tune_mm_mlp_adapter": false, + "use_cache": true, + "use_mm_proj": true, + "vocab_size": 32064 +} diff --git a/sft_full/smoe_perturbed/generation_config.json b/sft_full/smoe_perturbed/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..dad5c4578f0dc5969b38755d095fc30c368bb54a --- /dev/null +++ b/sft_full/smoe_perturbed/generation_config.json @@ -0,0 +1,12 @@ +{ + "_from_model_config": true, + "bos_token_id": 1, + "do_sample": true, + "eos_token_id": [ + 32007, + 32001, + 32000 + ], + "pad_token_id": 32000, + "transformers_version": "4.43.0" +} diff --git a/sft_full/smoe_perturbed/model-00001-of-00003.safetensors b/sft_full/smoe_perturbed/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..987cb3fe208b7b1309a98371f0ebc6459469e494 --- /dev/null +++ b/sft_full/smoe_perturbed/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:87fd7ac6ec78d179eaf9a74c6bcf515318e337ab9f99c8c6d5659630e3f884a9 +size 4972489328 diff --git a/sft_full/smoe_perturbed/model-00002-of-00003.safetensors b/sft_full/smoe_perturbed/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3dace7885a5e5586f8bea990abbc688effb2ac48 --- /dev/null +++ b/sft_full/smoe_perturbed/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea4992fcf804b3dd4e133110725b2bfeb819559139bddaeb852e6931cf4faa74 +size 4985533608 diff --git a/sft_full/smoe_perturbed/model-00003-of-00003.safetensors b/sft_full/smoe_perturbed/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..132132af46cd2424bf8a3b09c92c3c3ef802c17c --- /dev/null +++ b/sft_full/smoe_perturbed/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:369ab1a2c9903478d233348bcdee93fab707680031305c19354783c04fbf4004 +size 248943664 diff --git a/sft_full/smoe_perturbed/model.safetensors.index.json b/sft_full/smoe_perturbed/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..f5e0d563e520320e7e1cb47747945b2591e60790 --- /dev/null +++ b/sft_full/smoe_perturbed/model.safetensors.index.json @@ -0,0 +1,1033 @@ +{ + "metadata": { + "total_size": 10206819680 + }, + "weight_map": { + "lm_head.weight": "model-00003-of-00003.safetensors", + "model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.gate.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.gate.weight": "model-00003-of-00003.safetensors", + "model.norm.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.position_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.gate.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.gate.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors" + } +} diff --git a/sft_full/smoe_perturbed/special_tokens_map.json b/sft_full/smoe_perturbed/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..3e4d5a5bc1cb51753cc9ae0305ece0da60052b10 --- /dev/null +++ b/sft_full/smoe_perturbed/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/sft_full/smoe_perturbed/tokenizer.model b/sft_full/smoe_perturbed/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/sft_full/smoe_perturbed/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/sft_full/smoe_perturbed/tokenizer_config.json b/sft_full/smoe_perturbed/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d579bb0b91b24b214ea3c2e487e27a65017cdc4a --- /dev/null +++ b/sft_full/smoe_perturbed/tokenizer_config.json @@ -0,0 +1,132 @@ +{ + "add_bos_token": false, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": false + }, + "32000": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "32001": { + "content": "<|assistant|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32002": { + "content": "<|placeholder1|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32003": { + "content": "<|placeholder2|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32004": { + "content": "<|placeholder3|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32005": { + "content": "<|placeholder4|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32006": { + "content": "<|system|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32007": { + "content": "<|end|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32008": { + "content": "<|placeholder5|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32009": { + "content": "<|placeholder6|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32010": { + "content": "<|user|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "chat_template": "{% for message in messages %}{% if message['role'] == 'system' and message['content'] %}{{'<|system|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'user' %}{{'<|user|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>\n' + message['content'] + '<|end|>\n'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}", + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "legacy": false, + "model_max_length": 2048, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +} diff --git a/sft_full/smoe_perturbed/trainer_state.json b/sft_full/smoe_perturbed/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..f25ab15a045a435a1676de0d30715efd92c84c90 --- /dev/null +++ b/sft_full/smoe_perturbed/trainer_state.json @@ -0,0 +1,77658 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999549082382648, + "eval_steps": 500, + "global_step": 11088, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 9.018352347026198e-05, + "grad_norm": 56.46596562034072, + "learning_rate": 0.0, + "loss": 2.2384, + "step": 1 + }, + { + "epoch": 0.00018036704694052397, + "grad_norm": 57.19681447368813, + "learning_rate": 4.773623799730706e-07, + "loss": 2.17, + "step": 2 + }, + { + "epoch": 0.0002705505704107859, + "grad_norm": 34.88642111093781, + "learning_rate": 7.566014715123208e-07, + "loss": 1.8934, + "step": 3 + }, + { + "epoch": 0.00036073409388104793, + "grad_norm": 56.180543055124964, + "learning_rate": 9.547247599461412e-07, + "loss": 2.2095, + "step": 4 + }, + { + "epoch": 0.0004509176173513099, + "grad_norm": 40.23435527912138, + "learning_rate": 1.108401121501769e-06, + "loss": 1.81, + "step": 5 + }, + { + "epoch": 0.0005411011408215718, + "grad_norm": 43.33755081389435, + "learning_rate": 1.2339638514853914e-06, + "loss": 1.9881, + "step": 6 + }, + { + "epoch": 0.0006312846642918339, + "grad_norm": 35.24446917491984, + "learning_rate": 1.3401256270225321e-06, + "loss": 1.7166, + "step": 7 + }, + { + "epoch": 0.0007214681877620959, + "grad_norm": 27.078855298865108, + "learning_rate": 1.4320871399192119e-06, + "loss": 1.6518, + "step": 8 + }, + { + "epoch": 0.0008116517112323579, + "grad_norm": 16.916221174315478, + "learning_rate": 1.5132029430246416e-06, + "loss": 1.5269, + "step": 9 + }, + { + "epoch": 0.0009018352347026198, + "grad_norm": 16.524799552313507, + "learning_rate": 1.5857635014748399e-06, + "loss": 1.6367, + "step": 10 + }, + { + "epoch": 0.0009920187581728818, + "grad_norm": 18.01106430780813, + "learning_rate": 1.6514025108267924e-06, + "loss": 1.7116, + "step": 11 + }, + { + "epoch": 0.0010822022816431437, + "grad_norm": 45.28353600164744, + "learning_rate": 1.711326231458462e-06, + "loss": 1.5918, + "step": 12 + }, + { + "epoch": 0.0011723858051134058, + "grad_norm": 7.320265339279115, + "learning_rate": 1.7664507107987104e-06, + "loss": 1.4329, + "step": 13 + }, + { + "epoch": 0.0012625693285836677, + "grad_norm": 5.5485725491674005, + "learning_rate": 1.8174880069956024e-06, + "loss": 1.3986, + "step": 14 + }, + { + "epoch": 0.0013527528520539298, + "grad_norm": 4.576329214311269, + "learning_rate": 1.8650025930140899e-06, + "loss": 1.3552, + "step": 15 + }, + { + "epoch": 0.0014429363755241917, + "grad_norm": 5.682716453984549, + "learning_rate": 1.9094495198922823e-06, + "loss": 1.4691, + "step": 16 + }, + { + "epoch": 0.0015331198989944536, + "grad_norm": 4.71133567936599, + "learning_rate": 1.9512009899507514e-06, + "loss": 1.3176, + "step": 17 + }, + { + "epoch": 0.0016233034224647158, + "grad_norm": 4.377738967202798, + "learning_rate": 1.990565322997712e-06, + "loss": 1.3701, + "step": 18 + }, + { + "epoch": 0.0017134869459349777, + "grad_norm": 3.5128330928028197, + "learning_rate": 2.027800787770518e-06, + "loss": 1.2788, + "step": 19 + }, + { + "epoch": 0.0018036704694052396, + "grad_norm": 4.773242902769342, + "learning_rate": 2.06312588144791e-06, + "loss": 1.3721, + "step": 20 + }, + { + "epoch": 0.0018938539928755017, + "grad_norm": 3.199671525785859, + "learning_rate": 2.0967270985348526e-06, + "loss": 1.2822, + "step": 21 + }, + { + "epoch": 0.0019840375163457636, + "grad_norm": 7.0552149646216655, + "learning_rate": 2.128764890799863e-06, + "loss": 1.1954, + "step": 22 + }, + { + "epoch": 0.0020742210398160257, + "grad_norm": 3.2012968475357733, + "learning_rate": 2.1593783012990145e-06, + "loss": 1.1977, + "step": 23 + }, + { + "epoch": 0.0021644045632862874, + "grad_norm": 4.71301223844594, + "learning_rate": 2.188688611431533e-06, + "loss": 1.3463, + "step": 24 + }, + { + "epoch": 0.0022545880867565495, + "grad_norm": 3.4696990768155653, + "learning_rate": 2.216802243003538e-06, + "loss": 1.3246, + "step": 25 + }, + { + "epoch": 0.0023447716102268116, + "grad_norm": 3.275165984628516, + "learning_rate": 2.243813090771781e-06, + "loss": 1.1667, + "step": 26 + }, + { + "epoch": 0.0024349551336970737, + "grad_norm": 3.3083128466480294, + "learning_rate": 2.269804414536962e-06, + "loss": 1.3332, + "step": 27 + }, + { + "epoch": 0.0025251386571673354, + "grad_norm": 2.845881742528325, + "learning_rate": 2.2948503869686733e-06, + "loss": 1.2432, + "step": 28 + }, + { + "epoch": 0.0026153221806375975, + "grad_norm": 2.542986844843696, + "learning_rate": 2.3190173696980436e-06, + "loss": 0.9364, + "step": 29 + }, + { + "epoch": 0.0027055057041078597, + "grad_norm": 3.0526259702641356, + "learning_rate": 2.3423649729871604e-06, + "loss": 1.2448, + "step": 30 + }, + { + "epoch": 0.0027956892275781214, + "grad_norm": 2.1108374153315603, + "learning_rate": 2.364946941580084e-06, + "loss": 1.2417, + "step": 31 + }, + { + "epoch": 0.0028858727510483835, + "grad_norm": 2.5565182719958854, + "learning_rate": 2.3868118998653532e-06, + "loss": 1.3044, + "step": 32 + }, + { + "epoch": 0.0029760562745186456, + "grad_norm": 2.2979726439292776, + "learning_rate": 2.408003982339113e-06, + "loss": 0.8823, + "step": 33 + }, + { + "epoch": 0.0030662397979889073, + "grad_norm": 1.8249906095498323, + "learning_rate": 2.4285633699238223e-06, + "loss": 1.1527, + "step": 34 + }, + { + "epoch": 0.0031564233214591694, + "grad_norm": 2.855718557386898, + "learning_rate": 2.4485267485243007e-06, + "loss": 1.185, + "step": 35 + }, + { + "epoch": 0.0032466068449294315, + "grad_norm": 1.9481886471535736, + "learning_rate": 2.467927702970783e-06, + "loss": 1.221, + "step": 36 + }, + { + "epoch": 0.003336790368399693, + "grad_norm": 3.8177099431734045, + "learning_rate": 2.4867970569753584e-06, + "loss": 1.2404, + "step": 37 + }, + { + "epoch": 0.0034269738918699553, + "grad_norm": 2.3772353644960043, + "learning_rate": 2.5051631677435883e-06, + "loss": 1.0815, + "step": 38 + }, + { + "epoch": 0.0035171574153402174, + "grad_norm": 2.4253952530116196, + "learning_rate": 2.523052182311031e-06, + "loss": 1.0491, + "step": 39 + }, + { + "epoch": 0.003607340938810479, + "grad_norm": 2.9457664241143515, + "learning_rate": 2.540488261420981e-06, + "loss": 1.2527, + "step": 40 + }, + { + "epoch": 0.0036975244622807412, + "grad_norm": 1.8861247326086383, + "learning_rate": 2.557493775753984e-06, + "loss": 1.0524, + "step": 41 + }, + { + "epoch": 0.0037877079857510034, + "grad_norm": 1.9715418925699564, + "learning_rate": 2.5740894785079235e-06, + "loss": 1.167, + "step": 42 + }, + { + "epoch": 0.0038778915092212655, + "grad_norm": 2.0483686515591937, + "learning_rate": 2.5902946576685834e-06, + "loss": 1.172, + "step": 43 + }, + { + "epoch": 0.003968075032691527, + "grad_norm": 2.341844521643101, + "learning_rate": 2.606127270772933e-06, + "loss": 1.098, + "step": 44 + }, + { + "epoch": 0.004058258556161789, + "grad_norm": 3.349376187674, + "learning_rate": 2.62160406452641e-06, + "loss": 1.0737, + "step": 45 + }, + { + "epoch": 0.004148442079632051, + "grad_norm": 1.794282186958553, + "learning_rate": 2.636740681272085e-06, + "loss": 1.2357, + "step": 46 + }, + { + "epoch": 0.004238625603102313, + "grad_norm": 2.5635032550693633, + "learning_rate": 2.651551754008722e-06, + "loss": 1.2686, + "step": 47 + }, + { + "epoch": 0.004328809126572575, + "grad_norm": 1.666940841375674, + "learning_rate": 2.6660509914046035e-06, + "loss": 1.2056, + "step": 48 + }, + { + "epoch": 0.004418992650042837, + "grad_norm": 3.1125095229405337, + "learning_rate": 2.6802512540450642e-06, + "loss": 1.1636, + "step": 49 + }, + { + "epoch": 0.004509176173513099, + "grad_norm": 2.3270151449906527, + "learning_rate": 2.694164622976609e-06, + "loss": 1.1512, + "step": 50 + }, + { + "epoch": 0.0045993596969833616, + "grad_norm": 1.9419733520195954, + "learning_rate": 2.707802461463072e-06, + "loss": 1.0892, + "step": 51 + }, + { + "epoch": 0.004689543220453623, + "grad_norm": 1.839948323187565, + "learning_rate": 2.7211754707448516e-06, + "loss": 1.1327, + "step": 52 + }, + { + "epoch": 0.004779726743923885, + "grad_norm": 1.6102920437947255, + "learning_rate": 2.734293740486721e-06, + "loss": 1.1128, + "step": 53 + }, + { + "epoch": 0.0048699102673941475, + "grad_norm": 1.8210445781993325, + "learning_rate": 2.747166794510033e-06, + "loss": 1.1567, + "step": 54 + }, + { + "epoch": 0.004960093790864409, + "grad_norm": 1.9163341510209384, + "learning_rate": 2.759803632328562e-06, + "loss": 1.1251, + "step": 55 + }, + { + "epoch": 0.005050277314334671, + "grad_norm": 2.518355883487478, + "learning_rate": 2.772212766941744e-06, + "loss": 1.2448, + "step": 56 + }, + { + "epoch": 0.005140460837804933, + "grad_norm": 1.7531385207037316, + "learning_rate": 2.7844022592828385e-06, + "loss": 1.232, + "step": 57 + }, + { + "epoch": 0.005230644361275195, + "grad_norm": 2.0651862251234627, + "learning_rate": 2.7963797496711145e-06, + "loss": 1.0865, + "step": 58 + }, + { + "epoch": 0.005320827884745457, + "grad_norm": 1.783250187631969, + "learning_rate": 2.80815248657541e-06, + "loss": 1.1752, + "step": 59 + }, + { + "epoch": 0.005411011408215719, + "grad_norm": 1.3775061396411798, + "learning_rate": 2.819727352960231e-06, + "loss": 0.8555, + "step": 60 + }, + { + "epoch": 0.005501194931685981, + "grad_norm": 2.778003726882577, + "learning_rate": 2.8311108904541717e-06, + "loss": 1.1343, + "step": 61 + }, + { + "epoch": 0.005591378455156243, + "grad_norm": 4.792906271877832, + "learning_rate": 2.842309321553155e-06, + "loss": 1.083, + "step": 62 + }, + { + "epoch": 0.005681561978626505, + "grad_norm": 2.3098828739346042, + "learning_rate": 2.8533285700471737e-06, + "loss": 1.0694, + "step": 63 + }, + { + "epoch": 0.005771745502096767, + "grad_norm": 2.2783240797761395, + "learning_rate": 2.8641742798384237e-06, + "loss": 1.2063, + "step": 64 + }, + { + "epoch": 0.005861929025567029, + "grad_norm": 2.7875538226349845, + "learning_rate": 2.874851832300479e-06, + "loss": 1.1651, + "step": 65 + }, + { + "epoch": 0.005952112549037291, + "grad_norm": 1.9201119252205443, + "learning_rate": 2.8853663623121834e-06, + "loss": 1.1636, + "step": 66 + }, + { + "epoch": 0.006042296072507553, + "grad_norm": 2.3173222138028686, + "learning_rate": 2.895722773085839e-06, + "loss": 1.1919, + "step": 67 + }, + { + "epoch": 0.0061324795959778146, + "grad_norm": 1.9470253263333959, + "learning_rate": 2.905925749896893e-06, + "loss": 1.1596, + "step": 68 + }, + { + "epoch": 0.006222663119448077, + "grad_norm": 1.8810040664872256, + "learning_rate": 2.915979772811335e-06, + "loss": 1.1498, + "step": 69 + }, + { + "epoch": 0.006312846642918339, + "grad_norm": 1.8066592323788395, + "learning_rate": 2.925889128497372e-06, + "loss": 1.1111, + "step": 70 + }, + { + "epoch": 0.0064030301663886005, + "grad_norm": 1.9032398283193193, + "learning_rate": 2.9356579211992906e-06, + "loss": 1.2072, + "step": 71 + }, + { + "epoch": 0.006493213689858863, + "grad_norm": 2.0566611438390603, + "learning_rate": 2.9452900829438533e-06, + "loss": 1.0799, + "step": 72 + }, + { + "epoch": 0.006583397213329125, + "grad_norm": 2.1273353310319316, + "learning_rate": 2.954789383042727e-06, + "loss": 1.1125, + "step": 73 + }, + { + "epoch": 0.006673580736799386, + "grad_norm": 1.213947763016726, + "learning_rate": 2.9641594369484293e-06, + "loss": 0.9241, + "step": 74 + }, + { + "epoch": 0.006763764260269649, + "grad_norm": 1.789474506496087, + "learning_rate": 2.9734037145158586e-06, + "loss": 1.1362, + "step": 75 + }, + { + "epoch": 0.006853947783739911, + "grad_norm": 2.4025251614685033, + "learning_rate": 2.982525547716659e-06, + "loss": 1.1368, + "step": 76 + }, + { + "epoch": 0.006944131307210172, + "grad_norm": 2.2778899648480753, + "learning_rate": 2.9915281378493246e-06, + "loss": 1.1053, + "step": 77 + }, + { + "epoch": 0.007034314830680435, + "grad_norm": 2.8720429181475753, + "learning_rate": 3.000414562284102e-06, + "loss": 1.1316, + "step": 78 + }, + { + "epoch": 0.0071244983541506966, + "grad_norm": 1.898652931831183, + "learning_rate": 3.009187780778246e-06, + "loss": 1.1223, + "step": 79 + }, + { + "epoch": 0.007214681877620958, + "grad_norm": 4.281606172099889, + "learning_rate": 3.017850641394051e-06, + "loss": 1.1157, + "step": 80 + }, + { + "epoch": 0.007304865401091221, + "grad_norm": 2.2042813343971943, + "learning_rate": 3.0264058860492832e-06, + "loss": 0.9861, + "step": 81 + }, + { + "epoch": 0.0073950489245614825, + "grad_norm": 1.9787084818741858, + "learning_rate": 3.0348561557270548e-06, + "loss": 1.1728, + "step": 82 + }, + { + "epoch": 0.007485232448031745, + "grad_norm": 0.8814264066812668, + "learning_rate": 3.043203995369939e-06, + "loss": 0.8842, + "step": 83 + }, + { + "epoch": 0.007575415971502007, + "grad_norm": 1.8918919377345371, + "learning_rate": 3.051451858480994e-06, + "loss": 1.2171, + "step": 84 + }, + { + "epoch": 0.007665599494972268, + "grad_norm": 3.449434543654869, + "learning_rate": 3.05960211145252e-06, + "loss": 1.0467, + "step": 85 + }, + { + "epoch": 0.007755783018442531, + "grad_norm": 1.5978232298799802, + "learning_rate": 3.0676570376416543e-06, + "loss": 1.0496, + "step": 86 + }, + { + "epoch": 0.007845966541912792, + "grad_norm": 2.0651526299105982, + "learning_rate": 3.0756188412103647e-06, + "loss": 1.1446, + "step": 87 + }, + { + "epoch": 0.007936150065383054, + "grad_norm": 1.6486853101416385, + "learning_rate": 3.083489650746004e-06, + "loss": 1.122, + "step": 88 + }, + { + "epoch": 0.008026333588853317, + "grad_norm": 2.2089106566846675, + "learning_rate": 3.0912715226772975e-06, + "loss": 1.1519, + "step": 89 + }, + { + "epoch": 0.008116517112323578, + "grad_norm": 1.586833155990434, + "learning_rate": 3.098966444499481e-06, + "loss": 1.064, + "step": 90 + }, + { + "epoch": 0.00820670063579384, + "grad_norm": 7.521364173929171, + "learning_rate": 3.1065763378212426e-06, + "loss": 1.1314, + "step": 91 + }, + { + "epoch": 0.008296884159264103, + "grad_norm": 2.593854879074292, + "learning_rate": 3.1141030612451554e-06, + "loss": 1.1353, + "step": 92 + }, + { + "epoch": 0.008387067682734364, + "grad_norm": 0.8768656799625918, + "learning_rate": 3.1215484130924052e-06, + "loss": 0.9155, + "step": 93 + }, + { + "epoch": 0.008477251206204626, + "grad_norm": 2.2460198697088485, + "learning_rate": 3.128914133981793e-06, + "loss": 1.1387, + "step": 94 + }, + { + "epoch": 0.008567434729674889, + "grad_norm": 2.0934545963339213, + "learning_rate": 3.136201909272287e-06, + "loss": 1.1318, + "step": 95 + }, + { + "epoch": 0.00865761825314515, + "grad_norm": 2.273786065820643, + "learning_rate": 3.1434133713776735e-06, + "loss": 1.0939, + "step": 96 + }, + { + "epoch": 0.008747801776615412, + "grad_norm": 1.9478361619711362, + "learning_rate": 3.15055010196128e-06, + "loss": 1.122, + "step": 97 + }, + { + "epoch": 0.008837985300085675, + "grad_norm": 2.1279935507712477, + "learning_rate": 3.157613634018135e-06, + "loss": 1.1129, + "step": 98 + }, + { + "epoch": 0.008928168823555935, + "grad_norm": 2.4112155756413767, + "learning_rate": 3.1646054538514336e-06, + "loss": 1.1395, + "step": 99 + }, + { + "epoch": 0.009018352347026198, + "grad_norm": 1.7377199788426139, + "learning_rate": 3.1715270029496797e-06, + "loss": 1.1712, + "step": 100 + }, + { + "epoch": 0.00910853587049646, + "grad_norm": 1.5570105444588453, + "learning_rate": 3.1783796797704243e-06, + "loss": 1.1523, + "step": 101 + }, + { + "epoch": 0.009198719393966723, + "grad_norm": 1.9668957288611617, + "learning_rate": 3.185164841436142e-06, + "loss": 1.0777, + "step": 102 + }, + { + "epoch": 0.009288902917436984, + "grad_norm": 1.9896110479478366, + "learning_rate": 3.1918838053473723e-06, + "loss": 1.1132, + "step": 103 + }, + { + "epoch": 0.009379086440907246, + "grad_norm": 1.8396781960289843, + "learning_rate": 3.198537850717922e-06, + "loss": 1.0225, + "step": 104 + }, + { + "epoch": 0.009469269964377509, + "grad_norm": 1.9351001766799105, + "learning_rate": 3.205128220036622e-06, + "loss": 1.0766, + "step": 105 + }, + { + "epoch": 0.00955945348784777, + "grad_norm": 1.721309628586147, + "learning_rate": 3.2116561204597917e-06, + "loss": 1.075, + "step": 106 + }, + { + "epoch": 0.009649637011318032, + "grad_norm": 0.8838275711353218, + "learning_rate": 3.218122725138335e-06, + "loss": 0.9066, + "step": 107 + }, + { + "epoch": 0.009739820534788295, + "grad_norm": 2.7053858833279145, + "learning_rate": 3.224529174483104e-06, + "loss": 1.1108, + "step": 108 + }, + { + "epoch": 0.009830004058258556, + "grad_norm": 2.424601343165265, + "learning_rate": 3.2308765773719435e-06, + "loss": 1.0534, + "step": 109 + }, + { + "epoch": 0.009920187581728818, + "grad_norm": 1.7557920245703709, + "learning_rate": 3.2371660123016323e-06, + "loss": 1.1039, + "step": 110 + }, + { + "epoch": 0.010010371105199081, + "grad_norm": 1.7312565855045592, + "learning_rate": 3.2433985284876787e-06, + "loss": 1.0309, + "step": 111 + }, + { + "epoch": 0.010100554628669342, + "grad_norm": 1.6077173825850846, + "learning_rate": 3.2495751469148143e-06, + "loss": 1.1019, + "step": 112 + }, + { + "epoch": 0.010190738152139604, + "grad_norm": 5.928672894710963, + "learning_rate": 3.2556968613407816e-06, + "loss": 1.0738, + "step": 113 + }, + { + "epoch": 0.010280921675609867, + "grad_norm": 2.098202426494257, + "learning_rate": 3.2617646392559094e-06, + "loss": 1.1421, + "step": 114 + }, + { + "epoch": 0.010371105199080128, + "grad_norm": 2.1993880808062674, + "learning_rate": 3.2677794228007836e-06, + "loss": 1.1553, + "step": 115 + }, + { + "epoch": 0.01046128872255039, + "grad_norm": 9.528816287648219, + "learning_rate": 3.273742129644185e-06, + "loss": 1.0909, + "step": 116 + }, + { + "epoch": 0.010551472246020653, + "grad_norm": 2.032842862535067, + "learning_rate": 3.279653653823352e-06, + "loss": 1.1072, + "step": 117 + }, + { + "epoch": 0.010641655769490914, + "grad_norm": 1.871298187491614, + "learning_rate": 3.285514866548481e-06, + "loss": 1.1037, + "step": 118 + }, + { + "epoch": 0.010731839292961176, + "grad_norm": 2.9458436360554474, + "learning_rate": 3.2913266169732838e-06, + "loss": 1.1659, + "step": 119 + }, + { + "epoch": 0.010822022816431439, + "grad_norm": 2.176454066788474, + "learning_rate": 3.2970897329333017e-06, + "loss": 1.0836, + "step": 120 + }, + { + "epoch": 0.0109122063399017, + "grad_norm": 2.936876108997799, + "learning_rate": 3.302805021653585e-06, + "loss": 1.0794, + "step": 121 + }, + { + "epoch": 0.011002389863371962, + "grad_norm": 1.856313926546472, + "learning_rate": 3.3084732704272426e-06, + "loss": 1.0915, + "step": 122 + }, + { + "epoch": 0.011092573386842225, + "grad_norm": 4.454960387938985, + "learning_rate": 3.314095247266304e-06, + "loss": 1.0925, + "step": 123 + }, + { + "epoch": 0.011182756910312485, + "grad_norm": 2.132419088780752, + "learning_rate": 3.3196717015262255e-06, + "loss": 1.0584, + "step": 124 + }, + { + "epoch": 0.011272940433782748, + "grad_norm": 1.660738207971461, + "learning_rate": 3.325203364505307e-06, + "loss": 1.0833, + "step": 125 + }, + { + "epoch": 0.01136312395725301, + "grad_norm": 2.6575739534267204, + "learning_rate": 3.3306909500202442e-06, + "loss": 1.1329, + "step": 126 + }, + { + "epoch": 0.011453307480723271, + "grad_norm": 2.753357658124837, + "learning_rate": 3.3361351549589145e-06, + "loss": 1.1629, + "step": 127 + }, + { + "epoch": 0.011543491004193534, + "grad_norm": 2.517622729040241, + "learning_rate": 3.341536659811494e-06, + "loss": 1.0344, + "step": 128 + }, + { + "epoch": 0.011633674527663796, + "grad_norm": 2.1465971121369556, + "learning_rate": 3.346896129180904e-06, + "loss": 1.1203, + "step": 129 + }, + { + "epoch": 0.011723858051134057, + "grad_norm": 2.6005699853411417, + "learning_rate": 3.35221421227355e-06, + "loss": 1.1042, + "step": 130 + }, + { + "epoch": 0.01181404157460432, + "grad_norm": 3.036865224561485, + "learning_rate": 3.357491543371255e-06, + "loss": 1.1258, + "step": 131 + }, + { + "epoch": 0.011904225098074582, + "grad_norm": 2.0016353596894665, + "learning_rate": 3.3627287422852543e-06, + "loss": 1.0539, + "step": 132 + }, + { + "epoch": 0.011994408621544843, + "grad_norm": 2.7814120984791453, + "learning_rate": 3.3679264147930497e-06, + "loss": 1.1281, + "step": 133 + }, + { + "epoch": 0.012084592145015106, + "grad_norm": 1.8112066850557396, + "learning_rate": 3.37308515305891e-06, + "loss": 1.0456, + "step": 134 + }, + { + "epoch": 0.012174775668485368, + "grad_norm": 2.398935769477858, + "learning_rate": 3.3782055360387313e-06, + "loss": 1.1637, + "step": 135 + }, + { + "epoch": 0.012264959191955629, + "grad_norm": 1.5865499131471876, + "learning_rate": 3.3832881298699633e-06, + "loss": 1.0605, + "step": 136 + }, + { + "epoch": 0.012355142715425892, + "grad_norm": 1.7938541395415055, + "learning_rate": 3.388333488247249e-06, + "loss": 1.0734, + "step": 137 + }, + { + "epoch": 0.012445326238896154, + "grad_norm": 2.0503700061996124, + "learning_rate": 3.393342152784406e-06, + "loss": 1.1516, + "step": 138 + }, + { + "epoch": 0.012535509762366415, + "grad_norm": 1.685668233699622, + "learning_rate": 3.3983146533633376e-06, + "loss": 1.1099, + "step": 139 + }, + { + "epoch": 0.012625693285836678, + "grad_norm": 2.29516101646305, + "learning_rate": 3.403251508470442e-06, + "loss": 1.1637, + "step": 140 + }, + { + "epoch": 0.01271587680930694, + "grad_norm": 1.6892871576716075, + "learning_rate": 3.408153225521043e-06, + "loss": 1.0623, + "step": 141 + }, + { + "epoch": 0.012806060332777201, + "grad_norm": 1.9263544989775898, + "learning_rate": 3.413020301172361e-06, + "loss": 0.9899, + "step": 142 + }, + { + "epoch": 0.012896243856247463, + "grad_norm": 1.6761382462971464, + "learning_rate": 3.4178532216255024e-06, + "loss": 1.0825, + "step": 143 + }, + { + "epoch": 0.012986427379717726, + "grad_norm": 1.6159509178818816, + "learning_rate": 3.422652462916924e-06, + "loss": 1.1394, + "step": 144 + }, + { + "epoch": 0.013076610903187987, + "grad_norm": 2.175984398828004, + "learning_rate": 3.4274184911998124e-06, + "loss": 1.0033, + "step": 145 + }, + { + "epoch": 0.01316679442665825, + "grad_norm": 2.295520305120807, + "learning_rate": 3.4321517630157976e-06, + "loss": 1.1911, + "step": 146 + }, + { + "epoch": 0.013256977950128512, + "grad_norm": 3.0987260661983984, + "learning_rate": 3.4368527255573845e-06, + "loss": 1.1583, + "step": 147 + }, + { + "epoch": 0.013347161473598773, + "grad_norm": 1.8401829109291443, + "learning_rate": 3.4415218169214994e-06, + "loss": 1.0699, + "step": 148 + }, + { + "epoch": 0.013437344997069035, + "grad_norm": 1.7851759494220443, + "learning_rate": 3.4461594663544882e-06, + "loss": 1.0747, + "step": 149 + }, + { + "epoch": 0.013527528520539298, + "grad_norm": 2.0436449263040655, + "learning_rate": 3.450766094488929e-06, + "loss": 1.0411, + "step": 150 + }, + { + "epoch": 0.013617712044009559, + "grad_norm": 1.9260711761189082, + "learning_rate": 3.4553421135725735e-06, + "loss": 1.1409, + "step": 151 + }, + { + "epoch": 0.013707895567479821, + "grad_norm": 2.3950136081571762, + "learning_rate": 3.45988792768973e-06, + "loss": 1.11, + "step": 152 + }, + { + "epoch": 0.013798079090950084, + "grad_norm": 1.5238310476752766, + "learning_rate": 3.464403932975393e-06, + "loss": 1.1592, + "step": 153 + }, + { + "epoch": 0.013888262614420345, + "grad_norm": 2.1148069166422108, + "learning_rate": 3.468890517822395e-06, + "loss": 1.1489, + "step": 154 + }, + { + "epoch": 0.013978446137890607, + "grad_norm": 4.369009876298557, + "learning_rate": 3.473348063081853e-06, + "loss": 1.0872, + "step": 155 + }, + { + "epoch": 0.01406862966136087, + "grad_norm": 2.9652731362117164, + "learning_rate": 3.4777769422571727e-06, + "loss": 1.0481, + "step": 156 + }, + { + "epoch": 0.01415881318483113, + "grad_norm": 1.8972594448190792, + "learning_rate": 3.4821775216918497e-06, + "loss": 1.0906, + "step": 157 + }, + { + "epoch": 0.014248996708301393, + "grad_norm": 1.7095780218929622, + "learning_rate": 3.4865501607513164e-06, + "loss": 1.0418, + "step": 158 + }, + { + "epoch": 0.014339180231771656, + "grad_norm": 2.136413552422348, + "learning_rate": 3.4908952119990423e-06, + "loss": 1.1628, + "step": 159 + }, + { + "epoch": 0.014429363755241916, + "grad_norm": 0.914601908576971, + "learning_rate": 3.495213021367122e-06, + "loss": 0.8652, + "step": 160 + }, + { + "epoch": 0.014519547278712179, + "grad_norm": 2.3386111061531274, + "learning_rate": 3.4995039283215464e-06, + "loss": 1.1153, + "step": 161 + }, + { + "epoch": 0.014609730802182442, + "grad_norm": 1.7107315984703753, + "learning_rate": 3.5037682660223533e-06, + "loss": 1.1079, + "step": 162 + }, + { + "epoch": 0.014699914325652702, + "grad_norm": 1.9265311678699975, + "learning_rate": 3.508006361478857e-06, + "loss": 1.1224, + "step": 163 + }, + { + "epoch": 0.014790097849122965, + "grad_norm": 1.6625985166572663, + "learning_rate": 3.5122185357001253e-06, + "loss": 1.0748, + "step": 164 + }, + { + "epoch": 0.014880281372593228, + "grad_norm": 2.4351732935720536, + "learning_rate": 3.5164051038408817e-06, + "loss": 1.1469, + "step": 165 + }, + { + "epoch": 0.01497046489606349, + "grad_norm": 1.7406677323032147, + "learning_rate": 3.5205663753430093e-06, + "loss": 1.1322, + "step": 166 + }, + { + "epoch": 0.015060648419533751, + "grad_norm": 2.0593869384049115, + "learning_rate": 3.5247026540727915e-06, + "loss": 1.1886, + "step": 167 + }, + { + "epoch": 0.015150831943004013, + "grad_norm": 1.9963871391261236, + "learning_rate": 3.5288142384540645e-06, + "loss": 1.0382, + "step": 168 + }, + { + "epoch": 0.015241015466474276, + "grad_norm": 1.951858414998777, + "learning_rate": 3.532901421597421e-06, + "loss": 1.124, + "step": 169 + }, + { + "epoch": 0.015331198989944537, + "grad_norm": 2.2926194608923125, + "learning_rate": 3.5369644914255915e-06, + "loss": 1.0369, + "step": 170 + }, + { + "epoch": 0.0154213825134148, + "grad_norm": 2.1184305856121717, + "learning_rate": 3.5410037307951596e-06, + "loss": 1.091, + "step": 171 + }, + { + "epoch": 0.015511566036885062, + "grad_norm": 1.9369589142348786, + "learning_rate": 3.545019417614725e-06, + "loss": 1.1383, + "step": 172 + }, + { + "epoch": 0.015601749560355323, + "grad_norm": 1.785094412435884, + "learning_rate": 3.5490118249596387e-06, + "loss": 1.1335, + "step": 173 + }, + { + "epoch": 0.015691933083825584, + "grad_norm": 1.8222766806378523, + "learning_rate": 3.5529812211834352e-06, + "loss": 1.1436, + "step": 174 + }, + { + "epoch": 0.015782116607295848, + "grad_norm": 1.504778007407657, + "learning_rate": 3.5569278700260707e-06, + "loss": 1.0945, + "step": 175 + }, + { + "epoch": 0.01587230013076611, + "grad_norm": 2.05588058817344, + "learning_rate": 3.5608520307190746e-06, + "loss": 1.0752, + "step": 176 + }, + { + "epoch": 0.01596248365423637, + "grad_norm": 1.9546637615223035, + "learning_rate": 3.564753958087731e-06, + "loss": 1.0315, + "step": 177 + }, + { + "epoch": 0.016052667177706634, + "grad_norm": 1.9539631380373992, + "learning_rate": 3.5686339026503684e-06, + "loss": 1.1027, + "step": 178 + }, + { + "epoch": 0.016142850701176895, + "grad_norm": 4.4470792722128, + "learning_rate": 3.5724921107148806e-06, + "loss": 1.1783, + "step": 179 + }, + { + "epoch": 0.016233034224647155, + "grad_norm": 2.1599776623595184, + "learning_rate": 3.576328824472552e-06, + "loss": 1.1506, + "step": 180 + }, + { + "epoch": 0.01632321774811742, + "grad_norm": 1.740256158454782, + "learning_rate": 3.5801442820892838e-06, + "loss": 1.186, + "step": 181 + }, + { + "epoch": 0.01641340127158768, + "grad_norm": 1.4760172003542429, + "learning_rate": 3.583938717794313e-06, + "loss": 1.1155, + "step": 182 + }, + { + "epoch": 0.01650358479505794, + "grad_norm": 1.7600373154715832, + "learning_rate": 3.5877123619664928e-06, + "loss": 1.1341, + "step": 183 + }, + { + "epoch": 0.016593768318528206, + "grad_norm": 1.6689210982600127, + "learning_rate": 3.5914654412182268e-06, + "loss": 1.1129, + "step": 184 + }, + { + "epoch": 0.016683951841998466, + "grad_norm": 1.7228655698764905, + "learning_rate": 3.595198178477127e-06, + "loss": 1.1293, + "step": 185 + }, + { + "epoch": 0.016774135365468727, + "grad_norm": 1.8175157009532736, + "learning_rate": 3.5989107930654757e-06, + "loss": 1.0271, + "step": 186 + }, + { + "epoch": 0.01686431888893899, + "grad_norm": 3.002981135230017, + "learning_rate": 3.6026035007775437e-06, + "loss": 1.1582, + "step": 187 + }, + { + "epoch": 0.016954502412409252, + "grad_norm": 1.8076404011542933, + "learning_rate": 3.6062765139548636e-06, + "loss": 1.0638, + "step": 188 + }, + { + "epoch": 0.017044685935879513, + "grad_norm": 1.7883368607219927, + "learning_rate": 3.6099300415594945e-06, + "loss": 0.9613, + "step": 189 + }, + { + "epoch": 0.017134869459349777, + "grad_norm": 1.6730131749320127, + "learning_rate": 3.6135642892453575e-06, + "loss": 0.977, + "step": 190 + }, + { + "epoch": 0.01722505298282004, + "grad_norm": 2.6418447893742356, + "learning_rate": 3.6171794594277004e-06, + "loss": 1.1088, + "step": 191 + }, + { + "epoch": 0.0173152365062903, + "grad_norm": 1.7444163891519895, + "learning_rate": 3.620775751350745e-06, + "loss": 1.0448, + "step": 192 + }, + { + "epoch": 0.017405420029760563, + "grad_norm": 6.700837712085048, + "learning_rate": 3.6243533611535794e-06, + "loss": 1.1332, + "step": 193 + }, + { + "epoch": 0.017495603553230824, + "grad_norm": 1.6747087367740527, + "learning_rate": 3.627912481934351e-06, + "loss": 1.0873, + "step": 194 + }, + { + "epoch": 0.017585787076701085, + "grad_norm": 4.8366604019709865, + "learning_rate": 3.6314533038128e-06, + "loss": 1.0686, + "step": 195 + }, + { + "epoch": 0.01767597060017135, + "grad_norm": 0.8124067913290889, + "learning_rate": 3.6349760139912048e-06, + "loss": 0.8506, + "step": 196 + }, + { + "epoch": 0.01776615412364161, + "grad_norm": 1.8935733851296115, + "learning_rate": 3.638480796813769e-06, + "loss": 1.1543, + "step": 197 + }, + { + "epoch": 0.01785633764711187, + "grad_norm": 1.5417722287852675, + "learning_rate": 3.641967833824504e-06, + "loss": 1.0914, + "step": 198 + }, + { + "epoch": 0.017946521170582135, + "grad_norm": 1.4902770909978054, + "learning_rate": 3.645437303823663e-06, + "loss": 1.0861, + "step": 199 + }, + { + "epoch": 0.018036704694052396, + "grad_norm": 2.395837026399724, + "learning_rate": 3.64888938292275e-06, + "loss": 1.0692, + "step": 200 + }, + { + "epoch": 0.01812688821752266, + "grad_norm": 1.9011600265933188, + "learning_rate": 3.6523242445981603e-06, + "loss": 1.0945, + "step": 201 + }, + { + "epoch": 0.01821707174099292, + "grad_norm": 1.6175639870676835, + "learning_rate": 3.655742059743495e-06, + "loss": 1.0609, + "step": 202 + }, + { + "epoch": 0.018307255264463182, + "grad_norm": 1.792862766432912, + "learning_rate": 3.659142996720576e-06, + "loss": 1.1482, + "step": 203 + }, + { + "epoch": 0.018397438787933446, + "grad_norm": 2.2062812003288057, + "learning_rate": 3.6625272214092135e-06, + "loss": 1.0599, + "step": 204 + }, + { + "epoch": 0.018487622311403707, + "grad_norm": 0.6849927918201072, + "learning_rate": 3.6658948972557535e-06, + "loss": 0.8259, + "step": 205 + }, + { + "epoch": 0.018577805834873968, + "grad_norm": 2.5482322961085044, + "learning_rate": 3.6692461853204432e-06, + "loss": 1.0136, + "step": 206 + }, + { + "epoch": 0.018667989358344232, + "grad_norm": 2.24041412084727, + "learning_rate": 3.672581244323656e-06, + "loss": 1.0312, + "step": 207 + }, + { + "epoch": 0.018758172881814493, + "grad_norm": 1.8225225040631934, + "learning_rate": 3.6759002306909926e-06, + "loss": 0.9658, + "step": 208 + }, + { + "epoch": 0.018848356405284754, + "grad_norm": 1.5843585944596041, + "learning_rate": 3.67920329859731e-06, + "loss": 1.103, + "step": 209 + }, + { + "epoch": 0.018938539928755018, + "grad_norm": 1.7499991825646808, + "learning_rate": 3.6824906000096923e-06, + "loss": 1.0904, + "step": 210 + }, + { + "epoch": 0.01902872345222528, + "grad_norm": 1.9194785583244502, + "learning_rate": 3.6857622847294067e-06, + "loss": 1.0951, + "step": 211 + }, + { + "epoch": 0.01911890697569554, + "grad_norm": 2.893104108733761, + "learning_rate": 3.6890185004328626e-06, + "loss": 1.0882, + "step": 212 + }, + { + "epoch": 0.019209090499165804, + "grad_norm": 2.5574598289459725, + "learning_rate": 3.6922593927116113e-06, + "loss": 1.0268, + "step": 213 + }, + { + "epoch": 0.019299274022636065, + "grad_norm": 2.069358058665387, + "learning_rate": 3.695485105111406e-06, + "loss": 1.078, + "step": 214 + }, + { + "epoch": 0.019389457546106326, + "grad_norm": 1.8773885771582197, + "learning_rate": 3.698695779170352e-06, + "loss": 1.0587, + "step": 215 + }, + { + "epoch": 0.01947964106957659, + "grad_norm": 2.731420784854792, + "learning_rate": 3.7018915544561744e-06, + "loss": 1.0772, + "step": 216 + }, + { + "epoch": 0.01956982459304685, + "grad_norm": 2.054605692696221, + "learning_rate": 3.7050725686026164e-06, + "loss": 1.0366, + "step": 217 + }, + { + "epoch": 0.01966000811651711, + "grad_norm": 1.8863903573348682, + "learning_rate": 3.708238957345014e-06, + "loss": 1.1484, + "step": 218 + }, + { + "epoch": 0.019750191639987376, + "grad_norm": 1.7148225474513206, + "learning_rate": 3.7113908545550482e-06, + "loss": 1.1271, + "step": 219 + }, + { + "epoch": 0.019840375163457637, + "grad_norm": 2.061300796058711, + "learning_rate": 3.7145283922747028e-06, + "loss": 1.0812, + "step": 220 + }, + { + "epoch": 0.019930558686927898, + "grad_norm": 1.7092128753491407, + "learning_rate": 3.7176517007494612e-06, + "loss": 1.1711, + "step": 221 + }, + { + "epoch": 0.020020742210398162, + "grad_norm": 1.9727313641136965, + "learning_rate": 3.7207609084607496e-06, + "loss": 1.0264, + "step": 222 + }, + { + "epoch": 0.020110925733868423, + "grad_norm": 2.608361966968219, + "learning_rate": 3.723856142157645e-06, + "loss": 1.1339, + "step": 223 + }, + { + "epoch": 0.020201109257338683, + "grad_norm": 1.7540168982184408, + "learning_rate": 3.726937526887885e-06, + "loss": 1.0915, + "step": 224 + }, + { + "epoch": 0.020291292780808948, + "grad_norm": 2.7272903651331832, + "learning_rate": 3.7300051860281798e-06, + "loss": 1.0455, + "step": 225 + }, + { + "epoch": 0.02038147630427921, + "grad_norm": 2.047366833018473, + "learning_rate": 3.733059241313852e-06, + "loss": 1.1623, + "step": 226 + }, + { + "epoch": 0.02047165982774947, + "grad_norm": 1.8350837183260553, + "learning_rate": 3.736099812867827e-06, + "loss": 1.0737, + "step": 227 + }, + { + "epoch": 0.020561843351219734, + "grad_norm": 1.8692637754746213, + "learning_rate": 3.73912701922898e-06, + "loss": 1.1441, + "step": 228 + }, + { + "epoch": 0.020652026874689994, + "grad_norm": 0.8893344965898073, + "learning_rate": 3.742140977379868e-06, + "loss": 0.8561, + "step": 229 + }, + { + "epoch": 0.020742210398160255, + "grad_norm": 1.5612862259485119, + "learning_rate": 3.745141802773854e-06, + "loss": 1.1139, + "step": 230 + }, + { + "epoch": 0.02083239392163052, + "grad_norm": 2.1796467001319826, + "learning_rate": 3.748129609361645e-06, + "loss": 1.0762, + "step": 231 + }, + { + "epoch": 0.02092257744510078, + "grad_norm": 1.8608916051988194, + "learning_rate": 3.7511045096172555e-06, + "loss": 1.0583, + "step": 232 + }, + { + "epoch": 0.02101276096857104, + "grad_norm": 1.8393571538185804, + "learning_rate": 3.7540666145634137e-06, + "loss": 1.0783, + "step": 233 + }, + { + "epoch": 0.021102944492041305, + "grad_norm": 1.766578821781322, + "learning_rate": 3.7570160337964225e-06, + "loss": 1.0579, + "step": 234 + }, + { + "epoch": 0.021193128015511566, + "grad_norm": 3.5690257218134906, + "learning_rate": 3.7599528755104913e-06, + "loss": 1.1128, + "step": 235 + }, + { + "epoch": 0.021283311538981827, + "grad_norm": 2.410277624799097, + "learning_rate": 3.7628772465215515e-06, + "loss": 1.079, + "step": 236 + }, + { + "epoch": 0.02137349506245209, + "grad_norm": 1.856448523184033, + "learning_rate": 3.7657892522905666e-06, + "loss": 1.0754, + "step": 237 + }, + { + "epoch": 0.021463678585922352, + "grad_norm": 9.668408291776077, + "learning_rate": 3.7686889969463542e-06, + "loss": 1.0, + "step": 238 + }, + { + "epoch": 0.021553862109392613, + "grad_norm": 1.7618750526127802, + "learning_rate": 3.771576583307928e-06, + "loss": 1.0567, + "step": 239 + }, + { + "epoch": 0.021644045632862877, + "grad_norm": 1.9576618450193537, + "learning_rate": 3.7744521129063722e-06, + "loss": 1.1157, + "step": 240 + }, + { + "epoch": 0.021734229156333138, + "grad_norm": 2.9138396231583004, + "learning_rate": 3.7773156860062653e-06, + "loss": 1.1062, + "step": 241 + }, + { + "epoch": 0.0218244126798034, + "grad_norm": 2.1091110629505967, + "learning_rate": 3.7801674016266554e-06, + "loss": 1.1798, + "step": 242 + }, + { + "epoch": 0.021914596203273663, + "grad_norm": 0.6245926006037643, + "learning_rate": 3.7830073575616035e-06, + "loss": 0.802, + "step": 243 + }, + { + "epoch": 0.022004779726743924, + "grad_norm": 1.7342519630573892, + "learning_rate": 3.785835650400313e-06, + "loss": 1.1199, + "step": 244 + }, + { + "epoch": 0.022094963250214185, + "grad_norm": 2.1346783982279587, + "learning_rate": 3.7886523755468334e-06, + "loss": 0.9896, + "step": 245 + }, + { + "epoch": 0.02218514677368445, + "grad_norm": 1.4914712035480884, + "learning_rate": 3.7914576272393746e-06, + "loss": 1.0742, + "step": 246 + }, + { + "epoch": 0.02227533029715471, + "grad_norm": 2.0369732556167497, + "learning_rate": 3.7942514985692284e-06, + "loss": 1.1868, + "step": 247 + }, + { + "epoch": 0.02236551382062497, + "grad_norm": 2.0918929561486004, + "learning_rate": 3.797034081499296e-06, + "loss": 1.0767, + "step": 248 + }, + { + "epoch": 0.022455697344095235, + "grad_norm": 4.8602133270037005, + "learning_rate": 3.7998054668822595e-06, + "loss": 1.0597, + "step": 249 + }, + { + "epoch": 0.022545880867565496, + "grad_norm": 2.6786289626709077, + "learning_rate": 3.8025657444783776e-06, + "loss": 1.0583, + "step": 250 + }, + { + "epoch": 0.022636064391035757, + "grad_norm": 1.95569703693301, + "learning_rate": 3.80531500297293e-06, + "loss": 0.9912, + "step": 251 + }, + { + "epoch": 0.02272624791450602, + "grad_norm": 2.241307740624565, + "learning_rate": 3.8080533299933147e-06, + "loss": 1.0514, + "step": 252 + }, + { + "epoch": 0.022816431437976282, + "grad_norm": 2.6899933892207826, + "learning_rate": 3.8107808121258067e-06, + "loss": 1.0442, + "step": 253 + }, + { + "epoch": 0.022906614961446543, + "grad_norm": 2.0302436096143444, + "learning_rate": 3.813497534931985e-06, + "loss": 1.1447, + "step": 254 + }, + { + "epoch": 0.022996798484916807, + "grad_norm": 1.7780229884486258, + "learning_rate": 3.816203582964841e-06, + "loss": 1.1099, + "step": 255 + }, + { + "epoch": 0.023086982008387068, + "grad_norm": 1.9504320424085104, + "learning_rate": 3.818899039784565e-06, + "loss": 1.1369, + "step": 256 + }, + { + "epoch": 0.02317716553185733, + "grad_norm": 2.2950813400381604, + "learning_rate": 3.821583987974031e-06, + "loss": 1.1398, + "step": 257 + }, + { + "epoch": 0.023267349055327593, + "grad_norm": 1.9560513355913929, + "learning_rate": 3.8242585091539755e-06, + "loss": 1.0236, + "step": 258 + }, + { + "epoch": 0.023357532578797854, + "grad_norm": 1.9997346225151575, + "learning_rate": 3.8269226839978895e-06, + "loss": 1.1646, + "step": 259 + }, + { + "epoch": 0.023447716102268115, + "grad_norm": 2.4905255076448696, + "learning_rate": 3.82957659224662e-06, + "loss": 1.0388, + "step": 260 + }, + { + "epoch": 0.02353789962573838, + "grad_norm": 2.4467542011945604, + "learning_rate": 3.8322203127226855e-06, + "loss": 0.9668, + "step": 261 + }, + { + "epoch": 0.02362808314920864, + "grad_norm": 2.1796454969071233, + "learning_rate": 3.834853923344326e-06, + "loss": 1.1895, + "step": 262 + }, + { + "epoch": 0.0237182666726789, + "grad_norm": 3.064849575016769, + "learning_rate": 3.837477501139285e-06, + "loss": 1.0853, + "step": 263 + }, + { + "epoch": 0.023808450196149165, + "grad_norm": 2.2832083461365666, + "learning_rate": 3.840091122258324e-06, + "loss": 1.0958, + "step": 264 + }, + { + "epoch": 0.023898633719619426, + "grad_norm": 2.2056820146429175, + "learning_rate": 3.84269486198849e-06, + "loss": 1.0056, + "step": 265 + }, + { + "epoch": 0.023988817243089686, + "grad_norm": 3.1623871311959437, + "learning_rate": 3.845288794766121e-06, + "loss": 1.102, + "step": 266 + }, + { + "epoch": 0.02407900076655995, + "grad_norm": 1.8062999177258712, + "learning_rate": 3.847872994189619e-06, + "loss": 1.0801, + "step": 267 + }, + { + "epoch": 0.02416918429003021, + "grad_norm": 2.34267329597874, + "learning_rate": 3.8504475330319805e-06, + "loss": 1.1463, + "step": 268 + }, + { + "epoch": 0.024259367813500472, + "grad_norm": 1.9806732132383096, + "learning_rate": 3.853012483253093e-06, + "loss": 1.0629, + "step": 269 + }, + { + "epoch": 0.024349551336970737, + "grad_norm": 3.9261031137490567, + "learning_rate": 3.855567916011802e-06, + "loss": 1.0931, + "step": 270 + }, + { + "epoch": 0.024439734860440997, + "grad_norm": 2.3116557668013185, + "learning_rate": 3.858113901677755e-06, + "loss": 1.0874, + "step": 271 + }, + { + "epoch": 0.024529918383911258, + "grad_norm": 1.8426122872907975, + "learning_rate": 3.860650509843034e-06, + "loss": 1.0467, + "step": 272 + }, + { + "epoch": 0.024620101907381522, + "grad_norm": 2.4873859705260117, + "learning_rate": 3.863177809333563e-06, + "loss": 1.0697, + "step": 273 + }, + { + "epoch": 0.024710285430851783, + "grad_norm": 1.753152392168292, + "learning_rate": 3.86569586822032e-06, + "loss": 1.0564, + "step": 274 + }, + { + "epoch": 0.024800468954322044, + "grad_norm": 2.7366855123833584, + "learning_rate": 3.868204753830331e-06, + "loss": 1.0514, + "step": 275 + }, + { + "epoch": 0.02489065247779231, + "grad_norm": 1.8567634643008977, + "learning_rate": 3.870704532757476e-06, + "loss": 0.9995, + "step": 276 + }, + { + "epoch": 0.02498083600126257, + "grad_norm": 2.006623743724008, + "learning_rate": 3.8731952708730974e-06, + "loss": 1.1108, + "step": 277 + }, + { + "epoch": 0.02507101952473283, + "grad_norm": 2.0242736738285108, + "learning_rate": 3.8756770333364085e-06, + "loss": 1.1065, + "step": 278 + }, + { + "epoch": 0.025161203048203094, + "grad_norm": 2.229351986701322, + "learning_rate": 3.878149884604725e-06, + "loss": 1.0038, + "step": 279 + }, + { + "epoch": 0.025251386571673355, + "grad_norm": 2.333336852843037, + "learning_rate": 3.8806138884435125e-06, + "loss": 1.0589, + "step": 280 + }, + { + "epoch": 0.025341570095143616, + "grad_norm": 1.7395041975040346, + "learning_rate": 3.883069107936248e-06, + "loss": 1.0132, + "step": 281 + }, + { + "epoch": 0.02543175361861388, + "grad_norm": 2.202712108258012, + "learning_rate": 3.885515605494114e-06, + "loss": 1.14, + "step": 282 + }, + { + "epoch": 0.02552193714208414, + "grad_norm": 6.776130114678635, + "learning_rate": 3.8879534428655145e-06, + "loss": 1.1359, + "step": 283 + }, + { + "epoch": 0.025612120665554402, + "grad_norm": 1.908311260989724, + "learning_rate": 3.890382681145432e-06, + "loss": 1.1215, + "step": 284 + }, + { + "epoch": 0.025702304189024666, + "grad_norm": 2.225173463381826, + "learning_rate": 3.892803380784608e-06, + "loss": 1.1136, + "step": 285 + }, + { + "epoch": 0.025792487712494927, + "grad_norm": 0.9611735790516326, + "learning_rate": 3.8952156015985725e-06, + "loss": 0.8803, + "step": 286 + }, + { + "epoch": 0.025882671235965188, + "grad_norm": 1.8542534311403476, + "learning_rate": 3.897619402776516e-06, + "loss": 1.083, + "step": 287 + }, + { + "epoch": 0.025972854759435452, + "grad_norm": 2.7912421591717362, + "learning_rate": 3.900014842889995e-06, + "loss": 1.1585, + "step": 288 + }, + { + "epoch": 0.026063038282905713, + "grad_norm": 1.645628791685912, + "learning_rate": 3.902401979901503e-06, + "loss": 1.0014, + "step": 289 + }, + { + "epoch": 0.026153221806375974, + "grad_norm": 2.3145599469349074, + "learning_rate": 3.904780871172884e-06, + "loss": 1.1248, + "step": 290 + }, + { + "epoch": 0.026243405329846238, + "grad_norm": 3.5366063637106095, + "learning_rate": 3.907151573473601e-06, + "loss": 0.9546, + "step": 291 + }, + { + "epoch": 0.0263335888533165, + "grad_norm": 1.4811531035201948, + "learning_rate": 3.909514142988868e-06, + "loss": 1.0897, + "step": 292 + }, + { + "epoch": 0.02642377237678676, + "grad_norm": 2.2593915756693903, + "learning_rate": 3.911868635327639e-06, + "loss": 1.0458, + "step": 293 + }, + { + "epoch": 0.026513955900257024, + "grad_norm": 2.0449855324529547, + "learning_rate": 3.914215105530455e-06, + "loss": 1.0564, + "step": 294 + }, + { + "epoch": 0.026604139423727285, + "grad_norm": 1.6055483461868594, + "learning_rate": 3.916553608077179e-06, + "loss": 1.0709, + "step": 295 + }, + { + "epoch": 0.026694322947197546, + "grad_norm": 4.011547110952166, + "learning_rate": 3.91888419689457e-06, + "loss": 1.0655, + "step": 296 + }, + { + "epoch": 0.02678450647066781, + "grad_norm": 4.8537521949387825, + "learning_rate": 3.921206925363754e-06, + "loss": 1.0147, + "step": 297 + }, + { + "epoch": 0.02687468999413807, + "grad_norm": 2.9194444961459123, + "learning_rate": 3.923521846327559e-06, + "loss": 0.9956, + "step": 298 + }, + { + "epoch": 0.02696487351760833, + "grad_norm": 2.0642719171226074, + "learning_rate": 3.925829012097725e-06, + "loss": 1.0364, + "step": 299 + }, + { + "epoch": 0.027055057041078596, + "grad_norm": 2.212279111536356, + "learning_rate": 3.928128474462e-06, + "loss": 0.9967, + "step": 300 + }, + { + "epoch": 0.027145240564548857, + "grad_norm": 1.9686479693419578, + "learning_rate": 3.930420284691115e-06, + "loss": 1.0868, + "step": 301 + }, + { + "epoch": 0.027235424088019117, + "grad_norm": 1.9640035785828909, + "learning_rate": 3.932704493545644e-06, + "loss": 1.1071, + "step": 302 + }, + { + "epoch": 0.02732560761148938, + "grad_norm": 2.158180257408546, + "learning_rate": 3.934981151282745e-06, + "loss": 1.1606, + "step": 303 + }, + { + "epoch": 0.027415791134959643, + "grad_norm": 1.9122646062260458, + "learning_rate": 3.9372503076628006e-06, + "loss": 1.0728, + "step": 304 + }, + { + "epoch": 0.027505974658429903, + "grad_norm": 2.122540003710718, + "learning_rate": 3.939512011955941e-06, + "loss": 1.095, + "step": 305 + }, + { + "epoch": 0.027596158181900168, + "grad_norm": 1.8918383149344782, + "learning_rate": 3.941766312948463e-06, + "loss": 1.0703, + "step": 306 + }, + { + "epoch": 0.02768634170537043, + "grad_norm": 2.0226831392377282, + "learning_rate": 3.944013258949147e-06, + "loss": 1.0427, + "step": 307 + }, + { + "epoch": 0.02777652522884069, + "grad_norm": 2.0615897048281617, + "learning_rate": 3.946252897795465e-06, + "loss": 0.8936, + "step": 308 + }, + { + "epoch": 0.027866708752310954, + "grad_norm": 2.5631430214557867, + "learning_rate": 3.9484852768596935e-06, + "loss": 1.0857, + "step": 309 + }, + { + "epoch": 0.027956892275781214, + "grad_norm": 1.6385889407462262, + "learning_rate": 3.950710443054923e-06, + "loss": 0.9758, + "step": 310 + }, + { + "epoch": 0.028047075799251475, + "grad_norm": 2.8110477512826875, + "learning_rate": 3.952928442840981e-06, + "loss": 0.9842, + "step": 311 + }, + { + "epoch": 0.02813725932272174, + "grad_norm": 1.9145786757165197, + "learning_rate": 3.955139322230243e-06, + "loss": 1.0923, + "step": 312 + }, + { + "epoch": 0.028227442846192, + "grad_norm": 2.206969997940131, + "learning_rate": 3.957343126793365e-06, + "loss": 1.0056, + "step": 313 + }, + { + "epoch": 0.02831762636966226, + "grad_norm": 2.1496436600028, + "learning_rate": 3.959539901664921e-06, + "loss": 1.0518, + "step": 314 + }, + { + "epoch": 0.028407809893132525, + "grad_norm": 11.623468113533418, + "learning_rate": 3.9617296915489425e-06, + "loss": 1.1676, + "step": 315 + }, + { + "epoch": 0.028497993416602786, + "grad_norm": 6.129358414745871, + "learning_rate": 3.963912540724387e-06, + "loss": 1.0139, + "step": 316 + }, + { + "epoch": 0.028588176940073047, + "grad_norm": 2.249650716161636, + "learning_rate": 3.966088493050501e-06, + "loss": 1.0735, + "step": 317 + }, + { + "epoch": 0.02867836046354331, + "grad_norm": 2.5063149802899924, + "learning_rate": 3.968257591972113e-06, + "loss": 1.0533, + "step": 318 + }, + { + "epoch": 0.028768543987013572, + "grad_norm": 1.7962548554101996, + "learning_rate": 3.970419880524835e-06, + "loss": 1.1484, + "step": 319 + }, + { + "epoch": 0.028858727510483833, + "grad_norm": 1.7885359609783553, + "learning_rate": 3.972575401340192e-06, + "loss": 1.0358, + "step": 320 + }, + { + "epoch": 0.028948911033954097, + "grad_norm": 2.0813186313436542, + "learning_rate": 3.974724196650656e-06, + "loss": 1.1045, + "step": 321 + }, + { + "epoch": 0.029039094557424358, + "grad_norm": 2.019623567990896, + "learning_rate": 3.976866308294617e-06, + "loss": 1.0574, + "step": 322 + }, + { + "epoch": 0.02912927808089462, + "grad_norm": 1.4567364350959509, + "learning_rate": 3.979001777721269e-06, + "loss": 1.0237, + "step": 323 + }, + { + "epoch": 0.029219461604364883, + "grad_norm": 2.267060555099012, + "learning_rate": 3.981130645995424e-06, + "loss": 1.0354, + "step": 324 + }, + { + "epoch": 0.029309645127835144, + "grad_norm": 2.051313872426755, + "learning_rate": 3.983252953802248e-06, + "loss": 1.1366, + "step": 325 + }, + { + "epoch": 0.029399828651305405, + "grad_norm": 0.7336836056019788, + "learning_rate": 3.9853687414519285e-06, + "loss": 0.8729, + "step": 326 + }, + { + "epoch": 0.02949001217477567, + "grad_norm": 2.577260467765584, + "learning_rate": 3.987478048884265e-06, + "loss": 1.1103, + "step": 327 + }, + { + "epoch": 0.02958019569824593, + "grad_norm": 2.104355328351518, + "learning_rate": 3.989580915673196e-06, + "loss": 1.0386, + "step": 328 + }, + { + "epoch": 0.02967037922171619, + "grad_norm": 1.8715925407960536, + "learning_rate": 3.991677381031255e-06, + "loss": 1.0383, + "step": 329 + }, + { + "epoch": 0.029760562745186455, + "grad_norm": 1.8585692511118839, + "learning_rate": 3.993767483813953e-06, + "loss": 1.1347, + "step": 330 + }, + { + "epoch": 0.029850746268656716, + "grad_norm": 3.1079714880657905, + "learning_rate": 3.995851262524104e-06, + "loss": 0.9764, + "step": 331 + }, + { + "epoch": 0.02994092979212698, + "grad_norm": 1.755450344286298, + "learning_rate": 3.997928755316079e-06, + "loss": 1.0575, + "step": 332 + }, + { + "epoch": 0.03003111331559724, + "grad_norm": 0.8344580491128464, + "learning_rate": 4e-06, + "loss": 0.8578, + "step": 333 + }, + { + "epoch": 0.030121296839067502, + "grad_norm": 1.8387045927684451, + "learning_rate": 3.999999914674486e-06, + "loss": 1.0268, + "step": 334 + }, + { + "epoch": 0.030211480362537766, + "grad_norm": 1.635848075513958, + "learning_rate": 3.999999658697952e-06, + "loss": 1.0432, + "step": 335 + }, + { + "epoch": 0.030301663886008027, + "grad_norm": 0.6783487794517165, + "learning_rate": 3.9999992320704185e-06, + "loss": 0.7963, + "step": 336 + }, + { + "epoch": 0.030391847409478288, + "grad_norm": 2.3180005407258686, + "learning_rate": 3.999998634791922e-06, + "loss": 1.0763, + "step": 337 + }, + { + "epoch": 0.030482030932948552, + "grad_norm": 1.9776898814016042, + "learning_rate": 3.999997866862515e-06, + "loss": 1.043, + "step": 338 + }, + { + "epoch": 0.030572214456418813, + "grad_norm": 1.6945971548684389, + "learning_rate": 3.999996928282262e-06, + "loss": 1.1467, + "step": 339 + }, + { + "epoch": 0.030662397979889074, + "grad_norm": 2.7958917860884034, + "learning_rate": 3.999995819051244e-06, + "loss": 1.0555, + "step": 340 + }, + { + "epoch": 0.030752581503359338, + "grad_norm": 1.9079720895113346, + "learning_rate": 3.9999945391695536e-06, + "loss": 1.0189, + "step": 341 + }, + { + "epoch": 0.0308427650268296, + "grad_norm": 1.6979076720703532, + "learning_rate": 3.999993088637302e-06, + "loss": 1.1427, + "step": 342 + }, + { + "epoch": 0.03093294855029986, + "grad_norm": 2.114979962012779, + "learning_rate": 3.999991467454612e-06, + "loss": 1.1249, + "step": 343 + }, + { + "epoch": 0.031023132073770124, + "grad_norm": 1.835245464652438, + "learning_rate": 3.999989675621622e-06, + "loss": 0.9754, + "step": 344 + }, + { + "epoch": 0.031113315597240385, + "grad_norm": 4.278122030721098, + "learning_rate": 3.999987713138485e-06, + "loss": 1.0645, + "step": 345 + }, + { + "epoch": 0.031203499120710645, + "grad_norm": 0.7657542995340894, + "learning_rate": 3.999985580005369e-06, + "loss": 0.9047, + "step": 346 + }, + { + "epoch": 0.031293682644180906, + "grad_norm": 2.2891130148876395, + "learning_rate": 3.999983276222455e-06, + "loss": 1.1307, + "step": 347 + }, + { + "epoch": 0.03138386616765117, + "grad_norm": 1.6608572164458961, + "learning_rate": 3.999980801789941e-06, + "loss": 1.072, + "step": 348 + }, + { + "epoch": 0.031474049691121435, + "grad_norm": 2.2229850002989124, + "learning_rate": 3.999978156708036e-06, + "loss": 1.0662, + "step": 349 + }, + { + "epoch": 0.031564233214591696, + "grad_norm": 2.69798159920064, + "learning_rate": 3.9999753409769675e-06, + "loss": 1.1708, + "step": 350 + }, + { + "epoch": 0.031654416738061956, + "grad_norm": 0.7027044310029292, + "learning_rate": 3.999972354596975e-06, + "loss": 0.8469, + "step": 351 + }, + { + "epoch": 0.03174460026153222, + "grad_norm": 1.8512354050807298, + "learning_rate": 3.999969197568314e-06, + "loss": 1.0131, + "step": 352 + }, + { + "epoch": 0.03183478378500248, + "grad_norm": 2.1204707177318673, + "learning_rate": 3.999965869891253e-06, + "loss": 1.1128, + "step": 353 + }, + { + "epoch": 0.03192496730847274, + "grad_norm": 1.6260605065840532, + "learning_rate": 3.999962371566075e-06, + "loss": 1.0884, + "step": 354 + }, + { + "epoch": 0.03201515083194301, + "grad_norm": 4.450590141683789, + "learning_rate": 3.999958702593082e-06, + "loss": 1.1298, + "step": 355 + }, + { + "epoch": 0.03210533435541327, + "grad_norm": 1.500063576940418, + "learning_rate": 3.999954862972583e-06, + "loss": 1.0853, + "step": 356 + }, + { + "epoch": 0.03219551787888353, + "grad_norm": 2.2316021553194614, + "learning_rate": 3.999950852704908e-06, + "loss": 1.0092, + "step": 357 + }, + { + "epoch": 0.03228570140235379, + "grad_norm": 1.609871500704108, + "learning_rate": 3.9999466717903995e-06, + "loss": 1.1181, + "step": 358 + }, + { + "epoch": 0.03237588492582405, + "grad_norm": 0.7129460979460562, + "learning_rate": 3.999942320229413e-06, + "loss": 0.8323, + "step": 359 + }, + { + "epoch": 0.03246606844929431, + "grad_norm": 2.2499852709817922, + "learning_rate": 3.99993779802232e-06, + "loss": 1.193, + "step": 360 + }, + { + "epoch": 0.03255625197276458, + "grad_norm": 2.385328521654718, + "learning_rate": 3.999933105169506e-06, + "loss": 1.045, + "step": 361 + }, + { + "epoch": 0.03264643549623484, + "grad_norm": 1.5079029431666822, + "learning_rate": 3.999928241671373e-06, + "loss": 1.0609, + "step": 362 + }, + { + "epoch": 0.0327366190197051, + "grad_norm": 5.033783458430852, + "learning_rate": 3.999923207528334e-06, + "loss": 0.9593, + "step": 363 + }, + { + "epoch": 0.03282680254317536, + "grad_norm": 1.7421627299510845, + "learning_rate": 3.9999180027408196e-06, + "loss": 1.0549, + "step": 364 + }, + { + "epoch": 0.03291698606664562, + "grad_norm": 2.106496358437767, + "learning_rate": 3.9999126273092735e-06, + "loss": 1.0391, + "step": 365 + }, + { + "epoch": 0.03300716959011588, + "grad_norm": 2.741819351567207, + "learning_rate": 3.999907081234156e-06, + "loss": 1.1493, + "step": 366 + }, + { + "epoch": 0.03309735311358615, + "grad_norm": 2.5313990042828642, + "learning_rate": 3.999901364515938e-06, + "loss": 1.0289, + "step": 367 + }, + { + "epoch": 0.03318753663705641, + "grad_norm": 2.0427437361146277, + "learning_rate": 3.999895477155108e-06, + "loss": 1.1083, + "step": 368 + }, + { + "epoch": 0.03327772016052667, + "grad_norm": 0.7329860294722883, + "learning_rate": 3.999889419152169e-06, + "loss": 0.8133, + "step": 369 + }, + { + "epoch": 0.03336790368399693, + "grad_norm": 1.9016109990099936, + "learning_rate": 3.999883190507638e-06, + "loss": 1.0577, + "step": 370 + }, + { + "epoch": 0.033458087207467194, + "grad_norm": 2.262355049506075, + "learning_rate": 3.999876791222044e-06, + "loss": 1.0464, + "step": 371 + }, + { + "epoch": 0.033548270730937454, + "grad_norm": 2.328683158992302, + "learning_rate": 3.999870221295936e-06, + "loss": 1.0148, + "step": 372 + }, + { + "epoch": 0.03363845425440772, + "grad_norm": 1.7011064238718678, + "learning_rate": 3.999863480729875e-06, + "loss": 1.0249, + "step": 373 + }, + { + "epoch": 0.03372863777787798, + "grad_norm": 2.43199033865139, + "learning_rate": 3.999856569524433e-06, + "loss": 0.9949, + "step": 374 + }, + { + "epoch": 0.033818821301348244, + "grad_norm": 2.7776227293187645, + "learning_rate": 3.999849487680202e-06, + "loss": 1.0, + "step": 375 + }, + { + "epoch": 0.033909004824818505, + "grad_norm": 4.144638937396926, + "learning_rate": 3.999842235197786e-06, + "loss": 0.993, + "step": 376 + }, + { + "epoch": 0.033999188348288766, + "grad_norm": 3.0456412140406632, + "learning_rate": 3.999834812077803e-06, + "loss": 0.9929, + "step": 377 + }, + { + "epoch": 0.034089371871759026, + "grad_norm": 1.560712936789861, + "learning_rate": 3.999827218320886e-06, + "loss": 1.1118, + "step": 378 + }, + { + "epoch": 0.034179555395229294, + "grad_norm": 3.1571726441904646, + "learning_rate": 3.999819453927685e-06, + "loss": 0.97, + "step": 379 + }, + { + "epoch": 0.034269738918699555, + "grad_norm": 2.082153087722632, + "learning_rate": 3.999811518898861e-06, + "loss": 1.0582, + "step": 380 + }, + { + "epoch": 0.034359922442169816, + "grad_norm": 2.730051775570592, + "learning_rate": 3.999803413235092e-06, + "loss": 1.0696, + "step": 381 + }, + { + "epoch": 0.03445010596564008, + "grad_norm": 1.682966712962469, + "learning_rate": 3.999795136937068e-06, + "loss": 1.1644, + "step": 382 + }, + { + "epoch": 0.03454028948911034, + "grad_norm": 2.055318409182032, + "learning_rate": 3.999786690005496e-06, + "loss": 1.0489, + "step": 383 + }, + { + "epoch": 0.0346304730125806, + "grad_norm": 1.7163399184774784, + "learning_rate": 3.999778072441098e-06, + "loss": 1.0384, + "step": 384 + }, + { + "epoch": 0.034720656536050866, + "grad_norm": 7.178707948429013, + "learning_rate": 3.999769284244608e-06, + "loss": 1.0871, + "step": 385 + }, + { + "epoch": 0.03481084005952113, + "grad_norm": 1.971304674085797, + "learning_rate": 3.999760325416775e-06, + "loss": 1.0712, + "step": 386 + }, + { + "epoch": 0.03490102358299139, + "grad_norm": 1.7764281209484623, + "learning_rate": 3.999751195958366e-06, + "loss": 1.1437, + "step": 387 + }, + { + "epoch": 0.03499120710646165, + "grad_norm": 0.7926601316724315, + "learning_rate": 3.999741895870157e-06, + "loss": 0.837, + "step": 388 + }, + { + "epoch": 0.03508139062993191, + "grad_norm": 20.464073147712355, + "learning_rate": 3.999732425152944e-06, + "loss": 1.1024, + "step": 389 + }, + { + "epoch": 0.03517157415340217, + "grad_norm": 1.5509043916100318, + "learning_rate": 3.999722783807533e-06, + "loss": 1.1167, + "step": 390 + }, + { + "epoch": 0.03526175767687244, + "grad_norm": 2.7805352310503206, + "learning_rate": 3.999712971834748e-06, + "loss": 0.9919, + "step": 391 + }, + { + "epoch": 0.0353519412003427, + "grad_norm": 2.331057551317889, + "learning_rate": 3.999702989235427e-06, + "loss": 1.0707, + "step": 392 + }, + { + "epoch": 0.03544212472381296, + "grad_norm": 1.8175873884681935, + "learning_rate": 3.999692836010419e-06, + "loss": 1.1476, + "step": 393 + }, + { + "epoch": 0.03553230824728322, + "grad_norm": 2.104910187243699, + "learning_rate": 3.999682512160593e-06, + "loss": 1.2036, + "step": 394 + }, + { + "epoch": 0.03562249177075348, + "grad_norm": 0.8742953596731742, + "learning_rate": 3.99967201768683e-06, + "loss": 0.8155, + "step": 395 + }, + { + "epoch": 0.03571267529422374, + "grad_norm": 1.6772156025476268, + "learning_rate": 3.999661352590023e-06, + "loss": 1.0475, + "step": 396 + }, + { + "epoch": 0.03580285881769401, + "grad_norm": 1.664734753854108, + "learning_rate": 3.999650516871083e-06, + "loss": 1.0622, + "step": 397 + }, + { + "epoch": 0.03589304234116427, + "grad_norm": 1.8790889024679742, + "learning_rate": 3.9996395105309365e-06, + "loss": 1.0903, + "step": 398 + }, + { + "epoch": 0.03598322586463453, + "grad_norm": 1.7376454079798016, + "learning_rate": 3.99962833357052e-06, + "loss": 1.0837, + "step": 399 + }, + { + "epoch": 0.03607340938810479, + "grad_norm": 1.874910861121696, + "learning_rate": 3.999616985990789e-06, + "loss": 1.043, + "step": 400 + }, + { + "epoch": 0.03616359291157505, + "grad_norm": 1.4928283912878133, + "learning_rate": 3.9996054677927104e-06, + "loss": 0.9976, + "step": 401 + }, + { + "epoch": 0.03625377643504532, + "grad_norm": 1.9339145586929913, + "learning_rate": 3.9995937789772675e-06, + "loss": 1.0032, + "step": 402 + }, + { + "epoch": 0.03634395995851558, + "grad_norm": 0.7203210164061095, + "learning_rate": 3.999581919545458e-06, + "loss": 0.8415, + "step": 403 + }, + { + "epoch": 0.03643414348198584, + "grad_norm": 2.6264383825673687, + "learning_rate": 3.9995698894982935e-06, + "loss": 1.0835, + "step": 404 + }, + { + "epoch": 0.0365243270054561, + "grad_norm": 4.68987651345169, + "learning_rate": 3.9995576888368e-06, + "loss": 1.111, + "step": 405 + }, + { + "epoch": 0.036614510528926364, + "grad_norm": 2.9141109350029737, + "learning_rate": 3.9995453175620194e-06, + "loss": 1.052, + "step": 406 + }, + { + "epoch": 0.036704694052396625, + "grad_norm": 2.9232644434097157, + "learning_rate": 3.999532775675007e-06, + "loss": 0.9205, + "step": 407 + }, + { + "epoch": 0.03679487757586689, + "grad_norm": 5.6924470618747165, + "learning_rate": 3.9995200631768326e-06, + "loss": 1.0376, + "step": 408 + }, + { + "epoch": 0.03688506109933715, + "grad_norm": 2.984511926515651, + "learning_rate": 3.9995071800685815e-06, + "loss": 1.1911, + "step": 409 + }, + { + "epoch": 0.036975244622807414, + "grad_norm": 2.0097566090966383, + "learning_rate": 3.999494126351352e-06, + "loss": 1.0346, + "step": 410 + }, + { + "epoch": 0.037065428146277675, + "grad_norm": 2.5292585097536744, + "learning_rate": 3.99948090202626e-06, + "loss": 1.0658, + "step": 411 + }, + { + "epoch": 0.037155611669747936, + "grad_norm": 1.5916358538669255, + "learning_rate": 3.999467507094431e-06, + "loss": 1.0661, + "step": 412 + }, + { + "epoch": 0.0372457951932182, + "grad_norm": 1.5531523519828467, + "learning_rate": 3.999453941557011e-06, + "loss": 0.9678, + "step": 413 + }, + { + "epoch": 0.037335978716688464, + "grad_norm": 0.7383105156159058, + "learning_rate": 3.999440205415154e-06, + "loss": 0.7935, + "step": 414 + }, + { + "epoch": 0.037426162240158725, + "grad_norm": 4.104362418578202, + "learning_rate": 3.999426298670035e-06, + "loss": 1.1095, + "step": 415 + }, + { + "epoch": 0.037516345763628986, + "grad_norm": 1.6326384383150092, + "learning_rate": 3.9994122213228385e-06, + "loss": 1.1043, + "step": 416 + }, + { + "epoch": 0.03760652928709925, + "grad_norm": 1.9100103342066301, + "learning_rate": 3.9993979733747675e-06, + "loss": 1.0924, + "step": 417 + }, + { + "epoch": 0.03769671281056951, + "grad_norm": 2.076178765128501, + "learning_rate": 3.999383554827037e-06, + "loss": 1.0645, + "step": 418 + }, + { + "epoch": 0.03778689633403977, + "grad_norm": 1.8005543729310096, + "learning_rate": 3.999368965680876e-06, + "loss": 1.0219, + "step": 419 + }, + { + "epoch": 0.037877079857510036, + "grad_norm": 2.104364392146494, + "learning_rate": 3.999354205937531e-06, + "loss": 1.1054, + "step": 420 + }, + { + "epoch": 0.0379672633809803, + "grad_norm": 1.8044738643082188, + "learning_rate": 3.999339275598261e-06, + "loss": 1.1074, + "step": 421 + }, + { + "epoch": 0.03805744690445056, + "grad_norm": 1.9474018933090624, + "learning_rate": 3.99932417466434e-06, + "loss": 1.0441, + "step": 422 + }, + { + "epoch": 0.03814763042792082, + "grad_norm": 3.852513565346552, + "learning_rate": 3.999308903137056e-06, + "loss": 0.9768, + "step": 423 + }, + { + "epoch": 0.03823781395139108, + "grad_norm": 1.9212968158167996, + "learning_rate": 3.999293461017711e-06, + "loss": 1.1087, + "step": 424 + }, + { + "epoch": 0.03832799747486134, + "grad_norm": 2.416814821459349, + "learning_rate": 3.9992778483076255e-06, + "loss": 1.0478, + "step": 425 + }, + { + "epoch": 0.03841818099833161, + "grad_norm": 2.0658603228300803, + "learning_rate": 3.99926206500813e-06, + "loss": 1.1144, + "step": 426 + }, + { + "epoch": 0.03850836452180187, + "grad_norm": 2.4775867931202424, + "learning_rate": 3.999246111120571e-06, + "loss": 0.9899, + "step": 427 + }, + { + "epoch": 0.03859854804527213, + "grad_norm": 2.1641444662387843, + "learning_rate": 3.999229986646311e-06, + "loss": 1.1314, + "step": 428 + }, + { + "epoch": 0.03868873156874239, + "grad_norm": 2.987648651383112, + "learning_rate": 3.999213691586723e-06, + "loss": 1.0279, + "step": 429 + }, + { + "epoch": 0.03877891509221265, + "grad_norm": 1.8209507686824848, + "learning_rate": 3.9991972259432e-06, + "loss": 1.1044, + "step": 430 + }, + { + "epoch": 0.03886909861568291, + "grad_norm": 3.387917276843023, + "learning_rate": 3.999180589717147e-06, + "loss": 1.1446, + "step": 431 + }, + { + "epoch": 0.03895928213915318, + "grad_norm": 3.990637790118746, + "learning_rate": 3.999163782909983e-06, + "loss": 0.9717, + "step": 432 + }, + { + "epoch": 0.03904946566262344, + "grad_norm": 2.0672402672870107, + "learning_rate": 3.99914680552314e-06, + "loss": 1.091, + "step": 433 + }, + { + "epoch": 0.0391396491860937, + "grad_norm": 1.7374988638236084, + "learning_rate": 3.999129657558069e-06, + "loss": 0.9926, + "step": 434 + }, + { + "epoch": 0.03922983270956396, + "grad_norm": 0.7245304445118691, + "learning_rate": 3.999112339016234e-06, + "loss": 0.8071, + "step": 435 + }, + { + "epoch": 0.03932001623303422, + "grad_norm": 2.592968742644151, + "learning_rate": 3.999094849899109e-06, + "loss": 1.0739, + "step": 436 + }, + { + "epoch": 0.039410199756504484, + "grad_norm": 1.66257736484956, + "learning_rate": 3.99907719020819e-06, + "loss": 0.9788, + "step": 437 + }, + { + "epoch": 0.03950038327997475, + "grad_norm": 1.6874327469599182, + "learning_rate": 3.999059359944982e-06, + "loss": 1.0867, + "step": 438 + }, + { + "epoch": 0.03959056680344501, + "grad_norm": 1.8657838022195272, + "learning_rate": 3.999041359111007e-06, + "loss": 1.117, + "step": 439 + }, + { + "epoch": 0.03968075032691527, + "grad_norm": 1.9447599798344395, + "learning_rate": 3.999023187707801e-06, + "loss": 1.1341, + "step": 440 + }, + { + "epoch": 0.039770933850385534, + "grad_norm": 1.6354593971971212, + "learning_rate": 3.999004845736913e-06, + "loss": 1.1224, + "step": 441 + }, + { + "epoch": 0.039861117373855795, + "grad_norm": 2.570103877450764, + "learning_rate": 3.9989863331999096e-06, + "loss": 1.1389, + "step": 442 + }, + { + "epoch": 0.039951300897326056, + "grad_norm": 2.422892793476009, + "learning_rate": 3.99896765009837e-06, + "loss": 0.9933, + "step": 443 + }, + { + "epoch": 0.040041484420796324, + "grad_norm": 1.6707981955748885, + "learning_rate": 3.998948796433888e-06, + "loss": 1.027, + "step": 444 + }, + { + "epoch": 0.040131667944266584, + "grad_norm": 1.5766180416187412, + "learning_rate": 3.998929772208073e-06, + "loss": 1.0991, + "step": 445 + }, + { + "epoch": 0.040221851467736845, + "grad_norm": 0.6852754452242217, + "learning_rate": 3.998910577422547e-06, + "loss": 0.834, + "step": 446 + }, + { + "epoch": 0.040312034991207106, + "grad_norm": 1.6217776407115327, + "learning_rate": 3.99889121207895e-06, + "loss": 1.1038, + "step": 447 + }, + { + "epoch": 0.04040221851467737, + "grad_norm": 2.003361499649728, + "learning_rate": 3.9988716761789324e-06, + "loss": 1.0854, + "step": 448 + }, + { + "epoch": 0.04049240203814763, + "grad_norm": 2.0495786792164346, + "learning_rate": 3.998851969724161e-06, + "loss": 0.9991, + "step": 449 + }, + { + "epoch": 0.040582585561617895, + "grad_norm": 2.2503909725009494, + "learning_rate": 3.998832092716319e-06, + "loss": 1.1126, + "step": 450 + }, + { + "epoch": 0.040672769085088156, + "grad_norm": 1.700658219197726, + "learning_rate": 3.998812045157102e-06, + "loss": 1.0415, + "step": 451 + }, + { + "epoch": 0.04076295260855842, + "grad_norm": 1.838544836572899, + "learning_rate": 3.998791827048219e-06, + "loss": 1.1088, + "step": 452 + }, + { + "epoch": 0.04085313613202868, + "grad_norm": 1.5224962505676505, + "learning_rate": 3.998771438391396e-06, + "loss": 1.0064, + "step": 453 + }, + { + "epoch": 0.04094331965549894, + "grad_norm": 0.6399347727289287, + "learning_rate": 3.9987508791883725e-06, + "loss": 0.8122, + "step": 454 + }, + { + "epoch": 0.0410335031789692, + "grad_norm": 2.0189368666385206, + "learning_rate": 3.998730149440904e-06, + "loss": 1.1793, + "step": 455 + }, + { + "epoch": 0.04112368670243947, + "grad_norm": 2.6096708392713555, + "learning_rate": 3.998709249150758e-06, + "loss": 1.112, + "step": 456 + }, + { + "epoch": 0.04121387022590973, + "grad_norm": 1.671463461163601, + "learning_rate": 3.998688178319717e-06, + "loss": 1.0853, + "step": 457 + }, + { + "epoch": 0.04130405374937999, + "grad_norm": 2.2754883839313473, + "learning_rate": 3.9986669369495805e-06, + "loss": 1.0495, + "step": 458 + }, + { + "epoch": 0.04139423727285025, + "grad_norm": 2.000457949660981, + "learning_rate": 3.998645525042161e-06, + "loss": 1.117, + "step": 459 + }, + { + "epoch": 0.04148442079632051, + "grad_norm": 3.9439020336124506, + "learning_rate": 3.998623942599284e-06, + "loss": 1.1299, + "step": 460 + }, + { + "epoch": 0.04157460431979077, + "grad_norm": 4.967018065513227, + "learning_rate": 3.998602189622793e-06, + "loss": 1.0568, + "step": 461 + }, + { + "epoch": 0.04166478784326104, + "grad_norm": 2.7340381523824537, + "learning_rate": 3.998580266114542e-06, + "loss": 1.0989, + "step": 462 + }, + { + "epoch": 0.0417549713667313, + "grad_norm": 3.064474286893153, + "learning_rate": 3.998558172076404e-06, + "loss": 0.9674, + "step": 463 + }, + { + "epoch": 0.04184515489020156, + "grad_norm": 1.788578484374179, + "learning_rate": 3.998535907510262e-06, + "loss": 0.9978, + "step": 464 + }, + { + "epoch": 0.04193533841367182, + "grad_norm": 1.7278138196129122, + "learning_rate": 3.998513472418016e-06, + "loss": 1.0727, + "step": 465 + }, + { + "epoch": 0.04202552193714208, + "grad_norm": 1.7689099431595343, + "learning_rate": 3.998490866801582e-06, + "loss": 1.0951, + "step": 466 + }, + { + "epoch": 0.04211570546061234, + "grad_norm": 2.9713821788718935, + "learning_rate": 3.998468090662886e-06, + "loss": 1.0522, + "step": 467 + }, + { + "epoch": 0.04220588898408261, + "grad_norm": 1.5576879024896162, + "learning_rate": 3.998445144003874e-06, + "loss": 0.9578, + "step": 468 + }, + { + "epoch": 0.04229607250755287, + "grad_norm": 2.0178789651223634, + "learning_rate": 3.998422026826504e-06, + "loss": 1.0901, + "step": 469 + }, + { + "epoch": 0.04238625603102313, + "grad_norm": 2.4341049024909682, + "learning_rate": 3.998398739132746e-06, + "loss": 1.0935, + "step": 470 + }, + { + "epoch": 0.04247643955449339, + "grad_norm": 1.8922042556134229, + "learning_rate": 3.99837528092459e-06, + "loss": 1.0188, + "step": 471 + }, + { + "epoch": 0.042566623077963654, + "grad_norm": 1.795389041978784, + "learning_rate": 3.998351652204034e-06, + "loss": 1.1095, + "step": 472 + }, + { + "epoch": 0.042656806601433915, + "grad_norm": 2.0624594250936115, + "learning_rate": 3.998327852973098e-06, + "loss": 1.1061, + "step": 473 + }, + { + "epoch": 0.04274699012490418, + "grad_norm": 2.1088206375131033, + "learning_rate": 3.99830388323381e-06, + "loss": 1.0124, + "step": 474 + }, + { + "epoch": 0.042837173648374444, + "grad_norm": 7.151634093599884, + "learning_rate": 3.998279742988216e-06, + "loss": 1.093, + "step": 475 + }, + { + "epoch": 0.042927357171844704, + "grad_norm": 1.8614036969017107, + "learning_rate": 3.998255432238377e-06, + "loss": 1.0319, + "step": 476 + }, + { + "epoch": 0.043017540695314965, + "grad_norm": 2.1168253120783183, + "learning_rate": 3.9982309509863656e-06, + "loss": 1.135, + "step": 477 + }, + { + "epoch": 0.043107724218785226, + "grad_norm": 2.2168686103137207, + "learning_rate": 3.998206299234272e-06, + "loss": 1.0097, + "step": 478 + }, + { + "epoch": 0.04319790774225549, + "grad_norm": 1.8176790761551425, + "learning_rate": 3.998181476984198e-06, + "loss": 1.078, + "step": 479 + }, + { + "epoch": 0.043288091265725755, + "grad_norm": 2.272100206045824, + "learning_rate": 3.998156484238263e-06, + "loss": 0.9871, + "step": 480 + }, + { + "epoch": 0.043378274789196015, + "grad_norm": 2.486622112001151, + "learning_rate": 3.998131320998599e-06, + "loss": 1.0012, + "step": 481 + }, + { + "epoch": 0.043468458312666276, + "grad_norm": 1.9542076467136238, + "learning_rate": 3.998105987267353e-06, + "loss": 1.1095, + "step": 482 + }, + { + "epoch": 0.04355864183613654, + "grad_norm": 2.206006485681346, + "learning_rate": 3.998080483046687e-06, + "loss": 0.9895, + "step": 483 + }, + { + "epoch": 0.0436488253596068, + "grad_norm": 1.5274517674799224, + "learning_rate": 3.998054808338776e-06, + "loss": 1.0122, + "step": 484 + }, + { + "epoch": 0.04373900888307706, + "grad_norm": 1.6470024768924212, + "learning_rate": 3.998028963145812e-06, + "loss": 1.0248, + "step": 485 + }, + { + "epoch": 0.043829192406547327, + "grad_norm": 2.0375933853934693, + "learning_rate": 3.99800294747e-06, + "loss": 1.0761, + "step": 486 + }, + { + "epoch": 0.04391937593001759, + "grad_norm": 1.7019888211565666, + "learning_rate": 3.99797676131356e-06, + "loss": 1.0978, + "step": 487 + }, + { + "epoch": 0.04400955945348785, + "grad_norm": 2.204536905424064, + "learning_rate": 3.997950404678726e-06, + "loss": 0.9218, + "step": 488 + }, + { + "epoch": 0.04409974297695811, + "grad_norm": 6.818823443191213, + "learning_rate": 3.997923877567746e-06, + "loss": 1.1751, + "step": 489 + }, + { + "epoch": 0.04418992650042837, + "grad_norm": 1.7041238777863759, + "learning_rate": 3.9978971799828855e-06, + "loss": 0.8942, + "step": 490 + }, + { + "epoch": 0.04428011002389863, + "grad_norm": 2.12082621255315, + "learning_rate": 3.997870311926421e-06, + "loss": 1.0491, + "step": 491 + }, + { + "epoch": 0.0443702935473689, + "grad_norm": 1.9950261256292758, + "learning_rate": 3.997843273400645e-06, + "loss": 1.019, + "step": 492 + }, + { + "epoch": 0.04446047707083916, + "grad_norm": 1.8478147668564928, + "learning_rate": 3.997816064407865e-06, + "loss": 1.0401, + "step": 493 + }, + { + "epoch": 0.04455066059430942, + "grad_norm": 1.6035703401140093, + "learning_rate": 3.997788684950402e-06, + "loss": 1.0824, + "step": 494 + }, + { + "epoch": 0.04464084411777968, + "grad_norm": 2.217138753063437, + "learning_rate": 3.997761135030593e-06, + "loss": 1.1032, + "step": 495 + }, + { + "epoch": 0.04473102764124994, + "grad_norm": 1.6698080102125061, + "learning_rate": 3.997733414650789e-06, + "loss": 0.9631, + "step": 496 + }, + { + "epoch": 0.0448212111647202, + "grad_norm": 2.305993075649059, + "learning_rate": 3.9977055238133554e-06, + "loss": 1.022, + "step": 497 + }, + { + "epoch": 0.04491139468819047, + "grad_norm": 1.7570029195805774, + "learning_rate": 3.99767746252067e-06, + "loss": 0.9397, + "step": 498 + }, + { + "epoch": 0.04500157821166073, + "grad_norm": 1.818725221340584, + "learning_rate": 3.997649230775129e-06, + "loss": 1.0622, + "step": 499 + }, + { + "epoch": 0.04509176173513099, + "grad_norm": 2.8563439819099457, + "learning_rate": 3.9976208285791395e-06, + "loss": 1.0004, + "step": 500 + }, + { + "epoch": 0.04518194525860125, + "grad_norm": 0.7787658488343197, + "learning_rate": 3.997592255935127e-06, + "loss": 0.9086, + "step": 501 + }, + { + "epoch": 0.045272128782071513, + "grad_norm": 1.7912487015649599, + "learning_rate": 3.997563512845529e-06, + "loss": 1.0875, + "step": 502 + }, + { + "epoch": 0.045362312305541774, + "grad_norm": 1.9789620410812037, + "learning_rate": 3.9975345993127975e-06, + "loss": 1.018, + "step": 503 + }, + { + "epoch": 0.04545249582901204, + "grad_norm": 1.901324741416786, + "learning_rate": 3.9975055153393985e-06, + "loss": 1.074, + "step": 504 + }, + { + "epoch": 0.0455426793524823, + "grad_norm": 2.8855607493722726, + "learning_rate": 3.997476260927816e-06, + "loss": 1.0816, + "step": 505 + }, + { + "epoch": 0.045632862875952564, + "grad_norm": 2.9209795717062446, + "learning_rate": 3.997446836080545e-06, + "loss": 1.037, + "step": 506 + }, + { + "epoch": 0.045723046399422825, + "grad_norm": 3.53599836741276, + "learning_rate": 3.997417240800095e-06, + "loss": 1.1098, + "step": 507 + }, + { + "epoch": 0.045813229922893085, + "grad_norm": 1.7168841551379337, + "learning_rate": 3.997387475088994e-06, + "loss": 1.158, + "step": 508 + }, + { + "epoch": 0.045903413446363346, + "grad_norm": 3.3348293920798686, + "learning_rate": 3.99735753894978e-06, + "loss": 1.1602, + "step": 509 + }, + { + "epoch": 0.045993596969833614, + "grad_norm": 1.9593074148102922, + "learning_rate": 3.997327432385006e-06, + "loss": 1.0603, + "step": 510 + }, + { + "epoch": 0.046083780493303875, + "grad_norm": 1.7165963985666226, + "learning_rate": 3.997297155397244e-06, + "loss": 0.9934, + "step": 511 + }, + { + "epoch": 0.046173964016774136, + "grad_norm": 5.27772377164994, + "learning_rate": 3.997266707989074e-06, + "loss": 0.9801, + "step": 512 + }, + { + "epoch": 0.046264147540244396, + "grad_norm": 2.439868461002484, + "learning_rate": 3.997236090163097e-06, + "loss": 0.981, + "step": 513 + }, + { + "epoch": 0.04635433106371466, + "grad_norm": 3.0359284782177456, + "learning_rate": 3.9972053019219235e-06, + "loss": 1.1919, + "step": 514 + }, + { + "epoch": 0.04644451458718492, + "grad_norm": 2.2130946766441166, + "learning_rate": 3.997174343268181e-06, + "loss": 1.0781, + "step": 515 + }, + { + "epoch": 0.046534698110655186, + "grad_norm": 2.8252441739585223, + "learning_rate": 3.9971432142045115e-06, + "loss": 1.0699, + "step": 516 + }, + { + "epoch": 0.04662488163412545, + "grad_norm": 0.8134655350650936, + "learning_rate": 3.99711191473357e-06, + "loss": 0.9008, + "step": 517 + }, + { + "epoch": 0.04671506515759571, + "grad_norm": 1.7111926302269187, + "learning_rate": 3.99708044485803e-06, + "loss": 1.0777, + "step": 518 + }, + { + "epoch": 0.04680524868106597, + "grad_norm": 1.7273340356934057, + "learning_rate": 3.997048804580574e-06, + "loss": 1.0677, + "step": 519 + }, + { + "epoch": 0.04689543220453623, + "grad_norm": 2.418082929389677, + "learning_rate": 3.997016993903901e-06, + "loss": 1.0088, + "step": 520 + }, + { + "epoch": 0.04698561572800649, + "grad_norm": 1.694478968318969, + "learning_rate": 3.996985012830728e-06, + "loss": 1.1378, + "step": 521 + }, + { + "epoch": 0.04707579925147676, + "grad_norm": 1.5224629733088537, + "learning_rate": 3.996952861363782e-06, + "loss": 1.0043, + "step": 522 + }, + { + "epoch": 0.04716598277494702, + "grad_norm": 2.1885104434221962, + "learning_rate": 3.9969205395058064e-06, + "loss": 1.0662, + "step": 523 + }, + { + "epoch": 0.04725616629841728, + "grad_norm": 4.028371330169361, + "learning_rate": 3.99688804725956e-06, + "loss": 1.0564, + "step": 524 + }, + { + "epoch": 0.04734634982188754, + "grad_norm": 1.7380987521606974, + "learning_rate": 3.996855384627815e-06, + "loss": 1.0123, + "step": 525 + }, + { + "epoch": 0.0474365333453578, + "grad_norm": 4.9430236823000655, + "learning_rate": 3.996822551613357e-06, + "loss": 1.0948, + "step": 526 + }, + { + "epoch": 0.04752671686882806, + "grad_norm": 1.938303934477325, + "learning_rate": 3.996789548218989e-06, + "loss": 0.9895, + "step": 527 + }, + { + "epoch": 0.04761690039229833, + "grad_norm": 1.6727216484188243, + "learning_rate": 3.996756374447526e-06, + "loss": 0.9474, + "step": 528 + }, + { + "epoch": 0.04770708391576859, + "grad_norm": 2.605873545438161, + "learning_rate": 3.9967230303018005e-06, + "loss": 1.0614, + "step": 529 + }, + { + "epoch": 0.04779726743923885, + "grad_norm": 2.158649491630372, + "learning_rate": 3.996689515784655e-06, + "loss": 1.0447, + "step": 530 + }, + { + "epoch": 0.04788745096270911, + "grad_norm": 2.5088284059901595, + "learning_rate": 3.996655830898951e-06, + "loss": 1.0327, + "step": 531 + }, + { + "epoch": 0.04797763448617937, + "grad_norm": 1.8593928873179464, + "learning_rate": 3.996621975647562e-06, + "loss": 1.0608, + "step": 532 + }, + { + "epoch": 0.04806781800964964, + "grad_norm": 13.261987409833933, + "learning_rate": 3.996587950033377e-06, + "loss": 0.9917, + "step": 533 + }, + { + "epoch": 0.0481580015331199, + "grad_norm": 2.5613732070294786, + "learning_rate": 3.996553754059299e-06, + "loss": 0.9891, + "step": 534 + }, + { + "epoch": 0.04824818505659016, + "grad_norm": 1.9813687839459886, + "learning_rate": 3.996519387728245e-06, + "loss": 1.0595, + "step": 535 + }, + { + "epoch": 0.04833836858006042, + "grad_norm": 1.6209591729248787, + "learning_rate": 3.9964848510431495e-06, + "loss": 1.0426, + "step": 536 + }, + { + "epoch": 0.048428552103530684, + "grad_norm": 3.7150400071794447, + "learning_rate": 3.996450144006957e-06, + "loss": 1.1176, + "step": 537 + }, + { + "epoch": 0.048518735627000945, + "grad_norm": 2.026134681691237, + "learning_rate": 3.99641526662263e-06, + "loss": 1.037, + "step": 538 + }, + { + "epoch": 0.04860891915047121, + "grad_norm": 1.8382384470744302, + "learning_rate": 3.996380218893145e-06, + "loss": 1.1413, + "step": 539 + }, + { + "epoch": 0.04869910267394147, + "grad_norm": 2.0161884318799865, + "learning_rate": 3.996345000821491e-06, + "loss": 1.0855, + "step": 540 + }, + { + "epoch": 0.048789286197411734, + "grad_norm": 2.209771917319683, + "learning_rate": 3.996309612410674e-06, + "loss": 1.1692, + "step": 541 + }, + { + "epoch": 0.048879469720881995, + "grad_norm": 1.8290089483485417, + "learning_rate": 3.996274053663713e-06, + "loss": 0.9945, + "step": 542 + }, + { + "epoch": 0.048969653244352256, + "grad_norm": 0.7833877307861132, + "learning_rate": 3.996238324583643e-06, + "loss": 0.8563, + "step": 543 + }, + { + "epoch": 0.049059836767822516, + "grad_norm": 2.0773808573908763, + "learning_rate": 3.996202425173512e-06, + "loss": 1.1222, + "step": 544 + }, + { + "epoch": 0.049150020291292784, + "grad_norm": 3.6546344285155676, + "learning_rate": 3.996166355436383e-06, + "loss": 1.0164, + "step": 545 + }, + { + "epoch": 0.049240203814763045, + "grad_norm": 3.921878039598237, + "learning_rate": 3.996130115375333e-06, + "loss": 1.066, + "step": 546 + }, + { + "epoch": 0.049330387338233306, + "grad_norm": 0.8832916714782105, + "learning_rate": 3.996093704993456e-06, + "loss": 0.8713, + "step": 547 + }, + { + "epoch": 0.04942057086170357, + "grad_norm": 2.0038644886684662, + "learning_rate": 3.996057124293857e-06, + "loss": 1.1416, + "step": 548 + }, + { + "epoch": 0.04951075438517383, + "grad_norm": 1.4249121956794806, + "learning_rate": 3.996020373279659e-06, + "loss": 1.1082, + "step": 549 + }, + { + "epoch": 0.04960093790864409, + "grad_norm": 4.424702633278799, + "learning_rate": 3.995983451953996e-06, + "loss": 1.0666, + "step": 550 + }, + { + "epoch": 0.049691121432114356, + "grad_norm": 1.784338832829409, + "learning_rate": 3.99594636032002e-06, + "loss": 1.0824, + "step": 551 + }, + { + "epoch": 0.04978130495558462, + "grad_norm": 2.366531331486595, + "learning_rate": 3.995909098380894e-06, + "loss": 0.945, + "step": 552 + }, + { + "epoch": 0.04987148847905488, + "grad_norm": 1.8528459181195058, + "learning_rate": 3.995871666139799e-06, + "loss": 0.9426, + "step": 553 + }, + { + "epoch": 0.04996167200252514, + "grad_norm": 2.9175853553975544, + "learning_rate": 3.995834063599928e-06, + "loss": 1.0348, + "step": 554 + }, + { + "epoch": 0.0500518555259954, + "grad_norm": 1.783252861607859, + "learning_rate": 3.99579629076449e-06, + "loss": 1.0836, + "step": 555 + }, + { + "epoch": 0.05014203904946566, + "grad_norm": 3.375797036329815, + "learning_rate": 3.9957583476367084e-06, + "loss": 1.03, + "step": 556 + }, + { + "epoch": 0.05023222257293593, + "grad_norm": 1.9376002255019884, + "learning_rate": 3.995720234219819e-06, + "loss": 1.0596, + "step": 557 + }, + { + "epoch": 0.05032240609640619, + "grad_norm": 2.1974902228961386, + "learning_rate": 3.995681950517075e-06, + "loss": 0.9639, + "step": 558 + }, + { + "epoch": 0.05041258961987645, + "grad_norm": 2.7639393815157387, + "learning_rate": 3.995643496531743e-06, + "loss": 1.0762, + "step": 559 + }, + { + "epoch": 0.05050277314334671, + "grad_norm": 1.9548904984386262, + "learning_rate": 3.9956048722671044e-06, + "loss": 1.0661, + "step": 560 + }, + { + "epoch": 0.05059295666681697, + "grad_norm": 1.7879585958291009, + "learning_rate": 3.995566077726454e-06, + "loss": 1.0362, + "step": 561 + }, + { + "epoch": 0.05068314019028723, + "grad_norm": 1.720237712291223, + "learning_rate": 3.995527112913103e-06, + "loss": 1.0182, + "step": 562 + }, + { + "epoch": 0.0507733237137575, + "grad_norm": 2.4839089394309055, + "learning_rate": 3.995487977830375e-06, + "loss": 1.0151, + "step": 563 + }, + { + "epoch": 0.05086350723722776, + "grad_norm": 1.712660875212392, + "learning_rate": 3.9954486724816105e-06, + "loss": 1.1531, + "step": 564 + }, + { + "epoch": 0.05095369076069802, + "grad_norm": 2.3194573364610185, + "learning_rate": 3.995409196870161e-06, + "loss": 1.0979, + "step": 565 + }, + { + "epoch": 0.05104387428416828, + "grad_norm": 0.7916684317987406, + "learning_rate": 3.995369550999398e-06, + "loss": 0.8168, + "step": 566 + }, + { + "epoch": 0.05113405780763854, + "grad_norm": 3.4349901226406985, + "learning_rate": 3.995329734872702e-06, + "loss": 1.0813, + "step": 567 + }, + { + "epoch": 0.051224241331108804, + "grad_norm": 1.5597412932896269, + "learning_rate": 3.9952897484934706e-06, + "loss": 1.1728, + "step": 568 + }, + { + "epoch": 0.05131442485457907, + "grad_norm": 2.29655506215362, + "learning_rate": 3.995249591865115e-06, + "loss": 1.0601, + "step": 569 + }, + { + "epoch": 0.05140460837804933, + "grad_norm": 1.981111260499281, + "learning_rate": 3.995209264991063e-06, + "loss": 1.015, + "step": 570 + }, + { + "epoch": 0.05149479190151959, + "grad_norm": 2.556505681987128, + "learning_rate": 3.995168767874756e-06, + "loss": 1.1062, + "step": 571 + }, + { + "epoch": 0.051584975424989854, + "grad_norm": 1.6698791855380475, + "learning_rate": 3.995128100519648e-06, + "loss": 1.0461, + "step": 572 + }, + { + "epoch": 0.051675158948460115, + "grad_norm": 1.8627027337253534, + "learning_rate": 3.995087262929209e-06, + "loss": 1.0679, + "step": 573 + }, + { + "epoch": 0.051765342471930376, + "grad_norm": 1.8301689915200077, + "learning_rate": 3.995046255106925e-06, + "loss": 1.0693, + "step": 574 + }, + { + "epoch": 0.05185552599540064, + "grad_norm": 2.9706329798897233, + "learning_rate": 3.995005077056293e-06, + "loss": 1.1205, + "step": 575 + }, + { + "epoch": 0.051945709518870904, + "grad_norm": 2.812285181954546, + "learning_rate": 3.9949637287808284e-06, + "loss": 0.9476, + "step": 576 + }, + { + "epoch": 0.052035893042341165, + "grad_norm": 2.0872571107188347, + "learning_rate": 3.994922210284057e-06, + "loss": 1.0056, + "step": 577 + }, + { + "epoch": 0.052126076565811426, + "grad_norm": 2.2468798513983503, + "learning_rate": 3.994880521569524e-06, + "loss": 1.0895, + "step": 578 + }, + { + "epoch": 0.05221626008928169, + "grad_norm": 2.015673614625399, + "learning_rate": 3.994838662640785e-06, + "loss": 1.068, + "step": 579 + }, + { + "epoch": 0.05230644361275195, + "grad_norm": 2.9704490216871076, + "learning_rate": 3.9947966335014116e-06, + "loss": 1.0947, + "step": 580 + }, + { + "epoch": 0.052396627136222215, + "grad_norm": 2.005245958115628, + "learning_rate": 3.99475443415499e-06, + "loss": 1.0323, + "step": 581 + }, + { + "epoch": 0.052486810659692476, + "grad_norm": 1.7128568709037035, + "learning_rate": 3.994712064605121e-06, + "loss": 1.054, + "step": 582 + }, + { + "epoch": 0.05257699418316274, + "grad_norm": 1.9745637339402569, + "learning_rate": 3.99466952485542e-06, + "loss": 1.2023, + "step": 583 + }, + { + "epoch": 0.052667177706633, + "grad_norm": 2.4181621024331266, + "learning_rate": 3.994626814909518e-06, + "loss": 1.0711, + "step": 584 + }, + { + "epoch": 0.05275736123010326, + "grad_norm": 2.1632224976292354, + "learning_rate": 3.994583934771056e-06, + "loss": 1.0412, + "step": 585 + }, + { + "epoch": 0.05284754475357352, + "grad_norm": 1.6343065207378316, + "learning_rate": 3.9945408844436955e-06, + "loss": 1.1177, + "step": 586 + }, + { + "epoch": 0.05293772827704379, + "grad_norm": 1.9655962919191714, + "learning_rate": 3.994497663931109e-06, + "loss": 1.0867, + "step": 587 + }, + { + "epoch": 0.05302791180051405, + "grad_norm": 2.0244767160468697, + "learning_rate": 3.994454273236984e-06, + "loss": 1.0775, + "step": 588 + }, + { + "epoch": 0.05311809532398431, + "grad_norm": 1.6315425332803408, + "learning_rate": 3.994410712365023e-06, + "loss": 1.0639, + "step": 589 + }, + { + "epoch": 0.05320827884745457, + "grad_norm": 4.2239598636020785, + "learning_rate": 3.994366981318943e-06, + "loss": 1.1631, + "step": 590 + }, + { + "epoch": 0.05329846237092483, + "grad_norm": 1.7791670281843879, + "learning_rate": 3.9943230801024765e-06, + "loss": 1.0651, + "step": 591 + }, + { + "epoch": 0.05338864589439509, + "grad_norm": 2.0942588301735796, + "learning_rate": 3.9942790087193666e-06, + "loss": 1.101, + "step": 592 + }, + { + "epoch": 0.05347882941786536, + "grad_norm": 2.1735937329763204, + "learning_rate": 3.994234767173376e-06, + "loss": 1.0511, + "step": 593 + }, + { + "epoch": 0.05356901294133562, + "grad_norm": 2.18127350890162, + "learning_rate": 3.994190355468279e-06, + "loss": 0.9483, + "step": 594 + }, + { + "epoch": 0.05365919646480588, + "grad_norm": 1.6273602437663681, + "learning_rate": 3.994145773607865e-06, + "loss": 1.0611, + "step": 595 + }, + { + "epoch": 0.05374937998827614, + "grad_norm": 1.5926114671822473, + "learning_rate": 3.994101021595938e-06, + "loss": 1.0317, + "step": 596 + }, + { + "epoch": 0.0538395635117464, + "grad_norm": 1.888534903860494, + "learning_rate": 3.9940560994363165e-06, + "loss": 1.0769, + "step": 597 + }, + { + "epoch": 0.05392974703521666, + "grad_norm": 2.3089396932399606, + "learning_rate": 3.994011007132833e-06, + "loss": 0.9358, + "step": 598 + }, + { + "epoch": 0.05401993055868693, + "grad_norm": 0.7053826432680117, + "learning_rate": 3.993965744689337e-06, + "loss": 0.8865, + "step": 599 + }, + { + "epoch": 0.05411011408215719, + "grad_norm": 2.069049608368748, + "learning_rate": 3.993920312109687e-06, + "loss": 1.0426, + "step": 600 + }, + { + "epoch": 0.05420029760562745, + "grad_norm": 2.2894491969193504, + "learning_rate": 3.993874709397764e-06, + "loss": 1.0545, + "step": 601 + }, + { + "epoch": 0.05429048112909771, + "grad_norm": 2.0736348682748003, + "learning_rate": 3.993828936557454e-06, + "loss": 1.0477, + "step": 602 + }, + { + "epoch": 0.054380664652567974, + "grad_norm": 1.687914832660423, + "learning_rate": 3.993782993592667e-06, + "loss": 1.0189, + "step": 603 + }, + { + "epoch": 0.054470848176038235, + "grad_norm": 1.8517782994957572, + "learning_rate": 3.993736880507321e-06, + "loss": 1.0984, + "step": 604 + }, + { + "epoch": 0.0545610316995085, + "grad_norm": 2.5238239948975165, + "learning_rate": 3.99369059730535e-06, + "loss": 1.005, + "step": 605 + }, + { + "epoch": 0.05465121522297876, + "grad_norm": 1.854469174628475, + "learning_rate": 3.993644143990706e-06, + "loss": 1.0627, + "step": 606 + }, + { + "epoch": 0.054741398746449024, + "grad_norm": 2.1542658868464843, + "learning_rate": 3.99359752056735e-06, + "loss": 0.9793, + "step": 607 + }, + { + "epoch": 0.054831582269919285, + "grad_norm": 1.9627904762037012, + "learning_rate": 3.993550727039261e-06, + "loss": 0.9136, + "step": 608 + }, + { + "epoch": 0.054921765793389546, + "grad_norm": 3.995453635040189, + "learning_rate": 3.993503763410431e-06, + "loss": 1.077, + "step": 609 + }, + { + "epoch": 0.05501194931685981, + "grad_norm": 1.7092660900617938, + "learning_rate": 3.9934566296848686e-06, + "loss": 0.9728, + "step": 610 + }, + { + "epoch": 0.055102132840330074, + "grad_norm": 2.308007490202123, + "learning_rate": 3.993409325866595e-06, + "loss": 1.1393, + "step": 611 + }, + { + "epoch": 0.055192316363800335, + "grad_norm": 0.6326717644068377, + "learning_rate": 3.993361851959645e-06, + "loss": 0.835, + "step": 612 + }, + { + "epoch": 0.055282499887270596, + "grad_norm": 1.5984058067483597, + "learning_rate": 3.993314207968071e-06, + "loss": 1.1009, + "step": 613 + }, + { + "epoch": 0.05537268341074086, + "grad_norm": 2.006713685258989, + "learning_rate": 3.993266393895938e-06, + "loss": 1.0477, + "step": 614 + }, + { + "epoch": 0.05546286693421112, + "grad_norm": 2.862504757539885, + "learning_rate": 3.993218409747326e-06, + "loss": 1.0601, + "step": 615 + }, + { + "epoch": 0.05555305045768138, + "grad_norm": 1.5312422148837492, + "learning_rate": 3.993170255526328e-06, + "loss": 1.0533, + "step": 616 + }, + { + "epoch": 0.055643233981151646, + "grad_norm": 1.6815201099201633, + "learning_rate": 3.993121931237054e-06, + "loss": 1.088, + "step": 617 + }, + { + "epoch": 0.05573341750462191, + "grad_norm": 2.2532832244891905, + "learning_rate": 3.993073436883627e-06, + "loss": 1.0774, + "step": 618 + }, + { + "epoch": 0.05582360102809217, + "grad_norm": 2.211439635068523, + "learning_rate": 3.993024772470184e-06, + "loss": 1.0404, + "step": 619 + }, + { + "epoch": 0.05591378455156243, + "grad_norm": 151.0463892939848, + "learning_rate": 3.992975938000878e-06, + "loss": 1.0425, + "step": 620 + }, + { + "epoch": 0.05600396807503269, + "grad_norm": 0.6594431936788155, + "learning_rate": 3.992926933479876e-06, + "loss": 0.8252, + "step": 621 + }, + { + "epoch": 0.05609415159850295, + "grad_norm": 2.6968443125087225, + "learning_rate": 3.9928777589113595e-06, + "loss": 1.0839, + "step": 622 + }, + { + "epoch": 0.05618433512197322, + "grad_norm": 1.9541158375368772, + "learning_rate": 3.992828414299524e-06, + "loss": 1.0216, + "step": 623 + }, + { + "epoch": 0.05627451864544348, + "grad_norm": 2.5488893932710455, + "learning_rate": 3.992778899648579e-06, + "loss": 1.0584, + "step": 624 + }, + { + "epoch": 0.05636470216891374, + "grad_norm": 1.8077262627055137, + "learning_rate": 3.992729214962751e-06, + "loss": 1.0805, + "step": 625 + }, + { + "epoch": 0.056454885692384, + "grad_norm": 1.852441442287416, + "learning_rate": 3.992679360246279e-06, + "loss": 1.0948, + "step": 626 + }, + { + "epoch": 0.05654506921585426, + "grad_norm": 2.0267787138600144, + "learning_rate": 3.992629335503416e-06, + "loss": 1.0902, + "step": 627 + }, + { + "epoch": 0.05663525273932452, + "grad_norm": 1.5634266203814955, + "learning_rate": 3.9925791407384304e-06, + "loss": 1.0403, + "step": 628 + }, + { + "epoch": 0.05672543626279479, + "grad_norm": 2.5568759888558725, + "learning_rate": 3.992528775955606e-06, + "loss": 1.0988, + "step": 629 + }, + { + "epoch": 0.05681561978626505, + "grad_norm": 1.6674450804947263, + "learning_rate": 3.992478241159239e-06, + "loss": 1.0294, + "step": 630 + }, + { + "epoch": 0.05690580330973531, + "grad_norm": 1.9613464050717473, + "learning_rate": 3.992427536353643e-06, + "loss": 1.084, + "step": 631 + }, + { + "epoch": 0.05699598683320557, + "grad_norm": 2.0946381236793963, + "learning_rate": 3.992376661543143e-06, + "loss": 1.0437, + "step": 632 + }, + { + "epoch": 0.05708617035667583, + "grad_norm": 2.091864576766799, + "learning_rate": 3.992325616732081e-06, + "loss": 1.0347, + "step": 633 + }, + { + "epoch": 0.057176353880146094, + "grad_norm": 1.9095585361807021, + "learning_rate": 3.992274401924811e-06, + "loss": 1.1619, + "step": 634 + }, + { + "epoch": 0.05726653740361636, + "grad_norm": 2.1622441223114044, + "learning_rate": 3.992223017125704e-06, + "loss": 0.9777, + "step": 635 + }, + { + "epoch": 0.05735672092708662, + "grad_norm": 1.8274754079267792, + "learning_rate": 3.992171462339145e-06, + "loss": 1.0858, + "step": 636 + }, + { + "epoch": 0.057446904450556883, + "grad_norm": 2.131586749636999, + "learning_rate": 3.992119737569532e-06, + "loss": 1.1255, + "step": 637 + }, + { + "epoch": 0.057537087974027144, + "grad_norm": 2.000652802741015, + "learning_rate": 3.992067842821277e-06, + "loss": 1.0541, + "step": 638 + }, + { + "epoch": 0.057627271497497405, + "grad_norm": 2.2191636680423286, + "learning_rate": 3.99201577809881e-06, + "loss": 1.19, + "step": 639 + }, + { + "epoch": 0.057717455020967666, + "grad_norm": 2.4178714270410278, + "learning_rate": 3.991963543406574e-06, + "loss": 1.0789, + "step": 640 + }, + { + "epoch": 0.057807638544437934, + "grad_norm": 1.997187365247053, + "learning_rate": 3.991911138749024e-06, + "loss": 1.0565, + "step": 641 + }, + { + "epoch": 0.057897822067908195, + "grad_norm": 2.0704385970959107, + "learning_rate": 3.991858564130633e-06, + "loss": 1.0806, + "step": 642 + }, + { + "epoch": 0.057988005591378455, + "grad_norm": 2.1527473638977135, + "learning_rate": 3.991805819555885e-06, + "loss": 1.0378, + "step": 643 + }, + { + "epoch": 0.058078189114848716, + "grad_norm": 1.6077721029976892, + "learning_rate": 3.991752905029283e-06, + "loss": 1.0613, + "step": 644 + }, + { + "epoch": 0.05816837263831898, + "grad_norm": 1.4814578654914767, + "learning_rate": 3.991699820555341e-06, + "loss": 0.9347, + "step": 645 + }, + { + "epoch": 0.05825855616178924, + "grad_norm": 2.3780493986165934, + "learning_rate": 3.991646566138588e-06, + "loss": 1.0589, + "step": 646 + }, + { + "epoch": 0.058348739685259506, + "grad_norm": 1.808622949088506, + "learning_rate": 3.991593141783567e-06, + "loss": 0.9803, + "step": 647 + }, + { + "epoch": 0.058438923208729766, + "grad_norm": 5.795979638884632, + "learning_rate": 3.991539547494839e-06, + "loss": 1.08, + "step": 648 + }, + { + "epoch": 0.05852910673220003, + "grad_norm": 6.824914382565228, + "learning_rate": 3.991485783276974e-06, + "loss": 1.0545, + "step": 649 + }, + { + "epoch": 0.05861929025567029, + "grad_norm": 2.213586090240506, + "learning_rate": 3.991431849134563e-06, + "loss": 1.1099, + "step": 650 + }, + { + "epoch": 0.05870947377914055, + "grad_norm": 2.054569487573092, + "learning_rate": 3.991377745072205e-06, + "loss": 0.9512, + "step": 651 + }, + { + "epoch": 0.05879965730261081, + "grad_norm": 2.032713846778671, + "learning_rate": 3.991323471094517e-06, + "loss": 1.0881, + "step": 652 + }, + { + "epoch": 0.05888984082608108, + "grad_norm": 2.3586451936452026, + "learning_rate": 3.991269027206131e-06, + "loss": 1.0324, + "step": 653 + }, + { + "epoch": 0.05898002434955134, + "grad_norm": 0.8087096937415341, + "learning_rate": 3.9912144134116916e-06, + "loss": 0.8769, + "step": 654 + }, + { + "epoch": 0.0590702078730216, + "grad_norm": 2.531229278102874, + "learning_rate": 3.99115962971586e-06, + "loss": 0.9801, + "step": 655 + }, + { + "epoch": 0.05916039139649186, + "grad_norm": 2.246695529099587, + "learning_rate": 3.991104676123308e-06, + "loss": 1.0921, + "step": 656 + }, + { + "epoch": 0.05925057491996212, + "grad_norm": 0.6520627038838914, + "learning_rate": 3.991049552638727e-06, + "loss": 0.8114, + "step": 657 + }, + { + "epoch": 0.05934075844343238, + "grad_norm": 1.9799155040012741, + "learning_rate": 3.99099425926682e-06, + "loss": 1.019, + "step": 658 + }, + { + "epoch": 0.05943094196690265, + "grad_norm": 1.8468152444793202, + "learning_rate": 3.990938796012304e-06, + "loss": 1.0255, + "step": 659 + }, + { + "epoch": 0.05952112549037291, + "grad_norm": 2.2394809669934412, + "learning_rate": 3.990883162879912e-06, + "loss": 1.2092, + "step": 660 + }, + { + "epoch": 0.05961130901384317, + "grad_norm": 2.1640753900577865, + "learning_rate": 3.990827359874391e-06, + "loss": 1.1256, + "step": 661 + }, + { + "epoch": 0.05970149253731343, + "grad_norm": 1.9475508837730668, + "learning_rate": 3.990771387000503e-06, + "loss": 1.0583, + "step": 662 + }, + { + "epoch": 0.05979167606078369, + "grad_norm": 1.7192767636500108, + "learning_rate": 3.990715244263023e-06, + "loss": 1.0809, + "step": 663 + }, + { + "epoch": 0.05988185958425396, + "grad_norm": 0.718697006925712, + "learning_rate": 3.990658931666741e-06, + "loss": 0.8438, + "step": 664 + }, + { + "epoch": 0.05997204310772422, + "grad_norm": 3.1713752540809903, + "learning_rate": 3.990602449216463e-06, + "loss": 0.9786, + "step": 665 + }, + { + "epoch": 0.06006222663119448, + "grad_norm": 1.8046080997059084, + "learning_rate": 3.990545796917008e-06, + "loss": 0.9591, + "step": 666 + }, + { + "epoch": 0.06015241015466474, + "grad_norm": 1.594658461808517, + "learning_rate": 3.99048897477321e-06, + "loss": 1.106, + "step": 667 + }, + { + "epoch": 0.060242593678135004, + "grad_norm": 1.9174334056684577, + "learning_rate": 3.990431982789917e-06, + "loss": 1.0766, + "step": 668 + }, + { + "epoch": 0.060332777201605264, + "grad_norm": 1.6910598082926511, + "learning_rate": 3.990374820971992e-06, + "loss": 0.9894, + "step": 669 + }, + { + "epoch": 0.06042296072507553, + "grad_norm": 4.751078031579591, + "learning_rate": 3.990317489324312e-06, + "loss": 0.9475, + "step": 670 + }, + { + "epoch": 0.06051314424854579, + "grad_norm": 1.8797870562056935, + "learning_rate": 3.99025998785177e-06, + "loss": 1.039, + "step": 671 + }, + { + "epoch": 0.060603327772016054, + "grad_norm": 1.5658335411690165, + "learning_rate": 3.990202316559271e-06, + "loss": 0.9967, + "step": 672 + }, + { + "epoch": 0.060693511295486315, + "grad_norm": 2.027930261229212, + "learning_rate": 3.990144475451738e-06, + "loss": 1.1016, + "step": 673 + }, + { + "epoch": 0.060783694818956575, + "grad_norm": 1.4697345112430402, + "learning_rate": 3.9900864645341036e-06, + "loss": 1.0538, + "step": 674 + }, + { + "epoch": 0.060873878342426836, + "grad_norm": 2.121267294062992, + "learning_rate": 3.990028283811319e-06, + "loss": 1.0605, + "step": 675 + }, + { + "epoch": 0.060964061865897104, + "grad_norm": 2.021910455737773, + "learning_rate": 3.989969933288348e-06, + "loss": 1.0509, + "step": 676 + }, + { + "epoch": 0.061054245389367365, + "grad_norm": 0.6650433645235304, + "learning_rate": 3.98991141297017e-06, + "loss": 0.7881, + "step": 677 + }, + { + "epoch": 0.061144428912837626, + "grad_norm": 1.6699668476674467, + "learning_rate": 3.989852722861778e-06, + "loss": 1.0053, + "step": 678 + }, + { + "epoch": 0.061234612436307886, + "grad_norm": 0.7197705362827919, + "learning_rate": 3.98979386296818e-06, + "loss": 0.8167, + "step": 679 + }, + { + "epoch": 0.06132479595977815, + "grad_norm": 2.1053934455384717, + "learning_rate": 3.989734833294398e-06, + "loss": 1.0357, + "step": 680 + }, + { + "epoch": 0.06141497948324841, + "grad_norm": 2.776838769086083, + "learning_rate": 3.989675633845469e-06, + "loss": 1.12, + "step": 681 + }, + { + "epoch": 0.061505163006718676, + "grad_norm": 2.327601431532526, + "learning_rate": 3.989616264626443e-06, + "loss": 1.0337, + "step": 682 + }, + { + "epoch": 0.06159534653018894, + "grad_norm": 1.753042845407514, + "learning_rate": 3.989556725642388e-06, + "loss": 1.0905, + "step": 683 + }, + { + "epoch": 0.0616855300536592, + "grad_norm": 1.9752489029698186, + "learning_rate": 3.989497016898382e-06, + "loss": 1.1094, + "step": 684 + }, + { + "epoch": 0.06177571357712946, + "grad_norm": 1.7147394726445793, + "learning_rate": 3.98943713839952e-06, + "loss": 1.0502, + "step": 685 + }, + { + "epoch": 0.06186589710059972, + "grad_norm": 0.6832837519064494, + "learning_rate": 3.9893770901509125e-06, + "loss": 0.8448, + "step": 686 + }, + { + "epoch": 0.06195608062406998, + "grad_norm": 1.5913081442734867, + "learning_rate": 3.989316872157682e-06, + "loss": 1.0434, + "step": 687 + }, + { + "epoch": 0.06204626414754025, + "grad_norm": 1.9612720097801442, + "learning_rate": 3.989256484424968e-06, + "loss": 1.1285, + "step": 688 + }, + { + "epoch": 0.06213644767101051, + "grad_norm": 2.011183702740611, + "learning_rate": 3.98919592695792e-06, + "loss": 1.023, + "step": 689 + }, + { + "epoch": 0.06222663119448077, + "grad_norm": 1.961591816589343, + "learning_rate": 3.9891351997617096e-06, + "loss": 1.0831, + "step": 690 + }, + { + "epoch": 0.06231681471795103, + "grad_norm": 1.8217389325847084, + "learning_rate": 3.989074302841514e-06, + "loss": 0.9959, + "step": 691 + }, + { + "epoch": 0.06240699824142129, + "grad_norm": 2.8085684529784776, + "learning_rate": 3.989013236202533e-06, + "loss": 1.1708, + "step": 692 + }, + { + "epoch": 0.06249718176489155, + "grad_norm": 1.6656241096651656, + "learning_rate": 3.988951999849974e-06, + "loss": 1.0931, + "step": 693 + }, + { + "epoch": 0.06258736528836181, + "grad_norm": 1.9726291767944184, + "learning_rate": 3.988890593789064e-06, + "loss": 0.9694, + "step": 694 + }, + { + "epoch": 0.06267754881183207, + "grad_norm": 1.4982402333552796, + "learning_rate": 3.9888290180250415e-06, + "loss": 1.0417, + "step": 695 + }, + { + "epoch": 0.06276773233530233, + "grad_norm": 2.138114403656876, + "learning_rate": 3.988767272563161e-06, + "loss": 1.0199, + "step": 696 + }, + { + "epoch": 0.06285791585877261, + "grad_norm": 2.4356074688921434, + "learning_rate": 3.988705357408691e-06, + "loss": 0.9949, + "step": 697 + }, + { + "epoch": 0.06294809938224287, + "grad_norm": 8.133040455198685, + "learning_rate": 3.9886432725669146e-06, + "loss": 1.0175, + "step": 698 + }, + { + "epoch": 0.06303828290571313, + "grad_norm": 1.700904098090268, + "learning_rate": 3.988581018043128e-06, + "loss": 1.0139, + "step": 699 + }, + { + "epoch": 0.06312846642918339, + "grad_norm": 2.105068416435037, + "learning_rate": 3.988518593842645e-06, + "loss": 1.0657, + "step": 700 + }, + { + "epoch": 0.06321864995265365, + "grad_norm": 2.5001938744710634, + "learning_rate": 3.9884559999707906e-06, + "loss": 1.0517, + "step": 701 + }, + { + "epoch": 0.06330883347612391, + "grad_norm": 1.8361686398715809, + "learning_rate": 3.988393236432906e-06, + "loss": 1.0687, + "step": 702 + }, + { + "epoch": 0.06339901699959417, + "grad_norm": 6.112673130196875, + "learning_rate": 3.988330303234347e-06, + "loss": 1.1318, + "step": 703 + }, + { + "epoch": 0.06348920052306443, + "grad_norm": 1.6271291868787945, + "learning_rate": 3.988267200380483e-06, + "loss": 1.0064, + "step": 704 + }, + { + "epoch": 0.0635793840465347, + "grad_norm": 1.5232586877983014, + "learning_rate": 3.988203927876698e-06, + "loss": 1.1036, + "step": 705 + }, + { + "epoch": 0.06366956757000496, + "grad_norm": 1.6447788514243407, + "learning_rate": 3.988140485728391e-06, + "loss": 0.9413, + "step": 706 + }, + { + "epoch": 0.06375975109347522, + "grad_norm": 1.6752612323028637, + "learning_rate": 3.988076873940975e-06, + "loss": 1.0061, + "step": 707 + }, + { + "epoch": 0.06384993461694548, + "grad_norm": 2.4208953845092185, + "learning_rate": 3.9880130925198786e-06, + "loss": 1.0465, + "step": 708 + }, + { + "epoch": 0.06394011814041575, + "grad_norm": 2.289717648852912, + "learning_rate": 3.987949141470543e-06, + "loss": 1.0671, + "step": 709 + }, + { + "epoch": 0.06403030166388601, + "grad_norm": 2.817433586522552, + "learning_rate": 3.987885020798425e-06, + "loss": 1.0533, + "step": 710 + }, + { + "epoch": 0.06412048518735627, + "grad_norm": 1.7350788234867482, + "learning_rate": 3.987820730508996e-06, + "loss": 1.0234, + "step": 711 + }, + { + "epoch": 0.06421066871082654, + "grad_norm": 1.8007002872297548, + "learning_rate": 3.987756270607742e-06, + "loss": 1.1507, + "step": 712 + }, + { + "epoch": 0.0643008522342968, + "grad_norm": 0.6704901462075508, + "learning_rate": 3.987691641100162e-06, + "loss": 0.8036, + "step": 713 + }, + { + "epoch": 0.06439103575776706, + "grad_norm": 2.1866166647373473, + "learning_rate": 3.987626841991771e-06, + "loss": 1.1389, + "step": 714 + }, + { + "epoch": 0.06448121928123732, + "grad_norm": 1.8239317184182495, + "learning_rate": 3.987561873288099e-06, + "loss": 1.0915, + "step": 715 + }, + { + "epoch": 0.06457140280470758, + "grad_norm": 3.297263316928836, + "learning_rate": 3.987496734994688e-06, + "loss": 1.0023, + "step": 716 + }, + { + "epoch": 0.06466158632817784, + "grad_norm": 2.117861256954982, + "learning_rate": 3.987431427117097e-06, + "loss": 1.089, + "step": 717 + }, + { + "epoch": 0.0647517698516481, + "grad_norm": 2.1949468016206177, + "learning_rate": 3.9873659496608985e-06, + "loss": 1.0684, + "step": 718 + }, + { + "epoch": 0.06484195337511836, + "grad_norm": 2.123819808845453, + "learning_rate": 3.987300302631678e-06, + "loss": 1.0046, + "step": 719 + }, + { + "epoch": 0.06493213689858862, + "grad_norm": 2.0358366606793545, + "learning_rate": 3.987234486035039e-06, + "loss": 1.0915, + "step": 720 + }, + { + "epoch": 0.0650223204220589, + "grad_norm": 2.6122681802355574, + "learning_rate": 3.987168499876595e-06, + "loss": 1.103, + "step": 721 + }, + { + "epoch": 0.06511250394552916, + "grad_norm": 1.8641592072616304, + "learning_rate": 3.987102344161978e-06, + "loss": 1.0671, + "step": 722 + }, + { + "epoch": 0.06520268746899942, + "grad_norm": 1.7241044304322557, + "learning_rate": 3.987036018896832e-06, + "loss": 1.0915, + "step": 723 + }, + { + "epoch": 0.06529287099246968, + "grad_norm": 2.070701361070568, + "learning_rate": 3.986969524086817e-06, + "loss": 1.1155, + "step": 724 + }, + { + "epoch": 0.06538305451593994, + "grad_norm": 2.519848612824821, + "learning_rate": 3.986902859737605e-06, + "loss": 1.0294, + "step": 725 + }, + { + "epoch": 0.0654732380394102, + "grad_norm": 1.6695585513705316, + "learning_rate": 3.986836025854886e-06, + "loss": 0.9866, + "step": 726 + }, + { + "epoch": 0.06556342156288046, + "grad_norm": 2.321344627707122, + "learning_rate": 3.986769022444362e-06, + "loss": 1.0045, + "step": 727 + }, + { + "epoch": 0.06565360508635072, + "grad_norm": 1.95788348557122, + "learning_rate": 3.986701849511751e-06, + "loss": 1.0479, + "step": 728 + }, + { + "epoch": 0.06574378860982098, + "grad_norm": 4.984821359385444, + "learning_rate": 3.986634507062782e-06, + "loss": 1.0372, + "step": 729 + }, + { + "epoch": 0.06583397213329124, + "grad_norm": 2.044073619360307, + "learning_rate": 3.986566995103204e-06, + "loss": 1.0696, + "step": 730 + }, + { + "epoch": 0.0659241556567615, + "grad_norm": 1.6765338479301917, + "learning_rate": 3.986499313638776e-06, + "loss": 0.9905, + "step": 731 + }, + { + "epoch": 0.06601433918023177, + "grad_norm": 2.038810743670103, + "learning_rate": 3.986431462675272e-06, + "loss": 0.9832, + "step": 732 + }, + { + "epoch": 0.06610452270370204, + "grad_norm": 6.566873600728085, + "learning_rate": 3.9863634422184835e-06, + "loss": 1.0721, + "step": 733 + }, + { + "epoch": 0.0661947062271723, + "grad_norm": 4.00043818934715, + "learning_rate": 3.986295252274213e-06, + "loss": 1.0889, + "step": 734 + }, + { + "epoch": 0.06628488975064256, + "grad_norm": 1.5911583866138228, + "learning_rate": 3.9862268928482796e-06, + "loss": 1.1002, + "step": 735 + }, + { + "epoch": 0.06637507327411282, + "grad_norm": 1.683322043742982, + "learning_rate": 3.986158363946515e-06, + "loss": 0.9679, + "step": 736 + }, + { + "epoch": 0.06646525679758308, + "grad_norm": 1.750285125393037, + "learning_rate": 3.9860896655747685e-06, + "loss": 1.0575, + "step": 737 + }, + { + "epoch": 0.06655544032105334, + "grad_norm": 1.7104136740704405, + "learning_rate": 3.9860207977388994e-06, + "loss": 1.1294, + "step": 738 + }, + { + "epoch": 0.0666456238445236, + "grad_norm": 2.103722352306906, + "learning_rate": 3.9859517604447854e-06, + "loss": 0.9842, + "step": 739 + }, + { + "epoch": 0.06673580736799387, + "grad_norm": 1.7199815672618648, + "learning_rate": 3.985882553698317e-06, + "loss": 1.0222, + "step": 740 + }, + { + "epoch": 0.06682599089146413, + "grad_norm": 2.2825076673197158, + "learning_rate": 3.985813177505399e-06, + "loss": 1.1114, + "step": 741 + }, + { + "epoch": 0.06691617441493439, + "grad_norm": 1.5655359147666317, + "learning_rate": 3.985743631871951e-06, + "loss": 1.0043, + "step": 742 + }, + { + "epoch": 0.06700635793840465, + "grad_norm": 1.7114197212579942, + "learning_rate": 3.985673916803907e-06, + "loss": 1.0517, + "step": 743 + }, + { + "epoch": 0.06709654146187491, + "grad_norm": 1.8251567773236093, + "learning_rate": 3.985604032307215e-06, + "loss": 1.0751, + "step": 744 + }, + { + "epoch": 0.06718672498534518, + "grad_norm": 2.081853404509999, + "learning_rate": 3.985533978387839e-06, + "loss": 1.0642, + "step": 745 + }, + { + "epoch": 0.06727690850881544, + "grad_norm": 2.213160499216943, + "learning_rate": 3.985463755051756e-06, + "loss": 1.0953, + "step": 746 + }, + { + "epoch": 0.0673670920322857, + "grad_norm": 2.2325592132268217, + "learning_rate": 3.9853933623049575e-06, + "loss": 1.0505, + "step": 747 + }, + { + "epoch": 0.06745727555575597, + "grad_norm": 1.6112838831521008, + "learning_rate": 3.98532280015345e-06, + "loss": 1.05, + "step": 748 + }, + { + "epoch": 0.06754745907922623, + "grad_norm": 1.9994628304081363, + "learning_rate": 3.985252068603254e-06, + "loss": 0.979, + "step": 749 + }, + { + "epoch": 0.06763764260269649, + "grad_norm": 1.5731258716361078, + "learning_rate": 3.985181167660406e-06, + "loss": 1.0852, + "step": 750 + }, + { + "epoch": 0.06772782612616675, + "grad_norm": 1.6875443629332165, + "learning_rate": 3.985110097330953e-06, + "loss": 1.0777, + "step": 751 + }, + { + "epoch": 0.06781800964963701, + "grad_norm": 2.173545140357764, + "learning_rate": 3.985038857620962e-06, + "loss": 1.045, + "step": 752 + }, + { + "epoch": 0.06790819317310727, + "grad_norm": 0.6942000973823125, + "learning_rate": 3.9849674485365094e-06, + "loss": 0.8276, + "step": 753 + }, + { + "epoch": 0.06799837669657753, + "grad_norm": 4.681086514194022, + "learning_rate": 3.98489587008369e-06, + "loss": 0.9791, + "step": 754 + }, + { + "epoch": 0.06808856022004779, + "grad_norm": 2.086534507555699, + "learning_rate": 3.98482412226861e-06, + "loss": 0.8961, + "step": 755 + }, + { + "epoch": 0.06817874374351805, + "grad_norm": 1.770582413193235, + "learning_rate": 3.984752205097391e-06, + "loss": 1.0896, + "step": 756 + }, + { + "epoch": 0.06826892726698833, + "grad_norm": 1.9314631273511007, + "learning_rate": 3.984680118576171e-06, + "loss": 1.062, + "step": 757 + }, + { + "epoch": 0.06835911079045859, + "grad_norm": 3.453078463832001, + "learning_rate": 3.984607862711099e-06, + "loss": 1.0076, + "step": 758 + }, + { + "epoch": 0.06844929431392885, + "grad_norm": 2.153600640766451, + "learning_rate": 3.984535437508341e-06, + "loss": 1.0893, + "step": 759 + }, + { + "epoch": 0.06853947783739911, + "grad_norm": 0.6764693312773526, + "learning_rate": 3.984462842974078e-06, + "loss": 0.8324, + "step": 760 + }, + { + "epoch": 0.06862966136086937, + "grad_norm": 2.368140705426052, + "learning_rate": 3.984390079114502e-06, + "loss": 1.0655, + "step": 761 + }, + { + "epoch": 0.06871984488433963, + "grad_norm": 1.5296019424031106, + "learning_rate": 3.984317145935824e-06, + "loss": 1.1701, + "step": 762 + }, + { + "epoch": 0.06881002840780989, + "grad_norm": 2.9161019641437784, + "learning_rate": 3.984244043444264e-06, + "loss": 1.0283, + "step": 763 + }, + { + "epoch": 0.06890021193128015, + "grad_norm": 1.539589264083541, + "learning_rate": 3.984170771646062e-06, + "loss": 1.0031, + "step": 764 + }, + { + "epoch": 0.06899039545475041, + "grad_norm": 2.0699226444183205, + "learning_rate": 3.9840973305474695e-06, + "loss": 1.0957, + "step": 765 + }, + { + "epoch": 0.06908057897822067, + "grad_norm": 1.6257266107384596, + "learning_rate": 3.984023720154752e-06, + "loss": 1.1061, + "step": 766 + }, + { + "epoch": 0.06917076250169094, + "grad_norm": 2.4673013420487546, + "learning_rate": 3.9839499404741915e-06, + "loss": 1.0665, + "step": 767 + }, + { + "epoch": 0.0692609460251612, + "grad_norm": 0.7268342771135498, + "learning_rate": 3.983875991512082e-06, + "loss": 0.847, + "step": 768 + }, + { + "epoch": 0.06935112954863147, + "grad_norm": 1.6960625461570351, + "learning_rate": 3.9838018732747345e-06, + "loss": 1.07, + "step": 769 + }, + { + "epoch": 0.06944131307210173, + "grad_norm": 1.5498098410584997, + "learning_rate": 3.9837275857684716e-06, + "loss": 1.0449, + "step": 770 + }, + { + "epoch": 0.06953149659557199, + "grad_norm": 2.8068835658382505, + "learning_rate": 3.983653128999634e-06, + "loss": 1.1845, + "step": 771 + }, + { + "epoch": 0.06962168011904225, + "grad_norm": 1.9023131068462105, + "learning_rate": 3.983578502974574e-06, + "loss": 1.0852, + "step": 772 + }, + { + "epoch": 0.06971186364251251, + "grad_norm": 1.5766439758963806, + "learning_rate": 3.983503707699658e-06, + "loss": 1.0441, + "step": 773 + }, + { + "epoch": 0.06980204716598278, + "grad_norm": 2.9291661319940188, + "learning_rate": 3.983428743181268e-06, + "loss": 1.0841, + "step": 774 + }, + { + "epoch": 0.06989223068945304, + "grad_norm": 2.3933644275540353, + "learning_rate": 3.983353609425802e-06, + "loss": 1.1228, + "step": 775 + }, + { + "epoch": 0.0699824142129233, + "grad_norm": 1.5884986079482768, + "learning_rate": 3.983278306439671e-06, + "loss": 1.0499, + "step": 776 + }, + { + "epoch": 0.07007259773639356, + "grad_norm": 2.138072698957599, + "learning_rate": 3.983202834229297e-06, + "loss": 1.0856, + "step": 777 + }, + { + "epoch": 0.07016278125986382, + "grad_norm": 2.17416458747777, + "learning_rate": 3.983127192801123e-06, + "loss": 1.1339, + "step": 778 + }, + { + "epoch": 0.07025296478333408, + "grad_norm": 1.889019875877748, + "learning_rate": 3.983051382161602e-06, + "loss": 1.0905, + "step": 779 + }, + { + "epoch": 0.07034314830680434, + "grad_norm": 2.6670618261173242, + "learning_rate": 3.982975402317203e-06, + "loss": 1.076, + "step": 780 + }, + { + "epoch": 0.07043333183027461, + "grad_norm": 2.750470901399133, + "learning_rate": 3.982899253274409e-06, + "loss": 1.0726, + "step": 781 + }, + { + "epoch": 0.07052351535374488, + "grad_norm": 2.012949742833957, + "learning_rate": 3.982822935039717e-06, + "loss": 1.0384, + "step": 782 + }, + { + "epoch": 0.07061369887721514, + "grad_norm": 5.177813022495391, + "learning_rate": 3.982746447619638e-06, + "loss": 0.9859, + "step": 783 + }, + { + "epoch": 0.0707038824006854, + "grad_norm": 0.7075665802512395, + "learning_rate": 3.9826697910207e-06, + "loss": 0.822, + "step": 784 + }, + { + "epoch": 0.07079406592415566, + "grad_norm": 4.685786226431425, + "learning_rate": 3.982592965249442e-06, + "loss": 1.033, + "step": 785 + }, + { + "epoch": 0.07088424944762592, + "grad_norm": 2.3059516155463813, + "learning_rate": 3.982515970312422e-06, + "loss": 1.0986, + "step": 786 + }, + { + "epoch": 0.07097443297109618, + "grad_norm": 2.749588241962002, + "learning_rate": 3.982438806216207e-06, + "loss": 1.0234, + "step": 787 + }, + { + "epoch": 0.07106461649456644, + "grad_norm": 2.2887323030697027, + "learning_rate": 3.982361472967382e-06, + "loss": 1.0388, + "step": 788 + }, + { + "epoch": 0.0711548000180367, + "grad_norm": 2.0115973871159114, + "learning_rate": 3.982283970572546e-06, + "loss": 1.1272, + "step": 789 + }, + { + "epoch": 0.07124498354150696, + "grad_norm": 2.6124591988475743, + "learning_rate": 3.982206299038311e-06, + "loss": 1.026, + "step": 790 + }, + { + "epoch": 0.07133516706497722, + "grad_norm": 5.892261550992346, + "learning_rate": 3.9821284583713054e-06, + "loss": 1.13, + "step": 791 + }, + { + "epoch": 0.07142535058844748, + "grad_norm": 2.1550035052702468, + "learning_rate": 3.98205044857817e-06, + "loss": 1.0363, + "step": 792 + }, + { + "epoch": 0.07151553411191776, + "grad_norm": 0.7406739713678869, + "learning_rate": 3.981972269665561e-06, + "loss": 0.8505, + "step": 793 + }, + { + "epoch": 0.07160571763538802, + "grad_norm": 1.4461899027293994, + "learning_rate": 3.98189392164015e-06, + "loss": 1.0413, + "step": 794 + }, + { + "epoch": 0.07169590115885828, + "grad_norm": 1.7654923245704202, + "learning_rate": 3.981815404508621e-06, + "loss": 1.094, + "step": 795 + }, + { + "epoch": 0.07178608468232854, + "grad_norm": 1.7121908396078933, + "learning_rate": 3.981736718277674e-06, + "loss": 1.056, + "step": 796 + }, + { + "epoch": 0.0718762682057988, + "grad_norm": 2.2170846626245733, + "learning_rate": 3.9816578629540235e-06, + "loss": 1.0827, + "step": 797 + }, + { + "epoch": 0.07196645172926906, + "grad_norm": 2.2869429947220805, + "learning_rate": 3.981578838544398e-06, + "loss": 1.065, + "step": 798 + }, + { + "epoch": 0.07205663525273932, + "grad_norm": 0.6718861556790414, + "learning_rate": 3.981499645055539e-06, + "loss": 0.8046, + "step": 799 + }, + { + "epoch": 0.07214681877620958, + "grad_norm": 0.6576343875150449, + "learning_rate": 3.981420282494204e-06, + "loss": 0.8386, + "step": 800 + }, + { + "epoch": 0.07223700229967984, + "grad_norm": 0.7545098769019869, + "learning_rate": 3.981340750867166e-06, + "loss": 0.837, + "step": 801 + }, + { + "epoch": 0.0723271858231501, + "grad_norm": 1.559447167203947, + "learning_rate": 3.981261050181209e-06, + "loss": 1.037, + "step": 802 + }, + { + "epoch": 0.07241736934662037, + "grad_norm": 1.8540940502677201, + "learning_rate": 3.9811811804431355e-06, + "loss": 1.0009, + "step": 803 + }, + { + "epoch": 0.07250755287009064, + "grad_norm": 2.4148595784118103, + "learning_rate": 3.981101141659759e-06, + "loss": 1.0941, + "step": 804 + }, + { + "epoch": 0.0725977363935609, + "grad_norm": 1.7821518472138422, + "learning_rate": 3.98102093383791e-06, + "loss": 1.0768, + "step": 805 + }, + { + "epoch": 0.07268791991703116, + "grad_norm": 3.1945290945880322, + "learning_rate": 3.9809405569844315e-06, + "loss": 0.9708, + "step": 806 + }, + { + "epoch": 0.07277810344050142, + "grad_norm": 1.7439076779628595, + "learning_rate": 3.980860011106182e-06, + "loss": 1.0127, + "step": 807 + }, + { + "epoch": 0.07286828696397168, + "grad_norm": 1.8235924113156514, + "learning_rate": 3.980779296210033e-06, + "loss": 0.9881, + "step": 808 + }, + { + "epoch": 0.07295847048744195, + "grad_norm": 1.808037624997268, + "learning_rate": 3.980698412302874e-06, + "loss": 1.1233, + "step": 809 + }, + { + "epoch": 0.0730486540109122, + "grad_norm": 1.8753014640057475, + "learning_rate": 3.980617359391604e-06, + "loss": 1.0665, + "step": 810 + }, + { + "epoch": 0.07313883753438247, + "grad_norm": 2.6275977949677043, + "learning_rate": 3.98053613748314e-06, + "loss": 1.0067, + "step": 811 + }, + { + "epoch": 0.07322902105785273, + "grad_norm": 1.6397824258228515, + "learning_rate": 3.980454746584413e-06, + "loss": 0.9156, + "step": 812 + }, + { + "epoch": 0.07331920458132299, + "grad_norm": 2.257009290758041, + "learning_rate": 3.9803731867023665e-06, + "loss": 1.133, + "step": 813 + }, + { + "epoch": 0.07340938810479325, + "grad_norm": 2.045055366829871, + "learning_rate": 3.9802914578439596e-06, + "loss": 1.0359, + "step": 814 + }, + { + "epoch": 0.07349957162826351, + "grad_norm": 1.8084899345246512, + "learning_rate": 3.980209560016167e-06, + "loss": 1.1522, + "step": 815 + }, + { + "epoch": 0.07358975515173379, + "grad_norm": 1.7441619495826843, + "learning_rate": 3.980127493225975e-06, + "loss": 1.1439, + "step": 816 + }, + { + "epoch": 0.07367993867520405, + "grad_norm": 2.01346324354165, + "learning_rate": 3.980045257480387e-06, + "loss": 1.0781, + "step": 817 + }, + { + "epoch": 0.0737701221986743, + "grad_norm": 2.032468576974301, + "learning_rate": 3.9799628527864205e-06, + "loss": 1.0037, + "step": 818 + }, + { + "epoch": 0.07386030572214457, + "grad_norm": 1.9318073073504365, + "learning_rate": 3.979880279151106e-06, + "loss": 1.0338, + "step": 819 + }, + { + "epoch": 0.07395048924561483, + "grad_norm": 2.0807580544030424, + "learning_rate": 3.979797536581489e-06, + "loss": 0.9912, + "step": 820 + }, + { + "epoch": 0.07404067276908509, + "grad_norm": 2.3298058298762876, + "learning_rate": 3.97971462508463e-06, + "loss": 0.8917, + "step": 821 + }, + { + "epoch": 0.07413085629255535, + "grad_norm": 1.7702058765055673, + "learning_rate": 3.979631544667603e-06, + "loss": 1.0356, + "step": 822 + }, + { + "epoch": 0.07422103981602561, + "grad_norm": 1.8415987428017906, + "learning_rate": 3.979548295337496e-06, + "loss": 1.0602, + "step": 823 + }, + { + "epoch": 0.07431122333949587, + "grad_norm": 1.790558701597685, + "learning_rate": 3.9794648771014146e-06, + "loss": 1.0149, + "step": 824 + }, + { + "epoch": 0.07440140686296613, + "grad_norm": 1.9083699179839235, + "learning_rate": 3.9793812899664745e-06, + "loss": 1.0532, + "step": 825 + }, + { + "epoch": 0.0744915903864364, + "grad_norm": 1.983140696725841, + "learning_rate": 3.979297533939809e-06, + "loss": 1.007, + "step": 826 + }, + { + "epoch": 0.07458177390990665, + "grad_norm": 2.167478543783641, + "learning_rate": 3.979213609028564e-06, + "loss": 1.0267, + "step": 827 + }, + { + "epoch": 0.07467195743337693, + "grad_norm": 1.9415469205001934, + "learning_rate": 3.979129515239901e-06, + "loss": 1.0609, + "step": 828 + }, + { + "epoch": 0.07476214095684719, + "grad_norm": 1.666654332433201, + "learning_rate": 3.979045252580994e-06, + "loss": 1.0584, + "step": 829 + }, + { + "epoch": 0.07485232448031745, + "grad_norm": 1.7036803853624511, + "learning_rate": 3.978960821059034e-06, + "loss": 0.995, + "step": 830 + }, + { + "epoch": 0.07494250800378771, + "grad_norm": 2.540788926282528, + "learning_rate": 3.978876220681225e-06, + "loss": 1.0894, + "step": 831 + }, + { + "epoch": 0.07503269152725797, + "grad_norm": 2.3114425715945845, + "learning_rate": 3.978791451454786e-06, + "loss": 0.9765, + "step": 832 + }, + { + "epoch": 0.07512287505072823, + "grad_norm": 1.6801323589909927, + "learning_rate": 3.978706513386949e-06, + "loss": 1.0044, + "step": 833 + }, + { + "epoch": 0.0752130585741985, + "grad_norm": 4.512731555693292, + "learning_rate": 3.978621406484962e-06, + "loss": 1.0585, + "step": 834 + }, + { + "epoch": 0.07530324209766875, + "grad_norm": 2.8819577156792877, + "learning_rate": 3.978536130756086e-06, + "loss": 1.0426, + "step": 835 + }, + { + "epoch": 0.07539342562113902, + "grad_norm": 1.7218707795965462, + "learning_rate": 3.978450686207599e-06, + "loss": 1.1142, + "step": 836 + }, + { + "epoch": 0.07548360914460928, + "grad_norm": 2.5533323280823854, + "learning_rate": 3.978365072846789e-06, + "loss": 1.0314, + "step": 837 + }, + { + "epoch": 0.07557379266807954, + "grad_norm": 2.1446283037080516, + "learning_rate": 3.9782792906809625e-06, + "loss": 0.9983, + "step": 838 + }, + { + "epoch": 0.0756639761915498, + "grad_norm": 2.2613675861141167, + "learning_rate": 3.97819333971744e-06, + "loss": 1.0355, + "step": 839 + }, + { + "epoch": 0.07575415971502007, + "grad_norm": 2.614104045895755, + "learning_rate": 3.978107219963553e-06, + "loss": 1.0632, + "step": 840 + }, + { + "epoch": 0.07584434323849033, + "grad_norm": 2.449915639242359, + "learning_rate": 3.978020931426651e-06, + "loss": 0.9546, + "step": 841 + }, + { + "epoch": 0.0759345267619606, + "grad_norm": 1.5561602340608327, + "learning_rate": 3.977934474114096e-06, + "loss": 0.992, + "step": 842 + }, + { + "epoch": 0.07602471028543085, + "grad_norm": 0.642287794574355, + "learning_rate": 3.977847848033267e-06, + "loss": 0.8158, + "step": 843 + }, + { + "epoch": 0.07611489380890112, + "grad_norm": 1.9437440731044084, + "learning_rate": 3.977761053191553e-06, + "loss": 1.0891, + "step": 844 + }, + { + "epoch": 0.07620507733237138, + "grad_norm": 6.973918828265941, + "learning_rate": 3.977674089596361e-06, + "loss": 1.0298, + "step": 845 + }, + { + "epoch": 0.07629526085584164, + "grad_norm": 0.6879316622067977, + "learning_rate": 3.97758695725511e-06, + "loss": 0.813, + "step": 846 + }, + { + "epoch": 0.0763854443793119, + "grad_norm": 1.5161180775638705, + "learning_rate": 3.977499656175236e-06, + "loss": 1.0435, + "step": 847 + }, + { + "epoch": 0.07647562790278216, + "grad_norm": 2.7749345204644, + "learning_rate": 3.977412186364187e-06, + "loss": 0.9581, + "step": 848 + }, + { + "epoch": 0.07656581142625242, + "grad_norm": 2.0924851881662896, + "learning_rate": 3.977324547829428e-06, + "loss": 1.0736, + "step": 849 + }, + { + "epoch": 0.07665599494972268, + "grad_norm": 2.245868704758511, + "learning_rate": 3.977236740578435e-06, + "loss": 1.0682, + "step": 850 + }, + { + "epoch": 0.07674617847319294, + "grad_norm": 2.001601174283922, + "learning_rate": 3.9771487646187015e-06, + "loss": 1.033, + "step": 851 + }, + { + "epoch": 0.07683636199666322, + "grad_norm": 1.79456021024381, + "learning_rate": 3.9770606199577325e-06, + "loss": 1.0841, + "step": 852 + }, + { + "epoch": 0.07692654552013348, + "grad_norm": 2.454001397708411, + "learning_rate": 3.9769723066030505e-06, + "loss": 0.9814, + "step": 853 + }, + { + "epoch": 0.07701672904360374, + "grad_norm": 2.150411136262589, + "learning_rate": 3.976883824562191e-06, + "loss": 1.1034, + "step": 854 + }, + { + "epoch": 0.077106912567074, + "grad_norm": 3.052381811204901, + "learning_rate": 3.976795173842703e-06, + "loss": 1.0445, + "step": 855 + }, + { + "epoch": 0.07719709609054426, + "grad_norm": 2.111102604709464, + "learning_rate": 3.97670635445215e-06, + "loss": 1.024, + "step": 856 + }, + { + "epoch": 0.07728727961401452, + "grad_norm": 2.33626531131599, + "learning_rate": 3.976617366398112e-06, + "loss": 1.0188, + "step": 857 + }, + { + "epoch": 0.07737746313748478, + "grad_norm": 1.521835423763718, + "learning_rate": 3.976528209688181e-06, + "loss": 1.0775, + "step": 858 + }, + { + "epoch": 0.07746764666095504, + "grad_norm": 1.643126696632528, + "learning_rate": 3.976438884329965e-06, + "loss": 1.0511, + "step": 859 + }, + { + "epoch": 0.0775578301844253, + "grad_norm": 1.827209569498614, + "learning_rate": 3.976349390331085e-06, + "loss": 1.1793, + "step": 860 + }, + { + "epoch": 0.07764801370789556, + "grad_norm": 1.6596887985049156, + "learning_rate": 3.976259727699178e-06, + "loss": 1.0152, + "step": 861 + }, + { + "epoch": 0.07773819723136582, + "grad_norm": 2.4236189931056975, + "learning_rate": 3.976169896441895e-06, + "loss": 1.0043, + "step": 862 + }, + { + "epoch": 0.07782838075483609, + "grad_norm": 1.7841097980376257, + "learning_rate": 3.976079896566898e-06, + "loss": 1.0806, + "step": 863 + }, + { + "epoch": 0.07791856427830636, + "grad_norm": 1.7852664971897498, + "learning_rate": 3.97598972808187e-06, + "loss": 1.1096, + "step": 864 + }, + { + "epoch": 0.07800874780177662, + "grad_norm": 1.8398597538407921, + "learning_rate": 3.975899390994501e-06, + "loss": 1.0478, + "step": 865 + }, + { + "epoch": 0.07809893132524688, + "grad_norm": 1.497770162040134, + "learning_rate": 3.975808885312502e-06, + "loss": 1.1087, + "step": 866 + }, + { + "epoch": 0.07818911484871714, + "grad_norm": 1.9416376048957018, + "learning_rate": 3.975718211043594e-06, + "loss": 1.0892, + "step": 867 + }, + { + "epoch": 0.0782792983721874, + "grad_norm": 1.7319751666757717, + "learning_rate": 3.975627368195515e-06, + "loss": 0.9583, + "step": 868 + }, + { + "epoch": 0.07836948189565766, + "grad_norm": 1.4289546401484143, + "learning_rate": 3.975536356776015e-06, + "loss": 1.0128, + "step": 869 + }, + { + "epoch": 0.07845966541912792, + "grad_norm": 0.7395728383640449, + "learning_rate": 3.975445176792861e-06, + "loss": 0.8905, + "step": 870 + }, + { + "epoch": 0.07854984894259819, + "grad_norm": 1.852929684926577, + "learning_rate": 3.975353828253831e-06, + "loss": 1.0485, + "step": 871 + }, + { + "epoch": 0.07864003246606845, + "grad_norm": 1.8752423447713458, + "learning_rate": 3.97526231116672e-06, + "loss": 1.0505, + "step": 872 + }, + { + "epoch": 0.07873021598953871, + "grad_norm": 3.597509126419675, + "learning_rate": 3.975170625539338e-06, + "loss": 0.9877, + "step": 873 + }, + { + "epoch": 0.07882039951300897, + "grad_norm": 2.1075615894011013, + "learning_rate": 3.975078771379507e-06, + "loss": 1.0484, + "step": 874 + }, + { + "epoch": 0.07891058303647923, + "grad_norm": 1.9956833031103598, + "learning_rate": 3.974986748695064e-06, + "loss": 0.9569, + "step": 875 + }, + { + "epoch": 0.0790007665599495, + "grad_norm": 0.8661858900162438, + "learning_rate": 3.974894557493862e-06, + "loss": 0.8278, + "step": 876 + }, + { + "epoch": 0.07909095008341976, + "grad_norm": 1.7423478125469338, + "learning_rate": 3.974802197783768e-06, + "loss": 1.0924, + "step": 877 + }, + { + "epoch": 0.07918113360689003, + "grad_norm": 1.9474345203796666, + "learning_rate": 3.974709669572661e-06, + "loss": 1.1018, + "step": 878 + }, + { + "epoch": 0.07927131713036029, + "grad_norm": 2.0295687246478864, + "learning_rate": 3.974616972868436e-06, + "loss": 1.0395, + "step": 879 + }, + { + "epoch": 0.07936150065383055, + "grad_norm": 1.6428258756659417, + "learning_rate": 3.974524107679003e-06, + "loss": 1.054, + "step": 880 + }, + { + "epoch": 0.07945168417730081, + "grad_norm": 1.4421719633941446, + "learning_rate": 3.974431074012286e-06, + "loss": 1.039, + "step": 881 + }, + { + "epoch": 0.07954186770077107, + "grad_norm": 2.0553841805830997, + "learning_rate": 3.974337871876223e-06, + "loss": 1.0731, + "step": 882 + }, + { + "epoch": 0.07963205122424133, + "grad_norm": 1.6913488083175199, + "learning_rate": 3.974244501278766e-06, + "loss": 1.0146, + "step": 883 + }, + { + "epoch": 0.07972223474771159, + "grad_norm": 2.4319624967381324, + "learning_rate": 3.974150962227883e-06, + "loss": 0.995, + "step": 884 + }, + { + "epoch": 0.07981241827118185, + "grad_norm": 2.1123733121365316, + "learning_rate": 3.9740572547315535e-06, + "loss": 0.9745, + "step": 885 + }, + { + "epoch": 0.07990260179465211, + "grad_norm": 2.95186301418555, + "learning_rate": 3.973963378797775e-06, + "loss": 1.0863, + "step": 886 + }, + { + "epoch": 0.07999278531812237, + "grad_norm": 1.7129543731901757, + "learning_rate": 3.973869334434556e-06, + "loss": 1.0622, + "step": 887 + }, + { + "epoch": 0.08008296884159265, + "grad_norm": 2.1584820459258465, + "learning_rate": 3.973775121649922e-06, + "loss": 1.0616, + "step": 888 + }, + { + "epoch": 0.08017315236506291, + "grad_norm": 2.062908132226352, + "learning_rate": 3.973680740451911e-06, + "loss": 0.9632, + "step": 889 + }, + { + "epoch": 0.08026333588853317, + "grad_norm": 3.237951687561358, + "learning_rate": 3.9735861908485776e-06, + "loss": 1.0126, + "step": 890 + }, + { + "epoch": 0.08035351941200343, + "grad_norm": 1.6459367051637785, + "learning_rate": 3.973491472847987e-06, + "loss": 1.065, + "step": 891 + }, + { + "epoch": 0.08044370293547369, + "grad_norm": 3.9924972025862075, + "learning_rate": 3.973396586458222e-06, + "loss": 0.976, + "step": 892 + }, + { + "epoch": 0.08053388645894395, + "grad_norm": 4.755495004971572, + "learning_rate": 3.97330153168738e-06, + "loss": 1.031, + "step": 893 + }, + { + "epoch": 0.08062406998241421, + "grad_norm": 1.7400241986049956, + "learning_rate": 3.973206308543571e-06, + "loss": 0.9904, + "step": 894 + }, + { + "epoch": 0.08071425350588447, + "grad_norm": 2.0052649816132466, + "learning_rate": 3.973110917034918e-06, + "loss": 1.0159, + "step": 895 + }, + { + "epoch": 0.08080443702935473, + "grad_norm": 3.209333107144652, + "learning_rate": 3.973015357169563e-06, + "loss": 0.956, + "step": 896 + }, + { + "epoch": 0.080894620552825, + "grad_norm": 2.1066717840486655, + "learning_rate": 3.972919628955659e-06, + "loss": 0.9363, + "step": 897 + }, + { + "epoch": 0.08098480407629526, + "grad_norm": 2.3558637752754357, + "learning_rate": 3.972823732401373e-06, + "loss": 1.0278, + "step": 898 + }, + { + "epoch": 0.08107498759976552, + "grad_norm": 2.3176164473132066, + "learning_rate": 3.972727667514888e-06, + "loss": 1.0361, + "step": 899 + }, + { + "epoch": 0.08116517112323579, + "grad_norm": 1.6820732029964236, + "learning_rate": 3.972631434304402e-06, + "loss": 1.0068, + "step": 900 + }, + { + "epoch": 0.08125535464670605, + "grad_norm": 2.265255917708015, + "learning_rate": 3.972535032778124e-06, + "loss": 1.1578, + "step": 901 + }, + { + "epoch": 0.08134553817017631, + "grad_norm": 1.91345641213266, + "learning_rate": 3.97243846294428e-06, + "loss": 1.1296, + "step": 902 + }, + { + "epoch": 0.08143572169364657, + "grad_norm": 1.7683392482494455, + "learning_rate": 3.972341724811111e-06, + "loss": 1.1772, + "step": 903 + }, + { + "epoch": 0.08152590521711683, + "grad_norm": 2.3343040853859027, + "learning_rate": 3.972244818386872e-06, + "loss": 0.9861, + "step": 904 + }, + { + "epoch": 0.0816160887405871, + "grad_norm": 1.9278007890107824, + "learning_rate": 3.972147743679828e-06, + "loss": 0.9851, + "step": 905 + }, + { + "epoch": 0.08170627226405736, + "grad_norm": 1.714737873675767, + "learning_rate": 3.972050500698265e-06, + "loss": 1.0095, + "step": 906 + }, + { + "epoch": 0.08179645578752762, + "grad_norm": 3.618405625159854, + "learning_rate": 3.971953089450481e-06, + "loss": 1.035, + "step": 907 + }, + { + "epoch": 0.08188663931099788, + "grad_norm": 2.271559001454425, + "learning_rate": 3.971855509944784e-06, + "loss": 1.1129, + "step": 908 + }, + { + "epoch": 0.08197682283446814, + "grad_norm": 4.142032254596784, + "learning_rate": 3.971757762189504e-06, + "loss": 1.0278, + "step": 909 + }, + { + "epoch": 0.0820670063579384, + "grad_norm": 2.77981281652045, + "learning_rate": 3.9716598461929785e-06, + "loss": 1.0327, + "step": 910 + }, + { + "epoch": 0.08215718988140866, + "grad_norm": 2.274868963471161, + "learning_rate": 3.971561761963563e-06, + "loss": 1.0122, + "step": 911 + }, + { + "epoch": 0.08224737340487893, + "grad_norm": 2.1056389394849586, + "learning_rate": 3.971463509509628e-06, + "loss": 1.1143, + "step": 912 + }, + { + "epoch": 0.0823375569283492, + "grad_norm": 19.095178828494575, + "learning_rate": 3.9713650888395555e-06, + "loss": 1.0351, + "step": 913 + }, + { + "epoch": 0.08242774045181946, + "grad_norm": 2.0416260021078565, + "learning_rate": 3.9712664999617425e-06, + "loss": 1.0288, + "step": 914 + }, + { + "epoch": 0.08251792397528972, + "grad_norm": 1.965778833625363, + "learning_rate": 3.971167742884603e-06, + "loss": 1.0172, + "step": 915 + }, + { + "epoch": 0.08260810749875998, + "grad_norm": 1.8895230194974968, + "learning_rate": 3.971068817616564e-06, + "loss": 1.0885, + "step": 916 + }, + { + "epoch": 0.08269829102223024, + "grad_norm": 1.9416649260397163, + "learning_rate": 3.970969724166064e-06, + "loss": 1.0631, + "step": 917 + }, + { + "epoch": 0.0827884745457005, + "grad_norm": 0.8260872799825799, + "learning_rate": 3.970870462541559e-06, + "loss": 0.8154, + "step": 918 + }, + { + "epoch": 0.08287865806917076, + "grad_norm": 1.4968675649056116, + "learning_rate": 3.97077103275152e-06, + "loss": 1.0608, + "step": 919 + }, + { + "epoch": 0.08296884159264102, + "grad_norm": 3.107398703804637, + "learning_rate": 3.970671434804428e-06, + "loss": 1.0248, + "step": 920 + }, + { + "epoch": 0.08305902511611128, + "grad_norm": 2.223074660783621, + "learning_rate": 3.970571668708784e-06, + "loss": 1.0372, + "step": 921 + }, + { + "epoch": 0.08314920863958154, + "grad_norm": 0.7309082667330163, + "learning_rate": 3.9704717344731e-06, + "loss": 0.8265, + "step": 922 + }, + { + "epoch": 0.0832393921630518, + "grad_norm": 0.7497244964668135, + "learning_rate": 3.9703716321059026e-06, + "loss": 0.8424, + "step": 923 + }, + { + "epoch": 0.08332957568652208, + "grad_norm": 2.886091151172037, + "learning_rate": 3.9702713616157325e-06, + "loss": 1.1258, + "step": 924 + }, + { + "epoch": 0.08341975920999234, + "grad_norm": 2.2660949219689663, + "learning_rate": 3.9701709230111455e-06, + "loss": 1.0529, + "step": 925 + }, + { + "epoch": 0.0835099427334626, + "grad_norm": 2.2862542017110234, + "learning_rate": 3.970070316300713e-06, + "loss": 1.0046, + "step": 926 + }, + { + "epoch": 0.08360012625693286, + "grad_norm": 1.8550963339250308, + "learning_rate": 3.969969541493017e-06, + "loss": 1.0606, + "step": 927 + }, + { + "epoch": 0.08369030978040312, + "grad_norm": 11.814122971550367, + "learning_rate": 3.969868598596658e-06, + "loss": 1.0927, + "step": 928 + }, + { + "epoch": 0.08378049330387338, + "grad_norm": 1.546670765350655, + "learning_rate": 3.969767487620249e-06, + "loss": 0.9814, + "step": 929 + }, + { + "epoch": 0.08387067682734364, + "grad_norm": 1.568202104675998, + "learning_rate": 3.969666208572416e-06, + "loss": 1.0898, + "step": 930 + }, + { + "epoch": 0.0839608603508139, + "grad_norm": 2.8501083219589067, + "learning_rate": 3.969564761461802e-06, + "loss": 1.1038, + "step": 931 + }, + { + "epoch": 0.08405104387428416, + "grad_norm": 2.181428931474723, + "learning_rate": 3.969463146297062e-06, + "loss": 1.0193, + "step": 932 + }, + { + "epoch": 0.08414122739775443, + "grad_norm": 1.8995460444786814, + "learning_rate": 3.969361363086867e-06, + "loss": 1.0665, + "step": 933 + }, + { + "epoch": 0.08423141092122469, + "grad_norm": 2.7186842285619246, + "learning_rate": 3.9692594118399014e-06, + "loss": 1.0052, + "step": 934 + }, + { + "epoch": 0.08432159444469496, + "grad_norm": 1.9772097767900922, + "learning_rate": 3.969157292564865e-06, + "loss": 1.0741, + "step": 935 + }, + { + "epoch": 0.08441177796816522, + "grad_norm": 7.378724531659339, + "learning_rate": 3.96905500527047e-06, + "loss": 1.0267, + "step": 936 + }, + { + "epoch": 0.08450196149163548, + "grad_norm": 2.611588504338157, + "learning_rate": 3.968952549965445e-06, + "loss": 1.0277, + "step": 937 + }, + { + "epoch": 0.08459214501510574, + "grad_norm": 2.3234515294239557, + "learning_rate": 3.968849926658532e-06, + "loss": 1.0775, + "step": 938 + }, + { + "epoch": 0.084682328538576, + "grad_norm": 5.413825928554813, + "learning_rate": 3.9687471353584866e-06, + "loss": 0.9541, + "step": 939 + }, + { + "epoch": 0.08477251206204627, + "grad_norm": 1.8830796681394175, + "learning_rate": 3.9686441760740795e-06, + "loss": 1.0581, + "step": 940 + }, + { + "epoch": 0.08486269558551653, + "grad_norm": 2.5816440493941486, + "learning_rate": 3.968541048814098e-06, + "loss": 0.9855, + "step": 941 + }, + { + "epoch": 0.08495287910898679, + "grad_norm": 5.445233337615039, + "learning_rate": 3.968437753587339e-06, + "loss": 0.9733, + "step": 942 + }, + { + "epoch": 0.08504306263245705, + "grad_norm": 1.8122999804394835, + "learning_rate": 3.968334290402616e-06, + "loss": 1.0883, + "step": 943 + }, + { + "epoch": 0.08513324615592731, + "grad_norm": 1.7839338595741387, + "learning_rate": 3.968230659268759e-06, + "loss": 0.9377, + "step": 944 + }, + { + "epoch": 0.08522342967939757, + "grad_norm": 2.627115033346941, + "learning_rate": 3.968126860194609e-06, + "loss": 1.1137, + "step": 945 + }, + { + "epoch": 0.08531361320286783, + "grad_norm": 2.3103325454996035, + "learning_rate": 3.968022893189025e-06, + "loss": 1.0948, + "step": 946 + }, + { + "epoch": 0.0854037967263381, + "grad_norm": 1.8862888644128541, + "learning_rate": 3.967918758260874e-06, + "loss": 1.0811, + "step": 947 + }, + { + "epoch": 0.08549398024980837, + "grad_norm": 2.36194778528084, + "learning_rate": 3.967814455419044e-06, + "loss": 1.0921, + "step": 948 + }, + { + "epoch": 0.08558416377327863, + "grad_norm": 2.8620606197652307, + "learning_rate": 3.967709984672434e-06, + "loss": 1.0854, + "step": 949 + }, + { + "epoch": 0.08567434729674889, + "grad_norm": 3.2134526187341117, + "learning_rate": 3.967605346029959e-06, + "loss": 1.076, + "step": 950 + }, + { + "epoch": 0.08576453082021915, + "grad_norm": 1.453908647456337, + "learning_rate": 3.9675005395005466e-06, + "loss": 1.1207, + "step": 951 + }, + { + "epoch": 0.08585471434368941, + "grad_norm": 1.8860811218339544, + "learning_rate": 3.967395565093139e-06, + "loss": 1.0648, + "step": 952 + }, + { + "epoch": 0.08594489786715967, + "grad_norm": 2.6152100337564477, + "learning_rate": 3.967290422816693e-06, + "loss": 1.0515, + "step": 953 + }, + { + "epoch": 0.08603508139062993, + "grad_norm": 2.735004897453205, + "learning_rate": 3.967185112680183e-06, + "loss": 1.0137, + "step": 954 + }, + { + "epoch": 0.08612526491410019, + "grad_norm": 2.8877139718178855, + "learning_rate": 3.96707963469259e-06, + "loss": 1.1264, + "step": 955 + }, + { + "epoch": 0.08621544843757045, + "grad_norm": 1.9788598743039203, + "learning_rate": 3.966973988862917e-06, + "loss": 1.0352, + "step": 956 + }, + { + "epoch": 0.08630563196104071, + "grad_norm": 3.1405431632391965, + "learning_rate": 3.966868175200178e-06, + "loss": 1.0993, + "step": 957 + }, + { + "epoch": 0.08639581548451097, + "grad_norm": 2.167532442195496, + "learning_rate": 3.9667621937134e-06, + "loss": 1.1078, + "step": 958 + }, + { + "epoch": 0.08648599900798125, + "grad_norm": 2.3833391451820347, + "learning_rate": 3.966656044411627e-06, + "loss": 0.9549, + "step": 959 + }, + { + "epoch": 0.08657618253145151, + "grad_norm": 2.8470511925251514, + "learning_rate": 3.966549727303918e-06, + "loss": 1.02, + "step": 960 + }, + { + "epoch": 0.08666636605492177, + "grad_norm": 7.167497217119926, + "learning_rate": 3.966443242399341e-06, + "loss": 1.0808, + "step": 961 + }, + { + "epoch": 0.08675654957839203, + "grad_norm": 0.7157076425600298, + "learning_rate": 3.966336589706985e-06, + "loss": 0.8368, + "step": 962 + }, + { + "epoch": 0.08684673310186229, + "grad_norm": 1.7004401366485076, + "learning_rate": 3.966229769235948e-06, + "loss": 1.0595, + "step": 963 + }, + { + "epoch": 0.08693691662533255, + "grad_norm": 2.6037197390742297, + "learning_rate": 3.966122780995345e-06, + "loss": 1.1281, + "step": 964 + }, + { + "epoch": 0.08702710014880281, + "grad_norm": 2.730273762345508, + "learning_rate": 3.966015624994306e-06, + "loss": 1.0784, + "step": 965 + }, + { + "epoch": 0.08711728367227307, + "grad_norm": 1.9581049792754426, + "learning_rate": 3.9659083012419735e-06, + "loss": 1.0218, + "step": 966 + }, + { + "epoch": 0.08720746719574334, + "grad_norm": 2.0166602026885716, + "learning_rate": 3.965800809747505e-06, + "loss": 0.9799, + "step": 967 + }, + { + "epoch": 0.0872976507192136, + "grad_norm": 2.728461972953628, + "learning_rate": 3.965693150520071e-06, + "loss": 1.0147, + "step": 968 + }, + { + "epoch": 0.08738783424268386, + "grad_norm": 2.391978759428913, + "learning_rate": 3.96558532356886e-06, + "loss": 0.9834, + "step": 969 + }, + { + "epoch": 0.08747801776615412, + "grad_norm": 1.8635970193977045, + "learning_rate": 3.9654773289030704e-06, + "loss": 1.0885, + "step": 970 + }, + { + "epoch": 0.08756820128962439, + "grad_norm": 7.49928432864727, + "learning_rate": 3.9653691665319176e-06, + "loss": 1.0634, + "step": 971 + }, + { + "epoch": 0.08765838481309465, + "grad_norm": 4.273962570933356, + "learning_rate": 3.96526083646463e-06, + "loss": 0.9745, + "step": 972 + }, + { + "epoch": 0.08774856833656491, + "grad_norm": 3.48856310757624, + "learning_rate": 3.9651523387104526e-06, + "loss": 0.9966, + "step": 973 + }, + { + "epoch": 0.08783875186003517, + "grad_norm": 2.613977906750315, + "learning_rate": 3.965043673278641e-06, + "loss": 1.0586, + "step": 974 + }, + { + "epoch": 0.08792893538350544, + "grad_norm": 1.6281022756678372, + "learning_rate": 3.964934840178469e-06, + "loss": 1.0555, + "step": 975 + }, + { + "epoch": 0.0880191189069757, + "grad_norm": 1.8604232013355528, + "learning_rate": 3.964825839419221e-06, + "loss": 1.0983, + "step": 976 + }, + { + "epoch": 0.08810930243044596, + "grad_norm": 4.712885768491074, + "learning_rate": 3.964716671010199e-06, + "loss": 1.0337, + "step": 977 + }, + { + "epoch": 0.08819948595391622, + "grad_norm": 2.081120447474036, + "learning_rate": 3.9646073349607165e-06, + "loss": 1.0119, + "step": 978 + }, + { + "epoch": 0.08828966947738648, + "grad_norm": 3.6044122836409027, + "learning_rate": 3.964497831280105e-06, + "loss": 1.1805, + "step": 979 + }, + { + "epoch": 0.08837985300085674, + "grad_norm": 2.0751101774816063, + "learning_rate": 3.964388159977705e-06, + "loss": 1.0501, + "step": 980 + }, + { + "epoch": 0.088470036524327, + "grad_norm": 1.6941466651757813, + "learning_rate": 3.964278321062876e-06, + "loss": 0.9554, + "step": 981 + }, + { + "epoch": 0.08856022004779726, + "grad_norm": 3.1137607640739247, + "learning_rate": 3.96416831454499e-06, + "loss": 1.0688, + "step": 982 + }, + { + "epoch": 0.08865040357126754, + "grad_norm": 0.7036548313682756, + "learning_rate": 3.964058140433434e-06, + "loss": 0.8674, + "step": 983 + }, + { + "epoch": 0.0887405870947378, + "grad_norm": 3.200691154465231, + "learning_rate": 3.963947798737606e-06, + "loss": 1.1112, + "step": 984 + }, + { + "epoch": 0.08883077061820806, + "grad_norm": 2.0581702822868353, + "learning_rate": 3.963837289466923e-06, + "loss": 1.0699, + "step": 985 + }, + { + "epoch": 0.08892095414167832, + "grad_norm": 7.777948008294126, + "learning_rate": 3.9637266126308145e-06, + "loss": 1.0193, + "step": 986 + }, + { + "epoch": 0.08901113766514858, + "grad_norm": 0.6896130427695916, + "learning_rate": 3.963615768238724e-06, + "loss": 0.8568, + "step": 987 + }, + { + "epoch": 0.08910132118861884, + "grad_norm": 1.6157495122333625, + "learning_rate": 3.963504756300107e-06, + "loss": 0.9352, + "step": 988 + }, + { + "epoch": 0.0891915047120891, + "grad_norm": 3.6261327880935643, + "learning_rate": 3.96339357682444e-06, + "loss": 1.0862, + "step": 989 + }, + { + "epoch": 0.08928168823555936, + "grad_norm": 3.528630728413627, + "learning_rate": 3.963282229821206e-06, + "loss": 1.0642, + "step": 990 + }, + { + "epoch": 0.08937187175902962, + "grad_norm": 1.9156399824293981, + "learning_rate": 3.963170715299906e-06, + "loss": 1.0606, + "step": 991 + }, + { + "epoch": 0.08946205528249988, + "grad_norm": 4.506843661118788, + "learning_rate": 3.963059033270056e-06, + "loss": 0.9726, + "step": 992 + }, + { + "epoch": 0.08955223880597014, + "grad_norm": 1.6614941042055902, + "learning_rate": 3.9629471837411855e-06, + "loss": 0.9981, + "step": 993 + }, + { + "epoch": 0.0896424223294404, + "grad_norm": 1.8663501857226767, + "learning_rate": 3.962835166722838e-06, + "loss": 0.9942, + "step": 994 + }, + { + "epoch": 0.08973260585291068, + "grad_norm": 1.6701232431815207, + "learning_rate": 3.96272298222457e-06, + "loss": 1.0542, + "step": 995 + }, + { + "epoch": 0.08982278937638094, + "grad_norm": 2.3670432673159363, + "learning_rate": 3.962610630255956e-06, + "loss": 1.1716, + "step": 996 + }, + { + "epoch": 0.0899129728998512, + "grad_norm": 2.1878764237370523, + "learning_rate": 3.96249811082658e-06, + "loss": 1.0192, + "step": 997 + }, + { + "epoch": 0.09000315642332146, + "grad_norm": 0.8063893885218002, + "learning_rate": 3.962385423946046e-06, + "loss": 0.9115, + "step": 998 + }, + { + "epoch": 0.09009333994679172, + "grad_norm": 5.952763904423444, + "learning_rate": 3.962272569623966e-06, + "loss": 1.0219, + "step": 999 + }, + { + "epoch": 0.09018352347026198, + "grad_norm": 1.9300882641468655, + "learning_rate": 3.9621595478699704e-06, + "loss": 0.9396, + "step": 1000 + }, + { + "epoch": 0.09027370699373224, + "grad_norm": 1.714384882193025, + "learning_rate": 3.962046358693703e-06, + "loss": 1.0432, + "step": 1001 + }, + { + "epoch": 0.0903638905172025, + "grad_norm": 2.3300537722035224, + "learning_rate": 3.961933002104822e-06, + "loss": 1.128, + "step": 1002 + }, + { + "epoch": 0.09045407404067277, + "grad_norm": 3.1626214407420634, + "learning_rate": 3.961819478112999e-06, + "loss": 0.9791, + "step": 1003 + }, + { + "epoch": 0.09054425756414303, + "grad_norm": 1.6775596447729153, + "learning_rate": 3.961705786727921e-06, + "loss": 1.0699, + "step": 1004 + }, + { + "epoch": 0.09063444108761329, + "grad_norm": 2.3701398209480184, + "learning_rate": 3.961591927959288e-06, + "loss": 0.9722, + "step": 1005 + }, + { + "epoch": 0.09072462461108355, + "grad_norm": 0.7637478529654884, + "learning_rate": 3.961477901816816e-06, + "loss": 0.8326, + "step": 1006 + }, + { + "epoch": 0.09081480813455382, + "grad_norm": 2.007678550675331, + "learning_rate": 3.961363708310233e-06, + "loss": 1.1097, + "step": 1007 + }, + { + "epoch": 0.09090499165802408, + "grad_norm": 1.8123535886363125, + "learning_rate": 3.961249347449286e-06, + "loss": 1.0719, + "step": 1008 + }, + { + "epoch": 0.09099517518149434, + "grad_norm": 5.038520819359501, + "learning_rate": 3.961134819243728e-06, + "loss": 1.0343, + "step": 1009 + }, + { + "epoch": 0.0910853587049646, + "grad_norm": 3.7323763937901377, + "learning_rate": 3.961020123703335e-06, + "loss": 1.0562, + "step": 1010 + }, + { + "epoch": 0.09117554222843487, + "grad_norm": 2.1264875478660223, + "learning_rate": 3.960905260837892e-06, + "loss": 1.0611, + "step": 1011 + }, + { + "epoch": 0.09126572575190513, + "grad_norm": 1.8666186244209164, + "learning_rate": 3.960790230657199e-06, + "loss": 1.016, + "step": 1012 + }, + { + "epoch": 0.09135590927537539, + "grad_norm": 1.4320660592617924, + "learning_rate": 3.960675033171072e-06, + "loss": 0.9976, + "step": 1013 + }, + { + "epoch": 0.09144609279884565, + "grad_norm": 2.387307851359067, + "learning_rate": 3.960559668389341e-06, + "loss": 0.9928, + "step": 1014 + }, + { + "epoch": 0.09153627632231591, + "grad_norm": 3.308542730845569, + "learning_rate": 3.960444136321847e-06, + "loss": 1.0492, + "step": 1015 + }, + { + "epoch": 0.09162645984578617, + "grad_norm": 2.600879934769983, + "learning_rate": 3.960328436978451e-06, + "loss": 1.1321, + "step": 1016 + }, + { + "epoch": 0.09171664336925643, + "grad_norm": 1.1786589528071598, + "learning_rate": 3.960212570369024e-06, + "loss": 0.8322, + "step": 1017 + }, + { + "epoch": 0.09180682689272669, + "grad_norm": 2.1085391048489583, + "learning_rate": 3.9600965365034515e-06, + "loss": 0.9826, + "step": 1018 + }, + { + "epoch": 0.09189701041619697, + "grad_norm": 1.8145485186676935, + "learning_rate": 3.959980335391634e-06, + "loss": 1.0998, + "step": 1019 + }, + { + "epoch": 0.09198719393966723, + "grad_norm": 1.7065797116931487, + "learning_rate": 3.959863967043487e-06, + "loss": 0.987, + "step": 1020 + }, + { + "epoch": 0.09207737746313749, + "grad_norm": 2.0700781347667307, + "learning_rate": 3.9597474314689405e-06, + "loss": 1.077, + "step": 1021 + }, + { + "epoch": 0.09216756098660775, + "grad_norm": 1.8534256458048102, + "learning_rate": 3.959630728677937e-06, + "loss": 1.0656, + "step": 1022 + }, + { + "epoch": 0.09225774451007801, + "grad_norm": 2.2202946833325887, + "learning_rate": 3.959513858680434e-06, + "loss": 1.026, + "step": 1023 + }, + { + "epoch": 0.09234792803354827, + "grad_norm": 3.536297591783504, + "learning_rate": 3.959396821486405e-06, + "loss": 1.1376, + "step": 1024 + }, + { + "epoch": 0.09243811155701853, + "grad_norm": 0.831590311760131, + "learning_rate": 3.959279617105835e-06, + "loss": 0.8639, + "step": 1025 + }, + { + "epoch": 0.09252829508048879, + "grad_norm": 3.193757590566674, + "learning_rate": 3.9591622455487235e-06, + "loss": 1.0586, + "step": 1026 + }, + { + "epoch": 0.09261847860395905, + "grad_norm": 2.085545535175574, + "learning_rate": 3.959044706825087e-06, + "loss": 1.1216, + "step": 1027 + }, + { + "epoch": 0.09270866212742931, + "grad_norm": 1.9562419586503867, + "learning_rate": 3.958927000944954e-06, + "loss": 0.9524, + "step": 1028 + }, + { + "epoch": 0.09279884565089958, + "grad_norm": 2.0876100385525036, + "learning_rate": 3.958809127918368e-06, + "loss": 1.1328, + "step": 1029 + }, + { + "epoch": 0.09288902917436984, + "grad_norm": 1.6677230904200464, + "learning_rate": 3.958691087755387e-06, + "loss": 0.9844, + "step": 1030 + }, + { + "epoch": 0.09297921269784011, + "grad_norm": 2.0896358778772477, + "learning_rate": 3.958572880466081e-06, + "loss": 0.9214, + "step": 1031 + }, + { + "epoch": 0.09306939622131037, + "grad_norm": 1.7151188727461857, + "learning_rate": 3.9584545060605385e-06, + "loss": 1.032, + "step": 1032 + }, + { + "epoch": 0.09315957974478063, + "grad_norm": 1.7395194797636775, + "learning_rate": 3.958335964548859e-06, + "loss": 1.0112, + "step": 1033 + }, + { + "epoch": 0.0932497632682509, + "grad_norm": 2.734208543343026, + "learning_rate": 3.958217255941156e-06, + "loss": 1.0959, + "step": 1034 + }, + { + "epoch": 0.09333994679172115, + "grad_norm": 1.5774327638930514, + "learning_rate": 3.95809838024756e-06, + "loss": 0.9957, + "step": 1035 + }, + { + "epoch": 0.09343013031519141, + "grad_norm": 1.8080299767566317, + "learning_rate": 3.957979337478212e-06, + "loss": 1.0145, + "step": 1036 + }, + { + "epoch": 0.09352031383866168, + "grad_norm": 1.9549806248516972, + "learning_rate": 3.957860127643272e-06, + "loss": 1.0649, + "step": 1037 + }, + { + "epoch": 0.09361049736213194, + "grad_norm": 2.2720728183346557, + "learning_rate": 3.95774075075291e-06, + "loss": 1.0508, + "step": 1038 + }, + { + "epoch": 0.0937006808856022, + "grad_norm": 1.8448483703043654, + "learning_rate": 3.957621206817312e-06, + "loss": 1.104, + "step": 1039 + }, + { + "epoch": 0.09379086440907246, + "grad_norm": 2.1436255371619826, + "learning_rate": 3.957501495846679e-06, + "loss": 1.0328, + "step": 1040 + }, + { + "epoch": 0.09388104793254272, + "grad_norm": 0.6788188162494776, + "learning_rate": 3.957381617851225e-06, + "loss": 0.8371, + "step": 1041 + }, + { + "epoch": 0.09397123145601298, + "grad_norm": 1.5610561566316936, + "learning_rate": 3.9572615728411776e-06, + "loss": 0.969, + "step": 1042 + }, + { + "epoch": 0.09406141497948325, + "grad_norm": 5.068072795667266, + "learning_rate": 3.957141360826781e-06, + "loss": 0.9298, + "step": 1043 + }, + { + "epoch": 0.09415159850295352, + "grad_norm": 1.8042253174913756, + "learning_rate": 3.957020981818292e-06, + "loss": 1.0313, + "step": 1044 + }, + { + "epoch": 0.09424178202642378, + "grad_norm": 1.8247658278843288, + "learning_rate": 3.956900435825982e-06, + "loss": 1.0359, + "step": 1045 + }, + { + "epoch": 0.09433196554989404, + "grad_norm": 1.575721950672935, + "learning_rate": 3.9567797228601364e-06, + "loss": 0.9986, + "step": 1046 + }, + { + "epoch": 0.0944221490733643, + "grad_norm": 1.6744381517417055, + "learning_rate": 3.956658842931055e-06, + "loss": 1.1208, + "step": 1047 + }, + { + "epoch": 0.09451233259683456, + "grad_norm": 1.9639339577939408, + "learning_rate": 3.956537796049052e-06, + "loss": 1.0619, + "step": 1048 + }, + { + "epoch": 0.09460251612030482, + "grad_norm": 0.7743280081541906, + "learning_rate": 3.956416582224457e-06, + "loss": 0.8572, + "step": 1049 + }, + { + "epoch": 0.09469269964377508, + "grad_norm": 2.4197900472945664, + "learning_rate": 3.956295201467611e-06, + "loss": 1.084, + "step": 1050 + }, + { + "epoch": 0.09478288316724534, + "grad_norm": 1.7180944406310994, + "learning_rate": 3.956173653788872e-06, + "loss": 1.0767, + "step": 1051 + }, + { + "epoch": 0.0948730666907156, + "grad_norm": 2.456495557163351, + "learning_rate": 3.95605193919861e-06, + "loss": 1.0825, + "step": 1052 + }, + { + "epoch": 0.09496325021418586, + "grad_norm": 1.8833348985963692, + "learning_rate": 3.955930057707211e-06, + "loss": 1.0301, + "step": 1053 + }, + { + "epoch": 0.09505343373765612, + "grad_norm": 2.2098216553614995, + "learning_rate": 3.955808009325075e-06, + "loss": 1.0591, + "step": 1054 + }, + { + "epoch": 0.0951436172611264, + "grad_norm": 1.797607803189259, + "learning_rate": 3.955685794062615e-06, + "loss": 1.0164, + "step": 1055 + }, + { + "epoch": 0.09523380078459666, + "grad_norm": 2.275020087855596, + "learning_rate": 3.95556341193026e-06, + "loss": 1.1494, + "step": 1056 + }, + { + "epoch": 0.09532398430806692, + "grad_norm": 1.6830302488087228, + "learning_rate": 3.955440862938452e-06, + "loss": 1.1416, + "step": 1057 + }, + { + "epoch": 0.09541416783153718, + "grad_norm": 1.829439847817921, + "learning_rate": 3.955318147097647e-06, + "loss": 1.045, + "step": 1058 + }, + { + "epoch": 0.09550435135500744, + "grad_norm": 2.085809025040507, + "learning_rate": 3.955195264418316e-06, + "loss": 1.0103, + "step": 1059 + }, + { + "epoch": 0.0955945348784777, + "grad_norm": 1.5469593063654057, + "learning_rate": 3.955072214910944e-06, + "loss": 0.903, + "step": 1060 + }, + { + "epoch": 0.09568471840194796, + "grad_norm": 3.4406057026362644, + "learning_rate": 3.954948998586032e-06, + "loss": 0.9666, + "step": 1061 + }, + { + "epoch": 0.09577490192541822, + "grad_norm": 2.020201226966394, + "learning_rate": 3.954825615454089e-06, + "loss": 1.0547, + "step": 1062 + }, + { + "epoch": 0.09586508544888848, + "grad_norm": 2.188556198175972, + "learning_rate": 3.954702065525649e-06, + "loss": 1.0418, + "step": 1063 + }, + { + "epoch": 0.09595526897235875, + "grad_norm": 3.1448069196905855, + "learning_rate": 3.954578348811248e-06, + "loss": 1.0734, + "step": 1064 + }, + { + "epoch": 0.096045452495829, + "grad_norm": 3.0866322954487355, + "learning_rate": 3.954454465321447e-06, + "loss": 1.0237, + "step": 1065 + }, + { + "epoch": 0.09613563601929928, + "grad_norm": 4.187880996907946, + "learning_rate": 3.954330415066813e-06, + "loss": 1.1262, + "step": 1066 + }, + { + "epoch": 0.09622581954276954, + "grad_norm": 1.950225447802502, + "learning_rate": 3.954206198057932e-06, + "loss": 1.0282, + "step": 1067 + }, + { + "epoch": 0.0963160030662398, + "grad_norm": 2.381531217695232, + "learning_rate": 3.954081814305403e-06, + "loss": 1.1052, + "step": 1068 + }, + { + "epoch": 0.09640618658971006, + "grad_norm": 1.690583484866207, + "learning_rate": 3.953957263819839e-06, + "loss": 0.9833, + "step": 1069 + }, + { + "epoch": 0.09649637011318032, + "grad_norm": 2.5158996432110468, + "learning_rate": 3.953832546611867e-06, + "loss": 0.9805, + "step": 1070 + }, + { + "epoch": 0.09658655363665059, + "grad_norm": 2.2467903980780024, + "learning_rate": 3.953707662692129e-06, + "loss": 1.0719, + "step": 1071 + }, + { + "epoch": 0.09667673716012085, + "grad_norm": 2.188116804084149, + "learning_rate": 3.95358261207128e-06, + "loss": 0.9778, + "step": 1072 + }, + { + "epoch": 0.0967669206835911, + "grad_norm": 2.1235642912500357, + "learning_rate": 3.953457394759992e-06, + "loss": 1.0378, + "step": 1073 + }, + { + "epoch": 0.09685710420706137, + "grad_norm": 2.4168893996451692, + "learning_rate": 3.953332010768947e-06, + "loss": 1.0628, + "step": 1074 + }, + { + "epoch": 0.09694728773053163, + "grad_norm": 3.1063452293489413, + "learning_rate": 3.9532064601088436e-06, + "loss": 1.0102, + "step": 1075 + }, + { + "epoch": 0.09703747125400189, + "grad_norm": 4.375376221965402, + "learning_rate": 3.953080742790396e-06, + "loss": 1.1971, + "step": 1076 + }, + { + "epoch": 0.09712765477747215, + "grad_norm": 0.7557579977966975, + "learning_rate": 3.95295485882433e-06, + "loss": 0.8984, + "step": 1077 + }, + { + "epoch": 0.09721783830094242, + "grad_norm": 0.6842461878468612, + "learning_rate": 3.952828808221387e-06, + "loss": 0.867, + "step": 1078 + }, + { + "epoch": 0.09730802182441269, + "grad_norm": 2.422548754096169, + "learning_rate": 3.9527025909923225e-06, + "loss": 1.0222, + "step": 1079 + }, + { + "epoch": 0.09739820534788295, + "grad_norm": 1.7864694919367439, + "learning_rate": 3.952576207147906e-06, + "loss": 1.1386, + "step": 1080 + }, + { + "epoch": 0.09748838887135321, + "grad_norm": 1.984961055210226, + "learning_rate": 3.95244965669892e-06, + "loss": 1.0819, + "step": 1081 + }, + { + "epoch": 0.09757857239482347, + "grad_norm": 2.3764841562132633, + "learning_rate": 3.952322939656165e-06, + "loss": 1.0073, + "step": 1082 + }, + { + "epoch": 0.09766875591829373, + "grad_norm": 2.2173209087729475, + "learning_rate": 3.952196056030451e-06, + "loss": 1.0397, + "step": 1083 + }, + { + "epoch": 0.09775893944176399, + "grad_norm": 1.957440118470568, + "learning_rate": 3.952069005832605e-06, + "loss": 1.0152, + "step": 1084 + }, + { + "epoch": 0.09784912296523425, + "grad_norm": 2.745580416325438, + "learning_rate": 3.951941789073468e-06, + "loss": 1.0696, + "step": 1085 + }, + { + "epoch": 0.09793930648870451, + "grad_norm": 1.7473552337189984, + "learning_rate": 3.9518144057638955e-06, + "loss": 1.0706, + "step": 1086 + }, + { + "epoch": 0.09802949001217477, + "grad_norm": 3.7079467286168963, + "learning_rate": 3.951686855914755e-06, + "loss": 1.0387, + "step": 1087 + }, + { + "epoch": 0.09811967353564503, + "grad_norm": 3.844753475545568, + "learning_rate": 3.9515591395369305e-06, + "loss": 1.0768, + "step": 1088 + }, + { + "epoch": 0.0982098570591153, + "grad_norm": 1.6697840940510391, + "learning_rate": 3.95143125664132e-06, + "loss": 1.0289, + "step": 1089 + }, + { + "epoch": 0.09830004058258557, + "grad_norm": 2.1001512972507466, + "learning_rate": 3.951303207238833e-06, + "loss": 1.0689, + "step": 1090 + }, + { + "epoch": 0.09839022410605583, + "grad_norm": 1.5090036381601168, + "learning_rate": 3.951174991340399e-06, + "loss": 1.018, + "step": 1091 + }, + { + "epoch": 0.09848040762952609, + "grad_norm": 1.6868728602630059, + "learning_rate": 3.9510466089569546e-06, + "loss": 1.0057, + "step": 1092 + }, + { + "epoch": 0.09857059115299635, + "grad_norm": 1.7096771056957967, + "learning_rate": 3.950918060099456e-06, + "loss": 1.0121, + "step": 1093 + }, + { + "epoch": 0.09866077467646661, + "grad_norm": 1.8780766993451163, + "learning_rate": 3.950789344778871e-06, + "loss": 1.0998, + "step": 1094 + }, + { + "epoch": 0.09875095819993687, + "grad_norm": 1.9289215668216546, + "learning_rate": 3.950660463006184e-06, + "loss": 0.9874, + "step": 1095 + }, + { + "epoch": 0.09884114172340713, + "grad_norm": 1.8668946232002253, + "learning_rate": 3.950531414792389e-06, + "loss": 1.1144, + "step": 1096 + }, + { + "epoch": 0.0989313252468774, + "grad_norm": 2.4954478781898954, + "learning_rate": 3.950402200148498e-06, + "loss": 1.0529, + "step": 1097 + }, + { + "epoch": 0.09902150877034765, + "grad_norm": 2.239179873534612, + "learning_rate": 3.950272819085538e-06, + "loss": 1.0197, + "step": 1098 + }, + { + "epoch": 0.09911169229381792, + "grad_norm": 2.203010826669514, + "learning_rate": 3.9501432716145474e-06, + "loss": 1.1298, + "step": 1099 + }, + { + "epoch": 0.09920187581728818, + "grad_norm": 1.953161376614371, + "learning_rate": 3.950013557746579e-06, + "loss": 0.9394, + "step": 1100 + }, + { + "epoch": 0.09929205934075844, + "grad_norm": 2.434849300381867, + "learning_rate": 3.949883677492703e-06, + "loss": 1.0844, + "step": 1101 + }, + { + "epoch": 0.09938224286422871, + "grad_norm": 2.7660026507968296, + "learning_rate": 3.9497536308639994e-06, + "loss": 1.0592, + "step": 1102 + }, + { + "epoch": 0.09947242638769897, + "grad_norm": 1.8134875402493194, + "learning_rate": 3.949623417871565e-06, + "loss": 1.0942, + "step": 1103 + }, + { + "epoch": 0.09956260991116923, + "grad_norm": 2.656059707108372, + "learning_rate": 3.949493038526511e-06, + "loss": 1.0251, + "step": 1104 + }, + { + "epoch": 0.0996527934346395, + "grad_norm": 2.487955548156122, + "learning_rate": 3.949362492839961e-06, + "loss": 0.9974, + "step": 1105 + }, + { + "epoch": 0.09974297695810976, + "grad_norm": 2.346204261069753, + "learning_rate": 3.949231780823054e-06, + "loss": 1.0054, + "step": 1106 + }, + { + "epoch": 0.09983316048158002, + "grad_norm": 1.9216080767105257, + "learning_rate": 3.949100902486945e-06, + "loss": 0.9849, + "step": 1107 + }, + { + "epoch": 0.09992334400505028, + "grad_norm": 2.5171630138340486, + "learning_rate": 3.948969857842799e-06, + "loss": 1.013, + "step": 1108 + }, + { + "epoch": 0.10001352752852054, + "grad_norm": 1.6477086774679626, + "learning_rate": 3.948838646901798e-06, + "loss": 1.0335, + "step": 1109 + }, + { + "epoch": 0.1001037110519908, + "grad_norm": 4.274177445365676, + "learning_rate": 3.948707269675138e-06, + "loss": 1.032, + "step": 1110 + }, + { + "epoch": 0.10019389457546106, + "grad_norm": 2.309948725275806, + "learning_rate": 3.948575726174028e-06, + "loss": 1.0279, + "step": 1111 + }, + { + "epoch": 0.10028407809893132, + "grad_norm": 2.171500194235979, + "learning_rate": 3.9484440164096935e-06, + "loss": 1.1232, + "step": 1112 + }, + { + "epoch": 0.10037426162240158, + "grad_norm": 3.2438270927191075, + "learning_rate": 3.948312140393372e-06, + "loss": 1.0671, + "step": 1113 + }, + { + "epoch": 0.10046444514587186, + "grad_norm": 1.9139079128821532, + "learning_rate": 3.948180098136316e-06, + "loss": 1.0157, + "step": 1114 + }, + { + "epoch": 0.10055462866934212, + "grad_norm": 1.989245887087628, + "learning_rate": 3.948047889649791e-06, + "loss": 1.0109, + "step": 1115 + }, + { + "epoch": 0.10064481219281238, + "grad_norm": 4.014166778718613, + "learning_rate": 3.947915514945079e-06, + "loss": 1.024, + "step": 1116 + }, + { + "epoch": 0.10073499571628264, + "grad_norm": 2.580947625504897, + "learning_rate": 3.947782974033474e-06, + "loss": 1.1078, + "step": 1117 + }, + { + "epoch": 0.1008251792397529, + "grad_norm": 3.4100808815858783, + "learning_rate": 3.9476502669262866e-06, + "loss": 1.0079, + "step": 1118 + }, + { + "epoch": 0.10091536276322316, + "grad_norm": 4.009683570077983, + "learning_rate": 3.947517393634839e-06, + "loss": 1.0131, + "step": 1119 + }, + { + "epoch": 0.10100554628669342, + "grad_norm": 3.1979565115733752, + "learning_rate": 3.947384354170469e-06, + "loss": 1.0815, + "step": 1120 + }, + { + "epoch": 0.10109572981016368, + "grad_norm": 1.7624700936018927, + "learning_rate": 3.947251148544528e-06, + "loss": 1.0791, + "step": 1121 + }, + { + "epoch": 0.10118591333363394, + "grad_norm": 1.8386415088837487, + "learning_rate": 3.947117776768382e-06, + "loss": 1.0668, + "step": 1122 + }, + { + "epoch": 0.1012760968571042, + "grad_norm": 2.36394477546076, + "learning_rate": 3.9469842388534105e-06, + "loss": 1.0247, + "step": 1123 + }, + { + "epoch": 0.10136628038057446, + "grad_norm": 1.7050808516542846, + "learning_rate": 3.946850534811009e-06, + "loss": 1.0503, + "step": 1124 + }, + { + "epoch": 0.10145646390404472, + "grad_norm": 1.7853046914785737, + "learning_rate": 3.946716664652585e-06, + "loss": 1.0256, + "step": 1125 + }, + { + "epoch": 0.101546647427515, + "grad_norm": 3.5570027283836887, + "learning_rate": 3.94658262838956e-06, + "loss": 0.9464, + "step": 1126 + }, + { + "epoch": 0.10163683095098526, + "grad_norm": 1.6781653315920546, + "learning_rate": 3.946448426033373e-06, + "loss": 0.9605, + "step": 1127 + }, + { + "epoch": 0.10172701447445552, + "grad_norm": 1.719354141868723, + "learning_rate": 3.946314057595473e-06, + "loss": 1.0754, + "step": 1128 + }, + { + "epoch": 0.10181719799792578, + "grad_norm": 2.5280274965324008, + "learning_rate": 3.946179523087326e-06, + "loss": 1.1075, + "step": 1129 + }, + { + "epoch": 0.10190738152139604, + "grad_norm": 5.319732231055604, + "learning_rate": 3.9460448225204104e-06, + "loss": 0.9591, + "step": 1130 + }, + { + "epoch": 0.1019975650448663, + "grad_norm": 4.3703003846028885, + "learning_rate": 3.945909955906221e-06, + "loss": 1.0376, + "step": 1131 + }, + { + "epoch": 0.10208774856833656, + "grad_norm": 3.1202117857822995, + "learning_rate": 3.945774923256264e-06, + "loss": 1.0407, + "step": 1132 + }, + { + "epoch": 0.10217793209180683, + "grad_norm": 3.1172375208921306, + "learning_rate": 3.945639724582062e-06, + "loss": 1.1126, + "step": 1133 + }, + { + "epoch": 0.10226811561527709, + "grad_norm": 2.4601850055505268, + "learning_rate": 3.94550435989515e-06, + "loss": 1.0078, + "step": 1134 + }, + { + "epoch": 0.10235829913874735, + "grad_norm": 3.590586854152724, + "learning_rate": 3.945368829207079e-06, + "loss": 1.064, + "step": 1135 + }, + { + "epoch": 0.10244848266221761, + "grad_norm": 2.223234131681245, + "learning_rate": 3.945233132529414e-06, + "loss": 1.0017, + "step": 1136 + }, + { + "epoch": 0.10253866618568787, + "grad_norm": 1.7255279755801955, + "learning_rate": 3.9450972698737304e-06, + "loss": 1.0391, + "step": 1137 + }, + { + "epoch": 0.10262884970915814, + "grad_norm": 1.6362332372775743, + "learning_rate": 3.944961241251623e-06, + "loss": 1.0262, + "step": 1138 + }, + { + "epoch": 0.1027190332326284, + "grad_norm": 3.099463994847962, + "learning_rate": 3.9448250466746985e-06, + "loss": 1.0117, + "step": 1139 + }, + { + "epoch": 0.10280921675609866, + "grad_norm": 2.224235523349557, + "learning_rate": 3.944688686154578e-06, + "loss": 1.0681, + "step": 1140 + }, + { + "epoch": 0.10289940027956893, + "grad_norm": 2.6483669215337624, + "learning_rate": 3.944552159702894e-06, + "loss": 1.0026, + "step": 1141 + }, + { + "epoch": 0.10298958380303919, + "grad_norm": 1.8238611299621248, + "learning_rate": 3.944415467331299e-06, + "loss": 1.033, + "step": 1142 + }, + { + "epoch": 0.10307976732650945, + "grad_norm": 1.5066348205565367, + "learning_rate": 3.944278609051455e-06, + "loss": 0.9972, + "step": 1143 + }, + { + "epoch": 0.10316995084997971, + "grad_norm": 1.9965190994147797, + "learning_rate": 3.944141584875039e-06, + "loss": 0.9802, + "step": 1144 + }, + { + "epoch": 0.10326013437344997, + "grad_norm": 2.104730423544126, + "learning_rate": 3.944004394813743e-06, + "loss": 1.0455, + "step": 1145 + }, + { + "epoch": 0.10335031789692023, + "grad_norm": 3.45973036404494, + "learning_rate": 3.943867038879273e-06, + "loss": 1.0644, + "step": 1146 + }, + { + "epoch": 0.10344050142039049, + "grad_norm": 1.9655082291940527, + "learning_rate": 3.943729517083349e-06, + "loss": 1.0751, + "step": 1147 + }, + { + "epoch": 0.10353068494386075, + "grad_norm": 2.0185472463351597, + "learning_rate": 3.943591829437705e-06, + "loss": 1.073, + "step": 1148 + }, + { + "epoch": 0.10362086846733101, + "grad_norm": 2.38773155886982, + "learning_rate": 3.9434539759540895e-06, + "loss": 0.9818, + "step": 1149 + }, + { + "epoch": 0.10371105199080129, + "grad_norm": 1.5929481134579335, + "learning_rate": 3.943315956644264e-06, + "loss": 0.9827, + "step": 1150 + }, + { + "epoch": 0.10380123551427155, + "grad_norm": 53.76094092862633, + "learning_rate": 3.943177771520006e-06, + "loss": 1.1108, + "step": 1151 + }, + { + "epoch": 0.10389141903774181, + "grad_norm": 1.8986984513136878, + "learning_rate": 3.9430394205931065e-06, + "loss": 1.0877, + "step": 1152 + }, + { + "epoch": 0.10398160256121207, + "grad_norm": 2.1425737852123894, + "learning_rate": 3.942900903875369e-06, + "loss": 0.918, + "step": 1153 + }, + { + "epoch": 0.10407178608468233, + "grad_norm": 1.7021359012599528, + "learning_rate": 3.942762221378614e-06, + "loss": 1.0449, + "step": 1154 + }, + { + "epoch": 0.10416196960815259, + "grad_norm": 1.7639618188527668, + "learning_rate": 3.942623373114673e-06, + "loss": 1.0767, + "step": 1155 + }, + { + "epoch": 0.10425215313162285, + "grad_norm": 2.013765763943878, + "learning_rate": 3.942484359095396e-06, + "loss": 1.0286, + "step": 1156 + }, + { + "epoch": 0.10434233665509311, + "grad_norm": 1.9694273252808767, + "learning_rate": 3.942345179332642e-06, + "loss": 0.9768, + "step": 1157 + }, + { + "epoch": 0.10443252017856337, + "grad_norm": 1.6580634355960484, + "learning_rate": 3.942205833838287e-06, + "loss": 0.9902, + "step": 1158 + }, + { + "epoch": 0.10452270370203363, + "grad_norm": 5.045332445546476, + "learning_rate": 3.9420663226242204e-06, + "loss": 1.1358, + "step": 1159 + }, + { + "epoch": 0.1046128872255039, + "grad_norm": 2.3217213267920433, + "learning_rate": 3.941926645702348e-06, + "loss": 1.0878, + "step": 1160 + }, + { + "epoch": 0.10470307074897416, + "grad_norm": 9.662892448874489, + "learning_rate": 3.941786803084586e-06, + "loss": 1.0277, + "step": 1161 + }, + { + "epoch": 0.10479325427244443, + "grad_norm": 2.5704212107757094, + "learning_rate": 3.941646794782867e-06, + "loss": 0.9889, + "step": 1162 + }, + { + "epoch": 0.10488343779591469, + "grad_norm": 2.1972010624195297, + "learning_rate": 3.941506620809137e-06, + "loss": 1.0567, + "step": 1163 + }, + { + "epoch": 0.10497362131938495, + "grad_norm": 2.0060022409108997, + "learning_rate": 3.941366281175357e-06, + "loss": 1.0343, + "step": 1164 + }, + { + "epoch": 0.10506380484285521, + "grad_norm": 1.7523621556900915, + "learning_rate": 3.941225775893502e-06, + "loss": 0.9802, + "step": 1165 + }, + { + "epoch": 0.10515398836632547, + "grad_norm": 3.664014868843527, + "learning_rate": 3.941085104975559e-06, + "loss": 1.0163, + "step": 1166 + }, + { + "epoch": 0.10524417188979573, + "grad_norm": 0.7593022676651823, + "learning_rate": 3.9409442684335325e-06, + "loss": 0.8607, + "step": 1167 + }, + { + "epoch": 0.105334355413266, + "grad_norm": 2.0888767750481234, + "learning_rate": 3.940803266279438e-06, + "loss": 1.0372, + "step": 1168 + }, + { + "epoch": 0.10542453893673626, + "grad_norm": 1.8419109806705432, + "learning_rate": 3.9406620985253076e-06, + "loss": 0.977, + "step": 1169 + }, + { + "epoch": 0.10551472246020652, + "grad_norm": 4.609644788185114, + "learning_rate": 3.940520765183187e-06, + "loss": 1.0117, + "step": 1170 + }, + { + "epoch": 0.10560490598367678, + "grad_norm": 3.717047678373031, + "learning_rate": 3.940379266265134e-06, + "loss": 1.0754, + "step": 1171 + }, + { + "epoch": 0.10569508950714704, + "grad_norm": 2.8975935784435682, + "learning_rate": 3.940237601783223e-06, + "loss": 1.0049, + "step": 1172 + }, + { + "epoch": 0.1057852730306173, + "grad_norm": 2.8661706768892983, + "learning_rate": 3.940095771749542e-06, + "loss": 1.0406, + "step": 1173 + }, + { + "epoch": 0.10587545655408757, + "grad_norm": 2.398437301188408, + "learning_rate": 3.939953776176192e-06, + "loss": 1.0959, + "step": 1174 + }, + { + "epoch": 0.10596564007755783, + "grad_norm": 1.3595714646356318, + "learning_rate": 3.939811615075288e-06, + "loss": 1.0634, + "step": 1175 + }, + { + "epoch": 0.1060558236010281, + "grad_norm": 2.6128227618576796, + "learning_rate": 3.9396692884589616e-06, + "loss": 1.1026, + "step": 1176 + }, + { + "epoch": 0.10614600712449836, + "grad_norm": 2.533922456537901, + "learning_rate": 3.9395267963393565e-06, + "loss": 1.0212, + "step": 1177 + }, + { + "epoch": 0.10623619064796862, + "grad_norm": 2.107465996337833, + "learning_rate": 3.939384138728631e-06, + "loss": 1.0797, + "step": 1178 + }, + { + "epoch": 0.10632637417143888, + "grad_norm": 2.069782923497735, + "learning_rate": 3.939241315638956e-06, + "loss": 1.029, + "step": 1179 + }, + { + "epoch": 0.10641655769490914, + "grad_norm": 2.9305286878305346, + "learning_rate": 3.93909832708252e-06, + "loss": 1.0241, + "step": 1180 + }, + { + "epoch": 0.1065067412183794, + "grad_norm": 1.6045708663140446, + "learning_rate": 3.938955173071523e-06, + "loss": 1.0481, + "step": 1181 + }, + { + "epoch": 0.10659692474184966, + "grad_norm": 0.6031773677880485, + "learning_rate": 3.938811853618179e-06, + "loss": 0.8473, + "step": 1182 + }, + { + "epoch": 0.10668710826531992, + "grad_norm": 2.1885881169973067, + "learning_rate": 3.938668368734717e-06, + "loss": 1.0763, + "step": 1183 + }, + { + "epoch": 0.10677729178879018, + "grad_norm": 1.8760376919780721, + "learning_rate": 3.93852471843338e-06, + "loss": 0.9164, + "step": 1184 + }, + { + "epoch": 0.10686747531226044, + "grad_norm": 1.9813200495998957, + "learning_rate": 3.9383809027264254e-06, + "loss": 1.0256, + "step": 1185 + }, + { + "epoch": 0.10695765883573072, + "grad_norm": 1.7658491414352522, + "learning_rate": 3.938236921626124e-06, + "loss": 1.0807, + "step": 1186 + }, + { + "epoch": 0.10704784235920098, + "grad_norm": 2.5983779248939465, + "learning_rate": 3.938092775144761e-06, + "loss": 1.0404, + "step": 1187 + }, + { + "epoch": 0.10713802588267124, + "grad_norm": 1.6412659619173284, + "learning_rate": 3.9379484632946355e-06, + "loss": 0.9637, + "step": 1188 + }, + { + "epoch": 0.1072282094061415, + "grad_norm": 1.8274421394675693, + "learning_rate": 3.937803986088062e-06, + "loss": 1.0876, + "step": 1189 + }, + { + "epoch": 0.10731839292961176, + "grad_norm": 1.9426510540765325, + "learning_rate": 3.937659343537367e-06, + "loss": 0.9963, + "step": 1190 + }, + { + "epoch": 0.10740857645308202, + "grad_norm": 4.089384124199974, + "learning_rate": 3.937514535654893e-06, + "loss": 1.1076, + "step": 1191 + }, + { + "epoch": 0.10749875997655228, + "grad_norm": 3.6179819243966973, + "learning_rate": 3.937369562452996e-06, + "loss": 1.0488, + "step": 1192 + }, + { + "epoch": 0.10758894350002254, + "grad_norm": 2.0193595887059184, + "learning_rate": 3.937224423944044e-06, + "loss": 1.099, + "step": 1193 + }, + { + "epoch": 0.1076791270234928, + "grad_norm": 1.7419766011905726, + "learning_rate": 3.937079120140423e-06, + "loss": 0.9707, + "step": 1194 + }, + { + "epoch": 0.10776931054696307, + "grad_norm": 1.5905332827403011, + "learning_rate": 3.936933651054531e-06, + "loss": 1.0782, + "step": 1195 + }, + { + "epoch": 0.10785949407043333, + "grad_norm": 1.8713626548885307, + "learning_rate": 3.936788016698779e-06, + "loss": 1.1583, + "step": 1196 + }, + { + "epoch": 0.1079496775939036, + "grad_norm": 5.144453020594465, + "learning_rate": 3.936642217085594e-06, + "loss": 1.0738, + "step": 1197 + }, + { + "epoch": 0.10803986111737386, + "grad_norm": 2.318934689798429, + "learning_rate": 3.936496252227417e-06, + "loss": 0.9519, + "step": 1198 + }, + { + "epoch": 0.10813004464084412, + "grad_norm": 1.7876119938691952, + "learning_rate": 3.936350122136703e-06, + "loss": 1.0204, + "step": 1199 + }, + { + "epoch": 0.10822022816431438, + "grad_norm": 0.6860108286389419, + "learning_rate": 3.936203826825919e-06, + "loss": 0.8681, + "step": 1200 + }, + { + "epoch": 0.10831041168778464, + "grad_norm": 1.8802596390787392, + "learning_rate": 3.9360573663075475e-06, + "loss": 1.0015, + "step": 1201 + }, + { + "epoch": 0.1084005952112549, + "grad_norm": 2.117237611297242, + "learning_rate": 3.935910740594087e-06, + "loss": 1.0668, + "step": 1202 + }, + { + "epoch": 0.10849077873472517, + "grad_norm": 2.0746173367915857, + "learning_rate": 3.935763949698047e-06, + "loss": 1.0506, + "step": 1203 + }, + { + "epoch": 0.10858096225819543, + "grad_norm": 2.044361580618348, + "learning_rate": 3.935616993631954e-06, + "loss": 1.1085, + "step": 1204 + }, + { + "epoch": 0.10867114578166569, + "grad_norm": 1.907367441313514, + "learning_rate": 3.935469872408345e-06, + "loss": 1.032, + "step": 1205 + }, + { + "epoch": 0.10876132930513595, + "grad_norm": 1.7476877877293462, + "learning_rate": 3.935322586039776e-06, + "loss": 1.0654, + "step": 1206 + }, + { + "epoch": 0.10885151282860621, + "grad_norm": 6.601868752141913, + "learning_rate": 3.935175134538811e-06, + "loss": 0.9711, + "step": 1207 + }, + { + "epoch": 0.10894169635207647, + "grad_norm": 5.069543906619468, + "learning_rate": 3.935027517918034e-06, + "loss": 0.9813, + "step": 1208 + }, + { + "epoch": 0.10903187987554674, + "grad_norm": 5.519224208274752, + "learning_rate": 3.93487973619004e-06, + "loss": 1.0733, + "step": 1209 + }, + { + "epoch": 0.109122063399017, + "grad_norm": 2.387421100371287, + "learning_rate": 3.934731789367438e-06, + "loss": 1.0499, + "step": 1210 + }, + { + "epoch": 0.10921224692248727, + "grad_norm": 0.7934772383244226, + "learning_rate": 3.9345836774628505e-06, + "loss": 0.8458, + "step": 1211 + }, + { + "epoch": 0.10930243044595753, + "grad_norm": 1.8872778401779138, + "learning_rate": 3.934435400488917e-06, + "loss": 1.0294, + "step": 1212 + }, + { + "epoch": 0.10939261396942779, + "grad_norm": 2.479096184591484, + "learning_rate": 3.934286958458289e-06, + "loss": 1.075, + "step": 1213 + }, + { + "epoch": 0.10948279749289805, + "grad_norm": 1.7582230830166887, + "learning_rate": 3.934138351383632e-06, + "loss": 1.1051, + "step": 1214 + }, + { + "epoch": 0.10957298101636831, + "grad_norm": 1.6325815575923157, + "learning_rate": 3.933989579277626e-06, + "loss": 0.9744, + "step": 1215 + }, + { + "epoch": 0.10966316453983857, + "grad_norm": 1.7327398204094415, + "learning_rate": 3.933840642152966e-06, + "loss": 1.033, + "step": 1216 + }, + { + "epoch": 0.10975334806330883, + "grad_norm": 2.1342947144718925, + "learning_rate": 3.933691540022359e-06, + "loss": 1.0693, + "step": 1217 + }, + { + "epoch": 0.10984353158677909, + "grad_norm": 1.7440122619075353, + "learning_rate": 3.933542272898527e-06, + "loss": 1.0071, + "step": 1218 + }, + { + "epoch": 0.10993371511024935, + "grad_norm": 0.7517751904305362, + "learning_rate": 3.933392840794207e-06, + "loss": 0.8035, + "step": 1219 + }, + { + "epoch": 0.11002389863371961, + "grad_norm": 2.812695051423283, + "learning_rate": 3.93324324372215e-06, + "loss": 1.1466, + "step": 1220 + }, + { + "epoch": 0.11011408215718989, + "grad_norm": 1.8341269798208328, + "learning_rate": 3.9330934816951185e-06, + "loss": 1.0615, + "step": 1221 + }, + { + "epoch": 0.11020426568066015, + "grad_norm": 2.4542217354244173, + "learning_rate": 3.932943554725893e-06, + "loss": 1.0685, + "step": 1222 + }, + { + "epoch": 0.11029444920413041, + "grad_norm": 0.7535257515787956, + "learning_rate": 3.932793462827265e-06, + "loss": 0.8781, + "step": 1223 + }, + { + "epoch": 0.11038463272760067, + "grad_norm": 2.1586468408760773, + "learning_rate": 3.932643206012041e-06, + "loss": 1.0976, + "step": 1224 + }, + { + "epoch": 0.11047481625107093, + "grad_norm": 2.093680878822026, + "learning_rate": 3.932492784293043e-06, + "loss": 0.9662, + "step": 1225 + }, + { + "epoch": 0.11056499977454119, + "grad_norm": 2.3808638812829073, + "learning_rate": 3.932342197683104e-06, + "loss": 1.0267, + "step": 1226 + }, + { + "epoch": 0.11065518329801145, + "grad_norm": 2.0902011458782415, + "learning_rate": 3.932191446195075e-06, + "loss": 0.9176, + "step": 1227 + }, + { + "epoch": 0.11074536682148171, + "grad_norm": 1.995250247471121, + "learning_rate": 3.9320405298418175e-06, + "loss": 1.0414, + "step": 1228 + }, + { + "epoch": 0.11083555034495197, + "grad_norm": 5.963376487432214, + "learning_rate": 3.9318894486362076e-06, + "loss": 0.9639, + "step": 1229 + }, + { + "epoch": 0.11092573386842224, + "grad_norm": 2.5119407638745384, + "learning_rate": 3.9317382025911395e-06, + "loss": 1.1132, + "step": 1230 + }, + { + "epoch": 0.1110159173918925, + "grad_norm": 2.324625034244277, + "learning_rate": 3.9315867917195145e-06, + "loss": 1.141, + "step": 1231 + }, + { + "epoch": 0.11110610091536276, + "grad_norm": 1.6997184632488647, + "learning_rate": 3.931435216034256e-06, + "loss": 1.0411, + "step": 1232 + }, + { + "epoch": 0.11119628443883303, + "grad_norm": 2.2957773505260644, + "learning_rate": 3.931283475548293e-06, + "loss": 1.048, + "step": 1233 + }, + { + "epoch": 0.11128646796230329, + "grad_norm": 2.1072101928629516, + "learning_rate": 3.931131570274576e-06, + "loss": 1.0398, + "step": 1234 + }, + { + "epoch": 0.11137665148577355, + "grad_norm": 1.5753638407357304, + "learning_rate": 3.930979500226065e-06, + "loss": 1.0783, + "step": 1235 + }, + { + "epoch": 0.11146683500924381, + "grad_norm": 2.088295164964571, + "learning_rate": 3.930827265415736e-06, + "loss": 0.8797, + "step": 1236 + }, + { + "epoch": 0.11155701853271408, + "grad_norm": 1.9926701578118398, + "learning_rate": 3.930674865856578e-06, + "loss": 1.0516, + "step": 1237 + }, + { + "epoch": 0.11164720205618434, + "grad_norm": 1.9577187189168572, + "learning_rate": 3.930522301561595e-06, + "loss": 1.075, + "step": 1238 + }, + { + "epoch": 0.1117373855796546, + "grad_norm": 2.159187306529129, + "learning_rate": 3.930369572543804e-06, + "loss": 1.0825, + "step": 1239 + }, + { + "epoch": 0.11182756910312486, + "grad_norm": 1.5737345592174579, + "learning_rate": 3.930216678816237e-06, + "loss": 0.9836, + "step": 1240 + }, + { + "epoch": 0.11191775262659512, + "grad_norm": 0.673019764642576, + "learning_rate": 3.930063620391941e-06, + "loss": 0.8519, + "step": 1241 + }, + { + "epoch": 0.11200793615006538, + "grad_norm": 3.7378008463033416, + "learning_rate": 3.9299103972839735e-06, + "loss": 1.0448, + "step": 1242 + }, + { + "epoch": 0.11209811967353564, + "grad_norm": 0.6633405014190725, + "learning_rate": 3.92975700950541e-06, + "loss": 0.8157, + "step": 1243 + }, + { + "epoch": 0.1121883031970059, + "grad_norm": 1.8285995910702486, + "learning_rate": 3.929603457069338e-06, + "loss": 1.0764, + "step": 1244 + }, + { + "epoch": 0.11227848672047618, + "grad_norm": 1.9851216991409848, + "learning_rate": 3.929449739988859e-06, + "loss": 1.0082, + "step": 1245 + }, + { + "epoch": 0.11236867024394644, + "grad_norm": 1.4227950923398551, + "learning_rate": 3.929295858277089e-06, + "loss": 1.0361, + "step": 1246 + }, + { + "epoch": 0.1124588537674167, + "grad_norm": 2.749102619446891, + "learning_rate": 3.9291418119471585e-06, + "loss": 1.0438, + "step": 1247 + }, + { + "epoch": 0.11254903729088696, + "grad_norm": 10.131937923055991, + "learning_rate": 3.928987601012212e-06, + "loss": 1.0676, + "step": 1248 + }, + { + "epoch": 0.11263922081435722, + "grad_norm": 2.1796725146144262, + "learning_rate": 3.928833225485407e-06, + "loss": 1.0332, + "step": 1249 + }, + { + "epoch": 0.11272940433782748, + "grad_norm": 1.694721804606991, + "learning_rate": 3.928678685379915e-06, + "loss": 1.0284, + "step": 1250 + }, + { + "epoch": 0.11281958786129774, + "grad_norm": 1.8161578264889764, + "learning_rate": 3.928523980708924e-06, + "loss": 1.0875, + "step": 1251 + }, + { + "epoch": 0.112909771384768, + "grad_norm": 1.9257304898704541, + "learning_rate": 3.928369111485632e-06, + "loss": 0.999, + "step": 1252 + }, + { + "epoch": 0.11299995490823826, + "grad_norm": 2.3944840543243493, + "learning_rate": 3.928214077723255e-06, + "loss": 1.0334, + "step": 1253 + }, + { + "epoch": 0.11309013843170852, + "grad_norm": 4.210390979072522, + "learning_rate": 3.928058879435021e-06, + "loss": 1.0289, + "step": 1254 + }, + { + "epoch": 0.11318032195517878, + "grad_norm": 3.9525618924839137, + "learning_rate": 3.9279035166341725e-06, + "loss": 1.0294, + "step": 1255 + }, + { + "epoch": 0.11327050547864904, + "grad_norm": 2.7582411824609863, + "learning_rate": 3.927747989333965e-06, + "loss": 1.0251, + "step": 1256 + }, + { + "epoch": 0.11336068900211932, + "grad_norm": 1.8169021421753382, + "learning_rate": 3.927592297547669e-06, + "loss": 1.0013, + "step": 1257 + }, + { + "epoch": 0.11345087252558958, + "grad_norm": 2.032864794399359, + "learning_rate": 3.927436441288571e-06, + "loss": 1.0682, + "step": 1258 + }, + { + "epoch": 0.11354105604905984, + "grad_norm": 1.819913763162057, + "learning_rate": 3.927280420569968e-06, + "loss": 1.0579, + "step": 1259 + }, + { + "epoch": 0.1136312395725301, + "grad_norm": 1.6941858582442424, + "learning_rate": 3.927124235405171e-06, + "loss": 1.0671, + "step": 1260 + }, + { + "epoch": 0.11372142309600036, + "grad_norm": 1.9876743432118038, + "learning_rate": 3.92696788580751e-06, + "loss": 0.8971, + "step": 1261 + }, + { + "epoch": 0.11381160661947062, + "grad_norm": 1.9794201604896646, + "learning_rate": 3.9268113717903225e-06, + "loss": 0.88, + "step": 1262 + }, + { + "epoch": 0.11390179014294088, + "grad_norm": 2.557820117265641, + "learning_rate": 3.926654693366965e-06, + "loss": 1.1104, + "step": 1263 + }, + { + "epoch": 0.11399197366641114, + "grad_norm": 2.364841923178959, + "learning_rate": 3.926497850550805e-06, + "loss": 1.0526, + "step": 1264 + }, + { + "epoch": 0.1140821571898814, + "grad_norm": 2.580737830236362, + "learning_rate": 3.926340843355226e-06, + "loss": 1.0332, + "step": 1265 + }, + { + "epoch": 0.11417234071335167, + "grad_norm": 2.5488068912250106, + "learning_rate": 3.926183671793625e-06, + "loss": 1.0545, + "step": 1266 + }, + { + "epoch": 0.11426252423682193, + "grad_norm": 2.5627026361500795, + "learning_rate": 3.926026335879412e-06, + "loss": 1.0068, + "step": 1267 + }, + { + "epoch": 0.11435270776029219, + "grad_norm": 0.6347044695224863, + "learning_rate": 3.925868835626012e-06, + "loss": 0.7884, + "step": 1268 + }, + { + "epoch": 0.11444289128376246, + "grad_norm": 2.01224156965399, + "learning_rate": 3.925711171046864e-06, + "loss": 1.0511, + "step": 1269 + }, + { + "epoch": 0.11453307480723272, + "grad_norm": 1.7540648118966717, + "learning_rate": 3.925553342155421e-06, + "loss": 1.0008, + "step": 1270 + }, + { + "epoch": 0.11462325833070298, + "grad_norm": 0.7000817668365849, + "learning_rate": 3.9253953489651485e-06, + "loss": 0.8628, + "step": 1271 + }, + { + "epoch": 0.11471344185417325, + "grad_norm": 0.7131656899908726, + "learning_rate": 3.925237191489529e-06, + "loss": 0.8378, + "step": 1272 + }, + { + "epoch": 0.1148036253776435, + "grad_norm": 1.706404442137826, + "learning_rate": 3.925078869742056e-06, + "loss": 1.0086, + "step": 1273 + }, + { + "epoch": 0.11489380890111377, + "grad_norm": 2.3371717579527345, + "learning_rate": 3.92492038373624e-06, + "loss": 0.9742, + "step": 1274 + }, + { + "epoch": 0.11498399242458403, + "grad_norm": 1.8372286537831208, + "learning_rate": 3.924761733485602e-06, + "loss": 1.0911, + "step": 1275 + }, + { + "epoch": 0.11507417594805429, + "grad_norm": 1.6871471742607302, + "learning_rate": 3.92460291900368e-06, + "loss": 1.0359, + "step": 1276 + }, + { + "epoch": 0.11516435947152455, + "grad_norm": 1.7956792418655254, + "learning_rate": 3.924443940304025e-06, + "loss": 1.0613, + "step": 1277 + }, + { + "epoch": 0.11525454299499481, + "grad_norm": 1.938319924893939, + "learning_rate": 3.924284797400202e-06, + "loss": 1.0876, + "step": 1278 + }, + { + "epoch": 0.11534472651846507, + "grad_norm": 2.082773158738541, + "learning_rate": 3.924125490305789e-06, + "loss": 1.0619, + "step": 1279 + }, + { + "epoch": 0.11543491004193533, + "grad_norm": 2.103142240320426, + "learning_rate": 3.923966019034381e-06, + "loss": 1.0183, + "step": 1280 + }, + { + "epoch": 0.1155250935654056, + "grad_norm": 2.2165362637627744, + "learning_rate": 3.923806383599583e-06, + "loss": 1.0466, + "step": 1281 + }, + { + "epoch": 0.11561527708887587, + "grad_norm": 1.7878697846464677, + "learning_rate": 3.923646584015017e-06, + "loss": 1.0018, + "step": 1282 + }, + { + "epoch": 0.11570546061234613, + "grad_norm": 3.0252817133095644, + "learning_rate": 3.923486620294316e-06, + "loss": 1.0386, + "step": 1283 + }, + { + "epoch": 0.11579564413581639, + "grad_norm": 1.9257743169880273, + "learning_rate": 3.923326492451132e-06, + "loss": 1.0476, + "step": 1284 + }, + { + "epoch": 0.11588582765928665, + "grad_norm": 2.7396833935777867, + "learning_rate": 3.923166200499125e-06, + "loss": 1.0197, + "step": 1285 + }, + { + "epoch": 0.11597601118275691, + "grad_norm": 1.7632543126568287, + "learning_rate": 3.923005744451975e-06, + "loss": 1.0092, + "step": 1286 + }, + { + "epoch": 0.11606619470622717, + "grad_norm": 5.392381641714305, + "learning_rate": 3.9228451243233715e-06, + "loss": 0.9931, + "step": 1287 + }, + { + "epoch": 0.11615637822969743, + "grad_norm": 4.158271699396394, + "learning_rate": 3.9226843401270195e-06, + "loss": 1.069, + "step": 1288 + }, + { + "epoch": 0.1162465617531677, + "grad_norm": 2.1831667897232423, + "learning_rate": 3.9225233918766376e-06, + "loss": 1.0414, + "step": 1289 + }, + { + "epoch": 0.11633674527663795, + "grad_norm": 1.9034089977414486, + "learning_rate": 3.92236227958596e-06, + "loss": 1.0554, + "step": 1290 + }, + { + "epoch": 0.11642692880010821, + "grad_norm": 2.3956413744997067, + "learning_rate": 3.922201003268731e-06, + "loss": 1.0603, + "step": 1291 + }, + { + "epoch": 0.11651711232357848, + "grad_norm": 0.6965556229687869, + "learning_rate": 3.922039562938715e-06, + "loss": 0.8644, + "step": 1292 + }, + { + "epoch": 0.11660729584704875, + "grad_norm": 1.5408963349444265, + "learning_rate": 3.921877958609685e-06, + "loss": 1.0904, + "step": 1293 + }, + { + "epoch": 0.11669747937051901, + "grad_norm": 0.7816874233193046, + "learning_rate": 3.921716190295431e-06, + "loss": 0.9284, + "step": 1294 + }, + { + "epoch": 0.11678766289398927, + "grad_norm": 1.9892667415308012, + "learning_rate": 3.921554258009755e-06, + "loss": 0.9914, + "step": 1295 + }, + { + "epoch": 0.11687784641745953, + "grad_norm": 1.6684812522248726, + "learning_rate": 3.921392161766474e-06, + "loss": 1.0268, + "step": 1296 + }, + { + "epoch": 0.1169680299409298, + "grad_norm": 2.496171212822623, + "learning_rate": 3.92122990157942e-06, + "loss": 1.0166, + "step": 1297 + }, + { + "epoch": 0.11705821346440005, + "grad_norm": 2.6438817814297266, + "learning_rate": 3.921067477462437e-06, + "loss": 1.0828, + "step": 1298 + }, + { + "epoch": 0.11714839698787032, + "grad_norm": 2.4088845630824314, + "learning_rate": 3.920904889429385e-06, + "loss": 1.0912, + "step": 1299 + }, + { + "epoch": 0.11723858051134058, + "grad_norm": 1.9284345134192105, + "learning_rate": 3.920742137494135e-06, + "loss": 1.0762, + "step": 1300 + }, + { + "epoch": 0.11732876403481084, + "grad_norm": 1.6145166280769723, + "learning_rate": 3.920579221670575e-06, + "loss": 1.0954, + "step": 1301 + }, + { + "epoch": 0.1174189475582811, + "grad_norm": 2.9354123752870365, + "learning_rate": 3.920416141972606e-06, + "loss": 1.0396, + "step": 1302 + }, + { + "epoch": 0.11750913108175136, + "grad_norm": 1.594017604725878, + "learning_rate": 3.920252898414143e-06, + "loss": 1.0335, + "step": 1303 + }, + { + "epoch": 0.11759931460522162, + "grad_norm": 2.493378166451836, + "learning_rate": 3.920089491009114e-06, + "loss": 1.051, + "step": 1304 + }, + { + "epoch": 0.1176894981286919, + "grad_norm": 1.6751786150994683, + "learning_rate": 3.919925919771463e-06, + "loss": 1.1205, + "step": 1305 + }, + { + "epoch": 0.11777968165216215, + "grad_norm": 1.9847208157282192, + "learning_rate": 3.919762184715146e-06, + "loss": 0.951, + "step": 1306 + }, + { + "epoch": 0.11786986517563242, + "grad_norm": 2.9326965341621722, + "learning_rate": 3.919598285854134e-06, + "loss": 1.0379, + "step": 1307 + }, + { + "epoch": 0.11796004869910268, + "grad_norm": 2.1890320998374264, + "learning_rate": 3.919434223202411e-06, + "loss": 1.0953, + "step": 1308 + }, + { + "epoch": 0.11805023222257294, + "grad_norm": 2.1589503310949536, + "learning_rate": 3.919269996773977e-06, + "loss": 1.0908, + "step": 1309 + }, + { + "epoch": 0.1181404157460432, + "grad_norm": 1.3850290669307201, + "learning_rate": 3.919105606582844e-06, + "loss": 1.053, + "step": 1310 + }, + { + "epoch": 0.11823059926951346, + "grad_norm": 2.481084119503113, + "learning_rate": 3.918941052643039e-06, + "loss": 1.0109, + "step": 1311 + }, + { + "epoch": 0.11832078279298372, + "grad_norm": 1.7287680921069868, + "learning_rate": 3.918776334968602e-06, + "loss": 1.0512, + "step": 1312 + }, + { + "epoch": 0.11841096631645398, + "grad_norm": 1.9696618042543312, + "learning_rate": 3.918611453573589e-06, + "loss": 1.1484, + "step": 1313 + }, + { + "epoch": 0.11850114983992424, + "grad_norm": 4.065309931915314, + "learning_rate": 3.918446408472066e-06, + "loss": 0.9748, + "step": 1314 + }, + { + "epoch": 0.1185913333633945, + "grad_norm": 1.7405023336419534, + "learning_rate": 3.918281199678119e-06, + "loss": 1.0746, + "step": 1315 + }, + { + "epoch": 0.11868151688686476, + "grad_norm": 0.7483352224910936, + "learning_rate": 3.9181158272058414e-06, + "loss": 0.8466, + "step": 1316 + }, + { + "epoch": 0.11877170041033504, + "grad_norm": 1.5047822853094166, + "learning_rate": 3.9179502910693455e-06, + "loss": 1.0695, + "step": 1317 + }, + { + "epoch": 0.1188618839338053, + "grad_norm": 1.5117143340132528, + "learning_rate": 3.917784591282756e-06, + "loss": 0.9963, + "step": 1318 + }, + { + "epoch": 0.11895206745727556, + "grad_norm": 9.000363236620691, + "learning_rate": 3.9176187278602105e-06, + "loss": 1.0636, + "step": 1319 + }, + { + "epoch": 0.11904225098074582, + "grad_norm": 1.6595699935684716, + "learning_rate": 3.9174527008158606e-06, + "loss": 0.9795, + "step": 1320 + }, + { + "epoch": 0.11913243450421608, + "grad_norm": 1.6334517108422404, + "learning_rate": 3.917286510163874e-06, + "loss": 0.9576, + "step": 1321 + }, + { + "epoch": 0.11922261802768634, + "grad_norm": 1.1772203225060518, + "learning_rate": 3.917120155918431e-06, + "loss": 1.0127, + "step": 1322 + }, + { + "epoch": 0.1193128015511566, + "grad_norm": 0.7075195733722598, + "learning_rate": 3.916953638093725e-06, + "loss": 0.8773, + "step": 1323 + }, + { + "epoch": 0.11940298507462686, + "grad_norm": 2.02442101101527, + "learning_rate": 3.916786956703964e-06, + "loss": 1.0279, + "step": 1324 + }, + { + "epoch": 0.11949316859809712, + "grad_norm": 2.1082386488127716, + "learning_rate": 3.916620111763372e-06, + "loss": 1.0262, + "step": 1325 + }, + { + "epoch": 0.11958335212156739, + "grad_norm": 1.5419251594688226, + "learning_rate": 3.916453103286183e-06, + "loss": 1.0993, + "step": 1326 + }, + { + "epoch": 0.11967353564503765, + "grad_norm": 2.2512805261824567, + "learning_rate": 3.916285931286648e-06, + "loss": 1.082, + "step": 1327 + }, + { + "epoch": 0.11976371916850792, + "grad_norm": 1.6888778147975743, + "learning_rate": 3.916118595779031e-06, + "loss": 0.9865, + "step": 1328 + }, + { + "epoch": 0.11985390269197818, + "grad_norm": 1.8506108667310008, + "learning_rate": 3.915951096777611e-06, + "loss": 1.0098, + "step": 1329 + }, + { + "epoch": 0.11994408621544844, + "grad_norm": 1.5318362612360998, + "learning_rate": 3.915783434296678e-06, + "loss": 0.9864, + "step": 1330 + }, + { + "epoch": 0.1200342697389187, + "grad_norm": 3.88821360916496, + "learning_rate": 3.91561560835054e-06, + "loss": 1.0206, + "step": 1331 + }, + { + "epoch": 0.12012445326238896, + "grad_norm": 1.6964305935037365, + "learning_rate": 3.915447618953515e-06, + "loss": 1.0534, + "step": 1332 + }, + { + "epoch": 0.12021463678585922, + "grad_norm": 2.181678764809018, + "learning_rate": 3.915279466119937e-06, + "loss": 0.9936, + "step": 1333 + }, + { + "epoch": 0.12030482030932949, + "grad_norm": 0.7146752581029548, + "learning_rate": 3.9151111498641546e-06, + "loss": 0.8615, + "step": 1334 + }, + { + "epoch": 0.12039500383279975, + "grad_norm": 1.8421038374859449, + "learning_rate": 3.914942670200529e-06, + "loss": 0.9064, + "step": 1335 + }, + { + "epoch": 0.12048518735627001, + "grad_norm": 1.6638140425762507, + "learning_rate": 3.914774027143436e-06, + "loss": 1.0975, + "step": 1336 + }, + { + "epoch": 0.12057537087974027, + "grad_norm": 2.1234087876827097, + "learning_rate": 3.914605220707265e-06, + "loss": 1.0264, + "step": 1337 + }, + { + "epoch": 0.12066555440321053, + "grad_norm": 6.266027118973624, + "learning_rate": 3.9144362509064194e-06, + "loss": 1.1868, + "step": 1338 + }, + { + "epoch": 0.12075573792668079, + "grad_norm": 2.302328577111365, + "learning_rate": 3.914267117755317e-06, + "loss": 1.1046, + "step": 1339 + }, + { + "epoch": 0.12084592145015106, + "grad_norm": 1.5158920052896412, + "learning_rate": 3.914097821268389e-06, + "loss": 1.0033, + "step": 1340 + }, + { + "epoch": 0.12093610497362133, + "grad_norm": 1.5595239143898476, + "learning_rate": 3.913928361460081e-06, + "loss": 1.0495, + "step": 1341 + }, + { + "epoch": 0.12102628849709159, + "grad_norm": 1.552399605224685, + "learning_rate": 3.913758738344851e-06, + "loss": 1.0247, + "step": 1342 + }, + { + "epoch": 0.12111647202056185, + "grad_norm": 0.7293901282721656, + "learning_rate": 3.913588951937174e-06, + "loss": 0.8386, + "step": 1343 + }, + { + "epoch": 0.12120665554403211, + "grad_norm": 0.7442575197695285, + "learning_rate": 3.9134190022515355e-06, + "loss": 0.8387, + "step": 1344 + }, + { + "epoch": 0.12129683906750237, + "grad_norm": 2.046298928768152, + "learning_rate": 3.913248889302438e-06, + "loss": 0.9995, + "step": 1345 + }, + { + "epoch": 0.12138702259097263, + "grad_norm": 2.4813036857581605, + "learning_rate": 3.913078613104395e-06, + "loss": 0.974, + "step": 1346 + }, + { + "epoch": 0.12147720611444289, + "grad_norm": 1.7761993412791053, + "learning_rate": 3.912908173671936e-06, + "loss": 0.9204, + "step": 1347 + }, + { + "epoch": 0.12156738963791315, + "grad_norm": 1.834349379612394, + "learning_rate": 3.9127375710196044e-06, + "loss": 0.9701, + "step": 1348 + }, + { + "epoch": 0.12165757316138341, + "grad_norm": 3.0266690591418968, + "learning_rate": 3.912566805161957e-06, + "loss": 0.9819, + "step": 1349 + }, + { + "epoch": 0.12174775668485367, + "grad_norm": 5.117092079263012, + "learning_rate": 3.912395876113564e-06, + "loss": 1.0564, + "step": 1350 + }, + { + "epoch": 0.12183794020832393, + "grad_norm": 4.367733397511606, + "learning_rate": 3.912224783889009e-06, + "loss": 1.042, + "step": 1351 + }, + { + "epoch": 0.12192812373179421, + "grad_norm": 1.6925322768769473, + "learning_rate": 3.912053528502892e-06, + "loss": 1.0506, + "step": 1352 + }, + { + "epoch": 0.12201830725526447, + "grad_norm": 1.5137749618374785, + "learning_rate": 3.911882109969825e-06, + "loss": 0.9661, + "step": 1353 + }, + { + "epoch": 0.12210849077873473, + "grad_norm": 1.842380014702735, + "learning_rate": 3.911710528304435e-06, + "loss": 1.0446, + "step": 1354 + }, + { + "epoch": 0.12219867430220499, + "grad_norm": 2.026388601166976, + "learning_rate": 3.911538783521361e-06, + "loss": 0.9835, + "step": 1355 + }, + { + "epoch": 0.12228885782567525, + "grad_norm": 1.649157615406757, + "learning_rate": 3.9113668756352575e-06, + "loss": 1.0478, + "step": 1356 + }, + { + "epoch": 0.12237904134914551, + "grad_norm": 1.6374076904602721, + "learning_rate": 3.911194804660793e-06, + "loss": 0.9646, + "step": 1357 + }, + { + "epoch": 0.12246922487261577, + "grad_norm": 1.7662164153343465, + "learning_rate": 3.91102257061265e-06, + "loss": 1.1171, + "step": 1358 + }, + { + "epoch": 0.12255940839608603, + "grad_norm": 1.5157601994255145, + "learning_rate": 3.910850173505524e-06, + "loss": 1.0742, + "step": 1359 + }, + { + "epoch": 0.1226495919195563, + "grad_norm": 1.8659418651687552, + "learning_rate": 3.9106776133541255e-06, + "loss": 1.1121, + "step": 1360 + }, + { + "epoch": 0.12273977544302656, + "grad_norm": 1.8731882879278392, + "learning_rate": 3.9105048901731766e-06, + "loss": 0.9941, + "step": 1361 + }, + { + "epoch": 0.12282995896649682, + "grad_norm": 2.4250335769688363, + "learning_rate": 3.9103320039774165e-06, + "loss": 1.0272, + "step": 1362 + }, + { + "epoch": 0.12292014248996708, + "grad_norm": 2.1879527032262778, + "learning_rate": 3.9101589547815965e-06, + "loss": 1.0938, + "step": 1363 + }, + { + "epoch": 0.12301032601343735, + "grad_norm": 1.8816378399600846, + "learning_rate": 3.909985742600482e-06, + "loss": 1.0245, + "step": 1364 + }, + { + "epoch": 0.12310050953690761, + "grad_norm": 1.793049197584181, + "learning_rate": 3.909812367448852e-06, + "loss": 1.0004, + "step": 1365 + }, + { + "epoch": 0.12319069306037787, + "grad_norm": 2.061716335420976, + "learning_rate": 3.909638829341501e-06, + "loss": 1.0172, + "step": 1366 + }, + { + "epoch": 0.12328087658384813, + "grad_norm": 2.099391226857677, + "learning_rate": 3.909465128293234e-06, + "loss": 1.0926, + "step": 1367 + }, + { + "epoch": 0.1233710601073184, + "grad_norm": 1.7862393955906317, + "learning_rate": 3.9092912643188745e-06, + "loss": 1.0889, + "step": 1368 + }, + { + "epoch": 0.12346124363078866, + "grad_norm": 1.6863093237483797, + "learning_rate": 3.909117237433256e-06, + "loss": 1.0311, + "step": 1369 + }, + { + "epoch": 0.12355142715425892, + "grad_norm": 1.719839548345136, + "learning_rate": 3.908943047651229e-06, + "loss": 1.0795, + "step": 1370 + }, + { + "epoch": 0.12364161067772918, + "grad_norm": 1.6782635707934461, + "learning_rate": 3.908768694987655e-06, + "loss": 1.0267, + "step": 1371 + }, + { + "epoch": 0.12373179420119944, + "grad_norm": 2.11877517010077, + "learning_rate": 3.908594179457411e-06, + "loss": 0.9573, + "step": 1372 + }, + { + "epoch": 0.1238219777246697, + "grad_norm": 1.6936376358429153, + "learning_rate": 3.908419501075388e-06, + "loss": 1.0415, + "step": 1373 + }, + { + "epoch": 0.12391216124813996, + "grad_norm": 1.8295737500751674, + "learning_rate": 3.90824465985649e-06, + "loss": 1.1109, + "step": 1374 + }, + { + "epoch": 0.12400234477161022, + "grad_norm": 4.536852300145439, + "learning_rate": 3.908069655815636e-06, + "loss": 1.0773, + "step": 1375 + }, + { + "epoch": 0.1240925282950805, + "grad_norm": 2.1340901666485386, + "learning_rate": 3.907894488967758e-06, + "loss": 0.9724, + "step": 1376 + }, + { + "epoch": 0.12418271181855076, + "grad_norm": 1.829090939158452, + "learning_rate": 3.9077191593278005e-06, + "loss": 1.096, + "step": 1377 + }, + { + "epoch": 0.12427289534202102, + "grad_norm": 0.7285301882027132, + "learning_rate": 3.9075436669107265e-06, + "loss": 0.815, + "step": 1378 + }, + { + "epoch": 0.12436307886549128, + "grad_norm": 1.901467687520343, + "learning_rate": 3.90736801173151e-06, + "loss": 1.0477, + "step": 1379 + }, + { + "epoch": 0.12445326238896154, + "grad_norm": 1.4034049606543266, + "learning_rate": 3.907192193805136e-06, + "loss": 1.0481, + "step": 1380 + }, + { + "epoch": 0.1245434459124318, + "grad_norm": 5.9023938382016015, + "learning_rate": 3.907016213146608e-06, + "loss": 1.0881, + "step": 1381 + }, + { + "epoch": 0.12463362943590206, + "grad_norm": 1.6018965209991742, + "learning_rate": 3.906840069770942e-06, + "loss": 1.1489, + "step": 1382 + }, + { + "epoch": 0.12472381295937232, + "grad_norm": 1.5988786493217546, + "learning_rate": 3.906663763693167e-06, + "loss": 1.0688, + "step": 1383 + }, + { + "epoch": 0.12481399648284258, + "grad_norm": 1.7543430251797818, + "learning_rate": 3.906487294928327e-06, + "loss": 1.0211, + "step": 1384 + }, + { + "epoch": 0.12490418000631284, + "grad_norm": 2.061792078770389, + "learning_rate": 3.906310663491478e-06, + "loss": 1.0393, + "step": 1385 + }, + { + "epoch": 0.1249943635297831, + "grad_norm": 2.4900059254893687, + "learning_rate": 3.906133869397692e-06, + "loss": 1.0895, + "step": 1386 + }, + { + "epoch": 0.12508454705325336, + "grad_norm": 1.8972594448190792, + "learning_rate": 3.905956912662054e-06, + "loss": 1.0044, + "step": 1387 + }, + { + "epoch": 0.12517473057672363, + "grad_norm": 2.126689912116904, + "learning_rate": 3.905779793299662e-06, + "loss": 1.0995, + "step": 1388 + }, + { + "epoch": 0.12526491410019389, + "grad_norm": 1.9128529989551137, + "learning_rate": 3.905602511325631e-06, + "loss": 1.0718, + "step": 1389 + }, + { + "epoch": 0.12535509762366415, + "grad_norm": 2.9541866271327017, + "learning_rate": 3.905425066755086e-06, + "loss": 0.9989, + "step": 1390 + }, + { + "epoch": 0.1254452811471344, + "grad_norm": 1.3391919121363574, + "learning_rate": 3.905247459603168e-06, + "loss": 0.9941, + "step": 1391 + }, + { + "epoch": 0.12553546467060467, + "grad_norm": 2.054560204105419, + "learning_rate": 3.905069689885031e-06, + "loss": 0.9143, + "step": 1392 + }, + { + "epoch": 0.12562564819407493, + "grad_norm": 1.4595333430058057, + "learning_rate": 3.904891757615843e-06, + "loss": 0.9207, + "step": 1393 + }, + { + "epoch": 0.12571583171754522, + "grad_norm": 0.6106210342637662, + "learning_rate": 3.9047136628107874e-06, + "loss": 0.8006, + "step": 1394 + }, + { + "epoch": 0.12580601524101548, + "grad_norm": 1.7571933588442141, + "learning_rate": 3.904535405485059e-06, + "loss": 1.1672, + "step": 1395 + }, + { + "epoch": 0.12589619876448574, + "grad_norm": 1.9269476834229433, + "learning_rate": 3.90435698565387e-06, + "loss": 0.9858, + "step": 1396 + }, + { + "epoch": 0.125986382287956, + "grad_norm": 1.7416560983413247, + "learning_rate": 3.904178403332441e-06, + "loss": 0.9998, + "step": 1397 + }, + { + "epoch": 0.12607656581142626, + "grad_norm": 2.1705196185663236, + "learning_rate": 3.903999658536012e-06, + "loss": 1.124, + "step": 1398 + }, + { + "epoch": 0.12616674933489652, + "grad_norm": 1.9571831010394154, + "learning_rate": 3.903820751279833e-06, + "loss": 1.0384, + "step": 1399 + }, + { + "epoch": 0.12625693285836678, + "grad_norm": 2.2651748736244106, + "learning_rate": 3.90364168157917e-06, + "loss": 1.0655, + "step": 1400 + }, + { + "epoch": 0.12634711638183704, + "grad_norm": 1.8507232695579852, + "learning_rate": 3.903462449449302e-06, + "loss": 1.0691, + "step": 1401 + }, + { + "epoch": 0.1264372999053073, + "grad_norm": 1.9103908277765849, + "learning_rate": 3.903283054905522e-06, + "loss": 1.0075, + "step": 1402 + }, + { + "epoch": 0.12652748342877757, + "grad_norm": 1.62248394656946, + "learning_rate": 3.9031034979631385e-06, + "loss": 0.9474, + "step": 1403 + }, + { + "epoch": 0.12661766695224783, + "grad_norm": 1.5559688635900366, + "learning_rate": 3.902923778637469e-06, + "loss": 1.0696, + "step": 1404 + }, + { + "epoch": 0.1267078504757181, + "grad_norm": 1.6783862374133305, + "learning_rate": 3.902743896943852e-06, + "loss": 0.9668, + "step": 1405 + }, + { + "epoch": 0.12679803399918835, + "grad_norm": 0.7209402381768911, + "learning_rate": 3.902563852897633e-06, + "loss": 0.8704, + "step": 1406 + }, + { + "epoch": 0.1268882175226586, + "grad_norm": 0.6971400385577424, + "learning_rate": 3.9023836465141755e-06, + "loss": 0.8442, + "step": 1407 + }, + { + "epoch": 0.12697840104612887, + "grad_norm": 3.1129825692869804, + "learning_rate": 3.902203277808856e-06, + "loss": 1.0813, + "step": 1408 + }, + { + "epoch": 0.12706858456959913, + "grad_norm": 1.7855613465892097, + "learning_rate": 3.902022746797064e-06, + "loss": 1.0943, + "step": 1409 + }, + { + "epoch": 0.1271587680930694, + "grad_norm": 2.04655267638247, + "learning_rate": 3.9018420534942035e-06, + "loss": 1.1478, + "step": 1410 + }, + { + "epoch": 0.12724895161653965, + "grad_norm": 2.222762120412171, + "learning_rate": 3.9016611979156935e-06, + "loss": 1.0284, + "step": 1411 + }, + { + "epoch": 0.1273391351400099, + "grad_norm": 2.06875471131863, + "learning_rate": 3.9014801800769635e-06, + "loss": 0.9414, + "step": 1412 + }, + { + "epoch": 0.12742931866348017, + "grad_norm": 1.8566839796157404, + "learning_rate": 3.901298999993459e-06, + "loss": 1.0633, + "step": 1413 + }, + { + "epoch": 0.12751950218695043, + "grad_norm": 2.267857996765594, + "learning_rate": 3.901117657680642e-06, + "loss": 0.9701, + "step": 1414 + }, + { + "epoch": 0.1276096857104207, + "grad_norm": 1.6366867754232441, + "learning_rate": 3.900936153153982e-06, + "loss": 0.9677, + "step": 1415 + }, + { + "epoch": 0.12769986923389096, + "grad_norm": 0.7203859287659519, + "learning_rate": 3.900754486428968e-06, + "loss": 0.7974, + "step": 1416 + }, + { + "epoch": 0.12779005275736122, + "grad_norm": 1.883886197336121, + "learning_rate": 3.900572657521102e-06, + "loss": 1.1245, + "step": 1417 + }, + { + "epoch": 0.1278802362808315, + "grad_norm": 3.43218847340585, + "learning_rate": 3.900390666445896e-06, + "loss": 1.0183, + "step": 1418 + }, + { + "epoch": 0.12797041980430177, + "grad_norm": 1.9363342592882142, + "learning_rate": 3.9002085132188795e-06, + "loss": 1.0036, + "step": 1419 + }, + { + "epoch": 0.12806060332777203, + "grad_norm": 1.4799086182967214, + "learning_rate": 3.9000261978555964e-06, + "loss": 1.0793, + "step": 1420 + }, + { + "epoch": 0.1281507868512423, + "grad_norm": 1.9169735939185193, + "learning_rate": 3.8998437203716e-06, + "loss": 1.1039, + "step": 1421 + }, + { + "epoch": 0.12824097037471255, + "grad_norm": 1.6029971512348564, + "learning_rate": 3.899661080782462e-06, + "loss": 1.0596, + "step": 1422 + }, + { + "epoch": 0.1283311538981828, + "grad_norm": 1.8048906088145584, + "learning_rate": 3.899478279103767e-06, + "loss": 1.0656, + "step": 1423 + }, + { + "epoch": 0.12842133742165307, + "grad_norm": 2.0499434440068605, + "learning_rate": 3.8992953153511105e-06, + "loss": 1.0668, + "step": 1424 + }, + { + "epoch": 0.12851152094512333, + "grad_norm": 1.7797783912980143, + "learning_rate": 3.899112189540106e-06, + "loss": 1.0963, + "step": 1425 + }, + { + "epoch": 0.1286017044685936, + "grad_norm": 5.102247206814684, + "learning_rate": 3.898928901686377e-06, + "loss": 1.0714, + "step": 1426 + }, + { + "epoch": 0.12869188799206385, + "grad_norm": 0.7249466350251847, + "learning_rate": 3.898745451805564e-06, + "loss": 0.8026, + "step": 1427 + }, + { + "epoch": 0.1287820715155341, + "grad_norm": 1.6978018630207297, + "learning_rate": 3.898561839913319e-06, + "loss": 0.9665, + "step": 1428 + }, + { + "epoch": 0.12887225503900437, + "grad_norm": 2.2805817552869283, + "learning_rate": 3.89837806602531e-06, + "loss": 0.8844, + "step": 1429 + }, + { + "epoch": 0.12896243856247463, + "grad_norm": 1.891778956314033, + "learning_rate": 3.898194130157217e-06, + "loss": 1.0452, + "step": 1430 + }, + { + "epoch": 0.1290526220859449, + "grad_norm": 2.1666460280780355, + "learning_rate": 3.8980100323247335e-06, + "loss": 0.9712, + "step": 1431 + }, + { + "epoch": 0.12914280560941516, + "grad_norm": 2.26297620956855, + "learning_rate": 3.897825772543568e-06, + "loss": 0.9717, + "step": 1432 + }, + { + "epoch": 0.12923298913288542, + "grad_norm": 1.907714156381851, + "learning_rate": 3.897641350829444e-06, + "loss": 1.0905, + "step": 1433 + }, + { + "epoch": 0.12932317265635568, + "grad_norm": 2.3328060735165583, + "learning_rate": 3.897456767198096e-06, + "loss": 1.1197, + "step": 1434 + }, + { + "epoch": 0.12941335617982594, + "grad_norm": 2.1473543523174055, + "learning_rate": 3.897272021665275e-06, + "loss": 0.9774, + "step": 1435 + }, + { + "epoch": 0.1295035397032962, + "grad_norm": 1.569054246556726, + "learning_rate": 3.897087114246743e-06, + "loss": 1.1473, + "step": 1436 + }, + { + "epoch": 0.12959372322676646, + "grad_norm": 1.7984110362130037, + "learning_rate": 3.896902044958279e-06, + "loss": 1.1538, + "step": 1437 + }, + { + "epoch": 0.12968390675023672, + "grad_norm": 2.2062282485538067, + "learning_rate": 3.896716813815672e-06, + "loss": 1.089, + "step": 1438 + }, + { + "epoch": 0.12977409027370698, + "grad_norm": 2.561705582454865, + "learning_rate": 3.896531420834728e-06, + "loss": 0.9873, + "step": 1439 + }, + { + "epoch": 0.12986427379717724, + "grad_norm": 1.8404773180122802, + "learning_rate": 3.896345866031266e-06, + "loss": 0.9495, + "step": 1440 + }, + { + "epoch": 0.1299544573206475, + "grad_norm": 1.6406725921994039, + "learning_rate": 3.896160149421119e-06, + "loss": 1.099, + "step": 1441 + }, + { + "epoch": 0.1300446408441178, + "grad_norm": 1.969177018131789, + "learning_rate": 3.8959742710201314e-06, + "loss": 1.1106, + "step": 1442 + }, + { + "epoch": 0.13013482436758805, + "grad_norm": 2.2870230589230838, + "learning_rate": 3.895788230844166e-06, + "loss": 0.9348, + "step": 1443 + }, + { + "epoch": 0.13022500789105831, + "grad_norm": 1.8351967473796964, + "learning_rate": 3.895602028909095e-06, + "loss": 1.1957, + "step": 1444 + }, + { + "epoch": 0.13031519141452858, + "grad_norm": 0.6434209334402412, + "learning_rate": 3.895415665230807e-06, + "loss": 0.8354, + "step": 1445 + }, + { + "epoch": 0.13040537493799884, + "grad_norm": 1.3457960033062044, + "learning_rate": 3.895229139825203e-06, + "loss": 1.0635, + "step": 1446 + }, + { + "epoch": 0.1304955584614691, + "grad_norm": 1.5575044501472197, + "learning_rate": 3.895042452708198e-06, + "loss": 0.9503, + "step": 1447 + }, + { + "epoch": 0.13058574198493936, + "grad_norm": 1.939386249476417, + "learning_rate": 3.894855603895723e-06, + "loss": 0.945, + "step": 1448 + }, + { + "epoch": 0.13067592550840962, + "grad_norm": 2.1740310176682605, + "learning_rate": 3.894668593403718e-06, + "loss": 1.1171, + "step": 1449 + }, + { + "epoch": 0.13076610903187988, + "grad_norm": 1.775458709116351, + "learning_rate": 3.8944814212481425e-06, + "loss": 1.0436, + "step": 1450 + }, + { + "epoch": 0.13085629255535014, + "grad_norm": 1.5979417764944641, + "learning_rate": 3.894294087444966e-06, + "loss": 1.0168, + "step": 1451 + }, + { + "epoch": 0.1309464760788204, + "grad_norm": 1.9499435147638644, + "learning_rate": 3.894106592010173e-06, + "loss": 0.9297, + "step": 1452 + }, + { + "epoch": 0.13103665960229066, + "grad_norm": 1.5864157871578473, + "learning_rate": 3.893918934959762e-06, + "loss": 0.9143, + "step": 1453 + }, + { + "epoch": 0.13112684312576092, + "grad_norm": 3.0942776933170326, + "learning_rate": 3.893731116309743e-06, + "loss": 0.9758, + "step": 1454 + }, + { + "epoch": 0.13121702664923118, + "grad_norm": 1.606166344627598, + "learning_rate": 3.893543136076145e-06, + "loss": 1.0463, + "step": 1455 + }, + { + "epoch": 0.13130721017270144, + "grad_norm": 1.8526097808515456, + "learning_rate": 3.893354994275006e-06, + "loss": 1.0638, + "step": 1456 + }, + { + "epoch": 0.1313973936961717, + "grad_norm": 1.6331964798769083, + "learning_rate": 3.893166690922378e-06, + "loss": 1.0068, + "step": 1457 + }, + { + "epoch": 0.13148757721964197, + "grad_norm": 2.016247557078689, + "learning_rate": 3.892978226034329e-06, + "loss": 1.0676, + "step": 1458 + }, + { + "epoch": 0.13157776074311223, + "grad_norm": 2.3281881848664234, + "learning_rate": 3.89278959962694e-06, + "loss": 1.0871, + "step": 1459 + }, + { + "epoch": 0.1316679442665825, + "grad_norm": 2.1534505168968114, + "learning_rate": 3.8926008117163056e-06, + "loss": 1.0144, + "step": 1460 + }, + { + "epoch": 0.13175812779005275, + "grad_norm": 2.2698020576518143, + "learning_rate": 3.892411862318535e-06, + "loss": 1.0253, + "step": 1461 + }, + { + "epoch": 0.131848311313523, + "grad_norm": 1.507304051508065, + "learning_rate": 3.892222751449749e-06, + "loss": 1.0583, + "step": 1462 + }, + { + "epoch": 0.13193849483699327, + "grad_norm": 1.9516944834515544, + "learning_rate": 3.892033479126084e-06, + "loss": 0.9873, + "step": 1463 + }, + { + "epoch": 0.13202867836046353, + "grad_norm": 2.032610628277468, + "learning_rate": 3.891844045363691e-06, + "loss": 1.0193, + "step": 1464 + }, + { + "epoch": 0.13211886188393382, + "grad_norm": 1.7549453070397047, + "learning_rate": 3.891654450178732e-06, + "loss": 0.9579, + "step": 1465 + }, + { + "epoch": 0.13220904540740408, + "grad_norm": 2.235355615508012, + "learning_rate": 3.891464693587385e-06, + "loss": 1.0807, + "step": 1466 + }, + { + "epoch": 0.13229922893087434, + "grad_norm": 1.941694518395709, + "learning_rate": 3.89127477560584e-06, + "loss": 1.0784, + "step": 1467 + }, + { + "epoch": 0.1323894124543446, + "grad_norm": 26.12438438456792, + "learning_rate": 3.891084696250304e-06, + "loss": 1.074, + "step": 1468 + }, + { + "epoch": 0.13247959597781486, + "grad_norm": 1.8346333518376496, + "learning_rate": 3.890894455536993e-06, + "loss": 1.0808, + "step": 1469 + }, + { + "epoch": 0.13256977950128512, + "grad_norm": 2.769840454574745, + "learning_rate": 3.890704053482142e-06, + "loss": 1.085, + "step": 1470 + }, + { + "epoch": 0.13265996302475538, + "grad_norm": 3.510241239033776, + "learning_rate": 3.890513490101995e-06, + "loss": 0.9804, + "step": 1471 + }, + { + "epoch": 0.13275014654822564, + "grad_norm": 0.628671946031963, + "learning_rate": 3.890322765412814e-06, + "loss": 0.8445, + "step": 1472 + }, + { + "epoch": 0.1328403300716959, + "grad_norm": 1.61515555137552, + "learning_rate": 3.890131879430871e-06, + "loss": 1.0652, + "step": 1473 + }, + { + "epoch": 0.13293051359516617, + "grad_norm": 1.5783939698900211, + "learning_rate": 3.889940832172454e-06, + "loss": 1.0651, + "step": 1474 + }, + { + "epoch": 0.13302069711863643, + "grad_norm": 10.389434700384554, + "learning_rate": 3.889749623653864e-06, + "loss": 1.0845, + "step": 1475 + }, + { + "epoch": 0.1331108806421067, + "grad_norm": 1.475036882488963, + "learning_rate": 3.889558253891416e-06, + "loss": 0.9845, + "step": 1476 + }, + { + "epoch": 0.13320106416557695, + "grad_norm": 0.6303774761701245, + "learning_rate": 3.8893667229014385e-06, + "loss": 0.8197, + "step": 1477 + }, + { + "epoch": 0.1332912476890472, + "grad_norm": 3.355076510470854, + "learning_rate": 3.8891750307002746e-06, + "loss": 0.9457, + "step": 1478 + }, + { + "epoch": 0.13338143121251747, + "grad_norm": 1.34707832660368, + "learning_rate": 3.888983177304281e-06, + "loss": 0.9716, + "step": 1479 + }, + { + "epoch": 0.13347161473598773, + "grad_norm": 0.7414727634384496, + "learning_rate": 3.888791162729826e-06, + "loss": 0.8186, + "step": 1480 + }, + { + "epoch": 0.133561798259458, + "grad_norm": 2.408072932947989, + "learning_rate": 3.888598986993295e-06, + "loss": 1.0571, + "step": 1481 + }, + { + "epoch": 0.13365198178292825, + "grad_norm": 1.6015958456894446, + "learning_rate": 3.888406650111085e-06, + "loss": 1.0659, + "step": 1482 + }, + { + "epoch": 0.1337421653063985, + "grad_norm": 1.6630273742579007, + "learning_rate": 3.888214152099607e-06, + "loss": 1.0395, + "step": 1483 + }, + { + "epoch": 0.13383234882986877, + "grad_norm": 2.211588625514392, + "learning_rate": 3.888021492975285e-06, + "loss": 1.1093, + "step": 1484 + }, + { + "epoch": 0.13392253235333904, + "grad_norm": 2.2526276397952096, + "learning_rate": 3.88782867275456e-06, + "loss": 1.0898, + "step": 1485 + }, + { + "epoch": 0.1340127158768093, + "grad_norm": 1.4779664484843555, + "learning_rate": 3.8876356914538824e-06, + "loss": 1.0013, + "step": 1486 + }, + { + "epoch": 0.13410289940027956, + "grad_norm": 5.204384774411569, + "learning_rate": 3.88744254908972e-06, + "loss": 0.9918, + "step": 1487 + }, + { + "epoch": 0.13419308292374982, + "grad_norm": 1.901240975021822, + "learning_rate": 3.887249245678552e-06, + "loss": 1.0115, + "step": 1488 + }, + { + "epoch": 0.1342832664472201, + "grad_norm": 3.2018692160112394, + "learning_rate": 3.887055781236872e-06, + "loss": 1.111, + "step": 1489 + }, + { + "epoch": 0.13437344997069037, + "grad_norm": 1.7639516817598764, + "learning_rate": 3.886862155781186e-06, + "loss": 0.9071, + "step": 1490 + }, + { + "epoch": 0.13446363349416063, + "grad_norm": 0.6715854420033787, + "learning_rate": 3.886668369328019e-06, + "loss": 0.8403, + "step": 1491 + }, + { + "epoch": 0.1345538170176309, + "grad_norm": 3.2938509265123743, + "learning_rate": 3.886474421893904e-06, + "loss": 1.0634, + "step": 1492 + }, + { + "epoch": 0.13464400054110115, + "grad_norm": 1.34486196773968, + "learning_rate": 3.886280313495388e-06, + "loss": 0.962, + "step": 1493 + }, + { + "epoch": 0.1347341840645714, + "grad_norm": 2.2924750087295203, + "learning_rate": 3.886086044149035e-06, + "loss": 1.0017, + "step": 1494 + }, + { + "epoch": 0.13482436758804167, + "grad_norm": 2.7862103408311607, + "learning_rate": 3.885891613871421e-06, + "loss": 1.0446, + "step": 1495 + }, + { + "epoch": 0.13491455111151193, + "grad_norm": 1.9400059890654624, + "learning_rate": 3.885697022679136e-06, + "loss": 1.0824, + "step": 1496 + }, + { + "epoch": 0.1350047346349822, + "grad_norm": 1.7517833478608043, + "learning_rate": 3.885502270588784e-06, + "loss": 1.0671, + "step": 1497 + }, + { + "epoch": 0.13509491815845245, + "grad_norm": 1.7841591752934527, + "learning_rate": 3.885307357616981e-06, + "loss": 1.0277, + "step": 1498 + }, + { + "epoch": 0.13518510168192271, + "grad_norm": 1.860185398480641, + "learning_rate": 3.885112283780359e-06, + "loss": 1.0316, + "step": 1499 + }, + { + "epoch": 0.13527528520539298, + "grad_norm": 1.6697758125540907, + "learning_rate": 3.8849170490955624e-06, + "loss": 0.9278, + "step": 1500 + }, + { + "epoch": 0.13536546872886324, + "grad_norm": 8.918835765316649, + "learning_rate": 3.88472165357925e-06, + "loss": 1.0297, + "step": 1501 + }, + { + "epoch": 0.1354556522523335, + "grad_norm": 1.629518169848403, + "learning_rate": 3.884526097248093e-06, + "loss": 0.9907, + "step": 1502 + }, + { + "epoch": 0.13554583577580376, + "grad_norm": 1.8971904536327782, + "learning_rate": 3.884330380118779e-06, + "loss": 1.0519, + "step": 1503 + }, + { + "epoch": 0.13563601929927402, + "grad_norm": 1.7783823185793786, + "learning_rate": 3.884134502208007e-06, + "loss": 0.9893, + "step": 1504 + }, + { + "epoch": 0.13572620282274428, + "grad_norm": 1.9761379826435534, + "learning_rate": 3.88393846353249e-06, + "loss": 1.1403, + "step": 1505 + }, + { + "epoch": 0.13581638634621454, + "grad_norm": 1.9543106752648365, + "learning_rate": 3.883742264108955e-06, + "loss": 1.0878, + "step": 1506 + }, + { + "epoch": 0.1359065698696848, + "grad_norm": 1.5372854129560494, + "learning_rate": 3.883545903954145e-06, + "loss": 1.04, + "step": 1507 + }, + { + "epoch": 0.13599675339315506, + "grad_norm": 2.043766019332586, + "learning_rate": 3.883349383084811e-06, + "loss": 1.027, + "step": 1508 + }, + { + "epoch": 0.13608693691662532, + "grad_norm": 2.5920727373855112, + "learning_rate": 3.883152701517723e-06, + "loss": 0.9566, + "step": 1509 + }, + { + "epoch": 0.13617712044009558, + "grad_norm": 1.7692539427269847, + "learning_rate": 3.882955859269664e-06, + "loss": 1.0215, + "step": 1510 + }, + { + "epoch": 0.13626730396356584, + "grad_norm": 1.8414490129557413, + "learning_rate": 3.882758856357428e-06, + "loss": 1.0407, + "step": 1511 + }, + { + "epoch": 0.1363574874870361, + "grad_norm": 1.8688449924783102, + "learning_rate": 3.882561692797824e-06, + "loss": 1.025, + "step": 1512 + }, + { + "epoch": 0.1364476710105064, + "grad_norm": 1.6659363418085626, + "learning_rate": 3.882364368607677e-06, + "loss": 1.0377, + "step": 1513 + }, + { + "epoch": 0.13653785453397665, + "grad_norm": 1.969686497369239, + "learning_rate": 3.8821668838038225e-06, + "loss": 1.0478, + "step": 1514 + }, + { + "epoch": 0.13662803805744692, + "grad_norm": 2.8851844519717305, + "learning_rate": 3.881969238403111e-06, + "loss": 0.9451, + "step": 1515 + }, + { + "epoch": 0.13671822158091718, + "grad_norm": 2.869983442836945, + "learning_rate": 3.881771432422408e-06, + "loss": 1.1577, + "step": 1516 + }, + { + "epoch": 0.13680840510438744, + "grad_norm": 1.877081288101262, + "learning_rate": 3.88157346587859e-06, + "loss": 1.0922, + "step": 1517 + }, + { + "epoch": 0.1368985886278577, + "grad_norm": 1.541836402759432, + "learning_rate": 3.881375338788549e-06, + "loss": 1.0129, + "step": 1518 + }, + { + "epoch": 0.13698877215132796, + "grad_norm": 2.3284838099130822, + "learning_rate": 3.88117705116919e-06, + "loss": 1.0509, + "step": 1519 + }, + { + "epoch": 0.13707895567479822, + "grad_norm": 0.8093253985785333, + "learning_rate": 3.880978603037432e-06, + "loss": 0.8939, + "step": 1520 + }, + { + "epoch": 0.13716913919826848, + "grad_norm": 1.7308225937303359, + "learning_rate": 3.880779994410209e-06, + "loss": 1.0392, + "step": 1521 + }, + { + "epoch": 0.13725932272173874, + "grad_norm": 2.0453915644110583, + "learning_rate": 3.880581225304466e-06, + "loss": 0.9077, + "step": 1522 + }, + { + "epoch": 0.137349506245209, + "grad_norm": 1.6966477861958982, + "learning_rate": 3.880382295737163e-06, + "loss": 1.0644, + "step": 1523 + }, + { + "epoch": 0.13743968976867926, + "grad_norm": 2.68274214335162, + "learning_rate": 3.880183205725274e-06, + "loss": 1.0037, + "step": 1524 + }, + { + "epoch": 0.13752987329214952, + "grad_norm": 2.2312857873434826, + "learning_rate": 3.879983955285788e-06, + "loss": 0.9756, + "step": 1525 + }, + { + "epoch": 0.13762005681561978, + "grad_norm": 2.5124254433876327, + "learning_rate": 3.879784544435703e-06, + "loss": 1.0054, + "step": 1526 + }, + { + "epoch": 0.13771024033909005, + "grad_norm": 1.8016901267139043, + "learning_rate": 3.879584973192037e-06, + "loss": 0.9759, + "step": 1527 + }, + { + "epoch": 0.1378004238625603, + "grad_norm": 2.251885154207847, + "learning_rate": 3.8793852415718165e-06, + "loss": 0.968, + "step": 1528 + }, + { + "epoch": 0.13789060738603057, + "grad_norm": 2.3098087631266786, + "learning_rate": 3.879185349592085e-06, + "loss": 1.0419, + "step": 1529 + }, + { + "epoch": 0.13798079090950083, + "grad_norm": 1.6283210416711889, + "learning_rate": 3.878985297269897e-06, + "loss": 1.0582, + "step": 1530 + }, + { + "epoch": 0.1380709744329711, + "grad_norm": 1.9411119203408722, + "learning_rate": 3.878785084622323e-06, + "loss": 1.0738, + "step": 1531 + }, + { + "epoch": 0.13816115795644135, + "grad_norm": 2.1095681561230415, + "learning_rate": 3.878584711666447e-06, + "loss": 1.2187, + "step": 1532 + }, + { + "epoch": 0.1382513414799116, + "grad_norm": 1.9916900612110202, + "learning_rate": 3.8783841784193635e-06, + "loss": 1.0536, + "step": 1533 + }, + { + "epoch": 0.13834152500338187, + "grad_norm": 3.1656548908174353, + "learning_rate": 3.8781834848981855e-06, + "loss": 1.0141, + "step": 1534 + }, + { + "epoch": 0.13843170852685213, + "grad_norm": 2.0317697886999113, + "learning_rate": 3.877982631120037e-06, + "loss": 0.9729, + "step": 1535 + }, + { + "epoch": 0.1385218920503224, + "grad_norm": 3.501330803861767, + "learning_rate": 3.877781617102053e-06, + "loss": 1.0983, + "step": 1536 + }, + { + "epoch": 0.13861207557379268, + "grad_norm": 2.0270898325529387, + "learning_rate": 3.877580442861389e-06, + "loss": 1.0708, + "step": 1537 + }, + { + "epoch": 0.13870225909726294, + "grad_norm": 1.847006943793429, + "learning_rate": 3.877379108415209e-06, + "loss": 1.0626, + "step": 1538 + }, + { + "epoch": 0.1387924426207332, + "grad_norm": 1.645469343601707, + "learning_rate": 3.8771776137806915e-06, + "loss": 1.0614, + "step": 1539 + }, + { + "epoch": 0.13888262614420346, + "grad_norm": 1.9707238732144585, + "learning_rate": 3.8769759589750295e-06, + "loss": 0.9803, + "step": 1540 + }, + { + "epoch": 0.13897280966767372, + "grad_norm": 1.5222212411284863, + "learning_rate": 3.876774144015429e-06, + "loss": 1.0454, + "step": 1541 + }, + { + "epoch": 0.13906299319114399, + "grad_norm": 2.059907150398221, + "learning_rate": 3.87657216891911e-06, + "loss": 1.0709, + "step": 1542 + }, + { + "epoch": 0.13915317671461425, + "grad_norm": 2.993763481221277, + "learning_rate": 3.876370033703307e-06, + "loss": 1.0356, + "step": 1543 + }, + { + "epoch": 0.1392433602380845, + "grad_norm": 1.7067840187635175, + "learning_rate": 3.876167738385265e-06, + "loss": 0.9386, + "step": 1544 + }, + { + "epoch": 0.13933354376155477, + "grad_norm": 2.0612637832023464, + "learning_rate": 3.875965282982247e-06, + "loss": 1.0795, + "step": 1545 + }, + { + "epoch": 0.13942372728502503, + "grad_norm": 2.4887580833897736, + "learning_rate": 3.875762667511528e-06, + "loss": 1.0508, + "step": 1546 + }, + { + "epoch": 0.1395139108084953, + "grad_norm": 1.5481847747363569, + "learning_rate": 3.875559891990394e-06, + "loss": 1.0392, + "step": 1547 + }, + { + "epoch": 0.13960409433196555, + "grad_norm": 1.6440072210111538, + "learning_rate": 3.875356956436149e-06, + "loss": 1.0032, + "step": 1548 + }, + { + "epoch": 0.1396942778554358, + "grad_norm": 1.9148535514590028, + "learning_rate": 3.875153860866108e-06, + "loss": 1.0191, + "step": 1549 + }, + { + "epoch": 0.13978446137890607, + "grad_norm": 1.8079908119829038, + "learning_rate": 3.8749506052976e-06, + "loss": 1.0807, + "step": 1550 + }, + { + "epoch": 0.13987464490237633, + "grad_norm": 2.1390223385793345, + "learning_rate": 3.874747189747968e-06, + "loss": 1.0651, + "step": 1551 + }, + { + "epoch": 0.1399648284258466, + "grad_norm": 1.800070665879716, + "learning_rate": 3.874543614234568e-06, + "loss": 1.0483, + "step": 1552 + }, + { + "epoch": 0.14005501194931685, + "grad_norm": 1.9360884169652182, + "learning_rate": 3.874339878774771e-06, + "loss": 1.0183, + "step": 1553 + }, + { + "epoch": 0.14014519547278712, + "grad_norm": 1.8124329127030652, + "learning_rate": 3.874135983385961e-06, + "loss": 1.0453, + "step": 1554 + }, + { + "epoch": 0.14023537899625738, + "grad_norm": 1.7932263689982313, + "learning_rate": 3.873931928085535e-06, + "loss": 1.0387, + "step": 1555 + }, + { + "epoch": 0.14032556251972764, + "grad_norm": 1.8009274106898976, + "learning_rate": 3.873727712890904e-06, + "loss": 1.1251, + "step": 1556 + }, + { + "epoch": 0.1404157460431979, + "grad_norm": 2.4742441005630673, + "learning_rate": 3.873523337819493e-06, + "loss": 1.0071, + "step": 1557 + }, + { + "epoch": 0.14050592956666816, + "grad_norm": 1.5082369192230531, + "learning_rate": 3.873318802888739e-06, + "loss": 0.9669, + "step": 1558 + }, + { + "epoch": 0.14059611309013842, + "grad_norm": 0.5695790578192882, + "learning_rate": 3.873114108116097e-06, + "loss": 0.7683, + "step": 1559 + }, + { + "epoch": 0.14068629661360868, + "grad_norm": 1.7217677586662827, + "learning_rate": 3.872909253519031e-06, + "loss": 1.1247, + "step": 1560 + }, + { + "epoch": 0.14077648013707897, + "grad_norm": 2.4564039342270108, + "learning_rate": 3.8727042391150195e-06, + "loss": 1.1098, + "step": 1561 + }, + { + "epoch": 0.14086666366054923, + "grad_norm": 1.8883612455216976, + "learning_rate": 3.872499064921556e-06, + "loss": 0.9784, + "step": 1562 + }, + { + "epoch": 0.1409568471840195, + "grad_norm": 2.4381957406130166, + "learning_rate": 3.872293730956149e-06, + "loss": 1.0553, + "step": 1563 + }, + { + "epoch": 0.14104703070748975, + "grad_norm": 2.110510061544164, + "learning_rate": 3.872088237236316e-06, + "loss": 1.0342, + "step": 1564 + }, + { + "epoch": 0.14113721423096, + "grad_norm": 2.9018965077470984, + "learning_rate": 3.871882583779592e-06, + "loss": 0.9449, + "step": 1565 + }, + { + "epoch": 0.14122739775443027, + "grad_norm": 1.988385930331603, + "learning_rate": 3.871676770603525e-06, + "loss": 1.0406, + "step": 1566 + }, + { + "epoch": 0.14131758127790053, + "grad_norm": 1.9257899781313683, + "learning_rate": 3.871470797725676e-06, + "loss": 1.0317, + "step": 1567 + }, + { + "epoch": 0.1414077648013708, + "grad_norm": 2.259025591032816, + "learning_rate": 3.8712646651636185e-06, + "loss": 1.1094, + "step": 1568 + }, + { + "epoch": 0.14149794832484106, + "grad_norm": 4.047489546726387, + "learning_rate": 3.871058372934942e-06, + "loss": 1.0395, + "step": 1569 + }, + { + "epoch": 0.14158813184831132, + "grad_norm": 1.6371677119775248, + "learning_rate": 3.8708519210572485e-06, + "loss": 0.8683, + "step": 1570 + }, + { + "epoch": 0.14167831537178158, + "grad_norm": 2.2125656549227526, + "learning_rate": 3.870645309548153e-06, + "loss": 0.9785, + "step": 1571 + }, + { + "epoch": 0.14176849889525184, + "grad_norm": 0.6008937388101684, + "learning_rate": 3.870438538425284e-06, + "loss": 0.782, + "step": 1572 + }, + { + "epoch": 0.1418586824187221, + "grad_norm": 2.214320206703114, + "learning_rate": 3.870231607706287e-06, + "loss": 0.9898, + "step": 1573 + }, + { + "epoch": 0.14194886594219236, + "grad_norm": 1.9190825357464327, + "learning_rate": 3.870024517408817e-06, + "loss": 1.0421, + "step": 1574 + }, + { + "epoch": 0.14203904946566262, + "grad_norm": 3.6378304557530834, + "learning_rate": 3.8698172675505425e-06, + "loss": 0.9604, + "step": 1575 + }, + { + "epoch": 0.14212923298913288, + "grad_norm": 0.7389862488975846, + "learning_rate": 3.86960985814915e-06, + "loss": 0.7963, + "step": 1576 + }, + { + "epoch": 0.14221941651260314, + "grad_norm": 1.9397024126986855, + "learning_rate": 3.869402289222335e-06, + "loss": 0.9684, + "step": 1577 + }, + { + "epoch": 0.1423096000360734, + "grad_norm": 5.69333893917859, + "learning_rate": 3.869194560787808e-06, + "loss": 1.0168, + "step": 1578 + }, + { + "epoch": 0.14239978355954366, + "grad_norm": 1.9081406519648032, + "learning_rate": 3.868986672863296e-06, + "loss": 0.9654, + "step": 1579 + }, + { + "epoch": 0.14248996708301392, + "grad_norm": 1.70402978784198, + "learning_rate": 3.868778625466535e-06, + "loss": 1.1397, + "step": 1580 + }, + { + "epoch": 0.14258015060648419, + "grad_norm": 1.954448465174841, + "learning_rate": 3.868570418615278e-06, + "loss": 0.9586, + "step": 1581 + }, + { + "epoch": 0.14267033412995445, + "grad_norm": 0.8872504299804471, + "learning_rate": 3.8683620523272885e-06, + "loss": 0.8044, + "step": 1582 + }, + { + "epoch": 0.1427605176534247, + "grad_norm": 1.840605948435398, + "learning_rate": 3.8681535266203464e-06, + "loss": 1.1422, + "step": 1583 + }, + { + "epoch": 0.14285070117689497, + "grad_norm": 1.8441469444097136, + "learning_rate": 3.867944841512246e-06, + "loss": 1.0632, + "step": 1584 + }, + { + "epoch": 0.14294088470036526, + "grad_norm": 2.0648503638633198, + "learning_rate": 3.867735997020791e-06, + "loss": 1.0423, + "step": 1585 + }, + { + "epoch": 0.14303106822383552, + "grad_norm": 1.9425509050687886, + "learning_rate": 3.867526993163802e-06, + "loss": 0.9397, + "step": 1586 + }, + { + "epoch": 0.14312125174730578, + "grad_norm": 6.033564464434301, + "learning_rate": 3.867317829959113e-06, + "loss": 0.9124, + "step": 1587 + }, + { + "epoch": 0.14321143527077604, + "grad_norm": 2.238442939390619, + "learning_rate": 3.8671085074245704e-06, + "loss": 0.8311, + "step": 1588 + }, + { + "epoch": 0.1433016187942463, + "grad_norm": 4.8216175062510835, + "learning_rate": 3.866899025578035e-06, + "loss": 1.1021, + "step": 1589 + }, + { + "epoch": 0.14339180231771656, + "grad_norm": 1.9885572080835812, + "learning_rate": 3.86668938443738e-06, + "loss": 1.0043, + "step": 1590 + }, + { + "epoch": 0.14348198584118682, + "grad_norm": 9.39706144461902, + "learning_rate": 3.866479584020495e-06, + "loss": 1.0483, + "step": 1591 + }, + { + "epoch": 0.14357216936465708, + "grad_norm": 1.7747919229321505, + "learning_rate": 3.866269624345279e-06, + "loss": 1.0728, + "step": 1592 + }, + { + "epoch": 0.14366235288812734, + "grad_norm": 1.8100007683135872, + "learning_rate": 3.866059505429649e-06, + "loss": 1.1184, + "step": 1593 + }, + { + "epoch": 0.1437525364115976, + "grad_norm": 0.655132227320074, + "learning_rate": 3.865849227291532e-06, + "loss": 0.8323, + "step": 1594 + }, + { + "epoch": 0.14384271993506786, + "grad_norm": 2.0918874854543534, + "learning_rate": 3.865638789948872e-06, + "loss": 1.033, + "step": 1595 + }, + { + "epoch": 0.14393290345853813, + "grad_norm": 3.648246939867548, + "learning_rate": 3.865428193419622e-06, + "loss": 0.9942, + "step": 1596 + }, + { + "epoch": 0.14402308698200839, + "grad_norm": 2.325206283934374, + "learning_rate": 3.865217437721753e-06, + "loss": 1.0244, + "step": 1597 + }, + { + "epoch": 0.14411327050547865, + "grad_norm": 5.215911824434757, + "learning_rate": 3.865006522873249e-06, + "loss": 0.9653, + "step": 1598 + }, + { + "epoch": 0.1442034540289489, + "grad_norm": 1.8157967149748688, + "learning_rate": 3.864795448892103e-06, + "loss": 1.029, + "step": 1599 + }, + { + "epoch": 0.14429363755241917, + "grad_norm": 2.2410230636805686, + "learning_rate": 3.864584215796327e-06, + "loss": 1.0825, + "step": 1600 + }, + { + "epoch": 0.14438382107588943, + "grad_norm": 1.8839413120094657, + "learning_rate": 3.8643728236039455e-06, + "loss": 1.0563, + "step": 1601 + }, + { + "epoch": 0.1444740045993597, + "grad_norm": 1.7500585137530253, + "learning_rate": 3.864161272332994e-06, + "loss": 1.0209, + "step": 1602 + }, + { + "epoch": 0.14456418812282995, + "grad_norm": 1.496696490193835, + "learning_rate": 3.863949562001524e-06, + "loss": 1.0806, + "step": 1603 + }, + { + "epoch": 0.1446543716463002, + "grad_norm": 1.7494200699245128, + "learning_rate": 3.8637376926276005e-06, + "loss": 1.0083, + "step": 1604 + }, + { + "epoch": 0.14474455516977047, + "grad_norm": 2.390268137107833, + "learning_rate": 3.8635256642293e-06, + "loss": 1.0407, + "step": 1605 + }, + { + "epoch": 0.14483473869324073, + "grad_norm": 1.8448508903809158, + "learning_rate": 3.863313476824714e-06, + "loss": 1.0193, + "step": 1606 + }, + { + "epoch": 0.144924922216711, + "grad_norm": 2.072127787961233, + "learning_rate": 3.863101130431948e-06, + "loss": 1.0173, + "step": 1607 + }, + { + "epoch": 0.14501510574018128, + "grad_norm": 1.4828984998358552, + "learning_rate": 3.862888625069121e-06, + "loss": 1.0765, + "step": 1608 + }, + { + "epoch": 0.14510528926365154, + "grad_norm": 1.3503060205962287, + "learning_rate": 3.8626759607543645e-06, + "loss": 0.9953, + "step": 1609 + }, + { + "epoch": 0.1451954727871218, + "grad_norm": 2.3854629534321563, + "learning_rate": 3.862463137505825e-06, + "loss": 0.9933, + "step": 1610 + }, + { + "epoch": 0.14528565631059207, + "grad_norm": 1.9023970140210742, + "learning_rate": 3.862250155341659e-06, + "loss": 0.994, + "step": 1611 + }, + { + "epoch": 0.14537583983406233, + "grad_norm": 1.514915693600083, + "learning_rate": 3.862037014280043e-06, + "loss": 1.163, + "step": 1612 + }, + { + "epoch": 0.1454660233575326, + "grad_norm": 1.6571384422238737, + "learning_rate": 3.861823714339162e-06, + "loss": 1.0308, + "step": 1613 + }, + { + "epoch": 0.14555620688100285, + "grad_norm": 4.314341497466641, + "learning_rate": 3.861610255537215e-06, + "loss": 1.0442, + "step": 1614 + }, + { + "epoch": 0.1456463904044731, + "grad_norm": 2.5247809551142004, + "learning_rate": 3.8613966378924165e-06, + "loss": 1.0659, + "step": 1615 + }, + { + "epoch": 0.14573657392794337, + "grad_norm": 1.647360210132037, + "learning_rate": 3.861182861422993e-06, + "loss": 0.9824, + "step": 1616 + }, + { + "epoch": 0.14582675745141363, + "grad_norm": 1.7424973011406422, + "learning_rate": 3.860968926147185e-06, + "loss": 1.1062, + "step": 1617 + }, + { + "epoch": 0.1459169409748839, + "grad_norm": 1.6871623655150898, + "learning_rate": 3.860754832083247e-06, + "loss": 0.8675, + "step": 1618 + }, + { + "epoch": 0.14600712449835415, + "grad_norm": 1.5681688850985285, + "learning_rate": 3.8605405792494475e-06, + "loss": 1.1041, + "step": 1619 + }, + { + "epoch": 0.1460973080218244, + "grad_norm": 1.781597973480922, + "learning_rate": 3.860326167664066e-06, + "loss": 1.0478, + "step": 1620 + }, + { + "epoch": 0.14618749154529467, + "grad_norm": 2.656660070742177, + "learning_rate": 3.860111597345399e-06, + "loss": 1.0225, + "step": 1621 + }, + { + "epoch": 0.14627767506876493, + "grad_norm": 2.0657896914020033, + "learning_rate": 3.859896868311753e-06, + "loss": 0.8883, + "step": 1622 + }, + { + "epoch": 0.1463678585922352, + "grad_norm": 1.7447920006898778, + "learning_rate": 3.859681980581452e-06, + "loss": 1.1223, + "step": 1623 + }, + { + "epoch": 0.14645804211570546, + "grad_norm": 1.68010411973013, + "learning_rate": 3.859466934172829e-06, + "loss": 1.0076, + "step": 1624 + }, + { + "epoch": 0.14654822563917572, + "grad_norm": 2.432428806748996, + "learning_rate": 3.859251729104235e-06, + "loss": 0.964, + "step": 1625 + }, + { + "epoch": 0.14663840916264598, + "grad_norm": 1.6097818249315665, + "learning_rate": 3.859036365394031e-06, + "loss": 0.98, + "step": 1626 + }, + { + "epoch": 0.14672859268611624, + "grad_norm": 2.161464412328299, + "learning_rate": 3.858820843060594e-06, + "loss": 1.0178, + "step": 1627 + }, + { + "epoch": 0.1468187762095865, + "grad_norm": 1.772848829801243, + "learning_rate": 3.858605162122314e-06, + "loss": 1.0103, + "step": 1628 + }, + { + "epoch": 0.14690895973305676, + "grad_norm": 1.9532736759818607, + "learning_rate": 3.858389322597592e-06, + "loss": 0.9838, + "step": 1629 + }, + { + "epoch": 0.14699914325652702, + "grad_norm": 1.7307632918948799, + "learning_rate": 3.858173324504847e-06, + "loss": 1.0445, + "step": 1630 + }, + { + "epoch": 0.14708932677999728, + "grad_norm": 1.487248500431425, + "learning_rate": 3.857957167862508e-06, + "loss": 0.9942, + "step": 1631 + }, + { + "epoch": 0.14717951030346757, + "grad_norm": 0.6696417976552665, + "learning_rate": 3.857740852689018e-06, + "loss": 0.8167, + "step": 1632 + }, + { + "epoch": 0.14726969382693783, + "grad_norm": 2.18240067263618, + "learning_rate": 3.857524379002835e-06, + "loss": 1.0384, + "step": 1633 + }, + { + "epoch": 0.1473598773504081, + "grad_norm": 1.7506927072927567, + "learning_rate": 3.85730774682243e-06, + "loss": 1.0474, + "step": 1634 + }, + { + "epoch": 0.14745006087387835, + "grad_norm": 2.228086251327195, + "learning_rate": 3.8570909561662875e-06, + "loss": 1.0917, + "step": 1635 + }, + { + "epoch": 0.1475402443973486, + "grad_norm": 1.7077942982477614, + "learning_rate": 3.8568740070529045e-06, + "loss": 1.0937, + "step": 1636 + }, + { + "epoch": 0.14763042792081887, + "grad_norm": 1.765827673698502, + "learning_rate": 3.856656899500792e-06, + "loss": 0.9981, + "step": 1637 + }, + { + "epoch": 0.14772061144428913, + "grad_norm": 3.1804621172362366, + "learning_rate": 3.856439633528476e-06, + "loss": 0.9606, + "step": 1638 + }, + { + "epoch": 0.1478107949677594, + "grad_norm": 1.8444997265330294, + "learning_rate": 3.856222209154494e-06, + "loss": 1.0854, + "step": 1639 + }, + { + "epoch": 0.14790097849122966, + "grad_norm": 1.7404346629095082, + "learning_rate": 3.856004626397397e-06, + "loss": 1.0492, + "step": 1640 + }, + { + "epoch": 0.14799116201469992, + "grad_norm": 2.1535152840090763, + "learning_rate": 3.855786885275753e-06, + "loss": 1.0038, + "step": 1641 + }, + { + "epoch": 0.14808134553817018, + "grad_norm": 1.7929206791835137, + "learning_rate": 3.855568985808138e-06, + "loss": 0.9404, + "step": 1642 + }, + { + "epoch": 0.14817152906164044, + "grad_norm": 1.68399439600558, + "learning_rate": 3.855350928013145e-06, + "loss": 0.9979, + "step": 1643 + }, + { + "epoch": 0.1482617125851107, + "grad_norm": 2.1576814184903967, + "learning_rate": 3.8551327119093825e-06, + "loss": 1.0362, + "step": 1644 + }, + { + "epoch": 0.14835189610858096, + "grad_norm": 2.70214155531966, + "learning_rate": 3.854914337515467e-06, + "loss": 0.9607, + "step": 1645 + }, + { + "epoch": 0.14844207963205122, + "grad_norm": 2.3784537049351573, + "learning_rate": 3.8546958048500324e-06, + "loss": 0.9198, + "step": 1646 + }, + { + "epoch": 0.14853226315552148, + "grad_norm": 0.6780499210194748, + "learning_rate": 3.854477113931725e-06, + "loss": 0.8688, + "step": 1647 + }, + { + "epoch": 0.14862244667899174, + "grad_norm": 2.1318673635238095, + "learning_rate": 3.854258264779205e-06, + "loss": 1.0453, + "step": 1648 + }, + { + "epoch": 0.148712630202462, + "grad_norm": 5.7545662039848775, + "learning_rate": 3.854039257411145e-06, + "loss": 1.0084, + "step": 1649 + }, + { + "epoch": 0.14880281372593226, + "grad_norm": 1.9283859867676367, + "learning_rate": 3.853820091846232e-06, + "loss": 0.9825, + "step": 1650 + }, + { + "epoch": 0.14889299724940253, + "grad_norm": 1.9308089113673512, + "learning_rate": 3.853600768103169e-06, + "loss": 1.0281, + "step": 1651 + }, + { + "epoch": 0.1489831807728728, + "grad_norm": 3.397968970262055, + "learning_rate": 3.853381286200667e-06, + "loss": 1.0183, + "step": 1652 + }, + { + "epoch": 0.14907336429634305, + "grad_norm": 2.04060510038572, + "learning_rate": 3.853161646157453e-06, + "loss": 1.1187, + "step": 1653 + }, + { + "epoch": 0.1491635478198133, + "grad_norm": 1.7804402970521611, + "learning_rate": 3.852941847992269e-06, + "loss": 0.9893, + "step": 1654 + }, + { + "epoch": 0.14925373134328357, + "grad_norm": 2.2813400877840135, + "learning_rate": 3.852721891723871e-06, + "loss": 1.0428, + "step": 1655 + }, + { + "epoch": 0.14934391486675386, + "grad_norm": 2.07492139736348, + "learning_rate": 3.852501777371025e-06, + "loss": 1.0117, + "step": 1656 + }, + { + "epoch": 0.14943409839022412, + "grad_norm": 2.819496184585785, + "learning_rate": 3.8522815049525125e-06, + "loss": 1.0634, + "step": 1657 + }, + { + "epoch": 0.14952428191369438, + "grad_norm": 1.964141356130195, + "learning_rate": 3.852061074487129e-06, + "loss": 1.0812, + "step": 1658 + }, + { + "epoch": 0.14961446543716464, + "grad_norm": 1.682258731610193, + "learning_rate": 3.851840485993682e-06, + "loss": 0.9349, + "step": 1659 + }, + { + "epoch": 0.1497046489606349, + "grad_norm": 1.6258849155384398, + "learning_rate": 3.851619739490994e-06, + "loss": 1.0377, + "step": 1660 + }, + { + "epoch": 0.14979483248410516, + "grad_norm": 3.803613656089114, + "learning_rate": 3.8513988349978996e-06, + "loss": 1.0624, + "step": 1661 + }, + { + "epoch": 0.14988501600757542, + "grad_norm": 1.5642009631898386, + "learning_rate": 3.851177772533249e-06, + "loss": 1.0518, + "step": 1662 + }, + { + "epoch": 0.14997519953104568, + "grad_norm": 1.9876637277401765, + "learning_rate": 3.850956552115903e-06, + "loss": 1.2054, + "step": 1663 + }, + { + "epoch": 0.15006538305451594, + "grad_norm": 0.7045907319949573, + "learning_rate": 3.850735173764738e-06, + "loss": 0.8204, + "step": 1664 + }, + { + "epoch": 0.1501555665779862, + "grad_norm": 1.6533208893235851, + "learning_rate": 3.850513637498642e-06, + "loss": 1.0479, + "step": 1665 + }, + { + "epoch": 0.15024575010145647, + "grad_norm": 1.6934396973317785, + "learning_rate": 3.850291943336521e-06, + "loss": 1.0449, + "step": 1666 + }, + { + "epoch": 0.15033593362492673, + "grad_norm": 2.3446879226824513, + "learning_rate": 3.850070091297287e-06, + "loss": 1.0839, + "step": 1667 + }, + { + "epoch": 0.150426117148397, + "grad_norm": 0.7211851663390557, + "learning_rate": 3.8498480813998735e-06, + "loss": 0.9115, + "step": 1668 + }, + { + "epoch": 0.15051630067186725, + "grad_norm": 1.9655671200622604, + "learning_rate": 3.84962591366322e-06, + "loss": 1.0277, + "step": 1669 + }, + { + "epoch": 0.1506064841953375, + "grad_norm": 2.018243316272644, + "learning_rate": 3.8494035881062855e-06, + "loss": 1.079, + "step": 1670 + }, + { + "epoch": 0.15069666771880777, + "grad_norm": 2.8629388331235845, + "learning_rate": 3.84918110474804e-06, + "loss": 1.0078, + "step": 1671 + }, + { + "epoch": 0.15078685124227803, + "grad_norm": 1.7000476437792515, + "learning_rate": 3.8489584636074655e-06, + "loss": 1.0133, + "step": 1672 + }, + { + "epoch": 0.1508770347657483, + "grad_norm": 0.6997600144040734, + "learning_rate": 3.848735664703561e-06, + "loss": 0.8026, + "step": 1673 + }, + { + "epoch": 0.15096721828921855, + "grad_norm": 1.6920757422346986, + "learning_rate": 3.8485127080553346e-06, + "loss": 1.0275, + "step": 1674 + }, + { + "epoch": 0.1510574018126888, + "grad_norm": 3.286922238712794, + "learning_rate": 3.8482895936818115e-06, + "loss": 1.0264, + "step": 1675 + }, + { + "epoch": 0.15114758533615907, + "grad_norm": 2.5104911022837735, + "learning_rate": 3.848066321602029e-06, + "loss": 1.0461, + "step": 1676 + }, + { + "epoch": 0.15123776885962933, + "grad_norm": 2.614621398972907, + "learning_rate": 3.847842891835038e-06, + "loss": 0.9926, + "step": 1677 + }, + { + "epoch": 0.1513279523830996, + "grad_norm": 0.629252320789639, + "learning_rate": 3.847619304399902e-06, + "loss": 0.8184, + "step": 1678 + }, + { + "epoch": 0.15141813590656986, + "grad_norm": 2.668314474334247, + "learning_rate": 3.8473955593157e-06, + "loss": 1.0575, + "step": 1679 + }, + { + "epoch": 0.15150831943004014, + "grad_norm": 2.322764119383816, + "learning_rate": 3.847171656601522e-06, + "loss": 0.9931, + "step": 1680 + }, + { + "epoch": 0.1515985029535104, + "grad_norm": 1.6647558066431103, + "learning_rate": 3.846947596276473e-06, + "loss": 1.0537, + "step": 1681 + }, + { + "epoch": 0.15168868647698067, + "grad_norm": 1.9027260272869453, + "learning_rate": 3.846723378359672e-06, + "loss": 1.0786, + "step": 1682 + }, + { + "epoch": 0.15177887000045093, + "grad_norm": 2.7733271187321407, + "learning_rate": 3.846499002870249e-06, + "loss": 1.0347, + "step": 1683 + }, + { + "epoch": 0.1518690535239212, + "grad_norm": 9.479758130299759, + "learning_rate": 3.846274469827349e-06, + "loss": 1.048, + "step": 1684 + }, + { + "epoch": 0.15195923704739145, + "grad_norm": 2.0522636941050245, + "learning_rate": 3.846049779250132e-06, + "loss": 1.0402, + "step": 1685 + }, + { + "epoch": 0.1520494205708617, + "grad_norm": 0.8234026516049572, + "learning_rate": 3.845824931157769e-06, + "loss": 0.8616, + "step": 1686 + }, + { + "epoch": 0.15213960409433197, + "grad_norm": 1.4755626704403215, + "learning_rate": 3.845599925569444e-06, + "loss": 1.0135, + "step": 1687 + }, + { + "epoch": 0.15222978761780223, + "grad_norm": 2.1188039767161206, + "learning_rate": 3.845374762504357e-06, + "loss": 0.9864, + "step": 1688 + }, + { + "epoch": 0.1523199711412725, + "grad_norm": 1.7021796726462135, + "learning_rate": 3.8451494419817204e-06, + "loss": 1.0762, + "step": 1689 + }, + { + "epoch": 0.15241015466474275, + "grad_norm": 2.4952554503949753, + "learning_rate": 3.8449239640207594e-06, + "loss": 0.9575, + "step": 1690 + }, + { + "epoch": 0.152500338188213, + "grad_norm": 2.1576663907669356, + "learning_rate": 3.844698328640713e-06, + "loss": 1.0712, + "step": 1691 + }, + { + "epoch": 0.15259052171168327, + "grad_norm": 1.7938082190446214, + "learning_rate": 3.844472535860833e-06, + "loss": 1.0508, + "step": 1692 + }, + { + "epoch": 0.15268070523515354, + "grad_norm": 2.1843545369730117, + "learning_rate": 3.8442465857003864e-06, + "loss": 1.0795, + "step": 1693 + }, + { + "epoch": 0.1527708887586238, + "grad_norm": 1.5532723892246214, + "learning_rate": 3.844020478178653e-06, + "loss": 0.9771, + "step": 1694 + }, + { + "epoch": 0.15286107228209406, + "grad_norm": 1.8302485855820307, + "learning_rate": 3.843794213314923e-06, + "loss": 1.0568, + "step": 1695 + }, + { + "epoch": 0.15295125580556432, + "grad_norm": 1.769579619363412, + "learning_rate": 3.843567791128505e-06, + "loss": 1.0602, + "step": 1696 + }, + { + "epoch": 0.15304143932903458, + "grad_norm": 1.7127116859080576, + "learning_rate": 3.843341211638717e-06, + "loss": 1.0181, + "step": 1697 + }, + { + "epoch": 0.15313162285250484, + "grad_norm": 2.4183666783072972, + "learning_rate": 3.843114474864894e-06, + "loss": 0.9897, + "step": 1698 + }, + { + "epoch": 0.1532218063759751, + "grad_norm": 2.5616877129045306, + "learning_rate": 3.84288758082638e-06, + "loss": 1.0311, + "step": 1699 + }, + { + "epoch": 0.15331198989944536, + "grad_norm": 1.4478462074023286, + "learning_rate": 3.842660529542536e-06, + "loss": 1.1187, + "step": 1700 + }, + { + "epoch": 0.15340217342291562, + "grad_norm": 1.7912598820884265, + "learning_rate": 3.842433321032736e-06, + "loss": 1.0457, + "step": 1701 + }, + { + "epoch": 0.15349235694638588, + "grad_norm": 2.3459391225085344, + "learning_rate": 3.842205955316365e-06, + "loss": 0.9569, + "step": 1702 + }, + { + "epoch": 0.15358254046985614, + "grad_norm": 1.7271383714668362, + "learning_rate": 3.8419784324128256e-06, + "loss": 0.9351, + "step": 1703 + }, + { + "epoch": 0.15367272399332643, + "grad_norm": 2.0222032236364242, + "learning_rate": 3.841750752341529e-06, + "loss": 0.9974, + "step": 1704 + }, + { + "epoch": 0.1537629075167967, + "grad_norm": 1.7977688100523306, + "learning_rate": 3.841522915121902e-06, + "loss": 1.0352, + "step": 1705 + }, + { + "epoch": 0.15385309104026695, + "grad_norm": 1.7454121259841386, + "learning_rate": 3.841294920773387e-06, + "loss": 0.9586, + "step": 1706 + }, + { + "epoch": 0.15394327456373721, + "grad_norm": 2.1117913993280975, + "learning_rate": 3.841066769315436e-06, + "loss": 1.0414, + "step": 1707 + }, + { + "epoch": 0.15403345808720748, + "grad_norm": 1.7778778950328975, + "learning_rate": 3.840838460767518e-06, + "loss": 1.0374, + "step": 1708 + }, + { + "epoch": 0.15412364161067774, + "grad_norm": 1.9290208784785932, + "learning_rate": 3.840609995149111e-06, + "loss": 1.0014, + "step": 1709 + }, + { + "epoch": 0.154213825134148, + "grad_norm": 1.858783852346545, + "learning_rate": 3.84038137247971e-06, + "loss": 1.0322, + "step": 1710 + }, + { + "epoch": 0.15430400865761826, + "grad_norm": 2.0013130169488935, + "learning_rate": 3.840152592778823e-06, + "loss": 1.0171, + "step": 1711 + }, + { + "epoch": 0.15439419218108852, + "grad_norm": 1.4331194433411496, + "learning_rate": 3.83992365606597e-06, + "loss": 1.0736, + "step": 1712 + }, + { + "epoch": 0.15448437570455878, + "grad_norm": 1.626493941006426, + "learning_rate": 3.8396945623606855e-06, + "loss": 0.9865, + "step": 1713 + }, + { + "epoch": 0.15457455922802904, + "grad_norm": 1.7905912572534157, + "learning_rate": 3.8394653116825174e-06, + "loss": 0.9875, + "step": 1714 + }, + { + "epoch": 0.1546647427514993, + "grad_norm": 1.7854820304784642, + "learning_rate": 3.839235904051026e-06, + "loss": 1.0223, + "step": 1715 + }, + { + "epoch": 0.15475492627496956, + "grad_norm": 1.8117613767253808, + "learning_rate": 3.8390063394857855e-06, + "loss": 1.0018, + "step": 1716 + }, + { + "epoch": 0.15484510979843982, + "grad_norm": 1.5523849382009443, + "learning_rate": 3.838776618006385e-06, + "loss": 1.1125, + "step": 1717 + }, + { + "epoch": 0.15493529332191008, + "grad_norm": 1.7477118655936164, + "learning_rate": 3.838546739632423e-06, + "loss": 1.0377, + "step": 1718 + }, + { + "epoch": 0.15502547684538034, + "grad_norm": 1.8105031556547317, + "learning_rate": 3.838316704383517e-06, + "loss": 1.0755, + "step": 1719 + }, + { + "epoch": 0.1551156603688506, + "grad_norm": 1.705939880246446, + "learning_rate": 3.838086512279292e-06, + "loss": 1.0332, + "step": 1720 + }, + { + "epoch": 0.15520584389232087, + "grad_norm": 2.24697991180608, + "learning_rate": 3.837856163339391e-06, + "loss": 1.0277, + "step": 1721 + }, + { + "epoch": 0.15529602741579113, + "grad_norm": 2.549641610186237, + "learning_rate": 3.837625657583469e-06, + "loss": 0.9597, + "step": 1722 + }, + { + "epoch": 0.1553862109392614, + "grad_norm": 1.9412311190972251, + "learning_rate": 3.837394995031193e-06, + "loss": 0.9882, + "step": 1723 + }, + { + "epoch": 0.15547639446273165, + "grad_norm": 1.6278725690152005, + "learning_rate": 3.837164175702245e-06, + "loss": 1.0287, + "step": 1724 + }, + { + "epoch": 0.1555665779862019, + "grad_norm": 2.012645441332345, + "learning_rate": 3.836933199616319e-06, + "loss": 0.9391, + "step": 1725 + }, + { + "epoch": 0.15565676150967217, + "grad_norm": 2.384710337804815, + "learning_rate": 3.836702066793124e-06, + "loss": 1.0983, + "step": 1726 + }, + { + "epoch": 0.15574694503314246, + "grad_norm": 1.5657688280386943, + "learning_rate": 3.836470777252381e-06, + "loss": 1.0913, + "step": 1727 + }, + { + "epoch": 0.15583712855661272, + "grad_norm": 2.040135127061421, + "learning_rate": 3.836239331013825e-06, + "loss": 1.0425, + "step": 1728 + }, + { + "epoch": 0.15592731208008298, + "grad_norm": 1.9163720346748472, + "learning_rate": 3.836007728097205e-06, + "loss": 0.9904, + "step": 1729 + }, + { + "epoch": 0.15601749560355324, + "grad_norm": 1.5533789106703348, + "learning_rate": 3.835775968522282e-06, + "loss": 0.9816, + "step": 1730 + }, + { + "epoch": 0.1561076791270235, + "grad_norm": 1.9991506322694996, + "learning_rate": 3.83554405230883e-06, + "loss": 1.0488, + "step": 1731 + }, + { + "epoch": 0.15619786265049376, + "grad_norm": 4.348902658193712, + "learning_rate": 3.835311979476639e-06, + "loss": 1.014, + "step": 1732 + }, + { + "epoch": 0.15628804617396402, + "grad_norm": 1.6896023546583574, + "learning_rate": 3.83507975004551e-06, + "loss": 0.9231, + "step": 1733 + }, + { + "epoch": 0.15637822969743428, + "grad_norm": 1.6209344624773478, + "learning_rate": 3.834847364035258e-06, + "loss": 1.0112, + "step": 1734 + }, + { + "epoch": 0.15646841322090455, + "grad_norm": 2.4704453646731808, + "learning_rate": 3.834614821465712e-06, + "loss": 0.9743, + "step": 1735 + }, + { + "epoch": 0.1565585967443748, + "grad_norm": 1.8917991838238286, + "learning_rate": 3.834382122356713e-06, + "loss": 1.0612, + "step": 1736 + }, + { + "epoch": 0.15664878026784507, + "grad_norm": 1.6712432095397542, + "learning_rate": 3.834149266728117e-06, + "loss": 0.9879, + "step": 1737 + }, + { + "epoch": 0.15673896379131533, + "grad_norm": 0.6887144935649816, + "learning_rate": 3.833916254599792e-06, + "loss": 0.8575, + "step": 1738 + }, + { + "epoch": 0.1568291473147856, + "grad_norm": 0.6405862470861375, + "learning_rate": 3.83368308599162e-06, + "loss": 0.8037, + "step": 1739 + }, + { + "epoch": 0.15691933083825585, + "grad_norm": 0.6461687729440161, + "learning_rate": 3.833449760923498e-06, + "loss": 0.7759, + "step": 1740 + }, + { + "epoch": 0.1570095143617261, + "grad_norm": 1.6156682788947605, + "learning_rate": 3.83321627941533e-06, + "loss": 1.0645, + "step": 1741 + }, + { + "epoch": 0.15709969788519637, + "grad_norm": 2.2440843168387286, + "learning_rate": 3.832982641487043e-06, + "loss": 1.0833, + "step": 1742 + }, + { + "epoch": 0.15718988140866663, + "grad_norm": 2.0369622533088405, + "learning_rate": 3.832748847158568e-06, + "loss": 1.1183, + "step": 1743 + }, + { + "epoch": 0.1572800649321369, + "grad_norm": 2.1762990557966133, + "learning_rate": 3.832514896449858e-06, + "loss": 1.0508, + "step": 1744 + }, + { + "epoch": 0.15737024845560715, + "grad_norm": 1.7474190889981758, + "learning_rate": 3.832280789380871e-06, + "loss": 0.9669, + "step": 1745 + }, + { + "epoch": 0.15746043197907741, + "grad_norm": 2.0054876620118334, + "learning_rate": 3.832046525971584e-06, + "loss": 0.9293, + "step": 1746 + }, + { + "epoch": 0.15755061550254768, + "grad_norm": 2.0816235456233096, + "learning_rate": 3.831812106241987e-06, + "loss": 1.0864, + "step": 1747 + }, + { + "epoch": 0.15764079902601794, + "grad_norm": 1.848252633686168, + "learning_rate": 3.8315775302120796e-06, + "loss": 0.9338, + "step": 1748 + }, + { + "epoch": 0.1577309825494882, + "grad_norm": 1.8666973029188443, + "learning_rate": 3.831342797901878e-06, + "loss": 1.0399, + "step": 1749 + }, + { + "epoch": 0.15782116607295846, + "grad_norm": 0.6651887879176828, + "learning_rate": 3.831107909331411e-06, + "loss": 0.8168, + "step": 1750 + }, + { + "epoch": 0.15791134959642875, + "grad_norm": 1.9120404832097524, + "learning_rate": 3.830872864520721e-06, + "loss": 1.0694, + "step": 1751 + }, + { + "epoch": 0.158001533119899, + "grad_norm": 0.7149638924451396, + "learning_rate": 3.830637663489862e-06, + "loss": 0.8967, + "step": 1752 + }, + { + "epoch": 0.15809171664336927, + "grad_norm": 0.7808408047031281, + "learning_rate": 3.830402306258904e-06, + "loss": 0.9268, + "step": 1753 + }, + { + "epoch": 0.15818190016683953, + "grad_norm": 2.7508194309435314, + "learning_rate": 3.830166792847929e-06, + "loss": 1.0094, + "step": 1754 + }, + { + "epoch": 0.1582720836903098, + "grad_norm": 2.126023775206411, + "learning_rate": 3.829931123277031e-06, + "loss": 0.9673, + "step": 1755 + }, + { + "epoch": 0.15836226721378005, + "grad_norm": 2.537359705462027, + "learning_rate": 3.8296952975663204e-06, + "loss": 1.0154, + "step": 1756 + }, + { + "epoch": 0.1584524507372503, + "grad_norm": 1.574362774264246, + "learning_rate": 3.829459315735918e-06, + "loss": 1.0935, + "step": 1757 + }, + { + "epoch": 0.15854263426072057, + "grad_norm": 1.9075679912923675, + "learning_rate": 3.829223177805959e-06, + "loss": 0.9504, + "step": 1758 + }, + { + "epoch": 0.15863281778419083, + "grad_norm": 1.6441014106408003, + "learning_rate": 3.828986883796591e-06, + "loss": 1.0624, + "step": 1759 + }, + { + "epoch": 0.1587230013076611, + "grad_norm": 2.156070204163297, + "learning_rate": 3.828750433727979e-06, + "loss": 1.0151, + "step": 1760 + }, + { + "epoch": 0.15881318483113135, + "grad_norm": 2.218685202928438, + "learning_rate": 3.828513827620296e-06, + "loss": 1.0707, + "step": 1761 + }, + { + "epoch": 0.15890336835460162, + "grad_norm": 2.035428488419509, + "learning_rate": 3.82827706549373e-06, + "loss": 1.0774, + "step": 1762 + }, + { + "epoch": 0.15899355187807188, + "grad_norm": 2.824374414905235, + "learning_rate": 3.828040147368484e-06, + "loss": 1.0119, + "step": 1763 + }, + { + "epoch": 0.15908373540154214, + "grad_norm": 1.6049902729678491, + "learning_rate": 3.827803073264774e-06, + "loss": 1.0237, + "step": 1764 + }, + { + "epoch": 0.1591739189250124, + "grad_norm": 1.7834758570870815, + "learning_rate": 3.827565843202826e-06, + "loss": 1.0214, + "step": 1765 + }, + { + "epoch": 0.15926410244848266, + "grad_norm": 2.4723131565510013, + "learning_rate": 3.827328457202884e-06, + "loss": 0.8984, + "step": 1766 + }, + { + "epoch": 0.15935428597195292, + "grad_norm": 1.4538977424417767, + "learning_rate": 3.8270909152852014e-06, + "loss": 0.9569, + "step": 1767 + }, + { + "epoch": 0.15944446949542318, + "grad_norm": 1.820763462509758, + "learning_rate": 3.826853217470048e-06, + "loss": 1.0157, + "step": 1768 + }, + { + "epoch": 0.15953465301889344, + "grad_norm": 1.8746206535608958, + "learning_rate": 3.826615363777705e-06, + "loss": 1.1047, + "step": 1769 + }, + { + "epoch": 0.1596248365423637, + "grad_norm": 1.772397042796346, + "learning_rate": 3.826377354228468e-06, + "loss": 1.0009, + "step": 1770 + }, + { + "epoch": 0.15971502006583396, + "grad_norm": 0.6446419360747133, + "learning_rate": 3.826139188842643e-06, + "loss": 0.8292, + "step": 1771 + }, + { + "epoch": 0.15980520358930422, + "grad_norm": 1.604045970719365, + "learning_rate": 3.825900867640554e-06, + "loss": 1.0801, + "step": 1772 + }, + { + "epoch": 0.15989538711277448, + "grad_norm": 1.4389767732051315, + "learning_rate": 3.825662390642535e-06, + "loss": 1.0624, + "step": 1773 + }, + { + "epoch": 0.15998557063624474, + "grad_norm": 1.697340354547315, + "learning_rate": 3.825423757868934e-06, + "loss": 1.1319, + "step": 1774 + }, + { + "epoch": 0.16007575415971503, + "grad_norm": 1.650892842642115, + "learning_rate": 3.825184969340114e-06, + "loss": 1.0487, + "step": 1775 + }, + { + "epoch": 0.1601659376831853, + "grad_norm": 2.5620033899444143, + "learning_rate": 3.824946025076447e-06, + "loss": 1.052, + "step": 1776 + }, + { + "epoch": 0.16025612120665556, + "grad_norm": 1.9212533208933753, + "learning_rate": 3.824706925098323e-06, + "loss": 0.9739, + "step": 1777 + }, + { + "epoch": 0.16034630473012582, + "grad_norm": 2.6539764547733755, + "learning_rate": 3.824467669426143e-06, + "loss": 1.0872, + "step": 1778 + }, + { + "epoch": 0.16043648825359608, + "grad_norm": 2.269918648358574, + "learning_rate": 3.824228258080321e-06, + "loss": 0.9922, + "step": 1779 + }, + { + "epoch": 0.16052667177706634, + "grad_norm": 3.1531049714525103, + "learning_rate": 3.823988691081285e-06, + "loss": 0.9373, + "step": 1780 + }, + { + "epoch": 0.1606168553005366, + "grad_norm": 1.5328779325216875, + "learning_rate": 3.823748968449478e-06, + "loss": 1.0561, + "step": 1781 + }, + { + "epoch": 0.16070703882400686, + "grad_norm": 4.666511442236839, + "learning_rate": 3.823509090205352e-06, + "loss": 0.9165, + "step": 1782 + }, + { + "epoch": 0.16079722234747712, + "grad_norm": 1.8315241585652653, + "learning_rate": 3.823269056369376e-06, + "loss": 0.9774, + "step": 1783 + }, + { + "epoch": 0.16088740587094738, + "grad_norm": 1.6735756326435027, + "learning_rate": 3.8230288669620295e-06, + "loss": 1.0533, + "step": 1784 + }, + { + "epoch": 0.16097758939441764, + "grad_norm": 1.892487042482229, + "learning_rate": 3.822788522003809e-06, + "loss": 1.0256, + "step": 1785 + }, + { + "epoch": 0.1610677729178879, + "grad_norm": 2.2241706716997607, + "learning_rate": 3.822548021515221e-06, + "loss": 1.0939, + "step": 1786 + }, + { + "epoch": 0.16115795644135816, + "grad_norm": 2.8180670847708362, + "learning_rate": 3.822307365516787e-06, + "loss": 0.999, + "step": 1787 + }, + { + "epoch": 0.16124813996482842, + "grad_norm": 1.9487127285586903, + "learning_rate": 3.8220665540290395e-06, + "loss": 1.041, + "step": 1788 + }, + { + "epoch": 0.16133832348829868, + "grad_norm": 1.4964393634842692, + "learning_rate": 3.8218255870725265e-06, + "loss": 1.0878, + "step": 1789 + }, + { + "epoch": 0.16142850701176895, + "grad_norm": 1.7408042171679634, + "learning_rate": 3.82158446466781e-06, + "loss": 0.9718, + "step": 1790 + }, + { + "epoch": 0.1615186905352392, + "grad_norm": 1.6681509878764786, + "learning_rate": 3.821343186835462e-06, + "loss": 1.0356, + "step": 1791 + }, + { + "epoch": 0.16160887405870947, + "grad_norm": 1.5038019317513078, + "learning_rate": 3.821101753596072e-06, + "loss": 1.0957, + "step": 1792 + }, + { + "epoch": 0.16169905758217973, + "grad_norm": 2.5863885860638454, + "learning_rate": 3.820860164970237e-06, + "loss": 1.0095, + "step": 1793 + }, + { + "epoch": 0.16178924110565, + "grad_norm": 1.577444930891252, + "learning_rate": 3.820618420978574e-06, + "loss": 1.0924, + "step": 1794 + }, + { + "epoch": 0.16187942462912025, + "grad_norm": 2.2794921582140026, + "learning_rate": 3.820376521641708e-06, + "loss": 1.0204, + "step": 1795 + }, + { + "epoch": 0.1619696081525905, + "grad_norm": 1.560302409669081, + "learning_rate": 3.82013446698028e-06, + "loss": 1.0323, + "step": 1796 + }, + { + "epoch": 0.16205979167606077, + "grad_norm": 1.7905543741142143, + "learning_rate": 3.819892257014943e-06, + "loss": 0.9917, + "step": 1797 + }, + { + "epoch": 0.16214997519953103, + "grad_norm": 1.815574406298097, + "learning_rate": 3.819649891766364e-06, + "loss": 1.0509, + "step": 1798 + }, + { + "epoch": 0.16224015872300132, + "grad_norm": 2.0933265827256387, + "learning_rate": 3.819407371255222e-06, + "loss": 1.0162, + "step": 1799 + }, + { + "epoch": 0.16233034224647158, + "grad_norm": 1.7030487043457623, + "learning_rate": 3.819164695502212e-06, + "loss": 1.0336, + "step": 1800 + }, + { + "epoch": 0.16242052576994184, + "grad_norm": 1.6518768300144828, + "learning_rate": 3.818921864528039e-06, + "loss": 0.9684, + "step": 1801 + }, + { + "epoch": 0.1625107092934121, + "grad_norm": 1.7517270013500708, + "learning_rate": 3.818678878353423e-06, + "loss": 1.076, + "step": 1802 + }, + { + "epoch": 0.16260089281688236, + "grad_norm": 1.9322906727201334, + "learning_rate": 3.818435736999097e-06, + "loss": 0.9719, + "step": 1803 + }, + { + "epoch": 0.16269107634035262, + "grad_norm": 1.8030173760463861, + "learning_rate": 3.818192440485807e-06, + "loss": 1.0174, + "step": 1804 + }, + { + "epoch": 0.16278125986382289, + "grad_norm": 2.275242354510621, + "learning_rate": 3.817948988834314e-06, + "loss": 0.959, + "step": 1805 + }, + { + "epoch": 0.16287144338729315, + "grad_norm": 1.6669233919148592, + "learning_rate": 3.817705382065388e-06, + "loss": 1.0202, + "step": 1806 + }, + { + "epoch": 0.1629616269107634, + "grad_norm": 1.6031652850764995, + "learning_rate": 3.8174616201998155e-06, + "loss": 0.994, + "step": 1807 + }, + { + "epoch": 0.16305181043423367, + "grad_norm": 1.8745670454520684, + "learning_rate": 3.817217703258397e-06, + "loss": 1.0731, + "step": 1808 + }, + { + "epoch": 0.16314199395770393, + "grad_norm": 1.5629720356319496, + "learning_rate": 3.816973631261943e-06, + "loss": 1.0161, + "step": 1809 + }, + { + "epoch": 0.1632321774811742, + "grad_norm": 1.6227534978062237, + "learning_rate": 3.816729404231281e-06, + "loss": 1.053, + "step": 1810 + }, + { + "epoch": 0.16332236100464445, + "grad_norm": 2.0611777257422386, + "learning_rate": 3.816485022187249e-06, + "loss": 1.0907, + "step": 1811 + }, + { + "epoch": 0.1634125445281147, + "grad_norm": 1.8332306081718244, + "learning_rate": 3.816240485150698e-06, + "loss": 0.9761, + "step": 1812 + }, + { + "epoch": 0.16350272805158497, + "grad_norm": 1.6887431334180174, + "learning_rate": 3.815995793142495e-06, + "loss": 1.0194, + "step": 1813 + }, + { + "epoch": 0.16359291157505523, + "grad_norm": 1.7126103414363842, + "learning_rate": 3.815750946183518e-06, + "loss": 0.9658, + "step": 1814 + }, + { + "epoch": 0.1636830950985255, + "grad_norm": 1.707975497411749, + "learning_rate": 3.815505944294658e-06, + "loss": 1.0765, + "step": 1815 + }, + { + "epoch": 0.16377327862199575, + "grad_norm": 1.3256659687479388, + "learning_rate": 3.81526078749682e-06, + "loss": 0.9835, + "step": 1816 + }, + { + "epoch": 0.16386346214546602, + "grad_norm": 1.8557680551153468, + "learning_rate": 3.8150154758109225e-06, + "loss": 0.9656, + "step": 1817 + }, + { + "epoch": 0.16395364566893628, + "grad_norm": 1.5074835700156695, + "learning_rate": 3.814770009257896e-06, + "loss": 0.9971, + "step": 1818 + }, + { + "epoch": 0.16404382919240654, + "grad_norm": 1.7838897553024053, + "learning_rate": 3.814524387858687e-06, + "loss": 1.0361, + "step": 1819 + }, + { + "epoch": 0.1641340127158768, + "grad_norm": 1.7456469209457628, + "learning_rate": 3.814278611634251e-06, + "loss": 1.0438, + "step": 1820 + }, + { + "epoch": 0.16422419623934706, + "grad_norm": 1.6825256499876045, + "learning_rate": 3.8140326806055606e-06, + "loss": 0.9939, + "step": 1821 + }, + { + "epoch": 0.16431437976281732, + "grad_norm": 1.69325194408531, + "learning_rate": 3.8137865947935992e-06, + "loss": 1.0872, + "step": 1822 + }, + { + "epoch": 0.1644045632862876, + "grad_norm": 2.2373783227665793, + "learning_rate": 3.8135403542193646e-06, + "loss": 1.0428, + "step": 1823 + }, + { + "epoch": 0.16449474680975787, + "grad_norm": 2.2473859436831276, + "learning_rate": 3.813293958903867e-06, + "loss": 0.9572, + "step": 1824 + }, + { + "epoch": 0.16458493033322813, + "grad_norm": 1.7704411539868319, + "learning_rate": 3.8130474088681306e-06, + "loss": 0.9872, + "step": 1825 + }, + { + "epoch": 0.1646751138566984, + "grad_norm": 2.04264826111598, + "learning_rate": 3.8128007041331927e-06, + "loss": 0.9569, + "step": 1826 + }, + { + "epoch": 0.16476529738016865, + "grad_norm": 1.6125949388101233, + "learning_rate": 3.812553844720102e-06, + "loss": 0.9405, + "step": 1827 + }, + { + "epoch": 0.1648554809036389, + "grad_norm": 1.7625816921483617, + "learning_rate": 3.8123068306499236e-06, + "loss": 1.043, + "step": 1828 + }, + { + "epoch": 0.16494566442710917, + "grad_norm": 1.7805717247665043, + "learning_rate": 3.812059661943733e-06, + "loss": 1.0186, + "step": 1829 + }, + { + "epoch": 0.16503584795057943, + "grad_norm": 1.6617746161721882, + "learning_rate": 3.811812338622621e-06, + "loss": 0.991, + "step": 1830 + }, + { + "epoch": 0.1651260314740497, + "grad_norm": 1.5184129851239234, + "learning_rate": 3.81156486070769e-06, + "loss": 1.014, + "step": 1831 + }, + { + "epoch": 0.16521621499751996, + "grad_norm": 1.8869856809468022, + "learning_rate": 3.811317228220056e-06, + "loss": 1.065, + "step": 1832 + }, + { + "epoch": 0.16530639852099022, + "grad_norm": 2.507708776117484, + "learning_rate": 3.811069441180849e-06, + "loss": 1.062, + "step": 1833 + }, + { + "epoch": 0.16539658204446048, + "grad_norm": 1.7165287576606034, + "learning_rate": 3.8108214996112107e-06, + "loss": 1.0233, + "step": 1834 + }, + { + "epoch": 0.16548676556793074, + "grad_norm": 1.9455058296682937, + "learning_rate": 3.810573403532297e-06, + "loss": 1.0604, + "step": 1835 + }, + { + "epoch": 0.165576949091401, + "grad_norm": 1.9194126016404873, + "learning_rate": 3.8103251529652774e-06, + "loss": 1.003, + "step": 1836 + }, + { + "epoch": 0.16566713261487126, + "grad_norm": 1.876674031954056, + "learning_rate": 3.810076747931334e-06, + "loss": 0.9981, + "step": 1837 + }, + { + "epoch": 0.16575731613834152, + "grad_norm": 1.7689403364126903, + "learning_rate": 3.809828188451662e-06, + "loss": 0.9826, + "step": 1838 + }, + { + "epoch": 0.16584749966181178, + "grad_norm": 1.5443369533370963, + "learning_rate": 3.809579474547469e-06, + "loss": 1.0186, + "step": 1839 + }, + { + "epoch": 0.16593768318528204, + "grad_norm": 1.792246488150818, + "learning_rate": 3.809330606239977e-06, + "loss": 1.0711, + "step": 1840 + }, + { + "epoch": 0.1660278667087523, + "grad_norm": 1.4733632462400756, + "learning_rate": 3.809081583550422e-06, + "loss": 1.0873, + "step": 1841 + }, + { + "epoch": 0.16611805023222256, + "grad_norm": 2.751864668085758, + "learning_rate": 3.808832406500051e-06, + "loss": 1.1032, + "step": 1842 + }, + { + "epoch": 0.16620823375569282, + "grad_norm": 1.884653671181162, + "learning_rate": 3.8085830751101253e-06, + "loss": 1.0926, + "step": 1843 + }, + { + "epoch": 0.16629841727916309, + "grad_norm": 1.8061362392411924, + "learning_rate": 3.808333589401919e-06, + "loss": 1.0606, + "step": 1844 + }, + { + "epoch": 0.16638860080263335, + "grad_norm": 1.7152811591783863, + "learning_rate": 3.8080839493967194e-06, + "loss": 0.9564, + "step": 1845 + }, + { + "epoch": 0.1664787843261036, + "grad_norm": 0.6657612334337568, + "learning_rate": 3.807834155115828e-06, + "loss": 0.7833, + "step": 1846 + }, + { + "epoch": 0.1665689678495739, + "grad_norm": 0.6697880691937963, + "learning_rate": 3.8075842065805584e-06, + "loss": 0.8877, + "step": 1847 + }, + { + "epoch": 0.16665915137304416, + "grad_norm": 1.5652496558010647, + "learning_rate": 3.8073341038122374e-06, + "loss": 1.0093, + "step": 1848 + }, + { + "epoch": 0.16674933489651442, + "grad_norm": 3.3743531348955598, + "learning_rate": 3.8070838468322048e-06, + "loss": 1.092, + "step": 1849 + }, + { + "epoch": 0.16683951841998468, + "grad_norm": 2.3306086049074817, + "learning_rate": 3.8068334356618143e-06, + "loss": 0.9586, + "step": 1850 + }, + { + "epoch": 0.16692970194345494, + "grad_norm": 2.9499517275206637, + "learning_rate": 3.8065828703224324e-06, + "loss": 1.0633, + "step": 1851 + }, + { + "epoch": 0.1670198854669252, + "grad_norm": 1.7819008558331524, + "learning_rate": 3.8063321508354386e-06, + "loss": 1.0246, + "step": 1852 + }, + { + "epoch": 0.16711006899039546, + "grad_norm": 1.7897313629258524, + "learning_rate": 3.8060812772222255e-06, + "loss": 1.0019, + "step": 1853 + }, + { + "epoch": 0.16720025251386572, + "grad_norm": 1.904823770256357, + "learning_rate": 3.8058302495041993e-06, + "loss": 1.088, + "step": 1854 + }, + { + "epoch": 0.16729043603733598, + "grad_norm": 2.767598627912322, + "learning_rate": 3.805579067702779e-06, + "loss": 1.0351, + "step": 1855 + }, + { + "epoch": 0.16738061956080624, + "grad_norm": 2.4492406524700994, + "learning_rate": 3.8053277318393967e-06, + "loss": 1.0272, + "step": 1856 + }, + { + "epoch": 0.1674708030842765, + "grad_norm": 1.6171172490732904, + "learning_rate": 3.805076241935498e-06, + "loss": 1.0407, + "step": 1857 + }, + { + "epoch": 0.16756098660774676, + "grad_norm": 1.8489532757183982, + "learning_rate": 3.804824598012541e-06, + "loss": 0.8976, + "step": 1858 + }, + { + "epoch": 0.16765117013121703, + "grad_norm": 1.8485972808631346, + "learning_rate": 3.8045728000919975e-06, + "loss": 1.0622, + "step": 1859 + }, + { + "epoch": 0.1677413536546873, + "grad_norm": 2.995974223750505, + "learning_rate": 3.8043208481953524e-06, + "loss": 1.0569, + "step": 1860 + }, + { + "epoch": 0.16783153717815755, + "grad_norm": 1.9183760635871148, + "learning_rate": 3.804068742344104e-06, + "loss": 0.9503, + "step": 1861 + }, + { + "epoch": 0.1679217207016278, + "grad_norm": 2.588428231305277, + "learning_rate": 3.8038164825597628e-06, + "loss": 0.9904, + "step": 1862 + }, + { + "epoch": 0.16801190422509807, + "grad_norm": 2.291144143847026, + "learning_rate": 3.8035640688638537e-06, + "loss": 1.1055, + "step": 1863 + }, + { + "epoch": 0.16810208774856833, + "grad_norm": 2.096615097834898, + "learning_rate": 3.8033115012779125e-06, + "loss": 1.0992, + "step": 1864 + }, + { + "epoch": 0.1681922712720386, + "grad_norm": 1.5592967195299114, + "learning_rate": 3.8030587798234915e-06, + "loss": 1.0679, + "step": 1865 + }, + { + "epoch": 0.16828245479550885, + "grad_norm": 1.6582400763789975, + "learning_rate": 3.802805904522153e-06, + "loss": 0.9594, + "step": 1866 + }, + { + "epoch": 0.1683726383189791, + "grad_norm": 2.008970053108281, + "learning_rate": 3.8025528753954742e-06, + "loss": 1.0134, + "step": 1867 + }, + { + "epoch": 0.16846282184244937, + "grad_norm": 1.929788810757793, + "learning_rate": 3.802299692465045e-06, + "loss": 0.9671, + "step": 1868 + }, + { + "epoch": 0.16855300536591963, + "grad_norm": 1.8032748816225428, + "learning_rate": 3.802046355752468e-06, + "loss": 1.0054, + "step": 1869 + }, + { + "epoch": 0.16864318888938992, + "grad_norm": 2.081920857034689, + "learning_rate": 3.80179286527936e-06, + "loss": 1.0793, + "step": 1870 + }, + { + "epoch": 0.16873337241286018, + "grad_norm": 1.7674383118694803, + "learning_rate": 3.801539221067349e-06, + "loss": 0.9371, + "step": 1871 + }, + { + "epoch": 0.16882355593633044, + "grad_norm": 2.7403767141622084, + "learning_rate": 3.801285423138079e-06, + "loss": 0.9798, + "step": 1872 + }, + { + "epoch": 0.1689137394598007, + "grad_norm": 1.8075147668807758, + "learning_rate": 3.8010314715132037e-06, + "loss": 0.9779, + "step": 1873 + }, + { + "epoch": 0.16900392298327097, + "grad_norm": 2.3696019917690294, + "learning_rate": 3.800777366214393e-06, + "loss": 0.9708, + "step": 1874 + }, + { + "epoch": 0.16909410650674123, + "grad_norm": 1.9758036366706972, + "learning_rate": 3.800523107263328e-06, + "loss": 0.9981, + "step": 1875 + }, + { + "epoch": 0.1691842900302115, + "grad_norm": 1.6793985517502175, + "learning_rate": 3.800268694681703e-06, + "loss": 1.0777, + "step": 1876 + }, + { + "epoch": 0.16927447355368175, + "grad_norm": 0.6551485127329457, + "learning_rate": 3.800014128491227e-06, + "loss": 0.8373, + "step": 1877 + }, + { + "epoch": 0.169364657077152, + "grad_norm": 1.724959513976957, + "learning_rate": 3.79975940871362e-06, + "loss": 1.0763, + "step": 1878 + }, + { + "epoch": 0.16945484060062227, + "grad_norm": 2.6054218064959085, + "learning_rate": 3.799504535370617e-06, + "loss": 1.0286, + "step": 1879 + }, + { + "epoch": 0.16954502412409253, + "grad_norm": 2.919776366828048, + "learning_rate": 3.799249508483964e-06, + "loss": 0.9591, + "step": 1880 + }, + { + "epoch": 0.1696352076475628, + "grad_norm": 0.7224611173548582, + "learning_rate": 3.798994328075422e-06, + "loss": 0.8965, + "step": 1881 + }, + { + "epoch": 0.16972539117103305, + "grad_norm": 1.9238995069244498, + "learning_rate": 3.798738994166765e-06, + "loss": 0.9537, + "step": 1882 + }, + { + "epoch": 0.1698155746945033, + "grad_norm": 1.5446138135099472, + "learning_rate": 3.7984835067797788e-06, + "loss": 1.0264, + "step": 1883 + }, + { + "epoch": 0.16990575821797357, + "grad_norm": 3.2351368177392725, + "learning_rate": 3.798227865936263e-06, + "loss": 1.1146, + "step": 1884 + }, + { + "epoch": 0.16999594174144383, + "grad_norm": 2.288771679211482, + "learning_rate": 3.7979720716580297e-06, + "loss": 0.8881, + "step": 1885 + }, + { + "epoch": 0.1700861252649141, + "grad_norm": 1.678664068662912, + "learning_rate": 3.7977161239669057e-06, + "loss": 1.0089, + "step": 1886 + }, + { + "epoch": 0.17017630878838436, + "grad_norm": 2.3068170568836712, + "learning_rate": 3.7974600228847294e-06, + "loss": 1.0051, + "step": 1887 + }, + { + "epoch": 0.17026649231185462, + "grad_norm": 2.6117644191609375, + "learning_rate": 3.7972037684333534e-06, + "loss": 0.9646, + "step": 1888 + }, + { + "epoch": 0.17035667583532488, + "grad_norm": 1.6584446599707652, + "learning_rate": 3.796947360634642e-06, + "loss": 0.9485, + "step": 1889 + }, + { + "epoch": 0.17044685935879514, + "grad_norm": 1.6700138178270856, + "learning_rate": 3.796690799510473e-06, + "loss": 1.0711, + "step": 1890 + }, + { + "epoch": 0.1705370428822654, + "grad_norm": 2.2819375870082372, + "learning_rate": 3.7964340850827387e-06, + "loss": 0.985, + "step": 1891 + }, + { + "epoch": 0.17062722640573566, + "grad_norm": 0.6756396944704743, + "learning_rate": 3.7961772173733425e-06, + "loss": 0.8043, + "step": 1892 + }, + { + "epoch": 0.17071740992920592, + "grad_norm": 1.4680155277158577, + "learning_rate": 3.7959201964042024e-06, + "loss": 1.0116, + "step": 1893 + }, + { + "epoch": 0.1708075934526762, + "grad_norm": 1.6367778903955763, + "learning_rate": 3.795663022197248e-06, + "loss": 1.0517, + "step": 1894 + }, + { + "epoch": 0.17089777697614647, + "grad_norm": 1.7866582582522619, + "learning_rate": 3.7954056947744242e-06, + "loss": 1.0197, + "step": 1895 + }, + { + "epoch": 0.17098796049961673, + "grad_norm": 2.0333203841667373, + "learning_rate": 3.7951482141576863e-06, + "loss": 0.9952, + "step": 1896 + }, + { + "epoch": 0.171078144023087, + "grad_norm": 1.4581486812128683, + "learning_rate": 3.794890580369004e-06, + "loss": 1.0754, + "step": 1897 + }, + { + "epoch": 0.17116832754655725, + "grad_norm": 2.3475453482280817, + "learning_rate": 3.7946327934303612e-06, + "loss": 1.0553, + "step": 1898 + }, + { + "epoch": 0.1712585110700275, + "grad_norm": 1.6973053079230715, + "learning_rate": 3.794374853363752e-06, + "loss": 1.1169, + "step": 1899 + }, + { + "epoch": 0.17134869459349777, + "grad_norm": 1.4783669400758654, + "learning_rate": 3.794116760191187e-06, + "loss": 1.0457, + "step": 1900 + }, + { + "epoch": 0.17143887811696804, + "grad_norm": 2.193651867706705, + "learning_rate": 3.7938585139346877e-06, + "loss": 1.1173, + "step": 1901 + }, + { + "epoch": 0.1715290616404383, + "grad_norm": 1.7238595786338091, + "learning_rate": 3.793600114616288e-06, + "loss": 1.0002, + "step": 1902 + }, + { + "epoch": 0.17161924516390856, + "grad_norm": 0.7339989125272298, + "learning_rate": 3.793341562258037e-06, + "loss": 0.7943, + "step": 1903 + }, + { + "epoch": 0.17170942868737882, + "grad_norm": 43.666587615972894, + "learning_rate": 3.7930828568819953e-06, + "loss": 1.0331, + "step": 1904 + }, + { + "epoch": 0.17179961221084908, + "grad_norm": 1.8309096296708596, + "learning_rate": 3.7928239985102378e-06, + "loss": 1.0165, + "step": 1905 + }, + { + "epoch": 0.17188979573431934, + "grad_norm": 1.5203048972462136, + "learning_rate": 3.7925649871648505e-06, + "loss": 1.0225, + "step": 1906 + }, + { + "epoch": 0.1719799792577896, + "grad_norm": 1.789648634666924, + "learning_rate": 3.792305822867935e-06, + "loss": 0.9418, + "step": 1907 + }, + { + "epoch": 0.17207016278125986, + "grad_norm": 2.042703118934441, + "learning_rate": 3.792046505641604e-06, + "loss": 0.9808, + "step": 1908 + }, + { + "epoch": 0.17216034630473012, + "grad_norm": 1.881233661365213, + "learning_rate": 3.791787035507984e-06, + "loss": 0.9826, + "step": 1909 + }, + { + "epoch": 0.17225052982820038, + "grad_norm": 1.7889818065052976, + "learning_rate": 3.7915274124892136e-06, + "loss": 0.9835, + "step": 1910 + }, + { + "epoch": 0.17234071335167064, + "grad_norm": 1.4905246436926831, + "learning_rate": 3.7912676366074466e-06, + "loss": 1.0041, + "step": 1911 + }, + { + "epoch": 0.1724308968751409, + "grad_norm": 2.501058545122519, + "learning_rate": 3.7910077078848478e-06, + "loss": 0.982, + "step": 1912 + }, + { + "epoch": 0.17252108039861117, + "grad_norm": 2.076451251377437, + "learning_rate": 3.7907476263435957e-06, + "loss": 0.9872, + "step": 1913 + }, + { + "epoch": 0.17261126392208143, + "grad_norm": 1.785064093884279, + "learning_rate": 3.7904873920058826e-06, + "loss": 0.9897, + "step": 1914 + }, + { + "epoch": 0.1727014474455517, + "grad_norm": 2.234240761305724, + "learning_rate": 3.7902270048939114e-06, + "loss": 0.8964, + "step": 1915 + }, + { + "epoch": 0.17279163096902195, + "grad_norm": 2.1213035971457335, + "learning_rate": 3.7899664650299023e-06, + "loss": 0.9839, + "step": 1916 + }, + { + "epoch": 0.1728818144924922, + "grad_norm": 1.4183028907213335, + "learning_rate": 3.7897057724360836e-06, + "loss": 1.0309, + "step": 1917 + }, + { + "epoch": 0.1729719980159625, + "grad_norm": 3.9074310958556726, + "learning_rate": 3.7894449271347004e-06, + "loss": 1.0376, + "step": 1918 + }, + { + "epoch": 0.17306218153943276, + "grad_norm": 1.624296696394877, + "learning_rate": 3.789183929148009e-06, + "loss": 1.08, + "step": 1919 + }, + { + "epoch": 0.17315236506290302, + "grad_norm": 1.5278990533647663, + "learning_rate": 3.7889227784982795e-06, + "loss": 1.0466, + "step": 1920 + }, + { + "epoch": 0.17324254858637328, + "grad_norm": 1.3802911217769689, + "learning_rate": 3.7886614752077945e-06, + "loss": 1.0164, + "step": 1921 + }, + { + "epoch": 0.17333273210984354, + "grad_norm": 1.7659344697331216, + "learning_rate": 3.7884000192988495e-06, + "loss": 1.0157, + "step": 1922 + }, + { + "epoch": 0.1734229156333138, + "grad_norm": 1.77647818144679, + "learning_rate": 3.7881384107937546e-06, + "loss": 1.0723, + "step": 1923 + }, + { + "epoch": 0.17351309915678406, + "grad_norm": 1.8651510329875411, + "learning_rate": 3.78787664971483e-06, + "loss": 1.0106, + "step": 1924 + }, + { + "epoch": 0.17360328268025432, + "grad_norm": 1.370587812581102, + "learning_rate": 3.7876147360844115e-06, + "loss": 0.9903, + "step": 1925 + }, + { + "epoch": 0.17369346620372458, + "grad_norm": 2.2019879809197946, + "learning_rate": 3.7873526699248474e-06, + "loss": 1.0869, + "step": 1926 + }, + { + "epoch": 0.17378364972719484, + "grad_norm": 7.553322259412808, + "learning_rate": 3.7870904512584974e-06, + "loss": 1.0258, + "step": 1927 + }, + { + "epoch": 0.1738738332506651, + "grad_norm": 2.026361892877119, + "learning_rate": 3.7868280801077368e-06, + "loss": 1.0183, + "step": 1928 + }, + { + "epoch": 0.17396401677413537, + "grad_norm": 1.3884152203407276, + "learning_rate": 3.7865655564949517e-06, + "loss": 1.0119, + "step": 1929 + }, + { + "epoch": 0.17405420029760563, + "grad_norm": 1.6179535751331096, + "learning_rate": 3.786302880442542e-06, + "loss": 0.9902, + "step": 1930 + }, + { + "epoch": 0.1741443838210759, + "grad_norm": 1.5594245398226234, + "learning_rate": 3.7860400519729215e-06, + "loss": 1.0206, + "step": 1931 + }, + { + "epoch": 0.17423456734454615, + "grad_norm": 2.5334100807845252, + "learning_rate": 3.7857770711085157e-06, + "loss": 1.0395, + "step": 1932 + }, + { + "epoch": 0.1743247508680164, + "grad_norm": 1.6025446415722127, + "learning_rate": 3.785513937871763e-06, + "loss": 1.0686, + "step": 1933 + }, + { + "epoch": 0.17441493439148667, + "grad_norm": 1.7043329557262599, + "learning_rate": 3.785250652285116e-06, + "loss": 0.9894, + "step": 1934 + }, + { + "epoch": 0.17450511791495693, + "grad_norm": 1.8189234857305403, + "learning_rate": 3.78498721437104e-06, + "loss": 1.0955, + "step": 1935 + }, + { + "epoch": 0.1745953014384272, + "grad_norm": 2.020407156559684, + "learning_rate": 3.784723624152012e-06, + "loss": 1.0168, + "step": 1936 + }, + { + "epoch": 0.17468548496189745, + "grad_norm": 1.9869038845481417, + "learning_rate": 3.784459881650524e-06, + "loss": 0.9601, + "step": 1937 + }, + { + "epoch": 0.1747756684853677, + "grad_norm": 1.8241490222094414, + "learning_rate": 3.784195986889079e-06, + "loss": 0.9905, + "step": 1938 + }, + { + "epoch": 0.17486585200883797, + "grad_norm": 3.1087619809792972, + "learning_rate": 3.7839319398901946e-06, + "loss": 0.9403, + "step": 1939 + }, + { + "epoch": 0.17495603553230824, + "grad_norm": 1.8542560027242505, + "learning_rate": 3.7836677406764013e-06, + "loss": 1.0869, + "step": 1940 + }, + { + "epoch": 0.1750462190557785, + "grad_norm": 1.5683798972719423, + "learning_rate": 3.7834033892702407e-06, + "loss": 1.0644, + "step": 1941 + }, + { + "epoch": 0.17513640257924878, + "grad_norm": 1.915574204734443, + "learning_rate": 3.783138885694269e-06, + "loss": 0.9786, + "step": 1942 + }, + { + "epoch": 0.17522658610271905, + "grad_norm": 1.885197312326375, + "learning_rate": 3.7828742299710558e-06, + "loss": 1.0902, + "step": 1943 + }, + { + "epoch": 0.1753167696261893, + "grad_norm": 1.7165114650504547, + "learning_rate": 3.782609422123183e-06, + "loss": 0.9976, + "step": 1944 + }, + { + "epoch": 0.17540695314965957, + "grad_norm": 2.567764549852222, + "learning_rate": 3.7823444621732444e-06, + "loss": 1.0201, + "step": 1945 + }, + { + "epoch": 0.17549713667312983, + "grad_norm": 2.050599996640477, + "learning_rate": 3.782079350143849e-06, + "loss": 1.0352, + "step": 1946 + }, + { + "epoch": 0.1755873201966001, + "grad_norm": 4.025908013510984, + "learning_rate": 3.781814086057617e-06, + "loss": 1.1184, + "step": 1947 + }, + { + "epoch": 0.17567750372007035, + "grad_norm": 1.8218185677793604, + "learning_rate": 3.7815486699371826e-06, + "loss": 1.0063, + "step": 1948 + }, + { + "epoch": 0.1757676872435406, + "grad_norm": 3.4328951423292176, + "learning_rate": 3.7812831018051918e-06, + "loss": 0.9924, + "step": 1949 + }, + { + "epoch": 0.17585787076701087, + "grad_norm": 1.5641613329777992, + "learning_rate": 3.7810173816843058e-06, + "loss": 1.0279, + "step": 1950 + }, + { + "epoch": 0.17594805429048113, + "grad_norm": 1.8264165214636798, + "learning_rate": 3.7807515095971955e-06, + "loss": 1.0948, + "step": 1951 + }, + { + "epoch": 0.1760382378139514, + "grad_norm": 1.778345450360323, + "learning_rate": 3.7804854855665475e-06, + "loss": 1.0141, + "step": 1952 + }, + { + "epoch": 0.17612842133742165, + "grad_norm": 1.6085038280874322, + "learning_rate": 3.7802193096150606e-06, + "loss": 1.0304, + "step": 1953 + }, + { + "epoch": 0.17621860486089191, + "grad_norm": 1.713481040121707, + "learning_rate": 3.779952981765446e-06, + "loss": 0.9721, + "step": 1954 + }, + { + "epoch": 0.17630878838436218, + "grad_norm": 1.6370169065298026, + "learning_rate": 3.779686502040429e-06, + "loss": 1.0341, + "step": 1955 + }, + { + "epoch": 0.17639897190783244, + "grad_norm": 0.6999862124583854, + "learning_rate": 3.779419870462746e-06, + "loss": 0.8129, + "step": 1956 + }, + { + "epoch": 0.1764891554313027, + "grad_norm": 3.8760131003805043, + "learning_rate": 3.779153087055148e-06, + "loss": 1.0027, + "step": 1957 + }, + { + "epoch": 0.17657933895477296, + "grad_norm": 2.066890897686831, + "learning_rate": 3.7788861518403988e-06, + "loss": 1.0198, + "step": 1958 + }, + { + "epoch": 0.17666952247824322, + "grad_norm": 1.8427738903309923, + "learning_rate": 3.7786190648412742e-06, + "loss": 0.9131, + "step": 1959 + }, + { + "epoch": 0.17675970600171348, + "grad_norm": 0.6349093230527475, + "learning_rate": 3.778351826080564e-06, + "loss": 0.8254, + "step": 1960 + }, + { + "epoch": 0.17684988952518374, + "grad_norm": 2.066696867670868, + "learning_rate": 3.7780844355810704e-06, + "loss": 1.1236, + "step": 1961 + }, + { + "epoch": 0.176940073048654, + "grad_norm": 1.8602366654766975, + "learning_rate": 3.777816893365608e-06, + "loss": 1.0311, + "step": 1962 + }, + { + "epoch": 0.17703025657212426, + "grad_norm": 1.7428134768686605, + "learning_rate": 3.7775491994570057e-06, + "loss": 1.0953, + "step": 1963 + }, + { + "epoch": 0.17712044009559452, + "grad_norm": 1.6438254248426165, + "learning_rate": 3.777281353878105e-06, + "loss": 0.9946, + "step": 1964 + }, + { + "epoch": 0.17721062361906478, + "grad_norm": 0.6126958154512724, + "learning_rate": 3.777013356651758e-06, + "loss": 0.83, + "step": 1965 + }, + { + "epoch": 0.17730080714253507, + "grad_norm": 3.2522167568626426, + "learning_rate": 3.776745207800834e-06, + "loss": 1.0369, + "step": 1966 + }, + { + "epoch": 0.17739099066600533, + "grad_norm": 1.712591895540612, + "learning_rate": 3.7764769073482122e-06, + "loss": 1.0508, + "step": 1967 + }, + { + "epoch": 0.1774811741894756, + "grad_norm": 1.6726405449904567, + "learning_rate": 3.7762084553167846e-06, + "loss": 1.0907, + "step": 1968 + }, + { + "epoch": 0.17757135771294585, + "grad_norm": 1.6851463151337374, + "learning_rate": 3.775939851729458e-06, + "loss": 1.1143, + "step": 1969 + }, + { + "epoch": 0.17766154123641612, + "grad_norm": 2.4431891927144087, + "learning_rate": 3.775671096609151e-06, + "loss": 0.9819, + "step": 1970 + }, + { + "epoch": 0.17775172475988638, + "grad_norm": 3.5919119737115066, + "learning_rate": 3.775402189978795e-06, + "loss": 1.0016, + "step": 1971 + }, + { + "epoch": 0.17784190828335664, + "grad_norm": 1.545417175759137, + "learning_rate": 3.7751331318613343e-06, + "loss": 1.0014, + "step": 1972 + }, + { + "epoch": 0.1779320918068269, + "grad_norm": 1.8716033687514764, + "learning_rate": 3.774863922279727e-06, + "loss": 1.0402, + "step": 1973 + }, + { + "epoch": 0.17802227533029716, + "grad_norm": 1.425018942857621, + "learning_rate": 3.7745945612569435e-06, + "loss": 1.0772, + "step": 1974 + }, + { + "epoch": 0.17811245885376742, + "grad_norm": 1.77774097070076, + "learning_rate": 3.7743250488159674e-06, + "loss": 0.8993, + "step": 1975 + }, + { + "epoch": 0.17820264237723768, + "grad_norm": 7.556628761863911, + "learning_rate": 3.774055384979794e-06, + "loss": 0.9225, + "step": 1976 + }, + { + "epoch": 0.17829282590070794, + "grad_norm": 1.6218103502504742, + "learning_rate": 3.773785569771433e-06, + "loss": 1.0015, + "step": 1977 + }, + { + "epoch": 0.1783830094241782, + "grad_norm": 1.5352999010894908, + "learning_rate": 3.7735156032139066e-06, + "loss": 1.0077, + "step": 1978 + }, + { + "epoch": 0.17847319294764846, + "grad_norm": 1.6184210748717371, + "learning_rate": 3.773245485330251e-06, + "loss": 1.1262, + "step": 1979 + }, + { + "epoch": 0.17856337647111872, + "grad_norm": 1.6514515012109956, + "learning_rate": 3.7729752161435115e-06, + "loss": 0.99, + "step": 1980 + }, + { + "epoch": 0.17865355999458898, + "grad_norm": 1.725674016547189, + "learning_rate": 3.7727047956767514e-06, + "loss": 1.0856, + "step": 1981 + }, + { + "epoch": 0.17874374351805924, + "grad_norm": 2.6643242581551507, + "learning_rate": 3.7724342239530436e-06, + "loss": 1.1491, + "step": 1982 + }, + { + "epoch": 0.1788339270415295, + "grad_norm": 1.9008893792456483, + "learning_rate": 3.772163500995474e-06, + "loss": 0.9724, + "step": 1983 + }, + { + "epoch": 0.17892411056499977, + "grad_norm": 2.197612277342639, + "learning_rate": 3.7718926268271437e-06, + "loss": 1.1138, + "step": 1984 + }, + { + "epoch": 0.17901429408847003, + "grad_norm": 2.2976865826744404, + "learning_rate": 3.771621601471164e-06, + "loss": 1.0414, + "step": 1985 + }, + { + "epoch": 0.1791044776119403, + "grad_norm": 1.7275611440434435, + "learning_rate": 3.771350424950661e-06, + "loss": 1.067, + "step": 1986 + }, + { + "epoch": 0.17919466113541055, + "grad_norm": 3.1380568295177835, + "learning_rate": 3.771079097288772e-06, + "loss": 0.9662, + "step": 1987 + }, + { + "epoch": 0.1792848446588808, + "grad_norm": 1.5857287490337322, + "learning_rate": 3.770807618508649e-06, + "loss": 1.0389, + "step": 1988 + }, + { + "epoch": 0.17937502818235107, + "grad_norm": 1.479401860768859, + "learning_rate": 3.7705359886334555e-06, + "loss": 1.0132, + "step": 1989 + }, + { + "epoch": 0.17946521170582136, + "grad_norm": 2.456269696571641, + "learning_rate": 3.7702642076863694e-06, + "loss": 0.9999, + "step": 1990 + }, + { + "epoch": 0.17955539522929162, + "grad_norm": 1.832785119830252, + "learning_rate": 3.7699922756905795e-06, + "loss": 1.0781, + "step": 1991 + }, + { + "epoch": 0.17964557875276188, + "grad_norm": 1.6075147228842264, + "learning_rate": 3.7697201926692895e-06, + "loss": 0.9747, + "step": 1992 + }, + { + "epoch": 0.17973576227623214, + "grad_norm": 2.39046023929277, + "learning_rate": 3.7694479586457144e-06, + "loss": 1.0341, + "step": 1993 + }, + { + "epoch": 0.1798259457997024, + "grad_norm": 3.904524277003783, + "learning_rate": 3.7691755736430827e-06, + "loss": 1.0343, + "step": 1994 + }, + { + "epoch": 0.17991612932317266, + "grad_norm": 2.017206916745181, + "learning_rate": 3.768903037684636e-06, + "loss": 1.0685, + "step": 1995 + }, + { + "epoch": 0.18000631284664292, + "grad_norm": 1.5714413964379852, + "learning_rate": 3.7686303507936284e-06, + "loss": 1.0712, + "step": 1996 + }, + { + "epoch": 0.18009649637011318, + "grad_norm": 1.9471932630188875, + "learning_rate": 3.7683575129933272e-06, + "loss": 1.0663, + "step": 1997 + }, + { + "epoch": 0.18018667989358345, + "grad_norm": 1.823257225558966, + "learning_rate": 3.7680845243070128e-06, + "loss": 1.0304, + "step": 1998 + }, + { + "epoch": 0.1802768634170537, + "grad_norm": 0.7323680400136791, + "learning_rate": 3.7678113847579767e-06, + "loss": 0.7922, + "step": 1999 + }, + { + "epoch": 0.18036704694052397, + "grad_norm": 1.9940266814823775, + "learning_rate": 3.7675380943695264e-06, + "loss": 1.0914, + "step": 2000 + }, + { + "epoch": 0.18045723046399423, + "grad_norm": 1.741517558247813, + "learning_rate": 3.7672646531649795e-06, + "loss": 0.986, + "step": 2001 + }, + { + "epoch": 0.1805474139874645, + "grad_norm": 0.5804814692832659, + "learning_rate": 3.7669910611676682e-06, + "loss": 0.814, + "step": 2002 + }, + { + "epoch": 0.18063759751093475, + "grad_norm": 2.0290811550705175, + "learning_rate": 3.7667173184009356e-06, + "loss": 1.0125, + "step": 2003 + }, + { + "epoch": 0.180727781034405, + "grad_norm": 1.798212631577667, + "learning_rate": 3.7664434248881403e-06, + "loss": 1.0847, + "step": 2004 + }, + { + "epoch": 0.18081796455787527, + "grad_norm": 1.7728507798081683, + "learning_rate": 3.766169380652652e-06, + "loss": 1.0351, + "step": 2005 + }, + { + "epoch": 0.18090814808134553, + "grad_norm": 1.627241349535296, + "learning_rate": 3.7658951857178537e-06, + "loss": 1.0627, + "step": 2006 + }, + { + "epoch": 0.1809983316048158, + "grad_norm": 1.635681989330734, + "learning_rate": 3.7656208401071414e-06, + "loss": 1.0333, + "step": 2007 + }, + { + "epoch": 0.18108851512828605, + "grad_norm": 3.4824254350970305, + "learning_rate": 3.7653463438439225e-06, + "loss": 0.9586, + "step": 2008 + }, + { + "epoch": 0.18117869865175631, + "grad_norm": 1.885959511970561, + "learning_rate": 3.7650716969516203e-06, + "loss": 1.0382, + "step": 2009 + }, + { + "epoch": 0.18126888217522658, + "grad_norm": 1.8062240862719945, + "learning_rate": 3.764796899453668e-06, + "loss": 0.9231, + "step": 2010 + }, + { + "epoch": 0.18135906569869684, + "grad_norm": 2.637441849632494, + "learning_rate": 3.7645219513735134e-06, + "loss": 1.0643, + "step": 2011 + }, + { + "epoch": 0.1814492492221671, + "grad_norm": 1.739318880854303, + "learning_rate": 3.764246852734617e-06, + "loss": 1.0655, + "step": 2012 + }, + { + "epoch": 0.18153943274563739, + "grad_norm": 2.814246673741967, + "learning_rate": 3.7639716035604502e-06, + "loss": 0.9967, + "step": 2013 + }, + { + "epoch": 0.18162961626910765, + "grad_norm": 0.6440606219668265, + "learning_rate": 3.7636962038745e-06, + "loss": 0.7862, + "step": 2014 + }, + { + "epoch": 0.1817197997925779, + "grad_norm": 1.7948630969828898, + "learning_rate": 3.763420653700265e-06, + "loss": 0.9815, + "step": 2015 + }, + { + "epoch": 0.18180998331604817, + "grad_norm": 3.3218581248560817, + "learning_rate": 3.7631449530612565e-06, + "loss": 0.9814, + "step": 2016 + }, + { + "epoch": 0.18190016683951843, + "grad_norm": 1.6633087743629205, + "learning_rate": 3.762869101980999e-06, + "loss": 1.1134, + "step": 2017 + }, + { + "epoch": 0.1819903503629887, + "grad_norm": 1.8630126503613695, + "learning_rate": 3.7625931004830287e-06, + "loss": 1.0226, + "step": 2018 + }, + { + "epoch": 0.18208053388645895, + "grad_norm": 2.065291163351188, + "learning_rate": 3.7623169485908966e-06, + "loss": 0.9805, + "step": 2019 + }, + { + "epoch": 0.1821707174099292, + "grad_norm": 0.7263323511651261, + "learning_rate": 3.7620406463281647e-06, + "loss": 0.8326, + "step": 2020 + }, + { + "epoch": 0.18226090093339947, + "grad_norm": 2.4335140992871693, + "learning_rate": 3.7617641937184095e-06, + "loss": 1.0462, + "step": 2021 + }, + { + "epoch": 0.18235108445686973, + "grad_norm": 1.7683162602215534, + "learning_rate": 3.761487590785219e-06, + "loss": 1.0521, + "step": 2022 + }, + { + "epoch": 0.18244126798034, + "grad_norm": 2.2236230554312177, + "learning_rate": 3.7612108375521942e-06, + "loss": 1.0008, + "step": 2023 + }, + { + "epoch": 0.18253145150381025, + "grad_norm": 1.6433571726841152, + "learning_rate": 3.76093393404295e-06, + "loss": 1.0702, + "step": 2024 + }, + { + "epoch": 0.18262163502728052, + "grad_norm": 1.6564125844940294, + "learning_rate": 3.7606568802811126e-06, + "loss": 1.0143, + "step": 2025 + }, + { + "epoch": 0.18271181855075078, + "grad_norm": 2.1874711171014525, + "learning_rate": 3.760379676290322e-06, + "loss": 1.0306, + "step": 2026 + }, + { + "epoch": 0.18280200207422104, + "grad_norm": 2.938469909275037, + "learning_rate": 3.760102322094231e-06, + "loss": 0.9512, + "step": 2027 + }, + { + "epoch": 0.1828921855976913, + "grad_norm": 2.5075438168349455, + "learning_rate": 3.759824817716504e-06, + "loss": 0.9848, + "step": 2028 + }, + { + "epoch": 0.18298236912116156, + "grad_norm": 1.7506218895140597, + "learning_rate": 3.759547163180821e-06, + "loss": 1.0771, + "step": 2029 + }, + { + "epoch": 0.18307255264463182, + "grad_norm": 1.7227442020445463, + "learning_rate": 3.759269358510871e-06, + "loss": 0.8237, + "step": 2030 + }, + { + "epoch": 0.18316273616810208, + "grad_norm": 1.6464326065269368, + "learning_rate": 3.75899140373036e-06, + "loss": 0.9283, + "step": 2031 + }, + { + "epoch": 0.18325291969157234, + "grad_norm": 5.308863214049528, + "learning_rate": 3.7587132988630028e-06, + "loss": 1.0846, + "step": 2032 + }, + { + "epoch": 0.1833431032150426, + "grad_norm": 3.1020256868893883, + "learning_rate": 3.7584350439325295e-06, + "loss": 1.0579, + "step": 2033 + }, + { + "epoch": 0.18343328673851286, + "grad_norm": 1.9741170478834067, + "learning_rate": 3.758156638962682e-06, + "loss": 1.0145, + "step": 2034 + }, + { + "epoch": 0.18352347026198312, + "grad_norm": 3.651122320411684, + "learning_rate": 3.757878083977216e-06, + "loss": 0.9868, + "step": 2035 + }, + { + "epoch": 0.18361365378545338, + "grad_norm": 2.2406039610743527, + "learning_rate": 3.7575993789999e-06, + "loss": 1.0102, + "step": 2036 + }, + { + "epoch": 0.18370383730892367, + "grad_norm": 1.8877129314644328, + "learning_rate": 3.757320524054512e-06, + "loss": 1.0612, + "step": 2037 + }, + { + "epoch": 0.18379402083239393, + "grad_norm": 2.964023562081568, + "learning_rate": 3.757041519164848e-06, + "loss": 0.9569, + "step": 2038 + }, + { + "epoch": 0.1838842043558642, + "grad_norm": 1.9550845882526158, + "learning_rate": 3.7567623643547133e-06, + "loss": 1.0298, + "step": 2039 + }, + { + "epoch": 0.18397438787933446, + "grad_norm": 5.28785698589664, + "learning_rate": 3.756483059647927e-06, + "loss": 1.0237, + "step": 2040 + }, + { + "epoch": 0.18406457140280472, + "grad_norm": 2.3042888749212183, + "learning_rate": 3.756203605068321e-06, + "loss": 0.9919, + "step": 2041 + }, + { + "epoch": 0.18415475492627498, + "grad_norm": 2.5465698059329256, + "learning_rate": 3.7559240006397396e-06, + "loss": 1.0126, + "step": 2042 + }, + { + "epoch": 0.18424493844974524, + "grad_norm": 2.320424806079733, + "learning_rate": 3.7556442463860406e-06, + "loss": 1.1419, + "step": 2043 + }, + { + "epoch": 0.1843351219732155, + "grad_norm": 1.6148000592216385, + "learning_rate": 3.7553643423310934e-06, + "loss": 1.1331, + "step": 2044 + }, + { + "epoch": 0.18442530549668576, + "grad_norm": 1.9666696486477415, + "learning_rate": 3.755084288498782e-06, + "loss": 0.8825, + "step": 2045 + }, + { + "epoch": 0.18451548902015602, + "grad_norm": 2.016683610997331, + "learning_rate": 3.754804084913002e-06, + "loss": 1.0277, + "step": 2046 + }, + { + "epoch": 0.18460567254362628, + "grad_norm": 1.5189469103849527, + "learning_rate": 3.754523731597661e-06, + "loss": 1.0879, + "step": 2047 + }, + { + "epoch": 0.18469585606709654, + "grad_norm": 2.3862993586747296, + "learning_rate": 3.754243228576681e-06, + "loss": 0.9238, + "step": 2048 + }, + { + "epoch": 0.1847860395905668, + "grad_norm": 1.5907515200077549, + "learning_rate": 3.753962575873996e-06, + "loss": 1.0783, + "step": 2049 + }, + { + "epoch": 0.18487622311403706, + "grad_norm": 2.5252010454267237, + "learning_rate": 3.7536817735135527e-06, + "loss": 0.9833, + "step": 2050 + }, + { + "epoch": 0.18496640663750732, + "grad_norm": 1.40319182272924, + "learning_rate": 3.753400821519311e-06, + "loss": 1.0624, + "step": 2051 + }, + { + "epoch": 0.18505659016097759, + "grad_norm": 1.4792635800845717, + "learning_rate": 3.7531197199152426e-06, + "loss": 0.9567, + "step": 2052 + }, + { + "epoch": 0.18514677368444785, + "grad_norm": 0.7178861983600021, + "learning_rate": 3.7528384687253335e-06, + "loss": 0.8548, + "step": 2053 + }, + { + "epoch": 0.1852369572079181, + "grad_norm": 2.825169573395325, + "learning_rate": 3.7525570679735815e-06, + "loss": 1.1305, + "step": 2054 + }, + { + "epoch": 0.18532714073138837, + "grad_norm": 1.7330627288251417, + "learning_rate": 3.7522755176839965e-06, + "loss": 1.0376, + "step": 2055 + }, + { + "epoch": 0.18541732425485863, + "grad_norm": 1.75897978278638, + "learning_rate": 3.7519938178806027e-06, + "loss": 1.0324, + "step": 2056 + }, + { + "epoch": 0.1855075077783289, + "grad_norm": 2.1821084199747403, + "learning_rate": 3.7517119685874358e-06, + "loss": 0.9688, + "step": 2057 + }, + { + "epoch": 0.18559769130179915, + "grad_norm": 1.6132981712802432, + "learning_rate": 3.7514299698285447e-06, + "loss": 1.0135, + "step": 2058 + }, + { + "epoch": 0.1856878748252694, + "grad_norm": 1.4942594992029503, + "learning_rate": 3.751147821627991e-06, + "loss": 1.0029, + "step": 2059 + }, + { + "epoch": 0.18577805834873967, + "grad_norm": 2.394430385537509, + "learning_rate": 3.75086552400985e-06, + "loss": 1.0434, + "step": 2060 + }, + { + "epoch": 0.18586824187220996, + "grad_norm": 2.1173751015967426, + "learning_rate": 3.750583076998208e-06, + "loss": 0.9997, + "step": 2061 + }, + { + "epoch": 0.18595842539568022, + "grad_norm": 1.9353428799539065, + "learning_rate": 3.7503004806171655e-06, + "loss": 1.0012, + "step": 2062 + }, + { + "epoch": 0.18604860891915048, + "grad_norm": 1.8826512311193697, + "learning_rate": 3.7500177348908354e-06, + "loss": 0.9535, + "step": 2063 + }, + { + "epoch": 0.18613879244262074, + "grad_norm": 1.503567189229083, + "learning_rate": 3.749734839843342e-06, + "loss": 1.0458, + "step": 2064 + }, + { + "epoch": 0.186228975966091, + "grad_norm": 1.8552666805058748, + "learning_rate": 3.7494517954988245e-06, + "loss": 1.0239, + "step": 2065 + }, + { + "epoch": 0.18631915948956126, + "grad_norm": 1.7306322834251502, + "learning_rate": 3.749168601881433e-06, + "loss": 1.011, + "step": 2066 + }, + { + "epoch": 0.18640934301303153, + "grad_norm": 1.8473404660857564, + "learning_rate": 3.7488852590153315e-06, + "loss": 1.0039, + "step": 2067 + }, + { + "epoch": 0.1864995265365018, + "grad_norm": 2.3540620921266386, + "learning_rate": 3.748601766924697e-06, + "loss": 1.019, + "step": 2068 + }, + { + "epoch": 0.18658971005997205, + "grad_norm": 2.0727980178794083, + "learning_rate": 3.7483181256337176e-06, + "loss": 0.9882, + "step": 2069 + }, + { + "epoch": 0.1866798935834423, + "grad_norm": 1.5224808256698903, + "learning_rate": 3.7480343351665962e-06, + "loss": 1.0276, + "step": 2070 + }, + { + "epoch": 0.18677007710691257, + "grad_norm": 1.5460739373493677, + "learning_rate": 3.747750395547546e-06, + "loss": 1.0095, + "step": 2071 + }, + { + "epoch": 0.18686026063038283, + "grad_norm": 1.2658020178307592, + "learning_rate": 3.7474663068007956e-06, + "loss": 1.0085, + "step": 2072 + }, + { + "epoch": 0.1869504441538531, + "grad_norm": 1.2283232376032263, + "learning_rate": 3.747182068950584e-06, + "loss": 1.0061, + "step": 2073 + }, + { + "epoch": 0.18704062767732335, + "grad_norm": 1.9577609165760712, + "learning_rate": 3.7468976820211643e-06, + "loss": 1.0245, + "step": 2074 + }, + { + "epoch": 0.1871308112007936, + "grad_norm": 1.654032248056943, + "learning_rate": 3.746613146036803e-06, + "loss": 1.0237, + "step": 2075 + }, + { + "epoch": 0.18722099472426387, + "grad_norm": 1.7317226164183581, + "learning_rate": 3.7463284610217766e-06, + "loss": 1.0594, + "step": 2076 + }, + { + "epoch": 0.18731117824773413, + "grad_norm": 1.9236122281300423, + "learning_rate": 3.746043627000377e-06, + "loss": 0.9535, + "step": 2077 + }, + { + "epoch": 0.1874013617712044, + "grad_norm": 1.7541095981088908, + "learning_rate": 3.7457586439969076e-06, + "loss": 0.9699, + "step": 2078 + }, + { + "epoch": 0.18749154529467466, + "grad_norm": 1.479226670816679, + "learning_rate": 3.7454735120356842e-06, + "loss": 1.0119, + "step": 2079 + }, + { + "epoch": 0.18758172881814492, + "grad_norm": 1.4288640386771514, + "learning_rate": 3.7451882311410373e-06, + "loss": 1.0854, + "step": 2080 + }, + { + "epoch": 0.18767191234161518, + "grad_norm": 2.8045561608295215, + "learning_rate": 3.7449028013373074e-06, + "loss": 1.0575, + "step": 2081 + }, + { + "epoch": 0.18776209586508544, + "grad_norm": 1.6477800839131405, + "learning_rate": 3.7446172226488485e-06, + "loss": 1.0179, + "step": 2082 + }, + { + "epoch": 0.1878522793885557, + "grad_norm": 2.096307246155005, + "learning_rate": 3.7443314951000285e-06, + "loss": 1.0862, + "step": 2083 + }, + { + "epoch": 0.18794246291202596, + "grad_norm": 2.041519847443591, + "learning_rate": 3.7440456187152276e-06, + "loss": 1.0079, + "step": 2084 + }, + { + "epoch": 0.18803264643549625, + "grad_norm": 2.1642034050965706, + "learning_rate": 3.7437595935188377e-06, + "loss": 1.1057, + "step": 2085 + }, + { + "epoch": 0.1881228299589665, + "grad_norm": 2.620596689316508, + "learning_rate": 3.7434734195352647e-06, + "loss": 1.018, + "step": 2086 + }, + { + "epoch": 0.18821301348243677, + "grad_norm": 0.663643244619195, + "learning_rate": 3.743187096788926e-06, + "loss": 0.8193, + "step": 2087 + }, + { + "epoch": 0.18830319700590703, + "grad_norm": 1.6051373290494593, + "learning_rate": 3.7429006253042524e-06, + "loss": 1.0232, + "step": 2088 + }, + { + "epoch": 0.1883933805293773, + "grad_norm": 1.602929104441986, + "learning_rate": 3.7426140051056867e-06, + "loss": 0.9659, + "step": 2089 + }, + { + "epoch": 0.18848356405284755, + "grad_norm": 1.7611327285656586, + "learning_rate": 3.7423272362176856e-06, + "loss": 0.9598, + "step": 2090 + }, + { + "epoch": 0.1885737475763178, + "grad_norm": 1.805723281145109, + "learning_rate": 3.742040318664718e-06, + "loss": 1.0647, + "step": 2091 + }, + { + "epoch": 0.18866393109978807, + "grad_norm": 2.2556938606082264, + "learning_rate": 3.7417532524712643e-06, + "loss": 1.0746, + "step": 2092 + }, + { + "epoch": 0.18875411462325833, + "grad_norm": 1.7272668152803792, + "learning_rate": 3.7414660376618195e-06, + "loss": 1.0495, + "step": 2093 + }, + { + "epoch": 0.1888442981467286, + "grad_norm": 0.6139563956573668, + "learning_rate": 3.74117867426089e-06, + "loss": 0.828, + "step": 2094 + }, + { + "epoch": 0.18893448167019886, + "grad_norm": 1.8047226411114623, + "learning_rate": 3.7408911622929954e-06, + "loss": 0.996, + "step": 2095 + }, + { + "epoch": 0.18902466519366912, + "grad_norm": 1.543963456936987, + "learning_rate": 3.740603501782668e-06, + "loss": 1.0526, + "step": 2096 + }, + { + "epoch": 0.18911484871713938, + "grad_norm": 2.0509552800042696, + "learning_rate": 3.7403156927544516e-06, + "loss": 0.9593, + "step": 2097 + }, + { + "epoch": 0.18920503224060964, + "grad_norm": 2.0831132263894685, + "learning_rate": 3.740027735232904e-06, + "loss": 0.9477, + "step": 2098 + }, + { + "epoch": 0.1892952157640799, + "grad_norm": 2.237296055710924, + "learning_rate": 3.7397396292425966e-06, + "loss": 1.0051, + "step": 2099 + }, + { + "epoch": 0.18938539928755016, + "grad_norm": 2.072385045672463, + "learning_rate": 3.7394513748081105e-06, + "loss": 1.0013, + "step": 2100 + }, + { + "epoch": 0.18947558281102042, + "grad_norm": 1.939540280998936, + "learning_rate": 3.7391629719540418e-06, + "loss": 1.0533, + "step": 2101 + }, + { + "epoch": 0.18956576633449068, + "grad_norm": 1.9927143073569877, + "learning_rate": 3.7388744207049998e-06, + "loss": 1.0069, + "step": 2102 + }, + { + "epoch": 0.18965594985796094, + "grad_norm": 2.4831753119386835, + "learning_rate": 3.7385857210856027e-06, + "loss": 0.9896, + "step": 2103 + }, + { + "epoch": 0.1897461333814312, + "grad_norm": 2.4131379069803427, + "learning_rate": 3.738296873120486e-06, + "loss": 1.0007, + "step": 2104 + }, + { + "epoch": 0.18983631690490146, + "grad_norm": 2.6044786393535886, + "learning_rate": 3.7380078768342955e-06, + "loss": 0.9898, + "step": 2105 + }, + { + "epoch": 0.18992650042837173, + "grad_norm": 3.0219495793698985, + "learning_rate": 3.7377187322516895e-06, + "loss": 1.0359, + "step": 2106 + }, + { + "epoch": 0.19001668395184199, + "grad_norm": 2.587895415679528, + "learning_rate": 3.7374294393973395e-06, + "loss": 0.9697, + "step": 2107 + }, + { + "epoch": 0.19010686747531225, + "grad_norm": 1.8807142920703896, + "learning_rate": 3.7371399982959294e-06, + "loss": 1.0747, + "step": 2108 + }, + { + "epoch": 0.19019705099878254, + "grad_norm": 2.052742272968869, + "learning_rate": 3.7368504089721565e-06, + "loss": 0.9335, + "step": 2109 + }, + { + "epoch": 0.1902872345222528, + "grad_norm": 1.616751349417184, + "learning_rate": 3.73656067145073e-06, + "loss": 0.9688, + "step": 2110 + }, + { + "epoch": 0.19037741804572306, + "grad_norm": 1.5758824041275397, + "learning_rate": 3.736270785756371e-06, + "loss": 1.07, + "step": 2111 + }, + { + "epoch": 0.19046760156919332, + "grad_norm": 4.0805631050771165, + "learning_rate": 3.7359807519138156e-06, + "loss": 1.0232, + "step": 2112 + }, + { + "epoch": 0.19055778509266358, + "grad_norm": 2.8358680569472154, + "learning_rate": 3.73569056994781e-06, + "loss": 1.0163, + "step": 2113 + }, + { + "epoch": 0.19064796861613384, + "grad_norm": 2.0226137111842215, + "learning_rate": 3.7354002398831144e-06, + "loss": 1.0526, + "step": 2114 + }, + { + "epoch": 0.1907381521396041, + "grad_norm": 2.137522990677539, + "learning_rate": 3.7351097617445015e-06, + "loss": 1.0166, + "step": 2115 + }, + { + "epoch": 0.19082833566307436, + "grad_norm": 0.697669566711939, + "learning_rate": 3.7348191355567567e-06, + "loss": 0.8489, + "step": 2116 + }, + { + "epoch": 0.19091851918654462, + "grad_norm": 2.0214705757001834, + "learning_rate": 3.734528361344677e-06, + "loss": 1.0483, + "step": 2117 + }, + { + "epoch": 0.19100870271001488, + "grad_norm": 1.9908208609959819, + "learning_rate": 3.734237439133074e-06, + "loss": 1.0151, + "step": 2118 + }, + { + "epoch": 0.19109888623348514, + "grad_norm": 1.7530824217496193, + "learning_rate": 3.7339463689467702e-06, + "loss": 1.0629, + "step": 2119 + }, + { + "epoch": 0.1911890697569554, + "grad_norm": 1.5480526381372584, + "learning_rate": 3.733655150810601e-06, + "loss": 0.9886, + "step": 2120 + }, + { + "epoch": 0.19127925328042567, + "grad_norm": 2.132023815730557, + "learning_rate": 3.7333637847494154e-06, + "loss": 1.1521, + "step": 2121 + }, + { + "epoch": 0.19136943680389593, + "grad_norm": 1.875405204222062, + "learning_rate": 3.7330722707880734e-06, + "loss": 1.1705, + "step": 2122 + }, + { + "epoch": 0.1914596203273662, + "grad_norm": 3.0116296104370845, + "learning_rate": 3.7327806089514497e-06, + "loss": 0.9267, + "step": 2123 + }, + { + "epoch": 0.19154980385083645, + "grad_norm": 3.551541388701072, + "learning_rate": 3.7324887992644297e-06, + "loss": 1.1154, + "step": 2124 + }, + { + "epoch": 0.1916399873743067, + "grad_norm": 0.7144355650798755, + "learning_rate": 3.7321968417519123e-06, + "loss": 0.8234, + "step": 2125 + }, + { + "epoch": 0.19173017089777697, + "grad_norm": 3.190087857366898, + "learning_rate": 3.7319047364388097e-06, + "loss": 1.0067, + "step": 2126 + }, + { + "epoch": 0.19182035442124723, + "grad_norm": 0.6099382633877173, + "learning_rate": 3.7316124833500453e-06, + "loss": 0.8094, + "step": 2127 + }, + { + "epoch": 0.1919105379447175, + "grad_norm": 1.9439755790298971, + "learning_rate": 3.731320082510556e-06, + "loss": 1.0065, + "step": 2128 + }, + { + "epoch": 0.19200072146818775, + "grad_norm": 1.7660452418553458, + "learning_rate": 3.7310275339452906e-06, + "loss": 1.03, + "step": 2129 + }, + { + "epoch": 0.192090904991658, + "grad_norm": 2.0243456360164482, + "learning_rate": 3.7307348376792113e-06, + "loss": 1.0062, + "step": 2130 + }, + { + "epoch": 0.19218108851512827, + "grad_norm": 1.5355276963951137, + "learning_rate": 3.730441993737292e-06, + "loss": 1.0098, + "step": 2131 + }, + { + "epoch": 0.19227127203859856, + "grad_norm": 1.7402037543996915, + "learning_rate": 3.7301490021445205e-06, + "loss": 1.0017, + "step": 2132 + }, + { + "epoch": 0.19236145556206882, + "grad_norm": 2.2133448135507208, + "learning_rate": 3.7298558629258966e-06, + "loss": 0.9548, + "step": 2133 + }, + { + "epoch": 0.19245163908553908, + "grad_norm": 1.9258029155015444, + "learning_rate": 3.7295625761064314e-06, + "loss": 1.0529, + "step": 2134 + }, + { + "epoch": 0.19254182260900934, + "grad_norm": 1.932959805228065, + "learning_rate": 3.7292691417111504e-06, + "loss": 0.9817, + "step": 2135 + }, + { + "epoch": 0.1926320061324796, + "grad_norm": 1.8831734845452468, + "learning_rate": 3.728975559765092e-06, + "loss": 0.9717, + "step": 2136 + }, + { + "epoch": 0.19272218965594987, + "grad_norm": 2.027111238586155, + "learning_rate": 3.728681830293305e-06, + "loss": 1.0432, + "step": 2137 + }, + { + "epoch": 0.19281237317942013, + "grad_norm": 1.8209082156684657, + "learning_rate": 3.7283879533208523e-06, + "loss": 0.9718, + "step": 2138 + }, + { + "epoch": 0.1929025567028904, + "grad_norm": 1.6957022223374627, + "learning_rate": 3.7280939288728094e-06, + "loss": 1.0047, + "step": 2139 + }, + { + "epoch": 0.19299274022636065, + "grad_norm": 1.6093381858985347, + "learning_rate": 3.7277997569742637e-06, + "loss": 0.9795, + "step": 2140 + }, + { + "epoch": 0.1930829237498309, + "grad_norm": 2.833045009827758, + "learning_rate": 3.7275054376503155e-06, + "loss": 0.9816, + "step": 2141 + }, + { + "epoch": 0.19317310727330117, + "grad_norm": 2.934741734021949, + "learning_rate": 3.7272109709260783e-06, + "loss": 0.9576, + "step": 2142 + }, + { + "epoch": 0.19326329079677143, + "grad_norm": 0.7653919955330205, + "learning_rate": 3.7269163568266774e-06, + "loss": 0.8252, + "step": 2143 + }, + { + "epoch": 0.1933534743202417, + "grad_norm": 1.8968254120580907, + "learning_rate": 3.7266215953772512e-06, + "loss": 1.0009, + "step": 2144 + }, + { + "epoch": 0.19344365784371195, + "grad_norm": 2.742702251060069, + "learning_rate": 3.7263266866029492e-06, + "loss": 1.0488, + "step": 2145 + }, + { + "epoch": 0.1935338413671822, + "grad_norm": 1.5044045155462258, + "learning_rate": 3.726031630528936e-06, + "loss": 0.9843, + "step": 2146 + }, + { + "epoch": 0.19362402489065247, + "grad_norm": 22.370338337931226, + "learning_rate": 3.7257364271803865e-06, + "loss": 0.9077, + "step": 2147 + }, + { + "epoch": 0.19371420841412273, + "grad_norm": 1.746291249688437, + "learning_rate": 3.7254410765824896e-06, + "loss": 0.9395, + "step": 2148 + }, + { + "epoch": 0.193804391937593, + "grad_norm": 2.15530297311406, + "learning_rate": 3.725145578760446e-06, + "loss": 0.9355, + "step": 2149 + }, + { + "epoch": 0.19389457546106326, + "grad_norm": 3.0583436749220234, + "learning_rate": 3.7248499337394696e-06, + "loss": 0.9268, + "step": 2150 + }, + { + "epoch": 0.19398475898453352, + "grad_norm": 1.8488712630567812, + "learning_rate": 3.7245541415447848e-06, + "loss": 1.1033, + "step": 2151 + }, + { + "epoch": 0.19407494250800378, + "grad_norm": 1.6694234539213852, + "learning_rate": 3.724258202201633e-06, + "loss": 1.0366, + "step": 2152 + }, + { + "epoch": 0.19416512603147404, + "grad_norm": 1.7696814736437625, + "learning_rate": 3.7239621157352633e-06, + "loss": 1.0267, + "step": 2153 + }, + { + "epoch": 0.1942553095549443, + "grad_norm": 1.5948690337939457, + "learning_rate": 3.7236658821709403e-06, + "loss": 1.0073, + "step": 2154 + }, + { + "epoch": 0.19434549307841456, + "grad_norm": 1.8462050951208324, + "learning_rate": 3.7233695015339404e-06, + "loss": 1.0577, + "step": 2155 + }, + { + "epoch": 0.19443567660188485, + "grad_norm": 1.7120937116055697, + "learning_rate": 3.7230729738495513e-06, + "loss": 1.003, + "step": 2156 + }, + { + "epoch": 0.1945258601253551, + "grad_norm": 1.3857193045040703, + "learning_rate": 3.722776299143075e-06, + "loss": 1.0081, + "step": 2157 + }, + { + "epoch": 0.19461604364882537, + "grad_norm": 0.9369456241548844, + "learning_rate": 3.722479477439826e-06, + "loss": 0.8237, + "step": 2158 + }, + { + "epoch": 0.19470622717229563, + "grad_norm": 1.3337724330198917, + "learning_rate": 3.7221825087651306e-06, + "loss": 0.9915, + "step": 2159 + }, + { + "epoch": 0.1947964106957659, + "grad_norm": 1.9769025065377464, + "learning_rate": 3.7218853931443274e-06, + "loss": 0.9515, + "step": 2160 + }, + { + "epoch": 0.19488659421923615, + "grad_norm": 1.5639485320527764, + "learning_rate": 3.721588130602768e-06, + "loss": 0.9529, + "step": 2161 + }, + { + "epoch": 0.19497677774270641, + "grad_norm": 1.827293728669354, + "learning_rate": 3.7212907211658164e-06, + "loss": 0.9902, + "step": 2162 + }, + { + "epoch": 0.19506696126617667, + "grad_norm": 1.9004667762796907, + "learning_rate": 3.72099316485885e-06, + "loss": 0.9762, + "step": 2163 + }, + { + "epoch": 0.19515714478964694, + "grad_norm": 1.5617169516144207, + "learning_rate": 3.720695461707256e-06, + "loss": 1.0456, + "step": 2164 + }, + { + "epoch": 0.1952473283131172, + "grad_norm": 1.5059989340070419, + "learning_rate": 3.7203976117364383e-06, + "loss": 1.0284, + "step": 2165 + }, + { + "epoch": 0.19533751183658746, + "grad_norm": 1.8557300263205767, + "learning_rate": 3.7200996149718105e-06, + "loss": 0.9931, + "step": 2166 + }, + { + "epoch": 0.19542769536005772, + "grad_norm": 1.4420753311395187, + "learning_rate": 3.7198014714387985e-06, + "loss": 1.0015, + "step": 2167 + }, + { + "epoch": 0.19551787888352798, + "grad_norm": 1.625978688648197, + "learning_rate": 3.7195031811628422e-06, + "loss": 0.946, + "step": 2168 + }, + { + "epoch": 0.19560806240699824, + "grad_norm": 2.0878803477734666, + "learning_rate": 3.719204744169393e-06, + "loss": 1.0426, + "step": 2169 + }, + { + "epoch": 0.1956982459304685, + "grad_norm": 1.898565280208906, + "learning_rate": 3.718906160483916e-06, + "loss": 0.9855, + "step": 2170 + }, + { + "epoch": 0.19578842945393876, + "grad_norm": 1.570219957651897, + "learning_rate": 3.7186074301318868e-06, + "loss": 0.9825, + "step": 2171 + }, + { + "epoch": 0.19587861297740902, + "grad_norm": 1.490514006570948, + "learning_rate": 3.7183085531387957e-06, + "loss": 1.0848, + "step": 2172 + }, + { + "epoch": 0.19596879650087928, + "grad_norm": 1.5719303887317762, + "learning_rate": 3.7180095295301443e-06, + "loss": 0.9843, + "step": 2173 + }, + { + "epoch": 0.19605898002434954, + "grad_norm": 1.5685787213811355, + "learning_rate": 3.7177103593314465e-06, + "loss": 1.0754, + "step": 2174 + }, + { + "epoch": 0.1961491635478198, + "grad_norm": 1.7875713147429246, + "learning_rate": 3.7174110425682297e-06, + "loss": 1.0492, + "step": 2175 + }, + { + "epoch": 0.19623934707129007, + "grad_norm": 1.810012689198226, + "learning_rate": 3.7171115792660333e-06, + "loss": 1.1424, + "step": 2176 + }, + { + "epoch": 0.19632953059476033, + "grad_norm": 2.0643148818055583, + "learning_rate": 3.7168119694504083e-06, + "loss": 1.0753, + "step": 2177 + }, + { + "epoch": 0.1964197141182306, + "grad_norm": 1.8866956247369662, + "learning_rate": 3.71651221314692e-06, + "loss": 1.0063, + "step": 2178 + }, + { + "epoch": 0.19650989764170085, + "grad_norm": 1.9023468205546639, + "learning_rate": 3.716212310381145e-06, + "loss": 1.0109, + "step": 2179 + }, + { + "epoch": 0.19660008116517114, + "grad_norm": 2.406616975618261, + "learning_rate": 3.7159122611786725e-06, + "loss": 1.0572, + "step": 2180 + }, + { + "epoch": 0.1966902646886414, + "grad_norm": 5.19635962926032, + "learning_rate": 3.7156120655651045e-06, + "loss": 0.9968, + "step": 2181 + }, + { + "epoch": 0.19678044821211166, + "grad_norm": 1.4747858409260883, + "learning_rate": 3.7153117235660553e-06, + "loss": 0.9905, + "step": 2182 + }, + { + "epoch": 0.19687063173558192, + "grad_norm": 0.7318416891490481, + "learning_rate": 3.7150112352071514e-06, + "loss": 0.8227, + "step": 2183 + }, + { + "epoch": 0.19696081525905218, + "grad_norm": 1.7801533719754985, + "learning_rate": 3.7147106005140326e-06, + "loss": 0.9797, + "step": 2184 + }, + { + "epoch": 0.19705099878252244, + "grad_norm": 0.6816825307299637, + "learning_rate": 3.714409819512351e-06, + "loss": 0.832, + "step": 2185 + }, + { + "epoch": 0.1971411823059927, + "grad_norm": 2.239324143662959, + "learning_rate": 3.7141088922277695e-06, + "loss": 1.0377, + "step": 2186 + }, + { + "epoch": 0.19723136582946296, + "grad_norm": 1.891180035796542, + "learning_rate": 3.7138078186859664e-06, + "loss": 1.056, + "step": 2187 + }, + { + "epoch": 0.19732154935293322, + "grad_norm": 1.6031611209837844, + "learning_rate": 3.7135065989126303e-06, + "loss": 1.0265, + "step": 2188 + }, + { + "epoch": 0.19741173287640348, + "grad_norm": 0.7316307576375123, + "learning_rate": 3.713205232933463e-06, + "loss": 0.8863, + "step": 2189 + }, + { + "epoch": 0.19750191639987374, + "grad_norm": 1.5944335817874533, + "learning_rate": 3.7129037207741792e-06, + "loss": 1.0223, + "step": 2190 + }, + { + "epoch": 0.197592099923344, + "grad_norm": 2.447933949078094, + "learning_rate": 3.7126020624605046e-06, + "loss": 1.0241, + "step": 2191 + }, + { + "epoch": 0.19768228344681427, + "grad_norm": 2.196118841363797, + "learning_rate": 3.7123002580181785e-06, + "loss": 0.9699, + "step": 2192 + }, + { + "epoch": 0.19777246697028453, + "grad_norm": 1.4204363409089578, + "learning_rate": 3.7119983074729532e-06, + "loss": 1.084, + "step": 2193 + }, + { + "epoch": 0.1978626504937548, + "grad_norm": 1.8993042901883588, + "learning_rate": 3.7116962108505926e-06, + "loss": 1.0697, + "step": 2194 + }, + { + "epoch": 0.19795283401722505, + "grad_norm": 1.6197080273719413, + "learning_rate": 3.711393968176873e-06, + "loss": 1.1102, + "step": 2195 + }, + { + "epoch": 0.1980430175406953, + "grad_norm": 1.4307942093640023, + "learning_rate": 3.711091579477584e-06, + "loss": 1.023, + "step": 2196 + }, + { + "epoch": 0.19813320106416557, + "grad_norm": 2.399887292917581, + "learning_rate": 3.7107890447785255e-06, + "loss": 1.0448, + "step": 2197 + }, + { + "epoch": 0.19822338458763583, + "grad_norm": 2.092003392729187, + "learning_rate": 3.710486364105513e-06, + "loss": 1.0632, + "step": 2198 + }, + { + "epoch": 0.1983135681111061, + "grad_norm": 1.9310679575233716, + "learning_rate": 3.7101835374843728e-06, + "loss": 1.0115, + "step": 2199 + }, + { + "epoch": 0.19840375163457635, + "grad_norm": 1.5468682183011062, + "learning_rate": 3.7098805649409427e-06, + "loss": 1.0342, + "step": 2200 + }, + { + "epoch": 0.1984939351580466, + "grad_norm": 1.8809250357665348, + "learning_rate": 3.7095774465010748e-06, + "loss": 1.0826, + "step": 2201 + }, + { + "epoch": 0.19858411868151687, + "grad_norm": 2.378241284769387, + "learning_rate": 3.7092741821906328e-06, + "loss": 1.0149, + "step": 2202 + }, + { + "epoch": 0.19867430220498714, + "grad_norm": 2.1894836151422927, + "learning_rate": 3.708970772035493e-06, + "loss": 0.9937, + "step": 2203 + }, + { + "epoch": 0.19876448572845742, + "grad_norm": 1.552012303239684, + "learning_rate": 3.7086672160615427e-06, + "loss": 0.9833, + "step": 2204 + }, + { + "epoch": 0.19885466925192768, + "grad_norm": 1.5789606884344327, + "learning_rate": 3.7083635142946852e-06, + "loss": 1.001, + "step": 2205 + }, + { + "epoch": 0.19894485277539795, + "grad_norm": 1.6314050184233722, + "learning_rate": 3.7080596667608327e-06, + "loss": 1.0489, + "step": 2206 + }, + { + "epoch": 0.1990350362988682, + "grad_norm": 1.6830157285647795, + "learning_rate": 3.707755673485911e-06, + "loss": 1.0595, + "step": 2207 + }, + { + "epoch": 0.19912521982233847, + "grad_norm": 1.9500177920581294, + "learning_rate": 3.7074515344958584e-06, + "loss": 0.9262, + "step": 2208 + }, + { + "epoch": 0.19921540334580873, + "grad_norm": 0.658426218683019, + "learning_rate": 3.707147249816627e-06, + "loss": 0.8516, + "step": 2209 + }, + { + "epoch": 0.199305586869279, + "grad_norm": 1.5589205368226962, + "learning_rate": 3.706842819474178e-06, + "loss": 1.0416, + "step": 2210 + }, + { + "epoch": 0.19939577039274925, + "grad_norm": 1.8200190229784128, + "learning_rate": 3.706538243494489e-06, + "loss": 0.9514, + "step": 2211 + }, + { + "epoch": 0.1994859539162195, + "grad_norm": 1.9709298312520511, + "learning_rate": 3.706233521903547e-06, + "loss": 0.9913, + "step": 2212 + }, + { + "epoch": 0.19957613743968977, + "grad_norm": 1.6836713520760986, + "learning_rate": 3.705928654727353e-06, + "loss": 1.0973, + "step": 2213 + }, + { + "epoch": 0.19966632096316003, + "grad_norm": 1.38622199818817, + "learning_rate": 3.7056236419919195e-06, + "loss": 0.992, + "step": 2214 + }, + { + "epoch": 0.1997565044866303, + "grad_norm": 1.4854163281768942, + "learning_rate": 3.705318483723273e-06, + "loss": 1.0647, + "step": 2215 + }, + { + "epoch": 0.19984668801010055, + "grad_norm": 2.2578957090257545, + "learning_rate": 3.7050131799474493e-06, + "loss": 0.9934, + "step": 2216 + }, + { + "epoch": 0.19993687153357081, + "grad_norm": 2.163022008084637, + "learning_rate": 3.7047077306905e-06, + "loss": 1.014, + "step": 2217 + }, + { + "epoch": 0.20002705505704108, + "grad_norm": 1.5954845190011449, + "learning_rate": 3.704402135978488e-06, + "loss": 1.0517, + "step": 2218 + }, + { + "epoch": 0.20011723858051134, + "grad_norm": 2.3659138794136707, + "learning_rate": 3.7040963958374877e-06, + "loss": 1.0431, + "step": 2219 + }, + { + "epoch": 0.2002074221039816, + "grad_norm": 1.4854286068529612, + "learning_rate": 3.7037905102935864e-06, + "loss": 1.0476, + "step": 2220 + }, + { + "epoch": 0.20029760562745186, + "grad_norm": 1.6003754473304734, + "learning_rate": 3.7034844793728837e-06, + "loss": 1.0252, + "step": 2221 + }, + { + "epoch": 0.20038778915092212, + "grad_norm": 1.4898541966728658, + "learning_rate": 3.7031783031014933e-06, + "loss": 0.9032, + "step": 2222 + }, + { + "epoch": 0.20047797267439238, + "grad_norm": 1.5982784188350194, + "learning_rate": 3.702871981505538e-06, + "loss": 0.9918, + "step": 2223 + }, + { + "epoch": 0.20056815619786264, + "grad_norm": 1.9519430627844074, + "learning_rate": 3.7025655146111563e-06, + "loss": 0.9574, + "step": 2224 + }, + { + "epoch": 0.2006583397213329, + "grad_norm": 0.6815531327462262, + "learning_rate": 3.702258902444497e-06, + "loss": 0.9017, + "step": 2225 + }, + { + "epoch": 0.20074852324480316, + "grad_norm": 1.4883055997844024, + "learning_rate": 3.701952145031722e-06, + "loss": 1.0684, + "step": 2226 + }, + { + "epoch": 0.20083870676827342, + "grad_norm": 1.5910111623692555, + "learning_rate": 3.701645242399005e-06, + "loss": 1.099, + "step": 2227 + }, + { + "epoch": 0.2009288902917437, + "grad_norm": 1.5760998715353387, + "learning_rate": 3.701338194572533e-06, + "loss": 1.0861, + "step": 2228 + }, + { + "epoch": 0.20101907381521397, + "grad_norm": 1.729199635141596, + "learning_rate": 3.7010310015785056e-06, + "loss": 0.9999, + "step": 2229 + }, + { + "epoch": 0.20110925733868423, + "grad_norm": 1.7859694884364903, + "learning_rate": 3.700723663443134e-06, + "loss": 0.958, + "step": 2230 + }, + { + "epoch": 0.2011994408621545, + "grad_norm": 0.6986437516479814, + "learning_rate": 3.7004161801926416e-06, + "loss": 0.8288, + "step": 2231 + }, + { + "epoch": 0.20128962438562475, + "grad_norm": 1.956211794154111, + "learning_rate": 3.7001085518532643e-06, + "loss": 1.0394, + "step": 2232 + }, + { + "epoch": 0.20137980790909502, + "grad_norm": 1.6412512900818796, + "learning_rate": 3.6998007784512515e-06, + "loss": 1.0112, + "step": 2233 + }, + { + "epoch": 0.20146999143256528, + "grad_norm": 2.2093124618204367, + "learning_rate": 3.6994928600128637e-06, + "loss": 0.8912, + "step": 2234 + }, + { + "epoch": 0.20156017495603554, + "grad_norm": 1.86052777114149, + "learning_rate": 3.6991847965643742e-06, + "loss": 0.9822, + "step": 2235 + }, + { + "epoch": 0.2016503584795058, + "grad_norm": 1.8574559542541025, + "learning_rate": 3.698876588132068e-06, + "loss": 0.9701, + "step": 2236 + }, + { + "epoch": 0.20174054200297606, + "grad_norm": 1.6050925452150209, + "learning_rate": 3.6985682347422446e-06, + "loss": 0.9758, + "step": 2237 + }, + { + "epoch": 0.20183072552644632, + "grad_norm": 2.5032743945169864, + "learning_rate": 3.698259736421213e-06, + "loss": 1.0273, + "step": 2238 + }, + { + "epoch": 0.20192090904991658, + "grad_norm": 1.820273141067079, + "learning_rate": 3.697951093195297e-06, + "loss": 0.995, + "step": 2239 + }, + { + "epoch": 0.20201109257338684, + "grad_norm": 5.19467016852755, + "learning_rate": 3.6976423050908307e-06, + "loss": 0.9936, + "step": 2240 + }, + { + "epoch": 0.2021012760968571, + "grad_norm": 2.4078495604517784, + "learning_rate": 3.697333372134163e-06, + "loss": 1.0037, + "step": 2241 + }, + { + "epoch": 0.20219145962032736, + "grad_norm": 1.8898995008441628, + "learning_rate": 3.697024294351653e-06, + "loss": 1.1088, + "step": 2242 + }, + { + "epoch": 0.20228164314379762, + "grad_norm": 1.5319082441267369, + "learning_rate": 3.696715071769672e-06, + "loss": 0.9744, + "step": 2243 + }, + { + "epoch": 0.20237182666726788, + "grad_norm": 2.190826176046078, + "learning_rate": 3.696405704414606e-06, + "loss": 1.0155, + "step": 2244 + }, + { + "epoch": 0.20246201019073815, + "grad_norm": 1.6788342104971083, + "learning_rate": 3.6960961923128514e-06, + "loss": 1.1185, + "step": 2245 + }, + { + "epoch": 0.2025521937142084, + "grad_norm": 2.366744500433976, + "learning_rate": 3.6957865354908177e-06, + "loss": 0.9505, + "step": 2246 + }, + { + "epoch": 0.20264237723767867, + "grad_norm": 1.708685660277638, + "learning_rate": 3.6954767339749262e-06, + "loss": 1.0354, + "step": 2247 + }, + { + "epoch": 0.20273256076114893, + "grad_norm": 2.317953123959906, + "learning_rate": 3.6951667877916113e-06, + "loss": 0.9942, + "step": 2248 + }, + { + "epoch": 0.2028227442846192, + "grad_norm": 1.7221334273852096, + "learning_rate": 3.694856696967319e-06, + "loss": 1.0215, + "step": 2249 + }, + { + "epoch": 0.20291292780808945, + "grad_norm": 2.1240444278776045, + "learning_rate": 3.6945464615285077e-06, + "loss": 1.0024, + "step": 2250 + }, + { + "epoch": 0.2030031113315597, + "grad_norm": 0.742512240389233, + "learning_rate": 3.694236081501648e-06, + "loss": 0.9295, + "step": 2251 + }, + { + "epoch": 0.20309329485503, + "grad_norm": 1.637666487056247, + "learning_rate": 3.6939255569132246e-06, + "loss": 0.9752, + "step": 2252 + }, + { + "epoch": 0.20318347837850026, + "grad_norm": 0.826772160611732, + "learning_rate": 3.693614887789733e-06, + "loss": 0.8574, + "step": 2253 + }, + { + "epoch": 0.20327366190197052, + "grad_norm": 0.7813121770911957, + "learning_rate": 3.69330407415768e-06, + "loss": 0.8581, + "step": 2254 + }, + { + "epoch": 0.20336384542544078, + "grad_norm": 2.142639882563404, + "learning_rate": 3.6929931160435867e-06, + "loss": 1.0223, + "step": 2255 + }, + { + "epoch": 0.20345402894891104, + "grad_norm": 1.8405147551857324, + "learning_rate": 3.6926820134739858e-06, + "loss": 1.0739, + "step": 2256 + }, + { + "epoch": 0.2035442124723813, + "grad_norm": 1.6914757412561543, + "learning_rate": 3.692370766475422e-06, + "loss": 0.9465, + "step": 2257 + }, + { + "epoch": 0.20363439599585156, + "grad_norm": 2.170567949392718, + "learning_rate": 3.692059375074453e-06, + "loss": 1.1011, + "step": 2258 + }, + { + "epoch": 0.20372457951932182, + "grad_norm": 1.712724075136769, + "learning_rate": 3.6917478392976475e-06, + "loss": 1.0833, + "step": 2259 + }, + { + "epoch": 0.20381476304279209, + "grad_norm": 1.491920726955903, + "learning_rate": 3.691436159171589e-06, + "loss": 1.0738, + "step": 2260 + }, + { + "epoch": 0.20390494656626235, + "grad_norm": 1.8592212517267315, + "learning_rate": 3.6911243347228703e-06, + "loss": 1.001, + "step": 2261 + }, + { + "epoch": 0.2039951300897326, + "grad_norm": 0.633135324715935, + "learning_rate": 3.690812365978099e-06, + "loss": 0.7705, + "step": 2262 + }, + { + "epoch": 0.20408531361320287, + "grad_norm": 1.5091104405905533, + "learning_rate": 3.690500252963893e-06, + "loss": 0.9742, + "step": 2263 + }, + { + "epoch": 0.20417549713667313, + "grad_norm": 1.9216841316036057, + "learning_rate": 3.6901879957068846e-06, + "loss": 1.0032, + "step": 2264 + }, + { + "epoch": 0.2042656806601434, + "grad_norm": 1.667487260349456, + "learning_rate": 3.689875594233717e-06, + "loss": 1.1038, + "step": 2265 + }, + { + "epoch": 0.20435586418361365, + "grad_norm": 1.8023508641400627, + "learning_rate": 3.689563048571046e-06, + "loss": 0.9983, + "step": 2266 + }, + { + "epoch": 0.2044460477070839, + "grad_norm": 3.3364691924792713, + "learning_rate": 3.6892503587455395e-06, + "loss": 1.0125, + "step": 2267 + }, + { + "epoch": 0.20453623123055417, + "grad_norm": 2.531725144237252, + "learning_rate": 3.6889375247838766e-06, + "loss": 1.0252, + "step": 2268 + }, + { + "epoch": 0.20462641475402443, + "grad_norm": 1.9322564326635905, + "learning_rate": 3.688624546712753e-06, + "loss": 1.0134, + "step": 2269 + }, + { + "epoch": 0.2047165982774947, + "grad_norm": 3.0380890765613806, + "learning_rate": 3.688311424558871e-06, + "loss": 0.9681, + "step": 2270 + }, + { + "epoch": 0.20480678180096495, + "grad_norm": 2.2201514317870115, + "learning_rate": 3.6879981583489496e-06, + "loss": 0.9157, + "step": 2271 + }, + { + "epoch": 0.20489696532443522, + "grad_norm": 1.642922527481114, + "learning_rate": 3.687684748109718e-06, + "loss": 1.0298, + "step": 2272 + }, + { + "epoch": 0.20498714884790548, + "grad_norm": 2.451413182998843, + "learning_rate": 3.6873711938679174e-06, + "loss": 1.0777, + "step": 2273 + }, + { + "epoch": 0.20507733237137574, + "grad_norm": 2.1941159065461067, + "learning_rate": 3.6870574956503027e-06, + "loss": 1.0049, + "step": 2274 + }, + { + "epoch": 0.20516751589484603, + "grad_norm": 2.3262472068445983, + "learning_rate": 3.68674365348364e-06, + "loss": 1.0038, + "step": 2275 + }, + { + "epoch": 0.2052576994183163, + "grad_norm": 1.952803501371521, + "learning_rate": 3.6864296673947086e-06, + "loss": 1.0178, + "step": 2276 + }, + { + "epoch": 0.20534788294178655, + "grad_norm": 1.9736973196564718, + "learning_rate": 3.686115537410298e-06, + "loss": 0.9987, + "step": 2277 + }, + { + "epoch": 0.2054380664652568, + "grad_norm": 1.6663886633122467, + "learning_rate": 3.685801263557214e-06, + "loss": 0.962, + "step": 2278 + }, + { + "epoch": 0.20552824998872707, + "grad_norm": 2.1209895493008464, + "learning_rate": 3.68548684586227e-06, + "loss": 1.1072, + "step": 2279 + }, + { + "epoch": 0.20561843351219733, + "grad_norm": 2.098397065523679, + "learning_rate": 3.685172284352295e-06, + "loss": 1.0145, + "step": 2280 + }, + { + "epoch": 0.2057086170356676, + "grad_norm": 1.3178695559978524, + "learning_rate": 3.684857579054128e-06, + "loss": 0.9496, + "step": 2281 + }, + { + "epoch": 0.20579880055913785, + "grad_norm": 2.082313962139275, + "learning_rate": 3.6845427299946233e-06, + "loss": 1.0768, + "step": 2282 + }, + { + "epoch": 0.2058889840826081, + "grad_norm": 1.7869899021924325, + "learning_rate": 3.6842277372006434e-06, + "loss": 1.1121, + "step": 2283 + }, + { + "epoch": 0.20597916760607837, + "grad_norm": 1.362899289550529, + "learning_rate": 3.6839126006990664e-06, + "loss": 0.9847, + "step": 2284 + }, + { + "epoch": 0.20606935112954863, + "grad_norm": 1.743655351920886, + "learning_rate": 3.6835973205167818e-06, + "loss": 1.0679, + "step": 2285 + }, + { + "epoch": 0.2061595346530189, + "grad_norm": 2.2204916122748823, + "learning_rate": 3.6832818966806904e-06, + "loss": 1.0492, + "step": 2286 + }, + { + "epoch": 0.20624971817648916, + "grad_norm": 1.742049207458815, + "learning_rate": 3.682966329217706e-06, + "loss": 1.0974, + "step": 2287 + }, + { + "epoch": 0.20633990169995942, + "grad_norm": 1.8117682854450934, + "learning_rate": 3.6826506181547543e-06, + "loss": 1.0462, + "step": 2288 + }, + { + "epoch": 0.20643008522342968, + "grad_norm": 1.8868013920662206, + "learning_rate": 3.682334763518774e-06, + "loss": 1.0349, + "step": 2289 + }, + { + "epoch": 0.20652026874689994, + "grad_norm": 1.7863402332235054, + "learning_rate": 3.6820187653367158e-06, + "loss": 0.9753, + "step": 2290 + }, + { + "epoch": 0.2066104522703702, + "grad_norm": 1.778744793024437, + "learning_rate": 3.6817026236355412e-06, + "loss": 0.9058, + "step": 2291 + }, + { + "epoch": 0.20670063579384046, + "grad_norm": 2.260393614292525, + "learning_rate": 3.681386338442227e-06, + "loss": 0.9937, + "step": 2292 + }, + { + "epoch": 0.20679081931731072, + "grad_norm": 1.4593050398792962, + "learning_rate": 3.681069909783758e-06, + "loss": 0.9781, + "step": 2293 + }, + { + "epoch": 0.20688100284078098, + "grad_norm": 1.6208224517666199, + "learning_rate": 3.680753337687136e-06, + "loss": 1.0231, + "step": 2294 + }, + { + "epoch": 0.20697118636425124, + "grad_norm": 2.211443192844488, + "learning_rate": 3.680436622179371e-06, + "loss": 0.9417, + "step": 2295 + }, + { + "epoch": 0.2070613698877215, + "grad_norm": 2.056630191613493, + "learning_rate": 3.680119763287488e-06, + "loss": 0.9961, + "step": 2296 + }, + { + "epoch": 0.20715155341119176, + "grad_norm": 4.543873848466884, + "learning_rate": 3.6798027610385227e-06, + "loss": 0.9534, + "step": 2297 + }, + { + "epoch": 0.20724173693466202, + "grad_norm": 2.2690945072120536, + "learning_rate": 3.6794856154595235e-06, + "loss": 1.0628, + "step": 2298 + }, + { + "epoch": 0.2073319204581323, + "grad_norm": 1.8958145671656268, + "learning_rate": 3.6791683265775506e-06, + "loss": 1.0241, + "step": 2299 + }, + { + "epoch": 0.20742210398160257, + "grad_norm": 1.9901883734870152, + "learning_rate": 3.6788508944196773e-06, + "loss": 0.9694, + "step": 2300 + }, + { + "epoch": 0.20751228750507283, + "grad_norm": 1.397845009541499, + "learning_rate": 3.678533319012989e-06, + "loss": 1.0918, + "step": 2301 + }, + { + "epoch": 0.2076024710285431, + "grad_norm": 2.022386314676856, + "learning_rate": 3.6782156003845826e-06, + "loss": 0.9641, + "step": 2302 + }, + { + "epoch": 0.20769265455201336, + "grad_norm": 0.662335270068974, + "learning_rate": 3.6778977385615676e-06, + "loss": 0.8327, + "step": 2303 + }, + { + "epoch": 0.20778283807548362, + "grad_norm": 1.6848496003762654, + "learning_rate": 3.6775797335710656e-06, + "loss": 1.0651, + "step": 2304 + }, + { + "epoch": 0.20787302159895388, + "grad_norm": 2.24586700621812, + "learning_rate": 3.6772615854402105e-06, + "loss": 0.9653, + "step": 2305 + }, + { + "epoch": 0.20796320512242414, + "grad_norm": 1.633196844833843, + "learning_rate": 3.6769432941961487e-06, + "loss": 0.894, + "step": 2306 + }, + { + "epoch": 0.2080533886458944, + "grad_norm": 1.563760859072859, + "learning_rate": 3.676624859866038e-06, + "loss": 1.0136, + "step": 2307 + }, + { + "epoch": 0.20814357216936466, + "grad_norm": 3.2851440574610247, + "learning_rate": 3.67630628247705e-06, + "loss": 0.9351, + "step": 2308 + }, + { + "epoch": 0.20823375569283492, + "grad_norm": 2.306014787624882, + "learning_rate": 3.675987562056367e-06, + "loss": 1.0778, + "step": 2309 + }, + { + "epoch": 0.20832393921630518, + "grad_norm": 1.621470212311673, + "learning_rate": 3.675668698631184e-06, + "loss": 1.0511, + "step": 2310 + }, + { + "epoch": 0.20841412273977544, + "grad_norm": 1.6012578162688191, + "learning_rate": 3.675349692228708e-06, + "loss": 1.0778, + "step": 2311 + }, + { + "epoch": 0.2085043062632457, + "grad_norm": 1.8223579931205427, + "learning_rate": 3.6750305428761578e-06, + "loss": 0.988, + "step": 2312 + }, + { + "epoch": 0.20859448978671596, + "grad_norm": 2.1540994285651096, + "learning_rate": 3.674711250600766e-06, + "loss": 1.0225, + "step": 2313 + }, + { + "epoch": 0.20868467331018623, + "grad_norm": 0.6683572007750881, + "learning_rate": 3.6743918154297765e-06, + "loss": 0.814, + "step": 2314 + }, + { + "epoch": 0.20877485683365649, + "grad_norm": 1.8877474111125756, + "learning_rate": 3.6740722373904446e-06, + "loss": 1.0009, + "step": 2315 + }, + { + "epoch": 0.20886504035712675, + "grad_norm": 1.457837247485133, + "learning_rate": 3.6737525165100383e-06, + "loss": 0.9902, + "step": 2316 + }, + { + "epoch": 0.208955223880597, + "grad_norm": 1.8096795334769906, + "learning_rate": 3.6734326528158385e-06, + "loss": 1.0369, + "step": 2317 + }, + { + "epoch": 0.20904540740406727, + "grad_norm": 1.5261058424994598, + "learning_rate": 3.673112646335138e-06, + "loss": 0.9998, + "step": 2318 + }, + { + "epoch": 0.20913559092753753, + "grad_norm": 1.5638794722347908, + "learning_rate": 3.672792497095241e-06, + "loss": 0.9581, + "step": 2319 + }, + { + "epoch": 0.2092257744510078, + "grad_norm": 1.900157949507883, + "learning_rate": 3.672472205123464e-06, + "loss": 1.0104, + "step": 2320 + }, + { + "epoch": 0.20931595797447805, + "grad_norm": 3.2865750654299024, + "learning_rate": 3.6721517704471363e-06, + "loss": 0.9417, + "step": 2321 + }, + { + "epoch": 0.2094061414979483, + "grad_norm": 1.6822909029396929, + "learning_rate": 3.6718311930936e-06, + "loss": 1.0503, + "step": 2322 + }, + { + "epoch": 0.2094963250214186, + "grad_norm": 1.8021668504828763, + "learning_rate": 3.6715104730902074e-06, + "loss": 0.9687, + "step": 2323 + }, + { + "epoch": 0.20958650854488886, + "grad_norm": 17.019858934331133, + "learning_rate": 3.671189610464325e-06, + "loss": 1.0502, + "step": 2324 + }, + { + "epoch": 0.20967669206835912, + "grad_norm": 1.7585854420408489, + "learning_rate": 3.6708686052433303e-06, + "loss": 0.9628, + "step": 2325 + }, + { + "epoch": 0.20976687559182938, + "grad_norm": 1.8455188477576134, + "learning_rate": 3.6705474574546127e-06, + "loss": 1.0287, + "step": 2326 + }, + { + "epoch": 0.20985705911529964, + "grad_norm": 1.8657062992036837, + "learning_rate": 3.670226167125575e-06, + "loss": 1.0358, + "step": 2327 + }, + { + "epoch": 0.2099472426387699, + "grad_norm": 2.002367049434039, + "learning_rate": 3.6699047342836313e-06, + "loss": 1.0196, + "step": 2328 + }, + { + "epoch": 0.21003742616224017, + "grad_norm": 1.5830497571411313, + "learning_rate": 3.669583158956208e-06, + "loss": 1.0486, + "step": 2329 + }, + { + "epoch": 0.21012760968571043, + "grad_norm": 0.6810466069014796, + "learning_rate": 3.669261441170743e-06, + "loss": 0.8445, + "step": 2330 + }, + { + "epoch": 0.2102177932091807, + "grad_norm": 1.7034309182453666, + "learning_rate": 3.668939580954688e-06, + "loss": 1.0252, + "step": 2331 + }, + { + "epoch": 0.21030797673265095, + "grad_norm": 1.5436768271212684, + "learning_rate": 3.668617578335506e-06, + "loss": 0.9926, + "step": 2332 + }, + { + "epoch": 0.2103981602561212, + "grad_norm": 1.9409637253843919, + "learning_rate": 3.6682954333406707e-06, + "loss": 1.0445, + "step": 2333 + }, + { + "epoch": 0.21048834377959147, + "grad_norm": 1.8371411862012972, + "learning_rate": 3.6679731459976707e-06, + "loss": 1.1421, + "step": 2334 + }, + { + "epoch": 0.21057852730306173, + "grad_norm": 1.5340823234292165, + "learning_rate": 3.6676507163340046e-06, + "loss": 1.0112, + "step": 2335 + }, + { + "epoch": 0.210668710826532, + "grad_norm": 2.2946762667774125, + "learning_rate": 3.6673281443771842e-06, + "loss": 1.0445, + "step": 2336 + }, + { + "epoch": 0.21075889435000225, + "grad_norm": 1.5029502307065377, + "learning_rate": 3.667005430154733e-06, + "loss": 0.9051, + "step": 2337 + }, + { + "epoch": 0.2108490778734725, + "grad_norm": 0.6598318758959122, + "learning_rate": 3.666682573694186e-06, + "loss": 0.9183, + "step": 2338 + }, + { + "epoch": 0.21093926139694277, + "grad_norm": 0.6810042900602907, + "learning_rate": 3.6663595750230924e-06, + "loss": 0.7876, + "step": 2339 + }, + { + "epoch": 0.21102944492041303, + "grad_norm": 1.771179913283031, + "learning_rate": 3.666036434169012e-06, + "loss": 1.0682, + "step": 2340 + }, + { + "epoch": 0.2111196284438833, + "grad_norm": 1.5469636988014768, + "learning_rate": 3.665713151159516e-06, + "loss": 0.9963, + "step": 2341 + }, + { + "epoch": 0.21120981196735356, + "grad_norm": 2.093125093493683, + "learning_rate": 3.665389726022189e-06, + "loss": 1.1128, + "step": 2342 + }, + { + "epoch": 0.21129999549082382, + "grad_norm": 1.7343721131996668, + "learning_rate": 3.6650661587846283e-06, + "loss": 0.9932, + "step": 2343 + }, + { + "epoch": 0.21139017901429408, + "grad_norm": 1.7889272980222681, + "learning_rate": 3.6647424494744418e-06, + "loss": 0.981, + "step": 2344 + }, + { + "epoch": 0.21148036253776434, + "grad_norm": 1.514806860897099, + "learning_rate": 3.6644185981192503e-06, + "loss": 1.0653, + "step": 2345 + }, + { + "epoch": 0.2115705460612346, + "grad_norm": 6.432365962743577, + "learning_rate": 3.6640946047466868e-06, + "loss": 0.9383, + "step": 2346 + }, + { + "epoch": 0.2116607295847049, + "grad_norm": 1.7126019886025348, + "learning_rate": 3.6637704693843953e-06, + "loss": 0.9825, + "step": 2347 + }, + { + "epoch": 0.21175091310817515, + "grad_norm": 1.5892673356660572, + "learning_rate": 3.6634461920600337e-06, + "loss": 0.9465, + "step": 2348 + }, + { + "epoch": 0.2118410966316454, + "grad_norm": 2.1594295555543392, + "learning_rate": 3.66312177280127e-06, + "loss": 1.0365, + "step": 2349 + }, + { + "epoch": 0.21193128015511567, + "grad_norm": 1.665425776435319, + "learning_rate": 3.6627972116357872e-06, + "loss": 1.0539, + "step": 2350 + }, + { + "epoch": 0.21202146367858593, + "grad_norm": 1.9269659333080809, + "learning_rate": 3.662472508591278e-06, + "loss": 1.0787, + "step": 2351 + }, + { + "epoch": 0.2121116472020562, + "grad_norm": 2.130726335465042, + "learning_rate": 3.662147663695447e-06, + "loss": 1.0016, + "step": 2352 + }, + { + "epoch": 0.21220183072552645, + "grad_norm": 1.4318339593301537, + "learning_rate": 3.6618226769760127e-06, + "loss": 1.004, + "step": 2353 + }, + { + "epoch": 0.2122920142489967, + "grad_norm": 1.6538466524352002, + "learning_rate": 3.661497548460704e-06, + "loss": 0.8944, + "step": 2354 + }, + { + "epoch": 0.21238219777246697, + "grad_norm": 1.9820796276499208, + "learning_rate": 3.6611722781772635e-06, + "loss": 1.0303, + "step": 2355 + }, + { + "epoch": 0.21247238129593723, + "grad_norm": 0.8234537922877896, + "learning_rate": 3.6608468661534444e-06, + "loss": 0.882, + "step": 2356 + }, + { + "epoch": 0.2125625648194075, + "grad_norm": 2.4478178504857127, + "learning_rate": 3.660521312417013e-06, + "loss": 0.9421, + "step": 2357 + }, + { + "epoch": 0.21265274834287776, + "grad_norm": 4.0039548396441935, + "learning_rate": 3.660195616995747e-06, + "loss": 0.9866, + "step": 2358 + }, + { + "epoch": 0.21274293186634802, + "grad_norm": 1.849056108781846, + "learning_rate": 3.6598697799174367e-06, + "loss": 1.0422, + "step": 2359 + }, + { + "epoch": 0.21283311538981828, + "grad_norm": 2.0570320706016334, + "learning_rate": 3.6595438012098844e-06, + "loss": 1.0749, + "step": 2360 + }, + { + "epoch": 0.21292329891328854, + "grad_norm": 3.1038491686227374, + "learning_rate": 3.6592176809009045e-06, + "loss": 0.9871, + "step": 2361 + }, + { + "epoch": 0.2130134824367588, + "grad_norm": 1.51404371062943, + "learning_rate": 3.6588914190183227e-06, + "loss": 1.0534, + "step": 2362 + }, + { + "epoch": 0.21310366596022906, + "grad_norm": 1.7984516690091281, + "learning_rate": 3.658565015589978e-06, + "loss": 0.8879, + "step": 2363 + }, + { + "epoch": 0.21319384948369932, + "grad_norm": 3.4764673112846416, + "learning_rate": 3.6582384706437217e-06, + "loss": 1.0233, + "step": 2364 + }, + { + "epoch": 0.21328403300716958, + "grad_norm": 1.8934843651322728, + "learning_rate": 3.6579117842074156e-06, + "loss": 1.0279, + "step": 2365 + }, + { + "epoch": 0.21337421653063984, + "grad_norm": 1.6562391316759049, + "learning_rate": 3.657584956308934e-06, + "loss": 0.9508, + "step": 2366 + }, + { + "epoch": 0.2134644000541101, + "grad_norm": 2.1835844691425645, + "learning_rate": 3.6572579869761648e-06, + "loss": 1.0781, + "step": 2367 + }, + { + "epoch": 0.21355458357758036, + "grad_norm": 8.263484223303694, + "learning_rate": 3.6569308762370056e-06, + "loss": 1.0646, + "step": 2368 + }, + { + "epoch": 0.21364476710105063, + "grad_norm": 2.752439717144572, + "learning_rate": 3.6566036241193676e-06, + "loss": 1.0578, + "step": 2369 + }, + { + "epoch": 0.2137349506245209, + "grad_norm": 0.6170247261666542, + "learning_rate": 3.656276230651174e-06, + "loss": 0.8462, + "step": 2370 + }, + { + "epoch": 0.21382513414799117, + "grad_norm": 0.9392057160453758, + "learning_rate": 3.65594869586036e-06, + "loss": 0.8794, + "step": 2371 + }, + { + "epoch": 0.21391531767146144, + "grad_norm": 1.544601233541314, + "learning_rate": 3.6556210197748724e-06, + "loss": 0.9397, + "step": 2372 + }, + { + "epoch": 0.2140055011949317, + "grad_norm": 1.6677090802291288, + "learning_rate": 3.655293202422671e-06, + "loss": 0.9822, + "step": 2373 + }, + { + "epoch": 0.21409568471840196, + "grad_norm": 4.824941628849178, + "learning_rate": 3.654965243831725e-06, + "loss": 1.0274, + "step": 2374 + }, + { + "epoch": 0.21418586824187222, + "grad_norm": 1.6164740126479868, + "learning_rate": 3.65463714403002e-06, + "loss": 1.0567, + "step": 2375 + }, + { + "epoch": 0.21427605176534248, + "grad_norm": 1.5985758403676775, + "learning_rate": 3.65430890304555e-06, + "loss": 0.9705, + "step": 2376 + }, + { + "epoch": 0.21436623528881274, + "grad_norm": 1.4403364972503054, + "learning_rate": 3.653980520906323e-06, + "loss": 1.0418, + "step": 2377 + }, + { + "epoch": 0.214456418812283, + "grad_norm": 1.6103543014919475, + "learning_rate": 3.653651997640358e-06, + "loss": 0.9928, + "step": 2378 + }, + { + "epoch": 0.21454660233575326, + "grad_norm": 1.646992343724978, + "learning_rate": 3.653323333275686e-06, + "loss": 1.0192, + "step": 2379 + }, + { + "epoch": 0.21463678585922352, + "grad_norm": 1.5000994172528885, + "learning_rate": 3.652994527840351e-06, + "loss": 0.9953, + "step": 2380 + }, + { + "epoch": 0.21472696938269378, + "grad_norm": 1.7196433779924094, + "learning_rate": 3.6526655813624087e-06, + "loss": 0.9739, + "step": 2381 + }, + { + "epoch": 0.21481715290616404, + "grad_norm": 0.5814462320245527, + "learning_rate": 3.652336493869925e-06, + "loss": 0.8389, + "step": 2382 + }, + { + "epoch": 0.2149073364296343, + "grad_norm": 2.6109204312856105, + "learning_rate": 3.6520072653909823e-06, + "loss": 1.0171, + "step": 2383 + }, + { + "epoch": 0.21499751995310457, + "grad_norm": 1.9069701538550863, + "learning_rate": 3.6516778959536702e-06, + "loss": 1.0093, + "step": 2384 + }, + { + "epoch": 0.21508770347657483, + "grad_norm": 2.1187295963351422, + "learning_rate": 3.6513483855860923e-06, + "loss": 1.0627, + "step": 2385 + }, + { + "epoch": 0.2151778870000451, + "grad_norm": 1.8997993764864851, + "learning_rate": 3.6510187343163654e-06, + "loss": 0.9582, + "step": 2386 + }, + { + "epoch": 0.21526807052351535, + "grad_norm": 2.1401171673580754, + "learning_rate": 3.650688942172616e-06, + "loss": 1.0431, + "step": 2387 + }, + { + "epoch": 0.2153582540469856, + "grad_norm": 2.0357312581781715, + "learning_rate": 3.650359009182984e-06, + "loss": 0.9535, + "step": 2388 + }, + { + "epoch": 0.21544843757045587, + "grad_norm": 2.3681750362181315, + "learning_rate": 3.650028935375622e-06, + "loss": 1.096, + "step": 2389 + }, + { + "epoch": 0.21553862109392613, + "grad_norm": 2.6852278752301157, + "learning_rate": 3.6496987207786926e-06, + "loss": 0.9804, + "step": 2390 + }, + { + "epoch": 0.2156288046173964, + "grad_norm": 1.9057810472333636, + "learning_rate": 3.6493683654203724e-06, + "loss": 0.9607, + "step": 2391 + }, + { + "epoch": 0.21571898814086665, + "grad_norm": 2.8139520075920794, + "learning_rate": 3.6490378693288484e-06, + "loss": 0.9998, + "step": 2392 + }, + { + "epoch": 0.2158091716643369, + "grad_norm": 1.7154398861711875, + "learning_rate": 3.648707232532321e-06, + "loss": 1.0589, + "step": 2393 + }, + { + "epoch": 0.2158993551878072, + "grad_norm": 2.3579385186577224, + "learning_rate": 3.6483764550590017e-06, + "loss": 1.0115, + "step": 2394 + }, + { + "epoch": 0.21598953871127746, + "grad_norm": 1.8334566277168909, + "learning_rate": 3.6480455369371133e-06, + "loss": 1.029, + "step": 2395 + }, + { + "epoch": 0.21607972223474772, + "grad_norm": 6.917366850315639, + "learning_rate": 3.647714478194893e-06, + "loss": 1.018, + "step": 2396 + }, + { + "epoch": 0.21616990575821798, + "grad_norm": 2.260767919669971, + "learning_rate": 3.647383278860588e-06, + "loss": 0.956, + "step": 2397 + }, + { + "epoch": 0.21626008928168824, + "grad_norm": 1.9411409070077228, + "learning_rate": 3.6470519389624587e-06, + "loss": 1.0009, + "step": 2398 + }, + { + "epoch": 0.2163502728051585, + "grad_norm": 1.7782989953160135, + "learning_rate": 3.646720458528776e-06, + "loss": 1.0717, + "step": 2399 + }, + { + "epoch": 0.21644045632862877, + "grad_norm": 2.607691055861922, + "learning_rate": 3.6463888375878235e-06, + "loss": 0.9833, + "step": 2400 + }, + { + "epoch": 0.21653063985209903, + "grad_norm": 6.837078587797763, + "learning_rate": 3.646057076167897e-06, + "loss": 1.0235, + "step": 2401 + }, + { + "epoch": 0.2166208233755693, + "grad_norm": 1.9286289168675401, + "learning_rate": 3.645725174297305e-06, + "loss": 1.0164, + "step": 2402 + }, + { + "epoch": 0.21671100689903955, + "grad_norm": 2.2781207426054935, + "learning_rate": 3.645393132004367e-06, + "loss": 0.9897, + "step": 2403 + }, + { + "epoch": 0.2168011904225098, + "grad_norm": 3.9817417194440585, + "learning_rate": 3.6450609493174135e-06, + "loss": 0.9715, + "step": 2404 + }, + { + "epoch": 0.21689137394598007, + "grad_norm": 1.8683376201724515, + "learning_rate": 3.6447286262647896e-06, + "loss": 0.9724, + "step": 2405 + }, + { + "epoch": 0.21698155746945033, + "grad_norm": 2.734662983579246, + "learning_rate": 3.64439616287485e-06, + "loss": 0.9768, + "step": 2406 + }, + { + "epoch": 0.2170717409929206, + "grad_norm": 1.4664806934416357, + "learning_rate": 3.644063559175963e-06, + "loss": 0.9776, + "step": 2407 + }, + { + "epoch": 0.21716192451639085, + "grad_norm": 1.5701734187086902, + "learning_rate": 3.6437308151965074e-06, + "loss": 0.9929, + "step": 2408 + }, + { + "epoch": 0.2172521080398611, + "grad_norm": 1.652141297114476, + "learning_rate": 3.643397930964876e-06, + "loss": 1.0274, + "step": 2409 + }, + { + "epoch": 0.21734229156333137, + "grad_norm": 2.3533780512825175, + "learning_rate": 3.6430649065094707e-06, + "loss": 1.0831, + "step": 2410 + }, + { + "epoch": 0.21743247508680164, + "grad_norm": 1.8186407312033608, + "learning_rate": 3.6427317418587086e-06, + "loss": 1.0758, + "step": 2411 + }, + { + "epoch": 0.2175226586102719, + "grad_norm": 1.924681064887917, + "learning_rate": 3.6423984370410157e-06, + "loss": 0.952, + "step": 2412 + }, + { + "epoch": 0.21761284213374216, + "grad_norm": 1.7448735080332485, + "learning_rate": 3.6420649920848324e-06, + "loss": 0.9772, + "step": 2413 + }, + { + "epoch": 0.21770302565721242, + "grad_norm": 2.075237476593711, + "learning_rate": 3.6417314070186096e-06, + "loss": 1.0162, + "step": 2414 + }, + { + "epoch": 0.21779320918068268, + "grad_norm": 1.5900361415666113, + "learning_rate": 3.641397681870811e-06, + "loss": 1.0179, + "step": 2415 + }, + { + "epoch": 0.21788339270415294, + "grad_norm": 2.6945874994411994, + "learning_rate": 3.641063816669911e-06, + "loss": 1.0271, + "step": 2416 + }, + { + "epoch": 0.2179735762276232, + "grad_norm": 1.553291882888292, + "learning_rate": 3.640729811444398e-06, + "loss": 1.0715, + "step": 2417 + }, + { + "epoch": 0.2180637597510935, + "grad_norm": 1.87901689675506, + "learning_rate": 3.6403956662227706e-06, + "loss": 1.0127, + "step": 2418 + }, + { + "epoch": 0.21815394327456375, + "grad_norm": 1.8578459920342691, + "learning_rate": 3.6400613810335396e-06, + "loss": 1.0659, + "step": 2419 + }, + { + "epoch": 0.218244126798034, + "grad_norm": 1.8414336702994292, + "learning_rate": 3.639726955905228e-06, + "loss": 0.9313, + "step": 2420 + }, + { + "epoch": 0.21833431032150427, + "grad_norm": 1.670870755633496, + "learning_rate": 3.639392390866372e-06, + "loss": 1.0375, + "step": 2421 + }, + { + "epoch": 0.21842449384497453, + "grad_norm": 2.2842235259260515, + "learning_rate": 3.639057685945517e-06, + "loss": 0.9904, + "step": 2422 + }, + { + "epoch": 0.2185146773684448, + "grad_norm": 1.7932612693705872, + "learning_rate": 3.638722841171223e-06, + "loss": 0.9025, + "step": 2423 + }, + { + "epoch": 0.21860486089191505, + "grad_norm": 1.5457020946361395, + "learning_rate": 3.638387856572061e-06, + "loss": 1.0504, + "step": 2424 + }, + { + "epoch": 0.21869504441538531, + "grad_norm": 0.6591418446168629, + "learning_rate": 3.638052732176612e-06, + "loss": 0.8452, + "step": 2425 + }, + { + "epoch": 0.21878522793885558, + "grad_norm": 1.9166421750480322, + "learning_rate": 3.637717468013472e-06, + "loss": 1.0261, + "step": 2426 + }, + { + "epoch": 0.21887541146232584, + "grad_norm": 0.604144635840194, + "learning_rate": 3.6373820641112475e-06, + "loss": 0.7969, + "step": 2427 + }, + { + "epoch": 0.2189655949857961, + "grad_norm": 2.480703844358598, + "learning_rate": 3.6370465204985567e-06, + "loss": 1.0304, + "step": 2428 + }, + { + "epoch": 0.21905577850926636, + "grad_norm": 0.6278855707264891, + "learning_rate": 3.6367108372040304e-06, + "loss": 0.8576, + "step": 2429 + }, + { + "epoch": 0.21914596203273662, + "grad_norm": 1.7496539182415944, + "learning_rate": 3.6363750142563107e-06, + "loss": 1.0164, + "step": 2430 + }, + { + "epoch": 0.21923614555620688, + "grad_norm": 0.7244599386187076, + "learning_rate": 3.636039051684052e-06, + "loss": 0.8937, + "step": 2431 + }, + { + "epoch": 0.21932632907967714, + "grad_norm": 2.792416818143162, + "learning_rate": 3.6357029495159203e-06, + "loss": 0.9814, + "step": 2432 + }, + { + "epoch": 0.2194165126031474, + "grad_norm": 1.7698844910771356, + "learning_rate": 3.6353667077805934e-06, + "loss": 0.9889, + "step": 2433 + }, + { + "epoch": 0.21950669612661766, + "grad_norm": 1.9514706128053332, + "learning_rate": 3.6350303265067625e-06, + "loss": 1.0493, + "step": 2434 + }, + { + "epoch": 0.21959687965008792, + "grad_norm": 1.6070565866597357, + "learning_rate": 3.6346938057231285e-06, + "loss": 1.0462, + "step": 2435 + }, + { + "epoch": 0.21968706317355818, + "grad_norm": 1.6268165411981217, + "learning_rate": 3.6343571454584047e-06, + "loss": 0.9601, + "step": 2436 + }, + { + "epoch": 0.21977724669702844, + "grad_norm": 2.3146221630319874, + "learning_rate": 3.6340203457413176e-06, + "loss": 1.0551, + "step": 2437 + }, + { + "epoch": 0.2198674302204987, + "grad_norm": 2.264174424505533, + "learning_rate": 3.633683406600605e-06, + "loss": 1.0426, + "step": 2438 + }, + { + "epoch": 0.21995761374396897, + "grad_norm": 2.1048452837846905, + "learning_rate": 3.6333463280650165e-06, + "loss": 1.0766, + "step": 2439 + }, + { + "epoch": 0.22004779726743923, + "grad_norm": 0.7272946350363441, + "learning_rate": 3.6330091101633126e-06, + "loss": 0.8061, + "step": 2440 + }, + { + "epoch": 0.2201379807909095, + "grad_norm": 2.2095382090586346, + "learning_rate": 3.632671752924267e-06, + "loss": 1.0477, + "step": 2441 + }, + { + "epoch": 0.22022816431437978, + "grad_norm": 2.228318442230951, + "learning_rate": 3.632334256376665e-06, + "loss": 1.0198, + "step": 2442 + }, + { + "epoch": 0.22031834783785004, + "grad_norm": 3.395401925297499, + "learning_rate": 3.6319966205493044e-06, + "loss": 1.0727, + "step": 2443 + }, + { + "epoch": 0.2204085313613203, + "grad_norm": 0.6940875675530112, + "learning_rate": 3.6316588454709922e-06, + "loss": 0.8896, + "step": 2444 + }, + { + "epoch": 0.22049871488479056, + "grad_norm": 1.597412613103858, + "learning_rate": 3.6313209311705514e-06, + "loss": 0.9647, + "step": 2445 + }, + { + "epoch": 0.22058889840826082, + "grad_norm": 2.2638077386562885, + "learning_rate": 3.6309828776768133e-06, + "loss": 1.0543, + "step": 2446 + }, + { + "epoch": 0.22067908193173108, + "grad_norm": 2.898926891540654, + "learning_rate": 3.630644685018623e-06, + "loss": 0.9776, + "step": 2447 + }, + { + "epoch": 0.22076926545520134, + "grad_norm": 2.6743490799737413, + "learning_rate": 3.6303063532248367e-06, + "loss": 1.1006, + "step": 2448 + }, + { + "epoch": 0.2208594489786716, + "grad_norm": 1.6840677324342235, + "learning_rate": 3.6299678823243236e-06, + "loss": 1.0961, + "step": 2449 + }, + { + "epoch": 0.22094963250214186, + "grad_norm": 2.101536747090409, + "learning_rate": 3.629629272345963e-06, + "loss": 0.9225, + "step": 2450 + }, + { + "epoch": 0.22103981602561212, + "grad_norm": 1.992938150838665, + "learning_rate": 3.6292905233186468e-06, + "loss": 1.1345, + "step": 2451 + }, + { + "epoch": 0.22112999954908238, + "grad_norm": 1.6646146619360722, + "learning_rate": 3.6289516352712796e-06, + "loss": 0.9761, + "step": 2452 + }, + { + "epoch": 0.22122018307255265, + "grad_norm": 0.7101554341830865, + "learning_rate": 3.6286126082327764e-06, + "loss": 0.8407, + "step": 2453 + }, + { + "epoch": 0.2213103665960229, + "grad_norm": 1.9433106076049718, + "learning_rate": 3.628273442232066e-06, + "loss": 1.0475, + "step": 2454 + }, + { + "epoch": 0.22140055011949317, + "grad_norm": 1.501668955257477, + "learning_rate": 3.627934137298087e-06, + "loss": 1.0127, + "step": 2455 + }, + { + "epoch": 0.22149073364296343, + "grad_norm": 2.0779664402180233, + "learning_rate": 3.627594693459792e-06, + "loss": 0.9915, + "step": 2456 + }, + { + "epoch": 0.2215809171664337, + "grad_norm": 1.7865382214655252, + "learning_rate": 3.6272551107461424e-06, + "loss": 0.9764, + "step": 2457 + }, + { + "epoch": 0.22167110068990395, + "grad_norm": 1.8986695073277051, + "learning_rate": 3.6269153891861137e-06, + "loss": 1.011, + "step": 2458 + }, + { + "epoch": 0.2217612842133742, + "grad_norm": 2.4466205060209436, + "learning_rate": 3.6265755288086944e-06, + "loss": 0.9822, + "step": 2459 + }, + { + "epoch": 0.22185146773684447, + "grad_norm": 2.070365242916019, + "learning_rate": 3.626235529642881e-06, + "loss": 0.9946, + "step": 2460 + }, + { + "epoch": 0.22194165126031473, + "grad_norm": 2.460942053412206, + "learning_rate": 3.625895391717686e-06, + "loss": 0.9519, + "step": 2461 + }, + { + "epoch": 0.222031834783785, + "grad_norm": 2.7644969988989656, + "learning_rate": 3.625555115062131e-06, + "loss": 0.9833, + "step": 2462 + }, + { + "epoch": 0.22212201830725525, + "grad_norm": 1.6540240318369885, + "learning_rate": 3.6252146997052507e-06, + "loss": 1.0823, + "step": 2463 + }, + { + "epoch": 0.22221220183072551, + "grad_norm": 0.6656308223129445, + "learning_rate": 3.6248741456760898e-06, + "loss": 0.7977, + "step": 2464 + }, + { + "epoch": 0.22230238535419578, + "grad_norm": 1.8881827730029572, + "learning_rate": 3.624533453003708e-06, + "loss": 0.9999, + "step": 2465 + }, + { + "epoch": 0.22239256887766606, + "grad_norm": 1.65785362775088, + "learning_rate": 3.6241926217171745e-06, + "loss": 1.0209, + "step": 2466 + }, + { + "epoch": 0.22248275240113632, + "grad_norm": 1.9324474289073745, + "learning_rate": 3.6238516518455703e-06, + "loss": 0.931, + "step": 2467 + }, + { + "epoch": 0.22257293592460659, + "grad_norm": 3.1886972908858615, + "learning_rate": 3.62351054341799e-06, + "loss": 1.1379, + "step": 2468 + }, + { + "epoch": 0.22266311944807685, + "grad_norm": 1.691882975925547, + "learning_rate": 3.623169296463538e-06, + "loss": 1.0442, + "step": 2469 + }, + { + "epoch": 0.2227533029715471, + "grad_norm": 2.0627272509600068, + "learning_rate": 3.6228279110113316e-06, + "loss": 1.0134, + "step": 2470 + }, + { + "epoch": 0.22284348649501737, + "grad_norm": 2.0946819451833374, + "learning_rate": 3.6224863870904994e-06, + "loss": 1.0254, + "step": 2471 + }, + { + "epoch": 0.22293367001848763, + "grad_norm": 3.6625464582914256, + "learning_rate": 3.6221447247301827e-06, + "loss": 1.0316, + "step": 2472 + }, + { + "epoch": 0.2230238535419579, + "grad_norm": 1.6204388502071723, + "learning_rate": 3.6218029239595332e-06, + "loss": 0.9626, + "step": 2473 + }, + { + "epoch": 0.22311403706542815, + "grad_norm": 2.3705401457754633, + "learning_rate": 3.621460984807716e-06, + "loss": 1.0443, + "step": 2474 + }, + { + "epoch": 0.2232042205888984, + "grad_norm": 1.9429267440077524, + "learning_rate": 3.621118907303907e-06, + "loss": 0.9896, + "step": 2475 + }, + { + "epoch": 0.22329440411236867, + "grad_norm": 1.9848244225708827, + "learning_rate": 3.620776691477294e-06, + "loss": 1.0132, + "step": 2476 + }, + { + "epoch": 0.22338458763583893, + "grad_norm": 0.6701005747377783, + "learning_rate": 3.6204343373570765e-06, + "loss": 0.7877, + "step": 2477 + }, + { + "epoch": 0.2234747711593092, + "grad_norm": 1.810019143567091, + "learning_rate": 3.620091844972467e-06, + "loss": 1.0374, + "step": 2478 + }, + { + "epoch": 0.22356495468277945, + "grad_norm": 1.7017250957165795, + "learning_rate": 3.619749214352688e-06, + "loss": 1.0402, + "step": 2479 + }, + { + "epoch": 0.22365513820624972, + "grad_norm": 1.8025065533311206, + "learning_rate": 3.6194064455269744e-06, + "loss": 1.1168, + "step": 2480 + }, + { + "epoch": 0.22374532172971998, + "grad_norm": 4.649484109461055, + "learning_rate": 3.6190635385245737e-06, + "loss": 0.935, + "step": 2481 + }, + { + "epoch": 0.22383550525319024, + "grad_norm": 1.664490120975035, + "learning_rate": 3.618720493374745e-06, + "loss": 0.9524, + "step": 2482 + }, + { + "epoch": 0.2239256887766605, + "grad_norm": 1.720221288551474, + "learning_rate": 3.6183773101067575e-06, + "loss": 0.8746, + "step": 2483 + }, + { + "epoch": 0.22401587230013076, + "grad_norm": 1.6022400702045152, + "learning_rate": 3.6180339887498948e-06, + "loss": 0.9367, + "step": 2484 + }, + { + "epoch": 0.22410605582360102, + "grad_norm": 2.018287851444348, + "learning_rate": 3.61769052933345e-06, + "loss": 1.0195, + "step": 2485 + }, + { + "epoch": 0.22419623934707128, + "grad_norm": 2.5010149802728234, + "learning_rate": 3.6173469318867297e-06, + "loss": 1.1693, + "step": 2486 + }, + { + "epoch": 0.22428642287054154, + "grad_norm": 1.8211783122325746, + "learning_rate": 3.617003196439051e-06, + "loss": 1.0364, + "step": 2487 + }, + { + "epoch": 0.2243766063940118, + "grad_norm": 1.8152031964981294, + "learning_rate": 3.616659323019744e-06, + "loss": 1.0234, + "step": 2488 + }, + { + "epoch": 0.22446678991748206, + "grad_norm": 2.7195449193157795, + "learning_rate": 3.616315311658149e-06, + "loss": 1.0443, + "step": 2489 + }, + { + "epoch": 0.22455697344095235, + "grad_norm": 1.643827455385517, + "learning_rate": 3.6159711623836195e-06, + "loss": 1.0726, + "step": 2490 + }, + { + "epoch": 0.2246471569644226, + "grad_norm": 0.6481680252670419, + "learning_rate": 3.6156268752255203e-06, + "loss": 0.802, + "step": 2491 + }, + { + "epoch": 0.22473734048789287, + "grad_norm": 51.01447644189466, + "learning_rate": 3.615282450213227e-06, + "loss": 0.9794, + "step": 2492 + }, + { + "epoch": 0.22482752401136313, + "grad_norm": 1.625769139711736, + "learning_rate": 3.614937887376128e-06, + "loss": 1.0278, + "step": 2493 + }, + { + "epoch": 0.2249177075348334, + "grad_norm": 1.808354009636075, + "learning_rate": 3.614593186743625e-06, + "loss": 1.0319, + "step": 2494 + }, + { + "epoch": 0.22500789105830366, + "grad_norm": 1.8901655411253444, + "learning_rate": 3.614248348345128e-06, + "loss": 0.9521, + "step": 2495 + }, + { + "epoch": 0.22509807458177392, + "grad_norm": 1.8186401412661832, + "learning_rate": 3.6139033722100614e-06, + "loss": 1.0667, + "step": 2496 + }, + { + "epoch": 0.22518825810524418, + "grad_norm": 2.733376718579978, + "learning_rate": 3.6135582583678596e-06, + "loss": 1.0948, + "step": 2497 + }, + { + "epoch": 0.22527844162871444, + "grad_norm": 1.7219680484062576, + "learning_rate": 3.61321300684797e-06, + "loss": 1.0332, + "step": 2498 + }, + { + "epoch": 0.2253686251521847, + "grad_norm": 1.312502497716071, + "learning_rate": 3.6128676176798527e-06, + "loss": 1.0412, + "step": 2499 + }, + { + "epoch": 0.22545880867565496, + "grad_norm": 1.5092534902878862, + "learning_rate": 3.612522090892976e-06, + "loss": 1.0348, + "step": 2500 + }, + { + "epoch": 0.22554899219912522, + "grad_norm": 1.6144047628118845, + "learning_rate": 3.6121764265168232e-06, + "loss": 0.9818, + "step": 2501 + }, + { + "epoch": 0.22563917572259548, + "grad_norm": 1.610948182596923, + "learning_rate": 3.611830624580888e-06, + "loss": 1.0433, + "step": 2502 + }, + { + "epoch": 0.22572935924606574, + "grad_norm": 2.0083150865456654, + "learning_rate": 3.6114846851146767e-06, + "loss": 1.0251, + "step": 2503 + }, + { + "epoch": 0.225819542769536, + "grad_norm": 2.212814558696503, + "learning_rate": 3.6111386081477068e-06, + "loss": 0.9208, + "step": 2504 + }, + { + "epoch": 0.22590972629300626, + "grad_norm": 2.6772544202548114, + "learning_rate": 3.6107923937095066e-06, + "loss": 1.0342, + "step": 2505 + }, + { + "epoch": 0.22599990981647652, + "grad_norm": 2.2254083151613893, + "learning_rate": 3.6104460418296173e-06, + "loss": 0.9335, + "step": 2506 + }, + { + "epoch": 0.22609009333994678, + "grad_norm": 1.761672736199256, + "learning_rate": 3.6100995525375924e-06, + "loss": 1.0679, + "step": 2507 + }, + { + "epoch": 0.22618027686341705, + "grad_norm": 1.5846897722971578, + "learning_rate": 3.6097529258629952e-06, + "loss": 0.9613, + "step": 2508 + }, + { + "epoch": 0.2262704603868873, + "grad_norm": 2.4591713032107294, + "learning_rate": 3.6094061618354027e-06, + "loss": 1.0484, + "step": 2509 + }, + { + "epoch": 0.22636064391035757, + "grad_norm": 2.2065539358718658, + "learning_rate": 3.609059260484402e-06, + "loss": 1.0151, + "step": 2510 + }, + { + "epoch": 0.22645082743382783, + "grad_norm": 2.3625871632154194, + "learning_rate": 3.6087122218395935e-06, + "loss": 0.961, + "step": 2511 + }, + { + "epoch": 0.2265410109572981, + "grad_norm": 1.6331110047165185, + "learning_rate": 3.608365045930587e-06, + "loss": 0.9864, + "step": 2512 + }, + { + "epoch": 0.22663119448076835, + "grad_norm": 1.9269858533243966, + "learning_rate": 3.608017732787007e-06, + "loss": 1.0003, + "step": 2513 + }, + { + "epoch": 0.22672137800423864, + "grad_norm": 1.892197829579793, + "learning_rate": 3.6076702824384875e-06, + "loss": 0.9449, + "step": 2514 + }, + { + "epoch": 0.2268115615277089, + "grad_norm": 5.172949365314569, + "learning_rate": 3.607322694914675e-06, + "loss": 1.0313, + "step": 2515 + }, + { + "epoch": 0.22690174505117916, + "grad_norm": 1.7096045889088187, + "learning_rate": 3.606974970245227e-06, + "loss": 0.988, + "step": 2516 + }, + { + "epoch": 0.22699192857464942, + "grad_norm": 2.5826401447036544, + "learning_rate": 3.606627108459814e-06, + "loss": 0.9645, + "step": 2517 + }, + { + "epoch": 0.22708211209811968, + "grad_norm": 1.505872991082248, + "learning_rate": 3.6062791095881174e-06, + "loss": 0.9428, + "step": 2518 + }, + { + "epoch": 0.22717229562158994, + "grad_norm": 2.019361477767927, + "learning_rate": 3.6059309736598303e-06, + "loss": 1.088, + "step": 2519 + }, + { + "epoch": 0.2272624791450602, + "grad_norm": 2.156471683329553, + "learning_rate": 3.605582700704657e-06, + "loss": 0.9934, + "step": 2520 + }, + { + "epoch": 0.22735266266853046, + "grad_norm": 1.7961570964193807, + "learning_rate": 3.6052342907523146e-06, + "loss": 1.0221, + "step": 2521 + }, + { + "epoch": 0.22744284619200072, + "grad_norm": 2.1830344260031143, + "learning_rate": 3.604885743832532e-06, + "loss": 0.9466, + "step": 2522 + }, + { + "epoch": 0.22753302971547099, + "grad_norm": 0.6917256194196459, + "learning_rate": 3.6045370599750482e-06, + "loss": 0.8177, + "step": 2523 + }, + { + "epoch": 0.22762321323894125, + "grad_norm": 1.980345592393843, + "learning_rate": 3.604188239209615e-06, + "loss": 0.9956, + "step": 2524 + }, + { + "epoch": 0.2277133967624115, + "grad_norm": 2.4993758376116064, + "learning_rate": 3.603839281565996e-06, + "loss": 0.9415, + "step": 2525 + }, + { + "epoch": 0.22780358028588177, + "grad_norm": 1.9546677866672744, + "learning_rate": 3.603490187073966e-06, + "loss": 1.0315, + "step": 2526 + }, + { + "epoch": 0.22789376380935203, + "grad_norm": 2.215690114688011, + "learning_rate": 3.6031409557633117e-06, + "loss": 1.0762, + "step": 2527 + }, + { + "epoch": 0.2279839473328223, + "grad_norm": 1.820608614302082, + "learning_rate": 3.602791587663831e-06, + "loss": 1.0592, + "step": 2528 + }, + { + "epoch": 0.22807413085629255, + "grad_norm": 3.1229334292344717, + "learning_rate": 3.6024420828053348e-06, + "loss": 1.0556, + "step": 2529 + }, + { + "epoch": 0.2281643143797628, + "grad_norm": 1.707103736438319, + "learning_rate": 3.6020924412176445e-06, + "loss": 1.0905, + "step": 2530 + }, + { + "epoch": 0.22825449790323307, + "grad_norm": 3.709663124110431, + "learning_rate": 3.601742662930593e-06, + "loss": 0.9874, + "step": 2531 + }, + { + "epoch": 0.22834468142670333, + "grad_norm": 1.877161496729916, + "learning_rate": 3.6013927479740248e-06, + "loss": 0.9813, + "step": 2532 + }, + { + "epoch": 0.2284348649501736, + "grad_norm": 1.8970187816142572, + "learning_rate": 3.6010426963777985e-06, + "loss": 1.0947, + "step": 2533 + }, + { + "epoch": 0.22852504847364385, + "grad_norm": 1.7604795440941512, + "learning_rate": 3.6006925081717804e-06, + "loss": 1.0604, + "step": 2534 + }, + { + "epoch": 0.22861523199711412, + "grad_norm": 1.9179752761512252, + "learning_rate": 3.600342183385852e-06, + "loss": 0.9823, + "step": 2535 + }, + { + "epoch": 0.22870541552058438, + "grad_norm": 1.825744121861205, + "learning_rate": 3.5999917220499043e-06, + "loss": 0.955, + "step": 2536 + }, + { + "epoch": 0.22879559904405466, + "grad_norm": 4.247308664306443, + "learning_rate": 3.5996411241938404e-06, + "loss": 1.0175, + "step": 2537 + }, + { + "epoch": 0.22888578256752493, + "grad_norm": 3.1146835940414896, + "learning_rate": 3.5992903898475752e-06, + "loss": 0.9758, + "step": 2538 + }, + { + "epoch": 0.2289759660909952, + "grad_norm": 2.0604226025173467, + "learning_rate": 3.5989395190410365e-06, + "loss": 1.1166, + "step": 2539 + }, + { + "epoch": 0.22906614961446545, + "grad_norm": 2.1941704545481557, + "learning_rate": 3.598588511804161e-06, + "loss": 1.0246, + "step": 2540 + }, + { + "epoch": 0.2291563331379357, + "grad_norm": 2.2670878981960474, + "learning_rate": 3.5982373681668987e-06, + "loss": 1.0208, + "step": 2541 + }, + { + "epoch": 0.22924651666140597, + "grad_norm": 7.326900030046334, + "learning_rate": 3.597886088159212e-06, + "loss": 1.0255, + "step": 2542 + }, + { + "epoch": 0.22933670018487623, + "grad_norm": 1.4344799826107244, + "learning_rate": 3.597534671811074e-06, + "loss": 0.9482, + "step": 2543 + }, + { + "epoch": 0.2294268837083465, + "grad_norm": 2.3589081965585814, + "learning_rate": 3.5971831191524684e-06, + "loss": 0.9984, + "step": 2544 + }, + { + "epoch": 0.22951706723181675, + "grad_norm": 2.2429316013705174, + "learning_rate": 3.5968314302133925e-06, + "loss": 1.0662, + "step": 2545 + }, + { + "epoch": 0.229607250755287, + "grad_norm": 1.888047913105008, + "learning_rate": 3.596479605023854e-06, + "loss": 1.0135, + "step": 2546 + }, + { + "epoch": 0.22969743427875727, + "grad_norm": 2.1108047725985704, + "learning_rate": 3.596127643613873e-06, + "loss": 1.0314, + "step": 2547 + }, + { + "epoch": 0.22978761780222753, + "grad_norm": 1.9605218211182789, + "learning_rate": 3.59577554601348e-06, + "loss": 1.0381, + "step": 2548 + }, + { + "epoch": 0.2298778013256978, + "grad_norm": 1.6704033765345827, + "learning_rate": 3.595423312252719e-06, + "loss": 1.0664, + "step": 2549 + }, + { + "epoch": 0.22996798484916806, + "grad_norm": 2.2670135452343736, + "learning_rate": 3.5950709423616436e-06, + "loss": 1.0029, + "step": 2550 + }, + { + "epoch": 0.23005816837263832, + "grad_norm": 2.510828504113551, + "learning_rate": 3.5947184363703203e-06, + "loss": 0.9926, + "step": 2551 + }, + { + "epoch": 0.23014835189610858, + "grad_norm": 8.11747557612611, + "learning_rate": 3.5943657943088274e-06, + "loss": 0.9642, + "step": 2552 + }, + { + "epoch": 0.23023853541957884, + "grad_norm": 1.8855082101534404, + "learning_rate": 3.5940130162072525e-06, + "loss": 0.9414, + "step": 2553 + }, + { + "epoch": 0.2303287189430491, + "grad_norm": 6.731704784391486, + "learning_rate": 3.5936601020956985e-06, + "loss": 1.0236, + "step": 2554 + }, + { + "epoch": 0.23041890246651936, + "grad_norm": 1.7847216054055237, + "learning_rate": 3.5933070520042772e-06, + "loss": 1.0508, + "step": 2555 + }, + { + "epoch": 0.23050908598998962, + "grad_norm": 0.8611705402321357, + "learning_rate": 3.5929538659631133e-06, + "loss": 0.8679, + "step": 2556 + }, + { + "epoch": 0.23059926951345988, + "grad_norm": 1.7869836981936527, + "learning_rate": 3.592600544002341e-06, + "loss": 1.0522, + "step": 2557 + }, + { + "epoch": 0.23068945303693014, + "grad_norm": 1.6872055361935916, + "learning_rate": 3.5922470861521098e-06, + "loss": 1.0186, + "step": 2558 + }, + { + "epoch": 0.2307796365604004, + "grad_norm": 3.9882623595445135, + "learning_rate": 3.591893492442577e-06, + "loss": 0.9855, + "step": 2559 + }, + { + "epoch": 0.23086982008387066, + "grad_norm": 1.8020252887293489, + "learning_rate": 3.591539762903914e-06, + "loss": 1.0778, + "step": 2560 + }, + { + "epoch": 0.23096000360734095, + "grad_norm": 2.133488921611194, + "learning_rate": 3.591185897566303e-06, + "loss": 0.948, + "step": 2561 + }, + { + "epoch": 0.2310501871308112, + "grad_norm": 2.38581324005815, + "learning_rate": 3.590831896459937e-06, + "loss": 1.0193, + "step": 2562 + }, + { + "epoch": 0.23114037065428147, + "grad_norm": 2.6724798399090797, + "learning_rate": 3.5904777596150222e-06, + "loss": 1.0227, + "step": 2563 + }, + { + "epoch": 0.23123055417775173, + "grad_norm": 2.3374154509078657, + "learning_rate": 3.590123487061775e-06, + "loss": 1.0112, + "step": 2564 + }, + { + "epoch": 0.231320737701222, + "grad_norm": 2.2184717782963594, + "learning_rate": 3.589769078830424e-06, + "loss": 1.1125, + "step": 2565 + }, + { + "epoch": 0.23141092122469226, + "grad_norm": 2.04351542581962, + "learning_rate": 3.58941453495121e-06, + "loss": 0.9963, + "step": 2566 + }, + { + "epoch": 0.23150110474816252, + "grad_norm": 1.9770797233695976, + "learning_rate": 3.5890598554543834e-06, + "loss": 1.0263, + "step": 2567 + }, + { + "epoch": 0.23159128827163278, + "grad_norm": 2.2202946833325887, + "learning_rate": 3.5887050403702073e-06, + "loss": 1.0665, + "step": 2568 + }, + { + "epoch": 0.23168147179510304, + "grad_norm": 2.2412998688783006, + "learning_rate": 3.588350089728958e-06, + "loss": 0.9961, + "step": 2569 + }, + { + "epoch": 0.2317716553185733, + "grad_norm": 1.7481565302547883, + "learning_rate": 3.5879950035609204e-06, + "loss": 1.0428, + "step": 2570 + }, + { + "epoch": 0.23186183884204356, + "grad_norm": 1.836156239566569, + "learning_rate": 3.5876397818963933e-06, + "loss": 1.0134, + "step": 2571 + }, + { + "epoch": 0.23195202236551382, + "grad_norm": 1.4971664208743873, + "learning_rate": 3.5872844247656858e-06, + "loss": 1.0736, + "step": 2572 + }, + { + "epoch": 0.23204220588898408, + "grad_norm": 1.4280940279951493, + "learning_rate": 3.5869289321991195e-06, + "loss": 1.0221, + "step": 2573 + }, + { + "epoch": 0.23213238941245434, + "grad_norm": 1.5181322104988344, + "learning_rate": 3.5865733042270263e-06, + "loss": 1.0062, + "step": 2574 + }, + { + "epoch": 0.2322225729359246, + "grad_norm": 1.9800067189613166, + "learning_rate": 3.5862175408797498e-06, + "loss": 1.0452, + "step": 2575 + }, + { + "epoch": 0.23231275645939486, + "grad_norm": 1.5905254130487299, + "learning_rate": 3.585861642187647e-06, + "loss": 0.9549, + "step": 2576 + }, + { + "epoch": 0.23240293998286513, + "grad_norm": 1.430829368649505, + "learning_rate": 3.5855056081810845e-06, + "loss": 1.0161, + "step": 2577 + }, + { + "epoch": 0.2324931235063354, + "grad_norm": 1.941031222128199, + "learning_rate": 3.5851494388904406e-06, + "loss": 1.0156, + "step": 2578 + }, + { + "epoch": 0.23258330702980565, + "grad_norm": 2.6076874901285807, + "learning_rate": 3.5847931343461064e-06, + "loss": 0.9555, + "step": 2579 + }, + { + "epoch": 0.2326734905532759, + "grad_norm": 1.9275902227519093, + "learning_rate": 3.5844366945784835e-06, + "loss": 0.9446, + "step": 2580 + }, + { + "epoch": 0.23276367407674617, + "grad_norm": 2.3521931458465377, + "learning_rate": 3.5840801196179856e-06, + "loss": 1.0735, + "step": 2581 + }, + { + "epoch": 0.23285385760021643, + "grad_norm": 2.330220655381136, + "learning_rate": 3.583723409495037e-06, + "loss": 1.0727, + "step": 2582 + }, + { + "epoch": 0.2329440411236867, + "grad_norm": 1.7685492950980741, + "learning_rate": 3.5833665642400747e-06, + "loss": 1.074, + "step": 2583 + }, + { + "epoch": 0.23303422464715695, + "grad_norm": 1.6262426026958203, + "learning_rate": 3.5830095838835472e-06, + "loss": 0.9418, + "step": 2584 + }, + { + "epoch": 0.23312440817062724, + "grad_norm": 2.7357197098071726, + "learning_rate": 3.5826524684559125e-06, + "loss": 0.9865, + "step": 2585 + }, + { + "epoch": 0.2332145916940975, + "grad_norm": 2.2522180009419803, + "learning_rate": 3.5822952179876433e-06, + "loss": 1.097, + "step": 2586 + }, + { + "epoch": 0.23330477521756776, + "grad_norm": 1.99951887304123, + "learning_rate": 3.5819378325092205e-06, + "loss": 0.9881, + "step": 2587 + }, + { + "epoch": 0.23339495874103802, + "grad_norm": 1.488197304268041, + "learning_rate": 3.581580312051139e-06, + "loss": 1.0416, + "step": 2588 + }, + { + "epoch": 0.23348514226450828, + "grad_norm": 1.8096402726543066, + "learning_rate": 3.5812226566439057e-06, + "loss": 0.9422, + "step": 2589 + }, + { + "epoch": 0.23357532578797854, + "grad_norm": 1.9943847506818841, + "learning_rate": 3.580864866318036e-06, + "loss": 0.9932, + "step": 2590 + }, + { + "epoch": 0.2336655093114488, + "grad_norm": 2.03871017276821, + "learning_rate": 3.580506941104059e-06, + "loss": 1.0245, + "step": 2591 + }, + { + "epoch": 0.23375569283491907, + "grad_norm": 1.675304425032783, + "learning_rate": 3.580148881032515e-06, + "loss": 1.0084, + "step": 2592 + }, + { + "epoch": 0.23384587635838933, + "grad_norm": 2.3132242666619107, + "learning_rate": 3.5797906861339556e-06, + "loss": 1.0244, + "step": 2593 + }, + { + "epoch": 0.2339360598818596, + "grad_norm": 2.217929312209918, + "learning_rate": 3.5794323564389435e-06, + "loss": 1.0487, + "step": 2594 + }, + { + "epoch": 0.23402624340532985, + "grad_norm": 2.003639723990989, + "learning_rate": 3.579073891978055e-06, + "loss": 1.0256, + "step": 2595 + }, + { + "epoch": 0.2341164269288001, + "grad_norm": 1.6964868090965872, + "learning_rate": 3.5787152927818746e-06, + "loss": 0.9847, + "step": 2596 + }, + { + "epoch": 0.23420661045227037, + "grad_norm": 2.2143752260178404, + "learning_rate": 3.5783565588810003e-06, + "loss": 1.0806, + "step": 2597 + }, + { + "epoch": 0.23429679397574063, + "grad_norm": 1.6309583439749447, + "learning_rate": 3.5779976903060412e-06, + "loss": 0.9877, + "step": 2598 + }, + { + "epoch": 0.2343869774992109, + "grad_norm": 1.8195474351201575, + "learning_rate": 3.577638687087619e-06, + "loss": 1.0482, + "step": 2599 + }, + { + "epoch": 0.23447716102268115, + "grad_norm": 1.872295845626707, + "learning_rate": 3.577279549256364e-06, + "loss": 1.0272, + "step": 2600 + }, + { + "epoch": 0.2345673445461514, + "grad_norm": 1.911095386386221, + "learning_rate": 3.5769202768429213e-06, + "loss": 1.0624, + "step": 2601 + }, + { + "epoch": 0.23465752806962167, + "grad_norm": 2.5234138794729453, + "learning_rate": 3.5765608698779454e-06, + "loss": 0.9973, + "step": 2602 + }, + { + "epoch": 0.23474771159309193, + "grad_norm": 2.13820628493683, + "learning_rate": 3.5762013283921033e-06, + "loss": 1.0245, + "step": 2603 + }, + { + "epoch": 0.2348378951165622, + "grad_norm": 2.9350392200247506, + "learning_rate": 3.5758416524160728e-06, + "loss": 1.0232, + "step": 2604 + }, + { + "epoch": 0.23492807864003246, + "grad_norm": 1.9983646740442895, + "learning_rate": 3.5754818419805427e-06, + "loss": 1.0096, + "step": 2605 + }, + { + "epoch": 0.23501826216350272, + "grad_norm": 1.510747158681018, + "learning_rate": 3.575121897116216e-06, + "loss": 0.9949, + "step": 2606 + }, + { + "epoch": 0.23510844568697298, + "grad_norm": 2.2618158349265016, + "learning_rate": 3.574761817853803e-06, + "loss": 1.0561, + "step": 2607 + }, + { + "epoch": 0.23519862921044324, + "grad_norm": 1.7897398220411747, + "learning_rate": 3.5744016042240287e-06, + "loss": 0.9487, + "step": 2608 + }, + { + "epoch": 0.23528881273391353, + "grad_norm": 2.0915011954745295, + "learning_rate": 3.5740412562576286e-06, + "loss": 0.9499, + "step": 2609 + }, + { + "epoch": 0.2353789962573838, + "grad_norm": 3.398855179298787, + "learning_rate": 3.573680773985349e-06, + "loss": 1.0122, + "step": 2610 + }, + { + "epoch": 0.23546917978085405, + "grad_norm": 3.090609256715859, + "learning_rate": 3.5733201574379486e-06, + "loss": 1.0152, + "step": 2611 + }, + { + "epoch": 0.2355593633043243, + "grad_norm": 2.116882527475683, + "learning_rate": 3.5729594066461975e-06, + "loss": 0.9609, + "step": 2612 + }, + { + "epoch": 0.23564954682779457, + "grad_norm": 4.0461722098795985, + "learning_rate": 3.572598521640876e-06, + "loss": 1.0303, + "step": 2613 + }, + { + "epoch": 0.23573973035126483, + "grad_norm": 0.7985291982649851, + "learning_rate": 3.5722375024527782e-06, + "loss": 0.8717, + "step": 2614 + }, + { + "epoch": 0.2358299138747351, + "grad_norm": 1.5210692949881144, + "learning_rate": 3.571876349112707e-06, + "loss": 1.0399, + "step": 2615 + }, + { + "epoch": 0.23592009739820535, + "grad_norm": 2.3372077677623615, + "learning_rate": 3.5715150616514784e-06, + "loss": 1.014, + "step": 2616 + }, + { + "epoch": 0.2360102809216756, + "grad_norm": 3.109004779800957, + "learning_rate": 3.5711536400999196e-06, + "loss": 1.0712, + "step": 2617 + }, + { + "epoch": 0.23610046444514587, + "grad_norm": 0.7202942885208834, + "learning_rate": 3.570792084488869e-06, + "loss": 0.8746, + "step": 2618 + }, + { + "epoch": 0.23619064796861614, + "grad_norm": 0.714394767133978, + "learning_rate": 3.5704303948491764e-06, + "loss": 0.8864, + "step": 2619 + }, + { + "epoch": 0.2362808314920864, + "grad_norm": 1.8254753538843411, + "learning_rate": 3.5700685712117035e-06, + "loss": 1.0363, + "step": 2620 + }, + { + "epoch": 0.23637101501555666, + "grad_norm": 1.798394796094798, + "learning_rate": 3.5697066136073227e-06, + "loss": 1.0057, + "step": 2621 + }, + { + "epoch": 0.23646119853902692, + "grad_norm": 1.6334466752151673, + "learning_rate": 3.5693445220669184e-06, + "loss": 1.024, + "step": 2622 + }, + { + "epoch": 0.23655138206249718, + "grad_norm": 1.4197941739719537, + "learning_rate": 3.568982296621386e-06, + "loss": 1.0123, + "step": 2623 + }, + { + "epoch": 0.23664156558596744, + "grad_norm": 2.1227028716312817, + "learning_rate": 3.5686199373016325e-06, + "loss": 1.0903, + "step": 2624 + }, + { + "epoch": 0.2367317491094377, + "grad_norm": 1.6085167235079303, + "learning_rate": 3.568257444138577e-06, + "loss": 0.966, + "step": 2625 + }, + { + "epoch": 0.23682193263290796, + "grad_norm": 2.153767801787963, + "learning_rate": 3.5678948171631495e-06, + "loss": 1.0184, + "step": 2626 + }, + { + "epoch": 0.23691211615637822, + "grad_norm": 2.0987439168327287, + "learning_rate": 3.5675320564062908e-06, + "loss": 1.0235, + "step": 2627 + }, + { + "epoch": 0.23700229967984848, + "grad_norm": 1.8439184208902952, + "learning_rate": 3.5671691618989533e-06, + "loss": 0.9319, + "step": 2628 + }, + { + "epoch": 0.23709248320331874, + "grad_norm": 1.6930415681353859, + "learning_rate": 3.5668061336721024e-06, + "loss": 1.0288, + "step": 2629 + }, + { + "epoch": 0.237182666726789, + "grad_norm": 1.9357603938855683, + "learning_rate": 3.5664429717567117e-06, + "loss": 1.0091, + "step": 2630 + }, + { + "epoch": 0.23727285025025927, + "grad_norm": 1.4751332142044147, + "learning_rate": 3.56607967618377e-06, + "loss": 1.045, + "step": 2631 + }, + { + "epoch": 0.23736303377372953, + "grad_norm": 1.994146363850704, + "learning_rate": 3.5657162469842754e-06, + "loss": 1.0316, + "step": 2632 + }, + { + "epoch": 0.23745321729719981, + "grad_norm": 1.5645011294425772, + "learning_rate": 3.5653526841892374e-06, + "loss": 1.0192, + "step": 2633 + }, + { + "epoch": 0.23754340082067008, + "grad_norm": 2.4328990452929613, + "learning_rate": 3.564988987829676e-06, + "loss": 1.0506, + "step": 2634 + }, + { + "epoch": 0.23763358434414034, + "grad_norm": 1.9117355214994038, + "learning_rate": 3.564625157936626e-06, + "loss": 0.9572, + "step": 2635 + }, + { + "epoch": 0.2377237678676106, + "grad_norm": 1.8615867859321003, + "learning_rate": 3.56426119454113e-06, + "loss": 1.053, + "step": 2636 + }, + { + "epoch": 0.23781395139108086, + "grad_norm": 2.1729753300424615, + "learning_rate": 3.5638970976742436e-06, + "loss": 1.0904, + "step": 2637 + }, + { + "epoch": 0.23790413491455112, + "grad_norm": 3.652319858085291, + "learning_rate": 3.5635328673670335e-06, + "loss": 1.0398, + "step": 2638 + }, + { + "epoch": 0.23799431843802138, + "grad_norm": 1.5059885644972488, + "learning_rate": 3.5631685036505783e-06, + "loss": 1.1103, + "step": 2639 + }, + { + "epoch": 0.23808450196149164, + "grad_norm": 1.6274174534831594, + "learning_rate": 3.562804006555966e-06, + "loss": 1.0333, + "step": 2640 + }, + { + "epoch": 0.2381746854849619, + "grad_norm": 2.3696704092897245, + "learning_rate": 3.5624393761143e-06, + "loss": 0.958, + "step": 2641 + }, + { + "epoch": 0.23826486900843216, + "grad_norm": 2.0292563411420415, + "learning_rate": 3.5620746123566906e-06, + "loss": 0.9324, + "step": 2642 + }, + { + "epoch": 0.23835505253190242, + "grad_norm": 1.811766772108743, + "learning_rate": 3.5617097153142623e-06, + "loss": 0.965, + "step": 2643 + }, + { + "epoch": 0.23844523605537268, + "grad_norm": 1.8641277445704714, + "learning_rate": 3.5613446850181497e-06, + "loss": 1.0604, + "step": 2644 + }, + { + "epoch": 0.23853541957884294, + "grad_norm": 0.611014556903342, + "learning_rate": 3.5609795214994996e-06, + "loss": 0.7748, + "step": 2645 + }, + { + "epoch": 0.2386256031023132, + "grad_norm": 1.801489568658422, + "learning_rate": 3.560614224789469e-06, + "loss": 1.0733, + "step": 2646 + }, + { + "epoch": 0.23871578662578347, + "grad_norm": 1.7165184098947406, + "learning_rate": 3.5602487949192285e-06, + "loss": 0.9638, + "step": 2647 + }, + { + "epoch": 0.23880597014925373, + "grad_norm": 3.4720947360900563, + "learning_rate": 3.559883231919957e-06, + "loss": 0.9618, + "step": 2648 + }, + { + "epoch": 0.238896153672724, + "grad_norm": 10.360663378475813, + "learning_rate": 3.5595175358228473e-06, + "loss": 0.9934, + "step": 2649 + }, + { + "epoch": 0.23898633719619425, + "grad_norm": 1.775901326501311, + "learning_rate": 3.5591517066591027e-06, + "loss": 1.018, + "step": 2650 + }, + { + "epoch": 0.2390765207196645, + "grad_norm": 2.6875541149280147, + "learning_rate": 3.5587857444599364e-06, + "loss": 1.0119, + "step": 2651 + }, + { + "epoch": 0.23916670424313477, + "grad_norm": 1.7773825840323074, + "learning_rate": 3.5584196492565766e-06, + "loss": 0.9864, + "step": 2652 + }, + { + "epoch": 0.23925688776660503, + "grad_norm": 1.6117408177709043, + "learning_rate": 3.5580534210802587e-06, + "loss": 1.0843, + "step": 2653 + }, + { + "epoch": 0.2393470712900753, + "grad_norm": 1.5289329592750363, + "learning_rate": 3.557687059962232e-06, + "loss": 1.0622, + "step": 2654 + }, + { + "epoch": 0.23943725481354555, + "grad_norm": 1.4985801016343452, + "learning_rate": 3.5573205659337558e-06, + "loss": 1.0144, + "step": 2655 + }, + { + "epoch": 0.23952743833701584, + "grad_norm": 1.5919772742149372, + "learning_rate": 3.5569539390261025e-06, + "loss": 1.0074, + "step": 2656 + }, + { + "epoch": 0.2396176218604861, + "grad_norm": 1.5144644157533884, + "learning_rate": 3.5565871792705543e-06, + "loss": 1.0333, + "step": 2657 + }, + { + "epoch": 0.23970780538395636, + "grad_norm": 1.535548657455322, + "learning_rate": 3.5562202866984045e-06, + "loss": 1.0133, + "step": 2658 + }, + { + "epoch": 0.23979798890742662, + "grad_norm": 2.114036326447248, + "learning_rate": 3.5558532613409594e-06, + "loss": 1.0312, + "step": 2659 + }, + { + "epoch": 0.23988817243089688, + "grad_norm": 2.7886417189043082, + "learning_rate": 3.555486103229535e-06, + "loss": 0.9558, + "step": 2660 + }, + { + "epoch": 0.23997835595436715, + "grad_norm": 1.9035180949620365, + "learning_rate": 3.5551188123954595e-06, + "loss": 1.0426, + "step": 2661 + }, + { + "epoch": 0.2400685394778374, + "grad_norm": 1.9748979880147164, + "learning_rate": 3.5547513888700715e-06, + "loss": 0.912, + "step": 2662 + }, + { + "epoch": 0.24015872300130767, + "grad_norm": 1.6265974262749547, + "learning_rate": 3.5543838326847224e-06, + "loss": 1.052, + "step": 2663 + }, + { + "epoch": 0.24024890652477793, + "grad_norm": 2.1813094691173545, + "learning_rate": 3.5540161438707744e-06, + "loss": 0.9998, + "step": 2664 + }, + { + "epoch": 0.2403390900482482, + "grad_norm": 1.7536623642900153, + "learning_rate": 3.5536483224596e-06, + "loss": 1.1774, + "step": 2665 + }, + { + "epoch": 0.24042927357171845, + "grad_norm": 1.9930167473157814, + "learning_rate": 3.553280368482584e-06, + "loss": 0.9942, + "step": 2666 + }, + { + "epoch": 0.2405194570951887, + "grad_norm": 2.2933428327600254, + "learning_rate": 3.5529122819711227e-06, + "loss": 0.8402, + "step": 2667 + }, + { + "epoch": 0.24060964061865897, + "grad_norm": 1.5845952108564783, + "learning_rate": 3.5525440629566223e-06, + "loss": 0.9628, + "step": 2668 + }, + { + "epoch": 0.24069982414212923, + "grad_norm": 2.0007938955101774, + "learning_rate": 3.552175711470502e-06, + "loss": 1.0393, + "step": 2669 + }, + { + "epoch": 0.2407900076655995, + "grad_norm": 1.8040598649032646, + "learning_rate": 3.5518072275441912e-06, + "loss": 1.063, + "step": 2670 + }, + { + "epoch": 0.24088019118906975, + "grad_norm": 1.982510569425215, + "learning_rate": 3.551438611209131e-06, + "loss": 1.0469, + "step": 2671 + }, + { + "epoch": 0.24097037471254001, + "grad_norm": 1.6624495878678993, + "learning_rate": 3.551069862496774e-06, + "loss": 1.1127, + "step": 2672 + }, + { + "epoch": 0.24106055823601027, + "grad_norm": 0.655295677205612, + "learning_rate": 3.5507009814385846e-06, + "loss": 0.8359, + "step": 2673 + }, + { + "epoch": 0.24115074175948054, + "grad_norm": 0.6639100965755191, + "learning_rate": 3.550331968066036e-06, + "loss": 0.8273, + "step": 2674 + }, + { + "epoch": 0.2412409252829508, + "grad_norm": 1.7484350018658148, + "learning_rate": 3.549962822410616e-06, + "loss": 1.0341, + "step": 2675 + }, + { + "epoch": 0.24133110880642106, + "grad_norm": 0.6977087797784464, + "learning_rate": 3.5495935445038217e-06, + "loss": 0.8235, + "step": 2676 + }, + { + "epoch": 0.24142129232989132, + "grad_norm": 1.8164131410529567, + "learning_rate": 3.5492241343771612e-06, + "loss": 1.0591, + "step": 2677 + }, + { + "epoch": 0.24151147585336158, + "grad_norm": 1.9574985212805218, + "learning_rate": 3.548854592062156e-06, + "loss": 1.0179, + "step": 2678 + }, + { + "epoch": 0.24160165937683184, + "grad_norm": 2.5555607880893527, + "learning_rate": 3.548484917590336e-06, + "loss": 1.0003, + "step": 2679 + }, + { + "epoch": 0.24169184290030213, + "grad_norm": 1.8791461244237089, + "learning_rate": 3.5481151109932447e-06, + "loss": 0.9352, + "step": 2680 + }, + { + "epoch": 0.2417820264237724, + "grad_norm": 1.9518678204895739, + "learning_rate": 3.5477451723024364e-06, + "loss": 1.0416, + "step": 2681 + }, + { + "epoch": 0.24187220994724265, + "grad_norm": 0.5895628354942996, + "learning_rate": 3.5473751015494757e-06, + "loss": 0.808, + "step": 2682 + }, + { + "epoch": 0.2419623934707129, + "grad_norm": 3.0720734727139796, + "learning_rate": 3.547004898765939e-06, + "loss": 1.034, + "step": 2683 + }, + { + "epoch": 0.24205257699418317, + "grad_norm": 2.0618250349315312, + "learning_rate": 3.546634563983414e-06, + "loss": 1.0363, + "step": 2684 + }, + { + "epoch": 0.24214276051765343, + "grad_norm": 1.5997610241521583, + "learning_rate": 3.5462640972335002e-06, + "loss": 0.9823, + "step": 2685 + }, + { + "epoch": 0.2422329440411237, + "grad_norm": 1.8956829542387443, + "learning_rate": 3.5458934985478077e-06, + "loss": 1.0379, + "step": 2686 + }, + { + "epoch": 0.24232312756459395, + "grad_norm": 1.9386438254349283, + "learning_rate": 3.5455227679579577e-06, + "loss": 0.9979, + "step": 2687 + }, + { + "epoch": 0.24241331108806422, + "grad_norm": 1.822819959709276, + "learning_rate": 3.545151905495584e-06, + "loss": 1.1242, + "step": 2688 + }, + { + "epoch": 0.24250349461153448, + "grad_norm": 1.9202239224868658, + "learning_rate": 3.544780911192329e-06, + "loss": 0.9952, + "step": 2689 + }, + { + "epoch": 0.24259367813500474, + "grad_norm": 2.6671543470131263, + "learning_rate": 3.544409785079849e-06, + "loss": 0.9659, + "step": 2690 + }, + { + "epoch": 0.242683861658475, + "grad_norm": 1.5533176694061113, + "learning_rate": 3.5440385271898103e-06, + "loss": 0.9622, + "step": 2691 + }, + { + "epoch": 0.24277404518194526, + "grad_norm": 1.9758728995585624, + "learning_rate": 3.5436671375538903e-06, + "loss": 0.9434, + "step": 2692 + }, + { + "epoch": 0.24286422870541552, + "grad_norm": 1.9579937484882997, + "learning_rate": 3.543295616203779e-06, + "loss": 1.0164, + "step": 2693 + }, + { + "epoch": 0.24295441222888578, + "grad_norm": 1.768635504154139, + "learning_rate": 3.542923963171176e-06, + "loss": 1.1108, + "step": 2694 + }, + { + "epoch": 0.24304459575235604, + "grad_norm": 1.3673268710352122, + "learning_rate": 3.542552178487793e-06, + "loss": 0.9365, + "step": 2695 + }, + { + "epoch": 0.2431347792758263, + "grad_norm": 1.7699100855171794, + "learning_rate": 3.5421802621853523e-06, + "loss": 1.0084, + "step": 2696 + }, + { + "epoch": 0.24322496279929656, + "grad_norm": 2.0287959606819377, + "learning_rate": 3.5418082142955887e-06, + "loss": 1.0152, + "step": 2697 + }, + { + "epoch": 0.24331514632276682, + "grad_norm": 0.6702286043106555, + "learning_rate": 3.5414360348502463e-06, + "loss": 0.8523, + "step": 2698 + }, + { + "epoch": 0.24340532984623708, + "grad_norm": 1.5186558729602584, + "learning_rate": 3.5410637238810825e-06, + "loss": 0.9251, + "step": 2699 + }, + { + "epoch": 0.24349551336970734, + "grad_norm": 1.8738664379421746, + "learning_rate": 3.5406912814198635e-06, + "loss": 0.9722, + "step": 2700 + }, + { + "epoch": 0.2435856968931776, + "grad_norm": 3.566388868642612, + "learning_rate": 3.54031870749837e-06, + "loss": 0.9835, + "step": 2701 + }, + { + "epoch": 0.24367588041664787, + "grad_norm": 1.7625408411366825, + "learning_rate": 3.539946002148391e-06, + "loss": 1.0306, + "step": 2702 + }, + { + "epoch": 0.24376606394011813, + "grad_norm": 2.0584895127487726, + "learning_rate": 3.5395731654017277e-06, + "loss": 1.0191, + "step": 2703 + }, + { + "epoch": 0.24385624746358842, + "grad_norm": 1.9562255053464426, + "learning_rate": 3.5392001972901923e-06, + "loss": 1.0587, + "step": 2704 + }, + { + "epoch": 0.24394643098705868, + "grad_norm": 1.9193381336960824, + "learning_rate": 3.5388270978456098e-06, + "loss": 0.9813, + "step": 2705 + }, + { + "epoch": 0.24403661451052894, + "grad_norm": 3.119121821376086, + "learning_rate": 3.5384538670998137e-06, + "loss": 0.9862, + "step": 2706 + }, + { + "epoch": 0.2441267980339992, + "grad_norm": 1.7663883145892583, + "learning_rate": 3.538080505084651e-06, + "loss": 1.0403, + "step": 2707 + }, + { + "epoch": 0.24421698155746946, + "grad_norm": 1.583251934301513, + "learning_rate": 3.5377070118319788e-06, + "loss": 1.0328, + "step": 2708 + }, + { + "epoch": 0.24430716508093972, + "grad_norm": 2.3159727877861322, + "learning_rate": 3.5373333873736657e-06, + "loss": 0.9695, + "step": 2709 + }, + { + "epoch": 0.24439734860440998, + "grad_norm": 2.4847336036455103, + "learning_rate": 3.536959631741591e-06, + "loss": 1.0243, + "step": 2710 + }, + { + "epoch": 0.24448753212788024, + "grad_norm": 1.2463675649549537, + "learning_rate": 3.536585744967646e-06, + "loss": 1.0422, + "step": 2711 + }, + { + "epoch": 0.2445777156513505, + "grad_norm": 2.0840089719656087, + "learning_rate": 3.5362117270837326e-06, + "loss": 1.001, + "step": 2712 + }, + { + "epoch": 0.24466789917482076, + "grad_norm": 1.5174512576927361, + "learning_rate": 3.5358375781217634e-06, + "loss": 0.9171, + "step": 2713 + }, + { + "epoch": 0.24475808269829102, + "grad_norm": 1.8805548083618322, + "learning_rate": 3.535463298113664e-06, + "loss": 1.0508, + "step": 2714 + }, + { + "epoch": 0.24484826622176128, + "grad_norm": 1.787316415602571, + "learning_rate": 3.5350888870913697e-06, + "loss": 0.9807, + "step": 2715 + }, + { + "epoch": 0.24493844974523155, + "grad_norm": 1.7469304957983498, + "learning_rate": 3.5347143450868273e-06, + "loss": 1.0246, + "step": 2716 + }, + { + "epoch": 0.2450286332687018, + "grad_norm": 1.826316134216864, + "learning_rate": 3.534339672131994e-06, + "loss": 1.0368, + "step": 2717 + }, + { + "epoch": 0.24511881679217207, + "grad_norm": 1.5516930501206103, + "learning_rate": 3.5339648682588397e-06, + "loss": 0.9518, + "step": 2718 + }, + { + "epoch": 0.24520900031564233, + "grad_norm": 2.5411393812039926, + "learning_rate": 3.533589933499345e-06, + "loss": 1.0626, + "step": 2719 + }, + { + "epoch": 0.2452991838391126, + "grad_norm": 1.7847200691351992, + "learning_rate": 3.533214867885501e-06, + "loss": 0.9524, + "step": 2720 + }, + { + "epoch": 0.24538936736258285, + "grad_norm": 2.3731892358597193, + "learning_rate": 3.53283967144931e-06, + "loss": 0.9586, + "step": 2721 + }, + { + "epoch": 0.2454795508860531, + "grad_norm": 1.7689340017200772, + "learning_rate": 3.532464344222787e-06, + "loss": 1.0287, + "step": 2722 + }, + { + "epoch": 0.24556973440952337, + "grad_norm": 1.7297666343774578, + "learning_rate": 3.532088886237956e-06, + "loss": 1.0575, + "step": 2723 + }, + { + "epoch": 0.24565991793299363, + "grad_norm": 0.645196559734124, + "learning_rate": 3.5317132975268535e-06, + "loss": 0.765, + "step": 2724 + }, + { + "epoch": 0.2457501014564639, + "grad_norm": 3.318233063232037, + "learning_rate": 3.531337578121526e-06, + "loss": 1.0434, + "step": 2725 + }, + { + "epoch": 0.24584028497993415, + "grad_norm": 2.183066643984195, + "learning_rate": 3.530961728054033e-06, + "loss": 1.0564, + "step": 2726 + }, + { + "epoch": 0.24593046850340441, + "grad_norm": 1.4766279740929675, + "learning_rate": 3.5305857473564435e-06, + "loss": 0.9777, + "step": 2727 + }, + { + "epoch": 0.2460206520268747, + "grad_norm": 2.3149569063727813, + "learning_rate": 3.5302096360608385e-06, + "loss": 0.9549, + "step": 2728 + }, + { + "epoch": 0.24611083555034496, + "grad_norm": 1.6736731440794796, + "learning_rate": 3.5298333941993105e-06, + "loss": 1.0739, + "step": 2729 + }, + { + "epoch": 0.24620101907381522, + "grad_norm": 1.9698500814371669, + "learning_rate": 3.529457021803962e-06, + "loss": 1.0086, + "step": 2730 + }, + { + "epoch": 0.24629120259728549, + "grad_norm": 1.8629183945732775, + "learning_rate": 3.529080518906906e-06, + "loss": 1.032, + "step": 2731 + }, + { + "epoch": 0.24638138612075575, + "grad_norm": 2.4110505409465257, + "learning_rate": 3.5287038855402696e-06, + "loss": 0.9996, + "step": 2732 + }, + { + "epoch": 0.246471569644226, + "grad_norm": 1.899936915906375, + "learning_rate": 3.528327121736188e-06, + "loss": 1.0558, + "step": 2733 + }, + { + "epoch": 0.24656175316769627, + "grad_norm": 1.6289436464110671, + "learning_rate": 3.52795022752681e-06, + "loss": 1.0971, + "step": 2734 + }, + { + "epoch": 0.24665193669116653, + "grad_norm": 2.2013663990410928, + "learning_rate": 3.5275732029442925e-06, + "loss": 1.0838, + "step": 2735 + }, + { + "epoch": 0.2467421202146368, + "grad_norm": 2.3515603708656787, + "learning_rate": 3.5271960480208077e-06, + "loss": 0.9579, + "step": 2736 + }, + { + "epoch": 0.24683230373810705, + "grad_norm": 0.6834077200486751, + "learning_rate": 3.526818762788534e-06, + "loss": 0.8653, + "step": 2737 + }, + { + "epoch": 0.2469224872615773, + "grad_norm": 1.7135693935327003, + "learning_rate": 3.5264413472796653e-06, + "loss": 1.057, + "step": 2738 + }, + { + "epoch": 0.24701267078504757, + "grad_norm": 2.030248659247848, + "learning_rate": 3.5260638015264037e-06, + "loss": 0.999, + "step": 2739 + }, + { + "epoch": 0.24710285430851783, + "grad_norm": 1.3445105507302966, + "learning_rate": 3.5256861255609644e-06, + "loss": 0.99, + "step": 2740 + }, + { + "epoch": 0.2471930378319881, + "grad_norm": 2.205748598868124, + "learning_rate": 3.5253083194155723e-06, + "loss": 1.0303, + "step": 2741 + }, + { + "epoch": 0.24728322135545835, + "grad_norm": 1.4767585906758383, + "learning_rate": 3.5249303831224637e-06, + "loss": 1.0394, + "step": 2742 + }, + { + "epoch": 0.24737340487892862, + "grad_norm": 1.5061258161961923, + "learning_rate": 3.524552316713887e-06, + "loss": 1.0065, + "step": 2743 + }, + { + "epoch": 0.24746358840239888, + "grad_norm": 1.6805238992250897, + "learning_rate": 3.5241741202220995e-06, + "loss": 1.0766, + "step": 2744 + }, + { + "epoch": 0.24755377192586914, + "grad_norm": 1.6003508659938013, + "learning_rate": 3.5237957936793724e-06, + "loss": 1.0032, + "step": 2745 + }, + { + "epoch": 0.2476439554493394, + "grad_norm": 1.4261909627773341, + "learning_rate": 3.523417337117986e-06, + "loss": 0.9821, + "step": 2746 + }, + { + "epoch": 0.24773413897280966, + "grad_norm": 1.3179718126081645, + "learning_rate": 3.523038750570232e-06, + "loss": 0.9668, + "step": 2747 + }, + { + "epoch": 0.24782432249627992, + "grad_norm": 1.547314745116821, + "learning_rate": 3.522660034068414e-06, + "loss": 1.0459, + "step": 2748 + }, + { + "epoch": 0.24791450601975018, + "grad_norm": 1.8482149662519405, + "learning_rate": 3.5222811876448464e-06, + "loss": 1.029, + "step": 2749 + }, + { + "epoch": 0.24800468954322044, + "grad_norm": 1.879735941908786, + "learning_rate": 3.521902211331854e-06, + "loss": 1.0342, + "step": 2750 + }, + { + "epoch": 0.2480948730666907, + "grad_norm": 1.6336957377008603, + "learning_rate": 3.5215231051617726e-06, + "loss": 0.961, + "step": 2751 + }, + { + "epoch": 0.248185056590161, + "grad_norm": 1.8880454506858648, + "learning_rate": 3.521143869166951e-06, + "loss": 0.9166, + "step": 2752 + }, + { + "epoch": 0.24827524011363125, + "grad_norm": 2.8584923554458874, + "learning_rate": 3.5207645033797464e-06, + "loss": 1.0484, + "step": 2753 + }, + { + "epoch": 0.2483654236371015, + "grad_norm": 1.8079933174988798, + "learning_rate": 3.5203850078325293e-06, + "loss": 0.9426, + "step": 2754 + }, + { + "epoch": 0.24845560716057177, + "grad_norm": 1.814549438416939, + "learning_rate": 3.5200053825576797e-06, + "loss": 0.9247, + "step": 2755 + }, + { + "epoch": 0.24854579068404203, + "grad_norm": 1.63752577375556, + "learning_rate": 3.51962562758759e-06, + "loss": 1.0183, + "step": 2756 + }, + { + "epoch": 0.2486359742075123, + "grad_norm": 1.9552392731218553, + "learning_rate": 3.5192457429546627e-06, + "loss": 0.9895, + "step": 2757 + }, + { + "epoch": 0.24872615773098256, + "grad_norm": 1.6735831830379646, + "learning_rate": 3.5188657286913115e-06, + "loss": 0.9848, + "step": 2758 + }, + { + "epoch": 0.24881634125445282, + "grad_norm": 2.4510102116135872, + "learning_rate": 3.518485584829961e-06, + "loss": 0.981, + "step": 2759 + }, + { + "epoch": 0.24890652477792308, + "grad_norm": 1.461660603967763, + "learning_rate": 3.5181053114030485e-06, + "loss": 1.0281, + "step": 2760 + }, + { + "epoch": 0.24899670830139334, + "grad_norm": 1.9251059069543999, + "learning_rate": 3.5177249084430198e-06, + "loss": 1.0112, + "step": 2761 + }, + { + "epoch": 0.2490868918248636, + "grad_norm": 2.0277917620497465, + "learning_rate": 3.517344375982333e-06, + "loss": 1.0466, + "step": 2762 + }, + { + "epoch": 0.24917707534833386, + "grad_norm": 1.7691793533974771, + "learning_rate": 3.5169637140534565e-06, + "loss": 1.0503, + "step": 2763 + }, + { + "epoch": 0.24926725887180412, + "grad_norm": 1.9310003594586331, + "learning_rate": 3.5165829226888733e-06, + "loss": 0.9843, + "step": 2764 + }, + { + "epoch": 0.24935744239527438, + "grad_norm": 2.0429964081914234, + "learning_rate": 3.516202001921072e-06, + "loss": 1.0407, + "step": 2765 + }, + { + "epoch": 0.24944762591874464, + "grad_norm": 1.797446118689254, + "learning_rate": 3.515820951782555e-06, + "loss": 1.0415, + "step": 2766 + }, + { + "epoch": 0.2495378094422149, + "grad_norm": 2.319636596466513, + "learning_rate": 3.5154397723058366e-06, + "loss": 0.9415, + "step": 2767 + }, + { + "epoch": 0.24962799296568516, + "grad_norm": 2.0043533153687227, + "learning_rate": 3.5150584635234416e-06, + "loss": 1.0713, + "step": 2768 + }, + { + "epoch": 0.24971817648915542, + "grad_norm": 1.3377898909242871, + "learning_rate": 3.5146770254679035e-06, + "loss": 1.0606, + "step": 2769 + }, + { + "epoch": 0.24980836001262569, + "grad_norm": 2.0647724233107865, + "learning_rate": 3.51429545817177e-06, + "loss": 0.9405, + "step": 2770 + }, + { + "epoch": 0.24989854353609595, + "grad_norm": 1.8091294749598712, + "learning_rate": 3.5139137616675985e-06, + "loss": 1.0239, + "step": 2771 + }, + { + "epoch": 0.2499887270595662, + "grad_norm": 1.9093261662558259, + "learning_rate": 3.513531935987957e-06, + "loss": 0.9559, + "step": 2772 + }, + { + "epoch": 0.2500789105830365, + "grad_norm": 1.8477281229827027, + "learning_rate": 3.5131499811654253e-06, + "loss": 1.0354, + "step": 2773 + }, + { + "epoch": 0.25016909410650673, + "grad_norm": 1.5685950609268744, + "learning_rate": 3.512767897232594e-06, + "loss": 0.9814, + "step": 2774 + }, + { + "epoch": 0.250259277629977, + "grad_norm": 1.7411258353650967, + "learning_rate": 3.512385684222064e-06, + "loss": 1.0069, + "step": 2775 + }, + { + "epoch": 0.25034946115344725, + "grad_norm": 2.0759015342425577, + "learning_rate": 3.512003342166449e-06, + "loss": 1.1111, + "step": 2776 + }, + { + "epoch": 0.25043964467691754, + "grad_norm": 1.817672124318032, + "learning_rate": 3.511620871098371e-06, + "loss": 1.0545, + "step": 2777 + }, + { + "epoch": 0.25052982820038777, + "grad_norm": 1.6448092191732806, + "learning_rate": 3.511238271050465e-06, + "loss": 1.0454, + "step": 2778 + }, + { + "epoch": 0.25062001172385806, + "grad_norm": 1.4824309618926454, + "learning_rate": 3.5108555420553778e-06, + "loss": 0.9283, + "step": 2779 + }, + { + "epoch": 0.2507101952473283, + "grad_norm": 1.5427686344221188, + "learning_rate": 3.510472684145764e-06, + "loss": 0.8295, + "step": 2780 + }, + { + "epoch": 0.2508003787707986, + "grad_norm": 3.5531489469581796, + "learning_rate": 3.5100896973542926e-06, + "loss": 0.9059, + "step": 2781 + }, + { + "epoch": 0.2508905622942688, + "grad_norm": 2.860359731071897, + "learning_rate": 3.509706581713642e-06, + "loss": 1.0517, + "step": 2782 + }, + { + "epoch": 0.2509807458177391, + "grad_norm": 2.7699274766978594, + "learning_rate": 3.509323337256501e-06, + "loss": 1.0382, + "step": 2783 + }, + { + "epoch": 0.25107092934120934, + "grad_norm": 1.8218008350175667, + "learning_rate": 3.5089399640155703e-06, + "loss": 0.9656, + "step": 2784 + }, + { + "epoch": 0.2511611128646796, + "grad_norm": 1.409959309105779, + "learning_rate": 3.508556462023562e-06, + "loss": 1.0071, + "step": 2785 + }, + { + "epoch": 0.25125129638814986, + "grad_norm": 1.8984057890829884, + "learning_rate": 3.5081728313131984e-06, + "loss": 0.9994, + "step": 2786 + }, + { + "epoch": 0.25134147991162015, + "grad_norm": 2.313943876101853, + "learning_rate": 3.5077890719172125e-06, + "loss": 1.0024, + "step": 2787 + }, + { + "epoch": 0.25143166343509044, + "grad_norm": 1.582279666187953, + "learning_rate": 3.5074051838683497e-06, + "loss": 1.0135, + "step": 2788 + }, + { + "epoch": 0.25152184695856067, + "grad_norm": 1.643834562265916, + "learning_rate": 3.5070211671993643e-06, + "loss": 0.9386, + "step": 2789 + }, + { + "epoch": 0.25161203048203096, + "grad_norm": 1.6850989178564015, + "learning_rate": 3.5066370219430238e-06, + "loss": 1.0373, + "step": 2790 + }, + { + "epoch": 0.2517022140055012, + "grad_norm": 1.5657050258871812, + "learning_rate": 3.5062527481321044e-06, + "loss": 1.0135, + "step": 2791 + }, + { + "epoch": 0.2517923975289715, + "grad_norm": 2.379883012861604, + "learning_rate": 3.5058683457993954e-06, + "loss": 1.0271, + "step": 2792 + }, + { + "epoch": 0.2518825810524417, + "grad_norm": 0.7623576688033368, + "learning_rate": 3.5054838149776963e-06, + "loss": 0.7735, + "step": 2793 + }, + { + "epoch": 0.251972764575912, + "grad_norm": 2.4546394290017246, + "learning_rate": 3.505099155699816e-06, + "loss": 1.0471, + "step": 2794 + }, + { + "epoch": 0.25206294809938223, + "grad_norm": 1.5851752793780625, + "learning_rate": 3.5047143679985775e-06, + "loss": 0.8892, + "step": 2795 + }, + { + "epoch": 0.2521531316228525, + "grad_norm": 1.366289028856239, + "learning_rate": 3.5043294519068126e-06, + "loss": 1.0296, + "step": 2796 + }, + { + "epoch": 0.25224331514632276, + "grad_norm": 1.7640239917481664, + "learning_rate": 3.503944407457363e-06, + "loss": 0.9318, + "step": 2797 + }, + { + "epoch": 0.25233349866979304, + "grad_norm": 2.1185533685505136, + "learning_rate": 3.5035592346830846e-06, + "loss": 1.0259, + "step": 2798 + }, + { + "epoch": 0.2524236821932633, + "grad_norm": 2.4707300478708523, + "learning_rate": 3.503173933616841e-06, + "loss": 1.0815, + "step": 2799 + }, + { + "epoch": 0.25251386571673357, + "grad_norm": 1.8732166074156318, + "learning_rate": 3.50278850429151e-06, + "loss": 1.0871, + "step": 2800 + }, + { + "epoch": 0.2526040492402038, + "grad_norm": 1.9611182262530524, + "learning_rate": 3.502402946739977e-06, + "loss": 1.0362, + "step": 2801 + }, + { + "epoch": 0.2526942327636741, + "grad_norm": 1.54299625420502, + "learning_rate": 3.5020172609951405e-06, + "loss": 1.0375, + "step": 2802 + }, + { + "epoch": 0.2527844162871443, + "grad_norm": 1.7071303419599535, + "learning_rate": 3.501631447089909e-06, + "loss": 1.0341, + "step": 2803 + }, + { + "epoch": 0.2528745998106146, + "grad_norm": 4.285992522517529, + "learning_rate": 3.501245505057203e-06, + "loss": 0.9873, + "step": 2804 + }, + { + "epoch": 0.25296478333408484, + "grad_norm": 1.5172238126547999, + "learning_rate": 3.5008594349299526e-06, + "loss": 1.0879, + "step": 2805 + }, + { + "epoch": 0.25305496685755513, + "grad_norm": 1.7195577630681542, + "learning_rate": 3.500473236741099e-06, + "loss": 1.0021, + "step": 2806 + }, + { + "epoch": 0.25314515038102536, + "grad_norm": 1.5472954843245543, + "learning_rate": 3.500086910523596e-06, + "loss": 0.9706, + "step": 2807 + }, + { + "epoch": 0.25323533390449565, + "grad_norm": 1.4246369250692872, + "learning_rate": 3.499700456310406e-06, + "loss": 0.9624, + "step": 2808 + }, + { + "epoch": 0.2533255174279659, + "grad_norm": 1.4859498353399998, + "learning_rate": 3.499313874134504e-06, + "loss": 1.0649, + "step": 2809 + }, + { + "epoch": 0.2534157009514362, + "grad_norm": 1.8851113116033784, + "learning_rate": 3.498927164028875e-06, + "loss": 0.8836, + "step": 2810 + }, + { + "epoch": 0.25350588447490646, + "grad_norm": 1.6201054145518599, + "learning_rate": 3.498540326026515e-06, + "loss": 1.0498, + "step": 2811 + }, + { + "epoch": 0.2535960679983767, + "grad_norm": 1.5421938725023878, + "learning_rate": 3.4981533601604323e-06, + "loss": 1.0818, + "step": 2812 + }, + { + "epoch": 0.253686251521847, + "grad_norm": 2.6427321293989614, + "learning_rate": 3.4977662664636443e-06, + "loss": 1.0241, + "step": 2813 + }, + { + "epoch": 0.2537764350453172, + "grad_norm": 2.0786285937486553, + "learning_rate": 3.497379044969179e-06, + "loss": 0.9764, + "step": 2814 + }, + { + "epoch": 0.2538666185687875, + "grad_norm": 2.493084689235445, + "learning_rate": 3.4969916957100777e-06, + "loss": 1.0096, + "step": 2815 + }, + { + "epoch": 0.25395680209225774, + "grad_norm": 1.540008176930505, + "learning_rate": 3.4966042187193905e-06, + "loss": 1.0096, + "step": 2816 + }, + { + "epoch": 0.254046985615728, + "grad_norm": 1.4584541634048462, + "learning_rate": 3.496216614030179e-06, + "loss": 1.0131, + "step": 2817 + }, + { + "epoch": 0.25413716913919826, + "grad_norm": 1.9763141822888608, + "learning_rate": 3.495828881675516e-06, + "loss": 0.915, + "step": 2818 + }, + { + "epoch": 0.25422735266266855, + "grad_norm": 2.210803995044426, + "learning_rate": 3.4954410216884845e-06, + "loss": 1.0871, + "step": 2819 + }, + { + "epoch": 0.2543175361861388, + "grad_norm": 1.9241362504081898, + "learning_rate": 3.49505303410218e-06, + "loss": 1.0196, + "step": 2820 + }, + { + "epoch": 0.25440771970960907, + "grad_norm": 1.6349802184001911, + "learning_rate": 3.4946649189497067e-06, + "loss": 1.0438, + "step": 2821 + }, + { + "epoch": 0.2544979032330793, + "grad_norm": 3.6856238554742027, + "learning_rate": 3.4942766762641805e-06, + "loss": 1.0002, + "step": 2822 + }, + { + "epoch": 0.2545880867565496, + "grad_norm": 2.625565331754252, + "learning_rate": 3.49388830607873e-06, + "loss": 1.0336, + "step": 2823 + }, + { + "epoch": 0.2546782702800198, + "grad_norm": 1.8058957765653727, + "learning_rate": 3.493499808426491e-06, + "loss": 0.9768, + "step": 2824 + }, + { + "epoch": 0.2547684538034901, + "grad_norm": 2.540710008613709, + "learning_rate": 3.493111183340614e-06, + "loss": 1.0644, + "step": 2825 + }, + { + "epoch": 0.25485863732696035, + "grad_norm": 1.6731940796776783, + "learning_rate": 3.4927224308542576e-06, + "loss": 0.9167, + "step": 2826 + }, + { + "epoch": 0.25494882085043064, + "grad_norm": 1.5799024033885094, + "learning_rate": 3.4923335510005923e-06, + "loss": 1.0924, + "step": 2827 + }, + { + "epoch": 0.25503900437390087, + "grad_norm": 1.897923717911397, + "learning_rate": 3.4919445438128e-06, + "loss": 1.0333, + "step": 2828 + }, + { + "epoch": 0.25512918789737116, + "grad_norm": 1.8795837322965603, + "learning_rate": 3.491555409324073e-06, + "loss": 1.0917, + "step": 2829 + }, + { + "epoch": 0.2552193714208414, + "grad_norm": 0.6640988059497421, + "learning_rate": 3.4911661475676136e-06, + "loss": 0.7981, + "step": 2830 + }, + { + "epoch": 0.2553095549443117, + "grad_norm": 1.5737630406961234, + "learning_rate": 3.490776758576637e-06, + "loss": 0.9913, + "step": 2831 + }, + { + "epoch": 0.2553997384677819, + "grad_norm": 1.9381271700830007, + "learning_rate": 3.4903872423843668e-06, + "loss": 1.0607, + "step": 2832 + }, + { + "epoch": 0.2554899219912522, + "grad_norm": 1.7596035119514426, + "learning_rate": 3.4899975990240396e-06, + "loss": 0.9775, + "step": 2833 + }, + { + "epoch": 0.25558010551472243, + "grad_norm": 2.1301147390632993, + "learning_rate": 3.489607828528901e-06, + "loss": 1.0356, + "step": 2834 + }, + { + "epoch": 0.2556702890381927, + "grad_norm": 0.6576522650114208, + "learning_rate": 3.4892179309322093e-06, + "loss": 0.9062, + "step": 2835 + }, + { + "epoch": 0.255760472561663, + "grad_norm": 2.0039262380216596, + "learning_rate": 3.488827906267232e-06, + "loss": 0.9309, + "step": 2836 + }, + { + "epoch": 0.25585065608513324, + "grad_norm": 2.0096971743004857, + "learning_rate": 3.4884377545672485e-06, + "loss": 1.051, + "step": 2837 + }, + { + "epoch": 0.25594083960860353, + "grad_norm": 1.5988126640774563, + "learning_rate": 3.4880474758655485e-06, + "loss": 0.9845, + "step": 2838 + }, + { + "epoch": 0.25603102313207377, + "grad_norm": 2.0878860573500084, + "learning_rate": 3.487657070195433e-06, + "loss": 1.0817, + "step": 2839 + }, + { + "epoch": 0.25612120665554405, + "grad_norm": 2.8479650091957796, + "learning_rate": 3.487266537590213e-06, + "loss": 1.0896, + "step": 2840 + }, + { + "epoch": 0.2562113901790143, + "grad_norm": 1.6765699686842177, + "learning_rate": 3.4868758780832116e-06, + "loss": 0.9273, + "step": 2841 + }, + { + "epoch": 0.2563015737024846, + "grad_norm": 2.2165654133016948, + "learning_rate": 3.486485091707762e-06, + "loss": 1.0249, + "step": 2842 + }, + { + "epoch": 0.2563917572259548, + "grad_norm": 1.9564658317087782, + "learning_rate": 3.4860941784972077e-06, + "loss": 0.8889, + "step": 2843 + }, + { + "epoch": 0.2564819407494251, + "grad_norm": 2.154308827454377, + "learning_rate": 3.485703138484904e-06, + "loss": 1.1074, + "step": 2844 + }, + { + "epoch": 0.25657212427289533, + "grad_norm": 1.7612058988400905, + "learning_rate": 3.485311971704216e-06, + "loss": 1.0657, + "step": 2845 + }, + { + "epoch": 0.2566623077963656, + "grad_norm": 2.1884382143282295, + "learning_rate": 3.484920678188521e-06, + "loss": 1.043, + "step": 2846 + }, + { + "epoch": 0.25675249131983585, + "grad_norm": 1.8580521437818829, + "learning_rate": 3.4845292579712063e-06, + "loss": 1.0411, + "step": 2847 + }, + { + "epoch": 0.25684267484330614, + "grad_norm": 6.204664370845049, + "learning_rate": 3.484137711085669e-06, + "loss": 0.8963, + "step": 2848 + }, + { + "epoch": 0.2569328583667764, + "grad_norm": 1.4634742947566841, + "learning_rate": 3.4837460375653198e-06, + "loss": 0.9993, + "step": 2849 + }, + { + "epoch": 0.25702304189024666, + "grad_norm": 1.5505340148481657, + "learning_rate": 3.483354237443576e-06, + "loss": 1.0465, + "step": 2850 + }, + { + "epoch": 0.2571132254137169, + "grad_norm": 1.5212621563290163, + "learning_rate": 3.48296231075387e-06, + "loss": 0.9574, + "step": 2851 + }, + { + "epoch": 0.2572034089371872, + "grad_norm": 1.6287259887190157, + "learning_rate": 3.4825702575296433e-06, + "loss": 1.0315, + "step": 2852 + }, + { + "epoch": 0.2572935924606574, + "grad_norm": 0.6901186143814272, + "learning_rate": 3.482178077804347e-06, + "loss": 0.851, + "step": 2853 + }, + { + "epoch": 0.2573837759841277, + "grad_norm": 2.7449214163894236, + "learning_rate": 3.4817857716114443e-06, + "loss": 0.9781, + "step": 2854 + }, + { + "epoch": 0.25747395950759794, + "grad_norm": 1.7389333123466288, + "learning_rate": 3.4813933389844094e-06, + "loss": 0.9666, + "step": 2855 + }, + { + "epoch": 0.2575641430310682, + "grad_norm": 2.911124458800074, + "learning_rate": 3.4810007799567264e-06, + "loss": 1.018, + "step": 2856 + }, + { + "epoch": 0.25765432655453846, + "grad_norm": 2.3468735289791285, + "learning_rate": 3.480608094561891e-06, + "loss": 1.0325, + "step": 2857 + }, + { + "epoch": 0.25774451007800875, + "grad_norm": 3.6518521717565173, + "learning_rate": 3.4802152828334083e-06, + "loss": 1.0355, + "step": 2858 + }, + { + "epoch": 0.25783469360147904, + "grad_norm": 1.7800865389114637, + "learning_rate": 3.479822344804796e-06, + "loss": 0.9514, + "step": 2859 + }, + { + "epoch": 0.25792487712494927, + "grad_norm": 2.0607060809348705, + "learning_rate": 3.479429280509582e-06, + "loss": 1.0671, + "step": 2860 + }, + { + "epoch": 0.25801506064841956, + "grad_norm": 1.6727091054391579, + "learning_rate": 3.4790360899813038e-06, + "loss": 1.0551, + "step": 2861 + }, + { + "epoch": 0.2581052441718898, + "grad_norm": 2.489066151378609, + "learning_rate": 3.4786427732535115e-06, + "loss": 0.9468, + "step": 2862 + }, + { + "epoch": 0.2581954276953601, + "grad_norm": 2.0040080916494536, + "learning_rate": 3.478249330359764e-06, + "loss": 1.0583, + "step": 2863 + }, + { + "epoch": 0.2582856112188303, + "grad_norm": 1.638562095338517, + "learning_rate": 3.4778557613336333e-06, + "loss": 0.9487, + "step": 2864 + }, + { + "epoch": 0.2583757947423006, + "grad_norm": 1.700104230602559, + "learning_rate": 3.4774620662087004e-06, + "loss": 1.0313, + "step": 2865 + }, + { + "epoch": 0.25846597826577083, + "grad_norm": 3.2081496115921535, + "learning_rate": 3.477068245018557e-06, + "loss": 1.0172, + "step": 2866 + }, + { + "epoch": 0.2585561617892411, + "grad_norm": 1.612002403901156, + "learning_rate": 3.476674297796807e-06, + "loss": 1.0171, + "step": 2867 + }, + { + "epoch": 0.25864634531271136, + "grad_norm": 1.8018843112775103, + "learning_rate": 3.4762802245770627e-06, + "loss": 1.0576, + "step": 2868 + }, + { + "epoch": 0.25873652883618165, + "grad_norm": 3.609103634363954, + "learning_rate": 3.4758860253929497e-06, + "loss": 0.9646, + "step": 2869 + }, + { + "epoch": 0.2588267123596519, + "grad_norm": 1.7043361731856916, + "learning_rate": 3.4754917002781038e-06, + "loss": 1.0324, + "step": 2870 + }, + { + "epoch": 0.25891689588312217, + "grad_norm": 1.673689454787687, + "learning_rate": 3.475097249266169e-06, + "loss": 1.0519, + "step": 2871 + }, + { + "epoch": 0.2590070794065924, + "grad_norm": 1.6107652455540473, + "learning_rate": 3.4747026723908044e-06, + "loss": 0.9973, + "step": 2872 + }, + { + "epoch": 0.2590972629300627, + "grad_norm": 1.6167966950944799, + "learning_rate": 3.474307969685676e-06, + "loss": 0.9936, + "step": 2873 + }, + { + "epoch": 0.2591874464535329, + "grad_norm": 1.5119698880189383, + "learning_rate": 3.473913141184462e-06, + "loss": 0.9609, + "step": 2874 + }, + { + "epoch": 0.2592776299770032, + "grad_norm": 2.2554273844018597, + "learning_rate": 3.4735181869208523e-06, + "loss": 0.8876, + "step": 2875 + }, + { + "epoch": 0.25936781350047344, + "grad_norm": 2.1355459282682645, + "learning_rate": 3.473123106928546e-06, + "loss": 0.8964, + "step": 2876 + }, + { + "epoch": 0.25945799702394373, + "grad_norm": 1.6239487842460951, + "learning_rate": 3.4727279012412533e-06, + "loss": 0.9719, + "step": 2877 + }, + { + "epoch": 0.25954818054741396, + "grad_norm": 0.5933163715824128, + "learning_rate": 3.4723325698926953e-06, + "loss": 0.8133, + "step": 2878 + }, + { + "epoch": 0.25963836407088425, + "grad_norm": 1.838101479315968, + "learning_rate": 3.4719371129166045e-06, + "loss": 1.0184, + "step": 2879 + }, + { + "epoch": 0.2597285475943545, + "grad_norm": 2.0110612405854305, + "learning_rate": 3.471541530346723e-06, + "loss": 0.9359, + "step": 2880 + }, + { + "epoch": 0.2598187311178248, + "grad_norm": 2.5462887329091988, + "learning_rate": 3.4711458222168037e-06, + "loss": 1.0749, + "step": 2881 + }, + { + "epoch": 0.259908914641295, + "grad_norm": 1.5284541549820314, + "learning_rate": 3.4707499885606114e-06, + "loss": 1.0553, + "step": 2882 + }, + { + "epoch": 0.2599990981647653, + "grad_norm": 2.000561277786306, + "learning_rate": 3.4703540294119204e-06, + "loss": 1.0527, + "step": 2883 + }, + { + "epoch": 0.2600892816882356, + "grad_norm": 1.6321872810072937, + "learning_rate": 3.4699579448045163e-06, + "loss": 1.0658, + "step": 2884 + }, + { + "epoch": 0.2601794652117058, + "grad_norm": 1.4719253466635163, + "learning_rate": 3.4695617347721947e-06, + "loss": 0.9023, + "step": 2885 + }, + { + "epoch": 0.2602696487351761, + "grad_norm": 3.22372200653832, + "learning_rate": 3.469165399348763e-06, + "loss": 0.9842, + "step": 2886 + }, + { + "epoch": 0.26035983225864634, + "grad_norm": 2.1378836794935028, + "learning_rate": 3.4687689385680384e-06, + "loss": 0.994, + "step": 2887 + }, + { + "epoch": 0.26045001578211663, + "grad_norm": 1.5025243974885125, + "learning_rate": 3.4683723524638494e-06, + "loss": 0.9955, + "step": 2888 + }, + { + "epoch": 0.26054019930558686, + "grad_norm": 1.7089483813173891, + "learning_rate": 3.4679756410700354e-06, + "loss": 1.0835, + "step": 2889 + }, + { + "epoch": 0.26063038282905715, + "grad_norm": 1.97655043767935, + "learning_rate": 3.4675788044204445e-06, + "loss": 1.009, + "step": 2890 + }, + { + "epoch": 0.2607205663525274, + "grad_norm": 2.222774026495669, + "learning_rate": 3.467181842548938e-06, + "loss": 0.9489, + "step": 2891 + }, + { + "epoch": 0.26081074987599767, + "grad_norm": 1.5628063664490466, + "learning_rate": 3.466784755489387e-06, + "loss": 1.01, + "step": 2892 + }, + { + "epoch": 0.2609009333994679, + "grad_norm": 1.505446321709352, + "learning_rate": 3.4663875432756726e-06, + "loss": 1.1314, + "step": 2893 + }, + { + "epoch": 0.2609911169229382, + "grad_norm": 2.088066814471701, + "learning_rate": 3.465990205941687e-06, + "loss": 1.1806, + "step": 2894 + }, + { + "epoch": 0.2610813004464084, + "grad_norm": 2.101377911597496, + "learning_rate": 3.465592743521335e-06, + "loss": 0.9989, + "step": 2895 + }, + { + "epoch": 0.2611714839698787, + "grad_norm": 5.887762613777105, + "learning_rate": 3.465195156048528e-06, + "loss": 1.065, + "step": 2896 + }, + { + "epoch": 0.26126166749334895, + "grad_norm": 0.5682300276244227, + "learning_rate": 3.464797443557191e-06, + "loss": 0.8106, + "step": 2897 + }, + { + "epoch": 0.26135185101681924, + "grad_norm": 3.801920360399314, + "learning_rate": 3.46439960608126e-06, + "loss": 0.9911, + "step": 2898 + }, + { + "epoch": 0.26144203454028947, + "grad_norm": 3.923102319787675, + "learning_rate": 3.4640016436546797e-06, + "loss": 0.9493, + "step": 2899 + }, + { + "epoch": 0.26153221806375976, + "grad_norm": 3.0016171547214454, + "learning_rate": 3.4636035563114065e-06, + "loss": 0.994, + "step": 2900 + }, + { + "epoch": 0.26162240158723, + "grad_norm": 1.548281097981862, + "learning_rate": 3.4632053440854085e-06, + "loss": 0.9407, + "step": 2901 + }, + { + "epoch": 0.2617125851107003, + "grad_norm": 1.7065546343708933, + "learning_rate": 3.462807007010662e-06, + "loss": 1.0685, + "step": 2902 + }, + { + "epoch": 0.2618027686341705, + "grad_norm": 1.7885003018654901, + "learning_rate": 3.462408545121155e-06, + "loss": 1.0412, + "step": 2903 + }, + { + "epoch": 0.2618929521576408, + "grad_norm": 2.8909398912628776, + "learning_rate": 3.4620099584508883e-06, + "loss": 1.0736, + "step": 2904 + }, + { + "epoch": 0.26198313568111103, + "grad_norm": 1.5044975406452943, + "learning_rate": 3.46161124703387e-06, + "loss": 0.9739, + "step": 2905 + }, + { + "epoch": 0.2620733192045813, + "grad_norm": 1.9381233566164038, + "learning_rate": 3.461212410904122e-06, + "loss": 1.0079, + "step": 2906 + }, + { + "epoch": 0.2621635027280516, + "grad_norm": 2.09039755640336, + "learning_rate": 3.4608134500956726e-06, + "loss": 1.072, + "step": 2907 + }, + { + "epoch": 0.26225368625152184, + "grad_norm": 1.4710745576729916, + "learning_rate": 3.4604143646425655e-06, + "loss": 0.9995, + "step": 2908 + }, + { + "epoch": 0.26234386977499213, + "grad_norm": 1.3423447469095502, + "learning_rate": 3.460015154578852e-06, + "loss": 1.0838, + "step": 2909 + }, + { + "epoch": 0.26243405329846237, + "grad_norm": 1.6908054474094232, + "learning_rate": 3.459615819938595e-06, + "loss": 0.9872, + "step": 2910 + }, + { + "epoch": 0.26252423682193265, + "grad_norm": 1.7589940825856722, + "learning_rate": 3.4592163607558684e-06, + "loss": 0.9767, + "step": 2911 + }, + { + "epoch": 0.2626144203454029, + "grad_norm": 1.405946953227255, + "learning_rate": 3.4588167770647553e-06, + "loss": 1.0185, + "step": 2912 + }, + { + "epoch": 0.2627046038688732, + "grad_norm": 2.101808781710817, + "learning_rate": 3.458417068899351e-06, + "loss": 1.1057, + "step": 2913 + }, + { + "epoch": 0.2627947873923434, + "grad_norm": 1.8319456816579043, + "learning_rate": 3.4580172362937612e-06, + "loss": 0.9856, + "step": 2914 + }, + { + "epoch": 0.2628849709158137, + "grad_norm": 1.3810729815253708, + "learning_rate": 3.457617279282101e-06, + "loss": 0.9803, + "step": 2915 + }, + { + "epoch": 0.26297515443928393, + "grad_norm": 2.1420857812283396, + "learning_rate": 3.4572171978984975e-06, + "loss": 0.9925, + "step": 2916 + }, + { + "epoch": 0.2630653379627542, + "grad_norm": 1.7052443031178963, + "learning_rate": 3.456816992177088e-06, + "loss": 1.0101, + "step": 2917 + }, + { + "epoch": 0.26315552148622445, + "grad_norm": 1.5083078153921037, + "learning_rate": 3.4564166621520193e-06, + "loss": 0.9851, + "step": 2918 + }, + { + "epoch": 0.26324570500969474, + "grad_norm": 1.9419254092379412, + "learning_rate": 3.4560162078574507e-06, + "loss": 1.0058, + "step": 2919 + }, + { + "epoch": 0.263335888533165, + "grad_norm": 1.6731294578176512, + "learning_rate": 3.455615629327551e-06, + "loss": 1.0164, + "step": 2920 + }, + { + "epoch": 0.26342607205663526, + "grad_norm": 1.7809114636407124, + "learning_rate": 3.4552149265964994e-06, + "loss": 0.9881, + "step": 2921 + }, + { + "epoch": 0.2635162555801055, + "grad_norm": 1.9270130110273203, + "learning_rate": 3.4548140996984866e-06, + "loss": 0.9626, + "step": 2922 + }, + { + "epoch": 0.2636064391035758, + "grad_norm": 1.802167776551195, + "learning_rate": 3.4544131486677124e-06, + "loss": 1.0352, + "step": 2923 + }, + { + "epoch": 0.263696622627046, + "grad_norm": 1.6636147769857794, + "learning_rate": 3.454012073538389e-06, + "loss": 0.9525, + "step": 2924 + }, + { + "epoch": 0.2637868061505163, + "grad_norm": 1.8431907710890725, + "learning_rate": 3.453610874344738e-06, + "loss": 1.0196, + "step": 2925 + }, + { + "epoch": 0.26387698967398654, + "grad_norm": 1.7071283867126377, + "learning_rate": 3.453209551120993e-06, + "loss": 1.0515, + "step": 2926 + }, + { + "epoch": 0.26396717319745683, + "grad_norm": 1.7914105461458378, + "learning_rate": 3.452808103901395e-06, + "loss": 0.9861, + "step": 2927 + }, + { + "epoch": 0.26405735672092706, + "grad_norm": 1.940735238548473, + "learning_rate": 3.4524065327202e-06, + "loss": 1.0853, + "step": 2928 + }, + { + "epoch": 0.26414754024439735, + "grad_norm": 0.6446736496340383, + "learning_rate": 3.4520048376116702e-06, + "loss": 0.7633, + "step": 2929 + }, + { + "epoch": 0.26423772376786764, + "grad_norm": 1.927941215572647, + "learning_rate": 3.4516030186100817e-06, + "loss": 1.0498, + "step": 2930 + }, + { + "epoch": 0.26432790729133787, + "grad_norm": 1.6237351556793935, + "learning_rate": 3.4512010757497197e-06, + "loss": 0.9304, + "step": 2931 + }, + { + "epoch": 0.26441809081480816, + "grad_norm": 2.937795096651097, + "learning_rate": 3.4507990090648804e-06, + "loss": 0.9341, + "step": 2932 + }, + { + "epoch": 0.2645082743382784, + "grad_norm": 1.7518737162579539, + "learning_rate": 3.4503968185898696e-06, + "loss": 1.0388, + "step": 2933 + }, + { + "epoch": 0.2645984578617487, + "grad_norm": 1.6357633219250347, + "learning_rate": 3.4499945043590047e-06, + "loss": 1.0156, + "step": 2934 + }, + { + "epoch": 0.2646886413852189, + "grad_norm": 1.7189769941768853, + "learning_rate": 3.4495920664066137e-06, + "loss": 1.029, + "step": 2935 + }, + { + "epoch": 0.2647788249086892, + "grad_norm": 2.4864561373111362, + "learning_rate": 3.449189504767035e-06, + "loss": 0.9987, + "step": 2936 + }, + { + "epoch": 0.26486900843215944, + "grad_norm": 1.7294305653635875, + "learning_rate": 3.4487868194746163e-06, + "loss": 1.0376, + "step": 2937 + }, + { + "epoch": 0.2649591919556297, + "grad_norm": 1.5163054364441109, + "learning_rate": 3.4483840105637175e-06, + "loss": 1.0549, + "step": 2938 + }, + { + "epoch": 0.26504937547909996, + "grad_norm": 2.731317958379142, + "learning_rate": 3.4479810780687097e-06, + "loss": 0.9961, + "step": 2939 + }, + { + "epoch": 0.26513955900257025, + "grad_norm": 1.5102199645595902, + "learning_rate": 3.4475780220239714e-06, + "loss": 1.0015, + "step": 2940 + }, + { + "epoch": 0.2652297425260405, + "grad_norm": 1.343975181341921, + "learning_rate": 3.4471748424638948e-06, + "loss": 1.0007, + "step": 2941 + }, + { + "epoch": 0.26531992604951077, + "grad_norm": 0.5972879559051592, + "learning_rate": 3.4467715394228803e-06, + "loss": 0.8302, + "step": 2942 + }, + { + "epoch": 0.265410109572981, + "grad_norm": 1.4218930253783182, + "learning_rate": 3.4463681129353413e-06, + "loss": 1.0157, + "step": 2943 + }, + { + "epoch": 0.2655002930964513, + "grad_norm": 1.7345452439891658, + "learning_rate": 3.4459645630357e-06, + "loss": 1.0028, + "step": 2944 + }, + { + "epoch": 0.2655904766199215, + "grad_norm": 12.713210340550727, + "learning_rate": 3.4455608897583884e-06, + "loss": 1.0479, + "step": 2945 + }, + { + "epoch": 0.2656806601433918, + "grad_norm": 1.900302990797227, + "learning_rate": 3.4451570931378514e-06, + "loss": 1.014, + "step": 2946 + }, + { + "epoch": 0.26577084366686204, + "grad_norm": 1.5471401565325904, + "learning_rate": 3.444753173208543e-06, + "loss": 1.0199, + "step": 2947 + }, + { + "epoch": 0.26586102719033233, + "grad_norm": 1.8138031714208607, + "learning_rate": 3.444349130004927e-06, + "loss": 1.007, + "step": 2948 + }, + { + "epoch": 0.26595121071380257, + "grad_norm": 2.190180090375581, + "learning_rate": 3.4439449635614794e-06, + "loss": 0.9366, + "step": 2949 + }, + { + "epoch": 0.26604139423727285, + "grad_norm": 1.5103699337098389, + "learning_rate": 3.4435406739126854e-06, + "loss": 1.0392, + "step": 2950 + }, + { + "epoch": 0.2661315777607431, + "grad_norm": 9.936370977408382, + "learning_rate": 3.443136261093042e-06, + "loss": 0.8225, + "step": 2951 + }, + { + "epoch": 0.2662217612842134, + "grad_norm": 1.987554331102849, + "learning_rate": 3.4427317251370553e-06, + "loss": 0.9589, + "step": 2952 + }, + { + "epoch": 0.2663119448076836, + "grad_norm": 2.5047704957762558, + "learning_rate": 3.4423270660792422e-06, + "loss": 0.9271, + "step": 2953 + }, + { + "epoch": 0.2664021283311539, + "grad_norm": 1.7237644220956327, + "learning_rate": 3.4419222839541314e-06, + "loss": 0.9036, + "step": 2954 + }, + { + "epoch": 0.2664923118546242, + "grad_norm": 1.47456183956909, + "learning_rate": 3.4415173787962607e-06, + "loss": 0.9865, + "step": 2955 + }, + { + "epoch": 0.2665824953780944, + "grad_norm": 2.839870951232348, + "learning_rate": 3.4411123506401783e-06, + "loss": 0.8757, + "step": 2956 + }, + { + "epoch": 0.2666726789015647, + "grad_norm": 2.162119850746697, + "learning_rate": 3.440707199520444e-06, + "loss": 0.8615, + "step": 2957 + }, + { + "epoch": 0.26676286242503494, + "grad_norm": 1.4439767799782421, + "learning_rate": 3.440301925471628e-06, + "loss": 0.9732, + "step": 2958 + }, + { + "epoch": 0.26685304594850523, + "grad_norm": 1.7802572328098891, + "learning_rate": 3.43989652852831e-06, + "loss": 1.0475, + "step": 2959 + }, + { + "epoch": 0.26694322947197546, + "grad_norm": 9.717602628338183, + "learning_rate": 3.4394910087250804e-06, + "loss": 1.0727, + "step": 2960 + }, + { + "epoch": 0.26703341299544575, + "grad_norm": 1.4024590150009106, + "learning_rate": 3.4390853660965405e-06, + "loss": 1.0282, + "step": 2961 + }, + { + "epoch": 0.267123596518916, + "grad_norm": 2.00390315659788, + "learning_rate": 3.438679600677302e-06, + "loss": 1.0892, + "step": 2962 + }, + { + "epoch": 0.2672137800423863, + "grad_norm": 1.4423439676662542, + "learning_rate": 3.4382737125019874e-06, + "loss": 0.9671, + "step": 2963 + }, + { + "epoch": 0.2673039635658565, + "grad_norm": 2.4720099449698387, + "learning_rate": 3.4378677016052294e-06, + "loss": 0.9771, + "step": 2964 + }, + { + "epoch": 0.2673941470893268, + "grad_norm": 1.6457666391629358, + "learning_rate": 3.43746156802167e-06, + "loss": 1.0323, + "step": 2965 + }, + { + "epoch": 0.267484330612797, + "grad_norm": 1.7439890215729408, + "learning_rate": 3.4370553117859643e-06, + "loss": 0.948, + "step": 2966 + }, + { + "epoch": 0.2675745141362673, + "grad_norm": 1.6613191010144643, + "learning_rate": 3.4366489329327754e-06, + "loss": 1.0412, + "step": 2967 + }, + { + "epoch": 0.26766469765973755, + "grad_norm": 2.061884354694164, + "learning_rate": 3.4362424314967777e-06, + "loss": 1.0335, + "step": 2968 + }, + { + "epoch": 0.26775488118320784, + "grad_norm": 2.433293356196908, + "learning_rate": 3.4358358075126567e-06, + "loss": 1.0379, + "step": 2969 + }, + { + "epoch": 0.26784506470667807, + "grad_norm": 1.6328629481356998, + "learning_rate": 3.4354290610151077e-06, + "loss": 1.0308, + "step": 2970 + }, + { + "epoch": 0.26793524823014836, + "grad_norm": 1.877720956613565, + "learning_rate": 3.4350221920388354e-06, + "loss": 1.0167, + "step": 2971 + }, + { + "epoch": 0.2680254317536186, + "grad_norm": 1.501761832289538, + "learning_rate": 3.4346152006185574e-06, + "loss": 1.1113, + "step": 2972 + }, + { + "epoch": 0.2681156152770889, + "grad_norm": 1.8635101497651114, + "learning_rate": 3.4342080867890006e-06, + "loss": 1.1135, + "step": 2973 + }, + { + "epoch": 0.2682057988005591, + "grad_norm": 1.8155828763241273, + "learning_rate": 3.4338008505849016e-06, + "loss": 0.9311, + "step": 2974 + }, + { + "epoch": 0.2682959823240294, + "grad_norm": 1.5836050235476866, + "learning_rate": 3.433393492041008e-06, + "loss": 1.0199, + "step": 2975 + }, + { + "epoch": 0.26838616584749964, + "grad_norm": 1.5484658740049015, + "learning_rate": 3.432986011192078e-06, + "loss": 1.063, + "step": 2976 + }, + { + "epoch": 0.2684763493709699, + "grad_norm": 1.8437900862134613, + "learning_rate": 3.4325784080728796e-06, + "loss": 1.0779, + "step": 2977 + }, + { + "epoch": 0.2685665328944402, + "grad_norm": 1.5190654913600716, + "learning_rate": 3.4321706827181926e-06, + "loss": 1.0385, + "step": 2978 + }, + { + "epoch": 0.26865671641791045, + "grad_norm": 1.370611426568477, + "learning_rate": 3.4317628351628064e-06, + "loss": 0.9631, + "step": 2979 + }, + { + "epoch": 0.26874689994138073, + "grad_norm": 1.8544482185249687, + "learning_rate": 3.43135486544152e-06, + "loss": 0.9924, + "step": 2980 + }, + { + "epoch": 0.26883708346485097, + "grad_norm": 1.9281016640510698, + "learning_rate": 3.4309467735891442e-06, + "loss": 1.0507, + "step": 2981 + }, + { + "epoch": 0.26892726698832126, + "grad_norm": 2.1579274825744372, + "learning_rate": 3.4305385596405e-06, + "loss": 1.0974, + "step": 2982 + }, + { + "epoch": 0.2690174505117915, + "grad_norm": 1.6487742437123178, + "learning_rate": 3.4301302236304174e-06, + "loss": 0.9756, + "step": 2983 + }, + { + "epoch": 0.2691076340352618, + "grad_norm": 0.6013125854165413, + "learning_rate": 3.429721765593739e-06, + "loss": 0.7851, + "step": 2984 + }, + { + "epoch": 0.269197817558732, + "grad_norm": 1.4203846425282494, + "learning_rate": 3.4293131855653155e-06, + "loss": 1.0349, + "step": 2985 + }, + { + "epoch": 0.2692880010822023, + "grad_norm": 2.587822172591143, + "learning_rate": 3.4289044835800102e-06, + "loss": 0.9488, + "step": 2986 + }, + { + "epoch": 0.26937818460567253, + "grad_norm": 1.7966424542755437, + "learning_rate": 3.4284956596726953e-06, + "loss": 1.0146, + "step": 2987 + }, + { + "epoch": 0.2694683681291428, + "grad_norm": 1.5679689446312783, + "learning_rate": 3.4280867138782544e-06, + "loss": 1.0051, + "step": 2988 + }, + { + "epoch": 0.26955855165261305, + "grad_norm": 2.0443323081587397, + "learning_rate": 3.4276776462315803e-06, + "loss": 1.0134, + "step": 2989 + }, + { + "epoch": 0.26964873517608334, + "grad_norm": 2.4424631501368883, + "learning_rate": 3.427268456767578e-06, + "loss": 1.0122, + "step": 2990 + }, + { + "epoch": 0.2697389186995536, + "grad_norm": 1.735653380346365, + "learning_rate": 3.42685914552116e-06, + "loss": 1.052, + "step": 2991 + }, + { + "epoch": 0.26982910222302386, + "grad_norm": 1.6722041724038101, + "learning_rate": 3.426449712527253e-06, + "loss": 0.9856, + "step": 2992 + }, + { + "epoch": 0.2699192857464941, + "grad_norm": 1.6154627794177, + "learning_rate": 3.4260401578207904e-06, + "loss": 0.9242, + "step": 2993 + }, + { + "epoch": 0.2700094692699644, + "grad_norm": 3.1402387239132095, + "learning_rate": 3.4256304814367185e-06, + "loss": 0.9773, + "step": 2994 + }, + { + "epoch": 0.2700996527934346, + "grad_norm": 1.669841492263097, + "learning_rate": 3.4252206834099936e-06, + "loss": 0.9697, + "step": 2995 + }, + { + "epoch": 0.2701898363169049, + "grad_norm": 1.600842123144835, + "learning_rate": 3.424810763775581e-06, + "loss": 1.0083, + "step": 2996 + }, + { + "epoch": 0.27028001984037514, + "grad_norm": 1.5144574889207987, + "learning_rate": 3.4244007225684587e-06, + "loss": 0.9507, + "step": 2997 + }, + { + "epoch": 0.27037020336384543, + "grad_norm": 1.5159809441790533, + "learning_rate": 3.4239905598236115e-06, + "loss": 1.019, + "step": 2998 + }, + { + "epoch": 0.27046038688731566, + "grad_norm": 1.64023294759944, + "learning_rate": 3.4235802755760386e-06, + "loss": 1.0369, + "step": 2999 + }, + { + "epoch": 0.27055057041078595, + "grad_norm": 0.8431683054011594, + "learning_rate": 3.4231698698607464e-06, + "loss": 0.932, + "step": 3000 + }, + { + "epoch": 0.2706407539342562, + "grad_norm": 4.599633218820895, + "learning_rate": 3.4227593427127543e-06, + "loss": 1.0691, + "step": 3001 + }, + { + "epoch": 0.2707309374577265, + "grad_norm": 1.9713934453950666, + "learning_rate": 3.42234869416709e-06, + "loss": 0.9844, + "step": 3002 + }, + { + "epoch": 0.27082112098119676, + "grad_norm": 2.126395496590971, + "learning_rate": 3.421937924258792e-06, + "loss": 1.0121, + "step": 3003 + }, + { + "epoch": 0.270911304504667, + "grad_norm": 1.7900867445401956, + "learning_rate": 3.4215270330229096e-06, + "loss": 1.0545, + "step": 3004 + }, + { + "epoch": 0.2710014880281373, + "grad_norm": 1.8356328183852852, + "learning_rate": 3.421116020494503e-06, + "loss": 0.9756, + "step": 3005 + }, + { + "epoch": 0.2710916715516075, + "grad_norm": 1.4104074270301694, + "learning_rate": 3.420704886708642e-06, + "loss": 0.957, + "step": 3006 + }, + { + "epoch": 0.2711818550750778, + "grad_norm": 1.3201760215872496, + "learning_rate": 3.4202936317004056e-06, + "loss": 0.9608, + "step": 3007 + }, + { + "epoch": 0.27127203859854804, + "grad_norm": 1.9292663238139296, + "learning_rate": 3.4198822555048856e-06, + "loss": 1.0232, + "step": 3008 + }, + { + "epoch": 0.2713622221220183, + "grad_norm": 2.0729164877528845, + "learning_rate": 3.419470758157182e-06, + "loss": 1.0973, + "step": 3009 + }, + { + "epoch": 0.27145240564548856, + "grad_norm": 1.7121581857187098, + "learning_rate": 3.4190591396924068e-06, + "loss": 1.0056, + "step": 3010 + }, + { + "epoch": 0.27154258916895885, + "grad_norm": 0.672772894303546, + "learning_rate": 3.418647400145681e-06, + "loss": 0.8523, + "step": 3011 + }, + { + "epoch": 0.2716327726924291, + "grad_norm": 0.7064632220498924, + "learning_rate": 3.4182355395521367e-06, + "loss": 0.8129, + "step": 3012 + }, + { + "epoch": 0.27172295621589937, + "grad_norm": 1.5288467231900762, + "learning_rate": 3.417823557946916e-06, + "loss": 1.0096, + "step": 3013 + }, + { + "epoch": 0.2718131397393696, + "grad_norm": 1.5609473334108666, + "learning_rate": 3.417411455365172e-06, + "loss": 0.9053, + "step": 3014 + }, + { + "epoch": 0.2719033232628399, + "grad_norm": 2.246871680839538, + "learning_rate": 3.416999231842066e-06, + "loss": 0.9226, + "step": 3015 + }, + { + "epoch": 0.2719935067863101, + "grad_norm": 2.8404458101700683, + "learning_rate": 3.416586887412773e-06, + "loss": 1.0608, + "step": 3016 + }, + { + "epoch": 0.2720836903097804, + "grad_norm": 1.6340274944630315, + "learning_rate": 3.416174422112476e-06, + "loss": 0.9767, + "step": 3017 + }, + { + "epoch": 0.27217387383325065, + "grad_norm": 0.608310111355969, + "learning_rate": 3.4157618359763687e-06, + "loss": 0.7766, + "step": 3018 + }, + { + "epoch": 0.27226405735672093, + "grad_norm": 1.7988839239191112, + "learning_rate": 3.4153491290396542e-06, + "loss": 0.933, + "step": 3019 + }, + { + "epoch": 0.27235424088019117, + "grad_norm": 1.848904339355678, + "learning_rate": 3.4149363013375485e-06, + "loss": 1.025, + "step": 3020 + }, + { + "epoch": 0.27244442440366146, + "grad_norm": 0.6975572120623982, + "learning_rate": 3.414523352905276e-06, + "loss": 0.7932, + "step": 3021 + }, + { + "epoch": 0.2725346079271317, + "grad_norm": 1.5693972373246803, + "learning_rate": 3.414110283778071e-06, + "loss": 0.9846, + "step": 3022 + }, + { + "epoch": 0.272624791450602, + "grad_norm": 2.619053599528301, + "learning_rate": 3.4136970939911797e-06, + "loss": 0.9579, + "step": 3023 + }, + { + "epoch": 0.2727149749740722, + "grad_norm": 2.263981821795972, + "learning_rate": 3.413283783579857e-06, + "loss": 0.9801, + "step": 3024 + }, + { + "epoch": 0.2728051584975425, + "grad_norm": 2.346482476985717, + "learning_rate": 3.412870352579369e-06, + "loss": 0.9422, + "step": 3025 + }, + { + "epoch": 0.2728953420210128, + "grad_norm": 0.6504690678577623, + "learning_rate": 3.4124568010249915e-06, + "loss": 0.8337, + "step": 3026 + }, + { + "epoch": 0.272985525544483, + "grad_norm": 2.1588919114393663, + "learning_rate": 3.4120431289520124e-06, + "loss": 0.9939, + "step": 3027 + }, + { + "epoch": 0.2730757090679533, + "grad_norm": 1.833507276721888, + "learning_rate": 3.4116293363957276e-06, + "loss": 1.0178, + "step": 3028 + }, + { + "epoch": 0.27316589259142354, + "grad_norm": 1.845292819804031, + "learning_rate": 3.4112154233914438e-06, + "loss": 1.021, + "step": 3029 + }, + { + "epoch": 0.27325607611489383, + "grad_norm": 2.038596615239996, + "learning_rate": 3.410801389974479e-06, + "loss": 0.9175, + "step": 3030 + }, + { + "epoch": 0.27334625963836406, + "grad_norm": 2.856945569175145, + "learning_rate": 3.410387236180161e-06, + "loss": 1.0393, + "step": 3031 + }, + { + "epoch": 0.27343644316183435, + "grad_norm": 1.8966250460750458, + "learning_rate": 3.409972962043826e-06, + "loss": 0.8919, + "step": 3032 + }, + { + "epoch": 0.2735266266853046, + "grad_norm": 2.7006282852460095, + "learning_rate": 3.4095585676008234e-06, + "loss": 1.0654, + "step": 3033 + }, + { + "epoch": 0.2736168102087749, + "grad_norm": 1.7188018790998478, + "learning_rate": 3.4091440528865125e-06, + "loss": 1.0041, + "step": 3034 + }, + { + "epoch": 0.2737069937322451, + "grad_norm": 2.3001742379906718, + "learning_rate": 3.4087294179362606e-06, + "loss": 1.007, + "step": 3035 + }, + { + "epoch": 0.2737971772557154, + "grad_norm": 1.5565010113490734, + "learning_rate": 3.4083146627854474e-06, + "loss": 1.0541, + "step": 3036 + }, + { + "epoch": 0.27388736077918563, + "grad_norm": 1.911631570253859, + "learning_rate": 3.4078997874694614e-06, + "loss": 0.935, + "step": 3037 + }, + { + "epoch": 0.2739775443026559, + "grad_norm": 1.669998256431172, + "learning_rate": 3.407484792023703e-06, + "loss": 0.9008, + "step": 3038 + }, + { + "epoch": 0.27406772782612615, + "grad_norm": 2.624606148373891, + "learning_rate": 3.407069676483581e-06, + "loss": 0.9785, + "step": 3039 + }, + { + "epoch": 0.27415791134959644, + "grad_norm": 1.5682975025252708, + "learning_rate": 3.406654440884516e-06, + "loss": 1.0473, + "step": 3040 + }, + { + "epoch": 0.2742480948730667, + "grad_norm": 4.742435455124351, + "learning_rate": 3.4062390852619372e-06, + "loss": 1.0369, + "step": 3041 + }, + { + "epoch": 0.27433827839653696, + "grad_norm": 3.1778254575049742, + "learning_rate": 3.4058236096512867e-06, + "loss": 1.0059, + "step": 3042 + }, + { + "epoch": 0.2744284619200072, + "grad_norm": 2.2960586654199515, + "learning_rate": 3.405408014088013e-06, + "loss": 0.912, + "step": 3043 + }, + { + "epoch": 0.2745186454434775, + "grad_norm": 2.3568563596827454, + "learning_rate": 3.404992298607579e-06, + "loss": 0.8974, + "step": 3044 + }, + { + "epoch": 0.2746088289669477, + "grad_norm": 2.335660477864642, + "learning_rate": 3.4045764632454547e-06, + "loss": 0.9878, + "step": 3045 + }, + { + "epoch": 0.274699012490418, + "grad_norm": 3.2263131772713063, + "learning_rate": 3.4041605080371223e-06, + "loss": 0.9665, + "step": 3046 + }, + { + "epoch": 0.27478919601388824, + "grad_norm": 1.8746186186428824, + "learning_rate": 3.4037444330180726e-06, + "loss": 1.0206, + "step": 3047 + }, + { + "epoch": 0.2748793795373585, + "grad_norm": 2.1808900421353226, + "learning_rate": 3.403328238223808e-06, + "loss": 1.019, + "step": 3048 + }, + { + "epoch": 0.27496956306082876, + "grad_norm": 1.9037741545198703, + "learning_rate": 3.4029119236898395e-06, + "loss": 1.035, + "step": 3049 + }, + { + "epoch": 0.27505974658429905, + "grad_norm": 1.0631694928367053, + "learning_rate": 3.4024954894516906e-06, + "loss": 0.846, + "step": 3050 + }, + { + "epoch": 0.27514993010776934, + "grad_norm": 2.4864065632879564, + "learning_rate": 3.4020789355448933e-06, + "loss": 1.0107, + "step": 3051 + }, + { + "epoch": 0.27524011363123957, + "grad_norm": 1.5792801661035625, + "learning_rate": 3.40166226200499e-06, + "loss": 1.0788, + "step": 3052 + }, + { + "epoch": 0.27533029715470986, + "grad_norm": 2.051717257564154, + "learning_rate": 3.401245468867534e-06, + "loss": 0.9919, + "step": 3053 + }, + { + "epoch": 0.2754204806781801, + "grad_norm": 0.7097309059968817, + "learning_rate": 3.400828556168088e-06, + "loss": 0.8702, + "step": 3054 + }, + { + "epoch": 0.2755106642016504, + "grad_norm": 1.990544557967193, + "learning_rate": 3.4004115239422255e-06, + "loss": 1.0259, + "step": 3055 + }, + { + "epoch": 0.2756008477251206, + "grad_norm": 2.049457115303529, + "learning_rate": 3.3999943722255305e-06, + "loss": 0.9906, + "step": 3056 + }, + { + "epoch": 0.2756910312485909, + "grad_norm": 1.8510352563040409, + "learning_rate": 3.3995771010535955e-06, + "loss": 0.9055, + "step": 3057 + }, + { + "epoch": 0.27578121477206113, + "grad_norm": 1.1084638803126716, + "learning_rate": 3.3991597104620253e-06, + "loss": 0.7771, + "step": 3058 + }, + { + "epoch": 0.2758713982955314, + "grad_norm": 1.6553571201653081, + "learning_rate": 3.398742200486434e-06, + "loss": 0.9873, + "step": 3059 + }, + { + "epoch": 0.27596158181900166, + "grad_norm": 1.7437014309808498, + "learning_rate": 3.3983245711624453e-06, + "loss": 1.0272, + "step": 3060 + }, + { + "epoch": 0.27605176534247194, + "grad_norm": 0.7709001477814272, + "learning_rate": 3.3979068225256946e-06, + "loss": 0.8923, + "step": 3061 + }, + { + "epoch": 0.2761419488659422, + "grad_norm": 2.869509388263697, + "learning_rate": 3.3974889546118246e-06, + "loss": 0.9703, + "step": 3062 + }, + { + "epoch": 0.27623213238941247, + "grad_norm": 1.7364435318482099, + "learning_rate": 3.3970709674564918e-06, + "loss": 0.9519, + "step": 3063 + }, + { + "epoch": 0.2763223159128827, + "grad_norm": 1.7477558597013025, + "learning_rate": 3.3966528610953607e-06, + "loss": 1.0011, + "step": 3064 + }, + { + "epoch": 0.276412499436353, + "grad_norm": 1.910009959729176, + "learning_rate": 3.3962346355641067e-06, + "loss": 1.0121, + "step": 3065 + }, + { + "epoch": 0.2765026829598232, + "grad_norm": 1.7027231144735124, + "learning_rate": 3.3958162908984146e-06, + "loss": 1.0409, + "step": 3066 + }, + { + "epoch": 0.2765928664832935, + "grad_norm": 1.6686488603095773, + "learning_rate": 3.39539782713398e-06, + "loss": 1.003, + "step": 3067 + }, + { + "epoch": 0.27668305000676374, + "grad_norm": 3.4345576785291514, + "learning_rate": 3.394979244306509e-06, + "loss": 1.0338, + "step": 3068 + }, + { + "epoch": 0.27677323353023403, + "grad_norm": 1.592649566118408, + "learning_rate": 3.3945605424517166e-06, + "loss": 1.0166, + "step": 3069 + }, + { + "epoch": 0.27686341705370426, + "grad_norm": 1.6522876197894607, + "learning_rate": 3.3941417216053294e-06, + "loss": 1.0002, + "step": 3070 + }, + { + "epoch": 0.27695360057717455, + "grad_norm": 1.9330050718672158, + "learning_rate": 3.3937227818030835e-06, + "loss": 0.9643, + "step": 3071 + }, + { + "epoch": 0.2770437841006448, + "grad_norm": 2.7787132266927994, + "learning_rate": 3.393303723080725e-06, + "loss": 0.9668, + "step": 3072 + }, + { + "epoch": 0.2771339676241151, + "grad_norm": 2.121403736484733, + "learning_rate": 3.3928845454740097e-06, + "loss": 1.0179, + "step": 3073 + }, + { + "epoch": 0.27722415114758536, + "grad_norm": 1.9301276071887963, + "learning_rate": 3.392465249018705e-06, + "loss": 0.9918, + "step": 3074 + }, + { + "epoch": 0.2773143346710556, + "grad_norm": 1.8941148437960798, + "learning_rate": 3.3920458337505872e-06, + "loss": 1.0006, + "step": 3075 + }, + { + "epoch": 0.2774045181945259, + "grad_norm": 2.366903558704341, + "learning_rate": 3.391626299705443e-06, + "loss": 0.8393, + "step": 3076 + }, + { + "epoch": 0.2774947017179961, + "grad_norm": 2.0518486805348246, + "learning_rate": 3.39120664691907e-06, + "loss": 1.0153, + "step": 3077 + }, + { + "epoch": 0.2775848852414664, + "grad_norm": 1.9360733317090562, + "learning_rate": 3.390786875427275e-06, + "loss": 0.9813, + "step": 3078 + }, + { + "epoch": 0.27767506876493664, + "grad_norm": 1.6901886325727016, + "learning_rate": 3.390366985265875e-06, + "loss": 0.9991, + "step": 3079 + }, + { + "epoch": 0.2777652522884069, + "grad_norm": 1.7713208443762873, + "learning_rate": 3.389946976470697e-06, + "loss": 0.9618, + "step": 3080 + }, + { + "epoch": 0.27785543581187716, + "grad_norm": 1.5270095810855047, + "learning_rate": 3.3895268490775787e-06, + "loss": 1.0963, + "step": 3081 + }, + { + "epoch": 0.27794561933534745, + "grad_norm": 1.9377756230364689, + "learning_rate": 3.3891066031223685e-06, + "loss": 0.9623, + "step": 3082 + }, + { + "epoch": 0.2780358028588177, + "grad_norm": 1.5823000832543734, + "learning_rate": 3.3886862386409237e-06, + "loss": 1.035, + "step": 3083 + }, + { + "epoch": 0.27812598638228797, + "grad_norm": 1.9453739018688694, + "learning_rate": 3.388265755669111e-06, + "loss": 0.9638, + "step": 3084 + }, + { + "epoch": 0.2782161699057582, + "grad_norm": 1.7516772543130164, + "learning_rate": 3.3878451542428093e-06, + "loss": 1.0681, + "step": 3085 + }, + { + "epoch": 0.2783063534292285, + "grad_norm": 1.47203718778034, + "learning_rate": 3.387424434397907e-06, + "loss": 0.9891, + "step": 3086 + }, + { + "epoch": 0.2783965369526987, + "grad_norm": 2.169009177252286, + "learning_rate": 3.3870035961703013e-06, + "loss": 0.9144, + "step": 3087 + }, + { + "epoch": 0.278486720476169, + "grad_norm": 2.183729136847158, + "learning_rate": 3.3865826395959014e-06, + "loss": 1.1291, + "step": 3088 + }, + { + "epoch": 0.27857690399963925, + "grad_norm": 1.4690374539497548, + "learning_rate": 3.3861615647106253e-06, + "loss": 1.0897, + "step": 3089 + }, + { + "epoch": 0.27866708752310954, + "grad_norm": 2.0791563758773806, + "learning_rate": 3.3857403715504012e-06, + "loss": 1.0174, + "step": 3090 + }, + { + "epoch": 0.27875727104657977, + "grad_norm": 1.7679643914965055, + "learning_rate": 3.385319060151167e-06, + "loss": 0.9681, + "step": 3091 + }, + { + "epoch": 0.27884745457005006, + "grad_norm": 2.3322179262726372, + "learning_rate": 3.3848976305488728e-06, + "loss": 0.9941, + "step": 3092 + }, + { + "epoch": 0.2789376380935203, + "grad_norm": 3.5594956964487414, + "learning_rate": 3.384476082779476e-06, + "loss": 1.0462, + "step": 3093 + }, + { + "epoch": 0.2790278216169906, + "grad_norm": 1.9687435816100114, + "learning_rate": 3.3840544168789463e-06, + "loss": 0.9808, + "step": 3094 + }, + { + "epoch": 0.2791180051404608, + "grad_norm": 1.856432533922518, + "learning_rate": 3.3836326328832617e-06, + "loss": 1.1148, + "step": 3095 + }, + { + "epoch": 0.2792081886639311, + "grad_norm": 3.413491068140919, + "learning_rate": 3.383210730828412e-06, + "loss": 0.9597, + "step": 3096 + }, + { + "epoch": 0.2792983721874014, + "grad_norm": 1.5003444752923927, + "learning_rate": 3.3827887107503953e-06, + "loss": 1.0703, + "step": 3097 + }, + { + "epoch": 0.2793885557108716, + "grad_norm": 2.059180507245462, + "learning_rate": 3.3823665726852216e-06, + "loss": 0.9516, + "step": 3098 + }, + { + "epoch": 0.2794787392343419, + "grad_norm": 1.6385486360932402, + "learning_rate": 3.3819443166689095e-06, + "loss": 1.124, + "step": 3099 + }, + { + "epoch": 0.27956892275781214, + "grad_norm": 1.7750970410966433, + "learning_rate": 3.3815219427374886e-06, + "loss": 1.0008, + "step": 3100 + }, + { + "epoch": 0.27965910628128243, + "grad_norm": 1.6701260982807031, + "learning_rate": 3.3810994509269975e-06, + "loss": 1.0218, + "step": 3101 + }, + { + "epoch": 0.27974928980475267, + "grad_norm": 1.6701184608794555, + "learning_rate": 3.3806768412734864e-06, + "loss": 1.0361, + "step": 3102 + }, + { + "epoch": 0.27983947332822295, + "grad_norm": 2.1077981211721153, + "learning_rate": 3.380254113813014e-06, + "loss": 1.0871, + "step": 3103 + }, + { + "epoch": 0.2799296568516932, + "grad_norm": 1.554987049478162, + "learning_rate": 3.3798312685816496e-06, + "loss": 1.0369, + "step": 3104 + }, + { + "epoch": 0.2800198403751635, + "grad_norm": 2.1003751555900605, + "learning_rate": 3.3794083056154738e-06, + "loss": 1.0032, + "step": 3105 + }, + { + "epoch": 0.2801100238986337, + "grad_norm": 1.9104690763660894, + "learning_rate": 3.3789852249505746e-06, + "loss": 1.0146, + "step": 3106 + }, + { + "epoch": 0.280200207422104, + "grad_norm": 1.8016980003622354, + "learning_rate": 3.378562026623053e-06, + "loss": 1.0611, + "step": 3107 + }, + { + "epoch": 0.28029039094557423, + "grad_norm": 2.371020345792835, + "learning_rate": 3.3781387106690175e-06, + "loss": 1.0751, + "step": 3108 + }, + { + "epoch": 0.2803805744690445, + "grad_norm": 1.6105170040612684, + "learning_rate": 3.3777152771245885e-06, + "loss": 1.0358, + "step": 3109 + }, + { + "epoch": 0.28047075799251475, + "grad_norm": 1.98375548056102, + "learning_rate": 3.377291726025895e-06, + "loss": 1.04, + "step": 3110 + }, + { + "epoch": 0.28056094151598504, + "grad_norm": 2.4589562569753576, + "learning_rate": 3.3768680574090782e-06, + "loss": 1.1248, + "step": 3111 + }, + { + "epoch": 0.2806511250394553, + "grad_norm": 2.8134410555345797, + "learning_rate": 3.3764442713102857e-06, + "loss": 0.9849, + "step": 3112 + }, + { + "epoch": 0.28074130856292556, + "grad_norm": 2.2791798225687976, + "learning_rate": 3.3760203677656786e-06, + "loss": 1.0047, + "step": 3113 + }, + { + "epoch": 0.2808314920863958, + "grad_norm": 5.835731594635663, + "learning_rate": 3.3755963468114262e-06, + "loss": 1.0582, + "step": 3114 + }, + { + "epoch": 0.2809216756098661, + "grad_norm": 1.9874745589253433, + "learning_rate": 3.3751722084837095e-06, + "loss": 0.9132, + "step": 3115 + }, + { + "epoch": 0.2810118591333363, + "grad_norm": 1.9250468929573739, + "learning_rate": 3.3747479528187166e-06, + "loss": 0.986, + "step": 3116 + }, + { + "epoch": 0.2811020426568066, + "grad_norm": 1.7556208532331252, + "learning_rate": 3.3743235798526485e-06, + "loss": 1.059, + "step": 3117 + }, + { + "epoch": 0.28119222618027684, + "grad_norm": 1.4700698603683442, + "learning_rate": 3.373899089621714e-06, + "loss": 0.9407, + "step": 3118 + }, + { + "epoch": 0.2812824097037471, + "grad_norm": 2.2258632699697305, + "learning_rate": 3.373474482162134e-06, + "loss": 0.9843, + "step": 3119 + }, + { + "epoch": 0.28137259322721736, + "grad_norm": 2.3122770872159686, + "learning_rate": 3.3730497575101376e-06, + "loss": 1.009, + "step": 3120 + }, + { + "epoch": 0.28146277675068765, + "grad_norm": 1.884673848660124, + "learning_rate": 3.3726249157019654e-06, + "loss": 1.0026, + "step": 3121 + }, + { + "epoch": 0.28155296027415794, + "grad_norm": 15.395647776530366, + "learning_rate": 3.372199956773866e-06, + "loss": 1.092, + "step": 3122 + }, + { + "epoch": 0.28164314379762817, + "grad_norm": 2.9123923089415067, + "learning_rate": 3.371774880762101e-06, + "loss": 0.9603, + "step": 3123 + }, + { + "epoch": 0.28173332732109846, + "grad_norm": 2.033556171880461, + "learning_rate": 3.3713496877029392e-06, + "loss": 0.997, + "step": 3124 + }, + { + "epoch": 0.2818235108445687, + "grad_norm": 2.053598441663031, + "learning_rate": 3.37092437763266e-06, + "loss": 1.0708, + "step": 3125 + }, + { + "epoch": 0.281913694368039, + "grad_norm": 2.362852552639787, + "learning_rate": 3.3704989505875537e-06, + "loss": 1.0365, + "step": 3126 + }, + { + "epoch": 0.2820038778915092, + "grad_norm": 1.5090244146688856, + "learning_rate": 3.3700734066039205e-06, + "loss": 0.9787, + "step": 3127 + }, + { + "epoch": 0.2820940614149795, + "grad_norm": 1.9966914705997691, + "learning_rate": 3.36964774571807e-06, + "loss": 1.0366, + "step": 3128 + }, + { + "epoch": 0.28218424493844974, + "grad_norm": 1.753842223088368, + "learning_rate": 3.3692219679663206e-06, + "loss": 1.0007, + "step": 3129 + }, + { + "epoch": 0.28227442846192, + "grad_norm": 1.9914280060769085, + "learning_rate": 3.3687960733850043e-06, + "loss": 1.0202, + "step": 3130 + }, + { + "epoch": 0.28236461198539026, + "grad_norm": 1.733723973613132, + "learning_rate": 3.3683700620104586e-06, + "loss": 0.9402, + "step": 3131 + }, + { + "epoch": 0.28245479550886055, + "grad_norm": 1.8615318419022915, + "learning_rate": 3.3679439338790347e-06, + "loss": 1.0248, + "step": 3132 + }, + { + "epoch": 0.2825449790323308, + "grad_norm": 1.796562167592465, + "learning_rate": 3.3675176890270916e-06, + "loss": 1.1272, + "step": 3133 + }, + { + "epoch": 0.28263516255580107, + "grad_norm": 2.2207517593412893, + "learning_rate": 3.367091327490998e-06, + "loss": 1.0322, + "step": 3134 + }, + { + "epoch": 0.2827253460792713, + "grad_norm": 2.0337660245727136, + "learning_rate": 3.3666648493071347e-06, + "loss": 0.9884, + "step": 3135 + }, + { + "epoch": 0.2828155296027416, + "grad_norm": 1.7056397237453218, + "learning_rate": 3.3662382545118914e-06, + "loss": 0.9841, + "step": 3136 + }, + { + "epoch": 0.2829057131262118, + "grad_norm": 2.3198662017190075, + "learning_rate": 3.3658115431416663e-06, + "loss": 1.0903, + "step": 3137 + }, + { + "epoch": 0.2829958966496821, + "grad_norm": 0.6896935710976821, + "learning_rate": 3.36538471523287e-06, + "loss": 0.8459, + "step": 3138 + }, + { + "epoch": 0.28308608017315234, + "grad_norm": 1.5984553271778972, + "learning_rate": 3.3649577708219204e-06, + "loss": 1.0674, + "step": 3139 + }, + { + "epoch": 0.28317626369662263, + "grad_norm": 2.0135149890009254, + "learning_rate": 3.3645307099452477e-06, + "loss": 0.9476, + "step": 3140 + }, + { + "epoch": 0.28326644722009287, + "grad_norm": 4.312202553926342, + "learning_rate": 3.3641035326392907e-06, + "loss": 1.0045, + "step": 3141 + }, + { + "epoch": 0.28335663074356315, + "grad_norm": 1.8441206349406725, + "learning_rate": 3.363676238940499e-06, + "loss": 1.0125, + "step": 3142 + }, + { + "epoch": 0.2834468142670334, + "grad_norm": 1.9413800304952624, + "learning_rate": 3.363248828885331e-06, + "loss": 1.0414, + "step": 3143 + }, + { + "epoch": 0.2835369977905037, + "grad_norm": 1.6186552896372137, + "learning_rate": 3.3628213025102562e-06, + "loss": 1.0301, + "step": 3144 + }, + { + "epoch": 0.28362718131397396, + "grad_norm": 1.6615682902368856, + "learning_rate": 3.3623936598517536e-06, + "loss": 1.046, + "step": 3145 + }, + { + "epoch": 0.2837173648374442, + "grad_norm": 2.027514970491317, + "learning_rate": 3.3619659009463117e-06, + "loss": 0.8531, + "step": 3146 + }, + { + "epoch": 0.2838075483609145, + "grad_norm": 1.4821546311165246, + "learning_rate": 3.3615380258304287e-06, + "loss": 1.0602, + "step": 3147 + }, + { + "epoch": 0.2838977318843847, + "grad_norm": 1.8057071067915689, + "learning_rate": 3.3611100345406146e-06, + "loss": 0.9953, + "step": 3148 + }, + { + "epoch": 0.283987915407855, + "grad_norm": 1.9510915317738164, + "learning_rate": 3.3606819271133873e-06, + "loss": 0.9979, + "step": 3149 + }, + { + "epoch": 0.28407809893132524, + "grad_norm": 0.7354772290660262, + "learning_rate": 3.360253703585275e-06, + "loss": 0.8309, + "step": 3150 + }, + { + "epoch": 0.28416828245479553, + "grad_norm": 1.8448475302780836, + "learning_rate": 3.3598253639928164e-06, + "loss": 0.9476, + "step": 3151 + }, + { + "epoch": 0.28425846597826576, + "grad_norm": 1.7550206371735084, + "learning_rate": 3.3593969083725596e-06, + "loss": 0.9617, + "step": 3152 + }, + { + "epoch": 0.28434864950173605, + "grad_norm": 1.918981405214803, + "learning_rate": 3.358968336761063e-06, + "loss": 0.963, + "step": 3153 + }, + { + "epoch": 0.2844388330252063, + "grad_norm": 2.9488187202832212, + "learning_rate": 3.3585396491948945e-06, + "loss": 0.9711, + "step": 3154 + }, + { + "epoch": 0.28452901654867657, + "grad_norm": 1.858770512645337, + "learning_rate": 3.358110845710633e-06, + "loss": 1.0663, + "step": 3155 + }, + { + "epoch": 0.2846192000721468, + "grad_norm": 2.7999703950679256, + "learning_rate": 3.357681926344865e-06, + "loss": 0.9913, + "step": 3156 + }, + { + "epoch": 0.2847093835956171, + "grad_norm": 1.5291051829742632, + "learning_rate": 3.357252891134189e-06, + "loss": 1.0103, + "step": 3157 + }, + { + "epoch": 0.2847995671190873, + "grad_norm": 5.572021906911666, + "learning_rate": 3.356823740115212e-06, + "loss": 1.0521, + "step": 3158 + }, + { + "epoch": 0.2848897506425576, + "grad_norm": 1.4637929987923834, + "learning_rate": 3.3563944733245525e-06, + "loss": 0.9537, + "step": 3159 + }, + { + "epoch": 0.28497993416602785, + "grad_norm": 1.8979016713437384, + "learning_rate": 3.3559650907988375e-06, + "loss": 0.9994, + "step": 3160 + }, + { + "epoch": 0.28507011768949814, + "grad_norm": 2.3845498677096773, + "learning_rate": 3.3555355925747045e-06, + "loss": 0.8592, + "step": 3161 + }, + { + "epoch": 0.28516030121296837, + "grad_norm": 1.6570014688350634, + "learning_rate": 3.3551059786888e-06, + "loss": 0.9931, + "step": 3162 + }, + { + "epoch": 0.28525048473643866, + "grad_norm": 1.7337578715150526, + "learning_rate": 3.3546762491777807e-06, + "loss": 1.0593, + "step": 3163 + }, + { + "epoch": 0.2853406682599089, + "grad_norm": 1.8893541210357265, + "learning_rate": 3.3542464040783156e-06, + "loss": 0.9934, + "step": 3164 + }, + { + "epoch": 0.2854308517833792, + "grad_norm": 2.1039679279203685, + "learning_rate": 3.353816443427079e-06, + "loss": 1.0222, + "step": 3165 + }, + { + "epoch": 0.2855210353068494, + "grad_norm": 1.8979230898076005, + "learning_rate": 3.3533863672607597e-06, + "loss": 0.989, + "step": 3166 + }, + { + "epoch": 0.2856112188303197, + "grad_norm": 1.609558909750404, + "learning_rate": 3.352956175616052e-06, + "loss": 1.0367, + "step": 3167 + }, + { + "epoch": 0.28570140235378993, + "grad_norm": 1.7951098853136964, + "learning_rate": 3.352525868529664e-06, + "loss": 0.9742, + "step": 3168 + }, + { + "epoch": 0.2857915858772602, + "grad_norm": 1.7191124273864071, + "learning_rate": 3.3520954460383103e-06, + "loss": 0.9409, + "step": 3169 + }, + { + "epoch": 0.2858817694007305, + "grad_norm": 1.9001979749835198, + "learning_rate": 3.3516649081787182e-06, + "loss": 0.9466, + "step": 3170 + }, + { + "epoch": 0.28597195292420075, + "grad_norm": 1.5245067267690193, + "learning_rate": 3.3512342549876236e-06, + "loss": 1.1052, + "step": 3171 + }, + { + "epoch": 0.28606213644767103, + "grad_norm": 1.5684099201119244, + "learning_rate": 3.350803486501771e-06, + "loss": 0.9616, + "step": 3172 + }, + { + "epoch": 0.28615231997114127, + "grad_norm": 1.6627700156161564, + "learning_rate": 3.3503726027579175e-06, + "loss": 1.0306, + "step": 3173 + }, + { + "epoch": 0.28624250349461156, + "grad_norm": 1.4950502905153538, + "learning_rate": 3.349941603792827e-06, + "loss": 1.0627, + "step": 3174 + }, + { + "epoch": 0.2863326870180818, + "grad_norm": 1.9113224892501024, + "learning_rate": 3.3495104896432755e-06, + "loss": 1.0008, + "step": 3175 + }, + { + "epoch": 0.2864228705415521, + "grad_norm": 2.030171856488071, + "learning_rate": 3.3490792603460477e-06, + "loss": 0.9697, + "step": 3176 + }, + { + "epoch": 0.2865130540650223, + "grad_norm": 1.5899759373553484, + "learning_rate": 3.3486479159379393e-06, + "loss": 0.937, + "step": 3177 + }, + { + "epoch": 0.2866032375884926, + "grad_norm": 0.6798609588977027, + "learning_rate": 3.3482164564557537e-06, + "loss": 0.8133, + "step": 3178 + }, + { + "epoch": 0.28669342111196283, + "grad_norm": 3.1425634067302166, + "learning_rate": 3.3477848819363065e-06, + "loss": 1.0671, + "step": 3179 + }, + { + "epoch": 0.2867836046354331, + "grad_norm": 0.6718428182657127, + "learning_rate": 3.3473531924164213e-06, + "loss": 0.8429, + "step": 3180 + }, + { + "epoch": 0.28687378815890335, + "grad_norm": 2.2300834790671846, + "learning_rate": 3.3469213879329325e-06, + "loss": 0.9624, + "step": 3181 + }, + { + "epoch": 0.28696397168237364, + "grad_norm": 2.7564906769942024, + "learning_rate": 3.3464894685226837e-06, + "loss": 0.9651, + "step": 3182 + }, + { + "epoch": 0.2870541552058439, + "grad_norm": 3.4171419200813915, + "learning_rate": 3.34605743422253e-06, + "loss": 1.0862, + "step": 3183 + }, + { + "epoch": 0.28714433872931416, + "grad_norm": 1.9914485383660012, + "learning_rate": 3.345625285069333e-06, + "loss": 0.9763, + "step": 3184 + }, + { + "epoch": 0.2872345222527844, + "grad_norm": 2.4623551418085223, + "learning_rate": 3.345193021099967e-06, + "loss": 0.9872, + "step": 3185 + }, + { + "epoch": 0.2873247057762547, + "grad_norm": 1.8213658375448396, + "learning_rate": 3.3447606423513157e-06, + "loss": 0.9759, + "step": 3186 + }, + { + "epoch": 0.2874148892997249, + "grad_norm": 1.7791199915508615, + "learning_rate": 3.344328148860271e-06, + "loss": 0.9743, + "step": 3187 + }, + { + "epoch": 0.2875050728231952, + "grad_norm": 1.4844100144924544, + "learning_rate": 3.3438955406637365e-06, + "loss": 1.0193, + "step": 3188 + }, + { + "epoch": 0.28759525634666544, + "grad_norm": 2.2136352040592393, + "learning_rate": 3.343462817798624e-06, + "loss": 1.0295, + "step": 3189 + }, + { + "epoch": 0.28768543987013573, + "grad_norm": 4.590158524778069, + "learning_rate": 3.343029980301856e-06, + "loss": 0.9706, + "step": 3190 + }, + { + "epoch": 0.28777562339360596, + "grad_norm": 1.996388811075739, + "learning_rate": 3.342597028210365e-06, + "loss": 1.0318, + "step": 3191 + }, + { + "epoch": 0.28786580691707625, + "grad_norm": 1.7705672643898287, + "learning_rate": 3.342163961561092e-06, + "loss": 0.9508, + "step": 3192 + }, + { + "epoch": 0.28795599044054654, + "grad_norm": 1.7990677432695654, + "learning_rate": 3.34173078039099e-06, + "loss": 0.9737, + "step": 3193 + }, + { + "epoch": 0.28804617396401677, + "grad_norm": 2.224249565358573, + "learning_rate": 3.3412974847370193e-06, + "loss": 1.1708, + "step": 3194 + }, + { + "epoch": 0.28813635748748706, + "grad_norm": 2.1991863349978464, + "learning_rate": 3.3408640746361514e-06, + "loss": 1.0265, + "step": 3195 + }, + { + "epoch": 0.2882265410109573, + "grad_norm": 0.7649966016083075, + "learning_rate": 3.3404305501253663e-06, + "loss": 0.836, + "step": 3196 + }, + { + "epoch": 0.2883167245344276, + "grad_norm": 2.183198678132772, + "learning_rate": 3.3399969112416565e-06, + "loss": 1.0786, + "step": 3197 + }, + { + "epoch": 0.2884069080578978, + "grad_norm": 1.7582446436036323, + "learning_rate": 3.3395631580220213e-06, + "loss": 0.9433, + "step": 3198 + }, + { + "epoch": 0.2884970915813681, + "grad_norm": 2.6702642214335395, + "learning_rate": 3.3391292905034714e-06, + "loss": 0.9864, + "step": 3199 + }, + { + "epoch": 0.28858727510483834, + "grad_norm": 2.450655234553363, + "learning_rate": 3.338695308723027e-06, + "loss": 1.0666, + "step": 3200 + }, + { + "epoch": 0.2886774586283086, + "grad_norm": 1.857461473622487, + "learning_rate": 3.338261212717716e-06, + "loss": 1.0636, + "step": 3201 + }, + { + "epoch": 0.28876764215177886, + "grad_norm": 1.8705902377797767, + "learning_rate": 3.33782700252458e-06, + "loss": 1.037, + "step": 3202 + }, + { + "epoch": 0.28885782567524915, + "grad_norm": 2.5038354063203436, + "learning_rate": 3.337392678180668e-06, + "loss": 1.0475, + "step": 3203 + }, + { + "epoch": 0.2889480091987194, + "grad_norm": 2.2796633702779694, + "learning_rate": 3.3369582397230377e-06, + "loss": 1.0064, + "step": 3204 + }, + { + "epoch": 0.28903819272218967, + "grad_norm": 1.9751101499334454, + "learning_rate": 3.336523687188759e-06, + "loss": 1.0472, + "step": 3205 + }, + { + "epoch": 0.2891283762456599, + "grad_norm": 0.7138951579196764, + "learning_rate": 3.336089020614909e-06, + "loss": 0.8501, + "step": 3206 + }, + { + "epoch": 0.2892185597691302, + "grad_norm": 1.7072519120465273, + "learning_rate": 3.3356542400385774e-06, + "loss": 0.8821, + "step": 3207 + }, + { + "epoch": 0.2893087432926004, + "grad_norm": 1.5738175782593298, + "learning_rate": 3.3352193454968607e-06, + "loss": 0.9408, + "step": 3208 + }, + { + "epoch": 0.2893989268160707, + "grad_norm": 2.2861950722046025, + "learning_rate": 3.3347843370268675e-06, + "loss": 1.0718, + "step": 3209 + }, + { + "epoch": 0.28948911033954094, + "grad_norm": 1.5307599276056107, + "learning_rate": 3.334349214665715e-06, + "loss": 1.023, + "step": 3210 + }, + { + "epoch": 0.28957929386301123, + "grad_norm": 1.809963226893212, + "learning_rate": 3.3339139784505293e-06, + "loss": 1.0881, + "step": 3211 + }, + { + "epoch": 0.28966947738648147, + "grad_norm": 1.5042501477492978, + "learning_rate": 3.333478628418448e-06, + "loss": 1.075, + "step": 3212 + }, + { + "epoch": 0.28975966090995176, + "grad_norm": 1.4573188886447206, + "learning_rate": 3.333043164606618e-06, + "loss": 0.8972, + "step": 3213 + }, + { + "epoch": 0.289849844433422, + "grad_norm": 3.042455509960718, + "learning_rate": 3.3326075870521948e-06, + "loss": 1.0866, + "step": 3214 + }, + { + "epoch": 0.2899400279568923, + "grad_norm": 1.9519988208282537, + "learning_rate": 3.3321718957923437e-06, + "loss": 1.0081, + "step": 3215 + }, + { + "epoch": 0.29003021148036257, + "grad_norm": 4.094470011750486, + "learning_rate": 3.3317360908642413e-06, + "loss": 1.0032, + "step": 3216 + }, + { + "epoch": 0.2901203950038328, + "grad_norm": 1.9244010265903435, + "learning_rate": 3.331300172305072e-06, + "loss": 0.9597, + "step": 3217 + }, + { + "epoch": 0.2902105785273031, + "grad_norm": 1.5211266622328141, + "learning_rate": 3.330864140152032e-06, + "loss": 0.9976, + "step": 3218 + }, + { + "epoch": 0.2903007620507733, + "grad_norm": 1.7755798977552033, + "learning_rate": 3.330427994442325e-06, + "loss": 1.0232, + "step": 3219 + }, + { + "epoch": 0.2903909455742436, + "grad_norm": 2.157475993706913, + "learning_rate": 3.3299917352131657e-06, + "loss": 1.0437, + "step": 3220 + }, + { + "epoch": 0.29048112909771384, + "grad_norm": 2.609891646171281, + "learning_rate": 3.329555362501778e-06, + "loss": 0.9725, + "step": 3221 + }, + { + "epoch": 0.29057131262118413, + "grad_norm": 1.6944832597605213, + "learning_rate": 3.3291188763453954e-06, + "loss": 0.9748, + "step": 3222 + }, + { + "epoch": 0.29066149614465436, + "grad_norm": 5.53238765418042, + "learning_rate": 3.3286822767812618e-06, + "loss": 0.9951, + "step": 3223 + }, + { + "epoch": 0.29075167966812465, + "grad_norm": 1.6432513332735865, + "learning_rate": 3.32824556384663e-06, + "loss": 0.988, + "step": 3224 + }, + { + "epoch": 0.2908418631915949, + "grad_norm": 2.1601328512238602, + "learning_rate": 3.3278087375787628e-06, + "loss": 1.0078, + "step": 3225 + }, + { + "epoch": 0.2909320467150652, + "grad_norm": 2.759564154115412, + "learning_rate": 3.327371798014933e-06, + "loss": 1.0083, + "step": 3226 + }, + { + "epoch": 0.2910222302385354, + "grad_norm": 2.105360944886162, + "learning_rate": 3.3269347451924218e-06, + "loss": 1.0055, + "step": 3227 + }, + { + "epoch": 0.2911124137620057, + "grad_norm": 2.1795888352025368, + "learning_rate": 3.326497579148522e-06, + "loss": 1.0716, + "step": 3228 + }, + { + "epoch": 0.29120259728547593, + "grad_norm": 1.7885648210108347, + "learning_rate": 3.3260602999205345e-06, + "loss": 1.0046, + "step": 3229 + }, + { + "epoch": 0.2912927808089462, + "grad_norm": 4.348586648678448, + "learning_rate": 3.32562290754577e-06, + "loss": 1.0245, + "step": 3230 + }, + { + "epoch": 0.29138296433241645, + "grad_norm": 1.7232631730063581, + "learning_rate": 3.3251854020615494e-06, + "loss": 0.9938, + "step": 3231 + }, + { + "epoch": 0.29147314785588674, + "grad_norm": 1.6948530567807285, + "learning_rate": 3.324747783505204e-06, + "loss": 1.0102, + "step": 3232 + }, + { + "epoch": 0.29156333137935697, + "grad_norm": 2.020876880714012, + "learning_rate": 3.324310051914073e-06, + "loss": 0.9223, + "step": 3233 + }, + { + "epoch": 0.29165351490282726, + "grad_norm": 2.498504382032117, + "learning_rate": 3.3238722073255056e-06, + "loss": 0.9762, + "step": 3234 + }, + { + "epoch": 0.2917436984262975, + "grad_norm": 2.2799370527780205, + "learning_rate": 3.323434249776863e-06, + "loss": 0.9708, + "step": 3235 + }, + { + "epoch": 0.2918338819497678, + "grad_norm": 0.7169082929202059, + "learning_rate": 3.3229961793055117e-06, + "loss": 0.8268, + "step": 3236 + }, + { + "epoch": 0.291924065473238, + "grad_norm": 1.701525225550719, + "learning_rate": 3.3225579959488314e-06, + "loss": 0.9316, + "step": 3237 + }, + { + "epoch": 0.2920142489967083, + "grad_norm": 2.0761785354579194, + "learning_rate": 3.322119699744211e-06, + "loss": 1.0279, + "step": 3238 + }, + { + "epoch": 0.29210443252017854, + "grad_norm": 2.4236316831845723, + "learning_rate": 3.3216812907290476e-06, + "loss": 1.022, + "step": 3239 + }, + { + "epoch": 0.2921946160436488, + "grad_norm": 1.6386580529158565, + "learning_rate": 3.3212427689407484e-06, + "loss": 1.0258, + "step": 3240 + }, + { + "epoch": 0.2922847995671191, + "grad_norm": 0.6558953416787744, + "learning_rate": 3.3208041344167317e-06, + "loss": 0.8048, + "step": 3241 + }, + { + "epoch": 0.29237498309058935, + "grad_norm": 2.343428627869267, + "learning_rate": 3.3203653871944224e-06, + "loss": 0.9129, + "step": 3242 + }, + { + "epoch": 0.29246516661405964, + "grad_norm": 1.517610880965869, + "learning_rate": 3.3199265273112587e-06, + "loss": 0.9676, + "step": 3243 + }, + { + "epoch": 0.29255535013752987, + "grad_norm": 1.6624828595876866, + "learning_rate": 3.3194875548046852e-06, + "loss": 0.9464, + "step": 3244 + }, + { + "epoch": 0.29264553366100016, + "grad_norm": 1.6958977177025423, + "learning_rate": 3.319048469712158e-06, + "loss": 0.9843, + "step": 3245 + }, + { + "epoch": 0.2927357171844704, + "grad_norm": 2.666436711966387, + "learning_rate": 3.3186092720711423e-06, + "loss": 1.1019, + "step": 3246 + }, + { + "epoch": 0.2928259007079407, + "grad_norm": 7.3700389663308, + "learning_rate": 3.3181699619191125e-06, + "loss": 0.934, + "step": 3247 + }, + { + "epoch": 0.2929160842314109, + "grad_norm": 1.9551718400717608, + "learning_rate": 3.3177305392935536e-06, + "loss": 1.0479, + "step": 3248 + }, + { + "epoch": 0.2930062677548812, + "grad_norm": 0.8735345080284243, + "learning_rate": 3.3172910042319595e-06, + "loss": 0.8923, + "step": 3249 + }, + { + "epoch": 0.29309645127835143, + "grad_norm": 1.593543993033501, + "learning_rate": 3.316851356771833e-06, + "loss": 1.0377, + "step": 3250 + }, + { + "epoch": 0.2931866348018217, + "grad_norm": 1.862353079716699, + "learning_rate": 3.3164115969506876e-06, + "loss": 1.0354, + "step": 3251 + }, + { + "epoch": 0.29327681832529195, + "grad_norm": 2.10441061689363, + "learning_rate": 3.315971724806046e-06, + "loss": 1.0906, + "step": 3252 + }, + { + "epoch": 0.29336700184876224, + "grad_norm": 1.6888260752655817, + "learning_rate": 3.315531740375441e-06, + "loss": 0.951, + "step": 3253 + }, + { + "epoch": 0.2934571853722325, + "grad_norm": 2.1659998845869124, + "learning_rate": 3.315091643696414e-06, + "loss": 0.9738, + "step": 3254 + }, + { + "epoch": 0.29354736889570276, + "grad_norm": 1.5639040170227323, + "learning_rate": 3.3146514348065164e-06, + "loss": 0.9347, + "step": 3255 + }, + { + "epoch": 0.293637552419173, + "grad_norm": 2.5201723687697526, + "learning_rate": 3.31421111374331e-06, + "loss": 1.012, + "step": 3256 + }, + { + "epoch": 0.2937277359426433, + "grad_norm": 4.877513922002291, + "learning_rate": 3.3137706805443647e-06, + "loss": 1.0165, + "step": 3257 + }, + { + "epoch": 0.2938179194661135, + "grad_norm": 2.2154757558382423, + "learning_rate": 3.313330135247261e-06, + "loss": 1.0541, + "step": 3258 + }, + { + "epoch": 0.2939081029895838, + "grad_norm": 1.494008655579196, + "learning_rate": 3.312889477889588e-06, + "loss": 0.9802, + "step": 3259 + }, + { + "epoch": 0.29399828651305404, + "grad_norm": 1.5693983007453673, + "learning_rate": 3.3124487085089464e-06, + "loss": 1.0155, + "step": 3260 + }, + { + "epoch": 0.29408847003652433, + "grad_norm": 3.558500755594057, + "learning_rate": 3.312007827142943e-06, + "loss": 1.0402, + "step": 3261 + }, + { + "epoch": 0.29417865355999456, + "grad_norm": 2.545738481501729, + "learning_rate": 3.3115668338291983e-06, + "loss": 0.9239, + "step": 3262 + }, + { + "epoch": 0.29426883708346485, + "grad_norm": 1.6330850182130863, + "learning_rate": 3.3111257286053394e-06, + "loss": 1.0896, + "step": 3263 + }, + { + "epoch": 0.29435902060693514, + "grad_norm": 2.909434552919967, + "learning_rate": 3.3106845115090043e-06, + "loss": 1.0048, + "step": 3264 + }, + { + "epoch": 0.2944492041304054, + "grad_norm": 1.6927174338071673, + "learning_rate": 3.310243182577839e-06, + "loss": 0.9552, + "step": 3265 + }, + { + "epoch": 0.29453938765387566, + "grad_norm": 1.558960223793448, + "learning_rate": 3.3098017418495007e-06, + "loss": 0.8959, + "step": 3266 + }, + { + "epoch": 0.2946295711773459, + "grad_norm": 1.6959860030802267, + "learning_rate": 3.309360189361656e-06, + "loss": 0.9696, + "step": 3267 + }, + { + "epoch": 0.2947197547008162, + "grad_norm": 2.1560297314464054, + "learning_rate": 3.3089185251519797e-06, + "loss": 1.0843, + "step": 3268 + }, + { + "epoch": 0.2948099382242864, + "grad_norm": 2.2903720898240465, + "learning_rate": 3.3084767492581574e-06, + "loss": 0.9818, + "step": 3269 + }, + { + "epoch": 0.2949001217477567, + "grad_norm": 2.8680825980088973, + "learning_rate": 3.3080348617178846e-06, + "loss": 1.0211, + "step": 3270 + }, + { + "epoch": 0.29499030527122694, + "grad_norm": 1.644318555546789, + "learning_rate": 3.307592862568865e-06, + "loss": 0.9793, + "step": 3271 + }, + { + "epoch": 0.2950804887946972, + "grad_norm": 1.65289961030143, + "learning_rate": 3.307150751848812e-06, + "loss": 1.0707, + "step": 3272 + }, + { + "epoch": 0.29517067231816746, + "grad_norm": 1.5346887085229601, + "learning_rate": 3.3067085295954497e-06, + "loss": 0.9087, + "step": 3273 + }, + { + "epoch": 0.29526085584163775, + "grad_norm": 2.120332978387947, + "learning_rate": 3.3062661958465098e-06, + "loss": 0.867, + "step": 3274 + }, + { + "epoch": 0.295351039365108, + "grad_norm": 1.7747502110910083, + "learning_rate": 3.305823750639736e-06, + "loss": 0.8943, + "step": 3275 + }, + { + "epoch": 0.29544122288857827, + "grad_norm": 1.833343036221346, + "learning_rate": 3.3053811940128795e-06, + "loss": 0.9996, + "step": 3276 + }, + { + "epoch": 0.2955314064120485, + "grad_norm": 2.3261680827201983, + "learning_rate": 3.3049385260037016e-06, + "loss": 0.9815, + "step": 3277 + }, + { + "epoch": 0.2956215899355188, + "grad_norm": 1.55118776465534, + "learning_rate": 3.3044957466499736e-06, + "loss": 0.8963, + "step": 3278 + }, + { + "epoch": 0.295711773458989, + "grad_norm": 1.79739345866153, + "learning_rate": 3.304052855989475e-06, + "loss": 0.9227, + "step": 3279 + }, + { + "epoch": 0.2958019569824593, + "grad_norm": 2.4296400059415553, + "learning_rate": 3.3036098540599966e-06, + "loss": 1.0621, + "step": 3280 + }, + { + "epoch": 0.29589214050592955, + "grad_norm": 1.9496388565552314, + "learning_rate": 3.3031667408993373e-06, + "loss": 0.9641, + "step": 3281 + }, + { + "epoch": 0.29598232402939983, + "grad_norm": 1.7588539259841418, + "learning_rate": 3.302723516545306e-06, + "loss": 0.9972, + "step": 3282 + }, + { + "epoch": 0.29607250755287007, + "grad_norm": 2.2050084930048754, + "learning_rate": 3.302280181035722e-06, + "loss": 1.0288, + "step": 3283 + }, + { + "epoch": 0.29616269107634036, + "grad_norm": 0.6024621206461289, + "learning_rate": 3.3018367344084117e-06, + "loss": 0.8552, + "step": 3284 + }, + { + "epoch": 0.2962528745998106, + "grad_norm": 1.8604475502221731, + "learning_rate": 3.3013931767012125e-06, + "loss": 0.9926, + "step": 3285 + }, + { + "epoch": 0.2963430581232809, + "grad_norm": 1.7166804251381829, + "learning_rate": 3.300949507951972e-06, + "loss": 0.9977, + "step": 3286 + }, + { + "epoch": 0.2964332416467511, + "grad_norm": 2.653530119574583, + "learning_rate": 3.300505728198546e-06, + "loss": 1.0445, + "step": 3287 + }, + { + "epoch": 0.2965234251702214, + "grad_norm": 2.0271088862858675, + "learning_rate": 3.3000618374788e-06, + "loss": 0.9854, + "step": 3288 + }, + { + "epoch": 0.2966136086936917, + "grad_norm": 1.59052803628358, + "learning_rate": 3.2996178358306104e-06, + "loss": 1.0388, + "step": 3289 + }, + { + "epoch": 0.2967037922171619, + "grad_norm": 1.574425846988484, + "learning_rate": 3.2991737232918606e-06, + "loss": 0.9636, + "step": 3290 + }, + { + "epoch": 0.2967939757406322, + "grad_norm": 2.045233497813243, + "learning_rate": 3.298729499900445e-06, + "loss": 0.9541, + "step": 3291 + }, + { + "epoch": 0.29688415926410244, + "grad_norm": 3.233790690398666, + "learning_rate": 3.2982851656942677e-06, + "loss": 0.9471, + "step": 3292 + }, + { + "epoch": 0.29697434278757273, + "grad_norm": 3.4504437976610878, + "learning_rate": 3.2978407207112416e-06, + "loss": 1.0223, + "step": 3293 + }, + { + "epoch": 0.29706452631104296, + "grad_norm": 2.208483408980679, + "learning_rate": 3.2973961649892888e-06, + "loss": 1.0053, + "step": 3294 + }, + { + "epoch": 0.29715470983451325, + "grad_norm": 1.4616378492663074, + "learning_rate": 3.296951498566341e-06, + "loss": 1.0627, + "step": 3295 + }, + { + "epoch": 0.2972448933579835, + "grad_norm": 1.6537375196936055, + "learning_rate": 3.2965067214803404e-06, + "loss": 0.9961, + "step": 3296 + }, + { + "epoch": 0.2973350768814538, + "grad_norm": 1.2927417844040254, + "learning_rate": 3.2960618337692372e-06, + "loss": 1.0368, + "step": 3297 + }, + { + "epoch": 0.297425260404924, + "grad_norm": 1.6477075922397566, + "learning_rate": 3.2956168354709927e-06, + "loss": 0.9646, + "step": 3298 + }, + { + "epoch": 0.2975154439283943, + "grad_norm": 1.8772727543316352, + "learning_rate": 3.2951717266235754e-06, + "loss": 1.0493, + "step": 3299 + }, + { + "epoch": 0.29760562745186453, + "grad_norm": 1.9454692486342702, + "learning_rate": 3.294726507264964e-06, + "loss": 0.8771, + "step": 3300 + }, + { + "epoch": 0.2976958109753348, + "grad_norm": 2.0162468475865825, + "learning_rate": 3.2942811774331487e-06, + "loss": 0.9751, + "step": 3301 + }, + { + "epoch": 0.29778599449880505, + "grad_norm": 2.500074194760844, + "learning_rate": 3.293835737166127e-06, + "loss": 0.9942, + "step": 3302 + }, + { + "epoch": 0.29787617802227534, + "grad_norm": 1.8320591645869402, + "learning_rate": 3.293390186501906e-06, + "loss": 0.9468, + "step": 3303 + }, + { + "epoch": 0.2979663615457456, + "grad_norm": 1.7966981883754696, + "learning_rate": 3.2929445254785024e-06, + "loss": 1.0147, + "step": 3304 + }, + { + "epoch": 0.29805654506921586, + "grad_norm": 1.8650377739156718, + "learning_rate": 3.2924987541339423e-06, + "loss": 0.96, + "step": 3305 + }, + { + "epoch": 0.2981467285926861, + "grad_norm": 1.7666187021953925, + "learning_rate": 3.292052872506262e-06, + "loss": 1.0247, + "step": 3306 + }, + { + "epoch": 0.2982369121161564, + "grad_norm": 1.6594724436164756, + "learning_rate": 3.291606880633506e-06, + "loss": 1.0383, + "step": 3307 + }, + { + "epoch": 0.2983270956396266, + "grad_norm": 1.9366262988590481, + "learning_rate": 3.2911607785537297e-06, + "loss": 1.0548, + "step": 3308 + }, + { + "epoch": 0.2984172791630969, + "grad_norm": 4.058134109860501, + "learning_rate": 3.290714566304997e-06, + "loss": 1.0874, + "step": 3309 + }, + { + "epoch": 0.29850746268656714, + "grad_norm": 2.523472080052829, + "learning_rate": 3.2902682439253794e-06, + "loss": 1.0734, + "step": 3310 + }, + { + "epoch": 0.2985976462100374, + "grad_norm": 2.1628842227476586, + "learning_rate": 3.289821811452961e-06, + "loss": 1.0114, + "step": 3311 + }, + { + "epoch": 0.2986878297335077, + "grad_norm": 3.3951435129566727, + "learning_rate": 3.289375268925834e-06, + "loss": 1.0581, + "step": 3312 + }, + { + "epoch": 0.29877801325697795, + "grad_norm": 1.6176599374554255, + "learning_rate": 3.288928616382099e-06, + "loss": 0.9875, + "step": 3313 + }, + { + "epoch": 0.29886819678044824, + "grad_norm": 1.4228824777065672, + "learning_rate": 3.288481853859868e-06, + "loss": 1.0357, + "step": 3314 + }, + { + "epoch": 0.29895838030391847, + "grad_norm": 1.5946093467715017, + "learning_rate": 3.2880349813972604e-06, + "loss": 0.9779, + "step": 3315 + }, + { + "epoch": 0.29904856382738876, + "grad_norm": 1.761482848890336, + "learning_rate": 3.2875879990324052e-06, + "loss": 1.002, + "step": 3316 + }, + { + "epoch": 0.299138747350859, + "grad_norm": 1.9809078053410534, + "learning_rate": 3.287140906803443e-06, + "loss": 1.0494, + "step": 3317 + }, + { + "epoch": 0.2992289308743293, + "grad_norm": 2.0373764375068024, + "learning_rate": 3.2866937047485216e-06, + "loss": 0.9221, + "step": 3318 + }, + { + "epoch": 0.2993191143977995, + "grad_norm": 2.4736161380150095, + "learning_rate": 3.2862463929057985e-06, + "loss": 1.0581, + "step": 3319 + }, + { + "epoch": 0.2994092979212698, + "grad_norm": 1.6450012695168463, + "learning_rate": 3.285798971313441e-06, + "loss": 0.844, + "step": 3320 + }, + { + "epoch": 0.29949948144474003, + "grad_norm": 1.5844492576999207, + "learning_rate": 3.2853514400096248e-06, + "loss": 1.0226, + "step": 3321 + }, + { + "epoch": 0.2995896649682103, + "grad_norm": 1.7281714490059787, + "learning_rate": 3.2849037990325367e-06, + "loss": 1.0275, + "step": 3322 + }, + { + "epoch": 0.29967984849168056, + "grad_norm": 1.568363935528083, + "learning_rate": 3.2844560484203717e-06, + "loss": 0.9605, + "step": 3323 + }, + { + "epoch": 0.29977003201515084, + "grad_norm": 1.6650075601858612, + "learning_rate": 3.2840081882113333e-06, + "loss": 1.0841, + "step": 3324 + }, + { + "epoch": 0.2998602155386211, + "grad_norm": 1.8443236670852943, + "learning_rate": 3.283560218443638e-06, + "loss": 0.979, + "step": 3325 + }, + { + "epoch": 0.29995039906209137, + "grad_norm": 1.70832548682419, + "learning_rate": 3.2831121391555064e-06, + "loss": 1.018, + "step": 3326 + }, + { + "epoch": 0.3000405825855616, + "grad_norm": 1.4181446145030636, + "learning_rate": 3.2826639503851724e-06, + "loss": 0.956, + "step": 3327 + }, + { + "epoch": 0.3001307661090319, + "grad_norm": 1.528871050679459, + "learning_rate": 3.282215652170877e-06, + "loss": 1.0449, + "step": 3328 + }, + { + "epoch": 0.3002209496325021, + "grad_norm": 2.1053638892167945, + "learning_rate": 3.281767244550873e-06, + "loss": 1.039, + "step": 3329 + }, + { + "epoch": 0.3003111331559724, + "grad_norm": 0.6008697831459294, + "learning_rate": 3.2813187275634193e-06, + "loss": 0.7898, + "step": 3330 + }, + { + "epoch": 0.30040131667944264, + "grad_norm": 2.930405673693414, + "learning_rate": 3.280870101246787e-06, + "loss": 1.0277, + "step": 3331 + }, + { + "epoch": 0.30049150020291293, + "grad_norm": 2.4234669037344196, + "learning_rate": 3.280421365639255e-06, + "loss": 0.9191, + "step": 3332 + }, + { + "epoch": 0.30058168372638316, + "grad_norm": 1.6851659103350105, + "learning_rate": 3.279972520779112e-06, + "loss": 0.9886, + "step": 3333 + }, + { + "epoch": 0.30067186724985345, + "grad_norm": 1.727923172103794, + "learning_rate": 3.279523566704656e-06, + "loss": 1.0676, + "step": 3334 + }, + { + "epoch": 0.30076205077332374, + "grad_norm": 2.5476396033133026, + "learning_rate": 3.2790745034541935e-06, + "loss": 1.0548, + "step": 3335 + }, + { + "epoch": 0.300852234296794, + "grad_norm": 1.7977632400392238, + "learning_rate": 3.278625331066042e-06, + "loss": 1.0476, + "step": 3336 + }, + { + "epoch": 0.30094241782026426, + "grad_norm": 1.9688176945749536, + "learning_rate": 3.278176049578527e-06, + "loss": 1.0532, + "step": 3337 + }, + { + "epoch": 0.3010326013437345, + "grad_norm": 2.497913825314783, + "learning_rate": 3.2777266590299835e-06, + "loss": 0.9636, + "step": 3338 + }, + { + "epoch": 0.3011227848672048, + "grad_norm": 1.7344415325428442, + "learning_rate": 3.2772771594587562e-06, + "loss": 0.9389, + "step": 3339 + }, + { + "epoch": 0.301212968390675, + "grad_norm": 1.76295985722617, + "learning_rate": 3.2768275509031988e-06, + "loss": 1.1071, + "step": 3340 + }, + { + "epoch": 0.3013031519141453, + "grad_norm": 7.0020980415617275, + "learning_rate": 3.276377833401675e-06, + "loss": 0.9809, + "step": 3341 + }, + { + "epoch": 0.30139333543761554, + "grad_norm": 1.651525055633173, + "learning_rate": 3.2759280069925557e-06, + "loss": 1.064, + "step": 3342 + }, + { + "epoch": 0.30148351896108583, + "grad_norm": 2.110290554705859, + "learning_rate": 3.2754780717142233e-06, + "loss": 1.0432, + "step": 3343 + }, + { + "epoch": 0.30157370248455606, + "grad_norm": 1.717118754932949, + "learning_rate": 3.27502802760507e-06, + "loss": 0.9795, + "step": 3344 + }, + { + "epoch": 0.30166388600802635, + "grad_norm": 1.6695272057376729, + "learning_rate": 3.2745778747034943e-06, + "loss": 0.917, + "step": 3345 + }, + { + "epoch": 0.3017540695314966, + "grad_norm": 1.860467285401721, + "learning_rate": 3.274127613047906e-06, + "loss": 0.9798, + "step": 3346 + }, + { + "epoch": 0.30184425305496687, + "grad_norm": 1.6042377736899514, + "learning_rate": 3.273677242676725e-06, + "loss": 1.0203, + "step": 3347 + }, + { + "epoch": 0.3019344365784371, + "grad_norm": 1.9362033689140372, + "learning_rate": 3.2732267636283782e-06, + "loss": 1.0023, + "step": 3348 + }, + { + "epoch": 0.3020246201019074, + "grad_norm": 2.0285873567888406, + "learning_rate": 3.2727761759413034e-06, + "loss": 0.9633, + "step": 3349 + }, + { + "epoch": 0.3021148036253776, + "grad_norm": 1.8872876306769915, + "learning_rate": 3.2723254796539477e-06, + "loss": 1.0084, + "step": 3350 + }, + { + "epoch": 0.3022049871488479, + "grad_norm": 2.4058237069755153, + "learning_rate": 3.271874674804766e-06, + "loss": 0.9865, + "step": 3351 + }, + { + "epoch": 0.30229517067231815, + "grad_norm": 2.054083325576313, + "learning_rate": 3.2714237614322242e-06, + "loss": 1.0309, + "step": 3352 + }, + { + "epoch": 0.30238535419578844, + "grad_norm": 1.4625047243482137, + "learning_rate": 3.2709727395747974e-06, + "loss": 0.8347, + "step": 3353 + }, + { + "epoch": 0.30247553771925867, + "grad_norm": 0.677147801093327, + "learning_rate": 3.2705216092709673e-06, + "loss": 0.8094, + "step": 3354 + }, + { + "epoch": 0.30256572124272896, + "grad_norm": 2.083918819808605, + "learning_rate": 3.2700703705592282e-06, + "loss": 1.0951, + "step": 3355 + }, + { + "epoch": 0.3026559047661992, + "grad_norm": 1.7875799174550078, + "learning_rate": 3.269619023478082e-06, + "loss": 0.9119, + "step": 3356 + }, + { + "epoch": 0.3027460882896695, + "grad_norm": 1.7326516189396344, + "learning_rate": 3.26916756806604e-06, + "loss": 1.1058, + "step": 3357 + }, + { + "epoch": 0.3028362718131397, + "grad_norm": 1.611079674264247, + "learning_rate": 3.268716004361623e-06, + "loss": 1.0163, + "step": 3358 + }, + { + "epoch": 0.30292645533661, + "grad_norm": 0.6608856280016832, + "learning_rate": 3.268264332403361e-06, + "loss": 0.8251, + "step": 3359 + }, + { + "epoch": 0.3030166388600803, + "grad_norm": 1.517750537558674, + "learning_rate": 3.2678125522297933e-06, + "loss": 1.0373, + "step": 3360 + }, + { + "epoch": 0.3031068223835505, + "grad_norm": 2.563348769136078, + "learning_rate": 3.267360663879468e-06, + "loss": 0.9448, + "step": 3361 + }, + { + "epoch": 0.3031970059070208, + "grad_norm": 2.0701754542642004, + "learning_rate": 3.266908667390942e-06, + "loss": 0.9965, + "step": 3362 + }, + { + "epoch": 0.30328718943049104, + "grad_norm": 1.8309511038409443, + "learning_rate": 3.2664565628027833e-06, + "loss": 1.0599, + "step": 3363 + }, + { + "epoch": 0.30337737295396133, + "grad_norm": 4.171865856146346, + "learning_rate": 3.2660043501535675e-06, + "loss": 1.0054, + "step": 3364 + }, + { + "epoch": 0.30346755647743157, + "grad_norm": 1.684373290108855, + "learning_rate": 3.2655520294818797e-06, + "loss": 1.0732, + "step": 3365 + }, + { + "epoch": 0.30355774000090185, + "grad_norm": 2.152645138445149, + "learning_rate": 3.2650996008263146e-06, + "loss": 1.0788, + "step": 3366 + }, + { + "epoch": 0.3036479235243721, + "grad_norm": 1.8519367853750155, + "learning_rate": 3.2646470642254756e-06, + "loss": 0.9606, + "step": 3367 + }, + { + "epoch": 0.3037381070478424, + "grad_norm": 1.8273563562344068, + "learning_rate": 3.2641944197179767e-06, + "loss": 1.0256, + "step": 3368 + }, + { + "epoch": 0.3038282905713126, + "grad_norm": 0.6624708286196827, + "learning_rate": 3.2637416673424383e-06, + "loss": 0.8098, + "step": 3369 + }, + { + "epoch": 0.3039184740947829, + "grad_norm": 2.4886937061574343, + "learning_rate": 3.2632888071374937e-06, + "loss": 1.0181, + "step": 3370 + }, + { + "epoch": 0.30400865761825313, + "grad_norm": 1.9533608256066386, + "learning_rate": 3.2628358391417815e-06, + "loss": 1.0405, + "step": 3371 + }, + { + "epoch": 0.3040988411417234, + "grad_norm": 1.9760839915899924, + "learning_rate": 3.2623827633939526e-06, + "loss": 0.9967, + "step": 3372 + }, + { + "epoch": 0.30418902466519365, + "grad_norm": 1.5806808171008266, + "learning_rate": 3.2619295799326657e-06, + "loss": 0.9958, + "step": 3373 + }, + { + "epoch": 0.30427920818866394, + "grad_norm": 3.0009328663329518, + "learning_rate": 3.2614762887965883e-06, + "loss": 0.9034, + "step": 3374 + }, + { + "epoch": 0.3043693917121342, + "grad_norm": 1.8183855982937258, + "learning_rate": 3.2610228900243984e-06, + "loss": 0.9951, + "step": 3375 + }, + { + "epoch": 0.30445957523560446, + "grad_norm": 0.6062795563011486, + "learning_rate": 3.260569383654783e-06, + "loss": 0.8608, + "step": 3376 + }, + { + "epoch": 0.3045497587590747, + "grad_norm": 0.6876298825080753, + "learning_rate": 3.2601157697264365e-06, + "loss": 0.8308, + "step": 3377 + }, + { + "epoch": 0.304639942282545, + "grad_norm": 1.822076949218806, + "learning_rate": 3.2596620482780647e-06, + "loss": 0.9787, + "step": 3378 + }, + { + "epoch": 0.3047301258060152, + "grad_norm": 0.5719761816264259, + "learning_rate": 3.2592082193483803e-06, + "loss": 0.8305, + "step": 3379 + }, + { + "epoch": 0.3048203093294855, + "grad_norm": 1.9058933237050242, + "learning_rate": 3.258754282976109e-06, + "loss": 0.9921, + "step": 3380 + }, + { + "epoch": 0.30491049285295574, + "grad_norm": 1.859752328236637, + "learning_rate": 3.25830023919998e-06, + "loss": 1.0764, + "step": 3381 + }, + { + "epoch": 0.305000676376426, + "grad_norm": 2.1716890324074956, + "learning_rate": 3.2578460880587374e-06, + "loss": 0.959, + "step": 3382 + }, + { + "epoch": 0.3050908598998963, + "grad_norm": 2.1412656618916164, + "learning_rate": 3.2573918295911306e-06, + "loss": 1.0637, + "step": 3383 + }, + { + "epoch": 0.30518104342336655, + "grad_norm": 1.7609463709258313, + "learning_rate": 3.2569374638359196e-06, + "loss": 0.945, + "step": 3384 + }, + { + "epoch": 0.30527122694683684, + "grad_norm": 1.860391867694128, + "learning_rate": 3.2564829908318736e-06, + "loss": 1.0175, + "step": 3385 + }, + { + "epoch": 0.30536141047030707, + "grad_norm": 1.487446548399129, + "learning_rate": 3.2560284106177705e-06, + "loss": 1.0657, + "step": 3386 + }, + { + "epoch": 0.30545159399377736, + "grad_norm": 6.853210877290533, + "learning_rate": 3.2555737232323978e-06, + "loss": 0.9799, + "step": 3387 + }, + { + "epoch": 0.3055417775172476, + "grad_norm": 3.012088579757356, + "learning_rate": 3.255118928714552e-06, + "loss": 1.0366, + "step": 3388 + }, + { + "epoch": 0.3056319610407179, + "grad_norm": 1.7391905731449289, + "learning_rate": 3.2546640271030386e-06, + "loss": 0.9377, + "step": 3389 + }, + { + "epoch": 0.3057221445641881, + "grad_norm": 0.6567460410641701, + "learning_rate": 3.2542090184366717e-06, + "loss": 0.8622, + "step": 3390 + }, + { + "epoch": 0.3058123280876584, + "grad_norm": 1.9101959411616314, + "learning_rate": 3.253753902754276e-06, + "loss": 1.0831, + "step": 3391 + }, + { + "epoch": 0.30590251161112864, + "grad_norm": 1.499285686801616, + "learning_rate": 3.253298680094685e-06, + "loss": 0.9341, + "step": 3392 + }, + { + "epoch": 0.3059926951345989, + "grad_norm": 2.0260077100194733, + "learning_rate": 3.2528433504967394e-06, + "loss": 0.9488, + "step": 3393 + }, + { + "epoch": 0.30608287865806916, + "grad_norm": 1.6404453724209962, + "learning_rate": 3.252387913999291e-06, + "loss": 1.0164, + "step": 3394 + }, + { + "epoch": 0.30617306218153945, + "grad_norm": 1.8092411604482572, + "learning_rate": 3.2519323706411998e-06, + "loss": 1.0103, + "step": 3395 + }, + { + "epoch": 0.3062632457050097, + "grad_norm": 1.5066992252061513, + "learning_rate": 3.251476720461336e-06, + "loss": 0.8958, + "step": 3396 + }, + { + "epoch": 0.30635342922847997, + "grad_norm": 2.890862697416628, + "learning_rate": 3.251020963498578e-06, + "loss": 0.935, + "step": 3397 + }, + { + "epoch": 0.3064436127519502, + "grad_norm": 1.4643018203274911, + "learning_rate": 3.250565099791813e-06, + "loss": 1.0494, + "step": 3398 + }, + { + "epoch": 0.3065337962754205, + "grad_norm": 1.4919639539787561, + "learning_rate": 3.2501091293799387e-06, + "loss": 1.0597, + "step": 3399 + }, + { + "epoch": 0.3066239797988907, + "grad_norm": 1.8607897465212633, + "learning_rate": 3.24965305230186e-06, + "loss": 1.0168, + "step": 3400 + }, + { + "epoch": 0.306714163322361, + "grad_norm": 2.093947671697323, + "learning_rate": 3.249196868596492e-06, + "loss": 1.0784, + "step": 3401 + }, + { + "epoch": 0.30680434684583124, + "grad_norm": 1.930365810019423, + "learning_rate": 3.24874057830276e-06, + "loss": 0.9337, + "step": 3402 + }, + { + "epoch": 0.30689453036930153, + "grad_norm": 1.732108482390548, + "learning_rate": 3.2482841814595954e-06, + "loss": 0.9292, + "step": 3403 + }, + { + "epoch": 0.30698471389277177, + "grad_norm": 1.9764105093797983, + "learning_rate": 3.247827678105943e-06, + "loss": 0.9452, + "step": 3404 + }, + { + "epoch": 0.30707489741624205, + "grad_norm": 2.9773867609765574, + "learning_rate": 3.247371068280751e-06, + "loss": 1.0007, + "step": 3405 + }, + { + "epoch": 0.3071650809397123, + "grad_norm": 1.456342600878455, + "learning_rate": 3.2469143520229823e-06, + "loss": 0.8625, + "step": 3406 + }, + { + "epoch": 0.3072552644631826, + "grad_norm": 1.4371402124371364, + "learning_rate": 3.2464575293716054e-06, + "loss": 0.9568, + "step": 3407 + }, + { + "epoch": 0.30734544798665286, + "grad_norm": 2.5842244446893123, + "learning_rate": 3.2460006003655997e-06, + "loss": 0.9486, + "step": 3408 + }, + { + "epoch": 0.3074356315101231, + "grad_norm": 5.145068257196965, + "learning_rate": 3.245543565043952e-06, + "loss": 1.059, + "step": 3409 + }, + { + "epoch": 0.3075258150335934, + "grad_norm": 2.052438527718844, + "learning_rate": 3.2450864234456592e-06, + "loss": 0.9989, + "step": 3410 + }, + { + "epoch": 0.3076159985570636, + "grad_norm": 1.5556692732767408, + "learning_rate": 3.244629175609728e-06, + "loss": 0.9987, + "step": 3411 + }, + { + "epoch": 0.3077061820805339, + "grad_norm": 1.9598475614941162, + "learning_rate": 3.2441718215751726e-06, + "loss": 0.9889, + "step": 3412 + }, + { + "epoch": 0.30779636560400414, + "grad_norm": 1.822452057258868, + "learning_rate": 3.2437143613810173e-06, + "loss": 0.9638, + "step": 3413 + }, + { + "epoch": 0.30788654912747443, + "grad_norm": 1.5773572659492043, + "learning_rate": 3.2432567950662947e-06, + "loss": 1.0478, + "step": 3414 + }, + { + "epoch": 0.30797673265094466, + "grad_norm": 1.4409288065343437, + "learning_rate": 3.2427991226700468e-06, + "loss": 0.9912, + "step": 3415 + }, + { + "epoch": 0.30806691617441495, + "grad_norm": 2.44696905313432, + "learning_rate": 3.2423413442313246e-06, + "loss": 0.923, + "step": 3416 + }, + { + "epoch": 0.3081570996978852, + "grad_norm": 2.297248550452774, + "learning_rate": 3.2418834597891904e-06, + "loss": 0.9174, + "step": 3417 + }, + { + "epoch": 0.3082472832213555, + "grad_norm": 2.3130586310252284, + "learning_rate": 3.2414254693827098e-06, + "loss": 1.0904, + "step": 3418 + }, + { + "epoch": 0.3083374667448257, + "grad_norm": 1.6308263352753525, + "learning_rate": 3.2409673730509644e-06, + "loss": 0.9843, + "step": 3419 + }, + { + "epoch": 0.308427650268296, + "grad_norm": 3.5952544670610402, + "learning_rate": 3.2405091708330393e-06, + "loss": 0.982, + "step": 3420 + }, + { + "epoch": 0.3085178337917662, + "grad_norm": 1.8221321016152188, + "learning_rate": 3.2400508627680323e-06, + "loss": 0.9955, + "step": 3421 + }, + { + "epoch": 0.3086080173152365, + "grad_norm": 2.3290800657419912, + "learning_rate": 3.2395924488950474e-06, + "loss": 1.0934, + "step": 3422 + }, + { + "epoch": 0.30869820083870675, + "grad_norm": 5.0149206694911, + "learning_rate": 3.2391339292532004e-06, + "loss": 1.1286, + "step": 3423 + }, + { + "epoch": 0.30878838436217704, + "grad_norm": 4.447493569517928, + "learning_rate": 3.238675303881614e-06, + "loss": 0.9539, + "step": 3424 + }, + { + "epoch": 0.30887856788564727, + "grad_norm": 1.605174090964805, + "learning_rate": 3.2382165728194203e-06, + "loss": 0.9676, + "step": 3425 + }, + { + "epoch": 0.30896875140911756, + "grad_norm": 2.1050100870864825, + "learning_rate": 3.237757736105761e-06, + "loss": 0.9855, + "step": 3426 + }, + { + "epoch": 0.3090589349325878, + "grad_norm": 1.648977760623016, + "learning_rate": 3.2372987937797867e-06, + "loss": 1.0817, + "step": 3427 + }, + { + "epoch": 0.3091491184560581, + "grad_norm": 1.9113920929645716, + "learning_rate": 3.2368397458806573e-06, + "loss": 0.9089, + "step": 3428 + }, + { + "epoch": 0.3092393019795283, + "grad_norm": 2.3270578690936845, + "learning_rate": 3.2363805924475412e-06, + "loss": 1.0336, + "step": 3429 + }, + { + "epoch": 0.3093294855029986, + "grad_norm": 1.9196628151296529, + "learning_rate": 3.2359213335196153e-06, + "loss": 0.9423, + "step": 3430 + }, + { + "epoch": 0.3094196690264689, + "grad_norm": 1.761568456358531, + "learning_rate": 3.2354619691360663e-06, + "loss": 0.9527, + "step": 3431 + }, + { + "epoch": 0.3095098525499391, + "grad_norm": 1.9785002607104882, + "learning_rate": 3.2350024993360898e-06, + "loss": 1.0375, + "step": 3432 + }, + { + "epoch": 0.3096000360734094, + "grad_norm": 1.851684244394635, + "learning_rate": 3.2345429241588902e-06, + "loss": 1.001, + "step": 3433 + }, + { + "epoch": 0.30969021959687965, + "grad_norm": 1.8341524577280093, + "learning_rate": 3.234083243643681e-06, + "loss": 1.026, + "step": 3434 + }, + { + "epoch": 0.30978040312034993, + "grad_norm": 3.2997866243679184, + "learning_rate": 3.233623457829686e-06, + "loss": 1.0683, + "step": 3435 + }, + { + "epoch": 0.30987058664382017, + "grad_norm": 2.100876820391721, + "learning_rate": 3.2331635667561344e-06, + "loss": 0.969, + "step": 3436 + }, + { + "epoch": 0.30996077016729046, + "grad_norm": 1.770323249804458, + "learning_rate": 3.2327035704622674e-06, + "loss": 1.0341, + "step": 3437 + }, + { + "epoch": 0.3100509536907607, + "grad_norm": 1.5055807724248913, + "learning_rate": 3.2322434689873353e-06, + "loss": 0.898, + "step": 3438 + }, + { + "epoch": 0.310141137214231, + "grad_norm": 1.989246965771085, + "learning_rate": 3.2317832623705957e-06, + "loss": 1.0811, + "step": 3439 + }, + { + "epoch": 0.3102313207377012, + "grad_norm": 1.5284056423363643, + "learning_rate": 3.231322950651316e-06, + "loss": 0.9396, + "step": 3440 + }, + { + "epoch": 0.3103215042611715, + "grad_norm": 1.2556444521090773, + "learning_rate": 3.2308625338687735e-06, + "loss": 1.0278, + "step": 3441 + }, + { + "epoch": 0.31041168778464173, + "grad_norm": 3.180544576019778, + "learning_rate": 3.230402012062252e-06, + "loss": 1.1073, + "step": 3442 + }, + { + "epoch": 0.310501871308112, + "grad_norm": 0.6071611685155321, + "learning_rate": 3.2299413852710466e-06, + "loss": 0.8506, + "step": 3443 + }, + { + "epoch": 0.31059205483158225, + "grad_norm": 3.519707644632793, + "learning_rate": 3.2294806535344606e-06, + "loss": 0.9572, + "step": 3444 + }, + { + "epoch": 0.31068223835505254, + "grad_norm": 2.0793745830089447, + "learning_rate": 3.2290198168918056e-06, + "loss": 1.0414, + "step": 3445 + }, + { + "epoch": 0.3107724218785228, + "grad_norm": 1.8644641504255441, + "learning_rate": 3.2285588753824035e-06, + "loss": 1.0359, + "step": 3446 + }, + { + "epoch": 0.31086260540199306, + "grad_norm": 1.521904589856881, + "learning_rate": 3.228097829045584e-06, + "loss": 1.0119, + "step": 3447 + }, + { + "epoch": 0.3109527889254633, + "grad_norm": 1.3529302038360003, + "learning_rate": 3.227636677920685e-06, + "loss": 1.0445, + "step": 3448 + }, + { + "epoch": 0.3110429724489336, + "grad_norm": 1.6920229027778495, + "learning_rate": 3.2271754220470567e-06, + "loss": 1.0478, + "step": 3449 + }, + { + "epoch": 0.3111331559724038, + "grad_norm": 3.3583443413515375, + "learning_rate": 3.2267140614640547e-06, + "loss": 1.0787, + "step": 3450 + }, + { + "epoch": 0.3112233394958741, + "grad_norm": 1.7770551174677134, + "learning_rate": 3.2262525962110445e-06, + "loss": 1.0476, + "step": 3451 + }, + { + "epoch": 0.31131352301934434, + "grad_norm": 1.9281011694327326, + "learning_rate": 3.2257910263274015e-06, + "loss": 0.9271, + "step": 3452 + }, + { + "epoch": 0.31140370654281463, + "grad_norm": 1.6224277385025805, + "learning_rate": 3.225329351852509e-06, + "loss": 0.9463, + "step": 3453 + }, + { + "epoch": 0.3114938900662849, + "grad_norm": 1.998180276811914, + "learning_rate": 3.2248675728257596e-06, + "loss": 1.0038, + "step": 3454 + }, + { + "epoch": 0.31158407358975515, + "grad_norm": 1.7145448046125613, + "learning_rate": 3.2244056892865557e-06, + "loss": 1.0729, + "step": 3455 + }, + { + "epoch": 0.31167425711322544, + "grad_norm": 1.797257092527562, + "learning_rate": 3.2239437012743063e-06, + "loss": 0.9928, + "step": 3456 + }, + { + "epoch": 0.31176444063669567, + "grad_norm": 2.7030443829989923, + "learning_rate": 3.223481608828432e-06, + "loss": 1.0172, + "step": 3457 + }, + { + "epoch": 0.31185462416016596, + "grad_norm": 2.5822414941646707, + "learning_rate": 3.223019411988361e-06, + "loss": 1.0205, + "step": 3458 + }, + { + "epoch": 0.3119448076836362, + "grad_norm": 5.125104391383946, + "learning_rate": 3.22255711079353e-06, + "loss": 1.0031, + "step": 3459 + }, + { + "epoch": 0.3120349912071065, + "grad_norm": 3.0233513871578825, + "learning_rate": 3.222094705283385e-06, + "loss": 0.9661, + "step": 3460 + }, + { + "epoch": 0.3121251747305767, + "grad_norm": 1.9110264580420628, + "learning_rate": 3.2216321954973805e-06, + "loss": 0.9963, + "step": 3461 + }, + { + "epoch": 0.312215358254047, + "grad_norm": 1.6709975319689474, + "learning_rate": 3.2211695814749816e-06, + "loss": 0.99, + "step": 3462 + }, + { + "epoch": 0.31230554177751724, + "grad_norm": 3.1791177375064645, + "learning_rate": 3.220706863255661e-06, + "loss": 1.1056, + "step": 3463 + }, + { + "epoch": 0.3123957253009875, + "grad_norm": 2.3325788663265077, + "learning_rate": 3.2202440408788994e-06, + "loss": 1.0009, + "step": 3464 + }, + { + "epoch": 0.31248590882445776, + "grad_norm": 2.041082324530658, + "learning_rate": 3.2197811143841883e-06, + "loss": 1.073, + "step": 3465 + }, + { + "epoch": 0.31257609234792805, + "grad_norm": 1.5952270245794893, + "learning_rate": 3.2193180838110267e-06, + "loss": 0.9968, + "step": 3466 + }, + { + "epoch": 0.3126662758713983, + "grad_norm": 3.1813759409739077, + "learning_rate": 3.2188549491989225e-06, + "loss": 0.8983, + "step": 3467 + }, + { + "epoch": 0.31275645939486857, + "grad_norm": 1.6646178845514676, + "learning_rate": 3.2183917105873934e-06, + "loss": 1.0278, + "step": 3468 + }, + { + "epoch": 0.3128466429183388, + "grad_norm": 1.456532083458529, + "learning_rate": 3.217928368015966e-06, + "loss": 1.0648, + "step": 3469 + }, + { + "epoch": 0.3129368264418091, + "grad_norm": 1.592091763569006, + "learning_rate": 3.217464921524174e-06, + "loss": 1.0302, + "step": 3470 + }, + { + "epoch": 0.3130270099652793, + "grad_norm": 1.6372272002140844, + "learning_rate": 3.2170013711515616e-06, + "loss": 1.015, + "step": 3471 + }, + { + "epoch": 0.3131171934887496, + "grad_norm": 3.4633807993461097, + "learning_rate": 3.216537716937682e-06, + "loss": 0.9263, + "step": 3472 + }, + { + "epoch": 0.31320737701221985, + "grad_norm": 2.0327472743742225, + "learning_rate": 3.2160739589220968e-06, + "loss": 0.9765, + "step": 3473 + }, + { + "epoch": 0.31329756053569013, + "grad_norm": 2.7769510713194863, + "learning_rate": 3.215610097144376e-06, + "loss": 1.0586, + "step": 3474 + }, + { + "epoch": 0.31338774405916037, + "grad_norm": 1.798593512031275, + "learning_rate": 3.215146131644099e-06, + "loss": 0.9493, + "step": 3475 + }, + { + "epoch": 0.31347792758263066, + "grad_norm": 1.6358199462627867, + "learning_rate": 3.214682062460854e-06, + "loss": 1.0758, + "step": 3476 + }, + { + "epoch": 0.3135681111061009, + "grad_norm": 6.302869645763299, + "learning_rate": 3.2142178896342367e-06, + "loss": 0.9203, + "step": 3477 + }, + { + "epoch": 0.3136582946295712, + "grad_norm": 1.6644793780749045, + "learning_rate": 3.2137536132038552e-06, + "loss": 0.9946, + "step": 3478 + }, + { + "epoch": 0.31374847815304147, + "grad_norm": 1.5413916918245447, + "learning_rate": 3.2132892332093226e-06, + "loss": 1.0053, + "step": 3479 + }, + { + "epoch": 0.3138386616765117, + "grad_norm": 1.7394551288928146, + "learning_rate": 3.2128247496902623e-06, + "loss": 1.114, + "step": 3480 + }, + { + "epoch": 0.313928845199982, + "grad_norm": 1.3733838727287204, + "learning_rate": 3.2123601626863064e-06, + "loss": 1.0162, + "step": 3481 + }, + { + "epoch": 0.3140190287234522, + "grad_norm": 0.6623517383364586, + "learning_rate": 3.2118954722370974e-06, + "loss": 0.8083, + "step": 3482 + }, + { + "epoch": 0.3141092122469225, + "grad_norm": 1.5882979993035546, + "learning_rate": 3.2114306783822837e-06, + "loss": 0.9485, + "step": 3483 + }, + { + "epoch": 0.31419939577039274, + "grad_norm": 1.7808418475554304, + "learning_rate": 3.210965781161525e-06, + "loss": 0.9817, + "step": 3484 + }, + { + "epoch": 0.31428957929386303, + "grad_norm": 1.646527526217611, + "learning_rate": 3.2105007806144892e-06, + "loss": 1.1172, + "step": 3485 + }, + { + "epoch": 0.31437976281733326, + "grad_norm": 2.5424423493307393, + "learning_rate": 3.2100356767808513e-06, + "loss": 1.0052, + "step": 3486 + }, + { + "epoch": 0.31446994634080355, + "grad_norm": 2.384919182709687, + "learning_rate": 3.2095704697002977e-06, + "loss": 1.0627, + "step": 3487 + }, + { + "epoch": 0.3145601298642738, + "grad_norm": 1.6578264471360844, + "learning_rate": 3.209105159412522e-06, + "loss": 1.0038, + "step": 3488 + }, + { + "epoch": 0.3146503133877441, + "grad_norm": 2.0070587999733074, + "learning_rate": 3.208639745957228e-06, + "loss": 0.9643, + "step": 3489 + }, + { + "epoch": 0.3147404969112143, + "grad_norm": 1.4785801257396982, + "learning_rate": 3.2081742293741256e-06, + "loss": 1.0325, + "step": 3490 + }, + { + "epoch": 0.3148306804346846, + "grad_norm": 1.3435017999708685, + "learning_rate": 3.2077086097029366e-06, + "loss": 1.0653, + "step": 3491 + }, + { + "epoch": 0.31492086395815483, + "grad_norm": 1.9843015882474464, + "learning_rate": 3.2072428869833895e-06, + "loss": 1.0034, + "step": 3492 + }, + { + "epoch": 0.3150110474816251, + "grad_norm": 1.6182531264387032, + "learning_rate": 3.206777061255223e-06, + "loss": 1.0157, + "step": 3493 + }, + { + "epoch": 0.31510123100509535, + "grad_norm": 2.025068531525056, + "learning_rate": 3.206311132558183e-06, + "loss": 1.043, + "step": 3494 + }, + { + "epoch": 0.31519141452856564, + "grad_norm": 2.0358515337040557, + "learning_rate": 3.205845100932026e-06, + "loss": 1.0185, + "step": 3495 + }, + { + "epoch": 0.31528159805203587, + "grad_norm": 1.753694925129977, + "learning_rate": 3.205378966416516e-06, + "loss": 1.0338, + "step": 3496 + }, + { + "epoch": 0.31537178157550616, + "grad_norm": 1.7609353364188198, + "learning_rate": 3.204912729051426e-06, + "loss": 1.0024, + "step": 3497 + }, + { + "epoch": 0.3154619650989764, + "grad_norm": 2.0933584729687595, + "learning_rate": 3.2044463888765384e-06, + "loss": 0.9815, + "step": 3498 + }, + { + "epoch": 0.3155521486224467, + "grad_norm": 2.1407239013991237, + "learning_rate": 3.2039799459316436e-06, + "loss": 0.9029, + "step": 3499 + }, + { + "epoch": 0.3156423321459169, + "grad_norm": 1.8482557296055255, + "learning_rate": 3.2035134002565407e-06, + "loss": 1.0399, + "step": 3500 + }, + { + "epoch": 0.3157325156693872, + "grad_norm": 2.8233224968850505, + "learning_rate": 3.203046751891039e-06, + "loss": 1.0155, + "step": 3501 + }, + { + "epoch": 0.3158226991928575, + "grad_norm": 2.5111260790177186, + "learning_rate": 3.2025800008749545e-06, + "loss": 1.0176, + "step": 3502 + }, + { + "epoch": 0.3159128827163277, + "grad_norm": 3.057857965707358, + "learning_rate": 3.202113147248114e-06, + "loss": 0.9762, + "step": 3503 + }, + { + "epoch": 0.316003066239798, + "grad_norm": 3.5939285897371254, + "learning_rate": 3.20164619105035e-06, + "loss": 0.9761, + "step": 3504 + }, + { + "epoch": 0.31609324976326825, + "grad_norm": 1.7390778163019516, + "learning_rate": 3.201179132321508e-06, + "loss": 0.8606, + "step": 3505 + }, + { + "epoch": 0.31618343328673854, + "grad_norm": 1.8002221818069093, + "learning_rate": 3.200711971101439e-06, + "loss": 0.9422, + "step": 3506 + }, + { + "epoch": 0.31627361681020877, + "grad_norm": 2.22776028723231, + "learning_rate": 3.2002447074300047e-06, + "loss": 1.0055, + "step": 3507 + }, + { + "epoch": 0.31636380033367906, + "grad_norm": 1.7639769568030303, + "learning_rate": 3.1997773413470736e-06, + "loss": 1.0099, + "step": 3508 + }, + { + "epoch": 0.3164539838571493, + "grad_norm": 2.2557773591692567, + "learning_rate": 3.199309872892524e-06, + "loss": 1.0006, + "step": 3509 + }, + { + "epoch": 0.3165441673806196, + "grad_norm": 1.7457781000739687, + "learning_rate": 3.198842302106243e-06, + "loss": 1.0554, + "step": 3510 + }, + { + "epoch": 0.3166343509040898, + "grad_norm": 2.9100264757138325, + "learning_rate": 3.1983746290281265e-06, + "loss": 0.9923, + "step": 3511 + }, + { + "epoch": 0.3167245344275601, + "grad_norm": 1.5057205155963196, + "learning_rate": 3.197906853698079e-06, + "loss": 1.012, + "step": 3512 + }, + { + "epoch": 0.31681471795103033, + "grad_norm": 1.5883055798139614, + "learning_rate": 3.1974389761560137e-06, + "loss": 0.9541, + "step": 3513 + }, + { + "epoch": 0.3169049014745006, + "grad_norm": 1.7228882648505648, + "learning_rate": 3.1969709964418525e-06, + "loss": 0.9543, + "step": 3514 + }, + { + "epoch": 0.31699508499797086, + "grad_norm": 2.3011435982775605, + "learning_rate": 3.196502914595525e-06, + "loss": 1.0219, + "step": 3515 + }, + { + "epoch": 0.31708526852144114, + "grad_norm": 1.5855701190140348, + "learning_rate": 3.1960347306569723e-06, + "loss": 1.0249, + "step": 3516 + }, + { + "epoch": 0.3171754520449114, + "grad_norm": 2.2696488000363826, + "learning_rate": 3.195566444666141e-06, + "loss": 0.9535, + "step": 3517 + }, + { + "epoch": 0.31726563556838167, + "grad_norm": 1.6993850801988972, + "learning_rate": 3.1950980566629886e-06, + "loss": 1.111, + "step": 3518 + }, + { + "epoch": 0.3173558190918519, + "grad_norm": 1.4728417722373857, + "learning_rate": 3.1946295666874797e-06, + "loss": 1.0203, + "step": 3519 + }, + { + "epoch": 0.3174460026153222, + "grad_norm": 1.8868034770256645, + "learning_rate": 3.19416097477959e-06, + "loss": 1.0259, + "step": 3520 + }, + { + "epoch": 0.3175361861387924, + "grad_norm": 2.087070457442641, + "learning_rate": 3.1936922809793005e-06, + "loss": 1.0624, + "step": 3521 + }, + { + "epoch": 0.3176263696622627, + "grad_norm": 1.535768653693002, + "learning_rate": 3.193223485326604e-06, + "loss": 1.1272, + "step": 3522 + }, + { + "epoch": 0.31771655318573294, + "grad_norm": 2.0177371764661167, + "learning_rate": 3.1927545878615005e-06, + "loss": 1.0932, + "step": 3523 + }, + { + "epoch": 0.31780673670920323, + "grad_norm": 2.275884406085797, + "learning_rate": 3.192285588623999e-06, + "loss": 0.9575, + "step": 3524 + }, + { + "epoch": 0.31789692023267346, + "grad_norm": 1.37902343079189, + "learning_rate": 3.191816487654117e-06, + "loss": 0.9384, + "step": 3525 + }, + { + "epoch": 0.31798710375614375, + "grad_norm": 3.005796871811987, + "learning_rate": 3.19134728499188e-06, + "loss": 0.9691, + "step": 3526 + }, + { + "epoch": 0.31807728727961404, + "grad_norm": 2.439538445671922, + "learning_rate": 3.1908779806773235e-06, + "loss": 0.8958, + "step": 3527 + }, + { + "epoch": 0.3181674708030843, + "grad_norm": 1.5769487284064454, + "learning_rate": 3.190408574750492e-06, + "loss": 1.0641, + "step": 3528 + }, + { + "epoch": 0.31825765432655456, + "grad_norm": 3.227029761962961, + "learning_rate": 3.1899390672514367e-06, + "loss": 1.076, + "step": 3529 + }, + { + "epoch": 0.3183478378500248, + "grad_norm": 1.996924300323967, + "learning_rate": 3.189469458220219e-06, + "loss": 0.9276, + "step": 3530 + }, + { + "epoch": 0.3184380213734951, + "grad_norm": 1.4471606835582789, + "learning_rate": 3.1889997476969086e-06, + "loss": 1.0048, + "step": 3531 + }, + { + "epoch": 0.3185282048969653, + "grad_norm": 1.7579271745363945, + "learning_rate": 3.188529935721583e-06, + "loss": 0.9974, + "step": 3532 + }, + { + "epoch": 0.3186183884204356, + "grad_norm": 1.7954588907582123, + "learning_rate": 3.18806002233433e-06, + "loss": 1.0686, + "step": 3533 + }, + { + "epoch": 0.31870857194390584, + "grad_norm": 1.8447657874295535, + "learning_rate": 3.187590007575245e-06, + "loss": 1.0706, + "step": 3534 + }, + { + "epoch": 0.3187987554673761, + "grad_norm": 1.6895980508197033, + "learning_rate": 3.1871198914844327e-06, + "loss": 0.9449, + "step": 3535 + }, + { + "epoch": 0.31888893899084636, + "grad_norm": 1.7631159145874342, + "learning_rate": 3.1866496741020057e-06, + "loss": 0.9418, + "step": 3536 + }, + { + "epoch": 0.31897912251431665, + "grad_norm": 1.392762202740609, + "learning_rate": 3.186179355468085e-06, + "loss": 0.9728, + "step": 3537 + }, + { + "epoch": 0.3190693060377869, + "grad_norm": 2.2127964575479195, + "learning_rate": 3.1857089356228015e-06, + "loss": 0.9997, + "step": 3538 + }, + { + "epoch": 0.31915948956125717, + "grad_norm": 1.6142454060321796, + "learning_rate": 3.1852384146062933e-06, + "loss": 0.9537, + "step": 3539 + }, + { + "epoch": 0.3192496730847274, + "grad_norm": 0.5881403649215222, + "learning_rate": 3.184767792458708e-06, + "loss": 0.7952, + "step": 3540 + }, + { + "epoch": 0.3193398566081977, + "grad_norm": 2.1269586175965007, + "learning_rate": 3.1842970692202023e-06, + "loss": 1.1333, + "step": 3541 + }, + { + "epoch": 0.3194300401316679, + "grad_norm": 2.1094863297082136, + "learning_rate": 3.1838262449309403e-06, + "loss": 0.9874, + "step": 3542 + }, + { + "epoch": 0.3195202236551382, + "grad_norm": 1.9332025305576916, + "learning_rate": 3.1833553196310956e-06, + "loss": 0.9261, + "step": 3543 + }, + { + "epoch": 0.31961040717860845, + "grad_norm": 2.8444793101332917, + "learning_rate": 3.18288429336085e-06, + "loss": 0.8647, + "step": 3544 + }, + { + "epoch": 0.31970059070207874, + "grad_norm": 2.063796473995893, + "learning_rate": 3.182413166160394e-06, + "loss": 1.1139, + "step": 3545 + }, + { + "epoch": 0.31979077422554897, + "grad_norm": 1.956875305663103, + "learning_rate": 3.1819419380699275e-06, + "loss": 1.0469, + "step": 3546 + }, + { + "epoch": 0.31988095774901926, + "grad_norm": 1.4614423399757277, + "learning_rate": 3.181470609129658e-06, + "loss": 1.0131, + "step": 3547 + }, + { + "epoch": 0.3199711412724895, + "grad_norm": 2.040001663880511, + "learning_rate": 3.1809991793798e-06, + "loss": 1.0555, + "step": 3548 + }, + { + "epoch": 0.3200613247959598, + "grad_norm": 1.7739466133857986, + "learning_rate": 3.1805276488605806e-06, + "loss": 0.9622, + "step": 3549 + }, + { + "epoch": 0.32015150831943007, + "grad_norm": 1.9782421830648484, + "learning_rate": 3.1800560176122336e-06, + "loss": 0.9992, + "step": 3550 + }, + { + "epoch": 0.3202416918429003, + "grad_norm": 1.7195393916989783, + "learning_rate": 3.179584285675e-06, + "loss": 1.0113, + "step": 3551 + }, + { + "epoch": 0.3203318753663706, + "grad_norm": 5.268615235837203, + "learning_rate": 3.1791124530891315e-06, + "loss": 0.9193, + "step": 3552 + }, + { + "epoch": 0.3204220588898408, + "grad_norm": 9.314438995651372, + "learning_rate": 3.178640519894886e-06, + "loss": 1.0071, + "step": 3553 + }, + { + "epoch": 0.3205122424133111, + "grad_norm": 1.7655160245770443, + "learning_rate": 3.1781684861325324e-06, + "loss": 0.9659, + "step": 3554 + }, + { + "epoch": 0.32060242593678134, + "grad_norm": 1.8622902848117355, + "learning_rate": 3.177696351842348e-06, + "loss": 1.0144, + "step": 3555 + }, + { + "epoch": 0.32069260946025163, + "grad_norm": 0.6955702229034314, + "learning_rate": 3.1772241170646167e-06, + "loss": 0.7975, + "step": 3556 + }, + { + "epoch": 0.32078279298372187, + "grad_norm": 1.9193179479757263, + "learning_rate": 3.1767517818396334e-06, + "loss": 1.1103, + "step": 3557 + }, + { + "epoch": 0.32087297650719215, + "grad_norm": 1.6862803396064623, + "learning_rate": 3.1762793462076986e-06, + "loss": 1.1617, + "step": 3558 + }, + { + "epoch": 0.3209631600306624, + "grad_norm": 1.9583959637089192, + "learning_rate": 3.1758068102091236e-06, + "loss": 0.8903, + "step": 3559 + }, + { + "epoch": 0.3210533435541327, + "grad_norm": 1.7191355186201074, + "learning_rate": 3.175334173884229e-06, + "loss": 0.9719, + "step": 3560 + }, + { + "epoch": 0.3211435270776029, + "grad_norm": 2.8263528234715074, + "learning_rate": 3.174861437273342e-06, + "loss": 1.0035, + "step": 3561 + }, + { + "epoch": 0.3212337106010732, + "grad_norm": 1.7718088342870313, + "learning_rate": 3.174388600416799e-06, + "loss": 0.967, + "step": 3562 + }, + { + "epoch": 0.32132389412454343, + "grad_norm": 0.6591466146578805, + "learning_rate": 3.1739156633549445e-06, + "loss": 0.8074, + "step": 3563 + }, + { + "epoch": 0.3214140776480137, + "grad_norm": 1.7374786238418862, + "learning_rate": 3.173442626128133e-06, + "loss": 1.0697, + "step": 3564 + }, + { + "epoch": 0.32150426117148395, + "grad_norm": 2.082875722854215, + "learning_rate": 3.1729694887767265e-06, + "loss": 1.1138, + "step": 3565 + }, + { + "epoch": 0.32159444469495424, + "grad_norm": 2.1095771975211113, + "learning_rate": 3.172496251341096e-06, + "loss": 1.0045, + "step": 3566 + }, + { + "epoch": 0.3216846282184245, + "grad_norm": 1.7118798713294443, + "learning_rate": 3.172022913861619e-06, + "loss": 0.8605, + "step": 3567 + }, + { + "epoch": 0.32177481174189476, + "grad_norm": 2.384632553649319, + "learning_rate": 3.171549476378686e-06, + "loss": 0.9605, + "step": 3568 + }, + { + "epoch": 0.321864995265365, + "grad_norm": 1.8641952096490175, + "learning_rate": 3.1710759389326906e-06, + "loss": 0.9754, + "step": 3569 + }, + { + "epoch": 0.3219551787888353, + "grad_norm": 1.8860816906766364, + "learning_rate": 3.1706023015640396e-06, + "loss": 0.9963, + "step": 3570 + }, + { + "epoch": 0.3220453623123055, + "grad_norm": 1.8980173030542902, + "learning_rate": 3.1701285643131453e-06, + "loss": 1.0281, + "step": 3571 + }, + { + "epoch": 0.3221355458357758, + "grad_norm": 2.69371458592336, + "learning_rate": 3.16965472722043e-06, + "loss": 1.0125, + "step": 3572 + }, + { + "epoch": 0.32222572935924604, + "grad_norm": 2.862118264943714, + "learning_rate": 3.169180790326324e-06, + "loss": 0.9304, + "step": 3573 + }, + { + "epoch": 0.3223159128827163, + "grad_norm": 1.946888358130023, + "learning_rate": 3.168706753671266e-06, + "loss": 1.0063, + "step": 3574 + }, + { + "epoch": 0.3224060964061866, + "grad_norm": 1.4657123076049239, + "learning_rate": 3.168232617295704e-06, + "loss": 0.9304, + "step": 3575 + }, + { + "epoch": 0.32249627992965685, + "grad_norm": 2.392679964350477, + "learning_rate": 3.167758381240093e-06, + "loss": 1.0, + "step": 3576 + }, + { + "epoch": 0.32258646345312714, + "grad_norm": 1.6786914089769078, + "learning_rate": 3.1672840455448978e-06, + "loss": 1.0459, + "step": 3577 + }, + { + "epoch": 0.32267664697659737, + "grad_norm": 2.0166821923196863, + "learning_rate": 3.166809610250592e-06, + "loss": 0.9454, + "step": 3578 + }, + { + "epoch": 0.32276683050006766, + "grad_norm": 2.2370746016360097, + "learning_rate": 3.166335075397656e-06, + "loss": 0.9303, + "step": 3579 + }, + { + "epoch": 0.3228570140235379, + "grad_norm": 1.8171990090003263, + "learning_rate": 3.1658604410265808e-06, + "loss": 0.9693, + "step": 3580 + }, + { + "epoch": 0.3229471975470082, + "grad_norm": 0.850537374272933, + "learning_rate": 3.1653857071778644e-06, + "loss": 0.8373, + "step": 3581 + }, + { + "epoch": 0.3230373810704784, + "grad_norm": 3.719894233143454, + "learning_rate": 3.1649108738920133e-06, + "loss": 0.9958, + "step": 3582 + }, + { + "epoch": 0.3231275645939487, + "grad_norm": 1.6442849887757762, + "learning_rate": 3.1644359412095432e-06, + "loss": 1.0062, + "step": 3583 + }, + { + "epoch": 0.32321774811741893, + "grad_norm": 0.7346039273178593, + "learning_rate": 3.163960909170978e-06, + "loss": 0.847, + "step": 3584 + }, + { + "epoch": 0.3233079316408892, + "grad_norm": 1.5891040327707886, + "learning_rate": 3.1634857778168496e-06, + "loss": 0.9016, + "step": 3585 + }, + { + "epoch": 0.32339811516435946, + "grad_norm": 2.250217427238673, + "learning_rate": 3.1630105471877002e-06, + "loss": 1.0387, + "step": 3586 + }, + { + "epoch": 0.32348829868782975, + "grad_norm": 2.228421689861288, + "learning_rate": 3.162535217324077e-06, + "loss": 0.9802, + "step": 3587 + }, + { + "epoch": 0.3235784822113, + "grad_norm": 2.762897763161463, + "learning_rate": 3.1620597882665393e-06, + "loss": 1.0337, + "step": 3588 + }, + { + "epoch": 0.32366866573477027, + "grad_norm": 2.284910218125653, + "learning_rate": 3.1615842600556535e-06, + "loss": 1.0586, + "step": 3589 + }, + { + "epoch": 0.3237588492582405, + "grad_norm": 1.8426056883066317, + "learning_rate": 3.1611086327319932e-06, + "loss": 0.999, + "step": 3590 + }, + { + "epoch": 0.3238490327817108, + "grad_norm": 1.7599615904561614, + "learning_rate": 3.160632906336142e-06, + "loss": 1.0714, + "step": 3591 + }, + { + "epoch": 0.323939216305181, + "grad_norm": 2.1506529326704946, + "learning_rate": 3.160157080908692e-06, + "loss": 1.0259, + "step": 3592 + }, + { + "epoch": 0.3240293998286513, + "grad_norm": 1.7819801306510565, + "learning_rate": 3.1596811564902426e-06, + "loss": 1.0577, + "step": 3593 + }, + { + "epoch": 0.32411958335212154, + "grad_norm": 2.2101664462348207, + "learning_rate": 3.1592051331214023e-06, + "loss": 0.9403, + "step": 3594 + }, + { + "epoch": 0.32420976687559183, + "grad_norm": 1.8352308496064047, + "learning_rate": 3.158729010842789e-06, + "loss": 1.028, + "step": 3595 + }, + { + "epoch": 0.32429995039906206, + "grad_norm": 1.7833778655056638, + "learning_rate": 3.1582527896950266e-06, + "loss": 1.0575, + "step": 3596 + }, + { + "epoch": 0.32439013392253235, + "grad_norm": 1.7336984638016097, + "learning_rate": 3.157776469718749e-06, + "loss": 1.0197, + "step": 3597 + }, + { + "epoch": 0.32448031744600264, + "grad_norm": 2.433111887048152, + "learning_rate": 3.1573000509546004e-06, + "loss": 0.9949, + "step": 3598 + }, + { + "epoch": 0.3245705009694729, + "grad_norm": 1.8348491283416861, + "learning_rate": 3.1568235334432296e-06, + "loss": 1.0581, + "step": 3599 + }, + { + "epoch": 0.32466068449294316, + "grad_norm": 2.0891995301595414, + "learning_rate": 3.1563469172252964e-06, + "loss": 1.072, + "step": 3600 + }, + { + "epoch": 0.3247508680164134, + "grad_norm": 1.780670138557179, + "learning_rate": 3.155870202341468e-06, + "loss": 1.0756, + "step": 3601 + }, + { + "epoch": 0.3248410515398837, + "grad_norm": 2.016667887264334, + "learning_rate": 3.155393388832421e-06, + "loss": 0.9984, + "step": 3602 + }, + { + "epoch": 0.3249312350633539, + "grad_norm": 1.7950329834396783, + "learning_rate": 3.1549164767388386e-06, + "loss": 1.0266, + "step": 3603 + }, + { + "epoch": 0.3250214185868242, + "grad_norm": 6.061014524335755, + "learning_rate": 3.1544394661014145e-06, + "loss": 1.0057, + "step": 3604 + }, + { + "epoch": 0.32511160211029444, + "grad_norm": 1.8857003379694028, + "learning_rate": 3.15396235696085e-06, + "loss": 0.9226, + "step": 3605 + }, + { + "epoch": 0.32520178563376473, + "grad_norm": 3.214589809988885, + "learning_rate": 3.153485149357854e-06, + "loss": 0.9966, + "step": 3606 + }, + { + "epoch": 0.32529196915723496, + "grad_norm": 1.718749306418539, + "learning_rate": 3.153007843333145e-06, + "loss": 0.9908, + "step": 3607 + }, + { + "epoch": 0.32538215268070525, + "grad_norm": 1.8698989143181153, + "learning_rate": 3.152530438927449e-06, + "loss": 1.0367, + "step": 3608 + }, + { + "epoch": 0.3254723362041755, + "grad_norm": 3.257164760386885, + "learning_rate": 3.1520529361815008e-06, + "loss": 1.0478, + "step": 3609 + }, + { + "epoch": 0.32556251972764577, + "grad_norm": 2.4205572665760102, + "learning_rate": 3.151575335136044e-06, + "loss": 1.046, + "step": 3610 + }, + { + "epoch": 0.325652703251116, + "grad_norm": 1.7425040055917156, + "learning_rate": 3.1510976358318298e-06, + "loss": 0.9873, + "step": 3611 + }, + { + "epoch": 0.3257428867745863, + "grad_norm": 1.6454665906140502, + "learning_rate": 3.1506198383096186e-06, + "loss": 0.9584, + "step": 3612 + }, + { + "epoch": 0.3258330702980565, + "grad_norm": 1.659727225125415, + "learning_rate": 3.150141942610178e-06, + "loss": 1.023, + "step": 3613 + }, + { + "epoch": 0.3259232538215268, + "grad_norm": 0.841012824125184, + "learning_rate": 3.1496639487742853e-06, + "loss": 0.8229, + "step": 3614 + }, + { + "epoch": 0.32601343734499705, + "grad_norm": 1.811227417483243, + "learning_rate": 3.1491858568427247e-06, + "loss": 1.0511, + "step": 3615 + }, + { + "epoch": 0.32610362086846734, + "grad_norm": 1.8275840154453509, + "learning_rate": 3.1487076668562903e-06, + "loss": 1.058, + "step": 3616 + }, + { + "epoch": 0.32619380439193757, + "grad_norm": 4.618493529594922, + "learning_rate": 3.1482293788557847e-06, + "loss": 1.0502, + "step": 3617 + }, + { + "epoch": 0.32628398791540786, + "grad_norm": 3.97824738487302, + "learning_rate": 3.1477509928820165e-06, + "loss": 0.9712, + "step": 3618 + }, + { + "epoch": 0.3263741714388781, + "grad_norm": 1.85258700198663, + "learning_rate": 3.147272508975805e-06, + "loss": 1.0081, + "step": 3619 + }, + { + "epoch": 0.3264643549623484, + "grad_norm": 0.6758820342834958, + "learning_rate": 3.1467939271779775e-06, + "loss": 0.8435, + "step": 3620 + }, + { + "epoch": 0.32655453848581867, + "grad_norm": 1.4766443623450634, + "learning_rate": 3.146315247529368e-06, + "loss": 0.9726, + "step": 3621 + }, + { + "epoch": 0.3266447220092889, + "grad_norm": 1.9268765381526345, + "learning_rate": 3.1458364700708212e-06, + "loss": 0.9676, + "step": 3622 + }, + { + "epoch": 0.3267349055327592, + "grad_norm": 2.2450015389621956, + "learning_rate": 3.1453575948431892e-06, + "loss": 1.0341, + "step": 3623 + }, + { + "epoch": 0.3268250890562294, + "grad_norm": 1.6564787940577337, + "learning_rate": 3.144878621887331e-06, + "loss": 1.0113, + "step": 3624 + }, + { + "epoch": 0.3269152725796997, + "grad_norm": 1.932899365828392, + "learning_rate": 3.1443995512441167e-06, + "loss": 1.0474, + "step": 3625 + }, + { + "epoch": 0.32700545610316994, + "grad_norm": 3.6307072928242223, + "learning_rate": 3.1439203829544224e-06, + "loss": 1.0735, + "step": 3626 + }, + { + "epoch": 0.32709563962664023, + "grad_norm": 1.558412699114817, + "learning_rate": 3.143441117059133e-06, + "loss": 1.0887, + "step": 3627 + }, + { + "epoch": 0.32718582315011047, + "grad_norm": 1.9845148472806013, + "learning_rate": 3.142961753599143e-06, + "loss": 0.9543, + "step": 3628 + }, + { + "epoch": 0.32727600667358075, + "grad_norm": 1.7679428820165193, + "learning_rate": 3.1424822926153543e-06, + "loss": 0.974, + "step": 3629 + }, + { + "epoch": 0.327366190197051, + "grad_norm": 1.970942048938075, + "learning_rate": 3.142002734148676e-06, + "loss": 1.0534, + "step": 3630 + }, + { + "epoch": 0.3274563737205213, + "grad_norm": 1.8654068632410024, + "learning_rate": 3.141523078240028e-06, + "loss": 1.0306, + "step": 3631 + }, + { + "epoch": 0.3275465572439915, + "grad_norm": 1.9298513860260345, + "learning_rate": 3.1410433249303366e-06, + "loss": 1.0652, + "step": 3632 + }, + { + "epoch": 0.3276367407674618, + "grad_norm": 2.9410440695637683, + "learning_rate": 3.1405634742605366e-06, + "loss": 1.0964, + "step": 3633 + }, + { + "epoch": 0.32772692429093203, + "grad_norm": 1.921896461429015, + "learning_rate": 3.1400835262715727e-06, + "loss": 1.0492, + "step": 3634 + }, + { + "epoch": 0.3278171078144023, + "grad_norm": 1.9332164666052118, + "learning_rate": 3.139603481004396e-06, + "loss": 0.9963, + "step": 3635 + }, + { + "epoch": 0.32790729133787255, + "grad_norm": 2.0151844578221247, + "learning_rate": 3.139123338499966e-06, + "loss": 1.0209, + "step": 3636 + }, + { + "epoch": 0.32799747486134284, + "grad_norm": 2.0744147199165113, + "learning_rate": 3.1386430987992524e-06, + "loss": 1.0445, + "step": 3637 + }, + { + "epoch": 0.3280876583848131, + "grad_norm": 5.732192952200734, + "learning_rate": 3.1381627619432307e-06, + "loss": 0.9506, + "step": 3638 + }, + { + "epoch": 0.32817784190828336, + "grad_norm": 2.2532596288661733, + "learning_rate": 3.1376823279728864e-06, + "loss": 1.0135, + "step": 3639 + }, + { + "epoch": 0.3282680254317536, + "grad_norm": 1.9206305716098917, + "learning_rate": 3.1372017969292125e-06, + "loss": 1.093, + "step": 3640 + }, + { + "epoch": 0.3283582089552239, + "grad_norm": 2.668412491408365, + "learning_rate": 3.136721168853211e-06, + "loss": 0.9897, + "step": 3641 + }, + { + "epoch": 0.3284483924786941, + "grad_norm": 2.711503731026, + "learning_rate": 3.1362404437858924e-06, + "loss": 0.9713, + "step": 3642 + }, + { + "epoch": 0.3285385760021644, + "grad_norm": 1.6385481268223292, + "learning_rate": 3.135759621768273e-06, + "loss": 1.067, + "step": 3643 + }, + { + "epoch": 0.32862875952563464, + "grad_norm": 2.2917208809653156, + "learning_rate": 3.13527870284138e-06, + "loss": 1.0611, + "step": 3644 + }, + { + "epoch": 0.32871894304910493, + "grad_norm": 2.2218248793559647, + "learning_rate": 3.134797687046249e-06, + "loss": 1.0691, + "step": 3645 + }, + { + "epoch": 0.3288091265725752, + "grad_norm": 2.013529790089079, + "learning_rate": 3.1343165744239218e-06, + "loss": 0.947, + "step": 3646 + }, + { + "epoch": 0.32889931009604545, + "grad_norm": 1.8055239633910014, + "learning_rate": 3.13383536501545e-06, + "loss": 1.0272, + "step": 3647 + }, + { + "epoch": 0.32898949361951574, + "grad_norm": 2.8049428499489064, + "learning_rate": 3.133354058861893e-06, + "loss": 1.0327, + "step": 3648 + }, + { + "epoch": 0.32907967714298597, + "grad_norm": 2.1998257654726885, + "learning_rate": 3.132872656004318e-06, + "loss": 0.9809, + "step": 3649 + }, + { + "epoch": 0.32916986066645626, + "grad_norm": 2.2353461229250655, + "learning_rate": 3.132391156483802e-06, + "loss": 0.9123, + "step": 3650 + }, + { + "epoch": 0.3292600441899265, + "grad_norm": 2.7531275737287113, + "learning_rate": 3.131909560341428e-06, + "loss": 0.9694, + "step": 3651 + }, + { + "epoch": 0.3293502277133968, + "grad_norm": 1.8447326369258776, + "learning_rate": 3.1314278676182893e-06, + "loss": 0.9594, + "step": 3652 + }, + { + "epoch": 0.329440411236867, + "grad_norm": 1.6081217358137583, + "learning_rate": 3.130946078355486e-06, + "loss": 1.0793, + "step": 3653 + }, + { + "epoch": 0.3295305947603373, + "grad_norm": 1.7145379908311085, + "learning_rate": 3.130464192594128e-06, + "loss": 1.0452, + "step": 3654 + }, + { + "epoch": 0.32962077828380754, + "grad_norm": 2.0208990604159665, + "learning_rate": 3.1299822103753315e-06, + "loss": 0.9374, + "step": 3655 + }, + { + "epoch": 0.3297109618072778, + "grad_norm": 2.0769894587947673, + "learning_rate": 3.1295001317402217e-06, + "loss": 1.0108, + "step": 3656 + }, + { + "epoch": 0.32980114533074806, + "grad_norm": 2.9011239405984326, + "learning_rate": 3.1290179567299335e-06, + "loss": 0.9516, + "step": 3657 + }, + { + "epoch": 0.32989132885421835, + "grad_norm": 1.642067701795597, + "learning_rate": 3.128535685385607e-06, + "loss": 1.0404, + "step": 3658 + }, + { + "epoch": 0.3299815123776886, + "grad_norm": 1.820859311207301, + "learning_rate": 3.1280533177483935e-06, + "loss": 1.0907, + "step": 3659 + }, + { + "epoch": 0.33007169590115887, + "grad_norm": 1.9031573363314647, + "learning_rate": 3.127570853859451e-06, + "loss": 1.0229, + "step": 3660 + }, + { + "epoch": 0.3301618794246291, + "grad_norm": 18.631145103591827, + "learning_rate": 3.1270882937599456e-06, + "loss": 0.8975, + "step": 3661 + }, + { + "epoch": 0.3302520629480994, + "grad_norm": 2.7731590453579886, + "learning_rate": 3.1266056374910532e-06, + "loss": 1.0299, + "step": 3662 + }, + { + "epoch": 0.3303422464715696, + "grad_norm": 1.835546963324974, + "learning_rate": 3.126122885093955e-06, + "loss": 0.9085, + "step": 3663 + }, + { + "epoch": 0.3304324299950399, + "grad_norm": 1.768724337550038, + "learning_rate": 3.1256400366098427e-06, + "loss": 1.0035, + "step": 3664 + }, + { + "epoch": 0.33052261351851014, + "grad_norm": 3.275804924509758, + "learning_rate": 3.125157092079916e-06, + "loss": 0.9637, + "step": 3665 + }, + { + "epoch": 0.33061279704198043, + "grad_norm": 2.653629042123387, + "learning_rate": 3.1246740515453824e-06, + "loss": 0.9824, + "step": 3666 + }, + { + "epoch": 0.33070298056545067, + "grad_norm": 1.6152543017116525, + "learning_rate": 3.124190915047457e-06, + "loss": 0.9471, + "step": 3667 + }, + { + "epoch": 0.33079316408892095, + "grad_norm": 3.6842164760197376, + "learning_rate": 3.123707682627364e-06, + "loss": 0.9828, + "step": 3668 + }, + { + "epoch": 0.33088334761239124, + "grad_norm": 2.3209914938307756, + "learning_rate": 3.1232243543263356e-06, + "loss": 0.9911, + "step": 3669 + }, + { + "epoch": 0.3309735311358615, + "grad_norm": 1.9506996660203022, + "learning_rate": 3.1227409301856122e-06, + "loss": 0.911, + "step": 3670 + }, + { + "epoch": 0.33106371465933176, + "grad_norm": 2.106279149971575, + "learning_rate": 3.1222574102464413e-06, + "loss": 0.9916, + "step": 3671 + }, + { + "epoch": 0.331153898182802, + "grad_norm": 1.692981013318312, + "learning_rate": 3.12177379455008e-06, + "loss": 1.0378, + "step": 3672 + }, + { + "epoch": 0.3312440817062723, + "grad_norm": 1.6379999867999364, + "learning_rate": 3.121290083137794e-06, + "loss": 0.9991, + "step": 3673 + }, + { + "epoch": 0.3313342652297425, + "grad_norm": 2.0730255200608916, + "learning_rate": 3.1208062760508547e-06, + "loss": 0.9818, + "step": 3674 + }, + { + "epoch": 0.3314244487532128, + "grad_norm": 1.6184612178592972, + "learning_rate": 3.1203223733305438e-06, + "loss": 0.9506, + "step": 3675 + }, + { + "epoch": 0.33151463227668304, + "grad_norm": 0.6481711288634345, + "learning_rate": 3.1198383750181512e-06, + "loss": 0.8012, + "step": 3676 + }, + { + "epoch": 0.33160481580015333, + "grad_norm": 1.8435719937533586, + "learning_rate": 3.1193542811549734e-06, + "loss": 0.9306, + "step": 3677 + }, + { + "epoch": 0.33169499932362356, + "grad_norm": 1.7156041573125231, + "learning_rate": 3.1188700917823166e-06, + "loss": 0.9315, + "step": 3678 + }, + { + "epoch": 0.33178518284709385, + "grad_norm": 0.6916934349058445, + "learning_rate": 3.1183858069414937e-06, + "loss": 0.8642, + "step": 3679 + }, + { + "epoch": 0.3318753663705641, + "grad_norm": 2.163707938223445, + "learning_rate": 3.117901426673827e-06, + "loss": 1.0547, + "step": 3680 + }, + { + "epoch": 0.3319655498940344, + "grad_norm": 2.549123976334304, + "learning_rate": 3.1174169510206466e-06, + "loss": 0.965, + "step": 3681 + }, + { + "epoch": 0.3320557334175046, + "grad_norm": 2.002514093470895, + "learning_rate": 3.1169323800232908e-06, + "loss": 1.0068, + "step": 3682 + }, + { + "epoch": 0.3321459169409749, + "grad_norm": 1.918918040618071, + "learning_rate": 3.1164477137231054e-06, + "loss": 1.036, + "step": 3683 + }, + { + "epoch": 0.3322361004644451, + "grad_norm": 2.163397070421392, + "learning_rate": 3.115962952161445e-06, + "loss": 1.0119, + "step": 3684 + }, + { + "epoch": 0.3323262839879154, + "grad_norm": 1.9773306579035852, + "learning_rate": 3.1154780953796727e-06, + "loss": 1.0436, + "step": 3685 + }, + { + "epoch": 0.33241646751138565, + "grad_norm": 6.48202288147661, + "learning_rate": 3.114993143419158e-06, + "loss": 0.9113, + "step": 3686 + }, + { + "epoch": 0.33250665103485594, + "grad_norm": 1.541595465931859, + "learning_rate": 3.1145080963212806e-06, + "loss": 0.9154, + "step": 3687 + }, + { + "epoch": 0.33259683455832617, + "grad_norm": 2.036856440845553, + "learning_rate": 3.114022954127427e-06, + "loss": 1.0088, + "step": 3688 + }, + { + "epoch": 0.33268701808179646, + "grad_norm": 1.5381800256440348, + "learning_rate": 3.1135377168789923e-06, + "loss": 1.2128, + "step": 3689 + }, + { + "epoch": 0.3327772016052667, + "grad_norm": 2.9097177469547386, + "learning_rate": 3.1130523846173803e-06, + "loss": 1.0733, + "step": 3690 + }, + { + "epoch": 0.332867385128737, + "grad_norm": 1.8261891736521925, + "learning_rate": 3.1125669573840006e-06, + "loss": 0.9458, + "step": 3691 + }, + { + "epoch": 0.3329575686522072, + "grad_norm": 2.4127390174856878, + "learning_rate": 3.112081435220274e-06, + "loss": 0.9513, + "step": 3692 + }, + { + "epoch": 0.3330477521756775, + "grad_norm": 1.8357445148737774, + "learning_rate": 3.111595818167627e-06, + "loss": 0.9939, + "step": 3693 + }, + { + "epoch": 0.3331379356991478, + "grad_norm": 2.600612157502537, + "learning_rate": 3.1111101062674953e-06, + "loss": 1.0648, + "step": 3694 + }, + { + "epoch": 0.333228119222618, + "grad_norm": 1.8925768022562135, + "learning_rate": 3.1106242995613233e-06, + "loss": 0.9963, + "step": 3695 + }, + { + "epoch": 0.3333183027460883, + "grad_norm": 2.104395435348459, + "learning_rate": 3.1101383980905616e-06, + "loss": 1.0055, + "step": 3696 + }, + { + "epoch": 0.33340848626955855, + "grad_norm": 0.6665995658444474, + "learning_rate": 3.109652401896671e-06, + "loss": 0.789, + "step": 3697 + }, + { + "epoch": 0.33349866979302883, + "grad_norm": 1.4248605760329016, + "learning_rate": 3.109166311021119e-06, + "loss": 1.0514, + "step": 3698 + }, + { + "epoch": 0.33358885331649907, + "grad_norm": 2.6736416692487444, + "learning_rate": 3.1086801255053807e-06, + "loss": 1.0433, + "step": 3699 + }, + { + "epoch": 0.33367903683996936, + "grad_norm": 1.8161439767807432, + "learning_rate": 3.108193845390942e-06, + "loss": 1.0457, + "step": 3700 + }, + { + "epoch": 0.3337692203634396, + "grad_norm": 1.3811905394999322, + "learning_rate": 3.1077074707192933e-06, + "loss": 0.9829, + "step": 3701 + }, + { + "epoch": 0.3338594038869099, + "grad_norm": 1.4210554003456166, + "learning_rate": 3.1072210015319353e-06, + "loss": 1.0235, + "step": 3702 + }, + { + "epoch": 0.3339495874103801, + "grad_norm": 1.6248981737364983, + "learning_rate": 3.106734437870376e-06, + "loss": 0.9602, + "step": 3703 + }, + { + "epoch": 0.3340397709338504, + "grad_norm": 2.0148751690157654, + "learning_rate": 3.1062477797761327e-06, + "loss": 0.9544, + "step": 3704 + }, + { + "epoch": 0.33412995445732063, + "grad_norm": 1.6225868794320173, + "learning_rate": 3.105761027290729e-06, + "loss": 1.0513, + "step": 3705 + }, + { + "epoch": 0.3342201379807909, + "grad_norm": 2.3508157653404576, + "learning_rate": 3.105274180455697e-06, + "loss": 0.9752, + "step": 3706 + }, + { + "epoch": 0.33431032150426115, + "grad_norm": 1.8737298478053774, + "learning_rate": 3.1047872393125775e-06, + "loss": 1.0258, + "step": 3707 + }, + { + "epoch": 0.33440050502773144, + "grad_norm": 0.6285922528314741, + "learning_rate": 3.1043002039029186e-06, + "loss": 0.8171, + "step": 3708 + }, + { + "epoch": 0.3344906885512017, + "grad_norm": 2.125393101852536, + "learning_rate": 3.1038130742682782e-06, + "loss": 0.9967, + "step": 3709 + }, + { + "epoch": 0.33458087207467196, + "grad_norm": 1.5987612906289206, + "learning_rate": 3.103325850450219e-06, + "loss": 0.984, + "step": 3710 + }, + { + "epoch": 0.3346710555981422, + "grad_norm": 2.204054832385664, + "learning_rate": 3.1028385324903154e-06, + "loss": 1.0405, + "step": 3711 + }, + { + "epoch": 0.3347612391216125, + "grad_norm": 2.228131086386651, + "learning_rate": 3.1023511204301465e-06, + "loss": 1.0088, + "step": 3712 + }, + { + "epoch": 0.3348514226450827, + "grad_norm": 1.4603253882484506, + "learning_rate": 3.1018636143113022e-06, + "loss": 0.9656, + "step": 3713 + }, + { + "epoch": 0.334941606168553, + "grad_norm": 1.5735154012357253, + "learning_rate": 3.1013760141753787e-06, + "loss": 0.9087, + "step": 3714 + }, + { + "epoch": 0.33503178969202324, + "grad_norm": 2.2087936521403706, + "learning_rate": 3.100888320063981e-06, + "loss": 0.971, + "step": 3715 + }, + { + "epoch": 0.33512197321549353, + "grad_norm": 1.4496890227862198, + "learning_rate": 3.100400532018721e-06, + "loss": 1.0077, + "step": 3716 + }, + { + "epoch": 0.3352121567389638, + "grad_norm": 1.7300548876616313, + "learning_rate": 3.0999126500812204e-06, + "loss": 1.0727, + "step": 3717 + }, + { + "epoch": 0.33530234026243405, + "grad_norm": 0.7286324087837257, + "learning_rate": 3.0994246742931076e-06, + "loss": 0.8708, + "step": 3718 + }, + { + "epoch": 0.33539252378590434, + "grad_norm": 2.091422538138151, + "learning_rate": 3.098936604696019e-06, + "loss": 0.947, + "step": 3719 + }, + { + "epoch": 0.3354827073093746, + "grad_norm": 2.598333147239393, + "learning_rate": 3.0984484413316e-06, + "loss": 1.0323, + "step": 3720 + }, + { + "epoch": 0.33557289083284486, + "grad_norm": 2.0851864266239732, + "learning_rate": 3.0979601842415033e-06, + "loss": 0.9841, + "step": 3721 + }, + { + "epoch": 0.3356630743563151, + "grad_norm": 1.8733013724025007, + "learning_rate": 3.0974718334673896e-06, + "loss": 0.924, + "step": 3722 + }, + { + "epoch": 0.3357532578797854, + "grad_norm": 2.6753227787825167, + "learning_rate": 3.0969833890509282e-06, + "loss": 0.9952, + "step": 3723 + }, + { + "epoch": 0.3358434414032556, + "grad_norm": 1.8424496348522428, + "learning_rate": 3.096494851033795e-06, + "loss": 0.9971, + "step": 3724 + }, + { + "epoch": 0.3359336249267259, + "grad_norm": 1.881238350562846, + "learning_rate": 3.0960062194576747e-06, + "loss": 0.928, + "step": 3725 + }, + { + "epoch": 0.33602380845019614, + "grad_norm": 1.751741156199558, + "learning_rate": 3.0955174943642606e-06, + "loss": 0.9498, + "step": 3726 + }, + { + "epoch": 0.3361139919736664, + "grad_norm": 1.6327746811278878, + "learning_rate": 3.0950286757952534e-06, + "loss": 0.9229, + "step": 3727 + }, + { + "epoch": 0.33620417549713666, + "grad_norm": 1.7719397585167787, + "learning_rate": 3.0945397637923617e-06, + "loss": 1.0683, + "step": 3728 + }, + { + "epoch": 0.33629435902060695, + "grad_norm": 2.318644530990054, + "learning_rate": 3.0940507583973025e-06, + "loss": 1.0097, + "step": 3729 + }, + { + "epoch": 0.3363845425440772, + "grad_norm": 1.6580616381775721, + "learning_rate": 3.093561659651799e-06, + "loss": 0.9714, + "step": 3730 + }, + { + "epoch": 0.33647472606754747, + "grad_norm": 1.8227353323708764, + "learning_rate": 3.093072467597586e-06, + "loss": 0.9412, + "step": 3731 + }, + { + "epoch": 0.3365649095910177, + "grad_norm": 0.64157157839499, + "learning_rate": 3.092583182276402e-06, + "loss": 0.8226, + "step": 3732 + }, + { + "epoch": 0.336655093114488, + "grad_norm": 0.6261292031943476, + "learning_rate": 3.092093803729997e-06, + "loss": 0.827, + "step": 3733 + }, + { + "epoch": 0.3367452766379582, + "grad_norm": 1.644804145841346, + "learning_rate": 3.0916043320001264e-06, + "loss": 1.0173, + "step": 3734 + }, + { + "epoch": 0.3368354601614285, + "grad_norm": 4.5958504578221495, + "learning_rate": 3.0911147671285557e-06, + "loss": 1.0145, + "step": 3735 + }, + { + "epoch": 0.33692564368489875, + "grad_norm": 3.4812469811143916, + "learning_rate": 3.0906251091570565e-06, + "loss": 1.0336, + "step": 3736 + }, + { + "epoch": 0.33701582720836903, + "grad_norm": 2.334464536938793, + "learning_rate": 3.0901353581274094e-06, + "loss": 1.0149, + "step": 3737 + }, + { + "epoch": 0.33710601073183927, + "grad_norm": 1.6383601222883097, + "learning_rate": 3.089645514081402e-06, + "loss": 0.8827, + "step": 3738 + }, + { + "epoch": 0.33719619425530956, + "grad_norm": 3.9756083664698316, + "learning_rate": 3.0891555770608323e-06, + "loss": 0.9301, + "step": 3739 + }, + { + "epoch": 0.33728637777877984, + "grad_norm": 3.397625003898202, + "learning_rate": 3.088665547107503e-06, + "loss": 1.0282, + "step": 3740 + }, + { + "epoch": 0.3373765613022501, + "grad_norm": 1.5466281424169603, + "learning_rate": 3.0881754242632254e-06, + "loss": 1.0788, + "step": 3741 + }, + { + "epoch": 0.33746674482572037, + "grad_norm": 1.3501596391739283, + "learning_rate": 3.0876852085698213e-06, + "loss": 1.0488, + "step": 3742 + }, + { + "epoch": 0.3375569283491906, + "grad_norm": 2.9732208317341335, + "learning_rate": 3.087194900069117e-06, + "loss": 1.0322, + "step": 3743 + }, + { + "epoch": 0.3376471118726609, + "grad_norm": 1.862545739962745, + "learning_rate": 3.08670449880295e-06, + "loss": 1.0613, + "step": 3744 + }, + { + "epoch": 0.3377372953961311, + "grad_norm": 2.402898262349886, + "learning_rate": 3.086214004813163e-06, + "loss": 1.0256, + "step": 3745 + }, + { + "epoch": 0.3378274789196014, + "grad_norm": 1.5582747742501029, + "learning_rate": 3.0857234181416074e-06, + "loss": 1.0643, + "step": 3746 + }, + { + "epoch": 0.33791766244307164, + "grad_norm": 1.6587458824215846, + "learning_rate": 3.085232738830143e-06, + "loss": 1.0555, + "step": 3747 + }, + { + "epoch": 0.33800784596654193, + "grad_norm": 1.7504646501848105, + "learning_rate": 3.084741966920638e-06, + "loss": 1.0647, + "step": 3748 + }, + { + "epoch": 0.33809802949001216, + "grad_norm": 3.8016416667655295, + "learning_rate": 3.084251102454966e-06, + "loss": 0.9611, + "step": 3749 + }, + { + "epoch": 0.33818821301348245, + "grad_norm": 0.7327494791420134, + "learning_rate": 3.083760145475013e-06, + "loss": 0.8505, + "step": 3750 + }, + { + "epoch": 0.3382783965369527, + "grad_norm": 2.3380089190519366, + "learning_rate": 3.0832690960226678e-06, + "loss": 0.9455, + "step": 3751 + }, + { + "epoch": 0.338368580060423, + "grad_norm": 2.072822287502039, + "learning_rate": 3.08277795413983e-06, + "loss": 0.9698, + "step": 3752 + }, + { + "epoch": 0.3384587635838932, + "grad_norm": 1.7353337705061824, + "learning_rate": 3.0822867198684073e-06, + "loss": 1.055, + "step": 3753 + }, + { + "epoch": 0.3385489471073635, + "grad_norm": 1.9104592798681106, + "learning_rate": 3.081795393250314e-06, + "loss": 1.0116, + "step": 3754 + }, + { + "epoch": 0.33863913063083373, + "grad_norm": 4.865034040310397, + "learning_rate": 3.081303974327473e-06, + "loss": 0.9729, + "step": 3755 + }, + { + "epoch": 0.338729314154304, + "grad_norm": 1.9439213079294437, + "learning_rate": 3.080812463141814e-06, + "loss": 1.0011, + "step": 3756 + }, + { + "epoch": 0.33881949767777425, + "grad_norm": 1.7813145140626496, + "learning_rate": 3.080320859735276e-06, + "loss": 0.8785, + "step": 3757 + }, + { + "epoch": 0.33890968120124454, + "grad_norm": 1.912149212649195, + "learning_rate": 3.079829164149806e-06, + "loss": 1.0575, + "step": 3758 + }, + { + "epoch": 0.3389998647247148, + "grad_norm": 1.8121608548377135, + "learning_rate": 3.0793373764273573e-06, + "loss": 0.9941, + "step": 3759 + }, + { + "epoch": 0.33909004824818506, + "grad_norm": 1.9575474223973344, + "learning_rate": 3.078845496609892e-06, + "loss": 1.0272, + "step": 3760 + }, + { + "epoch": 0.3391802317716553, + "grad_norm": 2.2258284580182557, + "learning_rate": 3.078353524739381e-06, + "loss": 1.094, + "step": 3761 + }, + { + "epoch": 0.3392704152951256, + "grad_norm": 1.9770815322380049, + "learning_rate": 3.077861460857801e-06, + "loss": 1.0917, + "step": 3762 + }, + { + "epoch": 0.3393605988185958, + "grad_norm": 3.6848921525305527, + "learning_rate": 3.077369305007138e-06, + "loss": 1.0255, + "step": 3763 + }, + { + "epoch": 0.3394507823420661, + "grad_norm": 1.7420928655675485, + "learning_rate": 3.0768770572293852e-06, + "loss": 1.0449, + "step": 3764 + }, + { + "epoch": 0.3395409658655364, + "grad_norm": 2.029118519968366, + "learning_rate": 3.0763847175665437e-06, + "loss": 1.086, + "step": 3765 + }, + { + "epoch": 0.3396311493890066, + "grad_norm": 1.7302996895143474, + "learning_rate": 3.0758922860606237e-06, + "loss": 0.9478, + "step": 3766 + }, + { + "epoch": 0.3397213329124769, + "grad_norm": 2.1292352544657587, + "learning_rate": 3.0753997627536404e-06, + "loss": 0.9447, + "step": 3767 + }, + { + "epoch": 0.33981151643594715, + "grad_norm": 1.8316792554599186, + "learning_rate": 3.0749071476876203e-06, + "loss": 0.9547, + "step": 3768 + }, + { + "epoch": 0.33990169995941744, + "grad_norm": 3.1491044239744896, + "learning_rate": 3.0744144409045952e-06, + "loss": 0.9727, + "step": 3769 + }, + { + "epoch": 0.33999188348288767, + "grad_norm": 2.0448038819528422, + "learning_rate": 3.0739216424466056e-06, + "loss": 1.0278, + "step": 3770 + }, + { + "epoch": 0.34008206700635796, + "grad_norm": 2.8285608113902927, + "learning_rate": 3.0734287523557002e-06, + "loss": 0.9934, + "step": 3771 + }, + { + "epoch": 0.3401722505298282, + "grad_norm": 3.0307728500002007, + "learning_rate": 3.0729357706739348e-06, + "loss": 0.9617, + "step": 3772 + }, + { + "epoch": 0.3402624340532985, + "grad_norm": 1.6174873405798376, + "learning_rate": 3.0724426974433737e-06, + "loss": 0.9893, + "step": 3773 + }, + { + "epoch": 0.3403526175767687, + "grad_norm": 1.5215863841264914, + "learning_rate": 3.0719495327060874e-06, + "loss": 0.9812, + "step": 3774 + }, + { + "epoch": 0.340442801100239, + "grad_norm": 65.3367628378159, + "learning_rate": 3.071456276504157e-06, + "loss": 1.0142, + "step": 3775 + }, + { + "epoch": 0.34053298462370923, + "grad_norm": 2.470443241488807, + "learning_rate": 3.070962928879669e-06, + "loss": 1.0026, + "step": 3776 + }, + { + "epoch": 0.3406231681471795, + "grad_norm": 2.162168038527777, + "learning_rate": 3.0704694898747185e-06, + "loss": 1.0218, + "step": 3777 + }, + { + "epoch": 0.34071335167064976, + "grad_norm": 2.5189366311722834, + "learning_rate": 3.069975959531408e-06, + "loss": 1.1296, + "step": 3778 + }, + { + "epoch": 0.34080353519412004, + "grad_norm": 0.6781610250792272, + "learning_rate": 3.06948233789185e-06, + "loss": 0.8856, + "step": 3779 + }, + { + "epoch": 0.3408937187175903, + "grad_norm": 2.137199166527253, + "learning_rate": 3.0689886249981614e-06, + "loss": 1.0286, + "step": 3780 + }, + { + "epoch": 0.34098390224106057, + "grad_norm": 1.5453937258457149, + "learning_rate": 3.0684948208924693e-06, + "loss": 1.0352, + "step": 3781 + }, + { + "epoch": 0.3410740857645308, + "grad_norm": 1.7104039165782068, + "learning_rate": 3.068000925616907e-06, + "loss": 1.0238, + "step": 3782 + }, + { + "epoch": 0.3411642692880011, + "grad_norm": 2.845243292463853, + "learning_rate": 3.067506939213617e-06, + "loss": 0.9974, + "step": 3783 + }, + { + "epoch": 0.3412544528114713, + "grad_norm": 1.51883420671527, + "learning_rate": 3.0670128617247493e-06, + "loss": 1.0138, + "step": 3784 + }, + { + "epoch": 0.3413446363349416, + "grad_norm": 1.7774479091431645, + "learning_rate": 3.06651869319246e-06, + "loss": 0.9207, + "step": 3785 + }, + { + "epoch": 0.34143481985841184, + "grad_norm": 1.7222815558081108, + "learning_rate": 3.0660244336589154e-06, + "loss": 1.0033, + "step": 3786 + }, + { + "epoch": 0.34152500338188213, + "grad_norm": 1.7902540212073867, + "learning_rate": 3.065530083166288e-06, + "loss": 1.0169, + "step": 3787 + }, + { + "epoch": 0.3416151869053524, + "grad_norm": 1.5795947491954356, + "learning_rate": 3.0650356417567586e-06, + "loss": 0.9695, + "step": 3788 + }, + { + "epoch": 0.34170537042882265, + "grad_norm": 1.9266421732396093, + "learning_rate": 3.0645411094725156e-06, + "loss": 1.0536, + "step": 3789 + }, + { + "epoch": 0.34179555395229294, + "grad_norm": 1.6022949032752265, + "learning_rate": 3.0640464863557556e-06, + "loss": 0.999, + "step": 3790 + }, + { + "epoch": 0.3418857374757632, + "grad_norm": 1.4619626614959762, + "learning_rate": 3.063551772448682e-06, + "loss": 1.0406, + "step": 3791 + }, + { + "epoch": 0.34197592099923346, + "grad_norm": 1.929775467693817, + "learning_rate": 3.0630569677935075e-06, + "loss": 0.9348, + "step": 3792 + }, + { + "epoch": 0.3420661045227037, + "grad_norm": 1.9163747095186565, + "learning_rate": 3.06256207243245e-06, + "loss": 1.0892, + "step": 3793 + }, + { + "epoch": 0.342156288046174, + "grad_norm": 0.6030743661686214, + "learning_rate": 3.0620670864077385e-06, + "loss": 0.8054, + "step": 3794 + }, + { + "epoch": 0.3422464715696442, + "grad_norm": 0.6496922406293822, + "learning_rate": 3.0615720097616063e-06, + "loss": 0.8085, + "step": 3795 + }, + { + "epoch": 0.3423366550931145, + "grad_norm": 1.7747966918563705, + "learning_rate": 3.0610768425362967e-06, + "loss": 0.9613, + "step": 3796 + }, + { + "epoch": 0.34242683861658474, + "grad_norm": 1.6243887998986215, + "learning_rate": 3.0605815847740603e-06, + "loss": 0.985, + "step": 3797 + }, + { + "epoch": 0.342517022140055, + "grad_norm": 1.76984488631938, + "learning_rate": 3.0600862365171553e-06, + "loss": 0.9728, + "step": 3798 + }, + { + "epoch": 0.34260720566352526, + "grad_norm": 1.4811589788485837, + "learning_rate": 3.0595907978078474e-06, + "loss": 0.9863, + "step": 3799 + }, + { + "epoch": 0.34269738918699555, + "grad_norm": 1.5313561850340027, + "learning_rate": 3.05909526868841e-06, + "loss": 0.9758, + "step": 3800 + }, + { + "epoch": 0.3427875727104658, + "grad_norm": 2.7997012217377333, + "learning_rate": 3.0585996492011243e-06, + "loss": 1.0154, + "step": 3801 + }, + { + "epoch": 0.34287775623393607, + "grad_norm": 1.803235547463567, + "learning_rate": 3.05810393938828e-06, + "loss": 1.0502, + "step": 3802 + }, + { + "epoch": 0.3429679397574063, + "grad_norm": 1.510786927540292, + "learning_rate": 3.0576081392921723e-06, + "loss": 0.9208, + "step": 3803 + }, + { + "epoch": 0.3430581232808766, + "grad_norm": 1.6647446358086315, + "learning_rate": 3.057112248955107e-06, + "loss": 1.0519, + "step": 3804 + }, + { + "epoch": 0.3431483068043468, + "grad_norm": 2.1232901032906297, + "learning_rate": 3.0566162684193963e-06, + "loss": 1.0267, + "step": 3805 + }, + { + "epoch": 0.3432384903278171, + "grad_norm": 1.5286746263364732, + "learning_rate": 3.056120197727359e-06, + "loss": 0.9537, + "step": 3806 + }, + { + "epoch": 0.34332867385128735, + "grad_norm": 1.6053787540355364, + "learning_rate": 3.0556240369213236e-06, + "loss": 0.9982, + "step": 3807 + }, + { + "epoch": 0.34341885737475764, + "grad_norm": 2.113130856731234, + "learning_rate": 3.055127786043624e-06, + "loss": 0.9181, + "step": 3808 + }, + { + "epoch": 0.34350904089822787, + "grad_norm": 1.5402569468520666, + "learning_rate": 3.054631445136604e-06, + "loss": 1.092, + "step": 3809 + }, + { + "epoch": 0.34359922442169816, + "grad_norm": 2.719993832945845, + "learning_rate": 3.0541350142426147e-06, + "loss": 1.0177, + "step": 3810 + }, + { + "epoch": 0.3436894079451684, + "grad_norm": 1.5578580185831579, + "learning_rate": 3.053638493404012e-06, + "loss": 0.9613, + "step": 3811 + }, + { + "epoch": 0.3437795914686387, + "grad_norm": 1.5079969382267637, + "learning_rate": 3.0531418826631643e-06, + "loss": 0.9419, + "step": 3812 + }, + { + "epoch": 0.34386977499210897, + "grad_norm": 1.5866285800903965, + "learning_rate": 3.052645182062444e-06, + "loss": 1.0884, + "step": 3813 + }, + { + "epoch": 0.3439599585155792, + "grad_norm": 2.167379433205665, + "learning_rate": 3.0521483916442324e-06, + "loss": 1.049, + "step": 3814 + }, + { + "epoch": 0.3440501420390495, + "grad_norm": 2.507049064061778, + "learning_rate": 3.0516515114509183e-06, + "loss": 1.0011, + "step": 3815 + }, + { + "epoch": 0.3441403255625197, + "grad_norm": 1.542268541056104, + "learning_rate": 3.0511545415249e-06, + "loss": 0.9562, + "step": 3816 + }, + { + "epoch": 0.34423050908599, + "grad_norm": 2.6780023642451907, + "learning_rate": 3.050657481908579e-06, + "loss": 1.039, + "step": 3817 + }, + { + "epoch": 0.34432069260946024, + "grad_norm": 1.974483618047418, + "learning_rate": 3.0501603326443677e-06, + "loss": 1.1157, + "step": 3818 + }, + { + "epoch": 0.34441087613293053, + "grad_norm": 1.8746728611708556, + "learning_rate": 3.049663093774687e-06, + "loss": 0.9922, + "step": 3819 + }, + { + "epoch": 0.34450105965640077, + "grad_norm": 4.869986474803652, + "learning_rate": 3.0491657653419643e-06, + "loss": 0.9017, + "step": 3820 + }, + { + "epoch": 0.34459124317987105, + "grad_norm": 1.656515711936969, + "learning_rate": 3.0486683473886325e-06, + "loss": 1.0509, + "step": 3821 + }, + { + "epoch": 0.3446814267033413, + "grad_norm": 2.7179673043839574, + "learning_rate": 3.0481708399571355e-06, + "loss": 1.0655, + "step": 3822 + }, + { + "epoch": 0.3447716102268116, + "grad_norm": 1.856214192876737, + "learning_rate": 3.047673243089922e-06, + "loss": 1.0275, + "step": 3823 + }, + { + "epoch": 0.3448617937502818, + "grad_norm": 2.379603191442717, + "learning_rate": 3.047175556829451e-06, + "loss": 1.0345, + "step": 3824 + }, + { + "epoch": 0.3449519772737521, + "grad_norm": 2.1231226762787525, + "learning_rate": 3.046677781218188e-06, + "loss": 1.0529, + "step": 3825 + }, + { + "epoch": 0.34504216079722233, + "grad_norm": 1.5706015553301444, + "learning_rate": 3.0461799162986043e-06, + "loss": 0.9841, + "step": 3826 + }, + { + "epoch": 0.3451323443206926, + "grad_norm": 3.124283517718018, + "learning_rate": 3.045681962113183e-06, + "loss": 0.9314, + "step": 3827 + }, + { + "epoch": 0.34522252784416285, + "grad_norm": 2.720527681607751, + "learning_rate": 3.0451839187044095e-06, + "loss": 0.9581, + "step": 3828 + }, + { + "epoch": 0.34531271136763314, + "grad_norm": 2.016423503413206, + "learning_rate": 3.0446857861147816e-06, + "loss": 1.021, + "step": 3829 + }, + { + "epoch": 0.3454028948911034, + "grad_norm": 1.69685265682678, + "learning_rate": 3.044187564386802e-06, + "loss": 0.9801, + "step": 3830 + }, + { + "epoch": 0.34549307841457366, + "grad_norm": 2.0107969672145853, + "learning_rate": 3.0436892535629818e-06, + "loss": 1.0687, + "step": 3831 + }, + { + "epoch": 0.3455832619380439, + "grad_norm": 2.301162247774432, + "learning_rate": 3.0431908536858393e-06, + "loss": 1.0854, + "step": 3832 + }, + { + "epoch": 0.3456734454615142, + "grad_norm": 2.2215170999916745, + "learning_rate": 3.0426923647979016e-06, + "loss": 0.9961, + "step": 3833 + }, + { + "epoch": 0.3457636289849844, + "grad_norm": 1.5975980491576607, + "learning_rate": 3.0421937869417016e-06, + "loss": 0.9437, + "step": 3834 + }, + { + "epoch": 0.3458538125084547, + "grad_norm": 2.0913322495488202, + "learning_rate": 3.041695120159782e-06, + "loss": 1.0708, + "step": 3835 + }, + { + "epoch": 0.345943996031925, + "grad_norm": 3.8658419030904434, + "learning_rate": 3.04119636449469e-06, + "loss": 0.9387, + "step": 3836 + }, + { + "epoch": 0.3460341795553952, + "grad_norm": 1.4597605487999612, + "learning_rate": 3.040697519988983e-06, + "loss": 1.0361, + "step": 3837 + }, + { + "epoch": 0.3461243630788655, + "grad_norm": 0.6420107484553527, + "learning_rate": 3.040198586685226e-06, + "loss": 0.8283, + "step": 3838 + }, + { + "epoch": 0.34621454660233575, + "grad_norm": 20.975528809493472, + "learning_rate": 3.039699564625989e-06, + "loss": 0.959, + "step": 3839 + }, + { + "epoch": 0.34630473012580604, + "grad_norm": 0.6750290935039179, + "learning_rate": 3.039200453853853e-06, + "loss": 0.8223, + "step": 3840 + }, + { + "epoch": 0.34639491364927627, + "grad_norm": 2.123940708697802, + "learning_rate": 3.038701254411404e-06, + "loss": 1.0584, + "step": 3841 + }, + { + "epoch": 0.34648509717274656, + "grad_norm": 1.7255670085531538, + "learning_rate": 3.0382019663412367e-06, + "loss": 1.0242, + "step": 3842 + }, + { + "epoch": 0.3465752806962168, + "grad_norm": 2.2650189871760604, + "learning_rate": 3.0377025896859532e-06, + "loss": 1.0425, + "step": 3843 + }, + { + "epoch": 0.3466654642196871, + "grad_norm": 1.6338466308082715, + "learning_rate": 3.0372031244881627e-06, + "loss": 1.1137, + "step": 3844 + }, + { + "epoch": 0.3467556477431573, + "grad_norm": 2.494519711977603, + "learning_rate": 3.0367035707904826e-06, + "loss": 0.9645, + "step": 3845 + }, + { + "epoch": 0.3468458312666276, + "grad_norm": 1.7696200384302327, + "learning_rate": 3.036203928635537e-06, + "loss": 0.9207, + "step": 3846 + }, + { + "epoch": 0.34693601479009784, + "grad_norm": 2.029411540311913, + "learning_rate": 3.035704198065959e-06, + "loss": 1.0728, + "step": 3847 + }, + { + "epoch": 0.3470261983135681, + "grad_norm": 2.393606881825861, + "learning_rate": 3.0352043791243886e-06, + "loss": 0.996, + "step": 3848 + }, + { + "epoch": 0.34711638183703836, + "grad_norm": 0.6614677899102689, + "learning_rate": 3.034704471853472e-06, + "loss": 0.8263, + "step": 3849 + }, + { + "epoch": 0.34720656536050865, + "grad_norm": 1.6832790567441678, + "learning_rate": 3.0342044762958646e-06, + "loss": 1.0555, + "step": 3850 + }, + { + "epoch": 0.3472967488839789, + "grad_norm": 1.679607300728913, + "learning_rate": 3.0337043924942286e-06, + "loss": 1.0443, + "step": 3851 + }, + { + "epoch": 0.34738693240744917, + "grad_norm": 1.7522953511466457, + "learning_rate": 3.0332042204912343e-06, + "loss": 0.9929, + "step": 3852 + }, + { + "epoch": 0.3474771159309194, + "grad_norm": 0.6739906667847023, + "learning_rate": 3.0327039603295587e-06, + "loss": 0.831, + "step": 3853 + }, + { + "epoch": 0.3475672994543897, + "grad_norm": 2.1233742048361206, + "learning_rate": 3.032203612051887e-06, + "loss": 0.9676, + "step": 3854 + }, + { + "epoch": 0.3476574829778599, + "grad_norm": 1.65839398367435, + "learning_rate": 3.0317031757009116e-06, + "loss": 0.9863, + "step": 3855 + }, + { + "epoch": 0.3477476665013302, + "grad_norm": 1.8084447152544403, + "learning_rate": 3.0312026513193326e-06, + "loss": 0.9681, + "step": 3856 + }, + { + "epoch": 0.34783785002480044, + "grad_norm": 1.7592213050287728, + "learning_rate": 3.0307020389498573e-06, + "loss": 0.9803, + "step": 3857 + }, + { + "epoch": 0.34792803354827073, + "grad_norm": 1.744491968938393, + "learning_rate": 3.0302013386352004e-06, + "loss": 0.9365, + "step": 3858 + }, + { + "epoch": 0.348018217071741, + "grad_norm": 1.4229873665726676, + "learning_rate": 3.0297005504180854e-06, + "loss": 0.9982, + "step": 3859 + }, + { + "epoch": 0.34810840059521125, + "grad_norm": 0.6004197053437527, + "learning_rate": 3.0291996743412417e-06, + "loss": 0.7495, + "step": 3860 + }, + { + "epoch": 0.34819858411868154, + "grad_norm": 1.7761463869009828, + "learning_rate": 3.0286987104474063e-06, + "loss": 1.099, + "step": 3861 + }, + { + "epoch": 0.3482887676421518, + "grad_norm": 1.3224313201089233, + "learning_rate": 3.028197658779325e-06, + "loss": 0.9814, + "step": 3862 + }, + { + "epoch": 0.34837895116562206, + "grad_norm": 2.1606729451698072, + "learning_rate": 3.0276965193797503e-06, + "loss": 1.0254, + "step": 3863 + }, + { + "epoch": 0.3484691346890923, + "grad_norm": 1.468960199127217, + "learning_rate": 3.0271952922914423e-06, + "loss": 0.9689, + "step": 3864 + }, + { + "epoch": 0.3485593182125626, + "grad_norm": 1.6982843745883154, + "learning_rate": 3.0266939775571675e-06, + "loss": 0.9853, + "step": 3865 + }, + { + "epoch": 0.3486495017360328, + "grad_norm": 1.8147518376079526, + "learning_rate": 3.026192575219701e-06, + "loss": 0.996, + "step": 3866 + }, + { + "epoch": 0.3487396852595031, + "grad_norm": 1.823243037480553, + "learning_rate": 3.025691085321826e-06, + "loss": 0.8982, + "step": 3867 + }, + { + "epoch": 0.34882986878297334, + "grad_norm": 1.908311948142798, + "learning_rate": 3.025189507906332e-06, + "loss": 0.9877, + "step": 3868 + }, + { + "epoch": 0.34892005230644363, + "grad_norm": 1.724927516464821, + "learning_rate": 3.0246878430160166e-06, + "loss": 1.023, + "step": 3869 + }, + { + "epoch": 0.34901023582991386, + "grad_norm": 2.5341284127459636, + "learning_rate": 3.024186090693684e-06, + "loss": 0.9464, + "step": 3870 + }, + { + "epoch": 0.34910041935338415, + "grad_norm": 2.0178506081881613, + "learning_rate": 3.023684250982147e-06, + "loss": 1.0285, + "step": 3871 + }, + { + "epoch": 0.3491906028768544, + "grad_norm": 2.3628341882479957, + "learning_rate": 3.0231823239242252e-06, + "loss": 0.9252, + "step": 3872 + }, + { + "epoch": 0.34928078640032467, + "grad_norm": 1.7016726959948654, + "learning_rate": 3.0226803095627457e-06, + "loss": 0.9628, + "step": 3873 + }, + { + "epoch": 0.3493709699237949, + "grad_norm": 2.1439075779067998, + "learning_rate": 3.022178207940543e-06, + "loss": 0.9814, + "step": 3874 + }, + { + "epoch": 0.3494611534472652, + "grad_norm": 2.2471739717303625, + "learning_rate": 3.02167601910046e-06, + "loss": 1.0203, + "step": 3875 + }, + { + "epoch": 0.3495513369707354, + "grad_norm": 1.5667413747844363, + "learning_rate": 3.021173743085345e-06, + "loss": 1.0268, + "step": 3876 + }, + { + "epoch": 0.3496415204942057, + "grad_norm": 0.6496806121138113, + "learning_rate": 3.0206713799380557e-06, + "loss": 0.816, + "step": 3877 + }, + { + "epoch": 0.34973170401767595, + "grad_norm": 2.1102850187261457, + "learning_rate": 3.0201689297014565e-06, + "loss": 1.0148, + "step": 3878 + }, + { + "epoch": 0.34982188754114624, + "grad_norm": 1.295333497690559, + "learning_rate": 3.0196663924184187e-06, + "loss": 1.0451, + "step": 3879 + }, + { + "epoch": 0.34991207106461647, + "grad_norm": 2.212420178713839, + "learning_rate": 3.019163768131822e-06, + "loss": 1.0371, + "step": 3880 + }, + { + "epoch": 0.35000225458808676, + "grad_norm": 3.1838008883437814, + "learning_rate": 3.0186610568845533e-06, + "loss": 0.9226, + "step": 3881 + }, + { + "epoch": 0.350092438111557, + "grad_norm": 2.2274541842868336, + "learning_rate": 3.018158258719507e-06, + "loss": 0.9422, + "step": 3882 + }, + { + "epoch": 0.3501826216350273, + "grad_norm": 1.7137539470245964, + "learning_rate": 3.0176553736795827e-06, + "loss": 1.0293, + "step": 3883 + }, + { + "epoch": 0.35027280515849757, + "grad_norm": 0.5362117966246043, + "learning_rate": 3.017152401807691e-06, + "loss": 0.7541, + "step": 3884 + }, + { + "epoch": 0.3503629886819678, + "grad_norm": 1.8219644148331924, + "learning_rate": 3.0166493431467476e-06, + "loss": 0.9871, + "step": 3885 + }, + { + "epoch": 0.3504531722054381, + "grad_norm": 1.5095367060693905, + "learning_rate": 3.016146197739677e-06, + "loss": 0.9771, + "step": 3886 + }, + { + "epoch": 0.3505433557289083, + "grad_norm": 1.6497844237325825, + "learning_rate": 3.0156429656294097e-06, + "loss": 0.9383, + "step": 3887 + }, + { + "epoch": 0.3506335392523786, + "grad_norm": 1.570811481812908, + "learning_rate": 3.0151396468588844e-06, + "loss": 0.9438, + "step": 3888 + }, + { + "epoch": 0.35072372277584885, + "grad_norm": 2.8613951170698555, + "learning_rate": 3.014636241471047e-06, + "loss": 0.9556, + "step": 3889 + }, + { + "epoch": 0.35081390629931913, + "grad_norm": 11.556792298998348, + "learning_rate": 3.0141327495088514e-06, + "loss": 0.9973, + "step": 3890 + }, + { + "epoch": 0.35090408982278937, + "grad_norm": 2.081486671898071, + "learning_rate": 3.0136291710152566e-06, + "loss": 0.9602, + "step": 3891 + }, + { + "epoch": 0.35099427334625966, + "grad_norm": 1.6904407591212023, + "learning_rate": 3.0131255060332325e-06, + "loss": 1.0417, + "step": 3892 + }, + { + "epoch": 0.3510844568697299, + "grad_norm": 2.0781464539761347, + "learning_rate": 3.012621754605754e-06, + "loss": 1.0191, + "step": 3893 + }, + { + "epoch": 0.3511746403932002, + "grad_norm": 1.8826709868248244, + "learning_rate": 3.0121179167758035e-06, + "loss": 1.0271, + "step": 3894 + }, + { + "epoch": 0.3512648239166704, + "grad_norm": 1.9061895892467222, + "learning_rate": 3.0116139925863717e-06, + "loss": 1.08, + "step": 3895 + }, + { + "epoch": 0.3513550074401407, + "grad_norm": 2.527493267559741, + "learning_rate": 3.011109982080456e-06, + "loss": 1.0392, + "step": 3896 + }, + { + "epoch": 0.35144519096361093, + "grad_norm": 1.725858311726415, + "learning_rate": 3.0106058853010614e-06, + "loss": 0.9453, + "step": 3897 + }, + { + "epoch": 0.3515353744870812, + "grad_norm": 2.4010195672486283, + "learning_rate": 3.010101702291201e-06, + "loss": 0.9876, + "step": 3898 + }, + { + "epoch": 0.35162555801055145, + "grad_norm": 1.9319502193798221, + "learning_rate": 3.009597433093893e-06, + "loss": 0.996, + "step": 3899 + }, + { + "epoch": 0.35171574153402174, + "grad_norm": 6.669285291730309, + "learning_rate": 3.009093077752165e-06, + "loss": 1.0121, + "step": 3900 + }, + { + "epoch": 0.351805925057492, + "grad_norm": 1.5617528273360086, + "learning_rate": 3.008588636309052e-06, + "loss": 1.0322, + "step": 3901 + }, + { + "epoch": 0.35189610858096226, + "grad_norm": 1.8163270994599696, + "learning_rate": 3.0080841088075947e-06, + "loss": 0.9664, + "step": 3902 + }, + { + "epoch": 0.3519862921044325, + "grad_norm": 1.8029395551869156, + "learning_rate": 3.0075794952908436e-06, + "loss": 1.042, + "step": 3903 + }, + { + "epoch": 0.3520764756279028, + "grad_norm": 1.6751760532605893, + "learning_rate": 3.0070747958018528e-06, + "loss": 1.0145, + "step": 3904 + }, + { + "epoch": 0.352166659151373, + "grad_norm": 1.680627249492418, + "learning_rate": 3.0065700103836894e-06, + "loss": 1.015, + "step": 3905 + }, + { + "epoch": 0.3522568426748433, + "grad_norm": 1.7064707379205986, + "learning_rate": 3.0060651390794214e-06, + "loss": 1.0208, + "step": 3906 + }, + { + "epoch": 0.3523470261983136, + "grad_norm": 1.4415651972935861, + "learning_rate": 3.005560181932128e-06, + "loss": 0.9807, + "step": 3907 + }, + { + "epoch": 0.35243720972178383, + "grad_norm": 1.6119698651055177, + "learning_rate": 3.005055138984896e-06, + "loss": 0.9556, + "step": 3908 + }, + { + "epoch": 0.3525273932452541, + "grad_norm": 3.660847178316555, + "learning_rate": 3.0045500102808174e-06, + "loss": 0.9752, + "step": 3909 + }, + { + "epoch": 0.35261757676872435, + "grad_norm": 2.7748689637734754, + "learning_rate": 3.0040447958629927e-06, + "loss": 1.0836, + "step": 3910 + }, + { + "epoch": 0.35270776029219464, + "grad_norm": 1.619136575617863, + "learning_rate": 3.00353949577453e-06, + "loss": 1.025, + "step": 3911 + }, + { + "epoch": 0.35279794381566487, + "grad_norm": 2.2355599629127196, + "learning_rate": 3.003034110058544e-06, + "loss": 0.8364, + "step": 3912 + }, + { + "epoch": 0.35288812733913516, + "grad_norm": 1.4625824831664394, + "learning_rate": 3.002528638758157e-06, + "loss": 0.9835, + "step": 3913 + }, + { + "epoch": 0.3529783108626054, + "grad_norm": 2.5966088958536138, + "learning_rate": 3.0020230819164985e-06, + "loss": 0.9933, + "step": 3914 + }, + { + "epoch": 0.3530684943860757, + "grad_norm": 1.4207476663726377, + "learning_rate": 3.0015174395767064e-06, + "loss": 1.008, + "step": 3915 + }, + { + "epoch": 0.3531586779095459, + "grad_norm": 2.3217146519040024, + "learning_rate": 3.001011711781923e-06, + "loss": 1.0239, + "step": 3916 + }, + { + "epoch": 0.3532488614330162, + "grad_norm": 2.1247878249213117, + "learning_rate": 3.0005058985753017e-06, + "loss": 0.9579, + "step": 3917 + }, + { + "epoch": 0.35333904495648644, + "grad_norm": 1.8261394314255421, + "learning_rate": 3e-06, + "loss": 0.9898, + "step": 3918 + }, + { + "epoch": 0.3534292284799567, + "grad_norm": 0.649982266917788, + "learning_rate": 2.9994940160991843e-06, + "loss": 0.8423, + "step": 3919 + }, + { + "epoch": 0.35351941200342696, + "grad_norm": 0.6263186848449932, + "learning_rate": 2.9989879469160285e-06, + "loss": 0.8001, + "step": 3920 + }, + { + "epoch": 0.35360959552689725, + "grad_norm": 1.623909437590647, + "learning_rate": 2.9984817924937124e-06, + "loss": 1.0184, + "step": 3921 + }, + { + "epoch": 0.3536997790503675, + "grad_norm": 1.5430638537492336, + "learning_rate": 2.997975552875424e-06, + "loss": 0.9364, + "step": 3922 + }, + { + "epoch": 0.35378996257383777, + "grad_norm": 1.4686714313676255, + "learning_rate": 2.997469228104358e-06, + "loss": 0.9893, + "step": 3923 + }, + { + "epoch": 0.353880146097308, + "grad_norm": 2.45062186464957, + "learning_rate": 2.996962818223718e-06, + "loss": 0.8965, + "step": 3924 + }, + { + "epoch": 0.3539703296207783, + "grad_norm": 2.2165221728888063, + "learning_rate": 2.9964563232767135e-06, + "loss": 0.943, + "step": 3925 + }, + { + "epoch": 0.3540605131442485, + "grad_norm": 1.8737204318166185, + "learning_rate": 2.9959497433065617e-06, + "loss": 0.9957, + "step": 3926 + }, + { + "epoch": 0.3541506966677188, + "grad_norm": 1.835023433317399, + "learning_rate": 2.9954430783564848e-06, + "loss": 0.9195, + "step": 3927 + }, + { + "epoch": 0.35424088019118904, + "grad_norm": 2.240245549673797, + "learning_rate": 2.994936328469716e-06, + "loss": 1.025, + "step": 3928 + }, + { + "epoch": 0.35433106371465933, + "grad_norm": 2.4346263280760025, + "learning_rate": 2.994429493689494e-06, + "loss": 1.0021, + "step": 3929 + }, + { + "epoch": 0.35442124723812957, + "grad_norm": 1.4111494967520388, + "learning_rate": 2.9939225740590642e-06, + "loss": 1.024, + "step": 3930 + }, + { + "epoch": 0.35451143076159986, + "grad_norm": 1.5496017621264189, + "learning_rate": 2.99341556962168e-06, + "loss": 0.9548, + "step": 3931 + }, + { + "epoch": 0.35460161428507014, + "grad_norm": 1.5196362927371732, + "learning_rate": 2.992908480420602e-06, + "loss": 1.0412, + "step": 3932 + }, + { + "epoch": 0.3546917978085404, + "grad_norm": 1.7712564374312725, + "learning_rate": 2.9924013064990974e-06, + "loss": 0.9892, + "step": 3933 + }, + { + "epoch": 0.35478198133201067, + "grad_norm": 0.6476516097653764, + "learning_rate": 2.991894047900441e-06, + "loss": 0.8473, + "step": 3934 + }, + { + "epoch": 0.3548721648554809, + "grad_norm": 1.9636659523833067, + "learning_rate": 2.991386704667916e-06, + "loss": 1.0165, + "step": 3935 + }, + { + "epoch": 0.3549623483789512, + "grad_norm": 1.858125218539926, + "learning_rate": 2.9908792768448097e-06, + "loss": 1.0651, + "step": 3936 + }, + { + "epoch": 0.3550525319024214, + "grad_norm": 3.413657856231585, + "learning_rate": 2.990371764474421e-06, + "loss": 0.9971, + "step": 3937 + }, + { + "epoch": 0.3551427154258917, + "grad_norm": 1.705893969157752, + "learning_rate": 2.9898641676000518e-06, + "loss": 0.969, + "step": 3938 + }, + { + "epoch": 0.35523289894936194, + "grad_norm": 1.5504677405176757, + "learning_rate": 2.9893564862650138e-06, + "loss": 0.9964, + "step": 3939 + }, + { + "epoch": 0.35532308247283223, + "grad_norm": 1.8452598725775649, + "learning_rate": 2.9888487205126254e-06, + "loss": 1.0495, + "step": 3940 + }, + { + "epoch": 0.35541326599630246, + "grad_norm": 1.4786959781275497, + "learning_rate": 2.9883408703862115e-06, + "loss": 0.9956, + "step": 3941 + }, + { + "epoch": 0.35550344951977275, + "grad_norm": 1.9927235798336407, + "learning_rate": 2.987832935929105e-06, + "loss": 1.0783, + "step": 3942 + }, + { + "epoch": 0.355593633043243, + "grad_norm": 1.8584617487696686, + "learning_rate": 2.9873249171846454e-06, + "loss": 1.0618, + "step": 3943 + }, + { + "epoch": 0.3556838165667133, + "grad_norm": 2.0742855312710637, + "learning_rate": 2.98681681419618e-06, + "loss": 0.9867, + "step": 3944 + }, + { + "epoch": 0.3557740000901835, + "grad_norm": 1.9400327187168191, + "learning_rate": 2.9863086270070627e-06, + "loss": 1.0512, + "step": 3945 + }, + { + "epoch": 0.3558641836136538, + "grad_norm": 1.9331362404605965, + "learning_rate": 2.985800355660655e-06, + "loss": 1.0039, + "step": 3946 + }, + { + "epoch": 0.35595436713712403, + "grad_norm": 0.7307465784813006, + "learning_rate": 2.9852920002003252e-06, + "loss": 0.8457, + "step": 3947 + }, + { + "epoch": 0.3560445506605943, + "grad_norm": 1.9848500681832657, + "learning_rate": 2.9847835606694494e-06, + "loss": 1.0378, + "step": 3948 + }, + { + "epoch": 0.35613473418406455, + "grad_norm": 2.3944476114573217, + "learning_rate": 2.98427503711141e-06, + "loss": 0.9268, + "step": 3949 + }, + { + "epoch": 0.35622491770753484, + "grad_norm": 2.0068548271236906, + "learning_rate": 2.9837664295695973e-06, + "loss": 1.031, + "step": 3950 + }, + { + "epoch": 0.35631510123100507, + "grad_norm": 1.651317088102029, + "learning_rate": 2.983257738087408e-06, + "loss": 0.9926, + "step": 3951 + }, + { + "epoch": 0.35640528475447536, + "grad_norm": 2.947489695187222, + "learning_rate": 2.982748962708247e-06, + "loss": 1.0305, + "step": 3952 + }, + { + "epoch": 0.3564954682779456, + "grad_norm": 1.7064745102059227, + "learning_rate": 2.982240103475526e-06, + "loss": 1.0533, + "step": 3953 + }, + { + "epoch": 0.3565856518014159, + "grad_norm": 1.9742573198808582, + "learning_rate": 2.981731160432663e-06, + "loss": 1.002, + "step": 3954 + }, + { + "epoch": 0.35667583532488617, + "grad_norm": 2.038633104033516, + "learning_rate": 2.981222133623084e-06, + "loss": 0.9016, + "step": 3955 + }, + { + "epoch": 0.3567660188483564, + "grad_norm": 2.177782983622308, + "learning_rate": 2.980713023090222e-06, + "loss": 1.0074, + "step": 3956 + }, + { + "epoch": 0.3568562023718267, + "grad_norm": 1.6734427116305766, + "learning_rate": 2.980203828877518e-06, + "loss": 0.9741, + "step": 3957 + }, + { + "epoch": 0.3569463858952969, + "grad_norm": 2.7375991559469326, + "learning_rate": 2.9796945510284182e-06, + "loss": 1.0431, + "step": 3958 + }, + { + "epoch": 0.3570365694187672, + "grad_norm": 1.977398842443878, + "learning_rate": 2.9791851895863774e-06, + "loss": 1.056, + "step": 3959 + }, + { + "epoch": 0.35712675294223745, + "grad_norm": 1.9068336218761546, + "learning_rate": 2.978675744594857e-06, + "loss": 1.0314, + "step": 3960 + }, + { + "epoch": 0.35721693646570774, + "grad_norm": 1.4232579307475905, + "learning_rate": 2.978166216097326e-06, + "loss": 0.9576, + "step": 3961 + }, + { + "epoch": 0.35730711998917797, + "grad_norm": 1.5856124469649622, + "learning_rate": 2.9776566041372596e-06, + "loss": 0.9841, + "step": 3962 + }, + { + "epoch": 0.35739730351264826, + "grad_norm": 1.750406354319739, + "learning_rate": 2.977146908758141e-06, + "loss": 1.0589, + "step": 3963 + }, + { + "epoch": 0.3574874870361185, + "grad_norm": 1.8883894637211667, + "learning_rate": 2.9766371300034604e-06, + "loss": 1.0172, + "step": 3964 + }, + { + "epoch": 0.3575776705595888, + "grad_norm": 1.612529293060372, + "learning_rate": 2.9761272679167142e-06, + "loss": 1.0494, + "step": 3965 + }, + { + "epoch": 0.357667854083059, + "grad_norm": 1.5963725036596796, + "learning_rate": 2.9756173225414072e-06, + "loss": 0.8491, + "step": 3966 + }, + { + "epoch": 0.3577580376065293, + "grad_norm": 1.7865432259395895, + "learning_rate": 2.975107293921051e-06, + "loss": 1.1726, + "step": 3967 + }, + { + "epoch": 0.35784822112999953, + "grad_norm": 2.2008280582887156, + "learning_rate": 2.9745971820991643e-06, + "loss": 1.0423, + "step": 3968 + }, + { + "epoch": 0.3579384046534698, + "grad_norm": 1.9829717766027277, + "learning_rate": 2.9740869871192715e-06, + "loss": 1.0406, + "step": 3969 + }, + { + "epoch": 0.35802858817694005, + "grad_norm": 1.7727724416353547, + "learning_rate": 2.9735767090249065e-06, + "loss": 1.0621, + "step": 3970 + }, + { + "epoch": 0.35811877170041034, + "grad_norm": 1.7202384745706947, + "learning_rate": 2.973066347859608e-06, + "loss": 0.9141, + "step": 3971 + }, + { + "epoch": 0.3582089552238806, + "grad_norm": 2.5009787550934846, + "learning_rate": 2.972555903666923e-06, + "loss": 1.0145, + "step": 3972 + }, + { + "epoch": 0.35829913874735086, + "grad_norm": 2.8915720445027486, + "learning_rate": 2.972045376490406e-06, + "loss": 0.911, + "step": 3973 + }, + { + "epoch": 0.3583893222708211, + "grad_norm": 1.5626721096616696, + "learning_rate": 2.9715347663736177e-06, + "loss": 0.9276, + "step": 3974 + }, + { + "epoch": 0.3584795057942914, + "grad_norm": 1.746365178323128, + "learning_rate": 2.9710240733601266e-06, + "loss": 0.9987, + "step": 3975 + }, + { + "epoch": 0.3585696893177616, + "grad_norm": 2.0681089946718894, + "learning_rate": 2.970513297493507e-06, + "loss": 1.085, + "step": 3976 + }, + { + "epoch": 0.3586598728412319, + "grad_norm": 1.4351285363824842, + "learning_rate": 2.9700024388173416e-06, + "loss": 1.051, + "step": 3977 + }, + { + "epoch": 0.35875005636470214, + "grad_norm": 1.6156144162095156, + "learning_rate": 2.969491497375219e-06, + "loss": 1.0529, + "step": 3978 + }, + { + "epoch": 0.35884023988817243, + "grad_norm": 1.4056180593710212, + "learning_rate": 2.9689804732107364e-06, + "loss": 1.0219, + "step": 3979 + }, + { + "epoch": 0.3589304234116427, + "grad_norm": 5.375301796180789, + "learning_rate": 2.9684693663674968e-06, + "loss": 0.9609, + "step": 3980 + }, + { + "epoch": 0.35902060693511295, + "grad_norm": 1.8560071946941887, + "learning_rate": 2.9679581768891115e-06, + "loss": 1.0836, + "step": 3981 + }, + { + "epoch": 0.35911079045858324, + "grad_norm": 1.5756754985798793, + "learning_rate": 2.967446904819197e-06, + "loss": 0.9246, + "step": 3982 + }, + { + "epoch": 0.3592009739820535, + "grad_norm": 1.5566142808702417, + "learning_rate": 2.966935550201378e-06, + "loss": 1.0273, + "step": 3983 + }, + { + "epoch": 0.35929115750552376, + "grad_norm": 1.6047409895587346, + "learning_rate": 2.966424113079286e-06, + "loss": 1.0003, + "step": 3984 + }, + { + "epoch": 0.359381341028994, + "grad_norm": 5.889679769673954, + "learning_rate": 2.9659125934965596e-06, + "loss": 1.0532, + "step": 3985 + }, + { + "epoch": 0.3594715245524643, + "grad_norm": 1.5099581142957212, + "learning_rate": 2.9654009914968457e-06, + "loss": 1.0591, + "step": 3986 + }, + { + "epoch": 0.3595617080759345, + "grad_norm": 2.112239785504597, + "learning_rate": 2.9648893071237956e-06, + "loss": 0.893, + "step": 3987 + }, + { + "epoch": 0.3596518915994048, + "grad_norm": 2.633452929022317, + "learning_rate": 2.964377540421069e-06, + "loss": 0.9775, + "step": 3988 + }, + { + "epoch": 0.35974207512287504, + "grad_norm": 1.8171904153054257, + "learning_rate": 2.963865691432334e-06, + "loss": 1.0247, + "step": 3989 + }, + { + "epoch": 0.3598322586463453, + "grad_norm": 3.27519845142859, + "learning_rate": 2.963353760201263e-06, + "loss": 0.9161, + "step": 3990 + }, + { + "epoch": 0.35992244216981556, + "grad_norm": 1.872287250141614, + "learning_rate": 2.962841746771537e-06, + "loss": 0.9901, + "step": 3991 + }, + { + "epoch": 0.36001262569328585, + "grad_norm": 1.6083251343486833, + "learning_rate": 2.9623296511868445e-06, + "loss": 0.9993, + "step": 3992 + }, + { + "epoch": 0.3601028092167561, + "grad_norm": 1.6842679048084501, + "learning_rate": 2.96181747349088e-06, + "loss": 0.93, + "step": 3993 + }, + { + "epoch": 0.36019299274022637, + "grad_norm": 1.584750327773872, + "learning_rate": 2.961305213727345e-06, + "loss": 1.0202, + "step": 3994 + }, + { + "epoch": 0.3602831762636966, + "grad_norm": 0.6567374417324444, + "learning_rate": 2.960792871939949e-06, + "loss": 0.8397, + "step": 3995 + }, + { + "epoch": 0.3603733597871669, + "grad_norm": 1.6678254470194007, + "learning_rate": 2.9602804481724064e-06, + "loss": 1.0935, + "step": 3996 + }, + { + "epoch": 0.3604635433106371, + "grad_norm": 1.6148265614399908, + "learning_rate": 2.9597679424684427e-06, + "loss": 0.9982, + "step": 3997 + }, + { + "epoch": 0.3605537268341074, + "grad_norm": 1.816704772701221, + "learning_rate": 2.9592553548717848e-06, + "loss": 1.0161, + "step": 3998 + }, + { + "epoch": 0.36064391035757765, + "grad_norm": 1.5645663521662514, + "learning_rate": 2.958742685426171e-06, + "loss": 1.085, + "step": 3999 + }, + { + "epoch": 0.36073409388104793, + "grad_norm": 2.3592462567704646, + "learning_rate": 2.9582299341753446e-06, + "loss": 1.0685, + "step": 4000 + }, + { + "epoch": 0.36082427740451817, + "grad_norm": 1.800421490594001, + "learning_rate": 2.957717101163057e-06, + "loss": 0.9441, + "step": 4001 + }, + { + "epoch": 0.36091446092798846, + "grad_norm": 4.837732724449337, + "learning_rate": 2.9572041864330655e-06, + "loss": 1.0001, + "step": 4002 + }, + { + "epoch": 0.36100464445145874, + "grad_norm": 1.7711541352749312, + "learning_rate": 2.9566911900291346e-06, + "loss": 1.0257, + "step": 4003 + }, + { + "epoch": 0.361094827974929, + "grad_norm": 1.7989975044949893, + "learning_rate": 2.9561781119950368e-06, + "loss": 0.9192, + "step": 4004 + }, + { + "epoch": 0.36118501149839927, + "grad_norm": 1.6011637121412672, + "learning_rate": 2.9556649523745493e-06, + "loss": 0.9456, + "step": 4005 + }, + { + "epoch": 0.3612751950218695, + "grad_norm": 2.073571630421297, + "learning_rate": 2.955151711211459e-06, + "loss": 1.065, + "step": 4006 + }, + { + "epoch": 0.3613653785453398, + "grad_norm": 1.54159770845719, + "learning_rate": 2.9546383885495583e-06, + "loss": 0.9723, + "step": 4007 + }, + { + "epoch": 0.36145556206881, + "grad_norm": 1.6973533476008873, + "learning_rate": 2.9541249844326464e-06, + "loss": 0.9575, + "step": 4008 + }, + { + "epoch": 0.3615457455922803, + "grad_norm": 1.9784233772184698, + "learning_rate": 2.9536114989045295e-06, + "loss": 0.983, + "step": 4009 + }, + { + "epoch": 0.36163592911575054, + "grad_norm": 1.7606890390576329, + "learning_rate": 2.9530979320090216e-06, + "loss": 1.0616, + "step": 4010 + }, + { + "epoch": 0.36172611263922083, + "grad_norm": 1.979797731144244, + "learning_rate": 2.9525842837899422e-06, + "loss": 1.0838, + "step": 4011 + }, + { + "epoch": 0.36181629616269106, + "grad_norm": 1.935447467056545, + "learning_rate": 2.95207055429112e-06, + "loss": 1.0838, + "step": 4012 + }, + { + "epoch": 0.36190647968616135, + "grad_norm": 1.5656652814666638, + "learning_rate": 2.951556743556388e-06, + "loss": 0.9937, + "step": 4013 + }, + { + "epoch": 0.3619966632096316, + "grad_norm": 1.7902961708638574, + "learning_rate": 2.951042851629588e-06, + "loss": 1.0491, + "step": 4014 + }, + { + "epoch": 0.3620868467331019, + "grad_norm": 1.8438915909532838, + "learning_rate": 2.950528878554568e-06, + "loss": 0.9588, + "step": 4015 + }, + { + "epoch": 0.3621770302565721, + "grad_norm": 1.563770083169632, + "learning_rate": 2.950014824375183e-06, + "loss": 1.0435, + "step": 4016 + }, + { + "epoch": 0.3622672137800424, + "grad_norm": 0.6302671457596632, + "learning_rate": 2.949500689135295e-06, + "loss": 0.7647, + "step": 4017 + }, + { + "epoch": 0.36235739730351263, + "grad_norm": 1.8393823000432183, + "learning_rate": 2.9489864728787722e-06, + "loss": 1.0843, + "step": 4018 + }, + { + "epoch": 0.3624475808269829, + "grad_norm": 1.5253736663539939, + "learning_rate": 2.9484721756494915e-06, + "loss": 1.0646, + "step": 4019 + }, + { + "epoch": 0.36253776435045315, + "grad_norm": 1.9758069550700046, + "learning_rate": 2.9479577974913343e-06, + "loss": 1.0048, + "step": 4020 + }, + { + "epoch": 0.36262794787392344, + "grad_norm": 1.7815069046807455, + "learning_rate": 2.9474433384481908e-06, + "loss": 1.0009, + "step": 4021 + }, + { + "epoch": 0.3627181313973937, + "grad_norm": 1.5713132639479825, + "learning_rate": 2.9469287985639577e-06, + "loss": 1.0009, + "step": 4022 + }, + { + "epoch": 0.36280831492086396, + "grad_norm": 1.6812445232770128, + "learning_rate": 2.9464141778825384e-06, + "loss": 1.0501, + "step": 4023 + }, + { + "epoch": 0.3628984984443342, + "grad_norm": 1.6518356949008604, + "learning_rate": 2.9458994764478427e-06, + "loss": 1.0427, + "step": 4024 + }, + { + "epoch": 0.3629886819678045, + "grad_norm": 2.582245556685257, + "learning_rate": 2.9453846943037883e-06, + "loss": 1.0415, + "step": 4025 + }, + { + "epoch": 0.36307886549127477, + "grad_norm": 1.5361359166370416, + "learning_rate": 2.9448698314942987e-06, + "loss": 0.9511, + "step": 4026 + }, + { + "epoch": 0.363169049014745, + "grad_norm": 0.9150704263464083, + "learning_rate": 2.944354888063305e-06, + "loss": 0.8611, + "step": 4027 + }, + { + "epoch": 0.3632592325382153, + "grad_norm": 1.8047402114105988, + "learning_rate": 2.9438398640547453e-06, + "loss": 0.981, + "step": 4028 + }, + { + "epoch": 0.3633494160616855, + "grad_norm": 1.6081189930192443, + "learning_rate": 2.943324759512564e-06, + "loss": 0.9462, + "step": 4029 + }, + { + "epoch": 0.3634395995851558, + "grad_norm": 1.8480729969849834, + "learning_rate": 2.9428095744807134e-06, + "loss": 1.0305, + "step": 4030 + }, + { + "epoch": 0.36352978310862605, + "grad_norm": 1.6466572626645883, + "learning_rate": 2.942294309003151e-06, + "loss": 0.9587, + "step": 4031 + }, + { + "epoch": 0.36361996663209634, + "grad_norm": 1.8172557526068505, + "learning_rate": 2.941778963123843e-06, + "loss": 1.0204, + "step": 4032 + }, + { + "epoch": 0.36371015015556657, + "grad_norm": 2.3416516892937276, + "learning_rate": 2.94126353688676e-06, + "loss": 1.0047, + "step": 4033 + }, + { + "epoch": 0.36380033367903686, + "grad_norm": 2.4115710194392608, + "learning_rate": 2.9407480303358825e-06, + "loss": 0.9505, + "step": 4034 + }, + { + "epoch": 0.3638905172025071, + "grad_norm": 1.7063912385605637, + "learning_rate": 2.940232443515195e-06, + "loss": 1.0071, + "step": 4035 + }, + { + "epoch": 0.3639807007259774, + "grad_norm": 1.6633268350972439, + "learning_rate": 2.9397167764686916e-06, + "loss": 1.0284, + "step": 4036 + }, + { + "epoch": 0.3640708842494476, + "grad_norm": 1.4113121050810495, + "learning_rate": 2.9392010292403714e-06, + "loss": 0.9799, + "step": 4037 + }, + { + "epoch": 0.3641610677729179, + "grad_norm": 2.0826456333633625, + "learning_rate": 2.9386852018742404e-06, + "loss": 0.9921, + "step": 4038 + }, + { + "epoch": 0.36425125129638813, + "grad_norm": 1.5294685121296048, + "learning_rate": 2.938169294414312e-06, + "loss": 0.9772, + "step": 4039 + }, + { + "epoch": 0.3643414348198584, + "grad_norm": 1.636491782536375, + "learning_rate": 2.9376533069046067e-06, + "loss": 1.0387, + "step": 4040 + }, + { + "epoch": 0.36443161834332866, + "grad_norm": 0.7198815767597512, + "learning_rate": 2.9371372393891514e-06, + "loss": 0.8719, + "step": 4041 + }, + { + "epoch": 0.36452180186679894, + "grad_norm": 1.5063012015476638, + "learning_rate": 2.936621091911979e-06, + "loss": 1.076, + "step": 4042 + }, + { + "epoch": 0.3646119853902692, + "grad_norm": 1.771155952535836, + "learning_rate": 2.936104864517131e-06, + "loss": 0.9341, + "step": 4043 + }, + { + "epoch": 0.36470216891373947, + "grad_norm": 2.0301246460103437, + "learning_rate": 2.9355885572486535e-06, + "loss": 0.9641, + "step": 4044 + }, + { + "epoch": 0.3647923524372097, + "grad_norm": 2.0139629993468193, + "learning_rate": 2.9350721701506026e-06, + "loss": 1.0328, + "step": 4045 + }, + { + "epoch": 0.36488253596068, + "grad_norm": 1.6398189199311566, + "learning_rate": 2.9345557032670375e-06, + "loss": 0.9961, + "step": 4046 + }, + { + "epoch": 0.3649727194841502, + "grad_norm": 2.4215256962063596, + "learning_rate": 2.934039156642027e-06, + "loss": 1.0475, + "step": 4047 + }, + { + "epoch": 0.3650629030076205, + "grad_norm": 1.9197058493078625, + "learning_rate": 2.9335225303196454e-06, + "loss": 0.8621, + "step": 4048 + }, + { + "epoch": 0.36515308653109074, + "grad_norm": 1.8655355000713245, + "learning_rate": 2.933005824343974e-06, + "loss": 0.971, + "step": 4049 + }, + { + "epoch": 0.36524327005456103, + "grad_norm": 2.0830159390591008, + "learning_rate": 2.932489038759101e-06, + "loss": 0.9756, + "step": 4050 + }, + { + "epoch": 0.3653334535780313, + "grad_norm": 1.4545943512067205, + "learning_rate": 2.9319721736091215e-06, + "loss": 0.8997, + "step": 4051 + }, + { + "epoch": 0.36542363710150155, + "grad_norm": 1.3831634507176156, + "learning_rate": 2.9314552289381377e-06, + "loss": 0.9065, + "step": 4052 + }, + { + "epoch": 0.36551382062497184, + "grad_norm": 1.7945673180402766, + "learning_rate": 2.9309382047902574e-06, + "loss": 0.9708, + "step": 4053 + }, + { + "epoch": 0.3656040041484421, + "grad_norm": 1.4694909195179726, + "learning_rate": 2.9304211012095963e-06, + "loss": 1.047, + "step": 4054 + }, + { + "epoch": 0.36569418767191236, + "grad_norm": 1.8533062670924536, + "learning_rate": 2.929903918240277e-06, + "loss": 1.0582, + "step": 4055 + }, + { + "epoch": 0.3657843711953826, + "grad_norm": 0.6353451709443413, + "learning_rate": 2.9293866559264273e-06, + "loss": 0.8155, + "step": 4056 + }, + { + "epoch": 0.3658745547188529, + "grad_norm": 1.9202380148065763, + "learning_rate": 2.928869314312184e-06, + "loss": 1.0047, + "step": 4057 + }, + { + "epoch": 0.3659647382423231, + "grad_norm": 2.247517593875036, + "learning_rate": 2.9283518934416892e-06, + "loss": 1.007, + "step": 4058 + }, + { + "epoch": 0.3660549217657934, + "grad_norm": 2.124203701043144, + "learning_rate": 2.927834393359092e-06, + "loss": 1.0117, + "step": 4059 + }, + { + "epoch": 0.36614510528926364, + "grad_norm": 1.7190350382875002, + "learning_rate": 2.927316814108548e-06, + "loss": 1.0714, + "step": 4060 + }, + { + "epoch": 0.36623528881273393, + "grad_norm": 2.172899512286941, + "learning_rate": 2.92679915573422e-06, + "loss": 0.9125, + "step": 4061 + }, + { + "epoch": 0.36632547233620416, + "grad_norm": 1.5873083194143343, + "learning_rate": 2.926281418280278e-06, + "loss": 0.9417, + "step": 4062 + }, + { + "epoch": 0.36641565585967445, + "grad_norm": 1.5920409220271499, + "learning_rate": 2.925763601790899e-06, + "loss": 1.0052, + "step": 4063 + }, + { + "epoch": 0.3665058393831447, + "grad_norm": 2.1098578006502815, + "learning_rate": 2.9252457063102635e-06, + "loss": 0.9489, + "step": 4064 + }, + { + "epoch": 0.36659602290661497, + "grad_norm": 1.767227390962235, + "learning_rate": 2.9247277318825626e-06, + "loss": 0.9633, + "step": 4065 + }, + { + "epoch": 0.3666862064300852, + "grad_norm": 1.4698211335333626, + "learning_rate": 2.924209678551993e-06, + "loss": 1.0956, + "step": 4066 + }, + { + "epoch": 0.3667763899535555, + "grad_norm": 1.7344139713342228, + "learning_rate": 2.923691546362757e-06, + "loss": 1.0102, + "step": 4067 + }, + { + "epoch": 0.3668665734770257, + "grad_norm": 1.6690932331637565, + "learning_rate": 2.9231733353590663e-06, + "loss": 0.9598, + "step": 4068 + }, + { + "epoch": 0.366956757000496, + "grad_norm": 2.026052662902123, + "learning_rate": 2.922655045585136e-06, + "loss": 1.0719, + "step": 4069 + }, + { + "epoch": 0.36704694052396625, + "grad_norm": 0.6862716104408801, + "learning_rate": 2.92213667708519e-06, + "loss": 0.8569, + "step": 4070 + }, + { + "epoch": 0.36713712404743654, + "grad_norm": 1.5204405429784704, + "learning_rate": 2.921618229903457e-06, + "loss": 1.0084, + "step": 4071 + }, + { + "epoch": 0.36722730757090677, + "grad_norm": 2.295624476888936, + "learning_rate": 2.9210997040841752e-06, + "loss": 0.925, + "step": 4072 + }, + { + "epoch": 0.36731749109437706, + "grad_norm": 2.322773459999548, + "learning_rate": 2.9205810996715885e-06, + "loss": 0.9696, + "step": 4073 + }, + { + "epoch": 0.36740767461784735, + "grad_norm": 1.634696567038431, + "learning_rate": 2.9200624167099456e-06, + "loss": 0.9681, + "step": 4074 + }, + { + "epoch": 0.3674978581413176, + "grad_norm": 1.635707861696742, + "learning_rate": 2.919543655243505e-06, + "loss": 0.9152, + "step": 4075 + }, + { + "epoch": 0.36758804166478787, + "grad_norm": 1.5889192559749714, + "learning_rate": 2.919024815316529e-06, + "loss": 1.0266, + "step": 4076 + }, + { + "epoch": 0.3676782251882581, + "grad_norm": 1.8857969950727376, + "learning_rate": 2.9185058969732877e-06, + "loss": 1.1282, + "step": 4077 + }, + { + "epoch": 0.3677684087117284, + "grad_norm": 1.9282243254812033, + "learning_rate": 2.917986900258059e-06, + "loss": 1.0333, + "step": 4078 + }, + { + "epoch": 0.3678585922351986, + "grad_norm": 1.4592554537638704, + "learning_rate": 2.917467825215126e-06, + "loss": 0.996, + "step": 4079 + }, + { + "epoch": 0.3679487757586689, + "grad_norm": 2.0899765400277133, + "learning_rate": 2.9169486718887803e-06, + "loss": 0.9481, + "step": 4080 + }, + { + "epoch": 0.36803895928213914, + "grad_norm": 1.5284128179372587, + "learning_rate": 2.9164294403233173e-06, + "loss": 1.0528, + "step": 4081 + }, + { + "epoch": 0.36812914280560943, + "grad_norm": 1.7853511645461966, + "learning_rate": 2.915910130563041e-06, + "loss": 0.9321, + "step": 4082 + }, + { + "epoch": 0.36821932632907967, + "grad_norm": 1.91877533836906, + "learning_rate": 2.915390742652262e-06, + "loss": 0.9462, + "step": 4083 + }, + { + "epoch": 0.36830950985254995, + "grad_norm": 2.0747728200435005, + "learning_rate": 2.914871276635298e-06, + "loss": 0.9393, + "step": 4084 + }, + { + "epoch": 0.3683996933760202, + "grad_norm": 2.5160650970091565, + "learning_rate": 2.914351732556472e-06, + "loss": 0.9811, + "step": 4085 + }, + { + "epoch": 0.3684898768994905, + "grad_norm": 1.7885897482391269, + "learning_rate": 2.9138321104601144e-06, + "loss": 0.9242, + "step": 4086 + }, + { + "epoch": 0.3685800604229607, + "grad_norm": 6.6453384007842295, + "learning_rate": 2.9133124103905623e-06, + "loss": 0.9355, + "step": 4087 + }, + { + "epoch": 0.368670243946431, + "grad_norm": 1.9350818186965761, + "learning_rate": 2.9127926323921596e-06, + "loss": 0.9867, + "step": 4088 + }, + { + "epoch": 0.36876042746990123, + "grad_norm": 1.645884629749917, + "learning_rate": 2.912272776509256e-06, + "loss": 1.0179, + "step": 4089 + }, + { + "epoch": 0.3688506109933715, + "grad_norm": 2.136621671769227, + "learning_rate": 2.911752842786209e-06, + "loss": 0.9692, + "step": 4090 + }, + { + "epoch": 0.36894079451684175, + "grad_norm": 2.279695477644915, + "learning_rate": 2.911232831267383e-06, + "loss": 0.9434, + "step": 4091 + }, + { + "epoch": 0.36903097804031204, + "grad_norm": 2.48220298396303, + "learning_rate": 2.910712741997146e-06, + "loss": 1.0216, + "step": 4092 + }, + { + "epoch": 0.3691211615637823, + "grad_norm": 1.6638179115748206, + "learning_rate": 2.910192575019877e-06, + "loss": 0.9908, + "step": 4093 + }, + { + "epoch": 0.36921134508725256, + "grad_norm": 2.696562015764611, + "learning_rate": 2.9096723303799583e-06, + "loss": 1.0294, + "step": 4094 + }, + { + "epoch": 0.3693015286107228, + "grad_norm": 2.0118424757159112, + "learning_rate": 2.9091520081217805e-06, + "loss": 1.0078, + "step": 4095 + }, + { + "epoch": 0.3693917121341931, + "grad_norm": 1.6622224045366356, + "learning_rate": 2.908631608289741e-06, + "loss": 0.9442, + "step": 4096 + }, + { + "epoch": 0.3694818956576633, + "grad_norm": 1.9713847377595375, + "learning_rate": 2.9081111309282423e-06, + "loss": 0.9848, + "step": 4097 + }, + { + "epoch": 0.3695720791811336, + "grad_norm": 0.6439476607849419, + "learning_rate": 2.9075905760816942e-06, + "loss": 0.8687, + "step": 4098 + }, + { + "epoch": 0.3696622627046039, + "grad_norm": 1.9960785210134875, + "learning_rate": 2.907069943794514e-06, + "loss": 1.0395, + "step": 4099 + }, + { + "epoch": 0.3697524462280741, + "grad_norm": 1.5182251796593724, + "learning_rate": 2.906549234111125e-06, + "loss": 0.9711, + "step": 4100 + }, + { + "epoch": 0.3698426297515444, + "grad_norm": 1.592560642231617, + "learning_rate": 2.906028447075956e-06, + "loss": 1.055, + "step": 4101 + }, + { + "epoch": 0.36993281327501465, + "grad_norm": 1.647010004347878, + "learning_rate": 2.905507582733445e-06, + "loss": 1.0803, + "step": 4102 + }, + { + "epoch": 0.37002299679848494, + "grad_norm": 1.568638302903411, + "learning_rate": 2.904986641128033e-06, + "loss": 1.1151, + "step": 4103 + }, + { + "epoch": 0.37011318032195517, + "grad_norm": 1.5798912362194688, + "learning_rate": 2.9044656223041716e-06, + "loss": 0.9882, + "step": 4104 + }, + { + "epoch": 0.37020336384542546, + "grad_norm": 1.5515595988297397, + "learning_rate": 2.9039445263063157e-06, + "loss": 0.9562, + "step": 4105 + }, + { + "epoch": 0.3702935473688957, + "grad_norm": 2.180966565973846, + "learning_rate": 2.903423353178929e-06, + "loss": 0.8937, + "step": 4106 + }, + { + "epoch": 0.370383730892366, + "grad_norm": 1.7570376574584772, + "learning_rate": 2.9029021029664802e-06, + "loss": 0.9928, + "step": 4107 + }, + { + "epoch": 0.3704739144158362, + "grad_norm": 1.5565782867164124, + "learning_rate": 2.9023807757134455e-06, + "loss": 1.0523, + "step": 4108 + }, + { + "epoch": 0.3705640979393065, + "grad_norm": 2.008143416271896, + "learning_rate": 2.901859371464307e-06, + "loss": 1.0648, + "step": 4109 + }, + { + "epoch": 0.37065428146277674, + "grad_norm": 1.7799713497856473, + "learning_rate": 2.9013378902635535e-06, + "loss": 1.0361, + "step": 4110 + }, + { + "epoch": 0.370744464986247, + "grad_norm": 1.808817048428353, + "learning_rate": 2.9008163321556823e-06, + "loss": 0.967, + "step": 4111 + }, + { + "epoch": 0.37083464850971726, + "grad_norm": 2.2301953042897114, + "learning_rate": 2.900294697185194e-06, + "loss": 1.0136, + "step": 4112 + }, + { + "epoch": 0.37092483203318755, + "grad_norm": 2.502148372705828, + "learning_rate": 2.899772985396599e-06, + "loss": 1.0153, + "step": 4113 + }, + { + "epoch": 0.3710150155566578, + "grad_norm": 1.3807764095866042, + "learning_rate": 2.8992511968344104e-06, + "loss": 0.9883, + "step": 4114 + }, + { + "epoch": 0.37110519908012807, + "grad_norm": 1.617648146623868, + "learning_rate": 2.8987293315431523e-06, + "loss": 0.9447, + "step": 4115 + }, + { + "epoch": 0.3711953826035983, + "grad_norm": 2.2673868629300435, + "learning_rate": 2.898207389567351e-06, + "loss": 0.9304, + "step": 4116 + }, + { + "epoch": 0.3712855661270686, + "grad_norm": 1.5339454748915249, + "learning_rate": 2.897685370951543e-06, + "loss": 0.9914, + "step": 4117 + }, + { + "epoch": 0.3713757496505388, + "grad_norm": 1.868634353682673, + "learning_rate": 2.89716327574027e-06, + "loss": 0.8883, + "step": 4118 + }, + { + "epoch": 0.3714659331740091, + "grad_norm": 1.5106061443877137, + "learning_rate": 2.8966411039780787e-06, + "loss": 1.0865, + "step": 4119 + }, + { + "epoch": 0.37155611669747934, + "grad_norm": 1.9124251855914833, + "learning_rate": 2.8961188557095248e-06, + "loss": 1.0053, + "step": 4120 + }, + { + "epoch": 0.37164630022094963, + "grad_norm": 1.9679647423679545, + "learning_rate": 2.895596530979168e-06, + "loss": 0.9447, + "step": 4121 + }, + { + "epoch": 0.3717364837444199, + "grad_norm": 1.5609777283090804, + "learning_rate": 2.895074129831578e-06, + "loss": 1.0139, + "step": 4122 + }, + { + "epoch": 0.37182666726789015, + "grad_norm": 1.8168972869335458, + "learning_rate": 2.8945516523113275e-06, + "loss": 0.9118, + "step": 4123 + }, + { + "epoch": 0.37191685079136044, + "grad_norm": 0.6793941215414107, + "learning_rate": 2.894029098462998e-06, + "loss": 0.7894, + "step": 4124 + }, + { + "epoch": 0.3720070343148307, + "grad_norm": 1.6191524049631791, + "learning_rate": 2.8935064683311756e-06, + "loss": 1.0207, + "step": 4125 + }, + { + "epoch": 0.37209721783830096, + "grad_norm": 0.6774219815352344, + "learning_rate": 2.8929837619604544e-06, + "loss": 0.893, + "step": 4126 + }, + { + "epoch": 0.3721874013617712, + "grad_norm": 2.0370874888340724, + "learning_rate": 2.8924609793954346e-06, + "loss": 1.043, + "step": 4127 + }, + { + "epoch": 0.3722775848852415, + "grad_norm": 1.6951515367449266, + "learning_rate": 2.891938120680724e-06, + "loss": 0.9552, + "step": 4128 + }, + { + "epoch": 0.3723677684087117, + "grad_norm": 1.6733516696036745, + "learning_rate": 2.8914151858609343e-06, + "loss": 0.9908, + "step": 4129 + }, + { + "epoch": 0.372457951932182, + "grad_norm": 1.6413036577644513, + "learning_rate": 2.8908921749806858e-06, + "loss": 1.0197, + "step": 4130 + }, + { + "epoch": 0.37254813545565224, + "grad_norm": 1.5829505039316103, + "learning_rate": 2.890369088084605e-06, + "loss": 1.0115, + "step": 4131 + }, + { + "epoch": 0.37263831897912253, + "grad_norm": 1.5372880494946743, + "learning_rate": 2.889845925217323e-06, + "loss": 1.0239, + "step": 4132 + }, + { + "epoch": 0.37272850250259276, + "grad_norm": 1.6461795330100308, + "learning_rate": 2.8893226864234813e-06, + "loss": 1.0119, + "step": 4133 + }, + { + "epoch": 0.37281868602606305, + "grad_norm": 1.949153432726393, + "learning_rate": 2.8887993717477236e-06, + "loss": 1.0134, + "step": 4134 + }, + { + "epoch": 0.3729088695495333, + "grad_norm": 7.614654025799018, + "learning_rate": 2.8882759812347035e-06, + "loss": 0.8895, + "step": 4135 + }, + { + "epoch": 0.3729990530730036, + "grad_norm": 1.6982906920321221, + "learning_rate": 2.887752514929078e-06, + "loss": 1.0007, + "step": 4136 + }, + { + "epoch": 0.3730892365964738, + "grad_norm": 1.6674444370649513, + "learning_rate": 2.887228972875513e-06, + "loss": 0.9522, + "step": 4137 + }, + { + "epoch": 0.3731794201199441, + "grad_norm": 3.5844827331476155, + "learning_rate": 2.88670535511868e-06, + "loss": 1.0022, + "step": 4138 + }, + { + "epoch": 0.3732696036434143, + "grad_norm": 2.64727496623541, + "learning_rate": 2.886181661703257e-06, + "loss": 0.9485, + "step": 4139 + }, + { + "epoch": 0.3733597871668846, + "grad_norm": 1.9184539243421013, + "learning_rate": 2.8856578926739285e-06, + "loss": 0.9683, + "step": 4140 + }, + { + "epoch": 0.37344997069035485, + "grad_norm": 1.7628592375301027, + "learning_rate": 2.8851340480753846e-06, + "loss": 0.9577, + "step": 4141 + }, + { + "epoch": 0.37354015421382514, + "grad_norm": 1.9836664211307442, + "learning_rate": 2.8846101279523232e-06, + "loss": 1.0026, + "step": 4142 + }, + { + "epoch": 0.37363033773729537, + "grad_norm": 1.744209313828873, + "learning_rate": 2.8840861323494487e-06, + "loss": 0.9751, + "step": 4143 + }, + { + "epoch": 0.37372052126076566, + "grad_norm": 2.7639739717287575, + "learning_rate": 2.88356206131147e-06, + "loss": 1.0576, + "step": 4144 + }, + { + "epoch": 0.37381070478423595, + "grad_norm": 2.0689432474560165, + "learning_rate": 2.883037914883104e-06, + "loss": 0.9847, + "step": 4145 + }, + { + "epoch": 0.3739008883077062, + "grad_norm": 2.2276685677911794, + "learning_rate": 2.882513693109075e-06, + "loss": 0.9692, + "step": 4146 + }, + { + "epoch": 0.37399107183117647, + "grad_norm": 1.366218703209032, + "learning_rate": 2.8819893960341106e-06, + "loss": 0.9782, + "step": 4147 + }, + { + "epoch": 0.3740812553546467, + "grad_norm": 2.256460662897886, + "learning_rate": 2.881465023702948e-06, + "loss": 1.0586, + "step": 4148 + }, + { + "epoch": 0.374171438878117, + "grad_norm": 12.13784581486314, + "learning_rate": 2.8809405761603294e-06, + "loss": 1.0809, + "step": 4149 + }, + { + "epoch": 0.3742616224015872, + "grad_norm": 3.4423368237176275, + "learning_rate": 2.880416053451003e-06, + "loss": 0.9223, + "step": 4150 + }, + { + "epoch": 0.3743518059250575, + "grad_norm": 1.6791473939734611, + "learning_rate": 2.879891455619725e-06, + "loss": 1.0895, + "step": 4151 + }, + { + "epoch": 0.37444198944852775, + "grad_norm": 1.464334221279038, + "learning_rate": 2.879366782711256e-06, + "loss": 1.0022, + "step": 4152 + }, + { + "epoch": 0.37453217297199803, + "grad_norm": 1.744081707650248, + "learning_rate": 2.8788420347703643e-06, + "loss": 1.0205, + "step": 4153 + }, + { + "epoch": 0.37462235649546827, + "grad_norm": 0.6352549852718306, + "learning_rate": 2.8783172118418244e-06, + "loss": 0.8144, + "step": 4154 + }, + { + "epoch": 0.37471254001893856, + "grad_norm": 1.5352827413149082, + "learning_rate": 2.877792313970417e-06, + "loss": 1.0101, + "step": 4155 + }, + { + "epoch": 0.3748027235424088, + "grad_norm": 1.9477069626260421, + "learning_rate": 2.8772673412009293e-06, + "loss": 1.0582, + "step": 4156 + }, + { + "epoch": 0.3748929070658791, + "grad_norm": 2.221979396851591, + "learning_rate": 2.8767422935781545e-06, + "loss": 1.0468, + "step": 4157 + }, + { + "epoch": 0.3749830905893493, + "grad_norm": 1.4545942692530867, + "learning_rate": 2.8762171711468935e-06, + "loss": 1.0395, + "step": 4158 + }, + { + "epoch": 0.3750732741128196, + "grad_norm": 1.6179075250687451, + "learning_rate": 2.875691973951952e-06, + "loss": 1.0042, + "step": 4159 + }, + { + "epoch": 0.37516345763628983, + "grad_norm": 0.6420455627257179, + "learning_rate": 2.8751667020381425e-06, + "loss": 0.8849, + "step": 4160 + }, + { + "epoch": 0.3752536411597601, + "grad_norm": 1.4544522366672403, + "learning_rate": 2.8746413554502837e-06, + "loss": 1.065, + "step": 4161 + }, + { + "epoch": 0.37534382468323035, + "grad_norm": 1.5518460017616418, + "learning_rate": 2.8741159342332027e-06, + "loss": 0.9362, + "step": 4162 + }, + { + "epoch": 0.37543400820670064, + "grad_norm": 7.077688713185396, + "learning_rate": 2.87359043843173e-06, + "loss": 1.0599, + "step": 4163 + }, + { + "epoch": 0.3755241917301709, + "grad_norm": 1.7460682160902186, + "learning_rate": 2.873064868090704e-06, + "loss": 1.0413, + "step": 4164 + }, + { + "epoch": 0.37561437525364116, + "grad_norm": 1.6401913024232637, + "learning_rate": 2.8725392232549697e-06, + "loss": 1.0934, + "step": 4165 + }, + { + "epoch": 0.3757045587771114, + "grad_norm": 1.9741035817303905, + "learning_rate": 2.872013503969378e-06, + "loss": 1.078, + "step": 4166 + }, + { + "epoch": 0.3757947423005817, + "grad_norm": 1.5867708018267501, + "learning_rate": 2.8714877102787853e-06, + "loss": 1.0143, + "step": 4167 + }, + { + "epoch": 0.3758849258240519, + "grad_norm": 1.8130145164788092, + "learning_rate": 2.8709618422280564e-06, + "loss": 1.0085, + "step": 4168 + }, + { + "epoch": 0.3759751093475222, + "grad_norm": 1.8567749565545981, + "learning_rate": 2.8704358998620605e-06, + "loss": 0.9801, + "step": 4169 + }, + { + "epoch": 0.3760652928709925, + "grad_norm": 1.8663195264555053, + "learning_rate": 2.8699098832256735e-06, + "loss": 1.0315, + "step": 4170 + }, + { + "epoch": 0.37615547639446273, + "grad_norm": 1.9574814086459933, + "learning_rate": 2.86938379236378e-06, + "loss": 0.9166, + "step": 4171 + }, + { + "epoch": 0.376245659917933, + "grad_norm": 1.6017963517818947, + "learning_rate": 2.868857627321266e-06, + "loss": 0.9327, + "step": 4172 + }, + { + "epoch": 0.37633584344140325, + "grad_norm": 1.5035576750815889, + "learning_rate": 2.8683313881430296e-06, + "loss": 1.0084, + "step": 4173 + }, + { + "epoch": 0.37642602696487354, + "grad_norm": 1.5595500564931815, + "learning_rate": 2.8678050748739706e-06, + "loss": 0.9962, + "step": 4174 + }, + { + "epoch": 0.37651621048834377, + "grad_norm": 1.6228918926550102, + "learning_rate": 2.8672786875589976e-06, + "loss": 0.9762, + "step": 4175 + }, + { + "epoch": 0.37660639401181406, + "grad_norm": 1.5411791503165235, + "learning_rate": 2.866752226243025e-06, + "loss": 0.9565, + "step": 4176 + }, + { + "epoch": 0.3766965775352843, + "grad_norm": 1.6053869964500682, + "learning_rate": 2.8662256909709733e-06, + "loss": 1.0494, + "step": 4177 + }, + { + "epoch": 0.3767867610587546, + "grad_norm": 2.23440551736972, + "learning_rate": 2.865699081787769e-06, + "loss": 1.0397, + "step": 4178 + }, + { + "epoch": 0.3768769445822248, + "grad_norm": 1.5802324783133517, + "learning_rate": 2.8651723987383465e-06, + "loss": 0.992, + "step": 4179 + }, + { + "epoch": 0.3769671281056951, + "grad_norm": 2.036435126313048, + "learning_rate": 2.8646456418676437e-06, + "loss": 0.9781, + "step": 4180 + }, + { + "epoch": 0.37705731162916534, + "grad_norm": 0.6532143362903822, + "learning_rate": 2.8641188112206067e-06, + "loss": 0.8338, + "step": 4181 + }, + { + "epoch": 0.3771474951526356, + "grad_norm": 2.321054462009195, + "learning_rate": 2.863591906842189e-06, + "loss": 1.0349, + "step": 4182 + }, + { + "epoch": 0.37723767867610586, + "grad_norm": 1.5531054551182342, + "learning_rate": 2.863064928777347e-06, + "loss": 1.0345, + "step": 4183 + }, + { + "epoch": 0.37732786219957615, + "grad_norm": 1.7595517518610861, + "learning_rate": 2.862537877071047e-06, + "loss": 0.978, + "step": 4184 + }, + { + "epoch": 0.3774180457230464, + "grad_norm": 0.64165497765071, + "learning_rate": 2.8620107517682597e-06, + "loss": 0.7975, + "step": 4185 + }, + { + "epoch": 0.37750822924651667, + "grad_norm": 2.041038987605524, + "learning_rate": 2.8614835529139618e-06, + "loss": 0.9392, + "step": 4186 + }, + { + "epoch": 0.3775984127699869, + "grad_norm": 1.698082554988393, + "learning_rate": 2.8609562805531367e-06, + "loss": 0.9648, + "step": 4187 + }, + { + "epoch": 0.3776885962934572, + "grad_norm": 1.695793751560077, + "learning_rate": 2.8604289347307746e-06, + "loss": 1.0386, + "step": 4188 + }, + { + "epoch": 0.3777787798169274, + "grad_norm": 1.593465293436898, + "learning_rate": 2.859901515491871e-06, + "loss": 0.9993, + "step": 4189 + }, + { + "epoch": 0.3778689633403977, + "grad_norm": 1.5136071369724688, + "learning_rate": 2.8593740228814298e-06, + "loss": 0.8908, + "step": 4190 + }, + { + "epoch": 0.37795914686386795, + "grad_norm": 1.9995073069243148, + "learning_rate": 2.8588464569444574e-06, + "loss": 0.9886, + "step": 4191 + }, + { + "epoch": 0.37804933038733823, + "grad_norm": 1.517451179133842, + "learning_rate": 2.8583188177259697e-06, + "loss": 0.9942, + "step": 4192 + }, + { + "epoch": 0.3781395139108085, + "grad_norm": 1.7167246589617975, + "learning_rate": 2.857791105270988e-06, + "loss": 0.9889, + "step": 4193 + }, + { + "epoch": 0.37822969743427876, + "grad_norm": 1.5706464877254516, + "learning_rate": 2.857263319624539e-06, + "loss": 1.0312, + "step": 4194 + }, + { + "epoch": 0.37831988095774904, + "grad_norm": 1.4982351411095056, + "learning_rate": 2.856735460831657e-06, + "loss": 0.9854, + "step": 4195 + }, + { + "epoch": 0.3784100644812193, + "grad_norm": 1.7304137761531024, + "learning_rate": 2.856207528937382e-06, + "loss": 1.0141, + "step": 4196 + }, + { + "epoch": 0.37850024800468957, + "grad_norm": 1.527236193213761, + "learning_rate": 2.855679523986759e-06, + "loss": 0.9919, + "step": 4197 + }, + { + "epoch": 0.3785904315281598, + "grad_norm": 2.6418671705365946, + "learning_rate": 2.8551514460248406e-06, + "loss": 1.0472, + "step": 4198 + }, + { + "epoch": 0.3786806150516301, + "grad_norm": 2.3189999756675284, + "learning_rate": 2.8546232950966868e-06, + "loss": 1.0072, + "step": 4199 + }, + { + "epoch": 0.3787707985751003, + "grad_norm": 1.888621064284691, + "learning_rate": 2.85409507124736e-06, + "loss": 0.9513, + "step": 4200 + }, + { + "epoch": 0.3788609820985706, + "grad_norm": 1.6579411347986754, + "learning_rate": 2.8535667745219324e-06, + "loss": 0.9955, + "step": 4201 + }, + { + "epoch": 0.37895116562204084, + "grad_norm": 1.5888464796783635, + "learning_rate": 2.853038404965481e-06, + "loss": 0.9942, + "step": 4202 + }, + { + "epoch": 0.37904134914551113, + "grad_norm": 9.214253405475263, + "learning_rate": 2.8525099626230894e-06, + "loss": 1.0723, + "step": 4203 + }, + { + "epoch": 0.37913153266898136, + "grad_norm": 1.730362382982371, + "learning_rate": 2.8519814475398472e-06, + "loss": 0.9338, + "step": 4204 + }, + { + "epoch": 0.37922171619245165, + "grad_norm": 1.2793891232842634, + "learning_rate": 2.8514528597608502e-06, + "loss": 0.9801, + "step": 4205 + }, + { + "epoch": 0.3793118997159219, + "grad_norm": 1.2787753141252427, + "learning_rate": 2.8509241993312004e-06, + "loss": 1.0327, + "step": 4206 + }, + { + "epoch": 0.3794020832393922, + "grad_norm": 1.7288465624890628, + "learning_rate": 2.850395466296006e-06, + "loss": 1.0583, + "step": 4207 + }, + { + "epoch": 0.3794922667628624, + "grad_norm": 1.8433250325425912, + "learning_rate": 2.849866660700381e-06, + "loss": 1.0304, + "step": 4208 + }, + { + "epoch": 0.3795824502863327, + "grad_norm": 1.7294026485724145, + "learning_rate": 2.8493377825894464e-06, + "loss": 1.061, + "step": 4209 + }, + { + "epoch": 0.37967263380980293, + "grad_norm": 1.7341374415009974, + "learning_rate": 2.848808832008329e-06, + "loss": 1.0668, + "step": 4210 + }, + { + "epoch": 0.3797628173332732, + "grad_norm": 1.5179020402182515, + "learning_rate": 2.848279809002162e-06, + "loss": 1.0327, + "step": 4211 + }, + { + "epoch": 0.37985300085674345, + "grad_norm": 1.8360538526960477, + "learning_rate": 2.8477507136160842e-06, + "loss": 1.0198, + "step": 4212 + }, + { + "epoch": 0.37994318438021374, + "grad_norm": 0.6397962559550218, + "learning_rate": 2.847221545895241e-06, + "loss": 0.8266, + "step": 4213 + }, + { + "epoch": 0.38003336790368397, + "grad_norm": 1.819043548071088, + "learning_rate": 2.846692305884785e-06, + "loss": 0.9963, + "step": 4214 + }, + { + "epoch": 0.38012355142715426, + "grad_norm": 1.5183312549151122, + "learning_rate": 2.8461629936298718e-06, + "loss": 1.0443, + "step": 4215 + }, + { + "epoch": 0.3802137349506245, + "grad_norm": 1.7359234195553648, + "learning_rate": 2.845633609175666e-06, + "loss": 1.0564, + "step": 4216 + }, + { + "epoch": 0.3803039184740948, + "grad_norm": 1.6320426622779691, + "learning_rate": 2.8451041525673383e-06, + "loss": 1.0423, + "step": 4217 + }, + { + "epoch": 0.38039410199756507, + "grad_norm": 1.8073337194587036, + "learning_rate": 2.8445746238500647e-06, + "loss": 1.031, + "step": 4218 + }, + { + "epoch": 0.3804842855210353, + "grad_norm": 2.011843660791342, + "learning_rate": 2.844045023069027e-06, + "loss": 0.9829, + "step": 4219 + }, + { + "epoch": 0.3805744690445056, + "grad_norm": 1.7415032518495028, + "learning_rate": 2.8435153502694136e-06, + "loss": 1.0452, + "step": 4220 + }, + { + "epoch": 0.3806646525679758, + "grad_norm": 2.7813594667987496, + "learning_rate": 2.84298560549642e-06, + "loss": 0.9382, + "step": 4221 + }, + { + "epoch": 0.3807548360914461, + "grad_norm": 2.131779906322471, + "learning_rate": 2.8424557887952462e-06, + "loss": 1.0305, + "step": 4222 + }, + { + "epoch": 0.38084501961491635, + "grad_norm": 1.6764453914587296, + "learning_rate": 2.841925900211099e-06, + "loss": 0.9333, + "step": 4223 + }, + { + "epoch": 0.38093520313838664, + "grad_norm": 1.4627804479121074, + "learning_rate": 2.841395939789192e-06, + "loss": 1.0553, + "step": 4224 + }, + { + "epoch": 0.38102538666185687, + "grad_norm": 1.7589494885147676, + "learning_rate": 2.8408659075747435e-06, + "loss": 1.0276, + "step": 4225 + }, + { + "epoch": 0.38111557018532716, + "grad_norm": 1.982795447117648, + "learning_rate": 2.8403358036129796e-06, + "loss": 1.0108, + "step": 4226 + }, + { + "epoch": 0.3812057537087974, + "grad_norm": 1.714739889766633, + "learning_rate": 2.839805627949132e-06, + "loss": 1.0629, + "step": 4227 + }, + { + "epoch": 0.3812959372322677, + "grad_norm": 2.0394224821052545, + "learning_rate": 2.8392753806284367e-06, + "loss": 1.0635, + "step": 4228 + }, + { + "epoch": 0.3813861207557379, + "grad_norm": 2.053414069834334, + "learning_rate": 2.838745061696139e-06, + "loss": 0.96, + "step": 4229 + }, + { + "epoch": 0.3814763042792082, + "grad_norm": 1.7054477915784478, + "learning_rate": 2.838214671197487e-06, + "loss": 0.9455, + "step": 4230 + }, + { + "epoch": 0.38156648780267843, + "grad_norm": 1.4031577125593178, + "learning_rate": 2.8376842091777377e-06, + "loss": 1.1112, + "step": 4231 + }, + { + "epoch": 0.3816566713261487, + "grad_norm": 1.680810668154509, + "learning_rate": 2.8371536756821524e-06, + "loss": 0.9142, + "step": 4232 + }, + { + "epoch": 0.38174685484961896, + "grad_norm": 1.5064813143447784, + "learning_rate": 2.836623070756e-06, + "loss": 0.9917, + "step": 4233 + }, + { + "epoch": 0.38183703837308924, + "grad_norm": 1.5994113524741589, + "learning_rate": 2.8360923944445542e-06, + "loss": 1.0324, + "step": 4234 + }, + { + "epoch": 0.3819272218965595, + "grad_norm": 1.4719854390367475, + "learning_rate": 2.8355616467930947e-06, + "loss": 0.992, + "step": 4235 + }, + { + "epoch": 0.38201740542002977, + "grad_norm": 2.7331083143079526, + "learning_rate": 2.8350308278469085e-06, + "loss": 0.9838, + "step": 4236 + }, + { + "epoch": 0.3821075889435, + "grad_norm": 1.5415620596510193, + "learning_rate": 2.8344999376512877e-06, + "loss": 1.0406, + "step": 4237 + }, + { + "epoch": 0.3821977724669703, + "grad_norm": 1.6889148008857282, + "learning_rate": 2.8339689762515307e-06, + "loss": 0.9966, + "step": 4238 + }, + { + "epoch": 0.3822879559904405, + "grad_norm": 1.5716246631196884, + "learning_rate": 2.8334379436929424e-06, + "loss": 0.9415, + "step": 4239 + }, + { + "epoch": 0.3823781395139108, + "grad_norm": 2.0957873097315565, + "learning_rate": 2.832906840020833e-06, + "loss": 0.975, + "step": 4240 + }, + { + "epoch": 0.3824683230373811, + "grad_norm": 1.449723806056114, + "learning_rate": 2.83237566528052e-06, + "loss": 1.023, + "step": 4241 + }, + { + "epoch": 0.38255850656085133, + "grad_norm": 1.7449345848042475, + "learning_rate": 2.831844419517325e-06, + "loss": 0.9862, + "step": 4242 + }, + { + "epoch": 0.3826486900843216, + "grad_norm": 1.4871696265569998, + "learning_rate": 2.8313131027765774e-06, + "loss": 1.0425, + "step": 4243 + }, + { + "epoch": 0.38273887360779185, + "grad_norm": 1.5866710300679407, + "learning_rate": 2.8307817151036124e-06, + "loss": 0.9808, + "step": 4244 + }, + { + "epoch": 0.38282905713126214, + "grad_norm": 1.739983777156927, + "learning_rate": 2.8302502565437704e-06, + "loss": 0.9839, + "step": 4245 + }, + { + "epoch": 0.3829192406547324, + "grad_norm": 2.3224044257825835, + "learning_rate": 2.829718727142398e-06, + "loss": 0.9346, + "step": 4246 + }, + { + "epoch": 0.38300942417820266, + "grad_norm": 1.6396423303377925, + "learning_rate": 2.829187126944849e-06, + "loss": 1.0225, + "step": 4247 + }, + { + "epoch": 0.3830996077016729, + "grad_norm": 2.004766030680028, + "learning_rate": 2.8286554559964826e-06, + "loss": 0.9502, + "step": 4248 + }, + { + "epoch": 0.3831897912251432, + "grad_norm": 1.6565649344961242, + "learning_rate": 2.8281237143426637e-06, + "loss": 1.0287, + "step": 4249 + }, + { + "epoch": 0.3832799747486134, + "grad_norm": 1.6993612295464782, + "learning_rate": 2.8275919020287626e-06, + "loss": 0.9087, + "step": 4250 + }, + { + "epoch": 0.3833701582720837, + "grad_norm": 1.5603370955194433, + "learning_rate": 2.827060019100158e-06, + "loss": 1.0353, + "step": 4251 + }, + { + "epoch": 0.38346034179555394, + "grad_norm": 1.8271841905073607, + "learning_rate": 2.8265280656022315e-06, + "loss": 1.0036, + "step": 4252 + }, + { + "epoch": 0.3835505253190242, + "grad_norm": 1.745586756892717, + "learning_rate": 2.825996041580373e-06, + "loss": 0.8949, + "step": 4253 + }, + { + "epoch": 0.38364070884249446, + "grad_norm": 1.6454020389780264, + "learning_rate": 2.825463947079978e-06, + "loss": 1.0272, + "step": 4254 + }, + { + "epoch": 0.38373089236596475, + "grad_norm": 2.8386362229209823, + "learning_rate": 2.8249317821464483e-06, + "loss": 1.0121, + "step": 4255 + }, + { + "epoch": 0.383821075889435, + "grad_norm": 1.915327813897488, + "learning_rate": 2.824399546825189e-06, + "loss": 0.9381, + "step": 4256 + }, + { + "epoch": 0.38391125941290527, + "grad_norm": 3.7835030897149817, + "learning_rate": 2.823867241161616e-06, + "loss": 0.9613, + "step": 4257 + }, + { + "epoch": 0.3840014429363755, + "grad_norm": 1.8830425707542315, + "learning_rate": 2.8233348652011456e-06, + "loss": 1.058, + "step": 4258 + }, + { + "epoch": 0.3840916264598458, + "grad_norm": 2.240920822215548, + "learning_rate": 2.8228024189892057e-06, + "loss": 0.9214, + "step": 4259 + }, + { + "epoch": 0.384181809983316, + "grad_norm": 1.4121222488850615, + "learning_rate": 2.822269902571226e-06, + "loss": 1.0123, + "step": 4260 + }, + { + "epoch": 0.3842719935067863, + "grad_norm": 1.5765429583454638, + "learning_rate": 2.8217373159926446e-06, + "loss": 0.9899, + "step": 4261 + }, + { + "epoch": 0.38436217703025655, + "grad_norm": 1.2994327371236711, + "learning_rate": 2.8212046592989046e-06, + "loss": 0.9471, + "step": 4262 + }, + { + "epoch": 0.38445236055372684, + "grad_norm": 1.8146764907253334, + "learning_rate": 2.820671932535455e-06, + "loss": 1.0157, + "step": 4263 + }, + { + "epoch": 0.3845425440771971, + "grad_norm": 2.0223788876101505, + "learning_rate": 2.8201391357477506e-06, + "loss": 0.9964, + "step": 4264 + }, + { + "epoch": 0.38463272760066736, + "grad_norm": 1.5551960175783086, + "learning_rate": 2.8196062689812525e-06, + "loss": 0.9843, + "step": 4265 + }, + { + "epoch": 0.38472291112413765, + "grad_norm": 2.17119834842858, + "learning_rate": 2.819073332281429e-06, + "loss": 0.953, + "step": 4266 + }, + { + "epoch": 0.3848130946476079, + "grad_norm": 0.6131311706792406, + "learning_rate": 2.8185403256937524e-06, + "loss": 0.787, + "step": 4267 + }, + { + "epoch": 0.38490327817107817, + "grad_norm": 1.907639481730678, + "learning_rate": 2.8180072492637016e-06, + "loss": 0.9669, + "step": 4268 + }, + { + "epoch": 0.3849934616945484, + "grad_norm": 1.5847244509340366, + "learning_rate": 2.817474103036762e-06, + "loss": 0.9994, + "step": 4269 + }, + { + "epoch": 0.3850836452180187, + "grad_norm": 1.8891864063907062, + "learning_rate": 2.816940887058425e-06, + "loss": 1.0317, + "step": 4270 + }, + { + "epoch": 0.3851738287414889, + "grad_norm": 1.6817438357554824, + "learning_rate": 2.816407601374186e-06, + "loss": 1.0461, + "step": 4271 + }, + { + "epoch": 0.3852640122649592, + "grad_norm": 2.00535486519867, + "learning_rate": 2.815874246029549e-06, + "loss": 0.9399, + "step": 4272 + }, + { + "epoch": 0.38535419578842944, + "grad_norm": 1.7883407936165745, + "learning_rate": 2.815340821070023e-06, + "loss": 0.957, + "step": 4273 + }, + { + "epoch": 0.38544437931189973, + "grad_norm": 1.648716473589205, + "learning_rate": 2.814807326541122e-06, + "loss": 0.9843, + "step": 4274 + }, + { + "epoch": 0.38553456283536996, + "grad_norm": 1.5634963864100815, + "learning_rate": 2.8142737624883676e-06, + "loss": 0.9397, + "step": 4275 + }, + { + "epoch": 0.38562474635884025, + "grad_norm": 1.5060165857564405, + "learning_rate": 2.8137401289572854e-06, + "loss": 1.0069, + "step": 4276 + }, + { + "epoch": 0.3857149298823105, + "grad_norm": 2.2421588962338714, + "learning_rate": 2.8132064259934086e-06, + "loss": 1.0467, + "step": 4277 + }, + { + "epoch": 0.3858051134057808, + "grad_norm": 1.5939746025244699, + "learning_rate": 2.812672653642276e-06, + "loss": 1.0027, + "step": 4278 + }, + { + "epoch": 0.385895296929251, + "grad_norm": 1.4536928277487373, + "learning_rate": 2.812138811949431e-06, + "loss": 1.0157, + "step": 4279 + }, + { + "epoch": 0.3859854804527213, + "grad_norm": 1.502972042684124, + "learning_rate": 2.8116049009604247e-06, + "loss": 1.067, + "step": 4280 + }, + { + "epoch": 0.38607566397619153, + "grad_norm": 1.7348973630289346, + "learning_rate": 2.8110709207208132e-06, + "loss": 0.8893, + "step": 4281 + }, + { + "epoch": 0.3861658474996618, + "grad_norm": 1.6003986875197356, + "learning_rate": 2.810536871276158e-06, + "loss": 1.035, + "step": 4282 + }, + { + "epoch": 0.38625603102313205, + "grad_norm": 0.7194209491627755, + "learning_rate": 2.8100027526720283e-06, + "loss": 0.8669, + "step": 4283 + }, + { + "epoch": 0.38634621454660234, + "grad_norm": 8.045975658936536, + "learning_rate": 2.8094685649539974e-06, + "loss": 0.9319, + "step": 4284 + }, + { + "epoch": 0.3864363980700726, + "grad_norm": 0.7191295450943199, + "learning_rate": 2.8089343081676455e-06, + "loss": 0.839, + "step": 4285 + }, + { + "epoch": 0.38652658159354286, + "grad_norm": 2.1880353000171393, + "learning_rate": 2.8083999823585577e-06, + "loss": 0.9887, + "step": 4286 + }, + { + "epoch": 0.3866167651170131, + "grad_norm": 1.4433993487571886, + "learning_rate": 2.8078655875723254e-06, + "loss": 1.0358, + "step": 4287 + }, + { + "epoch": 0.3867069486404834, + "grad_norm": 3.7003882230245293, + "learning_rate": 2.807331123854547e-06, + "loss": 1.0901, + "step": 4288 + }, + { + "epoch": 0.38679713216395367, + "grad_norm": 2.221864690099476, + "learning_rate": 2.806796591250826e-06, + "loss": 0.8772, + "step": 4289 + }, + { + "epoch": 0.3868873156874239, + "grad_norm": 1.6257703129090233, + "learning_rate": 2.8062619898067707e-06, + "loss": 1.0716, + "step": 4290 + }, + { + "epoch": 0.3869774992108942, + "grad_norm": 1.7691633839995298, + "learning_rate": 2.8057273195679963e-06, + "loss": 1.0182, + "step": 4291 + }, + { + "epoch": 0.3870676827343644, + "grad_norm": 1.6275455637236274, + "learning_rate": 2.8051925805801253e-06, + "loss": 1.02, + "step": 4292 + }, + { + "epoch": 0.3871578662578347, + "grad_norm": 1.8106233154752545, + "learning_rate": 2.804657772888783e-06, + "loss": 1.0359, + "step": 4293 + }, + { + "epoch": 0.38724804978130495, + "grad_norm": 1.621829975657185, + "learning_rate": 2.804122896539602e-06, + "loss": 1.0032, + "step": 4294 + }, + { + "epoch": 0.38733823330477524, + "grad_norm": 1.4675233466713613, + "learning_rate": 2.8035879515782225e-06, + "loss": 0.9553, + "step": 4295 + }, + { + "epoch": 0.38742841682824547, + "grad_norm": 1.877130633007691, + "learning_rate": 2.803052938050288e-06, + "loss": 1.058, + "step": 4296 + }, + { + "epoch": 0.38751860035171576, + "grad_norm": 1.593311324240228, + "learning_rate": 2.802517856001449e-06, + "loss": 1.0523, + "step": 4297 + }, + { + "epoch": 0.387608783875186, + "grad_norm": 2.984862123049526, + "learning_rate": 2.801982705477361e-06, + "loss": 0.9701, + "step": 4298 + }, + { + "epoch": 0.3876989673986563, + "grad_norm": 1.3102428919441607, + "learning_rate": 2.8014474865236867e-06, + "loss": 0.9879, + "step": 4299 + }, + { + "epoch": 0.3877891509221265, + "grad_norm": 1.7993511222482022, + "learning_rate": 2.800912199186094e-06, + "loss": 0.9934, + "step": 4300 + }, + { + "epoch": 0.3878793344455968, + "grad_norm": 1.5517543555017739, + "learning_rate": 2.800376843510256e-06, + "loss": 1.0178, + "step": 4301 + }, + { + "epoch": 0.38796951796906703, + "grad_norm": 1.489321848008538, + "learning_rate": 2.799841419541852e-06, + "loss": 1.024, + "step": 4302 + }, + { + "epoch": 0.3880597014925373, + "grad_norm": 0.6392823547631209, + "learning_rate": 2.799305927326568e-06, + "loss": 0.8214, + "step": 4303 + }, + { + "epoch": 0.38814988501600756, + "grad_norm": 0.7228635748313738, + "learning_rate": 2.7987703669100955e-06, + "loss": 0.8317, + "step": 4304 + }, + { + "epoch": 0.38824006853947785, + "grad_norm": 1.9484297200822336, + "learning_rate": 2.79823473833813e-06, + "loss": 1.0125, + "step": 4305 + }, + { + "epoch": 0.3883302520629481, + "grad_norm": 2.035582514253049, + "learning_rate": 2.797699041656376e-06, + "loss": 0.9519, + "step": 4306 + }, + { + "epoch": 0.38842043558641837, + "grad_norm": 2.8457103302392484, + "learning_rate": 2.7971632769105412e-06, + "loss": 1.0169, + "step": 4307 + }, + { + "epoch": 0.3885106191098886, + "grad_norm": 8.96164626733285, + "learning_rate": 2.79662744414634e-06, + "loss": 0.9686, + "step": 4308 + }, + { + "epoch": 0.3886008026333589, + "grad_norm": 1.7095884116674807, + "learning_rate": 2.7960915434094923e-06, + "loss": 0.9481, + "step": 4309 + }, + { + "epoch": 0.3886909861568291, + "grad_norm": 1.770248503508876, + "learning_rate": 2.7955555747457256e-06, + "loss": 0.9795, + "step": 4310 + }, + { + "epoch": 0.3887811696802994, + "grad_norm": 1.5862101052800828, + "learning_rate": 2.79501953820077e-06, + "loss": 0.9981, + "step": 4311 + }, + { + "epoch": 0.3888713532037697, + "grad_norm": 1.5991245228970867, + "learning_rate": 2.7944834338203637e-06, + "loss": 1.0196, + "step": 4312 + }, + { + "epoch": 0.38896153672723993, + "grad_norm": 1.5704218128930063, + "learning_rate": 2.79394726165025e-06, + "loss": 0.9652, + "step": 4313 + }, + { + "epoch": 0.3890517202507102, + "grad_norm": 1.9295306431575299, + "learning_rate": 2.793411021736178e-06, + "loss": 0.9788, + "step": 4314 + }, + { + "epoch": 0.38914190377418045, + "grad_norm": 1.6468148584260283, + "learning_rate": 2.7928747141239027e-06, + "loss": 1.091, + "step": 4315 + }, + { + "epoch": 0.38923208729765074, + "grad_norm": 0.6474810751060359, + "learning_rate": 2.7923383388591856e-06, + "loss": 0.8258, + "step": 4316 + }, + { + "epoch": 0.389322270821121, + "grad_norm": 1.6861390171049546, + "learning_rate": 2.7918018959877923e-06, + "loss": 0.9666, + "step": 4317 + }, + { + "epoch": 0.38941245434459126, + "grad_norm": 1.8571121808023205, + "learning_rate": 2.791265385555495e-06, + "loss": 1.0274, + "step": 4318 + }, + { + "epoch": 0.3895026378680615, + "grad_norm": 1.7408750919405838, + "learning_rate": 2.790728807608072e-06, + "loss": 1.0662, + "step": 4319 + }, + { + "epoch": 0.3895928213915318, + "grad_norm": 2.0545084480043254, + "learning_rate": 2.790192162191307e-06, + "loss": 0.9915, + "step": 4320 + }, + { + "epoch": 0.389683004915002, + "grad_norm": 1.8046423175654904, + "learning_rate": 2.78965544935099e-06, + "loss": 1.0298, + "step": 4321 + }, + { + "epoch": 0.3897731884384723, + "grad_norm": 2.081760410349234, + "learning_rate": 2.789118669132916e-06, + "loss": 0.9829, + "step": 4322 + }, + { + "epoch": 0.38986337196194254, + "grad_norm": 4.437721354376793, + "learning_rate": 2.7885818215828856e-06, + "loss": 1.0643, + "step": 4323 + }, + { + "epoch": 0.38995355548541283, + "grad_norm": 2.7772784218703515, + "learning_rate": 2.7880449067467064e-06, + "loss": 0.9564, + "step": 4324 + }, + { + "epoch": 0.39004373900888306, + "grad_norm": 2.218759778498203, + "learning_rate": 2.78750792467019e-06, + "loss": 0.9526, + "step": 4325 + }, + { + "epoch": 0.39013392253235335, + "grad_norm": 1.7107863686060216, + "learning_rate": 2.786970875399156e-06, + "loss": 0.8699, + "step": 4326 + }, + { + "epoch": 0.3902241060558236, + "grad_norm": 1.425184820135764, + "learning_rate": 2.7864337589794267e-06, + "loss": 0.9334, + "step": 4327 + }, + { + "epoch": 0.39031428957929387, + "grad_norm": 2.107644847992027, + "learning_rate": 2.7858965754568335e-06, + "loss": 1.017, + "step": 4328 + }, + { + "epoch": 0.3904044731027641, + "grad_norm": 1.7907023681117107, + "learning_rate": 2.785359324877211e-06, + "loss": 0.9447, + "step": 4329 + }, + { + "epoch": 0.3904946566262344, + "grad_norm": 3.7008607275628322, + "learning_rate": 2.7848220072864e-06, + "loss": 1.0308, + "step": 4330 + }, + { + "epoch": 0.3905848401497046, + "grad_norm": 1.9417783200273964, + "learning_rate": 2.784284622730248e-06, + "loss": 0.9884, + "step": 4331 + }, + { + "epoch": 0.3906750236731749, + "grad_norm": 1.80471847969977, + "learning_rate": 2.7837471712546073e-06, + "loss": 1.0146, + "step": 4332 + }, + { + "epoch": 0.39076520719664515, + "grad_norm": 1.4656406523757324, + "learning_rate": 2.783209652905337e-06, + "loss": 0.9759, + "step": 4333 + }, + { + "epoch": 0.39085539072011544, + "grad_norm": 2.0855231284279165, + "learning_rate": 2.7826720677283e-06, + "loss": 0.9608, + "step": 4334 + }, + { + "epoch": 0.39094557424358567, + "grad_norm": 3.2110199186787765, + "learning_rate": 2.782134415769367e-06, + "loss": 0.9656, + "step": 4335 + }, + { + "epoch": 0.39103575776705596, + "grad_norm": 1.449617068928947, + "learning_rate": 2.7815966970744126e-06, + "loss": 1.043, + "step": 4336 + }, + { + "epoch": 0.39112594129052625, + "grad_norm": 1.774911075030106, + "learning_rate": 2.7810589116893184e-06, + "loss": 1.0685, + "step": 4337 + }, + { + "epoch": 0.3912161248139965, + "grad_norm": 2.0615685411485876, + "learning_rate": 2.780521059659972e-06, + "loss": 1.0027, + "step": 4338 + }, + { + "epoch": 0.39130630833746677, + "grad_norm": 1.600427886253517, + "learning_rate": 2.7799831410322637e-06, + "loss": 0.9817, + "step": 4339 + }, + { + "epoch": 0.391396491860937, + "grad_norm": 1.892221076598125, + "learning_rate": 2.779445155852094e-06, + "loss": 0.9553, + "step": 4340 + }, + { + "epoch": 0.3914866753844073, + "grad_norm": 0.657408123813183, + "learning_rate": 2.7789071041653655e-06, + "loss": 0.8635, + "step": 4341 + }, + { + "epoch": 0.3915768589078775, + "grad_norm": 1.532153038463624, + "learning_rate": 2.7783689860179875e-06, + "loss": 1.0557, + "step": 4342 + }, + { + "epoch": 0.3916670424313478, + "grad_norm": 1.4828841904573498, + "learning_rate": 2.7778308014558767e-06, + "loss": 0.9396, + "step": 4343 + }, + { + "epoch": 0.39175722595481804, + "grad_norm": 2.488425737195332, + "learning_rate": 2.7772925505249524e-06, + "loss": 1.074, + "step": 4344 + }, + { + "epoch": 0.39184740947828833, + "grad_norm": 1.3566674201004099, + "learning_rate": 2.7767542332711417e-06, + "loss": 0.967, + "step": 4345 + }, + { + "epoch": 0.39193759300175857, + "grad_norm": 1.717547533579665, + "learning_rate": 2.776215849740377e-06, + "loss": 0.9702, + "step": 4346 + }, + { + "epoch": 0.39202777652522885, + "grad_norm": 1.5476690238876079, + "learning_rate": 2.775677399978596e-06, + "loss": 1.0152, + "step": 4347 + }, + { + "epoch": 0.3921179600486991, + "grad_norm": 0.5882489199322448, + "learning_rate": 2.775138884031742e-06, + "loss": 0.81, + "step": 4348 + }, + { + "epoch": 0.3922081435721694, + "grad_norm": 1.607853142138476, + "learning_rate": 2.774600301945764e-06, + "loss": 0.957, + "step": 4349 + }, + { + "epoch": 0.3922983270956396, + "grad_norm": 1.4378305967237162, + "learning_rate": 2.774061653766618e-06, + "loss": 1.029, + "step": 4350 + }, + { + "epoch": 0.3923885106191099, + "grad_norm": 1.6772473020146668, + "learning_rate": 2.773522939540263e-06, + "loss": 0.9906, + "step": 4351 + }, + { + "epoch": 0.39247869414258013, + "grad_norm": 2.462668642572348, + "learning_rate": 2.7729841593126663e-06, + "loss": 0.9212, + "step": 4352 + }, + { + "epoch": 0.3925688776660504, + "grad_norm": 1.6321616449851766, + "learning_rate": 2.7724453131297988e-06, + "loss": 1.063, + "step": 4353 + }, + { + "epoch": 0.39265906118952065, + "grad_norm": 1.5952128260576082, + "learning_rate": 2.771906401037637e-06, + "loss": 1.0265, + "step": 4354 + }, + { + "epoch": 0.39274924471299094, + "grad_norm": 1.9822565019154272, + "learning_rate": 2.7713674230821664e-06, + "loss": 0.9541, + "step": 4355 + }, + { + "epoch": 0.3928394282364612, + "grad_norm": 1.542436570729017, + "learning_rate": 2.7708283793093724e-06, + "loss": 0.9515, + "step": 4356 + }, + { + "epoch": 0.39292961175993146, + "grad_norm": 1.675871235314374, + "learning_rate": 2.7702892697652514e-06, + "loss": 0.9711, + "step": 4357 + }, + { + "epoch": 0.3930197952834017, + "grad_norm": 1.571164028392252, + "learning_rate": 2.7697500944958024e-06, + "loss": 0.9775, + "step": 4358 + }, + { + "epoch": 0.393109978806872, + "grad_norm": 1.8043540877797586, + "learning_rate": 2.7692108535470312e-06, + "loss": 1.095, + "step": 4359 + }, + { + "epoch": 0.3932001623303423, + "grad_norm": 1.7837754800710242, + "learning_rate": 2.768671546964948e-06, + "loss": 1.0315, + "step": 4360 + }, + { + "epoch": 0.3932903458538125, + "grad_norm": 2.047869724526043, + "learning_rate": 2.7681321747955713e-06, + "loss": 1.0241, + "step": 4361 + }, + { + "epoch": 0.3933805293772828, + "grad_norm": 1.4674040938171706, + "learning_rate": 2.767592737084921e-06, + "loss": 1.0159, + "step": 4362 + }, + { + "epoch": 0.39347071290075303, + "grad_norm": 2.1709511527843914, + "learning_rate": 2.767053233879026e-06, + "loss": 1.0873, + "step": 4363 + }, + { + "epoch": 0.3935608964242233, + "grad_norm": 2.45388315718848, + "learning_rate": 2.76651366522392e-06, + "loss": 1.0729, + "step": 4364 + }, + { + "epoch": 0.39365107994769355, + "grad_norm": 1.589175447033992, + "learning_rate": 2.7659740311656413e-06, + "loss": 0.97, + "step": 4365 + }, + { + "epoch": 0.39374126347116384, + "grad_norm": 1.7905451199221567, + "learning_rate": 2.7654343317502352e-06, + "loss": 1.0191, + "step": 4366 + }, + { + "epoch": 0.39383144699463407, + "grad_norm": 1.8072824688621505, + "learning_rate": 2.7648945670237502e-06, + "loss": 0.9826, + "step": 4367 + }, + { + "epoch": 0.39392163051810436, + "grad_norm": 1.854580765105833, + "learning_rate": 2.7643547370322446e-06, + "loss": 1.0499, + "step": 4368 + }, + { + "epoch": 0.3940118140415746, + "grad_norm": 1.7471143909947475, + "learning_rate": 2.7638148418217775e-06, + "loss": 1.0328, + "step": 4369 + }, + { + "epoch": 0.3941019975650449, + "grad_norm": 1.6610748980617416, + "learning_rate": 2.7632748814384163e-06, + "loss": 1.0398, + "step": 4370 + }, + { + "epoch": 0.3941921810885151, + "grad_norm": 1.3022099802096736, + "learning_rate": 2.7627348559282335e-06, + "loss": 1.0267, + "step": 4371 + }, + { + "epoch": 0.3942823646119854, + "grad_norm": 1.5363335596488465, + "learning_rate": 2.7621947653373075e-06, + "loss": 0.9357, + "step": 4372 + }, + { + "epoch": 0.39437254813545564, + "grad_norm": 0.6589906774738444, + "learning_rate": 2.7616546097117213e-06, + "loss": 0.8015, + "step": 4373 + }, + { + "epoch": 0.3944627316589259, + "grad_norm": 1.7487486725175685, + "learning_rate": 2.761114389097564e-06, + "loss": 0.9941, + "step": 4374 + }, + { + "epoch": 0.39455291518239616, + "grad_norm": 1.8137172689185839, + "learning_rate": 2.7605741035409305e-06, + "loss": 1.0368, + "step": 4375 + }, + { + "epoch": 0.39464309870586645, + "grad_norm": 2.338644035381369, + "learning_rate": 2.76003375308792e-06, + "loss": 0.9764, + "step": 4376 + }, + { + "epoch": 0.3947332822293367, + "grad_norm": 1.2894151147546138, + "learning_rate": 2.75949333778464e-06, + "loss": 0.9617, + "step": 4377 + }, + { + "epoch": 0.39482346575280697, + "grad_norm": 1.486600873386512, + "learning_rate": 2.7589528576772e-06, + "loss": 0.8769, + "step": 4378 + }, + { + "epoch": 0.3949136492762772, + "grad_norm": 5.302681859776792, + "learning_rate": 2.758412312811717e-06, + "loss": 0.9856, + "step": 4379 + }, + { + "epoch": 0.3950038327997475, + "grad_norm": 0.7574088998030419, + "learning_rate": 2.7578717032343146e-06, + "loss": 0.8675, + "step": 4380 + }, + { + "epoch": 0.3950940163232177, + "grad_norm": 1.5165371853539091, + "learning_rate": 2.757331028991119e-06, + "loss": 1.0207, + "step": 4381 + }, + { + "epoch": 0.395184199846688, + "grad_norm": 1.862712013415352, + "learning_rate": 2.7567902901282642e-06, + "loss": 1.009, + "step": 4382 + }, + { + "epoch": 0.3952743833701583, + "grad_norm": 2.0752890603824485, + "learning_rate": 2.7562494866918892e-06, + "loss": 0.9848, + "step": 4383 + }, + { + "epoch": 0.39536456689362853, + "grad_norm": 1.4925375323153094, + "learning_rate": 2.7557086187281378e-06, + "loss": 0.9688, + "step": 4384 + }, + { + "epoch": 0.3954547504170988, + "grad_norm": 1.6716774797074012, + "learning_rate": 2.75516768628316e-06, + "loss": 1.0937, + "step": 4385 + }, + { + "epoch": 0.39554493394056905, + "grad_norm": 1.4468463903115607, + "learning_rate": 2.7546266894031114e-06, + "loss": 1.0681, + "step": 4386 + }, + { + "epoch": 0.39563511746403934, + "grad_norm": 1.524970432682759, + "learning_rate": 2.7540856281341526e-06, + "loss": 0.9448, + "step": 4387 + }, + { + "epoch": 0.3957253009875096, + "grad_norm": 1.5745923377562998, + "learning_rate": 2.7535445025224506e-06, + "loss": 0.9623, + "step": 4388 + }, + { + "epoch": 0.39581548451097986, + "grad_norm": 2.1810577351564295, + "learning_rate": 2.753003312614176e-06, + "loss": 0.9964, + "step": 4389 + }, + { + "epoch": 0.3959056680344501, + "grad_norm": 1.6730639784368657, + "learning_rate": 2.7524620584555065e-06, + "loss": 0.9987, + "step": 4390 + }, + { + "epoch": 0.3959958515579204, + "grad_norm": 2.296509278952874, + "learning_rate": 2.7519207400926253e-06, + "loss": 1.0359, + "step": 4391 + }, + { + "epoch": 0.3960860350813906, + "grad_norm": 1.559803123032933, + "learning_rate": 2.751379357571721e-06, + "loss": 0.9788, + "step": 4392 + }, + { + "epoch": 0.3961762186048609, + "grad_norm": 1.5145254177405065, + "learning_rate": 2.7508379109389865e-06, + "loss": 1.0627, + "step": 4393 + }, + { + "epoch": 0.39626640212833114, + "grad_norm": 1.7771722396428826, + "learning_rate": 2.750296400240622e-06, + "loss": 1.0126, + "step": 4394 + }, + { + "epoch": 0.39635658565180143, + "grad_norm": 1.3672118266529951, + "learning_rate": 2.7497548255228305e-06, + "loss": 1.0288, + "step": 4395 + }, + { + "epoch": 0.39644676917527166, + "grad_norm": 1.5469932896247227, + "learning_rate": 2.749213186831824e-06, + "loss": 0.9497, + "step": 4396 + }, + { + "epoch": 0.39653695269874195, + "grad_norm": 2.241697038447126, + "learning_rate": 2.7486714842138173e-06, + "loss": 1.0473, + "step": 4397 + }, + { + "epoch": 0.3966271362222122, + "grad_norm": 2.3094022467919686, + "learning_rate": 2.748129717715031e-06, + "loss": 0.9495, + "step": 4398 + }, + { + "epoch": 0.3967173197456825, + "grad_norm": 1.6293975308773334, + "learning_rate": 2.747587887381692e-06, + "loss": 1.0147, + "step": 4399 + }, + { + "epoch": 0.3968075032691527, + "grad_norm": 1.5871988176008789, + "learning_rate": 2.7470459932600328e-06, + "loss": 0.9481, + "step": 4400 + }, + { + "epoch": 0.396897686792623, + "grad_norm": 3.630111247898491, + "learning_rate": 2.7465040353962897e-06, + "loss": 0.9961, + "step": 4401 + }, + { + "epoch": 0.3969878703160932, + "grad_norm": 2.04645947613963, + "learning_rate": 2.745962013836706e-06, + "loss": 0.9879, + "step": 4402 + }, + { + "epoch": 0.3970780538395635, + "grad_norm": 2.078857178164438, + "learning_rate": 2.74541992862753e-06, + "loss": 0.9752, + "step": 4403 + }, + { + "epoch": 0.39716823736303375, + "grad_norm": 1.903680664523136, + "learning_rate": 2.744877779815016e-06, + "loss": 0.9976, + "step": 4404 + }, + { + "epoch": 0.39725842088650404, + "grad_norm": 1.9756372268814917, + "learning_rate": 2.7443355674454234e-06, + "loss": 0.9928, + "step": 4405 + }, + { + "epoch": 0.39734860440997427, + "grad_norm": 2.210081116581497, + "learning_rate": 2.743793291565015e-06, + "loss": 1.0359, + "step": 4406 + }, + { + "epoch": 0.39743878793344456, + "grad_norm": 2.068261624224613, + "learning_rate": 2.7432509522200617e-06, + "loss": 0.9202, + "step": 4407 + }, + { + "epoch": 0.39752897145691485, + "grad_norm": 2.652927879316591, + "learning_rate": 2.7427085494568383e-06, + "loss": 0.941, + "step": 4408 + }, + { + "epoch": 0.3976191549803851, + "grad_norm": 0.6571872240284528, + "learning_rate": 2.742166083321628e-06, + "loss": 0.8441, + "step": 4409 + }, + { + "epoch": 0.39770933850385537, + "grad_norm": 1.649398234479451, + "learning_rate": 2.7416235538607137e-06, + "loss": 1.0376, + "step": 4410 + }, + { + "epoch": 0.3977995220273256, + "grad_norm": 1.7546577457884334, + "learning_rate": 2.7410809611203894e-06, + "loss": 0.9734, + "step": 4411 + }, + { + "epoch": 0.3978897055507959, + "grad_norm": 1.652874007037148, + "learning_rate": 2.7405383051469507e-06, + "loss": 0.9555, + "step": 4412 + }, + { + "epoch": 0.3979798890742661, + "grad_norm": 2.1310518144656876, + "learning_rate": 2.7399955859867e-06, + "loss": 0.81, + "step": 4413 + }, + { + "epoch": 0.3980700725977364, + "grad_norm": 0.6252852742503504, + "learning_rate": 2.7394528036859465e-06, + "loss": 0.788, + "step": 4414 + }, + { + "epoch": 0.39816025612120665, + "grad_norm": 1.5941881063840544, + "learning_rate": 2.738909958291002e-06, + "loss": 0.9999, + "step": 4415 + }, + { + "epoch": 0.39825043964467693, + "grad_norm": 1.5273114369284067, + "learning_rate": 2.7383670498481863e-06, + "loss": 0.9103, + "step": 4416 + }, + { + "epoch": 0.39834062316814717, + "grad_norm": 0.5790923762758832, + "learning_rate": 2.737824078403822e-06, + "loss": 0.7467, + "step": 4417 + }, + { + "epoch": 0.39843080669161746, + "grad_norm": 1.5602156916695398, + "learning_rate": 2.737281044004239e-06, + "loss": 0.9731, + "step": 4418 + }, + { + "epoch": 0.3985209902150877, + "grad_norm": 1.5909197491127782, + "learning_rate": 2.736737946695772e-06, + "loss": 0.983, + "step": 4419 + }, + { + "epoch": 0.398611173738558, + "grad_norm": 1.8768659843357858, + "learning_rate": 2.736194786524761e-06, + "loss": 0.9924, + "step": 4420 + }, + { + "epoch": 0.3987013572620282, + "grad_norm": 1.6485043914384696, + "learning_rate": 2.7356515635375517e-06, + "loss": 0.9948, + "step": 4421 + }, + { + "epoch": 0.3987915407854985, + "grad_norm": 2.6075246497769364, + "learning_rate": 2.735108277780495e-06, + "loss": 0.9062, + "step": 4422 + }, + { + "epoch": 0.39888172430896873, + "grad_norm": 1.5188569678795771, + "learning_rate": 2.7345649292999456e-06, + "loss": 0.9857, + "step": 4423 + }, + { + "epoch": 0.398971907832439, + "grad_norm": 1.6740688297792234, + "learning_rate": 2.734021518142267e-06, + "loss": 1.0011, + "step": 4424 + }, + { + "epoch": 0.39906209135590925, + "grad_norm": 2.064496316882022, + "learning_rate": 2.733478044353825e-06, + "loss": 0.8277, + "step": 4425 + }, + { + "epoch": 0.39915227487937954, + "grad_norm": 1.7780554384056948, + "learning_rate": 2.7329345079809917e-06, + "loss": 1.0021, + "step": 4426 + }, + { + "epoch": 0.3992424584028498, + "grad_norm": 1.551229032732887, + "learning_rate": 2.7323909090701447e-06, + "loss": 0.9336, + "step": 4427 + }, + { + "epoch": 0.39933264192632006, + "grad_norm": 1.7320842564463899, + "learning_rate": 2.731847247667667e-06, + "loss": 1.0478, + "step": 4428 + }, + { + "epoch": 0.3994228254497903, + "grad_norm": 1.704260211544101, + "learning_rate": 2.731303523819947e-06, + "loss": 0.9749, + "step": 4429 + }, + { + "epoch": 0.3995130089732606, + "grad_norm": 1.8355281292270063, + "learning_rate": 2.7307597375733783e-06, + "loss": 1.0558, + "step": 4430 + }, + { + "epoch": 0.3996031924967309, + "grad_norm": 2.183796172066059, + "learning_rate": 2.7302158889743587e-06, + "loss": 0.9957, + "step": 4431 + }, + { + "epoch": 0.3996933760202011, + "grad_norm": 2.437220532951217, + "learning_rate": 2.7296719780692937e-06, + "loss": 0.9512, + "step": 4432 + }, + { + "epoch": 0.3997835595436714, + "grad_norm": 1.817734099663666, + "learning_rate": 2.7291280049045916e-06, + "loss": 0.9046, + "step": 4433 + }, + { + "epoch": 0.39987374306714163, + "grad_norm": 1.507275105112758, + "learning_rate": 2.7285839695266683e-06, + "loss": 0.9373, + "step": 4434 + }, + { + "epoch": 0.3999639265906119, + "grad_norm": 1.7807560285974693, + "learning_rate": 2.7280398719819423e-06, + "loss": 0.9457, + "step": 4435 + }, + { + "epoch": 0.40005411011408215, + "grad_norm": 1.9366731417488061, + "learning_rate": 2.727495712316841e-06, + "loss": 1.0624, + "step": 4436 + }, + { + "epoch": 0.40014429363755244, + "grad_norm": 1.7269060302332202, + "learning_rate": 2.7269514905777945e-06, + "loss": 0.9357, + "step": 4437 + }, + { + "epoch": 0.4002344771610227, + "grad_norm": 1.8322924202083783, + "learning_rate": 2.7264072068112377e-06, + "loss": 1.1146, + "step": 4438 + }, + { + "epoch": 0.40032466068449296, + "grad_norm": 2.0139498588131426, + "learning_rate": 2.7258628610636133e-06, + "loss": 0.9556, + "step": 4439 + }, + { + "epoch": 0.4004148442079632, + "grad_norm": 1.6975079924180387, + "learning_rate": 2.7253184533813667e-06, + "loss": 1.074, + "step": 4440 + }, + { + "epoch": 0.4005050277314335, + "grad_norm": 1.680371025645448, + "learning_rate": 2.72477398381095e-06, + "loss": 1.0048, + "step": 4441 + }, + { + "epoch": 0.4005952112549037, + "grad_norm": 1.8268813116882687, + "learning_rate": 2.724229452398821e-06, + "loss": 1.001, + "step": 4442 + }, + { + "epoch": 0.400685394778374, + "grad_norm": 1.7361695911306794, + "learning_rate": 2.7236848591914422e-06, + "loss": 1.014, + "step": 4443 + }, + { + "epoch": 0.40077557830184424, + "grad_norm": 2.110814373454793, + "learning_rate": 2.7231402042352803e-06, + "loss": 0.9794, + "step": 4444 + }, + { + "epoch": 0.4008657618253145, + "grad_norm": 1.37481701240158, + "learning_rate": 2.722595487576809e-06, + "loss": 1.0208, + "step": 4445 + }, + { + "epoch": 0.40095594534878476, + "grad_norm": 1.9036610016085873, + "learning_rate": 2.722050709262506e-06, + "loss": 1.04, + "step": 4446 + }, + { + "epoch": 0.40104612887225505, + "grad_norm": 1.7047592161986493, + "learning_rate": 2.7215058693388557e-06, + "loss": 0.9718, + "step": 4447 + }, + { + "epoch": 0.4011363123957253, + "grad_norm": 1.8242060070695518, + "learning_rate": 2.720960967852346e-06, + "loss": 0.8972, + "step": 4448 + }, + { + "epoch": 0.40122649591919557, + "grad_norm": 1.3665231452960993, + "learning_rate": 2.720416004849471e-06, + "loss": 0.9289, + "step": 4449 + }, + { + "epoch": 0.4013166794426658, + "grad_norm": 1.8314576379540863, + "learning_rate": 2.7198709803767304e-06, + "loss": 0.9481, + "step": 4450 + }, + { + "epoch": 0.4014068629661361, + "grad_norm": 1.5136317881536623, + "learning_rate": 2.7193258944806286e-06, + "loss": 0.9636, + "step": 4451 + }, + { + "epoch": 0.4014970464896063, + "grad_norm": 2.607156781497209, + "learning_rate": 2.718780747207675e-06, + "loss": 0.998, + "step": 4452 + }, + { + "epoch": 0.4015872300130766, + "grad_norm": 1.7173418692431954, + "learning_rate": 2.7182355386043847e-06, + "loss": 1.0589, + "step": 4453 + }, + { + "epoch": 0.40167741353654685, + "grad_norm": 1.5630514316260757, + "learning_rate": 2.717690268717278e-06, + "loss": 1.0566, + "step": 4454 + }, + { + "epoch": 0.40176759706001713, + "grad_norm": 1.7281196442395181, + "learning_rate": 2.7171449375928803e-06, + "loss": 1.019, + "step": 4455 + }, + { + "epoch": 0.4018577805834874, + "grad_norm": 2.1498833247618605, + "learning_rate": 2.716599545277722e-06, + "loss": 1.0249, + "step": 4456 + }, + { + "epoch": 0.40194796410695766, + "grad_norm": 1.7071672817473549, + "learning_rate": 2.7160540918183394e-06, + "loss": 0.9573, + "step": 4457 + }, + { + "epoch": 0.40203814763042794, + "grad_norm": 4.178779974124452, + "learning_rate": 2.715508577261273e-06, + "loss": 0.9105, + "step": 4458 + }, + { + "epoch": 0.4021283311538982, + "grad_norm": 1.7910061069514334, + "learning_rate": 2.7149630016530702e-06, + "loss": 0.974, + "step": 4459 + }, + { + "epoch": 0.40221851467736847, + "grad_norm": 1.8154437396494212, + "learning_rate": 2.7144173650402815e-06, + "loss": 1.0102, + "step": 4460 + }, + { + "epoch": 0.4023086982008387, + "grad_norm": 1.6392847308699992, + "learning_rate": 2.7138716674694636e-06, + "loss": 1.016, + "step": 4461 + }, + { + "epoch": 0.402398881724309, + "grad_norm": 1.820296913702384, + "learning_rate": 2.7133259089871795e-06, + "loss": 0.9395, + "step": 4462 + }, + { + "epoch": 0.4024890652477792, + "grad_norm": 1.8859163398472134, + "learning_rate": 2.712780089639995e-06, + "loss": 0.9939, + "step": 4463 + }, + { + "epoch": 0.4025792487712495, + "grad_norm": 1.6752023830845013, + "learning_rate": 2.712234209474483e-06, + "loss": 0.8737, + "step": 4464 + }, + { + "epoch": 0.40266943229471974, + "grad_norm": 1.52071062228738, + "learning_rate": 2.7116882685372218e-06, + "loss": 1.0368, + "step": 4465 + }, + { + "epoch": 0.40275961581819003, + "grad_norm": 1.827318323291996, + "learning_rate": 2.7111422668747927e-06, + "loss": 0.962, + "step": 4466 + }, + { + "epoch": 0.40284979934166026, + "grad_norm": 1.6544912825280107, + "learning_rate": 2.7105962045337846e-06, + "loss": 0.991, + "step": 4467 + }, + { + "epoch": 0.40293998286513055, + "grad_norm": 1.5489347211948834, + "learning_rate": 2.7100500815607898e-06, + "loss": 0.9877, + "step": 4468 + }, + { + "epoch": 0.4030301663886008, + "grad_norm": 2.126990564050087, + "learning_rate": 2.709503898002407e-06, + "loss": 1.0177, + "step": 4469 + }, + { + "epoch": 0.4031203499120711, + "grad_norm": 1.767984484755503, + "learning_rate": 2.708957653905239e-06, + "loss": 0.9902, + "step": 4470 + }, + { + "epoch": 0.4032105334355413, + "grad_norm": 2.272226332163598, + "learning_rate": 2.7084113493158956e-06, + "loss": 1.0063, + "step": 4471 + }, + { + "epoch": 0.4033007169590116, + "grad_norm": 2.1867930768659964, + "learning_rate": 2.7078649842809888e-06, + "loss": 1.0054, + "step": 4472 + }, + { + "epoch": 0.40339090048248183, + "grad_norm": 1.5698529230462797, + "learning_rate": 2.707318558847139e-06, + "loss": 1.0424, + "step": 4473 + }, + { + "epoch": 0.4034810840059521, + "grad_norm": 0.7215204334181006, + "learning_rate": 2.7067720730609697e-06, + "loss": 0.8171, + "step": 4474 + }, + { + "epoch": 0.40357126752942235, + "grad_norm": 1.629812596787426, + "learning_rate": 2.70622552696911e-06, + "loss": 0.9665, + "step": 4475 + }, + { + "epoch": 0.40366145105289264, + "grad_norm": 1.447079624676897, + "learning_rate": 2.7056789206181943e-06, + "loss": 0.9501, + "step": 4476 + }, + { + "epoch": 0.40375163457636287, + "grad_norm": 1.6422904870635817, + "learning_rate": 2.7051322540548615e-06, + "loss": 1.0484, + "step": 4477 + }, + { + "epoch": 0.40384181809983316, + "grad_norm": 1.7687074878557636, + "learning_rate": 2.704585527325757e-06, + "loss": 0.9658, + "step": 4478 + }, + { + "epoch": 0.40393200162330345, + "grad_norm": 1.7690042211936154, + "learning_rate": 2.7040387404775303e-06, + "loss": 0.9506, + "step": 4479 + }, + { + "epoch": 0.4040221851467737, + "grad_norm": 1.9256454946089099, + "learning_rate": 2.703491893556837e-06, + "loss": 1.0226, + "step": 4480 + }, + { + "epoch": 0.40411236867024397, + "grad_norm": 1.3398731995841044, + "learning_rate": 2.702944986610335e-06, + "loss": 0.968, + "step": 4481 + }, + { + "epoch": 0.4042025521937142, + "grad_norm": 1.8325074965673636, + "learning_rate": 2.7023980196846917e-06, + "loss": 1.0064, + "step": 4482 + }, + { + "epoch": 0.4042927357171845, + "grad_norm": 0.7059270761909948, + "learning_rate": 2.7018509928265763e-06, + "loss": 0.8353, + "step": 4483 + }, + { + "epoch": 0.4043829192406547, + "grad_norm": 1.6131347146345851, + "learning_rate": 2.7013039060826635e-06, + "loss": 1.0205, + "step": 4484 + }, + { + "epoch": 0.404473102764125, + "grad_norm": 1.9276873148978142, + "learning_rate": 2.7007567594996347e-06, + "loss": 1.0608, + "step": 4485 + }, + { + "epoch": 0.40456328628759525, + "grad_norm": 1.5615390874616402, + "learning_rate": 2.7002095531241757e-06, + "loss": 0.9018, + "step": 4486 + }, + { + "epoch": 0.40465346981106554, + "grad_norm": 2.061150774212815, + "learning_rate": 2.6996622870029767e-06, + "loss": 0.9672, + "step": 4487 + }, + { + "epoch": 0.40474365333453577, + "grad_norm": 1.658145036342675, + "learning_rate": 2.6991149611827335e-06, + "loss": 1.1025, + "step": 4488 + }, + { + "epoch": 0.40483383685800606, + "grad_norm": 1.8100821053866498, + "learning_rate": 2.6985675757101466e-06, + "loss": 1.0339, + "step": 4489 + }, + { + "epoch": 0.4049240203814763, + "grad_norm": 1.6003608475517979, + "learning_rate": 2.698020130631922e-06, + "loss": 0.9525, + "step": 4490 + }, + { + "epoch": 0.4050142039049466, + "grad_norm": 1.6953115155617102, + "learning_rate": 2.6974726259947713e-06, + "loss": 0.9257, + "step": 4491 + }, + { + "epoch": 0.4051043874284168, + "grad_norm": 1.8596284196764117, + "learning_rate": 2.6969250618454106e-06, + "loss": 0.9705, + "step": 4492 + }, + { + "epoch": 0.4051945709518871, + "grad_norm": 1.353413103192468, + "learning_rate": 2.696377438230561e-06, + "loss": 1.067, + "step": 4493 + }, + { + "epoch": 0.40528475447535733, + "grad_norm": 1.7093909951530406, + "learning_rate": 2.6958297551969484e-06, + "loss": 0.9304, + "step": 4494 + }, + { + "epoch": 0.4053749379988276, + "grad_norm": 2.5988700612175233, + "learning_rate": 2.695282012791304e-06, + "loss": 0.9587, + "step": 4495 + }, + { + "epoch": 0.40546512152229786, + "grad_norm": 2.038614274962004, + "learning_rate": 2.6947342110603646e-06, + "loss": 1.0474, + "step": 4496 + }, + { + "epoch": 0.40555530504576814, + "grad_norm": 1.5529973797011516, + "learning_rate": 2.6941863500508717e-06, + "loss": 1.0484, + "step": 4497 + }, + { + "epoch": 0.4056454885692384, + "grad_norm": 1.6891333834219122, + "learning_rate": 2.693638429809572e-06, + "loss": 1.0457, + "step": 4498 + }, + { + "epoch": 0.40573567209270867, + "grad_norm": 1.8356541191703437, + "learning_rate": 2.6930904503832167e-06, + "loss": 0.9803, + "step": 4499 + }, + { + "epoch": 0.4058258556161789, + "grad_norm": 1.48681303878929, + "learning_rate": 2.692542411818562e-06, + "loss": 0.9769, + "step": 4500 + }, + { + "epoch": 0.4059160391396492, + "grad_norm": 2.911215529215236, + "learning_rate": 2.69199431416237e-06, + "loss": 1.0031, + "step": 4501 + }, + { + "epoch": 0.4060062226631194, + "grad_norm": 5.4546022354407295, + "learning_rate": 2.691446157461408e-06, + "loss": 0.947, + "step": 4502 + }, + { + "epoch": 0.4060964061865897, + "grad_norm": 1.4949449555843355, + "learning_rate": 2.690897941762447e-06, + "loss": 1.058, + "step": 4503 + }, + { + "epoch": 0.40618658971006, + "grad_norm": 1.5096176489477076, + "learning_rate": 2.6903496671122642e-06, + "loss": 0.9717, + "step": 4504 + }, + { + "epoch": 0.40627677323353023, + "grad_norm": 0.6463620165138935, + "learning_rate": 2.689801333557641e-06, + "loss": 0.7665, + "step": 4505 + }, + { + "epoch": 0.4063669567570005, + "grad_norm": 2.2585653877918603, + "learning_rate": 2.689252941145365e-06, + "loss": 0.9946, + "step": 4506 + }, + { + "epoch": 0.40645714028047075, + "grad_norm": 1.4585814219578523, + "learning_rate": 2.6887044899222277e-06, + "loss": 0.9912, + "step": 4507 + }, + { + "epoch": 0.40654732380394104, + "grad_norm": 1.685676932151957, + "learning_rate": 2.688155979935025e-06, + "loss": 1.0394, + "step": 4508 + }, + { + "epoch": 0.4066375073274113, + "grad_norm": 1.7173148665950697, + "learning_rate": 2.68760741123056e-06, + "loss": 0.9548, + "step": 4509 + }, + { + "epoch": 0.40672769085088156, + "grad_norm": 2.039765803039653, + "learning_rate": 2.6870587838556394e-06, + "loss": 1.0095, + "step": 4510 + }, + { + "epoch": 0.4068178743743518, + "grad_norm": 1.883486488607748, + "learning_rate": 2.686510097857075e-06, + "loss": 1.0141, + "step": 4511 + }, + { + "epoch": 0.4069080578978221, + "grad_norm": 1.3305677053462694, + "learning_rate": 2.685961353281683e-06, + "loss": 0.9893, + "step": 4512 + }, + { + "epoch": 0.4069982414212923, + "grad_norm": 1.676180705231926, + "learning_rate": 2.6854125501762863e-06, + "loss": 0.9763, + "step": 4513 + }, + { + "epoch": 0.4070884249447626, + "grad_norm": 1.8301173382021534, + "learning_rate": 2.684863688587712e-06, + "loss": 0.9999, + "step": 4514 + }, + { + "epoch": 0.40717860846823284, + "grad_norm": 1.6156938077046428, + "learning_rate": 2.6843147685627916e-06, + "loss": 0.9618, + "step": 4515 + }, + { + "epoch": 0.4072687919917031, + "grad_norm": 3.1183503164307704, + "learning_rate": 2.683765790148361e-06, + "loss": 0.9669, + "step": 4516 + }, + { + "epoch": 0.40735897551517336, + "grad_norm": 2.108336694687085, + "learning_rate": 2.6832167533912637e-06, + "loss": 1.0463, + "step": 4517 + }, + { + "epoch": 0.40744915903864365, + "grad_norm": 2.8510114581554546, + "learning_rate": 2.682667658338345e-06, + "loss": 1.0689, + "step": 4518 + }, + { + "epoch": 0.4075393425621139, + "grad_norm": 1.8912277521916259, + "learning_rate": 2.682118505036458e-06, + "loss": 0.9691, + "step": 4519 + }, + { + "epoch": 0.40762952608558417, + "grad_norm": 0.6185577007060109, + "learning_rate": 2.681569293532459e-06, + "loss": 0.787, + "step": 4520 + }, + { + "epoch": 0.4077197096090544, + "grad_norm": 0.6458382503773543, + "learning_rate": 2.6810200238732102e-06, + "loss": 0.8148, + "step": 4521 + }, + { + "epoch": 0.4078098931325247, + "grad_norm": 1.631035307571632, + "learning_rate": 2.6804706961055776e-06, + "loss": 1.039, + "step": 4522 + }, + { + "epoch": 0.4079000766559949, + "grad_norm": 1.6623314822268056, + "learning_rate": 2.6799213102764326e-06, + "loss": 1.0293, + "step": 4523 + }, + { + "epoch": 0.4079902601794652, + "grad_norm": 2.2930333195031873, + "learning_rate": 2.679371866432653e-06, + "loss": 1.089, + "step": 4524 + }, + { + "epoch": 0.40808044370293545, + "grad_norm": 1.5704839053004847, + "learning_rate": 2.6788223646211194e-06, + "loss": 0.9678, + "step": 4525 + }, + { + "epoch": 0.40817062722640574, + "grad_norm": 1.472518469236441, + "learning_rate": 2.6782728048887183e-06, + "loss": 0.8886, + "step": 4526 + }, + { + "epoch": 0.408260810749876, + "grad_norm": 1.4372973091857772, + "learning_rate": 2.6777231872823416e-06, + "loss": 0.9808, + "step": 4527 + }, + { + "epoch": 0.40835099427334626, + "grad_norm": 1.8839247334827958, + "learning_rate": 2.6771735118488864e-06, + "loss": 0.9935, + "step": 4528 + }, + { + "epoch": 0.40844117779681655, + "grad_norm": 1.4607889844506239, + "learning_rate": 2.6766237786352523e-06, + "loss": 1.0296, + "step": 4529 + }, + { + "epoch": 0.4085313613202868, + "grad_norm": 1.3444386868752043, + "learning_rate": 2.676073987688347e-06, + "loss": 0.9211, + "step": 4530 + }, + { + "epoch": 0.40862154484375707, + "grad_norm": 1.7213170388546506, + "learning_rate": 2.6755241390550818e-06, + "loss": 0.9931, + "step": 4531 + }, + { + "epoch": 0.4087117283672273, + "grad_norm": 1.9145080047942193, + "learning_rate": 2.6749742327823716e-06, + "loss": 0.9954, + "step": 4532 + }, + { + "epoch": 0.4088019118906976, + "grad_norm": 1.543444982503984, + "learning_rate": 2.674424268917138e-06, + "loss": 1.0072, + "step": 4533 + }, + { + "epoch": 0.4088920954141678, + "grad_norm": 1.690231091201022, + "learning_rate": 2.6738742475063074e-06, + "loss": 1.0606, + "step": 4534 + }, + { + "epoch": 0.4089822789376381, + "grad_norm": 1.7951750965840714, + "learning_rate": 2.6733241685968104e-06, + "loss": 1.0339, + "step": 4535 + }, + { + "epoch": 0.40907246246110834, + "grad_norm": 2.1206285728831484, + "learning_rate": 2.6727740322355826e-06, + "loss": 1.0111, + "step": 4536 + }, + { + "epoch": 0.40916264598457863, + "grad_norm": 2.197971132396064, + "learning_rate": 2.6722238384695644e-06, + "loss": 0.9789, + "step": 4537 + }, + { + "epoch": 0.40925282950804887, + "grad_norm": 1.5528645778743222, + "learning_rate": 2.671673587345702e-06, + "loss": 1.0063, + "step": 4538 + }, + { + "epoch": 0.40934301303151915, + "grad_norm": 0.6460778147936039, + "learning_rate": 2.6711232789109455e-06, + "loss": 0.8048, + "step": 4539 + }, + { + "epoch": 0.4094331965549894, + "grad_norm": 1.5275676592437746, + "learning_rate": 2.6705729132122497e-06, + "loss": 0.9995, + "step": 4540 + }, + { + "epoch": 0.4095233800784597, + "grad_norm": 2.855951479748088, + "learning_rate": 2.670022490296576e-06, + "loss": 1.0692, + "step": 4541 + }, + { + "epoch": 0.4096135636019299, + "grad_norm": 1.6512105315903873, + "learning_rate": 2.669472010210889e-06, + "loss": 1.0535, + "step": 4542 + }, + { + "epoch": 0.4097037471254002, + "grad_norm": 1.9346011841499728, + "learning_rate": 2.668921473002159e-06, + "loss": 0.9716, + "step": 4543 + }, + { + "epoch": 0.40979393064887043, + "grad_norm": 1.6483988825175204, + "learning_rate": 2.6683708787173596e-06, + "loss": 1.0495, + "step": 4544 + }, + { + "epoch": 0.4098841141723407, + "grad_norm": 1.5749768876090486, + "learning_rate": 2.6678202274034718e-06, + "loss": 1.0133, + "step": 4545 + }, + { + "epoch": 0.40997429769581095, + "grad_norm": 1.885125413461436, + "learning_rate": 2.66726951910748e-06, + "loss": 1.0833, + "step": 4546 + }, + { + "epoch": 0.41006448121928124, + "grad_norm": 1.9091098157970334, + "learning_rate": 2.6667187538763737e-06, + "loss": 0.9964, + "step": 4547 + }, + { + "epoch": 0.4101546647427515, + "grad_norm": 1.469773443241096, + "learning_rate": 2.6661679317571473e-06, + "loss": 0.977, + "step": 4548 + }, + { + "epoch": 0.41024484826622176, + "grad_norm": 3.1039648481128634, + "learning_rate": 2.665617052796799e-06, + "loss": 0.9784, + "step": 4549 + }, + { + "epoch": 0.41033503178969205, + "grad_norm": 4.234452700869774, + "learning_rate": 2.6650661170423346e-06, + "loss": 0.8823, + "step": 4550 + }, + { + "epoch": 0.4104252153131623, + "grad_norm": 7.423409072542128, + "learning_rate": 2.6645151245407614e-06, + "loss": 1.0761, + "step": 4551 + }, + { + "epoch": 0.4105153988366326, + "grad_norm": 1.836401698089725, + "learning_rate": 2.6639640753390936e-06, + "loss": 0.9912, + "step": 4552 + }, + { + "epoch": 0.4106055823601028, + "grad_norm": 1.547887836596841, + "learning_rate": 2.66341296948435e-06, + "loss": 0.9035, + "step": 4553 + }, + { + "epoch": 0.4106957658835731, + "grad_norm": 1.347630376152777, + "learning_rate": 2.6628618070235534e-06, + "loss": 1.0034, + "step": 4554 + }, + { + "epoch": 0.4107859494070433, + "grad_norm": 1.7378562836395446, + "learning_rate": 2.662310588003733e-06, + "loss": 1.0002, + "step": 4555 + }, + { + "epoch": 0.4108761329305136, + "grad_norm": 2.136901624479671, + "learning_rate": 2.6617593124719205e-06, + "loss": 0.9661, + "step": 4556 + }, + { + "epoch": 0.41096631645398385, + "grad_norm": 1.4683876402240965, + "learning_rate": 2.661207980475155e-06, + "loss": 1.002, + "step": 4557 + }, + { + "epoch": 0.41105649997745414, + "grad_norm": 1.7628430080090032, + "learning_rate": 2.6606565920604793e-06, + "loss": 0.9507, + "step": 4558 + }, + { + "epoch": 0.41114668350092437, + "grad_norm": 1.4823925231520643, + "learning_rate": 2.66010514727494e-06, + "loss": 1.0736, + "step": 4559 + }, + { + "epoch": 0.41123686702439466, + "grad_norm": 1.8070977040039662, + "learning_rate": 2.659553646165589e-06, + "loss": 0.8889, + "step": 4560 + }, + { + "epoch": 0.4113270505478649, + "grad_norm": 1.5762366148027163, + "learning_rate": 2.659002088779485e-06, + "loss": 0.9682, + "step": 4561 + }, + { + "epoch": 0.4114172340713352, + "grad_norm": 0.5885478303524656, + "learning_rate": 2.6584504751636888e-06, + "loss": 0.7647, + "step": 4562 + }, + { + "epoch": 0.4115074175948054, + "grad_norm": 2.1504977248959634, + "learning_rate": 2.657898805365268e-06, + "loss": 0.9585, + "step": 4563 + }, + { + "epoch": 0.4115976011182757, + "grad_norm": 1.656064724905821, + "learning_rate": 2.657347079431293e-06, + "loss": 0.9618, + "step": 4564 + }, + { + "epoch": 0.41168778464174594, + "grad_norm": 1.4764711059058038, + "learning_rate": 2.6567952974088403e-06, + "loss": 0.9691, + "step": 4565 + }, + { + "epoch": 0.4117779681652162, + "grad_norm": 1.6499667193206422, + "learning_rate": 2.6562434593449917e-06, + "loss": 1.0414, + "step": 4566 + }, + { + "epoch": 0.41186815168868646, + "grad_norm": 1.7499993869235464, + "learning_rate": 2.6556915652868325e-06, + "loss": 1.0172, + "step": 4567 + }, + { + "epoch": 0.41195833521215675, + "grad_norm": 1.7806477783502899, + "learning_rate": 2.6551396152814534e-06, + "loss": 0.9955, + "step": 4568 + }, + { + "epoch": 0.412048518735627, + "grad_norm": 1.8230043735294748, + "learning_rate": 2.65458760937595e-06, + "loss": 0.9873, + "step": 4569 + }, + { + "epoch": 0.41213870225909727, + "grad_norm": 6.6910981097869815, + "learning_rate": 2.654035547617423e-06, + "loss": 0.943, + "step": 4570 + }, + { + "epoch": 0.4122288857825675, + "grad_norm": 3.667315050075964, + "learning_rate": 2.653483430052976e-06, + "loss": 0.8966, + "step": 4571 + }, + { + "epoch": 0.4123190693060378, + "grad_norm": 2.333541917560506, + "learning_rate": 2.6529312567297197e-06, + "loss": 0.9811, + "step": 4572 + }, + { + "epoch": 0.412409252829508, + "grad_norm": 1.7719812674167315, + "learning_rate": 2.652379027694768e-06, + "loss": 1.0734, + "step": 4573 + }, + { + "epoch": 0.4124994363529783, + "grad_norm": 1.943754437800306, + "learning_rate": 2.651826742995241e-06, + "loss": 1.0522, + "step": 4574 + }, + { + "epoch": 0.4125896198764486, + "grad_norm": 1.5214680779775458, + "learning_rate": 2.651274402678262e-06, + "loss": 0.9921, + "step": 4575 + }, + { + "epoch": 0.41267980339991883, + "grad_norm": 1.7238530091288293, + "learning_rate": 2.6507220067909597e-06, + "loss": 1.0232, + "step": 4576 + }, + { + "epoch": 0.4127699869233891, + "grad_norm": 1.7057866987082142, + "learning_rate": 2.650169555380468e-06, + "loss": 0.9524, + "step": 4577 + }, + { + "epoch": 0.41286017044685935, + "grad_norm": 2.4669057983609872, + "learning_rate": 2.6496170484939253e-06, + "loss": 0.9868, + "step": 4578 + }, + { + "epoch": 0.41295035397032964, + "grad_norm": 1.5889538423041187, + "learning_rate": 2.6490644861784735e-06, + "loss": 1.007, + "step": 4579 + }, + { + "epoch": 0.4130405374937999, + "grad_norm": 1.5957245656024, + "learning_rate": 2.648511868481261e-06, + "loss": 1.016, + "step": 4580 + }, + { + "epoch": 0.41313072101727016, + "grad_norm": 2.1858985488398845, + "learning_rate": 2.6479591954494397e-06, + "loss": 0.9925, + "step": 4581 + }, + { + "epoch": 0.4132209045407404, + "grad_norm": 1.9024134942337332, + "learning_rate": 2.647406467130167e-06, + "loss": 0.8906, + "step": 4582 + }, + { + "epoch": 0.4133110880642107, + "grad_norm": 1.9640898879394202, + "learning_rate": 2.646853683570605e-06, + "loss": 1.0149, + "step": 4583 + }, + { + "epoch": 0.4134012715876809, + "grad_norm": 1.6995993001714222, + "learning_rate": 2.6463008448179196e-06, + "loss": 0.9348, + "step": 4584 + }, + { + "epoch": 0.4134914551111512, + "grad_norm": 1.7434838095225842, + "learning_rate": 2.6457479509192828e-06, + "loss": 0.9468, + "step": 4585 + }, + { + "epoch": 0.41358163863462144, + "grad_norm": 2.239250678855024, + "learning_rate": 2.645195001921871e-06, + "loss": 0.9887, + "step": 4586 + }, + { + "epoch": 0.41367182215809173, + "grad_norm": 2.502368472641444, + "learning_rate": 2.644641997872863e-06, + "loss": 0.9733, + "step": 4587 + }, + { + "epoch": 0.41376200568156196, + "grad_norm": 2.163081197970259, + "learning_rate": 2.644088938819445e-06, + "loss": 0.9956, + "step": 4588 + }, + { + "epoch": 0.41385218920503225, + "grad_norm": 1.6313895271817191, + "learning_rate": 2.6435358248088077e-06, + "loss": 1.0786, + "step": 4589 + }, + { + "epoch": 0.4139423727285025, + "grad_norm": 1.5825215651261615, + "learning_rate": 2.642982655888146e-06, + "loss": 1.0362, + "step": 4590 + }, + { + "epoch": 0.41403255625197277, + "grad_norm": 1.7584760303658784, + "learning_rate": 2.6424294321046585e-06, + "loss": 0.9572, + "step": 4591 + }, + { + "epoch": 0.414122739775443, + "grad_norm": 0.7302861929037392, + "learning_rate": 2.641876153505549e-06, + "loss": 0.8111, + "step": 4592 + }, + { + "epoch": 0.4142129232989133, + "grad_norm": 1.926065667681799, + "learning_rate": 2.641322820138027e-06, + "loss": 1.0009, + "step": 4593 + }, + { + "epoch": 0.4143031068223835, + "grad_norm": 1.7057885856082822, + "learning_rate": 2.640769432049306e-06, + "loss": 0.945, + "step": 4594 + }, + { + "epoch": 0.4143932903458538, + "grad_norm": 1.6592129529432686, + "learning_rate": 2.6402159892866038e-06, + "loss": 0.9158, + "step": 4595 + }, + { + "epoch": 0.41448347386932405, + "grad_norm": 6.522267859214596, + "learning_rate": 2.639662491897143e-06, + "loss": 0.9808, + "step": 4596 + }, + { + "epoch": 0.41457365739279434, + "grad_norm": 1.5849560486629393, + "learning_rate": 2.639108939928152e-06, + "loss": 1.0045, + "step": 4597 + }, + { + "epoch": 0.4146638409162646, + "grad_norm": 2.5168881295721124, + "learning_rate": 2.638555333426862e-06, + "loss": 1.0063, + "step": 4598 + }, + { + "epoch": 0.41475402443973486, + "grad_norm": 1.389758450837371, + "learning_rate": 2.6380016724405093e-06, + "loss": 1.0258, + "step": 4599 + }, + { + "epoch": 0.41484420796320515, + "grad_norm": 2.087578288911629, + "learning_rate": 2.637447957016336e-06, + "loss": 0.9525, + "step": 4600 + }, + { + "epoch": 0.4149343914866754, + "grad_norm": 2.095464203965923, + "learning_rate": 2.636894187201589e-06, + "loss": 1.0519, + "step": 4601 + }, + { + "epoch": 0.41502457501014567, + "grad_norm": 1.5417709143506524, + "learning_rate": 2.6363403630435176e-06, + "loss": 0.8963, + "step": 4602 + }, + { + "epoch": 0.4151147585336159, + "grad_norm": 1.5825574211803526, + "learning_rate": 2.635786484589378e-06, + "loss": 0.9867, + "step": 4603 + }, + { + "epoch": 0.4152049420570862, + "grad_norm": 1.8537137696422694, + "learning_rate": 2.63523255188643e-06, + "loss": 0.892, + "step": 4604 + }, + { + "epoch": 0.4152951255805564, + "grad_norm": 1.632914635781189, + "learning_rate": 2.6346785649819375e-06, + "loss": 1.017, + "step": 4605 + }, + { + "epoch": 0.4153853091040267, + "grad_norm": 1.5954034493550493, + "learning_rate": 2.6341245239231706e-06, + "loss": 1.0214, + "step": 4606 + }, + { + "epoch": 0.41547549262749695, + "grad_norm": 1.3102440747172743, + "learning_rate": 2.6335704287574024e-06, + "loss": 0.9956, + "step": 4607 + }, + { + "epoch": 0.41556567615096723, + "grad_norm": 1.750933738648628, + "learning_rate": 2.6330162795319124e-06, + "loss": 1.0628, + "step": 4608 + }, + { + "epoch": 0.41565585967443747, + "grad_norm": 2.001453110194473, + "learning_rate": 2.632462076293983e-06, + "loss": 1.0494, + "step": 4609 + }, + { + "epoch": 0.41574604319790776, + "grad_norm": 1.2563921093628283, + "learning_rate": 2.6319078190909017e-06, + "loss": 0.9914, + "step": 4610 + }, + { + "epoch": 0.415836226721378, + "grad_norm": 1.9032958857543008, + "learning_rate": 2.6313535079699606e-06, + "loss": 0.9064, + "step": 4611 + }, + { + "epoch": 0.4159264102448483, + "grad_norm": 1.7808785302164734, + "learning_rate": 2.6307991429784572e-06, + "loss": 0.9707, + "step": 4612 + }, + { + "epoch": 0.4160165937683185, + "grad_norm": 1.5745923377562998, + "learning_rate": 2.6302447241636924e-06, + "loss": 1.0544, + "step": 4613 + }, + { + "epoch": 0.4161067772917888, + "grad_norm": 1.7107105540079293, + "learning_rate": 2.629690251572973e-06, + "loss": 1.0702, + "step": 4614 + }, + { + "epoch": 0.41619696081525903, + "grad_norm": 1.6491511097476732, + "learning_rate": 2.629135725253609e-06, + "loss": 0.9712, + "step": 4615 + }, + { + "epoch": 0.4162871443387293, + "grad_norm": 1.6751866564016116, + "learning_rate": 2.6285811452529162e-06, + "loss": 0.981, + "step": 4616 + }, + { + "epoch": 0.41637732786219955, + "grad_norm": 1.430900017871626, + "learning_rate": 2.6280265116182136e-06, + "loss": 1.0121, + "step": 4617 + }, + { + "epoch": 0.41646751138566984, + "grad_norm": 1.6013178940970223, + "learning_rate": 2.6274718243968266e-06, + "loss": 0.9985, + "step": 4618 + }, + { + "epoch": 0.4165576949091401, + "grad_norm": 1.677902550755965, + "learning_rate": 2.626917083636084e-06, + "loss": 0.9137, + "step": 4619 + }, + { + "epoch": 0.41664787843261036, + "grad_norm": 1.4022334913511971, + "learning_rate": 2.6263622893833183e-06, + "loss": 0.928, + "step": 4620 + }, + { + "epoch": 0.4167380619560806, + "grad_norm": 1.4613797747415145, + "learning_rate": 2.625807441685869e-06, + "loss": 0.939, + "step": 4621 + }, + { + "epoch": 0.4168282454795509, + "grad_norm": 0.6134922612175495, + "learning_rate": 2.625252540591078e-06, + "loss": 0.799, + "step": 4622 + }, + { + "epoch": 0.4169184290030212, + "grad_norm": 2.104178689853071, + "learning_rate": 2.6246975861462927e-06, + "loss": 1.0664, + "step": 4623 + }, + { + "epoch": 0.4170086125264914, + "grad_norm": 2.0756588406950027, + "learning_rate": 2.624142578398864e-06, + "loss": 1.0367, + "step": 4624 + }, + { + "epoch": 0.4170987960499617, + "grad_norm": 1.42781269156141, + "learning_rate": 2.6235875173961498e-06, + "loss": 0.9721, + "step": 4625 + }, + { + "epoch": 0.41718897957343193, + "grad_norm": 1.463995196933893, + "learning_rate": 2.62303240318551e-06, + "loss": 0.9217, + "step": 4626 + }, + { + "epoch": 0.4172791630969022, + "grad_norm": 7.29997497972355, + "learning_rate": 2.62247723581431e-06, + "loss": 1.0673, + "step": 4627 + }, + { + "epoch": 0.41736934662037245, + "grad_norm": 1.6962038149410814, + "learning_rate": 2.62192201532992e-06, + "loss": 0.9952, + "step": 4628 + }, + { + "epoch": 0.41745953014384274, + "grad_norm": 2.6285653197916496, + "learning_rate": 2.6213667417797145e-06, + "loss": 1.0102, + "step": 4629 + }, + { + "epoch": 0.41754971366731297, + "grad_norm": 1.5331060876935578, + "learning_rate": 2.6208114152110725e-06, + "loss": 1.0204, + "step": 4630 + }, + { + "epoch": 0.41763989719078326, + "grad_norm": 5.7934112482717435, + "learning_rate": 2.6202560356713774e-06, + "loss": 1.0095, + "step": 4631 + }, + { + "epoch": 0.4177300807142535, + "grad_norm": 0.6410534286742356, + "learning_rate": 2.619700603208017e-06, + "loss": 0.8764, + "step": 4632 + }, + { + "epoch": 0.4178202642377238, + "grad_norm": 1.5641702498630514, + "learning_rate": 2.6191451178683842e-06, + "loss": 0.9089, + "step": 4633 + }, + { + "epoch": 0.417910447761194, + "grad_norm": 2.0342700515139507, + "learning_rate": 2.6185895796998764e-06, + "loss": 1.0239, + "step": 4634 + }, + { + "epoch": 0.4180006312846643, + "grad_norm": 1.7655416823452197, + "learning_rate": 2.6180339887498946e-06, + "loss": 1.0635, + "step": 4635 + }, + { + "epoch": 0.41809081480813454, + "grad_norm": 2.3403710995363367, + "learning_rate": 2.617478345065846e-06, + "loss": 0.9305, + "step": 4636 + }, + { + "epoch": 0.4181809983316048, + "grad_norm": 1.9985505454605395, + "learning_rate": 2.616922648695139e-06, + "loss": 1.014, + "step": 4637 + }, + { + "epoch": 0.41827118185507506, + "grad_norm": 1.8810244732389716, + "learning_rate": 2.61636689968519e-06, + "loss": 0.8921, + "step": 4638 + }, + { + "epoch": 0.41836136537854535, + "grad_norm": 3.8502264278263203, + "learning_rate": 2.6158110980834186e-06, + "loss": 1.0582, + "step": 4639 + }, + { + "epoch": 0.4184515489020156, + "grad_norm": 1.797086222300104, + "learning_rate": 2.615255243937249e-06, + "loss": 1.0461, + "step": 4640 + }, + { + "epoch": 0.41854173242548587, + "grad_norm": 2.054818616969639, + "learning_rate": 2.61469933729411e-06, + "loss": 1.0163, + "step": 4641 + }, + { + "epoch": 0.4186319159489561, + "grad_norm": 3.751738971593589, + "learning_rate": 2.614143378201433e-06, + "loss": 0.9718, + "step": 4642 + }, + { + "epoch": 0.4187220994724264, + "grad_norm": 1.577387420216039, + "learning_rate": 2.6135873667066567e-06, + "loss": 0.9733, + "step": 4643 + }, + { + "epoch": 0.4188122829958966, + "grad_norm": 1.8996127537064467, + "learning_rate": 2.613031302857224e-06, + "loss": 0.9546, + "step": 4644 + }, + { + "epoch": 0.4189024665193669, + "grad_norm": 1.6820673916092546, + "learning_rate": 2.6124751867005792e-06, + "loss": 1.021, + "step": 4645 + }, + { + "epoch": 0.4189926500428372, + "grad_norm": 1.3486618933832841, + "learning_rate": 2.611919018284175e-06, + "loss": 1.0448, + "step": 4646 + }, + { + "epoch": 0.41908283356630743, + "grad_norm": 1.6741194588344406, + "learning_rate": 2.611362797655466e-06, + "loss": 1.043, + "step": 4647 + }, + { + "epoch": 0.4191730170897777, + "grad_norm": 3.3239519346138335, + "learning_rate": 2.6108065248619124e-06, + "loss": 1.0525, + "step": 4648 + }, + { + "epoch": 0.41926320061324795, + "grad_norm": 1.3868888643181199, + "learning_rate": 2.610250199950978e-06, + "loss": 0.9287, + "step": 4649 + }, + { + "epoch": 0.41935338413671824, + "grad_norm": 3.282583201996157, + "learning_rate": 2.609693822970131e-06, + "loss": 0.9637, + "step": 4650 + }, + { + "epoch": 0.4194435676601885, + "grad_norm": 1.8559365415658964, + "learning_rate": 2.609137393966846e-06, + "loss": 0.9594, + "step": 4651 + }, + { + "epoch": 0.41953375118365877, + "grad_norm": 2.28431558388231, + "learning_rate": 2.6085809129886e-06, + "loss": 0.9492, + "step": 4652 + }, + { + "epoch": 0.419623934707129, + "grad_norm": 2.5553762458238802, + "learning_rate": 2.608024380082874e-06, + "loss": 1.0072, + "step": 4653 + }, + { + "epoch": 0.4197141182305993, + "grad_norm": 1.6983424239933895, + "learning_rate": 2.6074677952971554e-06, + "loss": 0.9944, + "step": 4654 + }, + { + "epoch": 0.4198043017540695, + "grad_norm": 1.5090600421606648, + "learning_rate": 2.606911158678935e-06, + "loss": 0.9152, + "step": 4655 + }, + { + "epoch": 0.4198944852775398, + "grad_norm": 4.459369764745963, + "learning_rate": 2.606354470275708e-06, + "loss": 0.9884, + "step": 4656 + }, + { + "epoch": 0.41998466880101004, + "grad_norm": 3.5005825102941013, + "learning_rate": 2.6057977301349744e-06, + "loss": 0.9613, + "step": 4657 + }, + { + "epoch": 0.42007485232448033, + "grad_norm": 1.6223774802290902, + "learning_rate": 2.6052409383042383e-06, + "loss": 0.9567, + "step": 4658 + }, + { + "epoch": 0.42016503584795056, + "grad_norm": 1.6184861870318556, + "learning_rate": 2.6046840948310074e-06, + "loss": 0.9714, + "step": 4659 + }, + { + "epoch": 0.42025521937142085, + "grad_norm": 1.5511274359741807, + "learning_rate": 2.6041271997627962e-06, + "loss": 0.9934, + "step": 4660 + }, + { + "epoch": 0.4203454028948911, + "grad_norm": 1.747297516362699, + "learning_rate": 2.6035702531471202e-06, + "loss": 0.9827, + "step": 4661 + }, + { + "epoch": 0.4204355864183614, + "grad_norm": 1.4240517355650377, + "learning_rate": 2.6030132550315035e-06, + "loss": 0.9894, + "step": 4662 + }, + { + "epoch": 0.4205257699418316, + "grad_norm": 0.636805943796679, + "learning_rate": 2.60245620546347e-06, + "loss": 0.8501, + "step": 4663 + }, + { + "epoch": 0.4206159534653019, + "grad_norm": 1.6955450926813824, + "learning_rate": 2.6018991044905517e-06, + "loss": 0.969, + "step": 4664 + }, + { + "epoch": 0.42070613698877213, + "grad_norm": 1.960746178279539, + "learning_rate": 2.6013419521602825e-06, + "loss": 1.0261, + "step": 4665 + }, + { + "epoch": 0.4207963205122424, + "grad_norm": 1.7383707430658426, + "learning_rate": 2.600784748520202e-06, + "loss": 1.1212, + "step": 4666 + }, + { + "epoch": 0.42088650403571265, + "grad_norm": 1.4643678426140256, + "learning_rate": 2.6002274936178544e-06, + "loss": 0.9294, + "step": 4667 + }, + { + "epoch": 0.42097668755918294, + "grad_norm": 1.7631912336920812, + "learning_rate": 2.5996701875007873e-06, + "loss": 1.0008, + "step": 4668 + }, + { + "epoch": 0.4210668710826532, + "grad_norm": 2.1480498362963196, + "learning_rate": 2.5991128302165533e-06, + "loss": 0.9834, + "step": 4669 + }, + { + "epoch": 0.42115705460612346, + "grad_norm": 1.787416592232553, + "learning_rate": 2.5985554218127094e-06, + "loss": 1.0551, + "step": 4670 + }, + { + "epoch": 0.42124723812959375, + "grad_norm": 4.95688291647893, + "learning_rate": 2.597997962336816e-06, + "loss": 0.9258, + "step": 4671 + }, + { + "epoch": 0.421337421653064, + "grad_norm": 2.1268756666695277, + "learning_rate": 2.5974404518364393e-06, + "loss": 0.9843, + "step": 4672 + }, + { + "epoch": 0.42142760517653427, + "grad_norm": 1.6490458590745167, + "learning_rate": 2.596882890359149e-06, + "loss": 1.112, + "step": 4673 + }, + { + "epoch": 0.4215177887000045, + "grad_norm": 2.250401249181602, + "learning_rate": 2.5963252779525196e-06, + "loss": 0.9109, + "step": 4674 + }, + { + "epoch": 0.4216079722234748, + "grad_norm": 1.855364344821664, + "learning_rate": 2.595767614664129e-06, + "loss": 0.9956, + "step": 4675 + }, + { + "epoch": 0.421698155746945, + "grad_norm": 1.556671640046457, + "learning_rate": 2.5952099005415607e-06, + "loss": 0.9694, + "step": 4676 + }, + { + "epoch": 0.4217883392704153, + "grad_norm": 1.7956506289529583, + "learning_rate": 2.594652135632402e-06, + "loss": 0.9546, + "step": 4677 + }, + { + "epoch": 0.42187852279388555, + "grad_norm": 1.3174130808843214, + "learning_rate": 2.594094319984244e-06, + "loss": 0.9479, + "step": 4678 + }, + { + "epoch": 0.42196870631735584, + "grad_norm": 1.8019036955042032, + "learning_rate": 2.5935364536446825e-06, + "loss": 1.0518, + "step": 4679 + }, + { + "epoch": 0.42205888984082607, + "grad_norm": 1.9118480718137296, + "learning_rate": 2.5929785366613185e-06, + "loss": 0.9946, + "step": 4680 + }, + { + "epoch": 0.42214907336429636, + "grad_norm": 1.702619775152287, + "learning_rate": 2.592420569081756e-06, + "loss": 1.0287, + "step": 4681 + }, + { + "epoch": 0.4222392568877666, + "grad_norm": 1.7965908986937698, + "learning_rate": 2.5918625509536037e-06, + "loss": 0.9513, + "step": 4682 + }, + { + "epoch": 0.4223294404112369, + "grad_norm": 0.6174005309775182, + "learning_rate": 2.591304482324475e-06, + "loss": 0.8151, + "step": 4683 + }, + { + "epoch": 0.4224196239347071, + "grad_norm": 1.6325056892050531, + "learning_rate": 2.5907463632419878e-06, + "loss": 1.0064, + "step": 4684 + }, + { + "epoch": 0.4225098074581774, + "grad_norm": 4.007964549610265, + "learning_rate": 2.5901881937537632e-06, + "loss": 0.958, + "step": 4685 + }, + { + "epoch": 0.42259999098164763, + "grad_norm": 0.7008901258455328, + "learning_rate": 2.589629973907428e-06, + "loss": 0.849, + "step": 4686 + }, + { + "epoch": 0.4226901745051179, + "grad_norm": 2.0294995321464517, + "learning_rate": 2.589071703750612e-06, + "loss": 0.9426, + "step": 4687 + }, + { + "epoch": 0.42278035802858815, + "grad_norm": 1.3629240863395553, + "learning_rate": 2.5885133833309504e-06, + "loss": 1.0306, + "step": 4688 + }, + { + "epoch": 0.42287054155205844, + "grad_norm": 1.486416667451406, + "learning_rate": 2.5879550126960814e-06, + "loss": 1.0705, + "step": 4689 + }, + { + "epoch": 0.4229607250755287, + "grad_norm": 1.3843278932844951, + "learning_rate": 2.5873965918936494e-06, + "loss": 1.0476, + "step": 4690 + }, + { + "epoch": 0.42305090859899896, + "grad_norm": 2.257647551216783, + "learning_rate": 2.586838120971301e-06, + "loss": 1.0193, + "step": 4691 + }, + { + "epoch": 0.4231410921224692, + "grad_norm": 1.713981881067037, + "learning_rate": 2.586279599976689e-06, + "loss": 1.0545, + "step": 4692 + }, + { + "epoch": 0.4232312756459395, + "grad_norm": 1.6501433599210784, + "learning_rate": 2.585721028957468e-06, + "loss": 0.9902, + "step": 4693 + }, + { + "epoch": 0.4233214591694098, + "grad_norm": 1.6101136225536876, + "learning_rate": 2.585162407961299e-06, + "loss": 1.0044, + "step": 4694 + }, + { + "epoch": 0.42341164269288, + "grad_norm": 1.7549523035724734, + "learning_rate": 2.584603737035847e-06, + "loss": 0.9333, + "step": 4695 + }, + { + "epoch": 0.4235018262163503, + "grad_norm": 1.7698978271736572, + "learning_rate": 2.5840450162287806e-06, + "loss": 0.9888, + "step": 4696 + }, + { + "epoch": 0.42359200973982053, + "grad_norm": 1.9516898413741917, + "learning_rate": 2.583486245587774e-06, + "loss": 0.9042, + "step": 4697 + }, + { + "epoch": 0.4236821932632908, + "grad_norm": 1.6680799532053745, + "learning_rate": 2.5829274251605023e-06, + "loss": 1.1006, + "step": 4698 + }, + { + "epoch": 0.42377237678676105, + "grad_norm": 1.4962618661153424, + "learning_rate": 2.582368554994649e-06, + "loss": 0.9685, + "step": 4699 + }, + { + "epoch": 0.42386256031023134, + "grad_norm": 0.7480279905161585, + "learning_rate": 2.5818096351378994e-06, + "loss": 0.8499, + "step": 4700 + }, + { + "epoch": 0.4239527438337016, + "grad_norm": 1.8437198701110042, + "learning_rate": 2.5812506656379435e-06, + "loss": 1.0481, + "step": 4701 + }, + { + "epoch": 0.42404292735717186, + "grad_norm": 1.675232270491234, + "learning_rate": 2.580691646542476e-06, + "loss": 0.9446, + "step": 4702 + }, + { + "epoch": 0.4241331108806421, + "grad_norm": 1.4978703639797466, + "learning_rate": 2.5801325778991958e-06, + "loss": 1.0286, + "step": 4703 + }, + { + "epoch": 0.4242232944041124, + "grad_norm": 2.685158707742729, + "learning_rate": 2.5795734597558043e-06, + "loss": 0.847, + "step": 4704 + }, + { + "epoch": 0.4243134779275826, + "grad_norm": 2.413103326652238, + "learning_rate": 2.579014292160011e-06, + "loss": 0.9928, + "step": 4705 + }, + { + "epoch": 0.4244036614510529, + "grad_norm": 1.4210953304036618, + "learning_rate": 2.5784550751595236e-06, + "loss": 0.9751, + "step": 4706 + }, + { + "epoch": 0.42449384497452314, + "grad_norm": 1.372195027069457, + "learning_rate": 2.577895808802061e-06, + "loss": 1.0547, + "step": 4707 + }, + { + "epoch": 0.4245840284979934, + "grad_norm": 2.5775091187013337, + "learning_rate": 2.577336493135341e-06, + "loss": 0.9086, + "step": 4708 + }, + { + "epoch": 0.42467421202146366, + "grad_norm": 1.277956241201777, + "learning_rate": 2.576777128207088e-06, + "loss": 1.0882, + "step": 4709 + }, + { + "epoch": 0.42476439554493395, + "grad_norm": 1.7376441731087717, + "learning_rate": 2.5762177140650306e-06, + "loss": 0.9528, + "step": 4710 + }, + { + "epoch": 0.4248545790684042, + "grad_norm": 1.863022312429977, + "learning_rate": 2.5756582507569003e-06, + "loss": 1.0009, + "step": 4711 + }, + { + "epoch": 0.42494476259187447, + "grad_norm": 1.4964870005768536, + "learning_rate": 2.5750987383304335e-06, + "loss": 1.018, + "step": 4712 + }, + { + "epoch": 0.4250349461153447, + "grad_norm": 3.9216583313686524, + "learning_rate": 2.574539176833372e-06, + "loss": 0.9814, + "step": 4713 + }, + { + "epoch": 0.425125129638815, + "grad_norm": 1.9340627390518816, + "learning_rate": 2.5739795663134594e-06, + "loss": 0.9355, + "step": 4714 + }, + { + "epoch": 0.4252153131622852, + "grad_norm": 1.894933724666136, + "learning_rate": 2.5734199068184454e-06, + "loss": 1.0791, + "step": 4715 + }, + { + "epoch": 0.4253054966857555, + "grad_norm": 1.7121941119249302, + "learning_rate": 2.572860198396083e-06, + "loss": 0.9201, + "step": 4716 + }, + { + "epoch": 0.4253956802092258, + "grad_norm": 1.678764551004503, + "learning_rate": 2.57230044109413e-06, + "loss": 1.0173, + "step": 4717 + }, + { + "epoch": 0.42548586373269603, + "grad_norm": 2.201180215191886, + "learning_rate": 2.5717406349603483e-06, + "loss": 0.9723, + "step": 4718 + }, + { + "epoch": 0.4255760472561663, + "grad_norm": 1.695583831591055, + "learning_rate": 2.5711807800425026e-06, + "loss": 1.01, + "step": 4719 + }, + { + "epoch": 0.42566623077963656, + "grad_norm": 3.2497339873387108, + "learning_rate": 2.5706208763883633e-06, + "loss": 0.8296, + "step": 4720 + }, + { + "epoch": 0.42575641430310684, + "grad_norm": 1.7935746775155679, + "learning_rate": 2.570060924045704e-06, + "loss": 0.9474, + "step": 4721 + }, + { + "epoch": 0.4258465978265771, + "grad_norm": 18.436738596566503, + "learning_rate": 2.569500923062304e-06, + "loss": 1.0264, + "step": 4722 + }, + { + "epoch": 0.42593678135004737, + "grad_norm": 1.5593626186208032, + "learning_rate": 2.5689408734859445e-06, + "loss": 0.9144, + "step": 4723 + }, + { + "epoch": 0.4260269648735176, + "grad_norm": 2.3191375326322654, + "learning_rate": 2.5683807753644127e-06, + "loss": 0.9409, + "step": 4724 + }, + { + "epoch": 0.4261171483969879, + "grad_norm": 1.5460728578857044, + "learning_rate": 2.5678206287454996e-06, + "loss": 0.9704, + "step": 4725 + }, + { + "epoch": 0.4262073319204581, + "grad_norm": 2.5030860926429632, + "learning_rate": 2.567260433676999e-06, + "loss": 1.0309, + "step": 4726 + }, + { + "epoch": 0.4262975154439284, + "grad_norm": 1.8128594009284862, + "learning_rate": 2.5667001902067107e-06, + "loss": 0.9648, + "step": 4727 + }, + { + "epoch": 0.42638769896739864, + "grad_norm": 1.894522063242868, + "learning_rate": 2.566139898382437e-06, + "loss": 1.042, + "step": 4728 + }, + { + "epoch": 0.42647788249086893, + "grad_norm": 1.8049467486824158, + "learning_rate": 2.5655795582519853e-06, + "loss": 1.005, + "step": 4729 + }, + { + "epoch": 0.42656806601433916, + "grad_norm": 1.5300122143376302, + "learning_rate": 2.565019169863168e-06, + "loss": 1.0632, + "step": 4730 + }, + { + "epoch": 0.42665824953780945, + "grad_norm": 0.5837290278962078, + "learning_rate": 2.5644587332637994e-06, + "loss": 0.7845, + "step": 4731 + }, + { + "epoch": 0.4267484330612797, + "grad_norm": 1.5436007593580574, + "learning_rate": 2.5638982485016994e-06, + "loss": 1.0233, + "step": 4732 + }, + { + "epoch": 0.42683861658475, + "grad_norm": 0.7281631148680909, + "learning_rate": 2.5633377156246917e-06, + "loss": 0.7897, + "step": 4733 + }, + { + "epoch": 0.4269288001082202, + "grad_norm": 1.7224266262510586, + "learning_rate": 2.562777134680603e-06, + "loss": 0.9793, + "step": 4734 + }, + { + "epoch": 0.4270189836316905, + "grad_norm": 1.969965120771656, + "learning_rate": 2.562216505717267e-06, + "loss": 1.0004, + "step": 4735 + }, + { + "epoch": 0.42710916715516073, + "grad_norm": 1.7525764980843412, + "learning_rate": 2.561655828782518e-06, + "loss": 1.018, + "step": 4736 + }, + { + "epoch": 0.427199350678631, + "grad_norm": 6.9591726683935375, + "learning_rate": 2.561095103924197e-06, + "loss": 0.972, + "step": 4737 + }, + { + "epoch": 0.42728953420210125, + "grad_norm": 1.5106502571567093, + "learning_rate": 2.560534331190148e-06, + "loss": 0.9367, + "step": 4738 + }, + { + "epoch": 0.42737971772557154, + "grad_norm": 0.5979013376785323, + "learning_rate": 2.559973510628218e-06, + "loss": 0.786, + "step": 4739 + }, + { + "epoch": 0.4274699012490418, + "grad_norm": 2.2627701234530644, + "learning_rate": 2.5594126422862615e-06, + "loss": 0.9431, + "step": 4740 + }, + { + "epoch": 0.42756008477251206, + "grad_norm": 2.564160808977642, + "learning_rate": 2.558851726212134e-06, + "loss": 0.94, + "step": 4741 + }, + { + "epoch": 0.42765026829598235, + "grad_norm": 1.6862423766962769, + "learning_rate": 2.5582907624536953e-06, + "loss": 0.9657, + "step": 4742 + }, + { + "epoch": 0.4277404518194526, + "grad_norm": 1.9893699917743717, + "learning_rate": 2.557729751058811e-06, + "loss": 1.005, + "step": 4743 + }, + { + "epoch": 0.42783063534292287, + "grad_norm": 2.268934476724255, + "learning_rate": 2.557168692075348e-06, + "loss": 0.9664, + "step": 4744 + }, + { + "epoch": 0.4279208188663931, + "grad_norm": 3.135884213149269, + "learning_rate": 2.556607585551181e-06, + "loss": 0.934, + "step": 4745 + }, + { + "epoch": 0.4280110023898634, + "grad_norm": 0.726807460713973, + "learning_rate": 2.5560464315341844e-06, + "loss": 0.9151, + "step": 4746 + }, + { + "epoch": 0.4281011859133336, + "grad_norm": 3.6285256965424333, + "learning_rate": 2.555485230072242e-06, + "loss": 0.9839, + "step": 4747 + }, + { + "epoch": 0.4281913694368039, + "grad_norm": 1.8224258270659084, + "learning_rate": 2.5549239812132354e-06, + "loss": 1.0356, + "step": 4748 + }, + { + "epoch": 0.42828155296027415, + "grad_norm": 1.9052174382507552, + "learning_rate": 2.5543626850050556e-06, + "loss": 1.0592, + "step": 4749 + }, + { + "epoch": 0.42837173648374444, + "grad_norm": 1.8149167095861412, + "learning_rate": 2.5538013414955944e-06, + "loss": 1.043, + "step": 4750 + }, + { + "epoch": 0.42846192000721467, + "grad_norm": 1.7144173546043078, + "learning_rate": 2.5532399507327494e-06, + "loss": 0.9453, + "step": 4751 + }, + { + "epoch": 0.42855210353068496, + "grad_norm": 0.6018104537626408, + "learning_rate": 2.552678512764421e-06, + "loss": 0.8228, + "step": 4752 + }, + { + "epoch": 0.4286422870541552, + "grad_norm": 2.109143505815809, + "learning_rate": 2.5521170276385147e-06, + "loss": 1.0481, + "step": 4753 + }, + { + "epoch": 0.4287324705776255, + "grad_norm": 1.7343833167074316, + "learning_rate": 2.5515554954029394e-06, + "loss": 1.0427, + "step": 4754 + }, + { + "epoch": 0.4288226541010957, + "grad_norm": 1.5388519607838345, + "learning_rate": 2.550993916105608e-06, + "loss": 1.0216, + "step": 4755 + }, + { + "epoch": 0.428912837624566, + "grad_norm": 1.8899598015441061, + "learning_rate": 2.550432289794437e-06, + "loss": 0.9473, + "step": 4756 + }, + { + "epoch": 0.42900302114803623, + "grad_norm": 1.766546363453203, + "learning_rate": 2.5498706165173483e-06, + "loss": 1.0007, + "step": 4757 + }, + { + "epoch": 0.4290932046715065, + "grad_norm": 1.4430973708199597, + "learning_rate": 2.5493088963222668e-06, + "loss": 0.9688, + "step": 4758 + }, + { + "epoch": 0.42918338819497676, + "grad_norm": 1.6001197084706373, + "learning_rate": 2.548747129257121e-06, + "loss": 1.0394, + "step": 4759 + }, + { + "epoch": 0.42927357171844704, + "grad_norm": 1.4294129822041801, + "learning_rate": 2.548185315369845e-06, + "loss": 0.9109, + "step": 4760 + }, + { + "epoch": 0.4293637552419173, + "grad_norm": 1.6962326997958415, + "learning_rate": 2.5476234547083746e-06, + "loss": 1.0798, + "step": 4761 + }, + { + "epoch": 0.42945393876538757, + "grad_norm": 2.0452038881097425, + "learning_rate": 2.547061547320652e-06, + "loss": 1.0665, + "step": 4762 + }, + { + "epoch": 0.4295441222888578, + "grad_norm": 1.303806075721648, + "learning_rate": 2.5464995932546217e-06, + "loss": 0.9742, + "step": 4763 + }, + { + "epoch": 0.4296343058123281, + "grad_norm": 1.834699172533655, + "learning_rate": 2.545937592558232e-06, + "loss": 0.9435, + "step": 4764 + }, + { + "epoch": 0.4297244893357984, + "grad_norm": 1.497559310171617, + "learning_rate": 2.5453755452794374e-06, + "loss": 0.9399, + "step": 4765 + }, + { + "epoch": 0.4298146728592686, + "grad_norm": 2.651568876853383, + "learning_rate": 2.5448134514661938e-06, + "loss": 1.0321, + "step": 4766 + }, + { + "epoch": 0.4299048563827389, + "grad_norm": 1.4862015578810948, + "learning_rate": 2.5442513111664623e-06, + "loss": 0.987, + "step": 4767 + }, + { + "epoch": 0.42999503990620913, + "grad_norm": 2.853228523908233, + "learning_rate": 2.5436891244282084e-06, + "loss": 0.9841, + "step": 4768 + }, + { + "epoch": 0.4300852234296794, + "grad_norm": 1.6633225349401926, + "learning_rate": 2.5431268912994004e-06, + "loss": 1.0927, + "step": 4769 + }, + { + "epoch": 0.43017540695314965, + "grad_norm": 1.8632520117554987, + "learning_rate": 2.5425646118280108e-06, + "loss": 0.9598, + "step": 4770 + }, + { + "epoch": 0.43026559047661994, + "grad_norm": 0.5492983938019022, + "learning_rate": 2.5420022860620172e-06, + "loss": 0.7993, + "step": 4771 + }, + { + "epoch": 0.4303557740000902, + "grad_norm": 1.9256406040172303, + "learning_rate": 2.5414399140493995e-06, + "loss": 0.9235, + "step": 4772 + }, + { + "epoch": 0.43044595752356046, + "grad_norm": 2.332266790868717, + "learning_rate": 2.5408774958381436e-06, + "loss": 0.93, + "step": 4773 + }, + { + "epoch": 0.4305361410470307, + "grad_norm": 1.6173149464094854, + "learning_rate": 2.540315031476237e-06, + "loss": 1.0236, + "step": 4774 + }, + { + "epoch": 0.430626324570501, + "grad_norm": 3.2345802131456622, + "learning_rate": 2.5397525210116737e-06, + "loss": 0.9993, + "step": 4775 + }, + { + "epoch": 0.4307165080939712, + "grad_norm": 1.4109256156682795, + "learning_rate": 2.539189964492448e-06, + "loss": 0.9474, + "step": 4776 + }, + { + "epoch": 0.4308066916174415, + "grad_norm": 1.586645334818972, + "learning_rate": 2.5386273619665613e-06, + "loss": 1.0306, + "step": 4777 + }, + { + "epoch": 0.43089687514091174, + "grad_norm": 0.7013321589723275, + "learning_rate": 2.5380647134820186e-06, + "loss": 0.8207, + "step": 4778 + }, + { + "epoch": 0.43098705866438203, + "grad_norm": 1.4140760726698849, + "learning_rate": 2.5375020190868277e-06, + "loss": 1.0441, + "step": 4779 + }, + { + "epoch": 0.43107724218785226, + "grad_norm": 1.6761780738049195, + "learning_rate": 2.536939278829001e-06, + "loss": 0.9851, + "step": 4780 + }, + { + "epoch": 0.43116742571132255, + "grad_norm": 1.8945491200027058, + "learning_rate": 2.5363764927565536e-06, + "loss": 1.0521, + "step": 4781 + }, + { + "epoch": 0.4312576092347928, + "grad_norm": 2.3702652566284192, + "learning_rate": 2.5358136609175064e-06, + "loss": 0.9416, + "step": 4782 + }, + { + "epoch": 0.43134779275826307, + "grad_norm": 1.4752625086321451, + "learning_rate": 2.535250783359884e-06, + "loss": 0.9925, + "step": 4783 + }, + { + "epoch": 0.4314379762817333, + "grad_norm": 1.5323618530685317, + "learning_rate": 2.5346878601317124e-06, + "loss": 0.8826, + "step": 4784 + }, + { + "epoch": 0.4315281598052036, + "grad_norm": 0.5827347516929227, + "learning_rate": 2.534124891281025e-06, + "loss": 0.7843, + "step": 4785 + }, + { + "epoch": 0.4316183433286738, + "grad_norm": 1.8431641892658177, + "learning_rate": 2.533561876855857e-06, + "loss": 1.0015, + "step": 4786 + }, + { + "epoch": 0.4317085268521441, + "grad_norm": 2.278928751502567, + "learning_rate": 2.532998816904247e-06, + "loss": 0.9442, + "step": 4787 + }, + { + "epoch": 0.4317987103756144, + "grad_norm": 1.9622623785695028, + "learning_rate": 2.53243571147424e-06, + "loss": 0.8811, + "step": 4788 + }, + { + "epoch": 0.43188889389908464, + "grad_norm": 2.045083812870822, + "learning_rate": 2.5318725606138815e-06, + "loss": 0.9029, + "step": 4789 + }, + { + "epoch": 0.4319790774225549, + "grad_norm": 2.315963316796558, + "learning_rate": 2.5313093643712235e-06, + "loss": 0.9979, + "step": 4790 + }, + { + "epoch": 0.43206926094602516, + "grad_norm": 1.540813785770774, + "learning_rate": 2.530746122794321e-06, + "loss": 1.0357, + "step": 4791 + }, + { + "epoch": 0.43215944446949545, + "grad_norm": 2.5519928792896, + "learning_rate": 2.5301828359312323e-06, + "loss": 0.9207, + "step": 4792 + }, + { + "epoch": 0.4322496279929657, + "grad_norm": 1.7223086190172874, + "learning_rate": 2.529619503830021e-06, + "loss": 0.9515, + "step": 4793 + }, + { + "epoch": 0.43233981151643597, + "grad_norm": 2.533239359953116, + "learning_rate": 2.529056126538753e-06, + "loss": 0.9222, + "step": 4794 + }, + { + "epoch": 0.4324299950399062, + "grad_norm": 1.6971249362945264, + "learning_rate": 2.5284927041054995e-06, + "loss": 1.0197, + "step": 4795 + }, + { + "epoch": 0.4325201785633765, + "grad_norm": 0.6118154563805102, + "learning_rate": 2.5279292365783348e-06, + "loss": 0.7622, + "step": 4796 + }, + { + "epoch": 0.4326103620868467, + "grad_norm": 1.5754790077227383, + "learning_rate": 2.527365724005336e-06, + "loss": 1.0823, + "step": 4797 + }, + { + "epoch": 0.432700545610317, + "grad_norm": 2.2809066514075718, + "learning_rate": 2.526802166434586e-06, + "loss": 1.0467, + "step": 4798 + }, + { + "epoch": 0.43279072913378724, + "grad_norm": 1.4361449779761488, + "learning_rate": 2.5262385639141708e-06, + "loss": 0.9552, + "step": 4799 + }, + { + "epoch": 0.43288091265725753, + "grad_norm": 1.823362488499018, + "learning_rate": 2.525674916492179e-06, + "loss": 1.0612, + "step": 4800 + }, + { + "epoch": 0.43297109618072777, + "grad_norm": 1.3440049838842218, + "learning_rate": 2.5251112242167056e-06, + "loss": 0.996, + "step": 4801 + }, + { + "epoch": 0.43306127970419805, + "grad_norm": 2.023241306057644, + "learning_rate": 2.5245474871358464e-06, + "loss": 0.9475, + "step": 4802 + }, + { + "epoch": 0.4331514632276683, + "grad_norm": 1.7322627083687547, + "learning_rate": 2.5239837052977037e-06, + "loss": 0.9087, + "step": 4803 + }, + { + "epoch": 0.4332416467511386, + "grad_norm": 2.1772060694312585, + "learning_rate": 2.523419878750381e-06, + "loss": 1.1026, + "step": 4804 + }, + { + "epoch": 0.4333318302746088, + "grad_norm": 1.7602088700246006, + "learning_rate": 2.522856007541989e-06, + "loss": 0.9405, + "step": 4805 + }, + { + "epoch": 0.4334220137980791, + "grad_norm": 1.4814665559617286, + "learning_rate": 2.5222920917206397e-06, + "loss": 0.8997, + "step": 4806 + }, + { + "epoch": 0.43351219732154933, + "grad_norm": 0.6230989631715222, + "learning_rate": 2.5217281313344493e-06, + "loss": 0.7996, + "step": 4807 + }, + { + "epoch": 0.4336023808450196, + "grad_norm": 1.9238837684291243, + "learning_rate": 2.5211641264315372e-06, + "loss": 1.0334, + "step": 4808 + }, + { + "epoch": 0.43369256436848985, + "grad_norm": 4.1121283180610035, + "learning_rate": 2.5206000770600286e-06, + "loss": 0.9693, + "step": 4809 + }, + { + "epoch": 0.43378274789196014, + "grad_norm": 1.901594386464244, + "learning_rate": 2.520035983268051e-06, + "loss": 0.9474, + "step": 4810 + }, + { + "epoch": 0.4338729314154304, + "grad_norm": 1.6038485702942584, + "learning_rate": 2.5194718451037357e-06, + "loss": 0.9924, + "step": 4811 + }, + { + "epoch": 0.43396311493890066, + "grad_norm": 1.6558093078488054, + "learning_rate": 2.518907662615218e-06, + "loss": 1.0156, + "step": 4812 + }, + { + "epoch": 0.43405329846237095, + "grad_norm": 3.0428772350648297, + "learning_rate": 2.5183434358506373e-06, + "loss": 0.95, + "step": 4813 + }, + { + "epoch": 0.4341434819858412, + "grad_norm": 1.7495571666098366, + "learning_rate": 2.5177791648581368e-06, + "loss": 0.9726, + "step": 4814 + }, + { + "epoch": 0.4342336655093115, + "grad_norm": 1.469084762407804, + "learning_rate": 2.517214849685863e-06, + "loss": 0.9932, + "step": 4815 + }, + { + "epoch": 0.4343238490327817, + "grad_norm": 1.5680110634880569, + "learning_rate": 2.5166504903819663e-06, + "loss": 1.0074, + "step": 4816 + }, + { + "epoch": 0.434414032556252, + "grad_norm": 1.7943160702373786, + "learning_rate": 2.5160860869946014e-06, + "loss": 1.0593, + "step": 4817 + }, + { + "epoch": 0.4345042160797222, + "grad_norm": 1.5244071808781252, + "learning_rate": 2.5155216395719253e-06, + "loss": 1.0712, + "step": 4818 + }, + { + "epoch": 0.4345943996031925, + "grad_norm": 1.9752479976966906, + "learning_rate": 2.5149571481621e-06, + "loss": 0.9507, + "step": 4819 + }, + { + "epoch": 0.43468458312666275, + "grad_norm": 0.6020379107631793, + "learning_rate": 2.514392612813292e-06, + "loss": 0.7818, + "step": 4820 + }, + { + "epoch": 0.43477476665013304, + "grad_norm": 1.4105045807719094, + "learning_rate": 2.5138280335736695e-06, + "loss": 1.0439, + "step": 4821 + }, + { + "epoch": 0.43486495017360327, + "grad_norm": 2.0442716626856066, + "learning_rate": 2.5132634104914064e-06, + "loss": 1.0707, + "step": 4822 + }, + { + "epoch": 0.43495513369707356, + "grad_norm": 1.3628599723764838, + "learning_rate": 2.5126987436146794e-06, + "loss": 1.0028, + "step": 4823 + }, + { + "epoch": 0.4350453172205438, + "grad_norm": 1.997171308952912, + "learning_rate": 2.5121340329916675e-06, + "loss": 0.9687, + "step": 4824 + }, + { + "epoch": 0.4351355007440141, + "grad_norm": 1.510625872976724, + "learning_rate": 2.5115692786705566e-06, + "loss": 0.9125, + "step": 4825 + }, + { + "epoch": 0.4352256842674843, + "grad_norm": 2.224318273452837, + "learning_rate": 2.511004480699534e-06, + "loss": 1.0225, + "step": 4826 + }, + { + "epoch": 0.4353158677909546, + "grad_norm": 1.8960431859614923, + "learning_rate": 2.510439639126791e-06, + "loss": 0.8563, + "step": 4827 + }, + { + "epoch": 0.43540605131442484, + "grad_norm": 1.4982859036608756, + "learning_rate": 2.509874754000524e-06, + "loss": 0.9937, + "step": 4828 + }, + { + "epoch": 0.4354962348378951, + "grad_norm": 1.7494614998766853, + "learning_rate": 2.509309825368932e-06, + "loss": 0.9769, + "step": 4829 + }, + { + "epoch": 0.43558641836136536, + "grad_norm": 1.490006535662968, + "learning_rate": 2.5087448532802173e-06, + "loss": 0.9901, + "step": 4830 + }, + { + "epoch": 0.43567660188483565, + "grad_norm": 1.9168624639707537, + "learning_rate": 2.508179837782586e-06, + "loss": 0.9721, + "step": 4831 + }, + { + "epoch": 0.4357667854083059, + "grad_norm": 1.5314302143782523, + "learning_rate": 2.5076147789242493e-06, + "loss": 0.9588, + "step": 4832 + }, + { + "epoch": 0.43585696893177617, + "grad_norm": 2.114788991263568, + "learning_rate": 2.5070496767534202e-06, + "loss": 1.0025, + "step": 4833 + }, + { + "epoch": 0.4359471524552464, + "grad_norm": 2.0680013171990512, + "learning_rate": 2.506484531318317e-06, + "loss": 1.0153, + "step": 4834 + }, + { + "epoch": 0.4360373359787167, + "grad_norm": 0.7449561828810398, + "learning_rate": 2.5059193426671613e-06, + "loss": 0.8282, + "step": 4835 + }, + { + "epoch": 0.436127519502187, + "grad_norm": 1.9242452256444673, + "learning_rate": 2.5053541108481772e-06, + "loss": 1.1261, + "step": 4836 + }, + { + "epoch": 0.4362177030256572, + "grad_norm": 1.9201877289627773, + "learning_rate": 2.5047888359095935e-06, + "loss": 0.9708, + "step": 4837 + }, + { + "epoch": 0.4363078865491275, + "grad_norm": 1.4819843516070617, + "learning_rate": 2.5042235178996436e-06, + "loss": 0.9768, + "step": 4838 + }, + { + "epoch": 0.43639807007259773, + "grad_norm": 1.8247366911223386, + "learning_rate": 2.5036581568665627e-06, + "loss": 1.0521, + "step": 4839 + }, + { + "epoch": 0.436488253596068, + "grad_norm": 1.552713185291736, + "learning_rate": 2.503092752858591e-06, + "loss": 1.0542, + "step": 4840 + }, + { + "epoch": 0.43657843711953825, + "grad_norm": 0.9539979782802824, + "learning_rate": 2.502527305923971e-06, + "loss": 0.8448, + "step": 4841 + }, + { + "epoch": 0.43666862064300854, + "grad_norm": 1.8984071077653404, + "learning_rate": 2.5019618161109506e-06, + "loss": 0.9775, + "step": 4842 + }, + { + "epoch": 0.4367588041664788, + "grad_norm": 1.8115514706810054, + "learning_rate": 2.5013962834677804e-06, + "loss": 1.0073, + "step": 4843 + }, + { + "epoch": 0.43684898768994906, + "grad_norm": 2.1678864884449616, + "learning_rate": 2.500830708042715e-06, + "loss": 0.9775, + "step": 4844 + }, + { + "epoch": 0.4369391712134193, + "grad_norm": 1.6130006560000878, + "learning_rate": 2.500265089884011e-06, + "loss": 1.0559, + "step": 4845 + }, + { + "epoch": 0.4370293547368896, + "grad_norm": 2.723394613973375, + "learning_rate": 2.499699429039932e-06, + "loss": 0.9882, + "step": 4846 + }, + { + "epoch": 0.4371195382603598, + "grad_norm": 1.70774110747803, + "learning_rate": 2.4991337255587425e-06, + "loss": 0.9395, + "step": 4847 + }, + { + "epoch": 0.4372097217838301, + "grad_norm": 1.6076358916893472, + "learning_rate": 2.4985679794887106e-06, + "loss": 0.8576, + "step": 4848 + }, + { + "epoch": 0.43729990530730034, + "grad_norm": 2.019103958703223, + "learning_rate": 2.49800219087811e-06, + "loss": 1.0238, + "step": 4849 + }, + { + "epoch": 0.43739008883077063, + "grad_norm": 1.889953367879104, + "learning_rate": 2.4974363597752163e-06, + "loss": 1.0257, + "step": 4850 + }, + { + "epoch": 0.43748027235424086, + "grad_norm": 1.982411411802745, + "learning_rate": 2.4968704862283097e-06, + "loss": 1.0359, + "step": 4851 + }, + { + "epoch": 0.43757045587771115, + "grad_norm": 1.4571290175130647, + "learning_rate": 2.4963045702856737e-06, + "loss": 1.0606, + "step": 4852 + }, + { + "epoch": 0.4376606394011814, + "grad_norm": 1.4037269422585528, + "learning_rate": 2.4957386119955946e-06, + "loss": 0.933, + "step": 4853 + }, + { + "epoch": 0.4377508229246517, + "grad_norm": 0.6699642468278917, + "learning_rate": 2.495172611406364e-06, + "loss": 0.8054, + "step": 4854 + }, + { + "epoch": 0.4378410064481219, + "grad_norm": 2.0276730073429077, + "learning_rate": 2.4946065685662757e-06, + "loss": 1.0176, + "step": 4855 + }, + { + "epoch": 0.4379311899715922, + "grad_norm": 2.021249538178589, + "learning_rate": 2.4940404835236283e-06, + "loss": 0.9908, + "step": 4856 + }, + { + "epoch": 0.4380213734950624, + "grad_norm": 1.6682188753288394, + "learning_rate": 2.4934743563267223e-06, + "loss": 0.9806, + "step": 4857 + }, + { + "epoch": 0.4381115570185327, + "grad_norm": 2.3176019422707994, + "learning_rate": 2.4929081870238635e-06, + "loss": 0.9539, + "step": 4858 + }, + { + "epoch": 0.43820174054200295, + "grad_norm": 1.758166197054432, + "learning_rate": 2.49234197566336e-06, + "loss": 1.0269, + "step": 4859 + }, + { + "epoch": 0.43829192406547324, + "grad_norm": 1.5792680132443828, + "learning_rate": 2.4917757222935247e-06, + "loss": 1.0049, + "step": 4860 + }, + { + "epoch": 0.4383821075889435, + "grad_norm": 1.9991449077843122, + "learning_rate": 2.4912094269626725e-06, + "loss": 1.0084, + "step": 4861 + }, + { + "epoch": 0.43847229111241376, + "grad_norm": 2.42850733119462, + "learning_rate": 2.4906430897191245e-06, + "loss": 1.0293, + "step": 4862 + }, + { + "epoch": 0.43856247463588405, + "grad_norm": 1.9796172649687436, + "learning_rate": 2.490076710611202e-06, + "loss": 1.0286, + "step": 4863 + }, + { + "epoch": 0.4386526581593543, + "grad_norm": 2.2500638952719534, + "learning_rate": 2.4895102896872326e-06, + "loss": 1.0397, + "step": 4864 + }, + { + "epoch": 0.43874284168282457, + "grad_norm": 2.0560414309947155, + "learning_rate": 2.4889438269955457e-06, + "loss": 0.9241, + "step": 4865 + }, + { + "epoch": 0.4388330252062948, + "grad_norm": 1.7369999996310483, + "learning_rate": 2.4883773225844755e-06, + "loss": 0.9708, + "step": 4866 + }, + { + "epoch": 0.4389232087297651, + "grad_norm": 2.2686317223858055, + "learning_rate": 2.48781077650236e-06, + "loss": 0.9865, + "step": 4867 + }, + { + "epoch": 0.4390133922532353, + "grad_norm": 1.5122678089830603, + "learning_rate": 2.4872441887975386e-06, + "loss": 0.9526, + "step": 4868 + }, + { + "epoch": 0.4391035757767056, + "grad_norm": 1.2372961126811692, + "learning_rate": 2.486677559518356e-06, + "loss": 0.9317, + "step": 4869 + }, + { + "epoch": 0.43919375930017585, + "grad_norm": 1.7196833763787118, + "learning_rate": 2.4861108887131614e-06, + "loss": 0.9807, + "step": 4870 + }, + { + "epoch": 0.43928394282364613, + "grad_norm": 1.5975446965228566, + "learning_rate": 2.485544176430305e-06, + "loss": 0.9576, + "step": 4871 + }, + { + "epoch": 0.43937412634711637, + "grad_norm": 2.1699359388608146, + "learning_rate": 2.4849774227181425e-06, + "loss": 1.0674, + "step": 4872 + }, + { + "epoch": 0.43946430987058666, + "grad_norm": 1.461720221236858, + "learning_rate": 2.484410627625032e-06, + "loss": 0.9912, + "step": 4873 + }, + { + "epoch": 0.4395544933940569, + "grad_norm": 1.788674458228814, + "learning_rate": 2.4838437911993356e-06, + "loss": 1.0301, + "step": 4874 + }, + { + "epoch": 0.4396446769175272, + "grad_norm": 1.8745063767750791, + "learning_rate": 2.483276913489419e-06, + "loss": 1.0242, + "step": 4875 + }, + { + "epoch": 0.4397348604409974, + "grad_norm": 1.803584831220368, + "learning_rate": 2.4827099945436516e-06, + "loss": 0.8938, + "step": 4876 + }, + { + "epoch": 0.4398250439644677, + "grad_norm": 1.9102796268988858, + "learning_rate": 2.482143034410405e-06, + "loss": 1.0677, + "step": 4877 + }, + { + "epoch": 0.43991522748793793, + "grad_norm": 1.661733510857983, + "learning_rate": 2.4815760331380573e-06, + "loss": 1.0176, + "step": 4878 + }, + { + "epoch": 0.4400054110114082, + "grad_norm": 1.967293033692965, + "learning_rate": 2.481008990774987e-06, + "loss": 0.949, + "step": 4879 + }, + { + "epoch": 0.44009559453487845, + "grad_norm": 1.6491675184163186, + "learning_rate": 2.480441907369577e-06, + "loss": 0.987, + "step": 4880 + }, + { + "epoch": 0.44018577805834874, + "grad_norm": 1.6748770739868997, + "learning_rate": 2.479874782970214e-06, + "loss": 0.9345, + "step": 4881 + }, + { + "epoch": 0.440275961581819, + "grad_norm": 1.9563633431941705, + "learning_rate": 2.4793076176252887e-06, + "loss": 1.0109, + "step": 4882 + }, + { + "epoch": 0.44036614510528926, + "grad_norm": 2.665005245811925, + "learning_rate": 2.478740411383195e-06, + "loss": 0.8749, + "step": 4883 + }, + { + "epoch": 0.44045632862875955, + "grad_norm": 1.444066433268649, + "learning_rate": 2.4781731642923296e-06, + "loss": 0.98, + "step": 4884 + }, + { + "epoch": 0.4405465121522298, + "grad_norm": 1.8337117440381376, + "learning_rate": 2.477605876401093e-06, + "loss": 0.975, + "step": 4885 + }, + { + "epoch": 0.4406366956757001, + "grad_norm": 1.4953173024450213, + "learning_rate": 2.4770385477578894e-06, + "loss": 0.9434, + "step": 4886 + }, + { + "epoch": 0.4407268791991703, + "grad_norm": 1.3047076469282082, + "learning_rate": 2.476471178411127e-06, + "loss": 0.9452, + "step": 4887 + }, + { + "epoch": 0.4408170627226406, + "grad_norm": 1.7457480547098059, + "learning_rate": 2.475903768409216e-06, + "loss": 1.0124, + "step": 4888 + }, + { + "epoch": 0.44090724624611083, + "grad_norm": 1.8136434565562094, + "learning_rate": 2.475336317800572e-06, + "loss": 0.9617, + "step": 4889 + }, + { + "epoch": 0.4409974297695811, + "grad_norm": 1.914833442974638, + "learning_rate": 2.4747688266336118e-06, + "loss": 0.9797, + "step": 4890 + }, + { + "epoch": 0.44108761329305135, + "grad_norm": 1.6867590796493555, + "learning_rate": 2.4742012949567574e-06, + "loss": 1.0911, + "step": 4891 + }, + { + "epoch": 0.44117779681652164, + "grad_norm": 1.7939189977486159, + "learning_rate": 2.4736337228184338e-06, + "loss": 0.9736, + "step": 4892 + }, + { + "epoch": 0.44126798033999187, + "grad_norm": 1.7969333141650055, + "learning_rate": 2.4730661102670692e-06, + "loss": 1.0885, + "step": 4893 + }, + { + "epoch": 0.44135816386346216, + "grad_norm": 3.438523018936737, + "learning_rate": 2.472498457351096e-06, + "loss": 0.9929, + "step": 4894 + }, + { + "epoch": 0.4414483473869324, + "grad_norm": 2.0523316544552808, + "learning_rate": 2.4719307641189495e-06, + "loss": 0.9745, + "step": 4895 + }, + { + "epoch": 0.4415385309104027, + "grad_norm": 1.8890042254197221, + "learning_rate": 2.4713630306190673e-06, + "loss": 0.9535, + "step": 4896 + }, + { + "epoch": 0.4416287144338729, + "grad_norm": 1.9906537303917298, + "learning_rate": 2.4707952568998923e-06, + "loss": 1.0673, + "step": 4897 + }, + { + "epoch": 0.4417188979573432, + "grad_norm": 1.9364213709786064, + "learning_rate": 2.4702274430098703e-06, + "loss": 0.9623, + "step": 4898 + }, + { + "epoch": 0.44180908148081344, + "grad_norm": 1.678843157378442, + "learning_rate": 2.4696595889974497e-06, + "loss": 0.9453, + "step": 4899 + }, + { + "epoch": 0.4418992650042837, + "grad_norm": 1.446011513422799, + "learning_rate": 2.469091694911084e-06, + "loss": 1.045, + "step": 4900 + }, + { + "epoch": 0.44198944852775396, + "grad_norm": 1.5671974524108758, + "learning_rate": 2.4685237607992276e-06, + "loss": 0.9868, + "step": 4901 + }, + { + "epoch": 0.44207963205122425, + "grad_norm": 4.719769096222322, + "learning_rate": 2.4679557867103416e-06, + "loss": 1.0533, + "step": 4902 + }, + { + "epoch": 0.4421698155746945, + "grad_norm": 1.4240612786266655, + "learning_rate": 2.4673877726928865e-06, + "loss": 0.9934, + "step": 4903 + }, + { + "epoch": 0.44225999909816477, + "grad_norm": 1.9655406770070571, + "learning_rate": 2.46681971879533e-06, + "loss": 0.8771, + "step": 4904 + }, + { + "epoch": 0.442350182621635, + "grad_norm": 1.8650421842452471, + "learning_rate": 2.4662516250661407e-06, + "loss": 0.9158, + "step": 4905 + }, + { + "epoch": 0.4424403661451053, + "grad_norm": 1.858470664773192, + "learning_rate": 2.465683491553792e-06, + "loss": 0.9654, + "step": 4906 + }, + { + "epoch": 0.4425305496685755, + "grad_norm": 1.5583700148537223, + "learning_rate": 2.4651153183067604e-06, + "loss": 0.9915, + "step": 4907 + }, + { + "epoch": 0.4426207331920458, + "grad_norm": 1.4989944902538912, + "learning_rate": 2.4645471053735245e-06, + "loss": 0.9265, + "step": 4908 + }, + { + "epoch": 0.4427109167155161, + "grad_norm": 1.5326139642305534, + "learning_rate": 2.4639788528025684e-06, + "loss": 0.9246, + "step": 4909 + }, + { + "epoch": 0.44280110023898633, + "grad_norm": 1.599146961265019, + "learning_rate": 2.463410560642378e-06, + "loss": 0.9968, + "step": 4910 + }, + { + "epoch": 0.4428912837624566, + "grad_norm": 0.6300774325362845, + "learning_rate": 2.4628422289414448e-06, + "loss": 0.8246, + "step": 4911 + }, + { + "epoch": 0.44298146728592686, + "grad_norm": 0.6758686295835789, + "learning_rate": 2.4622738577482592e-06, + "loss": 0.8007, + "step": 4912 + }, + { + "epoch": 0.44307165080939714, + "grad_norm": 1.5957466035520185, + "learning_rate": 2.461705447111319e-06, + "loss": 0.9769, + "step": 4913 + }, + { + "epoch": 0.4431618343328674, + "grad_norm": 1.5415823200280996, + "learning_rate": 2.4611369970791246e-06, + "loss": 0.9595, + "step": 4914 + }, + { + "epoch": 0.44325201785633767, + "grad_norm": 2.1389763044959307, + "learning_rate": 2.460568507700179e-06, + "loss": 0.9898, + "step": 4915 + }, + { + "epoch": 0.4433422013798079, + "grad_norm": 1.802145484631671, + "learning_rate": 2.4599999790229887e-06, + "loss": 1.0293, + "step": 4916 + }, + { + "epoch": 0.4434323849032782, + "grad_norm": 1.8615429204910432, + "learning_rate": 2.459431411096064e-06, + "loss": 0.9502, + "step": 4917 + }, + { + "epoch": 0.4435225684267484, + "grad_norm": 1.6797208560914005, + "learning_rate": 2.458862803967918e-06, + "loss": 0.9601, + "step": 4918 + }, + { + "epoch": 0.4436127519502187, + "grad_norm": 1.556427944688165, + "learning_rate": 2.4582941576870667e-06, + "loss": 1.0074, + "step": 4919 + }, + { + "epoch": 0.44370293547368894, + "grad_norm": 1.747921253997033, + "learning_rate": 2.4577254723020315e-06, + "loss": 0.9689, + "step": 4920 + }, + { + "epoch": 0.44379311899715923, + "grad_norm": 2.1811635478466154, + "learning_rate": 2.457156747861335e-06, + "loss": 1.0066, + "step": 4921 + }, + { + "epoch": 0.44388330252062946, + "grad_norm": 2.094299358207496, + "learning_rate": 2.456587984413504e-06, + "loss": 1.0537, + "step": 4922 + }, + { + "epoch": 0.44397348604409975, + "grad_norm": 1.3686375802611919, + "learning_rate": 2.4560191820070683e-06, + "loss": 0.9199, + "step": 4923 + }, + { + "epoch": 0.44406366956757, + "grad_norm": 1.9400398465616495, + "learning_rate": 2.4554503406905617e-06, + "loss": 0.9692, + "step": 4924 + }, + { + "epoch": 0.4441538530910403, + "grad_norm": 1.4575364615065702, + "learning_rate": 2.454881460512521e-06, + "loss": 1.0301, + "step": 4925 + }, + { + "epoch": 0.4442440366145105, + "grad_norm": 1.854143042538907, + "learning_rate": 2.4543125415214856e-06, + "loss": 0.9574, + "step": 4926 + }, + { + "epoch": 0.4443342201379808, + "grad_norm": 1.7135498448633757, + "learning_rate": 2.4537435837659996e-06, + "loss": 1.0137, + "step": 4927 + }, + { + "epoch": 0.44442440366145103, + "grad_norm": 1.8807480126486666, + "learning_rate": 2.4531745872946085e-06, + "loss": 0.9418, + "step": 4928 + }, + { + "epoch": 0.4445145871849213, + "grad_norm": 1.7800118677408012, + "learning_rate": 2.4526055521558632e-06, + "loss": 1.053, + "step": 4929 + }, + { + "epoch": 0.44460477070839155, + "grad_norm": 1.711379320762908, + "learning_rate": 2.4520364783983164e-06, + "loss": 1.0032, + "step": 4930 + }, + { + "epoch": 0.44469495423186184, + "grad_norm": 1.4094780641075542, + "learning_rate": 2.451467366070525e-06, + "loss": 0.9731, + "step": 4931 + }, + { + "epoch": 0.4447851377553321, + "grad_norm": 2.8304888714826126, + "learning_rate": 2.450898215221048e-06, + "loss": 1.0614, + "step": 4932 + }, + { + "epoch": 0.44487532127880236, + "grad_norm": 1.6846618804370523, + "learning_rate": 2.4503290258984498e-06, + "loss": 1.1343, + "step": 4933 + }, + { + "epoch": 0.44496550480227265, + "grad_norm": 1.949043158991345, + "learning_rate": 2.4497597981512952e-06, + "loss": 0.8517, + "step": 4934 + }, + { + "epoch": 0.4450556883257429, + "grad_norm": 0.6368288986321533, + "learning_rate": 2.4491905320281555e-06, + "loss": 0.8289, + "step": 4935 + }, + { + "epoch": 0.44514587184921317, + "grad_norm": 1.626489909931434, + "learning_rate": 2.448621227577602e-06, + "loss": 0.9619, + "step": 4936 + }, + { + "epoch": 0.4452360553726834, + "grad_norm": 3.4301049884265087, + "learning_rate": 2.4480518848482123e-06, + "loss": 0.9345, + "step": 4937 + }, + { + "epoch": 0.4453262388961537, + "grad_norm": 2.0013856855877594, + "learning_rate": 2.447482503888565e-06, + "loss": 1.0034, + "step": 4938 + }, + { + "epoch": 0.4454164224196239, + "grad_norm": 2.0434953583435256, + "learning_rate": 2.4469130847472434e-06, + "loss": 0.9633, + "step": 4939 + }, + { + "epoch": 0.4455066059430942, + "grad_norm": 2.090770480307311, + "learning_rate": 2.4463436274728326e-06, + "loss": 1.0266, + "step": 4940 + }, + { + "epoch": 0.44559678946656445, + "grad_norm": 1.492319071296762, + "learning_rate": 2.4457741321139227e-06, + "loss": 0.837, + "step": 4941 + }, + { + "epoch": 0.44568697299003474, + "grad_norm": 1.5839182961035867, + "learning_rate": 2.4452045987191063e-06, + "loss": 1.0108, + "step": 4942 + }, + { + "epoch": 0.44577715651350497, + "grad_norm": 1.527148612354511, + "learning_rate": 2.4446350273369776e-06, + "loss": 0.9585, + "step": 4943 + }, + { + "epoch": 0.44586734003697526, + "grad_norm": 1.6984415315722332, + "learning_rate": 2.4440654180161374e-06, + "loss": 1.0259, + "step": 4944 + }, + { + "epoch": 0.4459575235604455, + "grad_norm": 1.595498192093081, + "learning_rate": 2.4434957708051875e-06, + "loss": 0.9446, + "step": 4945 + }, + { + "epoch": 0.4460477070839158, + "grad_norm": 1.7084813015132003, + "learning_rate": 2.4429260857527324e-06, + "loss": 0.9307, + "step": 4946 + }, + { + "epoch": 0.446137890607386, + "grad_norm": 1.9015909385592038, + "learning_rate": 2.4423563629073815e-06, + "loss": 0.8944, + "step": 4947 + }, + { + "epoch": 0.4462280741308563, + "grad_norm": 1.4273121596976266, + "learning_rate": 2.4417866023177466e-06, + "loss": 0.9912, + "step": 4948 + }, + { + "epoch": 0.44631825765432653, + "grad_norm": 1.7396286446462659, + "learning_rate": 2.441216804032443e-06, + "loss": 1.0213, + "step": 4949 + }, + { + "epoch": 0.4464084411777968, + "grad_norm": 1.4536236143401629, + "learning_rate": 2.440646968100089e-06, + "loss": 1.0017, + "step": 4950 + }, + { + "epoch": 0.44649862470126706, + "grad_norm": 0.6813778302363622, + "learning_rate": 2.4400770945693055e-06, + "loss": 0.8112, + "step": 4951 + }, + { + "epoch": 0.44658880822473734, + "grad_norm": 1.7690800308358916, + "learning_rate": 2.4395071834887177e-06, + "loss": 1.018, + "step": 4952 + }, + { + "epoch": 0.4466789917482076, + "grad_norm": 0.705587799909493, + "learning_rate": 2.438937234906954e-06, + "loss": 0.8726, + "step": 4953 + }, + { + "epoch": 0.44676917527167787, + "grad_norm": 3.198876368421696, + "learning_rate": 2.4383672488726447e-06, + "loss": 0.9788, + "step": 4954 + }, + { + "epoch": 0.44685935879514815, + "grad_norm": 2.009561570970029, + "learning_rate": 2.4377972254344256e-06, + "loss": 0.9077, + "step": 4955 + }, + { + "epoch": 0.4469495423186184, + "grad_norm": 2.5871874966087485, + "learning_rate": 2.437227164640932e-06, + "loss": 0.878, + "step": 4956 + }, + { + "epoch": 0.4470397258420887, + "grad_norm": 1.6777718909513315, + "learning_rate": 2.436657066540807e-06, + "loss": 0.9312, + "step": 4957 + }, + { + "epoch": 0.4471299093655589, + "grad_norm": 2.0728760017227432, + "learning_rate": 2.4360869311826927e-06, + "loss": 0.8616, + "step": 4958 + }, + { + "epoch": 0.4472200928890292, + "grad_norm": 1.8631178426365138, + "learning_rate": 2.4355167586152367e-06, + "loss": 1.0831, + "step": 4959 + }, + { + "epoch": 0.44731027641249943, + "grad_norm": 2.012473074217562, + "learning_rate": 2.4349465488870896e-06, + "loss": 0.959, + "step": 4960 + }, + { + "epoch": 0.4474004599359697, + "grad_norm": 2.190593276290724, + "learning_rate": 2.434376302046905e-06, + "loss": 1.0743, + "step": 4961 + }, + { + "epoch": 0.44749064345943995, + "grad_norm": 1.9422631320359665, + "learning_rate": 2.433806018143339e-06, + "loss": 1.0132, + "step": 4962 + }, + { + "epoch": 0.44758082698291024, + "grad_norm": 0.6712997546661079, + "learning_rate": 2.433235697225051e-06, + "loss": 0.8533, + "step": 4963 + }, + { + "epoch": 0.4476710105063805, + "grad_norm": 1.5640701034620224, + "learning_rate": 2.4326653393407048e-06, + "loss": 1.0054, + "step": 4964 + }, + { + "epoch": 0.44776119402985076, + "grad_norm": 1.6835578507492572, + "learning_rate": 2.432094944538966e-06, + "loss": 0.9581, + "step": 4965 + }, + { + "epoch": 0.447851377553321, + "grad_norm": 1.4433298069218525, + "learning_rate": 2.4315245128685047e-06, + "loss": 0.956, + "step": 4966 + }, + { + "epoch": 0.4479415610767913, + "grad_norm": 1.663919003885799, + "learning_rate": 2.4309540443779925e-06, + "loss": 0.991, + "step": 4967 + }, + { + "epoch": 0.4480317446002615, + "grad_norm": 1.6261465722634763, + "learning_rate": 2.4303835391161047e-06, + "loss": 0.9212, + "step": 4968 + }, + { + "epoch": 0.4481219281237318, + "grad_norm": 1.7602123239714054, + "learning_rate": 2.42981299713152e-06, + "loss": 0.9396, + "step": 4969 + }, + { + "epoch": 0.44821211164720204, + "grad_norm": 1.8464900201300025, + "learning_rate": 2.4292424184729204e-06, + "loss": 1.0225, + "step": 4970 + }, + { + "epoch": 0.4483022951706723, + "grad_norm": 1.3964543265997986, + "learning_rate": 2.4286718031889913e-06, + "loss": 0.9827, + "step": 4971 + }, + { + "epoch": 0.44839247869414256, + "grad_norm": 2.425305600917752, + "learning_rate": 2.4281011513284202e-06, + "loss": 0.9641, + "step": 4972 + }, + { + "epoch": 0.44848266221761285, + "grad_norm": 2.641982773440942, + "learning_rate": 2.4275304629398985e-06, + "loss": 0.9216, + "step": 4973 + }, + { + "epoch": 0.4485728457410831, + "grad_norm": 1.5838361075702594, + "learning_rate": 2.4269597380721194e-06, + "loss": 1.0625, + "step": 4974 + }, + { + "epoch": 0.44866302926455337, + "grad_norm": 1.5294195639942205, + "learning_rate": 2.426388976773782e-06, + "loss": 0.9829, + "step": 4975 + }, + { + "epoch": 0.4487532127880236, + "grad_norm": 1.7896285181998797, + "learning_rate": 2.425818179093586e-06, + "loss": 1.0919, + "step": 4976 + }, + { + "epoch": 0.4488433963114939, + "grad_norm": 1.463095636606818, + "learning_rate": 2.4252473450802346e-06, + "loss": 0.8964, + "step": 4977 + }, + { + "epoch": 0.4489335798349641, + "grad_norm": 2.1748747383966887, + "learning_rate": 2.4246764747824355e-06, + "loss": 1.0183, + "step": 4978 + }, + { + "epoch": 0.4490237633584344, + "grad_norm": 1.4102194394142853, + "learning_rate": 2.424105568248897e-06, + "loss": 0.9974, + "step": 4979 + }, + { + "epoch": 0.4491139468819047, + "grad_norm": 1.7555786179920299, + "learning_rate": 2.4235346255283337e-06, + "loss": 0.9276, + "step": 4980 + }, + { + "epoch": 0.44920413040537494, + "grad_norm": 1.6712414262961117, + "learning_rate": 2.42296364666946e-06, + "loss": 0.9986, + "step": 4981 + }, + { + "epoch": 0.4492943139288452, + "grad_norm": 1.868501336640045, + "learning_rate": 2.4223926317209965e-06, + "loss": 0.9392, + "step": 4982 + }, + { + "epoch": 0.44938449745231546, + "grad_norm": 1.4243334800892142, + "learning_rate": 2.4218215807316647e-06, + "loss": 0.9883, + "step": 4983 + }, + { + "epoch": 0.44947468097578575, + "grad_norm": 2.0569274064803027, + "learning_rate": 2.4212504937501894e-06, + "loss": 1.0764, + "step": 4984 + }, + { + "epoch": 0.449564864499256, + "grad_norm": 1.4668752189671825, + "learning_rate": 2.4206793708253e-06, + "loss": 1.0443, + "step": 4985 + }, + { + "epoch": 0.44965504802272627, + "grad_norm": 2.0053648520150307, + "learning_rate": 2.420108212005726e-06, + "loss": 0.9387, + "step": 4986 + }, + { + "epoch": 0.4497452315461965, + "grad_norm": 1.8119932321249441, + "learning_rate": 2.4195370173402034e-06, + "loss": 0.9862, + "step": 4987 + }, + { + "epoch": 0.4498354150696668, + "grad_norm": 1.8023768573658556, + "learning_rate": 2.4189657868774696e-06, + "loss": 1.1084, + "step": 4988 + }, + { + "epoch": 0.449925598593137, + "grad_norm": 1.9083661075204206, + "learning_rate": 2.418394520666264e-06, + "loss": 0.9264, + "step": 4989 + }, + { + "epoch": 0.4500157821166073, + "grad_norm": 1.9531903065253622, + "learning_rate": 2.4178232187553307e-06, + "loss": 0.9716, + "step": 4990 + }, + { + "epoch": 0.45010596564007754, + "grad_norm": 1.4992499065491376, + "learning_rate": 2.417251881193417e-06, + "loss": 1.0149, + "step": 4991 + }, + { + "epoch": 0.45019614916354783, + "grad_norm": 1.7289859110410972, + "learning_rate": 2.4166805080292723e-06, + "loss": 1.0472, + "step": 4992 + }, + { + "epoch": 0.45028633268701806, + "grad_norm": 1.6321746456393071, + "learning_rate": 2.4161090993116485e-06, + "loss": 0.9395, + "step": 4993 + }, + { + "epoch": 0.45037651621048835, + "grad_norm": 2.2064049296243105, + "learning_rate": 2.4155376550893026e-06, + "loss": 0.8671, + "step": 4994 + }, + { + "epoch": 0.4504666997339586, + "grad_norm": 1.91636967085628, + "learning_rate": 2.4149661754109926e-06, + "loss": 0.9973, + "step": 4995 + }, + { + "epoch": 0.4505568832574289, + "grad_norm": 1.9661405314864533, + "learning_rate": 2.41439466032548e-06, + "loss": 1.0462, + "step": 4996 + }, + { + "epoch": 0.4506470667808991, + "grad_norm": 1.9170081691299243, + "learning_rate": 2.41382310988153e-06, + "loss": 0.9341, + "step": 4997 + }, + { + "epoch": 0.4507372503043694, + "grad_norm": 1.7656002211730915, + "learning_rate": 2.413251524127911e-06, + "loss": 1.0477, + "step": 4998 + }, + { + "epoch": 0.45082743382783963, + "grad_norm": 1.6677342413042104, + "learning_rate": 2.412679903113393e-06, + "loss": 0.9563, + "step": 4999 + }, + { + "epoch": 0.4509176173513099, + "grad_norm": 2.2338695888175186, + "learning_rate": 2.4121082468867505e-06, + "loss": 0.9989, + "step": 5000 + }, + { + "epoch": 0.45100780087478015, + "grad_norm": 2.24613546581506, + "learning_rate": 2.4115365554967597e-06, + "loss": 0.9257, + "step": 5001 + }, + { + "epoch": 0.45109798439825044, + "grad_norm": 1.6726044106501594, + "learning_rate": 2.4109648289922006e-06, + "loss": 0.9885, + "step": 5002 + }, + { + "epoch": 0.45118816792172073, + "grad_norm": 2.668579299792038, + "learning_rate": 2.4103930674218565e-06, + "loss": 0.996, + "step": 5003 + }, + { + "epoch": 0.45127835144519096, + "grad_norm": 1.5883136106122517, + "learning_rate": 2.409821270834513e-06, + "loss": 0.964, + "step": 5004 + }, + { + "epoch": 0.45136853496866125, + "grad_norm": 1.9724542193303565, + "learning_rate": 2.409249439278959e-06, + "loss": 1.0456, + "step": 5005 + }, + { + "epoch": 0.4514587184921315, + "grad_norm": 0.6156769386882744, + "learning_rate": 2.408677572803986e-06, + "loss": 0.7888, + "step": 5006 + }, + { + "epoch": 0.45154890201560177, + "grad_norm": 1.736444492967625, + "learning_rate": 2.408105671458389e-06, + "loss": 0.9938, + "step": 5007 + }, + { + "epoch": 0.451639085539072, + "grad_norm": 1.5393870055163794, + "learning_rate": 2.4075337352909663e-06, + "loss": 0.9328, + "step": 5008 + }, + { + "epoch": 0.4517292690625423, + "grad_norm": 1.4410606735025338, + "learning_rate": 2.4069617643505177e-06, + "loss": 1.117, + "step": 5009 + }, + { + "epoch": 0.4518194525860125, + "grad_norm": 2.1487796025569934, + "learning_rate": 2.406389758685848e-06, + "loss": 1.0737, + "step": 5010 + }, + { + "epoch": 0.4519096361094828, + "grad_norm": 1.5808158821398866, + "learning_rate": 2.405817718345763e-06, + "loss": 1.0249, + "step": 5011 + }, + { + "epoch": 0.45199981963295305, + "grad_norm": 1.9919918787036173, + "learning_rate": 2.4052456433790726e-06, + "loss": 0.9439, + "step": 5012 + }, + { + "epoch": 0.45209000315642334, + "grad_norm": 1.666298952864516, + "learning_rate": 2.4046735338345897e-06, + "loss": 1.0454, + "step": 5013 + }, + { + "epoch": 0.45218018667989357, + "grad_norm": 1.7246718605965015, + "learning_rate": 2.404101389761129e-06, + "loss": 1.0045, + "step": 5014 + }, + { + "epoch": 0.45227037020336386, + "grad_norm": 0.6561073647758291, + "learning_rate": 2.4035292112075097e-06, + "loss": 0.8004, + "step": 5015 + }, + { + "epoch": 0.4523605537268341, + "grad_norm": 1.4955249631450538, + "learning_rate": 2.4029569982225534e-06, + "loss": 1.0481, + "step": 5016 + }, + { + "epoch": 0.4524507372503044, + "grad_norm": 0.6961008810069819, + "learning_rate": 2.402384750855084e-06, + "loss": 0.8726, + "step": 5017 + }, + { + "epoch": 0.4525409207737746, + "grad_norm": 1.5903560178667742, + "learning_rate": 2.4018124691539286e-06, + "loss": 0.9913, + "step": 5018 + }, + { + "epoch": 0.4526311042972449, + "grad_norm": 2.5586905657672707, + "learning_rate": 2.4012401531679178e-06, + "loss": 0.9801, + "step": 5019 + }, + { + "epoch": 0.45272128782071513, + "grad_norm": 1.9494695406557287, + "learning_rate": 2.4006678029458847e-06, + "loss": 0.9877, + "step": 5020 + }, + { + "epoch": 0.4528114713441854, + "grad_norm": 1.5800871783964985, + "learning_rate": 2.400095418536666e-06, + "loss": 0.896, + "step": 5021 + }, + { + "epoch": 0.45290165486765566, + "grad_norm": 1.644136576286147, + "learning_rate": 2.3995229999890996e-06, + "loss": 0.9694, + "step": 5022 + }, + { + "epoch": 0.45299183839112594, + "grad_norm": 2.3030696039253575, + "learning_rate": 2.398950547352028e-06, + "loss": 0.9311, + "step": 5023 + }, + { + "epoch": 0.4530820219145962, + "grad_norm": 2.681568875502496, + "learning_rate": 2.398378060674295e-06, + "loss": 0.9073, + "step": 5024 + }, + { + "epoch": 0.45317220543806647, + "grad_norm": 1.4300250712777844, + "learning_rate": 2.39780554000475e-06, + "loss": 0.9878, + "step": 5025 + }, + { + "epoch": 0.4532623889615367, + "grad_norm": 1.5678590805558144, + "learning_rate": 2.3972329853922434e-06, + "loss": 1.0222, + "step": 5026 + }, + { + "epoch": 0.453352572485007, + "grad_norm": 0.5957509756412434, + "learning_rate": 2.3966603968856278e-06, + "loss": 0.808, + "step": 5027 + }, + { + "epoch": 0.4534427560084773, + "grad_norm": 1.6222516072818332, + "learning_rate": 2.39608777453376e-06, + "loss": 1.0637, + "step": 5028 + }, + { + "epoch": 0.4535329395319475, + "grad_norm": 1.642793149175633, + "learning_rate": 2.3955151183854993e-06, + "loss": 1.0118, + "step": 5029 + }, + { + "epoch": 0.4536231230554178, + "grad_norm": 1.593636527416306, + "learning_rate": 2.3949424284897073e-06, + "loss": 0.9512, + "step": 5030 + }, + { + "epoch": 0.45371330657888803, + "grad_norm": 1.6082378925737657, + "learning_rate": 2.39436970489525e-06, + "loss": 0.8618, + "step": 5031 + }, + { + "epoch": 0.4538034901023583, + "grad_norm": 1.7380160354427623, + "learning_rate": 2.3937969476509955e-06, + "loss": 0.9883, + "step": 5032 + }, + { + "epoch": 0.45389367362582855, + "grad_norm": 1.542266144916425, + "learning_rate": 2.393224156805813e-06, + "loss": 1.0628, + "step": 5033 + }, + { + "epoch": 0.45398385714929884, + "grad_norm": 1.7199110097929649, + "learning_rate": 2.392651332408578e-06, + "loss": 0.9619, + "step": 5034 + }, + { + "epoch": 0.4540740406727691, + "grad_norm": 2.623860157130335, + "learning_rate": 2.3920784745081655e-06, + "loss": 0.9768, + "step": 5035 + }, + { + "epoch": 0.45416422419623936, + "grad_norm": 1.657234762842484, + "learning_rate": 2.391505583153456e-06, + "loss": 1.0218, + "step": 5036 + }, + { + "epoch": 0.4542544077197096, + "grad_norm": 1.5811958273552966, + "learning_rate": 2.3909326583933315e-06, + "loss": 0.9953, + "step": 5037 + }, + { + "epoch": 0.4543445912431799, + "grad_norm": 3.0289837732595126, + "learning_rate": 2.3903597002766777e-06, + "loss": 0.9564, + "step": 5038 + }, + { + "epoch": 0.4544347747666501, + "grad_norm": 1.462929983589986, + "learning_rate": 2.389786708852381e-06, + "loss": 0.9554, + "step": 5039 + }, + { + "epoch": 0.4545249582901204, + "grad_norm": 0.9434509232922239, + "learning_rate": 2.389213684169333e-06, + "loss": 0.9317, + "step": 5040 + }, + { + "epoch": 0.45461514181359064, + "grad_norm": 1.542268541056104, + "learning_rate": 2.388640626276428e-06, + "loss": 0.9525, + "step": 5041 + }, + { + "epoch": 0.45470532533706093, + "grad_norm": 0.6834249451525856, + "learning_rate": 2.388067535222561e-06, + "loss": 0.7729, + "step": 5042 + }, + { + "epoch": 0.45479550886053116, + "grad_norm": 1.960941937810124, + "learning_rate": 2.3874944110566332e-06, + "loss": 1.0065, + "step": 5043 + }, + { + "epoch": 0.45488569238400145, + "grad_norm": 0.7563722826874691, + "learning_rate": 2.3869212538275447e-06, + "loss": 0.9151, + "step": 5044 + }, + { + "epoch": 0.4549758759074717, + "grad_norm": 1.507928636258676, + "learning_rate": 2.386348063584202e-06, + "loss": 1.0058, + "step": 5045 + }, + { + "epoch": 0.45506605943094197, + "grad_norm": 1.521596333959397, + "learning_rate": 2.385774840375511e-06, + "loss": 1.0397, + "step": 5046 + }, + { + "epoch": 0.4551562429544122, + "grad_norm": 2.1173731873786745, + "learning_rate": 2.385201584250385e-06, + "loss": 0.9454, + "step": 5047 + }, + { + "epoch": 0.4552464264778825, + "grad_norm": 0.6384547386512204, + "learning_rate": 2.3846282952577346e-06, + "loss": 0.8497, + "step": 5048 + }, + { + "epoch": 0.4553366100013527, + "grad_norm": 2.080921901009109, + "learning_rate": 2.3840549734464785e-06, + "loss": 1.079, + "step": 5049 + }, + { + "epoch": 0.455426793524823, + "grad_norm": 2.2118540235712065, + "learning_rate": 2.3834816188655336e-06, + "loss": 0.9262, + "step": 5050 + }, + { + "epoch": 0.4555169770482933, + "grad_norm": 1.6876008745348994, + "learning_rate": 2.3829082315638224e-06, + "loss": 0.9346, + "step": 5051 + }, + { + "epoch": 0.45560716057176354, + "grad_norm": 1.7848931250112434, + "learning_rate": 2.3823348115902695e-06, + "loss": 0.8461, + "step": 5052 + }, + { + "epoch": 0.4556973440952338, + "grad_norm": 1.7003302786258154, + "learning_rate": 2.3817613589938026e-06, + "loss": 1.0093, + "step": 5053 + }, + { + "epoch": 0.45578752761870406, + "grad_norm": 1.9316185929726528, + "learning_rate": 2.3811878738233517e-06, + "loss": 1.0221, + "step": 5054 + }, + { + "epoch": 0.45587771114217435, + "grad_norm": 2.9558507863481176, + "learning_rate": 2.380614356127849e-06, + "loss": 0.9382, + "step": 5055 + }, + { + "epoch": 0.4559678946656446, + "grad_norm": 1.597558501221611, + "learning_rate": 2.3800408059562318e-06, + "loss": 0.9867, + "step": 5056 + }, + { + "epoch": 0.45605807818911487, + "grad_norm": 1.7184412592243061, + "learning_rate": 2.3794672233574365e-06, + "loss": 1.0038, + "step": 5057 + }, + { + "epoch": 0.4561482617125851, + "grad_norm": 4.486162102051573, + "learning_rate": 2.3788936083804058e-06, + "loss": 1.0023, + "step": 5058 + }, + { + "epoch": 0.4562384452360554, + "grad_norm": 1.6576530703466388, + "learning_rate": 2.378319961074083e-06, + "loss": 0.9876, + "step": 5059 + }, + { + "epoch": 0.4563286287595256, + "grad_norm": 1.8171837240033923, + "learning_rate": 2.377746281487415e-06, + "loss": 1.0267, + "step": 5060 + }, + { + "epoch": 0.4564188122829959, + "grad_norm": 1.6435441704162541, + "learning_rate": 2.377172569669352e-06, + "loss": 1.0791, + "step": 5061 + }, + { + "epoch": 0.45650899580646614, + "grad_norm": 1.7201608589270705, + "learning_rate": 2.376598825668845e-06, + "loss": 0.842, + "step": 5062 + }, + { + "epoch": 0.45659917932993643, + "grad_norm": 1.680603061701655, + "learning_rate": 2.3760250495348495e-06, + "loss": 0.9766, + "step": 5063 + }, + { + "epoch": 0.45668936285340667, + "grad_norm": 1.7539613031187817, + "learning_rate": 2.3754512413163236e-06, + "loss": 1.0353, + "step": 5064 + }, + { + "epoch": 0.45677954637687695, + "grad_norm": 1.9562627993033948, + "learning_rate": 2.3748774010622285e-06, + "loss": 0.971, + "step": 5065 + }, + { + "epoch": 0.4568697299003472, + "grad_norm": 1.8532074651011594, + "learning_rate": 2.3743035288215254e-06, + "loss": 1.0003, + "step": 5066 + }, + { + "epoch": 0.4569599134238175, + "grad_norm": 2.124372950769229, + "learning_rate": 2.3737296246431815e-06, + "loss": 0.9954, + "step": 5067 + }, + { + "epoch": 0.4570500969472877, + "grad_norm": 1.4283330786640618, + "learning_rate": 2.3731556885761656e-06, + "loss": 0.9619, + "step": 5068 + }, + { + "epoch": 0.457140280470758, + "grad_norm": 0.6342501616141324, + "learning_rate": 2.372581720669449e-06, + "loss": 0.8172, + "step": 5069 + }, + { + "epoch": 0.45723046399422823, + "grad_norm": 1.735089678906441, + "learning_rate": 2.3720077209720046e-06, + "loss": 1.0185, + "step": 5070 + }, + { + "epoch": 0.4573206475176985, + "grad_norm": 1.0072786558481936, + "learning_rate": 2.3714336895328112e-06, + "loss": 0.8103, + "step": 5071 + }, + { + "epoch": 0.45741083104116875, + "grad_norm": 1.9059194059274567, + "learning_rate": 2.370859626400847e-06, + "loss": 0.9438, + "step": 5072 + }, + { + "epoch": 0.45750101456463904, + "grad_norm": 1.6443374048369894, + "learning_rate": 2.3702855316250943e-06, + "loss": 0.9641, + "step": 5073 + }, + { + "epoch": 0.45759119808810933, + "grad_norm": 1.9823850130309013, + "learning_rate": 2.369711405254539e-06, + "loss": 0.9932, + "step": 5074 + }, + { + "epoch": 0.45768138161157956, + "grad_norm": 1.9301466916722394, + "learning_rate": 2.3691372473381673e-06, + "loss": 1.0418, + "step": 5075 + }, + { + "epoch": 0.45777156513504985, + "grad_norm": 1.5299648419248872, + "learning_rate": 2.3685630579249708e-06, + "loss": 1.0473, + "step": 5076 + }, + { + "epoch": 0.4578617486585201, + "grad_norm": 1.9247132719444782, + "learning_rate": 2.367988837063942e-06, + "loss": 0.9589, + "step": 5077 + }, + { + "epoch": 0.4579519321819904, + "grad_norm": 1.6654103153283577, + "learning_rate": 2.367414584804076e-06, + "loss": 1.0683, + "step": 5078 + }, + { + "epoch": 0.4580421157054606, + "grad_norm": 1.4528814952645748, + "learning_rate": 2.366840301194372e-06, + "loss": 1.0237, + "step": 5079 + }, + { + "epoch": 0.4581322992289309, + "grad_norm": 1.8713473026763061, + "learning_rate": 2.3662659862838308e-06, + "loss": 0.9766, + "step": 5080 + }, + { + "epoch": 0.45822248275240113, + "grad_norm": 1.994701041477907, + "learning_rate": 2.365691640121456e-06, + "loss": 1.0491, + "step": 5081 + }, + { + "epoch": 0.4583126662758714, + "grad_norm": 2.1669511486016537, + "learning_rate": 2.365117262756254e-06, + "loss": 0.9189, + "step": 5082 + }, + { + "epoch": 0.45840284979934165, + "grad_norm": 2.6413139934162326, + "learning_rate": 2.3645428542372342e-06, + "loss": 0.9869, + "step": 5083 + }, + { + "epoch": 0.45849303332281194, + "grad_norm": 0.6505441112237418, + "learning_rate": 2.3639684146134083e-06, + "loss": 0.7691, + "step": 5084 + }, + { + "epoch": 0.45858321684628217, + "grad_norm": 1.6357112870215655, + "learning_rate": 2.3633939439337897e-06, + "loss": 0.9527, + "step": 5085 + }, + { + "epoch": 0.45867340036975246, + "grad_norm": 1.5391949436885874, + "learning_rate": 2.362819442247396e-06, + "loss": 0.9634, + "step": 5086 + }, + { + "epoch": 0.4587635838932227, + "grad_norm": 0.8190442772169522, + "learning_rate": 2.3622449096032477e-06, + "loss": 0.7789, + "step": 5087 + }, + { + "epoch": 0.458853767416693, + "grad_norm": 1.7005916968200254, + "learning_rate": 2.361670346050366e-06, + "loss": 1.0263, + "step": 5088 + }, + { + "epoch": 0.4589439509401632, + "grad_norm": 2.027162753277987, + "learning_rate": 2.3610957516377757e-06, + "loss": 0.9283, + "step": 5089 + }, + { + "epoch": 0.4590341344636335, + "grad_norm": 2.026567196454538, + "learning_rate": 2.3605211264145048e-06, + "loss": 0.9948, + "step": 5090 + }, + { + "epoch": 0.45912431798710374, + "grad_norm": 1.4411309865892872, + "learning_rate": 2.3599464704295836e-06, + "loss": 0.9535, + "step": 5091 + }, + { + "epoch": 0.459214501510574, + "grad_norm": 1.8608222265918557, + "learning_rate": 2.359371783732045e-06, + "loss": 0.8824, + "step": 5092 + }, + { + "epoch": 0.45930468503404426, + "grad_norm": 1.7247898443717498, + "learning_rate": 2.358797066370924e-06, + "loss": 0.9726, + "step": 5093 + }, + { + "epoch": 0.45939486855751455, + "grad_norm": 1.6243839563384832, + "learning_rate": 2.3582223183952594e-06, + "loss": 0.9881, + "step": 5094 + }, + { + "epoch": 0.4594850520809848, + "grad_norm": 2.2089805974008923, + "learning_rate": 2.357647539854091e-06, + "loss": 1.0262, + "step": 5095 + }, + { + "epoch": 0.45957523560445507, + "grad_norm": 2.0535581552775497, + "learning_rate": 2.3570727307964624e-06, + "loss": 0.9967, + "step": 5096 + }, + { + "epoch": 0.4596654191279253, + "grad_norm": 1.494822148489657, + "learning_rate": 2.35649789127142e-06, + "loss": 1.0379, + "step": 5097 + }, + { + "epoch": 0.4597556026513956, + "grad_norm": 2.51424792016717, + "learning_rate": 2.3559230213280115e-06, + "loss": 0.954, + "step": 5098 + }, + { + "epoch": 0.4598457861748659, + "grad_norm": 2.9201266519958153, + "learning_rate": 2.3553481210152886e-06, + "loss": 0.9034, + "step": 5099 + }, + { + "epoch": 0.4599359696983361, + "grad_norm": 0.8206821153902919, + "learning_rate": 2.3547731903823043e-06, + "loss": 0.8345, + "step": 5100 + }, + { + "epoch": 0.4600261532218064, + "grad_norm": 1.632947706255392, + "learning_rate": 2.3541982294781155e-06, + "loss": 0.965, + "step": 5101 + }, + { + "epoch": 0.46011633674527663, + "grad_norm": 1.4876988991224243, + "learning_rate": 2.3536232383517804e-06, + "loss": 0.9877, + "step": 5102 + }, + { + "epoch": 0.4602065202687469, + "grad_norm": 1.5271260528348498, + "learning_rate": 2.3530482170523602e-06, + "loss": 0.981, + "step": 5103 + }, + { + "epoch": 0.46029670379221715, + "grad_norm": 1.48953202521607, + "learning_rate": 2.3524731656289206e-06, + "loss": 1.0855, + "step": 5104 + }, + { + "epoch": 0.46038688731568744, + "grad_norm": 1.4379423538930574, + "learning_rate": 2.351898084130526e-06, + "loss": 1.0786, + "step": 5105 + }, + { + "epoch": 0.4604770708391577, + "grad_norm": 4.529484687783505, + "learning_rate": 2.351322972606247e-06, + "loss": 1.0557, + "step": 5106 + }, + { + "epoch": 0.46056725436262796, + "grad_norm": 2.363110345382874, + "learning_rate": 2.350747831105155e-06, + "loss": 1.1413, + "step": 5107 + }, + { + "epoch": 0.4606574378860982, + "grad_norm": 1.7516353323203038, + "learning_rate": 2.350172659676323e-06, + "loss": 1.0681, + "step": 5108 + }, + { + "epoch": 0.4607476214095685, + "grad_norm": 1.6642177312380622, + "learning_rate": 2.3495974583688306e-06, + "loss": 1.028, + "step": 5109 + }, + { + "epoch": 0.4608378049330387, + "grad_norm": 1.42327133197032, + "learning_rate": 2.3490222272317543e-06, + "loss": 1.0604, + "step": 5110 + }, + { + "epoch": 0.460927988456509, + "grad_norm": 1.9373107940826797, + "learning_rate": 2.348446966314177e-06, + "loss": 1.0359, + "step": 5111 + }, + { + "epoch": 0.46101817197997924, + "grad_norm": 1.7687952393047184, + "learning_rate": 2.3478716756651837e-06, + "loss": 0.9629, + "step": 5112 + }, + { + "epoch": 0.46110835550344953, + "grad_norm": 1.8600796558387125, + "learning_rate": 2.347296355333861e-06, + "loss": 0.9831, + "step": 5113 + }, + { + "epoch": 0.46119853902691976, + "grad_norm": 1.866503027430403, + "learning_rate": 2.3467210053692972e-06, + "loss": 0.9994, + "step": 5114 + }, + { + "epoch": 0.46128872255039005, + "grad_norm": 1.666982318869075, + "learning_rate": 2.3461456258205866e-06, + "loss": 0.9941, + "step": 5115 + }, + { + "epoch": 0.4613789060738603, + "grad_norm": 2.1061425201431714, + "learning_rate": 2.345570216736822e-06, + "loss": 0.9297, + "step": 5116 + }, + { + "epoch": 0.4614690895973306, + "grad_norm": 1.6224119411136282, + "learning_rate": 2.3449947781671013e-06, + "loss": 1.0004, + "step": 5117 + }, + { + "epoch": 0.4615592731208008, + "grad_norm": 1.65543064507434, + "learning_rate": 2.3444193101605237e-06, + "loss": 1.0207, + "step": 5118 + }, + { + "epoch": 0.4616494566442711, + "grad_norm": 2.0862799088406683, + "learning_rate": 2.3438438127661913e-06, + "loss": 1.0105, + "step": 5119 + }, + { + "epoch": 0.4617396401677413, + "grad_norm": 7.079627949050194, + "learning_rate": 2.3432682860332096e-06, + "loss": 0.8167, + "step": 5120 + }, + { + "epoch": 0.4618298236912116, + "grad_norm": 1.7747823178768682, + "learning_rate": 2.342692730010684e-06, + "loss": 0.94, + "step": 5121 + }, + { + "epoch": 0.4619200072146819, + "grad_norm": 2.232247146349995, + "learning_rate": 2.342117144747726e-06, + "loss": 1.0009, + "step": 5122 + }, + { + "epoch": 0.46201019073815214, + "grad_norm": 3.319137544867348, + "learning_rate": 2.3415415302934457e-06, + "loss": 0.9934, + "step": 5123 + }, + { + "epoch": 0.4621003742616224, + "grad_norm": 1.70765280155766, + "learning_rate": 2.340965886696959e-06, + "loss": 1.0192, + "step": 5124 + }, + { + "epoch": 0.46219055778509266, + "grad_norm": 1.6436140896267308, + "learning_rate": 2.340390214007384e-06, + "loss": 0.9416, + "step": 5125 + }, + { + "epoch": 0.46228074130856295, + "grad_norm": 1.706588722674381, + "learning_rate": 2.339814512273838e-06, + "loss": 0.8746, + "step": 5126 + }, + { + "epoch": 0.4623709248320332, + "grad_norm": 1.577331267822818, + "learning_rate": 2.3392387815454447e-06, + "loss": 1.0962, + "step": 5127 + }, + { + "epoch": 0.46246110835550347, + "grad_norm": 2.384232995040521, + "learning_rate": 2.3386630218713273e-06, + "loss": 0.9843, + "step": 5128 + }, + { + "epoch": 0.4625512918789737, + "grad_norm": 2.1975784282554605, + "learning_rate": 2.3380872333006135e-06, + "loss": 1.0197, + "step": 5129 + }, + { + "epoch": 0.462641475402444, + "grad_norm": 1.6924900168458314, + "learning_rate": 2.3375114158824335e-06, + "loss": 1.0387, + "step": 5130 + }, + { + "epoch": 0.4627316589259142, + "grad_norm": 1.9388660414380994, + "learning_rate": 2.3369355696659184e-06, + "loss": 1.0271, + "step": 5131 + }, + { + "epoch": 0.4628218424493845, + "grad_norm": 1.6125392731608001, + "learning_rate": 2.336359694700202e-06, + "loss": 0.9792, + "step": 5132 + }, + { + "epoch": 0.46291202597285475, + "grad_norm": 1.5790900876979495, + "learning_rate": 2.335783791034422e-06, + "loss": 0.9938, + "step": 5133 + }, + { + "epoch": 0.46300220949632503, + "grad_norm": 1.6326552319877168, + "learning_rate": 2.3352078587177173e-06, + "loss": 0.9737, + "step": 5134 + }, + { + "epoch": 0.46309239301979527, + "grad_norm": 1.6118773478046786, + "learning_rate": 2.33463189779923e-06, + "loss": 0.9978, + "step": 5135 + }, + { + "epoch": 0.46318257654326556, + "grad_norm": 2.456215630599037, + "learning_rate": 2.334055908328104e-06, + "loss": 1.0625, + "step": 5136 + }, + { + "epoch": 0.4632727600667358, + "grad_norm": 1.4932803800396144, + "learning_rate": 2.3334798903534866e-06, + "loss": 0.9029, + "step": 5137 + }, + { + "epoch": 0.4633629435902061, + "grad_norm": 2.948460037962018, + "learning_rate": 2.3329038439245252e-06, + "loss": 0.8874, + "step": 5138 + }, + { + "epoch": 0.4634531271136763, + "grad_norm": 2.2250676005520296, + "learning_rate": 2.3323277690903724e-06, + "loss": 0.9933, + "step": 5139 + }, + { + "epoch": 0.4635433106371466, + "grad_norm": 1.451199147518795, + "learning_rate": 2.3317516659001827e-06, + "loss": 0.9649, + "step": 5140 + }, + { + "epoch": 0.46363349416061683, + "grad_norm": 1.7057087749009705, + "learning_rate": 2.331175534403111e-06, + "loss": 1.0061, + "step": 5141 + }, + { + "epoch": 0.4637236776840871, + "grad_norm": 1.7142663213790978, + "learning_rate": 2.3305993746483167e-06, + "loss": 1.0744, + "step": 5142 + }, + { + "epoch": 0.46381386120755735, + "grad_norm": 4.788484971039493, + "learning_rate": 2.3300231866849606e-06, + "loss": 0.9647, + "step": 5143 + }, + { + "epoch": 0.46390404473102764, + "grad_norm": 2.000717630384043, + "learning_rate": 2.3294469705622067e-06, + "loss": 1.0299, + "step": 5144 + }, + { + "epoch": 0.4639942282544979, + "grad_norm": 2.3356840576675673, + "learning_rate": 2.3288707263292203e-06, + "loss": 0.9686, + "step": 5145 + }, + { + "epoch": 0.46408441177796816, + "grad_norm": 1.3381118925576028, + "learning_rate": 2.3282944540351707e-06, + "loss": 1.0097, + "step": 5146 + }, + { + "epoch": 0.46417459530143845, + "grad_norm": 1.9008958386094053, + "learning_rate": 2.327718153729228e-06, + "loss": 0.9616, + "step": 5147 + }, + { + "epoch": 0.4642647788249087, + "grad_norm": 1.8407071754233228, + "learning_rate": 2.327141825460566e-06, + "loss": 1.0341, + "step": 5148 + }, + { + "epoch": 0.464354962348379, + "grad_norm": 0.6742036298380265, + "learning_rate": 2.326565469278358e-06, + "loss": 0.8216, + "step": 5149 + }, + { + "epoch": 0.4644451458718492, + "grad_norm": 0.7239094358621534, + "learning_rate": 2.3259890852317846e-06, + "loss": 0.8309, + "step": 5150 + }, + { + "epoch": 0.4645353293953195, + "grad_norm": 1.661996266049851, + "learning_rate": 2.3254126733700246e-06, + "loss": 0.9871, + "step": 5151 + }, + { + "epoch": 0.46462551291878973, + "grad_norm": 1.7614133447473586, + "learning_rate": 2.324836233742262e-06, + "loss": 1.0191, + "step": 5152 + }, + { + "epoch": 0.46471569644226, + "grad_norm": 2.014517783117969, + "learning_rate": 2.3242597663976793e-06, + "loss": 1.0308, + "step": 5153 + }, + { + "epoch": 0.46480587996573025, + "grad_norm": 1.8113221254592857, + "learning_rate": 2.3236832713854663e-06, + "loss": 0.9721, + "step": 5154 + }, + { + "epoch": 0.46489606348920054, + "grad_norm": 3.2929434090902916, + "learning_rate": 2.323106748754812e-06, + "loss": 0.9867, + "step": 5155 + }, + { + "epoch": 0.4649862470126708, + "grad_norm": 1.5561613831318886, + "learning_rate": 2.3225301985549077e-06, + "loss": 1.0845, + "step": 5156 + }, + { + "epoch": 0.46507643053614106, + "grad_norm": 1.6055628245466709, + "learning_rate": 2.321953620834948e-06, + "loss": 0.9233, + "step": 5157 + }, + { + "epoch": 0.4651666140596113, + "grad_norm": 1.4520339151052377, + "learning_rate": 2.3213770156441314e-06, + "loss": 0.9579, + "step": 5158 + }, + { + "epoch": 0.4652567975830816, + "grad_norm": 2.0247519211633405, + "learning_rate": 2.3208003830316554e-06, + "loss": 1.0387, + "step": 5159 + }, + { + "epoch": 0.4653469811065518, + "grad_norm": 2.8864250395024973, + "learning_rate": 2.3202237230467215e-06, + "loss": 1.0833, + "step": 5160 + }, + { + "epoch": 0.4654371646300221, + "grad_norm": 1.4953454439526266, + "learning_rate": 2.3196470357385338e-06, + "loss": 0.9546, + "step": 5161 + }, + { + "epoch": 0.46552734815349234, + "grad_norm": 1.4976719433077186, + "learning_rate": 2.319070321156299e-06, + "loss": 0.984, + "step": 5162 + }, + { + "epoch": 0.4656175316769626, + "grad_norm": 1.4477173464502706, + "learning_rate": 2.318493579349224e-06, + "loss": 1.1233, + "step": 5163 + }, + { + "epoch": 0.46570771520043286, + "grad_norm": 1.7240422699961624, + "learning_rate": 2.317916810366522e-06, + "loss": 0.9753, + "step": 5164 + }, + { + "epoch": 0.46579789872390315, + "grad_norm": 2.8080358892956676, + "learning_rate": 2.317340014257404e-06, + "loss": 0.9354, + "step": 5165 + }, + { + "epoch": 0.4658880822473734, + "grad_norm": 1.4033696664348834, + "learning_rate": 2.316763191071086e-06, + "loss": 0.9183, + "step": 5166 + }, + { + "epoch": 0.46597826577084367, + "grad_norm": 1.6609820298032931, + "learning_rate": 2.316186340856787e-06, + "loss": 0.9471, + "step": 5167 + }, + { + "epoch": 0.4660684492943139, + "grad_norm": 2.614808051393043, + "learning_rate": 2.315609463663725e-06, + "loss": 0.9417, + "step": 5168 + }, + { + "epoch": 0.4661586328177842, + "grad_norm": 1.7385079568193158, + "learning_rate": 2.315032559541123e-06, + "loss": 0.9603, + "step": 5169 + }, + { + "epoch": 0.4662488163412545, + "grad_norm": 1.676145073937118, + "learning_rate": 2.314455628538207e-06, + "loss": 1.0115, + "step": 5170 + }, + { + "epoch": 0.4663389998647247, + "grad_norm": 1.429711847150099, + "learning_rate": 2.3138786707042023e-06, + "loss": 0.9618, + "step": 5171 + }, + { + "epoch": 0.466429183388195, + "grad_norm": 1.3838819258603317, + "learning_rate": 2.3133016860883387e-06, + "loss": 1.0039, + "step": 5172 + }, + { + "epoch": 0.46651936691166523, + "grad_norm": 1.813282830444331, + "learning_rate": 2.3127246747398475e-06, + "loss": 0.9718, + "step": 5173 + }, + { + "epoch": 0.4666095504351355, + "grad_norm": 1.3922963332119362, + "learning_rate": 2.312147636707963e-06, + "loss": 0.9878, + "step": 5174 + }, + { + "epoch": 0.46669973395860576, + "grad_norm": 2.0435231260778983, + "learning_rate": 2.3115705720419214e-06, + "loss": 0.9711, + "step": 5175 + }, + { + "epoch": 0.46678991748207604, + "grad_norm": 1.6319501141749522, + "learning_rate": 2.31099348079096e-06, + "loss": 1.0114, + "step": 5176 + }, + { + "epoch": 0.4668801010055463, + "grad_norm": 1.3576227492305109, + "learning_rate": 2.31041636300432e-06, + "loss": 1.0311, + "step": 5177 + }, + { + "epoch": 0.46697028452901657, + "grad_norm": 3.2187234821893855, + "learning_rate": 2.3098392187312445e-06, + "loss": 1.0699, + "step": 5178 + }, + { + "epoch": 0.4670604680524868, + "grad_norm": 2.2247493388394353, + "learning_rate": 2.309262048020978e-06, + "loss": 0.9683, + "step": 5179 + }, + { + "epoch": 0.4671506515759571, + "grad_norm": 1.7719708398266731, + "learning_rate": 2.308684850922769e-06, + "loss": 1.0214, + "step": 5180 + }, + { + "epoch": 0.4672408350994273, + "grad_norm": 1.663867921154023, + "learning_rate": 2.3081076274858664e-06, + "loss": 0.9856, + "step": 5181 + }, + { + "epoch": 0.4673310186228976, + "grad_norm": 1.5022717597999233, + "learning_rate": 2.307530377759522e-06, + "loss": 1.0189, + "step": 5182 + }, + { + "epoch": 0.46742120214636784, + "grad_norm": 2.136207533359021, + "learning_rate": 2.30695310179299e-06, + "loss": 0.9169, + "step": 5183 + }, + { + "epoch": 0.46751138566983813, + "grad_norm": 1.4202463233227338, + "learning_rate": 2.3063757996355267e-06, + "loss": 0.9899, + "step": 5184 + }, + { + "epoch": 0.46760156919330836, + "grad_norm": 1.857246976333644, + "learning_rate": 2.3057984713363903e-06, + "loss": 1.0239, + "step": 5185 + }, + { + "epoch": 0.46769175271677865, + "grad_norm": 0.6888772218004231, + "learning_rate": 2.3052211169448436e-06, + "loss": 0.7941, + "step": 5186 + }, + { + "epoch": 0.4677819362402489, + "grad_norm": 1.8304338140080316, + "learning_rate": 2.3046437365101474e-06, + "loss": 0.9318, + "step": 5187 + }, + { + "epoch": 0.4678721197637192, + "grad_norm": 1.7813369328215964, + "learning_rate": 2.3040663300815673e-06, + "loss": 0.9687, + "step": 5188 + }, + { + "epoch": 0.4679623032871894, + "grad_norm": 0.7310727727937424, + "learning_rate": 2.3034888977083723e-06, + "loss": 0.8018, + "step": 5189 + }, + { + "epoch": 0.4680524868106597, + "grad_norm": 2.885667498504682, + "learning_rate": 2.30291143943983e-06, + "loss": 0.9, + "step": 5190 + }, + { + "epoch": 0.46814267033412993, + "grad_norm": 1.5208434884067616, + "learning_rate": 2.3023339553252145e-06, + "loss": 0.9475, + "step": 5191 + }, + { + "epoch": 0.4682328538576002, + "grad_norm": 1.7174906885597099, + "learning_rate": 2.301756445413799e-06, + "loss": 1.0251, + "step": 5192 + }, + { + "epoch": 0.4683230373810705, + "grad_norm": 1.7220505668180435, + "learning_rate": 2.3011789097548585e-06, + "loss": 0.9937, + "step": 5193 + }, + { + "epoch": 0.46841322090454074, + "grad_norm": 1.5890338156182657, + "learning_rate": 2.3006013483976738e-06, + "loss": 0.9881, + "step": 5194 + }, + { + "epoch": 0.468503404428011, + "grad_norm": 1.7498997250847155, + "learning_rate": 2.300023761391524e-06, + "loss": 0.9861, + "step": 5195 + }, + { + "epoch": 0.46859358795148126, + "grad_norm": 2.0003306592353316, + "learning_rate": 2.299446148785693e-06, + "loss": 1.039, + "step": 5196 + }, + { + "epoch": 0.46868377147495155, + "grad_norm": 2.0531888072232154, + "learning_rate": 2.2988685106294654e-06, + "loss": 0.9645, + "step": 5197 + }, + { + "epoch": 0.4687739549984218, + "grad_norm": 1.768090274020756, + "learning_rate": 2.2982908469721284e-06, + "loss": 0.8893, + "step": 5198 + }, + { + "epoch": 0.46886413852189207, + "grad_norm": 2.0219958261886366, + "learning_rate": 2.2977131578629714e-06, + "loss": 1.0571, + "step": 5199 + }, + { + "epoch": 0.4689543220453623, + "grad_norm": 1.6433803128123028, + "learning_rate": 2.297135443351286e-06, + "loss": 1.011, + "step": 5200 + }, + { + "epoch": 0.4690445055688326, + "grad_norm": 1.2939682062390983, + "learning_rate": 2.296557703486367e-06, + "loss": 0.9995, + "step": 5201 + }, + { + "epoch": 0.4691346890923028, + "grad_norm": 1.992070093622223, + "learning_rate": 2.295979938317509e-06, + "loss": 1.0198, + "step": 5202 + }, + { + "epoch": 0.4692248726157731, + "grad_norm": 3.2595877441730097, + "learning_rate": 2.295402147894011e-06, + "loss": 0.9894, + "step": 5203 + }, + { + "epoch": 0.46931505613924335, + "grad_norm": 25.591689477494448, + "learning_rate": 2.2948243322651723e-06, + "loss": 0.9179, + "step": 5204 + }, + { + "epoch": 0.46940523966271364, + "grad_norm": 2.904077353579572, + "learning_rate": 2.2942464914802962e-06, + "loss": 0.9948, + "step": 5205 + }, + { + "epoch": 0.46949542318618387, + "grad_norm": 2.7236504949116758, + "learning_rate": 2.293668625588687e-06, + "loss": 0.9679, + "step": 5206 + }, + { + "epoch": 0.46958560670965416, + "grad_norm": 2.439587799337476, + "learning_rate": 2.293090734639651e-06, + "loss": 1.0339, + "step": 5207 + }, + { + "epoch": 0.4696757902331244, + "grad_norm": 1.6635662646337213, + "learning_rate": 2.2925128186824983e-06, + "loss": 1.1065, + "step": 5208 + }, + { + "epoch": 0.4697659737565947, + "grad_norm": 1.9795789657263032, + "learning_rate": 2.2919348777665384e-06, + "loss": 1.092, + "step": 5209 + }, + { + "epoch": 0.4698561572800649, + "grad_norm": 1.9554535070016479, + "learning_rate": 2.2913569119410856e-06, + "loss": 0.9946, + "step": 5210 + }, + { + "epoch": 0.4699463408035352, + "grad_norm": 1.6086250706013923, + "learning_rate": 2.290778921255454e-06, + "loss": 1.0184, + "step": 5211 + }, + { + "epoch": 0.47003652432700543, + "grad_norm": 2.617300275963403, + "learning_rate": 2.2902009057589613e-06, + "loss": 1.0264, + "step": 5212 + }, + { + "epoch": 0.4701267078504757, + "grad_norm": 1.5092684974769057, + "learning_rate": 2.2896228655009276e-06, + "loss": 0.9482, + "step": 5213 + }, + { + "epoch": 0.47021689137394596, + "grad_norm": 1.7286185885867764, + "learning_rate": 2.289044800530674e-06, + "loss": 0.958, + "step": 5214 + }, + { + "epoch": 0.47030707489741624, + "grad_norm": 1.770191936646693, + "learning_rate": 2.2884667108975245e-06, + "loss": 0.9586, + "step": 5215 + }, + { + "epoch": 0.4703972584208865, + "grad_norm": 1.7006744112899896, + "learning_rate": 2.287888596650804e-06, + "loss": 1.0292, + "step": 5216 + }, + { + "epoch": 0.47048744194435677, + "grad_norm": 1.9904011215721968, + "learning_rate": 2.287310457839841e-06, + "loss": 0.9886, + "step": 5217 + }, + { + "epoch": 0.47057762546782705, + "grad_norm": 2.0654337276482355, + "learning_rate": 2.286732294513966e-06, + "loss": 1.0617, + "step": 5218 + }, + { + "epoch": 0.4706678089912973, + "grad_norm": 1.7778557679169318, + "learning_rate": 2.2861541067225106e-06, + "loss": 0.9953, + "step": 5219 + }, + { + "epoch": 0.4707579925147676, + "grad_norm": 1.8909125621678762, + "learning_rate": 2.2855758945148095e-06, + "loss": 0.9736, + "step": 5220 + }, + { + "epoch": 0.4708481760382378, + "grad_norm": 1.7318511332392361, + "learning_rate": 2.2849976579401977e-06, + "loss": 1.0544, + "step": 5221 + }, + { + "epoch": 0.4709383595617081, + "grad_norm": 2.1547326540347806, + "learning_rate": 2.284419397048014e-06, + "loss": 0.9841, + "step": 5222 + }, + { + "epoch": 0.47102854308517833, + "grad_norm": 1.7537805049143762, + "learning_rate": 2.2838411118875997e-06, + "loss": 0.9995, + "step": 5223 + }, + { + "epoch": 0.4711187266086486, + "grad_norm": 1.7477040215666158, + "learning_rate": 2.283262802508296e-06, + "loss": 0.9172, + "step": 5224 + }, + { + "epoch": 0.47120891013211885, + "grad_norm": 1.3600637948977656, + "learning_rate": 2.2826844689594492e-06, + "loss": 0.9715, + "step": 5225 + }, + { + "epoch": 0.47129909365558914, + "grad_norm": 1.7812011444351972, + "learning_rate": 2.282106111290404e-06, + "loss": 1.0681, + "step": 5226 + }, + { + "epoch": 0.4713892771790594, + "grad_norm": 0.6856920789524535, + "learning_rate": 2.2815277295505098e-06, + "loss": 0.8381, + "step": 5227 + }, + { + "epoch": 0.47147946070252966, + "grad_norm": 1.7679202259803177, + "learning_rate": 2.2809493237891174e-06, + "loss": 0.9394, + "step": 5228 + }, + { + "epoch": 0.4715696442259999, + "grad_norm": 1.9476753194145495, + "learning_rate": 2.2803708940555796e-06, + "loss": 1.0214, + "step": 5229 + }, + { + "epoch": 0.4716598277494702, + "grad_norm": 1.4450723809458421, + "learning_rate": 2.2797924403992514e-06, + "loss": 1.061, + "step": 5230 + }, + { + "epoch": 0.4717500112729404, + "grad_norm": 2.25504426307966, + "learning_rate": 2.2792139628694892e-06, + "loss": 1.0107, + "step": 5231 + }, + { + "epoch": 0.4718401947964107, + "grad_norm": 1.4461865230995365, + "learning_rate": 2.2786354615156524e-06, + "loss": 0.9512, + "step": 5232 + }, + { + "epoch": 0.47193037831988094, + "grad_norm": 2.2043411469462617, + "learning_rate": 2.2780569363871016e-06, + "loss": 1.0574, + "step": 5233 + }, + { + "epoch": 0.4720205618433512, + "grad_norm": 0.6214787471571572, + "learning_rate": 2.277478387533199e-06, + "loss": 0.8096, + "step": 5234 + }, + { + "epoch": 0.47211074536682146, + "grad_norm": 1.634842628114759, + "learning_rate": 2.276899815003311e-06, + "loss": 1.0302, + "step": 5235 + }, + { + "epoch": 0.47220092889029175, + "grad_norm": 1.9444163146330586, + "learning_rate": 2.2763212188468045e-06, + "loss": 0.9832, + "step": 5236 + }, + { + "epoch": 0.472291112413762, + "grad_norm": 1.3939733591421053, + "learning_rate": 2.2757425991130473e-06, + "loss": 1.0054, + "step": 5237 + }, + { + "epoch": 0.47238129593723227, + "grad_norm": 0.754988017776598, + "learning_rate": 2.2751639558514117e-06, + "loss": 0.8226, + "step": 5238 + }, + { + "epoch": 0.4724714794607025, + "grad_norm": 1.3720578013219622, + "learning_rate": 2.2745852891112697e-06, + "loss": 1.0342, + "step": 5239 + }, + { + "epoch": 0.4725616629841728, + "grad_norm": 2.5401181887365487, + "learning_rate": 2.274006598941997e-06, + "loss": 1.0003, + "step": 5240 + }, + { + "epoch": 0.4726518465076431, + "grad_norm": 8.386983013564784, + "learning_rate": 2.27342788539297e-06, + "loss": 0.9754, + "step": 5241 + }, + { + "epoch": 0.4727420300311133, + "grad_norm": 1.9387476195880928, + "learning_rate": 2.2728491485135684e-06, + "loss": 0.9512, + "step": 5242 + }, + { + "epoch": 0.4728322135545836, + "grad_norm": 1.9920232368515725, + "learning_rate": 2.272270388353173e-06, + "loss": 1.0448, + "step": 5243 + }, + { + "epoch": 0.47292239707805384, + "grad_norm": 3.0449624343333803, + "learning_rate": 2.2716916049611666e-06, + "loss": 0.9633, + "step": 5244 + }, + { + "epoch": 0.4730125806015241, + "grad_norm": 3.2609767412457575, + "learning_rate": 2.2711127983869346e-06, + "loss": 0.9754, + "step": 5245 + }, + { + "epoch": 0.47310276412499436, + "grad_norm": 1.4110752397560176, + "learning_rate": 2.270533968679864e-06, + "loss": 1.0011, + "step": 5246 + }, + { + "epoch": 0.47319294764846465, + "grad_norm": 1.53814817272403, + "learning_rate": 2.269955115889343e-06, + "loss": 0.9973, + "step": 5247 + }, + { + "epoch": 0.4732831311719349, + "grad_norm": 1.725670078991665, + "learning_rate": 2.269376240064763e-06, + "loss": 0.9566, + "step": 5248 + }, + { + "epoch": 0.47337331469540517, + "grad_norm": 2.357667014232502, + "learning_rate": 2.268797341255517e-06, + "loss": 1.0117, + "step": 5249 + }, + { + "epoch": 0.4734634982188754, + "grad_norm": 4.030737319401092, + "learning_rate": 2.268218419511e-06, + "loss": 0.9088, + "step": 5250 + }, + { + "epoch": 0.4735536817423457, + "grad_norm": 1.6941954980521337, + "learning_rate": 2.267639474880608e-06, + "loss": 0.9107, + "step": 5251 + }, + { + "epoch": 0.4736438652658159, + "grad_norm": 1.4240417738798545, + "learning_rate": 2.2670605074137407e-06, + "loss": 0.9732, + "step": 5252 + }, + { + "epoch": 0.4737340487892862, + "grad_norm": 3.009765625, + "learning_rate": 2.2664815171597983e-06, + "loss": 0.9742, + "step": 5253 + }, + { + "epoch": 0.47382423231275644, + "grad_norm": 1.6186802558160915, + "learning_rate": 2.265902504168183e-06, + "loss": 0.9363, + "step": 5254 + }, + { + "epoch": 0.47391441583622673, + "grad_norm": 1.7939316235613407, + "learning_rate": 2.2653234684883007e-06, + "loss": 0.9491, + "step": 5255 + }, + { + "epoch": 0.47400459935969697, + "grad_norm": 2.091876544022938, + "learning_rate": 2.264744410169556e-06, + "loss": 0.9106, + "step": 5256 + }, + { + "epoch": 0.47409478288316725, + "grad_norm": 2.417713651003421, + "learning_rate": 2.264165329261359e-06, + "loss": 0.9754, + "step": 5257 + }, + { + "epoch": 0.4741849664066375, + "grad_norm": 0.5733026735800046, + "learning_rate": 2.26358622581312e-06, + "loss": 0.8122, + "step": 5258 + }, + { + "epoch": 0.4742751499301078, + "grad_norm": 1.709110625916975, + "learning_rate": 2.2630070998742504e-06, + "loss": 1.0532, + "step": 5259 + }, + { + "epoch": 0.474365333453578, + "grad_norm": 1.3356417501467628, + "learning_rate": 2.262427951494165e-06, + "loss": 1.0129, + "step": 5260 + }, + { + "epoch": 0.4744555169770483, + "grad_norm": 0.67175733289986, + "learning_rate": 2.2618487807222794e-06, + "loss": 0.8524, + "step": 5261 + }, + { + "epoch": 0.47454570050051853, + "grad_norm": 1.6790330188405866, + "learning_rate": 2.261269587608012e-06, + "loss": 1.0742, + "step": 5262 + }, + { + "epoch": 0.4746358840239888, + "grad_norm": 1.3837008451502908, + "learning_rate": 2.260690372200783e-06, + "loss": 0.9974, + "step": 5263 + }, + { + "epoch": 0.47472606754745905, + "grad_norm": 1.4383896272471999, + "learning_rate": 2.2601111345500138e-06, + "loss": 1.0765, + "step": 5264 + }, + { + "epoch": 0.47481625107092934, + "grad_norm": 1.641801321571386, + "learning_rate": 2.2595318747051286e-06, + "loss": 0.981, + "step": 5265 + }, + { + "epoch": 0.47490643459439963, + "grad_norm": 1.6513534716775906, + "learning_rate": 2.258952592715553e-06, + "loss": 1.0183, + "step": 5266 + }, + { + "epoch": 0.47499661811786986, + "grad_norm": 0.7048138997804098, + "learning_rate": 2.2583732886307142e-06, + "loss": 0.8548, + "step": 5267 + }, + { + "epoch": 0.47508680164134015, + "grad_norm": 2.6583912017008555, + "learning_rate": 2.2577939625000414e-06, + "loss": 0.9407, + "step": 5268 + }, + { + "epoch": 0.4751769851648104, + "grad_norm": 0.679299232693793, + "learning_rate": 2.257214614372967e-06, + "loss": 0.8146, + "step": 5269 + }, + { + "epoch": 0.4752671686882807, + "grad_norm": 1.5129071278833006, + "learning_rate": 2.2566352442989227e-06, + "loss": 1.0172, + "step": 5270 + }, + { + "epoch": 0.4753573522117509, + "grad_norm": 3.9279332274221406, + "learning_rate": 2.256055852327344e-06, + "loss": 0.9979, + "step": 5271 + }, + { + "epoch": 0.4754475357352212, + "grad_norm": 2.335920557069094, + "learning_rate": 2.2554764385076685e-06, + "loss": 0.9392, + "step": 5272 + }, + { + "epoch": 0.4755377192586914, + "grad_norm": 1.5896092647305538, + "learning_rate": 2.2548970028893348e-06, + "loss": 1.0062, + "step": 5273 + }, + { + "epoch": 0.4756279027821617, + "grad_norm": 1.6299145548605425, + "learning_rate": 2.254317545521783e-06, + "loss": 1.0041, + "step": 5274 + }, + { + "epoch": 0.47571808630563195, + "grad_norm": 1.6491997569990475, + "learning_rate": 2.253738066454457e-06, + "loss": 1.0214, + "step": 5275 + }, + { + "epoch": 0.47580826982910224, + "grad_norm": 1.9969992776562833, + "learning_rate": 2.2531585657367986e-06, + "loss": 0.971, + "step": 5276 + }, + { + "epoch": 0.47589845335257247, + "grad_norm": 1.7469451671832106, + "learning_rate": 2.252579043418256e-06, + "loss": 0.9453, + "step": 5277 + }, + { + "epoch": 0.47598863687604276, + "grad_norm": 1.6188259210051927, + "learning_rate": 2.251999499548277e-06, + "loss": 0.9656, + "step": 5278 + }, + { + "epoch": 0.476078820399513, + "grad_norm": 2.067276942479933, + "learning_rate": 2.251419934176311e-06, + "loss": 0.9508, + "step": 5279 + }, + { + "epoch": 0.4761690039229833, + "grad_norm": 1.3835461066048946, + "learning_rate": 2.25084034735181e-06, + "loss": 1.0431, + "step": 5280 + }, + { + "epoch": 0.4762591874464535, + "grad_norm": 1.5913426037712666, + "learning_rate": 2.2502607391242274e-06, + "loss": 1.0427, + "step": 5281 + }, + { + "epoch": 0.4763493709699238, + "grad_norm": 0.6676910001396458, + "learning_rate": 2.2496811095430182e-06, + "loss": 0.8191, + "step": 5282 + }, + { + "epoch": 0.47643955449339404, + "grad_norm": 1.5650728786427988, + "learning_rate": 2.249101458657641e-06, + "loss": 0.9885, + "step": 5283 + }, + { + "epoch": 0.4765297380168643, + "grad_norm": 1.991659894902463, + "learning_rate": 2.2485217865175526e-06, + "loss": 1.0653, + "step": 5284 + }, + { + "epoch": 0.47661992154033456, + "grad_norm": 1.7656665898378794, + "learning_rate": 2.2479420931722156e-06, + "loss": 1.1264, + "step": 5285 + }, + { + "epoch": 0.47671010506380485, + "grad_norm": 0.5847417200944834, + "learning_rate": 2.2473623786710923e-06, + "loss": 0.8562, + "step": 5286 + }, + { + "epoch": 0.4768002885872751, + "grad_norm": 1.816750180077639, + "learning_rate": 2.2467826430636465e-06, + "loss": 0.9056, + "step": 5287 + }, + { + "epoch": 0.47689047211074537, + "grad_norm": 4.022358870131674, + "learning_rate": 2.246202886399345e-06, + "loss": 1.0175, + "step": 5288 + }, + { + "epoch": 0.47698065563421566, + "grad_norm": 1.5817337321516127, + "learning_rate": 2.2456231087276556e-06, + "loss": 0.9903, + "step": 5289 + }, + { + "epoch": 0.4770708391576859, + "grad_norm": 2.3431286814801386, + "learning_rate": 2.245043310098048e-06, + "loss": 0.9204, + "step": 5290 + }, + { + "epoch": 0.4771610226811562, + "grad_norm": 2.962658706470077, + "learning_rate": 2.244463490559995e-06, + "loss": 1.0362, + "step": 5291 + }, + { + "epoch": 0.4772512062046264, + "grad_norm": 1.8419646978526423, + "learning_rate": 2.2438836501629683e-06, + "loss": 1.0033, + "step": 5292 + }, + { + "epoch": 0.4773413897280967, + "grad_norm": 1.639037754101136, + "learning_rate": 2.2433037889564437e-06, + "loss": 1.0005, + "step": 5293 + }, + { + "epoch": 0.47743157325156693, + "grad_norm": 3.0411503148861176, + "learning_rate": 2.242723906989899e-06, + "loss": 0.9305, + "step": 5294 + }, + { + "epoch": 0.4775217567750372, + "grad_norm": 1.460955533003052, + "learning_rate": 2.2421440043128114e-06, + "loss": 0.9258, + "step": 5295 + }, + { + "epoch": 0.47761194029850745, + "grad_norm": 2.2659225794603834, + "learning_rate": 2.241564080974662e-06, + "loss": 0.9058, + "step": 5296 + }, + { + "epoch": 0.47770212382197774, + "grad_norm": 1.7291180094418577, + "learning_rate": 2.2409841370249343e-06, + "loss": 0.9408, + "step": 5297 + }, + { + "epoch": 0.477792307345448, + "grad_norm": 1.9737562077758652, + "learning_rate": 2.2404041725131106e-06, + "loss": 1.0778, + "step": 5298 + }, + { + "epoch": 0.47788249086891826, + "grad_norm": 1.769925576708982, + "learning_rate": 2.239824187488677e-06, + "loss": 0.9742, + "step": 5299 + }, + { + "epoch": 0.4779726743923885, + "grad_norm": 0.7308135825184789, + "learning_rate": 2.239244182001122e-06, + "loss": 0.8553, + "step": 5300 + }, + { + "epoch": 0.4780628579158588, + "grad_norm": 1.3532991222385793, + "learning_rate": 2.2386641560999336e-06, + "loss": 1.0063, + "step": 5301 + }, + { + "epoch": 0.478153041439329, + "grad_norm": 1.8145005596682457, + "learning_rate": 2.238084109834604e-06, + "loss": 1.0242, + "step": 5302 + }, + { + "epoch": 0.4782432249627993, + "grad_norm": 1.8807726054711134, + "learning_rate": 2.237504043254625e-06, + "loss": 0.9937, + "step": 5303 + }, + { + "epoch": 0.47833340848626954, + "grad_norm": 1.3851243430140294, + "learning_rate": 2.2369239564094915e-06, + "loss": 0.9636, + "step": 5304 + }, + { + "epoch": 0.47842359200973983, + "grad_norm": 1.532503043540434, + "learning_rate": 2.2363438493486995e-06, + "loss": 0.9231, + "step": 5305 + }, + { + "epoch": 0.47851377553321006, + "grad_norm": 1.6290603674562372, + "learning_rate": 2.235763722121747e-06, + "loss": 0.8811, + "step": 5306 + }, + { + "epoch": 0.47860395905668035, + "grad_norm": 1.7643565795239557, + "learning_rate": 2.2351835747781346e-06, + "loss": 1.0108, + "step": 5307 + }, + { + "epoch": 0.4786941425801506, + "grad_norm": 1.5726624929775102, + "learning_rate": 2.234603407367362e-06, + "loss": 1.0348, + "step": 5308 + }, + { + "epoch": 0.47878432610362087, + "grad_norm": 2.4356561191167403, + "learning_rate": 2.2340232199389337e-06, + "loss": 1.0051, + "step": 5309 + }, + { + "epoch": 0.4788745096270911, + "grad_norm": 1.360875715392727, + "learning_rate": 2.2334430125423538e-06, + "loss": 0.8746, + "step": 5310 + }, + { + "epoch": 0.4789646931505614, + "grad_norm": 2.3963639929396607, + "learning_rate": 2.232862785227128e-06, + "loss": 1.0415, + "step": 5311 + }, + { + "epoch": 0.4790548766740317, + "grad_norm": 1.7809290680465655, + "learning_rate": 2.232282538042766e-06, + "loss": 1.0262, + "step": 5312 + }, + { + "epoch": 0.4791450601975019, + "grad_norm": 2.917078334275715, + "learning_rate": 2.231702271038777e-06, + "loss": 1.042, + "step": 5313 + }, + { + "epoch": 0.4792352437209722, + "grad_norm": 1.9247857358513356, + "learning_rate": 2.231121984264673e-06, + "loss": 0.9672, + "step": 5314 + }, + { + "epoch": 0.47932542724444244, + "grad_norm": 1.8112147147859778, + "learning_rate": 2.2305416777699665e-06, + "loss": 1.021, + "step": 5315 + }, + { + "epoch": 0.4794156107679127, + "grad_norm": 2.123971914784586, + "learning_rate": 2.229961351604173e-06, + "loss": 0.9962, + "step": 5316 + }, + { + "epoch": 0.47950579429138296, + "grad_norm": 1.5179643961735523, + "learning_rate": 2.2293810058168085e-06, + "loss": 0.9515, + "step": 5317 + }, + { + "epoch": 0.47959597781485325, + "grad_norm": 1.3913614701739119, + "learning_rate": 2.2288006404573922e-06, + "loss": 1.0156, + "step": 5318 + }, + { + "epoch": 0.4796861613383235, + "grad_norm": 1.9207844935543161, + "learning_rate": 2.228220255575444e-06, + "loss": 0.9772, + "step": 5319 + }, + { + "epoch": 0.47977634486179377, + "grad_norm": 1.4452115410262367, + "learning_rate": 2.2276398512204847e-06, + "loss": 0.9395, + "step": 5320 + }, + { + "epoch": 0.479866528385264, + "grad_norm": 1.5290627720573553, + "learning_rate": 2.2270594274420382e-06, + "loss": 0.9998, + "step": 5321 + }, + { + "epoch": 0.4799567119087343, + "grad_norm": 0.6621502663917527, + "learning_rate": 2.22647898428963e-06, + "loss": 0.8454, + "step": 5322 + }, + { + "epoch": 0.4800468954322045, + "grad_norm": 1.4044190673297563, + "learning_rate": 2.225898521812785e-06, + "loss": 1.0369, + "step": 5323 + }, + { + "epoch": 0.4801370789556748, + "grad_norm": 1.6784267218227207, + "learning_rate": 2.2253180400610337e-06, + "loss": 1.0413, + "step": 5324 + }, + { + "epoch": 0.48022726247914505, + "grad_norm": 2.1135981352137296, + "learning_rate": 2.2247375390839037e-06, + "loss": 0.9725, + "step": 5325 + }, + { + "epoch": 0.48031744600261533, + "grad_norm": 2.038935164060144, + "learning_rate": 2.224157018930928e-06, + "loss": 0.9635, + "step": 5326 + }, + { + "epoch": 0.48040762952608557, + "grad_norm": 1.4268298330021596, + "learning_rate": 2.2235764796516395e-06, + "loss": 0.9945, + "step": 5327 + }, + { + "epoch": 0.48049781304955586, + "grad_norm": 1.4208569917172753, + "learning_rate": 2.222995921295573e-06, + "loss": 0.9917, + "step": 5328 + }, + { + "epoch": 0.4805879965730261, + "grad_norm": 2.027571178461567, + "learning_rate": 2.222415343912265e-06, + "loss": 0.97, + "step": 5329 + }, + { + "epoch": 0.4806781800964964, + "grad_norm": 2.1120741917265646, + "learning_rate": 2.221834747551254e-06, + "loss": 1.0601, + "step": 5330 + }, + { + "epoch": 0.4807683636199666, + "grad_norm": 1.7805566609511942, + "learning_rate": 2.221254132262078e-06, + "loss": 0.9668, + "step": 5331 + }, + { + "epoch": 0.4808585471434369, + "grad_norm": 1.8284051142569089, + "learning_rate": 2.2206734980942802e-06, + "loss": 1.0351, + "step": 5332 + }, + { + "epoch": 0.48094873066690713, + "grad_norm": 2.0397396205708618, + "learning_rate": 2.2200928450974024e-06, + "loss": 0.9732, + "step": 5333 + }, + { + "epoch": 0.4810389141903774, + "grad_norm": 1.658757309254322, + "learning_rate": 2.21951217332099e-06, + "loss": 1.0145, + "step": 5334 + }, + { + "epoch": 0.48112909771384765, + "grad_norm": 1.7781754444656632, + "learning_rate": 2.2189314828145883e-06, + "loss": 0.9065, + "step": 5335 + }, + { + "epoch": 0.48121928123731794, + "grad_norm": 1.5937084117773928, + "learning_rate": 2.2183507736277453e-06, + "loss": 0.9765, + "step": 5336 + }, + { + "epoch": 0.48130946476078823, + "grad_norm": 1.527695871334661, + "learning_rate": 2.2177700458100107e-06, + "loss": 1.0071, + "step": 5337 + }, + { + "epoch": 0.48139964828425846, + "grad_norm": 8.628579101992166, + "learning_rate": 2.2171892994109346e-06, + "loss": 0.9679, + "step": 5338 + }, + { + "epoch": 0.48148983180772875, + "grad_norm": 3.0572495898980248, + "learning_rate": 2.21660853448007e-06, + "loss": 1.0578, + "step": 5339 + }, + { + "epoch": 0.481580015331199, + "grad_norm": 2.0569544133469866, + "learning_rate": 2.2160277510669703e-06, + "loss": 0.9482, + "step": 5340 + }, + { + "epoch": 0.4816701988546693, + "grad_norm": 1.3543773731726683, + "learning_rate": 2.215446949221193e-06, + "loss": 0.9853, + "step": 5341 + }, + { + "epoch": 0.4817603823781395, + "grad_norm": 1.7931895400592093, + "learning_rate": 2.2148661289922924e-06, + "loss": 1.0006, + "step": 5342 + }, + { + "epoch": 0.4818505659016098, + "grad_norm": 1.8421838218386932, + "learning_rate": 2.21428529042983e-06, + "loss": 1.057, + "step": 5343 + }, + { + "epoch": 0.48194074942508003, + "grad_norm": 1.6921100517717897, + "learning_rate": 2.2137044335833647e-06, + "loss": 1.042, + "step": 5344 + }, + { + "epoch": 0.4820309329485503, + "grad_norm": 3.62184808439666, + "learning_rate": 2.213123558502459e-06, + "loss": 0.9764, + "step": 5345 + }, + { + "epoch": 0.48212111647202055, + "grad_norm": 1.6552127253108235, + "learning_rate": 2.2125426652366763e-06, + "loss": 1.0101, + "step": 5346 + }, + { + "epoch": 0.48221129999549084, + "grad_norm": 1.7655065040997941, + "learning_rate": 2.211961753835581e-06, + "loss": 1.0895, + "step": 5347 + }, + { + "epoch": 0.48230148351896107, + "grad_norm": 1.4295513317529305, + "learning_rate": 2.21138082434874e-06, + "loss": 0.9854, + "step": 5348 + }, + { + "epoch": 0.48239166704243136, + "grad_norm": 1.783926976706568, + "learning_rate": 2.210799876825722e-06, + "loss": 1.0216, + "step": 5349 + }, + { + "epoch": 0.4824818505659016, + "grad_norm": 1.6574150162626176, + "learning_rate": 2.210218911316096e-06, + "loss": 0.9525, + "step": 5350 + }, + { + "epoch": 0.4825720340893719, + "grad_norm": 1.7863150744000025, + "learning_rate": 2.2096379278694336e-06, + "loss": 0.9467, + "step": 5351 + }, + { + "epoch": 0.4826622176128421, + "grad_norm": 1.7185503063369332, + "learning_rate": 2.2090569265353074e-06, + "loss": 0.9379, + "step": 5352 + }, + { + "epoch": 0.4827524011363124, + "grad_norm": 2.012146661672512, + "learning_rate": 2.2084759073632912e-06, + "loss": 0.9903, + "step": 5353 + }, + { + "epoch": 0.48284258465978264, + "grad_norm": 1.5718688084990433, + "learning_rate": 2.2078948704029606e-06, + "loss": 0.9517, + "step": 5354 + }, + { + "epoch": 0.4829327681832529, + "grad_norm": 1.7349295889338996, + "learning_rate": 2.2073138157038935e-06, + "loss": 0.9145, + "step": 5355 + }, + { + "epoch": 0.48302295170672316, + "grad_norm": 1.6601676671252634, + "learning_rate": 2.2067327433156687e-06, + "loss": 0.9823, + "step": 5356 + }, + { + "epoch": 0.48311313523019345, + "grad_norm": 1.635827160805969, + "learning_rate": 2.2061516532878667e-06, + "loss": 1.0076, + "step": 5357 + }, + { + "epoch": 0.4832033187536637, + "grad_norm": 1.548533465724979, + "learning_rate": 2.2055705456700686e-06, + "loss": 1.0387, + "step": 5358 + }, + { + "epoch": 0.48329350227713397, + "grad_norm": 2.3389003169952702, + "learning_rate": 2.204989420511858e-06, + "loss": 0.8915, + "step": 5359 + }, + { + "epoch": 0.48338368580060426, + "grad_norm": 1.852022653049813, + "learning_rate": 2.20440827786282e-06, + "loss": 0.9397, + "step": 5360 + }, + { + "epoch": 0.4834738693240745, + "grad_norm": 1.566095792298449, + "learning_rate": 2.20382711777254e-06, + "loss": 0.9978, + "step": 5361 + }, + { + "epoch": 0.4835640528475448, + "grad_norm": 1.7112923868083425, + "learning_rate": 2.203245940290607e-06, + "loss": 1.0057, + "step": 5362 + }, + { + "epoch": 0.483654236371015, + "grad_norm": 1.941924427043102, + "learning_rate": 2.2026647454666097e-06, + "loss": 0.9882, + "step": 5363 + }, + { + "epoch": 0.4837444198944853, + "grad_norm": 2.004022368102579, + "learning_rate": 2.2020835333501384e-06, + "loss": 1.0307, + "step": 5364 + }, + { + "epoch": 0.48383460341795553, + "grad_norm": 1.6699254443035685, + "learning_rate": 2.2015023039907863e-06, + "loss": 1.049, + "step": 5365 + }, + { + "epoch": 0.4839247869414258, + "grad_norm": 1.5232978952460785, + "learning_rate": 2.2009210574381464e-06, + "loss": 0.978, + "step": 5366 + }, + { + "epoch": 0.48401497046489605, + "grad_norm": 7.4459539499388105, + "learning_rate": 2.2003397937418134e-06, + "loss": 0.9343, + "step": 5367 + }, + { + "epoch": 0.48410515398836634, + "grad_norm": 0.670599702960562, + "learning_rate": 2.1997585129513852e-06, + "loss": 0.8396, + "step": 5368 + }, + { + "epoch": 0.4841953375118366, + "grad_norm": 1.609312037745559, + "learning_rate": 2.1991772151164595e-06, + "loss": 1.0177, + "step": 5369 + }, + { + "epoch": 0.48428552103530687, + "grad_norm": 1.6452072819959258, + "learning_rate": 2.1985959002866346e-06, + "loss": 1.0371, + "step": 5370 + }, + { + "epoch": 0.4843757045587771, + "grad_norm": 10.95969775931736, + "learning_rate": 2.198014568511513e-06, + "loss": 1.0552, + "step": 5371 + }, + { + "epoch": 0.4844658880822474, + "grad_norm": 0.6839138916230154, + "learning_rate": 2.1974332198406965e-06, + "loss": 0.8122, + "step": 5372 + }, + { + "epoch": 0.4845560716057176, + "grad_norm": 2.122998697669501, + "learning_rate": 2.196851854323789e-06, + "loss": 0.9703, + "step": 5373 + }, + { + "epoch": 0.4846462551291879, + "grad_norm": 3.227241204580393, + "learning_rate": 2.196270472010396e-06, + "loss": 0.9753, + "step": 5374 + }, + { + "epoch": 0.48473643865265814, + "grad_norm": 1.9166418018663207, + "learning_rate": 2.195689072950124e-06, + "loss": 0.9885, + "step": 5375 + }, + { + "epoch": 0.48482662217612843, + "grad_norm": 0.5809628454120387, + "learning_rate": 2.195107657192581e-06, + "loss": 0.7999, + "step": 5376 + }, + { + "epoch": 0.48491680569959866, + "grad_norm": 0.5863243351965712, + "learning_rate": 2.194526224787378e-06, + "loss": 0.7539, + "step": 5377 + }, + { + "epoch": 0.48500698922306895, + "grad_norm": 1.76113854980888, + "learning_rate": 2.1939447757841236e-06, + "loss": 0.9596, + "step": 5378 + }, + { + "epoch": 0.4850971727465392, + "grad_norm": 1.709400758427168, + "learning_rate": 2.193363310232432e-06, + "loss": 1.0548, + "step": 5379 + }, + { + "epoch": 0.4851873562700095, + "grad_norm": 1.423594095805552, + "learning_rate": 2.192781828181917e-06, + "loss": 1.0917, + "step": 5380 + }, + { + "epoch": 0.4852775397934797, + "grad_norm": 1.3389966416605286, + "learning_rate": 2.192200329682193e-06, + "loss": 1.0166, + "step": 5381 + }, + { + "epoch": 0.48536772331695, + "grad_norm": 1.5277988702107954, + "learning_rate": 2.1916188147828767e-06, + "loss": 0.9894, + "step": 5382 + }, + { + "epoch": 0.48545790684042023, + "grad_norm": 1.5054135386186978, + "learning_rate": 2.191037283533587e-06, + "loss": 1.005, + "step": 5383 + }, + { + "epoch": 0.4855480903638905, + "grad_norm": 2.315242069515432, + "learning_rate": 2.1904557359839428e-06, + "loss": 0.9876, + "step": 5384 + }, + { + "epoch": 0.4856382738873608, + "grad_norm": 1.8432727131294628, + "learning_rate": 2.189874172183565e-06, + "loss": 0.972, + "step": 5385 + }, + { + "epoch": 0.48572845741083104, + "grad_norm": 1.574183461288328, + "learning_rate": 2.1892925921820763e-06, + "loss": 0.9645, + "step": 5386 + }, + { + "epoch": 0.4858186409343013, + "grad_norm": 1.6147265298983051, + "learning_rate": 2.1887109960290994e-06, + "loss": 0.9858, + "step": 5387 + }, + { + "epoch": 0.48590882445777156, + "grad_norm": 1.577131884225244, + "learning_rate": 2.18812938377426e-06, + "loss": 1.0882, + "step": 5388 + }, + { + "epoch": 0.48599900798124185, + "grad_norm": 1.7523198419063293, + "learning_rate": 2.187547755467184e-06, + "loss": 1.0264, + "step": 5389 + }, + { + "epoch": 0.4860891915047121, + "grad_norm": 1.8262208983017145, + "learning_rate": 2.1869661111574994e-06, + "loss": 0.8825, + "step": 5390 + }, + { + "epoch": 0.48617937502818237, + "grad_norm": 2.580166372656926, + "learning_rate": 2.1863844508948353e-06, + "loss": 1.0407, + "step": 5391 + }, + { + "epoch": 0.4862695585516526, + "grad_norm": 1.5112391457747665, + "learning_rate": 2.185802774728823e-06, + "loss": 1.0057, + "step": 5392 + }, + { + "epoch": 0.4863597420751229, + "grad_norm": 1.612591168687368, + "learning_rate": 2.1852210827090927e-06, + "loss": 0.9429, + "step": 5393 + }, + { + "epoch": 0.4864499255985931, + "grad_norm": 1.5278853995053445, + "learning_rate": 2.184639374885278e-06, + "loss": 0.9328, + "step": 5394 + }, + { + "epoch": 0.4865401091220634, + "grad_norm": 1.8759245818133552, + "learning_rate": 2.184057651307014e-06, + "loss": 1.0871, + "step": 5395 + }, + { + "epoch": 0.48663029264553365, + "grad_norm": 2.3248393792896422, + "learning_rate": 2.183475912023937e-06, + "loss": 1.0617, + "step": 5396 + }, + { + "epoch": 0.48672047616900393, + "grad_norm": 1.6330889600126832, + "learning_rate": 2.1828941570856826e-06, + "loss": 1.0045, + "step": 5397 + }, + { + "epoch": 0.48681065969247417, + "grad_norm": 0.6444558533004849, + "learning_rate": 2.1823123865418903e-06, + "loss": 0.8085, + "step": 5398 + }, + { + "epoch": 0.48690084321594446, + "grad_norm": 1.901183853754383, + "learning_rate": 2.1817306004422e-06, + "loss": 0.9661, + "step": 5399 + }, + { + "epoch": 0.4869910267394147, + "grad_norm": 1.628945256413402, + "learning_rate": 2.1811487988362527e-06, + "loss": 1.0126, + "step": 5400 + }, + { + "epoch": 0.487081210262885, + "grad_norm": 3.8706155398841635, + "learning_rate": 2.1805669817736917e-06, + "loss": 1.0612, + "step": 5401 + }, + { + "epoch": 0.4871713937863552, + "grad_norm": 1.997436072098825, + "learning_rate": 2.17998514930416e-06, + "loss": 0.8149, + "step": 5402 + }, + { + "epoch": 0.4872615773098255, + "grad_norm": 2.036704969633869, + "learning_rate": 2.1794033014773025e-06, + "loss": 0.9889, + "step": 5403 + }, + { + "epoch": 0.48735176083329573, + "grad_norm": 1.6419664983735223, + "learning_rate": 2.178821438342766e-06, + "loss": 1.0594, + "step": 5404 + }, + { + "epoch": 0.487441944356766, + "grad_norm": 1.6696368771789099, + "learning_rate": 2.1782395599501996e-06, + "loss": 0.9837, + "step": 5405 + }, + { + "epoch": 0.48753212788023625, + "grad_norm": 0.649522838230646, + "learning_rate": 2.1776576663492498e-06, + "loss": 0.8212, + "step": 5406 + }, + { + "epoch": 0.48762231140370654, + "grad_norm": 1.9065319540655932, + "learning_rate": 2.177075757589569e-06, + "loss": 1.0832, + "step": 5407 + }, + { + "epoch": 0.48771249492717683, + "grad_norm": 2.9012480319376808, + "learning_rate": 2.176493833720808e-06, + "loss": 0.9166, + "step": 5408 + }, + { + "epoch": 0.48780267845064706, + "grad_norm": 1.6345590258069747, + "learning_rate": 2.1759118947926195e-06, + "loss": 0.9573, + "step": 5409 + }, + { + "epoch": 0.48789286197411735, + "grad_norm": 1.7635009277228826, + "learning_rate": 2.1753299408546587e-06, + "loss": 1.0023, + "step": 5410 + }, + { + "epoch": 0.4879830454975876, + "grad_norm": 1.3259699222467545, + "learning_rate": 2.1747479719565803e-06, + "loss": 0.9964, + "step": 5411 + }, + { + "epoch": 0.4880732290210579, + "grad_norm": 1.4643548988984803, + "learning_rate": 2.174165988148042e-06, + "loss": 1.0562, + "step": 5412 + }, + { + "epoch": 0.4881634125445281, + "grad_norm": 1.4795472994066938, + "learning_rate": 2.1735839894787003e-06, + "loss": 0.9607, + "step": 5413 + }, + { + "epoch": 0.4882535960679984, + "grad_norm": 0.7659843146633037, + "learning_rate": 2.1730019759982163e-06, + "loss": 0.8658, + "step": 5414 + }, + { + "epoch": 0.48834377959146863, + "grad_norm": 1.7377627167261802, + "learning_rate": 2.172419947756249e-06, + "loss": 1.045, + "step": 5415 + }, + { + "epoch": 0.4884339631149389, + "grad_norm": 1.6963241301202139, + "learning_rate": 2.171837904802461e-06, + "loss": 1.0392, + "step": 5416 + }, + { + "epoch": 0.48852414663840915, + "grad_norm": 1.6750980576238006, + "learning_rate": 2.171255847186516e-06, + "loss": 0.9917, + "step": 5417 + }, + { + "epoch": 0.48861433016187944, + "grad_norm": 1.9078689954236194, + "learning_rate": 2.1706737749580783e-06, + "loss": 0.9801, + "step": 5418 + }, + { + "epoch": 0.4887045136853497, + "grad_norm": 1.71123421939682, + "learning_rate": 2.1700916881668127e-06, + "loss": 1.0655, + "step": 5419 + }, + { + "epoch": 0.48879469720881996, + "grad_norm": 2.1683075511239145, + "learning_rate": 2.1695095868623862e-06, + "loss": 1.0812, + "step": 5420 + }, + { + "epoch": 0.4888848807322902, + "grad_norm": 1.5093636711753777, + "learning_rate": 2.168927471094467e-06, + "loss": 0.925, + "step": 5421 + }, + { + "epoch": 0.4889750642557605, + "grad_norm": 2.292129805033157, + "learning_rate": 2.168345340912725e-06, + "loss": 1.0113, + "step": 5422 + }, + { + "epoch": 0.4890652477792307, + "grad_norm": 1.5768966426566149, + "learning_rate": 2.1677631963668298e-06, + "loss": 1.0142, + "step": 5423 + }, + { + "epoch": 0.489155431302701, + "grad_norm": 1.6679859026489243, + "learning_rate": 2.167181037506453e-06, + "loss": 0.9499, + "step": 5424 + }, + { + "epoch": 0.48924561482617124, + "grad_norm": 1.683334011921258, + "learning_rate": 2.1665988643812693e-06, + "loss": 0.9275, + "step": 5425 + }, + { + "epoch": 0.4893357983496415, + "grad_norm": 2.026885758194087, + "learning_rate": 2.166016677040951e-06, + "loss": 0.9906, + "step": 5426 + }, + { + "epoch": 0.48942598187311176, + "grad_norm": 1.5648106084428908, + "learning_rate": 2.165434475535175e-06, + "loss": 0.982, + "step": 5427 + }, + { + "epoch": 0.48951616539658205, + "grad_norm": 1.4758874066421737, + "learning_rate": 2.1648522599136173e-06, + "loss": 1.0341, + "step": 5428 + }, + { + "epoch": 0.4896063489200523, + "grad_norm": 1.4415876072868523, + "learning_rate": 2.164270030225956e-06, + "loss": 0.9638, + "step": 5429 + }, + { + "epoch": 0.48969653244352257, + "grad_norm": 1.4786244683660985, + "learning_rate": 2.16368778652187e-06, + "loss": 0.9015, + "step": 5430 + }, + { + "epoch": 0.4897867159669928, + "grad_norm": 1.5419611091211853, + "learning_rate": 2.163105528851039e-06, + "loss": 1.0323, + "step": 5431 + }, + { + "epoch": 0.4898768994904631, + "grad_norm": 2.2661757227344563, + "learning_rate": 2.1625232572631448e-06, + "loss": 0.9969, + "step": 5432 + }, + { + "epoch": 0.4899670830139334, + "grad_norm": 0.7351752445957483, + "learning_rate": 2.161940971807871e-06, + "loss": 0.8282, + "step": 5433 + }, + { + "epoch": 0.4900572665374036, + "grad_norm": 1.8280142726132325, + "learning_rate": 2.1613586725348994e-06, + "loss": 0.9794, + "step": 5434 + }, + { + "epoch": 0.4901474500608739, + "grad_norm": 1.8615307532512841, + "learning_rate": 2.1607763594939176e-06, + "loss": 1.0715, + "step": 5435 + }, + { + "epoch": 0.49023763358434413, + "grad_norm": 1.7820041465454153, + "learning_rate": 2.1601940327346093e-06, + "loss": 0.9042, + "step": 5436 + }, + { + "epoch": 0.4903278171078144, + "grad_norm": 1.3851265806752475, + "learning_rate": 2.159611692306663e-06, + "loss": 1.0621, + "step": 5437 + }, + { + "epoch": 0.49041800063128466, + "grad_norm": 1.8098947941253387, + "learning_rate": 2.1590293382597667e-06, + "loss": 0.9531, + "step": 5438 + }, + { + "epoch": 0.49050818415475494, + "grad_norm": 1.7457488741356875, + "learning_rate": 2.1584469706436102e-06, + "loss": 0.9319, + "step": 5439 + }, + { + "epoch": 0.4905983676782252, + "grad_norm": 2.458894493151035, + "learning_rate": 2.1578645895078855e-06, + "loss": 0.908, + "step": 5440 + }, + { + "epoch": 0.49068855120169547, + "grad_norm": 1.8912399805059472, + "learning_rate": 2.157282194902283e-06, + "loss": 1.0506, + "step": 5441 + }, + { + "epoch": 0.4907787347251657, + "grad_norm": 2.4445371525937944, + "learning_rate": 2.1566997868764965e-06, + "loss": 0.9378, + "step": 5442 + }, + { + "epoch": 0.490868918248636, + "grad_norm": 2.3658528106127767, + "learning_rate": 2.15611736548022e-06, + "loss": 0.9523, + "step": 5443 + }, + { + "epoch": 0.4909591017721062, + "grad_norm": 1.5768512835629882, + "learning_rate": 2.155534930763149e-06, + "loss": 1.0008, + "step": 5444 + }, + { + "epoch": 0.4910492852955765, + "grad_norm": 2.314044745748696, + "learning_rate": 2.1549524827749804e-06, + "loss": 0.8774, + "step": 5445 + }, + { + "epoch": 0.49113946881904674, + "grad_norm": 2.64429214726418, + "learning_rate": 2.1543700215654115e-06, + "loss": 0.8407, + "step": 5446 + }, + { + "epoch": 0.49122965234251703, + "grad_norm": 1.644575612246811, + "learning_rate": 2.153787547184141e-06, + "loss": 0.9858, + "step": 5447 + }, + { + "epoch": 0.49131983586598726, + "grad_norm": 1.6363908899089565, + "learning_rate": 2.1532050596808695e-06, + "loss": 0.9189, + "step": 5448 + }, + { + "epoch": 0.49141001938945755, + "grad_norm": 1.6382490119436293, + "learning_rate": 2.152622559105297e-06, + "loss": 1.0386, + "step": 5449 + }, + { + "epoch": 0.4915002029129278, + "grad_norm": 2.1658687834158137, + "learning_rate": 2.152040045507126e-06, + "loss": 1.0326, + "step": 5450 + }, + { + "epoch": 0.4915903864363981, + "grad_norm": 1.854278696797582, + "learning_rate": 2.1514575189360607e-06, + "loss": 1.0013, + "step": 5451 + }, + { + "epoch": 0.4916805699598683, + "grad_norm": 1.4952890806845522, + "learning_rate": 2.1508749794418043e-06, + "loss": 0.9495, + "step": 5452 + }, + { + "epoch": 0.4917707534833386, + "grad_norm": 1.4368109502865725, + "learning_rate": 2.1502924270740626e-06, + "loss": 0.9637, + "step": 5453 + }, + { + "epoch": 0.49186093700680883, + "grad_norm": 2.5457010196280807, + "learning_rate": 2.1497098618825427e-06, + "loss": 0.9964, + "step": 5454 + }, + { + "epoch": 0.4919511205302791, + "grad_norm": 1.6475876339425426, + "learning_rate": 2.1491272839169516e-06, + "loss": 1.0318, + "step": 5455 + }, + { + "epoch": 0.4920413040537494, + "grad_norm": 1.6744419249985565, + "learning_rate": 2.1485446932269986e-06, + "loss": 1.102, + "step": 5456 + }, + { + "epoch": 0.49213148757721964, + "grad_norm": 1.5604388566045821, + "learning_rate": 2.147962089862393e-06, + "loss": 1.0293, + "step": 5457 + }, + { + "epoch": 0.49222167110068993, + "grad_norm": 1.8880963400288522, + "learning_rate": 2.1473794738728462e-06, + "loss": 0.8956, + "step": 5458 + }, + { + "epoch": 0.49231185462416016, + "grad_norm": 1.94358755334424, + "learning_rate": 2.14679684530807e-06, + "loss": 1.0249, + "step": 5459 + }, + { + "epoch": 0.49240203814763045, + "grad_norm": 1.5856882284667277, + "learning_rate": 2.1462142042177774e-06, + "loss": 0.9725, + "step": 5460 + }, + { + "epoch": 0.4924922216711007, + "grad_norm": 1.7867264467648223, + "learning_rate": 2.145631550651683e-06, + "loss": 0.9895, + "step": 5461 + }, + { + "epoch": 0.49258240519457097, + "grad_norm": 1.7776662010041093, + "learning_rate": 2.1450488846595016e-06, + "loss": 1.0892, + "step": 5462 + }, + { + "epoch": 0.4926725887180412, + "grad_norm": 1.6634361986882225, + "learning_rate": 2.14446620629095e-06, + "loss": 0.9681, + "step": 5463 + }, + { + "epoch": 0.4927627722415115, + "grad_norm": 1.4593296280859318, + "learning_rate": 2.1438835155957445e-06, + "loss": 1.0599, + "step": 5464 + }, + { + "epoch": 0.4928529557649817, + "grad_norm": 1.5771203195055428, + "learning_rate": 2.143300812623604e-06, + "loss": 1.0335, + "step": 5465 + }, + { + "epoch": 0.492943139288452, + "grad_norm": 1.7433917754365804, + "learning_rate": 2.1427180974242485e-06, + "loss": 0.9726, + "step": 5466 + }, + { + "epoch": 0.49303332281192225, + "grad_norm": 1.6290148509023634, + "learning_rate": 2.142135370047398e-06, + "loss": 1.021, + "step": 5467 + }, + { + "epoch": 0.49312350633539254, + "grad_norm": 1.9281766590883094, + "learning_rate": 2.1415526305427735e-06, + "loss": 1.0203, + "step": 5468 + }, + { + "epoch": 0.49321368985886277, + "grad_norm": 1.2734562223029382, + "learning_rate": 2.140969878960098e-06, + "loss": 0.915, + "step": 5469 + }, + { + "epoch": 0.49330387338233306, + "grad_norm": 1.6225350831974836, + "learning_rate": 2.1403871153490956e-06, + "loss": 1.0493, + "step": 5470 + }, + { + "epoch": 0.4933940569058033, + "grad_norm": 2.23293444738933, + "learning_rate": 2.13980433975949e-06, + "loss": 1.1122, + "step": 5471 + }, + { + "epoch": 0.4934842404292736, + "grad_norm": 1.5647214737958515, + "learning_rate": 2.1392215522410076e-06, + "loss": 1.0332, + "step": 5472 + }, + { + "epoch": 0.4935744239527438, + "grad_norm": 4.110409682522985, + "learning_rate": 2.1386387528433743e-06, + "loss": 1.1042, + "step": 5473 + }, + { + "epoch": 0.4936646074762141, + "grad_norm": 1.9904498132484196, + "learning_rate": 2.1380559416163186e-06, + "loss": 1.0193, + "step": 5474 + }, + { + "epoch": 0.49375479099968433, + "grad_norm": 1.643889820838748, + "learning_rate": 2.1374731186095685e-06, + "loss": 1.0384, + "step": 5475 + }, + { + "epoch": 0.4938449745231546, + "grad_norm": 1.7054461139988875, + "learning_rate": 2.136890283872854e-06, + "loss": 0.9417, + "step": 5476 + }, + { + "epoch": 0.49393515804662486, + "grad_norm": 1.823359481070777, + "learning_rate": 2.136307437455906e-06, + "loss": 0.9721, + "step": 5477 + }, + { + "epoch": 0.49402534157009514, + "grad_norm": 1.2886537308353672, + "learning_rate": 2.135724579408456e-06, + "loss": 0.9371, + "step": 5478 + }, + { + "epoch": 0.49411552509356543, + "grad_norm": 3.007833267734716, + "learning_rate": 2.1351417097802356e-06, + "loss": 1.0918, + "step": 5479 + }, + { + "epoch": 0.49420570861703567, + "grad_norm": 2.397590413064408, + "learning_rate": 2.1345588286209798e-06, + "loss": 0.8763, + "step": 5480 + }, + { + "epoch": 0.49429589214050595, + "grad_norm": 2.306220524297157, + "learning_rate": 2.1339759359804227e-06, + "loss": 0.8352, + "step": 5481 + }, + { + "epoch": 0.4943860756639762, + "grad_norm": 1.674540026692367, + "learning_rate": 2.1333930319082997e-06, + "loss": 0.9686, + "step": 5482 + }, + { + "epoch": 0.4944762591874465, + "grad_norm": 1.3963414259085902, + "learning_rate": 2.132810116454348e-06, + "loss": 0.9969, + "step": 5483 + }, + { + "epoch": 0.4945664427109167, + "grad_norm": 1.7349641503799547, + "learning_rate": 2.132227189668305e-06, + "loss": 0.9874, + "step": 5484 + }, + { + "epoch": 0.494656626234387, + "grad_norm": 2.777614317429027, + "learning_rate": 2.1316442515999096e-06, + "loss": 0.9373, + "step": 5485 + }, + { + "epoch": 0.49474680975785723, + "grad_norm": 1.327547059814466, + "learning_rate": 2.1310613022989e-06, + "loss": 1.0203, + "step": 5486 + }, + { + "epoch": 0.4948369932813275, + "grad_norm": 2.4003298373863267, + "learning_rate": 2.130478341815017e-06, + "loss": 0.8397, + "step": 5487 + }, + { + "epoch": 0.49492717680479775, + "grad_norm": 1.6053453384079377, + "learning_rate": 2.1298953701980033e-06, + "loss": 0.8869, + "step": 5488 + }, + { + "epoch": 0.49501736032826804, + "grad_norm": 2.9438989621840523, + "learning_rate": 2.1293123874976003e-06, + "loss": 0.9982, + "step": 5489 + }, + { + "epoch": 0.4951075438517383, + "grad_norm": 1.4748092010715927, + "learning_rate": 2.1287293937635513e-06, + "loss": 0.9398, + "step": 5490 + }, + { + "epoch": 0.49519772737520856, + "grad_norm": 1.733775954708886, + "learning_rate": 2.1281463890456005e-06, + "loss": 1.0433, + "step": 5491 + }, + { + "epoch": 0.4952879108986788, + "grad_norm": 8.846138028542146, + "learning_rate": 2.127563373393493e-06, + "loss": 0.9847, + "step": 5492 + }, + { + "epoch": 0.4953780944221491, + "grad_norm": 1.6104441674546335, + "learning_rate": 2.1269803468569756e-06, + "loss": 0.9825, + "step": 5493 + }, + { + "epoch": 0.4954682779456193, + "grad_norm": 1.5154423554475196, + "learning_rate": 2.126397309485794e-06, + "loss": 1.0283, + "step": 5494 + }, + { + "epoch": 0.4955584614690896, + "grad_norm": 2.0385615292456856, + "learning_rate": 2.1258142613296983e-06, + "loss": 0.9506, + "step": 5495 + }, + { + "epoch": 0.49564864499255984, + "grad_norm": 3.145471884739323, + "learning_rate": 2.125231202438435e-06, + "loss": 0.9672, + "step": 5496 + }, + { + "epoch": 0.49573882851603013, + "grad_norm": 1.5210629468403774, + "learning_rate": 2.1246481328617553e-06, + "loss": 1.0247, + "step": 5497 + }, + { + "epoch": 0.49582901203950036, + "grad_norm": 1.948972942624436, + "learning_rate": 2.1240650526494096e-06, + "loss": 0.9737, + "step": 5498 + }, + { + "epoch": 0.49591919556297065, + "grad_norm": 2.3031437245491597, + "learning_rate": 2.1234819618511493e-06, + "loss": 0.9593, + "step": 5499 + }, + { + "epoch": 0.4960093790864409, + "grad_norm": 0.7691359690791183, + "learning_rate": 2.122898860516728e-06, + "loss": 0.8559, + "step": 5500 + }, + { + "epoch": 0.49609956260991117, + "grad_norm": 1.4521664154643277, + "learning_rate": 2.1223157486958976e-06, + "loss": 0.9782, + "step": 5501 + }, + { + "epoch": 0.4961897461333814, + "grad_norm": 1.3443729819509203, + "learning_rate": 2.1217326264384127e-06, + "loss": 0.8825, + "step": 5502 + }, + { + "epoch": 0.4962799296568517, + "grad_norm": 2.137951371570389, + "learning_rate": 2.1211494937940296e-06, + "loss": 1.0121, + "step": 5503 + }, + { + "epoch": 0.496370113180322, + "grad_norm": 1.319769256610285, + "learning_rate": 2.1205663508125034e-06, + "loss": 1.0025, + "step": 5504 + }, + { + "epoch": 0.4964602967037922, + "grad_norm": 1.5208409017435511, + "learning_rate": 2.1199831975435914e-06, + "loss": 0.9528, + "step": 5505 + }, + { + "epoch": 0.4965504802272625, + "grad_norm": 1.700936407354853, + "learning_rate": 2.1194000340370517e-06, + "loss": 1.0393, + "step": 5506 + }, + { + "epoch": 0.49664066375073274, + "grad_norm": 1.7226427557799493, + "learning_rate": 2.1188168603426423e-06, + "loss": 0.9408, + "step": 5507 + }, + { + "epoch": 0.496730847274203, + "grad_norm": 2.0430498562936523, + "learning_rate": 2.118233676510123e-06, + "loss": 0.9541, + "step": 5508 + }, + { + "epoch": 0.49682103079767326, + "grad_norm": 1.8232352568745485, + "learning_rate": 2.117650482589255e-06, + "loss": 1.0038, + "step": 5509 + }, + { + "epoch": 0.49691121432114355, + "grad_norm": 2.0319936711445568, + "learning_rate": 2.1170672786297988e-06, + "loss": 1.0297, + "step": 5510 + }, + { + "epoch": 0.4970013978446138, + "grad_norm": 2.980516425835102, + "learning_rate": 2.1164840646815174e-06, + "loss": 0.9574, + "step": 5511 + }, + { + "epoch": 0.49709158136808407, + "grad_norm": 2.3763282474836718, + "learning_rate": 2.1159008407941726e-06, + "loss": 1.0735, + "step": 5512 + }, + { + "epoch": 0.4971817648915543, + "grad_norm": 1.618434996130657, + "learning_rate": 2.1153176070175293e-06, + "loss": 0.9836, + "step": 5513 + }, + { + "epoch": 0.4972719484150246, + "grad_norm": 1.4916738055322973, + "learning_rate": 2.114734363401352e-06, + "loss": 0.8832, + "step": 5514 + }, + { + "epoch": 0.4973621319384948, + "grad_norm": 1.6872682765525306, + "learning_rate": 2.1141511099954056e-06, + "loss": 0.9879, + "step": 5515 + }, + { + "epoch": 0.4974523154619651, + "grad_norm": 1.621496311438803, + "learning_rate": 2.1135678468494576e-06, + "loss": 0.9724, + "step": 5516 + }, + { + "epoch": 0.49754249898543534, + "grad_norm": 1.8929169059326145, + "learning_rate": 2.112984574013275e-06, + "loss": 1.0283, + "step": 5517 + }, + { + "epoch": 0.49763268250890563, + "grad_norm": 1.354322448964329, + "learning_rate": 2.112401291536625e-06, + "loss": 0.9449, + "step": 5518 + }, + { + "epoch": 0.49772286603237587, + "grad_norm": 1.7342688038249112, + "learning_rate": 2.111817999469278e-06, + "loss": 1.0256, + "step": 5519 + }, + { + "epoch": 0.49781304955584615, + "grad_norm": 1.4841407189447235, + "learning_rate": 2.1112346978610016e-06, + "loss": 0.9398, + "step": 5520 + }, + { + "epoch": 0.4979032330793164, + "grad_norm": 3.4469762405094584, + "learning_rate": 2.1106513867615678e-06, + "loss": 0.9462, + "step": 5521 + }, + { + "epoch": 0.4979934166027867, + "grad_norm": 1.4186187179255583, + "learning_rate": 2.110068066220748e-06, + "loss": 0.9937, + "step": 5522 + }, + { + "epoch": 0.4980836001262569, + "grad_norm": 1.6568180045933978, + "learning_rate": 2.109484736288313e-06, + "loss": 1.0577, + "step": 5523 + }, + { + "epoch": 0.4981737836497272, + "grad_norm": 1.6780046416299161, + "learning_rate": 2.108901397014037e-06, + "loss": 1.0046, + "step": 5524 + }, + { + "epoch": 0.49826396717319743, + "grad_norm": 1.3767033776583273, + "learning_rate": 2.1083180484476934e-06, + "loss": 0.9936, + "step": 5525 + }, + { + "epoch": 0.4983541506966677, + "grad_norm": 1.7055503304960973, + "learning_rate": 2.1077346906390567e-06, + "loss": 1.0118, + "step": 5526 + }, + { + "epoch": 0.498444334220138, + "grad_norm": 1.7289715009495041, + "learning_rate": 2.107151323637902e-06, + "loss": 1.0094, + "step": 5527 + }, + { + "epoch": 0.49853451774360824, + "grad_norm": 1.6996397702640786, + "learning_rate": 2.106567947494006e-06, + "loss": 0.9764, + "step": 5528 + }, + { + "epoch": 0.49862470126707853, + "grad_norm": 1.9174287428167722, + "learning_rate": 2.1059845622571447e-06, + "loss": 1.0565, + "step": 5529 + }, + { + "epoch": 0.49871488479054876, + "grad_norm": 2.1305051066783482, + "learning_rate": 2.1054011679770956e-06, + "loss": 1.0312, + "step": 5530 + }, + { + "epoch": 0.49880506831401905, + "grad_norm": 1.743889358038122, + "learning_rate": 2.104817764703638e-06, + "loss": 1.0022, + "step": 5531 + }, + { + "epoch": 0.4988952518374893, + "grad_norm": 2.0368614740850237, + "learning_rate": 2.1042343524865516e-06, + "loss": 0.9934, + "step": 5532 + }, + { + "epoch": 0.4989854353609596, + "grad_norm": 2.0071621446405197, + "learning_rate": 2.103650931375615e-06, + "loss": 1.1007, + "step": 5533 + }, + { + "epoch": 0.4990756188844298, + "grad_norm": 1.7651022753382237, + "learning_rate": 2.1030675014206094e-06, + "loss": 0.9281, + "step": 5534 + }, + { + "epoch": 0.4991658024079001, + "grad_norm": 1.8564169298116358, + "learning_rate": 2.1024840626713166e-06, + "loss": 0.9239, + "step": 5535 + }, + { + "epoch": 0.4992559859313703, + "grad_norm": 2.013832300494602, + "learning_rate": 2.1019006151775177e-06, + "loss": 1.0916, + "step": 5536 + }, + { + "epoch": 0.4993461694548406, + "grad_norm": 2.0638120697132862, + "learning_rate": 2.101317158988997e-06, + "loss": 0.9814, + "step": 5537 + }, + { + "epoch": 0.49943635297831085, + "grad_norm": 4.777770128662926, + "learning_rate": 2.1007336941555374e-06, + "loss": 0.9631, + "step": 5538 + }, + { + "epoch": 0.49952653650178114, + "grad_norm": 1.5314696018242235, + "learning_rate": 2.1001502207269238e-06, + "loss": 0.9567, + "step": 5539 + }, + { + "epoch": 0.49961672002525137, + "grad_norm": 1.8736953646969643, + "learning_rate": 2.0995667387529407e-06, + "loss": 0.9476, + "step": 5540 + }, + { + "epoch": 0.49970690354872166, + "grad_norm": 1.5618335065333075, + "learning_rate": 2.098983248283375e-06, + "loss": 0.9404, + "step": 5541 + }, + { + "epoch": 0.4997970870721919, + "grad_norm": 3.1010608027111974, + "learning_rate": 2.098399749368012e-06, + "loss": 0.9938, + "step": 5542 + }, + { + "epoch": 0.4998872705956622, + "grad_norm": 0.649902160323509, + "learning_rate": 2.09781624205664e-06, + "loss": 0.8069, + "step": 5543 + }, + { + "epoch": 0.4999774541191324, + "grad_norm": 1.9207358356923254, + "learning_rate": 2.0972327263990477e-06, + "loss": 1.0403, + "step": 5544 + }, + { + "epoch": 0.5000676376426026, + "grad_norm": 1.9467767314082307, + "learning_rate": 2.0966492024450226e-06, + "loss": 0.9223, + "step": 5545 + }, + { + "epoch": 0.500157821166073, + "grad_norm": 1.7212302605321952, + "learning_rate": 2.0960656702443545e-06, + "loss": 1.1382, + "step": 5546 + }, + { + "epoch": 0.5002480046895432, + "grad_norm": 1.4655126239827674, + "learning_rate": 2.0954821298468343e-06, + "loss": 0.9891, + "step": 5547 + }, + { + "epoch": 0.5003381882130135, + "grad_norm": 1.454175015983914, + "learning_rate": 2.0948985813022513e-06, + "loss": 1.0506, + "step": 5548 + }, + { + "epoch": 0.5004283717364837, + "grad_norm": 1.5058980855159394, + "learning_rate": 2.094315024660399e-06, + "loss": 1.0246, + "step": 5549 + }, + { + "epoch": 0.500518555259954, + "grad_norm": 1.8255936140570588, + "learning_rate": 2.0937314599710676e-06, + "loss": 1.011, + "step": 5550 + }, + { + "epoch": 0.5006087387834243, + "grad_norm": 3.924260846169854, + "learning_rate": 2.0931478872840526e-06, + "loss": 0.9781, + "step": 5551 + }, + { + "epoch": 0.5006989223068945, + "grad_norm": 1.8018308547186423, + "learning_rate": 2.092564306649145e-06, + "loss": 0.9733, + "step": 5552 + }, + { + "epoch": 0.5007891058303648, + "grad_norm": 2.353472672066549, + "learning_rate": 2.091980718116141e-06, + "loss": 1.0275, + "step": 5553 + }, + { + "epoch": 0.5008792893538351, + "grad_norm": 1.5688556346815101, + "learning_rate": 2.091397121734835e-06, + "loss": 1.029, + "step": 5554 + }, + { + "epoch": 0.5009694728773053, + "grad_norm": 3.332974621227205, + "learning_rate": 2.090813517555022e-06, + "loss": 1.0166, + "step": 5555 + }, + { + "epoch": 0.5010596564007755, + "grad_norm": 1.7519589767443382, + "learning_rate": 2.0902299056265e-06, + "loss": 0.9318, + "step": 5556 + }, + { + "epoch": 0.5011498399242459, + "grad_norm": 2.1954004670570684, + "learning_rate": 2.0896462859990643e-06, + "loss": 0.9748, + "step": 5557 + }, + { + "epoch": 0.5012400234477161, + "grad_norm": 1.3571379220485589, + "learning_rate": 2.089062658722513e-06, + "loss": 1.052, + "step": 5558 + }, + { + "epoch": 0.5013302069711864, + "grad_norm": 1.8266242624614315, + "learning_rate": 2.0884790238466452e-06, + "loss": 1.0391, + "step": 5559 + }, + { + "epoch": 0.5014203904946566, + "grad_norm": 1.8687721455260191, + "learning_rate": 2.087895381421259e-06, + "loss": 1.0215, + "step": 5560 + }, + { + "epoch": 0.5015105740181269, + "grad_norm": 2.1168222710591924, + "learning_rate": 2.087311731496154e-06, + "loss": 0.9753, + "step": 5561 + }, + { + "epoch": 0.5016007575415972, + "grad_norm": 1.425854657517383, + "learning_rate": 2.08672807412113e-06, + "loss": 0.9814, + "step": 5562 + }, + { + "epoch": 0.5016909410650674, + "grad_norm": 1.8761979408844116, + "learning_rate": 2.08614440934599e-06, + "loss": 1.0805, + "step": 5563 + }, + { + "epoch": 0.5017811245885376, + "grad_norm": 1.9995527363390546, + "learning_rate": 2.0855607372205337e-06, + "loss": 0.9057, + "step": 5564 + }, + { + "epoch": 0.501871308112008, + "grad_norm": 2.963828253451031, + "learning_rate": 2.0849770577945623e-06, + "loss": 1.0064, + "step": 5565 + }, + { + "epoch": 0.5019614916354782, + "grad_norm": 1.360186674335406, + "learning_rate": 2.084393371117881e-06, + "loss": 1.0244, + "step": 5566 + }, + { + "epoch": 0.5020516751589484, + "grad_norm": 1.4940083364128998, + "learning_rate": 2.0838096772402902e-06, + "loss": 0.9237, + "step": 5567 + }, + { + "epoch": 0.5021418586824187, + "grad_norm": 1.855134504228218, + "learning_rate": 2.0832259762115973e-06, + "loss": 0.8785, + "step": 5568 + }, + { + "epoch": 0.502232042205889, + "grad_norm": 1.7455915373132886, + "learning_rate": 2.082642268081605e-06, + "loss": 1.0208, + "step": 5569 + }, + { + "epoch": 0.5023222257293593, + "grad_norm": 1.7667270698057949, + "learning_rate": 2.082058552900118e-06, + "loss": 0.9425, + "step": 5570 + }, + { + "epoch": 0.5024124092528295, + "grad_norm": 1.75217690628085, + "learning_rate": 2.081474830716944e-06, + "loss": 0.9638, + "step": 5571 + }, + { + "epoch": 0.5025025927762997, + "grad_norm": 1.904226511119897, + "learning_rate": 2.080891101581887e-06, + "loss": 1.0155, + "step": 5572 + }, + { + "epoch": 0.5025927762997701, + "grad_norm": 2.438961960005177, + "learning_rate": 2.080307365544755e-06, + "loss": 1.0349, + "step": 5573 + }, + { + "epoch": 0.5026829598232403, + "grad_norm": 2.144647091375504, + "learning_rate": 2.0797236226553567e-06, + "loss": 1.0286, + "step": 5574 + }, + { + "epoch": 0.5027731433467105, + "grad_norm": 1.8944955094889069, + "learning_rate": 2.079139872963499e-06, + "loss": 0.9298, + "step": 5575 + }, + { + "epoch": 0.5028633268701809, + "grad_norm": 1.8623459105787554, + "learning_rate": 2.078556116518991e-06, + "loss": 0.8923, + "step": 5576 + }, + { + "epoch": 0.5029535103936511, + "grad_norm": 1.8148921439541252, + "learning_rate": 2.077972353371642e-06, + "loss": 0.9645, + "step": 5577 + }, + { + "epoch": 0.5030436939171213, + "grad_norm": 1.5769600676014501, + "learning_rate": 2.077388583571262e-06, + "loss": 0.9383, + "step": 5578 + }, + { + "epoch": 0.5031338774405916, + "grad_norm": 0.6538081835061026, + "learning_rate": 2.0768048071676608e-06, + "loss": 0.7507, + "step": 5579 + }, + { + "epoch": 0.5032240609640619, + "grad_norm": 1.6639193621041872, + "learning_rate": 2.0762210242106505e-06, + "loss": 0.9517, + "step": 5580 + }, + { + "epoch": 0.5033142444875321, + "grad_norm": 1.7011940184946606, + "learning_rate": 2.0756372347500424e-06, + "loss": 1.0058, + "step": 5581 + }, + { + "epoch": 0.5034044280110024, + "grad_norm": 1.8193791825073768, + "learning_rate": 2.0750534388356473e-06, + "loss": 1.0945, + "step": 5582 + }, + { + "epoch": 0.5034946115344726, + "grad_norm": 2.4211645253248295, + "learning_rate": 2.07446963651728e-06, + "loss": 0.9355, + "step": 5583 + }, + { + "epoch": 0.503584795057943, + "grad_norm": 1.5652166782080366, + "learning_rate": 2.0738858278447516e-06, + "loss": 0.9281, + "step": 5584 + }, + { + "epoch": 0.5036749785814132, + "grad_norm": 1.5174408878835528, + "learning_rate": 2.073302012867878e-06, + "loss": 1.0816, + "step": 5585 + }, + { + "epoch": 0.5037651621048834, + "grad_norm": 1.4643467581382725, + "learning_rate": 2.0727181916364725e-06, + "loss": 1.0243, + "step": 5586 + }, + { + "epoch": 0.5038553456283537, + "grad_norm": 1.5205171421869759, + "learning_rate": 2.0721343642003493e-06, + "loss": 1.0294, + "step": 5587 + }, + { + "epoch": 0.503945529151824, + "grad_norm": 1.867386555934761, + "learning_rate": 2.0715505306093247e-06, + "loss": 1.0194, + "step": 5588 + }, + { + "epoch": 0.5040357126752942, + "grad_norm": 1.9680523923125952, + "learning_rate": 2.070966690913214e-06, + "loss": 1.0276, + "step": 5589 + }, + { + "epoch": 0.5041258961987645, + "grad_norm": 1.8892406092530614, + "learning_rate": 2.0703828451618346e-06, + "loss": 0.8793, + "step": 5590 + }, + { + "epoch": 0.5042160797222347, + "grad_norm": 1.941574120123753, + "learning_rate": 2.069798993405002e-06, + "loss": 0.9868, + "step": 5591 + }, + { + "epoch": 0.504306263245705, + "grad_norm": 4.5473984879063, + "learning_rate": 2.0692151356925345e-06, + "loss": 1.0297, + "step": 5592 + }, + { + "epoch": 0.5043964467691753, + "grad_norm": 2.295722205027749, + "learning_rate": 2.068631272074251e-06, + "loss": 1.014, + "step": 5593 + }, + { + "epoch": 0.5044866302926455, + "grad_norm": 1.5777820176678015, + "learning_rate": 2.0680474025999676e-06, + "loss": 1.1178, + "step": 5594 + }, + { + "epoch": 0.5045768138161157, + "grad_norm": 0.8667852138785269, + "learning_rate": 2.0674635273195055e-06, + "loss": 0.8684, + "step": 5595 + }, + { + "epoch": 0.5046669973395861, + "grad_norm": 1.8474624241787547, + "learning_rate": 2.066879646282682e-06, + "loss": 0.9895, + "step": 5596 + }, + { + "epoch": 0.5047571808630563, + "grad_norm": 1.7753397280675414, + "learning_rate": 2.0662957595393194e-06, + "loss": 0.9658, + "step": 5597 + }, + { + "epoch": 0.5048473643865266, + "grad_norm": 1.766317383708501, + "learning_rate": 2.0657118671392373e-06, + "loss": 1.0132, + "step": 5598 + }, + { + "epoch": 0.5049375479099969, + "grad_norm": 3.640327277718833, + "learning_rate": 2.0651279691322558e-06, + "loss": 1.0464, + "step": 5599 + }, + { + "epoch": 0.5050277314334671, + "grad_norm": 1.4666602506641735, + "learning_rate": 2.0645440655681973e-06, + "loss": 0.9479, + "step": 5600 + }, + { + "epoch": 0.5051179149569374, + "grad_norm": 1.8908002157436739, + "learning_rate": 2.0639601564968826e-06, + "loss": 1.0545, + "step": 5601 + }, + { + "epoch": 0.5052080984804076, + "grad_norm": 0.6854035317303531, + "learning_rate": 2.0633762419681355e-06, + "loss": 0.9026, + "step": 5602 + }, + { + "epoch": 0.5052982820038779, + "grad_norm": 1.3644248297657287, + "learning_rate": 2.062792322031777e-06, + "loss": 1.0124, + "step": 5603 + }, + { + "epoch": 0.5053884655273482, + "grad_norm": 1.7137224359055494, + "learning_rate": 2.062208396737632e-06, + "loss": 0.9494, + "step": 5604 + }, + { + "epoch": 0.5054786490508184, + "grad_norm": 1.4525582633750516, + "learning_rate": 2.0616244661355235e-06, + "loss": 1.013, + "step": 5605 + }, + { + "epoch": 0.5055688325742886, + "grad_norm": 2.133406895124222, + "learning_rate": 2.0610405302752752e-06, + "loss": 1.0247, + "step": 5606 + }, + { + "epoch": 0.505659016097759, + "grad_norm": 1.6285524418567814, + "learning_rate": 2.060456589206713e-06, + "loss": 0.9835, + "step": 5607 + }, + { + "epoch": 0.5057491996212292, + "grad_norm": 5.117779180666691, + "learning_rate": 2.0598726429796614e-06, + "loss": 0.9413, + "step": 5608 + }, + { + "epoch": 0.5058393831446995, + "grad_norm": 1.624268660666115, + "learning_rate": 2.059288691643945e-06, + "loss": 0.9475, + "step": 5609 + }, + { + "epoch": 0.5059295666681697, + "grad_norm": 1.9811374957275274, + "learning_rate": 2.0587047352493913e-06, + "loss": 1.0722, + "step": 5610 + }, + { + "epoch": 0.50601975019164, + "grad_norm": 5.36759811674148, + "learning_rate": 2.0581207738458248e-06, + "loss": 0.9662, + "step": 5611 + }, + { + "epoch": 0.5061099337151103, + "grad_norm": 1.862539787635559, + "learning_rate": 2.0575368074830743e-06, + "loss": 0.9476, + "step": 5612 + }, + { + "epoch": 0.5062001172385805, + "grad_norm": 3.1936215727603043, + "learning_rate": 2.0569528362109667e-06, + "loss": 0.906, + "step": 5613 + }, + { + "epoch": 0.5062903007620507, + "grad_norm": 1.5334878256270688, + "learning_rate": 2.056368860079327e-06, + "loss": 0.9896, + "step": 5614 + }, + { + "epoch": 0.5063804842855211, + "grad_norm": 1.9928228823621803, + "learning_rate": 2.0557848791379874e-06, + "loss": 0.9352, + "step": 5615 + }, + { + "epoch": 0.5064706678089913, + "grad_norm": 1.9098499887285418, + "learning_rate": 2.0552008934367734e-06, + "loss": 0.8959, + "step": 5616 + }, + { + "epoch": 0.5065608513324615, + "grad_norm": 1.8347057999556078, + "learning_rate": 2.0546169030255154e-06, + "loss": 1.0631, + "step": 5617 + }, + { + "epoch": 0.5066510348559318, + "grad_norm": 1.8366731509020948, + "learning_rate": 2.054032907954041e-06, + "loss": 1.0662, + "step": 5618 + }, + { + "epoch": 0.5067412183794021, + "grad_norm": 1.772186173929745, + "learning_rate": 2.053448908272182e-06, + "loss": 0.955, + "step": 5619 + }, + { + "epoch": 0.5068314019028723, + "grad_norm": 1.565940045545188, + "learning_rate": 2.0528649040297673e-06, + "loss": 1.0269, + "step": 5620 + }, + { + "epoch": 0.5069215854263426, + "grad_norm": 1.5442408472487454, + "learning_rate": 2.0522808952766266e-06, + "loss": 0.997, + "step": 5621 + }, + { + "epoch": 0.5070117689498129, + "grad_norm": 1.475084160165627, + "learning_rate": 2.0516968820625925e-06, + "loss": 1.0353, + "step": 5622 + }, + { + "epoch": 0.5071019524732832, + "grad_norm": 1.6988795150102924, + "learning_rate": 2.051112864437495e-06, + "loss": 0.9796, + "step": 5623 + }, + { + "epoch": 0.5071921359967534, + "grad_norm": 1.9207095202125584, + "learning_rate": 2.050528842451166e-06, + "loss": 0.9604, + "step": 5624 + }, + { + "epoch": 0.5072823195202236, + "grad_norm": 1.3228410513987026, + "learning_rate": 2.049944816153438e-06, + "loss": 0.874, + "step": 5625 + }, + { + "epoch": 0.507372503043694, + "grad_norm": 1.5055704000350398, + "learning_rate": 2.049360785594142e-06, + "loss": 1.0336, + "step": 5626 + }, + { + "epoch": 0.5074626865671642, + "grad_norm": 0.6183912152149381, + "learning_rate": 2.048776750823113e-06, + "loss": 0.7472, + "step": 5627 + }, + { + "epoch": 0.5075528700906344, + "grad_norm": 0.65119380444888, + "learning_rate": 2.0481927118901817e-06, + "loss": 0.8004, + "step": 5628 + }, + { + "epoch": 0.5076430536141047, + "grad_norm": 1.9376651939602176, + "learning_rate": 2.0476086688451824e-06, + "loss": 0.9306, + "step": 5629 + }, + { + "epoch": 0.507733237137575, + "grad_norm": 3.2163328705840466, + "learning_rate": 2.04702462173795e-06, + "loss": 0.9827, + "step": 5630 + }, + { + "epoch": 0.5078234206610452, + "grad_norm": 2.450178381206447, + "learning_rate": 2.0464405706183167e-06, + "loss": 0.9725, + "step": 5631 + }, + { + "epoch": 0.5079136041845155, + "grad_norm": 1.8405835391493108, + "learning_rate": 2.045856515536118e-06, + "loss": 1.009, + "step": 5632 + }, + { + "epoch": 0.5080037877079857, + "grad_norm": 8.078887136134433, + "learning_rate": 2.045272456541188e-06, + "loss": 1.0066, + "step": 5633 + }, + { + "epoch": 0.508093971231456, + "grad_norm": 1.498662908980952, + "learning_rate": 2.0446883936833635e-06, + "loss": 1.0609, + "step": 5634 + }, + { + "epoch": 0.5081841547549263, + "grad_norm": 1.8766700936122973, + "learning_rate": 2.0441043270124782e-06, + "loss": 0.9643, + "step": 5635 + }, + { + "epoch": 0.5082743382783965, + "grad_norm": 1.645305967544383, + "learning_rate": 2.0435202565783683e-06, + "loss": 1.0154, + "step": 5636 + }, + { + "epoch": 0.5083645218018668, + "grad_norm": 2.1651289202428505, + "learning_rate": 2.042936182430871e-06, + "loss": 0.9523, + "step": 5637 + }, + { + "epoch": 0.5084547053253371, + "grad_norm": 1.6021911875810888, + "learning_rate": 2.0423521046198206e-06, + "loss": 0.9977, + "step": 5638 + }, + { + "epoch": 0.5085448888488073, + "grad_norm": 1.678548595307953, + "learning_rate": 2.041768023195056e-06, + "loss": 0.9063, + "step": 5639 + }, + { + "epoch": 0.5086350723722776, + "grad_norm": 1.6491552300014063, + "learning_rate": 2.0411839382064126e-06, + "loss": 1.0573, + "step": 5640 + }, + { + "epoch": 0.5087252558957478, + "grad_norm": 1.665985572248727, + "learning_rate": 2.040599849703729e-06, + "loss": 0.8874, + "step": 5641 + }, + { + "epoch": 0.5088154394192181, + "grad_norm": 3.0835303853620726, + "learning_rate": 2.040015757736843e-06, + "loss": 0.9282, + "step": 5642 + }, + { + "epoch": 0.5089056229426884, + "grad_norm": 1.7502946605566714, + "learning_rate": 2.039431662355591e-06, + "loss": 0.9131, + "step": 5643 + }, + { + "epoch": 0.5089958064661586, + "grad_norm": 6.936368557880409, + "learning_rate": 2.0388475636098126e-06, + "loss": 1.0074, + "step": 5644 + }, + { + "epoch": 0.5090859899896288, + "grad_norm": 1.4500744701861672, + "learning_rate": 2.038263461549346e-06, + "loss": 1.0225, + "step": 5645 + }, + { + "epoch": 0.5091761735130992, + "grad_norm": 1.8636944390655834, + "learning_rate": 2.0376793562240297e-06, + "loss": 0.9938, + "step": 5646 + }, + { + "epoch": 0.5092663570365694, + "grad_norm": 2.008720578522395, + "learning_rate": 2.037095247683703e-06, + "loss": 1.0041, + "step": 5647 + }, + { + "epoch": 0.5093565405600397, + "grad_norm": 1.7410138885715343, + "learning_rate": 2.0365111359782046e-06, + "loss": 0.9509, + "step": 5648 + }, + { + "epoch": 0.50944672408351, + "grad_norm": 1.784492152121173, + "learning_rate": 2.0359270211573757e-06, + "loss": 1.0202, + "step": 5649 + }, + { + "epoch": 0.5095369076069802, + "grad_norm": 2.709388928229094, + "learning_rate": 2.0353429032710545e-06, + "loss": 1.0555, + "step": 5650 + }, + { + "epoch": 0.5096270911304505, + "grad_norm": 1.5272159766876698, + "learning_rate": 2.0347587823690825e-06, + "loss": 0.9688, + "step": 5651 + }, + { + "epoch": 0.5097172746539207, + "grad_norm": 1.583508816414161, + "learning_rate": 2.034174658501299e-06, + "loss": 0.9893, + "step": 5652 + }, + { + "epoch": 0.509807458177391, + "grad_norm": 1.5318091014071193, + "learning_rate": 2.0335905317175453e-06, + "loss": 1.0277, + "step": 5653 + }, + { + "epoch": 0.5098976417008613, + "grad_norm": 1.8514678544589886, + "learning_rate": 2.033006402067663e-06, + "loss": 0.9936, + "step": 5654 + }, + { + "epoch": 0.5099878252243315, + "grad_norm": 1.5227431059677994, + "learning_rate": 2.0324222696014912e-06, + "loss": 0.8959, + "step": 5655 + }, + { + "epoch": 0.5100780087478017, + "grad_norm": 1.6966185570936492, + "learning_rate": 2.0318381343688733e-06, + "loss": 1.027, + "step": 5656 + }, + { + "epoch": 0.5101681922712721, + "grad_norm": 1.571632248200572, + "learning_rate": 2.0312539964196505e-06, + "loss": 0.9724, + "step": 5657 + }, + { + "epoch": 0.5102583757947423, + "grad_norm": 1.5000261463429645, + "learning_rate": 2.030669855803664e-06, + "loss": 0.9705, + "step": 5658 + }, + { + "epoch": 0.5103485593182125, + "grad_norm": 1.455304879563203, + "learning_rate": 2.0300857125707563e-06, + "loss": 1.0408, + "step": 5659 + }, + { + "epoch": 0.5104387428416828, + "grad_norm": 1.9426431380745046, + "learning_rate": 2.0295015667707697e-06, + "loss": 1.0205, + "step": 5660 + }, + { + "epoch": 0.5105289263651531, + "grad_norm": 1.8686461556967373, + "learning_rate": 2.0289174184535472e-06, + "loss": 0.9978, + "step": 5661 + }, + { + "epoch": 0.5106191098886234, + "grad_norm": 1.640280623850272, + "learning_rate": 2.02833326766893e-06, + "loss": 0.9624, + "step": 5662 + }, + { + "epoch": 0.5107092934120936, + "grad_norm": 2.151304018094952, + "learning_rate": 2.027749114466763e-06, + "loss": 1.0569, + "step": 5663 + }, + { + "epoch": 0.5107994769355638, + "grad_norm": 1.6203210666764782, + "learning_rate": 2.027164958896889e-06, + "loss": 1.0381, + "step": 5664 + }, + { + "epoch": 0.5108896604590342, + "grad_norm": 1.841722246197309, + "learning_rate": 2.02658080100915e-06, + "loss": 1.1071, + "step": 5665 + }, + { + "epoch": 0.5109798439825044, + "grad_norm": 1.2960225256445774, + "learning_rate": 2.0259966408533915e-06, + "loss": 1.0679, + "step": 5666 + }, + { + "epoch": 0.5110700275059746, + "grad_norm": 1.6551156387957473, + "learning_rate": 2.025412478479455e-06, + "loss": 1.0792, + "step": 5667 + }, + { + "epoch": 0.5111602110294449, + "grad_norm": 3.1219326987676665, + "learning_rate": 2.0248283139371862e-06, + "loss": 1.036, + "step": 5668 + }, + { + "epoch": 0.5112503945529152, + "grad_norm": 2.324195669364128, + "learning_rate": 2.024244147276429e-06, + "loss": 0.9665, + "step": 5669 + }, + { + "epoch": 0.5113405780763854, + "grad_norm": 1.380187997728163, + "learning_rate": 2.023659978547027e-06, + "loss": 1.0587, + "step": 5670 + }, + { + "epoch": 0.5114307615998557, + "grad_norm": 1.4342667249981398, + "learning_rate": 2.023075807798826e-06, + "loss": 0.9891, + "step": 5671 + }, + { + "epoch": 0.511520945123326, + "grad_norm": 1.5069704224456353, + "learning_rate": 2.0224916350816696e-06, + "loss": 0.9152, + "step": 5672 + }, + { + "epoch": 0.5116111286467963, + "grad_norm": 1.3562357378794274, + "learning_rate": 2.0219074604454026e-06, + "loss": 1.0126, + "step": 5673 + }, + { + "epoch": 0.5117013121702665, + "grad_norm": 1.8000297384984394, + "learning_rate": 2.02132328393987e-06, + "loss": 0.9953, + "step": 5674 + }, + { + "epoch": 0.5117914956937367, + "grad_norm": 2.3358504365242014, + "learning_rate": 2.0207391056149174e-06, + "loss": 0.9451, + "step": 5675 + }, + { + "epoch": 0.5118816792172071, + "grad_norm": 1.6881454081256835, + "learning_rate": 2.020154925520391e-06, + "loss": 0.9603, + "step": 5676 + }, + { + "epoch": 0.5119718627406773, + "grad_norm": 2.00867037128082, + "learning_rate": 2.0195707437061332e-06, + "loss": 1.0896, + "step": 5677 + }, + { + "epoch": 0.5120620462641475, + "grad_norm": 1.71116789916926, + "learning_rate": 2.0189865602219934e-06, + "loss": 0.9882, + "step": 5678 + }, + { + "epoch": 0.5121522297876178, + "grad_norm": 4.4622249315472295, + "learning_rate": 2.0184023751178154e-06, + "loss": 1.0657, + "step": 5679 + }, + { + "epoch": 0.5122424133110881, + "grad_norm": 2.268453792014805, + "learning_rate": 2.017818188443444e-06, + "loss": 0.945, + "step": 5680 + }, + { + "epoch": 0.5123325968345583, + "grad_norm": 1.7987647029866956, + "learning_rate": 2.017234000248728e-06, + "loss": 1.0072, + "step": 5681 + }, + { + "epoch": 0.5124227803580286, + "grad_norm": 1.5852307027193602, + "learning_rate": 2.0166498105835108e-06, + "loss": 0.9313, + "step": 5682 + }, + { + "epoch": 0.5125129638814988, + "grad_norm": 1.277769198711362, + "learning_rate": 2.0160656194976407e-06, + "loss": 0.9272, + "step": 5683 + }, + { + "epoch": 0.5126031474049692, + "grad_norm": 1.625444278138632, + "learning_rate": 2.0154814270409634e-06, + "loss": 0.9446, + "step": 5684 + }, + { + "epoch": 0.5126933309284394, + "grad_norm": 1.8516737506134653, + "learning_rate": 2.0148972332633247e-06, + "loss": 0.9566, + "step": 5685 + }, + { + "epoch": 0.5127835144519096, + "grad_norm": 1.6157548243530155, + "learning_rate": 2.0143130382145733e-06, + "loss": 1.0509, + "step": 5686 + }, + { + "epoch": 0.5128736979753798, + "grad_norm": 2.488457450437017, + "learning_rate": 2.0137288419445533e-06, + "loss": 0.9654, + "step": 5687 + }, + { + "epoch": 0.5129638814988502, + "grad_norm": 1.9135923469932647, + "learning_rate": 2.0131446445031134e-06, + "loss": 1.0398, + "step": 5688 + }, + { + "epoch": 0.5130540650223204, + "grad_norm": 2.43959043802019, + "learning_rate": 2.0125604459400994e-06, + "loss": 1.0797, + "step": 5689 + }, + { + "epoch": 0.5131442485457907, + "grad_norm": 1.8614716449497108, + "learning_rate": 2.0119762463053596e-06, + "loss": 1.0507, + "step": 5690 + }, + { + "epoch": 0.5132344320692609, + "grad_norm": 3.7087764832387125, + "learning_rate": 2.0113920456487406e-06, + "loss": 0.9814, + "step": 5691 + }, + { + "epoch": 0.5133246155927312, + "grad_norm": 1.6425506190334094, + "learning_rate": 2.010807844020088e-06, + "loss": 0.9896, + "step": 5692 + }, + { + "epoch": 0.5134147991162015, + "grad_norm": 1.4601590943857312, + "learning_rate": 2.0102236414692524e-06, + "loss": 1.0476, + "step": 5693 + }, + { + "epoch": 0.5135049826396717, + "grad_norm": 1.391924004927521, + "learning_rate": 2.0096394380460777e-06, + "loss": 0.9798, + "step": 5694 + }, + { + "epoch": 0.513595166163142, + "grad_norm": 1.5093936042349898, + "learning_rate": 2.0090552338004136e-06, + "loss": 1.0279, + "step": 5695 + }, + { + "epoch": 0.5136853496866123, + "grad_norm": 1.49874428004302, + "learning_rate": 2.0084710287821077e-06, + "loss": 0.9921, + "step": 5696 + }, + { + "epoch": 0.5137755332100825, + "grad_norm": 1.6404076568508534, + "learning_rate": 2.007886823041006e-06, + "loss": 1.0325, + "step": 5697 + }, + { + "epoch": 0.5138657167335527, + "grad_norm": 1.5402599652820412, + "learning_rate": 2.0073026166269577e-06, + "loss": 1.0102, + "step": 5698 + }, + { + "epoch": 0.5139559002570231, + "grad_norm": 1.745129005123144, + "learning_rate": 2.0067184095898093e-06, + "loss": 1.0551, + "step": 5699 + }, + { + "epoch": 0.5140460837804933, + "grad_norm": 1.7078661938992874, + "learning_rate": 2.0061342019794094e-06, + "loss": 1.0068, + "step": 5700 + }, + { + "epoch": 0.5141362673039636, + "grad_norm": 1.4302901368479666, + "learning_rate": 2.0055499938456058e-06, + "loss": 1.0715, + "step": 5701 + }, + { + "epoch": 0.5142264508274338, + "grad_norm": 2.036648311355382, + "learning_rate": 2.0049657852382464e-06, + "loss": 0.9606, + "step": 5702 + }, + { + "epoch": 0.5143166343509041, + "grad_norm": 2.142102921678365, + "learning_rate": 2.0043815762071782e-06, + "loss": 0.9223, + "step": 5703 + }, + { + "epoch": 0.5144068178743744, + "grad_norm": 2.3275082174416273, + "learning_rate": 2.0037973668022492e-06, + "loss": 0.9629, + "step": 5704 + }, + { + "epoch": 0.5144970013978446, + "grad_norm": 1.7647994820698647, + "learning_rate": 2.003213157073309e-06, + "loss": 0.8785, + "step": 5705 + }, + { + "epoch": 0.5145871849213148, + "grad_norm": 2.5558044604932317, + "learning_rate": 2.002628947070204e-06, + "loss": 0.9925, + "step": 5706 + }, + { + "epoch": 0.5146773684447852, + "grad_norm": 1.8486480954078057, + "learning_rate": 2.002044736842783e-06, + "loss": 1.0597, + "step": 5707 + }, + { + "epoch": 0.5147675519682554, + "grad_norm": 1.3795493898139919, + "learning_rate": 2.001460526440894e-06, + "loss": 0.9772, + "step": 5708 + }, + { + "epoch": 0.5148577354917256, + "grad_norm": 0.6216859213767867, + "learning_rate": 2.0008763159143843e-06, + "loss": 0.8186, + "step": 5709 + }, + { + "epoch": 0.5149479190151959, + "grad_norm": 1.5251391957027793, + "learning_rate": 2.000292105313103e-06, + "loss": 1.0238, + "step": 5710 + }, + { + "epoch": 0.5150381025386662, + "grad_norm": 1.559195189725679, + "learning_rate": 1.999707894686897e-06, + "loss": 1.0319, + "step": 5711 + }, + { + "epoch": 0.5151282860621365, + "grad_norm": 1.3010348052838312, + "learning_rate": 1.9991236840856155e-06, + "loss": 0.9819, + "step": 5712 + }, + { + "epoch": 0.5152184695856067, + "grad_norm": 1.5846417776171846, + "learning_rate": 1.9985394735591065e-06, + "loss": 0.9686, + "step": 5713 + }, + { + "epoch": 0.5153086531090769, + "grad_norm": 1.634771094090874, + "learning_rate": 1.997955263157217e-06, + "loss": 1.04, + "step": 5714 + }, + { + "epoch": 0.5153988366325473, + "grad_norm": 1.878799467165252, + "learning_rate": 1.997371052929796e-06, + "loss": 1.0277, + "step": 5715 + }, + { + "epoch": 0.5154890201560175, + "grad_norm": 1.5166475445027483, + "learning_rate": 1.996786842926691e-06, + "loss": 0.9522, + "step": 5716 + }, + { + "epoch": 0.5155792036794877, + "grad_norm": 1.722391259535615, + "learning_rate": 1.9962026331977506e-06, + "loss": 1.0034, + "step": 5717 + }, + { + "epoch": 0.5156693872029581, + "grad_norm": 2.42044064285382, + "learning_rate": 1.9956184237928224e-06, + "loss": 1.0068, + "step": 5718 + }, + { + "epoch": 0.5157595707264283, + "grad_norm": 1.662482357648585, + "learning_rate": 1.995034214761754e-06, + "loss": 0.9817, + "step": 5719 + }, + { + "epoch": 0.5158497542498985, + "grad_norm": 1.2863001964327836, + "learning_rate": 1.9944500061543945e-06, + "loss": 0.9671, + "step": 5720 + }, + { + "epoch": 0.5159399377733688, + "grad_norm": 1.854043963654864, + "learning_rate": 1.99386579802059e-06, + "loss": 1.0136, + "step": 5721 + }, + { + "epoch": 0.5160301212968391, + "grad_norm": 1.788923366417099, + "learning_rate": 1.993281590410191e-06, + "loss": 0.9703, + "step": 5722 + }, + { + "epoch": 0.5161203048203094, + "grad_norm": 1.3845283939980886, + "learning_rate": 1.992697383373043e-06, + "loss": 0.9422, + "step": 5723 + }, + { + "epoch": 0.5162104883437796, + "grad_norm": 0.7742840536312902, + "learning_rate": 1.9921131769589937e-06, + "loss": 0.8581, + "step": 5724 + }, + { + "epoch": 0.5163006718672498, + "grad_norm": 1.7052598225023072, + "learning_rate": 1.991528971217893e-06, + "loss": 0.9936, + "step": 5725 + }, + { + "epoch": 0.5163908553907202, + "grad_norm": 3.4000443792251716, + "learning_rate": 1.9909447661995858e-06, + "loss": 0.9883, + "step": 5726 + }, + { + "epoch": 0.5164810389141904, + "grad_norm": 1.8144386709764713, + "learning_rate": 1.990360561953922e-06, + "loss": 1.0061, + "step": 5727 + }, + { + "epoch": 0.5165712224376606, + "grad_norm": 1.4261103002148559, + "learning_rate": 1.9897763585307483e-06, + "loss": 1.0166, + "step": 5728 + }, + { + "epoch": 0.5166614059611309, + "grad_norm": 2.197376299296051, + "learning_rate": 1.989192155979912e-06, + "loss": 1.0351, + "step": 5729 + }, + { + "epoch": 0.5167515894846012, + "grad_norm": 1.2673302923952314, + "learning_rate": 1.98860795435126e-06, + "loss": 1.0275, + "step": 5730 + }, + { + "epoch": 0.5168417730080714, + "grad_norm": 1.4473932083872956, + "learning_rate": 1.9880237536946406e-06, + "loss": 1.0031, + "step": 5731 + }, + { + "epoch": 0.5169319565315417, + "grad_norm": 2.0558185437797594, + "learning_rate": 1.987439554059901e-06, + "loss": 0.9608, + "step": 5732 + }, + { + "epoch": 0.5170221400550119, + "grad_norm": 1.5747369334612475, + "learning_rate": 1.9868553554968864e-06, + "loss": 0.9669, + "step": 5733 + }, + { + "epoch": 0.5171123235784822, + "grad_norm": 1.600677692420372, + "learning_rate": 1.986271158055447e-06, + "loss": 0.943, + "step": 5734 + }, + { + "epoch": 0.5172025071019525, + "grad_norm": 2.711960305522643, + "learning_rate": 1.9856869617854273e-06, + "loss": 0.9904, + "step": 5735 + }, + { + "epoch": 0.5172926906254227, + "grad_norm": 1.6204639360369042, + "learning_rate": 1.9851027667366746e-06, + "loss": 0.9749, + "step": 5736 + }, + { + "epoch": 0.517382874148893, + "grad_norm": 1.8155151806708578, + "learning_rate": 1.984518572959037e-06, + "loss": 1.0887, + "step": 5737 + }, + { + "epoch": 0.5174730576723633, + "grad_norm": 1.9056471825169052, + "learning_rate": 1.9839343805023587e-06, + "loss": 0.9123, + "step": 5738 + }, + { + "epoch": 0.5175632411958335, + "grad_norm": 1.913376790438177, + "learning_rate": 1.9833501894164886e-06, + "loss": 1.0809, + "step": 5739 + }, + { + "epoch": 0.5176534247193038, + "grad_norm": 1.6503263381987485, + "learning_rate": 1.982765999751273e-06, + "loss": 1.0321, + "step": 5740 + }, + { + "epoch": 0.5177436082427741, + "grad_norm": 1.7995282694129542, + "learning_rate": 1.9821818115565553e-06, + "loss": 1.0221, + "step": 5741 + }, + { + "epoch": 0.5178337917662443, + "grad_norm": 2.5210960083186835, + "learning_rate": 1.9815976248821853e-06, + "loss": 0.9025, + "step": 5742 + }, + { + "epoch": 0.5179239752897146, + "grad_norm": 1.4263727502715708, + "learning_rate": 1.981013439778007e-06, + "loss": 1.0073, + "step": 5743 + }, + { + "epoch": 0.5180141588131848, + "grad_norm": 2.1929950904147626, + "learning_rate": 1.9804292562938666e-06, + "loss": 1.0002, + "step": 5744 + }, + { + "epoch": 0.5181043423366551, + "grad_norm": 1.522135956184633, + "learning_rate": 1.97984507447961e-06, + "loss": 1.0322, + "step": 5745 + }, + { + "epoch": 0.5181945258601254, + "grad_norm": 1.795197143035798, + "learning_rate": 1.9792608943850824e-06, + "loss": 0.9323, + "step": 5746 + }, + { + "epoch": 0.5182847093835956, + "grad_norm": 1.8795746627514045, + "learning_rate": 1.9786767160601305e-06, + "loss": 0.9724, + "step": 5747 + }, + { + "epoch": 0.5183748929070658, + "grad_norm": 2.0981168614179384, + "learning_rate": 1.9780925395545977e-06, + "loss": 0.9255, + "step": 5748 + }, + { + "epoch": 0.5184650764305362, + "grad_norm": 1.4409006777649853, + "learning_rate": 1.9775083649183306e-06, + "loss": 0.8037, + "step": 5749 + }, + { + "epoch": 0.5185552599540064, + "grad_norm": 2.5829457843450223, + "learning_rate": 1.976924192201174e-06, + "loss": 1.0331, + "step": 5750 + }, + { + "epoch": 0.5186454434774767, + "grad_norm": 1.4543794529104486, + "learning_rate": 1.9763400214529723e-06, + "loss": 0.9255, + "step": 5751 + }, + { + "epoch": 0.5187356270009469, + "grad_norm": 1.5086222633579107, + "learning_rate": 1.9757558527235713e-06, + "loss": 0.8662, + "step": 5752 + }, + { + "epoch": 0.5188258105244172, + "grad_norm": 1.7554701733381801, + "learning_rate": 1.9751716860628136e-06, + "loss": 0.983, + "step": 5753 + }, + { + "epoch": 0.5189159940478875, + "grad_norm": 1.7700984626797394, + "learning_rate": 1.974587521520545e-06, + "loss": 1.0221, + "step": 5754 + }, + { + "epoch": 0.5190061775713577, + "grad_norm": 1.9758062310560855, + "learning_rate": 1.9740033591466088e-06, + "loss": 1.027, + "step": 5755 + }, + { + "epoch": 0.5190963610948279, + "grad_norm": 1.6760457859340363, + "learning_rate": 1.97341919899085e-06, + "loss": 1.1045, + "step": 5756 + }, + { + "epoch": 0.5191865446182983, + "grad_norm": 1.9056485587440721, + "learning_rate": 1.9728350411031114e-06, + "loss": 1.0196, + "step": 5757 + }, + { + "epoch": 0.5192767281417685, + "grad_norm": 1.9047444761421204, + "learning_rate": 1.9722508855332367e-06, + "loss": 1.0723, + "step": 5758 + }, + { + "epoch": 0.5193669116652387, + "grad_norm": 1.32688558753053, + "learning_rate": 1.97166673233107e-06, + "loss": 0.9273, + "step": 5759 + }, + { + "epoch": 0.519457095188709, + "grad_norm": 1.3461940185610082, + "learning_rate": 1.971082581546453e-06, + "loss": 1.0387, + "step": 5760 + }, + { + "epoch": 0.5195472787121793, + "grad_norm": 1.5749384368490413, + "learning_rate": 1.9704984332292306e-06, + "loss": 1.0206, + "step": 5761 + }, + { + "epoch": 0.5196374622356495, + "grad_norm": 1.8728350538464447, + "learning_rate": 1.9699142874292444e-06, + "loss": 1.0843, + "step": 5762 + }, + { + "epoch": 0.5197276457591198, + "grad_norm": 1.5884691148876726, + "learning_rate": 1.969330144196336e-06, + "loss": 1.0854, + "step": 5763 + }, + { + "epoch": 0.51981782928259, + "grad_norm": 1.871473811340138, + "learning_rate": 1.9687460035803497e-06, + "loss": 0.9396, + "step": 5764 + }, + { + "epoch": 0.5199080128060604, + "grad_norm": 3.7370455462400844, + "learning_rate": 1.9681618656311265e-06, + "loss": 0.9771, + "step": 5765 + }, + { + "epoch": 0.5199981963295306, + "grad_norm": 1.3418220173397257, + "learning_rate": 1.9675777303985086e-06, + "loss": 0.9047, + "step": 5766 + }, + { + "epoch": 0.5200883798530008, + "grad_norm": 1.728316921731298, + "learning_rate": 1.9669935979323376e-06, + "loss": 1.0273, + "step": 5767 + }, + { + "epoch": 0.5201785633764712, + "grad_norm": 2.309375301459142, + "learning_rate": 1.9664094682824545e-06, + "loss": 1.0528, + "step": 5768 + }, + { + "epoch": 0.5202687468999414, + "grad_norm": 1.5110001938603275, + "learning_rate": 1.965825341498701e-06, + "loss": 0.9574, + "step": 5769 + }, + { + "epoch": 0.5203589304234116, + "grad_norm": 2.036375299372343, + "learning_rate": 1.9652412176309177e-06, + "loss": 1.0682, + "step": 5770 + }, + { + "epoch": 0.5204491139468819, + "grad_norm": 1.4288754684611662, + "learning_rate": 1.9646570967289453e-06, + "loss": 0.8622, + "step": 5771 + }, + { + "epoch": 0.5205392974703522, + "grad_norm": 1.5813567807778808, + "learning_rate": 1.9640729788426246e-06, + "loss": 1.0327, + "step": 5772 + }, + { + "epoch": 0.5206294809938224, + "grad_norm": 1.5572490962747452, + "learning_rate": 1.963488864021795e-06, + "loss": 1.0209, + "step": 5773 + }, + { + "epoch": 0.5207196645172927, + "grad_norm": 1.5806392621067755, + "learning_rate": 1.962904752316298e-06, + "loss": 0.8831, + "step": 5774 + }, + { + "epoch": 0.5208098480407629, + "grad_norm": 1.9878170768158567, + "learning_rate": 1.9623206437759706e-06, + "loss": 0.9226, + "step": 5775 + }, + { + "epoch": 0.5209000315642333, + "grad_norm": 1.6930938829369737, + "learning_rate": 1.9617365384506545e-06, + "loss": 0.9549, + "step": 5776 + }, + { + "epoch": 0.5209902150877035, + "grad_norm": 1.4604455456166197, + "learning_rate": 1.9611524363901872e-06, + "loss": 0.9818, + "step": 5777 + }, + { + "epoch": 0.5210803986111737, + "grad_norm": 1.588582131125789, + "learning_rate": 1.960568337644409e-06, + "loss": 0.9881, + "step": 5778 + }, + { + "epoch": 0.521170582134644, + "grad_norm": 1.6496519704274464, + "learning_rate": 1.9599842422631576e-06, + "loss": 0.9763, + "step": 5779 + }, + { + "epoch": 0.5212607656581143, + "grad_norm": 1.702424071379113, + "learning_rate": 1.9594001502962703e-06, + "loss": 0.9832, + "step": 5780 + }, + { + "epoch": 0.5213509491815845, + "grad_norm": 1.3520069163416408, + "learning_rate": 1.9588160617935868e-06, + "loss": 1.0533, + "step": 5781 + }, + { + "epoch": 0.5214411327050548, + "grad_norm": 1.7155122952711357, + "learning_rate": 1.958231976804944e-06, + "loss": 1.0814, + "step": 5782 + }, + { + "epoch": 0.521531316228525, + "grad_norm": 3.9951741194229364, + "learning_rate": 1.957647895380179e-06, + "loss": 0.9147, + "step": 5783 + }, + { + "epoch": 0.5216214997519953, + "grad_norm": 2.1237527889707146, + "learning_rate": 1.9570638175691297e-06, + "loss": 0.9236, + "step": 5784 + }, + { + "epoch": 0.5217116832754656, + "grad_norm": 1.9309607872559624, + "learning_rate": 1.956479743421632e-06, + "loss": 1.0346, + "step": 5785 + }, + { + "epoch": 0.5218018667989358, + "grad_norm": 1.8919667298656007, + "learning_rate": 1.955895672987522e-06, + "loss": 0.9329, + "step": 5786 + }, + { + "epoch": 0.521892050322406, + "grad_norm": 1.3952127709369664, + "learning_rate": 1.9553116063166367e-06, + "loss": 0.8997, + "step": 5787 + }, + { + "epoch": 0.5219822338458764, + "grad_norm": 2.3621433008961388, + "learning_rate": 1.954727543458812e-06, + "loss": 0.9, + "step": 5788 + }, + { + "epoch": 0.5220724173693466, + "grad_norm": 2.352362613862061, + "learning_rate": 1.954143484463883e-06, + "loss": 1.024, + "step": 5789 + }, + { + "epoch": 0.5221626008928169, + "grad_norm": 1.7705286848518034, + "learning_rate": 1.9535594293816836e-06, + "loss": 0.9827, + "step": 5790 + }, + { + "epoch": 0.5222527844162872, + "grad_norm": 3.250016872655646, + "learning_rate": 1.952975378262051e-06, + "loss": 0.9367, + "step": 5791 + }, + { + "epoch": 0.5223429679397574, + "grad_norm": 1.7829677932064079, + "learning_rate": 1.952391331154817e-06, + "loss": 0.9474, + "step": 5792 + }, + { + "epoch": 0.5224331514632277, + "grad_norm": 1.6567738981997089, + "learning_rate": 1.9518072881098185e-06, + "loss": 1.0479, + "step": 5793 + }, + { + "epoch": 0.5225233349866979, + "grad_norm": 1.5917015374776042, + "learning_rate": 1.9512232491768867e-06, + "loss": 0.9809, + "step": 5794 + }, + { + "epoch": 0.5226135185101682, + "grad_norm": 2.3359792444077825, + "learning_rate": 1.9506392144058573e-06, + "loss": 0.985, + "step": 5795 + }, + { + "epoch": 0.5227037020336385, + "grad_norm": 2.2877897819034074, + "learning_rate": 1.9500551838465623e-06, + "loss": 0.9848, + "step": 5796 + }, + { + "epoch": 0.5227938855571087, + "grad_norm": 1.4245047930191295, + "learning_rate": 1.9494711575488337e-06, + "loss": 1.0038, + "step": 5797 + }, + { + "epoch": 0.5228840690805789, + "grad_norm": 2.0376972875271986, + "learning_rate": 1.948887135562505e-06, + "loss": 0.931, + "step": 5798 + }, + { + "epoch": 0.5229742526040493, + "grad_norm": 1.4731536753683305, + "learning_rate": 1.9483031179374074e-06, + "loss": 1.0428, + "step": 5799 + }, + { + "epoch": 0.5230644361275195, + "grad_norm": 1.4851839772324598, + "learning_rate": 1.9477191047233736e-06, + "loss": 1.0487, + "step": 5800 + }, + { + "epoch": 0.5231546196509897, + "grad_norm": 2.037147763001197, + "learning_rate": 1.9471350959702334e-06, + "loss": 0.8734, + "step": 5801 + }, + { + "epoch": 0.52324480317446, + "grad_norm": 3.1170721582474923, + "learning_rate": 1.9465510917278184e-06, + "loss": 0.9972, + "step": 5802 + }, + { + "epoch": 0.5233349866979303, + "grad_norm": 1.823936489576439, + "learning_rate": 1.9459670920459593e-06, + "loss": 0.9866, + "step": 5803 + }, + { + "epoch": 0.5234251702214006, + "grad_norm": 1.760574341283177, + "learning_rate": 1.945383096974485e-06, + "loss": 0.9083, + "step": 5804 + }, + { + "epoch": 0.5235153537448708, + "grad_norm": 2.521419036751492, + "learning_rate": 1.944799106563227e-06, + "loss": 0.9131, + "step": 5805 + }, + { + "epoch": 0.523605537268341, + "grad_norm": 3.882114892689172, + "learning_rate": 1.9442151208620133e-06, + "loss": 1.03, + "step": 5806 + }, + { + "epoch": 0.5236957207918114, + "grad_norm": 1.4539599737196458, + "learning_rate": 1.943631139920672e-06, + "loss": 1.0382, + "step": 5807 + }, + { + "epoch": 0.5237859043152816, + "grad_norm": 1.6511106827648208, + "learning_rate": 1.943047163789034e-06, + "loss": 0.9445, + "step": 5808 + }, + { + "epoch": 0.5238760878387518, + "grad_norm": 6.142790822529729, + "learning_rate": 1.942463192516925e-06, + "loss": 0.9474, + "step": 5809 + }, + { + "epoch": 0.5239662713622221, + "grad_norm": 1.6805197849442628, + "learning_rate": 1.9418792261541746e-06, + "loss": 1.0086, + "step": 5810 + }, + { + "epoch": 0.5240564548856924, + "grad_norm": 1.8402443872777405, + "learning_rate": 1.9412952647506094e-06, + "loss": 0.9457, + "step": 5811 + }, + { + "epoch": 0.5241466384091626, + "grad_norm": 1.4627684680983581, + "learning_rate": 1.9407113083560552e-06, + "loss": 0.9644, + "step": 5812 + }, + { + "epoch": 0.5242368219326329, + "grad_norm": 2.254601435826282, + "learning_rate": 1.940127357020339e-06, + "loss": 0.9684, + "step": 5813 + }, + { + "epoch": 0.5243270054561032, + "grad_norm": 1.617102136985061, + "learning_rate": 1.939543410793287e-06, + "loss": 0.9993, + "step": 5814 + }, + { + "epoch": 0.5244171889795735, + "grad_norm": 1.5151087088028887, + "learning_rate": 1.9389594697247246e-06, + "loss": 1.0237, + "step": 5815 + }, + { + "epoch": 0.5245073725030437, + "grad_norm": 1.9376447069826932, + "learning_rate": 1.9383755338644763e-06, + "loss": 1.0635, + "step": 5816 + }, + { + "epoch": 0.5245975560265139, + "grad_norm": 2.459061843342608, + "learning_rate": 1.937791603262368e-06, + "loss": 1.018, + "step": 5817 + }, + { + "epoch": 0.5246877395499843, + "grad_norm": 1.7592944192812041, + "learning_rate": 1.9372076779682235e-06, + "loss": 0.9434, + "step": 5818 + }, + { + "epoch": 0.5247779230734545, + "grad_norm": 1.832684365921247, + "learning_rate": 1.9366237580318648e-06, + "loss": 0.944, + "step": 5819 + }, + { + "epoch": 0.5248681065969247, + "grad_norm": 3.8428295242889168, + "learning_rate": 1.9360398435031176e-06, + "loss": 0.9951, + "step": 5820 + }, + { + "epoch": 0.524958290120395, + "grad_norm": 1.560017712076824, + "learning_rate": 1.9354559344318025e-06, + "loss": 0.9963, + "step": 5821 + }, + { + "epoch": 0.5250484736438653, + "grad_norm": 1.9487049595275088, + "learning_rate": 1.934872030867744e-06, + "loss": 1.0465, + "step": 5822 + }, + { + "epoch": 0.5251386571673355, + "grad_norm": 11.028312659457297, + "learning_rate": 1.934288132860763e-06, + "loss": 1.001, + "step": 5823 + }, + { + "epoch": 0.5252288406908058, + "grad_norm": 1.6923414644626178, + "learning_rate": 1.93370424046068e-06, + "loss": 0.9872, + "step": 5824 + }, + { + "epoch": 0.525319024214276, + "grad_norm": 1.5246026694062118, + "learning_rate": 1.9331203537173177e-06, + "loss": 1.0533, + "step": 5825 + }, + { + "epoch": 0.5254092077377464, + "grad_norm": 1.3610211632433438, + "learning_rate": 1.9325364726804947e-06, + "loss": 1.1052, + "step": 5826 + }, + { + "epoch": 0.5254993912612166, + "grad_norm": 1.5691397163857603, + "learning_rate": 1.9319525974000327e-06, + "loss": 0.9176, + "step": 5827 + }, + { + "epoch": 0.5255895747846868, + "grad_norm": 2.6283900758870873, + "learning_rate": 1.93136872792575e-06, + "loss": 1.0489, + "step": 5828 + }, + { + "epoch": 0.525679758308157, + "grad_norm": 1.625350034134024, + "learning_rate": 1.9307848643074653e-06, + "loss": 0.9404, + "step": 5829 + }, + { + "epoch": 0.5257699418316274, + "grad_norm": 1.5608370514740713, + "learning_rate": 1.9302010065949985e-06, + "loss": 1.0191, + "step": 5830 + }, + { + "epoch": 0.5258601253550976, + "grad_norm": 6.60617500197643, + "learning_rate": 1.9296171548381657e-06, + "loss": 0.9513, + "step": 5831 + }, + { + "epoch": 0.5259503088785679, + "grad_norm": 1.4851807666021208, + "learning_rate": 1.9290333090867862e-06, + "loss": 0.977, + "step": 5832 + }, + { + "epoch": 0.5260404924020381, + "grad_norm": 1.7030970019917713, + "learning_rate": 1.928449469390676e-06, + "loss": 0.9929, + "step": 5833 + }, + { + "epoch": 0.5261306759255084, + "grad_norm": 1.6430894053693077, + "learning_rate": 1.927865635799651e-06, + "loss": 0.9615, + "step": 5834 + }, + { + "epoch": 0.5262208594489787, + "grad_norm": 1.501726904848823, + "learning_rate": 1.927281808363528e-06, + "loss": 0.9172, + "step": 5835 + }, + { + "epoch": 0.5263110429724489, + "grad_norm": 1.4753737735770664, + "learning_rate": 1.9266979871321216e-06, + "loss": 1.0658, + "step": 5836 + }, + { + "epoch": 0.5264012264959192, + "grad_norm": 1.6591462058404427, + "learning_rate": 1.9261141721552482e-06, + "loss": 0.9407, + "step": 5837 + }, + { + "epoch": 0.5264914100193895, + "grad_norm": 1.527067427749686, + "learning_rate": 1.9255303634827204e-06, + "loss": 1.0569, + "step": 5838 + }, + { + "epoch": 0.5265815935428597, + "grad_norm": 1.5232974257017762, + "learning_rate": 1.924946561164352e-06, + "loss": 1.0497, + "step": 5839 + }, + { + "epoch": 0.52667177706633, + "grad_norm": 2.550027592827601, + "learning_rate": 1.9243627652499582e-06, + "loss": 0.938, + "step": 5840 + }, + { + "epoch": 0.5267619605898003, + "grad_norm": 1.8352749542131979, + "learning_rate": 1.9237789757893493e-06, + "loss": 1.0506, + "step": 5841 + }, + { + "epoch": 0.5268521441132705, + "grad_norm": 3.0775822969491213, + "learning_rate": 1.9231951928323395e-06, + "loss": 1.0331, + "step": 5842 + }, + { + "epoch": 0.5269423276367408, + "grad_norm": 3.3919520440040483, + "learning_rate": 1.922611416428738e-06, + "loss": 0.9414, + "step": 5843 + }, + { + "epoch": 0.527032511160211, + "grad_norm": 1.496466368644321, + "learning_rate": 1.922027646628358e-06, + "loss": 1.0123, + "step": 5844 + }, + { + "epoch": 0.5271226946836813, + "grad_norm": 0.6590232154782959, + "learning_rate": 1.9214438834810092e-06, + "loss": 0.8719, + "step": 5845 + }, + { + "epoch": 0.5272128782071516, + "grad_norm": 1.5845927282605385, + "learning_rate": 1.9208601270365008e-06, + "loss": 0.9191, + "step": 5846 + }, + { + "epoch": 0.5273030617306218, + "grad_norm": 4.78483098384314, + "learning_rate": 1.9202763773446435e-06, + "loss": 1.0693, + "step": 5847 + }, + { + "epoch": 0.527393245254092, + "grad_norm": 1.5862590294739527, + "learning_rate": 1.9196926344552444e-06, + "loss": 0.887, + "step": 5848 + }, + { + "epoch": 0.5274834287775624, + "grad_norm": 1.5990383447101553, + "learning_rate": 1.919108898418113e-06, + "loss": 1.0125, + "step": 5849 + }, + { + "epoch": 0.5275736123010326, + "grad_norm": 1.320314892648188, + "learning_rate": 1.918525169283057e-06, + "loss": 0.8849, + "step": 5850 + }, + { + "epoch": 0.5276637958245028, + "grad_norm": 3.103040520784866, + "learning_rate": 1.9179414470998817e-06, + "loss": 1.0185, + "step": 5851 + }, + { + "epoch": 0.5277539793479731, + "grad_norm": 1.5156878428851048, + "learning_rate": 1.917357731918395e-06, + "loss": 0.9482, + "step": 5852 + }, + { + "epoch": 0.5278441628714434, + "grad_norm": 1.4664037103890086, + "learning_rate": 1.9167740237884025e-06, + "loss": 1.1205, + "step": 5853 + }, + { + "epoch": 0.5279343463949137, + "grad_norm": 1.755376798278283, + "learning_rate": 1.916190322759709e-06, + "loss": 1.0305, + "step": 5854 + }, + { + "epoch": 0.5280245299183839, + "grad_norm": 1.666816688779219, + "learning_rate": 1.91560662888212e-06, + "loss": 1.0873, + "step": 5855 + }, + { + "epoch": 0.5281147134418541, + "grad_norm": 1.487438774451678, + "learning_rate": 1.915022942205438e-06, + "loss": 1.0369, + "step": 5856 + }, + { + "epoch": 0.5282048969653245, + "grad_norm": 2.6015789959358773, + "learning_rate": 1.914439262779468e-06, + "loss": 0.9835, + "step": 5857 + }, + { + "epoch": 0.5282950804887947, + "grad_norm": 1.9657890821421713, + "learning_rate": 1.9138555906540103e-06, + "loss": 0.9246, + "step": 5858 + }, + { + "epoch": 0.5283852640122649, + "grad_norm": 1.4973663415630616, + "learning_rate": 1.91327192587887e-06, + "loss": 0.8984, + "step": 5859 + }, + { + "epoch": 0.5284754475357353, + "grad_norm": 8.011385921444816, + "learning_rate": 1.912688268503846e-06, + "loss": 0.9638, + "step": 5860 + }, + { + "epoch": 0.5285656310592055, + "grad_norm": 1.7106924360514224, + "learning_rate": 1.912104618578741e-06, + "loss": 0.9497, + "step": 5861 + }, + { + "epoch": 0.5286558145826757, + "grad_norm": 1.7674335230907334, + "learning_rate": 1.9115209761533554e-06, + "loss": 0.9707, + "step": 5862 + }, + { + "epoch": 0.528745998106146, + "grad_norm": 1.8086434474897768, + "learning_rate": 1.9109373412774863e-06, + "loss": 0.9035, + "step": 5863 + }, + { + "epoch": 0.5288361816296163, + "grad_norm": 1.4485980655761004, + "learning_rate": 1.910353714000936e-06, + "loss": 0.9917, + "step": 5864 + }, + { + "epoch": 0.5289263651530866, + "grad_norm": 1.767719747865772, + "learning_rate": 1.9097700943734997e-06, + "loss": 0.9464, + "step": 5865 + }, + { + "epoch": 0.5290165486765568, + "grad_norm": 1.5571621316814288, + "learning_rate": 1.909186482444977e-06, + "loss": 0.8911, + "step": 5866 + }, + { + "epoch": 0.529106732200027, + "grad_norm": 1.602700029526285, + "learning_rate": 1.9086028782651652e-06, + "loss": 0.8927, + "step": 5867 + }, + { + "epoch": 0.5291969157234974, + "grad_norm": 1.8627954645502611, + "learning_rate": 1.908019281883859e-06, + "loss": 0.9852, + "step": 5868 + }, + { + "epoch": 0.5292870992469676, + "grad_norm": 1.5058492420216791, + "learning_rate": 1.9074356933508545e-06, + "loss": 0.9818, + "step": 5869 + }, + { + "epoch": 0.5293772827704378, + "grad_norm": 3.273811410019897, + "learning_rate": 1.9068521127159477e-06, + "loss": 1.02, + "step": 5870 + }, + { + "epoch": 0.5294674662939081, + "grad_norm": 3.0300641443997924, + "learning_rate": 1.9062685400289322e-06, + "loss": 0.9401, + "step": 5871 + }, + { + "epoch": 0.5295576498173784, + "grad_norm": 1.5841923023352305, + "learning_rate": 1.9056849753396018e-06, + "loss": 0.964, + "step": 5872 + }, + { + "epoch": 0.5296478333408486, + "grad_norm": 1.976215859877699, + "learning_rate": 1.9051014186977485e-06, + "loss": 1.0231, + "step": 5873 + }, + { + "epoch": 0.5297380168643189, + "grad_norm": 1.3769003134265072, + "learning_rate": 1.9045178701531664e-06, + "loss": 0.9917, + "step": 5874 + }, + { + "epoch": 0.5298282003877891, + "grad_norm": 1.838857657822658, + "learning_rate": 1.903934329755645e-06, + "loss": 0.9853, + "step": 5875 + }, + { + "epoch": 0.5299183839112594, + "grad_norm": 1.5024859967586828, + "learning_rate": 1.9033507975549775e-06, + "loss": 1.0234, + "step": 5876 + }, + { + "epoch": 0.5300085674347297, + "grad_norm": 1.5971498316333002, + "learning_rate": 1.9027672736009525e-06, + "loss": 0.9802, + "step": 5877 + }, + { + "epoch": 0.5300987509581999, + "grad_norm": 1.4381496785751613, + "learning_rate": 1.9021837579433593e-06, + "loss": 0.9359, + "step": 5878 + }, + { + "epoch": 0.5301889344816701, + "grad_norm": 1.5758701494187202, + "learning_rate": 1.901600250631988e-06, + "loss": 0.9612, + "step": 5879 + }, + { + "epoch": 0.5302791180051405, + "grad_norm": 2.078110200069186, + "learning_rate": 1.901016751716625e-06, + "loss": 0.9786, + "step": 5880 + }, + { + "epoch": 0.5303693015286107, + "grad_norm": 1.6104878402056682, + "learning_rate": 1.9004332612470593e-06, + "loss": 0.952, + "step": 5881 + }, + { + "epoch": 0.530459485052081, + "grad_norm": 1.1964872982277126, + "learning_rate": 1.8998497792730763e-06, + "loss": 1.0128, + "step": 5882 + }, + { + "epoch": 0.5305496685755512, + "grad_norm": 1.4894890477500362, + "learning_rate": 1.8992663058444629e-06, + "loss": 1.0397, + "step": 5883 + }, + { + "epoch": 0.5306398520990215, + "grad_norm": 1.4642018450334187, + "learning_rate": 1.8986828410110032e-06, + "loss": 1.024, + "step": 5884 + }, + { + "epoch": 0.5307300356224918, + "grad_norm": 2.1275856891951883, + "learning_rate": 1.8980993848224823e-06, + "loss": 0.9801, + "step": 5885 + }, + { + "epoch": 0.530820219145962, + "grad_norm": 1.543774126613364, + "learning_rate": 1.8975159373286843e-06, + "loss": 0.9465, + "step": 5886 + }, + { + "epoch": 0.5309104026694323, + "grad_norm": 2.0204515260329985, + "learning_rate": 1.8969324985793904e-06, + "loss": 1.0272, + "step": 5887 + }, + { + "epoch": 0.5310005861929026, + "grad_norm": 2.2675938963033584, + "learning_rate": 1.8963490686243851e-06, + "loss": 0.9804, + "step": 5888 + }, + { + "epoch": 0.5310907697163728, + "grad_norm": 1.6598331630699166, + "learning_rate": 1.8957656475134486e-06, + "loss": 0.9264, + "step": 5889 + }, + { + "epoch": 0.531180953239843, + "grad_norm": 1.9596662314052697, + "learning_rate": 1.895182235296361e-06, + "loss": 0.9638, + "step": 5890 + }, + { + "epoch": 0.5312711367633134, + "grad_norm": 1.4725778088307977, + "learning_rate": 1.8945988320229042e-06, + "loss": 0.9099, + "step": 5891 + }, + { + "epoch": 0.5313613202867836, + "grad_norm": 2.015038220064402, + "learning_rate": 1.8940154377428553e-06, + "loss": 1.0704, + "step": 5892 + }, + { + "epoch": 0.5314515038102539, + "grad_norm": 1.5156606296402018, + "learning_rate": 1.8934320525059944e-06, + "loss": 1.0182, + "step": 5893 + }, + { + "epoch": 0.5315416873337241, + "grad_norm": 2.6142349213662066, + "learning_rate": 1.8928486763620984e-06, + "loss": 0.8742, + "step": 5894 + }, + { + "epoch": 0.5316318708571944, + "grad_norm": 2.339215278605076, + "learning_rate": 1.892265309360943e-06, + "loss": 0.9978, + "step": 5895 + }, + { + "epoch": 0.5317220543806647, + "grad_norm": 1.5745754547743052, + "learning_rate": 1.8916819515523067e-06, + "loss": 0.9733, + "step": 5896 + }, + { + "epoch": 0.5318122379041349, + "grad_norm": 1.3654202183310669, + "learning_rate": 1.891098602985963e-06, + "loss": 0.904, + "step": 5897 + }, + { + "epoch": 0.5319024214276051, + "grad_norm": 1.594061278817794, + "learning_rate": 1.8905152637116872e-06, + "loss": 1.0108, + "step": 5898 + }, + { + "epoch": 0.5319926049510755, + "grad_norm": 1.7046973993029695, + "learning_rate": 1.8899319337792527e-06, + "loss": 1.0252, + "step": 5899 + }, + { + "epoch": 0.5320827884745457, + "grad_norm": 1.2428853695738133, + "learning_rate": 1.8893486132384325e-06, + "loss": 1.0259, + "step": 5900 + }, + { + "epoch": 0.5321729719980159, + "grad_norm": 1.897084008757216, + "learning_rate": 1.888765302138999e-06, + "loss": 0.9563, + "step": 5901 + }, + { + "epoch": 0.5322631555214862, + "grad_norm": 1.7915695814694392, + "learning_rate": 1.8881820005307224e-06, + "loss": 0.9322, + "step": 5902 + }, + { + "epoch": 0.5323533390449565, + "grad_norm": 1.651845437520347, + "learning_rate": 1.8875987084633748e-06, + "loss": 0.9204, + "step": 5903 + }, + { + "epoch": 0.5324435225684268, + "grad_norm": 1.7902573506002353, + "learning_rate": 1.8870154259867246e-06, + "loss": 1.0218, + "step": 5904 + }, + { + "epoch": 0.532533706091897, + "grad_norm": 2.7428614991726796, + "learning_rate": 1.886432153150542e-06, + "loss": 0.922, + "step": 5905 + }, + { + "epoch": 0.5326238896153672, + "grad_norm": 1.6039772252109321, + "learning_rate": 1.8858488900045944e-06, + "loss": 0.9656, + "step": 5906 + }, + { + "epoch": 0.5327140731388376, + "grad_norm": 6.292395576880658, + "learning_rate": 1.885265636598648e-06, + "loss": 0.9446, + "step": 5907 + }, + { + "epoch": 0.5328042566623078, + "grad_norm": 1.5490885608197793, + "learning_rate": 1.884682392982471e-06, + "loss": 1.0209, + "step": 5908 + }, + { + "epoch": 0.532894440185778, + "grad_norm": 4.41343265867302, + "learning_rate": 1.8840991592058274e-06, + "loss": 1.0382, + "step": 5909 + }, + { + "epoch": 0.5329846237092484, + "grad_norm": 1.4086336914086839, + "learning_rate": 1.8835159353184828e-06, + "loss": 1.0042, + "step": 5910 + }, + { + "epoch": 0.5330748072327186, + "grad_norm": 1.4714804086006406, + "learning_rate": 1.8829327213702013e-06, + "loss": 1.0628, + "step": 5911 + }, + { + "epoch": 0.5331649907561888, + "grad_norm": 3.4419403527703394, + "learning_rate": 1.8823495174107452e-06, + "loss": 1.0604, + "step": 5912 + }, + { + "epoch": 0.5332551742796591, + "grad_norm": 2.138416236683648, + "learning_rate": 1.8817663234898773e-06, + "loss": 0.9587, + "step": 5913 + }, + { + "epoch": 0.5333453578031294, + "grad_norm": 1.5881839120049794, + "learning_rate": 1.881183139657358e-06, + "loss": 0.9933, + "step": 5914 + }, + { + "epoch": 0.5334355413265996, + "grad_norm": 1.6707635197438506, + "learning_rate": 1.8805999659629488e-06, + "loss": 1.0016, + "step": 5915 + }, + { + "epoch": 0.5335257248500699, + "grad_norm": 0.6295886631867736, + "learning_rate": 1.880016802456409e-06, + "loss": 0.7733, + "step": 5916 + }, + { + "epoch": 0.5336159083735401, + "grad_norm": 2.244168778550828, + "learning_rate": 1.8794336491874964e-06, + "loss": 0.9266, + "step": 5917 + }, + { + "epoch": 0.5337060918970105, + "grad_norm": 1.6903987994090925, + "learning_rate": 1.8788505062059708e-06, + "loss": 0.9466, + "step": 5918 + }, + { + "epoch": 0.5337962754204807, + "grad_norm": 1.5359637827676396, + "learning_rate": 1.8782673735615869e-06, + "loss": 0.972, + "step": 5919 + }, + { + "epoch": 0.5338864589439509, + "grad_norm": 1.6024759060481133, + "learning_rate": 1.8776842513041026e-06, + "loss": 0.9458, + "step": 5920 + }, + { + "epoch": 0.5339766424674212, + "grad_norm": 2.4872021214007107, + "learning_rate": 1.8771011394832727e-06, + "loss": 0.952, + "step": 5921 + }, + { + "epoch": 0.5340668259908915, + "grad_norm": 1.7518841273777375, + "learning_rate": 1.8765180381488501e-06, + "loss": 0.9697, + "step": 5922 + }, + { + "epoch": 0.5341570095143617, + "grad_norm": 0.734623481430865, + "learning_rate": 1.8759349473505905e-06, + "loss": 0.904, + "step": 5923 + }, + { + "epoch": 0.534247193037832, + "grad_norm": 1.654962435062653, + "learning_rate": 1.8753518671382447e-06, + "loss": 0.9056, + "step": 5924 + }, + { + "epoch": 0.5343373765613022, + "grad_norm": 3.0454487910533357, + "learning_rate": 1.8747687975615649e-06, + "loss": 0.9651, + "step": 5925 + }, + { + "epoch": 0.5344275600847725, + "grad_norm": 1.7039954384711375, + "learning_rate": 1.874185738670302e-06, + "loss": 0.8768, + "step": 5926 + }, + { + "epoch": 0.5345177436082428, + "grad_norm": 1.8841731429913164, + "learning_rate": 1.8736026905142057e-06, + "loss": 0.9718, + "step": 5927 + }, + { + "epoch": 0.534607927131713, + "grad_norm": 1.7196753351833916, + "learning_rate": 1.873019653143025e-06, + "loss": 0.9616, + "step": 5928 + }, + { + "epoch": 0.5346981106551832, + "grad_norm": 1.5893889955421205, + "learning_rate": 1.8724366266065069e-06, + "loss": 0.9995, + "step": 5929 + }, + { + "epoch": 0.5347882941786536, + "grad_norm": 1.6089951150178696, + "learning_rate": 1.8718536109543998e-06, + "loss": 1.0593, + "step": 5930 + }, + { + "epoch": 0.5348784777021238, + "grad_norm": 1.5240632166028758, + "learning_rate": 1.8712706062364485e-06, + "loss": 0.9995, + "step": 5931 + }, + { + "epoch": 0.534968661225594, + "grad_norm": 1.4783054943251757, + "learning_rate": 1.8706876125024e-06, + "loss": 0.8747, + "step": 5932 + }, + { + "epoch": 0.5350588447490644, + "grad_norm": 1.7317342500957564, + "learning_rate": 1.870104629801997e-06, + "loss": 1.0267, + "step": 5933 + }, + { + "epoch": 0.5351490282725346, + "grad_norm": 1.8994552960860183, + "learning_rate": 1.8695216581849823e-06, + "loss": 0.9954, + "step": 5934 + }, + { + "epoch": 0.5352392117960049, + "grad_norm": 1.5019630461564804, + "learning_rate": 1.8689386977011003e-06, + "loss": 0.953, + "step": 5935 + }, + { + "epoch": 0.5353293953194751, + "grad_norm": 0.6825545340887843, + "learning_rate": 1.8683557484000903e-06, + "loss": 0.8123, + "step": 5936 + }, + { + "epoch": 0.5354195788429454, + "grad_norm": 1.7809888414036585, + "learning_rate": 1.8677728103316947e-06, + "loss": 1.0159, + "step": 5937 + }, + { + "epoch": 0.5355097623664157, + "grad_norm": 1.3339089055306923, + "learning_rate": 1.8671898835456518e-06, + "loss": 0.9376, + "step": 5938 + }, + { + "epoch": 0.5355999458898859, + "grad_norm": 1.946639745713383, + "learning_rate": 1.8666069680917003e-06, + "loss": 1.0856, + "step": 5939 + }, + { + "epoch": 0.5356901294133561, + "grad_norm": 1.7646230369643963, + "learning_rate": 1.8660240640195775e-06, + "loss": 0.9412, + "step": 5940 + }, + { + "epoch": 0.5357803129368265, + "grad_norm": 3.081989940758094, + "learning_rate": 1.8654411713790203e-06, + "loss": 1.0214, + "step": 5941 + }, + { + "epoch": 0.5358704964602967, + "grad_norm": 1.6409462750443071, + "learning_rate": 1.8648582902197648e-06, + "loss": 1.0081, + "step": 5942 + }, + { + "epoch": 0.535960679983767, + "grad_norm": 3.6098773821350627, + "learning_rate": 1.8642754205915452e-06, + "loss": 0.9885, + "step": 5943 + }, + { + "epoch": 0.5360508635072372, + "grad_norm": 1.7739139538640258, + "learning_rate": 1.8636925625440943e-06, + "loss": 0.9345, + "step": 5944 + }, + { + "epoch": 0.5361410470307075, + "grad_norm": 1.840898150915655, + "learning_rate": 1.863109716127146e-06, + "loss": 0.8919, + "step": 5945 + }, + { + "epoch": 0.5362312305541778, + "grad_norm": 1.556528659369221, + "learning_rate": 1.8625268813904311e-06, + "loss": 1.0267, + "step": 5946 + }, + { + "epoch": 0.536321414077648, + "grad_norm": 1.4848258688973113, + "learning_rate": 1.8619440583836814e-06, + "loss": 1.0126, + "step": 5947 + }, + { + "epoch": 0.5364115976011182, + "grad_norm": 1.8429621290061295, + "learning_rate": 1.8613612471566249e-06, + "loss": 0.9879, + "step": 5948 + }, + { + "epoch": 0.5365017811245886, + "grad_norm": 2.1604874479445213, + "learning_rate": 1.8607784477589922e-06, + "loss": 0.927, + "step": 5949 + }, + { + "epoch": 0.5365919646480588, + "grad_norm": 1.7622114283463775, + "learning_rate": 1.8601956602405103e-06, + "loss": 1.0029, + "step": 5950 + }, + { + "epoch": 0.536682148171529, + "grad_norm": 1.3894401817802098, + "learning_rate": 1.8596128846509043e-06, + "loss": 1.0281, + "step": 5951 + }, + { + "epoch": 0.5367723316949993, + "grad_norm": 1.3896348407050694, + "learning_rate": 1.859030121039902e-06, + "loss": 0.9404, + "step": 5952 + }, + { + "epoch": 0.5368625152184696, + "grad_norm": 1.6921604932776813, + "learning_rate": 1.8584473694572268e-06, + "loss": 0.9373, + "step": 5953 + }, + { + "epoch": 0.5369526987419398, + "grad_norm": 2.2008078002411513, + "learning_rate": 1.8578646299526026e-06, + "loss": 1.0083, + "step": 5954 + }, + { + "epoch": 0.5370428822654101, + "grad_norm": 1.4368964047557076, + "learning_rate": 1.8572819025757518e-06, + "loss": 0.9575, + "step": 5955 + }, + { + "epoch": 0.5371330657888804, + "grad_norm": 1.7566451880404448, + "learning_rate": 1.8566991873763959e-06, + "loss": 0.9986, + "step": 5956 + }, + { + "epoch": 0.5372232493123507, + "grad_norm": 1.5790904651598696, + "learning_rate": 1.856116484404256e-06, + "loss": 0.9315, + "step": 5957 + }, + { + "epoch": 0.5373134328358209, + "grad_norm": 1.8224966673434624, + "learning_rate": 1.8555337937090502e-06, + "loss": 0.9495, + "step": 5958 + }, + { + "epoch": 0.5374036163592911, + "grad_norm": 2.4265305594990685, + "learning_rate": 1.8549511153404984e-06, + "loss": 0.9677, + "step": 5959 + }, + { + "epoch": 0.5374937998827615, + "grad_norm": 1.5438252450002947, + "learning_rate": 1.854368449348317e-06, + "loss": 1.0927, + "step": 5960 + }, + { + "epoch": 0.5375839834062317, + "grad_norm": 1.8762288199493018, + "learning_rate": 1.853785795782222e-06, + "loss": 0.943, + "step": 5961 + }, + { + "epoch": 0.5376741669297019, + "grad_norm": 1.5999349372271316, + "learning_rate": 1.85320315469193e-06, + "loss": 0.8544, + "step": 5962 + }, + { + "epoch": 0.5377643504531722, + "grad_norm": 0.6551055920642451, + "learning_rate": 1.8526205261271534e-06, + "loss": 0.8185, + "step": 5963 + }, + { + "epoch": 0.5378545339766425, + "grad_norm": 1.4606335981750098, + "learning_rate": 1.852037910137607e-06, + "loss": 0.8976, + "step": 5964 + }, + { + "epoch": 0.5379447175001127, + "grad_norm": 1.383857375392369, + "learning_rate": 1.851455306773002e-06, + "loss": 1.0682, + "step": 5965 + }, + { + "epoch": 0.538034901023583, + "grad_norm": 1.4110202836836885, + "learning_rate": 1.8508727160830483e-06, + "loss": 0.9607, + "step": 5966 + }, + { + "epoch": 0.5381250845470532, + "grad_norm": 1.56884530070532, + "learning_rate": 1.8502901381174575e-06, + "loss": 1.0562, + "step": 5967 + }, + { + "epoch": 0.5382152680705236, + "grad_norm": 2.0551912683731333, + "learning_rate": 1.8497075729259372e-06, + "loss": 0.8929, + "step": 5968 + }, + { + "epoch": 0.5383054515939938, + "grad_norm": 0.7620264850003643, + "learning_rate": 1.8491250205581963e-06, + "loss": 0.8787, + "step": 5969 + }, + { + "epoch": 0.538395635117464, + "grad_norm": 1.662979848348495, + "learning_rate": 1.8485424810639393e-06, + "loss": 0.9884, + "step": 5970 + }, + { + "epoch": 0.5384858186409343, + "grad_norm": 1.7507476571914575, + "learning_rate": 1.847959954492874e-06, + "loss": 1.0413, + "step": 5971 + }, + { + "epoch": 0.5385760021644046, + "grad_norm": 2.1087408172099957, + "learning_rate": 1.8473774408947035e-06, + "loss": 1.0805, + "step": 5972 + }, + { + "epoch": 0.5386661856878748, + "grad_norm": 0.6446681021852234, + "learning_rate": 1.8467949403191308e-06, + "loss": 0.7897, + "step": 5973 + }, + { + "epoch": 0.5387563692113451, + "grad_norm": 1.4950770018276793, + "learning_rate": 1.8462124528158592e-06, + "loss": 1.0381, + "step": 5974 + }, + { + "epoch": 0.5388465527348153, + "grad_norm": 1.4726448362196531, + "learning_rate": 1.8456299784345881e-06, + "loss": 0.9724, + "step": 5975 + }, + { + "epoch": 0.5389367362582856, + "grad_norm": 1.7227880033941123, + "learning_rate": 1.8450475172250194e-06, + "loss": 1.0055, + "step": 5976 + }, + { + "epoch": 0.5390269197817559, + "grad_norm": 1.739959249786606, + "learning_rate": 1.844465069236851e-06, + "loss": 0.9649, + "step": 5977 + }, + { + "epoch": 0.5391171033052261, + "grad_norm": 1.6862441440760547, + "learning_rate": 1.8438826345197796e-06, + "loss": 0.892, + "step": 5978 + }, + { + "epoch": 0.5392072868286965, + "grad_norm": 1.8080190318042715, + "learning_rate": 1.8433002131235036e-06, + "loss": 0.9715, + "step": 5979 + }, + { + "epoch": 0.5392974703521667, + "grad_norm": 1.7274915862258076, + "learning_rate": 1.8427178050977167e-06, + "loss": 1.0073, + "step": 5980 + }, + { + "epoch": 0.5393876538756369, + "grad_norm": 1.4514444949962964, + "learning_rate": 1.8421354104921143e-06, + "loss": 1.0612, + "step": 5981 + }, + { + "epoch": 0.5394778373991072, + "grad_norm": 0.6122649938928881, + "learning_rate": 1.8415530293563894e-06, + "loss": 0.7857, + "step": 5982 + }, + { + "epoch": 0.5395680209225775, + "grad_norm": 1.4547854545220629, + "learning_rate": 1.8409706617402333e-06, + "loss": 0.9689, + "step": 5983 + }, + { + "epoch": 0.5396582044460477, + "grad_norm": 1.580962321532782, + "learning_rate": 1.8403883076933378e-06, + "loss": 1.0043, + "step": 5984 + }, + { + "epoch": 0.539748387969518, + "grad_norm": 1.620184218093261, + "learning_rate": 1.839805967265391e-06, + "loss": 0.9768, + "step": 5985 + }, + { + "epoch": 0.5398385714929882, + "grad_norm": 1.6271917527628388, + "learning_rate": 1.839223640506083e-06, + "loss": 1.0432, + "step": 5986 + }, + { + "epoch": 0.5399287550164585, + "grad_norm": 4.500241379092101, + "learning_rate": 1.8386413274650998e-06, + "loss": 0.9792, + "step": 5987 + }, + { + "epoch": 0.5400189385399288, + "grad_norm": 2.0096147219221914, + "learning_rate": 1.8380590281921294e-06, + "loss": 1.0326, + "step": 5988 + }, + { + "epoch": 0.540109122063399, + "grad_norm": 3.9147999691194664, + "learning_rate": 1.8374767427368552e-06, + "loss": 0.8895, + "step": 5989 + }, + { + "epoch": 0.5401993055868692, + "grad_norm": 5.511584306641228, + "learning_rate": 1.8368944711489608e-06, + "loss": 1.013, + "step": 5990 + }, + { + "epoch": 0.5402894891103396, + "grad_norm": 2.203383734910981, + "learning_rate": 1.8363122134781304e-06, + "loss": 1.0137, + "step": 5991 + }, + { + "epoch": 0.5403796726338098, + "grad_norm": 1.5402841899032442, + "learning_rate": 1.835729969774044e-06, + "loss": 1.0137, + "step": 5992 + }, + { + "epoch": 0.54046985615728, + "grad_norm": 1.554704445598015, + "learning_rate": 1.8351477400863823e-06, + "loss": 0.9947, + "step": 5993 + }, + { + "epoch": 0.5405600396807503, + "grad_norm": 1.4330111366951954, + "learning_rate": 1.8345655244648249e-06, + "loss": 0.9662, + "step": 5994 + }, + { + "epoch": 0.5406502232042206, + "grad_norm": 1.3055642659773765, + "learning_rate": 1.8339833229590486e-06, + "loss": 0.9757, + "step": 5995 + }, + { + "epoch": 0.5407404067276909, + "grad_norm": 1.7307539246363253, + "learning_rate": 1.833401135618731e-06, + "loss": 1.0587, + "step": 5996 + }, + { + "epoch": 0.5408305902511611, + "grad_norm": 1.4496092565809369, + "learning_rate": 1.8328189624935466e-06, + "loss": 0.9145, + "step": 5997 + }, + { + "epoch": 0.5409207737746313, + "grad_norm": 1.599441314550808, + "learning_rate": 1.832236803633171e-06, + "loss": 1.0224, + "step": 5998 + }, + { + "epoch": 0.5410109572981017, + "grad_norm": 2.139516835973102, + "learning_rate": 1.831654659087276e-06, + "loss": 0.9545, + "step": 5999 + }, + { + "epoch": 0.5411011408215719, + "grad_norm": 2.3938458256676953, + "learning_rate": 1.831072528905533e-06, + "loss": 1.0025, + "step": 6000 + }, + { + "epoch": 0.5411913243450421, + "grad_norm": 1.495804801109169, + "learning_rate": 1.8304904131376142e-06, + "loss": 1.0384, + "step": 6001 + }, + { + "epoch": 0.5412815078685124, + "grad_norm": 1.6913013734084648, + "learning_rate": 1.8299083118331874e-06, + "loss": 1.0026, + "step": 6002 + }, + { + "epoch": 0.5413716913919827, + "grad_norm": 2.0523223608650016, + "learning_rate": 1.8293262250419217e-06, + "loss": 1.1196, + "step": 6003 + }, + { + "epoch": 0.541461874915453, + "grad_norm": 1.7586345869531443, + "learning_rate": 1.828744152813484e-06, + "loss": 1.1016, + "step": 6004 + }, + { + "epoch": 0.5415520584389232, + "grad_norm": 1.826614407861529, + "learning_rate": 1.8281620951975382e-06, + "loss": 0.9393, + "step": 6005 + }, + { + "epoch": 0.5416422419623935, + "grad_norm": 1.619788027776656, + "learning_rate": 1.827580052243751e-06, + "loss": 0.9481, + "step": 6006 + }, + { + "epoch": 0.5417324254858638, + "grad_norm": 1.5411416353987382, + "learning_rate": 1.826998024001784e-06, + "loss": 1.0188, + "step": 6007 + }, + { + "epoch": 0.541822609009334, + "grad_norm": 3.2480895956582447, + "learning_rate": 1.8264160105212995e-06, + "loss": 1.0147, + "step": 6008 + }, + { + "epoch": 0.5419127925328042, + "grad_norm": 1.9224698766523605, + "learning_rate": 1.8258340118519582e-06, + "loss": 0.9539, + "step": 6009 + }, + { + "epoch": 0.5420029760562746, + "grad_norm": 4.720356042655426, + "learning_rate": 1.82525202804342e-06, + "loss": 0.9696, + "step": 6010 + }, + { + "epoch": 0.5420931595797448, + "grad_norm": 1.8207880143986246, + "learning_rate": 1.8246700591453415e-06, + "loss": 0.9669, + "step": 6011 + }, + { + "epoch": 0.542183343103215, + "grad_norm": 1.513848748486677, + "learning_rate": 1.8240881052073801e-06, + "loss": 0.9867, + "step": 6012 + }, + { + "epoch": 0.5422735266266853, + "grad_norm": 5.278780387991487, + "learning_rate": 1.8235061662791923e-06, + "loss": 0.9557, + "step": 6013 + }, + { + "epoch": 0.5423637101501556, + "grad_norm": 1.7727857560067544, + "learning_rate": 1.8229242424104309e-06, + "loss": 1.0553, + "step": 6014 + }, + { + "epoch": 0.5424538936736258, + "grad_norm": 0.6108518945372797, + "learning_rate": 1.8223423336507503e-06, + "loss": 0.7971, + "step": 6015 + }, + { + "epoch": 0.5425440771970961, + "grad_norm": 1.73464276420512, + "learning_rate": 1.8217604400498012e-06, + "loss": 0.9692, + "step": 6016 + }, + { + "epoch": 0.5426342607205663, + "grad_norm": 1.3903586743387553, + "learning_rate": 1.8211785616572333e-06, + "loss": 1.0351, + "step": 6017 + }, + { + "epoch": 0.5427244442440367, + "grad_norm": 1.658593445255433, + "learning_rate": 1.8205966985226975e-06, + "loss": 1.0545, + "step": 6018 + }, + { + "epoch": 0.5428146277675069, + "grad_norm": 1.316704509690243, + "learning_rate": 1.8200148506958397e-06, + "loss": 0.9942, + "step": 6019 + }, + { + "epoch": 0.5429048112909771, + "grad_norm": 0.6437528128701393, + "learning_rate": 1.819433018226308e-06, + "loss": 0.843, + "step": 6020 + }, + { + "epoch": 0.5429949948144474, + "grad_norm": 0.6001957067555983, + "learning_rate": 1.8188512011637471e-06, + "loss": 0.787, + "step": 6021 + }, + { + "epoch": 0.5430851783379177, + "grad_norm": 5.179975084517367, + "learning_rate": 1.8182693995578e-06, + "loss": 0.9669, + "step": 6022 + }, + { + "epoch": 0.5431753618613879, + "grad_norm": 3.3556127017743096, + "learning_rate": 1.8176876134581098e-06, + "loss": 0.8652, + "step": 6023 + }, + { + "epoch": 0.5432655453848582, + "grad_norm": 5.084633374346048, + "learning_rate": 1.8171058429143176e-06, + "loss": 0.9325, + "step": 6024 + }, + { + "epoch": 0.5433557289083284, + "grad_norm": 1.467863424033793, + "learning_rate": 1.8165240879760637e-06, + "loss": 1.0232, + "step": 6025 + }, + { + "epoch": 0.5434459124317987, + "grad_norm": 1.437957607907153, + "learning_rate": 1.8159423486929862e-06, + "loss": 1.0161, + "step": 6026 + }, + { + "epoch": 0.543536095955269, + "grad_norm": 1.609945399857163, + "learning_rate": 1.815360625114722e-06, + "loss": 0.9849, + "step": 6027 + }, + { + "epoch": 0.5436262794787392, + "grad_norm": 1.5105101808560333, + "learning_rate": 1.814778917290908e-06, + "loss": 0.9728, + "step": 6028 + }, + { + "epoch": 0.5437164630022095, + "grad_norm": 1.672545824321219, + "learning_rate": 1.8141972252711773e-06, + "loss": 1.004, + "step": 6029 + }, + { + "epoch": 0.5438066465256798, + "grad_norm": 2.093725175852722, + "learning_rate": 1.8136155491051645e-06, + "loss": 0.9813, + "step": 6030 + }, + { + "epoch": 0.54389683004915, + "grad_norm": 0.6773115922626616, + "learning_rate": 1.8130338888424998e-06, + "loss": 0.8309, + "step": 6031 + }, + { + "epoch": 0.5439870135726202, + "grad_norm": 1.611561447359207, + "learning_rate": 1.812452244532816e-06, + "loss": 1.0063, + "step": 6032 + }, + { + "epoch": 0.5440771970960906, + "grad_norm": 2.5213713794213435, + "learning_rate": 1.8118706162257405e-06, + "loss": 0.9251, + "step": 6033 + }, + { + "epoch": 0.5441673806195608, + "grad_norm": 1.614154200797063, + "learning_rate": 1.8112890039709002e-06, + "loss": 1.0677, + "step": 6034 + }, + { + "epoch": 0.5442575641430311, + "grad_norm": 1.954699743584725, + "learning_rate": 1.8107074078179238e-06, + "loss": 0.9923, + "step": 6035 + }, + { + "epoch": 0.5443477476665013, + "grad_norm": 1.4599251733182192, + "learning_rate": 1.8101258278164348e-06, + "loss": 0.9368, + "step": 6036 + }, + { + "epoch": 0.5444379311899716, + "grad_norm": 1.580558637777287, + "learning_rate": 1.8095442640160575e-06, + "loss": 1.0945, + "step": 6037 + }, + { + "epoch": 0.5445281147134419, + "grad_norm": 1.8487752546193643, + "learning_rate": 1.8089627164664132e-06, + "loss": 0.9527, + "step": 6038 + }, + { + "epoch": 0.5446182982369121, + "grad_norm": 1.3581162686863852, + "learning_rate": 1.8083811852171233e-06, + "loss": 0.91, + "step": 6039 + }, + { + "epoch": 0.5447084817603823, + "grad_norm": 1.8484535352022669, + "learning_rate": 1.8077996703178078e-06, + "loss": 0.9846, + "step": 6040 + }, + { + "epoch": 0.5447986652838527, + "grad_norm": 1.515241358072938, + "learning_rate": 1.8072181718180833e-06, + "loss": 0.9369, + "step": 6041 + }, + { + "epoch": 0.5448888488073229, + "grad_norm": 2.1955027650534675, + "learning_rate": 1.806636689767568e-06, + "loss": 1.0698, + "step": 6042 + }, + { + "epoch": 0.5449790323307931, + "grad_norm": 1.5033708684567457, + "learning_rate": 1.8060552242158769e-06, + "loss": 0.9263, + "step": 6043 + }, + { + "epoch": 0.5450692158542634, + "grad_norm": 1.6047920230153778, + "learning_rate": 1.8054737752126224e-06, + "loss": 0.9558, + "step": 6044 + }, + { + "epoch": 0.5451593993777337, + "grad_norm": 1.3836396755219427, + "learning_rate": 1.804892342807419e-06, + "loss": 0.9734, + "step": 6045 + }, + { + "epoch": 0.545249582901204, + "grad_norm": 1.9264100694336173, + "learning_rate": 1.8043109270498756e-06, + "loss": 1.0533, + "step": 6046 + }, + { + "epoch": 0.5453397664246742, + "grad_norm": 1.845691692982949, + "learning_rate": 1.803729527989604e-06, + "loss": 0.9973, + "step": 6047 + }, + { + "epoch": 0.5454299499481444, + "grad_norm": 1.422494019963411, + "learning_rate": 1.8031481456762112e-06, + "loss": 0.9769, + "step": 6048 + }, + { + "epoch": 0.5455201334716148, + "grad_norm": 2.7254230555840646, + "learning_rate": 1.8025667801593033e-06, + "loss": 0.914, + "step": 6049 + }, + { + "epoch": 0.545610316995085, + "grad_norm": 2.28031630547232, + "learning_rate": 1.8019854314884871e-06, + "loss": 0.9208, + "step": 6050 + }, + { + "epoch": 0.5457005005185552, + "grad_norm": 1.4868710061557233, + "learning_rate": 1.8014040997133652e-06, + "loss": 1.0367, + "step": 6051 + }, + { + "epoch": 0.5457906840420256, + "grad_norm": 0.7920781287661603, + "learning_rate": 1.8008227848835414e-06, + "loss": 0.8515, + "step": 6052 + }, + { + "epoch": 0.5458808675654958, + "grad_norm": 2.7453226713492564, + "learning_rate": 1.8002414870486144e-06, + "loss": 0.9337, + "step": 6053 + }, + { + "epoch": 0.545971051088966, + "grad_norm": 1.284797919967293, + "learning_rate": 1.7996602062581864e-06, + "loss": 0.9405, + "step": 6054 + }, + { + "epoch": 0.5460612346124363, + "grad_norm": 1.6250343319127198, + "learning_rate": 1.7990789425618544e-06, + "loss": 1.0345, + "step": 6055 + }, + { + "epoch": 0.5461514181359066, + "grad_norm": 1.3030233220013536, + "learning_rate": 1.7984976960092137e-06, + "loss": 0.9675, + "step": 6056 + }, + { + "epoch": 0.5462416016593769, + "grad_norm": 2.1041238484574087, + "learning_rate": 1.7979164666498617e-06, + "loss": 0.9537, + "step": 6057 + }, + { + "epoch": 0.5463317851828471, + "grad_norm": 0.6212015118471848, + "learning_rate": 1.7973352545333901e-06, + "loss": 0.7903, + "step": 6058 + }, + { + "epoch": 0.5464219687063173, + "grad_norm": 1.722161531233538, + "learning_rate": 1.796754059709393e-06, + "loss": 0.9065, + "step": 6059 + }, + { + "epoch": 0.5465121522297877, + "grad_norm": 2.244214992104396, + "learning_rate": 1.7961728822274603e-06, + "loss": 1.0474, + "step": 6060 + }, + { + "epoch": 0.5466023357532579, + "grad_norm": 1.5401444093694707, + "learning_rate": 1.7955917221371802e-06, + "loss": 0.9406, + "step": 6061 + }, + { + "epoch": 0.5466925192767281, + "grad_norm": 2.2622141452004754, + "learning_rate": 1.7950105794881422e-06, + "loss": 0.9042, + "step": 6062 + }, + { + "epoch": 0.5467827028001984, + "grad_norm": 1.4093543226666432, + "learning_rate": 1.7944294543299317e-06, + "loss": 0.9742, + "step": 6063 + }, + { + "epoch": 0.5468728863236687, + "grad_norm": 1.99683833558715, + "learning_rate": 1.7938483467121333e-06, + "loss": 1.0381, + "step": 6064 + }, + { + "epoch": 0.5469630698471389, + "grad_norm": 1.8039443562589494, + "learning_rate": 1.7932672566843313e-06, + "loss": 0.9664, + "step": 6065 + }, + { + "epoch": 0.5470532533706092, + "grad_norm": 1.6158245443046393, + "learning_rate": 1.7926861842961065e-06, + "loss": 0.9083, + "step": 6066 + }, + { + "epoch": 0.5471434368940794, + "grad_norm": 1.755702944227508, + "learning_rate": 1.7921051295970399e-06, + "loss": 0.9109, + "step": 6067 + }, + { + "epoch": 0.5472336204175497, + "grad_norm": 1.6821376938721841, + "learning_rate": 1.7915240926367092e-06, + "loss": 0.9504, + "step": 6068 + }, + { + "epoch": 0.54732380394102, + "grad_norm": 1.583614132057914, + "learning_rate": 1.7909430734646932e-06, + "loss": 1.0287, + "step": 6069 + }, + { + "epoch": 0.5474139874644902, + "grad_norm": 0.671912547104757, + "learning_rate": 1.790362072130567e-06, + "loss": 0.8236, + "step": 6070 + }, + { + "epoch": 0.5475041709879604, + "grad_norm": 1.669320838246692, + "learning_rate": 1.7897810886839037e-06, + "loss": 0.9712, + "step": 6071 + }, + { + "epoch": 0.5475943545114308, + "grad_norm": 1.8828711282933035, + "learning_rate": 1.7892001231742782e-06, + "loss": 0.9963, + "step": 6072 + }, + { + "epoch": 0.547684538034901, + "grad_norm": 4.154885663273772, + "learning_rate": 1.7886191756512594e-06, + "loss": 0.996, + "step": 6073 + }, + { + "epoch": 0.5477747215583713, + "grad_norm": 1.5652937517739767, + "learning_rate": 1.7880382461644192e-06, + "loss": 1.0658, + "step": 6074 + }, + { + "epoch": 0.5478649050818416, + "grad_norm": 1.745073809989188, + "learning_rate": 1.7874573347633235e-06, + "loss": 0.9663, + "step": 6075 + }, + { + "epoch": 0.5479550886053118, + "grad_norm": 2.2174666010632618, + "learning_rate": 1.7868764414975408e-06, + "loss": 1.0566, + "step": 6076 + }, + { + "epoch": 0.5480452721287821, + "grad_norm": 1.999906299303458, + "learning_rate": 1.7862955664166353e-06, + "loss": 0.9343, + "step": 6077 + }, + { + "epoch": 0.5481354556522523, + "grad_norm": 1.4149079719514612, + "learning_rate": 1.78571470957017e-06, + "loss": 1.0695, + "step": 6078 + }, + { + "epoch": 0.5482256391757226, + "grad_norm": 1.64462243776469, + "learning_rate": 1.7851338710077074e-06, + "loss": 1.0486, + "step": 6079 + }, + { + "epoch": 0.5483158226991929, + "grad_norm": 1.795030858297935, + "learning_rate": 1.7845530507788076e-06, + "loss": 1.0958, + "step": 6080 + }, + { + "epoch": 0.5484060062226631, + "grad_norm": 1.5634732076595799, + "learning_rate": 1.7839722489330298e-06, + "loss": 0.9557, + "step": 6081 + }, + { + "epoch": 0.5484961897461333, + "grad_norm": 1.5561721077541604, + "learning_rate": 1.7833914655199308e-06, + "loss": 0.9536, + "step": 6082 + }, + { + "epoch": 0.5485863732696037, + "grad_norm": 2.470627276194938, + "learning_rate": 1.7828107005890658e-06, + "loss": 1.0314, + "step": 6083 + }, + { + "epoch": 0.5486765567930739, + "grad_norm": 2.0591402142857156, + "learning_rate": 1.7822299541899898e-06, + "loss": 0.9072, + "step": 6084 + }, + { + "epoch": 0.5487667403165442, + "grad_norm": 1.7610305829227548, + "learning_rate": 1.7816492263722545e-06, + "loss": 0.9848, + "step": 6085 + }, + { + "epoch": 0.5488569238400144, + "grad_norm": 2.5261742835610717, + "learning_rate": 1.781068517185412e-06, + "loss": 1.0109, + "step": 6086 + }, + { + "epoch": 0.5489471073634847, + "grad_norm": 1.9253790321241009, + "learning_rate": 1.7804878266790104e-06, + "loss": 1.0508, + "step": 6087 + }, + { + "epoch": 0.549037290886955, + "grad_norm": 1.5343774271089272, + "learning_rate": 1.779907154902597e-06, + "loss": 0.9229, + "step": 6088 + }, + { + "epoch": 0.5491274744104252, + "grad_norm": 1.5183459368425865, + "learning_rate": 1.7793265019057198e-06, + "loss": 0.9158, + "step": 6089 + }, + { + "epoch": 0.5492176579338954, + "grad_norm": 0.7294809118418227, + "learning_rate": 1.7787458677379212e-06, + "loss": 0.8319, + "step": 6090 + }, + { + "epoch": 0.5493078414573658, + "grad_norm": 1.6542780671294715, + "learning_rate": 1.7781652524487463e-06, + "loss": 1.0429, + "step": 6091 + }, + { + "epoch": 0.549398024980836, + "grad_norm": 1.6164584520630567, + "learning_rate": 1.777584656087735e-06, + "loss": 0.9301, + "step": 6092 + }, + { + "epoch": 0.5494882085043062, + "grad_norm": 1.8671717942347434, + "learning_rate": 1.777004078704427e-06, + "loss": 0.9053, + "step": 6093 + }, + { + "epoch": 0.5495783920277765, + "grad_norm": 1.538382907838738, + "learning_rate": 1.7764235203483603e-06, + "loss": 1.0121, + "step": 6094 + }, + { + "epoch": 0.5496685755512468, + "grad_norm": 2.511183045303114, + "learning_rate": 1.775842981069072e-06, + "loss": 0.9573, + "step": 6095 + }, + { + "epoch": 0.549758759074717, + "grad_norm": 1.52174862879401, + "learning_rate": 1.7752624609160966e-06, + "loss": 0.9968, + "step": 6096 + }, + { + "epoch": 0.5498489425981873, + "grad_norm": 1.8198390885368145, + "learning_rate": 1.7746819599389665e-06, + "loss": 1.0236, + "step": 6097 + }, + { + "epoch": 0.5499391261216575, + "grad_norm": 1.4884859680190963, + "learning_rate": 1.774101478187215e-06, + "loss": 0.9894, + "step": 6098 + }, + { + "epoch": 0.5500293096451279, + "grad_norm": 2.112276808104562, + "learning_rate": 1.773521015710371e-06, + "loss": 1.0536, + "step": 6099 + }, + { + "epoch": 0.5501194931685981, + "grad_norm": 1.9069646527555593, + "learning_rate": 1.7729405725579614e-06, + "loss": 1.0671, + "step": 6100 + }, + { + "epoch": 0.5502096766920683, + "grad_norm": 1.4165613845898226, + "learning_rate": 1.7723601487795151e-06, + "loss": 0.9663, + "step": 6101 + }, + { + "epoch": 0.5502998602155387, + "grad_norm": 1.4038007387687848, + "learning_rate": 1.7717797444245557e-06, + "loss": 0.9355, + "step": 6102 + }, + { + "epoch": 0.5503900437390089, + "grad_norm": 1.4078730224064415, + "learning_rate": 1.7711993595426076e-06, + "loss": 0.921, + "step": 6103 + }, + { + "epoch": 0.5504802272624791, + "grad_norm": 2.2925045446852104, + "learning_rate": 1.7706189941831915e-06, + "loss": 1.0967, + "step": 6104 + }, + { + "epoch": 0.5505704107859494, + "grad_norm": 1.5429046231235126, + "learning_rate": 1.770038648395827e-06, + "loss": 0.948, + "step": 6105 + }, + { + "epoch": 0.5506605943094197, + "grad_norm": 0.6098209607701601, + "learning_rate": 1.7694583222300336e-06, + "loss": 0.8312, + "step": 6106 + }, + { + "epoch": 0.55075077783289, + "grad_norm": 1.7643375260052045, + "learning_rate": 1.7688780157353272e-06, + "loss": 0.9667, + "step": 6107 + }, + { + "epoch": 0.5508409613563602, + "grad_norm": 1.5972684282262866, + "learning_rate": 1.768297728961223e-06, + "loss": 0.9331, + "step": 6108 + }, + { + "epoch": 0.5509311448798304, + "grad_norm": 2.11498469660568, + "learning_rate": 1.7677174619572342e-06, + "loss": 0.992, + "step": 6109 + }, + { + "epoch": 0.5510213284033008, + "grad_norm": 1.5723072496673693, + "learning_rate": 1.7671372147728717e-06, + "loss": 1.0036, + "step": 6110 + }, + { + "epoch": 0.551111511926771, + "grad_norm": 1.67431605025605, + "learning_rate": 1.7665569874576471e-06, + "loss": 1.0531, + "step": 6111 + }, + { + "epoch": 0.5512016954502412, + "grad_norm": 2.5951380864560285, + "learning_rate": 1.7659767800610664e-06, + "loss": 0.9956, + "step": 6112 + }, + { + "epoch": 0.5512918789737115, + "grad_norm": 1.577333308388932, + "learning_rate": 1.7653965926326379e-06, + "loss": 0.8936, + "step": 6113 + }, + { + "epoch": 0.5513820624971818, + "grad_norm": 1.283006673374794, + "learning_rate": 1.764816425221866e-06, + "loss": 0.9466, + "step": 6114 + }, + { + "epoch": 0.551472246020652, + "grad_norm": 2.994181553709579, + "learning_rate": 1.7642362778782524e-06, + "loss": 0.9031, + "step": 6115 + }, + { + "epoch": 0.5515624295441223, + "grad_norm": 1.9346571955266083, + "learning_rate": 1.7636561506513005e-06, + "loss": 0.9568, + "step": 6116 + }, + { + "epoch": 0.5516526130675925, + "grad_norm": 2.202104805727678, + "learning_rate": 1.7630760435905083e-06, + "loss": 0.9622, + "step": 6117 + }, + { + "epoch": 0.5517427965910628, + "grad_norm": 1.7411872489888214, + "learning_rate": 1.762495956745375e-06, + "loss": 0.9469, + "step": 6118 + }, + { + "epoch": 0.5518329801145331, + "grad_norm": 1.879806968843425, + "learning_rate": 1.7619158901653962e-06, + "loss": 1.0274, + "step": 6119 + }, + { + "epoch": 0.5519231636380033, + "grad_norm": 1.941016543782171, + "learning_rate": 1.761335843900066e-06, + "loss": 0.9857, + "step": 6120 + }, + { + "epoch": 0.5520133471614735, + "grad_norm": 1.4877997793524376, + "learning_rate": 1.7607558179988785e-06, + "loss": 0.9735, + "step": 6121 + }, + { + "epoch": 0.5521035306849439, + "grad_norm": 1.3905088826252088, + "learning_rate": 1.760175812511323e-06, + "loss": 0.9163, + "step": 6122 + }, + { + "epoch": 0.5521937142084141, + "grad_norm": 1.787062080466235, + "learning_rate": 1.75959582748689e-06, + "loss": 0.9457, + "step": 6123 + }, + { + "epoch": 0.5522838977318844, + "grad_norm": 1.6061525396918772, + "learning_rate": 1.7590158629750657e-06, + "loss": 1.0167, + "step": 6124 + }, + { + "epoch": 0.5523740812553547, + "grad_norm": 1.4463688469722027, + "learning_rate": 1.7584359190253376e-06, + "loss": 1.0393, + "step": 6125 + }, + { + "epoch": 0.5524642647788249, + "grad_norm": 1.8892844625845402, + "learning_rate": 1.7578559956871892e-06, + "loss": 1.0477, + "step": 6126 + }, + { + "epoch": 0.5525544483022952, + "grad_norm": 1.8433116456400611, + "learning_rate": 1.7572760930101012e-06, + "loss": 0.985, + "step": 6127 + }, + { + "epoch": 0.5526446318257654, + "grad_norm": 0.674917341398778, + "learning_rate": 1.7566962110435563e-06, + "loss": 0.8703, + "step": 6128 + }, + { + "epoch": 0.5527348153492357, + "grad_norm": 1.424958961297425, + "learning_rate": 1.7561163498370313e-06, + "loss": 1.0062, + "step": 6129 + }, + { + "epoch": 0.552824998872706, + "grad_norm": 1.4312802807132565, + "learning_rate": 1.755536509440005e-06, + "loss": 0.9512, + "step": 6130 + }, + { + "epoch": 0.5529151823961762, + "grad_norm": 1.798971992563043, + "learning_rate": 1.7549566899019519e-06, + "loss": 0.981, + "step": 6131 + }, + { + "epoch": 0.5530053659196464, + "grad_norm": 1.5775655377976388, + "learning_rate": 1.754376891272344e-06, + "loss": 1.0064, + "step": 6132 + }, + { + "epoch": 0.5530955494431168, + "grad_norm": 1.36098148524509, + "learning_rate": 1.753797113600655e-06, + "loss": 0.9362, + "step": 6133 + }, + { + "epoch": 0.553185732966587, + "grad_norm": 1.6222831315460706, + "learning_rate": 1.7532173569363535e-06, + "loss": 1.1226, + "step": 6134 + }, + { + "epoch": 0.5532759164900573, + "grad_norm": 1.7307685953937897, + "learning_rate": 1.7526376213289077e-06, + "loss": 0.8055, + "step": 6135 + }, + { + "epoch": 0.5533661000135275, + "grad_norm": 1.9353049980719148, + "learning_rate": 1.7520579068277844e-06, + "loss": 0.9693, + "step": 6136 + }, + { + "epoch": 0.5534562835369978, + "grad_norm": 1.4770800778344582, + "learning_rate": 1.7514782134824472e-06, + "loss": 1.0158, + "step": 6137 + }, + { + "epoch": 0.5535464670604681, + "grad_norm": 1.6082594625608848, + "learning_rate": 1.7508985413423599e-06, + "loss": 0.9817, + "step": 6138 + }, + { + "epoch": 0.5536366505839383, + "grad_norm": 1.6267810011507253, + "learning_rate": 1.7503188904569814e-06, + "loss": 1.0539, + "step": 6139 + }, + { + "epoch": 0.5537268341074085, + "grad_norm": 1.6333724526502025, + "learning_rate": 1.7497392608757728e-06, + "loss": 0.9521, + "step": 6140 + }, + { + "epoch": 0.5538170176308789, + "grad_norm": 1.4227769108172623, + "learning_rate": 1.7491596526481897e-06, + "loss": 1.0032, + "step": 6141 + }, + { + "epoch": 0.5539072011543491, + "grad_norm": 1.686256303598712, + "learning_rate": 1.7485800658236888e-06, + "loss": 0.9109, + "step": 6142 + }, + { + "epoch": 0.5539973846778193, + "grad_norm": 1.5221541256457767, + "learning_rate": 1.7480005004517232e-06, + "loss": 1.0567, + "step": 6143 + }, + { + "epoch": 0.5540875682012896, + "grad_norm": 1.8159663498319356, + "learning_rate": 1.7474209565817435e-06, + "loss": 1.0483, + "step": 6144 + }, + { + "epoch": 0.5541777517247599, + "grad_norm": 1.5033611944764769, + "learning_rate": 1.7468414342632014e-06, + "loss": 0.9847, + "step": 6145 + }, + { + "epoch": 0.5542679352482301, + "grad_norm": 1.8586178047255892, + "learning_rate": 1.746261933545543e-06, + "loss": 1.0378, + "step": 6146 + }, + { + "epoch": 0.5543581187717004, + "grad_norm": 1.406003336996077, + "learning_rate": 1.7456824544782165e-06, + "loss": 0.9852, + "step": 6147 + }, + { + "epoch": 0.5544483022951707, + "grad_norm": 1.4665142655665608, + "learning_rate": 1.7451029971106653e-06, + "loss": 0.9896, + "step": 6148 + }, + { + "epoch": 0.554538485818641, + "grad_norm": 1.3980820273291141, + "learning_rate": 1.7445235614923313e-06, + "loss": 1.0152, + "step": 6149 + }, + { + "epoch": 0.5546286693421112, + "grad_norm": 1.4560325001715402, + "learning_rate": 1.7439441476726556e-06, + "loss": 0.9754, + "step": 6150 + }, + { + "epoch": 0.5547188528655814, + "grad_norm": 1.7179685723624283, + "learning_rate": 1.7433647557010776e-06, + "loss": 0.9983, + "step": 6151 + }, + { + "epoch": 0.5548090363890518, + "grad_norm": 1.9876342800051612, + "learning_rate": 1.7427853856270338e-06, + "loss": 0.9529, + "step": 6152 + }, + { + "epoch": 0.554899219912522, + "grad_norm": 1.9841551771582957, + "learning_rate": 1.7422060374999593e-06, + "loss": 0.9533, + "step": 6153 + }, + { + "epoch": 0.5549894034359922, + "grad_norm": 1.430165778867805, + "learning_rate": 1.7416267113692862e-06, + "loss": 0.9466, + "step": 6154 + }, + { + "epoch": 0.5550795869594625, + "grad_norm": 1.4573036736881475, + "learning_rate": 1.7410474072844475e-06, + "loss": 1.0206, + "step": 6155 + }, + { + "epoch": 0.5551697704829328, + "grad_norm": 1.4447427812413443, + "learning_rate": 1.740468125294871e-06, + "loss": 0.9717, + "step": 6156 + }, + { + "epoch": 0.555259954006403, + "grad_norm": 1.9027092365193274, + "learning_rate": 1.739888865449986e-06, + "loss": 0.9011, + "step": 6157 + }, + { + "epoch": 0.5553501375298733, + "grad_norm": 1.7068789346510613, + "learning_rate": 1.7393096277992174e-06, + "loss": 1.0132, + "step": 6158 + }, + { + "epoch": 0.5554403210533435, + "grad_norm": 1.566864859780609, + "learning_rate": 1.738730412391988e-06, + "loss": 0.9999, + "step": 6159 + }, + { + "epoch": 0.5555305045768139, + "grad_norm": 1.9263653904182152, + "learning_rate": 1.738151219277721e-06, + "loss": 1.0379, + "step": 6160 + }, + { + "epoch": 0.5556206881002841, + "grad_norm": 1.8098394664156146, + "learning_rate": 1.7375720485058349e-06, + "loss": 0.9481, + "step": 6161 + }, + { + "epoch": 0.5557108716237543, + "grad_norm": 1.8504394473270065, + "learning_rate": 1.7369929001257498e-06, + "loss": 1.0018, + "step": 6162 + }, + { + "epoch": 0.5558010551472246, + "grad_norm": 3.2410457209228483, + "learning_rate": 1.73641377418688e-06, + "loss": 0.9041, + "step": 6163 + }, + { + "epoch": 0.5558912386706949, + "grad_norm": 1.480810845540093, + "learning_rate": 1.7358346707386408e-06, + "loss": 0.92, + "step": 6164 + }, + { + "epoch": 0.5559814221941651, + "grad_norm": 1.6606985126572542, + "learning_rate": 1.7352555898304439e-06, + "loss": 1.0355, + "step": 6165 + }, + { + "epoch": 0.5560716057176354, + "grad_norm": 1.7192205305193409, + "learning_rate": 1.7346765315116996e-06, + "loss": 0.9127, + "step": 6166 + }, + { + "epoch": 0.5561617892411056, + "grad_norm": 1.5659495613091927, + "learning_rate": 1.734097495831817e-06, + "loss": 0.9498, + "step": 6167 + }, + { + "epoch": 0.5562519727645759, + "grad_norm": 2.0634028452727993, + "learning_rate": 1.7335184828402015e-06, + "loss": 1.0012, + "step": 6168 + }, + { + "epoch": 0.5563421562880462, + "grad_norm": 1.4291897934061004, + "learning_rate": 1.7329394925862595e-06, + "loss": 0.9638, + "step": 6169 + }, + { + "epoch": 0.5564323398115164, + "grad_norm": 1.8873521203727817, + "learning_rate": 1.7323605251193922e-06, + "loss": 1.0323, + "step": 6170 + }, + { + "epoch": 0.5565225233349868, + "grad_norm": 1.7773814438396496, + "learning_rate": 1.7317815804890001e-06, + "loss": 1.0025, + "step": 6171 + }, + { + "epoch": 0.556612706858457, + "grad_norm": 2.544241169366452, + "learning_rate": 1.731202658744483e-06, + "loss": 0.9398, + "step": 6172 + }, + { + "epoch": 0.5567028903819272, + "grad_norm": 1.4607836800445997, + "learning_rate": 1.7306237599352365e-06, + "loss": 1.049, + "step": 6173 + }, + { + "epoch": 0.5567930739053975, + "grad_norm": 1.3532382080787773, + "learning_rate": 1.730044884110657e-06, + "loss": 0.9719, + "step": 6174 + }, + { + "epoch": 0.5568832574288678, + "grad_norm": 1.7061998800105735, + "learning_rate": 1.7294660313201366e-06, + "loss": 0.9435, + "step": 6175 + }, + { + "epoch": 0.556973440952338, + "grad_norm": 3.0105044518225443, + "learning_rate": 1.7288872016130652e-06, + "loss": 0.9695, + "step": 6176 + }, + { + "epoch": 0.5570636244758083, + "grad_norm": 1.7931634136564831, + "learning_rate": 1.7283083950388334e-06, + "loss": 0.8992, + "step": 6177 + }, + { + "epoch": 0.5571538079992785, + "grad_norm": 1.4262170413034683, + "learning_rate": 1.727729611646827e-06, + "loss": 0.8905, + "step": 6178 + }, + { + "epoch": 0.5572439915227488, + "grad_norm": 3.193041304400312, + "learning_rate": 1.7271508514864318e-06, + "loss": 1.0415, + "step": 6179 + }, + { + "epoch": 0.5573341750462191, + "grad_norm": 1.9632300240041891, + "learning_rate": 1.7265721146070302e-06, + "loss": 0.9871, + "step": 6180 + }, + { + "epoch": 0.5574243585696893, + "grad_norm": 1.6349851034809877, + "learning_rate": 1.7259934010580035e-06, + "loss": 0.9833, + "step": 6181 + }, + { + "epoch": 0.5575145420931595, + "grad_norm": 1.581811583539102, + "learning_rate": 1.725414710888731e-06, + "loss": 0.9361, + "step": 6182 + }, + { + "epoch": 0.5576047256166299, + "grad_norm": 1.5092377720720238, + "learning_rate": 1.7248360441485885e-06, + "loss": 0.9634, + "step": 6183 + }, + { + "epoch": 0.5576949091401001, + "grad_norm": 2.4644362994006785, + "learning_rate": 1.7242574008869528e-06, + "loss": 1.0603, + "step": 6184 + }, + { + "epoch": 0.5577850926635703, + "grad_norm": 0.675566666772962, + "learning_rate": 1.7236787811531951e-06, + "loss": 0.8661, + "step": 6185 + }, + { + "epoch": 0.5578752761870406, + "grad_norm": 1.673461304111524, + "learning_rate": 1.7231001849966887e-06, + "loss": 0.9539, + "step": 6186 + }, + { + "epoch": 0.5579654597105109, + "grad_norm": 0.6486196606780739, + "learning_rate": 1.722521612466801e-06, + "loss": 0.8442, + "step": 6187 + }, + { + "epoch": 0.5580556432339812, + "grad_norm": 1.96674729467516, + "learning_rate": 1.7219430636128984e-06, + "loss": 0.9304, + "step": 6188 + }, + { + "epoch": 0.5581458267574514, + "grad_norm": 1.5844366178249418, + "learning_rate": 1.7213645384843479e-06, + "loss": 0.9291, + "step": 6189 + }, + { + "epoch": 0.5582360102809216, + "grad_norm": 1.736365542099827, + "learning_rate": 1.7207860371305108e-06, + "loss": 0.9783, + "step": 6190 + }, + { + "epoch": 0.558326193804392, + "grad_norm": 1.551405696878403, + "learning_rate": 1.7202075596007487e-06, + "loss": 0.9522, + "step": 6191 + }, + { + "epoch": 0.5584163773278622, + "grad_norm": 1.779004940868805, + "learning_rate": 1.7196291059444206e-06, + "loss": 0.9465, + "step": 6192 + }, + { + "epoch": 0.5585065608513324, + "grad_norm": 2.840579434775334, + "learning_rate": 1.7190506762108828e-06, + "loss": 0.9709, + "step": 6193 + }, + { + "epoch": 0.5585967443748028, + "grad_norm": 1.7817593565698262, + "learning_rate": 1.7184722704494907e-06, + "loss": 0.9879, + "step": 6194 + }, + { + "epoch": 0.558686927898273, + "grad_norm": 1.5316727502319485, + "learning_rate": 1.717893888709596e-06, + "loss": 0.9987, + "step": 6195 + }, + { + "epoch": 0.5587771114217432, + "grad_norm": 1.9865623493563571, + "learning_rate": 1.7173155310405515e-06, + "loss": 1.0119, + "step": 6196 + }, + { + "epoch": 0.5588672949452135, + "grad_norm": 1.406591882472953, + "learning_rate": 1.7167371974917043e-06, + "loss": 1.0627, + "step": 6197 + }, + { + "epoch": 0.5589574784686838, + "grad_norm": 2.0153319862950947, + "learning_rate": 1.7161588881124003e-06, + "loss": 1.0009, + "step": 6198 + }, + { + "epoch": 0.559047661992154, + "grad_norm": 1.8556785705789427, + "learning_rate": 1.7155806029519861e-06, + "loss": 0.9708, + "step": 6199 + }, + { + "epoch": 0.5591378455156243, + "grad_norm": 1.68696020464493, + "learning_rate": 1.7150023420598023e-06, + "loss": 1.0275, + "step": 6200 + }, + { + "epoch": 0.5592280290390945, + "grad_norm": 1.406416056683249, + "learning_rate": 1.714424105485191e-06, + "loss": 1.0375, + "step": 6201 + }, + { + "epoch": 0.5593182125625649, + "grad_norm": 2.020650586651478, + "learning_rate": 1.7138458932774896e-06, + "loss": 0.949, + "step": 6202 + }, + { + "epoch": 0.5594083960860351, + "grad_norm": 0.6568461616711767, + "learning_rate": 1.7132677054860335e-06, + "loss": 0.8444, + "step": 6203 + }, + { + "epoch": 0.5594985796095053, + "grad_norm": 1.718620156672184, + "learning_rate": 1.7126895421601586e-06, + "loss": 1.0089, + "step": 6204 + }, + { + "epoch": 0.5595887631329756, + "grad_norm": 1.6723563668880785, + "learning_rate": 1.712111403349196e-06, + "loss": 0.9278, + "step": 6205 + }, + { + "epoch": 0.5596789466564459, + "grad_norm": 1.4858013327349844, + "learning_rate": 1.7115332891024757e-06, + "loss": 1.0173, + "step": 6206 + }, + { + "epoch": 0.5597691301799161, + "grad_norm": 0.7135092781308385, + "learning_rate": 1.7109551994693257e-06, + "loss": 0.8937, + "step": 6207 + }, + { + "epoch": 0.5598593137033864, + "grad_norm": 2.0391125435370068, + "learning_rate": 1.7103771344990725e-06, + "loss": 0.9255, + "step": 6208 + }, + { + "epoch": 0.5599494972268566, + "grad_norm": 2.5169260202962436, + "learning_rate": 1.709799094241039e-06, + "loss": 0.8995, + "step": 6209 + }, + { + "epoch": 0.560039680750327, + "grad_norm": 1.7583718321263164, + "learning_rate": 1.709221078744546e-06, + "loss": 1.0186, + "step": 6210 + }, + { + "epoch": 0.5601298642737972, + "grad_norm": 1.3722786850366155, + "learning_rate": 1.7086430880589148e-06, + "loss": 0.9962, + "step": 6211 + }, + { + "epoch": 0.5602200477972674, + "grad_norm": 1.7616639393063849, + "learning_rate": 1.7080651222334612e-06, + "loss": 0.9877, + "step": 6212 + }, + { + "epoch": 0.5603102313207377, + "grad_norm": 1.8953402021102252, + "learning_rate": 1.7074871813175018e-06, + "loss": 1.0447, + "step": 6213 + }, + { + "epoch": 0.560400414844208, + "grad_norm": 1.524159343584397, + "learning_rate": 1.706909265360349e-06, + "loss": 0.8995, + "step": 6214 + }, + { + "epoch": 0.5604905983676782, + "grad_norm": 0.6950008894207979, + "learning_rate": 1.7063313744113128e-06, + "loss": 0.8336, + "step": 6215 + }, + { + "epoch": 0.5605807818911485, + "grad_norm": 0.6518474221179384, + "learning_rate": 1.7057535085197042e-06, + "loss": 0.8066, + "step": 6216 + }, + { + "epoch": 0.5606709654146187, + "grad_norm": 2.0774445531173282, + "learning_rate": 1.705175667734828e-06, + "loss": 0.9044, + "step": 6217 + }, + { + "epoch": 0.560761148938089, + "grad_norm": 2.0838141331588895, + "learning_rate": 1.7045978521059894e-06, + "loss": 0.9591, + "step": 6218 + }, + { + "epoch": 0.5608513324615593, + "grad_norm": 2.111310283010297, + "learning_rate": 1.7040200616824914e-06, + "loss": 0.9808, + "step": 6219 + }, + { + "epoch": 0.5609415159850295, + "grad_norm": 3.855931525781877, + "learning_rate": 1.7034422965136333e-06, + "loss": 0.9277, + "step": 6220 + }, + { + "epoch": 0.5610316995084998, + "grad_norm": 1.570795165322569, + "learning_rate": 1.7028645566487137e-06, + "loss": 1.0278, + "step": 6221 + }, + { + "epoch": 0.5611218830319701, + "grad_norm": 7.727867517682327, + "learning_rate": 1.7022868421370284e-06, + "loss": 0.9613, + "step": 6222 + }, + { + "epoch": 0.5612120665554403, + "grad_norm": 1.7505098690047864, + "learning_rate": 1.701709153027872e-06, + "loss": 1.0286, + "step": 6223 + }, + { + "epoch": 0.5613022500789105, + "grad_norm": 2.4196589021447363, + "learning_rate": 1.7011314893705353e-06, + "loss": 0.9503, + "step": 6224 + }, + { + "epoch": 0.5613924336023809, + "grad_norm": 1.4987037143766369, + "learning_rate": 1.700553851214307e-06, + "loss": 0.9888, + "step": 6225 + }, + { + "epoch": 0.5614826171258511, + "grad_norm": 2.0541172178969824, + "learning_rate": 1.699976238608476e-06, + "loss": 0.8504, + "step": 6226 + }, + { + "epoch": 0.5615728006493214, + "grad_norm": 1.7599694475859964, + "learning_rate": 1.699398651602326e-06, + "loss": 0.9313, + "step": 6227 + }, + { + "epoch": 0.5616629841727916, + "grad_norm": 1.4204005047315187, + "learning_rate": 1.6988210902451413e-06, + "loss": 1.0285, + "step": 6228 + }, + { + "epoch": 0.5617531676962619, + "grad_norm": 1.7900080951315827, + "learning_rate": 1.6982435545862011e-06, + "loss": 0.9639, + "step": 6229 + }, + { + "epoch": 0.5618433512197322, + "grad_norm": 1.7717268166987945, + "learning_rate": 1.6976660446747853e-06, + "loss": 0.9904, + "step": 6230 + }, + { + "epoch": 0.5619335347432024, + "grad_norm": 2.076711417686004, + "learning_rate": 1.6970885605601696e-06, + "loss": 0.9087, + "step": 6231 + }, + { + "epoch": 0.5620237182666726, + "grad_norm": 2.3621070656150858, + "learning_rate": 1.6965111022916282e-06, + "loss": 1.0907, + "step": 6232 + }, + { + "epoch": 0.562113901790143, + "grad_norm": 1.6127192000802542, + "learning_rate": 1.6959336699184323e-06, + "loss": 0.9585, + "step": 6233 + }, + { + "epoch": 0.5622040853136132, + "grad_norm": 1.3956149722814017, + "learning_rate": 1.6953562634898529e-06, + "loss": 0.9769, + "step": 6234 + }, + { + "epoch": 0.5622942688370834, + "grad_norm": 3.3793619014033904, + "learning_rate": 1.6947788830551569e-06, + "loss": 0.9538, + "step": 6235 + }, + { + "epoch": 0.5623844523605537, + "grad_norm": 0.7730273691022815, + "learning_rate": 1.6942015286636093e-06, + "loss": 0.8666, + "step": 6236 + }, + { + "epoch": 0.562474635884024, + "grad_norm": 1.7272594305375377, + "learning_rate": 1.6936242003644735e-06, + "loss": 0.9697, + "step": 6237 + }, + { + "epoch": 0.5625648194074943, + "grad_norm": 1.634036759629562, + "learning_rate": 1.6930468982070106e-06, + "loss": 1.0556, + "step": 6238 + }, + { + "epoch": 0.5626550029309645, + "grad_norm": 0.5974538871895356, + "learning_rate": 1.692469622240478e-06, + "loss": 0.7555, + "step": 6239 + }, + { + "epoch": 0.5627451864544347, + "grad_norm": 1.4690011803951415, + "learning_rate": 1.6918923725141339e-06, + "loss": 0.9402, + "step": 6240 + }, + { + "epoch": 0.5628353699779051, + "grad_norm": 1.4116544927382129, + "learning_rate": 1.6913151490772312e-06, + "loss": 1.0334, + "step": 6241 + }, + { + "epoch": 0.5629255535013753, + "grad_norm": 1.4344054374706652, + "learning_rate": 1.6907379519790215e-06, + "loss": 1.025, + "step": 6242 + }, + { + "epoch": 0.5630157370248455, + "grad_norm": 1.3866214771808885, + "learning_rate": 1.6901607812687558e-06, + "loss": 0.8917, + "step": 6243 + }, + { + "epoch": 0.5631059205483159, + "grad_norm": 1.7828951148667844, + "learning_rate": 1.6895836369956794e-06, + "loss": 0.9387, + "step": 6244 + }, + { + "epoch": 0.5631961040717861, + "grad_norm": 0.6621322627880014, + "learning_rate": 1.6890065192090402e-06, + "loss": 0.7794, + "step": 6245 + }, + { + "epoch": 0.5632862875952563, + "grad_norm": 1.981749051175276, + "learning_rate": 1.6884294279580793e-06, + "loss": 0.8825, + "step": 6246 + }, + { + "epoch": 0.5633764711187266, + "grad_norm": 1.918812366180301, + "learning_rate": 1.6878523632920371e-06, + "loss": 1.0067, + "step": 6247 + }, + { + "epoch": 0.5634666546421969, + "grad_norm": 1.6314229939250648, + "learning_rate": 1.6872753252601525e-06, + "loss": 1.0454, + "step": 6248 + }, + { + "epoch": 0.5635568381656672, + "grad_norm": 2.1733612902818766, + "learning_rate": 1.6866983139116616e-06, + "loss": 0.9245, + "step": 6249 + }, + { + "epoch": 0.5636470216891374, + "grad_norm": 2.4919380852430533, + "learning_rate": 1.6861213292957981e-06, + "loss": 0.9547, + "step": 6250 + }, + { + "epoch": 0.5637372052126076, + "grad_norm": 3.4344735434281355, + "learning_rate": 1.685544371461793e-06, + "loss": 1.0135, + "step": 6251 + }, + { + "epoch": 0.563827388736078, + "grad_norm": 1.7101818488842209, + "learning_rate": 1.6849674404588767e-06, + "loss": 1.0129, + "step": 6252 + }, + { + "epoch": 0.5639175722595482, + "grad_norm": 2.2099108789348207, + "learning_rate": 1.6843905363362758e-06, + "loss": 1.0152, + "step": 6253 + }, + { + "epoch": 0.5640077557830184, + "grad_norm": 1.7702006238180106, + "learning_rate": 1.6838136591432136e-06, + "loss": 0.9247, + "step": 6254 + }, + { + "epoch": 0.5640979393064887, + "grad_norm": 2.0580425067289485, + "learning_rate": 1.6832368089289139e-06, + "loss": 1.0069, + "step": 6255 + }, + { + "epoch": 0.564188122829959, + "grad_norm": 1.48528663347877, + "learning_rate": 1.682659985742596e-06, + "loss": 1.0009, + "step": 6256 + }, + { + "epoch": 0.5642783063534292, + "grad_norm": 1.5882040279921972, + "learning_rate": 1.6820831896334782e-06, + "loss": 1.0242, + "step": 6257 + }, + { + "epoch": 0.5643684898768995, + "grad_norm": 1.6709547986458018, + "learning_rate": 1.681506420650776e-06, + "loss": 1.0041, + "step": 6258 + }, + { + "epoch": 0.5644586734003697, + "grad_norm": 1.3738731188181181, + "learning_rate": 1.680929678843701e-06, + "loss": 0.9697, + "step": 6259 + }, + { + "epoch": 0.56454885692384, + "grad_norm": 1.5568273951581206, + "learning_rate": 1.6803529642614662e-06, + "loss": 1.0289, + "step": 6260 + }, + { + "epoch": 0.5646390404473103, + "grad_norm": 1.6358831270904617, + "learning_rate": 1.6797762769532785e-06, + "loss": 0.9869, + "step": 6261 + }, + { + "epoch": 0.5647292239707805, + "grad_norm": 1.4187909729540547, + "learning_rate": 1.679199616968345e-06, + "loss": 1.0175, + "step": 6262 + }, + { + "epoch": 0.5648194074942507, + "grad_norm": 1.6782943270764872, + "learning_rate": 1.6786229843558689e-06, + "loss": 1.0127, + "step": 6263 + }, + { + "epoch": 0.5649095910177211, + "grad_norm": 2.1306133180646603, + "learning_rate": 1.6780463791650514e-06, + "loss": 0.9629, + "step": 6264 + }, + { + "epoch": 0.5649997745411913, + "grad_norm": 1.8281722103446025, + "learning_rate": 1.6774698014450928e-06, + "loss": 0.9257, + "step": 6265 + }, + { + "epoch": 0.5650899580646616, + "grad_norm": 0.6492177795696695, + "learning_rate": 1.6768932512451883e-06, + "loss": 0.8152, + "step": 6266 + }, + { + "epoch": 0.5651801415881319, + "grad_norm": 1.4362789441305555, + "learning_rate": 1.676316728614534e-06, + "loss": 0.9329, + "step": 6267 + }, + { + "epoch": 0.5652703251116021, + "grad_norm": 0.5832635264408095, + "learning_rate": 1.675740233602321e-06, + "loss": 0.825, + "step": 6268 + }, + { + "epoch": 0.5653605086350724, + "grad_norm": 1.7326208643293193, + "learning_rate": 1.6751637662577385e-06, + "loss": 1.0322, + "step": 6269 + }, + { + "epoch": 0.5654506921585426, + "grad_norm": 0.8103411309965374, + "learning_rate": 1.6745873266299753e-06, + "loss": 0.7694, + "step": 6270 + }, + { + "epoch": 0.565540875682013, + "grad_norm": 1.4382290028246685, + "learning_rate": 1.6740109147682148e-06, + "loss": 0.9443, + "step": 6271 + }, + { + "epoch": 0.5656310592054832, + "grad_norm": 1.9204485181618471, + "learning_rate": 1.6734345307216418e-06, + "loss": 0.9973, + "step": 6272 + }, + { + "epoch": 0.5657212427289534, + "grad_norm": 1.9988031382412614, + "learning_rate": 1.6728581745394346e-06, + "loss": 0.9547, + "step": 6273 + }, + { + "epoch": 0.5658114262524236, + "grad_norm": 1.6207024894928255, + "learning_rate": 1.672281846270772e-06, + "loss": 1.0568, + "step": 6274 + }, + { + "epoch": 0.565901609775894, + "grad_norm": 2.454007907086779, + "learning_rate": 1.6717055459648295e-06, + "loss": 0.9772, + "step": 6275 + }, + { + "epoch": 0.5659917932993642, + "grad_norm": 1.911042676699649, + "learning_rate": 1.6711292736707793e-06, + "loss": 0.9321, + "step": 6276 + }, + { + "epoch": 0.5660819768228345, + "grad_norm": 2.0679347941597985, + "learning_rate": 1.6705530294377938e-06, + "loss": 1.0505, + "step": 6277 + }, + { + "epoch": 0.5661721603463047, + "grad_norm": 1.5680150168244755, + "learning_rate": 1.6699768133150395e-06, + "loss": 0.9743, + "step": 6278 + }, + { + "epoch": 0.566262343869775, + "grad_norm": 1.8772864070648332, + "learning_rate": 1.6694006253516837e-06, + "loss": 0.9724, + "step": 6279 + }, + { + "epoch": 0.5663525273932453, + "grad_norm": 2.005700570332479, + "learning_rate": 1.6688244655968896e-06, + "loss": 0.9803, + "step": 6280 + }, + { + "epoch": 0.5664427109167155, + "grad_norm": 3.128389122908705, + "learning_rate": 1.6682483340998175e-06, + "loss": 0.9016, + "step": 6281 + }, + { + "epoch": 0.5665328944401857, + "grad_norm": 2.601429428958729, + "learning_rate": 1.6676722309096276e-06, + "loss": 0.9425, + "step": 6282 + }, + { + "epoch": 0.5666230779636561, + "grad_norm": 0.7329833459874754, + "learning_rate": 1.6670961560754744e-06, + "loss": 0.8841, + "step": 6283 + }, + { + "epoch": 0.5667132614871263, + "grad_norm": 1.627476126148544, + "learning_rate": 1.6665201096465138e-06, + "loss": 0.9773, + "step": 6284 + }, + { + "epoch": 0.5668034450105965, + "grad_norm": 1.7810818107573605, + "learning_rate": 1.6659440916718961e-06, + "loss": 1.0256, + "step": 6285 + }, + { + "epoch": 0.5668936285340668, + "grad_norm": 6.644460059877303, + "learning_rate": 1.6653681022007696e-06, + "loss": 0.9976, + "step": 6286 + }, + { + "epoch": 0.5669838120575371, + "grad_norm": 1.6932177280902885, + "learning_rate": 1.6647921412822825e-06, + "loss": 0.914, + "step": 6287 + }, + { + "epoch": 0.5670739955810074, + "grad_norm": 1.57989727253686, + "learning_rate": 1.6642162089655782e-06, + "loss": 0.9829, + "step": 6288 + }, + { + "epoch": 0.5671641791044776, + "grad_norm": 2.1332380268543707, + "learning_rate": 1.663640305299798e-06, + "loss": 0.9488, + "step": 6289 + }, + { + "epoch": 0.5672543626279479, + "grad_norm": 0.6628220135457011, + "learning_rate": 1.6630644303340824e-06, + "loss": 0.8462, + "step": 6290 + }, + { + "epoch": 0.5673445461514182, + "grad_norm": 1.732493299110742, + "learning_rate": 1.662488584117567e-06, + "loss": 0.99, + "step": 6291 + }, + { + "epoch": 0.5674347296748884, + "grad_norm": 1.494036103625601, + "learning_rate": 1.6619127666993867e-06, + "loss": 1.0403, + "step": 6292 + }, + { + "epoch": 0.5675249131983586, + "grad_norm": 2.0336659076095054, + "learning_rate": 1.6613369781286727e-06, + "loss": 0.9842, + "step": 6293 + }, + { + "epoch": 0.567615096721829, + "grad_norm": 1.7181626443357296, + "learning_rate": 1.6607612184545562e-06, + "loss": 0.9179, + "step": 6294 + }, + { + "epoch": 0.5677052802452992, + "grad_norm": 1.6958459814601303, + "learning_rate": 1.6601854877261617e-06, + "loss": 1.051, + "step": 6295 + }, + { + "epoch": 0.5677954637687694, + "grad_norm": 1.8599683311801747, + "learning_rate": 1.6596097859926163e-06, + "loss": 0.9781, + "step": 6296 + }, + { + "epoch": 0.5678856472922397, + "grad_norm": 2.043666625582503, + "learning_rate": 1.6590341133030407e-06, + "loss": 1.0702, + "step": 6297 + }, + { + "epoch": 0.56797583081571, + "grad_norm": 1.7957523985619281, + "learning_rate": 1.658458469706554e-06, + "loss": 1.0548, + "step": 6298 + }, + { + "epoch": 0.5680660143391802, + "grad_norm": 1.4186340956880779, + "learning_rate": 1.6578828552522746e-06, + "loss": 0.9383, + "step": 6299 + }, + { + "epoch": 0.5681561978626505, + "grad_norm": 2.445067859636276, + "learning_rate": 1.6573072699893156e-06, + "loss": 1.0269, + "step": 6300 + }, + { + "epoch": 0.5682463813861207, + "grad_norm": 2.841213897999334, + "learning_rate": 1.6567317139667906e-06, + "loss": 0.956, + "step": 6301 + }, + { + "epoch": 0.5683365649095911, + "grad_norm": 1.8400948063640854, + "learning_rate": 1.6561561872338087e-06, + "loss": 0.9974, + "step": 6302 + }, + { + "epoch": 0.5684267484330613, + "grad_norm": 1.6350733240700397, + "learning_rate": 1.6555806898394764e-06, + "loss": 0.8939, + "step": 6303 + }, + { + "epoch": 0.5685169319565315, + "grad_norm": 1.8987146497076395, + "learning_rate": 1.6550052218328987e-06, + "loss": 0.9593, + "step": 6304 + }, + { + "epoch": 0.5686071154800018, + "grad_norm": 1.8244709763564053, + "learning_rate": 1.6544297832631777e-06, + "loss": 0.9549, + "step": 6305 + }, + { + "epoch": 0.5686972990034721, + "grad_norm": 2.2479299983871663, + "learning_rate": 1.6538543741794135e-06, + "loss": 0.9847, + "step": 6306 + }, + { + "epoch": 0.5687874825269423, + "grad_norm": 1.6166411873032023, + "learning_rate": 1.6532789946307028e-06, + "loss": 0.9504, + "step": 6307 + }, + { + "epoch": 0.5688776660504126, + "grad_norm": 0.6419389322565929, + "learning_rate": 1.6527036446661393e-06, + "loss": 0.8497, + "step": 6308 + }, + { + "epoch": 0.5689678495738828, + "grad_norm": 2.0709057227926415, + "learning_rate": 1.6521283243348165e-06, + "loss": 1.0146, + "step": 6309 + }, + { + "epoch": 0.5690580330973531, + "grad_norm": 1.5270307371372278, + "learning_rate": 1.6515530336858227e-06, + "loss": 0.9694, + "step": 6310 + }, + { + "epoch": 0.5691482166208234, + "grad_norm": 1.8082432582354564, + "learning_rate": 1.6509777727682457e-06, + "loss": 1.0299, + "step": 6311 + }, + { + "epoch": 0.5692384001442936, + "grad_norm": 2.8614959355121643, + "learning_rate": 1.65040254163117e-06, + "loss": 1.0212, + "step": 6312 + }, + { + "epoch": 0.569328583667764, + "grad_norm": 2.4794237233310676, + "learning_rate": 1.649827340323676e-06, + "loss": 0.8925, + "step": 6313 + }, + { + "epoch": 0.5694187671912342, + "grad_norm": 1.7023480244163152, + "learning_rate": 1.6492521688948454e-06, + "loss": 0.9477, + "step": 6314 + }, + { + "epoch": 0.5695089507147044, + "grad_norm": 1.717703205730989, + "learning_rate": 1.6486770273937526e-06, + "loss": 1.0047, + "step": 6315 + }, + { + "epoch": 0.5695991342381747, + "grad_norm": 2.0384907706535724, + "learning_rate": 1.6481019158694738e-06, + "loss": 0.959, + "step": 6316 + }, + { + "epoch": 0.569689317761645, + "grad_norm": 1.0233615526746527, + "learning_rate": 1.6475268343710792e-06, + "loss": 0.8894, + "step": 6317 + }, + { + "epoch": 0.5697795012851152, + "grad_norm": 1.6335610312090996, + "learning_rate": 1.6469517829476396e-06, + "loss": 0.9134, + "step": 6318 + }, + { + "epoch": 0.5698696848085855, + "grad_norm": 1.838120351900296, + "learning_rate": 1.64637676164822e-06, + "loss": 0.9623, + "step": 6319 + }, + { + "epoch": 0.5699598683320557, + "grad_norm": 1.6480053520532383, + "learning_rate": 1.6458017705218848e-06, + "loss": 1.0185, + "step": 6320 + }, + { + "epoch": 0.570050051855526, + "grad_norm": 2.901968971586075, + "learning_rate": 1.645226809617696e-06, + "loss": 1.0076, + "step": 6321 + }, + { + "epoch": 0.5701402353789963, + "grad_norm": 1.7973307156067377, + "learning_rate": 1.6446518789847112e-06, + "loss": 1.0208, + "step": 6322 + }, + { + "epoch": 0.5702304189024665, + "grad_norm": 2.4972351043712138, + "learning_rate": 1.6440769786719883e-06, + "loss": 1.0078, + "step": 6323 + }, + { + "epoch": 0.5703206024259367, + "grad_norm": 1.3795770412997854, + "learning_rate": 1.6435021087285803e-06, + "loss": 1.0207, + "step": 6324 + }, + { + "epoch": 0.5704107859494071, + "grad_norm": 1.7820551207138149, + "learning_rate": 1.642927269203537e-06, + "loss": 1.0174, + "step": 6325 + }, + { + "epoch": 0.5705009694728773, + "grad_norm": 1.5271655512903548, + "learning_rate": 1.642352460145909e-06, + "loss": 1.0328, + "step": 6326 + }, + { + "epoch": 0.5705911529963476, + "grad_norm": 1.5388709399297045, + "learning_rate": 1.6417776816047402e-06, + "loss": 0.9872, + "step": 6327 + }, + { + "epoch": 0.5706813365198178, + "grad_norm": 1.5375353987425704, + "learning_rate": 1.6412029336290755e-06, + "loss": 0.9714, + "step": 6328 + }, + { + "epoch": 0.5707715200432881, + "grad_norm": 2.0259893520109307, + "learning_rate": 1.6406282162679551e-06, + "loss": 0.984, + "step": 6329 + }, + { + "epoch": 0.5708617035667584, + "grad_norm": 1.3757129901370015, + "learning_rate": 1.6400535295704162e-06, + "loss": 0.9522, + "step": 6330 + }, + { + "epoch": 0.5709518870902286, + "grad_norm": 1.8918980497550903, + "learning_rate": 1.6394788735854955e-06, + "loss": 0.9216, + "step": 6331 + }, + { + "epoch": 0.5710420706136988, + "grad_norm": 1.660232937107484, + "learning_rate": 1.6389042483622246e-06, + "loss": 1.0489, + "step": 6332 + }, + { + "epoch": 0.5711322541371692, + "grad_norm": 1.6961829416376895, + "learning_rate": 1.638329653949635e-06, + "loss": 0.9766, + "step": 6333 + }, + { + "epoch": 0.5712224376606394, + "grad_norm": 1.579261219679867, + "learning_rate": 1.637755090396753e-06, + "loss": 0.9475, + "step": 6334 + }, + { + "epoch": 0.5713126211841096, + "grad_norm": 0.640393145403731, + "learning_rate": 1.6371805577526039e-06, + "loss": 0.8478, + "step": 6335 + }, + { + "epoch": 0.5714028047075799, + "grad_norm": 1.6250465092972417, + "learning_rate": 1.636606056066211e-06, + "loss": 0.9324, + "step": 6336 + }, + { + "epoch": 0.5714929882310502, + "grad_norm": 2.052967584576121, + "learning_rate": 1.636031585386592e-06, + "loss": 0.9123, + "step": 6337 + }, + { + "epoch": 0.5715831717545204, + "grad_norm": 3.7223096207816817, + "learning_rate": 1.635457145762766e-06, + "loss": 1.0443, + "step": 6338 + }, + { + "epoch": 0.5716733552779907, + "grad_norm": 2.1528496952418053, + "learning_rate": 1.6348827372437456e-06, + "loss": 1.0091, + "step": 6339 + }, + { + "epoch": 0.571763538801461, + "grad_norm": 1.4422722259464655, + "learning_rate": 1.634308359878544e-06, + "loss": 0.9862, + "step": 6340 + }, + { + "epoch": 0.5718537223249313, + "grad_norm": 1.5602959155365104, + "learning_rate": 1.6337340137161695e-06, + "loss": 1.0625, + "step": 6341 + }, + { + "epoch": 0.5719439058484015, + "grad_norm": 1.693251099254252, + "learning_rate": 1.6331596988056277e-06, + "loss": 0.9857, + "step": 6342 + }, + { + "epoch": 0.5720340893718717, + "grad_norm": 1.4214294909401006, + "learning_rate": 1.632585415195924e-06, + "loss": 1.0096, + "step": 6343 + }, + { + "epoch": 0.5721242728953421, + "grad_norm": 1.5508883268734248, + "learning_rate": 1.6320111629360583e-06, + "loss": 1.0143, + "step": 6344 + }, + { + "epoch": 0.5722144564188123, + "grad_norm": 1.9489508618677633, + "learning_rate": 1.631436942075029e-06, + "loss": 0.9605, + "step": 6345 + }, + { + "epoch": 0.5723046399422825, + "grad_norm": 1.7960372298657516, + "learning_rate": 1.630862752661833e-06, + "loss": 0.9604, + "step": 6346 + }, + { + "epoch": 0.5723948234657528, + "grad_norm": 3.940793068165787, + "learning_rate": 1.6302885947454612e-06, + "loss": 0.9143, + "step": 6347 + }, + { + "epoch": 0.5724850069892231, + "grad_norm": 1.9842717902308906, + "learning_rate": 1.6297144683749057e-06, + "loss": 1.0016, + "step": 6348 + }, + { + "epoch": 0.5725751905126933, + "grad_norm": 2.0520674677459794, + "learning_rate": 1.629140373599153e-06, + "loss": 0.991, + "step": 6349 + }, + { + "epoch": 0.5726653740361636, + "grad_norm": 1.9487058159576618, + "learning_rate": 1.628566310467189e-06, + "loss": 0.9981, + "step": 6350 + }, + { + "epoch": 0.5727555575596338, + "grad_norm": 1.7500909372952882, + "learning_rate": 1.6279922790279957e-06, + "loss": 0.9485, + "step": 6351 + }, + { + "epoch": 0.5728457410831042, + "grad_norm": 2.2809541066196677, + "learning_rate": 1.6274182793305512e-06, + "loss": 1.0, + "step": 6352 + }, + { + "epoch": 0.5729359246065744, + "grad_norm": 2.215298183856069, + "learning_rate": 1.626844311423835e-06, + "loss": 1.0148, + "step": 6353 + }, + { + "epoch": 0.5730261081300446, + "grad_norm": 1.9783484792196528, + "learning_rate": 1.6262703753568181e-06, + "loss": 1.0636, + "step": 6354 + }, + { + "epoch": 0.5731162916535149, + "grad_norm": 1.9854415670503287, + "learning_rate": 1.6256964711784747e-06, + "loss": 1.0624, + "step": 6355 + }, + { + "epoch": 0.5732064751769852, + "grad_norm": 1.6697834515225733, + "learning_rate": 1.6251225989377723e-06, + "loss": 0.9977, + "step": 6356 + }, + { + "epoch": 0.5732966587004554, + "grad_norm": 1.7057200967872732, + "learning_rate": 1.624548758683676e-06, + "loss": 0.9872, + "step": 6357 + }, + { + "epoch": 0.5733868422239257, + "grad_norm": 1.6599849118845407, + "learning_rate": 1.6239749504651505e-06, + "loss": 1.0775, + "step": 6358 + }, + { + "epoch": 0.5734770257473959, + "grad_norm": 1.538361675432568, + "learning_rate": 1.6234011743311552e-06, + "loss": 0.9635, + "step": 6359 + }, + { + "epoch": 0.5735672092708662, + "grad_norm": 1.8199290907324355, + "learning_rate": 1.6228274303306483e-06, + "loss": 0.9186, + "step": 6360 + }, + { + "epoch": 0.5736573927943365, + "grad_norm": 1.562605282097549, + "learning_rate": 1.6222537185125847e-06, + "loss": 1.07, + "step": 6361 + }, + { + "epoch": 0.5737475763178067, + "grad_norm": 2.774850834454512, + "learning_rate": 1.6216800389259172e-06, + "loss": 1.0043, + "step": 6362 + }, + { + "epoch": 0.573837759841277, + "grad_norm": 1.3054090777307559, + "learning_rate": 1.6211063916195949e-06, + "loss": 0.992, + "step": 6363 + }, + { + "epoch": 0.5739279433647473, + "grad_norm": 1.9104688267744079, + "learning_rate": 1.6205327766425633e-06, + "loss": 0.9958, + "step": 6364 + }, + { + "epoch": 0.5740181268882175, + "grad_norm": 1.6680901012007345, + "learning_rate": 1.6199591940437689e-06, + "loss": 1.0007, + "step": 6365 + }, + { + "epoch": 0.5741083104116878, + "grad_norm": 1.5065243609636725, + "learning_rate": 1.6193856438721505e-06, + "loss": 0.9463, + "step": 6366 + }, + { + "epoch": 0.5741984939351581, + "grad_norm": 2.0390491704395495, + "learning_rate": 1.6188121261766483e-06, + "loss": 0.9947, + "step": 6367 + }, + { + "epoch": 0.5742886774586283, + "grad_norm": 1.4087798780033214, + "learning_rate": 1.6182386410061976e-06, + "loss": 0.9182, + "step": 6368 + }, + { + "epoch": 0.5743788609820986, + "grad_norm": 1.765321823692449, + "learning_rate": 1.61766518840973e-06, + "loss": 0.9013, + "step": 6369 + }, + { + "epoch": 0.5744690445055688, + "grad_norm": 1.6190918844914348, + "learning_rate": 1.6170917684361779e-06, + "loss": 0.8586, + "step": 6370 + }, + { + "epoch": 0.5745592280290391, + "grad_norm": 1.9527358010659426, + "learning_rate": 1.6165183811344662e-06, + "loss": 1.0636, + "step": 6371 + }, + { + "epoch": 0.5746494115525094, + "grad_norm": 1.75361607111559, + "learning_rate": 1.6159450265535218e-06, + "loss": 0.9806, + "step": 6372 + }, + { + "epoch": 0.5747395950759796, + "grad_norm": 2.410210656737513, + "learning_rate": 1.6153717047422652e-06, + "loss": 1.0712, + "step": 6373 + }, + { + "epoch": 0.5748297785994498, + "grad_norm": 2.093037384377337, + "learning_rate": 1.6147984157496155e-06, + "loss": 0.875, + "step": 6374 + }, + { + "epoch": 0.5749199621229202, + "grad_norm": 1.5072687779577827, + "learning_rate": 1.6142251596244886e-06, + "loss": 0.9179, + "step": 6375 + }, + { + "epoch": 0.5750101456463904, + "grad_norm": 5.323536438018357, + "learning_rate": 1.6136519364157983e-06, + "loss": 1.0455, + "step": 6376 + }, + { + "epoch": 0.5751003291698606, + "grad_norm": 1.7517752498626913, + "learning_rate": 1.6130787461724555e-06, + "loss": 0.925, + "step": 6377 + }, + { + "epoch": 0.5751905126933309, + "grad_norm": 1.7346002930725266, + "learning_rate": 1.6125055889433679e-06, + "loss": 1.0628, + "step": 6378 + }, + { + "epoch": 0.5752806962168012, + "grad_norm": 1.3859899190560867, + "learning_rate": 1.6119324647774386e-06, + "loss": 1.0155, + "step": 6379 + }, + { + "epoch": 0.5753708797402715, + "grad_norm": 3.1353390368275154, + "learning_rate": 1.6113593737235724e-06, + "loss": 0.9968, + "step": 6380 + }, + { + "epoch": 0.5754610632637417, + "grad_norm": 2.196553921196369, + "learning_rate": 1.6107863158306665e-06, + "loss": 0.9388, + "step": 6381 + }, + { + "epoch": 0.5755512467872119, + "grad_norm": 1.7554853166070972, + "learning_rate": 1.610213291147619e-06, + "loss": 0.8865, + "step": 6382 + }, + { + "epoch": 0.5756414303106823, + "grad_norm": 1.401026369978017, + "learning_rate": 1.609640299723322e-06, + "loss": 1.0128, + "step": 6383 + }, + { + "epoch": 0.5757316138341525, + "grad_norm": 1.45197948301548, + "learning_rate": 1.609067341606668e-06, + "loss": 0.9693, + "step": 6384 + }, + { + "epoch": 0.5758217973576227, + "grad_norm": 1.8862401382865988, + "learning_rate": 1.6084944168465438e-06, + "loss": 0.9791, + "step": 6385 + }, + { + "epoch": 0.5759119808810931, + "grad_norm": 1.5676740049065654, + "learning_rate": 1.6079215254918339e-06, + "loss": 1.0113, + "step": 6386 + }, + { + "epoch": 0.5760021644045633, + "grad_norm": 2.184931964295143, + "learning_rate": 1.6073486675914222e-06, + "loss": 0.894, + "step": 6387 + }, + { + "epoch": 0.5760923479280335, + "grad_norm": 2.119112272995823, + "learning_rate": 1.606775843194187e-06, + "loss": 1.0135, + "step": 6388 + }, + { + "epoch": 0.5761825314515038, + "grad_norm": 2.701093935997103, + "learning_rate": 1.6062030523490053e-06, + "loss": 0.9779, + "step": 6389 + }, + { + "epoch": 0.5762727149749741, + "grad_norm": 1.5851483566215498, + "learning_rate": 1.60563029510475e-06, + "loss": 0.9117, + "step": 6390 + }, + { + "epoch": 0.5763628984984444, + "grad_norm": 0.6461793808195011, + "learning_rate": 1.6050575715102927e-06, + "loss": 0.8061, + "step": 6391 + }, + { + "epoch": 0.5764530820219146, + "grad_norm": 1.7215704923850357, + "learning_rate": 1.6044848816145014e-06, + "loss": 1.0434, + "step": 6392 + }, + { + "epoch": 0.5765432655453848, + "grad_norm": 1.8477252842464709, + "learning_rate": 1.60391222546624e-06, + "loss": 1.0169, + "step": 6393 + }, + { + "epoch": 0.5766334490688552, + "grad_norm": 1.5345246469277738, + "learning_rate": 1.6033396031143725e-06, + "loss": 0.982, + "step": 6394 + }, + { + "epoch": 0.5767236325923254, + "grad_norm": 1.9957063004025495, + "learning_rate": 1.602767014607757e-06, + "loss": 0.9918, + "step": 6395 + }, + { + "epoch": 0.5768138161157956, + "grad_norm": 1.5367787258776655, + "learning_rate": 1.6021944599952493e-06, + "loss": 1.0003, + "step": 6396 + }, + { + "epoch": 0.5769039996392659, + "grad_norm": 17.970127324930637, + "learning_rate": 1.6016219393257048e-06, + "loss": 0.9968, + "step": 6397 + }, + { + "epoch": 0.5769941831627362, + "grad_norm": 1.5431582563856172, + "learning_rate": 1.6010494526479722e-06, + "loss": 0.9393, + "step": 6398 + }, + { + "epoch": 0.5770843666862064, + "grad_norm": 2.2676315367040747, + "learning_rate": 1.6004770000109006e-06, + "loss": 0.9293, + "step": 6399 + }, + { + "epoch": 0.5771745502096767, + "grad_norm": 1.3809416448319873, + "learning_rate": 1.5999045814633348e-06, + "loss": 0.9557, + "step": 6400 + }, + { + "epoch": 0.5772647337331469, + "grad_norm": 1.8289248029247724, + "learning_rate": 1.5993321970541151e-06, + "loss": 0.9983, + "step": 6401 + }, + { + "epoch": 0.5773549172566173, + "grad_norm": 4.952164710305299, + "learning_rate": 1.5987598468320825e-06, + "loss": 0.9948, + "step": 6402 + }, + { + "epoch": 0.5774451007800875, + "grad_norm": 1.5939615146676847, + "learning_rate": 1.5981875308460717e-06, + "loss": 0.9743, + "step": 6403 + }, + { + "epoch": 0.5775352843035577, + "grad_norm": 2.0199006146824603, + "learning_rate": 1.5976152491449169e-06, + "loss": 0.9743, + "step": 6404 + }, + { + "epoch": 0.577625467827028, + "grad_norm": 2.0700986355856252, + "learning_rate": 1.5970430017774468e-06, + "loss": 1.0516, + "step": 6405 + }, + { + "epoch": 0.5777156513504983, + "grad_norm": 1.663762956621834, + "learning_rate": 1.5964707887924904e-06, + "loss": 1.0015, + "step": 6406 + }, + { + "epoch": 0.5778058348739685, + "grad_norm": 1.273508690090357, + "learning_rate": 1.5958986102388714e-06, + "loss": 0.8737, + "step": 6407 + }, + { + "epoch": 0.5778960183974388, + "grad_norm": 1.6037530572257037, + "learning_rate": 1.5953264661654104e-06, + "loss": 0.9456, + "step": 6408 + }, + { + "epoch": 0.5779862019209091, + "grad_norm": 1.5722241508747639, + "learning_rate": 1.5947543566209276e-06, + "loss": 1.0208, + "step": 6409 + }, + { + "epoch": 0.5780763854443793, + "grad_norm": 1.566119465057825, + "learning_rate": 1.5941822816542367e-06, + "loss": 0.9701, + "step": 6410 + }, + { + "epoch": 0.5781665689678496, + "grad_norm": 2.4074869320713836, + "learning_rate": 1.5936102413141519e-06, + "loss": 0.9451, + "step": 6411 + }, + { + "epoch": 0.5782567524913198, + "grad_norm": 2.0116982468236118, + "learning_rate": 1.5930382356494823e-06, + "loss": 0.9035, + "step": 6412 + }, + { + "epoch": 0.5783469360147901, + "grad_norm": 2.9995515806122452, + "learning_rate": 1.5924662647090335e-06, + "loss": 1.0389, + "step": 6413 + }, + { + "epoch": 0.5784371195382604, + "grad_norm": 1.5026749760049085, + "learning_rate": 1.5918943285416108e-06, + "loss": 1.004, + "step": 6414 + }, + { + "epoch": 0.5785273030617306, + "grad_norm": 1.4253383000670619, + "learning_rate": 1.5913224271960139e-06, + "loss": 0.9235, + "step": 6415 + }, + { + "epoch": 0.5786174865852008, + "grad_norm": 1.727125878571188, + "learning_rate": 1.590750560721041e-06, + "loss": 0.9969, + "step": 6416 + }, + { + "epoch": 0.5787076701086712, + "grad_norm": 1.401998875961724, + "learning_rate": 1.5901787291654874e-06, + "loss": 0.955, + "step": 6417 + }, + { + "epoch": 0.5787978536321414, + "grad_norm": 1.5377723204571634, + "learning_rate": 1.5896069325781435e-06, + "loss": 0.8613, + "step": 6418 + }, + { + "epoch": 0.5788880371556117, + "grad_norm": 1.8230239255321758, + "learning_rate": 1.5890351710077998e-06, + "loss": 1.0247, + "step": 6419 + }, + { + "epoch": 0.5789782206790819, + "grad_norm": 1.4224592412864905, + "learning_rate": 1.5884634445032406e-06, + "loss": 1.042, + "step": 6420 + }, + { + "epoch": 0.5790684042025522, + "grad_norm": 0.5944789627883194, + "learning_rate": 1.5878917531132501e-06, + "loss": 0.8345, + "step": 6421 + }, + { + "epoch": 0.5791585877260225, + "grad_norm": 2.3013093662840993, + "learning_rate": 1.5873200968866077e-06, + "loss": 0.8588, + "step": 6422 + }, + { + "epoch": 0.5792487712494927, + "grad_norm": 1.9089523297128426, + "learning_rate": 1.586748475872089e-06, + "loss": 0.9573, + "step": 6423 + }, + { + "epoch": 0.5793389547729629, + "grad_norm": 1.785517349282791, + "learning_rate": 1.58617689011847e-06, + "loss": 0.8558, + "step": 6424 + }, + { + "epoch": 0.5794291382964333, + "grad_norm": 2.04188441717792, + "learning_rate": 1.5856053396745198e-06, + "loss": 1.022, + "step": 6425 + }, + { + "epoch": 0.5795193218199035, + "grad_norm": 2.3163654909198623, + "learning_rate": 1.5850338245890078e-06, + "loss": 1.0503, + "step": 6426 + }, + { + "epoch": 0.5796095053433737, + "grad_norm": 1.7546332197209453, + "learning_rate": 1.5844623449106974e-06, + "loss": 0.9706, + "step": 6427 + }, + { + "epoch": 0.579699688866844, + "grad_norm": 2.129820796967967, + "learning_rate": 1.583890900688351e-06, + "loss": 1.0306, + "step": 6428 + }, + { + "epoch": 0.5797898723903143, + "grad_norm": 2.966411753928921, + "learning_rate": 1.583319491970728e-06, + "loss": 1.0622, + "step": 6429 + }, + { + "epoch": 0.5798800559137846, + "grad_norm": 1.5986710662086616, + "learning_rate": 1.5827481188065828e-06, + "loss": 1.0022, + "step": 6430 + }, + { + "epoch": 0.5799702394372548, + "grad_norm": 1.4897979457237285, + "learning_rate": 1.5821767812446689e-06, + "loss": 0.9551, + "step": 6431 + }, + { + "epoch": 0.5800604229607251, + "grad_norm": 1.7823582597959415, + "learning_rate": 1.581605479333736e-06, + "loss": 1.1006, + "step": 6432 + }, + { + "epoch": 0.5801506064841954, + "grad_norm": 1.8746491421811489, + "learning_rate": 1.5810342131225308e-06, + "loss": 0.9917, + "step": 6433 + }, + { + "epoch": 0.5802407900076656, + "grad_norm": 1.7867565369540481, + "learning_rate": 1.580462982659797e-06, + "loss": 1.0242, + "step": 6434 + }, + { + "epoch": 0.5803309735311358, + "grad_norm": 2.356746700156279, + "learning_rate": 1.5798917879942736e-06, + "loss": 1.063, + "step": 6435 + }, + { + "epoch": 0.5804211570546062, + "grad_norm": 1.3904807626737097, + "learning_rate": 1.5793206291747006e-06, + "loss": 0.9272, + "step": 6436 + }, + { + "epoch": 0.5805113405780764, + "grad_norm": 1.4831381212145185, + "learning_rate": 1.57874950624981e-06, + "loss": 0.9247, + "step": 6437 + }, + { + "epoch": 0.5806015241015466, + "grad_norm": 1.8124842807482757, + "learning_rate": 1.5781784192683351e-06, + "loss": 0.9833, + "step": 6438 + }, + { + "epoch": 0.5806917076250169, + "grad_norm": 5.160218162656116, + "learning_rate": 1.5776073682790033e-06, + "loss": 0.9484, + "step": 6439 + }, + { + "epoch": 0.5807818911484872, + "grad_norm": 1.809750411597304, + "learning_rate": 1.5770363533305393e-06, + "loss": 1.0126, + "step": 6440 + }, + { + "epoch": 0.5808720746719575, + "grad_norm": 1.8975689317850497, + "learning_rate": 1.5764653744716665e-06, + "loss": 0.9903, + "step": 6441 + }, + { + "epoch": 0.5809622581954277, + "grad_norm": 1.72699949508156, + "learning_rate": 1.575894431751103e-06, + "loss": 1.0124, + "step": 6442 + }, + { + "epoch": 0.5810524417188979, + "grad_norm": 1.8979996541287234, + "learning_rate": 1.575323525217565e-06, + "loss": 1.013, + "step": 6443 + }, + { + "epoch": 0.5811426252423683, + "grad_norm": 2.136189899168339, + "learning_rate": 1.574752654919766e-06, + "loss": 1.0715, + "step": 6444 + }, + { + "epoch": 0.5812328087658385, + "grad_norm": 2.491404060465681, + "learning_rate": 1.5741818209064146e-06, + "loss": 0.9604, + "step": 6445 + }, + { + "epoch": 0.5813229922893087, + "grad_norm": 5.53040130483578, + "learning_rate": 1.5736110232262183e-06, + "loss": 0.9931, + "step": 6446 + }, + { + "epoch": 0.581413175812779, + "grad_norm": 1.508996686260612, + "learning_rate": 1.5730402619278804e-06, + "loss": 1.0343, + "step": 6447 + }, + { + "epoch": 0.5815033593362493, + "grad_norm": 0.6519477921731204, + "learning_rate": 1.5724695370601024e-06, + "loss": 0.793, + "step": 6448 + }, + { + "epoch": 0.5815935428597195, + "grad_norm": 3.4463853614521036, + "learning_rate": 1.5718988486715798e-06, + "loss": 0.9081, + "step": 6449 + }, + { + "epoch": 0.5816837263831898, + "grad_norm": 1.4011404671510828, + "learning_rate": 1.5713281968110087e-06, + "loss": 0.9688, + "step": 6450 + }, + { + "epoch": 0.58177390990666, + "grad_norm": 2.089564566814804, + "learning_rate": 1.5707575815270796e-06, + "loss": 0.9666, + "step": 6451 + }, + { + "epoch": 0.5818640934301303, + "grad_norm": 1.9118228187186541, + "learning_rate": 1.57018700286848e-06, + "loss": 0.9882, + "step": 6452 + }, + { + "epoch": 0.5819542769536006, + "grad_norm": 1.6300866403546979, + "learning_rate": 1.5696164608838956e-06, + "loss": 1.0015, + "step": 6453 + }, + { + "epoch": 0.5820444604770708, + "grad_norm": 4.8507804793192735, + "learning_rate": 1.5690459556220073e-06, + "loss": 1.0031, + "step": 6454 + }, + { + "epoch": 0.582134644000541, + "grad_norm": 2.3667855000100584, + "learning_rate": 1.5684754871314949e-06, + "loss": 1.0482, + "step": 6455 + }, + { + "epoch": 0.5822248275240114, + "grad_norm": 1.9248686014883212, + "learning_rate": 1.5679050554610335e-06, + "loss": 0.9867, + "step": 6456 + }, + { + "epoch": 0.5823150110474816, + "grad_norm": 2.1524031227632197, + "learning_rate": 1.567334660659295e-06, + "loss": 0.9917, + "step": 6457 + }, + { + "epoch": 0.5824051945709519, + "grad_norm": 1.5440473818352516, + "learning_rate": 1.5667643027749488e-06, + "loss": 1.0008, + "step": 6458 + }, + { + "epoch": 0.5824953780944222, + "grad_norm": 1.9940546598222033, + "learning_rate": 1.5661939818566614e-06, + "loss": 1.0578, + "step": 6459 + }, + { + "epoch": 0.5825855616178924, + "grad_norm": 1.7464382847263136, + "learning_rate": 1.5656236979530956e-06, + "loss": 0.8552, + "step": 6460 + }, + { + "epoch": 0.5826757451413627, + "grad_norm": 1.917323918918121, + "learning_rate": 1.5650534511129106e-06, + "loss": 0.9293, + "step": 6461 + }, + { + "epoch": 0.5827659286648329, + "grad_norm": 2.595163626537764, + "learning_rate": 1.5644832413847635e-06, + "loss": 0.9725, + "step": 6462 + }, + { + "epoch": 0.5828561121883032, + "grad_norm": 1.558778527189213, + "learning_rate": 1.5639130688173082e-06, + "loss": 1.029, + "step": 6463 + }, + { + "epoch": 0.5829462957117735, + "grad_norm": 2.753073448711724, + "learning_rate": 1.5633429334591932e-06, + "loss": 0.9623, + "step": 6464 + }, + { + "epoch": 0.5830364792352437, + "grad_norm": 2.2497531967437108, + "learning_rate": 1.562772835359068e-06, + "loss": 1.0203, + "step": 6465 + }, + { + "epoch": 0.5831266627587139, + "grad_norm": 1.852359777129433, + "learning_rate": 1.5622027745655753e-06, + "loss": 0.9528, + "step": 6466 + }, + { + "epoch": 0.5832168462821843, + "grad_norm": 1.8845491116936635, + "learning_rate": 1.561632751127355e-06, + "loss": 0.9446, + "step": 6467 + }, + { + "epoch": 0.5833070298056545, + "grad_norm": 2.0054802912464336, + "learning_rate": 1.561062765093046e-06, + "loss": 1.0314, + "step": 6468 + }, + { + "epoch": 0.5833972133291248, + "grad_norm": 2.9484428951535495, + "learning_rate": 1.5604928165112817e-06, + "loss": 0.9788, + "step": 6469 + }, + { + "epoch": 0.583487396852595, + "grad_norm": 1.7927722698589794, + "learning_rate": 1.5599229054306945e-06, + "loss": 0.977, + "step": 6470 + }, + { + "epoch": 0.5835775803760653, + "grad_norm": 2.841649044425967, + "learning_rate": 1.5593530318999111e-06, + "loss": 0.9828, + "step": 6471 + }, + { + "epoch": 0.5836677638995356, + "grad_norm": 1.5836312950587883, + "learning_rate": 1.5587831959675572e-06, + "loss": 1.0163, + "step": 6472 + }, + { + "epoch": 0.5837579474230058, + "grad_norm": 1.7502248483350973, + "learning_rate": 1.5582133976822534e-06, + "loss": 0.9717, + "step": 6473 + }, + { + "epoch": 0.583848130946476, + "grad_norm": 1.4945596743092633, + "learning_rate": 1.5576436370926185e-06, + "loss": 0.9412, + "step": 6474 + }, + { + "epoch": 0.5839383144699464, + "grad_norm": 1.3122814995678218, + "learning_rate": 1.5570739142472679e-06, + "loss": 0.9663, + "step": 6475 + }, + { + "epoch": 0.5840284979934166, + "grad_norm": 2.000166647643933, + "learning_rate": 1.5565042291948127e-06, + "loss": 1.0503, + "step": 6476 + }, + { + "epoch": 0.5841186815168868, + "grad_norm": 1.522864209414067, + "learning_rate": 1.5559345819838624e-06, + "loss": 1.0191, + "step": 6477 + }, + { + "epoch": 0.5842088650403571, + "grad_norm": 2.1939126978694485, + "learning_rate": 1.5553649726630226e-06, + "loss": 0.9668, + "step": 6478 + }, + { + "epoch": 0.5842990485638274, + "grad_norm": 1.380658039269292, + "learning_rate": 1.5547954012808942e-06, + "loss": 0.915, + "step": 6479 + }, + { + "epoch": 0.5843892320872976, + "grad_norm": 2.0762879706119666, + "learning_rate": 1.5542258678860776e-06, + "loss": 0.9658, + "step": 6480 + }, + { + "epoch": 0.5844794156107679, + "grad_norm": 1.6392125180142072, + "learning_rate": 1.553656372527167e-06, + "loss": 1.1059, + "step": 6481 + }, + { + "epoch": 0.5845695991342382, + "grad_norm": 2.22970936904159, + "learning_rate": 1.5530869152527568e-06, + "loss": 1.0837, + "step": 6482 + }, + { + "epoch": 0.5846597826577085, + "grad_norm": 1.4644435488473497, + "learning_rate": 1.5525174961114353e-06, + "loss": 1.008, + "step": 6483 + }, + { + "epoch": 0.5847499661811787, + "grad_norm": 1.6142668957463369, + "learning_rate": 1.5519481151517875e-06, + "loss": 1.0232, + "step": 6484 + }, + { + "epoch": 0.5848401497046489, + "grad_norm": 2.9026910463686875, + "learning_rate": 1.551378772422398e-06, + "loss": 0.9859, + "step": 6485 + }, + { + "epoch": 0.5849303332281193, + "grad_norm": 1.8158965021040598, + "learning_rate": 1.5508094679718447e-06, + "loss": 1.0027, + "step": 6486 + }, + { + "epoch": 0.5850205167515895, + "grad_norm": 1.5899566685272242, + "learning_rate": 1.5502402018487048e-06, + "loss": 0.8888, + "step": 6487 + }, + { + "epoch": 0.5851107002750597, + "grad_norm": 1.84664560303685, + "learning_rate": 1.54967097410155e-06, + "loss": 0.9911, + "step": 6488 + }, + { + "epoch": 0.58520088379853, + "grad_norm": 2.225783790809196, + "learning_rate": 1.5491017847789519e-06, + "loss": 0.9138, + "step": 6489 + }, + { + "epoch": 0.5852910673220003, + "grad_norm": 1.5984718833648464, + "learning_rate": 1.5485326339294755e-06, + "loss": 0.9323, + "step": 6490 + }, + { + "epoch": 0.5853812508454705, + "grad_norm": 1.4114311985161858, + "learning_rate": 1.5479635216016832e-06, + "loss": 0.9534, + "step": 6491 + }, + { + "epoch": 0.5854714343689408, + "grad_norm": 2.248051859620614, + "learning_rate": 1.547394447844137e-06, + "loss": 0.9037, + "step": 6492 + }, + { + "epoch": 0.585561617892411, + "grad_norm": 1.8038951240682162, + "learning_rate": 1.546825412705391e-06, + "loss": 0.9053, + "step": 6493 + }, + { + "epoch": 0.5856518014158814, + "grad_norm": 1.6545582894657251, + "learning_rate": 1.5462564162340007e-06, + "loss": 1.0121, + "step": 6494 + }, + { + "epoch": 0.5857419849393516, + "grad_norm": 1.68882642820091, + "learning_rate": 1.5456874584785144e-06, + "loss": 0.9932, + "step": 6495 + }, + { + "epoch": 0.5858321684628218, + "grad_norm": 2.1129634145736893, + "learning_rate": 1.5451185394874785e-06, + "loss": 0.9603, + "step": 6496 + }, + { + "epoch": 0.5859223519862921, + "grad_norm": 1.6444448416655006, + "learning_rate": 1.5445496593094381e-06, + "loss": 0.9946, + "step": 6497 + }, + { + "epoch": 0.5860125355097624, + "grad_norm": 1.7901936915363523, + "learning_rate": 1.5439808179929316e-06, + "loss": 0.998, + "step": 6498 + }, + { + "epoch": 0.5861027190332326, + "grad_norm": 1.7063364670750722, + "learning_rate": 1.543412015586496e-06, + "loss": 0.9528, + "step": 6499 + }, + { + "epoch": 0.5861929025567029, + "grad_norm": 1.5910946285751482, + "learning_rate": 1.5428432521386655e-06, + "loss": 0.9546, + "step": 6500 + }, + { + "epoch": 0.5862830860801731, + "grad_norm": 3.2472067346811766, + "learning_rate": 1.5422745276979688e-06, + "loss": 1.02, + "step": 6501 + }, + { + "epoch": 0.5863732696036434, + "grad_norm": 1.7622282725063676, + "learning_rate": 1.5417058423129336e-06, + "loss": 0.9787, + "step": 6502 + }, + { + "epoch": 0.5864634531271137, + "grad_norm": 2.1011948955329376, + "learning_rate": 1.5411371960320822e-06, + "loss": 1.0242, + "step": 6503 + }, + { + "epoch": 0.5865536366505839, + "grad_norm": 1.600065843896265, + "learning_rate": 1.5405685889039363e-06, + "loss": 1.0218, + "step": 6504 + }, + { + "epoch": 0.5866438201740543, + "grad_norm": 2.7940909738714264, + "learning_rate": 1.5400000209770118e-06, + "loss": 0.983, + "step": 6505 + }, + { + "epoch": 0.5867340036975245, + "grad_norm": 1.8443635469273005, + "learning_rate": 1.5394314922998208e-06, + "loss": 1.0016, + "step": 6506 + }, + { + "epoch": 0.5868241872209947, + "grad_norm": 2.107770182168498, + "learning_rate": 1.5388630029208756e-06, + "loss": 1.0415, + "step": 6507 + }, + { + "epoch": 0.586914370744465, + "grad_norm": 1.9617706595799047, + "learning_rate": 1.5382945528886806e-06, + "loss": 1.0115, + "step": 6508 + }, + { + "epoch": 0.5870045542679353, + "grad_norm": 1.9100334892532564, + "learning_rate": 1.5377261422517412e-06, + "loss": 0.9679, + "step": 6509 + }, + { + "epoch": 0.5870947377914055, + "grad_norm": 1.5190136182273777, + "learning_rate": 1.5371577710585553e-06, + "loss": 0.9462, + "step": 6510 + }, + { + "epoch": 0.5871849213148758, + "grad_norm": 0.7076781880184585, + "learning_rate": 1.536589439357621e-06, + "loss": 0.8025, + "step": 6511 + }, + { + "epoch": 0.587275104838346, + "grad_norm": 1.56560360709249, + "learning_rate": 1.5360211471974315e-06, + "loss": 0.9937, + "step": 6512 + }, + { + "epoch": 0.5873652883618163, + "grad_norm": 1.5211565205573618, + "learning_rate": 1.5354528946264753e-06, + "loss": 1.044, + "step": 6513 + }, + { + "epoch": 0.5874554718852866, + "grad_norm": 1.3426740020001031, + "learning_rate": 1.5348846816932399e-06, + "loss": 0.9234, + "step": 6514 + }, + { + "epoch": 0.5875456554087568, + "grad_norm": 0.6257086312857508, + "learning_rate": 1.5343165084462077e-06, + "loss": 0.8119, + "step": 6515 + }, + { + "epoch": 0.587635838932227, + "grad_norm": 1.8450190815546215, + "learning_rate": 1.5337483749338595e-06, + "loss": 0.9411, + "step": 6516 + }, + { + "epoch": 0.5877260224556974, + "grad_norm": 1.8080001087484074, + "learning_rate": 1.5331802812046708e-06, + "loss": 0.9611, + "step": 6517 + }, + { + "epoch": 0.5878162059791676, + "grad_norm": 2.348647328924733, + "learning_rate": 1.5326122273071133e-06, + "loss": 0.8416, + "step": 6518 + }, + { + "epoch": 0.5879063895026378, + "grad_norm": 1.749814773021254, + "learning_rate": 1.532044213289659e-06, + "loss": 1.0283, + "step": 6519 + }, + { + "epoch": 0.5879965730261081, + "grad_norm": 1.640861639486137, + "learning_rate": 1.5314762392007718e-06, + "loss": 0.8762, + "step": 6520 + }, + { + "epoch": 0.5880867565495784, + "grad_norm": 1.6408487803111635, + "learning_rate": 1.530908305088916e-06, + "loss": 1.0378, + "step": 6521 + }, + { + "epoch": 0.5881769400730487, + "grad_norm": 2.038935164060144, + "learning_rate": 1.5303404110025501e-06, + "loss": 0.9807, + "step": 6522 + }, + { + "epoch": 0.5882671235965189, + "grad_norm": 2.177383681999779, + "learning_rate": 1.5297725569901293e-06, + "loss": 0.9442, + "step": 6523 + }, + { + "epoch": 0.5883573071199891, + "grad_norm": 1.855237572975212, + "learning_rate": 1.5292047431001077e-06, + "loss": 0.8813, + "step": 6524 + }, + { + "epoch": 0.5884474906434595, + "grad_norm": 0.7258253562509335, + "learning_rate": 1.5286369693809321e-06, + "loss": 0.8255, + "step": 6525 + }, + { + "epoch": 0.5885376741669297, + "grad_norm": 1.9504797760971815, + "learning_rate": 1.5280692358810506e-06, + "loss": 0.9978, + "step": 6526 + }, + { + "epoch": 0.5886278576903999, + "grad_norm": 1.6245908955804627, + "learning_rate": 1.527501542648904e-06, + "loss": 0.9327, + "step": 6527 + }, + { + "epoch": 0.5887180412138703, + "grad_norm": 1.653656350250042, + "learning_rate": 1.5269338897329308e-06, + "loss": 0.8944, + "step": 6528 + }, + { + "epoch": 0.5888082247373405, + "grad_norm": 1.8939937497633026, + "learning_rate": 1.5263662771815662e-06, + "loss": 1.013, + "step": 6529 + }, + { + "epoch": 0.5888984082608107, + "grad_norm": 1.7447953485098942, + "learning_rate": 1.5257987050432429e-06, + "loss": 1.0616, + "step": 6530 + }, + { + "epoch": 0.588988591784281, + "grad_norm": 1.688518428779865, + "learning_rate": 1.5252311733663887e-06, + "loss": 0.895, + "step": 6531 + }, + { + "epoch": 0.5890787753077513, + "grad_norm": 1.669135942727058, + "learning_rate": 1.5246636821994281e-06, + "loss": 0.9277, + "step": 6532 + }, + { + "epoch": 0.5891689588312216, + "grad_norm": 1.893363860392388, + "learning_rate": 1.524096231590784e-06, + "loss": 1.0338, + "step": 6533 + }, + { + "epoch": 0.5892591423546918, + "grad_norm": 1.6578400374991862, + "learning_rate": 1.5235288215888736e-06, + "loss": 1.0308, + "step": 6534 + }, + { + "epoch": 0.589349325878162, + "grad_norm": 1.86732686691817, + "learning_rate": 1.5229614522421102e-06, + "loss": 0.9888, + "step": 6535 + }, + { + "epoch": 0.5894395094016324, + "grad_norm": 1.7285504526196471, + "learning_rate": 1.5223941235989071e-06, + "loss": 0.8233, + "step": 6536 + }, + { + "epoch": 0.5895296929251026, + "grad_norm": 1.4277124155341427, + "learning_rate": 1.52182683570767e-06, + "loss": 0.9554, + "step": 6537 + }, + { + "epoch": 0.5896198764485728, + "grad_norm": 1.664382187487231, + "learning_rate": 1.5212595886168046e-06, + "loss": 1.0065, + "step": 6538 + }, + { + "epoch": 0.5897100599720431, + "grad_norm": 1.6927888430498148, + "learning_rate": 1.520692382374711e-06, + "loss": 1.0028, + "step": 6539 + }, + { + "epoch": 0.5898002434955134, + "grad_norm": 2.005494794984814, + "learning_rate": 1.5201252170297854e-06, + "loss": 1.0114, + "step": 6540 + }, + { + "epoch": 0.5898904270189836, + "grad_norm": 1.788564221153467, + "learning_rate": 1.5195580926304232e-06, + "loss": 0.8726, + "step": 6541 + }, + { + "epoch": 0.5899806105424539, + "grad_norm": 1.819996098210271, + "learning_rate": 1.5189910092250131e-06, + "loss": 1.0077, + "step": 6542 + }, + { + "epoch": 0.5900707940659241, + "grad_norm": 2.5921592890220087, + "learning_rate": 1.5184239668619427e-06, + "loss": 0.9948, + "step": 6543 + }, + { + "epoch": 0.5901609775893945, + "grad_norm": 1.60555384055416, + "learning_rate": 1.5178569655895946e-06, + "loss": 0.985, + "step": 6544 + }, + { + "epoch": 0.5902511611128647, + "grad_norm": 1.6166362467915618, + "learning_rate": 1.5172900054563487e-06, + "loss": 1.016, + "step": 6545 + }, + { + "epoch": 0.5903413446363349, + "grad_norm": 1.9654519446143004, + "learning_rate": 1.5167230865105814e-06, + "loss": 1.0035, + "step": 6546 + }, + { + "epoch": 0.5904315281598052, + "grad_norm": 2.161850001380021, + "learning_rate": 1.5161562088006644e-06, + "loss": 0.9023, + "step": 6547 + }, + { + "epoch": 0.5905217116832755, + "grad_norm": 2.0446261799703906, + "learning_rate": 1.5155893723749685e-06, + "loss": 1.0137, + "step": 6548 + }, + { + "epoch": 0.5906118952067457, + "grad_norm": 1.8789763566418205, + "learning_rate": 1.5150225772818582e-06, + "loss": 0.9948, + "step": 6549 + }, + { + "epoch": 0.590702078730216, + "grad_norm": 2.7191336350144946, + "learning_rate": 1.5144558235696949e-06, + "loss": 1.0368, + "step": 6550 + }, + { + "epoch": 0.5907922622536863, + "grad_norm": 1.8238098862602712, + "learning_rate": 1.5138891112868388e-06, + "loss": 0.9059, + "step": 6551 + }, + { + "epoch": 0.5908824457771565, + "grad_norm": 0.6854953800869333, + "learning_rate": 1.5133224404816433e-06, + "loss": 0.8155, + "step": 6552 + }, + { + "epoch": 0.5909726293006268, + "grad_norm": 1.8953468690729447, + "learning_rate": 1.5127558112024617e-06, + "loss": 0.9162, + "step": 6553 + }, + { + "epoch": 0.591062812824097, + "grad_norm": 0.606444599931332, + "learning_rate": 1.5121892234976404e-06, + "loss": 0.8293, + "step": 6554 + }, + { + "epoch": 0.5911529963475673, + "grad_norm": 1.7017922742828986, + "learning_rate": 1.5116226774155243e-06, + "loss": 0.9701, + "step": 6555 + }, + { + "epoch": 0.5912431798710376, + "grad_norm": 1.7656700331084687, + "learning_rate": 1.5110561730044547e-06, + "loss": 0.9576, + "step": 6556 + }, + { + "epoch": 0.5913333633945078, + "grad_norm": 2.1991324534856087, + "learning_rate": 1.510489710312768e-06, + "loss": 0.984, + "step": 6557 + }, + { + "epoch": 0.591423546917978, + "grad_norm": 2.0092989988770777, + "learning_rate": 1.5099232893887987e-06, + "loss": 0.9756, + "step": 6558 + }, + { + "epoch": 0.5915137304414484, + "grad_norm": 1.482448572614503, + "learning_rate": 1.5093569102808758e-06, + "loss": 0.9572, + "step": 6559 + }, + { + "epoch": 0.5916039139649186, + "grad_norm": 1.7207874359709663, + "learning_rate": 1.5087905730373275e-06, + "loss": 0.9869, + "step": 6560 + }, + { + "epoch": 0.5916940974883889, + "grad_norm": 0.6227062334510449, + "learning_rate": 1.508224277706476e-06, + "loss": 0.8276, + "step": 6561 + }, + { + "epoch": 0.5917842810118591, + "grad_norm": 1.8609097982373846, + "learning_rate": 1.5076580243366399e-06, + "loss": 0.9414, + "step": 6562 + }, + { + "epoch": 0.5918744645353294, + "grad_norm": 2.491191892352965, + "learning_rate": 1.507091812976137e-06, + "loss": 0.8976, + "step": 6563 + }, + { + "epoch": 0.5919646480587997, + "grad_norm": 1.8106045513163713, + "learning_rate": 1.5065256436732773e-06, + "loss": 0.9331, + "step": 6564 + }, + { + "epoch": 0.5920548315822699, + "grad_norm": 1.822283156986268, + "learning_rate": 1.5059595164763717e-06, + "loss": 0.915, + "step": 6565 + }, + { + "epoch": 0.5921450151057401, + "grad_norm": 0.5753335835138412, + "learning_rate": 1.5053934314337243e-06, + "loss": 0.8229, + "step": 6566 + }, + { + "epoch": 0.5922351986292105, + "grad_norm": 1.7016906998245194, + "learning_rate": 1.5048273885936356e-06, + "loss": 0.974, + "step": 6567 + }, + { + "epoch": 0.5923253821526807, + "grad_norm": 1.5242119801193519, + "learning_rate": 1.5042613880044053e-06, + "loss": 0.9011, + "step": 6568 + }, + { + "epoch": 0.592415565676151, + "grad_norm": 1.6703848927496163, + "learning_rate": 1.5036954297143265e-06, + "loss": 1.0095, + "step": 6569 + }, + { + "epoch": 0.5925057491996212, + "grad_norm": 1.3256050437853162, + "learning_rate": 1.50312951377169e-06, + "loss": 0.938, + "step": 6570 + }, + { + "epoch": 0.5925959327230915, + "grad_norm": 0.582077971765742, + "learning_rate": 1.502563640224784e-06, + "loss": 0.8245, + "step": 6571 + }, + { + "epoch": 0.5926861162465618, + "grad_norm": 2.3836391328375366, + "learning_rate": 1.5019978091218903e-06, + "loss": 0.9803, + "step": 6572 + }, + { + "epoch": 0.592776299770032, + "grad_norm": 1.4853412093604914, + "learning_rate": 1.50143202051129e-06, + "loss": 1.0475, + "step": 6573 + }, + { + "epoch": 0.5928664832935022, + "grad_norm": 1.8380973286185058, + "learning_rate": 1.500866274441258e-06, + "loss": 1.0428, + "step": 6574 + }, + { + "epoch": 0.5929566668169726, + "grad_norm": 1.7174189181671766, + "learning_rate": 1.5003005709600682e-06, + "loss": 0.9706, + "step": 6575 + }, + { + "epoch": 0.5930468503404428, + "grad_norm": 1.5838519133886608, + "learning_rate": 1.4997349101159885e-06, + "loss": 1.0618, + "step": 6576 + }, + { + "epoch": 0.593137033863913, + "grad_norm": 1.7189439837486697, + "learning_rate": 1.4991692919572854e-06, + "loss": 0.9385, + "step": 6577 + }, + { + "epoch": 0.5932272173873834, + "grad_norm": 1.7284216903766292, + "learning_rate": 1.4986037165322199e-06, + "loss": 0.9958, + "step": 6578 + }, + { + "epoch": 0.5933174009108536, + "grad_norm": 0.665406404339269, + "learning_rate": 1.498038183889049e-06, + "loss": 0.8525, + "step": 6579 + }, + { + "epoch": 0.5934075844343238, + "grad_norm": 1.6316577533300527, + "learning_rate": 1.4974726940760292e-06, + "loss": 0.9051, + "step": 6580 + }, + { + "epoch": 0.5934977679577941, + "grad_norm": 1.9338544573493845, + "learning_rate": 1.496907247141409e-06, + "loss": 1.059, + "step": 6581 + }, + { + "epoch": 0.5935879514812644, + "grad_norm": 0.7527538445512951, + "learning_rate": 1.4963418431334372e-06, + "loss": 0.8935, + "step": 6582 + }, + { + "epoch": 0.5936781350047347, + "grad_norm": 1.6801122084111444, + "learning_rate": 1.4957764821003566e-06, + "loss": 0.9677, + "step": 6583 + }, + { + "epoch": 0.5937683185282049, + "grad_norm": 1.741689568084596, + "learning_rate": 1.4952111640904063e-06, + "loss": 0.9907, + "step": 6584 + }, + { + "epoch": 0.5938585020516751, + "grad_norm": 2.034367091665996, + "learning_rate": 1.494645889151823e-06, + "loss": 0.9665, + "step": 6585 + }, + { + "epoch": 0.5939486855751455, + "grad_norm": 0.6547194070256043, + "learning_rate": 1.494080657332839e-06, + "loss": 0.7984, + "step": 6586 + }, + { + "epoch": 0.5940388690986157, + "grad_norm": 1.7542137776962923, + "learning_rate": 1.4935154686816832e-06, + "loss": 0.9425, + "step": 6587 + }, + { + "epoch": 0.5941290526220859, + "grad_norm": 1.8543007477132643, + "learning_rate": 1.4929503232465802e-06, + "loss": 1.0641, + "step": 6588 + }, + { + "epoch": 0.5942192361455562, + "grad_norm": 2.1011235228898757, + "learning_rate": 1.492385221075751e-06, + "loss": 0.8896, + "step": 6589 + }, + { + "epoch": 0.5943094196690265, + "grad_norm": 1.740974311774948, + "learning_rate": 1.4918201622174142e-06, + "loss": 1.0169, + "step": 6590 + }, + { + "epoch": 0.5943996031924967, + "grad_norm": 1.618702644038396, + "learning_rate": 1.4912551467197827e-06, + "loss": 1.0072, + "step": 6591 + }, + { + "epoch": 0.594489786715967, + "grad_norm": 2.727921286686293, + "learning_rate": 1.4906901746310678e-06, + "loss": 1.07, + "step": 6592 + }, + { + "epoch": 0.5945799702394372, + "grad_norm": 9.690607175592282, + "learning_rate": 1.4901252459994757e-06, + "loss": 0.9495, + "step": 6593 + }, + { + "epoch": 0.5946701537629075, + "grad_norm": 1.9659695443116016, + "learning_rate": 1.489560360873208e-06, + "loss": 0.9679, + "step": 6594 + }, + { + "epoch": 0.5947603372863778, + "grad_norm": 1.5371721150682456, + "learning_rate": 1.4889955193004659e-06, + "loss": 0.9651, + "step": 6595 + }, + { + "epoch": 0.594850520809848, + "grad_norm": 0.6295410411884714, + "learning_rate": 1.4884307213294428e-06, + "loss": 0.8119, + "step": 6596 + }, + { + "epoch": 0.5949407043333182, + "grad_norm": 1.6958426073025397, + "learning_rate": 1.4878659670083321e-06, + "loss": 0.97, + "step": 6597 + }, + { + "epoch": 0.5950308878567886, + "grad_norm": 2.020938818102233, + "learning_rate": 1.4873012563853208e-06, + "loss": 0.9054, + "step": 6598 + }, + { + "epoch": 0.5951210713802588, + "grad_norm": 1.8829393146275826, + "learning_rate": 1.4867365895085935e-06, + "loss": 0.9927, + "step": 6599 + }, + { + "epoch": 0.5952112549037291, + "grad_norm": 1.4492522284658584, + "learning_rate": 1.4861719664263301e-06, + "loss": 0.9636, + "step": 6600 + }, + { + "epoch": 0.5953014384271994, + "grad_norm": 1.7986830529644078, + "learning_rate": 1.485607387186708e-06, + "loss": 1.088, + "step": 6601 + }, + { + "epoch": 0.5953916219506696, + "grad_norm": 4.126970225207119, + "learning_rate": 1.4850428518379001e-06, + "loss": 0.981, + "step": 6602 + }, + { + "epoch": 0.5954818054741399, + "grad_norm": 1.446794234919749, + "learning_rate": 1.4844783604280746e-06, + "loss": 0.9776, + "step": 6603 + }, + { + "epoch": 0.5955719889976101, + "grad_norm": 2.1594266849410513, + "learning_rate": 1.483913913005399e-06, + "loss": 0.9196, + "step": 6604 + }, + { + "epoch": 0.5956621725210804, + "grad_norm": 1.9040936014304608, + "learning_rate": 1.483349509618034e-06, + "loss": 1.0404, + "step": 6605 + }, + { + "epoch": 0.5957523560445507, + "grad_norm": 1.6020559899717215, + "learning_rate": 1.4827851503141367e-06, + "loss": 0.9151, + "step": 6606 + }, + { + "epoch": 0.5958425395680209, + "grad_norm": 1.5786739942991679, + "learning_rate": 1.482220835141863e-06, + "loss": 0.9054, + "step": 6607 + }, + { + "epoch": 0.5959327230914911, + "grad_norm": 1.5517777092956235, + "learning_rate": 1.481656564149362e-06, + "loss": 1.0597, + "step": 6608 + }, + { + "epoch": 0.5960229066149615, + "grad_norm": 1.7791557046640405, + "learning_rate": 1.4810923373847818e-06, + "loss": 0.9769, + "step": 6609 + }, + { + "epoch": 0.5961130901384317, + "grad_norm": 1.5884865256156182, + "learning_rate": 1.4805281548962647e-06, + "loss": 0.9893, + "step": 6610 + }, + { + "epoch": 0.596203273661902, + "grad_norm": 1.7770483421228864, + "learning_rate": 1.4799640167319488e-06, + "loss": 0.9285, + "step": 6611 + }, + { + "epoch": 0.5962934571853722, + "grad_norm": 1.657539010027572, + "learning_rate": 1.4793999229399714e-06, + "loss": 0.9703, + "step": 6612 + }, + { + "epoch": 0.5963836407088425, + "grad_norm": 4.330616368216204, + "learning_rate": 1.4788358735684626e-06, + "loss": 1.0473, + "step": 6613 + }, + { + "epoch": 0.5964738242323128, + "grad_norm": 2.592284466457221, + "learning_rate": 1.4782718686655514e-06, + "loss": 1.0506, + "step": 6614 + }, + { + "epoch": 0.596564007755783, + "grad_norm": 2.072301866104809, + "learning_rate": 1.4777079082793605e-06, + "loss": 1.0349, + "step": 6615 + }, + { + "epoch": 0.5966541912792532, + "grad_norm": 2.2851345485080645, + "learning_rate": 1.4771439924580108e-06, + "loss": 0.9512, + "step": 6616 + }, + { + "epoch": 0.5967443748027236, + "grad_norm": 1.6572461281623445, + "learning_rate": 1.4765801212496189e-06, + "loss": 1.027, + "step": 6617 + }, + { + "epoch": 0.5968345583261938, + "grad_norm": 1.61061033965349, + "learning_rate": 1.4760162947022968e-06, + "loss": 0.8983, + "step": 6618 + }, + { + "epoch": 0.596924741849664, + "grad_norm": 2.3029868882899653, + "learning_rate": 1.475452512864154e-06, + "loss": 0.9729, + "step": 6619 + }, + { + "epoch": 0.5970149253731343, + "grad_norm": 1.515622011162082, + "learning_rate": 1.4748887757832945e-06, + "loss": 0.9926, + "step": 6620 + }, + { + "epoch": 0.5971051088966046, + "grad_norm": 1.772688720190546, + "learning_rate": 1.4743250835078209e-06, + "loss": 1.0147, + "step": 6621 + }, + { + "epoch": 0.5971952924200749, + "grad_norm": 1.8618848514456323, + "learning_rate": 1.4737614360858297e-06, + "loss": 0.9457, + "step": 6622 + }, + { + "epoch": 0.5972854759435451, + "grad_norm": 2.1849329463699925, + "learning_rate": 1.4731978335654133e-06, + "loss": 0.9388, + "step": 6623 + }, + { + "epoch": 0.5973756594670154, + "grad_norm": 2.105366493813537, + "learning_rate": 1.4726342759946638e-06, + "loss": 0.9218, + "step": 6624 + }, + { + "epoch": 0.5974658429904857, + "grad_norm": 1.7332123304582139, + "learning_rate": 1.4720707634216653e-06, + "loss": 0.982, + "step": 6625 + }, + { + "epoch": 0.5975560265139559, + "grad_norm": 2.4231874908819573, + "learning_rate": 1.4715072958945e-06, + "loss": 0.9905, + "step": 6626 + }, + { + "epoch": 0.5976462100374261, + "grad_norm": 1.9013661216572193, + "learning_rate": 1.470943873461247e-06, + "loss": 0.982, + "step": 6627 + }, + { + "epoch": 0.5977363935608965, + "grad_norm": 2.060689651821809, + "learning_rate": 1.470380496169979e-06, + "loss": 0.9443, + "step": 6628 + }, + { + "epoch": 0.5978265770843667, + "grad_norm": 2.0490760907724113, + "learning_rate": 1.4698171640687682e-06, + "loss": 0.973, + "step": 6629 + }, + { + "epoch": 0.5979167606078369, + "grad_norm": 1.5766148659635912, + "learning_rate": 1.4692538772056792e-06, + "loss": 0.9207, + "step": 6630 + }, + { + "epoch": 0.5980069441313072, + "grad_norm": 1.3987710650837628, + "learning_rate": 1.4686906356287772e-06, + "loss": 0.8956, + "step": 6631 + }, + { + "epoch": 0.5980971276547775, + "grad_norm": 1.6829993665788063, + "learning_rate": 1.4681274393861194e-06, + "loss": 1.0136, + "step": 6632 + }, + { + "epoch": 0.5981873111782477, + "grad_norm": 1.5023902286792725, + "learning_rate": 1.4675642885257603e-06, + "loss": 1.0779, + "step": 6633 + }, + { + "epoch": 0.598277494701718, + "grad_norm": 2.047391171200598, + "learning_rate": 1.4670011830957529e-06, + "loss": 1.0715, + "step": 6634 + }, + { + "epoch": 0.5983676782251882, + "grad_norm": 2.261631991974017, + "learning_rate": 1.4664381231441427e-06, + "loss": 0.9484, + "step": 6635 + }, + { + "epoch": 0.5984578617486586, + "grad_norm": 1.7947855867565905, + "learning_rate": 1.4658751087189746e-06, + "loss": 0.9935, + "step": 6636 + }, + { + "epoch": 0.5985480452721288, + "grad_norm": 2.2545275170767707, + "learning_rate": 1.4653121398682874e-06, + "loss": 1.0329, + "step": 6637 + }, + { + "epoch": 0.598638228795599, + "grad_norm": 1.2635832907332383, + "learning_rate": 1.4647492166401159e-06, + "loss": 0.9217, + "step": 6638 + }, + { + "epoch": 0.5987284123190693, + "grad_norm": 1.7017177402380208, + "learning_rate": 1.4641863390824934e-06, + "loss": 0.9666, + "step": 6639 + }, + { + "epoch": 0.5988185958425396, + "grad_norm": 1.513314599972924, + "learning_rate": 1.4636235072434465e-06, + "loss": 0.972, + "step": 6640 + }, + { + "epoch": 0.5989087793660098, + "grad_norm": 2.0995607734200457, + "learning_rate": 1.4630607211709994e-06, + "loss": 1.009, + "step": 6641 + }, + { + "epoch": 0.5989989628894801, + "grad_norm": 1.471835851452398, + "learning_rate": 1.4624979809131723e-06, + "loss": 0.9876, + "step": 6642 + }, + { + "epoch": 0.5990891464129503, + "grad_norm": 1.9355662139942678, + "learning_rate": 1.4619352865179814e-06, + "loss": 0.8924, + "step": 6643 + }, + { + "epoch": 0.5991793299364206, + "grad_norm": 1.6976629742464189, + "learning_rate": 1.4613726380334391e-06, + "loss": 0.9512, + "step": 6644 + }, + { + "epoch": 0.5992695134598909, + "grad_norm": 1.8522859600549728, + "learning_rate": 1.4608100355075522e-06, + "loss": 0.8361, + "step": 6645 + }, + { + "epoch": 0.5993596969833611, + "grad_norm": 0.7909688299960559, + "learning_rate": 1.460247478988327e-06, + "loss": 0.7932, + "step": 6646 + }, + { + "epoch": 0.5994498805068315, + "grad_norm": 1.8216106713784463, + "learning_rate": 1.4596849685237623e-06, + "loss": 1.0005, + "step": 6647 + }, + { + "epoch": 0.5995400640303017, + "grad_norm": 1.5712279123530657, + "learning_rate": 1.459122504161856e-06, + "loss": 0.8883, + "step": 6648 + }, + { + "epoch": 0.5996302475537719, + "grad_norm": 1.5345923089248186, + "learning_rate": 1.4585600859506001e-06, + "loss": 0.9519, + "step": 6649 + }, + { + "epoch": 0.5997204310772422, + "grad_norm": 1.5177737862274001, + "learning_rate": 1.4579977139379826e-06, + "loss": 1.0395, + "step": 6650 + }, + { + "epoch": 0.5998106146007125, + "grad_norm": 1.8342266146222097, + "learning_rate": 1.4574353881719895e-06, + "loss": 1.0449, + "step": 6651 + }, + { + "epoch": 0.5999007981241827, + "grad_norm": 2.5850481884004677, + "learning_rate": 1.4568731087005998e-06, + "loss": 0.9831, + "step": 6652 + }, + { + "epoch": 0.599990981647653, + "grad_norm": 1.3651483205761694, + "learning_rate": 1.4563108755717916e-06, + "loss": 1.0058, + "step": 6653 + }, + { + "epoch": 0.6000811651711232, + "grad_norm": 1.490284770148469, + "learning_rate": 1.455748688833538e-06, + "loss": 1.0122, + "step": 6654 + }, + { + "epoch": 0.6001713486945935, + "grad_norm": 2.4120304975084053, + "learning_rate": 1.4551865485338065e-06, + "loss": 0.8928, + "step": 6655 + }, + { + "epoch": 0.6002615322180638, + "grad_norm": 2.1593802025562554, + "learning_rate": 1.4546244547205629e-06, + "loss": 1.0476, + "step": 6656 + }, + { + "epoch": 0.600351715741534, + "grad_norm": 2.9252538065548555, + "learning_rate": 1.4540624074417678e-06, + "loss": 1.0059, + "step": 6657 + }, + { + "epoch": 0.6004418992650042, + "grad_norm": 1.4158144893518596, + "learning_rate": 1.453500406745379e-06, + "loss": 1.0384, + "step": 6658 + }, + { + "epoch": 0.6005320827884746, + "grad_norm": 2.366822973202836, + "learning_rate": 1.4529384526793486e-06, + "loss": 1.0572, + "step": 6659 + }, + { + "epoch": 0.6006222663119448, + "grad_norm": 1.7660818268556473, + "learning_rate": 1.4523765452916252e-06, + "loss": 1.036, + "step": 6660 + }, + { + "epoch": 0.600712449835415, + "grad_norm": 12.271365470984104, + "learning_rate": 1.4518146846301554e-06, + "loss": 1.0126, + "step": 6661 + }, + { + "epoch": 0.6008026333588853, + "grad_norm": 1.3753980580600056, + "learning_rate": 1.4512528707428787e-06, + "loss": 0.9602, + "step": 6662 + }, + { + "epoch": 0.6008928168823556, + "grad_norm": 2.061736109907316, + "learning_rate": 1.4506911036777335e-06, + "loss": 0.9662, + "step": 6663 + }, + { + "epoch": 0.6009830004058259, + "grad_norm": 0.6903872982755246, + "learning_rate": 1.450129383482651e-06, + "loss": 0.7853, + "step": 6664 + }, + { + "epoch": 0.6010731839292961, + "grad_norm": 1.6600540668231263, + "learning_rate": 1.4495677102055629e-06, + "loss": 0.9397, + "step": 6665 + }, + { + "epoch": 0.6011633674527663, + "grad_norm": 1.608564228159352, + "learning_rate": 1.4490060838943924e-06, + "loss": 1.0205, + "step": 6666 + }, + { + "epoch": 0.6012535509762367, + "grad_norm": 1.5595931670748302, + "learning_rate": 1.4484445045970609e-06, + "loss": 0.9848, + "step": 6667 + }, + { + "epoch": 0.6013437344997069, + "grad_norm": 1.7092992176503812, + "learning_rate": 1.447882972361485e-06, + "loss": 0.9778, + "step": 6668 + }, + { + "epoch": 0.6014339180231771, + "grad_norm": 1.6297062434785476, + "learning_rate": 1.4473214872355785e-06, + "loss": 0.9668, + "step": 6669 + }, + { + "epoch": 0.6015241015466475, + "grad_norm": 2.0118693767562643, + "learning_rate": 1.4467600492672508e-06, + "loss": 1.0166, + "step": 6670 + }, + { + "epoch": 0.6016142850701177, + "grad_norm": 1.7498656630044525, + "learning_rate": 1.4461986585044054e-06, + "loss": 0.9607, + "step": 6671 + }, + { + "epoch": 0.601704468593588, + "grad_norm": 1.5789731456509606, + "learning_rate": 1.4456373149949446e-06, + "loss": 0.9585, + "step": 6672 + }, + { + "epoch": 0.6017946521170582, + "grad_norm": 1.6762760027775239, + "learning_rate": 1.4450760187867648e-06, + "loss": 0.9914, + "step": 6673 + }, + { + "epoch": 0.6018848356405285, + "grad_norm": 1.7738112001298376, + "learning_rate": 1.4445147699277581e-06, + "loss": 0.9765, + "step": 6674 + }, + { + "epoch": 0.6019750191639988, + "grad_norm": 1.9258508882447394, + "learning_rate": 1.4439535684658154e-06, + "loss": 1.0326, + "step": 6675 + }, + { + "epoch": 0.602065202687469, + "grad_norm": 0.7678043721210935, + "learning_rate": 1.44339241444882e-06, + "loss": 0.8546, + "step": 6676 + }, + { + "epoch": 0.6021553862109392, + "grad_norm": 1.7077138134084, + "learning_rate": 1.4428313079246518e-06, + "loss": 0.9304, + "step": 6677 + }, + { + "epoch": 0.6022455697344096, + "grad_norm": 0.6630443397024925, + "learning_rate": 1.4422702489411896e-06, + "loss": 0.8325, + "step": 6678 + }, + { + "epoch": 0.6023357532578798, + "grad_norm": 2.0283884873593263, + "learning_rate": 1.4417092375463043e-06, + "loss": 0.9314, + "step": 6679 + }, + { + "epoch": 0.60242593678135, + "grad_norm": 2.279038180042741, + "learning_rate": 1.441148273787866e-06, + "loss": 1.0183, + "step": 6680 + }, + { + "epoch": 0.6025161203048203, + "grad_norm": 0.77750437241579, + "learning_rate": 1.4405873577137383e-06, + "loss": 0.8387, + "step": 6681 + }, + { + "epoch": 0.6026063038282906, + "grad_norm": 7.067417019767464, + "learning_rate": 1.4400264893717816e-06, + "loss": 0.9955, + "step": 6682 + }, + { + "epoch": 0.6026964873517608, + "grad_norm": 1.841870788838647, + "learning_rate": 1.4394656688098526e-06, + "loss": 1.033, + "step": 6683 + }, + { + "epoch": 0.6027866708752311, + "grad_norm": 15.107494469969408, + "learning_rate": 1.4389048960758032e-06, + "loss": 1.0064, + "step": 6684 + }, + { + "epoch": 0.6028768543987013, + "grad_norm": 2.0740726512886396, + "learning_rate": 1.4383441712174826e-06, + "loss": 1.0115, + "step": 6685 + }, + { + "epoch": 0.6029670379221717, + "grad_norm": 1.5720466413882876, + "learning_rate": 1.4377834942827333e-06, + "loss": 1.0495, + "step": 6686 + }, + { + "epoch": 0.6030572214456419, + "grad_norm": 1.8085417441353548, + "learning_rate": 1.437222865319397e-06, + "loss": 0.9243, + "step": 6687 + }, + { + "epoch": 0.6031474049691121, + "grad_norm": 1.4690621227011553, + "learning_rate": 1.4366622843753092e-06, + "loss": 0.9131, + "step": 6688 + }, + { + "epoch": 0.6032375884925824, + "grad_norm": 1.4418177237045853, + "learning_rate": 1.4361017514983006e-06, + "loss": 1.0035, + "step": 6689 + }, + { + "epoch": 0.6033277720160527, + "grad_norm": 1.4805668217125183, + "learning_rate": 1.4355412667362006e-06, + "loss": 1.0384, + "step": 6690 + }, + { + "epoch": 0.6034179555395229, + "grad_norm": 1.6361753886487975, + "learning_rate": 1.4349808301368311e-06, + "loss": 0.9394, + "step": 6691 + }, + { + "epoch": 0.6035081390629932, + "grad_norm": 2.1419208252109057, + "learning_rate": 1.4344204417480139e-06, + "loss": 1.0286, + "step": 6692 + }, + { + "epoch": 0.6035983225864634, + "grad_norm": 1.6748854014351768, + "learning_rate": 1.4338601016175628e-06, + "loss": 0.9347, + "step": 6693 + }, + { + "epoch": 0.6036885061099337, + "grad_norm": 1.641025675938297, + "learning_rate": 1.433299809793289e-06, + "loss": 0.9813, + "step": 6694 + }, + { + "epoch": 0.603778689633404, + "grad_norm": 2.109460334464586, + "learning_rate": 1.432739566323001e-06, + "loss": 0.8542, + "step": 6695 + }, + { + "epoch": 0.6038688731568742, + "grad_norm": 1.9787472196789582, + "learning_rate": 1.4321793712545004e-06, + "loss": 0.9511, + "step": 6696 + }, + { + "epoch": 0.6039590566803446, + "grad_norm": 2.2609643814608447, + "learning_rate": 1.4316192246355873e-06, + "loss": 1.0277, + "step": 6697 + }, + { + "epoch": 0.6040492402038148, + "grad_norm": 0.6129673683059527, + "learning_rate": 1.4310591265140555e-06, + "loss": 0.816, + "step": 6698 + }, + { + "epoch": 0.604139423727285, + "grad_norm": 1.422045854636625, + "learning_rate": 1.4304990769376963e-06, + "loss": 0.9893, + "step": 6699 + }, + { + "epoch": 0.6042296072507553, + "grad_norm": 1.4662714396776788, + "learning_rate": 1.4299390759542962e-06, + "loss": 0.9662, + "step": 6700 + }, + { + "epoch": 0.6043197907742256, + "grad_norm": 1.491687870762891, + "learning_rate": 1.4293791236116368e-06, + "loss": 0.9633, + "step": 6701 + }, + { + "epoch": 0.6044099742976958, + "grad_norm": 1.9986755634925888, + "learning_rate": 1.4288192199574978e-06, + "loss": 0.9495, + "step": 6702 + }, + { + "epoch": 0.6045001578211661, + "grad_norm": 0.6283477056269413, + "learning_rate": 1.4282593650396524e-06, + "loss": 0.7883, + "step": 6703 + }, + { + "epoch": 0.6045903413446363, + "grad_norm": 1.9620174154778305, + "learning_rate": 1.4276995589058695e-06, + "loss": 1.0818, + "step": 6704 + }, + { + "epoch": 0.6046805248681066, + "grad_norm": 0.5930904690706867, + "learning_rate": 1.4271398016039168e-06, + "loss": 0.7858, + "step": 6705 + }, + { + "epoch": 0.6047707083915769, + "grad_norm": 1.9786304015056295, + "learning_rate": 1.4265800931815542e-06, + "loss": 1.0269, + "step": 6706 + }, + { + "epoch": 0.6048608919150471, + "grad_norm": 1.5372571086536173, + "learning_rate": 1.4260204336865406e-06, + "loss": 0.9947, + "step": 6707 + }, + { + "epoch": 0.6049510754385173, + "grad_norm": 1.4774102900552484, + "learning_rate": 1.4254608231666281e-06, + "loss": 0.9808, + "step": 6708 + }, + { + "epoch": 0.6050412589619877, + "grad_norm": 1.6271006138364636, + "learning_rate": 1.4249012616695661e-06, + "loss": 0.99, + "step": 6709 + }, + { + "epoch": 0.6051314424854579, + "grad_norm": 2.0592942032427115, + "learning_rate": 1.4243417492431e-06, + "loss": 0.9885, + "step": 6710 + }, + { + "epoch": 0.6052216260089281, + "grad_norm": 1.5748142223691646, + "learning_rate": 1.4237822859349696e-06, + "loss": 0.9647, + "step": 6711 + }, + { + "epoch": 0.6053118095323984, + "grad_norm": 1.5822073379288577, + "learning_rate": 1.423222871792912e-06, + "loss": 0.9551, + "step": 6712 + }, + { + "epoch": 0.6054019930558687, + "grad_norm": 1.8432606193084469, + "learning_rate": 1.4226635068646586e-06, + "loss": 1.0395, + "step": 6713 + }, + { + "epoch": 0.605492176579339, + "grad_norm": 0.6330118218624232, + "learning_rate": 1.4221041911979393e-06, + "loss": 0.8093, + "step": 6714 + }, + { + "epoch": 0.6055823601028092, + "grad_norm": 1.7674877501930586, + "learning_rate": 1.4215449248404765e-06, + "loss": 1.0047, + "step": 6715 + }, + { + "epoch": 0.6056725436262794, + "grad_norm": 4.248235223913595, + "learning_rate": 1.4209857078399896e-06, + "loss": 1.0054, + "step": 6716 + }, + { + "epoch": 0.6057627271497498, + "grad_norm": 1.919206953775673, + "learning_rate": 1.4204265402441955e-06, + "loss": 0.9509, + "step": 6717 + }, + { + "epoch": 0.60585291067322, + "grad_norm": 1.4358462897255075, + "learning_rate": 1.419867422100804e-06, + "loss": 0.862, + "step": 6718 + }, + { + "epoch": 0.6059430941966902, + "grad_norm": 1.5960261976200076, + "learning_rate": 1.4193083534575236e-06, + "loss": 0.9697, + "step": 6719 + }, + { + "epoch": 0.6060332777201606, + "grad_norm": 1.9236190449917903, + "learning_rate": 1.4187493343620567e-06, + "loss": 1.0058, + "step": 6720 + }, + { + "epoch": 0.6061234612436308, + "grad_norm": 1.697625265878845, + "learning_rate": 1.4181903648621006e-06, + "loss": 1.0176, + "step": 6721 + }, + { + "epoch": 0.606213644767101, + "grad_norm": 1.5325657388299692, + "learning_rate": 1.4176314450053512e-06, + "loss": 1.017, + "step": 6722 + }, + { + "epoch": 0.6063038282905713, + "grad_norm": 1.8128396077818028, + "learning_rate": 1.4170725748394977e-06, + "loss": 0.9727, + "step": 6723 + }, + { + "epoch": 0.6063940118140416, + "grad_norm": 5.23212532550138, + "learning_rate": 1.4165137544122266e-06, + "loss": 0.8956, + "step": 6724 + }, + { + "epoch": 0.6064841953375119, + "grad_norm": 1.9992799654879783, + "learning_rate": 1.4159549837712194e-06, + "loss": 1.0499, + "step": 6725 + }, + { + "epoch": 0.6065743788609821, + "grad_norm": 1.3933783715709436, + "learning_rate": 1.415396262964153e-06, + "loss": 0.9023, + "step": 6726 + }, + { + "epoch": 0.6066645623844523, + "grad_norm": 1.951673960499717, + "learning_rate": 1.4148375920387016e-06, + "loss": 1.0566, + "step": 6727 + }, + { + "epoch": 0.6067547459079227, + "grad_norm": 0.6336618774889601, + "learning_rate": 1.4142789710425325e-06, + "loss": 0.8118, + "step": 6728 + }, + { + "epoch": 0.6068449294313929, + "grad_norm": 2.064283351359427, + "learning_rate": 1.4137204000233118e-06, + "loss": 0.9244, + "step": 6729 + }, + { + "epoch": 0.6069351129548631, + "grad_norm": 7.898875781840309, + "learning_rate": 1.4131618790286987e-06, + "loss": 0.9784, + "step": 6730 + }, + { + "epoch": 0.6070252964783334, + "grad_norm": 1.5180112007240532, + "learning_rate": 1.4126034081063506e-06, + "loss": 0.9714, + "step": 6731 + }, + { + "epoch": 0.6071154800018037, + "grad_norm": 2.1599587872776733, + "learning_rate": 1.4120449873039186e-06, + "loss": 0.9682, + "step": 6732 + }, + { + "epoch": 0.6072056635252739, + "grad_norm": 1.793897533663043, + "learning_rate": 1.4114866166690494e-06, + "loss": 1.0228, + "step": 6733 + }, + { + "epoch": 0.6072958470487442, + "grad_norm": 2.684911235144784, + "learning_rate": 1.4109282962493877e-06, + "loss": 0.9425, + "step": 6734 + }, + { + "epoch": 0.6073860305722144, + "grad_norm": 3.0285747560189775, + "learning_rate": 1.4103700260925716e-06, + "loss": 0.9751, + "step": 6735 + }, + { + "epoch": 0.6074762140956848, + "grad_norm": 1.8618589847076947, + "learning_rate": 1.4098118062462364e-06, + "loss": 1.0217, + "step": 6736 + }, + { + "epoch": 0.607566397619155, + "grad_norm": 1.5028559200037386, + "learning_rate": 1.4092536367580123e-06, + "loss": 0.9811, + "step": 6737 + }, + { + "epoch": 0.6076565811426252, + "grad_norm": 2.158200031813394, + "learning_rate": 1.4086955176755248e-06, + "loss": 0.8991, + "step": 6738 + }, + { + "epoch": 0.6077467646660955, + "grad_norm": 1.7231001854992583, + "learning_rate": 1.4081374490463964e-06, + "loss": 0.9251, + "step": 6739 + }, + { + "epoch": 0.6078369481895658, + "grad_norm": 2.0297880340181367, + "learning_rate": 1.4075794309182443e-06, + "loss": 0.9988, + "step": 6740 + }, + { + "epoch": 0.607927131713036, + "grad_norm": 2.1090274100550532, + "learning_rate": 1.407021463338682e-06, + "loss": 0.9937, + "step": 6741 + }, + { + "epoch": 0.6080173152365063, + "grad_norm": 0.5902151492185675, + "learning_rate": 1.4064635463553177e-06, + "loss": 0.7579, + "step": 6742 + }, + { + "epoch": 0.6081074987599766, + "grad_norm": 2.9793756762904526, + "learning_rate": 1.4059056800157563e-06, + "loss": 1.0139, + "step": 6743 + }, + { + "epoch": 0.6081976822834468, + "grad_norm": 2.0617129818343076, + "learning_rate": 1.4053478643675982e-06, + "loss": 0.9157, + "step": 6744 + }, + { + "epoch": 0.6082878658069171, + "grad_norm": 1.6674450804947263, + "learning_rate": 1.4047900994584389e-06, + "loss": 0.8812, + "step": 6745 + }, + { + "epoch": 0.6083780493303873, + "grad_norm": 1.7986717197544237, + "learning_rate": 1.404232385335871e-06, + "loss": 0.8963, + "step": 6746 + }, + { + "epoch": 0.6084682328538576, + "grad_norm": 2.654583486727549, + "learning_rate": 1.4036747220474806e-06, + "loss": 0.9814, + "step": 6747 + }, + { + "epoch": 0.6085584163773279, + "grad_norm": 2.2070756485785856, + "learning_rate": 1.4031171096408506e-06, + "loss": 0.9315, + "step": 6748 + }, + { + "epoch": 0.6086485999007981, + "grad_norm": 1.600836389205428, + "learning_rate": 1.4025595481635607e-06, + "loss": 1.0076, + "step": 6749 + }, + { + "epoch": 0.6087387834242683, + "grad_norm": 2.489694813635833, + "learning_rate": 1.4020020376631836e-06, + "loss": 1.0032, + "step": 6750 + }, + { + "epoch": 0.6088289669477387, + "grad_norm": 1.959535926160502, + "learning_rate": 1.4014445781872908e-06, + "loss": 0.9964, + "step": 6751 + }, + { + "epoch": 0.6089191504712089, + "grad_norm": 1.8500450773160086, + "learning_rate": 1.4008871697834465e-06, + "loss": 1.045, + "step": 6752 + }, + { + "epoch": 0.6090093339946792, + "grad_norm": 1.546109944741129, + "learning_rate": 1.400329812499213e-06, + "loss": 1.015, + "step": 6753 + }, + { + "epoch": 0.6090995175181494, + "grad_norm": 1.7364508775330931, + "learning_rate": 1.3997725063821458e-06, + "loss": 0.9968, + "step": 6754 + }, + { + "epoch": 0.6091897010416197, + "grad_norm": 4.5202264803294625, + "learning_rate": 1.3992152514797978e-06, + "loss": 0.9489, + "step": 6755 + }, + { + "epoch": 0.60927988456509, + "grad_norm": 0.634770038295956, + "learning_rate": 1.398658047839718e-06, + "loss": 0.8326, + "step": 6756 + }, + { + "epoch": 0.6093700680885602, + "grad_norm": 0.6011397870617435, + "learning_rate": 1.3981008955094481e-06, + "loss": 0.8027, + "step": 6757 + }, + { + "epoch": 0.6094602516120304, + "grad_norm": 1.7686888182070453, + "learning_rate": 1.39754379453653e-06, + "loss": 0.959, + "step": 6758 + }, + { + "epoch": 0.6095504351355008, + "grad_norm": 1.8580193586501004, + "learning_rate": 1.3969867449684972e-06, + "loss": 0.9938, + "step": 6759 + }, + { + "epoch": 0.609640618658971, + "grad_norm": 1.9537841295030205, + "learning_rate": 1.396429746852879e-06, + "loss": 0.9747, + "step": 6760 + }, + { + "epoch": 0.6097308021824412, + "grad_norm": 1.8552252358707737, + "learning_rate": 1.395872800237204e-06, + "loss": 0.9601, + "step": 6761 + }, + { + "epoch": 0.6098209857059115, + "grad_norm": 0.8476462297682339, + "learning_rate": 1.3953159051689918e-06, + "loss": 0.8344, + "step": 6762 + }, + { + "epoch": 0.6099111692293818, + "grad_norm": 3.8953399889818017, + "learning_rate": 1.3947590616957618e-06, + "loss": 1.0297, + "step": 6763 + }, + { + "epoch": 0.610001352752852, + "grad_norm": 1.7205873552188762, + "learning_rate": 1.3942022698650258e-06, + "loss": 1.0145, + "step": 6764 + }, + { + "epoch": 0.6100915362763223, + "grad_norm": 1.62282930791133, + "learning_rate": 1.3936455297242917e-06, + "loss": 1.0119, + "step": 6765 + }, + { + "epoch": 0.6101817197997926, + "grad_norm": 1.606536361757735, + "learning_rate": 1.3930888413210652e-06, + "loss": 1.0008, + "step": 6766 + }, + { + "epoch": 0.6102719033232629, + "grad_norm": 1.7167682667020543, + "learning_rate": 1.392532204702845e-06, + "loss": 0.9452, + "step": 6767 + }, + { + "epoch": 0.6103620868467331, + "grad_norm": 2.130264269056322, + "learning_rate": 1.3919756199171266e-06, + "loss": 0.9132, + "step": 6768 + }, + { + "epoch": 0.6104522703702033, + "grad_norm": 1.46958445118594, + "learning_rate": 1.3914190870114009e-06, + "loss": 1.0542, + "step": 6769 + }, + { + "epoch": 0.6105424538936737, + "grad_norm": 1.6655816679107838, + "learning_rate": 1.3908626060331541e-06, + "loss": 1.0346, + "step": 6770 + }, + { + "epoch": 0.6106326374171439, + "grad_norm": 1.6525401185860442, + "learning_rate": 1.3903061770298693e-06, + "loss": 1.0145, + "step": 6771 + }, + { + "epoch": 0.6107228209406141, + "grad_norm": 2.2842914738237834, + "learning_rate": 1.3897498000490223e-06, + "loss": 0.9596, + "step": 6772 + }, + { + "epoch": 0.6108130044640844, + "grad_norm": 1.5019848724700613, + "learning_rate": 1.3891934751380879e-06, + "loss": 1.0501, + "step": 6773 + }, + { + "epoch": 0.6109031879875547, + "grad_norm": 1.545761478258679, + "learning_rate": 1.3886372023445334e-06, + "loss": 1.002, + "step": 6774 + }, + { + "epoch": 0.610993371511025, + "grad_norm": 1.6313225916353076, + "learning_rate": 1.3880809817158246e-06, + "loss": 1.0533, + "step": 6775 + }, + { + "epoch": 0.6110835550344952, + "grad_norm": 1.72799740366829, + "learning_rate": 1.3875248132994206e-06, + "loss": 0.9742, + "step": 6776 + }, + { + "epoch": 0.6111737385579654, + "grad_norm": 1.9996777513292927, + "learning_rate": 1.386968697142776e-06, + "loss": 0.9053, + "step": 6777 + }, + { + "epoch": 0.6112639220814358, + "grad_norm": 1.7979451185725341, + "learning_rate": 1.386412633293343e-06, + "loss": 1.0298, + "step": 6778 + }, + { + "epoch": 0.611354105604906, + "grad_norm": 1.551029598615268, + "learning_rate": 1.3858566217985672e-06, + "loss": 0.9601, + "step": 6779 + }, + { + "epoch": 0.6114442891283762, + "grad_norm": 1.7434987834054276, + "learning_rate": 1.3853006627058905e-06, + "loss": 1.066, + "step": 6780 + }, + { + "epoch": 0.6115344726518465, + "grad_norm": 2.107512039928504, + "learning_rate": 1.3847447560627512e-06, + "loss": 0.8709, + "step": 6781 + }, + { + "epoch": 0.6116246561753168, + "grad_norm": 2.016073014496319, + "learning_rate": 1.3841889019165812e-06, + "loss": 1.0299, + "step": 6782 + }, + { + "epoch": 0.611714839698787, + "grad_norm": 2.371063483609084, + "learning_rate": 1.3836331003148106e-06, + "loss": 0.9388, + "step": 6783 + }, + { + "epoch": 0.6118050232222573, + "grad_norm": 1.6745817430875767, + "learning_rate": 1.3830773513048612e-06, + "loss": 1.0225, + "step": 6784 + }, + { + "epoch": 0.6118952067457275, + "grad_norm": 1.6924719855780634, + "learning_rate": 1.382521654934155e-06, + "loss": 0.9996, + "step": 6785 + }, + { + "epoch": 0.6119853902691978, + "grad_norm": 2.0774440940560166, + "learning_rate": 1.3819660112501057e-06, + "loss": 0.9303, + "step": 6786 + }, + { + "epoch": 0.6120755737926681, + "grad_norm": 1.4843656037685746, + "learning_rate": 1.3814104203001234e-06, + "loss": 0.9443, + "step": 6787 + }, + { + "epoch": 0.6121657573161383, + "grad_norm": 1.5584448262821342, + "learning_rate": 1.3808548821316156e-06, + "loss": 1.0226, + "step": 6788 + }, + { + "epoch": 0.6122559408396087, + "grad_norm": 1.3673421281838305, + "learning_rate": 1.3802993967919824e-06, + "loss": 0.9162, + "step": 6789 + }, + { + "epoch": 0.6123461243630789, + "grad_norm": 1.6334173369052642, + "learning_rate": 1.3797439643286227e-06, + "loss": 0.9417, + "step": 6790 + }, + { + "epoch": 0.6124363078865491, + "grad_norm": 1.517721868946439, + "learning_rate": 1.3791885847889277e-06, + "loss": 0.9682, + "step": 6791 + }, + { + "epoch": 0.6125264914100194, + "grad_norm": 1.7135453228974298, + "learning_rate": 1.3786332582202853e-06, + "loss": 1.0185, + "step": 6792 + }, + { + "epoch": 0.6126166749334897, + "grad_norm": 1.984530825766934, + "learning_rate": 1.3780779846700799e-06, + "loss": 0.9844, + "step": 6793 + }, + { + "epoch": 0.6127068584569599, + "grad_norm": 1.6683417803194502, + "learning_rate": 1.3775227641856899e-06, + "loss": 0.9963, + "step": 6794 + }, + { + "epoch": 0.6127970419804302, + "grad_norm": 1.6903356110538599, + "learning_rate": 1.37696759681449e-06, + "loss": 0.9842, + "step": 6795 + }, + { + "epoch": 0.6128872255039004, + "grad_norm": 0.6933523473249152, + "learning_rate": 1.37641248260385e-06, + "loss": 0.8432, + "step": 6796 + }, + { + "epoch": 0.6129774090273707, + "grad_norm": 1.7980991340184935, + "learning_rate": 1.375857421601136e-06, + "loss": 1.0353, + "step": 6797 + }, + { + "epoch": 0.613067592550841, + "grad_norm": 1.576789517499123, + "learning_rate": 1.3753024138537082e-06, + "loss": 0.9942, + "step": 6798 + }, + { + "epoch": 0.6131577760743112, + "grad_norm": 1.4644745628515565, + "learning_rate": 1.3747474594089221e-06, + "loss": 1.0502, + "step": 6799 + }, + { + "epoch": 0.6132479595977814, + "grad_norm": 1.686460598745193, + "learning_rate": 1.374192558314131e-06, + "loss": 0.9639, + "step": 6800 + }, + { + "epoch": 0.6133381431212518, + "grad_norm": 1.6984640615960425, + "learning_rate": 1.373637710616681e-06, + "loss": 1.047, + "step": 6801 + }, + { + "epoch": 0.613428326644722, + "grad_norm": 2.461296999876354, + "learning_rate": 1.373082916363916e-06, + "loss": 1.0947, + "step": 6802 + }, + { + "epoch": 0.6135185101681923, + "grad_norm": 1.3852308000446694, + "learning_rate": 1.3725281756031732e-06, + "loss": 1.0283, + "step": 6803 + }, + { + "epoch": 0.6136086936916625, + "grad_norm": 1.3816542435207573, + "learning_rate": 1.3719734883817858e-06, + "loss": 0.9848, + "step": 6804 + }, + { + "epoch": 0.6136988772151328, + "grad_norm": 2.194391023278033, + "learning_rate": 1.371418854747084e-06, + "loss": 1.0965, + "step": 6805 + }, + { + "epoch": 0.6137890607386031, + "grad_norm": 1.8131161004380283, + "learning_rate": 1.3708642747463905e-06, + "loss": 1.0209, + "step": 6806 + }, + { + "epoch": 0.6138792442620733, + "grad_norm": 1.5532127555140671, + "learning_rate": 1.370309748427027e-06, + "loss": 1.0188, + "step": 6807 + }, + { + "epoch": 0.6139694277855435, + "grad_norm": 1.9111862058203202, + "learning_rate": 1.3697552758363079e-06, + "loss": 1.0582, + "step": 6808 + }, + { + "epoch": 0.6140596113090139, + "grad_norm": 2.546499774650475, + "learning_rate": 1.3692008570215432e-06, + "loss": 0.9923, + "step": 6809 + }, + { + "epoch": 0.6141497948324841, + "grad_norm": 1.8583762427651922, + "learning_rate": 1.3686464920300398e-06, + "loss": 0.9418, + "step": 6810 + }, + { + "epoch": 0.6142399783559543, + "grad_norm": 1.7920114791621662, + "learning_rate": 1.3680921809090985e-06, + "loss": 0.9615, + "step": 6811 + }, + { + "epoch": 0.6143301618794246, + "grad_norm": 1.5962877451616904, + "learning_rate": 1.3675379237060175e-06, + "loss": 1.0282, + "step": 6812 + }, + { + "epoch": 0.6144203454028949, + "grad_norm": 1.5332931589309218, + "learning_rate": 1.366983720468088e-06, + "loss": 0.9818, + "step": 6813 + }, + { + "epoch": 0.6145105289263652, + "grad_norm": 1.333381036064037, + "learning_rate": 1.3664295712425972e-06, + "loss": 0.9941, + "step": 6814 + }, + { + "epoch": 0.6146007124498354, + "grad_norm": 1.6472782198754852, + "learning_rate": 1.3658754760768296e-06, + "loss": 0.9214, + "step": 6815 + }, + { + "epoch": 0.6146908959733057, + "grad_norm": 1.6845149022816364, + "learning_rate": 1.3653214350180621e-06, + "loss": 1.0225, + "step": 6816 + }, + { + "epoch": 0.614781079496776, + "grad_norm": 1.599917278535593, + "learning_rate": 1.3647674481135703e-06, + "loss": 1.0537, + "step": 6817 + }, + { + "epoch": 0.6148712630202462, + "grad_norm": 1.3302471484817027, + "learning_rate": 1.3642135154106217e-06, + "loss": 0.8863, + "step": 6818 + }, + { + "epoch": 0.6149614465437164, + "grad_norm": 2.131921714912826, + "learning_rate": 1.363659636956482e-06, + "loss": 0.9347, + "step": 6819 + }, + { + "epoch": 0.6150516300671868, + "grad_norm": 1.3621232325530261, + "learning_rate": 1.3631058127984112e-06, + "loss": 0.9555, + "step": 6820 + }, + { + "epoch": 0.615141813590657, + "grad_norm": 2.4907655874896197, + "learning_rate": 1.3625520429836632e-06, + "loss": 0.9638, + "step": 6821 + }, + { + "epoch": 0.6152319971141272, + "grad_norm": 1.715909865446335, + "learning_rate": 1.361998327559491e-06, + "loss": 0.8941, + "step": 6822 + }, + { + "epoch": 0.6153221806375975, + "grad_norm": 0.6437953794095509, + "learning_rate": 1.3614446665731385e-06, + "loss": 0.8183, + "step": 6823 + }, + { + "epoch": 0.6154123641610678, + "grad_norm": 2.0271406421094333, + "learning_rate": 1.3608910600718484e-06, + "loss": 0.9961, + "step": 6824 + }, + { + "epoch": 0.615502547684538, + "grad_norm": 3.1752783645793823, + "learning_rate": 1.360337508102857e-06, + "loss": 0.9894, + "step": 6825 + }, + { + "epoch": 0.6155927312080083, + "grad_norm": 2.2162373244110825, + "learning_rate": 1.3597840107133962e-06, + "loss": 1.0837, + "step": 6826 + }, + { + "epoch": 0.6156829147314785, + "grad_norm": 1.8045316026076048, + "learning_rate": 1.3592305679506944e-06, + "loss": 0.9968, + "step": 6827 + }, + { + "epoch": 0.6157730982549489, + "grad_norm": 1.613573098743597, + "learning_rate": 1.3586771798619726e-06, + "loss": 0.8922, + "step": 6828 + }, + { + "epoch": 0.6158632817784191, + "grad_norm": 1.4628062001047648, + "learning_rate": 1.358123846494451e-06, + "loss": 1.0528, + "step": 6829 + }, + { + "epoch": 0.6159534653018893, + "grad_norm": 5.502082777137637, + "learning_rate": 1.3575705678953422e-06, + "loss": 0.9954, + "step": 6830 + }, + { + "epoch": 0.6160436488253596, + "grad_norm": 2.0803862007773377, + "learning_rate": 1.357017344111854e-06, + "loss": 1.0273, + "step": 6831 + }, + { + "epoch": 0.6161338323488299, + "grad_norm": 1.741296377776152, + "learning_rate": 1.356464175191192e-06, + "loss": 1.0004, + "step": 6832 + }, + { + "epoch": 0.6162240158723001, + "grad_norm": 2.0737116709288994, + "learning_rate": 1.3559110611805542e-06, + "loss": 1.0115, + "step": 6833 + }, + { + "epoch": 0.6163141993957704, + "grad_norm": 1.8934168733492287, + "learning_rate": 1.3553580021271372e-06, + "loss": 1.0278, + "step": 6834 + }, + { + "epoch": 0.6164043829192406, + "grad_norm": 1.591043830095217, + "learning_rate": 1.3548049980781297e-06, + "loss": 0.9307, + "step": 6835 + }, + { + "epoch": 0.616494566442711, + "grad_norm": 2.072517918968273, + "learning_rate": 1.3542520490807166e-06, + "loss": 0.979, + "step": 6836 + }, + { + "epoch": 0.6165847499661812, + "grad_norm": 1.6737240699998754, + "learning_rate": 1.3536991551820802e-06, + "loss": 0.9753, + "step": 6837 + }, + { + "epoch": 0.6166749334896514, + "grad_norm": 1.4141241239323983, + "learning_rate": 1.3531463164293952e-06, + "loss": 0.9573, + "step": 6838 + }, + { + "epoch": 0.6167651170131218, + "grad_norm": 2.6986039119756846, + "learning_rate": 1.3525935328698332e-06, + "loss": 0.8938, + "step": 6839 + }, + { + "epoch": 0.616855300536592, + "grad_norm": 3.08575516174127, + "learning_rate": 1.3520408045505605e-06, + "loss": 0.9239, + "step": 6840 + }, + { + "epoch": 0.6169454840600622, + "grad_norm": 1.7495705213787023, + "learning_rate": 1.3514881315187396e-06, + "loss": 0.9559, + "step": 6841 + }, + { + "epoch": 0.6170356675835325, + "grad_norm": 1.4609786247136256, + "learning_rate": 1.3509355138215273e-06, + "loss": 1.0424, + "step": 6842 + }, + { + "epoch": 0.6171258511070028, + "grad_norm": 1.7021994919596988, + "learning_rate": 1.350382951506075e-06, + "loss": 1.0689, + "step": 6843 + }, + { + "epoch": 0.617216034630473, + "grad_norm": 0.8270708248111798, + "learning_rate": 1.349830444619532e-06, + "loss": 0.8041, + "step": 6844 + }, + { + "epoch": 0.6173062181539433, + "grad_norm": 1.4752717204276735, + "learning_rate": 1.3492779932090397e-06, + "loss": 1.0007, + "step": 6845 + }, + { + "epoch": 0.6173964016774135, + "grad_norm": 1.5067564275188963, + "learning_rate": 1.3487255973217377e-06, + "loss": 1.0133, + "step": 6846 + }, + { + "epoch": 0.6174865852008838, + "grad_norm": 2.0281803117683292, + "learning_rate": 1.3481732570047592e-06, + "loss": 1.0562, + "step": 6847 + }, + { + "epoch": 0.6175767687243541, + "grad_norm": 2.1571950638190973, + "learning_rate": 1.3476209723052314e-06, + "loss": 0.9247, + "step": 6848 + }, + { + "epoch": 0.6176669522478243, + "grad_norm": 3.2840648384895386, + "learning_rate": 1.3470687432702806e-06, + "loss": 1.0113, + "step": 6849 + }, + { + "epoch": 0.6177571357712945, + "grad_norm": 1.7156215980370983, + "learning_rate": 1.346516569947024e-06, + "loss": 0.9149, + "step": 6850 + }, + { + "epoch": 0.6178473192947649, + "grad_norm": 2.1093180330849273, + "learning_rate": 1.3459644523825774e-06, + "loss": 0.9754, + "step": 6851 + }, + { + "epoch": 0.6179375028182351, + "grad_norm": 1.681055195239808, + "learning_rate": 1.34541239062405e-06, + "loss": 1.0292, + "step": 6852 + }, + { + "epoch": 0.6180276863417054, + "grad_norm": 1.4446330354933634, + "learning_rate": 1.3448603847185464e-06, + "loss": 0.9869, + "step": 6853 + }, + { + "epoch": 0.6181178698651756, + "grad_norm": 2.1074223275312303, + "learning_rate": 1.344308434713168e-06, + "loss": 0.9282, + "step": 6854 + }, + { + "epoch": 0.6182080533886459, + "grad_norm": 2.132333778620229, + "learning_rate": 1.3437565406550083e-06, + "loss": 0.9178, + "step": 6855 + }, + { + "epoch": 0.6182982369121162, + "grad_norm": 1.9132398433418207, + "learning_rate": 1.34320470259116e-06, + "loss": 0.9704, + "step": 6856 + }, + { + "epoch": 0.6183884204355864, + "grad_norm": 16.426450547452138, + "learning_rate": 1.3426529205687078e-06, + "loss": 0.8932, + "step": 6857 + }, + { + "epoch": 0.6184786039590566, + "grad_norm": 2.4546771150657056, + "learning_rate": 1.3421011946347323e-06, + "loss": 1.0449, + "step": 6858 + }, + { + "epoch": 0.618568787482527, + "grad_norm": 1.5120278370227318, + "learning_rate": 1.3415495248363113e-06, + "loss": 0.9479, + "step": 6859 + }, + { + "epoch": 0.6186589710059972, + "grad_norm": 2.1364485937746203, + "learning_rate": 1.3409979112205148e-06, + "loss": 1.0103, + "step": 6860 + }, + { + "epoch": 0.6187491545294674, + "grad_norm": 2.049247473320189, + "learning_rate": 1.3404463538344107e-06, + "loss": 1.0131, + "step": 6861 + }, + { + "epoch": 0.6188393380529378, + "grad_norm": 1.8628510112394077, + "learning_rate": 1.3398948527250602e-06, + "loss": 1.0777, + "step": 6862 + }, + { + "epoch": 0.618929521576408, + "grad_norm": 0.7537737319778233, + "learning_rate": 1.3393434079395207e-06, + "loss": 0.8286, + "step": 6863 + }, + { + "epoch": 0.6190197050998782, + "grad_norm": 1.575221572912497, + "learning_rate": 1.3387920195248449e-06, + "loss": 0.9729, + "step": 6864 + }, + { + "epoch": 0.6191098886233485, + "grad_norm": 3.771654801787626, + "learning_rate": 1.3382406875280791e-06, + "loss": 1.003, + "step": 6865 + }, + { + "epoch": 0.6192000721468188, + "grad_norm": 1.5972985051316644, + "learning_rate": 1.3376894119962672e-06, + "loss": 1.0098, + "step": 6866 + }, + { + "epoch": 0.6192902556702891, + "grad_norm": 2.4425895569858813, + "learning_rate": 1.3371381929764464e-06, + "loss": 1.0151, + "step": 6867 + }, + { + "epoch": 0.6193804391937593, + "grad_norm": 0.6416004941516267, + "learning_rate": 1.3365870305156506e-06, + "loss": 0.8068, + "step": 6868 + }, + { + "epoch": 0.6194706227172295, + "grad_norm": 2.303983729609685, + "learning_rate": 1.3360359246609073e-06, + "loss": 0.9924, + "step": 6869 + }, + { + "epoch": 0.6195608062406999, + "grad_norm": 1.8584592471494847, + "learning_rate": 1.3354848754592387e-06, + "loss": 0.9825, + "step": 6870 + }, + { + "epoch": 0.6196509897641701, + "grad_norm": 1.8319279167786457, + "learning_rate": 1.334933882957666e-06, + "loss": 1.0752, + "step": 6871 + }, + { + "epoch": 0.6197411732876403, + "grad_norm": 1.4779051474047114, + "learning_rate": 1.3343829472032004e-06, + "loss": 0.9493, + "step": 6872 + }, + { + "epoch": 0.6198313568111106, + "grad_norm": 1.9606979041366905, + "learning_rate": 1.3338320682428527e-06, + "loss": 0.974, + "step": 6873 + }, + { + "epoch": 0.6199215403345809, + "grad_norm": 2.1009316330634467, + "learning_rate": 1.3332812461236263e-06, + "loss": 0.9489, + "step": 6874 + }, + { + "epoch": 0.6200117238580511, + "grad_norm": 1.3797407353845785, + "learning_rate": 1.3327304808925192e-06, + "loss": 0.9525, + "step": 6875 + }, + { + "epoch": 0.6201019073815214, + "grad_norm": 1.705007929629683, + "learning_rate": 1.332179772596528e-06, + "loss": 0.8475, + "step": 6876 + }, + { + "epoch": 0.6201920909049916, + "grad_norm": 1.7884880376324563, + "learning_rate": 1.3316291212826402e-06, + "loss": 0.9094, + "step": 6877 + }, + { + "epoch": 0.620282274428462, + "grad_norm": 10.16043295771315, + "learning_rate": 1.3310785269978413e-06, + "loss": 1.0349, + "step": 6878 + }, + { + "epoch": 0.6203724579519322, + "grad_norm": 1.8212618991984106, + "learning_rate": 1.3305279897891111e-06, + "loss": 1.028, + "step": 6879 + }, + { + "epoch": 0.6204626414754024, + "grad_norm": 1.712633520336937, + "learning_rate": 1.329977509703424e-06, + "loss": 0.8533, + "step": 6880 + }, + { + "epoch": 0.6205528249988727, + "grad_norm": 2.5624356145445923, + "learning_rate": 1.32942708678775e-06, + "loss": 1.0751, + "step": 6881 + }, + { + "epoch": 0.620643008522343, + "grad_norm": 1.781901926233929, + "learning_rate": 1.3288767210890548e-06, + "loss": 0.9997, + "step": 6882 + }, + { + "epoch": 0.6207331920458132, + "grad_norm": 2.0896301730818077, + "learning_rate": 1.3283264126542986e-06, + "loss": 0.9486, + "step": 6883 + }, + { + "epoch": 0.6208233755692835, + "grad_norm": 1.610155823553437, + "learning_rate": 1.3277761615304356e-06, + "loss": 0.9873, + "step": 6884 + }, + { + "epoch": 0.6209135590927538, + "grad_norm": 1.4249131996086875, + "learning_rate": 1.3272259677644178e-06, + "loss": 1.088, + "step": 6885 + }, + { + "epoch": 0.621003742616224, + "grad_norm": 2.4636500350204273, + "learning_rate": 1.32667583140319e-06, + "loss": 0.8574, + "step": 6886 + }, + { + "epoch": 0.6210939261396943, + "grad_norm": 2.645163974546756, + "learning_rate": 1.3261257524936924e-06, + "loss": 1.0288, + "step": 6887 + }, + { + "epoch": 0.6211841096631645, + "grad_norm": 1.636824356350204, + "learning_rate": 1.3255757310828619e-06, + "loss": 1.0166, + "step": 6888 + }, + { + "epoch": 0.6212742931866349, + "grad_norm": 2.057458206067466, + "learning_rate": 1.3250257672176282e-06, + "loss": 1.008, + "step": 6889 + }, + { + "epoch": 0.6213644767101051, + "grad_norm": 1.7422752187225352, + "learning_rate": 1.3244758609449183e-06, + "loss": 0.8924, + "step": 6890 + }, + { + "epoch": 0.6214546602335753, + "grad_norm": 1.8004640643287182, + "learning_rate": 1.323926012311653e-06, + "loss": 1.0029, + "step": 6891 + }, + { + "epoch": 0.6215448437570456, + "grad_norm": 13.38925238643941, + "learning_rate": 1.3233762213647476e-06, + "loss": 1.0117, + "step": 6892 + }, + { + "epoch": 0.6216350272805159, + "grad_norm": 1.7602430705717116, + "learning_rate": 1.3228264881511137e-06, + "loss": 0.9622, + "step": 6893 + }, + { + "epoch": 0.6217252108039861, + "grad_norm": 2.031569939872428, + "learning_rate": 1.322276812717658e-06, + "loss": 1.0717, + "step": 6894 + }, + { + "epoch": 0.6218153943274564, + "grad_norm": 1.771858688898379, + "learning_rate": 1.3217271951112818e-06, + "loss": 0.8947, + "step": 6895 + }, + { + "epoch": 0.6219055778509266, + "grad_norm": 2.0231022501819234, + "learning_rate": 1.321177635378881e-06, + "loss": 0.9632, + "step": 6896 + }, + { + "epoch": 0.6219957613743969, + "grad_norm": 2.245060903831801, + "learning_rate": 1.3206281335673475e-06, + "loss": 0.8816, + "step": 6897 + }, + { + "epoch": 0.6220859448978672, + "grad_norm": 2.176401156150372, + "learning_rate": 1.3200786897235677e-06, + "loss": 0.9654, + "step": 6898 + }, + { + "epoch": 0.6221761284213374, + "grad_norm": 1.958961494289359, + "learning_rate": 1.3195293038944227e-06, + "loss": 1.0028, + "step": 6899 + }, + { + "epoch": 0.6222663119448076, + "grad_norm": 1.3165975822158813, + "learning_rate": 1.3189799761267902e-06, + "loss": 0.9806, + "step": 6900 + }, + { + "epoch": 0.622356495468278, + "grad_norm": 2.1731562505827835, + "learning_rate": 1.3184307064675412e-06, + "loss": 0.9775, + "step": 6901 + }, + { + "epoch": 0.6224466789917482, + "grad_norm": 0.686115496400915, + "learning_rate": 1.3178814949635416e-06, + "loss": 0.8618, + "step": 6902 + }, + { + "epoch": 0.6225368625152184, + "grad_norm": 2.300826740948569, + "learning_rate": 1.3173323416616549e-06, + "loss": 0.9345, + "step": 6903 + }, + { + "epoch": 0.6226270460386887, + "grad_norm": 1.5739762566701188, + "learning_rate": 1.3167832466087361e-06, + "loss": 1.0413, + "step": 6904 + }, + { + "epoch": 0.622717229562159, + "grad_norm": 1.5528061568264406, + "learning_rate": 1.3162342098516388e-06, + "loss": 0.9365, + "step": 6905 + }, + { + "epoch": 0.6228074130856293, + "grad_norm": 1.8332538876515045, + "learning_rate": 1.3156852314372086e-06, + "loss": 0.9062, + "step": 6906 + }, + { + "epoch": 0.6228975966090995, + "grad_norm": 0.682290082060937, + "learning_rate": 1.3151363114122882e-06, + "loss": 0.7969, + "step": 6907 + }, + { + "epoch": 0.6229877801325698, + "grad_norm": 1.4388784350911856, + "learning_rate": 1.3145874498237133e-06, + "loss": 0.9996, + "step": 6908 + }, + { + "epoch": 0.6230779636560401, + "grad_norm": 3.2793064719372675, + "learning_rate": 1.3140386467183166e-06, + "loss": 1.0364, + "step": 6909 + }, + { + "epoch": 0.6231681471795103, + "grad_norm": 1.5875741565835524, + "learning_rate": 1.3134899021429258e-06, + "loss": 1.0287, + "step": 6910 + }, + { + "epoch": 0.6232583307029805, + "grad_norm": 2.0961158253647927, + "learning_rate": 1.3129412161443604e-06, + "loss": 0.969, + "step": 6911 + }, + { + "epoch": 0.6233485142264509, + "grad_norm": 2.0919508534524653, + "learning_rate": 1.3123925887694402e-06, + "loss": 1.0003, + "step": 6912 + }, + { + "epoch": 0.6234386977499211, + "grad_norm": 1.5511212108429595, + "learning_rate": 1.3118440200649756e-06, + "loss": 0.9982, + "step": 6913 + }, + { + "epoch": 0.6235288812733913, + "grad_norm": 2.2505509443739897, + "learning_rate": 1.3112955100777727e-06, + "loss": 0.946, + "step": 6914 + }, + { + "epoch": 0.6236190647968616, + "grad_norm": 1.4617158988627157, + "learning_rate": 1.3107470588546353e-06, + "loss": 1.0229, + "step": 6915 + }, + { + "epoch": 0.6237092483203319, + "grad_norm": 1.7381357196366194, + "learning_rate": 1.3101986664423583e-06, + "loss": 0.9558, + "step": 6916 + }, + { + "epoch": 0.6237994318438022, + "grad_norm": 1.6796393099237228, + "learning_rate": 1.3096503328877358e-06, + "loss": 0.9596, + "step": 6917 + }, + { + "epoch": 0.6238896153672724, + "grad_norm": 2.0017643537143712, + "learning_rate": 1.309102058237553e-06, + "loss": 1.0035, + "step": 6918 + }, + { + "epoch": 0.6239797988907426, + "grad_norm": 2.886640782522865, + "learning_rate": 1.3085538425385917e-06, + "loss": 1.038, + "step": 6919 + }, + { + "epoch": 0.624069982414213, + "grad_norm": 2.6889121315567213, + "learning_rate": 1.3080056858376298e-06, + "loss": 1.0147, + "step": 6920 + }, + { + "epoch": 0.6241601659376832, + "grad_norm": 1.7318067350944522, + "learning_rate": 1.3074575881814383e-06, + "loss": 0.974, + "step": 6921 + }, + { + "epoch": 0.6242503494611534, + "grad_norm": 2.9623503120044252, + "learning_rate": 1.3069095496167841e-06, + "loss": 0.9106, + "step": 6922 + }, + { + "epoch": 0.6243405329846237, + "grad_norm": 1.4897724200547908, + "learning_rate": 1.3063615701904285e-06, + "loss": 0.897, + "step": 6923 + }, + { + "epoch": 0.624430716508094, + "grad_norm": 2.3333914159175975, + "learning_rate": 1.3058136499491283e-06, + "loss": 1.0197, + "step": 6924 + }, + { + "epoch": 0.6245209000315642, + "grad_norm": 1.895440455659401, + "learning_rate": 1.3052657889396357e-06, + "loss": 0.9715, + "step": 6925 + }, + { + "epoch": 0.6246110835550345, + "grad_norm": 2.043611560297383, + "learning_rate": 1.304717987208696e-06, + "loss": 0.9694, + "step": 6926 + }, + { + "epoch": 0.6247012670785047, + "grad_norm": 1.694141880322555, + "learning_rate": 1.304170244803052e-06, + "loss": 0.9768, + "step": 6927 + }, + { + "epoch": 0.624791450601975, + "grad_norm": 1.7561523057366675, + "learning_rate": 1.3036225617694387e-06, + "loss": 0.955, + "step": 6928 + }, + { + "epoch": 0.6248816341254453, + "grad_norm": 1.5443227500830097, + "learning_rate": 1.3030749381545892e-06, + "loss": 1.0179, + "step": 6929 + }, + { + "epoch": 0.6249718176489155, + "grad_norm": 1.905155868461406, + "learning_rate": 1.3025273740052285e-06, + "loss": 0.9255, + "step": 6930 + }, + { + "epoch": 0.6250620011723858, + "grad_norm": 1.8213780112956035, + "learning_rate": 1.3019798693680774e-06, + "loss": 1.0326, + "step": 6931 + }, + { + "epoch": 0.6251521846958561, + "grad_norm": 2.021684159436171, + "learning_rate": 1.3014324242898536e-06, + "loss": 1.1027, + "step": 6932 + }, + { + "epoch": 0.6252423682193263, + "grad_norm": 1.837052805972755, + "learning_rate": 1.3008850388172668e-06, + "loss": 1.0425, + "step": 6933 + }, + { + "epoch": 0.6253325517427966, + "grad_norm": 1.5859448662948539, + "learning_rate": 1.3003377129970233e-06, + "loss": 0.9441, + "step": 6934 + }, + { + "epoch": 0.6254227352662669, + "grad_norm": 1.422043172092422, + "learning_rate": 1.2997904468758243e-06, + "loss": 0.9028, + "step": 6935 + }, + { + "epoch": 0.6255129187897371, + "grad_norm": 1.4710995974351058, + "learning_rate": 1.299243240500365e-06, + "loss": 1.0034, + "step": 6936 + }, + { + "epoch": 0.6256031023132074, + "grad_norm": 1.8179136522810038, + "learning_rate": 1.2986960939173368e-06, + "loss": 1.0036, + "step": 6937 + }, + { + "epoch": 0.6256932858366776, + "grad_norm": 1.954605822911201, + "learning_rate": 1.298149007173424e-06, + "loss": 1.0288, + "step": 6938 + }, + { + "epoch": 0.625783469360148, + "grad_norm": 2.131526908896153, + "learning_rate": 1.2976019803153087e-06, + "loss": 1.017, + "step": 6939 + }, + { + "epoch": 0.6258736528836182, + "grad_norm": 1.855659619602979, + "learning_rate": 1.2970550133896652e-06, + "loss": 0.9342, + "step": 6940 + }, + { + "epoch": 0.6259638364070884, + "grad_norm": 1.6594009656006081, + "learning_rate": 1.2965081064431634e-06, + "loss": 1.0643, + "step": 6941 + }, + { + "epoch": 0.6260540199305586, + "grad_norm": 3.0956800201227685, + "learning_rate": 1.2959612595224698e-06, + "loss": 1.0101, + "step": 6942 + }, + { + "epoch": 0.626144203454029, + "grad_norm": 1.9906319323169004, + "learning_rate": 1.2954144726742424e-06, + "loss": 0.9981, + "step": 6943 + }, + { + "epoch": 0.6262343869774992, + "grad_norm": 1.4464891747321125, + "learning_rate": 1.2948677459451385e-06, + "loss": 0.9813, + "step": 6944 + }, + { + "epoch": 0.6263245705009695, + "grad_norm": 1.8756889349143964, + "learning_rate": 1.2943210793818064e-06, + "loss": 1.0344, + "step": 6945 + }, + { + "epoch": 0.6264147540244397, + "grad_norm": 1.8209601956771935, + "learning_rate": 1.2937744730308899e-06, + "loss": 0.9662, + "step": 6946 + }, + { + "epoch": 0.62650493754791, + "grad_norm": 2.0705619229911765, + "learning_rate": 1.2932279269390305e-06, + "loss": 0.9827, + "step": 6947 + }, + { + "epoch": 0.6265951210713803, + "grad_norm": 1.8976413014790174, + "learning_rate": 1.292681441152861e-06, + "loss": 0.89, + "step": 6948 + }, + { + "epoch": 0.6266853045948505, + "grad_norm": 1.7733197278143793, + "learning_rate": 1.292135015719011e-06, + "loss": 0.9977, + "step": 6949 + }, + { + "epoch": 0.6267754881183207, + "grad_norm": 3.0428538857793743, + "learning_rate": 1.2915886506841046e-06, + "loss": 0.9446, + "step": 6950 + }, + { + "epoch": 0.6268656716417911, + "grad_norm": 1.5486960432923167, + "learning_rate": 1.2910423460947613e-06, + "loss": 0.9223, + "step": 6951 + }, + { + "epoch": 0.6269558551652613, + "grad_norm": 1.9550124547029932, + "learning_rate": 1.290496101997594e-06, + "loss": 0.9517, + "step": 6952 + }, + { + "epoch": 0.6270460386887315, + "grad_norm": 0.7172320964968095, + "learning_rate": 1.2899499184392105e-06, + "loss": 0.8184, + "step": 6953 + }, + { + "epoch": 0.6271362222122018, + "grad_norm": 2.1093874330507427, + "learning_rate": 1.289403795466216e-06, + "loss": 0.9635, + "step": 6954 + }, + { + "epoch": 0.6272264057356721, + "grad_norm": 1.911475414508166, + "learning_rate": 1.288857733125207e-06, + "loss": 0.9981, + "step": 6955 + }, + { + "epoch": 0.6273165892591424, + "grad_norm": 1.725796836136789, + "learning_rate": 1.2883117314627785e-06, + "loss": 0.9795, + "step": 6956 + }, + { + "epoch": 0.6274067727826126, + "grad_norm": 1.9045130210865988, + "learning_rate": 1.2877657905255168e-06, + "loss": 0.9223, + "step": 6957 + }, + { + "epoch": 0.6274969563060829, + "grad_norm": 1.472950711142377, + "learning_rate": 1.2872199103600046e-06, + "loss": 1.0478, + "step": 6958 + }, + { + "epoch": 0.6275871398295532, + "grad_norm": 1.4606058488862699, + "learning_rate": 1.286674091012821e-06, + "loss": 1.0536, + "step": 6959 + }, + { + "epoch": 0.6276773233530234, + "grad_norm": 1.573459413711097, + "learning_rate": 1.2861283325305356e-06, + "loss": 1.0451, + "step": 6960 + }, + { + "epoch": 0.6277675068764936, + "grad_norm": 1.36149322515645, + "learning_rate": 1.2855826349597185e-06, + "loss": 0.9842, + "step": 6961 + }, + { + "epoch": 0.627857690399964, + "grad_norm": 1.5675235102835772, + "learning_rate": 1.2850369983469302e-06, + "loss": 0.8433, + "step": 6962 + }, + { + "epoch": 0.6279478739234342, + "grad_norm": 1.565697640506048, + "learning_rate": 1.2844914227387266e-06, + "loss": 1.0123, + "step": 6963 + }, + { + "epoch": 0.6280380574469044, + "grad_norm": 1.353570581947059, + "learning_rate": 1.2839459081816606e-06, + "loss": 1.0042, + "step": 6964 + }, + { + "epoch": 0.6281282409703747, + "grad_norm": 1.6003393945741409, + "learning_rate": 1.283400454722278e-06, + "loss": 1.052, + "step": 6965 + }, + { + "epoch": 0.628218424493845, + "grad_norm": 1.854101637082092, + "learning_rate": 1.28285506240712e-06, + "loss": 1.0477, + "step": 6966 + }, + { + "epoch": 0.6283086080173153, + "grad_norm": 1.405564798208279, + "learning_rate": 1.2823097312827225e-06, + "loss": 1.0114, + "step": 6967 + }, + { + "epoch": 0.6283987915407855, + "grad_norm": 1.766345729406921, + "learning_rate": 1.2817644613956153e-06, + "loss": 0.9939, + "step": 6968 + }, + { + "epoch": 0.6284889750642557, + "grad_norm": 1.717437173396692, + "learning_rate": 1.2812192527923253e-06, + "loss": 1.0477, + "step": 6969 + }, + { + "epoch": 0.6285791585877261, + "grad_norm": 1.80556628476729, + "learning_rate": 1.2806741055193712e-06, + "loss": 1.0202, + "step": 6970 + }, + { + "epoch": 0.6286693421111963, + "grad_norm": 1.7197114509468796, + "learning_rate": 1.2801290196232695e-06, + "loss": 0.9451, + "step": 6971 + }, + { + "epoch": 0.6287595256346665, + "grad_norm": 2.0954485024869274, + "learning_rate": 1.2795839951505282e-06, + "loss": 0.9741, + "step": 6972 + }, + { + "epoch": 0.6288497091581368, + "grad_norm": 74.931687639476, + "learning_rate": 1.2790390321476538e-06, + "loss": 1.0049, + "step": 6973 + }, + { + "epoch": 0.6289398926816071, + "grad_norm": 1.5816117849097113, + "learning_rate": 1.2784941306611446e-06, + "loss": 0.9595, + "step": 6974 + }, + { + "epoch": 0.6290300762050773, + "grad_norm": 1.929644564872965, + "learning_rate": 1.2779492907374935e-06, + "loss": 1.0459, + "step": 6975 + }, + { + "epoch": 0.6291202597285476, + "grad_norm": 1.7805035014023025, + "learning_rate": 1.2774045124231911e-06, + "loss": 0.9745, + "step": 6976 + }, + { + "epoch": 0.6292104432520178, + "grad_norm": 1.5052435143747136, + "learning_rate": 1.2768597957647197e-06, + "loss": 0.8669, + "step": 6977 + }, + { + "epoch": 0.6293006267754881, + "grad_norm": 2.313586727174699, + "learning_rate": 1.2763151408085582e-06, + "loss": 0.9994, + "step": 6978 + }, + { + "epoch": 0.6293908102989584, + "grad_norm": 2.3100504918380698, + "learning_rate": 1.2757705476011788e-06, + "loss": 0.9529, + "step": 6979 + }, + { + "epoch": 0.6294809938224286, + "grad_norm": 3.0741214954934972, + "learning_rate": 1.27522601618905e-06, + "loss": 0.9345, + "step": 6980 + }, + { + "epoch": 0.629571177345899, + "grad_norm": 0.5646240609131428, + "learning_rate": 1.2746815466186337e-06, + "loss": 0.8404, + "step": 6981 + }, + { + "epoch": 0.6296613608693692, + "grad_norm": 1.8549723714123083, + "learning_rate": 1.274137138936387e-06, + "loss": 0.9889, + "step": 6982 + }, + { + "epoch": 0.6297515443928394, + "grad_norm": 1.950454106396992, + "learning_rate": 1.2735927931887625e-06, + "loss": 0.988, + "step": 6983 + }, + { + "epoch": 0.6298417279163097, + "grad_norm": 0.6911419077581841, + "learning_rate": 1.2730485094222061e-06, + "loss": 0.8233, + "step": 6984 + }, + { + "epoch": 0.62993191143978, + "grad_norm": 1.790071694201, + "learning_rate": 1.2725042876831586e-06, + "loss": 1.0228, + "step": 6985 + }, + { + "epoch": 0.6300220949632502, + "grad_norm": 1.7943498200045311, + "learning_rate": 1.2719601280180573e-06, + "loss": 0.8802, + "step": 6986 + }, + { + "epoch": 0.6301122784867205, + "grad_norm": 1.6971765633098814, + "learning_rate": 1.2714160304733317e-06, + "loss": 1.0524, + "step": 6987 + }, + { + "epoch": 0.6302024620101907, + "grad_norm": 2.0024731365406976, + "learning_rate": 1.2708719950954082e-06, + "loss": 0.9974, + "step": 6988 + }, + { + "epoch": 0.630292645533661, + "grad_norm": 2.0892229245430656, + "learning_rate": 1.2703280219307065e-06, + "loss": 1.0064, + "step": 6989 + }, + { + "epoch": 0.6303828290571313, + "grad_norm": 1.5437874855092715, + "learning_rate": 1.2697841110256411e-06, + "loss": 0.9168, + "step": 6990 + }, + { + "epoch": 0.6304730125806015, + "grad_norm": 1.7658885657814913, + "learning_rate": 1.2692402624266221e-06, + "loss": 1.0339, + "step": 6991 + }, + { + "epoch": 0.6305631961040717, + "grad_norm": 1.816277480944711, + "learning_rate": 1.2686964761800529e-06, + "loss": 0.8953, + "step": 6992 + }, + { + "epoch": 0.6306533796275421, + "grad_norm": 0.6550170806302953, + "learning_rate": 1.268152752332333e-06, + "loss": 0.8213, + "step": 6993 + }, + { + "epoch": 0.6307435631510123, + "grad_norm": 1.997640230892612, + "learning_rate": 1.2676090909298549e-06, + "loss": 0.8931, + "step": 6994 + }, + { + "epoch": 0.6308337466744826, + "grad_norm": 1.5847652970319093, + "learning_rate": 1.2670654920190086e-06, + "loss": 1.0334, + "step": 6995 + }, + { + "epoch": 0.6309239301979528, + "grad_norm": 1.7346155498126539, + "learning_rate": 1.2665219556461754e-06, + "loss": 1.0359, + "step": 6996 + }, + { + "epoch": 0.6310141137214231, + "grad_norm": 1.6776432816419087, + "learning_rate": 1.2659784818577329e-06, + "loss": 1.0874, + "step": 6997 + }, + { + "epoch": 0.6311042972448934, + "grad_norm": 2.665023048830828, + "learning_rate": 1.2654350707000542e-06, + "loss": 0.869, + "step": 6998 + }, + { + "epoch": 0.6311944807683636, + "grad_norm": 1.6151883950198689, + "learning_rate": 1.264891722219505e-06, + "loss": 0.9731, + "step": 6999 + }, + { + "epoch": 0.6312846642918338, + "grad_norm": 1.5286933419309252, + "learning_rate": 1.2643484364624483e-06, + "loss": 0.9787, + "step": 7000 + }, + { + "epoch": 0.6313748478153042, + "grad_norm": 1.7867377890242584, + "learning_rate": 1.2638052134752393e-06, + "loss": 1.0394, + "step": 7001 + }, + { + "epoch": 0.6314650313387744, + "grad_norm": 1.8961102699091388, + "learning_rate": 1.2632620533042277e-06, + "loss": 1.0624, + "step": 7002 + }, + { + "epoch": 0.6315552148622446, + "grad_norm": 1.7904457841434438, + "learning_rate": 1.2627189559957612e-06, + "loss": 0.9579, + "step": 7003 + }, + { + "epoch": 0.631645398385715, + "grad_norm": 1.6593210789887138, + "learning_rate": 1.262175921596178e-06, + "loss": 1.0443, + "step": 7004 + }, + { + "epoch": 0.6317355819091852, + "grad_norm": 1.8040045564809633, + "learning_rate": 1.2616329501518137e-06, + "loss": 0.8606, + "step": 7005 + }, + { + "epoch": 0.6318257654326555, + "grad_norm": 2.217399831098535, + "learning_rate": 1.2610900417089978e-06, + "loss": 0.9084, + "step": 7006 + }, + { + "epoch": 0.6319159489561257, + "grad_norm": 1.9904315465230256, + "learning_rate": 1.2605471963140535e-06, + "loss": 0.9968, + "step": 7007 + }, + { + "epoch": 0.632006132479596, + "grad_norm": 2.2718439448987566, + "learning_rate": 1.2600044140133e-06, + "loss": 1.0601, + "step": 7008 + }, + { + "epoch": 0.6320963160030663, + "grad_norm": 1.7937981177066757, + "learning_rate": 1.2594616948530493e-06, + "loss": 0.9375, + "step": 7009 + }, + { + "epoch": 0.6321864995265365, + "grad_norm": 1.6987867485826567, + "learning_rate": 1.258919038879611e-06, + "loss": 1.0083, + "step": 7010 + }, + { + "epoch": 0.6322766830500067, + "grad_norm": 3.218500072072117, + "learning_rate": 1.2583764461392859e-06, + "loss": 0.9818, + "step": 7011 + }, + { + "epoch": 0.6323668665734771, + "grad_norm": 1.5779671920585856, + "learning_rate": 1.2578339166783724e-06, + "loss": 0.9468, + "step": 7012 + }, + { + "epoch": 0.6324570500969473, + "grad_norm": 2.5075967761273517, + "learning_rate": 1.2572914505431613e-06, + "loss": 0.9204, + "step": 7013 + }, + { + "epoch": 0.6325472336204175, + "grad_norm": 1.4139290003696234, + "learning_rate": 1.2567490477799383e-06, + "loss": 0.8868, + "step": 7014 + }, + { + "epoch": 0.6326374171438878, + "grad_norm": 1.515138292327172, + "learning_rate": 1.2562067084349852e-06, + "loss": 1.0817, + "step": 7015 + }, + { + "epoch": 0.6327276006673581, + "grad_norm": 3.575350956262786, + "learning_rate": 1.2556644325545764e-06, + "loss": 0.9663, + "step": 7016 + }, + { + "epoch": 0.6328177841908283, + "grad_norm": 1.6681190440878053, + "learning_rate": 1.255122220184983e-06, + "loss": 0.9669, + "step": 7017 + }, + { + "epoch": 0.6329079677142986, + "grad_norm": 1.6068072528720052, + "learning_rate": 1.2545800713724694e-06, + "loss": 0.8931, + "step": 7018 + }, + { + "epoch": 0.6329981512377688, + "grad_norm": 1.6819778795363973, + "learning_rate": 1.254037986163294e-06, + "loss": 0.8486, + "step": 7019 + }, + { + "epoch": 0.6330883347612392, + "grad_norm": 1.7744654562461244, + "learning_rate": 1.2534959646037104e-06, + "loss": 0.9427, + "step": 7020 + }, + { + "epoch": 0.6331785182847094, + "grad_norm": 1.4613960892719604, + "learning_rate": 1.2529540067399675e-06, + "loss": 0.987, + "step": 7021 + }, + { + "epoch": 0.6332687018081796, + "grad_norm": 1.7005672321931604, + "learning_rate": 1.252412112618308e-06, + "loss": 0.9183, + "step": 7022 + }, + { + "epoch": 0.6333588853316499, + "grad_norm": 22.263204364032763, + "learning_rate": 1.2518702822849696e-06, + "loss": 0.869, + "step": 7023 + }, + { + "epoch": 0.6334490688551202, + "grad_norm": 2.5492892378035408, + "learning_rate": 1.2513285157861831e-06, + "loss": 0.9265, + "step": 7024 + }, + { + "epoch": 0.6335392523785904, + "grad_norm": 1.7762779309884067, + "learning_rate": 1.2507868131681764e-06, + "loss": 0.979, + "step": 7025 + }, + { + "epoch": 0.6336294359020607, + "grad_norm": 1.5675158292710345, + "learning_rate": 1.250245174477169e-06, + "loss": 0.9629, + "step": 7026 + }, + { + "epoch": 0.6337196194255309, + "grad_norm": 0.6524815670786033, + "learning_rate": 1.2497035997593783e-06, + "loss": 0.7561, + "step": 7027 + }, + { + "epoch": 0.6338098029490012, + "grad_norm": 1.5898021345760802, + "learning_rate": 1.2491620890610135e-06, + "loss": 0.9836, + "step": 7028 + }, + { + "epoch": 0.6338999864724715, + "grad_norm": 1.4888422351774444, + "learning_rate": 1.2486206424282788e-06, + "loss": 0.9673, + "step": 7029 + }, + { + "epoch": 0.6339901699959417, + "grad_norm": 1.8792393125489222, + "learning_rate": 1.2480792599073743e-06, + "loss": 1.0377, + "step": 7030 + }, + { + "epoch": 0.634080353519412, + "grad_norm": 1.5802336098802114, + "learning_rate": 1.247537941544493e-06, + "loss": 1.0252, + "step": 7031 + }, + { + "epoch": 0.6341705370428823, + "grad_norm": 1.847080713682284, + "learning_rate": 1.2469966873858242e-06, + "loss": 1.0149, + "step": 7032 + }, + { + "epoch": 0.6342607205663525, + "grad_norm": 1.7867324514993752, + "learning_rate": 1.2464554974775496e-06, + "loss": 0.9095, + "step": 7033 + }, + { + "epoch": 0.6343509040898228, + "grad_norm": 1.689395334655563, + "learning_rate": 1.2459143718658474e-06, + "loss": 0.9819, + "step": 7034 + }, + { + "epoch": 0.6344410876132931, + "grad_norm": 1.9726266990940993, + "learning_rate": 1.2453733105968886e-06, + "loss": 0.9128, + "step": 7035 + }, + { + "epoch": 0.6345312711367633, + "grad_norm": 0.6153686619086809, + "learning_rate": 1.2448323137168399e-06, + "loss": 0.8218, + "step": 7036 + }, + { + "epoch": 0.6346214546602336, + "grad_norm": 2.2241782824871748, + "learning_rate": 1.2442913812718625e-06, + "loss": 0.9051, + "step": 7037 + }, + { + "epoch": 0.6347116381837038, + "grad_norm": 1.6175750415499017, + "learning_rate": 1.2437505133081108e-06, + "loss": 0.9688, + "step": 7038 + }, + { + "epoch": 0.6348018217071741, + "grad_norm": 1.7318942224433176, + "learning_rate": 1.2432097098717358e-06, + "loss": 1.0027, + "step": 7039 + }, + { + "epoch": 0.6348920052306444, + "grad_norm": 1.8542114495287745, + "learning_rate": 1.2426689710088813e-06, + "loss": 0.9859, + "step": 7040 + }, + { + "epoch": 0.6349821887541146, + "grad_norm": 1.6145275557533056, + "learning_rate": 1.2421282967656855e-06, + "loss": 1.0417, + "step": 7041 + }, + { + "epoch": 0.6350723722775848, + "grad_norm": 1.896778087874088, + "learning_rate": 1.2415876871882827e-06, + "loss": 1.0084, + "step": 7042 + }, + { + "epoch": 0.6351625558010552, + "grad_norm": 0.6712323375416929, + "learning_rate": 1.2410471423227998e-06, + "loss": 0.8862, + "step": 7043 + }, + { + "epoch": 0.6352527393245254, + "grad_norm": 1.583316685773099, + "learning_rate": 1.24050666221536e-06, + "loss": 0.9751, + "step": 7044 + }, + { + "epoch": 0.6353429228479957, + "grad_norm": 1.4920796460224444, + "learning_rate": 1.23996624691208e-06, + "loss": 1.0483, + "step": 7045 + }, + { + "epoch": 0.6354331063714659, + "grad_norm": 1.5494635915106996, + "learning_rate": 1.2394258964590693e-06, + "loss": 0.9667, + "step": 7046 + }, + { + "epoch": 0.6355232898949362, + "grad_norm": 1.7868987079127032, + "learning_rate": 1.238885610902436e-06, + "loss": 0.9275, + "step": 7047 + }, + { + "epoch": 0.6356134734184065, + "grad_norm": 1.9355139859481836, + "learning_rate": 1.2383453902882787e-06, + "loss": 0.9965, + "step": 7048 + }, + { + "epoch": 0.6357036569418767, + "grad_norm": 2.074772590217285, + "learning_rate": 1.2378052346626927e-06, + "loss": 1.0498, + "step": 7049 + }, + { + "epoch": 0.6357938404653469, + "grad_norm": 1.7448822529369654, + "learning_rate": 1.2372651440717665e-06, + "loss": 0.9064, + "step": 7050 + }, + { + "epoch": 0.6358840239888173, + "grad_norm": 1.780222144463009, + "learning_rate": 1.236725118561584e-06, + "loss": 1.0171, + "step": 7051 + }, + { + "epoch": 0.6359742075122875, + "grad_norm": 1.4782896889684707, + "learning_rate": 1.2361851581782232e-06, + "loss": 0.9854, + "step": 7052 + }, + { + "epoch": 0.6360643910357577, + "grad_norm": 1.5132819085619453, + "learning_rate": 1.2356452629677554e-06, + "loss": 1.0435, + "step": 7053 + }, + { + "epoch": 0.6361545745592281, + "grad_norm": 1.557452402994905, + "learning_rate": 1.2351054329762494e-06, + "loss": 0.9185, + "step": 7054 + }, + { + "epoch": 0.6362447580826983, + "grad_norm": 2.354531982794396, + "learning_rate": 1.2345656682497648e-06, + "loss": 1.01, + "step": 7055 + }, + { + "epoch": 0.6363349416061685, + "grad_norm": 1.3100056788634467, + "learning_rate": 1.2340259688343583e-06, + "loss": 0.9589, + "step": 7056 + }, + { + "epoch": 0.6364251251296388, + "grad_norm": 1.543097382024468, + "learning_rate": 1.2334863347760803e-06, + "loss": 0.8846, + "step": 7057 + }, + { + "epoch": 0.6365153086531091, + "grad_norm": 0.6461053526642612, + "learning_rate": 1.2329467661209734e-06, + "loss": 0.7819, + "step": 7058 + }, + { + "epoch": 0.6366054921765794, + "grad_norm": 2.1959917097022044, + "learning_rate": 1.2324072629150788e-06, + "loss": 1.024, + "step": 7059 + }, + { + "epoch": 0.6366956757000496, + "grad_norm": 2.4557635458963207, + "learning_rate": 1.2318678252044287e-06, + "loss": 1.0296, + "step": 7060 + }, + { + "epoch": 0.6367858592235198, + "grad_norm": 2.5054226239736956, + "learning_rate": 1.2313284530350512e-06, + "loss": 1.0381, + "step": 7061 + }, + { + "epoch": 0.6368760427469902, + "grad_norm": 2.0507252131480973, + "learning_rate": 1.230789146452969e-06, + "loss": 0.9186, + "step": 7062 + }, + { + "epoch": 0.6369662262704604, + "grad_norm": 1.8236049624233917, + "learning_rate": 1.2302499055041974e-06, + "loss": 0.9798, + "step": 7063 + }, + { + "epoch": 0.6370564097939306, + "grad_norm": 0.7139372367253954, + "learning_rate": 1.2297107302347488e-06, + "loss": 0.819, + "step": 7064 + }, + { + "epoch": 0.6371465933174009, + "grad_norm": 1.5265389978479698, + "learning_rate": 1.2291716206906275e-06, + "loss": 1.0501, + "step": 7065 + }, + { + "epoch": 0.6372367768408712, + "grad_norm": 2.4858582105765445, + "learning_rate": 1.2286325769178345e-06, + "loss": 1.0466, + "step": 7066 + }, + { + "epoch": 0.6373269603643414, + "grad_norm": 1.618150728270913, + "learning_rate": 1.2280935989623633e-06, + "loss": 1.0114, + "step": 7067 + }, + { + "epoch": 0.6374171438878117, + "grad_norm": 1.8081288079947897, + "learning_rate": 1.2275546868702017e-06, + "loss": 0.9186, + "step": 7068 + }, + { + "epoch": 0.6375073274112819, + "grad_norm": 1.5628138417720852, + "learning_rate": 1.2270158406873341e-06, + "loss": 0.9214, + "step": 7069 + }, + { + "epoch": 0.6375975109347523, + "grad_norm": 1.387462362861508, + "learning_rate": 1.2264770604597363e-06, + "loss": 1.0229, + "step": 7070 + }, + { + "epoch": 0.6376876944582225, + "grad_norm": 4.38053565629917, + "learning_rate": 1.2259383462333819e-06, + "loss": 1.0849, + "step": 7071 + }, + { + "epoch": 0.6377778779816927, + "grad_norm": 7.125355761831546, + "learning_rate": 1.2253996980542359e-06, + "loss": 0.9367, + "step": 7072 + }, + { + "epoch": 0.637868061505163, + "grad_norm": 2.122854720669688, + "learning_rate": 1.2248611159682578e-06, + "loss": 0.9645, + "step": 7073 + }, + { + "epoch": 0.6379582450286333, + "grad_norm": 0.7015944671144644, + "learning_rate": 1.2243226000214044e-06, + "loss": 0.7999, + "step": 7074 + }, + { + "epoch": 0.6380484285521035, + "grad_norm": 2.022043933915585, + "learning_rate": 1.2237841502596232e-06, + "loss": 0.9807, + "step": 7075 + }, + { + "epoch": 0.6381386120755738, + "grad_norm": 1.915576818460778, + "learning_rate": 1.2232457667288583e-06, + "loss": 0.9885, + "step": 7076 + }, + { + "epoch": 0.6382287955990441, + "grad_norm": 1.9821715248519023, + "learning_rate": 1.2227074494750476e-06, + "loss": 0.9884, + "step": 7077 + }, + { + "epoch": 0.6383189791225143, + "grad_norm": 1.6209043093376292, + "learning_rate": 1.2221691985441238e-06, + "loss": 1.019, + "step": 7078 + }, + { + "epoch": 0.6384091626459846, + "grad_norm": 1.5884059993990838, + "learning_rate": 1.2216310139820128e-06, + "loss": 0.9313, + "step": 7079 + }, + { + "epoch": 0.6384993461694548, + "grad_norm": 1.6013437261257404, + "learning_rate": 1.2210928958346347e-06, + "loss": 1.0266, + "step": 7080 + }, + { + "epoch": 0.6385895296929252, + "grad_norm": 1.5459909706592285, + "learning_rate": 1.2205548441479065e-06, + "loss": 0.9108, + "step": 7081 + }, + { + "epoch": 0.6386797132163954, + "grad_norm": 1.480711904401207, + "learning_rate": 1.2200168589677357e-06, + "loss": 0.9188, + "step": 7082 + }, + { + "epoch": 0.6387698967398656, + "grad_norm": 1.5082366821065283, + "learning_rate": 1.2194789403400284e-06, + "loss": 0.9518, + "step": 7083 + }, + { + "epoch": 0.6388600802633358, + "grad_norm": 2.10823650012087, + "learning_rate": 1.2189410883106816e-06, + "loss": 0.9584, + "step": 7084 + }, + { + "epoch": 0.6389502637868062, + "grad_norm": 1.4540420427816088, + "learning_rate": 1.2184033029255872e-06, + "loss": 1.053, + "step": 7085 + }, + { + "epoch": 0.6390404473102764, + "grad_norm": 0.6472397968892212, + "learning_rate": 1.2178655842306334e-06, + "loss": 0.8176, + "step": 7086 + }, + { + "epoch": 0.6391306308337467, + "grad_norm": 2.5019655130572507, + "learning_rate": 1.2173279322716999e-06, + "loss": 0.9811, + "step": 7087 + }, + { + "epoch": 0.6392208143572169, + "grad_norm": 1.7152275750779724, + "learning_rate": 1.216790347094663e-06, + "loss": 0.9868, + "step": 7088 + }, + { + "epoch": 0.6393109978806872, + "grad_norm": 1.6221912759212611, + "learning_rate": 1.2162528287453927e-06, + "loss": 1.0245, + "step": 7089 + }, + { + "epoch": 0.6394011814041575, + "grad_norm": 2.9187839089492584, + "learning_rate": 1.215715377269752e-06, + "loss": 0.9911, + "step": 7090 + }, + { + "epoch": 0.6394913649276277, + "grad_norm": 1.4932177916316292, + "learning_rate": 1.2151779927136003e-06, + "loss": 0.9638, + "step": 7091 + }, + { + "epoch": 0.6395815484510979, + "grad_norm": 1.9455737815558523, + "learning_rate": 1.2146406751227893e-06, + "loss": 1.0575, + "step": 7092 + }, + { + "epoch": 0.6396717319745683, + "grad_norm": 1.6505705424999055, + "learning_rate": 1.214103424543167e-06, + "loss": 1.0144, + "step": 7093 + }, + { + "epoch": 0.6397619154980385, + "grad_norm": 1.5850587110746752, + "learning_rate": 1.2135662410205735e-06, + "loss": 0.9011, + "step": 7094 + }, + { + "epoch": 0.6398520990215087, + "grad_norm": 2.2538678514723607, + "learning_rate": 1.2130291246008444e-06, + "loss": 0.9, + "step": 7095 + }, + { + "epoch": 0.639942282544979, + "grad_norm": 1.4202983624685295, + "learning_rate": 1.21249207532981e-06, + "loss": 0.903, + "step": 7096 + }, + { + "epoch": 0.6400324660684493, + "grad_norm": 2.295816190439828, + "learning_rate": 1.2119550932532936e-06, + "loss": 0.9037, + "step": 7097 + }, + { + "epoch": 0.6401226495919196, + "grad_norm": 2.1721984841929673, + "learning_rate": 1.2114181784171144e-06, + "loss": 0.9557, + "step": 7098 + }, + { + "epoch": 0.6402128331153898, + "grad_norm": 1.4392814378563312, + "learning_rate": 1.2108813308670837e-06, + "loss": 0.8857, + "step": 7099 + }, + { + "epoch": 0.6403030166388601, + "grad_norm": 2.3785057294454606, + "learning_rate": 1.2103445506490099e-06, + "loss": 0.9846, + "step": 7100 + }, + { + "epoch": 0.6403932001623304, + "grad_norm": 0.7186405264371005, + "learning_rate": 1.209807837808693e-06, + "loss": 0.829, + "step": 7101 + }, + { + "epoch": 0.6404833836858006, + "grad_norm": 1.8792509845275858, + "learning_rate": 1.2092711923919282e-06, + "loss": 1.0253, + "step": 7102 + }, + { + "epoch": 0.6405735672092708, + "grad_norm": 1.3916373265004127, + "learning_rate": 1.2087346144445053e-06, + "loss": 0.9426, + "step": 7103 + }, + { + "epoch": 0.6406637507327412, + "grad_norm": 2.293005038076623, + "learning_rate": 1.2081981040122081e-06, + "loss": 0.9931, + "step": 7104 + }, + { + "epoch": 0.6407539342562114, + "grad_norm": 3.205810611949418, + "learning_rate": 1.2076616611408148e-06, + "loss": 0.934, + "step": 7105 + }, + { + "epoch": 0.6408441177796816, + "grad_norm": 1.719098905348805, + "learning_rate": 1.2071252858760972e-06, + "loss": 1.0125, + "step": 7106 + }, + { + "epoch": 0.6409343013031519, + "grad_norm": 1.8437343532092298, + "learning_rate": 1.2065889782638218e-06, + "loss": 0.9685, + "step": 7107 + }, + { + "epoch": 0.6410244848266222, + "grad_norm": 1.732741127573749, + "learning_rate": 1.2060527383497506e-06, + "loss": 0.9641, + "step": 7108 + }, + { + "epoch": 0.6411146683500925, + "grad_norm": 1.6522586160285933, + "learning_rate": 1.2055165661796363e-06, + "loss": 1.0948, + "step": 7109 + }, + { + "epoch": 0.6412048518735627, + "grad_norm": 2.0129332792743786, + "learning_rate": 1.2049804617992303e-06, + "loss": 1.0548, + "step": 7110 + }, + { + "epoch": 0.6412950353970329, + "grad_norm": 0.6192559697383585, + "learning_rate": 1.204444425254275e-06, + "loss": 0.8295, + "step": 7111 + }, + { + "epoch": 0.6413852189205033, + "grad_norm": 1.7171470102772723, + "learning_rate": 1.203908456590507e-06, + "loss": 0.9614, + "step": 7112 + }, + { + "epoch": 0.6414754024439735, + "grad_norm": 1.6500643255237322, + "learning_rate": 1.20337255585366e-06, + "loss": 0.9562, + "step": 7113 + }, + { + "epoch": 0.6415655859674437, + "grad_norm": 0.6082673153480302, + "learning_rate": 1.2028367230894582e-06, + "loss": 0.7755, + "step": 7114 + }, + { + "epoch": 0.641655769490914, + "grad_norm": 1.635799322696186, + "learning_rate": 1.2023009583436237e-06, + "loss": 0.9032, + "step": 7115 + }, + { + "epoch": 0.6417459530143843, + "grad_norm": 1.810160738725822, + "learning_rate": 1.2017652616618698e-06, + "loss": 1.0485, + "step": 7116 + }, + { + "epoch": 0.6418361365378545, + "grad_norm": 2.1752547717781368, + "learning_rate": 1.2012296330899048e-06, + "loss": 0.9753, + "step": 7117 + }, + { + "epoch": 0.6419263200613248, + "grad_norm": 1.8509168827184481, + "learning_rate": 1.200694072673432e-06, + "loss": 1.0151, + "step": 7118 + }, + { + "epoch": 0.642016503584795, + "grad_norm": 1.5235616780069274, + "learning_rate": 1.200158580458148e-06, + "loss": 1.016, + "step": 7119 + }, + { + "epoch": 0.6421066871082654, + "grad_norm": 1.5028878070431877, + "learning_rate": 1.1996231564897448e-06, + "loss": 0.9582, + "step": 7120 + }, + { + "epoch": 0.6421968706317356, + "grad_norm": 2.571779707745852, + "learning_rate": 1.1990878008139062e-06, + "loss": 0.9389, + "step": 7121 + }, + { + "epoch": 0.6422870541552058, + "grad_norm": 1.7586401453257408, + "learning_rate": 1.1985525134763132e-06, + "loss": 1.0134, + "step": 7122 + }, + { + "epoch": 0.6423772376786762, + "grad_norm": 2.296975373650528, + "learning_rate": 1.1980172945226393e-06, + "loss": 1.0283, + "step": 7123 + }, + { + "epoch": 0.6424674212021464, + "grad_norm": 2.694457518380903, + "learning_rate": 1.197482143998551e-06, + "loss": 0.9907, + "step": 7124 + }, + { + "epoch": 0.6425576047256166, + "grad_norm": 2.0179869541317403, + "learning_rate": 1.196947061949712e-06, + "loss": 0.9709, + "step": 7125 + }, + { + "epoch": 0.6426477882490869, + "grad_norm": 1.5448131498908084, + "learning_rate": 1.1964120484217768e-06, + "loss": 1.0407, + "step": 7126 + }, + { + "epoch": 0.6427379717725572, + "grad_norm": 0.6985732994081006, + "learning_rate": 1.1958771034603975e-06, + "loss": 0.8005, + "step": 7127 + }, + { + "epoch": 0.6428281552960274, + "grad_norm": 2.0119402420922556, + "learning_rate": 1.1953422271112175e-06, + "loss": 0.9231, + "step": 7128 + }, + { + "epoch": 0.6429183388194977, + "grad_norm": 1.917150505473795, + "learning_rate": 1.1948074194198748e-06, + "loss": 0.9669, + "step": 7129 + }, + { + "epoch": 0.6430085223429679, + "grad_norm": 2.2766957219821764, + "learning_rate": 1.1942726804320033e-06, + "loss": 0.8956, + "step": 7130 + }, + { + "epoch": 0.6430987058664382, + "grad_norm": 1.9605986161747482, + "learning_rate": 1.1937380101932295e-06, + "loss": 0.9252, + "step": 7131 + }, + { + "epoch": 0.6431888893899085, + "grad_norm": 1.4161078061320955, + "learning_rate": 1.1932034087491745e-06, + "loss": 0.9777, + "step": 7132 + }, + { + "epoch": 0.6432790729133787, + "grad_norm": 2.779927228482861, + "learning_rate": 1.1926688761454531e-06, + "loss": 0.897, + "step": 7133 + }, + { + "epoch": 0.643369256436849, + "grad_norm": 1.3586248322979098, + "learning_rate": 1.1921344124276746e-06, + "loss": 0.9988, + "step": 7134 + }, + { + "epoch": 0.6434594399603193, + "grad_norm": 1.6078013161139588, + "learning_rate": 1.1916000176414431e-06, + "loss": 1.0041, + "step": 7135 + }, + { + "epoch": 0.6435496234837895, + "grad_norm": 1.7154286979282414, + "learning_rate": 1.1910656918323546e-06, + "loss": 1.0183, + "step": 7136 + }, + { + "epoch": 0.6436398070072598, + "grad_norm": 1.4690161930487537, + "learning_rate": 1.1905314350460024e-06, + "loss": 1.0135, + "step": 7137 + }, + { + "epoch": 0.64372999053073, + "grad_norm": 1.6331807136594647, + "learning_rate": 1.1899972473279717e-06, + "loss": 1.0365, + "step": 7138 + }, + { + "epoch": 0.6438201740542003, + "grad_norm": 1.5320336808219894, + "learning_rate": 1.1894631287238414e-06, + "loss": 0.975, + "step": 7139 + }, + { + "epoch": 0.6439103575776706, + "grad_norm": 1.4542759264949312, + "learning_rate": 1.188929079279187e-06, + "loss": 0.9925, + "step": 7140 + }, + { + "epoch": 0.6440005411011408, + "grad_norm": 1.9461435281764765, + "learning_rate": 1.1883950990395751e-06, + "loss": 0.9151, + "step": 7141 + }, + { + "epoch": 0.644090724624611, + "grad_norm": 2.0511514965259, + "learning_rate": 1.187861188050569e-06, + "loss": 0.9209, + "step": 7142 + }, + { + "epoch": 0.6441809081480814, + "grad_norm": 3.883523488334819, + "learning_rate": 1.187327346357724e-06, + "loss": 0.9705, + "step": 7143 + }, + { + "epoch": 0.6442710916715516, + "grad_norm": 1.7820838850209444, + "learning_rate": 1.1867935740065912e-06, + "loss": 1.007, + "step": 7144 + }, + { + "epoch": 0.6443612751950218, + "grad_norm": 1.675342706949829, + "learning_rate": 1.1862598710427148e-06, + "loss": 0.8735, + "step": 7145 + }, + { + "epoch": 0.6444514587184921, + "grad_norm": 3.7668294286305257, + "learning_rate": 1.1857262375116328e-06, + "loss": 0.9432, + "step": 7146 + }, + { + "epoch": 0.6445416422419624, + "grad_norm": 1.7383581251654492, + "learning_rate": 1.1851926734588783e-06, + "loss": 0.9836, + "step": 7147 + }, + { + "epoch": 0.6446318257654327, + "grad_norm": 1.6572412367684142, + "learning_rate": 1.184659178929977e-06, + "loss": 1.0045, + "step": 7148 + }, + { + "epoch": 0.6447220092889029, + "grad_norm": 2.0934190630923046, + "learning_rate": 1.1841257539704513e-06, + "loss": 0.9768, + "step": 7149 + }, + { + "epoch": 0.6448121928123732, + "grad_norm": 1.5001601292651694, + "learning_rate": 1.1835923986258146e-06, + "loss": 0.9978, + "step": 7150 + }, + { + "epoch": 0.6449023763358435, + "grad_norm": 1.3908716475959833, + "learning_rate": 1.1830591129415754e-06, + "loss": 0.9413, + "step": 7151 + }, + { + "epoch": 0.6449925598593137, + "grad_norm": 7.954052824148561, + "learning_rate": 1.182525896963238e-06, + "loss": 0.9272, + "step": 7152 + }, + { + "epoch": 0.6450827433827839, + "grad_norm": 2.2963524048740003, + "learning_rate": 1.181992750736298e-06, + "loss": 0.9139, + "step": 7153 + }, + { + "epoch": 0.6451729269062543, + "grad_norm": 1.929647591985166, + "learning_rate": 1.1814596743062474e-06, + "loss": 0.9768, + "step": 7154 + }, + { + "epoch": 0.6452631104297245, + "grad_norm": 8.389456953586103, + "learning_rate": 1.1809266677185711e-06, + "loss": 1.0095, + "step": 7155 + }, + { + "epoch": 0.6453532939531947, + "grad_norm": 0.6528422797803104, + "learning_rate": 1.180393731018747e-06, + "loss": 0.8488, + "step": 7156 + }, + { + "epoch": 0.645443477476665, + "grad_norm": 0.566719047729313, + "learning_rate": 1.1798608642522498e-06, + "loss": 0.7883, + "step": 7157 + }, + { + "epoch": 0.6455336610001353, + "grad_norm": 1.869837520394858, + "learning_rate": 1.1793280674645454e-06, + "loss": 0.9676, + "step": 7158 + }, + { + "epoch": 0.6456238445236056, + "grad_norm": 1.5791071488866089, + "learning_rate": 1.1787953407010954e-06, + "loss": 1.0553, + "step": 7159 + }, + { + "epoch": 0.6457140280470758, + "grad_norm": 1.7323068195658233, + "learning_rate": 1.1782626840073554e-06, + "loss": 0.9857, + "step": 7160 + }, + { + "epoch": 0.645804211570546, + "grad_norm": 2.034791763673839, + "learning_rate": 1.1777300974287738e-06, + "loss": 0.8886, + "step": 7161 + }, + { + "epoch": 0.6458943950940164, + "grad_norm": 1.6688109511770446, + "learning_rate": 1.1771975810107947e-06, + "loss": 1.0085, + "step": 7162 + }, + { + "epoch": 0.6459845786174866, + "grad_norm": 1.6841429768210692, + "learning_rate": 1.1766651347988542e-06, + "loss": 1.0531, + "step": 7163 + }, + { + "epoch": 0.6460747621409568, + "grad_norm": 1.9701806577153387, + "learning_rate": 1.1761327588383848e-06, + "loss": 1.0933, + "step": 7164 + }, + { + "epoch": 0.6461649456644271, + "grad_norm": 1.5896500603039159, + "learning_rate": 1.1756004531748105e-06, + "loss": 0.9956, + "step": 7165 + }, + { + "epoch": 0.6462551291878974, + "grad_norm": 1.6441061236059888, + "learning_rate": 1.1750682178535521e-06, + "loss": 1.0219, + "step": 7166 + }, + { + "epoch": 0.6463453127113676, + "grad_norm": 1.6413607446244536, + "learning_rate": 1.1745360529200218e-06, + "loss": 0.9334, + "step": 7167 + }, + { + "epoch": 0.6464354962348379, + "grad_norm": 2.706832636253083, + "learning_rate": 1.1740039584196261e-06, + "loss": 1.0095, + "step": 7168 + }, + { + "epoch": 0.6465256797583081, + "grad_norm": 3.158928669268177, + "learning_rate": 1.1734719343977683e-06, + "loss": 1.0142, + "step": 7169 + }, + { + "epoch": 0.6466158632817784, + "grad_norm": 1.6383699450407236, + "learning_rate": 1.1729399808998416e-06, + "loss": 0.9282, + "step": 7170 + }, + { + "epoch": 0.6467060468052487, + "grad_norm": 1.759791231127869, + "learning_rate": 1.1724080979712368e-06, + "loss": 0.8722, + "step": 7171 + }, + { + "epoch": 0.6467962303287189, + "grad_norm": 1.6567215877898291, + "learning_rate": 1.1718762856573365e-06, + "loss": 1.0506, + "step": 7172 + }, + { + "epoch": 0.6468864138521893, + "grad_norm": 1.420640933899442, + "learning_rate": 1.1713445440035172e-06, + "loss": 0.9666, + "step": 7173 + }, + { + "epoch": 0.6469765973756595, + "grad_norm": 2.1790044117533194, + "learning_rate": 1.1708128730551506e-06, + "loss": 1.0242, + "step": 7174 + }, + { + "epoch": 0.6470667808991297, + "grad_norm": 1.8195463213495002, + "learning_rate": 1.1702812728576019e-06, + "loss": 0.9715, + "step": 7175 + }, + { + "epoch": 0.6471569644226, + "grad_norm": 1.6731604509726785, + "learning_rate": 1.1697497434562303e-06, + "loss": 1.0104, + "step": 7176 + }, + { + "epoch": 0.6472471479460703, + "grad_norm": 1.8276735710242293, + "learning_rate": 1.1692182848963885e-06, + "loss": 1.0183, + "step": 7177 + }, + { + "epoch": 0.6473373314695405, + "grad_norm": 2.015329147035872, + "learning_rate": 1.1686868972234227e-06, + "loss": 0.9446, + "step": 7178 + }, + { + "epoch": 0.6474275149930108, + "grad_norm": 1.2862546453215977, + "learning_rate": 1.1681555804826755e-06, + "loss": 0.9902, + "step": 7179 + }, + { + "epoch": 0.647517698516481, + "grad_norm": 1.4280351773318274, + "learning_rate": 1.1676243347194806e-06, + "loss": 0.9218, + "step": 7180 + }, + { + "epoch": 0.6476078820399513, + "grad_norm": 1.4018427131734206, + "learning_rate": 1.167093159979167e-06, + "loss": 1.0013, + "step": 7181 + }, + { + "epoch": 0.6476980655634216, + "grad_norm": 2.1952037852237973, + "learning_rate": 1.1665620563070575e-06, + "loss": 1.0119, + "step": 7182 + }, + { + "epoch": 0.6477882490868918, + "grad_norm": 2.0601455659267423, + "learning_rate": 1.1660310237484691e-06, + "loss": 1.0888, + "step": 7183 + }, + { + "epoch": 0.647878432610362, + "grad_norm": 1.5147482312184728, + "learning_rate": 1.165500062348713e-06, + "loss": 0.9271, + "step": 7184 + }, + { + "epoch": 0.6479686161338324, + "grad_norm": 2.822480950506646, + "learning_rate": 1.164969172153091e-06, + "loss": 1.026, + "step": 7185 + }, + { + "epoch": 0.6480587996573026, + "grad_norm": 6.613558407973262, + "learning_rate": 1.1644383532069055e-06, + "loss": 0.9992, + "step": 7186 + }, + { + "epoch": 0.6481489831807729, + "grad_norm": 1.3401572525218137, + "learning_rate": 1.1639076055554454e-06, + "loss": 1.0212, + "step": 7187 + }, + { + "epoch": 0.6482391667042431, + "grad_norm": 1.985091913435032, + "learning_rate": 1.163376929244e-06, + "loss": 0.9753, + "step": 7188 + }, + { + "epoch": 0.6483293502277134, + "grad_norm": 1.3162732156153456, + "learning_rate": 1.1628463243178472e-06, + "loss": 0.9459, + "step": 7189 + }, + { + "epoch": 0.6484195337511837, + "grad_norm": 3.3199194103249323, + "learning_rate": 1.1623157908222623e-06, + "loss": 1.0464, + "step": 7190 + }, + { + "epoch": 0.6485097172746539, + "grad_norm": 1.566697852182794, + "learning_rate": 1.1617853288025129e-06, + "loss": 1.0204, + "step": 7191 + }, + { + "epoch": 0.6485999007981241, + "grad_norm": 1.4453068088728747, + "learning_rate": 1.1612549383038612e-06, + "loss": 0.9568, + "step": 7192 + }, + { + "epoch": 0.6486900843215945, + "grad_norm": 1.3656354977970229, + "learning_rate": 1.1607246193715629e-06, + "loss": 0.9626, + "step": 7193 + }, + { + "epoch": 0.6487802678450647, + "grad_norm": 1.9710773456751176, + "learning_rate": 1.1601943720508684e-06, + "loss": 0.9981, + "step": 7194 + }, + { + "epoch": 0.6488704513685349, + "grad_norm": 1.4282875920067173, + "learning_rate": 1.1596641963870202e-06, + "loss": 1.0317, + "step": 7195 + }, + { + "epoch": 0.6489606348920053, + "grad_norm": 1.7397426676923886, + "learning_rate": 1.1591340924252561e-06, + "loss": 0.9955, + "step": 7196 + }, + { + "epoch": 0.6490508184154755, + "grad_norm": 0.7549868730325966, + "learning_rate": 1.158604060210808e-06, + "loss": 0.8605, + "step": 7197 + }, + { + "epoch": 0.6491410019389457, + "grad_norm": 1.8487740939764576, + "learning_rate": 1.1580740997889008e-06, + "loss": 1.0437, + "step": 7198 + }, + { + "epoch": 0.649231185462416, + "grad_norm": 2.1663141208322463, + "learning_rate": 1.1575442112047544e-06, + "loss": 0.9823, + "step": 7199 + }, + { + "epoch": 0.6493213689858863, + "grad_norm": 1.8853125850829857, + "learning_rate": 1.1570143945035797e-06, + "loss": 0.9195, + "step": 7200 + }, + { + "epoch": 0.6494115525093566, + "grad_norm": 3.5058319005713146, + "learning_rate": 1.1564846497305864e-06, + "loss": 0.9368, + "step": 7201 + }, + { + "epoch": 0.6495017360328268, + "grad_norm": 2.211994902289313, + "learning_rate": 1.1559549769309726e-06, + "loss": 0.9596, + "step": 7202 + }, + { + "epoch": 0.649591919556297, + "grad_norm": 1.446953331822857, + "learning_rate": 1.1554253761499358e-06, + "loss": 0.9769, + "step": 7203 + }, + { + "epoch": 0.6496821030797674, + "grad_norm": 1.8228935312892016, + "learning_rate": 1.1548958474326617e-06, + "loss": 1.0206, + "step": 7204 + }, + { + "epoch": 0.6497722866032376, + "grad_norm": 2.9760988530179184, + "learning_rate": 1.154366390824334e-06, + "loss": 0.9153, + "step": 7205 + }, + { + "epoch": 0.6498624701267078, + "grad_norm": 0.786499127641547, + "learning_rate": 1.1538370063701287e-06, + "loss": 0.7953, + "step": 7206 + }, + { + "epoch": 0.6499526536501781, + "grad_norm": 1.7252411770497589, + "learning_rate": 1.1533076941152153e-06, + "loss": 0.9427, + "step": 7207 + }, + { + "epoch": 0.6500428371736484, + "grad_norm": 1.8440887010759306, + "learning_rate": 1.1527784541047583e-06, + "loss": 0.9119, + "step": 7208 + }, + { + "epoch": 0.6501330206971186, + "grad_norm": 12.948814567598637, + "learning_rate": 1.1522492863839152e-06, + "loss": 0.9344, + "step": 7209 + }, + { + "epoch": 0.6502232042205889, + "grad_norm": 4.855184140077245, + "learning_rate": 1.1517201909978382e-06, + "loss": 0.9432, + "step": 7210 + }, + { + "epoch": 0.6503133877440591, + "grad_norm": 0.5988433368141769, + "learning_rate": 1.151191167991671e-06, + "loss": 0.8154, + "step": 7211 + }, + { + "epoch": 0.6504035712675295, + "grad_norm": 2.0869259439004315, + "learning_rate": 1.1506622174105536e-06, + "loss": 0.9899, + "step": 7212 + }, + { + "epoch": 0.6504937547909997, + "grad_norm": 1.7798612434439085, + "learning_rate": 1.1501333392996194e-06, + "loss": 1.0149, + "step": 7213 + }, + { + "epoch": 0.6505839383144699, + "grad_norm": 0.6570778575065118, + "learning_rate": 1.1496045337039943e-06, + "loss": 0.8218, + "step": 7214 + }, + { + "epoch": 0.6506741218379402, + "grad_norm": 2.7036883698739396, + "learning_rate": 1.1490758006687995e-06, + "loss": 0.9392, + "step": 7215 + }, + { + "epoch": 0.6507643053614105, + "grad_norm": 5.647943456631506, + "learning_rate": 1.1485471402391502e-06, + "loss": 0.96, + "step": 7216 + }, + { + "epoch": 0.6508544888848807, + "grad_norm": 1.8865267259012455, + "learning_rate": 1.1480185524601522e-06, + "loss": 0.955, + "step": 7217 + }, + { + "epoch": 0.650944672408351, + "grad_norm": 1.5099372716846906, + "learning_rate": 1.1474900373769108e-06, + "loss": 0.9943, + "step": 7218 + }, + { + "epoch": 0.6510348559318213, + "grad_norm": 0.5700178038175705, + "learning_rate": 1.1469615950345184e-06, + "loss": 0.7612, + "step": 7219 + }, + { + "epoch": 0.6511250394552915, + "grad_norm": 0.7055689194578402, + "learning_rate": 1.1464332254780678e-06, + "loss": 0.7993, + "step": 7220 + }, + { + "epoch": 0.6512152229787618, + "grad_norm": 1.816774261274819, + "learning_rate": 1.1459049287526404e-06, + "loss": 0.9662, + "step": 7221 + }, + { + "epoch": 0.651305406502232, + "grad_norm": 1.5601952912283898, + "learning_rate": 1.1453767049033137e-06, + "loss": 0.9405, + "step": 7222 + }, + { + "epoch": 0.6513955900257024, + "grad_norm": 1.6674407909582032, + "learning_rate": 1.1448485539751586e-06, + "loss": 1.0235, + "step": 7223 + }, + { + "epoch": 0.6514857735491726, + "grad_norm": 1.5175305215446508, + "learning_rate": 1.1443204760132408e-06, + "loss": 0.9509, + "step": 7224 + }, + { + "epoch": 0.6515759570726428, + "grad_norm": 1.504931608147904, + "learning_rate": 1.1437924710626185e-06, + "loss": 0.9653, + "step": 7225 + }, + { + "epoch": 0.651666140596113, + "grad_norm": 0.6503832934083766, + "learning_rate": 1.1432645391683429e-06, + "loss": 0.7835, + "step": 7226 + }, + { + "epoch": 0.6517563241195834, + "grad_norm": 1.9251830620853392, + "learning_rate": 1.1427366803754609e-06, + "loss": 1.0192, + "step": 7227 + }, + { + "epoch": 0.6518465076430536, + "grad_norm": 1.6006421678032776, + "learning_rate": 1.142208894729012e-06, + "loss": 0.9966, + "step": 7228 + }, + { + "epoch": 0.6519366911665239, + "grad_norm": 1.4809034207655583, + "learning_rate": 1.1416811822740301e-06, + "loss": 1.0206, + "step": 7229 + }, + { + "epoch": 0.6520268746899941, + "grad_norm": 1.724135267814684, + "learning_rate": 1.1411535430555428e-06, + "loss": 1.0338, + "step": 7230 + }, + { + "epoch": 0.6521170582134644, + "grad_norm": 1.5598450038400093, + "learning_rate": 1.1406259771185705e-06, + "loss": 1.0385, + "step": 7231 + }, + { + "epoch": 0.6522072417369347, + "grad_norm": 1.3920704048482762, + "learning_rate": 1.1400984845081282e-06, + "loss": 0.939, + "step": 7232 + }, + { + "epoch": 0.6522974252604049, + "grad_norm": 2.2457186974572596, + "learning_rate": 1.139571065269226e-06, + "loss": 1.0589, + "step": 7233 + }, + { + "epoch": 0.6523876087838751, + "grad_norm": 1.52673288602905, + "learning_rate": 1.139043719446863e-06, + "loss": 1.0263, + "step": 7234 + }, + { + "epoch": 0.6524777923073455, + "grad_norm": 1.6117428887353709, + "learning_rate": 1.1385164470860385e-06, + "loss": 1.0566, + "step": 7235 + }, + { + "epoch": 0.6525679758308157, + "grad_norm": 1.8877040272771546, + "learning_rate": 1.1379892482317403e-06, + "loss": 0.9821, + "step": 7236 + }, + { + "epoch": 0.652658159354286, + "grad_norm": 1.561806563045406, + "learning_rate": 1.1374621229289524e-06, + "loss": 0.9676, + "step": 7237 + }, + { + "epoch": 0.6527483428777562, + "grad_norm": 1.7311207939987656, + "learning_rate": 1.1369350712226525e-06, + "loss": 0.9662, + "step": 7238 + }, + { + "epoch": 0.6528385264012265, + "grad_norm": 2.623443596238124, + "learning_rate": 1.136408093157811e-06, + "loss": 0.9335, + "step": 7239 + }, + { + "epoch": 0.6529287099246968, + "grad_norm": 1.8392345934574272, + "learning_rate": 1.1358811887793935e-06, + "loss": 0.9569, + "step": 7240 + }, + { + "epoch": 0.653018893448167, + "grad_norm": 1.6775563759575984, + "learning_rate": 1.135354358132356e-06, + "loss": 1.017, + "step": 7241 + }, + { + "epoch": 0.6531090769716373, + "grad_norm": 1.646595219221696, + "learning_rate": 1.1348276012616542e-06, + "loss": 1.0227, + "step": 7242 + }, + { + "epoch": 0.6531992604951076, + "grad_norm": 1.7717624769862825, + "learning_rate": 1.134300918212231e-06, + "loss": 0.9395, + "step": 7243 + }, + { + "epoch": 0.6532894440185778, + "grad_norm": 2.216422566141037, + "learning_rate": 1.133774309029027e-06, + "loss": 0.9942, + "step": 7244 + }, + { + "epoch": 0.653379627542048, + "grad_norm": 1.6083180929196486, + "learning_rate": 1.133247773756975e-06, + "loss": 1.0187, + "step": 7245 + }, + { + "epoch": 0.6534698110655184, + "grad_norm": 0.657837990020025, + "learning_rate": 1.1327213124410024e-06, + "loss": 0.8545, + "step": 7246 + }, + { + "epoch": 0.6535599945889886, + "grad_norm": 1.7315226977031324, + "learning_rate": 1.1321949251260292e-06, + "loss": 1.0274, + "step": 7247 + }, + { + "epoch": 0.6536501781124588, + "grad_norm": 1.6313559136043265, + "learning_rate": 1.1316686118569712e-06, + "loss": 0.8689, + "step": 7248 + }, + { + "epoch": 0.6537403616359291, + "grad_norm": 1.638086589550663, + "learning_rate": 1.1311423726787335e-06, + "loss": 0.8964, + "step": 7249 + }, + { + "epoch": 0.6538305451593994, + "grad_norm": 1.6748442620142188, + "learning_rate": 1.130616207636221e-06, + "loss": 0.9519, + "step": 7250 + }, + { + "epoch": 0.6539207286828697, + "grad_norm": 1.806873071511822, + "learning_rate": 1.1300901167743263e-06, + "loss": 0.9896, + "step": 7251 + }, + { + "epoch": 0.6540109122063399, + "grad_norm": 1.2762142402055312, + "learning_rate": 1.12956410013794e-06, + "loss": 0.9813, + "step": 7252 + }, + { + "epoch": 0.6541010957298101, + "grad_norm": 2.077116068815144, + "learning_rate": 1.1290381577719436e-06, + "loss": 0.9738, + "step": 7253 + }, + { + "epoch": 0.6541912792532805, + "grad_norm": 1.9250984141967007, + "learning_rate": 1.1285122897212143e-06, + "loss": 0.8811, + "step": 7254 + }, + { + "epoch": 0.6542814627767507, + "grad_norm": 0.7393857172546393, + "learning_rate": 1.1279864960306228e-06, + "loss": 0.8551, + "step": 7255 + }, + { + "epoch": 0.6543716463002209, + "grad_norm": 1.8910949375199604, + "learning_rate": 1.1274607767450297e-06, + "loss": 1.0592, + "step": 7256 + }, + { + "epoch": 0.6544618298236912, + "grad_norm": 2.6612081590537424, + "learning_rate": 1.126935131909296e-06, + "loss": 0.9607, + "step": 7257 + }, + { + "epoch": 0.6545520133471615, + "grad_norm": 3.8930671105311823, + "learning_rate": 1.1264095615682693e-06, + "loss": 0.9304, + "step": 7258 + }, + { + "epoch": 0.6546421968706317, + "grad_norm": 1.593661810735025, + "learning_rate": 1.1258840657667973e-06, + "loss": 0.9037, + "step": 7259 + }, + { + "epoch": 0.654732380394102, + "grad_norm": 1.571579910397306, + "learning_rate": 1.125358644549716e-06, + "loss": 0.9131, + "step": 7260 + }, + { + "epoch": 0.6548225639175722, + "grad_norm": 1.2921652703177833, + "learning_rate": 1.1248332979618578e-06, + "loss": 0.8427, + "step": 7261 + }, + { + "epoch": 0.6549127474410426, + "grad_norm": 1.4249554476549482, + "learning_rate": 1.1243080260480482e-06, + "loss": 0.8767, + "step": 7262 + }, + { + "epoch": 0.6550029309645128, + "grad_norm": 1.8470143015524163, + "learning_rate": 1.1237828288531063e-06, + "loss": 1.0375, + "step": 7263 + }, + { + "epoch": 0.655093114487983, + "grad_norm": 1.5640423600979956, + "learning_rate": 1.1232577064218449e-06, + "loss": 0.9011, + "step": 7264 + }, + { + "epoch": 0.6551832980114533, + "grad_norm": 1.6276579906464936, + "learning_rate": 1.1227326587990711e-06, + "loss": 1.0592, + "step": 7265 + }, + { + "epoch": 0.6552734815349236, + "grad_norm": 2.857845710407087, + "learning_rate": 1.1222076860295832e-06, + "loss": 0.9667, + "step": 7266 + }, + { + "epoch": 0.6553636650583938, + "grad_norm": 2.043188138017149, + "learning_rate": 1.1216827881581756e-06, + "loss": 1.0002, + "step": 7267 + }, + { + "epoch": 0.6554538485818641, + "grad_norm": 4.586489174720584, + "learning_rate": 1.1211579652296355e-06, + "loss": 0.977, + "step": 7268 + }, + { + "epoch": 0.6555440321053344, + "grad_norm": 2.1227655443703544, + "learning_rate": 1.1206332172887438e-06, + "loss": 1.0658, + "step": 7269 + }, + { + "epoch": 0.6556342156288046, + "grad_norm": 1.8931944230436566, + "learning_rate": 1.1201085443802756e-06, + "loss": 0.997, + "step": 7270 + }, + { + "epoch": 0.6557243991522749, + "grad_norm": 1.5464809956312533, + "learning_rate": 1.1195839465489964e-06, + "loss": 0.9485, + "step": 7271 + }, + { + "epoch": 0.6558145826757451, + "grad_norm": 1.8230404693706765, + "learning_rate": 1.1190594238396708e-06, + "loss": 1.0238, + "step": 7272 + }, + { + "epoch": 0.6559047661992154, + "grad_norm": 8.416545703383807, + "learning_rate": 1.1185349762970515e-06, + "loss": 0.9538, + "step": 7273 + }, + { + "epoch": 0.6559949497226857, + "grad_norm": 1.4591186951849338, + "learning_rate": 1.1180106039658896e-06, + "loss": 0.9857, + "step": 7274 + }, + { + "epoch": 0.6560851332461559, + "grad_norm": 0.5812230226962778, + "learning_rate": 1.117486306890925e-06, + "loss": 0.795, + "step": 7275 + }, + { + "epoch": 0.6561753167696261, + "grad_norm": 1.9757392586212512, + "learning_rate": 1.116962085116896e-06, + "loss": 0.9363, + "step": 7276 + }, + { + "epoch": 0.6562655002930965, + "grad_norm": 0.6655907044507817, + "learning_rate": 1.1164379386885302e-06, + "loss": 0.7551, + "step": 7277 + }, + { + "epoch": 0.6563556838165667, + "grad_norm": 1.8628988133428699, + "learning_rate": 1.1159138676505516e-06, + "loss": 1.0267, + "step": 7278 + }, + { + "epoch": 0.656445867340037, + "grad_norm": 1.876337273872391, + "learning_rate": 1.1153898720476761e-06, + "loss": 1.0397, + "step": 7279 + }, + { + "epoch": 0.6565360508635072, + "grad_norm": 1.9507807586877073, + "learning_rate": 1.114865951924615e-06, + "loss": 0.9679, + "step": 7280 + }, + { + "epoch": 0.6566262343869775, + "grad_norm": 1.5001391505230464, + "learning_rate": 1.1143421073260721e-06, + "loss": 0.9247, + "step": 7281 + }, + { + "epoch": 0.6567164179104478, + "grad_norm": 2.7718152956892252, + "learning_rate": 1.1138183382967432e-06, + "loss": 0.9192, + "step": 7282 + }, + { + "epoch": 0.656806601433918, + "grad_norm": 1.7516794320514582, + "learning_rate": 1.11329464488132e-06, + "loss": 0.8939, + "step": 7283 + }, + { + "epoch": 0.6568967849573882, + "grad_norm": 1.5709067212002232, + "learning_rate": 1.112771027124487e-06, + "loss": 0.958, + "step": 7284 + }, + { + "epoch": 0.6569869684808586, + "grad_norm": 1.6037575914377198, + "learning_rate": 1.112247485070922e-06, + "loss": 0.9957, + "step": 7285 + }, + { + "epoch": 0.6570771520043288, + "grad_norm": 1.7029279542378852, + "learning_rate": 1.1117240187652968e-06, + "loss": 0.9843, + "step": 7286 + }, + { + "epoch": 0.657167335527799, + "grad_norm": 7.259309776430248, + "learning_rate": 1.1112006282522767e-06, + "loss": 1.0615, + "step": 7287 + }, + { + "epoch": 0.6572575190512693, + "grad_norm": 1.7953840621616486, + "learning_rate": 1.1106773135765183e-06, + "loss": 0.9643, + "step": 7288 + }, + { + "epoch": 0.6573477025747396, + "grad_norm": 1.4644650389358917, + "learning_rate": 1.110154074782677e-06, + "loss": 0.9591, + "step": 7289 + }, + { + "epoch": 0.6574378860982099, + "grad_norm": 1.621409043133857, + "learning_rate": 1.1096309119153948e-06, + "loss": 0.978, + "step": 7290 + }, + { + "epoch": 0.6575280696216801, + "grad_norm": 0.6271589422638135, + "learning_rate": 1.1091078250193145e-06, + "loss": 0.8562, + "step": 7291 + }, + { + "epoch": 0.6576182531451504, + "grad_norm": 5.855233236695487, + "learning_rate": 1.108584814139066e-06, + "loss": 0.9866, + "step": 7292 + }, + { + "epoch": 0.6577084366686207, + "grad_norm": 2.579492841893657, + "learning_rate": 1.108061879319276e-06, + "loss": 0.9801, + "step": 7293 + }, + { + "epoch": 0.6577986201920909, + "grad_norm": 3.0753399792802063, + "learning_rate": 1.1075390206045648e-06, + "loss": 1.0408, + "step": 7294 + }, + { + "epoch": 0.6578888037155611, + "grad_norm": 1.8404113153275197, + "learning_rate": 1.1070162380395454e-06, + "loss": 1.0274, + "step": 7295 + }, + { + "epoch": 0.6579789872390315, + "grad_norm": 1.5255419940550998, + "learning_rate": 1.1064935316688253e-06, + "loss": 0.9248, + "step": 7296 + }, + { + "epoch": 0.6580691707625017, + "grad_norm": 1.5681954912235425, + "learning_rate": 1.105970901537002e-06, + "loss": 1.0673, + "step": 7297 + }, + { + "epoch": 0.6581593542859719, + "grad_norm": 1.617806431370436, + "learning_rate": 1.1054483476886727e-06, + "loss": 1.026, + "step": 7298 + }, + { + "epoch": 0.6582495378094422, + "grad_norm": 1.432813801239654, + "learning_rate": 1.1049258701684222e-06, + "loss": 0.9048, + "step": 7299 + }, + { + "epoch": 0.6583397213329125, + "grad_norm": 2.006607584711928, + "learning_rate": 1.1044034690208315e-06, + "loss": 0.9315, + "step": 7300 + }, + { + "epoch": 0.6584299048563828, + "grad_norm": 2.1213705820081783, + "learning_rate": 1.1038811442904755e-06, + "loss": 1.0139, + "step": 7301 + }, + { + "epoch": 0.658520088379853, + "grad_norm": 1.8879847731380341, + "learning_rate": 1.103358896021921e-06, + "loss": 0.9161, + "step": 7302 + }, + { + "epoch": 0.6586102719033232, + "grad_norm": 1.841936480347487, + "learning_rate": 1.1028367242597298e-06, + "loss": 0.9584, + "step": 7303 + }, + { + "epoch": 0.6587004554267936, + "grad_norm": 1.833264096715548, + "learning_rate": 1.102314629048457e-06, + "loss": 1.0202, + "step": 7304 + }, + { + "epoch": 0.6587906389502638, + "grad_norm": 1.547744121492905, + "learning_rate": 1.1017926104326484e-06, + "loss": 1.0308, + "step": 7305 + }, + { + "epoch": 0.658880822473734, + "grad_norm": 1.6408523402058484, + "learning_rate": 1.1012706684568483e-06, + "loss": 1.0359, + "step": 7306 + }, + { + "epoch": 0.6589710059972043, + "grad_norm": 1.4041764548589717, + "learning_rate": 1.1007488031655894e-06, + "loss": 1.0014, + "step": 7307 + }, + { + "epoch": 0.6590611895206746, + "grad_norm": 2.0572764979579903, + "learning_rate": 1.1002270146034013e-06, + "loss": 0.9389, + "step": 7308 + }, + { + "epoch": 0.6591513730441448, + "grad_norm": 1.5475822912469477, + "learning_rate": 1.0997053028148052e-06, + "loss": 1.0107, + "step": 7309 + }, + { + "epoch": 0.6592415565676151, + "grad_norm": 1.3982746732742342, + "learning_rate": 1.0991836678443173e-06, + "loss": 0.9874, + "step": 7310 + }, + { + "epoch": 0.6593317400910853, + "grad_norm": 1.6012826070212678, + "learning_rate": 1.0986621097364465e-06, + "loss": 0.927, + "step": 7311 + }, + { + "epoch": 0.6594219236145556, + "grad_norm": 3.230541961836667, + "learning_rate": 1.0981406285356932e-06, + "loss": 0.8981, + "step": 7312 + }, + { + "epoch": 0.6595121071380259, + "grad_norm": 1.660406116306892, + "learning_rate": 1.0976192242865554e-06, + "loss": 0.904, + "step": 7313 + }, + { + "epoch": 0.6596022906614961, + "grad_norm": 2.226644949473999, + "learning_rate": 1.0970978970335202e-06, + "loss": 1.0471, + "step": 7314 + }, + { + "epoch": 0.6596924741849665, + "grad_norm": 1.6042081241287334, + "learning_rate": 1.0965766468210714e-06, + "loss": 1.0059, + "step": 7315 + }, + { + "epoch": 0.6597826577084367, + "grad_norm": 1.6138947361839997, + "learning_rate": 1.0960554736936843e-06, + "loss": 0.997, + "step": 7316 + }, + { + "epoch": 0.6598728412319069, + "grad_norm": 2.1807911037641445, + "learning_rate": 1.0955343776958283e-06, + "loss": 0.8944, + "step": 7317 + }, + { + "epoch": 0.6599630247553772, + "grad_norm": 1.8132383881400513, + "learning_rate": 1.0950133588719665e-06, + "loss": 0.9986, + "step": 7318 + }, + { + "epoch": 0.6600532082788475, + "grad_norm": 1.8643815412671965, + "learning_rate": 1.0944924172665551e-06, + "loss": 1.0064, + "step": 7319 + }, + { + "epoch": 0.6601433918023177, + "grad_norm": 1.8683806880538436, + "learning_rate": 1.0939715529240437e-06, + "loss": 1.0184, + "step": 7320 + }, + { + "epoch": 0.660233575325788, + "grad_norm": 1.619822249374255, + "learning_rate": 1.0934507658888755e-06, + "loss": 1.0154, + "step": 7321 + }, + { + "epoch": 0.6603237588492582, + "grad_norm": 1.6035342105062542, + "learning_rate": 1.092930056205486e-06, + "loss": 0.9548, + "step": 7322 + }, + { + "epoch": 0.6604139423727285, + "grad_norm": 2.1452276492304274, + "learning_rate": 1.092409423918306e-06, + "loss": 0.9505, + "step": 7323 + }, + { + "epoch": 0.6605041258961988, + "grad_norm": 1.4803837170500866, + "learning_rate": 1.0918888690717581e-06, + "loss": 0.9062, + "step": 7324 + }, + { + "epoch": 0.660594309419669, + "grad_norm": 2.69701360565137, + "learning_rate": 1.091368391710259e-06, + "loss": 1.0499, + "step": 7325 + }, + { + "epoch": 0.6606844929431392, + "grad_norm": 1.7979142873784393, + "learning_rate": 1.0908479918782198e-06, + "loss": 0.9294, + "step": 7326 + }, + { + "epoch": 0.6607746764666096, + "grad_norm": 2.1047065218580605, + "learning_rate": 1.0903276696200413e-06, + "loss": 1.0073, + "step": 7327 + }, + { + "epoch": 0.6608648599900798, + "grad_norm": 1.777498343315372, + "learning_rate": 1.0898074249801234e-06, + "loss": 0.9898, + "step": 7328 + }, + { + "epoch": 0.6609550435135501, + "grad_norm": 1.527948596344353, + "learning_rate": 1.0892872580028533e-06, + "loss": 1.0005, + "step": 7329 + }, + { + "epoch": 0.6610452270370203, + "grad_norm": 1.7642302278802937, + "learning_rate": 1.0887671687326178e-06, + "loss": 1.0161, + "step": 7330 + }, + { + "epoch": 0.6611354105604906, + "grad_norm": 1.6655510347415396, + "learning_rate": 1.0882471572137908e-06, + "loss": 0.9485, + "step": 7331 + }, + { + "epoch": 0.6612255940839609, + "grad_norm": 1.798847409620264, + "learning_rate": 1.087727223490744e-06, + "loss": 1.0293, + "step": 7332 + }, + { + "epoch": 0.6613157776074311, + "grad_norm": 1.6530387264591553, + "learning_rate": 1.0872073676078405e-06, + "loss": 1.0451, + "step": 7333 + }, + { + "epoch": 0.6614059611309013, + "grad_norm": 3.0547054514637964, + "learning_rate": 1.0866875896094375e-06, + "loss": 0.939, + "step": 7334 + }, + { + "epoch": 0.6614961446543717, + "grad_norm": 2.9134979110194412, + "learning_rate": 1.0861678895398854e-06, + "loss": 0.9994, + "step": 7335 + }, + { + "epoch": 0.6615863281778419, + "grad_norm": 1.8020772841982295, + "learning_rate": 1.0856482674435286e-06, + "loss": 0.9507, + "step": 7336 + }, + { + "epoch": 0.6616765117013121, + "grad_norm": 0.6703553868514716, + "learning_rate": 1.0851287233647024e-06, + "loss": 0.8115, + "step": 7337 + }, + { + "epoch": 0.6617666952247825, + "grad_norm": 1.6099543593435535, + "learning_rate": 1.084609257347738e-06, + "loss": 1.038, + "step": 7338 + }, + { + "epoch": 0.6618568787482527, + "grad_norm": 1.9632622058702227, + "learning_rate": 1.0840898694369594e-06, + "loss": 0.9622, + "step": 7339 + }, + { + "epoch": 0.661947062271723, + "grad_norm": 1.5078253869014953, + "learning_rate": 1.083570559676683e-06, + "loss": 0.9714, + "step": 7340 + }, + { + "epoch": 0.6620372457951932, + "grad_norm": 2.5113725438464116, + "learning_rate": 1.08305132811122e-06, + "loss": 0.961, + "step": 7341 + }, + { + "epoch": 0.6621274293186635, + "grad_norm": 1.530477913035078, + "learning_rate": 1.0825321747848735e-06, + "loss": 0.984, + "step": 7342 + }, + { + "epoch": 0.6622176128421338, + "grad_norm": 0.711971118456418, + "learning_rate": 1.0820130997419417e-06, + "loss": 0.8815, + "step": 7343 + }, + { + "epoch": 0.662307796365604, + "grad_norm": 1.8258318742647295, + "learning_rate": 1.0814941030267123e-06, + "loss": 0.9734, + "step": 7344 + }, + { + "epoch": 0.6623979798890742, + "grad_norm": 0.6007697591687241, + "learning_rate": 1.080975184683472e-06, + "loss": 0.7997, + "step": 7345 + }, + { + "epoch": 0.6624881634125446, + "grad_norm": 1.535854423503713, + "learning_rate": 1.0804563447564948e-06, + "loss": 1.046, + "step": 7346 + }, + { + "epoch": 0.6625783469360148, + "grad_norm": 1.8181052516941347, + "learning_rate": 1.0799375832900545e-06, + "loss": 0.9252, + "step": 7347 + }, + { + "epoch": 0.662668530459485, + "grad_norm": 2.4282190724416797, + "learning_rate": 1.0794189003284118e-06, + "loss": 1.022, + "step": 7348 + }, + { + "epoch": 0.6627587139829553, + "grad_norm": 1.9887526277516363, + "learning_rate": 1.0789002959158242e-06, + "loss": 0.9981, + "step": 7349 + }, + { + "epoch": 0.6628488975064256, + "grad_norm": 2.51443766182789, + "learning_rate": 1.0783817700965428e-06, + "loss": 0.9343, + "step": 7350 + }, + { + "epoch": 0.6629390810298958, + "grad_norm": 1.5059024393971947, + "learning_rate": 1.0778633229148102e-06, + "loss": 1.0334, + "step": 7351 + }, + { + "epoch": 0.6630292645533661, + "grad_norm": 1.8766728885653647, + "learning_rate": 1.0773449544148645e-06, + "loss": 0.9872, + "step": 7352 + }, + { + "epoch": 0.6631194480768363, + "grad_norm": 1.5539937220560762, + "learning_rate": 1.076826664640934e-06, + "loss": 1.0242, + "step": 7353 + }, + { + "epoch": 0.6632096316003067, + "grad_norm": 0.7428848604321011, + "learning_rate": 1.0763084536372424e-06, + "loss": 0.841, + "step": 7354 + }, + { + "epoch": 0.6632998151237769, + "grad_norm": 1.7603839969048907, + "learning_rate": 1.0757903214480068e-06, + "loss": 0.9945, + "step": 7355 + }, + { + "epoch": 0.6633899986472471, + "grad_norm": 1.7856620222345427, + "learning_rate": 1.0752722681174376e-06, + "loss": 0.9253, + "step": 7356 + }, + { + "epoch": 0.6634801821707174, + "grad_norm": 2.0936126663944328, + "learning_rate": 1.074754293689737e-06, + "loss": 0.9086, + "step": 7357 + }, + { + "epoch": 0.6635703656941877, + "grad_norm": 1.6471103188835385, + "learning_rate": 1.0742363982091023e-06, + "loss": 0.9903, + "step": 7358 + }, + { + "epoch": 0.6636605492176579, + "grad_norm": 2.190030078942658, + "learning_rate": 1.0737185817197215e-06, + "loss": 0.9407, + "step": 7359 + }, + { + "epoch": 0.6637507327411282, + "grad_norm": 1.7531413086049314, + "learning_rate": 1.0732008442657803e-06, + "loss": 0.9468, + "step": 7360 + }, + { + "epoch": 0.6638409162645985, + "grad_norm": 1.5768599774903413, + "learning_rate": 1.0726831858914516e-06, + "loss": 0.9673, + "step": 7361 + }, + { + "epoch": 0.6639310997880687, + "grad_norm": 2.9012850117738274, + "learning_rate": 1.0721656066409084e-06, + "loss": 1.0031, + "step": 7362 + }, + { + "epoch": 0.664021283311539, + "grad_norm": 1.8647765871058428, + "learning_rate": 1.0716481065583108e-06, + "loss": 0.993, + "step": 7363 + }, + { + "epoch": 0.6641114668350092, + "grad_norm": 0.6340320045127894, + "learning_rate": 1.071130685687816e-06, + "loss": 0.8112, + "step": 7364 + }, + { + "epoch": 0.6642016503584796, + "grad_norm": 1.5193649240188702, + "learning_rate": 1.0706133440735723e-06, + "loss": 0.9205, + "step": 7365 + }, + { + "epoch": 0.6642918338819498, + "grad_norm": 1.8035443140713825, + "learning_rate": 1.070096081759723e-06, + "loss": 0.9333, + "step": 7366 + }, + { + "epoch": 0.66438201740542, + "grad_norm": 1.727168188399522, + "learning_rate": 1.069578898790404e-06, + "loss": 1.0846, + "step": 7367 + }, + { + "epoch": 0.6644722009288903, + "grad_norm": 3.0077107201732547, + "learning_rate": 1.0690617952097424e-06, + "loss": 0.9662, + "step": 7368 + }, + { + "epoch": 0.6645623844523606, + "grad_norm": 1.8872271814294568, + "learning_rate": 1.068544771061863e-06, + "loss": 1.0315, + "step": 7369 + }, + { + "epoch": 0.6646525679758308, + "grad_norm": 1.9195353835934328, + "learning_rate": 1.0680278263908787e-06, + "loss": 1.0219, + "step": 7370 + }, + { + "epoch": 0.6647427514993011, + "grad_norm": 0.7376016789188641, + "learning_rate": 1.0675109612408991e-06, + "loss": 0.8211, + "step": 7371 + }, + { + "epoch": 0.6648329350227713, + "grad_norm": 2.188452703915296, + "learning_rate": 1.0669941756560264e-06, + "loss": 0.9536, + "step": 7372 + }, + { + "epoch": 0.6649231185462416, + "grad_norm": 1.8351762207815416, + "learning_rate": 1.0664774696803548e-06, + "loss": 0.9327, + "step": 7373 + }, + { + "epoch": 0.6650133020697119, + "grad_norm": 2.3439490678806574, + "learning_rate": 1.065960843357973e-06, + "loss": 0.9563, + "step": 7374 + }, + { + "epoch": 0.6651034855931821, + "grad_norm": 1.5822186394372897, + "learning_rate": 1.065444296732963e-06, + "loss": 0.9994, + "step": 7375 + }, + { + "epoch": 0.6651936691166523, + "grad_norm": 0.674253710816115, + "learning_rate": 1.064927829849397e-06, + "loss": 0.8233, + "step": 7376 + }, + { + "epoch": 0.6652838526401227, + "grad_norm": 1.956596828830352, + "learning_rate": 1.0644114427513465e-06, + "loss": 0.9803, + "step": 7377 + }, + { + "epoch": 0.6653740361635929, + "grad_norm": 2.6729514647250667, + "learning_rate": 1.0638951354828693e-06, + "loss": 0.9138, + "step": 7378 + }, + { + "epoch": 0.6654642196870632, + "grad_norm": 1.8490945971786261, + "learning_rate": 1.063378908088021e-06, + "loss": 1.0363, + "step": 7379 + }, + { + "epoch": 0.6655544032105334, + "grad_norm": 1.6041025256260046, + "learning_rate": 1.0628627606108486e-06, + "loss": 0.9754, + "step": 7380 + }, + { + "epoch": 0.6656445867340037, + "grad_norm": 2.905612178256439, + "learning_rate": 1.062346693095393e-06, + "loss": 1.049, + "step": 7381 + }, + { + "epoch": 0.665734770257474, + "grad_norm": 1.5752771194320005, + "learning_rate": 1.0618307055856882e-06, + "loss": 0.9209, + "step": 7382 + }, + { + "epoch": 0.6658249537809442, + "grad_norm": 2.553172090766139, + "learning_rate": 1.061314798125759e-06, + "loss": 1.0332, + "step": 7383 + }, + { + "epoch": 0.6659151373044144, + "grad_norm": 1.397702455476524, + "learning_rate": 1.0607989707596293e-06, + "loss": 0.9403, + "step": 7384 + }, + { + "epoch": 0.6660053208278848, + "grad_norm": 1.413856027608904, + "learning_rate": 1.0602832235313078e-06, + "loss": 1.11, + "step": 7385 + }, + { + "epoch": 0.666095504351355, + "grad_norm": 1.505299663302866, + "learning_rate": 1.0597675564848053e-06, + "loss": 0.9711, + "step": 7386 + }, + { + "epoch": 0.6661856878748252, + "grad_norm": 2.0899288552326567, + "learning_rate": 1.059251969664118e-06, + "loss": 1.0535, + "step": 7387 + }, + { + "epoch": 0.6662758713982956, + "grad_norm": 1.7701670197172645, + "learning_rate": 1.0587364631132402e-06, + "loss": 0.9887, + "step": 7388 + }, + { + "epoch": 0.6663660549217658, + "grad_norm": 1.5462265967630997, + "learning_rate": 1.0582210368761573e-06, + "loss": 1.0232, + "step": 7389 + }, + { + "epoch": 0.666456238445236, + "grad_norm": 2.19642214713226, + "learning_rate": 1.0577056909968485e-06, + "loss": 0.9868, + "step": 7390 + }, + { + "epoch": 0.6665464219687063, + "grad_norm": 1.9264457128911774, + "learning_rate": 1.0571904255192857e-06, + "loss": 1.0015, + "step": 7391 + }, + { + "epoch": 0.6666366054921766, + "grad_norm": 2.1874552040962985, + "learning_rate": 1.0566752404874354e-06, + "loss": 1.0146, + "step": 7392 + }, + { + "epoch": 0.6667267890156469, + "grad_norm": 1.7025335842410716, + "learning_rate": 1.0561601359452543e-06, + "loss": 0.9786, + "step": 7393 + }, + { + "epoch": 0.6668169725391171, + "grad_norm": 1.6537392497271923, + "learning_rate": 1.0556451119366947e-06, + "loss": 0.9205, + "step": 7394 + }, + { + "epoch": 0.6669071560625873, + "grad_norm": 1.5066267500009283, + "learning_rate": 1.0551301685057011e-06, + "loss": 0.9202, + "step": 7395 + }, + { + "epoch": 0.6669973395860577, + "grad_norm": 1.6784894352166506, + "learning_rate": 1.0546153056962117e-06, + "loss": 1.0656, + "step": 7396 + }, + { + "epoch": 0.6670875231095279, + "grad_norm": 1.8541470930230746, + "learning_rate": 1.0541005235521578e-06, + "loss": 0.9817, + "step": 7397 + }, + { + "epoch": 0.6671777066329981, + "grad_norm": 2.172208362490685, + "learning_rate": 1.0535858221174614e-06, + "loss": 0.9952, + "step": 7398 + }, + { + "epoch": 0.6672678901564684, + "grad_norm": 1.3540944984963768, + "learning_rate": 1.0530712014360426e-06, + "loss": 1.0218, + "step": 7399 + }, + { + "epoch": 0.6673580736799387, + "grad_norm": 1.6219339055001702, + "learning_rate": 1.0525566615518088e-06, + "loss": 1.0322, + "step": 7400 + }, + { + "epoch": 0.667448257203409, + "grad_norm": 1.523318789821012, + "learning_rate": 1.0520422025086662e-06, + "loss": 0.9962, + "step": 7401 + }, + { + "epoch": 0.6675384407268792, + "grad_norm": 1.8978364723031982, + "learning_rate": 1.0515278243505092e-06, + "loss": 1.0347, + "step": 7402 + }, + { + "epoch": 0.6676286242503494, + "grad_norm": 1.3644311640490416, + "learning_rate": 1.0510135271212278e-06, + "loss": 1.0123, + "step": 7403 + }, + { + "epoch": 0.6677188077738198, + "grad_norm": 1.7188099937372143, + "learning_rate": 1.0504993108647052e-06, + "loss": 1.0452, + "step": 7404 + }, + { + "epoch": 0.66780899129729, + "grad_norm": 1.5172171341332603, + "learning_rate": 1.0499851756248168e-06, + "loss": 0.9523, + "step": 7405 + }, + { + "epoch": 0.6678991748207602, + "grad_norm": 1.5646138197988635, + "learning_rate": 1.0494711214454316e-06, + "loss": 0.966, + "step": 7406 + }, + { + "epoch": 0.6679893583442305, + "grad_norm": 1.887037546460512, + "learning_rate": 1.0489571483704111e-06, + "loss": 0.9007, + "step": 7407 + }, + { + "epoch": 0.6680795418677008, + "grad_norm": 2.5472433373830876, + "learning_rate": 1.048443256443612e-06, + "loss": 0.9003, + "step": 7408 + }, + { + "epoch": 0.668169725391171, + "grad_norm": 1.6156988986618408, + "learning_rate": 1.0479294457088801e-06, + "loss": 1.0006, + "step": 7409 + }, + { + "epoch": 0.6682599089146413, + "grad_norm": 1.457055058425412, + "learning_rate": 1.0474157162100574e-06, + "loss": 1.1047, + "step": 7410 + }, + { + "epoch": 0.6683500924381116, + "grad_norm": 2.223933009078652, + "learning_rate": 1.0469020679909786e-06, + "loss": 0.9194, + "step": 7411 + }, + { + "epoch": 0.6684402759615818, + "grad_norm": 2.795374377159128, + "learning_rate": 1.0463885010954705e-06, + "loss": 0.9091, + "step": 7412 + }, + { + "epoch": 0.6685304594850521, + "grad_norm": 5.931952525138586, + "learning_rate": 1.0458750155673536e-06, + "loss": 1.052, + "step": 7413 + }, + { + "epoch": 0.6686206430085223, + "grad_norm": 1.6457548323955398, + "learning_rate": 1.0453616114504421e-06, + "loss": 0.9407, + "step": 7414 + }, + { + "epoch": 0.6687108265319927, + "grad_norm": 2.2023062841849113, + "learning_rate": 1.0448482887885406e-06, + "loss": 1.0116, + "step": 7415 + }, + { + "epoch": 0.6688010100554629, + "grad_norm": 2.0523142289389846, + "learning_rate": 1.044335047625451e-06, + "loss": 0.9889, + "step": 7416 + }, + { + "epoch": 0.6688911935789331, + "grad_norm": 2.5028548153183996, + "learning_rate": 1.0438218880049637e-06, + "loss": 0.9922, + "step": 7417 + }, + { + "epoch": 0.6689813771024034, + "grad_norm": 2.1166775360228836, + "learning_rate": 1.0433088099708653e-06, + "loss": 1.0445, + "step": 7418 + }, + { + "epoch": 0.6690715606258737, + "grad_norm": 0.5942346953610608, + "learning_rate": 1.0427958135669346e-06, + "loss": 0.7492, + "step": 7419 + }, + { + "epoch": 0.6691617441493439, + "grad_norm": 1.7968119071166782, + "learning_rate": 1.0422828988369428e-06, + "loss": 0.9178, + "step": 7420 + }, + { + "epoch": 0.6692519276728142, + "grad_norm": 1.5751931934850905, + "learning_rate": 1.041770065824655e-06, + "loss": 1.0502, + "step": 7421 + }, + { + "epoch": 0.6693421111962844, + "grad_norm": 1.6307475342919777, + "learning_rate": 1.0412573145738287e-06, + "loss": 0.8374, + "step": 7422 + }, + { + "epoch": 0.6694322947197547, + "grad_norm": 1.6516661641222474, + "learning_rate": 1.040744645128216e-06, + "loss": 0.9574, + "step": 7423 + }, + { + "epoch": 0.669522478243225, + "grad_norm": 1.6234200205885354, + "learning_rate": 1.040232057531558e-06, + "loss": 1.0411, + "step": 7424 + }, + { + "epoch": 0.6696126617666952, + "grad_norm": 2.7352609997336303, + "learning_rate": 1.0397195518275932e-06, + "loss": 0.9534, + "step": 7425 + }, + { + "epoch": 0.6697028452901654, + "grad_norm": 2.1366755674695534, + "learning_rate": 1.0392071280600512e-06, + "loss": 0.9563, + "step": 7426 + }, + { + "epoch": 0.6697930288136358, + "grad_norm": 1.5970509322087914, + "learning_rate": 1.0386947862726549e-06, + "loss": 0.9815, + "step": 7427 + }, + { + "epoch": 0.669883212337106, + "grad_norm": 1.8510038925492684, + "learning_rate": 1.0381825265091197e-06, + "loss": 0.9064, + "step": 7428 + }, + { + "epoch": 0.6699733958605762, + "grad_norm": 3.1451850543138007, + "learning_rate": 1.037670348813155e-06, + "loss": 0.905, + "step": 7429 + }, + { + "epoch": 0.6700635793840465, + "grad_norm": 1.578556794873635, + "learning_rate": 1.0371582532284624e-06, + "loss": 1.0135, + "step": 7430 + }, + { + "epoch": 0.6701537629075168, + "grad_norm": 1.6325763732430516, + "learning_rate": 1.0366462397987375e-06, + "loss": 0.9156, + "step": 7431 + }, + { + "epoch": 0.6702439464309871, + "grad_norm": 2.2034463850942374, + "learning_rate": 1.0361343085676665e-06, + "loss": 0.9892, + "step": 7432 + }, + { + "epoch": 0.6703341299544573, + "grad_norm": 1.4153348983370273, + "learning_rate": 1.0356224595789309e-06, + "loss": 1.0992, + "step": 7433 + }, + { + "epoch": 0.6704243134779276, + "grad_norm": 1.510070059458387, + "learning_rate": 1.0351106928762046e-06, + "loss": 1.0435, + "step": 7434 + }, + { + "epoch": 0.6705144970013979, + "grad_norm": 2.1485762117863083, + "learning_rate": 1.034599008503154e-06, + "loss": 1.0572, + "step": 7435 + }, + { + "epoch": 0.6706046805248681, + "grad_norm": 1.7593295862221334, + "learning_rate": 1.0340874065034406e-06, + "loss": 0.9728, + "step": 7436 + }, + { + "epoch": 0.6706948640483383, + "grad_norm": 2.2231922006303466, + "learning_rate": 1.0335758869207137e-06, + "loss": 0.9054, + "step": 7437 + }, + { + "epoch": 0.6707850475718087, + "grad_norm": 1.9752627837724948, + "learning_rate": 1.0330644497986227e-06, + "loss": 1.0046, + "step": 7438 + }, + { + "epoch": 0.6708752310952789, + "grad_norm": 1.5822776320008523, + "learning_rate": 1.0325530951808029e-06, + "loss": 1.0226, + "step": 7439 + }, + { + "epoch": 0.6709654146187491, + "grad_norm": 1.6722936373500572, + "learning_rate": 1.0320418231108887e-06, + "loss": 0.9899, + "step": 7440 + }, + { + "epoch": 0.6710555981422194, + "grad_norm": 2.5355317435770286, + "learning_rate": 1.0315306336325028e-06, + "loss": 1.0237, + "step": 7441 + }, + { + "epoch": 0.6711457816656897, + "grad_norm": 1.7145587102049746, + "learning_rate": 1.0310195267892635e-06, + "loss": 0.9415, + "step": 7442 + }, + { + "epoch": 0.67123596518916, + "grad_norm": 1.936481577333755, + "learning_rate": 1.030508502624781e-06, + "loss": 0.9263, + "step": 7443 + }, + { + "epoch": 0.6713261487126302, + "grad_norm": 1.5289747500583253, + "learning_rate": 1.0299975611826587e-06, + "loss": 0.935, + "step": 7444 + }, + { + "epoch": 0.6714163322361004, + "grad_norm": 1.4975112295605373, + "learning_rate": 1.0294867025064928e-06, + "loss": 0.9918, + "step": 7445 + }, + { + "epoch": 0.6715065157595708, + "grad_norm": 1.4871717908349802, + "learning_rate": 1.028975926639874e-06, + "loss": 0.9931, + "step": 7446 + }, + { + "epoch": 0.671596699283041, + "grad_norm": 3.1512117870636227, + "learning_rate": 1.0284652336263823e-06, + "loss": 0.9097, + "step": 7447 + }, + { + "epoch": 0.6716868828065112, + "grad_norm": 3.162410050270174, + "learning_rate": 1.0279546235095938e-06, + "loss": 0.9159, + "step": 7448 + }, + { + "epoch": 0.6717770663299815, + "grad_norm": 0.7300767713833216, + "learning_rate": 1.0274440963330768e-06, + "loss": 0.8459, + "step": 7449 + }, + { + "epoch": 0.6718672498534518, + "grad_norm": 1.7973894129280756, + "learning_rate": 1.0269336521403919e-06, + "loss": 0.9822, + "step": 7450 + }, + { + "epoch": 0.671957433376922, + "grad_norm": 1.8008557881384146, + "learning_rate": 1.0264232909750936e-06, + "loss": 1.1138, + "step": 7451 + }, + { + "epoch": 0.6720476169003923, + "grad_norm": 1.5715623123675537, + "learning_rate": 1.025913012880728e-06, + "loss": 1.0447, + "step": 7452 + }, + { + "epoch": 0.6721378004238625, + "grad_norm": 1.4901332595938044, + "learning_rate": 1.0254028179008362e-06, + "loss": 0.8942, + "step": 7453 + }, + { + "epoch": 0.6722279839473329, + "grad_norm": 1.453495757683021, + "learning_rate": 1.0248927060789483e-06, + "loss": 0.8739, + "step": 7454 + }, + { + "epoch": 0.6723181674708031, + "grad_norm": 3.1414857510418575, + "learning_rate": 1.0243826774585928e-06, + "loss": 0.9437, + "step": 7455 + }, + { + "epoch": 0.6724083509942733, + "grad_norm": 1.7428728474796116, + "learning_rate": 1.0238727320832854e-06, + "loss": 0.9646, + "step": 7456 + }, + { + "epoch": 0.6724985345177437, + "grad_norm": 4.125743452179634, + "learning_rate": 1.0233628699965403e-06, + "loss": 0.9853, + "step": 7457 + }, + { + "epoch": 0.6725887180412139, + "grad_norm": 1.6495944478404962, + "learning_rate": 1.0228530912418594e-06, + "loss": 1.0067, + "step": 7458 + }, + { + "epoch": 0.6726789015646841, + "grad_norm": 1.6100378060492737, + "learning_rate": 1.0223433958627404e-06, + "loss": 0.9514, + "step": 7459 + }, + { + "epoch": 0.6727690850881544, + "grad_norm": 4.638075825191897, + "learning_rate": 1.021833783902674e-06, + "loss": 1.0073, + "step": 7460 + }, + { + "epoch": 0.6728592686116247, + "grad_norm": 2.0739026304690644, + "learning_rate": 1.0213242554051427e-06, + "loss": 0.9874, + "step": 7461 + }, + { + "epoch": 0.6729494521350949, + "grad_norm": 1.6833954802771625, + "learning_rate": 1.0208148104136229e-06, + "loss": 0.9214, + "step": 7462 + }, + { + "epoch": 0.6730396356585652, + "grad_norm": 1.8114628948777955, + "learning_rate": 1.020305448971582e-06, + "loss": 0.9553, + "step": 7463 + }, + { + "epoch": 0.6731298191820354, + "grad_norm": 1.869182845289555, + "learning_rate": 1.0197961711224824e-06, + "loss": 0.9498, + "step": 7464 + }, + { + "epoch": 0.6732200027055057, + "grad_norm": 2.1545147755351475, + "learning_rate": 1.0192869769097777e-06, + "loss": 0.9935, + "step": 7465 + }, + { + "epoch": 0.673310186228976, + "grad_norm": 2.0307583727468526, + "learning_rate": 1.018777866376916e-06, + "loss": 0.9902, + "step": 7466 + }, + { + "epoch": 0.6734003697524462, + "grad_norm": 1.9140473657126482, + "learning_rate": 1.0182688395673374e-06, + "loss": 0.8533, + "step": 7467 + }, + { + "epoch": 0.6734905532759164, + "grad_norm": 1.5155234548628338, + "learning_rate": 1.017759896524475e-06, + "loss": 0.9482, + "step": 7468 + }, + { + "epoch": 0.6735807367993868, + "grad_norm": 2.129953333383757, + "learning_rate": 1.0172510372917528e-06, + "loss": 1.0255, + "step": 7469 + }, + { + "epoch": 0.673670920322857, + "grad_norm": 2.2411615769346533, + "learning_rate": 1.0167422619125925e-06, + "loss": 0.8862, + "step": 7470 + }, + { + "epoch": 0.6737611038463273, + "grad_norm": 2.117619769459709, + "learning_rate": 1.0162335704304026e-06, + "loss": 0.9886, + "step": 7471 + }, + { + "epoch": 0.6738512873697975, + "grad_norm": 0.6919730282595994, + "learning_rate": 1.0157249628885903e-06, + "loss": 0.8295, + "step": 7472 + }, + { + "epoch": 0.6739414708932678, + "grad_norm": 1.2066117332311659, + "learning_rate": 1.0152164393305506e-06, + "loss": 0.8943, + "step": 7473 + }, + { + "epoch": 0.6740316544167381, + "grad_norm": 1.7633714726343082, + "learning_rate": 1.0147079997996746e-06, + "loss": 0.9771, + "step": 7474 + }, + { + "epoch": 0.6741218379402083, + "grad_norm": 1.5671198639199415, + "learning_rate": 1.0141996443393446e-06, + "loss": 1.0137, + "step": 7475 + }, + { + "epoch": 0.6742120214636785, + "grad_norm": 1.5276082387931265, + "learning_rate": 1.0136913729929369e-06, + "loss": 0.9665, + "step": 7476 + }, + { + "epoch": 0.6743022049871489, + "grad_norm": 1.6733506722457483, + "learning_rate": 1.0131831858038203e-06, + "loss": 0.945, + "step": 7477 + }, + { + "epoch": 0.6743923885106191, + "grad_norm": 1.5218909605470023, + "learning_rate": 1.0126750828153538e-06, + "loss": 1.0471, + "step": 7478 + }, + { + "epoch": 0.6744825720340893, + "grad_norm": 2.34659698505694, + "learning_rate": 1.012167064070895e-06, + "loss": 1.0273, + "step": 7479 + }, + { + "epoch": 0.6745727555575597, + "grad_norm": 1.901320101757922, + "learning_rate": 1.0116591296137885e-06, + "loss": 1.0096, + "step": 7480 + }, + { + "epoch": 0.6746629390810299, + "grad_norm": 1.6988450615048059, + "learning_rate": 1.0111512794873746e-06, + "loss": 1.0109, + "step": 7481 + }, + { + "epoch": 0.6747531226045002, + "grad_norm": 1.4983859915527096, + "learning_rate": 1.010643513734986e-06, + "loss": 0.9831, + "step": 7482 + }, + { + "epoch": 0.6748433061279704, + "grad_norm": 1.5997518585427968, + "learning_rate": 1.010135832399948e-06, + "loss": 1.0238, + "step": 7483 + }, + { + "epoch": 0.6749334896514407, + "grad_norm": 1.5064963491492438, + "learning_rate": 1.0096282355255792e-06, + "loss": 0.9913, + "step": 7484 + }, + { + "epoch": 0.675023673174911, + "grad_norm": 2.127213951631195, + "learning_rate": 1.0091207231551905e-06, + "loss": 1.0385, + "step": 7485 + }, + { + "epoch": 0.6751138566983812, + "grad_norm": 1.6285201605397157, + "learning_rate": 1.0086132953320842e-06, + "loss": 0.9476, + "step": 7486 + }, + { + "epoch": 0.6752040402218514, + "grad_norm": 1.9599795490287861, + "learning_rate": 1.0081059520995591e-06, + "loss": 1.009, + "step": 7487 + }, + { + "epoch": 0.6752942237453218, + "grad_norm": 2.3510702131163956, + "learning_rate": 1.0075986935009028e-06, + "loss": 1.0528, + "step": 7488 + }, + { + "epoch": 0.675384407268792, + "grad_norm": 1.5178737671526752, + "learning_rate": 1.0070915195793982e-06, + "loss": 0.9943, + "step": 7489 + }, + { + "epoch": 0.6754745907922622, + "grad_norm": 1.8120446126342695, + "learning_rate": 1.0065844303783197e-06, + "loss": 0.9603, + "step": 7490 + }, + { + "epoch": 0.6755647743157325, + "grad_norm": 1.4735439063897036, + "learning_rate": 1.0060774259409356e-06, + "loss": 0.8997, + "step": 7491 + }, + { + "epoch": 0.6756549578392028, + "grad_norm": 1.4037618453493304, + "learning_rate": 1.0055705063105065e-06, + "loss": 1.0193, + "step": 7492 + }, + { + "epoch": 0.675745141362673, + "grad_norm": 1.672319299718197, + "learning_rate": 1.0050636715302837e-06, + "loss": 0.9774, + "step": 7493 + }, + { + "epoch": 0.6758353248861433, + "grad_norm": 1.7873970508939965, + "learning_rate": 1.0045569216435157e-06, + "loss": 0.9945, + "step": 7494 + }, + { + "epoch": 0.6759255084096135, + "grad_norm": 2.9807741501154665, + "learning_rate": 1.0040502566934384e-06, + "loss": 1.0669, + "step": 7495 + }, + { + "epoch": 0.6760156919330839, + "grad_norm": 1.7922728274585498, + "learning_rate": 1.0035436767232866e-06, + "loss": 0.9213, + "step": 7496 + }, + { + "epoch": 0.6761058754565541, + "grad_norm": 1.651999579764398, + "learning_rate": 1.0030371817762816e-06, + "loss": 0.9973, + "step": 7497 + }, + { + "epoch": 0.6761960589800243, + "grad_norm": 1.513803941441285, + "learning_rate": 1.0025307718956417e-06, + "loss": 0.9946, + "step": 7498 + }, + { + "epoch": 0.6762862425034946, + "grad_norm": 1.517583937860891, + "learning_rate": 1.0020244471245765e-06, + "loss": 1.0049, + "step": 7499 + }, + { + "epoch": 0.6763764260269649, + "grad_norm": 1.5966920811350034, + "learning_rate": 1.001518207506288e-06, + "loss": 0.9424, + "step": 7500 + }, + { + "epoch": 0.6764666095504351, + "grad_norm": 1.4765102639621477, + "learning_rate": 1.0010120530839717e-06, + "loss": 1.0511, + "step": 7501 + }, + { + "epoch": 0.6765567930739054, + "grad_norm": 1.3084933569815678, + "learning_rate": 1.0005059839008161e-06, + "loss": 1.0102, + "step": 7502 + }, + { + "epoch": 0.6766469765973756, + "grad_norm": 1.7220192075183167, + "learning_rate": 1.0000000000000004e-06, + "loss": 0.9357, + "step": 7503 + }, + { + "epoch": 0.676737160120846, + "grad_norm": 1.7347336819204644, + "learning_rate": 9.994941014246985e-07, + "loss": 1.0279, + "step": 7504 + }, + { + "epoch": 0.6768273436443162, + "grad_norm": 1.789936568452257, + "learning_rate": 9.989882882180766e-07, + "loss": 0.8747, + "step": 7505 + }, + { + "epoch": 0.6769175271677864, + "grad_norm": 2.060950766658814, + "learning_rate": 9.984825604232938e-07, + "loss": 1.0004, + "step": 7506 + }, + { + "epoch": 0.6770077106912568, + "grad_norm": 1.5547583482404395, + "learning_rate": 9.97976918083502e-07, + "loss": 0.9255, + "step": 7507 + }, + { + "epoch": 0.677097894214727, + "grad_norm": 1.3774366463217562, + "learning_rate": 9.974713612418427e-07, + "loss": 1.0549, + "step": 7508 + }, + { + "epoch": 0.6771880777381972, + "grad_norm": 2.4438517049323356, + "learning_rate": 9.969658899414563e-07, + "loss": 0.9959, + "step": 7509 + }, + { + "epoch": 0.6772782612616675, + "grad_norm": 1.673529474774449, + "learning_rate": 9.964605042254696e-07, + "loss": 0.9026, + "step": 7510 + }, + { + "epoch": 0.6773684447851378, + "grad_norm": 1.6160068348513428, + "learning_rate": 9.959552041370076e-07, + "loss": 0.9256, + "step": 7511 + }, + { + "epoch": 0.677458628308608, + "grad_norm": 1.4307656313935309, + "learning_rate": 9.954499897191824e-07, + "loss": 0.9038, + "step": 7512 + }, + { + "epoch": 0.6775488118320783, + "grad_norm": 2.217459397315192, + "learning_rate": 9.949448610151043e-07, + "loss": 0.938, + "step": 7513 + }, + { + "epoch": 0.6776389953555485, + "grad_norm": 1.643516317952115, + "learning_rate": 9.944398180678719e-07, + "loss": 0.9635, + "step": 7514 + }, + { + "epoch": 0.6777291788790188, + "grad_norm": 1.3636849141870253, + "learning_rate": 9.939348609205789e-07, + "loss": 0.935, + "step": 7515 + }, + { + "epoch": 0.6778193624024891, + "grad_norm": 1.3787075256141157, + "learning_rate": 9.93429989616311e-07, + "loss": 0.9911, + "step": 7516 + }, + { + "epoch": 0.6779095459259593, + "grad_norm": 1.612864147143557, + "learning_rate": 9.929252041981464e-07, + "loss": 0.902, + "step": 7517 + }, + { + "epoch": 0.6779997294494295, + "grad_norm": 1.7095529188214609, + "learning_rate": 9.924205047091572e-07, + "loss": 0.956, + "step": 7518 + }, + { + "epoch": 0.6780899129728999, + "grad_norm": 2.9596678898052624, + "learning_rate": 9.919158911924056e-07, + "loss": 1.1046, + "step": 7519 + }, + { + "epoch": 0.6781800964963701, + "grad_norm": 2.120398757068361, + "learning_rate": 9.914113636909483e-07, + "loss": 1.021, + "step": 7520 + }, + { + "epoch": 0.6782702800198404, + "grad_norm": 0.7385466567831754, + "learning_rate": 9.90906922247835e-07, + "loss": 0.8645, + "step": 7521 + }, + { + "epoch": 0.6783604635433106, + "grad_norm": 2.2212965414809687, + "learning_rate": 9.904025669061072e-07, + "loss": 1.0141, + "step": 7522 + }, + { + "epoch": 0.6784506470667809, + "grad_norm": 1.4526890746867458, + "learning_rate": 9.89898297708799e-07, + "loss": 0.9807, + "step": 7523 + }, + { + "epoch": 0.6785408305902512, + "grad_norm": 2.030634860281635, + "learning_rate": 9.893941146989388e-07, + "loss": 0.9323, + "step": 7524 + }, + { + "epoch": 0.6786310141137214, + "grad_norm": 2.0039138645079504, + "learning_rate": 9.888900179195437e-07, + "loss": 0.943, + "step": 7525 + }, + { + "epoch": 0.6787211976371916, + "grad_norm": 1.6393833366411856, + "learning_rate": 9.883860074136285e-07, + "loss": 1.0542, + "step": 7526 + }, + { + "epoch": 0.678811381160662, + "grad_norm": 15.952785447231172, + "learning_rate": 9.87882083224196e-07, + "loss": 0.9805, + "step": 7527 + }, + { + "epoch": 0.6789015646841322, + "grad_norm": 0.6982068459696676, + "learning_rate": 9.873782453942462e-07, + "loss": 0.8417, + "step": 7528 + }, + { + "epoch": 0.6789917482076024, + "grad_norm": 2.046612788287488, + "learning_rate": 9.868744939667676e-07, + "loss": 0.9093, + "step": 7529 + }, + { + "epoch": 0.6790819317310728, + "grad_norm": 1.7988466143822783, + "learning_rate": 9.863708289847432e-07, + "loss": 0.8656, + "step": 7530 + }, + { + "epoch": 0.679172115254543, + "grad_norm": 2.3427233672744485, + "learning_rate": 9.85867250491149e-07, + "loss": 0.8947, + "step": 7531 + }, + { + "epoch": 0.6792622987780133, + "grad_norm": 2.383823868037587, + "learning_rate": 9.853637585289528e-07, + "loss": 1.0912, + "step": 7532 + }, + { + "epoch": 0.6793524823014835, + "grad_norm": 1.5279262046769395, + "learning_rate": 9.848603531411159e-07, + "loss": 0.9754, + "step": 7533 + }, + { + "epoch": 0.6794426658249538, + "grad_norm": 2.0971152746399455, + "learning_rate": 9.843570343705899e-07, + "loss": 0.9908, + "step": 7534 + }, + { + "epoch": 0.6795328493484241, + "grad_norm": 1.3607056342480326, + "learning_rate": 9.83853802260323e-07, + "loss": 1.0545, + "step": 7535 + }, + { + "epoch": 0.6796230328718943, + "grad_norm": 2.251616956593632, + "learning_rate": 9.833506568532524e-07, + "loss": 0.9256, + "step": 7536 + }, + { + "epoch": 0.6797132163953645, + "grad_norm": 1.5586560076753468, + "learning_rate": 9.828475981923093e-07, + "loss": 0.831, + "step": 7537 + }, + { + "epoch": 0.6798033999188349, + "grad_norm": 2.3697357059341035, + "learning_rate": 9.823446263204175e-07, + "loss": 0.9612, + "step": 7538 + }, + { + "epoch": 0.6798935834423051, + "grad_norm": 2.142171148214112, + "learning_rate": 9.818417412804937e-07, + "loss": 0.9968, + "step": 7539 + }, + { + "epoch": 0.6799837669657753, + "grad_norm": 3.1922525616782744, + "learning_rate": 9.813389431154463e-07, + "loss": 1.0655, + "step": 7540 + }, + { + "epoch": 0.6800739504892456, + "grad_norm": 1.9174995547695712, + "learning_rate": 9.808362318681783e-07, + "loss": 1.0115, + "step": 7541 + }, + { + "epoch": 0.6801641340127159, + "grad_norm": 1.8998833946786984, + "learning_rate": 9.803336075815807e-07, + "loss": 1.0052, + "step": 7542 + }, + { + "epoch": 0.6802543175361861, + "grad_norm": 2.1346208779838656, + "learning_rate": 9.79831070298544e-07, + "loss": 0.9127, + "step": 7543 + }, + { + "epoch": 0.6803445010596564, + "grad_norm": 1.9628328077616763, + "learning_rate": 9.793286200619443e-07, + "loss": 0.952, + "step": 7544 + }, + { + "epoch": 0.6804346845831266, + "grad_norm": 1.473442454648309, + "learning_rate": 9.78826256914655e-07, + "loss": 1.0462, + "step": 7545 + }, + { + "epoch": 0.680524868106597, + "grad_norm": 1.7720364317108914, + "learning_rate": 9.7832398089954e-07, + "loss": 1.0357, + "step": 7546 + }, + { + "epoch": 0.6806150516300672, + "grad_norm": 1.7320562447760606, + "learning_rate": 9.778217920594565e-07, + "loss": 1.0854, + "step": 7547 + }, + { + "epoch": 0.6807052351535374, + "grad_norm": 1.630619163868272, + "learning_rate": 9.773196904372547e-07, + "loss": 1.0293, + "step": 7548 + }, + { + "epoch": 0.6807954186770077, + "grad_norm": 1.8980943032082098, + "learning_rate": 9.768176760757742e-07, + "loss": 1.0717, + "step": 7549 + }, + { + "epoch": 0.680885602200478, + "grad_norm": 1.782066559598534, + "learning_rate": 9.76315749017853e-07, + "loss": 0.963, + "step": 7550 + }, + { + "epoch": 0.6809757857239482, + "grad_norm": 2.051856233323325, + "learning_rate": 9.758139093063161e-07, + "loss": 0.9848, + "step": 7551 + }, + { + "epoch": 0.6810659692474185, + "grad_norm": 1.7126564901090915, + "learning_rate": 9.753121569839834e-07, + "loss": 1.0534, + "step": 7552 + }, + { + "epoch": 0.6811561527708888, + "grad_norm": 0.7459308386490455, + "learning_rate": 9.748104920936678e-07, + "loss": 0.7854, + "step": 7553 + }, + { + "epoch": 0.681246336294359, + "grad_norm": 13.449697520885751, + "learning_rate": 9.743089146781738e-07, + "loss": 0.9635, + "step": 7554 + }, + { + "epoch": 0.6813365198178293, + "grad_norm": 4.543152848873516, + "learning_rate": 9.738074247802988e-07, + "loss": 0.9105, + "step": 7555 + }, + { + "epoch": 0.6814267033412995, + "grad_norm": 1.9464056781107422, + "learning_rate": 9.733060224428325e-07, + "loss": 1.0012, + "step": 7556 + }, + { + "epoch": 0.6815168868647699, + "grad_norm": 1.7385711771031993, + "learning_rate": 9.728047077085577e-07, + "loss": 0.9937, + "step": 7557 + }, + { + "epoch": 0.6816070703882401, + "grad_norm": 2.162518773521122, + "learning_rate": 9.723034806202497e-07, + "loss": 0.8834, + "step": 7558 + }, + { + "epoch": 0.6816972539117103, + "grad_norm": 1.9742774873124014, + "learning_rate": 9.718023412206748e-07, + "loss": 1.0025, + "step": 7559 + }, + { + "epoch": 0.6817874374351806, + "grad_norm": 1.4017519751559173, + "learning_rate": 9.713012895525935e-07, + "loss": 1.0074, + "step": 7560 + }, + { + "epoch": 0.6818776209586509, + "grad_norm": 2.1029472468675743, + "learning_rate": 9.708003256587584e-07, + "loss": 0.9704, + "step": 7561 + }, + { + "epoch": 0.6819678044821211, + "grad_norm": 1.4566657295631402, + "learning_rate": 9.702994495819147e-07, + "loss": 1.0346, + "step": 7562 + }, + { + "epoch": 0.6820579880055914, + "grad_norm": 2.344021794136462, + "learning_rate": 9.697986613647999e-07, + "loss": 0.8897, + "step": 7563 + }, + { + "epoch": 0.6821481715290616, + "grad_norm": 0.69497848374522, + "learning_rate": 9.692979610501425e-07, + "loss": 0.8497, + "step": 7564 + }, + { + "epoch": 0.6822383550525319, + "grad_norm": 1.9225422390984168, + "learning_rate": 9.68797348680668e-07, + "loss": 1.058, + "step": 7565 + }, + { + "epoch": 0.6823285385760022, + "grad_norm": 1.4327102974120152, + "learning_rate": 9.682968242990878e-07, + "loss": 0.9542, + "step": 7566 + }, + { + "epoch": 0.6824187220994724, + "grad_norm": 1.4587943937710104, + "learning_rate": 9.677963879481132e-07, + "loss": 1.0348, + "step": 7567 + }, + { + "epoch": 0.6825089056229426, + "grad_norm": 1.953308951298389, + "learning_rate": 9.672960396704416e-07, + "loss": 0.9733, + "step": 7568 + }, + { + "epoch": 0.682599089146413, + "grad_norm": 1.4751786301633112, + "learning_rate": 9.667957795087657e-07, + "loss": 0.9886, + "step": 7569 + }, + { + "epoch": 0.6826892726698832, + "grad_norm": 1.9085234545934715, + "learning_rate": 9.662956075057712e-07, + "loss": 1.0504, + "step": 7570 + }, + { + "epoch": 0.6827794561933535, + "grad_norm": 3.584542462295776, + "learning_rate": 9.657955237041354e-07, + "loss": 1.0519, + "step": 7571 + }, + { + "epoch": 0.6828696397168237, + "grad_norm": 1.458898089818, + "learning_rate": 9.652955281465278e-07, + "loss": 0.9662, + "step": 7572 + }, + { + "epoch": 0.682959823240294, + "grad_norm": 1.7702773923162012, + "learning_rate": 9.64795620875612e-07, + "loss": 0.9887, + "step": 7573 + }, + { + "epoch": 0.6830500067637643, + "grad_norm": 1.9153171708916268, + "learning_rate": 9.64295801934041e-07, + "loss": 0.9305, + "step": 7574 + }, + { + "epoch": 0.6831401902872345, + "grad_norm": 0.6687927294832752, + "learning_rate": 9.63796071364463e-07, + "loss": 0.8243, + "step": 7575 + }, + { + "epoch": 0.6832303738107048, + "grad_norm": 1.890392572917708, + "learning_rate": 9.632964292095179e-07, + "loss": 1.0692, + "step": 7576 + }, + { + "epoch": 0.6833205573341751, + "grad_norm": 1.43253147776373, + "learning_rate": 9.627968755118374e-07, + "loss": 0.9732, + "step": 7577 + }, + { + "epoch": 0.6834107408576453, + "grad_norm": 0.6642367190539759, + "learning_rate": 9.622974103140468e-07, + "loss": 0.8036, + "step": 7578 + }, + { + "epoch": 0.6835009243811155, + "grad_norm": 1.612233484406895, + "learning_rate": 9.617980336587632e-07, + "loss": 1.0449, + "step": 7579 + }, + { + "epoch": 0.6835911079045859, + "grad_norm": 1.676328911911575, + "learning_rate": 9.612987455885964e-07, + "loss": 0.8665, + "step": 7580 + }, + { + "epoch": 0.6836812914280561, + "grad_norm": 1.6696570827848538, + "learning_rate": 9.607995461461467e-07, + "loss": 0.9887, + "step": 7581 + }, + { + "epoch": 0.6837714749515263, + "grad_norm": 1.7922748228448777, + "learning_rate": 9.603004353740111e-07, + "loss": 0.8878, + "step": 7582 + }, + { + "epoch": 0.6838616584749966, + "grad_norm": 0.6964356855231711, + "learning_rate": 9.598014133147738e-07, + "loss": 0.859, + "step": 7583 + }, + { + "epoch": 0.6839518419984669, + "grad_norm": 1.415418785856493, + "learning_rate": 9.59302480011017e-07, + "loss": 0.9463, + "step": 7584 + }, + { + "epoch": 0.6840420255219372, + "grad_norm": 0.6704528306019252, + "learning_rate": 9.588036355053102e-07, + "loss": 0.8355, + "step": 7585 + }, + { + "epoch": 0.6841322090454074, + "grad_norm": 2.3399195587155908, + "learning_rate": 9.583048798402182e-07, + "loss": 0.8508, + "step": 7586 + }, + { + "epoch": 0.6842223925688776, + "grad_norm": 1.7017152183523387, + "learning_rate": 9.57806213058298e-07, + "loss": 0.9803, + "step": 7587 + }, + { + "epoch": 0.684312576092348, + "grad_norm": 0.6946343747656708, + "learning_rate": 9.57307635202098e-07, + "loss": 0.8531, + "step": 7588 + }, + { + "epoch": 0.6844027596158182, + "grad_norm": 1.4940417687119338, + "learning_rate": 9.568091463141607e-07, + "loss": 1.0868, + "step": 7589 + }, + { + "epoch": 0.6844929431392884, + "grad_norm": 1.417796142833541, + "learning_rate": 9.563107464370187e-07, + "loss": 0.9507, + "step": 7590 + }, + { + "epoch": 0.6845831266627587, + "grad_norm": 1.7494991812015872, + "learning_rate": 9.558124356131982e-07, + "loss": 1.0022, + "step": 7591 + }, + { + "epoch": 0.684673310186229, + "grad_norm": 1.9604668527382514, + "learning_rate": 9.553142138852187e-07, + "loss": 0.9178, + "step": 7592 + }, + { + "epoch": 0.6847634937096992, + "grad_norm": 1.9831187559571115, + "learning_rate": 9.548160812955905e-07, + "loss": 0.8961, + "step": 7593 + }, + { + "epoch": 0.6848536772331695, + "grad_norm": 1.498851813516138, + "learning_rate": 9.543180378868175e-07, + "loss": 0.806, + "step": 7594 + }, + { + "epoch": 0.6849438607566397, + "grad_norm": 2.03969859290142, + "learning_rate": 9.538200837013962e-07, + "loss": 1.0196, + "step": 7595 + }, + { + "epoch": 0.68503404428011, + "grad_norm": 0.6552718911529352, + "learning_rate": 9.533222187818122e-07, + "loss": 0.8509, + "step": 7596 + }, + { + "epoch": 0.6851242278035803, + "grad_norm": 1.8137582817913873, + "learning_rate": 9.528244431705492e-07, + "loss": 0.9978, + "step": 7597 + }, + { + "epoch": 0.6852144113270505, + "grad_norm": 1.4424523999907288, + "learning_rate": 9.523267569100774e-07, + "loss": 0.9923, + "step": 7598 + }, + { + "epoch": 0.6853045948505209, + "grad_norm": 1.7750314279917063, + "learning_rate": 9.518291600428652e-07, + "loss": 0.9408, + "step": 7599 + }, + { + "epoch": 0.6853947783739911, + "grad_norm": 8.880790862873512, + "learning_rate": 9.513316526113677e-07, + "loss": 1.0073, + "step": 7600 + }, + { + "epoch": 0.6854849618974613, + "grad_norm": 1.555143969822801, + "learning_rate": 9.50834234658036e-07, + "loss": 1.0079, + "step": 7601 + }, + { + "epoch": 0.6855751454209316, + "grad_norm": 1.4751688521123376, + "learning_rate": 9.503369062253123e-07, + "loss": 0.9987, + "step": 7602 + }, + { + "epoch": 0.6856653289444019, + "grad_norm": 0.6401005901322626, + "learning_rate": 9.498396673556317e-07, + "loss": 0.8267, + "step": 7603 + }, + { + "epoch": 0.6857555124678721, + "grad_norm": 1.5584859021956308, + "learning_rate": 9.493425180914219e-07, + "loss": 1.0182, + "step": 7604 + }, + { + "epoch": 0.6858456959913424, + "grad_norm": 1.5822001802650978, + "learning_rate": 9.488454584751e-07, + "loss": 0.9158, + "step": 7605 + }, + { + "epoch": 0.6859358795148126, + "grad_norm": 1.899704122092533, + "learning_rate": 9.483484885490813e-07, + "loss": 0.9896, + "step": 7606 + }, + { + "epoch": 0.686026063038283, + "grad_norm": 1.8369467051483943, + "learning_rate": 9.478516083557675e-07, + "loss": 1.0005, + "step": 7607 + }, + { + "epoch": 0.6861162465617532, + "grad_norm": 2.1143326884836102, + "learning_rate": 9.473548179375561e-07, + "loss": 1.0082, + "step": 7608 + }, + { + "epoch": 0.6862064300852234, + "grad_norm": 0.6887773222099309, + "learning_rate": 9.468581173368358e-07, + "loss": 0.8546, + "step": 7609 + }, + { + "epoch": 0.6862966136086937, + "grad_norm": 2.055825966011436, + "learning_rate": 9.463615065959878e-07, + "loss": 0.9931, + "step": 7610 + }, + { + "epoch": 0.686386797132164, + "grad_norm": 1.630524487861046, + "learning_rate": 9.458649857573857e-07, + "loss": 1.0417, + "step": 7611 + }, + { + "epoch": 0.6864769806556342, + "grad_norm": 1.659157773614665, + "learning_rate": 9.453685548633963e-07, + "loss": 0.9925, + "step": 7612 + }, + { + "epoch": 0.6865671641791045, + "grad_norm": 1.6273760663732335, + "learning_rate": 9.448722139563756e-07, + "loss": 0.9346, + "step": 7613 + }, + { + "epoch": 0.6866573477025747, + "grad_norm": 1.643782130170323, + "learning_rate": 9.443759630786769e-07, + "loss": 0.8818, + "step": 7614 + }, + { + "epoch": 0.686747531226045, + "grad_norm": 1.6029243447767973, + "learning_rate": 9.438798022726408e-07, + "loss": 1.0565, + "step": 7615 + }, + { + "epoch": 0.6868377147495153, + "grad_norm": 1.7121602048471918, + "learning_rate": 9.433837315806037e-07, + "loss": 0.9386, + "step": 7616 + }, + { + "epoch": 0.6869278982729855, + "grad_norm": 1.8831886137439056, + "learning_rate": 9.428877510448925e-07, + "loss": 0.941, + "step": 7617 + }, + { + "epoch": 0.6870180817964557, + "grad_norm": 1.9569423145180018, + "learning_rate": 9.423918607078272e-07, + "loss": 0.9134, + "step": 7618 + }, + { + "epoch": 0.6871082653199261, + "grad_norm": 1.6400895471083827, + "learning_rate": 9.418960606117208e-07, + "loss": 0.9943, + "step": 7619 + }, + { + "epoch": 0.6871984488433963, + "grad_norm": 1.5691415396913955, + "learning_rate": 9.414003507988752e-07, + "loss": 0.9271, + "step": 7620 + }, + { + "epoch": 0.6872886323668665, + "grad_norm": 2.2694501484809995, + "learning_rate": 9.409047313115904e-07, + "loss": 0.9184, + "step": 7621 + }, + { + "epoch": 0.6873788158903368, + "grad_norm": 2.1082665816082122, + "learning_rate": 9.404092021921521e-07, + "loss": 0.9265, + "step": 7622 + }, + { + "epoch": 0.6874689994138071, + "grad_norm": 2.039100383538169, + "learning_rate": 9.399137634828447e-07, + "loss": 0.9667, + "step": 7623 + }, + { + "epoch": 0.6875591829372774, + "grad_norm": 1.611665595579268, + "learning_rate": 9.394184152259396e-07, + "loss": 1.006, + "step": 7624 + }, + { + "epoch": 0.6876493664607476, + "grad_norm": 1.46242655382193, + "learning_rate": 9.389231574637033e-07, + "loss": 1.0889, + "step": 7625 + }, + { + "epoch": 0.6877395499842179, + "grad_norm": 2.6328859403877627, + "learning_rate": 9.384279902383938e-07, + "loss": 1.0729, + "step": 7626 + }, + { + "epoch": 0.6878297335076882, + "grad_norm": 1.565001278593376, + "learning_rate": 9.379329135922615e-07, + "loss": 0.9995, + "step": 7627 + }, + { + "epoch": 0.6879199170311584, + "grad_norm": 1.574398513260969, + "learning_rate": 9.374379275675495e-07, + "loss": 0.9419, + "step": 7628 + }, + { + "epoch": 0.6880101005546286, + "grad_norm": 1.7358395003735596, + "learning_rate": 9.369430322064931e-07, + "loss": 0.9843, + "step": 7629 + }, + { + "epoch": 0.688100284078099, + "grad_norm": 1.6923961958050602, + "learning_rate": 9.364482275513179e-07, + "loss": 1.0635, + "step": 7630 + }, + { + "epoch": 0.6881904676015692, + "grad_norm": 1.7591565906206579, + "learning_rate": 9.359535136442444e-07, + "loss": 1.0505, + "step": 7631 + }, + { + "epoch": 0.6882806511250394, + "grad_norm": 1.5400683993470414, + "learning_rate": 9.354588905274843e-07, + "loss": 0.9176, + "step": 7632 + }, + { + "epoch": 0.6883708346485097, + "grad_norm": 1.5514003949415525, + "learning_rate": 9.349643582432414e-07, + "loss": 1.0214, + "step": 7633 + }, + { + "epoch": 0.68846101817198, + "grad_norm": 1.7150918350988893, + "learning_rate": 9.344699168337127e-07, + "loss": 0.9773, + "step": 7634 + }, + { + "epoch": 0.6885512016954503, + "grad_norm": 1.7733855386927584, + "learning_rate": 9.339755663410845e-07, + "loss": 0.9473, + "step": 7635 + }, + { + "epoch": 0.6886413852189205, + "grad_norm": 1.801672195840139, + "learning_rate": 9.334813068075405e-07, + "loss": 1.0192, + "step": 7636 + }, + { + "epoch": 0.6887315687423907, + "grad_norm": 2.61430906602494, + "learning_rate": 9.329871382752506e-07, + "loss": 0.988, + "step": 7637 + }, + { + "epoch": 0.6888217522658611, + "grad_norm": 1.5879372452652254, + "learning_rate": 9.32493060786383e-07, + "loss": 0.9751, + "step": 7638 + }, + { + "epoch": 0.6889119357893313, + "grad_norm": 1.3801405356207967, + "learning_rate": 9.31999074383093e-07, + "loss": 1.0451, + "step": 7639 + }, + { + "epoch": 0.6890021193128015, + "grad_norm": 1.9242478895439716, + "learning_rate": 9.315051791075308e-07, + "loss": 0.9778, + "step": 7640 + }, + { + "epoch": 0.6890923028362718, + "grad_norm": 4.460303801038512, + "learning_rate": 9.310113750018382e-07, + "loss": 0.9786, + "step": 7641 + }, + { + "epoch": 0.6891824863597421, + "grad_norm": 2.289755342065998, + "learning_rate": 9.305176621081496e-07, + "loss": 1.0552, + "step": 7642 + }, + { + "epoch": 0.6892726698832123, + "grad_norm": 1.8946266386263233, + "learning_rate": 9.300240404685911e-07, + "loss": 1.0185, + "step": 7643 + }, + { + "epoch": 0.6893628534066826, + "grad_norm": 2.6851613714786953, + "learning_rate": 9.295305101252812e-07, + "loss": 0.9873, + "step": 7644 + }, + { + "epoch": 0.6894530369301528, + "grad_norm": 1.7747186410908633, + "learning_rate": 9.290370711203314e-07, + "loss": 0.9673, + "step": 7645 + }, + { + "epoch": 0.6895432204536232, + "grad_norm": 1.5025534354052668, + "learning_rate": 9.285437234958433e-07, + "loss": 0.98, + "step": 7646 + }, + { + "epoch": 0.6896334039770934, + "grad_norm": 2.348836947826904, + "learning_rate": 9.280504672939124e-07, + "loss": 0.9787, + "step": 7647 + }, + { + "epoch": 0.6897235875005636, + "grad_norm": 2.008651498730746, + "learning_rate": 9.275573025566266e-07, + "loss": 1.0238, + "step": 7648 + }, + { + "epoch": 0.689813771024034, + "grad_norm": 1.3559441516059851, + "learning_rate": 9.27064229326065e-07, + "loss": 1.0188, + "step": 7649 + }, + { + "epoch": 0.6899039545475042, + "grad_norm": 2.0688051890592116, + "learning_rate": 9.265712476442995e-07, + "loss": 0.9505, + "step": 7650 + }, + { + "epoch": 0.6899941380709744, + "grad_norm": 1.4486366604171037, + "learning_rate": 9.260783575533949e-07, + "loss": 0.9146, + "step": 7651 + }, + { + "epoch": 0.6900843215944447, + "grad_norm": 1.7378169093207272, + "learning_rate": 9.255855590954045e-07, + "loss": 0.9339, + "step": 7652 + }, + { + "epoch": 0.690174505117915, + "grad_norm": 1.5480514830473304, + "learning_rate": 9.250928523123802e-07, + "loss": 1.0489, + "step": 7653 + }, + { + "epoch": 0.6902646886413852, + "grad_norm": 1.5438528883774962, + "learning_rate": 9.24600237246359e-07, + "loss": 0.996, + "step": 7654 + }, + { + "epoch": 0.6903548721648555, + "grad_norm": 1.9312656142705409, + "learning_rate": 9.241077139393769e-07, + "loss": 0.9945, + "step": 7655 + }, + { + "epoch": 0.6904450556883257, + "grad_norm": 2.6101064970132413, + "learning_rate": 9.236152824334564e-07, + "loss": 1.016, + "step": 7656 + }, + { + "epoch": 0.690535239211796, + "grad_norm": 1.9466162912282607, + "learning_rate": 9.231229427706151e-07, + "loss": 0.8831, + "step": 7657 + }, + { + "epoch": 0.6906254227352663, + "grad_norm": 2.2744896410599056, + "learning_rate": 9.226306949928622e-07, + "loss": 0.9938, + "step": 7658 + }, + { + "epoch": 0.6907156062587365, + "grad_norm": 1.5421256937088959, + "learning_rate": 9.221385391421988e-07, + "loss": 1.0602, + "step": 7659 + }, + { + "epoch": 0.6908057897822067, + "grad_norm": 1.9057006669778314, + "learning_rate": 9.216464752606192e-07, + "loss": 1.0055, + "step": 7660 + }, + { + "epoch": 0.6908959733056771, + "grad_norm": 1.5279731721869647, + "learning_rate": 9.211545033901078e-07, + "loss": 1.0634, + "step": 7661 + }, + { + "epoch": 0.6909861568291473, + "grad_norm": 1.457310463177643, + "learning_rate": 9.206626235726426e-07, + "loss": 1.014, + "step": 7662 + }, + { + "epoch": 0.6910763403526176, + "grad_norm": 1.8393490525131986, + "learning_rate": 9.20170835850194e-07, + "loss": 0.9229, + "step": 7663 + }, + { + "epoch": 0.6911665238760878, + "grad_norm": 0.6439485632571686, + "learning_rate": 9.196791402647237e-07, + "loss": 0.8095, + "step": 7664 + }, + { + "epoch": 0.6912567073995581, + "grad_norm": 1.900684863185493, + "learning_rate": 9.191875368581861e-07, + "loss": 0.9083, + "step": 7665 + }, + { + "epoch": 0.6913468909230284, + "grad_norm": 1.835919968541628, + "learning_rate": 9.186960256725271e-07, + "loss": 0.9953, + "step": 7666 + }, + { + "epoch": 0.6914370744464986, + "grad_norm": 1.639571441969188, + "learning_rate": 9.182046067496856e-07, + "loss": 1.0489, + "step": 7667 + }, + { + "epoch": 0.6915272579699688, + "grad_norm": 1.635454877986081, + "learning_rate": 9.177132801315927e-07, + "loss": 1.0736, + "step": 7668 + }, + { + "epoch": 0.6916174414934392, + "grad_norm": 1.500978468606467, + "learning_rate": 9.172220458601692e-07, + "loss": 0.9603, + "step": 7669 + }, + { + "epoch": 0.6917076250169094, + "grad_norm": 1.7174217640522258, + "learning_rate": 9.167309039773324e-07, + "loss": 0.9547, + "step": 7670 + }, + { + "epoch": 0.6917978085403796, + "grad_norm": 1.289820777508271, + "learning_rate": 9.162398545249872e-07, + "loss": 0.8856, + "step": 7671 + }, + { + "epoch": 0.69188799206385, + "grad_norm": 1.8914869881327852, + "learning_rate": 9.157488975450334e-07, + "loss": 1.0435, + "step": 7672 + }, + { + "epoch": 0.6919781755873202, + "grad_norm": 1.6036342711701401, + "learning_rate": 9.15258033079362e-07, + "loss": 0.8179, + "step": 7673 + }, + { + "epoch": 0.6920683591107905, + "grad_norm": 1.6404006804653897, + "learning_rate": 9.147672611698567e-07, + "loss": 0.9712, + "step": 7674 + }, + { + "epoch": 0.6921585426342607, + "grad_norm": 2.6047193830766484, + "learning_rate": 9.142765818583933e-07, + "loss": 0.9759, + "step": 7675 + }, + { + "epoch": 0.692248726157731, + "grad_norm": 1.5626156573405752, + "learning_rate": 9.13785995186837e-07, + "loss": 0.9632, + "step": 7676 + }, + { + "epoch": 0.6923389096812013, + "grad_norm": 2.2150201744039166, + "learning_rate": 9.132955011970502e-07, + "loss": 0.9529, + "step": 7677 + }, + { + "epoch": 0.6924290932046715, + "grad_norm": 1.3405303975051954, + "learning_rate": 9.128050999308827e-07, + "loss": 0.9738, + "step": 7678 + }, + { + "epoch": 0.6925192767281417, + "grad_norm": 1.8114055091331887, + "learning_rate": 9.123147914301789e-07, + "loss": 1.0729, + "step": 7679 + }, + { + "epoch": 0.6926094602516121, + "grad_norm": 1.9125153808972568, + "learning_rate": 9.118245757367745e-07, + "loss": 1.0308, + "step": 7680 + }, + { + "epoch": 0.6926996437750823, + "grad_norm": 0.6325256733716131, + "learning_rate": 9.113344528924973e-07, + "loss": 0.7969, + "step": 7681 + }, + { + "epoch": 0.6927898272985525, + "grad_norm": 3.4750129589689935, + "learning_rate": 9.108444229391676e-07, + "loss": 0.9634, + "step": 7682 + }, + { + "epoch": 0.6928800108220228, + "grad_norm": 1.7378658184013933, + "learning_rate": 9.103544859185979e-07, + "loss": 0.9599, + "step": 7683 + }, + { + "epoch": 0.6929701943454931, + "grad_norm": 1.614815488142605, + "learning_rate": 9.098646418725902e-07, + "loss": 1.0712, + "step": 7684 + }, + { + "epoch": 0.6930603778689634, + "grad_norm": 0.6407958244893124, + "learning_rate": 9.093748908429437e-07, + "loss": 0.8395, + "step": 7685 + }, + { + "epoch": 0.6931505613924336, + "grad_norm": 1.9657423266402119, + "learning_rate": 9.088852328714444e-07, + "loss": 0.9762, + "step": 7686 + }, + { + "epoch": 0.6932407449159038, + "grad_norm": 1.460711211768569, + "learning_rate": 9.083956679998735e-07, + "loss": 0.8948, + "step": 7687 + }, + { + "epoch": 0.6933309284393742, + "grad_norm": 1.8039637844315977, + "learning_rate": 9.079061962700032e-07, + "loss": 0.9555, + "step": 7688 + }, + { + "epoch": 0.6934211119628444, + "grad_norm": 1.3902163387285134, + "learning_rate": 9.074168177235979e-07, + "loss": 1.0078, + "step": 7689 + }, + { + "epoch": 0.6935112954863146, + "grad_norm": 1.637179143789293, + "learning_rate": 9.069275324024151e-07, + "loss": 1.0003, + "step": 7690 + }, + { + "epoch": 0.6936014790097849, + "grad_norm": 3.9833517518409094, + "learning_rate": 9.064383403482005e-07, + "loss": 0.9155, + "step": 7691 + }, + { + "epoch": 0.6936916625332552, + "grad_norm": 1.6825746075893246, + "learning_rate": 9.059492416026983e-07, + "loss": 0.9466, + "step": 7692 + }, + { + "epoch": 0.6937818460567254, + "grad_norm": 1.3980182466776025, + "learning_rate": 9.054602362076378e-07, + "loss": 0.9808, + "step": 7693 + }, + { + "epoch": 0.6938720295801957, + "grad_norm": 2.434842348106307, + "learning_rate": 9.049713242047468e-07, + "loss": 0.9738, + "step": 7694 + }, + { + "epoch": 0.693962213103666, + "grad_norm": 1.6664611292305038, + "learning_rate": 9.044825056357395e-07, + "loss": 0.952, + "step": 7695 + }, + { + "epoch": 0.6940523966271362, + "grad_norm": 1.7375759794239907, + "learning_rate": 9.039937805423255e-07, + "loss": 0.9355, + "step": 7696 + }, + { + "epoch": 0.6941425801506065, + "grad_norm": 0.5982682827978635, + "learning_rate": 9.035051489662051e-07, + "loss": 0.7546, + "step": 7697 + }, + { + "epoch": 0.6942327636740767, + "grad_norm": 1.6297612496327318, + "learning_rate": 9.030166109490718e-07, + "loss": 0.9334, + "step": 7698 + }, + { + "epoch": 0.6943229471975471, + "grad_norm": 5.355988525275772, + "learning_rate": 9.025281665326099e-07, + "loss": 0.9914, + "step": 7699 + }, + { + "epoch": 0.6944131307210173, + "grad_norm": 1.7727710967437944, + "learning_rate": 9.020398157584967e-07, + "loss": 0.9921, + "step": 7700 + }, + { + "epoch": 0.6945033142444875, + "grad_norm": 2.0988318418593788, + "learning_rate": 9.015515586684002e-07, + "loss": 0.887, + "step": 7701 + }, + { + "epoch": 0.6945934977679578, + "grad_norm": 0.6568925981133333, + "learning_rate": 9.010633953039812e-07, + "loss": 0.8296, + "step": 7702 + }, + { + "epoch": 0.6946836812914281, + "grad_norm": 1.5278530198652223, + "learning_rate": 9.005753257068929e-07, + "loss": 0.9975, + "step": 7703 + }, + { + "epoch": 0.6947738648148983, + "grad_norm": 2.052142405426202, + "learning_rate": 9.000873499187797e-07, + "loss": 0.9967, + "step": 7704 + }, + { + "epoch": 0.6948640483383686, + "grad_norm": 1.7653192576128625, + "learning_rate": 8.995994679812797e-07, + "loss": 1.0161, + "step": 7705 + }, + { + "epoch": 0.6949542318618388, + "grad_norm": 1.5286334512221007, + "learning_rate": 8.991116799360192e-07, + "loss": 1.006, + "step": 7706 + }, + { + "epoch": 0.6950444153853091, + "grad_norm": 1.5267439735334885, + "learning_rate": 8.986239858246217e-07, + "loss": 0.9758, + "step": 7707 + }, + { + "epoch": 0.6951345989087794, + "grad_norm": 2.125416995254588, + "learning_rate": 8.981363856886972e-07, + "loss": 1.0669, + "step": 7708 + }, + { + "epoch": 0.6952247824322496, + "grad_norm": 1.7689338669391368, + "learning_rate": 8.976488795698533e-07, + "loss": 0.9656, + "step": 7709 + }, + { + "epoch": 0.6953149659557198, + "grad_norm": 1.6456229964786304, + "learning_rate": 8.971614675096841e-07, + "loss": 0.999, + "step": 7710 + }, + { + "epoch": 0.6954051494791902, + "grad_norm": 3.1458740400160607, + "learning_rate": 8.966741495497807e-07, + "loss": 1.0483, + "step": 7711 + }, + { + "epoch": 0.6954953330026604, + "grad_norm": 1.4804970932313262, + "learning_rate": 8.961869257317218e-07, + "loss": 0.9508, + "step": 7712 + }, + { + "epoch": 0.6955855165261307, + "grad_norm": 1.6068109623757971, + "learning_rate": 8.956997960970809e-07, + "loss": 0.9626, + "step": 7713 + }, + { + "epoch": 0.6956757000496009, + "grad_norm": 2.1187997007589074, + "learning_rate": 8.952127606874224e-07, + "loss": 0.8926, + "step": 7714 + }, + { + "epoch": 0.6957658835730712, + "grad_norm": 1.5277529117217523, + "learning_rate": 8.947258195443028e-07, + "loss": 0.9881, + "step": 7715 + }, + { + "epoch": 0.6958560670965415, + "grad_norm": 1.7323764592940845, + "learning_rate": 8.942389727092716e-07, + "loss": 0.9629, + "step": 7716 + }, + { + "epoch": 0.6959462506200117, + "grad_norm": 0.6792227592200752, + "learning_rate": 8.937522202238677e-07, + "loss": 0.8009, + "step": 7717 + }, + { + "epoch": 0.696036434143482, + "grad_norm": 2.391439162099201, + "learning_rate": 8.932655621296239e-07, + "loss": 1.0014, + "step": 7718 + }, + { + "epoch": 0.6961266176669523, + "grad_norm": 1.6072210323151912, + "learning_rate": 8.927789984680649e-07, + "loss": 0.9803, + "step": 7719 + }, + { + "epoch": 0.6962168011904225, + "grad_norm": 1.6510036075028136, + "learning_rate": 8.922925292807068e-07, + "loss": 0.8539, + "step": 7720 + }, + { + "epoch": 0.6963069847138927, + "grad_norm": 1.6950255823596043, + "learning_rate": 8.91806154609058e-07, + "loss": 1.0671, + "step": 7721 + }, + { + "epoch": 0.6963971682373631, + "grad_norm": 1.9404381654397709, + "learning_rate": 8.913198744946195e-07, + "loss": 0.9509, + "step": 7722 + }, + { + "epoch": 0.6964873517608333, + "grad_norm": 1.4328188763958103, + "learning_rate": 8.908336889788807e-07, + "loss": 1.0071, + "step": 7723 + }, + { + "epoch": 0.6965775352843036, + "grad_norm": 3.1862956745140973, + "learning_rate": 8.903475981033293e-07, + "loss": 0.866, + "step": 7724 + }, + { + "epoch": 0.6966677188077738, + "grad_norm": 0.7387647710572434, + "learning_rate": 8.898616019094376e-07, + "loss": 0.8694, + "step": 7725 + }, + { + "epoch": 0.6967579023312441, + "grad_norm": 1.5715681531147936, + "learning_rate": 8.89375700438677e-07, + "loss": 1.0221, + "step": 7726 + }, + { + "epoch": 0.6968480858547144, + "grad_norm": 1.585153019254578, + "learning_rate": 8.888898937325047e-07, + "loss": 0.9469, + "step": 7727 + }, + { + "epoch": 0.6969382693781846, + "grad_norm": 2.080587663163676, + "learning_rate": 8.884041818323733e-07, + "loss": 0.9492, + "step": 7728 + }, + { + "epoch": 0.6970284529016548, + "grad_norm": 1.6301649613953442, + "learning_rate": 8.879185647797262e-07, + "loss": 1.0153, + "step": 7729 + }, + { + "epoch": 0.6971186364251252, + "grad_norm": 1.5985231169650946, + "learning_rate": 8.874330426159993e-07, + "loss": 0.9635, + "step": 7730 + }, + { + "epoch": 0.6972088199485954, + "grad_norm": 1.6188008834323684, + "learning_rate": 8.869476153826205e-07, + "loss": 1.0105, + "step": 7731 + }, + { + "epoch": 0.6972990034720656, + "grad_norm": 1.5124807088583243, + "learning_rate": 8.864622831210071e-07, + "loss": 0.9869, + "step": 7732 + }, + { + "epoch": 0.6973891869955359, + "grad_norm": 1.8094489627670194, + "learning_rate": 8.85977045872573e-07, + "loss": 0.9923, + "step": 7733 + }, + { + "epoch": 0.6974793705190062, + "grad_norm": 2.0198864504559646, + "learning_rate": 8.854919036787194e-07, + "loss": 0.9629, + "step": 7734 + }, + { + "epoch": 0.6975695540424764, + "grad_norm": 1.8699314911680553, + "learning_rate": 8.850068565808417e-07, + "loss": 0.9516, + "step": 7735 + }, + { + "epoch": 0.6976597375659467, + "grad_norm": 2.638179752926711, + "learning_rate": 8.845219046203271e-07, + "loss": 1.0348, + "step": 7736 + }, + { + "epoch": 0.6977499210894169, + "grad_norm": 1.834715481146609, + "learning_rate": 8.840370478385544e-07, + "loss": 1.0094, + "step": 7737 + }, + { + "epoch": 0.6978401046128873, + "grad_norm": 12.755214447647235, + "learning_rate": 8.83552286276894e-07, + "loss": 1.0037, + "step": 7738 + }, + { + "epoch": 0.6979302881363575, + "grad_norm": 2.393589251427053, + "learning_rate": 8.830676199767095e-07, + "loss": 1.0008, + "step": 7739 + }, + { + "epoch": 0.6980204716598277, + "grad_norm": 1.5677876837874276, + "learning_rate": 8.825830489793527e-07, + "loss": 0.967, + "step": 7740 + }, + { + "epoch": 0.698110655183298, + "grad_norm": 1.3459893574145125, + "learning_rate": 8.820985733261732e-07, + "loss": 0.9377, + "step": 7741 + }, + { + "epoch": 0.6982008387067683, + "grad_norm": 1.8066600241779434, + "learning_rate": 8.816141930585066e-07, + "loss": 0.8946, + "step": 7742 + }, + { + "epoch": 0.6982910222302385, + "grad_norm": 0.6873638711886196, + "learning_rate": 8.811299082176837e-07, + "loss": 0.8266, + "step": 7743 + }, + { + "epoch": 0.6983812057537088, + "grad_norm": 2.056507073655265, + "learning_rate": 8.806457188450265e-07, + "loss": 0.9495, + "step": 7744 + }, + { + "epoch": 0.6984713892771791, + "grad_norm": 4.99802741240349, + "learning_rate": 8.801616249818487e-07, + "loss": 1.0194, + "step": 7745 + }, + { + "epoch": 0.6985615728006493, + "grad_norm": 1.6805358873307548, + "learning_rate": 8.796776266694564e-07, + "loss": 0.9642, + "step": 7746 + }, + { + "epoch": 0.6986517563241196, + "grad_norm": 1.341296906588713, + "learning_rate": 8.79193723949145e-07, + "loss": 0.9319, + "step": 7747 + }, + { + "epoch": 0.6987419398475898, + "grad_norm": 1.8535283593724168, + "learning_rate": 8.787099168622063e-07, + "loss": 0.8785, + "step": 7748 + }, + { + "epoch": 0.6988321233710602, + "grad_norm": 1.7556052358186374, + "learning_rate": 8.782262054499199e-07, + "loss": 0.9199, + "step": 7749 + }, + { + "epoch": 0.6989223068945304, + "grad_norm": 1.483300713741234, + "learning_rate": 8.777425897535588e-07, + "loss": 1.012, + "step": 7750 + }, + { + "epoch": 0.6990124904180006, + "grad_norm": 1.6504648045228434, + "learning_rate": 8.77259069814388e-07, + "loss": 0.9807, + "step": 7751 + }, + { + "epoch": 0.6991026739414709, + "grad_norm": 1.815306759387038, + "learning_rate": 8.767756456736641e-07, + "loss": 0.9846, + "step": 7752 + }, + { + "epoch": 0.6991928574649412, + "grad_norm": 1.7912489012177766, + "learning_rate": 8.762923173726358e-07, + "loss": 0.9074, + "step": 7753 + }, + { + "epoch": 0.6992830409884114, + "grad_norm": 1.7393895420092989, + "learning_rate": 8.758090849525428e-07, + "loss": 1.0807, + "step": 7754 + }, + { + "epoch": 0.6993732245118817, + "grad_norm": 2.256859284448959, + "learning_rate": 8.753259484546174e-07, + "loss": 0.9226, + "step": 7755 + }, + { + "epoch": 0.6994634080353519, + "grad_norm": 1.6958888607947529, + "learning_rate": 8.748429079200841e-07, + "loss": 0.9372, + "step": 7756 + }, + { + "epoch": 0.6995535915588222, + "grad_norm": 1.9215068929818107, + "learning_rate": 8.743599633901575e-07, + "loss": 1.037, + "step": 7757 + }, + { + "epoch": 0.6996437750822925, + "grad_norm": 1.8703611209830928, + "learning_rate": 8.738771149060453e-07, + "loss": 0.9986, + "step": 7758 + }, + { + "epoch": 0.6997339586057627, + "grad_norm": 2.7381243477136827, + "learning_rate": 8.73394362508947e-07, + "loss": 0.951, + "step": 7759 + }, + { + "epoch": 0.6998241421292329, + "grad_norm": 1.41119135434904, + "learning_rate": 8.72911706240054e-07, + "loss": 0.945, + "step": 7760 + }, + { + "epoch": 0.6999143256527033, + "grad_norm": 1.7312630580588673, + "learning_rate": 8.724291461405493e-07, + "loss": 0.8607, + "step": 7761 + }, + { + "epoch": 0.7000045091761735, + "grad_norm": 2.0070753117198374, + "learning_rate": 8.71946682251606e-07, + "loss": 0.9467, + "step": 7762 + }, + { + "epoch": 0.7000946926996438, + "grad_norm": 1.5843018614611832, + "learning_rate": 8.714643146143932e-07, + "loss": 0.9823, + "step": 7763 + }, + { + "epoch": 0.700184876223114, + "grad_norm": 1.4239828394595193, + "learning_rate": 8.709820432700663e-07, + "loss": 0.9412, + "step": 7764 + }, + { + "epoch": 0.7002750597465843, + "grad_norm": 1.605488055407803, + "learning_rate": 8.704998682597784e-07, + "loss": 0.8959, + "step": 7765 + }, + { + "epoch": 0.7003652432700546, + "grad_norm": 1.6840248352584388, + "learning_rate": 8.700177896246688e-07, + "loss": 1.0022, + "step": 7766 + }, + { + "epoch": 0.7004554267935248, + "grad_norm": 2.3411565037117503, + "learning_rate": 8.695358074058721e-07, + "loss": 0.9119, + "step": 7767 + }, + { + "epoch": 0.7005456103169951, + "grad_norm": 1.8328190067627799, + "learning_rate": 8.690539216445136e-07, + "loss": 1.0181, + "step": 7768 + }, + { + "epoch": 0.7006357938404654, + "grad_norm": 1.7946868842281805, + "learning_rate": 8.685721323817106e-07, + "loss": 0.8855, + "step": 7769 + }, + { + "epoch": 0.7007259773639356, + "grad_norm": 2.946224162620056, + "learning_rate": 8.680904396585718e-07, + "loss": 1.0208, + "step": 7770 + }, + { + "epoch": 0.7008161608874058, + "grad_norm": 1.6782691822589477, + "learning_rate": 8.676088435161988e-07, + "loss": 0.9233, + "step": 7771 + }, + { + "epoch": 0.7009063444108762, + "grad_norm": 2.1196877893838257, + "learning_rate": 8.671273439956824e-07, + "loss": 1.0044, + "step": 7772 + }, + { + "epoch": 0.7009965279343464, + "grad_norm": 1.8391365911424051, + "learning_rate": 8.666459411381075e-07, + "loss": 0.9885, + "step": 7773 + }, + { + "epoch": 0.7010867114578166, + "grad_norm": 1.9277342513218678, + "learning_rate": 8.661646349845501e-07, + "loss": 1.0082, + "step": 7774 + }, + { + "epoch": 0.7011768949812869, + "grad_norm": 1.6043880937351829, + "learning_rate": 8.656834255760783e-07, + "loss": 1.0369, + "step": 7775 + }, + { + "epoch": 0.7012670785047572, + "grad_norm": 1.8839661795259315, + "learning_rate": 8.652023129537509e-07, + "loss": 1.0326, + "step": 7776 + }, + { + "epoch": 0.7013572620282275, + "grad_norm": 2.0437605364699967, + "learning_rate": 8.647212971586195e-07, + "loss": 1.008, + "step": 7777 + }, + { + "epoch": 0.7014474455516977, + "grad_norm": 1.4671555650706982, + "learning_rate": 8.642403782317275e-07, + "loss": 0.9332, + "step": 7778 + }, + { + "epoch": 0.7015376290751679, + "grad_norm": 1.6888516981787247, + "learning_rate": 8.637595562141075e-07, + "loss": 0.986, + "step": 7779 + }, + { + "epoch": 0.7016278125986383, + "grad_norm": 1.4705780502976413, + "learning_rate": 8.632788311467889e-07, + "loss": 0.9472, + "step": 7780 + }, + { + "epoch": 0.7017179961221085, + "grad_norm": 1.8464563840935928, + "learning_rate": 8.627982030707867e-07, + "loss": 0.9775, + "step": 7781 + }, + { + "epoch": 0.7018081796455787, + "grad_norm": 2.664803052427616, + "learning_rate": 8.623176720271139e-07, + "loss": 1.0452, + "step": 7782 + }, + { + "epoch": 0.701898363169049, + "grad_norm": 2.115553897806706, + "learning_rate": 8.618372380567696e-07, + "loss": 1.0541, + "step": 7783 + }, + { + "epoch": 0.7019885466925193, + "grad_norm": 1.91015269273298, + "learning_rate": 8.613569012007478e-07, + "loss": 0.9428, + "step": 7784 + }, + { + "epoch": 0.7020787302159895, + "grad_norm": 1.8255705633931245, + "learning_rate": 8.608766615000338e-07, + "loss": 1.0184, + "step": 7785 + }, + { + "epoch": 0.7021689137394598, + "grad_norm": 1.9433155150711716, + "learning_rate": 8.603965189956039e-07, + "loss": 1.0008, + "step": 7786 + }, + { + "epoch": 0.70225909726293, + "grad_norm": 2.011159874773916, + "learning_rate": 8.599164737284276e-07, + "loss": 0.9881, + "step": 7787 + }, + { + "epoch": 0.7023492807864004, + "grad_norm": 1.6174204193301787, + "learning_rate": 8.594365257394634e-07, + "loss": 0.9397, + "step": 7788 + }, + { + "epoch": 0.7024394643098706, + "grad_norm": 1.8519915634894133, + "learning_rate": 8.589566750696637e-07, + "loss": 0.9256, + "step": 7789 + }, + { + "epoch": 0.7025296478333408, + "grad_norm": 1.7078745698948534, + "learning_rate": 8.584769217599721e-07, + "loss": 0.916, + "step": 7790 + }, + { + "epoch": 0.7026198313568112, + "grad_norm": 1.758677494422564, + "learning_rate": 8.579972658513239e-07, + "loss": 0.9258, + "step": 7791 + }, + { + "epoch": 0.7027100148802814, + "grad_norm": 2.211378289485764, + "learning_rate": 8.57517707384646e-07, + "loss": 0.9892, + "step": 7792 + }, + { + "epoch": 0.7028001984037516, + "grad_norm": 1.5312093612572086, + "learning_rate": 8.570382464008574e-07, + "loss": 1.0203, + "step": 7793 + }, + { + "epoch": 0.7028903819272219, + "grad_norm": 1.7340984639521697, + "learning_rate": 8.565588829408665e-07, + "loss": 0.9519, + "step": 7794 + }, + { + "epoch": 0.7029805654506922, + "grad_norm": 2.8053278709994225, + "learning_rate": 8.560796170455782e-07, + "loss": 0.9629, + "step": 7795 + }, + { + "epoch": 0.7030707489741624, + "grad_norm": 1.6609025063331753, + "learning_rate": 8.556004487558828e-07, + "loss": 1.0201, + "step": 7796 + }, + { + "epoch": 0.7031609324976327, + "grad_norm": 1.4823872156325832, + "learning_rate": 8.55121378112669e-07, + "loss": 0.8978, + "step": 7797 + }, + { + "epoch": 0.7032511160211029, + "grad_norm": 1.6481412648514842, + "learning_rate": 8.546424051568111e-07, + "loss": 0.9594, + "step": 7798 + }, + { + "epoch": 0.7033412995445733, + "grad_norm": 3.0509509870955287, + "learning_rate": 8.541635299291785e-07, + "loss": 1.0342, + "step": 7799 + }, + { + "epoch": 0.7034314830680435, + "grad_norm": 0.5910904448071691, + "learning_rate": 8.536847524706317e-07, + "loss": 0.8366, + "step": 7800 + }, + { + "epoch": 0.7035216665915137, + "grad_norm": 1.821813725640134, + "learning_rate": 8.532060728220225e-07, + "loss": 0.9138, + "step": 7801 + }, + { + "epoch": 0.703611850114984, + "grad_norm": 2.028140108260909, + "learning_rate": 8.527274910241955e-07, + "loss": 0.9925, + "step": 7802 + }, + { + "epoch": 0.7037020336384543, + "grad_norm": 1.9280139291386769, + "learning_rate": 8.522490071179833e-07, + "loss": 0.9786, + "step": 7803 + }, + { + "epoch": 0.7037922171619245, + "grad_norm": 1.3515939874233642, + "learning_rate": 8.517706211442159e-07, + "loss": 0.8594, + "step": 7804 + }, + { + "epoch": 0.7038824006853948, + "grad_norm": 1.9080067653616035, + "learning_rate": 8.512923331437097e-07, + "loss": 1.0577, + "step": 7805 + }, + { + "epoch": 0.703972584208865, + "grad_norm": 1.6538336059012408, + "learning_rate": 8.508141431572755e-07, + "loss": 0.9483, + "step": 7806 + }, + { + "epoch": 0.7040627677323353, + "grad_norm": 1.9828908581652969, + "learning_rate": 8.503360512257152e-07, + "loss": 0.9661, + "step": 7807 + }, + { + "epoch": 0.7041529512558056, + "grad_norm": 1.8530399528144785, + "learning_rate": 8.498580573898219e-07, + "loss": 0.9473, + "step": 7808 + }, + { + "epoch": 0.7042431347792758, + "grad_norm": 2.392126073366307, + "learning_rate": 8.493801616903813e-07, + "loss": 0.9894, + "step": 7809 + }, + { + "epoch": 0.704333318302746, + "grad_norm": 1.664183634615937, + "learning_rate": 8.489023641681705e-07, + "loss": 0.9974, + "step": 7810 + }, + { + "epoch": 0.7044235018262164, + "grad_norm": 1.5013450314239387, + "learning_rate": 8.484246648639555e-07, + "loss": 1.0295, + "step": 7811 + }, + { + "epoch": 0.7045136853496866, + "grad_norm": 1.3468464505384885, + "learning_rate": 8.479470638184994e-07, + "loss": 0.8552, + "step": 7812 + }, + { + "epoch": 0.7046038688731568, + "grad_norm": 1.9921751134150774, + "learning_rate": 8.474695610725513e-07, + "loss": 0.9002, + "step": 7813 + }, + { + "epoch": 0.7046940523966272, + "grad_norm": 1.3754502339597219, + "learning_rate": 8.469921566668552e-07, + "loss": 1.0533, + "step": 7814 + }, + { + "epoch": 0.7047842359200974, + "grad_norm": 1.6627232709980875, + "learning_rate": 8.46514850642146e-07, + "loss": 1.03, + "step": 7815 + }, + { + "epoch": 0.7048744194435677, + "grad_norm": 2.216287885536793, + "learning_rate": 8.460376430391499e-07, + "loss": 0.9682, + "step": 7816 + }, + { + "epoch": 0.7049646029670379, + "grad_norm": 1.4452324097583553, + "learning_rate": 8.455605338985858e-07, + "loss": 1.01, + "step": 7817 + }, + { + "epoch": 0.7050547864905082, + "grad_norm": 19.03620352585395, + "learning_rate": 8.45083523261161e-07, + "loss": 0.8381, + "step": 7818 + }, + { + "epoch": 0.7051449700139785, + "grad_norm": 1.6188117085481386, + "learning_rate": 8.446066111675796e-07, + "loss": 0.9881, + "step": 7819 + }, + { + "epoch": 0.7052351535374487, + "grad_norm": 2.665518980612338, + "learning_rate": 8.441297976585314e-07, + "loss": 0.874, + "step": 7820 + }, + { + "epoch": 0.7053253370609189, + "grad_norm": 1.6561858686590767, + "learning_rate": 8.436530827747037e-07, + "loss": 1.0657, + "step": 7821 + }, + { + "epoch": 0.7054155205843893, + "grad_norm": 1.7809821479583774, + "learning_rate": 8.431764665567704e-07, + "loss": 1.1182, + "step": 7822 + }, + { + "epoch": 0.7055057041078595, + "grad_norm": 1.8779510163570363, + "learning_rate": 8.426999490453996e-07, + "loss": 0.9171, + "step": 7823 + }, + { + "epoch": 0.7055958876313297, + "grad_norm": 2.0827641154548013, + "learning_rate": 8.422235302812504e-07, + "loss": 0.9878, + "step": 7824 + }, + { + "epoch": 0.7056860711548, + "grad_norm": 1.5717176536846726, + "learning_rate": 8.417472103049734e-07, + "loss": 1.0534, + "step": 7825 + }, + { + "epoch": 0.7057762546782703, + "grad_norm": 1.6309736931303047, + "learning_rate": 8.412709891572112e-07, + "loss": 1.0174, + "step": 7826 + }, + { + "epoch": 0.7058664382017406, + "grad_norm": 1.9294094858402384, + "learning_rate": 8.407948668785978e-07, + "loss": 1.0206, + "step": 7827 + }, + { + "epoch": 0.7059566217252108, + "grad_norm": 1.6546336510156823, + "learning_rate": 8.403188435097576e-07, + "loss": 0.9711, + "step": 7828 + }, + { + "epoch": 0.706046805248681, + "grad_norm": 1.4729222226971084, + "learning_rate": 8.398429190913081e-07, + "loss": 1.0553, + "step": 7829 + }, + { + "epoch": 0.7061369887721514, + "grad_norm": 1.6813709428247796, + "learning_rate": 8.393670936638578e-07, + "loss": 0.9818, + "step": 7830 + }, + { + "epoch": 0.7062271722956216, + "grad_norm": 1.6541977169219892, + "learning_rate": 8.388913672680067e-07, + "loss": 0.9796, + "step": 7831 + }, + { + "epoch": 0.7063173558190918, + "grad_norm": 1.8148817001756132, + "learning_rate": 8.384157399443472e-07, + "loss": 0.9267, + "step": 7832 + }, + { + "epoch": 0.7064075393425621, + "grad_norm": 1.530190471062777, + "learning_rate": 8.379402117334601e-07, + "loss": 1.0063, + "step": 7833 + }, + { + "epoch": 0.7064977228660324, + "grad_norm": 1.7925239626294964, + "learning_rate": 8.374647826759232e-07, + "loss": 0.9816, + "step": 7834 + }, + { + "epoch": 0.7065879063895026, + "grad_norm": 1.613245337873976, + "learning_rate": 8.369894528122998e-07, + "loss": 0.9797, + "step": 7835 + }, + { + "epoch": 0.7066780899129729, + "grad_norm": 1.8915329951757065, + "learning_rate": 8.365142221831505e-07, + "loss": 0.8442, + "step": 7836 + }, + { + "epoch": 0.7067682734364432, + "grad_norm": 1.608524579288135, + "learning_rate": 8.360390908290222e-07, + "loss": 0.9403, + "step": 7837 + }, + { + "epoch": 0.7068584569599135, + "grad_norm": 3.221924336866885, + "learning_rate": 8.355640587904569e-07, + "loss": 0.9638, + "step": 7838 + }, + { + "epoch": 0.7069486404833837, + "grad_norm": 1.843642021508758, + "learning_rate": 8.350891261079866e-07, + "loss": 0.9916, + "step": 7839 + }, + { + "epoch": 0.7070388240068539, + "grad_norm": 1.7834037342020033, + "learning_rate": 8.346142928221356e-07, + "loss": 1.0287, + "step": 7840 + }, + { + "epoch": 0.7071290075303243, + "grad_norm": 1.8201060034724195, + "learning_rate": 8.341395589734189e-07, + "loss": 0.8219, + "step": 7841 + }, + { + "epoch": 0.7072191910537945, + "grad_norm": 3.074745453013969, + "learning_rate": 8.336649246023433e-07, + "loss": 0.9359, + "step": 7842 + }, + { + "epoch": 0.7073093745772647, + "grad_norm": 1.3835904793940292, + "learning_rate": 8.331903897494084e-07, + "loss": 0.8821, + "step": 7843 + }, + { + "epoch": 0.707399558100735, + "grad_norm": 1.9991809240640837, + "learning_rate": 8.327159544551024e-07, + "loss": 0.9886, + "step": 7844 + }, + { + "epoch": 0.7074897416242053, + "grad_norm": 2.263914844023509, + "learning_rate": 8.322416187599073e-07, + "loss": 0.9433, + "step": 7845 + }, + { + "epoch": 0.7075799251476755, + "grad_norm": 1.7559070347310304, + "learning_rate": 8.317673827042963e-07, + "loss": 1.0047, + "step": 7846 + }, + { + "epoch": 0.7076701086711458, + "grad_norm": 1.7632657382943653, + "learning_rate": 8.312932463287339e-07, + "loss": 0.9679, + "step": 7847 + }, + { + "epoch": 0.707760292194616, + "grad_norm": 2.0861480263758034, + "learning_rate": 8.308192096736759e-07, + "loss": 0.9497, + "step": 7848 + }, + { + "epoch": 0.7078504757180863, + "grad_norm": 1.7466810634010013, + "learning_rate": 8.303452727795703e-07, + "loss": 0.959, + "step": 7849 + }, + { + "epoch": 0.7079406592415566, + "grad_norm": 1.7531780268899466, + "learning_rate": 8.298714356868542e-07, + "loss": 1.0601, + "step": 7850 + }, + { + "epoch": 0.7080308427650268, + "grad_norm": 1.6807818728909745, + "learning_rate": 8.293976984359605e-07, + "loss": 0.9957, + "step": 7851 + }, + { + "epoch": 0.708121026288497, + "grad_norm": 2.1209748236543695, + "learning_rate": 8.289240610673092e-07, + "loss": 0.9826, + "step": 7852 + }, + { + "epoch": 0.7082112098119674, + "grad_norm": 1.6056090801752039, + "learning_rate": 8.284505236213144e-07, + "loss": 1.0127, + "step": 7853 + }, + { + "epoch": 0.7083013933354376, + "grad_norm": 1.8920402589268936, + "learning_rate": 8.279770861383806e-07, + "loss": 0.9725, + "step": 7854 + }, + { + "epoch": 0.7083915768589079, + "grad_norm": 1.9187609246654334, + "learning_rate": 8.275037486589042e-07, + "loss": 0.9744, + "step": 7855 + }, + { + "epoch": 0.7084817603823781, + "grad_norm": 0.6661325614945279, + "learning_rate": 8.270305112232739e-07, + "loss": 0.7958, + "step": 7856 + }, + { + "epoch": 0.7085719439058484, + "grad_norm": 2.005682501938925, + "learning_rate": 8.265573738718665e-07, + "loss": 0.9769, + "step": 7857 + }, + { + "epoch": 0.7086621274293187, + "grad_norm": 1.7085351669671027, + "learning_rate": 8.260843366450559e-07, + "loss": 0.9403, + "step": 7858 + }, + { + "epoch": 0.7087523109527889, + "grad_norm": 1.9468622124559665, + "learning_rate": 8.256113995832017e-07, + "loss": 0.9777, + "step": 7859 + }, + { + "epoch": 0.7088424944762591, + "grad_norm": 1.4306743113700733, + "learning_rate": 8.251385627266583e-07, + "loss": 1.0213, + "step": 7860 + }, + { + "epoch": 0.7089326779997295, + "grad_norm": 2.4030096852349065, + "learning_rate": 8.24665826115771e-07, + "loss": 0.9945, + "step": 7861 + }, + { + "epoch": 0.7090228615231997, + "grad_norm": 1.5533756875060953, + "learning_rate": 8.241931897908763e-07, + "loss": 0.9994, + "step": 7862 + }, + { + "epoch": 0.7091130450466699, + "grad_norm": 1.7512098626078194, + "learning_rate": 8.237206537923016e-07, + "loss": 1.0017, + "step": 7863 + }, + { + "epoch": 0.7092032285701403, + "grad_norm": 3.0327484517161523, + "learning_rate": 8.232482181603671e-07, + "loss": 0.9716, + "step": 7864 + }, + { + "epoch": 0.7092934120936105, + "grad_norm": 1.6710643763945576, + "learning_rate": 8.227758829353828e-07, + "loss": 0.9619, + "step": 7865 + }, + { + "epoch": 0.7093835956170808, + "grad_norm": 1.5876981236135772, + "learning_rate": 8.223036481576522e-07, + "loss": 0.9516, + "step": 7866 + }, + { + "epoch": 0.709473779140551, + "grad_norm": 4.466156106683094, + "learning_rate": 8.218315138674672e-07, + "loss": 0.9834, + "step": 7867 + }, + { + "epoch": 0.7095639626640213, + "grad_norm": 0.6213102383259199, + "learning_rate": 8.21359480105114e-07, + "loss": 0.8132, + "step": 7868 + }, + { + "epoch": 0.7096541461874916, + "grad_norm": 1.7388008626907205, + "learning_rate": 8.208875469108689e-07, + "loss": 1.0148, + "step": 7869 + }, + { + "epoch": 0.7097443297109618, + "grad_norm": 2.2220096552696553, + "learning_rate": 8.204157143249997e-07, + "loss": 0.966, + "step": 7870 + }, + { + "epoch": 0.709834513234432, + "grad_norm": 1.417276176529854, + "learning_rate": 8.199439823877668e-07, + "loss": 0.9533, + "step": 7871 + }, + { + "epoch": 0.7099246967579024, + "grad_norm": 1.4972180954863124, + "learning_rate": 8.194723511394186e-07, + "loss": 0.9657, + "step": 7872 + }, + { + "epoch": 0.7100148802813726, + "grad_norm": 1.6257486086221602, + "learning_rate": 8.190008206202002e-07, + "loss": 0.9288, + "step": 7873 + }, + { + "epoch": 0.7101050638048428, + "grad_norm": 3.3441690913163944, + "learning_rate": 8.185293908703423e-07, + "loss": 1.0244, + "step": 7874 + }, + { + "epoch": 0.7101952473283131, + "grad_norm": 1.8065632901522393, + "learning_rate": 8.180580619300727e-07, + "loss": 0.9434, + "step": 7875 + }, + { + "epoch": 0.7102854308517834, + "grad_norm": 1.274989627347154, + "learning_rate": 8.175868338396057e-07, + "loss": 0.9309, + "step": 7876 + }, + { + "epoch": 0.7103756143752537, + "grad_norm": 2.5046358042502654, + "learning_rate": 8.171157066391499e-07, + "loss": 0.9521, + "step": 7877 + }, + { + "epoch": 0.7104657978987239, + "grad_norm": 1.7719532136962082, + "learning_rate": 8.166446803689045e-07, + "loss": 1.0073, + "step": 7878 + }, + { + "epoch": 0.7105559814221941, + "grad_norm": 0.6683484387018593, + "learning_rate": 8.161737550690595e-07, + "loss": 0.82, + "step": 7879 + }, + { + "epoch": 0.7106461649456645, + "grad_norm": 1.442465622887308, + "learning_rate": 8.157029307797976e-07, + "loss": 1.0058, + "step": 7880 + }, + { + "epoch": 0.7107363484691347, + "grad_norm": 1.5189713179471473, + "learning_rate": 8.152322075412925e-07, + "loss": 1.0369, + "step": 7881 + }, + { + "epoch": 0.7108265319926049, + "grad_norm": 1.6094148501304928, + "learning_rate": 8.147615853937073e-07, + "loss": 0.9286, + "step": 7882 + }, + { + "epoch": 0.7109167155160752, + "grad_norm": 1.7502920724459639, + "learning_rate": 8.142910643771992e-07, + "loss": 0.9027, + "step": 7883 + }, + { + "epoch": 0.7110068990395455, + "grad_norm": 2.2002675890744503, + "learning_rate": 8.138206445319152e-07, + "loss": 0.9428, + "step": 7884 + }, + { + "epoch": 0.7110970825630157, + "grad_norm": 2.769707462962362, + "learning_rate": 8.133503258979944e-07, + "loss": 0.8798, + "step": 7885 + }, + { + "epoch": 0.711187266086486, + "grad_norm": 2.1107568806164827, + "learning_rate": 8.12880108515567e-07, + "loss": 0.9021, + "step": 7886 + }, + { + "epoch": 0.7112774496099563, + "grad_norm": 2.1241791205516884, + "learning_rate": 8.124099924247543e-07, + "loss": 0.9517, + "step": 7887 + }, + { + "epoch": 0.7113676331334265, + "grad_norm": 1.6858689938574922, + "learning_rate": 8.119399776656701e-07, + "loss": 0.8832, + "step": 7888 + }, + { + "epoch": 0.7114578166568968, + "grad_norm": 1.7621667127470642, + "learning_rate": 8.114700642784167e-07, + "loss": 0.9762, + "step": 7889 + }, + { + "epoch": 0.711548000180367, + "grad_norm": 1.8574024283657669, + "learning_rate": 8.110002523030921e-07, + "loss": 1.0368, + "step": 7890 + }, + { + "epoch": 0.7116381837038374, + "grad_norm": 4.730646560061395, + "learning_rate": 8.105305417797808e-07, + "loss": 0.9992, + "step": 7891 + }, + { + "epoch": 0.7117283672273076, + "grad_norm": 1.672586663913165, + "learning_rate": 8.100609327485635e-07, + "loss": 1.0156, + "step": 7892 + }, + { + "epoch": 0.7118185507507778, + "grad_norm": 2.2452506697283874, + "learning_rate": 8.095914252495082e-07, + "loss": 1.0356, + "step": 7893 + }, + { + "epoch": 0.7119087342742481, + "grad_norm": 1.8149055434309902, + "learning_rate": 8.091220193226762e-07, + "loss": 0.9524, + "step": 7894 + }, + { + "epoch": 0.7119989177977184, + "grad_norm": 1.487063252420427, + "learning_rate": 8.0865271500812e-07, + "loss": 0.9644, + "step": 7895 + }, + { + "epoch": 0.7120891013211886, + "grad_norm": 2.3112640170821974, + "learning_rate": 8.081835123458831e-07, + "loss": 1.0198, + "step": 7896 + }, + { + "epoch": 0.7121792848446589, + "grad_norm": 1.769384382270885, + "learning_rate": 8.077144113760013e-07, + "loss": 0.9628, + "step": 7897 + }, + { + "epoch": 0.7122694683681291, + "grad_norm": 1.8204299168516702, + "learning_rate": 8.072454121384995e-07, + "loss": 0.9594, + "step": 7898 + }, + { + "epoch": 0.7123596518915994, + "grad_norm": 1.4316661046611376, + "learning_rate": 8.067765146733958e-07, + "loss": 1.0215, + "step": 7899 + }, + { + "epoch": 0.7124498354150697, + "grad_norm": 2.839527390606041, + "learning_rate": 8.063077190206993e-07, + "loss": 1.0533, + "step": 7900 + }, + { + "epoch": 0.7125400189385399, + "grad_norm": 0.6004547968406405, + "learning_rate": 8.058390252204101e-07, + "loss": 0.7305, + "step": 7901 + }, + { + "epoch": 0.7126302024620101, + "grad_norm": 1.6468275986279381, + "learning_rate": 8.0537043331252e-07, + "loss": 0.9699, + "step": 7902 + }, + { + "epoch": 0.7127203859854805, + "grad_norm": 2.0159111115599053, + "learning_rate": 8.049019433370121e-07, + "loss": 0.9159, + "step": 7903 + }, + { + "epoch": 0.7128105695089507, + "grad_norm": 1.5528083063914604, + "learning_rate": 8.044335553338588e-07, + "loss": 0.9805, + "step": 7904 + }, + { + "epoch": 0.712900753032421, + "grad_norm": 5.028300682588164, + "learning_rate": 8.039652693430281e-07, + "loss": 0.9897, + "step": 7905 + }, + { + "epoch": 0.7129909365558912, + "grad_norm": 1.9918824205359087, + "learning_rate": 8.034970854044742e-07, + "loss": 0.9549, + "step": 7906 + }, + { + "epoch": 0.7130811200793615, + "grad_norm": 2.3386997999209713, + "learning_rate": 8.03029003558148e-07, + "loss": 1.0716, + "step": 7907 + }, + { + "epoch": 0.7131713036028318, + "grad_norm": 3.0520868572609685, + "learning_rate": 8.025610238439864e-07, + "loss": 1.0426, + "step": 7908 + }, + { + "epoch": 0.713261487126302, + "grad_norm": 1.7729036981427344, + "learning_rate": 8.020931463019207e-07, + "loss": 1.056, + "step": 7909 + }, + { + "epoch": 0.7133516706497723, + "grad_norm": 1.7965785569836834, + "learning_rate": 8.016253709718732e-07, + "loss": 0.9239, + "step": 7910 + }, + { + "epoch": 0.7134418541732426, + "grad_norm": 3.3315012824525385, + "learning_rate": 8.011576978937567e-07, + "loss": 0.8827, + "step": 7911 + }, + { + "epoch": 0.7135320376967128, + "grad_norm": 1.5228941119498585, + "learning_rate": 8.006901271074764e-07, + "loss": 0.9787, + "step": 7912 + }, + { + "epoch": 0.713622221220183, + "grad_norm": 2.2390987374673608, + "learning_rate": 8.002226586529261e-07, + "loss": 1.0189, + "step": 7913 + }, + { + "epoch": 0.7137124047436534, + "grad_norm": 1.4670036160061093, + "learning_rate": 7.997552925699956e-07, + "loss": 1.0576, + "step": 7914 + }, + { + "epoch": 0.7138025882671236, + "grad_norm": 1.9955699255727277, + "learning_rate": 7.992880288985606e-07, + "loss": 1.0716, + "step": 7915 + }, + { + "epoch": 0.7138927717905938, + "grad_norm": 2.6361691954787654, + "learning_rate": 7.988208676784918e-07, + "loss": 0.9862, + "step": 7916 + }, + { + "epoch": 0.7139829553140641, + "grad_norm": 1.964407536828573, + "learning_rate": 7.983538089496497e-07, + "loss": 0.9735, + "step": 7917 + }, + { + "epoch": 0.7140731388375344, + "grad_norm": 1.71617877409545, + "learning_rate": 7.978868527518864e-07, + "loss": 0.9356, + "step": 7918 + }, + { + "epoch": 0.7141633223610047, + "grad_norm": 1.4807746188764073, + "learning_rate": 7.974199991250455e-07, + "loss": 0.953, + "step": 7919 + }, + { + "epoch": 0.7142535058844749, + "grad_norm": 1.6475943628357808, + "learning_rate": 7.969532481089616e-07, + "loss": 1.077, + "step": 7920 + }, + { + "epoch": 0.7143436894079451, + "grad_norm": 1.8695710741863563, + "learning_rate": 7.964865997434589e-07, + "loss": 0.9925, + "step": 7921 + }, + { + "epoch": 0.7144338729314155, + "grad_norm": 1.7568759033678487, + "learning_rate": 7.96020054068357e-07, + "loss": 1.0531, + "step": 7922 + }, + { + "epoch": 0.7145240564548857, + "grad_norm": 1.268686195234927, + "learning_rate": 7.95553611123462e-07, + "loss": 0.9355, + "step": 7923 + }, + { + "epoch": 0.7146142399783559, + "grad_norm": 3.4635023683694346, + "learning_rate": 7.950872709485741e-07, + "loss": 1.1629, + "step": 7924 + }, + { + "epoch": 0.7147044235018262, + "grad_norm": 2.24936338531387, + "learning_rate": 7.946210335834842e-07, + "loss": 1.0327, + "step": 7925 + }, + { + "epoch": 0.7147946070252965, + "grad_norm": 1.3648174997391773, + "learning_rate": 7.94154899067974e-07, + "loss": 0.9028, + "step": 7926 + }, + { + "epoch": 0.7148847905487667, + "grad_norm": 2.0434679402523925, + "learning_rate": 7.936888674418177e-07, + "loss": 0.9094, + "step": 7927 + }, + { + "epoch": 0.714974974072237, + "grad_norm": 2.0625376842408416, + "learning_rate": 7.932229387447771e-07, + "loss": 0.9792, + "step": 7928 + }, + { + "epoch": 0.7150651575957072, + "grad_norm": 1.8277174014342632, + "learning_rate": 7.927571130166109e-07, + "loss": 0.9708, + "step": 7929 + }, + { + "epoch": 0.7151553411191776, + "grad_norm": 0.6086502656171904, + "learning_rate": 7.922913902970632e-07, + "loss": 0.8039, + "step": 7930 + }, + { + "epoch": 0.7152455246426478, + "grad_norm": 1.8273206065956078, + "learning_rate": 7.918257706258744e-07, + "loss": 0.9698, + "step": 7931 + }, + { + "epoch": 0.715335708166118, + "grad_norm": 1.8016793417447985, + "learning_rate": 7.913602540427724e-07, + "loss": 0.9809, + "step": 7932 + }, + { + "epoch": 0.7154258916895884, + "grad_norm": 1.7310760328011305, + "learning_rate": 7.908948405874775e-07, + "loss": 0.9741, + "step": 7933 + }, + { + "epoch": 0.7155160752130586, + "grad_norm": 1.7663234578250282, + "learning_rate": 7.904295302997019e-07, + "loss": 0.9106, + "step": 7934 + }, + { + "epoch": 0.7156062587365288, + "grad_norm": 1.6138462804413967, + "learning_rate": 7.899643232191484e-07, + "loss": 0.9922, + "step": 7935 + }, + { + "epoch": 0.7156964422599991, + "grad_norm": 1.6380424154423499, + "learning_rate": 7.894992193855108e-07, + "loss": 0.9607, + "step": 7936 + }, + { + "epoch": 0.7157866257834694, + "grad_norm": 1.5159303809765312, + "learning_rate": 7.890342188384751e-07, + "loss": 0.9279, + "step": 7937 + }, + { + "epoch": 0.7158768093069396, + "grad_norm": 1.7000317262045226, + "learning_rate": 7.885693216177165e-07, + "loss": 0.9222, + "step": 7938 + }, + { + "epoch": 0.7159669928304099, + "grad_norm": 2.0395610097638235, + "learning_rate": 7.88104527762903e-07, + "loss": 0.9753, + "step": 7939 + }, + { + "epoch": 0.7160571763538801, + "grad_norm": 1.8897581398833156, + "learning_rate": 7.876398373136936e-07, + "loss": 0.9351, + "step": 7940 + }, + { + "epoch": 0.7161473598773505, + "grad_norm": 1.4988163410486075, + "learning_rate": 7.87175250309738e-07, + "loss": 0.9747, + "step": 7941 + }, + { + "epoch": 0.7162375434008207, + "grad_norm": 1.5724650949506818, + "learning_rate": 7.867107667906785e-07, + "loss": 1.0543, + "step": 7942 + }, + { + "epoch": 0.7163277269242909, + "grad_norm": 1.6315645258141467, + "learning_rate": 7.862463867961446e-07, + "loss": 0.9772, + "step": 7943 + }, + { + "epoch": 0.7164179104477612, + "grad_norm": 1.3695842116844086, + "learning_rate": 7.857821103657632e-07, + "loss": 1.0426, + "step": 7944 + }, + { + "epoch": 0.7165080939712315, + "grad_norm": 1.4217427789603, + "learning_rate": 7.853179375391459e-07, + "loss": 1.0072, + "step": 7945 + }, + { + "epoch": 0.7165982774947017, + "grad_norm": 1.6739240549878796, + "learning_rate": 7.848538683559012e-07, + "loss": 0.9648, + "step": 7946 + }, + { + "epoch": 0.716688461018172, + "grad_norm": 2.511135573488361, + "learning_rate": 7.843899028556238e-07, + "loss": 1.0184, + "step": 7947 + }, + { + "epoch": 0.7167786445416422, + "grad_norm": 1.3392122966079902, + "learning_rate": 7.839260410779029e-07, + "loss": 0.9692, + "step": 7948 + }, + { + "epoch": 0.7168688280651125, + "grad_norm": 1.4327136256257176, + "learning_rate": 7.834622830623175e-07, + "loss": 1.0015, + "step": 7949 + }, + { + "epoch": 0.7169590115885828, + "grad_norm": 1.7413518295581223, + "learning_rate": 7.82998628848438e-07, + "loss": 0.9503, + "step": 7950 + }, + { + "epoch": 0.717049195112053, + "grad_norm": 1.4383166108350938, + "learning_rate": 7.825350784758261e-07, + "loss": 0.9686, + "step": 7951 + }, + { + "epoch": 0.7171393786355232, + "grad_norm": 1.6289477445956986, + "learning_rate": 7.820716319840342e-07, + "loss": 0.9422, + "step": 7952 + }, + { + "epoch": 0.7172295621589936, + "grad_norm": 2.9452606925151805, + "learning_rate": 7.816082894126069e-07, + "loss": 1.0566, + "step": 7953 + }, + { + "epoch": 0.7173197456824638, + "grad_norm": 1.9482951145105578, + "learning_rate": 7.811450508010778e-07, + "loss": 0.9485, + "step": 7954 + }, + { + "epoch": 0.717409929205934, + "grad_norm": 1.3970962571068641, + "learning_rate": 7.806819161889737e-07, + "loss": 0.9496, + "step": 7955 + }, + { + "epoch": 0.7175001127294043, + "grad_norm": 1.822373300133055, + "learning_rate": 7.802188856158119e-07, + "loss": 1.0081, + "step": 7956 + }, + { + "epoch": 0.7175902962528746, + "grad_norm": 1.7561766069491793, + "learning_rate": 7.797559591211002e-07, + "loss": 0.9677, + "step": 7957 + }, + { + "epoch": 0.7176804797763449, + "grad_norm": 1.693752995449669, + "learning_rate": 7.79293136744339e-07, + "loss": 0.9423, + "step": 7958 + }, + { + "epoch": 0.7177706632998151, + "grad_norm": 1.8777373994565238, + "learning_rate": 7.788304185250185e-07, + "loss": 0.9763, + "step": 7959 + }, + { + "epoch": 0.7178608468232854, + "grad_norm": 1.6071660706190865, + "learning_rate": 7.78367804502619e-07, + "loss": 1.0066, + "step": 7960 + }, + { + "epoch": 0.7179510303467557, + "grad_norm": 1.8796215321673169, + "learning_rate": 7.779052947166156e-07, + "loss": 0.986, + "step": 7961 + }, + { + "epoch": 0.7180412138702259, + "grad_norm": 2.1911286910970076, + "learning_rate": 7.774428892064697e-07, + "loss": 0.9318, + "step": 7962 + }, + { + "epoch": 0.7181313973936961, + "grad_norm": 2.2082290504832587, + "learning_rate": 7.769805880116391e-07, + "loss": 0.9675, + "step": 7963 + }, + { + "epoch": 0.7182215809171665, + "grad_norm": 2.1103678309472436, + "learning_rate": 7.765183911715678e-07, + "loss": 0.9556, + "step": 7964 + }, + { + "epoch": 0.7183117644406367, + "grad_norm": 1.447474496778179, + "learning_rate": 7.760562987256933e-07, + "loss": 0.8794, + "step": 7965 + }, + { + "epoch": 0.718401947964107, + "grad_norm": 1.7215047088523192, + "learning_rate": 7.755943107134444e-07, + "loss": 0.9803, + "step": 7966 + }, + { + "epoch": 0.7184921314875772, + "grad_norm": 0.6508440378955621, + "learning_rate": 7.751324271742401e-07, + "loss": 0.8051, + "step": 7967 + }, + { + "epoch": 0.7185823150110475, + "grad_norm": 1.5749267046388988, + "learning_rate": 7.746706481474916e-07, + "loss": 0.9906, + "step": 7968 + }, + { + "epoch": 0.7186724985345178, + "grad_norm": 0.59903336483466, + "learning_rate": 7.742089736725992e-07, + "loss": 0.7738, + "step": 7969 + }, + { + "epoch": 0.718762682057988, + "grad_norm": 1.7338592862295892, + "learning_rate": 7.737474037889559e-07, + "loss": 0.8992, + "step": 7970 + }, + { + "epoch": 0.7188528655814582, + "grad_norm": 1.973891855158845, + "learning_rate": 7.732859385359458e-07, + "loss": 0.9444, + "step": 7971 + }, + { + "epoch": 0.7189430491049286, + "grad_norm": 13.929025853607593, + "learning_rate": 7.728245779529434e-07, + "loss": 0.8974, + "step": 7972 + }, + { + "epoch": 0.7190332326283988, + "grad_norm": 1.5925507614932528, + "learning_rate": 7.723633220793146e-07, + "loss": 1.0088, + "step": 7973 + }, + { + "epoch": 0.719123416151869, + "grad_norm": 2.1599819671796476, + "learning_rate": 7.719021709544162e-07, + "loss": 0.8687, + "step": 7974 + }, + { + "epoch": 0.7192135996753393, + "grad_norm": 1.6998722617458795, + "learning_rate": 7.714411246175964e-07, + "loss": 1.0691, + "step": 7975 + }, + { + "epoch": 0.7193037831988096, + "grad_norm": 1.7665580377230523, + "learning_rate": 7.709801831081946e-07, + "loss": 0.8961, + "step": 7976 + }, + { + "epoch": 0.7193939667222798, + "grad_norm": 1.685668092261002, + "learning_rate": 7.705193464655391e-07, + "loss": 0.9334, + "step": 7977 + }, + { + "epoch": 0.7194841502457501, + "grad_norm": 1.2571911906264768, + "learning_rate": 7.700586147289534e-07, + "loss": 0.9532, + "step": 7978 + }, + { + "epoch": 0.7195743337692203, + "grad_norm": 2.1060526362015306, + "learning_rate": 7.695979879377481e-07, + "loss": 0.9332, + "step": 7979 + }, + { + "epoch": 0.7196645172926907, + "grad_norm": 1.677078634062002, + "learning_rate": 7.691374661312266e-07, + "loss": 0.9978, + "step": 7980 + }, + { + "epoch": 0.7197547008161609, + "grad_norm": 1.4611321793855565, + "learning_rate": 7.686770493486834e-07, + "loss": 0.9501, + "step": 7981 + }, + { + "epoch": 0.7198448843396311, + "grad_norm": 32.298407604540856, + "learning_rate": 7.68216737629404e-07, + "loss": 0.9192, + "step": 7982 + }, + { + "epoch": 0.7199350678631015, + "grad_norm": 1.550776330281869, + "learning_rate": 7.67756531012665e-07, + "loss": 1.0451, + "step": 7983 + }, + { + "epoch": 0.7200252513865717, + "grad_norm": 3.9554293102187033, + "learning_rate": 7.67296429537732e-07, + "loss": 1.0332, + "step": 7984 + }, + { + "epoch": 0.7201154349100419, + "grad_norm": 1.8258706563491953, + "learning_rate": 7.668364332438661e-07, + "loss": 1.041, + "step": 7985 + }, + { + "epoch": 0.7202056184335122, + "grad_norm": 1.627795236180711, + "learning_rate": 7.663765421703145e-07, + "loss": 0.9892, + "step": 7986 + }, + { + "epoch": 0.7202958019569825, + "grad_norm": 12.665301065281422, + "learning_rate": 7.659167563563187e-07, + "loss": 0.9159, + "step": 7987 + }, + { + "epoch": 0.7203859854804527, + "grad_norm": 1.5170023057702584, + "learning_rate": 7.654570758411096e-07, + "loss": 0.997, + "step": 7988 + }, + { + "epoch": 0.720476169003923, + "grad_norm": 1.8394276012659678, + "learning_rate": 7.649975006639103e-07, + "loss": 1.0379, + "step": 7989 + }, + { + "epoch": 0.7205663525273932, + "grad_norm": 0.6205186880741772, + "learning_rate": 7.645380308639337e-07, + "loss": 0.788, + "step": 7990 + }, + { + "epoch": 0.7206565360508635, + "grad_norm": 3.054940215679271, + "learning_rate": 7.640786664803853e-07, + "loss": 1.0239, + "step": 7991 + }, + { + "epoch": 0.7207467195743338, + "grad_norm": 1.4205408229634975, + "learning_rate": 7.636194075524587e-07, + "loss": 0.8964, + "step": 7992 + }, + { + "epoch": 0.720836903097804, + "grad_norm": 1.7564313421602622, + "learning_rate": 7.631602541193429e-07, + "loss": 0.9292, + "step": 7993 + }, + { + "epoch": 0.7209270866212742, + "grad_norm": 1.9223377942136641, + "learning_rate": 7.627012062202132e-07, + "loss": 1.032, + "step": 7994 + }, + { + "epoch": 0.7210172701447446, + "grad_norm": 2.0374677128954235, + "learning_rate": 7.622422638942391e-07, + "loss": 0.9741, + "step": 7995 + }, + { + "epoch": 0.7211074536682148, + "grad_norm": 1.8218048919746064, + "learning_rate": 7.617834271805801e-07, + "loss": 1.0416, + "step": 7996 + }, + { + "epoch": 0.7211976371916851, + "grad_norm": 0.6500325176468178, + "learning_rate": 7.613246961183863e-07, + "loss": 0.8022, + "step": 7997 + }, + { + "epoch": 0.7212878207151553, + "grad_norm": 1.564563990175812, + "learning_rate": 7.608660707468002e-07, + "loss": 0.9365, + "step": 7998 + }, + { + "epoch": 0.7213780042386256, + "grad_norm": 1.842210676647636, + "learning_rate": 7.604075511049522e-07, + "loss": 0.8699, + "step": 7999 + }, + { + "epoch": 0.7214681877620959, + "grad_norm": 2.051774777937355, + "learning_rate": 7.599491372319682e-07, + "loss": 0.9233, + "step": 8000 + }, + { + "epoch": 0.7215583712855661, + "grad_norm": 1.3398715091421052, + "learning_rate": 7.594908291669601e-07, + "loss": 0.9745, + "step": 8001 + }, + { + "epoch": 0.7216485548090363, + "grad_norm": 1.7220611582405971, + "learning_rate": 7.590326269490359e-07, + "loss": 1.0098, + "step": 8002 + }, + { + "epoch": 0.7217387383325067, + "grad_norm": 1.4037894020427264, + "learning_rate": 7.585745306172899e-07, + "loss": 0.933, + "step": 8003 + }, + { + "epoch": 0.7218289218559769, + "grad_norm": 2.1093466297643664, + "learning_rate": 7.5811654021081e-07, + "loss": 0.9607, + "step": 8004 + }, + { + "epoch": 0.7219191053794471, + "grad_norm": 1.4237749589535862, + "learning_rate": 7.576586557686748e-07, + "loss": 1.0054, + "step": 8005 + }, + { + "epoch": 0.7220092889029175, + "grad_norm": 1.839929922884094, + "learning_rate": 7.572008773299531e-07, + "loss": 0.9456, + "step": 8006 + }, + { + "epoch": 0.7220994724263877, + "grad_norm": 1.651392814126076, + "learning_rate": 7.567432049337055e-07, + "loss": 1.0133, + "step": 8007 + }, + { + "epoch": 0.722189655949858, + "grad_norm": 2.0316493888831717, + "learning_rate": 7.562856386189834e-07, + "loss": 0.9106, + "step": 8008 + }, + { + "epoch": 0.7222798394733282, + "grad_norm": 1.9744937610078466, + "learning_rate": 7.558281784248275e-07, + "loss": 0.9882, + "step": 8009 + }, + { + "epoch": 0.7223700229967985, + "grad_norm": 1.9257941255274447, + "learning_rate": 7.553708243902721e-07, + "loss": 0.9637, + "step": 8010 + }, + { + "epoch": 0.7224602065202688, + "grad_norm": 1.596924630605067, + "learning_rate": 7.549135765543404e-07, + "loss": 0.8667, + "step": 8011 + }, + { + "epoch": 0.722550390043739, + "grad_norm": 1.7207652675046123, + "learning_rate": 7.544564349560481e-07, + "loss": 0.9976, + "step": 8012 + }, + { + "epoch": 0.7226405735672092, + "grad_norm": 1.36715018085227, + "learning_rate": 7.539993996344009e-07, + "loss": 0.9596, + "step": 8013 + }, + { + "epoch": 0.7227307570906796, + "grad_norm": 1.538926714024123, + "learning_rate": 7.535424706283941e-07, + "loss": 0.9572, + "step": 8014 + }, + { + "epoch": 0.7228209406141498, + "grad_norm": 1.7544132171153866, + "learning_rate": 7.530856479770181e-07, + "loss": 0.9355, + "step": 8015 + }, + { + "epoch": 0.72291112413762, + "grad_norm": 3.731970743316095, + "learning_rate": 7.526289317192484e-07, + "loss": 1.0564, + "step": 8016 + }, + { + "epoch": 0.7230013076610903, + "grad_norm": 2.0285670241072062, + "learning_rate": 7.521723218940579e-07, + "loss": 0.9729, + "step": 8017 + }, + { + "epoch": 0.7230914911845606, + "grad_norm": 3.182121526204136, + "learning_rate": 7.517158185404038e-07, + "loss": 1.0373, + "step": 8018 + }, + { + "epoch": 0.7231816747080309, + "grad_norm": 1.788983672291945, + "learning_rate": 7.512594216972403e-07, + "loss": 0.9965, + "step": 8019 + }, + { + "epoch": 0.7232718582315011, + "grad_norm": 1.4189644672442498, + "learning_rate": 7.508031314035078e-07, + "loss": 1.0116, + "step": 8020 + }, + { + "epoch": 0.7233620417549713, + "grad_norm": 1.6256084770116779, + "learning_rate": 7.503469476981401e-07, + "loss": 0.9952, + "step": 8021 + }, + { + "epoch": 0.7234522252784417, + "grad_norm": 2.0002467480082253, + "learning_rate": 7.498908706200613e-07, + "loss": 0.8485, + "step": 8022 + }, + { + "epoch": 0.7235424088019119, + "grad_norm": 4.342051002248583, + "learning_rate": 7.494349002081866e-07, + "loss": 1.0232, + "step": 8023 + }, + { + "epoch": 0.7236325923253821, + "grad_norm": 1.5569874222822553, + "learning_rate": 7.489790365014224e-07, + "loss": 1.0054, + "step": 8024 + }, + { + "epoch": 0.7237227758488524, + "grad_norm": 1.7887454355796089, + "learning_rate": 7.485232795386642e-07, + "loss": 0.9938, + "step": 8025 + }, + { + "epoch": 0.7238129593723227, + "grad_norm": 1.4175464852597655, + "learning_rate": 7.480676293588002e-07, + "loss": 1.0008, + "step": 8026 + }, + { + "epoch": 0.7239031428957929, + "grad_norm": 0.7792081854123014, + "learning_rate": 7.476120860007093e-07, + "loss": 0.8332, + "step": 8027 + }, + { + "epoch": 0.7239933264192632, + "grad_norm": 1.7150300430913898, + "learning_rate": 7.471566495032608e-07, + "loss": 1.0196, + "step": 8028 + }, + { + "epoch": 0.7240835099427335, + "grad_norm": 2.4184174498798887, + "learning_rate": 7.467013199053152e-07, + "loss": 0.9096, + "step": 8029 + }, + { + "epoch": 0.7241736934662037, + "grad_norm": 1.6456346593124922, + "learning_rate": 7.46246097245724e-07, + "loss": 0.9717, + "step": 8030 + }, + { + "epoch": 0.724263876989674, + "grad_norm": 3.0286440314924152, + "learning_rate": 7.457909815633276e-07, + "loss": 0.9828, + "step": 8031 + }, + { + "epoch": 0.7243540605131442, + "grad_norm": 2.817321036205303, + "learning_rate": 7.453359728969618e-07, + "loss": 0.9692, + "step": 8032 + }, + { + "epoch": 0.7244442440366146, + "grad_norm": 1.954123707063899, + "learning_rate": 7.448810712854475e-07, + "loss": 0.9007, + "step": 8033 + }, + { + "epoch": 0.7245344275600848, + "grad_norm": 3.0206877750098964, + "learning_rate": 7.444262767676022e-07, + "loss": 1.0172, + "step": 8034 + }, + { + "epoch": 0.724624611083555, + "grad_norm": 1.82549141838838, + "learning_rate": 7.439715893822296e-07, + "loss": 1.0503, + "step": 8035 + }, + { + "epoch": 0.7247147946070253, + "grad_norm": 1.703972491866139, + "learning_rate": 7.435170091681264e-07, + "loss": 0.9965, + "step": 8036 + }, + { + "epoch": 0.7248049781304956, + "grad_norm": 1.680052180955774, + "learning_rate": 7.430625361640803e-07, + "loss": 0.9326, + "step": 8037 + }, + { + "epoch": 0.7248951616539658, + "grad_norm": 1.716851729383541, + "learning_rate": 7.426081704088694e-07, + "loss": 1.0186, + "step": 8038 + }, + { + "epoch": 0.7249853451774361, + "grad_norm": 2.307495165620412, + "learning_rate": 7.42153911941263e-07, + "loss": 1.048, + "step": 8039 + }, + { + "epoch": 0.7250755287009063, + "grad_norm": 1.6514857163350052, + "learning_rate": 7.416997608000192e-07, + "loss": 1.0161, + "step": 8040 + }, + { + "epoch": 0.7251657122243766, + "grad_norm": 1.7290185918564056, + "learning_rate": 7.412457170238918e-07, + "loss": 0.9581, + "step": 8041 + }, + { + "epoch": 0.7252558957478469, + "grad_norm": 1.6648004892314612, + "learning_rate": 7.407917806516193e-07, + "loss": 0.922, + "step": 8042 + }, + { + "epoch": 0.7253460792713171, + "grad_norm": 1.6175580176158733, + "learning_rate": 7.403379517219354e-07, + "loss": 0.9689, + "step": 8043 + }, + { + "epoch": 0.7254362627947873, + "grad_norm": 1.822816493597157, + "learning_rate": 7.398842302735636e-07, + "loss": 1.0045, + "step": 8044 + }, + { + "epoch": 0.7255264463182577, + "grad_norm": 0.6672757217720607, + "learning_rate": 7.394306163452171e-07, + "loss": 0.7827, + "step": 8045 + }, + { + "epoch": 0.7256166298417279, + "grad_norm": 1.3218747438831848, + "learning_rate": 7.38977109975601e-07, + "loss": 0.9541, + "step": 8046 + }, + { + "epoch": 0.7257068133651982, + "grad_norm": 1.7858829813701633, + "learning_rate": 7.385237112034119e-07, + "loss": 0.8945, + "step": 8047 + }, + { + "epoch": 0.7257969968886684, + "grad_norm": 2.685979102291296, + "learning_rate": 7.380704200673342e-07, + "loss": 1.0413, + "step": 8048 + }, + { + "epoch": 0.7258871804121387, + "grad_norm": 1.4162013280708499, + "learning_rate": 7.376172366060478e-07, + "loss": 0.9993, + "step": 8049 + }, + { + "epoch": 0.725977363935609, + "grad_norm": 0.6495873241615986, + "learning_rate": 7.371641608582187e-07, + "loss": 0.7482, + "step": 8050 + }, + { + "epoch": 0.7260675474590792, + "grad_norm": 2.2252407500929023, + "learning_rate": 7.367111928625067e-07, + "loss": 0.8429, + "step": 8051 + }, + { + "epoch": 0.7261577309825495, + "grad_norm": 2.7354783801990425, + "learning_rate": 7.362583326575613e-07, + "loss": 0.8905, + "step": 8052 + }, + { + "epoch": 0.7262479145060198, + "grad_norm": 1.9419270666906059, + "learning_rate": 7.358055802820234e-07, + "loss": 0.9615, + "step": 8053 + }, + { + "epoch": 0.72633809802949, + "grad_norm": 1.9047742039987297, + "learning_rate": 7.353529357745245e-07, + "loss": 0.9477, + "step": 8054 + }, + { + "epoch": 0.7264282815529602, + "grad_norm": 2.0453279195951777, + "learning_rate": 7.349003991736851e-07, + "loss": 1.0152, + "step": 8055 + }, + { + "epoch": 0.7265184650764306, + "grad_norm": 1.8493347286314445, + "learning_rate": 7.344479705181206e-07, + "loss": 0.8276, + "step": 8056 + }, + { + "epoch": 0.7266086485999008, + "grad_norm": 1.8549036711489746, + "learning_rate": 7.339956498464322e-07, + "loss": 0.9264, + "step": 8057 + }, + { + "epoch": 0.726698832123371, + "grad_norm": 1.837953345498433, + "learning_rate": 7.335434371972169e-07, + "loss": 1.0235, + "step": 8058 + }, + { + "epoch": 0.7267890156468413, + "grad_norm": 1.673990284065481, + "learning_rate": 7.33091332609058e-07, + "loss": 0.9635, + "step": 8059 + }, + { + "epoch": 0.7268791991703116, + "grad_norm": 1.7357570193848066, + "learning_rate": 7.326393361205323e-07, + "loss": 0.9926, + "step": 8060 + }, + { + "epoch": 0.7269693826937819, + "grad_norm": 1.7427923409901958, + "learning_rate": 7.321874477702068e-07, + "loss": 1.0541, + "step": 8061 + }, + { + "epoch": 0.7270595662172521, + "grad_norm": 0.7428473901959959, + "learning_rate": 7.317356675966386e-07, + "loss": 0.8288, + "step": 8062 + }, + { + "epoch": 0.7271497497407223, + "grad_norm": 2.0825965214190223, + "learning_rate": 7.312839956383765e-07, + "loss": 1.0119, + "step": 8063 + }, + { + "epoch": 0.7272399332641927, + "grad_norm": 2.226276151748748, + "learning_rate": 7.308324319339603e-07, + "loss": 0.9555, + "step": 8064 + }, + { + "epoch": 0.7273301167876629, + "grad_norm": 1.6774916378898672, + "learning_rate": 7.303809765219182e-07, + "loss": 0.9777, + "step": 8065 + }, + { + "epoch": 0.7274203003111331, + "grad_norm": 1.8419870256180324, + "learning_rate": 7.299296294407719e-07, + "loss": 1.0217, + "step": 8066 + }, + { + "epoch": 0.7275104838346034, + "grad_norm": 1.6691317289558598, + "learning_rate": 7.294783907290327e-07, + "loss": 1.018, + "step": 8067 + }, + { + "epoch": 0.7276006673580737, + "grad_norm": 1.6032846263511267, + "learning_rate": 7.290272604252028e-07, + "loss": 0.8451, + "step": 8068 + }, + { + "epoch": 0.727690850881544, + "grad_norm": 1.8650164891350112, + "learning_rate": 7.285762385677758e-07, + "loss": 0.8782, + "step": 8069 + }, + { + "epoch": 0.7277810344050142, + "grad_norm": 1.5655378945612435, + "learning_rate": 7.281253251952335e-07, + "loss": 1.0127, + "step": 8070 + }, + { + "epoch": 0.7278712179284844, + "grad_norm": 1.4428325925489276, + "learning_rate": 7.276745203460526e-07, + "loss": 0.9554, + "step": 8071 + }, + { + "epoch": 0.7279614014519548, + "grad_norm": 0.723324317399835, + "learning_rate": 7.272238240586959e-07, + "loss": 0.8688, + "step": 8072 + }, + { + "epoch": 0.728051584975425, + "grad_norm": 1.3205442507348262, + "learning_rate": 7.267732363716219e-07, + "loss": 0.9122, + "step": 8073 + }, + { + "epoch": 0.7281417684988952, + "grad_norm": 1.5957262838263606, + "learning_rate": 7.263227573232753e-07, + "loss": 0.9313, + "step": 8074 + }, + { + "epoch": 0.7282319520223655, + "grad_norm": 1.6017708991699575, + "learning_rate": 7.258723869520937e-07, + "loss": 1.0642, + "step": 8075 + }, + { + "epoch": 0.7283221355458358, + "grad_norm": 2.072486168202074, + "learning_rate": 7.254221252965059e-07, + "loss": 1.0224, + "step": 8076 + }, + { + "epoch": 0.728412319069306, + "grad_norm": 1.9198474889026347, + "learning_rate": 7.249719723949301e-07, + "loss": 0.8885, + "step": 8077 + }, + { + "epoch": 0.7285025025927763, + "grad_norm": 1.4441303264156813, + "learning_rate": 7.245219282857761e-07, + "loss": 0.9982, + "step": 8078 + }, + { + "epoch": 0.7285926861162466, + "grad_norm": 1.5490024463740142, + "learning_rate": 7.240719930074442e-07, + "loss": 1.0172, + "step": 8079 + }, + { + "epoch": 0.7286828696397168, + "grad_norm": 1.7016377385987482, + "learning_rate": 7.236221665983257e-07, + "loss": 1.0069, + "step": 8080 + }, + { + "epoch": 0.7287730531631871, + "grad_norm": 0.7228221805912147, + "learning_rate": 7.231724490968012e-07, + "loss": 0.7937, + "step": 8081 + }, + { + "epoch": 0.7288632366866573, + "grad_norm": 1.8168068064728011, + "learning_rate": 7.227228405412438e-07, + "loss": 0.9467, + "step": 8082 + }, + { + "epoch": 0.7289534202101277, + "grad_norm": 2.462276711966534, + "learning_rate": 7.222733409700165e-07, + "loss": 0.9359, + "step": 8083 + }, + { + "epoch": 0.7290436037335979, + "grad_norm": 1.4490894352472572, + "learning_rate": 7.21823950421473e-07, + "loss": 0.9864, + "step": 8084 + }, + { + "epoch": 0.7291337872570681, + "grad_norm": 1.7586553290832918, + "learning_rate": 7.213746689339577e-07, + "loss": 0.9429, + "step": 8085 + }, + { + "epoch": 0.7292239707805384, + "grad_norm": 1.75881671619015, + "learning_rate": 7.20925496545807e-07, + "loss": 0.9784, + "step": 8086 + }, + { + "epoch": 0.7293141543040087, + "grad_norm": 2.142119171588753, + "learning_rate": 7.20476433295344e-07, + "loss": 0.884, + "step": 8087 + }, + { + "epoch": 0.7294043378274789, + "grad_norm": 1.8170805305098368, + "learning_rate": 7.200274792208882e-07, + "loss": 1.0184, + "step": 8088 + }, + { + "epoch": 0.7294945213509492, + "grad_norm": 1.755619970513402, + "learning_rate": 7.195786343607444e-07, + "loss": 0.8697, + "step": 8089 + }, + { + "epoch": 0.7295847048744194, + "grad_norm": 3.2873083950817574, + "learning_rate": 7.191298987532131e-07, + "loss": 0.9677, + "step": 8090 + }, + { + "epoch": 0.7296748883978897, + "grad_norm": 1.4214443351052246, + "learning_rate": 7.186812724365805e-07, + "loss": 0.9827, + "step": 8091 + }, + { + "epoch": 0.72976507192136, + "grad_norm": 1.7435492423925352, + "learning_rate": 7.182327554491272e-07, + "loss": 0.9819, + "step": 8092 + }, + { + "epoch": 0.7298552554448302, + "grad_norm": 0.6793815538077972, + "learning_rate": 7.177843478291225e-07, + "loss": 0.8238, + "step": 8093 + }, + { + "epoch": 0.7299454389683004, + "grad_norm": 2.283744806493607, + "learning_rate": 7.173360496148276e-07, + "loss": 1.0053, + "step": 8094 + }, + { + "epoch": 0.7300356224917708, + "grad_norm": 2.5044524598334212, + "learning_rate": 7.168878608444939e-07, + "loss": 0.9898, + "step": 8095 + }, + { + "epoch": 0.730125806015241, + "grad_norm": 2.2819788565362447, + "learning_rate": 7.164397815563623e-07, + "loss": 0.9799, + "step": 8096 + }, + { + "epoch": 0.7302159895387113, + "grad_norm": 2.3956981012629743, + "learning_rate": 7.159918117886661e-07, + "loss": 0.9823, + "step": 8097 + }, + { + "epoch": 0.7303061730621815, + "grad_norm": 1.786708499160137, + "learning_rate": 7.155439515796284e-07, + "loss": 0.9946, + "step": 8098 + }, + { + "epoch": 0.7303963565856518, + "grad_norm": 2.3858421201696243, + "learning_rate": 7.150962009674633e-07, + "loss": 0.9735, + "step": 8099 + }, + { + "epoch": 0.7304865401091221, + "grad_norm": 1.4328868484759285, + "learning_rate": 7.146485599903751e-07, + "loss": 1.0246, + "step": 8100 + }, + { + "epoch": 0.7305767236325923, + "grad_norm": 1.5044245631790205, + "learning_rate": 7.142010286865592e-07, + "loss": 1.0497, + "step": 8101 + }, + { + "epoch": 0.7306669071560626, + "grad_norm": 0.55081171262653, + "learning_rate": 7.137536070942012e-07, + "loss": 0.7839, + "step": 8102 + }, + { + "epoch": 0.7307570906795329, + "grad_norm": 1.5790543793923004, + "learning_rate": 7.133062952514786e-07, + "loss": 0.9736, + "step": 8103 + }, + { + "epoch": 0.7308472742030031, + "grad_norm": 0.6370321567314681, + "learning_rate": 7.128590931965562e-07, + "loss": 0.8398, + "step": 8104 + }, + { + "epoch": 0.7309374577264733, + "grad_norm": 1.5619934786905878, + "learning_rate": 7.124120009675945e-07, + "loss": 1.0421, + "step": 8105 + }, + { + "epoch": 0.7310276412499437, + "grad_norm": 1.6979761249764351, + "learning_rate": 7.119650186027399e-07, + "loss": 1.0036, + "step": 8106 + }, + { + "epoch": 0.7311178247734139, + "grad_norm": 1.7720738348163287, + "learning_rate": 7.11518146140132e-07, + "loss": 1.098, + "step": 8107 + }, + { + "epoch": 0.7312080082968841, + "grad_norm": 2.2523851468254477, + "learning_rate": 7.110713836179007e-07, + "loss": 0.9373, + "step": 8108 + }, + { + "epoch": 0.7312981918203544, + "grad_norm": 2.319449935520802, + "learning_rate": 7.106247310741659e-07, + "loss": 0.9044, + "step": 8109 + }, + { + "epoch": 0.7313883753438247, + "grad_norm": 1.538826086795656, + "learning_rate": 7.101781885470393e-07, + "loss": 1.045, + "step": 8110 + }, + { + "epoch": 0.731478558867295, + "grad_norm": 1.8798872514152138, + "learning_rate": 7.097317560746203e-07, + "loss": 1.0127, + "step": 8111 + }, + { + "epoch": 0.7315687423907652, + "grad_norm": 1.8741691019877207, + "learning_rate": 7.092854336950036e-07, + "loss": 1.04, + "step": 8112 + }, + { + "epoch": 0.7316589259142354, + "grad_norm": 2.352331194278753, + "learning_rate": 7.0883922144627e-07, + "loss": 1.0302, + "step": 8113 + }, + { + "epoch": 0.7317491094377058, + "grad_norm": 1.5470949267093328, + "learning_rate": 7.083931193664934e-07, + "loss": 1.0156, + "step": 8114 + }, + { + "epoch": 0.731839292961176, + "grad_norm": 1.8876514222498806, + "learning_rate": 7.079471274937378e-07, + "loss": 1.0251, + "step": 8115 + }, + { + "epoch": 0.7319294764846462, + "grad_norm": 1.9882783721123574, + "learning_rate": 7.075012458660574e-07, + "loss": 0.9683, + "step": 8116 + }, + { + "epoch": 0.7320196600081165, + "grad_norm": 0.6687451585255714, + "learning_rate": 7.070554745214976e-07, + "loss": 0.8125, + "step": 8117 + }, + { + "epoch": 0.7321098435315868, + "grad_norm": 1.8599053916660617, + "learning_rate": 7.066098134980947e-07, + "loss": 0.9359, + "step": 8118 + }, + { + "epoch": 0.732200027055057, + "grad_norm": 1.634128460058942, + "learning_rate": 7.061642628338727e-07, + "loss": 0.9829, + "step": 8119 + }, + { + "epoch": 0.7322902105785273, + "grad_norm": 1.989062442464895, + "learning_rate": 7.057188225668513e-07, + "loss": 0.978, + "step": 8120 + }, + { + "epoch": 0.7323803941019975, + "grad_norm": 1.8061047558317558, + "learning_rate": 7.052734927350358e-07, + "loss": 1.128, + "step": 8121 + }, + { + "epoch": 0.7324705776254679, + "grad_norm": 1.7849201071247864, + "learning_rate": 7.048282733764252e-07, + "loss": 1.0042, + "step": 8122 + }, + { + "epoch": 0.7325607611489381, + "grad_norm": 1.5029106510733974, + "learning_rate": 7.043831645290077e-07, + "loss": 1.059, + "step": 8123 + }, + { + "epoch": 0.7326509446724083, + "grad_norm": 1.6611277172586834, + "learning_rate": 7.039381662307624e-07, + "loss": 0.9497, + "step": 8124 + }, + { + "epoch": 0.7327411281958787, + "grad_norm": 1.8478745055379608, + "learning_rate": 7.034932785196601e-07, + "loss": 0.9584, + "step": 8125 + }, + { + "epoch": 0.7328313117193489, + "grad_norm": 2.6906979186559066, + "learning_rate": 7.030485014336585e-07, + "loss": 0.963, + "step": 8126 + }, + { + "epoch": 0.7329214952428191, + "grad_norm": 1.6346521554668927, + "learning_rate": 7.026038350107118e-07, + "loss": 0.9407, + "step": 8127 + }, + { + "epoch": 0.7330116787662894, + "grad_norm": 1.6606155298757672, + "learning_rate": 7.021592792887579e-07, + "loss": 1.1089, + "step": 8128 + }, + { + "epoch": 0.7331018622897597, + "grad_norm": 1.6832968323907966, + "learning_rate": 7.01714834305732e-07, + "loss": 1.0292, + "step": 8129 + }, + { + "epoch": 0.7331920458132299, + "grad_norm": 2.4026665695513594, + "learning_rate": 7.012705000995544e-07, + "loss": 0.9391, + "step": 8130 + }, + { + "epoch": 0.7332822293367002, + "grad_norm": 1.3913044930314165, + "learning_rate": 7.008262767081392e-07, + "loss": 0.9728, + "step": 8131 + }, + { + "epoch": 0.7333724128601704, + "grad_norm": 2.2979355005188196, + "learning_rate": 7.003821641693892e-07, + "loss": 1.0307, + "step": 8132 + }, + { + "epoch": 0.7334625963836408, + "grad_norm": 1.5232053139619937, + "learning_rate": 6.999381625211993e-07, + "loss": 0.968, + "step": 8133 + }, + { + "epoch": 0.733552779907111, + "grad_norm": 1.7272493541085232, + "learning_rate": 6.994942718014536e-07, + "loss": 0.8976, + "step": 8134 + }, + { + "epoch": 0.7336429634305812, + "grad_norm": 2.091724155846758, + "learning_rate": 6.990504920480282e-07, + "loss": 0.8911, + "step": 8135 + }, + { + "epoch": 0.7337331469540515, + "grad_norm": 1.8750853836963621, + "learning_rate": 6.986068232987879e-07, + "loss": 0.976, + "step": 8136 + }, + { + "epoch": 0.7338233304775218, + "grad_norm": 1.4051194202674404, + "learning_rate": 6.981632655915888e-07, + "loss": 0.9315, + "step": 8137 + }, + { + "epoch": 0.733913514000992, + "grad_norm": 1.7166759808605554, + "learning_rate": 6.977198189642783e-07, + "loss": 1.0423, + "step": 8138 + }, + { + "epoch": 0.7340036975244623, + "grad_norm": 2.2692854152301702, + "learning_rate": 6.972764834546935e-07, + "loss": 1.0051, + "step": 8139 + }, + { + "epoch": 0.7340938810479325, + "grad_norm": 1.7657599608566914, + "learning_rate": 6.96833259100663e-07, + "loss": 0.9988, + "step": 8140 + }, + { + "epoch": 0.7341840645714028, + "grad_norm": 4.711006126093743, + "learning_rate": 6.96390145940003e-07, + "loss": 0.9513, + "step": 8141 + }, + { + "epoch": 0.7342742480948731, + "grad_norm": 1.7548597161286186, + "learning_rate": 6.959471440105253e-07, + "loss": 0.9293, + "step": 8142 + }, + { + "epoch": 0.7343644316183433, + "grad_norm": 6.317228755585049, + "learning_rate": 6.955042533500261e-07, + "loss": 1.0109, + "step": 8143 + }, + { + "epoch": 0.7344546151418135, + "grad_norm": 2.019031573520594, + "learning_rate": 6.950614739962986e-07, + "loss": 0.9413, + "step": 8144 + }, + { + "epoch": 0.7345447986652839, + "grad_norm": 1.4665918117985974, + "learning_rate": 6.946188059871198e-07, + "loss": 0.87, + "step": 8145 + }, + { + "epoch": 0.7346349821887541, + "grad_norm": 1.705091338745513, + "learning_rate": 6.941762493602638e-07, + "loss": 1.0977, + "step": 8146 + }, + { + "epoch": 0.7347251657122243, + "grad_norm": 1.4522758381824221, + "learning_rate": 6.937338041534899e-07, + "loss": 0.9216, + "step": 8147 + }, + { + "epoch": 0.7348153492356947, + "grad_norm": 1.4474646962922202, + "learning_rate": 6.932914704045505e-07, + "loss": 0.9775, + "step": 8148 + }, + { + "epoch": 0.7349055327591649, + "grad_norm": 1.8290351493090111, + "learning_rate": 6.928492481511878e-07, + "loss": 1.0084, + "step": 8149 + }, + { + "epoch": 0.7349957162826352, + "grad_norm": 2.0810871094890513, + "learning_rate": 6.924071374311349e-07, + "loss": 0.9305, + "step": 8150 + }, + { + "epoch": 0.7350858998061054, + "grad_norm": 0.6002768861134501, + "learning_rate": 6.919651382821157e-07, + "loss": 0.7953, + "step": 8151 + }, + { + "epoch": 0.7351760833295757, + "grad_norm": 1.7744956199869255, + "learning_rate": 6.915232507418425e-07, + "loss": 0.9506, + "step": 8152 + }, + { + "epoch": 0.735266266853046, + "grad_norm": 1.651469114148173, + "learning_rate": 6.910814748480204e-07, + "loss": 0.8645, + "step": 8153 + }, + { + "epoch": 0.7353564503765162, + "grad_norm": 1.7757490780320608, + "learning_rate": 6.906398106383445e-07, + "loss": 0.9969, + "step": 8154 + }, + { + "epoch": 0.7354466338999864, + "grad_norm": 1.4894744015146195, + "learning_rate": 6.901982581504994e-07, + "loss": 0.9751, + "step": 8155 + }, + { + "epoch": 0.7355368174234568, + "grad_norm": 2.319070811326812, + "learning_rate": 6.897568174221611e-07, + "loss": 0.8526, + "step": 8156 + }, + { + "epoch": 0.735627000946927, + "grad_norm": 3.105177456611512, + "learning_rate": 6.893154884909966e-07, + "loss": 1.0507, + "step": 8157 + }, + { + "epoch": 0.7357171844703972, + "grad_norm": 0.6039620140553912, + "learning_rate": 6.888742713946602e-07, + "loss": 0.8197, + "step": 8158 + }, + { + "epoch": 0.7358073679938675, + "grad_norm": 1.9546803499452374, + "learning_rate": 6.884331661708018e-07, + "loss": 0.9792, + "step": 8159 + }, + { + "epoch": 0.7358975515173378, + "grad_norm": 1.6063676902314372, + "learning_rate": 6.879921728570561e-07, + "loss": 0.9782, + "step": 8160 + }, + { + "epoch": 0.735987735040808, + "grad_norm": 1.6327156877792168, + "learning_rate": 6.875512914910539e-07, + "loss": 1.0577, + "step": 8161 + }, + { + "epoch": 0.7360779185642783, + "grad_norm": 1.445589523572806, + "learning_rate": 6.871105221104119e-07, + "loss": 0.9291, + "step": 8162 + }, + { + "epoch": 0.7361681020877485, + "grad_norm": 1.8832862226394762, + "learning_rate": 6.866698647527391e-07, + "loss": 0.9546, + "step": 8163 + }, + { + "epoch": 0.7362582856112189, + "grad_norm": 1.7119386435102388, + "learning_rate": 6.862293194556353e-07, + "loss": 1.0036, + "step": 8164 + }, + { + "epoch": 0.7363484691346891, + "grad_norm": 1.6609958096565611, + "learning_rate": 6.857888862566896e-07, + "loss": 0.9217, + "step": 8165 + }, + { + "epoch": 0.7364386526581593, + "grad_norm": 1.5216767920289165, + "learning_rate": 6.853485651934836e-07, + "loss": 0.9436, + "step": 8166 + }, + { + "epoch": 0.7365288361816296, + "grad_norm": 3.8299290979660157, + "learning_rate": 6.849083563035855e-07, + "loss": 0.8383, + "step": 8167 + }, + { + "epoch": 0.7366190197050999, + "grad_norm": 0.7086515880393859, + "learning_rate": 6.844682596245592e-07, + "loss": 0.8975, + "step": 8168 + }, + { + "epoch": 0.7367092032285701, + "grad_norm": 1.773433197957337, + "learning_rate": 6.840282751939539e-07, + "loss": 1.0382, + "step": 8169 + }, + { + "epoch": 0.7367993867520404, + "grad_norm": 1.5796554245225092, + "learning_rate": 6.835884030493126e-07, + "loss": 1.022, + "step": 8170 + }, + { + "epoch": 0.7368895702755107, + "grad_norm": 1.9097134129484814, + "learning_rate": 6.831486432281672e-07, + "loss": 0.9997, + "step": 8171 + }, + { + "epoch": 0.736979753798981, + "grad_norm": 1.83760546945996, + "learning_rate": 6.827089957680407e-07, + "loss": 0.9885, + "step": 8172 + }, + { + "epoch": 0.7370699373224512, + "grad_norm": 1.6474649896215727, + "learning_rate": 6.822694607064461e-07, + "loss": 1.0896, + "step": 8173 + }, + { + "epoch": 0.7371601208459214, + "grad_norm": 1.489895803405353, + "learning_rate": 6.818300380808877e-07, + "loss": 1.0039, + "step": 8174 + }, + { + "epoch": 0.7372503043693918, + "grad_norm": 1.596469429007755, + "learning_rate": 6.813907279288574e-07, + "loss": 0.9694, + "step": 8175 + }, + { + "epoch": 0.737340487892862, + "grad_norm": 1.8760897966134977, + "learning_rate": 6.809515302878422e-07, + "loss": 1.0059, + "step": 8176 + }, + { + "epoch": 0.7374306714163322, + "grad_norm": 1.8501115095194263, + "learning_rate": 6.80512445195315e-07, + "loss": 0.969, + "step": 8177 + }, + { + "epoch": 0.7375208549398025, + "grad_norm": 1.6381795186704158, + "learning_rate": 6.800734726887416e-07, + "loss": 0.9979, + "step": 8178 + }, + { + "epoch": 0.7376110384632728, + "grad_norm": 1.707479946080149, + "learning_rate": 6.796346128055775e-07, + "loss": 0.9812, + "step": 8179 + }, + { + "epoch": 0.737701221986743, + "grad_norm": 2.0599572662439436, + "learning_rate": 6.791958655832684e-07, + "loss": 0.9545, + "step": 8180 + }, + { + "epoch": 0.7377914055102133, + "grad_norm": 1.6633803711217718, + "learning_rate": 6.787572310592518e-07, + "loss": 0.9451, + "step": 8181 + }, + { + "epoch": 0.7378815890336835, + "grad_norm": 2.010325006427997, + "learning_rate": 6.783187092709521e-07, + "loss": 1.0615, + "step": 8182 + }, + { + "epoch": 0.7379717725571538, + "grad_norm": 1.623052089788366, + "learning_rate": 6.778803002557891e-07, + "loss": 0.9877, + "step": 8183 + }, + { + "epoch": 0.7380619560806241, + "grad_norm": 1.8433055665314513, + "learning_rate": 6.774420040511686e-07, + "loss": 1.0163, + "step": 8184 + }, + { + "epoch": 0.7381521396040943, + "grad_norm": 2.650286633806428, + "learning_rate": 6.770038206944886e-07, + "loss": 0.9808, + "step": 8185 + }, + { + "epoch": 0.7382423231275645, + "grad_norm": 1.2771470504638869, + "learning_rate": 6.765657502231375e-07, + "loss": 0.9941, + "step": 8186 + }, + { + "epoch": 0.7383325066510349, + "grad_norm": 2.1836540200202026, + "learning_rate": 6.761277926744939e-07, + "loss": 1.049, + "step": 8187 + }, + { + "epoch": 0.7384226901745051, + "grad_norm": 1.530717406779888, + "learning_rate": 6.756899480859268e-07, + "loss": 0.9719, + "step": 8188 + }, + { + "epoch": 0.7385128736979754, + "grad_norm": 3.702407316687302, + "learning_rate": 6.752522164947956e-07, + "loss": 1.0245, + "step": 8189 + }, + { + "epoch": 0.7386030572214456, + "grad_norm": 1.2549872566427247, + "learning_rate": 6.748145979384498e-07, + "loss": 1.0349, + "step": 8190 + }, + { + "epoch": 0.7386932407449159, + "grad_norm": 4.313525091878071, + "learning_rate": 6.743770924542303e-07, + "loss": 0.923, + "step": 8191 + }, + { + "epoch": 0.7387834242683862, + "grad_norm": 0.634262026043148, + "learning_rate": 6.739397000794658e-07, + "loss": 0.8066, + "step": 8192 + }, + { + "epoch": 0.7388736077918564, + "grad_norm": 1.7183991507583167, + "learning_rate": 6.735024208514782e-07, + "loss": 0.9529, + "step": 8193 + }, + { + "epoch": 0.7389637913153266, + "grad_norm": 1.463290354735506, + "learning_rate": 6.73065254807578e-07, + "loss": 0.8958, + "step": 8194 + }, + { + "epoch": 0.739053974838797, + "grad_norm": 2.4359719793339867, + "learning_rate": 6.726282019850669e-07, + "loss": 1.0346, + "step": 8195 + }, + { + "epoch": 0.7391441583622672, + "grad_norm": 1.3430157806878196, + "learning_rate": 6.721912624212376e-07, + "loss": 1.0472, + "step": 8196 + }, + { + "epoch": 0.7392343418857374, + "grad_norm": 3.44812270794492, + "learning_rate": 6.717544361533696e-07, + "loss": 0.8816, + "step": 8197 + }, + { + "epoch": 0.7393245254092078, + "grad_norm": 1.6613841105018954, + "learning_rate": 6.713177232187386e-07, + "loss": 0.9622, + "step": 8198 + }, + { + "epoch": 0.739414708932678, + "grad_norm": 2.7900423951841753, + "learning_rate": 6.708811236546041e-07, + "loss": 0.9711, + "step": 8199 + }, + { + "epoch": 0.7395048924561483, + "grad_norm": 2.006167439220728, + "learning_rate": 6.704446374982224e-07, + "loss": 0.9448, + "step": 8200 + }, + { + "epoch": 0.7395950759796185, + "grad_norm": 6.516522755054161, + "learning_rate": 6.700082647868346e-07, + "loss": 1.0164, + "step": 8201 + }, + { + "epoch": 0.7396852595030888, + "grad_norm": 1.5291142263264477, + "learning_rate": 6.695720055576751e-07, + "loss": 1.0325, + "step": 8202 + }, + { + "epoch": 0.7397754430265591, + "grad_norm": 1.7444462524426292, + "learning_rate": 6.691358598479679e-07, + "loss": 0.9902, + "step": 8203 + }, + { + "epoch": 0.7398656265500293, + "grad_norm": 1.4439652220587789, + "learning_rate": 6.686998276949276e-07, + "loss": 0.9741, + "step": 8204 + }, + { + "epoch": 0.7399558100734995, + "grad_norm": 0.6323253499024631, + "learning_rate": 6.682639091357587e-07, + "loss": 0.8222, + "step": 8205 + }, + { + "epoch": 0.7400459935969699, + "grad_norm": 2.62837937220582, + "learning_rate": 6.678281042076568e-07, + "loss": 0.9701, + "step": 8206 + }, + { + "epoch": 0.7401361771204401, + "grad_norm": 1.3118330305345745, + "learning_rate": 6.673924129478059e-07, + "loss": 0.983, + "step": 8207 + }, + { + "epoch": 0.7402263606439103, + "grad_norm": 1.4947882551448939, + "learning_rate": 6.669568353933824e-07, + "loss": 0.9017, + "step": 8208 + }, + { + "epoch": 0.7403165441673806, + "grad_norm": 1.5470466903811397, + "learning_rate": 6.665213715815519e-07, + "loss": 0.9161, + "step": 8209 + }, + { + "epoch": 0.7404067276908509, + "grad_norm": 2.3690798406516054, + "learning_rate": 6.660860215494706e-07, + "loss": 0.9954, + "step": 8210 + }, + { + "epoch": 0.7404969112143212, + "grad_norm": 1.7552845727981818, + "learning_rate": 6.656507853342852e-07, + "loss": 0.932, + "step": 8211 + }, + { + "epoch": 0.7405870947377914, + "grad_norm": 1.4938838563582653, + "learning_rate": 6.652156629731323e-07, + "loss": 0.9895, + "step": 8212 + }, + { + "epoch": 0.7406772782612616, + "grad_norm": 1.9708060170620083, + "learning_rate": 6.647806545031396e-07, + "loss": 1.0438, + "step": 8213 + }, + { + "epoch": 0.740767461784732, + "grad_norm": 1.7427441173952705, + "learning_rate": 6.643457599614224e-07, + "loss": 0.9034, + "step": 8214 + }, + { + "epoch": 0.7408576453082022, + "grad_norm": 1.8096682032598939, + "learning_rate": 6.63910979385091e-07, + "loss": 1.0354, + "step": 8215 + }, + { + "epoch": 0.7409478288316724, + "grad_norm": 2.1044660172749263, + "learning_rate": 6.634763128112409e-07, + "loss": 1.0653, + "step": 8216 + }, + { + "epoch": 0.7410380123551427, + "grad_norm": 2.3716636111181026, + "learning_rate": 6.630417602769622e-07, + "loss": 0.9729, + "step": 8217 + }, + { + "epoch": 0.741128195878613, + "grad_norm": 1.5072599989862623, + "learning_rate": 6.62607321819332e-07, + "loss": 0.9868, + "step": 8218 + }, + { + "epoch": 0.7412183794020832, + "grad_norm": 1.3630201641698683, + "learning_rate": 6.621729974754196e-07, + "loss": 0.9784, + "step": 8219 + }, + { + "epoch": 0.7413085629255535, + "grad_norm": 1.940941799185467, + "learning_rate": 6.617387872822835e-07, + "loss": 0.9298, + "step": 8220 + }, + { + "epoch": 0.7413987464490238, + "grad_norm": 1.7584505406708475, + "learning_rate": 6.613046912769731e-07, + "loss": 0.9498, + "step": 8221 + }, + { + "epoch": 0.741488929972494, + "grad_norm": 2.174267336336454, + "learning_rate": 6.608707094965289e-07, + "loss": 1.0463, + "step": 8222 + }, + { + "epoch": 0.7415791134959643, + "grad_norm": 3.8926365566126826, + "learning_rate": 6.604368419779787e-07, + "loss": 0.9163, + "step": 8223 + }, + { + "epoch": 0.7416692970194345, + "grad_norm": 2.307603136198227, + "learning_rate": 6.600030887583434e-07, + "loss": 0.993, + "step": 8224 + }, + { + "epoch": 0.7417594805429049, + "grad_norm": 1.6748562195971457, + "learning_rate": 6.595694498746336e-07, + "loss": 1.0875, + "step": 8225 + }, + { + "epoch": 0.7418496640663751, + "grad_norm": 2.0015511458087527, + "learning_rate": 6.591359253638491e-07, + "loss": 0.9526, + "step": 8226 + }, + { + "epoch": 0.7419398475898453, + "grad_norm": 2.905338923894304, + "learning_rate": 6.587025152629808e-07, + "loss": 0.8713, + "step": 8227 + }, + { + "epoch": 0.7420300311133156, + "grad_norm": 2.227531463195334, + "learning_rate": 6.582692196090107e-07, + "loss": 1.0189, + "step": 8228 + }, + { + "epoch": 0.7421202146367859, + "grad_norm": 1.5886218274845962, + "learning_rate": 6.578360384389074e-07, + "loss": 0.9976, + "step": 8229 + }, + { + "epoch": 0.7422103981602561, + "grad_norm": 2.1941838196884325, + "learning_rate": 6.574029717896355e-07, + "loss": 0.9595, + "step": 8230 + }, + { + "epoch": 0.7423005816837264, + "grad_norm": 1.6784426312180833, + "learning_rate": 6.569700196981436e-07, + "loss": 0.9721, + "step": 8231 + }, + { + "epoch": 0.7423907652071966, + "grad_norm": 2.116619076031112, + "learning_rate": 6.565371822013763e-07, + "loss": 0.8883, + "step": 8232 + }, + { + "epoch": 0.742480948730667, + "grad_norm": 1.8889169464957074, + "learning_rate": 6.561044593362636e-07, + "loss": 1.0354, + "step": 8233 + }, + { + "epoch": 0.7425711322541372, + "grad_norm": 1.4147427434902002, + "learning_rate": 6.556718511397288e-07, + "loss": 0.96, + "step": 8234 + }, + { + "epoch": 0.7426613157776074, + "grad_norm": 1.5559730773646183, + "learning_rate": 6.552393576486843e-07, + "loss": 1.0958, + "step": 8235 + }, + { + "epoch": 0.7427514993010776, + "grad_norm": 1.7949989808698645, + "learning_rate": 6.548069789000325e-07, + "loss": 0.9942, + "step": 8236 + }, + { + "epoch": 0.742841682824548, + "grad_norm": 1.3131375353876196, + "learning_rate": 6.543747149306673e-07, + "loss": 0.9667, + "step": 8237 + }, + { + "epoch": 0.7429318663480182, + "grad_norm": 1.6472937787992323, + "learning_rate": 6.5394256577747e-07, + "loss": 1.0298, + "step": 8238 + }, + { + "epoch": 0.7430220498714885, + "grad_norm": 1.560226923259238, + "learning_rate": 6.535105314773161e-07, + "loss": 1.0109, + "step": 8239 + }, + { + "epoch": 0.7431122333949587, + "grad_norm": 2.1598300791061256, + "learning_rate": 6.530786120670677e-07, + "loss": 0.9124, + "step": 8240 + }, + { + "epoch": 0.743202416918429, + "grad_norm": 1.7100552587894355, + "learning_rate": 6.526468075835787e-07, + "loss": 1.0393, + "step": 8241 + }, + { + "epoch": 0.7432926004418993, + "grad_norm": 1.498069633698944, + "learning_rate": 6.522151180636937e-07, + "loss": 1.0259, + "step": 8242 + }, + { + "epoch": 0.7433827839653695, + "grad_norm": 1.546894420155084, + "learning_rate": 6.517835435442461e-07, + "loss": 0.9488, + "step": 8243 + }, + { + "epoch": 0.7434729674888398, + "grad_norm": 2.1215546670955368, + "learning_rate": 6.513520840620606e-07, + "loss": 1.001, + "step": 8244 + }, + { + "epoch": 0.7435631510123101, + "grad_norm": 2.0019569359719824, + "learning_rate": 6.509207396539525e-07, + "loss": 0.9435, + "step": 8245 + }, + { + "epoch": 0.7436533345357803, + "grad_norm": 0.62728499901086, + "learning_rate": 6.50489510356724e-07, + "loss": 0.7888, + "step": 8246 + }, + { + "epoch": 0.7437435180592505, + "grad_norm": 2.0416342933838916, + "learning_rate": 6.500583962071732e-07, + "loss": 0.9101, + "step": 8247 + }, + { + "epoch": 0.7438337015827209, + "grad_norm": 1.6677854915245385, + "learning_rate": 6.496273972420827e-07, + "loss": 1.0061, + "step": 8248 + }, + { + "epoch": 0.7439238851061911, + "grad_norm": 2.015118794122334, + "learning_rate": 6.491965134982287e-07, + "loss": 1.0303, + "step": 8249 + }, + { + "epoch": 0.7440140686296614, + "grad_norm": 2.307173497006258, + "learning_rate": 6.487657450123765e-07, + "loss": 0.9501, + "step": 8250 + }, + { + "epoch": 0.7441042521531316, + "grad_norm": 1.535189174464047, + "learning_rate": 6.483350918212814e-07, + "loss": 0.995, + "step": 8251 + }, + { + "epoch": 0.7441944356766019, + "grad_norm": 0.7018294372444412, + "learning_rate": 6.479045539616898e-07, + "loss": 0.8212, + "step": 8252 + }, + { + "epoch": 0.7442846192000722, + "grad_norm": 2.347268375592549, + "learning_rate": 6.474741314703358e-07, + "loss": 0.9868, + "step": 8253 + }, + { + "epoch": 0.7443748027235424, + "grad_norm": 2.450577986912513, + "learning_rate": 6.47043824383948e-07, + "loss": 1.0391, + "step": 8254 + }, + { + "epoch": 0.7444649862470126, + "grad_norm": 2.322472077890666, + "learning_rate": 6.466136327392399e-07, + "loss": 1.0225, + "step": 8255 + }, + { + "epoch": 0.744555169770483, + "grad_norm": 1.9216855583795243, + "learning_rate": 6.461835565729206e-07, + "loss": 1.0181, + "step": 8256 + }, + { + "epoch": 0.7446453532939532, + "grad_norm": 5.007356191427971, + "learning_rate": 6.457535959216844e-07, + "loss": 1.012, + "step": 8257 + }, + { + "epoch": 0.7447355368174234, + "grad_norm": 2.2639019958595137, + "learning_rate": 6.453237508222186e-07, + "loss": 0.9415, + "step": 8258 + }, + { + "epoch": 0.7448257203408937, + "grad_norm": 1.5510800936080347, + "learning_rate": 6.448940213112e-07, + "loss": 1.0038, + "step": 8259 + }, + { + "epoch": 0.744915903864364, + "grad_norm": 1.6738969928902456, + "learning_rate": 6.444644074252954e-07, + "loss": 1.068, + "step": 8260 + }, + { + "epoch": 0.7450060873878342, + "grad_norm": 1.910601979305105, + "learning_rate": 6.440349092011628e-07, + "loss": 1.0531, + "step": 8261 + }, + { + "epoch": 0.7450962709113045, + "grad_norm": 1.5020195558140959, + "learning_rate": 6.436055266754475e-07, + "loss": 0.9741, + "step": 8262 + }, + { + "epoch": 0.7451864544347747, + "grad_norm": 1.8284586414502961, + "learning_rate": 6.431762598847879e-07, + "loss": 0.9554, + "step": 8263 + }, + { + "epoch": 0.7452766379582451, + "grad_norm": 1.7420618670543877, + "learning_rate": 6.427471088658111e-07, + "loss": 0.9551, + "step": 8264 + }, + { + "epoch": 0.7453668214817153, + "grad_norm": 2.350010153565893, + "learning_rate": 6.42318073655135e-07, + "loss": 0.9777, + "step": 8265 + }, + { + "epoch": 0.7454570050051855, + "grad_norm": 1.9771222916339923, + "learning_rate": 6.41889154289367e-07, + "loss": 0.9474, + "step": 8266 + }, + { + "epoch": 0.7455471885286559, + "grad_norm": 1.6466636333875222, + "learning_rate": 6.414603508051055e-07, + "loss": 1.0394, + "step": 8267 + }, + { + "epoch": 0.7456373720521261, + "grad_norm": 0.6343323622783151, + "learning_rate": 6.410316632389365e-07, + "loss": 0.8413, + "step": 8268 + }, + { + "epoch": 0.7457275555755963, + "grad_norm": 2.1948428483115094, + "learning_rate": 6.406030916274406e-07, + "loss": 0.9931, + "step": 8269 + }, + { + "epoch": 0.7458177390990666, + "grad_norm": 2.568578256773469, + "learning_rate": 6.401746360071831e-07, + "loss": 0.9411, + "step": 8270 + }, + { + "epoch": 0.7459079226225369, + "grad_norm": 2.288263070661719, + "learning_rate": 6.397462964147251e-07, + "loss": 0.945, + "step": 8271 + }, + { + "epoch": 0.7459981061460071, + "grad_norm": 1.7203715911057602, + "learning_rate": 6.393180728866128e-07, + "loss": 0.9734, + "step": 8272 + }, + { + "epoch": 0.7460882896694774, + "grad_norm": 1.5493892697026865, + "learning_rate": 6.388899654593853e-07, + "loss": 1.0165, + "step": 8273 + }, + { + "epoch": 0.7461784731929476, + "grad_norm": 2.44354057097548, + "learning_rate": 6.384619741695709e-07, + "loss": 1.007, + "step": 8274 + }, + { + "epoch": 0.746268656716418, + "grad_norm": 1.2398425346806032, + "learning_rate": 6.380340990536883e-07, + "loss": 0.9834, + "step": 8275 + }, + { + "epoch": 0.7463588402398882, + "grad_norm": 2.8276016847116274, + "learning_rate": 6.37606340148247e-07, + "loss": 0.9881, + "step": 8276 + }, + { + "epoch": 0.7464490237633584, + "grad_norm": 3.4090297450734295, + "learning_rate": 6.371786974897433e-07, + "loss": 0.8925, + "step": 8277 + }, + { + "epoch": 0.7465392072868287, + "grad_norm": 1.318598611106008, + "learning_rate": 6.367511711146691e-07, + "loss": 1.0111, + "step": 8278 + }, + { + "epoch": 0.746629390810299, + "grad_norm": 2.977192168830229, + "learning_rate": 6.363237610595014e-07, + "loss": 0.9299, + "step": 8279 + }, + { + "epoch": 0.7467195743337692, + "grad_norm": 2.1701743513812586, + "learning_rate": 6.358964673607094e-07, + "loss": 0.9629, + "step": 8280 + }, + { + "epoch": 0.7468097578572395, + "grad_norm": 1.49767998252217, + "learning_rate": 6.354692900547525e-07, + "loss": 1.0177, + "step": 8281 + }, + { + "epoch": 0.7468999413807097, + "grad_norm": 1.6648582024665248, + "learning_rate": 6.350422291780797e-07, + "loss": 0.9583, + "step": 8282 + }, + { + "epoch": 0.74699012490418, + "grad_norm": 2.045928503284452, + "learning_rate": 6.346152847671302e-07, + "loss": 1.0862, + "step": 8283 + }, + { + "epoch": 0.7470803084276503, + "grad_norm": 1.6804384194247886, + "learning_rate": 6.34188456858334e-07, + "loss": 0.9755, + "step": 8284 + }, + { + "epoch": 0.7471704919511205, + "grad_norm": 2.5542037459187323, + "learning_rate": 6.337617454881081e-07, + "loss": 0.9353, + "step": 8285 + }, + { + "epoch": 0.7472606754745907, + "grad_norm": 1.5970252546775079, + "learning_rate": 6.333351506928651e-07, + "loss": 0.8904, + "step": 8286 + }, + { + "epoch": 0.7473508589980611, + "grad_norm": 1.676936607181746, + "learning_rate": 6.329086725090018e-07, + "loss": 0.9554, + "step": 8287 + }, + { + "epoch": 0.7474410425215313, + "grad_norm": 1.7485429965485002, + "learning_rate": 6.324823109729087e-07, + "loss": 0.9443, + "step": 8288 + }, + { + "epoch": 0.7475312260450016, + "grad_norm": 2.237151867939369, + "learning_rate": 6.320560661209653e-07, + "loss": 0.9131, + "step": 8289 + }, + { + "epoch": 0.7476214095684719, + "grad_norm": 1.5508628074281714, + "learning_rate": 6.316299379895411e-07, + "loss": 0.865, + "step": 8290 + }, + { + "epoch": 0.7477115930919421, + "grad_norm": 4.423501514226293, + "learning_rate": 6.312039266149965e-07, + "loss": 1.0202, + "step": 8291 + }, + { + "epoch": 0.7478017766154124, + "grad_norm": 1.6050892773572918, + "learning_rate": 6.307780320336789e-07, + "loss": 1.0141, + "step": 8292 + }, + { + "epoch": 0.7478919601388826, + "grad_norm": 1.985660887271722, + "learning_rate": 6.303522542819306e-07, + "loss": 1.0397, + "step": 8293 + }, + { + "epoch": 0.7479821436623529, + "grad_norm": 3.3788076044578057, + "learning_rate": 6.299265933960796e-07, + "loss": 0.9589, + "step": 8294 + }, + { + "epoch": 0.7480723271858232, + "grad_norm": 8.483202774252307, + "learning_rate": 6.295010494124462e-07, + "loss": 0.9351, + "step": 8295 + }, + { + "epoch": 0.7481625107092934, + "grad_norm": 2.426999092004431, + "learning_rate": 6.290756223673399e-07, + "loss": 0.9584, + "step": 8296 + }, + { + "epoch": 0.7482526942327636, + "grad_norm": 1.4259173602626276, + "learning_rate": 6.28650312297061e-07, + "loss": 1.0172, + "step": 8297 + }, + { + "epoch": 0.748342877756234, + "grad_norm": 2.357089723260652, + "learning_rate": 6.282251192378987e-07, + "loss": 0.8976, + "step": 8298 + }, + { + "epoch": 0.7484330612797042, + "grad_norm": 3.090346805406401, + "learning_rate": 6.278000432261334e-07, + "loss": 0.9108, + "step": 8299 + }, + { + "epoch": 0.7485232448031744, + "grad_norm": 1.774863858441942, + "learning_rate": 6.273750842980345e-07, + "loss": 1.0082, + "step": 8300 + }, + { + "epoch": 0.7486134283266447, + "grad_norm": 1.7525564322317093, + "learning_rate": 6.269502424898625e-07, + "loss": 1.0085, + "step": 8301 + }, + { + "epoch": 0.748703611850115, + "grad_norm": 2.7240731754249152, + "learning_rate": 6.265255178378663e-07, + "loss": 1.0211, + "step": 8302 + }, + { + "epoch": 0.7487937953735853, + "grad_norm": 1.8315458325674092, + "learning_rate": 6.261009103782861e-07, + "loss": 0.8785, + "step": 8303 + }, + { + "epoch": 0.7488839788970555, + "grad_norm": 1.6876747605995348, + "learning_rate": 6.256764201473519e-07, + "loss": 1.028, + "step": 8304 + }, + { + "epoch": 0.7489741624205257, + "grad_norm": 0.6394647465548721, + "learning_rate": 6.252520471812835e-07, + "loss": 0.8095, + "step": 8305 + }, + { + "epoch": 0.7490643459439961, + "grad_norm": 2.248444549358629, + "learning_rate": 6.248277915162912e-07, + "loss": 0.9957, + "step": 8306 + }, + { + "epoch": 0.7491545294674663, + "grad_norm": 1.531007319312661, + "learning_rate": 6.244036531885731e-07, + "loss": 0.8368, + "step": 8307 + }, + { + "epoch": 0.7492447129909365, + "grad_norm": 1.5031711119242654, + "learning_rate": 6.239796322343216e-07, + "loss": 0.9165, + "step": 8308 + }, + { + "epoch": 0.7493348965144068, + "grad_norm": 1.8495980315782046, + "learning_rate": 6.235557286897137e-07, + "loss": 0.9178, + "step": 8309 + }, + { + "epoch": 0.7494250800378771, + "grad_norm": 2.2641623149217387, + "learning_rate": 6.231319425909223e-07, + "loss": 0.9795, + "step": 8310 + }, + { + "epoch": 0.7495152635613473, + "grad_norm": 1.9571943081903793, + "learning_rate": 6.227082739741045e-07, + "loss": 1.0141, + "step": 8311 + }, + { + "epoch": 0.7496054470848176, + "grad_norm": 1.5913402066129616, + "learning_rate": 6.222847228754113e-07, + "loss": 1.0075, + "step": 8312 + }, + { + "epoch": 0.7496956306082878, + "grad_norm": 3.533593404543428, + "learning_rate": 6.218612893309823e-07, + "loss": 1.077, + "step": 8313 + }, + { + "epoch": 0.7497858141317582, + "grad_norm": 1.7452195814733409, + "learning_rate": 6.214379733769468e-07, + "loss": 0.9227, + "step": 8314 + }, + { + "epoch": 0.7498759976552284, + "grad_norm": 1.791646964681959, + "learning_rate": 6.21014775049425e-07, + "loss": 1.0163, + "step": 8315 + }, + { + "epoch": 0.7499661811786986, + "grad_norm": 0.5784044878920088, + "learning_rate": 6.205916943845267e-07, + "loss": 0.8015, + "step": 8316 + }, + { + "epoch": 0.750056364702169, + "grad_norm": 1.6548111621730113, + "learning_rate": 6.201687314183504e-07, + "loss": 1.0258, + "step": 8317 + }, + { + "epoch": 0.7501465482256392, + "grad_norm": 0.6685619954526933, + "learning_rate": 6.197458861869862e-07, + "loss": 0.8485, + "step": 8318 + }, + { + "epoch": 0.7502367317491094, + "grad_norm": 1.9778103715315871, + "learning_rate": 6.193231587265138e-07, + "loss": 1.0265, + "step": 8319 + }, + { + "epoch": 0.7503269152725797, + "grad_norm": 1.7525931627703712, + "learning_rate": 6.189005490730024e-07, + "loss": 0.9664, + "step": 8320 + }, + { + "epoch": 0.75041709879605, + "grad_norm": 1.4939180894072728, + "learning_rate": 6.184780572625115e-07, + "loss": 1.0027, + "step": 8321 + }, + { + "epoch": 0.7505072823195202, + "grad_norm": 1.8092965064539868, + "learning_rate": 6.180556833310902e-07, + "loss": 0.8824, + "step": 8322 + }, + { + "epoch": 0.7505974658429905, + "grad_norm": 2.8997664785087114, + "learning_rate": 6.176334273147788e-07, + "loss": 1.0189, + "step": 8323 + }, + { + "epoch": 0.7506876493664607, + "grad_norm": 2.5476722639575473, + "learning_rate": 6.172112892496042e-07, + "loss": 0.9031, + "step": 8324 + }, + { + "epoch": 0.750777832889931, + "grad_norm": 1.702082183476421, + "learning_rate": 6.167892691715883e-07, + "loss": 1.0261, + "step": 8325 + }, + { + "epoch": 0.7508680164134013, + "grad_norm": 1.6653890560719158, + "learning_rate": 6.163673671167378e-07, + "loss": 0.9608, + "step": 8326 + }, + { + "epoch": 0.7509581999368715, + "grad_norm": 5.054769948673114, + "learning_rate": 6.15945583121054e-07, + "loss": 1.0217, + "step": 8327 + }, + { + "epoch": 0.7510483834603418, + "grad_norm": 1.9730798247672505, + "learning_rate": 6.15523917220524e-07, + "loss": 0.9332, + "step": 8328 + }, + { + "epoch": 0.7511385669838121, + "grad_norm": 5.1578359303466605, + "learning_rate": 6.151023694511273e-07, + "loss": 1.051, + "step": 8329 + }, + { + "epoch": 0.7512287505072823, + "grad_norm": 2.6055694057346823, + "learning_rate": 6.146809398488328e-07, + "loss": 0.924, + "step": 8330 + }, + { + "epoch": 0.7513189340307526, + "grad_norm": 1.567354595370034, + "learning_rate": 6.142596284495989e-07, + "loss": 1.0144, + "step": 8331 + }, + { + "epoch": 0.7514091175542228, + "grad_norm": 1.7102790854626324, + "learning_rate": 6.138384352893751e-07, + "loss": 0.9399, + "step": 8332 + }, + { + "epoch": 0.7514993010776931, + "grad_norm": 1.9241821582953111, + "learning_rate": 6.134173604040987e-07, + "loss": 0.9694, + "step": 8333 + }, + { + "epoch": 0.7515894846011634, + "grad_norm": 1.6596578410850527, + "learning_rate": 6.129964038296984e-07, + "loss": 0.9822, + "step": 8334 + }, + { + "epoch": 0.7516796681246336, + "grad_norm": 1.7308467684600715, + "learning_rate": 6.12575565602093e-07, + "loss": 1.0507, + "step": 8335 + }, + { + "epoch": 0.7517698516481038, + "grad_norm": 1.307490826313812, + "learning_rate": 6.121548457571905e-07, + "loss": 1.0533, + "step": 8336 + }, + { + "epoch": 0.7518600351715742, + "grad_norm": 2.0794782319742224, + "learning_rate": 6.11734244330889e-07, + "loss": 0.9256, + "step": 8337 + }, + { + "epoch": 0.7519502186950444, + "grad_norm": 1.7205128732347668, + "learning_rate": 6.113137613590773e-07, + "loss": 0.9938, + "step": 8338 + }, + { + "epoch": 0.7520404022185146, + "grad_norm": 1.6622449234626624, + "learning_rate": 6.108933968776313e-07, + "loss": 1.0394, + "step": 8339 + }, + { + "epoch": 0.752130585741985, + "grad_norm": 3.0479158628770797, + "learning_rate": 6.104731509224212e-07, + "loss": 0.9826, + "step": 8340 + }, + { + "epoch": 0.7522207692654552, + "grad_norm": 1.9729850268445916, + "learning_rate": 6.100530235293027e-07, + "loss": 0.9147, + "step": 8341 + }, + { + "epoch": 0.7523109527889255, + "grad_norm": 1.844452158530061, + "learning_rate": 6.096330147341253e-07, + "loss": 1.0047, + "step": 8342 + }, + { + "epoch": 0.7524011363123957, + "grad_norm": 5.935685051040569, + "learning_rate": 6.09213124572725e-07, + "loss": 1.0399, + "step": 8343 + }, + { + "epoch": 0.752491319835866, + "grad_norm": 1.471379786805658, + "learning_rate": 6.087933530809297e-07, + "loss": 1.0595, + "step": 8344 + }, + { + "epoch": 0.7525815033593363, + "grad_norm": 1.9097156601610055, + "learning_rate": 6.083737002945566e-07, + "loss": 0.9691, + "step": 8345 + }, + { + "epoch": 0.7526716868828065, + "grad_norm": 1.810549047771249, + "learning_rate": 6.079541662494126e-07, + "loss": 0.9329, + "step": 8346 + }, + { + "epoch": 0.7527618704062767, + "grad_norm": 1.9273121534710236, + "learning_rate": 6.075347509812954e-07, + "loss": 0.9896, + "step": 8347 + }, + { + "epoch": 0.7528520539297471, + "grad_norm": 1.594942955384162, + "learning_rate": 6.0711545452599e-07, + "loss": 0.9406, + "step": 8348 + }, + { + "epoch": 0.7529422374532173, + "grad_norm": 1.6065872639838286, + "learning_rate": 6.066962769192756e-07, + "loss": 0.9035, + "step": 8349 + }, + { + "epoch": 0.7530324209766875, + "grad_norm": 3.52021293516269, + "learning_rate": 6.062772181969167e-07, + "loss": 0.9158, + "step": 8350 + }, + { + "epoch": 0.7531226045001578, + "grad_norm": 1.5549136050102657, + "learning_rate": 6.058582783946706e-07, + "loss": 1.0484, + "step": 8351 + }, + { + "epoch": 0.7532127880236281, + "grad_norm": 1.8979494699880848, + "learning_rate": 6.054394575482833e-07, + "loss": 0.9728, + "step": 8352 + }, + { + "epoch": 0.7533029715470984, + "grad_norm": 1.587215566309998, + "learning_rate": 6.05020755693491e-07, + "loss": 0.8993, + "step": 8353 + }, + { + "epoch": 0.7533931550705686, + "grad_norm": 1.7326363449061015, + "learning_rate": 6.046021728660198e-07, + "loss": 0.8937, + "step": 8354 + }, + { + "epoch": 0.7534833385940388, + "grad_norm": 2.0520961651344396, + "learning_rate": 6.041837091015858e-07, + "loss": 1.0178, + "step": 8355 + }, + { + "epoch": 0.7535735221175092, + "grad_norm": 1.5434571857204922, + "learning_rate": 6.037653644358931e-07, + "loss": 1.032, + "step": 8356 + }, + { + "epoch": 0.7536637056409794, + "grad_norm": 1.9996986162079682, + "learning_rate": 6.033471389046393e-07, + "loss": 0.9915, + "step": 8357 + }, + { + "epoch": 0.7537538891644496, + "grad_norm": 1.4904014723702415, + "learning_rate": 6.029290325435084e-07, + "loss": 0.9414, + "step": 8358 + }, + { + "epoch": 0.7538440726879199, + "grad_norm": 1.9252697495548579, + "learning_rate": 6.025110453881756e-07, + "loss": 1.0474, + "step": 8359 + }, + { + "epoch": 0.7539342562113902, + "grad_norm": 16.580285998409753, + "learning_rate": 6.020931774743061e-07, + "loss": 0.924, + "step": 8360 + }, + { + "epoch": 0.7540244397348604, + "grad_norm": 1.999899384828766, + "learning_rate": 6.016754288375546e-07, + "loss": 0.9075, + "step": 8361 + }, + { + "epoch": 0.7541146232583307, + "grad_norm": 1.7733340464258314, + "learning_rate": 6.012577995135665e-07, + "loss": 0.9724, + "step": 8362 + }, + { + "epoch": 0.754204806781801, + "grad_norm": 3.7489413992472906, + "learning_rate": 6.008402895379743e-07, + "loss": 0.9252, + "step": 8363 + }, + { + "epoch": 0.7542949903052713, + "grad_norm": 1.988100294710967, + "learning_rate": 6.004228989464047e-07, + "loss": 1.0621, + "step": 8364 + }, + { + "epoch": 0.7543851738287415, + "grad_norm": 6.723685221859102, + "learning_rate": 6.000056277744692e-07, + "loss": 1.0015, + "step": 8365 + }, + { + "epoch": 0.7544753573522117, + "grad_norm": 2.2690997608121646, + "learning_rate": 5.995884760577745e-07, + "loss": 0.9524, + "step": 8366 + }, + { + "epoch": 0.7545655408756821, + "grad_norm": 0.6291668985589549, + "learning_rate": 5.99171443831912e-07, + "loss": 0.8009, + "step": 8367 + }, + { + "epoch": 0.7546557243991523, + "grad_norm": 1.5450525047738457, + "learning_rate": 5.98754531132466e-07, + "loss": 0.978, + "step": 8368 + }, + { + "epoch": 0.7547459079226225, + "grad_norm": 1.9105036444989894, + "learning_rate": 5.983377379950099e-07, + "loss": 0.9853, + "step": 8369 + }, + { + "epoch": 0.7548360914460928, + "grad_norm": 2.769326443221744, + "learning_rate": 5.979210644551067e-07, + "loss": 0.9206, + "step": 8370 + }, + { + "epoch": 0.7549262749695631, + "grad_norm": 2.2913546494280554, + "learning_rate": 5.975045105483091e-07, + "loss": 0.9756, + "step": 8371 + }, + { + "epoch": 0.7550164584930333, + "grad_norm": 0.6013005417024376, + "learning_rate": 5.970880763101607e-07, + "loss": 0.8246, + "step": 8372 + }, + { + "epoch": 0.7551066420165036, + "grad_norm": 1.6295909586004382, + "learning_rate": 5.966717617761925e-07, + "loss": 1.005, + "step": 8373 + }, + { + "epoch": 0.7551968255399738, + "grad_norm": 1.6398235725081058, + "learning_rate": 5.962555669819276e-07, + "loss": 0.9627, + "step": 8374 + }, + { + "epoch": 0.7552870090634441, + "grad_norm": 1.446278841857033, + "learning_rate": 5.958394919628777e-07, + "loss": 1.0021, + "step": 8375 + }, + { + "epoch": 0.7553771925869144, + "grad_norm": 1.5351083374311916, + "learning_rate": 5.954235367545451e-07, + "loss": 0.9199, + "step": 8376 + }, + { + "epoch": 0.7554673761103846, + "grad_norm": 1.4539371804909051, + "learning_rate": 5.950077013924213e-07, + "loss": 0.9728, + "step": 8377 + }, + { + "epoch": 0.7555575596338548, + "grad_norm": 2.338511703967612, + "learning_rate": 5.945919859119865e-07, + "loss": 1.0662, + "step": 8378 + }, + { + "epoch": 0.7556477431573252, + "grad_norm": 2.094573015752053, + "learning_rate": 5.94176390348714e-07, + "loss": 0.8737, + "step": 8379 + }, + { + "epoch": 0.7557379266807954, + "grad_norm": 1.3425420610313956, + "learning_rate": 5.937609147380622e-07, + "loss": 1.0882, + "step": 8380 + }, + { + "epoch": 0.7558281102042657, + "grad_norm": 1.6524444621531122, + "learning_rate": 5.933455591154844e-07, + "loss": 0.9259, + "step": 8381 + }, + { + "epoch": 0.7559182937277359, + "grad_norm": 3.3616731102017514, + "learning_rate": 5.929303235164191e-07, + "loss": 0.9517, + "step": 8382 + }, + { + "epoch": 0.7560084772512062, + "grad_norm": 1.724040126492875, + "learning_rate": 5.92515207976297e-07, + "loss": 0.954, + "step": 8383 + }, + { + "epoch": 0.7560986607746765, + "grad_norm": 1.4479489585879832, + "learning_rate": 5.921002125305383e-07, + "loss": 0.9696, + "step": 8384 + }, + { + "epoch": 0.7561888442981467, + "grad_norm": 3.3103066146223523, + "learning_rate": 5.916853372145525e-07, + "loss": 0.902, + "step": 8385 + }, + { + "epoch": 0.756279027821617, + "grad_norm": 2.02411007105405, + "learning_rate": 5.912705820637389e-07, + "loss": 0.9137, + "step": 8386 + }, + { + "epoch": 0.7563692113450873, + "grad_norm": 1.540780439911434, + "learning_rate": 5.908559471134871e-07, + "loss": 1.0479, + "step": 8387 + }, + { + "epoch": 0.7564593948685575, + "grad_norm": 1.7306599737389632, + "learning_rate": 5.904414323991764e-07, + "loss": 0.9196, + "step": 8388 + }, + { + "epoch": 0.7565495783920277, + "grad_norm": 1.7611563518635216, + "learning_rate": 5.900270379561743e-07, + "loss": 0.9897, + "step": 8389 + }, + { + "epoch": 0.7566397619154981, + "grad_norm": 1.5460444831419824, + "learning_rate": 5.896127638198399e-07, + "loss": 0.9953, + "step": 8390 + }, + { + "epoch": 0.7567299454389683, + "grad_norm": 2.2049811369765973, + "learning_rate": 5.89198610025521e-07, + "loss": 0.9837, + "step": 8391 + }, + { + "epoch": 0.7568201289624386, + "grad_norm": 1.769803731456974, + "learning_rate": 5.887845766085559e-07, + "loss": 1.0468, + "step": 8392 + }, + { + "epoch": 0.7569103124859088, + "grad_norm": 2.330354480710568, + "learning_rate": 5.883706636042722e-07, + "loss": 0.9381, + "step": 8393 + }, + { + "epoch": 0.7570004960093791, + "grad_norm": 2.5334460304524127, + "learning_rate": 5.879568710479879e-07, + "loss": 0.9256, + "step": 8394 + }, + { + "epoch": 0.7570906795328494, + "grad_norm": 1.6699661338250422, + "learning_rate": 5.875431989750078e-07, + "loss": 1.0714, + "step": 8395 + }, + { + "epoch": 0.7571808630563196, + "grad_norm": 1.5355235041487358, + "learning_rate": 5.871296474206313e-07, + "loss": 0.9149, + "step": 8396 + }, + { + "epoch": 0.7572710465797898, + "grad_norm": 1.626230727483976, + "learning_rate": 5.867162164201427e-07, + "loss": 0.986, + "step": 8397 + }, + { + "epoch": 0.7573612301032602, + "grad_norm": 1.7477979429509423, + "learning_rate": 5.863029060088205e-07, + "loss": 1.0108, + "step": 8398 + }, + { + "epoch": 0.7574514136267304, + "grad_norm": 2.4667846969340133, + "learning_rate": 5.858897162219289e-07, + "loss": 0.8817, + "step": 8399 + }, + { + "epoch": 0.7575415971502006, + "grad_norm": 2.935501067818753, + "learning_rate": 5.854766470947238e-07, + "loss": 1.0306, + "step": 8400 + }, + { + "epoch": 0.7576317806736709, + "grad_norm": 1.5341457312193423, + "learning_rate": 5.850636986624511e-07, + "loss": 0.9699, + "step": 8401 + }, + { + "epoch": 0.7577219641971412, + "grad_norm": 1.7106418441260935, + "learning_rate": 5.846508709603453e-07, + "loss": 0.9899, + "step": 8402 + }, + { + "epoch": 0.7578121477206115, + "grad_norm": 1.8276414802360832, + "learning_rate": 5.842381640236318e-07, + "loss": 0.9591, + "step": 8403 + }, + { + "epoch": 0.7579023312440817, + "grad_norm": 1.6328201658808161, + "learning_rate": 5.838255778875242e-07, + "loss": 0.9825, + "step": 8404 + }, + { + "epoch": 0.7579925147675519, + "grad_norm": 1.8803187907593788, + "learning_rate": 5.83413112587227e-07, + "loss": 1.0013, + "step": 8405 + }, + { + "epoch": 0.7580826982910223, + "grad_norm": 1.7892283400252842, + "learning_rate": 5.830007681579338e-07, + "loss": 0.9688, + "step": 8406 + }, + { + "epoch": 0.7581728818144925, + "grad_norm": 2.1302564346615624, + "learning_rate": 5.825885446348284e-07, + "loss": 0.9952, + "step": 8407 + }, + { + "epoch": 0.7582630653379627, + "grad_norm": 1.6771172308803348, + "learning_rate": 5.821764420530842e-07, + "loss": 0.9214, + "step": 8408 + }, + { + "epoch": 0.7583532488614331, + "grad_norm": 1.6644861819197045, + "learning_rate": 5.817644604478633e-07, + "loss": 1.001, + "step": 8409 + }, + { + "epoch": 0.7584434323849033, + "grad_norm": 1.9058040660128224, + "learning_rate": 5.81352599854319e-07, + "loss": 0.9263, + "step": 8410 + }, + { + "epoch": 0.7585336159083735, + "grad_norm": 1.8041999455555031, + "learning_rate": 5.809408603075938e-07, + "loss": 1.0146, + "step": 8411 + }, + { + "epoch": 0.7586237994318438, + "grad_norm": 1.5356647920712312, + "learning_rate": 5.805292418428176e-07, + "loss": 0.9597, + "step": 8412 + }, + { + "epoch": 0.7587139829553141, + "grad_norm": 3.93786885789354, + "learning_rate": 5.801177444951148e-07, + "loss": 0.9762, + "step": 8413 + }, + { + "epoch": 0.7588041664787843, + "grad_norm": 1.2586150363516757, + "learning_rate": 5.797063682995944e-07, + "loss": 1.0105, + "step": 8414 + }, + { + "epoch": 0.7588943500022546, + "grad_norm": 1.7909035351204525, + "learning_rate": 5.792951132913584e-07, + "loss": 1.0463, + "step": 8415 + }, + { + "epoch": 0.7589845335257248, + "grad_norm": 2.8305997190148786, + "learning_rate": 5.788839795054968e-07, + "loss": 1.0337, + "step": 8416 + }, + { + "epoch": 0.7590747170491952, + "grad_norm": 2.234224114291872, + "learning_rate": 5.784729669770898e-07, + "loss": 0.8949, + "step": 8417 + }, + { + "epoch": 0.7591649005726654, + "grad_norm": 1.5385737538677495, + "learning_rate": 5.780620757412084e-07, + "loss": 0.9792, + "step": 8418 + }, + { + "epoch": 0.7592550840961356, + "grad_norm": 1.9724825640972659, + "learning_rate": 5.776513058329098e-07, + "loss": 1.0001, + "step": 8419 + }, + { + "epoch": 0.7593452676196059, + "grad_norm": 1.4729882632511153, + "learning_rate": 5.772406572872459e-07, + "loss": 1.0031, + "step": 8420 + }, + { + "epoch": 0.7594354511430762, + "grad_norm": 2.1494663064059103, + "learning_rate": 5.768301301392535e-07, + "loss": 1.0132, + "step": 8421 + }, + { + "epoch": 0.7595256346665464, + "grad_norm": 1.497178921675944, + "learning_rate": 5.764197244239615e-07, + "loss": 0.9925, + "step": 8422 + }, + { + "epoch": 0.7596158181900167, + "grad_norm": 4.555036946889061, + "learning_rate": 5.760094401763884e-07, + "loss": 1.0172, + "step": 8423 + }, + { + "epoch": 0.7597060017134869, + "grad_norm": 1.4619571167257333, + "learning_rate": 5.755992774315414e-07, + "loss": 0.9634, + "step": 8424 + }, + { + "epoch": 0.7597961852369572, + "grad_norm": 2.6698891917462815, + "learning_rate": 5.751892362244183e-07, + "loss": 0.9712, + "step": 8425 + }, + { + "epoch": 0.7598863687604275, + "grad_norm": 3.678730250936493, + "learning_rate": 5.747793165900065e-07, + "loss": 1.0362, + "step": 8426 + }, + { + "epoch": 0.7599765522838977, + "grad_norm": 2.2930703343727004, + "learning_rate": 5.743695185632806e-07, + "loss": 1.0395, + "step": 8427 + }, + { + "epoch": 0.7600667358073679, + "grad_norm": 3.852565302112796, + "learning_rate": 5.739598421792098e-07, + "loss": 1.0213, + "step": 8428 + }, + { + "epoch": 0.7601569193308383, + "grad_norm": 1.5045440510379455, + "learning_rate": 5.735502874727474e-07, + "loss": 0.9638, + "step": 8429 + }, + { + "epoch": 0.7602471028543085, + "grad_norm": 3.012678218650857, + "learning_rate": 5.731408544788398e-07, + "loss": 1.112, + "step": 8430 + }, + { + "epoch": 0.7603372863777788, + "grad_norm": 2.0566167440594763, + "learning_rate": 5.727315432324225e-07, + "loss": 0.8952, + "step": 8431 + }, + { + "epoch": 0.760427469901249, + "grad_norm": 1.5604885887748141, + "learning_rate": 5.723223537684196e-07, + "loss": 0.9693, + "step": 8432 + }, + { + "epoch": 0.7605176534247193, + "grad_norm": 1.6355938015718434, + "learning_rate": 5.719132861217462e-07, + "loss": 1.0613, + "step": 8433 + }, + { + "epoch": 0.7606078369481896, + "grad_norm": 2.0228681904896746, + "learning_rate": 5.715043403273044e-07, + "loss": 0.9467, + "step": 8434 + }, + { + "epoch": 0.7606980204716598, + "grad_norm": 2.276014093600116, + "learning_rate": 5.710955164199902e-07, + "loss": 0.9387, + "step": 8435 + }, + { + "epoch": 0.7607882039951301, + "grad_norm": 1.6381122056213262, + "learning_rate": 5.706868144346841e-07, + "loss": 1.0005, + "step": 8436 + }, + { + "epoch": 0.7608783875186004, + "grad_norm": 1.6637395267744468, + "learning_rate": 5.702782344062613e-07, + "loss": 0.9747, + "step": 8437 + }, + { + "epoch": 0.7609685710420706, + "grad_norm": 1.8215449012878702, + "learning_rate": 5.698697763695826e-07, + "loss": 0.9588, + "step": 8438 + }, + { + "epoch": 0.7610587545655408, + "grad_norm": 1.6306883213999825, + "learning_rate": 5.694614403595002e-07, + "loss": 1.0017, + "step": 8439 + }, + { + "epoch": 0.7611489380890112, + "grad_norm": 1.9551894606863436, + "learning_rate": 5.690532264108554e-07, + "loss": 0.8884, + "step": 8440 + }, + { + "epoch": 0.7612391216124814, + "grad_norm": 1.579224911356262, + "learning_rate": 5.686451345584795e-07, + "loss": 1.0192, + "step": 8441 + }, + { + "epoch": 0.7613293051359517, + "grad_norm": 2.745469349270617, + "learning_rate": 5.682371648371933e-07, + "loss": 1.0806, + "step": 8442 + }, + { + "epoch": 0.7614194886594219, + "grad_norm": 1.8943456185232443, + "learning_rate": 5.678293172818074e-07, + "loss": 0.9392, + "step": 8443 + }, + { + "epoch": 0.7615096721828922, + "grad_norm": 2.322777463109077, + "learning_rate": 5.674215919271204e-07, + "loss": 0.9948, + "step": 8444 + }, + { + "epoch": 0.7615998557063625, + "grad_norm": 1.9377209937284916, + "learning_rate": 5.670139888079224e-07, + "loss": 0.9079, + "step": 8445 + }, + { + "epoch": 0.7616900392298327, + "grad_norm": 1.5135913852087604, + "learning_rate": 5.666065079589924e-07, + "loss": 1.041, + "step": 8446 + }, + { + "epoch": 0.7617802227533029, + "grad_norm": 2.083636554268331, + "learning_rate": 5.661991494150986e-07, + "loss": 0.9328, + "step": 8447 + }, + { + "epoch": 0.7618704062767733, + "grad_norm": 1.5087321742916056, + "learning_rate": 5.657919132109999e-07, + "loss": 1.028, + "step": 8448 + }, + { + "epoch": 0.7619605898002435, + "grad_norm": 1.7456872795514295, + "learning_rate": 5.653847993814421e-07, + "loss": 1.0122, + "step": 8449 + }, + { + "epoch": 0.7620507733237137, + "grad_norm": 1.8957608666735475, + "learning_rate": 5.649778079611647e-07, + "loss": 0.9931, + "step": 8450 + }, + { + "epoch": 0.762140956847184, + "grad_norm": 1.6524865920638157, + "learning_rate": 5.645709389848923e-07, + "loss": 0.9633, + "step": 8451 + }, + { + "epoch": 0.7622311403706543, + "grad_norm": 1.8297727283202927, + "learning_rate": 5.641641924873435e-07, + "loss": 0.9425, + "step": 8452 + }, + { + "epoch": 0.7623213238941245, + "grad_norm": 1.9121846855381008, + "learning_rate": 5.637575685032217e-07, + "loss": 0.9821, + "step": 8453 + }, + { + "epoch": 0.7624115074175948, + "grad_norm": 1.6440886493133962, + "learning_rate": 5.633510670672246e-07, + "loss": 1.0641, + "step": 8454 + }, + { + "epoch": 0.762501690941065, + "grad_norm": 0.6331364544192879, + "learning_rate": 5.629446882140354e-07, + "loss": 0.8421, + "step": 8455 + }, + { + "epoch": 0.7625918744645354, + "grad_norm": 1.4973411837570458, + "learning_rate": 5.625384319783295e-07, + "loss": 0.9916, + "step": 8456 + }, + { + "epoch": 0.7626820579880056, + "grad_norm": 2.275701490131401, + "learning_rate": 5.621322983947705e-07, + "loss": 0.9266, + "step": 8457 + }, + { + "epoch": 0.7627722415114758, + "grad_norm": 1.7369242997394885, + "learning_rate": 5.617262874980122e-07, + "loss": 1.0179, + "step": 8458 + }, + { + "epoch": 0.7628624250349462, + "grad_norm": 1.5534563413439304, + "learning_rate": 5.613203993226981e-07, + "loss": 1.0266, + "step": 8459 + }, + { + "epoch": 0.7629526085584164, + "grad_norm": 2.7354531042883004, + "learning_rate": 5.609146339034599e-07, + "loss": 1.0346, + "step": 8460 + }, + { + "epoch": 0.7630427920818866, + "grad_norm": 2.1262259593324413, + "learning_rate": 5.605089912749199e-07, + "loss": 0.9258, + "step": 8461 + }, + { + "epoch": 0.7631329756053569, + "grad_norm": 4.348443877653763, + "learning_rate": 5.601034714716901e-07, + "loss": 0.922, + "step": 8462 + }, + { + "epoch": 0.7632231591288272, + "grad_norm": 1.7708086872255682, + "learning_rate": 5.59698074528372e-07, + "loss": 0.9541, + "step": 8463 + }, + { + "epoch": 0.7633133426522974, + "grad_norm": 1.3698412181560642, + "learning_rate": 5.592928004795555e-07, + "loss": 0.9505, + "step": 8464 + }, + { + "epoch": 0.7634035261757677, + "grad_norm": 1.907091611589041, + "learning_rate": 5.58887649359822e-07, + "loss": 0.9871, + "step": 8465 + }, + { + "epoch": 0.7634937096992379, + "grad_norm": 4.207586366942254, + "learning_rate": 5.584826212037393e-07, + "loss": 0.8326, + "step": 8466 + }, + { + "epoch": 0.7635838932227083, + "grad_norm": 1.2681831126953806, + "learning_rate": 5.580777160458689e-07, + "loss": 0.9008, + "step": 8467 + }, + { + "epoch": 0.7636740767461785, + "grad_norm": 1.648251636024657, + "learning_rate": 5.576729339207574e-07, + "loss": 1.0047, + "step": 8468 + }, + { + "epoch": 0.7637642602696487, + "grad_norm": 1.5529626834411987, + "learning_rate": 5.572682748629449e-07, + "loss": 1.0403, + "step": 8469 + }, + { + "epoch": 0.763854443793119, + "grad_norm": 1.4952207563072963, + "learning_rate": 5.568637389069582e-07, + "loss": 0.9642, + "step": 8470 + }, + { + "epoch": 0.7639446273165893, + "grad_norm": 1.9382923720939025, + "learning_rate": 5.564593260873145e-07, + "loss": 0.9967, + "step": 8471 + }, + { + "epoch": 0.7640348108400595, + "grad_norm": 1.441286573119691, + "learning_rate": 5.560550364385206e-07, + "loss": 1.0196, + "step": 8472 + }, + { + "epoch": 0.7641249943635298, + "grad_norm": 8.658202023530835, + "learning_rate": 5.556508699950728e-07, + "loss": 1.0516, + "step": 8473 + }, + { + "epoch": 0.764215177887, + "grad_norm": 2.830380378204871, + "learning_rate": 5.552468267914577e-07, + "loss": 0.9329, + "step": 8474 + }, + { + "epoch": 0.7643053614104703, + "grad_norm": 1.951123608363531, + "learning_rate": 5.548429068621481e-07, + "loss": 0.8971, + "step": 8475 + }, + { + "epoch": 0.7643955449339406, + "grad_norm": 0.6295949825351106, + "learning_rate": 5.544391102416115e-07, + "loss": 0.812, + "step": 8476 + }, + { + "epoch": 0.7644857284574108, + "grad_norm": 1.852145783180009, + "learning_rate": 5.540354369643003e-07, + "loss": 1.0375, + "step": 8477 + }, + { + "epoch": 0.764575911980881, + "grad_norm": 3.7878584688971033, + "learning_rate": 5.536318870646586e-07, + "loss": 0.83, + "step": 8478 + }, + { + "epoch": 0.7646660955043514, + "grad_norm": 1.8781266528231795, + "learning_rate": 5.532284605771194e-07, + "loss": 0.8883, + "step": 8479 + }, + { + "epoch": 0.7647562790278216, + "grad_norm": 1.8147775875253827, + "learning_rate": 5.528251575361052e-07, + "loss": 0.9461, + "step": 8480 + }, + { + "epoch": 0.7648464625512919, + "grad_norm": 1.3638704005168996, + "learning_rate": 5.524219779760284e-07, + "loss": 0.9787, + "step": 8481 + }, + { + "epoch": 0.7649366460747622, + "grad_norm": 1.6995136575085548, + "learning_rate": 5.520189219312907e-07, + "loss": 1.003, + "step": 8482 + }, + { + "epoch": 0.7650268295982324, + "grad_norm": 1.7456876892786823, + "learning_rate": 5.516159894362817e-07, + "loss": 0.8674, + "step": 8483 + }, + { + "epoch": 0.7651170131217027, + "grad_norm": 1.790063036868913, + "learning_rate": 5.512131805253839e-07, + "loss": 0.9192, + "step": 8484 + }, + { + "epoch": 0.7652071966451729, + "grad_norm": 3.131243001050608, + "learning_rate": 5.508104952329653e-07, + "loss": 0.996, + "step": 8485 + }, + { + "epoch": 0.7652973801686432, + "grad_norm": 3.3243314945811675, + "learning_rate": 5.504079335933862e-07, + "loss": 1.0105, + "step": 8486 + }, + { + "epoch": 0.7653875636921135, + "grad_norm": 1.751490571232147, + "learning_rate": 5.500054956409952e-07, + "loss": 1.0072, + "step": 8487 + }, + { + "epoch": 0.7654777472155837, + "grad_norm": 1.7923474533951325, + "learning_rate": 5.496031814101303e-07, + "loss": 0.9429, + "step": 8488 + }, + { + "epoch": 0.7655679307390539, + "grad_norm": 1.4269609143545527, + "learning_rate": 5.492009909351203e-07, + "loss": 0.9869, + "step": 8489 + }, + { + "epoch": 0.7656581142625243, + "grad_norm": 1.7782648738827354, + "learning_rate": 5.4879892425028e-07, + "loss": 0.9047, + "step": 8490 + }, + { + "epoch": 0.7657482977859945, + "grad_norm": 2.1405763272508795, + "learning_rate": 5.483969813899184e-07, + "loss": 1.114, + "step": 8491 + }, + { + "epoch": 0.7658384813094647, + "grad_norm": 1.4643041812250617, + "learning_rate": 5.479951623883299e-07, + "loss": 1.0593, + "step": 8492 + }, + { + "epoch": 0.765928664832935, + "grad_norm": 1.838739407921838, + "learning_rate": 5.475934672798004e-07, + "loss": 0.9137, + "step": 8493 + }, + { + "epoch": 0.7660188483564053, + "grad_norm": 1.7076750006297683, + "learning_rate": 5.471918960986047e-07, + "loss": 0.9435, + "step": 8494 + }, + { + "epoch": 0.7661090318798756, + "grad_norm": 1.7560474944198068, + "learning_rate": 5.467904488790071e-07, + "loss": 1.077, + "step": 8495 + }, + { + "epoch": 0.7661992154033458, + "grad_norm": 1.7907297286851802, + "learning_rate": 5.463891256552615e-07, + "loss": 0.9854, + "step": 8496 + }, + { + "epoch": 0.766289398926816, + "grad_norm": 1.734769003164107, + "learning_rate": 5.459879264616107e-07, + "loss": 0.8901, + "step": 8497 + }, + { + "epoch": 0.7663795824502864, + "grad_norm": 3.283230719600723, + "learning_rate": 5.455868513322874e-07, + "loss": 0.8652, + "step": 8498 + }, + { + "epoch": 0.7664697659737566, + "grad_norm": 1.761784046933788, + "learning_rate": 5.451859003015143e-07, + "loss": 0.8711, + "step": 8499 + }, + { + "epoch": 0.7665599494972268, + "grad_norm": 1.7356703448802717, + "learning_rate": 5.447850734035009e-07, + "loss": 1.0157, + "step": 8500 + }, + { + "epoch": 0.7666501330206971, + "grad_norm": 1.5751766953429427, + "learning_rate": 5.443843706724494e-07, + "loss": 0.9628, + "step": 8501 + }, + { + "epoch": 0.7667403165441674, + "grad_norm": 1.875271459637981, + "learning_rate": 5.439837921425494e-07, + "loss": 0.9112, + "step": 8502 + }, + { + "epoch": 0.7668305000676376, + "grad_norm": 1.8968987527833838, + "learning_rate": 5.435833378479807e-07, + "loss": 0.97, + "step": 8503 + }, + { + "epoch": 0.7669206835911079, + "grad_norm": 1.5047481409769004, + "learning_rate": 5.431830078229128e-07, + "loss": 0.8422, + "step": 8504 + }, + { + "epoch": 0.7670108671145782, + "grad_norm": 1.6273688143641833, + "learning_rate": 5.427828021015022e-07, + "loss": 0.9975, + "step": 8505 + }, + { + "epoch": 0.7671010506380485, + "grad_norm": 2.307531328577565, + "learning_rate": 5.42382720717899e-07, + "loss": 1.009, + "step": 8506 + }, + { + "epoch": 0.7671912341615187, + "grad_norm": 3.16896517472589, + "learning_rate": 5.419827637062384e-07, + "loss": 1.0064, + "step": 8507 + }, + { + "epoch": 0.7672814176849889, + "grad_norm": 1.6613892049568908, + "learning_rate": 5.415829311006487e-07, + "loss": 1.1217, + "step": 8508 + }, + { + "epoch": 0.7673716012084593, + "grad_norm": 3.261116675813149, + "learning_rate": 5.411832229352447e-07, + "loss": 1.0598, + "step": 8509 + }, + { + "epoch": 0.7674617847319295, + "grad_norm": 1.6288061317028653, + "learning_rate": 5.407836392441319e-07, + "loss": 1.0747, + "step": 8510 + }, + { + "epoch": 0.7675519682553997, + "grad_norm": 3.1462271909072617, + "learning_rate": 5.403841800614049e-07, + "loss": 0.9971, + "step": 8511 + }, + { + "epoch": 0.76764215177887, + "grad_norm": 1.4646713765768502, + "learning_rate": 5.39984845421148e-07, + "loss": 0.9351, + "step": 8512 + }, + { + "epoch": 0.7677323353023403, + "grad_norm": 2.583028119012807, + "learning_rate": 5.395856353574344e-07, + "loss": 0.9666, + "step": 8513 + }, + { + "epoch": 0.7678225188258105, + "grad_norm": 0.5911690683601978, + "learning_rate": 5.391865499043275e-07, + "loss": 0.8035, + "step": 8514 + }, + { + "epoch": 0.7679127023492808, + "grad_norm": 1.4182138782825469, + "learning_rate": 5.387875890958788e-07, + "loss": 0.9849, + "step": 8515 + }, + { + "epoch": 0.768002885872751, + "grad_norm": 1.7035137747702274, + "learning_rate": 5.383887529661298e-07, + "loss": 1.0073, + "step": 8516 + }, + { + "epoch": 0.7680930693962214, + "grad_norm": 2.2502097985684006, + "learning_rate": 5.379900415491116e-07, + "loss": 0.9334, + "step": 8517 + }, + { + "epoch": 0.7681832529196916, + "grad_norm": 1.840043561253921, + "learning_rate": 5.375914548788447e-07, + "loss": 0.9315, + "step": 8518 + }, + { + "epoch": 0.7682734364431618, + "grad_norm": 1.5206874969094852, + "learning_rate": 5.371929929893384e-07, + "loss": 0.8627, + "step": 8519 + }, + { + "epoch": 0.768363619966632, + "grad_norm": 1.6374165724839518, + "learning_rate": 5.367946559145917e-07, + "loss": 0.944, + "step": 8520 + }, + { + "epoch": 0.7684538034901024, + "grad_norm": 3.7515513707909034, + "learning_rate": 5.363964436885935e-07, + "loss": 1.0, + "step": 8521 + }, + { + "epoch": 0.7685439870135726, + "grad_norm": 1.4159551778585173, + "learning_rate": 5.359983563453199e-07, + "loss": 1.0655, + "step": 8522 + }, + { + "epoch": 0.7686341705370429, + "grad_norm": 1.8156490593529153, + "learning_rate": 5.356003939187402e-07, + "loss": 1.0933, + "step": 8523 + }, + { + "epoch": 0.7687243540605131, + "grad_norm": 1.7430720115313747, + "learning_rate": 5.352025564428082e-07, + "loss": 0.9546, + "step": 8524 + }, + { + "epoch": 0.7688145375839834, + "grad_norm": 1.6890407874928821, + "learning_rate": 5.348048439514723e-07, + "loss": 1.0215, + "step": 8525 + }, + { + "epoch": 0.7689047211074537, + "grad_norm": 2.0482396825369684, + "learning_rate": 5.344072564786653e-07, + "loss": 0.8581, + "step": 8526 + }, + { + "epoch": 0.7689949046309239, + "grad_norm": 1.5054467968213598, + "learning_rate": 5.340097940583123e-07, + "loss": 0.9927, + "step": 8527 + }, + { + "epoch": 0.7690850881543942, + "grad_norm": 2.2401738179410593, + "learning_rate": 5.336124567243275e-07, + "loss": 0.9425, + "step": 8528 + }, + { + "epoch": 0.7691752716778645, + "grad_norm": 1.7919120916193956, + "learning_rate": 5.33215244510613e-07, + "loss": 0.9836, + "step": 8529 + }, + { + "epoch": 0.7692654552013347, + "grad_norm": 2.0434463555388693, + "learning_rate": 5.328181574510624e-07, + "loss": 0.9757, + "step": 8530 + }, + { + "epoch": 0.769355638724805, + "grad_norm": 1.9541148004650613, + "learning_rate": 5.324211955795559e-07, + "loss": 1.0323, + "step": 8531 + }, + { + "epoch": 0.7694458222482753, + "grad_norm": 1.9287284908220217, + "learning_rate": 5.320243589299651e-07, + "loss": 0.909, + "step": 8532 + }, + { + "epoch": 0.7695360057717455, + "grad_norm": 1.82882383634247, + "learning_rate": 5.316276475361505e-07, + "loss": 0.9347, + "step": 8533 + }, + { + "epoch": 0.7696261892952158, + "grad_norm": 2.182952841395702, + "learning_rate": 5.312310614319613e-07, + "loss": 1.0158, + "step": 8534 + }, + { + "epoch": 0.769716372818686, + "grad_norm": 2.027416426397851, + "learning_rate": 5.308346006512367e-07, + "loss": 0.952, + "step": 8535 + }, + { + "epoch": 0.7698065563421563, + "grad_norm": 1.5619272326665106, + "learning_rate": 5.30438265227805e-07, + "loss": 0.913, + "step": 8536 + }, + { + "epoch": 0.7698967398656266, + "grad_norm": 0.6084193046995762, + "learning_rate": 5.300420551954837e-07, + "loss": 0.7971, + "step": 8537 + }, + { + "epoch": 0.7699869233890968, + "grad_norm": 1.5363614930060479, + "learning_rate": 5.296459705880798e-07, + "loss": 0.9066, + "step": 8538 + }, + { + "epoch": 0.770077106912567, + "grad_norm": 1.7947300589495543, + "learning_rate": 5.292500114393881e-07, + "loss": 1.0345, + "step": 8539 + }, + { + "epoch": 0.7701672904360374, + "grad_norm": 1.6267935318539912, + "learning_rate": 5.288541777831963e-07, + "loss": 1.004, + "step": 8540 + }, + { + "epoch": 0.7702574739595076, + "grad_norm": 1.7057720227475466, + "learning_rate": 5.284584696532772e-07, + "loss": 1.0367, + "step": 8541 + }, + { + "epoch": 0.7703476574829778, + "grad_norm": 1.7148083659281872, + "learning_rate": 5.280628870833954e-07, + "loss": 0.9886, + "step": 8542 + }, + { + "epoch": 0.7704378410064481, + "grad_norm": 1.7231452230623434, + "learning_rate": 5.276674301073045e-07, + "loss": 1.0116, + "step": 8543 + }, + { + "epoch": 0.7705280245299184, + "grad_norm": 2.8995186570684433, + "learning_rate": 5.272720987587467e-07, + "loss": 0.9968, + "step": 8544 + }, + { + "epoch": 0.7706182080533887, + "grad_norm": 2.315679992764099, + "learning_rate": 5.268768930714545e-07, + "loss": 0.9376, + "step": 8545 + }, + { + "epoch": 0.7707083915768589, + "grad_norm": 1.8739236285261751, + "learning_rate": 5.264818130791473e-07, + "loss": 1.0059, + "step": 8546 + }, + { + "epoch": 0.7707985751003291, + "grad_norm": 8.175805045802687, + "learning_rate": 5.260868588155378e-07, + "loss": 0.9833, + "step": 8547 + }, + { + "epoch": 0.7708887586237995, + "grad_norm": 2.450382911187145, + "learning_rate": 5.256920303143242e-07, + "loss": 1.0089, + "step": 8548 + }, + { + "epoch": 0.7709789421472697, + "grad_norm": 1.554099120171356, + "learning_rate": 5.252973276091956e-07, + "loss": 1.0933, + "step": 8549 + }, + { + "epoch": 0.7710691256707399, + "grad_norm": 1.497525001151599, + "learning_rate": 5.249027507338307e-07, + "loss": 1.0511, + "step": 8550 + }, + { + "epoch": 0.7711593091942102, + "grad_norm": 1.389446745207143, + "learning_rate": 5.245082997218966e-07, + "loss": 0.951, + "step": 8551 + }, + { + "epoch": 0.7712494927176805, + "grad_norm": 1.9327870545373735, + "learning_rate": 5.241139746070499e-07, + "loss": 1.0061, + "step": 8552 + }, + { + "epoch": 0.7713396762411507, + "grad_norm": 1.8182209753024816, + "learning_rate": 5.237197754229376e-07, + "loss": 0.8644, + "step": 8553 + }, + { + "epoch": 0.771429859764621, + "grad_norm": 1.5369833442564855, + "learning_rate": 5.233257022031931e-07, + "loss": 0.8882, + "step": 8554 + }, + { + "epoch": 0.7715200432880913, + "grad_norm": 2.2245650048390124, + "learning_rate": 5.229317549814432e-07, + "loss": 0.9971, + "step": 8555 + }, + { + "epoch": 0.7716102268115616, + "grad_norm": 1.7345268939064664, + "learning_rate": 5.225379337912998e-07, + "loss": 1.0593, + "step": 8556 + }, + { + "epoch": 0.7717004103350318, + "grad_norm": 3.556202190594025, + "learning_rate": 5.221442386663663e-07, + "loss": 0.9946, + "step": 8557 + }, + { + "epoch": 0.771790593858502, + "grad_norm": 1.4740779095819379, + "learning_rate": 5.217506696402354e-07, + "loss": 1.0201, + "step": 8558 + }, + { + "epoch": 0.7718807773819724, + "grad_norm": 1.8918321397083235, + "learning_rate": 5.213572267464883e-07, + "loss": 0.998, + "step": 8559 + }, + { + "epoch": 0.7719709609054426, + "grad_norm": 1.7699819500320706, + "learning_rate": 5.209639100186965e-07, + "loss": 0.9593, + "step": 8560 + }, + { + "epoch": 0.7720611444289128, + "grad_norm": 0.6624147954167247, + "learning_rate": 5.205707194904179e-07, + "loss": 0.8774, + "step": 8561 + }, + { + "epoch": 0.7721513279523831, + "grad_norm": 1.7514599432343625, + "learning_rate": 5.201776551952042e-07, + "loss": 0.9084, + "step": 8562 + }, + { + "epoch": 0.7722415114758534, + "grad_norm": 1.8383699576475434, + "learning_rate": 5.197847171665914e-07, + "loss": 0.9481, + "step": 8563 + }, + { + "epoch": 0.7723316949993236, + "grad_norm": 1.7366183781759472, + "learning_rate": 5.193919054381095e-07, + "loss": 0.967, + "step": 8564 + }, + { + "epoch": 0.7724218785227939, + "grad_norm": 1.6699966146240641, + "learning_rate": 5.189992200432738e-07, + "loss": 0.9458, + "step": 8565 + }, + { + "epoch": 0.7725120620462641, + "grad_norm": 1.8500972696456446, + "learning_rate": 5.186066610155906e-07, + "loss": 0.9942, + "step": 8566 + }, + { + "epoch": 0.7726022455697344, + "grad_norm": 1.7403754146120345, + "learning_rate": 5.182142283885555e-07, + "loss": 1.0353, + "step": 8567 + }, + { + "epoch": 0.7726924290932047, + "grad_norm": 0.6516476875071551, + "learning_rate": 5.178219221956528e-07, + "loss": 0.7678, + "step": 8568 + }, + { + "epoch": 0.7727826126166749, + "grad_norm": 1.709917644168355, + "learning_rate": 5.174297424703565e-07, + "loss": 1.0026, + "step": 8569 + }, + { + "epoch": 0.7728727961401451, + "grad_norm": 4.215047836665779, + "learning_rate": 5.170376892461299e-07, + "loss": 0.9129, + "step": 8570 + }, + { + "epoch": 0.7729629796636155, + "grad_norm": 0.7183394296023075, + "learning_rate": 5.16645762556424e-07, + "loss": 0.8676, + "step": 8571 + }, + { + "epoch": 0.7730531631870857, + "grad_norm": 1.496381766780294, + "learning_rate": 5.162539624346809e-07, + "loss": 1.0293, + "step": 8572 + }, + { + "epoch": 0.773143346710556, + "grad_norm": 1.887683376979048, + "learning_rate": 5.158622889143309e-07, + "loss": 1.0471, + "step": 8573 + }, + { + "epoch": 0.7732335302340262, + "grad_norm": 2.0903951612642087, + "learning_rate": 5.154707420287939e-07, + "loss": 0.9406, + "step": 8574 + }, + { + "epoch": 0.7733237137574965, + "grad_norm": 1.5845387872611263, + "learning_rate": 5.150793218114793e-07, + "loss": 0.9062, + "step": 8575 + }, + { + "epoch": 0.7734138972809668, + "grad_norm": 2.80706669195549, + "learning_rate": 5.146880282957837e-07, + "loss": 0.972, + "step": 8576 + }, + { + "epoch": 0.773504080804437, + "grad_norm": 1.665299148606266, + "learning_rate": 5.142968615150964e-07, + "loss": 0.9684, + "step": 8577 + }, + { + "epoch": 0.7735942643279073, + "grad_norm": 2.7260153147602404, + "learning_rate": 5.139058215027921e-07, + "loss": 0.9135, + "step": 8578 + }, + { + "epoch": 0.7736844478513776, + "grad_norm": 1.392972358018585, + "learning_rate": 5.135149082922383e-07, + "loss": 0.9697, + "step": 8579 + }, + { + "epoch": 0.7737746313748478, + "grad_norm": 2.0367805895912054, + "learning_rate": 5.131241219167879e-07, + "loss": 0.9097, + "step": 8580 + }, + { + "epoch": 0.773864814898318, + "grad_norm": 0.8163741256015284, + "learning_rate": 5.127334624097869e-07, + "loss": 0.8588, + "step": 8581 + }, + { + "epoch": 0.7739549984217884, + "grad_norm": 3.901701208909826, + "learning_rate": 5.123429298045672e-07, + "loss": 1.036, + "step": 8582 + }, + { + "epoch": 0.7740451819452586, + "grad_norm": 1.824436673007004, + "learning_rate": 5.119525241344515e-07, + "loss": 0.8706, + "step": 8583 + }, + { + "epoch": 0.7741353654687289, + "grad_norm": 1.6616931219236493, + "learning_rate": 5.115622454327515e-07, + "loss": 0.9929, + "step": 8584 + }, + { + "epoch": 0.7742255489921991, + "grad_norm": 2.152967414570389, + "learning_rate": 5.11172093732768e-07, + "loss": 0.9677, + "step": 8585 + }, + { + "epoch": 0.7743157325156694, + "grad_norm": 1.804867359793274, + "learning_rate": 5.107820690677911e-07, + "loss": 1.0079, + "step": 8586 + }, + { + "epoch": 0.7744059160391397, + "grad_norm": 1.5463839435462152, + "learning_rate": 5.103921714710991e-07, + "loss": 0.9297, + "step": 8587 + }, + { + "epoch": 0.7744960995626099, + "grad_norm": 1.6394865173036746, + "learning_rate": 5.100024009759605e-07, + "loss": 0.9961, + "step": 8588 + }, + { + "epoch": 0.7745862830860801, + "grad_norm": 2.0357122851483833, + "learning_rate": 5.09612757615633e-07, + "loss": 1.0166, + "step": 8589 + }, + { + "epoch": 0.7746764666095505, + "grad_norm": 1.4507222974526075, + "learning_rate": 5.092232414233628e-07, + "loss": 0.9342, + "step": 8590 + }, + { + "epoch": 0.7747666501330207, + "grad_norm": 1.7543646334514107, + "learning_rate": 5.088338524323858e-07, + "loss": 0.9512, + "step": 8591 + }, + { + "epoch": 0.7748568336564909, + "grad_norm": 1.7499482964643447, + "learning_rate": 5.084445906759271e-07, + "loss": 1.0413, + "step": 8592 + }, + { + "epoch": 0.7749470171799612, + "grad_norm": 1.773874035854157, + "learning_rate": 5.080554561871995e-07, + "loss": 1.0455, + "step": 8593 + }, + { + "epoch": 0.7750372007034315, + "grad_norm": 1.431249380319786, + "learning_rate": 5.076664489994078e-07, + "loss": 1.0112, + "step": 8594 + }, + { + "epoch": 0.7751273842269017, + "grad_norm": 1.7411848527341924, + "learning_rate": 5.07277569145742e-07, + "loss": 0.9509, + "step": 8595 + }, + { + "epoch": 0.775217567750372, + "grad_norm": 1.6692979148854008, + "learning_rate": 5.068888166593861e-07, + "loss": 0.8989, + "step": 8596 + }, + { + "epoch": 0.7753077512738422, + "grad_norm": 1.8038600990071711, + "learning_rate": 5.065001915735087e-07, + "loss": 0.9684, + "step": 8597 + }, + { + "epoch": 0.7753979347973126, + "grad_norm": 1.4386842244985676, + "learning_rate": 5.061116939212702e-07, + "loss": 0.9474, + "step": 8598 + }, + { + "epoch": 0.7754881183207828, + "grad_norm": 1.7569814794538037, + "learning_rate": 5.05723323735819e-07, + "loss": 0.9786, + "step": 8599 + }, + { + "epoch": 0.775578301844253, + "grad_norm": 0.6471715081715189, + "learning_rate": 5.053350810502932e-07, + "loss": 0.8194, + "step": 8600 + }, + { + "epoch": 0.7756684853677234, + "grad_norm": 2.017033284598744, + "learning_rate": 5.049469658978202e-07, + "loss": 0.977, + "step": 8601 + }, + { + "epoch": 0.7757586688911936, + "grad_norm": 0.6453535671882995, + "learning_rate": 5.045589783115147e-07, + "loss": 0.7682, + "step": 8602 + }, + { + "epoch": 0.7758488524146638, + "grad_norm": 1.658397649671156, + "learning_rate": 5.041711183244842e-07, + "loss": 1.0167, + "step": 8603 + }, + { + "epoch": 0.7759390359381341, + "grad_norm": 1.953395489082428, + "learning_rate": 5.037833859698211e-07, + "loss": 0.9981, + "step": 8604 + }, + { + "epoch": 0.7760292194616044, + "grad_norm": 2.0532662585478803, + "learning_rate": 5.033957812806096e-07, + "loss": 1.0485, + "step": 8605 + }, + { + "epoch": 0.7761194029850746, + "grad_norm": 2.255950688086971, + "learning_rate": 5.030083042899223e-07, + "loss": 1.0353, + "step": 8606 + }, + { + "epoch": 0.7762095865085449, + "grad_norm": 2.176883000795314, + "learning_rate": 5.026209550308207e-07, + "loss": 0.9822, + "step": 8607 + }, + { + "epoch": 0.7762997700320151, + "grad_norm": 1.812604046170078, + "learning_rate": 5.022337335363558e-07, + "loss": 1.0387, + "step": 8608 + }, + { + "epoch": 0.7763899535554855, + "grad_norm": 1.760067116909813, + "learning_rate": 5.018466398395677e-07, + "loss": 1.0092, + "step": 8609 + }, + { + "epoch": 0.7764801370789557, + "grad_norm": 2.1542613492580194, + "learning_rate": 5.01459673973484e-07, + "loss": 0.9147, + "step": 8610 + }, + { + "epoch": 0.7765703206024259, + "grad_norm": 2.9430207663350085, + "learning_rate": 5.01072835971125e-07, + "loss": 1.0461, + "step": 8611 + }, + { + "epoch": 0.7766605041258962, + "grad_norm": 1.4585774172015056, + "learning_rate": 5.006861258654959e-07, + "loss": 0.8862, + "step": 8612 + }, + { + "epoch": 0.7767506876493665, + "grad_norm": 1.8428100517609887, + "learning_rate": 5.002995436895938e-07, + "loss": 1.0096, + "step": 8613 + }, + { + "epoch": 0.7768408711728367, + "grad_norm": 9.59594820665802, + "learning_rate": 4.999130894764039e-07, + "loss": 0.9401, + "step": 8614 + }, + { + "epoch": 0.776931054696307, + "grad_norm": 5.049353590281384, + "learning_rate": 4.995267632589006e-07, + "loss": 0.9214, + "step": 8615 + }, + { + "epoch": 0.7770212382197772, + "grad_norm": 0.7015004145416427, + "learning_rate": 4.99140565070048e-07, + "loss": 0.8174, + "step": 8616 + }, + { + "epoch": 0.7771114217432475, + "grad_norm": 2.140815447946302, + "learning_rate": 4.987544949427969e-07, + "loss": 1.0037, + "step": 8617 + }, + { + "epoch": 0.7772016052667178, + "grad_norm": 0.6683926046238565, + "learning_rate": 4.98368552910091e-07, + "loss": 0.8249, + "step": 8618 + }, + { + "epoch": 0.777291788790188, + "grad_norm": 1.7904148238097641, + "learning_rate": 4.979827390048596e-07, + "loss": 1.0946, + "step": 8619 + }, + { + "epoch": 0.7773819723136582, + "grad_norm": 1.430550903427971, + "learning_rate": 4.975970532600231e-07, + "loss": 0.9415, + "step": 8620 + }, + { + "epoch": 0.7774721558371286, + "grad_norm": 1.864737719177156, + "learning_rate": 4.972114957084901e-07, + "loss": 1.0002, + "step": 8621 + }, + { + "epoch": 0.7775623393605988, + "grad_norm": 1.4247008528337826, + "learning_rate": 4.968260663831585e-07, + "loss": 0.9634, + "step": 8622 + }, + { + "epoch": 0.777652522884069, + "grad_norm": 1.2735017631557497, + "learning_rate": 4.964407653169154e-07, + "loss": 1.0905, + "step": 8623 + }, + { + "epoch": 0.7777427064075394, + "grad_norm": 1.6403115836003033, + "learning_rate": 4.960555925426366e-07, + "loss": 0.9517, + "step": 8624 + }, + { + "epoch": 0.7778328899310096, + "grad_norm": 1.7794175342638747, + "learning_rate": 4.956705480931876e-07, + "loss": 1.0402, + "step": 8625 + }, + { + "epoch": 0.7779230734544799, + "grad_norm": 1.3363444729016623, + "learning_rate": 4.952856320014225e-07, + "loss": 0.9615, + "step": 8626 + }, + { + "epoch": 0.7780132569779501, + "grad_norm": 4.966683395835472, + "learning_rate": 4.949008443001838e-07, + "loss": 0.9182, + "step": 8627 + }, + { + "epoch": 0.7781034405014204, + "grad_norm": 1.5759175035069244, + "learning_rate": 4.945161850223041e-07, + "loss": 0.9959, + "step": 8628 + }, + { + "epoch": 0.7781936240248907, + "grad_norm": 1.7654300683366457, + "learning_rate": 4.941316542006044e-07, + "loss": 0.9125, + "step": 8629 + }, + { + "epoch": 0.7782838075483609, + "grad_norm": 1.5768261843037683, + "learning_rate": 4.937472518678956e-07, + "loss": 1.007, + "step": 8630 + }, + { + "epoch": 0.7783739910718311, + "grad_norm": 1.9402956942719651, + "learning_rate": 4.93362978056977e-07, + "loss": 0.9827, + "step": 8631 + }, + { + "epoch": 0.7784641745953015, + "grad_norm": 1.8978099649237452, + "learning_rate": 4.929788328006355e-07, + "loss": 0.9306, + "step": 8632 + }, + { + "epoch": 0.7785543581187717, + "grad_norm": 2.330808897448386, + "learning_rate": 4.925948161316506e-07, + "loss": 1.0524, + "step": 8633 + }, + { + "epoch": 0.778644541642242, + "grad_norm": 2.4623442005113727, + "learning_rate": 4.922109280827868e-07, + "loss": 0.9709, + "step": 8634 + }, + { + "epoch": 0.7787347251657122, + "grad_norm": 1.511459840983667, + "learning_rate": 4.918271686868016e-07, + "loss": 0.9759, + "step": 8635 + }, + { + "epoch": 0.7788249086891825, + "grad_norm": 1.838483822122439, + "learning_rate": 4.914435379764379e-07, + "loss": 1.0216, + "step": 8636 + }, + { + "epoch": 0.7789150922126528, + "grad_norm": 1.5952352447186062, + "learning_rate": 4.910600359844294e-07, + "loss": 0.9753, + "step": 8637 + }, + { + "epoch": 0.779005275736123, + "grad_norm": 1.702157261837066, + "learning_rate": 4.90676662743499e-07, + "loss": 0.8886, + "step": 8638 + }, + { + "epoch": 0.7790954592595932, + "grad_norm": 2.179036142246454, + "learning_rate": 4.902934182863581e-07, + "loss": 0.971, + "step": 8639 + }, + { + "epoch": 0.7791856427830636, + "grad_norm": 1.8331215548295983, + "learning_rate": 4.899103026457069e-07, + "loss": 0.8559, + "step": 8640 + }, + { + "epoch": 0.7792758263065338, + "grad_norm": 0.6022139898075121, + "learning_rate": 4.895273158542361e-07, + "loss": 0.7671, + "step": 8641 + }, + { + "epoch": 0.779366009830004, + "grad_norm": 2.9757309600674757, + "learning_rate": 4.891444579446227e-07, + "loss": 0.9948, + "step": 8642 + }, + { + "epoch": 0.7794561933534743, + "grad_norm": 1.9009660749343076, + "learning_rate": 4.887617289495349e-07, + "loss": 0.9407, + "step": 8643 + }, + { + "epoch": 0.7795463768769446, + "grad_norm": 1.6927370117888105, + "learning_rate": 4.883791289016292e-07, + "loss": 0.748, + "step": 8644 + }, + { + "epoch": 0.7796365604004148, + "grad_norm": 2.1360692461483595, + "learning_rate": 4.879966578335514e-07, + "loss": 0.9926, + "step": 8645 + }, + { + "epoch": 0.7797267439238851, + "grad_norm": 1.880112163870422, + "learning_rate": 4.876143157779358e-07, + "loss": 0.8759, + "step": 8646 + }, + { + "epoch": 0.7798169274473554, + "grad_norm": 2.281588803568008, + "learning_rate": 4.872321027674058e-07, + "loss": 0.8696, + "step": 8647 + }, + { + "epoch": 0.7799071109708257, + "grad_norm": 1.6008838239045207, + "learning_rate": 4.868500188345748e-07, + "loss": 1.0527, + "step": 8648 + }, + { + "epoch": 0.7799972944942959, + "grad_norm": 1.6060532297218013, + "learning_rate": 4.864680640120425e-07, + "loss": 1.0455, + "step": 8649 + }, + { + "epoch": 0.7800874780177661, + "grad_norm": 2.9284543303096724, + "learning_rate": 4.860862383324016e-07, + "loss": 0.9897, + "step": 8650 + }, + { + "epoch": 0.7801776615412365, + "grad_norm": 1.855624094031663, + "learning_rate": 4.857045418282295e-07, + "loss": 0.9786, + "step": 8651 + }, + { + "epoch": 0.7802678450647067, + "grad_norm": 1.8717334267091563, + "learning_rate": 4.853229745320966e-07, + "loss": 0.9324, + "step": 8652 + }, + { + "epoch": 0.7803580285881769, + "grad_norm": 1.475834500516792, + "learning_rate": 4.849415364765587e-07, + "loss": 0.9664, + "step": 8653 + }, + { + "epoch": 0.7804482121116472, + "grad_norm": 1.6522516175324355, + "learning_rate": 4.845602276941631e-07, + "loss": 1.0437, + "step": 8654 + }, + { + "epoch": 0.7805383956351175, + "grad_norm": 1.6771197186742488, + "learning_rate": 4.841790482174449e-07, + "loss": 1.0169, + "step": 8655 + }, + { + "epoch": 0.7806285791585877, + "grad_norm": 1.9674965031749085, + "learning_rate": 4.837979980789282e-07, + "loss": 0.9703, + "step": 8656 + }, + { + "epoch": 0.780718762682058, + "grad_norm": 1.5420589807018987, + "learning_rate": 4.834170773111273e-07, + "loss": 0.9263, + "step": 8657 + }, + { + "epoch": 0.7808089462055282, + "grad_norm": 1.871177910363431, + "learning_rate": 4.830362859465431e-07, + "loss": 0.9979, + "step": 8658 + }, + { + "epoch": 0.7808991297289986, + "grad_norm": 1.6259904190710248, + "learning_rate": 4.826556240176675e-07, + "loss": 1.0434, + "step": 8659 + }, + { + "epoch": 0.7809893132524688, + "grad_norm": 2.0849728300304857, + "learning_rate": 4.822750915569807e-07, + "loss": 0.9945, + "step": 8660 + }, + { + "epoch": 0.781079496775939, + "grad_norm": 2.002244762484773, + "learning_rate": 4.818946885969514e-07, + "loss": 1.0575, + "step": 8661 + }, + { + "epoch": 0.7811696802994093, + "grad_norm": 1.719985379223449, + "learning_rate": 4.815144151700383e-07, + "loss": 1.0263, + "step": 8662 + }, + { + "epoch": 0.7812598638228796, + "grad_norm": 2.0190117350403063, + "learning_rate": 4.811342713086885e-07, + "loss": 0.9816, + "step": 8663 + }, + { + "epoch": 0.7813500473463498, + "grad_norm": 1.9000778332880184, + "learning_rate": 4.807542570453367e-07, + "loss": 1.0173, + "step": 8664 + }, + { + "epoch": 0.7814402308698201, + "grad_norm": 1.6444258486208945, + "learning_rate": 4.803743724124098e-07, + "loss": 0.9497, + "step": 8665 + }, + { + "epoch": 0.7815304143932903, + "grad_norm": 1.949747934922642, + "learning_rate": 4.799946174423192e-07, + "loss": 0.9896, + "step": 8666 + }, + { + "epoch": 0.7816205979167606, + "grad_norm": 0.6672641093740791, + "learning_rate": 4.796149921674706e-07, + "loss": 0.7691, + "step": 8667 + }, + { + "epoch": 0.7817107814402309, + "grad_norm": 1.5233925837228715, + "learning_rate": 4.792354966202534e-07, + "loss": 0.9392, + "step": 8668 + }, + { + "epoch": 0.7818009649637011, + "grad_norm": 2.3160405248216116, + "learning_rate": 4.788561308330489e-07, + "loss": 0.8061, + "step": 8669 + }, + { + "epoch": 0.7818911484871713, + "grad_norm": 2.5944791021852587, + "learning_rate": 4.784768948382272e-07, + "loss": 0.9268, + "step": 8670 + }, + { + "epoch": 0.7819813320106417, + "grad_norm": 5.850863935793244, + "learning_rate": 4.780977886681461e-07, + "loss": 0.9859, + "step": 8671 + }, + { + "epoch": 0.7820715155341119, + "grad_norm": 2.0984435353366298, + "learning_rate": 4.777188123551541e-07, + "loss": 0.9297, + "step": 8672 + }, + { + "epoch": 0.7821616990575821, + "grad_norm": 2.3886368406632026, + "learning_rate": 4.773399659315856e-07, + "loss": 1.0126, + "step": 8673 + }, + { + "epoch": 0.7822518825810525, + "grad_norm": 2.0317023140479304, + "learning_rate": 4.769612494297681e-07, + "loss": 0.9662, + "step": 8674 + }, + { + "epoch": 0.7823420661045227, + "grad_norm": 0.6388426386175969, + "learning_rate": 4.765826628820142e-07, + "loss": 0.7917, + "step": 8675 + }, + { + "epoch": 0.782432249627993, + "grad_norm": 1.7074295382392088, + "learning_rate": 4.7620420632062775e-07, + "loss": 0.8892, + "step": 8676 + }, + { + "epoch": 0.7825224331514632, + "grad_norm": 2.1358103939654725, + "learning_rate": 4.758258797779002e-07, + "loss": 0.9368, + "step": 8677 + }, + { + "epoch": 0.7826126166749335, + "grad_norm": 0.6136117032744117, + "learning_rate": 4.7544768328611317e-07, + "loss": 0.8014, + "step": 8678 + }, + { + "epoch": 0.7827028001984038, + "grad_norm": 3.4647569097153714, + "learning_rate": 4.750696168775359e-07, + "loss": 0.9653, + "step": 8679 + }, + { + "epoch": 0.782792983721874, + "grad_norm": 0.6017390400831821, + "learning_rate": 4.746916805844279e-07, + "loss": 0.7887, + "step": 8680 + }, + { + "epoch": 0.7828831672453442, + "grad_norm": 1.554043200227464, + "learning_rate": 4.743138744390356e-07, + "loss": 1.0076, + "step": 8681 + }, + { + "epoch": 0.7829733507688146, + "grad_norm": 1.6597905014159764, + "learning_rate": 4.739361984735959e-07, + "loss": 1.0532, + "step": 8682 + }, + { + "epoch": 0.7830635342922848, + "grad_norm": 1.9220179062722724, + "learning_rate": 4.7355865272033455e-07, + "loss": 0.936, + "step": 8683 + }, + { + "epoch": 0.783153717815755, + "grad_norm": 1.574819823964332, + "learning_rate": 4.7318123721146563e-07, + "loss": 0.9696, + "step": 8684 + }, + { + "epoch": 0.7832439013392253, + "grad_norm": 1.6077820384825836, + "learning_rate": 4.728039519791924e-07, + "loss": 0.9672, + "step": 8685 + }, + { + "epoch": 0.7833340848626956, + "grad_norm": 0.6622945700941095, + "learning_rate": 4.72426797055707e-07, + "loss": 0.8353, + "step": 8686 + }, + { + "epoch": 0.7834242683861659, + "grad_norm": 1.59585260557219, + "learning_rate": 4.720497724731904e-07, + "loss": 0.9466, + "step": 8687 + }, + { + "epoch": 0.7835144519096361, + "grad_norm": 1.8237209253431694, + "learning_rate": 4.7167287826381153e-07, + "loss": 1.0183, + "step": 8688 + }, + { + "epoch": 0.7836046354331063, + "grad_norm": 1.3913880729880217, + "learning_rate": 4.712961144597307e-07, + "loss": 0.9891, + "step": 8689 + }, + { + "epoch": 0.7836948189565767, + "grad_norm": 2.720493765920778, + "learning_rate": 4.7091948109309343e-07, + "loss": 1.0199, + "step": 8690 + }, + { + "epoch": 0.7837850024800469, + "grad_norm": 1.96789526171825, + "learning_rate": 4.705429781960384e-07, + "loss": 0.9906, + "step": 8691 + }, + { + "epoch": 0.7838751860035171, + "grad_norm": 4.183007178694155, + "learning_rate": 4.7016660580068923e-07, + "loss": 0.9785, + "step": 8692 + }, + { + "epoch": 0.7839653695269874, + "grad_norm": 1.9111192144300984, + "learning_rate": 4.6979036393916093e-07, + "loss": 1.0139, + "step": 8693 + }, + { + "epoch": 0.7840555530504577, + "grad_norm": 1.873363734583493, + "learning_rate": 4.6941425264355603e-07, + "loss": 1.0592, + "step": 8694 + }, + { + "epoch": 0.7841457365739279, + "grad_norm": 2.15014558343575, + "learning_rate": 4.6903827194596666e-07, + "loss": 0.9929, + "step": 8695 + }, + { + "epoch": 0.7842359200973982, + "grad_norm": 1.4845234043824231, + "learning_rate": 4.686624218784743e-07, + "loss": 1.0405, + "step": 8696 + }, + { + "epoch": 0.7843261036208685, + "grad_norm": 3.0418777566759743, + "learning_rate": 4.6828670247314696e-07, + "loss": 0.9732, + "step": 8697 + }, + { + "epoch": 0.7844162871443388, + "grad_norm": 1.6514206781189262, + "learning_rate": 4.679111137620442e-07, + "loss": 0.997, + "step": 8698 + }, + { + "epoch": 0.784506470667809, + "grad_norm": 1.5982411253700797, + "learning_rate": 4.67535655777213e-07, + "loss": 0.9448, + "step": 8699 + }, + { + "epoch": 0.7845966541912792, + "grad_norm": 2.0490322247750288, + "learning_rate": 4.6716032855068956e-07, + "loss": 0.989, + "step": 8700 + }, + { + "epoch": 0.7846868377147496, + "grad_norm": 2.508765969294503, + "learning_rate": 4.6678513211449867e-07, + "loss": 1.0241, + "step": 8701 + }, + { + "epoch": 0.7847770212382198, + "grad_norm": 2.5575441959623175, + "learning_rate": 4.6641006650065516e-07, + "loss": 0.9476, + "step": 8702 + }, + { + "epoch": 0.78486720476169, + "grad_norm": 1.5576971625481544, + "learning_rate": 4.6603513174115973e-07, + "loss": 0.9703, + "step": 8703 + }, + { + "epoch": 0.7849573882851603, + "grad_norm": 1.9538143924930818, + "learning_rate": 4.6566032786800625e-07, + "loss": 0.9972, + "step": 8704 + }, + { + "epoch": 0.7850475718086306, + "grad_norm": 1.6880132812908586, + "learning_rate": 4.6528565491317274e-07, + "loss": 0.9612, + "step": 8705 + }, + { + "epoch": 0.7851377553321008, + "grad_norm": 1.9994224071934086, + "learning_rate": 4.649111129086305e-07, + "loss": 1.048, + "step": 8706 + }, + { + "epoch": 0.7852279388555711, + "grad_norm": 2.06709886564163, + "learning_rate": 4.6453670188633596e-07, + "loss": 1.0053, + "step": 8707 + }, + { + "epoch": 0.7853181223790413, + "grad_norm": 2.004887093577991, + "learning_rate": 4.641624218782365e-07, + "loss": 0.9957, + "step": 8708 + }, + { + "epoch": 0.7854083059025116, + "grad_norm": 2.8487166744490717, + "learning_rate": 4.6378827291626765e-07, + "loss": 0.9449, + "step": 8709 + }, + { + "epoch": 0.7854984894259819, + "grad_norm": 1.2019440936671941, + "learning_rate": 4.634142550323541e-07, + "loss": 1.0347, + "step": 8710 + }, + { + "epoch": 0.7855886729494521, + "grad_norm": 3.224807224869921, + "learning_rate": 4.6304036825840943e-07, + "loss": 1.0301, + "step": 8711 + }, + { + "epoch": 0.7856788564729223, + "grad_norm": 1.6561233903967905, + "learning_rate": 4.626666126263341e-07, + "loss": 0.8876, + "step": 8712 + }, + { + "epoch": 0.7857690399963927, + "grad_norm": 2.926923989589159, + "learning_rate": 4.622929881680213e-07, + "loss": 1.0039, + "step": 8713 + }, + { + "epoch": 0.7858592235198629, + "grad_norm": 1.6445738000869694, + "learning_rate": 4.6191949491534887e-07, + "loss": 0.9001, + "step": 8714 + }, + { + "epoch": 0.7859494070433332, + "grad_norm": 2.091633082369713, + "learning_rate": 4.6154613290018617e-07, + "loss": 0.9939, + "step": 8715 + }, + { + "epoch": 0.7860395905668034, + "grad_norm": 2.0253657870116966, + "learning_rate": 4.6117290215439043e-07, + "loss": 0.929, + "step": 8716 + }, + { + "epoch": 0.7861297740902737, + "grad_norm": 2.0707082694017993, + "learning_rate": 4.6079980270980744e-07, + "loss": 0.8458, + "step": 8717 + }, + { + "epoch": 0.786219957613744, + "grad_norm": 1.8109886840222431, + "learning_rate": 4.6042683459827245e-07, + "loss": 0.9534, + "step": 8718 + }, + { + "epoch": 0.7863101411372142, + "grad_norm": 0.7052299357213566, + "learning_rate": 4.600539978516098e-07, + "loss": 0.7998, + "step": 8719 + }, + { + "epoch": 0.7864003246606845, + "grad_norm": 7.668444081914608, + "learning_rate": 4.5968129250163004e-07, + "loss": 0.9444, + "step": 8720 + }, + { + "epoch": 0.7864905081841548, + "grad_norm": 2.00418047302101, + "learning_rate": 4.5930871858013653e-07, + "loss": 0.912, + "step": 8721 + }, + { + "epoch": 0.786580691707625, + "grad_norm": 0.656517927789348, + "learning_rate": 4.589362761189182e-07, + "loss": 0.8585, + "step": 8722 + }, + { + "epoch": 0.7866708752310952, + "grad_norm": 1.7816048653461225, + "learning_rate": 4.585639651497539e-07, + "loss": 1.0495, + "step": 8723 + }, + { + "epoch": 0.7867610587545656, + "grad_norm": 1.5192188245755054, + "learning_rate": 4.581917857044115e-07, + "loss": 0.9357, + "step": 8724 + }, + { + "epoch": 0.7868512422780358, + "grad_norm": 1.5216536813344155, + "learning_rate": 4.5781973781464734e-07, + "loss": 1.0187, + "step": 8725 + }, + { + "epoch": 0.7869414258015061, + "grad_norm": 1.5073354490187412, + "learning_rate": 4.574478215122073e-07, + "loss": 1.0191, + "step": 8726 + }, + { + "epoch": 0.7870316093249763, + "grad_norm": 1.9376902333052572, + "learning_rate": 4.5707603682882357e-07, + "loss": 0.9618, + "step": 8727 + }, + { + "epoch": 0.7871217928484466, + "grad_norm": 2.5872101662890783, + "learning_rate": 4.56704383796221e-07, + "loss": 0.8457, + "step": 8728 + }, + { + "epoch": 0.7872119763719169, + "grad_norm": 1.7895631714270939, + "learning_rate": 4.5633286244610956e-07, + "loss": 1.0162, + "step": 8729 + }, + { + "epoch": 0.7873021598953871, + "grad_norm": 2.792646909723844, + "learning_rate": 4.5596147281018993e-07, + "loss": 1.0486, + "step": 8730 + }, + { + "epoch": 0.7873923434188573, + "grad_norm": 2.279173650736688, + "learning_rate": 4.5559021492015137e-07, + "loss": 1.0796, + "step": 8731 + }, + { + "epoch": 0.7874825269423277, + "grad_norm": 1.5452697593710571, + "learning_rate": 4.552190888076712e-07, + "loss": 0.894, + "step": 8732 + }, + { + "epoch": 0.7875727104657979, + "grad_norm": 1.6465215893722132, + "learning_rate": 4.548480945044164e-07, + "loss": 0.9309, + "step": 8733 + }, + { + "epoch": 0.7876628939892681, + "grad_norm": 1.6054016247752632, + "learning_rate": 4.54477232042042e-07, + "loss": 0.9772, + "step": 8734 + }, + { + "epoch": 0.7877530775127384, + "grad_norm": 1.4714117078304187, + "learning_rate": 4.541065014521921e-07, + "loss": 0.9733, + "step": 8735 + }, + { + "epoch": 0.7878432610362087, + "grad_norm": 3.5580760859845006, + "learning_rate": 4.5373590276649996e-07, + "loss": 0.9279, + "step": 8736 + }, + { + "epoch": 0.787933444559679, + "grad_norm": 1.4724511932159918, + "learning_rate": 4.533654360165862e-07, + "loss": 0.9482, + "step": 8737 + }, + { + "epoch": 0.7880236280831492, + "grad_norm": 2.1143667426471002, + "learning_rate": 4.5299510123406115e-07, + "loss": 1.0389, + "step": 8738 + }, + { + "epoch": 0.7881138116066194, + "grad_norm": 2.418352876142132, + "learning_rate": 4.5262489845052456e-07, + "loss": 0.9489, + "step": 8739 + }, + { + "epoch": 0.7882039951300898, + "grad_norm": 1.6896288124420193, + "learning_rate": 4.5225482769756353e-07, + "loss": 0.9461, + "step": 8740 + }, + { + "epoch": 0.78829417865356, + "grad_norm": 22.456297074992406, + "learning_rate": 4.5188488900675545e-07, + "loss": 0.9124, + "step": 8741 + }, + { + "epoch": 0.7883843621770302, + "grad_norm": 1.7168979723504065, + "learning_rate": 4.5151508240966363e-07, + "loss": 1.0133, + "step": 8742 + }, + { + "epoch": 0.7884745457005006, + "grad_norm": 1.6135179840116451, + "learning_rate": 4.511454079378445e-07, + "loss": 1.0285, + "step": 8743 + }, + { + "epoch": 0.7885647292239708, + "grad_norm": 1.660559894633666, + "learning_rate": 4.507758656228382e-07, + "loss": 0.8798, + "step": 8744 + }, + { + "epoch": 0.788654912747441, + "grad_norm": 1.4630239347011043, + "learning_rate": 4.5040645549617864e-07, + "loss": 1.0447, + "step": 8745 + }, + { + "epoch": 0.7887450962709113, + "grad_norm": 1.7495018386242438, + "learning_rate": 4.5003717758938384e-07, + "loss": 0.9001, + "step": 8746 + }, + { + "epoch": 0.7888352797943816, + "grad_norm": 1.5468817046049066, + "learning_rate": 4.4966803193396365e-07, + "loss": 0.9212, + "step": 8747 + }, + { + "epoch": 0.7889254633178518, + "grad_norm": 2.073786746223207, + "learning_rate": 4.492990185614154e-07, + "loss": 0.9888, + "step": 8748 + }, + { + "epoch": 0.7890156468413221, + "grad_norm": 1.5007645724718766, + "learning_rate": 4.489301375032255e-07, + "loss": 0.9689, + "step": 8749 + }, + { + "epoch": 0.7891058303647923, + "grad_norm": 1.6450751123411176, + "learning_rate": 4.4856138879086857e-07, + "loss": 0.9871, + "step": 8750 + }, + { + "epoch": 0.7891960138882627, + "grad_norm": 1.720488899742798, + "learning_rate": 4.481927724558092e-07, + "loss": 0.9322, + "step": 8751 + }, + { + "epoch": 0.7892861974117329, + "grad_norm": 2.5370648316312105, + "learning_rate": 4.478242885294985e-07, + "loss": 0.9476, + "step": 8752 + }, + { + "epoch": 0.7893763809352031, + "grad_norm": 1.9368734269506584, + "learning_rate": 4.474559370433779e-07, + "loss": 0.9407, + "step": 8753 + }, + { + "epoch": 0.7894665644586734, + "grad_norm": 6.103922486439272, + "learning_rate": 4.470877180288777e-07, + "loss": 0.9257, + "step": 8754 + }, + { + "epoch": 0.7895567479821437, + "grad_norm": 1.8218800749143713, + "learning_rate": 4.4671963151741574e-07, + "loss": 0.9494, + "step": 8755 + }, + { + "epoch": 0.7896469315056139, + "grad_norm": 1.9091694473031002, + "learning_rate": 4.4635167754039973e-07, + "loss": 0.9602, + "step": 8756 + }, + { + "epoch": 0.7897371150290842, + "grad_norm": 1.637336850956753, + "learning_rate": 4.459838561292253e-07, + "loss": 1.0078, + "step": 8757 + }, + { + "epoch": 0.7898272985525544, + "grad_norm": 2.0447406852482954, + "learning_rate": 4.456161673152774e-07, + "loss": 0.8982, + "step": 8758 + }, + { + "epoch": 0.7899174820760247, + "grad_norm": 1.6261249463074534, + "learning_rate": 4.4524861112992806e-07, + "loss": 0.9704, + "step": 8759 + }, + { + "epoch": 0.790007665599495, + "grad_norm": 2.8734700651745273, + "learning_rate": 4.448811876045411e-07, + "loss": 0.9824, + "step": 8760 + }, + { + "epoch": 0.7900978491229652, + "grad_norm": 1.9556088332242199, + "learning_rate": 4.445138967704647e-07, + "loss": 1.05, + "step": 8761 + }, + { + "epoch": 0.7901880326464354, + "grad_norm": 1.8840495117873657, + "learning_rate": 4.4414673865904075e-07, + "loss": 0.9517, + "step": 8762 + }, + { + "epoch": 0.7902782161699058, + "grad_norm": 1.5435006685493389, + "learning_rate": 4.437797133015955e-07, + "loss": 0.9981, + "step": 8763 + }, + { + "epoch": 0.790368399693376, + "grad_norm": 2.1275849047706785, + "learning_rate": 4.4341282072944586e-07, + "loss": 0.8948, + "step": 8764 + }, + { + "epoch": 0.7904585832168463, + "grad_norm": 1.7626727919051317, + "learning_rate": 4.430460609738973e-07, + "loss": 1.0006, + "step": 8765 + }, + { + "epoch": 0.7905487667403166, + "grad_norm": 0.7093170478440389, + "learning_rate": 4.4267943406624386e-07, + "loss": 0.847, + "step": 8766 + }, + { + "epoch": 0.7906389502637868, + "grad_norm": 1.7891996905769407, + "learning_rate": 4.4231294003776853e-07, + "loss": 1.0922, + "step": 8767 + }, + { + "epoch": 0.7907291337872571, + "grad_norm": 1.9510702081871352, + "learning_rate": 4.419465789197416e-07, + "loss": 1.0202, + "step": 8768 + }, + { + "epoch": 0.7908193173107273, + "grad_norm": 1.5304858578187766, + "learning_rate": 4.415803507434237e-07, + "loss": 0.9731, + "step": 8769 + }, + { + "epoch": 0.7909095008341976, + "grad_norm": 1.7507107517495584, + "learning_rate": 4.4121425554006307e-07, + "loss": 1.0066, + "step": 8770 + }, + { + "epoch": 0.7909996843576679, + "grad_norm": 2.112321166645728, + "learning_rate": 4.4084829334089744e-07, + "loss": 0.9631, + "step": 8771 + }, + { + "epoch": 0.7910898678811381, + "grad_norm": 1.5893907206152016, + "learning_rate": 4.404824641771525e-07, + "loss": 0.9263, + "step": 8772 + }, + { + "epoch": 0.7911800514046083, + "grad_norm": 1.9554078455732187, + "learning_rate": 4.4011676808004327e-07, + "loss": 1.0445, + "step": 8773 + }, + { + "epoch": 0.7912702349280787, + "grad_norm": 1.7266397588368758, + "learning_rate": 4.3975120508077145e-07, + "loss": 0.991, + "step": 8774 + }, + { + "epoch": 0.7913604184515489, + "grad_norm": 1.6629746870820037, + "learning_rate": 4.39385775210531e-07, + "loss": 1.0271, + "step": 8775 + }, + { + "epoch": 0.7914506019750192, + "grad_norm": 1.7623135731898065, + "learning_rate": 4.390204785005003e-07, + "loss": 0.9942, + "step": 8776 + }, + { + "epoch": 0.7915407854984894, + "grad_norm": 1.4429041411718804, + "learning_rate": 4.386553149818504e-07, + "loss": 0.9818, + "step": 8777 + }, + { + "epoch": 0.7916309690219597, + "grad_norm": 2.099469699205802, + "learning_rate": 4.3829028468573793e-07, + "loss": 0.8469, + "step": 8778 + }, + { + "epoch": 0.79172115254543, + "grad_norm": 1.7688631055274433, + "learning_rate": 4.3792538764330935e-07, + "loss": 0.9162, + "step": 8779 + }, + { + "epoch": 0.7918113360689002, + "grad_norm": 5.4173397917531005, + "learning_rate": 4.3756062388569994e-07, + "loss": 0.9541, + "step": 8780 + }, + { + "epoch": 0.7919015195923704, + "grad_norm": 0.7337199597163715, + "learning_rate": 4.3719599344403346e-07, + "loss": 0.8166, + "step": 8781 + }, + { + "epoch": 0.7919917031158408, + "grad_norm": 1.639574059440369, + "learning_rate": 4.3683149634942243e-07, + "loss": 0.9591, + "step": 8782 + }, + { + "epoch": 0.792081886639311, + "grad_norm": 1.9981950721708213, + "learning_rate": 4.364671326329663e-07, + "loss": 0.8654, + "step": 8783 + }, + { + "epoch": 0.7921720701627812, + "grad_norm": 2.112956305886598, + "learning_rate": 4.3610290232575673e-07, + "loss": 0.9967, + "step": 8784 + }, + { + "epoch": 0.7922622536862515, + "grad_norm": 2.0356657887659866, + "learning_rate": 4.357388054588702e-07, + "loss": 0.9814, + "step": 8785 + }, + { + "epoch": 0.7923524372097218, + "grad_norm": 1.4727836574010347, + "learning_rate": 4.3537484206337405e-07, + "loss": 0.9222, + "step": 8786 + }, + { + "epoch": 0.792442620733192, + "grad_norm": 1.6719493671170198, + "learning_rate": 4.3501101217032366e-07, + "loss": 1.0088, + "step": 8787 + }, + { + "epoch": 0.7925328042566623, + "grad_norm": 1.8552687366497123, + "learning_rate": 4.346473158107629e-07, + "loss": 0.9058, + "step": 8788 + }, + { + "epoch": 0.7926229877801325, + "grad_norm": 2.701702559521179, + "learning_rate": 4.342837530157244e-07, + "loss": 0.9833, + "step": 8789 + }, + { + "epoch": 0.7927131713036029, + "grad_norm": 1.5592065815766627, + "learning_rate": 4.3392032381622987e-07, + "loss": 0.8826, + "step": 8790 + }, + { + "epoch": 0.7928033548270731, + "grad_norm": 1.8709838452643337, + "learning_rate": 4.3355702824328765e-07, + "loss": 0.9404, + "step": 8791 + }, + { + "epoch": 0.7928935383505433, + "grad_norm": 1.6467507215943933, + "learning_rate": 4.3319386632789823e-07, + "loss": 1.0222, + "step": 8792 + }, + { + "epoch": 0.7929837218740137, + "grad_norm": 1.9512360250105758, + "learning_rate": 4.328308381010466e-07, + "loss": 0.9754, + "step": 8793 + }, + { + "epoch": 0.7930739053974839, + "grad_norm": 1.577800755191699, + "learning_rate": 4.3246794359370933e-07, + "loss": 0.9881, + "step": 8794 + }, + { + "epoch": 0.7931640889209541, + "grad_norm": 2.2178104184224106, + "learning_rate": 4.3210518283685025e-07, + "loss": 0.9565, + "step": 8795 + }, + { + "epoch": 0.7932542724444244, + "grad_norm": 2.0506489447649314, + "learning_rate": 4.317425558614225e-07, + "loss": 0.9775, + "step": 8796 + }, + { + "epoch": 0.7933444559678947, + "grad_norm": 1.3116803107386135, + "learning_rate": 4.3138006269836744e-07, + "loss": 0.9432, + "step": 8797 + }, + { + "epoch": 0.793434639491365, + "grad_norm": 2.2791817054973005, + "learning_rate": 4.3101770337861376e-07, + "loss": 0.9682, + "step": 8798 + }, + { + "epoch": 0.7935248230148352, + "grad_norm": 0.6885237657313972, + "learning_rate": 4.30655477933082e-07, + "loss": 0.8301, + "step": 8799 + }, + { + "epoch": 0.7936150065383054, + "grad_norm": 1.340476862407859, + "learning_rate": 4.30293386392677e-07, + "loss": 1.0045, + "step": 8800 + }, + { + "epoch": 0.7937051900617758, + "grad_norm": 1.3620842868167173, + "learning_rate": 4.299314287882967e-07, + "loss": 0.9969, + "step": 8801 + }, + { + "epoch": 0.793795373585246, + "grad_norm": 1.6165756321212708, + "learning_rate": 4.2956960515082353e-07, + "loss": 1.0084, + "step": 8802 + }, + { + "epoch": 0.7938855571087162, + "grad_norm": 1.6867248732935836, + "learning_rate": 4.29207915511131e-07, + "loss": 0.9346, + "step": 8803 + }, + { + "epoch": 0.7939757406321865, + "grad_norm": 2.1468744314045827, + "learning_rate": 4.2884635990008024e-07, + "loss": 1.0358, + "step": 8804 + }, + { + "epoch": 0.7940659241556568, + "grad_norm": 1.608158207123159, + "learning_rate": 4.284849383485214e-07, + "loss": 0.9446, + "step": 8805 + }, + { + "epoch": 0.794156107679127, + "grad_norm": 2.154719818735934, + "learning_rate": 4.2812365088729296e-07, + "loss": 1.0071, + "step": 8806 + }, + { + "epoch": 0.7942462912025973, + "grad_norm": 1.6459895031711351, + "learning_rate": 4.2776249754722227e-07, + "loss": 1.0015, + "step": 8807 + }, + { + "epoch": 0.7943364747260675, + "grad_norm": 4.607943451560924, + "learning_rate": 4.27401478359124e-07, + "loss": 1.046, + "step": 8808 + }, + { + "epoch": 0.7944266582495378, + "grad_norm": 1.800661095439809, + "learning_rate": 4.2704059335380283e-07, + "loss": 1.0786, + "step": 8809 + }, + { + "epoch": 0.7945168417730081, + "grad_norm": 1.9472219754973343, + "learning_rate": 4.266798425620515e-07, + "loss": 1.0819, + "step": 8810 + }, + { + "epoch": 0.7946070252964783, + "grad_norm": 1.5740677451129077, + "learning_rate": 4.263192260146511e-07, + "loss": 0.9633, + "step": 8811 + }, + { + "epoch": 0.7946972088199485, + "grad_norm": 1.6397028193222574, + "learning_rate": 4.2595874374237216e-07, + "loss": 1.0255, + "step": 8812 + }, + { + "epoch": 0.7947873923434189, + "grad_norm": 1.6672952261445084, + "learning_rate": 4.255983957759712e-07, + "loss": 0.9774, + "step": 8813 + }, + { + "epoch": 0.7948775758668891, + "grad_norm": 1.7141222989440703, + "learning_rate": 4.2523818214619745e-07, + "loss": 0.9163, + "step": 8814 + }, + { + "epoch": 0.7949677593903594, + "grad_norm": 1.985230569300277, + "learning_rate": 4.24878102883784e-07, + "loss": 1.0258, + "step": 8815 + }, + { + "epoch": 0.7950579429138297, + "grad_norm": 1.9805840152100445, + "learning_rate": 4.24518158019457e-07, + "loss": 0.9846, + "step": 8816 + }, + { + "epoch": 0.7951481264372999, + "grad_norm": 1.7772000097482128, + "learning_rate": 4.241583475839274e-07, + "loss": 0.919, + "step": 8817 + }, + { + "epoch": 0.7952383099607702, + "grad_norm": 1.701095842136089, + "learning_rate": 4.237986716078965e-07, + "loss": 1.0222, + "step": 8818 + }, + { + "epoch": 0.7953284934842404, + "grad_norm": 1.6493806717035364, + "learning_rate": 4.2343913012205433e-07, + "loss": 0.9691, + "step": 8819 + }, + { + "epoch": 0.7954186770077107, + "grad_norm": 1.434710989767893, + "learning_rate": 4.230797231570784e-07, + "loss": 0.9537, + "step": 8820 + }, + { + "epoch": 0.795508860531181, + "grad_norm": 2.2262551614229236, + "learning_rate": 4.227204507436357e-07, + "loss": 0.9842, + "step": 8821 + }, + { + "epoch": 0.7955990440546512, + "grad_norm": 1.464795083827004, + "learning_rate": 4.223613129123811e-07, + "loss": 0.8982, + "step": 8822 + }, + { + "epoch": 0.7956892275781214, + "grad_norm": 1.4910864637545227, + "learning_rate": 4.220023096939589e-07, + "loss": 0.8854, + "step": 8823 + }, + { + "epoch": 0.7957794111015918, + "grad_norm": 2.6208844392662636, + "learning_rate": 4.21643441119e-07, + "loss": 0.9727, + "step": 8824 + }, + { + "epoch": 0.795869594625062, + "grad_norm": 0.6300364698722929, + "learning_rate": 4.212847072181256e-07, + "loss": 0.8663, + "step": 8825 + }, + { + "epoch": 0.7959597781485322, + "grad_norm": 1.560436946735413, + "learning_rate": 4.2092610802194505e-07, + "loss": 0.9524, + "step": 8826 + }, + { + "epoch": 0.7960499616720025, + "grad_norm": 1.4290127855800725, + "learning_rate": 4.2056764356105587e-07, + "loss": 0.8877, + "step": 8827 + }, + { + "epoch": 0.7961401451954728, + "grad_norm": 1.6329628177060922, + "learning_rate": 4.202093138660443e-07, + "loss": 1.0259, + "step": 8828 + }, + { + "epoch": 0.7962303287189431, + "grad_norm": 1.4955301443293132, + "learning_rate": 4.198511189674854e-07, + "loss": 1.0114, + "step": 8829 + }, + { + "epoch": 0.7963205122424133, + "grad_norm": 3.2574319228394617, + "learning_rate": 4.1949305889594066e-07, + "loss": 0.9981, + "step": 8830 + }, + { + "epoch": 0.7964106957658835, + "grad_norm": 1.393623248781525, + "learning_rate": 4.191351336819642e-07, + "loss": 0.9699, + "step": 8831 + }, + { + "epoch": 0.7965008792893539, + "grad_norm": 1.6531193493348089, + "learning_rate": 4.187773433560939e-07, + "loss": 0.9162, + "step": 8832 + }, + { + "epoch": 0.7965910628128241, + "grad_norm": 1.9043125875754012, + "learning_rate": 4.184196879488604e-07, + "loss": 0.9713, + "step": 8833 + }, + { + "epoch": 0.7966812463362943, + "grad_norm": 5.862479320888203, + "learning_rate": 4.1806216749077936e-07, + "loss": 1.0424, + "step": 8834 + }, + { + "epoch": 0.7967714298597646, + "grad_norm": 1.3810033655698788, + "learning_rate": 4.177047820123569e-07, + "loss": 1.0212, + "step": 8835 + }, + { + "epoch": 0.7968616133832349, + "grad_norm": 1.7442418461013485, + "learning_rate": 4.1734753154408733e-07, + "loss": 0.8498, + "step": 8836 + }, + { + "epoch": 0.7969517969067051, + "grad_norm": 1.7710142005855112, + "learning_rate": 4.169904161164528e-07, + "loss": 0.9622, + "step": 8837 + }, + { + "epoch": 0.7970419804301754, + "grad_norm": 2.5029795057895674, + "learning_rate": 4.1663343575992526e-07, + "loss": 1.0991, + "step": 8838 + }, + { + "epoch": 0.7971321639536457, + "grad_norm": 1.896116116851171, + "learning_rate": 4.1627659050496275e-07, + "loss": 0.9969, + "step": 8839 + }, + { + "epoch": 0.797222347477116, + "grad_norm": 1.709346780770699, + "learning_rate": 4.1591988038201453e-07, + "loss": 0.8944, + "step": 8840 + }, + { + "epoch": 0.7973125310005862, + "grad_norm": 1.6312354836238228, + "learning_rate": 4.155633054215164e-07, + "loss": 1.024, + "step": 8841 + }, + { + "epoch": 0.7974027145240564, + "grad_norm": 5.939538545505444, + "learning_rate": 4.152068656538934e-07, + "loss": 0.8883, + "step": 8842 + }, + { + "epoch": 0.7974928980475268, + "grad_norm": 1.4295903574141777, + "learning_rate": 4.148505611095594e-07, + "loss": 1.0368, + "step": 8843 + }, + { + "epoch": 0.797583081570997, + "grad_norm": 1.926867258141104, + "learning_rate": 4.1449439181891563e-07, + "loss": 0.8547, + "step": 8844 + }, + { + "epoch": 0.7976732650944672, + "grad_norm": 1.7027753418529374, + "learning_rate": 4.14138357812353e-07, + "loss": 0.9971, + "step": 8845 + }, + { + "epoch": 0.7977634486179375, + "grad_norm": 1.4525077082943205, + "learning_rate": 4.137824591202506e-07, + "loss": 1.0327, + "step": 8846 + }, + { + "epoch": 0.7978536321414078, + "grad_norm": 1.906571720678107, + "learning_rate": 4.134266957729737e-07, + "loss": 0.9642, + "step": 8847 + }, + { + "epoch": 0.797943815664878, + "grad_norm": 2.464579379115368, + "learning_rate": 4.1307106780088065e-07, + "loss": 1.0711, + "step": 8848 + }, + { + "epoch": 0.7980339991883483, + "grad_norm": 1.8957066615897236, + "learning_rate": 4.1271557523431387e-07, + "loss": 0.9888, + "step": 8849 + }, + { + "epoch": 0.7981241827118185, + "grad_norm": 1.8169572548213067, + "learning_rate": 4.1236021810360634e-07, + "loss": 0.9615, + "step": 8850 + }, + { + "epoch": 0.7982143662352889, + "grad_norm": 2.3648223644104993, + "learning_rate": 4.120049964390793e-07, + "loss": 0.9202, + "step": 8851 + }, + { + "epoch": 0.7983045497587591, + "grad_norm": 2.6685516926971227, + "learning_rate": 4.116499102710418e-07, + "loss": 0.9276, + "step": 8852 + }, + { + "epoch": 0.7983947332822293, + "grad_norm": 2.688043583743662, + "learning_rate": 4.112949596297928e-07, + "loss": 0.9549, + "step": 8853 + }, + { + "epoch": 0.7984849168056996, + "grad_norm": 2.026473547660486, + "learning_rate": 4.1094014454561664e-07, + "loss": 0.9228, + "step": 8854 + }, + { + "epoch": 0.7985751003291699, + "grad_norm": 2.4069122790211086, + "learning_rate": 4.1058546504879057e-07, + "loss": 0.8837, + "step": 8855 + }, + { + "epoch": 0.7986652838526401, + "grad_norm": 1.98642300629538, + "learning_rate": 4.1023092116957583e-07, + "loss": 1.0416, + "step": 8856 + }, + { + "epoch": 0.7987554673761104, + "grad_norm": 2.3149607170183018, + "learning_rate": 4.098765129382249e-07, + "loss": 1.0189, + "step": 8857 + }, + { + "epoch": 0.7988456508995806, + "grad_norm": 2.289616123990932, + "learning_rate": 4.0952224038497764e-07, + "loss": 1.0498, + "step": 8858 + }, + { + "epoch": 0.7989358344230509, + "grad_norm": 2.210670266328342, + "learning_rate": 4.091681035400627e-07, + "loss": 0.9759, + "step": 8859 + }, + { + "epoch": 0.7990260179465212, + "grad_norm": 1.5937624912146446, + "learning_rate": 4.088141024336971e-07, + "loss": 1.0122, + "step": 8860 + }, + { + "epoch": 0.7991162014699914, + "grad_norm": 1.583994384432963, + "learning_rate": 4.0846023709608636e-07, + "loss": 1.0584, + "step": 8861 + }, + { + "epoch": 0.7992063849934617, + "grad_norm": 2.0205816787337674, + "learning_rate": 4.081065075574226e-07, + "loss": 1.0542, + "step": 8862 + }, + { + "epoch": 0.799296568516932, + "grad_norm": 1.6431373614879712, + "learning_rate": 4.077529138478906e-07, + "loss": 0.9075, + "step": 8863 + }, + { + "epoch": 0.7993867520404022, + "grad_norm": 0.6877439022910152, + "learning_rate": 4.073994559976588e-07, + "loss": 0.7782, + "step": 8864 + }, + { + "epoch": 0.7994769355638724, + "grad_norm": 3.220109967077715, + "learning_rate": 4.0704613403688716e-07, + "loss": 0.9423, + "step": 8865 + }, + { + "epoch": 0.7995671190873428, + "grad_norm": 1.6715053256603574, + "learning_rate": 4.0669294799572264e-07, + "loss": 1.0337, + "step": 8866 + }, + { + "epoch": 0.799657302610813, + "grad_norm": 1.7700377827722793, + "learning_rate": 4.0633989790430113e-07, + "loss": 1.079, + "step": 8867 + }, + { + "epoch": 0.7997474861342833, + "grad_norm": 2.0549080733925775, + "learning_rate": 4.059869837927477e-07, + "loss": 0.9895, + "step": 8868 + }, + { + "epoch": 0.7998376696577535, + "grad_norm": 2.067312002433419, + "learning_rate": 4.056342056911728e-07, + "loss": 0.9624, + "step": 8869 + }, + { + "epoch": 0.7999278531812238, + "grad_norm": 2.0279858698681825, + "learning_rate": 4.052815636296798e-07, + "loss": 0.9953, + "step": 8870 + }, + { + "epoch": 0.8000180367046941, + "grad_norm": 1.4214263879029425, + "learning_rate": 4.0492905763835593e-07, + "loss": 1.0118, + "step": 8871 + }, + { + "epoch": 0.8001082202281643, + "grad_norm": 1.437991265648431, + "learning_rate": 4.0457668774728115e-07, + "loss": 1.0108, + "step": 8872 + }, + { + "epoch": 0.8001984037516345, + "grad_norm": 1.5176654726712593, + "learning_rate": 4.0422445398651985e-07, + "loss": 0.9956, + "step": 8873 + }, + { + "epoch": 0.8002885872751049, + "grad_norm": 7.394595218952045, + "learning_rate": 4.0387235638612706e-07, + "loss": 0.9006, + "step": 8874 + }, + { + "epoch": 0.8003787707985751, + "grad_norm": 1.76512429218482, + "learning_rate": 4.0352039497614586e-07, + "loss": 0.9835, + "step": 8875 + }, + { + "epoch": 0.8004689543220453, + "grad_norm": 2.354590915048256, + "learning_rate": 4.031685697866074e-07, + "loss": 0.97, + "step": 8876 + }, + { + "epoch": 0.8005591378455156, + "grad_norm": 2.1291568713575617, + "learning_rate": 4.0281688084753165e-07, + "loss": 1.0167, + "step": 8877 + }, + { + "epoch": 0.8006493213689859, + "grad_norm": 1.6893932883189715, + "learning_rate": 4.0246532818892675e-07, + "loss": 1.0307, + "step": 8878 + }, + { + "epoch": 0.8007395048924562, + "grad_norm": 1.8188501468103753, + "learning_rate": 4.0211391184078814e-07, + "loss": 1.0161, + "step": 8879 + }, + { + "epoch": 0.8008296884159264, + "grad_norm": 1.8363706037310215, + "learning_rate": 4.0176263183310135e-07, + "loss": 1.0129, + "step": 8880 + }, + { + "epoch": 0.8009198719393966, + "grad_norm": 1.463606818254743, + "learning_rate": 4.0141148819583925e-07, + "loss": 1.0369, + "step": 8881 + }, + { + "epoch": 0.801010055462867, + "grad_norm": 0.6097075948527167, + "learning_rate": 4.010604809589637e-07, + "loss": 0.8068, + "step": 8882 + }, + { + "epoch": 0.8011002389863372, + "grad_norm": 2.8476122105755284, + "learning_rate": 4.0070961015242475e-07, + "loss": 0.9848, + "step": 8883 + }, + { + "epoch": 0.8011904225098074, + "grad_norm": 1.396561499113218, + "learning_rate": 4.0035887580615933e-07, + "loss": 1.0499, + "step": 8884 + }, + { + "epoch": 0.8012806060332777, + "grad_norm": 1.642965554575731, + "learning_rate": 4.0000827795009594e-07, + "loss": 1.03, + "step": 8885 + }, + { + "epoch": 0.801370789556748, + "grad_norm": 2.0017237387638103, + "learning_rate": 3.996578166141475e-07, + "loss": 0.9413, + "step": 8886 + }, + { + "epoch": 0.8014609730802182, + "grad_norm": 2.21602538543924, + "learning_rate": 3.9930749182821955e-07, + "loss": 0.9893, + "step": 8887 + }, + { + "epoch": 0.8015511566036885, + "grad_norm": 0.6013310470202783, + "learning_rate": 3.9895730362220116e-07, + "loss": 0.8319, + "step": 8888 + }, + { + "epoch": 0.8016413401271588, + "grad_norm": 3.0207292911546135, + "learning_rate": 3.986072520259749e-07, + "loss": 1.0199, + "step": 8889 + }, + { + "epoch": 0.801731523650629, + "grad_norm": 1.944830409602091, + "learning_rate": 3.9825733706940736e-07, + "loss": 0.9693, + "step": 8890 + }, + { + "epoch": 0.8018217071740993, + "grad_norm": 1.465208532054115, + "learning_rate": 3.979075587823557e-07, + "loss": 0.9974, + "step": 8891 + }, + { + "epoch": 0.8019118906975695, + "grad_norm": 2.0099957541046085, + "learning_rate": 3.9755791719466504e-07, + "loss": 0.861, + "step": 8892 + }, + { + "epoch": 0.8020020742210399, + "grad_norm": 1.9480538413422537, + "learning_rate": 3.9720841233616875e-07, + "loss": 0.9399, + "step": 8893 + }, + { + "epoch": 0.8020922577445101, + "grad_norm": 1.6916391683639254, + "learning_rate": 3.968590442366888e-07, + "loss": 0.9553, + "step": 8894 + }, + { + "epoch": 0.8021824412679803, + "grad_norm": 1.512138525273627, + "learning_rate": 3.9650981292603423e-07, + "loss": 0.9577, + "step": 8895 + }, + { + "epoch": 0.8022726247914506, + "grad_norm": 2.0490658515785247, + "learning_rate": 3.961607184340041e-07, + "loss": 0.9566, + "step": 8896 + }, + { + "epoch": 0.8023628083149209, + "grad_norm": 1.3350154081448389, + "learning_rate": 3.9581176079038505e-07, + "loss": 0.879, + "step": 8897 + }, + { + "epoch": 0.8024529918383911, + "grad_norm": 1.9973149156820529, + "learning_rate": 3.954629400249516e-07, + "loss": 0.9247, + "step": 8898 + }, + { + "epoch": 0.8025431753618614, + "grad_norm": 1.4474320000733802, + "learning_rate": 3.9511425616746787e-07, + "loss": 1.0471, + "step": 8899 + }, + { + "epoch": 0.8026333588853316, + "grad_norm": 1.6033799442777998, + "learning_rate": 3.947657092476853e-07, + "loss": 1.0458, + "step": 8900 + }, + { + "epoch": 0.802723542408802, + "grad_norm": 2.8282606397948102, + "learning_rate": 3.944172992953425e-07, + "loss": 0.9933, + "step": 8901 + }, + { + "epoch": 0.8028137259322722, + "grad_norm": 1.7252819439006795, + "learning_rate": 3.9406902634017e-07, + "loss": 0.9346, + "step": 8902 + }, + { + "epoch": 0.8029039094557424, + "grad_norm": 1.4964599957891265, + "learning_rate": 3.9372089041188227e-07, + "loss": 1.0132, + "step": 8903 + }, + { + "epoch": 0.8029940929792126, + "grad_norm": 4.38235988081764, + "learning_rate": 3.9337289154018593e-07, + "loss": 0.8957, + "step": 8904 + }, + { + "epoch": 0.803084276502683, + "grad_norm": 1.482660528179424, + "learning_rate": 3.930250297547728e-07, + "loss": 0.9815, + "step": 8905 + }, + { + "epoch": 0.8031744600261532, + "grad_norm": 1.6780264514706933, + "learning_rate": 3.9267730508532513e-07, + "loss": 0.8612, + "step": 8906 + }, + { + "epoch": 0.8032646435496235, + "grad_norm": 1.4331714309510863, + "learning_rate": 3.923297175615121e-07, + "loss": 0.9689, + "step": 8907 + }, + { + "epoch": 0.8033548270730937, + "grad_norm": 1.4893048788669152, + "learning_rate": 3.9198226721299243e-07, + "loss": 1.0224, + "step": 8908 + }, + { + "epoch": 0.803445010596564, + "grad_norm": 1.9230031989711234, + "learning_rate": 3.916349540694128e-07, + "loss": 0.9609, + "step": 8909 + }, + { + "epoch": 0.8035351941200343, + "grad_norm": 1.8939488095714143, + "learning_rate": 3.912877781604063e-07, + "loss": 0.9895, + "step": 8910 + }, + { + "epoch": 0.8036253776435045, + "grad_norm": 1.6753346663967, + "learning_rate": 3.909407395155977e-07, + "loss": 1.0041, + "step": 8911 + }, + { + "epoch": 0.8037155611669748, + "grad_norm": 1.6124745862274072, + "learning_rate": 3.9059383816459725e-07, + "loss": 0.9222, + "step": 8912 + }, + { + "epoch": 0.8038057446904451, + "grad_norm": 2.888317610016774, + "learning_rate": 3.902470741370045e-07, + "loss": 1.0437, + "step": 8913 + }, + { + "epoch": 0.8038959282139153, + "grad_norm": 1.478363311460758, + "learning_rate": 3.8990044746240746e-07, + "loss": 0.9718, + "step": 8914 + }, + { + "epoch": 0.8039861117373855, + "grad_norm": 3.037502486616459, + "learning_rate": 3.8955395817038237e-07, + "loss": 1.08, + "step": 8915 + }, + { + "epoch": 0.8040762952608559, + "grad_norm": 1.6011003525285938, + "learning_rate": 3.892076062904934e-07, + "loss": 0.9494, + "step": 8916 + }, + { + "epoch": 0.8041664787843261, + "grad_norm": 1.738002386120035, + "learning_rate": 3.8886139185229384e-07, + "loss": 0.9382, + "step": 8917 + }, + { + "epoch": 0.8042566623077964, + "grad_norm": 2.9241992865201847, + "learning_rate": 3.8851531488532284e-07, + "loss": 0.8989, + "step": 8918 + }, + { + "epoch": 0.8043468458312666, + "grad_norm": 1.5141659035030801, + "learning_rate": 3.88169375419112e-07, + "loss": 1.005, + "step": 8919 + }, + { + "epoch": 0.8044370293547369, + "grad_norm": 1.7413412185512807, + "learning_rate": 3.8782357348317717e-07, + "loss": 0.9555, + "step": 8920 + }, + { + "epoch": 0.8045272128782072, + "grad_norm": 1.8715864261582562, + "learning_rate": 3.8747790910702437e-07, + "loss": 1.0231, + "step": 8921 + }, + { + "epoch": 0.8046173964016774, + "grad_norm": 1.701817281684993, + "learning_rate": 3.8713238232014776e-07, + "loss": 0.982, + "step": 8922 + }, + { + "epoch": 0.8047075799251476, + "grad_norm": 1.6715111737689412, + "learning_rate": 3.867869931520296e-07, + "loss": 1.0416, + "step": 8923 + }, + { + "epoch": 0.804797763448618, + "grad_norm": 1.8186354873106283, + "learning_rate": 3.864417416321406e-07, + "loss": 0.9949, + "step": 8924 + }, + { + "epoch": 0.8048879469720882, + "grad_norm": 1.938667683824555, + "learning_rate": 3.8609662778993847e-07, + "loss": 0.9669, + "step": 8925 + }, + { + "epoch": 0.8049781304955584, + "grad_norm": 0.5775166739083787, + "learning_rate": 3.85751651654872e-07, + "loss": 0.7944, + "step": 8926 + }, + { + "epoch": 0.8050683140190287, + "grad_norm": 1.733567745912801, + "learning_rate": 3.8540681325637505e-07, + "loss": 1.0937, + "step": 8927 + }, + { + "epoch": 0.805158497542499, + "grad_norm": 1.6438683557869946, + "learning_rate": 3.8506211262387155e-07, + "loss": 0.9926, + "step": 8928 + }, + { + "epoch": 0.8052486810659693, + "grad_norm": 2.7336525970954035, + "learning_rate": 3.847175497867732e-07, + "loss": 0.9891, + "step": 8929 + }, + { + "epoch": 0.8053388645894395, + "grad_norm": 1.7817434330178672, + "learning_rate": 3.843731247744801e-07, + "loss": 0.9317, + "step": 8930 + }, + { + "epoch": 0.8054290481129097, + "grad_norm": 1.7250756122768967, + "learning_rate": 3.8402883761638047e-07, + "loss": 1.026, + "step": 8931 + }, + { + "epoch": 0.8055192316363801, + "grad_norm": 0.662682344426125, + "learning_rate": 3.8368468834185076e-07, + "loss": 0.8508, + "step": 8932 + }, + { + "epoch": 0.8056094151598503, + "grad_norm": 2.81985114807217, + "learning_rate": 3.8334067698025583e-07, + "loss": 1.0317, + "step": 8933 + }, + { + "epoch": 0.8056995986833205, + "grad_norm": 1.7137265400355497, + "learning_rate": 3.8299680356094897e-07, + "loss": 0.9194, + "step": 8934 + }, + { + "epoch": 0.8057897822067909, + "grad_norm": 2.6679916269150965, + "learning_rate": 3.8265306811327024e-07, + "loss": 0.8695, + "step": 8935 + }, + { + "epoch": 0.8058799657302611, + "grad_norm": 2.604348189067407, + "learning_rate": 3.8230947066654994e-07, + "loss": 0.9535, + "step": 8936 + }, + { + "epoch": 0.8059701492537313, + "grad_norm": 2.1220233488964535, + "learning_rate": 3.819660112501053e-07, + "loss": 1.0424, + "step": 8937 + }, + { + "epoch": 0.8060603327772016, + "grad_norm": 0.6169173699759323, + "learning_rate": 3.816226898932422e-07, + "loss": 0.8345, + "step": 8938 + }, + { + "epoch": 0.8061505163006719, + "grad_norm": 2.7253615568190144, + "learning_rate": 3.812795066252557e-07, + "loss": 0.9745, + "step": 8939 + }, + { + "epoch": 0.8062406998241421, + "grad_norm": 2.294778606731454, + "learning_rate": 3.8093646147542577e-07, + "loss": 0.9936, + "step": 8940 + }, + { + "epoch": 0.8063308833476124, + "grad_norm": 2.197645908932133, + "learning_rate": 3.805935544730259e-07, + "loss": 0.9543, + "step": 8941 + }, + { + "epoch": 0.8064210668710826, + "grad_norm": 1.4232378286768972, + "learning_rate": 3.802507856473118e-07, + "loss": 0.9204, + "step": 8942 + }, + { + "epoch": 0.806511250394553, + "grad_norm": 2.203705515031162, + "learning_rate": 3.7990815502753317e-07, + "loss": 0.9883, + "step": 8943 + }, + { + "epoch": 0.8066014339180232, + "grad_norm": 1.5627250509313368, + "learning_rate": 3.795656626429231e-07, + "loss": 0.989, + "step": 8944 + }, + { + "epoch": 0.8066916174414934, + "grad_norm": 1.7894464605507516, + "learning_rate": 3.792233085227059e-07, + "loss": 0.9023, + "step": 8945 + }, + { + "epoch": 0.8067818009649637, + "grad_norm": 1.834422683745925, + "learning_rate": 3.788810926960928e-07, + "loss": 0.9934, + "step": 8946 + }, + { + "epoch": 0.806871984488434, + "grad_norm": 1.5692882328818145, + "learning_rate": 3.785390151922836e-07, + "loss": 0.9687, + "step": 8947 + }, + { + "epoch": 0.8069621680119042, + "grad_norm": 3.450623724021185, + "learning_rate": 3.781970760404665e-07, + "loss": 0.969, + "step": 8948 + }, + { + "epoch": 0.8070523515353745, + "grad_norm": 1.5163613329941208, + "learning_rate": 3.778552752698176e-07, + "loss": 1.0601, + "step": 8949 + }, + { + "epoch": 0.8071425350588447, + "grad_norm": 6.547156717070461, + "learning_rate": 3.775136129095007e-07, + "loss": 0.8946, + "step": 8950 + }, + { + "epoch": 0.807232718582315, + "grad_norm": 2.26918507757527, + "learning_rate": 3.771720889886685e-07, + "loss": 0.9565, + "step": 8951 + }, + { + "epoch": 0.8073229021057853, + "grad_norm": 0.6963789401691849, + "learning_rate": 3.7683070353646194e-07, + "loss": 0.8325, + "step": 8952 + }, + { + "epoch": 0.8074130856292555, + "grad_norm": 1.6872415697930294, + "learning_rate": 3.7648945658200983e-07, + "loss": 0.9449, + "step": 8953 + }, + { + "epoch": 0.8075032691527257, + "grad_norm": 1.892423042738269, + "learning_rate": 3.761483481544292e-07, + "loss": 0.8707, + "step": 8954 + }, + { + "epoch": 0.8075934526761961, + "grad_norm": 1.8782047223072076, + "learning_rate": 3.7580737828282525e-07, + "loss": 0.951, + "step": 8955 + }, + { + "epoch": 0.8076836361996663, + "grad_norm": 1.5970152522716117, + "learning_rate": 3.754665469962921e-07, + "loss": 0.9621, + "step": 8956 + }, + { + "epoch": 0.8077738197231366, + "grad_norm": 1.95178530713866, + "learning_rate": 3.7512585432390973e-07, + "loss": 0.9385, + "step": 8957 + }, + { + "epoch": 0.8078640032466069, + "grad_norm": 2.4363641293109826, + "learning_rate": 3.7478530029474987e-07, + "loss": 0.9213, + "step": 8958 + }, + { + "epoch": 0.8079541867700771, + "grad_norm": 0.6227906277151645, + "learning_rate": 3.7444488493786854e-07, + "loss": 0.8653, + "step": 8959 + }, + { + "epoch": 0.8080443702935474, + "grad_norm": 1.7569487081215465, + "learning_rate": 3.7410460828231405e-07, + "loss": 1.0338, + "step": 8960 + }, + { + "epoch": 0.8081345538170176, + "grad_norm": 1.3354288214707142, + "learning_rate": 3.737644703571188e-07, + "loss": 0.9176, + "step": 8961 + }, + { + "epoch": 0.8082247373404879, + "grad_norm": 2.261254033917508, + "learning_rate": 3.734244711913059e-07, + "loss": 1.0509, + "step": 8962 + }, + { + "epoch": 0.8083149208639582, + "grad_norm": 1.7881204048639032, + "learning_rate": 3.7308461081388584e-07, + "loss": 0.9609, + "step": 8963 + }, + { + "epoch": 0.8084051043874284, + "grad_norm": 1.6673641493790496, + "learning_rate": 3.727448892538576e-07, + "loss": 0.9125, + "step": 8964 + }, + { + "epoch": 0.8084952879108986, + "grad_norm": 1.5812876519707433, + "learning_rate": 3.724053065402086e-07, + "loss": 1.0651, + "step": 8965 + }, + { + "epoch": 0.808585471434369, + "grad_norm": 1.4740457228061055, + "learning_rate": 3.7206586270191285e-07, + "loss": 0.9492, + "step": 8966 + }, + { + "epoch": 0.8086756549578392, + "grad_norm": 1.5549457278087437, + "learning_rate": 3.7172655776793385e-07, + "loss": 1.03, + "step": 8967 + }, + { + "epoch": 0.8087658384813095, + "grad_norm": 2.0475375435429792, + "learning_rate": 3.7138739176722323e-07, + "loss": 0.8723, + "step": 8968 + }, + { + "epoch": 0.8088560220047797, + "grad_norm": 1.6798712341597923, + "learning_rate": 3.710483647287206e-07, + "loss": 0.9501, + "step": 8969 + }, + { + "epoch": 0.80894620552825, + "grad_norm": 1.8210289327050186, + "learning_rate": 3.707094766813532e-07, + "loss": 0.8832, + "step": 8970 + }, + { + "epoch": 0.8090363890517203, + "grad_norm": 1.498411211373891, + "learning_rate": 3.7037072765403754e-07, + "loss": 0.9451, + "step": 8971 + }, + { + "epoch": 0.8091265725751905, + "grad_norm": 2.0810262748822965, + "learning_rate": 3.700321176756762e-07, + "loss": 0.9381, + "step": 8972 + }, + { + "epoch": 0.8092167560986607, + "grad_norm": 1.6765062591135542, + "learning_rate": 3.69693646775163e-07, + "loss": 1.0316, + "step": 8973 + }, + { + "epoch": 0.8093069396221311, + "grad_norm": 1.9894661660837114, + "learning_rate": 3.693553149813764e-07, + "loss": 0.9633, + "step": 8974 + }, + { + "epoch": 0.8093971231456013, + "grad_norm": 0.5996325609526874, + "learning_rate": 3.690171223231866e-07, + "loss": 0.7884, + "step": 8975 + }, + { + "epoch": 0.8094873066690715, + "grad_norm": 1.325063700314346, + "learning_rate": 3.6867906882944854e-07, + "loss": 1.0737, + "step": 8976 + }, + { + "epoch": 0.8095774901925418, + "grad_norm": 1.6249467400846012, + "learning_rate": 3.6834115452900737e-07, + "loss": 0.981, + "step": 8977 + }, + { + "epoch": 0.8096676737160121, + "grad_norm": 1.369443895216482, + "learning_rate": 3.680033794506958e-07, + "loss": 0.8801, + "step": 8978 + }, + { + "epoch": 0.8097578572394823, + "grad_norm": 2.3128304116584166, + "learning_rate": 3.676657436233346e-07, + "loss": 0.8076, + "step": 8979 + }, + { + "epoch": 0.8098480407629526, + "grad_norm": 1.894304399574921, + "learning_rate": 3.6732824707573305e-07, + "loss": 0.9776, + "step": 8980 + }, + { + "epoch": 0.8099382242864229, + "grad_norm": 1.834699367458172, + "learning_rate": 3.6699088983668716e-07, + "loss": 1.0154, + "step": 8981 + }, + { + "epoch": 0.8100284078098932, + "grad_norm": 1.8649287267979913, + "learning_rate": 3.6665367193498376e-07, + "loss": 0.9336, + "step": 8982 + }, + { + "epoch": 0.8101185913333634, + "grad_norm": 4.104327100274164, + "learning_rate": 3.663165933993948e-07, + "loss": 0.9844, + "step": 8983 + }, + { + "epoch": 0.8102087748568336, + "grad_norm": 6.704118846140321, + "learning_rate": 3.659796542586822e-07, + "loss": 1.0021, + "step": 8984 + }, + { + "epoch": 0.810298958380304, + "grad_norm": 1.4344838884365632, + "learning_rate": 3.6564285454159526e-07, + "loss": 0.9363, + "step": 8985 + }, + { + "epoch": 0.8103891419037742, + "grad_norm": 5.336050454057025, + "learning_rate": 3.653061942768718e-07, + "loss": 1.0239, + "step": 8986 + }, + { + "epoch": 0.8104793254272444, + "grad_norm": 1.6603455200013564, + "learning_rate": 3.649696734932375e-07, + "loss": 0.9322, + "step": 8987 + }, + { + "epoch": 0.8105695089507147, + "grad_norm": 1.821918155950764, + "learning_rate": 3.646332922194064e-07, + "loss": 0.9168, + "step": 8988 + }, + { + "epoch": 0.810659692474185, + "grad_norm": 1.978347756135853, + "learning_rate": 3.6429705048407943e-07, + "loss": 0.9652, + "step": 8989 + }, + { + "epoch": 0.8107498759976552, + "grad_norm": 2.0014302384976483, + "learning_rate": 3.6396094831594804e-07, + "loss": 0.9716, + "step": 8990 + }, + { + "epoch": 0.8108400595211255, + "grad_norm": 1.841519250910508, + "learning_rate": 3.6362498574368926e-07, + "loss": 0.929, + "step": 8991 + }, + { + "epoch": 0.8109302430445957, + "grad_norm": 0.6972667576542906, + "learning_rate": 3.6328916279596935e-07, + "loss": 0.8595, + "step": 8992 + }, + { + "epoch": 0.811020426568066, + "grad_norm": 1.6830398816819254, + "learning_rate": 3.6295347950144305e-07, + "loss": 1.0236, + "step": 8993 + }, + { + "epoch": 0.8111106100915363, + "grad_norm": 2.2237575062136723, + "learning_rate": 3.626179358887522e-07, + "loss": 0.8287, + "step": 8994 + }, + { + "epoch": 0.8112007936150065, + "grad_norm": 1.858487342075021, + "learning_rate": 3.6228253198652816e-07, + "loss": 1.0368, + "step": 8995 + }, + { + "epoch": 0.8112909771384768, + "grad_norm": 2.021895362244416, + "learning_rate": 3.6194726782338767e-07, + "loss": 0.9304, + "step": 8996 + }, + { + "epoch": 0.8113811606619471, + "grad_norm": 2.0918891950478424, + "learning_rate": 3.6161214342793953e-07, + "loss": 0.9547, + "step": 8997 + }, + { + "epoch": 0.8114713441854173, + "grad_norm": 9.150100674101427, + "learning_rate": 3.612771588287764e-07, + "loss": 1.004, + "step": 8998 + }, + { + "epoch": 0.8115615277088876, + "grad_norm": 9.469260022619327, + "learning_rate": 3.609423140544827e-07, + "loss": 1.0464, + "step": 8999 + }, + { + "epoch": 0.8116517112323578, + "grad_norm": 2.069102037792687, + "learning_rate": 3.6060760913362787e-07, + "loss": 0.9961, + "step": 9000 + }, + { + "epoch": 0.8117418947558281, + "grad_norm": 1.6540640314708763, + "learning_rate": 3.6027304409477146e-07, + "loss": 0.9878, + "step": 9001 + }, + { + "epoch": 0.8118320782792984, + "grad_norm": 4.233941650738171, + "learning_rate": 3.599386189664604e-07, + "loss": 0.9497, + "step": 9002 + }, + { + "epoch": 0.8119222618027686, + "grad_norm": 1.8385422430136609, + "learning_rate": 3.5960433377722945e-07, + "loss": 0.9667, + "step": 9003 + }, + { + "epoch": 0.8120124453262388, + "grad_norm": 1.7500254084241191, + "learning_rate": 3.5927018855560174e-07, + "loss": 1.0354, + "step": 9004 + }, + { + "epoch": 0.8121026288497092, + "grad_norm": 1.6726900057212837, + "learning_rate": 3.5893618333008904e-07, + "loss": 0.9713, + "step": 9005 + }, + { + "epoch": 0.8121928123731794, + "grad_norm": 1.8921372851915272, + "learning_rate": 3.586023181291893e-07, + "loss": 1.0693, + "step": 9006 + }, + { + "epoch": 0.8122829958966497, + "grad_norm": 0.7471941278116622, + "learning_rate": 3.5826859298139044e-07, + "loss": 0.8472, + "step": 9007 + }, + { + "epoch": 0.81237317942012, + "grad_norm": 1.5713347338860515, + "learning_rate": 3.5793500791516773e-07, + "loss": 0.9607, + "step": 9008 + }, + { + "epoch": 0.8124633629435902, + "grad_norm": 2.1570475112429204, + "learning_rate": 3.5760156295898415e-07, + "loss": 0.9536, + "step": 9009 + }, + { + "epoch": 0.8125535464670605, + "grad_norm": 3.812027948891493, + "learning_rate": 3.5726825814129203e-07, + "loss": 0.9769, + "step": 9010 + }, + { + "epoch": 0.8126437299905307, + "grad_norm": 0.7003233937442217, + "learning_rate": 3.5693509349052886e-07, + "loss": 0.8501, + "step": 9011 + }, + { + "epoch": 0.812733913514001, + "grad_norm": 1.6563265081044816, + "learning_rate": 3.5660206903512433e-07, + "loss": 1.0501, + "step": 9012 + }, + { + "epoch": 0.8128240970374713, + "grad_norm": 2.1397774682254673, + "learning_rate": 3.56269184803492e-07, + "loss": 1.0304, + "step": 9013 + }, + { + "epoch": 0.8129142805609415, + "grad_norm": 1.6586150791056364, + "learning_rate": 3.5593644082403727e-07, + "loss": 1.0015, + "step": 9014 + }, + { + "epoch": 0.8130044640844117, + "grad_norm": 1.8397362550081284, + "learning_rate": 3.5560383712514994e-07, + "loss": 1.0783, + "step": 9015 + }, + { + "epoch": 0.8130946476078821, + "grad_norm": 1.6979363875228715, + "learning_rate": 3.5527137373521066e-07, + "loss": 0.9829, + "step": 9016 + }, + { + "epoch": 0.8131848311313523, + "grad_norm": 1.524624953506793, + "learning_rate": 3.5493905068258645e-07, + "loss": 0.9358, + "step": 9017 + }, + { + "epoch": 0.8132750146548225, + "grad_norm": 1.3902419773632444, + "learning_rate": 3.546068679956333e-07, + "loss": 0.9935, + "step": 9018 + }, + { + "epoch": 0.8133651981782928, + "grad_norm": 1.7798224634935913, + "learning_rate": 3.5427482570269487e-07, + "loss": 1.0437, + "step": 9019 + }, + { + "epoch": 0.8134553817017631, + "grad_norm": 2.648332801932427, + "learning_rate": 3.539429238321026e-07, + "loss": 0.83, + "step": 9020 + }, + { + "epoch": 0.8135455652252334, + "grad_norm": 1.3882637121566517, + "learning_rate": 3.536111624121769e-07, + "loss": 0.8311, + "step": 9021 + }, + { + "epoch": 0.8136357487487036, + "grad_norm": 1.6075796093961459, + "learning_rate": 3.532795414712244e-07, + "loss": 0.9392, + "step": 9022 + }, + { + "epoch": 0.8137259322721738, + "grad_norm": 1.748019596505035, + "learning_rate": 3.5294806103754124e-07, + "loss": 0.9957, + "step": 9023 + }, + { + "epoch": 0.8138161157956442, + "grad_norm": 1.5304640485098393, + "learning_rate": 3.526167211394115e-07, + "loss": 0.9924, + "step": 9024 + }, + { + "epoch": 0.8139062993191144, + "grad_norm": 1.452796078443696, + "learning_rate": 3.522855218051066e-07, + "loss": 0.9199, + "step": 9025 + }, + { + "epoch": 0.8139964828425846, + "grad_norm": 2.0571463489313615, + "learning_rate": 3.5195446306288633e-07, + "loss": 0.981, + "step": 9026 + }, + { + "epoch": 0.8140866663660549, + "grad_norm": 1.4250659559278325, + "learning_rate": 3.51623544940999e-07, + "loss": 0.9229, + "step": 9027 + }, + { + "epoch": 0.8141768498895252, + "grad_norm": 1.627481326744669, + "learning_rate": 3.5129276746767886e-07, + "loss": 0.9649, + "step": 9028 + }, + { + "epoch": 0.8142670334129954, + "grad_norm": 1.6184981190667875, + "learning_rate": 3.5096213067115165e-07, + "loss": 0.9536, + "step": 9029 + }, + { + "epoch": 0.8143572169364657, + "grad_norm": 1.8970352457097275, + "learning_rate": 3.506316345796272e-07, + "loss": 0.9973, + "step": 9030 + }, + { + "epoch": 0.814447400459936, + "grad_norm": 0.6226967093653836, + "learning_rate": 3.5030127922130714e-07, + "loss": 0.7727, + "step": 9031 + }, + { + "epoch": 0.8145375839834063, + "grad_norm": 1.5736230522990342, + "learning_rate": 3.4997106462437784e-07, + "loss": 0.8705, + "step": 9032 + }, + { + "epoch": 0.8146277675068765, + "grad_norm": 0.6531115142671143, + "learning_rate": 3.496409908170157e-07, + "loss": 0.7982, + "step": 9033 + }, + { + "epoch": 0.8147179510303467, + "grad_norm": 2.326835121388902, + "learning_rate": 3.493110578273839e-07, + "loss": 0.9442, + "step": 9034 + }, + { + "epoch": 0.8148081345538171, + "grad_norm": 1.7953993999541815, + "learning_rate": 3.489812656836346e-07, + "loss": 1.0357, + "step": 9035 + }, + { + "epoch": 0.8148983180772873, + "grad_norm": 1.438204965635323, + "learning_rate": 3.486516144139078e-07, + "loss": 0.8952, + "step": 9036 + }, + { + "epoch": 0.8149885016007575, + "grad_norm": 1.4815508026002433, + "learning_rate": 3.4832210404632957e-07, + "loss": 0.9517, + "step": 9037 + }, + { + "epoch": 0.8150786851242278, + "grad_norm": 1.4720565424767682, + "learning_rate": 3.479927346090179e-07, + "loss": 0.9889, + "step": 9038 + }, + { + "epoch": 0.8151688686476981, + "grad_norm": 1.6398319325741415, + "learning_rate": 3.4766350613007455e-07, + "loss": 0.9609, + "step": 9039 + }, + { + "epoch": 0.8152590521711683, + "grad_norm": 3.195766426059132, + "learning_rate": 3.4733441863759173e-07, + "loss": 0.9662, + "step": 9040 + }, + { + "epoch": 0.8153492356946386, + "grad_norm": 1.9763809542907345, + "learning_rate": 3.4700547215964916e-07, + "loss": 0.8707, + "step": 9041 + }, + { + "epoch": 0.8154394192181088, + "grad_norm": 1.7300469635779552, + "learning_rate": 3.46676666724314e-07, + "loss": 0.9584, + "step": 9042 + }, + { + "epoch": 0.8155296027415792, + "grad_norm": 1.5880949632610781, + "learning_rate": 3.463480023596421e-07, + "loss": 0.9707, + "step": 9043 + }, + { + "epoch": 0.8156197862650494, + "grad_norm": 1.3811607193896858, + "learning_rate": 3.460194790936772e-07, + "loss": 1.0512, + "step": 9044 + }, + { + "epoch": 0.8157099697885196, + "grad_norm": 1.5762333627498897, + "learning_rate": 3.456910969544495e-07, + "loss": 0.9733, + "step": 9045 + }, + { + "epoch": 0.8158001533119899, + "grad_norm": 1.8917310017584201, + "learning_rate": 3.4536285596997994e-07, + "loss": 1.0076, + "step": 9046 + }, + { + "epoch": 0.8158903368354602, + "grad_norm": 1.962688803854176, + "learning_rate": 3.450347561682747e-07, + "loss": 0.9367, + "step": 9047 + }, + { + "epoch": 0.8159805203589304, + "grad_norm": 2.076541153659565, + "learning_rate": 3.4470679757732945e-07, + "loss": 0.9156, + "step": 9048 + }, + { + "epoch": 0.8160707038824007, + "grad_norm": 1.6690286667992058, + "learning_rate": 3.4437898022512735e-07, + "loss": 1.061, + "step": 9049 + }, + { + "epoch": 0.8161608874058709, + "grad_norm": 1.687254216697444, + "learning_rate": 3.4405130413963977e-07, + "loss": 0.9817, + "step": 9050 + }, + { + "epoch": 0.8162510709293412, + "grad_norm": 1.2982571142608863, + "learning_rate": 3.437237693488262e-07, + "loss": 1.0503, + "step": 9051 + }, + { + "epoch": 0.8163412544528115, + "grad_norm": 1.3744609382965436, + "learning_rate": 3.433963758806322e-07, + "loss": 0.9973, + "step": 9052 + }, + { + "epoch": 0.8164314379762817, + "grad_norm": 1.6532989698858713, + "learning_rate": 3.430691237629948e-07, + "loss": 0.8424, + "step": 9053 + }, + { + "epoch": 0.816521621499752, + "grad_norm": 1.7776372310925002, + "learning_rate": 3.427420130238354e-07, + "loss": 0.9276, + "step": 9054 + }, + { + "epoch": 0.8166118050232223, + "grad_norm": 1.8695346014468628, + "learning_rate": 3.424150436910658e-07, + "loss": 1.0669, + "step": 9055 + }, + { + "epoch": 0.8167019885466925, + "grad_norm": 2.0407217005765634, + "learning_rate": 3.420882157925842e-07, + "loss": 1.0325, + "step": 9056 + }, + { + "epoch": 0.8167921720701627, + "grad_norm": 1.9959955060636887, + "learning_rate": 3.417615293562777e-07, + "loss": 0.9535, + "step": 9057 + }, + { + "epoch": 0.8168823555936331, + "grad_norm": 1.5883015268723042, + "learning_rate": 3.4143498441002105e-07, + "loss": 0.8927, + "step": 9058 + }, + { + "epoch": 0.8169725391171033, + "grad_norm": 1.73093781669078, + "learning_rate": 3.411085809816767e-07, + "loss": 0.9845, + "step": 9059 + }, + { + "epoch": 0.8170627226405736, + "grad_norm": 2.085383423785765, + "learning_rate": 3.407823190990953e-07, + "loss": 1.0919, + "step": 9060 + }, + { + "epoch": 0.8171529061640438, + "grad_norm": 5.465949513140439, + "learning_rate": 3.4045619879011577e-07, + "loss": 1.0629, + "step": 9061 + }, + { + "epoch": 0.8172430896875141, + "grad_norm": 1.8028658305436638, + "learning_rate": 3.4013022008256334e-07, + "loss": 0.9078, + "step": 9062 + }, + { + "epoch": 0.8173332732109844, + "grad_norm": 2.03533981579022, + "learning_rate": 3.398043830042532e-07, + "loss": 0.9174, + "step": 9063 + }, + { + "epoch": 0.8174234567344546, + "grad_norm": 1.818345673178606, + "learning_rate": 3.394786875829871e-07, + "loss": 0.9829, + "step": 9064 + }, + { + "epoch": 0.8175136402579248, + "grad_norm": 1.7937756553170958, + "learning_rate": 3.3915313384655564e-07, + "loss": 0.8625, + "step": 9065 + }, + { + "epoch": 0.8176038237813952, + "grad_norm": 1.3363103959625566, + "learning_rate": 3.388277218227369e-07, + "loss": 0.9956, + "step": 9066 + }, + { + "epoch": 0.8176940073048654, + "grad_norm": 2.0169479405158124, + "learning_rate": 3.3850245153929557e-07, + "loss": 0.9978, + "step": 9067 + }, + { + "epoch": 0.8177841908283356, + "grad_norm": 2.45183398054682, + "learning_rate": 3.381773230239875e-07, + "loss": 0.9403, + "step": 9068 + }, + { + "epoch": 0.8178743743518059, + "grad_norm": 6.884515177182306, + "learning_rate": 3.3785233630455247e-07, + "loss": 1.0059, + "step": 9069 + }, + { + "epoch": 0.8179645578752762, + "grad_norm": 7.694099384868789, + "learning_rate": 3.375274914087221e-07, + "loss": 0.9837, + "step": 9070 + }, + { + "epoch": 0.8180547413987465, + "grad_norm": 1.6478356442344442, + "learning_rate": 3.3720278836421234e-07, + "loss": 0.9527, + "step": 9071 + }, + { + "epoch": 0.8181449249222167, + "grad_norm": 2.531827695880669, + "learning_rate": 3.368782271987294e-07, + "loss": 1.0017, + "step": 9072 + }, + { + "epoch": 0.8182351084456869, + "grad_norm": 1.9700424553803344, + "learning_rate": 3.3655380793996636e-07, + "loss": 0.9864, + "step": 9073 + }, + { + "epoch": 0.8183252919691573, + "grad_norm": 1.4782997689348685, + "learning_rate": 3.362295306156047e-07, + "loss": 0.8988, + "step": 9074 + }, + { + "epoch": 0.8184154754926275, + "grad_norm": 1.5154695726119662, + "learning_rate": 3.3590539525331327e-07, + "loss": 0.9906, + "step": 9075 + }, + { + "epoch": 0.8185056590160977, + "grad_norm": 1.7313328774786638, + "learning_rate": 3.3558140188074967e-07, + "loss": 0.9292, + "step": 9076 + }, + { + "epoch": 0.8185958425395681, + "grad_norm": 1.9804191507904034, + "learning_rate": 3.3525755052555817e-07, + "loss": 1.0091, + "step": 9077 + }, + { + "epoch": 0.8186860260630383, + "grad_norm": 0.7346669286446471, + "learning_rate": 3.3493384121537147e-07, + "loss": 0.8858, + "step": 9078 + }, + { + "epoch": 0.8187762095865085, + "grad_norm": 1.7010181940082374, + "learning_rate": 3.3461027397781075e-07, + "loss": 0.9771, + "step": 9079 + }, + { + "epoch": 0.8188663931099788, + "grad_norm": 1.8212697536886988, + "learning_rate": 3.3428684884048397e-07, + "loss": 0.9653, + "step": 9080 + }, + { + "epoch": 0.8189565766334491, + "grad_norm": 1.636129049984669, + "learning_rate": 3.3396356583098826e-07, + "loss": 0.9811, + "step": 9081 + }, + { + "epoch": 0.8190467601569194, + "grad_norm": 1.5651357164128021, + "learning_rate": 3.3364042497690736e-07, + "loss": 1.0126, + "step": 9082 + }, + { + "epoch": 0.8191369436803896, + "grad_norm": 1.3373518032790186, + "learning_rate": 3.3331742630581405e-07, + "loss": 1.035, + "step": 9083 + }, + { + "epoch": 0.8192271272038598, + "grad_norm": 1.3674010191807193, + "learning_rate": 3.3299456984526717e-07, + "loss": 0.9281, + "step": 9084 + }, + { + "epoch": 0.8193173107273302, + "grad_norm": 1.6685837766461982, + "learning_rate": 3.3267185562281605e-07, + "loss": 1.0171, + "step": 9085 + }, + { + "epoch": 0.8194074942508004, + "grad_norm": 1.773444087482603, + "learning_rate": 3.3234928366599514e-07, + "loss": 0.9754, + "step": 9086 + }, + { + "epoch": 0.8194976777742706, + "grad_norm": 1.730240714244621, + "learning_rate": 3.3202685400232946e-07, + "loss": 1.011, + "step": 9087 + }, + { + "epoch": 0.8195878612977409, + "grad_norm": 2.5965730861145473, + "learning_rate": 3.317045666593292e-07, + "loss": 1.0061, + "step": 9088 + }, + { + "epoch": 0.8196780448212112, + "grad_norm": 2.1696872807551673, + "learning_rate": 3.3138242166449426e-07, + "loss": 0.9841, + "step": 9089 + }, + { + "epoch": 0.8197682283446814, + "grad_norm": 1.5015364566255331, + "learning_rate": 3.310604190453117e-07, + "loss": 1.0438, + "step": 9090 + }, + { + "epoch": 0.8198584118681517, + "grad_norm": 1.4965282635885908, + "learning_rate": 3.307385588292566e-07, + "loss": 0.9269, + "step": 9091 + }, + { + "epoch": 0.8199485953916219, + "grad_norm": 1.7944038980323471, + "learning_rate": 3.304168410437924e-07, + "loss": 0.9723, + "step": 9092 + }, + { + "epoch": 0.8200387789150922, + "grad_norm": 0.6263521826064661, + "learning_rate": 3.300952657163687e-07, + "loss": 0.8404, + "step": 9093 + }, + { + "epoch": 0.8201289624385625, + "grad_norm": 1.8130817137544386, + "learning_rate": 3.297738328744248e-07, + "loss": 0.9612, + "step": 9094 + }, + { + "epoch": 0.8202191459620327, + "grad_norm": 1.8734112684435313, + "learning_rate": 3.2945254254538714e-07, + "loss": 0.9699, + "step": 9095 + }, + { + "epoch": 0.820309329485503, + "grad_norm": 2.0136509181059696, + "learning_rate": 3.2913139475666963e-07, + "loss": 0.9447, + "step": 9096 + }, + { + "epoch": 0.8203995130089733, + "grad_norm": 2.5377776196077604, + "learning_rate": 3.288103895356749e-07, + "loss": 0.9291, + "step": 9097 + }, + { + "epoch": 0.8204896965324435, + "grad_norm": 1.6218317397270545, + "learning_rate": 3.284895269097927e-07, + "loss": 0.9134, + "step": 9098 + }, + { + "epoch": 0.8205798800559138, + "grad_norm": 0.6813390769817491, + "learning_rate": 3.281688069063999e-07, + "loss": 0.8189, + "step": 9099 + }, + { + "epoch": 0.8206700635793841, + "grad_norm": 1.9057880529782834, + "learning_rate": 3.2784822955286396e-07, + "loss": 0.8043, + "step": 9100 + }, + { + "epoch": 0.8207602471028543, + "grad_norm": 1.5237749557391718, + "learning_rate": 3.275277948765365e-07, + "loss": 1.0505, + "step": 9101 + }, + { + "epoch": 0.8208504306263246, + "grad_norm": 2.1319341283072997, + "learning_rate": 3.2720750290475964e-07, + "loss": 1.0218, + "step": 9102 + }, + { + "epoch": 0.8209406141497948, + "grad_norm": 1.8927047261980696, + "learning_rate": 3.268873536648622e-07, + "loss": 0.9822, + "step": 9103 + }, + { + "epoch": 0.8210307976732651, + "grad_norm": 1.650027624534785, + "learning_rate": 3.265673471841612e-07, + "loss": 0.9502, + "step": 9104 + }, + { + "epoch": 0.8211209811967354, + "grad_norm": 0.6284982530695536, + "learning_rate": 3.262474834899616e-07, + "loss": 0.7773, + "step": 9105 + }, + { + "epoch": 0.8212111647202056, + "grad_norm": 0.6716625631844408, + "learning_rate": 3.2592776260955534e-07, + "loss": 0.853, + "step": 9106 + }, + { + "epoch": 0.8213013482436758, + "grad_norm": 1.843983780861977, + "learning_rate": 3.256081845702239e-07, + "loss": 1.0193, + "step": 9107 + }, + { + "epoch": 0.8213915317671462, + "grad_norm": 1.4571789214135102, + "learning_rate": 3.2528874939923335e-07, + "loss": 1.0297, + "step": 9108 + }, + { + "epoch": 0.8214817152906164, + "grad_norm": 1.9713685317799006, + "learning_rate": 3.2496945712384217e-07, + "loss": 0.9361, + "step": 9109 + }, + { + "epoch": 0.8215718988140867, + "grad_norm": 1.7675373220212802, + "learning_rate": 3.246503077712923e-07, + "loss": 0.9575, + "step": 9110 + }, + { + "epoch": 0.8216620823375569, + "grad_norm": 1.6793183386798176, + "learning_rate": 3.2433130136881625e-07, + "loss": 1.0723, + "step": 9111 + }, + { + "epoch": 0.8217522658610272, + "grad_norm": 2.102942258425787, + "learning_rate": 3.2401243794363287e-07, + "loss": 0.9722, + "step": 9112 + }, + { + "epoch": 0.8218424493844975, + "grad_norm": 1.560124537069715, + "learning_rate": 3.236937175229495e-07, + "loss": 1.0356, + "step": 9113 + }, + { + "epoch": 0.8219326329079677, + "grad_norm": 1.9021667147435055, + "learning_rate": 3.233751401339615e-07, + "loss": 0.9492, + "step": 9114 + }, + { + "epoch": 0.8220228164314379, + "grad_norm": 1.7878811196526054, + "learning_rate": 3.2305670580385157e-07, + "loss": 1.0594, + "step": 9115 + }, + { + "epoch": 0.8221129999549083, + "grad_norm": 1.6021785388663938, + "learning_rate": 3.227384145597898e-07, + "loss": 0.9817, + "step": 9116 + }, + { + "epoch": 0.8222031834783785, + "grad_norm": 2.151306899546784, + "learning_rate": 3.224202664289346e-07, + "loss": 0.9629, + "step": 9117 + }, + { + "epoch": 0.8222933670018487, + "grad_norm": 1.6300058289874177, + "learning_rate": 3.2210226143843257e-07, + "loss": 0.9229, + "step": 9118 + }, + { + "epoch": 0.822383550525319, + "grad_norm": 0.6539410434886848, + "learning_rate": 3.217843996154173e-07, + "loss": 0.8504, + "step": 9119 + }, + { + "epoch": 0.8224737340487893, + "grad_norm": 0.6048046284477049, + "learning_rate": 3.2146668098701055e-07, + "loss": 0.7655, + "step": 9120 + }, + { + "epoch": 0.8225639175722596, + "grad_norm": 2.0019173967848434, + "learning_rate": 3.2114910558032215e-07, + "loss": 1.107, + "step": 9121 + }, + { + "epoch": 0.8226541010957298, + "grad_norm": 2.3371439086260146, + "learning_rate": 3.2083167342244945e-07, + "loss": 1.0075, + "step": 9122 + }, + { + "epoch": 0.8227442846192, + "grad_norm": 1.850362332405384, + "learning_rate": 3.205143845404763e-07, + "loss": 1.0796, + "step": 9123 + }, + { + "epoch": 0.8228344681426704, + "grad_norm": 1.9747877031014136, + "learning_rate": 3.201972389614773e-07, + "loss": 0.9553, + "step": 9124 + }, + { + "epoch": 0.8229246516661406, + "grad_norm": 2.4453616396320577, + "learning_rate": 3.198802367125115e-07, + "loss": 0.9714, + "step": 9125 + }, + { + "epoch": 0.8230148351896108, + "grad_norm": 3.0227136657262563, + "learning_rate": 3.195633778206288e-07, + "loss": 0.9885, + "step": 9126 + }, + { + "epoch": 0.8231050187130812, + "grad_norm": 2.1007154381353663, + "learning_rate": 3.19246662312864e-07, + "loss": 1.0571, + "step": 9127 + }, + { + "epoch": 0.8231952022365514, + "grad_norm": 0.6385555803070202, + "learning_rate": 3.189300902162417e-07, + "loss": 0.8149, + "step": 9128 + }, + { + "epoch": 0.8232853857600216, + "grad_norm": 1.8753124612491339, + "learning_rate": 3.1861366155777327e-07, + "loss": 0.9929, + "step": 9129 + }, + { + "epoch": 0.8233755692834919, + "grad_norm": 1.4232099365835273, + "learning_rate": 3.182973763644583e-07, + "loss": 0.9913, + "step": 9130 + }, + { + "epoch": 0.8234657528069622, + "grad_norm": 1.4316077339172366, + "learning_rate": 3.1798123466328463e-07, + "loss": 0.9051, + "step": 9131 + }, + { + "epoch": 0.8235559363304324, + "grad_norm": 2.475603080253762, + "learning_rate": 3.17665236481226e-07, + "loss": 0.9798, + "step": 9132 + }, + { + "epoch": 0.8236461198539027, + "grad_norm": 2.406659079123219, + "learning_rate": 3.1734938184524576e-07, + "loss": 0.94, + "step": 9133 + }, + { + "epoch": 0.8237363033773729, + "grad_norm": 4.309229840993605, + "learning_rate": 3.1703367078229427e-07, + "loss": 0.9901, + "step": 9134 + }, + { + "epoch": 0.8238264869008433, + "grad_norm": 1.9340358036163137, + "learning_rate": 3.167181033193096e-07, + "loss": 0.9425, + "step": 9135 + }, + { + "epoch": 0.8239166704243135, + "grad_norm": 1.683864208899052, + "learning_rate": 3.16402679483218e-07, + "loss": 0.9921, + "step": 9136 + }, + { + "epoch": 0.8240068539477837, + "grad_norm": 1.4985449409351865, + "learning_rate": 3.1608739930093366e-07, + "loss": 0.9102, + "step": 9137 + }, + { + "epoch": 0.824097037471254, + "grad_norm": 1.888056373700154, + "learning_rate": 3.157722627993562e-07, + "loss": 0.894, + "step": 9138 + }, + { + "epoch": 0.8241872209947243, + "grad_norm": 1.5679037874199535, + "learning_rate": 3.1545727000537727e-07, + "loss": 0.9424, + "step": 9139 + }, + { + "epoch": 0.8242774045181945, + "grad_norm": 1.678849903011413, + "learning_rate": 3.151424209458713e-07, + "loss": 0.9798, + "step": 9140 + }, + { + "epoch": 0.8243675880416648, + "grad_norm": 1.6505017124688326, + "learning_rate": 3.148277156477053e-07, + "loss": 0.9942, + "step": 9141 + }, + { + "epoch": 0.824457771565135, + "grad_norm": 1.8271377375588624, + "learning_rate": 3.145131541377299e-07, + "loss": 0.9813, + "step": 9142 + }, + { + "epoch": 0.8245479550886053, + "grad_norm": 1.8979156153564667, + "learning_rate": 3.1419873644278606e-07, + "loss": 0.9451, + "step": 9143 + }, + { + "epoch": 0.8246381386120756, + "grad_norm": 1.5316550828387063, + "learning_rate": 3.1388446258970147e-07, + "loss": 1.0041, + "step": 9144 + }, + { + "epoch": 0.8247283221355458, + "grad_norm": 1.6839724510866194, + "learning_rate": 3.1357033260529145e-07, + "loss": 0.9643, + "step": 9145 + }, + { + "epoch": 0.824818505659016, + "grad_norm": 1.4808809617425576, + "learning_rate": 3.1325634651636025e-07, + "loss": 1.0567, + "step": 9146 + }, + { + "epoch": 0.8249086891824864, + "grad_norm": 1.8122424896417193, + "learning_rate": 3.1294250434969694e-07, + "loss": 1.0201, + "step": 9147 + }, + { + "epoch": 0.8249988727059566, + "grad_norm": 1.916629673421147, + "learning_rate": 3.1262880613208274e-07, + "loss": 1.0184, + "step": 9148 + }, + { + "epoch": 0.8250890562294269, + "grad_norm": 1.5824718474042907, + "learning_rate": 3.123152518902823e-07, + "loss": 1.0187, + "step": 9149 + }, + { + "epoch": 0.8251792397528972, + "grad_norm": 1.4582534041526034, + "learning_rate": 3.1200184165105017e-07, + "loss": 0.9334, + "step": 9150 + }, + { + "epoch": 0.8252694232763674, + "grad_norm": 1.654707928496307, + "learning_rate": 3.116885754411287e-07, + "loss": 0.9898, + "step": 9151 + }, + { + "epoch": 0.8253596067998377, + "grad_norm": 2.1079134928924543, + "learning_rate": 3.1137545328724703e-07, + "loss": 0.9129, + "step": 9152 + }, + { + "epoch": 0.8254497903233079, + "grad_norm": 1.7403865794743432, + "learning_rate": 3.1106247521612285e-07, + "loss": 0.9817, + "step": 9153 + }, + { + "epoch": 0.8255399738467782, + "grad_norm": 1.7428108776482385, + "learning_rate": 3.107496412544612e-07, + "loss": 0.9984, + "step": 9154 + }, + { + "epoch": 0.8256301573702485, + "grad_norm": 1.3786351529082266, + "learning_rate": 3.1043695142895397e-07, + "loss": 1.0064, + "step": 9155 + }, + { + "epoch": 0.8257203408937187, + "grad_norm": 1.4487037257653432, + "learning_rate": 3.101244057662828e-07, + "loss": 0.9011, + "step": 9156 + }, + { + "epoch": 0.8258105244171889, + "grad_norm": 2.063554551997258, + "learning_rate": 3.098120042931152e-07, + "loss": 0.9698, + "step": 9157 + }, + { + "epoch": 0.8259007079406593, + "grad_norm": 1.5103470446446152, + "learning_rate": 3.0949974703610647e-07, + "loss": 1.0222, + "step": 9158 + }, + { + "epoch": 0.8259908914641295, + "grad_norm": 2.1296067511516723, + "learning_rate": 3.0918763402190107e-07, + "loss": 0.9693, + "step": 9159 + }, + { + "epoch": 0.8260810749875998, + "grad_norm": 1.731867102526166, + "learning_rate": 3.088756652771296e-07, + "loss": 0.9351, + "step": 9160 + }, + { + "epoch": 0.82617125851107, + "grad_norm": 2.1158488077762465, + "learning_rate": 3.0856384082841147e-07, + "loss": 1.0725, + "step": 9161 + }, + { + "epoch": 0.8262614420345403, + "grad_norm": 2.420427345028612, + "learning_rate": 3.0825216070235207e-07, + "loss": 0.9256, + "step": 9162 + }, + { + "epoch": 0.8263516255580106, + "grad_norm": 2.0346734175255516, + "learning_rate": 3.0794062492554764e-07, + "loss": 1.0925, + "step": 9163 + }, + { + "epoch": 0.8264418090814808, + "grad_norm": 1.8129135186283405, + "learning_rate": 3.076292335245783e-07, + "loss": 0.9959, + "step": 9164 + }, + { + "epoch": 0.826531992604951, + "grad_norm": 1.9901796282874422, + "learning_rate": 3.073179865260145e-07, + "loss": 1.0516, + "step": 9165 + }, + { + "epoch": 0.8266221761284214, + "grad_norm": 1.546337226868256, + "learning_rate": 3.070068839564135e-07, + "loss": 0.9852, + "step": 9166 + }, + { + "epoch": 0.8267123596518916, + "grad_norm": 5.27263394485438, + "learning_rate": 3.0669592584232006e-07, + "loss": 1.0165, + "step": 9167 + }, + { + "epoch": 0.8268025431753618, + "grad_norm": 1.5765324479141467, + "learning_rate": 3.063851122102672e-07, + "loss": 0.8916, + "step": 9168 + }, + { + "epoch": 0.8268927266988321, + "grad_norm": 2.081558946928031, + "learning_rate": 3.06074443086775e-07, + "loss": 0.8999, + "step": 9169 + }, + { + "epoch": 0.8269829102223024, + "grad_norm": 0.5975067101185588, + "learning_rate": 3.057639184983514e-07, + "loss": 0.7682, + "step": 9170 + }, + { + "epoch": 0.8270730937457726, + "grad_norm": 1.2929449165660016, + "learning_rate": 3.054535384714927e-07, + "loss": 0.9845, + "step": 9171 + }, + { + "epoch": 0.8271632772692429, + "grad_norm": 2.0671048632977334, + "learning_rate": 3.0514330303268135e-07, + "loss": 1.1457, + "step": 9172 + }, + { + "epoch": 0.8272534607927132, + "grad_norm": 1.8030866648600377, + "learning_rate": 3.0483321220838876e-07, + "loss": 0.933, + "step": 9173 + }, + { + "epoch": 0.8273436443161835, + "grad_norm": 4.827637681780589, + "learning_rate": 3.045232660250734e-07, + "loss": 0.8392, + "step": 9174 + }, + { + "epoch": 0.8274338278396537, + "grad_norm": 1.6283350247101345, + "learning_rate": 3.0421346450918185e-07, + "loss": 1.0092, + "step": 9175 + }, + { + "epoch": 0.8275240113631239, + "grad_norm": 1.4558768518789336, + "learning_rate": 3.039038076871485e-07, + "loss": 0.9285, + "step": 9176 + }, + { + "epoch": 0.8276141948865943, + "grad_norm": 1.4470788008844708, + "learning_rate": 3.035942955853934e-07, + "loss": 0.9973, + "step": 9177 + }, + { + "epoch": 0.8277043784100645, + "grad_norm": 1.7792799244313, + "learning_rate": 3.0328492823032804e-07, + "loss": 0.9569, + "step": 9178 + }, + { + "epoch": 0.8277945619335347, + "grad_norm": 2.22206351852619, + "learning_rate": 3.029757056483471e-07, + "loss": 0.9191, + "step": 9179 + }, + { + "epoch": 0.827884745457005, + "grad_norm": 2.0808321879970264, + "learning_rate": 3.026666278658372e-07, + "loss": 0.8697, + "step": 9180 + }, + { + "epoch": 0.8279749289804753, + "grad_norm": 1.5640842036095228, + "learning_rate": 3.023576949091691e-07, + "loss": 0.9772, + "step": 9181 + }, + { + "epoch": 0.8280651125039455, + "grad_norm": 2.190894626453128, + "learning_rate": 3.020489068047032e-07, + "loss": 0.9053, + "step": 9182 + }, + { + "epoch": 0.8281552960274158, + "grad_norm": 1.8674782243350199, + "learning_rate": 3.017402635787869e-07, + "loss": 0.9414, + "step": 9183 + }, + { + "epoch": 0.828245479550886, + "grad_norm": 1.7215832333260594, + "learning_rate": 3.0143176525775537e-07, + "loss": 0.952, + "step": 9184 + }, + { + "epoch": 0.8283356630743564, + "grad_norm": 1.5786773923485915, + "learning_rate": 3.0112341186793155e-07, + "loss": 0.925, + "step": 9185 + }, + { + "epoch": 0.8284258465978266, + "grad_norm": 1.4958262391130257, + "learning_rate": 3.008152034356264e-07, + "loss": 0.914, + "step": 9186 + }, + { + "epoch": 0.8285160301212968, + "grad_norm": 0.6310075520911387, + "learning_rate": 3.005071399871366e-07, + "loss": 0.8004, + "step": 9187 + }, + { + "epoch": 0.828606213644767, + "grad_norm": 1.902279392540256, + "learning_rate": 3.0019922154874853e-07, + "loss": 1.0397, + "step": 9188 + }, + { + "epoch": 0.8286963971682374, + "grad_norm": 1.5200170216109221, + "learning_rate": 2.998914481467356e-07, + "loss": 0.9356, + "step": 9189 + }, + { + "epoch": 0.8287865806917076, + "grad_norm": 1.9736205510698106, + "learning_rate": 2.9958381980735837e-07, + "loss": 0.9275, + "step": 9190 + }, + { + "epoch": 0.8288767642151779, + "grad_norm": 1.788042335666167, + "learning_rate": 2.992763365568658e-07, + "loss": 0.9072, + "step": 9191 + }, + { + "epoch": 0.8289669477386481, + "grad_norm": 1.6036886105640054, + "learning_rate": 2.98968998421494e-07, + "loss": 0.982, + "step": 9192 + }, + { + "epoch": 0.8290571312621184, + "grad_norm": 1.705967691842299, + "learning_rate": 2.98661805427467e-07, + "loss": 0.968, + "step": 9193 + }, + { + "epoch": 0.8291473147855887, + "grad_norm": 1.773068898079552, + "learning_rate": 2.9835475760099483e-07, + "loss": 0.9948, + "step": 9194 + }, + { + "epoch": 0.8292374983090589, + "grad_norm": 4.18825808822697, + "learning_rate": 2.9804785496827856e-07, + "loss": 0.9936, + "step": 9195 + }, + { + "epoch": 0.8293276818325293, + "grad_norm": 2.235150289410907, + "learning_rate": 2.977410975555028e-07, + "loss": 0.9758, + "step": 9196 + }, + { + "epoch": 0.8294178653559995, + "grad_norm": 1.769447577310873, + "learning_rate": 2.9743448538884376e-07, + "loss": 1.0385, + "step": 9197 + }, + { + "epoch": 0.8295080488794697, + "grad_norm": 1.5037638807409215, + "learning_rate": 2.9712801849446154e-07, + "loss": 0.9673, + "step": 9198 + }, + { + "epoch": 0.82959823240294, + "grad_norm": 1.6648004892314612, + "learning_rate": 2.9682169689850665e-07, + "loss": 0.9291, + "step": 9199 + }, + { + "epoch": 0.8296884159264103, + "grad_norm": 0.6499819688863651, + "learning_rate": 2.9651552062711573e-07, + "loss": 0.7638, + "step": 9200 + }, + { + "epoch": 0.8297785994498805, + "grad_norm": 2.349252472341466, + "learning_rate": 2.9620948970641333e-07, + "loss": 1.0591, + "step": 9201 + }, + { + "epoch": 0.8298687829733508, + "grad_norm": 1.3938738130734254, + "learning_rate": 2.959036041625125e-07, + "loss": 0.9934, + "step": 9202 + }, + { + "epoch": 0.829958966496821, + "grad_norm": 1.8461411052349797, + "learning_rate": 2.95597864021512e-07, + "loss": 0.8841, + "step": 9203 + }, + { + "epoch": 0.8300491500202913, + "grad_norm": 3.5375540847148494, + "learning_rate": 2.9529226930949966e-07, + "loss": 1.0278, + "step": 9204 + }, + { + "epoch": 0.8301393335437616, + "grad_norm": 1.8431989848596042, + "learning_rate": 2.949868200525505e-07, + "loss": 0.9912, + "step": 9205 + }, + { + "epoch": 0.8302295170672318, + "grad_norm": 1.799575236323551, + "learning_rate": 2.9468151627672734e-07, + "loss": 0.9716, + "step": 9206 + }, + { + "epoch": 0.830319700590702, + "grad_norm": 1.6722790238811132, + "learning_rate": 2.9437635800808026e-07, + "loss": 0.9364, + "step": 9207 + }, + { + "epoch": 0.8304098841141724, + "grad_norm": 2.254699461674568, + "learning_rate": 2.940713452726473e-07, + "loss": 0.9718, + "step": 9208 + }, + { + "epoch": 0.8305000676376426, + "grad_norm": 4.158765218217877, + "learning_rate": 2.937664780964526e-07, + "loss": 0.973, + "step": 9209 + }, + { + "epoch": 0.8305902511611128, + "grad_norm": 1.9545171430162387, + "learning_rate": 2.9346175650551133e-07, + "loss": 0.9888, + "step": 9210 + }, + { + "epoch": 0.8306804346845831, + "grad_norm": 1.5679044716987853, + "learning_rate": 2.931571805258215e-07, + "loss": 0.9993, + "step": 9211 + }, + { + "epoch": 0.8307706182080534, + "grad_norm": 8.132515116741157, + "learning_rate": 2.9285275018337353e-07, + "loss": 1.0249, + "step": 9212 + }, + { + "epoch": 0.8308608017315237, + "grad_norm": 1.7025856073845516, + "learning_rate": 2.9254846550414146e-07, + "loss": 1.08, + "step": 9213 + }, + { + "epoch": 0.8309509852549939, + "grad_norm": 1.8218871415585922, + "learning_rate": 2.922443265140893e-07, + "loss": 1.0272, + "step": 9214 + }, + { + "epoch": 0.8310411687784641, + "grad_norm": 1.8559393677426697, + "learning_rate": 2.919403332391674e-07, + "loss": 0.9755, + "step": 9215 + }, + { + "epoch": 0.8311313523019345, + "grad_norm": 1.7804969400381196, + "learning_rate": 2.9163648570531464e-07, + "loss": 1.0042, + "step": 9216 + }, + { + "epoch": 0.8312215358254047, + "grad_norm": 1.7510257848064394, + "learning_rate": 2.9133278393845717e-07, + "loss": 0.9182, + "step": 9217 + }, + { + "epoch": 0.8313117193488749, + "grad_norm": 1.4990894414927252, + "learning_rate": 2.9102922796450703e-07, + "loss": 1.0516, + "step": 9218 + }, + { + "epoch": 0.8314019028723453, + "grad_norm": 1.7470364681951014, + "learning_rate": 2.907258178093672e-07, + "loss": 0.9374, + "step": 9219 + }, + { + "epoch": 0.8314920863958155, + "grad_norm": 1.5104012674884004, + "learning_rate": 2.904225534989251e-07, + "loss": 0.8996, + "step": 9220 + }, + { + "epoch": 0.8315822699192857, + "grad_norm": 2.5565545495262505, + "learning_rate": 2.901194350590572e-07, + "loss": 0.8815, + "step": 9221 + }, + { + "epoch": 0.831672453442756, + "grad_norm": 2.1413197747312354, + "learning_rate": 2.898164625156274e-07, + "loss": 0.8632, + "step": 9222 + }, + { + "epoch": 0.8317626369662263, + "grad_norm": 1.5877129899968496, + "learning_rate": 2.8951363589448676e-07, + "loss": 0.9362, + "step": 9223 + }, + { + "epoch": 0.8318528204896966, + "grad_norm": 1.5451176997909497, + "learning_rate": 2.8921095522147434e-07, + "loss": 0.9752, + "step": 9224 + }, + { + "epoch": 0.8319430040131668, + "grad_norm": 1.7659792924403517, + "learning_rate": 2.8890842052241683e-07, + "loss": 1.0076, + "step": 9225 + }, + { + "epoch": 0.832033187536637, + "grad_norm": 0.6012163280884606, + "learning_rate": 2.886060318231267e-07, + "loss": 0.7641, + "step": 9226 + }, + { + "epoch": 0.8321233710601074, + "grad_norm": 1.645518099698927, + "learning_rate": 2.8830378914940755e-07, + "loss": 1.0055, + "step": 9227 + }, + { + "epoch": 0.8322135545835776, + "grad_norm": 1.7430300877480698, + "learning_rate": 2.8800169252704675e-07, + "loss": 0.9688, + "step": 9228 + }, + { + "epoch": 0.8323037381070478, + "grad_norm": 1.6808129377109573, + "learning_rate": 2.8769974198182143e-07, + "loss": 1.002, + "step": 9229 + }, + { + "epoch": 0.8323939216305181, + "grad_norm": 1.5276491294024577, + "learning_rate": 2.873979375394955e-07, + "loss": 0.9826, + "step": 9230 + }, + { + "epoch": 0.8324841051539884, + "grad_norm": 1.6104830288589431, + "learning_rate": 2.870962792258209e-07, + "loss": 1.0113, + "step": 9231 + }, + { + "epoch": 0.8325742886774586, + "grad_norm": 1.4422041176017328, + "learning_rate": 2.8679476706653716e-07, + "loss": 0.9904, + "step": 9232 + }, + { + "epoch": 0.8326644722009289, + "grad_norm": 1.8018694917963145, + "learning_rate": 2.864934010873692e-07, + "loss": 1.002, + "step": 9233 + }, + { + "epoch": 0.8327546557243991, + "grad_norm": 1.956810365801638, + "learning_rate": 2.8619218131403357e-07, + "loss": 0.9534, + "step": 9234 + }, + { + "epoch": 0.8328448392478695, + "grad_norm": 1.8548737867042764, + "learning_rate": 2.858911077722299e-07, + "loss": 1.0033, + "step": 9235 + }, + { + "epoch": 0.8329350227713397, + "grad_norm": 2.1400177922801564, + "learning_rate": 2.855901804876493e-07, + "loss": 0.8302, + "step": 9236 + }, + { + "epoch": 0.8330252062948099, + "grad_norm": 1.4598378821729394, + "learning_rate": 2.852893994859673e-07, + "loss": 1.0173, + "step": 9237 + }, + { + "epoch": 0.8331153898182801, + "grad_norm": 2.2825329452092418, + "learning_rate": 2.849887647928484e-07, + "loss": 0.9266, + "step": 9238 + }, + { + "epoch": 0.8332055733417505, + "grad_norm": 2.247883330891378, + "learning_rate": 2.8468827643394465e-07, + "loss": 0.8745, + "step": 9239 + }, + { + "epoch": 0.8332957568652207, + "grad_norm": 2.226234920564159, + "learning_rate": 2.843879344348954e-07, + "loss": 0.9932, + "step": 9240 + }, + { + "epoch": 0.833385940388691, + "grad_norm": 1.5607603688589913, + "learning_rate": 2.840877388213272e-07, + "loss": 0.9516, + "step": 9241 + }, + { + "epoch": 0.8334761239121612, + "grad_norm": 1.7894283403713311, + "learning_rate": 2.8378768961885515e-07, + "loss": 0.9799, + "step": 9242 + }, + { + "epoch": 0.8335663074356315, + "grad_norm": 1.7320899000204724, + "learning_rate": 2.8348778685307983e-07, + "loss": 1.0118, + "step": 9243 + }, + { + "epoch": 0.8336564909591018, + "grad_norm": 1.6424541630728726, + "learning_rate": 2.831880305495915e-07, + "loss": 1.0531, + "step": 9244 + }, + { + "epoch": 0.833746674482572, + "grad_norm": 0.815522367955348, + "learning_rate": 2.828884207339668e-07, + "loss": 0.8489, + "step": 9245 + }, + { + "epoch": 0.8338368580060423, + "grad_norm": 2.3198938473718425, + "learning_rate": 2.8258895743177014e-07, + "loss": 1.0539, + "step": 9246 + }, + { + "epoch": 0.8339270415295126, + "grad_norm": 1.8767554648459404, + "learning_rate": 2.8228964066855356e-07, + "loss": 1.0536, + "step": 9247 + }, + { + "epoch": 0.8340172250529828, + "grad_norm": 1.4499775424402037, + "learning_rate": 2.819904704698555e-07, + "loss": 0.9581, + "step": 9248 + }, + { + "epoch": 0.834107408576453, + "grad_norm": 0.63003814911006, + "learning_rate": 2.8169144686120437e-07, + "loss": 0.7645, + "step": 9249 + }, + { + "epoch": 0.8341975920999234, + "grad_norm": 1.6253663163681082, + "learning_rate": 2.8139256986811254e-07, + "loss": 0.8938, + "step": 9250 + }, + { + "epoch": 0.8342877756233936, + "grad_norm": 1.623009416168595, + "learning_rate": 2.8109383951608424e-07, + "loss": 1.0314, + "step": 9251 + }, + { + "epoch": 0.8343779591468639, + "grad_norm": 1.8313254357121254, + "learning_rate": 2.8079525583060683e-07, + "loss": 1.0457, + "step": 9252 + }, + { + "epoch": 0.8344681426703341, + "grad_norm": 1.338459734753153, + "learning_rate": 2.804968188371577e-07, + "loss": 0.9875, + "step": 9253 + }, + { + "epoch": 0.8345583261938044, + "grad_norm": 1.4415907496196825, + "learning_rate": 2.801985285612014e-07, + "loss": 0.9763, + "step": 9254 + }, + { + "epoch": 0.8346485097172747, + "grad_norm": 1.7275179467600832, + "learning_rate": 2.7990038502818934e-07, + "loss": 0.953, + "step": 9255 + }, + { + "epoch": 0.8347386932407449, + "grad_norm": 1.6591004368994207, + "learning_rate": 2.796023882635612e-07, + "loss": 0.956, + "step": 9256 + }, + { + "epoch": 0.8348288767642151, + "grad_norm": 2.240098997380787, + "learning_rate": 2.7930453829274323e-07, + "loss": 1.0273, + "step": 9257 + }, + { + "epoch": 0.8349190602876855, + "grad_norm": 1.8876384760099805, + "learning_rate": 2.7900683514115054e-07, + "loss": 1.0203, + "step": 9258 + }, + { + "epoch": 0.8350092438111557, + "grad_norm": 1.875280232151099, + "learning_rate": 2.787092788341836e-07, + "loss": 0.9158, + "step": 9259 + }, + { + "epoch": 0.8350994273346259, + "grad_norm": 1.7312660189001117, + "learning_rate": 2.7841186939723195e-07, + "loss": 0.9784, + "step": 9260 + }, + { + "epoch": 0.8351896108580962, + "grad_norm": 1.6385602037755875, + "learning_rate": 2.7811460685567255e-07, + "loss": 0.9558, + "step": 9261 + }, + { + "epoch": 0.8352797943815665, + "grad_norm": 2.7735109131458913, + "learning_rate": 2.778174912348692e-07, + "loss": 0.969, + "step": 9262 + }, + { + "epoch": 0.8353699779050368, + "grad_norm": 1.7569780191556588, + "learning_rate": 2.7752052256017354e-07, + "loss": 0.9465, + "step": 9263 + }, + { + "epoch": 0.835460161428507, + "grad_norm": 1.7321905177305341, + "learning_rate": 2.7722370085692493e-07, + "loss": 0.9498, + "step": 9264 + }, + { + "epoch": 0.8355503449519772, + "grad_norm": 2.2931864698462356, + "learning_rate": 2.769270261504486e-07, + "loss": 0.9524, + "step": 9265 + }, + { + "epoch": 0.8356405284754476, + "grad_norm": 1.4104139774058335, + "learning_rate": 2.7663049846606015e-07, + "loss": 0.9411, + "step": 9266 + }, + { + "epoch": 0.8357307119989178, + "grad_norm": 1.6947528249286856, + "learning_rate": 2.763341178290592e-07, + "loss": 0.9919, + "step": 9267 + }, + { + "epoch": 0.835820895522388, + "grad_norm": 1.5081328214844902, + "learning_rate": 2.7603788426473663e-07, + "loss": 1.004, + "step": 9268 + }, + { + "epoch": 0.8359110790458584, + "grad_norm": 2.282865083207648, + "learning_rate": 2.7574179779836695e-07, + "loss": 0.9126, + "step": 9269 + }, + { + "epoch": 0.8360012625693286, + "grad_norm": 2.388571960904811, + "learning_rate": 2.754458584552146e-07, + "loss": 0.938, + "step": 9270 + }, + { + "epoch": 0.8360914460927988, + "grad_norm": 1.8083902660613322, + "learning_rate": 2.751500662605308e-07, + "loss": 0.9649, + "step": 9271 + }, + { + "epoch": 0.8361816296162691, + "grad_norm": 1.550799468195295, + "learning_rate": 2.7485442123955383e-07, + "loss": 0.9357, + "step": 9272 + }, + { + "epoch": 0.8362718131397394, + "grad_norm": 2.5267512527572924, + "learning_rate": 2.7455892341751075e-07, + "loss": 0.8646, + "step": 9273 + }, + { + "epoch": 0.8363619966632097, + "grad_norm": 1.9353840256132944, + "learning_rate": 2.7426357281961365e-07, + "loss": 0.8554, + "step": 9274 + }, + { + "epoch": 0.8364521801866799, + "grad_norm": 1.809586254694009, + "learning_rate": 2.7396836947106416e-07, + "loss": 0.986, + "step": 9275 + }, + { + "epoch": 0.8365423637101501, + "grad_norm": 2.2731509732725717, + "learning_rate": 2.736733133970506e-07, + "loss": 1.0772, + "step": 9276 + }, + { + "epoch": 0.8366325472336205, + "grad_norm": 1.5569266292260966, + "learning_rate": 2.7337840462274896e-07, + "loss": 1.0199, + "step": 9277 + }, + { + "epoch": 0.8367227307570907, + "grad_norm": 1.6506845788050586, + "learning_rate": 2.730836431733221e-07, + "loss": 0.9466, + "step": 9278 + }, + { + "epoch": 0.8368129142805609, + "grad_norm": 1.6320456570381554, + "learning_rate": 2.727890290739212e-07, + "loss": 1.029, + "step": 9279 + }, + { + "epoch": 0.8369030978040312, + "grad_norm": 1.4764711059058038, + "learning_rate": 2.7249456234968395e-07, + "loss": 1.0104, + "step": 9280 + }, + { + "epoch": 0.8369932813275015, + "grad_norm": 2.591504331290357, + "learning_rate": 2.722002430257364e-07, + "loss": 1.0459, + "step": 9281 + }, + { + "epoch": 0.8370834648509717, + "grad_norm": 1.4875746123250042, + "learning_rate": 2.7190607112719035e-07, + "loss": 0.9423, + "step": 9282 + }, + { + "epoch": 0.837173648374442, + "grad_norm": 2.81669261815081, + "learning_rate": 2.716120466791476e-07, + "loss": 0.8971, + "step": 9283 + }, + { + "epoch": 0.8372638318979122, + "grad_norm": 1.7812441775578154, + "learning_rate": 2.7131816970669483e-07, + "loss": 0.9572, + "step": 9284 + }, + { + "epoch": 0.8373540154213825, + "grad_norm": 1.7862230448093837, + "learning_rate": 2.7102444023490777e-07, + "loss": 0.9829, + "step": 9285 + }, + { + "epoch": 0.8374441989448528, + "grad_norm": 1.3839436877431524, + "learning_rate": 2.70730858288849e-07, + "loss": 1.0333, + "step": 9286 + }, + { + "epoch": 0.837534382468323, + "grad_norm": 1.7502458263489669, + "learning_rate": 2.704374238935685e-07, + "loss": 0.9155, + "step": 9287 + }, + { + "epoch": 0.8376245659917932, + "grad_norm": 0.6688307615338914, + "learning_rate": 2.70144137074104e-07, + "loss": 0.811, + "step": 9288 + }, + { + "epoch": 0.8377147495152636, + "grad_norm": 1.5159107214222334, + "learning_rate": 2.6985099785547926e-07, + "loss": 0.9992, + "step": 9289 + }, + { + "epoch": 0.8378049330387338, + "grad_norm": 1.3791289325327158, + "learning_rate": 2.695580062627083e-07, + "loss": 0.9997, + "step": 9290 + }, + { + "epoch": 0.8378951165622041, + "grad_norm": 1.7568041134337957, + "learning_rate": 2.692651623207891e-07, + "loss": 0.9766, + "step": 9291 + }, + { + "epoch": 0.8379853000856744, + "grad_norm": 1.3450326785829383, + "learning_rate": 2.689724660547097e-07, + "loss": 0.964, + "step": 9292 + }, + { + "epoch": 0.8380754836091446, + "grad_norm": 0.6178606201557555, + "learning_rate": 2.686799174894441e-07, + "loss": 0.8077, + "step": 9293 + }, + { + "epoch": 0.8381656671326149, + "grad_norm": 0.6467000826168877, + "learning_rate": 2.683875166499545e-07, + "loss": 0.8082, + "step": 9294 + }, + { + "epoch": 0.8382558506560851, + "grad_norm": 6.581426738745974, + "learning_rate": 2.680952635611899e-07, + "loss": 0.9819, + "step": 9295 + }, + { + "epoch": 0.8383460341795554, + "grad_norm": 2.588478522550756, + "learning_rate": 2.678031582480875e-07, + "loss": 0.9036, + "step": 9296 + }, + { + "epoch": 0.8384362177030257, + "grad_norm": 1.6465601060856785, + "learning_rate": 2.6751120073557e-07, + "loss": 1.0192, + "step": 9297 + }, + { + "epoch": 0.8385264012264959, + "grad_norm": 1.833045727394069, + "learning_rate": 2.672193910485505e-07, + "loss": 0.9464, + "step": 9298 + }, + { + "epoch": 0.8386165847499661, + "grad_norm": 1.5923817314893838, + "learning_rate": 2.669277292119265e-07, + "loss": 0.9775, + "step": 9299 + }, + { + "epoch": 0.8387067682734365, + "grad_norm": 1.9437313164797398, + "learning_rate": 2.666362152505848e-07, + "loss": 0.9786, + "step": 9300 + }, + { + "epoch": 0.8387969517969067, + "grad_norm": 1.7879169244646775, + "learning_rate": 2.663448491893989e-07, + "loss": 0.9139, + "step": 9301 + }, + { + "epoch": 0.838887135320377, + "grad_norm": 1.6122012460429094, + "learning_rate": 2.6605363105322974e-07, + "loss": 0.9549, + "step": 9302 + }, + { + "epoch": 0.8389773188438472, + "grad_norm": 4.358628732117301, + "learning_rate": 2.657625608669263e-07, + "loss": 0.9734, + "step": 9303 + }, + { + "epoch": 0.8390675023673175, + "grad_norm": 1.4060971918794125, + "learning_rate": 2.654716386553224e-07, + "loss": 0.8921, + "step": 9304 + }, + { + "epoch": 0.8391576858907878, + "grad_norm": 2.2557518871510442, + "learning_rate": 2.651808644432436e-07, + "loss": 0.9739, + "step": 9305 + }, + { + "epoch": 0.839247869414258, + "grad_norm": 1.7580248217040102, + "learning_rate": 2.6489023825549807e-07, + "loss": 1.1091, + "step": 9306 + }, + { + "epoch": 0.8393380529377282, + "grad_norm": 2.0205447459544303, + "learning_rate": 2.6459976011688547e-07, + "loss": 0.9009, + "step": 9307 + }, + { + "epoch": 0.8394282364611986, + "grad_norm": 2.2022878801472476, + "learning_rate": 2.6430943005219e-07, + "loss": 0.9702, + "step": 9308 + }, + { + "epoch": 0.8395184199846688, + "grad_norm": 1.4290147876761867, + "learning_rate": 2.6401924808618447e-07, + "loss": 1.0032, + "step": 9309 + }, + { + "epoch": 0.839608603508139, + "grad_norm": 2.5917368050393272, + "learning_rate": 2.637292142436287e-07, + "loss": 0.9752, + "step": 9310 + }, + { + "epoch": 0.8396987870316093, + "grad_norm": 1.7997044055672622, + "learning_rate": 2.6343932854927e-07, + "loss": 0.9485, + "step": 9311 + }, + { + "epoch": 0.8397889705550796, + "grad_norm": 0.6845151740273838, + "learning_rate": 2.6314959102784316e-07, + "loss": 0.8065, + "step": 9312 + }, + { + "epoch": 0.8398791540785498, + "grad_norm": 1.9250585968214808, + "learning_rate": 2.6286000170407074e-07, + "loss": 0.9654, + "step": 9313 + }, + { + "epoch": 0.8399693376020201, + "grad_norm": 2.8683909870968427, + "learning_rate": 2.625705606026607e-07, + "loss": 0.9541, + "step": 9314 + }, + { + "epoch": 0.8400595211254904, + "grad_norm": 1.3753935077423594, + "learning_rate": 2.622812677483106e-07, + "loss": 1.0557, + "step": 9315 + }, + { + "epoch": 0.8401497046489607, + "grad_norm": 3.626793121933508, + "learning_rate": 2.6199212316570453e-07, + "loss": 1.0309, + "step": 9316 + }, + { + "epoch": 0.8402398881724309, + "grad_norm": 1.6652330033907161, + "learning_rate": 2.617031268795138e-07, + "loss": 0.9563, + "step": 9317 + }, + { + "epoch": 0.8403300716959011, + "grad_norm": 1.3416941241106464, + "learning_rate": 2.614142789143976e-07, + "loss": 1.0036, + "step": 9318 + }, + { + "epoch": 0.8404202552193715, + "grad_norm": 1.8017140783775873, + "learning_rate": 2.6112557929500047e-07, + "loss": 0.9623, + "step": 9319 + }, + { + "epoch": 0.8405104387428417, + "grad_norm": 1.5893672444600244, + "learning_rate": 2.6083702804595817e-07, + "loss": 0.999, + "step": 9320 + }, + { + "epoch": 0.8406006222663119, + "grad_norm": 2.0581999371298143, + "learning_rate": 2.6054862519188915e-07, + "loss": 0.9612, + "step": 9321 + }, + { + "epoch": 0.8406908057897822, + "grad_norm": 1.8243421231032588, + "learning_rate": 2.6026037075740357e-07, + "loss": 0.9111, + "step": 9322 + }, + { + "epoch": 0.8407809893132525, + "grad_norm": 2.1477193846014138, + "learning_rate": 2.5997226476709524e-07, + "loss": 0.9456, + "step": 9323 + }, + { + "epoch": 0.8408711728367227, + "grad_norm": 2.48728302435845, + "learning_rate": 2.5968430724554856e-07, + "loss": 0.9779, + "step": 9324 + }, + { + "epoch": 0.840961356360193, + "grad_norm": 2.450932294224049, + "learning_rate": 2.5939649821733225e-07, + "loss": 0.9112, + "step": 9325 + }, + { + "epoch": 0.8410515398836632, + "grad_norm": 2.9282362942221387, + "learning_rate": 2.5910883770700433e-07, + "loss": 0.9403, + "step": 9326 + }, + { + "epoch": 0.8411417234071336, + "grad_norm": 1.4437193466829674, + "learning_rate": 2.5882132573910965e-07, + "loss": 0.8183, + "step": 9327 + }, + { + "epoch": 0.8412319069306038, + "grad_norm": 1.3292231339485157, + "learning_rate": 2.585339623381801e-07, + "loss": 1.0151, + "step": 9328 + }, + { + "epoch": 0.841322090454074, + "grad_norm": 3.047525659191833, + "learning_rate": 2.582467475287358e-07, + "loss": 0.9531, + "step": 9329 + }, + { + "epoch": 0.8414122739775443, + "grad_norm": 2.0426180303268064, + "learning_rate": 2.5795968133528224e-07, + "loss": 0.9888, + "step": 9330 + }, + { + "epoch": 0.8415024575010146, + "grad_norm": 1.775770694350536, + "learning_rate": 2.576727637823144e-07, + "loss": 0.9762, + "step": 9331 + }, + { + "epoch": 0.8415926410244848, + "grad_norm": 3.7429258378733374, + "learning_rate": 2.5738599489431335e-07, + "loss": 1.0214, + "step": 9332 + }, + { + "epoch": 0.8416828245479551, + "grad_norm": 1.310082570815222, + "learning_rate": 2.5709937469574794e-07, + "loss": 1.0368, + "step": 9333 + }, + { + "epoch": 0.8417730080714253, + "grad_norm": 1.3501570345356828, + "learning_rate": 2.568129032110742e-07, + "loss": 1.0158, + "step": 9334 + }, + { + "epoch": 0.8418631915948956, + "grad_norm": 1.5437332770179277, + "learning_rate": 2.5652658046473565e-07, + "loss": 1.0517, + "step": 9335 + }, + { + "epoch": 0.8419533751183659, + "grad_norm": 2.8886566883897618, + "learning_rate": 2.5624040648116184e-07, + "loss": 0.9781, + "step": 9336 + }, + { + "epoch": 0.8420435586418361, + "grad_norm": 1.5877083348817036, + "learning_rate": 2.5595438128477245e-07, + "loss": 0.913, + "step": 9337 + }, + { + "epoch": 0.8421337421653065, + "grad_norm": 1.4567586117030495, + "learning_rate": 2.5566850489997096e-07, + "loss": 0.9817, + "step": 9338 + }, + { + "epoch": 0.8422239256887767, + "grad_norm": 5.203102088854869, + "learning_rate": 2.5538277735115166e-07, + "loss": 1.0056, + "step": 9339 + }, + { + "epoch": 0.8423141092122469, + "grad_norm": 1.928811928724618, + "learning_rate": 2.5509719866269306e-07, + "loss": 1.0304, + "step": 9340 + }, + { + "epoch": 0.8424042927357172, + "grad_norm": 4.152211227527199, + "learning_rate": 2.548117688589628e-07, + "loss": 1.0386, + "step": 9341 + }, + { + "epoch": 0.8424944762591875, + "grad_norm": 1.6596321984229996, + "learning_rate": 2.545264879643152e-07, + "loss": 0.9614, + "step": 9342 + }, + { + "epoch": 0.8425846597826577, + "grad_norm": 1.7208736824434598, + "learning_rate": 2.542413560030923e-07, + "loss": 0.935, + "step": 9343 + }, + { + "epoch": 0.842674843306128, + "grad_norm": 1.720310473809015, + "learning_rate": 2.53956372999623e-07, + "loss": 0.8709, + "step": 9344 + }, + { + "epoch": 0.8427650268295982, + "grad_norm": 0.643242003175062, + "learning_rate": 2.5367153897822293e-07, + "loss": 0.8463, + "step": 9345 + }, + { + "epoch": 0.8428552103530685, + "grad_norm": 1.4188012235695893, + "learning_rate": 2.5338685396319715e-07, + "loss": 0.9698, + "step": 9346 + }, + { + "epoch": 0.8429453938765388, + "grad_norm": 1.9221428785259798, + "learning_rate": 2.531023179788352e-07, + "loss": 0.9096, + "step": 9347 + }, + { + "epoch": 0.843035577400009, + "grad_norm": 1.8513831200604787, + "learning_rate": 2.528179310494158e-07, + "loss": 0.9483, + "step": 9348 + }, + { + "epoch": 0.8431257609234792, + "grad_norm": 1.9137246592791772, + "learning_rate": 2.5253369319920436e-07, + "loss": 0.9973, + "step": 9349 + }, + { + "epoch": 0.8432159444469496, + "grad_norm": 3.094472473248924, + "learning_rate": 2.522496044524538e-07, + "loss": 0.9771, + "step": 9350 + }, + { + "epoch": 0.8433061279704198, + "grad_norm": 1.7651546155515607, + "learning_rate": 2.5196566483340386e-07, + "loss": 1.0677, + "step": 9351 + }, + { + "epoch": 0.84339631149389, + "grad_norm": 1.812884454402273, + "learning_rate": 2.516818743662825e-07, + "loss": 0.9924, + "step": 9352 + }, + { + "epoch": 0.8434864950173603, + "grad_norm": 2.057659595923318, + "learning_rate": 2.5139823307530285e-07, + "loss": 0.9774, + "step": 9353 + }, + { + "epoch": 0.8435766785408306, + "grad_norm": 2.5135500862833196, + "learning_rate": 2.5111474098466836e-07, + "loss": 1.0168, + "step": 9354 + }, + { + "epoch": 0.8436668620643009, + "grad_norm": 1.8952781856540837, + "learning_rate": 2.50831398118567e-07, + "loss": 1.024, + "step": 9355 + }, + { + "epoch": 0.8437570455877711, + "grad_norm": 1.6119452387167066, + "learning_rate": 2.5054820450117576e-07, + "loss": 0.92, + "step": 9356 + }, + { + "epoch": 0.8438472291112413, + "grad_norm": 1.5601264473213283, + "learning_rate": 2.502651601566579e-07, + "loss": 0.9377, + "step": 9357 + }, + { + "epoch": 0.8439374126347117, + "grad_norm": 1.8345452406745895, + "learning_rate": 2.499822651091645e-07, + "loss": 0.933, + "step": 9358 + }, + { + "epoch": 0.8440275961581819, + "grad_norm": 1.937150492984658, + "learning_rate": 2.496995193828344e-07, + "loss": 0.9987, + "step": 9359 + }, + { + "epoch": 0.8441177796816521, + "grad_norm": 1.7743667655020103, + "learning_rate": 2.494169230017913e-07, + "loss": 0.9728, + "step": 9360 + }, + { + "epoch": 0.8442079632051224, + "grad_norm": 1.970687276362498, + "learning_rate": 2.491344759901499e-07, + "loss": 0.9374, + "step": 9361 + }, + { + "epoch": 0.8442981467285927, + "grad_norm": 1.7714792606906604, + "learning_rate": 2.488521783720088e-07, + "loss": 0.9706, + "step": 9362 + }, + { + "epoch": 0.844388330252063, + "grad_norm": 1.822598246057362, + "learning_rate": 2.4857003017145526e-07, + "loss": 1.0116, + "step": 9363 + }, + { + "epoch": 0.8444785137755332, + "grad_norm": 1.556011766943286, + "learning_rate": 2.482880314125644e-07, + "loss": 1.0135, + "step": 9364 + }, + { + "epoch": 0.8445686972990035, + "grad_norm": 1.8222136822296209, + "learning_rate": 2.4800618211939726e-07, + "loss": 0.969, + "step": 9365 + }, + { + "epoch": 0.8446588808224738, + "grad_norm": 1.740348358360264, + "learning_rate": 2.477244823160034e-07, + "loss": 0.9844, + "step": 9366 + }, + { + "epoch": 0.844749064345944, + "grad_norm": 2.6559429215594577, + "learning_rate": 2.474429320264184e-07, + "loss": 0.9975, + "step": 9367 + }, + { + "epoch": 0.8448392478694142, + "grad_norm": 1.4083844410728916, + "learning_rate": 2.47161531274666e-07, + "loss": 0.9363, + "step": 9368 + }, + { + "epoch": 0.8449294313928846, + "grad_norm": 1.5474126632769847, + "learning_rate": 2.4688028008475714e-07, + "loss": 0.9091, + "step": 9369 + }, + { + "epoch": 0.8450196149163548, + "grad_norm": 2.5999629054724536, + "learning_rate": 2.465991784806891e-07, + "loss": 0.9666, + "step": 9370 + }, + { + "epoch": 0.845109798439825, + "grad_norm": 3.476273359640646, + "learning_rate": 2.463182264864472e-07, + "loss": 1.0556, + "step": 9371 + }, + { + "epoch": 0.8451999819632953, + "grad_norm": 1.877246401100707, + "learning_rate": 2.460374241260039e-07, + "loss": 0.9026, + "step": 9372 + }, + { + "epoch": 0.8452901654867656, + "grad_norm": 1.8284099389403006, + "learning_rate": 2.4575677142331884e-07, + "loss": 1.0405, + "step": 9373 + }, + { + "epoch": 0.8453803490102358, + "grad_norm": 1.5152602395871495, + "learning_rate": 2.454762684023395e-07, + "loss": 0.9183, + "step": 9374 + }, + { + "epoch": 0.8454705325337061, + "grad_norm": 1.7244399474613528, + "learning_rate": 2.4519591508699823e-07, + "loss": 0.9253, + "step": 9375 + }, + { + "epoch": 0.8455607160571763, + "grad_norm": 1.9408495467499554, + "learning_rate": 2.4491571150121815e-07, + "loss": 1.029, + "step": 9376 + }, + { + "epoch": 0.8456508995806467, + "grad_norm": 6.205182942437805, + "learning_rate": 2.446356576689062e-07, + "loss": 0.9729, + "step": 9377 + }, + { + "epoch": 0.8457410831041169, + "grad_norm": 2.0466554247312336, + "learning_rate": 2.4435575361395976e-07, + "loss": 0.9843, + "step": 9378 + }, + { + "epoch": 0.8458312666275871, + "grad_norm": 5.968595972120911, + "learning_rate": 2.440759993602606e-07, + "loss": 0.9551, + "step": 9379 + }, + { + "epoch": 0.8459214501510574, + "grad_norm": 1.749504359765084, + "learning_rate": 2.437963949316793e-07, + "loss": 0.9204, + "step": 9380 + }, + { + "epoch": 0.8460116336745277, + "grad_norm": 1.6520843662997942, + "learning_rate": 2.435169403520729e-07, + "loss": 0.9704, + "step": 9381 + }, + { + "epoch": 0.8461018171979979, + "grad_norm": 2.3858287294347615, + "learning_rate": 2.4323763564528653e-07, + "loss": 1.0235, + "step": 9382 + }, + { + "epoch": 0.8461920007214682, + "grad_norm": 1.6006785861112642, + "learning_rate": 2.429584808351517e-07, + "loss": 0.9935, + "step": 9383 + }, + { + "epoch": 0.8462821842449384, + "grad_norm": 2.8930261485200073, + "learning_rate": 2.42679475945488e-07, + "loss": 1.0918, + "step": 9384 + }, + { + "epoch": 0.8463723677684087, + "grad_norm": 1.3455290882534352, + "learning_rate": 2.424006210001008e-07, + "loss": 0.9758, + "step": 9385 + }, + { + "epoch": 0.846462551291879, + "grad_norm": 2.816678313131538, + "learning_rate": 2.421219160227839e-07, + "loss": 1.0038, + "step": 9386 + }, + { + "epoch": 0.8465527348153492, + "grad_norm": 2.485062126288776, + "learning_rate": 2.4184336103731785e-07, + "loss": 0.9731, + "step": 9387 + }, + { + "epoch": 0.8466429183388196, + "grad_norm": 3.0781076498559243, + "learning_rate": 2.4156495606747065e-07, + "loss": 0.9944, + "step": 9388 + }, + { + "epoch": 0.8467331018622898, + "grad_norm": 2.0085496551172275, + "learning_rate": 2.412867011369972e-07, + "loss": 1.0888, + "step": 9389 + }, + { + "epoch": 0.84682328538576, + "grad_norm": 1.9923068141737053, + "learning_rate": 2.4100859626963997e-07, + "loss": 0.9843, + "step": 9390 + }, + { + "epoch": 0.8469134689092302, + "grad_norm": 1.6617551038286924, + "learning_rate": 2.407306414891288e-07, + "loss": 0.9744, + "step": 9391 + }, + { + "epoch": 0.8470036524327006, + "grad_norm": 1.4083595136131455, + "learning_rate": 2.4045283681917893e-07, + "loss": 1.0784, + "step": 9392 + }, + { + "epoch": 0.8470938359561708, + "grad_norm": 1.742704716701673, + "learning_rate": 2.4017518228349586e-07, + "loss": 1.0895, + "step": 9393 + }, + { + "epoch": 0.8471840194796411, + "grad_norm": 0.6177575581210578, + "learning_rate": 2.3989767790576887e-07, + "loss": 0.8282, + "step": 9394 + }, + { + "epoch": 0.8472742030031113, + "grad_norm": 3.0222873915240642, + "learning_rate": 2.396203237096781e-07, + "loss": 0.9425, + "step": 9395 + }, + { + "epoch": 0.8473643865265816, + "grad_norm": 1.9391895433930841, + "learning_rate": 2.393431197188873e-07, + "loss": 0.9688, + "step": 9396 + }, + { + "epoch": 0.8474545700500519, + "grad_norm": 2.3333177452474976, + "learning_rate": 2.3906606595705004e-07, + "loss": 1.0488, + "step": 9397 + }, + { + "epoch": 0.8475447535735221, + "grad_norm": 1.5277008653797481, + "learning_rate": 2.387891624478056e-07, + "loss": 1.0274, + "step": 9398 + }, + { + "epoch": 0.8476349370969923, + "grad_norm": 1.4883181750104808, + "learning_rate": 2.3851240921478075e-07, + "loss": 0.9973, + "step": 9399 + }, + { + "epoch": 0.8477251206204627, + "grad_norm": 1.6344359144015168, + "learning_rate": 2.3823580628159057e-07, + "loss": 0.9577, + "step": 9400 + }, + { + "epoch": 0.8478153041439329, + "grad_norm": 2.9748203255387176, + "learning_rate": 2.3795935367183517e-07, + "loss": 0.958, + "step": 9401 + }, + { + "epoch": 0.8479054876674031, + "grad_norm": 1.3801776330760414, + "learning_rate": 2.376830514091035e-07, + "loss": 0.9993, + "step": 9402 + }, + { + "epoch": 0.8479956711908734, + "grad_norm": 1.5596288623460837, + "learning_rate": 2.3740689951697135e-07, + "loss": 1.0573, + "step": 9403 + }, + { + "epoch": 0.8480858547143437, + "grad_norm": 1.4988019450319374, + "learning_rate": 2.371308980190012e-07, + "loss": 0.9763, + "step": 9404 + }, + { + "epoch": 0.848176038237814, + "grad_norm": 0.6224911163263028, + "learning_rate": 2.3685504693874337e-07, + "loss": 0.8341, + "step": 9405 + }, + { + "epoch": 0.8482662217612842, + "grad_norm": 1.9422497519155248, + "learning_rate": 2.3657934629973497e-07, + "loss": 1.0103, + "step": 9406 + }, + { + "epoch": 0.8483564052847544, + "grad_norm": 1.8267091664477353, + "learning_rate": 2.3630379612549944e-07, + "loss": 0.9811, + "step": 9407 + }, + { + "epoch": 0.8484465888082248, + "grad_norm": 1.037171097982328, + "learning_rate": 2.3602839643954997e-07, + "loss": 0.8581, + "step": 9408 + }, + { + "epoch": 0.848536772331695, + "grad_norm": 1.7901801070915786, + "learning_rate": 2.3575314726538308e-07, + "loss": 0.9805, + "step": 9409 + }, + { + "epoch": 0.8486269558551652, + "grad_norm": 4.835859800340051, + "learning_rate": 2.3547804862648645e-07, + "loss": 1.0095, + "step": 9410 + }, + { + "epoch": 0.8487171393786356, + "grad_norm": 1.9701165194835042, + "learning_rate": 2.3520310054633174e-07, + "loss": 1.0231, + "step": 9411 + }, + { + "epoch": 0.8488073229021058, + "grad_norm": 1.6761113621944017, + "learning_rate": 2.3492830304837973e-07, + "loss": 0.9345, + "step": 9412 + }, + { + "epoch": 0.848897506425576, + "grad_norm": 5.206933995774532, + "learning_rate": 2.3465365615607723e-07, + "loss": 0.9245, + "step": 9413 + }, + { + "epoch": 0.8489876899490463, + "grad_norm": 1.7555931492024948, + "learning_rate": 2.3437915989285884e-07, + "loss": 0.9701, + "step": 9414 + }, + { + "epoch": 0.8490778734725166, + "grad_norm": 2.102585554162599, + "learning_rate": 2.3410481428214647e-07, + "loss": 1.0413, + "step": 9415 + }, + { + "epoch": 0.8491680569959869, + "grad_norm": 1.519958592904921, + "learning_rate": 2.338306193473476e-07, + "loss": 1.0035, + "step": 9416 + }, + { + "epoch": 0.8492582405194571, + "grad_norm": 1.4308859382931791, + "learning_rate": 2.3355657511185957e-07, + "loss": 0.9794, + "step": 9417 + }, + { + "epoch": 0.8493484240429273, + "grad_norm": 0.6583971365862329, + "learning_rate": 2.3328268159906428e-07, + "loss": 0.7988, + "step": 9418 + }, + { + "epoch": 0.8494386075663977, + "grad_norm": 1.6799219721884708, + "learning_rate": 2.330089388323322e-07, + "loss": 1.0251, + "step": 9419 + }, + { + "epoch": 0.8495287910898679, + "grad_norm": 1.9871600574098853, + "learning_rate": 2.327353468350204e-07, + "loss": 1.0171, + "step": 9420 + }, + { + "epoch": 0.8496189746133381, + "grad_norm": 1.4944245511227434, + "learning_rate": 2.3246190563047352e-07, + "loss": 0.9687, + "step": 9421 + }, + { + "epoch": 0.8497091581368084, + "grad_norm": 1.7866503183247955, + "learning_rate": 2.3218861524202293e-07, + "loss": 0.9621, + "step": 9422 + }, + { + "epoch": 0.8497993416602787, + "grad_norm": 1.7588889661815565, + "learning_rate": 2.3191547569298775e-07, + "loss": 0.9885, + "step": 9423 + }, + { + "epoch": 0.8498895251837489, + "grad_norm": 4.407489791572679, + "learning_rate": 2.3164248700667245e-07, + "loss": 0.9663, + "step": 9424 + }, + { + "epoch": 0.8499797087072192, + "grad_norm": 1.9887475926392453, + "learning_rate": 2.313696492063717e-07, + "loss": 1.0102, + "step": 9425 + }, + { + "epoch": 0.8500698922306894, + "grad_norm": 2.5423092785384944, + "learning_rate": 2.3109696231536401e-07, + "loss": 0.9661, + "step": 9426 + }, + { + "epoch": 0.8501600757541597, + "grad_norm": 2.2019913374187325, + "learning_rate": 2.3082442635691722e-07, + "loss": 0.9391, + "step": 9427 + }, + { + "epoch": 0.85025025927763, + "grad_norm": 1.591355413524754, + "learning_rate": 2.305520413542854e-07, + "loss": 0.9931, + "step": 9428 + }, + { + "epoch": 0.8503404428011002, + "grad_norm": 1.5612417113646855, + "learning_rate": 2.3027980733071018e-07, + "loss": 0.9014, + "step": 9429 + }, + { + "epoch": 0.8504306263245704, + "grad_norm": 0.6205872923967746, + "learning_rate": 2.3000772430942027e-07, + "loss": 0.7817, + "step": 9430 + }, + { + "epoch": 0.8505208098480408, + "grad_norm": 2.6927738341598677, + "learning_rate": 2.2973579231363028e-07, + "loss": 0.9956, + "step": 9431 + }, + { + "epoch": 0.850610993371511, + "grad_norm": 1.8785813143820747, + "learning_rate": 2.2946401136654446e-07, + "loss": 1.0015, + "step": 9432 + }, + { + "epoch": 0.8507011768949813, + "grad_norm": 1.9298500888298151, + "learning_rate": 2.2919238149135077e-07, + "loss": 1.0032, + "step": 9433 + }, + { + "epoch": 0.8507913604184516, + "grad_norm": 1.7268733784698704, + "learning_rate": 2.289209027112282e-07, + "loss": 0.9829, + "step": 9434 + }, + { + "epoch": 0.8508815439419218, + "grad_norm": 1.9035727664755704, + "learning_rate": 2.2864957504933934e-07, + "loss": 1.0393, + "step": 9435 + }, + { + "epoch": 0.8509717274653921, + "grad_norm": 1.6422832283262305, + "learning_rate": 2.2837839852883589e-07, + "loss": 0.9601, + "step": 9436 + }, + { + "epoch": 0.8510619109888623, + "grad_norm": 1.692350410375766, + "learning_rate": 2.2810737317285623e-07, + "loss": 0.9708, + "step": 9437 + }, + { + "epoch": 0.8511520945123326, + "grad_norm": 1.499164428044857, + "learning_rate": 2.278364990045254e-07, + "loss": 0.9179, + "step": 9438 + }, + { + "epoch": 0.8512422780358029, + "grad_norm": 2.066920542761703, + "learning_rate": 2.2756577604695625e-07, + "loss": 0.9677, + "step": 9439 + }, + { + "epoch": 0.8513324615592731, + "grad_norm": 1.6157172702435538, + "learning_rate": 2.2729520432324855e-07, + "loss": 0.9963, + "step": 9440 + }, + { + "epoch": 0.8514226450827433, + "grad_norm": 1.729167785031367, + "learning_rate": 2.2702478385648826e-07, + "loss": 0.8916, + "step": 9441 + }, + { + "epoch": 0.8515128286062137, + "grad_norm": 1.570942387054412, + "learning_rate": 2.2675451466974938e-07, + "loss": 0.9729, + "step": 9442 + }, + { + "epoch": 0.8516030121296839, + "grad_norm": 1.8714006208144196, + "learning_rate": 2.26484396786093e-07, + "loss": 0.9091, + "step": 9443 + }, + { + "epoch": 0.8516931956531542, + "grad_norm": 0.6403830699468017, + "learning_rate": 2.2621443022856667e-07, + "loss": 0.8395, + "step": 9444 + }, + { + "epoch": 0.8517833791766244, + "grad_norm": 1.7449397769143735, + "learning_rate": 2.2594461502020646e-07, + "loss": 1.016, + "step": 9445 + }, + { + "epoch": 0.8518735627000947, + "grad_norm": 1.9272148568574774, + "learning_rate": 2.2567495118403278e-07, + "loss": 0.8576, + "step": 9446 + }, + { + "epoch": 0.851963746223565, + "grad_norm": 1.4756319858073834, + "learning_rate": 2.254054387430566e-07, + "loss": 0.9913, + "step": 9447 + }, + { + "epoch": 0.8520539297470352, + "grad_norm": 3.2832827130076927, + "learning_rate": 2.2513607772027243e-07, + "loss": 0.961, + "step": 9448 + }, + { + "epoch": 0.8521441132705054, + "grad_norm": 1.7924424273663502, + "learning_rate": 2.2486686813866562e-07, + "loss": 0.9397, + "step": 9449 + }, + { + "epoch": 0.8522342967939758, + "grad_norm": 2.2884113309958773, + "learning_rate": 2.2459781002120514e-07, + "loss": 1.0285, + "step": 9450 + }, + { + "epoch": 0.852324480317446, + "grad_norm": 1.4641837705953127, + "learning_rate": 2.243289033908491e-07, + "loss": 1.0419, + "step": 9451 + }, + { + "epoch": 0.8524146638409162, + "grad_norm": 2.2604567938786593, + "learning_rate": 2.2406014827054176e-07, + "loss": 0.9888, + "step": 9452 + }, + { + "epoch": 0.8525048473643865, + "grad_norm": 1.7286963761647, + "learning_rate": 2.2379154468321525e-07, + "loss": 1.0384, + "step": 9453 + }, + { + "epoch": 0.8525950308878568, + "grad_norm": 1.616946363669497, + "learning_rate": 2.2352309265178793e-07, + "loss": 1.0111, + "step": 9454 + }, + { + "epoch": 0.852685214411327, + "grad_norm": 2.5824408271266064, + "learning_rate": 2.2325479219916565e-07, + "loss": 1.052, + "step": 9455 + }, + { + "epoch": 0.8527753979347973, + "grad_norm": 1.855199212022049, + "learning_rate": 2.229866433482419e-07, + "loss": 1.0084, + "step": 9456 + }, + { + "epoch": 0.8528655814582676, + "grad_norm": 2.93748019597297, + "learning_rate": 2.2271864612189552e-07, + "loss": 1.0292, + "step": 9457 + }, + { + "epoch": 0.8529557649817379, + "grad_norm": 1.5988756669965911, + "learning_rate": 2.2245080054299415e-07, + "loss": 1.0098, + "step": 9458 + }, + { + "epoch": 0.8530459485052081, + "grad_norm": 8.4671372570709, + "learning_rate": 2.2218310663439198e-07, + "loss": 0.9306, + "step": 9459 + }, + { + "epoch": 0.8531361320286783, + "grad_norm": 1.7558130039630826, + "learning_rate": 2.2191556441892968e-07, + "loss": 1.002, + "step": 9460 + }, + { + "epoch": 0.8532263155521487, + "grad_norm": 1.8551012177113413, + "learning_rate": 2.216481739194358e-07, + "loss": 0.9853, + "step": 9461 + }, + { + "epoch": 0.8533164990756189, + "grad_norm": 1.8165677562816729, + "learning_rate": 2.2138093515872592e-07, + "loss": 0.9248, + "step": 9462 + }, + { + "epoch": 0.8534066825990891, + "grad_norm": 1.3948026761605026, + "learning_rate": 2.2111384815960087e-07, + "loss": 1.0591, + "step": 9463 + }, + { + "epoch": 0.8534968661225594, + "grad_norm": 1.4544005179533335, + "learning_rate": 2.208469129448518e-07, + "loss": 0.9016, + "step": 9464 + }, + { + "epoch": 0.8535870496460297, + "grad_norm": 0.620213322434131, + "learning_rate": 2.2058012953725357e-07, + "loss": 0.8052, + "step": 9465 + }, + { + "epoch": 0.8536772331695, + "grad_norm": 1.7520270188482168, + "learning_rate": 2.203134979595711e-07, + "loss": 0.9453, + "step": 9466 + }, + { + "epoch": 0.8537674166929702, + "grad_norm": 1.5932958367190553, + "learning_rate": 2.2004701823455374e-07, + "loss": 0.9297, + "step": 9467 + }, + { + "epoch": 0.8538576002164404, + "grad_norm": 2.011943797143695, + "learning_rate": 2.1978069038493906e-07, + "loss": 1.0098, + "step": 9468 + }, + { + "epoch": 0.8539477837399108, + "grad_norm": 1.3738002744991236, + "learning_rate": 2.1951451443345225e-07, + "loss": 0.9788, + "step": 9469 + }, + { + "epoch": 0.854037967263381, + "grad_norm": 2.1143855736874317, + "learning_rate": 2.1924849040280425e-07, + "loss": 0.964, + "step": 9470 + }, + { + "epoch": 0.8541281507868512, + "grad_norm": 1.8527218050352372, + "learning_rate": 2.1898261831569465e-07, + "loss": 1.0601, + "step": 9471 + }, + { + "epoch": 0.8542183343103215, + "grad_norm": 1.5052927734907053, + "learning_rate": 2.1871689819480798e-07, + "loss": 0.986, + "step": 9472 + }, + { + "epoch": 0.8543085178337918, + "grad_norm": 1.3264784985153815, + "learning_rate": 2.1845133006281745e-07, + "loss": 0.9624, + "step": 9473 + }, + { + "epoch": 0.854398701357262, + "grad_norm": 1.5030665047221956, + "learning_rate": 2.1818591394238294e-07, + "loss": 0.8336, + "step": 9474 + }, + { + "epoch": 0.8544888848807323, + "grad_norm": 2.926072313730586, + "learning_rate": 2.1792064985615076e-07, + "loss": 0.9566, + "step": 9475 + }, + { + "epoch": 0.8545790684042025, + "grad_norm": 1.3970595662436858, + "learning_rate": 2.1765553782675528e-07, + "loss": 0.948, + "step": 9476 + }, + { + "epoch": 0.8546692519276728, + "grad_norm": 1.454821017291232, + "learning_rate": 2.1739057787681703e-07, + "loss": 1.0515, + "step": 9477 + }, + { + "epoch": 0.8547594354511431, + "grad_norm": 1.8186244750866731, + "learning_rate": 2.1712577002894372e-07, + "loss": 1.0031, + "step": 9478 + }, + { + "epoch": 0.8548496189746133, + "grad_norm": 1.4611077031014024, + "learning_rate": 2.1686111430573105e-07, + "loss": 0.9495, + "step": 9479 + }, + { + "epoch": 0.8549398024980835, + "grad_norm": 4.17404221381504, + "learning_rate": 2.165966107297592e-07, + "loss": 1.0295, + "step": 9480 + }, + { + "epoch": 0.8550299860215539, + "grad_norm": 2.508749718392836, + "learning_rate": 2.16332259323599e-07, + "loss": 1.0055, + "step": 9481 + }, + { + "epoch": 0.8551201695450241, + "grad_norm": 2.11214022761155, + "learning_rate": 2.1606806010980504e-07, + "loss": 0.9248, + "step": 9482 + }, + { + "epoch": 0.8552103530684944, + "grad_norm": 1.7794586677563757, + "learning_rate": 2.1580401311092067e-07, + "loss": 0.9765, + "step": 9483 + }, + { + "epoch": 0.8553005365919647, + "grad_norm": 1.613232997514435, + "learning_rate": 2.1554011834947604e-07, + "loss": 0.9672, + "step": 9484 + }, + { + "epoch": 0.8553907201154349, + "grad_norm": 1.6006705428752706, + "learning_rate": 2.1527637584798764e-07, + "loss": 1.0071, + "step": 9485 + }, + { + "epoch": 0.8554809036389052, + "grad_norm": 1.6856986427274057, + "learning_rate": 2.150127856289603e-07, + "loss": 1.046, + "step": 9486 + }, + { + "epoch": 0.8555710871623754, + "grad_norm": 0.6108540899992394, + "learning_rate": 2.1474934771488363e-07, + "loss": 0.7939, + "step": 9487 + }, + { + "epoch": 0.8556612706858457, + "grad_norm": 1.7778055450469705, + "learning_rate": 2.1448606212823715e-07, + "loss": 1.0225, + "step": 9488 + }, + { + "epoch": 0.855751454209316, + "grad_norm": 1.6380697787865908, + "learning_rate": 2.1422292889148452e-07, + "loss": 0.9622, + "step": 9489 + }, + { + "epoch": 0.8558416377327862, + "grad_norm": 1.436138005417818, + "learning_rate": 2.139599480270784e-07, + "loss": 0.9965, + "step": 9490 + }, + { + "epoch": 0.8559318212562564, + "grad_norm": 21.375175609201875, + "learning_rate": 2.1369711955745773e-07, + "loss": 0.9634, + "step": 9491 + }, + { + "epoch": 0.8560220047797268, + "grad_norm": 1.7269506233989318, + "learning_rate": 2.1343444350504813e-07, + "loss": 0.9312, + "step": 9492 + }, + { + "epoch": 0.856112188303197, + "grad_norm": 1.9327405492146672, + "learning_rate": 2.1317191989226302e-07, + "loss": 0.9806, + "step": 9493 + }, + { + "epoch": 0.8562023718266673, + "grad_norm": 1.8418959654672058, + "learning_rate": 2.129095487415027e-07, + "loss": 0.9613, + "step": 9494 + }, + { + "epoch": 0.8562925553501375, + "grad_norm": 2.0851151920762785, + "learning_rate": 2.1264733007515257e-07, + "loss": 0.9901, + "step": 9495 + }, + { + "epoch": 0.8563827388736078, + "grad_norm": 2.197640918470728, + "learning_rate": 2.1238526391558852e-07, + "loss": 0.9605, + "step": 9496 + }, + { + "epoch": 0.8564729223970781, + "grad_norm": 2.0565788353951007, + "learning_rate": 2.1212335028517003e-07, + "loss": 1.1161, + "step": 9497 + }, + { + "epoch": 0.8565631059205483, + "grad_norm": 1.872062671150289, + "learning_rate": 2.1186158920624563e-07, + "loss": 0.9881, + "step": 9498 + }, + { + "epoch": 0.8566532894440185, + "grad_norm": 1.6910970286274363, + "learning_rate": 2.1159998070115015e-07, + "loss": 0.9819, + "step": 9499 + }, + { + "epoch": 0.8567434729674889, + "grad_norm": 1.835718409675177, + "learning_rate": 2.113385247922055e-07, + "loss": 0.9628, + "step": 9500 + }, + { + "epoch": 0.8568336564909591, + "grad_norm": 1.6873290363881535, + "learning_rate": 2.1107722150172068e-07, + "loss": 0.9841, + "step": 9501 + }, + { + "epoch": 0.8569238400144293, + "grad_norm": 3.3143577404605087, + "learning_rate": 2.108160708519906e-07, + "loss": 1.0159, + "step": 9502 + }, + { + "epoch": 0.8570140235378996, + "grad_norm": 1.705247309135816, + "learning_rate": 2.1055507286529984e-07, + "loss": 0.9596, + "step": 9503 + }, + { + "epoch": 0.8571042070613699, + "grad_norm": 2.3385686950707565, + "learning_rate": 2.1029422756391612e-07, + "loss": 1.0003, + "step": 9504 + }, + { + "epoch": 0.8571943905848401, + "grad_norm": 1.5868836385304745, + "learning_rate": 2.1003353497009812e-07, + "loss": 0.8918, + "step": 9505 + }, + { + "epoch": 0.8572845741083104, + "grad_norm": 1.7038060497711636, + "learning_rate": 2.0977299510608825e-07, + "loss": 0.9003, + "step": 9506 + }, + { + "epoch": 0.8573747576317807, + "grad_norm": 1.5441653476688182, + "learning_rate": 2.0951260799411784e-07, + "loss": 1.0147, + "step": 9507 + }, + { + "epoch": 0.857464941155251, + "grad_norm": 1.9710166235677082, + "learning_rate": 2.0925237365640424e-07, + "loss": 0.9699, + "step": 9508 + }, + { + "epoch": 0.8575551246787212, + "grad_norm": 0.6401802240248907, + "learning_rate": 2.0899229211515211e-07, + "loss": 0.8126, + "step": 9509 + }, + { + "epoch": 0.8576453082021914, + "grad_norm": 2.2266728959405686, + "learning_rate": 2.0873236339255306e-07, + "loss": 0.9048, + "step": 9510 + }, + { + "epoch": 0.8577354917256618, + "grad_norm": 2.0628742687614925, + "learning_rate": 2.0847258751078644e-07, + "loss": 0.9265, + "step": 9511 + }, + { + "epoch": 0.857825675249132, + "grad_norm": 2.0208375937295684, + "learning_rate": 2.082129644920163e-07, + "loss": 0.957, + "step": 9512 + }, + { + "epoch": 0.8579158587726022, + "grad_norm": 1.5502795890170666, + "learning_rate": 2.0795349435839605e-07, + "loss": 0.9906, + "step": 9513 + }, + { + "epoch": 0.8580060422960725, + "grad_norm": 2.1189486790230423, + "learning_rate": 2.0769417713206484e-07, + "loss": 0.9492, + "step": 9514 + }, + { + "epoch": 0.8580962258195428, + "grad_norm": 1.7286421045269067, + "learning_rate": 2.074350128351492e-07, + "loss": 1.0404, + "step": 9515 + }, + { + "epoch": 0.858186409343013, + "grad_norm": 1.5603771284498829, + "learning_rate": 2.0717600148976256e-07, + "loss": 0.8971, + "step": 9516 + }, + { + "epoch": 0.8582765928664833, + "grad_norm": 1.7312091423198712, + "learning_rate": 2.0691714311800436e-07, + "loss": 0.9073, + "step": 9517 + }, + { + "epoch": 0.8583667763899535, + "grad_norm": 1.6248055855287367, + "learning_rate": 2.066584377419631e-07, + "loss": 1.0381, + "step": 9518 + }, + { + "epoch": 0.8584569599134239, + "grad_norm": 1.5246451262011758, + "learning_rate": 2.0639988538371167e-07, + "loss": 1.002, + "step": 9519 + }, + { + "epoch": 0.8585471434368941, + "grad_norm": 2.006545086131439, + "learning_rate": 2.0614148606531258e-07, + "loss": 0.9458, + "step": 9520 + }, + { + "epoch": 0.8586373269603643, + "grad_norm": 5.552740557945907, + "learning_rate": 2.0588323980881285e-07, + "loss": 1.0042, + "step": 9521 + }, + { + "epoch": 0.8587275104838346, + "grad_norm": 2.006812770737633, + "learning_rate": 2.0562514663624752e-07, + "loss": 0.9648, + "step": 9522 + }, + { + "epoch": 0.8588176940073049, + "grad_norm": 1.4182378340208903, + "learning_rate": 2.0536720656963902e-07, + "loss": 1.0435, + "step": 9523 + }, + { + "epoch": 0.8589078775307751, + "grad_norm": 1.7913954403830197, + "learning_rate": 2.051094196309957e-07, + "loss": 1.0016, + "step": 9524 + }, + { + "epoch": 0.8589980610542454, + "grad_norm": 1.5163514274266174, + "learning_rate": 2.0485178584231378e-07, + "loss": 0.9393, + "step": 9525 + }, + { + "epoch": 0.8590882445777156, + "grad_norm": 0.671939735842471, + "learning_rate": 2.0459430522557587e-07, + "loss": 0.8142, + "step": 9526 + }, + { + "epoch": 0.8591784281011859, + "grad_norm": 1.8116186564218686, + "learning_rate": 2.0433697780275195e-07, + "loss": 0.984, + "step": 9527 + }, + { + "epoch": 0.8592686116246562, + "grad_norm": 1.504800664304539, + "learning_rate": 2.040798035957978e-07, + "loss": 0.9958, + "step": 9528 + }, + { + "epoch": 0.8593587951481264, + "grad_norm": 0.7175853664508347, + "learning_rate": 2.038227826266574e-07, + "loss": 0.8296, + "step": 9529 + }, + { + "epoch": 0.8594489786715968, + "grad_norm": 1.935774557847376, + "learning_rate": 2.0356591491726126e-07, + "loss": 0.9877, + "step": 9530 + }, + { + "epoch": 0.859539162195067, + "grad_norm": 1.806218740331357, + "learning_rate": 2.033092004895267e-07, + "loss": 1.0191, + "step": 9531 + }, + { + "epoch": 0.8596293457185372, + "grad_norm": 1.4817312690087732, + "learning_rate": 2.03052639365358e-07, + "loss": 0.996, + "step": 9532 + }, + { + "epoch": 0.8597195292420075, + "grad_norm": 1.5471901620723696, + "learning_rate": 2.0279623156664694e-07, + "loss": 0.9973, + "step": 9533 + }, + { + "epoch": 0.8598097127654778, + "grad_norm": 1.4859406095165364, + "learning_rate": 2.0253997711527005e-07, + "loss": 1.0361, + "step": 9534 + }, + { + "epoch": 0.859899896288948, + "grad_norm": 1.8822842565269957, + "learning_rate": 2.0228387603309428e-07, + "loss": 1.0275, + "step": 9535 + }, + { + "epoch": 0.8599900798124183, + "grad_norm": 1.630163937615164, + "learning_rate": 2.0202792834197035e-07, + "loss": 1.0011, + "step": 9536 + }, + { + "epoch": 0.8600802633358885, + "grad_norm": 1.695059761861793, + "learning_rate": 2.017721340637375e-07, + "loss": 0.9073, + "step": 9537 + }, + { + "epoch": 0.8601704468593588, + "grad_norm": 1.7568850634981144, + "learning_rate": 2.0151649322022134e-07, + "loss": 0.9657, + "step": 9538 + }, + { + "epoch": 0.8602606303828291, + "grad_norm": 2.797611246223742, + "learning_rate": 2.012610058332349e-07, + "loss": 0.7893, + "step": 9539 + }, + { + "epoch": 0.8603508139062993, + "grad_norm": 0.6682439984295774, + "learning_rate": 2.010056719245774e-07, + "loss": 0.8106, + "step": 9540 + }, + { + "epoch": 0.8604409974297695, + "grad_norm": 1.7314314736783039, + "learning_rate": 2.0075049151603563e-07, + "loss": 1.0031, + "step": 9541 + }, + { + "epoch": 0.8605311809532399, + "grad_norm": 2.1275794137910116, + "learning_rate": 2.0049546462938326e-07, + "loss": 0.984, + "step": 9542 + }, + { + "epoch": 0.8606213644767101, + "grad_norm": 2.660762320109594, + "learning_rate": 2.0024059128637961e-07, + "loss": 1.0006, + "step": 9543 + }, + { + "epoch": 0.8607115480001803, + "grad_norm": 1.5496631502398919, + "learning_rate": 1.99985871508773e-07, + "loss": 0.8453, + "step": 9544 + }, + { + "epoch": 0.8608017315236506, + "grad_norm": 1.4852321358546914, + "learning_rate": 1.9973130531829674e-07, + "loss": 1.0061, + "step": 9545 + }, + { + "epoch": 0.8608919150471209, + "grad_norm": 1.7201381972706002, + "learning_rate": 1.994768927366721e-07, + "loss": 0.926, + "step": 9546 + }, + { + "epoch": 0.8609820985705912, + "grad_norm": 1.6017384502243948, + "learning_rate": 1.992226337856069e-07, + "loss": 0.9501, + "step": 9547 + }, + { + "epoch": 0.8610722820940614, + "grad_norm": 2.2866549280641983, + "learning_rate": 1.9896852848679592e-07, + "loss": 0.9314, + "step": 9548 + }, + { + "epoch": 0.8611624656175316, + "grad_norm": 1.5634213592104806, + "learning_rate": 1.9871457686192094e-07, + "loss": 0.9889, + "step": 9549 + }, + { + "epoch": 0.861252649141002, + "grad_norm": 2.3229062774380256, + "learning_rate": 1.984607789326509e-07, + "loss": 0.9736, + "step": 9550 + }, + { + "epoch": 0.8613428326644722, + "grad_norm": 1.4442307006381112, + "learning_rate": 1.982071347206402e-07, + "loss": 0.9254, + "step": 9551 + }, + { + "epoch": 0.8614330161879424, + "grad_norm": 2.7780294738754927, + "learning_rate": 1.9795364424753202e-07, + "loss": 0.9553, + "step": 9552 + }, + { + "epoch": 0.8615231997114128, + "grad_norm": 2.201579316009345, + "learning_rate": 1.9770030753495505e-07, + "loss": 0.9464, + "step": 9553 + }, + { + "epoch": 0.861613383234883, + "grad_norm": 4.084424061323668, + "learning_rate": 1.9744712460452573e-07, + "loss": 0.9377, + "step": 9554 + }, + { + "epoch": 0.8617035667583532, + "grad_norm": 1.6534812384764612, + "learning_rate": 1.9719409547784703e-07, + "loss": 0.9607, + "step": 9555 + }, + { + "epoch": 0.8617937502818235, + "grad_norm": 2.0005500752255885, + "learning_rate": 1.9694122017650837e-07, + "loss": 1.0047, + "step": 9556 + }, + { + "epoch": 0.8618839338052938, + "grad_norm": 1.8699286223880118, + "learning_rate": 1.9668849872208738e-07, + "loss": 0.9918, + "step": 9557 + }, + { + "epoch": 0.8619741173287641, + "grad_norm": 1.5015034293822493, + "learning_rate": 1.9643593113614632e-07, + "loss": 0.8277, + "step": 9558 + }, + { + "epoch": 0.8620643008522343, + "grad_norm": 1.5492191470516463, + "learning_rate": 1.961835174402371e-07, + "loss": 1.0209, + "step": 9559 + }, + { + "epoch": 0.8621544843757045, + "grad_norm": 3.2654713863765736, + "learning_rate": 1.9593125765589535e-07, + "loss": 0.9467, + "step": 9560 + }, + { + "epoch": 0.8622446678991749, + "grad_norm": 1.8485894135118066, + "learning_rate": 1.9567915180464721e-07, + "loss": 0.9254, + "step": 9561 + }, + { + "epoch": 0.8623348514226451, + "grad_norm": 2.0143807289262217, + "learning_rate": 1.9542719990800217e-07, + "loss": 0.9646, + "step": 9562 + }, + { + "epoch": 0.8624250349461153, + "grad_norm": 1.681712717245528, + "learning_rate": 1.9517540198745896e-07, + "loss": 1.0487, + "step": 9563 + }, + { + "epoch": 0.8625152184695856, + "grad_norm": 2.0982536729634806, + "learning_rate": 1.94923758064502e-07, + "loss": 0.9748, + "step": 9564 + }, + { + "epoch": 0.8626054019930559, + "grad_norm": 1.4019163962862438, + "learning_rate": 1.9467226816060322e-07, + "loss": 0.946, + "step": 9565 + }, + { + "epoch": 0.8626955855165261, + "grad_norm": 1.8846895983290088, + "learning_rate": 1.9442093229722122e-07, + "loss": 0.9213, + "step": 9566 + }, + { + "epoch": 0.8627857690399964, + "grad_norm": 7.517191955527026, + "learning_rate": 1.9416975049580085e-07, + "loss": 1.0531, + "step": 9567 + }, + { + "epoch": 0.8628759525634666, + "grad_norm": 1.7480901787364962, + "learning_rate": 1.9391872277777456e-07, + "loss": 0.9621, + "step": 9568 + }, + { + "epoch": 0.862966136086937, + "grad_norm": 2.0731022302839412, + "learning_rate": 1.9366784916456158e-07, + "loss": 0.9671, + "step": 9569 + }, + { + "epoch": 0.8630563196104072, + "grad_norm": 1.648031247958435, + "learning_rate": 1.9341712967756774e-07, + "loss": 0.9323, + "step": 9570 + }, + { + "epoch": 0.8631465031338774, + "grad_norm": 1.7686550505801024, + "learning_rate": 1.9316656433818566e-07, + "loss": 0.8866, + "step": 9571 + }, + { + "epoch": 0.8632366866573477, + "grad_norm": 1.7475796719836323, + "learning_rate": 1.929161531677954e-07, + "loss": 0.878, + "step": 9572 + }, + { + "epoch": 0.863326870180818, + "grad_norm": 1.5562602000687396, + "learning_rate": 1.9266589618776251e-07, + "loss": 0.93, + "step": 9573 + }, + { + "epoch": 0.8634170537042882, + "grad_norm": 2.0191771678307755, + "learning_rate": 1.924157934194417e-07, + "loss": 1.0309, + "step": 9574 + }, + { + "epoch": 0.8635072372277585, + "grad_norm": 2.082442996194004, + "learning_rate": 1.9216584488417142e-07, + "loss": 0.8744, + "step": 9575 + }, + { + "epoch": 0.8635974207512288, + "grad_norm": 2.1332194739991635, + "learning_rate": 1.919160506032802e-07, + "loss": 1.0294, + "step": 9576 + }, + { + "epoch": 0.863687604274699, + "grad_norm": 1.4627793069816917, + "learning_rate": 1.916664105980812e-07, + "loss": 1.0387, + "step": 9577 + }, + { + "epoch": 0.8637777877981693, + "grad_norm": 2.0123131327621726, + "learning_rate": 1.914169248898747e-07, + "loss": 1.01, + "step": 9578 + }, + { + "epoch": 0.8638679713216395, + "grad_norm": 1.5616312282019427, + "learning_rate": 1.9116759349994882e-07, + "loss": 0.8915, + "step": 9579 + }, + { + "epoch": 0.8639581548451098, + "grad_norm": 1.5271734352471114, + "learning_rate": 1.9091841644957763e-07, + "loss": 1.0135, + "step": 9580 + }, + { + "epoch": 0.8640483383685801, + "grad_norm": 1.2903026279802603, + "learning_rate": 1.9066939376002278e-07, + "loss": 1.0621, + "step": 9581 + }, + { + "epoch": 0.8641385218920503, + "grad_norm": 2.4687831248704826, + "learning_rate": 1.9042052545253085e-07, + "loss": 0.8923, + "step": 9582 + }, + { + "epoch": 0.8642287054155205, + "grad_norm": 1.621385368866322, + "learning_rate": 1.901718115483384e-07, + "loss": 0.9567, + "step": 9583 + }, + { + "epoch": 0.8643188889389909, + "grad_norm": 1.5362688454200732, + "learning_rate": 1.8992325206866598e-07, + "loss": 0.9593, + "step": 9584 + }, + { + "epoch": 0.8644090724624611, + "grad_norm": 6.4011892286714795, + "learning_rate": 1.8967484703472225e-07, + "loss": 0.953, + "step": 9585 + }, + { + "epoch": 0.8644992559859314, + "grad_norm": 1.4378194868568441, + "learning_rate": 1.8942659646770288e-07, + "loss": 1.0239, + "step": 9586 + }, + { + "epoch": 0.8645894395094016, + "grad_norm": 1.9159258433547206, + "learning_rate": 1.8917850038878936e-07, + "loss": 1.0429, + "step": 9587 + }, + { + "epoch": 0.8646796230328719, + "grad_norm": 3.3586227639349975, + "learning_rate": 1.8893055881915121e-07, + "loss": 0.8862, + "step": 9588 + }, + { + "epoch": 0.8647698065563422, + "grad_norm": 1.9743830912473441, + "learning_rate": 1.886827717799442e-07, + "loss": 0.9227, + "step": 9589 + }, + { + "epoch": 0.8648599900798124, + "grad_norm": 1.9104157878023689, + "learning_rate": 1.884351392923096e-07, + "loss": 0.9942, + "step": 9590 + }, + { + "epoch": 0.8649501736032826, + "grad_norm": 2.771465277253175, + "learning_rate": 1.8818766137737896e-07, + "loss": 1.0169, + "step": 9591 + }, + { + "epoch": 0.865040357126753, + "grad_norm": 2.3124763899963976, + "learning_rate": 1.8794033805626653e-07, + "loss": 0.955, + "step": 9592 + }, + { + "epoch": 0.8651305406502232, + "grad_norm": 1.7259087338900327, + "learning_rate": 1.876931693500763e-07, + "loss": 0.9687, + "step": 9593 + }, + { + "epoch": 0.8652207241736934, + "grad_norm": 3.2101434992864015, + "learning_rate": 1.8744615527989783e-07, + "loss": 0.8221, + "step": 9594 + }, + { + "epoch": 0.8653109076971637, + "grad_norm": 1.8147191898659216, + "learning_rate": 1.871992958668076e-07, + "loss": 1.0639, + "step": 9595 + }, + { + "epoch": 0.865401091220634, + "grad_norm": 2.273277041091604, + "learning_rate": 1.8695259113186944e-07, + "loss": 1.0216, + "step": 9596 + }, + { + "epoch": 0.8654912747441043, + "grad_norm": 1.632928506481053, + "learning_rate": 1.8670604109613252e-07, + "loss": 0.9402, + "step": 9597 + }, + { + "epoch": 0.8655814582675745, + "grad_norm": 1.6161405703620482, + "learning_rate": 1.8645964578063533e-07, + "loss": 1.0222, + "step": 9598 + }, + { + "epoch": 0.8656716417910447, + "grad_norm": 1.8240732139023104, + "learning_rate": 1.862134052064006e-07, + "loss": 1.0369, + "step": 9599 + }, + { + "epoch": 0.8657618253145151, + "grad_norm": 2.0795297105688757, + "learning_rate": 1.8596731939443932e-07, + "loss": 1.0186, + "step": 9600 + }, + { + "epoch": 0.8658520088379853, + "grad_norm": 1.823586135729387, + "learning_rate": 1.857213883657487e-07, + "loss": 1.0058, + "step": 9601 + }, + { + "epoch": 0.8659421923614555, + "grad_norm": 0.6300745709124493, + "learning_rate": 1.8547561214131303e-07, + "loss": 0.787, + "step": 9602 + }, + { + "epoch": 0.8660323758849259, + "grad_norm": 1.8289987807105554, + "learning_rate": 1.8522999074210355e-07, + "loss": 1.0524, + "step": 9603 + }, + { + "epoch": 0.8661225594083961, + "grad_norm": 1.7695045721809364, + "learning_rate": 1.849845241890775e-07, + "loss": 1.0166, + "step": 9604 + }, + { + "epoch": 0.8662127429318663, + "grad_norm": 6.960385901177502, + "learning_rate": 1.8473921250317992e-07, + "loss": 0.9824, + "step": 9605 + }, + { + "epoch": 0.8663029264553366, + "grad_norm": 1.5914061271506381, + "learning_rate": 1.8449405570534225e-07, + "loss": 0.909, + "step": 9606 + }, + { + "epoch": 0.8663931099788069, + "grad_norm": 2.0987643648420513, + "learning_rate": 1.8424905381648204e-07, + "loss": 1.0195, + "step": 9607 + }, + { + "epoch": 0.8664832935022772, + "grad_norm": 1.4430706886588858, + "learning_rate": 1.8400420685750452e-07, + "loss": 0.9363, + "step": 9608 + }, + { + "epoch": 0.8665734770257474, + "grad_norm": 1.710685258500202, + "learning_rate": 1.8375951484930142e-07, + "loss": 0.9146, + "step": 9609 + }, + { + "epoch": 0.8666636605492176, + "grad_norm": 1.6818264848565072, + "learning_rate": 1.8351497781275094e-07, + "loss": 0.9806, + "step": 9610 + }, + { + "epoch": 0.866753844072688, + "grad_norm": 1.8308182790094174, + "learning_rate": 1.8327059576871907e-07, + "loss": 0.9688, + "step": 9611 + }, + { + "epoch": 0.8668440275961582, + "grad_norm": 1.4866309440209777, + "learning_rate": 1.8302636873805665e-07, + "loss": 1.0264, + "step": 9612 + }, + { + "epoch": 0.8669342111196284, + "grad_norm": 4.062513850261847, + "learning_rate": 1.8278229674160373e-07, + "loss": 1.0546, + "step": 9613 + }, + { + "epoch": 0.8670243946430987, + "grad_norm": 1.6103272074748247, + "learning_rate": 1.825383798001845e-07, + "loss": 0.91, + "step": 9614 + }, + { + "epoch": 0.867114578166569, + "grad_norm": 1.7750433822488274, + "learning_rate": 1.8229461793461297e-07, + "loss": 0.9211, + "step": 9615 + }, + { + "epoch": 0.8672047616900392, + "grad_norm": 1.5648116749805765, + "learning_rate": 1.8205101116568698e-07, + "loss": 0.9935, + "step": 9616 + }, + { + "epoch": 0.8672949452135095, + "grad_norm": 2.5764172070624376, + "learning_rate": 1.818075595141928e-07, + "loss": 1.024, + "step": 9617 + }, + { + "epoch": 0.8673851287369797, + "grad_norm": 1.8620356268870122, + "learning_rate": 1.8156426300090288e-07, + "loss": 0.9756, + "step": 9618 + }, + { + "epoch": 0.86747531226045, + "grad_norm": 1.5049038835420523, + "learning_rate": 1.8132112164657686e-07, + "loss": 0.9678, + "step": 9619 + }, + { + "epoch": 0.8675654957839203, + "grad_norm": 1.8712794584046348, + "learning_rate": 1.8107813547196106e-07, + "loss": 0.874, + "step": 9620 + }, + { + "epoch": 0.8676556793073905, + "grad_norm": 1.5392943851427772, + "learning_rate": 1.8083530449778817e-07, + "loss": 1.0156, + "step": 9621 + }, + { + "epoch": 0.8677458628308607, + "grad_norm": 2.9969465292945556, + "learning_rate": 1.8059262874477787e-07, + "loss": 0.8977, + "step": 9622 + }, + { + "epoch": 0.8678360463543311, + "grad_norm": 1.4849342497324394, + "learning_rate": 1.8035010823363627e-07, + "loss": 0.9622, + "step": 9623 + }, + { + "epoch": 0.8679262298778013, + "grad_norm": 15.991268159487316, + "learning_rate": 1.8010774298505705e-07, + "loss": 1.016, + "step": 9624 + }, + { + "epoch": 0.8680164134012716, + "grad_norm": 1.7845193407031905, + "learning_rate": 1.7986553301972007e-07, + "loss": 1.0735, + "step": 9625 + }, + { + "epoch": 0.8681065969247419, + "grad_norm": 1.6857931900839072, + "learning_rate": 1.7962347835829171e-07, + "loss": 0.9709, + "step": 9626 + }, + { + "epoch": 0.8681967804482121, + "grad_norm": 1.561244002026046, + "learning_rate": 1.793815790214257e-07, + "loss": 0.985, + "step": 9627 + }, + { + "epoch": 0.8682869639716824, + "grad_norm": 2.5650487180548587, + "learning_rate": 1.791398350297626e-07, + "loss": 0.9328, + "step": 9628 + }, + { + "epoch": 0.8683771474951526, + "grad_norm": 1.4316676867166482, + "learning_rate": 1.7889824640392813e-07, + "loss": 0.9846, + "step": 9629 + }, + { + "epoch": 0.868467331018623, + "grad_norm": 1.670894014139747, + "learning_rate": 1.7865681316453741e-07, + "loss": 0.9576, + "step": 9630 + }, + { + "epoch": 0.8685575145420932, + "grad_norm": 1.5860903060284945, + "learning_rate": 1.7841553533218968e-07, + "loss": 0.9787, + "step": 9631 + }, + { + "epoch": 0.8686476980655634, + "grad_norm": 2.074107366470035, + "learning_rate": 1.7817441292747292e-07, + "loss": 1.0787, + "step": 9632 + }, + { + "epoch": 0.8687378815890336, + "grad_norm": 2.0517750103396324, + "learning_rate": 1.779334459709607e-07, + "loss": 0.9723, + "step": 9633 + }, + { + "epoch": 0.868828065112504, + "grad_norm": 2.1255839050501977, + "learning_rate": 1.7769263448321347e-07, + "loss": 0.993, + "step": 9634 + }, + { + "epoch": 0.8689182486359742, + "grad_norm": 2.392148498576019, + "learning_rate": 1.7745197848477879e-07, + "loss": 0.9688, + "step": 9635 + }, + { + "epoch": 0.8690084321594445, + "grad_norm": 1.9128688282082136, + "learning_rate": 1.7721147799619063e-07, + "loss": 1.0777, + "step": 9636 + }, + { + "epoch": 0.8690986156829147, + "grad_norm": 0.5868672179253571, + "learning_rate": 1.769711330379704e-07, + "loss": 0.7843, + "step": 9637 + }, + { + "epoch": 0.869188799206385, + "grad_norm": 1.941416995596362, + "learning_rate": 1.767309436306248e-07, + "loss": 1.0261, + "step": 9638 + }, + { + "epoch": 0.8692789827298553, + "grad_norm": 1.8978747251910282, + "learning_rate": 1.764909097946483e-07, + "loss": 1.0422, + "step": 9639 + }, + { + "epoch": 0.8693691662533255, + "grad_norm": 2.0379021511004196, + "learning_rate": 1.7625103155052236e-07, + "loss": 1.0433, + "step": 9640 + }, + { + "epoch": 0.8694593497767957, + "grad_norm": 1.8034405384292613, + "learning_rate": 1.760113089187143e-07, + "loss": 0.9922, + "step": 9641 + }, + { + "epoch": 0.8695495333002661, + "grad_norm": 1.5322177709226348, + "learning_rate": 1.7577174191967868e-07, + "loss": 1.0005, + "step": 9642 + }, + { + "epoch": 0.8696397168237363, + "grad_norm": 2.4801297173876975, + "learning_rate": 1.755323305738574e-07, + "loss": 0.9911, + "step": 9643 + }, + { + "epoch": 0.8697299003472065, + "grad_norm": 1.3958782027515184, + "learning_rate": 1.7529307490167677e-07, + "loss": 1.0081, + "step": 9644 + }, + { + "epoch": 0.8698200838706768, + "grad_norm": 2.1678605336167016, + "learning_rate": 1.7505397492355288e-07, + "loss": 0.9589, + "step": 9645 + }, + { + "epoch": 0.8699102673941471, + "grad_norm": 1.8049249534116703, + "learning_rate": 1.7481503065988589e-07, + "loss": 1.0168, + "step": 9646 + }, + { + "epoch": 0.8700004509176174, + "grad_norm": 1.4223367130038833, + "learning_rate": 1.7457624213106526e-07, + "loss": 0.9414, + "step": 9647 + }, + { + "epoch": 0.8700906344410876, + "grad_norm": 2.241712566411607, + "learning_rate": 1.7433760935746465e-07, + "loss": 0.9812, + "step": 9648 + }, + { + "epoch": 0.8701808179645579, + "grad_norm": 1.7251372517701506, + "learning_rate": 1.740991323594456e-07, + "loss": 0.9235, + "step": 9649 + }, + { + "epoch": 0.8702710014880282, + "grad_norm": 1.6631434235238685, + "learning_rate": 1.7386081115735651e-07, + "loss": 0.9538, + "step": 9650 + }, + { + "epoch": 0.8703611850114984, + "grad_norm": 1.7539535550065506, + "learning_rate": 1.736226457715324e-07, + "loss": 0.9371, + "step": 9651 + }, + { + "epoch": 0.8704513685349686, + "grad_norm": 2.220255381354825, + "learning_rate": 1.7338463622229505e-07, + "loss": 0.9321, + "step": 9652 + }, + { + "epoch": 0.870541552058439, + "grad_norm": 1.662756465547707, + "learning_rate": 1.7314678252995152e-07, + "loss": 0.9286, + "step": 9653 + }, + { + "epoch": 0.8706317355819092, + "grad_norm": 2.834135802079517, + "learning_rate": 1.7290908471479847e-07, + "loss": 0.9084, + "step": 9654 + }, + { + "epoch": 0.8707219191053794, + "grad_norm": 1.9351439763997826, + "learning_rate": 1.7267154279711637e-07, + "loss": 0.9796, + "step": 9655 + }, + { + "epoch": 0.8708121026288497, + "grad_norm": 1.4282783275741695, + "learning_rate": 1.724341567971741e-07, + "loss": 1.0236, + "step": 9656 + }, + { + "epoch": 0.87090228615232, + "grad_norm": 1.8366379720496786, + "learning_rate": 1.7219692673522657e-07, + "loss": 1.0789, + "step": 9657 + }, + { + "epoch": 0.8709924696757902, + "grad_norm": 1.5150953330836674, + "learning_rate": 1.7195985263151558e-07, + "loss": 0.9977, + "step": 9658 + }, + { + "epoch": 0.8710826531992605, + "grad_norm": 1.7617528535709626, + "learning_rate": 1.7172293450626985e-07, + "loss": 1.027, + "step": 9659 + }, + { + "epoch": 0.8711728367227307, + "grad_norm": 1.5998399356409474, + "learning_rate": 1.7148617237970475e-07, + "loss": 0.9806, + "step": 9660 + }, + { + "epoch": 0.8712630202462011, + "grad_norm": 1.6334950603810925, + "learning_rate": 1.7124956627202102e-07, + "loss": 0.943, + "step": 9661 + }, + { + "epoch": 0.8713532037696713, + "grad_norm": 2.2147474049330027, + "learning_rate": 1.7101311620340852e-07, + "loss": 0.9961, + "step": 9662 + }, + { + "epoch": 0.8714433872931415, + "grad_norm": 1.7753112573792216, + "learning_rate": 1.7077682219404155e-07, + "loss": 0.9197, + "step": 9663 + }, + { + "epoch": 0.8715335708166118, + "grad_norm": 2.1421137178654486, + "learning_rate": 1.705406842640824e-07, + "loss": 1.1069, + "step": 9664 + }, + { + "epoch": 0.8716237543400821, + "grad_norm": 1.3802538114528615, + "learning_rate": 1.7030470243367946e-07, + "loss": 0.9691, + "step": 9665 + }, + { + "epoch": 0.8717139378635523, + "grad_norm": 1.3346018667578736, + "learning_rate": 1.7006887672296834e-07, + "loss": 0.9027, + "step": 9666 + }, + { + "epoch": 0.8718041213870226, + "grad_norm": 2.347606385474231, + "learning_rate": 1.6983320715207094e-07, + "loss": 0.9217, + "step": 9667 + }, + { + "epoch": 0.8718943049104928, + "grad_norm": 1.4996230923461735, + "learning_rate": 1.6959769374109523e-07, + "loss": 0.9985, + "step": 9668 + }, + { + "epoch": 0.8719844884339631, + "grad_norm": 1.5362895635813403, + "learning_rate": 1.6936233651013754e-07, + "loss": 0.9117, + "step": 9669 + }, + { + "epoch": 0.8720746719574334, + "grad_norm": 1.6210442592925485, + "learning_rate": 1.691271354792787e-07, + "loss": 1.01, + "step": 9670 + }, + { + "epoch": 0.8721648554809036, + "grad_norm": 1.9029930933447634, + "learning_rate": 1.6889209066858866e-07, + "loss": 0.8586, + "step": 9671 + }, + { + "epoch": 0.872255039004374, + "grad_norm": 1.6973638824361745, + "learning_rate": 1.6865720209812185e-07, + "loss": 1.0221, + "step": 9672 + }, + { + "epoch": 0.8723452225278442, + "grad_norm": 10.708301667835801, + "learning_rate": 1.684224697879204e-07, + "loss": 0.9477, + "step": 9673 + }, + { + "epoch": 0.8724354060513144, + "grad_norm": 2.0101794826026387, + "learning_rate": 1.6818789375801302e-07, + "loss": 1.0532, + "step": 9674 + }, + { + "epoch": 0.8725255895747847, + "grad_norm": 1.8031260683793682, + "learning_rate": 1.679534740284152e-07, + "loss": 0.9359, + "step": 9675 + }, + { + "epoch": 0.872615773098255, + "grad_norm": 1.6441583278552172, + "learning_rate": 1.6771921061912853e-07, + "loss": 0.9221, + "step": 9676 + }, + { + "epoch": 0.8727059566217252, + "grad_norm": 2.862859052103242, + "learning_rate": 1.6748510355014234e-07, + "loss": 0.9196, + "step": 9677 + }, + { + "epoch": 0.8727961401451955, + "grad_norm": 1.460385794682872, + "learning_rate": 1.6725115284143132e-07, + "loss": 0.9938, + "step": 9678 + }, + { + "epoch": 0.8728863236686657, + "grad_norm": 1.8554262176457454, + "learning_rate": 1.670173585129575e-07, + "loss": 0.9499, + "step": 9679 + }, + { + "epoch": 0.872976507192136, + "grad_norm": 1.7346348610795828, + "learning_rate": 1.667837205846696e-07, + "loss": 1.0407, + "step": 9680 + }, + { + "epoch": 0.8730666907156063, + "grad_norm": 2.6513554974995515, + "learning_rate": 1.6655023907650278e-07, + "loss": 0.9838, + "step": 9681 + }, + { + "epoch": 0.8731568742390765, + "grad_norm": 2.258115543500884, + "learning_rate": 1.6631691400837954e-07, + "loss": 1.0183, + "step": 9682 + }, + { + "epoch": 0.8732470577625467, + "grad_norm": 1.6597507116124024, + "learning_rate": 1.6608374540020752e-07, + "loss": 1.0254, + "step": 9683 + }, + { + "epoch": 0.8733372412860171, + "grad_norm": 0.6437283688216445, + "learning_rate": 1.658507332718828e-07, + "loss": 0.7635, + "step": 9684 + }, + { + "epoch": 0.8734274248094873, + "grad_norm": 1.5828089933365956, + "learning_rate": 1.656178776432864e-07, + "loss": 0.9115, + "step": 9685 + }, + { + "epoch": 0.8735176083329576, + "grad_norm": 1.9133299379396163, + "learning_rate": 1.6538517853428814e-07, + "loss": 0.9551, + "step": 9686 + }, + { + "epoch": 0.8736077918564278, + "grad_norm": 2.4981434604324404, + "learning_rate": 1.6515263596474194e-07, + "loss": 0.9591, + "step": 9687 + }, + { + "epoch": 0.8736979753798981, + "grad_norm": 2.5519008545286384, + "learning_rate": 1.6492024995449017e-07, + "loss": 0.962, + "step": 9688 + }, + { + "epoch": 0.8737881589033684, + "grad_norm": 10.278725543154847, + "learning_rate": 1.6468802052336116e-07, + "loss": 0.9442, + "step": 9689 + }, + { + "epoch": 0.8738783424268386, + "grad_norm": 1.5318725254718992, + "learning_rate": 1.6445594769116998e-07, + "loss": 0.9992, + "step": 9690 + }, + { + "epoch": 0.8739685259503088, + "grad_norm": 1.6525562771954532, + "learning_rate": 1.6422403147771836e-07, + "loss": 1.0469, + "step": 9691 + }, + { + "epoch": 0.8740587094737792, + "grad_norm": 1.5170006555443933, + "learning_rate": 1.6399227190279485e-07, + "loss": 0.9414, + "step": 9692 + }, + { + "epoch": 0.8741488929972494, + "grad_norm": 3.720460386192748, + "learning_rate": 1.637606689861748e-07, + "loss": 0.907, + "step": 9693 + }, + { + "epoch": 0.8742390765207196, + "grad_norm": 2.4904044059599255, + "learning_rate": 1.6352922274761883e-07, + "loss": 0.9387, + "step": 9694 + }, + { + "epoch": 0.87432926004419, + "grad_norm": 1.3632773150630444, + "learning_rate": 1.6329793320687602e-07, + "loss": 0.8443, + "step": 9695 + }, + { + "epoch": 0.8744194435676602, + "grad_norm": 1.811319361293867, + "learning_rate": 1.630668003836808e-07, + "loss": 1.0154, + "step": 9696 + }, + { + "epoch": 0.8745096270911304, + "grad_norm": 4.376264552876982, + "learning_rate": 1.62835824297755e-07, + "loss": 0.9807, + "step": 9697 + }, + { + "epoch": 0.8745998106146007, + "grad_norm": 1.5559403628514314, + "learning_rate": 1.626050049688066e-07, + "loss": 0.9867, + "step": 9698 + }, + { + "epoch": 0.874689994138071, + "grad_norm": 1.9059755721873517, + "learning_rate": 1.623743424165309e-07, + "loss": 0.9507, + "step": 9699 + }, + { + "epoch": 0.8747801776615413, + "grad_norm": 2.0229055522871358, + "learning_rate": 1.6214383666060826e-07, + "loss": 0.9228, + "step": 9700 + }, + { + "epoch": 0.8748703611850115, + "grad_norm": 1.863072605567242, + "learning_rate": 1.619134877207078e-07, + "loss": 0.9973, + "step": 9701 + }, + { + "epoch": 0.8749605447084817, + "grad_norm": 1.6225024617415975, + "learning_rate": 1.616832956164831e-07, + "loss": 0.9772, + "step": 9702 + }, + { + "epoch": 0.8750507282319521, + "grad_norm": 1.4121472787584541, + "learning_rate": 1.6145326036757667e-07, + "loss": 0.9458, + "step": 9703 + }, + { + "epoch": 0.8751409117554223, + "grad_norm": 2.5108458810267593, + "learning_rate": 1.612233819936155e-07, + "loss": 0.9405, + "step": 9704 + }, + { + "epoch": 0.8752310952788925, + "grad_norm": 1.6940207768171582, + "learning_rate": 1.6099366051421414e-07, + "loss": 0.9726, + "step": 9705 + }, + { + "epoch": 0.8753212788023628, + "grad_norm": 1.660750123558746, + "learning_rate": 1.6076409594897378e-07, + "loss": 1.0399, + "step": 9706 + }, + { + "epoch": 0.8754114623258331, + "grad_norm": 0.6594687725118291, + "learning_rate": 1.605346883174823e-07, + "loss": 0.7644, + "step": 9707 + }, + { + "epoch": 0.8755016458493033, + "grad_norm": 1.8239438750432704, + "learning_rate": 1.6030543763931427e-07, + "loss": 0.9529, + "step": 9708 + }, + { + "epoch": 0.8755918293727736, + "grad_norm": 2.2914905364814797, + "learning_rate": 1.600763439340298e-07, + "loss": 0.9943, + "step": 9709 + }, + { + "epoch": 0.8756820128962438, + "grad_norm": 2.556096333033072, + "learning_rate": 1.5984740722117707e-07, + "loss": 1.0327, + "step": 9710 + }, + { + "epoch": 0.8757721964197142, + "grad_norm": 1.3692007010825589, + "learning_rate": 1.5961862752028998e-07, + "loss": 0.9894, + "step": 9711 + }, + { + "epoch": 0.8758623799431844, + "grad_norm": 1.33896254318871, + "learning_rate": 1.5939000485088937e-07, + "loss": 0.8847, + "step": 9712 + }, + { + "epoch": 0.8759525634666546, + "grad_norm": 1.5133414615435223, + "learning_rate": 1.5916153923248254e-07, + "loss": 1.0745, + "step": 9713 + }, + { + "epoch": 0.8760427469901249, + "grad_norm": 1.5680103032299106, + "learning_rate": 1.5893323068456342e-07, + "loss": 1.0199, + "step": 9714 + }, + { + "epoch": 0.8761329305135952, + "grad_norm": 1.514418446180928, + "learning_rate": 1.5870507922661248e-07, + "loss": 1.0253, + "step": 9715 + }, + { + "epoch": 0.8762231140370654, + "grad_norm": 0.666972281520595, + "learning_rate": 1.5847708487809763e-07, + "loss": 0.8036, + "step": 9716 + }, + { + "epoch": 0.8763132975605357, + "grad_norm": 1.75838830629926, + "learning_rate": 1.5824924765847113e-07, + "loss": 1.0521, + "step": 9717 + }, + { + "epoch": 0.8764034810840059, + "grad_norm": 1.319205412489462, + "learning_rate": 1.5802156758717478e-07, + "loss": 0.957, + "step": 9718 + }, + { + "epoch": 0.8764936646074762, + "grad_norm": 2.5522963973394286, + "learning_rate": 1.5779404468363433e-07, + "loss": 0.9791, + "step": 9719 + }, + { + "epoch": 0.8765838481309465, + "grad_norm": 2.063091308910337, + "learning_rate": 1.5756667896726405e-07, + "loss": 1.0319, + "step": 9720 + }, + { + "epoch": 0.8766740316544167, + "grad_norm": 4.309525943184525, + "learning_rate": 1.5733947045746377e-07, + "loss": 0.9931, + "step": 9721 + }, + { + "epoch": 0.876764215177887, + "grad_norm": 2.2053301441664663, + "learning_rate": 1.5711241917362018e-07, + "loss": 1.0441, + "step": 9722 + }, + { + "epoch": 0.8768543987013573, + "grad_norm": 2.2968349323540718, + "learning_rate": 1.5688552513510688e-07, + "loss": 0.9147, + "step": 9723 + }, + { + "epoch": 0.8769445822248275, + "grad_norm": 1.732689046632767, + "learning_rate": 1.5665878836128266e-07, + "loss": 0.9999, + "step": 9724 + }, + { + "epoch": 0.8770347657482978, + "grad_norm": 1.6993834667829661, + "learning_rate": 1.5643220887149554e-07, + "loss": 1.0277, + "step": 9725 + }, + { + "epoch": 0.8771249492717681, + "grad_norm": 0.6611517673857404, + "learning_rate": 1.562057866850772e-07, + "loss": 0.8251, + "step": 9726 + }, + { + "epoch": 0.8772151327952383, + "grad_norm": 1.6631610559958263, + "learning_rate": 1.5597952182134777e-07, + "loss": 0.8545, + "step": 9727 + }, + { + "epoch": 0.8773053163187086, + "grad_norm": 1.9360466089667558, + "learning_rate": 1.557534142996133e-07, + "loss": 1.0337, + "step": 9728 + }, + { + "epoch": 0.8773954998421788, + "grad_norm": 1.8240258974654482, + "learning_rate": 1.5552746413916662e-07, + "loss": 1.0105, + "step": 9729 + }, + { + "epoch": 0.8774856833656491, + "grad_norm": 1.849533192275119, + "learning_rate": 1.5530167135928697e-07, + "loss": 0.9686, + "step": 9730 + }, + { + "epoch": 0.8775758668891194, + "grad_norm": 2.1183994104427772, + "learning_rate": 1.5507603597924068e-07, + "loss": 0.9194, + "step": 9731 + }, + { + "epoch": 0.8776660504125896, + "grad_norm": 5.89061301961232, + "learning_rate": 1.548505580182793e-07, + "loss": 0.9368, + "step": 9732 + }, + { + "epoch": 0.8777562339360598, + "grad_norm": 2.7057118468821018, + "learning_rate": 1.5462523749564271e-07, + "loss": 0.9733, + "step": 9733 + }, + { + "epoch": 0.8778464174595302, + "grad_norm": 1.5706819317992862, + "learning_rate": 1.5440007443055602e-07, + "loss": 0.9812, + "step": 9734 + }, + { + "epoch": 0.8779366009830004, + "grad_norm": 2.4988200264030676, + "learning_rate": 1.541750688422314e-07, + "loss": 0.9142, + "step": 9735 + }, + { + "epoch": 0.8780267845064706, + "grad_norm": 2.6783664661281663, + "learning_rate": 1.5395022074986797e-07, + "loss": 0.9806, + "step": 9736 + }, + { + "epoch": 0.8781169680299409, + "grad_norm": 1.641923445098339, + "learning_rate": 1.5372553017265033e-07, + "loss": 0.9224, + "step": 9737 + }, + { + "epoch": 0.8782071515534112, + "grad_norm": 3.4240973236744643, + "learning_rate": 1.5350099712975116e-07, + "loss": 1.0378, + "step": 9738 + }, + { + "epoch": 0.8782973350768815, + "grad_norm": 1.5938962981451217, + "learning_rate": 1.5327662164032785e-07, + "loss": 0.9713, + "step": 9739 + }, + { + "epoch": 0.8783875186003517, + "grad_norm": 1.9843083768458916, + "learning_rate": 1.5305240372352656e-07, + "loss": 0.8272, + "step": 9740 + }, + { + "epoch": 0.8784777021238219, + "grad_norm": 1.8789788309460953, + "learning_rate": 1.5282834339847738e-07, + "loss": 0.9899, + "step": 9741 + }, + { + "epoch": 0.8785678856472923, + "grad_norm": 1.498013372929449, + "learning_rate": 1.526044406842999e-07, + "loss": 0.959, + "step": 9742 + }, + { + "epoch": 0.8786580691707625, + "grad_norm": 1.852408043024357, + "learning_rate": 1.523806956000977e-07, + "loss": 0.9094, + "step": 9743 + }, + { + "epoch": 0.8787482526942327, + "grad_norm": 1.4589950785825163, + "learning_rate": 1.5215710816496197e-07, + "loss": 1.0272, + "step": 9744 + }, + { + "epoch": 0.8788384362177031, + "grad_norm": 1.6098609857192838, + "learning_rate": 1.5193367839797077e-07, + "loss": 0.9239, + "step": 9745 + }, + { + "epoch": 0.8789286197411733, + "grad_norm": 1.5958856223813507, + "learning_rate": 1.5171040631818842e-07, + "loss": 0.9787, + "step": 9746 + }, + { + "epoch": 0.8790188032646435, + "grad_norm": 1.6054732793665198, + "learning_rate": 1.5148729194466547e-07, + "loss": 0.9094, + "step": 9747 + }, + { + "epoch": 0.8791089867881138, + "grad_norm": 1.5906582931901434, + "learning_rate": 1.5126433529643956e-07, + "loss": 1.0468, + "step": 9748 + }, + { + "epoch": 0.8791991703115841, + "grad_norm": 2.0115342138890506, + "learning_rate": 1.5104153639253436e-07, + "loss": 0.977, + "step": 9749 + }, + { + "epoch": 0.8792893538350544, + "grad_norm": 0.6685364301965215, + "learning_rate": 1.5081889525196002e-07, + "loss": 0.8663, + "step": 9750 + }, + { + "epoch": 0.8793795373585246, + "grad_norm": 1.6570705324630859, + "learning_rate": 1.5059641189371398e-07, + "loss": 1.0324, + "step": 9751 + }, + { + "epoch": 0.8794697208819948, + "grad_norm": 1.8101891222830853, + "learning_rate": 1.503740863367795e-07, + "loss": 0.9961, + "step": 9752 + }, + { + "epoch": 0.8795599044054652, + "grad_norm": 1.8952389997679593, + "learning_rate": 1.50151918600127e-07, + "loss": 1.0041, + "step": 9753 + }, + { + "epoch": 0.8796500879289354, + "grad_norm": 1.5980662074070102, + "learning_rate": 1.4992990870271217e-07, + "loss": 1.0262, + "step": 9754 + }, + { + "epoch": 0.8797402714524056, + "grad_norm": 1.9185455760100627, + "learning_rate": 1.497080566634794e-07, + "loss": 0.9859, + "step": 9755 + }, + { + "epoch": 0.8798304549758759, + "grad_norm": 1.6140151305878014, + "learning_rate": 1.4948636250135693e-07, + "loss": 0.9886, + "step": 9756 + }, + { + "epoch": 0.8799206384993462, + "grad_norm": 9.986016223676549, + "learning_rate": 1.4926482623526249e-07, + "loss": 1.0402, + "step": 9757 + }, + { + "epoch": 0.8800108220228164, + "grad_norm": 1.9791339536941503, + "learning_rate": 1.4904344788409694e-07, + "loss": 0.9805, + "step": 9758 + }, + { + "epoch": 0.8801010055462867, + "grad_norm": 1.659298448538976, + "learning_rate": 1.4882222746675143e-07, + "loss": 0.9379, + "step": 9759 + }, + { + "epoch": 0.8801911890697569, + "grad_norm": 1.475701297918624, + "learning_rate": 1.4860116500210018e-07, + "loss": 1.0224, + "step": 9760 + }, + { + "epoch": 0.8802813725932273, + "grad_norm": 2.5345776195880836, + "learning_rate": 1.4838026050900632e-07, + "loss": 1.0518, + "step": 9761 + }, + { + "epoch": 0.8803715561166975, + "grad_norm": 1.4953778897360934, + "learning_rate": 1.481595140063181e-07, + "loss": 1.0281, + "step": 9762 + }, + { + "epoch": 0.8804617396401677, + "grad_norm": 1.9308153323784756, + "learning_rate": 1.4793892551287136e-07, + "loss": 1.0668, + "step": 9763 + }, + { + "epoch": 0.880551923163638, + "grad_norm": 2.2486091129429453, + "learning_rate": 1.4771849504748768e-07, + "loss": 0.9075, + "step": 9764 + }, + { + "epoch": 0.8806421066871083, + "grad_norm": 1.6240048662638926, + "learning_rate": 1.4749822262897517e-07, + "loss": 1.0566, + "step": 9765 + }, + { + "epoch": 0.8807322902105785, + "grad_norm": 1.647879845183312, + "learning_rate": 1.4727810827612895e-07, + "loss": 0.9212, + "step": 9766 + }, + { + "epoch": 0.8808224737340488, + "grad_norm": 1.8673654894408822, + "learning_rate": 1.470581520077303e-07, + "loss": 0.986, + "step": 9767 + }, + { + "epoch": 0.8809126572575191, + "grad_norm": 2.15249273265715, + "learning_rate": 1.4683835384254705e-07, + "loss": 0.9771, + "step": 9768 + }, + { + "epoch": 0.8810028407809893, + "grad_norm": 2.2471626193149628, + "learning_rate": 1.4661871379933376e-07, + "loss": 1.0623, + "step": 9769 + }, + { + "epoch": 0.8810930243044596, + "grad_norm": 1.724604605541784, + "learning_rate": 1.4639923189683169e-07, + "loss": 1.0296, + "step": 9770 + }, + { + "epoch": 0.8811832078279298, + "grad_norm": 14.933325020469894, + "learning_rate": 1.461799081537669e-07, + "loss": 1.0276, + "step": 9771 + }, + { + "epoch": 0.8812733913514001, + "grad_norm": 1.611027508060966, + "learning_rate": 1.4596074258885514e-07, + "loss": 1.0437, + "step": 9772 + }, + { + "epoch": 0.8813635748748704, + "grad_norm": 1.7610980037732265, + "learning_rate": 1.4574173522079502e-07, + "loss": 0.9516, + "step": 9773 + }, + { + "epoch": 0.8814537583983406, + "grad_norm": 2.5207905305953933, + "learning_rate": 1.4552288606827513e-07, + "loss": 0.9546, + "step": 9774 + }, + { + "epoch": 0.8815439419218108, + "grad_norm": 1.8983482057269598, + "learning_rate": 1.4530419514996761e-07, + "loss": 1.0557, + "step": 9775 + }, + { + "epoch": 0.8816341254452812, + "grad_norm": 0.6963210132004158, + "learning_rate": 1.4508566248453291e-07, + "loss": 0.8256, + "step": 9776 + }, + { + "epoch": 0.8817243089687514, + "grad_norm": 2.683829795528846, + "learning_rate": 1.448672880906172e-07, + "loss": 0.9773, + "step": 9777 + }, + { + "epoch": 0.8818144924922217, + "grad_norm": 1.6867305272845405, + "learning_rate": 1.4464907198685382e-07, + "loss": 0.9356, + "step": 9778 + }, + { + "epoch": 0.8819046760156919, + "grad_norm": 1.5267951155574662, + "learning_rate": 1.444310141918621e-07, + "loss": 0.9848, + "step": 9779 + }, + { + "epoch": 0.8819948595391622, + "grad_norm": 1.934803224163985, + "learning_rate": 1.4421311472424735e-07, + "loss": 1.0066, + "step": 9780 + }, + { + "epoch": 0.8820850430626325, + "grad_norm": 1.8771513993951636, + "learning_rate": 1.4399537360260273e-07, + "loss": 1.0161, + "step": 9781 + }, + { + "epoch": 0.8821752265861027, + "grad_norm": 2.0124761544467544, + "learning_rate": 1.4377779084550645e-07, + "loss": 1.0036, + "step": 9782 + }, + { + "epoch": 0.8822654101095729, + "grad_norm": 1.718757490661944, + "learning_rate": 1.4356036647152413e-07, + "loss": 0.9743, + "step": 9783 + }, + { + "epoch": 0.8823555936330433, + "grad_norm": 1.4461423398947726, + "learning_rate": 1.4334310049920785e-07, + "loss": 0.9596, + "step": 9784 + }, + { + "epoch": 0.8824457771565135, + "grad_norm": 0.5953841809548601, + "learning_rate": 1.431259929470956e-07, + "loss": 0.7658, + "step": 9785 + }, + { + "epoch": 0.8825359606799837, + "grad_norm": 1.6407429244031138, + "learning_rate": 1.4290904383371237e-07, + "loss": 1.0601, + "step": 9786 + }, + { + "epoch": 0.882626144203454, + "grad_norm": 1.9640525605186172, + "learning_rate": 1.4269225317756961e-07, + "loss": 0.963, + "step": 9787 + }, + { + "epoch": 0.8827163277269243, + "grad_norm": 1.9332308958582909, + "learning_rate": 1.424756209971645e-07, + "loss": 0.9414, + "step": 9788 + }, + { + "epoch": 0.8828065112503946, + "grad_norm": 1.8790818606213024, + "learning_rate": 1.4225914731098199e-07, + "loss": 0.9813, + "step": 9789 + }, + { + "epoch": 0.8828966947738648, + "grad_norm": 1.8527369255291255, + "learning_rate": 1.4204283213749248e-07, + "loss": 0.9142, + "step": 9790 + }, + { + "epoch": 0.8829868782973351, + "grad_norm": 1.5608116183151097, + "learning_rate": 1.4182667549515315e-07, + "loss": 1.0404, + "step": 9791 + }, + { + "epoch": 0.8830770618208054, + "grad_norm": 1.9156515570786399, + "learning_rate": 1.4161067740240752e-07, + "loss": 0.9609, + "step": 9792 + }, + { + "epoch": 0.8831672453442756, + "grad_norm": 1.5492004485787851, + "learning_rate": 1.4139483787768614e-07, + "loss": 0.9742, + "step": 9793 + }, + { + "epoch": 0.8832574288677458, + "grad_norm": 1.5372442358395586, + "learning_rate": 1.4117915693940584e-07, + "loss": 0.8896, + "step": 9794 + }, + { + "epoch": 0.8833476123912162, + "grad_norm": 1.4765944704500198, + "learning_rate": 1.409636346059684e-07, + "loss": 0.9732, + "step": 9795 + }, + { + "epoch": 0.8834377959146864, + "grad_norm": 2.0756105972337098, + "learning_rate": 1.4074827089576501e-07, + "loss": 1.015, + "step": 9796 + }, + { + "epoch": 0.8835279794381566, + "grad_norm": 1.8166385625129402, + "learning_rate": 1.4053306582717085e-07, + "loss": 0.9961, + "step": 9797 + }, + { + "epoch": 0.8836181629616269, + "grad_norm": 1.5959105712817778, + "learning_rate": 1.4031801941854827e-07, + "loss": 1.0044, + "step": 9798 + }, + { + "epoch": 0.8837083464850972, + "grad_norm": 1.4501460725278992, + "learning_rate": 1.401031316882466e-07, + "loss": 0.9472, + "step": 9799 + }, + { + "epoch": 0.8837985300085675, + "grad_norm": 1.7652867085954342, + "learning_rate": 1.39888402654601e-07, + "loss": 1.0001, + "step": 9800 + }, + { + "epoch": 0.8838887135320377, + "grad_norm": 0.6112559949350767, + "learning_rate": 1.3967383233593344e-07, + "loss": 0.824, + "step": 9801 + }, + { + "epoch": 0.8839788970555079, + "grad_norm": 1.7395206447734788, + "learning_rate": 1.3945942075055218e-07, + "loss": 0.9382, + "step": 9802 + }, + { + "epoch": 0.8840690805789783, + "grad_norm": 1.9309339320231174, + "learning_rate": 1.3924516791675212e-07, + "loss": 1.0054, + "step": 9803 + }, + { + "epoch": 0.8841592641024485, + "grad_norm": 1.540409331479234, + "learning_rate": 1.3903107385281487e-07, + "loss": 0.8907, + "step": 9804 + }, + { + "epoch": 0.8842494476259187, + "grad_norm": 1.7135262609405737, + "learning_rate": 1.3881713857700717e-07, + "loss": 0.9907, + "step": 9805 + }, + { + "epoch": 0.884339631149389, + "grad_norm": 1.676958430912946, + "learning_rate": 1.3860336210758372e-07, + "loss": 0.9579, + "step": 9806 + }, + { + "epoch": 0.8844298146728593, + "grad_norm": 1.609621195804135, + "learning_rate": 1.3838974446278506e-07, + "loss": 0.9377, + "step": 9807 + }, + { + "epoch": 0.8845199981963295, + "grad_norm": 1.9361181561259608, + "learning_rate": 1.3817628566083817e-07, + "loss": 0.9594, + "step": 9808 + }, + { + "epoch": 0.8846101817197998, + "grad_norm": 1.4070847047334645, + "learning_rate": 1.3796298571995712e-07, + "loss": 0.9916, + "step": 9809 + }, + { + "epoch": 0.88470036524327, + "grad_norm": 1.9446284918496584, + "learning_rate": 1.377498446583405e-07, + "loss": 0.8785, + "step": 9810 + }, + { + "epoch": 0.8847905487667403, + "grad_norm": 1.4497340024120404, + "learning_rate": 1.3753686249417596e-07, + "loss": 0.9536, + "step": 9811 + }, + { + "epoch": 0.8848807322902106, + "grad_norm": 2.147666210097346, + "learning_rate": 1.373240392456354e-07, + "loss": 0.9309, + "step": 9812 + }, + { + "epoch": 0.8849709158136808, + "grad_norm": 1.691098720441371, + "learning_rate": 1.37111374930879e-07, + "loss": 0.9153, + "step": 9813 + }, + { + "epoch": 0.885061099337151, + "grad_norm": 1.4491902885829706, + "learning_rate": 1.3689886956805176e-07, + "loss": 0.9091, + "step": 9814 + }, + { + "epoch": 0.8851512828606214, + "grad_norm": 1.408749711071347, + "learning_rate": 1.3668652317528585e-07, + "loss": 0.9696, + "step": 9815 + }, + { + "epoch": 0.8852414663840916, + "grad_norm": 1.866446503643851, + "learning_rate": 1.3647433577070012e-07, + "loss": 0.9995, + "step": 9816 + }, + { + "epoch": 0.8853316499075619, + "grad_norm": 0.6349353739374964, + "learning_rate": 1.3626230737239942e-07, + "loss": 0.7863, + "step": 9817 + }, + { + "epoch": 0.8854218334310322, + "grad_norm": 0.7191597972761335, + "learning_rate": 1.3605043799847527e-07, + "loss": 0.7936, + "step": 9818 + }, + { + "epoch": 0.8855120169545024, + "grad_norm": 1.7469556076750714, + "learning_rate": 1.3583872766700567e-07, + "loss": 0.917, + "step": 9819 + }, + { + "epoch": 0.8856022004779727, + "grad_norm": 2.442385448163062, + "learning_rate": 1.3562717639605437e-07, + "loss": 0.9599, + "step": 9820 + }, + { + "epoch": 0.8856923840014429, + "grad_norm": 1.8334704766852241, + "learning_rate": 1.3541578420367229e-07, + "loss": 0.9496, + "step": 9821 + }, + { + "epoch": 0.8857825675249132, + "grad_norm": 1.9731898426497407, + "learning_rate": 1.3520455110789697e-07, + "loss": 0.9929, + "step": 9822 + }, + { + "epoch": 0.8858727510483835, + "grad_norm": 1.4549971794808962, + "learning_rate": 1.3499347712675158e-07, + "loss": 0.9544, + "step": 9823 + }, + { + "epoch": 0.8859629345718537, + "grad_norm": 2.1611616061189065, + "learning_rate": 1.3478256227824635e-07, + "loss": 0.9426, + "step": 9824 + }, + { + "epoch": 0.8860531180953239, + "grad_norm": 2.1346588526878008, + "learning_rate": 1.3457180658037759e-07, + "loss": 0.8852, + "step": 9825 + }, + { + "epoch": 0.8861433016187943, + "grad_norm": 1.6785999414520998, + "learning_rate": 1.3436121005112843e-07, + "loss": 1.0107, + "step": 9826 + }, + { + "epoch": 0.8862334851422645, + "grad_norm": 1.5246290193426413, + "learning_rate": 1.3415077270846719e-07, + "loss": 0.9473, + "step": 9827 + }, + { + "epoch": 0.8863236686657348, + "grad_norm": 2.955842559028584, + "learning_rate": 1.3394049457035105e-07, + "loss": 1.0872, + "step": 9828 + }, + { + "epoch": 0.886413852189205, + "grad_norm": 1.7048537553506398, + "learning_rate": 1.3373037565472034e-07, + "loss": 0.9856, + "step": 9829 + }, + { + "epoch": 0.8865040357126753, + "grad_norm": 1.6183946315157782, + "learning_rate": 1.3352041597950537e-07, + "loss": 0.9559, + "step": 9830 + }, + { + "epoch": 0.8865942192361456, + "grad_norm": 1.6325952850855054, + "learning_rate": 1.333106155626196e-07, + "loss": 0.9682, + "step": 9831 + }, + { + "epoch": 0.8866844027596158, + "grad_norm": 3.93330244332377, + "learning_rate": 1.331009744219651e-07, + "loss": 0.9175, + "step": 9832 + }, + { + "epoch": 0.886774586283086, + "grad_norm": 0.7020295192307465, + "learning_rate": 1.3289149257542943e-07, + "loss": 0.848, + "step": 9833 + }, + { + "epoch": 0.8868647698065564, + "grad_norm": 1.5329088839870573, + "learning_rate": 1.3268217004088666e-07, + "loss": 0.8826, + "step": 9834 + }, + { + "epoch": 0.8869549533300266, + "grad_norm": 3.3995738267272766, + "learning_rate": 1.3247300683619788e-07, + "loss": 0.8678, + "step": 9835 + }, + { + "epoch": 0.8870451368534968, + "grad_norm": 2.241966954459886, + "learning_rate": 1.3226400297920903e-07, + "loss": 0.9607, + "step": 9836 + }, + { + "epoch": 0.8871353203769671, + "grad_norm": 2.071137461853939, + "learning_rate": 1.3205515848775428e-07, + "loss": 0.9635, + "step": 9837 + }, + { + "epoch": 0.8872255039004374, + "grad_norm": 1.6472331343502555, + "learning_rate": 1.3184647337965316e-07, + "loss": 0.9174, + "step": 9838 + }, + { + "epoch": 0.8873156874239077, + "grad_norm": 1.3699496894843035, + "learning_rate": 1.3163794767271163e-07, + "loss": 0.9252, + "step": 9839 + }, + { + "epoch": 0.8874058709473779, + "grad_norm": 2.4963402663688266, + "learning_rate": 1.314295813847226e-07, + "loss": 0.9481, + "step": 9840 + }, + { + "epoch": 0.8874960544708482, + "grad_norm": 1.8507033660590928, + "learning_rate": 1.3122137453346515e-07, + "loss": 1.0765, + "step": 9841 + }, + { + "epoch": 0.8875862379943185, + "grad_norm": 1.740798807284178, + "learning_rate": 1.3101332713670376e-07, + "loss": 1.0338, + "step": 9842 + }, + { + "epoch": 0.8876764215177887, + "grad_norm": 0.6726894986290336, + "learning_rate": 1.3080543921219133e-07, + "loss": 0.8428, + "step": 9843 + }, + { + "epoch": 0.8877666050412589, + "grad_norm": 1.6567756970152407, + "learning_rate": 1.3059771077766478e-07, + "loss": 0.9673, + "step": 9844 + }, + { + "epoch": 0.8878567885647293, + "grad_norm": 2.3508171852131454, + "learning_rate": 1.3039014185085018e-07, + "loss": 1.0172, + "step": 9845 + }, + { + "epoch": 0.8879469720881995, + "grad_norm": 1.7241728112375698, + "learning_rate": 1.301827324494571e-07, + "loss": 0.985, + "step": 9846 + }, + { + "epoch": 0.8880371556116697, + "grad_norm": 2.054098414692249, + "learning_rate": 1.2997548259118342e-07, + "loss": 1.0187, + "step": 9847 + }, + { + "epoch": 0.88812733913514, + "grad_norm": 1.6683595007572827, + "learning_rate": 1.2976839229371272e-07, + "loss": 1.0065, + "step": 9848 + }, + { + "epoch": 0.8882175226586103, + "grad_norm": 0.6226353018382672, + "learning_rate": 1.2956146157471515e-07, + "loss": 0.8321, + "step": 9849 + }, + { + "epoch": 0.8883077061820805, + "grad_norm": 2.230472811880605, + "learning_rate": 1.2935469045184745e-07, + "loss": 1.0018, + "step": 9850 + }, + { + "epoch": 0.8883978897055508, + "grad_norm": 2.1027481535024837, + "learning_rate": 1.291480789427517e-07, + "loss": 0.9529, + "step": 9851 + }, + { + "epoch": 0.888488073229021, + "grad_norm": 1.7581058511732004, + "learning_rate": 1.2894162706505807e-07, + "loss": 1.0159, + "step": 9852 + }, + { + "epoch": 0.8885782567524914, + "grad_norm": 3.2463793026478216, + "learning_rate": 1.2873533483638155e-07, + "loss": 0.9734, + "step": 9853 + }, + { + "epoch": 0.8886684402759616, + "grad_norm": 2.3492572422254923, + "learning_rate": 1.285292022743243e-07, + "loss": 0.9272, + "step": 9854 + }, + { + "epoch": 0.8887586237994318, + "grad_norm": 1.4988430647792697, + "learning_rate": 1.2832322939647467e-07, + "loss": 0.9253, + "step": 9855 + }, + { + "epoch": 0.8888488073229021, + "grad_norm": 1.6816198544875456, + "learning_rate": 1.281174162204075e-07, + "loss": 0.9633, + "step": 9856 + }, + { + "epoch": 0.8889389908463724, + "grad_norm": 2.6600179538264315, + "learning_rate": 1.2791176276368366e-07, + "loss": 0.8584, + "step": 9857 + }, + { + "epoch": 0.8890291743698426, + "grad_norm": 1.7938259627044806, + "learning_rate": 1.2770626904385128e-07, + "loss": 0.9909, + "step": 9858 + }, + { + "epoch": 0.8891193578933129, + "grad_norm": 1.5754537352679083, + "learning_rate": 1.2750093507844306e-07, + "loss": 0.8434, + "step": 9859 + }, + { + "epoch": 0.8892095414167831, + "grad_norm": 1.3020511877542398, + "learning_rate": 1.272957608849805e-07, + "loss": 1.0184, + "step": 9860 + }, + { + "epoch": 0.8892997249402534, + "grad_norm": 1.4337780077858466, + "learning_rate": 1.270907464809694e-07, + "loss": 0.9531, + "step": 9861 + }, + { + "epoch": 0.8893899084637237, + "grad_norm": 1.5436872523601939, + "learning_rate": 1.2688589188390285e-07, + "loss": 0.9795, + "step": 9862 + }, + { + "epoch": 0.8894800919871939, + "grad_norm": 1.4811813531734626, + "learning_rate": 1.2668119711126023e-07, + "loss": 0.99, + "step": 9863 + }, + { + "epoch": 0.8895702755106643, + "grad_norm": 1.4086885290265243, + "learning_rate": 1.2647666218050735e-07, + "loss": 0.9784, + "step": 9864 + }, + { + "epoch": 0.8896604590341345, + "grad_norm": 6.3408455048446495, + "learning_rate": 1.2627228710909643e-07, + "loss": 0.9412, + "step": 9865 + }, + { + "epoch": 0.8897506425576047, + "grad_norm": 1.6550119199064244, + "learning_rate": 1.260680719144649e-07, + "loss": 0.9425, + "step": 9866 + }, + { + "epoch": 0.889840826081075, + "grad_norm": 7.4055610690422675, + "learning_rate": 1.2586401661403877e-07, + "loss": 0.9064, + "step": 9867 + }, + { + "epoch": 0.8899310096045453, + "grad_norm": 1.7370309512203392, + "learning_rate": 1.2566012122522817e-07, + "loss": 0.9487, + "step": 9868 + }, + { + "epoch": 0.8900211931280155, + "grad_norm": 2.543326122762009, + "learning_rate": 1.254563857654316e-07, + "loss": 0.949, + "step": 9869 + }, + { + "epoch": 0.8901113766514858, + "grad_norm": 2.6608905422975178, + "learning_rate": 1.2525281025203205e-07, + "loss": 0.8651, + "step": 9870 + }, + { + "epoch": 0.890201560174956, + "grad_norm": 1.864440940946597, + "learning_rate": 1.2504939470240006e-07, + "loss": 0.9599, + "step": 9871 + }, + { + "epoch": 0.8902917436984263, + "grad_norm": 1.8359794449580173, + "learning_rate": 1.2484613913389196e-07, + "loss": 1.0453, + "step": 9872 + }, + { + "epoch": 0.8903819272218966, + "grad_norm": 1.6734855238918374, + "learning_rate": 1.2464304356385057e-07, + "loss": 0.8935, + "step": 9873 + }, + { + "epoch": 0.8904721107453668, + "grad_norm": 1.8708379327707598, + "learning_rate": 1.2444010800960558e-07, + "loss": 1.0368, + "step": 9874 + }, + { + "epoch": 0.890562294268837, + "grad_norm": 1.4674907724327582, + "learning_rate": 1.2423733248847267e-07, + "loss": 0.962, + "step": 9875 + }, + { + "epoch": 0.8906524777923074, + "grad_norm": 1.559111086236861, + "learning_rate": 1.2403471701775293e-07, + "loss": 0.918, + "step": 9876 + }, + { + "epoch": 0.8907426613157776, + "grad_norm": 8.643327783518327, + "learning_rate": 1.2383226161473515e-07, + "loss": 1.0157, + "step": 9877 + }, + { + "epoch": 0.8908328448392479, + "grad_norm": 1.6808105972308205, + "learning_rate": 1.2362996629669376e-07, + "loss": 1.0519, + "step": 9878 + }, + { + "epoch": 0.8909230283627181, + "grad_norm": 3.5284419414927126, + "learning_rate": 1.2342783108089007e-07, + "loss": 0.9909, + "step": 9879 + }, + { + "epoch": 0.8910132118861884, + "grad_norm": 1.8415403541412787, + "learning_rate": 1.2322585598457135e-07, + "loss": 1.0349, + "step": 9880 + }, + { + "epoch": 0.8911033954096587, + "grad_norm": 1.3336717901206927, + "learning_rate": 1.2302404102497034e-07, + "loss": 0.9516, + "step": 9881 + }, + { + "epoch": 0.8911935789331289, + "grad_norm": 4.434547409136242, + "learning_rate": 1.228223862193083e-07, + "loss": 0.907, + "step": 9882 + }, + { + "epoch": 0.8912837624565991, + "grad_norm": 2.4989437733064777, + "learning_rate": 1.2262089158479038e-07, + "loss": 0.9894, + "step": 9883 + }, + { + "epoch": 0.8913739459800695, + "grad_norm": 1.6031103330539849, + "learning_rate": 1.2241955713861042e-07, + "loss": 1.024, + "step": 9884 + }, + { + "epoch": 0.8914641295035397, + "grad_norm": 2.0137043163154225, + "learning_rate": 1.222183828979464e-07, + "loss": 0.9225, + "step": 9885 + }, + { + "epoch": 0.8915543130270099, + "grad_norm": 1.466333064517544, + "learning_rate": 1.2201736887996372e-07, + "loss": 0.9743, + "step": 9886 + }, + { + "epoch": 0.8916444965504803, + "grad_norm": 1.6841417735036108, + "learning_rate": 1.2181651510181444e-07, + "loss": 1.0578, + "step": 9887 + }, + { + "epoch": 0.8917346800739505, + "grad_norm": 1.5230400936487432, + "learning_rate": 1.2161582158063622e-07, + "loss": 1.0059, + "step": 9888 + }, + { + "epoch": 0.8918248635974207, + "grad_norm": 2.056508812660795, + "learning_rate": 1.214152883335533e-07, + "loss": 0.8854, + "step": 9889 + }, + { + "epoch": 0.891915047120891, + "grad_norm": 2.8589940885714205, + "learning_rate": 1.2121491537767648e-07, + "loss": 0.9301, + "step": 9890 + }, + { + "epoch": 0.8920052306443613, + "grad_norm": 1.4947924818863014, + "learning_rate": 1.2101470273010294e-07, + "loss": 1.0121, + "step": 9891 + }, + { + "epoch": 0.8920954141678316, + "grad_norm": 2.087292863082911, + "learning_rate": 1.2081465040791528e-07, + "loss": 0.9533, + "step": 9892 + }, + { + "epoch": 0.8921855976913018, + "grad_norm": 1.897489397323398, + "learning_rate": 1.2061475842818335e-07, + "loss": 0.9432, + "step": 9893 + }, + { + "epoch": 0.892275781214772, + "grad_norm": 1.4421241027189104, + "learning_rate": 1.2041502680796313e-07, + "loss": 0.9541, + "step": 9894 + }, + { + "epoch": 0.8923659647382424, + "grad_norm": 1.4198488324253737, + "learning_rate": 1.2021545556429648e-07, + "loss": 0.9844, + "step": 9895 + }, + { + "epoch": 0.8924561482617126, + "grad_norm": 1.2933284109470908, + "learning_rate": 1.2001604471421245e-07, + "loss": 0.9752, + "step": 9896 + }, + { + "epoch": 0.8925463317851828, + "grad_norm": 1.492455822851535, + "learning_rate": 1.1981679427472567e-07, + "loss": 1.0272, + "step": 9897 + }, + { + "epoch": 0.8926365153086531, + "grad_norm": 1.7991055120428971, + "learning_rate": 1.196177042628368e-07, + "loss": 1.0295, + "step": 9898 + }, + { + "epoch": 0.8927266988321234, + "grad_norm": 1.9840456471498746, + "learning_rate": 1.194187746955344e-07, + "loss": 0.9852, + "step": 9899 + }, + { + "epoch": 0.8928168823555936, + "grad_norm": 1.6186477776142232, + "learning_rate": 1.1922000558979094e-07, + "loss": 0.9552, + "step": 9900 + }, + { + "epoch": 0.8929070658790639, + "grad_norm": 1.6414597340530337, + "learning_rate": 1.1902139696256752e-07, + "loss": 1.0098, + "step": 9901 + }, + { + "epoch": 0.8929972494025341, + "grad_norm": 1.7401615560004926, + "learning_rate": 1.188229488308099e-07, + "loss": 0.9155, + "step": 9902 + }, + { + "epoch": 0.8930874329260045, + "grad_norm": 0.5589264099222067, + "learning_rate": 1.1862466121145098e-07, + "loss": 0.7938, + "step": 9903 + }, + { + "epoch": 0.8931776164494747, + "grad_norm": 7.92419279301674, + "learning_rate": 1.184265341214099e-07, + "loss": 0.9181, + "step": 9904 + }, + { + "epoch": 0.8932677999729449, + "grad_norm": 3.396411230298871, + "learning_rate": 1.182285675775918e-07, + "loss": 1.001, + "step": 9905 + }, + { + "epoch": 0.8933579834964152, + "grad_norm": 2.2926098934279255, + "learning_rate": 1.1803076159688851e-07, + "loss": 0.9604, + "step": 9906 + }, + { + "epoch": 0.8934481670198855, + "grad_norm": 1.7145555119287061, + "learning_rate": 1.1783311619617741e-07, + "loss": 1.0135, + "step": 9907 + }, + { + "epoch": 0.8935383505433557, + "grad_norm": 1.687024226052587, + "learning_rate": 1.1763563139232302e-07, + "loss": 0.9447, + "step": 9908 + }, + { + "epoch": 0.893628534066826, + "grad_norm": 2.308318094233076, + "learning_rate": 1.1743830720217562e-07, + "loss": 0.9185, + "step": 9909 + }, + { + "epoch": 0.8937187175902963, + "grad_norm": 3.0428867941165807, + "learning_rate": 1.1724114364257243e-07, + "loss": 0.9526, + "step": 9910 + }, + { + "epoch": 0.8938089011137665, + "grad_norm": 1.5220291278974305, + "learning_rate": 1.1704414073033619e-07, + "loss": 0.9858, + "step": 9911 + }, + { + "epoch": 0.8938990846372368, + "grad_norm": 1.6351685386374568, + "learning_rate": 1.1684729848227636e-07, + "loss": 1.0198, + "step": 9912 + }, + { + "epoch": 0.893989268160707, + "grad_norm": 2.1394276855471865, + "learning_rate": 1.1665061691518884e-07, + "loss": 1.0327, + "step": 9913 + }, + { + "epoch": 0.8940794516841774, + "grad_norm": 1.654340398907233, + "learning_rate": 1.1645409604585532e-07, + "loss": 0.9749, + "step": 9914 + }, + { + "epoch": 0.8941696352076476, + "grad_norm": 1.6751075937815543, + "learning_rate": 1.162577358910437e-07, + "loss": 0.9852, + "step": 9915 + }, + { + "epoch": 0.8942598187311178, + "grad_norm": 1.6401026302882444, + "learning_rate": 1.160615364675095e-07, + "loss": 0.9567, + "step": 9916 + }, + { + "epoch": 0.894350002254588, + "grad_norm": 0.610441985489461, + "learning_rate": 1.1586549779199262e-07, + "loss": 0.7956, + "step": 9917 + }, + { + "epoch": 0.8944401857780584, + "grad_norm": 1.388488414036097, + "learning_rate": 1.1566961988122037e-07, + "loss": 1.0353, + "step": 9918 + }, + { + "epoch": 0.8945303693015286, + "grad_norm": 1.6713325832737043, + "learning_rate": 1.1547390275190627e-07, + "loss": 0.9418, + "step": 9919 + }, + { + "epoch": 0.8946205528249989, + "grad_norm": 3.4232966287410167, + "learning_rate": 1.1527834642075007e-07, + "loss": 0.9148, + "step": 9920 + }, + { + "epoch": 0.8947107363484691, + "grad_norm": 4.991411653265877, + "learning_rate": 1.1508295090443797e-07, + "loss": 0.9615, + "step": 9921 + }, + { + "epoch": 0.8948009198719394, + "grad_norm": 1.9943091968570112, + "learning_rate": 1.148877162196411e-07, + "loss": 0.9273, + "step": 9922 + }, + { + "epoch": 0.8948911033954097, + "grad_norm": 2.3183700702680845, + "learning_rate": 1.1469264238301924e-07, + "loss": 0.9115, + "step": 9923 + }, + { + "epoch": 0.8949812869188799, + "grad_norm": 3.1889498536074914, + "learning_rate": 1.1449772941121638e-07, + "loss": 1.0037, + "step": 9924 + }, + { + "epoch": 0.8950714704423501, + "grad_norm": 1.8615513094363487, + "learning_rate": 1.1430297732086369e-07, + "loss": 0.9452, + "step": 9925 + }, + { + "epoch": 0.8951616539658205, + "grad_norm": 1.489683356850498, + "learning_rate": 1.1410838612857876e-07, + "loss": 0.9673, + "step": 9926 + }, + { + "epoch": 0.8952518374892907, + "grad_norm": 2.309488758800607, + "learning_rate": 1.1391395585096497e-07, + "loss": 1.0764, + "step": 9927 + }, + { + "epoch": 0.895342021012761, + "grad_norm": 1.5205375261778364, + "learning_rate": 1.1371968650461216e-07, + "loss": 0.9071, + "step": 9928 + }, + { + "epoch": 0.8954322045362312, + "grad_norm": 0.6770294216896372, + "learning_rate": 1.1352557810609687e-07, + "loss": 0.8098, + "step": 9929 + }, + { + "epoch": 0.8955223880597015, + "grad_norm": 1.6253792246900276, + "learning_rate": 1.1333163067198048e-07, + "loss": 0.9636, + "step": 9930 + }, + { + "epoch": 0.8956125715831718, + "grad_norm": 1.8829299446927077, + "learning_rate": 1.1313784421881311e-07, + "loss": 1.0407, + "step": 9931 + }, + { + "epoch": 0.895702755106642, + "grad_norm": 1.5103161043913012, + "learning_rate": 1.1294421876312865e-07, + "loss": 1.0129, + "step": 9932 + }, + { + "epoch": 0.8957929386301122, + "grad_norm": 1.7454596611963067, + "learning_rate": 1.1275075432144831e-07, + "loss": 0.953, + "step": 9933 + }, + { + "epoch": 0.8958831221535826, + "grad_norm": 1.9486061616657842, + "learning_rate": 1.1255745091028002e-07, + "loss": 0.9915, + "step": 9934 + }, + { + "epoch": 0.8959733056770528, + "grad_norm": 1.6850303663013513, + "learning_rate": 1.1236430854611723e-07, + "loss": 0.9864, + "step": 9935 + }, + { + "epoch": 0.896063489200523, + "grad_norm": 1.8659462733614307, + "learning_rate": 1.1217132724544032e-07, + "loss": 0.9429, + "step": 9936 + }, + { + "epoch": 0.8961536727239934, + "grad_norm": 3.013562223327195, + "learning_rate": 1.1197850702471434e-07, + "loss": 1.0307, + "step": 9937 + }, + { + "epoch": 0.8962438562474636, + "grad_norm": 3.9395182447673083, + "learning_rate": 1.1178584790039348e-07, + "loss": 1.0441, + "step": 9938 + }, + { + "epoch": 0.8963340397709338, + "grad_norm": 1.58624099308039, + "learning_rate": 1.1159334988891478e-07, + "loss": 0.9232, + "step": 9939 + }, + { + "epoch": 0.8964242232944041, + "grad_norm": 1.7491730370965144, + "learning_rate": 1.1140101300670446e-07, + "loss": 0.9542, + "step": 9940 + }, + { + "epoch": 0.8965144068178744, + "grad_norm": 1.5322792331062305, + "learning_rate": 1.1120883727017338e-07, + "loss": 0.8424, + "step": 9941 + }, + { + "epoch": 0.8966045903413447, + "grad_norm": 6.59360594388698, + "learning_rate": 1.1101682269571889e-07, + "loss": 1.0275, + "step": 9942 + }, + { + "epoch": 0.8966947738648149, + "grad_norm": 2.155919588730361, + "learning_rate": 1.1082496929972473e-07, + "loss": 0.9569, + "step": 9943 + }, + { + "epoch": 0.8967849573882851, + "grad_norm": 1.759343950936062, + "learning_rate": 1.1063327709856096e-07, + "loss": 0.945, + "step": 9944 + }, + { + "epoch": 0.8968751409117555, + "grad_norm": 1.915871959903635, + "learning_rate": 1.1044174610858403e-07, + "loss": 0.8622, + "step": 9945 + }, + { + "epoch": 0.8969653244352257, + "grad_norm": 1.6309910155749947, + "learning_rate": 1.1025037634613643e-07, + "loss": 0.931, + "step": 9946 + }, + { + "epoch": 0.8970555079586959, + "grad_norm": 1.6727879251021494, + "learning_rate": 1.1005916782754643e-07, + "loss": 1.0006, + "step": 9947 + }, + { + "epoch": 0.8971456914821662, + "grad_norm": 1.9986446079430589, + "learning_rate": 1.0986812056912919e-07, + "loss": 0.9641, + "step": 9948 + }, + { + "epoch": 0.8972358750056365, + "grad_norm": 1.7691905386283429, + "learning_rate": 1.0967723458718613e-07, + "loss": 0.8553, + "step": 9949 + }, + { + "epoch": 0.8973260585291067, + "grad_norm": 1.9241457914087667, + "learning_rate": 1.0948650989800445e-07, + "loss": 0.941, + "step": 9950 + }, + { + "epoch": 0.897416242052577, + "grad_norm": 1.4357089619378545, + "learning_rate": 1.0929594651785823e-07, + "loss": 0.9604, + "step": 9951 + }, + { + "epoch": 0.8975064255760472, + "grad_norm": 1.4814916614783173, + "learning_rate": 1.0910554446300646e-07, + "loss": 1.0276, + "step": 9952 + }, + { + "epoch": 0.8975966090995176, + "grad_norm": 4.5595131072056665, + "learning_rate": 1.089153037496966e-07, + "loss": 0.8751, + "step": 9953 + }, + { + "epoch": 0.8976867926229878, + "grad_norm": 1.7487428100307247, + "learning_rate": 1.0872522439415966e-07, + "loss": 0.9947, + "step": 9954 + }, + { + "epoch": 0.897776976146458, + "grad_norm": 2.799093719679692, + "learning_rate": 1.0853530641261554e-07, + "loss": 0.9234, + "step": 9955 + }, + { + "epoch": 0.8978671596699282, + "grad_norm": 1.7392947553409173, + "learning_rate": 1.083455498212682e-07, + "loss": 0.9885, + "step": 9956 + }, + { + "epoch": 0.8979573431933986, + "grad_norm": 1.9197193244582622, + "learning_rate": 1.0815595463630911e-07, + "loss": 1.0153, + "step": 9957 + }, + { + "epoch": 0.8980475267168688, + "grad_norm": 1.642236844237024, + "learning_rate": 1.0796652087391556e-07, + "loss": 0.8461, + "step": 9958 + }, + { + "epoch": 0.8981377102403391, + "grad_norm": 9.682152632025781, + "learning_rate": 1.0777724855025083e-07, + "loss": 1.0767, + "step": 9959 + }, + { + "epoch": 0.8982278937638094, + "grad_norm": 1.621090440861645, + "learning_rate": 1.075881376814649e-07, + "loss": 0.9755, + "step": 9960 + }, + { + "epoch": 0.8983180772872796, + "grad_norm": 1.6778241134642349, + "learning_rate": 1.0739918828369377e-07, + "loss": 1.0325, + "step": 9961 + }, + { + "epoch": 0.8984082608107499, + "grad_norm": 1.9275651140507912, + "learning_rate": 1.0721040037305983e-07, + "loss": 0.9808, + "step": 9962 + }, + { + "epoch": 0.8984984443342201, + "grad_norm": 2.432742831899259, + "learning_rate": 1.0702177396567114e-07, + "loss": 0.9266, + "step": 9963 + }, + { + "epoch": 0.8985886278576904, + "grad_norm": 0.6151586819058795, + "learning_rate": 1.0683330907762233e-07, + "loss": 0.7991, + "step": 9964 + }, + { + "epoch": 0.8986788113811607, + "grad_norm": 1.4338530012413062, + "learning_rate": 1.0664500572499435e-07, + "loss": 0.9078, + "step": 9965 + }, + { + "epoch": 0.8987689949046309, + "grad_norm": 3.0023571926335153, + "learning_rate": 1.0645686392385455e-07, + "loss": 0.9193, + "step": 9966 + }, + { + "epoch": 0.8988591784281011, + "grad_norm": 1.39722048692151, + "learning_rate": 1.0626888369025588e-07, + "loss": 0.9014, + "step": 9967 + }, + { + "epoch": 0.8989493619515715, + "grad_norm": 1.4470389287703884, + "learning_rate": 1.0608106504023817e-07, + "loss": 1.0069, + "step": 9968 + }, + { + "epoch": 0.8990395454750417, + "grad_norm": 1.4909969193742065, + "learning_rate": 1.0589340798982637e-07, + "loss": 0.9656, + "step": 9969 + }, + { + "epoch": 0.899129728998512, + "grad_norm": 1.7108811321292186, + "learning_rate": 1.057059125550337e-07, + "loss": 0.9199, + "step": 9970 + }, + { + "epoch": 0.8992199125219822, + "grad_norm": 1.3577593706313695, + "learning_rate": 1.0551857875185732e-07, + "loss": 0.9872, + "step": 9971 + }, + { + "epoch": 0.8993100960454525, + "grad_norm": 1.488846318662777, + "learning_rate": 1.0533140659628181e-07, + "loss": 0.8609, + "step": 9972 + }, + { + "epoch": 0.8994002795689228, + "grad_norm": 1.7928272598720252, + "learning_rate": 1.0514439610427772e-07, + "loss": 1.0352, + "step": 9973 + }, + { + "epoch": 0.899490463092393, + "grad_norm": 1.627681134247853, + "learning_rate": 1.0495754729180206e-07, + "loss": 1.0006, + "step": 9974 + }, + { + "epoch": 0.8995806466158632, + "grad_norm": 1.587796104029671, + "learning_rate": 1.0477086017479741e-07, + "loss": 1.0778, + "step": 9975 + }, + { + "epoch": 0.8996708301393336, + "grad_norm": 1.739510091126797, + "learning_rate": 1.0458433476919327e-07, + "loss": 0.9346, + "step": 9976 + }, + { + "epoch": 0.8997610136628038, + "grad_norm": 1.326339463893671, + "learning_rate": 1.0439797109090509e-07, + "loss": 1.0024, + "step": 9977 + }, + { + "epoch": 0.899851197186274, + "grad_norm": 2.1263732960524684, + "learning_rate": 1.0421176915583396e-07, + "loss": 0.9491, + "step": 9978 + }, + { + "epoch": 0.8999413807097443, + "grad_norm": 2.0704369849440645, + "learning_rate": 1.0402572897986828e-07, + "loss": 1.0067, + "step": 9979 + }, + { + "epoch": 0.9000315642332146, + "grad_norm": 0.5976747746806758, + "learning_rate": 1.0383985057888134e-07, + "loss": 0.805, + "step": 9980 + }, + { + "epoch": 0.9001217477566849, + "grad_norm": 1.60915543623248, + "learning_rate": 1.036541339687338e-07, + "loss": 0.8808, + "step": 9981 + }, + { + "epoch": 0.9002119312801551, + "grad_norm": 0.6485173049903159, + "learning_rate": 1.0346857916527186e-07, + "loss": 0.7755, + "step": 9982 + }, + { + "epoch": 0.9003021148036254, + "grad_norm": 1.9605516152819573, + "learning_rate": 1.0328318618432819e-07, + "loss": 0.9723, + "step": 9983 + }, + { + "epoch": 0.9003922983270957, + "grad_norm": 1.660037909385115, + "learning_rate": 1.0309795504172148e-07, + "loss": 1.0379, + "step": 9984 + }, + { + "epoch": 0.9004824818505659, + "grad_norm": 1.9798551733822247, + "learning_rate": 1.0291288575325685e-07, + "loss": 0.9517, + "step": 9985 + }, + { + "epoch": 0.9005726653740361, + "grad_norm": 1.787011849499114, + "learning_rate": 1.0272797833472502e-07, + "loss": 0.9369, + "step": 9986 + }, + { + "epoch": 0.9006628488975065, + "grad_norm": 0.6544845084272356, + "learning_rate": 1.0254323280190335e-07, + "loss": 0.7841, + "step": 9987 + }, + { + "epoch": 0.9007530324209767, + "grad_norm": 8.72354486349165, + "learning_rate": 1.023586491705557e-07, + "loss": 0.9545, + "step": 9988 + }, + { + "epoch": 0.9008432159444469, + "grad_norm": 3.429350886412213, + "learning_rate": 1.0217422745643145e-07, + "loss": 0.9128, + "step": 9989 + }, + { + "epoch": 0.9009333994679172, + "grad_norm": 2.094716546450413, + "learning_rate": 1.0198996767526691e-07, + "loss": 0.8958, + "step": 9990 + }, + { + "epoch": 0.9010235829913875, + "grad_norm": 2.077577102827854, + "learning_rate": 1.018058698427835e-07, + "loss": 0.9533, + "step": 9991 + }, + { + "epoch": 0.9011137665148578, + "grad_norm": 1.382143020207426, + "learning_rate": 1.0162193397469021e-07, + "loss": 1.0214, + "step": 9992 + }, + { + "epoch": 0.901203950038328, + "grad_norm": 1.908670295994797, + "learning_rate": 1.0143816008668049e-07, + "loss": 0.913, + "step": 9993 + }, + { + "epoch": 0.9012941335617982, + "grad_norm": 0.6441358335928671, + "learning_rate": 1.0125454819443624e-07, + "loss": 0.8235, + "step": 9994 + }, + { + "epoch": 0.9013843170852686, + "grad_norm": 1.4564607950650938, + "learning_rate": 1.0107109831362315e-07, + "loss": 0.8818, + "step": 9995 + }, + { + "epoch": 0.9014745006087388, + "grad_norm": 1.5853697413527468, + "learning_rate": 1.0088781045989447e-07, + "loss": 0.9839, + "step": 9996 + }, + { + "epoch": 0.901564684132209, + "grad_norm": 1.4681793078125425, + "learning_rate": 1.0070468464888926e-07, + "loss": 1.0243, + "step": 9997 + }, + { + "epoch": 0.9016548676556793, + "grad_norm": 2.4149610705031868, + "learning_rate": 1.0052172089623324e-07, + "loss": 0.9948, + "step": 9998 + }, + { + "epoch": 0.9017450511791496, + "grad_norm": 1.6301351983806096, + "learning_rate": 1.0033891921753746e-07, + "loss": 1.0157, + "step": 9999 + }, + { + "epoch": 0.9018352347026198, + "grad_norm": 1.7996003422511466, + "learning_rate": 1.0015627962839968e-07, + "loss": 0.8877, + "step": 10000 + }, + { + "epoch": 0.9019254182260901, + "grad_norm": 1.7920423453896612, + "learning_rate": 9.99738021444041e-08, + "loss": 0.9473, + "step": 10001 + }, + { + "epoch": 0.9020156017495603, + "grad_norm": 1.5177361640481748, + "learning_rate": 9.979148678112003e-08, + "loss": 0.9074, + "step": 10002 + }, + { + "epoch": 0.9021057852730306, + "grad_norm": 2.1074285498257943, + "learning_rate": 9.960933355410417e-08, + "loss": 1.0743, + "step": 10003 + }, + { + "epoch": 0.9021959687965009, + "grad_norm": 1.4541181226495734, + "learning_rate": 9.942734247889828e-08, + "loss": 0.8908, + "step": 10004 + }, + { + "epoch": 0.9022861523199711, + "grad_norm": 1.5192555469175024, + "learning_rate": 9.92455135710315e-08, + "loss": 0.98, + "step": 10005 + }, + { + "epoch": 0.9023763358434415, + "grad_norm": 2.524040504426793, + "learning_rate": 9.906384684601787e-08, + "loss": 0.9066, + "step": 10006 + }, + { + "epoch": 0.9024665193669117, + "grad_norm": 1.6246487164440384, + "learning_rate": 9.8882342319359e-08, + "loss": 1.0192, + "step": 10007 + }, + { + "epoch": 0.9025567028903819, + "grad_norm": 1.3938109944771926, + "learning_rate": 9.870100000654048e-08, + "loss": 0.9262, + "step": 10008 + }, + { + "epoch": 0.9026468864138522, + "grad_norm": 1.2987894841232792, + "learning_rate": 9.851981992303704e-08, + "loss": 0.947, + "step": 10009 + }, + { + "epoch": 0.9027370699373225, + "grad_norm": 1.4484291907671925, + "learning_rate": 9.833880208430678e-08, + "loss": 0.9797, + "step": 10010 + }, + { + "epoch": 0.9028272534607927, + "grad_norm": 1.872468001699606, + "learning_rate": 9.815794650579601e-08, + "loss": 0.966, + "step": 10011 + }, + { + "epoch": 0.902917436984263, + "grad_norm": 1.5531832831823895, + "learning_rate": 9.797725320293548e-08, + "loss": 0.9529, + "step": 10012 + }, + { + "epoch": 0.9030076205077332, + "grad_norm": 1.9108048724408842, + "learning_rate": 9.779672219114354e-08, + "loss": 1.0492, + "step": 10013 + }, + { + "epoch": 0.9030978040312035, + "grad_norm": 1.7644039421667088, + "learning_rate": 9.761635348582386e-08, + "loss": 0.9545, + "step": 10014 + }, + { + "epoch": 0.9031879875546738, + "grad_norm": 1.842875451042422, + "learning_rate": 9.743614710236658e-08, + "loss": 1.006, + "step": 10015 + }, + { + "epoch": 0.903278171078144, + "grad_norm": 2.316132038733891, + "learning_rate": 9.725610305614806e-08, + "loss": 0.9303, + "step": 10016 + }, + { + "epoch": 0.9033683546016142, + "grad_norm": 1.581950545901592, + "learning_rate": 9.707622136253002e-08, + "loss": 0.9678, + "step": 10017 + }, + { + "epoch": 0.9034585381250846, + "grad_norm": 1.4865204414735445, + "learning_rate": 9.689650203686173e-08, + "loss": 0.996, + "step": 10018 + }, + { + "epoch": 0.9035487216485548, + "grad_norm": 1.9338998879732323, + "learning_rate": 9.671694509447715e-08, + "loss": 0.9486, + "step": 10019 + }, + { + "epoch": 0.903638905172025, + "grad_norm": 2.193362201203099, + "learning_rate": 9.653755055069757e-08, + "loss": 0.9889, + "step": 10020 + }, + { + "epoch": 0.9037290886954953, + "grad_norm": 1.441578511021627, + "learning_rate": 9.635831842082987e-08, + "loss": 0.9508, + "step": 10021 + }, + { + "epoch": 0.9038192722189656, + "grad_norm": 3.4157011171953084, + "learning_rate": 9.617924872016691e-08, + "loss": 0.9894, + "step": 10022 + }, + { + "epoch": 0.9039094557424359, + "grad_norm": 1.6259901258114855, + "learning_rate": 9.600034146398806e-08, + "loss": 1.0131, + "step": 10023 + }, + { + "epoch": 0.9039996392659061, + "grad_norm": 0.6641498732587605, + "learning_rate": 9.582159666755863e-08, + "loss": 0.8174, + "step": 10024 + }, + { + "epoch": 0.9040898227893763, + "grad_norm": 1.574977417435799, + "learning_rate": 9.564301434612976e-08, + "loss": 0.941, + "step": 10025 + }, + { + "epoch": 0.9041800063128467, + "grad_norm": 1.5282501887117454, + "learning_rate": 9.546459451494015e-08, + "loss": 1.0168, + "step": 10026 + }, + { + "epoch": 0.9042701898363169, + "grad_norm": 0.6338829358942571, + "learning_rate": 9.528633718921231e-08, + "loss": 0.8421, + "step": 10027 + }, + { + "epoch": 0.9043603733597871, + "grad_norm": 1.2964972439937819, + "learning_rate": 9.510824238415672e-08, + "loss": 1.0242, + "step": 10028 + }, + { + "epoch": 0.9044505568832575, + "grad_norm": 1.4062485588914058, + "learning_rate": 9.493031011496944e-08, + "loss": 1.0072, + "step": 10029 + }, + { + "epoch": 0.9045407404067277, + "grad_norm": 2.9439880469224673, + "learning_rate": 9.475254039683234e-08, + "loss": 0.8283, + "step": 10030 + }, + { + "epoch": 0.904630923930198, + "grad_norm": 1.4636565827384522, + "learning_rate": 9.45749332449144e-08, + "loss": 0.9387, + "step": 10031 + }, + { + "epoch": 0.9047211074536682, + "grad_norm": 0.6489343692851349, + "learning_rate": 9.439748867436903e-08, + "loss": 0.8264, + "step": 10032 + }, + { + "epoch": 0.9048112909771385, + "grad_norm": 1.6997226713272457, + "learning_rate": 9.42202067003377e-08, + "loss": 0.9405, + "step": 10033 + }, + { + "epoch": 0.9049014745006088, + "grad_norm": 1.7670329733828447, + "learning_rate": 9.404308733794652e-08, + "loss": 1.106, + "step": 10034 + }, + { + "epoch": 0.904991658024079, + "grad_norm": 1.5340290930524423, + "learning_rate": 9.38661306023083e-08, + "loss": 0.9957, + "step": 10035 + }, + { + "epoch": 0.9050818415475492, + "grad_norm": 1.8714533004458092, + "learning_rate": 9.368933650852229e-08, + "loss": 0.9648, + "step": 10036 + }, + { + "epoch": 0.9051720250710196, + "grad_norm": 0.6943945236276869, + "learning_rate": 9.351270507167352e-08, + "loss": 0.8302, + "step": 10037 + }, + { + "epoch": 0.9052622085944898, + "grad_norm": 1.850035540780633, + "learning_rate": 9.333623630683285e-08, + "loss": 0.91, + "step": 10038 + }, + { + "epoch": 0.90535239211796, + "grad_norm": 1.8990404717751388, + "learning_rate": 9.315993022905799e-08, + "loss": 0.8243, + "step": 10039 + }, + { + "epoch": 0.9054425756414303, + "grad_norm": 2.227469490433157, + "learning_rate": 9.298378685339158e-08, + "loss": 0.952, + "step": 10040 + }, + { + "epoch": 0.9055327591649006, + "grad_norm": 2.508097409615381, + "learning_rate": 9.280780619486406e-08, + "loss": 1.0365, + "step": 10041 + }, + { + "epoch": 0.9056229426883708, + "grad_norm": 1.6550682458319372, + "learning_rate": 9.26319882684905e-08, + "loss": 1.0546, + "step": 10042 + }, + { + "epoch": 0.9057131262118411, + "grad_norm": 1.7365937346278086, + "learning_rate": 9.245633308927293e-08, + "loss": 0.9414, + "step": 10043 + }, + { + "epoch": 0.9058033097353113, + "grad_norm": 1.7496345002094351, + "learning_rate": 9.228084067219888e-08, + "loss": 0.9577, + "step": 10044 + }, + { + "epoch": 0.9058934932587817, + "grad_norm": 2.042675923646879, + "learning_rate": 9.210551103224284e-08, + "loss": 0.9214, + "step": 10045 + }, + { + "epoch": 0.9059836767822519, + "grad_norm": 2.0282218075352416, + "learning_rate": 9.193034418436463e-08, + "loss": 1.0252, + "step": 10046 + }, + { + "epoch": 0.9060738603057221, + "grad_norm": 0.6509857662361767, + "learning_rate": 9.175534014351005e-08, + "loss": 0.8504, + "step": 10047 + }, + { + "epoch": 0.9061640438291924, + "grad_norm": 1.8387552268798726, + "learning_rate": 9.158049892461228e-08, + "loss": 0.9216, + "step": 10048 + }, + { + "epoch": 0.9062542273526627, + "grad_norm": 1.5177565853885249, + "learning_rate": 9.140582054258871e-08, + "loss": 0.9109, + "step": 10049 + }, + { + "epoch": 0.9063444108761329, + "grad_norm": 1.6431113159027766, + "learning_rate": 9.123130501234499e-08, + "loss": 0.9807, + "step": 10050 + }, + { + "epoch": 0.9064345943996032, + "grad_norm": 3.206894605655101, + "learning_rate": 9.105695234877098e-08, + "loss": 1.0261, + "step": 10051 + }, + { + "epoch": 0.9065247779230734, + "grad_norm": 1.8283482602720533, + "learning_rate": 9.088276256674344e-08, + "loss": 1.0004, + "step": 10052 + }, + { + "epoch": 0.9066149614465437, + "grad_norm": 2.1015224523876506, + "learning_rate": 9.070873568112536e-08, + "loss": 0.9614, + "step": 10053 + }, + { + "epoch": 0.906705144970014, + "grad_norm": 1.619013175043664, + "learning_rate": 9.053487170676577e-08, + "loss": 1.0329, + "step": 10054 + }, + { + "epoch": 0.9067953284934842, + "grad_norm": 1.7846514031806417, + "learning_rate": 9.036117065849968e-08, + "loss": 1.0338, + "step": 10055 + }, + { + "epoch": 0.9068855120169546, + "grad_norm": 1.493238707947582, + "learning_rate": 9.018763255114837e-08, + "loss": 0.9553, + "step": 10056 + }, + { + "epoch": 0.9069756955404248, + "grad_norm": 4.131440654285477, + "learning_rate": 9.00142573995184e-08, + "loss": 0.9423, + "step": 10057 + }, + { + "epoch": 0.907065879063895, + "grad_norm": 1.6395068036467861, + "learning_rate": 8.984104521840375e-08, + "loss": 1.0526, + "step": 10058 + }, + { + "epoch": 0.9071560625873653, + "grad_norm": 1.63271108796072, + "learning_rate": 8.966799602258346e-08, + "loss": 0.9954, + "step": 10059 + }, + { + "epoch": 0.9072462461108356, + "grad_norm": 2.2112526820604828, + "learning_rate": 8.949510982682329e-08, + "loss": 1.0174, + "step": 10060 + }, + { + "epoch": 0.9073364296343058, + "grad_norm": 2.3223373876829836, + "learning_rate": 8.932238664587499e-08, + "loss": 1.0419, + "step": 10061 + }, + { + "epoch": 0.9074266131577761, + "grad_norm": 2.432116015433992, + "learning_rate": 8.914982649447567e-08, + "loss": 1.0676, + "step": 10062 + }, + { + "epoch": 0.9075167966812463, + "grad_norm": 1.4744441261920396, + "learning_rate": 8.897742938734975e-08, + "loss": 1.0412, + "step": 10063 + }, + { + "epoch": 0.9076069802047166, + "grad_norm": 2.056301165005061, + "learning_rate": 8.880519533920661e-08, + "loss": 1.0012, + "step": 10064 + }, + { + "epoch": 0.9076971637281869, + "grad_norm": 1.3852237433241277, + "learning_rate": 8.863312436474268e-08, + "loss": 0.9753, + "step": 10065 + }, + { + "epoch": 0.9077873472516571, + "grad_norm": 2.051538411192951, + "learning_rate": 8.846121647863936e-08, + "loss": 0.981, + "step": 10066 + }, + { + "epoch": 0.9078775307751273, + "grad_norm": 1.4164922083227334, + "learning_rate": 8.828947169556555e-08, + "loss": 0.979, + "step": 10067 + }, + { + "epoch": 0.9079677142985977, + "grad_norm": 1.5604260986341858, + "learning_rate": 8.81178900301749e-08, + "loss": 0.9393, + "step": 10068 + }, + { + "epoch": 0.9080578978220679, + "grad_norm": 1.541706583004637, + "learning_rate": 8.794647149710787e-08, + "loss": 0.891, + "step": 10069 + }, + { + "epoch": 0.9081480813455381, + "grad_norm": 1.794610163506769, + "learning_rate": 8.777521611099081e-08, + "loss": 0.9693, + "step": 10070 + }, + { + "epoch": 0.9082382648690084, + "grad_norm": 1.5840005556307888, + "learning_rate": 8.760412388643624e-08, + "loss": 0.9126, + "step": 10071 + }, + { + "epoch": 0.9083284483924787, + "grad_norm": 1.5456193393676119, + "learning_rate": 8.74331948380429e-08, + "loss": 0.9894, + "step": 10072 + }, + { + "epoch": 0.908418631915949, + "grad_norm": 1.60165449669085, + "learning_rate": 8.726242898039516e-08, + "loss": 0.9732, + "step": 10073 + }, + { + "epoch": 0.9085088154394192, + "grad_norm": 2.0469927571908855, + "learning_rate": 8.709182632806334e-08, + "loss": 0.9817, + "step": 10074 + }, + { + "epoch": 0.9085989989628894, + "grad_norm": 2.5695171575023537, + "learning_rate": 8.692138689560469e-08, + "loss": 0.9831, + "step": 10075 + }, + { + "epoch": 0.9086891824863598, + "grad_norm": 1.5878752347568668, + "learning_rate": 8.675111069756203e-08, + "loss": 0.9573, + "step": 10076 + }, + { + "epoch": 0.90877936600983, + "grad_norm": 1.5350406981798104, + "learning_rate": 8.658099774846395e-08, + "loss": 0.9464, + "step": 10077 + }, + { + "epoch": 0.9088695495333002, + "grad_norm": 1.4511945473708368, + "learning_rate": 8.641104806282595e-08, + "loss": 0.9194, + "step": 10078 + }, + { + "epoch": 0.9089597330567706, + "grad_norm": 1.7498506073536764, + "learning_rate": 8.624126165514845e-08, + "loss": 0.9845, + "step": 10079 + }, + { + "epoch": 0.9090499165802408, + "grad_norm": 1.8373257197557744, + "learning_rate": 8.607163853991917e-08, + "loss": 0.94, + "step": 10080 + }, + { + "epoch": 0.909140100103711, + "grad_norm": 1.5520025484024833, + "learning_rate": 8.590217873161054e-08, + "loss": 0.8207, + "step": 10081 + }, + { + "epoch": 0.9092302836271813, + "grad_norm": 1.6018160735569016, + "learning_rate": 8.573288224468255e-08, + "loss": 0.991, + "step": 10082 + }, + { + "epoch": 0.9093204671506516, + "grad_norm": 2.110106503840517, + "learning_rate": 8.556374909358011e-08, + "loss": 0.9037, + "step": 10083 + }, + { + "epoch": 0.9094106506741219, + "grad_norm": 1.6292859556554278, + "learning_rate": 8.539477929273476e-08, + "loss": 0.9518, + "step": 10084 + }, + { + "epoch": 0.9095008341975921, + "grad_norm": 2.7647313974991103, + "learning_rate": 8.522597285656386e-08, + "loss": 1.0062, + "step": 10085 + }, + { + "epoch": 0.9095910177210623, + "grad_norm": 3.2074873828714114, + "learning_rate": 8.505732979947078e-08, + "loss": 0.8264, + "step": 10086 + }, + { + "epoch": 0.9096812012445327, + "grad_norm": 1.3858014148531514, + "learning_rate": 8.488885013584557e-08, + "loss": 0.9908, + "step": 10087 + }, + { + "epoch": 0.9097713847680029, + "grad_norm": 1.5049838082339173, + "learning_rate": 8.472053388006295e-08, + "loss": 0.9149, + "step": 10088 + }, + { + "epoch": 0.9098615682914731, + "grad_norm": 1.911387415375034, + "learning_rate": 8.455238104648565e-08, + "loss": 1.0621, + "step": 10089 + }, + { + "epoch": 0.9099517518149434, + "grad_norm": 2.5879707755866423, + "learning_rate": 8.438439164946043e-08, + "loss": 0.846, + "step": 10090 + }, + { + "epoch": 0.9100419353384137, + "grad_norm": 1.5940021240670137, + "learning_rate": 8.42165657033218e-08, + "loss": 0.9877, + "step": 10091 + }, + { + "epoch": 0.9101321188618839, + "grad_norm": 1.5892853377144307, + "learning_rate": 8.4048903222389e-08, + "loss": 1.0109, + "step": 10092 + }, + { + "epoch": 0.9102223023853542, + "grad_norm": 1.5341475961146727, + "learning_rate": 8.388140422096856e-08, + "loss": 1.0267, + "step": 10093 + }, + { + "epoch": 0.9103124859088244, + "grad_norm": 1.7409019345793557, + "learning_rate": 8.371406871335173e-08, + "loss": 0.9863, + "step": 10094 + }, + { + "epoch": 0.9104026694322948, + "grad_norm": 9.799906625108212, + "learning_rate": 8.354689671381732e-08, + "loss": 0.9788, + "step": 10095 + }, + { + "epoch": 0.910492852955765, + "grad_norm": 1.400972721525889, + "learning_rate": 8.337988823662834e-08, + "loss": 0.9719, + "step": 10096 + }, + { + "epoch": 0.9105830364792352, + "grad_norm": 1.5226475159035915, + "learning_rate": 8.321304329603607e-08, + "loss": 0.9267, + "step": 10097 + }, + { + "epoch": 0.9106732200027055, + "grad_norm": 0.6255770880066983, + "learning_rate": 8.304636190627557e-08, + "loss": 0.8233, + "step": 10098 + }, + { + "epoch": 0.9107634035261758, + "grad_norm": 1.8281971191787516, + "learning_rate": 8.287984408156945e-08, + "loss": 0.9882, + "step": 10099 + }, + { + "epoch": 0.910853587049646, + "grad_norm": 2.742065687815959, + "learning_rate": 8.271348983612591e-08, + "loss": 0.8781, + "step": 10100 + }, + { + "epoch": 0.9109437705731163, + "grad_norm": 2.6975805497892584, + "learning_rate": 8.254729918413938e-08, + "loss": 0.936, + "step": 10101 + }, + { + "epoch": 0.9110339540965866, + "grad_norm": 1.8202543454028637, + "learning_rate": 8.238127213979006e-08, + "loss": 0.9448, + "step": 10102 + }, + { + "epoch": 0.9111241376200568, + "grad_norm": 1.8580875587732983, + "learning_rate": 8.221540871724398e-08, + "loss": 1.0441, + "step": 10103 + }, + { + "epoch": 0.9112143211435271, + "grad_norm": 4.456764634471971, + "learning_rate": 8.2049708930654e-08, + "loss": 0.8621, + "step": 10104 + }, + { + "epoch": 0.9113045046669973, + "grad_norm": 3.6809456976357136, + "learning_rate": 8.188417279415793e-08, + "loss": 0.9407, + "step": 10105 + }, + { + "epoch": 0.9113946881904677, + "grad_norm": 1.7434506477260887, + "learning_rate": 8.171880032188117e-08, + "loss": 1.0638, + "step": 10106 + }, + { + "epoch": 0.9114848717139379, + "grad_norm": 1.4173845921652823, + "learning_rate": 8.155359152793351e-08, + "loss": 0.9563, + "step": 10107 + }, + { + "epoch": 0.9115750552374081, + "grad_norm": 2.121569388717366, + "learning_rate": 8.138854642641147e-08, + "loss": 0.9606, + "step": 10108 + }, + { + "epoch": 0.9116652387608783, + "grad_norm": 1.6328483468095218, + "learning_rate": 8.122366503139777e-08, + "loss": 0.9012, + "step": 10109 + }, + { + "epoch": 0.9117554222843487, + "grad_norm": 1.816325655554071, + "learning_rate": 8.105894735696117e-08, + "loss": 0.9148, + "step": 10110 + }, + { + "epoch": 0.9118456058078189, + "grad_norm": 2.342221282857859, + "learning_rate": 8.089439341715576e-08, + "loss": 0.9507, + "step": 10111 + }, + { + "epoch": 0.9119357893312892, + "grad_norm": 1.9071632448967069, + "learning_rate": 8.073000322602319e-08, + "loss": 1.0034, + "step": 10112 + }, + { + "epoch": 0.9120259728547594, + "grad_norm": 1.495478730244645, + "learning_rate": 8.056577679758891e-08, + "loss": 0.9961, + "step": 10113 + }, + { + "epoch": 0.9121161563782297, + "grad_norm": 2.1503978316821226, + "learning_rate": 8.040171414586638e-08, + "loss": 0.9254, + "step": 10114 + }, + { + "epoch": 0.9122063399017, + "grad_norm": 1.9019906658260457, + "learning_rate": 8.023781528485419e-08, + "loss": 0.9395, + "step": 10115 + }, + { + "epoch": 0.9122965234251702, + "grad_norm": 1.9118983899494268, + "learning_rate": 8.00740802285369e-08, + "loss": 0.9994, + "step": 10116 + }, + { + "epoch": 0.9123867069486404, + "grad_norm": 1.5210796400609934, + "learning_rate": 7.99105089908858e-08, + "loss": 0.8408, + "step": 10117 + }, + { + "epoch": 0.9124768904721108, + "grad_norm": 1.5327881852923058, + "learning_rate": 7.974710158585685e-08, + "loss": 0.9504, + "step": 10118 + }, + { + "epoch": 0.912567073995581, + "grad_norm": 1.6496279065741812, + "learning_rate": 7.958385802739375e-08, + "loss": 0.9992, + "step": 10119 + }, + { + "epoch": 0.9126572575190512, + "grad_norm": 5.903260781466791, + "learning_rate": 7.942077832942452e-08, + "loss": 1.0343, + "step": 10120 + }, + { + "epoch": 0.9127474410425215, + "grad_norm": 0.7828327930127231, + "learning_rate": 7.925786250586508e-08, + "loss": 0.8171, + "step": 10121 + }, + { + "epoch": 0.9128376245659918, + "grad_norm": 0.6563191377460135, + "learning_rate": 7.909511057061524e-08, + "loss": 0.817, + "step": 10122 + }, + { + "epoch": 0.9129278080894621, + "grad_norm": 1.7538407957102, + "learning_rate": 7.893252253756234e-08, + "loss": 0.9587, + "step": 10123 + }, + { + "epoch": 0.9130179916129323, + "grad_norm": 1.8904459214323883, + "learning_rate": 7.877009842057925e-08, + "loss": 0.9029, + "step": 10124 + }, + { + "epoch": 0.9131081751364026, + "grad_norm": 2.125626527762712, + "learning_rate": 7.860783823352512e-08, + "loss": 1.0012, + "step": 10125 + }, + { + "epoch": 0.9131983586598729, + "grad_norm": 1.6179613114153306, + "learning_rate": 7.844574199024445e-08, + "loss": 0.9829, + "step": 10126 + }, + { + "epoch": 0.9132885421833431, + "grad_norm": 1.6674286372114613, + "learning_rate": 7.82838097045686e-08, + "loss": 1.0188, + "step": 10127 + }, + { + "epoch": 0.9133787257068133, + "grad_norm": 1.9708544065331408, + "learning_rate": 7.812204139031454e-08, + "loss": 0.9656, + "step": 10128 + }, + { + "epoch": 0.9134689092302837, + "grad_norm": 2.124702208353474, + "learning_rate": 7.796043706128474e-08, + "loss": 0.9918, + "step": 10129 + }, + { + "epoch": 0.9135590927537539, + "grad_norm": 1.7669072179349914, + "learning_rate": 7.779899673126844e-08, + "loss": 1.0003, + "step": 10130 + }, + { + "epoch": 0.9136492762772241, + "grad_norm": 1.5153990114973817, + "learning_rate": 7.76377204140406e-08, + "loss": 1.02, + "step": 10131 + }, + { + "epoch": 0.9137394598006944, + "grad_norm": 1.6575301639119435, + "learning_rate": 7.74766081233622e-08, + "loss": 0.965, + "step": 10132 + }, + { + "epoch": 0.9138296433241647, + "grad_norm": 1.81319019725076, + "learning_rate": 7.73156598729805e-08, + "loss": 1.0444, + "step": 10133 + }, + { + "epoch": 0.913919826847635, + "grad_norm": 1.8510000927975063, + "learning_rate": 7.715487567662849e-08, + "loss": 1.0273, + "step": 10134 + }, + { + "epoch": 0.9140100103711052, + "grad_norm": 3.0495399753535444, + "learning_rate": 7.69942555480243e-08, + "loss": 0.9432, + "step": 10135 + }, + { + "epoch": 0.9141001938945754, + "grad_norm": 2.3073356285528352, + "learning_rate": 7.68337995008741e-08, + "loss": 1.0793, + "step": 10136 + }, + { + "epoch": 0.9141903774180458, + "grad_norm": 1.9199319957571177, + "learning_rate": 7.667350754886803e-08, + "loss": 0.8293, + "step": 10137 + }, + { + "epoch": 0.914280560941516, + "grad_norm": 2.037553601642281, + "learning_rate": 7.651337970568361e-08, + "loss": 0.9673, + "step": 10138 + }, + { + "epoch": 0.9143707444649862, + "grad_norm": 3.75844982555941, + "learning_rate": 7.635341598498368e-08, + "loss": 0.9401, + "step": 10139 + }, + { + "epoch": 0.9144609279884565, + "grad_norm": 1.7131039215873147, + "learning_rate": 7.61936164004171e-08, + "loss": 0.9721, + "step": 10140 + }, + { + "epoch": 0.9145511115119268, + "grad_norm": 1.8273585742546474, + "learning_rate": 7.603398096561875e-08, + "loss": 1.0195, + "step": 10141 + }, + { + "epoch": 0.914641295035397, + "grad_norm": 1.4536088527603015, + "learning_rate": 7.587450969420994e-08, + "loss": 0.9778, + "step": 10142 + }, + { + "epoch": 0.9147314785588673, + "grad_norm": 0.6793996047729363, + "learning_rate": 7.571520259979757e-08, + "loss": 0.7972, + "step": 10143 + }, + { + "epoch": 0.9148216620823375, + "grad_norm": 2.722824113397033, + "learning_rate": 7.555605969597455e-08, + "loss": 0.8159, + "step": 10144 + }, + { + "epoch": 0.9149118456058078, + "grad_norm": 1.7273683350399478, + "learning_rate": 7.539708099631959e-08, + "loss": 1.056, + "step": 10145 + }, + { + "epoch": 0.9150020291292781, + "grad_norm": 1.5521937934774013, + "learning_rate": 7.52382665143978e-08, + "loss": 0.9823, + "step": 10146 + }, + { + "epoch": 0.9150922126527483, + "grad_norm": 1.7888193423348007, + "learning_rate": 7.507961626376014e-08, + "loss": 1.0297, + "step": 10147 + }, + { + "epoch": 0.9151823961762187, + "grad_norm": 1.8506995656903065, + "learning_rate": 7.492113025794378e-08, + "loss": 1.0121, + "step": 10148 + }, + { + "epoch": 0.9152725796996889, + "grad_norm": 2.7475044891903884, + "learning_rate": 7.476280851047101e-08, + "loss": 0.9245, + "step": 10149 + }, + { + "epoch": 0.9153627632231591, + "grad_norm": 1.7047103362822205, + "learning_rate": 7.460465103485125e-08, + "loss": 0.998, + "step": 10150 + }, + { + "epoch": 0.9154529467466294, + "grad_norm": 5.957298597996693, + "learning_rate": 7.444665784457948e-08, + "loss": 0.9668, + "step": 10151 + }, + { + "epoch": 0.9155431302700997, + "grad_norm": 1.6777570409875722, + "learning_rate": 7.42888289531356e-08, + "loss": 0.9331, + "step": 10152 + }, + { + "epoch": 0.9156333137935699, + "grad_norm": 1.8939364099190064, + "learning_rate": 7.41311643739877e-08, + "loss": 0.8967, + "step": 10153 + }, + { + "epoch": 0.9157234973170402, + "grad_norm": 1.5365311772379588, + "learning_rate": 7.39736641205877e-08, + "loss": 1.0144, + "step": 10154 + }, + { + "epoch": 0.9158136808405104, + "grad_norm": 1.8992252050028815, + "learning_rate": 7.381632820637462e-08, + "loss": 0.998, + "step": 10155 + }, + { + "epoch": 0.9159038643639807, + "grad_norm": 2.9777482041587406, + "learning_rate": 7.365915664477352e-08, + "loss": 0.9639, + "step": 10156 + }, + { + "epoch": 0.915994047887451, + "grad_norm": 1.6684183055390187, + "learning_rate": 7.350214944919474e-08, + "loss": 0.9789, + "step": 10157 + }, + { + "epoch": 0.9160842314109212, + "grad_norm": 1.646822603901435, + "learning_rate": 7.334530663303539e-08, + "loss": 1.0025, + "step": 10158 + }, + { + "epoch": 0.9161744149343914, + "grad_norm": 1.561819767702371, + "learning_rate": 7.318862820967742e-08, + "loss": 0.9837, + "step": 10159 + }, + { + "epoch": 0.9162645984578618, + "grad_norm": 1.6902393430033371, + "learning_rate": 7.303211419249056e-08, + "loss": 1.0049, + "step": 10160 + }, + { + "epoch": 0.916354781981332, + "grad_norm": 1.2296999512639455, + "learning_rate": 7.287576459482858e-08, + "loss": 0.9889, + "step": 10161 + }, + { + "epoch": 0.9164449655048023, + "grad_norm": 2.479693434286919, + "learning_rate": 7.271957943003259e-08, + "loss": 1.0136, + "step": 10162 + }, + { + "epoch": 0.9165351490282725, + "grad_norm": 2.0239643604020636, + "learning_rate": 7.256355871142883e-08, + "loss": 0.9734, + "step": 10163 + }, + { + "epoch": 0.9166253325517428, + "grad_norm": 1.6373310992078627, + "learning_rate": 7.240770245233019e-08, + "loss": 1.0249, + "step": 10164 + }, + { + "epoch": 0.9167155160752131, + "grad_norm": 1.7055810839799295, + "learning_rate": 7.225201066603492e-08, + "loss": 0.8913, + "step": 10165 + }, + { + "epoch": 0.9168056995986833, + "grad_norm": 1.5000340934693566, + "learning_rate": 7.209648336582774e-08, + "loss": 1.0862, + "step": 10166 + }, + { + "epoch": 0.9168958831221535, + "grad_norm": 1.631994234051523, + "learning_rate": 7.19411205649787e-08, + "loss": 0.9571, + "step": 10167 + }, + { + "epoch": 0.9169860666456239, + "grad_norm": 1.5450986430704796, + "learning_rate": 7.178592227674474e-08, + "loss": 0.9947, + "step": 10168 + }, + { + "epoch": 0.9170762501690941, + "grad_norm": 1.78484330053156, + "learning_rate": 7.163088851436771e-08, + "loss": 0.8945, + "step": 10169 + }, + { + "epoch": 0.9171664336925643, + "grad_norm": 1.5523121385662617, + "learning_rate": 7.147601929107639e-08, + "loss": 0.9408, + "step": 10170 + }, + { + "epoch": 0.9172566172160346, + "grad_norm": 1.4244848759157278, + "learning_rate": 7.132131462008461e-08, + "loss": 1.0191, + "step": 10171 + }, + { + "epoch": 0.9173468007395049, + "grad_norm": 1.7116442749034668, + "learning_rate": 7.116677451459297e-08, + "loss": 0.9783, + "step": 10172 + }, + { + "epoch": 0.9174369842629752, + "grad_norm": 1.7518000881415854, + "learning_rate": 7.101239898778799e-08, + "loss": 0.9991, + "step": 10173 + }, + { + "epoch": 0.9175271677864454, + "grad_norm": 1.614969031093095, + "learning_rate": 7.085818805284094e-08, + "loss": 1.0117, + "step": 10174 + }, + { + "epoch": 0.9176173513099157, + "grad_norm": 1.6767708228555747, + "learning_rate": 7.070414172291083e-08, + "loss": 1.0553, + "step": 10175 + }, + { + "epoch": 0.917707534833386, + "grad_norm": 1.4285680004487458, + "learning_rate": 7.055026001114095e-08, + "loss": 0.9197, + "step": 10176 + }, + { + "epoch": 0.9177977183568562, + "grad_norm": 0.6499812581954972, + "learning_rate": 7.039654293066211e-08, + "loss": 0.8018, + "step": 10177 + }, + { + "epoch": 0.9178879018803264, + "grad_norm": 1.721685503659348, + "learning_rate": 7.024299049459003e-08, + "loss": 0.8996, + "step": 10178 + }, + { + "epoch": 0.9179780854037968, + "grad_norm": 2.283593424256413, + "learning_rate": 7.008960271602627e-08, + "loss": 0.8888, + "step": 10179 + }, + { + "epoch": 0.918068268927267, + "grad_norm": 1.476301059082164, + "learning_rate": 6.993637960805921e-08, + "loss": 0.9396, + "step": 10180 + }, + { + "epoch": 0.9181584524507372, + "grad_norm": 2.2265725654240693, + "learning_rate": 6.97833211837624e-08, + "loss": 1.0017, + "step": 10181 + }, + { + "epoch": 0.9182486359742075, + "grad_norm": 2.7772214194611173, + "learning_rate": 6.963042745619562e-08, + "loss": 0.9234, + "step": 10182 + }, + { + "epoch": 0.9183388194976778, + "grad_norm": 1.445812077141579, + "learning_rate": 6.947769843840511e-08, + "loss": 0.945, + "step": 10183 + }, + { + "epoch": 0.918429003021148, + "grad_norm": 1.6583299352460898, + "learning_rate": 6.9325134143422e-08, + "loss": 1.0085, + "step": 10184 + }, + { + "epoch": 0.9185191865446183, + "grad_norm": 4.335287949305508, + "learning_rate": 6.917273458426387e-08, + "loss": 0.9544, + "step": 10185 + }, + { + "epoch": 0.9186093700680885, + "grad_norm": 0.6901740392664086, + "learning_rate": 6.902049977393476e-08, + "loss": 0.7906, + "step": 10186 + }, + { + "epoch": 0.9186995535915589, + "grad_norm": 1.9856539832267166, + "learning_rate": 6.886842972542362e-08, + "loss": 0.9837, + "step": 10187 + }, + { + "epoch": 0.9187897371150291, + "grad_norm": 1.7159610662333973, + "learning_rate": 6.871652445170672e-08, + "loss": 0.8666, + "step": 10188 + }, + { + "epoch": 0.9188799206384993, + "grad_norm": 1.781800837423522, + "learning_rate": 6.856478396574416e-08, + "loss": 1.0606, + "step": 10189 + }, + { + "epoch": 0.9189701041619696, + "grad_norm": 1.6443832222106978, + "learning_rate": 6.841320828048491e-08, + "loss": 1.0359, + "step": 10190 + }, + { + "epoch": 0.9190602876854399, + "grad_norm": 1.5734485038668022, + "learning_rate": 6.826179740886062e-08, + "loss": 0.9723, + "step": 10191 + }, + { + "epoch": 0.9191504712089101, + "grad_norm": 1.5764449590867902, + "learning_rate": 6.811055136379184e-08, + "loss": 1.0408, + "step": 10192 + }, + { + "epoch": 0.9192406547323804, + "grad_norm": 10.657898000774912, + "learning_rate": 6.79594701581827e-08, + "loss": 0.9528, + "step": 10193 + }, + { + "epoch": 0.9193308382558506, + "grad_norm": 1.8809539359179985, + "learning_rate": 6.780855380492511e-08, + "loss": 0.9824, + "step": 10194 + }, + { + "epoch": 0.919421021779321, + "grad_norm": 1.5737275902665266, + "learning_rate": 6.765780231689544e-08, + "loss": 0.9912, + "step": 10195 + }, + { + "epoch": 0.9195112053027912, + "grad_norm": 1.8491241882153273, + "learning_rate": 6.750721570695695e-08, + "loss": 1.0319, + "step": 10196 + }, + { + "epoch": 0.9196013888262614, + "grad_norm": 2.4406749881406977, + "learning_rate": 6.735679398795868e-08, + "loss": 0.961, + "step": 10197 + }, + { + "epoch": 0.9196915723497318, + "grad_norm": 2.347682959076176, + "learning_rate": 6.720653717273506e-08, + "loss": 0.9096, + "step": 10198 + }, + { + "epoch": 0.919781755873202, + "grad_norm": 1.6662631102587768, + "learning_rate": 6.705644527410714e-08, + "loss": 0.9749, + "step": 10199 + }, + { + "epoch": 0.9198719393966722, + "grad_norm": 1.8777781567707972, + "learning_rate": 6.690651830488136e-08, + "loss": 0.9445, + "step": 10200 + }, + { + "epoch": 0.9199621229201425, + "grad_norm": 2.243348143567893, + "learning_rate": 6.675675627785037e-08, + "loss": 0.9692, + "step": 10201 + }, + { + "epoch": 0.9200523064436128, + "grad_norm": 2.015886038503455, + "learning_rate": 6.660715920579263e-08, + "loss": 0.9969, + "step": 10202 + }, + { + "epoch": 0.920142489967083, + "grad_norm": 0.5911947529810945, + "learning_rate": 6.645772710147279e-08, + "loss": 0.8143, + "step": 10203 + }, + { + "epoch": 0.9202326734905533, + "grad_norm": 1.2345243918881756, + "learning_rate": 6.630845997764112e-08, + "loss": 0.9441, + "step": 10204 + }, + { + "epoch": 0.9203228570140235, + "grad_norm": 2.8744138037112577, + "learning_rate": 6.615935784703409e-08, + "loss": 0.9162, + "step": 10205 + }, + { + "epoch": 0.9204130405374938, + "grad_norm": 1.678059556420209, + "learning_rate": 6.601042072237328e-08, + "loss": 0.9025, + "step": 10206 + }, + { + "epoch": 0.9205032240609641, + "grad_norm": 1.4977931637135649, + "learning_rate": 6.586164861636767e-08, + "loss": 1.0044, + "step": 10207 + }, + { + "epoch": 0.9205934075844343, + "grad_norm": 1.9413074491293973, + "learning_rate": 6.571304154171065e-08, + "loss": 0.8883, + "step": 10208 + }, + { + "epoch": 0.9206835911079045, + "grad_norm": 1.6381223209479998, + "learning_rate": 6.556459951108273e-08, + "loss": 1.0026, + "step": 10209 + }, + { + "epoch": 0.9207737746313749, + "grad_norm": 3.3632225536581086, + "learning_rate": 6.541632253714957e-08, + "loss": 0.9556, + "step": 10210 + }, + { + "epoch": 0.9208639581548451, + "grad_norm": 1.9437754123412603, + "learning_rate": 6.526821063256261e-08, + "loss": 0.9033, + "step": 10211 + }, + { + "epoch": 0.9209541416783154, + "grad_norm": 1.7657588806706759, + "learning_rate": 6.512026380996016e-08, + "loss": 0.8816, + "step": 10212 + }, + { + "epoch": 0.9210443252017856, + "grad_norm": 1.793144799193312, + "learning_rate": 6.49724820819657e-08, + "loss": 0.9445, + "step": 10213 + }, + { + "epoch": 0.9211345087252559, + "grad_norm": 1.5039431084413881, + "learning_rate": 6.48248654611887e-08, + "loss": 1.0094, + "step": 10214 + }, + { + "epoch": 0.9212246922487262, + "grad_norm": 1.4008253797993315, + "learning_rate": 6.467741396022419e-08, + "loss": 1.0223, + "step": 10215 + }, + { + "epoch": 0.9213148757721964, + "grad_norm": 2.1251158963182277, + "learning_rate": 6.453012759165455e-08, + "loss": 0.9835, + "step": 10216 + }, + { + "epoch": 0.9214050592956666, + "grad_norm": 1.692758702237881, + "learning_rate": 6.438300636804639e-08, + "loss": 0.9424, + "step": 10217 + }, + { + "epoch": 0.921495242819137, + "grad_norm": 1.5348888678813903, + "learning_rate": 6.423605030195278e-08, + "loss": 1.0088, + "step": 10218 + }, + { + "epoch": 0.9215854263426072, + "grad_norm": 1.8575617179012862, + "learning_rate": 6.408925940591304e-08, + "loss": 1.0402, + "step": 10219 + }, + { + "epoch": 0.9216756098660774, + "grad_norm": 1.4858928749932852, + "learning_rate": 6.394263369245222e-08, + "loss": 0.9075, + "step": 10220 + }, + { + "epoch": 0.9217657933895478, + "grad_norm": 1.8486164976952757, + "learning_rate": 6.379617317408126e-08, + "loss": 0.9799, + "step": 10221 + }, + { + "epoch": 0.921855976913018, + "grad_norm": 1.829328092202407, + "learning_rate": 6.364987786329723e-08, + "loss": 0.9708, + "step": 10222 + }, + { + "epoch": 0.9219461604364882, + "grad_norm": 1.954151646679035, + "learning_rate": 6.350374777258193e-08, + "loss": 0.9668, + "step": 10223 + }, + { + "epoch": 0.9220363439599585, + "grad_norm": 1.6307945375631252, + "learning_rate": 6.335778291440519e-08, + "loss": 0.9553, + "step": 10224 + }, + { + "epoch": 0.9221265274834288, + "grad_norm": 1.4673005114369166, + "learning_rate": 6.321198330122057e-08, + "loss": 0.8821, + "step": 10225 + }, + { + "epoch": 0.9222167110068991, + "grad_norm": 1.6026735499250602, + "learning_rate": 6.306634894546902e-08, + "loss": 0.9113, + "step": 10226 + }, + { + "epoch": 0.9223068945303693, + "grad_norm": 1.752238885035226, + "learning_rate": 6.292087985957661e-08, + "loss": 0.9181, + "step": 10227 + }, + { + "epoch": 0.9223970780538395, + "grad_norm": 2.121206263022157, + "learning_rate": 6.277557605595585e-08, + "loss": 1.0137, + "step": 10228 + }, + { + "epoch": 0.9224872615773099, + "grad_norm": 1.4592304558411582, + "learning_rate": 6.263043754700481e-08, + "loss": 0.8996, + "step": 10229 + }, + { + "epoch": 0.9225774451007801, + "grad_norm": 1.7345233200850028, + "learning_rate": 6.248546434510671e-08, + "loss": 0.9742, + "step": 10230 + }, + { + "epoch": 0.9226676286242503, + "grad_norm": 2.2021432407089456, + "learning_rate": 6.234065646263298e-08, + "loss": 0.9064, + "step": 10231 + }, + { + "epoch": 0.9227578121477206, + "grad_norm": 1.7434397076136192, + "learning_rate": 6.219601391193796e-08, + "loss": 0.883, + "step": 10232 + }, + { + "epoch": 0.9228479956711909, + "grad_norm": 1.5462063201433676, + "learning_rate": 6.205153670536423e-08, + "loss": 0.9035, + "step": 10233 + }, + { + "epoch": 0.9229381791946611, + "grad_norm": 1.4448344496828542, + "learning_rate": 6.190722485523902e-08, + "loss": 1.0082, + "step": 10234 + }, + { + "epoch": 0.9230283627181314, + "grad_norm": 1.7174232911105518, + "learning_rate": 6.176307837387607e-08, + "loss": 0.9556, + "step": 10235 + }, + { + "epoch": 0.9231185462416016, + "grad_norm": 1.435005511495925, + "learning_rate": 6.16190972735744e-08, + "loss": 0.9256, + "step": 10236 + }, + { + "epoch": 0.923208729765072, + "grad_norm": 1.8451574740105774, + "learning_rate": 6.147528156661974e-08, + "loss": 0.9997, + "step": 10237 + }, + { + "epoch": 0.9232989132885422, + "grad_norm": 2.248535103233577, + "learning_rate": 6.133163126528273e-08, + "loss": 0.899, + "step": 10238 + }, + { + "epoch": 0.9233890968120124, + "grad_norm": 1.7402780101370379, + "learning_rate": 6.11881463818209e-08, + "loss": 0.9786, + "step": 10239 + }, + { + "epoch": 0.9234792803354827, + "grad_norm": 1.5416293354497739, + "learning_rate": 6.104482692847668e-08, + "loss": 0.9621, + "step": 10240 + }, + { + "epoch": 0.923569463858953, + "grad_norm": 1.4950450279318759, + "learning_rate": 6.090167291747917e-08, + "loss": 0.9434, + "step": 10241 + }, + { + "epoch": 0.9236596473824232, + "grad_norm": 1.6454150799167016, + "learning_rate": 6.075868436104303e-08, + "loss": 0.8724, + "step": 10242 + }, + { + "epoch": 0.9237498309058935, + "grad_norm": 2.7449820426534206, + "learning_rate": 6.061586127136875e-08, + "loss": 0.9662, + "step": 10243 + }, + { + "epoch": 0.9238400144293638, + "grad_norm": 1.521753172335789, + "learning_rate": 6.047320366064324e-08, + "loss": 0.9441, + "step": 10244 + }, + { + "epoch": 0.923930197952834, + "grad_norm": 1.5249426033600348, + "learning_rate": 6.033071154103786e-08, + "loss": 0.9826, + "step": 10245 + }, + { + "epoch": 0.9240203814763043, + "grad_norm": 1.69341590373405, + "learning_rate": 6.018838492471178e-08, + "loss": 0.9953, + "step": 10246 + }, + { + "epoch": 0.9241105649997745, + "grad_norm": 0.6234388404009567, + "learning_rate": 6.00462238238082e-08, + "loss": 0.8161, + "step": 10247 + }, + { + "epoch": 0.9242007485232449, + "grad_norm": 1.9481998448562834, + "learning_rate": 5.990422825045827e-08, + "loss": 0.9384, + "step": 10248 + }, + { + "epoch": 0.9242909320467151, + "grad_norm": 3.70901394431012, + "learning_rate": 5.976239821677675e-08, + "loss": 0.9528, + "step": 10249 + }, + { + "epoch": 0.9243811155701853, + "grad_norm": 1.6362770959579156, + "learning_rate": 5.962073373486598e-08, + "loss": 0.9184, + "step": 10250 + }, + { + "epoch": 0.9244712990936556, + "grad_norm": 1.4808154341875877, + "learning_rate": 5.947923481681316e-08, + "loss": 0.9365, + "step": 10251 + }, + { + "epoch": 0.9245614826171259, + "grad_norm": 1.4673984883696158, + "learning_rate": 5.933790147469198e-08, + "loss": 0.9462, + "step": 10252 + }, + { + "epoch": 0.9246516661405961, + "grad_norm": 2.0154805686125665, + "learning_rate": 5.9196733720561665e-08, + "loss": 0.9253, + "step": 10253 + }, + { + "epoch": 0.9247418496640664, + "grad_norm": 2.2330529630045715, + "learning_rate": 5.905573156646793e-08, + "loss": 1.0597, + "step": 10254 + }, + { + "epoch": 0.9248320331875366, + "grad_norm": 2.5011288955096105, + "learning_rate": 5.8914895024441134e-08, + "loss": 1.0489, + "step": 10255 + }, + { + "epoch": 0.9249222167110069, + "grad_norm": 1.7819452100266342, + "learning_rate": 5.877422410649857e-08, + "loss": 1.0102, + "step": 10256 + }, + { + "epoch": 0.9250124002344772, + "grad_norm": 1.7387078951050785, + "learning_rate": 5.863371882464285e-08, + "loss": 0.9881, + "step": 10257 + }, + { + "epoch": 0.9251025837579474, + "grad_norm": 1.7849960422204856, + "learning_rate": 5.849337919086283e-08, + "loss": 0.9564, + "step": 10258 + }, + { + "epoch": 0.9251927672814176, + "grad_norm": 2.333392131155421, + "learning_rate": 5.835320521713316e-08, + "loss": 0.9524, + "step": 10259 + }, + { + "epoch": 0.925282950804888, + "grad_norm": 1.474100148777792, + "learning_rate": 5.8213196915414264e-08, + "loss": 0.9593, + "step": 10260 + }, + { + "epoch": 0.9253731343283582, + "grad_norm": 2.3118505854785765, + "learning_rate": 5.807335429765237e-08, + "loss": 1.0302, + "step": 10261 + }, + { + "epoch": 0.9254633178518284, + "grad_norm": 2.0789544307454086, + "learning_rate": 5.7933677375779034e-08, + "loss": 1.0255, + "step": 10262 + }, + { + "epoch": 0.9255535013752987, + "grad_norm": 1.5230675663879252, + "learning_rate": 5.77941661617134e-08, + "loss": 1.0245, + "step": 10263 + }, + { + "epoch": 0.925643684898769, + "grad_norm": 1.8426194038158628, + "learning_rate": 5.765482066735816e-08, + "loss": 0.9098, + "step": 10264 + }, + { + "epoch": 0.9257338684222393, + "grad_norm": 1.599388396075467, + "learning_rate": 5.7515640904604256e-08, + "loss": 0.9745, + "step": 10265 + }, + { + "epoch": 0.9258240519457095, + "grad_norm": 2.0480041883619515, + "learning_rate": 5.7376626885326187e-08, + "loss": 0.9569, + "step": 10266 + }, + { + "epoch": 0.9259142354691798, + "grad_norm": 1.5719049833862204, + "learning_rate": 5.723777862138601e-08, + "loss": 1.0493, + "step": 10267 + }, + { + "epoch": 0.9260044189926501, + "grad_norm": 2.320983378721036, + "learning_rate": 5.7099096124630705e-08, + "loss": 1.0535, + "step": 10268 + }, + { + "epoch": 0.9260946025161203, + "grad_norm": 1.2488291502069269, + "learning_rate": 5.696057940689347e-08, + "loss": 0.9847, + "step": 10269 + }, + { + "epoch": 0.9261847860395905, + "grad_norm": 2.155431509162873, + "learning_rate": 5.6822228479993736e-08, + "loss": 0.9957, + "step": 10270 + }, + { + "epoch": 0.9262749695630609, + "grad_norm": 1.9369217409622368, + "learning_rate": 5.668404335573584e-08, + "loss": 1.0176, + "step": 10271 + }, + { + "epoch": 0.9263651530865311, + "grad_norm": 1.5626889686756078, + "learning_rate": 5.654602404591058e-08, + "loss": 0.9184, + "step": 10272 + }, + { + "epoch": 0.9264553366100013, + "grad_norm": 1.6704157940976547, + "learning_rate": 5.640817056229474e-08, + "loss": 1.032, + "step": 10273 + }, + { + "epoch": 0.9265455201334716, + "grad_norm": 2.049174756746067, + "learning_rate": 5.6270482916650706e-08, + "loss": 1.0132, + "step": 10274 + }, + { + "epoch": 0.9266357036569419, + "grad_norm": 2.2172198324823316, + "learning_rate": 5.613296112072663e-08, + "loss": 1.0501, + "step": 10275 + }, + { + "epoch": 0.9267258871804122, + "grad_norm": 2.1501209668572465, + "learning_rate": 5.59956051862569e-08, + "loss": 0.9794, + "step": 10276 + }, + { + "epoch": 0.9268160707038824, + "grad_norm": 1.5201374011930169, + "learning_rate": 5.585841512496081e-08, + "loss": 1.0395, + "step": 10277 + }, + { + "epoch": 0.9269062542273526, + "grad_norm": 1.7678888711655536, + "learning_rate": 5.5721390948545e-08, + "loss": 1.0864, + "step": 10278 + }, + { + "epoch": 0.926996437750823, + "grad_norm": 1.4636797947258298, + "learning_rate": 5.558453266870056e-08, + "loss": 1.016, + "step": 10279 + }, + { + "epoch": 0.9270866212742932, + "grad_norm": 1.883549336337382, + "learning_rate": 5.544784029710525e-08, + "loss": 0.9153, + "step": 10280 + }, + { + "epoch": 0.9271768047977634, + "grad_norm": 1.9932287748635609, + "learning_rate": 5.531131384542242e-08, + "loss": 1.0485, + "step": 10281 + }, + { + "epoch": 0.9272669883212337, + "grad_norm": 2.060554743024078, + "learning_rate": 5.51749533253012e-08, + "loss": 0.928, + "step": 10282 + }, + { + "epoch": 0.927357171844704, + "grad_norm": 1.3758942990318912, + "learning_rate": 5.503875874837649e-08, + "loss": 0.9157, + "step": 10283 + }, + { + "epoch": 0.9274473553681742, + "grad_norm": 4.003364102014802, + "learning_rate": 5.4902730126269225e-08, + "loss": 0.9443, + "step": 10284 + }, + { + "epoch": 0.9275375388916445, + "grad_norm": 2.494724238260289, + "learning_rate": 5.476686747058656e-08, + "loss": 1.0095, + "step": 10285 + }, + { + "epoch": 0.9276277224151147, + "grad_norm": 3.2130302793071572, + "learning_rate": 5.4631170792920124e-08, + "loss": 0.9533, + "step": 10286 + }, + { + "epoch": 0.927717905938585, + "grad_norm": 1.7383905612136332, + "learning_rate": 5.449564010484953e-08, + "loss": 0.9535, + "step": 10287 + }, + { + "epoch": 0.9278080894620553, + "grad_norm": 2.29460166482811, + "learning_rate": 5.436027541793775e-08, + "loss": 1.001, + "step": 10288 + }, + { + "epoch": 0.9278982729855255, + "grad_norm": 1.4607733159957084, + "learning_rate": 5.4225076743735554e-08, + "loss": 0.9522, + "step": 10289 + }, + { + "epoch": 0.9279884565089958, + "grad_norm": 2.2824933569854298, + "learning_rate": 5.409004409377882e-08, + "loss": 0.8958, + "step": 10290 + }, + { + "epoch": 0.9280786400324661, + "grad_norm": 1.4155368600652436, + "learning_rate": 5.3955177479589e-08, + "loss": 0.9017, + "step": 10291 + }, + { + "epoch": 0.9281688235559363, + "grad_norm": 1.6017194717520216, + "learning_rate": 5.3820476912674e-08, + "loss": 0.983, + "step": 10292 + }, + { + "epoch": 0.9282590070794066, + "grad_norm": 1.7811733697837864, + "learning_rate": 5.3685942404527063e-08, + "loss": 1.001, + "step": 10293 + }, + { + "epoch": 0.9283491906028769, + "grad_norm": 1.6717102825272312, + "learning_rate": 5.355157396662702e-08, + "loss": 0.9589, + "step": 10294 + }, + { + "epoch": 0.9284393741263471, + "grad_norm": 1.3888261282834007, + "learning_rate": 5.34173716104398e-08, + "loss": 0.9605, + "step": 10295 + }, + { + "epoch": 0.9285295576498174, + "grad_norm": 1.6546662153652743, + "learning_rate": 5.328333534741536e-08, + "loss": 0.9858, + "step": 10296 + }, + { + "epoch": 0.9286197411732876, + "grad_norm": 2.203052492504208, + "learning_rate": 5.314946518899099e-08, + "loss": 0.9472, + "step": 10297 + }, + { + "epoch": 0.928709924696758, + "grad_norm": 3.502491472929568, + "learning_rate": 5.301576114658912e-08, + "loss": 1.0491, + "step": 10298 + }, + { + "epoch": 0.9288001082202282, + "grad_norm": 2.3711149664091455, + "learning_rate": 5.288222323161795e-08, + "loss": 1.006, + "step": 10299 + }, + { + "epoch": 0.9288902917436984, + "grad_norm": 1.9306621483970343, + "learning_rate": 5.274885145547214e-08, + "loss": 1.0198, + "step": 10300 + }, + { + "epoch": 0.9289804752671686, + "grad_norm": 1.5179706787474077, + "learning_rate": 5.261564582953082e-08, + "loss": 0.9681, + "step": 10301 + }, + { + "epoch": 0.929070658790639, + "grad_norm": 2.392530193237991, + "learning_rate": 5.248260636516066e-08, + "loss": 0.9531, + "step": 10302 + }, + { + "epoch": 0.9291608423141092, + "grad_norm": 1.8199021691448634, + "learning_rate": 5.2349733073712824e-08, + "loss": 0.9193, + "step": 10303 + }, + { + "epoch": 0.9292510258375795, + "grad_norm": 1.7931453310377985, + "learning_rate": 5.221702596652533e-08, + "loss": 1.0648, + "step": 10304 + }, + { + "epoch": 0.9293412093610497, + "grad_norm": 2.6313946355723417, + "learning_rate": 5.208448505492091e-08, + "loss": 0.9167, + "step": 10305 + }, + { + "epoch": 0.92943139288452, + "grad_norm": 6.811470443902278, + "learning_rate": 5.1952110350208965e-08, + "loss": 0.9616, + "step": 10306 + }, + { + "epoch": 0.9295215764079903, + "grad_norm": 1.6217078826596045, + "learning_rate": 5.181990186368446e-08, + "loss": 0.9319, + "step": 10307 + }, + { + "epoch": 0.9296117599314605, + "grad_norm": 1.9706603575830466, + "learning_rate": 5.1687859606627915e-08, + "loss": 1.0751, + "step": 10308 + }, + { + "epoch": 0.9297019434549307, + "grad_norm": 1.7818565674373332, + "learning_rate": 5.1555983590306327e-08, + "loss": 0.892, + "step": 10309 + }, + { + "epoch": 0.9297921269784011, + "grad_norm": 0.6539508189117696, + "learning_rate": 5.1424273825971806e-08, + "loss": 0.8141, + "step": 10310 + }, + { + "epoch": 0.9298823105018713, + "grad_norm": 1.6074527260101374, + "learning_rate": 5.1292730324862475e-08, + "loss": 0.9411, + "step": 10311 + }, + { + "epoch": 0.9299724940253415, + "grad_norm": 1.6302316519762536, + "learning_rate": 5.116135309820224e-08, + "loss": 0.9908, + "step": 10312 + }, + { + "epoch": 0.9300626775488118, + "grad_norm": 1.808770387398869, + "learning_rate": 5.103014215720147e-08, + "loss": 1.0154, + "step": 10313 + }, + { + "epoch": 0.9301528610722821, + "grad_norm": 1.5820476013092108, + "learning_rate": 5.0899097513055214e-08, + "loss": 1.0329, + "step": 10314 + }, + { + "epoch": 0.9302430445957524, + "grad_norm": 1.6572695059486522, + "learning_rate": 5.076821917694563e-08, + "loss": 0.9645, + "step": 10315 + }, + { + "epoch": 0.9303332281192226, + "grad_norm": 1.7463739157573601, + "learning_rate": 5.063750716003889e-08, + "loss": 1.0251, + "step": 10316 + }, + { + "epoch": 0.9304234116426929, + "grad_norm": 2.9883418856184774, + "learning_rate": 5.050696147348921e-08, + "loss": 0.9592, + "step": 10317 + }, + { + "epoch": 0.9305135951661632, + "grad_norm": 2.213792124023819, + "learning_rate": 5.037658212843454e-08, + "loss": 0.9483, + "step": 10318 + }, + { + "epoch": 0.9306037786896334, + "grad_norm": 1.431412120449361, + "learning_rate": 5.0246369136000444e-08, + "loss": 0.8692, + "step": 10319 + }, + { + "epoch": 0.9306939622131036, + "grad_norm": 1.5342106904042414, + "learning_rate": 5.011632250729691e-08, + "loss": 1.0493, + "step": 10320 + }, + { + "epoch": 0.930784145736574, + "grad_norm": 1.8837617879092452, + "learning_rate": 4.998644225342019e-08, + "loss": 1.0668, + "step": 10321 + }, + { + "epoch": 0.9308743292600442, + "grad_norm": 1.9573079596939826, + "learning_rate": 4.9856728385452296e-08, + "loss": 0.9772, + "step": 10322 + }, + { + "epoch": 0.9309645127835144, + "grad_norm": 1.810704558747565, + "learning_rate": 4.9727180914461485e-08, + "loss": 1.0644, + "step": 10323 + }, + { + "epoch": 0.9310546963069847, + "grad_norm": 2.9296053048365516, + "learning_rate": 4.959779985150137e-08, + "loss": 0.9839, + "step": 10324 + }, + { + "epoch": 0.931144879830455, + "grad_norm": 1.8295033141484338, + "learning_rate": 4.9468585207611105e-08, + "loss": 0.9195, + "step": 10325 + }, + { + "epoch": 0.9312350633539253, + "grad_norm": 1.6356097631797928, + "learning_rate": 4.9339536993816764e-08, + "loss": 0.955, + "step": 10326 + }, + { + "epoch": 0.9313252468773955, + "grad_norm": 1.5617625976025107, + "learning_rate": 4.921065522112844e-08, + "loss": 0.9236, + "step": 10327 + }, + { + "epoch": 0.9314154304008657, + "grad_norm": 1.7375558775554711, + "learning_rate": 4.908193990054377e-08, + "loss": 1.0254, + "step": 10328 + }, + { + "epoch": 0.9315056139243361, + "grad_norm": 1.4942962764871022, + "learning_rate": 4.89533910430453e-08, + "loss": 0.9915, + "step": 10329 + }, + { + "epoch": 0.9315957974478063, + "grad_norm": 1.5747769788243915, + "learning_rate": 4.8825008659601376e-08, + "loss": 0.9645, + "step": 10330 + }, + { + "epoch": 0.9316859809712765, + "grad_norm": 2.684927218994058, + "learning_rate": 4.869679276116634e-08, + "loss": 0.987, + "step": 10331 + }, + { + "epoch": 0.9317761644947468, + "grad_norm": 1.6088427061506527, + "learning_rate": 4.856874335868055e-08, + "loss": 0.9929, + "step": 10332 + }, + { + "epoch": 0.9318663480182171, + "grad_norm": 1.7568562937432401, + "learning_rate": 4.844086046306928e-08, + "loss": 1.0829, + "step": 10333 + }, + { + "epoch": 0.9319565315416873, + "grad_norm": 1.685838941463257, + "learning_rate": 4.8313144085244896e-08, + "loss": 0.973, + "step": 10334 + }, + { + "epoch": 0.9320467150651576, + "grad_norm": 1.4250444572607823, + "learning_rate": 4.818559423610424e-08, + "loss": 0.8116, + "step": 10335 + }, + { + "epoch": 0.9321368985886278, + "grad_norm": 1.7765414596559763, + "learning_rate": 4.8058210926531284e-08, + "loss": 0.9205, + "step": 10336 + }, + { + "epoch": 0.9322270821120981, + "grad_norm": 1.7371387627605357, + "learning_rate": 4.7930994167394435e-08, + "loss": 1.0093, + "step": 10337 + }, + { + "epoch": 0.9323172656355684, + "grad_norm": 1.288636801971733, + "learning_rate": 4.7803943969548786e-08, + "loss": 1.0212, + "step": 10338 + }, + { + "epoch": 0.9324074491590386, + "grad_norm": 2.5887015054784808, + "learning_rate": 4.7677060343834784e-08, + "loss": 0.9206, + "step": 10339 + }, + { + "epoch": 0.932497632682509, + "grad_norm": 1.8018397201487464, + "learning_rate": 4.75503433010791e-08, + "loss": 0.9704, + "step": 10340 + }, + { + "epoch": 0.9325878162059792, + "grad_norm": 2.193904873415758, + "learning_rate": 4.742379285209419e-08, + "loss": 0.9581, + "step": 10341 + }, + { + "epoch": 0.9326779997294494, + "grad_norm": 2.5628297174975696, + "learning_rate": 4.72974090076772e-08, + "loss": 1.0773, + "step": 10342 + }, + { + "epoch": 0.9327681832529197, + "grad_norm": 1.398199220954461, + "learning_rate": 4.717119177861262e-08, + "loss": 0.9323, + "step": 10343 + }, + { + "epoch": 0.93285836677639, + "grad_norm": 1.7457709302044957, + "learning_rate": 4.70451411756696e-08, + "loss": 0.9898, + "step": 10344 + }, + { + "epoch": 0.9329485502998602, + "grad_norm": 1.396197565677096, + "learning_rate": 4.691925720960355e-08, + "loss": 1.0164, + "step": 10345 + }, + { + "epoch": 0.9330387338233305, + "grad_norm": 1.7854091206424145, + "learning_rate": 4.6793539891155645e-08, + "loss": 0.9958, + "step": 10346 + }, + { + "epoch": 0.9331289173468007, + "grad_norm": 1.8634720870522914, + "learning_rate": 4.6667989231052864e-08, + "loss": 0.953, + "step": 10347 + }, + { + "epoch": 0.933219100870271, + "grad_norm": 2.6541524346737084, + "learning_rate": 4.654260524000797e-08, + "loss": 0.9824, + "step": 10348 + }, + { + "epoch": 0.9333092843937413, + "grad_norm": 1.4671170511283091, + "learning_rate": 4.6417387928719076e-08, + "loss": 0.9634, + "step": 10349 + }, + { + "epoch": 0.9333994679172115, + "grad_norm": 1.5259421861878544, + "learning_rate": 4.629233730787052e-08, + "loss": 0.849, + "step": 10350 + }, + { + "epoch": 0.9334896514406817, + "grad_norm": 1.7337181290802968, + "learning_rate": 4.616745338813266e-08, + "loss": 0.9414, + "step": 10351 + }, + { + "epoch": 0.9335798349641521, + "grad_norm": 1.6928316589814902, + "learning_rate": 4.6042736180160744e-08, + "loss": 0.9936, + "step": 10352 + }, + { + "epoch": 0.9336700184876223, + "grad_norm": 5.160187668455639, + "learning_rate": 4.591818569459671e-08, + "loss": 1.0949, + "step": 10353 + }, + { + "epoch": 0.9337602020110926, + "grad_norm": 2.1628768372057485, + "learning_rate": 4.5793801942067614e-08, + "loss": 0.9998, + "step": 10354 + }, + { + "epoch": 0.9338503855345628, + "grad_norm": 1.8126984849311616, + "learning_rate": 4.566958493318673e-08, + "loss": 0.8505, + "step": 10355 + }, + { + "epoch": 0.9339405690580331, + "grad_norm": 0.6192619854610742, + "learning_rate": 4.554553467855316e-08, + "loss": 0.8363, + "step": 10356 + }, + { + "epoch": 0.9340307525815034, + "grad_norm": 1.5755997651139932, + "learning_rate": 4.5421651188751074e-08, + "loss": 0.9948, + "step": 10357 + }, + { + "epoch": 0.9341209361049736, + "grad_norm": 2.4020424258434234, + "learning_rate": 4.529793447435137e-08, + "loss": 0.992, + "step": 10358 + }, + { + "epoch": 0.9342111196284438, + "grad_norm": 5.008805722487647, + "learning_rate": 4.5174384545909824e-08, + "loss": 0.9796, + "step": 10359 + }, + { + "epoch": 0.9343013031519142, + "grad_norm": 1.9612989358528667, + "learning_rate": 4.505100141396867e-08, + "loss": 1.0144, + "step": 10360 + }, + { + "epoch": 0.9343914866753844, + "grad_norm": 1.9399769239921252, + "learning_rate": 4.492778508905548e-08, + "loss": 1.0107, + "step": 10361 + }, + { + "epoch": 0.9344816701988546, + "grad_norm": 1.726537091512702, + "learning_rate": 4.480473558168385e-08, + "loss": 0.9181, + "step": 10362 + }, + { + "epoch": 0.934571853722325, + "grad_norm": 1.7472848947009534, + "learning_rate": 4.4681852902352936e-08, + "loss": 0.8819, + "step": 10363 + }, + { + "epoch": 0.9346620372457952, + "grad_norm": 14.228046304245083, + "learning_rate": 4.455913706154812e-08, + "loss": 0.8883, + "step": 10364 + }, + { + "epoch": 0.9347522207692655, + "grad_norm": 1.7549209887768067, + "learning_rate": 4.443658806973949e-08, + "loss": 1.0043, + "step": 10365 + }, + { + "epoch": 0.9348424042927357, + "grad_norm": 0.6724549607081388, + "learning_rate": 4.431420593738444e-08, + "loss": 0.8377, + "step": 10366 + }, + { + "epoch": 0.934932587816206, + "grad_norm": 1.7529328157614845, + "learning_rate": 4.419199067492485e-08, + "loss": 0.891, + "step": 10367 + }, + { + "epoch": 0.9350227713396763, + "grad_norm": 1.8877273927854774, + "learning_rate": 4.4069942292788596e-08, + "loss": 0.9887, + "step": 10368 + }, + { + "epoch": 0.9351129548631465, + "grad_norm": 1.7277276435943436, + "learning_rate": 4.39480608013898e-08, + "loss": 0.9837, + "step": 10369 + }, + { + "epoch": 0.9352031383866167, + "grad_norm": 1.8065862534048052, + "learning_rate": 4.3826346211128126e-08, + "loss": 1.0676, + "step": 10370 + }, + { + "epoch": 0.9352933219100871, + "grad_norm": 1.4977917310937012, + "learning_rate": 4.370479853238884e-08, + "loss": 0.9808, + "step": 10371 + }, + { + "epoch": 0.9353835054335573, + "grad_norm": 1.7220077851149396, + "learning_rate": 4.3583417775542756e-08, + "loss": 0.9388, + "step": 10372 + }, + { + "epoch": 0.9354736889570275, + "grad_norm": 1.8919139913240313, + "learning_rate": 4.3462203950947575e-08, + "loss": 0.9303, + "step": 10373 + }, + { + "epoch": 0.9355638724804978, + "grad_norm": 2.0965466397149384, + "learning_rate": 4.3341157068944814e-08, + "loss": 0.9626, + "step": 10374 + }, + { + "epoch": 0.9356540560039681, + "grad_norm": 1.8376316127223482, + "learning_rate": 4.322027713986376e-08, + "loss": 0.9933, + "step": 10375 + }, + { + "epoch": 0.9357442395274383, + "grad_norm": 2.5825030520514356, + "learning_rate": 4.309956417401816e-08, + "loss": 1.0441, + "step": 10376 + }, + { + "epoch": 0.9358344230509086, + "grad_norm": 1.4134941012485784, + "learning_rate": 4.297901818170801e-08, + "loss": 0.9504, + "step": 10377 + }, + { + "epoch": 0.9359246065743788, + "grad_norm": 1.5159005769924925, + "learning_rate": 4.285863917321886e-08, + "loss": 0.9097, + "step": 10378 + }, + { + "epoch": 0.9360147900978492, + "grad_norm": 1.6226296376308331, + "learning_rate": 4.2738427158822253e-08, + "loss": 0.9135, + "step": 10379 + }, + { + "epoch": 0.9361049736213194, + "grad_norm": 2.296232483973015, + "learning_rate": 4.261838214877511e-08, + "loss": 0.9828, + "step": 10380 + }, + { + "epoch": 0.9361951571447896, + "grad_norm": 1.4671622277176462, + "learning_rate": 4.249850415332079e-08, + "loss": 0.9864, + "step": 10381 + }, + { + "epoch": 0.9362853406682599, + "grad_norm": 1.5223813039177907, + "learning_rate": 4.237879318268756e-08, + "loss": 1.0347, + "step": 10382 + }, + { + "epoch": 0.9363755241917302, + "grad_norm": 1.5370340680260093, + "learning_rate": 4.225924924708968e-08, + "loss": 0.9108, + "step": 10383 + }, + { + "epoch": 0.9364657077152004, + "grad_norm": 1.7890371129262836, + "learning_rate": 4.2139872356727665e-08, + "loss": 0.9863, + "step": 10384 + }, + { + "epoch": 0.9365558912386707, + "grad_norm": 1.6723964270209826, + "learning_rate": 4.202066252178738e-08, + "loss": 0.9417, + "step": 10385 + }, + { + "epoch": 0.936646074762141, + "grad_norm": 1.5128710394117766, + "learning_rate": 4.1901619752440445e-08, + "loss": 1.0681, + "step": 10386 + }, + { + "epoch": 0.9367362582856112, + "grad_norm": 1.6771399052084046, + "learning_rate": 4.178274405884363e-08, + "loss": 0.9758, + "step": 10387 + }, + { + "epoch": 0.9368264418090815, + "grad_norm": 1.554992109198957, + "learning_rate": 4.166403545114105e-08, + "loss": 0.9892, + "step": 10388 + }, + { + "epoch": 0.9369166253325517, + "grad_norm": 1.7910634141576158, + "learning_rate": 4.154549393946083e-08, + "loss": 1.0319, + "step": 10389 + }, + { + "epoch": 0.937006808856022, + "grad_norm": 1.4903455620280124, + "learning_rate": 4.14271195339182e-08, + "loss": 1.0061, + "step": 10390 + }, + { + "epoch": 0.9370969923794923, + "grad_norm": 2.2453236196689903, + "learning_rate": 4.1308912244613084e-08, + "loss": 0.9421, + "step": 10391 + }, + { + "epoch": 0.9371871759029625, + "grad_norm": 1.4268462919337366, + "learning_rate": 4.1190872081631636e-08, + "loss": 0.9325, + "step": 10392 + }, + { + "epoch": 0.9372773594264328, + "grad_norm": 1.9251709874495195, + "learning_rate": 4.107299905504558e-08, + "loss": 1.0409, + "step": 10393 + }, + { + "epoch": 0.9373675429499031, + "grad_norm": 1.6000098168548786, + "learning_rate": 4.095529317491286e-08, + "loss": 0.9707, + "step": 10394 + }, + { + "epoch": 0.9374577264733733, + "grad_norm": 1.9233000519493808, + "learning_rate": 4.0837754451276575e-08, + "loss": 0.9844, + "step": 10395 + }, + { + "epoch": 0.9375479099968436, + "grad_norm": 1.4552760456836444, + "learning_rate": 4.072038289416557e-08, + "loss": 0.9643, + "step": 10396 + }, + { + "epoch": 0.9376380935203138, + "grad_norm": 1.5297987152287718, + "learning_rate": 4.0603178513595185e-08, + "loss": 0.9815, + "step": 10397 + }, + { + "epoch": 0.9377282770437841, + "grad_norm": 1.5988890128584452, + "learning_rate": 4.0486141319565624e-08, + "loss": 0.9843, + "step": 10398 + }, + { + "epoch": 0.9378184605672544, + "grad_norm": 1.701747792421436, + "learning_rate": 4.0369271322062916e-08, + "loss": 0.9214, + "step": 10399 + }, + { + "epoch": 0.9379086440907246, + "grad_norm": 1.4734832303504175, + "learning_rate": 4.0252568531059295e-08, + "loss": 0.997, + "step": 10400 + }, + { + "epoch": 0.9379988276141948, + "grad_norm": 3.7235780037108444, + "learning_rate": 4.013603295651235e-08, + "loss": 0.9808, + "step": 10401 + }, + { + "epoch": 0.9380890111376652, + "grad_norm": 1.551533322093551, + "learning_rate": 4.001966460836592e-08, + "loss": 1.0042, + "step": 10402 + }, + { + "epoch": 0.9381791946611354, + "grad_norm": 1.6895622085900113, + "learning_rate": 3.990346349654894e-08, + "loss": 1.0509, + "step": 10403 + }, + { + "epoch": 0.9382693781846057, + "grad_norm": 1.4134404621519145, + "learning_rate": 3.9787429630975924e-08, + "loss": 0.9538, + "step": 10404 + }, + { + "epoch": 0.9383595617080759, + "grad_norm": 3.3664014567042053, + "learning_rate": 3.967156302154828e-08, + "loss": 0.9672, + "step": 10405 + }, + { + "epoch": 0.9384497452315462, + "grad_norm": 1.7109618163449343, + "learning_rate": 3.955586367815189e-08, + "loss": 0.9335, + "step": 10406 + }, + { + "epoch": 0.9385399287550165, + "grad_norm": 1.8239187120469231, + "learning_rate": 3.944033161065907e-08, + "loss": 1.0478, + "step": 10407 + }, + { + "epoch": 0.9386301122784867, + "grad_norm": 1.5902367558340842, + "learning_rate": 3.93249668289275e-08, + "loss": 0.8633, + "step": 10408 + }, + { + "epoch": 0.9387202958019569, + "grad_norm": 1.5560964209829267, + "learning_rate": 3.920976934280063e-08, + "loss": 0.905, + "step": 10409 + }, + { + "epoch": 0.9388104793254273, + "grad_norm": 1.6402913072103278, + "learning_rate": 3.909473916210815e-08, + "loss": 1.0635, + "step": 10410 + }, + { + "epoch": 0.9389006628488975, + "grad_norm": 1.9886724840326975, + "learning_rate": 3.897987629666488e-08, + "loss": 0.9371, + "step": 10411 + }, + { + "epoch": 0.9389908463723677, + "grad_norm": 1.3357821814836892, + "learning_rate": 3.886518075627143e-08, + "loss": 1.0513, + "step": 10412 + }, + { + "epoch": 0.9390810298958381, + "grad_norm": 2.1690366572038817, + "learning_rate": 3.875065255071419e-08, + "loss": 0.9766, + "step": 10413 + }, + { + "epoch": 0.9391712134193083, + "grad_norm": 1.7519322354741582, + "learning_rate": 3.863629168976579e-08, + "loss": 0.9482, + "step": 10414 + }, + { + "epoch": 0.9392613969427785, + "grad_norm": 1.6248700750202802, + "learning_rate": 3.852209818318375e-08, + "loss": 1.0367, + "step": 10415 + }, + { + "epoch": 0.9393515804662488, + "grad_norm": 1.83837436710927, + "learning_rate": 3.840807204071161e-08, + "loss": 0.8825, + "step": 10416 + }, + { + "epoch": 0.9394417639897191, + "grad_norm": 1.6114948718172548, + "learning_rate": 3.829421327207894e-08, + "loss": 0.9634, + "step": 10417 + }, + { + "epoch": 0.9395319475131894, + "grad_norm": 1.792621254750199, + "learning_rate": 3.8180521887000825e-08, + "loss": 0.9342, + "step": 10418 + }, + { + "epoch": 0.9396221310366596, + "grad_norm": 0.5878150359063049, + "learning_rate": 3.806699789517775e-08, + "loss": 0.8472, + "step": 10419 + }, + { + "epoch": 0.9397123145601298, + "grad_norm": 1.5082196096187792, + "learning_rate": 3.7953641306296635e-08, + "loss": 0.9606, + "step": 10420 + }, + { + "epoch": 0.9398024980836002, + "grad_norm": 1.9266788023706163, + "learning_rate": 3.784045213002951e-08, + "loss": 0.8243, + "step": 10421 + }, + { + "epoch": 0.9398926816070704, + "grad_norm": 2.0840551905947855, + "learning_rate": 3.7727430376033986e-08, + "loss": 0.9947, + "step": 10422 + }, + { + "epoch": 0.9399828651305406, + "grad_norm": 1.8707252728944777, + "learning_rate": 3.7614576053954126e-08, + "loss": 0.9835, + "step": 10423 + }, + { + "epoch": 0.9400730486540109, + "grad_norm": 1.9737874328658493, + "learning_rate": 3.75018891734189e-08, + "loss": 0.9746, + "step": 10424 + }, + { + "epoch": 0.9401632321774812, + "grad_norm": 1.8541637449205788, + "learning_rate": 3.738936974404372e-08, + "loss": 1.0257, + "step": 10425 + }, + { + "epoch": 0.9402534157009514, + "grad_norm": 1.3242130335681603, + "learning_rate": 3.7277017775429354e-08, + "loss": 1.044, + "step": 10426 + }, + { + "epoch": 0.9403435992244217, + "grad_norm": 1.7207348546751746, + "learning_rate": 3.7164833277162136e-08, + "loss": 1.0015, + "step": 10427 + }, + { + "epoch": 0.9404337827478919, + "grad_norm": 1.2411263209687564, + "learning_rate": 3.705281625881418e-08, + "loss": 0.9573, + "step": 10428 + }, + { + "epoch": 0.9405239662713623, + "grad_norm": 1.4947057914212833, + "learning_rate": 3.694096672994362e-08, + "loss": 0.8837, + "step": 10429 + }, + { + "epoch": 0.9406141497948325, + "grad_norm": 1.5492686238150546, + "learning_rate": 3.682928470009394e-08, + "loss": 1.0318, + "step": 10430 + }, + { + "epoch": 0.9407043333183027, + "grad_norm": 2.048931573778065, + "learning_rate": 3.6717770178794406e-08, + "loss": 0.936, + "step": 10431 + }, + { + "epoch": 0.940794516841773, + "grad_norm": 1.5055166366554904, + "learning_rate": 3.6606423175560287e-08, + "loss": 1.0849, + "step": 10432 + }, + { + "epoch": 0.9408847003652433, + "grad_norm": 0.6274058766083448, + "learning_rate": 3.649524369989221e-08, + "loss": 0.7822, + "step": 10433 + }, + { + "epoch": 0.9409748838887135, + "grad_norm": 1.6007770380027095, + "learning_rate": 3.638423176127636e-08, + "loss": 1.008, + "step": 10434 + }, + { + "epoch": 0.9410650674121838, + "grad_norm": 1.6671138958114609, + "learning_rate": 3.6273387369185396e-08, + "loss": 1.0143, + "step": 10435 + }, + { + "epoch": 0.9411552509356541, + "grad_norm": 2.2665176211702383, + "learning_rate": 3.616271053307685e-08, + "loss": 0.9223, + "step": 10436 + }, + { + "epoch": 0.9412454344591243, + "grad_norm": 2.1063661945112107, + "learning_rate": 3.6052201262394275e-08, + "loss": 0.9991, + "step": 10437 + }, + { + "epoch": 0.9413356179825946, + "grad_norm": 1.8514907114783352, + "learning_rate": 3.5941859566566816e-08, + "loss": 1.0351, + "step": 10438 + }, + { + "epoch": 0.9414258015060648, + "grad_norm": 0.6797160000142368, + "learning_rate": 3.583168545500981e-08, + "loss": 0.8039, + "step": 10439 + }, + { + "epoch": 0.9415159850295352, + "grad_norm": 1.921759253016984, + "learning_rate": 3.5721678937123746e-08, + "loss": 0.9227, + "step": 10440 + }, + { + "epoch": 0.9416061685530054, + "grad_norm": 1.5884297899550288, + "learning_rate": 3.561184002229467e-08, + "loss": 1.0609, + "step": 10441 + }, + { + "epoch": 0.9416963520764756, + "grad_norm": 1.497307745512501, + "learning_rate": 3.550216871989531e-08, + "loss": 0.9255, + "step": 10442 + }, + { + "epoch": 0.9417865355999459, + "grad_norm": 0.6406956377860641, + "learning_rate": 3.539266503928262e-08, + "loss": 0.8831, + "step": 10443 + }, + { + "epoch": 0.9418767191234162, + "grad_norm": 1.5987134201243414, + "learning_rate": 3.528332898980091e-08, + "loss": 0.9911, + "step": 10444 + }, + { + "epoch": 0.9419669026468864, + "grad_norm": 1.5580900903395343, + "learning_rate": 3.517416058077849e-08, + "loss": 0.9468, + "step": 10445 + }, + { + "epoch": 0.9420570861703567, + "grad_norm": 1.6666310147604875, + "learning_rate": 3.506515982153102e-08, + "loss": 0.8837, + "step": 10446 + }, + { + "epoch": 0.9421472696938269, + "grad_norm": 1.717022322259122, + "learning_rate": 3.495632672135862e-08, + "loss": 0.9839, + "step": 10447 + }, + { + "epoch": 0.9422374532172972, + "grad_norm": 1.4865046432684352, + "learning_rate": 3.4847661289547417e-08, + "loss": 1.0105, + "step": 10448 + }, + { + "epoch": 0.9423276367407675, + "grad_norm": 1.8174777244451636, + "learning_rate": 3.473916353536932e-08, + "loss": 0.9347, + "step": 10449 + }, + { + "epoch": 0.9424178202642377, + "grad_norm": 1.6291554944109794, + "learning_rate": 3.463083346808249e-08, + "loss": 0.9485, + "step": 10450 + }, + { + "epoch": 0.9425080037877079, + "grad_norm": 2.810238840534881, + "learning_rate": 3.452267109692975e-08, + "loss": 0.8744, + "step": 10451 + }, + { + "epoch": 0.9425981873111783, + "grad_norm": 4.616708695166437, + "learning_rate": 3.441467643114016e-08, + "loss": 0.9794, + "step": 10452 + }, + { + "epoch": 0.9426883708346485, + "grad_norm": 0.7230519216764256, + "learning_rate": 3.430684947992857e-08, + "loss": 0.7942, + "step": 10453 + }, + { + "epoch": 0.9427785543581187, + "grad_norm": 2.2186314725781977, + "learning_rate": 3.419919025249518e-08, + "loss": 0.9951, + "step": 10454 + }, + { + "epoch": 0.942868737881589, + "grad_norm": 5.071348581124592, + "learning_rate": 3.40916987580262e-08, + "loss": 0.9998, + "step": 10455 + }, + { + "epoch": 0.9429589214050593, + "grad_norm": 1.8997299755359842, + "learning_rate": 3.398437500569362e-08, + "loss": 0.9553, + "step": 10456 + }, + { + "epoch": 0.9430491049285296, + "grad_norm": 1.494001873280731, + "learning_rate": 3.3877219004654347e-08, + "loss": 0.8592, + "step": 10457 + }, + { + "epoch": 0.9431392884519998, + "grad_norm": 1.672834388479962, + "learning_rate": 3.3770230764051946e-08, + "loss": 1.0121, + "step": 10458 + }, + { + "epoch": 0.9432294719754701, + "grad_norm": 2.2547102474299714, + "learning_rate": 3.366341029301534e-08, + "loss": 0.9987, + "step": 10459 + }, + { + "epoch": 0.9433196554989404, + "grad_norm": 1.81598748743057, + "learning_rate": 3.355675760065857e-08, + "loss": 0.9857, + "step": 10460 + }, + { + "epoch": 0.9434098390224106, + "grad_norm": 0.5786763989753374, + "learning_rate": 3.345027269608236e-08, + "loss": 0.7621, + "step": 10461 + }, + { + "epoch": 0.9435000225458808, + "grad_norm": 0.6411325258619146, + "learning_rate": 3.334395558837211e-08, + "loss": 0.8292, + "step": 10462 + }, + { + "epoch": 0.9435902060693512, + "grad_norm": 1.3389928134158422, + "learning_rate": 3.3237806286599667e-08, + "loss": 0.887, + "step": 10463 + }, + { + "epoch": 0.9436803895928214, + "grad_norm": 1.6595673358033503, + "learning_rate": 3.313182479982224e-08, + "loss": 0.9682, + "step": 10464 + }, + { + "epoch": 0.9437705731162916, + "grad_norm": 1.890556081981712, + "learning_rate": 3.302601113708259e-08, + "loss": 1.0445, + "step": 10465 + }, + { + "epoch": 0.9438607566397619, + "grad_norm": 1.329024560002479, + "learning_rate": 3.292036530740972e-08, + "loss": 1.0265, + "step": 10466 + }, + { + "epoch": 0.9439509401632322, + "grad_norm": 1.957363625814764, + "learning_rate": 3.2814887319817294e-08, + "loss": 0.9715, + "step": 10467 + }, + { + "epoch": 0.9440411236867025, + "grad_norm": 1.7806522638035263, + "learning_rate": 3.270957718330591e-08, + "loss": 1.0202, + "step": 10468 + }, + { + "epoch": 0.9441313072101727, + "grad_norm": 1.3694193470615144, + "learning_rate": 3.260443490686082e-08, + "loss": 1.0244, + "step": 10469 + }, + { + "epoch": 0.9442214907336429, + "grad_norm": 1.7221885963288321, + "learning_rate": 3.249946049945351e-08, + "loss": 0.9916, + "step": 10470 + }, + { + "epoch": 0.9443116742571133, + "grad_norm": 1.8721544926382063, + "learning_rate": 3.239465397004082e-08, + "loss": 0.9933, + "step": 10471 + }, + { + "epoch": 0.9444018577805835, + "grad_norm": 1.4372062382945678, + "learning_rate": 3.229001532756559e-08, + "loss": 0.942, + "step": 10472 + }, + { + "epoch": 0.9444920413040537, + "grad_norm": 5.184044847788908, + "learning_rate": 3.218554458095602e-08, + "loss": 0.9753, + "step": 10473 + }, + { + "epoch": 0.944582224827524, + "grad_norm": 1.806928424078395, + "learning_rate": 3.20812417391263e-08, + "loss": 0.9821, + "step": 10474 + }, + { + "epoch": 0.9446724083509943, + "grad_norm": 1.3223441930330877, + "learning_rate": 3.1977106810975764e-08, + "loss": 0.9696, + "step": 10475 + }, + { + "epoch": 0.9447625918744645, + "grad_norm": 0.7763996237450759, + "learning_rate": 3.187313980539042e-08, + "loss": 0.8587, + "step": 10476 + }, + { + "epoch": 0.9448527753979348, + "grad_norm": 1.2999455972438856, + "learning_rate": 3.176934073124071e-08, + "loss": 1.0619, + "step": 10477 + }, + { + "epoch": 0.944942958921405, + "grad_norm": 1.9042869841489931, + "learning_rate": 3.166570959738357e-08, + "loss": 0.9787, + "step": 10478 + }, + { + "epoch": 0.9450331424448754, + "grad_norm": 1.5254758842760152, + "learning_rate": 3.1562246412661476e-08, + "loss": 0.916, + "step": 10479 + }, + { + "epoch": 0.9451233259683456, + "grad_norm": 1.951843146248377, + "learning_rate": 3.145895118590225e-08, + "loss": 1.0133, + "step": 10480 + }, + { + "epoch": 0.9452135094918158, + "grad_norm": 1.5830548024719153, + "learning_rate": 3.135582392591996e-08, + "loss": 0.8885, + "step": 10481 + }, + { + "epoch": 0.9453036930152862, + "grad_norm": 0.5967391713362088, + "learning_rate": 3.125286464151333e-08, + "loss": 0.773, + "step": 10482 + }, + { + "epoch": 0.9453938765387564, + "grad_norm": 2.976970013934838, + "learning_rate": 3.115007334146824e-08, + "loss": 0.953, + "step": 10483 + }, + { + "epoch": 0.9454840600622266, + "grad_norm": 0.6334093834574934, + "learning_rate": 3.104745003455478e-08, + "loss": 0.794, + "step": 10484 + }, + { + "epoch": 0.9455742435856969, + "grad_norm": 1.752246844826164, + "learning_rate": 3.094499472952972e-08, + "loss": 1.0028, + "step": 10485 + }, + { + "epoch": 0.9456644271091672, + "grad_norm": 2.0311022191208945, + "learning_rate": 3.084270743513495e-08, + "loss": 0.9426, + "step": 10486 + }, + { + "epoch": 0.9457546106326374, + "grad_norm": 1.6958562444815366, + "learning_rate": 3.074058816009817e-08, + "loss": 0.8675, + "step": 10487 + }, + { + "epoch": 0.9458447941561077, + "grad_norm": 1.77716613552774, + "learning_rate": 3.063863691313284e-08, + "loss": 0.9278, + "step": 10488 + }, + { + "epoch": 0.9459349776795779, + "grad_norm": 1.5782928377546936, + "learning_rate": 3.0536853702937794e-08, + "loss": 0.9184, + "step": 10489 + }, + { + "epoch": 0.9460251612030482, + "grad_norm": 1.8336886003924517, + "learning_rate": 3.043523853819807e-08, + "loss": 0.9786, + "step": 10490 + }, + { + "epoch": 0.9461153447265185, + "grad_norm": 1.8593398018719098, + "learning_rate": 3.0333791427583855e-08, + "loss": 0.9993, + "step": 10491 + }, + { + "epoch": 0.9462055282499887, + "grad_norm": 2.477859301953176, + "learning_rate": 3.023251237975111e-08, + "loss": 1.1037, + "step": 10492 + }, + { + "epoch": 0.946295711773459, + "grad_norm": 0.6618538422485326, + "learning_rate": 3.0131401403341584e-08, + "loss": 0.8307, + "step": 10493 + }, + { + "epoch": 0.9463858952969293, + "grad_norm": 1.4299512109236274, + "learning_rate": 3.00304585069826e-08, + "loss": 0.9573, + "step": 10494 + }, + { + "epoch": 0.9464760788203995, + "grad_norm": 1.63416799835405, + "learning_rate": 2.992968369928728e-08, + "loss": 0.9681, + "step": 10495 + }, + { + "epoch": 0.9465662623438698, + "grad_norm": 2.143489619421021, + "learning_rate": 2.982907698885429e-08, + "loss": 0.9946, + "step": 10496 + }, + { + "epoch": 0.94665644586734, + "grad_norm": 1.8866539227265302, + "learning_rate": 2.9728638384267645e-08, + "loss": 0.9963, + "step": 10497 + }, + { + "epoch": 0.9467466293908103, + "grad_norm": 2.2077515633523213, + "learning_rate": 2.962836789409784e-08, + "loss": 0.9852, + "step": 10498 + }, + { + "epoch": 0.9468368129142806, + "grad_norm": 1.5932824440069047, + "learning_rate": 2.95282655268998e-08, + "loss": 1.0742, + "step": 10499 + }, + { + "epoch": 0.9469269964377508, + "grad_norm": 1.802308533402593, + "learning_rate": 2.942833129121558e-08, + "loss": 0.9798, + "step": 10500 + }, + { + "epoch": 0.947017179961221, + "grad_norm": 1.8368207389654294, + "learning_rate": 2.9328565195571475e-08, + "loss": 0.9043, + "step": 10501 + }, + { + "epoch": 0.9471073634846914, + "grad_norm": 1.470653923194458, + "learning_rate": 2.9228967248480675e-08, + "loss": 1.046, + "step": 10502 + }, + { + "epoch": 0.9471975470081616, + "grad_norm": 3.4506154326900655, + "learning_rate": 2.912953745844082e-08, + "loss": 1.0236, + "step": 10503 + }, + { + "epoch": 0.9472877305316318, + "grad_norm": 1.691119092567937, + "learning_rate": 2.9030275833936247e-08, + "loss": 0.9664, + "step": 10504 + }, + { + "epoch": 0.9473779140551022, + "grad_norm": 2.1899285051354944, + "learning_rate": 2.893118238343617e-08, + "loss": 0.8804, + "step": 10505 + }, + { + "epoch": 0.9474680975785724, + "grad_norm": 2.0207341225512496, + "learning_rate": 2.8832257115396052e-08, + "loss": 0.8914, + "step": 10506 + }, + { + "epoch": 0.9475582811020427, + "grad_norm": 1.7443611714694154, + "learning_rate": 2.873350003825692e-08, + "loss": 0.9632, + "step": 10507 + }, + { + "epoch": 0.9476484646255129, + "grad_norm": 0.5592701424717403, + "learning_rate": 2.8634911160444696e-08, + "loss": 0.7091, + "step": 10508 + }, + { + "epoch": 0.9477386481489832, + "grad_norm": 1.7654803056420612, + "learning_rate": 2.853649049037199e-08, + "loss": 1.0382, + "step": 10509 + }, + { + "epoch": 0.9478288316724535, + "grad_norm": 1.875995371782332, + "learning_rate": 2.8438238036436525e-08, + "loss": 1.0284, + "step": 10510 + }, + { + "epoch": 0.9479190151959237, + "grad_norm": 1.7821447567911888, + "learning_rate": 2.834015380702137e-08, + "loss": 0.9744, + "step": 10511 + }, + { + "epoch": 0.9480091987193939, + "grad_norm": 0.5781689962193717, + "learning_rate": 2.824223781049606e-08, + "loss": 0.75, + "step": 10512 + }, + { + "epoch": 0.9480993822428643, + "grad_norm": 3.063907689021205, + "learning_rate": 2.8144490055215465e-08, + "loss": 1.043, + "step": 10513 + }, + { + "epoch": 0.9481895657663345, + "grad_norm": 1.4337463297709654, + "learning_rate": 2.8046910549519355e-08, + "loss": 1.014, + "step": 10514 + }, + { + "epoch": 0.9482797492898047, + "grad_norm": 1.8462863222757817, + "learning_rate": 2.794949930173418e-08, + "loss": 0.9852, + "step": 10515 + }, + { + "epoch": 0.948369932813275, + "grad_norm": 1.9317643576278887, + "learning_rate": 2.7852256320171296e-08, + "loss": 1.007, + "step": 10516 + }, + { + "epoch": 0.9484601163367453, + "grad_norm": 1.9880754705337738, + "learning_rate": 2.775518161312851e-08, + "loss": 0.8434, + "step": 10517 + }, + { + "epoch": 0.9485502998602156, + "grad_norm": 1.3674446518069059, + "learning_rate": 2.76582751888883e-08, + "loss": 1.0061, + "step": 10518 + }, + { + "epoch": 0.9486404833836858, + "grad_norm": 1.892392742891977, + "learning_rate": 2.756153705571962e-08, + "loss": 0.9992, + "step": 10519 + }, + { + "epoch": 0.948730666907156, + "grad_norm": 2.0527772327308234, + "learning_rate": 2.74649672218763e-08, + "loss": 0.9624, + "step": 10520 + }, + { + "epoch": 0.9488208504306264, + "grad_norm": 1.620222698728586, + "learning_rate": 2.7368565695598424e-08, + "loss": 0.9671, + "step": 10521 + }, + { + "epoch": 0.9489110339540966, + "grad_norm": 1.5409450732423917, + "learning_rate": 2.727233248511185e-08, + "loss": 0.9063, + "step": 10522 + }, + { + "epoch": 0.9490012174775668, + "grad_norm": 2.139626040182581, + "learning_rate": 2.71762675986269e-08, + "loss": 0.9352, + "step": 10523 + }, + { + "epoch": 0.9490914010010371, + "grad_norm": 1.639778500113416, + "learning_rate": 2.7080371044341242e-08, + "loss": 1.0193, + "step": 10524 + }, + { + "epoch": 0.9491815845245074, + "grad_norm": 2.0895584054247913, + "learning_rate": 2.6984642830436556e-08, + "loss": 1.0116, + "step": 10525 + }, + { + "epoch": 0.9492717680479776, + "grad_norm": 1.5410861735277106, + "learning_rate": 2.688908296508141e-08, + "loss": 0.963, + "step": 10526 + }, + { + "epoch": 0.9493619515714479, + "grad_norm": 1.9043824474964048, + "learning_rate": 2.679369145642929e-08, + "loss": 1.0332, + "step": 10527 + }, + { + "epoch": 0.9494521350949181, + "grad_norm": 1.7209971909486232, + "learning_rate": 2.669846831261946e-08, + "loss": 1.0014, + "step": 10528 + }, + { + "epoch": 0.9495423186183884, + "grad_norm": 1.8231348906187603, + "learning_rate": 2.6603413541776976e-08, + "loss": 0.9699, + "step": 10529 + }, + { + "epoch": 0.9496325021418587, + "grad_norm": 1.2928527595608297, + "learning_rate": 2.6508527152012683e-08, + "loss": 0.9308, + "step": 10530 + }, + { + "epoch": 0.9497226856653289, + "grad_norm": 1.6070597021600543, + "learning_rate": 2.641380915142233e-08, + "loss": 0.8822, + "step": 10531 + }, + { + "epoch": 0.9498128691887993, + "grad_norm": 2.084461733087303, + "learning_rate": 2.6319259548088334e-08, + "loss": 1.0034, + "step": 10532 + }, + { + "epoch": 0.9499030527122695, + "grad_norm": 1.7069430470320788, + "learning_rate": 2.6224878350077585e-08, + "loss": 0.9809, + "step": 10533 + }, + { + "epoch": 0.9499932362357397, + "grad_norm": 1.4559948382495727, + "learning_rate": 2.6130665565443633e-08, + "loss": 0.8942, + "step": 10534 + }, + { + "epoch": 0.95008341975921, + "grad_norm": 1.419472057764882, + "learning_rate": 2.603662120222494e-08, + "loss": 0.9838, + "step": 10535 + }, + { + "epoch": 0.9501736032826803, + "grad_norm": 1.782306224140026, + "learning_rate": 2.59427452684462e-08, + "loss": 0.8274, + "step": 10536 + }, + { + "epoch": 0.9502637868061505, + "grad_norm": 1.4970493700885565, + "learning_rate": 2.5849037772117443e-08, + "loss": 0.958, + "step": 10537 + }, + { + "epoch": 0.9503539703296208, + "grad_norm": 1.6576780964032691, + "learning_rate": 2.575549872123384e-08, + "loss": 1.0043, + "step": 10538 + }, + { + "epoch": 0.950444153853091, + "grad_norm": 1.4082879029043514, + "learning_rate": 2.5662128123776994e-08, + "loss": 0.9856, + "step": 10539 + }, + { + "epoch": 0.9505343373765613, + "grad_norm": 1.5581202349415861, + "learning_rate": 2.5568925987713875e-08, + "loss": 0.9763, + "step": 10540 + }, + { + "epoch": 0.9506245209000316, + "grad_norm": 2.0162851598030698, + "learning_rate": 2.5475892320996785e-08, + "loss": 0.952, + "step": 10541 + }, + { + "epoch": 0.9507147044235018, + "grad_norm": 0.6781760983178824, + "learning_rate": 2.5383027131564038e-08, + "loss": 0.7983, + "step": 10542 + }, + { + "epoch": 0.950804887946972, + "grad_norm": 1.7832033924048265, + "learning_rate": 2.52903304273393e-08, + "loss": 0.9492, + "step": 10543 + }, + { + "epoch": 0.9508950714704424, + "grad_norm": 1.5547090461809816, + "learning_rate": 2.519780221623202e-08, + "loss": 0.9621, + "step": 10544 + }, + { + "epoch": 0.9509852549939126, + "grad_norm": 1.9132987230717229, + "learning_rate": 2.510544250613722e-08, + "loss": 0.9369, + "step": 10545 + }, + { + "epoch": 0.9510754385173829, + "grad_norm": 1.6943120861298513, + "learning_rate": 2.501325130493548e-08, + "loss": 1.0783, + "step": 10546 + }, + { + "epoch": 0.9511656220408531, + "grad_norm": 0.7155296078866483, + "learning_rate": 2.4921228620493395e-08, + "loss": 0.8151, + "step": 10547 + }, + { + "epoch": 0.9512558055643234, + "grad_norm": 0.7351959185382179, + "learning_rate": 2.4829374460662244e-08, + "loss": 0.8518, + "step": 10548 + }, + { + "epoch": 0.9513459890877937, + "grad_norm": 2.078621597040419, + "learning_rate": 2.473768883327976e-08, + "loss": 0.9957, + "step": 10549 + }, + { + "epoch": 0.9514361726112639, + "grad_norm": 1.633024722092626, + "learning_rate": 2.464617174616923e-08, + "loss": 0.9425, + "step": 10550 + }, + { + "epoch": 0.9515263561347341, + "grad_norm": 1.6095195816427355, + "learning_rate": 2.455482320713953e-08, + "loss": 0.979, + "step": 10551 + }, + { + "epoch": 0.9516165396582045, + "grad_norm": 1.7635693356603586, + "learning_rate": 2.4463643223984643e-08, + "loss": 1.039, + "step": 10552 + }, + { + "epoch": 0.9517067231816747, + "grad_norm": 2.020419547069172, + "learning_rate": 2.4372631804484567e-08, + "loss": 1.0513, + "step": 10553 + }, + { + "epoch": 0.9517969067051449, + "grad_norm": 2.660447338607533, + "learning_rate": 2.4281788956405313e-08, + "loss": 1.0601, + "step": 10554 + }, + { + "epoch": 0.9518870902286153, + "grad_norm": 5.093294088169178, + "learning_rate": 2.4191114687497572e-08, + "loss": 0.9057, + "step": 10555 + }, + { + "epoch": 0.9519772737520855, + "grad_norm": 2.8211581581726453, + "learning_rate": 2.4100609005498706e-08, + "loss": 0.942, + "step": 10556 + }, + { + "epoch": 0.9520674572755558, + "grad_norm": 1.7957066593795092, + "learning_rate": 2.4010271918130764e-08, + "loss": 0.9143, + "step": 10557 + }, + { + "epoch": 0.952157640799026, + "grad_norm": 1.6158419553786698, + "learning_rate": 2.39201034331018e-08, + "loss": 0.9417, + "step": 10558 + }, + { + "epoch": 0.9522478243224963, + "grad_norm": 1.6261193748281233, + "learning_rate": 2.3830103558105663e-08, + "loss": 0.9728, + "step": 10559 + }, + { + "epoch": 0.9523380078459666, + "grad_norm": 3.1869980659969745, + "learning_rate": 2.374027230082154e-08, + "loss": 0.9869, + "step": 10560 + }, + { + "epoch": 0.9524281913694368, + "grad_norm": 0.6933341008267678, + "learning_rate": 2.365060966891441e-08, + "loss": 0.8324, + "step": 10561 + }, + { + "epoch": 0.952518374892907, + "grad_norm": 1.5918466010254535, + "learning_rate": 2.3561115670034827e-08, + "loss": 0.9757, + "step": 10562 + }, + { + "epoch": 0.9526085584163774, + "grad_norm": 1.693132607522844, + "learning_rate": 2.3471790311818675e-08, + "loss": 0.9978, + "step": 10563 + }, + { + "epoch": 0.9526987419398476, + "grad_norm": 3.7133025974757192, + "learning_rate": 2.338263360188808e-08, + "loss": 0.9685, + "step": 10564 + }, + { + "epoch": 0.9527889254633178, + "grad_norm": 2.2698487996903696, + "learning_rate": 2.329364554784985e-08, + "loss": 0.9203, + "step": 10565 + }, + { + "epoch": 0.9528791089867881, + "grad_norm": 1.6988019760971846, + "learning_rate": 2.3204826157297465e-08, + "loss": 1.0366, + "step": 10566 + }, + { + "epoch": 0.9529692925102584, + "grad_norm": 1.723688140932095, + "learning_rate": 2.3116175437809082e-08, + "loss": 0.914, + "step": 10567 + }, + { + "epoch": 0.9530594760337286, + "grad_norm": 0.6476086984440761, + "learning_rate": 2.30276933969491e-08, + "loss": 0.8327, + "step": 10568 + }, + { + "epoch": 0.9531496595571989, + "grad_norm": 1.7547216750066892, + "learning_rate": 2.2939380042267255e-08, + "loss": 1.0387, + "step": 10569 + }, + { + "epoch": 0.9532398430806691, + "grad_norm": 1.480608850457878, + "learning_rate": 2.2851235381298627e-08, + "loss": 0.8827, + "step": 10570 + }, + { + "epoch": 0.9533300266041395, + "grad_norm": 1.5709074800566158, + "learning_rate": 2.2763259421564986e-08, + "loss": 0.9061, + "step": 10571 + }, + { + "epoch": 0.9534202101276097, + "grad_norm": 1.6997770948689745, + "learning_rate": 2.2675452170571873e-08, + "loss": 0.9481, + "step": 10572 + }, + { + "epoch": 0.9535103936510799, + "grad_norm": 1.66468025883381, + "learning_rate": 2.2587813635812414e-08, + "loss": 0.9733, + "step": 10573 + }, + { + "epoch": 0.9536005771745502, + "grad_norm": 1.5628087310959151, + "learning_rate": 2.2500343824763958e-08, + "loss": 1.0657, + "step": 10574 + }, + { + "epoch": 0.9536907606980205, + "grad_norm": 1.7237168418566515, + "learning_rate": 2.2413042744890088e-08, + "loss": 0.9263, + "step": 10575 + }, + { + "epoch": 0.9537809442214907, + "grad_norm": 1.5226292740383154, + "learning_rate": 2.2325910403639514e-08, + "loss": 0.9684, + "step": 10576 + }, + { + "epoch": 0.953871127744961, + "grad_norm": 1.8297230183633897, + "learning_rate": 2.223894680844718e-08, + "loss": 0.934, + "step": 10577 + }, + { + "epoch": 0.9539613112684313, + "grad_norm": 1.381252637489528, + "learning_rate": 2.2152151966733146e-08, + "loss": 0.8169, + "step": 10578 + }, + { + "epoch": 0.9540514947919015, + "grad_norm": 1.3864561450156643, + "learning_rate": 2.2065525885903267e-08, + "loss": 1.0146, + "step": 10579 + }, + { + "epoch": 0.9541416783153718, + "grad_norm": 3.5460104539849437, + "learning_rate": 2.1979068573348747e-08, + "loss": 1.0257, + "step": 10580 + }, + { + "epoch": 0.954231861838842, + "grad_norm": 2.2864269929329586, + "learning_rate": 2.1892780036447013e-08, + "loss": 0.943, + "step": 10581 + }, + { + "epoch": 0.9543220453623124, + "grad_norm": 13.723522980919022, + "learning_rate": 2.1806660282560175e-08, + "loss": 1.122, + "step": 10582 + }, + { + "epoch": 0.9544122288857826, + "grad_norm": 1.684972848758358, + "learning_rate": 2.1720709319037024e-08, + "loss": 1.0035, + "step": 10583 + }, + { + "epoch": 0.9545024124092528, + "grad_norm": 1.413927777864501, + "learning_rate": 2.1634927153211023e-08, + "loss": 0.9438, + "step": 10584 + }, + { + "epoch": 0.954592595932723, + "grad_norm": 1.9150663896212854, + "learning_rate": 2.1549313792401437e-08, + "loss": 1.043, + "step": 10585 + }, + { + "epoch": 0.9546827794561934, + "grad_norm": 1.9139890695616548, + "learning_rate": 2.1463869243913746e-08, + "loss": 1.0689, + "step": 10586 + }, + { + "epoch": 0.9547729629796636, + "grad_norm": 1.6222142035723053, + "learning_rate": 2.1378593515037902e-08, + "loss": 0.9598, + "step": 10587 + }, + { + "epoch": 0.9548631465031339, + "grad_norm": 1.4244010201823034, + "learning_rate": 2.129348661305075e-08, + "loss": 1.035, + "step": 10588 + }, + { + "epoch": 0.9549533300266041, + "grad_norm": 1.5959732406017328, + "learning_rate": 2.1208548545213813e-08, + "loss": 1.0124, + "step": 10589 + }, + { + "epoch": 0.9550435135500744, + "grad_norm": 3.1892828349843603, + "learning_rate": 2.1123779318774404e-08, + "loss": 0.9337, + "step": 10590 + }, + { + "epoch": 0.9551336970735447, + "grad_norm": 1.4408937282197287, + "learning_rate": 2.1039178940965408e-08, + "loss": 0.9233, + "step": 10591 + }, + { + "epoch": 0.9552238805970149, + "grad_norm": 1.4522743606612467, + "learning_rate": 2.0954747419005712e-08, + "loss": 1.0413, + "step": 10592 + }, + { + "epoch": 0.9553140641204851, + "grad_norm": 1.7980216969930218, + "learning_rate": 2.087048476009934e-08, + "loss": 0.9588, + "step": 10593 + }, + { + "epoch": 0.9554042476439555, + "grad_norm": 2.2070047832691264, + "learning_rate": 2.0786390971435862e-08, + "loss": 0.9539, + "step": 10594 + }, + { + "epoch": 0.9554944311674257, + "grad_norm": 1.9555261119639271, + "learning_rate": 2.070246606019088e-08, + "loss": 0.9496, + "step": 10595 + }, + { + "epoch": 0.955584614690896, + "grad_norm": 38.73983769214918, + "learning_rate": 2.0618710033525112e-08, + "loss": 0.8907, + "step": 10596 + }, + { + "epoch": 0.9556747982143662, + "grad_norm": 1.6479686775471534, + "learning_rate": 2.053512289858528e-08, + "loss": 0.9447, + "step": 10597 + }, + { + "epoch": 0.9557649817378365, + "grad_norm": 2.138501415688361, + "learning_rate": 2.0451704662503456e-08, + "loss": 1.021, + "step": 10598 + }, + { + "epoch": 0.9558551652613068, + "grad_norm": 2.1691565756380493, + "learning_rate": 2.0368455332397282e-08, + "loss": 1.0605, + "step": 10599 + }, + { + "epoch": 0.955945348784777, + "grad_norm": 1.643793080856812, + "learning_rate": 2.0285374915369967e-08, + "loss": 0.9944, + "step": 10600 + }, + { + "epoch": 0.9560355323082473, + "grad_norm": 1.8851704375218887, + "learning_rate": 2.020246341851073e-08, + "loss": 1.018, + "step": 10601 + }, + { + "epoch": 0.9561257158317176, + "grad_norm": 1.765218704873293, + "learning_rate": 2.0119720848893463e-08, + "loss": 1.0243, + "step": 10602 + }, + { + "epoch": 0.9562158993551878, + "grad_norm": 1.535422420965391, + "learning_rate": 2.0037147213578964e-08, + "loss": 0.9289, + "step": 10603 + }, + { + "epoch": 0.956306082878658, + "grad_norm": 1.6384539817375423, + "learning_rate": 1.9954742519612265e-08, + "loss": 1.008, + "step": 10604 + }, + { + "epoch": 0.9563962664021284, + "grad_norm": 2.451483012965847, + "learning_rate": 1.9872506774024633e-08, + "loss": 0.9581, + "step": 10605 + }, + { + "epoch": 0.9564864499255986, + "grad_norm": 1.764819273624849, + "learning_rate": 1.979043998383334e-08, + "loss": 0.9929, + "step": 10606 + }, + { + "epoch": 0.9565766334490688, + "grad_norm": 1.3329296047198482, + "learning_rate": 1.970854215604034e-08, + "loss": 0.9213, + "step": 10607 + }, + { + "epoch": 0.9566668169725391, + "grad_norm": 1.8656028503432183, + "learning_rate": 1.9626813297633826e-08, + "loss": 0.9481, + "step": 10608 + }, + { + "epoch": 0.9567570004960094, + "grad_norm": 1.8656723706542913, + "learning_rate": 1.954525341558688e-08, + "loss": 0.8759, + "step": 10609 + }, + { + "epoch": 0.9568471840194797, + "grad_norm": 1.8983050641619552, + "learning_rate": 1.9463862516859498e-08, + "loss": 0.9465, + "step": 10610 + }, + { + "epoch": 0.9569373675429499, + "grad_norm": 1.7807832072796919, + "learning_rate": 1.938264060839545e-08, + "loss": 1.0634, + "step": 10611 + }, + { + "epoch": 0.9570275510664201, + "grad_norm": 2.031060664816248, + "learning_rate": 1.9301587697126086e-08, + "loss": 0.9556, + "step": 10612 + }, + { + "epoch": 0.9571177345898905, + "grad_norm": 1.7974374305541772, + "learning_rate": 1.9220703789966318e-08, + "loss": 0.9353, + "step": 10613 + }, + { + "epoch": 0.9572079181133607, + "grad_norm": 1.6021737025667382, + "learning_rate": 1.913998889381818e-08, + "loss": 0.9364, + "step": 10614 + }, + { + "epoch": 0.9572981016368309, + "grad_norm": 1.6849531098132338, + "learning_rate": 1.9059443015568387e-08, + "loss": 0.9881, + "step": 10615 + }, + { + "epoch": 0.9573882851603012, + "grad_norm": 2.0555165289991635, + "learning_rate": 1.8979066162089884e-08, + "loss": 0.992, + "step": 10616 + }, + { + "epoch": 0.9574784686837715, + "grad_norm": 1.5812041958292677, + "learning_rate": 1.889885834024052e-08, + "loss": 0.9832, + "step": 10617 + }, + { + "epoch": 0.9575686522072417, + "grad_norm": 1.5846435078581775, + "learning_rate": 1.8818819556864374e-08, + "loss": 0.9293, + "step": 10618 + }, + { + "epoch": 0.957658835730712, + "grad_norm": 6.838824725097542, + "learning_rate": 1.873894981879065e-08, + "loss": 0.9615, + "step": 10619 + }, + { + "epoch": 0.9577490192541822, + "grad_norm": 0.7064041391470662, + "learning_rate": 1.8659249132834342e-08, + "loss": 0.8271, + "step": 10620 + }, + { + "epoch": 0.9578392027776526, + "grad_norm": 1.3239842660176488, + "learning_rate": 1.857971750579579e-08, + "loss": 0.9607, + "step": 10621 + }, + { + "epoch": 0.9579293863011228, + "grad_norm": 1.7027588196916554, + "learning_rate": 1.8500354944461116e-08, + "loss": 0.8614, + "step": 10622 + }, + { + "epoch": 0.958019569824593, + "grad_norm": 2.2820549158882306, + "learning_rate": 1.8421161455602242e-08, + "loss": 0.9868, + "step": 10623 + }, + { + "epoch": 0.9581097533480634, + "grad_norm": 1.9276967146421056, + "learning_rate": 1.834213704597598e-08, + "loss": 1.0042, + "step": 10624 + }, + { + "epoch": 0.9581999368715336, + "grad_norm": 1.5893095651493205, + "learning_rate": 1.8263281722325385e-08, + "loss": 0.992, + "step": 10625 + }, + { + "epoch": 0.9582901203950038, + "grad_norm": 2.1211900777001675, + "learning_rate": 1.818459549137885e-08, + "loss": 0.9915, + "step": 10626 + }, + { + "epoch": 0.9583803039184741, + "grad_norm": 1.6155900667350336, + "learning_rate": 1.8106078359850117e-08, + "loss": 0.9687, + "step": 10627 + }, + { + "epoch": 0.9584704874419444, + "grad_norm": 1.5187222796135587, + "learning_rate": 1.802773033443894e-08, + "loss": 0.9947, + "step": 10628 + }, + { + "epoch": 0.9585606709654146, + "grad_norm": 1.5790155749479942, + "learning_rate": 1.7949551421830413e-08, + "loss": 0.945, + "step": 10629 + }, + { + "epoch": 0.9586508544888849, + "grad_norm": 4.080758950571682, + "learning_rate": 1.7871541628694752e-08, + "loss": 0.9482, + "step": 10630 + }, + { + "epoch": 0.9587410380123551, + "grad_norm": 1.6475483453398063, + "learning_rate": 1.779370096168864e-08, + "loss": 0.8875, + "step": 10631 + }, + { + "epoch": 0.9588312215358255, + "grad_norm": 1.8216792529835575, + "learning_rate": 1.771602942745387e-08, + "loss": 0.9431, + "step": 10632 + }, + { + "epoch": 0.9589214050592957, + "grad_norm": 2.6447575301894224, + "learning_rate": 1.763852703261759e-08, + "loss": 0.9616, + "step": 10633 + }, + { + "epoch": 0.9590115885827659, + "grad_norm": 1.7370925096156768, + "learning_rate": 1.756119378379295e-08, + "loss": 1.0426, + "step": 10634 + }, + { + "epoch": 0.9591017721062362, + "grad_norm": 1.6314220440051352, + "learning_rate": 1.7484029687578005e-08, + "loss": 1.0705, + "step": 10635 + }, + { + "epoch": 0.9591919556297065, + "grad_norm": 1.728564245542354, + "learning_rate": 1.740703475055727e-08, + "loss": 0.9533, + "step": 10636 + }, + { + "epoch": 0.9592821391531767, + "grad_norm": 1.7949965900399378, + "learning_rate": 1.7330208979300153e-08, + "loss": 1.0156, + "step": 10637 + }, + { + "epoch": 0.959372322676647, + "grad_norm": 1.4455134071723241, + "learning_rate": 1.725355238036208e-08, + "loss": 1.0223, + "step": 10638 + }, + { + "epoch": 0.9594625062001172, + "grad_norm": 1.9248616651891497, + "learning_rate": 1.7177064960283594e-08, + "loss": 1.0236, + "step": 10639 + }, + { + "epoch": 0.9595526897235875, + "grad_norm": 1.8943450521620926, + "learning_rate": 1.7100746725591253e-08, + "loss": 0.9747, + "step": 10640 + }, + { + "epoch": 0.9596428732470578, + "grad_norm": 2.138738873272774, + "learning_rate": 1.7024597682796517e-08, + "loss": 0.9691, + "step": 10641 + }, + { + "epoch": 0.959733056770528, + "grad_norm": 2.533981356641778, + "learning_rate": 1.6948617838397293e-08, + "loss": 1.056, + "step": 10642 + }, + { + "epoch": 0.9598232402939982, + "grad_norm": 1.3192818134311266, + "learning_rate": 1.6872807198876404e-08, + "loss": 1.0062, + "step": 10643 + }, + { + "epoch": 0.9599134238174686, + "grad_norm": 1.9547087085039854, + "learning_rate": 1.679716577070245e-08, + "loss": 1.0791, + "step": 10644 + }, + { + "epoch": 0.9600036073409388, + "grad_norm": 1.5688771382520077, + "learning_rate": 1.6721693560329596e-08, + "loss": 1.0665, + "step": 10645 + }, + { + "epoch": 0.960093790864409, + "grad_norm": 2.561267370236774, + "learning_rate": 1.6646390574197366e-08, + "loss": 0.8868, + "step": 10646 + }, + { + "epoch": 0.9601839743878793, + "grad_norm": 0.6871164509125358, + "learning_rate": 1.6571256818731504e-08, + "loss": 0.8334, + "step": 10647 + }, + { + "epoch": 0.9602741579113496, + "grad_norm": 1.7375673349629572, + "learning_rate": 1.6496292300342218e-08, + "loss": 0.913, + "step": 10648 + }, + { + "epoch": 0.9603643414348199, + "grad_norm": 2.135919564429258, + "learning_rate": 1.642149702542639e-08, + "loss": 0.9206, + "step": 10649 + }, + { + "epoch": 0.9604545249582901, + "grad_norm": 1.5572683870580704, + "learning_rate": 1.634687100036558e-08, + "loss": 0.9431, + "step": 10650 + }, + { + "epoch": 0.9605447084817604, + "grad_norm": 1.5847781599497333, + "learning_rate": 1.627241423152781e-08, + "loss": 0.9133, + "step": 10651 + }, + { + "epoch": 0.9606348920052307, + "grad_norm": 1.6761891684421273, + "learning_rate": 1.619812672526555e-08, + "loss": 1.0301, + "step": 10652 + }, + { + "epoch": 0.9607250755287009, + "grad_norm": 1.6801035521019299, + "learning_rate": 1.6124008487917727e-08, + "loss": 0.964, + "step": 10653 + }, + { + "epoch": 0.9608152590521711, + "grad_norm": 0.6408116836222683, + "learning_rate": 1.6050059525808623e-08, + "loss": 0.8318, + "step": 10654 + }, + { + "epoch": 0.9609054425756415, + "grad_norm": 0.6348783422308139, + "learning_rate": 1.597627984524763e-08, + "loss": 0.827, + "step": 10655 + }, + { + "epoch": 0.9609956260991117, + "grad_norm": 1.4700745636330803, + "learning_rate": 1.590266945253038e-08, + "loss": 1.0179, + "step": 10656 + }, + { + "epoch": 0.9610858096225819, + "grad_norm": 1.479271799927144, + "learning_rate": 1.582922835393763e-08, + "loss": 0.9951, + "step": 10657 + }, + { + "epoch": 0.9611759931460522, + "grad_norm": 2.1525926394138173, + "learning_rate": 1.5755956555735473e-08, + "loss": 0.9568, + "step": 10658 + }, + { + "epoch": 0.9612661766695225, + "grad_norm": 0.7200351503851893, + "learning_rate": 1.5682854064176244e-08, + "loss": 0.9016, + "step": 10659 + }, + { + "epoch": 0.9613563601929928, + "grad_norm": 1.501389178098513, + "learning_rate": 1.5609920885497395e-08, + "loss": 0.9102, + "step": 10660 + }, + { + "epoch": 0.961446543716463, + "grad_norm": 1.9020822960250827, + "learning_rate": 1.5537157025921732e-08, + "loss": 0.9986, + "step": 10661 + }, + { + "epoch": 0.9615367272399332, + "grad_norm": 1.9694471184670976, + "learning_rate": 1.5464562491658285e-08, + "loss": 0.9613, + "step": 10662 + }, + { + "epoch": 0.9616269107634036, + "grad_norm": 3.9990890180825356, + "learning_rate": 1.5392137288900764e-08, + "loss": 1.0171, + "step": 10663 + }, + { + "epoch": 0.9617170942868738, + "grad_norm": 0.5955095820965222, + "learning_rate": 1.531988142382934e-08, + "loss": 0.8165, + "step": 10664 + }, + { + "epoch": 0.961807277810344, + "grad_norm": 1.5644089291315064, + "learning_rate": 1.5247794902608634e-08, + "loss": 0.914, + "step": 10665 + }, + { + "epoch": 0.9618974613338143, + "grad_norm": 2.0444168589407927, + "learning_rate": 1.5175877731390398e-08, + "loss": 0.9371, + "step": 10666 + }, + { + "epoch": 0.9619876448572846, + "grad_norm": 1.6081505719378164, + "learning_rate": 1.510412991631016e-08, + "loss": 1.0058, + "step": 10667 + }, + { + "epoch": 0.9620778283807548, + "grad_norm": 1.318818686264915, + "learning_rate": 1.503255146349014e-08, + "loss": 1.0069, + "step": 10668 + }, + { + "epoch": 0.9621680119042251, + "grad_norm": 1.7880797373040715, + "learning_rate": 1.4961142379037893e-08, + "loss": 0.9411, + "step": 10669 + }, + { + "epoch": 0.9622581954276953, + "grad_norm": 1.6283591836044213, + "learning_rate": 1.4889902669046327e-08, + "loss": 1.0219, + "step": 10670 + }, + { + "epoch": 0.9623483789511657, + "grad_norm": 2.208196659777098, + "learning_rate": 1.4818832339594135e-08, + "loss": 0.9799, + "step": 10671 + }, + { + "epoch": 0.9624385624746359, + "grad_norm": 2.966545169642673, + "learning_rate": 1.474793139674535e-08, + "loss": 0.9309, + "step": 10672 + }, + { + "epoch": 0.9625287459981061, + "grad_norm": 1.7784099357263938, + "learning_rate": 1.4677199846549581e-08, + "loss": 0.9767, + "step": 10673 + }, + { + "epoch": 0.9626189295215765, + "grad_norm": 1.7323533381137148, + "learning_rate": 1.4606637695042224e-08, + "loss": 0.9795, + "step": 10674 + }, + { + "epoch": 0.9627091130450467, + "grad_norm": 1.6749149386315119, + "learning_rate": 1.4536244948243793e-08, + "loss": 1.0704, + "step": 10675 + }, + { + "epoch": 0.9627992965685169, + "grad_norm": 3.9980108560450045, + "learning_rate": 1.4466021612160595e-08, + "loss": 0.9467, + "step": 10676 + }, + { + "epoch": 0.9628894800919872, + "grad_norm": 2.3211757710902963, + "learning_rate": 1.4395967692784505e-08, + "loss": 0.9467, + "step": 10677 + }, + { + "epoch": 0.9629796636154575, + "grad_norm": 1.7153285564297285, + "learning_rate": 1.4326083196092963e-08, + "loss": 0.9069, + "step": 10678 + }, + { + "epoch": 0.9630698471389277, + "grad_norm": 2.0816764600822215, + "learning_rate": 1.42563681280492e-08, + "loss": 0.9908, + "step": 10679 + }, + { + "epoch": 0.963160030662398, + "grad_norm": 1.7712176036900666, + "learning_rate": 1.4186822494600902e-08, + "loss": 0.953, + "step": 10680 + }, + { + "epoch": 0.9632502141858682, + "grad_norm": 1.4861029756428312, + "learning_rate": 1.4117446301682877e-08, + "loss": 0.9657, + "step": 10681 + }, + { + "epoch": 0.9633403977093385, + "grad_norm": 1.546996680203743, + "learning_rate": 1.4048239555214392e-08, + "loss": 0.9316, + "step": 10682 + }, + { + "epoch": 0.9634305812328088, + "grad_norm": 1.5418601387213853, + "learning_rate": 1.3979202261100497e-08, + "loss": 1.0087, + "step": 10683 + }, + { + "epoch": 0.963520764756279, + "grad_norm": 3.146917766746386, + "learning_rate": 1.3910334425231817e-08, + "loss": 1.0402, + "step": 10684 + }, + { + "epoch": 0.9636109482797492, + "grad_norm": 1.5897497200867403, + "learning_rate": 1.384163605348454e-08, + "loss": 0.9769, + "step": 10685 + }, + { + "epoch": 0.9637011318032196, + "grad_norm": 1.7501617765582034, + "learning_rate": 1.3773107151720642e-08, + "loss": 1.0338, + "step": 10686 + }, + { + "epoch": 0.9637913153266898, + "grad_norm": 1.615998499067006, + "learning_rate": 1.3704747725787003e-08, + "loss": 1.0371, + "step": 10687 + }, + { + "epoch": 0.9638814988501601, + "grad_norm": 0.725647216509344, + "learning_rate": 1.3636557781516512e-08, + "loss": 0.8231, + "step": 10688 + }, + { + "epoch": 0.9639716823736303, + "grad_norm": 1.7881944041164686, + "learning_rate": 1.3568537324727847e-08, + "loss": 0.9823, + "step": 10689 + }, + { + "epoch": 0.9640618658971006, + "grad_norm": 2.4593837131990566, + "learning_rate": 1.3500686361224589e-08, + "loss": 0.9756, + "step": 10690 + }, + { + "epoch": 0.9641520494205709, + "grad_norm": 2.141171239505609, + "learning_rate": 1.3433004896796108e-08, + "loss": 0.9752, + "step": 10691 + }, + { + "epoch": 0.9642422329440411, + "grad_norm": 1.4732172780192319, + "learning_rate": 1.336549293721756e-08, + "loss": 0.8706, + "step": 10692 + }, + { + "epoch": 0.9643324164675113, + "grad_norm": 0.6590355609581141, + "learning_rate": 1.3298150488249227e-08, + "loss": 0.8089, + "step": 10693 + }, + { + "epoch": 0.9644225999909817, + "grad_norm": 2.327417150840102, + "learning_rate": 1.3230977555637401e-08, + "loss": 1.0543, + "step": 10694 + }, + { + "epoch": 0.9645127835144519, + "grad_norm": 1.6973539796928483, + "learning_rate": 1.3163974145113499e-08, + "loss": 0.9358, + "step": 10695 + }, + { + "epoch": 0.9646029670379221, + "grad_norm": 2.049956353816166, + "learning_rate": 1.3097140262394723e-08, + "loss": 0.8762, + "step": 10696 + }, + { + "epoch": 0.9646931505613925, + "grad_norm": 1.8821577144459334, + "learning_rate": 1.303047591318318e-08, + "loss": 1.0095, + "step": 10697 + }, + { + "epoch": 0.9647833340848627, + "grad_norm": 1.6029848063481247, + "learning_rate": 1.2963981103167875e-08, + "loss": 0.9132, + "step": 10698 + }, + { + "epoch": 0.964873517608333, + "grad_norm": 1.4373615032318419, + "learning_rate": 1.2897655838021825e-08, + "loss": 0.9609, + "step": 10699 + }, + { + "epoch": 0.9649637011318032, + "grad_norm": 1.7610221889815725, + "learning_rate": 1.2831500123404726e-08, + "loss": 0.9814, + "step": 10700 + }, + { + "epoch": 0.9650538846552735, + "grad_norm": 2.2076080378775047, + "learning_rate": 1.2765513964961172e-08, + "loss": 1.1279, + "step": 10701 + }, + { + "epoch": 0.9651440681787438, + "grad_norm": 1.5336718189114544, + "learning_rate": 1.2699697368321549e-08, + "loss": 0.9739, + "step": 10702 + }, + { + "epoch": 0.965234251702214, + "grad_norm": 1.430341227173979, + "learning_rate": 1.2634050339101366e-08, + "loss": 0.9636, + "step": 10703 + }, + { + "epoch": 0.9653244352256842, + "grad_norm": 1.4109485123078906, + "learning_rate": 1.2568572882902361e-08, + "loss": 0.9535, + "step": 10704 + }, + { + "epoch": 0.9654146187491546, + "grad_norm": 1.552021597161399, + "learning_rate": 1.2503265005311402e-08, + "loss": 0.9484, + "step": 10705 + }, + { + "epoch": 0.9655048022726248, + "grad_norm": 1.7550514068030145, + "learning_rate": 1.2438126711900698e-08, + "loss": 0.938, + "step": 10706 + }, + { + "epoch": 0.965594985796095, + "grad_norm": 1.5057186946638894, + "learning_rate": 1.2373158008228247e-08, + "loss": 0.8279, + "step": 10707 + }, + { + "epoch": 0.9656851693195653, + "grad_norm": 1.6162831454555138, + "learning_rate": 1.2308358899837833e-08, + "loss": 0.953, + "step": 10708 + }, + { + "epoch": 0.9657753528430356, + "grad_norm": 1.9558380814527043, + "learning_rate": 1.224372939225815e-08, + "loss": 0.9561, + "step": 10709 + }, + { + "epoch": 0.9658655363665059, + "grad_norm": 1.5435099364894767, + "learning_rate": 1.2179269491003674e-08, + "loss": 1.0596, + "step": 10710 + }, + { + "epoch": 0.9659557198899761, + "grad_norm": 1.7093838121371852, + "learning_rate": 1.2114979201574894e-08, + "loss": 0.8527, + "step": 10711 + }, + { + "epoch": 0.9660459034134463, + "grad_norm": 1.543099776873394, + "learning_rate": 1.2050858529456975e-08, + "loss": 0.9954, + "step": 10712 + }, + { + "epoch": 0.9661360869369167, + "grad_norm": 1.991422558699372, + "learning_rate": 1.1986907480121545e-08, + "loss": 0.9792, + "step": 10713 + }, + { + "epoch": 0.9662262704603869, + "grad_norm": 1.9996046033065686, + "learning_rate": 1.192312605902468e-08, + "loss": 0.9981, + "step": 10714 + }, + { + "epoch": 0.9663164539838571, + "grad_norm": 0.7952226824153333, + "learning_rate": 1.1859514271608917e-08, + "loss": 0.8383, + "step": 10715 + }, + { + "epoch": 0.9664066375073274, + "grad_norm": 2.3783440387523207, + "learning_rate": 1.1796072123301914e-08, + "loss": 1.0093, + "step": 10716 + }, + { + "epoch": 0.9664968210307977, + "grad_norm": 0.6375007040356507, + "learning_rate": 1.1732799619516897e-08, + "loss": 0.8184, + "step": 10717 + }, + { + "epoch": 0.9665870045542679, + "grad_norm": 1.7369481150202677, + "learning_rate": 1.1669696765652659e-08, + "loss": 0.9641, + "step": 10718 + }, + { + "epoch": 0.9666771880777382, + "grad_norm": 1.822436227611528, + "learning_rate": 1.1606763567093336e-08, + "loss": 1.0093, + "step": 10719 + }, + { + "epoch": 0.9667673716012085, + "grad_norm": 0.6774258089792703, + "learning_rate": 1.1544000029208857e-08, + "loss": 0.8091, + "step": 10720 + }, + { + "epoch": 0.9668575551246787, + "grad_norm": 1.8313180149193609, + "learning_rate": 1.148140615735449e-08, + "loss": 0.8394, + "step": 10721 + }, + { + "epoch": 0.966947738648149, + "grad_norm": 1.5458499324897894, + "learning_rate": 1.1418981956871076e-08, + "loss": 0.9525, + "step": 10722 + }, + { + "epoch": 0.9670379221716192, + "grad_norm": 2.0185244502149886, + "learning_rate": 1.1356727433085245e-08, + "loss": 0.9658, + "step": 10723 + }, + { + "epoch": 0.9671281056950896, + "grad_norm": 1.563309802967512, + "learning_rate": 1.1294642591308524e-08, + "loss": 1.004, + "step": 10724 + }, + { + "epoch": 0.9672182892185598, + "grad_norm": 1.3964492046489019, + "learning_rate": 1.1232727436838452e-08, + "loss": 0.966, + "step": 10725 + }, + { + "epoch": 0.96730847274203, + "grad_norm": 1.6803713803566254, + "learning_rate": 1.1170981974958138e-08, + "loss": 1.0234, + "step": 10726 + }, + { + "epoch": 0.9673986562655003, + "grad_norm": 1.87203121393215, + "learning_rate": 1.1109406210936035e-08, + "loss": 1.0398, + "step": 10727 + }, + { + "epoch": 0.9674888397889706, + "grad_norm": 1.9997534599937152, + "learning_rate": 1.1048000150025939e-08, + "loss": 1.0959, + "step": 10728 + }, + { + "epoch": 0.9675790233124408, + "grad_norm": 1.910181275498292, + "learning_rate": 1.0986763797467213e-08, + "loss": 0.8738, + "step": 10729 + }, + { + "epoch": 0.9676692068359111, + "grad_norm": 0.5849313107536149, + "learning_rate": 1.0925697158485459e-08, + "loss": 0.7803, + "step": 10730 + }, + { + "epoch": 0.9677593903593813, + "grad_norm": 1.8574658377626223, + "learning_rate": 1.0864800238290727e-08, + "loss": 0.9794, + "step": 10731 + }, + { + "epoch": 0.9678495738828516, + "grad_norm": 1.6410630867388927, + "learning_rate": 1.0804073042079309e-08, + "loss": 0.9415, + "step": 10732 + }, + { + "epoch": 0.9679397574063219, + "grad_norm": 0.6387495407804566, + "learning_rate": 1.0743515575032392e-08, + "loss": 0.825, + "step": 10733 + }, + { + "epoch": 0.9680299409297921, + "grad_norm": 1.945847893387466, + "learning_rate": 1.0683127842317619e-08, + "loss": 0.9862, + "step": 10734 + }, + { + "epoch": 0.9681201244532623, + "grad_norm": 2.456936153639635, + "learning_rate": 1.0622909849087314e-08, + "loss": 1.0088, + "step": 10735 + }, + { + "epoch": 0.9682103079767327, + "grad_norm": 1.9407244891772908, + "learning_rate": 1.0562861600479588e-08, + "loss": 1.0388, + "step": 10736 + }, + { + "epoch": 0.9683004915002029, + "grad_norm": 1.7725984041906002, + "learning_rate": 1.0502983101618345e-08, + "loss": 1.0691, + "step": 10737 + }, + { + "epoch": 0.9683906750236732, + "grad_norm": 1.661042459352882, + "learning_rate": 1.0443274357612386e-08, + "loss": 0.8561, + "step": 10738 + }, + { + "epoch": 0.9684808585471434, + "grad_norm": 1.9782356146910447, + "learning_rate": 1.0383735373556524e-08, + "loss": 0.9918, + "step": 10739 + }, + { + "epoch": 0.9685710420706137, + "grad_norm": 1.30328298109667, + "learning_rate": 1.0324366154531139e-08, + "loss": 1.0108, + "step": 10740 + }, + { + "epoch": 0.968661225594084, + "grad_norm": 1.6957274601509842, + "learning_rate": 1.0265166705601735e-08, + "loss": 0.945, + "step": 10741 + }, + { + "epoch": 0.9687514091175542, + "grad_norm": 0.67116373694167, + "learning_rate": 1.0206137031819606e-08, + "loss": 0.8481, + "step": 10742 + }, + { + "epoch": 0.9688415926410244, + "grad_norm": 1.4742722283767802, + "learning_rate": 1.0147277138221388e-08, + "loss": 0.9903, + "step": 10743 + }, + { + "epoch": 0.9689317761644948, + "grad_norm": 1.6301686908748487, + "learning_rate": 1.0088587029829287e-08, + "loss": 1.0142, + "step": 10744 + }, + { + "epoch": 0.969021959687965, + "grad_norm": 1.4317811738222828, + "learning_rate": 1.003006671165152e-08, + "loss": 1.0789, + "step": 10745 + }, + { + "epoch": 0.9691121432114352, + "grad_norm": 1.5733892560303764, + "learning_rate": 9.971716188680978e-09, + "loss": 0.9714, + "step": 10746 + }, + { + "epoch": 0.9692023267349056, + "grad_norm": 2.586871390532588, + "learning_rate": 9.91353546589635e-09, + "loss": 0.9417, + "step": 10747 + }, + { + "epoch": 0.9692925102583758, + "grad_norm": 1.7855558720213673, + "learning_rate": 9.855524548262106e-09, + "loss": 0.9146, + "step": 10748 + }, + { + "epoch": 0.969382693781846, + "grad_norm": 1.526957743616504, + "learning_rate": 9.797683440728288e-09, + "loss": 0.9231, + "step": 10749 + }, + { + "epoch": 0.9694728773053163, + "grad_norm": 1.6218094682041464, + "learning_rate": 9.740012148229836e-09, + "loss": 0.8936, + "step": 10750 + }, + { + "epoch": 0.9695630608287866, + "grad_norm": 1.4395295618433046, + "learning_rate": 9.682510675687705e-09, + "loss": 0.9276, + "step": 10751 + }, + { + "epoch": 0.9696532443522569, + "grad_norm": 1.8089686884619014, + "learning_rate": 9.625179028008191e-09, + "loss": 0.9062, + "step": 10752 + }, + { + "epoch": 0.9697434278757271, + "grad_norm": 1.560714922702506, + "learning_rate": 9.568017210083379e-09, + "loss": 0.8227, + "step": 10753 + }, + { + "epoch": 0.9698336113991973, + "grad_norm": 2.1816072931022985, + "learning_rate": 9.511025226790259e-09, + "loss": 0.9512, + "step": 10754 + }, + { + "epoch": 0.9699237949226677, + "grad_norm": 2.2552735728165283, + "learning_rate": 9.454203082992052e-09, + "loss": 0.9528, + "step": 10755 + }, + { + "epoch": 0.9700139784461379, + "grad_norm": 1.969211705882389, + "learning_rate": 9.3975507835371e-09, + "loss": 0.8863, + "step": 10756 + }, + { + "epoch": 0.9701041619696081, + "grad_norm": 0.6231084571945834, + "learning_rate": 9.341068333259094e-09, + "loss": 0.8019, + "step": 10757 + }, + { + "epoch": 0.9701943454930784, + "grad_norm": 1.643606764205787, + "learning_rate": 9.28475573697729e-09, + "loss": 0.9533, + "step": 10758 + }, + { + "epoch": 0.9702845290165487, + "grad_norm": 1.6734565313487657, + "learning_rate": 9.228612999497177e-09, + "loss": 0.9778, + "step": 10759 + }, + { + "epoch": 0.970374712540019, + "grad_norm": 1.868904569223409, + "learning_rate": 9.172640125608478e-09, + "loss": 0.9231, + "step": 10760 + }, + { + "epoch": 0.9704648960634892, + "grad_norm": 1.5434673035040567, + "learning_rate": 9.116837120087817e-09, + "loss": 0.9918, + "step": 10761 + }, + { + "epoch": 0.9705550795869594, + "grad_norm": 4.4237470674141255, + "learning_rate": 9.061203987695832e-09, + "loss": 1.0438, + "step": 10762 + }, + { + "epoch": 0.9706452631104298, + "grad_norm": 2.031074985887306, + "learning_rate": 9.005740733180055e-09, + "loss": 0.9215, + "step": 10763 + }, + { + "epoch": 0.9707354466339, + "grad_norm": 1.6649677518637092, + "learning_rate": 8.950447361272483e-09, + "loss": 1.0611, + "step": 10764 + }, + { + "epoch": 0.9708256301573702, + "grad_norm": 2.0702498516464596, + "learning_rate": 8.895323876691784e-09, + "loss": 0.9818, + "step": 10765 + }, + { + "epoch": 0.9709158136808405, + "grad_norm": 1.922707416174463, + "learning_rate": 8.840370284140419e-09, + "loss": 1.0259, + "step": 10766 + }, + { + "epoch": 0.9710059972043108, + "grad_norm": 1.9888654948420936, + "learning_rate": 8.78558658830797e-09, + "loss": 0.894, + "step": 10767 + }, + { + "epoch": 0.971096180727781, + "grad_norm": 0.7889838226676025, + "learning_rate": 8.730972793868696e-09, + "loss": 0.8141, + "step": 10768 + }, + { + "epoch": 0.9711863642512513, + "grad_norm": 1.6940620134976057, + "learning_rate": 8.67652890548265e-09, + "loss": 0.9403, + "step": 10769 + }, + { + "epoch": 0.9712765477747216, + "grad_norm": 1.5046222518131402, + "learning_rate": 8.622254927795004e-09, + "loss": 0.9592, + "step": 10770 + }, + { + "epoch": 0.9713667312981918, + "grad_norm": 3.2626907819015987, + "learning_rate": 8.568150865436941e-09, + "loss": 0.9274, + "step": 10771 + }, + { + "epoch": 0.9714569148216621, + "grad_norm": 1.8133771681162236, + "learning_rate": 8.514216723024991e-09, + "loss": 0.987, + "step": 10772 + }, + { + "epoch": 0.9715470983451323, + "grad_norm": 3.3848366264913157, + "learning_rate": 8.460452505161031e-09, + "loss": 0.8955, + "step": 10773 + }, + { + "epoch": 0.9716372818686027, + "grad_norm": 2.6808564330736173, + "learning_rate": 8.4068582164325e-09, + "loss": 0.9986, + "step": 10774 + }, + { + "epoch": 0.9717274653920729, + "grad_norm": 2.042489048381006, + "learning_rate": 8.353433861412406e-09, + "loss": 0.9567, + "step": 10775 + }, + { + "epoch": 0.9718176489155431, + "grad_norm": 2.111640899950614, + "learning_rate": 8.300179444658883e-09, + "loss": 0.9334, + "step": 10776 + }, + { + "epoch": 0.9719078324390134, + "grad_norm": 2.2129230547394796, + "learning_rate": 8.247094970716296e-09, + "loss": 1.0682, + "step": 10777 + }, + { + "epoch": 0.9719980159624837, + "grad_norm": 1.7581333121961995, + "learning_rate": 8.19418044411413e-09, + "loss": 1.0163, + "step": 10778 + }, + { + "epoch": 0.9720881994859539, + "grad_norm": 1.4707913112671112, + "learning_rate": 8.141435869367219e-09, + "loss": 0.977, + "step": 10779 + }, + { + "epoch": 0.9721783830094242, + "grad_norm": 1.4704228877628407, + "learning_rate": 8.088861250975742e-09, + "loss": 1.013, + "step": 10780 + }, + { + "epoch": 0.9722685665328944, + "grad_norm": 2.3416621763609418, + "learning_rate": 8.036456593426111e-09, + "loss": 0.9433, + "step": 10781 + }, + { + "epoch": 0.9723587500563647, + "grad_norm": 1.614616746454185, + "learning_rate": 7.984221901189415e-09, + "loss": 1.0118, + "step": 10782 + }, + { + "epoch": 0.972448933579835, + "grad_norm": 1.4725650182336314, + "learning_rate": 7.932157178722976e-09, + "loss": 0.8996, + "step": 10783 + }, + { + "epoch": 0.9725391171033052, + "grad_norm": 1.4733973897124546, + "learning_rate": 7.880262430468799e-09, + "loss": 0.9429, + "step": 10784 + }, + { + "epoch": 0.9726293006267754, + "grad_norm": 1.524875842183857, + "learning_rate": 7.828537660855339e-09, + "loss": 0.9082, + "step": 10785 + }, + { + "epoch": 0.9727194841502458, + "grad_norm": 1.9583631540287034, + "learning_rate": 7.776982874295512e-09, + "loss": 0.9919, + "step": 10786 + }, + { + "epoch": 0.972809667673716, + "grad_norm": 2.3390915414641995, + "learning_rate": 7.725598075188688e-09, + "loss": 0.8835, + "step": 10787 + }, + { + "epoch": 0.9728998511971862, + "grad_norm": 1.667063483047181, + "learning_rate": 7.674383267918916e-09, + "loss": 0.9258, + "step": 10788 + }, + { + "epoch": 0.9729900347206565, + "grad_norm": 1.4974004949967066, + "learning_rate": 7.623338456856476e-09, + "loss": 0.9623, + "step": 10789 + }, + { + "epoch": 0.9730802182441268, + "grad_norm": 1.5636094541381458, + "learning_rate": 7.572463646356554e-09, + "loss": 0.8686, + "step": 10790 + }, + { + "epoch": 0.9731704017675971, + "grad_norm": 1.6560889831583778, + "learning_rate": 7.521758840760339e-09, + "loss": 0.9428, + "step": 10791 + }, + { + "epoch": 0.9732605852910673, + "grad_norm": 10.906298618153633, + "learning_rate": 7.471224044393931e-09, + "loss": 0.9324, + "step": 10792 + }, + { + "epoch": 0.9733507688145376, + "grad_norm": 1.433307671132389, + "learning_rate": 7.420859261569434e-09, + "loss": 1.0343, + "step": 10793 + }, + { + "epoch": 0.9734409523380079, + "grad_norm": 2.565728177819797, + "learning_rate": 7.370664496584078e-09, + "loss": 1.0333, + "step": 10794 + }, + { + "epoch": 0.9735311358614781, + "grad_norm": 1.8748792609440978, + "learning_rate": 7.3206397537211026e-09, + "loss": 0.9003, + "step": 10795 + }, + { + "epoch": 0.9736213193849483, + "grad_norm": 1.9922502096441472, + "learning_rate": 7.270785037248428e-09, + "loss": 0.945, + "step": 10796 + }, + { + "epoch": 0.9737115029084187, + "grad_norm": 1.9476153978821453, + "learning_rate": 7.221100351420428e-09, + "loss": 0.9367, + "step": 10797 + }, + { + "epoch": 0.9738016864318889, + "grad_norm": 1.859350829408124, + "learning_rate": 7.171585700475935e-09, + "loss": 0.9052, + "step": 10798 + }, + { + "epoch": 0.9738918699553591, + "grad_norm": 3.895804882795653, + "learning_rate": 7.122241088640235e-09, + "loss": 1.0297, + "step": 10799 + }, + { + "epoch": 0.9739820534788294, + "grad_norm": 1.785047665555305, + "learning_rate": 7.073066520123516e-09, + "loss": 1.1026, + "step": 10800 + }, + { + "epoch": 0.9740722370022997, + "grad_norm": 6.024746407518108, + "learning_rate": 7.0240619991217555e-09, + "loss": 0.8686, + "step": 10801 + }, + { + "epoch": 0.97416242052577, + "grad_norm": 1.6275724443523214, + "learning_rate": 6.975227529816052e-09, + "loss": 1.0517, + "step": 10802 + }, + { + "epoch": 0.9742526040492402, + "grad_norm": 2.7249794110561516, + "learning_rate": 6.926563116373296e-09, + "loss": 1.0196, + "step": 10803 + }, + { + "epoch": 0.9743427875727104, + "grad_norm": 1.7108968790884604, + "learning_rate": 6.878068762945943e-09, + "loss": 0.9114, + "step": 10804 + }, + { + "epoch": 0.9744329710961808, + "grad_norm": 1.5118558757403497, + "learning_rate": 6.829744473671794e-09, + "loss": 1.0119, + "step": 10805 + }, + { + "epoch": 0.974523154619651, + "grad_norm": 0.5916418495704534, + "learning_rate": 6.781590252674219e-09, + "loss": 0.7148, + "step": 10806 + }, + { + "epoch": 0.9746133381431212, + "grad_norm": 1.3142268764093576, + "learning_rate": 6.733606104061484e-09, + "loss": 0.9836, + "step": 10807 + }, + { + "epoch": 0.9747035216665915, + "grad_norm": 1.3368181148230827, + "learning_rate": 6.6857920319283165e-09, + "loss": 0.9196, + "step": 10808 + }, + { + "epoch": 0.9747937051900618, + "grad_norm": 5.434968907355937, + "learning_rate": 6.638148040354563e-09, + "loss": 0.9503, + "step": 10809 + }, + { + "epoch": 0.974883888713532, + "grad_norm": 2.1101110233910387, + "learning_rate": 6.590674133405194e-09, + "loss": 0.9333, + "step": 10810 + }, + { + "epoch": 0.9749740722370023, + "grad_norm": 1.4760447406759223, + "learning_rate": 6.5433703151311914e-09, + "loss": 0.9863, + "step": 10811 + }, + { + "epoch": 0.9750642557604725, + "grad_norm": 3.7176187541697954, + "learning_rate": 6.49623658956866e-09, + "loss": 0.9766, + "step": 10812 + }, + { + "epoch": 0.9751544392839429, + "grad_norm": 1.6322753606816403, + "learning_rate": 6.44927296073905e-09, + "loss": 0.8512, + "step": 10813 + }, + { + "epoch": 0.9752446228074131, + "grad_norm": 1.522286787465968, + "learning_rate": 6.402479432649821e-09, + "loss": 1.059, + "step": 10814 + }, + { + "epoch": 0.9753348063308833, + "grad_norm": 1.6056541465765843, + "learning_rate": 6.355856009293781e-09, + "loss": 0.9683, + "step": 10815 + }, + { + "epoch": 0.9754249898543537, + "grad_norm": 3.886326510319483, + "learning_rate": 6.3094026946488575e-09, + "loss": 0.9483, + "step": 10816 + }, + { + "epoch": 0.9755151733778239, + "grad_norm": 1.426209100777093, + "learning_rate": 6.2631194926787704e-09, + "loss": 0.9779, + "step": 10817 + }, + { + "epoch": 0.9756053569012941, + "grad_norm": 1.755380058000407, + "learning_rate": 6.217006407332581e-09, + "loss": 0.8746, + "step": 10818 + }, + { + "epoch": 0.9756955404247644, + "grad_norm": 1.7459652573929394, + "learning_rate": 6.1710634425453654e-09, + "loss": 0.9217, + "step": 10819 + }, + { + "epoch": 0.9757857239482347, + "grad_norm": 2.0668421035054307, + "learning_rate": 6.1252906022366544e-09, + "loss": 0.9956, + "step": 10820 + }, + { + "epoch": 0.9758759074717049, + "grad_norm": 1.5693352538453318, + "learning_rate": 6.079687890312213e-09, + "loss": 0.952, + "step": 10821 + }, + { + "epoch": 0.9759660909951752, + "grad_norm": 4.703426313411677, + "learning_rate": 6.034255310663372e-09, + "loss": 0.9533, + "step": 10822 + }, + { + "epoch": 0.9760562745186454, + "grad_norm": 11.591912548127574, + "learning_rate": 5.988992867166143e-09, + "loss": 1.064, + "step": 10823 + }, + { + "epoch": 0.9761464580421157, + "grad_norm": 4.882880663586722, + "learning_rate": 5.943900563682991e-09, + "loss": 1.0131, + "step": 10824 + }, + { + "epoch": 0.976236641565586, + "grad_norm": 0.723392585694884, + "learning_rate": 5.898978404061506e-09, + "loss": 0.8925, + "step": 10825 + }, + { + "epoch": 0.9763268250890562, + "grad_norm": 1.9742487456409632, + "learning_rate": 5.85422639213462e-09, + "loss": 0.9694, + "step": 10826 + }, + { + "epoch": 0.9764170086125264, + "grad_norm": 1.529943492693626, + "learning_rate": 5.809644531720614e-09, + "loss": 0.985, + "step": 10827 + }, + { + "epoch": 0.9765071921359968, + "grad_norm": 11.322914345795722, + "learning_rate": 5.765232826623556e-09, + "loss": 1.0026, + "step": 10828 + }, + { + "epoch": 0.976597375659467, + "grad_norm": 1.9479255744064026, + "learning_rate": 5.720991280633081e-09, + "loss": 0.9537, + "step": 10829 + }, + { + "epoch": 0.9766875591829373, + "grad_norm": 2.776225319537539, + "learning_rate": 5.676919897523724e-09, + "loss": 0.9575, + "step": 10830 + }, + { + "epoch": 0.9767777427064075, + "grad_norm": 1.8842687396554658, + "learning_rate": 5.633018681056256e-09, + "loss": 0.9814, + "step": 10831 + }, + { + "epoch": 0.9768679262298778, + "grad_norm": 1.6471290638515912, + "learning_rate": 5.589287634976569e-09, + "loss": 0.9757, + "step": 10832 + }, + { + "epoch": 0.9769581097533481, + "grad_norm": 1.6152719403364881, + "learning_rate": 5.5457267630159014e-09, + "loss": 0.9448, + "step": 10833 + }, + { + "epoch": 0.9770482932768183, + "grad_norm": 2.02903556426321, + "learning_rate": 5.5023360688910555e-09, + "loss": 1.0088, + "step": 10834 + }, + { + "epoch": 0.9771384768002885, + "grad_norm": 1.510470720398173, + "learning_rate": 5.459115556304183e-09, + "loss": 1.023, + "step": 10835 + }, + { + "epoch": 0.9772286603237589, + "grad_norm": 2.294845307067106, + "learning_rate": 5.416065228943889e-09, + "loss": 0.9043, + "step": 10836 + }, + { + "epoch": 0.9773188438472291, + "grad_norm": 1.539506877124339, + "learning_rate": 5.373185090482568e-09, + "loss": 0.9251, + "step": 10837 + }, + { + "epoch": 0.9774090273706993, + "grad_norm": 1.6577804260315139, + "learning_rate": 5.330475144579516e-09, + "loss": 0.9492, + "step": 10838 + }, + { + "epoch": 0.9774992108941697, + "grad_norm": 1.5385914192945225, + "learning_rate": 5.2879353948787065e-09, + "loss": 1.059, + "step": 10839 + }, + { + "epoch": 0.9775893944176399, + "grad_norm": 2.288979798994005, + "learning_rate": 5.245565845010125e-09, + "loss": 1.0338, + "step": 10840 + }, + { + "epoch": 0.9776795779411102, + "grad_norm": 1.8929685459617036, + "learning_rate": 5.2033664985886575e-09, + "loss": 0.9056, + "step": 10841 + }, + { + "epoch": 0.9777697614645804, + "grad_norm": 2.350435005670514, + "learning_rate": 5.161337359215201e-09, + "loss": 1.121, + "step": 10842 + }, + { + "epoch": 0.9778599449880507, + "grad_norm": 1.5580411232453504, + "learning_rate": 5.119478430475999e-09, + "loss": 0.9847, + "step": 10843 + }, + { + "epoch": 0.977950128511521, + "grad_norm": 2.1122453163610655, + "learning_rate": 5.077789715942416e-09, + "loss": 0.9187, + "step": 10844 + }, + { + "epoch": 0.9780403120349912, + "grad_norm": 1.74959041706979, + "learning_rate": 5.036271219171606e-09, + "loss": 1.0007, + "step": 10845 + }, + { + "epoch": 0.9781304955584614, + "grad_norm": 2.0107398160500827, + "learning_rate": 4.994922943706514e-09, + "loss": 1.0062, + "step": 10846 + }, + { + "epoch": 0.9782206790819318, + "grad_norm": 2.178834153354097, + "learning_rate": 4.953744893074763e-09, + "loss": 0.9564, + "step": 10847 + }, + { + "epoch": 0.978310862605402, + "grad_norm": 1.9861291656325613, + "learning_rate": 4.912737070789985e-09, + "loss": 0.9018, + "step": 10848 + }, + { + "epoch": 0.9784010461288722, + "grad_norm": 1.72451771614617, + "learning_rate": 4.871899480351604e-09, + "loss": 1.0456, + "step": 10849 + }, + { + "epoch": 0.9784912296523425, + "grad_norm": 2.035511066456348, + "learning_rate": 4.831232125243501e-09, + "loss": 0.9866, + "step": 10850 + }, + { + "epoch": 0.9785814131758128, + "grad_norm": 1.5497182282318493, + "learning_rate": 4.7907350089360086e-09, + "loss": 1.0769, + "step": 10851 + }, + { + "epoch": 0.978671596699283, + "grad_norm": 1.6975572201687548, + "learning_rate": 4.750408134884365e-09, + "loss": 0.8696, + "step": 10852 + }, + { + "epoch": 0.9787617802227533, + "grad_norm": 1.673590448479721, + "learning_rate": 4.710251506529816e-09, + "loss": 0.8831, + "step": 10853 + }, + { + "epoch": 0.9788519637462235, + "grad_norm": 1.7118376711020176, + "learning_rate": 4.6702651272982894e-09, + "loss": 0.9884, + "step": 10854 + }, + { + "epoch": 0.9789421472696939, + "grad_norm": 1.8842572885569477, + "learning_rate": 4.630449000602166e-09, + "loss": 0.9467, + "step": 10855 + }, + { + "epoch": 0.9790323307931641, + "grad_norm": 5.990995007442355, + "learning_rate": 4.590803129838283e-09, + "loss": 1.0363, + "step": 10856 + }, + { + "epoch": 0.9791225143166343, + "grad_norm": 2.2907140196693407, + "learning_rate": 4.551327518389714e-09, + "loss": 0.9788, + "step": 10857 + }, + { + "epoch": 0.9792126978401046, + "grad_norm": 3.7839250166320872, + "learning_rate": 4.512022169624652e-09, + "loss": 0.9659, + "step": 10858 + }, + { + "epoch": 0.9793028813635749, + "grad_norm": 2.3048162909260714, + "learning_rate": 4.472887086896637e-09, + "loss": 0.9475, + "step": 10859 + }, + { + "epoch": 0.9793930648870451, + "grad_norm": 2.0700712243325867, + "learning_rate": 4.433922273545443e-09, + "loss": 0.9227, + "step": 10860 + }, + { + "epoch": 0.9794832484105154, + "grad_norm": 0.6129699937670222, + "learning_rate": 4.395127732895299e-09, + "loss": 0.8176, + "step": 10861 + }, + { + "epoch": 0.9795734319339856, + "grad_norm": 2.797624796520383, + "learning_rate": 4.356503468256445e-09, + "loss": 1.0491, + "step": 10862 + }, + { + "epoch": 0.979663615457456, + "grad_norm": 2.0595912651924366, + "learning_rate": 4.318049482924913e-09, + "loss": 0.9883, + "step": 10863 + }, + { + "epoch": 0.9797537989809262, + "grad_norm": 1.8739718479293292, + "learning_rate": 4.279765780181188e-09, + "loss": 0.9433, + "step": 10864 + }, + { + "epoch": 0.9798439825043964, + "grad_norm": 2.2271911804331523, + "learning_rate": 4.241652363291992e-09, + "loss": 0.9841, + "step": 10865 + }, + { + "epoch": 0.9799341660278668, + "grad_norm": 3.4382451030180694, + "learning_rate": 4.203709235509834e-09, + "loss": 0.9453, + "step": 10866 + }, + { + "epoch": 0.980024349551337, + "grad_norm": 1.5591424344312217, + "learning_rate": 4.165936400071679e-09, + "loss": 0.9953, + "step": 10867 + }, + { + "epoch": 0.9801145330748072, + "grad_norm": 2.0663342526446065, + "learning_rate": 4.12833386020095e-09, + "loss": 0.9244, + "step": 10868 + }, + { + "epoch": 0.9802047165982775, + "grad_norm": 1.9423619456678294, + "learning_rate": 4.090901619105746e-09, + "loss": 1.0173, + "step": 10869 + }, + { + "epoch": 0.9802949001217478, + "grad_norm": 1.9032670116839314, + "learning_rate": 4.053639679980181e-09, + "loss": 0.8676, + "step": 10870 + }, + { + "epoch": 0.980385083645218, + "grad_norm": 1.703613421198762, + "learning_rate": 4.01654804600371e-09, + "loss": 0.9806, + "step": 10871 + }, + { + "epoch": 0.9804752671686883, + "grad_norm": 1.5036088129165357, + "learning_rate": 3.9796267203409114e-09, + "loss": 0.9851, + "step": 10872 + }, + { + "epoch": 0.9805654506921585, + "grad_norm": 1.6381597252825255, + "learning_rate": 3.942875706142379e-09, + "loss": 0.9504, + "step": 10873 + }, + { + "epoch": 0.9806556342156288, + "grad_norm": 1.8243717235654688, + "learning_rate": 3.906295006543825e-09, + "loss": 1.0325, + "step": 10874 + }, + { + "epoch": 0.9807458177390991, + "grad_norm": 1.5323288678550975, + "learning_rate": 3.8698846246665305e-09, + "loss": 0.9517, + "step": 10875 + }, + { + "epoch": 0.9808360012625693, + "grad_norm": 2.3128588629978495, + "learning_rate": 3.833644563617344e-09, + "loss": 0.9777, + "step": 10876 + }, + { + "epoch": 0.9809261847860395, + "grad_norm": 2.002473850912937, + "learning_rate": 3.797574826488237e-09, + "loss": 0.99, + "step": 10877 + }, + { + "epoch": 0.9810163683095099, + "grad_norm": 1.9976999647290303, + "learning_rate": 3.761675416356969e-09, + "loss": 0.9032, + "step": 10878 + }, + { + "epoch": 0.9811065518329801, + "grad_norm": 1.4157510023482323, + "learning_rate": 3.725946336286867e-09, + "loss": 0.8827, + "step": 10879 + }, + { + "epoch": 0.9811967353564504, + "grad_norm": 1.366889485167755, + "learning_rate": 3.6903875893261604e-09, + "loss": 0.9482, + "step": 10880 + }, + { + "epoch": 0.9812869188799206, + "grad_norm": 1.8309819647016914, + "learning_rate": 3.6549991785093105e-09, + "loss": 0.9077, + "step": 10881 + }, + { + "epoch": 0.9813771024033909, + "grad_norm": 1.5805326923101533, + "learning_rate": 3.6197811068554575e-09, + "loss": 0.964, + "step": 10882 + }, + { + "epoch": 0.9814672859268612, + "grad_norm": 1.4257408658931876, + "learning_rate": 3.584733377369975e-09, + "loss": 1.0618, + "step": 10883 + }, + { + "epoch": 0.9815574694503314, + "grad_norm": 2.1671762478515006, + "learning_rate": 3.549855993043138e-09, + "loss": 0.9491, + "step": 10884 + }, + { + "epoch": 0.9816476529738016, + "grad_norm": 1.4346806618127237, + "learning_rate": 3.5151489568507887e-09, + "loss": 0.9593, + "step": 10885 + }, + { + "epoch": 0.981737836497272, + "grad_norm": 1.7959137045214992, + "learning_rate": 3.4806122717545572e-09, + "loss": 0.9901, + "step": 10886 + }, + { + "epoch": 0.9818280200207422, + "grad_norm": 1.542026434838773, + "learning_rate": 3.446245940701198e-09, + "loss": 0.9359, + "step": 10887 + }, + { + "epoch": 0.9819182035442124, + "grad_norm": 5.973110663203307, + "learning_rate": 3.41204996662281e-09, + "loss": 0.9105, + "step": 10888 + }, + { + "epoch": 0.9820083870676828, + "grad_norm": 1.3254641187884406, + "learning_rate": 3.3780243524375028e-09, + "loss": 0.9736, + "step": 10889 + }, + { + "epoch": 0.982098570591153, + "grad_norm": 1.5614933585967155, + "learning_rate": 3.3441691010485107e-09, + "loss": 0.9997, + "step": 10890 + }, + { + "epoch": 0.9821887541146233, + "grad_norm": 1.5651846900405646, + "learning_rate": 3.3104842153444113e-09, + "loss": 1.052, + "step": 10891 + }, + { + "epoch": 0.9822789376380935, + "grad_norm": 1.8755175829810407, + "learning_rate": 3.27696969819935e-09, + "loss": 0.9696, + "step": 10892 + }, + { + "epoch": 0.9823691211615638, + "grad_norm": 1.4967896757714805, + "learning_rate": 3.2436255524732615e-09, + "loss": 0.9297, + "step": 10893 + }, + { + "epoch": 0.9824593046850341, + "grad_norm": 1.5470224945957962, + "learning_rate": 3.210451781010759e-09, + "loss": 0.9561, + "step": 10894 + }, + { + "epoch": 0.9825494882085043, + "grad_norm": 1.4914445075857712, + "learning_rate": 3.1774483866426895e-09, + "loss": 0.838, + "step": 10895 + }, + { + "epoch": 0.9826396717319745, + "grad_norm": 1.5324926200054256, + "learning_rate": 3.144615372185244e-09, + "loss": 0.9397, + "step": 10896 + }, + { + "epoch": 0.9827298552554449, + "grad_norm": 1.6075520236791214, + "learning_rate": 3.1119527404399604e-09, + "loss": 1.0643, + "step": 10897 + }, + { + "epoch": 0.9828200387789151, + "grad_norm": 2.216860974431201, + "learning_rate": 3.0794604941932754e-09, + "loss": 0.9709, + "step": 10898 + }, + { + "epoch": 0.9829102223023853, + "grad_norm": 1.5710943752232565, + "learning_rate": 3.0471386362180827e-09, + "loss": 0.8726, + "step": 10899 + }, + { + "epoch": 0.9830004058258556, + "grad_norm": 1.7885416930311049, + "learning_rate": 3.0149871692719542e-09, + "loss": 1.0218, + "step": 10900 + }, + { + "epoch": 0.9830905893493259, + "grad_norm": 1.6219066374234334, + "learning_rate": 2.9830060960984728e-09, + "loss": 0.9583, + "step": 10901 + }, + { + "epoch": 0.9831807728727961, + "grad_norm": 0.6565717407985214, + "learning_rate": 2.9511954194263442e-09, + "loss": 0.8207, + "step": 10902 + }, + { + "epoch": 0.9832709563962664, + "grad_norm": 1.592852246119947, + "learning_rate": 2.9195551419698426e-09, + "loss": 0.9703, + "step": 10903 + }, + { + "epoch": 0.9833611399197366, + "grad_norm": 3.467380700180312, + "learning_rate": 2.888085266428808e-09, + "loss": 0.9344, + "step": 10904 + }, + { + "epoch": 0.983451323443207, + "grad_norm": 1.7087910046851462, + "learning_rate": 2.8567857954882037e-09, + "loss": 0.9277, + "step": 10905 + }, + { + "epoch": 0.9835415069666772, + "grad_norm": 1.8585250578826178, + "learning_rate": 2.82565673181856e-09, + "loss": 1.0281, + "step": 10906 + }, + { + "epoch": 0.9836316904901474, + "grad_norm": 1.4411519971283058, + "learning_rate": 2.7946980780764184e-09, + "loss": 0.9943, + "step": 10907 + }, + { + "epoch": 0.9837218740136177, + "grad_norm": 1.8666347180198293, + "learning_rate": 2.763909836903e-09, + "loss": 1.0309, + "step": 10908 + }, + { + "epoch": 0.983812057537088, + "grad_norm": 1.6284088180262009, + "learning_rate": 2.7332920109255364e-09, + "loss": 0.8896, + "step": 10909 + }, + { + "epoch": 0.9839022410605582, + "grad_norm": 1.5546828226757137, + "learning_rate": 2.702844602756382e-09, + "loss": 0.9534, + "step": 10910 + }, + { + "epoch": 0.9839924245840285, + "grad_norm": 2.27118091081375, + "learning_rate": 2.6725676149936814e-09, + "loss": 0.9498, + "step": 10911 + }, + { + "epoch": 0.9840826081074988, + "grad_norm": 1.8370563750070992, + "learning_rate": 2.642461050220479e-09, + "loss": 1.0412, + "step": 10912 + }, + { + "epoch": 0.984172791630969, + "grad_norm": 1.3839936465244322, + "learning_rate": 2.612524911005609e-09, + "loss": 0.9541, + "step": 10913 + }, + { + "epoch": 0.9842629751544393, + "grad_norm": 1.4921088073112614, + "learning_rate": 2.582759199903917e-09, + "loss": 0.8527, + "step": 10914 + }, + { + "epoch": 0.9843531586779095, + "grad_norm": 1.8579242079174605, + "learning_rate": 2.553163919454704e-09, + "loss": 0.9324, + "step": 10915 + }, + { + "epoch": 0.9844433422013799, + "grad_norm": 2.0324566118364746, + "learning_rate": 2.523739072183506e-09, + "loss": 0.9551, + "step": 10916 + }, + { + "epoch": 0.9845335257248501, + "grad_norm": 1.8252228744491326, + "learning_rate": 2.4944846606007597e-09, + "loss": 0.9332, + "step": 10917 + }, + { + "epoch": 0.9846237092483203, + "grad_norm": 0.6105712007153923, + "learning_rate": 2.46540068720269e-09, + "loss": 0.8513, + "step": 10918 + }, + { + "epoch": 0.9847138927717906, + "grad_norm": 1.4123576725508726, + "learning_rate": 2.4364871544708674e-09, + "loss": 1.0104, + "step": 10919 + }, + { + "epoch": 0.9848040762952609, + "grad_norm": 2.3236190919903086, + "learning_rate": 2.4077440648726523e-09, + "loss": 0.9277, + "step": 10920 + }, + { + "epoch": 0.9848942598187311, + "grad_norm": 1.7790083583230165, + "learning_rate": 2.379171420860082e-09, + "loss": 1.0346, + "step": 10921 + }, + { + "epoch": 0.9849844433422014, + "grad_norm": 1.8389135386276914, + "learning_rate": 2.3507692248714296e-09, + "loss": 0.8968, + "step": 10922 + }, + { + "epoch": 0.9850746268656716, + "grad_norm": 1.824799798409544, + "learning_rate": 2.322537479330089e-09, + "loss": 0.988, + "step": 10923 + }, + { + "epoch": 0.9851648103891419, + "grad_norm": 1.6185586617403618, + "learning_rate": 2.2944761866450223e-09, + "loss": 1.0474, + "step": 10924 + }, + { + "epoch": 0.9852549939126122, + "grad_norm": 1.6694258103659227, + "learning_rate": 2.266585349210315e-09, + "loss": 0.9561, + "step": 10925 + }, + { + "epoch": 0.9853451774360824, + "grad_norm": 2.116023345364181, + "learning_rate": 2.2388649694060623e-09, + "loss": 0.9282, + "step": 10926 + }, + { + "epoch": 0.9854353609595526, + "grad_norm": 2.639968158356504, + "learning_rate": 2.211315049597262e-09, + "loss": 1.0321, + "step": 10927 + }, + { + "epoch": 0.985525544483023, + "grad_norm": 1.713789213748875, + "learning_rate": 2.1839355921349224e-09, + "loss": 1.0068, + "step": 10928 + }, + { + "epoch": 0.9856157280064932, + "grad_norm": 2.181570791357984, + "learning_rate": 2.156726599354952e-09, + "loss": 0.9644, + "step": 10929 + }, + { + "epoch": 0.9857059115299635, + "grad_norm": 1.8515147271411472, + "learning_rate": 2.129688073578828e-09, + "loss": 0.9964, + "step": 10930 + }, + { + "epoch": 0.9857960950534337, + "grad_norm": 3.8722565075818847, + "learning_rate": 2.1028200171142597e-09, + "loss": 1.1072, + "step": 10931 + }, + { + "epoch": 0.985886278576904, + "grad_norm": 2.967999975929363, + "learning_rate": 2.076122432253191e-09, + "loss": 0.9545, + "step": 10932 + }, + { + "epoch": 0.9859764621003743, + "grad_norm": 2.0588669423727373, + "learning_rate": 2.0495953212738005e-09, + "loss": 1.0096, + "step": 10933 + }, + { + "epoch": 0.9860666456238445, + "grad_norm": 1.4529715836457642, + "learning_rate": 2.0232386864396102e-09, + "loss": 1.0557, + "step": 10934 + }, + { + "epoch": 0.9861568291473148, + "grad_norm": 1.500236174746193, + "learning_rate": 1.9970525299992656e-09, + "loss": 0.9977, + "step": 10935 + }, + { + "epoch": 0.9862470126707851, + "grad_norm": 2.249783505514651, + "learning_rate": 1.9710368541874245e-09, + "loss": 0.9359, + "step": 10936 + }, + { + "epoch": 0.9863371961942553, + "grad_norm": 0.5838586167895564, + "learning_rate": 1.945191661223644e-09, + "loss": 0.8229, + "step": 10937 + }, + { + "epoch": 0.9864273797177255, + "grad_norm": 2.3985464462735844, + "learning_rate": 1.9195169533132714e-09, + "loss": 1.0204, + "step": 10938 + }, + { + "epoch": 0.9865175632411959, + "grad_norm": 0.581367088132805, + "learning_rate": 1.894012732646999e-09, + "loss": 0.8299, + "step": 10939 + }, + { + "epoch": 0.9866077467646661, + "grad_norm": 1.901520349165479, + "learning_rate": 1.8686790014010854e-09, + "loss": 0.9239, + "step": 10940 + }, + { + "epoch": 0.9866979302881363, + "grad_norm": 0.6043856130889063, + "learning_rate": 1.8435157617369134e-09, + "loss": 0.7961, + "step": 10941 + }, + { + "epoch": 0.9867881138116066, + "grad_norm": 4.330241102593688, + "learning_rate": 1.818523015801876e-09, + "loss": 1.1132, + "step": 10942 + }, + { + "epoch": 0.9868782973350769, + "grad_norm": 1.8546222241427115, + "learning_rate": 1.7937007657282677e-09, + "loss": 0.9314, + "step": 10943 + }, + { + "epoch": 0.9869684808585472, + "grad_norm": 1.6767220512757912, + "learning_rate": 1.7690490136341718e-09, + "loss": 0.9147, + "step": 10944 + }, + { + "epoch": 0.9870586643820174, + "grad_norm": 1.449256341246162, + "learning_rate": 1.744567761622795e-09, + "loss": 0.9706, + "step": 10945 + }, + { + "epoch": 0.9871488479054876, + "grad_norm": 1.5043943727685734, + "learning_rate": 1.7202570117831327e-09, + "loss": 0.9661, + "step": 10946 + }, + { + "epoch": 0.987239031428958, + "grad_norm": 1.9618604091601184, + "learning_rate": 1.696116766189526e-09, + "loss": 0.9464, + "step": 10947 + }, + { + "epoch": 0.9873292149524282, + "grad_norm": 1.4819336741917195, + "learning_rate": 1.6721470269021042e-09, + "loss": 1.0176, + "step": 10948 + }, + { + "epoch": 0.9874193984758984, + "grad_norm": 2.1935665477764577, + "learning_rate": 1.6483477959654546e-09, + "loss": 0.9428, + "step": 10949 + }, + { + "epoch": 0.9875095819993687, + "grad_norm": 1.5046322345982794, + "learning_rate": 1.6247190754106187e-09, + "loss": 0.9564, + "step": 10950 + }, + { + "epoch": 0.987599765522839, + "grad_norm": 1.5819164846731293, + "learning_rate": 1.6012608672537619e-09, + "loss": 0.9378, + "step": 10951 + }, + { + "epoch": 0.9876899490463092, + "grad_norm": 2.294835125521312, + "learning_rate": 1.5779731734963942e-09, + "loss": 0.9811, + "step": 10952 + }, + { + "epoch": 0.9877801325697795, + "grad_norm": 2.62852169131972, + "learning_rate": 1.5548559961253705e-09, + "loss": 1.0355, + "step": 10953 + }, + { + "epoch": 0.9878703160932497, + "grad_norm": 1.4957785809767152, + "learning_rate": 1.5319093371135573e-09, + "loss": 0.8998, + "step": 10954 + }, + { + "epoch": 0.9879604996167201, + "grad_norm": 9.401513209189178, + "learning_rate": 1.5091331984184997e-09, + "loss": 1.062, + "step": 10955 + }, + { + "epoch": 0.9880506831401903, + "grad_norm": 3.0429635788527145, + "learning_rate": 1.486527581983754e-09, + "loss": 0.9749, + "step": 10956 + }, + { + "epoch": 0.9881408666636605, + "grad_norm": 1.8375822451084582, + "learning_rate": 1.4640924897382223e-09, + "loss": 1.0329, + "step": 10957 + }, + { + "epoch": 0.9882310501871309, + "grad_norm": 1.5377241017841945, + "learning_rate": 1.4418279235961506e-09, + "loss": 1.0131, + "step": 10958 + }, + { + "epoch": 0.9883212337106011, + "grad_norm": 1.7720472625474482, + "learning_rate": 1.4197338854573526e-09, + "loss": 0.878, + "step": 10959 + }, + { + "epoch": 0.9884114172340713, + "grad_norm": 2.08382157008896, + "learning_rate": 1.3978103772067651e-09, + "loss": 0.9289, + "step": 10960 + }, + { + "epoch": 0.9885016007575416, + "grad_norm": 1.896595693320727, + "learning_rate": 1.3760574007153358e-09, + "loss": 0.8974, + "step": 10961 + }, + { + "epoch": 0.9885917842810119, + "grad_norm": 2.749577749953536, + "learning_rate": 1.3544749578389137e-09, + "loss": 1.0547, + "step": 10962 + }, + { + "epoch": 0.9886819678044821, + "grad_norm": 1.632247315927952, + "learning_rate": 1.3330630504189143e-09, + "loss": 1.0214, + "step": 10963 + }, + { + "epoch": 0.9887721513279524, + "grad_norm": 2.2365124489062316, + "learning_rate": 1.3118216802827652e-09, + "loss": 0.9408, + "step": 10964 + }, + { + "epoch": 0.9888623348514226, + "grad_norm": 1.7670696728565198, + "learning_rate": 1.2907508492425722e-09, + "loss": 0.9429, + "step": 10965 + }, + { + "epoch": 0.988952518374893, + "grad_norm": 4.313770715116043, + "learning_rate": 1.2698505590962305e-09, + "loss": 0.9379, + "step": 10966 + }, + { + "epoch": 0.9890427018983632, + "grad_norm": 1.776945367118184, + "learning_rate": 1.2491208116272022e-09, + "loss": 0.968, + "step": 10967 + }, + { + "epoch": 0.9891328854218334, + "grad_norm": 1.76799918370638, + "learning_rate": 1.2285616086040728e-09, + "loss": 1.0019, + "step": 10968 + }, + { + "epoch": 0.9892230689453037, + "grad_norm": 2.1270699236898896, + "learning_rate": 1.2081729517812167e-09, + "loss": 1.037, + "step": 10969 + }, + { + "epoch": 0.989313252468774, + "grad_norm": 1.4736118605876762, + "learning_rate": 1.1879548428983533e-09, + "loss": 0.9682, + "step": 10970 + }, + { + "epoch": 0.9894034359922442, + "grad_norm": 1.6824880983291641, + "learning_rate": 1.167907283680547e-09, + "loss": 0.9231, + "step": 10971 + }, + { + "epoch": 0.9894936195157145, + "grad_norm": 2.9317305701662564, + "learning_rate": 1.1480302758382077e-09, + "loss": 0.9162, + "step": 10972 + }, + { + "epoch": 0.9895838030391847, + "grad_norm": 1.5282469905534715, + "learning_rate": 1.1283238210675338e-09, + "loss": 0.9656, + "step": 10973 + }, + { + "epoch": 0.989673986562655, + "grad_norm": 2.2951796106934674, + "learning_rate": 1.1087879210498474e-09, + "loss": 0.9013, + "step": 10974 + }, + { + "epoch": 0.9897641700861253, + "grad_norm": 1.5090427419980934, + "learning_rate": 1.0894225774522592e-09, + "loss": 1.0223, + "step": 10975 + }, + { + "epoch": 0.9898543536095955, + "grad_norm": 0.6496456794434795, + "learning_rate": 1.070227791927003e-09, + "loss": 0.8318, + "step": 10976 + }, + { + "epoch": 0.9899445371330657, + "grad_norm": 1.9497176699559482, + "learning_rate": 1.0512035661118802e-09, + "loss": 1.0743, + "step": 10977 + }, + { + "epoch": 0.9900347206565361, + "grad_norm": 1.6765073256999536, + "learning_rate": 1.0323499016300364e-09, + "loss": 0.9955, + "step": 10978 + }, + { + "epoch": 0.9901249041800063, + "grad_norm": 2.1396302745192695, + "learning_rate": 1.013666800090407e-09, + "loss": 0.9281, + "step": 10979 + }, + { + "epoch": 0.9902150877034765, + "grad_norm": 2.224427923316252, + "learning_rate": 9.951542630870502e-10, + "loss": 0.9606, + "step": 10980 + }, + { + "epoch": 0.9903052712269468, + "grad_norm": 2.8609404734205945, + "learning_rate": 9.768122921995915e-10, + "loss": 1.0109, + "step": 10981 + }, + { + "epoch": 0.9903954547504171, + "grad_norm": 2.1743786329835, + "learning_rate": 9.58640888992779e-10, + "loss": 0.9181, + "step": 10982 + }, + { + "epoch": 0.9904856382738874, + "grad_norm": 1.6852966335008814, + "learning_rate": 9.40640055017594e-10, + "loss": 0.8908, + "step": 10983 + }, + { + "epoch": 0.9905758217973576, + "grad_norm": 2.183597571518067, + "learning_rate": 9.228097918094757e-10, + "loss": 0.9738, + "step": 10984 + }, + { + "epoch": 0.9906660053208279, + "grad_norm": 1.7906078344457685, + "learning_rate": 9.051501008900952e-10, + "loss": 0.9402, + "step": 10985 + }, + { + "epoch": 0.9907561888442982, + "grad_norm": 3.575991865093357, + "learning_rate": 8.876609837662475e-10, + "loss": 0.9912, + "step": 10986 + }, + { + "epoch": 0.9908463723677684, + "grad_norm": 2.22103903451803, + "learning_rate": 8.70342441930294e-10, + "loss": 1.0603, + "step": 10987 + }, + { + "epoch": 0.9909365558912386, + "grad_norm": 1.9043700531812968, + "learning_rate": 8.531944768594979e-10, + "loss": 0.9451, + "step": 10988 + }, + { + "epoch": 0.991026739414709, + "grad_norm": 4.247188984854723, + "learning_rate": 8.362170900175769e-10, + "loss": 1.0556, + "step": 10989 + }, + { + "epoch": 0.9911169229381792, + "grad_norm": 1.622453307903844, + "learning_rate": 8.194102828527061e-10, + "loss": 0.9293, + "step": 10990 + }, + { + "epoch": 0.9912071064616494, + "grad_norm": 1.785644397730375, + "learning_rate": 8.027740567992936e-10, + "loss": 1.0171, + "step": 10991 + }, + { + "epoch": 0.9912972899851197, + "grad_norm": 1.575820221842191, + "learning_rate": 7.863084132766484e-10, + "loss": 0.9712, + "step": 10992 + }, + { + "epoch": 0.99138747350859, + "grad_norm": 1.823054789805309, + "learning_rate": 7.700133536896469e-10, + "loss": 0.9916, + "step": 10993 + }, + { + "epoch": 0.9914776570320603, + "grad_norm": 1.5035554551051777, + "learning_rate": 7.538888794287324e-10, + "loss": 0.9395, + "step": 10994 + }, + { + "epoch": 0.9915678405555305, + "grad_norm": 2.101400830031199, + "learning_rate": 7.379349918696931e-10, + "loss": 0.9971, + "step": 10995 + }, + { + "epoch": 0.9916580240790007, + "grad_norm": 1.7445664521427013, + "learning_rate": 7.221516923738846e-10, + "loss": 0.9481, + "step": 10996 + }, + { + "epoch": 0.9917482076024711, + "grad_norm": 13.60424119228885, + "learning_rate": 7.065389822880075e-10, + "loss": 1.0789, + "step": 10997 + }, + { + "epoch": 0.9918383911259413, + "grad_norm": 2.555461428009898, + "learning_rate": 6.910968629443292e-10, + "loss": 1.051, + "step": 10998 + }, + { + "epoch": 0.9919285746494115, + "grad_norm": 1.6728017502298052, + "learning_rate": 6.758253356602406e-10, + "loss": 0.968, + "step": 10999 + }, + { + "epoch": 0.9920187581728818, + "grad_norm": 1.691308280803388, + "learning_rate": 6.607244017389213e-10, + "loss": 1.0674, + "step": 11000 + }, + { + "epoch": 0.9921089416963521, + "grad_norm": 2.846742334216329, + "learning_rate": 6.457940624686742e-10, + "loss": 0.9885, + "step": 11001 + }, + { + "epoch": 0.9921991252198223, + "grad_norm": 1.8668323641985873, + "learning_rate": 6.310343191238132e-10, + "loss": 1.0582, + "step": 11002 + }, + { + "epoch": 0.9922893087432926, + "grad_norm": 2.6499690575862487, + "learning_rate": 6.164451729635534e-10, + "loss": 0.8687, + "step": 11003 + }, + { + "epoch": 0.9923794922667628, + "grad_norm": 1.5442958098563382, + "learning_rate": 6.020266252324546e-10, + "loss": 0.8429, + "step": 11004 + }, + { + "epoch": 0.9924696757902332, + "grad_norm": 1.6802704265336486, + "learning_rate": 5.877786771610882e-10, + "loss": 0.946, + "step": 11005 + }, + { + "epoch": 0.9925598593137034, + "grad_norm": 0.7130647211567686, + "learning_rate": 5.737013299651483e-10, + "loss": 0.8832, + "step": 11006 + }, + { + "epoch": 0.9926500428371736, + "grad_norm": 1.6422665331085673, + "learning_rate": 5.597945848458963e-10, + "loss": 0.9784, + "step": 11007 + }, + { + "epoch": 0.992740226360644, + "grad_norm": 1.7697535495197059, + "learning_rate": 5.460584429894944e-10, + "loss": 0.9484, + "step": 11008 + }, + { + "epoch": 0.9928304098841142, + "grad_norm": 1.6847334188473786, + "learning_rate": 5.32492905568338e-10, + "loss": 0.9754, + "step": 11009 + }, + { + "epoch": 0.9929205934075844, + "grad_norm": 1.6362368072141107, + "learning_rate": 5.190979737399459e-10, + "loss": 1.0043, + "step": 11010 + }, + { + "epoch": 0.9930107769310547, + "grad_norm": 1.9868738855657004, + "learning_rate": 5.058736486469594e-10, + "loss": 0.8672, + "step": 11011 + }, + { + "epoch": 0.993100960454525, + "grad_norm": 2.0165023668475346, + "learning_rate": 4.928199314180314e-10, + "loss": 1.008, + "step": 11012 + }, + { + "epoch": 0.9931911439779952, + "grad_norm": 2.396484573973606, + "learning_rate": 4.799368231669376e-10, + "loss": 0.8899, + "step": 11013 + }, + { + "epoch": 0.9932813275014655, + "grad_norm": 1.545873529711888, + "learning_rate": 4.672243249927988e-10, + "loss": 0.9608, + "step": 11014 + }, + { + "epoch": 0.9933715110249357, + "grad_norm": 3.095230981098984, + "learning_rate": 4.546824379803027e-10, + "loss": 1.0037, + "step": 11015 + }, + { + "epoch": 0.993461694548406, + "grad_norm": 1.9218096838295946, + "learning_rate": 4.4231116319970454e-10, + "loss": 1.0006, + "step": 11016 + }, + { + "epoch": 0.9935518780718763, + "grad_norm": 1.9205318812927235, + "learning_rate": 4.3011050170660423e-10, + "loss": 0.9773, + "step": 11017 + }, + { + "epoch": 0.9936420615953465, + "grad_norm": 2.9322802654652262, + "learning_rate": 4.18080454542169e-10, + "loss": 0.966, + "step": 11018 + }, + { + "epoch": 0.9937322451188167, + "grad_norm": 0.6614418603442841, + "learning_rate": 4.0622102273246694e-10, + "loss": 0.8128, + "step": 11019 + }, + { + "epoch": 0.9938224286422871, + "grad_norm": 1.4327899228046745, + "learning_rate": 3.945322072897994e-10, + "loss": 1.0355, + "step": 11020 + }, + { + "epoch": 0.9939126121657573, + "grad_norm": 1.9601531264765617, + "learning_rate": 3.830140092111467e-10, + "loss": 1.0317, + "step": 11021 + }, + { + "epoch": 0.9940027956892276, + "grad_norm": 1.629641799066522, + "learning_rate": 3.7166642947972225e-10, + "loss": 0.9676, + "step": 11022 + }, + { + "epoch": 0.9940929792126978, + "grad_norm": 2.014987933593667, + "learning_rate": 3.604894690634186e-10, + "loss": 0.9497, + "step": 11023 + }, + { + "epoch": 0.9941831627361681, + "grad_norm": 1.714130018464056, + "learning_rate": 3.494831289161393e-10, + "loss": 1.0588, + "step": 11024 + }, + { + "epoch": 0.9942733462596384, + "grad_norm": 1.666427507725442, + "learning_rate": 3.3864740997668897e-10, + "loss": 1.0115, + "step": 11025 + }, + { + "epoch": 0.9943635297831086, + "grad_norm": 1.5731901305836946, + "learning_rate": 3.279823131701054e-10, + "loss": 0.9995, + "step": 11026 + }, + { + "epoch": 0.9944537133065788, + "grad_norm": 1.4087659158289931, + "learning_rate": 3.1748783940610536e-10, + "loss": 0.9317, + "step": 11027 + }, + { + "epoch": 0.9945438968300492, + "grad_norm": 1.944818150469914, + "learning_rate": 3.071639895801947e-10, + "loss": 0.9345, + "step": 11028 + }, + { + "epoch": 0.9946340803535194, + "grad_norm": 4.201262447812027, + "learning_rate": 2.9701076457322447e-10, + "loss": 1.0364, + "step": 11029 + }, + { + "epoch": 0.9947242638769896, + "grad_norm": 1.7860733447804378, + "learning_rate": 2.870281652513906e-10, + "loss": 1.0138, + "step": 11030 + }, + { + "epoch": 0.99481444740046, + "grad_norm": 3.3719761335675438, + "learning_rate": 2.772161924669003e-10, + "loss": 0.9614, + "step": 11031 + }, + { + "epoch": 0.9949046309239302, + "grad_norm": 1.7567327955757157, + "learning_rate": 2.6757484705641765e-10, + "loss": 0.9592, + "step": 11032 + }, + { + "epoch": 0.9949948144474005, + "grad_norm": 1.7352500976026073, + "learning_rate": 2.58104129843062e-10, + "loss": 0.9729, + "step": 11033 + }, + { + "epoch": 0.9950849979708707, + "grad_norm": 2.2304652225713677, + "learning_rate": 2.4880404163463154e-10, + "loss": 0.9301, + "step": 11034 + }, + { + "epoch": 0.995175181494341, + "grad_norm": 2.835439759178404, + "learning_rate": 2.3967458322471377e-10, + "loss": 0.9648, + "step": 11035 + }, + { + "epoch": 0.9952653650178113, + "grad_norm": 1.6220149720344088, + "learning_rate": 2.307157553922412e-10, + "loss": 0.9874, + "step": 11036 + }, + { + "epoch": 0.9953555485412815, + "grad_norm": 1.428142692740129, + "learning_rate": 2.2192755890193538e-10, + "loss": 1.0216, + "step": 11037 + }, + { + "epoch": 0.9954457320647517, + "grad_norm": 0.622305330110098, + "learning_rate": 2.133099945034189e-10, + "loss": 0.7815, + "step": 11038 + }, + { + "epoch": 0.9955359155882221, + "grad_norm": 1.8481429831978324, + "learning_rate": 2.048630629318815e-10, + "loss": 0.9943, + "step": 11039 + }, + { + "epoch": 0.9956260991116923, + "grad_norm": 1.76701846878512, + "learning_rate": 1.965867649080799e-10, + "loss": 0.9713, + "step": 11040 + }, + { + "epoch": 0.9957162826351625, + "grad_norm": 0.6095086953635821, + "learning_rate": 1.8848110113856008e-10, + "loss": 0.8001, + "step": 11041 + }, + { + "epoch": 0.9958064661586328, + "grad_norm": 2.0146951822542984, + "learning_rate": 1.8054607231454687e-10, + "loss": 1.0143, + "step": 11042 + }, + { + "epoch": 0.9958966496821031, + "grad_norm": 1.6157088591838764, + "learning_rate": 1.7278167911327635e-10, + "loss": 0.9278, + "step": 11043 + }, + { + "epoch": 0.9959868332055734, + "grad_norm": 1.764906542954028, + "learning_rate": 1.6518792219710753e-10, + "loss": 1.0071, + "step": 11044 + }, + { + "epoch": 0.9960770167290436, + "grad_norm": 1.6055447822628812, + "learning_rate": 1.5776480221418865e-10, + "loss": 0.9816, + "step": 11045 + }, + { + "epoch": 0.9961672002525138, + "grad_norm": 2.454704990725301, + "learning_rate": 1.505123197977909e-10, + "loss": 0.9749, + "step": 11046 + }, + { + "epoch": 0.9962573837759842, + "grad_norm": 1.6736898109149139, + "learning_rate": 1.4343047556675258e-10, + "loss": 0.9089, + "step": 11047 + }, + { + "epoch": 0.9963475672994544, + "grad_norm": 2.0764961455961775, + "learning_rate": 1.3651927012503506e-10, + "loss": 0.9764, + "step": 11048 + }, + { + "epoch": 0.9964377508229246, + "grad_norm": 1.5878726071394726, + "learning_rate": 1.297787040630549e-10, + "loss": 0.9816, + "step": 11049 + }, + { + "epoch": 0.9965279343463949, + "grad_norm": 2.6925818722885757, + "learning_rate": 1.2320877795524153e-10, + "loss": 0.9969, + "step": 11050 + }, + { + "epoch": 0.9966181178698652, + "grad_norm": 1.7449078725197908, + "learning_rate": 1.1680949236247962e-10, + "loss": 0.9611, + "step": 11051 + }, + { + "epoch": 0.9967083013933354, + "grad_norm": 1.5214846100519315, + "learning_rate": 1.1058084783099886e-10, + "loss": 0.982, + "step": 11052 + }, + { + "epoch": 0.9967984849168057, + "grad_norm": 1.8188943864592626, + "learning_rate": 1.0452284489170793e-10, + "loss": 0.9875, + "step": 11053 + }, + { + "epoch": 0.996888668440276, + "grad_norm": 1.5791822612197115, + "learning_rate": 9.86354840621928e-11, + "loss": 0.9624, + "step": 11054 + }, + { + "epoch": 0.9969788519637462, + "grad_norm": 1.5577261668380777, + "learning_rate": 9.291876584427427e-11, + "loss": 0.851, + "step": 11055 + }, + { + "epoch": 0.9970690354872165, + "grad_norm": 1.8076711983548597, + "learning_rate": 8.737269072578435e-11, + "loss": 0.9488, + "step": 11056 + }, + { + "epoch": 0.9971592190106867, + "grad_norm": 1.492493283561789, + "learning_rate": 8.199725918012212e-11, + "loss": 0.9709, + "step": 11057 + }, + { + "epoch": 0.9972494025341571, + "grad_norm": 2.019409648252086, + "learning_rate": 7.679247166603175e-11, + "loss": 0.9823, + "step": 11058 + }, + { + "epoch": 0.9973395860576273, + "grad_norm": 2.1957797710489997, + "learning_rate": 7.17583286273804e-11, + "loss": 0.9832, + "step": 11059 + }, + { + "epoch": 0.9974297695810975, + "grad_norm": 1.434634711661875, + "learning_rate": 6.689483049360233e-11, + "loss": 0.9952, + "step": 11060 + }, + { + "epoch": 0.9975199531045678, + "grad_norm": 2.0953002428667755, + "learning_rate": 6.220197768014302e-11, + "loss": 0.9568, + "step": 11061 + }, + { + "epoch": 0.9976101366280381, + "grad_norm": 1.6491912275698606, + "learning_rate": 5.7679770587126806e-11, + "loss": 0.9956, + "step": 11062 + }, + { + "epoch": 0.9977003201515083, + "grad_norm": 0.6237782218908913, + "learning_rate": 5.33282096002452e-11, + "loss": 0.9019, + "step": 11063 + }, + { + "epoch": 0.9977905036749786, + "grad_norm": 1.6361068273609487, + "learning_rate": 4.914729509120086e-11, + "loss": 0.9425, + "step": 11064 + }, + { + "epoch": 0.9978806871984488, + "grad_norm": 1.4880993348928098, + "learning_rate": 4.513702741637537e-11, + "loss": 0.8981, + "step": 11065 + }, + { + "epoch": 0.9979708707219191, + "grad_norm": 1.544865468535106, + "learning_rate": 4.129740691816153e-11, + "loss": 1.0376, + "step": 11066 + }, + { + "epoch": 0.9980610542453894, + "grad_norm": 1.5469832719614771, + "learning_rate": 3.762843392429715e-11, + "loss": 0.9955, + "step": 11067 + }, + { + "epoch": 0.9981512377688596, + "grad_norm": 1.742835912083801, + "learning_rate": 3.413010874742106e-11, + "loss": 0.9576, + "step": 11068 + }, + { + "epoch": 0.9982414212923298, + "grad_norm": 1.9384989470527472, + "learning_rate": 3.080243168618324e-11, + "loss": 0.9708, + "step": 11069 + }, + { + "epoch": 0.9983316048158002, + "grad_norm": 2.5277767134816744, + "learning_rate": 2.7645403024800783e-11, + "loss": 0.9227, + "step": 11070 + }, + { + "epoch": 0.9984217883392704, + "grad_norm": 1.5375462532839619, + "learning_rate": 2.4659023032391756e-11, + "loss": 1.0475, + "step": 11071 + }, + { + "epoch": 0.9985119718627407, + "grad_norm": 1.3656885704066708, + "learning_rate": 2.1843291963863364e-11, + "loss": 1.0214, + "step": 11072 + }, + { + "epoch": 0.9986021553862109, + "grad_norm": 1.6793024375807426, + "learning_rate": 1.9198210059245822e-11, + "loss": 0.9945, + "step": 11073 + }, + { + "epoch": 0.9986923389096812, + "grad_norm": 1.745677241203695, + "learning_rate": 1.672377754458054e-11, + "loss": 0.9783, + "step": 11074 + }, + { + "epoch": 0.9987825224331515, + "grad_norm": 1.5507421224351907, + "learning_rate": 1.4419994630809895e-11, + "loss": 0.9691, + "step": 11075 + }, + { + "epoch": 0.9988727059566217, + "grad_norm": 1.8888057436439276, + "learning_rate": 1.2286861514443358e-11, + "loss": 0.928, + "step": 11076 + }, + { + "epoch": 0.998962889480092, + "grad_norm": 1.3578795176166378, + "learning_rate": 1.0324378377779553e-11, + "loss": 1.014, + "step": 11077 + }, + { + "epoch": 0.9990530730035623, + "grad_norm": 1.5242469397850906, + "learning_rate": 8.532545388018064e-12, + "loss": 0.9739, + "step": 11078 + }, + { + "epoch": 0.9991432565270325, + "grad_norm": 1.5119341714041878, + "learning_rate": 6.911362697925582e-12, + "loss": 0.8115, + "step": 11079 + }, + { + "epoch": 0.9992334400505027, + "grad_norm": 2.1985287168187635, + "learning_rate": 5.46083044605794e-12, + "loss": 0.9458, + "step": 11080 + }, + { + "epoch": 0.9993236235739731, + "grad_norm": 1.564178023516412, + "learning_rate": 4.1809487563160276e-12, + "loss": 1.0262, + "step": 11081 + }, + { + "epoch": 0.9994138070974433, + "grad_norm": 1.795595260676577, + "learning_rate": 3.0717177375017e-12, + "loss": 0.979, + "step": 11082 + }, + { + "epoch": 0.9995039906209136, + "grad_norm": 1.8731547494638137, + "learning_rate": 2.1331374846500495e-12, + "loss": 0.9738, + "step": 11083 + }, + { + "epoch": 0.9995941741443838, + "grad_norm": 1.8372216461481166, + "learning_rate": 1.3652080774750885e-12, + "loss": 1.0712, + "step": 11084 + }, + { + "epoch": 0.9996843576678541, + "grad_norm": 6.4318188533421115, + "learning_rate": 7.679295817020204e-13, + "loss": 0.8926, + "step": 11085 + }, + { + "epoch": 0.9997745411913244, + "grad_norm": 2.0557480312422878, + "learning_rate": 3.413020484011042e-13, + "loss": 0.954, + "step": 11086 + }, + { + "epoch": 0.9998647247147946, + "grad_norm": 1.9425940458687394, + "learning_rate": 8.53255139876552e-14, + "loss": 1.0194, + "step": 11087 + }, + { + "epoch": 0.9999549082382648, + "grad_norm": 2.0190556628417453, + "learning_rate": 0.0, + "loss": 0.9199, + "step": 11088 + }, + { + "epoch": 0.9999549082382648, + "step": 11088, + "total_flos": 1.353872062422188e+18, + "train_loss": 0.9886689228402158, + "train_runtime": 149279.455, + "train_samples_per_second": 4.457, + "train_steps_per_second": 0.074 + } + ], + "logging_steps": 1.0, + "max_steps": 11088, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.353872062422188e+18, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +} diff --git a/sft_full/smoe_perturbed/training_args.bin b/sft_full/smoe_perturbed/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..a52996e0e58cf1bbd1ba45e6213a6c65aeecfb25 --- /dev/null +++ b/sft_full/smoe_perturbed/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08aaf4412961001d21ae464e0d5be33c4cc8cb307123de1b06ef7f4872692493 +size 8184 diff --git a/sft_full/smoe_sigmoidgating/added_tokens.json b/sft_full/smoe_sigmoidgating/added_tokens.json new file mode 100644 index 0000000000000000000000000000000000000000..c9d3d3a1b74d87e381e471f7b33784015d2dc0ea --- /dev/null +++ b/sft_full/smoe_sigmoidgating/added_tokens.json @@ -0,0 +1,13 @@ +{ + "<|assistant|>": 32001, + "<|endoftext|>": 32000, + "<|end|>": 32007, + "<|placeholder1|>": 32002, + "<|placeholder2|>": 32003, + "<|placeholder3|>": 32004, + "<|placeholder4|>": 32005, + "<|placeholder5|>": 32008, + "<|placeholder6|>": 32009, + "<|system|>": 32006, + "<|user|>": 32010 +} diff --git a/sft_full/smoe_sigmoidgating/config.json b/sft_full/smoe_sigmoidgating/config.json new file mode 100644 index 0000000000000000000000000000000000000000..df93afba22e4e6ef6ce6d8589c80d0d60baa0503 --- /dev/null +++ b/sft_full/smoe_sigmoidgating/config.json @@ -0,0 +1,168 @@ +{ + "_name_or_path": "/cm/archive/namnv78/checkpoints/phi35-siglip224/pft", + "architectures": [ + "LlavaPhiForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "auto_map": { + "AutoConfig": "configuration_phi3.Phi3Config", + "AutoModelForCausalLM": "modeling_phi3.Phi3ForCausalLM" + }, + "balance_loss_coef": 0.1, + "bos_token_id": 1, + "clip_smoe": true, + "dropout": false, + "embd_pdrop": 0.0, + "eos_token_id": 32000, + "freeze_mm_mlp_adapter": false, + "hidden_act": "silu", + "hidden_size": 3072, + "image_aspect_ratio": "pad", + "initializer_range": 0.02, + "intermediate_size": 8192, + "local_rank": 0, + "max_position_embeddings": 131072, + "mlp_smoe": true, + "mm_hidden_size": 1152, + "mm_patch_merge_type": "flat", + "mm_projector_lr": null, + "mm_projector_type": "moe", + "mm_use_im_patch_token": false, + "mm_use_im_start_end": false, + "mm_vision_select_feature": "patch", + "mm_vision_select_layer": -2, + "mm_vision_tower": "google/siglip-so400m-patch14-224", + "model_type": "llava_phi", + "moe_name": "smoe_sigmoidgating", + "num_attention_heads": 32, + "num_experts": 4, + "num_hidden_layers": 32, + "num_key_value_heads": 32, + "num_layers": 3, + "num_selected": 2, + "original_max_position_embeddings": 4096, + "pad_token_id": 32000, + "resid_pdrop": 0.0, + "rms_norm_eps": 1e-05, + "rope_scaling": { + "long_factor": [ + 1.0800000429153442, + 1.1100000143051147, + 1.1399999856948853, + 1.340000033378601, + 1.5899999141693115, + 1.600000023841858, + 1.6200000047683716, + 2.620000123977661, + 3.2300000190734863, + 3.2300000190734863, + 4.789999961853027, + 7.400000095367432, + 7.700000286102295, + 9.09000015258789, + 12.199999809265137, + 17.670000076293945, + 24.46000099182129, + 28.57000160217285, + 30.420001983642578, + 30.840002059936523, + 32.590003967285156, + 32.93000411987305, + 42.320003509521484, + 44.96000289916992, + 50.340003967285156, + 50.45000457763672, + 57.55000305175781, + 57.93000411987305, + 58.21000289916992, + 60.1400032043457, + 62.61000442504883, + 62.62000274658203, + 62.71000289916992, + 63.1400032043457, + 63.1400032043457, + 63.77000427246094, + 63.93000411987305, + 63.96000289916992, + 63.970001220703125, + 64.02999877929688, + 64.06999969482422, + 64.08000183105469, + 64.12000274658203, + 64.41000366210938, + 64.4800033569336, + 64.51000213623047, + 64.52999877929688, + 64.83999633789062 + ], + "short_factor": [ + 1.0, + 1.0199999809265137, + 1.0299999713897705, + 1.0299999713897705, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0499999523162842, + 1.0699999332427979, + 1.0999999046325684, + 1.1099998950958252, + 1.1599998474121094, + 1.1599998474121094, + 1.1699998378753662, + 1.2899998426437378, + 1.339999794960022, + 1.679999828338623, + 1.7899998426437378, + 1.8199998140335083, + 1.8499997854232788, + 1.8799997568130493, + 1.9099997282028198, + 1.9399996995925903, + 1.9899996519088745, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0199997425079346, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0299997329711914, + 2.0799996852874756, + 2.0899996757507324, + 2.189999580383301, + 2.2199995517730713, + 2.5899994373321533, + 2.729999542236328, + 2.749999523162842, + 2.8399994373321533 + ], + "type": "longrope" + }, + "rope_theta": 10000.0, + "router_z_loss_coef": 0.01, + "scales": [ + 1, + 3 + ], + "sliding_window": 262144, + "tie_word_embeddings": false, + "tokenizer_model_max_length": 2048, + "tokenizer_padding_side": "right", + "torch_dtype": "bfloat16", + "training": true, + "transformers_version": "4.43.0", + "tune_mm_mlp_adapter": false, + "use_cache": true, + "use_mm_proj": true, + "vocab_size": 32064 +} diff --git a/sft_full/smoe_sigmoidgating/generation_config.json b/sft_full/smoe_sigmoidgating/generation_config.json new file mode 100644 index 0000000000000000000000000000000000000000..dad5c4578f0dc5969b38755d095fc30c368bb54a --- /dev/null +++ b/sft_full/smoe_sigmoidgating/generation_config.json @@ -0,0 +1,12 @@ +{ + "_from_model_config": true, + "bos_token_id": 1, + "do_sample": true, + "eos_token_id": [ + 32007, + 32001, + 32000 + ], + "pad_token_id": 32000, + "transformers_version": "4.43.0" +} diff --git a/sft_full/smoe_sigmoidgating/model-00001-of-00003.safetensors b/sft_full/smoe_sigmoidgating/model-00001-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..bf41c5acda3b9e0db23466bdf42fd15c192cdfcf --- /dev/null +++ b/sft_full/smoe_sigmoidgating/model-00001-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e587beda01a33653098b85be9516b41b8ec89fbf3f4752b38c1594c9b6aca959 +size 4972489328 diff --git a/sft_full/smoe_sigmoidgating/model-00002-of-00003.safetensors b/sft_full/smoe_sigmoidgating/model-00002-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..05f034d2d0df5a3c1b0c6af7512d08a09f8887f4 --- /dev/null +++ b/sft_full/smoe_sigmoidgating/model-00002-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cc2a3a7f989c4e5d89514d1e28d5677812802ba89e73fb9e86dffb769c3e45ee +size 4985529728 diff --git a/sft_full/smoe_sigmoidgating/model-00003-of-00003.safetensors b/sft_full/smoe_sigmoidgating/model-00003-of-00003.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6f94f2d5651401c98550ae76a04f6170be30454e --- /dev/null +++ b/sft_full/smoe_sigmoidgating/model-00003-of-00003.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:261fd4bbdc640fe87db6fc90bf94160c7dcc4061c41329d2a035e3362e82b5d1 +size 248943552 diff --git a/sft_full/smoe_sigmoidgating/model.safetensors.index.json b/sft_full/smoe_sigmoidgating/model.safetensors.index.json new file mode 100644 index 0000000000000000000000000000000000000000..e8756dc8afc086e56dc085a5915e675a3782a3e2 --- /dev/null +++ b/sft_full/smoe_sigmoidgating/model.safetensors.index.json @@ -0,0 +1,1005 @@ +{ + "metadata": { + "total_size": 10206819456 + }, + "weight_map": { + "lm_head.weight": "model-00003-of-00003.safetensors", + "model.embed_tokens.weight": "model-00001-of-00003.safetensors", + "model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.21.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.30.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.input_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.down_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.post_attention_layernorm.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.o_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00003.safetensors", + "model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.input_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.down_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.post_attention_layernorm.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.o_proj.weight": "model-00001-of-00003.safetensors", + "model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.0.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.1.2.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.bias": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.0.weight": "model-00002-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.2.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.0.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.bias": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.experts.3.2.weight": "model-00003-of-00003.safetensors", + "model.mm_projector.moelayer.gate.fc.weight": "model-00003-of-00003.safetensors", + "model.norm.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.patch_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.embeddings.position_embedding.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.0.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.1.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.2.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.25.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.26.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.3.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.4.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.5.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.6.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.7.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.8.self_attn.v_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.layer_norm2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.0.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.1.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.2.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc1.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.experts.3.fc2.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.moelayer.gate.fc.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.out_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.bias": "model-00002-of-00003.safetensors", + "model.vision_tower.vision_model.encoder.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors" + } +} diff --git a/sft_full/smoe_sigmoidgating/special_tokens_map.json b/sft_full/smoe_sigmoidgating/special_tokens_map.json new file mode 100644 index 0000000000000000000000000000000000000000..3e4d5a5bc1cb51753cc9ae0305ece0da60052b10 --- /dev/null +++ b/sft_full/smoe_sigmoidgating/special_tokens_map.json @@ -0,0 +1,24 @@ +{ + "bos_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "eos_token": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + }, + "pad_token": "", + "unk_token": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false + } +} diff --git a/sft_full/smoe_sigmoidgating/tokenizer.model b/sft_full/smoe_sigmoidgating/tokenizer.model new file mode 100644 index 0000000000000000000000000000000000000000..6c00c742ce03c627d6cd5b795984876fa49fa899 --- /dev/null +++ b/sft_full/smoe_sigmoidgating/tokenizer.model @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347 +size 499723 diff --git a/sft_full/smoe_sigmoidgating/tokenizer_config.json b/sft_full/smoe_sigmoidgating/tokenizer_config.json new file mode 100644 index 0000000000000000000000000000000000000000..d579bb0b91b24b214ea3c2e487e27a65017cdc4a --- /dev/null +++ b/sft_full/smoe_sigmoidgating/tokenizer_config.json @@ -0,0 +1,132 @@ +{ + "add_bos_token": false, + "add_eos_token": false, + "add_prefix_space": true, + "added_tokens_decoder": { + "0": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "1": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "2": { + "content": "", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": false + }, + "32000": { + "content": "<|endoftext|>", + "lstrip": false, + "normalized": false, + "rstrip": false, + "single_word": false, + "special": true + }, + "32001": { + "content": "<|assistant|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32002": { + "content": "<|placeholder1|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32003": { + "content": "<|placeholder2|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32004": { + "content": "<|placeholder3|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32005": { + "content": "<|placeholder4|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32006": { + "content": "<|system|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32007": { + "content": "<|end|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32008": { + "content": "<|placeholder5|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32009": { + "content": "<|placeholder6|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + }, + "32010": { + "content": "<|user|>", + "lstrip": false, + "normalized": false, + "rstrip": true, + "single_word": false, + "special": true + } + }, + "bos_token": "", + "chat_template": "{% for message in messages %}{% if message['role'] == 'system' and message['content'] %}{{'<|system|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'user' %}{{'<|user|>\n' + message['content'] + '<|end|>\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>\n' + message['content'] + '<|end|>\n'}}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>\n' }}{% else %}{{ eos_token }}{% endif %}", + "clean_up_tokenization_spaces": false, + "eos_token": "<|endoftext|>", + "legacy": false, + "model_max_length": 2048, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": {}, + "spaces_between_special_tokens": false, + "tokenizer_class": "LlamaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +} diff --git a/sft_full/smoe_sigmoidgating/trainer_state.json b/sft_full/smoe_sigmoidgating/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..d575f16fdada71bb4fe754c0bc7ad9e8bc9c931d --- /dev/null +++ b/sft_full/smoe_sigmoidgating/trainer_state.json @@ -0,0 +1,77658 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999549082382648, + "eval_steps": 500, + "global_step": 11088, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 9.018352347026198e-05, + "grad_norm": 66.02296371930943, + "learning_rate": 0.0, + "loss": 2.1194, + "step": 1 + }, + { + "epoch": 0.00018036704694052397, + "grad_norm": 60.49348856036594, + "learning_rate": 4.773623799730706e-07, + "loss": 2.0397, + "step": 2 + }, + { + "epoch": 0.0002705505704107859, + "grad_norm": 36.67498721881168, + "learning_rate": 7.566014715123208e-07, + "loss": 1.7394, + "step": 3 + }, + { + "epoch": 0.00036073409388104793, + "grad_norm": 52.48251662012074, + "learning_rate": 9.547247599461412e-07, + "loss": 2.1386, + "step": 4 + }, + { + "epoch": 0.0004509176173513099, + "grad_norm": 45.11166700533244, + "learning_rate": 1.108401121501769e-06, + "loss": 1.6264, + "step": 5 + }, + { + "epoch": 0.0005411011408215718, + "grad_norm": 48.47355764760438, + "learning_rate": 1.2339638514853914e-06, + "loss": 1.8425, + "step": 6 + }, + { + "epoch": 0.0006312846642918339, + "grad_norm": 38.37405046412793, + "learning_rate": 1.3401256270225321e-06, + "loss": 1.5285, + "step": 7 + }, + { + "epoch": 0.0007214681877620959, + "grad_norm": 24.79345710337142, + "learning_rate": 1.4320871399192119e-06, + "loss": 1.4702, + "step": 8 + }, + { + "epoch": 0.0008116517112323579, + "grad_norm": 14.442534654657283, + "learning_rate": 1.5132029430246416e-06, + "loss": 1.3272, + "step": 9 + }, + { + "epoch": 0.0009018352347026198, + "grad_norm": 16.00263764574746, + "learning_rate": 1.5857635014748399e-06, + "loss": 1.4375, + "step": 10 + }, + { + "epoch": 0.0009920187581728818, + "grad_norm": 16.700138406693952, + "learning_rate": 1.6514025108267924e-06, + "loss": 1.5854, + "step": 11 + }, + { + "epoch": 0.0010822022816431437, + "grad_norm": 11.664910175159141, + "learning_rate": 1.711326231458462e-06, + "loss": 1.4252, + "step": 12 + }, + { + "epoch": 0.0011723858051134058, + "grad_norm": 5.90237509556948, + "learning_rate": 1.7664507107987104e-06, + "loss": 1.3125, + "step": 13 + }, + { + "epoch": 0.0012625693285836677, + "grad_norm": 4.731143465442639, + "learning_rate": 1.8174880069956024e-06, + "loss": 1.263, + "step": 14 + }, + { + "epoch": 0.0013527528520539298, + "grad_norm": 4.038155016602303, + "learning_rate": 1.8650025930140899e-06, + "loss": 1.2224, + "step": 15 + }, + { + "epoch": 0.0014429363755241917, + "grad_norm": 4.534637447783109, + "learning_rate": 1.9094495198922823e-06, + "loss": 1.3112, + "step": 16 + }, + { + "epoch": 0.0015331198989944536, + "grad_norm": 3.7201595680305015, + "learning_rate": 1.9512009899507514e-06, + "loss": 1.1645, + "step": 17 + }, + { + "epoch": 0.0016233034224647158, + "grad_norm": 3.2848393744415185, + "learning_rate": 1.990565322997712e-06, + "loss": 1.2201, + "step": 18 + }, + { + "epoch": 0.0017134869459349777, + "grad_norm": 2.991184953109061, + "learning_rate": 2.027800787770518e-06, + "loss": 1.1098, + "step": 19 + }, + { + "epoch": 0.0018036704694052396, + "grad_norm": 3.4548499860643003, + "learning_rate": 2.06312588144791e-06, + "loss": 1.2607, + "step": 20 + }, + { + "epoch": 0.0018938539928755017, + "grad_norm": 2.9046790686654798, + "learning_rate": 2.0967270985348526e-06, + "loss": 1.1201, + "step": 21 + }, + { + "epoch": 0.0019840375163457636, + "grad_norm": 4.576912468748122, + "learning_rate": 2.128764890799863e-06, + "loss": 1.0585, + "step": 22 + }, + { + "epoch": 0.0020742210398160257, + "grad_norm": 2.256751738578352, + "learning_rate": 2.1593783012990145e-06, + "loss": 1.0679, + "step": 23 + }, + { + "epoch": 0.0021644045632862874, + "grad_norm": 2.18881414993715, + "learning_rate": 2.188688611431533e-06, + "loss": 1.189, + "step": 24 + }, + { + "epoch": 0.0022545880867565495, + "grad_norm": 2.2330777330411493, + "learning_rate": 2.216802243003538e-06, + "loss": 1.2215, + "step": 25 + }, + { + "epoch": 0.0023447716102268116, + "grad_norm": 2.0820329676866156, + "learning_rate": 2.243813090771781e-06, + "loss": 1.0678, + "step": 26 + }, + { + "epoch": 0.0024349551336970737, + "grad_norm": 2.2198059363214564, + "learning_rate": 2.269804414536962e-06, + "loss": 1.1955, + "step": 27 + }, + { + "epoch": 0.0025251386571673354, + "grad_norm": 1.8676160376473958, + "learning_rate": 2.2948503869686733e-06, + "loss": 1.1228, + "step": 28 + }, + { + "epoch": 0.0026153221806375975, + "grad_norm": 2.4714909077885547, + "learning_rate": 2.3190173696980436e-06, + "loss": 0.698, + "step": 29 + }, + { + "epoch": 0.0027055057041078597, + "grad_norm": 2.140920688227387, + "learning_rate": 2.3423649729871604e-06, + "loss": 1.1151, + "step": 30 + }, + { + "epoch": 0.0027956892275781214, + "grad_norm": 1.7424667203070374, + "learning_rate": 2.364946941580084e-06, + "loss": 1.1525, + "step": 31 + }, + { + "epoch": 0.0028858727510483835, + "grad_norm": 1.9299507116042631, + "learning_rate": 2.3868118998653532e-06, + "loss": 1.2207, + "step": 32 + }, + { + "epoch": 0.0029760562745186456, + "grad_norm": 2.2759694685203846, + "learning_rate": 2.408003982339113e-06, + "loss": 0.6451, + "step": 33 + }, + { + "epoch": 0.0030662397979889073, + "grad_norm": 1.509968298648748, + "learning_rate": 2.4285633699238223e-06, + "loss": 1.0106, + "step": 34 + }, + { + "epoch": 0.0031564233214591694, + "grad_norm": 2.0406752014995275, + "learning_rate": 2.4485267485243007e-06, + "loss": 1.1093, + "step": 35 + }, + { + "epoch": 0.0032466068449294315, + "grad_norm": 1.532303428302214, + "learning_rate": 2.467927702970783e-06, + "loss": 1.0668, + "step": 36 + }, + { + "epoch": 0.003336790368399693, + "grad_norm": 1.4794440031860727, + "learning_rate": 2.4867970569753584e-06, + "loss": 1.0816, + "step": 37 + }, + { + "epoch": 0.0034269738918699553, + "grad_norm": 1.887697333321474, + "learning_rate": 2.5051631677435883e-06, + "loss": 0.9768, + "step": 38 + }, + { + "epoch": 0.0035171574153402174, + "grad_norm": 1.6430300568276772, + "learning_rate": 2.523052182311031e-06, + "loss": 0.9466, + "step": 39 + }, + { + "epoch": 0.003607340938810479, + "grad_norm": 1.7602849907319529, + "learning_rate": 2.540488261420981e-06, + "loss": 1.1248, + "step": 40 + }, + { + "epoch": 0.0036975244622807412, + "grad_norm": 1.7329230201912291, + "learning_rate": 2.557493775753984e-06, + "loss": 0.9319, + "step": 41 + }, + { + "epoch": 0.0037877079857510034, + "grad_norm": 1.429045319294377, + "learning_rate": 2.5740894785079235e-06, + "loss": 1.0066, + "step": 42 + }, + { + "epoch": 0.0038778915092212655, + "grad_norm": 1.3916859811976787, + "learning_rate": 2.5902946576685834e-06, + "loss": 1.0534, + "step": 43 + }, + { + "epoch": 0.003968075032691527, + "grad_norm": 1.7406517067851646, + "learning_rate": 2.606127270772933e-06, + "loss": 0.9967, + "step": 44 + }, + { + "epoch": 0.004058258556161789, + "grad_norm": 1.7641069756333576, + "learning_rate": 2.62160406452641e-06, + "loss": 0.9453, + "step": 45 + }, + { + "epoch": 0.004148442079632051, + "grad_norm": 1.5659129443324444, + "learning_rate": 2.636740681272085e-06, + "loss": 1.0826, + "step": 46 + }, + { + "epoch": 0.004238625603102313, + "grad_norm": 1.6587246814856667, + "learning_rate": 2.651551754008722e-06, + "loss": 1.1542, + "step": 47 + }, + { + "epoch": 0.004328809126572575, + "grad_norm": 1.5336112675805589, + "learning_rate": 2.6660509914046035e-06, + "loss": 1.0475, + "step": 48 + }, + { + "epoch": 0.004418992650042837, + "grad_norm": 1.6370489474781165, + "learning_rate": 2.6802512540450642e-06, + "loss": 1.0133, + "step": 49 + }, + { + "epoch": 0.004509176173513099, + "grad_norm": 1.733145613164302, + "learning_rate": 2.694164622976609e-06, + "loss": 1.0843, + "step": 50 + }, + { + "epoch": 0.0045993596969833616, + "grad_norm": 1.4998985892029721, + "learning_rate": 2.707802461463072e-06, + "loss": 0.9705, + "step": 51 + }, + { + "epoch": 0.004689543220453623, + "grad_norm": 1.50826964094594, + "learning_rate": 2.7211754707448516e-06, + "loss": 0.9997, + "step": 52 + }, + { + "epoch": 0.004779726743923885, + "grad_norm": 1.372029085999296, + "learning_rate": 2.734293740486721e-06, + "loss": 0.9664, + "step": 53 + }, + { + "epoch": 0.0048699102673941475, + "grad_norm": 1.4251831472389183, + "learning_rate": 2.747166794510033e-06, + "loss": 1.0324, + "step": 54 + }, + { + "epoch": 0.004960093790864409, + "grad_norm": 1.5558223703682974, + "learning_rate": 2.759803632328562e-06, + "loss": 0.9857, + "step": 55 + }, + { + "epoch": 0.005050277314334671, + "grad_norm": 1.5265646115525635, + "learning_rate": 2.772212766941744e-06, + "loss": 1.1301, + "step": 56 + }, + { + "epoch": 0.005140460837804933, + "grad_norm": 1.529281596647021, + "learning_rate": 2.7844022592828385e-06, + "loss": 1.0702, + "step": 57 + }, + { + "epoch": 0.005230644361275195, + "grad_norm": 1.5177975057859154, + "learning_rate": 2.7963797496711145e-06, + "loss": 0.9553, + "step": 58 + }, + { + "epoch": 0.005320827884745457, + "grad_norm": 1.5686136802017456, + "learning_rate": 2.80815248657541e-06, + "loss": 1.055, + "step": 59 + }, + { + "epoch": 0.005411011408215719, + "grad_norm": 1.416505252767802, + "learning_rate": 2.819727352960231e-06, + "loss": 0.6207, + "step": 60 + }, + { + "epoch": 0.005501194931685981, + "grad_norm": 1.7102699545209337, + "learning_rate": 2.8311108904541717e-06, + "loss": 1.0041, + "step": 61 + }, + { + "epoch": 0.005591378455156243, + "grad_norm": 1.5901352523762278, + "learning_rate": 2.842309321553155e-06, + "loss": 0.9974, + "step": 62 + }, + { + "epoch": 0.005681561978626505, + "grad_norm": 1.3391872833048108, + "learning_rate": 2.8533285700471737e-06, + "loss": 0.9267, + "step": 63 + }, + { + "epoch": 0.005771745502096767, + "grad_norm": 1.4323162654729502, + "learning_rate": 2.8641742798384237e-06, + "loss": 1.0717, + "step": 64 + }, + { + "epoch": 0.005861929025567029, + "grad_norm": 1.37883509811669, + "learning_rate": 2.874851832300479e-06, + "loss": 1.0083, + "step": 65 + }, + { + "epoch": 0.005952112549037291, + "grad_norm": 1.609741909830849, + "learning_rate": 2.8853663623121834e-06, + "loss": 1.0416, + "step": 66 + }, + { + "epoch": 0.006042296072507553, + "grad_norm": 1.479463663864893, + "learning_rate": 2.895722773085839e-06, + "loss": 1.0427, + "step": 67 + }, + { + "epoch": 0.0061324795959778146, + "grad_norm": 1.3251863618462696, + "learning_rate": 2.905925749896893e-06, + "loss": 1.0159, + "step": 68 + }, + { + "epoch": 0.006222663119448077, + "grad_norm": 1.4638206062260204, + "learning_rate": 2.915979772811335e-06, + "loss": 1.0086, + "step": 69 + }, + { + "epoch": 0.006312846642918339, + "grad_norm": 1.4380954670101382, + "learning_rate": 2.925889128497372e-06, + "loss": 0.9551, + "step": 70 + }, + { + "epoch": 0.0064030301663886005, + "grad_norm": 1.6057626868193275, + "learning_rate": 2.9356579211992906e-06, + "loss": 1.0809, + "step": 71 + }, + { + "epoch": 0.006493213689858863, + "grad_norm": 1.3981864320221378, + "learning_rate": 2.9452900829438533e-06, + "loss": 0.9879, + "step": 72 + }, + { + "epoch": 0.006583397213329125, + "grad_norm": 1.6462499920056457, + "learning_rate": 2.954789383042727e-06, + "loss": 1.0289, + "step": 73 + }, + { + "epoch": 0.006673580736799386, + "grad_norm": 1.298145177736712, + "learning_rate": 2.9641594369484293e-06, + "loss": 0.6913, + "step": 74 + }, + { + "epoch": 0.006763764260269649, + "grad_norm": 1.5941823485115454, + "learning_rate": 2.9734037145158586e-06, + "loss": 0.9994, + "step": 75 + }, + { + "epoch": 0.006853947783739911, + "grad_norm": 1.585123388740365, + "learning_rate": 2.982525547716659e-06, + "loss": 0.9789, + "step": 76 + }, + { + "epoch": 0.006944131307210172, + "grad_norm": 2.3170044819636617, + "learning_rate": 2.9915281378493246e-06, + "loss": 1.0538, + "step": 77 + }, + { + "epoch": 0.007034314830680435, + "grad_norm": 1.6136549546835155, + "learning_rate": 3.000414562284102e-06, + "loss": 1.033, + "step": 78 + }, + { + "epoch": 0.0071244983541506966, + "grad_norm": 1.3822294396099242, + "learning_rate": 3.009187780778246e-06, + "loss": 0.9971, + "step": 79 + }, + { + "epoch": 0.007214681877620958, + "grad_norm": 1.8241264761024305, + "learning_rate": 3.017850641394051e-06, + "loss": 1.0066, + "step": 80 + }, + { + "epoch": 0.007304865401091221, + "grad_norm": 1.6106731031322925, + "learning_rate": 3.0264058860492832e-06, + "loss": 0.9055, + "step": 81 + }, + { + "epoch": 0.0073950489245614825, + "grad_norm": 1.4644992270632053, + "learning_rate": 3.0348561557270548e-06, + "loss": 1.0557, + "step": 82 + }, + { + "epoch": 0.007485232448031745, + "grad_norm": 0.9382787331222585, + "learning_rate": 3.043203995369939e-06, + "loss": 0.6496, + "step": 83 + }, + { + "epoch": 0.007575415971502007, + "grad_norm": 1.5454446363896404, + "learning_rate": 3.051451858480994e-06, + "loss": 1.0919, + "step": 84 + }, + { + "epoch": 0.007665599494972268, + "grad_norm": 1.7436981494263468, + "learning_rate": 3.05960211145252e-06, + "loss": 0.9249, + "step": 85 + }, + { + "epoch": 0.007755783018442531, + "grad_norm": 1.3440523029322848, + "learning_rate": 3.0676570376416543e-06, + "loss": 0.8969, + "step": 86 + }, + { + "epoch": 0.007845966541912792, + "grad_norm": 1.6008230595786073, + "learning_rate": 3.0756188412103647e-06, + "loss": 1.0483, + "step": 87 + }, + { + "epoch": 0.007936150065383054, + "grad_norm": 1.383891702855259, + "learning_rate": 3.083489650746004e-06, + "loss": 0.978, + "step": 88 + }, + { + "epoch": 0.008026333588853317, + "grad_norm": 1.4356016810328944, + "learning_rate": 3.0912715226772975e-06, + "loss": 1.033, + "step": 89 + }, + { + "epoch": 0.008116517112323578, + "grad_norm": 1.253676063586227, + "learning_rate": 3.098966444499481e-06, + "loss": 0.9033, + "step": 90 + }, + { + "epoch": 0.00820670063579384, + "grad_norm": 1.419167256440096, + "learning_rate": 3.1065763378212426e-06, + "loss": 1.0298, + "step": 91 + }, + { + "epoch": 0.008296884159264103, + "grad_norm": 1.309482921440906, + "learning_rate": 3.1141030612451554e-06, + "loss": 0.9717, + "step": 92 + }, + { + "epoch": 0.008387067682734364, + "grad_norm": 0.9250591375159557, + "learning_rate": 3.1215484130924052e-06, + "loss": 0.6803, + "step": 93 + }, + { + "epoch": 0.008477251206204626, + "grad_norm": 1.5707042452170916, + "learning_rate": 3.128914133981793e-06, + "loss": 1.0117, + "step": 94 + }, + { + "epoch": 0.008567434729674889, + "grad_norm": 1.538473878407507, + "learning_rate": 3.136201909272287e-06, + "loss": 1.0416, + "step": 95 + }, + { + "epoch": 0.00865761825314515, + "grad_norm": 1.6884272112013075, + "learning_rate": 3.1434133713776735e-06, + "loss": 0.9467, + "step": 96 + }, + { + "epoch": 0.008747801776615412, + "grad_norm": 1.4869600774694374, + "learning_rate": 3.15055010196128e-06, + "loss": 1.0108, + "step": 97 + }, + { + "epoch": 0.008837985300085675, + "grad_norm": 1.626966973253826, + "learning_rate": 3.157613634018135e-06, + "loss": 0.9812, + "step": 98 + }, + { + "epoch": 0.008928168823555935, + "grad_norm": 1.2750057089434699, + "learning_rate": 3.1646054538514336e-06, + "loss": 0.978, + "step": 99 + }, + { + "epoch": 0.009018352347026198, + "grad_norm": 1.5881112521378564, + "learning_rate": 3.1715270029496797e-06, + "loss": 1.027, + "step": 100 + }, + { + "epoch": 0.00910853587049646, + "grad_norm": 1.3067705805405634, + "learning_rate": 3.1783796797704243e-06, + "loss": 1.0168, + "step": 101 + }, + { + "epoch": 0.009198719393966723, + "grad_norm": 1.6077350297508795, + "learning_rate": 3.185164841436142e-06, + "loss": 0.9562, + "step": 102 + }, + { + "epoch": 0.009288902917436984, + "grad_norm": 2.130492461142175, + "learning_rate": 3.1918838053473723e-06, + "loss": 0.9869, + "step": 103 + }, + { + "epoch": 0.009379086440907246, + "grad_norm": 1.3983632499546934, + "learning_rate": 3.198537850717922e-06, + "loss": 0.8763, + "step": 104 + }, + { + "epoch": 0.009469269964377509, + "grad_norm": 1.4128274740544848, + "learning_rate": 3.205128220036622e-06, + "loss": 0.9601, + "step": 105 + }, + { + "epoch": 0.00955945348784777, + "grad_norm": 1.2900550488991016, + "learning_rate": 3.2116561204597917e-06, + "loss": 0.9292, + "step": 106 + }, + { + "epoch": 0.009649637011318032, + "grad_norm": 0.9087963832787456, + "learning_rate": 3.218122725138335e-06, + "loss": 0.672, + "step": 107 + }, + { + "epoch": 0.009739820534788295, + "grad_norm": 1.6345539206604685, + "learning_rate": 3.224529174483104e-06, + "loss": 0.974, + "step": 108 + }, + { + "epoch": 0.009830004058258556, + "grad_norm": 1.713343283195153, + "learning_rate": 3.2308765773719435e-06, + "loss": 0.9356, + "step": 109 + }, + { + "epoch": 0.009920187581728818, + "grad_norm": 1.3499036984180564, + "learning_rate": 3.2371660123016323e-06, + "loss": 0.9814, + "step": 110 + }, + { + "epoch": 0.010010371105199081, + "grad_norm": 1.4695111189684726, + "learning_rate": 3.2433985284876787e-06, + "loss": 0.9445, + "step": 111 + }, + { + "epoch": 0.010100554628669342, + "grad_norm": 1.3054492576874321, + "learning_rate": 3.2495751469148143e-06, + "loss": 0.9712, + "step": 112 + }, + { + "epoch": 0.010190738152139604, + "grad_norm": 2.122659965570853, + "learning_rate": 3.2556968613407816e-06, + "loss": 0.9999, + "step": 113 + }, + { + "epoch": 0.010280921675609867, + "grad_norm": 1.5224278160735303, + "learning_rate": 3.2617646392559094e-06, + "loss": 1.0274, + "step": 114 + }, + { + "epoch": 0.010371105199080128, + "grad_norm": 1.5665664400682378, + "learning_rate": 3.2677794228007836e-06, + "loss": 0.9938, + "step": 115 + }, + { + "epoch": 0.01046128872255039, + "grad_norm": 1.5009374867839762, + "learning_rate": 3.273742129644185e-06, + "loss": 1.0027, + "step": 116 + }, + { + "epoch": 0.010551472246020653, + "grad_norm": 1.483008950729265, + "learning_rate": 3.279653653823352e-06, + "loss": 0.9702, + "step": 117 + }, + { + "epoch": 0.010641655769490914, + "grad_norm": 1.7124109064023514, + "learning_rate": 3.285514866548481e-06, + "loss": 0.9553, + "step": 118 + }, + { + "epoch": 0.010731839292961176, + "grad_norm": 1.6421754323005993, + "learning_rate": 3.2913266169732838e-06, + "loss": 1.0205, + "step": 119 + }, + { + "epoch": 0.010822022816431439, + "grad_norm": 1.6933978823251077, + "learning_rate": 3.2970897329333017e-06, + "loss": 0.96, + "step": 120 + }, + { + "epoch": 0.0109122063399017, + "grad_norm": 1.5044091114698361, + "learning_rate": 3.302805021653585e-06, + "loss": 0.977, + "step": 121 + }, + { + "epoch": 0.011002389863371962, + "grad_norm": 1.4929744382996442, + "learning_rate": 3.3084732704272426e-06, + "loss": 0.9945, + "step": 122 + }, + { + "epoch": 0.011092573386842225, + "grad_norm": 1.5214451208180368, + "learning_rate": 3.314095247266304e-06, + "loss": 0.9832, + "step": 123 + }, + { + "epoch": 0.011182756910312485, + "grad_norm": 4.41161241979121, + "learning_rate": 3.3196717015262255e-06, + "loss": 0.9796, + "step": 124 + }, + { + "epoch": 0.011272940433782748, + "grad_norm": 1.3804973467642803, + "learning_rate": 3.325203364505307e-06, + "loss": 0.9445, + "step": 125 + }, + { + "epoch": 0.01136312395725301, + "grad_norm": 1.6193718647309976, + "learning_rate": 3.3306909500202442e-06, + "loss": 1.0167, + "step": 126 + }, + { + "epoch": 0.011453307480723271, + "grad_norm": 1.840943220573555, + "learning_rate": 3.3361351549589145e-06, + "loss": 1.0127, + "step": 127 + }, + { + "epoch": 0.011543491004193534, + "grad_norm": 1.7126562812943686, + "learning_rate": 3.341536659811494e-06, + "loss": 0.942, + "step": 128 + }, + { + "epoch": 0.011633674527663796, + "grad_norm": 1.5179987144163214, + "learning_rate": 3.346896129180904e-06, + "loss": 1.0228, + "step": 129 + }, + { + "epoch": 0.011723858051134057, + "grad_norm": 1.1848101518554086, + "learning_rate": 3.35221421227355e-06, + "loss": 0.9521, + "step": 130 + }, + { + "epoch": 0.01181404157460432, + "grad_norm": 1.700822095682767, + "learning_rate": 3.357491543371255e-06, + "loss": 1.0363, + "step": 131 + }, + { + "epoch": 0.011904225098074582, + "grad_norm": 1.3564199139034572, + "learning_rate": 3.3627287422852543e-06, + "loss": 0.923, + "step": 132 + }, + { + "epoch": 0.011994408621544843, + "grad_norm": 1.2296430936015161, + "learning_rate": 3.3679264147930497e-06, + "loss": 0.9523, + "step": 133 + }, + { + "epoch": 0.012084592145015106, + "grad_norm": 1.6067241577425437, + "learning_rate": 3.37308515305891e-06, + "loss": 0.9083, + "step": 134 + }, + { + "epoch": 0.012174775668485368, + "grad_norm": 1.3291846592003231, + "learning_rate": 3.3782055360387313e-06, + "loss": 1.0255, + "step": 135 + }, + { + "epoch": 0.012264959191955629, + "grad_norm": 1.4234698228065954, + "learning_rate": 3.3832881298699633e-06, + "loss": 0.9396, + "step": 136 + }, + { + "epoch": 0.012355142715425892, + "grad_norm": 1.3603456035628638, + "learning_rate": 3.388333488247249e-06, + "loss": 0.9408, + "step": 137 + }, + { + "epoch": 0.012445326238896154, + "grad_norm": 1.3832218835738874, + "learning_rate": 3.393342152784406e-06, + "loss": 0.9931, + "step": 138 + }, + { + "epoch": 0.012535509762366415, + "grad_norm": 1.5337012775520038, + "learning_rate": 3.3983146533633376e-06, + "loss": 0.9874, + "step": 139 + }, + { + "epoch": 0.012625693285836678, + "grad_norm": 1.8065828221327407, + "learning_rate": 3.403251508470442e-06, + "loss": 1.0248, + "step": 140 + }, + { + "epoch": 0.01271587680930694, + "grad_norm": 1.4716125351457232, + "learning_rate": 3.408153225521043e-06, + "loss": 0.9081, + "step": 141 + }, + { + "epoch": 0.012806060332777201, + "grad_norm": 1.5394326166792025, + "learning_rate": 3.413020301172361e-06, + "loss": 0.8845, + "step": 142 + }, + { + "epoch": 0.012896243856247463, + "grad_norm": 1.3258926929218728, + "learning_rate": 3.4178532216255024e-06, + "loss": 0.9521, + "step": 143 + }, + { + "epoch": 0.012986427379717726, + "grad_norm": 1.4256441233287624, + "learning_rate": 3.422652462916924e-06, + "loss": 1.0253, + "step": 144 + }, + { + "epoch": 0.013076610903187987, + "grad_norm": 1.3345485851295062, + "learning_rate": 3.4274184911998124e-06, + "loss": 0.8527, + "step": 145 + }, + { + "epoch": 0.01316679442665825, + "grad_norm": 1.5979471478130869, + "learning_rate": 3.4321517630157976e-06, + "loss": 1.0822, + "step": 146 + }, + { + "epoch": 0.013256977950128512, + "grad_norm": 1.5910008224446817, + "learning_rate": 3.4368527255573845e-06, + "loss": 1.0337, + "step": 147 + }, + { + "epoch": 0.013347161473598773, + "grad_norm": 1.3902575832572395, + "learning_rate": 3.4415218169214994e-06, + "loss": 0.9371, + "step": 148 + }, + { + "epoch": 0.013437344997069035, + "grad_norm": 1.2876660730602671, + "learning_rate": 3.4461594663544882e-06, + "loss": 0.9571, + "step": 149 + }, + { + "epoch": 0.013527528520539298, + "grad_norm": 1.5190416346476383, + "learning_rate": 3.450766094488929e-06, + "loss": 0.934, + "step": 150 + }, + { + "epoch": 0.013617712044009559, + "grad_norm": 1.5130216912569732, + "learning_rate": 3.4553421135725735e-06, + "loss": 1.0081, + "step": 151 + }, + { + "epoch": 0.013707895567479821, + "grad_norm": 1.3633965386116058, + "learning_rate": 3.45988792768973e-06, + "loss": 0.968, + "step": 152 + }, + { + "epoch": 0.013798079090950084, + "grad_norm": 1.350288275563416, + "learning_rate": 3.464403932975393e-06, + "loss": 1.0017, + "step": 153 + }, + { + "epoch": 0.013888262614420345, + "grad_norm": 1.5173813386679915, + "learning_rate": 3.468890517822395e-06, + "loss": 1.0006, + "step": 154 + }, + { + "epoch": 0.013978446137890607, + "grad_norm": 1.7178274279551795, + "learning_rate": 3.473348063081853e-06, + "loss": 1.0086, + "step": 155 + }, + { + "epoch": 0.01406862966136087, + "grad_norm": 1.2847037402819743, + "learning_rate": 3.4777769422571727e-06, + "loss": 0.8852, + "step": 156 + }, + { + "epoch": 0.01415881318483113, + "grad_norm": 1.3896423468385941, + "learning_rate": 3.4821775216918497e-06, + "loss": 0.9487, + "step": 157 + }, + { + "epoch": 0.014248996708301393, + "grad_norm": 1.3453327438833975, + "learning_rate": 3.4865501607513164e-06, + "loss": 0.8987, + "step": 158 + }, + { + "epoch": 0.014339180231771656, + "grad_norm": 1.5046818306282326, + "learning_rate": 3.4908952119990423e-06, + "loss": 1.0259, + "step": 159 + }, + { + "epoch": 0.014429363755241916, + "grad_norm": 0.9475109608501866, + "learning_rate": 3.495213021367122e-06, + "loss": 0.6297, + "step": 160 + }, + { + "epoch": 0.014519547278712179, + "grad_norm": 1.6736653804114394, + "learning_rate": 3.4995039283215464e-06, + "loss": 1.0623, + "step": 161 + }, + { + "epoch": 0.014609730802182442, + "grad_norm": 1.494153869171285, + "learning_rate": 3.5037682660223533e-06, + "loss": 0.9835, + "step": 162 + }, + { + "epoch": 0.014699914325652702, + "grad_norm": 1.5899601924159634, + "learning_rate": 3.508006361478857e-06, + "loss": 1.0224, + "step": 163 + }, + { + "epoch": 0.014790097849122965, + "grad_norm": 1.4361330249982842, + "learning_rate": 3.5122185357001253e-06, + "loss": 0.9507, + "step": 164 + }, + { + "epoch": 0.014880281372593228, + "grad_norm": 1.9340598421237254, + "learning_rate": 3.5164051038408817e-06, + "loss": 1.0753, + "step": 165 + }, + { + "epoch": 0.01497046489606349, + "grad_norm": 1.4550099606517142, + "learning_rate": 3.5205663753430093e-06, + "loss": 0.9908, + "step": 166 + }, + { + "epoch": 0.015060648419533751, + "grad_norm": 1.3785686565952964, + "learning_rate": 3.5247026540727915e-06, + "loss": 1.055, + "step": 167 + }, + { + "epoch": 0.015150831943004013, + "grad_norm": 1.446334477548058, + "learning_rate": 3.5288142384540645e-06, + "loss": 0.877, + "step": 168 + }, + { + "epoch": 0.015241015466474276, + "grad_norm": 1.3613614881414315, + "learning_rate": 3.532901421597421e-06, + "loss": 0.9799, + "step": 169 + }, + { + "epoch": 0.015331198989944537, + "grad_norm": 2.0360050591291663, + "learning_rate": 3.5369644914255915e-06, + "loss": 0.9765, + "step": 170 + }, + { + "epoch": 0.0154213825134148, + "grad_norm": 1.4611267130509888, + "learning_rate": 3.5410037307951596e-06, + "loss": 0.9807, + "step": 171 + }, + { + "epoch": 0.015511566036885062, + "grad_norm": 1.240548017597279, + "learning_rate": 3.545019417614725e-06, + "loss": 0.9877, + "step": 172 + }, + { + "epoch": 0.015601749560355323, + "grad_norm": 1.4009077535207348, + "learning_rate": 3.5490118249596387e-06, + "loss": 1.021, + "step": 173 + }, + { + "epoch": 0.015691933083825584, + "grad_norm": 1.388231510780462, + "learning_rate": 3.5529812211834352e-06, + "loss": 1.0069, + "step": 174 + }, + { + "epoch": 0.015782116607295848, + "grad_norm": 1.356025339671894, + "learning_rate": 3.5569278700260707e-06, + "loss": 0.971, + "step": 175 + }, + { + "epoch": 0.01587230013076611, + "grad_norm": 1.3586156192925642, + "learning_rate": 3.5608520307190746e-06, + "loss": 0.9631, + "step": 176 + }, + { + "epoch": 0.01596248365423637, + "grad_norm": 2.988045716820703, + "learning_rate": 3.564753958087731e-06, + "loss": 0.9307, + "step": 177 + }, + { + "epoch": 0.016052667177706634, + "grad_norm": 1.3237516766646504, + "learning_rate": 3.5686339026503684e-06, + "loss": 0.9394, + "step": 178 + }, + { + "epoch": 0.016142850701176895, + "grad_norm": 1.3934036525902387, + "learning_rate": 3.5724921107148806e-06, + "loss": 1.0542, + "step": 179 + }, + { + "epoch": 0.016233034224647155, + "grad_norm": 1.4700284223338957, + "learning_rate": 3.576328824472552e-06, + "loss": 1.0231, + "step": 180 + }, + { + "epoch": 0.01632321774811742, + "grad_norm": 1.527209732146749, + "learning_rate": 3.5801442820892838e-06, + "loss": 1.0392, + "step": 181 + }, + { + "epoch": 0.01641340127158768, + "grad_norm": 1.2098239824250996, + "learning_rate": 3.583938717794313e-06, + "loss": 0.9668, + "step": 182 + }, + { + "epoch": 0.01650358479505794, + "grad_norm": 1.3215529804874606, + "learning_rate": 3.5877123619664928e-06, + "loss": 0.9875, + "step": 183 + }, + { + "epoch": 0.016593768318528206, + "grad_norm": 1.4172414380358462, + "learning_rate": 3.5914654412182268e-06, + "loss": 0.9483, + "step": 184 + }, + { + "epoch": 0.016683951841998466, + "grad_norm": 1.346879021800115, + "learning_rate": 3.595198178477127e-06, + "loss": 1.0047, + "step": 185 + }, + { + "epoch": 0.016774135365468727, + "grad_norm": 1.3312247958953913, + "learning_rate": 3.5989107930654757e-06, + "loss": 0.9068, + "step": 186 + }, + { + "epoch": 0.01686431888893899, + "grad_norm": 1.3847372592877778, + "learning_rate": 3.6026035007775437e-06, + "loss": 1.0045, + "step": 187 + }, + { + "epoch": 0.016954502412409252, + "grad_norm": 1.594654948302073, + "learning_rate": 3.6062765139548636e-06, + "loss": 0.9863, + "step": 188 + }, + { + "epoch": 0.017044685935879513, + "grad_norm": 1.4829616845416485, + "learning_rate": 3.6099300415594945e-06, + "loss": 0.8718, + "step": 189 + }, + { + "epoch": 0.017134869459349777, + "grad_norm": 1.2584916171043654, + "learning_rate": 3.6135642892453575e-06, + "loss": 0.8647, + "step": 190 + }, + { + "epoch": 0.01722505298282004, + "grad_norm": 1.258116833375595, + "learning_rate": 3.6171794594277004e-06, + "loss": 0.9538, + "step": 191 + }, + { + "epoch": 0.0173152365062903, + "grad_norm": 1.395854119482943, + "learning_rate": 3.620775751350745e-06, + "loss": 0.9283, + "step": 192 + }, + { + "epoch": 0.017405420029760563, + "grad_norm": 1.7938841766256675, + "learning_rate": 3.6243533611535794e-06, + "loss": 1.0456, + "step": 193 + }, + { + "epoch": 0.017495603553230824, + "grad_norm": 1.4689516781259577, + "learning_rate": 3.627912481934351e-06, + "loss": 0.9406, + "step": 194 + }, + { + "epoch": 0.017585787076701085, + "grad_norm": 1.475713495867052, + "learning_rate": 3.6314533038128e-06, + "loss": 0.9445, + "step": 195 + }, + { + "epoch": 0.01767597060017135, + "grad_norm": 0.833829521436084, + "learning_rate": 3.6349760139912048e-06, + "loss": 0.6147, + "step": 196 + }, + { + "epoch": 0.01776615412364161, + "grad_norm": 1.6006458171189377, + "learning_rate": 3.638480796813769e-06, + "loss": 1.0549, + "step": 197 + }, + { + "epoch": 0.01785633764711187, + "grad_norm": 1.1963856187117423, + "learning_rate": 3.641967833824504e-06, + "loss": 0.9579, + "step": 198 + }, + { + "epoch": 0.017946521170582135, + "grad_norm": 1.216434136905871, + "learning_rate": 3.645437303823663e-06, + "loss": 0.9203, + "step": 199 + }, + { + "epoch": 0.018036704694052396, + "grad_norm": 1.5177188842378768, + "learning_rate": 3.64888938292275e-06, + "loss": 0.9405, + "step": 200 + }, + { + "epoch": 0.01812688821752266, + "grad_norm": 1.391308905630317, + "learning_rate": 3.6523242445981603e-06, + "loss": 0.9531, + "step": 201 + }, + { + "epoch": 0.01821707174099292, + "grad_norm": 1.1862753526331964, + "learning_rate": 3.655742059743495e-06, + "loss": 0.9241, + "step": 202 + }, + { + "epoch": 0.018307255264463182, + "grad_norm": 1.4581264439921908, + "learning_rate": 3.659142996720576e-06, + "loss": 0.9758, + "step": 203 + }, + { + "epoch": 0.018397438787933446, + "grad_norm": 1.5206787954050505, + "learning_rate": 3.6625272214092135e-06, + "loss": 0.9354, + "step": 204 + }, + { + "epoch": 0.018487622311403707, + "grad_norm": 0.7063738680553531, + "learning_rate": 3.6658948972557535e-06, + "loss": 0.5904, + "step": 205 + }, + { + "epoch": 0.018577805834873968, + "grad_norm": 1.4875942456818547, + "learning_rate": 3.6692461853204432e-06, + "loss": 0.8882, + "step": 206 + }, + { + "epoch": 0.018667989358344232, + "grad_norm": 1.4506692953056708, + "learning_rate": 3.672581244323656e-06, + "loss": 0.8877, + "step": 207 + }, + { + "epoch": 0.018758172881814493, + "grad_norm": 1.4888580085784715, + "learning_rate": 3.6759002306909926e-06, + "loss": 0.8934, + "step": 208 + }, + { + "epoch": 0.018848356405284754, + "grad_norm": 1.331568349464172, + "learning_rate": 3.67920329859731e-06, + "loss": 0.9301, + "step": 209 + }, + { + "epoch": 0.018938539928755018, + "grad_norm": 1.4571598599614628, + "learning_rate": 3.6824906000096923e-06, + "loss": 0.9465, + "step": 210 + }, + { + "epoch": 0.01902872345222528, + "grad_norm": 1.61711238373347, + "learning_rate": 3.6857622847294067e-06, + "loss": 0.9711, + "step": 211 + }, + { + "epoch": 0.01911890697569554, + "grad_norm": 1.6931927344882085, + "learning_rate": 3.6890185004328626e-06, + "loss": 1.0121, + "step": 212 + }, + { + "epoch": 0.019209090499165804, + "grad_norm": 1.4884032352146828, + "learning_rate": 3.6922593927116113e-06, + "loss": 0.8794, + "step": 213 + }, + { + "epoch": 0.019299274022636065, + "grad_norm": 2.07804801626234, + "learning_rate": 3.695485105111406e-06, + "loss": 0.9607, + "step": 214 + }, + { + "epoch": 0.019389457546106326, + "grad_norm": 1.2850059259033122, + "learning_rate": 3.698695779170352e-06, + "loss": 0.8821, + "step": 215 + }, + { + "epoch": 0.01947964106957659, + "grad_norm": 1.6144344466278264, + "learning_rate": 3.7018915544561744e-06, + "loss": 0.9855, + "step": 216 + }, + { + "epoch": 0.01956982459304685, + "grad_norm": 1.5677244960758614, + "learning_rate": 3.7050725686026164e-06, + "loss": 0.9486, + "step": 217 + }, + { + "epoch": 0.01966000811651711, + "grad_norm": 1.6149337470339022, + "learning_rate": 3.708238957345014e-06, + "loss": 1.0304, + "step": 218 + }, + { + "epoch": 0.019750191639987376, + "grad_norm": 1.360131853247465, + "learning_rate": 3.7113908545550482e-06, + "loss": 0.9645, + "step": 219 + }, + { + "epoch": 0.019840375163457637, + "grad_norm": 1.4955724699469295, + "learning_rate": 3.7145283922747028e-06, + "loss": 0.9663, + "step": 220 + }, + { + "epoch": 0.019930558686927898, + "grad_norm": 1.3760414081181473, + "learning_rate": 3.7176517007494612e-06, + "loss": 1.023, + "step": 221 + }, + { + "epoch": 0.020020742210398162, + "grad_norm": 1.4613194093951618, + "learning_rate": 3.7207609084607496e-06, + "loss": 0.8795, + "step": 222 + }, + { + "epoch": 0.020110925733868423, + "grad_norm": 1.5767625271620258, + "learning_rate": 3.723856142157645e-06, + "loss": 0.9929, + "step": 223 + }, + { + "epoch": 0.020201109257338683, + "grad_norm": 1.3462833211276368, + "learning_rate": 3.726937526887885e-06, + "loss": 0.926, + "step": 224 + }, + { + "epoch": 0.020291292780808948, + "grad_norm": 1.4897786614813215, + "learning_rate": 3.7300051860281798e-06, + "loss": 0.9234, + "step": 225 + }, + { + "epoch": 0.02038147630427921, + "grad_norm": 1.7381438125969448, + "learning_rate": 3.733059241313852e-06, + "loss": 1.056, + "step": 226 + }, + { + "epoch": 0.02047165982774947, + "grad_norm": 1.4383429667893206, + "learning_rate": 3.736099812867827e-06, + "loss": 0.9423, + "step": 227 + }, + { + "epoch": 0.020561843351219734, + "grad_norm": 1.6350804690002878, + "learning_rate": 3.73912701922898e-06, + "loss": 1.0147, + "step": 228 + }, + { + "epoch": 0.020652026874689994, + "grad_norm": 0.9426568616159745, + "learning_rate": 3.742140977379868e-06, + "loss": 0.6216, + "step": 229 + }, + { + "epoch": 0.020742210398160255, + "grad_norm": 1.3622617653268492, + "learning_rate": 3.745141802773854e-06, + "loss": 1.0029, + "step": 230 + }, + { + "epoch": 0.02083239392163052, + "grad_norm": 1.5892746115184941, + "learning_rate": 3.748129609361645e-06, + "loss": 0.9385, + "step": 231 + }, + { + "epoch": 0.02092257744510078, + "grad_norm": 1.2419449191480154, + "learning_rate": 3.7511045096172555e-06, + "loss": 0.9252, + "step": 232 + }, + { + "epoch": 0.02101276096857104, + "grad_norm": 1.6828892200659438, + "learning_rate": 3.7540666145634137e-06, + "loss": 0.9576, + "step": 233 + }, + { + "epoch": 0.021102944492041305, + "grad_norm": 1.4056937495101143, + "learning_rate": 3.7570160337964225e-06, + "loss": 0.9191, + "step": 234 + }, + { + "epoch": 0.021193128015511566, + "grad_norm": 2.4936149120789346, + "learning_rate": 3.7599528755104913e-06, + "loss": 1.0097, + "step": 235 + }, + { + "epoch": 0.021283311538981827, + "grad_norm": 1.478197111342628, + "learning_rate": 3.7628772465215515e-06, + "loss": 0.9327, + "step": 236 + }, + { + "epoch": 0.02137349506245209, + "grad_norm": 1.4359883362713461, + "learning_rate": 3.7657892522905666e-06, + "loss": 0.9353, + "step": 237 + }, + { + "epoch": 0.021463678585922352, + "grad_norm": 1.5381320522600845, + "learning_rate": 3.7686889969463542e-06, + "loss": 0.8809, + "step": 238 + }, + { + "epoch": 0.021553862109392613, + "grad_norm": 1.5795800328134266, + "learning_rate": 3.771576583307928e-06, + "loss": 0.9183, + "step": 239 + }, + { + "epoch": 0.021644045632862877, + "grad_norm": 1.8773776555740083, + "learning_rate": 3.7744521129063722e-06, + "loss": 0.9739, + "step": 240 + }, + { + "epoch": 0.021734229156333138, + "grad_norm": 1.4258280707209616, + "learning_rate": 3.7773156860062653e-06, + "loss": 0.9626, + "step": 241 + }, + { + "epoch": 0.0218244126798034, + "grad_norm": 1.639646038260417, + "learning_rate": 3.7801674016266554e-06, + "loss": 1.0741, + "step": 242 + }, + { + "epoch": 0.021914596203273663, + "grad_norm": 0.6332671098426979, + "learning_rate": 3.7830073575616035e-06, + "loss": 0.5657, + "step": 243 + }, + { + "epoch": 0.022004779726743924, + "grad_norm": 1.5322241506563634, + "learning_rate": 3.785835650400313e-06, + "loss": 0.9382, + "step": 244 + }, + { + "epoch": 0.022094963250214185, + "grad_norm": 1.800688635705629, + "learning_rate": 3.7886523755468334e-06, + "loss": 0.8538, + "step": 245 + }, + { + "epoch": 0.02218514677368445, + "grad_norm": 1.2779937863325717, + "learning_rate": 3.7914576272393746e-06, + "loss": 0.8944, + "step": 246 + }, + { + "epoch": 0.02227533029715471, + "grad_norm": 1.8514721683405804, + "learning_rate": 3.7942514985692284e-06, + "loss": 1.0779, + "step": 247 + }, + { + "epoch": 0.02236551382062497, + "grad_norm": 1.4833327800592542, + "learning_rate": 3.797034081499296e-06, + "loss": 0.967, + "step": 248 + }, + { + "epoch": 0.022455697344095235, + "grad_norm": 1.407537930182467, + "learning_rate": 3.7998054668822595e-06, + "loss": 0.9291, + "step": 249 + }, + { + "epoch": 0.022545880867565496, + "grad_norm": 2.862891197959248, + "learning_rate": 3.8025657444783776e-06, + "loss": 0.9063, + "step": 250 + }, + { + "epoch": 0.022636064391035757, + "grad_norm": 1.6068861150783726, + "learning_rate": 3.80531500297293e-06, + "loss": 0.8569, + "step": 251 + }, + { + "epoch": 0.02272624791450602, + "grad_norm": 1.971047166295798, + "learning_rate": 3.8080533299933147e-06, + "loss": 1.0156, + "step": 252 + }, + { + "epoch": 0.022816431437976282, + "grad_norm": 1.4523154844393016, + "learning_rate": 3.8107808121258067e-06, + "loss": 0.8925, + "step": 253 + }, + { + "epoch": 0.022906614961446543, + "grad_norm": 1.5321602743189668, + "learning_rate": 3.813497534931985e-06, + "loss": 1.0204, + "step": 254 + }, + { + "epoch": 0.022996798484916807, + "grad_norm": 1.7138495202526003, + "learning_rate": 3.816203582964841e-06, + "loss": 0.9673, + "step": 255 + }, + { + "epoch": 0.023086982008387068, + "grad_norm": 1.7209257746044708, + "learning_rate": 3.818899039784565e-06, + "loss": 1.0049, + "step": 256 + }, + { + "epoch": 0.02317716553185733, + "grad_norm": 1.5308093001227439, + "learning_rate": 3.821583987974031e-06, + "loss": 0.9565, + "step": 257 + }, + { + "epoch": 0.023267349055327593, + "grad_norm": 1.668556485046183, + "learning_rate": 3.8242585091539755e-06, + "loss": 0.8794, + "step": 258 + }, + { + "epoch": 0.023357532578797854, + "grad_norm": 1.4973716756005326, + "learning_rate": 3.8269226839978895e-06, + "loss": 1.0111, + "step": 259 + }, + { + "epoch": 0.023447716102268115, + "grad_norm": 1.8197710271522671, + "learning_rate": 3.82957659224662e-06, + "loss": 0.9, + "step": 260 + }, + { + "epoch": 0.02353789962573838, + "grad_norm": 1.8785648154681587, + "learning_rate": 3.8322203127226855e-06, + "loss": 0.876, + "step": 261 + }, + { + "epoch": 0.02362808314920864, + "grad_norm": 1.658874160270832, + "learning_rate": 3.834853923344326e-06, + "loss": 1.0282, + "step": 262 + }, + { + "epoch": 0.0237182666726789, + "grad_norm": 1.4107223600187546, + "learning_rate": 3.837477501139285e-06, + "loss": 0.9277, + "step": 263 + }, + { + "epoch": 0.023808450196149165, + "grad_norm": 1.8620267919749685, + "learning_rate": 3.840091122258324e-06, + "loss": 0.9439, + "step": 264 + }, + { + "epoch": 0.023898633719619426, + "grad_norm": 1.5838467200657278, + "learning_rate": 3.84269486198849e-06, + "loss": 0.8345, + "step": 265 + }, + { + "epoch": 0.023988817243089686, + "grad_norm": 1.618260861288885, + "learning_rate": 3.845288794766121e-06, + "loss": 0.9555, + "step": 266 + }, + { + "epoch": 0.02407900076655995, + "grad_norm": 1.9133911200955995, + "learning_rate": 3.847872994189619e-06, + "loss": 0.9385, + "step": 267 + }, + { + "epoch": 0.02416918429003021, + "grad_norm": 1.8240604699551435, + "learning_rate": 3.8504475330319805e-06, + "loss": 0.9761, + "step": 268 + }, + { + "epoch": 0.024259367813500472, + "grad_norm": 2.285918488491891, + "learning_rate": 3.853012483253093e-06, + "loss": 0.8726, + "step": 269 + }, + { + "epoch": 0.024349551336970737, + "grad_norm": 2.0221708007651187, + "learning_rate": 3.855567916011802e-06, + "loss": 0.928, + "step": 270 + }, + { + "epoch": 0.024439734860440997, + "grad_norm": 1.934305202722428, + "learning_rate": 3.858113901677755e-06, + "loss": 0.9461, + "step": 271 + }, + { + "epoch": 0.024529918383911258, + "grad_norm": 1.7048422179391345, + "learning_rate": 3.860650509843034e-06, + "loss": 0.8741, + "step": 272 + }, + { + "epoch": 0.024620101907381522, + "grad_norm": 2.3919794571482402, + "learning_rate": 3.863177809333563e-06, + "loss": 0.9472, + "step": 273 + }, + { + "epoch": 0.024710285430851783, + "grad_norm": 1.5169453326479605, + "learning_rate": 3.86569586822032e-06, + "loss": 0.8834, + "step": 274 + }, + { + "epoch": 0.024800468954322044, + "grad_norm": 2.0658375871600936, + "learning_rate": 3.868204753830331e-06, + "loss": 0.9185, + "step": 275 + }, + { + "epoch": 0.02489065247779231, + "grad_norm": 1.799615710350628, + "learning_rate": 3.870704532757476e-06, + "loss": 0.842, + "step": 276 + }, + { + "epoch": 0.02498083600126257, + "grad_norm": 1.7194584860131636, + "learning_rate": 3.8731952708730974e-06, + "loss": 0.9383, + "step": 277 + }, + { + "epoch": 0.02507101952473283, + "grad_norm": 1.698936561742791, + "learning_rate": 3.8756770333364085e-06, + "loss": 0.9234, + "step": 278 + }, + { + "epoch": 0.025161203048203094, + "grad_norm": 2.1086614461507143, + "learning_rate": 3.878149884604725e-06, + "loss": 0.8173, + "step": 279 + }, + { + "epoch": 0.025251386571673355, + "grad_norm": 2.221787214087483, + "learning_rate": 3.8806138884435125e-06, + "loss": 0.9067, + "step": 280 + }, + { + "epoch": 0.025341570095143616, + "grad_norm": 1.6782569648669752, + "learning_rate": 3.883069107936248e-06, + "loss": 0.8248, + "step": 281 + }, + { + "epoch": 0.02543175361861388, + "grad_norm": 2.419364661842831, + "learning_rate": 3.885515605494114e-06, + "loss": 0.9568, + "step": 282 + }, + { + "epoch": 0.02552193714208414, + "grad_norm": 2.386031381175261, + "learning_rate": 3.8879534428655145e-06, + "loss": 0.9673, + "step": 283 + }, + { + "epoch": 0.025612120665554402, + "grad_norm": 1.8500671142570069, + "learning_rate": 3.890382681145432e-06, + "loss": 0.9302, + "step": 284 + }, + { + "epoch": 0.025702304189024666, + "grad_norm": 1.7906887209540838, + "learning_rate": 3.892803380784608e-06, + "loss": 0.9322, + "step": 285 + }, + { + "epoch": 0.025792487712494927, + "grad_norm": 0.9745743922986421, + "learning_rate": 3.8952156015985725e-06, + "loss": 0.6442, + "step": 286 + }, + { + "epoch": 0.025882671235965188, + "grad_norm": 2.032525351766248, + "learning_rate": 3.897619402776516e-06, + "loss": 0.9015, + "step": 287 + }, + { + "epoch": 0.025972854759435452, + "grad_norm": 1.82678904176837, + "learning_rate": 3.900014842889995e-06, + "loss": 0.9667, + "step": 288 + }, + { + "epoch": 0.026063038282905713, + "grad_norm": 1.525166239823479, + "learning_rate": 3.902401979901503e-06, + "loss": 0.7947, + "step": 289 + }, + { + "epoch": 0.026153221806375974, + "grad_norm": 2.0276112756637983, + "learning_rate": 3.904780871172884e-06, + "loss": 0.9362, + "step": 290 + }, + { + "epoch": 0.026243405329846238, + "grad_norm": 2.895952748276716, + "learning_rate": 3.907151573473601e-06, + "loss": 0.7918, + "step": 291 + }, + { + "epoch": 0.0263335888533165, + "grad_norm": 1.3264601201785882, + "learning_rate": 3.909514142988868e-06, + "loss": 0.8897, + "step": 292 + }, + { + "epoch": 0.02642377237678676, + "grad_norm": 2.09489250326114, + "learning_rate": 3.911868635327639e-06, + "loss": 0.8606, + "step": 293 + }, + { + "epoch": 0.026513955900257024, + "grad_norm": 1.754605907740964, + "learning_rate": 3.914215105530455e-06, + "loss": 0.8629, + "step": 294 + }, + { + "epoch": 0.026604139423727285, + "grad_norm": 1.7411451428916294, + "learning_rate": 3.916553608077179e-06, + "loss": 0.8703, + "step": 295 + }, + { + "epoch": 0.026694322947197546, + "grad_norm": 1.7601091767174113, + "learning_rate": 3.91888419689457e-06, + "loss": 0.868, + "step": 296 + }, + { + "epoch": 0.02678450647066781, + "grad_norm": 2.179170171128246, + "learning_rate": 3.921206925363754e-06, + "loss": 0.819, + "step": 297 + }, + { + "epoch": 0.02687468999413807, + "grad_norm": 3.107441977022896, + "learning_rate": 3.923521846327559e-06, + "loss": 0.8202, + "step": 298 + }, + { + "epoch": 0.02696487351760833, + "grad_norm": 2.0139712861257757, + "learning_rate": 3.925829012097725e-06, + "loss": 0.8682, + "step": 299 + }, + { + "epoch": 0.027055057041078596, + "grad_norm": 2.5510507064499253, + "learning_rate": 3.928128474462e-06, + "loss": 0.8231, + "step": 300 + }, + { + "epoch": 0.027145240564548857, + "grad_norm": 4.1324686581622565, + "learning_rate": 3.930420284691115e-06, + "loss": 0.9135, + "step": 301 + }, + { + "epoch": 0.027235424088019117, + "grad_norm": 2.4759197183144055, + "learning_rate": 3.932704493545644e-06, + "loss": 0.9113, + "step": 302 + }, + { + "epoch": 0.02732560761148938, + "grad_norm": 2.0995137605409218, + "learning_rate": 3.934981151282745e-06, + "loss": 0.9583, + "step": 303 + }, + { + "epoch": 0.027415791134959643, + "grad_norm": 2.1715264349410313, + "learning_rate": 3.9372503076628006e-06, + "loss": 0.8935, + "step": 304 + }, + { + "epoch": 0.027505974658429903, + "grad_norm": 2.070101514897862, + "learning_rate": 3.939512011955941e-06, + "loss": 0.8924, + "step": 305 + }, + { + "epoch": 0.027596158181900168, + "grad_norm": 1.9765793268143375, + "learning_rate": 3.941766312948463e-06, + "loss": 0.8891, + "step": 306 + }, + { + "epoch": 0.02768634170537043, + "grad_norm": 1.8934481011896798, + "learning_rate": 3.944013258949147e-06, + "loss": 0.8377, + "step": 307 + }, + { + "epoch": 0.02777652522884069, + "grad_norm": 2.150158667808711, + "learning_rate": 3.946252897795465e-06, + "loss": 0.6945, + "step": 308 + }, + { + "epoch": 0.027866708752310954, + "grad_norm": 2.913384161737814, + "learning_rate": 3.9484852768596935e-06, + "loss": 0.8945, + "step": 309 + }, + { + "epoch": 0.027956892275781214, + "grad_norm": 2.8575747299768635, + "learning_rate": 3.950710443054923e-06, + "loss": 0.7795, + "step": 310 + }, + { + "epoch": 0.028047075799251475, + "grad_norm": 2.6780516855894465, + "learning_rate": 3.952928442840981e-06, + "loss": 0.7842, + "step": 311 + }, + { + "epoch": 0.02813725932272174, + "grad_norm": 2.047786713438822, + "learning_rate": 3.955139322230243e-06, + "loss": 0.8938, + "step": 312 + }, + { + "epoch": 0.028227442846192, + "grad_norm": 2.288866262519546, + "learning_rate": 3.957343126793365e-06, + "loss": 0.8049, + "step": 313 + }, + { + "epoch": 0.02831762636966226, + "grad_norm": 1.8699047794680197, + "learning_rate": 3.959539901664921e-06, + "loss": 0.8445, + "step": 314 + }, + { + "epoch": 0.028407809893132525, + "grad_norm": 2.378029848217825, + "learning_rate": 3.9617296915489425e-06, + "loss": 0.9702, + "step": 315 + }, + { + "epoch": 0.028497993416602786, + "grad_norm": 2.575708632505666, + "learning_rate": 3.963912540724387e-06, + "loss": 0.8352, + "step": 316 + }, + { + "epoch": 0.028588176940073047, + "grad_norm": 2.2626042716093298, + "learning_rate": 3.966088493050501e-06, + "loss": 0.8872, + "step": 317 + }, + { + "epoch": 0.02867836046354331, + "grad_norm": 1.9481359620163745, + "learning_rate": 3.968257591972113e-06, + "loss": 0.8551, + "step": 318 + }, + { + "epoch": 0.028768543987013572, + "grad_norm": 1.6713824393295347, + "learning_rate": 3.970419880524835e-06, + "loss": 0.9345, + "step": 319 + }, + { + "epoch": 0.028858727510483833, + "grad_norm": 2.24591488333318, + "learning_rate": 3.972575401340192e-06, + "loss": 0.837, + "step": 320 + }, + { + "epoch": 0.028948911033954097, + "grad_norm": 1.660227910907417, + "learning_rate": 3.974724196650656e-06, + "loss": 0.8896, + "step": 321 + }, + { + "epoch": 0.029039094557424358, + "grad_norm": 2.186956828664131, + "learning_rate": 3.976866308294617e-06, + "loss": 0.856, + "step": 322 + }, + { + "epoch": 0.02912927808089462, + "grad_norm": 2.1929385561982775, + "learning_rate": 3.979001777721269e-06, + "loss": 0.8125, + "step": 323 + }, + { + "epoch": 0.029219461604364883, + "grad_norm": 2.0359351485290764, + "learning_rate": 3.981130645995424e-06, + "loss": 0.8263, + "step": 324 + }, + { + "epoch": 0.029309645127835144, + "grad_norm": 1.8751139924048161, + "learning_rate": 3.983252953802248e-06, + "loss": 0.9346, + "step": 325 + }, + { + "epoch": 0.029399828651305405, + "grad_norm": 0.7312465341599974, + "learning_rate": 3.9853687414519285e-06, + "loss": 0.6369, + "step": 326 + }, + { + "epoch": 0.02949001217477567, + "grad_norm": 2.0357174383344594, + "learning_rate": 3.987478048884265e-06, + "loss": 0.9061, + "step": 327 + }, + { + "epoch": 0.02958019569824593, + "grad_norm": 2.3297199699856215, + "learning_rate": 3.989580915673196e-06, + "loss": 0.8356, + "step": 328 + }, + { + "epoch": 0.02967037922171619, + "grad_norm": 1.795595659015598, + "learning_rate": 3.991677381031255e-06, + "loss": 0.8337, + "step": 329 + }, + { + "epoch": 0.029760562745186455, + "grad_norm": 1.7753236126396639, + "learning_rate": 3.993767483813953e-06, + "loss": 0.9294, + "step": 330 + }, + { + "epoch": 0.029850746268656716, + "grad_norm": 1.9416936588730564, + "learning_rate": 3.995851262524104e-06, + "loss": 0.7664, + "step": 331 + }, + { + "epoch": 0.02994092979212698, + "grad_norm": 2.0092006293260702, + "learning_rate": 3.997928755316079e-06, + "loss": 0.8621, + "step": 332 + }, + { + "epoch": 0.03003111331559724, + "grad_norm": 0.8739310615355309, + "learning_rate": 4e-06, + "loss": 0.6243, + "step": 333 + }, + { + "epoch": 0.030121296839067502, + "grad_norm": 1.6216363573263988, + "learning_rate": 3.999999914674486e-06, + "loss": 0.8219, + "step": 334 + }, + { + "epoch": 0.030211480362537766, + "grad_norm": 1.6674952672522323, + "learning_rate": 3.999999658697952e-06, + "loss": 0.8427, + "step": 335 + }, + { + "epoch": 0.030301663886008027, + "grad_norm": 0.6927725659373812, + "learning_rate": 3.9999992320704185e-06, + "loss": 0.5606, + "step": 336 + }, + { + "epoch": 0.030391847409478288, + "grad_norm": 2.0799820712123798, + "learning_rate": 3.999998634791922e-06, + "loss": 0.8753, + "step": 337 + }, + { + "epoch": 0.030482030932948552, + "grad_norm": 1.6800526066899595, + "learning_rate": 3.999997866862515e-06, + "loss": 0.8357, + "step": 338 + }, + { + "epoch": 0.030572214456418813, + "grad_norm": 1.6903582491178364, + "learning_rate": 3.999996928282262e-06, + "loss": 0.939, + "step": 339 + }, + { + "epoch": 0.030662397979889074, + "grad_norm": 2.52736544710306, + "learning_rate": 3.999995819051244e-06, + "loss": 0.8753, + "step": 340 + }, + { + "epoch": 0.030752581503359338, + "grad_norm": 1.612868064450966, + "learning_rate": 3.9999945391695536e-06, + "loss": 0.8099, + "step": 341 + }, + { + "epoch": 0.0308427650268296, + "grad_norm": 3.3506347510379704, + "learning_rate": 3.999993088637302e-06, + "loss": 0.9329, + "step": 342 + }, + { + "epoch": 0.03093294855029986, + "grad_norm": 1.9688104287213493, + "learning_rate": 3.999991467454612e-06, + "loss": 0.9132, + "step": 343 + }, + { + "epoch": 0.031023132073770124, + "grad_norm": 1.5433872862435192, + "learning_rate": 3.999989675621622e-06, + "loss": 0.7652, + "step": 344 + }, + { + "epoch": 0.031113315597240385, + "grad_norm": 1.8884769563973371, + "learning_rate": 3.999987713138485e-06, + "loss": 0.8526, + "step": 345 + }, + { + "epoch": 0.031203499120710645, + "grad_norm": 0.8007130663616862, + "learning_rate": 3.999985580005369e-06, + "loss": 0.6682, + "step": 346 + }, + { + "epoch": 0.031293682644180906, + "grad_norm": 2.036671372869687, + "learning_rate": 3.999983276222455e-06, + "loss": 0.9187, + "step": 347 + }, + { + "epoch": 0.03138386616765117, + "grad_norm": 1.675634062022911, + "learning_rate": 3.999980801789941e-06, + "loss": 0.8477, + "step": 348 + }, + { + "epoch": 0.031474049691121435, + "grad_norm": 1.8567067082171984, + "learning_rate": 3.999978156708036e-06, + "loss": 0.8443, + "step": 349 + }, + { + "epoch": 0.031564233214591696, + "grad_norm": 4.197359008602281, + "learning_rate": 3.9999753409769675e-06, + "loss": 0.9703, + "step": 350 + }, + { + "epoch": 0.031654416738061956, + "grad_norm": 0.7125797795921733, + "learning_rate": 3.999972354596975e-06, + "loss": 0.6108, + "step": 351 + }, + { + "epoch": 0.03174460026153222, + "grad_norm": 2.162600026546031, + "learning_rate": 3.999969197568314e-06, + "loss": 0.8002, + "step": 352 + }, + { + "epoch": 0.03183478378500248, + "grad_norm": 1.9631677839929806, + "learning_rate": 3.999965869891253e-06, + "loss": 0.9036, + "step": 353 + }, + { + "epoch": 0.03192496730847274, + "grad_norm": 1.5692896002318222, + "learning_rate": 3.999962371566075e-06, + "loss": 0.872, + "step": 354 + }, + { + "epoch": 0.03201515083194301, + "grad_norm": 1.67128907391244, + "learning_rate": 3.999958702593082e-06, + "loss": 0.9183, + "step": 355 + }, + { + "epoch": 0.03210533435541327, + "grad_norm": 1.5645375508781854, + "learning_rate": 3.999954862972583e-06, + "loss": 0.8732, + "step": 356 + }, + { + "epoch": 0.03219551787888353, + "grad_norm": 1.7661385253587063, + "learning_rate": 3.999950852704908e-06, + "loss": 0.7944, + "step": 357 + }, + { + "epoch": 0.03228570140235379, + "grad_norm": 2.2771460838315525, + "learning_rate": 3.9999466717903995e-06, + "loss": 0.8986, + "step": 358 + }, + { + "epoch": 0.03237588492582405, + "grad_norm": 0.7222988997820593, + "learning_rate": 3.999942320229413e-06, + "loss": 0.5969, + "step": 359 + }, + { + "epoch": 0.03246606844929431, + "grad_norm": 2.0724289925884767, + "learning_rate": 3.99993779802232e-06, + "loss": 0.9823, + "step": 360 + }, + { + "epoch": 0.03255625197276458, + "grad_norm": 1.8721182612575284, + "learning_rate": 3.999933105169506e-06, + "loss": 0.8329, + "step": 361 + }, + { + "epoch": 0.03264643549623484, + "grad_norm": 1.6248669936693554, + "learning_rate": 3.999928241671373e-06, + "loss": 0.8416, + "step": 362 + }, + { + "epoch": 0.0327366190197051, + "grad_norm": 1.8539875744112113, + "learning_rate": 3.999923207528334e-06, + "loss": 0.7488, + "step": 363 + }, + { + "epoch": 0.03282680254317536, + "grad_norm": 1.458617545997769, + "learning_rate": 3.9999180027408196e-06, + "loss": 0.8356, + "step": 364 + }, + { + "epoch": 0.03291698606664562, + "grad_norm": 1.862287660309876, + "learning_rate": 3.9999126273092735e-06, + "loss": 0.8277, + "step": 365 + }, + { + "epoch": 0.03300716959011588, + "grad_norm": 2.553540453499073, + "learning_rate": 3.999907081234156e-06, + "loss": 0.9366, + "step": 366 + }, + { + "epoch": 0.03309735311358615, + "grad_norm": 2.214333342554147, + "learning_rate": 3.999901364515938e-06, + "loss": 0.8254, + "step": 367 + }, + { + "epoch": 0.03318753663705641, + "grad_norm": 1.8901283936057294, + "learning_rate": 3.999895477155108e-06, + "loss": 0.8944, + "step": 368 + }, + { + "epoch": 0.03327772016052667, + "grad_norm": 0.7537759065367093, + "learning_rate": 3.999889419152169e-06, + "loss": 0.5777, + "step": 369 + }, + { + "epoch": 0.03336790368399693, + "grad_norm": 2.3043396768120736, + "learning_rate": 3.999883190507638e-06, + "loss": 0.8581, + "step": 370 + }, + { + "epoch": 0.033458087207467194, + "grad_norm": 1.742729684258471, + "learning_rate": 3.999876791222044e-06, + "loss": 0.8391, + "step": 371 + }, + { + "epoch": 0.033548270730937454, + "grad_norm": 1.8944406389413886, + "learning_rate": 3.999870221295936e-06, + "loss": 0.8026, + "step": 372 + }, + { + "epoch": 0.03363845425440772, + "grad_norm": 3.422360877810274, + "learning_rate": 3.999863480729875e-06, + "loss": 0.8227, + "step": 373 + }, + { + "epoch": 0.03372863777787798, + "grad_norm": 2.3489861553656217, + "learning_rate": 3.999856569524433e-06, + "loss": 0.7878, + "step": 374 + }, + { + "epoch": 0.033818821301348244, + "grad_norm": 1.5372453990500783, + "learning_rate": 3.999849487680202e-06, + "loss": 0.789, + "step": 375 + }, + { + "epoch": 0.033909004824818505, + "grad_norm": 1.8360390493025918, + "learning_rate": 3.999842235197786e-06, + "loss": 0.7783, + "step": 376 + }, + { + "epoch": 0.033999188348288766, + "grad_norm": 2.3475313327961085, + "learning_rate": 3.999834812077803e-06, + "loss": 0.7807, + "step": 377 + }, + { + "epoch": 0.034089371871759026, + "grad_norm": 2.279556481908581, + "learning_rate": 3.999827218320886e-06, + "loss": 0.9011, + "step": 378 + }, + { + "epoch": 0.034179555395229294, + "grad_norm": 2.0361931149268813, + "learning_rate": 3.999819453927685e-06, + "loss": 0.7596, + "step": 379 + }, + { + "epoch": 0.034269738918699555, + "grad_norm": 2.151370955440981, + "learning_rate": 3.999811518898861e-06, + "loss": 0.8475, + "step": 380 + }, + { + "epoch": 0.034359922442169816, + "grad_norm": 2.329265136093088, + "learning_rate": 3.999803413235092e-06, + "loss": 0.8758, + "step": 381 + }, + { + "epoch": 0.03445010596564008, + "grad_norm": 1.7340002256435085, + "learning_rate": 3.999795136937068e-06, + "loss": 0.954, + "step": 382 + }, + { + "epoch": 0.03454028948911034, + "grad_norm": 1.8729045603031658, + "learning_rate": 3.999786690005496e-06, + "loss": 0.8385, + "step": 383 + }, + { + "epoch": 0.0346304730125806, + "grad_norm": 1.757243288916031, + "learning_rate": 3.999778072441098e-06, + "loss": 0.8339, + "step": 384 + }, + { + "epoch": 0.034720656536050866, + "grad_norm": 2.7958331165417865, + "learning_rate": 3.999769284244608e-06, + "loss": 0.9044, + "step": 385 + }, + { + "epoch": 0.03481084005952113, + "grad_norm": 1.944800129405323, + "learning_rate": 3.999760325416775e-06, + "loss": 0.8599, + "step": 386 + }, + { + "epoch": 0.03490102358299139, + "grad_norm": 1.7009794386978352, + "learning_rate": 3.999751195958366e-06, + "loss": 0.9315, + "step": 387 + }, + { + "epoch": 0.03499120710646165, + "grad_norm": 0.7875190974750641, + "learning_rate": 3.999741895870157e-06, + "loss": 0.6016, + "step": 388 + }, + { + "epoch": 0.03508139062993191, + "grad_norm": 2.1661190906025576, + "learning_rate": 3.999732425152944e-06, + "loss": 0.8912, + "step": 389 + }, + { + "epoch": 0.03517157415340217, + "grad_norm": 1.9187623536150245, + "learning_rate": 3.999722783807533e-06, + "loss": 0.8998, + "step": 390 + }, + { + "epoch": 0.03526175767687244, + "grad_norm": 2.4497649878066667, + "learning_rate": 3.999712971834748e-06, + "loss": 0.7829, + "step": 391 + }, + { + "epoch": 0.0353519412003427, + "grad_norm": 1.7239733999771387, + "learning_rate": 3.999702989235427e-06, + "loss": 0.8606, + "step": 392 + }, + { + "epoch": 0.03544212472381296, + "grad_norm": 1.7810330175116218, + "learning_rate": 3.999692836010419e-06, + "loss": 0.9324, + "step": 393 + }, + { + "epoch": 0.03553230824728322, + "grad_norm": 2.2095055138268163, + "learning_rate": 3.999682512160593e-06, + "loss": 0.9836, + "step": 394 + }, + { + "epoch": 0.03562249177075348, + "grad_norm": 0.7859839636110727, + "learning_rate": 3.99967201768683e-06, + "loss": 0.5794, + "step": 395 + }, + { + "epoch": 0.03571267529422374, + "grad_norm": 4.14063444316435, + "learning_rate": 3.999661352590023e-06, + "loss": 0.8388, + "step": 396 + }, + { + "epoch": 0.03580285881769401, + "grad_norm": 2.39954114739633, + "learning_rate": 3.999650516871083e-06, + "loss": 0.8498, + "step": 397 + }, + { + "epoch": 0.03589304234116427, + "grad_norm": 1.7793098725598955, + "learning_rate": 3.9996395105309365e-06, + "loss": 0.8731, + "step": 398 + }, + { + "epoch": 0.03598322586463453, + "grad_norm": 1.6904958340701537, + "learning_rate": 3.99962833357052e-06, + "loss": 0.8707, + "step": 399 + }, + { + "epoch": 0.03607340938810479, + "grad_norm": 1.728918203251367, + "learning_rate": 3.999616985990789e-06, + "loss": 0.8263, + "step": 400 + }, + { + "epoch": 0.03616359291157505, + "grad_norm": 2.0619292191901133, + "learning_rate": 3.9996054677927104e-06, + "loss": 0.7821, + "step": 401 + }, + { + "epoch": 0.03625377643504532, + "grad_norm": 2.3056830890521605, + "learning_rate": 3.9995937789772675e-06, + "loss": 0.79, + "step": 402 + }, + { + "epoch": 0.03634395995851558, + "grad_norm": 0.7259559150163184, + "learning_rate": 3.999581919545458e-06, + "loss": 0.6044, + "step": 403 + }, + { + "epoch": 0.03643414348198584, + "grad_norm": 2.496224413393452, + "learning_rate": 3.9995698894982935e-06, + "loss": 0.8853, + "step": 404 + }, + { + "epoch": 0.0365243270054561, + "grad_norm": 1.8960485301305527, + "learning_rate": 3.9995576888368e-06, + "loss": 0.8916, + "step": 405 + }, + { + "epoch": 0.036614510528926364, + "grad_norm": 1.9223219189058856, + "learning_rate": 3.9995453175620194e-06, + "loss": 0.8384, + "step": 406 + }, + { + "epoch": 0.036704694052396625, + "grad_norm": 2.290525001747497, + "learning_rate": 3.999532775675007e-06, + "loss": 0.7109, + "step": 407 + }, + { + "epoch": 0.03679487757586689, + "grad_norm": 2.0704884580296468, + "learning_rate": 3.9995200631768326e-06, + "loss": 0.8258, + "step": 408 + }, + { + "epoch": 0.03688506109933715, + "grad_norm": 3.0568133125640258, + "learning_rate": 3.9995071800685815e-06, + "loss": 0.9871, + "step": 409 + }, + { + "epoch": 0.036975244622807414, + "grad_norm": 2.429635786383193, + "learning_rate": 3.999494126351352e-06, + "loss": 0.8227, + "step": 410 + }, + { + "epoch": 0.037065428146277675, + "grad_norm": 2.3336107792981657, + "learning_rate": 3.99948090202626e-06, + "loss": 0.8568, + "step": 411 + }, + { + "epoch": 0.037155611669747936, + "grad_norm": 2.107864177910211, + "learning_rate": 3.999467507094431e-06, + "loss": 0.8458, + "step": 412 + }, + { + "epoch": 0.0372457951932182, + "grad_norm": 1.4791098928525483, + "learning_rate": 3.999453941557011e-06, + "loss": 0.7535, + "step": 413 + }, + { + "epoch": 0.037335978716688464, + "grad_norm": 0.7561830128408047, + "learning_rate": 3.999440205415154e-06, + "loss": 0.5571, + "step": 414 + }, + { + "epoch": 0.037426162240158725, + "grad_norm": 3.031022840030591, + "learning_rate": 3.999426298670035e-06, + "loss": 0.8963, + "step": 415 + }, + { + "epoch": 0.037516345763628986, + "grad_norm": 1.542620501270663, + "learning_rate": 3.9994122213228385e-06, + "loss": 0.8878, + "step": 416 + }, + { + "epoch": 0.03760652928709925, + "grad_norm": 2.1571180282650766, + "learning_rate": 3.9993979733747675e-06, + "loss": 0.8744, + "step": 417 + }, + { + "epoch": 0.03769671281056951, + "grad_norm": 1.7283635476664687, + "learning_rate": 3.999383554827037e-06, + "loss": 0.8438, + "step": 418 + }, + { + "epoch": 0.03778689633403977, + "grad_norm": 1.7775892151066777, + "learning_rate": 3.999368965680876e-06, + "loss": 0.8146, + "step": 419 + }, + { + "epoch": 0.037877079857510036, + "grad_norm": 1.836823334955398, + "learning_rate": 3.999354205937531e-06, + "loss": 0.8963, + "step": 420 + }, + { + "epoch": 0.0379672633809803, + "grad_norm": 1.7478124024397796, + "learning_rate": 3.999339275598261e-06, + "loss": 0.8958, + "step": 421 + }, + { + "epoch": 0.03805744690445056, + "grad_norm": 1.5404200883819958, + "learning_rate": 3.99932417466434e-06, + "loss": 0.8204, + "step": 422 + }, + { + "epoch": 0.03814763042792082, + "grad_norm": 2.4257078297879127, + "learning_rate": 3.999308903137056e-06, + "loss": 0.7853, + "step": 423 + }, + { + "epoch": 0.03823781395139108, + "grad_norm": 1.8520390022158713, + "learning_rate": 3.999293461017711e-06, + "loss": 0.8946, + "step": 424 + }, + { + "epoch": 0.03832799747486134, + "grad_norm": 2.3063609109483045, + "learning_rate": 3.9992778483076255e-06, + "loss": 0.8438, + "step": 425 + }, + { + "epoch": 0.03841818099833161, + "grad_norm": 1.5706215929084737, + "learning_rate": 3.99926206500813e-06, + "loss": 0.9004, + "step": 426 + }, + { + "epoch": 0.03850836452180187, + "grad_norm": 1.773656755917356, + "learning_rate": 3.999246111120571e-06, + "loss": 0.778, + "step": 427 + }, + { + "epoch": 0.03859854804527213, + "grad_norm": 1.7163962459836972, + "learning_rate": 3.999229986646311e-06, + "loss": 0.9115, + "step": 428 + }, + { + "epoch": 0.03868873156874239, + "grad_norm": 1.944612737214957, + "learning_rate": 3.999213691586723e-06, + "loss": 0.8086, + "step": 429 + }, + { + "epoch": 0.03877891509221265, + "grad_norm": 1.617926461013881, + "learning_rate": 3.9991972259432e-06, + "loss": 0.8782, + "step": 430 + }, + { + "epoch": 0.03886909861568291, + "grad_norm": 1.6581188670252378, + "learning_rate": 3.999180589717147e-06, + "loss": 0.9223, + "step": 431 + }, + { + "epoch": 0.03895928213915318, + "grad_norm": 1.8315448562665226, + "learning_rate": 3.999163782909983e-06, + "loss": 0.7662, + "step": 432 + }, + { + "epoch": 0.03904946566262344, + "grad_norm": 1.7965416640544671, + "learning_rate": 3.99914680552314e-06, + "loss": 0.8788, + "step": 433 + }, + { + "epoch": 0.0391396491860937, + "grad_norm": 2.0808215321715555, + "learning_rate": 3.999129657558069e-06, + "loss": 0.7789, + "step": 434 + }, + { + "epoch": 0.03922983270956396, + "grad_norm": 0.7411148350752504, + "learning_rate": 3.999112339016234e-06, + "loss": 0.5712, + "step": 435 + }, + { + "epoch": 0.03932001623303422, + "grad_norm": 1.7480224607670083, + "learning_rate": 3.999094849899109e-06, + "loss": 0.869, + "step": 436 + }, + { + "epoch": 0.039410199756504484, + "grad_norm": 3.3918561677847503, + "learning_rate": 3.99907719020819e-06, + "loss": 0.7643, + "step": 437 + }, + { + "epoch": 0.03950038327997475, + "grad_norm": 1.6112468167222498, + "learning_rate": 3.999059359944982e-06, + "loss": 0.867, + "step": 438 + }, + { + "epoch": 0.03959056680344501, + "grad_norm": 1.6020761550061249, + "learning_rate": 3.999041359111007e-06, + "loss": 0.8966, + "step": 439 + }, + { + "epoch": 0.03968075032691527, + "grad_norm": 1.8507177945068165, + "learning_rate": 3.999023187707801e-06, + "loss": 0.9186, + "step": 440 + }, + { + "epoch": 0.039770933850385534, + "grad_norm": 1.6887668516728656, + "learning_rate": 3.999004845736913e-06, + "loss": 0.8941, + "step": 441 + }, + { + "epoch": 0.039861117373855795, + "grad_norm": 1.6852581532757447, + "learning_rate": 3.9989863331999096e-06, + "loss": 0.9161, + "step": 442 + }, + { + "epoch": 0.039951300897326056, + "grad_norm": 2.2104286713653187, + "learning_rate": 3.99896765009837e-06, + "loss": 0.7807, + "step": 443 + }, + { + "epoch": 0.040041484420796324, + "grad_norm": 1.6734589533494786, + "learning_rate": 3.998948796433888e-06, + "loss": 0.8156, + "step": 444 + }, + { + "epoch": 0.040131667944266584, + "grad_norm": 1.4613982101475411, + "learning_rate": 3.998929772208073e-06, + "loss": 0.8857, + "step": 445 + }, + { + "epoch": 0.040221851467736845, + "grad_norm": 0.6900600755149858, + "learning_rate": 3.998910577422547e-06, + "loss": 0.5982, + "step": 446 + }, + { + "epoch": 0.040312034991207106, + "grad_norm": 1.5001723667292606, + "learning_rate": 3.99889121207895e-06, + "loss": 0.8889, + "step": 447 + }, + { + "epoch": 0.04040221851467737, + "grad_norm": 1.6929605931823575, + "learning_rate": 3.9988716761789324e-06, + "loss": 0.8672, + "step": 448 + }, + { + "epoch": 0.04049240203814763, + "grad_norm": 2.310451940938551, + "learning_rate": 3.998851969724161e-06, + "loss": 0.7806, + "step": 449 + }, + { + "epoch": 0.040582585561617895, + "grad_norm": 1.7410390173162418, + "learning_rate": 3.998832092716319e-06, + "loss": 0.8917, + "step": 450 + }, + { + "epoch": 0.040672769085088156, + "grad_norm": 1.436971650260061, + "learning_rate": 3.998812045157102e-06, + "loss": 0.8233, + "step": 451 + }, + { + "epoch": 0.04076295260855842, + "grad_norm": 2.251734700656339, + "learning_rate": 3.998791827048219e-06, + "loss": 0.8982, + "step": 452 + }, + { + "epoch": 0.04085313613202868, + "grad_norm": 1.6461360105820793, + "learning_rate": 3.998771438391396e-06, + "loss": 0.7886, + "step": 453 + }, + { + "epoch": 0.04094331965549894, + "grad_norm": 0.6401343443802504, + "learning_rate": 3.9987508791883725e-06, + "loss": 0.5753, + "step": 454 + }, + { + "epoch": 0.0410335031789692, + "grad_norm": 1.7171653378195875, + "learning_rate": 3.998730149440904e-06, + "loss": 0.9693, + "step": 455 + }, + { + "epoch": 0.04112368670243947, + "grad_norm": 1.7657855475697222, + "learning_rate": 3.998709249150758e-06, + "loss": 0.8915, + "step": 456 + }, + { + "epoch": 0.04121387022590973, + "grad_norm": 1.676987505084299, + "learning_rate": 3.998688178319717e-06, + "loss": 0.8593, + "step": 457 + }, + { + "epoch": 0.04130405374937999, + "grad_norm": 1.9063794451679184, + "learning_rate": 3.9986669369495805e-06, + "loss": 0.8345, + "step": 458 + }, + { + "epoch": 0.04139423727285025, + "grad_norm": 1.9296495688722037, + "learning_rate": 3.998645525042161e-06, + "loss": 0.898, + "step": 459 + }, + { + "epoch": 0.04148442079632051, + "grad_norm": 1.8814874949043496, + "learning_rate": 3.998623942599284e-06, + "loss": 0.9153, + "step": 460 + }, + { + "epoch": 0.04157460431979077, + "grad_norm": 2.6838369023361057, + "learning_rate": 3.998602189622793e-06, + "loss": 0.841, + "step": 461 + }, + { + "epoch": 0.04166478784326104, + "grad_norm": 2.098349799328171, + "learning_rate": 3.998580266114542e-06, + "loss": 0.8703, + "step": 462 + }, + { + "epoch": 0.0417549713667313, + "grad_norm": 1.5703505326405247, + "learning_rate": 3.998558172076404e-06, + "loss": 0.7465, + "step": 463 + }, + { + "epoch": 0.04184515489020156, + "grad_norm": 1.777146347318016, + "learning_rate": 3.998535907510262e-06, + "loss": 0.7778, + "step": 464 + }, + { + "epoch": 0.04193533841367182, + "grad_norm": 1.5818557453528546, + "learning_rate": 3.998513472418016e-06, + "loss": 0.8545, + "step": 465 + }, + { + "epoch": 0.04202552193714208, + "grad_norm": 1.662703124522791, + "learning_rate": 3.998490866801582e-06, + "loss": 0.8742, + "step": 466 + }, + { + "epoch": 0.04211570546061234, + "grad_norm": 1.5932659087153762, + "learning_rate": 3.998468090662886e-06, + "loss": 0.8347, + "step": 467 + }, + { + "epoch": 0.04220588898408261, + "grad_norm": 1.5787189234727803, + "learning_rate": 3.998445144003874e-06, + "loss": 0.7357, + "step": 468 + }, + { + "epoch": 0.04229607250755287, + "grad_norm": 4.023486564855566, + "learning_rate": 3.998422026826504e-06, + "loss": 0.8682, + "step": 469 + }, + { + "epoch": 0.04238625603102313, + "grad_norm": 2.3886356428990143, + "learning_rate": 3.998398739132746e-06, + "loss": 0.8855, + "step": 470 + }, + { + "epoch": 0.04247643955449339, + "grad_norm": 1.807790161627835, + "learning_rate": 3.99837528092459e-06, + "loss": 0.796, + "step": 471 + }, + { + "epoch": 0.042566623077963654, + "grad_norm": 2.1765222024028357, + "learning_rate": 3.998351652204034e-06, + "loss": 0.8858, + "step": 472 + }, + { + "epoch": 0.042656806601433915, + "grad_norm": 1.55019184894134, + "learning_rate": 3.998327852973098e-06, + "loss": 0.8812, + "step": 473 + }, + { + "epoch": 0.04274699012490418, + "grad_norm": 2.0415001107093693, + "learning_rate": 3.99830388323381e-06, + "loss": 0.7994, + "step": 474 + }, + { + "epoch": 0.042837173648374444, + "grad_norm": 1.7293730080292882, + "learning_rate": 3.998279742988216e-06, + "loss": 0.8719, + "step": 475 + }, + { + "epoch": 0.042927357171844704, + "grad_norm": 1.7427501368672393, + "learning_rate": 3.998255432238377e-06, + "loss": 0.8141, + "step": 476 + }, + { + "epoch": 0.043017540695314965, + "grad_norm": 1.7233512324222064, + "learning_rate": 3.9982309509863656e-06, + "loss": 0.9178, + "step": 477 + }, + { + "epoch": 0.043107724218785226, + "grad_norm": 2.3935942317788594, + "learning_rate": 3.998206299234272e-06, + "loss": 0.8017, + "step": 478 + }, + { + "epoch": 0.04319790774225549, + "grad_norm": 1.4936757280727115, + "learning_rate": 3.998181476984198e-06, + "loss": 0.8552, + "step": 479 + }, + { + "epoch": 0.043288091265725755, + "grad_norm": 1.6533230524102371, + "learning_rate": 3.998156484238263e-06, + "loss": 0.7704, + "step": 480 + }, + { + "epoch": 0.043378274789196015, + "grad_norm": 1.7555088121828069, + "learning_rate": 3.998131320998599e-06, + "loss": 0.7896, + "step": 481 + }, + { + "epoch": 0.043468458312666276, + "grad_norm": 2.268720839972459, + "learning_rate": 3.998105987267353e-06, + "loss": 0.8985, + "step": 482 + }, + { + "epoch": 0.04355864183613654, + "grad_norm": 1.5681973916437704, + "learning_rate": 3.998080483046687e-06, + "loss": 0.7666, + "step": 483 + }, + { + "epoch": 0.0436488253596068, + "grad_norm": 1.3614820177177467, + "learning_rate": 3.998054808338776e-06, + "loss": 0.7859, + "step": 484 + }, + { + "epoch": 0.04373900888307706, + "grad_norm": 1.621588721027491, + "learning_rate": 3.998028963145812e-06, + "loss": 0.7972, + "step": 485 + }, + { + "epoch": 0.043829192406547327, + "grad_norm": 1.72909022549108, + "learning_rate": 3.99800294747e-06, + "loss": 0.8671, + "step": 486 + }, + { + "epoch": 0.04391937593001759, + "grad_norm": 1.5929100216623746, + "learning_rate": 3.99797676131356e-06, + "loss": 0.8685, + "step": 487 + }, + { + "epoch": 0.04400955945348785, + "grad_norm": 1.906919455621144, + "learning_rate": 3.997950404678726e-06, + "loss": 0.7045, + "step": 488 + }, + { + "epoch": 0.04409974297695811, + "grad_norm": 1.8366586121222719, + "learning_rate": 3.997923877567746e-06, + "loss": 0.9575, + "step": 489 + }, + { + "epoch": 0.04418992650042837, + "grad_norm": 1.7430952640922779, + "learning_rate": 3.9978971799828855e-06, + "loss": 0.6756, + "step": 490 + }, + { + "epoch": 0.04428011002389863, + "grad_norm": 3.5080685936236136, + "learning_rate": 3.997870311926421e-06, + "loss": 0.8365, + "step": 491 + }, + { + "epoch": 0.0443702935473689, + "grad_norm": 1.6890376820536286, + "learning_rate": 3.997843273400645e-06, + "loss": 0.8014, + "step": 492 + }, + { + "epoch": 0.04446047707083916, + "grad_norm": 1.5266252862050302, + "learning_rate": 3.997816064407865e-06, + "loss": 0.8249, + "step": 493 + }, + { + "epoch": 0.04455066059430942, + "grad_norm": 1.5019932060697585, + "learning_rate": 3.997788684950402e-06, + "loss": 0.864, + "step": 494 + }, + { + "epoch": 0.04464084411777968, + "grad_norm": 1.9667687513505654, + "learning_rate": 3.997761135030593e-06, + "loss": 0.8893, + "step": 495 + }, + { + "epoch": 0.04473102764124994, + "grad_norm": 1.6816552988934041, + "learning_rate": 3.997733414650789e-06, + "loss": 0.7412, + "step": 496 + }, + { + "epoch": 0.0448212111647202, + "grad_norm": 1.989355550246059, + "learning_rate": 3.9977055238133554e-06, + "loss": 0.7959, + "step": 497 + }, + { + "epoch": 0.04491139468819047, + "grad_norm": 1.4009238362544774, + "learning_rate": 3.99767746252067e-06, + "loss": 0.7154, + "step": 498 + }, + { + "epoch": 0.04500157821166073, + "grad_norm": 1.6227864080542982, + "learning_rate": 3.997649230775129e-06, + "loss": 0.8468, + "step": 499 + }, + { + "epoch": 0.04509176173513099, + "grad_norm": 1.6453127782229109, + "learning_rate": 3.9976208285791395e-06, + "loss": 0.7835, + "step": 500 + }, + { + "epoch": 0.04518194525860125, + "grad_norm": 0.7900732007095296, + "learning_rate": 3.997592255935127e-06, + "loss": 0.6725, + "step": 501 + }, + { + "epoch": 0.045272128782071513, + "grad_norm": 1.6945748548230712, + "learning_rate": 3.997563512845529e-06, + "loss": 0.8687, + "step": 502 + }, + { + "epoch": 0.045362312305541774, + "grad_norm": 2.4048426091645823, + "learning_rate": 3.9975345993127975e-06, + "loss": 0.8145, + "step": 503 + }, + { + "epoch": 0.04545249582901204, + "grad_norm": 1.8881849195713925, + "learning_rate": 3.9975055153393985e-06, + "loss": 0.8555, + "step": 504 + }, + { + "epoch": 0.0455426793524823, + "grad_norm": 1.689357159485432, + "learning_rate": 3.997476260927816e-06, + "loss": 0.8566, + "step": 505 + }, + { + "epoch": 0.045632862875952564, + "grad_norm": 2.0262849427963174, + "learning_rate": 3.997446836080545e-06, + "loss": 0.816, + "step": 506 + }, + { + "epoch": 0.045723046399422825, + "grad_norm": 2.246194800716018, + "learning_rate": 3.997417240800095e-06, + "loss": 0.8958, + "step": 507 + }, + { + "epoch": 0.045813229922893085, + "grad_norm": 1.586056709507168, + "learning_rate": 3.997387475088994e-06, + "loss": 0.9375, + "step": 508 + }, + { + "epoch": 0.045903413446363346, + "grad_norm": 2.001964319709554, + "learning_rate": 3.99735753894978e-06, + "loss": 0.9435, + "step": 509 + }, + { + "epoch": 0.045993596969833614, + "grad_norm": 1.742041611657312, + "learning_rate": 3.997327432385006e-06, + "loss": 0.8439, + "step": 510 + }, + { + "epoch": 0.046083780493303875, + "grad_norm": 1.89039717633589, + "learning_rate": 3.997297155397244e-06, + "loss": 0.774, + "step": 511 + }, + { + "epoch": 0.046173964016774136, + "grad_norm": 2.129421234838152, + "learning_rate": 3.997266707989074e-06, + "loss": 0.7554, + "step": 512 + }, + { + "epoch": 0.046264147540244396, + "grad_norm": 2.213583505272492, + "learning_rate": 3.997236090163097e-06, + "loss": 0.7697, + "step": 513 + }, + { + "epoch": 0.04635433106371466, + "grad_norm": 1.6968998663048032, + "learning_rate": 3.9972053019219235e-06, + "loss": 0.9742, + "step": 514 + }, + { + "epoch": 0.04644451458718492, + "grad_norm": 1.90599927660437, + "learning_rate": 3.997174343268181e-06, + "loss": 0.8593, + "step": 515 + }, + { + "epoch": 0.046534698110655186, + "grad_norm": 1.5922765465450686, + "learning_rate": 3.9971432142045115e-06, + "loss": 0.8442, + "step": 516 + }, + { + "epoch": 0.04662488163412545, + "grad_norm": 0.8090849043542454, + "learning_rate": 3.99711191473357e-06, + "loss": 0.664, + "step": 517 + }, + { + "epoch": 0.04671506515759571, + "grad_norm": 2.2575871444222577, + "learning_rate": 3.99708044485803e-06, + "loss": 0.8565, + "step": 518 + }, + { + "epoch": 0.04680524868106597, + "grad_norm": 2.288845533692851, + "learning_rate": 3.997048804580574e-06, + "loss": 0.8466, + "step": 519 + }, + { + "epoch": 0.04689543220453623, + "grad_norm": 1.7631798075717449, + "learning_rate": 3.997016993903901e-06, + "loss": 0.7902, + "step": 520 + }, + { + "epoch": 0.04698561572800649, + "grad_norm": 2.714043149291669, + "learning_rate": 3.996985012830728e-06, + "loss": 0.9104, + "step": 521 + }, + { + "epoch": 0.04707579925147676, + "grad_norm": 1.5366921543288699, + "learning_rate": 3.996952861363782e-06, + "loss": 0.7825, + "step": 522 + }, + { + "epoch": 0.04716598277494702, + "grad_norm": 1.533259416263613, + "learning_rate": 3.9969205395058064e-06, + "loss": 0.8476, + "step": 523 + }, + { + "epoch": 0.04725616629841728, + "grad_norm": 1.8802267339621457, + "learning_rate": 3.99688804725956e-06, + "loss": 0.841, + "step": 524 + }, + { + "epoch": 0.04734634982188754, + "grad_norm": 1.2209738469333915, + "learning_rate": 3.996855384627815e-06, + "loss": 0.7902, + "step": 525 + }, + { + "epoch": 0.0474365333453578, + "grad_norm": 1.719372723216328, + "learning_rate": 3.996822551613357e-06, + "loss": 0.8712, + "step": 526 + }, + { + "epoch": 0.04752671686882806, + "grad_norm": 1.6092417393297989, + "learning_rate": 3.996789548218989e-06, + "loss": 0.767, + "step": 527 + }, + { + "epoch": 0.04761690039229833, + "grad_norm": 1.521565700801069, + "learning_rate": 3.996756374447526e-06, + "loss": 0.7276, + "step": 528 + }, + { + "epoch": 0.04770708391576859, + "grad_norm": 2.1085290413390485, + "learning_rate": 3.9967230303018005e-06, + "loss": 0.8555, + "step": 529 + }, + { + "epoch": 0.04779726743923885, + "grad_norm": 1.8467497910659154, + "learning_rate": 3.996689515784655e-06, + "loss": 0.8268, + "step": 530 + }, + { + "epoch": 0.04788745096270911, + "grad_norm": 2.0968909545837233, + "learning_rate": 3.996655830898951e-06, + "loss": 0.8188, + "step": 531 + }, + { + "epoch": 0.04797763448617937, + "grad_norm": 2.2989452141020297, + "learning_rate": 3.996621975647562e-06, + "loss": 0.84, + "step": 532 + }, + { + "epoch": 0.04806781800964964, + "grad_norm": 1.3745344847831549, + "learning_rate": 3.996587950033377e-06, + "loss": 0.7687, + "step": 533 + }, + { + "epoch": 0.0481580015331199, + "grad_norm": 2.231731852157988, + "learning_rate": 3.996553754059299e-06, + "loss": 0.7754, + "step": 534 + }, + { + "epoch": 0.04824818505659016, + "grad_norm": 1.840579717885437, + "learning_rate": 3.996519387728245e-06, + "loss": 0.8495, + "step": 535 + }, + { + "epoch": 0.04833836858006042, + "grad_norm": 1.4833632384010988, + "learning_rate": 3.9964848510431495e-06, + "loss": 0.8132, + "step": 536 + }, + { + "epoch": 0.048428552103530684, + "grad_norm": 1.7273216132268565, + "learning_rate": 3.996450144006957e-06, + "loss": 0.899, + "step": 537 + }, + { + "epoch": 0.048518735627000945, + "grad_norm": 1.987908348983456, + "learning_rate": 3.99641526662263e-06, + "loss": 0.8282, + "step": 538 + }, + { + "epoch": 0.04860891915047121, + "grad_norm": 1.8802827800971542, + "learning_rate": 3.996380218893145e-06, + "loss": 0.9246, + "step": 539 + }, + { + "epoch": 0.04869910267394147, + "grad_norm": 1.782510746285466, + "learning_rate": 3.996345000821491e-06, + "loss": 0.8636, + "step": 540 + }, + { + "epoch": 0.048789286197411734, + "grad_norm": 2.2180641081259007, + "learning_rate": 3.996309612410674e-06, + "loss": 0.9471, + "step": 541 + }, + { + "epoch": 0.048879469720881995, + "grad_norm": 2.1336527416408746, + "learning_rate": 3.996274053663713e-06, + "loss": 0.7724, + "step": 542 + }, + { + "epoch": 0.048969653244352256, + "grad_norm": 0.8012232607279285, + "learning_rate": 3.996238324583643e-06, + "loss": 0.6195, + "step": 543 + }, + { + "epoch": 0.049059836767822516, + "grad_norm": 1.980479704997746, + "learning_rate": 3.996202425173512e-06, + "loss": 0.9155, + "step": 544 + }, + { + "epoch": 0.049150020291292784, + "grad_norm": 1.7999723193901285, + "learning_rate": 3.996166355436383e-06, + "loss": 0.8045, + "step": 545 + }, + { + "epoch": 0.049240203814763045, + "grad_norm": 2.4904765890170286, + "learning_rate": 3.996130115375333e-06, + "loss": 0.8485, + "step": 546 + }, + { + "epoch": 0.049330387338233306, + "grad_norm": 0.8651560108528268, + "learning_rate": 3.996093704993456e-06, + "loss": 0.6346, + "step": 547 + }, + { + "epoch": 0.04942057086170357, + "grad_norm": 1.727971878304182, + "learning_rate": 3.996057124293857e-06, + "loss": 0.9237, + "step": 548 + }, + { + "epoch": 0.04951075438517383, + "grad_norm": 1.4708433451685503, + "learning_rate": 3.996020373279659e-06, + "loss": 0.8834, + "step": 549 + }, + { + "epoch": 0.04960093790864409, + "grad_norm": 1.5166024271317158, + "learning_rate": 3.995983451953996e-06, + "loss": 0.8517, + "step": 550 + }, + { + "epoch": 0.049691121432114356, + "grad_norm": 1.7908374360778496, + "learning_rate": 3.99594636032002e-06, + "loss": 0.8503, + "step": 551 + }, + { + "epoch": 0.04978130495558462, + "grad_norm": 2.0957994821088595, + "learning_rate": 3.995909098380894e-06, + "loss": 0.7208, + "step": 552 + }, + { + "epoch": 0.04987148847905488, + "grad_norm": 1.4340279984287998, + "learning_rate": 3.995871666139799e-06, + "loss": 0.7155, + "step": 553 + }, + { + "epoch": 0.04996167200252514, + "grad_norm": 1.5635742309486447, + "learning_rate": 3.995834063599928e-06, + "loss": 0.8114, + "step": 554 + }, + { + "epoch": 0.0500518555259954, + "grad_norm": 1.491609950914281, + "learning_rate": 3.99579629076449e-06, + "loss": 0.8562, + "step": 555 + }, + { + "epoch": 0.05014203904946566, + "grad_norm": 1.832556220281204, + "learning_rate": 3.9957583476367084e-06, + "loss": 0.8072, + "step": 556 + }, + { + "epoch": 0.05023222257293593, + "grad_norm": 1.8742227850835185, + "learning_rate": 3.995720234219819e-06, + "loss": 0.8367, + "step": 557 + }, + { + "epoch": 0.05032240609640619, + "grad_norm": 1.728277950864836, + "learning_rate": 3.995681950517075e-06, + "loss": 0.752, + "step": 558 + }, + { + "epoch": 0.05041258961987645, + "grad_norm": 1.7823634097718142, + "learning_rate": 3.995643496531743e-06, + "loss": 0.8604, + "step": 559 + }, + { + "epoch": 0.05050277314334671, + "grad_norm": 2.792938445667108, + "learning_rate": 3.9956048722671044e-06, + "loss": 0.8475, + "step": 560 + }, + { + "epoch": 0.05059295666681697, + "grad_norm": 1.6479876297606844, + "learning_rate": 3.995566077726454e-06, + "loss": 0.8138, + "step": 561 + }, + { + "epoch": 0.05068314019028723, + "grad_norm": 2.184221262835891, + "learning_rate": 3.995527112913103e-06, + "loss": 0.8045, + "step": 562 + }, + { + "epoch": 0.0507733237137575, + "grad_norm": 1.791495322380805, + "learning_rate": 3.995487977830375e-06, + "loss": 0.7857, + "step": 563 + }, + { + "epoch": 0.05086350723722776, + "grad_norm": 2.084293131671774, + "learning_rate": 3.9954486724816105e-06, + "loss": 0.9276, + "step": 564 + }, + { + "epoch": 0.05095369076069802, + "grad_norm": 2.8962231014683595, + "learning_rate": 3.995409196870161e-06, + "loss": 0.8838, + "step": 565 + }, + { + "epoch": 0.05104387428416828, + "grad_norm": 0.7994102092084325, + "learning_rate": 3.995369550999398e-06, + "loss": 0.5797, + "step": 566 + }, + { + "epoch": 0.05113405780763854, + "grad_norm": 1.9728904056379009, + "learning_rate": 3.995329734872702e-06, + "loss": 0.868, + "step": 567 + }, + { + "epoch": 0.051224241331108804, + "grad_norm": 1.4884323083505504, + "learning_rate": 3.9952897484934706e-06, + "loss": 0.9494, + "step": 568 + }, + { + "epoch": 0.05131442485457907, + "grad_norm": 2.080915714029, + "learning_rate": 3.995249591865115e-06, + "loss": 0.8501, + "step": 569 + }, + { + "epoch": 0.05140460837804933, + "grad_norm": 1.9138205239066803, + "learning_rate": 3.995209264991063e-06, + "loss": 0.7977, + "step": 570 + }, + { + "epoch": 0.05149479190151959, + "grad_norm": 1.7931643443745693, + "learning_rate": 3.995168767874756e-06, + "loss": 0.8893, + "step": 571 + }, + { + "epoch": 0.051584975424989854, + "grad_norm": 1.567336341430447, + "learning_rate": 3.995128100519648e-06, + "loss": 0.828, + "step": 572 + }, + { + "epoch": 0.051675158948460115, + "grad_norm": 1.9177060073542818, + "learning_rate": 3.995087262929209e-06, + "loss": 0.8471, + "step": 573 + }, + { + "epoch": 0.051765342471930376, + "grad_norm": 1.9009914095432163, + "learning_rate": 3.995046255106925e-06, + "loss": 0.848, + "step": 574 + }, + { + "epoch": 0.05185552599540064, + "grad_norm": 1.9603700462686873, + "learning_rate": 3.995005077056293e-06, + "loss": 0.9041, + "step": 575 + }, + { + "epoch": 0.051945709518870904, + "grad_norm": 3.763564563086439, + "learning_rate": 3.9949637287808284e-06, + "loss": 0.7331, + "step": 576 + }, + { + "epoch": 0.052035893042341165, + "grad_norm": 2.1628181929031802, + "learning_rate": 3.994922210284057e-06, + "loss": 0.7827, + "step": 577 + }, + { + "epoch": 0.052126076565811426, + "grad_norm": 2.1884743836446092, + "learning_rate": 3.994880521569524e-06, + "loss": 0.8757, + "step": 578 + }, + { + "epoch": 0.05221626008928169, + "grad_norm": 1.626217825920974, + "learning_rate": 3.994838662640785e-06, + "loss": 0.845, + "step": 579 + }, + { + "epoch": 0.05230644361275195, + "grad_norm": 2.1452130899644866, + "learning_rate": 3.9947966335014116e-06, + "loss": 0.8805, + "step": 580 + }, + { + "epoch": 0.052396627136222215, + "grad_norm": 2.186385824652905, + "learning_rate": 3.99475443415499e-06, + "loss": 0.824, + "step": 581 + }, + { + "epoch": 0.052486810659692476, + "grad_norm": 2.0761129634677533, + "learning_rate": 3.994712064605121e-06, + "loss": 0.8325, + "step": 582 + }, + { + "epoch": 0.05257699418316274, + "grad_norm": 1.9113799312079638, + "learning_rate": 3.99466952485542e-06, + "loss": 0.9855, + "step": 583 + }, + { + "epoch": 0.052667177706633, + "grad_norm": 2.505178428889106, + "learning_rate": 3.994626814909518e-06, + "loss": 0.8521, + "step": 584 + }, + { + "epoch": 0.05275736123010326, + "grad_norm": 1.9172445821930515, + "learning_rate": 3.994583934771056e-06, + "loss": 0.824, + "step": 585 + }, + { + "epoch": 0.05284754475357352, + "grad_norm": 1.7567572923310508, + "learning_rate": 3.9945408844436955e-06, + "loss": 0.8944, + "step": 586 + }, + { + "epoch": 0.05293772827704379, + "grad_norm": 2.1834533318698646, + "learning_rate": 3.994497663931109e-06, + "loss": 0.8709, + "step": 587 + }, + { + "epoch": 0.05302791180051405, + "grad_norm": 1.823421982255428, + "learning_rate": 3.994454273236984e-06, + "loss": 0.856, + "step": 588 + }, + { + "epoch": 0.05311809532398431, + "grad_norm": 1.635426377512616, + "learning_rate": 3.994410712365023e-06, + "loss": 0.8508, + "step": 589 + }, + { + "epoch": 0.05320827884745457, + "grad_norm": 2.128904346150238, + "learning_rate": 3.994366981318943e-06, + "loss": 0.9349, + "step": 590 + }, + { + "epoch": 0.05329846237092483, + "grad_norm": 1.764295837269606, + "learning_rate": 3.9943230801024765e-06, + "loss": 0.8419, + "step": 591 + }, + { + "epoch": 0.05338864589439509, + "grad_norm": 1.7542788103490554, + "learning_rate": 3.9942790087193666e-06, + "loss": 0.8746, + "step": 592 + }, + { + "epoch": 0.05347882941786536, + "grad_norm": 1.790959447813763, + "learning_rate": 3.994234767173376e-06, + "loss": 0.8313, + "step": 593 + }, + { + "epoch": 0.05356901294133562, + "grad_norm": 3.972048732700636, + "learning_rate": 3.994190355468279e-06, + "loss": 0.7353, + "step": 594 + }, + { + "epoch": 0.05365919646480588, + "grad_norm": 1.7800880791806661, + "learning_rate": 3.994145773607865e-06, + "loss": 0.8415, + "step": 595 + }, + { + "epoch": 0.05374937998827614, + "grad_norm": 1.7544651967486198, + "learning_rate": 3.994101021595938e-06, + "loss": 0.8093, + "step": 596 + }, + { + "epoch": 0.0538395635117464, + "grad_norm": 1.7213610150253411, + "learning_rate": 3.9940560994363165e-06, + "loss": 0.8512, + "step": 597 + }, + { + "epoch": 0.05392974703521666, + "grad_norm": 1.9336557089386568, + "learning_rate": 3.994011007132833e-06, + "loss": 0.7232, + "step": 598 + }, + { + "epoch": 0.05401993055868693, + "grad_norm": 0.7428637585830059, + "learning_rate": 3.993965744689337e-06, + "loss": 0.6501, + "step": 599 + }, + { + "epoch": 0.05411011408215719, + "grad_norm": 1.7786994877584226, + "learning_rate": 3.993920312109687e-06, + "loss": 0.8293, + "step": 600 + }, + { + "epoch": 0.05420029760562745, + "grad_norm": 1.7108092238184023, + "learning_rate": 3.993874709397764e-06, + "loss": 0.8267, + "step": 601 + }, + { + "epoch": 0.05429048112909771, + "grad_norm": 2.2318663485788526, + "learning_rate": 3.993828936557454e-06, + "loss": 0.8254, + "step": 602 + }, + { + "epoch": 0.054380664652567974, + "grad_norm": 1.672554163376511, + "learning_rate": 3.993782993592667e-06, + "loss": 0.8008, + "step": 603 + }, + { + "epoch": 0.054470848176038235, + "grad_norm": 1.7808783294010428, + "learning_rate": 3.993736880507321e-06, + "loss": 0.8848, + "step": 604 + }, + { + "epoch": 0.0545610316995085, + "grad_norm": 1.9859915929310963, + "learning_rate": 3.99369059730535e-06, + "loss": 0.7803, + "step": 605 + }, + { + "epoch": 0.05465121522297876, + "grad_norm": 1.3370359043416136, + "learning_rate": 3.993644143990706e-06, + "loss": 0.8339, + "step": 606 + }, + { + "epoch": 0.054741398746449024, + "grad_norm": 1.7798471782746081, + "learning_rate": 3.99359752056735e-06, + "loss": 0.7608, + "step": 607 + }, + { + "epoch": 0.054831582269919285, + "grad_norm": 2.1472597535365194, + "learning_rate": 3.993550727039261e-06, + "loss": 0.6934, + "step": 608 + }, + { + "epoch": 0.054921765793389546, + "grad_norm": 2.0592832044133718, + "learning_rate": 3.993503763410431e-06, + "loss": 0.8592, + "step": 609 + }, + { + "epoch": 0.05501194931685981, + "grad_norm": 1.702607732492725, + "learning_rate": 3.9934566296848686e-06, + "loss": 0.7506, + "step": 610 + }, + { + "epoch": 0.055102132840330074, + "grad_norm": 1.861834078028839, + "learning_rate": 3.993409325866595e-06, + "loss": 0.9209, + "step": 611 + }, + { + "epoch": 0.055192316363800335, + "grad_norm": 0.6357401014255689, + "learning_rate": 3.993361851959645e-06, + "loss": 0.5977, + "step": 612 + }, + { + "epoch": 0.055282499887270596, + "grad_norm": 1.5874277263619687, + "learning_rate": 3.993314207968071e-06, + "loss": 0.8773, + "step": 613 + }, + { + "epoch": 0.05537268341074086, + "grad_norm": 2.01224156965399, + "learning_rate": 3.993266393895938e-06, + "loss": 0.8194, + "step": 614 + }, + { + "epoch": 0.05546286693421112, + "grad_norm": 3.1137719431762063, + "learning_rate": 3.993218409747326e-06, + "loss": 0.8463, + "step": 615 + }, + { + "epoch": 0.05555305045768138, + "grad_norm": 1.8291744252288726, + "learning_rate": 3.993170255526328e-06, + "loss": 0.8318, + "step": 616 + }, + { + "epoch": 0.055643233981151646, + "grad_norm": 1.9759351013545496, + "learning_rate": 3.993121931237054e-06, + "loss": 0.8668, + "step": 617 + }, + { + "epoch": 0.05573341750462191, + "grad_norm": 1.8419701989214157, + "learning_rate": 3.993073436883627e-06, + "loss": 0.857, + "step": 618 + }, + { + "epoch": 0.05582360102809217, + "grad_norm": 1.5280376916450762, + "learning_rate": 3.993024772470184e-06, + "loss": 0.8192, + "step": 619 + }, + { + "epoch": 0.05591378455156243, + "grad_norm": 1.832770485164712, + "learning_rate": 3.992975938000878e-06, + "loss": 0.8205, + "step": 620 + }, + { + "epoch": 0.05600396807503269, + "grad_norm": 0.6685435850005554, + "learning_rate": 3.992926933479876e-06, + "loss": 0.5881, + "step": 621 + }, + { + "epoch": 0.05609415159850295, + "grad_norm": 1.7332149440731228, + "learning_rate": 3.9928777589113595e-06, + "loss": 0.8577, + "step": 622 + }, + { + "epoch": 0.05618433512197322, + "grad_norm": 1.7779604334510597, + "learning_rate": 3.992828414299524e-06, + "loss": 0.7927, + "step": 623 + }, + { + "epoch": 0.05627451864544348, + "grad_norm": 1.705718419475526, + "learning_rate": 3.992778899648579e-06, + "loss": 0.8315, + "step": 624 + }, + { + "epoch": 0.05636470216891374, + "grad_norm": 1.8068514974062475, + "learning_rate": 3.992729214962751e-06, + "loss": 0.8605, + "step": 625 + }, + { + "epoch": 0.056454885692384, + "grad_norm": 2.3477876597867144, + "learning_rate": 3.992679360246279e-06, + "loss": 0.8783, + "step": 626 + }, + { + "epoch": 0.05654506921585426, + "grad_norm": 1.8680049112569872, + "learning_rate": 3.992629335503416e-06, + "loss": 0.8685, + "step": 627 + }, + { + "epoch": 0.05663525273932452, + "grad_norm": 1.4769895228774297, + "learning_rate": 3.9925791407384304e-06, + "loss": 0.817, + "step": 628 + }, + { + "epoch": 0.05672543626279479, + "grad_norm": 1.9601477138170735, + "learning_rate": 3.992528775955606e-06, + "loss": 0.8807, + "step": 629 + }, + { + "epoch": 0.05681561978626505, + "grad_norm": 1.834799555918954, + "learning_rate": 3.992478241159239e-06, + "loss": 0.8097, + "step": 630 + }, + { + "epoch": 0.05690580330973531, + "grad_norm": 1.9274769836318202, + "learning_rate": 3.992427536353643e-06, + "loss": 0.8545, + "step": 631 + }, + { + "epoch": 0.05699598683320557, + "grad_norm": 2.037084328779993, + "learning_rate": 3.992376661543143e-06, + "loss": 0.8321, + "step": 632 + }, + { + "epoch": 0.05708617035667583, + "grad_norm": 2.374928925103232, + "learning_rate": 3.992325616732081e-06, + "loss": 0.8283, + "step": 633 + }, + { + "epoch": 0.057176353880146094, + "grad_norm": 1.8797772900578635, + "learning_rate": 3.992274401924811e-06, + "loss": 0.9358, + "step": 634 + }, + { + "epoch": 0.05726653740361636, + "grad_norm": 2.652929317235903, + "learning_rate": 3.992223017125704e-06, + "loss": 0.7571, + "step": 635 + }, + { + "epoch": 0.05735672092708662, + "grad_norm": 1.6428134672507837, + "learning_rate": 3.992171462339145e-06, + "loss": 0.8609, + "step": 636 + }, + { + "epoch": 0.057446904450556883, + "grad_norm": 1.7740674349494636, + "learning_rate": 3.992119737569532e-06, + "loss": 0.9037, + "step": 637 + }, + { + "epoch": 0.057537087974027144, + "grad_norm": 2.10979677857831, + "learning_rate": 3.992067842821277e-06, + "loss": 0.8318, + "step": 638 + }, + { + "epoch": 0.057627271497497405, + "grad_norm": 2.580640825337182, + "learning_rate": 3.99201577809881e-06, + "loss": 0.9665, + "step": 639 + }, + { + "epoch": 0.057717455020967666, + "grad_norm": 2.230763003419031, + "learning_rate": 3.991963543406574e-06, + "loss": 0.8587, + "step": 640 + }, + { + "epoch": 0.057807638544437934, + "grad_norm": 1.95546832082746, + "learning_rate": 3.991911138749024e-06, + "loss": 0.8328, + "step": 641 + }, + { + "epoch": 0.057897822067908195, + "grad_norm": 2.333976509368284, + "learning_rate": 3.991858564130633e-06, + "loss": 0.8571, + "step": 642 + }, + { + "epoch": 0.057988005591378455, + "grad_norm": 2.7628781745939697, + "learning_rate": 3.991805819555885e-06, + "loss": 0.8195, + "step": 643 + }, + { + "epoch": 0.058078189114848716, + "grad_norm": 1.7445112392099567, + "learning_rate": 3.991752905029283e-06, + "loss": 0.8321, + "step": 644 + }, + { + "epoch": 0.05816837263831898, + "grad_norm": 1.9291595478077563, + "learning_rate": 3.991699820555341e-06, + "loss": 0.7067, + "step": 645 + }, + { + "epoch": 0.05825855616178924, + "grad_norm": 3.1286110423218454, + "learning_rate": 3.991646566138588e-06, + "loss": 0.8364, + "step": 646 + }, + { + "epoch": 0.058348739685259506, + "grad_norm": 1.9094123874131825, + "learning_rate": 3.991593141783567e-06, + "loss": 0.7599, + "step": 647 + }, + { + "epoch": 0.058438923208729766, + "grad_norm": 1.7525230339811226, + "learning_rate": 3.991539547494839e-06, + "loss": 0.8499, + "step": 648 + }, + { + "epoch": 0.05852910673220003, + "grad_norm": 2.9434714799072443, + "learning_rate": 3.991485783276974e-06, + "loss": 0.8328, + "step": 649 + }, + { + "epoch": 0.05861929025567029, + "grad_norm": 1.5456762581920414, + "learning_rate": 3.991431849134563e-06, + "loss": 0.8852, + "step": 650 + }, + { + "epoch": 0.05870947377914055, + "grad_norm": 1.9824992648648876, + "learning_rate": 3.991377745072205e-06, + "loss": 0.7286, + "step": 651 + }, + { + "epoch": 0.05879965730261081, + "grad_norm": 2.116281463335079, + "learning_rate": 3.991323471094517e-06, + "loss": 0.8662, + "step": 652 + }, + { + "epoch": 0.05888984082608108, + "grad_norm": 1.8225462473586196, + "learning_rate": 3.991269027206131e-06, + "loss": 0.8026, + "step": 653 + }, + { + "epoch": 0.05898002434955134, + "grad_norm": 0.8019105716941335, + "learning_rate": 3.9912144134116916e-06, + "loss": 0.6398, + "step": 654 + }, + { + "epoch": 0.0590702078730216, + "grad_norm": 1.7605578875657484, + "learning_rate": 3.99115962971586e-06, + "loss": 0.7611, + "step": 655 + }, + { + "epoch": 0.05916039139649186, + "grad_norm": 2.7374882876350295, + "learning_rate": 3.991104676123308e-06, + "loss": 0.8625, + "step": 656 + }, + { + "epoch": 0.05925057491996212, + "grad_norm": 0.6548607972424014, + "learning_rate": 3.991049552638727e-06, + "loss": 0.5742, + "step": 657 + }, + { + "epoch": 0.05934075844343238, + "grad_norm": 2.291141022017147, + "learning_rate": 3.99099425926682e-06, + "loss": 0.7994, + "step": 658 + }, + { + "epoch": 0.05943094196690265, + "grad_norm": 1.9313723972879013, + "learning_rate": 3.990938796012304e-06, + "loss": 0.809, + "step": 659 + }, + { + "epoch": 0.05952112549037291, + "grad_norm": 1.9397629474060785, + "learning_rate": 3.990883162879912e-06, + "loss": 0.9845, + "step": 660 + }, + { + "epoch": 0.05961130901384317, + "grad_norm": 2.6320230902264803, + "learning_rate": 3.990827359874391e-06, + "loss": 0.9007, + "step": 661 + }, + { + "epoch": 0.05970149253731343, + "grad_norm": 1.9510592102576365, + "learning_rate": 3.990771387000503e-06, + "loss": 0.8328, + "step": 662 + }, + { + "epoch": 0.05979167606078369, + "grad_norm": 1.9596075890340592, + "learning_rate": 3.990715244263023e-06, + "loss": 0.8602, + "step": 663 + }, + { + "epoch": 0.05988185958425396, + "grad_norm": 0.7431389181567802, + "learning_rate": 3.990658931666741e-06, + "loss": 0.6072, + "step": 664 + }, + { + "epoch": 0.05997204310772422, + "grad_norm": 1.722046828653354, + "learning_rate": 3.990602449216463e-06, + "loss": 0.7643, + "step": 665 + }, + { + "epoch": 0.06006222663119448, + "grad_norm": 1.9546608951261348, + "learning_rate": 3.990545796917008e-06, + "loss": 0.7372, + "step": 666 + }, + { + "epoch": 0.06015241015466474, + "grad_norm": 3.591405618006593, + "learning_rate": 3.99048897477321e-06, + "loss": 0.8827, + "step": 667 + }, + { + "epoch": 0.060242593678135004, + "grad_norm": 2.5639228592157437, + "learning_rate": 3.990431982789917e-06, + "loss": 0.8519, + "step": 668 + }, + { + "epoch": 0.060332777201605264, + "grad_norm": 2.0153998907196904, + "learning_rate": 3.990374820971992e-06, + "loss": 0.7686, + "step": 669 + }, + { + "epoch": 0.06042296072507553, + "grad_norm": 1.9782652625906805, + "learning_rate": 3.990317489324312e-06, + "loss": 0.7244, + "step": 670 + }, + { + "epoch": 0.06051314424854579, + "grad_norm": 2.2595426800608034, + "learning_rate": 3.99025998785177e-06, + "loss": 0.8255, + "step": 671 + }, + { + "epoch": 0.060603327772016054, + "grad_norm": 1.656942834433219, + "learning_rate": 3.990202316559271e-06, + "loss": 0.7771, + "step": 672 + }, + { + "epoch": 0.060693511295486315, + "grad_norm": 2.163553226190896, + "learning_rate": 3.990144475451738e-06, + "loss": 0.8777, + "step": 673 + }, + { + "epoch": 0.060783694818956575, + "grad_norm": 1.638942618845282, + "learning_rate": 3.9900864645341036e-06, + "loss": 0.8296, + "step": 674 + }, + { + "epoch": 0.060873878342426836, + "grad_norm": 1.8491157428914857, + "learning_rate": 3.990028283811319e-06, + "loss": 0.8429, + "step": 675 + }, + { + "epoch": 0.060964061865897104, + "grad_norm": 1.684283405141982, + "learning_rate": 3.989969933288348e-06, + "loss": 0.8231, + "step": 676 + }, + { + "epoch": 0.061054245389367365, + "grad_norm": 0.6732877471886547, + "learning_rate": 3.98991141297017e-06, + "loss": 0.5512, + "step": 677 + }, + { + "epoch": 0.061144428912837626, + "grad_norm": 1.8272788543072225, + "learning_rate": 3.989852722861778e-06, + "loss": 0.7819, + "step": 678 + }, + { + "epoch": 0.061234612436307886, + "grad_norm": 0.7261433776878155, + "learning_rate": 3.98979386296818e-06, + "loss": 0.5798, + "step": 679 + }, + { + "epoch": 0.06132479595977815, + "grad_norm": 2.136228069195081, + "learning_rate": 3.989734833294398e-06, + "loss": 0.8103, + "step": 680 + }, + { + "epoch": 0.06141497948324841, + "grad_norm": 1.6748698853016701, + "learning_rate": 3.989675633845469e-06, + "loss": 0.8959, + "step": 681 + }, + { + "epoch": 0.061505163006718676, + "grad_norm": 2.017504503323092, + "learning_rate": 3.989616264626443e-06, + "loss": 0.8122, + "step": 682 + }, + { + "epoch": 0.06159534653018894, + "grad_norm": 1.5708216510746416, + "learning_rate": 3.989556725642388e-06, + "loss": 0.8646, + "step": 683 + }, + { + "epoch": 0.0616855300536592, + "grad_norm": 1.7333669396345983, + "learning_rate": 3.989497016898382e-06, + "loss": 0.886, + "step": 684 + }, + { + "epoch": 0.06177571357712946, + "grad_norm": 2.127131458999647, + "learning_rate": 3.98943713839952e-06, + "loss": 0.8247, + "step": 685 + }, + { + "epoch": 0.06186589710059972, + "grad_norm": 0.6880080773217647, + "learning_rate": 3.9893770901509125e-06, + "loss": 0.6085, + "step": 686 + }, + { + "epoch": 0.06195608062406998, + "grad_norm": 1.5143573611653471, + "learning_rate": 3.989316872157682e-06, + "loss": 0.8172, + "step": 687 + }, + { + "epoch": 0.06204626414754025, + "grad_norm": 5.221632475503089, + "learning_rate": 3.989256484424968e-06, + "loss": 0.9045, + "step": 688 + }, + { + "epoch": 0.06213644767101051, + "grad_norm": 1.995454211717879, + "learning_rate": 3.98919592695792e-06, + "loss": 0.8056, + "step": 689 + }, + { + "epoch": 0.06222663119448077, + "grad_norm": 2.6436476703436353, + "learning_rate": 3.9891351997617096e-06, + "loss": 0.8572, + "step": 690 + }, + { + "epoch": 0.06231681471795103, + "grad_norm": 2.012062532253287, + "learning_rate": 3.989074302841514e-06, + "loss": 0.7802, + "step": 691 + }, + { + "epoch": 0.06240699824142129, + "grad_norm": 2.411883608488095, + "learning_rate": 3.989013236202533e-06, + "loss": 0.9462, + "step": 692 + }, + { + "epoch": 0.06249718176489155, + "grad_norm": 1.6076191332678664, + "learning_rate": 3.988951999849974e-06, + "loss": 0.8696, + "step": 693 + }, + { + "epoch": 0.06258736528836181, + "grad_norm": 2.061658861129569, + "learning_rate": 3.988890593789064e-06, + "loss": 0.7526, + "step": 694 + }, + { + "epoch": 0.06267754881183207, + "grad_norm": 1.5125538493256976, + "learning_rate": 3.9888290180250415e-06, + "loss": 0.8176, + "step": 695 + }, + { + "epoch": 0.06276773233530233, + "grad_norm": 2.509113295270546, + "learning_rate": 3.988767272563161e-06, + "loss": 0.7952, + "step": 696 + }, + { + "epoch": 0.06285791585877261, + "grad_norm": 1.8831659515516348, + "learning_rate": 3.988705357408691e-06, + "loss": 0.7735, + "step": 697 + }, + { + "epoch": 0.06294809938224287, + "grad_norm": 1.787059679017872, + "learning_rate": 3.9886432725669146e-06, + "loss": 0.79, + "step": 698 + }, + { + "epoch": 0.06303828290571313, + "grad_norm": 1.9917711009759282, + "learning_rate": 3.988581018043128e-06, + "loss": 0.7898, + "step": 699 + }, + { + "epoch": 0.06312846642918339, + "grad_norm": 2.734750950717665, + "learning_rate": 3.988518593842645e-06, + "loss": 0.8509, + "step": 700 + }, + { + "epoch": 0.06321864995265365, + "grad_norm": 2.0274848668372116, + "learning_rate": 3.9884559999707906e-06, + "loss": 0.8273, + "step": 701 + }, + { + "epoch": 0.06330883347612391, + "grad_norm": 2.586999680804044, + "learning_rate": 3.988393236432906e-06, + "loss": 0.8412, + "step": 702 + }, + { + "epoch": 0.06339901699959417, + "grad_norm": 1.9405658215331518, + "learning_rate": 3.988330303234347e-06, + "loss": 0.9076, + "step": 703 + }, + { + "epoch": 0.06348920052306443, + "grad_norm": 2.4087705415523883, + "learning_rate": 3.988267200380483e-06, + "loss": 0.778, + "step": 704 + }, + { + "epoch": 0.0635793840465347, + "grad_norm": 1.5623168838011565, + "learning_rate": 3.988203927876698e-06, + "loss": 0.8803, + "step": 705 + }, + { + "epoch": 0.06366956757000496, + "grad_norm": 1.7236324666917147, + "learning_rate": 3.988140485728391e-06, + "loss": 0.7203, + "step": 706 + }, + { + "epoch": 0.06375975109347522, + "grad_norm": 1.8810013413452362, + "learning_rate": 3.988076873940975e-06, + "loss": 0.7784, + "step": 707 + }, + { + "epoch": 0.06384993461694548, + "grad_norm": 2.095210121378825, + "learning_rate": 3.9880130925198786e-06, + "loss": 0.8233, + "step": 708 + }, + { + "epoch": 0.06394011814041575, + "grad_norm": 2.8906517233773177, + "learning_rate": 3.987949141470543e-06, + "loss": 0.8418, + "step": 709 + }, + { + "epoch": 0.06403030166388601, + "grad_norm": 1.6901167608026002, + "learning_rate": 3.987885020798425e-06, + "loss": 0.834, + "step": 710 + }, + { + "epoch": 0.06412048518735627, + "grad_norm": 3.3026835540885937, + "learning_rate": 3.987820730508996e-06, + "loss": 0.8001, + "step": 711 + }, + { + "epoch": 0.06421066871082654, + "grad_norm": 1.916183479881443, + "learning_rate": 3.987756270607742e-06, + "loss": 0.9289, + "step": 712 + }, + { + "epoch": 0.0643008522342968, + "grad_norm": 0.6637457597097813, + "learning_rate": 3.987691641100162e-06, + "loss": 0.566, + "step": 713 + }, + { + "epoch": 0.06439103575776706, + "grad_norm": 2.247994164612008, + "learning_rate": 3.987626841991771e-06, + "loss": 0.9219, + "step": 714 + }, + { + "epoch": 0.06448121928123732, + "grad_norm": 2.519627106491717, + "learning_rate": 3.987561873288099e-06, + "loss": 0.8665, + "step": 715 + }, + { + "epoch": 0.06457140280470758, + "grad_norm": 1.8942974772208299, + "learning_rate": 3.987496734994688e-06, + "loss": 0.7806, + "step": 716 + }, + { + "epoch": 0.06466158632817784, + "grad_norm": 1.993756203955681, + "learning_rate": 3.987431427117097e-06, + "loss": 0.8642, + "step": 717 + }, + { + "epoch": 0.0647517698516481, + "grad_norm": 2.3122811084972206, + "learning_rate": 3.9873659496608985e-06, + "loss": 0.8462, + "step": 718 + }, + { + "epoch": 0.06484195337511836, + "grad_norm": 1.9117016615764462, + "learning_rate": 3.987300302631678e-06, + "loss": 0.7817, + "step": 719 + }, + { + "epoch": 0.06493213689858862, + "grad_norm": 2.609139437521719, + "learning_rate": 3.987234486035039e-06, + "loss": 0.87, + "step": 720 + }, + { + "epoch": 0.0650223204220589, + "grad_norm": 2.6841518926662946, + "learning_rate": 3.987168499876595e-06, + "loss": 0.8891, + "step": 721 + }, + { + "epoch": 0.06511250394552916, + "grad_norm": 2.1852982475943006, + "learning_rate": 3.987102344161978e-06, + "loss": 0.8429, + "step": 722 + }, + { + "epoch": 0.06520268746899942, + "grad_norm": 1.8537108757638805, + "learning_rate": 3.987036018896832e-06, + "loss": 0.8651, + "step": 723 + }, + { + "epoch": 0.06529287099246968, + "grad_norm": 1.8283000112375727, + "learning_rate": 3.986969524086817e-06, + "loss": 0.8949, + "step": 724 + }, + { + "epoch": 0.06538305451593994, + "grad_norm": 2.385191683313922, + "learning_rate": 3.986902859737605e-06, + "loss": 0.8122, + "step": 725 + }, + { + "epoch": 0.0654732380394102, + "grad_norm": 1.7210860589436077, + "learning_rate": 3.986836025854886e-06, + "loss": 0.7639, + "step": 726 + }, + { + "epoch": 0.06556342156288046, + "grad_norm": 1.888697500755251, + "learning_rate": 3.986769022444362e-06, + "loss": 0.7908, + "step": 727 + }, + { + "epoch": 0.06565360508635072, + "grad_norm": 6.748365416233447, + "learning_rate": 3.986701849511751e-06, + "loss": 0.8231, + "step": 728 + }, + { + "epoch": 0.06574378860982098, + "grad_norm": 2.310280327404583, + "learning_rate": 3.986634507062782e-06, + "loss": 0.8071, + "step": 729 + }, + { + "epoch": 0.06583397213329124, + "grad_norm": 1.853660457246493, + "learning_rate": 3.986566995103204e-06, + "loss": 0.8512, + "step": 730 + }, + { + "epoch": 0.0659241556567615, + "grad_norm": 2.236333348963604, + "learning_rate": 3.986499313638776e-06, + "loss": 0.7657, + "step": 731 + }, + { + "epoch": 0.06601433918023177, + "grad_norm": 1.7119414985055477, + "learning_rate": 3.986431462675272e-06, + "loss": 0.7605, + "step": 732 + }, + { + "epoch": 0.06610452270370204, + "grad_norm": 1.4489237441380083, + "learning_rate": 3.9863634422184835e-06, + "loss": 0.8427, + "step": 733 + }, + { + "epoch": 0.0661947062271723, + "grad_norm": 1.8408795658359824, + "learning_rate": 3.986295252274213e-06, + "loss": 0.862, + "step": 734 + }, + { + "epoch": 0.06628488975064256, + "grad_norm": 1.809876944537132, + "learning_rate": 3.9862268928482796e-06, + "loss": 0.8716, + "step": 735 + }, + { + "epoch": 0.06637507327411282, + "grad_norm": 1.6573051114818735, + "learning_rate": 3.986158363946515e-06, + "loss": 0.7371, + "step": 736 + }, + { + "epoch": 0.06646525679758308, + "grad_norm": 1.9504754367192105, + "learning_rate": 3.9860896655747685e-06, + "loss": 0.8358, + "step": 737 + }, + { + "epoch": 0.06655544032105334, + "grad_norm": 2.1319035979371153, + "learning_rate": 3.9860207977388994e-06, + "loss": 0.9007, + "step": 738 + }, + { + "epoch": 0.0666456238445236, + "grad_norm": 1.7414419862786406, + "learning_rate": 3.9859517604447854e-06, + "loss": 0.7613, + "step": 739 + }, + { + "epoch": 0.06673580736799387, + "grad_norm": 1.6207494898719814, + "learning_rate": 3.985882553698317e-06, + "loss": 0.7956, + "step": 740 + }, + { + "epoch": 0.06682599089146413, + "grad_norm": 2.0499625179306706, + "learning_rate": 3.985813177505399e-06, + "loss": 0.8946, + "step": 741 + }, + { + "epoch": 0.06691617441493439, + "grad_norm": 1.774769891679272, + "learning_rate": 3.985743631871951e-06, + "loss": 0.7754, + "step": 742 + }, + { + "epoch": 0.06700635793840465, + "grad_norm": 1.9450910204695762, + "learning_rate": 3.985673916803907e-06, + "loss": 0.8234, + "step": 743 + }, + { + "epoch": 0.06709654146187491, + "grad_norm": 1.9624831958221394, + "learning_rate": 3.985604032307215e-06, + "loss": 0.8521, + "step": 744 + }, + { + "epoch": 0.06718672498534518, + "grad_norm": 1.9689685155309056, + "learning_rate": 3.985533978387839e-06, + "loss": 0.8347, + "step": 745 + }, + { + "epoch": 0.06727690850881544, + "grad_norm": 3.3627922243106663, + "learning_rate": 3.985463755051756e-06, + "loss": 0.871, + "step": 746 + }, + { + "epoch": 0.0673670920322857, + "grad_norm": 8.13859195912588, + "learning_rate": 3.9853933623049575e-06, + "loss": 0.8249, + "step": 747 + }, + { + "epoch": 0.06745727555575597, + "grad_norm": 1.6698542709497457, + "learning_rate": 3.98532280015345e-06, + "loss": 0.8263, + "step": 748 + }, + { + "epoch": 0.06754745907922623, + "grad_norm": 1.722761293698909, + "learning_rate": 3.985252068603254e-06, + "loss": 0.7534, + "step": 749 + }, + { + "epoch": 0.06763764260269649, + "grad_norm": 1.6884906121767624, + "learning_rate": 3.985181167660406e-06, + "loss": 0.8602, + "step": 750 + }, + { + "epoch": 0.06772782612616675, + "grad_norm": 1.8883263983042213, + "learning_rate": 3.985110097330953e-06, + "loss": 0.8485, + "step": 751 + }, + { + "epoch": 0.06781800964963701, + "grad_norm": 1.62872254870359, + "learning_rate": 3.985038857620962e-06, + "loss": 0.8179, + "step": 752 + }, + { + "epoch": 0.06790819317310727, + "grad_norm": 0.6909351360223669, + "learning_rate": 3.9849674485365094e-06, + "loss": 0.5906, + "step": 753 + }, + { + "epoch": 0.06799837669657753, + "grad_norm": 2.3906944115229596, + "learning_rate": 3.98489587008369e-06, + "loss": 0.7594, + "step": 754 + }, + { + "epoch": 0.06808856022004779, + "grad_norm": 2.506198637095939, + "learning_rate": 3.98482412226861e-06, + "loss": 0.6741, + "step": 755 + }, + { + "epoch": 0.06817874374351805, + "grad_norm": 1.756351457156368, + "learning_rate": 3.984752205097391e-06, + "loss": 0.8582, + "step": 756 + }, + { + "epoch": 0.06826892726698833, + "grad_norm": 1.6852087077260136, + "learning_rate": 3.984680118576171e-06, + "loss": 0.8348, + "step": 757 + }, + { + "epoch": 0.06835911079045859, + "grad_norm": 1.729882272231625, + "learning_rate": 3.984607862711099e-06, + "loss": 0.7861, + "step": 758 + }, + { + "epoch": 0.06844929431392885, + "grad_norm": 4.822446378624342, + "learning_rate": 3.984535437508341e-06, + "loss": 0.8734, + "step": 759 + }, + { + "epoch": 0.06853947783739911, + "grad_norm": 0.6694648228991199, + "learning_rate": 3.984462842974078e-06, + "loss": 0.5951, + "step": 760 + }, + { + "epoch": 0.06862966136086937, + "grad_norm": 2.1871514178876748, + "learning_rate": 3.984390079114502e-06, + "loss": 0.838, + "step": 761 + }, + { + "epoch": 0.06871984488433963, + "grad_norm": 1.5138826088186927, + "learning_rate": 3.984317145935824e-06, + "loss": 0.9445, + "step": 762 + }, + { + "epoch": 0.06881002840780989, + "grad_norm": 1.4845295875764049, + "learning_rate": 3.984244043444264e-06, + "loss": 0.7998, + "step": 763 + }, + { + "epoch": 0.06890021193128015, + "grad_norm": 1.5461117952054164, + "learning_rate": 3.984170771646062e-06, + "loss": 0.7782, + "step": 764 + }, + { + "epoch": 0.06899039545475041, + "grad_norm": 2.2933022875495674, + "learning_rate": 3.9840973305474695e-06, + "loss": 0.8698, + "step": 765 + }, + { + "epoch": 0.06908057897822067, + "grad_norm": 4.040288447667622, + "learning_rate": 3.984023720154752e-06, + "loss": 0.8798, + "step": 766 + }, + { + "epoch": 0.06917076250169094, + "grad_norm": 1.3777421570902317, + "learning_rate": 3.9839499404741915e-06, + "loss": 0.839, + "step": 767 + }, + { + "epoch": 0.0692609460251612, + "grad_norm": 0.7470464166546997, + "learning_rate": 3.983875991512082e-06, + "loss": 0.6111, + "step": 768 + }, + { + "epoch": 0.06935112954863147, + "grad_norm": 2.653211314096975, + "learning_rate": 3.9838018732747345e-06, + "loss": 0.841, + "step": 769 + }, + { + "epoch": 0.06944131307210173, + "grad_norm": 1.736215731541454, + "learning_rate": 3.9837275857684716e-06, + "loss": 0.8112, + "step": 770 + }, + { + "epoch": 0.06953149659557199, + "grad_norm": 2.4760489427724153, + "learning_rate": 3.983653128999634e-06, + "loss": 0.959, + "step": 771 + }, + { + "epoch": 0.06962168011904225, + "grad_norm": 1.8628927341608421, + "learning_rate": 3.983578502974574e-06, + "loss": 0.8535, + "step": 772 + }, + { + "epoch": 0.06971186364251251, + "grad_norm": 2.1889798199806947, + "learning_rate": 3.983503707699658e-06, + "loss": 0.8206, + "step": 773 + }, + { + "epoch": 0.06980204716598278, + "grad_norm": 1.7089228504550382, + "learning_rate": 3.983428743181268e-06, + "loss": 0.8576, + "step": 774 + }, + { + "epoch": 0.06989223068945304, + "grad_norm": 1.8530087516565648, + "learning_rate": 3.983353609425802e-06, + "loss": 0.8971, + "step": 775 + }, + { + "epoch": 0.0699824142129233, + "grad_norm": 1.7180115240374447, + "learning_rate": 3.983278306439671e-06, + "loss": 0.8244, + "step": 776 + }, + { + "epoch": 0.07007259773639356, + "grad_norm": 2.1275032109775363, + "learning_rate": 3.983202834229297e-06, + "loss": 0.863, + "step": 777 + }, + { + "epoch": 0.07016278125986382, + "grad_norm": 2.180627326297148, + "learning_rate": 3.983127192801123e-06, + "loss": 0.911, + "step": 778 + }, + { + "epoch": 0.07025296478333408, + "grad_norm": 1.9171060459333933, + "learning_rate": 3.983051382161602e-06, + "loss": 0.8642, + "step": 779 + }, + { + "epoch": 0.07034314830680434, + "grad_norm": 3.8043325521032183, + "learning_rate": 3.982975402317203e-06, + "loss": 0.8492, + "step": 780 + }, + { + "epoch": 0.07043333183027461, + "grad_norm": 3.213420121642022, + "learning_rate": 3.982899253274409e-06, + "loss": 0.8474, + "step": 781 + }, + { + "epoch": 0.07052351535374488, + "grad_norm": 3.8049411571629665, + "learning_rate": 3.982822935039717e-06, + "loss": 0.8157, + "step": 782 + }, + { + "epoch": 0.07061369887721514, + "grad_norm": 1.9981635722427702, + "learning_rate": 3.982746447619638e-06, + "loss": 0.7602, + "step": 783 + }, + { + "epoch": 0.0707038824006854, + "grad_norm": 0.71591560593375, + "learning_rate": 3.9826697910207e-06, + "loss": 0.5852, + "step": 784 + }, + { + "epoch": 0.07079406592415566, + "grad_norm": 2.101797324749399, + "learning_rate": 3.982592965249442e-06, + "loss": 0.8138, + "step": 785 + }, + { + "epoch": 0.07088424944762592, + "grad_norm": 3.11479933041832, + "learning_rate": 3.982515970312422e-06, + "loss": 0.8692, + "step": 786 + }, + { + "epoch": 0.07097443297109618, + "grad_norm": 3.4446339879814847, + "learning_rate": 3.982438806216207e-06, + "loss": 0.8044, + "step": 787 + }, + { + "epoch": 0.07106461649456644, + "grad_norm": 2.103873304809591, + "learning_rate": 3.982361472967382e-06, + "loss": 0.821, + "step": 788 + }, + { + "epoch": 0.0711548000180367, + "grad_norm": 5.855052442047048, + "learning_rate": 3.982283970572546e-06, + "loss": 0.8975, + "step": 789 + }, + { + "epoch": 0.07124498354150696, + "grad_norm": 1.9264049332560298, + "learning_rate": 3.982206299038311e-06, + "loss": 0.8003, + "step": 790 + }, + { + "epoch": 0.07133516706497722, + "grad_norm": 1.6867793628426164, + "learning_rate": 3.9821284583713054e-06, + "loss": 0.9019, + "step": 791 + }, + { + "epoch": 0.07142535058844748, + "grad_norm": 2.0316311992066485, + "learning_rate": 3.98205044857817e-06, + "loss": 0.819, + "step": 792 + }, + { + "epoch": 0.07151553411191776, + "grad_norm": 0.7577438323381275, + "learning_rate": 3.981972269665561e-06, + "loss": 0.6138, + "step": 793 + }, + { + "epoch": 0.07160571763538802, + "grad_norm": 1.7396372788644103, + "learning_rate": 3.98189392164015e-06, + "loss": 0.8169, + "step": 794 + }, + { + "epoch": 0.07169590115885828, + "grad_norm": 1.8142342656160142, + "learning_rate": 3.981815404508621e-06, + "loss": 0.8679, + "step": 795 + }, + { + "epoch": 0.07178608468232854, + "grad_norm": 1.8516448441234408, + "learning_rate": 3.981736718277674e-06, + "loss": 0.8345, + "step": 796 + }, + { + "epoch": 0.0718762682057988, + "grad_norm": 1.8412112197883854, + "learning_rate": 3.9816578629540235e-06, + "loss": 0.8644, + "step": 797 + }, + { + "epoch": 0.07196645172926906, + "grad_norm": 1.8587403055642422, + "learning_rate": 3.981578838544398e-06, + "loss": 0.838, + "step": 798 + }, + { + "epoch": 0.07205663525273932, + "grad_norm": 0.7021785936525011, + "learning_rate": 3.981499645055539e-06, + "loss": 0.5673, + "step": 799 + }, + { + "epoch": 0.07214681877620958, + "grad_norm": 0.6721793306103447, + "learning_rate": 3.981420282494204e-06, + "loss": 0.6018, + "step": 800 + }, + { + "epoch": 0.07223700229967984, + "grad_norm": 0.744937900156803, + "learning_rate": 3.981340750867166e-06, + "loss": 0.6008, + "step": 801 + }, + { + "epoch": 0.0723271858231501, + "grad_norm": 1.7952699873682416, + "learning_rate": 3.981261050181209e-06, + "loss": 0.8136, + "step": 802 + }, + { + "epoch": 0.07241736934662037, + "grad_norm": 2.396917103543075, + "learning_rate": 3.9811811804431355e-06, + "loss": 0.7879, + "step": 803 + }, + { + "epoch": 0.07250755287009064, + "grad_norm": 1.6803996860671846, + "learning_rate": 3.981101141659759e-06, + "loss": 0.8725, + "step": 804 + }, + { + "epoch": 0.0725977363935609, + "grad_norm": 1.5823516897938592, + "learning_rate": 3.98102093383791e-06, + "loss": 0.8571, + "step": 805 + }, + { + "epoch": 0.07268791991703116, + "grad_norm": 2.257830662637803, + "learning_rate": 3.9809405569844315e-06, + "loss": 0.7472, + "step": 806 + }, + { + "epoch": 0.07277810344050142, + "grad_norm": 1.9989751335685104, + "learning_rate": 3.980860011106182e-06, + "loss": 0.7863, + "step": 807 + }, + { + "epoch": 0.07286828696397168, + "grad_norm": 1.3728768689760145, + "learning_rate": 3.980779296210033e-06, + "loss": 0.76, + "step": 808 + }, + { + "epoch": 0.07295847048744195, + "grad_norm": 3.498334624672031, + "learning_rate": 3.980698412302874e-06, + "loss": 0.9013, + "step": 809 + }, + { + "epoch": 0.0730486540109122, + "grad_norm": 2.5473547173608924, + "learning_rate": 3.980617359391604e-06, + "loss": 0.8351, + "step": 810 + }, + { + "epoch": 0.07313883753438247, + "grad_norm": 1.7025935892605257, + "learning_rate": 3.98053613748314e-06, + "loss": 0.7823, + "step": 811 + }, + { + "epoch": 0.07322902105785273, + "grad_norm": 1.7576583455626273, + "learning_rate": 3.980454746584413e-06, + "loss": 0.6868, + "step": 812 + }, + { + "epoch": 0.07331920458132299, + "grad_norm": 1.977148881260446, + "learning_rate": 3.9803731867023665e-06, + "loss": 0.9047, + "step": 813 + }, + { + "epoch": 0.07340938810479325, + "grad_norm": 1.8815122681358933, + "learning_rate": 3.9802914578439596e-06, + "loss": 0.806, + "step": 814 + }, + { + "epoch": 0.07349957162826351, + "grad_norm": 2.0497633960303348, + "learning_rate": 3.980209560016167e-06, + "loss": 0.9214, + "step": 815 + }, + { + "epoch": 0.07358975515173379, + "grad_norm": 1.8551412514223498, + "learning_rate": 3.980127493225975e-06, + "loss": 0.9165, + "step": 816 + }, + { + "epoch": 0.07367993867520405, + "grad_norm": 2.0042133058869718, + "learning_rate": 3.980045257480387e-06, + "loss": 0.8553, + "step": 817 + }, + { + "epoch": 0.0737701221986743, + "grad_norm": 2.1178214049718482, + "learning_rate": 3.9799628527864205e-06, + "loss": 0.7782, + "step": 818 + }, + { + "epoch": 0.07386030572214457, + "grad_norm": 2.26259394498069, + "learning_rate": 3.979880279151106e-06, + "loss": 0.8096, + "step": 819 + }, + { + "epoch": 0.07395048924561483, + "grad_norm": 2.422919952399097, + "learning_rate": 3.979797536581489e-06, + "loss": 0.7694, + "step": 820 + }, + { + "epoch": 0.07404067276908509, + "grad_norm": 2.2416129090435035, + "learning_rate": 3.97971462508463e-06, + "loss": 0.6755, + "step": 821 + }, + { + "epoch": 0.07413085629255535, + "grad_norm": 1.51701220708775, + "learning_rate": 3.979631544667603e-06, + "loss": 0.8068, + "step": 822 + }, + { + "epoch": 0.07422103981602561, + "grad_norm": 1.6789894250260677, + "learning_rate": 3.979548295337496e-06, + "loss": 0.8295, + "step": 823 + }, + { + "epoch": 0.07431122333949587, + "grad_norm": 1.54470194759477, + "learning_rate": 3.9794648771014146e-06, + "loss": 0.7913, + "step": 824 + }, + { + "epoch": 0.07440140686296613, + "grad_norm": 1.6467602047370218, + "learning_rate": 3.9793812899664745e-06, + "loss": 0.8276, + "step": 825 + }, + { + "epoch": 0.0744915903864364, + "grad_norm": 1.885778283554552, + "learning_rate": 3.979297533939809e-06, + "loss": 0.7822, + "step": 826 + }, + { + "epoch": 0.07458177390990665, + "grad_norm": 1.9379799156171797, + "learning_rate": 3.979213609028564e-06, + "loss": 0.7966, + "step": 827 + }, + { + "epoch": 0.07467195743337693, + "grad_norm": 1.8125564961014176, + "learning_rate": 3.979129515239901e-06, + "loss": 0.8344, + "step": 828 + }, + { + "epoch": 0.07476214095684719, + "grad_norm": 2.019210347222539, + "learning_rate": 3.979045252580994e-06, + "loss": 0.8315, + "step": 829 + }, + { + "epoch": 0.07485232448031745, + "grad_norm": 2.1998873248290147, + "learning_rate": 3.978960821059034e-06, + "loss": 0.768, + "step": 830 + }, + { + "epoch": 0.07494250800378771, + "grad_norm": 2.0945534374480492, + "learning_rate": 3.978876220681225e-06, + "loss": 0.8599, + "step": 831 + }, + { + "epoch": 0.07503269152725797, + "grad_norm": 2.5737953590371974, + "learning_rate": 3.978791451454786e-06, + "loss": 0.7508, + "step": 832 + }, + { + "epoch": 0.07512287505072823, + "grad_norm": 2.968340393720566, + "learning_rate": 3.978706513386949e-06, + "loss": 0.7816, + "step": 833 + }, + { + "epoch": 0.0752130585741985, + "grad_norm": 2.401555820671726, + "learning_rate": 3.978621406484962e-06, + "loss": 0.8331, + "step": 834 + }, + { + "epoch": 0.07530324209766875, + "grad_norm": 2.0651302327983987, + "learning_rate": 3.978536130756086e-06, + "loss": 0.8089, + "step": 835 + }, + { + "epoch": 0.07539342562113902, + "grad_norm": 1.801824834140493, + "learning_rate": 3.978450686207599e-06, + "loss": 0.8847, + "step": 836 + }, + { + "epoch": 0.07548360914460928, + "grad_norm": 2.208460306320008, + "learning_rate": 3.978365072846789e-06, + "loss": 0.8069, + "step": 837 + }, + { + "epoch": 0.07557379266807954, + "grad_norm": 2.883696795213754, + "learning_rate": 3.9782792906809625e-06, + "loss": 0.7708, + "step": 838 + }, + { + "epoch": 0.0756639761915498, + "grad_norm": 2.356446020740414, + "learning_rate": 3.97819333971744e-06, + "loss": 0.8073, + "step": 839 + }, + { + "epoch": 0.07575415971502007, + "grad_norm": 2.9350639143943, + "learning_rate": 3.978107219963553e-06, + "loss": 0.8434, + "step": 840 + }, + { + "epoch": 0.07584434323849033, + "grad_norm": 1.9073250037598866, + "learning_rate": 3.978020931426651e-06, + "loss": 0.7307, + "step": 841 + }, + { + "epoch": 0.0759345267619606, + "grad_norm": 1.6191135307594342, + "learning_rate": 3.977934474114096e-06, + "loss": 0.762, + "step": 842 + }, + { + "epoch": 0.07602471028543085, + "grad_norm": 0.6477291650531511, + "learning_rate": 3.977847848033267e-06, + "loss": 0.5796, + "step": 843 + }, + { + "epoch": 0.07611489380890112, + "grad_norm": 1.7352505784928909, + "learning_rate": 3.977761053191553e-06, + "loss": 0.8634, + "step": 844 + }, + { + "epoch": 0.07620507733237138, + "grad_norm": 1.7619670684456237, + "learning_rate": 3.977674089596361e-06, + "loss": 0.7996, + "step": 845 + }, + { + "epoch": 0.07629526085584164, + "grad_norm": 0.6967902179001819, + "learning_rate": 3.97758695725511e-06, + "loss": 0.5765, + "step": 846 + }, + { + "epoch": 0.0763854443793119, + "grad_norm": 1.391986009464441, + "learning_rate": 3.977499656175236e-06, + "loss": 0.8222, + "step": 847 + }, + { + "epoch": 0.07647562790278216, + "grad_norm": 2.1411211315955003, + "learning_rate": 3.977412186364187e-06, + "loss": 0.7333, + "step": 848 + }, + { + "epoch": 0.07656581142625242, + "grad_norm": 2.0877199022884696, + "learning_rate": 3.977324547829428e-06, + "loss": 0.8476, + "step": 849 + }, + { + "epoch": 0.07665599494972268, + "grad_norm": 2.1087064460638865, + "learning_rate": 3.977236740578435e-06, + "loss": 0.8494, + "step": 850 + }, + { + "epoch": 0.07674617847319294, + "grad_norm": 1.9763583956192183, + "learning_rate": 3.9771487646187015e-06, + "loss": 0.8113, + "step": 851 + }, + { + "epoch": 0.07683636199666322, + "grad_norm": 2.057342090874008, + "learning_rate": 3.9770606199577325e-06, + "loss": 0.8544, + "step": 852 + }, + { + "epoch": 0.07692654552013348, + "grad_norm": 2.1305871329798935, + "learning_rate": 3.9769723066030505e-06, + "loss": 0.7592, + "step": 853 + }, + { + "epoch": 0.07701672904360374, + "grad_norm": 3.245814783111046, + "learning_rate": 3.976883824562191e-06, + "loss": 0.8809, + "step": 854 + }, + { + "epoch": 0.077106912567074, + "grad_norm": 2.019683655055366, + "learning_rate": 3.976795173842703e-06, + "loss": 0.8218, + "step": 855 + }, + { + "epoch": 0.07719709609054426, + "grad_norm": 3.9276258395776162, + "learning_rate": 3.97670635445215e-06, + "loss": 0.7965, + "step": 856 + }, + { + "epoch": 0.07728727961401452, + "grad_norm": 2.1331035708835637, + "learning_rate": 3.976617366398112e-06, + "loss": 0.7981, + "step": 857 + }, + { + "epoch": 0.07737746313748478, + "grad_norm": 1.4170855669502946, + "learning_rate": 3.976528209688181e-06, + "loss": 0.8508, + "step": 858 + }, + { + "epoch": 0.07746764666095504, + "grad_norm": 1.5412347635349832, + "learning_rate": 3.976438884329965e-06, + "loss": 0.8237, + "step": 859 + }, + { + "epoch": 0.0775578301844253, + "grad_norm": 1.6380246581339695, + "learning_rate": 3.976349390331085e-06, + "loss": 0.9441, + "step": 860 + }, + { + "epoch": 0.07764801370789556, + "grad_norm": 1.8544534254316922, + "learning_rate": 3.976259727699178e-06, + "loss": 0.789, + "step": 861 + }, + { + "epoch": 0.07773819723136582, + "grad_norm": 2.1033622664543765, + "learning_rate": 3.976169896441895e-06, + "loss": 0.7899, + "step": 862 + }, + { + "epoch": 0.07782838075483609, + "grad_norm": 1.895933532861596, + "learning_rate": 3.976079896566898e-06, + "loss": 0.852, + "step": 863 + }, + { + "epoch": 0.07791856427830636, + "grad_norm": 1.8346526499566729, + "learning_rate": 3.97598972808187e-06, + "loss": 0.885, + "step": 864 + }, + { + "epoch": 0.07800874780177662, + "grad_norm": 3.259857926110476, + "learning_rate": 3.975899390994501e-06, + "loss": 0.8176, + "step": 865 + }, + { + "epoch": 0.07809893132524688, + "grad_norm": 1.5285564788576238, + "learning_rate": 3.975808885312502e-06, + "loss": 0.8837, + "step": 866 + }, + { + "epoch": 0.07818911484871714, + "grad_norm": 1.954711330887561, + "learning_rate": 3.975718211043594e-06, + "loss": 0.872, + "step": 867 + }, + { + "epoch": 0.0782792983721874, + "grad_norm": 1.9025072971359447, + "learning_rate": 3.975627368195515e-06, + "loss": 0.7366, + "step": 868 + }, + { + "epoch": 0.07836948189565766, + "grad_norm": 1.370249866198147, + "learning_rate": 3.975536356776015e-06, + "loss": 0.7837, + "step": 869 + }, + { + "epoch": 0.07845966541912792, + "grad_norm": 0.7403482233834734, + "learning_rate": 3.975445176792861e-06, + "loss": 0.6535, + "step": 870 + }, + { + "epoch": 0.07854984894259819, + "grad_norm": 1.714803569209996, + "learning_rate": 3.975353828253831e-06, + "loss": 0.8266, + "step": 871 + }, + { + "epoch": 0.07864003246606845, + "grad_norm": 1.7583749507058777, + "learning_rate": 3.97526231116672e-06, + "loss": 0.8258, + "step": 872 + }, + { + "epoch": 0.07873021598953871, + "grad_norm": 1.7364422274710098, + "learning_rate": 3.975170625539338e-06, + "loss": 0.7644, + "step": 873 + }, + { + "epoch": 0.07882039951300897, + "grad_norm": 2.3603687340992057, + "learning_rate": 3.975078771379507e-06, + "loss": 0.8161, + "step": 874 + }, + { + "epoch": 0.07891058303647923, + "grad_norm": 2.0281879526956965, + "learning_rate": 3.974986748695064e-06, + "loss": 0.7388, + "step": 875 + }, + { + "epoch": 0.0790007665599495, + "grad_norm": 0.72295765133054, + "learning_rate": 3.974894557493862e-06, + "loss": 0.5909, + "step": 876 + }, + { + "epoch": 0.07909095008341976, + "grad_norm": 1.5883735776261294, + "learning_rate": 3.974802197783768e-06, + "loss": 0.8654, + "step": 877 + }, + { + "epoch": 0.07918113360689003, + "grad_norm": 1.9303558057213517, + "learning_rate": 3.974709669572661e-06, + "loss": 0.883, + "step": 878 + }, + { + "epoch": 0.07927131713036029, + "grad_norm": 2.966432650762324, + "learning_rate": 3.974616972868436e-06, + "loss": 0.812, + "step": 879 + }, + { + "epoch": 0.07936150065383055, + "grad_norm": 1.9056762707432502, + "learning_rate": 3.974524107679003e-06, + "loss": 0.8242, + "step": 880 + }, + { + "epoch": 0.07945168417730081, + "grad_norm": 2.5288975463000902, + "learning_rate": 3.974431074012286e-06, + "loss": 0.8132, + "step": 881 + }, + { + "epoch": 0.07954186770077107, + "grad_norm": 2.774534970058829, + "learning_rate": 3.974337871876223e-06, + "loss": 0.8528, + "step": 882 + }, + { + "epoch": 0.07963205122424133, + "grad_norm": 1.5582141844270192, + "learning_rate": 3.974244501278766e-06, + "loss": 0.7875, + "step": 883 + }, + { + "epoch": 0.07972223474771159, + "grad_norm": 1.683864916849633, + "learning_rate": 3.974150962227883e-06, + "loss": 0.7675, + "step": 884 + }, + { + "epoch": 0.07981241827118185, + "grad_norm": 1.7924750818199153, + "learning_rate": 3.9740572547315535e-06, + "loss": 0.7584, + "step": 885 + }, + { + "epoch": 0.07990260179465211, + "grad_norm": 2.401918649666445, + "learning_rate": 3.973963378797775e-06, + "loss": 0.8517, + "step": 886 + }, + { + "epoch": 0.07999278531812237, + "grad_norm": 1.58465313699093, + "learning_rate": 3.973869334434556e-06, + "loss": 0.8334, + "step": 887 + }, + { + "epoch": 0.08008296884159265, + "grad_norm": 1.7707109371596157, + "learning_rate": 3.973775121649922e-06, + "loss": 0.8426, + "step": 888 + }, + { + "epoch": 0.08017315236506291, + "grad_norm": 1.5776199439103533, + "learning_rate": 3.973680740451911e-06, + "loss": 0.7329, + "step": 889 + }, + { + "epoch": 0.08026333588853317, + "grad_norm": 3.8966953459490035, + "learning_rate": 3.9735861908485776e-06, + "loss": 0.7842, + "step": 890 + }, + { + "epoch": 0.08035351941200343, + "grad_norm": 1.8461004243359564, + "learning_rate": 3.973491472847987e-06, + "loss": 0.8348, + "step": 891 + }, + { + "epoch": 0.08044370293547369, + "grad_norm": 1.7464748709088793, + "learning_rate": 3.973396586458222e-06, + "loss": 0.7519, + "step": 892 + }, + { + "epoch": 0.08053388645894395, + "grad_norm": 1.6353020921489307, + "learning_rate": 3.97330153168738e-06, + "loss": 0.8012, + "step": 893 + }, + { + "epoch": 0.08062406998241421, + "grad_norm": 1.9121353724332324, + "learning_rate": 3.973206308543571e-06, + "loss": 0.764, + "step": 894 + }, + { + "epoch": 0.08071425350588447, + "grad_norm": 2.319075848903772, + "learning_rate": 3.973110917034918e-06, + "loss": 0.7951, + "step": 895 + }, + { + "epoch": 0.08080443702935473, + "grad_norm": 1.566290948671965, + "learning_rate": 3.973015357169563e-06, + "loss": 0.7267, + "step": 896 + }, + { + "epoch": 0.080894620552825, + "grad_norm": 1.7048194225814182, + "learning_rate": 3.972919628955659e-06, + "loss": 0.7108, + "step": 897 + }, + { + "epoch": 0.08098480407629526, + "grad_norm": 2.2099821904703316, + "learning_rate": 3.972823732401373e-06, + "loss": 0.8078, + "step": 898 + }, + { + "epoch": 0.08107498759976552, + "grad_norm": 1.778825548810047, + "learning_rate": 3.972727667514888e-06, + "loss": 0.8133, + "step": 899 + }, + { + "epoch": 0.08116517112323579, + "grad_norm": 1.3935047293077363, + "learning_rate": 3.972631434304402e-06, + "loss": 0.7773, + "step": 900 + }, + { + "epoch": 0.08125535464670605, + "grad_norm": 1.9839990214606917, + "learning_rate": 3.972535032778124e-06, + "loss": 0.9289, + "step": 901 + }, + { + "epoch": 0.08134553817017631, + "grad_norm": 1.824508807302446, + "learning_rate": 3.97243846294428e-06, + "loss": 0.8985, + "step": 902 + }, + { + "epoch": 0.08143572169364657, + "grad_norm": 1.555325861086489, + "learning_rate": 3.972341724811111e-06, + "loss": 0.9542, + "step": 903 + }, + { + "epoch": 0.08152590521711683, + "grad_norm": 2.0905787806398313, + "learning_rate": 3.972244818386872e-06, + "loss": 0.7609, + "step": 904 + }, + { + "epoch": 0.0816160887405871, + "grad_norm": 2.183660025094129, + "learning_rate": 3.972147743679828e-06, + "loss": 0.7601, + "step": 905 + }, + { + "epoch": 0.08170627226405736, + "grad_norm": 1.38323985247369, + "learning_rate": 3.972050500698265e-06, + "loss": 0.782, + "step": 906 + }, + { + "epoch": 0.08179645578752762, + "grad_norm": 1.9373714650495544, + "learning_rate": 3.971953089450481e-06, + "loss": 0.8147, + "step": 907 + }, + { + "epoch": 0.08188663931099788, + "grad_norm": 1.9303865595093956, + "learning_rate": 3.971855509944784e-06, + "loss": 0.8912, + "step": 908 + }, + { + "epoch": 0.08197682283446814, + "grad_norm": 1.599561827981158, + "learning_rate": 3.971757762189504e-06, + "loss": 0.8, + "step": 909 + }, + { + "epoch": 0.0820670063579384, + "grad_norm": 1.9341871795925216, + "learning_rate": 3.9716598461929785e-06, + "loss": 0.8136, + "step": 910 + }, + { + "epoch": 0.08215718988140866, + "grad_norm": 1.6892713504089567, + "learning_rate": 3.971561761963563e-06, + "loss": 0.7879, + "step": 911 + }, + { + "epoch": 0.08224737340487893, + "grad_norm": 2.103183050615655, + "learning_rate": 3.971463509509628e-06, + "loss": 0.8931, + "step": 912 + }, + { + "epoch": 0.0823375569283492, + "grad_norm": 2.2388269847623765, + "learning_rate": 3.9713650888395555e-06, + "loss": 0.8284, + "step": 913 + }, + { + "epoch": 0.08242774045181946, + "grad_norm": 2.0124476029108607, + "learning_rate": 3.9712664999617425e-06, + "loss": 0.8067, + "step": 914 + }, + { + "epoch": 0.08251792397528972, + "grad_norm": 1.5462583603629076, + "learning_rate": 3.971167742884603e-06, + "loss": 0.7875, + "step": 915 + }, + { + "epoch": 0.08260810749875998, + "grad_norm": 1.8160474197187457, + "learning_rate": 3.971068817616564e-06, + "loss": 0.8667, + "step": 916 + }, + { + "epoch": 0.08269829102223024, + "grad_norm": 1.5752108266294669, + "learning_rate": 3.970969724166064e-06, + "loss": 0.8395, + "step": 917 + }, + { + "epoch": 0.0827884745457005, + "grad_norm": 0.8056944685313467, + "learning_rate": 3.970870462541559e-06, + "loss": 0.5794, + "step": 918 + }, + { + "epoch": 0.08287865806917076, + "grad_norm": 1.5550671597908436, + "learning_rate": 3.97077103275152e-06, + "loss": 0.8304, + "step": 919 + }, + { + "epoch": 0.08296884159264102, + "grad_norm": 1.434756605128791, + "learning_rate": 3.970671434804428e-06, + "loss": 0.8001, + "step": 920 + }, + { + "epoch": 0.08305902511611128, + "grad_norm": 2.509719646344402, + "learning_rate": 3.970571668708784e-06, + "loss": 0.8108, + "step": 921 + }, + { + "epoch": 0.08314920863958154, + "grad_norm": 0.6531359037374347, + "learning_rate": 3.9704717344731e-06, + "loss": 0.5887, + "step": 922 + }, + { + "epoch": 0.0832393921630518, + "grad_norm": 0.7141530603457684, + "learning_rate": 3.9703716321059026e-06, + "loss": 0.6049, + "step": 923 + }, + { + "epoch": 0.08332957568652208, + "grad_norm": 1.755845796492269, + "learning_rate": 3.9702713616157325e-06, + "loss": 0.905, + "step": 924 + }, + { + "epoch": 0.08341975920999234, + "grad_norm": 2.1310124329195763, + "learning_rate": 3.9701709230111455e-06, + "loss": 0.8307, + "step": 925 + }, + { + "epoch": 0.0835099427334626, + "grad_norm": 1.6059931805799599, + "learning_rate": 3.970070316300713e-06, + "loss": 0.7793, + "step": 926 + }, + { + "epoch": 0.08360012625693286, + "grad_norm": 2.067670640834097, + "learning_rate": 3.969969541493017e-06, + "loss": 0.844, + "step": 927 + }, + { + "epoch": 0.08369030978040312, + "grad_norm": 1.5710326104619108, + "learning_rate": 3.969868598596658e-06, + "loss": 0.8675, + "step": 928 + }, + { + "epoch": 0.08378049330387338, + "grad_norm": 1.279687393803295, + "learning_rate": 3.969767487620249e-06, + "loss": 0.7566, + "step": 929 + }, + { + "epoch": 0.08387067682734364, + "grad_norm": 1.3457535289964038, + "learning_rate": 3.969666208572416e-06, + "loss": 0.8635, + "step": 930 + }, + { + "epoch": 0.0839608603508139, + "grad_norm": 2.03249379737724, + "learning_rate": 3.969564761461802e-06, + "loss": 0.881, + "step": 931 + }, + { + "epoch": 0.08405104387428416, + "grad_norm": 1.7483140453563215, + "learning_rate": 3.969463146297062e-06, + "loss": 0.8, + "step": 932 + }, + { + "epoch": 0.08414122739775443, + "grad_norm": 5.076331564316786, + "learning_rate": 3.969361363086867e-06, + "loss": 0.8416, + "step": 933 + }, + { + "epoch": 0.08423141092122469, + "grad_norm": 1.8081767382110816, + "learning_rate": 3.9692594118399014e-06, + "loss": 0.7838, + "step": 934 + }, + { + "epoch": 0.08432159444469496, + "grad_norm": 1.8701071796199251, + "learning_rate": 3.969157292564865e-06, + "loss": 0.847, + "step": 935 + }, + { + "epoch": 0.08441177796816522, + "grad_norm": 2.2500196032199957, + "learning_rate": 3.96905500527047e-06, + "loss": 0.8016, + "step": 936 + }, + { + "epoch": 0.08450196149163548, + "grad_norm": 1.6774828259289634, + "learning_rate": 3.968952549965445e-06, + "loss": 0.8082, + "step": 937 + }, + { + "epoch": 0.08459214501510574, + "grad_norm": 1.7048285127912788, + "learning_rate": 3.968849926658532e-06, + "loss": 0.8483, + "step": 938 + }, + { + "epoch": 0.084682328538576, + "grad_norm": 2.1083784221708015, + "learning_rate": 3.9687471353584866e-06, + "loss": 0.7277, + "step": 939 + }, + { + "epoch": 0.08477251206204627, + "grad_norm": 1.4900794992424062, + "learning_rate": 3.9686441760740795e-06, + "loss": 0.8292, + "step": 940 + }, + { + "epoch": 0.08486269558551653, + "grad_norm": 2.1965503393040926, + "learning_rate": 3.968541048814098e-06, + "loss": 0.7567, + "step": 941 + }, + { + "epoch": 0.08495287910898679, + "grad_norm": 2.422102492938234, + "learning_rate": 3.968437753587339e-06, + "loss": 0.7506, + "step": 942 + }, + { + "epoch": 0.08504306263245705, + "grad_norm": 2.0104000769475885, + "learning_rate": 3.968334290402616e-06, + "loss": 0.8616, + "step": 943 + }, + { + "epoch": 0.08513324615592731, + "grad_norm": 1.504460775024406, + "learning_rate": 3.968230659268759e-06, + "loss": 0.7099, + "step": 944 + }, + { + "epoch": 0.08522342967939757, + "grad_norm": 1.9576374264899923, + "learning_rate": 3.968126860194609e-06, + "loss": 0.893, + "step": 945 + }, + { + "epoch": 0.08531361320286783, + "grad_norm": 1.7234514611129192, + "learning_rate": 3.968022893189025e-06, + "loss": 0.8655, + "step": 946 + }, + { + "epoch": 0.0854037967263381, + "grad_norm": 1.8422047880362928, + "learning_rate": 3.967918758260874e-06, + "loss": 0.8519, + "step": 947 + }, + { + "epoch": 0.08549398024980837, + "grad_norm": 1.6092468506985353, + "learning_rate": 3.967814455419044e-06, + "loss": 0.8659, + "step": 948 + }, + { + "epoch": 0.08558416377327863, + "grad_norm": 2.640105518052992, + "learning_rate": 3.967709984672434e-06, + "loss": 0.8644, + "step": 949 + }, + { + "epoch": 0.08567434729674889, + "grad_norm": 2.191656142621706, + "learning_rate": 3.967605346029959e-06, + "loss": 0.8483, + "step": 950 + }, + { + "epoch": 0.08576453082021915, + "grad_norm": 1.5369393667895621, + "learning_rate": 3.9675005395005466e-06, + "loss": 0.8938, + "step": 951 + }, + { + "epoch": 0.08585471434368941, + "grad_norm": 1.6515449053820512, + "learning_rate": 3.967395565093139e-06, + "loss": 0.8484, + "step": 952 + }, + { + "epoch": 0.08594489786715967, + "grad_norm": 1.5186886841755123, + "learning_rate": 3.967290422816693e-06, + "loss": 0.8294, + "step": 953 + }, + { + "epoch": 0.08603508139062993, + "grad_norm": 2.2190208806984737, + "learning_rate": 3.967185112680183e-06, + "loss": 0.7858, + "step": 954 + }, + { + "epoch": 0.08612526491410019, + "grad_norm": 2.025139994550519, + "learning_rate": 3.96707963469259e-06, + "loss": 0.8995, + "step": 955 + }, + { + "epoch": 0.08621544843757045, + "grad_norm": 1.9918148515020628, + "learning_rate": 3.966973988862917e-06, + "loss": 0.8133, + "step": 956 + }, + { + "epoch": 0.08630563196104071, + "grad_norm": 3.007841511375916, + "learning_rate": 3.966868175200178e-06, + "loss": 0.8713, + "step": 957 + }, + { + "epoch": 0.08639581548451097, + "grad_norm": 1.8554084205858252, + "learning_rate": 3.9667621937134e-06, + "loss": 0.8763, + "step": 958 + }, + { + "epoch": 0.08648599900798125, + "grad_norm": 1.8346667498008207, + "learning_rate": 3.966656044411627e-06, + "loss": 0.7429, + "step": 959 + }, + { + "epoch": 0.08657618253145151, + "grad_norm": 1.485226597692598, + "learning_rate": 3.966549727303918e-06, + "loss": 0.791, + "step": 960 + }, + { + "epoch": 0.08666636605492177, + "grad_norm": 3.810293465873927, + "learning_rate": 3.966443242399341e-06, + "loss": 0.8646, + "step": 961 + }, + { + "epoch": 0.08675654957839203, + "grad_norm": 0.7027180235626528, + "learning_rate": 3.966336589706985e-06, + "loss": 0.5998, + "step": 962 + }, + { + "epoch": 0.08684673310186229, + "grad_norm": 2.078315439318011, + "learning_rate": 3.966229769235948e-06, + "loss": 0.8384, + "step": 963 + }, + { + "epoch": 0.08693691662533255, + "grad_norm": 1.7713836338456517, + "learning_rate": 3.966122780995345e-06, + "loss": 0.9013, + "step": 964 + }, + { + "epoch": 0.08702710014880281, + "grad_norm": 2.8089218266366043, + "learning_rate": 3.966015624994306e-06, + "loss": 0.8668, + "step": 965 + }, + { + "epoch": 0.08711728367227307, + "grad_norm": 1.8935073445280688, + "learning_rate": 3.9659083012419735e-06, + "loss": 0.7949, + "step": 966 + }, + { + "epoch": 0.08720746719574334, + "grad_norm": 1.6956979339808387, + "learning_rate": 3.965800809747505e-06, + "loss": 0.7556, + "step": 967 + }, + { + "epoch": 0.0872976507192136, + "grad_norm": 1.8832960338893692, + "learning_rate": 3.965693150520071e-06, + "loss": 0.7873, + "step": 968 + }, + { + "epoch": 0.08738783424268386, + "grad_norm": 2.402091259597269, + "learning_rate": 3.96558532356886e-06, + "loss": 0.7616, + "step": 969 + }, + { + "epoch": 0.08747801776615412, + "grad_norm": 2.003310682049156, + "learning_rate": 3.9654773289030704e-06, + "loss": 0.8687, + "step": 970 + }, + { + "epoch": 0.08756820128962439, + "grad_norm": 1.9267385088363977, + "learning_rate": 3.9653691665319176e-06, + "loss": 0.8416, + "step": 971 + }, + { + "epoch": 0.08765838481309465, + "grad_norm": 1.8747969835363318, + "learning_rate": 3.96526083646463e-06, + "loss": 0.7502, + "step": 972 + }, + { + "epoch": 0.08774856833656491, + "grad_norm": 1.6112577665746999, + "learning_rate": 3.9651523387104526e-06, + "loss": 0.7722, + "step": 973 + }, + { + "epoch": 0.08783875186003517, + "grad_norm": 2.1219039128486825, + "learning_rate": 3.965043673278641e-06, + "loss": 0.8405, + "step": 974 + }, + { + "epoch": 0.08792893538350544, + "grad_norm": 1.7110878516896644, + "learning_rate": 3.964934840178469e-06, + "loss": 0.8277, + "step": 975 + }, + { + "epoch": 0.0880191189069757, + "grad_norm": 1.8843566134490493, + "learning_rate": 3.964825839419221e-06, + "loss": 0.8691, + "step": 976 + }, + { + "epoch": 0.08810930243044596, + "grad_norm": 7.047617938681065, + "learning_rate": 3.964716671010199e-06, + "loss": 0.8121, + "step": 977 + }, + { + "epoch": 0.08819948595391622, + "grad_norm": 1.489416455433491, + "learning_rate": 3.9646073349607165e-06, + "loss": 0.7854, + "step": 978 + }, + { + "epoch": 0.08828966947738648, + "grad_norm": 2.144005771871171, + "learning_rate": 3.964497831280105e-06, + "loss": 0.9522, + "step": 979 + }, + { + "epoch": 0.08837985300085674, + "grad_norm": 1.8322053025454488, + "learning_rate": 3.964388159977705e-06, + "loss": 0.8261, + "step": 980 + }, + { + "epoch": 0.088470036524327, + "grad_norm": 1.6218157160219653, + "learning_rate": 3.964278321062876e-06, + "loss": 0.7302, + "step": 981 + }, + { + "epoch": 0.08856022004779726, + "grad_norm": 1.8770977365475352, + "learning_rate": 3.96416831454499e-06, + "loss": 0.8458, + "step": 982 + }, + { + "epoch": 0.08865040357126754, + "grad_norm": 0.7100886632737706, + "learning_rate": 3.964058140433434e-06, + "loss": 0.6312, + "step": 983 + }, + { + "epoch": 0.0887405870947378, + "grad_norm": 1.9453305163181156, + "learning_rate": 3.963947798737606e-06, + "loss": 0.8846, + "step": 984 + }, + { + "epoch": 0.08883077061820806, + "grad_norm": 1.9006130484318433, + "learning_rate": 3.963837289466923e-06, + "loss": 0.8502, + "step": 985 + }, + { + "epoch": 0.08892095414167832, + "grad_norm": 2.1783041443896414, + "learning_rate": 3.9637266126308145e-06, + "loss": 0.7875, + "step": 986 + }, + { + "epoch": 0.08901113766514858, + "grad_norm": 0.6949610732861418, + "learning_rate": 3.963615768238724e-06, + "loss": 0.6201, + "step": 987 + }, + { + "epoch": 0.08910132118861884, + "grad_norm": 1.7052715668074365, + "learning_rate": 3.963504756300107e-06, + "loss": 0.7173, + "step": 988 + }, + { + "epoch": 0.0891915047120891, + "grad_norm": 1.7143454555478723, + "learning_rate": 3.96339357682444e-06, + "loss": 0.8582, + "step": 989 + }, + { + "epoch": 0.08928168823555936, + "grad_norm": 1.5773394300714356, + "learning_rate": 3.963282229821206e-06, + "loss": 0.8389, + "step": 990 + }, + { + "epoch": 0.08937187175902962, + "grad_norm": 1.5388623412436382, + "learning_rate": 3.963170715299906e-06, + "loss": 0.8362, + "step": 991 + }, + { + "epoch": 0.08946205528249988, + "grad_norm": 1.6571563543982093, + "learning_rate": 3.963059033270056e-06, + "loss": 0.7492, + "step": 992 + }, + { + "epoch": 0.08955223880597014, + "grad_norm": 1.9620636521726307, + "learning_rate": 3.9629471837411855e-06, + "loss": 0.7682, + "step": 993 + }, + { + "epoch": 0.0896424223294404, + "grad_norm": 2.3781736150877957, + "learning_rate": 3.962835166722838e-06, + "loss": 0.7686, + "step": 994 + }, + { + "epoch": 0.08973260585291068, + "grad_norm": 1.6461010324890075, + "learning_rate": 3.96272298222457e-06, + "loss": 0.8295, + "step": 995 + }, + { + "epoch": 0.08982278937638094, + "grad_norm": 4.095798591370701, + "learning_rate": 3.962610630255956e-06, + "loss": 0.9481, + "step": 996 + }, + { + "epoch": 0.0899129728998512, + "grad_norm": 1.7907907725454517, + "learning_rate": 3.96249811082658e-06, + "loss": 0.7971, + "step": 997 + }, + { + "epoch": 0.09000315642332146, + "grad_norm": 0.8050398934590617, + "learning_rate": 3.962385423946046e-06, + "loss": 0.6741, + "step": 998 + }, + { + "epoch": 0.09009333994679172, + "grad_norm": 1.4901138196906083, + "learning_rate": 3.962272569623966e-06, + "loss": 0.8011, + "step": 999 + }, + { + "epoch": 0.09018352347026198, + "grad_norm": 3.5058359809456245, + "learning_rate": 3.9621595478699704e-06, + "loss": 0.7143, + "step": 1000 + }, + { + "epoch": 0.09027370699373224, + "grad_norm": 1.8166663855222898, + "learning_rate": 3.962046358693703e-06, + "loss": 0.8135, + "step": 1001 + }, + { + "epoch": 0.0903638905172025, + "grad_norm": 6.6414989659266155, + "learning_rate": 3.961933002104822e-06, + "loss": 0.8946, + "step": 1002 + }, + { + "epoch": 0.09045407404067277, + "grad_norm": 1.6985256842467165, + "learning_rate": 3.961819478112999e-06, + "loss": 0.7484, + "step": 1003 + }, + { + "epoch": 0.09054425756414303, + "grad_norm": 2.266388652618795, + "learning_rate": 3.961705786727921e-06, + "loss": 0.8431, + "step": 1004 + }, + { + "epoch": 0.09063444108761329, + "grad_norm": 1.6941088785403502, + "learning_rate": 3.961591927959288e-06, + "loss": 0.7503, + "step": 1005 + }, + { + "epoch": 0.09072462461108355, + "grad_norm": 0.7210607700927277, + "learning_rate": 3.961477901816816e-06, + "loss": 0.5955, + "step": 1006 + }, + { + "epoch": 0.09081480813455382, + "grad_norm": 1.7746512003157078, + "learning_rate": 3.961363708310233e-06, + "loss": 0.8871, + "step": 1007 + }, + { + "epoch": 0.09090499165802408, + "grad_norm": 1.7748558657514995, + "learning_rate": 3.961249347449286e-06, + "loss": 0.8506, + "step": 1008 + }, + { + "epoch": 0.09099517518149434, + "grad_norm": 1.6559431403575697, + "learning_rate": 3.961134819243728e-06, + "loss": 0.8092, + "step": 1009 + }, + { + "epoch": 0.0910853587049646, + "grad_norm": 2.4866101269093814, + "learning_rate": 3.961020123703335e-06, + "loss": 0.8276, + "step": 1010 + }, + { + "epoch": 0.09117554222843487, + "grad_norm": 1.6411891149077622, + "learning_rate": 3.960905260837892e-06, + "loss": 0.8323, + "step": 1011 + }, + { + "epoch": 0.09126572575190513, + "grad_norm": 2.0490212872182743, + "learning_rate": 3.960790230657199e-06, + "loss": 0.7947, + "step": 1012 + }, + { + "epoch": 0.09135590927537539, + "grad_norm": 2.2221009645228103, + "learning_rate": 3.960675033171072e-06, + "loss": 0.7707, + "step": 1013 + }, + { + "epoch": 0.09144609279884565, + "grad_norm": 1.8291298476607043, + "learning_rate": 3.960559668389341e-06, + "loss": 0.7627, + "step": 1014 + }, + { + "epoch": 0.09153627632231591, + "grad_norm": 1.7750736704837158, + "learning_rate": 3.960444136321847e-06, + "loss": 0.8238, + "step": 1015 + }, + { + "epoch": 0.09162645984578617, + "grad_norm": 1.8568537311497877, + "learning_rate": 3.960328436978451e-06, + "loss": 0.9081, + "step": 1016 + }, + { + "epoch": 0.09171664336925643, + "grad_norm": 0.6657227350396939, + "learning_rate": 3.960212570369024e-06, + "loss": 0.5959, + "step": 1017 + }, + { + "epoch": 0.09180682689272669, + "grad_norm": 2.0548830120636605, + "learning_rate": 3.9600965365034515e-06, + "loss": 0.7541, + "step": 1018 + }, + { + "epoch": 0.09189701041619697, + "grad_norm": 1.6437626943351908, + "learning_rate": 3.959980335391634e-06, + "loss": 0.8746, + "step": 1019 + }, + { + "epoch": 0.09198719393966723, + "grad_norm": 1.7112494754720793, + "learning_rate": 3.959863967043487e-06, + "loss": 0.7586, + "step": 1020 + }, + { + "epoch": 0.09207737746313749, + "grad_norm": 1.5437112687537997, + "learning_rate": 3.9597474314689405e-06, + "loss": 0.8462, + "step": 1021 + }, + { + "epoch": 0.09216756098660775, + "grad_norm": 1.728413965716415, + "learning_rate": 3.959630728677937e-06, + "loss": 0.8378, + "step": 1022 + }, + { + "epoch": 0.09225774451007801, + "grad_norm": 2.141603900370646, + "learning_rate": 3.959513858680434e-06, + "loss": 0.8002, + "step": 1023 + }, + { + "epoch": 0.09234792803354827, + "grad_norm": 2.696242815979931, + "learning_rate": 3.959396821486405e-06, + "loss": 0.9106, + "step": 1024 + }, + { + "epoch": 0.09243811155701853, + "grad_norm": 0.8238641780743866, + "learning_rate": 3.959279617105835e-06, + "loss": 0.6251, + "step": 1025 + }, + { + "epoch": 0.09252829508048879, + "grad_norm": 2.0712681130334953, + "learning_rate": 3.9591622455487235e-06, + "loss": 0.8318, + "step": 1026 + }, + { + "epoch": 0.09261847860395905, + "grad_norm": 1.888360993007942, + "learning_rate": 3.959044706825087e-06, + "loss": 0.8947, + "step": 1027 + }, + { + "epoch": 0.09270866212742931, + "grad_norm": 1.8605473132536423, + "learning_rate": 3.958927000944954e-06, + "loss": 0.7246, + "step": 1028 + }, + { + "epoch": 0.09279884565089958, + "grad_norm": 1.6673667232205904, + "learning_rate": 3.958809127918368e-06, + "loss": 0.906, + "step": 1029 + }, + { + "epoch": 0.09288902917436984, + "grad_norm": 1.4799127264323546, + "learning_rate": 3.958691087755387e-06, + "loss": 0.7568, + "step": 1030 + }, + { + "epoch": 0.09297921269784011, + "grad_norm": 2.1575393139206716, + "learning_rate": 3.958572880466081e-06, + "loss": 0.7012, + "step": 1031 + }, + { + "epoch": 0.09306939622131037, + "grad_norm": 1.3884302457542237, + "learning_rate": 3.9584545060605385e-06, + "loss": 0.8055, + "step": 1032 + }, + { + "epoch": 0.09315957974478063, + "grad_norm": 1.608154426599025, + "learning_rate": 3.958335964548859e-06, + "loss": 0.7819, + "step": 1033 + }, + { + "epoch": 0.0932497632682509, + "grad_norm": 1.800693600852969, + "learning_rate": 3.958217255941156e-06, + "loss": 0.8743, + "step": 1034 + }, + { + "epoch": 0.09333994679172115, + "grad_norm": 1.4190343630650988, + "learning_rate": 3.95809838024756e-06, + "loss": 0.7677, + "step": 1035 + }, + { + "epoch": 0.09343013031519141, + "grad_norm": 1.6040110409373287, + "learning_rate": 3.957979337478212e-06, + "loss": 0.7809, + "step": 1036 + }, + { + "epoch": 0.09352031383866168, + "grad_norm": 1.9285438020937555, + "learning_rate": 3.957860127643272e-06, + "loss": 0.843, + "step": 1037 + }, + { + "epoch": 0.09361049736213194, + "grad_norm": 1.7251395321127445, + "learning_rate": 3.95774075075291e-06, + "loss": 0.8253, + "step": 1038 + }, + { + "epoch": 0.0937006808856022, + "grad_norm": 1.5539715522558117, + "learning_rate": 3.957621206817312e-06, + "loss": 0.8764, + "step": 1039 + }, + { + "epoch": 0.09379086440907246, + "grad_norm": 2.0952904571352304, + "learning_rate": 3.957501495846679e-06, + "loss": 0.8005, + "step": 1040 + }, + { + "epoch": 0.09388104793254272, + "grad_norm": 0.6860211462579138, + "learning_rate": 3.957381617851225e-06, + "loss": 0.601, + "step": 1041 + }, + { + "epoch": 0.09397123145601298, + "grad_norm": 2.8859090740660656, + "learning_rate": 3.9572615728411776e-06, + "loss": 0.7512, + "step": 1042 + }, + { + "epoch": 0.09406141497948325, + "grad_norm": 2.5442555068200945, + "learning_rate": 3.957141360826781e-06, + "loss": 0.7081, + "step": 1043 + }, + { + "epoch": 0.09415159850295352, + "grad_norm": 2.503601531770344, + "learning_rate": 3.957020981818292e-06, + "loss": 0.7997, + "step": 1044 + }, + { + "epoch": 0.09424178202642378, + "grad_norm": 1.7698597718850047, + "learning_rate": 3.956900435825982e-06, + "loss": 0.8107, + "step": 1045 + }, + { + "epoch": 0.09433196554989404, + "grad_norm": 1.4896631108330687, + "learning_rate": 3.9567797228601364e-06, + "loss": 0.7644, + "step": 1046 + }, + { + "epoch": 0.0944221490733643, + "grad_norm": 1.8105102667044684, + "learning_rate": 3.956658842931055e-06, + "loss": 0.8905, + "step": 1047 + }, + { + "epoch": 0.09451233259683456, + "grad_norm": 1.5413528672886763, + "learning_rate": 3.956537796049052e-06, + "loss": 0.8297, + "step": 1048 + }, + { + "epoch": 0.09460251612030482, + "grad_norm": 1.488758321077424, + "learning_rate": 3.956416582224457e-06, + "loss": 0.6197, + "step": 1049 + }, + { + "epoch": 0.09469269964377508, + "grad_norm": 1.890092128344963, + "learning_rate": 3.956295201467611e-06, + "loss": 0.8575, + "step": 1050 + }, + { + "epoch": 0.09478288316724534, + "grad_norm": 1.4525055744372763, + "learning_rate": 3.956173653788872e-06, + "loss": 0.8479, + "step": 1051 + }, + { + "epoch": 0.0948730666907156, + "grad_norm": 1.962940667117086, + "learning_rate": 3.95605193919861e-06, + "loss": 0.8613, + "step": 1052 + }, + { + "epoch": 0.09496325021418586, + "grad_norm": 2.1854992026958353, + "learning_rate": 3.955930057707211e-06, + "loss": 0.8123, + "step": 1053 + }, + { + "epoch": 0.09505343373765612, + "grad_norm": 2.2333528541546874, + "learning_rate": 3.955808009325075e-06, + "loss": 0.8312, + "step": 1054 + }, + { + "epoch": 0.0951436172611264, + "grad_norm": 1.945446392798215, + "learning_rate": 3.955685794062615e-06, + "loss": 0.7978, + "step": 1055 + }, + { + "epoch": 0.09523380078459666, + "grad_norm": 1.758914449523048, + "learning_rate": 3.95556341193026e-06, + "loss": 0.9204, + "step": 1056 + }, + { + "epoch": 0.09532398430806692, + "grad_norm": 1.523092299256297, + "learning_rate": 3.955440862938452e-06, + "loss": 0.9152, + "step": 1057 + }, + { + "epoch": 0.09541416783153718, + "grad_norm": 2.5158207029430155, + "learning_rate": 3.955318147097647e-06, + "loss": 0.8207, + "step": 1058 + }, + { + "epoch": 0.09550435135500744, + "grad_norm": 2.051462173042515, + "learning_rate": 3.955195264418316e-06, + "loss": 0.7886, + "step": 1059 + }, + { + "epoch": 0.0955945348784777, + "grad_norm": 3.0663454474963103, + "learning_rate": 3.955072214910944e-06, + "loss": 0.6775, + "step": 1060 + }, + { + "epoch": 0.09568471840194796, + "grad_norm": 2.4565634956935316, + "learning_rate": 3.954948998586032e-06, + "loss": 0.7507, + "step": 1061 + }, + { + "epoch": 0.09577490192541822, + "grad_norm": 1.5270841333960519, + "learning_rate": 3.954825615454089e-06, + "loss": 0.8284, + "step": 1062 + }, + { + "epoch": 0.09586508544888848, + "grad_norm": 1.6330091002164595, + "learning_rate": 3.954702065525649e-06, + "loss": 0.8077, + "step": 1063 + }, + { + "epoch": 0.09595526897235875, + "grad_norm": 2.2629520828696337, + "learning_rate": 3.954578348811248e-06, + "loss": 0.8486, + "step": 1064 + }, + { + "epoch": 0.096045452495829, + "grad_norm": 2.7815044158126994, + "learning_rate": 3.954454465321447e-06, + "loss": 0.7969, + "step": 1065 + }, + { + "epoch": 0.09613563601929928, + "grad_norm": 2.379053070598853, + "learning_rate": 3.954330415066813e-06, + "loss": 0.9008, + "step": 1066 + }, + { + "epoch": 0.09622581954276954, + "grad_norm": 2.056779152170802, + "learning_rate": 3.954206198057932e-06, + "loss": 0.8091, + "step": 1067 + }, + { + "epoch": 0.0963160030662398, + "grad_norm": 1.722872835068407, + "learning_rate": 3.954081814305403e-06, + "loss": 0.8795, + "step": 1068 + }, + { + "epoch": 0.09640618658971006, + "grad_norm": 2.1000056402948113, + "learning_rate": 3.953957263819839e-06, + "loss": 0.7537, + "step": 1069 + }, + { + "epoch": 0.09649637011318032, + "grad_norm": 1.7042844133185966, + "learning_rate": 3.953832546611867e-06, + "loss": 0.7558, + "step": 1070 + }, + { + "epoch": 0.09658655363665059, + "grad_norm": 2.0424581148482392, + "learning_rate": 3.953707662692129e-06, + "loss": 0.8489, + "step": 1071 + }, + { + "epoch": 0.09667673716012085, + "grad_norm": 1.746217386191025, + "learning_rate": 3.95358261207128e-06, + "loss": 0.753, + "step": 1072 + }, + { + "epoch": 0.0967669206835911, + "grad_norm": 4.09176168130748, + "learning_rate": 3.953457394759992e-06, + "loss": 0.8086, + "step": 1073 + }, + { + "epoch": 0.09685710420706137, + "grad_norm": 2.043417186644846, + "learning_rate": 3.953332010768947e-06, + "loss": 0.8328, + "step": 1074 + }, + { + "epoch": 0.09694728773053163, + "grad_norm": 2.6416750293041757, + "learning_rate": 3.9532064601088436e-06, + "loss": 0.7869, + "step": 1075 + }, + { + "epoch": 0.09703747125400189, + "grad_norm": 2.181637565043736, + "learning_rate": 3.953080742790396e-06, + "loss": 0.9729, + "step": 1076 + }, + { + "epoch": 0.09712765477747215, + "grad_norm": 0.7671044904936222, + "learning_rate": 3.95295485882433e-06, + "loss": 0.6616, + "step": 1077 + }, + { + "epoch": 0.09721783830094242, + "grad_norm": 0.6803872682683745, + "learning_rate": 3.952828808221387e-06, + "loss": 0.6291, + "step": 1078 + }, + { + "epoch": 0.09730802182441269, + "grad_norm": 2.199422877647784, + "learning_rate": 3.9527025909923225e-06, + "loss": 0.801, + "step": 1079 + }, + { + "epoch": 0.09739820534788295, + "grad_norm": 1.5430424539853158, + "learning_rate": 3.952576207147906e-06, + "loss": 0.905, + "step": 1080 + }, + { + "epoch": 0.09748838887135321, + "grad_norm": 1.656465912186938, + "learning_rate": 3.95244965669892e-06, + "loss": 0.8561, + "step": 1081 + }, + { + "epoch": 0.09757857239482347, + "grad_norm": 1.709533951811856, + "learning_rate": 3.952322939656165e-06, + "loss": 0.7756, + "step": 1082 + }, + { + "epoch": 0.09766875591829373, + "grad_norm": 1.7201797780629824, + "learning_rate": 3.952196056030451e-06, + "loss": 0.812, + "step": 1083 + }, + { + "epoch": 0.09775893944176399, + "grad_norm": 1.6818667446970121, + "learning_rate": 3.952069005832605e-06, + "loss": 0.7854, + "step": 1084 + }, + { + "epoch": 0.09784912296523425, + "grad_norm": 1.6040855816581894, + "learning_rate": 3.951941789073468e-06, + "loss": 0.8463, + "step": 1085 + }, + { + "epoch": 0.09793930648870451, + "grad_norm": 1.7380529360901908, + "learning_rate": 3.9518144057638955e-06, + "loss": 0.8463, + "step": 1086 + }, + { + "epoch": 0.09802949001217477, + "grad_norm": 2.786375316678553, + "learning_rate": 3.951686855914755e-06, + "loss": 0.8152, + "step": 1087 + }, + { + "epoch": 0.09811967353564503, + "grad_norm": 2.029356205709007, + "learning_rate": 3.9515591395369305e-06, + "loss": 0.8448, + "step": 1088 + }, + { + "epoch": 0.0982098570591153, + "grad_norm": 2.2007918753348847, + "learning_rate": 3.95143125664132e-06, + "loss": 0.8041, + "step": 1089 + }, + { + "epoch": 0.09830004058258557, + "grad_norm": 3.5742638423555846, + "learning_rate": 3.951303207238833e-06, + "loss": 0.8439, + "step": 1090 + }, + { + "epoch": 0.09839022410605583, + "grad_norm": 1.5812799624291956, + "learning_rate": 3.951174991340399e-06, + "loss": 0.7854, + "step": 1091 + }, + { + "epoch": 0.09848040762952609, + "grad_norm": 1.8181655731342419, + "learning_rate": 3.9510466089569546e-06, + "loss": 0.7776, + "step": 1092 + }, + { + "epoch": 0.09857059115299635, + "grad_norm": 1.8788076521605768, + "learning_rate": 3.950918060099456e-06, + "loss": 0.7869, + "step": 1093 + }, + { + "epoch": 0.09866077467646661, + "grad_norm": 1.734260211617017, + "learning_rate": 3.950789344778871e-06, + "loss": 0.8684, + "step": 1094 + }, + { + "epoch": 0.09875095819993687, + "grad_norm": 2.1757433355510534, + "learning_rate": 3.950660463006184e-06, + "loss": 0.7605, + "step": 1095 + }, + { + "epoch": 0.09884114172340713, + "grad_norm": 1.816295398898252, + "learning_rate": 3.950531414792389e-06, + "loss": 0.8884, + "step": 1096 + }, + { + "epoch": 0.0989313252468774, + "grad_norm": 1.6626659855406547, + "learning_rate": 3.950402200148498e-06, + "loss": 0.8231, + "step": 1097 + }, + { + "epoch": 0.09902150877034765, + "grad_norm": 1.6863552025058524, + "learning_rate": 3.950272819085538e-06, + "loss": 0.7957, + "step": 1098 + }, + { + "epoch": 0.09911169229381792, + "grad_norm": 2.173140781313469, + "learning_rate": 3.9501432716145474e-06, + "loss": 0.9055, + "step": 1099 + }, + { + "epoch": 0.09920187581728818, + "grad_norm": 1.8671021382120452, + "learning_rate": 3.950013557746579e-06, + "loss": 0.7146, + "step": 1100 + }, + { + "epoch": 0.09929205934075844, + "grad_norm": 1.5111837697247534, + "learning_rate": 3.949883677492703e-06, + "loss": 0.8536, + "step": 1101 + }, + { + "epoch": 0.09938224286422871, + "grad_norm": 1.8361021576761059, + "learning_rate": 3.9497536308639994e-06, + "loss": 0.8329, + "step": 1102 + }, + { + "epoch": 0.09947242638769897, + "grad_norm": 1.644450206073485, + "learning_rate": 3.949623417871565e-06, + "loss": 0.8655, + "step": 1103 + }, + { + "epoch": 0.09956260991116923, + "grad_norm": 2.0590197938446564, + "learning_rate": 3.949493038526511e-06, + "loss": 0.7993, + "step": 1104 + }, + { + "epoch": 0.0996527934346395, + "grad_norm": 1.7464619020139425, + "learning_rate": 3.949362492839961e-06, + "loss": 0.7689, + "step": 1105 + }, + { + "epoch": 0.09974297695810976, + "grad_norm": 1.7133434919261472, + "learning_rate": 3.949231780823054e-06, + "loss": 0.7812, + "step": 1106 + }, + { + "epoch": 0.09983316048158002, + "grad_norm": 1.7538896657588703, + "learning_rate": 3.949100902486945e-06, + "loss": 0.758, + "step": 1107 + }, + { + "epoch": 0.09992334400505028, + "grad_norm": 1.6564459774743314, + "learning_rate": 3.948969857842799e-06, + "loss": 0.7866, + "step": 1108 + }, + { + "epoch": 0.10001352752852054, + "grad_norm": 1.61804559765518, + "learning_rate": 3.948838646901798e-06, + "loss": 0.8092, + "step": 1109 + }, + { + "epoch": 0.1001037110519908, + "grad_norm": 1.9200325174359612, + "learning_rate": 3.948707269675138e-06, + "loss": 0.8117, + "step": 1110 + }, + { + "epoch": 0.10019389457546106, + "grad_norm": 1.7070970326194337, + "learning_rate": 3.948575726174028e-06, + "loss": 0.8076, + "step": 1111 + }, + { + "epoch": 0.10028407809893132, + "grad_norm": 1.8340772578857405, + "learning_rate": 3.9484440164096935e-06, + "loss": 0.8951, + "step": 1112 + }, + { + "epoch": 0.10037426162240158, + "grad_norm": 1.6881336152903064, + "learning_rate": 3.948312140393372e-06, + "loss": 0.8386, + "step": 1113 + }, + { + "epoch": 0.10046444514587186, + "grad_norm": 2.1471540466820507, + "learning_rate": 3.948180098136316e-06, + "loss": 0.786, + "step": 1114 + }, + { + "epoch": 0.10055462866934212, + "grad_norm": 12.60181496279188, + "learning_rate": 3.948047889649791e-06, + "loss": 0.7869, + "step": 1115 + }, + { + "epoch": 0.10064481219281238, + "grad_norm": 1.605243305053913, + "learning_rate": 3.947915514945079e-06, + "loss": 0.8014, + "step": 1116 + }, + { + "epoch": 0.10073499571628264, + "grad_norm": 1.8496622885654919, + "learning_rate": 3.947782974033474e-06, + "loss": 0.8788, + "step": 1117 + }, + { + "epoch": 0.1008251792397529, + "grad_norm": 2.2260072249974767, + "learning_rate": 3.9476502669262866e-06, + "loss": 0.7776, + "step": 1118 + }, + { + "epoch": 0.10091536276322316, + "grad_norm": 1.9566792002032256, + "learning_rate": 3.947517393634839e-06, + "loss": 0.7926, + "step": 1119 + }, + { + "epoch": 0.10100554628669342, + "grad_norm": 1.7301476311949133, + "learning_rate": 3.947384354170469e-06, + "loss": 0.8547, + "step": 1120 + }, + { + "epoch": 0.10109572981016368, + "grad_norm": 1.58137306367766, + "learning_rate": 3.947251148544528e-06, + "loss": 0.8463, + "step": 1121 + }, + { + "epoch": 0.10118591333363394, + "grad_norm": 1.8588116217059258, + "learning_rate": 3.947117776768382e-06, + "loss": 0.8473, + "step": 1122 + }, + { + "epoch": 0.1012760968571042, + "grad_norm": 1.860930105045721, + "learning_rate": 3.9469842388534105e-06, + "loss": 0.7923, + "step": 1123 + }, + { + "epoch": 0.10136628038057446, + "grad_norm": 1.8248671496387254, + "learning_rate": 3.946850534811009e-06, + "loss": 0.8241, + "step": 1124 + }, + { + "epoch": 0.10145646390404472, + "grad_norm": 1.8215613276722777, + "learning_rate": 3.946716664652585e-06, + "loss": 0.8041, + "step": 1125 + }, + { + "epoch": 0.101546647427515, + "grad_norm": 4.287347836477379, + "learning_rate": 3.94658262838956e-06, + "loss": 0.7416, + "step": 1126 + }, + { + "epoch": 0.10163683095098526, + "grad_norm": 1.9838256676321824, + "learning_rate": 3.946448426033373e-06, + "loss": 0.737, + "step": 1127 + }, + { + "epoch": 0.10172701447445552, + "grad_norm": 1.6521097653340218, + "learning_rate": 3.946314057595473e-06, + "loss": 0.8495, + "step": 1128 + }, + { + "epoch": 0.10181719799792578, + "grad_norm": 1.9118424600436503, + "learning_rate": 3.946179523087326e-06, + "loss": 0.8766, + "step": 1129 + }, + { + "epoch": 0.10190738152139604, + "grad_norm": 1.5299766072498264, + "learning_rate": 3.9460448225204104e-06, + "loss": 0.7288, + "step": 1130 + }, + { + "epoch": 0.1019975650448663, + "grad_norm": 2.5206823277498533, + "learning_rate": 3.945909955906221e-06, + "loss": 0.8129, + "step": 1131 + }, + { + "epoch": 0.10208774856833656, + "grad_norm": 1.618320676215216, + "learning_rate": 3.945774923256264e-06, + "loss": 0.8146, + "step": 1132 + }, + { + "epoch": 0.10217793209180683, + "grad_norm": 2.869357667611798, + "learning_rate": 3.945639724582062e-06, + "loss": 0.8898, + "step": 1133 + }, + { + "epoch": 0.10226811561527709, + "grad_norm": 2.2792692599548214, + "learning_rate": 3.94550435989515e-06, + "loss": 0.7877, + "step": 1134 + }, + { + "epoch": 0.10235829913874735, + "grad_norm": 1.4660411772520026, + "learning_rate": 3.945368829207079e-06, + "loss": 0.839, + "step": 1135 + }, + { + "epoch": 0.10244848266221761, + "grad_norm": 2.0488686207568483, + "learning_rate": 3.945233132529414e-06, + "loss": 0.7715, + "step": 1136 + }, + { + "epoch": 0.10253866618568787, + "grad_norm": 2.082492111758969, + "learning_rate": 3.9450972698737304e-06, + "loss": 0.8159, + "step": 1137 + }, + { + "epoch": 0.10262884970915814, + "grad_norm": 1.8789302324759063, + "learning_rate": 3.944961241251623e-06, + "loss": 0.793, + "step": 1138 + }, + { + "epoch": 0.1027190332326284, + "grad_norm": 1.4568337314280622, + "learning_rate": 3.9448250466746985e-06, + "loss": 0.7828, + "step": 1139 + }, + { + "epoch": 0.10280921675609866, + "grad_norm": 1.7334879763406008, + "learning_rate": 3.944688686154578e-06, + "loss": 0.8474, + "step": 1140 + }, + { + "epoch": 0.10289940027956893, + "grad_norm": 2.047357749795026, + "learning_rate": 3.944552159702894e-06, + "loss": 0.7728, + "step": 1141 + }, + { + "epoch": 0.10298958380303919, + "grad_norm": 1.7035522624945838, + "learning_rate": 3.944415467331299e-06, + "loss": 0.8026, + "step": 1142 + }, + { + "epoch": 0.10307976732650945, + "grad_norm": 3.2848444551428035, + "learning_rate": 3.944278609051455e-06, + "loss": 0.7682, + "step": 1143 + }, + { + "epoch": 0.10316995084997971, + "grad_norm": 2.449180395869449, + "learning_rate": 3.944141584875039e-06, + "loss": 0.7521, + "step": 1144 + }, + { + "epoch": 0.10326013437344997, + "grad_norm": 1.7137031672316063, + "learning_rate": 3.944004394813743e-06, + "loss": 0.8241, + "step": 1145 + }, + { + "epoch": 0.10335031789692023, + "grad_norm": 1.6202849427163304, + "learning_rate": 3.943867038879273e-06, + "loss": 0.8362, + "step": 1146 + }, + { + "epoch": 0.10344050142039049, + "grad_norm": 3.044026926590757, + "learning_rate": 3.943729517083349e-06, + "loss": 0.8559, + "step": 1147 + }, + { + "epoch": 0.10353068494386075, + "grad_norm": 1.5858930008381467, + "learning_rate": 3.943591829437705e-06, + "loss": 0.8502, + "step": 1148 + }, + { + "epoch": 0.10362086846733101, + "grad_norm": 1.93039952782737, + "learning_rate": 3.9434539759540895e-06, + "loss": 0.7556, + "step": 1149 + }, + { + "epoch": 0.10371105199080129, + "grad_norm": 1.6952073728250752, + "learning_rate": 3.943315956644264e-06, + "loss": 0.7497, + "step": 1150 + }, + { + "epoch": 0.10380123551427155, + "grad_norm": 1.7546521748004247, + "learning_rate": 3.943177771520006e-06, + "loss": 0.883, + "step": 1151 + }, + { + "epoch": 0.10389141903774181, + "grad_norm": 1.9236393095199769, + "learning_rate": 3.9430394205931065e-06, + "loss": 0.8602, + "step": 1152 + }, + { + "epoch": 0.10398160256121207, + "grad_norm": 1.8011524537759607, + "learning_rate": 3.942900903875369e-06, + "loss": 0.6915, + "step": 1153 + }, + { + "epoch": 0.10407178608468233, + "grad_norm": 2.1788063592674045, + "learning_rate": 3.942762221378614e-06, + "loss": 0.8178, + "step": 1154 + }, + { + "epoch": 0.10416196960815259, + "grad_norm": 1.4476711513289668, + "learning_rate": 3.942623373114673e-06, + "loss": 0.8444, + "step": 1155 + }, + { + "epoch": 0.10425215313162285, + "grad_norm": 2.0124311352433106, + "learning_rate": 3.942484359095396e-06, + "loss": 0.7998, + "step": 1156 + }, + { + "epoch": 0.10434233665509311, + "grad_norm": 1.9529594046011036, + "learning_rate": 3.942345179332642e-06, + "loss": 0.7458, + "step": 1157 + }, + { + "epoch": 0.10443252017856337, + "grad_norm": 1.705888169038851, + "learning_rate": 3.942205833838287e-06, + "loss": 0.7694, + "step": 1158 + }, + { + "epoch": 0.10452270370203363, + "grad_norm": 1.8531559393167119, + "learning_rate": 3.9420663226242204e-06, + "loss": 0.9101, + "step": 1159 + }, + { + "epoch": 0.1046128872255039, + "grad_norm": 2.396901586352816, + "learning_rate": 3.941926645702348e-06, + "loss": 0.8629, + "step": 1160 + }, + { + "epoch": 0.10470307074897416, + "grad_norm": 3.0148428263271128, + "learning_rate": 3.941786803084586e-06, + "loss": 0.7974, + "step": 1161 + }, + { + "epoch": 0.10479325427244443, + "grad_norm": 2.0678580075404995, + "learning_rate": 3.941646794782867e-06, + "loss": 0.7633, + "step": 1162 + }, + { + "epoch": 0.10488343779591469, + "grad_norm": 2.0963818534460184, + "learning_rate": 3.941506620809137e-06, + "loss": 0.8259, + "step": 1163 + }, + { + "epoch": 0.10497362131938495, + "grad_norm": 2.137991182878261, + "learning_rate": 3.941366281175357e-06, + "loss": 0.7961, + "step": 1164 + }, + { + "epoch": 0.10506380484285521, + "grad_norm": 1.5030015478373646, + "learning_rate": 3.941225775893502e-06, + "loss": 0.7527, + "step": 1165 + }, + { + "epoch": 0.10515398836632547, + "grad_norm": 1.4540044932761678, + "learning_rate": 3.941085104975559e-06, + "loss": 0.7874, + "step": 1166 + }, + { + "epoch": 0.10524417188979573, + "grad_norm": 0.7624734514336108, + "learning_rate": 3.9409442684335325e-06, + "loss": 0.6244, + "step": 1167 + }, + { + "epoch": 0.105334355413266, + "grad_norm": 1.944216745153747, + "learning_rate": 3.940803266279438e-06, + "loss": 0.8076, + "step": 1168 + }, + { + "epoch": 0.10542453893673626, + "grad_norm": 3.6583046641924515, + "learning_rate": 3.9406620985253076e-06, + "loss": 0.7563, + "step": 1169 + }, + { + "epoch": 0.10551472246020652, + "grad_norm": 2.6424689541728, + "learning_rate": 3.940520765183187e-06, + "loss": 0.784, + "step": 1170 + }, + { + "epoch": 0.10560490598367678, + "grad_norm": 8.162750343117356, + "learning_rate": 3.940379266265134e-06, + "loss": 0.8576, + "step": 1171 + }, + { + "epoch": 0.10569508950714704, + "grad_norm": 2.199526289310811, + "learning_rate": 3.940237601783223e-06, + "loss": 0.7797, + "step": 1172 + }, + { + "epoch": 0.1057852730306173, + "grad_norm": 1.4932407037639448, + "learning_rate": 3.940095771749542e-06, + "loss": 0.8108, + "step": 1173 + }, + { + "epoch": 0.10587545655408757, + "grad_norm": 1.934748264468221, + "learning_rate": 3.939953776176192e-06, + "loss": 0.8734, + "step": 1174 + }, + { + "epoch": 0.10596564007755783, + "grad_norm": 1.2721147747050532, + "learning_rate": 3.939811615075288e-06, + "loss": 0.8376, + "step": 1175 + }, + { + "epoch": 0.1060558236010281, + "grad_norm": 5.164858388815096, + "learning_rate": 3.9396692884589616e-06, + "loss": 0.8763, + "step": 1176 + }, + { + "epoch": 0.10614600712449836, + "grad_norm": 2.1357235447645992, + "learning_rate": 3.9395267963393565e-06, + "loss": 0.7957, + "step": 1177 + }, + { + "epoch": 0.10623619064796862, + "grad_norm": 2.88082511234922, + "learning_rate": 3.939384138728631e-06, + "loss": 0.8521, + "step": 1178 + }, + { + "epoch": 0.10632637417143888, + "grad_norm": 1.5832118774142014, + "learning_rate": 3.939241315638956e-06, + "loss": 0.7988, + "step": 1179 + }, + { + "epoch": 0.10641655769490914, + "grad_norm": 2.2402816275258526, + "learning_rate": 3.93909832708252e-06, + "loss": 0.7998, + "step": 1180 + }, + { + "epoch": 0.1065067412183794, + "grad_norm": 1.4901295796315894, + "learning_rate": 3.938955173071523e-06, + "loss": 0.8188, + "step": 1181 + }, + { + "epoch": 0.10659692474184966, + "grad_norm": 0.6150090607309072, + "learning_rate": 3.938811853618179e-06, + "loss": 0.6097, + "step": 1182 + }, + { + "epoch": 0.10668710826531992, + "grad_norm": 1.922722668296897, + "learning_rate": 3.938668368734717e-06, + "loss": 0.852, + "step": 1183 + }, + { + "epoch": 0.10677729178879018, + "grad_norm": 2.066298945332435, + "learning_rate": 3.93852471843338e-06, + "loss": 0.6919, + "step": 1184 + }, + { + "epoch": 0.10686747531226044, + "grad_norm": 1.755149621481491, + "learning_rate": 3.9383809027264254e-06, + "loss": 0.7926, + "step": 1185 + }, + { + "epoch": 0.10695765883573072, + "grad_norm": 2.2853918231273354, + "learning_rate": 3.938236921626124e-06, + "loss": 0.852, + "step": 1186 + }, + { + "epoch": 0.10704784235920098, + "grad_norm": 2.660521270758688, + "learning_rate": 3.938092775144761e-06, + "loss": 0.8246, + "step": 1187 + }, + { + "epoch": 0.10713802588267124, + "grad_norm": 1.5604094443601852, + "learning_rate": 3.9379484632946355e-06, + "loss": 0.734, + "step": 1188 + }, + { + "epoch": 0.1072282094061415, + "grad_norm": 1.6495672756697282, + "learning_rate": 3.937803986088062e-06, + "loss": 0.8566, + "step": 1189 + }, + { + "epoch": 0.10731839292961176, + "grad_norm": 2.080058868748896, + "learning_rate": 3.937659343537367e-06, + "loss": 0.7744, + "step": 1190 + }, + { + "epoch": 0.10740857645308202, + "grad_norm": 1.757781236158608, + "learning_rate": 3.937514535654893e-06, + "loss": 0.8851, + "step": 1191 + }, + { + "epoch": 0.10749875997655228, + "grad_norm": 3.2778136651899623, + "learning_rate": 3.937369562452996e-06, + "loss": 0.8223, + "step": 1192 + }, + { + "epoch": 0.10758894350002254, + "grad_norm": 2.545549668045743, + "learning_rate": 3.937224423944044e-06, + "loss": 0.8715, + "step": 1193 + }, + { + "epoch": 0.1076791270234928, + "grad_norm": 1.7342002711904192, + "learning_rate": 3.937079120140423e-06, + "loss": 0.748, + "step": 1194 + }, + { + "epoch": 0.10776931054696307, + "grad_norm": 1.931732082998214, + "learning_rate": 3.936933651054531e-06, + "loss": 0.8483, + "step": 1195 + }, + { + "epoch": 0.10785949407043333, + "grad_norm": 1.6202422697909664, + "learning_rate": 3.936788016698779e-06, + "loss": 0.9269, + "step": 1196 + }, + { + "epoch": 0.1079496775939036, + "grad_norm": 2.557364737922809, + "learning_rate": 3.936642217085594e-06, + "loss": 0.846, + "step": 1197 + }, + { + "epoch": 0.10803986111737386, + "grad_norm": 2.011073807229307, + "learning_rate": 3.936496252227417e-06, + "loss": 0.7261, + "step": 1198 + }, + { + "epoch": 0.10813004464084412, + "grad_norm": 5.085533583111134, + "learning_rate": 3.936350122136703e-06, + "loss": 0.798, + "step": 1199 + }, + { + "epoch": 0.10822022816431438, + "grad_norm": 0.6954516796587367, + "learning_rate": 3.936203826825919e-06, + "loss": 0.6312, + "step": 1200 + }, + { + "epoch": 0.10831041168778464, + "grad_norm": 1.9861171013920338, + "learning_rate": 3.9360573663075475e-06, + "loss": 0.7719, + "step": 1201 + }, + { + "epoch": 0.1084005952112549, + "grad_norm": 2.413190171735779, + "learning_rate": 3.935910740594087e-06, + "loss": 0.8443, + "step": 1202 + }, + { + "epoch": 0.10849077873472517, + "grad_norm": 2.167095386442514, + "learning_rate": 3.935763949698047e-06, + "loss": 0.8214, + "step": 1203 + }, + { + "epoch": 0.10858096225819543, + "grad_norm": 1.716766461308037, + "learning_rate": 3.935616993631954e-06, + "loss": 0.8814, + "step": 1204 + }, + { + "epoch": 0.10867114578166569, + "grad_norm": 4.52158793584155, + "learning_rate": 3.935469872408345e-06, + "loss": 0.8057, + "step": 1205 + }, + { + "epoch": 0.10876132930513595, + "grad_norm": 2.1389339478055303, + "learning_rate": 3.935322586039776e-06, + "loss": 0.832, + "step": 1206 + }, + { + "epoch": 0.10885151282860621, + "grad_norm": 1.5377067364898578, + "learning_rate": 3.935175134538811e-06, + "loss": 0.7436, + "step": 1207 + }, + { + "epoch": 0.10894169635207647, + "grad_norm": 2.2107961225295845, + "learning_rate": 3.935027517918034e-06, + "loss": 0.7566, + "step": 1208 + }, + { + "epoch": 0.10903187987554674, + "grad_norm": 3.0338072067795796, + "learning_rate": 3.93487973619004e-06, + "loss": 0.8454, + "step": 1209 + }, + { + "epoch": 0.109122063399017, + "grad_norm": 4.9065938452486, + "learning_rate": 3.934731789367438e-06, + "loss": 0.8277, + "step": 1210 + }, + { + "epoch": 0.10921224692248727, + "grad_norm": 1.0232223401053406, + "learning_rate": 3.9345836774628505e-06, + "loss": 0.6088, + "step": 1211 + }, + { + "epoch": 0.10930243044595753, + "grad_norm": 1.5553231018330502, + "learning_rate": 3.934435400488917e-06, + "loss": 0.8014, + "step": 1212 + }, + { + "epoch": 0.10939261396942779, + "grad_norm": 1.622464402547424, + "learning_rate": 3.934286958458289e-06, + "loss": 0.8433, + "step": 1213 + }, + { + "epoch": 0.10948279749289805, + "grad_norm": 1.6065308707506765, + "learning_rate": 3.934138351383632e-06, + "loss": 0.8724, + "step": 1214 + }, + { + "epoch": 0.10957298101636831, + "grad_norm": 3.5324139322539394, + "learning_rate": 3.933989579277626e-06, + "loss": 0.7473, + "step": 1215 + }, + { + "epoch": 0.10966316453983857, + "grad_norm": 1.8002604561030466, + "learning_rate": 3.933840642152966e-06, + "loss": 0.8071, + "step": 1216 + }, + { + "epoch": 0.10975334806330883, + "grad_norm": 2.2195103376800485, + "learning_rate": 3.933691540022359e-06, + "loss": 0.8399, + "step": 1217 + }, + { + "epoch": 0.10984353158677909, + "grad_norm": 1.7733536082855927, + "learning_rate": 3.933542272898527e-06, + "loss": 0.7824, + "step": 1218 + }, + { + "epoch": 0.10993371511024935, + "grad_norm": 0.7526594535469616, + "learning_rate": 3.933392840794207e-06, + "loss": 0.567, + "step": 1219 + }, + { + "epoch": 0.11002389863371961, + "grad_norm": 1.7884463119489713, + "learning_rate": 3.93324324372215e-06, + "loss": 0.9197, + "step": 1220 + }, + { + "epoch": 0.11011408215718989, + "grad_norm": 1.699932781461319, + "learning_rate": 3.9330934816951185e-06, + "loss": 0.8382, + "step": 1221 + }, + { + "epoch": 0.11020426568066015, + "grad_norm": 1.7797666697856787, + "learning_rate": 3.932943554725893e-06, + "loss": 0.8445, + "step": 1222 + }, + { + "epoch": 0.11029444920413041, + "grad_norm": 0.7070543696380228, + "learning_rate": 3.932793462827265e-06, + "loss": 0.641, + "step": 1223 + }, + { + "epoch": 0.11038463272760067, + "grad_norm": 1.8073221766583536, + "learning_rate": 3.932643206012041e-06, + "loss": 0.8711, + "step": 1224 + }, + { + "epoch": 0.11047481625107093, + "grad_norm": 2.115730037363798, + "learning_rate": 3.932492784293043e-06, + "loss": 0.7381, + "step": 1225 + }, + { + "epoch": 0.11056499977454119, + "grad_norm": 1.5144156123945, + "learning_rate": 3.932342197683104e-06, + "loss": 0.8032, + "step": 1226 + }, + { + "epoch": 0.11065518329801145, + "grad_norm": 2.056875709943901, + "learning_rate": 3.932191446195075e-06, + "loss": 0.6854, + "step": 1227 + }, + { + "epoch": 0.11074536682148171, + "grad_norm": 2.428398845454799, + "learning_rate": 3.9320405298418175e-06, + "loss": 0.8129, + "step": 1228 + }, + { + "epoch": 0.11083555034495197, + "grad_norm": 1.6947060480356637, + "learning_rate": 3.9318894486362076e-06, + "loss": 0.7381, + "step": 1229 + }, + { + "epoch": 0.11092573386842224, + "grad_norm": 1.9142695723648127, + "learning_rate": 3.9317382025911395e-06, + "loss": 0.8857, + "step": 1230 + }, + { + "epoch": 0.1110159173918925, + "grad_norm": 1.8092078861418082, + "learning_rate": 3.9315867917195145e-06, + "loss": 0.9117, + "step": 1231 + }, + { + "epoch": 0.11110610091536276, + "grad_norm": 1.7184998763758348, + "learning_rate": 3.931435216034256e-06, + "loss": 0.8121, + "step": 1232 + }, + { + "epoch": 0.11119628443883303, + "grad_norm": 3.2721052349301103, + "learning_rate": 3.931283475548293e-06, + "loss": 0.8253, + "step": 1233 + }, + { + "epoch": 0.11128646796230329, + "grad_norm": 1.352898615340822, + "learning_rate": 3.931131570274576e-06, + "loss": 0.8169, + "step": 1234 + }, + { + "epoch": 0.11137665148577355, + "grad_norm": 1.5574705431557452, + "learning_rate": 3.930979500226065e-06, + "loss": 0.847, + "step": 1235 + }, + { + "epoch": 0.11146683500924381, + "grad_norm": 2.8065173627084223, + "learning_rate": 3.930827265415736e-06, + "loss": 0.65, + "step": 1236 + }, + { + "epoch": 0.11155701853271408, + "grad_norm": 1.7343415265676416, + "learning_rate": 3.930674865856578e-06, + "loss": 0.8236, + "step": 1237 + }, + { + "epoch": 0.11164720205618434, + "grad_norm": 1.9646313288931916, + "learning_rate": 3.930522301561595e-06, + "loss": 0.8486, + "step": 1238 + }, + { + "epoch": 0.1117373855796546, + "grad_norm": 2.4088857507783, + "learning_rate": 3.930369572543804e-06, + "loss": 0.8575, + "step": 1239 + }, + { + "epoch": 0.11182756910312486, + "grad_norm": 1.635268194501253, + "learning_rate": 3.930216678816237e-06, + "loss": 0.7594, + "step": 1240 + }, + { + "epoch": 0.11191775262659512, + "grad_norm": 0.6841361591454287, + "learning_rate": 3.930063620391941e-06, + "loss": 0.6152, + "step": 1241 + }, + { + "epoch": 0.11200793615006538, + "grad_norm": 1.8249518740853583, + "learning_rate": 3.9299103972839735e-06, + "loss": 0.8247, + "step": 1242 + }, + { + "epoch": 0.11209811967353564, + "grad_norm": 0.654133448901395, + "learning_rate": 3.92975700950541e-06, + "loss": 0.5802, + "step": 1243 + }, + { + "epoch": 0.1121883031970059, + "grad_norm": 1.8952893185750657, + "learning_rate": 3.929603457069338e-06, + "loss": 0.8419, + "step": 1244 + }, + { + "epoch": 0.11227848672047618, + "grad_norm": 1.6489425536328652, + "learning_rate": 3.929449739988859e-06, + "loss": 0.7771, + "step": 1245 + }, + { + "epoch": 0.11236867024394644, + "grad_norm": 1.327081068067948, + "learning_rate": 3.929295858277089e-06, + "loss": 0.81, + "step": 1246 + }, + { + "epoch": 0.1124588537674167, + "grad_norm": 1.7679343186083951, + "learning_rate": 3.9291418119471585e-06, + "loss": 0.8103, + "step": 1247 + }, + { + "epoch": 0.11254903729088696, + "grad_norm": 2.304006805760875, + "learning_rate": 3.928987601012212e-06, + "loss": 0.8425, + "step": 1248 + }, + { + "epoch": 0.11263922081435722, + "grad_norm": 1.8496378621502787, + "learning_rate": 3.928833225485407e-06, + "loss": 0.8017, + "step": 1249 + }, + { + "epoch": 0.11272940433782748, + "grad_norm": 1.7314402176232853, + "learning_rate": 3.928678685379915e-06, + "loss": 0.8035, + "step": 1250 + }, + { + "epoch": 0.11281958786129774, + "grad_norm": 1.6686715782683559, + "learning_rate": 3.928523980708924e-06, + "loss": 0.8562, + "step": 1251 + }, + { + "epoch": 0.112909771384768, + "grad_norm": 2.170087998535248, + "learning_rate": 3.928369111485632e-06, + "loss": 0.7648, + "step": 1252 + }, + { + "epoch": 0.11299995490823826, + "grad_norm": 1.9537821770359522, + "learning_rate": 3.928214077723255e-06, + "loss": 0.7995, + "step": 1253 + }, + { + "epoch": 0.11309013843170852, + "grad_norm": 1.8641644509918638, + "learning_rate": 3.928058879435021e-06, + "loss": 0.7971, + "step": 1254 + }, + { + "epoch": 0.11318032195517878, + "grad_norm": 1.7154414844856516, + "learning_rate": 3.9279035166341725e-06, + "loss": 0.795, + "step": 1255 + }, + { + "epoch": 0.11327050547864904, + "grad_norm": 12.151428364752473, + "learning_rate": 3.927747989333965e-06, + "loss": 0.794, + "step": 1256 + }, + { + "epoch": 0.11336068900211932, + "grad_norm": 2.311780663035396, + "learning_rate": 3.927592297547669e-06, + "loss": 0.7842, + "step": 1257 + }, + { + "epoch": 0.11345087252558958, + "grad_norm": 1.878153057174443, + "learning_rate": 3.927436441288571e-06, + "loss": 0.8376, + "step": 1258 + }, + { + "epoch": 0.11354105604905984, + "grad_norm": 2.220719121843494, + "learning_rate": 3.927280420569968e-06, + "loss": 0.826, + "step": 1259 + }, + { + "epoch": 0.1136312395725301, + "grad_norm": 1.8236060083451373, + "learning_rate": 3.927124235405171e-06, + "loss": 0.8384, + "step": 1260 + }, + { + "epoch": 0.11372142309600036, + "grad_norm": 1.9168336698779853, + "learning_rate": 3.92696788580751e-06, + "loss": 0.6738, + "step": 1261 + }, + { + "epoch": 0.11381160661947062, + "grad_norm": 1.9619456583443862, + "learning_rate": 3.9268113717903225e-06, + "loss": 0.651, + "step": 1262 + }, + { + "epoch": 0.11390179014294088, + "grad_norm": 1.9431846041661947, + "learning_rate": 3.926654693366965e-06, + "loss": 0.8827, + "step": 1263 + }, + { + "epoch": 0.11399197366641114, + "grad_norm": 2.346361358552524, + "learning_rate": 3.926497850550805e-06, + "loss": 0.8225, + "step": 1264 + }, + { + "epoch": 0.1140821571898814, + "grad_norm": 1.9270844605588624, + "learning_rate": 3.926340843355226e-06, + "loss": 0.8116, + "step": 1265 + }, + { + "epoch": 0.11417234071335167, + "grad_norm": 1.905730379918635, + "learning_rate": 3.926183671793625e-06, + "loss": 0.8235, + "step": 1266 + }, + { + "epoch": 0.11426252423682193, + "grad_norm": 1.5562212102593995, + "learning_rate": 3.926026335879412e-06, + "loss": 0.7794, + "step": 1267 + }, + { + "epoch": 0.11435270776029219, + "grad_norm": 0.6332168464656507, + "learning_rate": 3.925868835626012e-06, + "loss": 0.5517, + "step": 1268 + }, + { + "epoch": 0.11444289128376246, + "grad_norm": 2.008914748855751, + "learning_rate": 3.925711171046864e-06, + "loss": 0.821, + "step": 1269 + }, + { + "epoch": 0.11453307480723272, + "grad_norm": 1.4501718024714692, + "learning_rate": 3.925553342155421e-06, + "loss": 0.7659, + "step": 1270 + }, + { + "epoch": 0.11462325833070298, + "grad_norm": 0.6937499115058912, + "learning_rate": 3.9253953489651485e-06, + "loss": 0.6257, + "step": 1271 + }, + { + "epoch": 0.11471344185417325, + "grad_norm": 0.7380979199331464, + "learning_rate": 3.925237191489529e-06, + "loss": 0.601, + "step": 1272 + }, + { + "epoch": 0.1148036253776435, + "grad_norm": 1.5977257151530126, + "learning_rate": 3.925078869742056e-06, + "loss": 0.7789, + "step": 1273 + }, + { + "epoch": 0.11489380890111377, + "grad_norm": 1.6639194337478556, + "learning_rate": 3.92492038373624e-06, + "loss": 0.7461, + "step": 1274 + }, + { + "epoch": 0.11498399242458403, + "grad_norm": 2.182736141368373, + "learning_rate": 3.924761733485602e-06, + "loss": 0.8671, + "step": 1275 + }, + { + "epoch": 0.11507417594805429, + "grad_norm": 1.4887661682070903, + "learning_rate": 3.92460291900368e-06, + "loss": 0.7994, + "step": 1276 + }, + { + "epoch": 0.11516435947152455, + "grad_norm": 2.0714132585570244, + "learning_rate": 3.924443940304025e-06, + "loss": 0.8303, + "step": 1277 + }, + { + "epoch": 0.11525454299499481, + "grad_norm": 1.9127354595373054, + "learning_rate": 3.924284797400202e-06, + "loss": 0.8649, + "step": 1278 + }, + { + "epoch": 0.11534472651846507, + "grad_norm": 2.3049707271016286, + "learning_rate": 3.924125490305789e-06, + "loss": 0.8362, + "step": 1279 + }, + { + "epoch": 0.11543491004193533, + "grad_norm": 2.0652291707853028, + "learning_rate": 3.923966019034381e-06, + "loss": 0.7904, + "step": 1280 + }, + { + "epoch": 0.1155250935654056, + "grad_norm": 1.8572237408494512, + "learning_rate": 3.923806383599583e-06, + "loss": 0.8221, + "step": 1281 + }, + { + "epoch": 0.11561527708887587, + "grad_norm": 1.65427597735288, + "learning_rate": 3.923646584015017e-06, + "loss": 0.778, + "step": 1282 + }, + { + "epoch": 0.11570546061234613, + "grad_norm": 2.104845057242062, + "learning_rate": 3.923486620294316e-06, + "loss": 0.8136, + "step": 1283 + }, + { + "epoch": 0.11579564413581639, + "grad_norm": 1.8671336146284465, + "learning_rate": 3.923326492451132e-06, + "loss": 0.8257, + "step": 1284 + }, + { + "epoch": 0.11588582765928665, + "grad_norm": 2.1911890802875247, + "learning_rate": 3.923166200499125e-06, + "loss": 0.7877, + "step": 1285 + }, + { + "epoch": 0.11597601118275691, + "grad_norm": 1.6052405573411679, + "learning_rate": 3.923005744451975e-06, + "loss": 0.7763, + "step": 1286 + }, + { + "epoch": 0.11606619470622717, + "grad_norm": 2.097298191987902, + "learning_rate": 3.9228451243233715e-06, + "loss": 0.7585, + "step": 1287 + }, + { + "epoch": 0.11615637822969743, + "grad_norm": 2.2373497640796782, + "learning_rate": 3.9226843401270195e-06, + "loss": 0.8383, + "step": 1288 + }, + { + "epoch": 0.1162465617531677, + "grad_norm": 1.7945916969040066, + "learning_rate": 3.9225233918766376e-06, + "loss": 0.8136, + "step": 1289 + }, + { + "epoch": 0.11633674527663795, + "grad_norm": 1.5547368794176046, + "learning_rate": 3.92236227958596e-06, + "loss": 0.8225, + "step": 1290 + }, + { + "epoch": 0.11642692880010821, + "grad_norm": 1.9531301269463959, + "learning_rate": 3.922201003268731e-06, + "loss": 0.8307, + "step": 1291 + }, + { + "epoch": 0.11651711232357848, + "grad_norm": 0.6919338131712399, + "learning_rate": 3.922039562938715e-06, + "loss": 0.6274, + "step": 1292 + }, + { + "epoch": 0.11660729584704875, + "grad_norm": 1.6609191577720377, + "learning_rate": 3.921877958609685e-06, + "loss": 0.8572, + "step": 1293 + }, + { + "epoch": 0.11669747937051901, + "grad_norm": 0.7834649540184359, + "learning_rate": 3.921716190295431e-06, + "loss": 0.6919, + "step": 1294 + }, + { + "epoch": 0.11678766289398927, + "grad_norm": 1.820819767683927, + "learning_rate": 3.921554258009755e-06, + "loss": 0.7677, + "step": 1295 + }, + { + "epoch": 0.11687784641745953, + "grad_norm": 2.2510190351538024, + "learning_rate": 3.921392161766474e-06, + "loss": 0.7955, + "step": 1296 + }, + { + "epoch": 0.1169680299409298, + "grad_norm": 2.151171467138376, + "learning_rate": 3.92122990157942e-06, + "loss": 0.792, + "step": 1297 + }, + { + "epoch": 0.11705821346440005, + "grad_norm": 1.8695852932609054, + "learning_rate": 3.921067477462437e-06, + "loss": 0.8542, + "step": 1298 + }, + { + "epoch": 0.11714839698787032, + "grad_norm": 2.2264716949685424, + "learning_rate": 3.920904889429385e-06, + "loss": 0.866, + "step": 1299 + }, + { + "epoch": 0.11723858051134058, + "grad_norm": 2.3372244973435192, + "learning_rate": 3.920742137494135e-06, + "loss": 0.8518, + "step": 1300 + }, + { + "epoch": 0.11732876403481084, + "grad_norm": 1.738068093935826, + "learning_rate": 3.920579221670575e-06, + "loss": 0.8681, + "step": 1301 + }, + { + "epoch": 0.1174189475582811, + "grad_norm": 1.762171042294946, + "learning_rate": 3.920416141972606e-06, + "loss": 0.8142, + "step": 1302 + }, + { + "epoch": 0.11750913108175136, + "grad_norm": 1.506508614273334, + "learning_rate": 3.920252898414143e-06, + "loss": 0.8041, + "step": 1303 + }, + { + "epoch": 0.11759931460522162, + "grad_norm": 2.351370566003107, + "learning_rate": 3.920089491009114e-06, + "loss": 0.8233, + "step": 1304 + }, + { + "epoch": 0.1176894981286919, + "grad_norm": 1.8303031009577984, + "learning_rate": 3.919925919771463e-06, + "loss": 0.8904, + "step": 1305 + }, + { + "epoch": 0.11777968165216215, + "grad_norm": 1.8846780232839733, + "learning_rate": 3.919762184715146e-06, + "loss": 0.7248, + "step": 1306 + }, + { + "epoch": 0.11786986517563242, + "grad_norm": 2.1567456325399674, + "learning_rate": 3.919598285854134e-06, + "loss": 0.8067, + "step": 1307 + }, + { + "epoch": 0.11796004869910268, + "grad_norm": 1.9170608391617006, + "learning_rate": 3.919434223202411e-06, + "loss": 0.8636, + "step": 1308 + }, + { + "epoch": 0.11805023222257294, + "grad_norm": 1.8961647777983295, + "learning_rate": 3.919269996773977e-06, + "loss": 0.8614, + "step": 1309 + }, + { + "epoch": 0.1181404157460432, + "grad_norm": 1.4804040094477469, + "learning_rate": 3.919105606582844e-06, + "loss": 0.8242, + "step": 1310 + }, + { + "epoch": 0.11823059926951346, + "grad_norm": 1.806022645573882, + "learning_rate": 3.918941052643039e-06, + "loss": 0.7847, + "step": 1311 + }, + { + "epoch": 0.11832078279298372, + "grad_norm": 1.6891635889226988, + "learning_rate": 3.918776334968602e-06, + "loss": 0.821, + "step": 1312 + }, + { + "epoch": 0.11841096631645398, + "grad_norm": 1.9454683295039539, + "learning_rate": 3.918611453573589e-06, + "loss": 0.9191, + "step": 1313 + }, + { + "epoch": 0.11850114983992424, + "grad_norm": 2.009582333228031, + "learning_rate": 3.918446408472066e-06, + "loss": 0.7448, + "step": 1314 + }, + { + "epoch": 0.1185913333633945, + "grad_norm": 1.5979166354992136, + "learning_rate": 3.918281199678119e-06, + "loss": 0.8487, + "step": 1315 + }, + { + "epoch": 0.11868151688686476, + "grad_norm": 0.7330946210467959, + "learning_rate": 3.9181158272058414e-06, + "loss": 0.6094, + "step": 1316 + }, + { + "epoch": 0.11877170041033504, + "grad_norm": 1.5890086837252797, + "learning_rate": 3.9179502910693455e-06, + "loss": 0.8375, + "step": 1317 + }, + { + "epoch": 0.1188618839338053, + "grad_norm": 1.752158264602047, + "learning_rate": 3.917784591282756e-06, + "loss": 0.7672, + "step": 1318 + }, + { + "epoch": 0.11895206745727556, + "grad_norm": 1.7641386003864012, + "learning_rate": 3.9176187278602105e-06, + "loss": 0.8323, + "step": 1319 + }, + { + "epoch": 0.11904225098074582, + "grad_norm": 1.51256385857574, + "learning_rate": 3.9174527008158606e-06, + "loss": 0.7514, + "step": 1320 + }, + { + "epoch": 0.11913243450421608, + "grad_norm": 1.5129963998140312, + "learning_rate": 3.917286510163874e-06, + "loss": 0.733, + "step": 1321 + }, + { + "epoch": 0.11922261802768634, + "grad_norm": 1.316759826084641, + "learning_rate": 3.917120155918431e-06, + "loss": 0.7797, + "step": 1322 + }, + { + "epoch": 0.1193128015511566, + "grad_norm": 0.7407789818729613, + "learning_rate": 3.916953638093725e-06, + "loss": 0.6411, + "step": 1323 + }, + { + "epoch": 0.11940298507462686, + "grad_norm": 1.9296382635221312, + "learning_rate": 3.916786956703964e-06, + "loss": 0.8027, + "step": 1324 + }, + { + "epoch": 0.11949316859809712, + "grad_norm": 3.8251576659176307, + "learning_rate": 3.916620111763372e-06, + "loss": 0.8077, + "step": 1325 + }, + { + "epoch": 0.11958335212156739, + "grad_norm": 1.6842952957121144, + "learning_rate": 3.916453103286183e-06, + "loss": 0.8699, + "step": 1326 + }, + { + "epoch": 0.11967353564503765, + "grad_norm": 2.3803307789054013, + "learning_rate": 3.916285931286648e-06, + "loss": 0.8544, + "step": 1327 + }, + { + "epoch": 0.11976371916850792, + "grad_norm": 1.724605918872174, + "learning_rate": 3.916118595779031e-06, + "loss": 0.7558, + "step": 1328 + }, + { + "epoch": 0.11985390269197818, + "grad_norm": 1.7433076003359356, + "learning_rate": 3.915951096777611e-06, + "loss": 0.7867, + "step": 1329 + }, + { + "epoch": 0.11994408621544844, + "grad_norm": 1.7584441003968767, + "learning_rate": 3.915783434296678e-06, + "loss": 0.7634, + "step": 1330 + }, + { + "epoch": 0.1200342697389187, + "grad_norm": 2.0520821069497512, + "learning_rate": 3.91561560835054e-06, + "loss": 0.7908, + "step": 1331 + }, + { + "epoch": 0.12012445326238896, + "grad_norm": 1.599295001874651, + "learning_rate": 3.915447618953515e-06, + "loss": 0.8261, + "step": 1332 + }, + { + "epoch": 0.12021463678585922, + "grad_norm": 2.231341884227758, + "learning_rate": 3.915279466119937e-06, + "loss": 0.7653, + "step": 1333 + }, + { + "epoch": 0.12030482030932949, + "grad_norm": 0.725687751455233, + "learning_rate": 3.9151111498641546e-06, + "loss": 0.6257, + "step": 1334 + }, + { + "epoch": 0.12039500383279975, + "grad_norm": 2.095293529404707, + "learning_rate": 3.914942670200529e-06, + "loss": 0.6913, + "step": 1335 + }, + { + "epoch": 0.12048518735627001, + "grad_norm": 2.097787863582317, + "learning_rate": 3.914774027143436e-06, + "loss": 0.8662, + "step": 1336 + }, + { + "epoch": 0.12057537087974027, + "grad_norm": 3.175436191182632, + "learning_rate": 3.914605220707265e-06, + "loss": 0.7978, + "step": 1337 + }, + { + "epoch": 0.12066555440321053, + "grad_norm": 1.808361986112001, + "learning_rate": 3.9144362509064194e-06, + "loss": 0.9561, + "step": 1338 + }, + { + "epoch": 0.12075573792668079, + "grad_norm": 3.4757545475310274, + "learning_rate": 3.914267117755317e-06, + "loss": 0.8783, + "step": 1339 + }, + { + "epoch": 0.12084592145015106, + "grad_norm": 1.859310052843358, + "learning_rate": 3.914097821268389e-06, + "loss": 0.7738, + "step": 1340 + }, + { + "epoch": 0.12093610497362133, + "grad_norm": 1.7276414632772523, + "learning_rate": 3.913928361460081e-06, + "loss": 0.8231, + "step": 1341 + }, + { + "epoch": 0.12102628849709159, + "grad_norm": 1.9711116371144526, + "learning_rate": 3.913758738344851e-06, + "loss": 0.7919, + "step": 1342 + }, + { + "epoch": 0.12111647202056185, + "grad_norm": 0.7460283180824441, + "learning_rate": 3.913588951937174e-06, + "loss": 0.6028, + "step": 1343 + }, + { + "epoch": 0.12120665554403211, + "grad_norm": 0.7113735831753825, + "learning_rate": 3.9134190022515355e-06, + "loss": 0.6021, + "step": 1344 + }, + { + "epoch": 0.12129683906750237, + "grad_norm": 1.7684453533380442, + "learning_rate": 3.913248889302438e-06, + "loss": 0.77, + "step": 1345 + }, + { + "epoch": 0.12138702259097263, + "grad_norm": 1.7139208141316353, + "learning_rate": 3.913078613104395e-06, + "loss": 0.7441, + "step": 1346 + }, + { + "epoch": 0.12147720611444289, + "grad_norm": 1.6727519364762689, + "learning_rate": 3.912908173671936e-06, + "loss": 0.6937, + "step": 1347 + }, + { + "epoch": 0.12156738963791315, + "grad_norm": 3.0352395143220177, + "learning_rate": 3.9127375710196044e-06, + "loss": 0.7329, + "step": 1348 + }, + { + "epoch": 0.12165757316138341, + "grad_norm": 1.9852167581931974, + "learning_rate": 3.912566805161957e-06, + "loss": 0.751, + "step": 1349 + }, + { + "epoch": 0.12174775668485367, + "grad_norm": 3.2113516510302493, + "learning_rate": 3.912395876113564e-06, + "loss": 0.8278, + "step": 1350 + }, + { + "epoch": 0.12183794020832393, + "grad_norm": 1.4400123534732323, + "learning_rate": 3.912224783889009e-06, + "loss": 0.8101, + "step": 1351 + }, + { + "epoch": 0.12192812373179421, + "grad_norm": 1.6968490739109852, + "learning_rate": 3.912053528502892e-06, + "loss": 0.818, + "step": 1352 + }, + { + "epoch": 0.12201830725526447, + "grad_norm": 1.600464085984857, + "learning_rate": 3.911882109969825e-06, + "loss": 0.7317, + "step": 1353 + }, + { + "epoch": 0.12210849077873473, + "grad_norm": 1.8084068119338443, + "learning_rate": 3.911710528304435e-06, + "loss": 0.8199, + "step": 1354 + }, + { + "epoch": 0.12219867430220499, + "grad_norm": 8.581959685494743, + "learning_rate": 3.911538783521361e-06, + "loss": 0.7607, + "step": 1355 + }, + { + "epoch": 0.12228885782567525, + "grad_norm": 2.2019727141987766, + "learning_rate": 3.9113668756352575e-06, + "loss": 0.8246, + "step": 1356 + }, + { + "epoch": 0.12237904134914551, + "grad_norm": 1.403871304645486, + "learning_rate": 3.911194804660793e-06, + "loss": 0.7349, + "step": 1357 + }, + { + "epoch": 0.12246922487261577, + "grad_norm": 1.7097462726880925, + "learning_rate": 3.91102257061265e-06, + "loss": 0.8896, + "step": 1358 + }, + { + "epoch": 0.12255940839608603, + "grad_norm": 1.4897646582442932, + "learning_rate": 3.910850173505524e-06, + "loss": 0.8482, + "step": 1359 + }, + { + "epoch": 0.1226495919195563, + "grad_norm": 2.5134248767779104, + "learning_rate": 3.9106776133541255e-06, + "loss": 0.8838, + "step": 1360 + }, + { + "epoch": 0.12273977544302656, + "grad_norm": 1.8355981391393557, + "learning_rate": 3.9105048901731766e-06, + "loss": 0.7633, + "step": 1361 + }, + { + "epoch": 0.12282995896649682, + "grad_norm": 2.2390043944587874, + "learning_rate": 3.9103320039774165e-06, + "loss": 0.7952, + "step": 1362 + }, + { + "epoch": 0.12292014248996708, + "grad_norm": 1.6737409499979963, + "learning_rate": 3.9101589547815965e-06, + "loss": 0.8637, + "step": 1363 + }, + { + "epoch": 0.12301032601343735, + "grad_norm": 1.426190377676727, + "learning_rate": 3.909985742600482e-06, + "loss": 0.7953, + "step": 1364 + }, + { + "epoch": 0.12310050953690761, + "grad_norm": 1.6482431740141392, + "learning_rate": 3.909812367448852e-06, + "loss": 0.7713, + "step": 1365 + }, + { + "epoch": 0.12319069306037787, + "grad_norm": 2.5192084527049228, + "learning_rate": 3.909638829341501e-06, + "loss": 0.785, + "step": 1366 + }, + { + "epoch": 0.12328087658384813, + "grad_norm": 1.5866199396903693, + "learning_rate": 3.909465128293234e-06, + "loss": 0.8632, + "step": 1367 + }, + { + "epoch": 0.1233710601073184, + "grad_norm": 1.9970400483839184, + "learning_rate": 3.9092912643188745e-06, + "loss": 0.8648, + "step": 1368 + }, + { + "epoch": 0.12346124363078866, + "grad_norm": 5.553943181770318, + "learning_rate": 3.909117237433256e-06, + "loss": 0.8044, + "step": 1369 + }, + { + "epoch": 0.12355142715425892, + "grad_norm": 2.9651314617173576, + "learning_rate": 3.908943047651229e-06, + "loss": 0.852, + "step": 1370 + }, + { + "epoch": 0.12364161067772918, + "grad_norm": 1.6507701550249414, + "learning_rate": 3.908768694987655e-06, + "loss": 0.793, + "step": 1371 + }, + { + "epoch": 0.12373179420119944, + "grad_norm": 2.500468496293895, + "learning_rate": 3.908594179457411e-06, + "loss": 0.7288, + "step": 1372 + }, + { + "epoch": 0.1238219777246697, + "grad_norm": 1.760711923551785, + "learning_rate": 3.908419501075388e-06, + "loss": 0.809, + "step": 1373 + }, + { + "epoch": 0.12391216124813996, + "grad_norm": 2.229173915768064, + "learning_rate": 3.90824465985649e-06, + "loss": 0.8815, + "step": 1374 + }, + { + "epoch": 0.12400234477161022, + "grad_norm": 1.764840618534358, + "learning_rate": 3.908069655815636e-06, + "loss": 0.8491, + "step": 1375 + }, + { + "epoch": 0.1240925282950805, + "grad_norm": 2.094486050258579, + "learning_rate": 3.907894488967758e-06, + "loss": 0.7469, + "step": 1376 + }, + { + "epoch": 0.12418271181855076, + "grad_norm": 2.5062833027517746, + "learning_rate": 3.9077191593278005e-06, + "loss": 0.8672, + "step": 1377 + }, + { + "epoch": 0.12427289534202102, + "grad_norm": 0.7155432275420278, + "learning_rate": 3.9075436669107265e-06, + "loss": 0.5779, + "step": 1378 + }, + { + "epoch": 0.12436307886549128, + "grad_norm": 1.8234228975288267, + "learning_rate": 3.90736801173151e-06, + "loss": 0.8242, + "step": 1379 + }, + { + "epoch": 0.12445326238896154, + "grad_norm": 1.4001963120152443, + "learning_rate": 3.907192193805136e-06, + "loss": 0.8192, + "step": 1380 + }, + { + "epoch": 0.1245434459124318, + "grad_norm": 1.9588583452145847, + "learning_rate": 3.907016213146608e-06, + "loss": 0.8651, + "step": 1381 + }, + { + "epoch": 0.12463362943590206, + "grad_norm": 1.7540457145449142, + "learning_rate": 3.906840069770942e-06, + "loss": 0.9213, + "step": 1382 + }, + { + "epoch": 0.12472381295937232, + "grad_norm": 5.416763754732614, + "learning_rate": 3.906663763693167e-06, + "loss": 0.8357, + "step": 1383 + }, + { + "epoch": 0.12481399648284258, + "grad_norm": 2.537911867437698, + "learning_rate": 3.906487294928327e-06, + "loss": 0.791, + "step": 1384 + }, + { + "epoch": 0.12490418000631284, + "grad_norm": 1.9679482053477104, + "learning_rate": 3.906310663491478e-06, + "loss": 0.8095, + "step": 1385 + }, + { + "epoch": 0.1249943635297831, + "grad_norm": 1.7318644180442628, + "learning_rate": 3.906133869397692e-06, + "loss": 0.8587, + "step": 1386 + }, + { + "epoch": 0.12508454705325336, + "grad_norm": 1.784626821729833, + "learning_rate": 3.905956912662054e-06, + "loss": 0.7709, + "step": 1387 + }, + { + "epoch": 0.12517473057672363, + "grad_norm": 3.2159390814216366, + "learning_rate": 3.905779793299662e-06, + "loss": 0.8758, + "step": 1388 + }, + { + "epoch": 0.12526491410019389, + "grad_norm": 1.781204089192122, + "learning_rate": 3.905602511325631e-06, + "loss": 0.838, + "step": 1389 + }, + { + "epoch": 0.12535509762366415, + "grad_norm": 1.8463949854490127, + "learning_rate": 3.905425066755086e-06, + "loss": 0.7733, + "step": 1390 + }, + { + "epoch": 0.1254452811471344, + "grad_norm": 1.5501171006299115, + "learning_rate": 3.905247459603168e-06, + "loss": 0.7633, + "step": 1391 + }, + { + "epoch": 0.12553546467060467, + "grad_norm": 2.176340575724349, + "learning_rate": 3.905069689885031e-06, + "loss": 0.6872, + "step": 1392 + }, + { + "epoch": 0.12562564819407493, + "grad_norm": 1.462061322273834, + "learning_rate": 3.904891757615843e-06, + "loss": 0.6918, + "step": 1393 + }, + { + "epoch": 0.12571583171754522, + "grad_norm": 0.6161450140703891, + "learning_rate": 3.9047136628107874e-06, + "loss": 0.5642, + "step": 1394 + }, + { + "epoch": 0.12580601524101548, + "grad_norm": 1.72451163303326, + "learning_rate": 3.904535405485059e-06, + "loss": 0.9325, + "step": 1395 + }, + { + "epoch": 0.12589619876448574, + "grad_norm": 4.325220999831997, + "learning_rate": 3.90435698565387e-06, + "loss": 0.754, + "step": 1396 + }, + { + "epoch": 0.125986382287956, + "grad_norm": 4.054917757115152, + "learning_rate": 3.904178403332441e-06, + "loss": 0.7692, + "step": 1397 + }, + { + "epoch": 0.12607656581142626, + "grad_norm": 2.2332298705584783, + "learning_rate": 3.903999658536012e-06, + "loss": 0.8942, + "step": 1398 + }, + { + "epoch": 0.12616674933489652, + "grad_norm": 2.1198789936295506, + "learning_rate": 3.903820751279833e-06, + "loss": 0.807, + "step": 1399 + }, + { + "epoch": 0.12625693285836678, + "grad_norm": 2.801191100455003, + "learning_rate": 3.90364168157917e-06, + "loss": 0.8444, + "step": 1400 + }, + { + "epoch": 0.12634711638183704, + "grad_norm": 1.7414901090259471, + "learning_rate": 3.903462449449302e-06, + "loss": 0.841, + "step": 1401 + }, + { + "epoch": 0.1264372999053073, + "grad_norm": 3.012458047535132, + "learning_rate": 3.903283054905522e-06, + "loss": 0.7786, + "step": 1402 + }, + { + "epoch": 0.12652748342877757, + "grad_norm": 2.4167082772016, + "learning_rate": 3.9031034979631385e-06, + "loss": 0.7236, + "step": 1403 + }, + { + "epoch": 0.12661766695224783, + "grad_norm": 1.6986210615951916, + "learning_rate": 3.902923778637469e-06, + "loss": 0.8402, + "step": 1404 + }, + { + "epoch": 0.1267078504757181, + "grad_norm": 1.6287014693014605, + "learning_rate": 3.902743896943852e-06, + "loss": 0.7405, + "step": 1405 + }, + { + "epoch": 0.12679803399918835, + "grad_norm": 0.7239077891171432, + "learning_rate": 3.902563852897633e-06, + "loss": 0.6326, + "step": 1406 + }, + { + "epoch": 0.1268882175226586, + "grad_norm": 0.702148734611268, + "learning_rate": 3.9023836465141755e-06, + "loss": 0.6071, + "step": 1407 + }, + { + "epoch": 0.12697840104612887, + "grad_norm": 1.8613506688036672, + "learning_rate": 3.902203277808856e-06, + "loss": 0.8548, + "step": 1408 + }, + { + "epoch": 0.12706858456959913, + "grad_norm": 1.6721361615850094, + "learning_rate": 3.902022746797064e-06, + "loss": 0.8643, + "step": 1409 + }, + { + "epoch": 0.1271587680930694, + "grad_norm": 1.9560498119946728, + "learning_rate": 3.9018420534942035e-06, + "loss": 0.9179, + "step": 1410 + }, + { + "epoch": 0.12724895161653965, + "grad_norm": 2.168404310307514, + "learning_rate": 3.9016611979156935e-06, + "loss": 0.804, + "step": 1411 + }, + { + "epoch": 0.1273391351400099, + "grad_norm": 2.7122938297516166, + "learning_rate": 3.9014801800769635e-06, + "loss": 0.7234, + "step": 1412 + }, + { + "epoch": 0.12742931866348017, + "grad_norm": 1.7816844878749174, + "learning_rate": 3.901298999993459e-06, + "loss": 0.8373, + "step": 1413 + }, + { + "epoch": 0.12751950218695043, + "grad_norm": 2.781361352641635, + "learning_rate": 3.901117657680642e-06, + "loss": 0.7422, + "step": 1414 + }, + { + "epoch": 0.1276096857104207, + "grad_norm": 1.6435813788310976, + "learning_rate": 3.900936153153982e-06, + "loss": 0.7398, + "step": 1415 + }, + { + "epoch": 0.12769986923389096, + "grad_norm": 0.7122390068279678, + "learning_rate": 3.900754486428968e-06, + "loss": 0.5602, + "step": 1416 + }, + { + "epoch": 0.12779005275736122, + "grad_norm": 3.1963035333750485, + "learning_rate": 3.900572657521102e-06, + "loss": 0.8955, + "step": 1417 + }, + { + "epoch": 0.1278802362808315, + "grad_norm": 2.1097871730913917, + "learning_rate": 3.900390666445896e-06, + "loss": 0.7853, + "step": 1418 + }, + { + "epoch": 0.12797041980430177, + "grad_norm": 7.843081016432398, + "learning_rate": 3.9002085132188795e-06, + "loss": 0.7763, + "step": 1419 + }, + { + "epoch": 0.12806060332777203, + "grad_norm": 1.7983403077695912, + "learning_rate": 3.9000261978555964e-06, + "loss": 0.8471, + "step": 1420 + }, + { + "epoch": 0.1281507868512423, + "grad_norm": 1.7464432675910382, + "learning_rate": 3.8998437203716e-06, + "loss": 0.8671, + "step": 1421 + }, + { + "epoch": 0.12824097037471255, + "grad_norm": 4.971004910104614, + "learning_rate": 3.899661080782462e-06, + "loss": 0.8326, + "step": 1422 + }, + { + "epoch": 0.1283311538981828, + "grad_norm": 1.5861949993485203, + "learning_rate": 3.899478279103767e-06, + "loss": 0.8421, + "step": 1423 + }, + { + "epoch": 0.12842133742165307, + "grad_norm": 2.0428848395246892, + "learning_rate": 3.8992953153511105e-06, + "loss": 0.8385, + "step": 1424 + }, + { + "epoch": 0.12851152094512333, + "grad_norm": 1.6801911774350897, + "learning_rate": 3.899112189540106e-06, + "loss": 0.8693, + "step": 1425 + }, + { + "epoch": 0.1286017044685936, + "grad_norm": 2.172704414785762, + "learning_rate": 3.898928901686377e-06, + "loss": 0.846, + "step": 1426 + }, + { + "epoch": 0.12869188799206385, + "grad_norm": 0.7092917960689188, + "learning_rate": 3.898745451805564e-06, + "loss": 0.5647, + "step": 1427 + }, + { + "epoch": 0.1287820715155341, + "grad_norm": 1.704563128639838, + "learning_rate": 3.898561839913319e-06, + "loss": 0.7402, + "step": 1428 + }, + { + "epoch": 0.12887225503900437, + "grad_norm": 1.9718856666143443, + "learning_rate": 3.89837806602531e-06, + "loss": 0.6582, + "step": 1429 + }, + { + "epoch": 0.12896243856247463, + "grad_norm": 1.8493374359761015, + "learning_rate": 3.898194130157217e-06, + "loss": 0.8168, + "step": 1430 + }, + { + "epoch": 0.1290526220859449, + "grad_norm": 1.8485330513436296, + "learning_rate": 3.8980100323247335e-06, + "loss": 0.7478, + "step": 1431 + }, + { + "epoch": 0.12914280560941516, + "grad_norm": 1.728960882910426, + "learning_rate": 3.897825772543568e-06, + "loss": 0.7431, + "step": 1432 + }, + { + "epoch": 0.12923298913288542, + "grad_norm": 1.972809557103585, + "learning_rate": 3.897641350829444e-06, + "loss": 0.8666, + "step": 1433 + }, + { + "epoch": 0.12932317265635568, + "grad_norm": 1.9126427192044215, + "learning_rate": 3.897456767198096e-06, + "loss": 0.8906, + "step": 1434 + }, + { + "epoch": 0.12941335617982594, + "grad_norm": 2.7613772946015307, + "learning_rate": 3.897272021665275e-06, + "loss": 0.746, + "step": 1435 + }, + { + "epoch": 0.1295035397032962, + "grad_norm": 1.6608629584967984, + "learning_rate": 3.897087114246743e-06, + "loss": 0.914, + "step": 1436 + }, + { + "epoch": 0.12959372322676646, + "grad_norm": 2.01532938364096, + "learning_rate": 3.896902044958279e-06, + "loss": 0.9237, + "step": 1437 + }, + { + "epoch": 0.12968390675023672, + "grad_norm": 3.088018946722935, + "learning_rate": 3.896716813815672e-06, + "loss": 0.8692, + "step": 1438 + }, + { + "epoch": 0.12977409027370698, + "grad_norm": 1.6037413871456558, + "learning_rate": 3.896531420834728e-06, + "loss": 0.7547, + "step": 1439 + }, + { + "epoch": 0.12986427379717724, + "grad_norm": 1.590274911662281, + "learning_rate": 3.896345866031266e-06, + "loss": 0.7145, + "step": 1440 + }, + { + "epoch": 0.1299544573206475, + "grad_norm": 1.7132185966700795, + "learning_rate": 3.896160149421119e-06, + "loss": 0.8685, + "step": 1441 + }, + { + "epoch": 0.1300446408441178, + "grad_norm": 1.9930540107724017, + "learning_rate": 3.8959742710201314e-06, + "loss": 0.8769, + "step": 1442 + }, + { + "epoch": 0.13013482436758805, + "grad_norm": 2.060219515428322, + "learning_rate": 3.895788230844166e-06, + "loss": 0.7112, + "step": 1443 + }, + { + "epoch": 0.13022500789105831, + "grad_norm": 1.799069664857879, + "learning_rate": 3.895602028909095e-06, + "loss": 0.9684, + "step": 1444 + }, + { + "epoch": 0.13031519141452858, + "grad_norm": 0.6429899339674072, + "learning_rate": 3.895415665230807e-06, + "loss": 0.5987, + "step": 1445 + }, + { + "epoch": 0.13040537493799884, + "grad_norm": 1.2620819797216851, + "learning_rate": 3.895229139825203e-06, + "loss": 0.8323, + "step": 1446 + }, + { + "epoch": 0.1304955584614691, + "grad_norm": 1.7352784012038092, + "learning_rate": 3.895042452708198e-06, + "loss": 0.726, + "step": 1447 + }, + { + "epoch": 0.13058574198493936, + "grad_norm": 1.7749841474443058, + "learning_rate": 3.894855603895723e-06, + "loss": 0.718, + "step": 1448 + }, + { + "epoch": 0.13067592550840962, + "grad_norm": 1.7871624047588055, + "learning_rate": 3.894668593403718e-06, + "loss": 0.8871, + "step": 1449 + }, + { + "epoch": 0.13076610903187988, + "grad_norm": 1.7714885471887887, + "learning_rate": 3.8944814212481425e-06, + "loss": 0.8123, + "step": 1450 + }, + { + "epoch": 0.13085629255535014, + "grad_norm": 1.9636451903017984, + "learning_rate": 3.894294087444966e-06, + "loss": 0.7904, + "step": 1451 + }, + { + "epoch": 0.1309464760788204, + "grad_norm": 1.964874146000754, + "learning_rate": 3.894106592010173e-06, + "loss": 0.7062, + "step": 1452 + }, + { + "epoch": 0.13103665960229066, + "grad_norm": 1.5972570093024696, + "learning_rate": 3.893918934959762e-06, + "loss": 0.6865, + "step": 1453 + }, + { + "epoch": 0.13112684312576092, + "grad_norm": 3.6593563870967314, + "learning_rate": 3.893731116309743e-06, + "loss": 0.7462, + "step": 1454 + }, + { + "epoch": 0.13121702664923118, + "grad_norm": 1.965725649635681, + "learning_rate": 3.893543136076145e-06, + "loss": 0.8158, + "step": 1455 + }, + { + "epoch": 0.13130721017270144, + "grad_norm": 2.289553124249041, + "learning_rate": 3.893354994275006e-06, + "loss": 0.8361, + "step": 1456 + }, + { + "epoch": 0.1313973936961717, + "grad_norm": 13.046424720897557, + "learning_rate": 3.893166690922378e-06, + "loss": 0.7763, + "step": 1457 + }, + { + "epoch": 0.13148757721964197, + "grad_norm": 2.3249650029655604, + "learning_rate": 3.892978226034329e-06, + "loss": 0.8439, + "step": 1458 + }, + { + "epoch": 0.13157776074311223, + "grad_norm": 2.0545878222985148, + "learning_rate": 3.89278959962694e-06, + "loss": 0.852, + "step": 1459 + }, + { + "epoch": 0.1316679442665825, + "grad_norm": 1.7388090211132754, + "learning_rate": 3.8926008117163056e-06, + "loss": 0.7839, + "step": 1460 + }, + { + "epoch": 0.13175812779005275, + "grad_norm": 1.8281655592372374, + "learning_rate": 3.892411862318535e-06, + "loss": 0.7922, + "step": 1461 + }, + { + "epoch": 0.131848311313523, + "grad_norm": 1.8632765795891086, + "learning_rate": 3.892222751449749e-06, + "loss": 0.8275, + "step": 1462 + }, + { + "epoch": 0.13193849483699327, + "grad_norm": 2.0037198045041524, + "learning_rate": 3.892033479126084e-06, + "loss": 0.76, + "step": 1463 + }, + { + "epoch": 0.13202867836046353, + "grad_norm": 2.805715048592591, + "learning_rate": 3.891844045363691e-06, + "loss": 0.7912, + "step": 1464 + }, + { + "epoch": 0.13211886188393382, + "grad_norm": 1.861595174679734, + "learning_rate": 3.891654450178732e-06, + "loss": 0.733, + "step": 1465 + }, + { + "epoch": 0.13220904540740408, + "grad_norm": 1.5001048210394616, + "learning_rate": 3.891464693587385e-06, + "loss": 0.8529, + "step": 1466 + }, + { + "epoch": 0.13229922893087434, + "grad_norm": 2.5280823844269693, + "learning_rate": 3.89127477560584e-06, + "loss": 0.8481, + "step": 1467 + }, + { + "epoch": 0.1323894124543446, + "grad_norm": 1.6533911160913426, + "learning_rate": 3.891084696250304e-06, + "loss": 0.8432, + "step": 1468 + }, + { + "epoch": 0.13247959597781486, + "grad_norm": 2.3374354430071067, + "learning_rate": 3.890894455536993e-06, + "loss": 0.849, + "step": 1469 + }, + { + "epoch": 0.13256977950128512, + "grad_norm": 2.9386857257208816, + "learning_rate": 3.890704053482142e-06, + "loss": 0.8649, + "step": 1470 + }, + { + "epoch": 0.13265996302475538, + "grad_norm": 1.7490773493531913, + "learning_rate": 3.890513490101995e-06, + "loss": 0.7571, + "step": 1471 + }, + { + "epoch": 0.13275014654822564, + "grad_norm": 0.6184931356763047, + "learning_rate": 3.890322765412814e-06, + "loss": 0.6064, + "step": 1472 + }, + { + "epoch": 0.1328403300716959, + "grad_norm": 1.6345547228988324, + "learning_rate": 3.890131879430871e-06, + "loss": 0.833, + "step": 1473 + }, + { + "epoch": 0.13293051359516617, + "grad_norm": 1.64694920468591, + "learning_rate": 3.889940832172454e-06, + "loss": 0.8339, + "step": 1474 + }, + { + "epoch": 0.13302069711863643, + "grad_norm": 6.649351171520216, + "learning_rate": 3.889749623653864e-06, + "loss": 0.8515, + "step": 1475 + }, + { + "epoch": 0.1331108806421067, + "grad_norm": 1.437467906427788, + "learning_rate": 3.889558253891416e-06, + "loss": 0.7483, + "step": 1476 + }, + { + "epoch": 0.13320106416557695, + "grad_norm": 0.6294075289050791, + "learning_rate": 3.8893667229014385e-06, + "loss": 0.5826, + "step": 1477 + }, + { + "epoch": 0.1332912476890472, + "grad_norm": 1.9506132530926958, + "learning_rate": 3.8891750307002746e-06, + "loss": 0.7159, + "step": 1478 + }, + { + "epoch": 0.13338143121251747, + "grad_norm": 1.3745487946681432, + "learning_rate": 3.888983177304281e-06, + "loss": 0.7409, + "step": 1479 + }, + { + "epoch": 0.13347161473598773, + "grad_norm": 0.7477515212757446, + "learning_rate": 3.888791162729826e-06, + "loss": 0.5809, + "step": 1480 + }, + { + "epoch": 0.133561798259458, + "grad_norm": 3.48335476979743, + "learning_rate": 3.888598986993295e-06, + "loss": 0.8329, + "step": 1481 + }, + { + "epoch": 0.13365198178292825, + "grad_norm": 1.7194074587044403, + "learning_rate": 3.888406650111085e-06, + "loss": 0.8381, + "step": 1482 + }, + { + "epoch": 0.1337421653063985, + "grad_norm": 2.0364792636533973, + "learning_rate": 3.888214152099607e-06, + "loss": 0.8091, + "step": 1483 + }, + { + "epoch": 0.13383234882986877, + "grad_norm": 5.0280036641352615, + "learning_rate": 3.888021492975285e-06, + "loss": 0.8815, + "step": 1484 + }, + { + "epoch": 0.13392253235333904, + "grad_norm": 2.2608921470917736, + "learning_rate": 3.88782867275456e-06, + "loss": 0.8611, + "step": 1485 + }, + { + "epoch": 0.1340127158768093, + "grad_norm": 1.464475132656513, + "learning_rate": 3.8876356914538824e-06, + "loss": 0.7706, + "step": 1486 + }, + { + "epoch": 0.13410289940027956, + "grad_norm": 2.0513775639733276, + "learning_rate": 3.88744254908972e-06, + "loss": 0.7649, + "step": 1487 + }, + { + "epoch": 0.13419308292374982, + "grad_norm": 1.9408420533502608, + "learning_rate": 3.887249245678552e-06, + "loss": 0.7837, + "step": 1488 + }, + { + "epoch": 0.1342832664472201, + "grad_norm": 1.5861049620061367, + "learning_rate": 3.887055781236872e-06, + "loss": 0.8815, + "step": 1489 + }, + { + "epoch": 0.13437344997069037, + "grad_norm": 1.8831548735651116, + "learning_rate": 3.886862155781186e-06, + "loss": 0.6807, + "step": 1490 + }, + { + "epoch": 0.13446363349416063, + "grad_norm": 0.6801621325479685, + "learning_rate": 3.886668369328019e-06, + "loss": 0.605, + "step": 1491 + }, + { + "epoch": 0.1345538170176309, + "grad_norm": 2.3152470124454747, + "learning_rate": 3.886474421893904e-06, + "loss": 0.8379, + "step": 1492 + }, + { + "epoch": 0.13464400054110115, + "grad_norm": 1.598036070353316, + "learning_rate": 3.886280313495388e-06, + "loss": 0.7282, + "step": 1493 + }, + { + "epoch": 0.1347341840645714, + "grad_norm": 2.718472959938582, + "learning_rate": 3.886086044149035e-06, + "loss": 0.7774, + "step": 1494 + }, + { + "epoch": 0.13482436758804167, + "grad_norm": 1.507166037061642, + "learning_rate": 3.885891613871421e-06, + "loss": 0.8152, + "step": 1495 + }, + { + "epoch": 0.13491455111151193, + "grad_norm": 2.277341132710492, + "learning_rate": 3.885697022679136e-06, + "loss": 0.8575, + "step": 1496 + }, + { + "epoch": 0.1350047346349822, + "grad_norm": 1.7333109572995908, + "learning_rate": 3.885502270588784e-06, + "loss": 0.8349, + "step": 1497 + }, + { + "epoch": 0.13509491815845245, + "grad_norm": 1.9317125204841024, + "learning_rate": 3.885307357616981e-06, + "loss": 0.8036, + "step": 1498 + }, + { + "epoch": 0.13518510168192271, + "grad_norm": 1.7489518023655428, + "learning_rate": 3.885112283780359e-06, + "loss": 0.8034, + "step": 1499 + }, + { + "epoch": 0.13527528520539298, + "grad_norm": 2.9502150311737574, + "learning_rate": 3.8849170490955624e-06, + "loss": 0.6941, + "step": 1500 + }, + { + "epoch": 0.13536546872886324, + "grad_norm": 2.0999016420854346, + "learning_rate": 3.88472165357925e-06, + "loss": 0.7993, + "step": 1501 + }, + { + "epoch": 0.1354556522523335, + "grad_norm": 1.9164894754631858, + "learning_rate": 3.884526097248093e-06, + "loss": 0.761, + "step": 1502 + }, + { + "epoch": 0.13554583577580376, + "grad_norm": 2.034158004227629, + "learning_rate": 3.884330380118779e-06, + "loss": 0.8198, + "step": 1503 + }, + { + "epoch": 0.13563601929927402, + "grad_norm": 3.7673933380358045, + "learning_rate": 3.884134502208007e-06, + "loss": 0.7574, + "step": 1504 + }, + { + "epoch": 0.13572620282274428, + "grad_norm": 3.2656596058741316, + "learning_rate": 3.88393846353249e-06, + "loss": 0.9053, + "step": 1505 + }, + { + "epoch": 0.13581638634621454, + "grad_norm": 1.5555355500266943, + "learning_rate": 3.883742264108955e-06, + "loss": 0.8571, + "step": 1506 + }, + { + "epoch": 0.1359065698696848, + "grad_norm": 1.7951256238921551, + "learning_rate": 3.883545903954145e-06, + "loss": 0.8104, + "step": 1507 + }, + { + "epoch": 0.13599675339315506, + "grad_norm": 1.9134640750787095, + "learning_rate": 3.883349383084811e-06, + "loss": 0.8008, + "step": 1508 + }, + { + "epoch": 0.13608693691662532, + "grad_norm": 2.0050804460706417, + "learning_rate": 3.883152701517723e-06, + "loss": 0.7288, + "step": 1509 + }, + { + "epoch": 0.13617712044009558, + "grad_norm": 1.849651718863408, + "learning_rate": 3.882955859269664e-06, + "loss": 0.7923, + "step": 1510 + }, + { + "epoch": 0.13626730396356584, + "grad_norm": 1.6707992658051387, + "learning_rate": 3.882758856357428e-06, + "loss": 0.8097, + "step": 1511 + }, + { + "epoch": 0.1363574874870361, + "grad_norm": 2.23645498923727, + "learning_rate": 3.882561692797824e-06, + "loss": 0.8007, + "step": 1512 + }, + { + "epoch": 0.1364476710105064, + "grad_norm": 2.194916930484175, + "learning_rate": 3.882364368607677e-06, + "loss": 0.8066, + "step": 1513 + }, + { + "epoch": 0.13653785453397665, + "grad_norm": 4.870045222516969, + "learning_rate": 3.8821668838038225e-06, + "loss": 0.8177, + "step": 1514 + }, + { + "epoch": 0.13662803805744692, + "grad_norm": 3.3474850338180366, + "learning_rate": 3.881969238403111e-06, + "loss": 0.7179, + "step": 1515 + }, + { + "epoch": 0.13671822158091718, + "grad_norm": 3.2990424818880775, + "learning_rate": 3.881771432422408e-06, + "loss": 0.9335, + "step": 1516 + }, + { + "epoch": 0.13680840510438744, + "grad_norm": 1.7514118221649333, + "learning_rate": 3.88157346587859e-06, + "loss": 0.8645, + "step": 1517 + }, + { + "epoch": 0.1368985886278577, + "grad_norm": 1.540601628805243, + "learning_rate": 3.881375338788549e-06, + "loss": 0.7798, + "step": 1518 + }, + { + "epoch": 0.13698877215132796, + "grad_norm": 3.1506402046307422, + "learning_rate": 3.88117705116919e-06, + "loss": 0.8263, + "step": 1519 + }, + { + "epoch": 0.13707895567479822, + "grad_norm": 0.8156739011991138, + "learning_rate": 3.880978603037432e-06, + "loss": 0.6563, + "step": 1520 + }, + { + "epoch": 0.13716913919826848, + "grad_norm": 1.7247230086254237, + "learning_rate": 3.880779994410209e-06, + "loss": 0.8075, + "step": 1521 + }, + { + "epoch": 0.13725932272173874, + "grad_norm": 2.300659694601744, + "learning_rate": 3.880581225304466e-06, + "loss": 0.6833, + "step": 1522 + }, + { + "epoch": 0.137349506245209, + "grad_norm": 1.92741896997078, + "learning_rate": 3.880382295737163e-06, + "loss": 0.8317, + "step": 1523 + }, + { + "epoch": 0.13743968976867926, + "grad_norm": 2.7332764080756844, + "learning_rate": 3.880183205725274e-06, + "loss": 0.7861, + "step": 1524 + }, + { + "epoch": 0.13752987329214952, + "grad_norm": 2.172953824840771, + "learning_rate": 3.879983955285788e-06, + "loss": 0.7419, + "step": 1525 + }, + { + "epoch": 0.13762005681561978, + "grad_norm": 4.982931567851168, + "learning_rate": 3.879784544435703e-06, + "loss": 0.7812, + "step": 1526 + }, + { + "epoch": 0.13771024033909005, + "grad_norm": 2.457854746984734, + "learning_rate": 3.879584973192037e-06, + "loss": 0.7509, + "step": 1527 + }, + { + "epoch": 0.1378004238625603, + "grad_norm": 1.9022515057204252, + "learning_rate": 3.8793852415718165e-06, + "loss": 0.7402, + "step": 1528 + }, + { + "epoch": 0.13789060738603057, + "grad_norm": 2.108731998350938, + "learning_rate": 3.879185349592085e-06, + "loss": 0.8135, + "step": 1529 + }, + { + "epoch": 0.13798079090950083, + "grad_norm": 4.261639084670156, + "learning_rate": 3.878985297269897e-06, + "loss": 0.8272, + "step": 1530 + }, + { + "epoch": 0.1380709744329711, + "grad_norm": 2.849131428390104, + "learning_rate": 3.878785084622323e-06, + "loss": 0.8472, + "step": 1531 + }, + { + "epoch": 0.13816115795644135, + "grad_norm": 2.466141686978156, + "learning_rate": 3.878584711666447e-06, + "loss": 0.9883, + "step": 1532 + }, + { + "epoch": 0.1382513414799116, + "grad_norm": 1.8980262216660007, + "learning_rate": 3.8783841784193635e-06, + "loss": 0.8243, + "step": 1533 + }, + { + "epoch": 0.13834152500338187, + "grad_norm": 1.808298239242894, + "learning_rate": 3.8781834848981855e-06, + "loss": 0.7829, + "step": 1534 + }, + { + "epoch": 0.13843170852685213, + "grad_norm": 7.7301825363893215, + "learning_rate": 3.877982631120037e-06, + "loss": 0.7434, + "step": 1535 + }, + { + "epoch": 0.1385218920503224, + "grad_norm": 2.414530983314309, + "learning_rate": 3.877781617102053e-06, + "loss": 0.8698, + "step": 1536 + }, + { + "epoch": 0.13861207557379268, + "grad_norm": 1.8357123054209572, + "learning_rate": 3.877580442861389e-06, + "loss": 0.8399, + "step": 1537 + }, + { + "epoch": 0.13870225909726294, + "grad_norm": 3.262791915052601, + "learning_rate": 3.877379108415209e-06, + "loss": 0.8351, + "step": 1538 + }, + { + "epoch": 0.1387924426207332, + "grad_norm": 1.526318999034394, + "learning_rate": 3.8771776137806915e-06, + "loss": 0.8331, + "step": 1539 + }, + { + "epoch": 0.13888262614420346, + "grad_norm": 2.123604821355983, + "learning_rate": 3.8769759589750295e-06, + "loss": 0.7554, + "step": 1540 + }, + { + "epoch": 0.13897280966767372, + "grad_norm": 1.8494345110013461, + "learning_rate": 3.876774144015429e-06, + "loss": 0.807, + "step": 1541 + }, + { + "epoch": 0.13906299319114399, + "grad_norm": 4.053700938955689, + "learning_rate": 3.87657216891911e-06, + "loss": 0.8464, + "step": 1542 + }, + { + "epoch": 0.13915317671461425, + "grad_norm": 2.539195083677824, + "learning_rate": 3.876370033703307e-06, + "loss": 0.8029, + "step": 1543 + }, + { + "epoch": 0.1392433602380845, + "grad_norm": 1.774498710229856, + "learning_rate": 3.876167738385265e-06, + "loss": 0.708, + "step": 1544 + }, + { + "epoch": 0.13933354376155477, + "grad_norm": 1.8071990928429789, + "learning_rate": 3.875965282982247e-06, + "loss": 0.8532, + "step": 1545 + }, + { + "epoch": 0.13942372728502503, + "grad_norm": 3.062037530351363, + "learning_rate": 3.875762667511528e-06, + "loss": 0.8241, + "step": 1546 + }, + { + "epoch": 0.1395139108084953, + "grad_norm": 1.4819758250520885, + "learning_rate": 3.875559891990394e-06, + "loss": 0.8142, + "step": 1547 + }, + { + "epoch": 0.13960409433196555, + "grad_norm": 1.7356522127407665, + "learning_rate": 3.875356956436149e-06, + "loss": 0.7723, + "step": 1548 + }, + { + "epoch": 0.1396942778554358, + "grad_norm": 2.1743378431495746, + "learning_rate": 3.875153860866108e-06, + "loss": 0.7908, + "step": 1549 + }, + { + "epoch": 0.13978446137890607, + "grad_norm": 2.797595991408738, + "learning_rate": 3.8749506052976e-06, + "loss": 0.8504, + "step": 1550 + }, + { + "epoch": 0.13987464490237633, + "grad_norm": 1.8893835231981209, + "learning_rate": 3.874747189747968e-06, + "loss": 0.8344, + "step": 1551 + }, + { + "epoch": 0.1399648284258466, + "grad_norm": 2.004724762972378, + "learning_rate": 3.874543614234568e-06, + "loss": 0.8177, + "step": 1552 + }, + { + "epoch": 0.14005501194931685, + "grad_norm": 1.6241022711342015, + "learning_rate": 3.874339878774771e-06, + "loss": 0.7893, + "step": 1553 + }, + { + "epoch": 0.14014519547278712, + "grad_norm": 1.7801095088747843, + "learning_rate": 3.874135983385961e-06, + "loss": 0.8135, + "step": 1554 + }, + { + "epoch": 0.14023537899625738, + "grad_norm": 1.8205767918551263, + "learning_rate": 3.873931928085535e-06, + "loss": 0.8075, + "step": 1555 + }, + { + "epoch": 0.14032556251972764, + "grad_norm": 2.299004741615692, + "learning_rate": 3.873727712890904e-06, + "loss": 0.8955, + "step": 1556 + }, + { + "epoch": 0.1404157460431979, + "grad_norm": 2.0604355623814286, + "learning_rate": 3.873523337819493e-06, + "loss": 0.781, + "step": 1557 + }, + { + "epoch": 0.14050592956666816, + "grad_norm": 1.6466302592130482, + "learning_rate": 3.873318802888739e-06, + "loss": 0.7371, + "step": 1558 + }, + { + "epoch": 0.14059611309013842, + "grad_norm": 0.5891346995985618, + "learning_rate": 3.873114108116097e-06, + "loss": 0.5318, + "step": 1559 + }, + { + "epoch": 0.14068629661360868, + "grad_norm": 2.0709798636938013, + "learning_rate": 3.872909253519031e-06, + "loss": 0.9014, + "step": 1560 + }, + { + "epoch": 0.14077648013707897, + "grad_norm": 2.5259561635651333, + "learning_rate": 3.8727042391150195e-06, + "loss": 0.8835, + "step": 1561 + }, + { + "epoch": 0.14086666366054923, + "grad_norm": 1.7806748247924944, + "learning_rate": 3.872499064921556e-06, + "loss": 0.7503, + "step": 1562 + }, + { + "epoch": 0.1409568471840195, + "grad_norm": 1.9098411877501118, + "learning_rate": 3.872293730956149e-06, + "loss": 0.8248, + "step": 1563 + }, + { + "epoch": 0.14104703070748975, + "grad_norm": 1.923005926582301, + "learning_rate": 3.872088237236316e-06, + "loss": 0.8066, + "step": 1564 + }, + { + "epoch": 0.14113721423096, + "grad_norm": 1.9004497773706053, + "learning_rate": 3.871882583779592e-06, + "loss": 0.7169, + "step": 1565 + }, + { + "epoch": 0.14122739775443027, + "grad_norm": 1.8803514406756652, + "learning_rate": 3.871676770603525e-06, + "loss": 0.8085, + "step": 1566 + }, + { + "epoch": 0.14131758127790053, + "grad_norm": 6.051775383350521, + "learning_rate": 3.871470797725676e-06, + "loss": 0.8069, + "step": 1567 + }, + { + "epoch": 0.1414077648013708, + "grad_norm": 1.7528946641913798, + "learning_rate": 3.8712646651636185e-06, + "loss": 0.881, + "step": 1568 + }, + { + "epoch": 0.14149794832484106, + "grad_norm": 1.5106450489183125, + "learning_rate": 3.871058372934942e-06, + "loss": 0.8049, + "step": 1569 + }, + { + "epoch": 0.14158813184831132, + "grad_norm": 2.3099148709022654, + "learning_rate": 3.8708519210572485e-06, + "loss": 0.6448, + "step": 1570 + }, + { + "epoch": 0.14167831537178158, + "grad_norm": 1.8631509859306925, + "learning_rate": 3.870645309548153e-06, + "loss": 0.7547, + "step": 1571 + }, + { + "epoch": 0.14176849889525184, + "grad_norm": 0.6076472315139877, + "learning_rate": 3.870438538425284e-06, + "loss": 0.5448, + "step": 1572 + }, + { + "epoch": 0.1418586824187221, + "grad_norm": 2.2737179088372605, + "learning_rate": 3.870231607706287e-06, + "loss": 0.7616, + "step": 1573 + }, + { + "epoch": 0.14194886594219236, + "grad_norm": 1.8930802599360896, + "learning_rate": 3.870024517408817e-06, + "loss": 0.8156, + "step": 1574 + }, + { + "epoch": 0.14203904946566262, + "grad_norm": 3.083165722664464, + "learning_rate": 3.8698172675505425e-06, + "loss": 0.7368, + "step": 1575 + }, + { + "epoch": 0.14212923298913288, + "grad_norm": 0.7479887459183716, + "learning_rate": 3.86960985814915e-06, + "loss": 0.5592, + "step": 1576 + }, + { + "epoch": 0.14221941651260314, + "grad_norm": 1.7298153575861623, + "learning_rate": 3.869402289222335e-06, + "loss": 0.7395, + "step": 1577 + }, + { + "epoch": 0.1423096000360734, + "grad_norm": 2.7093978159294694, + "learning_rate": 3.869194560787808e-06, + "loss": 0.7885, + "step": 1578 + }, + { + "epoch": 0.14239978355954366, + "grad_norm": 1.7633901985796845, + "learning_rate": 3.868986672863296e-06, + "loss": 0.7369, + "step": 1579 + }, + { + "epoch": 0.14248996708301392, + "grad_norm": 1.582426045453959, + "learning_rate": 3.868778625466535e-06, + "loss": 0.9078, + "step": 1580 + }, + { + "epoch": 0.14258015060648419, + "grad_norm": 1.7607998703633754, + "learning_rate": 3.868570418615278e-06, + "loss": 0.7346, + "step": 1581 + }, + { + "epoch": 0.14267033412995445, + "grad_norm": 0.7530433974634446, + "learning_rate": 3.8683620523272885e-06, + "loss": 0.5675, + "step": 1582 + }, + { + "epoch": 0.1427605176534247, + "grad_norm": 2.0086319138214868, + "learning_rate": 3.8681535266203464e-06, + "loss": 0.9171, + "step": 1583 + }, + { + "epoch": 0.14285070117689497, + "grad_norm": 2.0247773554346193, + "learning_rate": 3.867944841512246e-06, + "loss": 0.8331, + "step": 1584 + }, + { + "epoch": 0.14294088470036526, + "grad_norm": 2.045024821869947, + "learning_rate": 3.867735997020791e-06, + "loss": 0.8107, + "step": 1585 + }, + { + "epoch": 0.14303106822383552, + "grad_norm": 2.0254624297933534, + "learning_rate": 3.867526993163802e-06, + "loss": 0.7124, + "step": 1586 + }, + { + "epoch": 0.14312125174730578, + "grad_norm": 1.6264720264964485, + "learning_rate": 3.867317829959113e-06, + "loss": 0.6844, + "step": 1587 + }, + { + "epoch": 0.14321143527077604, + "grad_norm": 2.2828297827847766, + "learning_rate": 3.8671085074245704e-06, + "loss": 0.6068, + "step": 1588 + }, + { + "epoch": 0.1433016187942463, + "grad_norm": 4.326427360895644, + "learning_rate": 3.866899025578035e-06, + "loss": 0.876, + "step": 1589 + }, + { + "epoch": 0.14339180231771656, + "grad_norm": 1.8592424105004215, + "learning_rate": 3.86668938443738e-06, + "loss": 0.7739, + "step": 1590 + }, + { + "epoch": 0.14348198584118682, + "grad_norm": 2.9173951102183566, + "learning_rate": 3.866479584020495e-06, + "loss": 0.8209, + "step": 1591 + }, + { + "epoch": 0.14357216936465708, + "grad_norm": 1.8240502747333032, + "learning_rate": 3.866269624345279e-06, + "loss": 0.8407, + "step": 1592 + }, + { + "epoch": 0.14366235288812734, + "grad_norm": 1.6823981126427323, + "learning_rate": 3.866059505429649e-06, + "loss": 0.8871, + "step": 1593 + }, + { + "epoch": 0.1437525364115976, + "grad_norm": 0.6557160657725851, + "learning_rate": 3.865849227291532e-06, + "loss": 0.5951, + "step": 1594 + }, + { + "epoch": 0.14384271993506786, + "grad_norm": 1.9429428804437021, + "learning_rate": 3.865638789948872e-06, + "loss": 0.7962, + "step": 1595 + }, + { + "epoch": 0.14393290345853813, + "grad_norm": 2.027884997257801, + "learning_rate": 3.865428193419622e-06, + "loss": 0.762, + "step": 1596 + }, + { + "epoch": 0.14402308698200839, + "grad_norm": 1.866326552582051, + "learning_rate": 3.865217437721753e-06, + "loss": 0.7969, + "step": 1597 + }, + { + "epoch": 0.14411327050547865, + "grad_norm": 2.3466095836516856, + "learning_rate": 3.865006522873249e-06, + "loss": 0.7361, + "step": 1598 + }, + { + "epoch": 0.1442034540289489, + "grad_norm": 1.572969532598193, + "learning_rate": 3.864795448892103e-06, + "loss": 0.8013, + "step": 1599 + }, + { + "epoch": 0.14429363755241917, + "grad_norm": 2.109818475517216, + "learning_rate": 3.864584215796327e-06, + "loss": 0.8538, + "step": 1600 + }, + { + "epoch": 0.14438382107588943, + "grad_norm": 1.7942813896972287, + "learning_rate": 3.8643728236039455e-06, + "loss": 0.8266, + "step": 1601 + }, + { + "epoch": 0.1444740045993597, + "grad_norm": 2.3728027218150336, + "learning_rate": 3.864161272332994e-06, + "loss": 0.7867, + "step": 1602 + }, + { + "epoch": 0.14456418812282995, + "grad_norm": 1.8474805558850056, + "learning_rate": 3.863949562001524e-06, + "loss": 0.8459, + "step": 1603 + }, + { + "epoch": 0.1446543716463002, + "grad_norm": 1.7638473338975127, + "learning_rate": 3.8637376926276005e-06, + "loss": 0.7765, + "step": 1604 + }, + { + "epoch": 0.14474455516977047, + "grad_norm": 7.806194232530734, + "learning_rate": 3.8635256642293e-06, + "loss": 0.8227, + "step": 1605 + }, + { + "epoch": 0.14483473869324073, + "grad_norm": 1.9575592364316319, + "learning_rate": 3.863313476824714e-06, + "loss": 0.7833, + "step": 1606 + }, + { + "epoch": 0.144924922216711, + "grad_norm": 1.9656699777378717, + "learning_rate": 3.863101130431948e-06, + "loss": 0.7886, + "step": 1607 + }, + { + "epoch": 0.14501510574018128, + "grad_norm": 1.5966366823336908, + "learning_rate": 3.862888625069121e-06, + "loss": 0.8469, + "step": 1608 + }, + { + "epoch": 0.14510528926365154, + "grad_norm": 1.6295535770444536, + "learning_rate": 3.8626759607543645e-06, + "loss": 0.7644, + "step": 1609 + }, + { + "epoch": 0.1451954727871218, + "grad_norm": 1.8534868758678282, + "learning_rate": 3.862463137505825e-06, + "loss": 0.7677, + "step": 1610 + }, + { + "epoch": 0.14528565631059207, + "grad_norm": 2.868099389858626, + "learning_rate": 3.862250155341659e-06, + "loss": 0.7707, + "step": 1611 + }, + { + "epoch": 0.14537583983406233, + "grad_norm": 1.7445816900577948, + "learning_rate": 3.862037014280043e-06, + "loss": 0.9309, + "step": 1612 + }, + { + "epoch": 0.1454660233575326, + "grad_norm": 1.7356376519461514, + "learning_rate": 3.861823714339162e-06, + "loss": 0.7999, + "step": 1613 + }, + { + "epoch": 0.14555620688100285, + "grad_norm": 1.596755764983237, + "learning_rate": 3.861610255537215e-06, + "loss": 0.8152, + "step": 1614 + }, + { + "epoch": 0.1456463904044731, + "grad_norm": 1.7439182733538785, + "learning_rate": 3.8613966378924165e-06, + "loss": 0.8393, + "step": 1615 + }, + { + "epoch": 0.14573657392794337, + "grad_norm": 1.978753906843771, + "learning_rate": 3.861182861422993e-06, + "loss": 0.7515, + "step": 1616 + }, + { + "epoch": 0.14582675745141363, + "grad_norm": 1.667485473089123, + "learning_rate": 3.860968926147185e-06, + "loss": 0.8772, + "step": 1617 + }, + { + "epoch": 0.1459169409748839, + "grad_norm": 1.4660127984824454, + "learning_rate": 3.860754832083247e-06, + "loss": 0.6391, + "step": 1618 + }, + { + "epoch": 0.14600712449835415, + "grad_norm": 2.2755596310960677, + "learning_rate": 3.8605405792494475e-06, + "loss": 0.8741, + "step": 1619 + }, + { + "epoch": 0.1460973080218244, + "grad_norm": 1.891321038363625, + "learning_rate": 3.860326167664066e-06, + "loss": 0.8204, + "step": 1620 + }, + { + "epoch": 0.14618749154529467, + "grad_norm": 1.7403773325075875, + "learning_rate": 3.860111597345399e-06, + "loss": 0.7913, + "step": 1621 + }, + { + "epoch": 0.14627767506876493, + "grad_norm": 2.0464002917819677, + "learning_rate": 3.859896868311753e-06, + "loss": 0.6642, + "step": 1622 + }, + { + "epoch": 0.1463678585922352, + "grad_norm": 1.5958787501606233, + "learning_rate": 3.859681980581452e-06, + "loss": 0.8953, + "step": 1623 + }, + { + "epoch": 0.14645804211570546, + "grad_norm": 1.7413679170903285, + "learning_rate": 3.859466934172829e-06, + "loss": 0.777, + "step": 1624 + }, + { + "epoch": 0.14654822563917572, + "grad_norm": 1.8630739492583952, + "learning_rate": 3.859251729104235e-06, + "loss": 0.7335, + "step": 1625 + }, + { + "epoch": 0.14663840916264598, + "grad_norm": 1.5994150791311719, + "learning_rate": 3.859036365394031e-06, + "loss": 0.7554, + "step": 1626 + }, + { + "epoch": 0.14672859268611624, + "grad_norm": 1.9192458366377705, + "learning_rate": 3.858820843060594e-06, + "loss": 0.7868, + "step": 1627 + }, + { + "epoch": 0.1468187762095865, + "grad_norm": 1.6663639826897922, + "learning_rate": 3.858605162122314e-06, + "loss": 0.7769, + "step": 1628 + }, + { + "epoch": 0.14690895973305676, + "grad_norm": 1.7586827137431638, + "learning_rate": 3.858389322597592e-06, + "loss": 0.76, + "step": 1629 + }, + { + "epoch": 0.14699914325652702, + "grad_norm": 2.1273422795969026, + "learning_rate": 3.858173324504847e-06, + "loss": 0.819, + "step": 1630 + }, + { + "epoch": 0.14708932677999728, + "grad_norm": 3.5869919633182437, + "learning_rate": 3.857957167862508e-06, + "loss": 0.7645, + "step": 1631 + }, + { + "epoch": 0.14717951030346757, + "grad_norm": 0.6742616889514658, + "learning_rate": 3.857740852689018e-06, + "loss": 0.58, + "step": 1632 + }, + { + "epoch": 0.14726969382693783, + "grad_norm": 3.4033692103284245, + "learning_rate": 3.857524379002835e-06, + "loss": 0.8055, + "step": 1633 + }, + { + "epoch": 0.1473598773504081, + "grad_norm": 2.652870901136526, + "learning_rate": 3.85730774682243e-06, + "loss": 0.8199, + "step": 1634 + }, + { + "epoch": 0.14745006087387835, + "grad_norm": 1.4536590415202357, + "learning_rate": 3.8570909561662875e-06, + "loss": 0.86, + "step": 1635 + }, + { + "epoch": 0.1475402443973486, + "grad_norm": 1.5380859375, + "learning_rate": 3.8568740070529045e-06, + "loss": 0.862, + "step": 1636 + }, + { + "epoch": 0.14763042792081887, + "grad_norm": 1.7113963169494706, + "learning_rate": 3.856656899500792e-06, + "loss": 0.7676, + "step": 1637 + }, + { + "epoch": 0.14772061144428913, + "grad_norm": 1.73117485020997, + "learning_rate": 3.856439633528476e-06, + "loss": 0.7313, + "step": 1638 + }, + { + "epoch": 0.1478107949677594, + "grad_norm": 2.904330696570889, + "learning_rate": 3.856222209154494e-06, + "loss": 0.8551, + "step": 1639 + }, + { + "epoch": 0.14790097849122966, + "grad_norm": 1.7271673601583426, + "learning_rate": 3.856004626397397e-06, + "loss": 0.8248, + "step": 1640 + }, + { + "epoch": 0.14799116201469992, + "grad_norm": 1.787106306563281, + "learning_rate": 3.855786885275753e-06, + "loss": 0.7765, + "step": 1641 + }, + { + "epoch": 0.14808134553817018, + "grad_norm": 3.389729324382424, + "learning_rate": 3.855568985808138e-06, + "loss": 0.7184, + "step": 1642 + }, + { + "epoch": 0.14817152906164044, + "grad_norm": 3.1304959605638283, + "learning_rate": 3.855350928013145e-06, + "loss": 0.7724, + "step": 1643 + }, + { + "epoch": 0.1482617125851107, + "grad_norm": 2.1135707240945383, + "learning_rate": 3.8551327119093825e-06, + "loss": 0.8019, + "step": 1644 + }, + { + "epoch": 0.14835189610858096, + "grad_norm": 1.9394568129213425, + "learning_rate": 3.854914337515467e-06, + "loss": 0.7313, + "step": 1645 + }, + { + "epoch": 0.14844207963205122, + "grad_norm": 1.7108423912055069, + "learning_rate": 3.8546958048500324e-06, + "loss": 0.6957, + "step": 1646 + }, + { + "epoch": 0.14853226315552148, + "grad_norm": 0.6857827805010739, + "learning_rate": 3.854477113931725e-06, + "loss": 0.6321, + "step": 1647 + }, + { + "epoch": 0.14862244667899174, + "grad_norm": 2.0107200143340944, + "learning_rate": 3.854258264779205e-06, + "loss": 0.821, + "step": 1648 + }, + { + "epoch": 0.148712630202462, + "grad_norm": 1.7287125814695323, + "learning_rate": 3.854039257411145e-06, + "loss": 0.7774, + "step": 1649 + }, + { + "epoch": 0.14880281372593226, + "grad_norm": 2.2272274697470507, + "learning_rate": 3.853820091846232e-06, + "loss": 0.7503, + "step": 1650 + }, + { + "epoch": 0.14889299724940253, + "grad_norm": 1.9317987915330697, + "learning_rate": 3.853600768103169e-06, + "loss": 0.8006, + "step": 1651 + }, + { + "epoch": 0.1489831807728728, + "grad_norm": 2.215092182523155, + "learning_rate": 3.853381286200667e-06, + "loss": 0.784, + "step": 1652 + }, + { + "epoch": 0.14907336429634305, + "grad_norm": 1.854954248671297, + "learning_rate": 3.853161646157453e-06, + "loss": 0.8907, + "step": 1653 + }, + { + "epoch": 0.1491635478198133, + "grad_norm": 1.691011096351712, + "learning_rate": 3.852941847992269e-06, + "loss": 0.768, + "step": 1654 + }, + { + "epoch": 0.14925373134328357, + "grad_norm": 3.6071302691594775, + "learning_rate": 3.852721891723871e-06, + "loss": 0.8145, + "step": 1655 + }, + { + "epoch": 0.14934391486675386, + "grad_norm": 1.8408771698364474, + "learning_rate": 3.852501777371025e-06, + "loss": 0.7836, + "step": 1656 + }, + { + "epoch": 0.14943409839022412, + "grad_norm": 1.8847071188903521, + "learning_rate": 3.8522815049525125e-06, + "loss": 0.8369, + "step": 1657 + }, + { + "epoch": 0.14952428191369438, + "grad_norm": 2.2319135646217636, + "learning_rate": 3.852061074487129e-06, + "loss": 0.8503, + "step": 1658 + }, + { + "epoch": 0.14961446543716464, + "grad_norm": 2.089475110776845, + "learning_rate": 3.851840485993682e-06, + "loss": 0.7068, + "step": 1659 + }, + { + "epoch": 0.1497046489606349, + "grad_norm": 2.0466673068605763, + "learning_rate": 3.851619739490994e-06, + "loss": 0.8054, + "step": 1660 + }, + { + "epoch": 0.14979483248410516, + "grad_norm": 4.575070348313411, + "learning_rate": 3.8513988349978996e-06, + "loss": 0.8249, + "step": 1661 + }, + { + "epoch": 0.14988501600757542, + "grad_norm": 3.0423464255022163, + "learning_rate": 3.851177772533249e-06, + "loss": 0.8226, + "step": 1662 + }, + { + "epoch": 0.14997519953104568, + "grad_norm": 2.285696216920552, + "learning_rate": 3.850956552115903e-06, + "loss": 0.9791, + "step": 1663 + }, + { + "epoch": 0.15006538305451594, + "grad_norm": 0.7082651236360519, + "learning_rate": 3.850735173764738e-06, + "loss": 0.5829, + "step": 1664 + }, + { + "epoch": 0.1501555665779862, + "grad_norm": 2.6729860728393744, + "learning_rate": 3.850513637498642e-06, + "loss": 0.8201, + "step": 1665 + }, + { + "epoch": 0.15024575010145647, + "grad_norm": 1.812281628312052, + "learning_rate": 3.850291943336521e-06, + "loss": 0.8225, + "step": 1666 + }, + { + "epoch": 0.15033593362492673, + "grad_norm": 1.9694417918799716, + "learning_rate": 3.850070091297287e-06, + "loss": 0.8547, + "step": 1667 + }, + { + "epoch": 0.150426117148397, + "grad_norm": 0.7381297768511749, + "learning_rate": 3.8498480813998735e-06, + "loss": 0.6744, + "step": 1668 + }, + { + "epoch": 0.15051630067186725, + "grad_norm": 1.8278278207096077, + "learning_rate": 3.84962591366322e-06, + "loss": 0.7956, + "step": 1669 + }, + { + "epoch": 0.1506064841953375, + "grad_norm": 2.1738362410703163, + "learning_rate": 3.8494035881062855e-06, + "loss": 0.8518, + "step": 1670 + }, + { + "epoch": 0.15069666771880777, + "grad_norm": 6.2888001570865235, + "learning_rate": 3.84918110474804e-06, + "loss": 0.7786, + "step": 1671 + }, + { + "epoch": 0.15078685124227803, + "grad_norm": 1.6916943452321491, + "learning_rate": 3.8489584636074655e-06, + "loss": 0.7871, + "step": 1672 + }, + { + "epoch": 0.1508770347657483, + "grad_norm": 0.7054585833149604, + "learning_rate": 3.848735664703561e-06, + "loss": 0.5664, + "step": 1673 + }, + { + "epoch": 0.15096721828921855, + "grad_norm": 1.8779648545930432, + "learning_rate": 3.8485127080553346e-06, + "loss": 0.7956, + "step": 1674 + }, + { + "epoch": 0.1510574018126888, + "grad_norm": 2.1470687667714805, + "learning_rate": 3.8482895936818115e-06, + "loss": 0.7987, + "step": 1675 + }, + { + "epoch": 0.15114758533615907, + "grad_norm": 2.0342541121091675, + "learning_rate": 3.848066321602029e-06, + "loss": 0.8141, + "step": 1676 + }, + { + "epoch": 0.15123776885962933, + "grad_norm": 1.9203233735635687, + "learning_rate": 3.847842891835038e-06, + "loss": 0.7649, + "step": 1677 + }, + { + "epoch": 0.1513279523830996, + "grad_norm": 0.6318773023221929, + "learning_rate": 3.847619304399902e-06, + "loss": 0.5813, + "step": 1678 + }, + { + "epoch": 0.15141813590656986, + "grad_norm": 2.167835018397892, + "learning_rate": 3.8473955593157e-06, + "loss": 0.8252, + "step": 1679 + }, + { + "epoch": 0.15150831943004014, + "grad_norm": 2.16077236336184, + "learning_rate": 3.847171656601522e-06, + "loss": 0.763, + "step": 1680 + }, + { + "epoch": 0.1515985029535104, + "grad_norm": 1.78990067083861, + "learning_rate": 3.846947596276473e-06, + "loss": 0.8284, + "step": 1681 + }, + { + "epoch": 0.15168868647698067, + "grad_norm": 4.164387244753732, + "learning_rate": 3.846723378359672e-06, + "loss": 0.8448, + "step": 1682 + }, + { + "epoch": 0.15177887000045093, + "grad_norm": 2.08790227246225, + "learning_rate": 3.846499002870249e-06, + "loss": 0.8072, + "step": 1683 + }, + { + "epoch": 0.1518690535239212, + "grad_norm": 2.3634450808399343, + "learning_rate": 3.846274469827349e-06, + "loss": 0.8174, + "step": 1684 + }, + { + "epoch": 0.15195923704739145, + "grad_norm": 2.0331346425186343, + "learning_rate": 3.846049779250132e-06, + "loss": 0.8089, + "step": 1685 + }, + { + "epoch": 0.1520494205708617, + "grad_norm": 0.7878108152971485, + "learning_rate": 3.845824931157769e-06, + "loss": 0.6239, + "step": 1686 + }, + { + "epoch": 0.15213960409433197, + "grad_norm": 1.7414825107858318, + "learning_rate": 3.845599925569444e-06, + "loss": 0.7906, + "step": 1687 + }, + { + "epoch": 0.15222978761780223, + "grad_norm": 2.2498895829976084, + "learning_rate": 3.845374762504357e-06, + "loss": 0.7636, + "step": 1688 + }, + { + "epoch": 0.1523199711412725, + "grad_norm": 2.23294950242436, + "learning_rate": 3.8451494419817204e-06, + "loss": 0.8573, + "step": 1689 + }, + { + "epoch": 0.15241015466474275, + "grad_norm": 2.42841671402768, + "learning_rate": 3.8449239640207594e-06, + "loss": 0.7376, + "step": 1690 + }, + { + "epoch": 0.152500338188213, + "grad_norm": 1.9063072196002087, + "learning_rate": 3.844698328640713e-06, + "loss": 0.8433, + "step": 1691 + }, + { + "epoch": 0.15259052171168327, + "grad_norm": 1.5615535921640862, + "learning_rate": 3.844472535860833e-06, + "loss": 0.8164, + "step": 1692 + }, + { + "epoch": 0.15268070523515354, + "grad_norm": 2.1528141456184624, + "learning_rate": 3.8442465857003864e-06, + "loss": 0.8478, + "step": 1693 + }, + { + "epoch": 0.1527708887586238, + "grad_norm": 1.6014354376282933, + "learning_rate": 3.844020478178653e-06, + "loss": 0.7427, + "step": 1694 + }, + { + "epoch": 0.15286107228209406, + "grad_norm": 2.031276761392221, + "learning_rate": 3.843794213314923e-06, + "loss": 0.8314, + "step": 1695 + }, + { + "epoch": 0.15295125580556432, + "grad_norm": 2.03025359143589, + "learning_rate": 3.843567791128505e-06, + "loss": 0.8286, + "step": 1696 + }, + { + "epoch": 0.15304143932903458, + "grad_norm": 2.365930809099206, + "learning_rate": 3.843341211638717e-06, + "loss": 0.7857, + "step": 1697 + }, + { + "epoch": 0.15313162285250484, + "grad_norm": 1.7559715294582843, + "learning_rate": 3.843114474864894e-06, + "loss": 0.7606, + "step": 1698 + }, + { + "epoch": 0.1532218063759751, + "grad_norm": 2.0138008083798407, + "learning_rate": 3.84288758082638e-06, + "loss": 0.7965, + "step": 1699 + }, + { + "epoch": 0.15331198989944536, + "grad_norm": 2.189397370498847, + "learning_rate": 3.842660529542536e-06, + "loss": 0.8866, + "step": 1700 + }, + { + "epoch": 0.15340217342291562, + "grad_norm": 1.6378396504624104, + "learning_rate": 3.842433321032736e-06, + "loss": 0.8143, + "step": 1701 + }, + { + "epoch": 0.15349235694638588, + "grad_norm": 1.7441063820957885, + "learning_rate": 3.842205955316365e-06, + "loss": 0.7216, + "step": 1702 + }, + { + "epoch": 0.15358254046985614, + "grad_norm": 1.7867959670114189, + "learning_rate": 3.8419784324128256e-06, + "loss": 0.704, + "step": 1703 + }, + { + "epoch": 0.15367272399332643, + "grad_norm": 1.6848333977083454, + "learning_rate": 3.841750752341529e-06, + "loss": 0.7608, + "step": 1704 + }, + { + "epoch": 0.1537629075167967, + "grad_norm": 1.8034938811771333, + "learning_rate": 3.841522915121902e-06, + "loss": 0.8093, + "step": 1705 + }, + { + "epoch": 0.15385309104026695, + "grad_norm": 2.2249951651874262, + "learning_rate": 3.841294920773387e-06, + "loss": 0.733, + "step": 1706 + }, + { + "epoch": 0.15394327456373721, + "grad_norm": 2.2382438616600813, + "learning_rate": 3.841066769315436e-06, + "loss": 0.8155, + "step": 1707 + }, + { + "epoch": 0.15403345808720748, + "grad_norm": 3.4338871783839418, + "learning_rate": 3.840838460767518e-06, + "loss": 0.8107, + "step": 1708 + }, + { + "epoch": 0.15412364161067774, + "grad_norm": 1.7243742042641983, + "learning_rate": 3.840609995149111e-06, + "loss": 0.7748, + "step": 1709 + }, + { + "epoch": 0.154213825134148, + "grad_norm": 1.8872140427788573, + "learning_rate": 3.84038137247971e-06, + "loss": 0.8085, + "step": 1710 + }, + { + "epoch": 0.15430400865761826, + "grad_norm": 2.067059995794655, + "learning_rate": 3.840152592778823e-06, + "loss": 0.7862, + "step": 1711 + }, + { + "epoch": 0.15439419218108852, + "grad_norm": 1.5999776302204387, + "learning_rate": 3.83992365606597e-06, + "loss": 0.8427, + "step": 1712 + }, + { + "epoch": 0.15448437570455878, + "grad_norm": 1.7925447116313813, + "learning_rate": 3.8396945623606855e-06, + "loss": 0.7536, + "step": 1713 + }, + { + "epoch": 0.15457455922802904, + "grad_norm": 1.7770677960138828, + "learning_rate": 3.8394653116825174e-06, + "loss": 0.7611, + "step": 1714 + }, + { + "epoch": 0.1546647427514993, + "grad_norm": 1.6047805833423483, + "learning_rate": 3.839235904051026e-06, + "loss": 0.789, + "step": 1715 + }, + { + "epoch": 0.15475492627496956, + "grad_norm": 1.943017241298888, + "learning_rate": 3.8390063394857855e-06, + "loss": 0.7796, + "step": 1716 + }, + { + "epoch": 0.15484510979843982, + "grad_norm": 1.704465985613107, + "learning_rate": 3.838776618006385e-06, + "loss": 0.884, + "step": 1717 + }, + { + "epoch": 0.15493529332191008, + "grad_norm": 1.8003475966208435, + "learning_rate": 3.838546739632423e-06, + "loss": 0.8102, + "step": 1718 + }, + { + "epoch": 0.15502547684538034, + "grad_norm": 2.4193470220542066, + "learning_rate": 3.838316704383517e-06, + "loss": 0.8436, + "step": 1719 + }, + { + "epoch": 0.1551156603688506, + "grad_norm": 1.732424971482048, + "learning_rate": 3.838086512279292e-06, + "loss": 0.8028, + "step": 1720 + }, + { + "epoch": 0.15520584389232087, + "grad_norm": 2.1112498675584046, + "learning_rate": 3.837856163339391e-06, + "loss": 0.7995, + "step": 1721 + }, + { + "epoch": 0.15529602741579113, + "grad_norm": 1.490724095663406, + "learning_rate": 3.837625657583469e-06, + "loss": 0.7283, + "step": 1722 + }, + { + "epoch": 0.1553862109392614, + "grad_norm": 1.6237927867723083, + "learning_rate": 3.837394995031193e-06, + "loss": 0.7624, + "step": 1723 + }, + { + "epoch": 0.15547639446273165, + "grad_norm": 1.8141619856281486, + "learning_rate": 3.837164175702245e-06, + "loss": 0.7998, + "step": 1724 + }, + { + "epoch": 0.1555665779862019, + "grad_norm": 1.9543681346551922, + "learning_rate": 3.836933199616319e-06, + "loss": 0.7141, + "step": 1725 + }, + { + "epoch": 0.15565676150967217, + "grad_norm": 2.6985180355174014, + "learning_rate": 3.836702066793124e-06, + "loss": 0.8678, + "step": 1726 + }, + { + "epoch": 0.15574694503314246, + "grad_norm": 1.996724008693109, + "learning_rate": 3.836470777252381e-06, + "loss": 0.8546, + "step": 1727 + }, + { + "epoch": 0.15583712855661272, + "grad_norm": 2.747712397605032, + "learning_rate": 3.836239331013825e-06, + "loss": 0.8153, + "step": 1728 + }, + { + "epoch": 0.15592731208008298, + "grad_norm": 2.3557016437842013, + "learning_rate": 3.836007728097205e-06, + "loss": 0.769, + "step": 1729 + }, + { + "epoch": 0.15601749560355324, + "grad_norm": 1.7190517507444827, + "learning_rate": 3.835775968522282e-06, + "loss": 0.7499, + "step": 1730 + }, + { + "epoch": 0.1561076791270235, + "grad_norm": 3.352444619354999, + "learning_rate": 3.83554405230883e-06, + "loss": 0.8227, + "step": 1731 + }, + { + "epoch": 0.15619786265049376, + "grad_norm": 2.3304357134388916, + "learning_rate": 3.835311979476639e-06, + "loss": 0.7795, + "step": 1732 + }, + { + "epoch": 0.15628804617396402, + "grad_norm": 4.098462602754447, + "learning_rate": 3.83507975004551e-06, + "loss": 0.6895, + "step": 1733 + }, + { + "epoch": 0.15637822969743428, + "grad_norm": 1.6242123675714095, + "learning_rate": 3.834847364035258e-06, + "loss": 0.773, + "step": 1734 + }, + { + "epoch": 0.15646841322090455, + "grad_norm": 6.976503720094625, + "learning_rate": 3.834614821465712e-06, + "loss": 0.7433, + "step": 1735 + }, + { + "epoch": 0.1565585967443748, + "grad_norm": 2.0652132394343004, + "learning_rate": 3.834382122356713e-06, + "loss": 0.8335, + "step": 1736 + }, + { + "epoch": 0.15664878026784507, + "grad_norm": 1.4886943414054372, + "learning_rate": 3.834149266728117e-06, + "loss": 0.7595, + "step": 1737 + }, + { + "epoch": 0.15673896379131533, + "grad_norm": 0.6909105711631911, + "learning_rate": 3.833916254599792e-06, + "loss": 0.6201, + "step": 1738 + }, + { + "epoch": 0.1568291473147856, + "grad_norm": 0.6437938749305808, + "learning_rate": 3.83368308599162e-06, + "loss": 0.5663, + "step": 1739 + }, + { + "epoch": 0.15691933083825585, + "grad_norm": 0.5789914211368956, + "learning_rate": 3.833449760923498e-06, + "loss": 0.5382, + "step": 1740 + }, + { + "epoch": 0.1570095143617261, + "grad_norm": 1.6597318937196306, + "learning_rate": 3.83321627941533e-06, + "loss": 0.8359, + "step": 1741 + }, + { + "epoch": 0.15709969788519637, + "grad_norm": 2.6669633124180647, + "learning_rate": 3.832982641487043e-06, + "loss": 0.851, + "step": 1742 + }, + { + "epoch": 0.15718988140866663, + "grad_norm": 1.7966746342941766, + "learning_rate": 3.832748847158568e-06, + "loss": 0.8869, + "step": 1743 + }, + { + "epoch": 0.1572800649321369, + "grad_norm": 1.589728573837862, + "learning_rate": 3.832514896449858e-06, + "loss": 0.8169, + "step": 1744 + }, + { + "epoch": 0.15737024845560715, + "grad_norm": 1.7838221934625111, + "learning_rate": 3.832280789380871e-06, + "loss": 0.7319, + "step": 1745 + }, + { + "epoch": 0.15746043197907741, + "grad_norm": 1.8994087278001883, + "learning_rate": 3.832046525971584e-06, + "loss": 0.7017, + "step": 1746 + }, + { + "epoch": 0.15755061550254768, + "grad_norm": 2.0248799135369993, + "learning_rate": 3.831812106241987e-06, + "loss": 0.8551, + "step": 1747 + }, + { + "epoch": 0.15764079902601794, + "grad_norm": 1.6246801941962905, + "learning_rate": 3.8315775302120796e-06, + "loss": 0.6985, + "step": 1748 + }, + { + "epoch": 0.1577309825494882, + "grad_norm": 1.5818803126540966, + "learning_rate": 3.831342797901878e-06, + "loss": 0.8087, + "step": 1749 + }, + { + "epoch": 0.15782116607295846, + "grad_norm": 0.6316044897950266, + "learning_rate": 3.831107909331411e-06, + "loss": 0.5787, + "step": 1750 + }, + { + "epoch": 0.15791134959642875, + "grad_norm": 3.8834743741554005, + "learning_rate": 3.830872864520721e-06, + "loss": 0.8388, + "step": 1751 + }, + { + "epoch": 0.158001533119899, + "grad_norm": 0.7127983974214, + "learning_rate": 3.830637663489862e-06, + "loss": 0.6589, + "step": 1752 + }, + { + "epoch": 0.15809171664336927, + "grad_norm": 0.7746234625123497, + "learning_rate": 3.830402306258904e-06, + "loss": 0.6892, + "step": 1753 + }, + { + "epoch": 0.15818190016683953, + "grad_norm": 3.5548141289258623, + "learning_rate": 3.830166792847929e-06, + "loss": 0.7796, + "step": 1754 + }, + { + "epoch": 0.1582720836903098, + "grad_norm": 1.8337207153727921, + "learning_rate": 3.829931123277031e-06, + "loss": 0.733, + "step": 1755 + }, + { + "epoch": 0.15836226721378005, + "grad_norm": 1.7608866619093029, + "learning_rate": 3.8296952975663204e-06, + "loss": 0.7865, + "step": 1756 + }, + { + "epoch": 0.1584524507372503, + "grad_norm": 1.6341369222099429, + "learning_rate": 3.829459315735918e-06, + "loss": 0.8594, + "step": 1757 + }, + { + "epoch": 0.15854263426072057, + "grad_norm": 1.7734146452788067, + "learning_rate": 3.829223177805959e-06, + "loss": 0.7182, + "step": 1758 + }, + { + "epoch": 0.15863281778419083, + "grad_norm": 1.738532161819992, + "learning_rate": 3.828986883796591e-06, + "loss": 0.8322, + "step": 1759 + }, + { + "epoch": 0.1587230013076611, + "grad_norm": 2.255975735128056, + "learning_rate": 3.828750433727979e-06, + "loss": 0.7874, + "step": 1760 + }, + { + "epoch": 0.15881318483113135, + "grad_norm": 1.7718762486905872, + "learning_rate": 3.828513827620296e-06, + "loss": 0.8437, + "step": 1761 + }, + { + "epoch": 0.15890336835460162, + "grad_norm": 2.001303248176575, + "learning_rate": 3.82827706549373e-06, + "loss": 0.8505, + "step": 1762 + }, + { + "epoch": 0.15899355187807188, + "grad_norm": 2.4985961787331465, + "learning_rate": 3.828040147368484e-06, + "loss": 0.7882, + "step": 1763 + }, + { + "epoch": 0.15908373540154214, + "grad_norm": 4.060957395674377, + "learning_rate": 3.827803073264774e-06, + "loss": 0.7918, + "step": 1764 + }, + { + "epoch": 0.1591739189250124, + "grad_norm": 1.8157766912390134, + "learning_rate": 3.827565843202826e-06, + "loss": 0.7927, + "step": 1765 + }, + { + "epoch": 0.15926410244848266, + "grad_norm": 2.070283939956308, + "learning_rate": 3.827328457202884e-06, + "loss": 0.6664, + "step": 1766 + }, + { + "epoch": 0.15935428597195292, + "grad_norm": 1.4993601069917235, + "learning_rate": 3.8270909152852014e-06, + "loss": 0.7257, + "step": 1767 + }, + { + "epoch": 0.15944446949542318, + "grad_norm": 2.024705879609624, + "learning_rate": 3.826853217470048e-06, + "loss": 0.7897, + "step": 1768 + }, + { + "epoch": 0.15953465301889344, + "grad_norm": 1.7037628799141948, + "learning_rate": 3.826615363777705e-06, + "loss": 0.8759, + "step": 1769 + }, + { + "epoch": 0.1596248365423637, + "grad_norm": 1.8805649508127773, + "learning_rate": 3.826377354228468e-06, + "loss": 0.7732, + "step": 1770 + }, + { + "epoch": 0.15971502006583396, + "grad_norm": 0.6481596799674829, + "learning_rate": 3.826139188842643e-06, + "loss": 0.5921, + "step": 1771 + }, + { + "epoch": 0.15980520358930422, + "grad_norm": 1.474433292208436, + "learning_rate": 3.825900867640554e-06, + "loss": 0.8473, + "step": 1772 + }, + { + "epoch": 0.15989538711277448, + "grad_norm": 1.4320446656842694, + "learning_rate": 3.825662390642535e-06, + "loss": 0.8312, + "step": 1773 + }, + { + "epoch": 0.15998557063624474, + "grad_norm": 1.6201145385941274, + "learning_rate": 3.825423757868934e-06, + "loss": 0.8974, + "step": 1774 + }, + { + "epoch": 0.16007575415971503, + "grad_norm": 2.1085834289501713, + "learning_rate": 3.825184969340114e-06, + "loss": 0.8238, + "step": 1775 + }, + { + "epoch": 0.1601659376831853, + "grad_norm": 2.0868410589491866, + "learning_rate": 3.824946025076447e-06, + "loss": 0.8184, + "step": 1776 + }, + { + "epoch": 0.16025612120665556, + "grad_norm": 1.8955091331984386, + "learning_rate": 3.824706925098323e-06, + "loss": 0.7438, + "step": 1777 + }, + { + "epoch": 0.16034630473012582, + "grad_norm": 2.3513414652919655, + "learning_rate": 3.824467669426143e-06, + "loss": 0.8646, + "step": 1778 + }, + { + "epoch": 0.16043648825359608, + "grad_norm": 1.5818040472702588, + "learning_rate": 3.824228258080321e-06, + "loss": 0.7678, + "step": 1779 + }, + { + "epoch": 0.16052667177706634, + "grad_norm": 1.7089169908590436, + "learning_rate": 3.823988691081285e-06, + "loss": 0.711, + "step": 1780 + }, + { + "epoch": 0.1606168553005366, + "grad_norm": 2.4618473369267337, + "learning_rate": 3.823748968449478e-06, + "loss": 0.8252, + "step": 1781 + }, + { + "epoch": 0.16070703882400686, + "grad_norm": 2.4408416339302903, + "learning_rate": 3.823509090205352e-06, + "loss": 0.6814, + "step": 1782 + }, + { + "epoch": 0.16079722234747712, + "grad_norm": 2.51752253456978, + "learning_rate": 3.823269056369376e-06, + "loss": 0.7506, + "step": 1783 + }, + { + "epoch": 0.16088740587094738, + "grad_norm": 1.8287774250427713, + "learning_rate": 3.8230288669620295e-06, + "loss": 0.8252, + "step": 1784 + }, + { + "epoch": 0.16097758939441764, + "grad_norm": 1.6823726748713945, + "learning_rate": 3.822788522003809e-06, + "loss": 0.8021, + "step": 1785 + }, + { + "epoch": 0.1610677729178879, + "grad_norm": 2.290850152871767, + "learning_rate": 3.822548021515221e-06, + "loss": 0.8725, + "step": 1786 + }, + { + "epoch": 0.16115795644135816, + "grad_norm": 1.705370411503829, + "learning_rate": 3.822307365516787e-06, + "loss": 0.7691, + "step": 1787 + }, + { + "epoch": 0.16124813996482842, + "grad_norm": 1.5950354086075689, + "learning_rate": 3.8220665540290395e-06, + "loss": 0.8146, + "step": 1788 + }, + { + "epoch": 0.16133832348829868, + "grad_norm": 1.5362416863585104, + "learning_rate": 3.8218255870725265e-06, + "loss": 0.8567, + "step": 1789 + }, + { + "epoch": 0.16142850701176895, + "grad_norm": 2.6467964241768116, + "learning_rate": 3.82158446466781e-06, + "loss": 0.7372, + "step": 1790 + }, + { + "epoch": 0.1615186905352392, + "grad_norm": 1.5630187127279989, + "learning_rate": 3.821343186835462e-06, + "loss": 0.8074, + "step": 1791 + }, + { + "epoch": 0.16160887405870947, + "grad_norm": 1.847700961259527, + "learning_rate": 3.821101753596072e-06, + "loss": 0.8624, + "step": 1792 + }, + { + "epoch": 0.16169905758217973, + "grad_norm": 3.4751655429406934, + "learning_rate": 3.820860164970237e-06, + "loss": 0.7871, + "step": 1793 + }, + { + "epoch": 0.16178924110565, + "grad_norm": 1.4451377144366324, + "learning_rate": 3.820618420978574e-06, + "loss": 0.8579, + "step": 1794 + }, + { + "epoch": 0.16187942462912025, + "grad_norm": 2.1057870373041694, + "learning_rate": 3.820376521641708e-06, + "loss": 0.7932, + "step": 1795 + }, + { + "epoch": 0.1619696081525905, + "grad_norm": 1.6130002864735697, + "learning_rate": 3.82013446698028e-06, + "loss": 0.8062, + "step": 1796 + }, + { + "epoch": 0.16205979167606077, + "grad_norm": 1.5543421836244926, + "learning_rate": 3.819892257014943e-06, + "loss": 0.7549, + "step": 1797 + }, + { + "epoch": 0.16214997519953103, + "grad_norm": 2.454058912812097, + "learning_rate": 3.819649891766364e-06, + "loss": 0.8242, + "step": 1798 + }, + { + "epoch": 0.16224015872300132, + "grad_norm": 1.5098856376136158, + "learning_rate": 3.819407371255222e-06, + "loss": 0.7866, + "step": 1799 + }, + { + "epoch": 0.16233034224647158, + "grad_norm": 5.144092630907935, + "learning_rate": 3.819164695502212e-06, + "loss": 0.8027, + "step": 1800 + }, + { + "epoch": 0.16242052576994184, + "grad_norm": 1.689651389423138, + "learning_rate": 3.818921864528039e-06, + "loss": 0.7352, + "step": 1801 + }, + { + "epoch": 0.1625107092934121, + "grad_norm": 4.6409515368728576, + "learning_rate": 3.818678878353423e-06, + "loss": 0.8477, + "step": 1802 + }, + { + "epoch": 0.16260089281688236, + "grad_norm": 3.4430378089178304, + "learning_rate": 3.818435736999097e-06, + "loss": 0.7401, + "step": 1803 + }, + { + "epoch": 0.16269107634035262, + "grad_norm": 3.2111413895595873, + "learning_rate": 3.818192440485807e-06, + "loss": 0.7838, + "step": 1804 + }, + { + "epoch": 0.16278125986382289, + "grad_norm": 1.9663994092977644, + "learning_rate": 3.817948988834314e-06, + "loss": 0.7313, + "step": 1805 + }, + { + "epoch": 0.16287144338729315, + "grad_norm": 1.9924532245616307, + "learning_rate": 3.817705382065388e-06, + "loss": 0.7928, + "step": 1806 + }, + { + "epoch": 0.1629616269107634, + "grad_norm": 1.3979761652026588, + "learning_rate": 3.8174616201998155e-06, + "loss": 0.7606, + "step": 1807 + }, + { + "epoch": 0.16305181043423367, + "grad_norm": 13.917321347734044, + "learning_rate": 3.817217703258397e-06, + "loss": 0.8457, + "step": 1808 + }, + { + "epoch": 0.16314199395770393, + "grad_norm": 1.665098915254889, + "learning_rate": 3.816973631261943e-06, + "loss": 0.7824, + "step": 1809 + }, + { + "epoch": 0.1632321774811742, + "grad_norm": 2.02553870488394, + "learning_rate": 3.816729404231281e-06, + "loss": 0.8277, + "step": 1810 + }, + { + "epoch": 0.16332236100464445, + "grad_norm": 2.191673548098979, + "learning_rate": 3.816485022187249e-06, + "loss": 0.8576, + "step": 1811 + }, + { + "epoch": 0.1634125445281147, + "grad_norm": 2.1390451880591113, + "learning_rate": 3.816240485150698e-06, + "loss": 0.7492, + "step": 1812 + }, + { + "epoch": 0.16350272805158497, + "grad_norm": 1.4383111406708666, + "learning_rate": 3.815995793142495e-06, + "loss": 0.7951, + "step": 1813 + }, + { + "epoch": 0.16359291157505523, + "grad_norm": 1.65932186925281, + "learning_rate": 3.815750946183518e-06, + "loss": 0.7337, + "step": 1814 + }, + { + "epoch": 0.1636830950985255, + "grad_norm": 2.2508280607875215, + "learning_rate": 3.815505944294658e-06, + "loss": 0.8463, + "step": 1815 + }, + { + "epoch": 0.16377327862199575, + "grad_norm": 1.9494085123961358, + "learning_rate": 3.81526078749682e-06, + "loss": 0.7522, + "step": 1816 + }, + { + "epoch": 0.16386346214546602, + "grad_norm": 3.434822565685054, + "learning_rate": 3.8150154758109225e-06, + "loss": 0.7326, + "step": 1817 + }, + { + "epoch": 0.16395364566893628, + "grad_norm": 1.5300836597583072, + "learning_rate": 3.814770009257896e-06, + "loss": 0.7634, + "step": 1818 + }, + { + "epoch": 0.16404382919240654, + "grad_norm": 1.9010055190181332, + "learning_rate": 3.814524387858687e-06, + "loss": 0.7997, + "step": 1819 + }, + { + "epoch": 0.1641340127158768, + "grad_norm": 1.719593257383982, + "learning_rate": 3.814278611634251e-06, + "loss": 0.8151, + "step": 1820 + }, + { + "epoch": 0.16422419623934706, + "grad_norm": 1.795281674059836, + "learning_rate": 3.8140326806055606e-06, + "loss": 0.7704, + "step": 1821 + }, + { + "epoch": 0.16431437976281732, + "grad_norm": 1.7340502735668664, + "learning_rate": 3.8137865947935992e-06, + "loss": 0.854, + "step": 1822 + }, + { + "epoch": 0.1644045632862876, + "grad_norm": 1.7547315257418301, + "learning_rate": 3.8135403542193646e-06, + "loss": 0.8157, + "step": 1823 + }, + { + "epoch": 0.16449474680975787, + "grad_norm": 2.2036201515826037, + "learning_rate": 3.813293958903867e-06, + "loss": 0.733, + "step": 1824 + }, + { + "epoch": 0.16458493033322813, + "grad_norm": 1.7717383222662058, + "learning_rate": 3.8130474088681306e-06, + "loss": 0.7645, + "step": 1825 + }, + { + "epoch": 0.1646751138566984, + "grad_norm": 2.0426866617435935, + "learning_rate": 3.8128007041331927e-06, + "loss": 0.7171, + "step": 1826 + }, + { + "epoch": 0.16476529738016865, + "grad_norm": 1.6387542231059418, + "learning_rate": 3.812553844720102e-06, + "loss": 0.7072, + "step": 1827 + }, + { + "epoch": 0.1648554809036389, + "grad_norm": 1.6798439130480516, + "learning_rate": 3.8123068306499236e-06, + "loss": 0.8082, + "step": 1828 + }, + { + "epoch": 0.16494566442710917, + "grad_norm": 1.9582989331193281, + "learning_rate": 3.812059661943733e-06, + "loss": 0.7892, + "step": 1829 + }, + { + "epoch": 0.16503584795057943, + "grad_norm": 5.641021799095109, + "learning_rate": 3.811812338622621e-06, + "loss": 0.7622, + "step": 1830 + }, + { + "epoch": 0.1651260314740497, + "grad_norm": 1.4913536258323594, + "learning_rate": 3.81156486070769e-06, + "loss": 0.7814, + "step": 1831 + }, + { + "epoch": 0.16521621499751996, + "grad_norm": 1.6580678212889532, + "learning_rate": 3.811317228220056e-06, + "loss": 0.8402, + "step": 1832 + }, + { + "epoch": 0.16530639852099022, + "grad_norm": 3.885574801229505, + "learning_rate": 3.811069441180849e-06, + "loss": 0.8393, + "step": 1833 + }, + { + "epoch": 0.16539658204446048, + "grad_norm": 1.8278389079233655, + "learning_rate": 3.8108214996112107e-06, + "loss": 0.7966, + "step": 1834 + }, + { + "epoch": 0.16548676556793074, + "grad_norm": 2.060255968431239, + "learning_rate": 3.810573403532297e-06, + "loss": 0.8277, + "step": 1835 + }, + { + "epoch": 0.165576949091401, + "grad_norm": 5.442969366746771, + "learning_rate": 3.8103251529652774e-06, + "loss": 0.7687, + "step": 1836 + }, + { + "epoch": 0.16566713261487126, + "grad_norm": 1.7881571382089814, + "learning_rate": 3.810076747931334e-06, + "loss": 0.7704, + "step": 1837 + }, + { + "epoch": 0.16575731613834152, + "grad_norm": 4.394679033973415, + "learning_rate": 3.809828188451662e-06, + "loss": 0.7538, + "step": 1838 + }, + { + "epoch": 0.16584749966181178, + "grad_norm": 1.9392636793410727, + "learning_rate": 3.809579474547469e-06, + "loss": 0.7884, + "step": 1839 + }, + { + "epoch": 0.16593768318528204, + "grad_norm": 1.646525281802965, + "learning_rate": 3.809330606239977e-06, + "loss": 0.8382, + "step": 1840 + }, + { + "epoch": 0.1660278667087523, + "grad_norm": 1.8707307531133177, + "learning_rate": 3.809081583550422e-06, + "loss": 0.8574, + "step": 1841 + }, + { + "epoch": 0.16611805023222256, + "grad_norm": 2.182486101348773, + "learning_rate": 3.808832406500051e-06, + "loss": 0.8763, + "step": 1842 + }, + { + "epoch": 0.16620823375569282, + "grad_norm": 2.172912788814353, + "learning_rate": 3.8085830751101253e-06, + "loss": 0.8671, + "step": 1843 + }, + { + "epoch": 0.16629841727916309, + "grad_norm": 1.6572521704525618, + "learning_rate": 3.808333589401919e-06, + "loss": 0.8267, + "step": 1844 + }, + { + "epoch": 0.16638860080263335, + "grad_norm": 1.7135508188237099, + "learning_rate": 3.8080839493967194e-06, + "loss": 0.7253, + "step": 1845 + }, + { + "epoch": 0.1664787843261036, + "grad_norm": 0.6466345252532504, + "learning_rate": 3.807834155115828e-06, + "loss": 0.5461, + "step": 1846 + }, + { + "epoch": 0.1665689678495739, + "grad_norm": 0.6520370173433466, + "learning_rate": 3.8075842065805584e-06, + "loss": 0.6501, + "step": 1847 + }, + { + "epoch": 0.16665915137304416, + "grad_norm": 3.033315367872533, + "learning_rate": 3.8073341038122374e-06, + "loss": 0.7821, + "step": 1848 + }, + { + "epoch": 0.16674933489651442, + "grad_norm": 1.9841197893526934, + "learning_rate": 3.8070838468322048e-06, + "loss": 0.8607, + "step": 1849 + }, + { + "epoch": 0.16683951841998468, + "grad_norm": 1.9374302882297154, + "learning_rate": 3.8068334356618143e-06, + "loss": 0.7296, + "step": 1850 + }, + { + "epoch": 0.16692970194345494, + "grad_norm": 3.374178468534884, + "learning_rate": 3.8065828703224324e-06, + "loss": 0.835, + "step": 1851 + }, + { + "epoch": 0.1670198854669252, + "grad_norm": 2.346306284177047, + "learning_rate": 3.8063321508354386e-06, + "loss": 0.7929, + "step": 1852 + }, + { + "epoch": 0.16711006899039546, + "grad_norm": 1.5509918607969226, + "learning_rate": 3.8060812772222255e-06, + "loss": 0.7773, + "step": 1853 + }, + { + "epoch": 0.16720025251386572, + "grad_norm": 1.4254160792262118, + "learning_rate": 3.8058302495041993e-06, + "loss": 0.8586, + "step": 1854 + }, + { + "epoch": 0.16729043603733598, + "grad_norm": 1.92684708942859, + "learning_rate": 3.805579067702779e-06, + "loss": 0.8079, + "step": 1855 + }, + { + "epoch": 0.16738061956080624, + "grad_norm": 1.7634965338417385, + "learning_rate": 3.8053277318393967e-06, + "loss": 0.7956, + "step": 1856 + }, + { + "epoch": 0.1674708030842765, + "grad_norm": 1.5521921038650548, + "learning_rate": 3.805076241935498e-06, + "loss": 0.8101, + "step": 1857 + }, + { + "epoch": 0.16756098660774676, + "grad_norm": 1.8914261058598676, + "learning_rate": 3.804824598012541e-06, + "loss": 0.6681, + "step": 1858 + }, + { + "epoch": 0.16765117013121703, + "grad_norm": 1.5163820086924042, + "learning_rate": 3.8045728000919975e-06, + "loss": 0.8338, + "step": 1859 + }, + { + "epoch": 0.1677413536546873, + "grad_norm": 1.3923050236787848, + "learning_rate": 3.8043208481953524e-06, + "loss": 0.826, + "step": 1860 + }, + { + "epoch": 0.16783153717815755, + "grad_norm": 15.23022729995841, + "learning_rate": 3.804068742344104e-06, + "loss": 0.7204, + "step": 1861 + }, + { + "epoch": 0.1679217207016278, + "grad_norm": 1.7907154825993357, + "learning_rate": 3.8038164825597628e-06, + "loss": 0.7613, + "step": 1862 + }, + { + "epoch": 0.16801190422509807, + "grad_norm": 2.6009888345844656, + "learning_rate": 3.8035640688638537e-06, + "loss": 0.8784, + "step": 1863 + }, + { + "epoch": 0.16810208774856833, + "grad_norm": 1.6776590563528673, + "learning_rate": 3.8033115012779125e-06, + "loss": 0.8677, + "step": 1864 + }, + { + "epoch": 0.1681922712720386, + "grad_norm": 1.4876050639514324, + "learning_rate": 3.8030587798234915e-06, + "loss": 0.8347, + "step": 1865 + }, + { + "epoch": 0.16828245479550885, + "grad_norm": 2.3833620532068265, + "learning_rate": 3.802805904522153e-06, + "loss": 0.7334, + "step": 1866 + }, + { + "epoch": 0.1683726383189791, + "grad_norm": 1.4407125318170577, + "learning_rate": 3.8025528753954742e-06, + "loss": 0.783, + "step": 1867 + }, + { + "epoch": 0.16846282184244937, + "grad_norm": 2.2002985795020504, + "learning_rate": 3.802299692465045e-06, + "loss": 0.7475, + "step": 1868 + }, + { + "epoch": 0.16855300536591963, + "grad_norm": 1.9018606085456864, + "learning_rate": 3.802046355752468e-06, + "loss": 0.7833, + "step": 1869 + }, + { + "epoch": 0.16864318888938992, + "grad_norm": 2.3211874805221284, + "learning_rate": 3.80179286527936e-06, + "loss": 0.8529, + "step": 1870 + }, + { + "epoch": 0.16873337241286018, + "grad_norm": 1.7204016638991881, + "learning_rate": 3.801539221067349e-06, + "loss": 0.7073, + "step": 1871 + }, + { + "epoch": 0.16882355593633044, + "grad_norm": 1.5100603494521094, + "learning_rate": 3.801285423138079e-06, + "loss": 0.746, + "step": 1872 + }, + { + "epoch": 0.1689137394598007, + "grad_norm": 3.4470851083667386, + "learning_rate": 3.8010314715132037e-06, + "loss": 0.7487, + "step": 1873 + }, + { + "epoch": 0.16900392298327097, + "grad_norm": 2.0085346986272357, + "learning_rate": 3.800777366214393e-06, + "loss": 0.7503, + "step": 1874 + }, + { + "epoch": 0.16909410650674123, + "grad_norm": 2.2079217515224463, + "learning_rate": 3.800523107263328e-06, + "loss": 0.7662, + "step": 1875 + }, + { + "epoch": 0.1691842900302115, + "grad_norm": 1.5849629682432693, + "learning_rate": 3.800268694681703e-06, + "loss": 0.8456, + "step": 1876 + }, + { + "epoch": 0.16927447355368175, + "grad_norm": 0.6560381592840218, + "learning_rate": 3.800014128491227e-06, + "loss": 0.5989, + "step": 1877 + }, + { + "epoch": 0.169364657077152, + "grad_norm": 1.5503503311238072, + "learning_rate": 3.79975940871362e-06, + "loss": 0.8468, + "step": 1878 + }, + { + "epoch": 0.16945484060062227, + "grad_norm": 3.051637653884463, + "learning_rate": 3.799504535370617e-06, + "loss": 0.8015, + "step": 1879 + }, + { + "epoch": 0.16954502412409253, + "grad_norm": 2.332875774570441, + "learning_rate": 3.799249508483964e-06, + "loss": 0.7304, + "step": 1880 + }, + { + "epoch": 0.1696352076475628, + "grad_norm": 0.7260923197374896, + "learning_rate": 3.798994328075422e-06, + "loss": 0.6595, + "step": 1881 + }, + { + "epoch": 0.16972539117103305, + "grad_norm": 1.7239796232846996, + "learning_rate": 3.798738994166765e-06, + "loss": 0.7213, + "step": 1882 + }, + { + "epoch": 0.1698155746945033, + "grad_norm": 1.728450795482702, + "learning_rate": 3.7984835067797788e-06, + "loss": 0.792, + "step": 1883 + }, + { + "epoch": 0.16990575821797357, + "grad_norm": 2.346599829906827, + "learning_rate": 3.798227865936263e-06, + "loss": 0.8836, + "step": 1884 + }, + { + "epoch": 0.16999594174144383, + "grad_norm": 2.0789369990136595, + "learning_rate": 3.7979720716580297e-06, + "loss": 0.6567, + "step": 1885 + }, + { + "epoch": 0.1700861252649141, + "grad_norm": 1.821842647415728, + "learning_rate": 3.7977161239669057e-06, + "loss": 0.7827, + "step": 1886 + }, + { + "epoch": 0.17017630878838436, + "grad_norm": 2.0390007623556, + "learning_rate": 3.7974600228847294e-06, + "loss": 0.7753, + "step": 1887 + }, + { + "epoch": 0.17026649231185462, + "grad_norm": 2.386409758250087, + "learning_rate": 3.7972037684333534e-06, + "loss": 0.7334, + "step": 1888 + }, + { + "epoch": 0.17035667583532488, + "grad_norm": 1.4859247249210552, + "learning_rate": 3.796947360634642e-06, + "loss": 0.719, + "step": 1889 + }, + { + "epoch": 0.17044685935879514, + "grad_norm": 2.1454915882551107, + "learning_rate": 3.796690799510473e-06, + "loss": 0.8479, + "step": 1890 + }, + { + "epoch": 0.1705370428822654, + "grad_norm": 1.5517491315787313, + "learning_rate": 3.7964340850827387e-06, + "loss": 0.7546, + "step": 1891 + }, + { + "epoch": 0.17062722640573566, + "grad_norm": 0.6797795945838627, + "learning_rate": 3.7961772173733425e-06, + "loss": 0.5677, + "step": 1892 + }, + { + "epoch": 0.17071740992920592, + "grad_norm": 1.4742709346201504, + "learning_rate": 3.7959201964042024e-06, + "loss": 0.7759, + "step": 1893 + }, + { + "epoch": 0.1708075934526762, + "grad_norm": 1.5886931133515512, + "learning_rate": 3.795663022197248e-06, + "loss": 0.8216, + "step": 1894 + }, + { + "epoch": 0.17089777697614647, + "grad_norm": 2.139682088750095, + "learning_rate": 3.7954056947744242e-06, + "loss": 0.7964, + "step": 1895 + }, + { + "epoch": 0.17098796049961673, + "grad_norm": 2.8399636349566126, + "learning_rate": 3.7951482141576863e-06, + "loss": 0.7645, + "step": 1896 + }, + { + "epoch": 0.171078144023087, + "grad_norm": 1.85486151144878, + "learning_rate": 3.794890580369004e-06, + "loss": 0.8434, + "step": 1897 + }, + { + "epoch": 0.17116832754655725, + "grad_norm": 1.6899856345564965, + "learning_rate": 3.7946327934303612e-06, + "loss": 0.8205, + "step": 1898 + }, + { + "epoch": 0.1712585110700275, + "grad_norm": 1.6609363832227768, + "learning_rate": 3.794374853363752e-06, + "loss": 0.8846, + "step": 1899 + }, + { + "epoch": 0.17134869459349777, + "grad_norm": 1.6858388000389588, + "learning_rate": 3.794116760191187e-06, + "loss": 0.812, + "step": 1900 + }, + { + "epoch": 0.17143887811696804, + "grad_norm": 1.7534208241935814, + "learning_rate": 3.7938585139346877e-06, + "loss": 0.8898, + "step": 1901 + }, + { + "epoch": 0.1715290616404383, + "grad_norm": 2.039127626511732, + "learning_rate": 3.793600114616288e-06, + "loss": 0.774, + "step": 1902 + }, + { + "epoch": 0.17161924516390856, + "grad_norm": 0.7403012046741297, + "learning_rate": 3.793341562258037e-06, + "loss": 0.5576, + "step": 1903 + }, + { + "epoch": 0.17170942868737882, + "grad_norm": 1.6213752961694101, + "learning_rate": 3.7930828568819953e-06, + "loss": 0.806, + "step": 1904 + }, + { + "epoch": 0.17179961221084908, + "grad_norm": 2.1084074839141267, + "learning_rate": 3.7928239985102378e-06, + "loss": 0.7867, + "step": 1905 + }, + { + "epoch": 0.17188979573431934, + "grad_norm": 3.241010410855906, + "learning_rate": 3.7925649871648505e-06, + "loss": 0.7919, + "step": 1906 + }, + { + "epoch": 0.1719799792577896, + "grad_norm": 1.628442858458586, + "learning_rate": 3.792305822867935e-06, + "loss": 0.7125, + "step": 1907 + }, + { + "epoch": 0.17207016278125986, + "grad_norm": 2.2633741049635994, + "learning_rate": 3.792046505641604e-06, + "loss": 0.7565, + "step": 1908 + }, + { + "epoch": 0.17216034630473012, + "grad_norm": 1.7019214402165639, + "learning_rate": 3.791787035507984e-06, + "loss": 0.748, + "step": 1909 + }, + { + "epoch": 0.17225052982820038, + "grad_norm": 3.175022611762801, + "learning_rate": 3.7915274124892136e-06, + "loss": 0.7498, + "step": 1910 + }, + { + "epoch": 0.17234071335167064, + "grad_norm": 4.334072612398862, + "learning_rate": 3.7912676366074466e-06, + "loss": 0.7723, + "step": 1911 + }, + { + "epoch": 0.1724308968751409, + "grad_norm": 2.48167253254884, + "learning_rate": 3.7910077078848478e-06, + "loss": 0.755, + "step": 1912 + }, + { + "epoch": 0.17252108039861117, + "grad_norm": 2.1293614452077314, + "learning_rate": 3.7907476263435957e-06, + "loss": 0.7541, + "step": 1913 + }, + { + "epoch": 0.17261126392208143, + "grad_norm": 3.5252356159654474, + "learning_rate": 3.7904873920058826e-06, + "loss": 0.759, + "step": 1914 + }, + { + "epoch": 0.1727014474455517, + "grad_norm": 1.994172188474495, + "learning_rate": 3.7902270048939114e-06, + "loss": 0.6678, + "step": 1915 + }, + { + "epoch": 0.17279163096902195, + "grad_norm": 1.9749616088033075, + "learning_rate": 3.7899664650299023e-06, + "loss": 0.7525, + "step": 1916 + }, + { + "epoch": 0.1728818144924922, + "grad_norm": 1.785737658756425, + "learning_rate": 3.7897057724360836e-06, + "loss": 0.8003, + "step": 1917 + }, + { + "epoch": 0.1729719980159625, + "grad_norm": 2.852790296461683, + "learning_rate": 3.7894449271347004e-06, + "loss": 0.8113, + "step": 1918 + }, + { + "epoch": 0.17306218153943276, + "grad_norm": 2.1416871714321633, + "learning_rate": 3.789183929148009e-06, + "loss": 0.849, + "step": 1919 + }, + { + "epoch": 0.17315236506290302, + "grad_norm": 4.396810606787935, + "learning_rate": 3.7889227784982795e-06, + "loss": 0.8173, + "step": 1920 + }, + { + "epoch": 0.17324254858637328, + "grad_norm": 1.6428598350597619, + "learning_rate": 3.7886614752077945e-06, + "loss": 0.7837, + "step": 1921 + }, + { + "epoch": 0.17333273210984354, + "grad_norm": 2.0110841213028716, + "learning_rate": 3.7884000192988495e-06, + "loss": 0.7929, + "step": 1922 + }, + { + "epoch": 0.1734229156333138, + "grad_norm": 2.405335388748717, + "learning_rate": 3.7881384107937546e-06, + "loss": 0.8448, + "step": 1923 + }, + { + "epoch": 0.17351309915678406, + "grad_norm": 1.7997824987822761, + "learning_rate": 3.78787664971483e-06, + "loss": 0.7821, + "step": 1924 + }, + { + "epoch": 0.17360328268025432, + "grad_norm": 1.499352553828891, + "learning_rate": 3.7876147360844115e-06, + "loss": 0.7567, + "step": 1925 + }, + { + "epoch": 0.17369346620372458, + "grad_norm": 2.228698241552273, + "learning_rate": 3.7873526699248474e-06, + "loss": 0.8581, + "step": 1926 + }, + { + "epoch": 0.17378364972719484, + "grad_norm": 1.4406388884100403, + "learning_rate": 3.7870904512584974e-06, + "loss": 0.7947, + "step": 1927 + }, + { + "epoch": 0.1738738332506651, + "grad_norm": 1.8298498640332064, + "learning_rate": 3.7868280801077368e-06, + "loss": 0.79, + "step": 1928 + }, + { + "epoch": 0.17396401677413537, + "grad_norm": 1.495615511719464, + "learning_rate": 3.7865655564949517e-06, + "loss": 0.7795, + "step": 1929 + }, + { + "epoch": 0.17405420029760563, + "grad_norm": 1.4500885278708744, + "learning_rate": 3.786302880442542e-06, + "loss": 0.7592, + "step": 1930 + }, + { + "epoch": 0.1741443838210759, + "grad_norm": 1.8617578832921229, + "learning_rate": 3.7860400519729215e-06, + "loss": 0.7858, + "step": 1931 + }, + { + "epoch": 0.17423456734454615, + "grad_norm": 1.9105235489704528, + "learning_rate": 3.7857770711085157e-06, + "loss": 0.8215, + "step": 1932 + }, + { + "epoch": 0.1743247508680164, + "grad_norm": 2.8314201590160235, + "learning_rate": 3.785513937871763e-06, + "loss": 0.8352, + "step": 1933 + }, + { + "epoch": 0.17441493439148667, + "grad_norm": 1.7787904992388774, + "learning_rate": 3.785250652285116e-06, + "loss": 0.7613, + "step": 1934 + }, + { + "epoch": 0.17450511791495693, + "grad_norm": 1.6435171158157278, + "learning_rate": 3.78498721437104e-06, + "loss": 0.863, + "step": 1935 + }, + { + "epoch": 0.1745953014384272, + "grad_norm": 2.123320308367885, + "learning_rate": 3.784723624152012e-06, + "loss": 0.7947, + "step": 1936 + }, + { + "epoch": 0.17468548496189745, + "grad_norm": 3.7299726385999925, + "learning_rate": 3.784459881650524e-06, + "loss": 0.7404, + "step": 1937 + }, + { + "epoch": 0.1747756684853677, + "grad_norm": 1.6603442276390008, + "learning_rate": 3.784195986889079e-06, + "loss": 0.7604, + "step": 1938 + }, + { + "epoch": 0.17486585200883797, + "grad_norm": 1.7346545157426034, + "learning_rate": 3.7839319398901946e-06, + "loss": 0.7157, + "step": 1939 + }, + { + "epoch": 0.17495603553230824, + "grad_norm": 5.843414460203121, + "learning_rate": 3.7836677406764013e-06, + "loss": 0.8578, + "step": 1940 + }, + { + "epoch": 0.1750462190557785, + "grad_norm": 1.4921570620003821, + "learning_rate": 3.7834033892702407e-06, + "loss": 0.8362, + "step": 1941 + }, + { + "epoch": 0.17513640257924878, + "grad_norm": 1.697029334368308, + "learning_rate": 3.783138885694269e-06, + "loss": 0.7574, + "step": 1942 + }, + { + "epoch": 0.17522658610271905, + "grad_norm": 1.8337959948429117, + "learning_rate": 3.7828742299710558e-06, + "loss": 0.8701, + "step": 1943 + }, + { + "epoch": 0.1753167696261893, + "grad_norm": 1.6958554009478732, + "learning_rate": 3.782609422123183e-06, + "loss": 0.7732, + "step": 1944 + }, + { + "epoch": 0.17540695314965957, + "grad_norm": 1.8959093253307497, + "learning_rate": 3.7823444621732444e-06, + "loss": 0.7953, + "step": 1945 + }, + { + "epoch": 0.17549713667312983, + "grad_norm": 2.2108540333875055, + "learning_rate": 3.782079350143849e-06, + "loss": 0.8065, + "step": 1946 + }, + { + "epoch": 0.1755873201966001, + "grad_norm": 2.5404455555587204, + "learning_rate": 3.781814086057617e-06, + "loss": 0.889, + "step": 1947 + }, + { + "epoch": 0.17567750372007035, + "grad_norm": 3.926536798601781, + "learning_rate": 3.7815486699371826e-06, + "loss": 0.7767, + "step": 1948 + }, + { + "epoch": 0.1757676872435406, + "grad_norm": 1.5548665312619125, + "learning_rate": 3.7812831018051918e-06, + "loss": 0.7635, + "step": 1949 + }, + { + "epoch": 0.17585787076701087, + "grad_norm": 1.5018497823375263, + "learning_rate": 3.7810173816843058e-06, + "loss": 0.7987, + "step": 1950 + }, + { + "epoch": 0.17594805429048113, + "grad_norm": 1.5707540318365838, + "learning_rate": 3.7807515095971955e-06, + "loss": 0.8627, + "step": 1951 + }, + { + "epoch": 0.1760382378139514, + "grad_norm": 1.5084198830292799, + "learning_rate": 3.7804854855665475e-06, + "loss": 0.783, + "step": 1952 + }, + { + "epoch": 0.17612842133742165, + "grad_norm": 3.1161587578383636, + "learning_rate": 3.7802193096150606e-06, + "loss": 0.7988, + "step": 1953 + }, + { + "epoch": 0.17621860486089191, + "grad_norm": 1.933345216133607, + "learning_rate": 3.779952981765446e-06, + "loss": 0.7475, + "step": 1954 + }, + { + "epoch": 0.17630878838436218, + "grad_norm": 1.4438968633291516, + "learning_rate": 3.779686502040429e-06, + "loss": 0.806, + "step": 1955 + }, + { + "epoch": 0.17639897190783244, + "grad_norm": 0.7003870532598444, + "learning_rate": 3.779419870462746e-06, + "loss": 0.5762, + "step": 1956 + }, + { + "epoch": 0.1764891554313027, + "grad_norm": 2.0088694124745223, + "learning_rate": 3.779153087055148e-06, + "loss": 0.7702, + "step": 1957 + }, + { + "epoch": 0.17657933895477296, + "grad_norm": 2.062091324516388, + "learning_rate": 3.7788861518403988e-06, + "loss": 0.7999, + "step": 1958 + }, + { + "epoch": 0.17666952247824322, + "grad_norm": 1.6624198291923196, + "learning_rate": 3.7786190648412742e-06, + "loss": 0.6825, + "step": 1959 + }, + { + "epoch": 0.17675970600171348, + "grad_norm": 0.6475048021558217, + "learning_rate": 3.778351826080564e-06, + "loss": 0.5871, + "step": 1960 + }, + { + "epoch": 0.17684988952518374, + "grad_norm": 1.5335005745005146, + "learning_rate": 3.7780844355810704e-06, + "loss": 0.8926, + "step": 1961 + }, + { + "epoch": 0.176940073048654, + "grad_norm": 1.5129242262690366, + "learning_rate": 3.777816893365608e-06, + "loss": 0.8012, + "step": 1962 + }, + { + "epoch": 0.17703025657212426, + "grad_norm": 1.5400705666908816, + "learning_rate": 3.7775491994570057e-06, + "loss": 0.8644, + "step": 1963 + }, + { + "epoch": 0.17712044009559452, + "grad_norm": 1.6617502974442064, + "learning_rate": 3.777281353878105e-06, + "loss": 0.7624, + "step": 1964 + }, + { + "epoch": 0.17721062361906478, + "grad_norm": 0.6201857401284572, + "learning_rate": 3.777013356651758e-06, + "loss": 0.594, + "step": 1965 + }, + { + "epoch": 0.17730080714253507, + "grad_norm": 1.9058160757004317, + "learning_rate": 3.776745207800834e-06, + "loss": 0.8053, + "step": 1966 + }, + { + "epoch": 0.17739099066600533, + "grad_norm": 2.0542461661017173, + "learning_rate": 3.7764769073482122e-06, + "loss": 0.8181, + "step": 1967 + }, + { + "epoch": 0.1774811741894756, + "grad_norm": 1.5307643665248234, + "learning_rate": 3.7762084553167846e-06, + "loss": 0.8571, + "step": 1968 + }, + { + "epoch": 0.17757135771294585, + "grad_norm": 2.0830254390809535, + "learning_rate": 3.775939851729458e-06, + "loss": 0.8799, + "step": 1969 + }, + { + "epoch": 0.17766154123641612, + "grad_norm": 4.031717199413893, + "learning_rate": 3.775671096609151e-06, + "loss": 0.7484, + "step": 1970 + }, + { + "epoch": 0.17775172475988638, + "grad_norm": 1.716536744148651, + "learning_rate": 3.775402189978795e-06, + "loss": 0.7741, + "step": 1971 + }, + { + "epoch": 0.17784190828335664, + "grad_norm": 1.3417337061242907, + "learning_rate": 3.7751331318613343e-06, + "loss": 0.7717, + "step": 1972 + }, + { + "epoch": 0.1779320918068269, + "grad_norm": 2.230659971040541, + "learning_rate": 3.774863922279727e-06, + "loss": 0.8114, + "step": 1973 + }, + { + "epoch": 0.17802227533029716, + "grad_norm": 1.4662095683345395, + "learning_rate": 3.7745945612569435e-06, + "loss": 0.8474, + "step": 1974 + }, + { + "epoch": 0.17811245885376742, + "grad_norm": 1.4991641894933745, + "learning_rate": 3.7743250488159674e-06, + "loss": 0.6698, + "step": 1975 + }, + { + "epoch": 0.17820264237723768, + "grad_norm": 1.7583309511045013, + "learning_rate": 3.774055384979794e-06, + "loss": 0.6951, + "step": 1976 + }, + { + "epoch": 0.17829282590070794, + "grad_norm": 1.6091744010867166, + "learning_rate": 3.773785569771433e-06, + "loss": 0.7727, + "step": 1977 + }, + { + "epoch": 0.1783830094241782, + "grad_norm": 1.670161215601429, + "learning_rate": 3.7735156032139066e-06, + "loss": 0.7808, + "step": 1978 + }, + { + "epoch": 0.17847319294764846, + "grad_norm": 1.6323006297605684, + "learning_rate": 3.773245485330251e-06, + "loss": 0.8937, + "step": 1979 + }, + { + "epoch": 0.17856337647111872, + "grad_norm": 2.144810170307968, + "learning_rate": 3.7729752161435115e-06, + "loss": 0.7595, + "step": 1980 + }, + { + "epoch": 0.17865355999458898, + "grad_norm": 1.739805431997465, + "learning_rate": 3.7727047956767514e-06, + "loss": 0.8573, + "step": 1981 + }, + { + "epoch": 0.17874374351805924, + "grad_norm": 1.4871701876664083, + "learning_rate": 3.7724342239530436e-06, + "loss": 0.9164, + "step": 1982 + }, + { + "epoch": 0.1788339270415295, + "grad_norm": 2.1016986334858037, + "learning_rate": 3.772163500995474e-06, + "loss": 0.7408, + "step": 1983 + }, + { + "epoch": 0.17892411056499977, + "grad_norm": 2.0319578844647053, + "learning_rate": 3.7718926268271437e-06, + "loss": 0.8883, + "step": 1984 + }, + { + "epoch": 0.17901429408847003, + "grad_norm": 1.8494292255069675, + "learning_rate": 3.771621601471164e-06, + "loss": 0.8156, + "step": 1985 + }, + { + "epoch": 0.1791044776119403, + "grad_norm": 2.9965469833196656, + "learning_rate": 3.771350424950661e-06, + "loss": 0.8376, + "step": 1986 + }, + { + "epoch": 0.17919466113541055, + "grad_norm": 1.713701219483555, + "learning_rate": 3.771079097288772e-06, + "loss": 0.7394, + "step": 1987 + }, + { + "epoch": 0.1792848446588808, + "grad_norm": 1.751238452837911, + "learning_rate": 3.770807618508649e-06, + "loss": 0.8065, + "step": 1988 + }, + { + "epoch": 0.17937502818235107, + "grad_norm": 1.703230268392377, + "learning_rate": 3.7705359886334555e-06, + "loss": 0.7802, + "step": 1989 + }, + { + "epoch": 0.17946521170582136, + "grad_norm": 1.8486484178306601, + "learning_rate": 3.7702642076863694e-06, + "loss": 0.772, + "step": 1990 + }, + { + "epoch": 0.17955539522929162, + "grad_norm": 1.5442447842421414, + "learning_rate": 3.7699922756905795e-06, + "loss": 0.8482, + "step": 1991 + }, + { + "epoch": 0.17964557875276188, + "grad_norm": 2.028270707972517, + "learning_rate": 3.7697201926692895e-06, + "loss": 0.7474, + "step": 1992 + }, + { + "epoch": 0.17973576227623214, + "grad_norm": 1.8279310596329261, + "learning_rate": 3.7694479586457144e-06, + "loss": 0.8032, + "step": 1993 + }, + { + "epoch": 0.1798259457997024, + "grad_norm": 2.9665585108840404, + "learning_rate": 3.7691755736430827e-06, + "loss": 0.8058, + "step": 1994 + }, + { + "epoch": 0.17991612932317266, + "grad_norm": 1.5894371468787856, + "learning_rate": 3.768903037684636e-06, + "loss": 0.8322, + "step": 1995 + }, + { + "epoch": 0.18000631284664292, + "grad_norm": 1.6456826861362084, + "learning_rate": 3.7686303507936284e-06, + "loss": 0.8363, + "step": 1996 + }, + { + "epoch": 0.18009649637011318, + "grad_norm": 2.6556930855749328, + "learning_rate": 3.7683575129933272e-06, + "loss": 0.8344, + "step": 1997 + }, + { + "epoch": 0.18018667989358345, + "grad_norm": 1.5508840992835025, + "learning_rate": 3.7680845243070128e-06, + "loss": 0.7987, + "step": 1998 + }, + { + "epoch": 0.1802768634170537, + "grad_norm": 0.7069361290778869, + "learning_rate": 3.7678113847579767e-06, + "loss": 0.5541, + "step": 1999 + }, + { + "epoch": 0.18036704694052397, + "grad_norm": 2.001604985926004, + "learning_rate": 3.7675380943695264e-06, + "loss": 0.8623, + "step": 2000 + }, + { + "epoch": 0.18045723046399423, + "grad_norm": 1.655653216250084, + "learning_rate": 3.7672646531649795e-06, + "loss": 0.7569, + "step": 2001 + }, + { + "epoch": 0.1805474139874645, + "grad_norm": 0.5906039511749632, + "learning_rate": 3.7669910611676682e-06, + "loss": 0.5771, + "step": 2002 + }, + { + "epoch": 0.18063759751093475, + "grad_norm": 2.123099093948709, + "learning_rate": 3.7667173184009356e-06, + "loss": 0.78, + "step": 2003 + }, + { + "epoch": 0.180727781034405, + "grad_norm": 3.6720757612171027, + "learning_rate": 3.7664434248881403e-06, + "loss": 0.853, + "step": 2004 + }, + { + "epoch": 0.18081796455787527, + "grad_norm": 2.28540705421209, + "learning_rate": 3.766169380652652e-06, + "loss": 0.8053, + "step": 2005 + }, + { + "epoch": 0.18090814808134553, + "grad_norm": 1.8132679726354093, + "learning_rate": 3.7658951857178537e-06, + "loss": 0.8355, + "step": 2006 + }, + { + "epoch": 0.1809983316048158, + "grad_norm": 1.6232716832735965, + "learning_rate": 3.7656208401071414e-06, + "loss": 0.7999, + "step": 2007 + }, + { + "epoch": 0.18108851512828605, + "grad_norm": 2.4131552957580897, + "learning_rate": 3.7653463438439225e-06, + "loss": 0.7355, + "step": 2008 + }, + { + "epoch": 0.18117869865175631, + "grad_norm": 1.8206674778150895, + "learning_rate": 3.7650716969516203e-06, + "loss": 0.8091, + "step": 2009 + }, + { + "epoch": 0.18126888217522658, + "grad_norm": 1.8482250281759527, + "learning_rate": 3.764796899453668e-06, + "loss": 0.697, + "step": 2010 + }, + { + "epoch": 0.18135906569869684, + "grad_norm": 1.6638508693460432, + "learning_rate": 3.7645219513735134e-06, + "loss": 0.8321, + "step": 2011 + }, + { + "epoch": 0.1814492492221671, + "grad_norm": 1.6518410353068063, + "learning_rate": 3.764246852734617e-06, + "loss": 0.8372, + "step": 2012 + }, + { + "epoch": 0.18153943274563739, + "grad_norm": 1.8496208472535582, + "learning_rate": 3.7639716035604502e-06, + "loss": 0.7665, + "step": 2013 + }, + { + "epoch": 0.18162961626910765, + "grad_norm": 0.6421790695565218, + "learning_rate": 3.7636962038745e-06, + "loss": 0.5487, + "step": 2014 + }, + { + "epoch": 0.1817197997925779, + "grad_norm": 1.893253296528244, + "learning_rate": 3.763420653700265e-06, + "loss": 0.7499, + "step": 2015 + }, + { + "epoch": 0.18180998331604817, + "grad_norm": 1.5455904935262086, + "learning_rate": 3.7631449530612565e-06, + "loss": 0.7524, + "step": 2016 + }, + { + "epoch": 0.18190016683951843, + "grad_norm": 1.526149897811352, + "learning_rate": 3.762869101980999e-06, + "loss": 0.8812, + "step": 2017 + }, + { + "epoch": 0.1819903503629887, + "grad_norm": 1.899810294669234, + "learning_rate": 3.7625931004830287e-06, + "loss": 0.7986, + "step": 2018 + }, + { + "epoch": 0.18208053388645895, + "grad_norm": 1.6338272227079274, + "learning_rate": 3.7623169485908966e-06, + "loss": 0.7498, + "step": 2019 + }, + { + "epoch": 0.1821707174099292, + "grad_norm": 0.7217517037393025, + "learning_rate": 3.7620406463281647e-06, + "loss": 0.5952, + "step": 2020 + }, + { + "epoch": 0.18226090093339947, + "grad_norm": 4.851952087190936, + "learning_rate": 3.7617641937184095e-06, + "loss": 0.8129, + "step": 2021 + }, + { + "epoch": 0.18235108445686973, + "grad_norm": 1.96781311752409, + "learning_rate": 3.761487590785219e-06, + "loss": 0.826, + "step": 2022 + }, + { + "epoch": 0.18244126798034, + "grad_norm": 1.6193108371403786, + "learning_rate": 3.7612108375521942e-06, + "loss": 0.7695, + "step": 2023 + }, + { + "epoch": 0.18253145150381025, + "grad_norm": 1.717492493192434, + "learning_rate": 3.76093393404295e-06, + "loss": 0.842, + "step": 2024 + }, + { + "epoch": 0.18262163502728052, + "grad_norm": 1.855026867098717, + "learning_rate": 3.7606568802811126e-06, + "loss": 0.7866, + "step": 2025 + }, + { + "epoch": 0.18271181855075078, + "grad_norm": 2.062212028051169, + "learning_rate": 3.760379676290322e-06, + "loss": 0.804, + "step": 2026 + }, + { + "epoch": 0.18280200207422104, + "grad_norm": 1.6250115907695852, + "learning_rate": 3.760102322094231e-06, + "loss": 0.7296, + "step": 2027 + }, + { + "epoch": 0.1828921855976913, + "grad_norm": 2.1908800441927423, + "learning_rate": 3.759824817716504e-06, + "loss": 0.7588, + "step": 2028 + }, + { + "epoch": 0.18298236912116156, + "grad_norm": 1.3245003099912782, + "learning_rate": 3.759547163180821e-06, + "loss": 0.8428, + "step": 2029 + }, + { + "epoch": 0.18307255264463182, + "grad_norm": 1.8776375339486646, + "learning_rate": 3.759269358510871e-06, + "loss": 0.5932, + "step": 2030 + }, + { + "epoch": 0.18316273616810208, + "grad_norm": 1.8907915152370467, + "learning_rate": 3.75899140373036e-06, + "loss": 0.6979, + "step": 2031 + }, + { + "epoch": 0.18325291969157234, + "grad_norm": 2.3967983347979924, + "learning_rate": 3.7587132988630028e-06, + "loss": 0.8574, + "step": 2032 + }, + { + "epoch": 0.1833431032150426, + "grad_norm": 2.1773492994398578, + "learning_rate": 3.7584350439325295e-06, + "loss": 0.8284, + "step": 2033 + }, + { + "epoch": 0.18343328673851286, + "grad_norm": 1.8884922324806879, + "learning_rate": 3.758156638962682e-06, + "loss": 0.7832, + "step": 2034 + }, + { + "epoch": 0.18352347026198312, + "grad_norm": 1.8230079700674497, + "learning_rate": 3.757878083977216e-06, + "loss": 0.7531, + "step": 2035 + }, + { + "epoch": 0.18361365378545338, + "grad_norm": 2.14173726609925, + "learning_rate": 3.7575993789999e-06, + "loss": 0.7799, + "step": 2036 + }, + { + "epoch": 0.18370383730892367, + "grad_norm": 4.8285178607990735, + "learning_rate": 3.757320524054512e-06, + "loss": 0.8271, + "step": 2037 + }, + { + "epoch": 0.18379402083239393, + "grad_norm": 1.6169875016681938, + "learning_rate": 3.757041519164848e-06, + "loss": 0.727, + "step": 2038 + }, + { + "epoch": 0.1838842043558642, + "grad_norm": 1.8443753749804372, + "learning_rate": 3.7567623643547133e-06, + "loss": 0.7975, + "step": 2039 + }, + { + "epoch": 0.18397438787933446, + "grad_norm": 1.7807078289290688, + "learning_rate": 3.756483059647927e-06, + "loss": 0.7953, + "step": 2040 + }, + { + "epoch": 0.18406457140280472, + "grad_norm": 1.455873740380981, + "learning_rate": 3.756203605068321e-06, + "loss": 0.7614, + "step": 2041 + }, + { + "epoch": 0.18415475492627498, + "grad_norm": 1.8253851027010277, + "learning_rate": 3.7559240006397396e-06, + "loss": 0.7845, + "step": 2042 + }, + { + "epoch": 0.18424493844974524, + "grad_norm": 1.6830735254756348, + "learning_rate": 3.7556442463860406e-06, + "loss": 0.9088, + "step": 2043 + }, + { + "epoch": 0.1843351219732155, + "grad_norm": 2.030948675253733, + "learning_rate": 3.7553643423310934e-06, + "loss": 0.8991, + "step": 2044 + }, + { + "epoch": 0.18442530549668576, + "grad_norm": 3.6282041135273113, + "learning_rate": 3.755084288498782e-06, + "loss": 0.6492, + "step": 2045 + }, + { + "epoch": 0.18451548902015602, + "grad_norm": 1.8068485284724727, + "learning_rate": 3.754804084913002e-06, + "loss": 0.8011, + "step": 2046 + }, + { + "epoch": 0.18460567254362628, + "grad_norm": 1.5689528163187683, + "learning_rate": 3.754523731597661e-06, + "loss": 0.8575, + "step": 2047 + }, + { + "epoch": 0.18469585606709654, + "grad_norm": 1.855705744013546, + "learning_rate": 3.754243228576681e-06, + "loss": 0.6928, + "step": 2048 + }, + { + "epoch": 0.1847860395905668, + "grad_norm": 2.1075990335493757, + "learning_rate": 3.753962575873996e-06, + "loss": 0.851, + "step": 2049 + }, + { + "epoch": 0.18487622311403706, + "grad_norm": 1.9587231782343917, + "learning_rate": 3.7536817735135527e-06, + "loss": 0.7504, + "step": 2050 + }, + { + "epoch": 0.18496640663750732, + "grad_norm": 1.4657281672559463, + "learning_rate": 3.753400821519311e-06, + "loss": 0.8352, + "step": 2051 + }, + { + "epoch": 0.18505659016097759, + "grad_norm": 1.5950724780796477, + "learning_rate": 3.7531197199152426e-06, + "loss": 0.7266, + "step": 2052 + }, + { + "epoch": 0.18514677368444785, + "grad_norm": 0.7116055546175485, + "learning_rate": 3.7528384687253335e-06, + "loss": 0.6184, + "step": 2053 + }, + { + "epoch": 0.1852369572079181, + "grad_norm": 1.720258709502268, + "learning_rate": 3.7525570679735815e-06, + "loss": 0.9038, + "step": 2054 + }, + { + "epoch": 0.18532714073138837, + "grad_norm": 1.6879521929673862, + "learning_rate": 3.7522755176839965e-06, + "loss": 0.8073, + "step": 2055 + }, + { + "epoch": 0.18541732425485863, + "grad_norm": 2.643609521619146, + "learning_rate": 3.7519938178806027e-06, + "loss": 0.8039, + "step": 2056 + }, + { + "epoch": 0.1855075077783289, + "grad_norm": 2.5528757746439914, + "learning_rate": 3.7517119685874358e-06, + "loss": 0.7511, + "step": 2057 + }, + { + "epoch": 0.18559769130179915, + "grad_norm": 1.5083368210023351, + "learning_rate": 3.7514299698285447e-06, + "loss": 0.7826, + "step": 2058 + }, + { + "epoch": 0.1856878748252694, + "grad_norm": 1.5911344120928914, + "learning_rate": 3.751147821627991e-06, + "loss": 0.7719, + "step": 2059 + }, + { + "epoch": 0.18577805834873967, + "grad_norm": 2.0948629125932308, + "learning_rate": 3.75086552400985e-06, + "loss": 0.8102, + "step": 2060 + }, + { + "epoch": 0.18586824187220996, + "grad_norm": 1.9203938304790102, + "learning_rate": 3.750583076998208e-06, + "loss": 0.7689, + "step": 2061 + }, + { + "epoch": 0.18595842539568022, + "grad_norm": 1.6122433923984298, + "learning_rate": 3.7503004806171655e-06, + "loss": 0.7686, + "step": 2062 + }, + { + "epoch": 0.18604860891915048, + "grad_norm": 2.3232638408971473, + "learning_rate": 3.7500177348908354e-06, + "loss": 0.7251, + "step": 2063 + }, + { + "epoch": 0.18613879244262074, + "grad_norm": 1.4404532737740547, + "learning_rate": 3.749734839843342e-06, + "loss": 0.8153, + "step": 2064 + }, + { + "epoch": 0.186228975966091, + "grad_norm": 2.53325846544698, + "learning_rate": 3.7494517954988245e-06, + "loss": 0.7944, + "step": 2065 + }, + { + "epoch": 0.18631915948956126, + "grad_norm": 1.7823643461294651, + "learning_rate": 3.749168601881433e-06, + "loss": 0.7894, + "step": 2066 + }, + { + "epoch": 0.18640934301303153, + "grad_norm": 2.2748315476313574, + "learning_rate": 3.7488852590153315e-06, + "loss": 0.7762, + "step": 2067 + }, + { + "epoch": 0.1864995265365018, + "grad_norm": 1.5369882305640254, + "learning_rate": 3.748601766924697e-06, + "loss": 0.7844, + "step": 2068 + }, + { + "epoch": 0.18658971005997205, + "grad_norm": 1.8036269998150034, + "learning_rate": 3.7483181256337176e-06, + "loss": 0.757, + "step": 2069 + }, + { + "epoch": 0.1866798935834423, + "grad_norm": 1.6834301791643567, + "learning_rate": 3.7480343351665962e-06, + "loss": 0.7937, + "step": 2070 + }, + { + "epoch": 0.18677007710691257, + "grad_norm": 1.6744898375017407, + "learning_rate": 3.747750395547546e-06, + "loss": 0.7781, + "step": 2071 + }, + { + "epoch": 0.18686026063038283, + "grad_norm": 1.3069130195914294, + "learning_rate": 3.7474663068007956e-06, + "loss": 0.7753, + "step": 2072 + }, + { + "epoch": 0.1869504441538531, + "grad_norm": 1.2183199881950386, + "learning_rate": 3.747182068950584e-06, + "loss": 0.7707, + "step": 2073 + }, + { + "epoch": 0.18704062767732335, + "grad_norm": 3.0129230629574297, + "learning_rate": 3.7468976820211643e-06, + "loss": 0.7916, + "step": 2074 + }, + { + "epoch": 0.1871308112007936, + "grad_norm": 2.448171583389344, + "learning_rate": 3.746613146036803e-06, + "loss": 0.7914, + "step": 2075 + }, + { + "epoch": 0.18722099472426387, + "grad_norm": 1.6152299468084705, + "learning_rate": 3.7463284610217766e-06, + "loss": 0.8353, + "step": 2076 + }, + { + "epoch": 0.18731117824773413, + "grad_norm": 1.8276677008002138, + "learning_rate": 3.746043627000377e-06, + "loss": 0.729, + "step": 2077 + }, + { + "epoch": 0.1874013617712044, + "grad_norm": 1.6450071393817547, + "learning_rate": 3.7457586439969076e-06, + "loss": 0.7394, + "step": 2078 + }, + { + "epoch": 0.18749154529467466, + "grad_norm": 1.3342192011697402, + "learning_rate": 3.7454735120356842e-06, + "loss": 0.7809, + "step": 2079 + }, + { + "epoch": 0.18758172881814492, + "grad_norm": 1.7373919666103241, + "learning_rate": 3.7451882311410373e-06, + "loss": 0.8605, + "step": 2080 + }, + { + "epoch": 0.18767191234161518, + "grad_norm": 1.678889382172456, + "learning_rate": 3.7449028013373074e-06, + "loss": 0.8269, + "step": 2081 + }, + { + "epoch": 0.18776209586508544, + "grad_norm": 2.036206229004654, + "learning_rate": 3.7446172226488485e-06, + "loss": 0.7872, + "step": 2082 + }, + { + "epoch": 0.1878522793885557, + "grad_norm": 2.7436625542750255, + "learning_rate": 3.7443314951000285e-06, + "loss": 0.8535, + "step": 2083 + }, + { + "epoch": 0.18794246291202596, + "grad_norm": 1.7788481999819719, + "learning_rate": 3.7440456187152276e-06, + "loss": 0.782, + "step": 2084 + }, + { + "epoch": 0.18803264643549625, + "grad_norm": 1.9668843951635766, + "learning_rate": 3.7437595935188377e-06, + "loss": 0.8767, + "step": 2085 + }, + { + "epoch": 0.1881228299589665, + "grad_norm": 1.990119189522082, + "learning_rate": 3.7434734195352647e-06, + "loss": 0.7906, + "step": 2086 + }, + { + "epoch": 0.18821301348243677, + "grad_norm": 0.6639479370029406, + "learning_rate": 3.743187096788926e-06, + "loss": 0.5809, + "step": 2087 + }, + { + "epoch": 0.18830319700590703, + "grad_norm": 1.606389730587654, + "learning_rate": 3.7429006253042524e-06, + "loss": 0.7869, + "step": 2088 + }, + { + "epoch": 0.1883933805293773, + "grad_norm": 1.6492781823100804, + "learning_rate": 3.7426140051056867e-06, + "loss": 0.733, + "step": 2089 + }, + { + "epoch": 0.18848356405284755, + "grad_norm": 1.8221293538485055, + "learning_rate": 3.7423272362176856e-06, + "loss": 0.7329, + "step": 2090 + }, + { + "epoch": 0.1885737475763178, + "grad_norm": 1.5470755090711708, + "learning_rate": 3.742040318664718e-06, + "loss": 0.833, + "step": 2091 + }, + { + "epoch": 0.18866393109978807, + "grad_norm": 2.2456939606837056, + "learning_rate": 3.7417532524712643e-06, + "loss": 0.8473, + "step": 2092 + }, + { + "epoch": 0.18875411462325833, + "grad_norm": 1.4831819257007355, + "learning_rate": 3.7414660376618195e-06, + "loss": 0.8221, + "step": 2093 + }, + { + "epoch": 0.1888442981467286, + "grad_norm": 0.6474685093133445, + "learning_rate": 3.74117867426089e-06, + "loss": 0.5904, + "step": 2094 + }, + { + "epoch": 0.18893448167019886, + "grad_norm": 2.3269231369445365, + "learning_rate": 3.7408911622929954e-06, + "loss": 0.7689, + "step": 2095 + }, + { + "epoch": 0.18902466519366912, + "grad_norm": 1.4911049316140499, + "learning_rate": 3.740603501782668e-06, + "loss": 0.8283, + "step": 2096 + }, + { + "epoch": 0.18911484871713938, + "grad_norm": 1.7186676005638866, + "learning_rate": 3.7403156927544516e-06, + "loss": 0.7354, + "step": 2097 + }, + { + "epoch": 0.18920503224060964, + "grad_norm": 2.109076245662036, + "learning_rate": 3.740027735232904e-06, + "loss": 0.7216, + "step": 2098 + }, + { + "epoch": 0.1892952157640799, + "grad_norm": 1.7644599514116204, + "learning_rate": 3.7397396292425966e-06, + "loss": 0.7757, + "step": 2099 + }, + { + "epoch": 0.18938539928755016, + "grad_norm": 1.835966199285328, + "learning_rate": 3.7394513748081105e-06, + "loss": 0.7685, + "step": 2100 + }, + { + "epoch": 0.18947558281102042, + "grad_norm": 1.6892364186566204, + "learning_rate": 3.7391629719540418e-06, + "loss": 0.8186, + "step": 2101 + }, + { + "epoch": 0.18956576633449068, + "grad_norm": 1.4591275187058623, + "learning_rate": 3.7388744207049998e-06, + "loss": 0.7788, + "step": 2102 + }, + { + "epoch": 0.18965594985796094, + "grad_norm": 2.1642435046355524, + "learning_rate": 3.7385857210856027e-06, + "loss": 0.7663, + "step": 2103 + }, + { + "epoch": 0.1897461333814312, + "grad_norm": 3.240717469591017, + "learning_rate": 3.738296873120486e-06, + "loss": 0.7707, + "step": 2104 + }, + { + "epoch": 0.18983631690490146, + "grad_norm": 1.713027583250358, + "learning_rate": 3.7380078768342955e-06, + "loss": 0.7603, + "step": 2105 + }, + { + "epoch": 0.18992650042837173, + "grad_norm": 2.067038311408102, + "learning_rate": 3.7377187322516895e-06, + "loss": 0.8061, + "step": 2106 + }, + { + "epoch": 0.19001668395184199, + "grad_norm": 1.5931452183359, + "learning_rate": 3.7374294393973395e-06, + "loss": 0.7461, + "step": 2107 + }, + { + "epoch": 0.19010686747531225, + "grad_norm": 1.5223356517164066, + "learning_rate": 3.7371399982959294e-06, + "loss": 0.8369, + "step": 2108 + }, + { + "epoch": 0.19019705099878254, + "grad_norm": 2.233518636427191, + "learning_rate": 3.7368504089721565e-06, + "loss": 0.7066, + "step": 2109 + }, + { + "epoch": 0.1902872345222528, + "grad_norm": 1.5502567509302876, + "learning_rate": 3.73656067145073e-06, + "loss": 0.7343, + "step": 2110 + }, + { + "epoch": 0.19037741804572306, + "grad_norm": 1.6593853046794125, + "learning_rate": 3.736270785756371e-06, + "loss": 0.8398, + "step": 2111 + }, + { + "epoch": 0.19046760156919332, + "grad_norm": 2.0361482688262273, + "learning_rate": 3.7359807519138156e-06, + "loss": 0.7994, + "step": 2112 + }, + { + "epoch": 0.19055778509266358, + "grad_norm": 1.4423269417285172, + "learning_rate": 3.73569056994781e-06, + "loss": 0.7818, + "step": 2113 + }, + { + "epoch": 0.19064796861613384, + "grad_norm": 2.107695073051502, + "learning_rate": 3.7354002398831144e-06, + "loss": 0.8287, + "step": 2114 + }, + { + "epoch": 0.1907381521396041, + "grad_norm": 1.9259340794335003, + "learning_rate": 3.7351097617445015e-06, + "loss": 0.7919, + "step": 2115 + }, + { + "epoch": 0.19082833566307436, + "grad_norm": 0.6950818868807335, + "learning_rate": 3.7348191355567567e-06, + "loss": 0.6118, + "step": 2116 + }, + { + "epoch": 0.19091851918654462, + "grad_norm": 1.6233173608614233, + "learning_rate": 3.734528361344677e-06, + "loss": 0.8154, + "step": 2117 + }, + { + "epoch": 0.19100870271001488, + "grad_norm": 1.7106957112346917, + "learning_rate": 3.734237439133074e-06, + "loss": 0.7843, + "step": 2118 + }, + { + "epoch": 0.19109888623348514, + "grad_norm": 1.733332277566637, + "learning_rate": 3.7339463689467702e-06, + "loss": 0.8372, + "step": 2119 + }, + { + "epoch": 0.1911890697569554, + "grad_norm": 1.7954280168704697, + "learning_rate": 3.733655150810601e-06, + "loss": 0.7575, + "step": 2120 + }, + { + "epoch": 0.19127925328042567, + "grad_norm": 2.0711228422224703, + "learning_rate": 3.7333637847494154e-06, + "loss": 0.9205, + "step": 2121 + }, + { + "epoch": 0.19136943680389593, + "grad_norm": 1.7393777539268787, + "learning_rate": 3.7330722707880734e-06, + "loss": 0.9419, + "step": 2122 + }, + { + "epoch": 0.1914596203273662, + "grad_norm": 1.39998485693235, + "learning_rate": 3.7327806089514497e-06, + "loss": 0.6945, + "step": 2123 + }, + { + "epoch": 0.19154980385083645, + "grad_norm": 1.5006513770731278, + "learning_rate": 3.7324887992644297e-06, + "loss": 0.884, + "step": 2124 + }, + { + "epoch": 0.1916399873743067, + "grad_norm": 0.7271809560680491, + "learning_rate": 3.7321968417519123e-06, + "loss": 0.5864, + "step": 2125 + }, + { + "epoch": 0.19173017089777697, + "grad_norm": 1.9991969045896711, + "learning_rate": 3.7319047364388097e-06, + "loss": 0.7749, + "step": 2126 + }, + { + "epoch": 0.19182035442124723, + "grad_norm": 0.6351733732085492, + "learning_rate": 3.7316124833500453e-06, + "loss": 0.5723, + "step": 2127 + }, + { + "epoch": 0.1919105379447175, + "grad_norm": 2.2876420023883344, + "learning_rate": 3.731320082510556e-06, + "loss": 0.7734, + "step": 2128 + }, + { + "epoch": 0.19200072146818775, + "grad_norm": 2.165769599171518, + "learning_rate": 3.7310275339452906e-06, + "loss": 0.8022, + "step": 2129 + }, + { + "epoch": 0.192090904991658, + "grad_norm": 6.0468459806312715, + "learning_rate": 3.7307348376792113e-06, + "loss": 0.7778, + "step": 2130 + }, + { + "epoch": 0.19218108851512827, + "grad_norm": 1.532207267644606, + "learning_rate": 3.730441993737292e-06, + "loss": 0.7823, + "step": 2131 + }, + { + "epoch": 0.19227127203859856, + "grad_norm": 2.038036804630576, + "learning_rate": 3.7301490021445205e-06, + "loss": 0.7735, + "step": 2132 + }, + { + "epoch": 0.19236145556206882, + "grad_norm": 2.1095765194176, + "learning_rate": 3.7298558629258966e-06, + "loss": 0.7326, + "step": 2133 + }, + { + "epoch": 0.19245163908553908, + "grad_norm": 1.5598184081293283, + "learning_rate": 3.7295625761064314e-06, + "loss": 0.825, + "step": 2134 + }, + { + "epoch": 0.19254182260900934, + "grad_norm": 3.4688580470397885, + "learning_rate": 3.7292691417111504e-06, + "loss": 0.755, + "step": 2135 + }, + { + "epoch": 0.1926320061324796, + "grad_norm": 1.6013258596373463, + "learning_rate": 3.728975559765092e-06, + "loss": 0.7468, + "step": 2136 + }, + { + "epoch": 0.19272218965594987, + "grad_norm": 1.9910765897694958, + "learning_rate": 3.728681830293305e-06, + "loss": 0.8163, + "step": 2137 + }, + { + "epoch": 0.19281237317942013, + "grad_norm": 1.6241678161121187, + "learning_rate": 3.7283879533208523e-06, + "loss": 0.7474, + "step": 2138 + }, + { + "epoch": 0.1929025567028904, + "grad_norm": 1.3743417204487536, + "learning_rate": 3.7280939288728094e-06, + "loss": 0.7722, + "step": 2139 + }, + { + "epoch": 0.19299274022636065, + "grad_norm": 1.860878600910839, + "learning_rate": 3.7277997569742637e-06, + "loss": 0.7513, + "step": 2140 + }, + { + "epoch": 0.1930829237498309, + "grad_norm": 4.647079961960151, + "learning_rate": 3.7275054376503155e-06, + "loss": 0.7565, + "step": 2141 + }, + { + "epoch": 0.19317310727330117, + "grad_norm": 1.9867619253635653, + "learning_rate": 3.7272109709260783e-06, + "loss": 0.7306, + "step": 2142 + }, + { + "epoch": 0.19326329079677143, + "grad_norm": 0.7681947517229645, + "learning_rate": 3.7269163568266774e-06, + "loss": 0.5883, + "step": 2143 + }, + { + "epoch": 0.1933534743202417, + "grad_norm": 1.7172389237094834, + "learning_rate": 3.7266215953772512e-06, + "loss": 0.774, + "step": 2144 + }, + { + "epoch": 0.19344365784371195, + "grad_norm": 2.9015492806161745, + "learning_rate": 3.7263266866029492e-06, + "loss": 0.8206, + "step": 2145 + }, + { + "epoch": 0.1935338413671822, + "grad_norm": 1.4411250307902217, + "learning_rate": 3.726031630528936e-06, + "loss": 0.7586, + "step": 2146 + }, + { + "epoch": 0.19362402489065247, + "grad_norm": 2.6835791797048056, + "learning_rate": 3.7257364271803865e-06, + "loss": 0.6715, + "step": 2147 + }, + { + "epoch": 0.19371420841412273, + "grad_norm": 2.077801672083547, + "learning_rate": 3.7254410765824896e-06, + "loss": 0.7126, + "step": 2148 + }, + { + "epoch": 0.193804391937593, + "grad_norm": 1.8966545869323037, + "learning_rate": 3.725145578760446e-06, + "loss": 0.7029, + "step": 2149 + }, + { + "epoch": 0.19389457546106326, + "grad_norm": 2.4336604664874266, + "learning_rate": 3.7248499337394696e-06, + "loss": 0.7034, + "step": 2150 + }, + { + "epoch": 0.19398475898453352, + "grad_norm": 1.9306239894322164, + "learning_rate": 3.7245541415447848e-06, + "loss": 0.8737, + "step": 2151 + }, + { + "epoch": 0.19407494250800378, + "grad_norm": 1.740058521843184, + "learning_rate": 3.724258202201633e-06, + "loss": 0.8033, + "step": 2152 + }, + { + "epoch": 0.19416512603147404, + "grad_norm": 1.8002761496536015, + "learning_rate": 3.7239621157352633e-06, + "loss": 0.7983, + "step": 2153 + }, + { + "epoch": 0.1942553095549443, + "grad_norm": 1.9122735207290993, + "learning_rate": 3.7236658821709403e-06, + "loss": 0.7718, + "step": 2154 + }, + { + "epoch": 0.19434549307841456, + "grad_norm": 1.7341207369444906, + "learning_rate": 3.7233695015339404e-06, + "loss": 0.8259, + "step": 2155 + }, + { + "epoch": 0.19443567660188485, + "grad_norm": 1.6506077369674086, + "learning_rate": 3.7230729738495513e-06, + "loss": 0.7686, + "step": 2156 + }, + { + "epoch": 0.1945258601253551, + "grad_norm": 1.8122480809320896, + "learning_rate": 3.722776299143075e-06, + "loss": 0.7791, + "step": 2157 + }, + { + "epoch": 0.19461604364882537, + "grad_norm": 0.6666803929783444, + "learning_rate": 3.722479477439826e-06, + "loss": 0.5869, + "step": 2158 + }, + { + "epoch": 0.19470622717229563, + "grad_norm": 1.3170184967115066, + "learning_rate": 3.7221825087651306e-06, + "loss": 0.7624, + "step": 2159 + }, + { + "epoch": 0.1947964106957659, + "grad_norm": 2.1866747798226425, + "learning_rate": 3.7218853931443274e-06, + "loss": 0.7215, + "step": 2160 + }, + { + "epoch": 0.19488659421923615, + "grad_norm": 1.3110941669389347, + "learning_rate": 3.721588130602768e-06, + "loss": 0.7245, + "step": 2161 + }, + { + "epoch": 0.19497677774270641, + "grad_norm": 2.02281597710565, + "learning_rate": 3.7212907211658164e-06, + "loss": 0.7617, + "step": 2162 + }, + { + "epoch": 0.19506696126617667, + "grad_norm": 1.5197291698857516, + "learning_rate": 3.72099316485885e-06, + "loss": 0.752, + "step": 2163 + }, + { + "epoch": 0.19515714478964694, + "grad_norm": 1.5027953168997599, + "learning_rate": 3.720695461707256e-06, + "loss": 0.8165, + "step": 2164 + }, + { + "epoch": 0.1952473283131172, + "grad_norm": 3.545116644098424, + "learning_rate": 3.7203976117364383e-06, + "loss": 0.8052, + "step": 2165 + }, + { + "epoch": 0.19533751183658746, + "grad_norm": 1.9952013742065897, + "learning_rate": 3.7200996149718105e-06, + "loss": 0.7683, + "step": 2166 + }, + { + "epoch": 0.19542769536005772, + "grad_norm": 1.3723233787139133, + "learning_rate": 3.7198014714387985e-06, + "loss": 0.7692, + "step": 2167 + }, + { + "epoch": 0.19551787888352798, + "grad_norm": 2.1864247540468074, + "learning_rate": 3.7195031811628422e-06, + "loss": 0.7129, + "step": 2168 + }, + { + "epoch": 0.19560806240699824, + "grad_norm": 2.6111054307681942, + "learning_rate": 3.719204744169393e-06, + "loss": 0.813, + "step": 2169 + }, + { + "epoch": 0.1956982459304685, + "grad_norm": 1.9171408674791393, + "learning_rate": 3.718906160483916e-06, + "loss": 0.7577, + "step": 2170 + }, + { + "epoch": 0.19578842945393876, + "grad_norm": 1.5633104130027902, + "learning_rate": 3.7186074301318868e-06, + "loss": 0.7541, + "step": 2171 + }, + { + "epoch": 0.19587861297740902, + "grad_norm": 1.55497302016621, + "learning_rate": 3.7183085531387957e-06, + "loss": 0.8552, + "step": 2172 + }, + { + "epoch": 0.19596879650087928, + "grad_norm": 1.6657539332087337, + "learning_rate": 3.7180095295301443e-06, + "loss": 0.7562, + "step": 2173 + }, + { + "epoch": 0.19605898002434954, + "grad_norm": 4.0363212447318695, + "learning_rate": 3.7177103593314465e-06, + "loss": 0.8446, + "step": 2174 + }, + { + "epoch": 0.1961491635478198, + "grad_norm": 1.7299861880947303, + "learning_rate": 3.7174110425682297e-06, + "loss": 0.8157, + "step": 2175 + }, + { + "epoch": 0.19623934707129007, + "grad_norm": 1.9121002103529414, + "learning_rate": 3.7171115792660333e-06, + "loss": 0.9035, + "step": 2176 + }, + { + "epoch": 0.19632953059476033, + "grad_norm": 1.632596234322789, + "learning_rate": 3.7168119694504083e-06, + "loss": 0.8487, + "step": 2177 + }, + { + "epoch": 0.1964197141182306, + "grad_norm": 1.8932866048867003, + "learning_rate": 3.71651221314692e-06, + "loss": 0.7792, + "step": 2178 + }, + { + "epoch": 0.19650989764170085, + "grad_norm": 1.6391589926362418, + "learning_rate": 3.716212310381145e-06, + "loss": 0.7801, + "step": 2179 + }, + { + "epoch": 0.19660008116517114, + "grad_norm": 1.5998946274391608, + "learning_rate": 3.7159122611786725e-06, + "loss": 0.8204, + "step": 2180 + }, + { + "epoch": 0.1966902646886414, + "grad_norm": 1.6361541866870388, + "learning_rate": 3.7156120655651045e-06, + "loss": 0.7675, + "step": 2181 + }, + { + "epoch": 0.19678044821211166, + "grad_norm": 1.3583878188331822, + "learning_rate": 3.7153117235660553e-06, + "loss": 0.7615, + "step": 2182 + }, + { + "epoch": 0.19687063173558192, + "grad_norm": 0.7091888889397645, + "learning_rate": 3.7150112352071514e-06, + "loss": 0.5862, + "step": 2183 + }, + { + "epoch": 0.19696081525905218, + "grad_norm": 1.9025739651912832, + "learning_rate": 3.7147106005140326e-06, + "loss": 0.747, + "step": 2184 + }, + { + "epoch": 0.19705099878252244, + "grad_norm": 0.6773159483381553, + "learning_rate": 3.714409819512351e-06, + "loss": 0.5947, + "step": 2185 + }, + { + "epoch": 0.1971411823059927, + "grad_norm": 2.472638604702295, + "learning_rate": 3.7141088922277695e-06, + "loss": 0.8075, + "step": 2186 + }, + { + "epoch": 0.19723136582946296, + "grad_norm": 1.764095082417144, + "learning_rate": 3.7138078186859664e-06, + "loss": 0.8289, + "step": 2187 + }, + { + "epoch": 0.19732154935293322, + "grad_norm": 1.5935946370113092, + "learning_rate": 3.7135065989126303e-06, + "loss": 0.7943, + "step": 2188 + }, + { + "epoch": 0.19741173287640348, + "grad_norm": 0.7530651242842608, + "learning_rate": 3.713205232933463e-06, + "loss": 0.6498, + "step": 2189 + }, + { + "epoch": 0.19750191639987374, + "grad_norm": 1.5537810630485969, + "learning_rate": 3.7129037207741792e-06, + "loss": 0.7884, + "step": 2190 + }, + { + "epoch": 0.197592099923344, + "grad_norm": 1.8396713921224226, + "learning_rate": 3.7126020624605046e-06, + "loss": 0.7949, + "step": 2191 + }, + { + "epoch": 0.19768228344681427, + "grad_norm": 1.8594205193597715, + "learning_rate": 3.7123002580181785e-06, + "loss": 0.7393, + "step": 2192 + }, + { + "epoch": 0.19777246697028453, + "grad_norm": 1.3597794731243722, + "learning_rate": 3.7119983074729532e-06, + "loss": 0.8537, + "step": 2193 + }, + { + "epoch": 0.1978626504937548, + "grad_norm": 1.6885199819776975, + "learning_rate": 3.7116962108505926e-06, + "loss": 0.8368, + "step": 2194 + }, + { + "epoch": 0.19795283401722505, + "grad_norm": 1.3979867816212819, + "learning_rate": 3.711393968176873e-06, + "loss": 0.8786, + "step": 2195 + }, + { + "epoch": 0.1980430175406953, + "grad_norm": 1.7572194094934719, + "learning_rate": 3.711091579477584e-06, + "loss": 0.788, + "step": 2196 + }, + { + "epoch": 0.19813320106416557, + "grad_norm": 2.0546398084784174, + "learning_rate": 3.7107890447785255e-06, + "loss": 0.8178, + "step": 2197 + }, + { + "epoch": 0.19822338458763583, + "grad_norm": 1.8683554854881896, + "learning_rate": 3.710486364105513e-06, + "loss": 0.835, + "step": 2198 + }, + { + "epoch": 0.1983135681111061, + "grad_norm": 2.100319420272853, + "learning_rate": 3.7101835374843728e-06, + "loss": 0.7808, + "step": 2199 + }, + { + "epoch": 0.19840375163457635, + "grad_norm": 1.5033615909523015, + "learning_rate": 3.7098805649409427e-06, + "loss": 0.807, + "step": 2200 + }, + { + "epoch": 0.1984939351580466, + "grad_norm": 1.7666153957316677, + "learning_rate": 3.7095774465010748e-06, + "loss": 0.8499, + "step": 2201 + }, + { + "epoch": 0.19858411868151687, + "grad_norm": 1.656575440876357, + "learning_rate": 3.7092741821906328e-06, + "loss": 0.7855, + "step": 2202 + }, + { + "epoch": 0.19867430220498714, + "grad_norm": 1.8529411366228514, + "learning_rate": 3.708970772035493e-06, + "loss": 0.7643, + "step": 2203 + }, + { + "epoch": 0.19876448572845742, + "grad_norm": 1.3986446717314431, + "learning_rate": 3.7086672160615427e-06, + "loss": 0.7528, + "step": 2204 + }, + { + "epoch": 0.19885466925192768, + "grad_norm": 2.8371522261797306, + "learning_rate": 3.7083635142946852e-06, + "loss": 0.7667, + "step": 2205 + }, + { + "epoch": 0.19894485277539795, + "grad_norm": 1.6787402653642696, + "learning_rate": 3.7080596667608327e-06, + "loss": 0.82, + "step": 2206 + }, + { + "epoch": 0.1990350362988682, + "grad_norm": 1.5636135710825299, + "learning_rate": 3.707755673485911e-06, + "loss": 0.829, + "step": 2207 + }, + { + "epoch": 0.19912521982233847, + "grad_norm": 1.6605409423320128, + "learning_rate": 3.7074515344958584e-06, + "loss": 0.6966, + "step": 2208 + }, + { + "epoch": 0.19921540334580873, + "grad_norm": 0.6597692269316712, + "learning_rate": 3.707147249816627e-06, + "loss": 0.614, + "step": 2209 + }, + { + "epoch": 0.199305586869279, + "grad_norm": 1.5551845196967997, + "learning_rate": 3.706842819474178e-06, + "loss": 0.8062, + "step": 2210 + }, + { + "epoch": 0.19939577039274925, + "grad_norm": 1.7533089825108055, + "learning_rate": 3.706538243494489e-06, + "loss": 0.7293, + "step": 2211 + }, + { + "epoch": 0.1994859539162195, + "grad_norm": 1.7237739656482245, + "learning_rate": 3.706233521903547e-06, + "loss": 0.7613, + "step": 2212 + }, + { + "epoch": 0.19957613743968977, + "grad_norm": 2.209471954851408, + "learning_rate": 3.705928654727353e-06, + "loss": 0.8654, + "step": 2213 + }, + { + "epoch": 0.19966632096316003, + "grad_norm": 6.2177101684438085, + "learning_rate": 3.7056236419919195e-06, + "loss": 0.7678, + "step": 2214 + }, + { + "epoch": 0.1997565044866303, + "grad_norm": 1.8847180612521748, + "learning_rate": 3.705318483723273e-06, + "loss": 0.8341, + "step": 2215 + }, + { + "epoch": 0.19984668801010055, + "grad_norm": 1.6320317057914335, + "learning_rate": 3.7050131799474493e-06, + "loss": 0.7619, + "step": 2216 + }, + { + "epoch": 0.19993687153357081, + "grad_norm": 1.4271222222962636, + "learning_rate": 3.7047077306905e-06, + "loss": 0.786, + "step": 2217 + }, + { + "epoch": 0.20002705505704108, + "grad_norm": 1.5177843893871212, + "learning_rate": 3.704402135978488e-06, + "loss": 0.8185, + "step": 2218 + }, + { + "epoch": 0.20011723858051134, + "grad_norm": 2.3937459283301084, + "learning_rate": 3.7040963958374877e-06, + "loss": 0.8096, + "step": 2219 + }, + { + "epoch": 0.2002074221039816, + "grad_norm": 1.4204588324742216, + "learning_rate": 3.7037905102935864e-06, + "loss": 0.8178, + "step": 2220 + }, + { + "epoch": 0.20029760562745186, + "grad_norm": 1.5988069974210275, + "learning_rate": 3.7034844793728837e-06, + "loss": 0.7964, + "step": 2221 + }, + { + "epoch": 0.20038778915092212, + "grad_norm": 1.540051447516782, + "learning_rate": 3.7031783031014933e-06, + "loss": 0.6715, + "step": 2222 + }, + { + "epoch": 0.20047797267439238, + "grad_norm": 2.218759026307564, + "learning_rate": 3.702871981505538e-06, + "loss": 0.7625, + "step": 2223 + }, + { + "epoch": 0.20056815619786264, + "grad_norm": 2.1063854366417685, + "learning_rate": 3.7025655146111563e-06, + "loss": 0.7296, + "step": 2224 + }, + { + "epoch": 0.2006583397213329, + "grad_norm": 0.6834960649721717, + "learning_rate": 3.702258902444497e-06, + "loss": 0.6651, + "step": 2225 + }, + { + "epoch": 0.20074852324480316, + "grad_norm": 1.4329365151418962, + "learning_rate": 3.701952145031722e-06, + "loss": 0.842, + "step": 2226 + }, + { + "epoch": 0.20083870676827342, + "grad_norm": 1.6758620591675861, + "learning_rate": 3.701645242399005e-06, + "loss": 0.8646, + "step": 2227 + }, + { + "epoch": 0.2009288902917437, + "grad_norm": 2.028931571088565, + "learning_rate": 3.701338194572533e-06, + "loss": 0.8527, + "step": 2228 + }, + { + "epoch": 0.20101907381521397, + "grad_norm": 1.4824292731822968, + "learning_rate": 3.7010310015785056e-06, + "loss": 0.7674, + "step": 2229 + }, + { + "epoch": 0.20110925733868423, + "grad_norm": 1.708317252602915, + "learning_rate": 3.700723663443134e-06, + "loss": 0.731, + "step": 2230 + }, + { + "epoch": 0.2011994408621545, + "grad_norm": 0.6965001282592568, + "learning_rate": 3.7004161801926416e-06, + "loss": 0.5919, + "step": 2231 + }, + { + "epoch": 0.20128962438562475, + "grad_norm": 1.7730960601284391, + "learning_rate": 3.7001085518532643e-06, + "loss": 0.8091, + "step": 2232 + }, + { + "epoch": 0.20137980790909502, + "grad_norm": 1.671137210232659, + "learning_rate": 3.6998007784512515e-06, + "loss": 0.7809, + "step": 2233 + }, + { + "epoch": 0.20146999143256528, + "grad_norm": 1.7368107780341757, + "learning_rate": 3.6994928600128637e-06, + "loss": 0.6587, + "step": 2234 + }, + { + "epoch": 0.20156017495603554, + "grad_norm": 1.6089868910867755, + "learning_rate": 3.6991847965643742e-06, + "loss": 0.7535, + "step": 2235 + }, + { + "epoch": 0.2016503584795058, + "grad_norm": 1.6735232063236773, + "learning_rate": 3.698876588132068e-06, + "loss": 0.7434, + "step": 2236 + }, + { + "epoch": 0.20174054200297606, + "grad_norm": 1.5857752825077547, + "learning_rate": 3.6985682347422446e-06, + "loss": 0.7447, + "step": 2237 + }, + { + "epoch": 0.20183072552644632, + "grad_norm": 2.3198566438516113, + "learning_rate": 3.698259736421213e-06, + "loss": 0.7943, + "step": 2238 + }, + { + "epoch": 0.20192090904991658, + "grad_norm": 1.74427198575349, + "learning_rate": 3.697951093195297e-06, + "loss": 0.7671, + "step": 2239 + }, + { + "epoch": 0.20201109257338684, + "grad_norm": 2.240714090829265, + "learning_rate": 3.6976423050908307e-06, + "loss": 0.7648, + "step": 2240 + }, + { + "epoch": 0.2021012760968571, + "grad_norm": 1.4638191403576777, + "learning_rate": 3.697333372134163e-06, + "loss": 0.7728, + "step": 2241 + }, + { + "epoch": 0.20219145962032736, + "grad_norm": 3.557783786287568, + "learning_rate": 3.697024294351653e-06, + "loss": 0.8828, + "step": 2242 + }, + { + "epoch": 0.20228164314379762, + "grad_norm": 1.6941679857545675, + "learning_rate": 3.696715071769672e-06, + "loss": 0.7446, + "step": 2243 + }, + { + "epoch": 0.20237182666726788, + "grad_norm": 1.735685729577368, + "learning_rate": 3.696405704414606e-06, + "loss": 0.7849, + "step": 2244 + }, + { + "epoch": 0.20246201019073815, + "grad_norm": 1.452268696816149, + "learning_rate": 3.6960961923128514e-06, + "loss": 0.8869, + "step": 2245 + }, + { + "epoch": 0.2025521937142084, + "grad_norm": 1.985127584167109, + "learning_rate": 3.6957865354908177e-06, + "loss": 0.7268, + "step": 2246 + }, + { + "epoch": 0.20264237723767867, + "grad_norm": 1.6664811587408825, + "learning_rate": 3.6954767339749262e-06, + "loss": 0.8022, + "step": 2247 + }, + { + "epoch": 0.20273256076114893, + "grad_norm": 2.16537479865409, + "learning_rate": 3.6951667877916113e-06, + "loss": 0.7686, + "step": 2248 + }, + { + "epoch": 0.2028227442846192, + "grad_norm": 1.6058044082710325, + "learning_rate": 3.694856696967319e-06, + "loss": 0.7912, + "step": 2249 + }, + { + "epoch": 0.20291292780808945, + "grad_norm": 2.286173589148037, + "learning_rate": 3.6945464615285077e-06, + "loss": 0.7686, + "step": 2250 + }, + { + "epoch": 0.2030031113315597, + "grad_norm": 0.704109075455919, + "learning_rate": 3.694236081501648e-06, + "loss": 0.693, + "step": 2251 + }, + { + "epoch": 0.20309329485503, + "grad_norm": 4.398930090090089, + "learning_rate": 3.6939255569132246e-06, + "loss": 0.744, + "step": 2252 + }, + { + "epoch": 0.20318347837850026, + "grad_norm": 0.8445169354124615, + "learning_rate": 3.693614887789733e-06, + "loss": 0.6198, + "step": 2253 + }, + { + "epoch": 0.20327366190197052, + "grad_norm": 0.8806294820006804, + "learning_rate": 3.69330407415768e-06, + "loss": 0.6208, + "step": 2254 + }, + { + "epoch": 0.20336384542544078, + "grad_norm": 1.778030028262343, + "learning_rate": 3.6929931160435867e-06, + "loss": 0.7865, + "step": 2255 + }, + { + "epoch": 0.20345402894891104, + "grad_norm": 1.9419300746566435, + "learning_rate": 3.6926820134739858e-06, + "loss": 0.84, + "step": 2256 + }, + { + "epoch": 0.2035442124723813, + "grad_norm": 1.692178034733923, + "learning_rate": 3.692370766475422e-06, + "loss": 0.7179, + "step": 2257 + }, + { + "epoch": 0.20363439599585156, + "grad_norm": 1.7415209123624673, + "learning_rate": 3.692059375074453e-06, + "loss": 0.8708, + "step": 2258 + }, + { + "epoch": 0.20372457951932182, + "grad_norm": 1.9662114079805362, + "learning_rate": 3.6917478392976475e-06, + "loss": 0.8493, + "step": 2259 + }, + { + "epoch": 0.20381476304279209, + "grad_norm": 1.4879611414480798, + "learning_rate": 3.691436159171589e-06, + "loss": 0.8453, + "step": 2260 + }, + { + "epoch": 0.20390494656626235, + "grad_norm": 1.6030983608478997, + "learning_rate": 3.6911243347228703e-06, + "loss": 0.7664, + "step": 2261 + }, + { + "epoch": 0.2039951300897326, + "grad_norm": 0.6408816498035114, + "learning_rate": 3.690812365978099e-06, + "loss": 0.5335, + "step": 2262 + }, + { + "epoch": 0.20408531361320287, + "grad_norm": 1.5551620602563865, + "learning_rate": 3.690500252963893e-06, + "loss": 0.7397, + "step": 2263 + }, + { + "epoch": 0.20417549713667313, + "grad_norm": 1.710105310398514, + "learning_rate": 3.6901879957068846e-06, + "loss": 0.7721, + "step": 2264 + }, + { + "epoch": 0.2042656806601434, + "grad_norm": 1.419747909950079, + "learning_rate": 3.689875594233717e-06, + "loss": 0.872, + "step": 2265 + }, + { + "epoch": 0.20435586418361365, + "grad_norm": 1.70115211378986, + "learning_rate": 3.689563048571046e-06, + "loss": 0.7695, + "step": 2266 + }, + { + "epoch": 0.2044460477070839, + "grad_norm": 1.394668102227804, + "learning_rate": 3.6892503587455395e-06, + "loss": 0.7776, + "step": 2267 + }, + { + "epoch": 0.20453623123055417, + "grad_norm": 1.4653296720084723, + "learning_rate": 3.6889375247838766e-06, + "loss": 0.7917, + "step": 2268 + }, + { + "epoch": 0.20462641475402443, + "grad_norm": 1.5724481133177457, + "learning_rate": 3.688624546712753e-06, + "loss": 0.7856, + "step": 2269 + }, + { + "epoch": 0.2047165982774947, + "grad_norm": 1.5615049627092685, + "learning_rate": 3.688311424558871e-06, + "loss": 0.746, + "step": 2270 + }, + { + "epoch": 0.20480678180096495, + "grad_norm": 3.326020484296946, + "learning_rate": 3.6879981583489496e-06, + "loss": 0.6943, + "step": 2271 + }, + { + "epoch": 0.20489696532443522, + "grad_norm": 1.8694964700947625, + "learning_rate": 3.687684748109718e-06, + "loss": 0.8043, + "step": 2272 + }, + { + "epoch": 0.20498714884790548, + "grad_norm": 2.01496841027272, + "learning_rate": 3.6873711938679174e-06, + "loss": 0.8457, + "step": 2273 + }, + { + "epoch": 0.20507733237137574, + "grad_norm": 1.935340970478414, + "learning_rate": 3.6870574956503027e-06, + "loss": 0.7769, + "step": 2274 + }, + { + "epoch": 0.20516751589484603, + "grad_norm": 1.7752659316385528, + "learning_rate": 3.68674365348364e-06, + "loss": 0.7775, + "step": 2275 + }, + { + "epoch": 0.2052576994183163, + "grad_norm": 1.643548014599496, + "learning_rate": 3.6864296673947086e-06, + "loss": 0.7904, + "step": 2276 + }, + { + "epoch": 0.20534788294178655, + "grad_norm": 1.6124538119302891, + "learning_rate": 3.686115537410298e-06, + "loss": 0.7671, + "step": 2277 + }, + { + "epoch": 0.2054380664652568, + "grad_norm": 1.7093059128323984, + "learning_rate": 3.685801263557214e-06, + "loss": 0.733, + "step": 2278 + }, + { + "epoch": 0.20552824998872707, + "grad_norm": 1.846294974242261, + "learning_rate": 3.68548684586227e-06, + "loss": 0.8767, + "step": 2279 + }, + { + "epoch": 0.20561843351219733, + "grad_norm": 1.76573909964742, + "learning_rate": 3.685172284352295e-06, + "loss": 0.7881, + "step": 2280 + }, + { + "epoch": 0.2057086170356676, + "grad_norm": 1.2878269169386227, + "learning_rate": 3.684857579054128e-06, + "loss": 0.7157, + "step": 2281 + }, + { + "epoch": 0.20579880055913785, + "grad_norm": 1.7709167984384746, + "learning_rate": 3.6845427299946233e-06, + "loss": 0.8491, + "step": 2282 + }, + { + "epoch": 0.2058889840826081, + "grad_norm": 1.65182999363762, + "learning_rate": 3.6842277372006434e-06, + "loss": 0.8785, + "step": 2283 + }, + { + "epoch": 0.20597916760607837, + "grad_norm": 1.4769792725458781, + "learning_rate": 3.6839126006990664e-06, + "loss": 0.7546, + "step": 2284 + }, + { + "epoch": 0.20606935112954863, + "grad_norm": 1.6944590587020483, + "learning_rate": 3.6835973205167818e-06, + "loss": 0.8363, + "step": 2285 + }, + { + "epoch": 0.2061595346530189, + "grad_norm": 1.9572419987728278, + "learning_rate": 3.6832818966806904e-06, + "loss": 0.8238, + "step": 2286 + }, + { + "epoch": 0.20624971817648916, + "grad_norm": 1.589332592121267, + "learning_rate": 3.682966329217706e-06, + "loss": 0.8631, + "step": 2287 + }, + { + "epoch": 0.20633990169995942, + "grad_norm": 2.091581559697809, + "learning_rate": 3.6826506181547543e-06, + "loss": 0.8192, + "step": 2288 + }, + { + "epoch": 0.20643008522342968, + "grad_norm": 1.6855049349096722, + "learning_rate": 3.682334763518774e-06, + "loss": 0.8077, + "step": 2289 + }, + { + "epoch": 0.20652026874689994, + "grad_norm": 1.7291777124013812, + "learning_rate": 3.6820187653367158e-06, + "loss": 0.7487, + "step": 2290 + }, + { + "epoch": 0.2066104522703702, + "grad_norm": 2.3412633292841667, + "learning_rate": 3.6817026236355412e-06, + "loss": 0.672, + "step": 2291 + }, + { + "epoch": 0.20670063579384046, + "grad_norm": 1.9822525929301604, + "learning_rate": 3.681386338442227e-06, + "loss": 0.7681, + "step": 2292 + }, + { + "epoch": 0.20679081931731072, + "grad_norm": 1.3939836212192083, + "learning_rate": 3.681069909783758e-06, + "loss": 0.7441, + "step": 2293 + }, + { + "epoch": 0.20688100284078098, + "grad_norm": 1.6451956885914105, + "learning_rate": 3.680753337687136e-06, + "loss": 0.7935, + "step": 2294 + }, + { + "epoch": 0.20697118636425124, + "grad_norm": 1.77329875386495, + "learning_rate": 3.680436622179371e-06, + "loss": 0.7131, + "step": 2295 + }, + { + "epoch": 0.2070613698877215, + "grad_norm": 2.4647081340836814, + "learning_rate": 3.680119763287488e-06, + "loss": 0.7712, + "step": 2296 + }, + { + "epoch": 0.20715155341119176, + "grad_norm": 1.891362322329112, + "learning_rate": 3.6798027610385227e-06, + "loss": 0.7244, + "step": 2297 + }, + { + "epoch": 0.20724173693466202, + "grad_norm": 2.015029464392669, + "learning_rate": 3.6794856154595235e-06, + "loss": 0.8305, + "step": 2298 + }, + { + "epoch": 0.2073319204581323, + "grad_norm": 2.007458013568158, + "learning_rate": 3.6791683265775506e-06, + "loss": 0.7945, + "step": 2299 + }, + { + "epoch": 0.20742210398160257, + "grad_norm": 1.950909510052261, + "learning_rate": 3.6788508944196773e-06, + "loss": 0.7411, + "step": 2300 + }, + { + "epoch": 0.20751228750507283, + "grad_norm": 1.6084268980860823, + "learning_rate": 3.678533319012989e-06, + "loss": 0.8556, + "step": 2301 + }, + { + "epoch": 0.2076024710285431, + "grad_norm": 3.6161096668408206, + "learning_rate": 3.6782156003845826e-06, + "loss": 0.7305, + "step": 2302 + }, + { + "epoch": 0.20769265455201336, + "grad_norm": 0.6465087609431344, + "learning_rate": 3.6778977385615676e-06, + "loss": 0.5951, + "step": 2303 + }, + { + "epoch": 0.20778283807548362, + "grad_norm": 1.7158018316360304, + "learning_rate": 3.6775797335710656e-06, + "loss": 0.8317, + "step": 2304 + }, + { + "epoch": 0.20787302159895388, + "grad_norm": 1.8550492304312596, + "learning_rate": 3.6772615854402105e-06, + "loss": 0.7365, + "step": 2305 + }, + { + "epoch": 0.20796320512242414, + "grad_norm": 1.6716963057522654, + "learning_rate": 3.6769432941961487e-06, + "loss": 0.665, + "step": 2306 + }, + { + "epoch": 0.2080533886458944, + "grad_norm": 1.7827526246296146, + "learning_rate": 3.676624859866038e-06, + "loss": 0.7832, + "step": 2307 + }, + { + "epoch": 0.20814357216936466, + "grad_norm": 2.2639341161327864, + "learning_rate": 3.67630628247705e-06, + "loss": 0.7074, + "step": 2308 + }, + { + "epoch": 0.20823375569283492, + "grad_norm": 1.778045783934901, + "learning_rate": 3.675987562056367e-06, + "loss": 0.8457, + "step": 2309 + }, + { + "epoch": 0.20832393921630518, + "grad_norm": 2.142654681859249, + "learning_rate": 3.675668698631184e-06, + "loss": 0.8196, + "step": 2310 + }, + { + "epoch": 0.20841412273977544, + "grad_norm": 1.4909470280295054, + "learning_rate": 3.675349692228708e-06, + "loss": 0.8478, + "step": 2311 + }, + { + "epoch": 0.2085043062632457, + "grad_norm": 2.452478894717429, + "learning_rate": 3.6750305428761578e-06, + "loss": 0.7601, + "step": 2312 + }, + { + "epoch": 0.20859448978671596, + "grad_norm": 2.1482911216114497, + "learning_rate": 3.674711250600766e-06, + "loss": 0.7915, + "step": 2313 + }, + { + "epoch": 0.20868467331018623, + "grad_norm": 0.8581935043386061, + "learning_rate": 3.6743918154297765e-06, + "loss": 0.5764, + "step": 2314 + }, + { + "epoch": 0.20877485683365649, + "grad_norm": 1.6940737650640456, + "learning_rate": 3.6740722373904446e-06, + "loss": 0.7783, + "step": 2315 + }, + { + "epoch": 0.20886504035712675, + "grad_norm": 1.4529557488431344, + "learning_rate": 3.6737525165100383e-06, + "loss": 0.7616, + "step": 2316 + }, + { + "epoch": 0.208955223880597, + "grad_norm": 2.133577873201638, + "learning_rate": 3.6734326528158385e-06, + "loss": 0.8062, + "step": 2317 + }, + { + "epoch": 0.20904540740406727, + "grad_norm": 1.3664011846658946, + "learning_rate": 3.673112646335138e-06, + "loss": 0.7674, + "step": 2318 + }, + { + "epoch": 0.20913559092753753, + "grad_norm": 1.574284933133039, + "learning_rate": 3.672792497095241e-06, + "loss": 0.7371, + "step": 2319 + }, + { + "epoch": 0.2092257744510078, + "grad_norm": 2.0969946473234784, + "learning_rate": 3.672472205123464e-06, + "loss": 0.7906, + "step": 2320 + }, + { + "epoch": 0.20931595797447805, + "grad_norm": 1.9369552215575903, + "learning_rate": 3.6721517704471363e-06, + "loss": 0.7103, + "step": 2321 + }, + { + "epoch": 0.2094061414979483, + "grad_norm": 1.6002658712502602, + "learning_rate": 3.6718311930936e-06, + "loss": 0.818, + "step": 2322 + }, + { + "epoch": 0.2094963250214186, + "grad_norm": 1.7610776288686802, + "learning_rate": 3.6715104730902074e-06, + "loss": 0.7318, + "step": 2323 + }, + { + "epoch": 0.20958650854488886, + "grad_norm": 1.7038031811436065, + "learning_rate": 3.671189610464325e-06, + "loss": 0.8276, + "step": 2324 + }, + { + "epoch": 0.20967669206835912, + "grad_norm": 1.8799146456314497, + "learning_rate": 3.6708686052433303e-06, + "loss": 0.7299, + "step": 2325 + }, + { + "epoch": 0.20976687559182938, + "grad_norm": 1.5859966548918074, + "learning_rate": 3.6705474574546127e-06, + "loss": 0.7984, + "step": 2326 + }, + { + "epoch": 0.20985705911529964, + "grad_norm": 3.10641138897109, + "learning_rate": 3.670226167125575e-06, + "loss": 0.8031, + "step": 2327 + }, + { + "epoch": 0.2099472426387699, + "grad_norm": 1.8873648790916433, + "learning_rate": 3.6699047342836313e-06, + "loss": 0.7862, + "step": 2328 + }, + { + "epoch": 0.21003742616224017, + "grad_norm": 1.552430628211037, + "learning_rate": 3.669583158956208e-06, + "loss": 0.8174, + "step": 2329 + }, + { + "epoch": 0.21012760968571043, + "grad_norm": 0.6728826108725644, + "learning_rate": 3.669261441170743e-06, + "loss": 0.6077, + "step": 2330 + }, + { + "epoch": 0.2102177932091807, + "grad_norm": 2.047513905758616, + "learning_rate": 3.668939580954688e-06, + "loss": 0.7934, + "step": 2331 + }, + { + "epoch": 0.21030797673265095, + "grad_norm": 1.4802552726044802, + "learning_rate": 3.668617578335506e-06, + "loss": 0.7586, + "step": 2332 + }, + { + "epoch": 0.2103981602561212, + "grad_norm": 2.764401181239551, + "learning_rate": 3.6682954333406707e-06, + "loss": 0.8126, + "step": 2333 + }, + { + "epoch": 0.21048834377959147, + "grad_norm": 1.9461707860644923, + "learning_rate": 3.6679731459976707e-06, + "loss": 0.9244, + "step": 2334 + }, + { + "epoch": 0.21057852730306173, + "grad_norm": 1.5279439932060703, + "learning_rate": 3.6676507163340046e-06, + "loss": 0.7794, + "step": 2335 + }, + { + "epoch": 0.210668710826532, + "grad_norm": 4.044549104246979, + "learning_rate": 3.6673281443771842e-06, + "loss": 0.8182, + "step": 2336 + }, + { + "epoch": 0.21075889435000225, + "grad_norm": 1.979999554373951, + "learning_rate": 3.667005430154733e-06, + "loss": 0.674, + "step": 2337 + }, + { + "epoch": 0.2108490778734725, + "grad_norm": 0.6644859310351425, + "learning_rate": 3.666682573694186e-06, + "loss": 0.6814, + "step": 2338 + }, + { + "epoch": 0.21093926139694277, + "grad_norm": 0.6642342289296792, + "learning_rate": 3.6663595750230924e-06, + "loss": 0.5508, + "step": 2339 + }, + { + "epoch": 0.21102944492041303, + "grad_norm": 1.9157292796590846, + "learning_rate": 3.666036434169012e-06, + "loss": 0.8388, + "step": 2340 + }, + { + "epoch": 0.2111196284438833, + "grad_norm": 1.4162494755716186, + "learning_rate": 3.665713151159516e-06, + "loss": 0.764, + "step": 2341 + }, + { + "epoch": 0.21120981196735356, + "grad_norm": 2.057613132003894, + "learning_rate": 3.665389726022189e-06, + "loss": 0.8824, + "step": 2342 + }, + { + "epoch": 0.21129999549082382, + "grad_norm": 1.6405319914112464, + "learning_rate": 3.6650661587846283e-06, + "loss": 0.7535, + "step": 2343 + }, + { + "epoch": 0.21139017901429408, + "grad_norm": 1.7891216018442997, + "learning_rate": 3.6647424494744418e-06, + "loss": 0.7457, + "step": 2344 + }, + { + "epoch": 0.21148036253776434, + "grad_norm": 1.6700489375088134, + "learning_rate": 3.6644185981192503e-06, + "loss": 0.8358, + "step": 2345 + }, + { + "epoch": 0.2115705460612346, + "grad_norm": 1.9558236361332144, + "learning_rate": 3.6640946047466868e-06, + "loss": 0.7199, + "step": 2346 + }, + { + "epoch": 0.2116607295847049, + "grad_norm": 1.6374092193366334, + "learning_rate": 3.6637704693843953e-06, + "loss": 0.749, + "step": 2347 + }, + { + "epoch": 0.21175091310817515, + "grad_norm": 1.4805893660154583, + "learning_rate": 3.6634461920600337e-06, + "loss": 0.7173, + "step": 2348 + }, + { + "epoch": 0.2118410966316454, + "grad_norm": 3.9328197941430174, + "learning_rate": 3.66312177280127e-06, + "loss": 0.8079, + "step": 2349 + }, + { + "epoch": 0.21193128015511567, + "grad_norm": 1.4198208737558609, + "learning_rate": 3.6627972116357872e-06, + "loss": 0.8204, + "step": 2350 + }, + { + "epoch": 0.21202146367858593, + "grad_norm": 2.1043863716871893, + "learning_rate": 3.662472508591278e-06, + "loss": 0.8468, + "step": 2351 + }, + { + "epoch": 0.2121116472020562, + "grad_norm": 1.9323147329374362, + "learning_rate": 3.662147663695447e-06, + "loss": 0.7722, + "step": 2352 + }, + { + "epoch": 0.21220183072552645, + "grad_norm": 3.7293148473073012, + "learning_rate": 3.6618226769760127e-06, + "loss": 0.7729, + "step": 2353 + }, + { + "epoch": 0.2122920142489967, + "grad_norm": 2.405040982188678, + "learning_rate": 3.661497548460704e-06, + "loss": 0.6711, + "step": 2354 + }, + { + "epoch": 0.21238219777246697, + "grad_norm": 3.259521767812909, + "learning_rate": 3.6611722781772635e-06, + "loss": 0.8028, + "step": 2355 + }, + { + "epoch": 0.21247238129593723, + "grad_norm": 0.8301129502115647, + "learning_rate": 3.6608468661534444e-06, + "loss": 0.6451, + "step": 2356 + }, + { + "epoch": 0.2125625648194075, + "grad_norm": 2.015603826840404, + "learning_rate": 3.660521312417013e-06, + "loss": 0.7132, + "step": 2357 + }, + { + "epoch": 0.21265274834287776, + "grad_norm": 2.01257661472227, + "learning_rate": 3.660195616995747e-06, + "loss": 0.7502, + "step": 2358 + }, + { + "epoch": 0.21274293186634802, + "grad_norm": 2.4261031129250625, + "learning_rate": 3.6598697799174367e-06, + "loss": 0.813, + "step": 2359 + }, + { + "epoch": 0.21283311538981828, + "grad_norm": 2.1669502684022905, + "learning_rate": 3.6595438012098844e-06, + "loss": 0.8494, + "step": 2360 + }, + { + "epoch": 0.21292329891328854, + "grad_norm": 2.0455040453691558, + "learning_rate": 3.6592176809009045e-06, + "loss": 0.7584, + "step": 2361 + }, + { + "epoch": 0.2130134824367588, + "grad_norm": 1.7477158216982265, + "learning_rate": 3.6588914190183227e-06, + "loss": 0.8175, + "step": 2362 + }, + { + "epoch": 0.21310366596022906, + "grad_norm": 1.948639135626706, + "learning_rate": 3.658565015589978e-06, + "loss": 0.656, + "step": 2363 + }, + { + "epoch": 0.21319384948369932, + "grad_norm": 2.7029197482644323, + "learning_rate": 3.6582384706437217e-06, + "loss": 0.7875, + "step": 2364 + }, + { + "epoch": 0.21328403300716958, + "grad_norm": 1.9873278417339717, + "learning_rate": 3.6579117842074156e-06, + "loss": 0.8023, + "step": 2365 + }, + { + "epoch": 0.21337421653063984, + "grad_norm": 1.7097802276371217, + "learning_rate": 3.657584956308934e-06, + "loss": 0.7225, + "step": 2366 + }, + { + "epoch": 0.2134644000541101, + "grad_norm": 3.45914193930197, + "learning_rate": 3.6572579869761648e-06, + "loss": 0.8465, + "step": 2367 + }, + { + "epoch": 0.21355458357758036, + "grad_norm": 2.645270330247302, + "learning_rate": 3.6569308762370056e-06, + "loss": 0.834, + "step": 2368 + }, + { + "epoch": 0.21364476710105063, + "grad_norm": 2.0693325962617553, + "learning_rate": 3.6566036241193676e-06, + "loss": 0.8368, + "step": 2369 + }, + { + "epoch": 0.2137349506245209, + "grad_norm": 0.6199434292272539, + "learning_rate": 3.656276230651174e-06, + "loss": 0.6085, + "step": 2370 + }, + { + "epoch": 0.21382513414799117, + "grad_norm": 0.7116851229213816, + "learning_rate": 3.65594869586036e-06, + "loss": 0.6429, + "step": 2371 + }, + { + "epoch": 0.21391531767146144, + "grad_norm": 1.760779153496109, + "learning_rate": 3.6556210197748724e-06, + "loss": 0.705, + "step": 2372 + }, + { + "epoch": 0.2140055011949317, + "grad_norm": 1.6355824315643326, + "learning_rate": 3.655293202422671e-06, + "loss": 0.7527, + "step": 2373 + }, + { + "epoch": 0.21409568471840196, + "grad_norm": 1.3654329649381574, + "learning_rate": 3.654965243831725e-06, + "loss": 0.7965, + "step": 2374 + }, + { + "epoch": 0.21418586824187222, + "grad_norm": 1.4859167023335473, + "learning_rate": 3.65463714403002e-06, + "loss": 0.8259, + "step": 2375 + }, + { + "epoch": 0.21427605176534248, + "grad_norm": 1.4677444424687245, + "learning_rate": 3.65430890304555e-06, + "loss": 0.7378, + "step": 2376 + }, + { + "epoch": 0.21436623528881274, + "grad_norm": 1.504087125131577, + "learning_rate": 3.653980520906323e-06, + "loss": 0.8112, + "step": 2377 + }, + { + "epoch": 0.214456418812283, + "grad_norm": 2.178271527608976, + "learning_rate": 3.653651997640358e-06, + "loss": 0.763, + "step": 2378 + }, + { + "epoch": 0.21454660233575326, + "grad_norm": 1.6934052739418184, + "learning_rate": 3.653323333275686e-06, + "loss": 0.7858, + "step": 2379 + }, + { + "epoch": 0.21463678585922352, + "grad_norm": 1.6440029428321135, + "learning_rate": 3.652994527840351e-06, + "loss": 0.762, + "step": 2380 + }, + { + "epoch": 0.21472696938269378, + "grad_norm": 1.8219803794365508, + "learning_rate": 3.6526655813624087e-06, + "loss": 0.7467, + "step": 2381 + }, + { + "epoch": 0.21481715290616404, + "grad_norm": 0.5774253529446612, + "learning_rate": 3.652336493869925e-06, + "loss": 0.6027, + "step": 2382 + }, + { + "epoch": 0.2149073364296343, + "grad_norm": 1.6477865949851986, + "learning_rate": 3.6520072653909823e-06, + "loss": 0.7865, + "step": 2383 + }, + { + "epoch": 0.21499751995310457, + "grad_norm": 1.809351916651823, + "learning_rate": 3.6516778959536702e-06, + "loss": 0.7797, + "step": 2384 + }, + { + "epoch": 0.21508770347657483, + "grad_norm": 1.7038871389766386, + "learning_rate": 3.6513483855860923e-06, + "loss": 0.8307, + "step": 2385 + }, + { + "epoch": 0.2151778870000451, + "grad_norm": 1.6042332408351243, + "learning_rate": 3.6510187343163654e-06, + "loss": 0.7284, + "step": 2386 + }, + { + "epoch": 0.21526807052351535, + "grad_norm": 1.9424650501853067, + "learning_rate": 3.650688942172616e-06, + "loss": 0.818, + "step": 2387 + }, + { + "epoch": 0.2153582540469856, + "grad_norm": 1.9280893603822575, + "learning_rate": 3.650359009182984e-06, + "loss": 0.7218, + "step": 2388 + }, + { + "epoch": 0.21544843757045587, + "grad_norm": 2.6967744704657197, + "learning_rate": 3.650028935375622e-06, + "loss": 0.866, + "step": 2389 + }, + { + "epoch": 0.21553862109392613, + "grad_norm": 1.540421481356716, + "learning_rate": 3.6496987207786926e-06, + "loss": 0.745, + "step": 2390 + }, + { + "epoch": 0.2156288046173964, + "grad_norm": 1.8270129872796408, + "learning_rate": 3.6493683654203724e-06, + "loss": 0.7333, + "step": 2391 + }, + { + "epoch": 0.21571898814086665, + "grad_norm": 2.014740269204931, + "learning_rate": 3.6490378693288484e-06, + "loss": 0.767, + "step": 2392 + }, + { + "epoch": 0.2158091716643369, + "grad_norm": 1.824716896150916, + "learning_rate": 3.648707232532321e-06, + "loss": 0.8309, + "step": 2393 + }, + { + "epoch": 0.2158993551878072, + "grad_norm": 5.127591664105451, + "learning_rate": 3.6483764550590017e-06, + "loss": 0.7921, + "step": 2394 + }, + { + "epoch": 0.21598953871127746, + "grad_norm": 1.699419031726695, + "learning_rate": 3.6480455369371133e-06, + "loss": 0.7984, + "step": 2395 + }, + { + "epoch": 0.21607972223474772, + "grad_norm": 5.608646951251193, + "learning_rate": 3.647714478194893e-06, + "loss": 0.7857, + "step": 2396 + }, + { + "epoch": 0.21616990575821798, + "grad_norm": 2.0918793933595468, + "learning_rate": 3.647383278860588e-06, + "loss": 0.7277, + "step": 2397 + }, + { + "epoch": 0.21626008928168824, + "grad_norm": 1.8359676277448231, + "learning_rate": 3.6470519389624587e-06, + "loss": 0.769, + "step": 2398 + }, + { + "epoch": 0.2163502728051585, + "grad_norm": 1.5194869243364255, + "learning_rate": 3.646720458528776e-06, + "loss": 0.8399, + "step": 2399 + }, + { + "epoch": 0.21644045632862877, + "grad_norm": 2.452738057130886, + "learning_rate": 3.6463888375878235e-06, + "loss": 0.7523, + "step": 2400 + }, + { + "epoch": 0.21653063985209903, + "grad_norm": 2.1472754092563493, + "learning_rate": 3.646057076167897e-06, + "loss": 0.7969, + "step": 2401 + }, + { + "epoch": 0.2166208233755693, + "grad_norm": 1.6554313651846264, + "learning_rate": 3.645725174297305e-06, + "loss": 0.7877, + "step": 2402 + }, + { + "epoch": 0.21671100689903955, + "grad_norm": 1.600260954676497, + "learning_rate": 3.645393132004367e-06, + "loss": 0.7647, + "step": 2403 + }, + { + "epoch": 0.2168011904225098, + "grad_norm": 1.6541645668202227, + "learning_rate": 3.6450609493174135e-06, + "loss": 0.7442, + "step": 2404 + }, + { + "epoch": 0.21689137394598007, + "grad_norm": 1.748039032476297, + "learning_rate": 3.6447286262647896e-06, + "loss": 0.7435, + "step": 2405 + }, + { + "epoch": 0.21698155746945033, + "grad_norm": 1.6858178691120174, + "learning_rate": 3.64439616287485e-06, + "loss": 0.7477, + "step": 2406 + }, + { + "epoch": 0.2170717409929206, + "grad_norm": 1.9473012540230332, + "learning_rate": 3.644063559175963e-06, + "loss": 0.7449, + "step": 2407 + }, + { + "epoch": 0.21716192451639085, + "grad_norm": 1.996438491225366, + "learning_rate": 3.6437308151965074e-06, + "loss": 0.762, + "step": 2408 + }, + { + "epoch": 0.2172521080398611, + "grad_norm": 1.3527239621472424, + "learning_rate": 3.643397930964876e-06, + "loss": 0.7941, + "step": 2409 + }, + { + "epoch": 0.21734229156333137, + "grad_norm": 2.0887296469956387, + "learning_rate": 3.6430649065094707e-06, + "loss": 0.8476, + "step": 2410 + }, + { + "epoch": 0.21743247508680164, + "grad_norm": 1.9967884862841405, + "learning_rate": 3.6427317418587086e-06, + "loss": 0.8529, + "step": 2411 + }, + { + "epoch": 0.2175226586102719, + "grad_norm": 5.010896634636142, + "learning_rate": 3.6423984370410157e-06, + "loss": 0.7202, + "step": 2412 + }, + { + "epoch": 0.21761284213374216, + "grad_norm": 1.7147628313233256, + "learning_rate": 3.6420649920848324e-06, + "loss": 0.7501, + "step": 2413 + }, + { + "epoch": 0.21770302565721242, + "grad_norm": 1.677064986355652, + "learning_rate": 3.6417314070186096e-06, + "loss": 0.7912, + "step": 2414 + }, + { + "epoch": 0.21779320918068268, + "grad_norm": 1.4553510781110797, + "learning_rate": 3.641397681870811e-06, + "loss": 0.7886, + "step": 2415 + }, + { + "epoch": 0.21788339270415294, + "grad_norm": 1.5337247507866387, + "learning_rate": 3.641063816669911e-06, + "loss": 0.7988, + "step": 2416 + }, + { + "epoch": 0.2179735762276232, + "grad_norm": 1.6445827158941335, + "learning_rate": 3.640729811444398e-06, + "loss": 0.8408, + "step": 2417 + }, + { + "epoch": 0.2180637597510935, + "grad_norm": 1.7815572908596664, + "learning_rate": 3.6403956662227706e-06, + "loss": 0.7842, + "step": 2418 + }, + { + "epoch": 0.21815394327456375, + "grad_norm": 2.213614740100529, + "learning_rate": 3.6400613810335396e-06, + "loss": 0.8373, + "step": 2419 + }, + { + "epoch": 0.218244126798034, + "grad_norm": 7.699322797051902, + "learning_rate": 3.639726955905228e-06, + "loss": 0.7006, + "step": 2420 + }, + { + "epoch": 0.21833431032150427, + "grad_norm": 1.8886230209959294, + "learning_rate": 3.639392390866372e-06, + "loss": 0.8107, + "step": 2421 + }, + { + "epoch": 0.21842449384497453, + "grad_norm": 1.6113384084662385, + "learning_rate": 3.639057685945517e-06, + "loss": 0.755, + "step": 2422 + }, + { + "epoch": 0.2185146773684448, + "grad_norm": 1.79679273334097, + "learning_rate": 3.638722841171223e-06, + "loss": 0.6791, + "step": 2423 + }, + { + "epoch": 0.21860486089191505, + "grad_norm": 1.6875043091895574, + "learning_rate": 3.638387856572061e-06, + "loss": 0.8192, + "step": 2424 + }, + { + "epoch": 0.21869504441538531, + "grad_norm": 0.6549012083432998, + "learning_rate": 3.638052732176612e-06, + "loss": 0.6084, + "step": 2425 + }, + { + "epoch": 0.21878522793885558, + "grad_norm": 1.7668056760236037, + "learning_rate": 3.637717468013472e-06, + "loss": 0.7964, + "step": 2426 + }, + { + "epoch": 0.21887541146232584, + "grad_norm": 0.6019398632729984, + "learning_rate": 3.6373820641112475e-06, + "loss": 0.5599, + "step": 2427 + }, + { + "epoch": 0.2189655949857961, + "grad_norm": 1.7719359237719323, + "learning_rate": 3.6370465204985567e-06, + "loss": 0.7939, + "step": 2428 + }, + { + "epoch": 0.21905577850926636, + "grad_norm": 0.6394753957329539, + "learning_rate": 3.6367108372040304e-06, + "loss": 0.6211, + "step": 2429 + }, + { + "epoch": 0.21914596203273662, + "grad_norm": 1.8940070931443804, + "learning_rate": 3.6363750142563107e-06, + "loss": 0.786, + "step": 2430 + }, + { + "epoch": 0.21923614555620688, + "grad_norm": 0.7317433380355434, + "learning_rate": 3.636039051684052e-06, + "loss": 0.6569, + "step": 2431 + }, + { + "epoch": 0.21932632907967714, + "grad_norm": 2.2076937871011415, + "learning_rate": 3.6357029495159203e-06, + "loss": 0.7536, + "step": 2432 + }, + { + "epoch": 0.2194165126031474, + "grad_norm": 1.5747072583575417, + "learning_rate": 3.6353667077805934e-06, + "loss": 0.7596, + "step": 2433 + }, + { + "epoch": 0.21950669612661766, + "grad_norm": 2.212993191852948, + "learning_rate": 3.6350303265067625e-06, + "loss": 0.8265, + "step": 2434 + }, + { + "epoch": 0.21959687965008792, + "grad_norm": 1.9978432947819935, + "learning_rate": 3.6346938057231285e-06, + "loss": 0.8143, + "step": 2435 + }, + { + "epoch": 0.21968706317355818, + "grad_norm": 1.561798472275575, + "learning_rate": 3.6343571454584047e-06, + "loss": 0.7288, + "step": 2436 + }, + { + "epoch": 0.21977724669702844, + "grad_norm": 1.7624163885101778, + "learning_rate": 3.6340203457413176e-06, + "loss": 0.8224, + "step": 2437 + }, + { + "epoch": 0.2198674302204987, + "grad_norm": 1.5723465987495, + "learning_rate": 3.633683406600605e-06, + "loss": 0.803, + "step": 2438 + }, + { + "epoch": 0.21995761374396897, + "grad_norm": 2.0028314812817705, + "learning_rate": 3.6333463280650165e-06, + "loss": 0.8435, + "step": 2439 + }, + { + "epoch": 0.22004779726743923, + "grad_norm": 0.7043119161293506, + "learning_rate": 3.6330091101633126e-06, + "loss": 0.5694, + "step": 2440 + }, + { + "epoch": 0.2201379807909095, + "grad_norm": 1.6906125362592572, + "learning_rate": 3.632671752924267e-06, + "loss": 0.8172, + "step": 2441 + }, + { + "epoch": 0.22022816431437978, + "grad_norm": 2.437365992848276, + "learning_rate": 3.632334256376665e-06, + "loss": 0.7907, + "step": 2442 + }, + { + "epoch": 0.22031834783785004, + "grad_norm": 1.4456815506181606, + "learning_rate": 3.6319966205493044e-06, + "loss": 0.8409, + "step": 2443 + }, + { + "epoch": 0.2204085313613203, + "grad_norm": 0.7022628797026995, + "learning_rate": 3.6316588454709922e-06, + "loss": 0.6534, + "step": 2444 + }, + { + "epoch": 0.22049871488479056, + "grad_norm": 1.5439864653206594, + "learning_rate": 3.6313209311705514e-06, + "loss": 0.7364, + "step": 2445 + }, + { + "epoch": 0.22058889840826082, + "grad_norm": 1.7198446775877336, + "learning_rate": 3.6309828776768133e-06, + "loss": 0.8259, + "step": 2446 + }, + { + "epoch": 0.22067908193173108, + "grad_norm": 1.6184719715914602, + "learning_rate": 3.630644685018623e-06, + "loss": 0.7487, + "step": 2447 + }, + { + "epoch": 0.22076926545520134, + "grad_norm": 2.243767689924716, + "learning_rate": 3.6303063532248367e-06, + "loss": 0.864, + "step": 2448 + }, + { + "epoch": 0.2208594489786716, + "grad_norm": 1.5392351392555548, + "learning_rate": 3.6299678823243236e-06, + "loss": 0.8607, + "step": 2449 + }, + { + "epoch": 0.22094963250214186, + "grad_norm": 1.733113629220941, + "learning_rate": 3.629629272345963e-06, + "loss": 0.6916, + "step": 2450 + }, + { + "epoch": 0.22103981602561212, + "grad_norm": 1.660484227712928, + "learning_rate": 3.6292905233186468e-06, + "loss": 0.9046, + "step": 2451 + }, + { + "epoch": 0.22112999954908238, + "grad_norm": 1.627854261384407, + "learning_rate": 3.6289516352712796e-06, + "loss": 0.7423, + "step": 2452 + }, + { + "epoch": 0.22122018307255265, + "grad_norm": 0.7101881668422059, + "learning_rate": 3.6286126082327764e-06, + "loss": 0.6045, + "step": 2453 + }, + { + "epoch": 0.2213103665960229, + "grad_norm": 1.559599129082749, + "learning_rate": 3.628273442232066e-06, + "loss": 0.8148, + "step": 2454 + }, + { + "epoch": 0.22140055011949317, + "grad_norm": 1.3947025482371078, + "learning_rate": 3.627934137298087e-06, + "loss": 0.7793, + "step": 2455 + }, + { + "epoch": 0.22149073364296343, + "grad_norm": 1.809847831546924, + "learning_rate": 3.627594693459792e-06, + "loss": 0.7591, + "step": 2456 + }, + { + "epoch": 0.2215809171664337, + "grad_norm": 1.571724176470495, + "learning_rate": 3.6272551107461424e-06, + "loss": 0.7443, + "step": 2457 + }, + { + "epoch": 0.22167110068990395, + "grad_norm": 2.160704172464278, + "learning_rate": 3.6269153891861137e-06, + "loss": 0.7764, + "step": 2458 + }, + { + "epoch": 0.2217612842133742, + "grad_norm": 1.9917734950104204, + "learning_rate": 3.6265755288086944e-06, + "loss": 0.7511, + "step": 2459 + }, + { + "epoch": 0.22185146773684447, + "grad_norm": 2.222356524647664, + "learning_rate": 3.626235529642881e-06, + "loss": 0.7674, + "step": 2460 + }, + { + "epoch": 0.22194165126031473, + "grad_norm": 9.723965830596557, + "learning_rate": 3.625895391717686e-06, + "loss": 0.7156, + "step": 2461 + }, + { + "epoch": 0.222031834783785, + "grad_norm": 1.456223169169655, + "learning_rate": 3.625555115062131e-06, + "loss": 0.7557, + "step": 2462 + }, + { + "epoch": 0.22212201830725525, + "grad_norm": 1.8038971726828104, + "learning_rate": 3.6252146997052507e-06, + "loss": 0.8559, + "step": 2463 + }, + { + "epoch": 0.22221220183072551, + "grad_norm": 0.6586762717101058, + "learning_rate": 3.6248741456760898e-06, + "loss": 0.5603, + "step": 2464 + }, + { + "epoch": 0.22230238535419578, + "grad_norm": 2.4366293110138835, + "learning_rate": 3.624533453003708e-06, + "loss": 0.7773, + "step": 2465 + }, + { + "epoch": 0.22239256887766606, + "grad_norm": 1.8830968238502754, + "learning_rate": 3.6241926217171745e-06, + "loss": 0.7928, + "step": 2466 + }, + { + "epoch": 0.22248275240113632, + "grad_norm": 2.1050448583684407, + "learning_rate": 3.6238516518455703e-06, + "loss": 0.6987, + "step": 2467 + }, + { + "epoch": 0.22257293592460659, + "grad_norm": 3.38193153380624, + "learning_rate": 3.62351054341799e-06, + "loss": 0.9089, + "step": 2468 + }, + { + "epoch": 0.22266311944807685, + "grad_norm": 1.776175649591533, + "learning_rate": 3.623169296463538e-06, + "loss": 0.8111, + "step": 2469 + }, + { + "epoch": 0.2227533029715471, + "grad_norm": 1.5697559490160997, + "learning_rate": 3.6228279110113316e-06, + "loss": 0.7858, + "step": 2470 + }, + { + "epoch": 0.22284348649501737, + "grad_norm": 1.8604695920973544, + "learning_rate": 3.6224863870904994e-06, + "loss": 0.7895, + "step": 2471 + }, + { + "epoch": 0.22293367001848763, + "grad_norm": 1.8299140978803725, + "learning_rate": 3.6221447247301827e-06, + "loss": 0.7946, + "step": 2472 + }, + { + "epoch": 0.2230238535419579, + "grad_norm": 1.554926714872134, + "learning_rate": 3.6218029239595332e-06, + "loss": 0.7343, + "step": 2473 + }, + { + "epoch": 0.22311403706542815, + "grad_norm": 1.592889291559842, + "learning_rate": 3.621460984807716e-06, + "loss": 0.8084, + "step": 2474 + }, + { + "epoch": 0.2232042205888984, + "grad_norm": 2.1881117918895625, + "learning_rate": 3.621118907303907e-06, + "loss": 0.7592, + "step": 2475 + }, + { + "epoch": 0.22329440411236867, + "grad_norm": 2.0679859836553462, + "learning_rate": 3.620776691477294e-06, + "loss": 0.7847, + "step": 2476 + }, + { + "epoch": 0.22338458763583893, + "grad_norm": 0.6747266666339614, + "learning_rate": 3.6204343373570765e-06, + "loss": 0.551, + "step": 2477 + }, + { + "epoch": 0.2234747711593092, + "grad_norm": 1.5076500997893296, + "learning_rate": 3.620091844972467e-06, + "loss": 0.8075, + "step": 2478 + }, + { + "epoch": 0.22356495468277945, + "grad_norm": 1.5988209403422053, + "learning_rate": 3.619749214352688e-06, + "loss": 0.8108, + "step": 2479 + }, + { + "epoch": 0.22365513820624972, + "grad_norm": 1.9511202479836232, + "learning_rate": 3.6194064455269744e-06, + "loss": 0.8899, + "step": 2480 + }, + { + "epoch": 0.22374532172971998, + "grad_norm": 2.1045260610310432, + "learning_rate": 3.6190635385245737e-06, + "loss": 0.7069, + "step": 2481 + }, + { + "epoch": 0.22383550525319024, + "grad_norm": 1.822931656548767, + "learning_rate": 3.618720493374745e-06, + "loss": 0.7193, + "step": 2482 + }, + { + "epoch": 0.2239256887766605, + "grad_norm": 2.337113100559301, + "learning_rate": 3.6183773101067575e-06, + "loss": 0.6424, + "step": 2483 + }, + { + "epoch": 0.22401587230013076, + "grad_norm": 1.504187461025957, + "learning_rate": 3.6180339887498948e-06, + "loss": 0.7075, + "step": 2484 + }, + { + "epoch": 0.22410605582360102, + "grad_norm": 1.810678619233421, + "learning_rate": 3.61769052933345e-06, + "loss": 0.786, + "step": 2485 + }, + { + "epoch": 0.22419623934707128, + "grad_norm": 8.78483188110047, + "learning_rate": 3.6173469318867297e-06, + "loss": 0.9383, + "step": 2486 + }, + { + "epoch": 0.22428642287054154, + "grad_norm": 1.6840125180456493, + "learning_rate": 3.617003196439051e-06, + "loss": 0.8094, + "step": 2487 + }, + { + "epoch": 0.2243766063940118, + "grad_norm": 1.6011339312019974, + "learning_rate": 3.616659323019744e-06, + "loss": 0.7923, + "step": 2488 + }, + { + "epoch": 0.22446678991748206, + "grad_norm": 1.8847352020496388, + "learning_rate": 3.616315311658149e-06, + "loss": 0.8135, + "step": 2489 + }, + { + "epoch": 0.22455697344095235, + "grad_norm": 1.885967033806039, + "learning_rate": 3.6159711623836195e-06, + "loss": 0.844, + "step": 2490 + }, + { + "epoch": 0.2246471569644226, + "grad_norm": 0.8258303394000318, + "learning_rate": 3.6156268752255203e-06, + "loss": 0.5651, + "step": 2491 + }, + { + "epoch": 0.22473734048789287, + "grad_norm": 1.4774082728537345, + "learning_rate": 3.615282450213227e-06, + "loss": 0.7479, + "step": 2492 + }, + { + "epoch": 0.22482752401136313, + "grad_norm": 1.9174424204828966, + "learning_rate": 3.614937887376128e-06, + "loss": 0.7978, + "step": 2493 + }, + { + "epoch": 0.2249177075348334, + "grad_norm": 2.1594688604909154, + "learning_rate": 3.614593186743625e-06, + "loss": 0.7997, + "step": 2494 + }, + { + "epoch": 0.22500789105830366, + "grad_norm": 2.045983273151595, + "learning_rate": 3.614248348345128e-06, + "loss": 0.7269, + "step": 2495 + }, + { + "epoch": 0.22509807458177392, + "grad_norm": 1.7422910924656103, + "learning_rate": 3.6139033722100614e-06, + "loss": 0.8364, + "step": 2496 + }, + { + "epoch": 0.22518825810524418, + "grad_norm": 2.165307523457714, + "learning_rate": 3.6135582583678596e-06, + "loss": 0.8608, + "step": 2497 + }, + { + "epoch": 0.22527844162871444, + "grad_norm": 1.8198457700754678, + "learning_rate": 3.61321300684797e-06, + "loss": 0.8048, + "step": 2498 + }, + { + "epoch": 0.2253686251521847, + "grad_norm": 1.2230972128602977, + "learning_rate": 3.6128676176798527e-06, + "loss": 0.8079, + "step": 2499 + }, + { + "epoch": 0.22545880867565496, + "grad_norm": 1.8355395595991701, + "learning_rate": 3.612522090892976e-06, + "loss": 0.8058, + "step": 2500 + }, + { + "epoch": 0.22554899219912522, + "grad_norm": 1.3623958652843513, + "learning_rate": 3.6121764265168232e-06, + "loss": 0.7468, + "step": 2501 + }, + { + "epoch": 0.22563917572259548, + "grad_norm": 1.4569949229930217, + "learning_rate": 3.611830624580888e-06, + "loss": 0.8121, + "step": 2502 + }, + { + "epoch": 0.22572935924606574, + "grad_norm": 1.7520841041348547, + "learning_rate": 3.6114846851146767e-06, + "loss": 0.795, + "step": 2503 + }, + { + "epoch": 0.225819542769536, + "grad_norm": 1.647333001108315, + "learning_rate": 3.6111386081477068e-06, + "loss": 0.6835, + "step": 2504 + }, + { + "epoch": 0.22590972629300626, + "grad_norm": 2.442032536857725, + "learning_rate": 3.6107923937095066e-06, + "loss": 0.8046, + "step": 2505 + }, + { + "epoch": 0.22599990981647652, + "grad_norm": 3.910553171779405, + "learning_rate": 3.6104460418296173e-06, + "loss": 0.7162, + "step": 2506 + }, + { + "epoch": 0.22609009333994678, + "grad_norm": 1.8922867841054567, + "learning_rate": 3.6100995525375924e-06, + "loss": 0.8363, + "step": 2507 + }, + { + "epoch": 0.22618027686341705, + "grad_norm": 2.0563052230886614, + "learning_rate": 3.6097529258629952e-06, + "loss": 0.7299, + "step": 2508 + }, + { + "epoch": 0.2262704603868873, + "grad_norm": 1.70081018047408, + "learning_rate": 3.6094061618354027e-06, + "loss": 0.816, + "step": 2509 + }, + { + "epoch": 0.22636064391035757, + "grad_norm": 1.9378741733657776, + "learning_rate": 3.609059260484402e-06, + "loss": 0.7825, + "step": 2510 + }, + { + "epoch": 0.22645082743382783, + "grad_norm": 2.8892730025520894, + "learning_rate": 3.6087122218395935e-06, + "loss": 0.7307, + "step": 2511 + }, + { + "epoch": 0.2265410109572981, + "grad_norm": 1.857194086300967, + "learning_rate": 3.608365045930587e-06, + "loss": 0.7537, + "step": 2512 + }, + { + "epoch": 0.22663119448076835, + "grad_norm": 1.8718700351553719, + "learning_rate": 3.608017732787007e-06, + "loss": 0.7756, + "step": 2513 + }, + { + "epoch": 0.22672137800423864, + "grad_norm": 1.589232231085208, + "learning_rate": 3.6076702824384875e-06, + "loss": 0.7156, + "step": 2514 + }, + { + "epoch": 0.2268115615277089, + "grad_norm": 1.6209641738029517, + "learning_rate": 3.607322694914675e-06, + "loss": 0.8008, + "step": 2515 + }, + { + "epoch": 0.22690174505117916, + "grad_norm": 2.0001356555709005, + "learning_rate": 3.606974970245227e-06, + "loss": 0.7579, + "step": 2516 + }, + { + "epoch": 0.22699192857464942, + "grad_norm": 1.9578139515937605, + "learning_rate": 3.606627108459814e-06, + "loss": 0.7359, + "step": 2517 + }, + { + "epoch": 0.22708211209811968, + "grad_norm": 1.3586419858699854, + "learning_rate": 3.6062791095881174e-06, + "loss": 0.7127, + "step": 2518 + }, + { + "epoch": 0.22717229562158994, + "grad_norm": 1.869995814436038, + "learning_rate": 3.6059309736598303e-06, + "loss": 0.8607, + "step": 2519 + }, + { + "epoch": 0.2272624791450602, + "grad_norm": 1.8695601069339067, + "learning_rate": 3.605582700704657e-06, + "loss": 0.7605, + "step": 2520 + }, + { + "epoch": 0.22735266266853046, + "grad_norm": 1.8518383609315636, + "learning_rate": 3.6052342907523146e-06, + "loss": 0.7875, + "step": 2521 + }, + { + "epoch": 0.22744284619200072, + "grad_norm": 1.9951739496866072, + "learning_rate": 3.604885743832532e-06, + "loss": 0.7186, + "step": 2522 + }, + { + "epoch": 0.22753302971547099, + "grad_norm": 0.6920781077104031, + "learning_rate": 3.6045370599750482e-06, + "loss": 0.5812, + "step": 2523 + }, + { + "epoch": 0.22762321323894125, + "grad_norm": 1.6079585684662565, + "learning_rate": 3.604188239209615e-06, + "loss": 0.7723, + "step": 2524 + }, + { + "epoch": 0.2277133967624115, + "grad_norm": 2.927347372687577, + "learning_rate": 3.603839281565996e-06, + "loss": 0.7095, + "step": 2525 + }, + { + "epoch": 0.22780358028588177, + "grad_norm": 1.7530988776288838, + "learning_rate": 3.603490187073966e-06, + "loss": 0.8, + "step": 2526 + }, + { + "epoch": 0.22789376380935203, + "grad_norm": 1.818054567023814, + "learning_rate": 3.6031409557633117e-06, + "loss": 0.8507, + "step": 2527 + }, + { + "epoch": 0.2279839473328223, + "grad_norm": 1.8055398752884728, + "learning_rate": 3.602791587663831e-06, + "loss": 0.8244, + "step": 2528 + }, + { + "epoch": 0.22807413085629255, + "grad_norm": 2.1704084535813064, + "learning_rate": 3.6024420828053348e-06, + "loss": 0.8233, + "step": 2529 + }, + { + "epoch": 0.2281643143797628, + "grad_norm": 1.5203321841809267, + "learning_rate": 3.6020924412176445e-06, + "loss": 0.859, + "step": 2530 + }, + { + "epoch": 0.22825449790323307, + "grad_norm": 1.5847324998800736, + "learning_rate": 3.601742662930593e-06, + "loss": 0.7588, + "step": 2531 + }, + { + "epoch": 0.22834468142670333, + "grad_norm": 1.8484708188069827, + "learning_rate": 3.6013927479740248e-06, + "loss": 0.7491, + "step": 2532 + }, + { + "epoch": 0.2284348649501736, + "grad_norm": 1.8670232213746625, + "learning_rate": 3.6010426963777985e-06, + "loss": 0.8596, + "step": 2533 + }, + { + "epoch": 0.22852504847364385, + "grad_norm": 1.6449445264095077, + "learning_rate": 3.6006925081717804e-06, + "loss": 0.8331, + "step": 2534 + }, + { + "epoch": 0.22861523199711412, + "grad_norm": 1.469184405428376, + "learning_rate": 3.600342183385852e-06, + "loss": 0.7517, + "step": 2535 + }, + { + "epoch": 0.22870541552058438, + "grad_norm": 1.4923573342384926, + "learning_rate": 3.5999917220499043e-06, + "loss": 0.7256, + "step": 2536 + }, + { + "epoch": 0.22879559904405466, + "grad_norm": 1.753588539353044, + "learning_rate": 3.5996411241938404e-06, + "loss": 0.7905, + "step": 2537 + }, + { + "epoch": 0.22888578256752493, + "grad_norm": 1.6548084967647967, + "learning_rate": 3.5992903898475752e-06, + "loss": 0.7457, + "step": 2538 + }, + { + "epoch": 0.2289759660909952, + "grad_norm": 1.4471714745912896, + "learning_rate": 3.5989395190410365e-06, + "loss": 0.8856, + "step": 2539 + }, + { + "epoch": 0.22906614961446545, + "grad_norm": 2.079528105464905, + "learning_rate": 3.598588511804161e-06, + "loss": 0.7915, + "step": 2540 + }, + { + "epoch": 0.2291563331379357, + "grad_norm": 2.1656387042428045, + "learning_rate": 3.5982373681668987e-06, + "loss": 0.7917, + "step": 2541 + }, + { + "epoch": 0.22924651666140597, + "grad_norm": 4.1544155575577575, + "learning_rate": 3.597886088159212e-06, + "loss": 0.7889, + "step": 2542 + }, + { + "epoch": 0.22933670018487623, + "grad_norm": 1.5314116100742527, + "learning_rate": 3.597534671811074e-06, + "loss": 0.7153, + "step": 2543 + }, + { + "epoch": 0.2294268837083465, + "grad_norm": 1.6401538717378576, + "learning_rate": 3.5971831191524684e-06, + "loss": 0.7657, + "step": 2544 + }, + { + "epoch": 0.22951706723181675, + "grad_norm": 1.668762089831959, + "learning_rate": 3.5968314302133925e-06, + "loss": 0.8324, + "step": 2545 + }, + { + "epoch": 0.229607250755287, + "grad_norm": 1.8611172119281028, + "learning_rate": 3.596479605023854e-06, + "loss": 0.7859, + "step": 2546 + }, + { + "epoch": 0.22969743427875727, + "grad_norm": 5.737417635191417, + "learning_rate": 3.596127643613873e-06, + "loss": 0.8002, + "step": 2547 + }, + { + "epoch": 0.22978761780222753, + "grad_norm": 1.5230865856863025, + "learning_rate": 3.59577554601348e-06, + "loss": 0.8075, + "step": 2548 + }, + { + "epoch": 0.2298778013256978, + "grad_norm": 1.46292175341339, + "learning_rate": 3.595423312252719e-06, + "loss": 0.8328, + "step": 2549 + }, + { + "epoch": 0.22996798484916806, + "grad_norm": 2.525004369901898, + "learning_rate": 3.5950709423616436e-06, + "loss": 0.7732, + "step": 2550 + }, + { + "epoch": 0.23005816837263832, + "grad_norm": 2.3400611840098686, + "learning_rate": 3.5947184363703203e-06, + "loss": 0.7679, + "step": 2551 + }, + { + "epoch": 0.23014835189610858, + "grad_norm": 1.5607707563661015, + "learning_rate": 3.5943657943088274e-06, + "loss": 0.7378, + "step": 2552 + }, + { + "epoch": 0.23023853541957884, + "grad_norm": 1.926741107417763, + "learning_rate": 3.5940130162072525e-06, + "loss": 0.7103, + "step": 2553 + }, + { + "epoch": 0.2303287189430491, + "grad_norm": 1.9088030119143926, + "learning_rate": 3.5936601020956985e-06, + "loss": 0.7858, + "step": 2554 + }, + { + "epoch": 0.23041890246651936, + "grad_norm": 1.6462442713980328, + "learning_rate": 3.5933070520042772e-06, + "loss": 0.8206, + "step": 2555 + }, + { + "epoch": 0.23050908598998962, + "grad_norm": 0.7627721300893873, + "learning_rate": 3.5929538659631133e-06, + "loss": 0.6309, + "step": 2556 + }, + { + "epoch": 0.23059926951345988, + "grad_norm": 2.3181569783176994, + "learning_rate": 3.592600544002341e-06, + "loss": 0.8156, + "step": 2557 + }, + { + "epoch": 0.23068945303693014, + "grad_norm": 2.5070819681992402, + "learning_rate": 3.5922470861521098e-06, + "loss": 0.7985, + "step": 2558 + }, + { + "epoch": 0.2307796365604004, + "grad_norm": 2.082706649620345, + "learning_rate": 3.591893492442577e-06, + "loss": 0.7516, + "step": 2559 + }, + { + "epoch": 0.23086982008387066, + "grad_norm": 1.9292660766540726, + "learning_rate": 3.591539762903914e-06, + "loss": 0.8484, + "step": 2560 + }, + { + "epoch": 0.23096000360734095, + "grad_norm": 1.6781836586473122, + "learning_rate": 3.591185897566303e-06, + "loss": 0.7151, + "step": 2561 + }, + { + "epoch": 0.2310501871308112, + "grad_norm": 2.2821595978288105, + "learning_rate": 3.590831896459937e-06, + "loss": 0.7844, + "step": 2562 + }, + { + "epoch": 0.23114037065428147, + "grad_norm": 2.0638656718729647, + "learning_rate": 3.5904777596150222e-06, + "loss": 0.7944, + "step": 2563 + }, + { + "epoch": 0.23123055417775173, + "grad_norm": 1.8223873641551482, + "learning_rate": 3.590123487061775e-06, + "loss": 0.7806, + "step": 2564 + }, + { + "epoch": 0.231320737701222, + "grad_norm": 2.0745868823139917, + "learning_rate": 3.589769078830424e-06, + "loss": 0.8837, + "step": 2565 + }, + { + "epoch": 0.23141092122469226, + "grad_norm": 1.9587714401853689, + "learning_rate": 3.58941453495121e-06, + "loss": 0.7655, + "step": 2566 + }, + { + "epoch": 0.23150110474816252, + "grad_norm": 1.6619638453245287, + "learning_rate": 3.5890598554543834e-06, + "loss": 0.7981, + "step": 2567 + }, + { + "epoch": 0.23159128827163278, + "grad_norm": 1.7437213936376879, + "learning_rate": 3.5887050403702073e-06, + "loss": 0.8373, + "step": 2568 + }, + { + "epoch": 0.23168147179510304, + "grad_norm": 1.845829260012895, + "learning_rate": 3.588350089728958e-06, + "loss": 0.7581, + "step": 2569 + }, + { + "epoch": 0.2317716553185733, + "grad_norm": 1.832827787343667, + "learning_rate": 3.5879950035609204e-06, + "loss": 0.8108, + "step": 2570 + }, + { + "epoch": 0.23186183884204356, + "grad_norm": 1.8386236141914625, + "learning_rate": 3.5876397818963933e-06, + "loss": 0.7821, + "step": 2571 + }, + { + "epoch": 0.23195202236551382, + "grad_norm": 1.6189730457420703, + "learning_rate": 3.5872844247656858e-06, + "loss": 0.8369, + "step": 2572 + }, + { + "epoch": 0.23204220588898408, + "grad_norm": 2.3074000026737598, + "learning_rate": 3.5869289321991195e-06, + "loss": 0.7893, + "step": 2573 + }, + { + "epoch": 0.23213238941245434, + "grad_norm": 1.5424860331986527, + "learning_rate": 3.5865733042270263e-06, + "loss": 0.78, + "step": 2574 + }, + { + "epoch": 0.2322225729359246, + "grad_norm": 1.784737235211049, + "learning_rate": 3.5862175408797498e-06, + "loss": 0.8147, + "step": 2575 + }, + { + "epoch": 0.23231275645939486, + "grad_norm": 8.848775884593401, + "learning_rate": 3.585861642187647e-06, + "loss": 0.7258, + "step": 2576 + }, + { + "epoch": 0.23240293998286513, + "grad_norm": 1.4157032589196346, + "learning_rate": 3.5855056081810845e-06, + "loss": 0.7816, + "step": 2577 + }, + { + "epoch": 0.2324931235063354, + "grad_norm": 1.5888716141370036, + "learning_rate": 3.5851494388904406e-06, + "loss": 0.784, + "step": 2578 + }, + { + "epoch": 0.23258330702980565, + "grad_norm": 1.5908693197158998, + "learning_rate": 3.5847931343461064e-06, + "loss": 0.7219, + "step": 2579 + }, + { + "epoch": 0.2326734905532759, + "grad_norm": 2.5935564083792015, + "learning_rate": 3.5844366945784835e-06, + "loss": 0.7148, + "step": 2580 + }, + { + "epoch": 0.23276367407674617, + "grad_norm": 1.8382200296553586, + "learning_rate": 3.5840801196179856e-06, + "loss": 0.846, + "step": 2581 + }, + { + "epoch": 0.23285385760021643, + "grad_norm": 1.6966831977028916, + "learning_rate": 3.583723409495037e-06, + "loss": 0.8372, + "step": 2582 + }, + { + "epoch": 0.2329440411236867, + "grad_norm": 4.011748702876253, + "learning_rate": 3.5833665642400747e-06, + "loss": 0.8378, + "step": 2583 + }, + { + "epoch": 0.23303422464715695, + "grad_norm": 2.533173571953191, + "learning_rate": 3.5830095838835472e-06, + "loss": 0.7112, + "step": 2584 + }, + { + "epoch": 0.23312440817062724, + "grad_norm": 1.8051526679782135, + "learning_rate": 3.5826524684559125e-06, + "loss": 0.7552, + "step": 2585 + }, + { + "epoch": 0.2332145916940975, + "grad_norm": 2.296416983059263, + "learning_rate": 3.5822952179876433e-06, + "loss": 0.8688, + "step": 2586 + }, + { + "epoch": 0.23330477521756776, + "grad_norm": 1.5288180287306674, + "learning_rate": 3.5819378325092205e-06, + "loss": 0.7616, + "step": 2587 + }, + { + "epoch": 0.23339495874103802, + "grad_norm": 1.8666165169155458, + "learning_rate": 3.581580312051139e-06, + "loss": 0.8072, + "step": 2588 + }, + { + "epoch": 0.23348514226450828, + "grad_norm": 1.6438148369711414, + "learning_rate": 3.5812226566439057e-06, + "loss": 0.7137, + "step": 2589 + }, + { + "epoch": 0.23357532578797854, + "grad_norm": 1.7034695478023425, + "learning_rate": 3.580864866318036e-06, + "loss": 0.7618, + "step": 2590 + }, + { + "epoch": 0.2336655093114488, + "grad_norm": 1.9109562795719066, + "learning_rate": 3.580506941104059e-06, + "loss": 0.7894, + "step": 2591 + }, + { + "epoch": 0.23375569283491907, + "grad_norm": 1.5285170943352813, + "learning_rate": 3.580148881032515e-06, + "loss": 0.7735, + "step": 2592 + }, + { + "epoch": 0.23384587635838933, + "grad_norm": 1.4981279612334089, + "learning_rate": 3.5797906861339556e-06, + "loss": 0.7921, + "step": 2593 + }, + { + "epoch": 0.2339360598818596, + "grad_norm": 2.1702415855691934, + "learning_rate": 3.5794323564389435e-06, + "loss": 0.8131, + "step": 2594 + }, + { + "epoch": 0.23402624340532985, + "grad_norm": 2.177950587466137, + "learning_rate": 3.579073891978055e-06, + "loss": 0.797, + "step": 2595 + }, + { + "epoch": 0.2341164269288001, + "grad_norm": 2.0788693349858667, + "learning_rate": 3.5787152927818746e-06, + "loss": 0.7563, + "step": 2596 + }, + { + "epoch": 0.23420661045227037, + "grad_norm": 1.5505138714418027, + "learning_rate": 3.5783565588810003e-06, + "loss": 0.8482, + "step": 2597 + }, + { + "epoch": 0.23429679397574063, + "grad_norm": 1.6371933424252516, + "learning_rate": 3.5779976903060412e-06, + "loss": 0.7565, + "step": 2598 + }, + { + "epoch": 0.2343869774992109, + "grad_norm": 1.748020142079106, + "learning_rate": 3.577638687087619e-06, + "loss": 0.8211, + "step": 2599 + }, + { + "epoch": 0.23447716102268115, + "grad_norm": 1.9470088563706864, + "learning_rate": 3.577279549256364e-06, + "loss": 0.7939, + "step": 2600 + }, + { + "epoch": 0.2345673445461514, + "grad_norm": 1.620086283328665, + "learning_rate": 3.5769202768429213e-06, + "loss": 0.8304, + "step": 2601 + }, + { + "epoch": 0.23465752806962167, + "grad_norm": 1.7808238406442285, + "learning_rate": 3.5765608698779454e-06, + "loss": 0.7637, + "step": 2602 + }, + { + "epoch": 0.23474771159309193, + "grad_norm": 1.808031691020277, + "learning_rate": 3.5762013283921033e-06, + "loss": 0.7985, + "step": 2603 + }, + { + "epoch": 0.2348378951165622, + "grad_norm": 1.80721842006332, + "learning_rate": 3.5758416524160728e-06, + "loss": 0.7907, + "step": 2604 + }, + { + "epoch": 0.23492807864003246, + "grad_norm": 1.5700050926125901, + "learning_rate": 3.5754818419805427e-06, + "loss": 0.7809, + "step": 2605 + }, + { + "epoch": 0.23501826216350272, + "grad_norm": 1.6776073261157411, + "learning_rate": 3.575121897116216e-06, + "loss": 0.7632, + "step": 2606 + }, + { + "epoch": 0.23510844568697298, + "grad_norm": 1.473549812057435, + "learning_rate": 3.574761817853803e-06, + "loss": 0.8193, + "step": 2607 + }, + { + "epoch": 0.23519862921044324, + "grad_norm": 21.360498695981224, + "learning_rate": 3.5744016042240287e-06, + "loss": 0.7232, + "step": 2608 + }, + { + "epoch": 0.23528881273391353, + "grad_norm": 2.1367573568981344, + "learning_rate": 3.5740412562576286e-06, + "loss": 0.7207, + "step": 2609 + }, + { + "epoch": 0.2353789962573838, + "grad_norm": 2.1181785825703914, + "learning_rate": 3.573680773985349e-06, + "loss": 0.7859, + "step": 2610 + }, + { + "epoch": 0.23546917978085405, + "grad_norm": 1.8825509297243412, + "learning_rate": 3.5733201574379486e-06, + "loss": 0.7825, + "step": 2611 + }, + { + "epoch": 0.2355593633043243, + "grad_norm": 1.684272859268469, + "learning_rate": 3.5729594066461975e-06, + "loss": 0.7236, + "step": 2612 + }, + { + "epoch": 0.23564954682779457, + "grad_norm": 1.4158712379241911, + "learning_rate": 3.572598521640876e-06, + "loss": 0.7994, + "step": 2613 + }, + { + "epoch": 0.23573973035126483, + "grad_norm": 0.7822531934912534, + "learning_rate": 3.5722375024527782e-06, + "loss": 0.6365, + "step": 2614 + }, + { + "epoch": 0.2358299138747351, + "grad_norm": 1.3724646602312496, + "learning_rate": 3.571876349112707e-06, + "loss": 0.8068, + "step": 2615 + }, + { + "epoch": 0.23592009739820535, + "grad_norm": 1.604724646624007, + "learning_rate": 3.5715150616514784e-06, + "loss": 0.7821, + "step": 2616 + }, + { + "epoch": 0.2360102809216756, + "grad_norm": 2.163819557627764, + "learning_rate": 3.5711536400999196e-06, + "loss": 0.8388, + "step": 2617 + }, + { + "epoch": 0.23610046444514587, + "grad_norm": 0.7185124128990965, + "learning_rate": 3.570792084488869e-06, + "loss": 0.6378, + "step": 2618 + }, + { + "epoch": 0.23619064796861614, + "grad_norm": 0.7176442555547073, + "learning_rate": 3.5704303948491764e-06, + "loss": 0.6494, + "step": 2619 + }, + { + "epoch": 0.2362808314920864, + "grad_norm": 2.0073842582733805, + "learning_rate": 3.5700685712117035e-06, + "loss": 0.8116, + "step": 2620 + }, + { + "epoch": 0.23637101501555666, + "grad_norm": 2.1175491756057045, + "learning_rate": 3.5697066136073227e-06, + "loss": 0.7759, + "step": 2621 + }, + { + "epoch": 0.23646119853902692, + "grad_norm": 1.458772983582437, + "learning_rate": 3.5693445220669184e-06, + "loss": 0.7985, + "step": 2622 + }, + { + "epoch": 0.23655138206249718, + "grad_norm": 1.3658541471855523, + "learning_rate": 3.568982296621386e-06, + "loss": 0.7815, + "step": 2623 + }, + { + "epoch": 0.23664156558596744, + "grad_norm": 1.7863963554760272, + "learning_rate": 3.5686199373016325e-06, + "loss": 0.8598, + "step": 2624 + }, + { + "epoch": 0.2367317491094377, + "grad_norm": 2.1486647607075073, + "learning_rate": 3.568257444138577e-06, + "loss": 0.7357, + "step": 2625 + }, + { + "epoch": 0.23682193263290796, + "grad_norm": 2.35166358096837, + "learning_rate": 3.5678948171631495e-06, + "loss": 0.7914, + "step": 2626 + }, + { + "epoch": 0.23691211615637822, + "grad_norm": 1.5659417203238457, + "learning_rate": 3.5675320564062908e-06, + "loss": 0.795, + "step": 2627 + }, + { + "epoch": 0.23700229967984848, + "grad_norm": 1.819395169795312, + "learning_rate": 3.5671691618989533e-06, + "loss": 0.6986, + "step": 2628 + }, + { + "epoch": 0.23709248320331874, + "grad_norm": 1.7357118283051998, + "learning_rate": 3.5668061336721024e-06, + "loss": 0.7979, + "step": 2629 + }, + { + "epoch": 0.237182666726789, + "grad_norm": 1.6411173490447206, + "learning_rate": 3.5664429717567117e-06, + "loss": 0.7828, + "step": 2630 + }, + { + "epoch": 0.23727285025025927, + "grad_norm": 1.7209313855011492, + "learning_rate": 3.56607967618377e-06, + "loss": 0.8169, + "step": 2631 + }, + { + "epoch": 0.23736303377372953, + "grad_norm": 2.132832286030304, + "learning_rate": 3.5657162469842754e-06, + "loss": 0.7984, + "step": 2632 + }, + { + "epoch": 0.23745321729719981, + "grad_norm": 1.8044074695864674, + "learning_rate": 3.5653526841892374e-06, + "loss": 0.79, + "step": 2633 + }, + { + "epoch": 0.23754340082067008, + "grad_norm": 2.062780418984483, + "learning_rate": 3.564988987829676e-06, + "loss": 0.8181, + "step": 2634 + }, + { + "epoch": 0.23763358434414034, + "grad_norm": 1.656981324734654, + "learning_rate": 3.564625157936626e-06, + "loss": 0.7333, + "step": 2635 + }, + { + "epoch": 0.2377237678676106, + "grad_norm": 1.7335132141797944, + "learning_rate": 3.56426119454113e-06, + "loss": 0.8249, + "step": 2636 + }, + { + "epoch": 0.23781395139108086, + "grad_norm": 1.9311203058143644, + "learning_rate": 3.5638970976742436e-06, + "loss": 0.8636, + "step": 2637 + }, + { + "epoch": 0.23790413491455112, + "grad_norm": 2.0009165094392576, + "learning_rate": 3.5635328673670335e-06, + "loss": 0.8114, + "step": 2638 + }, + { + "epoch": 0.23799431843802138, + "grad_norm": 1.4780246821550613, + "learning_rate": 3.5631685036505783e-06, + "loss": 0.8785, + "step": 2639 + }, + { + "epoch": 0.23808450196149164, + "grad_norm": 1.3668394245110749, + "learning_rate": 3.562804006555966e-06, + "loss": 0.8, + "step": 2640 + }, + { + "epoch": 0.2381746854849619, + "grad_norm": 2.8031411331035705, + "learning_rate": 3.5624393761143e-06, + "loss": 0.727, + "step": 2641 + }, + { + "epoch": 0.23826486900843216, + "grad_norm": 5.072261300597406, + "learning_rate": 3.5620746123566906e-06, + "loss": 0.703, + "step": 2642 + }, + { + "epoch": 0.23835505253190242, + "grad_norm": 1.7398233152987206, + "learning_rate": 3.5617097153142623e-06, + "loss": 0.7408, + "step": 2643 + }, + { + "epoch": 0.23844523605537268, + "grad_norm": 1.647791008030521, + "learning_rate": 3.5613446850181497e-06, + "loss": 0.8245, + "step": 2644 + }, + { + "epoch": 0.23853541957884294, + "grad_norm": 0.6105868198948975, + "learning_rate": 3.5609795214994996e-06, + "loss": 0.5374, + "step": 2645 + }, + { + "epoch": 0.2386256031023132, + "grad_norm": 1.6570398858334092, + "learning_rate": 3.560614224789469e-06, + "loss": 0.8436, + "step": 2646 + }, + { + "epoch": 0.23871578662578347, + "grad_norm": 1.8392840463469373, + "learning_rate": 3.5602487949192285e-06, + "loss": 0.7329, + "step": 2647 + }, + { + "epoch": 0.23880597014925373, + "grad_norm": 3.377496290677827, + "learning_rate": 3.559883231919957e-06, + "loss": 0.7319, + "step": 2648 + }, + { + "epoch": 0.238896153672724, + "grad_norm": 1.7819253410900446, + "learning_rate": 3.5595175358228473e-06, + "loss": 0.7608, + "step": 2649 + }, + { + "epoch": 0.23898633719619425, + "grad_norm": 1.7320777181363511, + "learning_rate": 3.5591517066591027e-06, + "loss": 0.7852, + "step": 2650 + }, + { + "epoch": 0.2390765207196645, + "grad_norm": 2.238691096039453, + "learning_rate": 3.5587857444599364e-06, + "loss": 0.7793, + "step": 2651 + }, + { + "epoch": 0.23916670424313477, + "grad_norm": 1.6042190477150424, + "learning_rate": 3.5584196492565766e-06, + "loss": 0.7577, + "step": 2652 + }, + { + "epoch": 0.23925688776660503, + "grad_norm": 3.0803914512477015, + "learning_rate": 3.5580534210802587e-06, + "loss": 0.852, + "step": 2653 + }, + { + "epoch": 0.2393470712900753, + "grad_norm": 1.5355986523104868, + "learning_rate": 3.557687059962232e-06, + "loss": 0.8312, + "step": 2654 + }, + { + "epoch": 0.23943725481354555, + "grad_norm": 1.5096711083519005, + "learning_rate": 3.5573205659337558e-06, + "loss": 0.7808, + "step": 2655 + }, + { + "epoch": 0.23952743833701584, + "grad_norm": 1.5790846522362956, + "learning_rate": 3.5569539390261025e-06, + "loss": 0.7745, + "step": 2656 + }, + { + "epoch": 0.2396176218604861, + "grad_norm": 1.2912469099807327, + "learning_rate": 3.5565871792705543e-06, + "loss": 0.8001, + "step": 2657 + }, + { + "epoch": 0.23970780538395636, + "grad_norm": 1.4787719181916137, + "learning_rate": 3.5562202866984045e-06, + "loss": 0.7775, + "step": 2658 + }, + { + "epoch": 0.23979798890742662, + "grad_norm": 1.8825022968212366, + "learning_rate": 3.5558532613409594e-06, + "loss": 0.801, + "step": 2659 + }, + { + "epoch": 0.23988817243089688, + "grad_norm": 1.6007054710789903, + "learning_rate": 3.555486103229535e-06, + "loss": 0.727, + "step": 2660 + }, + { + "epoch": 0.23997835595436715, + "grad_norm": 5.936917888816948, + "learning_rate": 3.5551188123954595e-06, + "loss": 0.8132, + "step": 2661 + }, + { + "epoch": 0.2400685394778374, + "grad_norm": 1.6058730755599429, + "learning_rate": 3.5547513888700715e-06, + "loss": 0.6852, + "step": 2662 + }, + { + "epoch": 0.24015872300130767, + "grad_norm": 1.551000392237404, + "learning_rate": 3.5543838326847224e-06, + "loss": 0.8186, + "step": 2663 + }, + { + "epoch": 0.24024890652477793, + "grad_norm": 2.1784164386979374, + "learning_rate": 3.5540161438707744e-06, + "loss": 0.7637, + "step": 2664 + }, + { + "epoch": 0.2403390900482482, + "grad_norm": 1.6985574772912777, + "learning_rate": 3.5536483224596e-06, + "loss": 0.9424, + "step": 2665 + }, + { + "epoch": 0.24042927357171845, + "grad_norm": 2.0445523661425673, + "learning_rate": 3.553280368482584e-06, + "loss": 0.7644, + "step": 2666 + }, + { + "epoch": 0.2405194570951887, + "grad_norm": 1.6836985402785538, + "learning_rate": 3.5529122819711227e-06, + "loss": 0.6113, + "step": 2667 + }, + { + "epoch": 0.24060964061865897, + "grad_norm": 1.7098630553616918, + "learning_rate": 3.5525440629566223e-06, + "loss": 0.731, + "step": 2668 + }, + { + "epoch": 0.24069982414212923, + "grad_norm": 2.427063926870231, + "learning_rate": 3.552175711470502e-06, + "loss": 0.8073, + "step": 2669 + }, + { + "epoch": 0.2407900076655995, + "grad_norm": 3.9951004777669374, + "learning_rate": 3.5518072275441912e-06, + "loss": 0.8263, + "step": 2670 + }, + { + "epoch": 0.24088019118906975, + "grad_norm": 1.534058233998456, + "learning_rate": 3.551438611209131e-06, + "loss": 0.8139, + "step": 2671 + }, + { + "epoch": 0.24097037471254001, + "grad_norm": 2.3367540852315187, + "learning_rate": 3.551069862496774e-06, + "loss": 0.8834, + "step": 2672 + }, + { + "epoch": 0.24106055823601027, + "grad_norm": 0.648316797259696, + "learning_rate": 3.5507009814385846e-06, + "loss": 0.599, + "step": 2673 + }, + { + "epoch": 0.24115074175948054, + "grad_norm": 0.6542378640964303, + "learning_rate": 3.550331968066036e-06, + "loss": 0.5904, + "step": 2674 + }, + { + "epoch": 0.2412409252829508, + "grad_norm": 1.9801757114164127, + "learning_rate": 3.549962822410616e-06, + "loss": 0.8038, + "step": 2675 + }, + { + "epoch": 0.24133110880642106, + "grad_norm": 0.6948041879496969, + "learning_rate": 3.5495935445038217e-06, + "loss": 0.5868, + "step": 2676 + }, + { + "epoch": 0.24142129232989132, + "grad_norm": 1.840771937050869, + "learning_rate": 3.5492241343771612e-06, + "loss": 0.8268, + "step": 2677 + }, + { + "epoch": 0.24151147585336158, + "grad_norm": 2.7333405199964376, + "learning_rate": 3.548854592062156e-06, + "loss": 0.7817, + "step": 2678 + }, + { + "epoch": 0.24160165937683184, + "grad_norm": 3.1324024265007604, + "learning_rate": 3.548484917590336e-06, + "loss": 0.7675, + "step": 2679 + }, + { + "epoch": 0.24169184290030213, + "grad_norm": 2.019546479217542, + "learning_rate": 3.5481151109932447e-06, + "loss": 0.7032, + "step": 2680 + }, + { + "epoch": 0.2417820264237724, + "grad_norm": 1.7438568191898556, + "learning_rate": 3.5477451723024364e-06, + "loss": 0.8131, + "step": 2681 + }, + { + "epoch": 0.24187220994724265, + "grad_norm": 0.5888908470044281, + "learning_rate": 3.5473751015494757e-06, + "loss": 0.5711, + "step": 2682 + }, + { + "epoch": 0.2419623934707129, + "grad_norm": 2.0818862721422313, + "learning_rate": 3.547004898765939e-06, + "loss": 0.8044, + "step": 2683 + }, + { + "epoch": 0.24205257699418317, + "grad_norm": 1.3047209866732008, + "learning_rate": 3.546634563983414e-06, + "loss": 0.8065, + "step": 2684 + }, + { + "epoch": 0.24214276051765343, + "grad_norm": 1.8343827827457477, + "learning_rate": 3.5462640972335002e-06, + "loss": 0.7509, + "step": 2685 + }, + { + "epoch": 0.2422329440411237, + "grad_norm": 1.4876597150440327, + "learning_rate": 3.5458934985478077e-06, + "loss": 0.8088, + "step": 2686 + }, + { + "epoch": 0.24232312756459395, + "grad_norm": 2.7081644347907456, + "learning_rate": 3.5455227679579577e-06, + "loss": 0.7624, + "step": 2687 + }, + { + "epoch": 0.24241331108806422, + "grad_norm": 2.795392032214718, + "learning_rate": 3.545151905495584e-06, + "loss": 0.8919, + "step": 2688 + }, + { + "epoch": 0.24250349461153448, + "grad_norm": 2.41389874640548, + "learning_rate": 3.544780911192329e-06, + "loss": 0.7737, + "step": 2689 + }, + { + "epoch": 0.24259367813500474, + "grad_norm": 1.7094771889890992, + "learning_rate": 3.544409785079849e-06, + "loss": 0.7356, + "step": 2690 + }, + { + "epoch": 0.242683861658475, + "grad_norm": 1.4399617718549638, + "learning_rate": 3.5440385271898103e-06, + "loss": 0.728, + "step": 2691 + }, + { + "epoch": 0.24277404518194526, + "grad_norm": 1.6602644582152473, + "learning_rate": 3.5436671375538903e-06, + "loss": 0.7122, + "step": 2692 + }, + { + "epoch": 0.24286422870541552, + "grad_norm": 2.352540684022895, + "learning_rate": 3.543295616203779e-06, + "loss": 0.7894, + "step": 2693 + }, + { + "epoch": 0.24295441222888578, + "grad_norm": 1.8069568584086406, + "learning_rate": 3.542923963171176e-06, + "loss": 0.8822, + "step": 2694 + }, + { + "epoch": 0.24304459575235604, + "grad_norm": 1.9423059109768976, + "learning_rate": 3.542552178487793e-06, + "loss": 0.7037, + "step": 2695 + }, + { + "epoch": 0.2431347792758263, + "grad_norm": 2.1842005232955044, + "learning_rate": 3.5421802621853523e-06, + "loss": 0.7765, + "step": 2696 + }, + { + "epoch": 0.24322496279929656, + "grad_norm": 1.825827369221717, + "learning_rate": 3.5418082142955887e-06, + "loss": 0.783, + "step": 2697 + }, + { + "epoch": 0.24331514632276682, + "grad_norm": 0.6958463740887595, + "learning_rate": 3.5414360348502463e-06, + "loss": 0.6159, + "step": 2698 + }, + { + "epoch": 0.24340532984623708, + "grad_norm": 1.5948700054852, + "learning_rate": 3.5410637238810825e-06, + "loss": 0.6938, + "step": 2699 + }, + { + "epoch": 0.24349551336970734, + "grad_norm": 1.7417022987383406, + "learning_rate": 3.5406912814198635e-06, + "loss": 0.7428, + "step": 2700 + }, + { + "epoch": 0.2435856968931776, + "grad_norm": 1.7537554227890604, + "learning_rate": 3.54031870749837e-06, + "loss": 0.75, + "step": 2701 + }, + { + "epoch": 0.24367588041664787, + "grad_norm": 1.4242152144429512, + "learning_rate": 3.539946002148391e-06, + "loss": 0.7935, + "step": 2702 + }, + { + "epoch": 0.24376606394011813, + "grad_norm": 2.553666870437187, + "learning_rate": 3.5395731654017277e-06, + "loss": 0.792, + "step": 2703 + }, + { + "epoch": 0.24385624746358842, + "grad_norm": 2.1549704251837767, + "learning_rate": 3.5392001972901923e-06, + "loss": 0.8261, + "step": 2704 + }, + { + "epoch": 0.24394643098705868, + "grad_norm": 1.6447143452762163, + "learning_rate": 3.5388270978456098e-06, + "loss": 0.7468, + "step": 2705 + }, + { + "epoch": 0.24403661451052894, + "grad_norm": 1.8028598795441024, + "learning_rate": 3.5384538670998137e-06, + "loss": 0.7591, + "step": 2706 + }, + { + "epoch": 0.2441267980339992, + "grad_norm": 1.8252699639253054, + "learning_rate": 3.538080505084651e-06, + "loss": 0.809, + "step": 2707 + }, + { + "epoch": 0.24421698155746946, + "grad_norm": 1.4372291309741305, + "learning_rate": 3.5377070118319788e-06, + "loss": 0.7978, + "step": 2708 + }, + { + "epoch": 0.24430716508093972, + "grad_norm": 2.3651280272914903, + "learning_rate": 3.5373333873736657e-06, + "loss": 0.7472, + "step": 2709 + }, + { + "epoch": 0.24439734860440998, + "grad_norm": 1.4925414459505804, + "learning_rate": 3.536959631741591e-06, + "loss": 0.7928, + "step": 2710 + }, + { + "epoch": 0.24448753212788024, + "grad_norm": 1.7872564536685298, + "learning_rate": 3.536585744967646e-06, + "loss": 0.8066, + "step": 2711 + }, + { + "epoch": 0.2445777156513505, + "grad_norm": 1.9617413093002014, + "learning_rate": 3.5362117270837326e-06, + "loss": 0.7712, + "step": 2712 + }, + { + "epoch": 0.24466789917482076, + "grad_norm": 1.5424571287605795, + "learning_rate": 3.5358375781217634e-06, + "loss": 0.686, + "step": 2713 + }, + { + "epoch": 0.24475808269829102, + "grad_norm": 1.5693658660902403, + "learning_rate": 3.535463298113664e-06, + "loss": 0.8224, + "step": 2714 + }, + { + "epoch": 0.24484826622176128, + "grad_norm": 1.6133617168614744, + "learning_rate": 3.5350888870913697e-06, + "loss": 0.7458, + "step": 2715 + }, + { + "epoch": 0.24493844974523155, + "grad_norm": 1.668994882739195, + "learning_rate": 3.5347143450868273e-06, + "loss": 0.7921, + "step": 2716 + }, + { + "epoch": 0.2450286332687018, + "grad_norm": 1.8712802865655032, + "learning_rate": 3.534339672131994e-06, + "loss": 0.8113, + "step": 2717 + }, + { + "epoch": 0.24511881679217207, + "grad_norm": 2.201385243998123, + "learning_rate": 3.5339648682588397e-06, + "loss": 0.7203, + "step": 2718 + }, + { + "epoch": 0.24520900031564233, + "grad_norm": 1.7375263074900393, + "learning_rate": 3.533589933499345e-06, + "loss": 0.8334, + "step": 2719 + }, + { + "epoch": 0.2452991838391126, + "grad_norm": 2.3582661024919034, + "learning_rate": 3.533214867885501e-06, + "loss": 0.7212, + "step": 2720 + }, + { + "epoch": 0.24538936736258285, + "grad_norm": 1.9400433490275306, + "learning_rate": 3.53283967144931e-06, + "loss": 0.722, + "step": 2721 + }, + { + "epoch": 0.2454795508860531, + "grad_norm": 2.2621536495597083, + "learning_rate": 3.532464344222787e-06, + "loss": 0.7983, + "step": 2722 + }, + { + "epoch": 0.24556973440952337, + "grad_norm": 2.0266027253922716, + "learning_rate": 3.532088886237956e-06, + "loss": 0.8294, + "step": 2723 + }, + { + "epoch": 0.24565991793299363, + "grad_norm": 0.6365730727473645, + "learning_rate": 3.5317132975268535e-06, + "loss": 0.5272, + "step": 2724 + }, + { + "epoch": 0.2457501014564639, + "grad_norm": 1.5160744384114409, + "learning_rate": 3.531337578121526e-06, + "loss": 0.8106, + "step": 2725 + }, + { + "epoch": 0.24584028497993415, + "grad_norm": 2.1297646008288917, + "learning_rate": 3.530961728054033e-06, + "loss": 0.8183, + "step": 2726 + }, + { + "epoch": 0.24593046850340441, + "grad_norm": 1.5864163883080271, + "learning_rate": 3.5305857473564435e-06, + "loss": 0.7458, + "step": 2727 + }, + { + "epoch": 0.2460206520268747, + "grad_norm": 2.00299765528724, + "learning_rate": 3.5302096360608385e-06, + "loss": 0.7226, + "step": 2728 + }, + { + "epoch": 0.24611083555034496, + "grad_norm": 1.6292786389804876, + "learning_rate": 3.5298333941993105e-06, + "loss": 0.8365, + "step": 2729 + }, + { + "epoch": 0.24620101907381522, + "grad_norm": 1.804656651881407, + "learning_rate": 3.529457021803962e-06, + "loss": 0.7802, + "step": 2730 + }, + { + "epoch": 0.24629120259728549, + "grad_norm": 1.914875963143209, + "learning_rate": 3.529080518906906e-06, + "loss": 0.8036, + "step": 2731 + }, + { + "epoch": 0.24638138612075575, + "grad_norm": 1.641896145978136, + "learning_rate": 3.5287038855402696e-06, + "loss": 0.7714, + "step": 2732 + }, + { + "epoch": 0.246471569644226, + "grad_norm": 1.884083868662609, + "learning_rate": 3.528327121736188e-06, + "loss": 0.8246, + "step": 2733 + }, + { + "epoch": 0.24656175316769627, + "grad_norm": 1.6774892927757599, + "learning_rate": 3.52795022752681e-06, + "loss": 0.8665, + "step": 2734 + }, + { + "epoch": 0.24665193669116653, + "grad_norm": 2.1986396528511896, + "learning_rate": 3.5275732029442925e-06, + "loss": 0.8593, + "step": 2735 + }, + { + "epoch": 0.2467421202146368, + "grad_norm": 1.992763540723244, + "learning_rate": 3.5271960480208077e-06, + "loss": 0.722, + "step": 2736 + }, + { + "epoch": 0.24683230373810705, + "grad_norm": 0.6906701819713651, + "learning_rate": 3.526818762788534e-06, + "loss": 0.629, + "step": 2737 + }, + { + "epoch": 0.2469224872615773, + "grad_norm": 1.637462001097568, + "learning_rate": 3.5264413472796653e-06, + "loss": 0.8227, + "step": 2738 + }, + { + "epoch": 0.24701267078504757, + "grad_norm": 2.063100091737042, + "learning_rate": 3.5260638015264037e-06, + "loss": 0.7687, + "step": 2739 + }, + { + "epoch": 0.24710285430851783, + "grad_norm": 2.510825655427781, + "learning_rate": 3.5256861255609644e-06, + "loss": 0.7542, + "step": 2740 + }, + { + "epoch": 0.2471930378319881, + "grad_norm": 1.7071664438026255, + "learning_rate": 3.5253083194155723e-06, + "loss": 0.7961, + "step": 2741 + }, + { + "epoch": 0.24728322135545835, + "grad_norm": 1.4709735839515363, + "learning_rate": 3.5249303831224637e-06, + "loss": 0.8081, + "step": 2742 + }, + { + "epoch": 0.24737340487892862, + "grad_norm": 1.779334661457304, + "learning_rate": 3.524552316713887e-06, + "loss": 0.7747, + "step": 2743 + }, + { + "epoch": 0.24746358840239888, + "grad_norm": 1.857762318567865, + "learning_rate": 3.5241741202220995e-06, + "loss": 0.8497, + "step": 2744 + }, + { + "epoch": 0.24755377192586914, + "grad_norm": 1.62017663958126, + "learning_rate": 3.5237957936793724e-06, + "loss": 0.7731, + "step": 2745 + }, + { + "epoch": 0.2476439554493394, + "grad_norm": 1.5622466835674136, + "learning_rate": 3.523417337117986e-06, + "loss": 0.7467, + "step": 2746 + }, + { + "epoch": 0.24773413897280966, + "grad_norm": 1.6814420540745159, + "learning_rate": 3.523038750570232e-06, + "loss": 0.7333, + "step": 2747 + }, + { + "epoch": 0.24782432249627992, + "grad_norm": 1.6434720721818263, + "learning_rate": 3.522660034068414e-06, + "loss": 0.8152, + "step": 2748 + }, + { + "epoch": 0.24791450601975018, + "grad_norm": 1.787604658357462, + "learning_rate": 3.5222811876448464e-06, + "loss": 0.8019, + "step": 2749 + }, + { + "epoch": 0.24800468954322044, + "grad_norm": 1.3013278397026604, + "learning_rate": 3.521902211331854e-06, + "loss": 0.802, + "step": 2750 + }, + { + "epoch": 0.2480948730666907, + "grad_norm": 1.5395527944603076, + "learning_rate": 3.5215231051617726e-06, + "loss": 0.7352, + "step": 2751 + }, + { + "epoch": 0.248185056590161, + "grad_norm": 1.7927331042383934, + "learning_rate": 3.521143869166951e-06, + "loss": 0.6878, + "step": 2752 + }, + { + "epoch": 0.24827524011363125, + "grad_norm": 2.142602160587608, + "learning_rate": 3.5207645033797464e-06, + "loss": 0.8223, + "step": 2753 + }, + { + "epoch": 0.2483654236371015, + "grad_norm": 1.7625397589778051, + "learning_rate": 3.5203850078325293e-06, + "loss": 0.7124, + "step": 2754 + }, + { + "epoch": 0.24845560716057177, + "grad_norm": 1.7597845925379019, + "learning_rate": 3.5200053825576797e-06, + "loss": 0.7058, + "step": 2755 + }, + { + "epoch": 0.24854579068404203, + "grad_norm": 4.740357298764122, + "learning_rate": 3.51962562758759e-06, + "loss": 0.7867, + "step": 2756 + }, + { + "epoch": 0.2486359742075123, + "grad_norm": 1.6582598457475466, + "learning_rate": 3.5192457429546627e-06, + "loss": 0.7563, + "step": 2757 + }, + { + "epoch": 0.24872615773098256, + "grad_norm": 1.6656447933659915, + "learning_rate": 3.5188657286913115e-06, + "loss": 0.7536, + "step": 2758 + }, + { + "epoch": 0.24881634125445282, + "grad_norm": 2.174595617651999, + "learning_rate": 3.518485584829961e-06, + "loss": 0.7484, + "step": 2759 + }, + { + "epoch": 0.24890652477792308, + "grad_norm": 1.6017850395598543, + "learning_rate": 3.5181053114030485e-06, + "loss": 0.7952, + "step": 2760 + }, + { + "epoch": 0.24899670830139334, + "grad_norm": 1.5651910115635865, + "learning_rate": 3.5177249084430198e-06, + "loss": 0.7808, + "step": 2761 + }, + { + "epoch": 0.2490868918248636, + "grad_norm": 1.8621504129121358, + "learning_rate": 3.517344375982333e-06, + "loss": 0.8109, + "step": 2762 + }, + { + "epoch": 0.24917707534833386, + "grad_norm": 1.8246937037494955, + "learning_rate": 3.5169637140534565e-06, + "loss": 0.8179, + "step": 2763 + }, + { + "epoch": 0.24926725887180412, + "grad_norm": 1.76930952893572, + "learning_rate": 3.5165829226888733e-06, + "loss": 0.7506, + "step": 2764 + }, + { + "epoch": 0.24935744239527438, + "grad_norm": 3.7923724317627148, + "learning_rate": 3.516202001921072e-06, + "loss": 0.81, + "step": 2765 + }, + { + "epoch": 0.24944762591874464, + "grad_norm": 1.5957395813231454, + "learning_rate": 3.515820951782555e-06, + "loss": 0.8091, + "step": 2766 + }, + { + "epoch": 0.2495378094422149, + "grad_norm": 2.149313675209968, + "learning_rate": 3.5154397723058366e-06, + "loss": 0.7185, + "step": 2767 + }, + { + "epoch": 0.24962799296568516, + "grad_norm": 1.6603683515704364, + "learning_rate": 3.5150584635234416e-06, + "loss": 0.8432, + "step": 2768 + }, + { + "epoch": 0.24971817648915542, + "grad_norm": 1.3693099196136913, + "learning_rate": 3.5146770254679035e-06, + "loss": 0.8239, + "step": 2769 + }, + { + "epoch": 0.24980836001262569, + "grad_norm": 1.6226930381781186, + "learning_rate": 3.51429545817177e-06, + "loss": 0.703, + "step": 2770 + }, + { + "epoch": 0.24989854353609595, + "grad_norm": 1.8028198750927444, + "learning_rate": 3.5139137616675985e-06, + "loss": 0.7942, + "step": 2771 + }, + { + "epoch": 0.2499887270595662, + "grad_norm": 1.6403728472989052, + "learning_rate": 3.513531935987957e-06, + "loss": 0.7259, + "step": 2772 + }, + { + "epoch": 0.2500789105830365, + "grad_norm": 2.188942896579553, + "learning_rate": 3.5131499811654253e-06, + "loss": 0.8022, + "step": 2773 + }, + { + "epoch": 0.25016909410650673, + "grad_norm": 1.5184223276822089, + "learning_rate": 3.512767897232594e-06, + "loss": 0.7469, + "step": 2774 + }, + { + "epoch": 0.250259277629977, + "grad_norm": 1.8736706789823547, + "learning_rate": 3.512385684222064e-06, + "loss": 0.7772, + "step": 2775 + }, + { + "epoch": 0.25034946115344725, + "grad_norm": 1.8017646272525256, + "learning_rate": 3.512003342166449e-06, + "loss": 0.8773, + "step": 2776 + }, + { + "epoch": 0.25043964467691754, + "grad_norm": 1.6192155735537241, + "learning_rate": 3.511620871098371e-06, + "loss": 0.8183, + "step": 2777 + }, + { + "epoch": 0.25052982820038777, + "grad_norm": 1.6385049837261882, + "learning_rate": 3.511238271050465e-06, + "loss": 0.815, + "step": 2778 + }, + { + "epoch": 0.25062001172385806, + "grad_norm": 1.9806468514861815, + "learning_rate": 3.5108555420553778e-06, + "loss": 0.6974, + "step": 2779 + }, + { + "epoch": 0.2507101952473283, + "grad_norm": 1.7316736026707007, + "learning_rate": 3.510472684145764e-06, + "loss": 0.6023, + "step": 2780 + }, + { + "epoch": 0.2508003787707986, + "grad_norm": 1.6539820131617728, + "learning_rate": 3.5100896973542926e-06, + "loss": 0.6746, + "step": 2781 + }, + { + "epoch": 0.2508905622942688, + "grad_norm": 2.0385493659600336, + "learning_rate": 3.509706581713642e-06, + "loss": 0.8222, + "step": 2782 + }, + { + "epoch": 0.2509807458177391, + "grad_norm": 1.6010242581170797, + "learning_rate": 3.509323337256501e-06, + "loss": 0.808, + "step": 2783 + }, + { + "epoch": 0.25107092934120934, + "grad_norm": 1.736134023727113, + "learning_rate": 3.5089399640155703e-06, + "loss": 0.7338, + "step": 2784 + }, + { + "epoch": 0.2511611128646796, + "grad_norm": 1.5255744227457202, + "learning_rate": 3.508556462023562e-06, + "loss": 0.7743, + "step": 2785 + }, + { + "epoch": 0.25125129638814986, + "grad_norm": 2.4750867231183613, + "learning_rate": 3.5081728313131984e-06, + "loss": 0.7666, + "step": 2786 + }, + { + "epoch": 0.25134147991162015, + "grad_norm": 3.4344563274108726, + "learning_rate": 3.5077890719172125e-06, + "loss": 0.7794, + "step": 2787 + }, + { + "epoch": 0.25143166343509044, + "grad_norm": 1.6499747389969348, + "learning_rate": 3.5074051838683497e-06, + "loss": 0.7854, + "step": 2788 + }, + { + "epoch": 0.25152184695856067, + "grad_norm": 1.8825314260547201, + "learning_rate": 3.5070211671993643e-06, + "loss": 0.7054, + "step": 2789 + }, + { + "epoch": 0.25161203048203096, + "grad_norm": 2.456068471941934, + "learning_rate": 3.5066370219430238e-06, + "loss": 0.8035, + "step": 2790 + }, + { + "epoch": 0.2517022140055012, + "grad_norm": 1.5614330463597086, + "learning_rate": 3.5062527481321044e-06, + "loss": 0.7811, + "step": 2791 + }, + { + "epoch": 0.2517923975289715, + "grad_norm": 1.98865150346814, + "learning_rate": 3.5058683457993954e-06, + "loss": 0.7969, + "step": 2792 + }, + { + "epoch": 0.2518825810524417, + "grad_norm": 0.7782945808168066, + "learning_rate": 3.5054838149776963e-06, + "loss": 0.5368, + "step": 2793 + }, + { + "epoch": 0.251972764575912, + "grad_norm": 1.5475752045286442, + "learning_rate": 3.505099155699816e-06, + "loss": 0.8192, + "step": 2794 + }, + { + "epoch": 0.25206294809938223, + "grad_norm": 1.6366477350019273, + "learning_rate": 3.5047143679985775e-06, + "loss": 0.6609, + "step": 2795 + }, + { + "epoch": 0.2521531316228525, + "grad_norm": 1.5067872826365891, + "learning_rate": 3.5043294519068126e-06, + "loss": 0.7966, + "step": 2796 + }, + { + "epoch": 0.25224331514632276, + "grad_norm": 1.9897586152788524, + "learning_rate": 3.503944407457363e-06, + "loss": 0.7065, + "step": 2797 + }, + { + "epoch": 0.25233349866979304, + "grad_norm": 1.7535785462360145, + "learning_rate": 3.5035592346830846e-06, + "loss": 0.797, + "step": 2798 + }, + { + "epoch": 0.2524236821932633, + "grad_norm": 1.9214194773652777, + "learning_rate": 3.503173933616841e-06, + "loss": 0.8516, + "step": 2799 + }, + { + "epoch": 0.25251386571673357, + "grad_norm": 1.6211386800767211, + "learning_rate": 3.50278850429151e-06, + "loss": 0.8566, + "step": 2800 + }, + { + "epoch": 0.2526040492402038, + "grad_norm": 1.6607149507957804, + "learning_rate": 3.502402946739977e-06, + "loss": 0.8047, + "step": 2801 + }, + { + "epoch": 0.2526942327636741, + "grad_norm": 1.739497070305647, + "learning_rate": 3.5020172609951405e-06, + "loss": 0.8015, + "step": 2802 + }, + { + "epoch": 0.2527844162871443, + "grad_norm": 1.4848652080409235, + "learning_rate": 3.501631447089909e-06, + "loss": 0.7978, + "step": 2803 + }, + { + "epoch": 0.2528745998106146, + "grad_norm": 2.0515751346568054, + "learning_rate": 3.501245505057203e-06, + "loss": 0.7584, + "step": 2804 + }, + { + "epoch": 0.25296478333408484, + "grad_norm": 1.7600820174397056, + "learning_rate": 3.5008594349299526e-06, + "loss": 0.8568, + "step": 2805 + }, + { + "epoch": 0.25305496685755513, + "grad_norm": 1.6754899916707116, + "learning_rate": 3.500473236741099e-06, + "loss": 0.7672, + "step": 2806 + }, + { + "epoch": 0.25314515038102536, + "grad_norm": 1.495526637068085, + "learning_rate": 3.500086910523596e-06, + "loss": 0.7417, + "step": 2807 + }, + { + "epoch": 0.25323533390449565, + "grad_norm": 1.6097855275809134, + "learning_rate": 3.499700456310406e-06, + "loss": 0.7302, + "step": 2808 + }, + { + "epoch": 0.2533255174279659, + "grad_norm": 2.2167541768671506, + "learning_rate": 3.499313874134504e-06, + "loss": 0.8319, + "step": 2809 + }, + { + "epoch": 0.2534157009514362, + "grad_norm": 2.1633446118962545, + "learning_rate": 3.498927164028875e-06, + "loss": 0.6543, + "step": 2810 + }, + { + "epoch": 0.25350588447490646, + "grad_norm": 1.3127741754354811, + "learning_rate": 3.498540326026515e-06, + "loss": 0.8171, + "step": 2811 + }, + { + "epoch": 0.2535960679983767, + "grad_norm": 1.5674822148921805, + "learning_rate": 3.4981533601604323e-06, + "loss": 0.8525, + "step": 2812 + }, + { + "epoch": 0.253686251521847, + "grad_norm": 1.7593048542464638, + "learning_rate": 3.4977662664636443e-06, + "loss": 0.7942, + "step": 2813 + }, + { + "epoch": 0.2537764350453172, + "grad_norm": 1.6888485923917755, + "learning_rate": 3.497379044969179e-06, + "loss": 0.7431, + "step": 2814 + }, + { + "epoch": 0.2538666185687875, + "grad_norm": 1.6847399286211668, + "learning_rate": 3.4969916957100777e-06, + "loss": 0.776, + "step": 2815 + }, + { + "epoch": 0.25395680209225774, + "grad_norm": 1.4922655494609638, + "learning_rate": 3.4966042187193905e-06, + "loss": 0.7748, + "step": 2816 + }, + { + "epoch": 0.254046985615728, + "grad_norm": 1.4409379069009842, + "learning_rate": 3.496216614030179e-06, + "loss": 0.7782, + "step": 2817 + }, + { + "epoch": 0.25413716913919826, + "grad_norm": 1.598753461438832, + "learning_rate": 3.495828881675516e-06, + "loss": 0.6808, + "step": 2818 + }, + { + "epoch": 0.25422735266266855, + "grad_norm": 6.788710186665443, + "learning_rate": 3.4954410216884845e-06, + "loss": 0.8511, + "step": 2819 + }, + { + "epoch": 0.2543175361861388, + "grad_norm": 1.6711133844293076, + "learning_rate": 3.49505303410218e-06, + "loss": 0.793, + "step": 2820 + }, + { + "epoch": 0.25440771970960907, + "grad_norm": 3.724896008364841, + "learning_rate": 3.4946649189497067e-06, + "loss": 0.8094, + "step": 2821 + }, + { + "epoch": 0.2544979032330793, + "grad_norm": 19.58305551318338, + "learning_rate": 3.4942766762641805e-06, + "loss": 0.7743, + "step": 2822 + }, + { + "epoch": 0.2545880867565496, + "grad_norm": 2.025445479383144, + "learning_rate": 3.49388830607873e-06, + "loss": 0.8118, + "step": 2823 + }, + { + "epoch": 0.2546782702800198, + "grad_norm": 1.2732861200703645, + "learning_rate": 3.493499808426491e-06, + "loss": 0.748, + "step": 2824 + }, + { + "epoch": 0.2547684538034901, + "grad_norm": 1.8112744101070097, + "learning_rate": 3.493111183340614e-06, + "loss": 0.8328, + "step": 2825 + }, + { + "epoch": 0.25485863732696035, + "grad_norm": 1.5913551887931254, + "learning_rate": 3.4927224308542576e-06, + "loss": 0.6825, + "step": 2826 + }, + { + "epoch": 0.25494882085043064, + "grad_norm": 1.5628758550631057, + "learning_rate": 3.4923335510005923e-06, + "loss": 0.8623, + "step": 2827 + }, + { + "epoch": 0.25503900437390087, + "grad_norm": 1.5322513809285014, + "learning_rate": 3.4919445438128e-06, + "loss": 0.8026, + "step": 2828 + }, + { + "epoch": 0.25512918789737116, + "grad_norm": 1.4550974594982233, + "learning_rate": 3.491555409324073e-06, + "loss": 0.8617, + "step": 2829 + }, + { + "epoch": 0.2552193714208414, + "grad_norm": 0.6677975128377142, + "learning_rate": 3.4911661475676136e-06, + "loss": 0.5608, + "step": 2830 + }, + { + "epoch": 0.2553095549443117, + "grad_norm": 1.5265856175931984, + "learning_rate": 3.490776758576637e-06, + "loss": 0.7575, + "step": 2831 + }, + { + "epoch": 0.2553997384677819, + "grad_norm": 2.9740689463624217, + "learning_rate": 3.4903872423843668e-06, + "loss": 0.8303, + "step": 2832 + }, + { + "epoch": 0.2554899219912522, + "grad_norm": 1.6706439325506457, + "learning_rate": 3.4899975990240396e-06, + "loss": 0.7472, + "step": 2833 + }, + { + "epoch": 0.25558010551472243, + "grad_norm": 2.0299992158258737, + "learning_rate": 3.489607828528901e-06, + "loss": 0.8057, + "step": 2834 + }, + { + "epoch": 0.2556702890381927, + "grad_norm": 0.6616549437474907, + "learning_rate": 3.4892179309322093e-06, + "loss": 0.6698, + "step": 2835 + }, + { + "epoch": 0.255760472561663, + "grad_norm": 1.6484926765934098, + "learning_rate": 3.488827906267232e-06, + "loss": 0.6939, + "step": 2836 + }, + { + "epoch": 0.25585065608513324, + "grad_norm": 1.5931096005808547, + "learning_rate": 3.4884377545672485e-06, + "loss": 0.8194, + "step": 2837 + }, + { + "epoch": 0.25594083960860353, + "grad_norm": 1.4786276932325093, + "learning_rate": 3.4880474758655485e-06, + "loss": 0.7566, + "step": 2838 + }, + { + "epoch": 0.25603102313207377, + "grad_norm": 1.599564585448756, + "learning_rate": 3.487657070195433e-06, + "loss": 0.8491, + "step": 2839 + }, + { + "epoch": 0.25612120665554405, + "grad_norm": 1.5947977249890881, + "learning_rate": 3.487266537590213e-06, + "loss": 0.857, + "step": 2840 + }, + { + "epoch": 0.2562113901790143, + "grad_norm": 1.495781529768349, + "learning_rate": 3.4868758780832116e-06, + "loss": 0.6981, + "step": 2841 + }, + { + "epoch": 0.2563015737024846, + "grad_norm": 1.7733837237166041, + "learning_rate": 3.486485091707762e-06, + "loss": 0.793, + "step": 2842 + }, + { + "epoch": 0.2563917572259548, + "grad_norm": 1.812134475682181, + "learning_rate": 3.4860941784972077e-06, + "loss": 0.6603, + "step": 2843 + }, + { + "epoch": 0.2564819407494251, + "grad_norm": 2.63202789116201, + "learning_rate": 3.485703138484904e-06, + "loss": 0.8779, + "step": 2844 + }, + { + "epoch": 0.25657212427289533, + "grad_norm": 1.7330921686957974, + "learning_rate": 3.485311971704216e-06, + "loss": 0.8311, + "step": 2845 + }, + { + "epoch": 0.2566623077963656, + "grad_norm": 1.4717707311544852, + "learning_rate": 3.484920678188521e-06, + "loss": 0.8144, + "step": 2846 + }, + { + "epoch": 0.25675249131983585, + "grad_norm": 2.023665485278477, + "learning_rate": 3.4845292579712063e-06, + "loss": 0.8056, + "step": 2847 + }, + { + "epoch": 0.25684267484330614, + "grad_norm": 2.3691041948327474, + "learning_rate": 3.484137711085669e-06, + "loss": 0.6647, + "step": 2848 + }, + { + "epoch": 0.2569328583667764, + "grad_norm": 2.3731291580063303, + "learning_rate": 3.4837460375653198e-06, + "loss": 0.7686, + "step": 2849 + }, + { + "epoch": 0.25702304189024666, + "grad_norm": 3.218370730082226, + "learning_rate": 3.483354237443576e-06, + "loss": 0.8099, + "step": 2850 + }, + { + "epoch": 0.2571132254137169, + "grad_norm": 1.465773955952114, + "learning_rate": 3.48296231075387e-06, + "loss": 0.7248, + "step": 2851 + }, + { + "epoch": 0.2572034089371872, + "grad_norm": 1.5571086184835852, + "learning_rate": 3.4825702575296433e-06, + "loss": 0.7992, + "step": 2852 + }, + { + "epoch": 0.2572935924606574, + "grad_norm": 0.6914814191571244, + "learning_rate": 3.482178077804347e-06, + "loss": 0.6136, + "step": 2853 + }, + { + "epoch": 0.2573837759841277, + "grad_norm": 1.8440519829123536, + "learning_rate": 3.4817857716114443e-06, + "loss": 0.7482, + "step": 2854 + }, + { + "epoch": 0.25747395950759794, + "grad_norm": 3.0815204927207778, + "learning_rate": 3.4813933389844094e-06, + "loss": 0.7376, + "step": 2855 + }, + { + "epoch": 0.2575641430310682, + "grad_norm": 1.739230738862311, + "learning_rate": 3.4810007799567264e-06, + "loss": 0.7911, + "step": 2856 + }, + { + "epoch": 0.25765432655453846, + "grad_norm": 2.0218195393191634, + "learning_rate": 3.480608094561891e-06, + "loss": 0.7966, + "step": 2857 + }, + { + "epoch": 0.25774451007800875, + "grad_norm": 1.6572287204887983, + "learning_rate": 3.4802152828334083e-06, + "loss": 0.7992, + "step": 2858 + }, + { + "epoch": 0.25783469360147904, + "grad_norm": 2.3985099657275932, + "learning_rate": 3.479822344804796e-06, + "loss": 0.7252, + "step": 2859 + }, + { + "epoch": 0.25792487712494927, + "grad_norm": 1.7879007890515315, + "learning_rate": 3.479429280509582e-06, + "loss": 0.8405, + "step": 2860 + }, + { + "epoch": 0.25801506064841956, + "grad_norm": 1.9432954557248794, + "learning_rate": 3.4790360899813038e-06, + "loss": 0.8269, + "step": 2861 + }, + { + "epoch": 0.2581052441718898, + "grad_norm": 2.0077424389679344, + "learning_rate": 3.4786427732535115e-06, + "loss": 0.7175, + "step": 2862 + }, + { + "epoch": 0.2581954276953601, + "grad_norm": 2.1862456949059665, + "learning_rate": 3.478249330359764e-06, + "loss": 0.8235, + "step": 2863 + }, + { + "epoch": 0.2582856112188303, + "grad_norm": 1.539692860975895, + "learning_rate": 3.4778557613336333e-06, + "loss": 0.713, + "step": 2864 + }, + { + "epoch": 0.2583757947423006, + "grad_norm": 1.786889701641349, + "learning_rate": 3.4774620662087004e-06, + "loss": 0.7986, + "step": 2865 + }, + { + "epoch": 0.25846597826577083, + "grad_norm": 2.063610124936806, + "learning_rate": 3.477068245018557e-06, + "loss": 0.7823, + "step": 2866 + }, + { + "epoch": 0.2585561617892411, + "grad_norm": 1.6836692987825619, + "learning_rate": 3.476674297796807e-06, + "loss": 0.7867, + "step": 2867 + }, + { + "epoch": 0.25864634531271136, + "grad_norm": 1.7662695324378614, + "learning_rate": 3.4762802245770627e-06, + "loss": 0.8239, + "step": 2868 + }, + { + "epoch": 0.25873652883618165, + "grad_norm": 1.5216237544220195, + "learning_rate": 3.4758860253929497e-06, + "loss": 0.7285, + "step": 2869 + }, + { + "epoch": 0.2588267123596519, + "grad_norm": 1.6549904550467283, + "learning_rate": 3.4754917002781038e-06, + "loss": 0.804, + "step": 2870 + }, + { + "epoch": 0.25891689588312217, + "grad_norm": 2.956395835585946, + "learning_rate": 3.475097249266169e-06, + "loss": 0.8233, + "step": 2871 + }, + { + "epoch": 0.2590070794065924, + "grad_norm": 1.648183721741051, + "learning_rate": 3.4747026723908044e-06, + "loss": 0.7681, + "step": 2872 + }, + { + "epoch": 0.2590972629300627, + "grad_norm": 1.6213501509157082, + "learning_rate": 3.474307969685676e-06, + "loss": 0.7637, + "step": 2873 + }, + { + "epoch": 0.2591874464535329, + "grad_norm": 1.5399725687382977, + "learning_rate": 3.473913141184462e-06, + "loss": 0.7332, + "step": 2874 + }, + { + "epoch": 0.2592776299770032, + "grad_norm": 1.5855253087973828, + "learning_rate": 3.4735181869208523e-06, + "loss": 0.6548, + "step": 2875 + }, + { + "epoch": 0.25936781350047344, + "grad_norm": 2.3280623254724473, + "learning_rate": 3.473123106928546e-06, + "loss": 0.6582, + "step": 2876 + }, + { + "epoch": 0.25945799702394373, + "grad_norm": 1.86331509413462, + "learning_rate": 3.4727279012412533e-06, + "loss": 0.7399, + "step": 2877 + }, + { + "epoch": 0.25954818054741396, + "grad_norm": 0.6337612951276601, + "learning_rate": 3.4723325698926953e-06, + "loss": 0.5761, + "step": 2878 + }, + { + "epoch": 0.25963836407088425, + "grad_norm": 1.6060431350910447, + "learning_rate": 3.4719371129166045e-06, + "loss": 0.7855, + "step": 2879 + }, + { + "epoch": 0.2597285475943545, + "grad_norm": 1.746100918543905, + "learning_rate": 3.471541530346723e-06, + "loss": 0.7058, + "step": 2880 + }, + { + "epoch": 0.2598187311178248, + "grad_norm": 1.8983262896832729, + "learning_rate": 3.4711458222168037e-06, + "loss": 0.8489, + "step": 2881 + }, + { + "epoch": 0.259908914641295, + "grad_norm": 1.5521642249957996, + "learning_rate": 3.4707499885606114e-06, + "loss": 0.8233, + "step": 2882 + }, + { + "epoch": 0.2599990981647653, + "grad_norm": 1.6090985403286502, + "learning_rate": 3.4703540294119204e-06, + "loss": 0.8162, + "step": 2883 + }, + { + "epoch": 0.2600892816882356, + "grad_norm": 1.876680002972571, + "learning_rate": 3.4699579448045163e-06, + "loss": 0.8397, + "step": 2884 + }, + { + "epoch": 0.2601794652117058, + "grad_norm": 1.4821256761326567, + "learning_rate": 3.4695617347721947e-06, + "loss": 0.6692, + "step": 2885 + }, + { + "epoch": 0.2602696487351761, + "grad_norm": 1.7795241176238883, + "learning_rate": 3.469165399348763e-06, + "loss": 0.7524, + "step": 2886 + }, + { + "epoch": 0.26035983225864634, + "grad_norm": 1.946280365047363, + "learning_rate": 3.4687689385680384e-06, + "loss": 0.7665, + "step": 2887 + }, + { + "epoch": 0.26045001578211663, + "grad_norm": 1.6235642325873763, + "learning_rate": 3.4683723524638494e-06, + "loss": 0.7636, + "step": 2888 + }, + { + "epoch": 0.26054019930558686, + "grad_norm": 1.7024534808958094, + "learning_rate": 3.4679756410700354e-06, + "loss": 0.8499, + "step": 2889 + }, + { + "epoch": 0.26063038282905715, + "grad_norm": 2.131799925614869, + "learning_rate": 3.4675788044204445e-06, + "loss": 0.7858, + "step": 2890 + }, + { + "epoch": 0.2607205663525274, + "grad_norm": 1.5104880042055941, + "learning_rate": 3.467181842548938e-06, + "loss": 0.7128, + "step": 2891 + }, + { + "epoch": 0.26081074987599767, + "grad_norm": 1.6134553311988298, + "learning_rate": 3.466784755489387e-06, + "loss": 0.7819, + "step": 2892 + }, + { + "epoch": 0.2609009333994679, + "grad_norm": 1.5790443386450497, + "learning_rate": 3.4663875432756726e-06, + "loss": 0.8997, + "step": 2893 + }, + { + "epoch": 0.2609911169229382, + "grad_norm": 3.1452960298640438, + "learning_rate": 3.465990205941687e-06, + "loss": 0.9555, + "step": 2894 + }, + { + "epoch": 0.2610813004464084, + "grad_norm": 3.41570586364459, + "learning_rate": 3.465592743521335e-06, + "loss": 0.7665, + "step": 2895 + }, + { + "epoch": 0.2611714839698787, + "grad_norm": 1.8121712485142376, + "learning_rate": 3.465195156048528e-06, + "loss": 0.8328, + "step": 2896 + }, + { + "epoch": 0.26126166749334895, + "grad_norm": 0.5703837990283788, + "learning_rate": 3.464797443557191e-06, + "loss": 0.574, + "step": 2897 + }, + { + "epoch": 0.26135185101681924, + "grad_norm": 2.0802891297393655, + "learning_rate": 3.46439960608126e-06, + "loss": 0.7572, + "step": 2898 + }, + { + "epoch": 0.26144203454028947, + "grad_norm": 2.1616058176731796, + "learning_rate": 3.4640016436546797e-06, + "loss": 0.7269, + "step": 2899 + }, + { + "epoch": 0.26153221806375976, + "grad_norm": 1.78431758754849, + "learning_rate": 3.4636035563114065e-06, + "loss": 0.7615, + "step": 2900 + }, + { + "epoch": 0.26162240158723, + "grad_norm": 1.5764111570035568, + "learning_rate": 3.4632053440854085e-06, + "loss": 0.7103, + "step": 2901 + }, + { + "epoch": 0.2617125851107003, + "grad_norm": 1.802607539086621, + "learning_rate": 3.462807007010662e-06, + "loss": 0.8377, + "step": 2902 + }, + { + "epoch": 0.2618027686341705, + "grad_norm": 2.299090815259968, + "learning_rate": 3.462408545121155e-06, + "loss": 0.8119, + "step": 2903 + }, + { + "epoch": 0.2618929521576408, + "grad_norm": 4.501049978891102, + "learning_rate": 3.4620099584508883e-06, + "loss": 0.8422, + "step": 2904 + }, + { + "epoch": 0.26198313568111103, + "grad_norm": 1.7732157967749929, + "learning_rate": 3.46161124703387e-06, + "loss": 0.7423, + "step": 2905 + }, + { + "epoch": 0.2620733192045813, + "grad_norm": 2.417628053194902, + "learning_rate": 3.461212410904122e-06, + "loss": 0.7847, + "step": 2906 + }, + { + "epoch": 0.2621635027280516, + "grad_norm": 1.6143231664260382, + "learning_rate": 3.4608134500956726e-06, + "loss": 0.8422, + "step": 2907 + }, + { + "epoch": 0.26225368625152184, + "grad_norm": 1.5722813954954418, + "learning_rate": 3.4604143646425655e-06, + "loss": 0.7698, + "step": 2908 + }, + { + "epoch": 0.26234386977499213, + "grad_norm": 1.5457544603130728, + "learning_rate": 3.460015154578852e-06, + "loss": 0.8513, + "step": 2909 + }, + { + "epoch": 0.26243405329846237, + "grad_norm": 1.6627115846382334, + "learning_rate": 3.459615819938595e-06, + "loss": 0.7567, + "step": 2910 + }, + { + "epoch": 0.26252423682193265, + "grad_norm": 1.6624916793500295, + "learning_rate": 3.4592163607558684e-06, + "loss": 0.7502, + "step": 2911 + }, + { + "epoch": 0.2626144203454029, + "grad_norm": 1.3518075886151746, + "learning_rate": 3.4588167770647553e-06, + "loss": 0.7845, + "step": 2912 + }, + { + "epoch": 0.2627046038688732, + "grad_norm": 1.735185657099165, + "learning_rate": 3.458417068899351e-06, + "loss": 0.8724, + "step": 2913 + }, + { + "epoch": 0.2627947873923434, + "grad_norm": 1.4337645384792221, + "learning_rate": 3.4580172362937612e-06, + "loss": 0.7511, + "step": 2914 + }, + { + "epoch": 0.2628849709158137, + "grad_norm": 1.2916255093241906, + "learning_rate": 3.457617279282101e-06, + "loss": 0.746, + "step": 2915 + }, + { + "epoch": 0.26297515443928393, + "grad_norm": 2.3398029916107856, + "learning_rate": 3.4572171978984975e-06, + "loss": 0.763, + "step": 2916 + }, + { + "epoch": 0.2630653379627542, + "grad_norm": 1.5497415357870261, + "learning_rate": 3.456816992177088e-06, + "loss": 0.7847, + "step": 2917 + }, + { + "epoch": 0.26315552148622445, + "grad_norm": 1.3756818814543543, + "learning_rate": 3.4564166621520193e-06, + "loss": 0.7515, + "step": 2918 + }, + { + "epoch": 0.26324570500969474, + "grad_norm": 1.467333171141386, + "learning_rate": 3.4560162078574507e-06, + "loss": 0.7765, + "step": 2919 + }, + { + "epoch": 0.263335888533165, + "grad_norm": 1.639530070828045, + "learning_rate": 3.455615629327551e-06, + "loss": 0.7849, + "step": 2920 + }, + { + "epoch": 0.26342607205663526, + "grad_norm": 1.8942525442347231, + "learning_rate": 3.4552149265964994e-06, + "loss": 0.7635, + "step": 2921 + }, + { + "epoch": 0.2635162555801055, + "grad_norm": 1.6114304388681855, + "learning_rate": 3.4548140996984866e-06, + "loss": 0.7299, + "step": 2922 + }, + { + "epoch": 0.2636064391035758, + "grad_norm": 1.5877878453773666, + "learning_rate": 3.4544131486677124e-06, + "loss": 0.8069, + "step": 2923 + }, + { + "epoch": 0.263696622627046, + "grad_norm": 1.5165334908560177, + "learning_rate": 3.454012073538389e-06, + "loss": 0.7208, + "step": 2924 + }, + { + "epoch": 0.2637868061505163, + "grad_norm": 1.4548197062368344, + "learning_rate": 3.453610874344738e-06, + "loss": 0.7921, + "step": 2925 + }, + { + "epoch": 0.26387698967398654, + "grad_norm": 1.6897279372543577, + "learning_rate": 3.453209551120993e-06, + "loss": 0.8246, + "step": 2926 + }, + { + "epoch": 0.26396717319745683, + "grad_norm": 2.888907257249912, + "learning_rate": 3.452808103901395e-06, + "loss": 0.756, + "step": 2927 + }, + { + "epoch": 0.26405735672092706, + "grad_norm": 1.9411892990378596, + "learning_rate": 3.4524065327202e-06, + "loss": 0.8563, + "step": 2928 + }, + { + "epoch": 0.26414754024439735, + "grad_norm": 0.6494796834503622, + "learning_rate": 3.4520048376116702e-06, + "loss": 0.5261, + "step": 2929 + }, + { + "epoch": 0.26423772376786764, + "grad_norm": 1.6947696361583084, + "learning_rate": 3.4516030186100817e-06, + "loss": 0.8216, + "step": 2930 + }, + { + "epoch": 0.26432790729133787, + "grad_norm": 1.3666286525246598, + "learning_rate": 3.4512010757497197e-06, + "loss": 0.696, + "step": 2931 + }, + { + "epoch": 0.26441809081480816, + "grad_norm": 1.7535927541200227, + "learning_rate": 3.4507990090648804e-06, + "loss": 0.6991, + "step": 2932 + }, + { + "epoch": 0.2645082743382784, + "grad_norm": 1.7223483478685837, + "learning_rate": 3.4503968185898696e-06, + "loss": 0.8053, + "step": 2933 + }, + { + "epoch": 0.2645984578617487, + "grad_norm": 1.5454899916434333, + "learning_rate": 3.4499945043590047e-06, + "loss": 0.7875, + "step": 2934 + }, + { + "epoch": 0.2646886413852189, + "grad_norm": 2.0066225555657873, + "learning_rate": 3.4495920664066137e-06, + "loss": 0.7973, + "step": 2935 + }, + { + "epoch": 0.2647788249086892, + "grad_norm": 1.9730548720272316, + "learning_rate": 3.449189504767035e-06, + "loss": 0.7799, + "step": 2936 + }, + { + "epoch": 0.26486900843215944, + "grad_norm": 3.491886407558037, + "learning_rate": 3.4487868194746163e-06, + "loss": 0.8086, + "step": 2937 + }, + { + "epoch": 0.2649591919556297, + "grad_norm": 1.7195893752288727, + "learning_rate": 3.4483840105637175e-06, + "loss": 0.823, + "step": 2938 + }, + { + "epoch": 0.26504937547909996, + "grad_norm": 1.8737100614499198, + "learning_rate": 3.4479810780687097e-06, + "loss": 0.7686, + "step": 2939 + }, + { + "epoch": 0.26513955900257025, + "grad_norm": 1.4222864247440563, + "learning_rate": 3.4475780220239714e-06, + "loss": 0.7745, + "step": 2940 + }, + { + "epoch": 0.2652297425260405, + "grad_norm": 1.4148198412723991, + "learning_rate": 3.4471748424638948e-06, + "loss": 0.7718, + "step": 2941 + }, + { + "epoch": 0.26531992604951077, + "grad_norm": 2.1401433472440416, + "learning_rate": 3.4467715394228803e-06, + "loss": 0.5925, + "step": 2942 + }, + { + "epoch": 0.265410109572981, + "grad_norm": 1.3860109053894196, + "learning_rate": 3.4463681129353413e-06, + "loss": 0.785, + "step": 2943 + }, + { + "epoch": 0.2655002930964513, + "grad_norm": 1.4298987727583004, + "learning_rate": 3.4459645630357e-06, + "loss": 0.7724, + "step": 2944 + }, + { + "epoch": 0.2655904766199215, + "grad_norm": 1.6054510779231912, + "learning_rate": 3.4455608897583884e-06, + "loss": 0.8185, + "step": 2945 + }, + { + "epoch": 0.2656806601433918, + "grad_norm": 2.329997289107551, + "learning_rate": 3.4451570931378514e-06, + "loss": 0.783, + "step": 2946 + }, + { + "epoch": 0.26577084366686204, + "grad_norm": 1.4167985948205633, + "learning_rate": 3.444753173208543e-06, + "loss": 0.7888, + "step": 2947 + }, + { + "epoch": 0.26586102719033233, + "grad_norm": 1.9526926401329072, + "learning_rate": 3.444349130004927e-06, + "loss": 0.7794, + "step": 2948 + }, + { + "epoch": 0.26595121071380257, + "grad_norm": 1.7125389233920174, + "learning_rate": 3.4439449635614794e-06, + "loss": 0.7051, + "step": 2949 + }, + { + "epoch": 0.26604139423727285, + "grad_norm": 1.5888902959098603, + "learning_rate": 3.4435406739126854e-06, + "loss": 0.8091, + "step": 2950 + }, + { + "epoch": 0.2661315777607431, + "grad_norm": 2.102452993513839, + "learning_rate": 3.443136261093042e-06, + "loss": 0.6004, + "step": 2951 + }, + { + "epoch": 0.2662217612842134, + "grad_norm": 1.5598897111249286, + "learning_rate": 3.4427317251370553e-06, + "loss": 0.7261, + "step": 2952 + }, + { + "epoch": 0.2663119448076836, + "grad_norm": 2.7615642153657007, + "learning_rate": 3.4423270660792422e-06, + "loss": 0.6973, + "step": 2953 + }, + { + "epoch": 0.2664021283311539, + "grad_norm": 1.6679182201298532, + "learning_rate": 3.4419222839541314e-06, + "loss": 0.6672, + "step": 2954 + }, + { + "epoch": 0.2664923118546242, + "grad_norm": 1.419665957637364, + "learning_rate": 3.4415173787962607e-06, + "loss": 0.7564, + "step": 2955 + }, + { + "epoch": 0.2665824953780944, + "grad_norm": 2.1394839621859014, + "learning_rate": 3.4411123506401783e-06, + "loss": 0.6444, + "step": 2956 + }, + { + "epoch": 0.2666726789015647, + "grad_norm": 1.9098023006719773, + "learning_rate": 3.440707199520444e-06, + "loss": 0.6337, + "step": 2957 + }, + { + "epoch": 0.26676286242503494, + "grad_norm": 1.55866778587473, + "learning_rate": 3.440301925471628e-06, + "loss": 0.742, + "step": 2958 + }, + { + "epoch": 0.26685304594850523, + "grad_norm": 2.20605879433272, + "learning_rate": 3.43989652852831e-06, + "loss": 0.8156, + "step": 2959 + }, + { + "epoch": 0.26694322947197546, + "grad_norm": 1.5616178692766567, + "learning_rate": 3.4394910087250804e-06, + "loss": 0.8403, + "step": 2960 + }, + { + "epoch": 0.26703341299544575, + "grad_norm": 1.491301028584865, + "learning_rate": 3.4390853660965405e-06, + "loss": 0.7946, + "step": 2961 + }, + { + "epoch": 0.267123596518916, + "grad_norm": 1.8953052945169093, + "learning_rate": 3.438679600677302e-06, + "loss": 0.8596, + "step": 2962 + }, + { + "epoch": 0.2672137800423863, + "grad_norm": 2.6789725103274367, + "learning_rate": 3.4382737125019874e-06, + "loss": 0.735, + "step": 2963 + }, + { + "epoch": 0.2673039635658565, + "grad_norm": 2.7511823019999766, + "learning_rate": 3.4378677016052294e-06, + "loss": 0.7485, + "step": 2964 + }, + { + "epoch": 0.2673941470893268, + "grad_norm": 1.6964419070415218, + "learning_rate": 3.43746156802167e-06, + "loss": 0.7954, + "step": 2965 + }, + { + "epoch": 0.267484330612797, + "grad_norm": 1.5268925539398166, + "learning_rate": 3.4370553117859643e-06, + "loss": 0.7162, + "step": 2966 + }, + { + "epoch": 0.2675745141362673, + "grad_norm": 1.7465973880850945, + "learning_rate": 3.4366489329327754e-06, + "loss": 0.8069, + "step": 2967 + }, + { + "epoch": 0.26766469765973755, + "grad_norm": 2.2111887435234614, + "learning_rate": 3.4362424314967777e-06, + "loss": 0.8036, + "step": 2968 + }, + { + "epoch": 0.26775488118320784, + "grad_norm": 2.389017599418646, + "learning_rate": 3.4358358075126567e-06, + "loss": 0.8065, + "step": 2969 + }, + { + "epoch": 0.26784506470667807, + "grad_norm": 1.6593583646755825, + "learning_rate": 3.4354290610151077e-06, + "loss": 0.8011, + "step": 2970 + }, + { + "epoch": 0.26793524823014836, + "grad_norm": 1.3078643132272378, + "learning_rate": 3.4350221920388354e-06, + "loss": 0.7865, + "step": 2971 + }, + { + "epoch": 0.2680254317536186, + "grad_norm": 1.494879087578896, + "learning_rate": 3.4346152006185574e-06, + "loss": 0.8779, + "step": 2972 + }, + { + "epoch": 0.2681156152770889, + "grad_norm": 1.7248046349728576, + "learning_rate": 3.4342080867890006e-06, + "loss": 0.8772, + "step": 2973 + }, + { + "epoch": 0.2682057988005591, + "grad_norm": 3.8021745784035486, + "learning_rate": 3.4338008505849016e-06, + "loss": 0.6987, + "step": 2974 + }, + { + "epoch": 0.2682959823240294, + "grad_norm": 1.6191242065333502, + "learning_rate": 3.433393492041008e-06, + "loss": 0.7879, + "step": 2975 + }, + { + "epoch": 0.26838616584749964, + "grad_norm": 1.3956499074104762, + "learning_rate": 3.432986011192078e-06, + "loss": 0.8295, + "step": 2976 + }, + { + "epoch": 0.2684763493709699, + "grad_norm": 1.6479983354948953, + "learning_rate": 3.4325784080728796e-06, + "loss": 0.8472, + "step": 2977 + }, + { + "epoch": 0.2685665328944402, + "grad_norm": 1.5031663536045792, + "learning_rate": 3.4321706827181926e-06, + "loss": 0.8099, + "step": 2978 + }, + { + "epoch": 0.26865671641791045, + "grad_norm": 1.2105842813508372, + "learning_rate": 3.4317628351628064e-06, + "loss": 0.7311, + "step": 2979 + }, + { + "epoch": 0.26874689994138073, + "grad_norm": 1.588518644906075, + "learning_rate": 3.43135486544152e-06, + "loss": 0.7599, + "step": 2980 + }, + { + "epoch": 0.26883708346485097, + "grad_norm": 1.3954527559091008, + "learning_rate": 3.4309467735891442e-06, + "loss": 0.8148, + "step": 2981 + }, + { + "epoch": 0.26892726698832126, + "grad_norm": 1.774430723642197, + "learning_rate": 3.4305385596405e-06, + "loss": 0.8648, + "step": 2982 + }, + { + "epoch": 0.2690174505117915, + "grad_norm": 1.954508177218308, + "learning_rate": 3.4301302236304174e-06, + "loss": 0.7461, + "step": 2983 + }, + { + "epoch": 0.2691076340352618, + "grad_norm": 0.6631524076559451, + "learning_rate": 3.429721765593739e-06, + "loss": 0.5481, + "step": 2984 + }, + { + "epoch": 0.269197817558732, + "grad_norm": 1.7765018690997534, + "learning_rate": 3.4293131855653155e-06, + "loss": 0.8051, + "step": 2985 + }, + { + "epoch": 0.2692880010822023, + "grad_norm": 1.6434922367953602, + "learning_rate": 3.4289044835800102e-06, + "loss": 0.7175, + "step": 2986 + }, + { + "epoch": 0.26937818460567253, + "grad_norm": 1.7630308555720102, + "learning_rate": 3.4284956596726953e-06, + "loss": 0.7843, + "step": 2987 + }, + { + "epoch": 0.2694683681291428, + "grad_norm": 1.4715279625618538, + "learning_rate": 3.4280867138782544e-06, + "loss": 0.7726, + "step": 2988 + }, + { + "epoch": 0.26955855165261305, + "grad_norm": 2.282852132820049, + "learning_rate": 3.4276776462315803e-06, + "loss": 0.7856, + "step": 2989 + }, + { + "epoch": 0.26964873517608334, + "grad_norm": 1.4308721085132035, + "learning_rate": 3.427268456767578e-06, + "loss": 0.7802, + "step": 2990 + }, + { + "epoch": 0.2697389186995536, + "grad_norm": 2.609105627352526, + "learning_rate": 3.42685914552116e-06, + "loss": 0.8111, + "step": 2991 + }, + { + "epoch": 0.26982910222302386, + "grad_norm": 1.5599530633565513, + "learning_rate": 3.426449712527253e-06, + "loss": 0.7505, + "step": 2992 + }, + { + "epoch": 0.2699192857464941, + "grad_norm": 1.3984789708983767, + "learning_rate": 3.4260401578207904e-06, + "loss": 0.6951, + "step": 2993 + }, + { + "epoch": 0.2700094692699644, + "grad_norm": 1.9617696873209425, + "learning_rate": 3.4256304814367185e-06, + "loss": 0.7505, + "step": 2994 + }, + { + "epoch": 0.2700996527934346, + "grad_norm": 1.8240580518628107, + "learning_rate": 3.4252206834099936e-06, + "loss": 0.7394, + "step": 2995 + }, + { + "epoch": 0.2701898363169049, + "grad_norm": 1.739700046971669, + "learning_rate": 3.424810763775581e-06, + "loss": 0.776, + "step": 2996 + }, + { + "epoch": 0.27028001984037514, + "grad_norm": 1.794372540946931, + "learning_rate": 3.4244007225684587e-06, + "loss": 0.716, + "step": 2997 + }, + { + "epoch": 0.27037020336384543, + "grad_norm": 1.6738477102622773, + "learning_rate": 3.4239905598236115e-06, + "loss": 0.7864, + "step": 2998 + }, + { + "epoch": 0.27046038688731566, + "grad_norm": 1.5183555153497623, + "learning_rate": 3.4235802755760386e-06, + "loss": 0.8077, + "step": 2999 + }, + { + "epoch": 0.27055057041078595, + "grad_norm": 0.8496752833967948, + "learning_rate": 3.4231698698607464e-06, + "loss": 0.6954, + "step": 3000 + }, + { + "epoch": 0.2706407539342562, + "grad_norm": 1.6677857774348785, + "learning_rate": 3.4227593427127543e-06, + "loss": 0.8347, + "step": 3001 + }, + { + "epoch": 0.2707309374577265, + "grad_norm": 1.811444994932368, + "learning_rate": 3.42234869416709e-06, + "loss": 0.747, + "step": 3002 + }, + { + "epoch": 0.27082112098119676, + "grad_norm": 1.9316872185158787, + "learning_rate": 3.421937924258792e-06, + "loss": 0.7821, + "step": 3003 + }, + { + "epoch": 0.270911304504667, + "grad_norm": 2.055461201206266, + "learning_rate": 3.4215270330229096e-06, + "loss": 0.8242, + "step": 3004 + }, + { + "epoch": 0.2710014880281373, + "grad_norm": 2.1133775953517318, + "learning_rate": 3.421116020494503e-06, + "loss": 0.7436, + "step": 3005 + }, + { + "epoch": 0.2710916715516075, + "grad_norm": 1.4942517607003443, + "learning_rate": 3.420704886708642e-06, + "loss": 0.7256, + "step": 3006 + }, + { + "epoch": 0.2711818550750778, + "grad_norm": 1.142517557739638, + "learning_rate": 3.4202936317004056e-06, + "loss": 0.7296, + "step": 3007 + }, + { + "epoch": 0.27127203859854804, + "grad_norm": 1.8882678762584946, + "learning_rate": 3.4198822555048856e-06, + "loss": 0.797, + "step": 3008 + }, + { + "epoch": 0.2713622221220183, + "grad_norm": 1.7985721700111246, + "learning_rate": 3.419470758157182e-06, + "loss": 0.8652, + "step": 3009 + }, + { + "epoch": 0.27145240564548856, + "grad_norm": 1.6776002912487455, + "learning_rate": 3.4190591396924068e-06, + "loss": 0.7802, + "step": 3010 + }, + { + "epoch": 0.27154258916895885, + "grad_norm": 0.6829778021616402, + "learning_rate": 3.418647400145681e-06, + "loss": 0.616, + "step": 3011 + }, + { + "epoch": 0.2716327726924291, + "grad_norm": 0.69978516296184, + "learning_rate": 3.4182355395521367e-06, + "loss": 0.5765, + "step": 3012 + }, + { + "epoch": 0.27172295621589937, + "grad_norm": 1.6283472506196164, + "learning_rate": 3.417823557946916e-06, + "loss": 0.7761, + "step": 3013 + }, + { + "epoch": 0.2718131397393696, + "grad_norm": 1.4518491825299438, + "learning_rate": 3.417411455365172e-06, + "loss": 0.6707, + "step": 3014 + }, + { + "epoch": 0.2719033232628399, + "grad_norm": 1.8596022010283102, + "learning_rate": 3.416999231842066e-06, + "loss": 0.6895, + "step": 3015 + }, + { + "epoch": 0.2719935067863101, + "grad_norm": 1.8209455314421144, + "learning_rate": 3.416586887412773e-06, + "loss": 0.8287, + "step": 3016 + }, + { + "epoch": 0.2720836903097804, + "grad_norm": 2.403247793406342, + "learning_rate": 3.416174422112476e-06, + "loss": 0.7439, + "step": 3017 + }, + { + "epoch": 0.27217387383325065, + "grad_norm": 0.6193885908358305, + "learning_rate": 3.4157618359763687e-06, + "loss": 0.5407, + "step": 3018 + }, + { + "epoch": 0.27226405735672093, + "grad_norm": 1.3975621472701967, + "learning_rate": 3.4153491290396542e-06, + "loss": 0.6995, + "step": 3019 + }, + { + "epoch": 0.27235424088019117, + "grad_norm": 1.705041139950931, + "learning_rate": 3.4149363013375485e-06, + "loss": 0.7917, + "step": 3020 + }, + { + "epoch": 0.27244442440366146, + "grad_norm": 0.697458854849512, + "learning_rate": 3.414523352905276e-06, + "loss": 0.5558, + "step": 3021 + }, + { + "epoch": 0.2725346079271317, + "grad_norm": 2.4816200768286416, + "learning_rate": 3.414110283778071e-06, + "loss": 0.7525, + "step": 3022 + }, + { + "epoch": 0.272624791450602, + "grad_norm": 1.9082387961857479, + "learning_rate": 3.4136970939911797e-06, + "loss": 0.7305, + "step": 3023 + }, + { + "epoch": 0.2727149749740722, + "grad_norm": 1.536191014025758, + "learning_rate": 3.413283783579857e-06, + "loss": 0.7431, + "step": 3024 + }, + { + "epoch": 0.2728051584975425, + "grad_norm": 3.485462378281751, + "learning_rate": 3.412870352579369e-06, + "loss": 0.7214, + "step": 3025 + }, + { + "epoch": 0.2728953420210128, + "grad_norm": 0.6620239607819564, + "learning_rate": 3.4124568010249915e-06, + "loss": 0.5969, + "step": 3026 + }, + { + "epoch": 0.272985525544483, + "grad_norm": 1.5730685061835306, + "learning_rate": 3.4120431289520124e-06, + "loss": 0.7653, + "step": 3027 + }, + { + "epoch": 0.2730757090679533, + "grad_norm": 1.6859541983214996, + "learning_rate": 3.4116293363957276e-06, + "loss": 0.7845, + "step": 3028 + }, + { + "epoch": 0.27316589259142354, + "grad_norm": 1.8203412491652424, + "learning_rate": 3.4112154233914438e-06, + "loss": 0.7869, + "step": 3029 + }, + { + "epoch": 0.27325607611489383, + "grad_norm": 1.8507962472226704, + "learning_rate": 3.410801389974479e-06, + "loss": 0.6839, + "step": 3030 + }, + { + "epoch": 0.27334625963836406, + "grad_norm": 2.5305178489997746, + "learning_rate": 3.410387236180161e-06, + "loss": 0.8116, + "step": 3031 + }, + { + "epoch": 0.27343644316183435, + "grad_norm": 1.760147781453288, + "learning_rate": 3.409972962043826e-06, + "loss": 0.6614, + "step": 3032 + }, + { + "epoch": 0.2735266266853046, + "grad_norm": 2.133342746872796, + "learning_rate": 3.4095585676008234e-06, + "loss": 0.8343, + "step": 3033 + }, + { + "epoch": 0.2736168102087749, + "grad_norm": 1.493446258563939, + "learning_rate": 3.4091440528865125e-06, + "loss": 0.7733, + "step": 3034 + }, + { + "epoch": 0.2737069937322451, + "grad_norm": 3.226661957692685, + "learning_rate": 3.4087294179362606e-06, + "loss": 0.7782, + "step": 3035 + }, + { + "epoch": 0.2737971772557154, + "grad_norm": 4.771528146225245, + "learning_rate": 3.4083146627854474e-06, + "loss": 0.8239, + "step": 3036 + }, + { + "epoch": 0.27388736077918563, + "grad_norm": 2.0564386716045777, + "learning_rate": 3.4078997874694614e-06, + "loss": 0.7097, + "step": 3037 + }, + { + "epoch": 0.2739775443026559, + "grad_norm": 1.6857974329224796, + "learning_rate": 3.407484792023703e-06, + "loss": 0.6731, + "step": 3038 + }, + { + "epoch": 0.27406772782612615, + "grad_norm": 1.8316874557694522, + "learning_rate": 3.407069676483581e-06, + "loss": 0.7495, + "step": 3039 + }, + { + "epoch": 0.27415791134959644, + "grad_norm": 1.4044570937564522, + "learning_rate": 3.406654440884516e-06, + "loss": 0.8156, + "step": 3040 + }, + { + "epoch": 0.2742480948730667, + "grad_norm": 1.712464300187275, + "learning_rate": 3.4062390852619372e-06, + "loss": 0.8092, + "step": 3041 + }, + { + "epoch": 0.27433827839653696, + "grad_norm": 1.7688549509352427, + "learning_rate": 3.4058236096512867e-06, + "loss": 0.7733, + "step": 3042 + }, + { + "epoch": 0.2744284619200072, + "grad_norm": 1.9821698409100823, + "learning_rate": 3.405408014088013e-06, + "loss": 0.6749, + "step": 3043 + }, + { + "epoch": 0.2745186454434775, + "grad_norm": 2.104845963412431, + "learning_rate": 3.404992298607579e-06, + "loss": 0.665, + "step": 3044 + }, + { + "epoch": 0.2746088289669477, + "grad_norm": 2.3520193068172532, + "learning_rate": 3.4045764632454547e-06, + "loss": 0.7544, + "step": 3045 + }, + { + "epoch": 0.274699012490418, + "grad_norm": 1.781534205735451, + "learning_rate": 3.4041605080371223e-06, + "loss": 0.7348, + "step": 3046 + }, + { + "epoch": 0.27478919601388824, + "grad_norm": 3.953647382998306, + "learning_rate": 3.4037444330180726e-06, + "loss": 0.7886, + "step": 3047 + }, + { + "epoch": 0.2748793795373585, + "grad_norm": 1.982792621389654, + "learning_rate": 3.403328238223808e-06, + "loss": 0.7914, + "step": 3048 + }, + { + "epoch": 0.27496956306082876, + "grad_norm": 4.601680216554567, + "learning_rate": 3.4029119236898395e-06, + "loss": 0.8085, + "step": 3049 + }, + { + "epoch": 0.27505974658429905, + "grad_norm": 1.069364420193586, + "learning_rate": 3.4024954894516906e-06, + "loss": 0.6095, + "step": 3050 + }, + { + "epoch": 0.27514993010776934, + "grad_norm": 1.9867373844762393, + "learning_rate": 3.4020789355448933e-06, + "loss": 0.7798, + "step": 3051 + }, + { + "epoch": 0.27524011363123957, + "grad_norm": 1.593180984944069, + "learning_rate": 3.40166226200499e-06, + "loss": 0.8482, + "step": 3052 + }, + { + "epoch": 0.27533029715470986, + "grad_norm": 2.10036584755442, + "learning_rate": 3.401245468867534e-06, + "loss": 0.7636, + "step": 3053 + }, + { + "epoch": 0.2754204806781801, + "grad_norm": 0.712993457641119, + "learning_rate": 3.400828556168088e-06, + "loss": 0.6333, + "step": 3054 + }, + { + "epoch": 0.2755106642016504, + "grad_norm": 2.023760324302195, + "learning_rate": 3.4004115239422255e-06, + "loss": 0.7903, + "step": 3055 + }, + { + "epoch": 0.2756008477251206, + "grad_norm": 1.9008481767680543, + "learning_rate": 3.3999943722255305e-06, + "loss": 0.7642, + "step": 3056 + }, + { + "epoch": 0.2756910312485909, + "grad_norm": 1.6153268476259703, + "learning_rate": 3.3995771010535955e-06, + "loss": 0.6757, + "step": 3057 + }, + { + "epoch": 0.27578121477206113, + "grad_norm": 0.7013232139542087, + "learning_rate": 3.3991597104620253e-06, + "loss": 0.5397, + "step": 3058 + }, + { + "epoch": 0.2758713982955314, + "grad_norm": 1.6670084920665733, + "learning_rate": 3.398742200486434e-06, + "loss": 0.7548, + "step": 3059 + }, + { + "epoch": 0.27596158181900166, + "grad_norm": 1.7505892715158478, + "learning_rate": 3.3983245711624453e-06, + "loss": 0.7946, + "step": 3060 + }, + { + "epoch": 0.27605176534247194, + "grad_norm": 0.7926906229481142, + "learning_rate": 3.3979068225256946e-06, + "loss": 0.6554, + "step": 3061 + }, + { + "epoch": 0.2761419488659422, + "grad_norm": 1.8968703469434602, + "learning_rate": 3.3974889546118246e-06, + "loss": 0.7473, + "step": 3062 + }, + { + "epoch": 0.27623213238941247, + "grad_norm": 1.5552659995171727, + "learning_rate": 3.3970709674564918e-06, + "loss": 0.7165, + "step": 3063 + }, + { + "epoch": 0.2763223159128827, + "grad_norm": 2.011541325420816, + "learning_rate": 3.3966528610953607e-06, + "loss": 0.7697, + "step": 3064 + }, + { + "epoch": 0.276412499436353, + "grad_norm": 1.7132061414349813, + "learning_rate": 3.3962346355641067e-06, + "loss": 0.7805, + "step": 3065 + }, + { + "epoch": 0.2765026829598232, + "grad_norm": 1.6399190201188616, + "learning_rate": 3.3958162908984146e-06, + "loss": 0.8071, + "step": 3066 + }, + { + "epoch": 0.2765928664832935, + "grad_norm": 1.6024590936638117, + "learning_rate": 3.39539782713398e-06, + "loss": 0.7672, + "step": 3067 + }, + { + "epoch": 0.27668305000676374, + "grad_norm": 1.8996743148946054, + "learning_rate": 3.394979244306509e-06, + "loss": 0.8019, + "step": 3068 + }, + { + "epoch": 0.27677323353023403, + "grad_norm": 1.5788941727845947, + "learning_rate": 3.3945605424517166e-06, + "loss": 0.7839, + "step": 3069 + }, + { + "epoch": 0.27686341705370426, + "grad_norm": 2.173862234226727, + "learning_rate": 3.3941417216053294e-06, + "loss": 0.771, + "step": 3070 + }, + { + "epoch": 0.27695360057717455, + "grad_norm": 3.393369205812354, + "learning_rate": 3.3937227818030835e-06, + "loss": 0.7402, + "step": 3071 + }, + { + "epoch": 0.2770437841006448, + "grad_norm": 1.7857335866157509, + "learning_rate": 3.393303723080725e-06, + "loss": 0.734, + "step": 3072 + }, + { + "epoch": 0.2771339676241151, + "grad_norm": 1.7196973790601253, + "learning_rate": 3.3928845454740097e-06, + "loss": 0.7875, + "step": 3073 + }, + { + "epoch": 0.27722415114758536, + "grad_norm": 1.8155019170172637, + "learning_rate": 3.392465249018705e-06, + "loss": 0.7612, + "step": 3074 + }, + { + "epoch": 0.2773143346710556, + "grad_norm": 1.6762074459855913, + "learning_rate": 3.3920458337505872e-06, + "loss": 0.7721, + "step": 3075 + }, + { + "epoch": 0.2774045181945259, + "grad_norm": 2.231985134546963, + "learning_rate": 3.391626299705443e-06, + "loss": 0.6169, + "step": 3076 + }, + { + "epoch": 0.2774947017179961, + "grad_norm": 2.1325691279915975, + "learning_rate": 3.39120664691907e-06, + "loss": 0.7831, + "step": 3077 + }, + { + "epoch": 0.2775848852414664, + "grad_norm": 1.6797194366967496, + "learning_rate": 3.390786875427275e-06, + "loss": 0.7457, + "step": 3078 + }, + { + "epoch": 0.27767506876493664, + "grad_norm": 2.0868840159643867, + "learning_rate": 3.390366985265875e-06, + "loss": 0.7661, + "step": 3079 + }, + { + "epoch": 0.2777652522884069, + "grad_norm": 1.5222639992744158, + "learning_rate": 3.389946976470697e-06, + "loss": 0.7285, + "step": 3080 + }, + { + "epoch": 0.27785543581187716, + "grad_norm": 1.5570000553020842, + "learning_rate": 3.3895268490775787e-06, + "loss": 0.8654, + "step": 3081 + }, + { + "epoch": 0.27794561933534745, + "grad_norm": 1.7006484057334443, + "learning_rate": 3.3891066031223685e-06, + "loss": 0.7291, + "step": 3082 + }, + { + "epoch": 0.2780358028588177, + "grad_norm": 1.6912106583693398, + "learning_rate": 3.3886862386409237e-06, + "loss": 0.8053, + "step": 3083 + }, + { + "epoch": 0.27812598638228797, + "grad_norm": 2.4403002377569707, + "learning_rate": 3.388265755669111e-06, + "loss": 0.7395, + "step": 3084 + }, + { + "epoch": 0.2782161699057582, + "grad_norm": 2.0149646239097874, + "learning_rate": 3.3878451542428093e-06, + "loss": 0.8383, + "step": 3085 + }, + { + "epoch": 0.2783063534292285, + "grad_norm": 1.43948649933219, + "learning_rate": 3.387424434397907e-06, + "loss": 0.7556, + "step": 3086 + }, + { + "epoch": 0.2783965369526987, + "grad_norm": 5.21198460068554, + "learning_rate": 3.3870035961703013e-06, + "loss": 0.6869, + "step": 3087 + }, + { + "epoch": 0.278486720476169, + "grad_norm": 6.566484386153943, + "learning_rate": 3.3865826395959014e-06, + "loss": 0.8976, + "step": 3088 + }, + { + "epoch": 0.27857690399963925, + "grad_norm": 1.724814310998412, + "learning_rate": 3.3861615647106253e-06, + "loss": 0.8588, + "step": 3089 + }, + { + "epoch": 0.27866708752310954, + "grad_norm": 2.430067811498569, + "learning_rate": 3.3857403715504012e-06, + "loss": 0.7892, + "step": 3090 + }, + { + "epoch": 0.27875727104657977, + "grad_norm": 1.6828762570118412, + "learning_rate": 3.385319060151167e-06, + "loss": 0.7381, + "step": 3091 + }, + { + "epoch": 0.27884745457005006, + "grad_norm": 2.1694221092742665, + "learning_rate": 3.3848976305488728e-06, + "loss": 0.7696, + "step": 3092 + }, + { + "epoch": 0.2789376380935203, + "grad_norm": 2.3052220645937442, + "learning_rate": 3.384476082779476e-06, + "loss": 0.8165, + "step": 3093 + }, + { + "epoch": 0.2790278216169906, + "grad_norm": 2.5416448519244796, + "learning_rate": 3.3840544168789463e-06, + "loss": 0.7485, + "step": 3094 + }, + { + "epoch": 0.2791180051404608, + "grad_norm": 1.5991931788380624, + "learning_rate": 3.3836326328832617e-06, + "loss": 0.8766, + "step": 3095 + }, + { + "epoch": 0.2792081886639311, + "grad_norm": 2.1788382020694237, + "learning_rate": 3.383210730828412e-06, + "loss": 0.728, + "step": 3096 + }, + { + "epoch": 0.2792983721874014, + "grad_norm": 1.5059485106056778, + "learning_rate": 3.3827887107503953e-06, + "loss": 0.8409, + "step": 3097 + }, + { + "epoch": 0.2793885557108716, + "grad_norm": 2.8383575286401435, + "learning_rate": 3.3823665726852216e-06, + "loss": 0.7191, + "step": 3098 + }, + { + "epoch": 0.2794787392343419, + "grad_norm": 1.501774056701877, + "learning_rate": 3.3819443166689095e-06, + "loss": 0.8969, + "step": 3099 + }, + { + "epoch": 0.27956892275781214, + "grad_norm": 1.547713543728277, + "learning_rate": 3.3815219427374886e-06, + "loss": 0.7678, + "step": 3100 + }, + { + "epoch": 0.27965910628128243, + "grad_norm": 1.5865743326253443, + "learning_rate": 3.3810994509269975e-06, + "loss": 0.7925, + "step": 3101 + }, + { + "epoch": 0.27974928980475267, + "grad_norm": 1.758848300510241, + "learning_rate": 3.3806768412734864e-06, + "loss": 0.8032, + "step": 3102 + }, + { + "epoch": 0.27983947332822295, + "grad_norm": 1.5919833395867333, + "learning_rate": 3.380254113813014e-06, + "loss": 0.854, + "step": 3103 + }, + { + "epoch": 0.2799296568516932, + "grad_norm": 1.6459451066061017, + "learning_rate": 3.3798312685816496e-06, + "loss": 0.804, + "step": 3104 + }, + { + "epoch": 0.2800198403751635, + "grad_norm": 2.577274344007833, + "learning_rate": 3.3794083056154738e-06, + "loss": 0.7743, + "step": 3105 + }, + { + "epoch": 0.2801100238986337, + "grad_norm": 1.6180255579680418, + "learning_rate": 3.3789852249505746e-06, + "loss": 0.7821, + "step": 3106 + }, + { + "epoch": 0.280200207422104, + "grad_norm": 1.7051366422384522, + "learning_rate": 3.378562026623053e-06, + "loss": 0.8299, + "step": 3107 + }, + { + "epoch": 0.28029039094557423, + "grad_norm": 1.6591662518079089, + "learning_rate": 3.3781387106690175e-06, + "loss": 0.8473, + "step": 3108 + }, + { + "epoch": 0.2803805744690445, + "grad_norm": 1.4680750493191244, + "learning_rate": 3.3777152771245885e-06, + "loss": 0.8011, + "step": 3109 + }, + { + "epoch": 0.28047075799251475, + "grad_norm": 1.7052380114353307, + "learning_rate": 3.377291726025895e-06, + "loss": 0.8022, + "step": 3110 + }, + { + "epoch": 0.28056094151598504, + "grad_norm": 1.9522540782837507, + "learning_rate": 3.3768680574090782e-06, + "loss": 0.8878, + "step": 3111 + }, + { + "epoch": 0.2806511250394553, + "grad_norm": 2.0426065915331466, + "learning_rate": 3.3764442713102857e-06, + "loss": 0.7496, + "step": 3112 + }, + { + "epoch": 0.28074130856292556, + "grad_norm": 1.512972447403512, + "learning_rate": 3.3760203677656786e-06, + "loss": 0.7711, + "step": 3113 + }, + { + "epoch": 0.2808314920863958, + "grad_norm": 1.8521933466788627, + "learning_rate": 3.3755963468114262e-06, + "loss": 0.8293, + "step": 3114 + }, + { + "epoch": 0.2809216756098661, + "grad_norm": 1.6432319637059156, + "learning_rate": 3.3751722084837095e-06, + "loss": 0.6758, + "step": 3115 + }, + { + "epoch": 0.2810118591333363, + "grad_norm": 1.6214190420995456, + "learning_rate": 3.3747479528187166e-06, + "loss": 0.7532, + "step": 3116 + }, + { + "epoch": 0.2811020426568066, + "grad_norm": 1.8664001337001472, + "learning_rate": 3.3743235798526485e-06, + "loss": 0.8264, + "step": 3117 + }, + { + "epoch": 0.28119222618027684, + "grad_norm": 1.4449662077229286, + "learning_rate": 3.373899089621714e-06, + "loss": 0.7053, + "step": 3118 + }, + { + "epoch": 0.2812824097037471, + "grad_norm": 1.7806435606749993, + "learning_rate": 3.373474482162134e-06, + "loss": 0.754, + "step": 3119 + }, + { + "epoch": 0.28137259322721736, + "grad_norm": 2.797924079421263, + "learning_rate": 3.3730497575101376e-06, + "loss": 0.7791, + "step": 3120 + }, + { + "epoch": 0.28146277675068765, + "grad_norm": 1.615453850481619, + "learning_rate": 3.3726249157019654e-06, + "loss": 0.7705, + "step": 3121 + }, + { + "epoch": 0.28155296027415794, + "grad_norm": 2.081430201705143, + "learning_rate": 3.372199956773866e-06, + "loss": 0.8607, + "step": 3122 + }, + { + "epoch": 0.28164314379762817, + "grad_norm": 1.7586193352307822, + "learning_rate": 3.371774880762101e-06, + "loss": 0.7282, + "step": 3123 + }, + { + "epoch": 0.28173332732109846, + "grad_norm": 1.8150752615759316, + "learning_rate": 3.3713496877029392e-06, + "loss": 0.7631, + "step": 3124 + }, + { + "epoch": 0.2818235108445687, + "grad_norm": 1.993117470689669, + "learning_rate": 3.37092437763266e-06, + "loss": 0.8383, + "step": 3125 + }, + { + "epoch": 0.281913694368039, + "grad_norm": 1.8872671021527834, + "learning_rate": 3.3704989505875537e-06, + "loss": 0.8096, + "step": 3126 + }, + { + "epoch": 0.2820038778915092, + "grad_norm": 1.6175804213705853, + "learning_rate": 3.3700734066039205e-06, + "loss": 0.7445, + "step": 3127 + }, + { + "epoch": 0.2820940614149795, + "grad_norm": 2.7586888982093054, + "learning_rate": 3.36964774571807e-06, + "loss": 0.8065, + "step": 3128 + }, + { + "epoch": 0.28218424493844974, + "grad_norm": 2.0337240557665788, + "learning_rate": 3.3692219679663206e-06, + "loss": 0.774, + "step": 3129 + }, + { + "epoch": 0.28227442846192, + "grad_norm": 2.041017494018283, + "learning_rate": 3.3687960733850043e-06, + "loss": 0.7955, + "step": 3130 + }, + { + "epoch": 0.28236461198539026, + "grad_norm": 1.7150734853988014, + "learning_rate": 3.3683700620104586e-06, + "loss": 0.7154, + "step": 3131 + }, + { + "epoch": 0.28245479550886055, + "grad_norm": 1.9391137447548161, + "learning_rate": 3.3679439338790347e-06, + "loss": 0.79, + "step": 3132 + }, + { + "epoch": 0.2825449790323308, + "grad_norm": 1.5794405599859358, + "learning_rate": 3.3675176890270916e-06, + "loss": 0.8988, + "step": 3133 + }, + { + "epoch": 0.28263516255580107, + "grad_norm": 1.5381543728574842, + "learning_rate": 3.367091327490998e-06, + "loss": 0.8003, + "step": 3134 + }, + { + "epoch": 0.2827253460792713, + "grad_norm": 1.7242129119649763, + "learning_rate": 3.3666648493071347e-06, + "loss": 0.767, + "step": 3135 + }, + { + "epoch": 0.2828155296027416, + "grad_norm": 1.703825220475237, + "learning_rate": 3.3662382545118914e-06, + "loss": 0.7546, + "step": 3136 + }, + { + "epoch": 0.2829057131262118, + "grad_norm": 1.7056289604591135, + "learning_rate": 3.3658115431416663e-06, + "loss": 0.8574, + "step": 3137 + }, + { + "epoch": 0.2829958966496821, + "grad_norm": 0.6979761643930739, + "learning_rate": 3.36538471523287e-06, + "loss": 0.609, + "step": 3138 + }, + { + "epoch": 0.28308608017315234, + "grad_norm": 1.493458630862733, + "learning_rate": 3.3649577708219204e-06, + "loss": 0.8322, + "step": 3139 + }, + { + "epoch": 0.28317626369662263, + "grad_norm": 1.678664423734779, + "learning_rate": 3.3645307099452477e-06, + "loss": 0.7244, + "step": 3140 + }, + { + "epoch": 0.28326644722009287, + "grad_norm": 1.8340350094306435, + "learning_rate": 3.3641035326392907e-06, + "loss": 0.7746, + "step": 3141 + }, + { + "epoch": 0.28335663074356315, + "grad_norm": 2.112186620819049, + "learning_rate": 3.363676238940499e-06, + "loss": 0.7902, + "step": 3142 + }, + { + "epoch": 0.2834468142670334, + "grad_norm": 1.808640217853548, + "learning_rate": 3.363248828885331e-06, + "loss": 0.8049, + "step": 3143 + }, + { + "epoch": 0.2835369977905037, + "grad_norm": 1.4391575254861617, + "learning_rate": 3.3628213025102562e-06, + "loss": 0.802, + "step": 3144 + }, + { + "epoch": 0.28362718131397396, + "grad_norm": 1.7801137947826504, + "learning_rate": 3.3623936598517536e-06, + "loss": 0.8164, + "step": 3145 + }, + { + "epoch": 0.2837173648374442, + "grad_norm": 2.0708438983036257, + "learning_rate": 3.3619659009463117e-06, + "loss": 0.6257, + "step": 3146 + }, + { + "epoch": 0.2838075483609145, + "grad_norm": 1.5771144993204123, + "learning_rate": 3.3615380258304287e-06, + "loss": 0.8261, + "step": 3147 + }, + { + "epoch": 0.2838977318843847, + "grad_norm": 1.8503924829829244, + "learning_rate": 3.3611100345406146e-06, + "loss": 0.7647, + "step": 3148 + }, + { + "epoch": 0.283987915407855, + "grad_norm": 1.6396919140211963, + "learning_rate": 3.3606819271133873e-06, + "loss": 0.765, + "step": 3149 + }, + { + "epoch": 0.28407809893132524, + "grad_norm": 0.7189409168277952, + "learning_rate": 3.360253703585275e-06, + "loss": 0.5934, + "step": 3150 + }, + { + "epoch": 0.28416828245479553, + "grad_norm": 1.907551868078504, + "learning_rate": 3.3598253639928164e-06, + "loss": 0.7193, + "step": 3151 + }, + { + "epoch": 0.28425846597826576, + "grad_norm": 1.8914223873138585, + "learning_rate": 3.3593969083725596e-06, + "loss": 0.7344, + "step": 3152 + }, + { + "epoch": 0.28434864950173605, + "grad_norm": 3.028544368777242, + "learning_rate": 3.358968336761063e-06, + "loss": 0.7351, + "step": 3153 + }, + { + "epoch": 0.2844388330252063, + "grad_norm": 1.6699928313232792, + "learning_rate": 3.3585396491948945e-06, + "loss": 0.7399, + "step": 3154 + }, + { + "epoch": 0.28452901654867657, + "grad_norm": 1.7438057538488545, + "learning_rate": 3.358110845710633e-06, + "loss": 0.8327, + "step": 3155 + }, + { + "epoch": 0.2846192000721468, + "grad_norm": 1.9111084856094147, + "learning_rate": 3.357681926344865e-06, + "loss": 0.7648, + "step": 3156 + }, + { + "epoch": 0.2847093835956171, + "grad_norm": 1.60842600870039, + "learning_rate": 3.357252891134189e-06, + "loss": 0.7744, + "step": 3157 + }, + { + "epoch": 0.2847995671190873, + "grad_norm": 1.8382579018941665, + "learning_rate": 3.356823740115212e-06, + "loss": 0.8232, + "step": 3158 + }, + { + "epoch": 0.2848897506425576, + "grad_norm": 1.2743467115823703, + "learning_rate": 3.3563944733245525e-06, + "loss": 0.7197, + "step": 3159 + }, + { + "epoch": 0.28497993416602785, + "grad_norm": 1.8552697004663516, + "learning_rate": 3.3559650907988375e-06, + "loss": 0.767, + "step": 3160 + }, + { + "epoch": 0.28507011768949814, + "grad_norm": 2.804781941388495, + "learning_rate": 3.3555355925747045e-06, + "loss": 0.6213, + "step": 3161 + }, + { + "epoch": 0.28516030121296837, + "grad_norm": 1.7517501390254848, + "learning_rate": 3.3551059786888e-06, + "loss": 0.7633, + "step": 3162 + }, + { + "epoch": 0.28525048473643866, + "grad_norm": 1.7958554900086652, + "learning_rate": 3.3546762491777807e-06, + "loss": 0.8224, + "step": 3163 + }, + { + "epoch": 0.2853406682599089, + "grad_norm": 1.6528378733048528, + "learning_rate": 3.3542464040783156e-06, + "loss": 0.7697, + "step": 3164 + }, + { + "epoch": 0.2854308517833792, + "grad_norm": 1.9745488218838885, + "learning_rate": 3.353816443427079e-06, + "loss": 0.7923, + "step": 3165 + }, + { + "epoch": 0.2855210353068494, + "grad_norm": 1.893726484700362, + "learning_rate": 3.3533863672607597e-06, + "loss": 0.7552, + "step": 3166 + }, + { + "epoch": 0.2856112188303197, + "grad_norm": 1.613282062816646, + "learning_rate": 3.352956175616052e-06, + "loss": 0.8033, + "step": 3167 + }, + { + "epoch": 0.28570140235378993, + "grad_norm": 1.5984057321682426, + "learning_rate": 3.352525868529664e-06, + "loss": 0.7415, + "step": 3168 + }, + { + "epoch": 0.2857915858772602, + "grad_norm": 1.911132126361176, + "learning_rate": 3.3520954460383103e-06, + "loss": 0.7179, + "step": 3169 + }, + { + "epoch": 0.2858817694007305, + "grad_norm": 1.783152785327074, + "learning_rate": 3.3516649081787182e-06, + "loss": 0.7114, + "step": 3170 + }, + { + "epoch": 0.28597195292420075, + "grad_norm": 2.2011418717107247, + "learning_rate": 3.3512342549876236e-06, + "loss": 0.8743, + "step": 3171 + }, + { + "epoch": 0.28606213644767103, + "grad_norm": 1.8050992421687981, + "learning_rate": 3.350803486501771e-06, + "loss": 0.7303, + "step": 3172 + }, + { + "epoch": 0.28615231997114127, + "grad_norm": 1.6610583917182538, + "learning_rate": 3.3503726027579175e-06, + "loss": 0.8011, + "step": 3173 + }, + { + "epoch": 0.28624250349461156, + "grad_norm": 1.4256173653672024, + "learning_rate": 3.349941603792827e-06, + "loss": 0.8273, + "step": 3174 + }, + { + "epoch": 0.2863326870180818, + "grad_norm": 1.6443668382964793, + "learning_rate": 3.3495104896432755e-06, + "loss": 0.7697, + "step": 3175 + }, + { + "epoch": 0.2864228705415521, + "grad_norm": 1.5729734734725476, + "learning_rate": 3.3490792603460477e-06, + "loss": 0.7385, + "step": 3176 + }, + { + "epoch": 0.2865130540650223, + "grad_norm": 2.0108165310367028, + "learning_rate": 3.3486479159379393e-06, + "loss": 0.7075, + "step": 3177 + }, + { + "epoch": 0.2866032375884926, + "grad_norm": 0.6730158900004742, + "learning_rate": 3.3482164564557537e-06, + "loss": 0.5773, + "step": 3178 + }, + { + "epoch": 0.28669342111196283, + "grad_norm": 2.174970438087648, + "learning_rate": 3.3477848819363065e-06, + "loss": 0.8309, + "step": 3179 + }, + { + "epoch": 0.2867836046354331, + "grad_norm": 0.6594642985472572, + "learning_rate": 3.3473531924164213e-06, + "loss": 0.6053, + "step": 3180 + }, + { + "epoch": 0.28687378815890335, + "grad_norm": 3.6344628307852282, + "learning_rate": 3.3469213879329325e-06, + "loss": 0.7278, + "step": 3181 + }, + { + "epoch": 0.28696397168237364, + "grad_norm": 1.8474888796336157, + "learning_rate": 3.3464894685226837e-06, + "loss": 0.7346, + "step": 3182 + }, + { + "epoch": 0.2870541552058439, + "grad_norm": 1.6282125409217671, + "learning_rate": 3.34605743422253e-06, + "loss": 0.8484, + "step": 3183 + }, + { + "epoch": 0.28714433872931416, + "grad_norm": 1.7997884599571954, + "learning_rate": 3.345625285069333e-06, + "loss": 0.7428, + "step": 3184 + }, + { + "epoch": 0.2872345222527844, + "grad_norm": 1.898101086106323, + "learning_rate": 3.345193021099967e-06, + "loss": 0.7533, + "step": 3185 + }, + { + "epoch": 0.2873247057762547, + "grad_norm": 1.9773784657011546, + "learning_rate": 3.3447606423513157e-06, + "loss": 0.7425, + "step": 3186 + }, + { + "epoch": 0.2874148892997249, + "grad_norm": 3.5478336870334846, + "learning_rate": 3.344328148860271e-06, + "loss": 0.7438, + "step": 3187 + }, + { + "epoch": 0.2875050728231952, + "grad_norm": 1.396363025007805, + "learning_rate": 3.3438955406637365e-06, + "loss": 0.7841, + "step": 3188 + }, + { + "epoch": 0.28759525634666544, + "grad_norm": 1.5570191960459154, + "learning_rate": 3.343462817798624e-06, + "loss": 0.7967, + "step": 3189 + }, + { + "epoch": 0.28768543987013573, + "grad_norm": 1.8848747261371486, + "learning_rate": 3.343029980301856e-06, + "loss": 0.739, + "step": 3190 + }, + { + "epoch": 0.28777562339360596, + "grad_norm": 2.0456113918934573, + "learning_rate": 3.342597028210365e-06, + "loss": 0.8079, + "step": 3191 + }, + { + "epoch": 0.28786580691707625, + "grad_norm": 2.046581218107899, + "learning_rate": 3.342163961561092e-06, + "loss": 0.7187, + "step": 3192 + }, + { + "epoch": 0.28795599044054654, + "grad_norm": 1.5848367563653938, + "learning_rate": 3.34173078039099e-06, + "loss": 0.7456, + "step": 3193 + }, + { + "epoch": 0.28804617396401677, + "grad_norm": 1.8504781646347295, + "learning_rate": 3.3412974847370193e-06, + "loss": 0.9401, + "step": 3194 + }, + { + "epoch": 0.28813635748748706, + "grad_norm": 1.4152890781545842, + "learning_rate": 3.3408640746361514e-06, + "loss": 0.7953, + "step": 3195 + }, + { + "epoch": 0.2882265410109573, + "grad_norm": 0.7557638339594585, + "learning_rate": 3.3404305501253663e-06, + "loss": 0.5997, + "step": 3196 + }, + { + "epoch": 0.2883167245344276, + "grad_norm": 1.5214029664968627, + "learning_rate": 3.3399969112416565e-06, + "loss": 0.8489, + "step": 3197 + }, + { + "epoch": 0.2884069080578978, + "grad_norm": 1.6806453369090462, + "learning_rate": 3.3395631580220213e-06, + "loss": 0.7097, + "step": 3198 + }, + { + "epoch": 0.2884970915813681, + "grad_norm": 1.6529344445283742, + "learning_rate": 3.3391292905034714e-06, + "loss": 0.7523, + "step": 3199 + }, + { + "epoch": 0.28858727510483834, + "grad_norm": 1.6184152558973361, + "learning_rate": 3.338695308723027e-06, + "loss": 0.8396, + "step": 3200 + }, + { + "epoch": 0.2886774586283086, + "grad_norm": 1.7173220164378287, + "learning_rate": 3.338261212717716e-06, + "loss": 0.8288, + "step": 3201 + }, + { + "epoch": 0.28876764215177886, + "grad_norm": 1.815106129656642, + "learning_rate": 3.33782700252458e-06, + "loss": 0.8071, + "step": 3202 + }, + { + "epoch": 0.28885782567524915, + "grad_norm": 2.129984563314889, + "learning_rate": 3.337392678180668e-06, + "loss": 0.8185, + "step": 3203 + }, + { + "epoch": 0.2889480091987194, + "grad_norm": 1.9262757198490577, + "learning_rate": 3.3369582397230377e-06, + "loss": 0.7776, + "step": 3204 + }, + { + "epoch": 0.28903819272218967, + "grad_norm": 1.7504173189940035, + "learning_rate": 3.336523687188759e-06, + "loss": 0.8175, + "step": 3205 + }, + { + "epoch": 0.2891283762456599, + "grad_norm": 0.7194434013956463, + "learning_rate": 3.336089020614909e-06, + "loss": 0.6138, + "step": 3206 + }, + { + "epoch": 0.2892185597691302, + "grad_norm": 1.6896321990084189, + "learning_rate": 3.3356542400385774e-06, + "loss": 0.6548, + "step": 3207 + }, + { + "epoch": 0.2893087432926004, + "grad_norm": 2.0311408380373015, + "learning_rate": 3.3352193454968607e-06, + "loss": 0.71, + "step": 3208 + }, + { + "epoch": 0.2893989268160707, + "grad_norm": 2.2526769607924386, + "learning_rate": 3.3347843370268675e-06, + "loss": 0.8359, + "step": 3209 + }, + { + "epoch": 0.28948911033954094, + "grad_norm": 1.4205065839078457, + "learning_rate": 3.334349214665715e-06, + "loss": 0.7887, + "step": 3210 + }, + { + "epoch": 0.28957929386301123, + "grad_norm": 1.5590267488192546, + "learning_rate": 3.3339139784505293e-06, + "loss": 0.8586, + "step": 3211 + }, + { + "epoch": 0.28966947738648147, + "grad_norm": 2.060110846935222, + "learning_rate": 3.333478628418448e-06, + "loss": 0.8421, + "step": 3212 + }, + { + "epoch": 0.28975966090995176, + "grad_norm": 2.03658673466207, + "learning_rate": 3.333043164606618e-06, + "loss": 0.668, + "step": 3213 + }, + { + "epoch": 0.289849844433422, + "grad_norm": 2.0728616243910953, + "learning_rate": 3.3326075870521948e-06, + "loss": 0.8586, + "step": 3214 + }, + { + "epoch": 0.2899400279568923, + "grad_norm": 1.8828300380285685, + "learning_rate": 3.3321718957923437e-06, + "loss": 0.7752, + "step": 3215 + }, + { + "epoch": 0.29003021148036257, + "grad_norm": 1.619550222901257, + "learning_rate": 3.3317360908642413e-06, + "loss": 0.7718, + "step": 3216 + }, + { + "epoch": 0.2901203950038328, + "grad_norm": 2.326727428444431, + "learning_rate": 3.331300172305072e-06, + "loss": 0.7318, + "step": 3217 + }, + { + "epoch": 0.2902105785273031, + "grad_norm": 1.6781546762353599, + "learning_rate": 3.330864140152032e-06, + "loss": 0.765, + "step": 3218 + }, + { + "epoch": 0.2903007620507733, + "grad_norm": 1.8240623652144607, + "learning_rate": 3.330427994442325e-06, + "loss": 0.7914, + "step": 3219 + }, + { + "epoch": 0.2903909455742436, + "grad_norm": 1.883553260295258, + "learning_rate": 3.3299917352131657e-06, + "loss": 0.8136, + "step": 3220 + }, + { + "epoch": 0.29048112909771384, + "grad_norm": 2.5444288617933593, + "learning_rate": 3.329555362501778e-06, + "loss": 0.7438, + "step": 3221 + }, + { + "epoch": 0.29057131262118413, + "grad_norm": 1.8809539992950357, + "learning_rate": 3.3291188763453954e-06, + "loss": 0.7411, + "step": 3222 + }, + { + "epoch": 0.29066149614465436, + "grad_norm": 2.7382492956019364, + "learning_rate": 3.3286822767812618e-06, + "loss": 0.7659, + "step": 3223 + }, + { + "epoch": 0.29075167966812465, + "grad_norm": 4.203395905680617, + "learning_rate": 3.32824556384663e-06, + "loss": 0.758, + "step": 3224 + }, + { + "epoch": 0.2908418631915949, + "grad_norm": 2.0974476744316553, + "learning_rate": 3.3278087375787628e-06, + "loss": 0.7788, + "step": 3225 + }, + { + "epoch": 0.2909320467150652, + "grad_norm": 2.312078489086663, + "learning_rate": 3.327371798014933e-06, + "loss": 0.7804, + "step": 3226 + }, + { + "epoch": 0.2910222302385354, + "grad_norm": 1.7816202548517932, + "learning_rate": 3.3269347451924218e-06, + "loss": 0.7757, + "step": 3227 + }, + { + "epoch": 0.2911124137620057, + "grad_norm": 1.6044356463285896, + "learning_rate": 3.326497579148522e-06, + "loss": 0.8398, + "step": 3228 + }, + { + "epoch": 0.29120259728547593, + "grad_norm": 1.7788563757779188, + "learning_rate": 3.3260602999205345e-06, + "loss": 0.7759, + "step": 3229 + }, + { + "epoch": 0.2912927808089462, + "grad_norm": 1.5548132457391712, + "learning_rate": 3.32562290754577e-06, + "loss": 0.787, + "step": 3230 + }, + { + "epoch": 0.29138296433241645, + "grad_norm": 1.7533592271645397, + "learning_rate": 3.3251854020615494e-06, + "loss": 0.762, + "step": 3231 + }, + { + "epoch": 0.29147314785588674, + "grad_norm": 1.9085165213513657, + "learning_rate": 3.324747783505204e-06, + "loss": 0.7824, + "step": 3232 + }, + { + "epoch": 0.29156333137935697, + "grad_norm": 1.8033385416450431, + "learning_rate": 3.324310051914073e-06, + "loss": 0.6942, + "step": 3233 + }, + { + "epoch": 0.29165351490282726, + "grad_norm": 2.1922270789621368, + "learning_rate": 3.3238722073255056e-06, + "loss": 0.7466, + "step": 3234 + }, + { + "epoch": 0.2917436984262975, + "grad_norm": 1.9113440691674477, + "learning_rate": 3.323434249776863e-06, + "loss": 0.7374, + "step": 3235 + }, + { + "epoch": 0.2918338819497678, + "grad_norm": 0.7221660704262153, + "learning_rate": 3.3229961793055117e-06, + "loss": 0.5899, + "step": 3236 + }, + { + "epoch": 0.291924065473238, + "grad_norm": 1.5226100924421089, + "learning_rate": 3.3225579959488314e-06, + "loss": 0.7034, + "step": 3237 + }, + { + "epoch": 0.2920142489967083, + "grad_norm": 1.841370485692443, + "learning_rate": 3.322119699744211e-06, + "loss": 0.7966, + "step": 3238 + }, + { + "epoch": 0.29210443252017854, + "grad_norm": 1.5137618893346234, + "learning_rate": 3.3216812907290476e-06, + "loss": 0.7885, + "step": 3239 + }, + { + "epoch": 0.2921946160436488, + "grad_norm": 2.366459800741477, + "learning_rate": 3.3212427689407484e-06, + "loss": 0.7965, + "step": 3240 + }, + { + "epoch": 0.2922847995671191, + "grad_norm": 0.6517131978769983, + "learning_rate": 3.3208041344167317e-06, + "loss": 0.5679, + "step": 3241 + }, + { + "epoch": 0.29237498309058935, + "grad_norm": 1.651396856599791, + "learning_rate": 3.3203653871944224e-06, + "loss": 0.6845, + "step": 3242 + }, + { + "epoch": 0.29246516661405964, + "grad_norm": 1.5141955841751817, + "learning_rate": 3.3199265273112587e-06, + "loss": 0.7336, + "step": 3243 + }, + { + "epoch": 0.29255535013752987, + "grad_norm": 1.8320884451651325, + "learning_rate": 3.3194875548046852e-06, + "loss": 0.7149, + "step": 3244 + }, + { + "epoch": 0.29264553366100016, + "grad_norm": 1.7167662529932204, + "learning_rate": 3.319048469712158e-06, + "loss": 0.7632, + "step": 3245 + }, + { + "epoch": 0.2927357171844704, + "grad_norm": 2.177421129902762, + "learning_rate": 3.3186092720711423e-06, + "loss": 0.8719, + "step": 3246 + }, + { + "epoch": 0.2928259007079407, + "grad_norm": 1.8406799748602918, + "learning_rate": 3.3181699619191125e-06, + "loss": 0.7054, + "step": 3247 + }, + { + "epoch": 0.2929160842314109, + "grad_norm": 1.5890257134394654, + "learning_rate": 3.3177305392935536e-06, + "loss": 0.8144, + "step": 3248 + }, + { + "epoch": 0.2930062677548812, + "grad_norm": 0.731535148451496, + "learning_rate": 3.3172910042319595e-06, + "loss": 0.6558, + "step": 3249 + }, + { + "epoch": 0.29309645127835143, + "grad_norm": 1.4467559205193319, + "learning_rate": 3.316851356771833e-06, + "loss": 0.8063, + "step": 3250 + }, + { + "epoch": 0.2931866348018217, + "grad_norm": 1.4215929359436557, + "learning_rate": 3.3164115969506876e-06, + "loss": 0.8047, + "step": 3251 + }, + { + "epoch": 0.29327681832529195, + "grad_norm": 2.0342371177532663, + "learning_rate": 3.315971724806046e-06, + "loss": 0.863, + "step": 3252 + }, + { + "epoch": 0.29336700184876224, + "grad_norm": 2.0120773440258914, + "learning_rate": 3.315531740375441e-06, + "loss": 0.7185, + "step": 3253 + }, + { + "epoch": 0.2934571853722325, + "grad_norm": 1.4610618497588665, + "learning_rate": 3.315091643696414e-06, + "loss": 0.7437, + "step": 3254 + }, + { + "epoch": 0.29354736889570276, + "grad_norm": 19.059345121871562, + "learning_rate": 3.3146514348065164e-06, + "loss": 0.7039, + "step": 3255 + }, + { + "epoch": 0.293637552419173, + "grad_norm": 2.7811004941382635, + "learning_rate": 3.31421111374331e-06, + "loss": 0.7819, + "step": 3256 + }, + { + "epoch": 0.2937277359426433, + "grad_norm": 2.4957292317463766, + "learning_rate": 3.3137706805443647e-06, + "loss": 0.789, + "step": 3257 + }, + { + "epoch": 0.2938179194661135, + "grad_norm": 1.764470558511084, + "learning_rate": 3.313330135247261e-06, + "loss": 0.8197, + "step": 3258 + }, + { + "epoch": 0.2939081029895838, + "grad_norm": 1.3395857938009224, + "learning_rate": 3.312889477889588e-06, + "loss": 0.7471, + "step": 3259 + }, + { + "epoch": 0.29399828651305404, + "grad_norm": 1.7816592633076014, + "learning_rate": 3.3124487085089464e-06, + "loss": 0.7885, + "step": 3260 + }, + { + "epoch": 0.29408847003652433, + "grad_norm": 1.9170559888532712, + "learning_rate": 3.312007827142943e-06, + "loss": 0.8089, + "step": 3261 + }, + { + "epoch": 0.29417865355999456, + "grad_norm": 1.4374683210779928, + "learning_rate": 3.3115668338291983e-06, + "loss": 0.6954, + "step": 3262 + }, + { + "epoch": 0.29426883708346485, + "grad_norm": 1.6523157572798763, + "learning_rate": 3.3111257286053394e-06, + "loss": 0.8534, + "step": 3263 + }, + { + "epoch": 0.29435902060693514, + "grad_norm": 1.5756030184745686, + "learning_rate": 3.3106845115090043e-06, + "loss": 0.7752, + "step": 3264 + }, + { + "epoch": 0.2944492041304054, + "grad_norm": 1.6141019123948779, + "learning_rate": 3.310243182577839e-06, + "loss": 0.7307, + "step": 3265 + }, + { + "epoch": 0.29453938765387566, + "grad_norm": 1.5304954382383849, + "learning_rate": 3.3098017418495007e-06, + "loss": 0.667, + "step": 3266 + }, + { + "epoch": 0.2946295711773459, + "grad_norm": 2.0118653475484534, + "learning_rate": 3.309360189361656e-06, + "loss": 0.7429, + "step": 3267 + }, + { + "epoch": 0.2947197547008162, + "grad_norm": 1.8414750369159156, + "learning_rate": 3.3089185251519797e-06, + "loss": 0.8524, + "step": 3268 + }, + { + "epoch": 0.2948099382242864, + "grad_norm": 2.313959125319348, + "learning_rate": 3.3084767492581574e-06, + "loss": 0.7513, + "step": 3269 + }, + { + "epoch": 0.2949001217477567, + "grad_norm": 2.7048001253247054, + "learning_rate": 3.3080348617178846e-06, + "loss": 0.7957, + "step": 3270 + }, + { + "epoch": 0.29499030527122694, + "grad_norm": 1.5628271141963168, + "learning_rate": 3.307592862568865e-06, + "loss": 0.7477, + "step": 3271 + }, + { + "epoch": 0.2950804887946972, + "grad_norm": 1.570144416581519, + "learning_rate": 3.307150751848812e-06, + "loss": 0.8361, + "step": 3272 + }, + { + "epoch": 0.29517067231816746, + "grad_norm": 1.7691835310220982, + "learning_rate": 3.3067085295954497e-06, + "loss": 0.6805, + "step": 3273 + }, + { + "epoch": 0.29526085584163775, + "grad_norm": 1.75557467960586, + "learning_rate": 3.3062661958465098e-06, + "loss": 0.6322, + "step": 3274 + }, + { + "epoch": 0.295351039365108, + "grad_norm": 1.981217763740791, + "learning_rate": 3.305823750639736e-06, + "loss": 0.6642, + "step": 3275 + }, + { + "epoch": 0.29544122288857827, + "grad_norm": 1.8172299722052527, + "learning_rate": 3.3053811940128795e-06, + "loss": 0.7653, + "step": 3276 + }, + { + "epoch": 0.2955314064120485, + "grad_norm": 1.623728841830243, + "learning_rate": 3.3049385260037016e-06, + "loss": 0.7482, + "step": 3277 + }, + { + "epoch": 0.2956215899355188, + "grad_norm": 2.5394926205697033, + "learning_rate": 3.3044957466499736e-06, + "loss": 0.6602, + "step": 3278 + }, + { + "epoch": 0.295711773458989, + "grad_norm": 2.3340287080648348, + "learning_rate": 3.304052855989475e-06, + "loss": 0.6896, + "step": 3279 + }, + { + "epoch": 0.2958019569824593, + "grad_norm": 1.9450529607118041, + "learning_rate": 3.3036098540599966e-06, + "loss": 0.8309, + "step": 3280 + }, + { + "epoch": 0.29589214050592955, + "grad_norm": 2.735214976312518, + "learning_rate": 3.3031667408993373e-06, + "loss": 0.732, + "step": 3281 + }, + { + "epoch": 0.29598232402939983, + "grad_norm": 1.7916739116196936, + "learning_rate": 3.302723516545306e-06, + "loss": 0.765, + "step": 3282 + }, + { + "epoch": 0.29607250755287007, + "grad_norm": 2.526569607547159, + "learning_rate": 3.302280181035722e-06, + "loss": 0.795, + "step": 3283 + }, + { + "epoch": 0.29616269107634036, + "grad_norm": 0.6073947667392906, + "learning_rate": 3.3018367344084117e-06, + "loss": 0.6183, + "step": 3284 + }, + { + "epoch": 0.2962528745998106, + "grad_norm": 1.9939893764186065, + "learning_rate": 3.3013931767012125e-06, + "loss": 0.7584, + "step": 3285 + }, + { + "epoch": 0.2963430581232809, + "grad_norm": 1.680459488293864, + "learning_rate": 3.300949507951972e-06, + "loss": 0.7605, + "step": 3286 + }, + { + "epoch": 0.2964332416467511, + "grad_norm": 2.3486883398776426, + "learning_rate": 3.300505728198546e-06, + "loss": 0.8152, + "step": 3287 + }, + { + "epoch": 0.2965234251702214, + "grad_norm": 2.6220013884679636, + "learning_rate": 3.3000618374788e-06, + "loss": 0.7562, + "step": 3288 + }, + { + "epoch": 0.2966136086936917, + "grad_norm": 1.5123897512309106, + "learning_rate": 3.2996178358306104e-06, + "loss": 0.806, + "step": 3289 + }, + { + "epoch": 0.2967037922171619, + "grad_norm": 1.496061001013345, + "learning_rate": 3.2991737232918606e-06, + "loss": 0.7344, + "step": 3290 + }, + { + "epoch": 0.2967939757406322, + "grad_norm": 2.033624874643262, + "learning_rate": 3.298729499900445e-06, + "loss": 0.7235, + "step": 3291 + }, + { + "epoch": 0.29688415926410244, + "grad_norm": 3.9204381005064146, + "learning_rate": 3.2982851656942677e-06, + "loss": 0.7193, + "step": 3292 + }, + { + "epoch": 0.29697434278757273, + "grad_norm": 1.3780659392626802, + "learning_rate": 3.2978407207112416e-06, + "loss": 0.7926, + "step": 3293 + }, + { + "epoch": 0.29706452631104296, + "grad_norm": 2.1352019287315054, + "learning_rate": 3.2973961649892888e-06, + "loss": 0.7734, + "step": 3294 + }, + { + "epoch": 0.29715470983451325, + "grad_norm": 1.658890113461051, + "learning_rate": 3.296951498566341e-06, + "loss": 0.8335, + "step": 3295 + }, + { + "epoch": 0.2972448933579835, + "grad_norm": 2.361062059329265, + "learning_rate": 3.2965067214803404e-06, + "loss": 0.7668, + "step": 3296 + }, + { + "epoch": 0.2973350768814538, + "grad_norm": 1.2288096065557994, + "learning_rate": 3.2960618337692372e-06, + "loss": 0.8012, + "step": 3297 + }, + { + "epoch": 0.297425260404924, + "grad_norm": 1.601337100659549, + "learning_rate": 3.2956168354709927e-06, + "loss": 0.7294, + "step": 3298 + }, + { + "epoch": 0.2975154439283943, + "grad_norm": 3.0828280851160628, + "learning_rate": 3.2951717266235754e-06, + "loss": 0.8204, + "step": 3299 + }, + { + "epoch": 0.29760562745186453, + "grad_norm": 1.8492036755184718, + "learning_rate": 3.294726507264964e-06, + "loss": 0.6441, + "step": 3300 + }, + { + "epoch": 0.2976958109753348, + "grad_norm": 2.947826658207015, + "learning_rate": 3.2942811774331487e-06, + "loss": 0.7399, + "step": 3301 + }, + { + "epoch": 0.29778599449880505, + "grad_norm": 2.1953120655860334, + "learning_rate": 3.293835737166127e-06, + "loss": 0.7699, + "step": 3302 + }, + { + "epoch": 0.29787617802227534, + "grad_norm": 1.601192450222501, + "learning_rate": 3.293390186501906e-06, + "loss": 0.718, + "step": 3303 + }, + { + "epoch": 0.2979663615457456, + "grad_norm": 1.7529825271675727, + "learning_rate": 3.2929445254785024e-06, + "loss": 0.7796, + "step": 3304 + }, + { + "epoch": 0.29805654506921586, + "grad_norm": 1.9813982044709706, + "learning_rate": 3.2924987541339423e-06, + "loss": 0.7283, + "step": 3305 + }, + { + "epoch": 0.2981467285926861, + "grad_norm": 2.8143239358957963, + "learning_rate": 3.292052872506262e-06, + "loss": 0.7888, + "step": 3306 + }, + { + "epoch": 0.2982369121161564, + "grad_norm": 1.6400790078042613, + "learning_rate": 3.291606880633506e-06, + "loss": 0.8076, + "step": 3307 + }, + { + "epoch": 0.2983270956396266, + "grad_norm": 2.318480722261255, + "learning_rate": 3.2911607785537297e-06, + "loss": 0.8208, + "step": 3308 + }, + { + "epoch": 0.2984172791630969, + "grad_norm": 2.1341857960340302, + "learning_rate": 3.290714566304997e-06, + "loss": 0.8565, + "step": 3309 + }, + { + "epoch": 0.29850746268656714, + "grad_norm": 1.9736201886618479, + "learning_rate": 3.2902682439253794e-06, + "loss": 0.8317, + "step": 3310 + }, + { + "epoch": 0.2985976462100374, + "grad_norm": 1.985335830798776, + "learning_rate": 3.289821811452961e-06, + "loss": 0.7848, + "step": 3311 + }, + { + "epoch": 0.2986878297335077, + "grad_norm": 1.8335833451276937, + "learning_rate": 3.289375268925834e-06, + "loss": 0.8244, + "step": 3312 + }, + { + "epoch": 0.29877801325697795, + "grad_norm": 10.4328632220386, + "learning_rate": 3.288928616382099e-06, + "loss": 0.7557, + "step": 3313 + }, + { + "epoch": 0.29886819678044824, + "grad_norm": 1.4369942978046983, + "learning_rate": 3.288481853859868e-06, + "loss": 0.8035, + "step": 3314 + }, + { + "epoch": 0.29895838030391847, + "grad_norm": 1.4915981227250177, + "learning_rate": 3.2880349813972604e-06, + "loss": 0.7426, + "step": 3315 + }, + { + "epoch": 0.29904856382738876, + "grad_norm": 1.6510184092844091, + "learning_rate": 3.2875879990324052e-06, + "loss": 0.7679, + "step": 3316 + }, + { + "epoch": 0.299138747350859, + "grad_norm": 1.728227597824665, + "learning_rate": 3.287140906803443e-06, + "loss": 0.8216, + "step": 3317 + }, + { + "epoch": 0.2992289308743293, + "grad_norm": 2.0460107740638733, + "learning_rate": 3.2866937047485216e-06, + "loss": 0.6901, + "step": 3318 + }, + { + "epoch": 0.2993191143977995, + "grad_norm": 2.330324912931478, + "learning_rate": 3.2862463929057985e-06, + "loss": 0.829, + "step": 3319 + }, + { + "epoch": 0.2994092979212698, + "grad_norm": 1.9644410344432752, + "learning_rate": 3.285798971313441e-06, + "loss": 0.6189, + "step": 3320 + }, + { + "epoch": 0.29949948144474003, + "grad_norm": 1.4600702659472578, + "learning_rate": 3.2853514400096248e-06, + "loss": 0.7906, + "step": 3321 + }, + { + "epoch": 0.2995896649682103, + "grad_norm": 1.6542870747569807, + "learning_rate": 3.2849037990325367e-06, + "loss": 0.7975, + "step": 3322 + }, + { + "epoch": 0.29967984849168056, + "grad_norm": 1.6776298517035222, + "learning_rate": 3.2844560484203717e-06, + "loss": 0.726, + "step": 3323 + }, + { + "epoch": 0.29977003201515084, + "grad_norm": 1.4508586971747044, + "learning_rate": 3.2840081882113333e-06, + "loss": 0.8505, + "step": 3324 + }, + { + "epoch": 0.2998602155386211, + "grad_norm": 1.7173659560800258, + "learning_rate": 3.283560218443638e-06, + "loss": 0.7414, + "step": 3325 + }, + { + "epoch": 0.29995039906209137, + "grad_norm": 1.437479599517691, + "learning_rate": 3.2831121391555064e-06, + "loss": 0.7876, + "step": 3326 + }, + { + "epoch": 0.3000405825855616, + "grad_norm": 1.6592827866501865, + "learning_rate": 3.2826639503851724e-06, + "loss": 0.7203, + "step": 3327 + }, + { + "epoch": 0.3001307661090319, + "grad_norm": 1.6542150844275014, + "learning_rate": 3.282215652170877e-06, + "loss": 0.8127, + "step": 3328 + }, + { + "epoch": 0.3002209496325021, + "grad_norm": 6.904609588922422, + "learning_rate": 3.281767244550873e-06, + "loss": 0.8063, + "step": 3329 + }, + { + "epoch": 0.3003111331559724, + "grad_norm": 0.6038090014406198, + "learning_rate": 3.2813187275634193e-06, + "loss": 0.5532, + "step": 3330 + }, + { + "epoch": 0.30040131667944264, + "grad_norm": 1.5915933865093643, + "learning_rate": 3.280870101246787e-06, + "loss": 0.7964, + "step": 3331 + }, + { + "epoch": 0.30049150020291293, + "grad_norm": 2.7973239474428513, + "learning_rate": 3.280421365639255e-06, + "loss": 0.6849, + "step": 3332 + }, + { + "epoch": 0.30058168372638316, + "grad_norm": 1.6131451343815464, + "learning_rate": 3.279972520779112e-06, + "loss": 0.761, + "step": 3333 + }, + { + "epoch": 0.30067186724985345, + "grad_norm": 1.4476314602256368, + "learning_rate": 3.279523566704656e-06, + "loss": 0.8347, + "step": 3334 + }, + { + "epoch": 0.30076205077332374, + "grad_norm": 9.717888797002075, + "learning_rate": 3.2790745034541935e-06, + "loss": 0.8248, + "step": 3335 + }, + { + "epoch": 0.300852234296794, + "grad_norm": 2.344785436155655, + "learning_rate": 3.278625331066042e-06, + "loss": 0.8175, + "step": 3336 + }, + { + "epoch": 0.30094241782026426, + "grad_norm": 1.8541754461644944, + "learning_rate": 3.278176049578527e-06, + "loss": 0.8164, + "step": 3337 + }, + { + "epoch": 0.3010326013437345, + "grad_norm": 2.3016440776551614, + "learning_rate": 3.2777266590299835e-06, + "loss": 0.7297, + "step": 3338 + }, + { + "epoch": 0.3011227848672048, + "grad_norm": 2.1096120065419024, + "learning_rate": 3.2772771594587562e-06, + "loss": 0.7058, + "step": 3339 + }, + { + "epoch": 0.301212968390675, + "grad_norm": 1.665204081929812, + "learning_rate": 3.2768275509031988e-06, + "loss": 0.8737, + "step": 3340 + }, + { + "epoch": 0.3013031519141453, + "grad_norm": 1.6414683036567566, + "learning_rate": 3.276377833401675e-06, + "loss": 0.7479, + "step": 3341 + }, + { + "epoch": 0.30139333543761554, + "grad_norm": 1.6515936986598372, + "learning_rate": 3.2759280069925557e-06, + "loss": 0.8339, + "step": 3342 + }, + { + "epoch": 0.30148351896108583, + "grad_norm": 1.5998014118215533, + "learning_rate": 3.2754780717142233e-06, + "loss": 0.8092, + "step": 3343 + }, + { + "epoch": 0.30157370248455606, + "grad_norm": 2.7255916233201622, + "learning_rate": 3.27502802760507e-06, + "loss": 0.7449, + "step": 3344 + }, + { + "epoch": 0.30166388600802635, + "grad_norm": 1.8358666588934094, + "learning_rate": 3.2745778747034943e-06, + "loss": 0.6837, + "step": 3345 + }, + { + "epoch": 0.3017540695314966, + "grad_norm": 1.8138009368242929, + "learning_rate": 3.274127613047906e-06, + "loss": 0.7534, + "step": 3346 + }, + { + "epoch": 0.30184425305496687, + "grad_norm": 2.411404328259901, + "learning_rate": 3.273677242676725e-06, + "loss": 0.7899, + "step": 3347 + }, + { + "epoch": 0.3019344365784371, + "grad_norm": 2.7212653089310854, + "learning_rate": 3.2732267636283782e-06, + "loss": 0.7696, + "step": 3348 + }, + { + "epoch": 0.3020246201019074, + "grad_norm": 5.864722819953456, + "learning_rate": 3.2727761759413034e-06, + "loss": 0.7351, + "step": 3349 + }, + { + "epoch": 0.3021148036253776, + "grad_norm": 2.543451266356372, + "learning_rate": 3.2723254796539477e-06, + "loss": 0.7903, + "step": 3350 + }, + { + "epoch": 0.3022049871488479, + "grad_norm": 1.7923155947523937, + "learning_rate": 3.271874674804766e-06, + "loss": 0.7578, + "step": 3351 + }, + { + "epoch": 0.30229517067231815, + "grad_norm": 1.7514915240945985, + "learning_rate": 3.2714237614322242e-06, + "loss": 0.7979, + "step": 3352 + }, + { + "epoch": 0.30238535419578844, + "grad_norm": 1.8947909147720006, + "learning_rate": 3.2709727395747974e-06, + "loss": 0.6032, + "step": 3353 + }, + { + "epoch": 0.30247553771925867, + "grad_norm": 0.6758856720262612, + "learning_rate": 3.2705216092709673e-06, + "loss": 0.5725, + "step": 3354 + }, + { + "epoch": 0.30256572124272896, + "grad_norm": 2.1217171611705496, + "learning_rate": 3.2700703705592282e-06, + "loss": 0.8636, + "step": 3355 + }, + { + "epoch": 0.3026559047661992, + "grad_norm": 1.9431082251736647, + "learning_rate": 3.269619023478082e-06, + "loss": 0.6782, + "step": 3356 + }, + { + "epoch": 0.3027460882896695, + "grad_norm": 1.6072344572212003, + "learning_rate": 3.26916756806604e-06, + "loss": 0.8709, + "step": 3357 + }, + { + "epoch": 0.3028362718131397, + "grad_norm": 1.7918138295733919, + "learning_rate": 3.268716004361623e-06, + "loss": 0.7928, + "step": 3358 + }, + { + "epoch": 0.30292645533661, + "grad_norm": 0.6625695219089738, + "learning_rate": 3.268264332403361e-06, + "loss": 0.5879, + "step": 3359 + }, + { + "epoch": 0.3030166388600803, + "grad_norm": 1.4455774837589659, + "learning_rate": 3.2678125522297933e-06, + "loss": 0.8048, + "step": 3360 + }, + { + "epoch": 0.3031068223835505, + "grad_norm": 2.1321032116725984, + "learning_rate": 3.267360663879468e-06, + "loss": 0.7141, + "step": 3361 + }, + { + "epoch": 0.3031970059070208, + "grad_norm": 1.6305924796812632, + "learning_rate": 3.266908667390942e-06, + "loss": 0.7653, + "step": 3362 + }, + { + "epoch": 0.30328718943049104, + "grad_norm": 1.9499561696139551, + "learning_rate": 3.2664565628027833e-06, + "loss": 0.8258, + "step": 3363 + }, + { + "epoch": 0.30337737295396133, + "grad_norm": 3.1171653192252045, + "learning_rate": 3.2660043501535675e-06, + "loss": 0.7692, + "step": 3364 + }, + { + "epoch": 0.30346755647743157, + "grad_norm": 2.0914280100485936, + "learning_rate": 3.2655520294818797e-06, + "loss": 0.8454, + "step": 3365 + }, + { + "epoch": 0.30355774000090185, + "grad_norm": 4.785411743816965, + "learning_rate": 3.2650996008263146e-06, + "loss": 0.8472, + "step": 3366 + }, + { + "epoch": 0.3036479235243721, + "grad_norm": 1.8390568631846627, + "learning_rate": 3.2646470642254756e-06, + "loss": 0.7362, + "step": 3367 + }, + { + "epoch": 0.3037381070478424, + "grad_norm": 1.802601587234352, + "learning_rate": 3.2641944197179767e-06, + "loss": 0.797, + "step": 3368 + }, + { + "epoch": 0.3038282905713126, + "grad_norm": 0.6641577484037853, + "learning_rate": 3.2637416673424383e-06, + "loss": 0.5734, + "step": 3369 + }, + { + "epoch": 0.3039184740947829, + "grad_norm": 3.2541922893484165, + "learning_rate": 3.2632888071374937e-06, + "loss": 0.7844, + "step": 3370 + }, + { + "epoch": 0.30400865761825313, + "grad_norm": 2.2429730571092943, + "learning_rate": 3.2628358391417815e-06, + "loss": 0.8071, + "step": 3371 + }, + { + "epoch": 0.3040988411417234, + "grad_norm": 2.1035985903729224, + "learning_rate": 3.2623827633939526e-06, + "loss": 0.7646, + "step": 3372 + }, + { + "epoch": 0.30418902466519365, + "grad_norm": 2.4362225241601685, + "learning_rate": 3.2619295799326657e-06, + "loss": 0.7675, + "step": 3373 + }, + { + "epoch": 0.30427920818866394, + "grad_norm": 1.8333268454465799, + "learning_rate": 3.2614762887965883e-06, + "loss": 0.6712, + "step": 3374 + }, + { + "epoch": 0.3043693917121342, + "grad_norm": 1.672882418283916, + "learning_rate": 3.2610228900243984e-06, + "loss": 0.7642, + "step": 3375 + }, + { + "epoch": 0.30445957523560446, + "grad_norm": 0.6146936371520668, + "learning_rate": 3.260569383654783e-06, + "loss": 0.6241, + "step": 3376 + }, + { + "epoch": 0.3045497587590747, + "grad_norm": 0.6937626055697748, + "learning_rate": 3.2601157697264365e-06, + "loss": 0.5941, + "step": 3377 + }, + { + "epoch": 0.304639942282545, + "grad_norm": 1.9662728241252363, + "learning_rate": 3.2596620482780647e-06, + "loss": 0.7553, + "step": 3378 + }, + { + "epoch": 0.3047301258060152, + "grad_norm": 0.5727040821203052, + "learning_rate": 3.2592082193483803e-06, + "loss": 0.5936, + "step": 3379 + }, + { + "epoch": 0.3048203093294855, + "grad_norm": 1.5097630983161263, + "learning_rate": 3.258754282976109e-06, + "loss": 0.7589, + "step": 3380 + }, + { + "epoch": 0.30491049285295574, + "grad_norm": 1.8603400282614877, + "learning_rate": 3.25830023919998e-06, + "loss": 0.847, + "step": 3381 + }, + { + "epoch": 0.305000676376426, + "grad_norm": 2.2985675829415175, + "learning_rate": 3.2578460880587374e-06, + "loss": 0.7255, + "step": 3382 + }, + { + "epoch": 0.3050908598998963, + "grad_norm": 1.7473388601946078, + "learning_rate": 3.2573918295911306e-06, + "loss": 0.8314, + "step": 3383 + }, + { + "epoch": 0.30518104342336655, + "grad_norm": 1.6039137537688268, + "learning_rate": 3.2569374638359196e-06, + "loss": 0.7068, + "step": 3384 + }, + { + "epoch": 0.30527122694683684, + "grad_norm": 1.3784568421695989, + "learning_rate": 3.2564829908318736e-06, + "loss": 0.7854, + "step": 3385 + }, + { + "epoch": 0.30536141047030707, + "grad_norm": 1.870316887693429, + "learning_rate": 3.2560284106177705e-06, + "loss": 0.832, + "step": 3386 + }, + { + "epoch": 0.30545159399377736, + "grad_norm": 4.058424093392184, + "learning_rate": 3.2555737232323978e-06, + "loss": 0.7428, + "step": 3387 + }, + { + "epoch": 0.3055417775172476, + "grad_norm": 1.422488824166911, + "learning_rate": 3.255118928714552e-06, + "loss": 0.8063, + "step": 3388 + }, + { + "epoch": 0.3056319610407179, + "grad_norm": 2.144046360361557, + "learning_rate": 3.2546640271030386e-06, + "loss": 0.7016, + "step": 3389 + }, + { + "epoch": 0.3057221445641881, + "grad_norm": 0.6365461993295698, + "learning_rate": 3.2542090184366717e-06, + "loss": 0.6249, + "step": 3390 + }, + { + "epoch": 0.3058123280876584, + "grad_norm": 2.5064249448529092, + "learning_rate": 3.253753902754276e-06, + "loss": 0.8569, + "step": 3391 + }, + { + "epoch": 0.30590251161112864, + "grad_norm": 1.5710715362025514, + "learning_rate": 3.253298680094685e-06, + "loss": 0.7056, + "step": 3392 + }, + { + "epoch": 0.3059926951345989, + "grad_norm": 1.8348998038770248, + "learning_rate": 3.2528433504967394e-06, + "loss": 0.7191, + "step": 3393 + }, + { + "epoch": 0.30608287865806916, + "grad_norm": 1.6761504080114946, + "learning_rate": 3.252387913999291e-06, + "loss": 0.7848, + "step": 3394 + }, + { + "epoch": 0.30617306218153945, + "grad_norm": 1.6291069071834907, + "learning_rate": 3.2519323706411998e-06, + "loss": 0.7776, + "step": 3395 + }, + { + "epoch": 0.3062632457050097, + "grad_norm": 1.396699176620882, + "learning_rate": 3.251476720461336e-06, + "loss": 0.6632, + "step": 3396 + }, + { + "epoch": 0.30635342922847997, + "grad_norm": 2.3491411384674064, + "learning_rate": 3.251020963498578e-06, + "loss": 0.7057, + "step": 3397 + }, + { + "epoch": 0.3064436127519502, + "grad_norm": 1.5015479683545017, + "learning_rate": 3.250565099791813e-06, + "loss": 0.8145, + "step": 3398 + }, + { + "epoch": 0.3065337962754205, + "grad_norm": 1.4302568813159102, + "learning_rate": 3.2501091293799387e-06, + "loss": 0.8273, + "step": 3399 + }, + { + "epoch": 0.3066239797988907, + "grad_norm": 2.125788934897734, + "learning_rate": 3.24965305230186e-06, + "loss": 0.7875, + "step": 3400 + }, + { + "epoch": 0.306714163322361, + "grad_norm": 2.0496924425482246, + "learning_rate": 3.249196868596492e-06, + "loss": 0.8422, + "step": 3401 + }, + { + "epoch": 0.30680434684583124, + "grad_norm": 2.136711162435111, + "learning_rate": 3.24874057830276e-06, + "loss": 0.7036, + "step": 3402 + }, + { + "epoch": 0.30689453036930153, + "grad_norm": 1.6701083252089155, + "learning_rate": 3.2482841814595954e-06, + "loss": 0.6957, + "step": 3403 + }, + { + "epoch": 0.30698471389277177, + "grad_norm": 1.870451305213413, + "learning_rate": 3.247827678105943e-06, + "loss": 0.7162, + "step": 3404 + }, + { + "epoch": 0.30707489741624205, + "grad_norm": 1.9534418688273514, + "learning_rate": 3.247371068280751e-06, + "loss": 0.769, + "step": 3405 + }, + { + "epoch": 0.3071650809397123, + "grad_norm": 1.3691606506537441, + "learning_rate": 3.2469143520229823e-06, + "loss": 0.6349, + "step": 3406 + }, + { + "epoch": 0.3072552644631826, + "grad_norm": 1.4264649306625221, + "learning_rate": 3.2464575293716054e-06, + "loss": 0.7229, + "step": 3407 + }, + { + "epoch": 0.30734544798665286, + "grad_norm": 2.289967328800539, + "learning_rate": 3.2460006003655997e-06, + "loss": 0.7125, + "step": 3408 + }, + { + "epoch": 0.3074356315101231, + "grad_norm": 2.8594061844589636, + "learning_rate": 3.245543565043952e-06, + "loss": 0.8293, + "step": 3409 + }, + { + "epoch": 0.3075258150335934, + "grad_norm": 2.138756709397637, + "learning_rate": 3.2450864234456592e-06, + "loss": 0.7688, + "step": 3410 + }, + { + "epoch": 0.3076159985570636, + "grad_norm": 1.930644859639106, + "learning_rate": 3.244629175609728e-06, + "loss": 0.7647, + "step": 3411 + }, + { + "epoch": 0.3077061820805339, + "grad_norm": 1.5021917701693828, + "learning_rate": 3.2441718215751726e-06, + "loss": 0.7589, + "step": 3412 + }, + { + "epoch": 0.30779636560400414, + "grad_norm": 1.7953493358528285, + "learning_rate": 3.2437143613810173e-06, + "loss": 0.736, + "step": 3413 + }, + { + "epoch": 0.30788654912747443, + "grad_norm": 1.6104482386933452, + "learning_rate": 3.2432567950662947e-06, + "loss": 0.8156, + "step": 3414 + }, + { + "epoch": 0.30797673265094466, + "grad_norm": 1.5047509929693634, + "learning_rate": 3.2427991226700468e-06, + "loss": 0.7568, + "step": 3415 + }, + { + "epoch": 0.30806691617441495, + "grad_norm": 1.7592593193983697, + "learning_rate": 3.2423413442313246e-06, + "loss": 0.6956, + "step": 3416 + }, + { + "epoch": 0.3081570996978852, + "grad_norm": 1.8002561519408031, + "learning_rate": 3.2418834597891904e-06, + "loss": 0.692, + "step": 3417 + }, + { + "epoch": 0.3082472832213555, + "grad_norm": 2.050456749810397, + "learning_rate": 3.2414254693827098e-06, + "loss": 0.8551, + "step": 3418 + }, + { + "epoch": 0.3083374667448257, + "grad_norm": 1.5928422175092425, + "learning_rate": 3.2409673730509644e-06, + "loss": 0.7521, + "step": 3419 + }, + { + "epoch": 0.308427650268296, + "grad_norm": 2.3221440648377802, + "learning_rate": 3.2405091708330393e-06, + "loss": 0.7414, + "step": 3420 + }, + { + "epoch": 0.3085178337917662, + "grad_norm": 6.670952404833123, + "learning_rate": 3.2400508627680323e-06, + "loss": 0.7632, + "step": 3421 + }, + { + "epoch": 0.3086080173152365, + "grad_norm": 10.206820462806277, + "learning_rate": 3.2395924488950474e-06, + "loss": 0.8599, + "step": 3422 + }, + { + "epoch": 0.30869820083870675, + "grad_norm": 1.825115563378629, + "learning_rate": 3.2391339292532004e-06, + "loss": 0.8963, + "step": 3423 + }, + { + "epoch": 0.30878838436217704, + "grad_norm": 2.7574682263975925, + "learning_rate": 3.238675303881614e-06, + "loss": 0.7229, + "step": 3424 + }, + { + "epoch": 0.30887856788564727, + "grad_norm": 1.7773152443412914, + "learning_rate": 3.2382165728194203e-06, + "loss": 0.7411, + "step": 3425 + }, + { + "epoch": 0.30896875140911756, + "grad_norm": 1.5432165019399389, + "learning_rate": 3.237757736105761e-06, + "loss": 0.752, + "step": 3426 + }, + { + "epoch": 0.3090589349325878, + "grad_norm": 1.6794845813339874, + "learning_rate": 3.2372987937797867e-06, + "loss": 0.8488, + "step": 3427 + }, + { + "epoch": 0.3091491184560581, + "grad_norm": 1.3873810811674667, + "learning_rate": 3.2368397458806573e-06, + "loss": 0.6719, + "step": 3428 + }, + { + "epoch": 0.3092393019795283, + "grad_norm": 1.8400482906348643, + "learning_rate": 3.2363805924475412e-06, + "loss": 0.8023, + "step": 3429 + }, + { + "epoch": 0.3093294855029986, + "grad_norm": 1.8256461789301042, + "learning_rate": 3.2359213335196153e-06, + "loss": 0.7132, + "step": 3430 + }, + { + "epoch": 0.3094196690264689, + "grad_norm": 2.679807877130616, + "learning_rate": 3.2354619691360663e-06, + "loss": 0.7216, + "step": 3431 + }, + { + "epoch": 0.3095098525499391, + "grad_norm": 1.4013392854365123, + "learning_rate": 3.2350024993360898e-06, + "loss": 0.8091, + "step": 3432 + }, + { + "epoch": 0.3096000360734094, + "grad_norm": 1.733058945627931, + "learning_rate": 3.2345429241588902e-06, + "loss": 0.765, + "step": 3433 + }, + { + "epoch": 0.30969021959687965, + "grad_norm": 2.4887664178204236, + "learning_rate": 3.234083243643681e-06, + "loss": 0.7944, + "step": 3434 + }, + { + "epoch": 0.30978040312034993, + "grad_norm": 1.469140751581813, + "learning_rate": 3.233623457829686e-06, + "loss": 0.837, + "step": 3435 + }, + { + "epoch": 0.30987058664382017, + "grad_norm": 2.5556304777837213, + "learning_rate": 3.2331635667561344e-06, + "loss": 0.7428, + "step": 3436 + }, + { + "epoch": 0.30996077016729046, + "grad_norm": 1.764563114571362, + "learning_rate": 3.2327035704622674e-06, + "loss": 0.7977, + "step": 3437 + }, + { + "epoch": 0.3100509536907607, + "grad_norm": 1.3921331307255191, + "learning_rate": 3.2322434689873353e-06, + "loss": 0.6616, + "step": 3438 + }, + { + "epoch": 0.310141137214231, + "grad_norm": 1.7170964696773996, + "learning_rate": 3.2317832623705957e-06, + "loss": 0.8482, + "step": 3439 + }, + { + "epoch": 0.3102313207377012, + "grad_norm": 2.0666725261153966, + "learning_rate": 3.231322950651316e-06, + "loss": 0.7139, + "step": 3440 + }, + { + "epoch": 0.3103215042611715, + "grad_norm": 1.8971998788061872, + "learning_rate": 3.2308625338687735e-06, + "loss": 0.7932, + "step": 3441 + }, + { + "epoch": 0.31041168778464173, + "grad_norm": 2.9217075565114943, + "learning_rate": 3.230402012062252e-06, + "loss": 0.8813, + "step": 3442 + }, + { + "epoch": 0.310501871308112, + "grad_norm": 0.6171380276710208, + "learning_rate": 3.2299413852710466e-06, + "loss": 0.6138, + "step": 3443 + }, + { + "epoch": 0.31059205483158225, + "grad_norm": 1.691984857347153, + "learning_rate": 3.2294806535344606e-06, + "loss": 0.7314, + "step": 3444 + }, + { + "epoch": 0.31068223835505254, + "grad_norm": 2.035432353849164, + "learning_rate": 3.2290198168918056e-06, + "loss": 0.8099, + "step": 3445 + }, + { + "epoch": 0.3107724218785228, + "grad_norm": 4.289361106052988, + "learning_rate": 3.2285588753824035e-06, + "loss": 0.8062, + "step": 3446 + }, + { + "epoch": 0.31086260540199306, + "grad_norm": 1.7968226549568238, + "learning_rate": 3.228097829045584e-06, + "loss": 0.7804, + "step": 3447 + }, + { + "epoch": 0.3109527889254633, + "grad_norm": 1.3481724289574344, + "learning_rate": 3.227636677920685e-06, + "loss": 0.8102, + "step": 3448 + }, + { + "epoch": 0.3110429724489336, + "grad_norm": 1.8054339693978938, + "learning_rate": 3.2271754220470567e-06, + "loss": 0.8158, + "step": 3449 + }, + { + "epoch": 0.3111331559724038, + "grad_norm": 1.9218911891348442, + "learning_rate": 3.2267140614640547e-06, + "loss": 0.8512, + "step": 3450 + }, + { + "epoch": 0.3112233394958741, + "grad_norm": 1.7033555460985121, + "learning_rate": 3.2262525962110445e-06, + "loss": 0.8162, + "step": 3451 + }, + { + "epoch": 0.31131352301934434, + "grad_norm": 1.9159423316368265, + "learning_rate": 3.2257910263274015e-06, + "loss": 0.6971, + "step": 3452 + }, + { + "epoch": 0.31140370654281463, + "grad_norm": 2.7627415685196235, + "learning_rate": 3.225329351852509e-06, + "loss": 0.7166, + "step": 3453 + }, + { + "epoch": 0.3114938900662849, + "grad_norm": 1.9342223099453075, + "learning_rate": 3.2248675728257596e-06, + "loss": 0.772, + "step": 3454 + }, + { + "epoch": 0.31158407358975515, + "grad_norm": 1.5870232835009912, + "learning_rate": 3.2244056892865557e-06, + "loss": 0.8428, + "step": 3455 + }, + { + "epoch": 0.31167425711322544, + "grad_norm": 1.60222437138148, + "learning_rate": 3.2239437012743063e-06, + "loss": 0.7603, + "step": 3456 + }, + { + "epoch": 0.31176444063669567, + "grad_norm": 1.8592816497706388, + "learning_rate": 3.223481608828432e-06, + "loss": 0.7845, + "step": 3457 + }, + { + "epoch": 0.31185462416016596, + "grad_norm": 1.978210666723923, + "learning_rate": 3.223019411988361e-06, + "loss": 0.7896, + "step": 3458 + }, + { + "epoch": 0.3119448076836362, + "grad_norm": 1.6697271936272422, + "learning_rate": 3.22255711079353e-06, + "loss": 0.7687, + "step": 3459 + }, + { + "epoch": 0.3120349912071065, + "grad_norm": 1.5258077327150805, + "learning_rate": 3.222094705283385e-06, + "loss": 0.7291, + "step": 3460 + }, + { + "epoch": 0.3121251747305767, + "grad_norm": 2.1644978550129053, + "learning_rate": 3.2216321954973805e-06, + "loss": 0.7662, + "step": 3461 + }, + { + "epoch": 0.312215358254047, + "grad_norm": 2.269068028882353, + "learning_rate": 3.2211695814749816e-06, + "loss": 0.764, + "step": 3462 + }, + { + "epoch": 0.31230554177751724, + "grad_norm": 2.4907185880339497, + "learning_rate": 3.220706863255661e-06, + "loss": 0.88, + "step": 3463 + }, + { + "epoch": 0.3123957253009875, + "grad_norm": 2.453081847376226, + "learning_rate": 3.2202440408788994e-06, + "loss": 0.7684, + "step": 3464 + }, + { + "epoch": 0.31248590882445776, + "grad_norm": 1.7838083600105505, + "learning_rate": 3.2197811143841883e-06, + "loss": 0.8404, + "step": 3465 + }, + { + "epoch": 0.31257609234792805, + "grad_norm": 1.6986120785266572, + "learning_rate": 3.2193180838110267e-06, + "loss": 0.764, + "step": 3466 + }, + { + "epoch": 0.3126662758713983, + "grad_norm": 2.821373483696048, + "learning_rate": 3.2188549491989225e-06, + "loss": 0.6647, + "step": 3467 + }, + { + "epoch": 0.31275645939486857, + "grad_norm": 1.7277068061535137, + "learning_rate": 3.2183917105873934e-06, + "loss": 0.7933, + "step": 3468 + }, + { + "epoch": 0.3128466429183388, + "grad_norm": 1.7784104049460603, + "learning_rate": 3.217928368015966e-06, + "loss": 0.8274, + "step": 3469 + }, + { + "epoch": 0.3129368264418091, + "grad_norm": 1.4614916887598393, + "learning_rate": 3.217464921524174e-06, + "loss": 0.7984, + "step": 3470 + }, + { + "epoch": 0.3130270099652793, + "grad_norm": 1.6188385132851693, + "learning_rate": 3.2170013711515616e-06, + "loss": 0.7834, + "step": 3471 + }, + { + "epoch": 0.3131171934887496, + "grad_norm": 2.1575894825436803, + "learning_rate": 3.216537716937682e-06, + "loss": 0.6943, + "step": 3472 + }, + { + "epoch": 0.31320737701221985, + "grad_norm": 1.799659230506505, + "learning_rate": 3.2160739589220968e-06, + "loss": 0.7448, + "step": 3473 + }, + { + "epoch": 0.31329756053569013, + "grad_norm": 1.5548057319502089, + "learning_rate": 3.215610097144376e-06, + "loss": 0.825, + "step": 3474 + }, + { + "epoch": 0.31338774405916037, + "grad_norm": 1.9883093691413218, + "learning_rate": 3.215146131644099e-06, + "loss": 0.7161, + "step": 3475 + }, + { + "epoch": 0.31347792758263066, + "grad_norm": 1.7192286431807025, + "learning_rate": 3.214682062460854e-06, + "loss": 0.8443, + "step": 3476 + }, + { + "epoch": 0.3135681111061009, + "grad_norm": 2.0598873583542536, + "learning_rate": 3.2142178896342367e-06, + "loss": 0.689, + "step": 3477 + }, + { + "epoch": 0.3136582946295712, + "grad_norm": 1.6144760916585632, + "learning_rate": 3.2137536132038552e-06, + "loss": 0.7567, + "step": 3478 + }, + { + "epoch": 0.31374847815304147, + "grad_norm": 1.8821151518340167, + "learning_rate": 3.2132892332093226e-06, + "loss": 0.7725, + "step": 3479 + }, + { + "epoch": 0.3138386616765117, + "grad_norm": 1.7334733285978423, + "learning_rate": 3.2128247496902623e-06, + "loss": 0.8848, + "step": 3480 + }, + { + "epoch": 0.313928845199982, + "grad_norm": 1.8704144672333562, + "learning_rate": 3.2123601626863064e-06, + "loss": 0.7849, + "step": 3481 + }, + { + "epoch": 0.3140190287234522, + "grad_norm": 0.6416929925951208, + "learning_rate": 3.2118954722370974e-06, + "loss": 0.5713, + "step": 3482 + }, + { + "epoch": 0.3141092122469225, + "grad_norm": 1.9384996850006344, + "learning_rate": 3.2114306783822837e-06, + "loss": 0.7219, + "step": 3483 + }, + { + "epoch": 0.31419939577039274, + "grad_norm": 1.850467470750509, + "learning_rate": 3.210965781161525e-06, + "loss": 0.7444, + "step": 3484 + }, + { + "epoch": 0.31428957929386303, + "grad_norm": 1.9180598033416336, + "learning_rate": 3.2105007806144892e-06, + "loss": 0.8831, + "step": 3485 + }, + { + "epoch": 0.31437976281733326, + "grad_norm": 1.855578417297714, + "learning_rate": 3.2100356767808513e-06, + "loss": 0.7685, + "step": 3486 + }, + { + "epoch": 0.31446994634080355, + "grad_norm": 1.9395656649081516, + "learning_rate": 3.2095704697002977e-06, + "loss": 0.8322, + "step": 3487 + }, + { + "epoch": 0.3145601298642738, + "grad_norm": 1.381545181148605, + "learning_rate": 3.209105159412522e-06, + "loss": 0.7697, + "step": 3488 + }, + { + "epoch": 0.3146503133877441, + "grad_norm": 3.0993296852191548, + "learning_rate": 3.208639745957228e-06, + "loss": 0.7416, + "step": 3489 + }, + { + "epoch": 0.3147404969112143, + "grad_norm": 1.460608868687373, + "learning_rate": 3.2081742293741256e-06, + "loss": 0.7981, + "step": 3490 + }, + { + "epoch": 0.3148306804346846, + "grad_norm": 1.4836180513710817, + "learning_rate": 3.2077086097029366e-06, + "loss": 0.833, + "step": 3491 + }, + { + "epoch": 0.31492086395815483, + "grad_norm": 1.955778958583732, + "learning_rate": 3.2072428869833895e-06, + "loss": 0.7722, + "step": 3492 + }, + { + "epoch": 0.3150110474816251, + "grad_norm": 4.056885345967661, + "learning_rate": 3.206777061255223e-06, + "loss": 0.7869, + "step": 3493 + }, + { + "epoch": 0.31510123100509535, + "grad_norm": 1.865593073841704, + "learning_rate": 3.206311132558183e-06, + "loss": 0.8085, + "step": 3494 + }, + { + "epoch": 0.31519141452856564, + "grad_norm": 2.0080554384732343, + "learning_rate": 3.205845100932026e-06, + "loss": 0.7858, + "step": 3495 + }, + { + "epoch": 0.31528159805203587, + "grad_norm": 1.6193466147522948, + "learning_rate": 3.205378966416516e-06, + "loss": 0.8017, + "step": 3496 + }, + { + "epoch": 0.31537178157550616, + "grad_norm": 3.0896522279355976, + "learning_rate": 3.204912729051426e-06, + "loss": 0.7692, + "step": 3497 + }, + { + "epoch": 0.3154619650989764, + "grad_norm": 2.0692670377716835, + "learning_rate": 3.2044463888765384e-06, + "loss": 0.7449, + "step": 3498 + }, + { + "epoch": 0.3155521486224467, + "grad_norm": 1.8894874997075009, + "learning_rate": 3.2039799459316436e-06, + "loss": 0.6765, + "step": 3499 + }, + { + "epoch": 0.3156423321459169, + "grad_norm": 1.9733411147741737, + "learning_rate": 3.2035134002565407e-06, + "loss": 0.8054, + "step": 3500 + }, + { + "epoch": 0.3157325156693872, + "grad_norm": 1.966530108425279, + "learning_rate": 3.203046751891039e-06, + "loss": 0.7857, + "step": 3501 + }, + { + "epoch": 0.3158226991928575, + "grad_norm": 1.8010643593759719, + "learning_rate": 3.2025800008749545e-06, + "loss": 0.7866, + "step": 3502 + }, + { + "epoch": 0.3159128827163277, + "grad_norm": 2.139044742217938, + "learning_rate": 3.202113147248114e-06, + "loss": 0.7512, + "step": 3503 + }, + { + "epoch": 0.316003066239798, + "grad_norm": 2.1241199690708084, + "learning_rate": 3.20164619105035e-06, + "loss": 0.745, + "step": 3504 + }, + { + "epoch": 0.31609324976326825, + "grad_norm": 1.5197093241475894, + "learning_rate": 3.201179132321508e-06, + "loss": 0.6263, + "step": 3505 + }, + { + "epoch": 0.31618343328673854, + "grad_norm": 1.645266189720445, + "learning_rate": 3.200711971101439e-06, + "loss": 0.7157, + "step": 3506 + }, + { + "epoch": 0.31627361681020877, + "grad_norm": 2.490850682007284, + "learning_rate": 3.2002447074300047e-06, + "loss": 0.7736, + "step": 3507 + }, + { + "epoch": 0.31636380033367906, + "grad_norm": 1.7756122580866311, + "learning_rate": 3.1997773413470736e-06, + "loss": 0.7771, + "step": 3508 + }, + { + "epoch": 0.3164539838571493, + "grad_norm": 5.399448945927514, + "learning_rate": 3.199309872892524e-06, + "loss": 0.7663, + "step": 3509 + }, + { + "epoch": 0.3165441673806196, + "grad_norm": 1.9693833800645726, + "learning_rate": 3.198842302106243e-06, + "loss": 0.8235, + "step": 3510 + }, + { + "epoch": 0.3166343509040898, + "grad_norm": 2.0920545631207763, + "learning_rate": 3.1983746290281265e-06, + "loss": 0.7525, + "step": 3511 + }, + { + "epoch": 0.3167245344275601, + "grad_norm": 1.3469437194428568, + "learning_rate": 3.197906853698079e-06, + "loss": 0.782, + "step": 3512 + }, + { + "epoch": 0.31681471795103033, + "grad_norm": 1.5707214733874524, + "learning_rate": 3.1974389761560137e-06, + "loss": 0.7234, + "step": 3513 + }, + { + "epoch": 0.3169049014745006, + "grad_norm": 1.9180348184863973, + "learning_rate": 3.1969709964418525e-06, + "loss": 0.7253, + "step": 3514 + }, + { + "epoch": 0.31699508499797086, + "grad_norm": 2.7478523538007957, + "learning_rate": 3.196502914595525e-06, + "loss": 0.7932, + "step": 3515 + }, + { + "epoch": 0.31708526852144114, + "grad_norm": 1.6617162219316615, + "learning_rate": 3.1960347306569723e-06, + "loss": 0.7922, + "step": 3516 + }, + { + "epoch": 0.3171754520449114, + "grad_norm": 2.287759038714517, + "learning_rate": 3.195566444666141e-06, + "loss": 0.7232, + "step": 3517 + }, + { + "epoch": 0.31726563556838167, + "grad_norm": 1.708946916442159, + "learning_rate": 3.1950980566629886e-06, + "loss": 0.8806, + "step": 3518 + }, + { + "epoch": 0.3173558190918519, + "grad_norm": 1.5621566395192394, + "learning_rate": 3.1946295666874797e-06, + "loss": 0.7888, + "step": 3519 + }, + { + "epoch": 0.3174460026153222, + "grad_norm": 1.5557412261887127, + "learning_rate": 3.19416097477959e-06, + "loss": 0.7945, + "step": 3520 + }, + { + "epoch": 0.3175361861387924, + "grad_norm": 4.4100027133647846, + "learning_rate": 3.1936922809793005e-06, + "loss": 0.8314, + "step": 3521 + }, + { + "epoch": 0.3176263696622627, + "grad_norm": 4.219871993451138, + "learning_rate": 3.193223485326604e-06, + "loss": 0.8916, + "step": 3522 + }, + { + "epoch": 0.31771655318573294, + "grad_norm": 2.8310410445622463, + "learning_rate": 3.1927545878615005e-06, + "loss": 0.8564, + "step": 3523 + }, + { + "epoch": 0.31780673670920323, + "grad_norm": 2.372980363329738, + "learning_rate": 3.192285588623999e-06, + "loss": 0.7298, + "step": 3524 + }, + { + "epoch": 0.31789692023267346, + "grad_norm": 1.3929702185400017, + "learning_rate": 3.191816487654117e-06, + "loss": 0.7068, + "step": 3525 + }, + { + "epoch": 0.31798710375614375, + "grad_norm": 3.3623110693284626, + "learning_rate": 3.19134728499188e-06, + "loss": 0.745, + "step": 3526 + }, + { + "epoch": 0.31807728727961404, + "grad_norm": 5.616578134052181, + "learning_rate": 3.1908779806773235e-06, + "loss": 0.6633, + "step": 3527 + }, + { + "epoch": 0.3181674708030843, + "grad_norm": 1.5724437162578984, + "learning_rate": 3.190408574750492e-06, + "loss": 0.8289, + "step": 3528 + }, + { + "epoch": 0.31825765432655456, + "grad_norm": 1.4637017032464046, + "learning_rate": 3.1899390672514367e-06, + "loss": 0.8426, + "step": 3529 + }, + { + "epoch": 0.3183478378500248, + "grad_norm": 2.0135965712459383, + "learning_rate": 3.189469458220219e-06, + "loss": 0.6951, + "step": 3530 + }, + { + "epoch": 0.3184380213734951, + "grad_norm": 1.5722794241921623, + "learning_rate": 3.1889997476969086e-06, + "loss": 0.7721, + "step": 3531 + }, + { + "epoch": 0.3185282048969653, + "grad_norm": 4.2932730819391764, + "learning_rate": 3.188529935721583e-06, + "loss": 0.7648, + "step": 3532 + }, + { + "epoch": 0.3186183884204356, + "grad_norm": 1.753137568736453, + "learning_rate": 3.18806002233433e-06, + "loss": 0.8344, + "step": 3533 + }, + { + "epoch": 0.31870857194390584, + "grad_norm": 1.6587718981150472, + "learning_rate": 3.187590007575245e-06, + "loss": 0.8373, + "step": 3534 + }, + { + "epoch": 0.3187987554673761, + "grad_norm": 1.866666777077172, + "learning_rate": 3.1871198914844327e-06, + "loss": 0.7125, + "step": 3535 + }, + { + "epoch": 0.31888893899084636, + "grad_norm": 1.5963193340971453, + "learning_rate": 3.1866496741020057e-06, + "loss": 0.7144, + "step": 3536 + }, + { + "epoch": 0.31897912251431665, + "grad_norm": 1.2858128595030425, + "learning_rate": 3.186179355468085e-06, + "loss": 0.7387, + "step": 3537 + }, + { + "epoch": 0.3190693060377869, + "grad_norm": 1.8713531632867069, + "learning_rate": 3.1857089356228015e-06, + "loss": 0.7684, + "step": 3538 + }, + { + "epoch": 0.31915948956125717, + "grad_norm": 1.662978558033374, + "learning_rate": 3.1852384146062933e-06, + "loss": 0.723, + "step": 3539 + }, + { + "epoch": 0.3192496730847274, + "grad_norm": 0.6124292517693577, + "learning_rate": 3.184767792458708e-06, + "loss": 0.5574, + "step": 3540 + }, + { + "epoch": 0.3193398566081977, + "grad_norm": 11.843531624182367, + "learning_rate": 3.1842970692202023e-06, + "loss": 0.9008, + "step": 3541 + }, + { + "epoch": 0.3194300401316679, + "grad_norm": 2.311765915111379, + "learning_rate": 3.1838262449309403e-06, + "loss": 0.7529, + "step": 3542 + }, + { + "epoch": 0.3195202236551382, + "grad_norm": 1.7865386885503645, + "learning_rate": 3.1833553196310956e-06, + "loss": 0.6889, + "step": 3543 + }, + { + "epoch": 0.31961040717860845, + "grad_norm": 1.635795606054059, + "learning_rate": 3.18288429336085e-06, + "loss": 0.6294, + "step": 3544 + }, + { + "epoch": 0.31970059070207874, + "grad_norm": 1.7949286492918275, + "learning_rate": 3.182413166160394e-06, + "loss": 0.8787, + "step": 3545 + }, + { + "epoch": 0.31979077422554897, + "grad_norm": 1.5736743373725646, + "learning_rate": 3.1819419380699275e-06, + "loss": 0.8164, + "step": 3546 + }, + { + "epoch": 0.31988095774901926, + "grad_norm": 1.6811424164225481, + "learning_rate": 3.181470609129658e-06, + "loss": 0.7782, + "step": 3547 + }, + { + "epoch": 0.3199711412724895, + "grad_norm": 2.561042091789346, + "learning_rate": 3.1809991793798e-06, + "loss": 0.8257, + "step": 3548 + }, + { + "epoch": 0.3200613247959598, + "grad_norm": 1.9911097584565616, + "learning_rate": 3.1805276488605806e-06, + "loss": 0.7332, + "step": 3549 + }, + { + "epoch": 0.32015150831943007, + "grad_norm": 1.8599705744025716, + "learning_rate": 3.1800560176122336e-06, + "loss": 0.7651, + "step": 3550 + }, + { + "epoch": 0.3202416918429003, + "grad_norm": 1.637520677858209, + "learning_rate": 3.179584285675e-06, + "loss": 0.7852, + "step": 3551 + }, + { + "epoch": 0.3203318753663706, + "grad_norm": 1.8820574500895615, + "learning_rate": 3.1791124530891315e-06, + "loss": 0.692, + "step": 3552 + }, + { + "epoch": 0.3204220588898408, + "grad_norm": 1.9489359985028694, + "learning_rate": 3.178640519894886e-06, + "loss": 0.7738, + "step": 3553 + }, + { + "epoch": 0.3205122424133111, + "grad_norm": 1.5504382928933969, + "learning_rate": 3.1781684861325324e-06, + "loss": 0.7298, + "step": 3554 + }, + { + "epoch": 0.32060242593678134, + "grad_norm": 1.6984410402599732, + "learning_rate": 3.177696351842348e-06, + "loss": 0.7779, + "step": 3555 + }, + { + "epoch": 0.32069260946025163, + "grad_norm": 0.7159855796224921, + "learning_rate": 3.1772241170646167e-06, + "loss": 0.5604, + "step": 3556 + }, + { + "epoch": 0.32078279298372187, + "grad_norm": 1.8978614718147186, + "learning_rate": 3.1767517818396334e-06, + "loss": 0.8783, + "step": 3557 + }, + { + "epoch": 0.32087297650719215, + "grad_norm": 5.153159585238701, + "learning_rate": 3.1762793462076986e-06, + "loss": 0.9283, + "step": 3558 + }, + { + "epoch": 0.3209631600306624, + "grad_norm": 2.11956473470659, + "learning_rate": 3.1758068102091236e-06, + "loss": 0.6641, + "step": 3559 + }, + { + "epoch": 0.3210533435541327, + "grad_norm": 3.024324351117406, + "learning_rate": 3.175334173884229e-06, + "loss": 0.7414, + "step": 3560 + }, + { + "epoch": 0.3211435270776029, + "grad_norm": 1.4322768147016431, + "learning_rate": 3.174861437273342e-06, + "loss": 0.7722, + "step": 3561 + }, + { + "epoch": 0.3212337106010732, + "grad_norm": 1.590076476363195, + "learning_rate": 3.174388600416799e-06, + "loss": 0.733, + "step": 3562 + }, + { + "epoch": 0.32132389412454343, + "grad_norm": 0.6654711872458359, + "learning_rate": 3.1739156633549445e-06, + "loss": 0.571, + "step": 3563 + }, + { + "epoch": 0.3214140776480137, + "grad_norm": 1.932760348036586, + "learning_rate": 3.173442626128133e-06, + "loss": 0.8371, + "step": 3564 + }, + { + "epoch": 0.32150426117148395, + "grad_norm": 2.8718757239206956, + "learning_rate": 3.1729694887767265e-06, + "loss": 0.8844, + "step": 3565 + }, + { + "epoch": 0.32159444469495424, + "grad_norm": 2.9487829833795582, + "learning_rate": 3.172496251341096e-06, + "loss": 0.7744, + "step": 3566 + }, + { + "epoch": 0.3216846282184245, + "grad_norm": 1.6512901608523436, + "learning_rate": 3.172022913861619e-06, + "loss": 0.6271, + "step": 3567 + }, + { + "epoch": 0.32177481174189476, + "grad_norm": 1.9094574631085712, + "learning_rate": 3.171549476378686e-06, + "loss": 0.7335, + "step": 3568 + }, + { + "epoch": 0.321864995265365, + "grad_norm": 1.7211745067876039, + "learning_rate": 3.1710759389326906e-06, + "loss": 0.7444, + "step": 3569 + }, + { + "epoch": 0.3219551787888353, + "grad_norm": 1.7940679766769552, + "learning_rate": 3.1706023015640396e-06, + "loss": 0.7651, + "step": 3570 + }, + { + "epoch": 0.3220453623123055, + "grad_norm": 1.855252737220356, + "learning_rate": 3.1701285643131453e-06, + "loss": 0.7908, + "step": 3571 + }, + { + "epoch": 0.3221355458357758, + "grad_norm": 2.0313577476681033, + "learning_rate": 3.16965472722043e-06, + "loss": 0.7856, + "step": 3572 + }, + { + "epoch": 0.32222572935924604, + "grad_norm": 2.833825536124172, + "learning_rate": 3.169180790326324e-06, + "loss": 0.6989, + "step": 3573 + }, + { + "epoch": 0.3223159128827163, + "grad_norm": 1.5199392207441946, + "learning_rate": 3.168706753671266e-06, + "loss": 0.777, + "step": 3574 + }, + { + "epoch": 0.3224060964061866, + "grad_norm": 2.265306016236766, + "learning_rate": 3.168232617295704e-06, + "loss": 0.6966, + "step": 3575 + }, + { + "epoch": 0.32249627992965685, + "grad_norm": 1.8879580642964806, + "learning_rate": 3.167758381240093e-06, + "loss": 0.7676, + "step": 3576 + }, + { + "epoch": 0.32258646345312714, + "grad_norm": 1.6601515825991375, + "learning_rate": 3.1672840455448978e-06, + "loss": 0.8161, + "step": 3577 + }, + { + "epoch": 0.32267664697659737, + "grad_norm": 1.6620439635336817, + "learning_rate": 3.166809610250592e-06, + "loss": 0.7158, + "step": 3578 + }, + { + "epoch": 0.32276683050006766, + "grad_norm": 1.5785751456159587, + "learning_rate": 3.166335075397656e-06, + "loss": 0.7023, + "step": 3579 + }, + { + "epoch": 0.3228570140235379, + "grad_norm": 1.9013871249195453, + "learning_rate": 3.1658604410265808e-06, + "loss": 0.7361, + "step": 3580 + }, + { + "epoch": 0.3229471975470082, + "grad_norm": 0.8849388441410629, + "learning_rate": 3.1653857071778644e-06, + "loss": 0.5998, + "step": 3581 + }, + { + "epoch": 0.3230373810704784, + "grad_norm": 1.829491976382334, + "learning_rate": 3.1649108738920133e-06, + "loss": 0.7644, + "step": 3582 + }, + { + "epoch": 0.3231275645939487, + "grad_norm": 1.6840137214554676, + "learning_rate": 3.1644359412095432e-06, + "loss": 0.7753, + "step": 3583 + }, + { + "epoch": 0.32321774811741893, + "grad_norm": 0.7552423052984548, + "learning_rate": 3.163960909170978e-06, + "loss": 0.6094, + "step": 3584 + }, + { + "epoch": 0.3233079316408892, + "grad_norm": 1.6067766862347395, + "learning_rate": 3.1634857778168496e-06, + "loss": 0.6702, + "step": 3585 + }, + { + "epoch": 0.32339811516435946, + "grad_norm": 2.6709027222314807, + "learning_rate": 3.1630105471877002e-06, + "loss": 0.8073, + "step": 3586 + }, + { + "epoch": 0.32348829868782975, + "grad_norm": 2.1039650949548983, + "learning_rate": 3.162535217324077e-06, + "loss": 0.7561, + "step": 3587 + }, + { + "epoch": 0.3235784822113, + "grad_norm": 4.012961844130987, + "learning_rate": 3.1620597882665393e-06, + "loss": 0.8015, + "step": 3588 + }, + { + "epoch": 0.32366866573477027, + "grad_norm": 1.8802528552193576, + "learning_rate": 3.1615842600556535e-06, + "loss": 0.8307, + "step": 3589 + }, + { + "epoch": 0.3237588492582405, + "grad_norm": 1.6061129797296347, + "learning_rate": 3.1611086327319932e-06, + "loss": 0.767, + "step": 3590 + }, + { + "epoch": 0.3238490327817108, + "grad_norm": 1.6758353840369462, + "learning_rate": 3.160632906336142e-06, + "loss": 0.8393, + "step": 3591 + }, + { + "epoch": 0.323939216305181, + "grad_norm": 2.275583309806922, + "learning_rate": 3.160157080908692e-06, + "loss": 0.7939, + "step": 3592 + }, + { + "epoch": 0.3240293998286513, + "grad_norm": 2.072022160385811, + "learning_rate": 3.1596811564902426e-06, + "loss": 0.8296, + "step": 3593 + }, + { + "epoch": 0.32411958335212154, + "grad_norm": 2.3654828370285443, + "learning_rate": 3.1592051331214023e-06, + "loss": 0.7166, + "step": 3594 + }, + { + "epoch": 0.32420976687559183, + "grad_norm": 1.6798209203915044, + "learning_rate": 3.158729010842789e-06, + "loss": 0.7971, + "step": 3595 + }, + { + "epoch": 0.32429995039906206, + "grad_norm": 1.6915240169695318, + "learning_rate": 3.1582527896950266e-06, + "loss": 0.8182, + "step": 3596 + }, + { + "epoch": 0.32439013392253235, + "grad_norm": 3.4347335780963335, + "learning_rate": 3.157776469718749e-06, + "loss": 0.7885, + "step": 3597 + }, + { + "epoch": 0.32448031744600264, + "grad_norm": 1.8914107274188021, + "learning_rate": 3.1573000509546004e-06, + "loss": 0.7603, + "step": 3598 + }, + { + "epoch": 0.3245705009694729, + "grad_norm": 2.083796742080352, + "learning_rate": 3.1568235334432296e-06, + "loss": 0.8241, + "step": 3599 + }, + { + "epoch": 0.32466068449294316, + "grad_norm": 1.8261160612501834, + "learning_rate": 3.1563469172252964e-06, + "loss": 0.8444, + "step": 3600 + }, + { + "epoch": 0.3247508680164134, + "grad_norm": 1.9717596754879254, + "learning_rate": 3.155870202341468e-06, + "loss": 0.8471, + "step": 3601 + }, + { + "epoch": 0.3248410515398837, + "grad_norm": 2.802430160972599, + "learning_rate": 3.155393388832421e-06, + "loss": 0.7641, + "step": 3602 + }, + { + "epoch": 0.3249312350633539, + "grad_norm": 1.8748460706468666, + "learning_rate": 3.1549164767388386e-06, + "loss": 0.8023, + "step": 3603 + }, + { + "epoch": 0.3250214185868242, + "grad_norm": 2.0915734664194052, + "learning_rate": 3.1544394661014145e-06, + "loss": 0.7741, + "step": 3604 + }, + { + "epoch": 0.32511160211029444, + "grad_norm": 1.4384870871931255, + "learning_rate": 3.15396235696085e-06, + "loss": 0.6884, + "step": 3605 + }, + { + "epoch": 0.32520178563376473, + "grad_norm": 1.5572598134060947, + "learning_rate": 3.153485149357854e-06, + "loss": 0.7635, + "step": 3606 + }, + { + "epoch": 0.32529196915723496, + "grad_norm": 1.727340729985558, + "learning_rate": 3.153007843333145e-06, + "loss": 0.7642, + "step": 3607 + }, + { + "epoch": 0.32538215268070525, + "grad_norm": 1.8106868488149561, + "learning_rate": 3.152530438927449e-06, + "loss": 0.8041, + "step": 3608 + }, + { + "epoch": 0.3254723362041755, + "grad_norm": 1.8391882504209431, + "learning_rate": 3.1520529361815008e-06, + "loss": 0.8178, + "step": 3609 + }, + { + "epoch": 0.32556251972764577, + "grad_norm": 1.56128821113195, + "learning_rate": 3.151575335136044e-06, + "loss": 0.8182, + "step": 3610 + }, + { + "epoch": 0.325652703251116, + "grad_norm": 2.9209456164172902, + "learning_rate": 3.1510976358318298e-06, + "loss": 0.7563, + "step": 3611 + }, + { + "epoch": 0.3257428867745863, + "grad_norm": 1.6619098332869415, + "learning_rate": 3.1506198383096186e-06, + "loss": 0.7272, + "step": 3612 + }, + { + "epoch": 0.3258330702980565, + "grad_norm": 1.5923777637848489, + "learning_rate": 3.150141942610178e-06, + "loss": 0.7913, + "step": 3613 + }, + { + "epoch": 0.3259232538215268, + "grad_norm": 0.6977464530018196, + "learning_rate": 3.1496639487742853e-06, + "loss": 0.5873, + "step": 3614 + }, + { + "epoch": 0.32601343734499705, + "grad_norm": 2.242426630221165, + "learning_rate": 3.1491858568427247e-06, + "loss": 0.8227, + "step": 3615 + }, + { + "epoch": 0.32610362086846734, + "grad_norm": 2.0437631029181462, + "learning_rate": 3.1487076668562903e-06, + "loss": 0.8304, + "step": 3616 + }, + { + "epoch": 0.32619380439193757, + "grad_norm": 3.2022015985122874, + "learning_rate": 3.1482293788557847e-06, + "loss": 0.813, + "step": 3617 + }, + { + "epoch": 0.32628398791540786, + "grad_norm": 1.6986253425720754, + "learning_rate": 3.1477509928820165e-06, + "loss": 0.7404, + "step": 3618 + }, + { + "epoch": 0.3263741714388781, + "grad_norm": 1.680546172915382, + "learning_rate": 3.147272508975805e-06, + "loss": 0.7724, + "step": 3619 + }, + { + "epoch": 0.3264643549623484, + "grad_norm": 0.6838621430078339, + "learning_rate": 3.1467939271779775e-06, + "loss": 0.6065, + "step": 3620 + }, + { + "epoch": 0.32655453848581867, + "grad_norm": 1.8206742872644153, + "learning_rate": 3.146315247529368e-06, + "loss": 0.735, + "step": 3621 + }, + { + "epoch": 0.3266447220092889, + "grad_norm": 1.8689515786549777, + "learning_rate": 3.1458364700708212e-06, + "loss": 0.7355, + "step": 3622 + }, + { + "epoch": 0.3267349055327592, + "grad_norm": 6.424411838788453, + "learning_rate": 3.1453575948431892e-06, + "loss": 0.7994, + "step": 3623 + }, + { + "epoch": 0.3268250890562294, + "grad_norm": 2.088230658481007, + "learning_rate": 3.144878621887331e-06, + "loss": 0.7794, + "step": 3624 + }, + { + "epoch": 0.3269152725796997, + "grad_norm": 6.850352268489014, + "learning_rate": 3.1443995512441167e-06, + "loss": 0.8172, + "step": 3625 + }, + { + "epoch": 0.32700545610316994, + "grad_norm": 1.5791835445149642, + "learning_rate": 3.1439203829544224e-06, + "loss": 0.8405, + "step": 3626 + }, + { + "epoch": 0.32709563962664023, + "grad_norm": 1.5156918540446263, + "learning_rate": 3.143441117059133e-06, + "loss": 0.8554, + "step": 3627 + }, + { + "epoch": 0.32718582315011047, + "grad_norm": 1.7920841869052728, + "learning_rate": 3.142961753599143e-06, + "loss": 0.7274, + "step": 3628 + }, + { + "epoch": 0.32727600667358075, + "grad_norm": 1.5360316918054102, + "learning_rate": 3.1424822926153543e-06, + "loss": 0.7425, + "step": 3629 + }, + { + "epoch": 0.327366190197051, + "grad_norm": 1.9764858426996668, + "learning_rate": 3.142002734148676e-06, + "loss": 0.8172, + "step": 3630 + }, + { + "epoch": 0.3274563737205213, + "grad_norm": 1.916400897908503, + "learning_rate": 3.141523078240028e-06, + "loss": 0.8085, + "step": 3631 + }, + { + "epoch": 0.3275465572439915, + "grad_norm": 2.2641830591257266, + "learning_rate": 3.1410433249303366e-06, + "loss": 0.8359, + "step": 3632 + }, + { + "epoch": 0.3276367407674618, + "grad_norm": 2.5296345017779642, + "learning_rate": 3.1405634742605366e-06, + "loss": 0.8671, + "step": 3633 + }, + { + "epoch": 0.32772692429093203, + "grad_norm": 1.9794627386461092, + "learning_rate": 3.1400835262715727e-06, + "loss": 0.8143, + "step": 3634 + }, + { + "epoch": 0.3278171078144023, + "grad_norm": 2.0938679676130247, + "learning_rate": 3.139603481004396e-06, + "loss": 0.7696, + "step": 3635 + }, + { + "epoch": 0.32790729133787255, + "grad_norm": 1.7746335336332622, + "learning_rate": 3.139123338499966e-06, + "loss": 0.7887, + "step": 3636 + }, + { + "epoch": 0.32799747486134284, + "grad_norm": 2.555681694299371, + "learning_rate": 3.1386430987992524e-06, + "loss": 0.8149, + "step": 3637 + }, + { + "epoch": 0.3280876583848131, + "grad_norm": 1.7406890309471468, + "learning_rate": 3.1381627619432307e-06, + "loss": 0.7237, + "step": 3638 + }, + { + "epoch": 0.32817784190828336, + "grad_norm": 2.668652738634723, + "learning_rate": 3.1376823279728864e-06, + "loss": 0.7881, + "step": 3639 + }, + { + "epoch": 0.3282680254317536, + "grad_norm": 1.7090792384385571, + "learning_rate": 3.1372017969292125e-06, + "loss": 0.8577, + "step": 3640 + }, + { + "epoch": 0.3283582089552239, + "grad_norm": 6.904800470053425, + "learning_rate": 3.136721168853211e-06, + "loss": 0.7595, + "step": 3641 + }, + { + "epoch": 0.3284483924786941, + "grad_norm": 2.3196286821828975, + "learning_rate": 3.1362404437858924e-06, + "loss": 0.7425, + "step": 3642 + }, + { + "epoch": 0.3285385760021644, + "grad_norm": 1.8444410419177393, + "learning_rate": 3.135759621768273e-06, + "loss": 0.8344, + "step": 3643 + }, + { + "epoch": 0.32862875952563464, + "grad_norm": 2.3624915955409165, + "learning_rate": 3.13527870284138e-06, + "loss": 0.8274, + "step": 3644 + }, + { + "epoch": 0.32871894304910493, + "grad_norm": 2.60888502808819, + "learning_rate": 3.134797687046249e-06, + "loss": 0.8399, + "step": 3645 + }, + { + "epoch": 0.3288091265725752, + "grad_norm": 2.175304860640194, + "learning_rate": 3.1343165744239218e-06, + "loss": 0.7184, + "step": 3646 + }, + { + "epoch": 0.32889931009604545, + "grad_norm": 1.869521593514486, + "learning_rate": 3.13383536501545e-06, + "loss": 0.7981, + "step": 3647 + }, + { + "epoch": 0.32898949361951574, + "grad_norm": 1.944429863769117, + "learning_rate": 3.133354058861893e-06, + "loss": 0.8001, + "step": 3648 + }, + { + "epoch": 0.32907967714298597, + "grad_norm": 1.7350606851655868, + "learning_rate": 3.132872656004318e-06, + "loss": 0.746, + "step": 3649 + }, + { + "epoch": 0.32916986066645626, + "grad_norm": 1.6217950613795542, + "learning_rate": 3.132391156483802e-06, + "loss": 0.6811, + "step": 3650 + }, + { + "epoch": 0.3292600441899265, + "grad_norm": 5.2550740018474125, + "learning_rate": 3.131909560341428e-06, + "loss": 0.7428, + "step": 3651 + }, + { + "epoch": 0.3293502277133968, + "grad_norm": 3.7563655232518176, + "learning_rate": 3.1314278676182893e-06, + "loss": 0.7246, + "step": 3652 + }, + { + "epoch": 0.329440411236867, + "grad_norm": 1.690661753256745, + "learning_rate": 3.130946078355486e-06, + "loss": 0.8462, + "step": 3653 + }, + { + "epoch": 0.3295305947603373, + "grad_norm": 2.2140557503715153, + "learning_rate": 3.130464192594128e-06, + "loss": 0.8134, + "step": 3654 + }, + { + "epoch": 0.32962077828380754, + "grad_norm": 1.5904849397341423, + "learning_rate": 3.1299822103753315e-06, + "loss": 0.7055, + "step": 3655 + }, + { + "epoch": 0.3297109618072778, + "grad_norm": 1.8425583948951612, + "learning_rate": 3.1295001317402217e-06, + "loss": 0.7798, + "step": 3656 + }, + { + "epoch": 0.32980114533074806, + "grad_norm": 2.5842120819217964, + "learning_rate": 3.1290179567299335e-06, + "loss": 0.7215, + "step": 3657 + }, + { + "epoch": 0.32989132885421835, + "grad_norm": 1.6455493231271923, + "learning_rate": 3.128535685385607e-06, + "loss": 0.8071, + "step": 3658 + }, + { + "epoch": 0.3299815123776886, + "grad_norm": 1.6562418667575582, + "learning_rate": 3.1280533177483935e-06, + "loss": 0.8585, + "step": 3659 + }, + { + "epoch": 0.33007169590115887, + "grad_norm": 1.4064844995587702, + "learning_rate": 3.127570853859451e-06, + "loss": 0.7891, + "step": 3660 + }, + { + "epoch": 0.3301618794246291, + "grad_norm": 2.2540773958120948, + "learning_rate": 3.1270882937599456e-06, + "loss": 0.6676, + "step": 3661 + }, + { + "epoch": 0.3302520629480994, + "grad_norm": 3.1090064669026787, + "learning_rate": 3.1266056374910532e-06, + "loss": 0.8012, + "step": 3662 + }, + { + "epoch": 0.3303422464715696, + "grad_norm": 1.966161388478841, + "learning_rate": 3.126122885093955e-06, + "loss": 0.6761, + "step": 3663 + }, + { + "epoch": 0.3304324299950399, + "grad_norm": 1.8480950574413404, + "learning_rate": 3.1256400366098427e-06, + "loss": 0.7744, + "step": 3664 + }, + { + "epoch": 0.33052261351851014, + "grad_norm": 1.7543554601610598, + "learning_rate": 3.125157092079916e-06, + "loss": 0.7327, + "step": 3665 + }, + { + "epoch": 0.33061279704198043, + "grad_norm": 4.331899161083609, + "learning_rate": 3.1246740515453824e-06, + "loss": 0.7511, + "step": 3666 + }, + { + "epoch": 0.33070298056545067, + "grad_norm": 1.5534186624794137, + "learning_rate": 3.124190915047457e-06, + "loss": 0.716, + "step": 3667 + }, + { + "epoch": 0.33079316408892095, + "grad_norm": 1.9169673752891503, + "learning_rate": 3.123707682627364e-06, + "loss": 0.7461, + "step": 3668 + }, + { + "epoch": 0.33088334761239124, + "grad_norm": 2.0669257334898252, + "learning_rate": 3.1232243543263356e-06, + "loss": 0.7584, + "step": 3669 + }, + { + "epoch": 0.3309735311358615, + "grad_norm": 1.7503522790830175, + "learning_rate": 3.1227409301856122e-06, + "loss": 0.6829, + "step": 3670 + }, + { + "epoch": 0.33106371465933176, + "grad_norm": 1.5124413786317827, + "learning_rate": 3.1222574102464413e-06, + "loss": 0.753, + "step": 3671 + }, + { + "epoch": 0.331153898182802, + "grad_norm": 1.5080676875822219, + "learning_rate": 3.12177379455008e-06, + "loss": 0.8046, + "step": 3672 + }, + { + "epoch": 0.3312440817062723, + "grad_norm": 1.9143515855115119, + "learning_rate": 3.121290083137794e-06, + "loss": 0.7673, + "step": 3673 + }, + { + "epoch": 0.3313342652297425, + "grad_norm": 2.1411137823367428, + "learning_rate": 3.1208062760508547e-06, + "loss": 0.7493, + "step": 3674 + }, + { + "epoch": 0.3314244487532128, + "grad_norm": 1.5096012238170553, + "learning_rate": 3.1203223733305438e-06, + "loss": 0.7143, + "step": 3675 + }, + { + "epoch": 0.33151463227668304, + "grad_norm": 2.293757912105042, + "learning_rate": 3.1198383750181512e-06, + "loss": 0.5646, + "step": 3676 + }, + { + "epoch": 0.33160481580015333, + "grad_norm": 1.5933149155279835, + "learning_rate": 3.1193542811549734e-06, + "loss": 0.6939, + "step": 3677 + }, + { + "epoch": 0.33169499932362356, + "grad_norm": 1.7228301429779271, + "learning_rate": 3.1188700917823166e-06, + "loss": 0.6996, + "step": 3678 + }, + { + "epoch": 0.33178518284709385, + "grad_norm": 0.7026733219243546, + "learning_rate": 3.1183858069414937e-06, + "loss": 0.6263, + "step": 3679 + }, + { + "epoch": 0.3318753663705641, + "grad_norm": 2.1559210263723356, + "learning_rate": 3.117901426673827e-06, + "loss": 0.827, + "step": 3680 + }, + { + "epoch": 0.3319655498940344, + "grad_norm": 2.188283180636365, + "learning_rate": 3.1174169510206466e-06, + "loss": 0.7323, + "step": 3681 + }, + { + "epoch": 0.3320557334175046, + "grad_norm": 1.6561780949944598, + "learning_rate": 3.1169323800232908e-06, + "loss": 0.7723, + "step": 3682 + }, + { + "epoch": 0.3321459169409749, + "grad_norm": 1.6879806540451778, + "learning_rate": 3.1164477137231054e-06, + "loss": 0.8074, + "step": 3683 + }, + { + "epoch": 0.3322361004644451, + "grad_norm": 2.376958340816997, + "learning_rate": 3.115962952161445e-06, + "loss": 0.7796, + "step": 3684 + }, + { + "epoch": 0.3323262839879154, + "grad_norm": 2.085904581903182, + "learning_rate": 3.1154780953796727e-06, + "loss": 0.8143, + "step": 3685 + }, + { + "epoch": 0.33241646751138565, + "grad_norm": 3.6268344052409605, + "learning_rate": 3.114993143419158e-06, + "loss": 0.6803, + "step": 3686 + }, + { + "epoch": 0.33250665103485594, + "grad_norm": 1.3743029908540807, + "learning_rate": 3.1145080963212806e-06, + "loss": 0.6814, + "step": 3687 + }, + { + "epoch": 0.33259683455832617, + "grad_norm": 1.5731011676157631, + "learning_rate": 3.114022954127427e-06, + "loss": 0.7724, + "step": 3688 + }, + { + "epoch": 0.33268701808179646, + "grad_norm": 1.6332768414258236, + "learning_rate": 3.1135377168789923e-06, + "loss": 0.9808, + "step": 3689 + }, + { + "epoch": 0.3327772016052667, + "grad_norm": 1.8876297609301897, + "learning_rate": 3.1130523846173803e-06, + "loss": 0.838, + "step": 3690 + }, + { + "epoch": 0.332867385128737, + "grad_norm": 1.8849333535309458, + "learning_rate": 3.1125669573840006e-06, + "loss": 0.7163, + "step": 3691 + }, + { + "epoch": 0.3329575686522072, + "grad_norm": 2.850064487731552, + "learning_rate": 3.112081435220274e-06, + "loss": 0.7151, + "step": 3692 + }, + { + "epoch": 0.3330477521756775, + "grad_norm": 1.5949058829016622, + "learning_rate": 3.111595818167627e-06, + "loss": 0.7618, + "step": 3693 + }, + { + "epoch": 0.3331379356991478, + "grad_norm": 2.1613945886488604, + "learning_rate": 3.1111101062674953e-06, + "loss": 0.835, + "step": 3694 + }, + { + "epoch": 0.333228119222618, + "grad_norm": 1.603105573926869, + "learning_rate": 3.1106242995613233e-06, + "loss": 0.7592, + "step": 3695 + }, + { + "epoch": 0.3333183027460883, + "grad_norm": 1.6132335886715359, + "learning_rate": 3.1101383980905616e-06, + "loss": 0.7725, + "step": 3696 + }, + { + "epoch": 0.33340848626955855, + "grad_norm": 0.647811886593445, + "learning_rate": 3.109652401896671e-06, + "loss": 0.5516, + "step": 3697 + }, + { + "epoch": 0.33349866979302883, + "grad_norm": 1.457512988102985, + "learning_rate": 3.109166311021119e-06, + "loss": 0.8172, + "step": 3698 + }, + { + "epoch": 0.33358885331649907, + "grad_norm": 2.0367023942917677, + "learning_rate": 3.1086801255053807e-06, + "loss": 0.8067, + "step": 3699 + }, + { + "epoch": 0.33367903683996936, + "grad_norm": 1.4265620353840032, + "learning_rate": 3.108193845390942e-06, + "loss": 0.8084, + "step": 3700 + }, + { + "epoch": 0.3337692203634396, + "grad_norm": 1.3145982228608208, + "learning_rate": 3.1077074707192933e-06, + "loss": 0.7483, + "step": 3701 + }, + { + "epoch": 0.3338594038869099, + "grad_norm": 1.4653680702259733, + "learning_rate": 3.1072210015319353e-06, + "loss": 0.792, + "step": 3702 + }, + { + "epoch": 0.3339495874103801, + "grad_norm": 1.8315342470966696, + "learning_rate": 3.106734437870376e-06, + "loss": 0.7335, + "step": 3703 + }, + { + "epoch": 0.3340397709338504, + "grad_norm": 1.9125933555805936, + "learning_rate": 3.1062477797761327e-06, + "loss": 0.7187, + "step": 3704 + }, + { + "epoch": 0.33412995445732063, + "grad_norm": 2.0576929658714573, + "learning_rate": 3.105761027290729e-06, + "loss": 0.8193, + "step": 3705 + }, + { + "epoch": 0.3342201379807909, + "grad_norm": 1.9586010269741971, + "learning_rate": 3.105274180455697e-06, + "loss": 0.7414, + "step": 3706 + }, + { + "epoch": 0.33431032150426115, + "grad_norm": 1.6813948359729922, + "learning_rate": 3.1047872393125775e-06, + "loss": 0.7964, + "step": 3707 + }, + { + "epoch": 0.33440050502773144, + "grad_norm": 0.6245203084705402, + "learning_rate": 3.1043002039029186e-06, + "loss": 0.5792, + "step": 3708 + }, + { + "epoch": 0.3344906885512017, + "grad_norm": 1.4046773806197912, + "learning_rate": 3.1038130742682782e-06, + "loss": 0.7617, + "step": 3709 + }, + { + "epoch": 0.33458087207467196, + "grad_norm": 1.417757633350792, + "learning_rate": 3.103325850450219e-06, + "loss": 0.7481, + "step": 3710 + }, + { + "epoch": 0.3346710555981422, + "grad_norm": 1.8477414778878174, + "learning_rate": 3.1028385324903154e-06, + "loss": 0.8134, + "step": 3711 + }, + { + "epoch": 0.3347612391216125, + "grad_norm": 1.6339207588741589, + "learning_rate": 3.1023511204301465e-06, + "loss": 0.7794, + "step": 3712 + }, + { + "epoch": 0.3348514226450827, + "grad_norm": 1.5698846642231437, + "learning_rate": 3.1018636143113022e-06, + "loss": 0.7309, + "step": 3713 + }, + { + "epoch": 0.334941606168553, + "grad_norm": 1.7129282756917128, + "learning_rate": 3.1013760141753787e-06, + "loss": 0.6811, + "step": 3714 + }, + { + "epoch": 0.33503178969202324, + "grad_norm": 1.574762368819344, + "learning_rate": 3.100888320063981e-06, + "loss": 0.7405, + "step": 3715 + }, + { + "epoch": 0.33512197321549353, + "grad_norm": 1.4546015631084208, + "learning_rate": 3.100400532018721e-06, + "loss": 0.7773, + "step": 3716 + }, + { + "epoch": 0.3352121567389638, + "grad_norm": 1.98607598639005, + "learning_rate": 3.0999126500812204e-06, + "loss": 0.8407, + "step": 3717 + }, + { + "epoch": 0.33530234026243405, + "grad_norm": 0.7213932444503347, + "learning_rate": 3.0994246742931076e-06, + "loss": 0.6342, + "step": 3718 + }, + { + "epoch": 0.33539252378590434, + "grad_norm": 1.888553840554485, + "learning_rate": 3.098936604696019e-06, + "loss": 0.7123, + "step": 3719 + }, + { + "epoch": 0.3354827073093746, + "grad_norm": 1.8342226501314194, + "learning_rate": 3.0984484413316e-06, + "loss": 0.8004, + "step": 3720 + }, + { + "epoch": 0.33557289083284486, + "grad_norm": 2.177963066922882, + "learning_rate": 3.0979601842415033e-06, + "loss": 0.7523, + "step": 3721 + }, + { + "epoch": 0.3356630743563151, + "grad_norm": 3.5797074329448986, + "learning_rate": 3.0974718334673896e-06, + "loss": 0.6923, + "step": 3722 + }, + { + "epoch": 0.3357532578797854, + "grad_norm": 1.5103933749102108, + "learning_rate": 3.0969833890509282e-06, + "loss": 0.7618, + "step": 3723 + }, + { + "epoch": 0.3358434414032556, + "grad_norm": 2.518808849773638, + "learning_rate": 3.096494851033795e-06, + "loss": 0.7698, + "step": 3724 + }, + { + "epoch": 0.3359336249267259, + "grad_norm": 1.790930160430209, + "learning_rate": 3.0960062194576747e-06, + "loss": 0.7001, + "step": 3725 + }, + { + "epoch": 0.33602380845019614, + "grad_norm": 1.784437439689543, + "learning_rate": 3.0955174943642606e-06, + "loss": 0.7231, + "step": 3726 + }, + { + "epoch": 0.3361139919736664, + "grad_norm": 1.9617282443346635, + "learning_rate": 3.0950286757952534e-06, + "loss": 0.6928, + "step": 3727 + }, + { + "epoch": 0.33620417549713666, + "grad_norm": 1.8266934389509961, + "learning_rate": 3.0945397637923617e-06, + "loss": 0.834, + "step": 3728 + }, + { + "epoch": 0.33629435902060695, + "grad_norm": 2.005714596999458, + "learning_rate": 3.0940507583973025e-06, + "loss": 0.7795, + "step": 3729 + }, + { + "epoch": 0.3363845425440772, + "grad_norm": 1.6475714266030068, + "learning_rate": 3.093561659651799e-06, + "loss": 0.7415, + "step": 3730 + }, + { + "epoch": 0.33647472606754747, + "grad_norm": 1.8124707975994756, + "learning_rate": 3.093072467597586e-06, + "loss": 0.7082, + "step": 3731 + }, + { + "epoch": 0.3365649095910177, + "grad_norm": 0.6420087291697477, + "learning_rate": 3.092583182276402e-06, + "loss": 0.5844, + "step": 3732 + }, + { + "epoch": 0.336655093114488, + "grad_norm": 0.6360278681626408, + "learning_rate": 3.092093803729997e-06, + "loss": 0.5905, + "step": 3733 + }, + { + "epoch": 0.3367452766379582, + "grad_norm": 1.6311897354394214, + "learning_rate": 3.0916043320001264e-06, + "loss": 0.7819, + "step": 3734 + }, + { + "epoch": 0.3368354601614285, + "grad_norm": 1.7275538295634398, + "learning_rate": 3.0911147671285557e-06, + "loss": 0.782, + "step": 3735 + }, + { + "epoch": 0.33692564368489875, + "grad_norm": 2.3174779770348946, + "learning_rate": 3.0906251091570565e-06, + "loss": 0.799, + "step": 3736 + }, + { + "epoch": 0.33701582720836903, + "grad_norm": 1.7598836268618545, + "learning_rate": 3.0901353581274094e-06, + "loss": 0.782, + "step": 3737 + }, + { + "epoch": 0.33710601073183927, + "grad_norm": 1.949922423164538, + "learning_rate": 3.089645514081402e-06, + "loss": 0.6499, + "step": 3738 + }, + { + "epoch": 0.33719619425530956, + "grad_norm": 2.72661802568708, + "learning_rate": 3.0891555770608323e-06, + "loss": 0.702, + "step": 3739 + }, + { + "epoch": 0.33728637777877984, + "grad_norm": 1.7134897365257122, + "learning_rate": 3.088665547107503e-06, + "loss": 0.793, + "step": 3740 + }, + { + "epoch": 0.3373765613022501, + "grad_norm": 1.4262781400854962, + "learning_rate": 3.0881754242632254e-06, + "loss": 0.8478, + "step": 3741 + }, + { + "epoch": 0.33746674482572037, + "grad_norm": 1.353262917623807, + "learning_rate": 3.0876852085698213e-06, + "loss": 0.8137, + "step": 3742 + }, + { + "epoch": 0.3375569283491906, + "grad_norm": 1.8636973814057416, + "learning_rate": 3.087194900069117e-06, + "loss": 0.7978, + "step": 3743 + }, + { + "epoch": 0.3376471118726609, + "grad_norm": 2.3100056985892152, + "learning_rate": 3.08670449880295e-06, + "loss": 0.8263, + "step": 3744 + }, + { + "epoch": 0.3377372953961311, + "grad_norm": 1.550241371557704, + "learning_rate": 3.086214004813163e-06, + "loss": 0.7912, + "step": 3745 + }, + { + "epoch": 0.3378274789196014, + "grad_norm": 1.6128003602142915, + "learning_rate": 3.0857234181416074e-06, + "loss": 0.8347, + "step": 3746 + }, + { + "epoch": 0.33791766244307164, + "grad_norm": 1.6386306266458177, + "learning_rate": 3.085232738830143e-06, + "loss": 0.8216, + "step": 3747 + }, + { + "epoch": 0.33800784596654193, + "grad_norm": 1.8652602588395477, + "learning_rate": 3.084741966920638e-06, + "loss": 0.8316, + "step": 3748 + }, + { + "epoch": 0.33809802949001216, + "grad_norm": 1.5111621551603398, + "learning_rate": 3.084251102454966e-06, + "loss": 0.7294, + "step": 3749 + }, + { + "epoch": 0.33818821301348245, + "grad_norm": 0.7563495475888231, + "learning_rate": 3.083760145475013e-06, + "loss": 0.6144, + "step": 3750 + }, + { + "epoch": 0.3382783965369527, + "grad_norm": 1.7947713064439972, + "learning_rate": 3.0832690960226678e-06, + "loss": 0.7136, + "step": 3751 + }, + { + "epoch": 0.338368580060423, + "grad_norm": 2.0759053243095713, + "learning_rate": 3.08277795413983e-06, + "loss": 0.7444, + "step": 3752 + }, + { + "epoch": 0.3384587635838932, + "grad_norm": 1.5343044723182016, + "learning_rate": 3.0822867198684073e-06, + "loss": 0.8241, + "step": 3753 + }, + { + "epoch": 0.3385489471073635, + "grad_norm": 2.2536556428525807, + "learning_rate": 3.081795393250314e-06, + "loss": 0.781, + "step": 3754 + }, + { + "epoch": 0.33863913063083373, + "grad_norm": 4.93409147794896, + "learning_rate": 3.081303974327473e-06, + "loss": 0.7472, + "step": 3755 + }, + { + "epoch": 0.338729314154304, + "grad_norm": 2.1403441975881288, + "learning_rate": 3.080812463141814e-06, + "loss": 0.7704, + "step": 3756 + }, + { + "epoch": 0.33881949767777425, + "grad_norm": 7.484848994972381, + "learning_rate": 3.080320859735276e-06, + "loss": 0.6462, + "step": 3757 + }, + { + "epoch": 0.33890968120124454, + "grad_norm": 2.00155019287328, + "learning_rate": 3.079829164149806e-06, + "loss": 0.8223, + "step": 3758 + }, + { + "epoch": 0.3389998647247148, + "grad_norm": 1.7602331829570703, + "learning_rate": 3.0793373764273573e-06, + "loss": 0.7622, + "step": 3759 + }, + { + "epoch": 0.33909004824818506, + "grad_norm": 5.009126154667501, + "learning_rate": 3.078845496609892e-06, + "loss": 0.7974, + "step": 3760 + }, + { + "epoch": 0.3391802317716553, + "grad_norm": 1.58990665845575, + "learning_rate": 3.078353524739381e-06, + "loss": 0.8605, + "step": 3761 + }, + { + "epoch": 0.3392704152951256, + "grad_norm": 1.7083289758890297, + "learning_rate": 3.077861460857801e-06, + "loss": 0.8614, + "step": 3762 + }, + { + "epoch": 0.3393605988185958, + "grad_norm": 1.694414525037147, + "learning_rate": 3.077369305007138e-06, + "loss": 0.7899, + "step": 3763 + }, + { + "epoch": 0.3394507823420661, + "grad_norm": 1.8178276817103585, + "learning_rate": 3.0768770572293852e-06, + "loss": 0.8135, + "step": 3764 + }, + { + "epoch": 0.3395409658655364, + "grad_norm": 1.9001166684319555, + "learning_rate": 3.0763847175665437e-06, + "loss": 0.8559, + "step": 3765 + }, + { + "epoch": 0.3396311493890066, + "grad_norm": 3.019480400101428, + "learning_rate": 3.0758922860606237e-06, + "loss": 0.7147, + "step": 3766 + }, + { + "epoch": 0.3397213329124769, + "grad_norm": 1.6455403401134083, + "learning_rate": 3.0753997627536404e-06, + "loss": 0.7128, + "step": 3767 + }, + { + "epoch": 0.33981151643594715, + "grad_norm": 2.0237500748330963, + "learning_rate": 3.0749071476876203e-06, + "loss": 0.7219, + "step": 3768 + }, + { + "epoch": 0.33990169995941744, + "grad_norm": 1.5760992664502698, + "learning_rate": 3.0744144409045952e-06, + "loss": 0.7377, + "step": 3769 + }, + { + "epoch": 0.33999188348288767, + "grad_norm": 2.196798995590258, + "learning_rate": 3.0739216424466056e-06, + "loss": 0.7935, + "step": 3770 + }, + { + "epoch": 0.34008206700635796, + "grad_norm": 2.008685326760605, + "learning_rate": 3.0734287523557002e-06, + "loss": 0.7545, + "step": 3771 + }, + { + "epoch": 0.3401722505298282, + "grad_norm": 2.306720005513866, + "learning_rate": 3.0729357706739348e-06, + "loss": 0.7234, + "step": 3772 + }, + { + "epoch": 0.3402624340532985, + "grad_norm": 1.5174224263054559, + "learning_rate": 3.0724426974433737e-06, + "loss": 0.7578, + "step": 3773 + }, + { + "epoch": 0.3403526175767687, + "grad_norm": 1.4012916039476324, + "learning_rate": 3.0719495327060874e-06, + "loss": 0.7471, + "step": 3774 + }, + { + "epoch": 0.340442801100239, + "grad_norm": 1.6285077163237067, + "learning_rate": 3.071456276504157e-06, + "loss": 0.7815, + "step": 3775 + }, + { + "epoch": 0.34053298462370923, + "grad_norm": 1.9745372302481066, + "learning_rate": 3.070962928879669e-06, + "loss": 0.7709, + "step": 3776 + }, + { + "epoch": 0.3406231681471795, + "grad_norm": 1.8143904462650267, + "learning_rate": 3.0704694898747185e-06, + "loss": 0.7873, + "step": 3777 + }, + { + "epoch": 0.34071335167064976, + "grad_norm": 1.9700628474904383, + "learning_rate": 3.069975959531408e-06, + "loss": 0.9003, + "step": 3778 + }, + { + "epoch": 0.34080353519412004, + "grad_norm": 0.6711072971455442, + "learning_rate": 3.06948233789185e-06, + "loss": 0.6486, + "step": 3779 + }, + { + "epoch": 0.3408937187175903, + "grad_norm": 2.2240853433426144, + "learning_rate": 3.0689886249981614e-06, + "loss": 0.7928, + "step": 3780 + }, + { + "epoch": 0.34098390224106057, + "grad_norm": 2.5564212806977444, + "learning_rate": 3.0684948208924693e-06, + "loss": 0.8004, + "step": 3781 + }, + { + "epoch": 0.3410740857645308, + "grad_norm": 2.036659900676937, + "learning_rate": 3.068000925616907e-06, + "loss": 0.793, + "step": 3782 + }, + { + "epoch": 0.3411642692880011, + "grad_norm": 2.168789435036373, + "learning_rate": 3.067506939213617e-06, + "loss": 0.7664, + "step": 3783 + }, + { + "epoch": 0.3412544528114713, + "grad_norm": 1.3637662096047984, + "learning_rate": 3.0670128617247493e-06, + "loss": 0.7777, + "step": 3784 + }, + { + "epoch": 0.3413446363349416, + "grad_norm": 1.520665311894419, + "learning_rate": 3.06651869319246e-06, + "loss": 0.687, + "step": 3785 + }, + { + "epoch": 0.34143481985841184, + "grad_norm": 1.6996750492987236, + "learning_rate": 3.0660244336589154e-06, + "loss": 0.7717, + "step": 3786 + }, + { + "epoch": 0.34152500338188213, + "grad_norm": 1.5898681938672934, + "learning_rate": 3.065530083166288e-06, + "loss": 0.7913, + "step": 3787 + }, + { + "epoch": 0.3416151869053524, + "grad_norm": 1.6800545224924586, + "learning_rate": 3.0650356417567586e-06, + "loss": 0.7381, + "step": 3788 + }, + { + "epoch": 0.34170537042882265, + "grad_norm": 1.6482346396347427, + "learning_rate": 3.0645411094725156e-06, + "loss": 0.8251, + "step": 3789 + }, + { + "epoch": 0.34179555395229294, + "grad_norm": 2.115252071176265, + "learning_rate": 3.0640464863557556e-06, + "loss": 0.7712, + "step": 3790 + }, + { + "epoch": 0.3418857374757632, + "grad_norm": 1.5223931278496619, + "learning_rate": 3.063551772448682e-06, + "loss": 0.8097, + "step": 3791 + }, + { + "epoch": 0.34197592099923346, + "grad_norm": 2.1902179726247892, + "learning_rate": 3.0630569677935075e-06, + "loss": 0.7068, + "step": 3792 + }, + { + "epoch": 0.3420661045227037, + "grad_norm": 2.3301878117519004, + "learning_rate": 3.06256207243245e-06, + "loss": 0.8589, + "step": 3793 + }, + { + "epoch": 0.342156288046174, + "grad_norm": 0.6350948243040347, + "learning_rate": 3.0620670864077385e-06, + "loss": 0.5689, + "step": 3794 + }, + { + "epoch": 0.3422464715696442, + "grad_norm": 0.6331674498686833, + "learning_rate": 3.0615720097616063e-06, + "loss": 0.5717, + "step": 3795 + }, + { + "epoch": 0.3423366550931145, + "grad_norm": 1.5866454850847955, + "learning_rate": 3.0610768425362967e-06, + "loss": 0.7297, + "step": 3796 + }, + { + "epoch": 0.34242683861658474, + "grad_norm": 1.9890988810314352, + "learning_rate": 3.0605815847740603e-06, + "loss": 0.7522, + "step": 3797 + }, + { + "epoch": 0.342517022140055, + "grad_norm": 2.090564525054596, + "learning_rate": 3.0600862365171553e-06, + "loss": 0.743, + "step": 3798 + }, + { + "epoch": 0.34260720566352526, + "grad_norm": 1.4781238031684918, + "learning_rate": 3.0595907978078474e-06, + "loss": 0.7549, + "step": 3799 + }, + { + "epoch": 0.34269738918699555, + "grad_norm": 1.710776682920855, + "learning_rate": 3.05909526868841e-06, + "loss": 0.744, + "step": 3800 + }, + { + "epoch": 0.3427875727104658, + "grad_norm": 1.7134776311192952, + "learning_rate": 3.0585996492011243e-06, + "loss": 0.7842, + "step": 3801 + }, + { + "epoch": 0.34287775623393607, + "grad_norm": 1.9013900716273429, + "learning_rate": 3.05810393938828e-06, + "loss": 0.8192, + "step": 3802 + }, + { + "epoch": 0.3429679397574063, + "grad_norm": 1.7625373917529439, + "learning_rate": 3.0576081392921723e-06, + "loss": 0.6871, + "step": 3803 + }, + { + "epoch": 0.3430581232808766, + "grad_norm": 1.6356634775838994, + "learning_rate": 3.057112248955107e-06, + "loss": 0.8163, + "step": 3804 + }, + { + "epoch": 0.3431483068043468, + "grad_norm": 4.042500020005339, + "learning_rate": 3.0566162684193963e-06, + "loss": 0.7948, + "step": 3805 + }, + { + "epoch": 0.3432384903278171, + "grad_norm": 2.0113076984481797, + "learning_rate": 3.056120197727359e-06, + "loss": 0.7224, + "step": 3806 + }, + { + "epoch": 0.34332867385128735, + "grad_norm": 1.669183579027665, + "learning_rate": 3.0556240369213236e-06, + "loss": 0.7607, + "step": 3807 + }, + { + "epoch": 0.34341885737475764, + "grad_norm": 2.0550637716529145, + "learning_rate": 3.055127786043624e-06, + "loss": 0.6865, + "step": 3808 + }, + { + "epoch": 0.34350904089822787, + "grad_norm": 1.476346358369199, + "learning_rate": 3.054631445136604e-06, + "loss": 0.8569, + "step": 3809 + }, + { + "epoch": 0.34359922442169816, + "grad_norm": 11.305717331720514, + "learning_rate": 3.0541350142426147e-06, + "loss": 0.7866, + "step": 3810 + }, + { + "epoch": 0.3436894079451684, + "grad_norm": 1.7669170007345396, + "learning_rate": 3.053638493404012e-06, + "loss": 0.7273, + "step": 3811 + }, + { + "epoch": 0.3437795914686387, + "grad_norm": 1.5937562082207133, + "learning_rate": 3.0531418826631643e-06, + "loss": 0.7114, + "step": 3812 + }, + { + "epoch": 0.34386977499210897, + "grad_norm": 1.7549808328348073, + "learning_rate": 3.052645182062444e-06, + "loss": 0.8522, + "step": 3813 + }, + { + "epoch": 0.3439599585155792, + "grad_norm": 2.4896592856171047, + "learning_rate": 3.0521483916442324e-06, + "loss": 0.816, + "step": 3814 + }, + { + "epoch": 0.3440501420390495, + "grad_norm": 1.795620555029047, + "learning_rate": 3.0516515114509183e-06, + "loss": 0.7663, + "step": 3815 + }, + { + "epoch": 0.3441403255625197, + "grad_norm": 1.3956260764514616, + "learning_rate": 3.0511545415249e-06, + "loss": 0.7206, + "step": 3816 + }, + { + "epoch": 0.34423050908599, + "grad_norm": 1.7077741948445007, + "learning_rate": 3.050657481908579e-06, + "loss": 0.8079, + "step": 3817 + }, + { + "epoch": 0.34432069260946024, + "grad_norm": 1.6870307976489898, + "learning_rate": 3.0501603326443677e-06, + "loss": 0.8834, + "step": 3818 + }, + { + "epoch": 0.34441087613293053, + "grad_norm": 1.3577350502181496, + "learning_rate": 3.049663093774687e-06, + "loss": 0.7548, + "step": 3819 + }, + { + "epoch": 0.34450105965640077, + "grad_norm": 2.14693784219201, + "learning_rate": 3.0491657653419643e-06, + "loss": 0.6724, + "step": 3820 + }, + { + "epoch": 0.34459124317987105, + "grad_norm": 1.5972182739135299, + "learning_rate": 3.0486683473886325e-06, + "loss": 0.819, + "step": 3821 + }, + { + "epoch": 0.3446814267033413, + "grad_norm": 2.3690650468738887, + "learning_rate": 3.0481708399571355e-06, + "loss": 0.8415, + "step": 3822 + }, + { + "epoch": 0.3447716102268116, + "grad_norm": 1.8931091004664578, + "learning_rate": 3.047673243089922e-06, + "loss": 0.7998, + "step": 3823 + }, + { + "epoch": 0.3448617937502818, + "grad_norm": 1.5515711235736402, + "learning_rate": 3.047175556829451e-06, + "loss": 0.8053, + "step": 3824 + }, + { + "epoch": 0.3449519772737521, + "grad_norm": 2.487600571789379, + "learning_rate": 3.046677781218188e-06, + "loss": 0.8208, + "step": 3825 + }, + { + "epoch": 0.34504216079722233, + "grad_norm": 1.5870003732537814, + "learning_rate": 3.0461799162986043e-06, + "loss": 0.7521, + "step": 3826 + }, + { + "epoch": 0.3451323443206926, + "grad_norm": 6.200382270103203, + "learning_rate": 3.045681962113183e-06, + "loss": 0.6987, + "step": 3827 + }, + { + "epoch": 0.34522252784416285, + "grad_norm": 1.701764604600241, + "learning_rate": 3.0451839187044095e-06, + "loss": 0.7226, + "step": 3828 + }, + { + "epoch": 0.34531271136763314, + "grad_norm": 4.250727647311248, + "learning_rate": 3.0446857861147816e-06, + "loss": 0.7902, + "step": 3829 + }, + { + "epoch": 0.3454028948911034, + "grad_norm": 3.319632570547889, + "learning_rate": 3.044187564386802e-06, + "loss": 0.7456, + "step": 3830 + }, + { + "epoch": 0.34549307841457366, + "grad_norm": 1.9374993847261497, + "learning_rate": 3.0436892535629818e-06, + "loss": 0.8367, + "step": 3831 + }, + { + "epoch": 0.3455832619380439, + "grad_norm": 1.7933026835978887, + "learning_rate": 3.0431908536858393e-06, + "loss": 0.856, + "step": 3832 + }, + { + "epoch": 0.3456734454615142, + "grad_norm": 1.9302950995168775, + "learning_rate": 3.0426923647979016e-06, + "loss": 0.762, + "step": 3833 + }, + { + "epoch": 0.3457636289849844, + "grad_norm": 1.9884061942722895, + "learning_rate": 3.0421937869417016e-06, + "loss": 0.7148, + "step": 3834 + }, + { + "epoch": 0.3458538125084547, + "grad_norm": 1.9787169765432595, + "learning_rate": 3.041695120159782e-06, + "loss": 0.8379, + "step": 3835 + }, + { + "epoch": 0.345943996031925, + "grad_norm": 5.49650046247468, + "learning_rate": 3.04119636449469e-06, + "loss": 0.7068, + "step": 3836 + }, + { + "epoch": 0.3460341795553952, + "grad_norm": 2.3050694036068275, + "learning_rate": 3.040697519988983e-06, + "loss": 0.8075, + "step": 3837 + }, + { + "epoch": 0.3461243630788655, + "grad_norm": 0.6433766977077665, + "learning_rate": 3.040198586685226e-06, + "loss": 0.5914, + "step": 3838 + }, + { + "epoch": 0.34621454660233575, + "grad_norm": 1.8566991962525836, + "learning_rate": 3.039699564625989e-06, + "loss": 0.7269, + "step": 3839 + }, + { + "epoch": 0.34630473012580604, + "grad_norm": 0.6771407151971583, + "learning_rate": 3.039200453853853e-06, + "loss": 0.5853, + "step": 3840 + }, + { + "epoch": 0.34639491364927627, + "grad_norm": 1.8275285709720426, + "learning_rate": 3.038701254411404e-06, + "loss": 0.8233, + "step": 3841 + }, + { + "epoch": 0.34648509717274656, + "grad_norm": 1.8239000846701727, + "learning_rate": 3.0382019663412367e-06, + "loss": 0.7951, + "step": 3842 + }, + { + "epoch": 0.3465752806962168, + "grad_norm": 2.4830221655382867, + "learning_rate": 3.0377025896859532e-06, + "loss": 0.8127, + "step": 3843 + }, + { + "epoch": 0.3466654642196871, + "grad_norm": 1.591685435132854, + "learning_rate": 3.0372031244881627e-06, + "loss": 0.8778, + "step": 3844 + }, + { + "epoch": 0.3467556477431573, + "grad_norm": 1.5607025488979234, + "learning_rate": 3.0367035707904826e-06, + "loss": 0.7336, + "step": 3845 + }, + { + "epoch": 0.3468458312666276, + "grad_norm": 1.6888355339068835, + "learning_rate": 3.036203928635537e-06, + "loss": 0.689, + "step": 3846 + }, + { + "epoch": 0.34693601479009784, + "grad_norm": 1.8448922449915288, + "learning_rate": 3.035704198065959e-06, + "loss": 0.8429, + "step": 3847 + }, + { + "epoch": 0.3470261983135681, + "grad_norm": 2.858135684402352, + "learning_rate": 3.0352043791243886e-06, + "loss": 0.7545, + "step": 3848 + }, + { + "epoch": 0.34711638183703836, + "grad_norm": 0.6966976555939377, + "learning_rate": 3.034704471853472e-06, + "loss": 0.5896, + "step": 3849 + }, + { + "epoch": 0.34720656536050865, + "grad_norm": 1.42371467379048, + "learning_rate": 3.0342044762958646e-06, + "loss": 0.8219, + "step": 3850 + }, + { + "epoch": 0.3472967488839789, + "grad_norm": 1.9510916539713492, + "learning_rate": 3.0337043924942286e-06, + "loss": 0.8107, + "step": 3851 + }, + { + "epoch": 0.34738693240744917, + "grad_norm": 1.5924388503666262, + "learning_rate": 3.0332042204912343e-06, + "loss": 0.763, + "step": 3852 + }, + { + "epoch": 0.3474771159309194, + "grad_norm": 0.6788640569957899, + "learning_rate": 3.0327039603295587e-06, + "loss": 0.5935, + "step": 3853 + }, + { + "epoch": 0.3475672994543897, + "grad_norm": 2.3076394007804653, + "learning_rate": 3.032203612051887e-06, + "loss": 0.733, + "step": 3854 + }, + { + "epoch": 0.3476574829778599, + "grad_norm": 2.3474526213881006, + "learning_rate": 3.0317031757009116e-06, + "loss": 0.757, + "step": 3855 + }, + { + "epoch": 0.3477476665013302, + "grad_norm": 1.9114316960972513, + "learning_rate": 3.0312026513193326e-06, + "loss": 0.7351, + "step": 3856 + }, + { + "epoch": 0.34783785002480044, + "grad_norm": 1.6362854012899117, + "learning_rate": 3.0307020389498573e-06, + "loss": 0.7473, + "step": 3857 + }, + { + "epoch": 0.34792803354827073, + "grad_norm": 1.5870570849011272, + "learning_rate": 3.0302013386352004e-06, + "loss": 0.703, + "step": 3858 + }, + { + "epoch": 0.348018217071741, + "grad_norm": 1.4050122641981821, + "learning_rate": 3.0297005504180854e-06, + "loss": 0.7662, + "step": 3859 + }, + { + "epoch": 0.34810840059521125, + "grad_norm": 0.6053607475183085, + "learning_rate": 3.0291996743412417e-06, + "loss": 0.5131, + "step": 3860 + }, + { + "epoch": 0.34819858411868154, + "grad_norm": 1.5243840333765566, + "learning_rate": 3.0286987104474063e-06, + "loss": 0.8667, + "step": 3861 + }, + { + "epoch": 0.3482887676421518, + "grad_norm": 1.51316311069349, + "learning_rate": 3.028197658779325e-06, + "loss": 0.7423, + "step": 3862 + }, + { + "epoch": 0.34837895116562206, + "grad_norm": 1.8913508511497212, + "learning_rate": 3.0276965193797503e-06, + "loss": 0.7893, + "step": 3863 + }, + { + "epoch": 0.3484691346890923, + "grad_norm": 1.4541443561638687, + "learning_rate": 3.0271952922914423e-06, + "loss": 0.7352, + "step": 3864 + }, + { + "epoch": 0.3485593182125626, + "grad_norm": 1.786378471287323, + "learning_rate": 3.0266939775571675e-06, + "loss": 0.7518, + "step": 3865 + }, + { + "epoch": 0.3486495017360328, + "grad_norm": 1.6819174225275693, + "learning_rate": 3.026192575219701e-06, + "loss": 0.7601, + "step": 3866 + }, + { + "epoch": 0.3487396852595031, + "grad_norm": 1.4383976662851312, + "learning_rate": 3.025691085321826e-06, + "loss": 0.6671, + "step": 3867 + }, + { + "epoch": 0.34882986878297334, + "grad_norm": 2.269992065919967, + "learning_rate": 3.025189507906332e-06, + "loss": 0.7542, + "step": 3868 + }, + { + "epoch": 0.34892005230644363, + "grad_norm": 1.8981544063778533, + "learning_rate": 3.0246878430160166e-06, + "loss": 0.7901, + "step": 3869 + }, + { + "epoch": 0.34901023582991386, + "grad_norm": 2.0061051646208767, + "learning_rate": 3.024186090693684e-06, + "loss": 0.7073, + "step": 3870 + }, + { + "epoch": 0.34910041935338415, + "grad_norm": 2.366387864825721, + "learning_rate": 3.023684250982147e-06, + "loss": 0.7976, + "step": 3871 + }, + { + "epoch": 0.3491906028768544, + "grad_norm": 1.7574278007860198, + "learning_rate": 3.0231823239242252e-06, + "loss": 0.6909, + "step": 3872 + }, + { + "epoch": 0.34928078640032467, + "grad_norm": 2.2333257385618213, + "learning_rate": 3.0226803095627457e-06, + "loss": 0.7259, + "step": 3873 + }, + { + "epoch": 0.3493709699237949, + "grad_norm": 2.2151184449915533, + "learning_rate": 3.022178207940543e-06, + "loss": 0.7548, + "step": 3874 + }, + { + "epoch": 0.3494611534472652, + "grad_norm": 1.647978877047933, + "learning_rate": 3.02167601910046e-06, + "loss": 0.7806, + "step": 3875 + }, + { + "epoch": 0.3495513369707354, + "grad_norm": 1.9058930735141117, + "learning_rate": 3.021173743085345e-06, + "loss": 0.7944, + "step": 3876 + }, + { + "epoch": 0.3496415204942057, + "grad_norm": 0.6497504030191918, + "learning_rate": 3.0206713799380557e-06, + "loss": 0.5796, + "step": 3877 + }, + { + "epoch": 0.34973170401767595, + "grad_norm": 1.6975755485386257, + "learning_rate": 3.0201689297014565e-06, + "loss": 0.7782, + "step": 3878 + }, + { + "epoch": 0.34982188754114624, + "grad_norm": 1.4164843815980313, + "learning_rate": 3.0196663924184187e-06, + "loss": 0.8161, + "step": 3879 + }, + { + "epoch": 0.34991207106461647, + "grad_norm": 2.2530310665448905, + "learning_rate": 3.019163768131822e-06, + "loss": 0.8096, + "step": 3880 + }, + { + "epoch": 0.35000225458808676, + "grad_norm": 1.7288300826251966, + "learning_rate": 3.0186610568845533e-06, + "loss": 0.691, + "step": 3881 + }, + { + "epoch": 0.350092438111557, + "grad_norm": 7.858194736567825, + "learning_rate": 3.018158258719507e-06, + "loss": 0.7078, + "step": 3882 + }, + { + "epoch": 0.3501826216350273, + "grad_norm": 1.4293048117167846, + "learning_rate": 3.0176553736795827e-06, + "loss": 0.7964, + "step": 3883 + }, + { + "epoch": 0.35027280515849757, + "grad_norm": 0.5320081630184491, + "learning_rate": 3.017152401807691e-06, + "loss": 0.5174, + "step": 3884 + }, + { + "epoch": 0.3503629886819678, + "grad_norm": 3.019882280442722, + "learning_rate": 3.0166493431467476e-06, + "loss": 0.7559, + "step": 3885 + }, + { + "epoch": 0.3504531722054381, + "grad_norm": 1.5331698468721298, + "learning_rate": 3.016146197739677e-06, + "loss": 0.7442, + "step": 3886 + }, + { + "epoch": 0.3505433557289083, + "grad_norm": 2.328654286684855, + "learning_rate": 3.0156429656294097e-06, + "loss": 0.7046, + "step": 3887 + }, + { + "epoch": 0.3506335392523786, + "grad_norm": 1.933880224102865, + "learning_rate": 3.0151396468588844e-06, + "loss": 0.7107, + "step": 3888 + }, + { + "epoch": 0.35072372277584885, + "grad_norm": 1.904236464887156, + "learning_rate": 3.014636241471047e-06, + "loss": 0.724, + "step": 3889 + }, + { + "epoch": 0.35081390629931913, + "grad_norm": 3.4565398993157195, + "learning_rate": 3.0141327495088514e-06, + "loss": 0.7661, + "step": 3890 + }, + { + "epoch": 0.35090408982278937, + "grad_norm": 2.1733606320794423, + "learning_rate": 3.0136291710152566e-06, + "loss": 0.7279, + "step": 3891 + }, + { + "epoch": 0.35099427334625966, + "grad_norm": 3.063857264358891, + "learning_rate": 3.0131255060332325e-06, + "loss": 0.8063, + "step": 3892 + }, + { + "epoch": 0.3510844568697299, + "grad_norm": 1.5684805285260888, + "learning_rate": 3.012621754605754e-06, + "loss": 0.7911, + "step": 3893 + }, + { + "epoch": 0.3511746403932002, + "grad_norm": 1.7387811863376577, + "learning_rate": 3.0121179167758035e-06, + "loss": 0.791, + "step": 3894 + }, + { + "epoch": 0.3512648239166704, + "grad_norm": 2.021038739753756, + "learning_rate": 3.0116139925863717e-06, + "loss": 0.8431, + "step": 3895 + }, + { + "epoch": 0.3513550074401407, + "grad_norm": 1.6462940182952384, + "learning_rate": 3.011109982080456e-06, + "loss": 0.8078, + "step": 3896 + }, + { + "epoch": 0.35144519096361093, + "grad_norm": 1.8043077077497052, + "learning_rate": 3.0106058853010614e-06, + "loss": 0.7147, + "step": 3897 + }, + { + "epoch": 0.3515353744870812, + "grad_norm": 1.8306459179142545, + "learning_rate": 3.010101702291201e-06, + "loss": 0.7525, + "step": 3898 + }, + { + "epoch": 0.35162555801055145, + "grad_norm": 1.7068084641251426, + "learning_rate": 3.009597433093893e-06, + "loss": 0.7629, + "step": 3899 + }, + { + "epoch": 0.35171574153402174, + "grad_norm": 1.7838544710999018, + "learning_rate": 3.009093077752165e-06, + "loss": 0.7779, + "step": 3900 + }, + { + "epoch": 0.351805925057492, + "grad_norm": 1.7368450275043685, + "learning_rate": 3.008588636309052e-06, + "loss": 0.7961, + "step": 3901 + }, + { + "epoch": 0.35189610858096226, + "grad_norm": 2.0133576171033662, + "learning_rate": 3.0080841088075947e-06, + "loss": 0.7354, + "step": 3902 + }, + { + "epoch": 0.3519862921044325, + "grad_norm": 1.8627046536650083, + "learning_rate": 3.0075794952908436e-06, + "loss": 0.8148, + "step": 3903 + }, + { + "epoch": 0.3520764756279028, + "grad_norm": 2.4131942224540577, + "learning_rate": 3.0070747958018528e-06, + "loss": 0.7848, + "step": 3904 + }, + { + "epoch": 0.352166659151373, + "grad_norm": 1.4286032979680146, + "learning_rate": 3.0065700103836894e-06, + "loss": 0.7835, + "step": 3905 + }, + { + "epoch": 0.3522568426748433, + "grad_norm": 2.1722185700178, + "learning_rate": 3.0060651390794214e-06, + "loss": 0.7874, + "step": 3906 + }, + { + "epoch": 0.3523470261983136, + "grad_norm": 1.7211910599512064, + "learning_rate": 3.005560181932128e-06, + "loss": 0.749, + "step": 3907 + }, + { + "epoch": 0.35243720972178383, + "grad_norm": 1.6080483459160502, + "learning_rate": 3.005055138984896e-06, + "loss": 0.7257, + "step": 3908 + }, + { + "epoch": 0.3525273932452541, + "grad_norm": 3.2005055207066477, + "learning_rate": 3.0045500102808174e-06, + "loss": 0.7432, + "step": 3909 + }, + { + "epoch": 0.35261757676872435, + "grad_norm": 2.2276685677911794, + "learning_rate": 3.0040447958629927e-06, + "loss": 0.8517, + "step": 3910 + }, + { + "epoch": 0.35270776029219464, + "grad_norm": 1.972209555028553, + "learning_rate": 3.00353949577453e-06, + "loss": 0.7868, + "step": 3911 + }, + { + "epoch": 0.35279794381566487, + "grad_norm": 2.3357756185415885, + "learning_rate": 3.003034110058544e-06, + "loss": 0.6054, + "step": 3912 + }, + { + "epoch": 0.35288812733913516, + "grad_norm": 1.7170346109571029, + "learning_rate": 3.002528638758157e-06, + "loss": 0.7454, + "step": 3913 + }, + { + "epoch": 0.3529783108626054, + "grad_norm": 2.6747453229652742, + "learning_rate": 3.0020230819164985e-06, + "loss": 0.7591, + "step": 3914 + }, + { + "epoch": 0.3530684943860757, + "grad_norm": 1.6466845552474088, + "learning_rate": 3.0015174395767064e-06, + "loss": 0.7713, + "step": 3915 + }, + { + "epoch": 0.3531586779095459, + "grad_norm": 1.6724373415398717, + "learning_rate": 3.001011711781923e-06, + "loss": 0.7839, + "step": 3916 + }, + { + "epoch": 0.3532488614330162, + "grad_norm": 1.6084131866686753, + "learning_rate": 3.0005058985753017e-06, + "loss": 0.7214, + "step": 3917 + }, + { + "epoch": 0.35333904495648644, + "grad_norm": 1.4927057295051362, + "learning_rate": 3e-06, + "loss": 0.7543, + "step": 3918 + }, + { + "epoch": 0.3534292284799567, + "grad_norm": 0.6546599789755019, + "learning_rate": 2.9994940160991843e-06, + "loss": 0.6063, + "step": 3919 + }, + { + "epoch": 0.35351941200342696, + "grad_norm": 0.627977313049508, + "learning_rate": 2.9989879469160285e-06, + "loss": 0.5633, + "step": 3920 + }, + { + "epoch": 0.35360959552689725, + "grad_norm": 3.8476965267958563, + "learning_rate": 2.9984817924937124e-06, + "loss": 0.7833, + "step": 3921 + }, + { + "epoch": 0.3536997790503675, + "grad_norm": 4.060877784210687, + "learning_rate": 2.997975552875424e-06, + "loss": 0.7079, + "step": 3922 + }, + { + "epoch": 0.35378996257383777, + "grad_norm": 1.477511307984534, + "learning_rate": 2.997469228104358e-06, + "loss": 0.756, + "step": 3923 + }, + { + "epoch": 0.353880146097308, + "grad_norm": 1.6365554473183048, + "learning_rate": 2.996962818223718e-06, + "loss": 0.6607, + "step": 3924 + }, + { + "epoch": 0.3539703296207783, + "grad_norm": 2.3986845106733785, + "learning_rate": 2.9964563232767135e-06, + "loss": 0.7144, + "step": 3925 + }, + { + "epoch": 0.3540605131442485, + "grad_norm": 1.94739699614059, + "learning_rate": 2.9959497433065617e-06, + "loss": 0.7633, + "step": 3926 + }, + { + "epoch": 0.3541506966677188, + "grad_norm": 1.6184512742762511, + "learning_rate": 2.9954430783564848e-06, + "loss": 0.6859, + "step": 3927 + }, + { + "epoch": 0.35424088019118904, + "grad_norm": 2.2362540286919694, + "learning_rate": 2.994936328469716e-06, + "loss": 0.7949, + "step": 3928 + }, + { + "epoch": 0.35433106371465933, + "grad_norm": 1.661633505201773, + "learning_rate": 2.994429493689494e-06, + "loss": 0.7657, + "step": 3929 + }, + { + "epoch": 0.35442124723812957, + "grad_norm": 1.466035160022973, + "learning_rate": 2.9939225740590642e-06, + "loss": 0.7903, + "step": 3930 + }, + { + "epoch": 0.35451143076159986, + "grad_norm": 6.549997682061404, + "learning_rate": 2.99341556962168e-06, + "loss": 0.725, + "step": 3931 + }, + { + "epoch": 0.35460161428507014, + "grad_norm": 1.6250652153427017, + "learning_rate": 2.992908480420602e-06, + "loss": 0.8109, + "step": 3932 + }, + { + "epoch": 0.3546917978085404, + "grad_norm": 2.3144690919478244, + "learning_rate": 2.9924013064990974e-06, + "loss": 0.755, + "step": 3933 + }, + { + "epoch": 0.35478198133201067, + "grad_norm": 0.6605673418534416, + "learning_rate": 2.991894047900441e-06, + "loss": 0.6113, + "step": 3934 + }, + { + "epoch": 0.3548721648554809, + "grad_norm": 1.6974309526875742, + "learning_rate": 2.991386704667916e-06, + "loss": 0.7816, + "step": 3935 + }, + { + "epoch": 0.3549623483789512, + "grad_norm": 1.8932299362523868, + "learning_rate": 2.9908792768448097e-06, + "loss": 0.8282, + "step": 3936 + }, + { + "epoch": 0.3550525319024214, + "grad_norm": 3.519170575390782, + "learning_rate": 2.990371764474421e-06, + "loss": 0.7631, + "step": 3937 + }, + { + "epoch": 0.3551427154258917, + "grad_norm": 1.969837070253148, + "learning_rate": 2.9898641676000518e-06, + "loss": 0.7335, + "step": 3938 + }, + { + "epoch": 0.35523289894936194, + "grad_norm": 1.5754317161219924, + "learning_rate": 2.9893564862650138e-06, + "loss": 0.7654, + "step": 3939 + }, + { + "epoch": 0.35532308247283223, + "grad_norm": 4.084889381433103, + "learning_rate": 2.9888487205126254e-06, + "loss": 0.8185, + "step": 3940 + }, + { + "epoch": 0.35541326599630246, + "grad_norm": 1.46546886076003, + "learning_rate": 2.9883408703862115e-06, + "loss": 0.7605, + "step": 3941 + }, + { + "epoch": 0.35550344951977275, + "grad_norm": 1.8845475935458562, + "learning_rate": 2.987832935929105e-06, + "loss": 0.847, + "step": 3942 + }, + { + "epoch": 0.355593633043243, + "grad_norm": 1.610256287096836, + "learning_rate": 2.9873249171846454e-06, + "loss": 0.8308, + "step": 3943 + }, + { + "epoch": 0.3556838165667133, + "grad_norm": 1.8179076849661426, + "learning_rate": 2.98681681419618e-06, + "loss": 0.7597, + "step": 3944 + }, + { + "epoch": 0.3557740000901835, + "grad_norm": 1.8068652204145312, + "learning_rate": 2.9863086270070627e-06, + "loss": 0.8199, + "step": 3945 + }, + { + "epoch": 0.3558641836136538, + "grad_norm": 2.355891504512968, + "learning_rate": 2.985800355660655e-06, + "loss": 0.7732, + "step": 3946 + }, + { + "epoch": 0.35595436713712403, + "grad_norm": 0.6746701494241366, + "learning_rate": 2.9852920002003252e-06, + "loss": 0.609, + "step": 3947 + }, + { + "epoch": 0.3560445506605943, + "grad_norm": 1.880764238876888, + "learning_rate": 2.9847835606694494e-06, + "loss": 0.8056, + "step": 3948 + }, + { + "epoch": 0.35613473418406455, + "grad_norm": 2.2398529127057967, + "learning_rate": 2.98427503711141e-06, + "loss": 0.694, + "step": 3949 + }, + { + "epoch": 0.35622491770753484, + "grad_norm": 1.7408341424239222, + "learning_rate": 2.9837664295695973e-06, + "loss": 0.7994, + "step": 3950 + }, + { + "epoch": 0.35631510123100507, + "grad_norm": 2.495903951147903, + "learning_rate": 2.983257738087408e-06, + "loss": 0.7582, + "step": 3951 + }, + { + "epoch": 0.35640528475447536, + "grad_norm": 1.9626933591796678, + "learning_rate": 2.982748962708247e-06, + "loss": 0.799, + "step": 3952 + }, + { + "epoch": 0.3564954682779456, + "grad_norm": 1.5867892829548498, + "learning_rate": 2.982240103475526e-06, + "loss": 0.8173, + "step": 3953 + }, + { + "epoch": 0.3565856518014159, + "grad_norm": 2.1359972528947173, + "learning_rate": 2.981731160432663e-06, + "loss": 0.7684, + "step": 3954 + }, + { + "epoch": 0.35667583532488617, + "grad_norm": 1.568291269536216, + "learning_rate": 2.981222133623084e-06, + "loss": 0.6706, + "step": 3955 + }, + { + "epoch": 0.3567660188483564, + "grad_norm": 1.6146055240541803, + "learning_rate": 2.980713023090222e-06, + "loss": 0.7742, + "step": 3956 + }, + { + "epoch": 0.3568562023718267, + "grad_norm": 2.935092994997301, + "learning_rate": 2.980203828877518e-06, + "loss": 0.7391, + "step": 3957 + }, + { + "epoch": 0.3569463858952969, + "grad_norm": 2.1808988971740124, + "learning_rate": 2.9796945510284182e-06, + "loss": 0.8063, + "step": 3958 + }, + { + "epoch": 0.3570365694187672, + "grad_norm": 3.3446381717335125, + "learning_rate": 2.9791851895863774e-06, + "loss": 0.8226, + "step": 3959 + }, + { + "epoch": 0.35712675294223745, + "grad_norm": 2.435781019468845, + "learning_rate": 2.978675744594857e-06, + "loss": 0.7965, + "step": 3960 + }, + { + "epoch": 0.35721693646570774, + "grad_norm": 1.3551782302191067, + "learning_rate": 2.978166216097326e-06, + "loss": 0.7235, + "step": 3961 + }, + { + "epoch": 0.35730711998917797, + "grad_norm": 1.966049097641396, + "learning_rate": 2.9776566041372596e-06, + "loss": 0.7541, + "step": 3962 + }, + { + "epoch": 0.35739730351264826, + "grad_norm": 1.7623873032055821, + "learning_rate": 2.977146908758141e-06, + "loss": 0.8239, + "step": 3963 + }, + { + "epoch": 0.3574874870361185, + "grad_norm": 1.5324071287855134, + "learning_rate": 2.9766371300034604e-06, + "loss": 0.7827, + "step": 3964 + }, + { + "epoch": 0.3575776705595888, + "grad_norm": 1.6983636919002978, + "learning_rate": 2.9761272679167142e-06, + "loss": 0.8174, + "step": 3965 + }, + { + "epoch": 0.357667854083059, + "grad_norm": 1.9253630580619676, + "learning_rate": 2.9756173225414072e-06, + "loss": 0.6153, + "step": 3966 + }, + { + "epoch": 0.3577580376065293, + "grad_norm": 1.84107304869384, + "learning_rate": 2.975107293921051e-06, + "loss": 0.9427, + "step": 3967 + }, + { + "epoch": 0.35784822112999953, + "grad_norm": 2.3272802881435015, + "learning_rate": 2.9745971820991643e-06, + "loss": 0.8067, + "step": 3968 + }, + { + "epoch": 0.3579384046534698, + "grad_norm": 1.6451469954002924, + "learning_rate": 2.9740869871192715e-06, + "loss": 0.8088, + "step": 3969 + }, + { + "epoch": 0.35802858817694005, + "grad_norm": 1.9442630986011207, + "learning_rate": 2.9735767090249065e-06, + "loss": 0.8307, + "step": 3970 + }, + { + "epoch": 0.35811877170041034, + "grad_norm": 1.5280280178007877, + "learning_rate": 2.973066347859608e-06, + "loss": 0.6808, + "step": 3971 + }, + { + "epoch": 0.3582089552238806, + "grad_norm": 1.3314453357787743, + "learning_rate": 2.972555903666923e-06, + "loss": 0.7793, + "step": 3972 + }, + { + "epoch": 0.35829913874735086, + "grad_norm": 2.0201601565483536, + "learning_rate": 2.972045376490406e-06, + "loss": 0.6837, + "step": 3973 + }, + { + "epoch": 0.3583893222708211, + "grad_norm": 1.4330930745045387, + "learning_rate": 2.9715347663736177e-06, + "loss": 0.6934, + "step": 3974 + }, + { + "epoch": 0.3584795057942914, + "grad_norm": 1.5130553337361419, + "learning_rate": 2.9710240733601266e-06, + "loss": 0.767, + "step": 3975 + }, + { + "epoch": 0.3585696893177616, + "grad_norm": 1.85666356216025, + "learning_rate": 2.970513297493507e-06, + "loss": 0.8574, + "step": 3976 + }, + { + "epoch": 0.3586598728412319, + "grad_norm": 1.6794850781917214, + "learning_rate": 2.9700024388173416e-06, + "loss": 0.8193, + "step": 3977 + }, + { + "epoch": 0.35875005636470214, + "grad_norm": 2.1912324941907495, + "learning_rate": 2.969491497375219e-06, + "loss": 0.8192, + "step": 3978 + }, + { + "epoch": 0.35884023988817243, + "grad_norm": 1.4052039176281472, + "learning_rate": 2.9689804732107364e-06, + "loss": 0.7908, + "step": 3979 + }, + { + "epoch": 0.3589304234116427, + "grad_norm": 1.8955195100827236, + "learning_rate": 2.9684693663674968e-06, + "loss": 0.7244, + "step": 3980 + }, + { + "epoch": 0.35902060693511295, + "grad_norm": 1.785563683289885, + "learning_rate": 2.9679581768891115e-06, + "loss": 0.8527, + "step": 3981 + }, + { + "epoch": 0.35911079045858324, + "grad_norm": 1.6820731321259692, + "learning_rate": 2.967446904819197e-06, + "loss": 0.6903, + "step": 3982 + }, + { + "epoch": 0.3592009739820535, + "grad_norm": 1.6485573964561866, + "learning_rate": 2.966935550201378e-06, + "loss": 0.7959, + "step": 3983 + }, + { + "epoch": 0.35929115750552376, + "grad_norm": 1.5210343406607756, + "learning_rate": 2.966424113079286e-06, + "loss": 0.7671, + "step": 3984 + }, + { + "epoch": 0.359381341028994, + "grad_norm": 1.752140371125257, + "learning_rate": 2.9659125934965596e-06, + "loss": 0.8176, + "step": 3985 + }, + { + "epoch": 0.3594715245524643, + "grad_norm": 1.6694010318281733, + "learning_rate": 2.9654009914968457e-06, + "loss": 0.8301, + "step": 3986 + }, + { + "epoch": 0.3595617080759345, + "grad_norm": 2.0279423706398747, + "learning_rate": 2.9648893071237956e-06, + "loss": 0.6661, + "step": 3987 + }, + { + "epoch": 0.3596518915994048, + "grad_norm": 2.2196066908350027, + "learning_rate": 2.964377540421069e-06, + "loss": 0.7444, + "step": 3988 + }, + { + "epoch": 0.35974207512287504, + "grad_norm": 1.9155356829731809, + "learning_rate": 2.963865691432334e-06, + "loss": 0.7944, + "step": 3989 + }, + { + "epoch": 0.3598322586463453, + "grad_norm": 2.1977550453194947, + "learning_rate": 2.963353760201263e-06, + "loss": 0.6891, + "step": 3990 + }, + { + "epoch": 0.35992244216981556, + "grad_norm": 1.826532110087057, + "learning_rate": 2.962841746771537e-06, + "loss": 0.7559, + "step": 3991 + }, + { + "epoch": 0.36001262569328585, + "grad_norm": 1.9525157130704447, + "learning_rate": 2.9623296511868445e-06, + "loss": 0.7708, + "step": 3992 + }, + { + "epoch": 0.3601028092167561, + "grad_norm": 1.4802049386694485, + "learning_rate": 2.96181747349088e-06, + "loss": 0.6972, + "step": 3993 + }, + { + "epoch": 0.36019299274022637, + "grad_norm": 1.6011618508487908, + "learning_rate": 2.961305213727345e-06, + "loss": 0.7906, + "step": 3994 + }, + { + "epoch": 0.3602831762636966, + "grad_norm": 0.6503728686592332, + "learning_rate": 2.960792871939949e-06, + "loss": 0.6033, + "step": 3995 + }, + { + "epoch": 0.3603733597871669, + "grad_norm": 2.0153503230812384, + "learning_rate": 2.9602804481724064e-06, + "loss": 0.8643, + "step": 3996 + }, + { + "epoch": 0.3604635433106371, + "grad_norm": 1.7750468073301309, + "learning_rate": 2.9597679424684427e-06, + "loss": 0.7681, + "step": 3997 + }, + { + "epoch": 0.3605537268341074, + "grad_norm": 1.8010450323046567, + "learning_rate": 2.9592553548717848e-06, + "loss": 0.7813, + "step": 3998 + }, + { + "epoch": 0.36064391035757765, + "grad_norm": 2.6694967828657274, + "learning_rate": 2.958742685426171e-06, + "loss": 0.8547, + "step": 3999 + }, + { + "epoch": 0.36073409388104793, + "grad_norm": 2.0655965967577283, + "learning_rate": 2.9582299341753446e-06, + "loss": 0.841, + "step": 4000 + }, + { + "epoch": 0.36082427740451817, + "grad_norm": 11.948538744340247, + "learning_rate": 2.957717101163057e-06, + "loss": 0.7101, + "step": 4001 + }, + { + "epoch": 0.36091446092798846, + "grad_norm": 1.9704792360549022, + "learning_rate": 2.9572041864330655e-06, + "loss": 0.7696, + "step": 4002 + }, + { + "epoch": 0.36100464445145874, + "grad_norm": 1.6453129231306585, + "learning_rate": 2.9566911900291346e-06, + "loss": 0.7922, + "step": 4003 + }, + { + "epoch": 0.361094827974929, + "grad_norm": 1.6419427574996759, + "learning_rate": 2.9561781119950368e-06, + "loss": 0.6865, + "step": 4004 + }, + { + "epoch": 0.36118501149839927, + "grad_norm": 1.5033006910401807, + "learning_rate": 2.9556649523745493e-06, + "loss": 0.7158, + "step": 4005 + }, + { + "epoch": 0.3612751950218695, + "grad_norm": 1.8998835201698583, + "learning_rate": 2.955151711211459e-06, + "loss": 0.832, + "step": 4006 + }, + { + "epoch": 0.3613653785453398, + "grad_norm": 1.4903430024199331, + "learning_rate": 2.9546383885495583e-06, + "loss": 0.7396, + "step": 4007 + }, + { + "epoch": 0.36145556206881, + "grad_norm": 1.9815419917541084, + "learning_rate": 2.9541249844326464e-06, + "loss": 0.7222, + "step": 4008 + }, + { + "epoch": 0.3615457455922803, + "grad_norm": 2.076301290761606, + "learning_rate": 2.9536114989045295e-06, + "loss": 0.7528, + "step": 4009 + }, + { + "epoch": 0.36163592911575054, + "grad_norm": 2.0499236720692346, + "learning_rate": 2.9530979320090216e-06, + "loss": 0.8318, + "step": 4010 + }, + { + "epoch": 0.36172611263922083, + "grad_norm": 2.537378592006156, + "learning_rate": 2.9525842837899422e-06, + "loss": 0.8526, + "step": 4011 + }, + { + "epoch": 0.36181629616269106, + "grad_norm": 1.6071109586910675, + "learning_rate": 2.95207055429112e-06, + "loss": 0.8529, + "step": 4012 + }, + { + "epoch": 0.36190647968616135, + "grad_norm": 1.578191548067974, + "learning_rate": 2.951556743556388e-06, + "loss": 0.764, + "step": 4013 + }, + { + "epoch": 0.3619966632096316, + "grad_norm": 1.914772493677817, + "learning_rate": 2.951042851629588e-06, + "loss": 0.8126, + "step": 4014 + }, + { + "epoch": 0.3620868467331019, + "grad_norm": 1.7851637291282056, + "learning_rate": 2.950528878554568e-06, + "loss": 0.7292, + "step": 4015 + }, + { + "epoch": 0.3621770302565721, + "grad_norm": 1.425498453600368, + "learning_rate": 2.950014824375183e-06, + "loss": 0.8058, + "step": 4016 + }, + { + "epoch": 0.3622672137800424, + "grad_norm": 0.6324994525605497, + "learning_rate": 2.949500689135295e-06, + "loss": 0.5278, + "step": 4017 + }, + { + "epoch": 0.36235739730351263, + "grad_norm": 1.5589980745872563, + "learning_rate": 2.9489864728787722e-06, + "loss": 0.8525, + "step": 4018 + }, + { + "epoch": 0.3624475808269829, + "grad_norm": 1.6030595435151282, + "learning_rate": 2.9484721756494915e-06, + "loss": 0.8314, + "step": 4019 + }, + { + "epoch": 0.36253776435045315, + "grad_norm": 1.6318895569717837, + "learning_rate": 2.9479577974913343e-06, + "loss": 0.7735, + "step": 4020 + }, + { + "epoch": 0.36262794787392344, + "grad_norm": 1.5975620829617971, + "learning_rate": 2.9474433384481908e-06, + "loss": 0.7689, + "step": 4021 + }, + { + "epoch": 0.3627181313973937, + "grad_norm": 2.030786549443199, + "learning_rate": 2.9469287985639577e-06, + "loss": 0.7635, + "step": 4022 + }, + { + "epoch": 0.36280831492086396, + "grad_norm": 1.5400013649922055, + "learning_rate": 2.9464141778825384e-06, + "loss": 0.8154, + "step": 4023 + }, + { + "epoch": 0.3628984984443342, + "grad_norm": 1.6182229233363388, + "learning_rate": 2.9458994764478427e-06, + "loss": 0.8081, + "step": 4024 + }, + { + "epoch": 0.3629886819678045, + "grad_norm": 1.657404083652743, + "learning_rate": 2.9453846943037883e-06, + "loss": 0.8099, + "step": 4025 + }, + { + "epoch": 0.36307886549127477, + "grad_norm": 1.4354426547255357, + "learning_rate": 2.9448698314942987e-06, + "loss": 0.7192, + "step": 4026 + }, + { + "epoch": 0.363169049014745, + "grad_norm": 0.7419732838728211, + "learning_rate": 2.944354888063305e-06, + "loss": 0.6234, + "step": 4027 + }, + { + "epoch": 0.3632592325382153, + "grad_norm": 1.6448431376190364, + "learning_rate": 2.9438398640547453e-06, + "loss": 0.748, + "step": 4028 + }, + { + "epoch": 0.3633494160616855, + "grad_norm": 1.7061061840441, + "learning_rate": 2.943324759512564e-06, + "loss": 0.7137, + "step": 4029 + }, + { + "epoch": 0.3634395995851558, + "grad_norm": 1.4059612825497256, + "learning_rate": 2.9428095744807134e-06, + "loss": 0.7941, + "step": 4030 + }, + { + "epoch": 0.36352978310862605, + "grad_norm": 1.49343851584739, + "learning_rate": 2.942294309003151e-06, + "loss": 0.7283, + "step": 4031 + }, + { + "epoch": 0.36361996663209634, + "grad_norm": 1.6637193209530126, + "learning_rate": 2.941778963123843e-06, + "loss": 0.7893, + "step": 4032 + }, + { + "epoch": 0.36371015015556657, + "grad_norm": 5.240623456833031, + "learning_rate": 2.94126353688676e-06, + "loss": 0.7739, + "step": 4033 + }, + { + "epoch": 0.36380033367903686, + "grad_norm": 1.8169746412020285, + "learning_rate": 2.9407480303358825e-06, + "loss": 0.7221, + "step": 4034 + }, + { + "epoch": 0.3638905172025071, + "grad_norm": 1.5780199979087872, + "learning_rate": 2.940232443515195e-06, + "loss": 0.7774, + "step": 4035 + }, + { + "epoch": 0.3639807007259774, + "grad_norm": 1.5054253374461173, + "learning_rate": 2.9397167764686916e-06, + "loss": 0.7959, + "step": 4036 + }, + { + "epoch": 0.3640708842494476, + "grad_norm": 1.9557974270533067, + "learning_rate": 2.9392010292403714e-06, + "loss": 0.7444, + "step": 4037 + }, + { + "epoch": 0.3641610677729179, + "grad_norm": 4.356988506576021, + "learning_rate": 2.9386852018742404e-06, + "loss": 0.7659, + "step": 4038 + }, + { + "epoch": 0.36425125129638813, + "grad_norm": 1.4545014947783237, + "learning_rate": 2.938169294414312e-06, + "loss": 0.7469, + "step": 4039 + }, + { + "epoch": 0.3643414348198584, + "grad_norm": 1.4869421994969325, + "learning_rate": 2.9376533069046067e-06, + "loss": 0.8012, + "step": 4040 + }, + { + "epoch": 0.36443161834332866, + "grad_norm": 0.7376508089821185, + "learning_rate": 2.9371372393891514e-06, + "loss": 0.6353, + "step": 4041 + }, + { + "epoch": 0.36452180186679894, + "grad_norm": 1.3862600508137193, + "learning_rate": 2.936621091911979e-06, + "loss": 0.8417, + "step": 4042 + }, + { + "epoch": 0.3646119853902692, + "grad_norm": 1.9430699425250213, + "learning_rate": 2.936104864517131e-06, + "loss": 0.7018, + "step": 4043 + }, + { + "epoch": 0.36470216891373947, + "grad_norm": 2.7099475647192572, + "learning_rate": 2.9355885572486535e-06, + "loss": 0.7287, + "step": 4044 + }, + { + "epoch": 0.3647923524372097, + "grad_norm": 1.6527388440772492, + "learning_rate": 2.9350721701506026e-06, + "loss": 0.8004, + "step": 4045 + }, + { + "epoch": 0.36488253596068, + "grad_norm": 2.0918257112055945, + "learning_rate": 2.9345557032670375e-06, + "loss": 0.7682, + "step": 4046 + }, + { + "epoch": 0.3649727194841502, + "grad_norm": 2.034698493422788, + "learning_rate": 2.934039156642027e-06, + "loss": 0.8131, + "step": 4047 + }, + { + "epoch": 0.3650629030076205, + "grad_norm": 1.9719488405164718, + "learning_rate": 2.9335225303196454e-06, + "loss": 0.6302, + "step": 4048 + }, + { + "epoch": 0.36515308653109074, + "grad_norm": 1.73739800462561, + "learning_rate": 2.933005824343974e-06, + "loss": 0.7365, + "step": 4049 + }, + { + "epoch": 0.36524327005456103, + "grad_norm": 8.465988328247532, + "learning_rate": 2.932489038759101e-06, + "loss": 0.7396, + "step": 4050 + }, + { + "epoch": 0.3653334535780313, + "grad_norm": 1.4402457840759626, + "learning_rate": 2.9319721736091215e-06, + "loss": 0.6643, + "step": 4051 + }, + { + "epoch": 0.36542363710150155, + "grad_norm": 1.3338080494912201, + "learning_rate": 2.9314552289381377e-06, + "loss": 0.6738, + "step": 4052 + }, + { + "epoch": 0.36551382062497184, + "grad_norm": 2.019743976420191, + "learning_rate": 2.9309382047902574e-06, + "loss": 0.7394, + "step": 4053 + }, + { + "epoch": 0.3656040041484421, + "grad_norm": 1.8188759042479261, + "learning_rate": 2.9304211012095963e-06, + "loss": 0.8135, + "step": 4054 + }, + { + "epoch": 0.36569418767191236, + "grad_norm": 1.8407210346032434, + "learning_rate": 2.929903918240277e-06, + "loss": 0.8283, + "step": 4055 + }, + { + "epoch": 0.3657843711953826, + "grad_norm": 0.6444680616442345, + "learning_rate": 2.9293866559264273e-06, + "loss": 0.5794, + "step": 4056 + }, + { + "epoch": 0.3658745547188529, + "grad_norm": 1.879552844925049, + "learning_rate": 2.928869314312184e-06, + "loss": 0.7735, + "step": 4057 + }, + { + "epoch": 0.3659647382423231, + "grad_norm": 2.2506012643008986, + "learning_rate": 2.9283518934416892e-06, + "loss": 0.7701, + "step": 4058 + }, + { + "epoch": 0.3660549217657934, + "grad_norm": 1.6853172879652372, + "learning_rate": 2.927834393359092e-06, + "loss": 0.7829, + "step": 4059 + }, + { + "epoch": 0.36614510528926364, + "grad_norm": 1.616856195541758, + "learning_rate": 2.927316814108548e-06, + "loss": 0.8448, + "step": 4060 + }, + { + "epoch": 0.36623528881273393, + "grad_norm": 1.8248595066077857, + "learning_rate": 2.92679915573422e-06, + "loss": 0.6783, + "step": 4061 + }, + { + "epoch": 0.36632547233620416, + "grad_norm": 1.571788037664576, + "learning_rate": 2.926281418280278e-06, + "loss": 0.7147, + "step": 4062 + }, + { + "epoch": 0.36641565585967445, + "grad_norm": 1.7690758529668578, + "learning_rate": 2.925763601790899e-06, + "loss": 0.7775, + "step": 4063 + }, + { + "epoch": 0.3665058393831447, + "grad_norm": 1.7283629269152385, + "learning_rate": 2.9252457063102635e-06, + "loss": 0.7156, + "step": 4064 + }, + { + "epoch": 0.36659602290661497, + "grad_norm": 1.7229720537661317, + "learning_rate": 2.9247277318825626e-06, + "loss": 0.7294, + "step": 4065 + }, + { + "epoch": 0.3666862064300852, + "grad_norm": 1.4885087928140666, + "learning_rate": 2.924209678551993e-06, + "loss": 0.8652, + "step": 4066 + }, + { + "epoch": 0.3667763899535555, + "grad_norm": 1.988481612686655, + "learning_rate": 2.923691546362757e-06, + "loss": 0.7803, + "step": 4067 + }, + { + "epoch": 0.3668665734770257, + "grad_norm": 1.6185945296283357, + "learning_rate": 2.9231733353590663e-06, + "loss": 0.7283, + "step": 4068 + }, + { + "epoch": 0.366956757000496, + "grad_norm": 1.7914374966397977, + "learning_rate": 2.922655045585136e-06, + "loss": 0.8466, + "step": 4069 + }, + { + "epoch": 0.36704694052396625, + "grad_norm": 0.7030483415988716, + "learning_rate": 2.92213667708519e-06, + "loss": 0.621, + "step": 4070 + }, + { + "epoch": 0.36713712404743654, + "grad_norm": 1.635029214370524, + "learning_rate": 2.921618229903457e-06, + "loss": 0.7795, + "step": 4071 + }, + { + "epoch": 0.36722730757090677, + "grad_norm": 1.553435468440242, + "learning_rate": 2.9210997040841752e-06, + "loss": 0.692, + "step": 4072 + }, + { + "epoch": 0.36731749109437706, + "grad_norm": 1.927821504272269, + "learning_rate": 2.9205810996715885e-06, + "loss": 0.7316, + "step": 4073 + }, + { + "epoch": 0.36740767461784735, + "grad_norm": 2.811871267614285, + "learning_rate": 2.9200624167099456e-06, + "loss": 0.729, + "step": 4074 + }, + { + "epoch": 0.3674978581413176, + "grad_norm": 1.6278776951149265, + "learning_rate": 2.919543655243505e-06, + "loss": 0.6825, + "step": 4075 + }, + { + "epoch": 0.36758804166478787, + "grad_norm": 1.793242190584531, + "learning_rate": 2.919024815316529e-06, + "loss": 0.7962, + "step": 4076 + }, + { + "epoch": 0.3676782251882581, + "grad_norm": 1.8231082125346278, + "learning_rate": 2.9185058969732877e-06, + "loss": 0.8945, + "step": 4077 + }, + { + "epoch": 0.3677684087117284, + "grad_norm": 1.5032664337554114, + "learning_rate": 2.917986900258059e-06, + "loss": 0.8024, + "step": 4078 + }, + { + "epoch": 0.3678585922351986, + "grad_norm": 1.7894364678274406, + "learning_rate": 2.917467825215126e-06, + "loss": 0.7665, + "step": 4079 + }, + { + "epoch": 0.3679487757586689, + "grad_norm": 1.5387922330463475, + "learning_rate": 2.9169486718887803e-06, + "loss": 0.7176, + "step": 4080 + }, + { + "epoch": 0.36803895928213914, + "grad_norm": 1.4223526372488366, + "learning_rate": 2.9164294403233173e-06, + "loss": 0.818, + "step": 4081 + }, + { + "epoch": 0.36812914280560943, + "grad_norm": 1.6605292406135146, + "learning_rate": 2.915910130563041e-06, + "loss": 0.7108, + "step": 4082 + }, + { + "epoch": 0.36821932632907967, + "grad_norm": 2.2682650213258184, + "learning_rate": 2.915390742652262e-06, + "loss": 0.7172, + "step": 4083 + }, + { + "epoch": 0.36830950985254995, + "grad_norm": 2.1513781588391945, + "learning_rate": 2.914871276635298e-06, + "loss": 0.7042, + "step": 4084 + }, + { + "epoch": 0.3683996933760202, + "grad_norm": 2.34829139686467, + "learning_rate": 2.914351732556472e-06, + "loss": 0.7483, + "step": 4085 + }, + { + "epoch": 0.3684898768994905, + "grad_norm": 1.3756190553681436, + "learning_rate": 2.9138321104601144e-06, + "loss": 0.6927, + "step": 4086 + }, + { + "epoch": 0.3685800604229607, + "grad_norm": 1.6621619461657073, + "learning_rate": 2.9133124103905623e-06, + "loss": 0.7015, + "step": 4087 + }, + { + "epoch": 0.368670243946431, + "grad_norm": 2.8019382954514422, + "learning_rate": 2.9127926323921596e-06, + "loss": 0.7599, + "step": 4088 + }, + { + "epoch": 0.36876042746990123, + "grad_norm": 1.8983273572328103, + "learning_rate": 2.912272776509256e-06, + "loss": 0.7884, + "step": 4089 + }, + { + "epoch": 0.3688506109933715, + "grad_norm": 1.6817592175853318, + "learning_rate": 2.911752842786209e-06, + "loss": 0.7364, + "step": 4090 + }, + { + "epoch": 0.36894079451684175, + "grad_norm": 1.7932776224542224, + "learning_rate": 2.911232831267383e-06, + "loss": 0.7146, + "step": 4091 + }, + { + "epoch": 0.36903097804031204, + "grad_norm": 1.7083861955356945, + "learning_rate": 2.910712741997146e-06, + "loss": 0.7902, + "step": 4092 + }, + { + "epoch": 0.3691211615637823, + "grad_norm": 2.525053280560953, + "learning_rate": 2.910192575019877e-06, + "loss": 0.7579, + "step": 4093 + }, + { + "epoch": 0.36921134508725256, + "grad_norm": 1.4632564642375774, + "learning_rate": 2.9096723303799583e-06, + "loss": 0.7967, + "step": 4094 + }, + { + "epoch": 0.3693015286107228, + "grad_norm": 2.228378786504619, + "learning_rate": 2.9091520081217805e-06, + "loss": 0.776, + "step": 4095 + }, + { + "epoch": 0.3693917121341931, + "grad_norm": 1.754929887398134, + "learning_rate": 2.908631608289741e-06, + "loss": 0.7148, + "step": 4096 + }, + { + "epoch": 0.3694818956576633, + "grad_norm": 1.8258764670619372, + "learning_rate": 2.9081111309282423e-06, + "loss": 0.7599, + "step": 4097 + }, + { + "epoch": 0.3695720791811336, + "grad_norm": 0.6421856362717215, + "learning_rate": 2.9075905760816942e-06, + "loss": 0.6326, + "step": 4098 + }, + { + "epoch": 0.3696622627046039, + "grad_norm": 3.1635369511932274, + "learning_rate": 2.907069943794514e-06, + "loss": 0.8068, + "step": 4099 + }, + { + "epoch": 0.3697524462280741, + "grad_norm": 1.6090270471806742, + "learning_rate": 2.906549234111125e-06, + "loss": 0.736, + "step": 4100 + }, + { + "epoch": 0.3698426297515444, + "grad_norm": 2.472224434364653, + "learning_rate": 2.906028447075956e-06, + "loss": 0.8226, + "step": 4101 + }, + { + "epoch": 0.36993281327501465, + "grad_norm": 1.5311949583787738, + "learning_rate": 2.905507582733445e-06, + "loss": 0.8446, + "step": 4102 + }, + { + "epoch": 0.37002299679848494, + "grad_norm": 1.656350834494106, + "learning_rate": 2.904986641128033e-06, + "loss": 0.8813, + "step": 4103 + }, + { + "epoch": 0.37011318032195517, + "grad_norm": 1.3892296637927066, + "learning_rate": 2.9044656223041716e-06, + "loss": 0.7547, + "step": 4104 + }, + { + "epoch": 0.37020336384542546, + "grad_norm": 1.3917113357882662, + "learning_rate": 2.9039445263063157e-06, + "loss": 0.7225, + "step": 4105 + }, + { + "epoch": 0.3702935473688957, + "grad_norm": 2.2137766156139405, + "learning_rate": 2.903423353178929e-06, + "loss": 0.6683, + "step": 4106 + }, + { + "epoch": 0.370383730892366, + "grad_norm": 1.4883747221662205, + "learning_rate": 2.9029021029664802e-06, + "loss": 0.7647, + "step": 4107 + }, + { + "epoch": 0.3704739144158362, + "grad_norm": 1.520053018849917, + "learning_rate": 2.9023807757134455e-06, + "loss": 0.8207, + "step": 4108 + }, + { + "epoch": 0.3705640979393065, + "grad_norm": 1.7089450330293035, + "learning_rate": 2.901859371464307e-06, + "loss": 0.8294, + "step": 4109 + }, + { + "epoch": 0.37065428146277674, + "grad_norm": 2.7068366879429746, + "learning_rate": 2.9013378902635535e-06, + "loss": 0.8077, + "step": 4110 + }, + { + "epoch": 0.370744464986247, + "grad_norm": 2.4532744732651333, + "learning_rate": 2.9008163321556823e-06, + "loss": 0.738, + "step": 4111 + }, + { + "epoch": 0.37083464850971726, + "grad_norm": 1.689023636927516, + "learning_rate": 2.900294697185194e-06, + "loss": 0.7783, + "step": 4112 + }, + { + "epoch": 0.37092483203318755, + "grad_norm": 1.5166680591003956, + "learning_rate": 2.899772985396599e-06, + "loss": 0.7841, + "step": 4113 + }, + { + "epoch": 0.3710150155566578, + "grad_norm": 1.6485064162180667, + "learning_rate": 2.8992511968344104e-06, + "loss": 0.7584, + "step": 4114 + }, + { + "epoch": 0.37110519908012807, + "grad_norm": 1.6433964889466577, + "learning_rate": 2.8987293315431523e-06, + "loss": 0.712, + "step": 4115 + }, + { + "epoch": 0.3711953826035983, + "grad_norm": 1.4399775011901108, + "learning_rate": 2.898207389567351e-06, + "loss": 0.6962, + "step": 4116 + }, + { + "epoch": 0.3712855661270686, + "grad_norm": 2.2927634880044363, + "learning_rate": 2.897685370951543e-06, + "loss": 0.7578, + "step": 4117 + }, + { + "epoch": 0.3713757496505388, + "grad_norm": 1.7927712059485035, + "learning_rate": 2.89716327574027e-06, + "loss": 0.6577, + "step": 4118 + }, + { + "epoch": 0.3714659331740091, + "grad_norm": 1.6265611485485725, + "learning_rate": 2.8966411039780787e-06, + "loss": 0.8515, + "step": 4119 + }, + { + "epoch": 0.37155611669747934, + "grad_norm": 1.7536791546107526, + "learning_rate": 2.8961188557095248e-06, + "loss": 0.7713, + "step": 4120 + }, + { + "epoch": 0.37164630022094963, + "grad_norm": 1.8217995263198201, + "learning_rate": 2.895596530979168e-06, + "loss": 0.7123, + "step": 4121 + }, + { + "epoch": 0.3717364837444199, + "grad_norm": 1.6961436541510768, + "learning_rate": 2.895074129831578e-06, + "loss": 0.7791, + "step": 4122 + }, + { + "epoch": 0.37182666726789015, + "grad_norm": 1.7313967040113531, + "learning_rate": 2.8945516523113275e-06, + "loss": 0.6792, + "step": 4123 + }, + { + "epoch": 0.37191685079136044, + "grad_norm": 0.6769461101470094, + "learning_rate": 2.894029098462998e-06, + "loss": 0.5517, + "step": 4124 + }, + { + "epoch": 0.3720070343148307, + "grad_norm": 1.5455857886721345, + "learning_rate": 2.8935064683311756e-06, + "loss": 0.7928, + "step": 4125 + }, + { + "epoch": 0.37209721783830096, + "grad_norm": 0.6776881133459085, + "learning_rate": 2.8929837619604544e-06, + "loss": 0.6562, + "step": 4126 + }, + { + "epoch": 0.3721874013617712, + "grad_norm": 1.9586536740667817, + "learning_rate": 2.8924609793954346e-06, + "loss": 0.8142, + "step": 4127 + }, + { + "epoch": 0.3722775848852415, + "grad_norm": 3.104158405745265, + "learning_rate": 2.891938120680724e-06, + "loss": 0.7223, + "step": 4128 + }, + { + "epoch": 0.3723677684087117, + "grad_norm": 1.5067713013468162, + "learning_rate": 2.8914151858609343e-06, + "loss": 0.7574, + "step": 4129 + }, + { + "epoch": 0.372457951932182, + "grad_norm": 1.8360003521488824, + "learning_rate": 2.8908921749806858e-06, + "loss": 0.7893, + "step": 4130 + }, + { + "epoch": 0.37254813545565224, + "grad_norm": 3.809888852317952, + "learning_rate": 2.890369088084605e-06, + "loss": 0.7778, + "step": 4131 + }, + { + "epoch": 0.37263831897912253, + "grad_norm": 1.6562371163496912, + "learning_rate": 2.889845925217323e-06, + "loss": 0.7907, + "step": 4132 + }, + { + "epoch": 0.37272850250259276, + "grad_norm": 1.718620156672184, + "learning_rate": 2.8893226864234813e-06, + "loss": 0.7772, + "step": 4133 + }, + { + "epoch": 0.37281868602606305, + "grad_norm": 1.490451941850792, + "learning_rate": 2.8887993717477236e-06, + "loss": 0.7777, + "step": 4134 + }, + { + "epoch": 0.3729088695495333, + "grad_norm": 1.7773107504663175, + "learning_rate": 2.8882759812347035e-06, + "loss": 0.6613, + "step": 4135 + }, + { + "epoch": 0.3729990530730036, + "grad_norm": 1.5633708053170248, + "learning_rate": 2.887752514929078e-06, + "loss": 0.7743, + "step": 4136 + }, + { + "epoch": 0.3730892365964738, + "grad_norm": 1.7622248901578428, + "learning_rate": 2.887228972875513e-06, + "loss": 0.7246, + "step": 4137 + }, + { + "epoch": 0.3731794201199441, + "grad_norm": 1.8167115313854985, + "learning_rate": 2.88670535511868e-06, + "loss": 0.7681, + "step": 4138 + }, + { + "epoch": 0.3732696036434143, + "grad_norm": 1.5812731775085413, + "learning_rate": 2.886181661703257e-06, + "loss": 0.7133, + "step": 4139 + }, + { + "epoch": 0.3733597871668846, + "grad_norm": 1.735465179656747, + "learning_rate": 2.8856578926739285e-06, + "loss": 0.7336, + "step": 4140 + }, + { + "epoch": 0.37344997069035485, + "grad_norm": 1.672605479724688, + "learning_rate": 2.8851340480753846e-06, + "loss": 0.7236, + "step": 4141 + }, + { + "epoch": 0.37354015421382514, + "grad_norm": 2.254052116130021, + "learning_rate": 2.8846101279523232e-06, + "loss": 0.7711, + "step": 4142 + }, + { + "epoch": 0.37363033773729537, + "grad_norm": 1.641953357442954, + "learning_rate": 2.8840861323494487e-06, + "loss": 0.7468, + "step": 4143 + }, + { + "epoch": 0.37372052126076566, + "grad_norm": 2.164651398113332, + "learning_rate": 2.88356206131147e-06, + "loss": 0.8307, + "step": 4144 + }, + { + "epoch": 0.37381070478423595, + "grad_norm": 1.5819537862014952, + "learning_rate": 2.883037914883104e-06, + "loss": 0.746, + "step": 4145 + }, + { + "epoch": 0.3739008883077062, + "grad_norm": 2.5144903811067696, + "learning_rate": 2.882513693109075e-06, + "loss": 0.7382, + "step": 4146 + }, + { + "epoch": 0.37399107183117647, + "grad_norm": 1.3859164643868493, + "learning_rate": 2.8819893960341106e-06, + "loss": 0.7451, + "step": 4147 + }, + { + "epoch": 0.3740812553546467, + "grad_norm": 1.8190585553064715, + "learning_rate": 2.881465023702948e-06, + "loss": 0.8207, + "step": 4148 + }, + { + "epoch": 0.374171438878117, + "grad_norm": 1.6837726683814689, + "learning_rate": 2.8809405761603294e-06, + "loss": 0.8521, + "step": 4149 + }, + { + "epoch": 0.3742616224015872, + "grad_norm": 1.9720778419885105, + "learning_rate": 2.880416053451003e-06, + "loss": 0.6934, + "step": 4150 + }, + { + "epoch": 0.3743518059250575, + "grad_norm": 1.77623424063986, + "learning_rate": 2.879891455619725e-06, + "loss": 0.8562, + "step": 4151 + }, + { + "epoch": 0.37444198944852775, + "grad_norm": 1.5378075920015808, + "learning_rate": 2.879366782711256e-06, + "loss": 0.7709, + "step": 4152 + }, + { + "epoch": 0.37453217297199803, + "grad_norm": 1.5777280704894618, + "learning_rate": 2.8788420347703643e-06, + "loss": 0.7882, + "step": 4153 + }, + { + "epoch": 0.37462235649546827, + "grad_norm": 0.6380562804878637, + "learning_rate": 2.8783172118418244e-06, + "loss": 0.5779, + "step": 4154 + }, + { + "epoch": 0.37471254001893856, + "grad_norm": 1.44688725641386, + "learning_rate": 2.877792313970417e-06, + "loss": 0.7732, + "step": 4155 + }, + { + "epoch": 0.3748027235424088, + "grad_norm": 1.6286633353842728, + "learning_rate": 2.8772673412009293e-06, + "loss": 0.8208, + "step": 4156 + }, + { + "epoch": 0.3748929070658791, + "grad_norm": 1.6618659337640387, + "learning_rate": 2.8767422935781545e-06, + "loss": 0.8186, + "step": 4157 + }, + { + "epoch": 0.3749830905893493, + "grad_norm": 1.5312690733675385, + "learning_rate": 2.8762171711468935e-06, + "loss": 0.8047, + "step": 4158 + }, + { + "epoch": 0.3750732741128196, + "grad_norm": 1.4987730730458513, + "learning_rate": 2.875691973951952e-06, + "loss": 0.7708, + "step": 4159 + }, + { + "epoch": 0.37516345763628983, + "grad_norm": 0.6468324674914897, + "learning_rate": 2.8751667020381425e-06, + "loss": 0.6479, + "step": 4160 + }, + { + "epoch": 0.3752536411597601, + "grad_norm": 1.9726966778228936, + "learning_rate": 2.8746413554502837e-06, + "loss": 0.8336, + "step": 4161 + }, + { + "epoch": 0.37534382468323035, + "grad_norm": 1.5771323377419355, + "learning_rate": 2.8741159342332027e-06, + "loss": 0.7052, + "step": 4162 + }, + { + "epoch": 0.37543400820670064, + "grad_norm": 2.4578755054776975, + "learning_rate": 2.87359043843173e-06, + "loss": 0.8311, + "step": 4163 + }, + { + "epoch": 0.3755241917301709, + "grad_norm": 1.8333487582280314, + "learning_rate": 2.873064868090704e-06, + "loss": 0.8093, + "step": 4164 + }, + { + "epoch": 0.37561437525364116, + "grad_norm": 1.9692232683301396, + "learning_rate": 2.8725392232549697e-06, + "loss": 0.8643, + "step": 4165 + }, + { + "epoch": 0.3757045587771114, + "grad_norm": 1.7436771610046058, + "learning_rate": 2.872013503969378e-06, + "loss": 0.8387, + "step": 4166 + }, + { + "epoch": 0.3757947423005817, + "grad_norm": 1.4022587402419573, + "learning_rate": 2.8714877102787853e-06, + "loss": 0.7772, + "step": 4167 + }, + { + "epoch": 0.3758849258240519, + "grad_norm": 1.9304901183615717, + "learning_rate": 2.8709618422280564e-06, + "loss": 0.7767, + "step": 4168 + }, + { + "epoch": 0.3759751093475222, + "grad_norm": 2.0176653330771357, + "learning_rate": 2.8704358998620605e-06, + "loss": 0.7476, + "step": 4169 + }, + { + "epoch": 0.3760652928709925, + "grad_norm": 1.8103548048868836, + "learning_rate": 2.8699098832256735e-06, + "loss": 0.8028, + "step": 4170 + }, + { + "epoch": 0.37615547639446273, + "grad_norm": 1.9034847151330738, + "learning_rate": 2.86938379236378e-06, + "loss": 0.6898, + "step": 4171 + }, + { + "epoch": 0.376245659917933, + "grad_norm": 2.0398934377625606, + "learning_rate": 2.868857627321266e-06, + "loss": 0.6998, + "step": 4172 + }, + { + "epoch": 0.37633584344140325, + "grad_norm": 2.2585958949909823, + "learning_rate": 2.8683313881430296e-06, + "loss": 0.7725, + "step": 4173 + }, + { + "epoch": 0.37642602696487354, + "grad_norm": 2.9516145587488163, + "learning_rate": 2.8678050748739706e-06, + "loss": 0.7666, + "step": 4174 + }, + { + "epoch": 0.37651621048834377, + "grad_norm": 1.5221134790080224, + "learning_rate": 2.8672786875589976e-06, + "loss": 0.7418, + "step": 4175 + }, + { + "epoch": 0.37660639401181406, + "grad_norm": 1.5287448086342428, + "learning_rate": 2.866752226243025e-06, + "loss": 0.7216, + "step": 4176 + }, + { + "epoch": 0.3766965775352843, + "grad_norm": 1.6416680789431517, + "learning_rate": 2.8662256909709733e-06, + "loss": 0.8154, + "step": 4177 + }, + { + "epoch": 0.3767867610587546, + "grad_norm": 1.8163080004300338, + "learning_rate": 2.865699081787769e-06, + "loss": 0.811, + "step": 4178 + }, + { + "epoch": 0.3768769445822248, + "grad_norm": 2.061799710770366, + "learning_rate": 2.8651723987383465e-06, + "loss": 0.7494, + "step": 4179 + }, + { + "epoch": 0.3769671281056951, + "grad_norm": 1.7793432369954663, + "learning_rate": 2.8646456418676437e-06, + "loss": 0.7477, + "step": 4180 + }, + { + "epoch": 0.37705731162916534, + "grad_norm": 0.6567900116127285, + "learning_rate": 2.8641188112206067e-06, + "loss": 0.5963, + "step": 4181 + }, + { + "epoch": 0.3771474951526356, + "grad_norm": 2.3858713996489502, + "learning_rate": 2.863591906842189e-06, + "loss": 0.8026, + "step": 4182 + }, + { + "epoch": 0.37723767867610586, + "grad_norm": 1.5025593857287642, + "learning_rate": 2.863064928777347e-06, + "loss": 0.8051, + "step": 4183 + }, + { + "epoch": 0.37732786219957615, + "grad_norm": 1.9638821806636462, + "learning_rate": 2.862537877071047e-06, + "loss": 0.7451, + "step": 4184 + }, + { + "epoch": 0.3774180457230464, + "grad_norm": 0.6518708530853708, + "learning_rate": 2.8620107517682597e-06, + "loss": 0.5608, + "step": 4185 + }, + { + "epoch": 0.37750822924651667, + "grad_norm": 1.3907835366393464, + "learning_rate": 2.8614835529139618e-06, + "loss": 0.7014, + "step": 4186 + }, + { + "epoch": 0.3775984127699869, + "grad_norm": 1.7648804031355287, + "learning_rate": 2.8609562805531367e-06, + "loss": 0.7329, + "step": 4187 + }, + { + "epoch": 0.3776885962934572, + "grad_norm": 1.6955119072854021, + "learning_rate": 2.8604289347307746e-06, + "loss": 0.8062, + "step": 4188 + }, + { + "epoch": 0.3777787798169274, + "grad_norm": 1.8226494584200246, + "learning_rate": 2.859901515491871e-06, + "loss": 0.7656, + "step": 4189 + }, + { + "epoch": 0.3778689633403977, + "grad_norm": 1.3965467319037965, + "learning_rate": 2.8593740228814298e-06, + "loss": 0.6573, + "step": 4190 + }, + { + "epoch": 0.37795914686386795, + "grad_norm": 1.703485713172168, + "learning_rate": 2.8588464569444574e-06, + "loss": 0.7605, + "step": 4191 + }, + { + "epoch": 0.37804933038733823, + "grad_norm": 1.6114233370307505, + "learning_rate": 2.8583188177259697e-06, + "loss": 0.7587, + "step": 4192 + }, + { + "epoch": 0.3781395139108085, + "grad_norm": 2.1025333927147196, + "learning_rate": 2.857791105270988e-06, + "loss": 0.7554, + "step": 4193 + }, + { + "epoch": 0.37822969743427876, + "grad_norm": 1.5492795500412524, + "learning_rate": 2.857263319624539e-06, + "loss": 0.7994, + "step": 4194 + }, + { + "epoch": 0.37831988095774904, + "grad_norm": 5.507341859878708, + "learning_rate": 2.856735460831657e-06, + "loss": 0.7537, + "step": 4195 + }, + { + "epoch": 0.3784100644812193, + "grad_norm": 1.5551153773100619, + "learning_rate": 2.856207528937382e-06, + "loss": 0.7825, + "step": 4196 + }, + { + "epoch": 0.37850024800468957, + "grad_norm": 1.4087451838603111, + "learning_rate": 2.855679523986759e-06, + "loss": 0.7552, + "step": 4197 + }, + { + "epoch": 0.3785904315281598, + "grad_norm": 7.603088915669098, + "learning_rate": 2.8551514460248406e-06, + "loss": 0.8184, + "step": 4198 + }, + { + "epoch": 0.3786806150516301, + "grad_norm": 1.6187274621992584, + "learning_rate": 2.8546232950966868e-06, + "loss": 0.7738, + "step": 4199 + }, + { + "epoch": 0.3787707985751003, + "grad_norm": 2.0617623599558375, + "learning_rate": 2.85409507124736e-06, + "loss": 0.718, + "step": 4200 + }, + { + "epoch": 0.3788609820985706, + "grad_norm": 4.193141496290121, + "learning_rate": 2.8535667745219324e-06, + "loss": 0.7644, + "step": 4201 + }, + { + "epoch": 0.37895116562204084, + "grad_norm": 1.4914576957705892, + "learning_rate": 2.853038404965481e-06, + "loss": 0.7557, + "step": 4202 + }, + { + "epoch": 0.37904134914551113, + "grad_norm": 2.9098157440265133, + "learning_rate": 2.8525099626230894e-06, + "loss": 0.8393, + "step": 4203 + }, + { + "epoch": 0.37913153266898136, + "grad_norm": 2.1425838000936124, + "learning_rate": 2.8519814475398472e-06, + "loss": 0.6971, + "step": 4204 + }, + { + "epoch": 0.37922171619245165, + "grad_norm": 1.3440489769055524, + "learning_rate": 2.8514528597608502e-06, + "loss": 0.7455, + "step": 4205 + }, + { + "epoch": 0.3793118997159219, + "grad_norm": 1.2590583649333313, + "learning_rate": 2.8509241993312004e-06, + "loss": 0.7969, + "step": 4206 + }, + { + "epoch": 0.3794020832393922, + "grad_norm": 1.759543554115483, + "learning_rate": 2.850395466296006e-06, + "loss": 0.8217, + "step": 4207 + }, + { + "epoch": 0.3794922667628624, + "grad_norm": 1.7501180472749782, + "learning_rate": 2.849866660700381e-06, + "loss": 0.7948, + "step": 4208 + }, + { + "epoch": 0.3795824502863327, + "grad_norm": 1.8899568370149142, + "learning_rate": 2.8493377825894464e-06, + "loss": 0.8281, + "step": 4209 + }, + { + "epoch": 0.37967263380980293, + "grad_norm": 2.3493981016687044, + "learning_rate": 2.848808832008329e-06, + "loss": 0.8298, + "step": 4210 + }, + { + "epoch": 0.3797628173332732, + "grad_norm": 1.6367576430622839, + "learning_rate": 2.848279809002162e-06, + "loss": 0.7986, + "step": 4211 + }, + { + "epoch": 0.37985300085674345, + "grad_norm": 1.85753572678639, + "learning_rate": 2.8477507136160842e-06, + "loss": 0.7894, + "step": 4212 + }, + { + "epoch": 0.37994318438021374, + "grad_norm": 0.6397162015706787, + "learning_rate": 2.847221545895241e-06, + "loss": 0.589, + "step": 4213 + }, + { + "epoch": 0.38003336790368397, + "grad_norm": 1.493184420719415, + "learning_rate": 2.846692305884785e-06, + "loss": 0.7654, + "step": 4214 + }, + { + "epoch": 0.38012355142715426, + "grad_norm": 1.3838090915347645, + "learning_rate": 2.8461629936298718e-06, + "loss": 0.816, + "step": 4215 + }, + { + "epoch": 0.3802137349506245, + "grad_norm": 4.282827935787694, + "learning_rate": 2.845633609175666e-06, + "loss": 0.8241, + "step": 4216 + }, + { + "epoch": 0.3803039184740948, + "grad_norm": 1.7032959178590097, + "learning_rate": 2.8451041525673383e-06, + "loss": 0.8076, + "step": 4217 + }, + { + "epoch": 0.38039410199756507, + "grad_norm": 6.643560357675259, + "learning_rate": 2.8445746238500647e-06, + "loss": 0.8054, + "step": 4218 + }, + { + "epoch": 0.3804842855210353, + "grad_norm": 1.897454843398442, + "learning_rate": 2.844045023069027e-06, + "loss": 0.749, + "step": 4219 + }, + { + "epoch": 0.3805744690445056, + "grad_norm": 1.6589036951101979, + "learning_rate": 2.8435153502694136e-06, + "loss": 0.8145, + "step": 4220 + }, + { + "epoch": 0.3806646525679758, + "grad_norm": 1.74485526656959, + "learning_rate": 2.84298560549642e-06, + "loss": 0.7061, + "step": 4221 + }, + { + "epoch": 0.3807548360914461, + "grad_norm": 1.6780994093664607, + "learning_rate": 2.8424557887952462e-06, + "loss": 0.794, + "step": 4222 + }, + { + "epoch": 0.38084501961491635, + "grad_norm": 1.672982606581927, + "learning_rate": 2.841925900211099e-06, + "loss": 0.6988, + "step": 4223 + }, + { + "epoch": 0.38093520313838664, + "grad_norm": 1.575400011037972, + "learning_rate": 2.841395939789192e-06, + "loss": 0.8252, + "step": 4224 + }, + { + "epoch": 0.38102538666185687, + "grad_norm": 1.4835076458267549, + "learning_rate": 2.8408659075747435e-06, + "loss": 0.7921, + "step": 4225 + }, + { + "epoch": 0.38111557018532716, + "grad_norm": 1.7479777231870137, + "learning_rate": 2.8403358036129796e-06, + "loss": 0.7797, + "step": 4226 + }, + { + "epoch": 0.3812057537087974, + "grad_norm": 1.591039409501817, + "learning_rate": 2.839805627949132e-06, + "loss": 0.8298, + "step": 4227 + }, + { + "epoch": 0.3812959372322677, + "grad_norm": 3.0308343661368324, + "learning_rate": 2.8392753806284367e-06, + "loss": 0.8291, + "step": 4228 + }, + { + "epoch": 0.3813861207557379, + "grad_norm": 4.620692747618641, + "learning_rate": 2.838745061696139e-06, + "loss": 0.7268, + "step": 4229 + }, + { + "epoch": 0.3814763042792082, + "grad_norm": 1.6315275548068797, + "learning_rate": 2.838214671197487e-06, + "loss": 0.7132, + "step": 4230 + }, + { + "epoch": 0.38156648780267843, + "grad_norm": 1.3790538157750005, + "learning_rate": 2.8376842091777377e-06, + "loss": 0.883, + "step": 4231 + }, + { + "epoch": 0.3816566713261487, + "grad_norm": 1.5936201454068908, + "learning_rate": 2.8371536756821524e-06, + "loss": 0.6836, + "step": 4232 + }, + { + "epoch": 0.38174685484961896, + "grad_norm": 1.5554795284463012, + "learning_rate": 2.836623070756e-06, + "loss": 0.7547, + "step": 4233 + }, + { + "epoch": 0.38183703837308924, + "grad_norm": 1.6403930500098163, + "learning_rate": 2.8360923944445542e-06, + "loss": 0.8013, + "step": 4234 + }, + { + "epoch": 0.3819272218965595, + "grad_norm": 1.5189888189480636, + "learning_rate": 2.8355616467930947e-06, + "loss": 0.7614, + "step": 4235 + }, + { + "epoch": 0.38201740542002977, + "grad_norm": 1.967026698029724, + "learning_rate": 2.8350308278469085e-06, + "loss": 0.7533, + "step": 4236 + }, + { + "epoch": 0.3821075889435, + "grad_norm": 4.324212133717838, + "learning_rate": 2.8344999376512877e-06, + "loss": 0.8089, + "step": 4237 + }, + { + "epoch": 0.3821977724669703, + "grad_norm": 1.4350394876875459, + "learning_rate": 2.8339689762515307e-06, + "loss": 0.7671, + "step": 4238 + }, + { + "epoch": 0.3822879559904405, + "grad_norm": 1.4551260511454978, + "learning_rate": 2.8334379436929424e-06, + "loss": 0.703, + "step": 4239 + }, + { + "epoch": 0.3823781395139108, + "grad_norm": 2.245965200974957, + "learning_rate": 2.832906840020833e-06, + "loss": 0.7431, + "step": 4240 + }, + { + "epoch": 0.3824683230373811, + "grad_norm": 1.573249082828426, + "learning_rate": 2.83237566528052e-06, + "loss": 0.7863, + "step": 4241 + }, + { + "epoch": 0.38255850656085133, + "grad_norm": 1.5459455529923336, + "learning_rate": 2.831844419517325e-06, + "loss": 0.7576, + "step": 4242 + }, + { + "epoch": 0.3826486900843216, + "grad_norm": 1.9347320596859114, + "learning_rate": 2.8313131027765774e-06, + "loss": 0.809, + "step": 4243 + }, + { + "epoch": 0.38273887360779185, + "grad_norm": 1.5819219104046465, + "learning_rate": 2.8307817151036124e-06, + "loss": 0.7488, + "step": 4244 + }, + { + "epoch": 0.38282905713126214, + "grad_norm": 3.002621459028877, + "learning_rate": 2.8302502565437704e-06, + "loss": 0.7501, + "step": 4245 + }, + { + "epoch": 0.3829192406547324, + "grad_norm": 1.7483499104477893, + "learning_rate": 2.829718727142398e-06, + "loss": 0.7034, + "step": 4246 + }, + { + "epoch": 0.38300942417820266, + "grad_norm": 1.4934032340946943, + "learning_rate": 2.829187126944849e-06, + "loss": 0.7873, + "step": 4247 + }, + { + "epoch": 0.3830996077016729, + "grad_norm": 2.007081726318544, + "learning_rate": 2.8286554559964826e-06, + "loss": 0.7211, + "step": 4248 + }, + { + "epoch": 0.3831897912251432, + "grad_norm": 1.733197061366027, + "learning_rate": 2.8281237143426637e-06, + "loss": 0.7938, + "step": 4249 + }, + { + "epoch": 0.3832799747486134, + "grad_norm": 1.5660842983216015, + "learning_rate": 2.8275919020287626e-06, + "loss": 0.6784, + "step": 4250 + }, + { + "epoch": 0.3833701582720837, + "grad_norm": 1.719704241718789, + "learning_rate": 2.827060019100158e-06, + "loss": 0.8025, + "step": 4251 + }, + { + "epoch": 0.38346034179555394, + "grad_norm": 1.478794731695373, + "learning_rate": 2.8265280656022315e-06, + "loss": 0.7678, + "step": 4252 + }, + { + "epoch": 0.3835505253190242, + "grad_norm": 1.6896827144847821, + "learning_rate": 2.825996041580373e-06, + "loss": 0.6642, + "step": 4253 + }, + { + "epoch": 0.38364070884249446, + "grad_norm": 3.0032239121612543, + "learning_rate": 2.825463947079978e-06, + "loss": 0.7966, + "step": 4254 + }, + { + "epoch": 0.38373089236596475, + "grad_norm": 1.546001534509329, + "learning_rate": 2.8249317821464483e-06, + "loss": 0.7751, + "step": 4255 + }, + { + "epoch": 0.383821075889435, + "grad_norm": 1.5290853809646023, + "learning_rate": 2.824399546825189e-06, + "loss": 0.7075, + "step": 4256 + }, + { + "epoch": 0.38391125941290527, + "grad_norm": 2.243622111551622, + "learning_rate": 2.823867241161616e-06, + "loss": 0.7326, + "step": 4257 + }, + { + "epoch": 0.3840014429363755, + "grad_norm": 1.7607810491672977, + "learning_rate": 2.8233348652011456e-06, + "loss": 0.8256, + "step": 4258 + }, + { + "epoch": 0.3840916264598458, + "grad_norm": 3.782437350907163, + "learning_rate": 2.8228024189892057e-06, + "loss": 0.6851, + "step": 4259 + }, + { + "epoch": 0.384181809983316, + "grad_norm": 1.2947207563669736, + "learning_rate": 2.822269902571226e-06, + "loss": 0.7768, + "step": 4260 + }, + { + "epoch": 0.3842719935067863, + "grad_norm": 1.5574782737086064, + "learning_rate": 2.8217373159926446e-06, + "loss": 0.7539, + "step": 4261 + }, + { + "epoch": 0.38436217703025655, + "grad_norm": 1.4482435045088415, + "learning_rate": 2.8212046592989046e-06, + "loss": 0.7119, + "step": 4262 + }, + { + "epoch": 0.38445236055372684, + "grad_norm": 1.5873688500965093, + "learning_rate": 2.820671932535455e-06, + "loss": 0.7856, + "step": 4263 + }, + { + "epoch": 0.3845425440771971, + "grad_norm": 2.660948782340842, + "learning_rate": 2.8201391357477506e-06, + "loss": 0.7594, + "step": 4264 + }, + { + "epoch": 0.38463272760066736, + "grad_norm": 1.3720602775016706, + "learning_rate": 2.8196062689812525e-06, + "loss": 0.7514, + "step": 4265 + }, + { + "epoch": 0.38472291112413765, + "grad_norm": 3.9392116398857064, + "learning_rate": 2.819073332281429e-06, + "loss": 0.7186, + "step": 4266 + }, + { + "epoch": 0.3848130946476079, + "grad_norm": 0.619319733426057, + "learning_rate": 2.8185403256937524e-06, + "loss": 0.55, + "step": 4267 + }, + { + "epoch": 0.38490327817107817, + "grad_norm": 1.9864082432795924, + "learning_rate": 2.8180072492637016e-06, + "loss": 0.7379, + "step": 4268 + }, + { + "epoch": 0.3849934616945484, + "grad_norm": 2.4250998407634476, + "learning_rate": 2.817474103036762e-06, + "loss": 0.7687, + "step": 4269 + }, + { + "epoch": 0.3850836452180187, + "grad_norm": 1.938604347763924, + "learning_rate": 2.816940887058425e-06, + "loss": 0.7985, + "step": 4270 + }, + { + "epoch": 0.3851738287414889, + "grad_norm": 1.6421026206107892, + "learning_rate": 2.816407601374186e-06, + "loss": 0.809, + "step": 4271 + }, + { + "epoch": 0.3852640122649592, + "grad_norm": 1.7661920497752936, + "learning_rate": 2.815874246029549e-06, + "loss": 0.7079, + "step": 4272 + }, + { + "epoch": 0.38535419578842944, + "grad_norm": 1.5013947360410804, + "learning_rate": 2.815340821070023e-06, + "loss": 0.7236, + "step": 4273 + }, + { + "epoch": 0.38544437931189973, + "grad_norm": 1.7118174759056644, + "learning_rate": 2.814807326541122e-06, + "loss": 0.7488, + "step": 4274 + }, + { + "epoch": 0.38553456283536996, + "grad_norm": 1.3853654732483731, + "learning_rate": 2.8142737624883676e-06, + "loss": 0.7019, + "step": 4275 + }, + { + "epoch": 0.38562474635884025, + "grad_norm": 1.6035284861932841, + "learning_rate": 2.8137401289572854e-06, + "loss": 0.7715, + "step": 4276 + }, + { + "epoch": 0.3857149298823105, + "grad_norm": 1.5037065645890841, + "learning_rate": 2.8132064259934086e-06, + "loss": 0.8142, + "step": 4277 + }, + { + "epoch": 0.3858051134057808, + "grad_norm": 1.471343733038804, + "learning_rate": 2.812672653642276e-06, + "loss": 0.7664, + "step": 4278 + }, + { + "epoch": 0.385895296929251, + "grad_norm": 1.561511604498044, + "learning_rate": 2.812138811949431e-06, + "loss": 0.7825, + "step": 4279 + }, + { + "epoch": 0.3859854804527213, + "grad_norm": 1.4979499477178535, + "learning_rate": 2.8116049009604247e-06, + "loss": 0.833, + "step": 4280 + }, + { + "epoch": 0.38607566397619153, + "grad_norm": 1.6394722658157497, + "learning_rate": 2.8110709207208132e-06, + "loss": 0.659, + "step": 4281 + }, + { + "epoch": 0.3861658474996618, + "grad_norm": 3.6378996639329886, + "learning_rate": 2.810536871276158e-06, + "loss": 0.8099, + "step": 4282 + }, + { + "epoch": 0.38625603102313205, + "grad_norm": 0.7330966130295101, + "learning_rate": 2.8100027526720283e-06, + "loss": 0.6308, + "step": 4283 + }, + { + "epoch": 0.38634621454660234, + "grad_norm": 1.6508596261767916, + "learning_rate": 2.8094685649539974e-06, + "loss": 0.6985, + "step": 4284 + }, + { + "epoch": 0.3864363980700726, + "grad_norm": 0.7209194861361948, + "learning_rate": 2.8089343081676455e-06, + "loss": 0.6023, + "step": 4285 + }, + { + "epoch": 0.38652658159354286, + "grad_norm": 1.9656842900538918, + "learning_rate": 2.8083999823585577e-06, + "loss": 0.7577, + "step": 4286 + }, + { + "epoch": 0.3866167651170131, + "grad_norm": 13.109445419395191, + "learning_rate": 2.8078655875723254e-06, + "loss": 0.7954, + "step": 4287 + }, + { + "epoch": 0.3867069486404834, + "grad_norm": 2.2367492012831742, + "learning_rate": 2.807331123854547e-06, + "loss": 0.8585, + "step": 4288 + }, + { + "epoch": 0.38679713216395367, + "grad_norm": 1.9224122699815867, + "learning_rate": 2.806796591250826e-06, + "loss": 0.6444, + "step": 4289 + }, + { + "epoch": 0.3868873156874239, + "grad_norm": 2.1368424904540833, + "learning_rate": 2.8062619898067707e-06, + "loss": 0.8406, + "step": 4290 + }, + { + "epoch": 0.3869774992108942, + "grad_norm": 1.6225264135952706, + "learning_rate": 2.8057273195679963e-06, + "loss": 0.7883, + "step": 4291 + }, + { + "epoch": 0.3870676827343644, + "grad_norm": 1.7844432517124644, + "learning_rate": 2.8051925805801253e-06, + "loss": 0.7859, + "step": 4292 + }, + { + "epoch": 0.3871578662578347, + "grad_norm": 1.452383775246741, + "learning_rate": 2.804657772888783e-06, + "loss": 0.7984, + "step": 4293 + }, + { + "epoch": 0.38724804978130495, + "grad_norm": 1.9273558827711448, + "learning_rate": 2.804122896539602e-06, + "loss": 0.7711, + "step": 4294 + }, + { + "epoch": 0.38733823330477524, + "grad_norm": 1.5342993443749542, + "learning_rate": 2.8035879515782225e-06, + "loss": 0.7242, + "step": 4295 + }, + { + "epoch": 0.38742841682824547, + "grad_norm": 2.169464090507709, + "learning_rate": 2.803052938050288e-06, + "loss": 0.8207, + "step": 4296 + }, + { + "epoch": 0.38751860035171576, + "grad_norm": 14.775805642097723, + "learning_rate": 2.802517856001449e-06, + "loss": 0.8177, + "step": 4297 + }, + { + "epoch": 0.387608783875186, + "grad_norm": 6.5681938538466555, + "learning_rate": 2.801982705477361e-06, + "loss": 0.7379, + "step": 4298 + }, + { + "epoch": 0.3876989673986563, + "grad_norm": 1.3765833581389877, + "learning_rate": 2.8014474865236867e-06, + "loss": 0.7568, + "step": 4299 + }, + { + "epoch": 0.3877891509221265, + "grad_norm": 1.610377767799591, + "learning_rate": 2.800912199186094e-06, + "loss": 0.7593, + "step": 4300 + }, + { + "epoch": 0.3878793344455968, + "grad_norm": 1.4658356017065723, + "learning_rate": 2.800376843510256e-06, + "loss": 0.7811, + "step": 4301 + }, + { + "epoch": 0.38796951796906703, + "grad_norm": 1.5726190584411763, + "learning_rate": 2.799841419541852e-06, + "loss": 0.791, + "step": 4302 + }, + { + "epoch": 0.3880597014925373, + "grad_norm": 0.6439585597880062, + "learning_rate": 2.799305927326568e-06, + "loss": 0.5852, + "step": 4303 + }, + { + "epoch": 0.38814988501600756, + "grad_norm": 0.7254181330266334, + "learning_rate": 2.7987703669100955e-06, + "loss": 0.5946, + "step": 4304 + }, + { + "epoch": 0.38824006853947785, + "grad_norm": 1.5102464865037588, + "learning_rate": 2.79823473833813e-06, + "loss": 0.7801, + "step": 4305 + }, + { + "epoch": 0.3883302520629481, + "grad_norm": 1.6231197336234116, + "learning_rate": 2.797699041656376e-06, + "loss": 0.7193, + "step": 4306 + }, + { + "epoch": 0.38842043558641837, + "grad_norm": 2.1687548062933617, + "learning_rate": 2.7971632769105412e-06, + "loss": 0.7856, + "step": 4307 + }, + { + "epoch": 0.3885106191098886, + "grad_norm": 2.687320703247618, + "learning_rate": 2.79662744414634e-06, + "loss": 0.7376, + "step": 4308 + }, + { + "epoch": 0.3886008026333589, + "grad_norm": 1.5794930146857764, + "learning_rate": 2.7960915434094923e-06, + "loss": 0.7153, + "step": 4309 + }, + { + "epoch": 0.3886909861568291, + "grad_norm": 2.174786598873754, + "learning_rate": 2.7955555747457256e-06, + "loss": 0.7444, + "step": 4310 + }, + { + "epoch": 0.3887811696802994, + "grad_norm": 1.544495341646203, + "learning_rate": 2.79501953820077e-06, + "loss": 0.7675, + "step": 4311 + }, + { + "epoch": 0.3888713532037697, + "grad_norm": 1.6975661385956708, + "learning_rate": 2.7944834338203637e-06, + "loss": 0.7839, + "step": 4312 + }, + { + "epoch": 0.38896153672723993, + "grad_norm": 8.758802509414503, + "learning_rate": 2.79394726165025e-06, + "loss": 0.7329, + "step": 4313 + }, + { + "epoch": 0.3890517202507102, + "grad_norm": 7.8116254393304425, + "learning_rate": 2.793411021736178e-06, + "loss": 0.752, + "step": 4314 + }, + { + "epoch": 0.38914190377418045, + "grad_norm": 1.5958390102169302, + "learning_rate": 2.7928747141239027e-06, + "loss": 0.8616, + "step": 4315 + }, + { + "epoch": 0.38923208729765074, + "grad_norm": 0.6488620103856075, + "learning_rate": 2.7923383388591856e-06, + "loss": 0.5874, + "step": 4316 + }, + { + "epoch": 0.389322270821121, + "grad_norm": 1.9384337605476132, + "learning_rate": 2.7918018959877923e-06, + "loss": 0.7365, + "step": 4317 + }, + { + "epoch": 0.38941245434459126, + "grad_norm": 1.5875545582444819, + "learning_rate": 2.791265385555495e-06, + "loss": 0.795, + "step": 4318 + }, + { + "epoch": 0.3895026378680615, + "grad_norm": 1.8280435527688343, + "learning_rate": 2.790728807608072e-06, + "loss": 0.8316, + "step": 4319 + }, + { + "epoch": 0.3895928213915318, + "grad_norm": 2.0382008102060527, + "learning_rate": 2.790192162191307e-06, + "loss": 0.758, + "step": 4320 + }, + { + "epoch": 0.389683004915002, + "grad_norm": 2.0729562829096255, + "learning_rate": 2.78965544935099e-06, + "loss": 0.7979, + "step": 4321 + }, + { + "epoch": 0.3897731884384723, + "grad_norm": 2.062397694217822, + "learning_rate": 2.789118669132916e-06, + "loss": 0.7495, + "step": 4322 + }, + { + "epoch": 0.38986337196194254, + "grad_norm": 2.190242791674182, + "learning_rate": 2.7885818215828856e-06, + "loss": 0.8401, + "step": 4323 + }, + { + "epoch": 0.38995355548541283, + "grad_norm": 2.2635154637170216, + "learning_rate": 2.7880449067467064e-06, + "loss": 0.7263, + "step": 4324 + }, + { + "epoch": 0.39004373900888306, + "grad_norm": 1.5889570683224654, + "learning_rate": 2.78750792467019e-06, + "loss": 0.725, + "step": 4325 + }, + { + "epoch": 0.39013392253235335, + "grad_norm": 2.2980545025662766, + "learning_rate": 2.786970875399156e-06, + "loss": 0.6341, + "step": 4326 + }, + { + "epoch": 0.3902241060558236, + "grad_norm": 1.3357907487899883, + "learning_rate": 2.7864337589794267e-06, + "loss": 0.699, + "step": 4327 + }, + { + "epoch": 0.39031428957929387, + "grad_norm": 1.8454317085076275, + "learning_rate": 2.7858965754568335e-06, + "loss": 0.7816, + "step": 4328 + }, + { + "epoch": 0.3904044731027641, + "grad_norm": 1.7846138628939254, + "learning_rate": 2.785359324877211e-06, + "loss": 0.706, + "step": 4329 + }, + { + "epoch": 0.3904946566262344, + "grad_norm": 1.533492178912845, + "learning_rate": 2.7848220072864e-06, + "loss": 0.7955, + "step": 4330 + }, + { + "epoch": 0.3905848401497046, + "grad_norm": 1.7342680477123247, + "learning_rate": 2.784284622730248e-06, + "loss": 0.7557, + "step": 4331 + }, + { + "epoch": 0.3906750236731749, + "grad_norm": 1.8771750867676744, + "learning_rate": 2.7837471712546073e-06, + "loss": 0.7833, + "step": 4332 + }, + { + "epoch": 0.39076520719664515, + "grad_norm": 1.383362870686178, + "learning_rate": 2.783209652905337e-06, + "loss": 0.7378, + "step": 4333 + }, + { + "epoch": 0.39085539072011544, + "grad_norm": 1.6241877066180945, + "learning_rate": 2.7826720677283e-06, + "loss": 0.7261, + "step": 4334 + }, + { + "epoch": 0.39094557424358567, + "grad_norm": 1.6445400935499301, + "learning_rate": 2.782134415769367e-06, + "loss": 0.7347, + "step": 4335 + }, + { + "epoch": 0.39103575776705596, + "grad_norm": 1.5236133180546472, + "learning_rate": 2.7815966970744126e-06, + "loss": 0.8125, + "step": 4336 + }, + { + "epoch": 0.39112594129052625, + "grad_norm": 1.8501171796649019, + "learning_rate": 2.7810589116893184e-06, + "loss": 0.8313, + "step": 4337 + }, + { + "epoch": 0.3912161248139965, + "grad_norm": 4.395532654999505, + "learning_rate": 2.780521059659972e-06, + "loss": 0.7703, + "step": 4338 + }, + { + "epoch": 0.39130630833746677, + "grad_norm": 3.1690718568243006, + "learning_rate": 2.7799831410322637e-06, + "loss": 0.7542, + "step": 4339 + }, + { + "epoch": 0.391396491860937, + "grad_norm": 3.7343092397861284, + "learning_rate": 2.779445155852094e-06, + "loss": 0.7192, + "step": 4340 + }, + { + "epoch": 0.3914866753844073, + "grad_norm": 0.6516770708086107, + "learning_rate": 2.7789071041653655e-06, + "loss": 0.6268, + "step": 4341 + }, + { + "epoch": 0.3915768589078775, + "grad_norm": 1.5389535158559708, + "learning_rate": 2.7783689860179875e-06, + "loss": 0.8203, + "step": 4342 + }, + { + "epoch": 0.3916670424313478, + "grad_norm": 4.816209824664548, + "learning_rate": 2.7778308014558767e-06, + "loss": 0.7036, + "step": 4343 + }, + { + "epoch": 0.39175722595481804, + "grad_norm": 1.6820076465500917, + "learning_rate": 2.7772925505249524e-06, + "loss": 0.8423, + "step": 4344 + }, + { + "epoch": 0.39184740947828833, + "grad_norm": 1.8385593604370205, + "learning_rate": 2.7767542332711417e-06, + "loss": 0.7357, + "step": 4345 + }, + { + "epoch": 0.39193759300175857, + "grad_norm": 1.8822779866242005, + "learning_rate": 2.776215849740377e-06, + "loss": 0.7408, + "step": 4346 + }, + { + "epoch": 0.39202777652522885, + "grad_norm": 1.461844096678707, + "learning_rate": 2.775677399978596e-06, + "loss": 0.7826, + "step": 4347 + }, + { + "epoch": 0.3921179600486991, + "grad_norm": 0.5675844500636229, + "learning_rate": 2.775138884031742e-06, + "loss": 0.5729, + "step": 4348 + }, + { + "epoch": 0.3922081435721694, + "grad_norm": 1.7342787707327383, + "learning_rate": 2.774600301945764e-06, + "loss": 0.727, + "step": 4349 + }, + { + "epoch": 0.3922983270956396, + "grad_norm": 10.538369327765036, + "learning_rate": 2.774061653766618e-06, + "loss": 0.7957, + "step": 4350 + }, + { + "epoch": 0.3923885106191099, + "grad_norm": 1.9393995878853507, + "learning_rate": 2.773522939540263e-06, + "loss": 0.7606, + "step": 4351 + }, + { + "epoch": 0.39247869414258013, + "grad_norm": 1.8830910630939484, + "learning_rate": 2.7729841593126663e-06, + "loss": 0.6913, + "step": 4352 + }, + { + "epoch": 0.3925688776660504, + "grad_norm": 1.4884309468114076, + "learning_rate": 2.7724453131297988e-06, + "loss": 0.8303, + "step": 4353 + }, + { + "epoch": 0.39265906118952065, + "grad_norm": 1.5625461571551595, + "learning_rate": 2.771906401037637e-06, + "loss": 0.7926, + "step": 4354 + }, + { + "epoch": 0.39274924471299094, + "grad_norm": 20.571216718400265, + "learning_rate": 2.7713674230821664e-06, + "loss": 0.7226, + "step": 4355 + }, + { + "epoch": 0.3928394282364612, + "grad_norm": 1.3895571176704964, + "learning_rate": 2.7708283793093724e-06, + "loss": 0.7178, + "step": 4356 + }, + { + "epoch": 0.39292961175993146, + "grad_norm": 1.5923165999942692, + "learning_rate": 2.7702892697652514e-06, + "loss": 0.7391, + "step": 4357 + }, + { + "epoch": 0.3930197952834017, + "grad_norm": 1.7052069022192646, + "learning_rate": 2.7697500944958024e-06, + "loss": 0.7474, + "step": 4358 + }, + { + "epoch": 0.393109978806872, + "grad_norm": 2.172641756022425, + "learning_rate": 2.7692108535470312e-06, + "loss": 0.8661, + "step": 4359 + }, + { + "epoch": 0.3932001623303423, + "grad_norm": 2.2783904246739834, + "learning_rate": 2.768671546964948e-06, + "loss": 0.8008, + "step": 4360 + }, + { + "epoch": 0.3932903458538125, + "grad_norm": 1.6706470008254848, + "learning_rate": 2.7681321747955713e-06, + "loss": 0.7936, + "step": 4361 + }, + { + "epoch": 0.3933805293772828, + "grad_norm": 2.458552582441275, + "learning_rate": 2.767592737084921e-06, + "loss": 0.7798, + "step": 4362 + }, + { + "epoch": 0.39347071290075303, + "grad_norm": 2.324270039481356, + "learning_rate": 2.767053233879026e-06, + "loss": 0.8619, + "step": 4363 + }, + { + "epoch": 0.3935608964242233, + "grad_norm": 1.7335984837251237, + "learning_rate": 2.76651366522392e-06, + "loss": 0.8438, + "step": 4364 + }, + { + "epoch": 0.39365107994769355, + "grad_norm": 1.5991977259787211, + "learning_rate": 2.7659740311656413e-06, + "loss": 0.7375, + "step": 4365 + }, + { + "epoch": 0.39374126347116384, + "grad_norm": 2.55796552327345, + "learning_rate": 2.7654343317502352e-06, + "loss": 0.7879, + "step": 4366 + }, + { + "epoch": 0.39383144699463407, + "grad_norm": 5.477815623450336, + "learning_rate": 2.7648945670237502e-06, + "loss": 0.7465, + "step": 4367 + }, + { + "epoch": 0.39392163051810436, + "grad_norm": 2.0984863684581647, + "learning_rate": 2.7643547370322446e-06, + "loss": 0.8218, + "step": 4368 + }, + { + "epoch": 0.3940118140415746, + "grad_norm": 1.8230039157877689, + "learning_rate": 2.7638148418217775e-06, + "loss": 0.8057, + "step": 4369 + }, + { + "epoch": 0.3941019975650449, + "grad_norm": 1.6703094568037495, + "learning_rate": 2.7632748814384163e-06, + "loss": 0.8081, + "step": 4370 + }, + { + "epoch": 0.3941921810885151, + "grad_norm": 1.2934024233323307, + "learning_rate": 2.7627348559282335e-06, + "loss": 0.7928, + "step": 4371 + }, + { + "epoch": 0.3942823646119854, + "grad_norm": 1.8088940233151993, + "learning_rate": 2.7621947653373075e-06, + "loss": 0.7049, + "step": 4372 + }, + { + "epoch": 0.39437254813545564, + "grad_norm": 0.6475567178665946, + "learning_rate": 2.7616546097117213e-06, + "loss": 0.5649, + "step": 4373 + }, + { + "epoch": 0.3944627316589259, + "grad_norm": 1.5942120349820752, + "learning_rate": 2.761114389097564e-06, + "loss": 0.7609, + "step": 4374 + }, + { + "epoch": 0.39455291518239616, + "grad_norm": 1.8949619708107843, + "learning_rate": 2.7605741035409305e-06, + "loss": 0.798, + "step": 4375 + }, + { + "epoch": 0.39464309870586645, + "grad_norm": 3.9138502722060116, + "learning_rate": 2.76003375308792e-06, + "loss": 0.7475, + "step": 4376 + }, + { + "epoch": 0.3947332822293367, + "grad_norm": 1.8452527662359077, + "learning_rate": 2.75949333778464e-06, + "loss": 0.7325, + "step": 4377 + }, + { + "epoch": 0.39482346575280697, + "grad_norm": 1.5128987756061913, + "learning_rate": 2.7589528576772e-06, + "loss": 0.6474, + "step": 4378 + }, + { + "epoch": 0.3949136492762772, + "grad_norm": 1.9877031906195208, + "learning_rate": 2.758412312811717e-06, + "loss": 0.7551, + "step": 4379 + }, + { + "epoch": 0.3950038327997475, + "grad_norm": 3.105811142190096, + "learning_rate": 2.7578717032343146e-06, + "loss": 0.6308, + "step": 4380 + }, + { + "epoch": 0.3950940163232177, + "grad_norm": 1.7721345122758032, + "learning_rate": 2.757331028991119e-06, + "loss": 0.793, + "step": 4381 + }, + { + "epoch": 0.395184199846688, + "grad_norm": 3.3423916918854473, + "learning_rate": 2.7567902901282642e-06, + "loss": 0.7846, + "step": 4382 + }, + { + "epoch": 0.3952743833701583, + "grad_norm": 1.838833087882517, + "learning_rate": 2.7562494866918892e-06, + "loss": 0.7561, + "step": 4383 + }, + { + "epoch": 0.39536456689362853, + "grad_norm": 1.4541163190780784, + "learning_rate": 2.7557086187281378e-06, + "loss": 0.7348, + "step": 4384 + }, + { + "epoch": 0.3954547504170988, + "grad_norm": 3.8619812107745277, + "learning_rate": 2.75516768628316e-06, + "loss": 0.8638, + "step": 4385 + }, + { + "epoch": 0.39554493394056905, + "grad_norm": 1.6053154122666105, + "learning_rate": 2.7546266894031114e-06, + "loss": 0.8326, + "step": 4386 + }, + { + "epoch": 0.39563511746403934, + "grad_norm": 2.5416994457528377, + "learning_rate": 2.7540856281341526e-06, + "loss": 0.7113, + "step": 4387 + }, + { + "epoch": 0.3957253009875096, + "grad_norm": 1.6969198877850136, + "learning_rate": 2.7535445025224506e-06, + "loss": 0.7303, + "step": 4388 + }, + { + "epoch": 0.39581548451097986, + "grad_norm": 2.101685474300506, + "learning_rate": 2.753003312614176e-06, + "loss": 0.7658, + "step": 4389 + }, + { + "epoch": 0.3959056680344501, + "grad_norm": 1.664145812335489, + "learning_rate": 2.7524620584555065e-06, + "loss": 0.7707, + "step": 4390 + }, + { + "epoch": 0.3959958515579204, + "grad_norm": 3.963892930749123, + "learning_rate": 2.7519207400926253e-06, + "loss": 0.8034, + "step": 4391 + }, + { + "epoch": 0.3960860350813906, + "grad_norm": 1.563569427724961, + "learning_rate": 2.751379357571721e-06, + "loss": 0.7465, + "step": 4392 + }, + { + "epoch": 0.3961762186048609, + "grad_norm": 1.524206192535807, + "learning_rate": 2.7508379109389865e-06, + "loss": 0.8247, + "step": 4393 + }, + { + "epoch": 0.39626640212833114, + "grad_norm": 1.793986179341098, + "learning_rate": 2.750296400240622e-06, + "loss": 0.7783, + "step": 4394 + }, + { + "epoch": 0.39635658565180143, + "grad_norm": 1.3655122356865321, + "learning_rate": 2.7497548255228305e-06, + "loss": 0.7964, + "step": 4395 + }, + { + "epoch": 0.39644676917527166, + "grad_norm": 2.0867554850256576, + "learning_rate": 2.749213186831824e-06, + "loss": 0.7141, + "step": 4396 + }, + { + "epoch": 0.39653695269874195, + "grad_norm": 1.7613526363532723, + "learning_rate": 2.7486714842138173e-06, + "loss": 0.814, + "step": 4397 + }, + { + "epoch": 0.3966271362222122, + "grad_norm": 1.790205145007733, + "learning_rate": 2.748129717715031e-06, + "loss": 0.7152, + "step": 4398 + }, + { + "epoch": 0.3967173197456825, + "grad_norm": 1.5488139629317852, + "learning_rate": 2.747587887381692e-06, + "loss": 0.7837, + "step": 4399 + }, + { + "epoch": 0.3968075032691527, + "grad_norm": 2.3406965583486854, + "learning_rate": 2.7470459932600328e-06, + "loss": 0.7141, + "step": 4400 + }, + { + "epoch": 0.396897686792623, + "grad_norm": 1.979309585497352, + "learning_rate": 2.7465040353962897e-06, + "loss": 0.7645, + "step": 4401 + }, + { + "epoch": 0.3969878703160932, + "grad_norm": 1.5930315530098227, + "learning_rate": 2.745962013836706e-06, + "loss": 0.7588, + "step": 4402 + }, + { + "epoch": 0.3970780538395635, + "grad_norm": 2.084190637307533, + "learning_rate": 2.74541992862753e-06, + "loss": 0.7413, + "step": 4403 + }, + { + "epoch": 0.39716823736303375, + "grad_norm": 2.550738814791067, + "learning_rate": 2.744877779815016e-06, + "loss": 0.763, + "step": 4404 + }, + { + "epoch": 0.39725842088650404, + "grad_norm": 1.5249726996557789, + "learning_rate": 2.7443355674454234e-06, + "loss": 0.7616, + "step": 4405 + }, + { + "epoch": 0.39734860440997427, + "grad_norm": 2.184309239959838, + "learning_rate": 2.743793291565015e-06, + "loss": 0.8067, + "step": 4406 + }, + { + "epoch": 0.39743878793344456, + "grad_norm": 1.9246621120219727, + "learning_rate": 2.7432509522200617e-06, + "loss": 0.6891, + "step": 4407 + }, + { + "epoch": 0.39752897145691485, + "grad_norm": 1.7727150811040684, + "learning_rate": 2.7427085494568383e-06, + "loss": 0.7075, + "step": 4408 + }, + { + "epoch": 0.3976191549803851, + "grad_norm": 0.6624994844758529, + "learning_rate": 2.742166083321628e-06, + "loss": 0.6066, + "step": 4409 + }, + { + "epoch": 0.39770933850385537, + "grad_norm": 1.6754915569460387, + "learning_rate": 2.7416235538607137e-06, + "loss": 0.8031, + "step": 4410 + }, + { + "epoch": 0.3977995220273256, + "grad_norm": 1.4149030853080393, + "learning_rate": 2.7410809611203894e-06, + "loss": 0.741, + "step": 4411 + }, + { + "epoch": 0.3978897055507959, + "grad_norm": 1.6159177946483785, + "learning_rate": 2.7405383051469507e-06, + "loss": 0.723, + "step": 4412 + }, + { + "epoch": 0.3979798890742661, + "grad_norm": 1.6404586707678546, + "learning_rate": 2.7399955859867e-06, + "loss": 0.5772, + "step": 4413 + }, + { + "epoch": 0.3980700725977364, + "grad_norm": 0.6102157807951163, + "learning_rate": 2.7394528036859465e-06, + "loss": 0.5503, + "step": 4414 + }, + { + "epoch": 0.39816025612120665, + "grad_norm": 1.5756825345712275, + "learning_rate": 2.738909958291002e-06, + "loss": 0.7629, + "step": 4415 + }, + { + "epoch": 0.39825043964467693, + "grad_norm": 1.672624081512101, + "learning_rate": 2.7383670498481863e-06, + "loss": 0.6763, + "step": 4416 + }, + { + "epoch": 0.39834062316814717, + "grad_norm": 0.5848958483928385, + "learning_rate": 2.737824078403822e-06, + "loss": 0.5104, + "step": 4417 + }, + { + "epoch": 0.39843080669161746, + "grad_norm": 1.634204034269454, + "learning_rate": 2.737281044004239e-06, + "loss": 0.7449, + "step": 4418 + }, + { + "epoch": 0.3985209902150877, + "grad_norm": 1.6522166967581051, + "learning_rate": 2.736737946695772e-06, + "loss": 0.7466, + "step": 4419 + }, + { + "epoch": 0.398611173738558, + "grad_norm": 2.154485229115769, + "learning_rate": 2.736194786524761e-06, + "loss": 0.7578, + "step": 4420 + }, + { + "epoch": 0.3987013572620282, + "grad_norm": 1.5610860912328708, + "learning_rate": 2.7356515635375517e-06, + "loss": 0.7615, + "step": 4421 + }, + { + "epoch": 0.3987915407854985, + "grad_norm": 4.601928696366387, + "learning_rate": 2.735108277780495e-06, + "loss": 0.675, + "step": 4422 + }, + { + "epoch": 0.39888172430896873, + "grad_norm": 1.6438848897030085, + "learning_rate": 2.7345649292999456e-06, + "loss": 0.7532, + "step": 4423 + }, + { + "epoch": 0.398971907832439, + "grad_norm": 2.8911335265847686, + "learning_rate": 2.734021518142267e-06, + "loss": 0.7665, + "step": 4424 + }, + { + "epoch": 0.39906209135590925, + "grad_norm": 4.274714472643972, + "learning_rate": 2.733478044353825e-06, + "loss": 0.5978, + "step": 4425 + }, + { + "epoch": 0.39915227487937954, + "grad_norm": 2.3422660707575313, + "learning_rate": 2.7329345079809917e-06, + "loss": 0.7633, + "step": 4426 + }, + { + "epoch": 0.3992424584028498, + "grad_norm": 3.26090640613919, + "learning_rate": 2.7323909090701447e-06, + "loss": 0.7043, + "step": 4427 + }, + { + "epoch": 0.39933264192632006, + "grad_norm": 2.759889852288518, + "learning_rate": 2.731847247667667e-06, + "loss": 0.8152, + "step": 4428 + }, + { + "epoch": 0.3994228254497903, + "grad_norm": 2.0256965428247296, + "learning_rate": 2.731303523819947e-06, + "loss": 0.7372, + "step": 4429 + }, + { + "epoch": 0.3995130089732606, + "grad_norm": 2.1763489015248023, + "learning_rate": 2.7307597375733783e-06, + "loss": 0.8231, + "step": 4430 + }, + { + "epoch": 0.3996031924967309, + "grad_norm": 1.8426593852366233, + "learning_rate": 2.7302158889743587e-06, + "loss": 0.7658, + "step": 4431 + }, + { + "epoch": 0.3996933760202011, + "grad_norm": 1.6292384698496976, + "learning_rate": 2.7296719780692937e-06, + "loss": 0.7224, + "step": 4432 + }, + { + "epoch": 0.3997835595436714, + "grad_norm": 1.486825546462053, + "learning_rate": 2.7291280049045916e-06, + "loss": 0.6719, + "step": 4433 + }, + { + "epoch": 0.39987374306714163, + "grad_norm": 1.577447273594417, + "learning_rate": 2.7285839695266683e-06, + "loss": 0.7011, + "step": 4434 + }, + { + "epoch": 0.3999639265906119, + "grad_norm": 4.2549757160430675, + "learning_rate": 2.7280398719819423e-06, + "loss": 0.7085, + "step": 4435 + }, + { + "epoch": 0.40005411011408215, + "grad_norm": 2.0099679976615388, + "learning_rate": 2.727495712316841e-06, + "loss": 0.8239, + "step": 4436 + }, + { + "epoch": 0.40014429363755244, + "grad_norm": 2.0059999827102555, + "learning_rate": 2.7269514905777945e-06, + "loss": 0.7113, + "step": 4437 + }, + { + "epoch": 0.4002344771610227, + "grad_norm": 1.5418064810098682, + "learning_rate": 2.7264072068112377e-06, + "loss": 0.8832, + "step": 4438 + }, + { + "epoch": 0.40032466068449296, + "grad_norm": 2.1676614629495408, + "learning_rate": 2.7258628610636133e-06, + "loss": 0.7248, + "step": 4439 + }, + { + "epoch": 0.4004148442079632, + "grad_norm": 1.970489641616803, + "learning_rate": 2.7253184533813667e-06, + "loss": 0.8366, + "step": 4440 + }, + { + "epoch": 0.4005050277314335, + "grad_norm": 1.7132505345407028, + "learning_rate": 2.72477398381095e-06, + "loss": 0.7735, + "step": 4441 + }, + { + "epoch": 0.4005952112549037, + "grad_norm": 1.3459105311044015, + "learning_rate": 2.724229452398821e-06, + "loss": 0.7636, + "step": 4442 + }, + { + "epoch": 0.400685394778374, + "grad_norm": 1.6654709420877323, + "learning_rate": 2.7236848591914422e-06, + "loss": 0.7803, + "step": 4443 + }, + { + "epoch": 0.40077557830184424, + "grad_norm": 1.5192903067781818, + "learning_rate": 2.7231402042352803e-06, + "loss": 0.7488, + "step": 4444 + }, + { + "epoch": 0.4008657618253145, + "grad_norm": 1.476670841501045, + "learning_rate": 2.722595487576809e-06, + "loss": 0.7896, + "step": 4445 + }, + { + "epoch": 0.40095594534878476, + "grad_norm": 1.6392905484842684, + "learning_rate": 2.722050709262506e-06, + "loss": 0.8087, + "step": 4446 + }, + { + "epoch": 0.40104612887225505, + "grad_norm": 1.3450815567475336, + "learning_rate": 2.7215058693388557e-06, + "loss": 0.74, + "step": 4447 + }, + { + "epoch": 0.4011363123957253, + "grad_norm": 2.0557496549134475, + "learning_rate": 2.720960967852346e-06, + "loss": 0.6695, + "step": 4448 + }, + { + "epoch": 0.40122649591919557, + "grad_norm": 1.5571165039487274, + "learning_rate": 2.720416004849471e-06, + "loss": 0.6914, + "step": 4449 + }, + { + "epoch": 0.4013166794426658, + "grad_norm": 2.442105759163683, + "learning_rate": 2.7198709803767304e-06, + "loss": 0.713, + "step": 4450 + }, + { + "epoch": 0.4014068629661361, + "grad_norm": 1.3547238108554565, + "learning_rate": 2.7193258944806286e-06, + "loss": 0.7289, + "step": 4451 + }, + { + "epoch": 0.4014970464896063, + "grad_norm": 1.5609624545648615, + "learning_rate": 2.718780747207675e-06, + "loss": 0.7638, + "step": 4452 + }, + { + "epoch": 0.4015872300130766, + "grad_norm": 2.609175897186174, + "learning_rate": 2.7182355386043847e-06, + "loss": 0.8234, + "step": 4453 + }, + { + "epoch": 0.40167741353654685, + "grad_norm": 1.916312191844199, + "learning_rate": 2.717690268717278e-06, + "loss": 0.8257, + "step": 4454 + }, + { + "epoch": 0.40176759706001713, + "grad_norm": 1.4892427637559345, + "learning_rate": 2.7171449375928803e-06, + "loss": 0.7877, + "step": 4455 + }, + { + "epoch": 0.4018577805834874, + "grad_norm": 2.1045327450413245, + "learning_rate": 2.716599545277722e-06, + "loss": 0.7942, + "step": 4456 + }, + { + "epoch": 0.40194796410695766, + "grad_norm": 1.5862351312083252, + "learning_rate": 2.7160540918183394e-06, + "loss": 0.7263, + "step": 4457 + }, + { + "epoch": 0.40203814763042794, + "grad_norm": 2.8463223738551124, + "learning_rate": 2.715508577261273e-06, + "loss": 0.676, + "step": 4458 + }, + { + "epoch": 0.4021283311538982, + "grad_norm": 2.346242672680656, + "learning_rate": 2.7149630016530702e-06, + "loss": 0.7381, + "step": 4459 + }, + { + "epoch": 0.40221851467736847, + "grad_norm": 1.5288970151689596, + "learning_rate": 2.7144173650402815e-06, + "loss": 0.7795, + "step": 4460 + }, + { + "epoch": 0.4023086982008387, + "grad_norm": 1.442792685713285, + "learning_rate": 2.7138716674694636e-06, + "loss": 0.7832, + "step": 4461 + }, + { + "epoch": 0.402398881724309, + "grad_norm": 2.162954769936716, + "learning_rate": 2.7133259089871795e-06, + "loss": 0.7143, + "step": 4462 + }, + { + "epoch": 0.4024890652477792, + "grad_norm": 1.9714834825102932, + "learning_rate": 2.712780089639995e-06, + "loss": 0.758, + "step": 4463 + }, + { + "epoch": 0.4025792487712495, + "grad_norm": 1.5384261466503713, + "learning_rate": 2.712234209474483e-06, + "loss": 0.6356, + "step": 4464 + }, + { + "epoch": 0.40266943229471974, + "grad_norm": 2.080813167883227, + "learning_rate": 2.7116882685372218e-06, + "loss": 0.805, + "step": 4465 + }, + { + "epoch": 0.40275961581819003, + "grad_norm": 1.548013056564468, + "learning_rate": 2.7111422668747927e-06, + "loss": 0.7242, + "step": 4466 + }, + { + "epoch": 0.40284979934166026, + "grad_norm": 1.4618962044093387, + "learning_rate": 2.7105962045337846e-06, + "loss": 0.7575, + "step": 4467 + }, + { + "epoch": 0.40293998286513055, + "grad_norm": 1.5860421282451356, + "learning_rate": 2.7100500815607898e-06, + "loss": 0.7535, + "step": 4468 + }, + { + "epoch": 0.4030301663886008, + "grad_norm": 1.6092181082980836, + "learning_rate": 2.709503898002407e-06, + "loss": 0.7782, + "step": 4469 + }, + { + "epoch": 0.4031203499120711, + "grad_norm": 1.7129145656465479, + "learning_rate": 2.708957653905239e-06, + "loss": 0.7549, + "step": 4470 + }, + { + "epoch": 0.4032105334355413, + "grad_norm": 1.4580594032598426, + "learning_rate": 2.7084113493158956e-06, + "loss": 0.7715, + "step": 4471 + }, + { + "epoch": 0.4033007169590116, + "grad_norm": 1.3694977777441462, + "learning_rate": 2.7078649842809888e-06, + "loss": 0.7716, + "step": 4472 + }, + { + "epoch": 0.40339090048248183, + "grad_norm": 1.50764503932647, + "learning_rate": 2.707318558847139e-06, + "loss": 0.8083, + "step": 4473 + }, + { + "epoch": 0.4034810840059521, + "grad_norm": 0.7429217671682636, + "learning_rate": 2.7067720730609697e-06, + "loss": 0.5804, + "step": 4474 + }, + { + "epoch": 0.40357126752942235, + "grad_norm": 1.6951004106605838, + "learning_rate": 2.70622552696911e-06, + "loss": 0.7311, + "step": 4475 + }, + { + "epoch": 0.40366145105289264, + "grad_norm": 1.4424593420265468, + "learning_rate": 2.7056789206181943e-06, + "loss": 0.7143, + "step": 4476 + }, + { + "epoch": 0.40375163457636287, + "grad_norm": 6.172781962138272, + "learning_rate": 2.7051322540548615e-06, + "loss": 0.8157, + "step": 4477 + }, + { + "epoch": 0.40384181809983316, + "grad_norm": 1.7978505015952873, + "learning_rate": 2.704585527325757e-06, + "loss": 0.7341, + "step": 4478 + }, + { + "epoch": 0.40393200162330345, + "grad_norm": 1.9122411041551566, + "learning_rate": 2.7040387404775303e-06, + "loss": 0.7216, + "step": 4479 + }, + { + "epoch": 0.4040221851467737, + "grad_norm": 1.5684017113925086, + "learning_rate": 2.703491893556837e-06, + "loss": 0.7934, + "step": 4480 + }, + { + "epoch": 0.40411236867024397, + "grad_norm": 1.4035488465536816, + "learning_rate": 2.702944986610335e-06, + "loss": 0.7363, + "step": 4481 + }, + { + "epoch": 0.4042025521937142, + "grad_norm": 1.7794629552318189, + "learning_rate": 2.7023980196846917e-06, + "loss": 0.7751, + "step": 4482 + }, + { + "epoch": 0.4042927357171845, + "grad_norm": 0.7164085646064666, + "learning_rate": 2.7018509928265763e-06, + "loss": 0.5981, + "step": 4483 + }, + { + "epoch": 0.4043829192406547, + "grad_norm": 1.8286827736294815, + "learning_rate": 2.7013039060826635e-06, + "loss": 0.7882, + "step": 4484 + }, + { + "epoch": 0.404473102764125, + "grad_norm": 1.9571962572535613, + "learning_rate": 2.7007567594996347e-06, + "loss": 0.8318, + "step": 4485 + }, + { + "epoch": 0.40456328628759525, + "grad_norm": 1.7355829791929236, + "learning_rate": 2.7002095531241757e-06, + "loss": 0.6676, + "step": 4486 + }, + { + "epoch": 0.40465346981106554, + "grad_norm": 2.0103273783672377, + "learning_rate": 2.6996622870029767e-06, + "loss": 0.7344, + "step": 4487 + }, + { + "epoch": 0.40474365333453577, + "grad_norm": 1.8085788536694178, + "learning_rate": 2.6991149611827335e-06, + "loss": 0.8716, + "step": 4488 + }, + { + "epoch": 0.40483383685800606, + "grad_norm": 1.740328220016273, + "learning_rate": 2.6985675757101466e-06, + "loss": 0.798, + "step": 4489 + }, + { + "epoch": 0.4049240203814763, + "grad_norm": 2.75522498622734, + "learning_rate": 2.698020130631922e-06, + "loss": 0.7201, + "step": 4490 + }, + { + "epoch": 0.4050142039049466, + "grad_norm": 1.7097110620184859, + "learning_rate": 2.6974726259947713e-06, + "loss": 0.6933, + "step": 4491 + }, + { + "epoch": 0.4051043874284168, + "grad_norm": 3.5073956961971295, + "learning_rate": 2.6969250618454106e-06, + "loss": 0.7431, + "step": 4492 + }, + { + "epoch": 0.4051945709518871, + "grad_norm": 1.467730391449207, + "learning_rate": 2.696377438230561e-06, + "loss": 0.8327, + "step": 4493 + }, + { + "epoch": 0.40528475447535733, + "grad_norm": 2.1876323114979948, + "learning_rate": 2.6958297551969484e-06, + "loss": 0.6986, + "step": 4494 + }, + { + "epoch": 0.4053749379988276, + "grad_norm": 2.3885664709982333, + "learning_rate": 2.695282012791304e-06, + "loss": 0.7291, + "step": 4495 + }, + { + "epoch": 0.40546512152229786, + "grad_norm": 2.4502134113395915, + "learning_rate": 2.6947342110603646e-06, + "loss": 0.8149, + "step": 4496 + }, + { + "epoch": 0.40555530504576814, + "grad_norm": 1.6950665836140868, + "learning_rate": 2.6941863500508717e-06, + "loss": 0.8166, + "step": 4497 + }, + { + "epoch": 0.4056454885692384, + "grad_norm": 1.6761794250787632, + "learning_rate": 2.693638429809572e-06, + "loss": 0.8163, + "step": 4498 + }, + { + "epoch": 0.40573567209270867, + "grad_norm": 1.8772946621576683, + "learning_rate": 2.6930904503832167e-06, + "loss": 0.7452, + "step": 4499 + }, + { + "epoch": 0.4058258556161789, + "grad_norm": 1.551112449504842, + "learning_rate": 2.692542411818562e-06, + "loss": 0.7442, + "step": 4500 + }, + { + "epoch": 0.4059160391396492, + "grad_norm": 1.7178535204204255, + "learning_rate": 2.69199431416237e-06, + "loss": 0.7756, + "step": 4501 + }, + { + "epoch": 0.4060062226631194, + "grad_norm": 1.4949232657141753, + "learning_rate": 2.691446157461408e-06, + "loss": 0.7161, + "step": 4502 + }, + { + "epoch": 0.4060964061865897, + "grad_norm": 1.8148268529308174, + "learning_rate": 2.690897941762447e-06, + "loss": 0.8221, + "step": 4503 + }, + { + "epoch": 0.40618658971006, + "grad_norm": 1.7319197588298918, + "learning_rate": 2.6903496671122642e-06, + "loss": 0.7402, + "step": 4504 + }, + { + "epoch": 0.40627677323353023, + "grad_norm": 0.6427196358460255, + "learning_rate": 2.689801333557641e-06, + "loss": 0.5293, + "step": 4505 + }, + { + "epoch": 0.4063669567570005, + "grad_norm": 1.8862073375005446, + "learning_rate": 2.689252941145365e-06, + "loss": 0.7638, + "step": 4506 + }, + { + "epoch": 0.40645714028047075, + "grad_norm": 1.381717097156213, + "learning_rate": 2.6887044899222277e-06, + "loss": 0.7572, + "step": 4507 + }, + { + "epoch": 0.40654732380394104, + "grad_norm": 1.6036011908837204, + "learning_rate": 2.688155979935025e-06, + "loss": 0.8071, + "step": 4508 + }, + { + "epoch": 0.4066375073274113, + "grad_norm": 1.6147360534632669, + "learning_rate": 2.68760741123056e-06, + "loss": 0.7192, + "step": 4509 + }, + { + "epoch": 0.40672769085088156, + "grad_norm": 2.6590483342684523, + "learning_rate": 2.6870587838556394e-06, + "loss": 0.7802, + "step": 4510 + }, + { + "epoch": 0.4068178743743518, + "grad_norm": 1.8082141848687605, + "learning_rate": 2.686510097857075e-06, + "loss": 0.7772, + "step": 4511 + }, + { + "epoch": 0.4069080578978221, + "grad_norm": 1.4458378842123711, + "learning_rate": 2.685961353281683e-06, + "loss": 0.7544, + "step": 4512 + }, + { + "epoch": 0.4069982414212923, + "grad_norm": 1.821447387031801, + "learning_rate": 2.6854125501762863e-06, + "loss": 0.7463, + "step": 4513 + }, + { + "epoch": 0.4070884249447626, + "grad_norm": 1.9094289319402205, + "learning_rate": 2.684863688587712e-06, + "loss": 0.7659, + "step": 4514 + }, + { + "epoch": 0.40717860846823284, + "grad_norm": 1.8381889012036219, + "learning_rate": 2.6843147685627916e-06, + "loss": 0.7328, + "step": 4515 + }, + { + "epoch": 0.4072687919917031, + "grad_norm": 2.626737382956701, + "learning_rate": 2.683765790148361e-06, + "loss": 0.7427, + "step": 4516 + }, + { + "epoch": 0.40735897551517336, + "grad_norm": 1.751151931924835, + "learning_rate": 2.6832167533912637e-06, + "loss": 0.8132, + "step": 4517 + }, + { + "epoch": 0.40744915903864365, + "grad_norm": 2.655345897520648, + "learning_rate": 2.682667658338345e-06, + "loss": 0.8356, + "step": 4518 + }, + { + "epoch": 0.4075393425621139, + "grad_norm": 1.6474769288504043, + "learning_rate": 2.682118505036458e-06, + "loss": 0.732, + "step": 4519 + }, + { + "epoch": 0.40762952608558417, + "grad_norm": 0.6274306715703802, + "learning_rate": 2.681569293532459e-06, + "loss": 0.5496, + "step": 4520 + }, + { + "epoch": 0.4077197096090544, + "grad_norm": 0.6567347643449629, + "learning_rate": 2.6810200238732102e-06, + "loss": 0.5783, + "step": 4521 + }, + { + "epoch": 0.4078098931325247, + "grad_norm": 1.3736884624523475, + "learning_rate": 2.6804706961055776e-06, + "loss": 0.8038, + "step": 4522 + }, + { + "epoch": 0.4079000766559949, + "grad_norm": 1.447677821303844, + "learning_rate": 2.6799213102764326e-06, + "loss": 0.795, + "step": 4523 + }, + { + "epoch": 0.4079902601794652, + "grad_norm": 2.0797376757042345, + "learning_rate": 2.679371866432653e-06, + "loss": 0.8535, + "step": 4524 + }, + { + "epoch": 0.40808044370293545, + "grad_norm": 1.6732010618240667, + "learning_rate": 2.6788223646211194e-06, + "loss": 0.7313, + "step": 4525 + }, + { + "epoch": 0.40817062722640574, + "grad_norm": 1.5417765586796035, + "learning_rate": 2.6782728048887183e-06, + "loss": 0.6596, + "step": 4526 + }, + { + "epoch": 0.408260810749876, + "grad_norm": 1.4628498800025038, + "learning_rate": 2.6777231872823416e-06, + "loss": 0.7466, + "step": 4527 + }, + { + "epoch": 0.40835099427334626, + "grad_norm": 1.8672890735276102, + "learning_rate": 2.6771735118488864e-06, + "loss": 0.7613, + "step": 4528 + }, + { + "epoch": 0.40844117779681655, + "grad_norm": 1.4799014491699678, + "learning_rate": 2.6766237786352523e-06, + "loss": 0.7974, + "step": 4529 + }, + { + "epoch": 0.4085313613202868, + "grad_norm": 1.8815375479003695, + "learning_rate": 2.676073987688347e-06, + "loss": 0.6917, + "step": 4530 + }, + { + "epoch": 0.40862154484375707, + "grad_norm": 1.5961141815595654, + "learning_rate": 2.6755241390550818e-06, + "loss": 0.7617, + "step": 4531 + }, + { + "epoch": 0.4087117283672273, + "grad_norm": 1.765197634668103, + "learning_rate": 2.6749742327823716e-06, + "loss": 0.7614, + "step": 4532 + }, + { + "epoch": 0.4088019118906976, + "grad_norm": 2.8748981623650125, + "learning_rate": 2.674424268917138e-06, + "loss": 0.7753, + "step": 4533 + }, + { + "epoch": 0.4088920954141678, + "grad_norm": 2.0251843780596173, + "learning_rate": 2.6738742475063074e-06, + "loss": 0.8217, + "step": 4534 + }, + { + "epoch": 0.4089822789376381, + "grad_norm": 1.8700121339618496, + "learning_rate": 2.6733241685968104e-06, + "loss": 0.8042, + "step": 4535 + }, + { + "epoch": 0.40907246246110834, + "grad_norm": 1.9799996747873982, + "learning_rate": 2.6727740322355826e-06, + "loss": 0.7832, + "step": 4536 + }, + { + "epoch": 0.40916264598457863, + "grad_norm": 2.3995992723339548, + "learning_rate": 2.6722238384695644e-06, + "loss": 0.746, + "step": 4537 + }, + { + "epoch": 0.40925282950804887, + "grad_norm": 1.5293714717049505, + "learning_rate": 2.671673587345702e-06, + "loss": 0.7758, + "step": 4538 + }, + { + "epoch": 0.40934301303151915, + "grad_norm": 0.6621824466139853, + "learning_rate": 2.6711232789109455e-06, + "loss": 0.5685, + "step": 4539 + }, + { + "epoch": 0.4094331965549894, + "grad_norm": 1.9897924649504553, + "learning_rate": 2.6705729132122497e-06, + "loss": 0.7641, + "step": 4540 + }, + { + "epoch": 0.4095233800784597, + "grad_norm": 1.853188617508479, + "learning_rate": 2.670022490296576e-06, + "loss": 0.8383, + "step": 4541 + }, + { + "epoch": 0.4096135636019299, + "grad_norm": 2.6009340185811043, + "learning_rate": 2.669472010210889e-06, + "loss": 0.8202, + "step": 4542 + }, + { + "epoch": 0.4097037471254002, + "grad_norm": 1.7650290639715904, + "learning_rate": 2.668921473002159e-06, + "loss": 0.7416, + "step": 4543 + }, + { + "epoch": 0.40979393064887043, + "grad_norm": 1.613206986387539, + "learning_rate": 2.6683708787173596e-06, + "loss": 0.8142, + "step": 4544 + }, + { + "epoch": 0.4098841141723407, + "grad_norm": 1.5535649986948334, + "learning_rate": 2.6678202274034718e-06, + "loss": 0.7792, + "step": 4545 + }, + { + "epoch": 0.40997429769581095, + "grad_norm": 1.6472791606517931, + "learning_rate": 2.66726951910748e-06, + "loss": 0.8523, + "step": 4546 + }, + { + "epoch": 0.41006448121928124, + "grad_norm": 1.5056165605801783, + "learning_rate": 2.6667187538763737e-06, + "loss": 0.7626, + "step": 4547 + }, + { + "epoch": 0.4101546647427515, + "grad_norm": 1.796354865382512, + "learning_rate": 2.6661679317571473e-06, + "loss": 0.7428, + "step": 4548 + }, + { + "epoch": 0.41024484826622176, + "grad_norm": 2.6118299619758285, + "learning_rate": 2.665617052796799e-06, + "loss": 0.7434, + "step": 4549 + }, + { + "epoch": 0.41033503178969205, + "grad_norm": 1.9658478433165956, + "learning_rate": 2.6650661170423346e-06, + "loss": 0.6463, + "step": 4550 + }, + { + "epoch": 0.4104252153131623, + "grad_norm": 1.8169939300486109, + "learning_rate": 2.6645151245407614e-06, + "loss": 0.8483, + "step": 4551 + }, + { + "epoch": 0.4105153988366326, + "grad_norm": 1.530274372307468, + "learning_rate": 2.6639640753390936e-06, + "loss": 0.7567, + "step": 4552 + }, + { + "epoch": 0.4106055823601028, + "grad_norm": 1.475091191075289, + "learning_rate": 2.66341296948435e-06, + "loss": 0.6706, + "step": 4553 + }, + { + "epoch": 0.4106957658835731, + "grad_norm": 2.443683898500085, + "learning_rate": 2.6628618070235534e-06, + "loss": 0.7721, + "step": 4554 + }, + { + "epoch": 0.4107859494070433, + "grad_norm": 2.431421967139311, + "learning_rate": 2.662310588003733e-06, + "loss": 0.7668, + "step": 4555 + }, + { + "epoch": 0.4108761329305136, + "grad_norm": 2.3453565050809, + "learning_rate": 2.6617593124719205e-06, + "loss": 0.7373, + "step": 4556 + }, + { + "epoch": 0.41096631645398385, + "grad_norm": 1.4085628986418166, + "learning_rate": 2.661207980475155e-06, + "loss": 0.7697, + "step": 4557 + }, + { + "epoch": 0.41105649997745414, + "grad_norm": 2.0395745697716423, + "learning_rate": 2.6606565920604793e-06, + "loss": 0.722, + "step": 4558 + }, + { + "epoch": 0.41114668350092437, + "grad_norm": 2.1598689352114078, + "learning_rate": 2.66010514727494e-06, + "loss": 0.8397, + "step": 4559 + }, + { + "epoch": 0.41123686702439466, + "grad_norm": 1.7889307631586433, + "learning_rate": 2.659553646165589e-06, + "loss": 0.6586, + "step": 4560 + }, + { + "epoch": 0.4113270505478649, + "grad_norm": 1.9621066069692612, + "learning_rate": 2.659002088779485e-06, + "loss": 0.7357, + "step": 4561 + }, + { + "epoch": 0.4114172340713352, + "grad_norm": 0.644402739689033, + "learning_rate": 2.6584504751636888e-06, + "loss": 0.5276, + "step": 4562 + }, + { + "epoch": 0.4115074175948054, + "grad_norm": 1.6188111194280546, + "learning_rate": 2.657898805365268e-06, + "loss": 0.7281, + "step": 4563 + }, + { + "epoch": 0.4115976011182757, + "grad_norm": 1.7990768873611163, + "learning_rate": 2.657347079431293e-06, + "loss": 0.7303, + "step": 4564 + }, + { + "epoch": 0.41168778464174594, + "grad_norm": 1.304087975091153, + "learning_rate": 2.6567952974088403e-06, + "loss": 0.7333, + "step": 4565 + }, + { + "epoch": 0.4117779681652162, + "grad_norm": 1.5251896099345947, + "learning_rate": 2.6562434593449917e-06, + "loss": 0.8128, + "step": 4566 + }, + { + "epoch": 0.41186815168868646, + "grad_norm": 1.9120571920348466, + "learning_rate": 2.6556915652868325e-06, + "loss": 0.7854, + "step": 4567 + }, + { + "epoch": 0.41195833521215675, + "grad_norm": 1.8206985129834246, + "learning_rate": 2.6551396152814534e-06, + "loss": 0.7673, + "step": 4568 + }, + { + "epoch": 0.412048518735627, + "grad_norm": 1.7944836833995406, + "learning_rate": 2.65458760937595e-06, + "loss": 0.7544, + "step": 4569 + }, + { + "epoch": 0.41213870225909727, + "grad_norm": 1.4735727064086073, + "learning_rate": 2.654035547617423e-06, + "loss": 0.709, + "step": 4570 + }, + { + "epoch": 0.4122288857825675, + "grad_norm": 1.9311359853328962, + "learning_rate": 2.653483430052976e-06, + "loss": 0.6663, + "step": 4571 + }, + { + "epoch": 0.4123190693060378, + "grad_norm": 2.641201972001209, + "learning_rate": 2.6529312567297197e-06, + "loss": 0.7466, + "step": 4572 + }, + { + "epoch": 0.412409252829508, + "grad_norm": 1.814080962909383, + "learning_rate": 2.652379027694768e-06, + "loss": 0.8379, + "step": 4573 + }, + { + "epoch": 0.4124994363529783, + "grad_norm": 2.908577919834205, + "learning_rate": 2.651826742995241e-06, + "loss": 0.8188, + "step": 4574 + }, + { + "epoch": 0.4125896198764486, + "grad_norm": 1.9220218757392729, + "learning_rate": 2.651274402678262e-06, + "loss": 0.7544, + "step": 4575 + }, + { + "epoch": 0.41267980339991883, + "grad_norm": 1.5980265218613061, + "learning_rate": 2.6507220067909597e-06, + "loss": 0.7864, + "step": 4576 + }, + { + "epoch": 0.4127699869233891, + "grad_norm": 1.6428427828877903, + "learning_rate": 2.650169555380468e-06, + "loss": 0.7203, + "step": 4577 + }, + { + "epoch": 0.41286017044685935, + "grad_norm": 1.9658546350006068, + "learning_rate": 2.6496170484939253e-06, + "loss": 0.7546, + "step": 4578 + }, + { + "epoch": 0.41295035397032964, + "grad_norm": 1.956001178163825, + "learning_rate": 2.6490644861784735e-06, + "loss": 0.7753, + "step": 4579 + }, + { + "epoch": 0.4130405374937999, + "grad_norm": 2.0405230790253652, + "learning_rate": 2.648511868481261e-06, + "loss": 0.7803, + "step": 4580 + }, + { + "epoch": 0.41313072101727016, + "grad_norm": 1.730351222333378, + "learning_rate": 2.6479591954494397e-06, + "loss": 0.7605, + "step": 4581 + }, + { + "epoch": 0.4132209045407404, + "grad_norm": 2.2658657603757044, + "learning_rate": 2.647406467130167e-06, + "loss": 0.6598, + "step": 4582 + }, + { + "epoch": 0.4133110880642107, + "grad_norm": 1.8223391535440243, + "learning_rate": 2.646853683570605e-06, + "loss": 0.7832, + "step": 4583 + }, + { + "epoch": 0.4134012715876809, + "grad_norm": 1.5484342326756255, + "learning_rate": 2.6463008448179196e-06, + "loss": 0.6988, + "step": 4584 + }, + { + "epoch": 0.4134914551111512, + "grad_norm": 1.8607189546593972, + "learning_rate": 2.6457479509192828e-06, + "loss": 0.709, + "step": 4585 + }, + { + "epoch": 0.41358163863462144, + "grad_norm": 2.087828618298443, + "learning_rate": 2.645195001921871e-06, + "loss": 0.7542, + "step": 4586 + }, + { + "epoch": 0.41367182215809173, + "grad_norm": 1.632853165261595, + "learning_rate": 2.644641997872863e-06, + "loss": 0.7364, + "step": 4587 + }, + { + "epoch": 0.41376200568156196, + "grad_norm": 2.7392425808678573, + "learning_rate": 2.644088938819445e-06, + "loss": 0.7636, + "step": 4588 + }, + { + "epoch": 0.41385218920503225, + "grad_norm": 1.8696401281411066, + "learning_rate": 2.6435358248088077e-06, + "loss": 0.8442, + "step": 4589 + }, + { + "epoch": 0.4139423727285025, + "grad_norm": 2.9621895027817895, + "learning_rate": 2.642982655888146e-06, + "loss": 0.8013, + "step": 4590 + }, + { + "epoch": 0.41403255625197277, + "grad_norm": 2.7215733392727155, + "learning_rate": 2.6424294321046585e-06, + "loss": 0.721, + "step": 4591 + }, + { + "epoch": 0.414122739775443, + "grad_norm": 0.7416275321265258, + "learning_rate": 2.641876153505549e-06, + "loss": 0.5746, + "step": 4592 + }, + { + "epoch": 0.4142129232989133, + "grad_norm": 1.8684496584003976, + "learning_rate": 2.641322820138027e-06, + "loss": 0.7661, + "step": 4593 + }, + { + "epoch": 0.4143031068223835, + "grad_norm": 2.1505680132330194, + "learning_rate": 2.640769432049306e-06, + "loss": 0.7121, + "step": 4594 + }, + { + "epoch": 0.4143932903458538, + "grad_norm": 1.6827106328113117, + "learning_rate": 2.6402159892866038e-06, + "loss": 0.679, + "step": 4595 + }, + { + "epoch": 0.41448347386932405, + "grad_norm": 1.362415902578398, + "learning_rate": 2.639662491897143e-06, + "loss": 0.747, + "step": 4596 + }, + { + "epoch": 0.41457365739279434, + "grad_norm": 1.6539257943967653, + "learning_rate": 2.639108939928152e-06, + "loss": 0.7714, + "step": 4597 + }, + { + "epoch": 0.4146638409162646, + "grad_norm": 1.820494548551616, + "learning_rate": 2.638555333426862e-06, + "loss": 0.7714, + "step": 4598 + }, + { + "epoch": 0.41475402443973486, + "grad_norm": 1.4230501956571107, + "learning_rate": 2.6380016724405093e-06, + "loss": 0.7916, + "step": 4599 + }, + { + "epoch": 0.41484420796320515, + "grad_norm": 1.6887767341807371, + "learning_rate": 2.637447957016336e-06, + "loss": 0.7139, + "step": 4600 + }, + { + "epoch": 0.4149343914866754, + "grad_norm": 1.9921025276762052, + "learning_rate": 2.636894187201589e-06, + "loss": 0.8215, + "step": 4601 + }, + { + "epoch": 0.41502457501014567, + "grad_norm": 1.9001775859359202, + "learning_rate": 2.6363403630435176e-06, + "loss": 0.6676, + "step": 4602 + }, + { + "epoch": 0.4151147585336159, + "grad_norm": 1.5156891012892346, + "learning_rate": 2.635786484589378e-06, + "loss": 0.7501, + "step": 4603 + }, + { + "epoch": 0.4152049420570862, + "grad_norm": 1.9732965317829962, + "learning_rate": 2.63523255188643e-06, + "loss": 0.6644, + "step": 4604 + }, + { + "epoch": 0.4152951255805564, + "grad_norm": 1.4254096396075182, + "learning_rate": 2.6346785649819375e-06, + "loss": 0.7818, + "step": 4605 + }, + { + "epoch": 0.4153853091040267, + "grad_norm": 1.6117428887353709, + "learning_rate": 2.6341245239231706e-06, + "loss": 0.7843, + "step": 4606 + }, + { + "epoch": 0.41547549262749695, + "grad_norm": 1.7698569429636166, + "learning_rate": 2.6335704287574024e-06, + "loss": 0.7595, + "step": 4607 + }, + { + "epoch": 0.41556567615096723, + "grad_norm": 1.913642183203138, + "learning_rate": 2.6330162795319124e-06, + "loss": 0.8332, + "step": 4608 + }, + { + "epoch": 0.41565585967443747, + "grad_norm": 1.930608614490511, + "learning_rate": 2.632462076293983e-06, + "loss": 0.812, + "step": 4609 + }, + { + "epoch": 0.41574604319790776, + "grad_norm": 1.447083496495018, + "learning_rate": 2.6319078190909017e-06, + "loss": 0.7599, + "step": 4610 + }, + { + "epoch": 0.415836226721378, + "grad_norm": 2.3992331114040724, + "learning_rate": 2.6313535079699606e-06, + "loss": 0.6758, + "step": 4611 + }, + { + "epoch": 0.4159264102448483, + "grad_norm": 2.130704851430604, + "learning_rate": 2.6307991429784572e-06, + "loss": 0.7378, + "step": 4612 + }, + { + "epoch": 0.4160165937683185, + "grad_norm": 2.162915197656209, + "learning_rate": 2.6302447241636924e-06, + "loss": 0.824, + "step": 4613 + }, + { + "epoch": 0.4161067772917888, + "grad_norm": 2.1675529012417196, + "learning_rate": 2.629690251572973e-06, + "loss": 0.8373, + "step": 4614 + }, + { + "epoch": 0.41619696081525903, + "grad_norm": 2.73036361732061, + "learning_rate": 2.629135725253609e-06, + "loss": 0.7367, + "step": 4615 + }, + { + "epoch": 0.4162871443387293, + "grad_norm": 1.5686416466972863, + "learning_rate": 2.6285811452529162e-06, + "loss": 0.7467, + "step": 4616 + }, + { + "epoch": 0.41637732786219955, + "grad_norm": 1.5977210145988325, + "learning_rate": 2.6280265116182136e-06, + "loss": 0.7777, + "step": 4617 + }, + { + "epoch": 0.41646751138566984, + "grad_norm": 1.5611341992108034, + "learning_rate": 2.6274718243968266e-06, + "loss": 0.7646, + "step": 4618 + }, + { + "epoch": 0.4165576949091401, + "grad_norm": 2.662971250710491, + "learning_rate": 2.626917083636084e-06, + "loss": 0.6801, + "step": 4619 + }, + { + "epoch": 0.41664787843261036, + "grad_norm": 1.479494121285865, + "learning_rate": 2.6263622893833183e-06, + "loss": 0.6907, + "step": 4620 + }, + { + "epoch": 0.4167380619560806, + "grad_norm": 1.8253185542887405, + "learning_rate": 2.625807441685869e-06, + "loss": 0.7061, + "step": 4621 + }, + { + "epoch": 0.4168282454795509, + "grad_norm": 0.6216690470042266, + "learning_rate": 2.625252540591078e-06, + "loss": 0.5612, + "step": 4622 + }, + { + "epoch": 0.4169184290030212, + "grad_norm": 1.839149813981733, + "learning_rate": 2.6246975861462927e-06, + "loss": 0.8372, + "step": 4623 + }, + { + "epoch": 0.4170086125264914, + "grad_norm": 1.8331016553087498, + "learning_rate": 2.624142578398864e-06, + "loss": 0.8015, + "step": 4624 + }, + { + "epoch": 0.4170987960499617, + "grad_norm": 1.4409252491026683, + "learning_rate": 2.6235875173961498e-06, + "loss": 0.7436, + "step": 4625 + }, + { + "epoch": 0.41718897957343193, + "grad_norm": 1.66401829907094, + "learning_rate": 2.62303240318551e-06, + "loss": 0.6893, + "step": 4626 + }, + { + "epoch": 0.4172791630969022, + "grad_norm": 7.551650153704507, + "learning_rate": 2.62247723581431e-06, + "loss": 0.8369, + "step": 4627 + }, + { + "epoch": 0.41736934662037245, + "grad_norm": 1.7736363236478982, + "learning_rate": 2.62192201532992e-06, + "loss": 0.7626, + "step": 4628 + }, + { + "epoch": 0.41745953014384274, + "grad_norm": 1.7331222271131859, + "learning_rate": 2.6213667417797145e-06, + "loss": 0.7775, + "step": 4629 + }, + { + "epoch": 0.41754971366731297, + "grad_norm": 1.8611713355396313, + "learning_rate": 2.6208114152110725e-06, + "loss": 0.7881, + "step": 4630 + }, + { + "epoch": 0.41763989719078326, + "grad_norm": 1.62415174205663, + "learning_rate": 2.6202560356713774e-06, + "loss": 0.7786, + "step": 4631 + }, + { + "epoch": 0.4177300807142535, + "grad_norm": 0.6445611195434359, + "learning_rate": 2.619700603208017e-06, + "loss": 0.6392, + "step": 4632 + }, + { + "epoch": 0.4178202642377238, + "grad_norm": 1.411042460680688, + "learning_rate": 2.6191451178683842e-06, + "loss": 0.6737, + "step": 4633 + }, + { + "epoch": 0.417910447761194, + "grad_norm": 1.5904499369815692, + "learning_rate": 2.6185895796998764e-06, + "loss": 0.7924, + "step": 4634 + }, + { + "epoch": 0.4180006312846643, + "grad_norm": 1.8815364708246, + "learning_rate": 2.6180339887498946e-06, + "loss": 0.8336, + "step": 4635 + }, + { + "epoch": 0.41809081480813454, + "grad_norm": 2.3231611137466994, + "learning_rate": 2.617478345065846e-06, + "loss": 0.7008, + "step": 4636 + }, + { + "epoch": 0.4181809983316048, + "grad_norm": 1.5660106130754328, + "learning_rate": 2.616922648695139e-06, + "loss": 0.7798, + "step": 4637 + }, + { + "epoch": 0.41827118185507506, + "grad_norm": 1.818785194502153, + "learning_rate": 2.61636689968519e-06, + "loss": 0.6614, + "step": 4638 + }, + { + "epoch": 0.41836136537854535, + "grad_norm": 1.9293806318541327, + "learning_rate": 2.6158110980834186e-06, + "loss": 0.8229, + "step": 4639 + }, + { + "epoch": 0.4184515489020156, + "grad_norm": 1.87486800682869, + "learning_rate": 2.615255243937249e-06, + "loss": 0.8116, + "step": 4640 + }, + { + "epoch": 0.41854173242548587, + "grad_norm": 1.8585740616322024, + "learning_rate": 2.61469933729411e-06, + "loss": 0.7854, + "step": 4641 + }, + { + "epoch": 0.4186319159489561, + "grad_norm": 2.465719270052306, + "learning_rate": 2.614143378201433e-06, + "loss": 0.7434, + "step": 4642 + }, + { + "epoch": 0.4187220994724264, + "grad_norm": 1.6276444412334798, + "learning_rate": 2.6135873667066567e-06, + "loss": 0.7467, + "step": 4643 + }, + { + "epoch": 0.4188122829958966, + "grad_norm": 1.730707638614589, + "learning_rate": 2.613031302857224e-06, + "loss": 0.7193, + "step": 4644 + }, + { + "epoch": 0.4189024665193669, + "grad_norm": 1.8444736159876245, + "learning_rate": 2.6124751867005792e-06, + "loss": 0.7861, + "step": 4645 + }, + { + "epoch": 0.4189926500428372, + "grad_norm": 1.9458603910842567, + "learning_rate": 2.611919018284175e-06, + "loss": 0.8105, + "step": 4646 + }, + { + "epoch": 0.41908283356630743, + "grad_norm": 2.5699930195398193, + "learning_rate": 2.611362797655466e-06, + "loss": 0.8129, + "step": 4647 + }, + { + "epoch": 0.4191730170897777, + "grad_norm": 2.063981304549779, + "learning_rate": 2.6108065248619124e-06, + "loss": 0.8218, + "step": 4648 + }, + { + "epoch": 0.41926320061324795, + "grad_norm": 2.1911266236898497, + "learning_rate": 2.610250199950978e-06, + "loss": 0.695, + "step": 4649 + }, + { + "epoch": 0.41935338413671824, + "grad_norm": 81.05570217738385, + "learning_rate": 2.609693822970131e-06, + "loss": 0.7367, + "step": 4650 + }, + { + "epoch": 0.4194435676601885, + "grad_norm": 3.4384768398537235, + "learning_rate": 2.609137393966846e-06, + "loss": 0.7285, + "step": 4651 + }, + { + "epoch": 0.41953375118365877, + "grad_norm": 2.2984649967750537, + "learning_rate": 2.6085809129886e-06, + "loss": 0.7152, + "step": 4652 + }, + { + "epoch": 0.419623934707129, + "grad_norm": 2.7169382482256808, + "learning_rate": 2.608024380082874e-06, + "loss": 0.7758, + "step": 4653 + }, + { + "epoch": 0.4197141182305993, + "grad_norm": 2.095722237506045, + "learning_rate": 2.6074677952971554e-06, + "loss": 0.7639, + "step": 4654 + }, + { + "epoch": 0.4198043017540695, + "grad_norm": 1.794067710891489, + "learning_rate": 2.606911158678935e-06, + "loss": 0.6877, + "step": 4655 + }, + { + "epoch": 0.4198944852775398, + "grad_norm": 2.0973573041897247, + "learning_rate": 2.606354470275708e-06, + "loss": 0.7502, + "step": 4656 + }, + { + "epoch": 0.41998466880101004, + "grad_norm": 1.6263370148750715, + "learning_rate": 2.6057977301349744e-06, + "loss": 0.7334, + "step": 4657 + }, + { + "epoch": 0.42007485232448033, + "grad_norm": 1.624175889778601, + "learning_rate": 2.6052409383042383e-06, + "loss": 0.7287, + "step": 4658 + }, + { + "epoch": 0.42016503584795056, + "grad_norm": 2.7918317281026757, + "learning_rate": 2.6046840948310074e-06, + "loss": 0.7366, + "step": 4659 + }, + { + "epoch": 0.42025521937142085, + "grad_norm": 1.6022451295080753, + "learning_rate": 2.6041271997627962e-06, + "loss": 0.7642, + "step": 4660 + }, + { + "epoch": 0.4203454028948911, + "grad_norm": 2.0401191166153136, + "learning_rate": 2.6035702531471202e-06, + "loss": 0.7506, + "step": 4661 + }, + { + "epoch": 0.4204355864183614, + "grad_norm": 1.572800141454052, + "learning_rate": 2.6030132550315035e-06, + "loss": 0.7555, + "step": 4662 + }, + { + "epoch": 0.4205257699418316, + "grad_norm": 0.6451676434734771, + "learning_rate": 2.60245620546347e-06, + "loss": 0.6144, + "step": 4663 + }, + { + "epoch": 0.4206159534653019, + "grad_norm": 1.859833220110991, + "learning_rate": 2.6018991044905517e-06, + "loss": 0.7357, + "step": 4664 + }, + { + "epoch": 0.42070613698877213, + "grad_norm": 1.6409497620809674, + "learning_rate": 2.6013419521602825e-06, + "loss": 0.7925, + "step": 4665 + }, + { + "epoch": 0.4207963205122424, + "grad_norm": 1.6890809459543625, + "learning_rate": 2.600784748520202e-06, + "loss": 0.8896, + "step": 4666 + }, + { + "epoch": 0.42088650403571265, + "grad_norm": 1.596707461054317, + "learning_rate": 2.6002274936178544e-06, + "loss": 0.6974, + "step": 4667 + }, + { + "epoch": 0.42097668755918294, + "grad_norm": 1.629709754565136, + "learning_rate": 2.5996701875007873e-06, + "loss": 0.7692, + "step": 4668 + }, + { + "epoch": 0.4210668710826532, + "grad_norm": 1.7896959940657315, + "learning_rate": 2.5991128302165533e-06, + "loss": 0.7524, + "step": 4669 + }, + { + "epoch": 0.42115705460612346, + "grad_norm": 2.7687137403331215, + "learning_rate": 2.5985554218127094e-06, + "loss": 0.8216, + "step": 4670 + }, + { + "epoch": 0.42124723812959375, + "grad_norm": 5.336312276655973, + "learning_rate": 2.597997962336816e-06, + "loss": 0.694, + "step": 4671 + }, + { + "epoch": 0.421337421653064, + "grad_norm": 1.7140681917831517, + "learning_rate": 2.5974404518364393e-06, + "loss": 0.7531, + "step": 4672 + }, + { + "epoch": 0.42142760517653427, + "grad_norm": 1.734939689467469, + "learning_rate": 2.596882890359149e-06, + "loss": 0.885, + "step": 4673 + }, + { + "epoch": 0.4215177887000045, + "grad_norm": 1.9550011130912714, + "learning_rate": 2.5963252779525196e-06, + "loss": 0.6817, + "step": 4674 + }, + { + "epoch": 0.4216079722234748, + "grad_norm": 2.5596137190496284, + "learning_rate": 2.595767614664129e-06, + "loss": 0.7586, + "step": 4675 + }, + { + "epoch": 0.421698155746945, + "grad_norm": 1.532797440232268, + "learning_rate": 2.5952099005415607e-06, + "loss": 0.7386, + "step": 4676 + }, + { + "epoch": 0.4217883392704153, + "grad_norm": 2.079936793724139, + "learning_rate": 2.594652135632402e-06, + "loss": 0.7217, + "step": 4677 + }, + { + "epoch": 0.42187852279388555, + "grad_norm": 1.6396607244598072, + "learning_rate": 2.594094319984244e-06, + "loss": 0.7155, + "step": 4678 + }, + { + "epoch": 0.42196870631735584, + "grad_norm": 1.758340713822964, + "learning_rate": 2.5935364536446825e-06, + "loss": 0.8165, + "step": 4679 + }, + { + "epoch": 0.42205888984082607, + "grad_norm": 3.537792390527477, + "learning_rate": 2.5929785366613185e-06, + "loss": 0.758, + "step": 4680 + }, + { + "epoch": 0.42214907336429636, + "grad_norm": 2.2850680864069193, + "learning_rate": 2.592420569081756e-06, + "loss": 0.7974, + "step": 4681 + }, + { + "epoch": 0.4222392568877666, + "grad_norm": 1.6908416863027236, + "learning_rate": 2.5918625509536037e-06, + "loss": 0.7223, + "step": 4682 + }, + { + "epoch": 0.4223294404112369, + "grad_norm": 0.633040963834648, + "learning_rate": 2.591304482324475e-06, + "loss": 0.579, + "step": 4683 + }, + { + "epoch": 0.4224196239347071, + "grad_norm": 2.3129475134779343, + "learning_rate": 2.5907463632419878e-06, + "loss": 0.7724, + "step": 4684 + }, + { + "epoch": 0.4225098074581774, + "grad_norm": 3.4847340762289525, + "learning_rate": 2.5901881937537632e-06, + "loss": 0.7284, + "step": 4685 + }, + { + "epoch": 0.42259999098164763, + "grad_norm": 0.8033686450132945, + "learning_rate": 2.589629973907428e-06, + "loss": 0.6121, + "step": 4686 + }, + { + "epoch": 0.4226901745051179, + "grad_norm": 2.027543780210471, + "learning_rate": 2.589071703750612e-06, + "loss": 0.7095, + "step": 4687 + }, + { + "epoch": 0.42278035802858815, + "grad_norm": 1.6337676106724521, + "learning_rate": 2.5885133833309504e-06, + "loss": 0.7973, + "step": 4688 + }, + { + "epoch": 0.42287054155205844, + "grad_norm": 1.5030070205037207, + "learning_rate": 2.5879550126960814e-06, + "loss": 0.84, + "step": 4689 + }, + { + "epoch": 0.4229607250755287, + "grad_norm": 1.605487312896411, + "learning_rate": 2.5873965918936494e-06, + "loss": 0.8141, + "step": 4690 + }, + { + "epoch": 0.42305090859899896, + "grad_norm": 1.7439374816083002, + "learning_rate": 2.586838120971301e-06, + "loss": 0.7859, + "step": 4691 + }, + { + "epoch": 0.4231410921224692, + "grad_norm": 1.6011543312051577, + "learning_rate": 2.586279599976689e-06, + "loss": 0.8195, + "step": 4692 + }, + { + "epoch": 0.4232312756459395, + "grad_norm": 1.602518085014666, + "learning_rate": 2.585721028957468e-06, + "loss": 0.7613, + "step": 4693 + }, + { + "epoch": 0.4233214591694098, + "grad_norm": 1.6803869875743007, + "learning_rate": 2.585162407961299e-06, + "loss": 0.7724, + "step": 4694 + }, + { + "epoch": 0.42341164269288, + "grad_norm": 2.489462866854615, + "learning_rate": 2.584603737035847e-06, + "loss": 0.7029, + "step": 4695 + }, + { + "epoch": 0.4235018262163503, + "grad_norm": 2.096766448302435, + "learning_rate": 2.5840450162287806e-06, + "loss": 0.7599, + "step": 4696 + }, + { + "epoch": 0.42359200973982053, + "grad_norm": 1.8946002750996942, + "learning_rate": 2.583486245587774e-06, + "loss": 0.6732, + "step": 4697 + }, + { + "epoch": 0.4236821932632908, + "grad_norm": 1.7254727061436255, + "learning_rate": 2.5829274251605023e-06, + "loss": 0.8663, + "step": 4698 + }, + { + "epoch": 0.42377237678676105, + "grad_norm": 3.2038175997112006, + "learning_rate": 2.582368554994649e-06, + "loss": 0.7328, + "step": 4699 + }, + { + "epoch": 0.42386256031023134, + "grad_norm": 0.747932445249884, + "learning_rate": 2.5818096351378994e-06, + "loss": 0.6144, + "step": 4700 + }, + { + "epoch": 0.4239527438337016, + "grad_norm": 2.3896281838192057, + "learning_rate": 2.5812506656379435e-06, + "loss": 0.8176, + "step": 4701 + }, + { + "epoch": 0.42404292735717186, + "grad_norm": 1.7060467916916242, + "learning_rate": 2.580691646542476e-06, + "loss": 0.7124, + "step": 4702 + }, + { + "epoch": 0.4241331108806421, + "grad_norm": 1.6060667386136933, + "learning_rate": 2.5801325778991958e-06, + "loss": 0.7952, + "step": 4703 + }, + { + "epoch": 0.4242232944041124, + "grad_norm": 2.752697488627783, + "learning_rate": 2.5795734597558043e-06, + "loss": 0.6201, + "step": 4704 + }, + { + "epoch": 0.4243134779275826, + "grad_norm": 2.267097678531449, + "learning_rate": 2.579014292160011e-06, + "loss": 0.7606, + "step": 4705 + }, + { + "epoch": 0.4244036614510529, + "grad_norm": 1.3544414045939357, + "learning_rate": 2.5784550751595236e-06, + "loss": 0.7414, + "step": 4706 + }, + { + "epoch": 0.42449384497452314, + "grad_norm": 6.521935058083258, + "learning_rate": 2.577895808802061e-06, + "loss": 0.8198, + "step": 4707 + }, + { + "epoch": 0.4245840284979934, + "grad_norm": 1.6406812385729364, + "learning_rate": 2.577336493135341e-06, + "loss": 0.6786, + "step": 4708 + }, + { + "epoch": 0.42467421202146366, + "grad_norm": 1.5253252120407577, + "learning_rate": 2.576777128207088e-06, + "loss": 0.8532, + "step": 4709 + }, + { + "epoch": 0.42476439554493395, + "grad_norm": 1.3849215614747585, + "learning_rate": 2.5762177140650306e-06, + "loss": 0.719, + "step": 4710 + }, + { + "epoch": 0.4248545790684042, + "grad_norm": 1.587513483545908, + "learning_rate": 2.5756582507569003e-06, + "loss": 0.768, + "step": 4711 + }, + { + "epoch": 0.42494476259187447, + "grad_norm": 1.598720802118056, + "learning_rate": 2.5750987383304335e-06, + "loss": 0.7815, + "step": 4712 + }, + { + "epoch": 0.4250349461153447, + "grad_norm": 2.6920105985061578, + "learning_rate": 2.574539176833372e-06, + "loss": 0.7452, + "step": 4713 + }, + { + "epoch": 0.425125129638815, + "grad_norm": 2.241303911129836, + "learning_rate": 2.5739795663134594e-06, + "loss": 0.7122, + "step": 4714 + }, + { + "epoch": 0.4252153131622852, + "grad_norm": 1.5339979310550191, + "learning_rate": 2.5734199068184454e-06, + "loss": 0.8453, + "step": 4715 + }, + { + "epoch": 0.4253054966857555, + "grad_norm": 1.9211771248452565, + "learning_rate": 2.572860198396083e-06, + "loss": 0.6868, + "step": 4716 + }, + { + "epoch": 0.4253956802092258, + "grad_norm": 1.4911409872520154, + "learning_rate": 2.57230044109413e-06, + "loss": 0.7837, + "step": 4717 + }, + { + "epoch": 0.42548586373269603, + "grad_norm": 2.443858631578127, + "learning_rate": 2.5717406349603483e-06, + "loss": 0.7421, + "step": 4718 + }, + { + "epoch": 0.4255760472561663, + "grad_norm": 1.50988792723522, + "learning_rate": 2.5711807800425026e-06, + "loss": 0.7773, + "step": 4719 + }, + { + "epoch": 0.42566623077963656, + "grad_norm": 2.3891868504058746, + "learning_rate": 2.5706208763883633e-06, + "loss": 0.5986, + "step": 4720 + }, + { + "epoch": 0.42575641430310684, + "grad_norm": 2.1954395623955008, + "learning_rate": 2.570060924045704e-06, + "loss": 0.7129, + "step": 4721 + }, + { + "epoch": 0.4258465978265771, + "grad_norm": 1.7485773570942396, + "learning_rate": 2.569500923062304e-06, + "loss": 0.7932, + "step": 4722 + }, + { + "epoch": 0.42593678135004737, + "grad_norm": 1.6482396300822972, + "learning_rate": 2.5689408734859445e-06, + "loss": 0.6816, + "step": 4723 + }, + { + "epoch": 0.4260269648735176, + "grad_norm": 1.6205683948474807, + "learning_rate": 2.5683807753644127e-06, + "loss": 0.7099, + "step": 4724 + }, + { + "epoch": 0.4261171483969879, + "grad_norm": 1.6284350255619064, + "learning_rate": 2.5678206287454996e-06, + "loss": 0.7331, + "step": 4725 + }, + { + "epoch": 0.4262073319204581, + "grad_norm": 1.9078236947445857, + "learning_rate": 2.567260433676999e-06, + "loss": 0.7986, + "step": 4726 + }, + { + "epoch": 0.4262975154439284, + "grad_norm": 3.4831711951915607, + "learning_rate": 2.5667001902067107e-06, + "loss": 0.7341, + "step": 4727 + }, + { + "epoch": 0.42638769896739864, + "grad_norm": 1.7853001509415793, + "learning_rate": 2.566139898382437e-06, + "loss": 0.8143, + "step": 4728 + }, + { + "epoch": 0.42647788249086893, + "grad_norm": 3.173313359697244, + "learning_rate": 2.5655795582519853e-06, + "loss": 0.7766, + "step": 4729 + }, + { + "epoch": 0.42656806601433916, + "grad_norm": 1.5716655462725555, + "learning_rate": 2.565019169863168e-06, + "loss": 0.827, + "step": 4730 + }, + { + "epoch": 0.42665824953780945, + "grad_norm": 0.5874292331036345, + "learning_rate": 2.5644587332637994e-06, + "loss": 0.5483, + "step": 4731 + }, + { + "epoch": 0.4267484330612797, + "grad_norm": 1.8953857382565953, + "learning_rate": 2.5638982485016994e-06, + "loss": 0.7892, + "step": 4732 + }, + { + "epoch": 0.42683861658475, + "grad_norm": 0.7155511826336495, + "learning_rate": 2.5633377156246917e-06, + "loss": 0.553, + "step": 4733 + }, + { + "epoch": 0.4269288001082202, + "grad_norm": 2.0102948825561744, + "learning_rate": 2.562777134680603e-06, + "loss": 0.7447, + "step": 4734 + }, + { + "epoch": 0.4270189836316905, + "grad_norm": 1.9928727112243523, + "learning_rate": 2.562216505717267e-06, + "loss": 0.7742, + "step": 4735 + }, + { + "epoch": 0.42710916715516073, + "grad_norm": 2.1850917093335562, + "learning_rate": 2.561655828782518e-06, + "loss": 0.783, + "step": 4736 + }, + { + "epoch": 0.427199350678631, + "grad_norm": 2.250563656882791, + "learning_rate": 2.561095103924197e-06, + "loss": 0.7406, + "step": 4737 + }, + { + "epoch": 0.42728953420210125, + "grad_norm": 1.6857469424508957, + "learning_rate": 2.560534331190148e-06, + "loss": 0.7114, + "step": 4738 + }, + { + "epoch": 0.42737971772557154, + "grad_norm": 0.578678484755721, + "learning_rate": 2.559973510628218e-06, + "loss": 0.5494, + "step": 4739 + }, + { + "epoch": 0.4274699012490418, + "grad_norm": 1.7881470049477786, + "learning_rate": 2.5594126422862615e-06, + "loss": 0.7081, + "step": 4740 + }, + { + "epoch": 0.42756008477251206, + "grad_norm": 2.794236627649639, + "learning_rate": 2.558851726212134e-06, + "loss": 0.7113, + "step": 4741 + }, + { + "epoch": 0.42765026829598235, + "grad_norm": 3.1141729860664444, + "learning_rate": 2.5582907624536953e-06, + "loss": 0.7325, + "step": 4742 + }, + { + "epoch": 0.4277404518194526, + "grad_norm": 1.7682934067245812, + "learning_rate": 2.557729751058811e-06, + "loss": 0.7689, + "step": 4743 + }, + { + "epoch": 0.42783063534292287, + "grad_norm": 1.5678668359195533, + "learning_rate": 2.557168692075348e-06, + "loss": 0.7309, + "step": 4744 + }, + { + "epoch": 0.4279208188663931, + "grad_norm": 1.3957079456245027, + "learning_rate": 2.556607585551181e-06, + "loss": 0.7018, + "step": 4745 + }, + { + "epoch": 0.4280110023898634, + "grad_norm": 0.7220299971337872, + "learning_rate": 2.5560464315341844e-06, + "loss": 0.678, + "step": 4746 + }, + { + "epoch": 0.4281011859133336, + "grad_norm": 1.7126879512364612, + "learning_rate": 2.555485230072242e-06, + "loss": 0.7509, + "step": 4747 + }, + { + "epoch": 0.4281913694368039, + "grad_norm": 1.8624611895313083, + "learning_rate": 2.5549239812132354e-06, + "loss": 0.807, + "step": 4748 + }, + { + "epoch": 0.42828155296027415, + "grad_norm": 1.6493445337162451, + "learning_rate": 2.5543626850050556e-06, + "loss": 0.8195, + "step": 4749 + }, + { + "epoch": 0.42837173648374444, + "grad_norm": 1.4657324777981036, + "learning_rate": 2.5538013414955944e-06, + "loss": 0.8119, + "step": 4750 + }, + { + "epoch": 0.42846192000721467, + "grad_norm": 1.694533138177364, + "learning_rate": 2.5532399507327494e-06, + "loss": 0.7131, + "step": 4751 + }, + { + "epoch": 0.42855210353068496, + "grad_norm": 0.6368757417129824, + "learning_rate": 2.552678512764421e-06, + "loss": 0.587, + "step": 4752 + }, + { + "epoch": 0.4286422870541552, + "grad_norm": 1.993148093414256, + "learning_rate": 2.5521170276385147e-06, + "loss": 0.8138, + "step": 4753 + }, + { + "epoch": 0.4287324705776255, + "grad_norm": 2.3256122931379566, + "learning_rate": 2.5515554954029394e-06, + "loss": 0.8143, + "step": 4754 + }, + { + "epoch": 0.4288226541010957, + "grad_norm": 1.820675269394428, + "learning_rate": 2.550993916105608e-06, + "loss": 0.7913, + "step": 4755 + }, + { + "epoch": 0.428912837624566, + "grad_norm": 1.473412600293735, + "learning_rate": 2.550432289794437e-06, + "loss": 0.7141, + "step": 4756 + }, + { + "epoch": 0.42900302114803623, + "grad_norm": 1.8497701734114833, + "learning_rate": 2.5498706165173483e-06, + "loss": 0.768, + "step": 4757 + }, + { + "epoch": 0.4290932046715065, + "grad_norm": 1.323311328841118, + "learning_rate": 2.5493088963222668e-06, + "loss": 0.7327, + "step": 4758 + }, + { + "epoch": 0.42918338819497676, + "grad_norm": 1.8032336303145626, + "learning_rate": 2.548747129257121e-06, + "loss": 0.8086, + "step": 4759 + }, + { + "epoch": 0.42927357171844704, + "grad_norm": 1.903779476987046, + "learning_rate": 2.548185315369845e-06, + "loss": 0.6778, + "step": 4760 + }, + { + "epoch": 0.4293637552419173, + "grad_norm": 1.4974035998156832, + "learning_rate": 2.5476234547083746e-06, + "loss": 0.845, + "step": 4761 + }, + { + "epoch": 0.42945393876538757, + "grad_norm": 1.5937947285677359, + "learning_rate": 2.547061547320652e-06, + "loss": 0.8314, + "step": 4762 + }, + { + "epoch": 0.4295441222888578, + "grad_norm": 1.371695275145484, + "learning_rate": 2.5464995932546217e-06, + "loss": 0.7408, + "step": 4763 + }, + { + "epoch": 0.4296343058123281, + "grad_norm": 2.078206340373873, + "learning_rate": 2.545937592558232e-06, + "loss": 0.7136, + "step": 4764 + }, + { + "epoch": 0.4297244893357984, + "grad_norm": 1.5109669789985967, + "learning_rate": 2.5453755452794374e-06, + "loss": 0.7076, + "step": 4765 + }, + { + "epoch": 0.4298146728592686, + "grad_norm": 1.9748774647424787, + "learning_rate": 2.5448134514661938e-06, + "loss": 0.7997, + "step": 4766 + }, + { + "epoch": 0.4299048563827389, + "grad_norm": 1.4000405016217694, + "learning_rate": 2.5442513111664623e-06, + "loss": 0.7558, + "step": 4767 + }, + { + "epoch": 0.42999503990620913, + "grad_norm": 2.012071182341733, + "learning_rate": 2.5436891244282084e-06, + "loss": 0.7591, + "step": 4768 + }, + { + "epoch": 0.4300852234296794, + "grad_norm": 1.703439246068132, + "learning_rate": 2.5431268912994004e-06, + "loss": 0.8534, + "step": 4769 + }, + { + "epoch": 0.43017540695314965, + "grad_norm": 1.6231701157990182, + "learning_rate": 2.5425646118280108e-06, + "loss": 0.7278, + "step": 4770 + }, + { + "epoch": 0.43026559047661994, + "grad_norm": 0.5588498628799613, + "learning_rate": 2.5420022860620172e-06, + "loss": 0.5628, + "step": 4771 + }, + { + "epoch": 0.4303557740000902, + "grad_norm": 2.1998905761555747, + "learning_rate": 2.5414399140493995e-06, + "loss": 0.6871, + "step": 4772 + }, + { + "epoch": 0.43044595752356046, + "grad_norm": 2.1411158980350313, + "learning_rate": 2.5408774958381436e-06, + "loss": 0.6954, + "step": 4773 + }, + { + "epoch": 0.4305361410470307, + "grad_norm": 2.0113987344070825, + "learning_rate": 2.540315031476237e-06, + "loss": 0.7937, + "step": 4774 + }, + { + "epoch": 0.430626324570501, + "grad_norm": 1.5860715912755536, + "learning_rate": 2.5397525210116737e-06, + "loss": 0.7643, + "step": 4775 + }, + { + "epoch": 0.4307165080939712, + "grad_norm": 1.5220411895311488, + "learning_rate": 2.539189964492448e-06, + "loss": 0.7123, + "step": 4776 + }, + { + "epoch": 0.4308066916174415, + "grad_norm": 1.5886893615435616, + "learning_rate": 2.5386273619665613e-06, + "loss": 0.798, + "step": 4777 + }, + { + "epoch": 0.43089687514091174, + "grad_norm": 0.6840446075288248, + "learning_rate": 2.5380647134820186e-06, + "loss": 0.5838, + "step": 4778 + }, + { + "epoch": 0.43098705866438203, + "grad_norm": 1.4293651947098833, + "learning_rate": 2.5375020190868277e-06, + "loss": 0.8133, + "step": 4779 + }, + { + "epoch": 0.43107724218785226, + "grad_norm": 1.8089411424558586, + "learning_rate": 2.536939278829001e-06, + "loss": 0.7511, + "step": 4780 + }, + { + "epoch": 0.43116742571132255, + "grad_norm": 1.7719144625028482, + "learning_rate": 2.5363764927565536e-06, + "loss": 0.8153, + "step": 4781 + }, + { + "epoch": 0.4312576092347928, + "grad_norm": 2.5982721272670353, + "learning_rate": 2.5358136609175064e-06, + "loss": 0.7177, + "step": 4782 + }, + { + "epoch": 0.43134779275826307, + "grad_norm": 1.4496201116215166, + "learning_rate": 2.535250783359884e-06, + "loss": 0.7574, + "step": 4783 + }, + { + "epoch": 0.4314379762817333, + "grad_norm": 1.3683118719926206, + "learning_rate": 2.5346878601317124e-06, + "loss": 0.6495, + "step": 4784 + }, + { + "epoch": 0.4315281598052036, + "grad_norm": 0.5813197452990623, + "learning_rate": 2.534124891281025e-06, + "loss": 0.5476, + "step": 4785 + }, + { + "epoch": 0.4316183433286738, + "grad_norm": 2.19438265728246, + "learning_rate": 2.533561876855857e-06, + "loss": 0.7731, + "step": 4786 + }, + { + "epoch": 0.4317085268521441, + "grad_norm": 11.382800435391685, + "learning_rate": 2.532998816904247e-06, + "loss": 0.7098, + "step": 4787 + }, + { + "epoch": 0.4317987103756144, + "grad_norm": 1.6383750382965683, + "learning_rate": 2.53243571147424e-06, + "loss": 0.6537, + "step": 4788 + }, + { + "epoch": 0.43188889389908464, + "grad_norm": 2.0748309652579677, + "learning_rate": 2.5318725606138815e-06, + "loss": 0.6719, + "step": 4789 + }, + { + "epoch": 0.4319790774225549, + "grad_norm": 1.6463955349558572, + "learning_rate": 2.5313093643712235e-06, + "loss": 0.7617, + "step": 4790 + }, + { + "epoch": 0.43206926094602516, + "grad_norm": 1.5301002545639693, + "learning_rate": 2.530746122794321e-06, + "loss": 0.8024, + "step": 4791 + }, + { + "epoch": 0.43215944446949545, + "grad_norm": 1.474306916803103, + "learning_rate": 2.5301828359312323e-06, + "loss": 0.6873, + "step": 4792 + }, + { + "epoch": 0.4322496279929657, + "grad_norm": 1.7584450494913648, + "learning_rate": 2.529619503830021e-06, + "loss": 0.7121, + "step": 4793 + }, + { + "epoch": 0.43233981151643597, + "grad_norm": 3.1113455964081016, + "learning_rate": 2.529056126538753e-06, + "loss": 0.6924, + "step": 4794 + }, + { + "epoch": 0.4324299950399062, + "grad_norm": 2.8332670054432394, + "learning_rate": 2.5284927041054995e-06, + "loss": 0.7888, + "step": 4795 + }, + { + "epoch": 0.4325201785633765, + "grad_norm": 0.6147225567081831, + "learning_rate": 2.5279292365783348e-06, + "loss": 0.5251, + "step": 4796 + }, + { + "epoch": 0.4326103620868467, + "grad_norm": 1.6738056194701998, + "learning_rate": 2.527365724005336e-06, + "loss": 0.849, + "step": 4797 + }, + { + "epoch": 0.432700545610317, + "grad_norm": 1.7003391825077987, + "learning_rate": 2.526802166434586e-06, + "loss": 0.8099, + "step": 4798 + }, + { + "epoch": 0.43279072913378724, + "grad_norm": 1.4794362677654076, + "learning_rate": 2.5262385639141708e-06, + "loss": 0.7246, + "step": 4799 + }, + { + "epoch": 0.43288091265725753, + "grad_norm": 1.6221803263838341, + "learning_rate": 2.525674916492179e-06, + "loss": 0.828, + "step": 4800 + }, + { + "epoch": 0.43297109618072777, + "grad_norm": 1.594143239294674, + "learning_rate": 2.5251112242167056e-06, + "loss": 0.7621, + "step": 4801 + }, + { + "epoch": 0.43306127970419805, + "grad_norm": 1.631701954170072, + "learning_rate": 2.5245474871358464e-06, + "loss": 0.7155, + "step": 4802 + }, + { + "epoch": 0.4331514632276683, + "grad_norm": 1.866648193109017, + "learning_rate": 2.5239837052977037e-06, + "loss": 0.6741, + "step": 4803 + }, + { + "epoch": 0.4332416467511386, + "grad_norm": 1.7490235738152515, + "learning_rate": 2.523419878750381e-06, + "loss": 0.8647, + "step": 4804 + }, + { + "epoch": 0.4333318302746088, + "grad_norm": 2.0523009854476473, + "learning_rate": 2.522856007541989e-06, + "loss": 0.7118, + "step": 4805 + }, + { + "epoch": 0.4334220137980791, + "grad_norm": 1.8123793726275765, + "learning_rate": 2.5222920917206397e-06, + "loss": 0.6651, + "step": 4806 + }, + { + "epoch": 0.43351219732154933, + "grad_norm": 0.6078948359876132, + "learning_rate": 2.5217281313344493e-06, + "loss": 0.5626, + "step": 4807 + }, + { + "epoch": 0.4336023808450196, + "grad_norm": 1.7215564356939614, + "learning_rate": 2.5211641264315372e-06, + "loss": 0.8007, + "step": 4808 + }, + { + "epoch": 0.43369256436848985, + "grad_norm": 1.6571721083192255, + "learning_rate": 2.5206000770600286e-06, + "loss": 0.7352, + "step": 4809 + }, + { + "epoch": 0.43378274789196014, + "grad_norm": 3.735241649730947, + "learning_rate": 2.520035983268051e-06, + "loss": 0.7134, + "step": 4810 + }, + { + "epoch": 0.4338729314154304, + "grad_norm": 4.057817548264693, + "learning_rate": 2.5194718451037357e-06, + "loss": 0.7566, + "step": 4811 + }, + { + "epoch": 0.43396311493890066, + "grad_norm": 3.0227258125557177, + "learning_rate": 2.518907662615218e-06, + "loss": 0.7825, + "step": 4812 + }, + { + "epoch": 0.43405329846237095, + "grad_norm": 1.9216424444656577, + "learning_rate": 2.5183434358506373e-06, + "loss": 0.7205, + "step": 4813 + }, + { + "epoch": 0.4341434819858412, + "grad_norm": 1.6875649898871372, + "learning_rate": 2.5177791648581368e-06, + "loss": 0.7363, + "step": 4814 + }, + { + "epoch": 0.4342336655093115, + "grad_norm": 1.5410757306927205, + "learning_rate": 2.517214849685863e-06, + "loss": 0.7577, + "step": 4815 + }, + { + "epoch": 0.4343238490327817, + "grad_norm": 1.5636737229781579, + "learning_rate": 2.5166504903819663e-06, + "loss": 0.7739, + "step": 4816 + }, + { + "epoch": 0.434414032556252, + "grad_norm": 2.299099007643699, + "learning_rate": 2.5160860869946014e-06, + "loss": 0.823, + "step": 4817 + }, + { + "epoch": 0.4345042160797222, + "grad_norm": 1.4793050011793654, + "learning_rate": 2.5155216395719253e-06, + "loss": 0.8399, + "step": 4818 + }, + { + "epoch": 0.4345943996031925, + "grad_norm": 1.574677961214233, + "learning_rate": 2.5149571481621e-06, + "loss": 0.7164, + "step": 4819 + }, + { + "epoch": 0.43468458312666275, + "grad_norm": 0.596887405626087, + "learning_rate": 2.514392612813292e-06, + "loss": 0.5455, + "step": 4820 + }, + { + "epoch": 0.43477476665013304, + "grad_norm": 1.7034548518784716, + "learning_rate": 2.5138280335736695e-06, + "loss": 0.8143, + "step": 4821 + }, + { + "epoch": 0.43486495017360327, + "grad_norm": 1.8018892731301388, + "learning_rate": 2.5132634104914064e-06, + "loss": 0.8382, + "step": 4822 + }, + { + "epoch": 0.43495513369707356, + "grad_norm": 1.6202558075303841, + "learning_rate": 2.5126987436146794e-06, + "loss": 0.7686, + "step": 4823 + }, + { + "epoch": 0.4350453172205438, + "grad_norm": 1.7199499623095176, + "learning_rate": 2.5121340329916675e-06, + "loss": 0.7392, + "step": 4824 + }, + { + "epoch": 0.4351355007440141, + "grad_norm": 1.5407080978051007, + "learning_rate": 2.5115692786705566e-06, + "loss": 0.6751, + "step": 4825 + }, + { + "epoch": 0.4352256842674843, + "grad_norm": 1.8663535071130506, + "learning_rate": 2.511004480699534e-06, + "loss": 0.7931, + "step": 4826 + }, + { + "epoch": 0.4353158677909546, + "grad_norm": 0.7169196416097047, + "learning_rate": 2.510439639126791e-06, + "loss": 0.6195, + "step": 4827 + }, + { + "epoch": 0.43540605131442484, + "grad_norm": 1.47843410816501, + "learning_rate": 2.509874754000524e-06, + "loss": 0.7555, + "step": 4828 + }, + { + "epoch": 0.4354962348378951, + "grad_norm": 2.354636480212811, + "learning_rate": 2.509309825368932e-06, + "loss": 0.7487, + "step": 4829 + }, + { + "epoch": 0.43558641836136536, + "grad_norm": 1.6213979412688317, + "learning_rate": 2.5087448532802173e-06, + "loss": 0.759, + "step": 4830 + }, + { + "epoch": 0.43567660188483565, + "grad_norm": 3.826184960616023, + "learning_rate": 2.508179837782586e-06, + "loss": 0.7397, + "step": 4831 + }, + { + "epoch": 0.4357667854083059, + "grad_norm": 1.6666694958980706, + "learning_rate": 2.5076147789242493e-06, + "loss": 0.7318, + "step": 4832 + }, + { + "epoch": 0.43585696893177617, + "grad_norm": 3.8739137665264827, + "learning_rate": 2.5070496767534202e-06, + "loss": 0.7736, + "step": 4833 + }, + { + "epoch": 0.4359471524552464, + "grad_norm": 1.885435943785977, + "learning_rate": 2.506484531318317e-06, + "loss": 0.7878, + "step": 4834 + }, + { + "epoch": 0.4360373359787167, + "grad_norm": 0.9036706193644962, + "learning_rate": 2.5059193426671613e-06, + "loss": 0.5918, + "step": 4835 + }, + { + "epoch": 0.436127519502187, + "grad_norm": 2.2231664625091923, + "learning_rate": 2.5053541108481772e-06, + "loss": 0.8894, + "step": 4836 + }, + { + "epoch": 0.4362177030256572, + "grad_norm": 2.1657951387230185, + "learning_rate": 2.5047888359095935e-06, + "loss": 0.7409, + "step": 4837 + }, + { + "epoch": 0.4363078865491275, + "grad_norm": 1.5848011775421127, + "learning_rate": 2.5042235178996436e-06, + "loss": 0.7404, + "step": 4838 + }, + { + "epoch": 0.43639807007259773, + "grad_norm": 1.8229085068293005, + "learning_rate": 2.5036581568665627e-06, + "loss": 0.8195, + "step": 4839 + }, + { + "epoch": 0.436488253596068, + "grad_norm": 1.6527319918718786, + "learning_rate": 2.503092752858591e-06, + "loss": 0.8227, + "step": 4840 + }, + { + "epoch": 0.43657843711953825, + "grad_norm": 0.8324760358528752, + "learning_rate": 2.502527305923971e-06, + "loss": 0.6079, + "step": 4841 + }, + { + "epoch": 0.43666862064300854, + "grad_norm": 2.0034299288082864, + "learning_rate": 2.5019618161109506e-06, + "loss": 0.7514, + "step": 4842 + }, + { + "epoch": 0.4367588041664788, + "grad_norm": 1.6836513146562955, + "learning_rate": 2.5013962834677804e-06, + "loss": 0.7742, + "step": 4843 + }, + { + "epoch": 0.43684898768994906, + "grad_norm": 2.076210229758021, + "learning_rate": 2.500830708042715e-06, + "loss": 0.7461, + "step": 4844 + }, + { + "epoch": 0.4369391712134193, + "grad_norm": 1.5433925384689124, + "learning_rate": 2.500265089884011e-06, + "loss": 0.8184, + "step": 4845 + }, + { + "epoch": 0.4370293547368896, + "grad_norm": 1.8176923895048276, + "learning_rate": 2.499699429039932e-06, + "loss": 0.7593, + "step": 4846 + }, + { + "epoch": 0.4371195382603598, + "grad_norm": 1.9073729412513492, + "learning_rate": 2.4991337255587425e-06, + "loss": 0.7081, + "step": 4847 + }, + { + "epoch": 0.4372097217838301, + "grad_norm": 1.697389727621772, + "learning_rate": 2.4985679794887106e-06, + "loss": 0.6266, + "step": 4848 + }, + { + "epoch": 0.43729990530730034, + "grad_norm": 1.7357449319164686, + "learning_rate": 2.49800219087811e-06, + "loss": 0.7874, + "step": 4849 + }, + { + "epoch": 0.43739008883077063, + "grad_norm": 1.8487986607624718, + "learning_rate": 2.4974363597752163e-06, + "loss": 0.7916, + "step": 4850 + }, + { + "epoch": 0.43748027235424086, + "grad_norm": 2.122293206914568, + "learning_rate": 2.4968704862283097e-06, + "loss": 0.8053, + "step": 4851 + }, + { + "epoch": 0.43757045587771115, + "grad_norm": 1.8313000486649795, + "learning_rate": 2.4963045702856737e-06, + "loss": 0.8264, + "step": 4852 + }, + { + "epoch": 0.4376606394011814, + "grad_norm": 2.719483375292302, + "learning_rate": 2.4957386119955946e-06, + "loss": 0.6993, + "step": 4853 + }, + { + "epoch": 0.4377508229246517, + "grad_norm": 0.662600332436045, + "learning_rate": 2.495172611406364e-06, + "loss": 0.5688, + "step": 4854 + }, + { + "epoch": 0.4378410064481219, + "grad_norm": 2.0805654322083202, + "learning_rate": 2.4946065685662757e-06, + "loss": 0.7853, + "step": 4855 + }, + { + "epoch": 0.4379311899715922, + "grad_norm": 1.8537115831567927, + "learning_rate": 2.4940404835236283e-06, + "loss": 0.758, + "step": 4856 + }, + { + "epoch": 0.4380213734950624, + "grad_norm": 1.6872191725285934, + "learning_rate": 2.4934743563267223e-06, + "loss": 0.7501, + "step": 4857 + }, + { + "epoch": 0.4381115570185327, + "grad_norm": 3.486166044683052, + "learning_rate": 2.4929081870238635e-06, + "loss": 0.7216, + "step": 4858 + }, + { + "epoch": 0.43820174054200295, + "grad_norm": 1.4965800398895805, + "learning_rate": 2.49234197566336e-06, + "loss": 0.7938, + "step": 4859 + }, + { + "epoch": 0.43829192406547324, + "grad_norm": 1.571487290969711, + "learning_rate": 2.4917757222935247e-06, + "loss": 0.7774, + "step": 4860 + }, + { + "epoch": 0.4383821075889435, + "grad_norm": 1.7624155091940519, + "learning_rate": 2.4912094269626725e-06, + "loss": 0.7773, + "step": 4861 + }, + { + "epoch": 0.43847229111241376, + "grad_norm": 1.7998272071129127, + "learning_rate": 2.4906430897191245e-06, + "loss": 0.7949, + "step": 4862 + }, + { + "epoch": 0.43856247463588405, + "grad_norm": 1.6540186264068657, + "learning_rate": 2.490076710611202e-06, + "loss": 0.7943, + "step": 4863 + }, + { + "epoch": 0.4386526581593543, + "grad_norm": 3.7429192132285425, + "learning_rate": 2.4895102896872326e-06, + "loss": 0.8059, + "step": 4864 + }, + { + "epoch": 0.43874284168282457, + "grad_norm": 1.757587808447614, + "learning_rate": 2.4889438269955457e-06, + "loss": 0.6879, + "step": 4865 + }, + { + "epoch": 0.4388330252062948, + "grad_norm": 1.842895374397899, + "learning_rate": 2.4883773225844755e-06, + "loss": 0.7411, + "step": 4866 + }, + { + "epoch": 0.4389232087297651, + "grad_norm": 1.7881880043020808, + "learning_rate": 2.48781077650236e-06, + "loss": 0.7489, + "step": 4867 + }, + { + "epoch": 0.4390133922532353, + "grad_norm": 2.2100102397629695, + "learning_rate": 2.4872441887975386e-06, + "loss": 0.7193, + "step": 4868 + }, + { + "epoch": 0.4391035757767056, + "grad_norm": 1.3123564868758106, + "learning_rate": 2.486677559518356e-06, + "loss": 0.6958, + "step": 4869 + }, + { + "epoch": 0.43919375930017585, + "grad_norm": 1.7049329068807142, + "learning_rate": 2.4861108887131614e-06, + "loss": 0.7446, + "step": 4870 + }, + { + "epoch": 0.43928394282364613, + "grad_norm": 1.442302559574482, + "learning_rate": 2.485544176430305e-06, + "loss": 0.7255, + "step": 4871 + }, + { + "epoch": 0.43937412634711637, + "grad_norm": 2.404385128927054, + "learning_rate": 2.4849774227181425e-06, + "loss": 0.8366, + "step": 4872 + }, + { + "epoch": 0.43946430987058666, + "grad_norm": 1.4153511540379122, + "learning_rate": 2.484410627625032e-06, + "loss": 0.7521, + "step": 4873 + }, + { + "epoch": 0.4395544933940569, + "grad_norm": 1.889353490083021, + "learning_rate": 2.4838437911993356e-06, + "loss": 0.7955, + "step": 4874 + }, + { + "epoch": 0.4396446769175272, + "grad_norm": 1.7682038774961948, + "learning_rate": 2.483276913489419e-06, + "loss": 0.7926, + "step": 4875 + }, + { + "epoch": 0.4397348604409974, + "grad_norm": 2.0037715636084807, + "learning_rate": 2.4827099945436516e-06, + "loss": 0.656, + "step": 4876 + }, + { + "epoch": 0.4398250439644677, + "grad_norm": 2.1969019881390937, + "learning_rate": 2.482143034410405e-06, + "loss": 0.8365, + "step": 4877 + }, + { + "epoch": 0.43991522748793793, + "grad_norm": 1.8168805559333856, + "learning_rate": 2.4815760331380573e-06, + "loss": 0.7862, + "step": 4878 + }, + { + "epoch": 0.4400054110114082, + "grad_norm": 1.4628367598717584, + "learning_rate": 2.481008990774987e-06, + "loss": 0.7163, + "step": 4879 + }, + { + "epoch": 0.44009559453487845, + "grad_norm": 1.8740249641734272, + "learning_rate": 2.480441907369577e-06, + "loss": 0.7538, + "step": 4880 + }, + { + "epoch": 0.44018577805834874, + "grad_norm": 2.062579182347521, + "learning_rate": 2.479874782970214e-06, + "loss": 0.7009, + "step": 4881 + }, + { + "epoch": 0.440275961581819, + "grad_norm": 2.241239553364429, + "learning_rate": 2.4793076176252887e-06, + "loss": 0.7836, + "step": 4882 + }, + { + "epoch": 0.44036614510528926, + "grad_norm": 1.5839286822630638, + "learning_rate": 2.478740411383195e-06, + "loss": 0.6392, + "step": 4883 + }, + { + "epoch": 0.44045632862875955, + "grad_norm": 1.6281891852263923, + "learning_rate": 2.4781731642923296e-06, + "loss": 0.7504, + "step": 4884 + }, + { + "epoch": 0.4405465121522298, + "grad_norm": 1.5113764727784609, + "learning_rate": 2.477605876401093e-06, + "loss": 0.7403, + "step": 4885 + }, + { + "epoch": 0.4406366956757001, + "grad_norm": 1.542682785382435, + "learning_rate": 2.4770385477578894e-06, + "loss": 0.7064, + "step": 4886 + }, + { + "epoch": 0.4407268791991703, + "grad_norm": 1.3659292480107779, + "learning_rate": 2.476471178411127e-06, + "loss": 0.713, + "step": 4887 + }, + { + "epoch": 0.4408170627226406, + "grad_norm": 4.107317391235275, + "learning_rate": 2.475903768409216e-06, + "loss": 0.7794, + "step": 4888 + }, + { + "epoch": 0.44090724624611083, + "grad_norm": 4.035934447059555, + "learning_rate": 2.475336317800572e-06, + "loss": 0.7295, + "step": 4889 + }, + { + "epoch": 0.4409974297695811, + "grad_norm": 1.8601194542294013, + "learning_rate": 2.4747688266336118e-06, + "loss": 0.7438, + "step": 4890 + }, + { + "epoch": 0.44108761329305135, + "grad_norm": 2.2246561020388897, + "learning_rate": 2.4742012949567574e-06, + "loss": 0.8557, + "step": 4891 + }, + { + "epoch": 0.44117779681652164, + "grad_norm": 1.6920797579659521, + "learning_rate": 2.4736337228184338e-06, + "loss": 0.7329, + "step": 4892 + }, + { + "epoch": 0.44126798033999187, + "grad_norm": 1.7179016794085586, + "learning_rate": 2.4730661102670692e-06, + "loss": 0.8549, + "step": 4893 + }, + { + "epoch": 0.44135816386346216, + "grad_norm": 1.6444700687290725, + "learning_rate": 2.472498457351096e-06, + "loss": 0.7567, + "step": 4894 + }, + { + "epoch": 0.4414483473869324, + "grad_norm": 1.9860744258038654, + "learning_rate": 2.4719307641189495e-06, + "loss": 0.7446, + "step": 4895 + }, + { + "epoch": 0.4415385309104027, + "grad_norm": 2.1014592595604333, + "learning_rate": 2.4713630306190673e-06, + "loss": 0.7197, + "step": 4896 + }, + { + "epoch": 0.4416287144338729, + "grad_norm": 2.212260036063366, + "learning_rate": 2.4707952568998923e-06, + "loss": 0.8372, + "step": 4897 + }, + { + "epoch": 0.4417188979573432, + "grad_norm": 3.066204244192166, + "learning_rate": 2.4702274430098703e-06, + "loss": 0.7304, + "step": 4898 + }, + { + "epoch": 0.44180908148081344, + "grad_norm": 2.408591976158932, + "learning_rate": 2.4696595889974497e-06, + "loss": 0.7108, + "step": 4899 + }, + { + "epoch": 0.4418992650042837, + "grad_norm": 2.6795391786860527, + "learning_rate": 2.469091694911084e-06, + "loss": 0.8101, + "step": 4900 + }, + { + "epoch": 0.44198944852775396, + "grad_norm": 2.1020565090478, + "learning_rate": 2.4685237607992276e-06, + "loss": 0.7517, + "step": 4901 + }, + { + "epoch": 0.44207963205122425, + "grad_norm": 1.9575649607318857, + "learning_rate": 2.4679557867103416e-06, + "loss": 0.8189, + "step": 4902 + }, + { + "epoch": 0.4421698155746945, + "grad_norm": 1.5202567518624122, + "learning_rate": 2.4673877726928865e-06, + "loss": 0.7615, + "step": 4903 + }, + { + "epoch": 0.44225999909816477, + "grad_norm": 4.790100084824099, + "learning_rate": 2.46681971879533e-06, + "loss": 0.6552, + "step": 4904 + }, + { + "epoch": 0.442350182621635, + "grad_norm": 1.8403706373222475, + "learning_rate": 2.4662516250661407e-06, + "loss": 0.6869, + "step": 4905 + }, + { + "epoch": 0.4424403661451053, + "grad_norm": 2.22295690021613, + "learning_rate": 2.465683491553792e-06, + "loss": 0.727, + "step": 4906 + }, + { + "epoch": 0.4425305496685755, + "grad_norm": 1.734882589759389, + "learning_rate": 2.4651153183067604e-06, + "loss": 0.7631, + "step": 4907 + }, + { + "epoch": 0.4426207331920458, + "grad_norm": 2.17166949061427, + "learning_rate": 2.4645471053735245e-06, + "loss": 0.6953, + "step": 4908 + }, + { + "epoch": 0.4427109167155161, + "grad_norm": 2.0447347385943733, + "learning_rate": 2.4639788528025684e-06, + "loss": 0.6945, + "step": 4909 + }, + { + "epoch": 0.44280110023898633, + "grad_norm": 1.3711475242220426, + "learning_rate": 2.463410560642378e-06, + "loss": 0.7597, + "step": 4910 + }, + { + "epoch": 0.4428912837624566, + "grad_norm": 0.6348510449962333, + "learning_rate": 2.4628422289414448e-06, + "loss": 0.5883, + "step": 4911 + }, + { + "epoch": 0.44298146728592686, + "grad_norm": 0.670764337766765, + "learning_rate": 2.4622738577482592e-06, + "loss": 0.564, + "step": 4912 + }, + { + "epoch": 0.44307165080939714, + "grad_norm": 2.107955002545151, + "learning_rate": 2.461705447111319e-06, + "loss": 0.7404, + "step": 4913 + }, + { + "epoch": 0.4431618343328674, + "grad_norm": 1.612989865790906, + "learning_rate": 2.4611369970791246e-06, + "loss": 0.7274, + "step": 4914 + }, + { + "epoch": 0.44325201785633767, + "grad_norm": 1.9713319468957238, + "learning_rate": 2.460568507700179e-06, + "loss": 0.7584, + "step": 4915 + }, + { + "epoch": 0.4433422013798079, + "grad_norm": 2.1392531628341804, + "learning_rate": 2.4599999790229887e-06, + "loss": 0.798, + "step": 4916 + }, + { + "epoch": 0.4434323849032782, + "grad_norm": 2.032795479517763, + "learning_rate": 2.459431411096064e-06, + "loss": 0.7227, + "step": 4917 + }, + { + "epoch": 0.4435225684267484, + "grad_norm": 1.854759193069581, + "learning_rate": 2.458862803967918e-06, + "loss": 0.726, + "step": 4918 + }, + { + "epoch": 0.4436127519502187, + "grad_norm": 3.9377303283172918, + "learning_rate": 2.4582941576870667e-06, + "loss": 0.7692, + "step": 4919 + }, + { + "epoch": 0.44370293547368894, + "grad_norm": 2.5967233000697716, + "learning_rate": 2.4577254723020315e-06, + "loss": 0.7347, + "step": 4920 + }, + { + "epoch": 0.44379311899715923, + "grad_norm": 1.559876948663164, + "learning_rate": 2.457156747861335e-06, + "loss": 0.768, + "step": 4921 + }, + { + "epoch": 0.44388330252062946, + "grad_norm": 2.916406093038437, + "learning_rate": 2.456587984413504e-06, + "loss": 0.8214, + "step": 4922 + }, + { + "epoch": 0.44397348604409975, + "grad_norm": 1.957611667960579, + "learning_rate": 2.4560191820070683e-06, + "loss": 0.6909, + "step": 4923 + }, + { + "epoch": 0.44406366956757, + "grad_norm": 2.7033939117561765, + "learning_rate": 2.4554503406905617e-06, + "loss": 0.7403, + "step": 4924 + }, + { + "epoch": 0.4441538530910403, + "grad_norm": 1.676036966371396, + "learning_rate": 2.454881460512521e-06, + "loss": 0.7987, + "step": 4925 + }, + { + "epoch": 0.4442440366145105, + "grad_norm": 1.337950366837342, + "learning_rate": 2.4543125415214856e-06, + "loss": 0.728, + "step": 4926 + }, + { + "epoch": 0.4443342201379808, + "grad_norm": 2.120878935837625, + "learning_rate": 2.4537435837659996e-06, + "loss": 0.7813, + "step": 4927 + }, + { + "epoch": 0.44442440366145103, + "grad_norm": 2.0427908883797823, + "learning_rate": 2.4531745872946085e-06, + "loss": 0.7065, + "step": 4928 + }, + { + "epoch": 0.4445145871849213, + "grad_norm": 2.08112480084846, + "learning_rate": 2.4526055521558632e-06, + "loss": 0.8173, + "step": 4929 + }, + { + "epoch": 0.44460477070839155, + "grad_norm": 2.441476366180646, + "learning_rate": 2.4520364783983164e-06, + "loss": 0.7773, + "step": 4930 + }, + { + "epoch": 0.44469495423186184, + "grad_norm": 1.5227687052365593, + "learning_rate": 2.451467366070525e-06, + "loss": 0.7389, + "step": 4931 + }, + { + "epoch": 0.4447851377553321, + "grad_norm": 2.4860000573708367, + "learning_rate": 2.450898215221048e-06, + "loss": 0.8317, + "step": 4932 + }, + { + "epoch": 0.44487532127880236, + "grad_norm": 2.2438416443707, + "learning_rate": 2.4503290258984498e-06, + "loss": 0.9011, + "step": 4933 + }, + { + "epoch": 0.44496550480227265, + "grad_norm": 3.0152483294752557, + "learning_rate": 2.4497597981512952e-06, + "loss": 0.6166, + "step": 4934 + }, + { + "epoch": 0.4450556883257429, + "grad_norm": 0.6462180519692389, + "learning_rate": 2.4491905320281555e-06, + "loss": 0.5919, + "step": 4935 + }, + { + "epoch": 0.44514587184921317, + "grad_norm": 2.1107888463760642, + "learning_rate": 2.448621227577602e-06, + "loss": 0.7343, + "step": 4936 + }, + { + "epoch": 0.4452360553726834, + "grad_norm": 2.764162528221138, + "learning_rate": 2.4480518848482123e-06, + "loss": 0.7013, + "step": 4937 + }, + { + "epoch": 0.4453262388961537, + "grad_norm": 1.7101881223786963, + "learning_rate": 2.447482503888565e-06, + "loss": 0.7669, + "step": 4938 + }, + { + "epoch": 0.4454164224196239, + "grad_norm": 2.0700442734189997, + "learning_rate": 2.4469130847472434e-06, + "loss": 0.7366, + "step": 4939 + }, + { + "epoch": 0.4455066059430942, + "grad_norm": 1.755168299308039, + "learning_rate": 2.4463436274728326e-06, + "loss": 0.7946, + "step": 4940 + }, + { + "epoch": 0.44559678946656445, + "grad_norm": 2.157526716329892, + "learning_rate": 2.4457741321139227e-06, + "loss": 0.6027, + "step": 4941 + }, + { + "epoch": 0.44568697299003474, + "grad_norm": 1.6543344180408257, + "learning_rate": 2.4452045987191063e-06, + "loss": 0.7771, + "step": 4942 + }, + { + "epoch": 0.44577715651350497, + "grad_norm": 1.5725565955040322, + "learning_rate": 2.4446350273369776e-06, + "loss": 0.7309, + "step": 4943 + }, + { + "epoch": 0.44586734003697526, + "grad_norm": 1.5482008675260766, + "learning_rate": 2.4440654180161374e-06, + "loss": 0.7893, + "step": 4944 + }, + { + "epoch": 0.4459575235604455, + "grad_norm": 1.5819757145682625, + "learning_rate": 2.4434957708051875e-06, + "loss": 0.7125, + "step": 4945 + }, + { + "epoch": 0.4460477070839158, + "grad_norm": 1.81031200280521, + "learning_rate": 2.4429260857527324e-06, + "loss": 0.6942, + "step": 4946 + }, + { + "epoch": 0.446137890607386, + "grad_norm": 1.9967620387833818, + "learning_rate": 2.4423563629073815e-06, + "loss": 0.6626, + "step": 4947 + }, + { + "epoch": 0.4462280741308563, + "grad_norm": 1.4502840881918282, + "learning_rate": 2.4417866023177466e-06, + "loss": 0.7554, + "step": 4948 + }, + { + "epoch": 0.44631825765432653, + "grad_norm": 3.196446597572882, + "learning_rate": 2.441216804032443e-06, + "loss": 0.7893, + "step": 4949 + }, + { + "epoch": 0.4464084411777968, + "grad_norm": 1.6726275737786267, + "learning_rate": 2.440646968100089e-06, + "loss": 0.77, + "step": 4950 + }, + { + "epoch": 0.44649862470126706, + "grad_norm": 0.6779135431484611, + "learning_rate": 2.4400770945693055e-06, + "loss": 0.5746, + "step": 4951 + }, + { + "epoch": 0.44658880822473734, + "grad_norm": 2.074190588578543, + "learning_rate": 2.4395071834887177e-06, + "loss": 0.7921, + "step": 4952 + }, + { + "epoch": 0.4466789917482076, + "grad_norm": 0.6985695878306072, + "learning_rate": 2.438937234906954e-06, + "loss": 0.6356, + "step": 4953 + }, + { + "epoch": 0.44676917527167787, + "grad_norm": 1.8176374303184017, + "learning_rate": 2.4383672488726447e-06, + "loss": 0.7477, + "step": 4954 + }, + { + "epoch": 0.44685935879514815, + "grad_norm": 2.104604908342881, + "learning_rate": 2.4377972254344256e-06, + "loss": 0.681, + "step": 4955 + }, + { + "epoch": 0.4469495423186184, + "grad_norm": 1.8874274713589667, + "learning_rate": 2.437227164640932e-06, + "loss": 0.6411, + "step": 4956 + }, + { + "epoch": 0.4470397258420887, + "grad_norm": 3.576903288717885, + "learning_rate": 2.436657066540807e-06, + "loss": 0.7019, + "step": 4957 + }, + { + "epoch": 0.4471299093655589, + "grad_norm": 11.704205818332598, + "learning_rate": 2.4360869311826927e-06, + "loss": 0.6335, + "step": 4958 + }, + { + "epoch": 0.4472200928890292, + "grad_norm": 3.6239898852781907, + "learning_rate": 2.4355167586152367e-06, + "loss": 0.849, + "step": 4959 + }, + { + "epoch": 0.44731027641249943, + "grad_norm": 2.110827249828698, + "learning_rate": 2.4349465488870896e-06, + "loss": 0.7222, + "step": 4960 + }, + { + "epoch": 0.4474004599359697, + "grad_norm": 2.3084659960619662, + "learning_rate": 2.434376302046905e-06, + "loss": 0.8412, + "step": 4961 + }, + { + "epoch": 0.44749064345943995, + "grad_norm": 1.6662071627743038, + "learning_rate": 2.433806018143339e-06, + "loss": 0.7764, + "step": 4962 + }, + { + "epoch": 0.44758082698291024, + "grad_norm": 0.6751617723454089, + "learning_rate": 2.433235697225051e-06, + "loss": 0.6162, + "step": 4963 + }, + { + "epoch": 0.4476710105063805, + "grad_norm": 1.8256739299607168, + "learning_rate": 2.4326653393407048e-06, + "loss": 0.7741, + "step": 4964 + }, + { + "epoch": 0.44776119402985076, + "grad_norm": 2.73404286138484, + "learning_rate": 2.432094944538966e-06, + "loss": 0.7195, + "step": 4965 + }, + { + "epoch": 0.447851377553321, + "grad_norm": 1.6408070054597066, + "learning_rate": 2.4315245128685047e-06, + "loss": 0.7235, + "step": 4966 + }, + { + "epoch": 0.4479415610767913, + "grad_norm": 1.8312146414510166, + "learning_rate": 2.4309540443779925e-06, + "loss": 0.755, + "step": 4967 + }, + { + "epoch": 0.4480317446002615, + "grad_norm": 1.5488470588671674, + "learning_rate": 2.4303835391161047e-06, + "loss": 0.6902, + "step": 4968 + }, + { + "epoch": 0.4481219281237318, + "grad_norm": 1.676674984645245, + "learning_rate": 2.42981299713152e-06, + "loss": 0.7064, + "step": 4969 + }, + { + "epoch": 0.44821211164720204, + "grad_norm": 2.706274766346151, + "learning_rate": 2.4292424184729204e-06, + "loss": 0.79, + "step": 4970 + }, + { + "epoch": 0.4483022951706723, + "grad_norm": 3.2089134946809046, + "learning_rate": 2.4286718031889913e-06, + "loss": 0.7457, + "step": 4971 + }, + { + "epoch": 0.44839247869414256, + "grad_norm": 2.0493848715146252, + "learning_rate": 2.4281011513284202e-06, + "loss": 0.727, + "step": 4972 + }, + { + "epoch": 0.44848266221761285, + "grad_norm": 2.70848346196295, + "learning_rate": 2.4275304629398985e-06, + "loss": 0.6909, + "step": 4973 + }, + { + "epoch": 0.4485728457410831, + "grad_norm": 1.433694279859272, + "learning_rate": 2.4269597380721194e-06, + "loss": 0.8325, + "step": 4974 + }, + { + "epoch": 0.44866302926455337, + "grad_norm": 2.0224566936218147, + "learning_rate": 2.426388976773782e-06, + "loss": 0.7535, + "step": 4975 + }, + { + "epoch": 0.4487532127880236, + "grad_norm": 1.7317373478073317, + "learning_rate": 2.425818179093586e-06, + "loss": 0.8632, + "step": 4976 + }, + { + "epoch": 0.4488433963114939, + "grad_norm": 1.5377676692041171, + "learning_rate": 2.4252473450802346e-06, + "loss": 0.6641, + "step": 4977 + }, + { + "epoch": 0.4489335798349641, + "grad_norm": 4.670661896353538, + "learning_rate": 2.4246764747824355e-06, + "loss": 0.7907, + "step": 4978 + }, + { + "epoch": 0.4490237633584344, + "grad_norm": 1.5932706224132538, + "learning_rate": 2.424105568248897e-06, + "loss": 0.7652, + "step": 4979 + }, + { + "epoch": 0.4491139468819047, + "grad_norm": 1.816482575156326, + "learning_rate": 2.4235346255283337e-06, + "loss": 0.6939, + "step": 4980 + }, + { + "epoch": 0.44920413040537494, + "grad_norm": 1.830805647126603, + "learning_rate": 2.42296364666946e-06, + "loss": 0.7648, + "step": 4981 + }, + { + "epoch": 0.4492943139288452, + "grad_norm": 5.7308481789006915, + "learning_rate": 2.4223926317209965e-06, + "loss": 0.7047, + "step": 4982 + }, + { + "epoch": 0.44938449745231546, + "grad_norm": 1.5986320668045757, + "learning_rate": 2.4218215807316647e-06, + "loss": 0.7549, + "step": 4983 + }, + { + "epoch": 0.44947468097578575, + "grad_norm": 2.058227738157063, + "learning_rate": 2.4212504937501894e-06, + "loss": 0.8427, + "step": 4984 + }, + { + "epoch": 0.449564864499256, + "grad_norm": 1.4677424119828892, + "learning_rate": 2.4206793708253e-06, + "loss": 0.8108, + "step": 4985 + }, + { + "epoch": 0.44965504802272627, + "grad_norm": 1.9340324135463576, + "learning_rate": 2.420108212005726e-06, + "loss": 0.7118, + "step": 4986 + }, + { + "epoch": 0.4497452315461965, + "grad_norm": 2.0114354795253706, + "learning_rate": 2.4195370173402034e-06, + "loss": 0.7491, + "step": 4987 + }, + { + "epoch": 0.4498354150696668, + "grad_norm": 2.560543057390127, + "learning_rate": 2.4189657868774696e-06, + "loss": 0.8722, + "step": 4988 + }, + { + "epoch": 0.449925598593137, + "grad_norm": 1.809497846211577, + "learning_rate": 2.418394520666264e-06, + "loss": 0.6918, + "step": 4989 + }, + { + "epoch": 0.4500157821166073, + "grad_norm": 1.9767975799326516, + "learning_rate": 2.4178232187553307e-06, + "loss": 0.74, + "step": 4990 + }, + { + "epoch": 0.45010596564007754, + "grad_norm": 1.5057593880198739, + "learning_rate": 2.417251881193417e-06, + "loss": 0.7792, + "step": 4991 + }, + { + "epoch": 0.45019614916354783, + "grad_norm": 2.9146758960334607, + "learning_rate": 2.4166805080292723e-06, + "loss": 0.8099, + "step": 4992 + }, + { + "epoch": 0.45028633268701806, + "grad_norm": 3.7861626608879537, + "learning_rate": 2.4161090993116485e-06, + "loss": 0.702, + "step": 4993 + }, + { + "epoch": 0.45037651621048835, + "grad_norm": 1.9782324811472343, + "learning_rate": 2.4155376550893026e-06, + "loss": 0.6347, + "step": 4994 + }, + { + "epoch": 0.4504666997339586, + "grad_norm": 2.6063053555178777, + "learning_rate": 2.4149661754109926e-06, + "loss": 0.7627, + "step": 4995 + }, + { + "epoch": 0.4505568832574289, + "grad_norm": 2.034700602596316, + "learning_rate": 2.41439466032548e-06, + "loss": 0.8098, + "step": 4996 + }, + { + "epoch": 0.4506470667808991, + "grad_norm": 1.7246208492947142, + "learning_rate": 2.41382310988153e-06, + "loss": 0.7033, + "step": 4997 + }, + { + "epoch": 0.4507372503043694, + "grad_norm": 2.612269640535931, + "learning_rate": 2.413251524127911e-06, + "loss": 0.8149, + "step": 4998 + }, + { + "epoch": 0.45082743382783963, + "grad_norm": 1.8607822510398448, + "learning_rate": 2.412679903113393e-06, + "loss": 0.7274, + "step": 4999 + }, + { + "epoch": 0.4509176173513099, + "grad_norm": 2.140006539878478, + "learning_rate": 2.4121082468867505e-06, + "loss": 0.7715, + "step": 5000 + }, + { + "epoch": 0.45100780087478015, + "grad_norm": 2.5658991528954505, + "learning_rate": 2.4115365554967597e-06, + "loss": 0.6946, + "step": 5001 + }, + { + "epoch": 0.45109798439825044, + "grad_norm": 1.9334555836644647, + "learning_rate": 2.4109648289922006e-06, + "loss": 0.7569, + "step": 5002 + }, + { + "epoch": 0.45118816792172073, + "grad_norm": 3.6148998697218757, + "learning_rate": 2.4103930674218565e-06, + "loss": 0.7599, + "step": 5003 + }, + { + "epoch": 0.45127835144519096, + "grad_norm": 1.5230615396471168, + "learning_rate": 2.409821270834513e-06, + "loss": 0.7287, + "step": 5004 + }, + { + "epoch": 0.45136853496866125, + "grad_norm": 2.5578560967012756, + "learning_rate": 2.409249439278959e-06, + "loss": 0.8098, + "step": 5005 + }, + { + "epoch": 0.4514587184921315, + "grad_norm": 0.6296396185226986, + "learning_rate": 2.408677572803986e-06, + "loss": 0.553, + "step": 5006 + }, + { + "epoch": 0.45154890201560177, + "grad_norm": 1.7052113763886527, + "learning_rate": 2.408105671458389e-06, + "loss": 0.7627, + "step": 5007 + }, + { + "epoch": 0.451639085539072, + "grad_norm": 3.9267496764946177, + "learning_rate": 2.4075337352909663e-06, + "loss": 0.6963, + "step": 5008 + }, + { + "epoch": 0.4517292690625423, + "grad_norm": 1.7257983557845022, + "learning_rate": 2.4069617643505177e-06, + "loss": 0.8815, + "step": 5009 + }, + { + "epoch": 0.4518194525860125, + "grad_norm": 1.5163605468403338, + "learning_rate": 2.406389758685848e-06, + "loss": 0.8425, + "step": 5010 + }, + { + "epoch": 0.4519096361094828, + "grad_norm": 1.565210585274158, + "learning_rate": 2.405817718345763e-06, + "loss": 0.7929, + "step": 5011 + }, + { + "epoch": 0.45199981963295305, + "grad_norm": 1.9633487299696997, + "learning_rate": 2.4052456433790726e-06, + "loss": 0.7105, + "step": 5012 + }, + { + "epoch": 0.45209000315642334, + "grad_norm": 1.6737625305071395, + "learning_rate": 2.4046735338345897e-06, + "loss": 0.8115, + "step": 5013 + }, + { + "epoch": 0.45218018667989357, + "grad_norm": 1.9328583523022833, + "learning_rate": 2.404101389761129e-06, + "loss": 0.7692, + "step": 5014 + }, + { + "epoch": 0.45227037020336386, + "grad_norm": 0.6595097824394968, + "learning_rate": 2.4035292112075097e-06, + "loss": 0.5635, + "step": 5015 + }, + { + "epoch": 0.4523605537268341, + "grad_norm": 1.613960400257467, + "learning_rate": 2.4029569982225534e-06, + "loss": 0.8177, + "step": 5016 + }, + { + "epoch": 0.4524507372503044, + "grad_norm": 0.704443754514807, + "learning_rate": 2.402384750855084e-06, + "loss": 0.6361, + "step": 5017 + }, + { + "epoch": 0.4525409207737746, + "grad_norm": 1.5767748505385197, + "learning_rate": 2.4018124691539286e-06, + "loss": 0.7605, + "step": 5018 + }, + { + "epoch": 0.4526311042972449, + "grad_norm": 2.216003652480107, + "learning_rate": 2.4012401531679178e-06, + "loss": 0.7522, + "step": 5019 + }, + { + "epoch": 0.45272128782071513, + "grad_norm": 1.9290554231519739, + "learning_rate": 2.4006678029458847e-06, + "loss": 0.7566, + "step": 5020 + }, + { + "epoch": 0.4528114713441854, + "grad_norm": 1.831255522840068, + "learning_rate": 2.400095418536666e-06, + "loss": 0.6645, + "step": 5021 + }, + { + "epoch": 0.45290165486765566, + "grad_norm": 1.588479021014904, + "learning_rate": 2.3995229999890996e-06, + "loss": 0.7352, + "step": 5022 + }, + { + "epoch": 0.45299183839112594, + "grad_norm": 5.26115268304592, + "learning_rate": 2.398950547352028e-06, + "loss": 0.6986, + "step": 5023 + }, + { + "epoch": 0.4530820219145962, + "grad_norm": 3.4338507962925138, + "learning_rate": 2.398378060674295e-06, + "loss": 0.6793, + "step": 5024 + }, + { + "epoch": 0.45317220543806647, + "grad_norm": 1.6013484904889794, + "learning_rate": 2.39780554000475e-06, + "loss": 0.758, + "step": 5025 + }, + { + "epoch": 0.4532623889615367, + "grad_norm": 1.6085038280874322, + "learning_rate": 2.3972329853922434e-06, + "loss": 0.7875, + "step": 5026 + }, + { + "epoch": 0.453352572485007, + "grad_norm": 0.6064687285148824, + "learning_rate": 2.3966603968856278e-06, + "loss": 0.571, + "step": 5027 + }, + { + "epoch": 0.4534427560084773, + "grad_norm": 1.718095273246056, + "learning_rate": 2.39608777453376e-06, + "loss": 0.8286, + "step": 5028 + }, + { + "epoch": 0.4535329395319475, + "grad_norm": 8.527284699749735, + "learning_rate": 2.3955151183854993e-06, + "loss": 0.7836, + "step": 5029 + }, + { + "epoch": 0.4536231230554178, + "grad_norm": 2.7910366391244414, + "learning_rate": 2.3949424284897073e-06, + "loss": 0.715, + "step": 5030 + }, + { + "epoch": 0.45371330657888803, + "grad_norm": 1.6897402128051016, + "learning_rate": 2.39436970489525e-06, + "loss": 0.6264, + "step": 5031 + }, + { + "epoch": 0.4538034901023583, + "grad_norm": 1.6031066149871327, + "learning_rate": 2.3937969476509955e-06, + "loss": 0.7597, + "step": 5032 + }, + { + "epoch": 0.45389367362582855, + "grad_norm": 2.67127894841504, + "learning_rate": 2.393224156805813e-06, + "loss": 0.8257, + "step": 5033 + }, + { + "epoch": 0.45398385714929884, + "grad_norm": 3.056605836712976, + "learning_rate": 2.392651332408578e-06, + "loss": 0.7362, + "step": 5034 + }, + { + "epoch": 0.4540740406727691, + "grad_norm": 4.222952776721396, + "learning_rate": 2.3920784745081655e-06, + "loss": 0.7442, + "step": 5035 + }, + { + "epoch": 0.45416422419623936, + "grad_norm": 1.7589727345013038, + "learning_rate": 2.391505583153456e-06, + "loss": 0.7934, + "step": 5036 + }, + { + "epoch": 0.4542544077197096, + "grad_norm": 3.698548269285117, + "learning_rate": 2.3909326583933315e-06, + "loss": 0.7612, + "step": 5037 + }, + { + "epoch": 0.4543445912431799, + "grad_norm": 2.362008046594362, + "learning_rate": 2.3903597002766777e-06, + "loss": 0.7259, + "step": 5038 + }, + { + "epoch": 0.4544347747666501, + "grad_norm": 2.317712733803904, + "learning_rate": 2.389786708852381e-06, + "loss": 0.7198, + "step": 5039 + }, + { + "epoch": 0.4545249582901204, + "grad_norm": 0.9481279989960829, + "learning_rate": 2.389213684169333e-06, + "loss": 0.6964, + "step": 5040 + }, + { + "epoch": 0.45461514181359064, + "grad_norm": 1.7504919586781456, + "learning_rate": 2.388640626276428e-06, + "loss": 0.7185, + "step": 5041 + }, + { + "epoch": 0.45470532533706093, + "grad_norm": 0.682914679976113, + "learning_rate": 2.388067535222561e-06, + "loss": 0.5353, + "step": 5042 + }, + { + "epoch": 0.45479550886053116, + "grad_norm": 1.8246990608937514, + "learning_rate": 2.3874944110566332e-06, + "loss": 0.7753, + "step": 5043 + }, + { + "epoch": 0.45488569238400145, + "grad_norm": 0.7695288488670106, + "learning_rate": 2.3869212538275447e-06, + "loss": 0.6782, + "step": 5044 + }, + { + "epoch": 0.4549758759074717, + "grad_norm": 3.887964526324666, + "learning_rate": 2.386348063584202e-06, + "loss": 0.7683, + "step": 5045 + }, + { + "epoch": 0.45506605943094197, + "grad_norm": 1.843655729308395, + "learning_rate": 2.385774840375511e-06, + "loss": 0.8101, + "step": 5046 + }, + { + "epoch": 0.4551562429544122, + "grad_norm": 1.5841619013447588, + "learning_rate": 2.385201584250385e-06, + "loss": 0.7113, + "step": 5047 + }, + { + "epoch": 0.4552464264778825, + "grad_norm": 0.6324232811820859, + "learning_rate": 2.3846282952577346e-06, + "loss": 0.6128, + "step": 5048 + }, + { + "epoch": 0.4553366100013527, + "grad_norm": 1.6118578970490982, + "learning_rate": 2.3840549734464785e-06, + "loss": 0.8437, + "step": 5049 + }, + { + "epoch": 0.455426793524823, + "grad_norm": 5.7320169283077185, + "learning_rate": 2.3834816188655336e-06, + "loss": 0.7011, + "step": 5050 + }, + { + "epoch": 0.4555169770482933, + "grad_norm": 1.7733536755081063, + "learning_rate": 2.3829082315638224e-06, + "loss": 0.7017, + "step": 5051 + }, + { + "epoch": 0.45560716057176354, + "grad_norm": 4.876484375938716, + "learning_rate": 2.3823348115902695e-06, + "loss": 0.6154, + "step": 5052 + }, + { + "epoch": 0.4556973440952338, + "grad_norm": 1.7657237067524598, + "learning_rate": 2.3817613589938026e-06, + "loss": 0.7752, + "step": 5053 + }, + { + "epoch": 0.45578752761870406, + "grad_norm": 1.8408529506390414, + "learning_rate": 2.3811878738233517e-06, + "loss": 0.7911, + "step": 5054 + }, + { + "epoch": 0.45587771114217435, + "grad_norm": 1.9319813180063647, + "learning_rate": 2.380614356127849e-06, + "loss": 0.7023, + "step": 5055 + }, + { + "epoch": 0.4559678946656446, + "grad_norm": 1.8597035478384438, + "learning_rate": 2.3800408059562318e-06, + "loss": 0.75, + "step": 5056 + }, + { + "epoch": 0.45605807818911487, + "grad_norm": 2.2070687349900227, + "learning_rate": 2.3794672233574365e-06, + "loss": 0.7676, + "step": 5057 + }, + { + "epoch": 0.4561482617125851, + "grad_norm": 1.8795107307300274, + "learning_rate": 2.3788936083804058e-06, + "loss": 0.7663, + "step": 5058 + }, + { + "epoch": 0.4562384452360554, + "grad_norm": 1.9147033864080263, + "learning_rate": 2.378319961074083e-06, + "loss": 0.7557, + "step": 5059 + }, + { + "epoch": 0.4563286287595256, + "grad_norm": 26.919084029010673, + "learning_rate": 2.377746281487415e-06, + "loss": 0.7919, + "step": 5060 + }, + { + "epoch": 0.4564188122829959, + "grad_norm": 1.5423732719111822, + "learning_rate": 2.377172569669352e-06, + "loss": 0.845, + "step": 5061 + }, + { + "epoch": 0.45650899580646614, + "grad_norm": 1.9191743437392212, + "learning_rate": 2.376598825668845e-06, + "loss": 0.6119, + "step": 5062 + }, + { + "epoch": 0.45659917932993643, + "grad_norm": 1.8387733796148453, + "learning_rate": 2.3760250495348495e-06, + "loss": 0.7444, + "step": 5063 + }, + { + "epoch": 0.45668936285340667, + "grad_norm": 2.139672505997673, + "learning_rate": 2.3754512413163236e-06, + "loss": 0.8017, + "step": 5064 + }, + { + "epoch": 0.45677954637687695, + "grad_norm": 1.8062842765116576, + "learning_rate": 2.3748774010622285e-06, + "loss": 0.7418, + "step": 5065 + }, + { + "epoch": 0.4568697299003472, + "grad_norm": 1.6049643510798939, + "learning_rate": 2.3743035288215254e-06, + "loss": 0.7663, + "step": 5066 + }, + { + "epoch": 0.4569599134238175, + "grad_norm": 1.7619412233073066, + "learning_rate": 2.3737296246431815e-06, + "loss": 0.7672, + "step": 5067 + }, + { + "epoch": 0.4570500969472877, + "grad_norm": 1.6066745211909186, + "learning_rate": 2.3731556885761656e-06, + "loss": 0.7309, + "step": 5068 + }, + { + "epoch": 0.457140280470758, + "grad_norm": 0.6470807830856908, + "learning_rate": 2.372581720669449e-06, + "loss": 0.5797, + "step": 5069 + }, + { + "epoch": 0.45723046399422823, + "grad_norm": 2.7765820366768357, + "learning_rate": 2.3720077209720046e-06, + "loss": 0.7854, + "step": 5070 + }, + { + "epoch": 0.4573206475176985, + "grad_norm": 0.6914095258904562, + "learning_rate": 2.3714336895328112e-06, + "loss": 0.5735, + "step": 5071 + }, + { + "epoch": 0.45741083104116875, + "grad_norm": 1.9127683663165156, + "learning_rate": 2.370859626400847e-06, + "loss": 0.7089, + "step": 5072 + }, + { + "epoch": 0.45750101456463904, + "grad_norm": 1.561785954398152, + "learning_rate": 2.3702855316250943e-06, + "loss": 0.733, + "step": 5073 + }, + { + "epoch": 0.45759119808810933, + "grad_norm": 2.093472704388518, + "learning_rate": 2.369711405254539e-06, + "loss": 0.7588, + "step": 5074 + }, + { + "epoch": 0.45768138161157956, + "grad_norm": 2.558549673851034, + "learning_rate": 2.3691372473381673e-06, + "loss": 0.8116, + "step": 5075 + }, + { + "epoch": 0.45777156513504985, + "grad_norm": 1.700846696820141, + "learning_rate": 2.3685630579249708e-06, + "loss": 0.8158, + "step": 5076 + }, + { + "epoch": 0.4578617486585201, + "grad_norm": 2.0349399795866265, + "learning_rate": 2.367988837063942e-06, + "loss": 0.7321, + "step": 5077 + }, + { + "epoch": 0.4579519321819904, + "grad_norm": 1.7640560910266454, + "learning_rate": 2.367414584804076e-06, + "loss": 0.8333, + "step": 5078 + }, + { + "epoch": 0.4580421157054606, + "grad_norm": 1.7081783115792626, + "learning_rate": 2.366840301194372e-06, + "loss": 0.7923, + "step": 5079 + }, + { + "epoch": 0.4581322992289309, + "grad_norm": 1.8043228375486493, + "learning_rate": 2.3662659862838308e-06, + "loss": 0.7443, + "step": 5080 + }, + { + "epoch": 0.45822248275240113, + "grad_norm": 1.7340361118010286, + "learning_rate": 2.365691640121456e-06, + "loss": 0.8139, + "step": 5081 + }, + { + "epoch": 0.4583126662758714, + "grad_norm": 2.0934960510202085, + "learning_rate": 2.365117262756254e-06, + "loss": 0.6871, + "step": 5082 + }, + { + "epoch": 0.45840284979934165, + "grad_norm": 1.909112188604585, + "learning_rate": 2.3645428542372342e-06, + "loss": 0.7478, + "step": 5083 + }, + { + "epoch": 0.45849303332281194, + "grad_norm": 0.6157330868326834, + "learning_rate": 2.3639684146134083e-06, + "loss": 0.5316, + "step": 5084 + }, + { + "epoch": 0.45858321684628217, + "grad_norm": 1.9973069179063108, + "learning_rate": 2.3633939439337897e-06, + "loss": 0.722, + "step": 5085 + }, + { + "epoch": 0.45867340036975246, + "grad_norm": 1.5769259741758317, + "learning_rate": 2.362819442247396e-06, + "loss": 0.7259, + "step": 5086 + }, + { + "epoch": 0.4587635838932227, + "grad_norm": 0.6270223799153452, + "learning_rate": 2.3622449096032477e-06, + "loss": 0.5419, + "step": 5087 + }, + { + "epoch": 0.458853767416693, + "grad_norm": 1.729958280269787, + "learning_rate": 2.361670346050366e-06, + "loss": 0.7946, + "step": 5088 + }, + { + "epoch": 0.4589439509401632, + "grad_norm": 4.439574710307108, + "learning_rate": 2.3610957516377757e-06, + "loss": 0.6951, + "step": 5089 + }, + { + "epoch": 0.4590341344636335, + "grad_norm": 1.6944608878638918, + "learning_rate": 2.3605211264145048e-06, + "loss": 0.7569, + "step": 5090 + }, + { + "epoch": 0.45912431798710374, + "grad_norm": 1.3940055561556979, + "learning_rate": 2.3599464704295836e-06, + "loss": 0.7233, + "step": 5091 + }, + { + "epoch": 0.459214501510574, + "grad_norm": 1.8418876164434317, + "learning_rate": 2.359371783732045e-06, + "loss": 0.6465, + "step": 5092 + }, + { + "epoch": 0.45930468503404426, + "grad_norm": 1.899417012268757, + "learning_rate": 2.358797066370924e-06, + "loss": 0.7395, + "step": 5093 + }, + { + "epoch": 0.45939486855751455, + "grad_norm": 1.7242729921714781, + "learning_rate": 2.3582223183952594e-06, + "loss": 0.7524, + "step": 5094 + }, + { + "epoch": 0.4594850520809848, + "grad_norm": 1.7437496281011589, + "learning_rate": 2.357647539854091e-06, + "loss": 0.7864, + "step": 5095 + }, + { + "epoch": 0.45957523560445507, + "grad_norm": 1.8146686733892492, + "learning_rate": 2.3570727307964624e-06, + "loss": 0.7627, + "step": 5096 + }, + { + "epoch": 0.4596654191279253, + "grad_norm": 1.6574381758953434, + "learning_rate": 2.35649789127142e-06, + "loss": 0.8059, + "step": 5097 + }, + { + "epoch": 0.4597556026513956, + "grad_norm": 1.8272312946789606, + "learning_rate": 2.3559230213280115e-06, + "loss": 0.7262, + "step": 5098 + }, + { + "epoch": 0.4598457861748659, + "grad_norm": 4.683758272235489, + "learning_rate": 2.3553481210152886e-06, + "loss": 0.6711, + "step": 5099 + }, + { + "epoch": 0.4599359696983361, + "grad_norm": 0.8371202177127867, + "learning_rate": 2.3547731903823043e-06, + "loss": 0.5973, + "step": 5100 + }, + { + "epoch": 0.4600261532218064, + "grad_norm": 4.567309835950835, + "learning_rate": 2.3541982294781155e-06, + "loss": 0.7304, + "step": 5101 + }, + { + "epoch": 0.46011633674527663, + "grad_norm": 7.309680101567774, + "learning_rate": 2.3536232383517804e-06, + "loss": 0.7503, + "step": 5102 + }, + { + "epoch": 0.4602065202687469, + "grad_norm": 1.4499400521086596, + "learning_rate": 2.3530482170523602e-06, + "loss": 0.7497, + "step": 5103 + }, + { + "epoch": 0.46029670379221715, + "grad_norm": 1.7588377950589995, + "learning_rate": 2.3524731656289206e-06, + "loss": 0.8493, + "step": 5104 + }, + { + "epoch": 0.46038688731568744, + "grad_norm": 1.5341179129282116, + "learning_rate": 2.351898084130526e-06, + "loss": 0.8462, + "step": 5105 + }, + { + "epoch": 0.4604770708391577, + "grad_norm": 1.8204540148560986, + "learning_rate": 2.351322972606247e-06, + "loss": 0.8234, + "step": 5106 + }, + { + "epoch": 0.46056725436262796, + "grad_norm": 12.403274448346922, + "learning_rate": 2.350747831105155e-06, + "loss": 0.9059, + "step": 5107 + }, + { + "epoch": 0.4606574378860982, + "grad_norm": 2.2875370502077117, + "learning_rate": 2.350172659676323e-06, + "loss": 0.8358, + "step": 5108 + }, + { + "epoch": 0.4607476214095685, + "grad_norm": 1.6955019234162931, + "learning_rate": 2.3495974583688306e-06, + "loss": 0.7926, + "step": 5109 + }, + { + "epoch": 0.4608378049330387, + "grad_norm": 1.6119953045835809, + "learning_rate": 2.3490222272317543e-06, + "loss": 0.8263, + "step": 5110 + }, + { + "epoch": 0.460927988456509, + "grad_norm": 2.4617224031964224, + "learning_rate": 2.348446966314177e-06, + "loss": 0.8011, + "step": 5111 + }, + { + "epoch": 0.46101817197997924, + "grad_norm": 1.9440945408873718, + "learning_rate": 2.3478716756651837e-06, + "loss": 0.7282, + "step": 5112 + }, + { + "epoch": 0.46110835550344953, + "grad_norm": 1.8252085710251118, + "learning_rate": 2.347296355333861e-06, + "loss": 0.7499, + "step": 5113 + }, + { + "epoch": 0.46119853902691976, + "grad_norm": 2.293364768459688, + "learning_rate": 2.3467210053692972e-06, + "loss": 0.7655, + "step": 5114 + }, + { + "epoch": 0.46128872255039005, + "grad_norm": 2.072010308589846, + "learning_rate": 2.3461456258205866e-06, + "loss": 0.7583, + "step": 5115 + }, + { + "epoch": 0.4613789060738603, + "grad_norm": 1.8179966024864354, + "learning_rate": 2.345570216736822e-06, + "loss": 0.6993, + "step": 5116 + }, + { + "epoch": 0.4614690895973306, + "grad_norm": 1.8171771638794405, + "learning_rate": 2.3449947781671013e-06, + "loss": 0.7679, + "step": 5117 + }, + { + "epoch": 0.4615592731208008, + "grad_norm": 2.248245933064978, + "learning_rate": 2.3444193101605237e-06, + "loss": 0.7866, + "step": 5118 + }, + { + "epoch": 0.4616494566442711, + "grad_norm": 1.7505059192146613, + "learning_rate": 2.3438438127661913e-06, + "loss": 0.7757, + "step": 5119 + }, + { + "epoch": 0.4617396401677413, + "grad_norm": 1.647939308419275, + "learning_rate": 2.3432682860332096e-06, + "loss": 0.5848, + "step": 5120 + }, + { + "epoch": 0.4618298236912116, + "grad_norm": 1.8456502916958728, + "learning_rate": 2.342692730010684e-06, + "loss": 0.7069, + "step": 5121 + }, + { + "epoch": 0.4619200072146819, + "grad_norm": 2.633023035704221, + "learning_rate": 2.342117144747726e-06, + "loss": 0.77, + "step": 5122 + }, + { + "epoch": 0.46201019073815214, + "grad_norm": 1.8919678010035423, + "learning_rate": 2.3415415302934457e-06, + "loss": 0.759, + "step": 5123 + }, + { + "epoch": 0.4621003742616224, + "grad_norm": 1.6511447605553695, + "learning_rate": 2.340965886696959e-06, + "loss": 0.7855, + "step": 5124 + }, + { + "epoch": 0.46219055778509266, + "grad_norm": 1.8268297611759996, + "learning_rate": 2.340390214007384e-06, + "loss": 0.7059, + "step": 5125 + }, + { + "epoch": 0.46228074130856295, + "grad_norm": 1.9566141319804102, + "learning_rate": 2.339814512273838e-06, + "loss": 0.6408, + "step": 5126 + }, + { + "epoch": 0.4623709248320332, + "grad_norm": 1.7368929344478499, + "learning_rate": 2.3392387815454447e-06, + "loss": 0.863, + "step": 5127 + }, + { + "epoch": 0.46246110835550347, + "grad_norm": 1.9165909240794374, + "learning_rate": 2.3386630218713273e-06, + "loss": 0.7522, + "step": 5128 + }, + { + "epoch": 0.4625512918789737, + "grad_norm": 1.7565389809478626, + "learning_rate": 2.3380872333006135e-06, + "loss": 0.7883, + "step": 5129 + }, + { + "epoch": 0.462641475402444, + "grad_norm": 12.901712206387387, + "learning_rate": 2.3375114158824335e-06, + "loss": 0.8051, + "step": 5130 + }, + { + "epoch": 0.4627316589259142, + "grad_norm": 1.7018974850192272, + "learning_rate": 2.3369355696659184e-06, + "loss": 0.7919, + "step": 5131 + }, + { + "epoch": 0.4628218424493845, + "grad_norm": 1.8583922153033974, + "learning_rate": 2.336359694700202e-06, + "loss": 0.7443, + "step": 5132 + }, + { + "epoch": 0.46291202597285475, + "grad_norm": 1.6603041639070268, + "learning_rate": 2.335783791034422e-06, + "loss": 0.7605, + "step": 5133 + }, + { + "epoch": 0.46300220949632503, + "grad_norm": 1.805276287521924, + "learning_rate": 2.3352078587177173e-06, + "loss": 0.7412, + "step": 5134 + }, + { + "epoch": 0.46309239301979527, + "grad_norm": 3.3434416726821974, + "learning_rate": 2.33463189779923e-06, + "loss": 0.7693, + "step": 5135 + }, + { + "epoch": 0.46318257654326556, + "grad_norm": 1.5130357944013653, + "learning_rate": 2.334055908328104e-06, + "loss": 0.828, + "step": 5136 + }, + { + "epoch": 0.4632727600667358, + "grad_norm": 1.560712249357972, + "learning_rate": 2.3334798903534866e-06, + "loss": 0.6723, + "step": 5137 + }, + { + "epoch": 0.4633629435902061, + "grad_norm": 1.4486433259464961, + "learning_rate": 2.3329038439245252e-06, + "loss": 0.653, + "step": 5138 + }, + { + "epoch": 0.4634531271136763, + "grad_norm": 2.291352464348347, + "learning_rate": 2.3323277690903724e-06, + "loss": 0.7573, + "step": 5139 + }, + { + "epoch": 0.4635433106371466, + "grad_norm": 1.5612656867870918, + "learning_rate": 2.3317516659001827e-06, + "loss": 0.7314, + "step": 5140 + }, + { + "epoch": 0.46363349416061683, + "grad_norm": 1.53947342552152, + "learning_rate": 2.331175534403111e-06, + "loss": 0.768, + "step": 5141 + }, + { + "epoch": 0.4637236776840871, + "grad_norm": 2.3106398446862735, + "learning_rate": 2.3305993746483167e-06, + "loss": 0.8431, + "step": 5142 + }, + { + "epoch": 0.46381386120755735, + "grad_norm": 2.3868751786784554, + "learning_rate": 2.3300231866849606e-06, + "loss": 0.7356, + "step": 5143 + }, + { + "epoch": 0.46390404473102764, + "grad_norm": 1.9637959228818291, + "learning_rate": 2.3294469705622067e-06, + "loss": 0.797, + "step": 5144 + }, + { + "epoch": 0.4639942282544979, + "grad_norm": 1.926138761503037, + "learning_rate": 2.3288707263292203e-06, + "loss": 0.7357, + "step": 5145 + }, + { + "epoch": 0.46408441177796816, + "grad_norm": 1.391063706708694, + "learning_rate": 2.3282944540351707e-06, + "loss": 0.7772, + "step": 5146 + }, + { + "epoch": 0.46417459530143845, + "grad_norm": 2.1350517395325475, + "learning_rate": 2.327718153729228e-06, + "loss": 0.7317, + "step": 5147 + }, + { + "epoch": 0.4642647788249087, + "grad_norm": 1.6712036210831216, + "learning_rate": 2.327141825460566e-06, + "loss": 0.8032, + "step": 5148 + }, + { + "epoch": 0.464354962348379, + "grad_norm": 0.6713492310201739, + "learning_rate": 2.326565469278358e-06, + "loss": 0.5852, + "step": 5149 + }, + { + "epoch": 0.4644451458718492, + "grad_norm": 0.7133006549715077, + "learning_rate": 2.3259890852317846e-06, + "loss": 0.5944, + "step": 5150 + }, + { + "epoch": 0.4645353293953195, + "grad_norm": 1.8299783294728744, + "learning_rate": 2.3254126733700246e-06, + "loss": 0.7586, + "step": 5151 + }, + { + "epoch": 0.46462551291878973, + "grad_norm": 2.1459212670935086, + "learning_rate": 2.324836233742262e-06, + "loss": 0.7868, + "step": 5152 + }, + { + "epoch": 0.46471569644226, + "grad_norm": 1.945126199035883, + "learning_rate": 2.3242597663976793e-06, + "loss": 0.7951, + "step": 5153 + }, + { + "epoch": 0.46480587996573025, + "grad_norm": 1.8791618569841881, + "learning_rate": 2.3236832713854663e-06, + "loss": 0.7374, + "step": 5154 + }, + { + "epoch": 0.46489606348920054, + "grad_norm": 1.6229263426166205, + "learning_rate": 2.323106748754812e-06, + "loss": 0.7534, + "step": 5155 + }, + { + "epoch": 0.4649862470126708, + "grad_norm": 2.363326546731449, + "learning_rate": 2.3225301985549077e-06, + "loss": 0.8557, + "step": 5156 + }, + { + "epoch": 0.46507643053614106, + "grad_norm": 2.8089163094937937, + "learning_rate": 2.321953620834948e-06, + "loss": 0.6922, + "step": 5157 + }, + { + "epoch": 0.4651666140596113, + "grad_norm": 1.506155576160175, + "learning_rate": 2.3213770156441314e-06, + "loss": 0.7207, + "step": 5158 + }, + { + "epoch": 0.4652567975830816, + "grad_norm": 2.5179021728879554, + "learning_rate": 2.3208003830316554e-06, + "loss": 0.8114, + "step": 5159 + }, + { + "epoch": 0.4653469811065518, + "grad_norm": 1.852317109031026, + "learning_rate": 2.3202237230467215e-06, + "loss": 0.8474, + "step": 5160 + }, + { + "epoch": 0.4654371646300221, + "grad_norm": 2.0193843825041724, + "learning_rate": 2.3196470357385338e-06, + "loss": 0.7161, + "step": 5161 + }, + { + "epoch": 0.46552734815349234, + "grad_norm": 1.4973971513383013, + "learning_rate": 2.319070321156299e-06, + "loss": 0.7524, + "step": 5162 + }, + { + "epoch": 0.4656175316769626, + "grad_norm": 1.9992407311218332, + "learning_rate": 2.318493579349224e-06, + "loss": 0.8882, + "step": 5163 + }, + { + "epoch": 0.46570771520043286, + "grad_norm": 1.7698173375893929, + "learning_rate": 2.317916810366522e-06, + "loss": 0.7468, + "step": 5164 + }, + { + "epoch": 0.46579789872390315, + "grad_norm": 1.827624586688293, + "learning_rate": 2.317340014257404e-06, + "loss": 0.7002, + "step": 5165 + }, + { + "epoch": 0.4658880822473734, + "grad_norm": 1.5307832123191991, + "learning_rate": 2.316763191071086e-06, + "loss": 0.6851, + "step": 5166 + }, + { + "epoch": 0.46597826577084367, + "grad_norm": 1.5915360125830502, + "learning_rate": 2.316186340856787e-06, + "loss": 0.7164, + "step": 5167 + }, + { + "epoch": 0.4660684492943139, + "grad_norm": 2.5061528783991704, + "learning_rate": 2.315609463663725e-06, + "loss": 0.7142, + "step": 5168 + }, + { + "epoch": 0.4661586328177842, + "grad_norm": 1.939900971765873, + "learning_rate": 2.315032559541123e-06, + "loss": 0.7286, + "step": 5169 + }, + { + "epoch": 0.4662488163412545, + "grad_norm": 1.8663979620767481, + "learning_rate": 2.314455628538207e-06, + "loss": 0.7828, + "step": 5170 + }, + { + "epoch": 0.4663389998647247, + "grad_norm": 1.9649466455413538, + "learning_rate": 2.3138786707042023e-06, + "loss": 0.7279, + "step": 5171 + }, + { + "epoch": 0.466429183388195, + "grad_norm": 1.6074471639727352, + "learning_rate": 2.3133016860883387e-06, + "loss": 0.7642, + "step": 5172 + }, + { + "epoch": 0.46651936691166523, + "grad_norm": 1.8597975819691877, + "learning_rate": 2.3127246747398475e-06, + "loss": 0.7362, + "step": 5173 + }, + { + "epoch": 0.4666095504351355, + "grad_norm": 6.990854828918541, + "learning_rate": 2.312147636707963e-06, + "loss": 0.7554, + "step": 5174 + }, + { + "epoch": 0.46669973395860576, + "grad_norm": 1.8986734000369665, + "learning_rate": 2.3115705720419214e-06, + "loss": 0.7446, + "step": 5175 + }, + { + "epoch": 0.46678991748207604, + "grad_norm": 1.6436549228106065, + "learning_rate": 2.31099348079096e-06, + "loss": 0.7791, + "step": 5176 + }, + { + "epoch": 0.4668801010055463, + "grad_norm": 2.51964215175012, + "learning_rate": 2.31041636300432e-06, + "loss": 0.8006, + "step": 5177 + }, + { + "epoch": 0.46697028452901657, + "grad_norm": 1.9421401910398286, + "learning_rate": 2.3098392187312445e-06, + "loss": 0.838, + "step": 5178 + }, + { + "epoch": 0.4670604680524868, + "grad_norm": 3.982347997529258, + "learning_rate": 2.309262048020978e-06, + "loss": 0.7343, + "step": 5179 + }, + { + "epoch": 0.4671506515759571, + "grad_norm": 1.8631578960413029, + "learning_rate": 2.308684850922769e-06, + "loss": 0.7877, + "step": 5180 + }, + { + "epoch": 0.4672408350994273, + "grad_norm": 1.8611526326471124, + "learning_rate": 2.3081076274858664e-06, + "loss": 0.7562, + "step": 5181 + }, + { + "epoch": 0.4673310186228976, + "grad_norm": 1.562209827181892, + "learning_rate": 2.307530377759522e-06, + "loss": 0.7863, + "step": 5182 + }, + { + "epoch": 0.46742120214636784, + "grad_norm": 2.128515477144597, + "learning_rate": 2.30695310179299e-06, + "loss": 0.6847, + "step": 5183 + }, + { + "epoch": 0.46751138566983813, + "grad_norm": 1.5004186840656464, + "learning_rate": 2.3063757996355267e-06, + "loss": 0.7531, + "step": 5184 + }, + { + "epoch": 0.46760156919330836, + "grad_norm": 2.3742308123960916, + "learning_rate": 2.3057984713363903e-06, + "loss": 0.7873, + "step": 5185 + }, + { + "epoch": 0.46769175271677865, + "grad_norm": 0.677430868211191, + "learning_rate": 2.3052211169448436e-06, + "loss": 0.5571, + "step": 5186 + }, + { + "epoch": 0.4677819362402489, + "grad_norm": 1.7155369637002933, + "learning_rate": 2.3046437365101474e-06, + "loss": 0.7027, + "step": 5187 + }, + { + "epoch": 0.4678721197637192, + "grad_norm": 1.644151004859096, + "learning_rate": 2.3040663300815673e-06, + "loss": 0.7318, + "step": 5188 + }, + { + "epoch": 0.4679623032871894, + "grad_norm": 0.7233234521598183, + "learning_rate": 2.3034888977083723e-06, + "loss": 0.5657, + "step": 5189 + }, + { + "epoch": 0.4680524868106597, + "grad_norm": 4.036990078592381, + "learning_rate": 2.30291143943983e-06, + "loss": 0.6716, + "step": 5190 + }, + { + "epoch": 0.46814267033412993, + "grad_norm": 1.6047322981096228, + "learning_rate": 2.3023339553252145e-06, + "loss": 0.7172, + "step": 5191 + }, + { + "epoch": 0.4682328538576002, + "grad_norm": 2.8546488742063048, + "learning_rate": 2.301756445413799e-06, + "loss": 0.7924, + "step": 5192 + }, + { + "epoch": 0.4683230373810705, + "grad_norm": 2.264853721088779, + "learning_rate": 2.3011789097548585e-06, + "loss": 0.759, + "step": 5193 + }, + { + "epoch": 0.46841322090454074, + "grad_norm": 1.609930664725658, + "learning_rate": 2.3006013483976738e-06, + "loss": 0.7555, + "step": 5194 + }, + { + "epoch": 0.468503404428011, + "grad_norm": 3.3592563342050106, + "learning_rate": 2.300023761391524e-06, + "loss": 0.7533, + "step": 5195 + }, + { + "epoch": 0.46859358795148126, + "grad_norm": 2.94381003672938, + "learning_rate": 2.299446148785693e-06, + "loss": 0.8099, + "step": 5196 + }, + { + "epoch": 0.46868377147495155, + "grad_norm": 1.9242431193027139, + "learning_rate": 2.2988685106294654e-06, + "loss": 0.7371, + "step": 5197 + }, + { + "epoch": 0.4687739549984218, + "grad_norm": 2.3484051060679056, + "learning_rate": 2.2982908469721284e-06, + "loss": 0.6569, + "step": 5198 + }, + { + "epoch": 0.46886413852189207, + "grad_norm": 2.1864757864880833, + "learning_rate": 2.2977131578629714e-06, + "loss": 0.8263, + "step": 5199 + }, + { + "epoch": 0.4689543220453623, + "grad_norm": 1.9783643267398996, + "learning_rate": 2.297135443351286e-06, + "loss": 0.7799, + "step": 5200 + }, + { + "epoch": 0.4690445055688326, + "grad_norm": 2.2707190761927905, + "learning_rate": 2.296557703486367e-06, + "loss": 0.7664, + "step": 5201 + }, + { + "epoch": 0.4691346890923028, + "grad_norm": 2.2551050551613896, + "learning_rate": 2.295979938317509e-06, + "loss": 0.7819, + "step": 5202 + }, + { + "epoch": 0.4692248726157731, + "grad_norm": 2.174090127159397, + "learning_rate": 2.295402147894011e-06, + "loss": 0.763, + "step": 5203 + }, + { + "epoch": 0.46931505613924335, + "grad_norm": 1.3941411773774603, + "learning_rate": 2.2948243322651723e-06, + "loss": 0.681, + "step": 5204 + }, + { + "epoch": 0.46940523966271364, + "grad_norm": 1.7247503791097591, + "learning_rate": 2.2942464914802962e-06, + "loss": 0.7609, + "step": 5205 + }, + { + "epoch": 0.46949542318618387, + "grad_norm": 1.736346456046902, + "learning_rate": 2.293668625588687e-06, + "loss": 0.7325, + "step": 5206 + }, + { + "epoch": 0.46958560670965416, + "grad_norm": 2.2302699225906055, + "learning_rate": 2.293090734639651e-06, + "loss": 0.8022, + "step": 5207 + }, + { + "epoch": 0.4696757902331244, + "grad_norm": 1.5474421684786048, + "learning_rate": 2.2925128186824983e-06, + "loss": 0.8735, + "step": 5208 + }, + { + "epoch": 0.4697659737565947, + "grad_norm": 2.31657483675883, + "learning_rate": 2.2919348777665384e-06, + "loss": 0.8627, + "step": 5209 + }, + { + "epoch": 0.4698561572800649, + "grad_norm": 1.6496019634682841, + "learning_rate": 2.2913569119410856e-06, + "loss": 0.7623, + "step": 5210 + }, + { + "epoch": 0.4699463408035352, + "grad_norm": 2.785631650181108, + "learning_rate": 2.290778921255454e-06, + "loss": 0.7828, + "step": 5211 + }, + { + "epoch": 0.47003652432700543, + "grad_norm": 1.5052285462532553, + "learning_rate": 2.2902009057589613e-06, + "loss": 0.795, + "step": 5212 + }, + { + "epoch": 0.4701267078504757, + "grad_norm": 1.4391385566750818, + "learning_rate": 2.2896228655009276e-06, + "loss": 0.7138, + "step": 5213 + }, + { + "epoch": 0.47021689137394596, + "grad_norm": 2.5223834313489397, + "learning_rate": 2.289044800530674e-06, + "loss": 0.7281, + "step": 5214 + }, + { + "epoch": 0.47030707489741624, + "grad_norm": 2.5441806325603085, + "learning_rate": 2.2884667108975245e-06, + "loss": 0.7231, + "step": 5215 + }, + { + "epoch": 0.4703972584208865, + "grad_norm": 1.8255100295272724, + "learning_rate": 2.287888596650804e-06, + "loss": 0.7997, + "step": 5216 + }, + { + "epoch": 0.47048744194435677, + "grad_norm": 1.8651605561513407, + "learning_rate": 2.287310457839841e-06, + "loss": 0.7618, + "step": 5217 + }, + { + "epoch": 0.47057762546782705, + "grad_norm": 1.8330227054078325, + "learning_rate": 2.286732294513966e-06, + "loss": 0.8308, + "step": 5218 + }, + { + "epoch": 0.4706678089912973, + "grad_norm": 1.8626170384033558, + "learning_rate": 2.2861541067225106e-06, + "loss": 0.7706, + "step": 5219 + }, + { + "epoch": 0.4707579925147676, + "grad_norm": 14.596749432464087, + "learning_rate": 2.2855758945148095e-06, + "loss": 0.7411, + "step": 5220 + }, + { + "epoch": 0.4708481760382378, + "grad_norm": 1.6547587176766305, + "learning_rate": 2.2849976579401977e-06, + "loss": 0.8156, + "step": 5221 + }, + { + "epoch": 0.4709383595617081, + "grad_norm": 1.9094706984292245, + "learning_rate": 2.284419397048014e-06, + "loss": 0.7541, + "step": 5222 + }, + { + "epoch": 0.47102854308517833, + "grad_norm": 1.6588841489979802, + "learning_rate": 2.2838411118875997e-06, + "loss": 0.7628, + "step": 5223 + }, + { + "epoch": 0.4711187266086486, + "grad_norm": 1.420286863660102, + "learning_rate": 2.283262802508296e-06, + "loss": 0.68, + "step": 5224 + }, + { + "epoch": 0.47120891013211885, + "grad_norm": 1.3769575403020966, + "learning_rate": 2.2826844689594492e-06, + "loss": 0.7338, + "step": 5225 + }, + { + "epoch": 0.47129909365558914, + "grad_norm": 1.6460877797254316, + "learning_rate": 2.282106111290404e-06, + "loss": 0.8369, + "step": 5226 + }, + { + "epoch": 0.4713892771790594, + "grad_norm": 0.6912644451004795, + "learning_rate": 2.2815277295505098e-06, + "loss": 0.6011, + "step": 5227 + }, + { + "epoch": 0.47147946070252966, + "grad_norm": 1.7454210730841808, + "learning_rate": 2.2809493237891174e-06, + "loss": 0.7078, + "step": 5228 + }, + { + "epoch": 0.4715696442259999, + "grad_norm": 1.4779956462625639, + "learning_rate": 2.2803708940555796e-06, + "loss": 0.7903, + "step": 5229 + }, + { + "epoch": 0.4716598277494702, + "grad_norm": 1.9228861560102382, + "learning_rate": 2.2797924403992514e-06, + "loss": 0.8252, + "step": 5230 + }, + { + "epoch": 0.4717500112729404, + "grad_norm": 1.852143079940415, + "learning_rate": 2.2792139628694892e-06, + "loss": 0.7771, + "step": 5231 + }, + { + "epoch": 0.4718401947964107, + "grad_norm": 8.779559098211275, + "learning_rate": 2.2786354615156524e-06, + "loss": 0.7186, + "step": 5232 + }, + { + "epoch": 0.47193037831988094, + "grad_norm": 1.9056982899225536, + "learning_rate": 2.2780569363871016e-06, + "loss": 0.8282, + "step": 5233 + }, + { + "epoch": 0.4720205618433512, + "grad_norm": 0.6161290520913814, + "learning_rate": 2.277478387533199e-06, + "loss": 0.5734, + "step": 5234 + }, + { + "epoch": 0.47211074536682146, + "grad_norm": 2.932042185241815, + "learning_rate": 2.276899815003311e-06, + "loss": 0.7969, + "step": 5235 + }, + { + "epoch": 0.47220092889029175, + "grad_norm": 1.8285004971894445, + "learning_rate": 2.2763212188468045e-06, + "loss": 0.7529, + "step": 5236 + }, + { + "epoch": 0.472291112413762, + "grad_norm": 1.2781977700407547, + "learning_rate": 2.2757425991130473e-06, + "loss": 0.7726, + "step": 5237 + }, + { + "epoch": 0.47238129593723227, + "grad_norm": 0.7746012246246073, + "learning_rate": 2.2751639558514117e-06, + "loss": 0.5859, + "step": 5238 + }, + { + "epoch": 0.4724714794607025, + "grad_norm": 1.3927085355284416, + "learning_rate": 2.2745852891112697e-06, + "loss": 0.7991, + "step": 5239 + }, + { + "epoch": 0.4725616629841728, + "grad_norm": 1.798299805037246, + "learning_rate": 2.274006598941997e-06, + "loss": 0.7656, + "step": 5240 + }, + { + "epoch": 0.4726518465076431, + "grad_norm": 1.5771305236743864, + "learning_rate": 2.27342788539297e-06, + "loss": 0.7402, + "step": 5241 + }, + { + "epoch": 0.4727420300311133, + "grad_norm": 1.5372967345311055, + "learning_rate": 2.2728491485135684e-06, + "loss": 0.7156, + "step": 5242 + }, + { + "epoch": 0.4728322135545836, + "grad_norm": 2.004543745873328, + "learning_rate": 2.272270388353173e-06, + "loss": 0.8148, + "step": 5243 + }, + { + "epoch": 0.47292239707805384, + "grad_norm": 1.8590449192474574, + "learning_rate": 2.2716916049611666e-06, + "loss": 0.7344, + "step": 5244 + }, + { + "epoch": 0.4730125806015241, + "grad_norm": 2.198052593445445, + "learning_rate": 2.2711127983869346e-06, + "loss": 0.7392, + "step": 5245 + }, + { + "epoch": 0.47310276412499436, + "grad_norm": 1.4909679762012795, + "learning_rate": 2.270533968679864e-06, + "loss": 0.7746, + "step": 5246 + }, + { + "epoch": 0.47319294764846465, + "grad_norm": 1.5177751999863094, + "learning_rate": 2.269955115889343e-06, + "loss": 0.7655, + "step": 5247 + }, + { + "epoch": 0.4732831311719349, + "grad_norm": 1.6178857153008857, + "learning_rate": 2.269376240064763e-06, + "loss": 0.7243, + "step": 5248 + }, + { + "epoch": 0.47337331469540517, + "grad_norm": 1.6051219556351966, + "learning_rate": 2.268797341255517e-06, + "loss": 0.7789, + "step": 5249 + }, + { + "epoch": 0.4734634982188754, + "grad_norm": 1.8883847922810282, + "learning_rate": 2.268218419511e-06, + "loss": 0.6752, + "step": 5250 + }, + { + "epoch": 0.4735536817423457, + "grad_norm": 3.83119632751805, + "learning_rate": 2.267639474880608e-06, + "loss": 0.6814, + "step": 5251 + }, + { + "epoch": 0.4736438652658159, + "grad_norm": 1.7112306665926453, + "learning_rate": 2.2670605074137407e-06, + "loss": 0.7366, + "step": 5252 + }, + { + "epoch": 0.4737340487892862, + "grad_norm": 1.5788514382098708, + "learning_rate": 2.2664815171597983e-06, + "loss": 0.7428, + "step": 5253 + }, + { + "epoch": 0.47382423231275644, + "grad_norm": 1.684442293029581, + "learning_rate": 2.265902504168183e-06, + "loss": 0.7069, + "step": 5254 + }, + { + "epoch": 0.47391441583622673, + "grad_norm": 1.7581974540505425, + "learning_rate": 2.2653234684883007e-06, + "loss": 0.718, + "step": 5255 + }, + { + "epoch": 0.47400459935969697, + "grad_norm": 2.0644564741354694, + "learning_rate": 2.264744410169556e-06, + "loss": 0.6741, + "step": 5256 + }, + { + "epoch": 0.47409478288316725, + "grad_norm": 1.787109041474525, + "learning_rate": 2.264165329261359e-06, + "loss": 0.7366, + "step": 5257 + }, + { + "epoch": 0.4741849664066375, + "grad_norm": 0.5790731799421794, + "learning_rate": 2.26358622581312e-06, + "loss": 0.5756, + "step": 5258 + }, + { + "epoch": 0.4742751499301078, + "grad_norm": 1.4616951023556062, + "learning_rate": 2.2630070998742504e-06, + "loss": 0.8258, + "step": 5259 + }, + { + "epoch": 0.474365333453578, + "grad_norm": 1.3539962905906229, + "learning_rate": 2.262427951494165e-06, + "loss": 0.7839, + "step": 5260 + }, + { + "epoch": 0.4744555169770483, + "grad_norm": 0.6578086553900719, + "learning_rate": 2.2618487807222794e-06, + "loss": 0.6162, + "step": 5261 + }, + { + "epoch": 0.47454570050051853, + "grad_norm": 1.6101616723766579, + "learning_rate": 2.261269587608012e-06, + "loss": 0.8364, + "step": 5262 + }, + { + "epoch": 0.4746358840239888, + "grad_norm": 2.3561925582562284, + "learning_rate": 2.260690372200783e-06, + "loss": 0.7602, + "step": 5263 + }, + { + "epoch": 0.47472606754745905, + "grad_norm": 1.4402768225295257, + "learning_rate": 2.2601111345500138e-06, + "loss": 0.8479, + "step": 5264 + }, + { + "epoch": 0.47481625107092934, + "grad_norm": 1.4860129705250642, + "learning_rate": 2.2595318747051286e-06, + "loss": 0.7524, + "step": 5265 + }, + { + "epoch": 0.47490643459439963, + "grad_norm": 1.7740472761948702, + "learning_rate": 2.258952592715553e-06, + "loss": 0.7865, + "step": 5266 + }, + { + "epoch": 0.47499661811786986, + "grad_norm": 0.7048476415724183, + "learning_rate": 2.2583732886307142e-06, + "loss": 0.6175, + "step": 5267 + }, + { + "epoch": 0.47508680164134015, + "grad_norm": 1.426705758296014, + "learning_rate": 2.2577939625000414e-06, + "loss": 0.7082, + "step": 5268 + }, + { + "epoch": 0.4751769851648104, + "grad_norm": 0.7084533131655693, + "learning_rate": 2.257214614372967e-06, + "loss": 0.5775, + "step": 5269 + }, + { + "epoch": 0.4752671686882807, + "grad_norm": 1.604822404608261, + "learning_rate": 2.2566352442989227e-06, + "loss": 0.7833, + "step": 5270 + }, + { + "epoch": 0.4753573522117509, + "grad_norm": 3.533021735600917, + "learning_rate": 2.256055852327344e-06, + "loss": 0.7641, + "step": 5271 + }, + { + "epoch": 0.4754475357352212, + "grad_norm": 1.533116429301463, + "learning_rate": 2.2554764385076685e-06, + "loss": 0.7097, + "step": 5272 + }, + { + "epoch": 0.4755377192586914, + "grad_norm": 1.7140291055992336, + "learning_rate": 2.2548970028893348e-06, + "loss": 0.7728, + "step": 5273 + }, + { + "epoch": 0.4756279027821617, + "grad_norm": 1.7983758380570027, + "learning_rate": 2.254317545521783e-06, + "loss": 0.7731, + "step": 5274 + }, + { + "epoch": 0.47571808630563195, + "grad_norm": 1.7798377344557093, + "learning_rate": 2.253738066454457e-06, + "loss": 0.7954, + "step": 5275 + }, + { + "epoch": 0.47580826982910224, + "grad_norm": 1.6179414917230592, + "learning_rate": 2.2531585657367986e-06, + "loss": 0.737, + "step": 5276 + }, + { + "epoch": 0.47589845335257247, + "grad_norm": 1.5988034184697768, + "learning_rate": 2.252579043418256e-06, + "loss": 0.7128, + "step": 5277 + }, + { + "epoch": 0.47598863687604276, + "grad_norm": 1.5727250275067146, + "learning_rate": 2.251999499548277e-06, + "loss": 0.7373, + "step": 5278 + }, + { + "epoch": 0.476078820399513, + "grad_norm": 2.6483903278662213, + "learning_rate": 2.251419934176311e-06, + "loss": 0.7194, + "step": 5279 + }, + { + "epoch": 0.4761690039229833, + "grad_norm": 1.4285846062461156, + "learning_rate": 2.25084034735181e-06, + "loss": 0.8117, + "step": 5280 + }, + { + "epoch": 0.4762591874464535, + "grad_norm": 1.4858063071305627, + "learning_rate": 2.2502607391242274e-06, + "loss": 0.8086, + "step": 5281 + }, + { + "epoch": 0.4763493709699238, + "grad_norm": 0.6663599550505543, + "learning_rate": 2.2496811095430182e-06, + "loss": 0.5831, + "step": 5282 + }, + { + "epoch": 0.47643955449339404, + "grad_norm": 2.5311970646350495, + "learning_rate": 2.249101458657641e-06, + "loss": 0.757, + "step": 5283 + }, + { + "epoch": 0.4765297380168643, + "grad_norm": 3.5909612035097656, + "learning_rate": 2.2485217865175526e-06, + "loss": 0.8282, + "step": 5284 + }, + { + "epoch": 0.47661992154033456, + "grad_norm": 3.0297663882501906, + "learning_rate": 2.2479420931722156e-06, + "loss": 0.8977, + "step": 5285 + }, + { + "epoch": 0.47671010506380485, + "grad_norm": 0.5860969326550182, + "learning_rate": 2.2473623786710923e-06, + "loss": 0.6191, + "step": 5286 + }, + { + "epoch": 0.4768002885872751, + "grad_norm": 1.9316683961050318, + "learning_rate": 2.2467826430636465e-06, + "loss": 0.6737, + "step": 5287 + }, + { + "epoch": 0.47689047211074537, + "grad_norm": 1.630140756134697, + "learning_rate": 2.246202886399345e-06, + "loss": 0.7827, + "step": 5288 + }, + { + "epoch": 0.47698065563421566, + "grad_norm": 2.0048659259365564, + "learning_rate": 2.2456231087276556e-06, + "loss": 0.7579, + "step": 5289 + }, + { + "epoch": 0.4770708391576859, + "grad_norm": 1.5242317672322239, + "learning_rate": 2.245043310098048e-06, + "loss": 0.6863, + "step": 5290 + }, + { + "epoch": 0.4771610226811562, + "grad_norm": 1.785189037731411, + "learning_rate": 2.244463490559995e-06, + "loss": 0.8028, + "step": 5291 + }, + { + "epoch": 0.4772512062046264, + "grad_norm": 1.7029549750056114, + "learning_rate": 2.2438836501629683e-06, + "loss": 0.7749, + "step": 5292 + }, + { + "epoch": 0.4773413897280967, + "grad_norm": 1.4565883914567166, + "learning_rate": 2.2433037889564437e-06, + "loss": 0.7661, + "step": 5293 + }, + { + "epoch": 0.47743157325156693, + "grad_norm": 2.1005811704530983, + "learning_rate": 2.242723906989899e-06, + "loss": 0.7015, + "step": 5294 + }, + { + "epoch": 0.4775217567750372, + "grad_norm": 1.3146733047157853, + "learning_rate": 2.2421440043128114e-06, + "loss": 0.6912, + "step": 5295 + }, + { + "epoch": 0.47761194029850745, + "grad_norm": 2.411566965994482, + "learning_rate": 2.241564080974662e-06, + "loss": 0.6677, + "step": 5296 + }, + { + "epoch": 0.47770212382197774, + "grad_norm": 2.111362227624098, + "learning_rate": 2.2409841370249343e-06, + "loss": 0.7177, + "step": 5297 + }, + { + "epoch": 0.477792307345448, + "grad_norm": 1.7653584236847506, + "learning_rate": 2.2404041725131106e-06, + "loss": 0.848, + "step": 5298 + }, + { + "epoch": 0.47788249086891826, + "grad_norm": 2.173995485401116, + "learning_rate": 2.239824187488677e-06, + "loss": 0.7404, + "step": 5299 + }, + { + "epoch": 0.4779726743923885, + "grad_norm": 0.7344004038211236, + "learning_rate": 2.239244182001122e-06, + "loss": 0.6189, + "step": 5300 + }, + { + "epoch": 0.4780628579158588, + "grad_norm": 1.5307621081289378, + "learning_rate": 2.2386641560999336e-06, + "loss": 0.776, + "step": 5301 + }, + { + "epoch": 0.478153041439329, + "grad_norm": 1.8829789464039974, + "learning_rate": 2.238084109834604e-06, + "loss": 0.7918, + "step": 5302 + }, + { + "epoch": 0.4782432249627993, + "grad_norm": 3.1384506669870293, + "learning_rate": 2.237504043254625e-06, + "loss": 0.7663, + "step": 5303 + }, + { + "epoch": 0.47833340848626954, + "grad_norm": 1.4636986083823473, + "learning_rate": 2.2369239564094915e-06, + "loss": 0.7311, + "step": 5304 + }, + { + "epoch": 0.47842359200973983, + "grad_norm": 12.43692950756624, + "learning_rate": 2.2363438493486995e-06, + "loss": 0.6863, + "step": 5305 + }, + { + "epoch": 0.47851377553321006, + "grad_norm": 2.07524299118022, + "learning_rate": 2.235763722121747e-06, + "loss": 0.6427, + "step": 5306 + }, + { + "epoch": 0.47860395905668035, + "grad_norm": 1.5396320818804543, + "learning_rate": 2.2351835747781346e-06, + "loss": 0.7764, + "step": 5307 + }, + { + "epoch": 0.4786941425801506, + "grad_norm": 1.5153844583504568, + "learning_rate": 2.234603407367362e-06, + "loss": 0.7975, + "step": 5308 + }, + { + "epoch": 0.47878432610362087, + "grad_norm": 1.5970012189398253, + "learning_rate": 2.2340232199389337e-06, + "loss": 0.7746, + "step": 5309 + }, + { + "epoch": 0.4788745096270911, + "grad_norm": 1.686009631942803, + "learning_rate": 2.2334430125423538e-06, + "loss": 0.6422, + "step": 5310 + }, + { + "epoch": 0.4789646931505614, + "grad_norm": 1.7909941926794932, + "learning_rate": 2.232862785227128e-06, + "loss": 0.8056, + "step": 5311 + }, + { + "epoch": 0.4790548766740317, + "grad_norm": 1.71883406000052, + "learning_rate": 2.232282538042766e-06, + "loss": 0.7926, + "step": 5312 + }, + { + "epoch": 0.4791450601975019, + "grad_norm": 1.7153707402718545, + "learning_rate": 2.231702271038777e-06, + "loss": 0.8058, + "step": 5313 + }, + { + "epoch": 0.4792352437209722, + "grad_norm": 1.7694877299163345, + "learning_rate": 2.231121984264673e-06, + "loss": 0.7373, + "step": 5314 + }, + { + "epoch": 0.47932542724444244, + "grad_norm": 2.3723001694609933, + "learning_rate": 2.2305416777699665e-06, + "loss": 0.7931, + "step": 5315 + }, + { + "epoch": 0.4794156107679127, + "grad_norm": 1.717543577393768, + "learning_rate": 2.229961351604173e-06, + "loss": 0.7655, + "step": 5316 + }, + { + "epoch": 0.47950579429138296, + "grad_norm": 1.697517543134644, + "learning_rate": 2.2293810058168085e-06, + "loss": 0.7193, + "step": 5317 + }, + { + "epoch": 0.47959597781485325, + "grad_norm": 1.549307249987717, + "learning_rate": 2.2288006404573922e-06, + "loss": 0.7841, + "step": 5318 + }, + { + "epoch": 0.4796861613383235, + "grad_norm": 1.7790095644817592, + "learning_rate": 2.228220255575444e-06, + "loss": 0.7439, + "step": 5319 + }, + { + "epoch": 0.47977634486179377, + "grad_norm": 6.30242145566559, + "learning_rate": 2.2276398512204847e-06, + "loss": 0.7066, + "step": 5320 + }, + { + "epoch": 0.479866528385264, + "grad_norm": 1.6636060348348085, + "learning_rate": 2.2270594274420382e-06, + "loss": 0.7684, + "step": 5321 + }, + { + "epoch": 0.4799567119087343, + "grad_norm": 0.6728958978981159, + "learning_rate": 2.22647898428963e-06, + "loss": 0.6082, + "step": 5322 + }, + { + "epoch": 0.4800468954322045, + "grad_norm": 1.3385895400138617, + "learning_rate": 2.225898521812785e-06, + "loss": 0.8036, + "step": 5323 + }, + { + "epoch": 0.4801370789556748, + "grad_norm": 1.570545539834516, + "learning_rate": 2.2253180400610337e-06, + "loss": 0.8076, + "step": 5324 + }, + { + "epoch": 0.48022726247914505, + "grad_norm": 3.76107031441492, + "learning_rate": 2.2247375390839037e-06, + "loss": 0.745, + "step": 5325 + }, + { + "epoch": 0.48031744600261533, + "grad_norm": 1.7542027008390746, + "learning_rate": 2.224157018930928e-06, + "loss": 0.7269, + "step": 5326 + }, + { + "epoch": 0.48040762952608557, + "grad_norm": 1.9377924175479864, + "learning_rate": 2.2235764796516395e-06, + "loss": 0.7596, + "step": 5327 + }, + { + "epoch": 0.48049781304955586, + "grad_norm": 1.5693100343761044, + "learning_rate": 2.222995921295573e-06, + "loss": 0.7583, + "step": 5328 + }, + { + "epoch": 0.4805879965730261, + "grad_norm": 3.0041205400443256, + "learning_rate": 2.222415343912265e-06, + "loss": 0.7381, + "step": 5329 + }, + { + "epoch": 0.4806781800964964, + "grad_norm": 1.9183128032806507, + "learning_rate": 2.221834747551254e-06, + "loss": 0.8292, + "step": 5330 + }, + { + "epoch": 0.4807683636199666, + "grad_norm": 1.5827601131344478, + "learning_rate": 2.221254132262078e-06, + "loss": 0.7323, + "step": 5331 + }, + { + "epoch": 0.4808585471434369, + "grad_norm": 1.7258949197451376, + "learning_rate": 2.2206734980942802e-06, + "loss": 0.8011, + "step": 5332 + }, + { + "epoch": 0.48094873066690713, + "grad_norm": 4.019437768440412, + "learning_rate": 2.2200928450974024e-06, + "loss": 0.7336, + "step": 5333 + }, + { + "epoch": 0.4810389141903774, + "grad_norm": 1.6150417374452541, + "learning_rate": 2.21951217332099e-06, + "loss": 0.7857, + "step": 5334 + }, + { + "epoch": 0.48112909771384765, + "grad_norm": 2.0650019642448783, + "learning_rate": 2.2189314828145883e-06, + "loss": 0.6723, + "step": 5335 + }, + { + "epoch": 0.48121928123731794, + "grad_norm": 1.540034495499926, + "learning_rate": 2.2183507736277453e-06, + "loss": 0.7458, + "step": 5336 + }, + { + "epoch": 0.48130946476078823, + "grad_norm": 1.573791597405905, + "learning_rate": 2.2177700458100107e-06, + "loss": 0.7732, + "step": 5337 + }, + { + "epoch": 0.48139964828425846, + "grad_norm": 2.1889059725555717, + "learning_rate": 2.2171892994109346e-06, + "loss": 0.7354, + "step": 5338 + }, + { + "epoch": 0.48148983180772875, + "grad_norm": 1.618645568189062, + "learning_rate": 2.21660853448007e-06, + "loss": 0.8274, + "step": 5339 + }, + { + "epoch": 0.481580015331199, + "grad_norm": 2.549367235250916, + "learning_rate": 2.2160277510669703e-06, + "loss": 0.7169, + "step": 5340 + }, + { + "epoch": 0.4816701988546693, + "grad_norm": 1.37801788502009, + "learning_rate": 2.215446949221193e-06, + "loss": 0.7495, + "step": 5341 + }, + { + "epoch": 0.4817603823781395, + "grad_norm": 1.666847441724027, + "learning_rate": 2.2148661289922924e-06, + "loss": 0.7711, + "step": 5342 + }, + { + "epoch": 0.4818505659016098, + "grad_norm": 2.614956853156868, + "learning_rate": 2.21428529042983e-06, + "loss": 0.8252, + "step": 5343 + }, + { + "epoch": 0.48194074942508003, + "grad_norm": 3.1919003201076657, + "learning_rate": 2.2137044335833647e-06, + "loss": 0.8097, + "step": 5344 + }, + { + "epoch": 0.4820309329485503, + "grad_norm": 2.1394753814871303, + "learning_rate": 2.213123558502459e-06, + "loss": 0.7426, + "step": 5345 + }, + { + "epoch": 0.48212111647202055, + "grad_norm": 1.7831005054373252, + "learning_rate": 2.2125426652366763e-06, + "loss": 0.7773, + "step": 5346 + }, + { + "epoch": 0.48221129999549084, + "grad_norm": 1.6529738935842555, + "learning_rate": 2.211961753835581e-06, + "loss": 0.8535, + "step": 5347 + }, + { + "epoch": 0.48230148351896107, + "grad_norm": 1.3217616057401396, + "learning_rate": 2.21138082434874e-06, + "loss": 0.7511, + "step": 5348 + }, + { + "epoch": 0.48239166704243136, + "grad_norm": 3.9748453506040033, + "learning_rate": 2.210799876825722e-06, + "loss": 0.7921, + "step": 5349 + }, + { + "epoch": 0.4824818505659016, + "grad_norm": 1.6328909093109083, + "learning_rate": 2.210218911316096e-06, + "loss": 0.7216, + "step": 5350 + }, + { + "epoch": 0.4825720340893719, + "grad_norm": 1.7354648362061762, + "learning_rate": 2.2096379278694336e-06, + "loss": 0.7105, + "step": 5351 + }, + { + "epoch": 0.4826622176128421, + "grad_norm": 1.5730483482094386, + "learning_rate": 2.2090569265353074e-06, + "loss": 0.7047, + "step": 5352 + }, + { + "epoch": 0.4827524011363124, + "grad_norm": 2.7367048528363607, + "learning_rate": 2.2084759073632912e-06, + "loss": 0.761, + "step": 5353 + }, + { + "epoch": 0.48284258465978264, + "grad_norm": 2.2501502516699485, + "learning_rate": 2.2078948704029606e-06, + "loss": 0.7203, + "step": 5354 + }, + { + "epoch": 0.4829327681832529, + "grad_norm": 1.6325890055018808, + "learning_rate": 2.2073138157038935e-06, + "loss": 0.6847, + "step": 5355 + }, + { + "epoch": 0.48302295170672316, + "grad_norm": 1.6437932258985275, + "learning_rate": 2.2067327433156687e-06, + "loss": 0.7505, + "step": 5356 + }, + { + "epoch": 0.48311313523019345, + "grad_norm": 1.7084917677317337, + "learning_rate": 2.2061516532878667e-06, + "loss": 0.7756, + "step": 5357 + }, + { + "epoch": 0.4832033187536637, + "grad_norm": 1.5102430923523693, + "learning_rate": 2.2055705456700686e-06, + "loss": 0.8048, + "step": 5358 + }, + { + "epoch": 0.48329350227713397, + "grad_norm": 1.9438634170855336, + "learning_rate": 2.204989420511858e-06, + "loss": 0.6561, + "step": 5359 + }, + { + "epoch": 0.48338368580060426, + "grad_norm": 1.665299148606266, + "learning_rate": 2.20440827786282e-06, + "loss": 0.7049, + "step": 5360 + }, + { + "epoch": 0.4834738693240745, + "grad_norm": 1.6166132400310311, + "learning_rate": 2.20382711777254e-06, + "loss": 0.7628, + "step": 5361 + }, + { + "epoch": 0.4835640528475448, + "grad_norm": 1.679896922695907, + "learning_rate": 2.203245940290607e-06, + "loss": 0.7674, + "step": 5362 + }, + { + "epoch": 0.483654236371015, + "grad_norm": 1.9398864692227509, + "learning_rate": 2.2026647454666097e-06, + "loss": 0.7576, + "step": 5363 + }, + { + "epoch": 0.4837444198944853, + "grad_norm": 1.8691450256487185, + "learning_rate": 2.2020835333501384e-06, + "loss": 0.7936, + "step": 5364 + }, + { + "epoch": 0.48383460341795553, + "grad_norm": 1.59860290984079, + "learning_rate": 2.2015023039907863e-06, + "loss": 0.8177, + "step": 5365 + }, + { + "epoch": 0.4839247869414258, + "grad_norm": 1.4297271889769398, + "learning_rate": 2.2009210574381464e-06, + "loss": 0.7394, + "step": 5366 + }, + { + "epoch": 0.48401497046489605, + "grad_norm": 1.4620087312001904, + "learning_rate": 2.2003397937418134e-06, + "loss": 0.7049, + "step": 5367 + }, + { + "epoch": 0.48410515398836634, + "grad_norm": 0.6786063353802954, + "learning_rate": 2.1997585129513852e-06, + "loss": 0.603, + "step": 5368 + }, + { + "epoch": 0.4841953375118366, + "grad_norm": 1.9083950293727883, + "learning_rate": 2.1991772151164595e-06, + "loss": 0.7814, + "step": 5369 + }, + { + "epoch": 0.48428552103530687, + "grad_norm": 1.4723013290808649, + "learning_rate": 2.1985959002866346e-06, + "loss": 0.8039, + "step": 5370 + }, + { + "epoch": 0.4843757045587771, + "grad_norm": 2.623527568023062, + "learning_rate": 2.198014568511513e-06, + "loss": 0.8302, + "step": 5371 + }, + { + "epoch": 0.4844658880822474, + "grad_norm": 0.6655347995802217, + "learning_rate": 2.1974332198406965e-06, + "loss": 0.5754, + "step": 5372 + }, + { + "epoch": 0.4845560716057176, + "grad_norm": 1.54996536277709, + "learning_rate": 2.196851854323789e-06, + "loss": 0.7421, + "step": 5373 + }, + { + "epoch": 0.4846462551291879, + "grad_norm": 1.7785171156888933, + "learning_rate": 2.196270472010396e-06, + "loss": 0.7431, + "step": 5374 + }, + { + "epoch": 0.48473643865265814, + "grad_norm": 1.5744518930906133, + "learning_rate": 2.195689072950124e-06, + "loss": 0.7574, + "step": 5375 + }, + { + "epoch": 0.48482662217612843, + "grad_norm": 0.5868952235595739, + "learning_rate": 2.195107657192581e-06, + "loss": 0.5633, + "step": 5376 + }, + { + "epoch": 0.48491680569959866, + "grad_norm": 0.596459555718487, + "learning_rate": 2.194526224787378e-06, + "loss": 0.5175, + "step": 5377 + }, + { + "epoch": 0.48500698922306895, + "grad_norm": 2.145895046593582, + "learning_rate": 2.1939447757841236e-06, + "loss": 0.7264, + "step": 5378 + }, + { + "epoch": 0.4850971727465392, + "grad_norm": 2.257328073876629, + "learning_rate": 2.193363310232432e-06, + "loss": 0.8197, + "step": 5379 + }, + { + "epoch": 0.4851873562700095, + "grad_norm": 1.324523080609585, + "learning_rate": 2.192781828181917e-06, + "loss": 0.8575, + "step": 5380 + }, + { + "epoch": 0.4852775397934797, + "grad_norm": 1.3454647212909014, + "learning_rate": 2.192200329682193e-06, + "loss": 0.7805, + "step": 5381 + }, + { + "epoch": 0.48536772331695, + "grad_norm": 1.4803833144203347, + "learning_rate": 2.1916188147828767e-06, + "loss": 0.7566, + "step": 5382 + }, + { + "epoch": 0.48545790684042023, + "grad_norm": 1.4884414386396818, + "learning_rate": 2.191037283533587e-06, + "loss": 0.7718, + "step": 5383 + }, + { + "epoch": 0.4855480903638905, + "grad_norm": 2.365041735968377, + "learning_rate": 2.1904557359839428e-06, + "loss": 0.7561, + "step": 5384 + }, + { + "epoch": 0.4856382738873608, + "grad_norm": 1.9681943684597676, + "learning_rate": 2.189874172183565e-06, + "loss": 0.7339, + "step": 5385 + }, + { + "epoch": 0.48572845741083104, + "grad_norm": 1.7328880052824904, + "learning_rate": 2.1892925921820763e-06, + "loss": 0.7272, + "step": 5386 + }, + { + "epoch": 0.4858186409343013, + "grad_norm": 1.6981141457265185, + "learning_rate": 2.1887109960290994e-06, + "loss": 0.7559, + "step": 5387 + }, + { + "epoch": 0.48590882445777156, + "grad_norm": 2.3557234036526022, + "learning_rate": 2.18812938377426e-06, + "loss": 0.8523, + "step": 5388 + }, + { + "epoch": 0.48599900798124185, + "grad_norm": 1.6948657172245882, + "learning_rate": 2.187547755467184e-06, + "loss": 0.7939, + "step": 5389 + }, + { + "epoch": 0.4860891915047121, + "grad_norm": 1.7736956034646982, + "learning_rate": 2.1869661111574994e-06, + "loss": 0.6525, + "step": 5390 + }, + { + "epoch": 0.48617937502818237, + "grad_norm": 1.8084252033834023, + "learning_rate": 2.1863844508948353e-06, + "loss": 0.8056, + "step": 5391 + }, + { + "epoch": 0.4862695585516526, + "grad_norm": 1.45525810613506, + "learning_rate": 2.185802774728823e-06, + "loss": 0.7736, + "step": 5392 + }, + { + "epoch": 0.4863597420751229, + "grad_norm": 1.6898251515289586, + "learning_rate": 2.1852210827090927e-06, + "loss": 0.7102, + "step": 5393 + }, + { + "epoch": 0.4864499255985931, + "grad_norm": 1.509758913486659, + "learning_rate": 2.184639374885278e-06, + "loss": 0.6983, + "step": 5394 + }, + { + "epoch": 0.4865401091220634, + "grad_norm": 1.9501462270245158, + "learning_rate": 2.184057651307014e-06, + "loss": 0.8551, + "step": 5395 + }, + { + "epoch": 0.48663029264553365, + "grad_norm": 1.9945103884426656, + "learning_rate": 2.183475912023937e-06, + "loss": 0.8318, + "step": 5396 + }, + { + "epoch": 0.48672047616900393, + "grad_norm": 1.875598176270532, + "learning_rate": 2.1828941570856826e-06, + "loss": 0.7745, + "step": 5397 + }, + { + "epoch": 0.48681065969247417, + "grad_norm": 0.6432893289906996, + "learning_rate": 2.1823123865418903e-06, + "loss": 0.5727, + "step": 5398 + }, + { + "epoch": 0.48690084321594446, + "grad_norm": 2.136998466858874, + "learning_rate": 2.1817306004422e-06, + "loss": 0.7312, + "step": 5399 + }, + { + "epoch": 0.4869910267394147, + "grad_norm": 1.681166312818634, + "learning_rate": 2.1811487988362527e-06, + "loss": 0.7833, + "step": 5400 + }, + { + "epoch": 0.487081210262885, + "grad_norm": 1.7857545480276373, + "learning_rate": 2.1805669817736917e-06, + "loss": 0.8309, + "step": 5401 + }, + { + "epoch": 0.4871713937863552, + "grad_norm": 1.7640534555289482, + "learning_rate": 2.17998514930416e-06, + "loss": 0.5781, + "step": 5402 + }, + { + "epoch": 0.4872615773098255, + "grad_norm": 2.442748263967395, + "learning_rate": 2.1794033014773025e-06, + "loss": 0.7608, + "step": 5403 + }, + { + "epoch": 0.48735176083329573, + "grad_norm": 1.589032315217901, + "learning_rate": 2.178821438342766e-06, + "loss": 0.824, + "step": 5404 + }, + { + "epoch": 0.487441944356766, + "grad_norm": 1.7231494431116237, + "learning_rate": 2.1782395599501996e-06, + "loss": 0.7477, + "step": 5405 + }, + { + "epoch": 0.48753212788023625, + "grad_norm": 0.6561488573152324, + "learning_rate": 2.1776576663492498e-06, + "loss": 0.5844, + "step": 5406 + }, + { + "epoch": 0.48762231140370654, + "grad_norm": 2.202786899893581, + "learning_rate": 2.177075757589569e-06, + "loss": 0.8442, + "step": 5407 + }, + { + "epoch": 0.48771249492717683, + "grad_norm": 1.5880635860598091, + "learning_rate": 2.176493833720808e-06, + "loss": 0.6841, + "step": 5408 + }, + { + "epoch": 0.48780267845064706, + "grad_norm": 1.6179869514005185, + "learning_rate": 2.1759118947926195e-06, + "loss": 0.7245, + "step": 5409 + }, + { + "epoch": 0.48789286197411735, + "grad_norm": 1.8044194274231964, + "learning_rate": 2.1753299408546587e-06, + "loss": 0.7687, + "step": 5410 + }, + { + "epoch": 0.4879830454975876, + "grad_norm": 1.526250111097977, + "learning_rate": 2.1747479719565803e-06, + "loss": 0.7626, + "step": 5411 + }, + { + "epoch": 0.4880732290210579, + "grad_norm": 1.5137663780983304, + "learning_rate": 2.174165988148042e-06, + "loss": 0.821, + "step": 5412 + }, + { + "epoch": 0.4881634125445281, + "grad_norm": 1.4914880681163272, + "learning_rate": 2.1735839894787003e-06, + "loss": 0.7241, + "step": 5413 + }, + { + "epoch": 0.4882535960679984, + "grad_norm": 0.7666466183390729, + "learning_rate": 2.1730019759982163e-06, + "loss": 0.6291, + "step": 5414 + }, + { + "epoch": 0.48834377959146863, + "grad_norm": 1.5257501508177336, + "learning_rate": 2.172419947756249e-06, + "loss": 0.8084, + "step": 5415 + }, + { + "epoch": 0.4884339631149389, + "grad_norm": 2.002145570018554, + "learning_rate": 2.171837904802461e-06, + "loss": 0.809, + "step": 5416 + }, + { + "epoch": 0.48852414663840915, + "grad_norm": 2.0704085417729012, + "learning_rate": 2.171255847186516e-06, + "loss": 0.7538, + "step": 5417 + }, + { + "epoch": 0.48861433016187944, + "grad_norm": 1.8524569511817124, + "learning_rate": 2.1706737749580783e-06, + "loss": 0.7436, + "step": 5418 + }, + { + "epoch": 0.4887045136853497, + "grad_norm": 1.5739045314652569, + "learning_rate": 2.1700916881668127e-06, + "loss": 0.8291, + "step": 5419 + }, + { + "epoch": 0.48879469720881996, + "grad_norm": 1.9131339798834637, + "learning_rate": 2.1695095868623862e-06, + "loss": 0.8484, + "step": 5420 + }, + { + "epoch": 0.4888848807322902, + "grad_norm": 1.6032133200783087, + "learning_rate": 2.168927471094467e-06, + "loss": 0.6936, + "step": 5421 + }, + { + "epoch": 0.4889750642557605, + "grad_norm": 1.940382382373495, + "learning_rate": 2.168345340912725e-06, + "loss": 0.7837, + "step": 5422 + }, + { + "epoch": 0.4890652477792307, + "grad_norm": 1.9940163987460127, + "learning_rate": 2.1677631963668298e-06, + "loss": 0.7781, + "step": 5423 + }, + { + "epoch": 0.489155431302701, + "grad_norm": 2.0502467448907957, + "learning_rate": 2.167181037506453e-06, + "loss": 0.7161, + "step": 5424 + }, + { + "epoch": 0.48924561482617124, + "grad_norm": 1.5055932033628643, + "learning_rate": 2.1665988643812693e-06, + "loss": 0.6971, + "step": 5425 + }, + { + "epoch": 0.4893357983496415, + "grad_norm": 1.7567910850783361, + "learning_rate": 2.166016677040951e-06, + "loss": 0.757, + "step": 5426 + }, + { + "epoch": 0.48942598187311176, + "grad_norm": 1.7338986129159415, + "learning_rate": 2.165434475535175e-06, + "loss": 0.7493, + "step": 5427 + }, + { + "epoch": 0.48951616539658205, + "grad_norm": 2.287745490737421, + "learning_rate": 2.1648522599136173e-06, + "loss": 0.8013, + "step": 5428 + }, + { + "epoch": 0.4896063489200523, + "grad_norm": 1.382349486340728, + "learning_rate": 2.164270030225956e-06, + "loss": 0.7308, + "step": 5429 + }, + { + "epoch": 0.48969653244352257, + "grad_norm": 1.358377331721259, + "learning_rate": 2.16368778652187e-06, + "loss": 0.6683, + "step": 5430 + }, + { + "epoch": 0.4897867159669928, + "grad_norm": 1.6918970677740726, + "learning_rate": 2.163105528851039e-06, + "loss": 0.7993, + "step": 5431 + }, + { + "epoch": 0.4898768994904631, + "grad_norm": 1.5663311338607457, + "learning_rate": 2.1625232572631448e-06, + "loss": 0.7619, + "step": 5432 + }, + { + "epoch": 0.4899670830139334, + "grad_norm": 0.7454986595689274, + "learning_rate": 2.161940971807871e-06, + "loss": 0.5915, + "step": 5433 + }, + { + "epoch": 0.4900572665374036, + "grad_norm": 2.3291476263133544, + "learning_rate": 2.1613586725348994e-06, + "loss": 0.7467, + "step": 5434 + }, + { + "epoch": 0.4901474500608739, + "grad_norm": 1.5544757794701565, + "learning_rate": 2.1607763594939176e-06, + "loss": 0.8378, + "step": 5435 + }, + { + "epoch": 0.49023763358434413, + "grad_norm": 1.5083620325101543, + "learning_rate": 2.1601940327346093e-06, + "loss": 0.6686, + "step": 5436 + }, + { + "epoch": 0.4903278171078144, + "grad_norm": 1.4936912908149629, + "learning_rate": 2.159611692306663e-06, + "loss": 0.8317, + "step": 5437 + }, + { + "epoch": 0.49041800063128466, + "grad_norm": 2.2283601698285382, + "learning_rate": 2.1590293382597667e-06, + "loss": 0.7183, + "step": 5438 + }, + { + "epoch": 0.49050818415475494, + "grad_norm": 1.7981629109819701, + "learning_rate": 2.1584469706436102e-06, + "loss": 0.7008, + "step": 5439 + }, + { + "epoch": 0.4905983676782252, + "grad_norm": 1.5037373237360963, + "learning_rate": 2.1578645895078855e-06, + "loss": 0.6719, + "step": 5440 + }, + { + "epoch": 0.49068855120169547, + "grad_norm": 1.742450369260817, + "learning_rate": 2.157282194902283e-06, + "loss": 0.8169, + "step": 5441 + }, + { + "epoch": 0.4907787347251657, + "grad_norm": 1.8081239291945135, + "learning_rate": 2.1566997868764965e-06, + "loss": 0.6965, + "step": 5442 + }, + { + "epoch": 0.490868918248636, + "grad_norm": 1.7331210578023466, + "learning_rate": 2.15611736548022e-06, + "loss": 0.7214, + "step": 5443 + }, + { + "epoch": 0.4909591017721062, + "grad_norm": 1.6212941241234158, + "learning_rate": 2.155534930763149e-06, + "loss": 0.7706, + "step": 5444 + }, + { + "epoch": 0.4910492852955765, + "grad_norm": 2.317489396516478, + "learning_rate": 2.1549524827749804e-06, + "loss": 0.6421, + "step": 5445 + }, + { + "epoch": 0.49113946881904674, + "grad_norm": 4.205524879745709, + "learning_rate": 2.1543700215654115e-06, + "loss": 0.6093, + "step": 5446 + }, + { + "epoch": 0.49122965234251703, + "grad_norm": 1.5111164795711622, + "learning_rate": 2.153787547184141e-06, + "loss": 0.7482, + "step": 5447 + }, + { + "epoch": 0.49131983586598726, + "grad_norm": 1.5048809905539344, + "learning_rate": 2.1532050596808695e-06, + "loss": 0.6848, + "step": 5448 + }, + { + "epoch": 0.49141001938945755, + "grad_norm": 1.857792284882002, + "learning_rate": 2.152622559105297e-06, + "loss": 0.806, + "step": 5449 + }, + { + "epoch": 0.4915002029129278, + "grad_norm": 1.7587136226489772, + "learning_rate": 2.152040045507126e-06, + "loss": 0.8048, + "step": 5450 + }, + { + "epoch": 0.4915903864363981, + "grad_norm": 2.760449669598678, + "learning_rate": 2.1514575189360607e-06, + "loss": 0.7726, + "step": 5451 + }, + { + "epoch": 0.4916805699598683, + "grad_norm": 1.6466023865533161, + "learning_rate": 2.1508749794418043e-06, + "loss": 0.7165, + "step": 5452 + }, + { + "epoch": 0.4917707534833386, + "grad_norm": 1.619929766721983, + "learning_rate": 2.1502924270740626e-06, + "loss": 0.7272, + "step": 5453 + }, + { + "epoch": 0.49186093700680883, + "grad_norm": 1.8859857434515679, + "learning_rate": 2.1497098618825427e-06, + "loss": 0.7632, + "step": 5454 + }, + { + "epoch": 0.4919511205302791, + "grad_norm": 1.789929775267056, + "learning_rate": 2.1491272839169516e-06, + "loss": 0.8031, + "step": 5455 + }, + { + "epoch": 0.4920413040537494, + "grad_norm": 1.5731439826192863, + "learning_rate": 2.1485446932269986e-06, + "loss": 0.8734, + "step": 5456 + }, + { + "epoch": 0.49213148757721964, + "grad_norm": 1.6481865425177964, + "learning_rate": 2.147962089862393e-06, + "loss": 0.7953, + "step": 5457 + }, + { + "epoch": 0.49222167110068993, + "grad_norm": 2.3003013123223495, + "learning_rate": 2.1473794738728462e-06, + "loss": 0.6641, + "step": 5458 + }, + { + "epoch": 0.49231185462416016, + "grad_norm": 1.9044173764959025, + "learning_rate": 2.14679684530807e-06, + "loss": 0.7894, + "step": 5459 + }, + { + "epoch": 0.49240203814763045, + "grad_norm": 1.6288834897287656, + "learning_rate": 2.1462142042177774e-06, + "loss": 0.738, + "step": 5460 + }, + { + "epoch": 0.4924922216711007, + "grad_norm": 1.5247597458242272, + "learning_rate": 2.145631550651683e-06, + "loss": 0.753, + "step": 5461 + }, + { + "epoch": 0.49258240519457097, + "grad_norm": 1.7055859765285033, + "learning_rate": 2.1450488846595016e-06, + "loss": 0.8558, + "step": 5462 + }, + { + "epoch": 0.4926725887180412, + "grad_norm": 1.7056243476013748, + "learning_rate": 2.14446620629095e-06, + "loss": 0.7397, + "step": 5463 + }, + { + "epoch": 0.4927627722415115, + "grad_norm": 1.517949082290843, + "learning_rate": 2.1438835155957445e-06, + "loss": 0.8237, + "step": 5464 + }, + { + "epoch": 0.4928529557649817, + "grad_norm": 1.7904772765787749, + "learning_rate": 2.143300812623604e-06, + "loss": 0.7966, + "step": 5465 + }, + { + "epoch": 0.492943139288452, + "grad_norm": 1.71062456169304, + "learning_rate": 2.1427180974242485e-06, + "loss": 0.7435, + "step": 5466 + }, + { + "epoch": 0.49303332281192225, + "grad_norm": 45.17411789658294, + "learning_rate": 2.142135370047398e-06, + "loss": 0.7844, + "step": 5467 + }, + { + "epoch": 0.49312350633539254, + "grad_norm": 1.9371602775776662, + "learning_rate": 2.1415526305427735e-06, + "loss": 0.7903, + "step": 5468 + }, + { + "epoch": 0.49321368985886277, + "grad_norm": 1.3648215612519563, + "learning_rate": 2.140969878960098e-06, + "loss": 0.681, + "step": 5469 + }, + { + "epoch": 0.49330387338233306, + "grad_norm": 2.467745516233764, + "learning_rate": 2.1403871153490956e-06, + "loss": 0.8174, + "step": 5470 + }, + { + "epoch": 0.4933940569058033, + "grad_norm": 3.9427318631086945, + "learning_rate": 2.13980433975949e-06, + "loss": 0.8781, + "step": 5471 + }, + { + "epoch": 0.4934842404292736, + "grad_norm": 3.8637292677373396, + "learning_rate": 2.1392215522410076e-06, + "loss": 0.8045, + "step": 5472 + }, + { + "epoch": 0.4935744239527438, + "grad_norm": 2.1354761503044903, + "learning_rate": 2.1386387528433743e-06, + "loss": 0.8739, + "step": 5473 + }, + { + "epoch": 0.4936646074762141, + "grad_norm": 2.0481805496202656, + "learning_rate": 2.1380559416163186e-06, + "loss": 0.7901, + "step": 5474 + }, + { + "epoch": 0.49375479099968433, + "grad_norm": 1.4711694472048134, + "learning_rate": 2.1374731186095685e-06, + "loss": 0.8058, + "step": 5475 + }, + { + "epoch": 0.4938449745231546, + "grad_norm": 1.6361552795783816, + "learning_rate": 2.136890283872854e-06, + "loss": 0.7129, + "step": 5476 + }, + { + "epoch": 0.49393515804662486, + "grad_norm": 2.206862937788408, + "learning_rate": 2.136307437455906e-06, + "loss": 0.7402, + "step": 5477 + }, + { + "epoch": 0.49402534157009514, + "grad_norm": 1.422756215437372, + "learning_rate": 2.135724579408456e-06, + "loss": 0.7047, + "step": 5478 + }, + { + "epoch": 0.49411552509356543, + "grad_norm": 5.2957944414865175, + "learning_rate": 2.1351417097802356e-06, + "loss": 0.8546, + "step": 5479 + }, + { + "epoch": 0.49420570861703567, + "grad_norm": 1.9453120097577195, + "learning_rate": 2.1345588286209798e-06, + "loss": 0.6427, + "step": 5480 + }, + { + "epoch": 0.49429589214050595, + "grad_norm": 2.1329703359924235, + "learning_rate": 2.1339759359804227e-06, + "loss": 0.6058, + "step": 5481 + }, + { + "epoch": 0.4943860756639762, + "grad_norm": 1.8780961223292731, + "learning_rate": 2.1333930319082997e-06, + "loss": 0.7358, + "step": 5482 + }, + { + "epoch": 0.4944762591874465, + "grad_norm": 1.5028758296524825, + "learning_rate": 2.132810116454348e-06, + "loss": 0.7626, + "step": 5483 + }, + { + "epoch": 0.4945664427109167, + "grad_norm": 3.0850185909360253, + "learning_rate": 2.132227189668305e-06, + "loss": 0.7513, + "step": 5484 + }, + { + "epoch": 0.494656626234387, + "grad_norm": 1.7050684768036737, + "learning_rate": 2.1316442515999096e-06, + "loss": 0.701, + "step": 5485 + }, + { + "epoch": 0.49474680975785723, + "grad_norm": 1.6594169137262615, + "learning_rate": 2.1310613022989e-06, + "loss": 0.7873, + "step": 5486 + }, + { + "epoch": 0.4948369932813275, + "grad_norm": 1.6666074821772066, + "learning_rate": 2.130478341815017e-06, + "loss": 0.6064, + "step": 5487 + }, + { + "epoch": 0.49492717680479775, + "grad_norm": 1.719026577977623, + "learning_rate": 2.1298953701980033e-06, + "loss": 0.6603, + "step": 5488 + }, + { + "epoch": 0.49501736032826804, + "grad_norm": 2.3740841957015824, + "learning_rate": 2.1293123874976003e-06, + "loss": 0.7637, + "step": 5489 + }, + { + "epoch": 0.4951075438517383, + "grad_norm": 1.7236436708448701, + "learning_rate": 2.1287293937635513e-06, + "loss": 0.7099, + "step": 5490 + }, + { + "epoch": 0.49519772737520856, + "grad_norm": 1.719648577142021, + "learning_rate": 2.1281463890456005e-06, + "loss": 0.8115, + "step": 5491 + }, + { + "epoch": 0.4952879108986788, + "grad_norm": 1.697896578929314, + "learning_rate": 2.127563373393493e-06, + "loss": 0.7496, + "step": 5492 + }, + { + "epoch": 0.4953780944221491, + "grad_norm": 1.9086633632860945, + "learning_rate": 2.1269803468569756e-06, + "loss": 0.7512, + "step": 5493 + }, + { + "epoch": 0.4954682779456193, + "grad_norm": 1.620633347177543, + "learning_rate": 2.126397309485794e-06, + "loss": 0.7941, + "step": 5494 + }, + { + "epoch": 0.4955584614690896, + "grad_norm": 2.0040287924733047, + "learning_rate": 2.1258142613296983e-06, + "loss": 0.72, + "step": 5495 + }, + { + "epoch": 0.49564864499255984, + "grad_norm": 1.4043661426762297, + "learning_rate": 2.125231202438435e-06, + "loss": 0.7282, + "step": 5496 + }, + { + "epoch": 0.49573882851603013, + "grad_norm": 2.1263489648930416, + "learning_rate": 2.1246481328617553e-06, + "loss": 0.7904, + "step": 5497 + }, + { + "epoch": 0.49582901203950036, + "grad_norm": 2.007939791076875, + "learning_rate": 2.1240650526494096e-06, + "loss": 0.7404, + "step": 5498 + }, + { + "epoch": 0.49591919556297065, + "grad_norm": 1.7280446590904517, + "learning_rate": 2.1234819618511493e-06, + "loss": 0.7316, + "step": 5499 + }, + { + "epoch": 0.4960093790864409, + "grad_norm": 0.6980849509162584, + "learning_rate": 2.122898860516728e-06, + "loss": 0.6191, + "step": 5500 + }, + { + "epoch": 0.49609956260991117, + "grad_norm": 1.753839028478478, + "learning_rate": 2.1223157486958976e-06, + "loss": 0.7441, + "step": 5501 + }, + { + "epoch": 0.4961897461333814, + "grad_norm": 1.432386508460318, + "learning_rate": 2.1217326264384127e-06, + "loss": 0.6459, + "step": 5502 + }, + { + "epoch": 0.4962799296568517, + "grad_norm": 2.242340082698936, + "learning_rate": 2.1211494937940296e-06, + "loss": 0.7752, + "step": 5503 + }, + { + "epoch": 0.496370113180322, + "grad_norm": 1.5492932461877917, + "learning_rate": 2.1205663508125034e-06, + "loss": 0.7696, + "step": 5504 + }, + { + "epoch": 0.4964602967037922, + "grad_norm": 1.4543143706559079, + "learning_rate": 2.1199831975435914e-06, + "loss": 0.7186, + "step": 5505 + }, + { + "epoch": 0.4965504802272625, + "grad_norm": 2.5843381978266406, + "learning_rate": 2.1194000340370517e-06, + "loss": 0.8033, + "step": 5506 + }, + { + "epoch": 0.49664066375073274, + "grad_norm": 1.7923370777824346, + "learning_rate": 2.1188168603426423e-06, + "loss": 0.7035, + "step": 5507 + }, + { + "epoch": 0.496730847274203, + "grad_norm": 2.492469317248322, + "learning_rate": 2.118233676510123e-06, + "loss": 0.7219, + "step": 5508 + }, + { + "epoch": 0.49682103079767326, + "grad_norm": 1.6742755376500937, + "learning_rate": 2.117650482589255e-06, + "loss": 0.7665, + "step": 5509 + }, + { + "epoch": 0.49691121432114355, + "grad_norm": 2.3823151616610216, + "learning_rate": 2.1170672786297988e-06, + "loss": 0.7956, + "step": 5510 + }, + { + "epoch": 0.4970013978446138, + "grad_norm": 1.407248121595459, + "learning_rate": 2.1164840646815174e-06, + "loss": 0.7271, + "step": 5511 + }, + { + "epoch": 0.49709158136808407, + "grad_norm": 2.9874212087033274, + "learning_rate": 2.1159008407941726e-06, + "loss": 0.8382, + "step": 5512 + }, + { + "epoch": 0.4971817648915543, + "grad_norm": 1.7462629197884483, + "learning_rate": 2.1153176070175293e-06, + "loss": 0.7505, + "step": 5513 + }, + { + "epoch": 0.4972719484150246, + "grad_norm": 1.478245820154235, + "learning_rate": 2.114734363401352e-06, + "loss": 0.6478, + "step": 5514 + }, + { + "epoch": 0.4973621319384948, + "grad_norm": 2.2104461447439694, + "learning_rate": 2.1141511099954056e-06, + "loss": 0.7546, + "step": 5515 + }, + { + "epoch": 0.4974523154619651, + "grad_norm": 1.626563933534018, + "learning_rate": 2.1135678468494576e-06, + "loss": 0.7388, + "step": 5516 + }, + { + "epoch": 0.49754249898543534, + "grad_norm": 2.1298024382330207, + "learning_rate": 2.112984574013275e-06, + "loss": 0.794, + "step": 5517 + }, + { + "epoch": 0.49763268250890563, + "grad_norm": 1.7205011636807122, + "learning_rate": 2.112401291536625e-06, + "loss": 0.7142, + "step": 5518 + }, + { + "epoch": 0.49772286603237587, + "grad_norm": 1.5354883354089277, + "learning_rate": 2.111817999469278e-06, + "loss": 0.792, + "step": 5519 + }, + { + "epoch": 0.49781304955584615, + "grad_norm": 1.841890205292061, + "learning_rate": 2.1112346978610016e-06, + "loss": 0.7125, + "step": 5520 + }, + { + "epoch": 0.4979032330793164, + "grad_norm": 1.5715626157835294, + "learning_rate": 2.1106513867615678e-06, + "loss": 0.7064, + "step": 5521 + }, + { + "epoch": 0.4979934166027867, + "grad_norm": 2.668833021529796, + "learning_rate": 2.110068066220748e-06, + "loss": 0.7572, + "step": 5522 + }, + { + "epoch": 0.4980836001262569, + "grad_norm": 1.5640141589214085, + "learning_rate": 2.109484736288313e-06, + "loss": 0.8238, + "step": 5523 + }, + { + "epoch": 0.4981737836497272, + "grad_norm": 1.8018669777650909, + "learning_rate": 2.108901397014037e-06, + "loss": 0.774, + "step": 5524 + }, + { + "epoch": 0.49826396717319743, + "grad_norm": 1.4289834211814616, + "learning_rate": 2.1083180484476934e-06, + "loss": 0.7596, + "step": 5525 + }, + { + "epoch": 0.4983541506966677, + "grad_norm": 1.9571155522324273, + "learning_rate": 2.1077346906390567e-06, + "loss": 0.7791, + "step": 5526 + }, + { + "epoch": 0.498444334220138, + "grad_norm": 1.7699954200944, + "learning_rate": 2.107151323637902e-06, + "loss": 0.7742, + "step": 5527 + }, + { + "epoch": 0.49853451774360824, + "grad_norm": 1.8450342651741587, + "learning_rate": 2.106567947494006e-06, + "loss": 0.7397, + "step": 5528 + }, + { + "epoch": 0.49862470126707853, + "grad_norm": 3.160108115604137, + "learning_rate": 2.1059845622571447e-06, + "loss": 0.8242, + "step": 5529 + }, + { + "epoch": 0.49871488479054876, + "grad_norm": 1.551723395818774, + "learning_rate": 2.1054011679770956e-06, + "loss": 0.7955, + "step": 5530 + }, + { + "epoch": 0.49880506831401905, + "grad_norm": 1.6905946965066598, + "learning_rate": 2.104817764703638e-06, + "loss": 0.766, + "step": 5531 + }, + { + "epoch": 0.4988952518374893, + "grad_norm": 1.953855759289754, + "learning_rate": 2.1042343524865516e-06, + "loss": 0.7622, + "step": 5532 + }, + { + "epoch": 0.4989854353609596, + "grad_norm": 1.5379627771871014, + "learning_rate": 2.103650931375615e-06, + "loss": 0.8655, + "step": 5533 + }, + { + "epoch": 0.4990756188844298, + "grad_norm": 1.8795404137246798, + "learning_rate": 2.1030675014206094e-06, + "loss": 0.6934, + "step": 5534 + }, + { + "epoch": 0.4991658024079001, + "grad_norm": 2.899934037872453, + "learning_rate": 2.1024840626713166e-06, + "loss": 0.6912, + "step": 5535 + }, + { + "epoch": 0.4992559859313703, + "grad_norm": 1.5670617461025658, + "learning_rate": 2.1019006151775177e-06, + "loss": 0.8582, + "step": 5536 + }, + { + "epoch": 0.4993461694548406, + "grad_norm": 1.8576427693390845, + "learning_rate": 2.101317158988997e-06, + "loss": 0.7491, + "step": 5537 + }, + { + "epoch": 0.49943635297831085, + "grad_norm": 3.185424877195253, + "learning_rate": 2.1007336941555374e-06, + "loss": 0.7318, + "step": 5538 + }, + { + "epoch": 0.49952653650178114, + "grad_norm": 1.5276137793787257, + "learning_rate": 2.1001502207269238e-06, + "loss": 0.723, + "step": 5539 + }, + { + "epoch": 0.49961672002525137, + "grad_norm": 2.127526744490642, + "learning_rate": 2.0995667387529407e-06, + "loss": 0.7147, + "step": 5540 + }, + { + "epoch": 0.49970690354872166, + "grad_norm": 1.6652742370802556, + "learning_rate": 2.098983248283375e-06, + "loss": 0.7115, + "step": 5541 + }, + { + "epoch": 0.4997970870721919, + "grad_norm": 1.7944015064095422, + "learning_rate": 2.098399749368012e-06, + "loss": 0.7608, + "step": 5542 + }, + { + "epoch": 0.4998872705956622, + "grad_norm": 0.6311311638271007, + "learning_rate": 2.09781624205664e-06, + "loss": 0.5704, + "step": 5543 + }, + { + "epoch": 0.4999774541191324, + "grad_norm": 1.5933513516773279, + "learning_rate": 2.0972327263990477e-06, + "loss": 0.8035, + "step": 5544 + }, + { + "epoch": 0.5000676376426026, + "grad_norm": 1.7740752968016673, + "learning_rate": 2.0966492024450226e-06, + "loss": 0.6921, + "step": 5545 + }, + { + "epoch": 0.500157821166073, + "grad_norm": 4.528819096391124, + "learning_rate": 2.0960656702443545e-06, + "loss": 0.9054, + "step": 5546 + }, + { + "epoch": 0.5002480046895432, + "grad_norm": 1.6345286134844996, + "learning_rate": 2.0954821298468343e-06, + "loss": 0.7512, + "step": 5547 + }, + { + "epoch": 0.5003381882130135, + "grad_norm": 1.5249268905124864, + "learning_rate": 2.0948985813022513e-06, + "loss": 0.8152, + "step": 5548 + }, + { + "epoch": 0.5004283717364837, + "grad_norm": 4.355364742662315, + "learning_rate": 2.094315024660399e-06, + "loss": 0.7961, + "step": 5549 + }, + { + "epoch": 0.500518555259954, + "grad_norm": 3.3162641910460713, + "learning_rate": 2.0937314599710676e-06, + "loss": 0.7824, + "step": 5550 + }, + { + "epoch": 0.5006087387834243, + "grad_norm": 3.7782733645151185, + "learning_rate": 2.0931478872840526e-06, + "loss": 0.7472, + "step": 5551 + }, + { + "epoch": 0.5006989223068945, + "grad_norm": 2.011174930288195, + "learning_rate": 2.092564306649145e-06, + "loss": 0.7393, + "step": 5552 + }, + { + "epoch": 0.5007891058303648, + "grad_norm": 1.7282555335966232, + "learning_rate": 2.091980718116141e-06, + "loss": 0.7922, + "step": 5553 + }, + { + "epoch": 0.5008792893538351, + "grad_norm": 1.718664271210707, + "learning_rate": 2.091397121734835e-06, + "loss": 0.7955, + "step": 5554 + }, + { + "epoch": 0.5009694728773053, + "grad_norm": 1.6864420082035705, + "learning_rate": 2.090813517555022e-06, + "loss": 0.7856, + "step": 5555 + }, + { + "epoch": 0.5010596564007755, + "grad_norm": 1.7743472819969959, + "learning_rate": 2.0902299056265e-06, + "loss": 0.7001, + "step": 5556 + }, + { + "epoch": 0.5011498399242459, + "grad_norm": 1.7673667486267843, + "learning_rate": 2.0896462859990643e-06, + "loss": 0.7422, + "step": 5557 + }, + { + "epoch": 0.5012400234477161, + "grad_norm": 1.5746207280129436, + "learning_rate": 2.089062658722513e-06, + "loss": 0.8185, + "step": 5558 + }, + { + "epoch": 0.5013302069711864, + "grad_norm": 1.7927821774949693, + "learning_rate": 2.0884790238466452e-06, + "loss": 0.8078, + "step": 5559 + }, + { + "epoch": 0.5014203904946566, + "grad_norm": 2.0058314424555252, + "learning_rate": 2.087895381421259e-06, + "loss": 0.7943, + "step": 5560 + }, + { + "epoch": 0.5015105740181269, + "grad_norm": 2.0204608482186854, + "learning_rate": 2.087311731496154e-06, + "loss": 0.7429, + "step": 5561 + }, + { + "epoch": 0.5016007575415972, + "grad_norm": 1.3860883542505071, + "learning_rate": 2.08672807412113e-06, + "loss": 0.7496, + "step": 5562 + }, + { + "epoch": 0.5016909410650674, + "grad_norm": 2.3583657840971877, + "learning_rate": 2.08614440934599e-06, + "loss": 0.8489, + "step": 5563 + }, + { + "epoch": 0.5017811245885376, + "grad_norm": 2.6729702851727066, + "learning_rate": 2.0855607372205337e-06, + "loss": 0.6681, + "step": 5564 + }, + { + "epoch": 0.501871308112008, + "grad_norm": 3.7234747868382585, + "learning_rate": 2.0849770577945623e-06, + "loss": 0.7758, + "step": 5565 + }, + { + "epoch": 0.5019614916354782, + "grad_norm": 1.2968065634445483, + "learning_rate": 2.084393371117881e-06, + "loss": 0.7887, + "step": 5566 + }, + { + "epoch": 0.5020516751589484, + "grad_norm": 1.6171959770828168, + "learning_rate": 2.0838096772402902e-06, + "loss": 0.6896, + "step": 5567 + }, + { + "epoch": 0.5021418586824187, + "grad_norm": 1.9195877358769955, + "learning_rate": 2.0832259762115973e-06, + "loss": 0.647, + "step": 5568 + }, + { + "epoch": 0.502232042205889, + "grad_norm": 1.3308708528561728, + "learning_rate": 2.082642268081605e-06, + "loss": 0.7845, + "step": 5569 + }, + { + "epoch": 0.5023222257293593, + "grad_norm": 1.9166826026364583, + "learning_rate": 2.082058552900118e-06, + "loss": 0.7115, + "step": 5570 + }, + { + "epoch": 0.5024124092528295, + "grad_norm": 1.6263453709685334, + "learning_rate": 2.081474830716944e-06, + "loss": 0.7318, + "step": 5571 + }, + { + "epoch": 0.5025025927762997, + "grad_norm": 2.117689798039126, + "learning_rate": 2.080891101581887e-06, + "loss": 0.7841, + "step": 5572 + }, + { + "epoch": 0.5025927762997701, + "grad_norm": 12.618052771220475, + "learning_rate": 2.080307365544755e-06, + "loss": 0.7975, + "step": 5573 + }, + { + "epoch": 0.5026829598232403, + "grad_norm": 1.740167995432766, + "learning_rate": 2.0797236226553567e-06, + "loss": 0.792, + "step": 5574 + }, + { + "epoch": 0.5027731433467105, + "grad_norm": 1.7909839423485874, + "learning_rate": 2.079139872963499e-06, + "loss": 0.6959, + "step": 5575 + }, + { + "epoch": 0.5028633268701809, + "grad_norm": 2.07489853116781, + "learning_rate": 2.078556116518991e-06, + "loss": 0.6636, + "step": 5576 + }, + { + "epoch": 0.5029535103936511, + "grad_norm": 1.8762868915466122, + "learning_rate": 2.077972353371642e-06, + "loss": 0.7311, + "step": 5577 + }, + { + "epoch": 0.5030436939171213, + "grad_norm": 1.465401911888578, + "learning_rate": 2.077388583571262e-06, + "loss": 0.7042, + "step": 5578 + }, + { + "epoch": 0.5031338774405916, + "grad_norm": 0.6566010172175192, + "learning_rate": 2.0768048071676608e-06, + "loss": 0.5143, + "step": 5579 + }, + { + "epoch": 0.5032240609640619, + "grad_norm": 1.7865208057865032, + "learning_rate": 2.0762210242106505e-06, + "loss": 0.7192, + "step": 5580 + }, + { + "epoch": 0.5033142444875321, + "grad_norm": 2.021811048861234, + "learning_rate": 2.0756372347500424e-06, + "loss": 0.7668, + "step": 5581 + }, + { + "epoch": 0.5034044280110024, + "grad_norm": 1.6604546491936505, + "learning_rate": 2.0750534388356473e-06, + "loss": 0.862, + "step": 5582 + }, + { + "epoch": 0.5034946115344726, + "grad_norm": 3.1217290926297556, + "learning_rate": 2.07446963651728e-06, + "loss": 0.7017, + "step": 5583 + }, + { + "epoch": 0.503584795057943, + "grad_norm": 1.7224725811330428, + "learning_rate": 2.0738858278447516e-06, + "loss": 0.6951, + "step": 5584 + }, + { + "epoch": 0.5036749785814132, + "grad_norm": 1.760276322347735, + "learning_rate": 2.073302012867878e-06, + "loss": 0.8458, + "step": 5585 + }, + { + "epoch": 0.5037651621048834, + "grad_norm": 1.4517735585488234, + "learning_rate": 2.0727181916364725e-06, + "loss": 0.7908, + "step": 5586 + }, + { + "epoch": 0.5038553456283537, + "grad_norm": 1.4269330114935883, + "learning_rate": 2.0721343642003493e-06, + "loss": 0.7946, + "step": 5587 + }, + { + "epoch": 0.503945529151824, + "grad_norm": 1.8073583878248984, + "learning_rate": 2.0715505306093247e-06, + "loss": 0.7866, + "step": 5588 + }, + { + "epoch": 0.5040357126752942, + "grad_norm": 2.075634029912101, + "learning_rate": 2.070966690913214e-06, + "loss": 0.785, + "step": 5589 + }, + { + "epoch": 0.5041258961987645, + "grad_norm": 2.3547513005169733, + "learning_rate": 2.0703828451618346e-06, + "loss": 0.6431, + "step": 5590 + }, + { + "epoch": 0.5042160797222347, + "grad_norm": 1.3753779065389968, + "learning_rate": 2.069798993405002e-06, + "loss": 0.7516, + "step": 5591 + }, + { + "epoch": 0.504306263245705, + "grad_norm": 1.6670021990948631, + "learning_rate": 2.0692151356925345e-06, + "loss": 0.7964, + "step": 5592 + }, + { + "epoch": 0.5043964467691753, + "grad_norm": 1.7034846634774858, + "learning_rate": 2.068631272074251e-06, + "loss": 0.783, + "step": 5593 + }, + { + "epoch": 0.5044866302926455, + "grad_norm": 1.8928685393594435, + "learning_rate": 2.0680474025999676e-06, + "loss": 0.8835, + "step": 5594 + }, + { + "epoch": 0.5045768138161157, + "grad_norm": 0.8743635315105472, + "learning_rate": 2.0674635273195055e-06, + "loss": 0.632, + "step": 5595 + }, + { + "epoch": 0.5046669973395861, + "grad_norm": 2.3617005665817645, + "learning_rate": 2.066879646282682e-06, + "loss": 0.7551, + "step": 5596 + }, + { + "epoch": 0.5047571808630563, + "grad_norm": 1.7380110284172257, + "learning_rate": 2.0662957595393194e-06, + "loss": 0.7325, + "step": 5597 + }, + { + "epoch": 0.5048473643865266, + "grad_norm": 1.6815745554493466, + "learning_rate": 2.0657118671392373e-06, + "loss": 0.786, + "step": 5598 + }, + { + "epoch": 0.5049375479099969, + "grad_norm": 1.875204583768808, + "learning_rate": 2.0651279691322558e-06, + "loss": 0.8151, + "step": 5599 + }, + { + "epoch": 0.5050277314334671, + "grad_norm": 1.610077047459184, + "learning_rate": 2.0645440655681973e-06, + "loss": 0.7108, + "step": 5600 + }, + { + "epoch": 0.5051179149569374, + "grad_norm": 2.217501651802007, + "learning_rate": 2.0639601564968826e-06, + "loss": 0.8233, + "step": 5601 + }, + { + "epoch": 0.5052080984804076, + "grad_norm": 0.6949159584415388, + "learning_rate": 2.0633762419681355e-06, + "loss": 0.6664, + "step": 5602 + }, + { + "epoch": 0.5052982820038779, + "grad_norm": 1.4034698130612986, + "learning_rate": 2.062792322031777e-06, + "loss": 0.7734, + "step": 5603 + }, + { + "epoch": 0.5053884655273482, + "grad_norm": 1.7955980490478671, + "learning_rate": 2.062208396737632e-06, + "loss": 0.7214, + "step": 5604 + }, + { + "epoch": 0.5054786490508184, + "grad_norm": 1.4365802808104742, + "learning_rate": 2.0616244661355235e-06, + "loss": 0.7784, + "step": 5605 + }, + { + "epoch": 0.5055688325742886, + "grad_norm": 2.2439217589366818, + "learning_rate": 2.0610405302752752e-06, + "loss": 0.7943, + "step": 5606 + }, + { + "epoch": 0.505659016097759, + "grad_norm": 1.8538694536405085, + "learning_rate": 2.060456589206713e-06, + "loss": 0.7528, + "step": 5607 + }, + { + "epoch": 0.5057491996212292, + "grad_norm": 2.4700131625237893, + "learning_rate": 2.0598726429796614e-06, + "loss": 0.7041, + "step": 5608 + }, + { + "epoch": 0.5058393831446995, + "grad_norm": 1.5069300782221564, + "learning_rate": 2.059288691643945e-06, + "loss": 0.7147, + "step": 5609 + }, + { + "epoch": 0.5059295666681697, + "grad_norm": 1.701913805409575, + "learning_rate": 2.0587047352493913e-06, + "loss": 0.8427, + "step": 5610 + }, + { + "epoch": 0.50601975019164, + "grad_norm": 2.443489540834876, + "learning_rate": 2.0581207738458248e-06, + "loss": 0.7364, + "step": 5611 + }, + { + "epoch": 0.5061099337151103, + "grad_norm": 2.755493399301873, + "learning_rate": 2.0575368074830743e-06, + "loss": 0.7174, + "step": 5612 + }, + { + "epoch": 0.5062001172385805, + "grad_norm": 1.6814151838760252, + "learning_rate": 2.0569528362109667e-06, + "loss": 0.6728, + "step": 5613 + }, + { + "epoch": 0.5062903007620507, + "grad_norm": 1.5377009996977422, + "learning_rate": 2.056368860079327e-06, + "loss": 0.7553, + "step": 5614 + }, + { + "epoch": 0.5063804842855211, + "grad_norm": 5.229047290103025, + "learning_rate": 2.0557848791379874e-06, + "loss": 0.6953, + "step": 5615 + }, + { + "epoch": 0.5064706678089913, + "grad_norm": 1.6626973171304946, + "learning_rate": 2.0552008934367734e-06, + "loss": 0.6638, + "step": 5616 + }, + { + "epoch": 0.5065608513324615, + "grad_norm": 1.9940376217773823, + "learning_rate": 2.0546169030255154e-06, + "loss": 0.8347, + "step": 5617 + }, + { + "epoch": 0.5066510348559318, + "grad_norm": 1.7458283566181685, + "learning_rate": 2.054032907954041e-06, + "loss": 0.8344, + "step": 5618 + }, + { + "epoch": 0.5067412183794021, + "grad_norm": 1.47969981332427, + "learning_rate": 2.053448908272182e-06, + "loss": 0.724, + "step": 5619 + }, + { + "epoch": 0.5068314019028723, + "grad_norm": 1.6567661992471645, + "learning_rate": 2.0528649040297673e-06, + "loss": 0.797, + "step": 5620 + }, + { + "epoch": 0.5069215854263426, + "grad_norm": 1.7953475430806027, + "learning_rate": 2.0522808952766266e-06, + "loss": 0.7601, + "step": 5621 + }, + { + "epoch": 0.5070117689498129, + "grad_norm": 1.5261562247995801, + "learning_rate": 2.0516968820625925e-06, + "loss": 0.8038, + "step": 5622 + }, + { + "epoch": 0.5071019524732832, + "grad_norm": 2.8399977189713037, + "learning_rate": 2.051112864437495e-06, + "loss": 0.7482, + "step": 5623 + }, + { + "epoch": 0.5071921359967534, + "grad_norm": 3.1059276697563805, + "learning_rate": 2.050528842451166e-06, + "loss": 0.7261, + "step": 5624 + }, + { + "epoch": 0.5072823195202236, + "grad_norm": 1.5504571302109809, + "learning_rate": 2.049944816153438e-06, + "loss": 0.6416, + "step": 5625 + }, + { + "epoch": 0.507372503043694, + "grad_norm": 1.4969625714196475, + "learning_rate": 2.049360785594142e-06, + "loss": 0.8022, + "step": 5626 + }, + { + "epoch": 0.5074626865671642, + "grad_norm": 0.6153682986833532, + "learning_rate": 2.048776750823113e-06, + "loss": 0.5106, + "step": 5627 + }, + { + "epoch": 0.5075528700906344, + "grad_norm": 0.6711138472473221, + "learning_rate": 2.0481927118901817e-06, + "loss": 0.564, + "step": 5628 + }, + { + "epoch": 0.5076430536141047, + "grad_norm": 1.799556158234075, + "learning_rate": 2.0476086688451824e-06, + "loss": 0.7037, + "step": 5629 + }, + { + "epoch": 0.507733237137575, + "grad_norm": 1.7821053576080272, + "learning_rate": 2.04702462173795e-06, + "loss": 0.7459, + "step": 5630 + }, + { + "epoch": 0.5078234206610452, + "grad_norm": 2.101839068624891, + "learning_rate": 2.0464405706183167e-06, + "loss": 0.7379, + "step": 5631 + }, + { + "epoch": 0.5079136041845155, + "grad_norm": 1.916735779165819, + "learning_rate": 2.045856515536118e-06, + "loss": 0.7735, + "step": 5632 + }, + { + "epoch": 0.5080037877079857, + "grad_norm": 1.6097940436420917, + "learning_rate": 2.045272456541188e-06, + "loss": 0.7746, + "step": 5633 + }, + { + "epoch": 0.508093971231456, + "grad_norm": 1.588536280221614, + "learning_rate": 2.0446883936833635e-06, + "loss": 0.8261, + "step": 5634 + }, + { + "epoch": 0.5081841547549263, + "grad_norm": 2.2283258248628024, + "learning_rate": 2.0441043270124782e-06, + "loss": 0.7334, + "step": 5635 + }, + { + "epoch": 0.5082743382783965, + "grad_norm": 1.9993479380992663, + "learning_rate": 2.0435202565783683e-06, + "loss": 0.786, + "step": 5636 + }, + { + "epoch": 0.5083645218018668, + "grad_norm": 1.5314630632669535, + "learning_rate": 2.042936182430871e-06, + "loss": 0.72, + "step": 5637 + }, + { + "epoch": 0.5084547053253371, + "grad_norm": 1.572506183678821, + "learning_rate": 2.0423521046198206e-06, + "loss": 0.7624, + "step": 5638 + }, + { + "epoch": 0.5085448888488073, + "grad_norm": 1.5236665993601946, + "learning_rate": 2.041768023195056e-06, + "loss": 0.6751, + "step": 5639 + }, + { + "epoch": 0.5086350723722776, + "grad_norm": 1.667741317788182, + "learning_rate": 2.0411839382064126e-06, + "loss": 0.828, + "step": 5640 + }, + { + "epoch": 0.5087252558957478, + "grad_norm": 1.7200951600919434, + "learning_rate": 2.040599849703729e-06, + "loss": 0.6533, + "step": 5641 + }, + { + "epoch": 0.5088154394192181, + "grad_norm": 2.0225350860457003, + "learning_rate": 2.040015757736843e-06, + "loss": 0.698, + "step": 5642 + }, + { + "epoch": 0.5089056229426884, + "grad_norm": 1.5391229143242424, + "learning_rate": 2.039431662355591e-06, + "loss": 0.6835, + "step": 5643 + }, + { + "epoch": 0.5089958064661586, + "grad_norm": 1.680360313332617, + "learning_rate": 2.0388475636098126e-06, + "loss": 0.7743, + "step": 5644 + }, + { + "epoch": 0.5090859899896288, + "grad_norm": 1.474186271952582, + "learning_rate": 2.038263461549346e-06, + "loss": 0.7906, + "step": 5645 + }, + { + "epoch": 0.5091761735130992, + "grad_norm": 1.8439544305776971, + "learning_rate": 2.0376793562240297e-06, + "loss": 0.7606, + "step": 5646 + }, + { + "epoch": 0.5092663570365694, + "grad_norm": 1.5884207841082554, + "learning_rate": 2.037095247683703e-06, + "loss": 0.7753, + "step": 5647 + }, + { + "epoch": 0.5093565405600397, + "grad_norm": 1.5549754733883603, + "learning_rate": 2.0365111359782046e-06, + "loss": 0.7136, + "step": 5648 + }, + { + "epoch": 0.50944672408351, + "grad_norm": 1.7424881337882396, + "learning_rate": 2.0359270211573757e-06, + "loss": 0.7856, + "step": 5649 + }, + { + "epoch": 0.5095369076069802, + "grad_norm": 2.0724005767571128, + "learning_rate": 2.0353429032710545e-06, + "loss": 0.8205, + "step": 5650 + }, + { + "epoch": 0.5096270911304505, + "grad_norm": 2.8592510926128734, + "learning_rate": 2.0347587823690825e-06, + "loss": 0.7309, + "step": 5651 + }, + { + "epoch": 0.5097172746539207, + "grad_norm": 1.6494762889988326, + "learning_rate": 2.034174658501299e-06, + "loss": 0.7565, + "step": 5652 + }, + { + "epoch": 0.509807458177391, + "grad_norm": 2.3336424509285125, + "learning_rate": 2.0335905317175453e-06, + "loss": 0.7938, + "step": 5653 + }, + { + "epoch": 0.5098976417008613, + "grad_norm": 2.3933209943674663, + "learning_rate": 2.033006402067663e-06, + "loss": 0.7631, + "step": 5654 + }, + { + "epoch": 0.5099878252243315, + "grad_norm": 1.8396660137784757, + "learning_rate": 2.0324222696014912e-06, + "loss": 0.6597, + "step": 5655 + }, + { + "epoch": 0.5100780087478017, + "grad_norm": 1.7485974004341112, + "learning_rate": 2.0318381343688733e-06, + "loss": 0.7997, + "step": 5656 + }, + { + "epoch": 0.5101681922712721, + "grad_norm": 1.6641375027799143, + "learning_rate": 2.0312539964196505e-06, + "loss": 0.7399, + "step": 5657 + }, + { + "epoch": 0.5102583757947423, + "grad_norm": 1.57320513406377, + "learning_rate": 2.030669855803664e-06, + "loss": 0.7367, + "step": 5658 + }, + { + "epoch": 0.5103485593182125, + "grad_norm": 1.5185432261926803, + "learning_rate": 2.0300857125707563e-06, + "loss": 0.8079, + "step": 5659 + }, + { + "epoch": 0.5104387428416828, + "grad_norm": 1.7775122929401486, + "learning_rate": 2.0295015667707697e-06, + "loss": 0.7838, + "step": 5660 + }, + { + "epoch": 0.5105289263651531, + "grad_norm": 1.5049212312838085, + "learning_rate": 2.0289174184535472e-06, + "loss": 0.7627, + "step": 5661 + }, + { + "epoch": 0.5106191098886234, + "grad_norm": 1.7663165063343875, + "learning_rate": 2.02833326766893e-06, + "loss": 0.7307, + "step": 5662 + }, + { + "epoch": 0.5107092934120936, + "grad_norm": 1.7900615717778021, + "learning_rate": 2.027749114466763e-06, + "loss": 0.8203, + "step": 5663 + }, + { + "epoch": 0.5107994769355638, + "grad_norm": 2.1341378701603464, + "learning_rate": 2.027164958896889e-06, + "loss": 0.8009, + "step": 5664 + }, + { + "epoch": 0.5108896604590342, + "grad_norm": 1.8204301787882158, + "learning_rate": 2.02658080100915e-06, + "loss": 0.8777, + "step": 5665 + }, + { + "epoch": 0.5109798439825044, + "grad_norm": 1.8217069990062265, + "learning_rate": 2.0259966408533915e-06, + "loss": 0.8322, + "step": 5666 + }, + { + "epoch": 0.5110700275059746, + "grad_norm": 1.653120431010345, + "learning_rate": 2.025412478479455e-06, + "loss": 0.8432, + "step": 5667 + }, + { + "epoch": 0.5111602110294449, + "grad_norm": 2.2115535888664835, + "learning_rate": 2.0248283139371862e-06, + "loss": 0.8064, + "step": 5668 + }, + { + "epoch": 0.5112503945529152, + "grad_norm": 5.339922510220643, + "learning_rate": 2.024244147276429e-06, + "loss": 0.7307, + "step": 5669 + }, + { + "epoch": 0.5113405780763854, + "grad_norm": 1.4118606960455224, + "learning_rate": 2.023659978547027e-06, + "loss": 0.8279, + "step": 5670 + }, + { + "epoch": 0.5114307615998557, + "grad_norm": 1.7204771207370753, + "learning_rate": 2.023075807798826e-06, + "loss": 0.76, + "step": 5671 + }, + { + "epoch": 0.511520945123326, + "grad_norm": 1.4928049936918366, + "learning_rate": 2.0224916350816696e-06, + "loss": 0.6804, + "step": 5672 + }, + { + "epoch": 0.5116111286467963, + "grad_norm": 1.8157361835309649, + "learning_rate": 2.0219074604454026e-06, + "loss": 0.7792, + "step": 5673 + }, + { + "epoch": 0.5117013121702665, + "grad_norm": 1.9379593089411276, + "learning_rate": 2.02132328393987e-06, + "loss": 0.7547, + "step": 5674 + }, + { + "epoch": 0.5117914956937367, + "grad_norm": 1.7748114015559906, + "learning_rate": 2.0207391056149174e-06, + "loss": 0.7183, + "step": 5675 + }, + { + "epoch": 0.5118816792172071, + "grad_norm": 2.1390430703127117, + "learning_rate": 2.020154925520391e-06, + "loss": 0.7325, + "step": 5676 + }, + { + "epoch": 0.5119718627406773, + "grad_norm": 1.5094676050562037, + "learning_rate": 2.0195707437061332e-06, + "loss": 0.856, + "step": 5677 + }, + { + "epoch": 0.5120620462641475, + "grad_norm": 14.551334406432026, + "learning_rate": 2.0189865602219934e-06, + "loss": 0.7603, + "step": 5678 + }, + { + "epoch": 0.5121522297876178, + "grad_norm": 1.865602466951967, + "learning_rate": 2.0184023751178154e-06, + "loss": 0.8342, + "step": 5679 + }, + { + "epoch": 0.5122424133110881, + "grad_norm": 1.5607643405610496, + "learning_rate": 2.017818188443444e-06, + "loss": 0.7129, + "step": 5680 + }, + { + "epoch": 0.5123325968345583, + "grad_norm": 3.257383762021831, + "learning_rate": 2.017234000248728e-06, + "loss": 0.7727, + "step": 5681 + }, + { + "epoch": 0.5124227803580286, + "grad_norm": 1.5331874190576922, + "learning_rate": 2.0166498105835108e-06, + "loss": 0.696, + "step": 5682 + }, + { + "epoch": 0.5125129638814988, + "grad_norm": 1.3884687530001751, + "learning_rate": 2.0160656194976407e-06, + "loss": 0.6937, + "step": 5683 + }, + { + "epoch": 0.5126031474049692, + "grad_norm": 1.6284020098543333, + "learning_rate": 2.0154814270409634e-06, + "loss": 0.7147, + "step": 5684 + }, + { + "epoch": 0.5126933309284394, + "grad_norm": 1.7324196042429805, + "learning_rate": 2.0148972332633247e-06, + "loss": 0.7272, + "step": 5685 + }, + { + "epoch": 0.5127835144519096, + "grad_norm": 2.1257345388622526, + "learning_rate": 2.0143130382145733e-06, + "loss": 0.8247, + "step": 5686 + }, + { + "epoch": 0.5128736979753798, + "grad_norm": 2.1740300306686815, + "learning_rate": 2.0137288419445533e-06, + "loss": 0.7329, + "step": 5687 + }, + { + "epoch": 0.5129638814988502, + "grad_norm": 2.0541222088421716, + "learning_rate": 2.0131446445031134e-06, + "loss": 0.8046, + "step": 5688 + }, + { + "epoch": 0.5130540650223204, + "grad_norm": 1.8252984390801341, + "learning_rate": 2.0125604459400994e-06, + "loss": 0.8407, + "step": 5689 + }, + { + "epoch": 0.5131442485457907, + "grad_norm": 1.726289131997183, + "learning_rate": 2.0119762463053596e-06, + "loss": 0.8198, + "step": 5690 + }, + { + "epoch": 0.5132344320692609, + "grad_norm": 1.7436928852530986, + "learning_rate": 2.0113920456487406e-06, + "loss": 0.7521, + "step": 5691 + }, + { + "epoch": 0.5133246155927312, + "grad_norm": 1.9214502501011812, + "learning_rate": 2.010807844020088e-06, + "loss": 0.7549, + "step": 5692 + }, + { + "epoch": 0.5134147991162015, + "grad_norm": 1.5173693971191862, + "learning_rate": 2.0102236414692524e-06, + "loss": 0.8119, + "step": 5693 + }, + { + "epoch": 0.5135049826396717, + "grad_norm": 1.4907792721295328, + "learning_rate": 2.0096394380460777e-06, + "loss": 0.7471, + "step": 5694 + }, + { + "epoch": 0.513595166163142, + "grad_norm": 1.6243624536925823, + "learning_rate": 2.0090552338004136e-06, + "loss": 0.7919, + "step": 5695 + }, + { + "epoch": 0.5136853496866123, + "grad_norm": 8.211516355368977, + "learning_rate": 2.0084710287821077e-06, + "loss": 0.7594, + "step": 5696 + }, + { + "epoch": 0.5137755332100825, + "grad_norm": 1.6538339663041137, + "learning_rate": 2.007886823041006e-06, + "loss": 0.7987, + "step": 5697 + }, + { + "epoch": 0.5138657167335527, + "grad_norm": 1.5896765318713948, + "learning_rate": 2.0073026166269577e-06, + "loss": 0.7749, + "step": 5698 + }, + { + "epoch": 0.5139559002570231, + "grad_norm": 1.351088159487167, + "learning_rate": 2.0067184095898093e-06, + "loss": 0.8192, + "step": 5699 + }, + { + "epoch": 0.5140460837804933, + "grad_norm": 1.6834153083002799, + "learning_rate": 2.0061342019794094e-06, + "loss": 0.7697, + "step": 5700 + }, + { + "epoch": 0.5141362673039636, + "grad_norm": 1.460513946006923, + "learning_rate": 2.0055499938456058e-06, + "loss": 0.833, + "step": 5701 + }, + { + "epoch": 0.5142264508274338, + "grad_norm": 2.392454357449276, + "learning_rate": 2.0049657852382464e-06, + "loss": 0.7308, + "step": 5702 + }, + { + "epoch": 0.5143166343509041, + "grad_norm": 1.8760042044545704, + "learning_rate": 2.0043815762071782e-06, + "loss": 0.6936, + "step": 5703 + }, + { + "epoch": 0.5144068178743744, + "grad_norm": 1.737496462482172, + "learning_rate": 2.0037973668022492e-06, + "loss": 0.7284, + "step": 5704 + }, + { + "epoch": 0.5144970013978446, + "grad_norm": 1.7937674810641002, + "learning_rate": 2.003213157073309e-06, + "loss": 0.6436, + "step": 5705 + }, + { + "epoch": 0.5145871849213148, + "grad_norm": 1.5755856923606402, + "learning_rate": 2.002628947070204e-06, + "loss": 0.7593, + "step": 5706 + }, + { + "epoch": 0.5146773684447852, + "grad_norm": 1.6831286998414894, + "learning_rate": 2.002044736842783e-06, + "loss": 0.8257, + "step": 5707 + }, + { + "epoch": 0.5147675519682554, + "grad_norm": 1.5507983151513038, + "learning_rate": 2.001460526440894e-06, + "loss": 0.7397, + "step": 5708 + }, + { + "epoch": 0.5148577354917256, + "grad_norm": 0.6152722786020652, + "learning_rate": 2.0008763159143843e-06, + "loss": 0.5807, + "step": 5709 + }, + { + "epoch": 0.5149479190151959, + "grad_norm": 1.7405826720975859, + "learning_rate": 2.000292105313103e-06, + "loss": 0.7906, + "step": 5710 + }, + { + "epoch": 0.5150381025386662, + "grad_norm": 1.7066260234394104, + "learning_rate": 1.999707894686897e-06, + "loss": 0.7987, + "step": 5711 + }, + { + "epoch": 0.5151282860621365, + "grad_norm": 1.4305701527678942, + "learning_rate": 1.9991236840856155e-06, + "loss": 0.7484, + "step": 5712 + }, + { + "epoch": 0.5152184695856067, + "grad_norm": 2.096792032388382, + "learning_rate": 1.9985394735591065e-06, + "loss": 0.7334, + "step": 5713 + }, + { + "epoch": 0.5153086531090769, + "grad_norm": 5.567655081950773, + "learning_rate": 1.997955263157217e-06, + "loss": 0.8084, + "step": 5714 + }, + { + "epoch": 0.5153988366325473, + "grad_norm": 1.7084342027363377, + "learning_rate": 1.997371052929796e-06, + "loss": 0.795, + "step": 5715 + }, + { + "epoch": 0.5154890201560175, + "grad_norm": 1.693340508202999, + "learning_rate": 1.996786842926691e-06, + "loss": 0.716, + "step": 5716 + }, + { + "epoch": 0.5155792036794877, + "grad_norm": 1.619265341010666, + "learning_rate": 1.9962026331977506e-06, + "loss": 0.7707, + "step": 5717 + }, + { + "epoch": 0.5156693872029581, + "grad_norm": 2.0053012446556817, + "learning_rate": 1.9956184237928224e-06, + "loss": 0.7763, + "step": 5718 + }, + { + "epoch": 0.5157595707264283, + "grad_norm": 3.7913842707401115, + "learning_rate": 1.995034214761754e-06, + "loss": 0.752, + "step": 5719 + }, + { + "epoch": 0.5158497542498985, + "grad_norm": 1.2802327933068613, + "learning_rate": 1.9944500061543945e-06, + "loss": 0.735, + "step": 5720 + }, + { + "epoch": 0.5159399377733688, + "grad_norm": 1.6972711034056436, + "learning_rate": 1.99386579802059e-06, + "loss": 0.7798, + "step": 5721 + }, + { + "epoch": 0.5160301212968391, + "grad_norm": 1.864240994984549, + "learning_rate": 1.993281590410191e-06, + "loss": 0.7381, + "step": 5722 + }, + { + "epoch": 0.5161203048203094, + "grad_norm": 1.5359387138726686, + "learning_rate": 1.992697383373043e-06, + "loss": 0.7089, + "step": 5723 + }, + { + "epoch": 0.5162104883437796, + "grad_norm": 0.7803907246579572, + "learning_rate": 1.9921131769589937e-06, + "loss": 0.6222, + "step": 5724 + }, + { + "epoch": 0.5163006718672498, + "grad_norm": 1.9579245837399875, + "learning_rate": 1.991528971217893e-06, + "loss": 0.7657, + "step": 5725 + }, + { + "epoch": 0.5163908553907202, + "grad_norm": 1.8381647762909485, + "learning_rate": 1.9909447661995858e-06, + "loss": 0.7575, + "step": 5726 + }, + { + "epoch": 0.5164810389141904, + "grad_norm": 1.8221299426560071, + "learning_rate": 1.990360561953922e-06, + "loss": 0.7789, + "step": 5727 + }, + { + "epoch": 0.5165712224376606, + "grad_norm": 1.370356565351992, + "learning_rate": 1.9897763585307483e-06, + "loss": 0.7823, + "step": 5728 + }, + { + "epoch": 0.5166614059611309, + "grad_norm": 1.5344163504504054, + "learning_rate": 1.989192155979912e-06, + "loss": 0.8001, + "step": 5729 + }, + { + "epoch": 0.5167515894846012, + "grad_norm": 1.6375249729727406, + "learning_rate": 1.98860795435126e-06, + "loss": 0.7927, + "step": 5730 + }, + { + "epoch": 0.5168417730080714, + "grad_norm": 1.617819105270781, + "learning_rate": 1.9880237536946406e-06, + "loss": 0.7674, + "step": 5731 + }, + { + "epoch": 0.5169319565315417, + "grad_norm": 1.830591542988925, + "learning_rate": 1.987439554059901e-06, + "loss": 0.7313, + "step": 5732 + }, + { + "epoch": 0.5170221400550119, + "grad_norm": 1.6075209521030185, + "learning_rate": 1.9868553554968864e-06, + "loss": 0.7327, + "step": 5733 + }, + { + "epoch": 0.5171123235784822, + "grad_norm": 2.4932172316108887, + "learning_rate": 1.986271158055447e-06, + "loss": 0.712, + "step": 5734 + }, + { + "epoch": 0.5172025071019525, + "grad_norm": 1.599263918955766, + "learning_rate": 1.9856869617854273e-06, + "loss": 0.7514, + "step": 5735 + }, + { + "epoch": 0.5172926906254227, + "grad_norm": 1.5885206711012196, + "learning_rate": 1.9851027667366746e-06, + "loss": 0.738, + "step": 5736 + }, + { + "epoch": 0.517382874148893, + "grad_norm": 1.511190159372547, + "learning_rate": 1.984518572959037e-06, + "loss": 0.8594, + "step": 5737 + }, + { + "epoch": 0.5174730576723633, + "grad_norm": 1.6920181823727487, + "learning_rate": 1.9839343805023587e-06, + "loss": 0.68, + "step": 5738 + }, + { + "epoch": 0.5175632411958335, + "grad_norm": 2.5644226304832, + "learning_rate": 1.9833501894164886e-06, + "loss": 0.8532, + "step": 5739 + }, + { + "epoch": 0.5176534247193038, + "grad_norm": 1.5579037776365219, + "learning_rate": 1.982765999751273e-06, + "loss": 0.8001, + "step": 5740 + }, + { + "epoch": 0.5177436082427741, + "grad_norm": 1.8026283042839868, + "learning_rate": 1.9821818115565553e-06, + "loss": 0.7948, + "step": 5741 + }, + { + "epoch": 0.5178337917662443, + "grad_norm": 2.4923379786871007, + "learning_rate": 1.9815976248821853e-06, + "loss": 0.666, + "step": 5742 + }, + { + "epoch": 0.5179239752897146, + "grad_norm": 6.664568793505634, + "learning_rate": 1.981013439778007e-06, + "loss": 0.7736, + "step": 5743 + }, + { + "epoch": 0.5180141588131848, + "grad_norm": 1.6828222077921156, + "learning_rate": 1.9804292562938666e-06, + "loss": 0.7731, + "step": 5744 + }, + { + "epoch": 0.5181043423366551, + "grad_norm": 1.6497211249527468, + "learning_rate": 1.97984507447961e-06, + "loss": 0.7955, + "step": 5745 + }, + { + "epoch": 0.5181945258601254, + "grad_norm": 1.6948271025751538, + "learning_rate": 1.9792608943850824e-06, + "loss": 0.7009, + "step": 5746 + }, + { + "epoch": 0.5182847093835956, + "grad_norm": 1.626598452299895, + "learning_rate": 1.9786767160601305e-06, + "loss": 0.741, + "step": 5747 + }, + { + "epoch": 0.5183748929070658, + "grad_norm": 2.0043115633521302, + "learning_rate": 1.9780925395545977e-06, + "loss": 0.6939, + "step": 5748 + }, + { + "epoch": 0.5184650764305362, + "grad_norm": 0.6360342406751408, + "learning_rate": 1.9775083649183306e-06, + "loss": 0.5669, + "step": 5749 + }, + { + "epoch": 0.5185552599540064, + "grad_norm": 1.9338117996726942, + "learning_rate": 1.976924192201174e-06, + "loss": 0.8018, + "step": 5750 + }, + { + "epoch": 0.5186454434774767, + "grad_norm": 1.5212085557117745, + "learning_rate": 1.9763400214529723e-06, + "loss": 0.6905, + "step": 5751 + }, + { + "epoch": 0.5187356270009469, + "grad_norm": 1.5559154625853027, + "learning_rate": 1.9757558527235713e-06, + "loss": 0.6348, + "step": 5752 + }, + { + "epoch": 0.5188258105244172, + "grad_norm": 1.5695410964072134, + "learning_rate": 1.9751716860628136e-06, + "loss": 0.7499, + "step": 5753 + }, + { + "epoch": 0.5189159940478875, + "grad_norm": 2.3191287942038117, + "learning_rate": 1.974587521520545e-06, + "loss": 0.7879, + "step": 5754 + }, + { + "epoch": 0.5190061775713577, + "grad_norm": 1.704646489588009, + "learning_rate": 1.9740033591466088e-06, + "loss": 0.7954, + "step": 5755 + }, + { + "epoch": 0.5190963610948279, + "grad_norm": 1.715535295986685, + "learning_rate": 1.97341919899085e-06, + "loss": 0.8776, + "step": 5756 + }, + { + "epoch": 0.5191865446182983, + "grad_norm": 1.617934565825393, + "learning_rate": 1.9728350411031114e-06, + "loss": 0.7847, + "step": 5757 + }, + { + "epoch": 0.5192767281417685, + "grad_norm": 1.7136224729583471, + "learning_rate": 1.9722508855332367e-06, + "loss": 0.8372, + "step": 5758 + }, + { + "epoch": 0.5193669116652387, + "grad_norm": 1.40091060417748, + "learning_rate": 1.97166673233107e-06, + "loss": 0.6913, + "step": 5759 + }, + { + "epoch": 0.519457095188709, + "grad_norm": 5.734005450988244, + "learning_rate": 1.971082581546453e-06, + "loss": 0.8027, + "step": 5760 + }, + { + "epoch": 0.5195472787121793, + "grad_norm": 1.7076310908652443, + "learning_rate": 1.9704984332292306e-06, + "loss": 0.7861, + "step": 5761 + }, + { + "epoch": 0.5196374622356495, + "grad_norm": 1.6792871752234986, + "learning_rate": 1.9699142874292444e-06, + "loss": 0.8514, + "step": 5762 + }, + { + "epoch": 0.5197276457591198, + "grad_norm": 1.5258353118772543, + "learning_rate": 1.969330144196336e-06, + "loss": 0.8542, + "step": 5763 + }, + { + "epoch": 0.51981782928259, + "grad_norm": 2.126608408166814, + "learning_rate": 1.9687460035803497e-06, + "loss": 0.7083, + "step": 5764 + }, + { + "epoch": 0.5199080128060604, + "grad_norm": 1.8100954087544516, + "learning_rate": 1.9681618656311265e-06, + "loss": 0.744, + "step": 5765 + }, + { + "epoch": 0.5199981963295306, + "grad_norm": 1.5300922298682116, + "learning_rate": 1.9675777303985086e-06, + "loss": 0.6743, + "step": 5766 + }, + { + "epoch": 0.5200883798530008, + "grad_norm": 2.8866957894574865, + "learning_rate": 1.9669935979323376e-06, + "loss": 0.7928, + "step": 5767 + }, + { + "epoch": 0.5201785633764712, + "grad_norm": 1.6061759931681407, + "learning_rate": 1.9664094682824545e-06, + "loss": 0.8197, + "step": 5768 + }, + { + "epoch": 0.5202687468999414, + "grad_norm": 9.229587611350029, + "learning_rate": 1.965825341498701e-06, + "loss": 0.7213, + "step": 5769 + }, + { + "epoch": 0.5203589304234116, + "grad_norm": 2.7300756168722384, + "learning_rate": 1.9652412176309177e-06, + "loss": 0.8373, + "step": 5770 + }, + { + "epoch": 0.5204491139468819, + "grad_norm": 1.5593188900685717, + "learning_rate": 1.9646570967289453e-06, + "loss": 0.631, + "step": 5771 + }, + { + "epoch": 0.5205392974703522, + "grad_norm": 1.4999996821085275, + "learning_rate": 1.9640729788426246e-06, + "loss": 0.7977, + "step": 5772 + }, + { + "epoch": 0.5206294809938224, + "grad_norm": 1.4494613074316034, + "learning_rate": 1.963488864021795e-06, + "loss": 0.7935, + "step": 5773 + }, + { + "epoch": 0.5207196645172927, + "grad_norm": 2.0553214252006113, + "learning_rate": 1.962904752316298e-06, + "loss": 0.6552, + "step": 5774 + }, + { + "epoch": 0.5208098480407629, + "grad_norm": 8.089018038649057, + "learning_rate": 1.9623206437759706e-06, + "loss": 0.6927, + "step": 5775 + }, + { + "epoch": 0.5209000315642333, + "grad_norm": 1.443397862149675, + "learning_rate": 1.9617365384506545e-06, + "loss": 0.7205, + "step": 5776 + }, + { + "epoch": 0.5209902150877035, + "grad_norm": 1.3604879536822672, + "learning_rate": 1.9611524363901872e-06, + "loss": 0.7489, + "step": 5777 + }, + { + "epoch": 0.5210803986111737, + "grad_norm": 2.7995059463048113, + "learning_rate": 1.960568337644409e-06, + "loss": 0.7505, + "step": 5778 + }, + { + "epoch": 0.521170582134644, + "grad_norm": 1.428803551071641, + "learning_rate": 1.9599842422631576e-06, + "loss": 0.7463, + "step": 5779 + }, + { + "epoch": 0.5212607656581143, + "grad_norm": 1.6059934774910054, + "learning_rate": 1.9594001502962703e-06, + "loss": 0.7565, + "step": 5780 + }, + { + "epoch": 0.5213509491815845, + "grad_norm": 1.3242154191687, + "learning_rate": 1.9588160617935868e-06, + "loss": 0.8207, + "step": 5781 + }, + { + "epoch": 0.5214411327050548, + "grad_norm": 1.5767023453630724, + "learning_rate": 1.958231976804944e-06, + "loss": 0.8477, + "step": 5782 + }, + { + "epoch": 0.521531316228525, + "grad_norm": 2.162242688873559, + "learning_rate": 1.957647895380179e-06, + "loss": 0.6913, + "step": 5783 + }, + { + "epoch": 0.5216214997519953, + "grad_norm": 1.3129410683799545, + "learning_rate": 1.9570638175691297e-06, + "loss": 0.6879, + "step": 5784 + }, + { + "epoch": 0.5217116832754656, + "grad_norm": 2.707823534203977, + "learning_rate": 1.956479743421632e-06, + "loss": 0.8019, + "step": 5785 + }, + { + "epoch": 0.5218018667989358, + "grad_norm": 1.5484566357274172, + "learning_rate": 1.955895672987522e-06, + "loss": 0.7013, + "step": 5786 + }, + { + "epoch": 0.521892050322406, + "grad_norm": 1.6485383784807974, + "learning_rate": 1.9553116063166367e-06, + "loss": 0.6675, + "step": 5787 + }, + { + "epoch": 0.5219822338458764, + "grad_norm": 2.5235762897506486, + "learning_rate": 1.954727543458812e-06, + "loss": 0.6662, + "step": 5788 + }, + { + "epoch": 0.5220724173693466, + "grad_norm": 2.956706786401376, + "learning_rate": 1.954143484463883e-06, + "loss": 0.7902, + "step": 5789 + }, + { + "epoch": 0.5221626008928169, + "grad_norm": 1.628119848320072, + "learning_rate": 1.9535594293816836e-06, + "loss": 0.7448, + "step": 5790 + }, + { + "epoch": 0.5222527844162872, + "grad_norm": 1.9840150041012854, + "learning_rate": 1.952975378262051e-06, + "loss": 0.7103, + "step": 5791 + }, + { + "epoch": 0.5223429679397574, + "grad_norm": 2.0744082836619793, + "learning_rate": 1.952391331154817e-06, + "loss": 0.7161, + "step": 5792 + }, + { + "epoch": 0.5224331514632277, + "grad_norm": 1.5689377721745157, + "learning_rate": 1.9518072881098185e-06, + "loss": 0.8131, + "step": 5793 + }, + { + "epoch": 0.5225233349866979, + "grad_norm": 1.507735966929117, + "learning_rate": 1.9512232491768867e-06, + "loss": 0.7495, + "step": 5794 + }, + { + "epoch": 0.5226135185101682, + "grad_norm": 1.4567389719205057, + "learning_rate": 1.9506392144058573e-06, + "loss": 0.7469, + "step": 5795 + }, + { + "epoch": 0.5227037020336385, + "grad_norm": 2.0611741399367602, + "learning_rate": 1.9500551838465623e-06, + "loss": 0.7546, + "step": 5796 + }, + { + "epoch": 0.5227938855571087, + "grad_norm": 1.4691628220901487, + "learning_rate": 1.9494711575488337e-06, + "loss": 0.7712, + "step": 5797 + }, + { + "epoch": 0.5228840690805789, + "grad_norm": 1.8126424536136005, + "learning_rate": 1.948887135562505e-06, + "loss": 0.7031, + "step": 5798 + }, + { + "epoch": 0.5229742526040493, + "grad_norm": 1.4611362587263863, + "learning_rate": 1.9483031179374074e-06, + "loss": 0.8054, + "step": 5799 + }, + { + "epoch": 0.5230644361275195, + "grad_norm": 1.9466184346002524, + "learning_rate": 1.9477191047233736e-06, + "loss": 0.8165, + "step": 5800 + }, + { + "epoch": 0.5231546196509897, + "grad_norm": 1.8637043534540934, + "learning_rate": 1.9471350959702334e-06, + "loss": 0.6406, + "step": 5801 + }, + { + "epoch": 0.52324480317446, + "grad_norm": 3.4068866099844026, + "learning_rate": 1.9465510917278184e-06, + "loss": 0.7631, + "step": 5802 + }, + { + "epoch": 0.5233349866979303, + "grad_norm": 1.791363498262456, + "learning_rate": 1.9459670920459593e-06, + "loss": 0.7524, + "step": 5803 + }, + { + "epoch": 0.5234251702214006, + "grad_norm": 2.0380077922848026, + "learning_rate": 1.945383096974485e-06, + "loss": 0.6696, + "step": 5804 + }, + { + "epoch": 0.5235153537448708, + "grad_norm": 2.4563978194390605, + "learning_rate": 1.944799106563227e-06, + "loss": 0.6829, + "step": 5805 + }, + { + "epoch": 0.523605537268341, + "grad_norm": 3.5333278248102205, + "learning_rate": 1.9442151208620133e-06, + "loss": 0.7941, + "step": 5806 + }, + { + "epoch": 0.5236957207918114, + "grad_norm": 1.8733566712162242, + "learning_rate": 1.943631139920672e-06, + "loss": 0.8017, + "step": 5807 + }, + { + "epoch": 0.5237859043152816, + "grad_norm": 9.38660699124585, + "learning_rate": 1.943047163789034e-06, + "loss": 0.7108, + "step": 5808 + }, + { + "epoch": 0.5238760878387518, + "grad_norm": 1.5128158018962252, + "learning_rate": 1.942463192516925e-06, + "loss": 0.7165, + "step": 5809 + }, + { + "epoch": 0.5239662713622221, + "grad_norm": 1.7049575885408834, + "learning_rate": 1.9418792261541746e-06, + "loss": 0.7787, + "step": 5810 + }, + { + "epoch": 0.5240564548856924, + "grad_norm": 1.6546907102501467, + "learning_rate": 1.9412952647506094e-06, + "loss": 0.7178, + "step": 5811 + }, + { + "epoch": 0.5241466384091626, + "grad_norm": 1.568766805855145, + "learning_rate": 1.9407113083560552e-06, + "loss": 0.7301, + "step": 5812 + }, + { + "epoch": 0.5242368219326329, + "grad_norm": 1.723998431366563, + "learning_rate": 1.940127357020339e-06, + "loss": 0.7373, + "step": 5813 + }, + { + "epoch": 0.5243270054561032, + "grad_norm": 2.9664958226020035, + "learning_rate": 1.939543410793287e-06, + "loss": 0.7668, + "step": 5814 + }, + { + "epoch": 0.5244171889795735, + "grad_norm": 2.1055679439469883, + "learning_rate": 1.9389594697247246e-06, + "loss": 0.7918, + "step": 5815 + }, + { + "epoch": 0.5245073725030437, + "grad_norm": 2.9201171809682465, + "learning_rate": 1.9383755338644763e-06, + "loss": 0.8293, + "step": 5816 + }, + { + "epoch": 0.5245975560265139, + "grad_norm": 1.970775349619589, + "learning_rate": 1.937791603262368e-06, + "loss": 0.7908, + "step": 5817 + }, + { + "epoch": 0.5246877395499843, + "grad_norm": 2.672178463127051, + "learning_rate": 1.9372076779682235e-06, + "loss": 0.7079, + "step": 5818 + }, + { + "epoch": 0.5247779230734545, + "grad_norm": 2.110576372295259, + "learning_rate": 1.9366237580318648e-06, + "loss": 0.7083, + "step": 5819 + }, + { + "epoch": 0.5248681065969247, + "grad_norm": 1.3470007145178131, + "learning_rate": 1.9360398435031176e-06, + "loss": 0.7601, + "step": 5820 + }, + { + "epoch": 0.524958290120395, + "grad_norm": 3.4853804296664594, + "learning_rate": 1.9354559344318025e-06, + "loss": 0.7708, + "step": 5821 + }, + { + "epoch": 0.5250484736438653, + "grad_norm": 1.9653581131809705, + "learning_rate": 1.934872030867744e-06, + "loss": 0.8097, + "step": 5822 + }, + { + "epoch": 0.5251386571673355, + "grad_norm": 1.8186303089517168, + "learning_rate": 1.934288132860763e-06, + "loss": 0.7693, + "step": 5823 + }, + { + "epoch": 0.5252288406908058, + "grad_norm": 1.527837492999186, + "learning_rate": 1.93370424046068e-06, + "loss": 0.7568, + "step": 5824 + }, + { + "epoch": 0.525319024214276, + "grad_norm": 4.369689988721646, + "learning_rate": 1.9331203537173177e-06, + "loss": 0.819, + "step": 5825 + }, + { + "epoch": 0.5254092077377464, + "grad_norm": 1.3366689631489141, + "learning_rate": 1.9325364726804947e-06, + "loss": 0.8703, + "step": 5826 + }, + { + "epoch": 0.5254993912612166, + "grad_norm": 1.8815325426606861, + "learning_rate": 1.9319525974000327e-06, + "loss": 0.6847, + "step": 5827 + }, + { + "epoch": 0.5255895747846868, + "grad_norm": 12.782328590453929, + "learning_rate": 1.93136872792575e-06, + "loss": 0.8141, + "step": 5828 + }, + { + "epoch": 0.525679758308157, + "grad_norm": 1.5174791458431471, + "learning_rate": 1.9307848643074653e-06, + "loss": 0.7107, + "step": 5829 + }, + { + "epoch": 0.5257699418316274, + "grad_norm": 1.7289992868081117, + "learning_rate": 1.9302010065949985e-06, + "loss": 0.7931, + "step": 5830 + }, + { + "epoch": 0.5258601253550976, + "grad_norm": 3.5088844615475367, + "learning_rate": 1.9296171548381657e-06, + "loss": 0.7167, + "step": 5831 + }, + { + "epoch": 0.5259503088785679, + "grad_norm": 1.581665072036652, + "learning_rate": 1.9290333090867862e-06, + "loss": 0.7449, + "step": 5832 + }, + { + "epoch": 0.5260404924020381, + "grad_norm": 2.1622872352671076, + "learning_rate": 1.928449469390676e-06, + "loss": 0.7606, + "step": 5833 + }, + { + "epoch": 0.5261306759255084, + "grad_norm": 1.5939060209610871, + "learning_rate": 1.927865635799651e-06, + "loss": 0.7272, + "step": 5834 + }, + { + "epoch": 0.5262208594489787, + "grad_norm": 1.5013911630803631, + "learning_rate": 1.927281808363528e-06, + "loss": 0.6871, + "step": 5835 + }, + { + "epoch": 0.5263110429724489, + "grad_norm": 1.5797493761208992, + "learning_rate": 1.9266979871321216e-06, + "loss": 0.8295, + "step": 5836 + }, + { + "epoch": 0.5264012264959192, + "grad_norm": 1.5951001301490186, + "learning_rate": 1.9261141721552482e-06, + "loss": 0.7066, + "step": 5837 + }, + { + "epoch": 0.5264914100193895, + "grad_norm": 1.682591327948958, + "learning_rate": 1.9255303634827204e-06, + "loss": 0.8234, + "step": 5838 + }, + { + "epoch": 0.5265815935428597, + "grad_norm": 1.670247221367861, + "learning_rate": 1.924946561164352e-06, + "loss": 0.8184, + "step": 5839 + }, + { + "epoch": 0.52667177706633, + "grad_norm": 1.7417111280080644, + "learning_rate": 1.9243627652499582e-06, + "loss": 0.7091, + "step": 5840 + }, + { + "epoch": 0.5267619605898003, + "grad_norm": 2.043612493619661, + "learning_rate": 1.9237789757893493e-06, + "loss": 0.8176, + "step": 5841 + }, + { + "epoch": 0.5268521441132705, + "grad_norm": 35.246043314618134, + "learning_rate": 1.9231951928323395e-06, + "loss": 0.7997, + "step": 5842 + }, + { + "epoch": 0.5269423276367408, + "grad_norm": 1.673727061404359, + "learning_rate": 1.922611416428738e-06, + "loss": 0.7113, + "step": 5843 + }, + { + "epoch": 0.527032511160211, + "grad_norm": 1.5082252214300553, + "learning_rate": 1.922027646628358e-06, + "loss": 0.7794, + "step": 5844 + }, + { + "epoch": 0.5271226946836813, + "grad_norm": 0.6695421438321302, + "learning_rate": 1.9214438834810092e-06, + "loss": 0.6355, + "step": 5845 + }, + { + "epoch": 0.5272128782071516, + "grad_norm": 3.361433525139699, + "learning_rate": 1.9208601270365008e-06, + "loss": 0.6846, + "step": 5846 + }, + { + "epoch": 0.5273030617306218, + "grad_norm": 1.5493177142830517, + "learning_rate": 1.9202763773446435e-06, + "loss": 0.8344, + "step": 5847 + }, + { + "epoch": 0.527393245254092, + "grad_norm": 1.5963436041311914, + "learning_rate": 1.9196926344552444e-06, + "loss": 0.6519, + "step": 5848 + }, + { + "epoch": 0.5274834287775624, + "grad_norm": 1.7580908660884005, + "learning_rate": 1.919108898418113e-06, + "loss": 0.783, + "step": 5849 + }, + { + "epoch": 0.5275736123010326, + "grad_norm": 1.390243906672816, + "learning_rate": 1.918525169283057e-06, + "loss": 0.6549, + "step": 5850 + }, + { + "epoch": 0.5276637958245028, + "grad_norm": 1.8130888804401881, + "learning_rate": 1.9179414470998817e-06, + "loss": 0.7836, + "step": 5851 + }, + { + "epoch": 0.5277539793479731, + "grad_norm": 1.5768360879748546, + "learning_rate": 1.917357731918395e-06, + "loss": 0.7171, + "step": 5852 + }, + { + "epoch": 0.5278441628714434, + "grad_norm": 1.5751897122401222, + "learning_rate": 1.9167740237884025e-06, + "loss": 0.8878, + "step": 5853 + }, + { + "epoch": 0.5279343463949137, + "grad_norm": 1.58128275177702, + "learning_rate": 1.916190322759709e-06, + "loss": 0.7933, + "step": 5854 + }, + { + "epoch": 0.5280245299183839, + "grad_norm": 1.6153038960155932, + "learning_rate": 1.91560662888212e-06, + "loss": 0.86, + "step": 5855 + }, + { + "epoch": 0.5281147134418541, + "grad_norm": 2.2988102862784925, + "learning_rate": 1.915022942205438e-06, + "loss": 0.8027, + "step": 5856 + }, + { + "epoch": 0.5282048969653245, + "grad_norm": 2.212531280296458, + "learning_rate": 1.914439262779468e-06, + "loss": 0.7501, + "step": 5857 + }, + { + "epoch": 0.5282950804887947, + "grad_norm": 1.9278956445334217, + "learning_rate": 1.9138555906540103e-06, + "loss": 0.6915, + "step": 5858 + }, + { + "epoch": 0.5283852640122649, + "grad_norm": 2.7196419666769374, + "learning_rate": 1.91327192587887e-06, + "loss": 0.6654, + "step": 5859 + }, + { + "epoch": 0.5284754475357353, + "grad_norm": 21.927741942553475, + "learning_rate": 1.912688268503846e-06, + "loss": 0.7248, + "step": 5860 + }, + { + "epoch": 0.5285656310592055, + "grad_norm": 1.788655130574981, + "learning_rate": 1.912104618578741e-06, + "loss": 0.7239, + "step": 5861 + }, + { + "epoch": 0.5286558145826757, + "grad_norm": 1.748596650517369, + "learning_rate": 1.9115209761533554e-06, + "loss": 0.7314, + "step": 5862 + }, + { + "epoch": 0.528745998106146, + "grad_norm": 1.7859405197158076, + "learning_rate": 1.9109373412774863e-06, + "loss": 0.6695, + "step": 5863 + }, + { + "epoch": 0.5288361816296163, + "grad_norm": 1.4292833768087851, + "learning_rate": 1.910353714000936e-06, + "loss": 0.7564, + "step": 5864 + }, + { + "epoch": 0.5289263651530866, + "grad_norm": 1.5927185760162041, + "learning_rate": 1.9097700943734997e-06, + "loss": 0.7168, + "step": 5865 + }, + { + "epoch": 0.5290165486765568, + "grad_norm": 1.525467835253842, + "learning_rate": 1.909186482444977e-06, + "loss": 0.655, + "step": 5866 + }, + { + "epoch": 0.529106732200027, + "grad_norm": 2.2186572633087045, + "learning_rate": 1.9086028782651652e-06, + "loss": 0.6555, + "step": 5867 + }, + { + "epoch": 0.5291969157234974, + "grad_norm": 1.475442209066669, + "learning_rate": 1.908019281883859e-06, + "loss": 0.747, + "step": 5868 + }, + { + "epoch": 0.5292870992469676, + "grad_norm": 1.6403594029109547, + "learning_rate": 1.9074356933508545e-06, + "loss": 0.7465, + "step": 5869 + }, + { + "epoch": 0.5293772827704378, + "grad_norm": 2.842090331980875, + "learning_rate": 1.9068521127159477e-06, + "loss": 0.7904, + "step": 5870 + }, + { + "epoch": 0.5294674662939081, + "grad_norm": 1.9684201524318008, + "learning_rate": 1.9062685400289322e-06, + "loss": 0.7054, + "step": 5871 + }, + { + "epoch": 0.5295576498173784, + "grad_norm": 1.660435336720774, + "learning_rate": 1.9056849753396018e-06, + "loss": 0.7287, + "step": 5872 + }, + { + "epoch": 0.5296478333408486, + "grad_norm": 2.140010327822283, + "learning_rate": 1.9051014186977485e-06, + "loss": 0.7903, + "step": 5873 + }, + { + "epoch": 0.5297380168643189, + "grad_norm": 1.5433569311201307, + "learning_rate": 1.9045178701531664e-06, + "loss": 0.7569, + "step": 5874 + }, + { + "epoch": 0.5298282003877891, + "grad_norm": 1.3837560247274603, + "learning_rate": 1.903934329755645e-06, + "loss": 0.7481, + "step": 5875 + }, + { + "epoch": 0.5299183839112594, + "grad_norm": 2.1015598906887263, + "learning_rate": 1.9033507975549775e-06, + "loss": 0.7899, + "step": 5876 + }, + { + "epoch": 0.5300085674347297, + "grad_norm": 1.790181505495065, + "learning_rate": 1.9027672736009525e-06, + "loss": 0.746, + "step": 5877 + }, + { + "epoch": 0.5300987509581999, + "grad_norm": 1.4657864804973102, + "learning_rate": 1.9021837579433593e-06, + "loss": 0.7021, + "step": 5878 + }, + { + "epoch": 0.5301889344816701, + "grad_norm": 2.539963219143547, + "learning_rate": 1.901600250631988e-06, + "loss": 0.7285, + "step": 5879 + }, + { + "epoch": 0.5302791180051405, + "grad_norm": 3.0806533576277744, + "learning_rate": 1.901016751716625e-06, + "loss": 0.7442, + "step": 5880 + }, + { + "epoch": 0.5303693015286107, + "grad_norm": 7.103232343729431, + "learning_rate": 1.9004332612470593e-06, + "loss": 0.7177, + "step": 5881 + }, + { + "epoch": 0.530459485052081, + "grad_norm": 1.446844495282846, + "learning_rate": 1.8998497792730763e-06, + "loss": 0.7811, + "step": 5882 + }, + { + "epoch": 0.5305496685755512, + "grad_norm": 1.5265531322929775, + "learning_rate": 1.8992663058444629e-06, + "loss": 0.8027, + "step": 5883 + }, + { + "epoch": 0.5306398520990215, + "grad_norm": 1.611814039541204, + "learning_rate": 1.8986828410110032e-06, + "loss": 0.7887, + "step": 5884 + }, + { + "epoch": 0.5307300356224918, + "grad_norm": 2.0390574721935737, + "learning_rate": 1.8980993848224823e-06, + "loss": 0.7515, + "step": 5885 + }, + { + "epoch": 0.530820219145962, + "grad_norm": 1.9546629077026076, + "learning_rate": 1.8975159373286843e-06, + "loss": 0.7124, + "step": 5886 + }, + { + "epoch": 0.5309104026694323, + "grad_norm": 1.7468628693570594, + "learning_rate": 1.8969324985793904e-06, + "loss": 0.7914, + "step": 5887 + }, + { + "epoch": 0.5310005861929026, + "grad_norm": 3.20708269456683, + "learning_rate": 1.8963490686243851e-06, + "loss": 0.7443, + "step": 5888 + }, + { + "epoch": 0.5310907697163728, + "grad_norm": 2.1243140291519764, + "learning_rate": 1.8957656475134486e-06, + "loss": 0.6898, + "step": 5889 + }, + { + "epoch": 0.531180953239843, + "grad_norm": 2.0078323303905807, + "learning_rate": 1.895182235296361e-06, + "loss": 0.7341, + "step": 5890 + }, + { + "epoch": 0.5312711367633134, + "grad_norm": 3.4891230644986297, + "learning_rate": 1.8945988320229042e-06, + "loss": 0.6786, + "step": 5891 + }, + { + "epoch": 0.5313613202867836, + "grad_norm": 1.8542777324658386, + "learning_rate": 1.8940154377428553e-06, + "loss": 0.8346, + "step": 5892 + }, + { + "epoch": 0.5314515038102539, + "grad_norm": 1.5505312470673798, + "learning_rate": 1.8934320525059944e-06, + "loss": 0.7803, + "step": 5893 + }, + { + "epoch": 0.5315416873337241, + "grad_norm": 2.0409221719035684, + "learning_rate": 1.8928486763620984e-06, + "loss": 0.6442, + "step": 5894 + }, + { + "epoch": 0.5316318708571944, + "grad_norm": 1.6512589737838177, + "learning_rate": 1.892265309360943e-06, + "loss": 0.767, + "step": 5895 + }, + { + "epoch": 0.5317220543806647, + "grad_norm": 1.7134106324096734, + "learning_rate": 1.8916819515523067e-06, + "loss": 0.7389, + "step": 5896 + }, + { + "epoch": 0.5318122379041349, + "grad_norm": 8.03638955806065, + "learning_rate": 1.891098602985963e-06, + "loss": 0.6696, + "step": 5897 + }, + { + "epoch": 0.5319024214276051, + "grad_norm": 1.9443209162332549, + "learning_rate": 1.8905152637116872e-06, + "loss": 0.7761, + "step": 5898 + }, + { + "epoch": 0.5319926049510755, + "grad_norm": 1.9355181740981933, + "learning_rate": 1.8899319337792527e-06, + "loss": 0.7906, + "step": 5899 + }, + { + "epoch": 0.5320827884745457, + "grad_norm": 1.3015040316964426, + "learning_rate": 1.8893486132384325e-06, + "loss": 0.7909, + "step": 5900 + }, + { + "epoch": 0.5321729719980159, + "grad_norm": 1.9135160327814846, + "learning_rate": 1.888765302138999e-06, + "loss": 0.7235, + "step": 5901 + }, + { + "epoch": 0.5322631555214862, + "grad_norm": 2.1612939856373763, + "learning_rate": 1.8881820005307224e-06, + "loss": 0.6986, + "step": 5902 + }, + { + "epoch": 0.5323533390449565, + "grad_norm": 1.6046635077149176, + "learning_rate": 1.8875987084633748e-06, + "loss": 0.6863, + "step": 5903 + }, + { + "epoch": 0.5324435225684268, + "grad_norm": 3.137028243020736, + "learning_rate": 1.8870154259867246e-06, + "loss": 0.7887, + "step": 5904 + }, + { + "epoch": 0.532533706091897, + "grad_norm": 1.7355776904055875, + "learning_rate": 1.886432153150542e-06, + "loss": 0.693, + "step": 5905 + }, + { + "epoch": 0.5326238896153672, + "grad_norm": 1.580087932843865, + "learning_rate": 1.8858488900045944e-06, + "loss": 0.7313, + "step": 5906 + }, + { + "epoch": 0.5327140731388376, + "grad_norm": 2.41562342871927, + "learning_rate": 1.885265636598648e-06, + "loss": 0.7033, + "step": 5907 + }, + { + "epoch": 0.5328042566623078, + "grad_norm": 1.5036875379976125, + "learning_rate": 1.884682392982471e-06, + "loss": 0.7839, + "step": 5908 + }, + { + "epoch": 0.532894440185778, + "grad_norm": 2.40809016028517, + "learning_rate": 1.8840991592058274e-06, + "loss": 0.805, + "step": 5909 + }, + { + "epoch": 0.5329846237092484, + "grad_norm": 1.572976808050828, + "learning_rate": 1.8835159353184828e-06, + "loss": 0.7691, + "step": 5910 + }, + { + "epoch": 0.5330748072327186, + "grad_norm": 1.6769228872313893, + "learning_rate": 1.8829327213702013e-06, + "loss": 0.8209, + "step": 5911 + }, + { + "epoch": 0.5331649907561888, + "grad_norm": 1.694608832377649, + "learning_rate": 1.8823495174107452e-06, + "loss": 0.826, + "step": 5912 + }, + { + "epoch": 0.5332551742796591, + "grad_norm": 4.3309497635613585, + "learning_rate": 1.8817663234898773e-06, + "loss": 0.7284, + "step": 5913 + }, + { + "epoch": 0.5333453578031294, + "grad_norm": 1.9641141655559446, + "learning_rate": 1.881183139657358e-06, + "loss": 0.7631, + "step": 5914 + }, + { + "epoch": 0.5334355413265996, + "grad_norm": 1.7193701578929725, + "learning_rate": 1.8805999659629488e-06, + "loss": 0.769, + "step": 5915 + }, + { + "epoch": 0.5335257248500699, + "grad_norm": 0.6198778787058753, + "learning_rate": 1.880016802456409e-06, + "loss": 0.5363, + "step": 5916 + }, + { + "epoch": 0.5336159083735401, + "grad_norm": 1.92277623576824, + "learning_rate": 1.8794336491874964e-06, + "loss": 0.695, + "step": 5917 + }, + { + "epoch": 0.5337060918970105, + "grad_norm": 2.0703220583137254, + "learning_rate": 1.8788505062059708e-06, + "loss": 0.7121, + "step": 5918 + }, + { + "epoch": 0.5337962754204807, + "grad_norm": 1.6029142304413444, + "learning_rate": 1.8782673735615869e-06, + "loss": 0.7387, + "step": 5919 + }, + { + "epoch": 0.5338864589439509, + "grad_norm": 1.6892530025078936, + "learning_rate": 1.8776842513041026e-06, + "loss": 0.7088, + "step": 5920 + }, + { + "epoch": 0.5339766424674212, + "grad_norm": 2.1255019100075634, + "learning_rate": 1.8771011394832727e-06, + "loss": 0.7205, + "step": 5921 + }, + { + "epoch": 0.5340668259908915, + "grad_norm": 9.847535002104626, + "learning_rate": 1.8765180381488501e-06, + "loss": 0.7333, + "step": 5922 + }, + { + "epoch": 0.5341570095143617, + "grad_norm": 0.7320563052519546, + "learning_rate": 1.8759349473505905e-06, + "loss": 0.6678, + "step": 5923 + }, + { + "epoch": 0.534247193037832, + "grad_norm": 1.8588513189842477, + "learning_rate": 1.8753518671382447e-06, + "loss": 0.676, + "step": 5924 + }, + { + "epoch": 0.5343373765613022, + "grad_norm": 1.7058275810757202, + "learning_rate": 1.8747687975615649e-06, + "loss": 0.7318, + "step": 5925 + }, + { + "epoch": 0.5344275600847725, + "grad_norm": 1.6091206914470226, + "learning_rate": 1.874185738670302e-06, + "loss": 0.6433, + "step": 5926 + }, + { + "epoch": 0.5345177436082428, + "grad_norm": 2.1330576326119717, + "learning_rate": 1.8736026905142057e-06, + "loss": 0.739, + "step": 5927 + }, + { + "epoch": 0.534607927131713, + "grad_norm": 1.5577758325606055, + "learning_rate": 1.873019653143025e-06, + "loss": 0.7284, + "step": 5928 + }, + { + "epoch": 0.5346981106551832, + "grad_norm": 1.4655555724920915, + "learning_rate": 1.8724366266065069e-06, + "loss": 0.7613, + "step": 5929 + }, + { + "epoch": 0.5347882941786536, + "grad_norm": 1.7653110191216101, + "learning_rate": 1.8718536109543998e-06, + "loss": 0.826, + "step": 5930 + }, + { + "epoch": 0.5348784777021238, + "grad_norm": 2.080595455401084, + "learning_rate": 1.8712706062364485e-06, + "loss": 0.7672, + "step": 5931 + }, + { + "epoch": 0.534968661225594, + "grad_norm": 1.7304750877247093, + "learning_rate": 1.8706876125024e-06, + "loss": 0.6389, + "step": 5932 + }, + { + "epoch": 0.5350588447490644, + "grad_norm": 1.7900861451928238, + "learning_rate": 1.870104629801997e-06, + "loss": 0.7946, + "step": 5933 + }, + { + "epoch": 0.5351490282725346, + "grad_norm": 1.7791354695547246, + "learning_rate": 1.8695216581849823e-06, + "loss": 0.7621, + "step": 5934 + }, + { + "epoch": 0.5352392117960049, + "grad_norm": 1.6967179761421651, + "learning_rate": 1.8689386977011003e-06, + "loss": 0.7206, + "step": 5935 + }, + { + "epoch": 0.5353293953194751, + "grad_norm": 0.6778095434536786, + "learning_rate": 1.8683557484000903e-06, + "loss": 0.576, + "step": 5936 + }, + { + "epoch": 0.5354195788429454, + "grad_norm": 1.692647711817616, + "learning_rate": 1.8677728103316947e-06, + "loss": 0.7811, + "step": 5937 + }, + { + "epoch": 0.5355097623664157, + "grad_norm": 1.4298001437811567, + "learning_rate": 1.8671898835456518e-06, + "loss": 0.7072, + "step": 5938 + }, + { + "epoch": 0.5355999458898859, + "grad_norm": 2.7771539422054508, + "learning_rate": 1.8666069680917003e-06, + "loss": 0.8547, + "step": 5939 + }, + { + "epoch": 0.5356901294133561, + "grad_norm": 2.8036510709604636, + "learning_rate": 1.8660240640195775e-06, + "loss": 0.7123, + "step": 5940 + }, + { + "epoch": 0.5357803129368265, + "grad_norm": 1.4961108811794563, + "learning_rate": 1.8654411713790203e-06, + "loss": 0.7841, + "step": 5941 + }, + { + "epoch": 0.5358704964602967, + "grad_norm": 1.8703644352508537, + "learning_rate": 1.8648582902197648e-06, + "loss": 0.773, + "step": 5942 + }, + { + "epoch": 0.535960679983767, + "grad_norm": 1.786239862753611, + "learning_rate": 1.8642754205915452e-06, + "loss": 0.7581, + "step": 5943 + }, + { + "epoch": 0.5360508635072372, + "grad_norm": 1.5001644203353715, + "learning_rate": 1.8636925625440943e-06, + "loss": 0.7007, + "step": 5944 + }, + { + "epoch": 0.5361410470307075, + "grad_norm": 2.0795126276130422, + "learning_rate": 1.863109716127146e-06, + "loss": 0.6577, + "step": 5945 + }, + { + "epoch": 0.5362312305541778, + "grad_norm": 1.4673300026944351, + "learning_rate": 1.8625268813904311e-06, + "loss": 0.795, + "step": 5946 + }, + { + "epoch": 0.536321414077648, + "grad_norm": 1.6361324744277137, + "learning_rate": 1.8619440583836814e-06, + "loss": 0.7753, + "step": 5947 + }, + { + "epoch": 0.5364115976011182, + "grad_norm": 1.6833881155233785, + "learning_rate": 1.8613612471566249e-06, + "loss": 0.7602, + "step": 5948 + }, + { + "epoch": 0.5365017811245886, + "grad_norm": 1.8667932836414634, + "learning_rate": 1.8607784477589922e-06, + "loss": 0.698, + "step": 5949 + }, + { + "epoch": 0.5365919646480588, + "grad_norm": 1.5624761961077425, + "learning_rate": 1.8601956602405103e-06, + "loss": 0.7664, + "step": 5950 + }, + { + "epoch": 0.536682148171529, + "grad_norm": 1.5820185154531834, + "learning_rate": 1.8596128846509043e-06, + "loss": 0.794, + "step": 5951 + }, + { + "epoch": 0.5367723316949993, + "grad_norm": 1.849135663479041, + "learning_rate": 1.859030121039902e-06, + "loss": 0.7085, + "step": 5952 + }, + { + "epoch": 0.5368625152184696, + "grad_norm": 1.6039793805202593, + "learning_rate": 1.8584473694572268e-06, + "loss": 0.7055, + "step": 5953 + }, + { + "epoch": 0.5369526987419398, + "grad_norm": 1.7359488280001396, + "learning_rate": 1.8578646299526026e-06, + "loss": 0.7701, + "step": 5954 + }, + { + "epoch": 0.5370428822654101, + "grad_norm": 1.7011768503039688, + "learning_rate": 1.8572819025757518e-06, + "loss": 0.7211, + "step": 5955 + }, + { + "epoch": 0.5371330657888804, + "grad_norm": 1.8232463720158194, + "learning_rate": 1.8566991873763959e-06, + "loss": 0.7664, + "step": 5956 + }, + { + "epoch": 0.5372232493123507, + "grad_norm": 1.5991438303488597, + "learning_rate": 1.856116484404256e-06, + "loss": 0.6975, + "step": 5957 + }, + { + "epoch": 0.5373134328358209, + "grad_norm": 1.808812632817523, + "learning_rate": 1.8555337937090502e-06, + "loss": 0.7228, + "step": 5958 + }, + { + "epoch": 0.5374036163592911, + "grad_norm": 1.8207051913682022, + "learning_rate": 1.8549511153404984e-06, + "loss": 0.7358, + "step": 5959 + }, + { + "epoch": 0.5374937998827615, + "grad_norm": 2.5686097229758484, + "learning_rate": 1.854368449348317e-06, + "loss": 0.8599, + "step": 5960 + }, + { + "epoch": 0.5375839834062317, + "grad_norm": 2.9496643134037437, + "learning_rate": 1.853785795782222e-06, + "loss": 0.7122, + "step": 5961 + }, + { + "epoch": 0.5376741669297019, + "grad_norm": 2.0845965815271894, + "learning_rate": 1.85320315469193e-06, + "loss": 0.6215, + "step": 5962 + }, + { + "epoch": 0.5377643504531722, + "grad_norm": 0.6523635998292804, + "learning_rate": 1.8526205261271534e-06, + "loss": 0.5821, + "step": 5963 + }, + { + "epoch": 0.5378545339766425, + "grad_norm": 3.1785109684124673, + "learning_rate": 1.852037910137607e-06, + "loss": 0.6678, + "step": 5964 + }, + { + "epoch": 0.5379447175001127, + "grad_norm": 11.666779290518523, + "learning_rate": 1.851455306773002e-06, + "loss": 0.8357, + "step": 5965 + }, + { + "epoch": 0.538034901023583, + "grad_norm": 1.3610271192227008, + "learning_rate": 1.8508727160830483e-06, + "loss": 0.7254, + "step": 5966 + }, + { + "epoch": 0.5381250845470532, + "grad_norm": 2.18985131460102, + "learning_rate": 1.8502901381174575e-06, + "loss": 0.8199, + "step": 5967 + }, + { + "epoch": 0.5382152680705236, + "grad_norm": 2.0092857091829197, + "learning_rate": 1.8497075729259372e-06, + "loss": 0.6585, + "step": 5968 + }, + { + "epoch": 0.5383054515939938, + "grad_norm": 1.5977419804559967, + "learning_rate": 1.8491250205581963e-06, + "loss": 0.6424, + "step": 5969 + }, + { + "epoch": 0.538395635117464, + "grad_norm": 1.4128448976726924, + "learning_rate": 1.8485424810639393e-06, + "loss": 0.7559, + "step": 5970 + }, + { + "epoch": 0.5384858186409343, + "grad_norm": 2.0164516439431024, + "learning_rate": 1.847959954492874e-06, + "loss": 0.8107, + "step": 5971 + }, + { + "epoch": 0.5385760021644046, + "grad_norm": 2.393399492303937, + "learning_rate": 1.8473774408947035e-06, + "loss": 0.8425, + "step": 5972 + }, + { + "epoch": 0.5386661856878748, + "grad_norm": 0.6574421000642023, + "learning_rate": 1.8467949403191308e-06, + "loss": 0.5532, + "step": 5973 + }, + { + "epoch": 0.5387563692113451, + "grad_norm": 2.1889013978543685, + "learning_rate": 1.8462124528158592e-06, + "loss": 0.8077, + "step": 5974 + }, + { + "epoch": 0.5388465527348153, + "grad_norm": 1.6935117096452468, + "learning_rate": 1.8456299784345881e-06, + "loss": 0.7364, + "step": 5975 + }, + { + "epoch": 0.5389367362582856, + "grad_norm": 1.7937767186318208, + "learning_rate": 1.8450475172250194e-06, + "loss": 0.7744, + "step": 5976 + }, + { + "epoch": 0.5390269197817559, + "grad_norm": 1.7455920836462346, + "learning_rate": 1.844465069236851e-06, + "loss": 0.7351, + "step": 5977 + }, + { + "epoch": 0.5391171033052261, + "grad_norm": 1.539507264291413, + "learning_rate": 1.8438826345197796e-06, + "loss": 0.6593, + "step": 5978 + }, + { + "epoch": 0.5392072868286965, + "grad_norm": 4.554216782601558, + "learning_rate": 1.8433002131235036e-06, + "loss": 0.7452, + "step": 5979 + }, + { + "epoch": 0.5392974703521667, + "grad_norm": 1.826546076822319, + "learning_rate": 1.8427178050977167e-06, + "loss": 0.773, + "step": 5980 + }, + { + "epoch": 0.5393876538756369, + "grad_norm": 1.4470104244782012, + "learning_rate": 1.8421354104921143e-06, + "loss": 0.8296, + "step": 5981 + }, + { + "epoch": 0.5394778373991072, + "grad_norm": 0.6047861496843124, + "learning_rate": 1.8415530293563894e-06, + "loss": 0.549, + "step": 5982 + }, + { + "epoch": 0.5395680209225775, + "grad_norm": 1.6577558329799882, + "learning_rate": 1.8409706617402333e-06, + "loss": 0.7331, + "step": 5983 + }, + { + "epoch": 0.5396582044460477, + "grad_norm": 1.6559423484810454, + "learning_rate": 1.8403883076933378e-06, + "loss": 0.7722, + "step": 5984 + }, + { + "epoch": 0.539748387969518, + "grad_norm": 1.4662177800559664, + "learning_rate": 1.839805967265391e-06, + "loss": 0.745, + "step": 5985 + }, + { + "epoch": 0.5398385714929882, + "grad_norm": 2.0860464231934097, + "learning_rate": 1.839223640506083e-06, + "loss": 0.8086, + "step": 5986 + }, + { + "epoch": 0.5399287550164585, + "grad_norm": 1.9553249938770532, + "learning_rate": 1.8386413274650998e-06, + "loss": 0.7439, + "step": 5987 + }, + { + "epoch": 0.5400189385399288, + "grad_norm": 1.6763559347497485, + "learning_rate": 1.8380590281921294e-06, + "loss": 0.7993, + "step": 5988 + }, + { + "epoch": 0.540109122063399, + "grad_norm": 1.4734860619544963, + "learning_rate": 1.8374767427368552e-06, + "loss": 0.6595, + "step": 5989 + }, + { + "epoch": 0.5401993055868692, + "grad_norm": 2.0293582029502204, + "learning_rate": 1.8368944711489608e-06, + "loss": 0.783, + "step": 5990 + }, + { + "epoch": 0.5402894891103396, + "grad_norm": 2.1946024442146865, + "learning_rate": 1.8363122134781304e-06, + "loss": 0.7811, + "step": 5991 + }, + { + "epoch": 0.5403796726338098, + "grad_norm": 1.7206684156677705, + "learning_rate": 1.835729969774044e-06, + "loss": 0.78, + "step": 5992 + }, + { + "epoch": 0.54046985615728, + "grad_norm": 1.5612915706674027, + "learning_rate": 1.8351477400863823e-06, + "loss": 0.7601, + "step": 5993 + }, + { + "epoch": 0.5405600396807503, + "grad_norm": 1.751137227709752, + "learning_rate": 1.8345655244648249e-06, + "loss": 0.7308, + "step": 5994 + }, + { + "epoch": 0.5406502232042206, + "grad_norm": 1.3389545748827882, + "learning_rate": 1.8339833229590486e-06, + "loss": 0.7421, + "step": 5995 + }, + { + "epoch": 0.5407404067276909, + "grad_norm": 2.133374709483079, + "learning_rate": 1.833401135618731e-06, + "loss": 0.8259, + "step": 5996 + }, + { + "epoch": 0.5408305902511611, + "grad_norm": 1.4879725178570016, + "learning_rate": 1.8328189624935466e-06, + "loss": 0.682, + "step": 5997 + }, + { + "epoch": 0.5409207737746313, + "grad_norm": 1.7897881115008627, + "learning_rate": 1.832236803633171e-06, + "loss": 0.7861, + "step": 5998 + }, + { + "epoch": 0.5410109572981017, + "grad_norm": 1.6670172163744204, + "learning_rate": 1.831654659087276e-06, + "loss": 0.7165, + "step": 5999 + }, + { + "epoch": 0.5411011408215719, + "grad_norm": 1.6870141212898957, + "learning_rate": 1.831072528905533e-06, + "loss": 0.7701, + "step": 6000 + }, + { + "epoch": 0.5411913243450421, + "grad_norm": 1.4717046359655395, + "learning_rate": 1.8304904131376142e-06, + "loss": 0.8087, + "step": 6001 + }, + { + "epoch": 0.5412815078685124, + "grad_norm": 1.6650724256646436, + "learning_rate": 1.8299083118331874e-06, + "loss": 0.7701, + "step": 6002 + }, + { + "epoch": 0.5413716913919827, + "grad_norm": 1.921468738322449, + "learning_rate": 1.8293262250419217e-06, + "loss": 0.8856, + "step": 6003 + }, + { + "epoch": 0.541461874915453, + "grad_norm": 1.719676374995248, + "learning_rate": 1.828744152813484e-06, + "loss": 0.8672, + "step": 6004 + }, + { + "epoch": 0.5415520584389232, + "grad_norm": 1.6763879349158115, + "learning_rate": 1.8281620951975382e-06, + "loss": 0.7042, + "step": 6005 + }, + { + "epoch": 0.5416422419623935, + "grad_norm": 4.1511120696501225, + "learning_rate": 1.827580052243751e-06, + "loss": 0.7129, + "step": 6006 + }, + { + "epoch": 0.5417324254858638, + "grad_norm": 1.5269181617135874, + "learning_rate": 1.826998024001784e-06, + "loss": 0.7872, + "step": 6007 + }, + { + "epoch": 0.541822609009334, + "grad_norm": 5.187885683728164, + "learning_rate": 1.8264160105212995e-06, + "loss": 0.7777, + "step": 6008 + }, + { + "epoch": 0.5419127925328042, + "grad_norm": 2.131721972338864, + "learning_rate": 1.8258340118519582e-06, + "loss": 0.7223, + "step": 6009 + }, + { + "epoch": 0.5420029760562746, + "grad_norm": 1.8507907079774628, + "learning_rate": 1.82525202804342e-06, + "loss": 0.7383, + "step": 6010 + }, + { + "epoch": 0.5420931595797448, + "grad_norm": 1.8144856461319612, + "learning_rate": 1.8246700591453415e-06, + "loss": 0.7362, + "step": 6011 + }, + { + "epoch": 0.542183343103215, + "grad_norm": 1.5580699681105319, + "learning_rate": 1.8240881052073801e-06, + "loss": 0.7576, + "step": 6012 + }, + { + "epoch": 0.5422735266266853, + "grad_norm": 2.5085158267967524, + "learning_rate": 1.8235061662791923e-06, + "loss": 0.7206, + "step": 6013 + }, + { + "epoch": 0.5423637101501556, + "grad_norm": 1.7898949431379214, + "learning_rate": 1.8229242424104309e-06, + "loss": 0.8229, + "step": 6014 + }, + { + "epoch": 0.5424538936736258, + "grad_norm": 0.6202798464850159, + "learning_rate": 1.8223423336507503e-06, + "loss": 0.5598, + "step": 6015 + }, + { + "epoch": 0.5425440771970961, + "grad_norm": 1.8583916379851306, + "learning_rate": 1.8217604400498012e-06, + "loss": 0.7368, + "step": 6016 + }, + { + "epoch": 0.5426342607205663, + "grad_norm": 1.8035447767519572, + "learning_rate": 1.8211785616572333e-06, + "loss": 0.7994, + "step": 6017 + }, + { + "epoch": 0.5427244442440367, + "grad_norm": 2.0454132451598483, + "learning_rate": 1.8205966985226975e-06, + "loss": 0.8221, + "step": 6018 + }, + { + "epoch": 0.5428146277675069, + "grad_norm": 1.3721693552998564, + "learning_rate": 1.8200148506958397e-06, + "loss": 0.7618, + "step": 6019 + }, + { + "epoch": 0.5429048112909771, + "grad_norm": 0.6545130586042683, + "learning_rate": 1.819433018226308e-06, + "loss": 0.6068, + "step": 6020 + }, + { + "epoch": 0.5429949948144474, + "grad_norm": 0.5999981244375796, + "learning_rate": 1.8188512011637471e-06, + "loss": 0.551, + "step": 6021 + }, + { + "epoch": 0.5430851783379177, + "grad_norm": 1.7203174726118102, + "learning_rate": 1.8182693995578e-06, + "loss": 0.7303, + "step": 6022 + }, + { + "epoch": 0.5431753618613879, + "grad_norm": 1.4696158433949984, + "learning_rate": 1.8176876134581098e-06, + "loss": 0.6342, + "step": 6023 + }, + { + "epoch": 0.5432655453848582, + "grad_norm": 2.0345380726512303, + "learning_rate": 1.8171058429143176e-06, + "loss": 0.6997, + "step": 6024 + }, + { + "epoch": 0.5433557289083284, + "grad_norm": 1.4670297816191296, + "learning_rate": 1.8165240879760637e-06, + "loss": 0.7911, + "step": 6025 + }, + { + "epoch": 0.5434459124317987, + "grad_norm": 1.4696659721989476, + "learning_rate": 1.8159423486929862e-06, + "loss": 0.7829, + "step": 6026 + }, + { + "epoch": 0.543536095955269, + "grad_norm": 8.198224912969442, + "learning_rate": 1.815360625114722e-06, + "loss": 0.7523, + "step": 6027 + }, + { + "epoch": 0.5436262794787392, + "grad_norm": 1.5260361637714972, + "learning_rate": 1.814778917290908e-06, + "loss": 0.7369, + "step": 6028 + }, + { + "epoch": 0.5437164630022095, + "grad_norm": 2.026352597839044, + "learning_rate": 1.8141972252711773e-06, + "loss": 0.7697, + "step": 6029 + }, + { + "epoch": 0.5438066465256798, + "grad_norm": 1.897569685650254, + "learning_rate": 1.8136155491051645e-06, + "loss": 0.7484, + "step": 6030 + }, + { + "epoch": 0.54389683004915, + "grad_norm": 0.672190902431785, + "learning_rate": 1.8130338888424998e-06, + "loss": 0.5944, + "step": 6031 + }, + { + "epoch": 0.5439870135726202, + "grad_norm": 4.45765535897948, + "learning_rate": 1.812452244532816e-06, + "loss": 0.7737, + "step": 6032 + }, + { + "epoch": 0.5440771970960906, + "grad_norm": 1.6157276733357433, + "learning_rate": 1.8118706162257405e-06, + "loss": 0.6897, + "step": 6033 + }, + { + "epoch": 0.5441673806195608, + "grad_norm": 1.5537333411921836, + "learning_rate": 1.8112890039709002e-06, + "loss": 0.8324, + "step": 6034 + }, + { + "epoch": 0.5442575641430311, + "grad_norm": 1.7048730541189894, + "learning_rate": 1.8107074078179238e-06, + "loss": 0.7591, + "step": 6035 + }, + { + "epoch": 0.5443477476665013, + "grad_norm": 1.4624930683199948, + "learning_rate": 1.8101258278164348e-06, + "loss": 0.7045, + "step": 6036 + }, + { + "epoch": 0.5444379311899716, + "grad_norm": 1.5916498596076303, + "learning_rate": 1.8095442640160575e-06, + "loss": 0.86, + "step": 6037 + }, + { + "epoch": 0.5445281147134419, + "grad_norm": 2.1819016890367307, + "learning_rate": 1.8089627164664132e-06, + "loss": 0.7174, + "step": 6038 + }, + { + "epoch": 0.5446182982369121, + "grad_norm": 1.3265066272125743, + "learning_rate": 1.8083811852171233e-06, + "loss": 0.6729, + "step": 6039 + }, + { + "epoch": 0.5447084817603823, + "grad_norm": 1.613747443943969, + "learning_rate": 1.8077996703178078e-06, + "loss": 0.7524, + "step": 6040 + }, + { + "epoch": 0.5447986652838527, + "grad_norm": 1.5433198553503351, + "learning_rate": 1.8072181718180833e-06, + "loss": 0.7013, + "step": 6041 + }, + { + "epoch": 0.5448888488073229, + "grad_norm": 1.5826161751431416, + "learning_rate": 1.806636689767568e-06, + "loss": 0.8437, + "step": 6042 + }, + { + "epoch": 0.5449790323307931, + "grad_norm": 1.6855417828355874, + "learning_rate": 1.8060552242158769e-06, + "loss": 0.6902, + "step": 6043 + }, + { + "epoch": 0.5450692158542634, + "grad_norm": 1.5491712077161293, + "learning_rate": 1.8054737752126224e-06, + "loss": 0.7273, + "step": 6044 + }, + { + "epoch": 0.5451593993777337, + "grad_norm": 1.5142897396985247, + "learning_rate": 1.804892342807419e-06, + "loss": 0.7401, + "step": 6045 + }, + { + "epoch": 0.545249582901204, + "grad_norm": 2.3236603394643254, + "learning_rate": 1.8043109270498756e-06, + "loss": 0.8255, + "step": 6046 + }, + { + "epoch": 0.5453397664246742, + "grad_norm": 1.37275482153823, + "learning_rate": 1.803729527989604e-06, + "loss": 0.7631, + "step": 6047 + }, + { + "epoch": 0.5454299499481444, + "grad_norm": 2.714433685729385, + "learning_rate": 1.8031481456762112e-06, + "loss": 0.7489, + "step": 6048 + }, + { + "epoch": 0.5455201334716148, + "grad_norm": 1.8843733147001558, + "learning_rate": 1.8025667801593033e-06, + "loss": 0.6831, + "step": 6049 + }, + { + "epoch": 0.545610316995085, + "grad_norm": 2.1249325685460985, + "learning_rate": 1.8019854314884871e-06, + "loss": 0.6923, + "step": 6050 + }, + { + "epoch": 0.5457005005185552, + "grad_norm": 1.777475943208085, + "learning_rate": 1.8014040997133652e-06, + "loss": 0.8067, + "step": 6051 + }, + { + "epoch": 0.5457906840420256, + "grad_norm": 0.7577321511371772, + "learning_rate": 1.8008227848835414e-06, + "loss": 0.6148, + "step": 6052 + }, + { + "epoch": 0.5458808675654958, + "grad_norm": 1.9892616477933083, + "learning_rate": 1.8002414870486144e-06, + "loss": 0.7067, + "step": 6053 + }, + { + "epoch": 0.545971051088966, + "grad_norm": 1.3395605204979661, + "learning_rate": 1.7996602062581864e-06, + "loss": 0.7082, + "step": 6054 + }, + { + "epoch": 0.5460612346124363, + "grad_norm": 1.6287204261372557, + "learning_rate": 1.7990789425618544e-06, + "loss": 0.8031, + "step": 6055 + }, + { + "epoch": 0.5461514181359066, + "grad_norm": 1.379731792978853, + "learning_rate": 1.7984976960092137e-06, + "loss": 0.7357, + "step": 6056 + }, + { + "epoch": 0.5462416016593769, + "grad_norm": 2.10895415452423, + "learning_rate": 1.7979164666498617e-06, + "loss": 0.7228, + "step": 6057 + }, + { + "epoch": 0.5463317851828471, + "grad_norm": 0.617789807556522, + "learning_rate": 1.7973352545333901e-06, + "loss": 0.5543, + "step": 6058 + }, + { + "epoch": 0.5464219687063173, + "grad_norm": 1.674995931933573, + "learning_rate": 1.796754059709393e-06, + "loss": 0.6756, + "step": 6059 + }, + { + "epoch": 0.5465121522297877, + "grad_norm": 1.914161337032823, + "learning_rate": 1.7961728822274603e-06, + "loss": 0.8131, + "step": 6060 + }, + { + "epoch": 0.5466023357532579, + "grad_norm": 1.6117055370031437, + "learning_rate": 1.7955917221371802e-06, + "loss": 0.7117, + "step": 6061 + }, + { + "epoch": 0.5466925192767281, + "grad_norm": 1.831625562049854, + "learning_rate": 1.7950105794881422e-06, + "loss": 0.6692, + "step": 6062 + }, + { + "epoch": 0.5467827028001984, + "grad_norm": 1.3909163437731855, + "learning_rate": 1.7944294543299317e-06, + "loss": 0.7347, + "step": 6063 + }, + { + "epoch": 0.5468728863236687, + "grad_norm": 2.3540687765738895, + "learning_rate": 1.7938483467121333e-06, + "loss": 0.8097, + "step": 6064 + }, + { + "epoch": 0.5469630698471389, + "grad_norm": 1.7205486249895319, + "learning_rate": 1.7932672566843313e-06, + "loss": 0.7342, + "step": 6065 + }, + { + "epoch": 0.5470532533706092, + "grad_norm": 17.366491635856715, + "learning_rate": 1.7926861842961065e-06, + "loss": 0.6776, + "step": 6066 + }, + { + "epoch": 0.5471434368940794, + "grad_norm": 1.9459272891552128, + "learning_rate": 1.7921051295970399e-06, + "loss": 0.6754, + "step": 6067 + }, + { + "epoch": 0.5472336204175497, + "grad_norm": 1.5653718877010694, + "learning_rate": 1.7915240926367092e-06, + "loss": 0.7217, + "step": 6068 + }, + { + "epoch": 0.54732380394102, + "grad_norm": 1.6076428619548055, + "learning_rate": 1.7909430734646932e-06, + "loss": 0.794, + "step": 6069 + }, + { + "epoch": 0.5474139874644902, + "grad_norm": 0.6824228121877173, + "learning_rate": 1.790362072130567e-06, + "loss": 0.5872, + "step": 6070 + }, + { + "epoch": 0.5475041709879604, + "grad_norm": 1.5784462328004458, + "learning_rate": 1.7897810886839037e-06, + "loss": 0.7406, + "step": 6071 + }, + { + "epoch": 0.5475943545114308, + "grad_norm": 1.474863437214445, + "learning_rate": 1.7892001231742782e-06, + "loss": 0.7674, + "step": 6072 + }, + { + "epoch": 0.547684538034901, + "grad_norm": 1.614722469446552, + "learning_rate": 1.7886191756512594e-06, + "loss": 0.761, + "step": 6073 + }, + { + "epoch": 0.5477747215583713, + "grad_norm": 2.4778525665734845, + "learning_rate": 1.7880382461644192e-06, + "loss": 0.832, + "step": 6074 + }, + { + "epoch": 0.5478649050818416, + "grad_norm": 2.570726633768676, + "learning_rate": 1.7874573347633235e-06, + "loss": 0.7281, + "step": 6075 + }, + { + "epoch": 0.5479550886053118, + "grad_norm": 1.730676022897145, + "learning_rate": 1.7868764414975408e-06, + "loss": 0.8256, + "step": 6076 + }, + { + "epoch": 0.5480452721287821, + "grad_norm": 2.052283094981543, + "learning_rate": 1.7862955664166353e-06, + "loss": 0.6936, + "step": 6077 + }, + { + "epoch": 0.5481354556522523, + "grad_norm": 1.3198540247050623, + "learning_rate": 1.78571470957017e-06, + "loss": 0.8338, + "step": 6078 + }, + { + "epoch": 0.5482256391757226, + "grad_norm": 1.696045185449859, + "learning_rate": 1.7851338710077074e-06, + "loss": 0.823, + "step": 6079 + }, + { + "epoch": 0.5483158226991929, + "grad_norm": 2.0711996229142677, + "learning_rate": 1.7845530507788076e-06, + "loss": 0.8596, + "step": 6080 + }, + { + "epoch": 0.5484060062226631, + "grad_norm": 1.705157126293029, + "learning_rate": 1.7839722489330298e-06, + "loss": 0.7238, + "step": 6081 + }, + { + "epoch": 0.5484961897461333, + "grad_norm": 1.5396646009305939, + "learning_rate": 1.7833914655199308e-06, + "loss": 0.7231, + "step": 6082 + }, + { + "epoch": 0.5485863732696037, + "grad_norm": 2.120720874475305, + "learning_rate": 1.7828107005890658e-06, + "loss": 0.7988, + "step": 6083 + }, + { + "epoch": 0.5486765567930739, + "grad_norm": 1.9352573212836346, + "learning_rate": 1.7822299541899898e-06, + "loss": 0.6738, + "step": 6084 + }, + { + "epoch": 0.5487667403165442, + "grad_norm": 3.522211034765704, + "learning_rate": 1.7816492263722545e-06, + "loss": 0.7493, + "step": 6085 + }, + { + "epoch": 0.5488569238400144, + "grad_norm": 2.0119734223280736, + "learning_rate": 1.781068517185412e-06, + "loss": 0.7799, + "step": 6086 + }, + { + "epoch": 0.5489471073634847, + "grad_norm": 1.6794603770863799, + "learning_rate": 1.7804878266790104e-06, + "loss": 0.8182, + "step": 6087 + }, + { + "epoch": 0.549037290886955, + "grad_norm": 2.174672582251484, + "learning_rate": 1.779907154902597e-06, + "loss": 0.6831, + "step": 6088 + }, + { + "epoch": 0.5491274744104252, + "grad_norm": 2.4626030992332133, + "learning_rate": 1.7793265019057198e-06, + "loss": 0.6816, + "step": 6089 + }, + { + "epoch": 0.5492176579338954, + "grad_norm": 0.726395495510104, + "learning_rate": 1.7787458677379212e-06, + "loss": 0.5946, + "step": 6090 + }, + { + "epoch": 0.5493078414573658, + "grad_norm": 1.6430608922177163, + "learning_rate": 1.7781652524487463e-06, + "loss": 0.8101, + "step": 6091 + }, + { + "epoch": 0.549398024980836, + "grad_norm": 3.722057378417864, + "learning_rate": 1.777584656087735e-06, + "loss": 0.7026, + "step": 6092 + }, + { + "epoch": 0.5494882085043062, + "grad_norm": 1.8530976574892482, + "learning_rate": 1.777004078704427e-06, + "loss": 0.6757, + "step": 6093 + }, + { + "epoch": 0.5495783920277765, + "grad_norm": 1.6853946692214377, + "learning_rate": 1.7764235203483603e-06, + "loss": 0.7717, + "step": 6094 + }, + { + "epoch": 0.5496685755512468, + "grad_norm": 2.929327940175239, + "learning_rate": 1.775842981069072e-06, + "loss": 0.7224, + "step": 6095 + }, + { + "epoch": 0.549758759074717, + "grad_norm": 1.4756285928232482, + "learning_rate": 1.7752624609160966e-06, + "loss": 0.761, + "step": 6096 + }, + { + "epoch": 0.5498489425981873, + "grad_norm": 2.171735031777194, + "learning_rate": 1.7746819599389665e-06, + "loss": 0.791, + "step": 6097 + }, + { + "epoch": 0.5499391261216575, + "grad_norm": 2.984507133394837, + "learning_rate": 1.774101478187215e-06, + "loss": 0.7545, + "step": 6098 + }, + { + "epoch": 0.5500293096451279, + "grad_norm": 1.7339919758471336, + "learning_rate": 1.773521015710371e-06, + "loss": 0.8219, + "step": 6099 + }, + { + "epoch": 0.5501194931685981, + "grad_norm": 2.5979861860194333, + "learning_rate": 1.7729405725579614e-06, + "loss": 0.8342, + "step": 6100 + }, + { + "epoch": 0.5502096766920683, + "grad_norm": 1.5337776808351025, + "learning_rate": 1.7723601487795151e-06, + "loss": 0.7294, + "step": 6101 + }, + { + "epoch": 0.5502998602155387, + "grad_norm": 1.6130838712158715, + "learning_rate": 1.7717797444245557e-06, + "loss": 0.7012, + "step": 6102 + }, + { + "epoch": 0.5503900437390089, + "grad_norm": 1.2999099700270713, + "learning_rate": 1.7711993595426076e-06, + "loss": 0.6911, + "step": 6103 + }, + { + "epoch": 0.5504802272624791, + "grad_norm": 1.9838228433716194, + "learning_rate": 1.7706189941831915e-06, + "loss": 0.8624, + "step": 6104 + }, + { + "epoch": 0.5505704107859494, + "grad_norm": 1.8445286803690144, + "learning_rate": 1.770038648395827e-06, + "loss": 0.7126, + "step": 6105 + }, + { + "epoch": 0.5506605943094197, + "grad_norm": 0.6103100815982435, + "learning_rate": 1.7694583222300336e-06, + "loss": 0.5949, + "step": 6106 + }, + { + "epoch": 0.55075077783289, + "grad_norm": 1.6148849533722747, + "learning_rate": 1.7688780157353272e-06, + "loss": 0.7354, + "step": 6107 + }, + { + "epoch": 0.5508409613563602, + "grad_norm": 1.8137710323994016, + "learning_rate": 1.768297728961223e-06, + "loss": 0.6963, + "step": 6108 + }, + { + "epoch": 0.5509311448798304, + "grad_norm": 2.194298995572439, + "learning_rate": 1.7677174619572342e-06, + "loss": 0.7588, + "step": 6109 + }, + { + "epoch": 0.5510213284033008, + "grad_norm": 2.734150904600625, + "learning_rate": 1.7671372147728717e-06, + "loss": 0.774, + "step": 6110 + }, + { + "epoch": 0.551111511926771, + "grad_norm": 1.7978424121762662, + "learning_rate": 1.7665569874576471e-06, + "loss": 0.8208, + "step": 6111 + }, + { + "epoch": 0.5512016954502412, + "grad_norm": 3.0656794956352655, + "learning_rate": 1.7659767800610664e-06, + "loss": 0.7582, + "step": 6112 + }, + { + "epoch": 0.5512918789737115, + "grad_norm": 3.8475747036885286, + "learning_rate": 1.7653965926326379e-06, + "loss": 0.6626, + "step": 6113 + }, + { + "epoch": 0.5513820624971818, + "grad_norm": 1.7698568082529615, + "learning_rate": 1.764816425221866e-06, + "loss": 0.7108, + "step": 6114 + }, + { + "epoch": 0.551472246020652, + "grad_norm": 1.4877907252506057, + "learning_rate": 1.7642362778782524e-06, + "loss": 0.67, + "step": 6115 + }, + { + "epoch": 0.5515624295441223, + "grad_norm": 1.89304241394472, + "learning_rate": 1.7636561506513005e-06, + "loss": 0.7337, + "step": 6116 + }, + { + "epoch": 0.5516526130675925, + "grad_norm": 1.412726514090141, + "learning_rate": 1.7630760435905083e-06, + "loss": 0.7278, + "step": 6117 + }, + { + "epoch": 0.5517427965910628, + "grad_norm": 1.726847560359443, + "learning_rate": 1.762495956745375e-06, + "loss": 0.7094, + "step": 6118 + }, + { + "epoch": 0.5518329801145331, + "grad_norm": 1.7223955506435178, + "learning_rate": 1.7619158901653962e-06, + "loss": 0.8038, + "step": 6119 + }, + { + "epoch": 0.5519231636380033, + "grad_norm": 1.6400553122938941, + "learning_rate": 1.761335843900066e-06, + "loss": 0.7519, + "step": 6120 + }, + { + "epoch": 0.5520133471614735, + "grad_norm": 1.4729896390645054, + "learning_rate": 1.7607558179988785e-06, + "loss": 0.7375, + "step": 6121 + }, + { + "epoch": 0.5521035306849439, + "grad_norm": 1.6295975423473166, + "learning_rate": 1.760175812511323e-06, + "loss": 0.6814, + "step": 6122 + }, + { + "epoch": 0.5521937142084141, + "grad_norm": 1.5972645472940161, + "learning_rate": 1.75959582748689e-06, + "loss": 0.7142, + "step": 6123 + }, + { + "epoch": 0.5522838977318844, + "grad_norm": 1.658479665245724, + "learning_rate": 1.7590158629750657e-06, + "loss": 0.7879, + "step": 6124 + }, + { + "epoch": 0.5523740812553547, + "grad_norm": 1.6054903571909365, + "learning_rate": 1.7584359190253376e-06, + "loss": 0.8048, + "step": 6125 + }, + { + "epoch": 0.5524642647788249, + "grad_norm": 2.493405609443262, + "learning_rate": 1.7578559956871892e-06, + "loss": 0.8177, + "step": 6126 + }, + { + "epoch": 0.5525544483022952, + "grad_norm": 1.4657633018716383, + "learning_rate": 1.7572760930101012e-06, + "loss": 0.7501, + "step": 6127 + }, + { + "epoch": 0.5526446318257654, + "grad_norm": 0.6863415885564379, + "learning_rate": 1.7566962110435563e-06, + "loss": 0.6342, + "step": 6128 + }, + { + "epoch": 0.5527348153492357, + "grad_norm": 1.559700709023143, + "learning_rate": 1.7561163498370313e-06, + "loss": 0.7702, + "step": 6129 + }, + { + "epoch": 0.552824998872706, + "grad_norm": 1.5429221617027102, + "learning_rate": 1.755536509440005e-06, + "loss": 0.7175, + "step": 6130 + }, + { + "epoch": 0.5529151823961762, + "grad_norm": 3.125153499648026, + "learning_rate": 1.7549566899019519e-06, + "loss": 0.7531, + "step": 6131 + }, + { + "epoch": 0.5530053659196464, + "grad_norm": 1.618078087944056, + "learning_rate": 1.754376891272344e-06, + "loss": 0.7682, + "step": 6132 + }, + { + "epoch": 0.5530955494431168, + "grad_norm": 1.709786642048687, + "learning_rate": 1.753797113600655e-06, + "loss": 0.703, + "step": 6133 + }, + { + "epoch": 0.553185732966587, + "grad_norm": 1.8543190054158523, + "learning_rate": 1.7532173569363535e-06, + "loss": 0.8901, + "step": 6134 + }, + { + "epoch": 0.5532759164900573, + "grad_norm": 1.9803557655263837, + "learning_rate": 1.7526376213289077e-06, + "loss": 0.5705, + "step": 6135 + }, + { + "epoch": 0.5533661000135275, + "grad_norm": 1.9409924685977213, + "learning_rate": 1.7520579068277844e-06, + "loss": 0.7378, + "step": 6136 + }, + { + "epoch": 0.5534562835369978, + "grad_norm": 1.6021702055409697, + "learning_rate": 1.7514782134824472e-06, + "loss": 0.7808, + "step": 6137 + }, + { + "epoch": 0.5535464670604681, + "grad_norm": 1.6871177805615887, + "learning_rate": 1.7508985413423599e-06, + "loss": 0.7474, + "step": 6138 + }, + { + "epoch": 0.5536366505839383, + "grad_norm": 1.5773766886883989, + "learning_rate": 1.7503188904569814e-06, + "loss": 0.8261, + "step": 6139 + }, + { + "epoch": 0.5537268341074085, + "grad_norm": 2.8335732938092777, + "learning_rate": 1.7497392608757728e-06, + "loss": 0.7229, + "step": 6140 + }, + { + "epoch": 0.5538170176308789, + "grad_norm": 1.6642938730635128, + "learning_rate": 1.7491596526481897e-06, + "loss": 0.7708, + "step": 6141 + }, + { + "epoch": 0.5539072011543491, + "grad_norm": 1.646690708685022, + "learning_rate": 1.7485800658236888e-06, + "loss": 0.6819, + "step": 6142 + }, + { + "epoch": 0.5539973846778193, + "grad_norm": 2.603810298695786, + "learning_rate": 1.7480005004517232e-06, + "loss": 0.8264, + "step": 6143 + }, + { + "epoch": 0.5540875682012896, + "grad_norm": 1.6757403459869846, + "learning_rate": 1.7474209565817435e-06, + "loss": 0.8185, + "step": 6144 + }, + { + "epoch": 0.5541777517247599, + "grad_norm": 1.3923724906796722, + "learning_rate": 1.7468414342632014e-06, + "loss": 0.7546, + "step": 6145 + }, + { + "epoch": 0.5542679352482301, + "grad_norm": 1.7569347309284644, + "learning_rate": 1.746261933545543e-06, + "loss": 0.807, + "step": 6146 + }, + { + "epoch": 0.5543581187717004, + "grad_norm": 1.4004868137837911, + "learning_rate": 1.7456824544782165e-06, + "loss": 0.7482, + "step": 6147 + }, + { + "epoch": 0.5544483022951707, + "grad_norm": 1.8707991907732577, + "learning_rate": 1.7451029971106653e-06, + "loss": 0.7582, + "step": 6148 + }, + { + "epoch": 0.554538485818641, + "grad_norm": 1.4152309585096252, + "learning_rate": 1.7445235614923313e-06, + "loss": 0.7817, + "step": 6149 + }, + { + "epoch": 0.5546286693421112, + "grad_norm": 1.7937155770112407, + "learning_rate": 1.7439441476726556e-06, + "loss": 0.7415, + "step": 6150 + }, + { + "epoch": 0.5547188528655814, + "grad_norm": 3.7768087235007193, + "learning_rate": 1.7433647557010776e-06, + "loss": 0.761, + "step": 6151 + }, + { + "epoch": 0.5548090363890518, + "grad_norm": 2.3237669433501362, + "learning_rate": 1.7427853856270338e-06, + "loss": 0.7169, + "step": 6152 + }, + { + "epoch": 0.554899219912522, + "grad_norm": 2.2254038154970455, + "learning_rate": 1.7422060374999593e-06, + "loss": 0.7172, + "step": 6153 + }, + { + "epoch": 0.5549894034359922, + "grad_norm": 1.502633802451149, + "learning_rate": 1.7416267113692862e-06, + "loss": 0.7136, + "step": 6154 + }, + { + "epoch": 0.5550795869594625, + "grad_norm": 1.5000328219319632, + "learning_rate": 1.7410474072844475e-06, + "loss": 0.784, + "step": 6155 + }, + { + "epoch": 0.5551697704829328, + "grad_norm": 1.4855798750248117, + "learning_rate": 1.740468125294871e-06, + "loss": 0.7363, + "step": 6156 + }, + { + "epoch": 0.555259954006403, + "grad_norm": 2.251271206651575, + "learning_rate": 1.739888865449986e-06, + "loss": 0.6694, + "step": 6157 + }, + { + "epoch": 0.5553501375298733, + "grad_norm": 6.476842858075463, + "learning_rate": 1.7393096277992174e-06, + "loss": 0.7773, + "step": 6158 + }, + { + "epoch": 0.5554403210533435, + "grad_norm": 1.5893667944344931, + "learning_rate": 1.738730412391988e-06, + "loss": 0.7663, + "step": 6159 + }, + { + "epoch": 0.5555305045768139, + "grad_norm": 3.9261855794970875, + "learning_rate": 1.738151219277721e-06, + "loss": 0.8044, + "step": 6160 + }, + { + "epoch": 0.5556206881002841, + "grad_norm": 2.926956898050334, + "learning_rate": 1.7375720485058349e-06, + "loss": 0.713, + "step": 6161 + }, + { + "epoch": 0.5557108716237543, + "grad_norm": 2.050157085075758, + "learning_rate": 1.7369929001257498e-06, + "loss": 0.7738, + "step": 6162 + }, + { + "epoch": 0.5558010551472246, + "grad_norm": 1.961153360469456, + "learning_rate": 1.73641377418688e-06, + "loss": 0.6678, + "step": 6163 + }, + { + "epoch": 0.5558912386706949, + "grad_norm": 1.8013000402971628, + "learning_rate": 1.7358346707386408e-06, + "loss": 0.6878, + "step": 6164 + }, + { + "epoch": 0.5559814221941651, + "grad_norm": 1.8785542180124073, + "learning_rate": 1.7352555898304439e-06, + "loss": 0.8052, + "step": 6165 + }, + { + "epoch": 0.5560716057176354, + "grad_norm": 1.7650704652374132, + "learning_rate": 1.7346765315116996e-06, + "loss": 0.6794, + "step": 6166 + }, + { + "epoch": 0.5561617892411056, + "grad_norm": 1.5678659995586426, + "learning_rate": 1.734097495831817e-06, + "loss": 0.7146, + "step": 6167 + }, + { + "epoch": 0.5562519727645759, + "grad_norm": 1.8206116262744194, + "learning_rate": 1.7335184828402015e-06, + "loss": 0.7641, + "step": 6168 + }, + { + "epoch": 0.5563421562880462, + "grad_norm": 1.4093627810741503, + "learning_rate": 1.7329394925862595e-06, + "loss": 0.7327, + "step": 6169 + }, + { + "epoch": 0.5564323398115164, + "grad_norm": 2.1686785111751266, + "learning_rate": 1.7323605251193922e-06, + "loss": 0.7994, + "step": 6170 + }, + { + "epoch": 0.5565225233349868, + "grad_norm": 2.0228507468905255, + "learning_rate": 1.7317815804890001e-06, + "loss": 0.7738, + "step": 6171 + }, + { + "epoch": 0.556612706858457, + "grad_norm": 1.7175117887622644, + "learning_rate": 1.731202658744483e-06, + "loss": 0.709, + "step": 6172 + }, + { + "epoch": 0.5567028903819272, + "grad_norm": 1.4656775784364289, + "learning_rate": 1.7306237599352365e-06, + "loss": 0.8147, + "step": 6173 + }, + { + "epoch": 0.5567930739053975, + "grad_norm": 1.419687789658339, + "learning_rate": 1.730044884110657e-06, + "loss": 0.7409, + "step": 6174 + }, + { + "epoch": 0.5568832574288678, + "grad_norm": 1.670462394818588, + "learning_rate": 1.7294660313201366e-06, + "loss": 0.714, + "step": 6175 + }, + { + "epoch": 0.556973440952338, + "grad_norm": 2.436936337759859, + "learning_rate": 1.7288872016130652e-06, + "loss": 0.7351, + "step": 6176 + }, + { + "epoch": 0.5570636244758083, + "grad_norm": 1.9525650442428035, + "learning_rate": 1.7283083950388334e-06, + "loss": 0.666, + "step": 6177 + }, + { + "epoch": 0.5571538079992785, + "grad_norm": 1.5656136578969464, + "learning_rate": 1.727729611646827e-06, + "loss": 0.658, + "step": 6178 + }, + { + "epoch": 0.5572439915227488, + "grad_norm": 1.5842849314593233, + "learning_rate": 1.7271508514864318e-06, + "loss": 0.8071, + "step": 6179 + }, + { + "epoch": 0.5573341750462191, + "grad_norm": 1.9457387801837789, + "learning_rate": 1.7265721146070302e-06, + "loss": 0.7561, + "step": 6180 + }, + { + "epoch": 0.5574243585696893, + "grad_norm": 1.6756252402931846, + "learning_rate": 1.7259934010580035e-06, + "loss": 0.752, + "step": 6181 + }, + { + "epoch": 0.5575145420931595, + "grad_norm": 1.6295423843569616, + "learning_rate": 1.725414710888731e-06, + "loss": 0.7031, + "step": 6182 + }, + { + "epoch": 0.5576047256166299, + "grad_norm": 1.5093671462839962, + "learning_rate": 1.7248360441485885e-06, + "loss": 0.728, + "step": 6183 + }, + { + "epoch": 0.5576949091401001, + "grad_norm": 1.6067929341070408, + "learning_rate": 1.7242574008869528e-06, + "loss": 0.8271, + "step": 6184 + }, + { + "epoch": 0.5577850926635703, + "grad_norm": 0.6808862091046614, + "learning_rate": 1.7236787811531951e-06, + "loss": 0.6287, + "step": 6185 + }, + { + "epoch": 0.5578752761870406, + "grad_norm": 1.579398896882299, + "learning_rate": 1.7231001849966887e-06, + "loss": 0.7211, + "step": 6186 + }, + { + "epoch": 0.5579654597105109, + "grad_norm": 0.6473598487614537, + "learning_rate": 1.722521612466801e-06, + "loss": 0.6076, + "step": 6187 + }, + { + "epoch": 0.5580556432339812, + "grad_norm": 1.9033324631234143, + "learning_rate": 1.7219430636128984e-06, + "loss": 0.694, + "step": 6188 + }, + { + "epoch": 0.5581458267574514, + "grad_norm": 1.6817134969874348, + "learning_rate": 1.7213645384843479e-06, + "loss": 0.6963, + "step": 6189 + }, + { + "epoch": 0.5582360102809216, + "grad_norm": 1.712295202553361, + "learning_rate": 1.7207860371305108e-06, + "loss": 0.7444, + "step": 6190 + }, + { + "epoch": 0.558326193804392, + "grad_norm": 2.6211812534876815, + "learning_rate": 1.7202075596007487e-06, + "loss": 0.7197, + "step": 6191 + }, + { + "epoch": 0.5584163773278622, + "grad_norm": 1.664742487562638, + "learning_rate": 1.7196291059444206e-06, + "loss": 0.7076, + "step": 6192 + }, + { + "epoch": 0.5585065608513324, + "grad_norm": 2.628390529431941, + "learning_rate": 1.7190506762108828e-06, + "loss": 0.7399, + "step": 6193 + }, + { + "epoch": 0.5585967443748028, + "grad_norm": 1.6994357266506797, + "learning_rate": 1.7184722704494907e-06, + "loss": 0.7529, + "step": 6194 + }, + { + "epoch": 0.558686927898273, + "grad_norm": 1.5550130378691733, + "learning_rate": 1.717893888709596e-06, + "loss": 0.7656, + "step": 6195 + }, + { + "epoch": 0.5587771114217432, + "grad_norm": 1.6603388427850256, + "learning_rate": 1.7173155310405515e-06, + "loss": 0.7786, + "step": 6196 + }, + { + "epoch": 0.5588672949452135, + "grad_norm": 1.4071873401944186, + "learning_rate": 1.7167371974917043e-06, + "loss": 0.8269, + "step": 6197 + }, + { + "epoch": 0.5589574784686838, + "grad_norm": 2.1630377701677146, + "learning_rate": 1.7161588881124003e-06, + "loss": 0.7727, + "step": 6198 + }, + { + "epoch": 0.559047661992154, + "grad_norm": 1.9849740273148109, + "learning_rate": 1.7155806029519861e-06, + "loss": 0.7368, + "step": 6199 + }, + { + "epoch": 0.5591378455156243, + "grad_norm": 1.7790856178399224, + "learning_rate": 1.7150023420598023e-06, + "loss": 0.7921, + "step": 6200 + }, + { + "epoch": 0.5592280290390945, + "grad_norm": 1.4938415626969674, + "learning_rate": 1.714424105485191e-06, + "loss": 0.8057, + "step": 6201 + }, + { + "epoch": 0.5593182125625649, + "grad_norm": 3.486018866491778, + "learning_rate": 1.7138458932774896e-06, + "loss": 0.7129, + "step": 6202 + }, + { + "epoch": 0.5594083960860351, + "grad_norm": 0.655209829594572, + "learning_rate": 1.7132677054860335e-06, + "loss": 0.6073, + "step": 6203 + }, + { + "epoch": 0.5594985796095053, + "grad_norm": 1.8474322257818883, + "learning_rate": 1.7126895421601586e-06, + "loss": 0.7742, + "step": 6204 + }, + { + "epoch": 0.5595887631329756, + "grad_norm": 2.208426731461758, + "learning_rate": 1.712111403349196e-06, + "loss": 0.6907, + "step": 6205 + }, + { + "epoch": 0.5596789466564459, + "grad_norm": 1.7983383854019337, + "learning_rate": 1.7115332891024757e-06, + "loss": 0.787, + "step": 6206 + }, + { + "epoch": 0.5597691301799161, + "grad_norm": 0.7180063298034053, + "learning_rate": 1.7109551994693257e-06, + "loss": 0.6571, + "step": 6207 + }, + { + "epoch": 0.5598593137033864, + "grad_norm": 2.285399960294833, + "learning_rate": 1.7103771344990725e-06, + "loss": 0.6962, + "step": 6208 + }, + { + "epoch": 0.5599494972268566, + "grad_norm": 1.7267016186484392, + "learning_rate": 1.709799094241039e-06, + "loss": 0.6715, + "step": 6209 + }, + { + "epoch": 0.560039680750327, + "grad_norm": 20.591530015966214, + "learning_rate": 1.709221078744546e-06, + "loss": 0.784, + "step": 6210 + }, + { + "epoch": 0.5601298642737972, + "grad_norm": 1.5126154800341172, + "learning_rate": 1.7086430880589148e-06, + "loss": 0.7628, + "step": 6211 + }, + { + "epoch": 0.5602200477972674, + "grad_norm": 1.870089076057482, + "learning_rate": 1.7080651222334612e-06, + "loss": 0.7564, + "step": 6212 + }, + { + "epoch": 0.5603102313207377, + "grad_norm": 2.048309522365715, + "learning_rate": 1.7074871813175018e-06, + "loss": 0.8104, + "step": 6213 + }, + { + "epoch": 0.560400414844208, + "grad_norm": 1.7379191160706835, + "learning_rate": 1.706909265360349e-06, + "loss": 0.6701, + "step": 6214 + }, + { + "epoch": 0.5604905983676782, + "grad_norm": 0.7050425977553285, + "learning_rate": 1.7063313744113128e-06, + "loss": 0.5964, + "step": 6215 + }, + { + "epoch": 0.5605807818911485, + "grad_norm": 0.6420772652790047, + "learning_rate": 1.7057535085197042e-06, + "loss": 0.5699, + "step": 6216 + }, + { + "epoch": 0.5606709654146187, + "grad_norm": 3.641774925060219, + "learning_rate": 1.705175667734828e-06, + "loss": 0.6727, + "step": 6217 + }, + { + "epoch": 0.560761148938089, + "grad_norm": 3.7473358863487523, + "learning_rate": 1.7045978521059894e-06, + "loss": 0.7202, + "step": 6218 + }, + { + "epoch": 0.5608513324615593, + "grad_norm": 1.3822082233708155, + "learning_rate": 1.7040200616824914e-06, + "loss": 0.7516, + "step": 6219 + }, + { + "epoch": 0.5609415159850295, + "grad_norm": 1.7495657518300953, + "learning_rate": 1.7034422965136333e-06, + "loss": 0.696, + "step": 6220 + }, + { + "epoch": 0.5610316995084998, + "grad_norm": 1.9343339216218367, + "learning_rate": 1.7028645566487137e-06, + "loss": 0.7908, + "step": 6221 + }, + { + "epoch": 0.5611218830319701, + "grad_norm": 1.7057804789117292, + "learning_rate": 1.7022868421370284e-06, + "loss": 0.7244, + "step": 6222 + }, + { + "epoch": 0.5612120665554403, + "grad_norm": 2.0630931579295924, + "learning_rate": 1.701709153027872e-06, + "loss": 0.7966, + "step": 6223 + }, + { + "epoch": 0.5613022500789105, + "grad_norm": 2.2312846119651084, + "learning_rate": 1.7011314893705353e-06, + "loss": 0.7139, + "step": 6224 + }, + { + "epoch": 0.5613924336023809, + "grad_norm": 1.47976684035347, + "learning_rate": 1.700553851214307e-06, + "loss": 0.7547, + "step": 6225 + }, + { + "epoch": 0.5614826171258511, + "grad_norm": 1.5019067724885824, + "learning_rate": 1.699976238608476e-06, + "loss": 0.6198, + "step": 6226 + }, + { + "epoch": 0.5615728006493214, + "grad_norm": 3.85348914437493, + "learning_rate": 1.699398651602326e-06, + "loss": 0.6998, + "step": 6227 + }, + { + "epoch": 0.5616629841727916, + "grad_norm": 1.4224062754991436, + "learning_rate": 1.6988210902451413e-06, + "loss": 0.7908, + "step": 6228 + }, + { + "epoch": 0.5617531676962619, + "grad_norm": 1.7845361079023043, + "learning_rate": 1.6982435545862011e-06, + "loss": 0.7331, + "step": 6229 + }, + { + "epoch": 0.5618433512197322, + "grad_norm": 1.643587688946566, + "learning_rate": 1.6976660446747853e-06, + "loss": 0.7597, + "step": 6230 + }, + { + "epoch": 0.5619335347432024, + "grad_norm": 2.072998722567351, + "learning_rate": 1.6970885605601696e-06, + "loss": 0.6799, + "step": 6231 + }, + { + "epoch": 0.5620237182666726, + "grad_norm": 1.7010659886723951, + "learning_rate": 1.6965111022916282e-06, + "loss": 0.8592, + "step": 6232 + }, + { + "epoch": 0.562113901790143, + "grad_norm": 2.071811924775472, + "learning_rate": 1.6959336699184323e-06, + "loss": 0.7256, + "step": 6233 + }, + { + "epoch": 0.5622040853136132, + "grad_norm": 2.322559642867699, + "learning_rate": 1.6953562634898529e-06, + "loss": 0.7426, + "step": 6234 + }, + { + "epoch": 0.5622942688370834, + "grad_norm": 2.4274072280898125, + "learning_rate": 1.6947788830551569e-06, + "loss": 0.722, + "step": 6235 + }, + { + "epoch": 0.5623844523605537, + "grad_norm": 0.7703284766840419, + "learning_rate": 1.6942015286636093e-06, + "loss": 0.6306, + "step": 6236 + }, + { + "epoch": 0.562474635884024, + "grad_norm": 1.6691095886273184, + "learning_rate": 1.6936242003644735e-06, + "loss": 0.7362, + "step": 6237 + }, + { + "epoch": 0.5625648194074943, + "grad_norm": 2.3260188464764893, + "learning_rate": 1.6930468982070106e-06, + "loss": 0.8203, + "step": 6238 + }, + { + "epoch": 0.5626550029309645, + "grad_norm": 0.5958387877030132, + "learning_rate": 1.692469622240478e-06, + "loss": 0.5184, + "step": 6239 + }, + { + "epoch": 0.5627451864544347, + "grad_norm": 1.6286017775994228, + "learning_rate": 1.6918923725141339e-06, + "loss": 0.7058, + "step": 6240 + }, + { + "epoch": 0.5628353699779051, + "grad_norm": 1.6916767282922747, + "learning_rate": 1.6913151490772312e-06, + "loss": 0.8018, + "step": 6241 + }, + { + "epoch": 0.5629255535013753, + "grad_norm": 1.5652178206304985, + "learning_rate": 1.6907379519790215e-06, + "loss": 0.7881, + "step": 6242 + }, + { + "epoch": 0.5630157370248455, + "grad_norm": 1.965333790274834, + "learning_rate": 1.6901607812687558e-06, + "loss": 0.6615, + "step": 6243 + }, + { + "epoch": 0.5631059205483159, + "grad_norm": 1.4171849125646083, + "learning_rate": 1.6895836369956794e-06, + "loss": 0.7061, + "step": 6244 + }, + { + "epoch": 0.5631961040717861, + "grad_norm": 0.6691290164615314, + "learning_rate": 1.6890065192090402e-06, + "loss": 0.543, + "step": 6245 + }, + { + "epoch": 0.5632862875952563, + "grad_norm": 2.67770054056396, + "learning_rate": 1.6884294279580793e-06, + "loss": 0.6506, + "step": 6246 + }, + { + "epoch": 0.5633764711187266, + "grad_norm": 8.855871944394172, + "learning_rate": 1.6878523632920371e-06, + "loss": 0.775, + "step": 6247 + }, + { + "epoch": 0.5634666546421969, + "grad_norm": 1.6586604302146988, + "learning_rate": 1.6872753252601525e-06, + "loss": 0.8171, + "step": 6248 + }, + { + "epoch": 0.5635568381656672, + "grad_norm": 3.6147291756852, + "learning_rate": 1.6866983139116616e-06, + "loss": 0.6894, + "step": 6249 + }, + { + "epoch": 0.5636470216891374, + "grad_norm": 1.676593290332436, + "learning_rate": 1.6861213292957981e-06, + "loss": 0.7185, + "step": 6250 + }, + { + "epoch": 0.5637372052126076, + "grad_norm": 2.118632594093677, + "learning_rate": 1.685544371461793e-06, + "loss": 0.783, + "step": 6251 + }, + { + "epoch": 0.563827388736078, + "grad_norm": 2.4882044997423023, + "learning_rate": 1.6849674404588767e-06, + "loss": 0.7719, + "step": 6252 + }, + { + "epoch": 0.5639175722595482, + "grad_norm": 2.1712402953181917, + "learning_rate": 1.6843905363362758e-06, + "loss": 0.7859, + "step": 6253 + }, + { + "epoch": 0.5640077557830184, + "grad_norm": 1.9004904239556801, + "learning_rate": 1.6838136591432136e-06, + "loss": 0.6929, + "step": 6254 + }, + { + "epoch": 0.5640979393064887, + "grad_norm": 2.112284709185168, + "learning_rate": 1.6832368089289139e-06, + "loss": 0.7733, + "step": 6255 + }, + { + "epoch": 0.564188122829959, + "grad_norm": 1.5598237578776804, + "learning_rate": 1.682659985742596e-06, + "loss": 0.7647, + "step": 6256 + }, + { + "epoch": 0.5642783063534292, + "grad_norm": 2.5349493427218652, + "learning_rate": 1.6820831896334782e-06, + "loss": 0.7892, + "step": 6257 + }, + { + "epoch": 0.5643684898768995, + "grad_norm": 1.6231568961375114, + "learning_rate": 1.681506420650776e-06, + "loss": 0.7684, + "step": 6258 + }, + { + "epoch": 0.5644586734003697, + "grad_norm": 1.4858100780322423, + "learning_rate": 1.680929678843701e-06, + "loss": 0.7355, + "step": 6259 + }, + { + "epoch": 0.56454885692384, + "grad_norm": 1.632868204581161, + "learning_rate": 1.6803529642614662e-06, + "loss": 0.7966, + "step": 6260 + }, + { + "epoch": 0.5646390404473103, + "grad_norm": 1.7200898236839661, + "learning_rate": 1.6797762769532785e-06, + "loss": 0.7603, + "step": 6261 + }, + { + "epoch": 0.5647292239707805, + "grad_norm": 1.664684770315835, + "learning_rate": 1.679199616968345e-06, + "loss": 0.784, + "step": 6262 + }, + { + "epoch": 0.5648194074942507, + "grad_norm": 1.536215458009033, + "learning_rate": 1.6786229843558689e-06, + "loss": 0.7773, + "step": 6263 + }, + { + "epoch": 0.5649095910177211, + "grad_norm": 2.046520988776128, + "learning_rate": 1.6780463791650514e-06, + "loss": 0.7304, + "step": 6264 + }, + { + "epoch": 0.5649997745411913, + "grad_norm": 1.7457394507148265, + "learning_rate": 1.6774698014450928e-06, + "loss": 0.6942, + "step": 6265 + }, + { + "epoch": 0.5650899580646616, + "grad_norm": 0.6543944556505632, + "learning_rate": 1.6768932512451883e-06, + "loss": 0.5787, + "step": 6266 + }, + { + "epoch": 0.5651801415881319, + "grad_norm": 3.636267678121775, + "learning_rate": 1.676316728614534e-06, + "loss": 0.7031, + "step": 6267 + }, + { + "epoch": 0.5652703251116021, + "grad_norm": 0.5712201720866509, + "learning_rate": 1.675740233602321e-06, + "loss": 0.5885, + "step": 6268 + }, + { + "epoch": 0.5653605086350724, + "grad_norm": 1.5584262385017875, + "learning_rate": 1.6751637662577385e-06, + "loss": 0.7958, + "step": 6269 + }, + { + "epoch": 0.5654506921585426, + "grad_norm": 0.7936242311816112, + "learning_rate": 1.6745873266299753e-06, + "loss": 0.5322, + "step": 6270 + }, + { + "epoch": 0.565540875682013, + "grad_norm": 1.434584023531525, + "learning_rate": 1.6740109147682148e-06, + "loss": 0.7099, + "step": 6271 + }, + { + "epoch": 0.5656310592054832, + "grad_norm": 1.5693842483424232, + "learning_rate": 1.6734345307216418e-06, + "loss": 0.7647, + "step": 6272 + }, + { + "epoch": 0.5657212427289534, + "grad_norm": 2.158710016301115, + "learning_rate": 1.6728581745394346e-06, + "loss": 0.7253, + "step": 6273 + }, + { + "epoch": 0.5658114262524236, + "grad_norm": 1.9807841935707267, + "learning_rate": 1.672281846270772e-06, + "loss": 0.8269, + "step": 6274 + }, + { + "epoch": 0.565901609775894, + "grad_norm": 1.7764531513739643, + "learning_rate": 1.6717055459648295e-06, + "loss": 0.7462, + "step": 6275 + }, + { + "epoch": 0.5659917932993642, + "grad_norm": 1.7955361727430568, + "learning_rate": 1.6711292736707793e-06, + "loss": 0.7042, + "step": 6276 + }, + { + "epoch": 0.5660819768228345, + "grad_norm": 2.161329947339339, + "learning_rate": 1.6705530294377938e-06, + "loss": 0.8202, + "step": 6277 + }, + { + "epoch": 0.5661721603463047, + "grad_norm": 2.6776285074064208, + "learning_rate": 1.6699768133150395e-06, + "loss": 0.7389, + "step": 6278 + }, + { + "epoch": 0.566262343869775, + "grad_norm": 2.0274975668707946, + "learning_rate": 1.6694006253516837e-06, + "loss": 0.7371, + "step": 6279 + }, + { + "epoch": 0.5663525273932453, + "grad_norm": 1.7870401337764585, + "learning_rate": 1.6688244655968896e-06, + "loss": 0.7468, + "step": 6280 + }, + { + "epoch": 0.5664427109167155, + "grad_norm": 4.117068670827294, + "learning_rate": 1.6682483340998175e-06, + "loss": 0.6763, + "step": 6281 + }, + { + "epoch": 0.5665328944401857, + "grad_norm": 1.6248342722966578, + "learning_rate": 1.6676722309096276e-06, + "loss": 0.7109, + "step": 6282 + }, + { + "epoch": 0.5666230779636561, + "grad_norm": 0.7356486942361342, + "learning_rate": 1.6670961560754744e-06, + "loss": 0.6477, + "step": 6283 + }, + { + "epoch": 0.5667132614871263, + "grad_norm": 1.6172932023596345, + "learning_rate": 1.6665201096465138e-06, + "loss": 0.7443, + "step": 6284 + }, + { + "epoch": 0.5668034450105965, + "grad_norm": 1.6999242260819056, + "learning_rate": 1.6659440916718961e-06, + "loss": 0.7917, + "step": 6285 + }, + { + "epoch": 0.5668936285340668, + "grad_norm": 1.705518947392063, + "learning_rate": 1.6653681022007696e-06, + "loss": 0.7585, + "step": 6286 + }, + { + "epoch": 0.5669838120575371, + "grad_norm": 2.3748876645475194, + "learning_rate": 1.6647921412822825e-06, + "loss": 0.678, + "step": 6287 + }, + { + "epoch": 0.5670739955810074, + "grad_norm": 1.9254057171803383, + "learning_rate": 1.6642162089655782e-06, + "loss": 0.749, + "step": 6288 + }, + { + "epoch": 0.5671641791044776, + "grad_norm": 2.365152119750556, + "learning_rate": 1.663640305299798e-06, + "loss": 0.7131, + "step": 6289 + }, + { + "epoch": 0.5672543626279479, + "grad_norm": 0.6597982033816389, + "learning_rate": 1.6630644303340824e-06, + "loss": 0.6105, + "step": 6290 + }, + { + "epoch": 0.5673445461514182, + "grad_norm": 2.8213102736687503, + "learning_rate": 1.662488584117567e-06, + "loss": 0.7562, + "step": 6291 + }, + { + "epoch": 0.5674347296748884, + "grad_norm": 1.5669378962330462, + "learning_rate": 1.6619127666993867e-06, + "loss": 0.8063, + "step": 6292 + }, + { + "epoch": 0.5675249131983586, + "grad_norm": 2.1387336338828313, + "learning_rate": 1.6613369781286727e-06, + "loss": 0.7504, + "step": 6293 + }, + { + "epoch": 0.567615096721829, + "grad_norm": 2.8204421755917872, + "learning_rate": 1.6607612184545562e-06, + "loss": 0.6838, + "step": 6294 + }, + { + "epoch": 0.5677052802452992, + "grad_norm": 2.129623880086807, + "learning_rate": 1.6601854877261617e-06, + "loss": 0.822, + "step": 6295 + }, + { + "epoch": 0.5677954637687694, + "grad_norm": 1.957405587522351, + "learning_rate": 1.6596097859926163e-06, + "loss": 0.7473, + "step": 6296 + }, + { + "epoch": 0.5678856472922397, + "grad_norm": 2.379243372689994, + "learning_rate": 1.6590341133030407e-06, + "loss": 0.8406, + "step": 6297 + }, + { + "epoch": 0.56797583081571, + "grad_norm": 2.663726337959671, + "learning_rate": 1.658458469706554e-06, + "loss": 0.8239, + "step": 6298 + }, + { + "epoch": 0.5680660143391802, + "grad_norm": 1.5623023861852503, + "learning_rate": 1.6578828552522746e-06, + "loss": 0.7017, + "step": 6299 + }, + { + "epoch": 0.5681561978626505, + "grad_norm": 1.8852200765669413, + "learning_rate": 1.6573072699893156e-06, + "loss": 0.795, + "step": 6300 + }, + { + "epoch": 0.5682463813861207, + "grad_norm": 1.9827938238275904, + "learning_rate": 1.6567317139667906e-06, + "loss": 0.722, + "step": 6301 + }, + { + "epoch": 0.5683365649095911, + "grad_norm": 1.8378846576849195, + "learning_rate": 1.6561561872338087e-06, + "loss": 0.7665, + "step": 6302 + }, + { + "epoch": 0.5684267484330613, + "grad_norm": 1.7719708398266731, + "learning_rate": 1.6555806898394764e-06, + "loss": 0.6612, + "step": 6303 + }, + { + "epoch": 0.5685169319565315, + "grad_norm": 1.7767061200631078, + "learning_rate": 1.6550052218328987e-06, + "loss": 0.7243, + "step": 6304 + }, + { + "epoch": 0.5686071154800018, + "grad_norm": 1.5797828803995475, + "learning_rate": 1.6544297832631777e-06, + "loss": 0.7223, + "step": 6305 + }, + { + "epoch": 0.5686972990034721, + "grad_norm": 2.013007660329601, + "learning_rate": 1.6538543741794135e-06, + "loss": 0.7514, + "step": 6306 + }, + { + "epoch": 0.5687874825269423, + "grad_norm": 2.269717919556677, + "learning_rate": 1.6532789946307028e-06, + "loss": 0.7175, + "step": 6307 + }, + { + "epoch": 0.5688776660504126, + "grad_norm": 0.6403318059867636, + "learning_rate": 1.6527036446661393e-06, + "loss": 0.6134, + "step": 6308 + }, + { + "epoch": 0.5689678495738828, + "grad_norm": 1.6527511058475348, + "learning_rate": 1.6521283243348165e-06, + "loss": 0.7796, + "step": 6309 + }, + { + "epoch": 0.5690580330973531, + "grad_norm": 1.9036171663546322, + "learning_rate": 1.6515530336858227e-06, + "loss": 0.7414, + "step": 6310 + }, + { + "epoch": 0.5691482166208234, + "grad_norm": 2.2712497737634574, + "learning_rate": 1.6509777727682457e-06, + "loss": 0.7993, + "step": 6311 + }, + { + "epoch": 0.5692384001442936, + "grad_norm": 1.7821307096441366, + "learning_rate": 1.65040254163117e-06, + "loss": 0.7897, + "step": 6312 + }, + { + "epoch": 0.569328583667764, + "grad_norm": 2.1532394843815235, + "learning_rate": 1.649827340323676e-06, + "loss": 0.6554, + "step": 6313 + }, + { + "epoch": 0.5694187671912342, + "grad_norm": 2.202849350587338, + "learning_rate": 1.6492521688948454e-06, + "loss": 0.7127, + "step": 6314 + }, + { + "epoch": 0.5695089507147044, + "grad_norm": 1.9380555125613113, + "learning_rate": 1.6486770273937526e-06, + "loss": 0.7698, + "step": 6315 + }, + { + "epoch": 0.5695991342381747, + "grad_norm": 1.4098460524636507, + "learning_rate": 1.6481019158694738e-06, + "loss": 0.7283, + "step": 6316 + }, + { + "epoch": 0.569689317761645, + "grad_norm": 0.754646056875564, + "learning_rate": 1.6475268343710792e-06, + "loss": 0.653, + "step": 6317 + }, + { + "epoch": 0.5697795012851152, + "grad_norm": 1.8186112341100553, + "learning_rate": 1.6469517829476396e-06, + "loss": 0.6789, + "step": 6318 + }, + { + "epoch": 0.5698696848085855, + "grad_norm": 2.1673216807790334, + "learning_rate": 1.64637676164822e-06, + "loss": 0.7336, + "step": 6319 + }, + { + "epoch": 0.5699598683320557, + "grad_norm": 1.4202485056477967, + "learning_rate": 1.6458017705218848e-06, + "loss": 0.7832, + "step": 6320 + }, + { + "epoch": 0.570050051855526, + "grad_norm": 1.6237933740850194, + "learning_rate": 1.645226809617696e-06, + "loss": 0.7753, + "step": 6321 + }, + { + "epoch": 0.5701402353789963, + "grad_norm": 2.316440112427712, + "learning_rate": 1.6446518789847112e-06, + "loss": 0.7815, + "step": 6322 + }, + { + "epoch": 0.5702304189024665, + "grad_norm": 2.049429195300558, + "learning_rate": 1.6440769786719883e-06, + "loss": 0.7735, + "step": 6323 + }, + { + "epoch": 0.5703206024259367, + "grad_norm": 1.4284822146624059, + "learning_rate": 1.6435021087285803e-06, + "loss": 0.7882, + "step": 6324 + }, + { + "epoch": 0.5704107859494071, + "grad_norm": 1.6980263922249934, + "learning_rate": 1.642927269203537e-06, + "loss": 0.7796, + "step": 6325 + }, + { + "epoch": 0.5705009694728773, + "grad_norm": 2.247738867637841, + "learning_rate": 1.642352460145909e-06, + "loss": 0.8022, + "step": 6326 + }, + { + "epoch": 0.5705911529963476, + "grad_norm": 1.7317366594274606, + "learning_rate": 1.6417776816047402e-06, + "loss": 0.7535, + "step": 6327 + }, + { + "epoch": 0.5706813365198178, + "grad_norm": 1.61677464914356, + "learning_rate": 1.6412029336290755e-06, + "loss": 0.7363, + "step": 6328 + }, + { + "epoch": 0.5707715200432881, + "grad_norm": 2.953739324381711, + "learning_rate": 1.6406282162679551e-06, + "loss": 0.7493, + "step": 6329 + }, + { + "epoch": 0.5708617035667584, + "grad_norm": 1.8516732355797296, + "learning_rate": 1.6400535295704162e-06, + "loss": 0.7232, + "step": 6330 + }, + { + "epoch": 0.5709518870902286, + "grad_norm": 1.5737503907413013, + "learning_rate": 1.6394788735854955e-06, + "loss": 0.6883, + "step": 6331 + }, + { + "epoch": 0.5710420706136988, + "grad_norm": 1.8476811542405243, + "learning_rate": 1.6389042483622246e-06, + "loss": 0.8155, + "step": 6332 + }, + { + "epoch": 0.5711322541371692, + "grad_norm": 1.5223876465681596, + "learning_rate": 1.638329653949635e-06, + "loss": 0.7439, + "step": 6333 + }, + { + "epoch": 0.5712224376606394, + "grad_norm": 1.4666369232860919, + "learning_rate": 1.637755090396753e-06, + "loss": 0.7155, + "step": 6334 + }, + { + "epoch": 0.5713126211841096, + "grad_norm": 0.6475832954427672, + "learning_rate": 1.6371805577526039e-06, + "loss": 0.6107, + "step": 6335 + }, + { + "epoch": 0.5714028047075799, + "grad_norm": 1.6437539191253703, + "learning_rate": 1.636606056066211e-06, + "loss": 0.6984, + "step": 6336 + }, + { + "epoch": 0.5714929882310502, + "grad_norm": 1.949129947328989, + "learning_rate": 1.636031585386592e-06, + "loss": 0.6818, + "step": 6337 + }, + { + "epoch": 0.5715831717545204, + "grad_norm": 1.428259464669914, + "learning_rate": 1.635457145762766e-06, + "loss": 0.8101, + "step": 6338 + }, + { + "epoch": 0.5716733552779907, + "grad_norm": 6.416277522717916, + "learning_rate": 1.6348827372437456e-06, + "loss": 0.7811, + "step": 6339 + }, + { + "epoch": 0.571763538801461, + "grad_norm": 1.5270841333960519, + "learning_rate": 1.634308359878544e-06, + "loss": 0.7546, + "step": 6340 + }, + { + "epoch": 0.5718537223249313, + "grad_norm": 1.5804826103202332, + "learning_rate": 1.6337340137161695e-06, + "loss": 0.8246, + "step": 6341 + }, + { + "epoch": 0.5719439058484015, + "grad_norm": 1.7057430897695318, + "learning_rate": 1.6331596988056277e-06, + "loss": 0.756, + "step": 6342 + }, + { + "epoch": 0.5720340893718717, + "grad_norm": 1.444361193205913, + "learning_rate": 1.632585415195924e-06, + "loss": 0.7767, + "step": 6343 + }, + { + "epoch": 0.5721242728953421, + "grad_norm": 1.72125969501056, + "learning_rate": 1.6320111629360583e-06, + "loss": 0.7801, + "step": 6344 + }, + { + "epoch": 0.5722144564188123, + "grad_norm": 1.9608883794582332, + "learning_rate": 1.631436942075029e-06, + "loss": 0.7323, + "step": 6345 + }, + { + "epoch": 0.5723046399422825, + "grad_norm": 1.9503555193842987, + "learning_rate": 1.630862752661833e-06, + "loss": 0.7269, + "step": 6346 + }, + { + "epoch": 0.5723948234657528, + "grad_norm": 2.1889114186159664, + "learning_rate": 1.6302885947454612e-06, + "loss": 0.6768, + "step": 6347 + }, + { + "epoch": 0.5724850069892231, + "grad_norm": 2.454424179979356, + "learning_rate": 1.6297144683749057e-06, + "loss": 0.7644, + "step": 6348 + }, + { + "epoch": 0.5725751905126933, + "grad_norm": 2.0982041309424018, + "learning_rate": 1.629140373599153e-06, + "loss": 0.7566, + "step": 6349 + }, + { + "epoch": 0.5726653740361636, + "grad_norm": 1.6850667293930304, + "learning_rate": 1.628566310467189e-06, + "loss": 0.7632, + "step": 6350 + }, + { + "epoch": 0.5727555575596338, + "grad_norm": 1.697155982932819, + "learning_rate": 1.6279922790279957e-06, + "loss": 0.7124, + "step": 6351 + }, + { + "epoch": 0.5728457410831042, + "grad_norm": 1.7921441868650754, + "learning_rate": 1.6274182793305512e-06, + "loss": 0.7695, + "step": 6352 + }, + { + "epoch": 0.5729359246065744, + "grad_norm": 2.2525568315828344, + "learning_rate": 1.626844311423835e-06, + "loss": 0.7835, + "step": 6353 + }, + { + "epoch": 0.5730261081300446, + "grad_norm": 2.0597755470955668, + "learning_rate": 1.6262703753568181e-06, + "loss": 0.83, + "step": 6354 + }, + { + "epoch": 0.5731162916535149, + "grad_norm": 1.9368019076918197, + "learning_rate": 1.6256964711784747e-06, + "loss": 0.8242, + "step": 6355 + }, + { + "epoch": 0.5732064751769852, + "grad_norm": 1.6031752491115916, + "learning_rate": 1.6251225989377723e-06, + "loss": 0.765, + "step": 6356 + }, + { + "epoch": 0.5732966587004554, + "grad_norm": 1.9796291881665595, + "learning_rate": 1.624548758683676e-06, + "loss": 0.7589, + "step": 6357 + }, + { + "epoch": 0.5733868422239257, + "grad_norm": 1.618190435955491, + "learning_rate": 1.6239749504651505e-06, + "loss": 0.8433, + "step": 6358 + }, + { + "epoch": 0.5734770257473959, + "grad_norm": 1.5330742848682168, + "learning_rate": 1.6234011743311552e-06, + "loss": 0.7316, + "step": 6359 + }, + { + "epoch": 0.5735672092708662, + "grad_norm": 1.772451656096567, + "learning_rate": 1.6228274303306483e-06, + "loss": 0.6837, + "step": 6360 + }, + { + "epoch": 0.5736573927943365, + "grad_norm": 1.5978187534352404, + "learning_rate": 1.6222537185125847e-06, + "loss": 0.8356, + "step": 6361 + }, + { + "epoch": 0.5737475763178067, + "grad_norm": 1.699440145867811, + "learning_rate": 1.6216800389259172e-06, + "loss": 0.7699, + "step": 6362 + }, + { + "epoch": 0.573837759841277, + "grad_norm": 1.4489954856133545, + "learning_rate": 1.6211063916195949e-06, + "loss": 0.7579, + "step": 6363 + }, + { + "epoch": 0.5739279433647473, + "grad_norm": 1.8517019484918689, + "learning_rate": 1.6205327766425633e-06, + "loss": 0.7647, + "step": 6364 + }, + { + "epoch": 0.5740181268882175, + "grad_norm": 1.4474820735789855, + "learning_rate": 1.6199591940437689e-06, + "loss": 0.7648, + "step": 6365 + }, + { + "epoch": 0.5741083104116878, + "grad_norm": 1.5686728044339282, + "learning_rate": 1.6193856438721505e-06, + "loss": 0.7118, + "step": 6366 + }, + { + "epoch": 0.5741984939351581, + "grad_norm": 1.7941908982349348, + "learning_rate": 1.6188121261766483e-06, + "loss": 0.7636, + "step": 6367 + }, + { + "epoch": 0.5742886774586283, + "grad_norm": 1.7346158934313818, + "learning_rate": 1.6182386410061976e-06, + "loss": 0.6882, + "step": 6368 + }, + { + "epoch": 0.5743788609820986, + "grad_norm": 9.921571645830102, + "learning_rate": 1.61766518840973e-06, + "loss": 0.663, + "step": 6369 + }, + { + "epoch": 0.5744690445055688, + "grad_norm": 1.7938840437193233, + "learning_rate": 1.6170917684361779e-06, + "loss": 0.6272, + "step": 6370 + }, + { + "epoch": 0.5745592280290391, + "grad_norm": 1.71039666811937, + "learning_rate": 1.6165183811344662e-06, + "loss": 0.8308, + "step": 6371 + }, + { + "epoch": 0.5746494115525094, + "grad_norm": 4.01729492613619, + "learning_rate": 1.6159450265535218e-06, + "loss": 0.7517, + "step": 6372 + }, + { + "epoch": 0.5747395950759796, + "grad_norm": 1.486918468761943, + "learning_rate": 1.6153717047422652e-06, + "loss": 0.8411, + "step": 6373 + }, + { + "epoch": 0.5748297785994498, + "grad_norm": 2.3385284242205504, + "learning_rate": 1.6147984157496155e-06, + "loss": 0.6445, + "step": 6374 + }, + { + "epoch": 0.5749199621229202, + "grad_norm": 1.4846766918176781, + "learning_rate": 1.6142251596244886e-06, + "loss": 0.6792, + "step": 6375 + }, + { + "epoch": 0.5750101456463904, + "grad_norm": 2.2625291390413045, + "learning_rate": 1.6136519364157983e-06, + "loss": 0.8115, + "step": 6376 + }, + { + "epoch": 0.5751003291698606, + "grad_norm": 1.5885217217199066, + "learning_rate": 1.6130787461724555e-06, + "loss": 0.6939, + "step": 6377 + }, + { + "epoch": 0.5751905126933309, + "grad_norm": 1.5075949081986721, + "learning_rate": 1.6125055889433679e-06, + "loss": 0.8289, + "step": 6378 + }, + { + "epoch": 0.5752806962168012, + "grad_norm": 3.1028018656400596, + "learning_rate": 1.6119324647774386e-06, + "loss": 0.7772, + "step": 6379 + }, + { + "epoch": 0.5753708797402715, + "grad_norm": 2.1281480471445557, + "learning_rate": 1.6113593737235724e-06, + "loss": 0.7665, + "step": 6380 + }, + { + "epoch": 0.5754610632637417, + "grad_norm": 1.579580938440894, + "learning_rate": 1.6107863158306665e-06, + "loss": 0.7056, + "step": 6381 + }, + { + "epoch": 0.5755512467872119, + "grad_norm": 2.096867759453315, + "learning_rate": 1.610213291147619e-06, + "loss": 0.6492, + "step": 6382 + }, + { + "epoch": 0.5756414303106823, + "grad_norm": 1.4177953861070556, + "learning_rate": 1.609640299723322e-06, + "loss": 0.7753, + "step": 6383 + }, + { + "epoch": 0.5757316138341525, + "grad_norm": 1.5043865279170836, + "learning_rate": 1.609067341606668e-06, + "loss": 0.7438, + "step": 6384 + }, + { + "epoch": 0.5758217973576227, + "grad_norm": 1.8269085871936543, + "learning_rate": 1.6084944168465438e-06, + "loss": 0.7459, + "step": 6385 + }, + { + "epoch": 0.5759119808810931, + "grad_norm": 1.8809350494657804, + "learning_rate": 1.6079215254918339e-06, + "loss": 0.7767, + "step": 6386 + }, + { + "epoch": 0.5760021644045633, + "grad_norm": 1.931749115183556, + "learning_rate": 1.6073486675914222e-06, + "loss": 0.6621, + "step": 6387 + }, + { + "epoch": 0.5760923479280335, + "grad_norm": 4.274691047435521, + "learning_rate": 1.606775843194187e-06, + "loss": 0.7847, + "step": 6388 + }, + { + "epoch": 0.5761825314515038, + "grad_norm": 2.0407073303719088, + "learning_rate": 1.6062030523490053e-06, + "loss": 0.7412, + "step": 6389 + }, + { + "epoch": 0.5762727149749741, + "grad_norm": 1.6532365267366256, + "learning_rate": 1.60563029510475e-06, + "loss": 0.6818, + "step": 6390 + }, + { + "epoch": 0.5763628984984444, + "grad_norm": 0.6572564899041455, + "learning_rate": 1.6050575715102927e-06, + "loss": 0.5695, + "step": 6391 + }, + { + "epoch": 0.5764530820219146, + "grad_norm": 2.3267290679557355, + "learning_rate": 1.6044848816145014e-06, + "loss": 0.808, + "step": 6392 + }, + { + "epoch": 0.5765432655453848, + "grad_norm": 2.126493490137845, + "learning_rate": 1.60391222546624e-06, + "loss": 0.7811, + "step": 6393 + }, + { + "epoch": 0.5766334490688552, + "grad_norm": 1.7816742508919816, + "learning_rate": 1.6033396031143725e-06, + "loss": 0.7508, + "step": 6394 + }, + { + "epoch": 0.5767236325923254, + "grad_norm": 2.0165851287541647, + "learning_rate": 1.602767014607757e-06, + "loss": 0.7563, + "step": 6395 + }, + { + "epoch": 0.5768138161157956, + "grad_norm": 3.8659610537458327, + "learning_rate": 1.6021944599952493e-06, + "loss": 0.7645, + "step": 6396 + }, + { + "epoch": 0.5769039996392659, + "grad_norm": 2.2711691535161043, + "learning_rate": 1.6016219393257048e-06, + "loss": 0.7694, + "step": 6397 + }, + { + "epoch": 0.5769941831627362, + "grad_norm": 1.9601294079702978, + "learning_rate": 1.6010494526479722e-06, + "loss": 0.7079, + "step": 6398 + }, + { + "epoch": 0.5770843666862064, + "grad_norm": 2.205407657783866, + "learning_rate": 1.6004770000109006e-06, + "loss": 0.7013, + "step": 6399 + }, + { + "epoch": 0.5771745502096767, + "grad_norm": 1.5991658211779218, + "learning_rate": 1.5999045814633348e-06, + "loss": 0.7229, + "step": 6400 + }, + { + "epoch": 0.5772647337331469, + "grad_norm": 1.4319136334539402, + "learning_rate": 1.5993321970541151e-06, + "loss": 0.7652, + "step": 6401 + }, + { + "epoch": 0.5773549172566173, + "grad_norm": 3.3304416670605224, + "learning_rate": 1.5987598468320825e-06, + "loss": 0.7548, + "step": 6402 + }, + { + "epoch": 0.5774451007800875, + "grad_norm": 1.5162963953176731, + "learning_rate": 1.5981875308460717e-06, + "loss": 0.7438, + "step": 6403 + }, + { + "epoch": 0.5775352843035577, + "grad_norm": 2.7553861064322764, + "learning_rate": 1.5976152491449169e-06, + "loss": 0.742, + "step": 6404 + }, + { + "epoch": 0.577625467827028, + "grad_norm": 1.814452796500744, + "learning_rate": 1.5970430017774468e-06, + "loss": 0.818, + "step": 6405 + }, + { + "epoch": 0.5777156513504983, + "grad_norm": 8.108271823295592, + "learning_rate": 1.5964707887924904e-06, + "loss": 0.7675, + "step": 6406 + }, + { + "epoch": 0.5778058348739685, + "grad_norm": 1.3304807078909973, + "learning_rate": 1.5958986102388714e-06, + "loss": 0.6418, + "step": 6407 + }, + { + "epoch": 0.5778960183974388, + "grad_norm": 1.6746344921950405, + "learning_rate": 1.5953264661654104e-06, + "loss": 0.7093, + "step": 6408 + }, + { + "epoch": 0.5779862019209091, + "grad_norm": 2.016053146903051, + "learning_rate": 1.5947543566209276e-06, + "loss": 0.786, + "step": 6409 + }, + { + "epoch": 0.5780763854443793, + "grad_norm": 2.6221692398702974, + "learning_rate": 1.5941822816542367e-06, + "loss": 0.7356, + "step": 6410 + }, + { + "epoch": 0.5781665689678496, + "grad_norm": 2.0790595914475465, + "learning_rate": 1.5936102413141519e-06, + "loss": 0.715, + "step": 6411 + }, + { + "epoch": 0.5782567524913198, + "grad_norm": 1.892251127198607, + "learning_rate": 1.5930382356494823e-06, + "loss": 0.671, + "step": 6412 + }, + { + "epoch": 0.5783469360147901, + "grad_norm": 1.5802102994393188, + "learning_rate": 1.5924662647090335e-06, + "loss": 0.801, + "step": 6413 + }, + { + "epoch": 0.5784371195382604, + "grad_norm": 2.4616842438783326, + "learning_rate": 1.5918943285416108e-06, + "loss": 0.7727, + "step": 6414 + }, + { + "epoch": 0.5785273030617306, + "grad_norm": 1.4619636399826597, + "learning_rate": 1.5913224271960139e-06, + "loss": 0.6922, + "step": 6415 + }, + { + "epoch": 0.5786174865852008, + "grad_norm": 1.6340350816899125, + "learning_rate": 1.590750560721041e-06, + "loss": 0.7689, + "step": 6416 + }, + { + "epoch": 0.5787076701086712, + "grad_norm": 1.8412692304281602, + "learning_rate": 1.5901787291654874e-06, + "loss": 0.7217, + "step": 6417 + }, + { + "epoch": 0.5787978536321414, + "grad_norm": 2.5087018203860736, + "learning_rate": 1.5896069325781435e-06, + "loss": 0.6262, + "step": 6418 + }, + { + "epoch": 0.5788880371556117, + "grad_norm": 1.8475587589278561, + "learning_rate": 1.5890351710077998e-06, + "loss": 0.7954, + "step": 6419 + }, + { + "epoch": 0.5789782206790819, + "grad_norm": 1.6100198879553682, + "learning_rate": 1.5884634445032406e-06, + "loss": 0.8089, + "step": 6420 + }, + { + "epoch": 0.5790684042025522, + "grad_norm": 0.6011001990348906, + "learning_rate": 1.5878917531132501e-06, + "loss": 0.5981, + "step": 6421 + }, + { + "epoch": 0.5791585877260225, + "grad_norm": 1.807817791112544, + "learning_rate": 1.5873200968866077e-06, + "loss": 0.6319, + "step": 6422 + }, + { + "epoch": 0.5792487712494927, + "grad_norm": 2.2973694561055433, + "learning_rate": 1.586748475872089e-06, + "loss": 0.7312, + "step": 6423 + }, + { + "epoch": 0.5793389547729629, + "grad_norm": 1.861707938831617, + "learning_rate": 1.58617689011847e-06, + "loss": 0.6201, + "step": 6424 + }, + { + "epoch": 0.5794291382964333, + "grad_norm": 1.7471239434630907, + "learning_rate": 1.5856053396745198e-06, + "loss": 0.7885, + "step": 6425 + }, + { + "epoch": 0.5795193218199035, + "grad_norm": 2.4296061511389246, + "learning_rate": 1.5850338245890078e-06, + "loss": 0.8162, + "step": 6426 + }, + { + "epoch": 0.5796095053433737, + "grad_norm": 1.8205684105419875, + "learning_rate": 1.5844623449106974e-06, + "loss": 0.7393, + "step": 6427 + }, + { + "epoch": 0.579699688866844, + "grad_norm": 2.047917340868053, + "learning_rate": 1.583890900688351e-06, + "loss": 0.7924, + "step": 6428 + }, + { + "epoch": 0.5797898723903143, + "grad_norm": 2.2620830341347578, + "learning_rate": 1.583319491970728e-06, + "loss": 0.8281, + "step": 6429 + }, + { + "epoch": 0.5798800559137846, + "grad_norm": 1.4358234580230442, + "learning_rate": 1.5827481188065828e-06, + "loss": 0.7657, + "step": 6430 + }, + { + "epoch": 0.5799702394372548, + "grad_norm": 2.0438599256531544, + "learning_rate": 1.5821767812446689e-06, + "loss": 0.7263, + "step": 6431 + }, + { + "epoch": 0.5800604229607251, + "grad_norm": 1.7069830636702235, + "learning_rate": 1.581605479333736e-06, + "loss": 0.8658, + "step": 6432 + }, + { + "epoch": 0.5801506064841954, + "grad_norm": 2.134003359161851, + "learning_rate": 1.5810342131225308e-06, + "loss": 0.7608, + "step": 6433 + }, + { + "epoch": 0.5802407900076656, + "grad_norm": 2.940693681585768, + "learning_rate": 1.580462982659797e-06, + "loss": 0.7997, + "step": 6434 + }, + { + "epoch": 0.5803309735311358, + "grad_norm": 2.9019622346614655, + "learning_rate": 1.5798917879942736e-06, + "loss": 0.8304, + "step": 6435 + }, + { + "epoch": 0.5804211570546062, + "grad_norm": 1.4009202623296015, + "learning_rate": 1.5793206291747006e-06, + "loss": 0.6963, + "step": 6436 + }, + { + "epoch": 0.5805113405780764, + "grad_norm": 1.3859399462208453, + "learning_rate": 1.57874950624981e-06, + "loss": 0.6882, + "step": 6437 + }, + { + "epoch": 0.5806015241015466, + "grad_norm": 2.0142805953574334, + "learning_rate": 1.5781784192683351e-06, + "loss": 0.7535, + "step": 6438 + }, + { + "epoch": 0.5806917076250169, + "grad_norm": 1.3505560153782732, + "learning_rate": 1.5776073682790033e-06, + "loss": 0.7131, + "step": 6439 + }, + { + "epoch": 0.5807818911484872, + "grad_norm": 2.450896982531484, + "learning_rate": 1.5770363533305393e-06, + "loss": 0.7742, + "step": 6440 + }, + { + "epoch": 0.5808720746719575, + "grad_norm": 2.005336080337902, + "learning_rate": 1.5764653744716665e-06, + "loss": 0.7594, + "step": 6441 + }, + { + "epoch": 0.5809622581954277, + "grad_norm": 1.7064550199753215, + "learning_rate": 1.575894431751103e-06, + "loss": 0.7835, + "step": 6442 + }, + { + "epoch": 0.5810524417188979, + "grad_norm": 2.973898829696057, + "learning_rate": 1.575323525217565e-06, + "loss": 0.7817, + "step": 6443 + }, + { + "epoch": 0.5811426252423683, + "grad_norm": 1.6231536646483227, + "learning_rate": 1.574752654919766e-06, + "loss": 0.8368, + "step": 6444 + }, + { + "epoch": 0.5812328087658385, + "grad_norm": 1.5957643830997197, + "learning_rate": 1.5741818209064146e-06, + "loss": 0.7218, + "step": 6445 + }, + { + "epoch": 0.5813229922893087, + "grad_norm": 1.8334830902085486, + "learning_rate": 1.5736110232262183e-06, + "loss": 0.761, + "step": 6446 + }, + { + "epoch": 0.581413175812779, + "grad_norm": 1.7046029913468181, + "learning_rate": 1.5730402619278804e-06, + "loss": 0.8011, + "step": 6447 + }, + { + "epoch": 0.5815033593362493, + "grad_norm": 0.6598917865905084, + "learning_rate": 1.5724695370601024e-06, + "loss": 0.5564, + "step": 6448 + }, + { + "epoch": 0.5815935428597195, + "grad_norm": 1.670333650917191, + "learning_rate": 1.5718988486715798e-06, + "loss": 0.6776, + "step": 6449 + }, + { + "epoch": 0.5816837263831898, + "grad_norm": 1.622309952407461, + "learning_rate": 1.5713281968110087e-06, + "loss": 0.7374, + "step": 6450 + }, + { + "epoch": 0.58177390990666, + "grad_norm": 1.6398034354784221, + "learning_rate": 1.5707575815270796e-06, + "loss": 0.7364, + "step": 6451 + }, + { + "epoch": 0.5818640934301303, + "grad_norm": 2.0648190725282958, + "learning_rate": 1.57018700286848e-06, + "loss": 0.7587, + "step": 6452 + }, + { + "epoch": 0.5819542769536006, + "grad_norm": 2.4742622162084773, + "learning_rate": 1.5696164608838956e-06, + "loss": 0.7701, + "step": 6453 + }, + { + "epoch": 0.5820444604770708, + "grad_norm": 2.1255605743613946, + "learning_rate": 1.5690459556220073e-06, + "loss": 0.7743, + "step": 6454 + }, + { + "epoch": 0.582134644000541, + "grad_norm": 1.847662121237203, + "learning_rate": 1.5684754871314949e-06, + "loss": 0.8122, + "step": 6455 + }, + { + "epoch": 0.5822248275240114, + "grad_norm": 1.6313138956585955, + "learning_rate": 1.5679050554610335e-06, + "loss": 0.7548, + "step": 6456 + }, + { + "epoch": 0.5823150110474816, + "grad_norm": 1.4166662365781841, + "learning_rate": 1.567334660659295e-06, + "loss": 0.7577, + "step": 6457 + }, + { + "epoch": 0.5824051945709519, + "grad_norm": 1.622168641896608, + "learning_rate": 1.5667643027749488e-06, + "loss": 0.7701, + "step": 6458 + }, + { + "epoch": 0.5824953780944222, + "grad_norm": 1.9004545446076828, + "learning_rate": 1.5661939818566614e-06, + "loss": 0.829, + "step": 6459 + }, + { + "epoch": 0.5825855616178924, + "grad_norm": 3.49921531055347, + "learning_rate": 1.5656236979530956e-06, + "loss": 0.6199, + "step": 6460 + }, + { + "epoch": 0.5826757451413627, + "grad_norm": 1.5585555834387463, + "learning_rate": 1.5650534511129106e-06, + "loss": 0.6966, + "step": 6461 + }, + { + "epoch": 0.5827659286648329, + "grad_norm": 1.8945095414953108, + "learning_rate": 1.5644832413847635e-06, + "loss": 0.7394, + "step": 6462 + }, + { + "epoch": 0.5828561121883032, + "grad_norm": 1.314496383975436, + "learning_rate": 1.5639130688173082e-06, + "loss": 0.7969, + "step": 6463 + }, + { + "epoch": 0.5829462957117735, + "grad_norm": 3.0640010658341987, + "learning_rate": 1.5633429334591932e-06, + "loss": 0.7286, + "step": 6464 + }, + { + "epoch": 0.5830364792352437, + "grad_norm": 1.8715564259271151, + "learning_rate": 1.562772835359068e-06, + "loss": 0.7845, + "step": 6465 + }, + { + "epoch": 0.5831266627587139, + "grad_norm": 2.0342790759745215, + "learning_rate": 1.5622027745655753e-06, + "loss": 0.7209, + "step": 6466 + }, + { + "epoch": 0.5832168462821843, + "grad_norm": 1.6533215382498778, + "learning_rate": 1.561632751127355e-06, + "loss": 0.7099, + "step": 6467 + }, + { + "epoch": 0.5833070298056545, + "grad_norm": 1.8188089210430736, + "learning_rate": 1.561062765093046e-06, + "loss": 0.7967, + "step": 6468 + }, + { + "epoch": 0.5833972133291248, + "grad_norm": 1.9064027068461076, + "learning_rate": 1.5604928165112817e-06, + "loss": 0.749, + "step": 6469 + }, + { + "epoch": 0.583487396852595, + "grad_norm": 1.8098166103085165, + "learning_rate": 1.5599229054306945e-06, + "loss": 0.7467, + "step": 6470 + }, + { + "epoch": 0.5835775803760653, + "grad_norm": 1.6815330125356784, + "learning_rate": 1.5593530318999111e-06, + "loss": 0.7474, + "step": 6471 + }, + { + "epoch": 0.5836677638995356, + "grad_norm": 1.7837605769719973, + "learning_rate": 1.5587831959675572e-06, + "loss": 0.784, + "step": 6472 + }, + { + "epoch": 0.5837579474230058, + "grad_norm": 1.5113952448284624, + "learning_rate": 1.5582133976822534e-06, + "loss": 0.7403, + "step": 6473 + }, + { + "epoch": 0.583848130946476, + "grad_norm": 1.5956461229730454, + "learning_rate": 1.5576436370926185e-06, + "loss": 0.7073, + "step": 6474 + }, + { + "epoch": 0.5839383144699464, + "grad_norm": 1.5682335752055487, + "learning_rate": 1.5570739142472679e-06, + "loss": 0.7353, + "step": 6475 + }, + { + "epoch": 0.5840284979934166, + "grad_norm": 3.4266207459139864, + "learning_rate": 1.5565042291948127e-06, + "loss": 0.8165, + "step": 6476 + }, + { + "epoch": 0.5841186815168868, + "grad_norm": 1.636524780724513, + "learning_rate": 1.5559345819838624e-06, + "loss": 0.7917, + "step": 6477 + }, + { + "epoch": 0.5842088650403571, + "grad_norm": 1.330068579820169, + "learning_rate": 1.5553649726630226e-06, + "loss": 0.7312, + "step": 6478 + }, + { + "epoch": 0.5842990485638274, + "grad_norm": 1.5291733965955165, + "learning_rate": 1.5547954012808942e-06, + "loss": 0.6837, + "step": 6479 + }, + { + "epoch": 0.5843892320872976, + "grad_norm": 2.4342782663657387, + "learning_rate": 1.5542258678860776e-06, + "loss": 0.7311, + "step": 6480 + }, + { + "epoch": 0.5844794156107679, + "grad_norm": 1.7355368219600034, + "learning_rate": 1.553656372527167e-06, + "loss": 0.8708, + "step": 6481 + }, + { + "epoch": 0.5845695991342382, + "grad_norm": 2.6877632012229777, + "learning_rate": 1.5530869152527568e-06, + "loss": 0.8562, + "step": 6482 + }, + { + "epoch": 0.5846597826577085, + "grad_norm": 1.5499242917921752, + "learning_rate": 1.5525174961114353e-06, + "loss": 0.7728, + "step": 6483 + }, + { + "epoch": 0.5847499661811787, + "grad_norm": 1.720170283870604, + "learning_rate": 1.5519481151517875e-06, + "loss": 0.7915, + "step": 6484 + }, + { + "epoch": 0.5848401497046489, + "grad_norm": 2.2162508792090136, + "learning_rate": 1.551378772422398e-06, + "loss": 0.7612, + "step": 6485 + }, + { + "epoch": 0.5849303332281193, + "grad_norm": 1.4621960120249908, + "learning_rate": 1.5508094679718447e-06, + "loss": 0.7646, + "step": 6486 + }, + { + "epoch": 0.5850205167515895, + "grad_norm": 1.761205289664416, + "learning_rate": 1.5502402018487048e-06, + "loss": 0.6583, + "step": 6487 + }, + { + "epoch": 0.5851107002750597, + "grad_norm": 2.1653259114849197, + "learning_rate": 1.54967097410155e-06, + "loss": 0.759, + "step": 6488 + }, + { + "epoch": 0.58520088379853, + "grad_norm": 1.5498299936152666, + "learning_rate": 1.5491017847789519e-06, + "loss": 0.6775, + "step": 6489 + }, + { + "epoch": 0.5852910673220003, + "grad_norm": 1.7496986129634025, + "learning_rate": 1.5485326339294755e-06, + "loss": 0.6989, + "step": 6490 + }, + { + "epoch": 0.5853812508454705, + "grad_norm": 1.5938830600618796, + "learning_rate": 1.5479635216016832e-06, + "loss": 0.7245, + "step": 6491 + }, + { + "epoch": 0.5854714343689408, + "grad_norm": 2.1300819440311103, + "learning_rate": 1.547394447844137e-06, + "loss": 0.678, + "step": 6492 + }, + { + "epoch": 0.585561617892411, + "grad_norm": 1.6130921481606815, + "learning_rate": 1.546825412705391e-06, + "loss": 0.6713, + "step": 6493 + }, + { + "epoch": 0.5856518014158814, + "grad_norm": 2.9275915354060498, + "learning_rate": 1.5462564162340007e-06, + "loss": 0.78, + "step": 6494 + }, + { + "epoch": 0.5857419849393516, + "grad_norm": 1.9512448225618595, + "learning_rate": 1.5456874584785144e-06, + "loss": 0.7625, + "step": 6495 + }, + { + "epoch": 0.5858321684628218, + "grad_norm": 1.81897139365177, + "learning_rate": 1.5451185394874785e-06, + "loss": 0.7301, + "step": 6496 + }, + { + "epoch": 0.5859223519862921, + "grad_norm": 1.728083359283895, + "learning_rate": 1.5445496593094381e-06, + "loss": 0.7646, + "step": 6497 + }, + { + "epoch": 0.5860125355097624, + "grad_norm": 1.8607492577498654, + "learning_rate": 1.5439808179929316e-06, + "loss": 0.7694, + "step": 6498 + }, + { + "epoch": 0.5861027190332326, + "grad_norm": 1.6411615856612791, + "learning_rate": 1.543412015586496e-06, + "loss": 0.7246, + "step": 6499 + }, + { + "epoch": 0.5861929025567029, + "grad_norm": 1.766819845012738, + "learning_rate": 1.5428432521386655e-06, + "loss": 0.7208, + "step": 6500 + }, + { + "epoch": 0.5862830860801731, + "grad_norm": 1.6949161470540168, + "learning_rate": 1.5422745276979688e-06, + "loss": 0.7868, + "step": 6501 + }, + { + "epoch": 0.5863732696036434, + "grad_norm": 1.509071733482426, + "learning_rate": 1.5417058423129336e-06, + "loss": 0.7442, + "step": 6502 + }, + { + "epoch": 0.5864634531271137, + "grad_norm": 1.3741373910941244, + "learning_rate": 1.5411371960320822e-06, + "loss": 0.7862, + "step": 6503 + }, + { + "epoch": 0.5865536366505839, + "grad_norm": 1.4557636054463021, + "learning_rate": 1.5405685889039363e-06, + "loss": 0.7851, + "step": 6504 + }, + { + "epoch": 0.5866438201740543, + "grad_norm": 2.0387095880391577, + "learning_rate": 1.5400000209770118e-06, + "loss": 0.7525, + "step": 6505 + }, + { + "epoch": 0.5867340036975245, + "grad_norm": 1.9248573919209673, + "learning_rate": 1.5394314922998208e-06, + "loss": 0.7681, + "step": 6506 + }, + { + "epoch": 0.5868241872209947, + "grad_norm": 2.3286306356982007, + "learning_rate": 1.5388630029208756e-06, + "loss": 0.8031, + "step": 6507 + }, + { + "epoch": 0.586914370744465, + "grad_norm": 2.183064896580467, + "learning_rate": 1.5382945528886806e-06, + "loss": 0.7762, + "step": 6508 + }, + { + "epoch": 0.5870045542679353, + "grad_norm": 2.4701857435989787, + "learning_rate": 1.5377261422517412e-06, + "loss": 0.745, + "step": 6509 + }, + { + "epoch": 0.5870947377914055, + "grad_norm": 2.1946187399470567, + "learning_rate": 1.5371577710585553e-06, + "loss": 0.7128, + "step": 6510 + }, + { + "epoch": 0.5871849213148758, + "grad_norm": 0.7276551838826665, + "learning_rate": 1.536589439357621e-06, + "loss": 0.5653, + "step": 6511 + }, + { + "epoch": 0.587275104838346, + "grad_norm": 1.6034194229763412, + "learning_rate": 1.5360211471974315e-06, + "loss": 0.7576, + "step": 6512 + }, + { + "epoch": 0.5873652883618163, + "grad_norm": 1.6234040125604705, + "learning_rate": 1.5354528946264753e-06, + "loss": 0.8094, + "step": 6513 + }, + { + "epoch": 0.5874554718852866, + "grad_norm": 2.302795668183414, + "learning_rate": 1.5348846816932399e-06, + "loss": 0.6901, + "step": 6514 + }, + { + "epoch": 0.5875456554087568, + "grad_norm": 0.6535260570762879, + "learning_rate": 1.5343165084462077e-06, + "loss": 0.5756, + "step": 6515 + }, + { + "epoch": 0.587635838932227, + "grad_norm": 1.8038293028256027, + "learning_rate": 1.5337483749338595e-06, + "loss": 0.7073, + "step": 6516 + }, + { + "epoch": 0.5877260224556974, + "grad_norm": 1.526095531498125, + "learning_rate": 1.5331802812046708e-06, + "loss": 0.7259, + "step": 6517 + }, + { + "epoch": 0.5878162059791676, + "grad_norm": 15.006699655350554, + "learning_rate": 1.5326122273071133e-06, + "loss": 0.6056, + "step": 6518 + }, + { + "epoch": 0.5879063895026378, + "grad_norm": 1.909634946119652, + "learning_rate": 1.532044213289659e-06, + "loss": 0.7904, + "step": 6519 + }, + { + "epoch": 0.5879965730261081, + "grad_norm": 2.0201849405212116, + "learning_rate": 1.5314762392007718e-06, + "loss": 0.6411, + "step": 6520 + }, + { + "epoch": 0.5880867565495784, + "grad_norm": 1.4893015970766477, + "learning_rate": 1.530908305088916e-06, + "loss": 0.8058, + "step": 6521 + }, + { + "epoch": 0.5881769400730487, + "grad_norm": 1.8084193365972285, + "learning_rate": 1.5303404110025501e-06, + "loss": 0.7504, + "step": 6522 + }, + { + "epoch": 0.5882671235965189, + "grad_norm": 9.718481127586044, + "learning_rate": 1.5297725569901293e-06, + "loss": 0.7105, + "step": 6523 + }, + { + "epoch": 0.5883573071199891, + "grad_norm": 2.967204123114475, + "learning_rate": 1.5292047431001077e-06, + "loss": 0.6499, + "step": 6524 + }, + { + "epoch": 0.5884474906434595, + "grad_norm": 0.6857281958603906, + "learning_rate": 1.5286369693809321e-06, + "loss": 0.5887, + "step": 6525 + }, + { + "epoch": 0.5885376741669297, + "grad_norm": 2.1728198514188817, + "learning_rate": 1.5280692358810506e-06, + "loss": 0.762, + "step": 6526 + }, + { + "epoch": 0.5886278576903999, + "grad_norm": 2.1826538902115393, + "learning_rate": 1.527501542648904e-06, + "loss": 0.7, + "step": 6527 + }, + { + "epoch": 0.5887180412138703, + "grad_norm": 1.970770026622399, + "learning_rate": 1.5269338897329308e-06, + "loss": 0.6599, + "step": 6528 + }, + { + "epoch": 0.5888082247373405, + "grad_norm": 1.5555496509000943, + "learning_rate": 1.5263662771815662e-06, + "loss": 0.781, + "step": 6529 + }, + { + "epoch": 0.5888984082608107, + "grad_norm": 1.81933305446296, + "learning_rate": 1.5257987050432429e-06, + "loss": 0.8273, + "step": 6530 + }, + { + "epoch": 0.588988591784281, + "grad_norm": 1.8511576793453977, + "learning_rate": 1.5252311733663887e-06, + "loss": 0.6626, + "step": 6531 + }, + { + "epoch": 0.5890787753077513, + "grad_norm": 2.0226337500860865, + "learning_rate": 1.5246636821994281e-06, + "loss": 0.6961, + "step": 6532 + }, + { + "epoch": 0.5891689588312216, + "grad_norm": 1.5098272905673413, + "learning_rate": 1.524096231590784e-06, + "loss": 0.7994, + "step": 6533 + }, + { + "epoch": 0.5892591423546918, + "grad_norm": 1.8458711093570452, + "learning_rate": 1.5235288215888736e-06, + "loss": 0.7965, + "step": 6534 + }, + { + "epoch": 0.589349325878162, + "grad_norm": 1.700309876641674, + "learning_rate": 1.5229614522421102e-06, + "loss": 0.7567, + "step": 6535 + }, + { + "epoch": 0.5894395094016324, + "grad_norm": 4.0188326956159255, + "learning_rate": 1.5223941235989071e-06, + "loss": 0.5904, + "step": 6536 + }, + { + "epoch": 0.5895296929251026, + "grad_norm": 1.6233683978033366, + "learning_rate": 1.52182683570767e-06, + "loss": 0.7235, + "step": 6537 + }, + { + "epoch": 0.5896198764485728, + "grad_norm": 1.7092669967201322, + "learning_rate": 1.5212595886168046e-06, + "loss": 0.7786, + "step": 6538 + }, + { + "epoch": 0.5897100599720431, + "grad_norm": 1.727623591880702, + "learning_rate": 1.520692382374711e-06, + "loss": 0.7666, + "step": 6539 + }, + { + "epoch": 0.5898002434955134, + "grad_norm": 2.8542543012618853, + "learning_rate": 1.5201252170297854e-06, + "loss": 0.778, + "step": 6540 + }, + { + "epoch": 0.5898904270189836, + "grad_norm": 1.8151537442836523, + "learning_rate": 1.5195580926304232e-06, + "loss": 0.6388, + "step": 6541 + }, + { + "epoch": 0.5899806105424539, + "grad_norm": 2.8947636849468408, + "learning_rate": 1.5189910092250131e-06, + "loss": 0.7729, + "step": 6542 + }, + { + "epoch": 0.5900707940659241, + "grad_norm": 2.634892488200883, + "learning_rate": 1.5184239668619427e-06, + "loss": 0.7618, + "step": 6543 + }, + { + "epoch": 0.5901609775893945, + "grad_norm": 1.6976721027755861, + "learning_rate": 1.5178569655895946e-06, + "loss": 0.755, + "step": 6544 + }, + { + "epoch": 0.5902511611128647, + "grad_norm": 1.5275794430316707, + "learning_rate": 1.5172900054563487e-06, + "loss": 0.7826, + "step": 6545 + }, + { + "epoch": 0.5903413446363349, + "grad_norm": 1.59162649167206, + "learning_rate": 1.5167230865105814e-06, + "loss": 0.77, + "step": 6546 + }, + { + "epoch": 0.5904315281598052, + "grad_norm": 1.866405243392296, + "learning_rate": 1.5161562088006644e-06, + "loss": 0.6706, + "step": 6547 + }, + { + "epoch": 0.5905217116832755, + "grad_norm": 2.1241360198252357, + "learning_rate": 1.5155893723749685e-06, + "loss": 0.7834, + "step": 6548 + }, + { + "epoch": 0.5906118952067457, + "grad_norm": 3.169597691419474, + "learning_rate": 1.5150225772818582e-06, + "loss": 0.7611, + "step": 6549 + }, + { + "epoch": 0.590702078730216, + "grad_norm": 1.7217247622210388, + "learning_rate": 1.5144558235696949e-06, + "loss": 0.8024, + "step": 6550 + }, + { + "epoch": 0.5907922622536863, + "grad_norm": 1.6117898545364924, + "learning_rate": 1.5138891112868388e-06, + "loss": 0.6697, + "step": 6551 + }, + { + "epoch": 0.5908824457771565, + "grad_norm": 0.7005524728618798, + "learning_rate": 1.5133224404816433e-06, + "loss": 0.5796, + "step": 6552 + }, + { + "epoch": 0.5909726293006268, + "grad_norm": 3.023214800220369, + "learning_rate": 1.5127558112024617e-06, + "loss": 0.6768, + "step": 6553 + }, + { + "epoch": 0.591062812824097, + "grad_norm": 0.6812261944732447, + "learning_rate": 1.5121892234976404e-06, + "loss": 0.5929, + "step": 6554 + }, + { + "epoch": 0.5911529963475673, + "grad_norm": 1.6525250418838877, + "learning_rate": 1.5116226774155243e-06, + "loss": 0.7356, + "step": 6555 + }, + { + "epoch": 0.5912431798710376, + "grad_norm": 1.8455953899608724, + "learning_rate": 1.5110561730044547e-06, + "loss": 0.7263, + "step": 6556 + }, + { + "epoch": 0.5913333633945078, + "grad_norm": 1.8143648222592812, + "learning_rate": 1.510489710312768e-06, + "loss": 0.7532, + "step": 6557 + }, + { + "epoch": 0.591423546917978, + "grad_norm": 2.4333880048210226, + "learning_rate": 1.5099232893887987e-06, + "loss": 0.7443, + "step": 6558 + }, + { + "epoch": 0.5915137304414484, + "grad_norm": 1.552384477454498, + "learning_rate": 1.5093569102808758e-06, + "loss": 0.7237, + "step": 6559 + }, + { + "epoch": 0.5916039139649186, + "grad_norm": 1.8562554227704815, + "learning_rate": 1.5087905730373275e-06, + "loss": 0.7602, + "step": 6560 + }, + { + "epoch": 0.5916940974883889, + "grad_norm": 0.616512653948017, + "learning_rate": 1.508224277706476e-06, + "loss": 0.5905, + "step": 6561 + }, + { + "epoch": 0.5917842810118591, + "grad_norm": 4.507771880232933, + "learning_rate": 1.5076580243366399e-06, + "loss": 0.7056, + "step": 6562 + }, + { + "epoch": 0.5918744645353294, + "grad_norm": 3.3673639450398722, + "learning_rate": 1.507091812976137e-06, + "loss": 0.6612, + "step": 6563 + }, + { + "epoch": 0.5919646480587997, + "grad_norm": 1.6770454386661173, + "learning_rate": 1.5065256436732773e-06, + "loss": 0.7018, + "step": 6564 + }, + { + "epoch": 0.5920548315822699, + "grad_norm": 1.946705086091616, + "learning_rate": 1.5059595164763717e-06, + "loss": 0.6817, + "step": 6565 + }, + { + "epoch": 0.5921450151057401, + "grad_norm": 0.5747369516989698, + "learning_rate": 1.5053934314337243e-06, + "loss": 0.586, + "step": 6566 + }, + { + "epoch": 0.5922351986292105, + "grad_norm": 1.8454934621740067, + "learning_rate": 1.5048273885936356e-06, + "loss": 0.7409, + "step": 6567 + }, + { + "epoch": 0.5923253821526807, + "grad_norm": 1.443262904844436, + "learning_rate": 1.5042613880044053e-06, + "loss": 0.6698, + "step": 6568 + }, + { + "epoch": 0.592415565676151, + "grad_norm": 1.6994508782044304, + "learning_rate": 1.5036954297143265e-06, + "loss": 0.7768, + "step": 6569 + }, + { + "epoch": 0.5925057491996212, + "grad_norm": 1.9379247384431928, + "learning_rate": 1.50312951377169e-06, + "loss": 0.7064, + "step": 6570 + }, + { + "epoch": 0.5925959327230915, + "grad_norm": 0.5868589911750479, + "learning_rate": 1.502563640224784e-06, + "loss": 0.5885, + "step": 6571 + }, + { + "epoch": 0.5926861162465618, + "grad_norm": 1.8776161380409109, + "learning_rate": 1.5019978091218903e-06, + "loss": 0.7453, + "step": 6572 + }, + { + "epoch": 0.592776299770032, + "grad_norm": 1.3690009161484233, + "learning_rate": 1.50143202051129e-06, + "loss": 0.8093, + "step": 6573 + }, + { + "epoch": 0.5928664832935022, + "grad_norm": 1.8870406419211232, + "learning_rate": 1.500866274441258e-06, + "loss": 0.8082, + "step": 6574 + }, + { + "epoch": 0.5929566668169726, + "grad_norm": 2.5195974888337433, + "learning_rate": 1.5003005709600682e-06, + "loss": 0.7319, + "step": 6575 + }, + { + "epoch": 0.5930468503404428, + "grad_norm": 1.9751437678099983, + "learning_rate": 1.4997349101159885e-06, + "loss": 0.8253, + "step": 6576 + }, + { + "epoch": 0.593137033863913, + "grad_norm": 2.166563276122661, + "learning_rate": 1.4991692919572854e-06, + "loss": 0.7085, + "step": 6577 + }, + { + "epoch": 0.5932272173873834, + "grad_norm": 2.1233213189392495, + "learning_rate": 1.4986037165322199e-06, + "loss": 0.7637, + "step": 6578 + }, + { + "epoch": 0.5933174009108536, + "grad_norm": 0.7640100301189009, + "learning_rate": 1.498038183889049e-06, + "loss": 0.6156, + "step": 6579 + }, + { + "epoch": 0.5934075844343238, + "grad_norm": 2.4696331134702834, + "learning_rate": 1.4974726940760292e-06, + "loss": 0.6691, + "step": 6580 + }, + { + "epoch": 0.5934977679577941, + "grad_norm": 2.1867775950362645, + "learning_rate": 1.496907247141409e-06, + "loss": 0.8242, + "step": 6581 + }, + { + "epoch": 0.5935879514812644, + "grad_norm": 0.7464137564431993, + "learning_rate": 1.4963418431334372e-06, + "loss": 0.6576, + "step": 6582 + }, + { + "epoch": 0.5936781350047347, + "grad_norm": 2.373610742711673, + "learning_rate": 1.4957764821003566e-06, + "loss": 0.7326, + "step": 6583 + }, + { + "epoch": 0.5937683185282049, + "grad_norm": 1.626726480476083, + "learning_rate": 1.4952111640904063e-06, + "loss": 0.7558, + "step": 6584 + }, + { + "epoch": 0.5938585020516751, + "grad_norm": 1.5156525284931521, + "learning_rate": 1.494645889151823e-06, + "loss": 0.7343, + "step": 6585 + }, + { + "epoch": 0.5939486855751455, + "grad_norm": 0.6510335820013972, + "learning_rate": 1.494080657332839e-06, + "loss": 0.5624, + "step": 6586 + }, + { + "epoch": 0.5940388690986157, + "grad_norm": 1.98425352670924, + "learning_rate": 1.4935154686816832e-06, + "loss": 0.7037, + "step": 6587 + }, + { + "epoch": 0.5941290526220859, + "grad_norm": 1.959570419568546, + "learning_rate": 1.4929503232465802e-06, + "loss": 0.8232, + "step": 6588 + }, + { + "epoch": 0.5942192361455562, + "grad_norm": 1.658442000402221, + "learning_rate": 1.492385221075751e-06, + "loss": 0.6644, + "step": 6589 + }, + { + "epoch": 0.5943094196690265, + "grad_norm": 3.2342588247176, + "learning_rate": 1.4918201622174142e-06, + "loss": 0.7814, + "step": 6590 + }, + { + "epoch": 0.5943996031924967, + "grad_norm": 1.7810797359003234, + "learning_rate": 1.4912551467197827e-06, + "loss": 0.7687, + "step": 6591 + }, + { + "epoch": 0.594489786715967, + "grad_norm": 2.3088097898861273, + "learning_rate": 1.4906901746310678e-06, + "loss": 0.8384, + "step": 6592 + }, + { + "epoch": 0.5945799702394372, + "grad_norm": 1.6361948417534704, + "learning_rate": 1.4901252459994757e-06, + "loss": 0.7138, + "step": 6593 + }, + { + "epoch": 0.5946701537629075, + "grad_norm": 1.6594706477241346, + "learning_rate": 1.489560360873208e-06, + "loss": 0.7335, + "step": 6594 + }, + { + "epoch": 0.5947603372863778, + "grad_norm": 1.7163298473968538, + "learning_rate": 1.4889955193004659e-06, + "loss": 0.728, + "step": 6595 + }, + { + "epoch": 0.594850520809848, + "grad_norm": 0.6444066707549958, + "learning_rate": 1.4884307213294428e-06, + "loss": 0.5746, + "step": 6596 + }, + { + "epoch": 0.5949407043333182, + "grad_norm": 2.3376259713729555, + "learning_rate": 1.4878659670083321e-06, + "loss": 0.737, + "step": 6597 + }, + { + "epoch": 0.5950308878567886, + "grad_norm": 1.643655212918408, + "learning_rate": 1.4873012563853208e-06, + "loss": 0.6714, + "step": 6598 + }, + { + "epoch": 0.5951210713802588, + "grad_norm": 2.214361444400013, + "learning_rate": 1.4867365895085935e-06, + "loss": 0.753, + "step": 6599 + }, + { + "epoch": 0.5952112549037291, + "grad_norm": 2.3889858635236254, + "learning_rate": 1.4861719664263301e-06, + "loss": 0.7268, + "step": 6600 + }, + { + "epoch": 0.5953014384271994, + "grad_norm": 2.6065721820413206, + "learning_rate": 1.485607387186708e-06, + "loss": 0.852, + "step": 6601 + }, + { + "epoch": 0.5953916219506696, + "grad_norm": 1.8768440715059278, + "learning_rate": 1.4850428518379001e-06, + "loss": 0.7435, + "step": 6602 + }, + { + "epoch": 0.5954818054741399, + "grad_norm": 1.3705413227135181, + "learning_rate": 1.4844783604280746e-06, + "loss": 0.7448, + "step": 6603 + }, + { + "epoch": 0.5955719889976101, + "grad_norm": 3.045776325602086, + "learning_rate": 1.483913913005399e-06, + "loss": 0.6923, + "step": 6604 + }, + { + "epoch": 0.5956621725210804, + "grad_norm": 2.621691480722469, + "learning_rate": 1.483349509618034e-06, + "loss": 0.81, + "step": 6605 + }, + { + "epoch": 0.5957523560445507, + "grad_norm": 1.6530049761808603, + "learning_rate": 1.4827851503141367e-06, + "loss": 0.6819, + "step": 6606 + }, + { + "epoch": 0.5958425395680209, + "grad_norm": 1.5720601391827462, + "learning_rate": 1.482220835141863e-06, + "loss": 0.6756, + "step": 6607 + }, + { + "epoch": 0.5959327230914911, + "grad_norm": 1.6877540114632965, + "learning_rate": 1.481656564149362e-06, + "loss": 0.828, + "step": 6608 + }, + { + "epoch": 0.5960229066149615, + "grad_norm": 2.0172231090425456, + "learning_rate": 1.4810923373847818e-06, + "loss": 0.7464, + "step": 6609 + }, + { + "epoch": 0.5961130901384317, + "grad_norm": 1.6856362682697148, + "learning_rate": 1.4805281548962647e-06, + "loss": 0.7554, + "step": 6610 + }, + { + "epoch": 0.596203273661902, + "grad_norm": 2.1873700784520715, + "learning_rate": 1.4799640167319488e-06, + "loss": 0.6938, + "step": 6611 + }, + { + "epoch": 0.5962934571853722, + "grad_norm": 1.5697807056809632, + "learning_rate": 1.4793999229399714e-06, + "loss": 0.7371, + "step": 6612 + }, + { + "epoch": 0.5963836407088425, + "grad_norm": 1.6073468956777948, + "learning_rate": 1.4788358735684626e-06, + "loss": 0.8142, + "step": 6613 + }, + { + "epoch": 0.5964738242323128, + "grad_norm": 1.5243185772215304, + "learning_rate": 1.4782718686655514e-06, + "loss": 0.8168, + "step": 6614 + }, + { + "epoch": 0.596564007755783, + "grad_norm": 1.8697798222952093, + "learning_rate": 1.4777079082793605e-06, + "loss": 0.7997, + "step": 6615 + }, + { + "epoch": 0.5966541912792532, + "grad_norm": 1.9464314012019435, + "learning_rate": 1.4771439924580108e-06, + "loss": 0.7203, + "step": 6616 + }, + { + "epoch": 0.5967443748027236, + "grad_norm": 1.31386445284633, + "learning_rate": 1.4765801212496189e-06, + "loss": 0.7903, + "step": 6617 + }, + { + "epoch": 0.5968345583261938, + "grad_norm": 1.8825202176856208, + "learning_rate": 1.4760162947022968e-06, + "loss": 0.6694, + "step": 6618 + }, + { + "epoch": 0.596924741849664, + "grad_norm": 1.6917309878794449, + "learning_rate": 1.475452512864154e-06, + "loss": 0.7373, + "step": 6619 + }, + { + "epoch": 0.5970149253731343, + "grad_norm": 1.4148814322194563, + "learning_rate": 1.4748887757832945e-06, + "loss": 0.7561, + "step": 6620 + }, + { + "epoch": 0.5971051088966046, + "grad_norm": 1.58173991216943, + "learning_rate": 1.4743250835078209e-06, + "loss": 0.7834, + "step": 6621 + }, + { + "epoch": 0.5971952924200749, + "grad_norm": 1.791308662960665, + "learning_rate": 1.4737614360858297e-06, + "loss": 0.7102, + "step": 6622 + }, + { + "epoch": 0.5972854759435451, + "grad_norm": 2.329088664469004, + "learning_rate": 1.4731978335654133e-06, + "loss": 0.703, + "step": 6623 + }, + { + "epoch": 0.5973756594670154, + "grad_norm": 2.167885058737933, + "learning_rate": 1.4726342759946638e-06, + "loss": 0.6882, + "step": 6624 + }, + { + "epoch": 0.5974658429904857, + "grad_norm": 1.6392605148287702, + "learning_rate": 1.4720707634216653e-06, + "loss": 0.7496, + "step": 6625 + }, + { + "epoch": 0.5975560265139559, + "grad_norm": 1.969666827633672, + "learning_rate": 1.4715072958945e-06, + "loss": 0.7569, + "step": 6626 + }, + { + "epoch": 0.5976462100374261, + "grad_norm": 1.9832399982107385, + "learning_rate": 1.470943873461247e-06, + "loss": 0.7484, + "step": 6627 + }, + { + "epoch": 0.5977363935608965, + "grad_norm": 2.8994834637197497, + "learning_rate": 1.470380496169979e-06, + "loss": 0.7116, + "step": 6628 + }, + { + "epoch": 0.5978265770843667, + "grad_norm": 1.9224746512935698, + "learning_rate": 1.4698171640687682e-06, + "loss": 0.7395, + "step": 6629 + }, + { + "epoch": 0.5979167606078369, + "grad_norm": 1.899038086383192, + "learning_rate": 1.4692538772056792e-06, + "loss": 0.6841, + "step": 6630 + }, + { + "epoch": 0.5980069441313072, + "grad_norm": 1.4024094165101781, + "learning_rate": 1.4686906356287772e-06, + "loss": 0.6625, + "step": 6631 + }, + { + "epoch": 0.5980971276547775, + "grad_norm": 1.5615824484872587, + "learning_rate": 1.4681274393861194e-06, + "loss": 0.78, + "step": 6632 + }, + { + "epoch": 0.5981873111782477, + "grad_norm": 1.8855062502097621, + "learning_rate": 1.4675642885257603e-06, + "loss": 0.8486, + "step": 6633 + }, + { + "epoch": 0.598277494701718, + "grad_norm": 2.7662398522316285, + "learning_rate": 1.4670011830957529e-06, + "loss": 0.8355, + "step": 6634 + }, + { + "epoch": 0.5983676782251882, + "grad_norm": 2.2837917851476117, + "learning_rate": 1.4664381231441427e-06, + "loss": 0.7161, + "step": 6635 + }, + { + "epoch": 0.5984578617486586, + "grad_norm": 1.7189893382497983, + "learning_rate": 1.4658751087189746e-06, + "loss": 0.7611, + "step": 6636 + }, + { + "epoch": 0.5985480452721288, + "grad_norm": 1.6090390493610696, + "learning_rate": 1.4653121398682874e-06, + "loss": 0.7989, + "step": 6637 + }, + { + "epoch": 0.598638228795599, + "grad_norm": 12.142294593410426, + "learning_rate": 1.4647492166401159e-06, + "loss": 0.6867, + "step": 6638 + }, + { + "epoch": 0.5987284123190693, + "grad_norm": 1.7118382282075058, + "learning_rate": 1.4641863390824934e-06, + "loss": 0.7374, + "step": 6639 + }, + { + "epoch": 0.5988185958425396, + "grad_norm": 2.0124516309529152, + "learning_rate": 1.4636235072434465e-06, + "loss": 0.7398, + "step": 6640 + }, + { + "epoch": 0.5989087793660098, + "grad_norm": 1.6757613316488162, + "learning_rate": 1.4630607211709994e-06, + "loss": 0.7752, + "step": 6641 + }, + { + "epoch": 0.5989989628894801, + "grad_norm": 2.831307491030363, + "learning_rate": 1.4624979809131723e-06, + "loss": 0.7554, + "step": 6642 + }, + { + "epoch": 0.5990891464129503, + "grad_norm": 1.473650447469448, + "learning_rate": 1.4619352865179814e-06, + "loss": 0.658, + "step": 6643 + }, + { + "epoch": 0.5991793299364206, + "grad_norm": 2.232884156430059, + "learning_rate": 1.4613726380334391e-06, + "loss": 0.7195, + "step": 6644 + }, + { + "epoch": 0.5992695134598909, + "grad_norm": 2.0980651570576345, + "learning_rate": 1.4608100355075522e-06, + "loss": 0.6032, + "step": 6645 + }, + { + "epoch": 0.5993596969833611, + "grad_norm": 0.6221361829056511, + "learning_rate": 1.460247478988327e-06, + "loss": 0.5561, + "step": 6646 + }, + { + "epoch": 0.5994498805068315, + "grad_norm": 1.5361363822570568, + "learning_rate": 1.4596849685237623e-06, + "loss": 0.7654, + "step": 6647 + }, + { + "epoch": 0.5995400640303017, + "grad_norm": 2.0554158475441047, + "learning_rate": 1.459122504161856e-06, + "loss": 0.6572, + "step": 6648 + }, + { + "epoch": 0.5996302475537719, + "grad_norm": 1.9181466262905844, + "learning_rate": 1.4585600859506001e-06, + "loss": 0.72, + "step": 6649 + }, + { + "epoch": 0.5997204310772422, + "grad_norm": 1.6401663729549787, + "learning_rate": 1.4579977139379826e-06, + "loss": 0.8087, + "step": 6650 + }, + { + "epoch": 0.5998106146007125, + "grad_norm": 1.9621922707290842, + "learning_rate": 1.4574353881719895e-06, + "loss": 0.8066, + "step": 6651 + }, + { + "epoch": 0.5999007981241827, + "grad_norm": 1.7224381150864998, + "learning_rate": 1.4568731087005998e-06, + "loss": 0.753, + "step": 6652 + }, + { + "epoch": 0.599990981647653, + "grad_norm": 1.2657984391041157, + "learning_rate": 1.4563108755717916e-06, + "loss": 0.7718, + "step": 6653 + }, + { + "epoch": 0.6000811651711232, + "grad_norm": 1.892683059727988, + "learning_rate": 1.455748688833538e-06, + "loss": 0.7771, + "step": 6654 + }, + { + "epoch": 0.6001713486945935, + "grad_norm": 1.741291517103679, + "learning_rate": 1.4551865485338065e-06, + "loss": 0.6619, + "step": 6655 + }, + { + "epoch": 0.6002615322180638, + "grad_norm": 1.5403822454201475, + "learning_rate": 1.4546244547205629e-06, + "loss": 0.8179, + "step": 6656 + }, + { + "epoch": 0.600351715741534, + "grad_norm": 1.8130623175362444, + "learning_rate": 1.4540624074417678e-06, + "loss": 0.7751, + "step": 6657 + }, + { + "epoch": 0.6004418992650042, + "grad_norm": 1.4328431703369529, + "learning_rate": 1.453500406745379e-06, + "loss": 0.8083, + "step": 6658 + }, + { + "epoch": 0.6005320827884746, + "grad_norm": 3.2083339113177156, + "learning_rate": 1.4529384526793486e-06, + "loss": 0.8275, + "step": 6659 + }, + { + "epoch": 0.6006222663119448, + "grad_norm": 1.636036733042389, + "learning_rate": 1.4523765452916252e-06, + "loss": 0.8052, + "step": 6660 + }, + { + "epoch": 0.600712449835415, + "grad_norm": 1.9960933916722403, + "learning_rate": 1.4518146846301554e-06, + "loss": 0.7801, + "step": 6661 + }, + { + "epoch": 0.6008026333588853, + "grad_norm": 2.373700338429722, + "learning_rate": 1.4512528707428787e-06, + "loss": 0.7255, + "step": 6662 + }, + { + "epoch": 0.6008928168823556, + "grad_norm": 2.5766911073769965, + "learning_rate": 1.4506911036777335e-06, + "loss": 0.7309, + "step": 6663 + }, + { + "epoch": 0.6009830004058259, + "grad_norm": 0.67243331067996, + "learning_rate": 1.450129383482651e-06, + "loss": 0.5488, + "step": 6664 + }, + { + "epoch": 0.6010731839292961, + "grad_norm": 1.685577498386537, + "learning_rate": 1.4495677102055629e-06, + "loss": 0.7052, + "step": 6665 + }, + { + "epoch": 0.6011633674527663, + "grad_norm": 1.8949680100209305, + "learning_rate": 1.4490060838943924e-06, + "loss": 0.7857, + "step": 6666 + }, + { + "epoch": 0.6012535509762367, + "grad_norm": 2.5295152724544523, + "learning_rate": 1.4484445045970609e-06, + "loss": 0.7531, + "step": 6667 + }, + { + "epoch": 0.6013437344997069, + "grad_norm": 2.740630400596874, + "learning_rate": 1.447882972361485e-06, + "loss": 0.7451, + "step": 6668 + }, + { + "epoch": 0.6014339180231771, + "grad_norm": 1.5601831425243522, + "learning_rate": 1.4473214872355785e-06, + "loss": 0.7376, + "step": 6669 + }, + { + "epoch": 0.6015241015466475, + "grad_norm": 2.0933876293373372, + "learning_rate": 1.4467600492672508e-06, + "loss": 0.7828, + "step": 6670 + }, + { + "epoch": 0.6016142850701177, + "grad_norm": 1.6742675631728197, + "learning_rate": 1.4461986585044054e-06, + "loss": 0.7268, + "step": 6671 + }, + { + "epoch": 0.601704468593588, + "grad_norm": 1.807269936320095, + "learning_rate": 1.4456373149949446e-06, + "loss": 0.7255, + "step": 6672 + }, + { + "epoch": 0.6017946521170582, + "grad_norm": 4.307109393776301, + "learning_rate": 1.4450760187867648e-06, + "loss": 0.7582, + "step": 6673 + }, + { + "epoch": 0.6018848356405285, + "grad_norm": 1.851546661695511, + "learning_rate": 1.4445147699277581e-06, + "loss": 0.7446, + "step": 6674 + }, + { + "epoch": 0.6019750191639988, + "grad_norm": 1.6879875750287792, + "learning_rate": 1.4439535684658154e-06, + "loss": 0.7963, + "step": 6675 + }, + { + "epoch": 0.602065202687469, + "grad_norm": 0.767904935185912, + "learning_rate": 1.44339241444882e-06, + "loss": 0.6179, + "step": 6676 + }, + { + "epoch": 0.6021553862109392, + "grad_norm": 2.11274867653842, + "learning_rate": 1.4428313079246518e-06, + "loss": 0.7008, + "step": 6677 + }, + { + "epoch": 0.6022455697344096, + "grad_norm": 0.6639165381291842, + "learning_rate": 1.4422702489411896e-06, + "loss": 0.5958, + "step": 6678 + }, + { + "epoch": 0.6023357532578798, + "grad_norm": 1.9964479255371812, + "learning_rate": 1.4417092375463043e-06, + "loss": 0.7019, + "step": 6679 + }, + { + "epoch": 0.60242593678135, + "grad_norm": 1.6759969221282593, + "learning_rate": 1.441148273787866e-06, + "loss": 0.7875, + "step": 6680 + }, + { + "epoch": 0.6025161203048203, + "grad_norm": 0.8160537615858412, + "learning_rate": 1.4405873577137383e-06, + "loss": 0.6022, + "step": 6681 + }, + { + "epoch": 0.6026063038282906, + "grad_norm": 3.152256015469008, + "learning_rate": 1.4400264893717816e-06, + "loss": 0.7586, + "step": 6682 + }, + { + "epoch": 0.6026964873517608, + "grad_norm": 1.6160499884456017, + "learning_rate": 1.4394656688098526e-06, + "loss": 0.7972, + "step": 6683 + }, + { + "epoch": 0.6027866708752311, + "grad_norm": 1.4890471963005971, + "learning_rate": 1.4389048960758032e-06, + "loss": 0.7717, + "step": 6684 + }, + { + "epoch": 0.6028768543987013, + "grad_norm": 2.787452211526319, + "learning_rate": 1.4383441712174826e-06, + "loss": 0.7859, + "step": 6685 + }, + { + "epoch": 0.6029670379221717, + "grad_norm": 3.1507971466442832, + "learning_rate": 1.4377834942827333e-06, + "loss": 0.8136, + "step": 6686 + }, + { + "epoch": 0.6030572214456419, + "grad_norm": 2.3891250791767717, + "learning_rate": 1.437222865319397e-06, + "loss": 0.6905, + "step": 6687 + }, + { + "epoch": 0.6031474049691121, + "grad_norm": 1.5294730327382484, + "learning_rate": 1.4366622843753092e-06, + "loss": 0.6801, + "step": 6688 + }, + { + "epoch": 0.6032375884925824, + "grad_norm": 1.3100022208944815, + "learning_rate": 1.4361017514983006e-06, + "loss": 0.7684, + "step": 6689 + }, + { + "epoch": 0.6033277720160527, + "grad_norm": 1.6559440042224371, + "learning_rate": 1.4355412667362006e-06, + "loss": 0.8053, + "step": 6690 + }, + { + "epoch": 0.6034179555395229, + "grad_norm": 2.1358668775401237, + "learning_rate": 1.4349808301368311e-06, + "loss": 0.7105, + "step": 6691 + }, + { + "epoch": 0.6035081390629932, + "grad_norm": 3.1036239422668794, + "learning_rate": 1.4344204417480139e-06, + "loss": 0.7927, + "step": 6692 + }, + { + "epoch": 0.6035983225864634, + "grad_norm": 1.811526925219857, + "learning_rate": 1.4338601016175628e-06, + "loss": 0.7025, + "step": 6693 + }, + { + "epoch": 0.6036885061099337, + "grad_norm": 1.57181276232325, + "learning_rate": 1.433299809793289e-06, + "loss": 0.7486, + "step": 6694 + }, + { + "epoch": 0.603778689633404, + "grad_norm": 3.435939087635348, + "learning_rate": 1.432739566323001e-06, + "loss": 0.6337, + "step": 6695 + }, + { + "epoch": 0.6038688731568742, + "grad_norm": 2.1121392116904953, + "learning_rate": 1.4321793712545004e-06, + "loss": 0.7146, + "step": 6696 + }, + { + "epoch": 0.6039590566803446, + "grad_norm": 13.754515859999172, + "learning_rate": 1.4316192246355873e-06, + "loss": 0.7974, + "step": 6697 + }, + { + "epoch": 0.6040492402038148, + "grad_norm": 0.6113445779457174, + "learning_rate": 1.4310591265140555e-06, + "loss": 0.5798, + "step": 6698 + }, + { + "epoch": 0.604139423727285, + "grad_norm": 1.4504624451460482, + "learning_rate": 1.4304990769376963e-06, + "loss": 0.7562, + "step": 6699 + }, + { + "epoch": 0.6042296072507553, + "grad_norm": 1.4637492655972952, + "learning_rate": 1.4299390759542962e-06, + "loss": 0.7319, + "step": 6700 + }, + { + "epoch": 0.6043197907742256, + "grad_norm": 1.624641672387997, + "learning_rate": 1.4293791236116368e-06, + "loss": 0.735, + "step": 6701 + }, + { + "epoch": 0.6044099742976958, + "grad_norm": 1.9460243240279018, + "learning_rate": 1.4288192199574978e-06, + "loss": 0.7154, + "step": 6702 + }, + { + "epoch": 0.6045001578211661, + "grad_norm": 0.666889657536086, + "learning_rate": 1.4282593650396524e-06, + "loss": 0.5523, + "step": 6703 + }, + { + "epoch": 0.6045903413446363, + "grad_norm": 2.16165214190558, + "learning_rate": 1.4276995589058695e-06, + "loss": 0.8511, + "step": 6704 + }, + { + "epoch": 0.6046805248681066, + "grad_norm": 0.5936577875910899, + "learning_rate": 1.4271398016039168e-06, + "loss": 0.5484, + "step": 6705 + }, + { + "epoch": 0.6047707083915769, + "grad_norm": 3.269650255450688, + "learning_rate": 1.4265800931815542e-06, + "loss": 0.7897, + "step": 6706 + }, + { + "epoch": 0.6048608919150471, + "grad_norm": 1.752491267641, + "learning_rate": 1.4260204336865406e-06, + "loss": 0.7598, + "step": 6707 + }, + { + "epoch": 0.6049510754385173, + "grad_norm": 2.488200858597599, + "learning_rate": 1.4254608231666281e-06, + "loss": 0.7495, + "step": 6708 + }, + { + "epoch": 0.6050412589619877, + "grad_norm": 2.2612593057334474, + "learning_rate": 1.4249012616695661e-06, + "loss": 0.759, + "step": 6709 + }, + { + "epoch": 0.6051314424854579, + "grad_norm": 1.6969088584374188, + "learning_rate": 1.4243417492431e-06, + "loss": 0.7561, + "step": 6710 + }, + { + "epoch": 0.6052216260089281, + "grad_norm": 2.099549758419608, + "learning_rate": 1.4237822859349696e-06, + "loss": 0.7337, + "step": 6711 + }, + { + "epoch": 0.6053118095323984, + "grad_norm": 1.445287673324567, + "learning_rate": 1.423222871792912e-06, + "loss": 0.7216, + "step": 6712 + }, + { + "epoch": 0.6054019930558687, + "grad_norm": 1.7287272005929173, + "learning_rate": 1.4226635068646586e-06, + "loss": 0.8031, + "step": 6713 + }, + { + "epoch": 0.605492176579339, + "grad_norm": 0.6253552380950097, + "learning_rate": 1.4221041911979393e-06, + "loss": 0.5724, + "step": 6714 + }, + { + "epoch": 0.6055823601028092, + "grad_norm": 1.7137058106004206, + "learning_rate": 1.4215449248404765e-06, + "loss": 0.7697, + "step": 6715 + }, + { + "epoch": 0.6056725436262794, + "grad_norm": 17.85842318931304, + "learning_rate": 1.4209857078399896e-06, + "loss": 0.7715, + "step": 6716 + }, + { + "epoch": 0.6057627271497498, + "grad_norm": 2.091243211150337, + "learning_rate": 1.4204265402441955e-06, + "loss": 0.7171, + "step": 6717 + }, + { + "epoch": 0.60585291067322, + "grad_norm": 1.6802866732013866, + "learning_rate": 1.419867422100804e-06, + "loss": 0.6233, + "step": 6718 + }, + { + "epoch": 0.6059430941966902, + "grad_norm": 1.7476569566652782, + "learning_rate": 1.4193083534575236e-06, + "loss": 0.7383, + "step": 6719 + }, + { + "epoch": 0.6060332777201606, + "grad_norm": 1.9008515005931235, + "learning_rate": 1.4187493343620567e-06, + "loss": 0.7746, + "step": 6720 + }, + { + "epoch": 0.6061234612436308, + "grad_norm": 1.6437660303530806, + "learning_rate": 1.4181903648621006e-06, + "loss": 0.7814, + "step": 6721 + }, + { + "epoch": 0.606213644767101, + "grad_norm": 1.5452228547140348, + "learning_rate": 1.4176314450053512e-06, + "loss": 0.7778, + "step": 6722 + }, + { + "epoch": 0.6063038282905713, + "grad_norm": 2.0085964231054145, + "learning_rate": 1.4170725748394977e-06, + "loss": 0.7365, + "step": 6723 + }, + { + "epoch": 0.6063940118140416, + "grad_norm": 1.4391919006420857, + "learning_rate": 1.4165137544122266e-06, + "loss": 0.6624, + "step": 6724 + }, + { + "epoch": 0.6064841953375119, + "grad_norm": 1.7679649309158376, + "learning_rate": 1.4159549837712194e-06, + "loss": 0.8148, + "step": 6725 + }, + { + "epoch": 0.6065743788609821, + "grad_norm": 1.514471972261899, + "learning_rate": 1.415396262964153e-06, + "loss": 0.6663, + "step": 6726 + }, + { + "epoch": 0.6066645623844523, + "grad_norm": 1.8087177275314001, + "learning_rate": 1.4148375920387016e-06, + "loss": 0.826, + "step": 6727 + }, + { + "epoch": 0.6067547459079227, + "grad_norm": 0.6496008582827107, + "learning_rate": 1.4142789710425325e-06, + "loss": 0.576, + "step": 6728 + }, + { + "epoch": 0.6068449294313929, + "grad_norm": 20.860361048791496, + "learning_rate": 1.4137204000233118e-06, + "loss": 0.6911, + "step": 6729 + }, + { + "epoch": 0.6069351129548631, + "grad_norm": 1.773772691147383, + "learning_rate": 1.4131618790286987e-06, + "loss": 0.7429, + "step": 6730 + }, + { + "epoch": 0.6070252964783334, + "grad_norm": 1.866458638838371, + "learning_rate": 1.4126034081063506e-06, + "loss": 0.7354, + "step": 6731 + }, + { + "epoch": 0.6071154800018037, + "grad_norm": 1.7291304879444436, + "learning_rate": 1.4120449873039186e-06, + "loss": 0.7333, + "step": 6732 + }, + { + "epoch": 0.6072056635252739, + "grad_norm": 2.716195846979814, + "learning_rate": 1.4114866166690494e-06, + "loss": 0.7861, + "step": 6733 + }, + { + "epoch": 0.6072958470487442, + "grad_norm": 1.583455214912331, + "learning_rate": 1.4109282962493877e-06, + "loss": 0.702, + "step": 6734 + }, + { + "epoch": 0.6073860305722144, + "grad_norm": 1.7712116136611238, + "learning_rate": 1.4103700260925716e-06, + "loss": 0.7387, + "step": 6735 + }, + { + "epoch": 0.6074762140956848, + "grad_norm": 1.9996100880106844, + "learning_rate": 1.4098118062462364e-06, + "loss": 0.7863, + "step": 6736 + }, + { + "epoch": 0.607566397619155, + "grad_norm": 1.48281923378962, + "learning_rate": 1.4092536367580123e-06, + "loss": 0.7467, + "step": 6737 + }, + { + "epoch": 0.6076565811426252, + "grad_norm": 1.7590592095850348, + "learning_rate": 1.4086955176755248e-06, + "loss": 0.6683, + "step": 6738 + }, + { + "epoch": 0.6077467646660955, + "grad_norm": 1.8414770437234413, + "learning_rate": 1.4081374490463964e-06, + "loss": 0.6902, + "step": 6739 + }, + { + "epoch": 0.6078369481895658, + "grad_norm": 1.706679947448754, + "learning_rate": 1.4075794309182443e-06, + "loss": 0.7611, + "step": 6740 + }, + { + "epoch": 0.607927131713036, + "grad_norm": 2.0549375432679646, + "learning_rate": 1.407021463338682e-06, + "loss": 0.7676, + "step": 6741 + }, + { + "epoch": 0.6080173152365063, + "grad_norm": 0.5906742135994495, + "learning_rate": 1.4064635463553177e-06, + "loss": 0.5213, + "step": 6742 + }, + { + "epoch": 0.6081074987599766, + "grad_norm": 1.7127265808125791, + "learning_rate": 1.4059056800157563e-06, + "loss": 0.78, + "step": 6743 + }, + { + "epoch": 0.6081976822834468, + "grad_norm": 1.5470990876000887, + "learning_rate": 1.4053478643675982e-06, + "loss": 0.6812, + "step": 6744 + }, + { + "epoch": 0.6082878658069171, + "grad_norm": 2.1714420024198313, + "learning_rate": 1.4047900994584389e-06, + "loss": 0.6494, + "step": 6745 + }, + { + "epoch": 0.6083780493303873, + "grad_norm": 2.342001198961324, + "learning_rate": 1.404232385335871e-06, + "loss": 0.6634, + "step": 6746 + }, + { + "epoch": 0.6084682328538576, + "grad_norm": 7.092474869047222, + "learning_rate": 1.4036747220474806e-06, + "loss": 0.7487, + "step": 6747 + }, + { + "epoch": 0.6085584163773279, + "grad_norm": 2.1748816447009323, + "learning_rate": 1.4031171096408506e-06, + "loss": 0.7049, + "step": 6748 + }, + { + "epoch": 0.6086485999007981, + "grad_norm": 2.3424541705848343, + "learning_rate": 1.4025595481635607e-06, + "loss": 0.7753, + "step": 6749 + }, + { + "epoch": 0.6087387834242683, + "grad_norm": 1.8541157176129317, + "learning_rate": 1.4020020376631836e-06, + "loss": 0.7711, + "step": 6750 + }, + { + "epoch": 0.6088289669477387, + "grad_norm": 2.3369400781760743, + "learning_rate": 1.4014445781872908e-06, + "loss": 0.7626, + "step": 6751 + }, + { + "epoch": 0.6089191504712089, + "grad_norm": 1.7547961316064544, + "learning_rate": 1.4008871697834465e-06, + "loss": 0.8127, + "step": 6752 + }, + { + "epoch": 0.6090093339946792, + "grad_norm": 2.488289106946485, + "learning_rate": 1.400329812499213e-06, + "loss": 0.7832, + "step": 6753 + }, + { + "epoch": 0.6090995175181494, + "grad_norm": 1.7105462310573762, + "learning_rate": 1.3997725063821458e-06, + "loss": 0.7637, + "step": 6754 + }, + { + "epoch": 0.6091897010416197, + "grad_norm": 1.6924942428964267, + "learning_rate": 1.3992152514797978e-06, + "loss": 0.7131, + "step": 6755 + }, + { + "epoch": 0.60927988456509, + "grad_norm": 0.6495673665532686, + "learning_rate": 1.398658047839718e-06, + "loss": 0.596, + "step": 6756 + }, + { + "epoch": 0.6093700680885602, + "grad_norm": 0.6032795367133832, + "learning_rate": 1.3981008955094481e-06, + "loss": 0.5656, + "step": 6757 + }, + { + "epoch": 0.6094602516120304, + "grad_norm": 8.502887908422828, + "learning_rate": 1.39754379453653e-06, + "loss": 0.7223, + "step": 6758 + }, + { + "epoch": 0.6095504351355008, + "grad_norm": 3.419823854829234, + "learning_rate": 1.3969867449684972e-06, + "loss": 0.7582, + "step": 6759 + }, + { + "epoch": 0.609640618658971, + "grad_norm": 2.16689822597946, + "learning_rate": 1.396429746852879e-06, + "loss": 0.7402, + "step": 6760 + }, + { + "epoch": 0.6097308021824412, + "grad_norm": 1.6777769356459524, + "learning_rate": 1.395872800237204e-06, + "loss": 0.7259, + "step": 6761 + }, + { + "epoch": 0.6098209857059115, + "grad_norm": 0.7000383656069757, + "learning_rate": 1.3953159051689918e-06, + "loss": 0.5975, + "step": 6762 + }, + { + "epoch": 0.6099111692293818, + "grad_norm": 3.1044636185379404, + "learning_rate": 1.3947590616957618e-06, + "loss": 0.7924, + "step": 6763 + }, + { + "epoch": 0.610001352752852, + "grad_norm": 5.162868067295333, + "learning_rate": 1.3942022698650258e-06, + "loss": 0.783, + "step": 6764 + }, + { + "epoch": 0.6100915362763223, + "grad_norm": 4.37963817512862, + "learning_rate": 1.3936455297242917e-06, + "loss": 0.7823, + "step": 6765 + }, + { + "epoch": 0.6101817197997926, + "grad_norm": 9.52873421908026, + "learning_rate": 1.3930888413210652e-06, + "loss": 0.7704, + "step": 6766 + }, + { + "epoch": 0.6102719033232629, + "grad_norm": 2.7039773291873277, + "learning_rate": 1.392532204702845e-06, + "loss": 0.7114, + "step": 6767 + }, + { + "epoch": 0.6103620868467331, + "grad_norm": 2.4281496535219325, + "learning_rate": 1.3919756199171266e-06, + "loss": 0.681, + "step": 6768 + }, + { + "epoch": 0.6104522703702033, + "grad_norm": 1.9948377745024861, + "learning_rate": 1.3914190870114009e-06, + "loss": 0.8246, + "step": 6769 + }, + { + "epoch": 0.6105424538936737, + "grad_norm": 1.6998099164719067, + "learning_rate": 1.3908626060331541e-06, + "loss": 0.7999, + "step": 6770 + }, + { + "epoch": 0.6106326374171439, + "grad_norm": 2.6021299001183325, + "learning_rate": 1.3903061770298693e-06, + "loss": 0.7855, + "step": 6771 + }, + { + "epoch": 0.6107228209406141, + "grad_norm": 1.7021539001902446, + "learning_rate": 1.3897498000490223e-06, + "loss": 0.7218, + "step": 6772 + }, + { + "epoch": 0.6108130044640844, + "grad_norm": 1.6499277041231255, + "learning_rate": 1.3891934751380879e-06, + "loss": 0.8177, + "step": 6773 + }, + { + "epoch": 0.6109031879875547, + "grad_norm": 1.735628929147473, + "learning_rate": 1.3886372023445334e-06, + "loss": 0.7694, + "step": 6774 + }, + { + "epoch": 0.610993371511025, + "grad_norm": 1.4705179004117965, + "learning_rate": 1.3880809817158246e-06, + "loss": 0.8187, + "step": 6775 + }, + { + "epoch": 0.6110835550344952, + "grad_norm": 1.8495313231163724, + "learning_rate": 1.3875248132994206e-06, + "loss": 0.7404, + "step": 6776 + }, + { + "epoch": 0.6111737385579654, + "grad_norm": 1.8079637126277144, + "learning_rate": 1.386968697142776e-06, + "loss": 0.6694, + "step": 6777 + }, + { + "epoch": 0.6112639220814358, + "grad_norm": 1.5001939807076907, + "learning_rate": 1.386412633293343e-06, + "loss": 0.7994, + "step": 6778 + }, + { + "epoch": 0.611354105604906, + "grad_norm": 1.8786997850195932, + "learning_rate": 1.3858566217985672e-06, + "loss": 0.7248, + "step": 6779 + }, + { + "epoch": 0.6114442891283762, + "grad_norm": 8.882383472076599, + "learning_rate": 1.3853006627058905e-06, + "loss": 0.836, + "step": 6780 + }, + { + "epoch": 0.6115344726518465, + "grad_norm": 2.318687409329912, + "learning_rate": 1.3847447560627512e-06, + "loss": 0.6364, + "step": 6781 + }, + { + "epoch": 0.6116246561753168, + "grad_norm": 1.695124391588167, + "learning_rate": 1.3841889019165812e-06, + "loss": 0.7917, + "step": 6782 + }, + { + "epoch": 0.611714839698787, + "grad_norm": 2.269491434979124, + "learning_rate": 1.3836331003148106e-06, + "loss": 0.7125, + "step": 6783 + }, + { + "epoch": 0.6118050232222573, + "grad_norm": 1.4142909418700296, + "learning_rate": 1.3830773513048612e-06, + "loss": 0.7874, + "step": 6784 + }, + { + "epoch": 0.6118952067457275, + "grad_norm": 1.613584845473919, + "learning_rate": 1.382521654934155e-06, + "loss": 0.7664, + "step": 6785 + }, + { + "epoch": 0.6119853902691978, + "grad_norm": 1.578694684531231, + "learning_rate": 1.3819660112501057e-06, + "loss": 0.6924, + "step": 6786 + }, + { + "epoch": 0.6120755737926681, + "grad_norm": 1.6228940228444233, + "learning_rate": 1.3814104203001234e-06, + "loss": 0.7132, + "step": 6787 + }, + { + "epoch": 0.6121657573161383, + "grad_norm": 1.6045773298411596, + "learning_rate": 1.3808548821316156e-06, + "loss": 0.7881, + "step": 6788 + }, + { + "epoch": 0.6122559408396087, + "grad_norm": 1.5628217747288558, + "learning_rate": 1.3802993967919824e-06, + "loss": 0.685, + "step": 6789 + }, + { + "epoch": 0.6123461243630789, + "grad_norm": 2.329854949725728, + "learning_rate": 1.3797439643286227e-06, + "loss": 0.7118, + "step": 6790 + }, + { + "epoch": 0.6124363078865491, + "grad_norm": 2.0746410105294846, + "learning_rate": 1.3791885847889277e-06, + "loss": 0.7358, + "step": 6791 + }, + { + "epoch": 0.6125264914100194, + "grad_norm": 1.5678602210528876, + "learning_rate": 1.3786332582202853e-06, + "loss": 0.7902, + "step": 6792 + }, + { + "epoch": 0.6126166749334897, + "grad_norm": 1.6612860930162274, + "learning_rate": 1.3780779846700799e-06, + "loss": 0.7505, + "step": 6793 + }, + { + "epoch": 0.6127068584569599, + "grad_norm": 1.497774221184643, + "learning_rate": 1.3775227641856899e-06, + "loss": 0.7633, + "step": 6794 + }, + { + "epoch": 0.6127970419804302, + "grad_norm": 1.6949937230671925, + "learning_rate": 1.37696759681449e-06, + "loss": 0.7513, + "step": 6795 + }, + { + "epoch": 0.6128872255039004, + "grad_norm": 0.8304401640399627, + "learning_rate": 1.37641248260385e-06, + "loss": 0.6073, + "step": 6796 + }, + { + "epoch": 0.6129774090273707, + "grad_norm": 2.475249799972848, + "learning_rate": 1.375857421601136e-06, + "loss": 0.8008, + "step": 6797 + }, + { + "epoch": 0.613067592550841, + "grad_norm": 1.671052819705325, + "learning_rate": 1.3753024138537082e-06, + "loss": 0.7603, + "step": 6798 + }, + { + "epoch": 0.6131577760743112, + "grad_norm": 1.5926316769475446, + "learning_rate": 1.3747474594089221e-06, + "loss": 0.8187, + "step": 6799 + }, + { + "epoch": 0.6132479595977814, + "grad_norm": 1.775575265211368, + "learning_rate": 1.374192558314131e-06, + "loss": 0.733, + "step": 6800 + }, + { + "epoch": 0.6133381431212518, + "grad_norm": 1.719625492797701, + "learning_rate": 1.373637710616681e-06, + "loss": 0.8161, + "step": 6801 + }, + { + "epoch": 0.613428326644722, + "grad_norm": 3.9793724819035785, + "learning_rate": 1.373082916363916e-06, + "loss": 0.8647, + "step": 6802 + }, + { + "epoch": 0.6135185101681923, + "grad_norm": 1.9896753965584546, + "learning_rate": 1.3725281756031732e-06, + "loss": 0.7957, + "step": 6803 + }, + { + "epoch": 0.6136086936916625, + "grad_norm": 1.8222353361337598, + "learning_rate": 1.3719734883817858e-06, + "loss": 0.7536, + "step": 6804 + }, + { + "epoch": 0.6136988772151328, + "grad_norm": 1.6613702621157054, + "learning_rate": 1.371418854747084e-06, + "loss": 0.8682, + "step": 6805 + }, + { + "epoch": 0.6137890607386031, + "grad_norm": 1.9482057190352462, + "learning_rate": 1.3708642747463905e-06, + "loss": 0.7876, + "step": 6806 + }, + { + "epoch": 0.6138792442620733, + "grad_norm": 3.093002999798528, + "learning_rate": 1.370309748427027e-06, + "loss": 0.7811, + "step": 6807 + }, + { + "epoch": 0.6139694277855435, + "grad_norm": 1.618384024588633, + "learning_rate": 1.3697552758363079e-06, + "loss": 0.8281, + "step": 6808 + }, + { + "epoch": 0.6140596113090139, + "grad_norm": 4.328825415390266, + "learning_rate": 1.3692008570215432e-06, + "loss": 0.7576, + "step": 6809 + }, + { + "epoch": 0.6141497948324841, + "grad_norm": 1.8284190666848636, + "learning_rate": 1.3686464920300398e-06, + "loss": 0.7022, + "step": 6810 + }, + { + "epoch": 0.6142399783559543, + "grad_norm": 2.2461662479807933, + "learning_rate": 1.3680921809090985e-06, + "loss": 0.7339, + "step": 6811 + }, + { + "epoch": 0.6143301618794246, + "grad_norm": 2.1711816572942366, + "learning_rate": 1.3675379237060175e-06, + "loss": 0.7979, + "step": 6812 + }, + { + "epoch": 0.6144203454028949, + "grad_norm": 1.7683153164251655, + "learning_rate": 1.366983720468088e-06, + "loss": 0.7448, + "step": 6813 + }, + { + "epoch": 0.6145105289263652, + "grad_norm": 2.9787933720397772, + "learning_rate": 1.3664295712425972e-06, + "loss": 0.7566, + "step": 6814 + }, + { + "epoch": 0.6146007124498354, + "grad_norm": 1.7486297828957407, + "learning_rate": 1.3658754760768296e-06, + "loss": 0.6912, + "step": 6815 + }, + { + "epoch": 0.6146908959733057, + "grad_norm": 1.9833611128435031, + "learning_rate": 1.3653214350180621e-06, + "loss": 0.785, + "step": 6816 + }, + { + "epoch": 0.614781079496776, + "grad_norm": 1.5443198939780978, + "learning_rate": 1.3647674481135703e-06, + "loss": 0.8189, + "step": 6817 + }, + { + "epoch": 0.6148712630202462, + "grad_norm": 5.202166128155343, + "learning_rate": 1.3642135154106217e-06, + "loss": 0.652, + "step": 6818 + }, + { + "epoch": 0.6149614465437164, + "grad_norm": 2.0783600638393143, + "learning_rate": 1.363659636956482e-06, + "loss": 0.7019, + "step": 6819 + }, + { + "epoch": 0.6150516300671868, + "grad_norm": 2.0113470530401854, + "learning_rate": 1.3631058127984112e-06, + "loss": 0.7218, + "step": 6820 + }, + { + "epoch": 0.615141813590657, + "grad_norm": 5.819236215504225, + "learning_rate": 1.3625520429836632e-06, + "loss": 0.7331, + "step": 6821 + }, + { + "epoch": 0.6152319971141272, + "grad_norm": 1.755721955655647, + "learning_rate": 1.361998327559491e-06, + "loss": 0.6626, + "step": 6822 + }, + { + "epoch": 0.6153221806375975, + "grad_norm": 0.6558529242902088, + "learning_rate": 1.3614446665731385e-06, + "loss": 0.5813, + "step": 6823 + }, + { + "epoch": 0.6154123641610678, + "grad_norm": 1.7364255450867339, + "learning_rate": 1.3608910600718484e-06, + "loss": 0.7655, + "step": 6824 + }, + { + "epoch": 0.615502547684538, + "grad_norm": 1.6665326620907577, + "learning_rate": 1.360337508102857e-06, + "loss": 0.7587, + "step": 6825 + }, + { + "epoch": 0.6155927312080083, + "grad_norm": 2.49146004227289, + "learning_rate": 1.3597840107133962e-06, + "loss": 0.8535, + "step": 6826 + }, + { + "epoch": 0.6156829147314785, + "grad_norm": 1.905363783656887, + "learning_rate": 1.3592305679506944e-06, + "loss": 0.7595, + "step": 6827 + }, + { + "epoch": 0.6157730982549489, + "grad_norm": 1.4324269548736452, + "learning_rate": 1.3586771798619726e-06, + "loss": 0.6588, + "step": 6828 + }, + { + "epoch": 0.6158632817784191, + "grad_norm": 1.5274982813249525, + "learning_rate": 1.358123846494451e-06, + "loss": 0.8134, + "step": 6829 + }, + { + "epoch": 0.6159534653018893, + "grad_norm": 2.316751438300552, + "learning_rate": 1.3575705678953422e-06, + "loss": 0.761, + "step": 6830 + }, + { + "epoch": 0.6160436488253596, + "grad_norm": 1.9160413827827125, + "learning_rate": 1.357017344111854e-06, + "loss": 0.7945, + "step": 6831 + }, + { + "epoch": 0.6161338323488299, + "grad_norm": 1.725054397278163, + "learning_rate": 1.356464175191192e-06, + "loss": 0.7657, + "step": 6832 + }, + { + "epoch": 0.6162240158723001, + "grad_norm": 1.8160577911516507, + "learning_rate": 1.3559110611805542e-06, + "loss": 0.7822, + "step": 6833 + }, + { + "epoch": 0.6163141993957704, + "grad_norm": 2.0895133354453392, + "learning_rate": 1.3553580021271372e-06, + "loss": 0.7936, + "step": 6834 + }, + { + "epoch": 0.6164043829192406, + "grad_norm": 1.4132603430269124, + "learning_rate": 1.3548049980781297e-06, + "loss": 0.6979, + "step": 6835 + }, + { + "epoch": 0.616494566442711, + "grad_norm": 1.6878634873733291, + "learning_rate": 1.3542520490807166e-06, + "loss": 0.7482, + "step": 6836 + }, + { + "epoch": 0.6165847499661812, + "grad_norm": 2.123686328312925, + "learning_rate": 1.3536991551820802e-06, + "loss": 0.7424, + "step": 6837 + }, + { + "epoch": 0.6166749334896514, + "grad_norm": 1.5040791994202507, + "learning_rate": 1.3531463164293952e-06, + "loss": 0.7245, + "step": 6838 + }, + { + "epoch": 0.6167651170131218, + "grad_norm": 2.1527073824731597, + "learning_rate": 1.3525935328698332e-06, + "loss": 0.6612, + "step": 6839 + }, + { + "epoch": 0.616855300536592, + "grad_norm": 1.9586666986782695, + "learning_rate": 1.3520408045505605e-06, + "loss": 0.6926, + "step": 6840 + }, + { + "epoch": 0.6169454840600622, + "grad_norm": 1.722423788635323, + "learning_rate": 1.3514881315187396e-06, + "loss": 0.7207, + "step": 6841 + }, + { + "epoch": 0.6170356675835325, + "grad_norm": 1.3814722236246015, + "learning_rate": 1.3509355138215273e-06, + "loss": 0.8137, + "step": 6842 + }, + { + "epoch": 0.6171258511070028, + "grad_norm": 2.134441829322132, + "learning_rate": 1.350382951506075e-06, + "loss": 0.835, + "step": 6843 + }, + { + "epoch": 0.617216034630473, + "grad_norm": 0.6408604445094525, + "learning_rate": 1.349830444619532e-06, + "loss": 0.5677, + "step": 6844 + }, + { + "epoch": 0.6173062181539433, + "grad_norm": 2.176448589556431, + "learning_rate": 1.3492779932090397e-06, + "loss": 0.7697, + "step": 6845 + }, + { + "epoch": 0.6173964016774135, + "grad_norm": 1.7334991855624926, + "learning_rate": 1.3487255973217377e-06, + "loss": 0.7793, + "step": 6846 + }, + { + "epoch": 0.6174865852008838, + "grad_norm": 1.6094284048756353, + "learning_rate": 1.3481732570047592e-06, + "loss": 0.8206, + "step": 6847 + }, + { + "epoch": 0.6175767687243541, + "grad_norm": 4.666576838764133, + "learning_rate": 1.3476209723052314e-06, + "loss": 0.6956, + "step": 6848 + }, + { + "epoch": 0.6176669522478243, + "grad_norm": 1.851117945873826, + "learning_rate": 1.3470687432702806e-06, + "loss": 0.777, + "step": 6849 + }, + { + "epoch": 0.6177571357712945, + "grad_norm": 2.0558614531862953, + "learning_rate": 1.346516569947024e-06, + "loss": 0.6886, + "step": 6850 + }, + { + "epoch": 0.6178473192947649, + "grad_norm": 1.701773080677411, + "learning_rate": 1.3459644523825774e-06, + "loss": 0.7396, + "step": 6851 + }, + { + "epoch": 0.6179375028182351, + "grad_norm": 1.9289131000655189, + "learning_rate": 1.34541239062405e-06, + "loss": 0.799, + "step": 6852 + }, + { + "epoch": 0.6180276863417054, + "grad_norm": 2.6730623339839132, + "learning_rate": 1.3448603847185464e-06, + "loss": 0.7516, + "step": 6853 + }, + { + "epoch": 0.6181178698651756, + "grad_norm": 1.443214089149545, + "learning_rate": 1.344308434713168e-06, + "loss": 0.695, + "step": 6854 + }, + { + "epoch": 0.6182080533886459, + "grad_norm": 2.2147572011032874, + "learning_rate": 1.3437565406550083e-06, + "loss": 0.6898, + "step": 6855 + }, + { + "epoch": 0.6182982369121162, + "grad_norm": 2.1632845475270552, + "learning_rate": 1.34320470259116e-06, + "loss": 0.7397, + "step": 6856 + }, + { + "epoch": 0.6183884204355864, + "grad_norm": 2.186377973262577, + "learning_rate": 1.3426529205687078e-06, + "loss": 0.6581, + "step": 6857 + }, + { + "epoch": 0.6184786039590566, + "grad_norm": 1.7206623882273517, + "learning_rate": 1.3421011946347323e-06, + "loss": 0.8096, + "step": 6858 + }, + { + "epoch": 0.618568787482527, + "grad_norm": 1.7212015181440554, + "learning_rate": 1.3415495248363113e-06, + "loss": 0.7164, + "step": 6859 + }, + { + "epoch": 0.6186589710059972, + "grad_norm": 2.109070819539324, + "learning_rate": 1.3409979112205148e-06, + "loss": 0.7752, + "step": 6860 + }, + { + "epoch": 0.6187491545294674, + "grad_norm": 6.792563849167672, + "learning_rate": 1.3404463538344107e-06, + "loss": 0.7786, + "step": 6861 + }, + { + "epoch": 0.6188393380529378, + "grad_norm": 2.285360108762116, + "learning_rate": 1.3398948527250602e-06, + "loss": 0.8406, + "step": 6862 + }, + { + "epoch": 0.618929521576408, + "grad_norm": 0.785658376890602, + "learning_rate": 1.3393434079395207e-06, + "loss": 0.5923, + "step": 6863 + }, + { + "epoch": 0.6190197050998782, + "grad_norm": 2.0940516240370197, + "learning_rate": 1.3387920195248449e-06, + "loss": 0.7423, + "step": 6864 + }, + { + "epoch": 0.6191098886233485, + "grad_norm": 1.5459748548656382, + "learning_rate": 1.3382406875280791e-06, + "loss": 0.7679, + "step": 6865 + }, + { + "epoch": 0.6192000721468188, + "grad_norm": 1.581432615448018, + "learning_rate": 1.3376894119962672e-06, + "loss": 0.7755, + "step": 6866 + }, + { + "epoch": 0.6192902556702891, + "grad_norm": 2.242198026944746, + "learning_rate": 1.3371381929764464e-06, + "loss": 0.7786, + "step": 6867 + }, + { + "epoch": 0.6193804391937593, + "grad_norm": 0.6362661145311125, + "learning_rate": 1.3365870305156506e-06, + "loss": 0.57, + "step": 6868 + }, + { + "epoch": 0.6194706227172295, + "grad_norm": 2.4686183652912757, + "learning_rate": 1.3360359246609073e-06, + "loss": 0.7646, + "step": 6869 + }, + { + "epoch": 0.6195608062406999, + "grad_norm": 1.7359786996149946, + "learning_rate": 1.3354848754592387e-06, + "loss": 0.7538, + "step": 6870 + }, + { + "epoch": 0.6196509897641701, + "grad_norm": 1.6659600985398053, + "learning_rate": 1.334933882957666e-06, + "loss": 0.8407, + "step": 6871 + }, + { + "epoch": 0.6197411732876403, + "grad_norm": 1.4494855691665813, + "learning_rate": 1.3343829472032004e-06, + "loss": 0.7181, + "step": 6872 + }, + { + "epoch": 0.6198313568111106, + "grad_norm": 2.2215190317942963, + "learning_rate": 1.3338320682428527e-06, + "loss": 0.7423, + "step": 6873 + }, + { + "epoch": 0.6199215403345809, + "grad_norm": 1.487506895289748, + "learning_rate": 1.3332812461236263e-06, + "loss": 0.7144, + "step": 6874 + }, + { + "epoch": 0.6200117238580511, + "grad_norm": 1.4214624498085837, + "learning_rate": 1.3327304808925192e-06, + "loss": 0.7202, + "step": 6875 + }, + { + "epoch": 0.6201019073815214, + "grad_norm": 2.2689931103470204, + "learning_rate": 1.332179772596528e-06, + "loss": 0.6169, + "step": 6876 + }, + { + "epoch": 0.6201920909049916, + "grad_norm": 1.9891547961520766, + "learning_rate": 1.3316291212826402e-06, + "loss": 0.6774, + "step": 6877 + }, + { + "epoch": 0.620282274428462, + "grad_norm": 2.484630835453985, + "learning_rate": 1.3310785269978413e-06, + "loss": 0.8064, + "step": 6878 + }, + { + "epoch": 0.6203724579519322, + "grad_norm": 1.7143143725270413, + "learning_rate": 1.3305279897891111e-06, + "loss": 0.799, + "step": 6879 + }, + { + "epoch": 0.6204626414754024, + "grad_norm": 1.935555189559431, + "learning_rate": 1.329977509703424e-06, + "loss": 0.617, + "step": 6880 + }, + { + "epoch": 0.6205528249988727, + "grad_norm": 1.815757061198372, + "learning_rate": 1.32942708678775e-06, + "loss": 0.8414, + "step": 6881 + }, + { + "epoch": 0.620643008522343, + "grad_norm": 1.7175476723930398, + "learning_rate": 1.3288767210890548e-06, + "loss": 0.7655, + "step": 6882 + }, + { + "epoch": 0.6207331920458132, + "grad_norm": 1.7044818617778004, + "learning_rate": 1.3283264126542986e-06, + "loss": 0.7162, + "step": 6883 + }, + { + "epoch": 0.6208233755692835, + "grad_norm": 1.5726298982370974, + "learning_rate": 1.3277761615304356e-06, + "loss": 0.7544, + "step": 6884 + }, + { + "epoch": 0.6209135590927538, + "grad_norm": 1.4694247217823815, + "learning_rate": 1.3272259677644178e-06, + "loss": 0.8529, + "step": 6885 + }, + { + "epoch": 0.621003742616224, + "grad_norm": 1.771984832965705, + "learning_rate": 1.32667583140319e-06, + "loss": 0.6272, + "step": 6886 + }, + { + "epoch": 0.6210939261396943, + "grad_norm": 2.5441763218349656, + "learning_rate": 1.3261257524936924e-06, + "loss": 0.7986, + "step": 6887 + }, + { + "epoch": 0.6211841096631645, + "grad_norm": 1.4537822918573193, + "learning_rate": 1.3255757310828619e-06, + "loss": 0.787, + "step": 6888 + }, + { + "epoch": 0.6212742931866349, + "grad_norm": 3.1243296094887345, + "learning_rate": 1.3250257672176282e-06, + "loss": 0.7762, + "step": 6889 + }, + { + "epoch": 0.6213644767101051, + "grad_norm": 1.7028131461951728, + "learning_rate": 1.3244758609449183e-06, + "loss": 0.6593, + "step": 6890 + }, + { + "epoch": 0.6214546602335753, + "grad_norm": 2.0340583754108086, + "learning_rate": 1.323926012311653e-06, + "loss": 0.7691, + "step": 6891 + }, + { + "epoch": 0.6215448437570456, + "grad_norm": 2.865548561932081, + "learning_rate": 1.3233762213647476e-06, + "loss": 0.7831, + "step": 6892 + }, + { + "epoch": 0.6216350272805159, + "grad_norm": 1.6819620035803535, + "learning_rate": 1.3228264881511137e-06, + "loss": 0.7301, + "step": 6893 + }, + { + "epoch": 0.6217252108039861, + "grad_norm": 2.376156575525092, + "learning_rate": 1.322276812717658e-06, + "loss": 0.8386, + "step": 6894 + }, + { + "epoch": 0.6218153943274564, + "grad_norm": 1.5362006364655936, + "learning_rate": 1.3217271951112818e-06, + "loss": 0.6562, + "step": 6895 + }, + { + "epoch": 0.6219055778509266, + "grad_norm": 11.431560370735292, + "learning_rate": 1.321177635378881e-06, + "loss": 0.7341, + "step": 6896 + }, + { + "epoch": 0.6219957613743969, + "grad_norm": 2.1109833413771937, + "learning_rate": 1.3206281335673475e-06, + "loss": 0.6528, + "step": 6897 + }, + { + "epoch": 0.6220859448978672, + "grad_norm": 1.9980459203932226, + "learning_rate": 1.3200786897235677e-06, + "loss": 0.728, + "step": 6898 + }, + { + "epoch": 0.6221761284213374, + "grad_norm": 1.7318021231288827, + "learning_rate": 1.3195293038944227e-06, + "loss": 0.7725, + "step": 6899 + }, + { + "epoch": 0.6222663119448076, + "grad_norm": 3.4737279416180433, + "learning_rate": 1.3189799761267902e-06, + "loss": 0.7478, + "step": 6900 + }, + { + "epoch": 0.622356495468278, + "grad_norm": 2.3873512940756045, + "learning_rate": 1.3184307064675412e-06, + "loss": 0.7479, + "step": 6901 + }, + { + "epoch": 0.6224466789917482, + "grad_norm": 0.7069838070626991, + "learning_rate": 1.3178814949635416e-06, + "loss": 0.6249, + "step": 6902 + }, + { + "epoch": 0.6225368625152184, + "grad_norm": 1.8758556956524841, + "learning_rate": 1.3173323416616549e-06, + "loss": 0.7078, + "step": 6903 + }, + { + "epoch": 0.6226270460386887, + "grad_norm": 1.5829282878312914, + "learning_rate": 1.3167832466087361e-06, + "loss": 0.8033, + "step": 6904 + }, + { + "epoch": 0.622717229562159, + "grad_norm": 1.3307213929229835, + "learning_rate": 1.3162342098516388e-06, + "loss": 0.6988, + "step": 6905 + }, + { + "epoch": 0.6228074130856293, + "grad_norm": 1.5254432971960004, + "learning_rate": 1.3156852314372086e-06, + "loss": 0.6712, + "step": 6906 + }, + { + "epoch": 0.6228975966090995, + "grad_norm": 0.6836749873255576, + "learning_rate": 1.3151363114122882e-06, + "loss": 0.5619, + "step": 6907 + }, + { + "epoch": 0.6229877801325698, + "grad_norm": 1.863839503696974, + "learning_rate": 1.3145874498237133e-06, + "loss": 0.767, + "step": 6908 + }, + { + "epoch": 0.6230779636560401, + "grad_norm": 1.655283952077593, + "learning_rate": 1.3140386467183166e-06, + "loss": 0.8005, + "step": 6909 + }, + { + "epoch": 0.6231681471795103, + "grad_norm": 5.523614818527098, + "learning_rate": 1.3134899021429258e-06, + "loss": 0.796, + "step": 6910 + }, + { + "epoch": 0.6232583307029805, + "grad_norm": 4.376454356831196, + "learning_rate": 1.3129412161443604e-06, + "loss": 0.7323, + "step": 6911 + }, + { + "epoch": 0.6233485142264509, + "grad_norm": 1.797457326984686, + "learning_rate": 1.3123925887694402e-06, + "loss": 0.7636, + "step": 6912 + }, + { + "epoch": 0.6234386977499211, + "grad_norm": 2.1443681498556817, + "learning_rate": 1.3118440200649756e-06, + "loss": 0.7618, + "step": 6913 + }, + { + "epoch": 0.6235288812733913, + "grad_norm": 6.907174027020804, + "learning_rate": 1.3112955100777727e-06, + "loss": 0.7141, + "step": 6914 + }, + { + "epoch": 0.6236190647968616, + "grad_norm": 1.3919557354948564, + "learning_rate": 1.3107470588546353e-06, + "loss": 0.7859, + "step": 6915 + }, + { + "epoch": 0.6237092483203319, + "grad_norm": 2.4779961224172333, + "learning_rate": 1.3101986664423583e-06, + "loss": 0.7227, + "step": 6916 + }, + { + "epoch": 0.6237994318438022, + "grad_norm": 2.070026421124927, + "learning_rate": 1.3096503328877358e-06, + "loss": 0.728, + "step": 6917 + }, + { + "epoch": 0.6238896153672724, + "grad_norm": 1.9359027061999563, + "learning_rate": 1.309102058237553e-06, + "loss": 0.7713, + "step": 6918 + }, + { + "epoch": 0.6239797988907426, + "grad_norm": 1.4332972747495774, + "learning_rate": 1.3085538425385917e-06, + "loss": 0.8037, + "step": 6919 + }, + { + "epoch": 0.624069982414213, + "grad_norm": 1.8787389986143768, + "learning_rate": 1.3080056858376298e-06, + "loss": 0.7831, + "step": 6920 + }, + { + "epoch": 0.6241601659376832, + "grad_norm": 2.052481856271745, + "learning_rate": 1.3074575881814383e-06, + "loss": 0.7389, + "step": 6921 + }, + { + "epoch": 0.6242503494611534, + "grad_norm": 1.9574930403818325, + "learning_rate": 1.3069095496167841e-06, + "loss": 0.6787, + "step": 6922 + }, + { + "epoch": 0.6243405329846237, + "grad_norm": 1.482121574130864, + "learning_rate": 1.3063615701904285e-06, + "loss": 0.6614, + "step": 6923 + }, + { + "epoch": 0.624430716508094, + "grad_norm": 1.8214173463534293, + "learning_rate": 1.3058136499491283e-06, + "loss": 0.7849, + "step": 6924 + }, + { + "epoch": 0.6245209000315642, + "grad_norm": 2.6457755515529953, + "learning_rate": 1.3052657889396357e-06, + "loss": 0.7371, + "step": 6925 + }, + { + "epoch": 0.6246110835550345, + "grad_norm": 1.5761668832692797, + "learning_rate": 1.304717987208696e-06, + "loss": 0.7368, + "step": 6926 + }, + { + "epoch": 0.6247012670785047, + "grad_norm": 2.9996355153239, + "learning_rate": 1.304170244803052e-06, + "loss": 0.7393, + "step": 6927 + }, + { + "epoch": 0.624791450601975, + "grad_norm": 2.6253476593856293, + "learning_rate": 1.3036225617694387e-06, + "loss": 0.7236, + "step": 6928 + }, + { + "epoch": 0.6248816341254453, + "grad_norm": 3.072836578532399, + "learning_rate": 1.3030749381545892e-06, + "loss": 0.781, + "step": 6929 + }, + { + "epoch": 0.6249718176489155, + "grad_norm": 1.792012477001112, + "learning_rate": 1.3025273740052285e-06, + "loss": 0.692, + "step": 6930 + }, + { + "epoch": 0.6250620011723858, + "grad_norm": 1.4710260976347378, + "learning_rate": 1.3019798693680774e-06, + "loss": 0.8006, + "step": 6931 + }, + { + "epoch": 0.6251521846958561, + "grad_norm": 1.6112758188716763, + "learning_rate": 1.3014324242898536e-06, + "loss": 0.8694, + "step": 6932 + }, + { + "epoch": 0.6252423682193263, + "grad_norm": 2.0294601771246135, + "learning_rate": 1.3008850388172668e-06, + "loss": 0.8109, + "step": 6933 + }, + { + "epoch": 0.6253325517427966, + "grad_norm": 1.6852929552820457, + "learning_rate": 1.3003377129970233e-06, + "loss": 0.7106, + "step": 6934 + }, + { + "epoch": 0.6254227352662669, + "grad_norm": 1.390881889703999, + "learning_rate": 1.2997904468758243e-06, + "loss": 0.6711, + "step": 6935 + }, + { + "epoch": 0.6255129187897371, + "grad_norm": 1.620223508062919, + "learning_rate": 1.299243240500365e-06, + "loss": 0.772, + "step": 6936 + }, + { + "epoch": 0.6256031023132074, + "grad_norm": 1.818724500339726, + "learning_rate": 1.2986960939173368e-06, + "loss": 0.7697, + "step": 6937 + }, + { + "epoch": 0.6256932858366776, + "grad_norm": 1.7365725230486682, + "learning_rate": 1.298149007173424e-06, + "loss": 0.798, + "step": 6938 + }, + { + "epoch": 0.625783469360148, + "grad_norm": 1.9514663367176706, + "learning_rate": 1.2976019803153087e-06, + "loss": 0.7846, + "step": 6939 + }, + { + "epoch": 0.6258736528836182, + "grad_norm": 1.6481137070686636, + "learning_rate": 1.2970550133896652e-06, + "loss": 0.7001, + "step": 6940 + }, + { + "epoch": 0.6259638364070884, + "grad_norm": 1.5796699892728086, + "learning_rate": 1.2965081064431634e-06, + "loss": 0.8351, + "step": 6941 + }, + { + "epoch": 0.6260540199305586, + "grad_norm": 2.12807108044762, + "learning_rate": 1.2959612595224698e-06, + "loss": 0.7779, + "step": 6942 + }, + { + "epoch": 0.626144203454029, + "grad_norm": 2.0882418473590465, + "learning_rate": 1.2954144726742424e-06, + "loss": 0.7637, + "step": 6943 + }, + { + "epoch": 0.6262343869774992, + "grad_norm": 1.4925271491515957, + "learning_rate": 1.2948677459451385e-06, + "loss": 0.7496, + "step": 6944 + }, + { + "epoch": 0.6263245705009695, + "grad_norm": 1.8620513759740809, + "learning_rate": 1.2943210793818064e-06, + "loss": 0.8002, + "step": 6945 + }, + { + "epoch": 0.6264147540244397, + "grad_norm": 1.8408651250347854, + "learning_rate": 1.2937744730308899e-06, + "loss": 0.7345, + "step": 6946 + }, + { + "epoch": 0.62650493754791, + "grad_norm": 4.498975849169036, + "learning_rate": 1.2932279269390305e-06, + "loss": 0.7544, + "step": 6947 + }, + { + "epoch": 0.6265951210713803, + "grad_norm": 1.723372814247636, + "learning_rate": 1.292681441152861e-06, + "loss": 0.6576, + "step": 6948 + }, + { + "epoch": 0.6266853045948505, + "grad_norm": 1.7336391228148658, + "learning_rate": 1.292135015719011e-06, + "loss": 0.762, + "step": 6949 + }, + { + "epoch": 0.6267754881183207, + "grad_norm": 2.0884089883595918, + "learning_rate": 1.2915886506841046e-06, + "loss": 0.7139, + "step": 6950 + }, + { + "epoch": 0.6268656716417911, + "grad_norm": 1.5145177040765498, + "learning_rate": 1.2910423460947613e-06, + "loss": 0.6908, + "step": 6951 + }, + { + "epoch": 0.6269558551652613, + "grad_norm": 1.785055746175036, + "learning_rate": 1.290496101997594e-06, + "loss": 0.7231, + "step": 6952 + }, + { + "epoch": 0.6270460386887315, + "grad_norm": 0.7374868812040796, + "learning_rate": 1.2899499184392105e-06, + "loss": 0.5818, + "step": 6953 + }, + { + "epoch": 0.6271362222122018, + "grad_norm": 1.6734154280374318, + "learning_rate": 1.289403795466216e-06, + "loss": 0.7276, + "step": 6954 + }, + { + "epoch": 0.6272264057356721, + "grad_norm": 1.9229263283379028, + "learning_rate": 1.288857733125207e-06, + "loss": 0.7661, + "step": 6955 + }, + { + "epoch": 0.6273165892591424, + "grad_norm": 3.1669575574347792, + "learning_rate": 1.2883117314627785e-06, + "loss": 0.7482, + "step": 6956 + }, + { + "epoch": 0.6274067727826126, + "grad_norm": 1.6772819859469308, + "learning_rate": 1.2877657905255168e-06, + "loss": 0.6952, + "step": 6957 + }, + { + "epoch": 0.6274969563060829, + "grad_norm": 1.6308994311136227, + "learning_rate": 1.2872199103600046e-06, + "loss": 0.8126, + "step": 6958 + }, + { + "epoch": 0.6275871398295532, + "grad_norm": 1.5457519153359145, + "learning_rate": 1.286674091012821e-06, + "loss": 0.8223, + "step": 6959 + }, + { + "epoch": 0.6276773233530234, + "grad_norm": 1.5540929069436213, + "learning_rate": 1.2861283325305356e-06, + "loss": 0.8083, + "step": 6960 + }, + { + "epoch": 0.6277675068764936, + "grad_norm": 1.4331470593943623, + "learning_rate": 1.2855826349597185e-06, + "loss": 0.7521, + "step": 6961 + }, + { + "epoch": 0.627857690399964, + "grad_norm": 1.355082871975011, + "learning_rate": 1.2850369983469302e-06, + "loss": 0.6071, + "step": 6962 + }, + { + "epoch": 0.6279478739234342, + "grad_norm": 3.685090263638011, + "learning_rate": 1.2844914227387266e-06, + "loss": 0.7791, + "step": 6963 + }, + { + "epoch": 0.6280380574469044, + "grad_norm": 8.672398456892832, + "learning_rate": 1.2839459081816606e-06, + "loss": 0.7744, + "step": 6964 + }, + { + "epoch": 0.6281282409703747, + "grad_norm": 1.8052702124002553, + "learning_rate": 1.283400454722278e-06, + "loss": 0.8182, + "step": 6965 + }, + { + "epoch": 0.628218424493845, + "grad_norm": 2.0710377700360203, + "learning_rate": 1.28285506240712e-06, + "loss": 0.8115, + "step": 6966 + }, + { + "epoch": 0.6283086080173153, + "grad_norm": 1.590420854879609, + "learning_rate": 1.2823097312827225e-06, + "loss": 0.7772, + "step": 6967 + }, + { + "epoch": 0.6283987915407855, + "grad_norm": 3.083559921461817, + "learning_rate": 1.2817644613956153e-06, + "loss": 0.7632, + "step": 6968 + }, + { + "epoch": 0.6284889750642557, + "grad_norm": 3.2399370955023956, + "learning_rate": 1.2812192527923253e-06, + "loss": 0.8175, + "step": 6969 + }, + { + "epoch": 0.6285791585877261, + "grad_norm": 1.79843821322562, + "learning_rate": 1.2806741055193712e-06, + "loss": 0.7888, + "step": 6970 + }, + { + "epoch": 0.6286693421111963, + "grad_norm": 1.873621115702179, + "learning_rate": 1.2801290196232695e-06, + "loss": 0.7109, + "step": 6971 + }, + { + "epoch": 0.6287595256346665, + "grad_norm": 1.7901840359366918, + "learning_rate": 1.2795839951505282e-06, + "loss": 0.7434, + "step": 6972 + }, + { + "epoch": 0.6288497091581368, + "grad_norm": 6.927886075041217, + "learning_rate": 1.2790390321476538e-06, + "loss": 0.7715, + "step": 6973 + }, + { + "epoch": 0.6289398926816071, + "grad_norm": 1.552975656248596, + "learning_rate": 1.2784941306611446e-06, + "loss": 0.7216, + "step": 6974 + }, + { + "epoch": 0.6290300762050773, + "grad_norm": 1.7941887056538874, + "learning_rate": 1.2779492907374935e-06, + "loss": 0.8136, + "step": 6975 + }, + { + "epoch": 0.6291202597285476, + "grad_norm": 1.8442652354190923, + "learning_rate": 1.2774045124231911e-06, + "loss": 0.7426, + "step": 6976 + }, + { + "epoch": 0.6292104432520178, + "grad_norm": 1.8942270565629626, + "learning_rate": 1.2768597957647197e-06, + "loss": 0.6352, + "step": 6977 + }, + { + "epoch": 0.6293006267754881, + "grad_norm": 1.8481778140579705, + "learning_rate": 1.2763151408085582e-06, + "loss": 0.7631, + "step": 6978 + }, + { + "epoch": 0.6293908102989584, + "grad_norm": 2.2465703044038916, + "learning_rate": 1.2757705476011788e-06, + "loss": 0.7191, + "step": 6979 + }, + { + "epoch": 0.6294809938224286, + "grad_norm": 1.6273074273768313, + "learning_rate": 1.27522601618905e-06, + "loss": 0.6971, + "step": 6980 + }, + { + "epoch": 0.629571177345899, + "grad_norm": 0.5689746109254185, + "learning_rate": 1.2746815466186337e-06, + "loss": 0.6046, + "step": 6981 + }, + { + "epoch": 0.6296613608693692, + "grad_norm": 2.076220794417205, + "learning_rate": 1.274137138936387e-06, + "loss": 0.7528, + "step": 6982 + }, + { + "epoch": 0.6297515443928394, + "grad_norm": 1.8769298794167337, + "learning_rate": 1.2735927931887625e-06, + "loss": 0.7521, + "step": 6983 + }, + { + "epoch": 0.6298417279163097, + "grad_norm": 0.6938080746167903, + "learning_rate": 1.2730485094222061e-06, + "loss": 0.5864, + "step": 6984 + }, + { + "epoch": 0.62993191143978, + "grad_norm": 2.0156815395261773, + "learning_rate": 1.2725042876831586e-06, + "loss": 0.7889, + "step": 6985 + }, + { + "epoch": 0.6300220949632502, + "grad_norm": 2.090528942689718, + "learning_rate": 1.2719601280180573e-06, + "loss": 0.6493, + "step": 6986 + }, + { + "epoch": 0.6301122784867205, + "grad_norm": 1.5229258925353732, + "learning_rate": 1.2714160304733317e-06, + "loss": 0.8168, + "step": 6987 + }, + { + "epoch": 0.6302024620101907, + "grad_norm": 2.219203419091548, + "learning_rate": 1.2708719950954082e-06, + "loss": 0.7645, + "step": 6988 + }, + { + "epoch": 0.630292645533661, + "grad_norm": 2.3129830757969976, + "learning_rate": 1.2703280219307065e-06, + "loss": 0.7754, + "step": 6989 + }, + { + "epoch": 0.6303828290571313, + "grad_norm": 1.6666035481263075, + "learning_rate": 1.2697841110256411e-06, + "loss": 0.6816, + "step": 6990 + }, + { + "epoch": 0.6304730125806015, + "grad_norm": 1.5141347263615528, + "learning_rate": 1.2692402624266221e-06, + "loss": 0.7958, + "step": 6991 + }, + { + "epoch": 0.6305631961040717, + "grad_norm": 1.8308193859228303, + "learning_rate": 1.2686964761800529e-06, + "loss": 0.6654, + "step": 6992 + }, + { + "epoch": 0.6306533796275421, + "grad_norm": 0.6368682311412425, + "learning_rate": 1.268152752332333e-06, + "loss": 0.5844, + "step": 6993 + }, + { + "epoch": 0.6307435631510123, + "grad_norm": 1.9018280771655505, + "learning_rate": 1.2676090909298549e-06, + "loss": 0.6633, + "step": 6994 + }, + { + "epoch": 0.6308337466744826, + "grad_norm": 1.9033315236449895, + "learning_rate": 1.2670654920190086e-06, + "loss": 0.803, + "step": 6995 + }, + { + "epoch": 0.6309239301979528, + "grad_norm": 1.3780174092271318, + "learning_rate": 1.2665219556461754e-06, + "loss": 0.8014, + "step": 6996 + }, + { + "epoch": 0.6310141137214231, + "grad_norm": 1.745213297293115, + "learning_rate": 1.2659784818577329e-06, + "loss": 0.8512, + "step": 6997 + }, + { + "epoch": 0.6311042972448934, + "grad_norm": 1.7907527618365604, + "learning_rate": 1.2654350707000542e-06, + "loss": 0.6348, + "step": 6998 + }, + { + "epoch": 0.6311944807683636, + "grad_norm": 1.60237934402221, + "learning_rate": 1.264891722219505e-06, + "loss": 0.7365, + "step": 6999 + }, + { + "epoch": 0.6312846642918338, + "grad_norm": 2.2329444840906283, + "learning_rate": 1.2643484364624483e-06, + "loss": 0.7483, + "step": 7000 + }, + { + "epoch": 0.6313748478153042, + "grad_norm": 1.7349106932199987, + "learning_rate": 1.2638052134752393e-06, + "loss": 0.8028, + "step": 7001 + }, + { + "epoch": 0.6314650313387744, + "grad_norm": 2.091622823531812, + "learning_rate": 1.2632620533042277e-06, + "loss": 0.8236, + "step": 7002 + }, + { + "epoch": 0.6315552148622446, + "grad_norm": 2.149130858465315, + "learning_rate": 1.2627189559957612e-06, + "loss": 0.7173, + "step": 7003 + }, + { + "epoch": 0.631645398385715, + "grad_norm": 2.2096014400485773, + "learning_rate": 1.262175921596178e-06, + "loss": 0.8118, + "step": 7004 + }, + { + "epoch": 0.6317355819091852, + "grad_norm": 1.8134152962355572, + "learning_rate": 1.2616329501518137e-06, + "loss": 0.6265, + "step": 7005 + }, + { + "epoch": 0.6318257654326555, + "grad_norm": 1.7771959180449752, + "learning_rate": 1.2610900417089978e-06, + "loss": 0.6735, + "step": 7006 + }, + { + "epoch": 0.6319159489561257, + "grad_norm": 1.7875795173297824, + "learning_rate": 1.2605471963140535e-06, + "loss": 0.7646, + "step": 7007 + }, + { + "epoch": 0.632006132479596, + "grad_norm": 2.230213905647036, + "learning_rate": 1.2600044140133e-06, + "loss": 0.8218, + "step": 7008 + }, + { + "epoch": 0.6320963160030663, + "grad_norm": 1.6734897979343295, + "learning_rate": 1.2594616948530493e-06, + "loss": 0.7025, + "step": 7009 + }, + { + "epoch": 0.6321864995265365, + "grad_norm": 1.9068668180507093, + "learning_rate": 1.258919038879611e-06, + "loss": 0.7736, + "step": 7010 + }, + { + "epoch": 0.6322766830500067, + "grad_norm": 1.7946278992377256, + "learning_rate": 1.2583764461392859e-06, + "loss": 0.7538, + "step": 7011 + }, + { + "epoch": 0.6323668665734771, + "grad_norm": 1.7452937604125236, + "learning_rate": 1.2578339166783724e-06, + "loss": 0.7124, + "step": 7012 + }, + { + "epoch": 0.6324570500969473, + "grad_norm": 1.6402992288405454, + "learning_rate": 1.2572914505431613e-06, + "loss": 0.6872, + "step": 7013 + }, + { + "epoch": 0.6325472336204175, + "grad_norm": 1.5200054929031706, + "learning_rate": 1.2567490477799383e-06, + "loss": 0.652, + "step": 7014 + }, + { + "epoch": 0.6326374171438878, + "grad_norm": 9.969867703431465, + "learning_rate": 1.2562067084349852e-06, + "loss": 0.8455, + "step": 7015 + }, + { + "epoch": 0.6327276006673581, + "grad_norm": 2.5714122889018363, + "learning_rate": 1.2556644325545764e-06, + "loss": 0.735, + "step": 7016 + }, + { + "epoch": 0.6328177841908283, + "grad_norm": 2.075258730648615, + "learning_rate": 1.255122220184983e-06, + "loss": 0.7298, + "step": 7017 + }, + { + "epoch": 0.6329079677142986, + "grad_norm": 2.639720874826933, + "learning_rate": 1.2545800713724694e-06, + "loss": 0.6619, + "step": 7018 + }, + { + "epoch": 0.6329981512377688, + "grad_norm": 1.4202475823568332, + "learning_rate": 1.254037986163294e-06, + "loss": 0.6217, + "step": 7019 + }, + { + "epoch": 0.6330883347612392, + "grad_norm": 1.74924398168074, + "learning_rate": 1.2534959646037104e-06, + "loss": 0.7127, + "step": 7020 + }, + { + "epoch": 0.6331785182847094, + "grad_norm": 1.4360105842529671, + "learning_rate": 1.2529540067399675e-06, + "loss": 0.7565, + "step": 7021 + }, + { + "epoch": 0.6332687018081796, + "grad_norm": 1.7039168030885588, + "learning_rate": 1.252412112618308e-06, + "loss": 0.6907, + "step": 7022 + }, + { + "epoch": 0.6333588853316499, + "grad_norm": 1.9622977960495342, + "learning_rate": 1.2518702822849696e-06, + "loss": 0.6388, + "step": 7023 + }, + { + "epoch": 0.6334490688551202, + "grad_norm": 2.094762301195709, + "learning_rate": 1.2513285157861831e-06, + "loss": 0.6967, + "step": 7024 + }, + { + "epoch": 0.6335392523785904, + "grad_norm": 1.5485413178747551, + "learning_rate": 1.2507868131681764e-06, + "loss": 0.7458, + "step": 7025 + }, + { + "epoch": 0.6336294359020607, + "grad_norm": 1.6528679487419415, + "learning_rate": 1.250245174477169e-06, + "loss": 0.7327, + "step": 7026 + }, + { + "epoch": 0.6337196194255309, + "grad_norm": 0.652347975848553, + "learning_rate": 1.2497035997593783e-06, + "loss": 0.5202, + "step": 7027 + }, + { + "epoch": 0.6338098029490012, + "grad_norm": 2.5663426128608964, + "learning_rate": 1.2491620890610135e-06, + "loss": 0.7499, + "step": 7028 + }, + { + "epoch": 0.6338999864724715, + "grad_norm": 1.492187020666235, + "learning_rate": 1.2486206424282788e-06, + "loss": 0.7336, + "step": 7029 + }, + { + "epoch": 0.6339901699959417, + "grad_norm": 1.6548370236135914, + "learning_rate": 1.2480792599073743e-06, + "loss": 0.807, + "step": 7030 + }, + { + "epoch": 0.634080353519412, + "grad_norm": 2.808588486880065, + "learning_rate": 1.247537941544493e-06, + "loss": 0.7898, + "step": 7031 + }, + { + "epoch": 0.6341705370428823, + "grad_norm": 2.9121745439346216, + "learning_rate": 1.2469966873858242e-06, + "loss": 0.7843, + "step": 7032 + }, + { + "epoch": 0.6342607205663525, + "grad_norm": 2.0359528313364734, + "learning_rate": 1.2464554974775496e-06, + "loss": 0.6762, + "step": 7033 + }, + { + "epoch": 0.6343509040898228, + "grad_norm": 1.843310934256047, + "learning_rate": 1.2459143718658474e-06, + "loss": 0.7497, + "step": 7034 + }, + { + "epoch": 0.6344410876132931, + "grad_norm": 1.752294398721299, + "learning_rate": 1.2453733105968886e-06, + "loss": 0.6856, + "step": 7035 + }, + { + "epoch": 0.6345312711367633, + "grad_norm": 0.610389403123852, + "learning_rate": 1.2448323137168399e-06, + "loss": 0.5861, + "step": 7036 + }, + { + "epoch": 0.6346214546602336, + "grad_norm": 1.5326532434758158, + "learning_rate": 1.2442913812718625e-06, + "loss": 0.6721, + "step": 7037 + }, + { + "epoch": 0.6347116381837038, + "grad_norm": 2.252646161755552, + "learning_rate": 1.2437505133081108e-06, + "loss": 0.736, + "step": 7038 + }, + { + "epoch": 0.6348018217071741, + "grad_norm": 1.6353139014869544, + "learning_rate": 1.2432097098717358e-06, + "loss": 0.7668, + "step": 7039 + }, + { + "epoch": 0.6348920052306444, + "grad_norm": 1.9795452425094313, + "learning_rate": 1.2426689710088813e-06, + "loss": 0.7536, + "step": 7040 + }, + { + "epoch": 0.6349821887541146, + "grad_norm": 1.7274402441352654, + "learning_rate": 1.2421282967656855e-06, + "loss": 0.8046, + "step": 7041 + }, + { + "epoch": 0.6350723722775848, + "grad_norm": 1.714554399483388, + "learning_rate": 1.2415876871882827e-06, + "loss": 0.779, + "step": 7042 + }, + { + "epoch": 0.6351625558010552, + "grad_norm": 0.6730348423232592, + "learning_rate": 1.2410471423227998e-06, + "loss": 0.6497, + "step": 7043 + }, + { + "epoch": 0.6352527393245254, + "grad_norm": 2.271609275841117, + "learning_rate": 1.24050666221536e-06, + "loss": 0.7413, + "step": 7044 + }, + { + "epoch": 0.6353429228479957, + "grad_norm": 1.585089771724895, + "learning_rate": 1.23996624691208e-06, + "loss": 0.8165, + "step": 7045 + }, + { + "epoch": 0.6354331063714659, + "grad_norm": 1.63675858988561, + "learning_rate": 1.2394258964590693e-06, + "loss": 0.7342, + "step": 7046 + }, + { + "epoch": 0.6355232898949362, + "grad_norm": 1.9180731657550119, + "learning_rate": 1.238885610902436e-06, + "loss": 0.698, + "step": 7047 + }, + { + "epoch": 0.6356134734184065, + "grad_norm": 2.7042207646848717, + "learning_rate": 1.2383453902882787e-06, + "loss": 0.7615, + "step": 7048 + }, + { + "epoch": 0.6357036569418767, + "grad_norm": 1.5108189628045283, + "learning_rate": 1.2378052346626927e-06, + "loss": 0.813, + "step": 7049 + }, + { + "epoch": 0.6357938404653469, + "grad_norm": 2.1116919331994772, + "learning_rate": 1.2372651440717665e-06, + "loss": 0.6754, + "step": 7050 + }, + { + "epoch": 0.6358840239888173, + "grad_norm": 1.8079170295742515, + "learning_rate": 1.236725118561584e-06, + "loss": 0.7845, + "step": 7051 + }, + { + "epoch": 0.6359742075122875, + "grad_norm": 1.356919054171047, + "learning_rate": 1.2361851581782232e-06, + "loss": 0.7499, + "step": 7052 + }, + { + "epoch": 0.6360643910357577, + "grad_norm": 1.7947827307031616, + "learning_rate": 1.2356452629677554e-06, + "loss": 0.8114, + "step": 7053 + }, + { + "epoch": 0.6361545745592281, + "grad_norm": 1.5082912179255334, + "learning_rate": 1.2351054329762494e-06, + "loss": 0.6842, + "step": 7054 + }, + { + "epoch": 0.6362447580826983, + "grad_norm": 1.5869493685976348, + "learning_rate": 1.2345656682497648e-06, + "loss": 0.7766, + "step": 7055 + }, + { + "epoch": 0.6363349416061685, + "grad_norm": 1.339388845196829, + "learning_rate": 1.2340259688343583e-06, + "loss": 0.7265, + "step": 7056 + }, + { + "epoch": 0.6364251251296388, + "grad_norm": 1.4901349395735333, + "learning_rate": 1.2334863347760803e-06, + "loss": 0.6455, + "step": 7057 + }, + { + "epoch": 0.6365153086531091, + "grad_norm": 0.6446439008813764, + "learning_rate": 1.2329467661209734e-06, + "loss": 0.5451, + "step": 7058 + }, + { + "epoch": 0.6366054921765794, + "grad_norm": 1.671470165259147, + "learning_rate": 1.2324072629150788e-06, + "loss": 0.7925, + "step": 7059 + }, + { + "epoch": 0.6366956757000496, + "grad_norm": 1.9327214903440357, + "learning_rate": 1.2318678252044287e-06, + "loss": 0.7956, + "step": 7060 + }, + { + "epoch": 0.6367858592235198, + "grad_norm": 2.128473248325064, + "learning_rate": 1.2313284530350512e-06, + "loss": 0.8073, + "step": 7061 + }, + { + "epoch": 0.6368760427469902, + "grad_norm": 1.7507674714158632, + "learning_rate": 1.230789146452969e-06, + "loss": 0.6833, + "step": 7062 + }, + { + "epoch": 0.6369662262704604, + "grad_norm": 2.0842742575301907, + "learning_rate": 1.2302499055041974e-06, + "loss": 0.7531, + "step": 7063 + }, + { + "epoch": 0.6370564097939306, + "grad_norm": 0.6695767728804219, + "learning_rate": 1.2297107302347488e-06, + "loss": 0.5825, + "step": 7064 + }, + { + "epoch": 0.6371465933174009, + "grad_norm": 1.8873876171886605, + "learning_rate": 1.2291716206906275e-06, + "loss": 0.8172, + "step": 7065 + }, + { + "epoch": 0.6372367768408712, + "grad_norm": 2.074470576572535, + "learning_rate": 1.2286325769178345e-06, + "loss": 0.8185, + "step": 7066 + }, + { + "epoch": 0.6373269603643414, + "grad_norm": 1.7005500576746415, + "learning_rate": 1.2280935989623633e-06, + "loss": 0.7808, + "step": 7067 + }, + { + "epoch": 0.6374171438878117, + "grad_norm": 1.8544909018722506, + "learning_rate": 1.2275546868702017e-06, + "loss": 0.6867, + "step": 7068 + }, + { + "epoch": 0.6375073274112819, + "grad_norm": 2.0538281866643358, + "learning_rate": 1.2270158406873341e-06, + "loss": 0.6873, + "step": 7069 + }, + { + "epoch": 0.6375975109347523, + "grad_norm": 1.8238308675928794, + "learning_rate": 1.2264770604597363e-06, + "loss": 0.7897, + "step": 7070 + }, + { + "epoch": 0.6376876944582225, + "grad_norm": 1.6743441735449032, + "learning_rate": 1.2259383462333819e-06, + "loss": 0.8519, + "step": 7071 + }, + { + "epoch": 0.6377778779816927, + "grad_norm": 1.9173615964951622, + "learning_rate": 1.2253996980542359e-06, + "loss": 0.7008, + "step": 7072 + }, + { + "epoch": 0.637868061505163, + "grad_norm": 1.8528797598507902, + "learning_rate": 1.2248611159682578e-06, + "loss": 0.733, + "step": 7073 + }, + { + "epoch": 0.6379582450286333, + "grad_norm": 0.7354733390332661, + "learning_rate": 1.2243226000214044e-06, + "loss": 0.5647, + "step": 7074 + }, + { + "epoch": 0.6380484285521035, + "grad_norm": 3.0586882244149707, + "learning_rate": 1.2237841502596232e-06, + "loss": 0.7498, + "step": 7075 + }, + { + "epoch": 0.6381386120755738, + "grad_norm": 1.5375116735495928, + "learning_rate": 1.2232457667288583e-06, + "loss": 0.751, + "step": 7076 + }, + { + "epoch": 0.6382287955990441, + "grad_norm": 2.5111338644862946, + "learning_rate": 1.2227074494750476e-06, + "loss": 0.7566, + "step": 7077 + }, + { + "epoch": 0.6383189791225143, + "grad_norm": 2.5923163806775777, + "learning_rate": 1.2221691985441238e-06, + "loss": 0.7868, + "step": 7078 + }, + { + "epoch": 0.6384091626459846, + "grad_norm": 3.6220154150008677, + "learning_rate": 1.2216310139820128e-06, + "loss": 0.6962, + "step": 7079 + }, + { + "epoch": 0.6384993461694548, + "grad_norm": 1.9703035431551656, + "learning_rate": 1.2210928958346347e-06, + "loss": 0.7888, + "step": 7080 + }, + { + "epoch": 0.6385895296929252, + "grad_norm": 2.4355223041876575, + "learning_rate": 1.2205548441479065e-06, + "loss": 0.6779, + "step": 7081 + }, + { + "epoch": 0.6386797132163954, + "grad_norm": 1.650837529636206, + "learning_rate": 1.2200168589677357e-06, + "loss": 0.6807, + "step": 7082 + }, + { + "epoch": 0.6387698967398656, + "grad_norm": 1.922734262309025, + "learning_rate": 1.2194789403400284e-06, + "loss": 0.7164, + "step": 7083 + }, + { + "epoch": 0.6388600802633358, + "grad_norm": 1.9321058926044443, + "learning_rate": 1.2189410883106816e-06, + "loss": 0.7267, + "step": 7084 + }, + { + "epoch": 0.6389502637868062, + "grad_norm": 2.347397775750083, + "learning_rate": 1.2184033029255872e-06, + "loss": 0.8162, + "step": 7085 + }, + { + "epoch": 0.6390404473102764, + "grad_norm": 0.6458739657592171, + "learning_rate": 1.2178655842306334e-06, + "loss": 0.5815, + "step": 7086 + }, + { + "epoch": 0.6391306308337467, + "grad_norm": 1.8230931732465054, + "learning_rate": 1.2173279322716999e-06, + "loss": 0.7443, + "step": 7087 + }, + { + "epoch": 0.6392208143572169, + "grad_norm": 1.7957907681240703, + "learning_rate": 1.216790347094663e-06, + "loss": 0.7456, + "step": 7088 + }, + { + "epoch": 0.6393109978806872, + "grad_norm": 1.6486511092068314, + "learning_rate": 1.2162528287453927e-06, + "loss": 0.7959, + "step": 7089 + }, + { + "epoch": 0.6394011814041575, + "grad_norm": 2.2059297496715167, + "learning_rate": 1.215715377269752e-06, + "loss": 0.7542, + "step": 7090 + }, + { + "epoch": 0.6394913649276277, + "grad_norm": 1.4748225380129576, + "learning_rate": 1.2151779927136003e-06, + "loss": 0.7286, + "step": 7091 + }, + { + "epoch": 0.6395815484510979, + "grad_norm": 3.1014391973089674, + "learning_rate": 1.2146406751227893e-06, + "loss": 0.824, + "step": 7092 + }, + { + "epoch": 0.6396717319745683, + "grad_norm": 1.6050839299393842, + "learning_rate": 1.214103424543167e-06, + "loss": 0.78, + "step": 7093 + }, + { + "epoch": 0.6397619154980385, + "grad_norm": 1.7003510309083703, + "learning_rate": 1.2135662410205735e-06, + "loss": 0.6666, + "step": 7094 + }, + { + "epoch": 0.6398520990215087, + "grad_norm": 2.4817784013395943, + "learning_rate": 1.2130291246008444e-06, + "loss": 0.6659, + "step": 7095 + }, + { + "epoch": 0.639942282544979, + "grad_norm": 2.021100790146465, + "learning_rate": 1.21249207532981e-06, + "loss": 0.6703, + "step": 7096 + }, + { + "epoch": 0.6400324660684493, + "grad_norm": 15.691788896411119, + "learning_rate": 1.2119550932532936e-06, + "loss": 0.6697, + "step": 7097 + }, + { + "epoch": 0.6401226495919196, + "grad_norm": 2.5113811829682473, + "learning_rate": 1.2114181784171144e-06, + "loss": 0.7262, + "step": 7098 + }, + { + "epoch": 0.6402128331153898, + "grad_norm": 1.3931419654247863, + "learning_rate": 1.2108813308670837e-06, + "loss": 0.6489, + "step": 7099 + }, + { + "epoch": 0.6403030166388601, + "grad_norm": 2.787104156852815, + "learning_rate": 1.2103445506490099e-06, + "loss": 0.7481, + "step": 7100 + }, + { + "epoch": 0.6403932001623304, + "grad_norm": 0.710951333382972, + "learning_rate": 1.209807837808693e-06, + "loss": 0.5925, + "step": 7101 + }, + { + "epoch": 0.6404833836858006, + "grad_norm": 2.428303707817435, + "learning_rate": 1.2092711923919282e-06, + "loss": 0.79, + "step": 7102 + }, + { + "epoch": 0.6405735672092708, + "grad_norm": 1.8919520489138735, + "learning_rate": 1.2087346144445053e-06, + "loss": 0.7099, + "step": 7103 + }, + { + "epoch": 0.6406637507327412, + "grad_norm": 2.0013768701885906, + "learning_rate": 1.2081981040122081e-06, + "loss": 0.7567, + "step": 7104 + }, + { + "epoch": 0.6407539342562114, + "grad_norm": 6.648650079683851, + "learning_rate": 1.2076616611408148e-06, + "loss": 0.7013, + "step": 7105 + }, + { + "epoch": 0.6408441177796816, + "grad_norm": 1.7747526291955242, + "learning_rate": 1.2071252858760972e-06, + "loss": 0.7787, + "step": 7106 + }, + { + "epoch": 0.6409343013031519, + "grad_norm": 2.063238185215038, + "learning_rate": 1.2065889782638218e-06, + "loss": 0.7341, + "step": 7107 + }, + { + "epoch": 0.6410244848266222, + "grad_norm": 1.57913221189141, + "learning_rate": 1.2060527383497506e-06, + "loss": 0.7295, + "step": 7108 + }, + { + "epoch": 0.6411146683500925, + "grad_norm": 3.187306716145028, + "learning_rate": 1.2055165661796363e-06, + "loss": 0.8616, + "step": 7109 + }, + { + "epoch": 0.6412048518735627, + "grad_norm": 1.9440685415978887, + "learning_rate": 1.2049804617992303e-06, + "loss": 0.8198, + "step": 7110 + }, + { + "epoch": 0.6412950353970329, + "grad_norm": 0.6290057322797372, + "learning_rate": 1.204444425254275e-06, + "loss": 0.5927, + "step": 7111 + }, + { + "epoch": 0.6413852189205033, + "grad_norm": 1.7793623308186974, + "learning_rate": 1.203908456590507e-06, + "loss": 0.7351, + "step": 7112 + }, + { + "epoch": 0.6414754024439735, + "grad_norm": 1.7960453274153447, + "learning_rate": 1.20337255585366e-06, + "loss": 0.7245, + "step": 7113 + }, + { + "epoch": 0.6415655859674437, + "grad_norm": 0.6159831984941037, + "learning_rate": 1.2028367230894582e-06, + "loss": 0.5384, + "step": 7114 + }, + { + "epoch": 0.641655769490914, + "grad_norm": 1.7785622915491754, + "learning_rate": 1.2023009583436237e-06, + "loss": 0.6681, + "step": 7115 + }, + { + "epoch": 0.6417459530143843, + "grad_norm": 2.0093247474093774, + "learning_rate": 1.2017652616618698e-06, + "loss": 0.811, + "step": 7116 + }, + { + "epoch": 0.6418361365378545, + "grad_norm": 2.062060338178142, + "learning_rate": 1.2012296330899048e-06, + "loss": 0.7409, + "step": 7117 + }, + { + "epoch": 0.6419263200613248, + "grad_norm": 1.934718257787105, + "learning_rate": 1.200694072673432e-06, + "loss": 0.7849, + "step": 7118 + }, + { + "epoch": 0.642016503584795, + "grad_norm": 2.4574768923723758, + "learning_rate": 1.200158580458148e-06, + "loss": 0.7758, + "step": 7119 + }, + { + "epoch": 0.6421066871082654, + "grad_norm": 1.5752743194499663, + "learning_rate": 1.1996231564897448e-06, + "loss": 0.7233, + "step": 7120 + }, + { + "epoch": 0.6421968706317356, + "grad_norm": 1.7694033140708334, + "learning_rate": 1.1990878008139062e-06, + "loss": 0.7069, + "step": 7121 + }, + { + "epoch": 0.6422870541552058, + "grad_norm": 1.5442620760158479, + "learning_rate": 1.1985525134763132e-06, + "loss": 0.7793, + "step": 7122 + }, + { + "epoch": 0.6423772376786762, + "grad_norm": 3.593301496299823, + "learning_rate": 1.1980172945226393e-06, + "loss": 0.7912, + "step": 7123 + }, + { + "epoch": 0.6424674212021464, + "grad_norm": 2.152296117681309, + "learning_rate": 1.197482143998551e-06, + "loss": 0.755, + "step": 7124 + }, + { + "epoch": 0.6425576047256166, + "grad_norm": 1.6514866547144949, + "learning_rate": 1.196947061949712e-06, + "loss": 0.7344, + "step": 7125 + }, + { + "epoch": 0.6426477882490869, + "grad_norm": 1.588256268321235, + "learning_rate": 1.1964120484217768e-06, + "loss": 0.8046, + "step": 7126 + }, + { + "epoch": 0.6427379717725572, + "grad_norm": 0.7029326493692918, + "learning_rate": 1.1958771034603975e-06, + "loss": 0.5627, + "step": 7127 + }, + { + "epoch": 0.6428281552960274, + "grad_norm": 1.69025041587135, + "learning_rate": 1.1953422271112175e-06, + "loss": 0.6912, + "step": 7128 + }, + { + "epoch": 0.6429183388194977, + "grad_norm": 2.059055110184607, + "learning_rate": 1.1948074194198748e-06, + "loss": 0.734, + "step": 7129 + }, + { + "epoch": 0.6430085223429679, + "grad_norm": 1.7579125269934295, + "learning_rate": 1.1942726804320033e-06, + "loss": 0.6641, + "step": 7130 + }, + { + "epoch": 0.6430987058664382, + "grad_norm": 2.2996213933061393, + "learning_rate": 1.1937380101932295e-06, + "loss": 0.6889, + "step": 7131 + }, + { + "epoch": 0.6431888893899085, + "grad_norm": 1.5391958730777129, + "learning_rate": 1.1932034087491745e-06, + "loss": 0.7422, + "step": 7132 + }, + { + "epoch": 0.6432790729133787, + "grad_norm": 1.8936711512098823, + "learning_rate": 1.1926688761454531e-06, + "loss": 0.6664, + "step": 7133 + }, + { + "epoch": 0.643369256436849, + "grad_norm": 1.2929894944108635, + "learning_rate": 1.1921344124276746e-06, + "loss": 0.7666, + "step": 7134 + }, + { + "epoch": 0.6434594399603193, + "grad_norm": 1.6635690593275585, + "learning_rate": 1.1916000176414431e-06, + "loss": 0.7698, + "step": 7135 + }, + { + "epoch": 0.6435496234837895, + "grad_norm": 1.756430391977659, + "learning_rate": 1.1910656918323546e-06, + "loss": 0.7818, + "step": 7136 + }, + { + "epoch": 0.6436398070072598, + "grad_norm": 2.520954623076125, + "learning_rate": 1.1905314350460024e-06, + "loss": 0.7777, + "step": 7137 + }, + { + "epoch": 0.64372999053073, + "grad_norm": 1.5887493143747984, + "learning_rate": 1.1899972473279717e-06, + "loss": 0.803, + "step": 7138 + }, + { + "epoch": 0.6438201740542003, + "grad_norm": 1.471678715473302, + "learning_rate": 1.1894631287238414e-06, + "loss": 0.7424, + "step": 7139 + }, + { + "epoch": 0.6439103575776706, + "grad_norm": 1.5140861485763182, + "learning_rate": 1.188929079279187e-06, + "loss": 0.7576, + "step": 7140 + }, + { + "epoch": 0.6440005411011408, + "grad_norm": 2.1023762201802034, + "learning_rate": 1.1883950990395751e-06, + "loss": 0.687, + "step": 7141 + }, + { + "epoch": 0.644090724624611, + "grad_norm": 2.307828153377643, + "learning_rate": 1.187861188050569e-06, + "loss": 0.6924, + "step": 7142 + }, + { + "epoch": 0.6441809081480814, + "grad_norm": 1.6136002860145144, + "learning_rate": 1.187327346357724e-06, + "loss": 0.7372, + "step": 7143 + }, + { + "epoch": 0.6442710916715516, + "grad_norm": 1.744385090234618, + "learning_rate": 1.1867935740065912e-06, + "loss": 0.774, + "step": 7144 + }, + { + "epoch": 0.6443612751950218, + "grad_norm": 1.998565338078352, + "learning_rate": 1.1862598710427148e-06, + "loss": 0.6413, + "step": 7145 + }, + { + "epoch": 0.6444514587184921, + "grad_norm": 2.207082022023857, + "learning_rate": 1.1857262375116328e-06, + "loss": 0.7152, + "step": 7146 + }, + { + "epoch": 0.6445416422419624, + "grad_norm": 1.6384177482617839, + "learning_rate": 1.1851926734588783e-06, + "loss": 0.7527, + "step": 7147 + }, + { + "epoch": 0.6446318257654327, + "grad_norm": 1.9319919309151412, + "learning_rate": 1.184659178929977e-06, + "loss": 0.7712, + "step": 7148 + }, + { + "epoch": 0.6447220092889029, + "grad_norm": 1.816647552541334, + "learning_rate": 1.1841257539704513e-06, + "loss": 0.7439, + "step": 7149 + }, + { + "epoch": 0.6448121928123732, + "grad_norm": 1.5262820561172301, + "learning_rate": 1.1835923986258146e-06, + "loss": 0.7695, + "step": 7150 + }, + { + "epoch": 0.6449023763358435, + "grad_norm": 1.5546088270668257, + "learning_rate": 1.1830591129415754e-06, + "loss": 0.7081, + "step": 7151 + }, + { + "epoch": 0.6449925598593137, + "grad_norm": 1.7108605772519636, + "learning_rate": 1.182525896963238e-06, + "loss": 0.6948, + "step": 7152 + }, + { + "epoch": 0.6450827433827839, + "grad_norm": 3.5131664580562814, + "learning_rate": 1.181992750736298e-06, + "loss": 0.6842, + "step": 7153 + }, + { + "epoch": 0.6451729269062543, + "grad_norm": 2.1002321251234672, + "learning_rate": 1.1814596743062474e-06, + "loss": 0.7433, + "step": 7154 + }, + { + "epoch": 0.6452631104297245, + "grad_norm": 4.413315755419189, + "learning_rate": 1.1809266677185711e-06, + "loss": 0.7753, + "step": 7155 + }, + { + "epoch": 0.6453532939531947, + "grad_norm": 0.6538390422451557, + "learning_rate": 1.180393731018747e-06, + "loss": 0.6128, + "step": 7156 + }, + { + "epoch": 0.645443477476665, + "grad_norm": 0.5638548640565091, + "learning_rate": 1.1798608642522498e-06, + "loss": 0.5514, + "step": 7157 + }, + { + "epoch": 0.6455336610001353, + "grad_norm": 1.994467713135571, + "learning_rate": 1.1793280674645454e-06, + "loss": 0.7305, + "step": 7158 + }, + { + "epoch": 0.6456238445236056, + "grad_norm": 1.6851617366469775, + "learning_rate": 1.1787953407010954e-06, + "loss": 0.8235, + "step": 7159 + }, + { + "epoch": 0.6457140280470758, + "grad_norm": 2.055237091012776, + "learning_rate": 1.1782626840073554e-06, + "loss": 0.7509, + "step": 7160 + }, + { + "epoch": 0.645804211570546, + "grad_norm": 2.7725546707803295, + "learning_rate": 1.1777300974287738e-06, + "loss": 0.6548, + "step": 7161 + }, + { + "epoch": 0.6458943950940164, + "grad_norm": 1.7004307425577778, + "learning_rate": 1.1771975810107947e-06, + "loss": 0.781, + "step": 7162 + }, + { + "epoch": 0.6459845786174866, + "grad_norm": 2.3611797983488287, + "learning_rate": 1.1766651347988542e-06, + "loss": 0.8166, + "step": 7163 + }, + { + "epoch": 0.6460747621409568, + "grad_norm": 1.9286521574310396, + "learning_rate": 1.1761327588383848e-06, + "loss": 0.8599, + "step": 7164 + }, + { + "epoch": 0.6461649456644271, + "grad_norm": 2.045084046033457, + "learning_rate": 1.1756004531748105e-06, + "loss": 0.7658, + "step": 7165 + }, + { + "epoch": 0.6462551291878974, + "grad_norm": 1.6442389511611823, + "learning_rate": 1.1750682178535521e-06, + "loss": 0.7861, + "step": 7166 + }, + { + "epoch": 0.6463453127113676, + "grad_norm": 1.733237297197532, + "learning_rate": 1.1745360529200218e-06, + "loss": 0.6987, + "step": 7167 + }, + { + "epoch": 0.6464354962348379, + "grad_norm": 1.567242634526998, + "learning_rate": 1.1740039584196261e-06, + "loss": 0.7751, + "step": 7168 + }, + { + "epoch": 0.6465256797583081, + "grad_norm": 8.174875091625063, + "learning_rate": 1.1734719343977683e-06, + "loss": 0.7819, + "step": 7169 + }, + { + "epoch": 0.6466158632817784, + "grad_norm": 1.9293892201306473, + "learning_rate": 1.1729399808998416e-06, + "loss": 0.6913, + "step": 7170 + }, + { + "epoch": 0.6467060468052487, + "grad_norm": 1.834073683054153, + "learning_rate": 1.1724080979712368e-06, + "loss": 0.6377, + "step": 7171 + }, + { + "epoch": 0.6467962303287189, + "grad_norm": 1.528820134048986, + "learning_rate": 1.1718762856573365e-06, + "loss": 0.8173, + "step": 7172 + }, + { + "epoch": 0.6468864138521893, + "grad_norm": 1.3897503448886652, + "learning_rate": 1.1713445440035172e-06, + "loss": 0.7362, + "step": 7173 + }, + { + "epoch": 0.6469765973756595, + "grad_norm": 1.7479408956613327, + "learning_rate": 1.1708128730551506e-06, + "loss": 0.7863, + "step": 7174 + }, + { + "epoch": 0.6470667808991297, + "grad_norm": 1.6230088285721957, + "learning_rate": 1.1702812728576019e-06, + "loss": 0.7405, + "step": 7175 + }, + { + "epoch": 0.6471569644226, + "grad_norm": 1.663546415006501, + "learning_rate": 1.1697497434562303e-06, + "loss": 0.7757, + "step": 7176 + }, + { + "epoch": 0.6472471479460703, + "grad_norm": 2.4514872921749093, + "learning_rate": 1.1692182848963885e-06, + "loss": 0.7856, + "step": 7177 + }, + { + "epoch": 0.6473373314695405, + "grad_norm": 2.174181913666237, + "learning_rate": 1.1686868972234227e-06, + "loss": 0.714, + "step": 7178 + }, + { + "epoch": 0.6474275149930108, + "grad_norm": 2.190783733543086, + "learning_rate": 1.1681555804826755e-06, + "loss": 0.7533, + "step": 7179 + }, + { + "epoch": 0.647517698516481, + "grad_norm": 1.5495811450214874, + "learning_rate": 1.1676243347194806e-06, + "loss": 0.6894, + "step": 7180 + }, + { + "epoch": 0.6476078820399513, + "grad_norm": 1.3605435928011613, + "learning_rate": 1.167093159979167e-06, + "loss": 0.7712, + "step": 7181 + }, + { + "epoch": 0.6476980655634216, + "grad_norm": 1.5162269733941625, + "learning_rate": 1.1665620563070575e-06, + "loss": 0.7773, + "step": 7182 + }, + { + "epoch": 0.6477882490868918, + "grad_norm": 1.8236970666076024, + "learning_rate": 1.1660310237484691e-06, + "loss": 0.8652, + "step": 7183 + }, + { + "epoch": 0.647878432610362, + "grad_norm": 1.4612176799871006, + "learning_rate": 1.165500062348713e-06, + "loss": 0.691, + "step": 7184 + }, + { + "epoch": 0.6479686161338324, + "grad_norm": 1.694761406414015, + "learning_rate": 1.164969172153091e-06, + "loss": 0.7957, + "step": 7185 + }, + { + "epoch": 0.6480587996573026, + "grad_norm": 2.044225477618535, + "learning_rate": 1.1644383532069055e-06, + "loss": 0.7663, + "step": 7186 + }, + { + "epoch": 0.6481489831807729, + "grad_norm": 2.363442659777192, + "learning_rate": 1.1639076055554454e-06, + "loss": 0.7887, + "step": 7187 + }, + { + "epoch": 0.6482391667042431, + "grad_norm": 2.218261987449769, + "learning_rate": 1.163376929244e-06, + "loss": 0.737, + "step": 7188 + }, + { + "epoch": 0.6483293502277134, + "grad_norm": 2.8507580602261453, + "learning_rate": 1.1628463243178472e-06, + "loss": 0.7097, + "step": 7189 + }, + { + "epoch": 0.6484195337511837, + "grad_norm": 2.022019408550746, + "learning_rate": 1.1623157908222623e-06, + "loss": 0.8115, + "step": 7190 + }, + { + "epoch": 0.6485097172746539, + "grad_norm": 1.8667664631778809, + "learning_rate": 1.1617853288025129e-06, + "loss": 0.7884, + "step": 7191 + }, + { + "epoch": 0.6485999007981241, + "grad_norm": 1.9213095973046792, + "learning_rate": 1.1612549383038612e-06, + "loss": 0.7252, + "step": 7192 + }, + { + "epoch": 0.6486900843215945, + "grad_norm": 1.3766610776004224, + "learning_rate": 1.1607246193715629e-06, + "loss": 0.7291, + "step": 7193 + }, + { + "epoch": 0.6487802678450647, + "grad_norm": 1.8478177989959217, + "learning_rate": 1.1601943720508684e-06, + "loss": 0.7678, + "step": 7194 + }, + { + "epoch": 0.6488704513685349, + "grad_norm": 1.551978967558523, + "learning_rate": 1.1596641963870202e-06, + "loss": 0.8003, + "step": 7195 + }, + { + "epoch": 0.6489606348920053, + "grad_norm": 1.8040407020815372, + "learning_rate": 1.1591340924252561e-06, + "loss": 0.7616, + "step": 7196 + }, + { + "epoch": 0.6490508184154755, + "grad_norm": 0.7660273837481122, + "learning_rate": 1.158604060210808e-06, + "loss": 0.6237, + "step": 7197 + }, + { + "epoch": 0.6491410019389457, + "grad_norm": 2.0933124597484687, + "learning_rate": 1.1580740997889008e-06, + "loss": 0.8142, + "step": 7198 + }, + { + "epoch": 0.649231185462416, + "grad_norm": 1.505101350812513, + "learning_rate": 1.1575442112047544e-06, + "loss": 0.7459, + "step": 7199 + }, + { + "epoch": 0.6493213689858863, + "grad_norm": 2.1334490262887034, + "learning_rate": 1.1570143945035797e-06, + "loss": 0.6887, + "step": 7200 + }, + { + "epoch": 0.6494115525093566, + "grad_norm": 2.7837843759826897, + "learning_rate": 1.1564846497305864e-06, + "loss": 0.7069, + "step": 7201 + }, + { + "epoch": 0.6495017360328268, + "grad_norm": 1.6720985191879345, + "learning_rate": 1.1559549769309726e-06, + "loss": 0.7269, + "step": 7202 + }, + { + "epoch": 0.649591919556297, + "grad_norm": 1.5115936779555237, + "learning_rate": 1.1554253761499358e-06, + "loss": 0.7419, + "step": 7203 + }, + { + "epoch": 0.6496821030797674, + "grad_norm": 1.92183803119614, + "learning_rate": 1.1548958474326617e-06, + "loss": 0.7879, + "step": 7204 + }, + { + "epoch": 0.6497722866032376, + "grad_norm": 1.7161168128315751, + "learning_rate": 1.154366390824334e-06, + "loss": 0.6797, + "step": 7205 + }, + { + "epoch": 0.6498624701267078, + "grad_norm": 0.7288248668838676, + "learning_rate": 1.1538370063701287e-06, + "loss": 0.5591, + "step": 7206 + }, + { + "epoch": 0.6499526536501781, + "grad_norm": 1.435018055366944, + "learning_rate": 1.1533076941152153e-06, + "loss": 0.7048, + "step": 7207 + }, + { + "epoch": 0.6500428371736484, + "grad_norm": 2.228757826726653, + "learning_rate": 1.1527784541047583e-06, + "loss": 0.6751, + "step": 7208 + }, + { + "epoch": 0.6501330206971186, + "grad_norm": 2.383704446918449, + "learning_rate": 1.1522492863839152e-06, + "loss": 0.7064, + "step": 7209 + }, + { + "epoch": 0.6502232042205889, + "grad_norm": 1.842859990887643, + "learning_rate": 1.1517201909978382e-06, + "loss": 0.7124, + "step": 7210 + }, + { + "epoch": 0.6503133877440591, + "grad_norm": 0.6004379462292783, + "learning_rate": 1.151191167991671e-06, + "loss": 0.5787, + "step": 7211 + }, + { + "epoch": 0.6504035712675295, + "grad_norm": 2.3858514136951614, + "learning_rate": 1.1506622174105536e-06, + "loss": 0.7569, + "step": 7212 + }, + { + "epoch": 0.6504937547909997, + "grad_norm": 2.173566530018969, + "learning_rate": 1.1501333392996194e-06, + "loss": 0.7801, + "step": 7213 + }, + { + "epoch": 0.6505839383144699, + "grad_norm": 0.655680113804178, + "learning_rate": 1.1496045337039943e-06, + "loss": 0.5847, + "step": 7214 + }, + { + "epoch": 0.6506741218379402, + "grad_norm": 2.0031359643978273, + "learning_rate": 1.1490758006687995e-06, + "loss": 0.7064, + "step": 7215 + }, + { + "epoch": 0.6507643053614105, + "grad_norm": 1.5168074095402186, + "learning_rate": 1.1485471402391502e-06, + "loss": 0.7303, + "step": 7216 + }, + { + "epoch": 0.6508544888848807, + "grad_norm": 2.154830465352581, + "learning_rate": 1.1480185524601522e-06, + "loss": 0.7233, + "step": 7217 + }, + { + "epoch": 0.650944672408351, + "grad_norm": 1.9864866780454387, + "learning_rate": 1.1474900373769108e-06, + "loss": 0.7546, + "step": 7218 + }, + { + "epoch": 0.6510348559318213, + "grad_norm": 0.5702648142905999, + "learning_rate": 1.1469615950345184e-06, + "loss": 0.525, + "step": 7219 + }, + { + "epoch": 0.6511250394552915, + "grad_norm": 0.7017425297623882, + "learning_rate": 1.1464332254780678e-06, + "loss": 0.5629, + "step": 7220 + }, + { + "epoch": 0.6512152229787618, + "grad_norm": 3.518582062584321, + "learning_rate": 1.1459049287526404e-06, + "loss": 0.7337, + "step": 7221 + }, + { + "epoch": 0.651305406502232, + "grad_norm": 2.900656290442424, + "learning_rate": 1.1453767049033137e-06, + "loss": 0.709, + "step": 7222 + }, + { + "epoch": 0.6513955900257024, + "grad_norm": 3.5043532682819603, + "learning_rate": 1.1448485539751586e-06, + "loss": 0.7886, + "step": 7223 + }, + { + "epoch": 0.6514857735491726, + "grad_norm": 1.5188024974901926, + "learning_rate": 1.1443204760132408e-06, + "loss": 0.723, + "step": 7224 + }, + { + "epoch": 0.6515759570726428, + "grad_norm": 1.4242059235196647, + "learning_rate": 1.1437924710626185e-06, + "loss": 0.7278, + "step": 7225 + }, + { + "epoch": 0.651666140596113, + "grad_norm": 0.651444369065691, + "learning_rate": 1.1432645391683429e-06, + "loss": 0.547, + "step": 7226 + }, + { + "epoch": 0.6517563241195834, + "grad_norm": 2.2728867535556025, + "learning_rate": 1.1427366803754609e-06, + "loss": 0.7847, + "step": 7227 + }, + { + "epoch": 0.6518465076430536, + "grad_norm": 1.469688440379759, + "learning_rate": 1.142208894729012e-06, + "loss": 0.7615, + "step": 7228 + }, + { + "epoch": 0.6519366911665239, + "grad_norm": 1.5037060889272316, + "learning_rate": 1.1416811822740301e-06, + "loss": 0.7897, + "step": 7229 + }, + { + "epoch": 0.6520268746899941, + "grad_norm": 1.4681874273221476, + "learning_rate": 1.1411535430555428e-06, + "loss": 0.7985, + "step": 7230 + }, + { + "epoch": 0.6521170582134644, + "grad_norm": 1.542009813764838, + "learning_rate": 1.1406259771185705e-06, + "loss": 0.8058, + "step": 7231 + }, + { + "epoch": 0.6522072417369347, + "grad_norm": 1.5518185775891635, + "learning_rate": 1.1400984845081282e-06, + "loss": 0.7092, + "step": 7232 + }, + { + "epoch": 0.6522974252604049, + "grad_norm": 6.437114222550362, + "learning_rate": 1.139571065269226e-06, + "loss": 0.8329, + "step": 7233 + }, + { + "epoch": 0.6523876087838751, + "grad_norm": 1.5373023952873666, + "learning_rate": 1.139043719446863e-06, + "loss": 0.7975, + "step": 7234 + }, + { + "epoch": 0.6524777923073455, + "grad_norm": 2.2454759892649663, + "learning_rate": 1.1385164470860385e-06, + "loss": 0.8209, + "step": 7235 + }, + { + "epoch": 0.6525679758308157, + "grad_norm": 1.9757612813390748, + "learning_rate": 1.1379892482317403e-06, + "loss": 0.7497, + "step": 7236 + }, + { + "epoch": 0.652658159354286, + "grad_norm": 2.0737204087751926, + "learning_rate": 1.1374621229289524e-06, + "loss": 0.7382, + "step": 7237 + }, + { + "epoch": 0.6527483428777562, + "grad_norm": 3.0185175646273565, + "learning_rate": 1.1369350712226525e-06, + "loss": 0.7366, + "step": 7238 + }, + { + "epoch": 0.6528385264012265, + "grad_norm": 1.5072860984813636, + "learning_rate": 1.136408093157811e-06, + "loss": 0.6995, + "step": 7239 + }, + { + "epoch": 0.6529287099246968, + "grad_norm": 3.591009139718323, + "learning_rate": 1.1358811887793935e-06, + "loss": 0.7234, + "step": 7240 + }, + { + "epoch": 0.653018893448167, + "grad_norm": 1.6827571055752923, + "learning_rate": 1.135354358132356e-06, + "loss": 0.789, + "step": 7241 + }, + { + "epoch": 0.6531090769716373, + "grad_norm": 1.664227616262595, + "learning_rate": 1.1348276012616542e-06, + "loss": 0.7893, + "step": 7242 + }, + { + "epoch": 0.6531992604951076, + "grad_norm": 3.7452904692229865, + "learning_rate": 1.134300918212231e-06, + "loss": 0.7057, + "step": 7243 + }, + { + "epoch": 0.6532894440185778, + "grad_norm": 1.858166790956427, + "learning_rate": 1.133774309029027e-06, + "loss": 0.7591, + "step": 7244 + }, + { + "epoch": 0.653379627542048, + "grad_norm": 5.089972470252153, + "learning_rate": 1.133247773756975e-06, + "loss": 0.7937, + "step": 7245 + }, + { + "epoch": 0.6534698110655184, + "grad_norm": 0.6555317626782362, + "learning_rate": 1.1327213124410024e-06, + "loss": 0.6177, + "step": 7246 + }, + { + "epoch": 0.6535599945889886, + "grad_norm": 1.6455080298358535, + "learning_rate": 1.1321949251260292e-06, + "loss": 0.7921, + "step": 7247 + }, + { + "epoch": 0.6536501781124588, + "grad_norm": 6.156424118135728, + "learning_rate": 1.1316686118569712e-06, + "loss": 0.6367, + "step": 7248 + }, + { + "epoch": 0.6537403616359291, + "grad_norm": 1.6044814886663992, + "learning_rate": 1.1311423726787335e-06, + "loss": 0.66, + "step": 7249 + }, + { + "epoch": 0.6538305451593994, + "grad_norm": 2.788221558380568, + "learning_rate": 1.130616207636221e-06, + "loss": 0.7189, + "step": 7250 + }, + { + "epoch": 0.6539207286828697, + "grad_norm": 2.017634846129114, + "learning_rate": 1.1300901167743263e-06, + "loss": 0.756, + "step": 7251 + }, + { + "epoch": 0.6540109122063399, + "grad_norm": 1.4195614951674378, + "learning_rate": 1.12956410013794e-06, + "loss": 0.7474, + "step": 7252 + }, + { + "epoch": 0.6541010957298101, + "grad_norm": 3.857583807914354, + "learning_rate": 1.1290381577719436e-06, + "loss": 0.7369, + "step": 7253 + }, + { + "epoch": 0.6541912792532805, + "grad_norm": 1.8111529112726719, + "learning_rate": 1.1285122897212143e-06, + "loss": 0.6545, + "step": 7254 + }, + { + "epoch": 0.6542814627767507, + "grad_norm": 0.7441724635244085, + "learning_rate": 1.1279864960306228e-06, + "loss": 0.6183, + "step": 7255 + }, + { + "epoch": 0.6543716463002209, + "grad_norm": 1.6317227025812004, + "learning_rate": 1.1274607767450297e-06, + "loss": 0.8242, + "step": 7256 + }, + { + "epoch": 0.6544618298236912, + "grad_norm": 3.0056193493246393, + "learning_rate": 1.126935131909296e-06, + "loss": 0.724, + "step": 7257 + }, + { + "epoch": 0.6545520133471615, + "grad_norm": 1.4126213696754417, + "learning_rate": 1.1264095615682693e-06, + "loss": 0.694, + "step": 7258 + }, + { + "epoch": 0.6546421968706317, + "grad_norm": 3.0015382002204825, + "learning_rate": 1.1258840657667973e-06, + "loss": 0.6647, + "step": 7259 + }, + { + "epoch": 0.654732380394102, + "grad_norm": 1.4701260552887023, + "learning_rate": 1.125358644549716e-06, + "loss": 0.6798, + "step": 7260 + }, + { + "epoch": 0.6548225639175722, + "grad_norm": 1.4314922344331567, + "learning_rate": 1.1248332979618578e-06, + "loss": 0.6108, + "step": 7261 + }, + { + "epoch": 0.6549127474410426, + "grad_norm": 1.491029299882437, + "learning_rate": 1.1243080260480482e-06, + "loss": 0.6387, + "step": 7262 + }, + { + "epoch": 0.6550029309645128, + "grad_norm": 1.9965835836231431, + "learning_rate": 1.1237828288531063e-06, + "loss": 0.8052, + "step": 7263 + }, + { + "epoch": 0.655093114487983, + "grad_norm": 2.1752695683891483, + "learning_rate": 1.1232577064218449e-06, + "loss": 0.6668, + "step": 7264 + }, + { + "epoch": 0.6551832980114533, + "grad_norm": 2.1892073371670855, + "learning_rate": 1.1227326587990711e-06, + "loss": 0.8235, + "step": 7265 + }, + { + "epoch": 0.6552734815349236, + "grad_norm": 1.4263670671513369, + "learning_rate": 1.1222076860295832e-06, + "loss": 0.7314, + "step": 7266 + }, + { + "epoch": 0.6553636650583938, + "grad_norm": 1.6729027272075583, + "learning_rate": 1.1216827881581756e-06, + "loss": 0.7754, + "step": 7267 + }, + { + "epoch": 0.6554538485818641, + "grad_norm": 2.7310662004792463, + "learning_rate": 1.1211579652296355e-06, + "loss": 0.7404, + "step": 7268 + }, + { + "epoch": 0.6555440321053344, + "grad_norm": 2.307233845579697, + "learning_rate": 1.1206332172887438e-06, + "loss": 0.8316, + "step": 7269 + }, + { + "epoch": 0.6556342156288046, + "grad_norm": 1.6713488454982666, + "learning_rate": 1.1201085443802756e-06, + "loss": 0.7585, + "step": 7270 + }, + { + "epoch": 0.6557243991522749, + "grad_norm": 2.131021047694976, + "learning_rate": 1.1195839465489964e-06, + "loss": 0.7152, + "step": 7271 + }, + { + "epoch": 0.6558145826757451, + "grad_norm": 1.7118902471336184, + "learning_rate": 1.1190594238396708e-06, + "loss": 0.7928, + "step": 7272 + }, + { + "epoch": 0.6559047661992154, + "grad_norm": 1.5931041381236244, + "learning_rate": 1.1185349762970515e-06, + "loss": 0.7217, + "step": 7273 + }, + { + "epoch": 0.6559949497226857, + "grad_norm": 2.8979829087263473, + "learning_rate": 1.1180106039658896e-06, + "loss": 0.7518, + "step": 7274 + }, + { + "epoch": 0.6560851332461559, + "grad_norm": 0.5844929244654967, + "learning_rate": 1.117486306890925e-06, + "loss": 0.5584, + "step": 7275 + }, + { + "epoch": 0.6561753167696261, + "grad_norm": 2.501243091518426, + "learning_rate": 1.116962085116896e-06, + "loss": 0.7015, + "step": 7276 + }, + { + "epoch": 0.6562655002930965, + "grad_norm": 0.6653814344766861, + "learning_rate": 1.1164379386885302e-06, + "loss": 0.5178, + "step": 7277 + }, + { + "epoch": 0.6563556838165667, + "grad_norm": 2.2407803788837453, + "learning_rate": 1.1159138676505516e-06, + "loss": 0.7958, + "step": 7278 + }, + { + "epoch": 0.656445867340037, + "grad_norm": 1.9361217272574003, + "learning_rate": 1.1153898720476761e-06, + "loss": 0.8045, + "step": 7279 + }, + { + "epoch": 0.6565360508635072, + "grad_norm": 1.6751973306368644, + "learning_rate": 1.114865951924615e-06, + "loss": 0.7337, + "step": 7280 + }, + { + "epoch": 0.6566262343869775, + "grad_norm": 2.035880108344241, + "learning_rate": 1.1143421073260721e-06, + "loss": 0.6935, + "step": 7281 + }, + { + "epoch": 0.6567164179104478, + "grad_norm": 2.131308336102794, + "learning_rate": 1.1138183382967432e-06, + "loss": 0.6883, + "step": 7282 + }, + { + "epoch": 0.656806601433918, + "grad_norm": 2.491451238379727, + "learning_rate": 1.11329464488132e-06, + "loss": 0.6602, + "step": 7283 + }, + { + "epoch": 0.6568967849573882, + "grad_norm": 1.5044110132272233, + "learning_rate": 1.112771027124487e-06, + "loss": 0.7229, + "step": 7284 + }, + { + "epoch": 0.6569869684808586, + "grad_norm": 1.7606069097107508, + "learning_rate": 1.112247485070922e-06, + "loss": 0.7587, + "step": 7285 + }, + { + "epoch": 0.6570771520043288, + "grad_norm": 1.746055244174837, + "learning_rate": 1.1117240187652968e-06, + "loss": 0.7513, + "step": 7286 + }, + { + "epoch": 0.657167335527799, + "grad_norm": 1.9556135879139736, + "learning_rate": 1.1112006282522767e-06, + "loss": 0.8273, + "step": 7287 + }, + { + "epoch": 0.6572575190512693, + "grad_norm": 1.7560085279957147, + "learning_rate": 1.1106773135765183e-06, + "loss": 0.7302, + "step": 7288 + }, + { + "epoch": 0.6573477025747396, + "grad_norm": 1.6512189061993523, + "learning_rate": 1.110154074782677e-06, + "loss": 0.726, + "step": 7289 + }, + { + "epoch": 0.6574378860982099, + "grad_norm": 1.7803813087647822, + "learning_rate": 1.1096309119153948e-06, + "loss": 0.7438, + "step": 7290 + }, + { + "epoch": 0.6575280696216801, + "grad_norm": 0.6324729950609383, + "learning_rate": 1.1091078250193145e-06, + "loss": 0.6199, + "step": 7291 + }, + { + "epoch": 0.6576182531451504, + "grad_norm": 2.708000416957085, + "learning_rate": 1.108584814139066e-06, + "loss": 0.754, + "step": 7292 + }, + { + "epoch": 0.6577084366686207, + "grad_norm": 1.6239731552038144, + "learning_rate": 1.108061879319276e-06, + "loss": 0.7462, + "step": 7293 + }, + { + "epoch": 0.6577986201920909, + "grad_norm": 2.0103040146437365, + "learning_rate": 1.1075390206045648e-06, + "loss": 0.8049, + "step": 7294 + }, + { + "epoch": 0.6578888037155611, + "grad_norm": 1.8065587370588663, + "learning_rate": 1.1070162380395454e-06, + "loss": 0.7915, + "step": 7295 + }, + { + "epoch": 0.6579789872390315, + "grad_norm": 1.5475833696577577, + "learning_rate": 1.1064935316688253e-06, + "loss": 0.6909, + "step": 7296 + }, + { + "epoch": 0.6580691707625017, + "grad_norm": 1.6703082435203966, + "learning_rate": 1.105970901537002e-06, + "loss": 0.8345, + "step": 7297 + }, + { + "epoch": 0.6581593542859719, + "grad_norm": 1.7116363352335673, + "learning_rate": 1.1054483476886727e-06, + "loss": 0.7935, + "step": 7298 + }, + { + "epoch": 0.6582495378094422, + "grad_norm": 1.413316266336198, + "learning_rate": 1.1049258701684222e-06, + "loss": 0.6678, + "step": 7299 + }, + { + "epoch": 0.6583397213329125, + "grad_norm": 1.6524687014110333, + "learning_rate": 1.1044034690208315e-06, + "loss": 0.6944, + "step": 7300 + }, + { + "epoch": 0.6584299048563828, + "grad_norm": 2.3873038565868225, + "learning_rate": 1.1038811442904755e-06, + "loss": 0.782, + "step": 7301 + }, + { + "epoch": 0.658520088379853, + "grad_norm": 1.7191897436612642, + "learning_rate": 1.103358896021921e-06, + "loss": 0.6794, + "step": 7302 + }, + { + "epoch": 0.6586102719033232, + "grad_norm": 1.7622131195343704, + "learning_rate": 1.1028367242597298e-06, + "loss": 0.731, + "step": 7303 + }, + { + "epoch": 0.6587004554267936, + "grad_norm": 2.046658802990712, + "learning_rate": 1.102314629048457e-06, + "loss": 0.7862, + "step": 7304 + }, + { + "epoch": 0.6587906389502638, + "grad_norm": 1.5360054598701316, + "learning_rate": 1.1017926104326484e-06, + "loss": 0.7965, + "step": 7305 + }, + { + "epoch": 0.658880822473734, + "grad_norm": 1.891380915576502, + "learning_rate": 1.1012706684568483e-06, + "loss": 0.8012, + "step": 7306 + }, + { + "epoch": 0.6589710059972043, + "grad_norm": 1.7089292680847414, + "learning_rate": 1.1007488031655894e-06, + "loss": 0.7685, + "step": 7307 + }, + { + "epoch": 0.6590611895206746, + "grad_norm": 2.4278874736458196, + "learning_rate": 1.1002270146034013e-06, + "loss": 0.7037, + "step": 7308 + }, + { + "epoch": 0.6591513730441448, + "grad_norm": 1.5422581062270817, + "learning_rate": 1.0997053028148052e-06, + "loss": 0.7765, + "step": 7309 + }, + { + "epoch": 0.6592415565676151, + "grad_norm": 1.4273251052582463, + "learning_rate": 1.0991836678443173e-06, + "loss": 0.7512, + "step": 7310 + }, + { + "epoch": 0.6593317400910853, + "grad_norm": 2.8191618963102174, + "learning_rate": 1.0986621097364465e-06, + "loss": 0.691, + "step": 7311 + }, + { + "epoch": 0.6594219236145556, + "grad_norm": 3.0695161441888583, + "learning_rate": 1.0981406285356932e-06, + "loss": 0.6632, + "step": 7312 + }, + { + "epoch": 0.6595121071380259, + "grad_norm": 3.576602928563074, + "learning_rate": 1.0976192242865554e-06, + "loss": 0.6753, + "step": 7313 + }, + { + "epoch": 0.6596022906614961, + "grad_norm": 1.7390737719997382, + "learning_rate": 1.0970978970335202e-06, + "loss": 0.8145, + "step": 7314 + }, + { + "epoch": 0.6596924741849665, + "grad_norm": 2.5007574841205145, + "learning_rate": 1.0965766468210714e-06, + "loss": 0.7688, + "step": 7315 + }, + { + "epoch": 0.6597826577084367, + "grad_norm": 1.6700341616351573, + "learning_rate": 1.0960554736936843e-06, + "loss": 0.7592, + "step": 7316 + }, + { + "epoch": 0.6598728412319069, + "grad_norm": 3.0554719585966597, + "learning_rate": 1.0955343776958283e-06, + "loss": 0.6604, + "step": 7317 + }, + { + "epoch": 0.6599630247553772, + "grad_norm": 1.830511898715857, + "learning_rate": 1.0950133588719665e-06, + "loss": 0.7653, + "step": 7318 + }, + { + "epoch": 0.6600532082788475, + "grad_norm": 1.77670135626407, + "learning_rate": 1.0944924172665551e-06, + "loss": 0.7697, + "step": 7319 + }, + { + "epoch": 0.6601433918023177, + "grad_norm": 2.8272097460699137, + "learning_rate": 1.0939715529240437e-06, + "loss": 0.7863, + "step": 7320 + }, + { + "epoch": 0.660233575325788, + "grad_norm": 1.7205822281902814, + "learning_rate": 1.0934507658888755e-06, + "loss": 0.7801, + "step": 7321 + }, + { + "epoch": 0.6603237588492582, + "grad_norm": 1.417999519455671, + "learning_rate": 1.092930056205486e-06, + "loss": 0.7233, + "step": 7322 + }, + { + "epoch": 0.6604139423727285, + "grad_norm": 2.044278193823002, + "learning_rate": 1.092409423918306e-06, + "loss": 0.7268, + "step": 7323 + }, + { + "epoch": 0.6605041258961988, + "grad_norm": 1.4026458329864275, + "learning_rate": 1.0918888690717581e-06, + "loss": 0.6772, + "step": 7324 + }, + { + "epoch": 0.660594309419669, + "grad_norm": 1.705312950857032, + "learning_rate": 1.091368391710259e-06, + "loss": 0.8127, + "step": 7325 + }, + { + "epoch": 0.6606844929431392, + "grad_norm": 1.6308298439504452, + "learning_rate": 1.0908479918782198e-06, + "loss": 0.6966, + "step": 7326 + }, + { + "epoch": 0.6607746764666096, + "grad_norm": 1.506008591043556, + "learning_rate": 1.0903276696200413e-06, + "loss": 0.7728, + "step": 7327 + }, + { + "epoch": 0.6608648599900798, + "grad_norm": 2.771437146548571, + "learning_rate": 1.0898074249801234e-06, + "loss": 0.755, + "step": 7328 + }, + { + "epoch": 0.6609550435135501, + "grad_norm": 1.6772470177171763, + "learning_rate": 1.0892872580028533e-06, + "loss": 0.7661, + "step": 7329 + }, + { + "epoch": 0.6610452270370203, + "grad_norm": 1.6086367052520756, + "learning_rate": 1.0887671687326178e-06, + "loss": 0.7789, + "step": 7330 + }, + { + "epoch": 0.6611354105604906, + "grad_norm": 3.975113340477365, + "learning_rate": 1.0882471572137908e-06, + "loss": 0.7138, + "step": 7331 + }, + { + "epoch": 0.6612255940839609, + "grad_norm": 1.6135367498288025, + "learning_rate": 1.087727223490744e-06, + "loss": 0.7962, + "step": 7332 + }, + { + "epoch": 0.6613157776074311, + "grad_norm": 1.7281871073980617, + "learning_rate": 1.0872073676078405e-06, + "loss": 0.8068, + "step": 7333 + }, + { + "epoch": 0.6614059611309013, + "grad_norm": 2.369911162888371, + "learning_rate": 1.0866875896094375e-06, + "loss": 0.698, + "step": 7334 + }, + { + "epoch": 0.6614961446543717, + "grad_norm": 1.5626677613797004, + "learning_rate": 1.0861678895398854e-06, + "loss": 0.7698, + "step": 7335 + }, + { + "epoch": 0.6615863281778419, + "grad_norm": 2.146613550160841, + "learning_rate": 1.0856482674435286e-06, + "loss": 0.7147, + "step": 7336 + }, + { + "epoch": 0.6616765117013121, + "grad_norm": 0.7533045527001712, + "learning_rate": 1.0851287233647024e-06, + "loss": 0.5748, + "step": 7337 + }, + { + "epoch": 0.6617666952247825, + "grad_norm": 1.6843740686191113, + "learning_rate": 1.084609257347738e-06, + "loss": 0.8061, + "step": 7338 + }, + { + "epoch": 0.6618568787482527, + "grad_norm": 8.447604987764098, + "learning_rate": 1.0840898694369594e-06, + "loss": 0.7245, + "step": 7339 + }, + { + "epoch": 0.661947062271723, + "grad_norm": 1.4779257964734196, + "learning_rate": 1.083570559676683e-06, + "loss": 0.7401, + "step": 7340 + }, + { + "epoch": 0.6620372457951932, + "grad_norm": 1.6385503094108331, + "learning_rate": 1.08305132811122e-06, + "loss": 0.7276, + "step": 7341 + }, + { + "epoch": 0.6621274293186635, + "grad_norm": 1.9680926724223207, + "learning_rate": 1.0825321747848735e-06, + "loss": 0.7515, + "step": 7342 + }, + { + "epoch": 0.6622176128421338, + "grad_norm": 0.7312517916013501, + "learning_rate": 1.0820130997419417e-06, + "loss": 0.6447, + "step": 7343 + }, + { + "epoch": 0.662307796365604, + "grad_norm": 1.809611287607797, + "learning_rate": 1.0814941030267123e-06, + "loss": 0.7431, + "step": 7344 + }, + { + "epoch": 0.6623979798890742, + "grad_norm": 0.6044983254954556, + "learning_rate": 1.080975184683472e-06, + "loss": 0.5631, + "step": 7345 + }, + { + "epoch": 0.6624881634125446, + "grad_norm": 1.8090990979500303, + "learning_rate": 1.0804563447564948e-06, + "loss": 0.8084, + "step": 7346 + }, + { + "epoch": 0.6625783469360148, + "grad_norm": 1.7104149982972396, + "learning_rate": 1.0799375832900545e-06, + "loss": 0.6938, + "step": 7347 + }, + { + "epoch": 0.662668530459485, + "grad_norm": 2.010150186801607, + "learning_rate": 1.0794189003284118e-06, + "loss": 0.783, + "step": 7348 + }, + { + "epoch": 0.6627587139829553, + "grad_norm": 2.1040282125213277, + "learning_rate": 1.0789002959158242e-06, + "loss": 0.768, + "step": 7349 + }, + { + "epoch": 0.6628488975064256, + "grad_norm": 1.9926419207335087, + "learning_rate": 1.0783817700965428e-06, + "loss": 0.7047, + "step": 7350 + }, + { + "epoch": 0.6629390810298958, + "grad_norm": 1.4730462081612288, + "learning_rate": 1.0778633229148102e-06, + "loss": 0.7961, + "step": 7351 + }, + { + "epoch": 0.6630292645533661, + "grad_norm": 2.0348434354273204, + "learning_rate": 1.0773449544148645e-06, + "loss": 0.7514, + "step": 7352 + }, + { + "epoch": 0.6631194480768363, + "grad_norm": 1.4918395430622768, + "learning_rate": 1.076826664640934e-06, + "loss": 0.7897, + "step": 7353 + }, + { + "epoch": 0.6632096316003067, + "grad_norm": 0.7456170201733913, + "learning_rate": 1.0763084536372424e-06, + "loss": 0.604, + "step": 7354 + }, + { + "epoch": 0.6632998151237769, + "grad_norm": 1.8760941174207264, + "learning_rate": 1.0757903214480068e-06, + "loss": 0.7585, + "step": 7355 + }, + { + "epoch": 0.6633899986472471, + "grad_norm": 2.080031588754672, + "learning_rate": 1.0752722681174376e-06, + "loss": 0.6999, + "step": 7356 + }, + { + "epoch": 0.6634801821707174, + "grad_norm": 2.2136805472095546, + "learning_rate": 1.074754293689737e-06, + "loss": 0.673, + "step": 7357 + }, + { + "epoch": 0.6635703656941877, + "grad_norm": 1.7429176476962418, + "learning_rate": 1.0742363982091023e-06, + "loss": 0.7546, + "step": 7358 + }, + { + "epoch": 0.6636605492176579, + "grad_norm": 1.7818788455746357, + "learning_rate": 1.0737185817197215e-06, + "loss": 0.7076, + "step": 7359 + }, + { + "epoch": 0.6637507327411282, + "grad_norm": 1.7283797560919043, + "learning_rate": 1.0732008442657803e-06, + "loss": 0.7108, + "step": 7360 + }, + { + "epoch": 0.6638409162645985, + "grad_norm": 1.7865921355947798, + "learning_rate": 1.0726831858914516e-06, + "loss": 0.7352, + "step": 7361 + }, + { + "epoch": 0.6639310997880687, + "grad_norm": 1.7705842310409339, + "learning_rate": 1.0721656066409084e-06, + "loss": 0.7691, + "step": 7362 + }, + { + "epoch": 0.664021283311539, + "grad_norm": 2.6812281503209467, + "learning_rate": 1.0716481065583108e-06, + "loss": 0.7567, + "step": 7363 + }, + { + "epoch": 0.6641114668350092, + "grad_norm": 0.6493951780398657, + "learning_rate": 1.071130685687816e-06, + "loss": 0.5749, + "step": 7364 + }, + { + "epoch": 0.6642016503584796, + "grad_norm": 1.9003945769124826, + "learning_rate": 1.0706133440735723e-06, + "loss": 0.6878, + "step": 7365 + }, + { + "epoch": 0.6642918338819498, + "grad_norm": 3.1531424757262405, + "learning_rate": 1.070096081759723e-06, + "loss": 0.7009, + "step": 7366 + }, + { + "epoch": 0.66438201740542, + "grad_norm": 1.8111870054843728, + "learning_rate": 1.069578898790404e-06, + "loss": 0.8463, + "step": 7367 + }, + { + "epoch": 0.6644722009288903, + "grad_norm": 1.7634237966185873, + "learning_rate": 1.0690617952097424e-06, + "loss": 0.7333, + "step": 7368 + }, + { + "epoch": 0.6645623844523606, + "grad_norm": 1.7863136729697071, + "learning_rate": 1.068544771061863e-06, + "loss": 0.7962, + "step": 7369 + }, + { + "epoch": 0.6646525679758308, + "grad_norm": 2.3953602088710833, + "learning_rate": 1.0680278263908787e-06, + "loss": 0.7905, + "step": 7370 + }, + { + "epoch": 0.6647427514993011, + "grad_norm": 0.7907541478639516, + "learning_rate": 1.0675109612408991e-06, + "loss": 0.5847, + "step": 7371 + }, + { + "epoch": 0.6648329350227713, + "grad_norm": 3.4654640924026934, + "learning_rate": 1.0669941756560264e-06, + "loss": 0.721, + "step": 7372 + }, + { + "epoch": 0.6649231185462416, + "grad_norm": 2.4786465417121204, + "learning_rate": 1.0664774696803548e-06, + "loss": 0.6982, + "step": 7373 + }, + { + "epoch": 0.6650133020697119, + "grad_norm": 2.2542684121946857, + "learning_rate": 1.065960843357973e-06, + "loss": 0.7205, + "step": 7374 + }, + { + "epoch": 0.6651034855931821, + "grad_norm": 1.6192548134069502, + "learning_rate": 1.065444296732963e-06, + "loss": 0.766, + "step": 7375 + }, + { + "epoch": 0.6651936691166523, + "grad_norm": 0.6720577479382237, + "learning_rate": 1.064927829849397e-06, + "loss": 0.5871, + "step": 7376 + }, + { + "epoch": 0.6652838526401227, + "grad_norm": 3.1714038217037865, + "learning_rate": 1.0644114427513465e-06, + "loss": 0.7477, + "step": 7377 + }, + { + "epoch": 0.6653740361635929, + "grad_norm": 2.496178853907913, + "learning_rate": 1.0638951354828693e-06, + "loss": 0.6786, + "step": 7378 + }, + { + "epoch": 0.6654642196870632, + "grad_norm": 1.531673450697084, + "learning_rate": 1.063378908088021e-06, + "loss": 0.8032, + "step": 7379 + }, + { + "epoch": 0.6655544032105334, + "grad_norm": 1.750440133468767, + "learning_rate": 1.0628627606108486e-06, + "loss": 0.7392, + "step": 7380 + }, + { + "epoch": 0.6656445867340037, + "grad_norm": 2.204091286286689, + "learning_rate": 1.062346693095393e-06, + "loss": 0.8148, + "step": 7381 + }, + { + "epoch": 0.665734770257474, + "grad_norm": 1.54465857575764, + "learning_rate": 1.0618307055856882e-06, + "loss": 0.686, + "step": 7382 + }, + { + "epoch": 0.6658249537809442, + "grad_norm": 1.731636979133456, + "learning_rate": 1.061314798125759e-06, + "loss": 0.7985, + "step": 7383 + }, + { + "epoch": 0.6659151373044144, + "grad_norm": 1.4495550623972546, + "learning_rate": 1.0607989707596293e-06, + "loss": 0.7079, + "step": 7384 + }, + { + "epoch": 0.6660053208278848, + "grad_norm": 46.359261775224056, + "learning_rate": 1.0602832235313078e-06, + "loss": 0.8729, + "step": 7385 + }, + { + "epoch": 0.666095504351355, + "grad_norm": 1.463836160627453, + "learning_rate": 1.0597675564848053e-06, + "loss": 0.7361, + "step": 7386 + }, + { + "epoch": 0.6661856878748252, + "grad_norm": 1.8377917077649906, + "learning_rate": 1.059251969664118e-06, + "loss": 0.8165, + "step": 7387 + }, + { + "epoch": 0.6662758713982956, + "grad_norm": 13.113978235871006, + "learning_rate": 1.0587364631132402e-06, + "loss": 0.7576, + "step": 7388 + }, + { + "epoch": 0.6663660549217658, + "grad_norm": 1.4633887630788742, + "learning_rate": 1.0582210368761573e-06, + "loss": 0.7848, + "step": 7389 + }, + { + "epoch": 0.666456238445236, + "grad_norm": 8.438867528789455, + "learning_rate": 1.0577056909968485e-06, + "loss": 0.7509, + "step": 7390 + }, + { + "epoch": 0.6665464219687063, + "grad_norm": 1.7443784613525426, + "learning_rate": 1.0571904255192857e-06, + "loss": 0.7665, + "step": 7391 + }, + { + "epoch": 0.6666366054921766, + "grad_norm": 2.0164828581891308, + "learning_rate": 1.0566752404874354e-06, + "loss": 0.7821, + "step": 7392 + }, + { + "epoch": 0.6667267890156469, + "grad_norm": 1.8120099424632563, + "learning_rate": 1.0561601359452543e-06, + "loss": 0.7463, + "step": 7393 + }, + { + "epoch": 0.6668169725391171, + "grad_norm": 2.241118172780649, + "learning_rate": 1.0556451119366947e-06, + "loss": 0.6844, + "step": 7394 + }, + { + "epoch": 0.6669071560625873, + "grad_norm": 1.6174826974547678, + "learning_rate": 1.0551301685057011e-06, + "loss": 0.6896, + "step": 7395 + }, + { + "epoch": 0.6669973395860577, + "grad_norm": 1.965393656836619, + "learning_rate": 1.0546153056962117e-06, + "loss": 0.8349, + "step": 7396 + }, + { + "epoch": 0.6670875231095279, + "grad_norm": 1.8943070426489006, + "learning_rate": 1.0541005235521578e-06, + "loss": 0.7453, + "step": 7397 + }, + { + "epoch": 0.6671777066329981, + "grad_norm": 2.1615660000991004, + "learning_rate": 1.0535858221174614e-06, + "loss": 0.7654, + "step": 7398 + }, + { + "epoch": 0.6672678901564684, + "grad_norm": 1.4935077199413922, + "learning_rate": 1.0530712014360426e-06, + "loss": 0.7875, + "step": 7399 + }, + { + "epoch": 0.6673580736799387, + "grad_norm": 1.6693774668833488, + "learning_rate": 1.0525566615518088e-06, + "loss": 0.7992, + "step": 7400 + }, + { + "epoch": 0.667448257203409, + "grad_norm": 1.639330253150366, + "learning_rate": 1.0520422025086662e-06, + "loss": 0.7641, + "step": 7401 + }, + { + "epoch": 0.6675384407268792, + "grad_norm": 1.834981336576674, + "learning_rate": 1.0515278243505092e-06, + "loss": 0.8013, + "step": 7402 + }, + { + "epoch": 0.6676286242503494, + "grad_norm": 1.8746776303684736, + "learning_rate": 1.0510135271212278e-06, + "loss": 0.7787, + "step": 7403 + }, + { + "epoch": 0.6677188077738198, + "grad_norm": 1.3605395623272607, + "learning_rate": 1.0504993108647052e-06, + "loss": 0.8124, + "step": 7404 + }, + { + "epoch": 0.66780899129729, + "grad_norm": 1.5094417012326664, + "learning_rate": 1.0499851756248168e-06, + "loss": 0.725, + "step": 7405 + }, + { + "epoch": 0.6678991748207602, + "grad_norm": 1.6424339857146222, + "learning_rate": 1.0494711214454316e-06, + "loss": 0.7328, + "step": 7406 + }, + { + "epoch": 0.6679893583442305, + "grad_norm": 1.8937583998692111, + "learning_rate": 1.0489571483704111e-06, + "loss": 0.6646, + "step": 7407 + }, + { + "epoch": 0.6680795418677008, + "grad_norm": 1.7972970218150488, + "learning_rate": 1.048443256443612e-06, + "loss": 0.6678, + "step": 7408 + }, + { + "epoch": 0.668169725391171, + "grad_norm": 2.111868620651468, + "learning_rate": 1.0479294457088801e-06, + "loss": 0.7672, + "step": 7409 + }, + { + "epoch": 0.6682599089146413, + "grad_norm": 4.2317262367441755, + "learning_rate": 1.0474157162100574e-06, + "loss": 0.8703, + "step": 7410 + }, + { + "epoch": 0.6683500924381116, + "grad_norm": 1.7422535289346046, + "learning_rate": 1.0469020679909786e-06, + "loss": 0.6891, + "step": 7411 + }, + { + "epoch": 0.6684402759615818, + "grad_norm": 1.7366833832949706, + "learning_rate": 1.0463885010954705e-06, + "loss": 0.6726, + "step": 7412 + }, + { + "epoch": 0.6685304594850521, + "grad_norm": 1.8249533111661793, + "learning_rate": 1.0458750155673536e-06, + "loss": 0.8238, + "step": 7413 + }, + { + "epoch": 0.6686206430085223, + "grad_norm": 1.7870149848066361, + "learning_rate": 1.0453616114504421e-06, + "loss": 0.7037, + "step": 7414 + }, + { + "epoch": 0.6687108265319927, + "grad_norm": 1.5373406242325647, + "learning_rate": 1.0448482887885406e-06, + "loss": 0.7792, + "step": 7415 + }, + { + "epoch": 0.6688010100554629, + "grad_norm": 1.933656202136174, + "learning_rate": 1.044335047625451e-06, + "loss": 0.7578, + "step": 7416 + }, + { + "epoch": 0.6688911935789331, + "grad_norm": 2.540698747866596, + "learning_rate": 1.0438218880049637e-06, + "loss": 0.7634, + "step": 7417 + }, + { + "epoch": 0.6689813771024034, + "grad_norm": 1.9212747892668014, + "learning_rate": 1.0433088099708653e-06, + "loss": 0.8091, + "step": 7418 + }, + { + "epoch": 0.6690715606258737, + "grad_norm": 0.5977219439535025, + "learning_rate": 1.0427958135669346e-06, + "loss": 0.5126, + "step": 7419 + }, + { + "epoch": 0.6691617441493439, + "grad_norm": 1.7808583816220975, + "learning_rate": 1.0422828988369428e-06, + "loss": 0.6865, + "step": 7420 + }, + { + "epoch": 0.6692519276728142, + "grad_norm": 10.123615311493106, + "learning_rate": 1.041770065824655e-06, + "loss": 0.8197, + "step": 7421 + }, + { + "epoch": 0.6693421111962844, + "grad_norm": 1.6749335859426564, + "learning_rate": 1.0412573145738287e-06, + "loss": 0.6061, + "step": 7422 + }, + { + "epoch": 0.6694322947197547, + "grad_norm": 1.989693909879353, + "learning_rate": 1.040744645128216e-06, + "loss": 0.7242, + "step": 7423 + }, + { + "epoch": 0.669522478243225, + "grad_norm": 2.8137664486689418, + "learning_rate": 1.040232057531558e-06, + "loss": 0.8086, + "step": 7424 + }, + { + "epoch": 0.6696126617666952, + "grad_norm": 2.619943835551925, + "learning_rate": 1.0397195518275932e-06, + "loss": 0.7268, + "step": 7425 + }, + { + "epoch": 0.6697028452901654, + "grad_norm": 2.7426393142131236, + "learning_rate": 1.0392071280600512e-06, + "loss": 0.7192, + "step": 7426 + }, + { + "epoch": 0.6697930288136358, + "grad_norm": 1.7799655231608722, + "learning_rate": 1.0386947862726549e-06, + "loss": 0.7479, + "step": 7427 + }, + { + "epoch": 0.669883212337106, + "grad_norm": 1.9343493902085704, + "learning_rate": 1.0381825265091197e-06, + "loss": 0.6831, + "step": 7428 + }, + { + "epoch": 0.6699733958605762, + "grad_norm": 1.6855475822438406, + "learning_rate": 1.037670348813155e-06, + "loss": 0.6714, + "step": 7429 + }, + { + "epoch": 0.6700635793840465, + "grad_norm": 1.7524496372185847, + "learning_rate": 1.0371582532284624e-06, + "loss": 0.7805, + "step": 7430 + }, + { + "epoch": 0.6701537629075168, + "grad_norm": 1.577911891334036, + "learning_rate": 1.0366462397987375e-06, + "loss": 0.6831, + "step": 7431 + }, + { + "epoch": 0.6702439464309871, + "grad_norm": 2.044704655256544, + "learning_rate": 1.0361343085676665e-06, + "loss": 0.7569, + "step": 7432 + }, + { + "epoch": 0.6703341299544573, + "grad_norm": 1.8669947440916759, + "learning_rate": 1.0356224595789309e-06, + "loss": 0.8649, + "step": 7433 + }, + { + "epoch": 0.6704243134779276, + "grad_norm": 2.858031243764146, + "learning_rate": 1.0351106928762046e-06, + "loss": 0.8108, + "step": 7434 + }, + { + "epoch": 0.6705144970013979, + "grad_norm": 1.8893655412432768, + "learning_rate": 1.034599008503154e-06, + "loss": 0.8238, + "step": 7435 + }, + { + "epoch": 0.6706046805248681, + "grad_norm": 1.4495401771453909, + "learning_rate": 1.0340874065034406e-06, + "loss": 0.7406, + "step": 7436 + }, + { + "epoch": 0.6706948640483383, + "grad_norm": 2.0896628043014864, + "learning_rate": 1.0335758869207137e-06, + "loss": 0.6758, + "step": 7437 + }, + { + "epoch": 0.6707850475718087, + "grad_norm": 1.8890417736800131, + "learning_rate": 1.0330644497986227e-06, + "loss": 0.7738, + "step": 7438 + }, + { + "epoch": 0.6708752310952789, + "grad_norm": 1.8898256992520412, + "learning_rate": 1.0325530951808029e-06, + "loss": 0.7936, + "step": 7439 + }, + { + "epoch": 0.6709654146187491, + "grad_norm": 1.743146281859597, + "learning_rate": 1.0320418231108887e-06, + "loss": 0.7572, + "step": 7440 + }, + { + "epoch": 0.6710555981422194, + "grad_norm": 2.0490722510806996, + "learning_rate": 1.0315306336325028e-06, + "loss": 0.7942, + "step": 7441 + }, + { + "epoch": 0.6711457816656897, + "grad_norm": 2.511653157708135, + "learning_rate": 1.0310195267892635e-06, + "loss": 0.7076, + "step": 7442 + }, + { + "epoch": 0.67123596518916, + "grad_norm": 3.2344043380217578, + "learning_rate": 1.030508502624781e-06, + "loss": 0.697, + "step": 7443 + }, + { + "epoch": 0.6713261487126302, + "grad_norm": 1.8462756686792685, + "learning_rate": 1.0299975611826587e-06, + "loss": 0.6988, + "step": 7444 + }, + { + "epoch": 0.6714163322361004, + "grad_norm": 2.077568266443584, + "learning_rate": 1.0294867025064928e-06, + "loss": 0.7569, + "step": 7445 + }, + { + "epoch": 0.6715065157595708, + "grad_norm": 1.6339705161816276, + "learning_rate": 1.028975926639874e-06, + "loss": 0.7553, + "step": 7446 + }, + { + "epoch": 0.671596699283041, + "grad_norm": 2.8589839146719638, + "learning_rate": 1.0284652336263823e-06, + "loss": 0.6786, + "step": 7447 + }, + { + "epoch": 0.6716868828065112, + "grad_norm": 1.9983247655070286, + "learning_rate": 1.0279546235095938e-06, + "loss": 0.6809, + "step": 7448 + }, + { + "epoch": 0.6717770663299815, + "grad_norm": 0.739149319991025, + "learning_rate": 1.0274440963330768e-06, + "loss": 0.6089, + "step": 7449 + }, + { + "epoch": 0.6718672498534518, + "grad_norm": 1.6318299472043702, + "learning_rate": 1.0269336521403919e-06, + "loss": 0.7502, + "step": 7450 + }, + { + "epoch": 0.671957433376922, + "grad_norm": 1.893934143983905, + "learning_rate": 1.0264232909750936e-06, + "loss": 0.8777, + "step": 7451 + }, + { + "epoch": 0.6720476169003923, + "grad_norm": 3.971843685143753, + "learning_rate": 1.025913012880728e-06, + "loss": 0.8089, + "step": 7452 + }, + { + "epoch": 0.6721378004238625, + "grad_norm": 1.5346214391759192, + "learning_rate": 1.0254028179008362e-06, + "loss": 0.6612, + "step": 7453 + }, + { + "epoch": 0.6722279839473329, + "grad_norm": 2.492371077039461, + "learning_rate": 1.0248927060789483e-06, + "loss": 0.6418, + "step": 7454 + }, + { + "epoch": 0.6723181674708031, + "grad_norm": 1.9932361909337037, + "learning_rate": 1.0243826774585928e-06, + "loss": 0.7069, + "step": 7455 + }, + { + "epoch": 0.6724083509942733, + "grad_norm": 2.6817388662414925, + "learning_rate": 1.0238727320832854e-06, + "loss": 0.7372, + "step": 7456 + }, + { + "epoch": 0.6724985345177437, + "grad_norm": 1.8302847339516402, + "learning_rate": 1.0233628699965403e-06, + "loss": 0.756, + "step": 7457 + }, + { + "epoch": 0.6725887180412139, + "grad_norm": 1.5328266823644268, + "learning_rate": 1.0228530912418594e-06, + "loss": 0.7729, + "step": 7458 + }, + { + "epoch": 0.6726789015646841, + "grad_norm": 1.820015813528256, + "learning_rate": 1.0223433958627404e-06, + "loss": 0.7231, + "step": 7459 + }, + { + "epoch": 0.6727690850881544, + "grad_norm": 2.13954781487639, + "learning_rate": 1.021833783902674e-06, + "loss": 0.7686, + "step": 7460 + }, + { + "epoch": 0.6728592686116247, + "grad_norm": 1.9698926243831467, + "learning_rate": 1.0213242554051427e-06, + "loss": 0.7597, + "step": 7461 + }, + { + "epoch": 0.6729494521350949, + "grad_norm": 1.568731166528974, + "learning_rate": 1.0208148104136229e-06, + "loss": 0.6892, + "step": 7462 + }, + { + "epoch": 0.6730396356585652, + "grad_norm": 1.8564067196434035, + "learning_rate": 1.020305448971582e-06, + "loss": 0.7248, + "step": 7463 + }, + { + "epoch": 0.6731298191820354, + "grad_norm": 2.24216102292049, + "learning_rate": 1.0197961711224824e-06, + "loss": 0.715, + "step": 7464 + }, + { + "epoch": 0.6732200027055057, + "grad_norm": 2.0274061954120266, + "learning_rate": 1.0192869769097777e-06, + "loss": 0.7592, + "step": 7465 + }, + { + "epoch": 0.673310186228976, + "grad_norm": 1.5409051543697243, + "learning_rate": 1.018777866376916e-06, + "loss": 0.7621, + "step": 7466 + }, + { + "epoch": 0.6734003697524462, + "grad_norm": 1.7551507761165324, + "learning_rate": 1.0182688395673374e-06, + "loss": 0.6211, + "step": 7467 + }, + { + "epoch": 0.6734905532759164, + "grad_norm": 2.2472481323601086, + "learning_rate": 1.017759896524475e-06, + "loss": 0.7169, + "step": 7468 + }, + { + "epoch": 0.6735807367993868, + "grad_norm": 3.967618405474801, + "learning_rate": 1.0172510372917528e-06, + "loss": 0.7889, + "step": 7469 + }, + { + "epoch": 0.673670920322857, + "grad_norm": 3.519029791161414, + "learning_rate": 1.0167422619125925e-06, + "loss": 0.6539, + "step": 7470 + }, + { + "epoch": 0.6737611038463273, + "grad_norm": 1.7738616704971557, + "learning_rate": 1.0162335704304026e-06, + "loss": 0.75, + "step": 7471 + }, + { + "epoch": 0.6738512873697975, + "grad_norm": 0.7258981928732334, + "learning_rate": 1.0157249628885903e-06, + "loss": 0.5934, + "step": 7472 + }, + { + "epoch": 0.6739414708932678, + "grad_norm": 1.3019590343910061, + "learning_rate": 1.0152164393305506e-06, + "loss": 0.6562, + "step": 7473 + }, + { + "epoch": 0.6740316544167381, + "grad_norm": 1.811185689117454, + "learning_rate": 1.0147079997996746e-06, + "loss": 0.7418, + "step": 7474 + }, + { + "epoch": 0.6741218379402083, + "grad_norm": 1.6767990471596959, + "learning_rate": 1.0141996443393446e-06, + "loss": 0.7821, + "step": 7475 + }, + { + "epoch": 0.6742120214636785, + "grad_norm": 1.9951306313217207, + "learning_rate": 1.0136913729929369e-06, + "loss": 0.7377, + "step": 7476 + }, + { + "epoch": 0.6743022049871489, + "grad_norm": 1.5214581273058887, + "learning_rate": 1.0131831858038203e-06, + "loss": 0.7099, + "step": 7477 + }, + { + "epoch": 0.6743923885106191, + "grad_norm": 1.463976712802513, + "learning_rate": 1.0126750828153538e-06, + "loss": 0.8133, + "step": 7478 + }, + { + "epoch": 0.6744825720340893, + "grad_norm": 1.7590222074441637, + "learning_rate": 1.012167064070895e-06, + "loss": 0.7943, + "step": 7479 + }, + { + "epoch": 0.6745727555575597, + "grad_norm": 3.70009273206647, + "learning_rate": 1.0116591296137885e-06, + "loss": 0.7716, + "step": 7480 + }, + { + "epoch": 0.6746629390810299, + "grad_norm": 1.7810300724718469, + "learning_rate": 1.0111512794873746e-06, + "loss": 0.7742, + "step": 7481 + }, + { + "epoch": 0.6747531226045002, + "grad_norm": 1.5100291664894032, + "learning_rate": 1.010643513734986e-06, + "loss": 0.7486, + "step": 7482 + }, + { + "epoch": 0.6748433061279704, + "grad_norm": 1.8052561470944544, + "learning_rate": 1.010135832399948e-06, + "loss": 0.7894, + "step": 7483 + }, + { + "epoch": 0.6749334896514407, + "grad_norm": 1.3543621460111446, + "learning_rate": 1.0096282355255792e-06, + "loss": 0.7562, + "step": 7484 + }, + { + "epoch": 0.675023673174911, + "grad_norm": 9.626571725111226, + "learning_rate": 1.0091207231551905e-06, + "loss": 0.8054, + "step": 7485 + }, + { + "epoch": 0.6751138566983812, + "grad_norm": 1.5508190698119753, + "learning_rate": 1.0086132953320842e-06, + "loss": 0.7157, + "step": 7486 + }, + { + "epoch": 0.6752040402218514, + "grad_norm": 1.9920288621159714, + "learning_rate": 1.0081059520995591e-06, + "loss": 0.7778, + "step": 7487 + }, + { + "epoch": 0.6752942237453218, + "grad_norm": 1.8774130234972588, + "learning_rate": 1.0075986935009028e-06, + "loss": 0.8173, + "step": 7488 + }, + { + "epoch": 0.675384407268792, + "grad_norm": 1.738718659305654, + "learning_rate": 1.0070915195793982e-06, + "loss": 0.7621, + "step": 7489 + }, + { + "epoch": 0.6754745907922622, + "grad_norm": 2.0298264430226407, + "learning_rate": 1.0065844303783197e-06, + "loss": 0.7338, + "step": 7490 + }, + { + "epoch": 0.6755647743157325, + "grad_norm": 1.6615681467467922, + "learning_rate": 1.0060774259409356e-06, + "loss": 0.6668, + "step": 7491 + }, + { + "epoch": 0.6756549578392028, + "grad_norm": 1.5569949255427564, + "learning_rate": 1.0055705063105065e-06, + "loss": 0.7828, + "step": 7492 + }, + { + "epoch": 0.675745141362673, + "grad_norm": 2.1449534516519044, + "learning_rate": 1.0050636715302837e-06, + "loss": 0.7455, + "step": 7493 + }, + { + "epoch": 0.6758353248861433, + "grad_norm": 1.8980758385294314, + "learning_rate": 1.0045569216435157e-06, + "loss": 0.7676, + "step": 7494 + }, + { + "epoch": 0.6759255084096135, + "grad_norm": 1.5652313011525767, + "learning_rate": 1.0040502566934384e-06, + "loss": 0.8343, + "step": 7495 + }, + { + "epoch": 0.6760156919330839, + "grad_norm": 2.3869016486730725, + "learning_rate": 1.0035436767232866e-06, + "loss": 0.688, + "step": 7496 + }, + { + "epoch": 0.6761058754565541, + "grad_norm": 3.301587104615414, + "learning_rate": 1.0030371817762816e-06, + "loss": 0.7639, + "step": 7497 + }, + { + "epoch": 0.6761960589800243, + "grad_norm": 1.6084963427671928, + "learning_rate": 1.0025307718956417e-06, + "loss": 0.7574, + "step": 7498 + }, + { + "epoch": 0.6762862425034946, + "grad_norm": 1.9274866317898964, + "learning_rate": 1.0020244471245765e-06, + "loss": 0.7727, + "step": 7499 + }, + { + "epoch": 0.6763764260269649, + "grad_norm": 1.8583659150670078, + "learning_rate": 1.001518207506288e-06, + "loss": 0.7038, + "step": 7500 + }, + { + "epoch": 0.6764666095504351, + "grad_norm": 1.5213871387363822, + "learning_rate": 1.0010120530839717e-06, + "loss": 0.8131, + "step": 7501 + }, + { + "epoch": 0.6765567930739054, + "grad_norm": 1.2207205076887362, + "learning_rate": 1.0005059839008161e-06, + "loss": 0.7755, + "step": 7502 + }, + { + "epoch": 0.6766469765973756, + "grad_norm": 1.86450602905739, + "learning_rate": 1.0000000000000004e-06, + "loss": 0.7044, + "step": 7503 + }, + { + "epoch": 0.676737160120846, + "grad_norm": 1.713455924659141, + "learning_rate": 9.994941014246985e-07, + "loss": 0.7944, + "step": 7504 + }, + { + "epoch": 0.6768273436443162, + "grad_norm": 1.6489211543557383, + "learning_rate": 9.989882882180766e-07, + "loss": 0.6411, + "step": 7505 + }, + { + "epoch": 0.6769175271677864, + "grad_norm": 1.8212919424407525, + "learning_rate": 9.984825604232938e-07, + "loss": 0.769, + "step": 7506 + }, + { + "epoch": 0.6770077106912568, + "grad_norm": 1.8161118135437622, + "learning_rate": 9.97976918083502e-07, + "loss": 0.6867, + "step": 7507 + }, + { + "epoch": 0.677097894214727, + "grad_norm": 3.1127616802961824, + "learning_rate": 9.974713612418427e-07, + "loss": 0.8202, + "step": 7508 + }, + { + "epoch": 0.6771880777381972, + "grad_norm": 2.30758164580649, + "learning_rate": 9.969658899414563e-07, + "loss": 0.766, + "step": 7509 + }, + { + "epoch": 0.6772782612616675, + "grad_norm": 2.0632817491894038, + "learning_rate": 9.964605042254696e-07, + "loss": 0.6746, + "step": 7510 + }, + { + "epoch": 0.6773684447851378, + "grad_norm": 4.3198809278820525, + "learning_rate": 9.959552041370076e-07, + "loss": 0.6925, + "step": 7511 + }, + { + "epoch": 0.677458628308608, + "grad_norm": 3.3585896838266267, + "learning_rate": 9.954499897191824e-07, + "loss": 0.6708, + "step": 7512 + }, + { + "epoch": 0.6775488118320783, + "grad_norm": 2.225427170798743, + "learning_rate": 9.949448610151043e-07, + "loss": 0.7048, + "step": 7513 + }, + { + "epoch": 0.6776389953555485, + "grad_norm": 1.8625481080876263, + "learning_rate": 9.944398180678719e-07, + "loss": 0.734, + "step": 7514 + }, + { + "epoch": 0.6777291788790188, + "grad_norm": 1.7675608597223083, + "learning_rate": 9.939348609205789e-07, + "loss": 0.7011, + "step": 7515 + }, + { + "epoch": 0.6778193624024891, + "grad_norm": 1.4113788746594882, + "learning_rate": 9.93429989616311e-07, + "loss": 0.7582, + "step": 7516 + }, + { + "epoch": 0.6779095459259593, + "grad_norm": 1.6369283537382497, + "learning_rate": 9.929252041981464e-07, + "loss": 0.6656, + "step": 7517 + }, + { + "epoch": 0.6779997294494295, + "grad_norm": 1.8592397175789612, + "learning_rate": 9.924205047091572e-07, + "loss": 0.7249, + "step": 7518 + }, + { + "epoch": 0.6780899129728999, + "grad_norm": 1.7072593135068581, + "learning_rate": 9.919158911924056e-07, + "loss": 0.8724, + "step": 7519 + }, + { + "epoch": 0.6781800964963701, + "grad_norm": 1.7293998913340194, + "learning_rate": 9.914113636909483e-07, + "loss": 0.7858, + "step": 7520 + }, + { + "epoch": 0.6782702800198404, + "grad_norm": 0.7678688023084371, + "learning_rate": 9.90906922247835e-07, + "loss": 0.6279, + "step": 7521 + }, + { + "epoch": 0.6783604635433106, + "grad_norm": 1.6827193465523222, + "learning_rate": 9.904025669061072e-07, + "loss": 0.7796, + "step": 7522 + }, + { + "epoch": 0.6784506470667809, + "grad_norm": 2.127541985103106, + "learning_rate": 9.89898297708799e-07, + "loss": 0.7456, + "step": 7523 + }, + { + "epoch": 0.6785408305902512, + "grad_norm": 4.305368469986456, + "learning_rate": 9.893941146989388e-07, + "loss": 0.7079, + "step": 7524 + }, + { + "epoch": 0.6786310141137214, + "grad_norm": 1.8896492576104487, + "learning_rate": 9.888900179195437e-07, + "loss": 0.7117, + "step": 7525 + }, + { + "epoch": 0.6787211976371916, + "grad_norm": 1.8166612671704792, + "learning_rate": 9.883860074136285e-07, + "loss": 0.8196, + "step": 7526 + }, + { + "epoch": 0.678811381160662, + "grad_norm": 3.5836603917340843, + "learning_rate": 9.87882083224196e-07, + "loss": 0.7458, + "step": 7527 + }, + { + "epoch": 0.6789015646841322, + "grad_norm": 0.6865507421584949, + "learning_rate": 9.873782453942462e-07, + "loss": 0.6045, + "step": 7528 + }, + { + "epoch": 0.6789917482076024, + "grad_norm": 1.7513936488004265, + "learning_rate": 9.868744939667676e-07, + "loss": 0.6771, + "step": 7529 + }, + { + "epoch": 0.6790819317310728, + "grad_norm": 1.4955743829409012, + "learning_rate": 9.863708289847432e-07, + "loss": 0.6325, + "step": 7530 + }, + { + "epoch": 0.679172115254543, + "grad_norm": 2.0068016030746, + "learning_rate": 9.85867250491149e-07, + "loss": 0.6638, + "step": 7531 + }, + { + "epoch": 0.6792622987780133, + "grad_norm": 2.0275995170356382, + "learning_rate": 9.853637585289528e-07, + "loss": 0.857, + "step": 7532 + }, + { + "epoch": 0.6793524823014835, + "grad_norm": 1.676071888722489, + "learning_rate": 9.848603531411159e-07, + "loss": 0.7393, + "step": 7533 + }, + { + "epoch": 0.6794426658249538, + "grad_norm": 2.023726512638336, + "learning_rate": 9.843570343705899e-07, + "loss": 0.761, + "step": 7534 + }, + { + "epoch": 0.6795328493484241, + "grad_norm": 1.436832687730596, + "learning_rate": 9.83853802260323e-07, + "loss": 0.8227, + "step": 7535 + }, + { + "epoch": 0.6796230328718943, + "grad_norm": 1.7751193369118372, + "learning_rate": 9.833506568532524e-07, + "loss": 0.6984, + "step": 7536 + }, + { + "epoch": 0.6797132163953645, + "grad_norm": 1.760445483571889, + "learning_rate": 9.828475981923093e-07, + "loss": 0.5987, + "step": 7537 + }, + { + "epoch": 0.6798033999188349, + "grad_norm": 2.3592881950931943, + "learning_rate": 9.823446263204175e-07, + "loss": 0.7288, + "step": 7538 + }, + { + "epoch": 0.6798935834423051, + "grad_norm": 1.7229149033869153, + "learning_rate": 9.818417412804937e-07, + "loss": 0.7614, + "step": 7539 + }, + { + "epoch": 0.6799837669657753, + "grad_norm": 2.591422541937736, + "learning_rate": 9.813389431154463e-07, + "loss": 0.8303, + "step": 7540 + }, + { + "epoch": 0.6800739504892456, + "grad_norm": 7.833954427999978, + "learning_rate": 9.808362318681783e-07, + "loss": 0.7799, + "step": 7541 + }, + { + "epoch": 0.6801641340127159, + "grad_norm": 2.9097061116317047, + "learning_rate": 9.803336075815807e-07, + "loss": 0.7724, + "step": 7542 + }, + { + "epoch": 0.6802543175361861, + "grad_norm": 3.006156961186985, + "learning_rate": 9.79831070298544e-07, + "loss": 0.6797, + "step": 7543 + }, + { + "epoch": 0.6803445010596564, + "grad_norm": 1.7763498063204424, + "learning_rate": 9.793286200619443e-07, + "loss": 0.7243, + "step": 7544 + }, + { + "epoch": 0.6804346845831266, + "grad_norm": 1.6648809005480372, + "learning_rate": 9.78826256914655e-07, + "loss": 0.8114, + "step": 7545 + }, + { + "epoch": 0.680524868106597, + "grad_norm": 1.7264062456912461, + "learning_rate": 9.7832398089954e-07, + "loss": 0.8008, + "step": 7546 + }, + { + "epoch": 0.6806150516300672, + "grad_norm": 1.562183271731633, + "learning_rate": 9.778217920594565e-07, + "loss": 0.8538, + "step": 7547 + }, + { + "epoch": 0.6807052351535374, + "grad_norm": 1.9270926879105899, + "learning_rate": 9.773196904372547e-07, + "loss": 0.7941, + "step": 7548 + }, + { + "epoch": 0.6807954186770077, + "grad_norm": 5.2100505388558025, + "learning_rate": 9.768176760757742e-07, + "loss": 0.8379, + "step": 7549 + }, + { + "epoch": 0.680885602200478, + "grad_norm": 1.9853647722291725, + "learning_rate": 9.76315749017853e-07, + "loss": 0.7286, + "step": 7550 + }, + { + "epoch": 0.6809757857239482, + "grad_norm": 1.5398055088785498, + "learning_rate": 9.758139093063161e-07, + "loss": 0.7531, + "step": 7551 + }, + { + "epoch": 0.6810659692474185, + "grad_norm": 2.5986647405028442, + "learning_rate": 9.753121569839834e-07, + "loss": 0.822, + "step": 7552 + }, + { + "epoch": 0.6811561527708888, + "grad_norm": 0.747112557207858, + "learning_rate": 9.748104920936678e-07, + "loss": 0.5485, + "step": 7553 + }, + { + "epoch": 0.681246336294359, + "grad_norm": 1.7144954388486322, + "learning_rate": 9.743089146781738e-07, + "loss": 0.7313, + "step": 7554 + }, + { + "epoch": 0.6813365198178293, + "grad_norm": 2.7421605470207626, + "learning_rate": 9.738074247802988e-07, + "loss": 0.6777, + "step": 7555 + }, + { + "epoch": 0.6814267033412995, + "grad_norm": 1.675882901040851, + "learning_rate": 9.733060224428325e-07, + "loss": 0.7683, + "step": 7556 + }, + { + "epoch": 0.6815168868647699, + "grad_norm": 2.3294465071478, + "learning_rate": 9.728047077085577e-07, + "loss": 0.7614, + "step": 7557 + }, + { + "epoch": 0.6816070703882401, + "grad_norm": 3.756335945859899, + "learning_rate": 9.723034806202497e-07, + "loss": 0.6527, + "step": 7558 + }, + { + "epoch": 0.6816972539117103, + "grad_norm": 2.230063166044482, + "learning_rate": 9.718023412206748e-07, + "loss": 0.7698, + "step": 7559 + }, + { + "epoch": 0.6817874374351806, + "grad_norm": 1.3463264870765086, + "learning_rate": 9.713012895525935e-07, + "loss": 0.7755, + "step": 7560 + }, + { + "epoch": 0.6818776209586509, + "grad_norm": 2.062312146503996, + "learning_rate": 9.708003256587584e-07, + "loss": 0.739, + "step": 7561 + }, + { + "epoch": 0.6819678044821211, + "grad_norm": 1.594924568736436, + "learning_rate": 9.702994495819147e-07, + "loss": 0.798, + "step": 7562 + }, + { + "epoch": 0.6820579880055914, + "grad_norm": 1.931857105617773, + "learning_rate": 9.697986613647999e-07, + "loss": 0.6583, + "step": 7563 + }, + { + "epoch": 0.6821481715290616, + "grad_norm": 0.6997212783204223, + "learning_rate": 9.692979610501425e-07, + "loss": 0.6134, + "step": 7564 + }, + { + "epoch": 0.6822383550525319, + "grad_norm": 1.860307603852365, + "learning_rate": 9.68797348680668e-07, + "loss": 0.824, + "step": 7565 + }, + { + "epoch": 0.6823285385760022, + "grad_norm": 1.5610086571833974, + "learning_rate": 9.682968242990878e-07, + "loss": 0.7201, + "step": 7566 + }, + { + "epoch": 0.6824187220994724, + "grad_norm": 1.5337863080267033, + "learning_rate": 9.677963879481132e-07, + "loss": 0.8014, + "step": 7567 + }, + { + "epoch": 0.6825089056229426, + "grad_norm": 2.2291908144014316, + "learning_rate": 9.672960396704416e-07, + "loss": 0.747, + "step": 7568 + }, + { + "epoch": 0.682599089146413, + "grad_norm": 1.7964741135262896, + "learning_rate": 9.667957795087657e-07, + "loss": 0.7524, + "step": 7569 + }, + { + "epoch": 0.6826892726698832, + "grad_norm": 2.0948486861617996, + "learning_rate": 9.662956075057712e-07, + "loss": 0.8158, + "step": 7570 + }, + { + "epoch": 0.6827794561933535, + "grad_norm": 2.5459828130497035, + "learning_rate": 9.657955237041354e-07, + "loss": 0.8161, + "step": 7571 + }, + { + "epoch": 0.6828696397168237, + "grad_norm": 1.768071867548007, + "learning_rate": 9.652955281465278e-07, + "loss": 0.736, + "step": 7572 + }, + { + "epoch": 0.682959823240294, + "grad_norm": 1.8188315330661218, + "learning_rate": 9.64795620875612e-07, + "loss": 0.7556, + "step": 7573 + }, + { + "epoch": 0.6830500067637643, + "grad_norm": 1.5531508936728735, + "learning_rate": 9.64295801934041e-07, + "loss": 0.6964, + "step": 7574 + }, + { + "epoch": 0.6831401902872345, + "grad_norm": 0.6662645965322753, + "learning_rate": 9.63796071364463e-07, + "loss": 0.5883, + "step": 7575 + }, + { + "epoch": 0.6832303738107048, + "grad_norm": 1.9973899380341178, + "learning_rate": 9.632964292095179e-07, + "loss": 0.8345, + "step": 7576 + }, + { + "epoch": 0.6833205573341751, + "grad_norm": 1.41562081982667, + "learning_rate": 9.627968755118374e-07, + "loss": 0.7399, + "step": 7577 + }, + { + "epoch": 0.6834107408576453, + "grad_norm": 0.6668812560384196, + "learning_rate": 9.622974103140468e-07, + "loss": 0.5684, + "step": 7578 + }, + { + "epoch": 0.6835009243811155, + "grad_norm": 2.322551225275938, + "learning_rate": 9.617980336587632e-07, + "loss": 0.8145, + "step": 7579 + }, + { + "epoch": 0.6835911079045859, + "grad_norm": 2.863869724976285, + "learning_rate": 9.612987455885964e-07, + "loss": 0.6328, + "step": 7580 + }, + { + "epoch": 0.6836812914280561, + "grad_norm": 1.7213512503348696, + "learning_rate": 9.607995461461467e-07, + "loss": 0.7559, + "step": 7581 + }, + { + "epoch": 0.6837714749515263, + "grad_norm": 1.9017586875647892, + "learning_rate": 9.603004353740111e-07, + "loss": 0.6483, + "step": 7582 + }, + { + "epoch": 0.6838616584749966, + "grad_norm": 0.6945301969882804, + "learning_rate": 9.598014133147738e-07, + "loss": 0.6227, + "step": 7583 + }, + { + "epoch": 0.6839518419984669, + "grad_norm": 1.5776379277405612, + "learning_rate": 9.59302480011017e-07, + "loss": 0.713, + "step": 7584 + }, + { + "epoch": 0.6840420255219372, + "grad_norm": 0.6744665466266948, + "learning_rate": 9.588036355053102e-07, + "loss": 0.5987, + "step": 7585 + }, + { + "epoch": 0.6841322090454074, + "grad_norm": 2.0150046170089495, + "learning_rate": 9.583048798402182e-07, + "loss": 0.6224, + "step": 7586 + }, + { + "epoch": 0.6842223925688776, + "grad_norm": 2.326164700411375, + "learning_rate": 9.57806213058298e-07, + "loss": 0.7469, + "step": 7587 + }, + { + "epoch": 0.684312576092348, + "grad_norm": 0.6958657538880862, + "learning_rate": 9.57307635202098e-07, + "loss": 0.6164, + "step": 7588 + }, + { + "epoch": 0.6844027596158182, + "grad_norm": 1.6039352332616703, + "learning_rate": 9.568091463141607e-07, + "loss": 0.8549, + "step": 7589 + }, + { + "epoch": 0.6844929431392884, + "grad_norm": 1.5426636985831583, + "learning_rate": 9.563107464370187e-07, + "loss": 0.718, + "step": 7590 + }, + { + "epoch": 0.6845831266627587, + "grad_norm": 1.6583668119159825, + "learning_rate": 9.558124356131982e-07, + "loss": 0.7665, + "step": 7591 + }, + { + "epoch": 0.684673310186229, + "grad_norm": 1.8396069805602568, + "learning_rate": 9.553142138852187e-07, + "loss": 0.6875, + "step": 7592 + }, + { + "epoch": 0.6847634937096992, + "grad_norm": 1.920320890453312, + "learning_rate": 9.548160812955905e-07, + "loss": 0.6582, + "step": 7593 + }, + { + "epoch": 0.6848536772331695, + "grad_norm": 1.96163952180182, + "learning_rate": 9.543180378868175e-07, + "loss": 0.5754, + "step": 7594 + }, + { + "epoch": 0.6849438607566397, + "grad_norm": 1.8695252918574092, + "learning_rate": 9.538200837013962e-07, + "loss": 0.7892, + "step": 7595 + }, + { + "epoch": 0.68503404428011, + "grad_norm": 0.6615449192916768, + "learning_rate": 9.533222187818122e-07, + "loss": 0.6155, + "step": 7596 + }, + { + "epoch": 0.6851242278035803, + "grad_norm": 4.326083696900453, + "learning_rate": 9.528244431705492e-07, + "loss": 0.7599, + "step": 7597 + }, + { + "epoch": 0.6852144113270505, + "grad_norm": 4.6300135064803465, + "learning_rate": 9.523267569100774e-07, + "loss": 0.758, + "step": 7598 + }, + { + "epoch": 0.6853045948505209, + "grad_norm": 2.6118581686141655, + "learning_rate": 9.518291600428652e-07, + "loss": 0.7078, + "step": 7599 + }, + { + "epoch": 0.6853947783739911, + "grad_norm": 2.248480071512426, + "learning_rate": 9.513316526113677e-07, + "loss": 0.7776, + "step": 7600 + }, + { + "epoch": 0.6854849618974613, + "grad_norm": 1.8230536127875274, + "learning_rate": 9.50834234658036e-07, + "loss": 0.7746, + "step": 7601 + }, + { + "epoch": 0.6855751454209316, + "grad_norm": 1.9413807673480068, + "learning_rate": 9.503369062253123e-07, + "loss": 0.7674, + "step": 7602 + }, + { + "epoch": 0.6856653289444019, + "grad_norm": 0.6458212687278152, + "learning_rate": 9.498396673556317e-07, + "loss": 0.5908, + "step": 7603 + }, + { + "epoch": 0.6857555124678721, + "grad_norm": 1.4405160858622612, + "learning_rate": 9.493425180914219e-07, + "loss": 0.7859, + "step": 7604 + }, + { + "epoch": 0.6858456959913424, + "grad_norm": 3.06252319949958, + "learning_rate": 9.488454584751e-07, + "loss": 0.6811, + "step": 7605 + }, + { + "epoch": 0.6859358795148126, + "grad_norm": 1.492092988381479, + "learning_rate": 9.483484885490813e-07, + "loss": 0.7551, + "step": 7606 + }, + { + "epoch": 0.686026063038283, + "grad_norm": 2.3074083722108405, + "learning_rate": 9.478516083557675e-07, + "loss": 0.7635, + "step": 7607 + }, + { + "epoch": 0.6861162465617532, + "grad_norm": 5.250366198165967, + "learning_rate": 9.473548179375561e-07, + "loss": 0.7674, + "step": 7608 + }, + { + "epoch": 0.6862064300852234, + "grad_norm": 0.681249446168727, + "learning_rate": 9.468581173368358e-07, + "loss": 0.6171, + "step": 7609 + }, + { + "epoch": 0.6862966136086937, + "grad_norm": 1.628320236361624, + "learning_rate": 9.463615065959878e-07, + "loss": 0.7587, + "step": 7610 + }, + { + "epoch": 0.686386797132164, + "grad_norm": 4.435818205731258, + "learning_rate": 9.458649857573857e-07, + "loss": 0.8015, + "step": 7611 + }, + { + "epoch": 0.6864769806556342, + "grad_norm": 1.5362746651565136, + "learning_rate": 9.453685548633963e-07, + "loss": 0.7578, + "step": 7612 + }, + { + "epoch": 0.6865671641791045, + "grad_norm": 2.0899605691648975, + "learning_rate": 9.448722139563756e-07, + "loss": 0.7018, + "step": 7613 + }, + { + "epoch": 0.6866573477025747, + "grad_norm": 2.3700533097142458, + "learning_rate": 9.443759630786769e-07, + "loss": 0.652, + "step": 7614 + }, + { + "epoch": 0.686747531226045, + "grad_norm": 1.6829193959690576, + "learning_rate": 9.438798022726408e-07, + "loss": 0.8233, + "step": 7615 + }, + { + "epoch": 0.6868377147495153, + "grad_norm": 2.1985111487225004, + "learning_rate": 9.433837315806037e-07, + "loss": 0.708, + "step": 7616 + }, + { + "epoch": 0.6869278982729855, + "grad_norm": 1.8127579998681058, + "learning_rate": 9.428877510448925e-07, + "loss": 0.7087, + "step": 7617 + }, + { + "epoch": 0.6870180817964557, + "grad_norm": 1.9789839074597086, + "learning_rate": 9.423918607078272e-07, + "loss": 0.6811, + "step": 7618 + }, + { + "epoch": 0.6871082653199261, + "grad_norm": 1.5653348002765373, + "learning_rate": 9.418960606117208e-07, + "loss": 0.7593, + "step": 7619 + }, + { + "epoch": 0.6871984488433963, + "grad_norm": 1.92157234351912, + "learning_rate": 9.414003507988752e-07, + "loss": 0.6937, + "step": 7620 + }, + { + "epoch": 0.6872886323668665, + "grad_norm": 1.6408895370107974, + "learning_rate": 9.409047313115904e-07, + "loss": 0.6848, + "step": 7621 + }, + { + "epoch": 0.6873788158903368, + "grad_norm": 1.994492039367901, + "learning_rate": 9.404092021921521e-07, + "loss": 0.6937, + "step": 7622 + }, + { + "epoch": 0.6874689994138071, + "grad_norm": 1.8214878986217393, + "learning_rate": 9.399137634828447e-07, + "loss": 0.7307, + "step": 7623 + }, + { + "epoch": 0.6875591829372774, + "grad_norm": 1.5999101255924173, + "learning_rate": 9.394184152259396e-07, + "loss": 0.7761, + "step": 7624 + }, + { + "epoch": 0.6876493664607476, + "grad_norm": 1.5964540467952884, + "learning_rate": 9.389231574637033e-07, + "loss": 0.8563, + "step": 7625 + }, + { + "epoch": 0.6877395499842179, + "grad_norm": 1.7255008246907044, + "learning_rate": 9.384279902383938e-07, + "loss": 0.8357, + "step": 7626 + }, + { + "epoch": 0.6878297335076882, + "grad_norm": 6.028921043315562, + "learning_rate": 9.379329135922615e-07, + "loss": 0.7717, + "step": 7627 + }, + { + "epoch": 0.6879199170311584, + "grad_norm": 1.6380905920881095, + "learning_rate": 9.374379275675495e-07, + "loss": 0.7098, + "step": 7628 + }, + { + "epoch": 0.6880101005546286, + "grad_norm": 2.6619012293673157, + "learning_rate": 9.369430322064931e-07, + "loss": 0.7453, + "step": 7629 + }, + { + "epoch": 0.688100284078099, + "grad_norm": 12.818785377129998, + "learning_rate": 9.364482275513179e-07, + "loss": 0.8324, + "step": 7630 + }, + { + "epoch": 0.6881904676015692, + "grad_norm": 2.330854927520319, + "learning_rate": 9.359535136442444e-07, + "loss": 0.8154, + "step": 7631 + }, + { + "epoch": 0.6882806511250394, + "grad_norm": 1.6784336822017474, + "learning_rate": 9.354588905274843e-07, + "loss": 0.6863, + "step": 7632 + }, + { + "epoch": 0.6883708346485097, + "grad_norm": 1.5858589491159323, + "learning_rate": 9.349643582432414e-07, + "loss": 0.7849, + "step": 7633 + }, + { + "epoch": 0.68846101817198, + "grad_norm": 1.6279234631470314, + "learning_rate": 9.344699168337127e-07, + "loss": 0.7457, + "step": 7634 + }, + { + "epoch": 0.6885512016954503, + "grad_norm": 1.5630353392000242, + "learning_rate": 9.339755663410845e-07, + "loss": 0.7189, + "step": 7635 + }, + { + "epoch": 0.6886413852189205, + "grad_norm": 1.7298986731606785, + "learning_rate": 9.334813068075405e-07, + "loss": 0.7853, + "step": 7636 + }, + { + "epoch": 0.6887315687423907, + "grad_norm": 1.3895087315990946, + "learning_rate": 9.329871382752506e-07, + "loss": 0.757, + "step": 7637 + }, + { + "epoch": 0.6888217522658611, + "grad_norm": 1.6201973884326968, + "learning_rate": 9.32493060786383e-07, + "loss": 0.7422, + "step": 7638 + }, + { + "epoch": 0.6889119357893313, + "grad_norm": 1.510888238713308, + "learning_rate": 9.31999074383093e-07, + "loss": 0.8097, + "step": 7639 + }, + { + "epoch": 0.6890021193128015, + "grad_norm": 2.7477233305837627, + "learning_rate": 9.315051791075308e-07, + "loss": 0.7486, + "step": 7640 + }, + { + "epoch": 0.6890923028362718, + "grad_norm": 1.5001215090809965, + "learning_rate": 9.310113750018382e-07, + "loss": 0.7438, + "step": 7641 + }, + { + "epoch": 0.6891824863597421, + "grad_norm": 1.916506269903001, + "learning_rate": 9.305176621081496e-07, + "loss": 0.8205, + "step": 7642 + }, + { + "epoch": 0.6892726698832123, + "grad_norm": 1.5677929303060099, + "learning_rate": 9.300240404685911e-07, + "loss": 0.7851, + "step": 7643 + }, + { + "epoch": 0.6893628534066826, + "grad_norm": 1.8322445352842118, + "learning_rate": 9.295305101252812e-07, + "loss": 0.754, + "step": 7644 + }, + { + "epoch": 0.6894530369301528, + "grad_norm": 1.9913120515378124, + "learning_rate": 9.290370711203314e-07, + "loss": 0.7358, + "step": 7645 + }, + { + "epoch": 0.6895432204536232, + "grad_norm": 2.8011276903392766, + "learning_rate": 9.285437234958433e-07, + "loss": 0.7457, + "step": 7646 + }, + { + "epoch": 0.6896334039770934, + "grad_norm": 3.2203144604687597, + "learning_rate": 9.280504672939124e-07, + "loss": 0.744, + "step": 7647 + }, + { + "epoch": 0.6897235875005636, + "grad_norm": 3.094556606991068, + "learning_rate": 9.275573025566266e-07, + "loss": 0.7908, + "step": 7648 + }, + { + "epoch": 0.689813771024034, + "grad_norm": 3.0439316838301735, + "learning_rate": 9.27064229326065e-07, + "loss": 0.7825, + "step": 7649 + }, + { + "epoch": 0.6899039545475042, + "grad_norm": 1.7341832922756135, + "learning_rate": 9.265712476442995e-07, + "loss": 0.7183, + "step": 7650 + }, + { + "epoch": 0.6899941380709744, + "grad_norm": 2.000765058101152, + "learning_rate": 9.260783575533949e-07, + "loss": 0.6821, + "step": 7651 + }, + { + "epoch": 0.6900843215944447, + "grad_norm": 1.5725915417005771, + "learning_rate": 9.255855590954045e-07, + "loss": 0.7011, + "step": 7652 + }, + { + "epoch": 0.690174505117915, + "grad_norm": 3.674729556558247, + "learning_rate": 9.250928523123802e-07, + "loss": 0.8178, + "step": 7653 + }, + { + "epoch": 0.6902646886413852, + "grad_norm": 1.4580168880274726, + "learning_rate": 9.24600237246359e-07, + "loss": 0.7593, + "step": 7654 + }, + { + "epoch": 0.6903548721648555, + "grad_norm": 1.9629096945979783, + "learning_rate": 9.241077139393769e-07, + "loss": 0.7611, + "step": 7655 + }, + { + "epoch": 0.6904450556883257, + "grad_norm": 5.453635292989031, + "learning_rate": 9.236152824334564e-07, + "loss": 0.7816, + "step": 7656 + }, + { + "epoch": 0.690535239211796, + "grad_norm": 11.321063262285564, + "learning_rate": 9.231229427706151e-07, + "loss": 0.6498, + "step": 7657 + }, + { + "epoch": 0.6906254227352663, + "grad_norm": 1.7419626407318047, + "learning_rate": 9.226306949928622e-07, + "loss": 0.7568, + "step": 7658 + }, + { + "epoch": 0.6907156062587365, + "grad_norm": 1.6382547604698197, + "learning_rate": 9.221385391421988e-07, + "loss": 0.827, + "step": 7659 + }, + { + "epoch": 0.6908057897822067, + "grad_norm": 2.673736102552709, + "learning_rate": 9.216464752606192e-07, + "loss": 0.7673, + "step": 7660 + }, + { + "epoch": 0.6908959733056771, + "grad_norm": 1.659135715739429, + "learning_rate": 9.211545033901078e-07, + "loss": 0.8289, + "step": 7661 + }, + { + "epoch": 0.6909861568291473, + "grad_norm": 1.7601540123137547, + "learning_rate": 9.206626235726426e-07, + "loss": 0.7836, + "step": 7662 + }, + { + "epoch": 0.6910763403526176, + "grad_norm": 2.5802486112117955, + "learning_rate": 9.20170835850194e-07, + "loss": 0.689, + "step": 7663 + }, + { + "epoch": 0.6911665238760878, + "grad_norm": 0.943544989543737, + "learning_rate": 9.196791402647237e-07, + "loss": 0.573, + "step": 7664 + }, + { + "epoch": 0.6912567073995581, + "grad_norm": 1.77463017492849, + "learning_rate": 9.191875368581861e-07, + "loss": 0.676, + "step": 7665 + }, + { + "epoch": 0.6913468909230284, + "grad_norm": 1.7972673070808936, + "learning_rate": 9.186960256725271e-07, + "loss": 0.7652, + "step": 7666 + }, + { + "epoch": 0.6914370744464986, + "grad_norm": 1.711442916283693, + "learning_rate": 9.182046067496856e-07, + "loss": 0.8182, + "step": 7667 + }, + { + "epoch": 0.6915272579699688, + "grad_norm": 2.43275449434978, + "learning_rate": 9.177132801315927e-07, + "loss": 0.8366, + "step": 7668 + }, + { + "epoch": 0.6916174414934392, + "grad_norm": 2.1977925800411726, + "learning_rate": 9.172220458601692e-07, + "loss": 0.7258, + "step": 7669 + }, + { + "epoch": 0.6917076250169094, + "grad_norm": 2.6869699709555546, + "learning_rate": 9.167309039773324e-07, + "loss": 0.7177, + "step": 7670 + }, + { + "epoch": 0.6917978085403796, + "grad_norm": 1.3032652360942794, + "learning_rate": 9.162398545249872e-07, + "loss": 0.6477, + "step": 7671 + }, + { + "epoch": 0.69188799206385, + "grad_norm": 1.302248016115707, + "learning_rate": 9.157488975450334e-07, + "loss": 0.8068, + "step": 7672 + }, + { + "epoch": 0.6919781755873202, + "grad_norm": 1.5770530459251728, + "learning_rate": 9.15258033079362e-07, + "loss": 0.5798, + "step": 7673 + }, + { + "epoch": 0.6920683591107905, + "grad_norm": 1.6502883427994888, + "learning_rate": 9.147672611698567e-07, + "loss": 0.7409, + "step": 7674 + }, + { + "epoch": 0.6921585426342607, + "grad_norm": 1.899131867694056, + "learning_rate": 9.142765818583933e-07, + "loss": 0.7467, + "step": 7675 + }, + { + "epoch": 0.692248726157731, + "grad_norm": 1.5361086000157933, + "learning_rate": 9.13785995186837e-07, + "loss": 0.7298, + "step": 7676 + }, + { + "epoch": 0.6923389096812013, + "grad_norm": 2.8403890681724837, + "learning_rate": 9.132955011970502e-07, + "loss": 0.7173, + "step": 7677 + }, + { + "epoch": 0.6924290932046715, + "grad_norm": 1.2708099967606572, + "learning_rate": 9.128050999308827e-07, + "loss": 0.7392, + "step": 7678 + }, + { + "epoch": 0.6925192767281417, + "grad_norm": 3.0077362447212783, + "learning_rate": 9.123147914301789e-07, + "loss": 0.8432, + "step": 7679 + }, + { + "epoch": 0.6926094602516121, + "grad_norm": 2.3137509854323373, + "learning_rate": 9.118245757367745e-07, + "loss": 0.8006, + "step": 7680 + }, + { + "epoch": 0.6926996437750823, + "grad_norm": 0.6194301133025718, + "learning_rate": 9.113344528924973e-07, + "loss": 0.5605, + "step": 7681 + }, + { + "epoch": 0.6927898272985525, + "grad_norm": 1.4153365828742877, + "learning_rate": 9.108444229391676e-07, + "loss": 0.7324, + "step": 7682 + }, + { + "epoch": 0.6928800108220228, + "grad_norm": 1.844930044879387, + "learning_rate": 9.103544859185979e-07, + "loss": 0.7277, + "step": 7683 + }, + { + "epoch": 0.6929701943454931, + "grad_norm": 2.3239698773662627, + "learning_rate": 9.098646418725902e-07, + "loss": 0.8383, + "step": 7684 + }, + { + "epoch": 0.6930603778689634, + "grad_norm": 0.6421167868411412, + "learning_rate": 9.093748908429437e-07, + "loss": 0.6031, + "step": 7685 + }, + { + "epoch": 0.6931505613924336, + "grad_norm": 1.6224450052401427, + "learning_rate": 9.088852328714444e-07, + "loss": 0.7398, + "step": 7686 + }, + { + "epoch": 0.6932407449159038, + "grad_norm": 1.4756024181119691, + "learning_rate": 9.083956679998735e-07, + "loss": 0.6602, + "step": 7687 + }, + { + "epoch": 0.6933309284393742, + "grad_norm": 1.3548476585390756, + "learning_rate": 9.079061962700032e-07, + "loss": 0.7171, + "step": 7688 + }, + { + "epoch": 0.6934211119628444, + "grad_norm": 1.4796215038627254, + "learning_rate": 9.074168177235979e-07, + "loss": 0.772, + "step": 7689 + }, + { + "epoch": 0.6935112954863146, + "grad_norm": 1.6670693467378273, + "learning_rate": 9.069275324024151e-07, + "loss": 0.7625, + "step": 7690 + }, + { + "epoch": 0.6936014790097849, + "grad_norm": 14.924092296812864, + "learning_rate": 9.064383403482005e-07, + "loss": 0.6801, + "step": 7691 + }, + { + "epoch": 0.6936916625332552, + "grad_norm": 1.4385363947693253, + "learning_rate": 9.059492416026983e-07, + "loss": 0.7123, + "step": 7692 + }, + { + "epoch": 0.6937818460567254, + "grad_norm": 1.5325257572666735, + "learning_rate": 9.054602362076378e-07, + "loss": 0.7459, + "step": 7693 + }, + { + "epoch": 0.6938720295801957, + "grad_norm": 4.277062811835561, + "learning_rate": 9.049713242047468e-07, + "loss": 0.7407, + "step": 7694 + }, + { + "epoch": 0.693962213103666, + "grad_norm": 2.608912261318088, + "learning_rate": 9.044825056357395e-07, + "loss": 0.7214, + "step": 7695 + }, + { + "epoch": 0.6940523966271362, + "grad_norm": 1.5265025288135763, + "learning_rate": 9.039937805423255e-07, + "loss": 0.7062, + "step": 7696 + }, + { + "epoch": 0.6941425801506065, + "grad_norm": 0.5980486859195134, + "learning_rate": 9.035051489662051e-07, + "loss": 0.5174, + "step": 7697 + }, + { + "epoch": 0.6942327636740767, + "grad_norm": 1.8052520529472913, + "learning_rate": 9.030166109490718e-07, + "loss": 0.7014, + "step": 7698 + }, + { + "epoch": 0.6943229471975471, + "grad_norm": 5.681680573617721, + "learning_rate": 9.025281665326099e-07, + "loss": 0.7621, + "step": 7699 + }, + { + "epoch": 0.6944131307210173, + "grad_norm": 2.478716181583522, + "learning_rate": 9.020398157584967e-07, + "loss": 0.7619, + "step": 7700 + }, + { + "epoch": 0.6945033142444875, + "grad_norm": 1.7341746309028163, + "learning_rate": 9.015515586684002e-07, + "loss": 0.6526, + "step": 7701 + }, + { + "epoch": 0.6945934977679578, + "grad_norm": 0.6599069158340537, + "learning_rate": 9.010633953039812e-07, + "loss": 0.5937, + "step": 7702 + }, + { + "epoch": 0.6946836812914281, + "grad_norm": 1.476526492048135, + "learning_rate": 9.005753257068929e-07, + "loss": 0.7617, + "step": 7703 + }, + { + "epoch": 0.6947738648148983, + "grad_norm": 1.7979385545561541, + "learning_rate": 9.000873499187797e-07, + "loss": 0.7647, + "step": 7704 + }, + { + "epoch": 0.6948640483383686, + "grad_norm": 1.5538580133729294, + "learning_rate": 8.995994679812797e-07, + "loss": 0.7814, + "step": 7705 + }, + { + "epoch": 0.6949542318618388, + "grad_norm": 9.819264406828596, + "learning_rate": 8.991116799360192e-07, + "loss": 0.7738, + "step": 7706 + }, + { + "epoch": 0.6950444153853091, + "grad_norm": 1.673337849019456, + "learning_rate": 8.986239858246217e-07, + "loss": 0.746, + "step": 7707 + }, + { + "epoch": 0.6951345989087794, + "grad_norm": 1.9109610205986187, + "learning_rate": 8.981363856886972e-07, + "loss": 0.8345, + "step": 7708 + }, + { + "epoch": 0.6952247824322496, + "grad_norm": 1.7431387592263665, + "learning_rate": 8.976488795698533e-07, + "loss": 0.7345, + "step": 7709 + }, + { + "epoch": 0.6953149659557198, + "grad_norm": 1.452888141319604, + "learning_rate": 8.971614675096841e-07, + "loss": 0.7634, + "step": 7710 + }, + { + "epoch": 0.6954051494791902, + "grad_norm": 2.45701931463839, + "learning_rate": 8.966741495497807e-07, + "loss": 0.817, + "step": 7711 + }, + { + "epoch": 0.6954953330026604, + "grad_norm": 1.572056499339473, + "learning_rate": 8.961869257317218e-07, + "loss": 0.7163, + "step": 7712 + }, + { + "epoch": 0.6955855165261307, + "grad_norm": 1.7493154004115974, + "learning_rate": 8.956997960970809e-07, + "loss": 0.7351, + "step": 7713 + }, + { + "epoch": 0.6956757000496009, + "grad_norm": 9.105234137367862, + "learning_rate": 8.952127606874224e-07, + "loss": 0.6623, + "step": 7714 + }, + { + "epoch": 0.6957658835730712, + "grad_norm": 1.590419205878515, + "learning_rate": 8.947258195443028e-07, + "loss": 0.7525, + "step": 7715 + }, + { + "epoch": 0.6958560670965415, + "grad_norm": 2.089625381041602, + "learning_rate": 8.942389727092716e-07, + "loss": 0.7366, + "step": 7716 + }, + { + "epoch": 0.6959462506200117, + "grad_norm": 0.6774862973210491, + "learning_rate": 8.937522202238677e-07, + "loss": 0.5648, + "step": 7717 + }, + { + "epoch": 0.696036434143482, + "grad_norm": 2.302586105012382, + "learning_rate": 8.932655621296239e-07, + "loss": 0.7675, + "step": 7718 + }, + { + "epoch": 0.6961266176669523, + "grad_norm": 1.4292483463060928, + "learning_rate": 8.927789984680649e-07, + "loss": 0.7483, + "step": 7719 + }, + { + "epoch": 0.6962168011904225, + "grad_norm": 1.7516891637870375, + "learning_rate": 8.922925292807068e-07, + "loss": 0.6221, + "step": 7720 + }, + { + "epoch": 0.6963069847138927, + "grad_norm": 1.5968854392470055, + "learning_rate": 8.91806154609058e-07, + "loss": 0.8338, + "step": 7721 + }, + { + "epoch": 0.6963971682373631, + "grad_norm": 1.6777220826592394, + "learning_rate": 8.913198744946195e-07, + "loss": 0.7212, + "step": 7722 + }, + { + "epoch": 0.6964873517608333, + "grad_norm": 1.6494246867025733, + "learning_rate": 8.908336889788807e-07, + "loss": 0.774, + "step": 7723 + }, + { + "epoch": 0.6965775352843036, + "grad_norm": 1.727338935641753, + "learning_rate": 8.903475981033293e-07, + "loss": 0.6291, + "step": 7724 + }, + { + "epoch": 0.6966677188077738, + "grad_norm": 0.7413665648311599, + "learning_rate": 8.898616019094376e-07, + "loss": 0.6325, + "step": 7725 + }, + { + "epoch": 0.6967579023312441, + "grad_norm": 1.9793827608159336, + "learning_rate": 8.89375700438677e-07, + "loss": 0.791, + "step": 7726 + }, + { + "epoch": 0.6968480858547144, + "grad_norm": 1.903898508512453, + "learning_rate": 8.888898937325047e-07, + "loss": 0.7146, + "step": 7727 + }, + { + "epoch": 0.6969382693781846, + "grad_norm": 2.0406105917265673, + "learning_rate": 8.884041818323733e-07, + "loss": 0.7174, + "step": 7728 + }, + { + "epoch": 0.6970284529016548, + "grad_norm": 2.9132630425127433, + "learning_rate": 8.879185647797262e-07, + "loss": 0.7867, + "step": 7729 + }, + { + "epoch": 0.6971186364251252, + "grad_norm": 1.689535890878188, + "learning_rate": 8.874330426159993e-07, + "loss": 0.7325, + "step": 7730 + }, + { + "epoch": 0.6972088199485954, + "grad_norm": 1.4513281921560837, + "learning_rate": 8.869476153826205e-07, + "loss": 0.773, + "step": 7731 + }, + { + "epoch": 0.6972990034720656, + "grad_norm": 2.5893740220819605, + "learning_rate": 8.864622831210071e-07, + "loss": 0.7502, + "step": 7732 + }, + { + "epoch": 0.6973891869955359, + "grad_norm": 2.1524856437532214, + "learning_rate": 8.85977045872573e-07, + "loss": 0.7569, + "step": 7733 + }, + { + "epoch": 0.6974793705190062, + "grad_norm": 1.7399612366534958, + "learning_rate": 8.854919036787194e-07, + "loss": 0.7307, + "step": 7734 + }, + { + "epoch": 0.6975695540424764, + "grad_norm": 1.694201619653487, + "learning_rate": 8.850068565808417e-07, + "loss": 0.7158, + "step": 7735 + }, + { + "epoch": 0.6976597375659467, + "grad_norm": 2.0321840926154544, + "learning_rate": 8.845219046203271e-07, + "loss": 0.7957, + "step": 7736 + }, + { + "epoch": 0.6977499210894169, + "grad_norm": 2.4798917805224945, + "learning_rate": 8.840370478385544e-07, + "loss": 0.7772, + "step": 7737 + }, + { + "epoch": 0.6978401046128873, + "grad_norm": 1.6996037891026987, + "learning_rate": 8.83552286276894e-07, + "loss": 0.7664, + "step": 7738 + }, + { + "epoch": 0.6979302881363575, + "grad_norm": 1.9316641378948434, + "learning_rate": 8.830676199767095e-07, + "loss": 0.7713, + "step": 7739 + }, + { + "epoch": 0.6980204716598277, + "grad_norm": 1.5594556523872987, + "learning_rate": 8.825830489793527e-07, + "loss": 0.7356, + "step": 7740 + }, + { + "epoch": 0.698110655183298, + "grad_norm": 1.3750551819565635, + "learning_rate": 8.820985733261732e-07, + "loss": 0.708, + "step": 7741 + }, + { + "epoch": 0.6982008387067683, + "grad_norm": 1.7337355938610641, + "learning_rate": 8.816141930585066e-07, + "loss": 0.6627, + "step": 7742 + }, + { + "epoch": 0.6982910222302385, + "grad_norm": 0.6925422043798745, + "learning_rate": 8.811299082176837e-07, + "loss": 0.5901, + "step": 7743 + }, + { + "epoch": 0.6983812057537088, + "grad_norm": 2.483939366563507, + "learning_rate": 8.806457188450265e-07, + "loss": 0.7115, + "step": 7744 + }, + { + "epoch": 0.6984713892771791, + "grad_norm": 1.800349980344737, + "learning_rate": 8.801616249818487e-07, + "loss": 0.7873, + "step": 7745 + }, + { + "epoch": 0.6985615728006493, + "grad_norm": 1.8390823376152026, + "learning_rate": 8.796776266694564e-07, + "loss": 0.73, + "step": 7746 + }, + { + "epoch": 0.6986517563241196, + "grad_norm": 1.7736156222927417, + "learning_rate": 8.79193723949145e-07, + "loss": 0.6959, + "step": 7747 + }, + { + "epoch": 0.6987419398475898, + "grad_norm": 1.685769571420678, + "learning_rate": 8.787099168622063e-07, + "loss": 0.6441, + "step": 7748 + }, + { + "epoch": 0.6988321233710602, + "grad_norm": 1.6639565447534255, + "learning_rate": 8.782262054499199e-07, + "loss": 0.6913, + "step": 7749 + }, + { + "epoch": 0.6989223068945304, + "grad_norm": 1.310776123371528, + "learning_rate": 8.777425897535588e-07, + "loss": 0.781, + "step": 7750 + }, + { + "epoch": 0.6990124904180006, + "grad_norm": 8.034225684447307, + "learning_rate": 8.77259069814388e-07, + "loss": 0.7559, + "step": 7751 + }, + { + "epoch": 0.6991026739414709, + "grad_norm": 2.8578942639150005, + "learning_rate": 8.767756456736641e-07, + "loss": 0.7494, + "step": 7752 + }, + { + "epoch": 0.6991928574649412, + "grad_norm": 2.2645213628979612, + "learning_rate": 8.762923173726358e-07, + "loss": 0.6753, + "step": 7753 + }, + { + "epoch": 0.6992830409884114, + "grad_norm": 1.850431201273301, + "learning_rate": 8.758090849525428e-07, + "loss": 0.8467, + "step": 7754 + }, + { + "epoch": 0.6993732245118817, + "grad_norm": 1.783863693174741, + "learning_rate": 8.753259484546174e-07, + "loss": 0.6921, + "step": 7755 + }, + { + "epoch": 0.6994634080353519, + "grad_norm": 1.7706128450643674, + "learning_rate": 8.748429079200841e-07, + "loss": 0.7095, + "step": 7756 + }, + { + "epoch": 0.6995535915588222, + "grad_norm": 1.9089507060773026, + "learning_rate": 8.743599633901575e-07, + "loss": 0.8113, + "step": 7757 + }, + { + "epoch": 0.6996437750822925, + "grad_norm": 1.9139105288840548, + "learning_rate": 8.738771149060453e-07, + "loss": 0.7632, + "step": 7758 + }, + { + "epoch": 0.6997339586057627, + "grad_norm": 2.117011594327045, + "learning_rate": 8.73394362508947e-07, + "loss": 0.7181, + "step": 7759 + }, + { + "epoch": 0.6998241421292329, + "grad_norm": 1.409846179295886, + "learning_rate": 8.72911706240054e-07, + "loss": 0.7121, + "step": 7760 + }, + { + "epoch": 0.6999143256527033, + "grad_norm": 3.6176214431592837, + "learning_rate": 8.724291461405493e-07, + "loss": 0.6288, + "step": 7761 + }, + { + "epoch": 0.7000045091761735, + "grad_norm": 2.048322908060947, + "learning_rate": 8.71946682251606e-07, + "loss": 0.7187, + "step": 7762 + }, + { + "epoch": 0.7000946926996438, + "grad_norm": 1.569615982981236, + "learning_rate": 8.714643146143932e-07, + "loss": 0.7472, + "step": 7763 + }, + { + "epoch": 0.700184876223114, + "grad_norm": 1.4763676751998054, + "learning_rate": 8.709820432700663e-07, + "loss": 0.7058, + "step": 7764 + }, + { + "epoch": 0.7002750597465843, + "grad_norm": 1.4410524011491275, + "learning_rate": 8.704998682597784e-07, + "loss": 0.6637, + "step": 7765 + }, + { + "epoch": 0.7003652432700546, + "grad_norm": 2.1417694373897977, + "learning_rate": 8.700177896246688e-07, + "loss": 0.7695, + "step": 7766 + }, + { + "epoch": 0.7004554267935248, + "grad_norm": 1.9294196186152108, + "learning_rate": 8.695358074058721e-07, + "loss": 0.6788, + "step": 7767 + }, + { + "epoch": 0.7005456103169951, + "grad_norm": 2.0506598736402872, + "learning_rate": 8.690539216445136e-07, + "loss": 0.789, + "step": 7768 + }, + { + "epoch": 0.7006357938404654, + "grad_norm": 10.915983300016999, + "learning_rate": 8.685721323817106e-07, + "loss": 0.6517, + "step": 7769 + }, + { + "epoch": 0.7007259773639356, + "grad_norm": 1.6053048674387322, + "learning_rate": 8.680904396585718e-07, + "loss": 0.7835, + "step": 7770 + }, + { + "epoch": 0.7008161608874058, + "grad_norm": 1.7094423911916719, + "learning_rate": 8.676088435161988e-07, + "loss": 0.688, + "step": 7771 + }, + { + "epoch": 0.7009063444108762, + "grad_norm": 1.8494049249299032, + "learning_rate": 8.671273439956824e-07, + "loss": 0.7687, + "step": 7772 + }, + { + "epoch": 0.7009965279343464, + "grad_norm": 1.4691355584728427, + "learning_rate": 8.666459411381075e-07, + "loss": 0.752, + "step": 7773 + }, + { + "epoch": 0.7010867114578166, + "grad_norm": 1.3575042480065074, + "learning_rate": 8.661646349845501e-07, + "loss": 0.7739, + "step": 7774 + }, + { + "epoch": 0.7011768949812869, + "grad_norm": 2.4838641139224755, + "learning_rate": 8.656834255760783e-07, + "loss": 0.8087, + "step": 7775 + }, + { + "epoch": 0.7012670785047572, + "grad_norm": 1.6484294005280877, + "learning_rate": 8.652023129537509e-07, + "loss": 0.7998, + "step": 7776 + }, + { + "epoch": 0.7013572620282275, + "grad_norm": 3.8040101625324807, + "learning_rate": 8.647212971586195e-07, + "loss": 0.7779, + "step": 7777 + }, + { + "epoch": 0.7014474455516977, + "grad_norm": 1.7147611628568247, + "learning_rate": 8.642403782317275e-07, + "loss": 0.7051, + "step": 7778 + }, + { + "epoch": 0.7015376290751679, + "grad_norm": 1.7592526788012184, + "learning_rate": 8.637595562141075e-07, + "loss": 0.7538, + "step": 7779 + }, + { + "epoch": 0.7016278125986383, + "grad_norm": 1.5066811067316808, + "learning_rate": 8.632788311467889e-07, + "loss": 0.7145, + "step": 7780 + }, + { + "epoch": 0.7017179961221085, + "grad_norm": 1.495093347320748, + "learning_rate": 8.627982030707867e-07, + "loss": 0.7422, + "step": 7781 + }, + { + "epoch": 0.7018081796455787, + "grad_norm": 1.9442199335214994, + "learning_rate": 8.623176720271139e-07, + "loss": 0.8144, + "step": 7782 + }, + { + "epoch": 0.701898363169049, + "grad_norm": 2.8446392303900714, + "learning_rate": 8.618372380567696e-07, + "loss": 0.8189, + "step": 7783 + }, + { + "epoch": 0.7019885466925193, + "grad_norm": 1.9518370387147725, + "learning_rate": 8.613569012007478e-07, + "loss": 0.7063, + "step": 7784 + }, + { + "epoch": 0.7020787302159895, + "grad_norm": 2.143916140866378, + "learning_rate": 8.608766615000338e-07, + "loss": 0.7874, + "step": 7785 + }, + { + "epoch": 0.7021689137394598, + "grad_norm": 1.8109406307479745, + "learning_rate": 8.603965189956039e-07, + "loss": 0.7716, + "step": 7786 + }, + { + "epoch": 0.70225909726293, + "grad_norm": 1.739966649141446, + "learning_rate": 8.599164737284276e-07, + "loss": 0.754, + "step": 7787 + }, + { + "epoch": 0.7023492807864004, + "grad_norm": 1.7617418917808634, + "learning_rate": 8.594365257394634e-07, + "loss": 0.7076, + "step": 7788 + }, + { + "epoch": 0.7024394643098706, + "grad_norm": 1.7084190610759646, + "learning_rate": 8.589566750696637e-07, + "loss": 0.6947, + "step": 7789 + }, + { + "epoch": 0.7025296478333408, + "grad_norm": 2.0565099719969986, + "learning_rate": 8.584769217599721e-07, + "loss": 0.6828, + "step": 7790 + }, + { + "epoch": 0.7026198313568112, + "grad_norm": 2.2902894361065713, + "learning_rate": 8.579972658513239e-07, + "loss": 0.6945, + "step": 7791 + }, + { + "epoch": 0.7027100148802814, + "grad_norm": 2.14078927627951, + "learning_rate": 8.57517707384646e-07, + "loss": 0.7529, + "step": 7792 + }, + { + "epoch": 0.7028001984037516, + "grad_norm": 3.2714990962821764, + "learning_rate": 8.570382464008574e-07, + "loss": 0.7881, + "step": 7793 + }, + { + "epoch": 0.7028903819272219, + "grad_norm": 1.7087019855978214, + "learning_rate": 8.565588829408665e-07, + "loss": 0.7127, + "step": 7794 + }, + { + "epoch": 0.7029805654506922, + "grad_norm": 2.079324935159034, + "learning_rate": 8.560796170455782e-07, + "loss": 0.7246, + "step": 7795 + }, + { + "epoch": 0.7030707489741624, + "grad_norm": 1.8936846227660213, + "learning_rate": 8.556004487558828e-07, + "loss": 0.7857, + "step": 7796 + }, + { + "epoch": 0.7031609324976327, + "grad_norm": 1.5546022324685789, + "learning_rate": 8.55121378112669e-07, + "loss": 0.6633, + "step": 7797 + }, + { + "epoch": 0.7032511160211029, + "grad_norm": 2.0989160146935513, + "learning_rate": 8.546424051568111e-07, + "loss": 0.7268, + "step": 7798 + }, + { + "epoch": 0.7033412995445733, + "grad_norm": 1.6778587144244501, + "learning_rate": 8.541635299291785e-07, + "loss": 0.7969, + "step": 7799 + }, + { + "epoch": 0.7034314830680435, + "grad_norm": 0.5849091979821394, + "learning_rate": 8.536847524706317e-07, + "loss": 0.6002, + "step": 7800 + }, + { + "epoch": 0.7035216665915137, + "grad_norm": 1.6044841633814877, + "learning_rate": 8.532060728220225e-07, + "loss": 0.6787, + "step": 7801 + }, + { + "epoch": 0.703611850114984, + "grad_norm": 3.464138918366258, + "learning_rate": 8.527274910241955e-07, + "loss": 0.7563, + "step": 7802 + }, + { + "epoch": 0.7037020336384543, + "grad_norm": 1.7408930327297139, + "learning_rate": 8.522490071179833e-07, + "loss": 0.7421, + "step": 7803 + }, + { + "epoch": 0.7037922171619245, + "grad_norm": 3.6868575231484186, + "learning_rate": 8.517706211442159e-07, + "loss": 0.6296, + "step": 7804 + }, + { + "epoch": 0.7038824006853948, + "grad_norm": 3.6856882849543986, + "learning_rate": 8.512923331437097e-07, + "loss": 0.826, + "step": 7805 + }, + { + "epoch": 0.703972584208865, + "grad_norm": 1.5963129864889058, + "learning_rate": 8.508141431572755e-07, + "loss": 0.7168, + "step": 7806 + }, + { + "epoch": 0.7040627677323353, + "grad_norm": 2.1530122636090527, + "learning_rate": 8.503360512257152e-07, + "loss": 0.7347, + "step": 7807 + }, + { + "epoch": 0.7041529512558056, + "grad_norm": 2.973967615080789, + "learning_rate": 8.498580573898219e-07, + "loss": 0.715, + "step": 7808 + }, + { + "epoch": 0.7042431347792758, + "grad_norm": 1.9980881016450263, + "learning_rate": 8.493801616903813e-07, + "loss": 0.7579, + "step": 7809 + }, + { + "epoch": 0.704333318302746, + "grad_norm": 2.070220484578115, + "learning_rate": 8.489023641681705e-07, + "loss": 0.769, + "step": 7810 + }, + { + "epoch": 0.7044235018262164, + "grad_norm": 1.5601066570012072, + "learning_rate": 8.484246648639555e-07, + "loss": 0.7953, + "step": 7811 + }, + { + "epoch": 0.7045136853496866, + "grad_norm": 1.5264239651498095, + "learning_rate": 8.479470638184994e-07, + "loss": 0.6264, + "step": 7812 + }, + { + "epoch": 0.7046038688731568, + "grad_norm": 4.296077251940271, + "learning_rate": 8.474695610725513e-07, + "loss": 0.6684, + "step": 7813 + }, + { + "epoch": 0.7046940523966272, + "grad_norm": 1.4244371741577648, + "learning_rate": 8.469921566668552e-07, + "loss": 0.8204, + "step": 7814 + }, + { + "epoch": 0.7047842359200974, + "grad_norm": 1.7895586416948788, + "learning_rate": 8.46514850642146e-07, + "loss": 0.7973, + "step": 7815 + }, + { + "epoch": 0.7048744194435677, + "grad_norm": 1.9158625643661267, + "learning_rate": 8.460376430391499e-07, + "loss": 0.735, + "step": 7816 + }, + { + "epoch": 0.7049646029670379, + "grad_norm": 1.2970434332246563, + "learning_rate": 8.455605338985858e-07, + "loss": 0.7762, + "step": 7817 + }, + { + "epoch": 0.7050547864905082, + "grad_norm": 1.6055933400466929, + "learning_rate": 8.45083523261161e-07, + "loss": 0.6062, + "step": 7818 + }, + { + "epoch": 0.7051449700139785, + "grad_norm": 1.5124569059175506, + "learning_rate": 8.446066111675796e-07, + "loss": 0.7524, + "step": 7819 + }, + { + "epoch": 0.7052351535374487, + "grad_norm": 2.20240111667038, + "learning_rate": 8.441297976585314e-07, + "loss": 0.6433, + "step": 7820 + }, + { + "epoch": 0.7053253370609189, + "grad_norm": 2.031498937758164, + "learning_rate": 8.436530827747037e-07, + "loss": 0.8302, + "step": 7821 + }, + { + "epoch": 0.7054155205843893, + "grad_norm": 1.8986935540976133, + "learning_rate": 8.431764665567704e-07, + "loss": 0.8806, + "step": 7822 + }, + { + "epoch": 0.7055057041078595, + "grad_norm": 2.434258481960087, + "learning_rate": 8.426999490453996e-07, + "loss": 0.6881, + "step": 7823 + }, + { + "epoch": 0.7055958876313297, + "grad_norm": 1.9815611826566375, + "learning_rate": 8.422235302812504e-07, + "loss": 0.756, + "step": 7824 + }, + { + "epoch": 0.7056860711548, + "grad_norm": 1.9858758612190455, + "learning_rate": 8.417472103049734e-07, + "loss": 0.8177, + "step": 7825 + }, + { + "epoch": 0.7057762546782703, + "grad_norm": 1.875936528602309, + "learning_rate": 8.412709891572112e-07, + "loss": 0.7833, + "step": 7826 + }, + { + "epoch": 0.7058664382017406, + "grad_norm": 1.9580033071565996, + "learning_rate": 8.407948668785978e-07, + "loss": 0.7837, + "step": 7827 + }, + { + "epoch": 0.7059566217252108, + "grad_norm": 2.227935260756705, + "learning_rate": 8.403188435097576e-07, + "loss": 0.7397, + "step": 7828 + }, + { + "epoch": 0.706046805248681, + "grad_norm": 1.571036935591832, + "learning_rate": 8.398429190913081e-07, + "loss": 0.8184, + "step": 7829 + }, + { + "epoch": 0.7061369887721514, + "grad_norm": 1.9047894745489053, + "learning_rate": 8.393670936638578e-07, + "loss": 0.7469, + "step": 7830 + }, + { + "epoch": 0.7062271722956216, + "grad_norm": 11.781278493831591, + "learning_rate": 8.388913672680067e-07, + "loss": 0.7428, + "step": 7831 + }, + { + "epoch": 0.7063173558190918, + "grad_norm": 2.105999202115885, + "learning_rate": 8.384157399443472e-07, + "loss": 0.6981, + "step": 7832 + }, + { + "epoch": 0.7064075393425621, + "grad_norm": 1.6710645904065675, + "learning_rate": 8.379402117334601e-07, + "loss": 0.7758, + "step": 7833 + }, + { + "epoch": 0.7064977228660324, + "grad_norm": 1.8283877061718066, + "learning_rate": 8.374647826759232e-07, + "loss": 0.7462, + "step": 7834 + }, + { + "epoch": 0.7065879063895026, + "grad_norm": 1.552598095557487, + "learning_rate": 8.369894528122998e-07, + "loss": 0.7412, + "step": 7835 + }, + { + "epoch": 0.7066780899129729, + "grad_norm": 1.7695084795634144, + "learning_rate": 8.365142221831505e-07, + "loss": 0.6081, + "step": 7836 + }, + { + "epoch": 0.7067682734364432, + "grad_norm": 1.9703500089168293, + "learning_rate": 8.360390908290222e-07, + "loss": 0.711, + "step": 7837 + }, + { + "epoch": 0.7068584569599135, + "grad_norm": 1.6435263274857492, + "learning_rate": 8.355640587904569e-07, + "loss": 0.7278, + "step": 7838 + }, + { + "epoch": 0.7069486404833837, + "grad_norm": 4.6054277491018425, + "learning_rate": 8.350891261079866e-07, + "loss": 0.7545, + "step": 7839 + }, + { + "epoch": 0.7070388240068539, + "grad_norm": 1.684942497381564, + "learning_rate": 8.346142928221356e-07, + "loss": 0.799, + "step": 7840 + }, + { + "epoch": 0.7071290075303243, + "grad_norm": 1.8557672200319382, + "learning_rate": 8.341395589734189e-07, + "loss": 0.5938, + "step": 7841 + }, + { + "epoch": 0.7072191910537945, + "grad_norm": 2.2635865610945545, + "learning_rate": 8.336649246023433e-07, + "loss": 0.7072, + "step": 7842 + }, + { + "epoch": 0.7073093745772647, + "grad_norm": 1.6387281078561085, + "learning_rate": 8.331903897494084e-07, + "loss": 0.6525, + "step": 7843 + }, + { + "epoch": 0.707399558100735, + "grad_norm": 2.0408907473385294, + "learning_rate": 8.327159544551024e-07, + "loss": 0.7613, + "step": 7844 + }, + { + "epoch": 0.7074897416242053, + "grad_norm": 5.322825909442413, + "learning_rate": 8.322416187599073e-07, + "loss": 0.7077, + "step": 7845 + }, + { + "epoch": 0.7075799251476755, + "grad_norm": 1.7746569100423908, + "learning_rate": 8.317673827042963e-07, + "loss": 0.7745, + "step": 7846 + }, + { + "epoch": 0.7076701086711458, + "grad_norm": 1.881423247714941, + "learning_rate": 8.312932463287339e-07, + "loss": 0.7379, + "step": 7847 + }, + { + "epoch": 0.707760292194616, + "grad_norm": 2.2071709242652955, + "learning_rate": 8.308192096736759e-07, + "loss": 0.7105, + "step": 7848 + }, + { + "epoch": 0.7078504757180863, + "grad_norm": 3.0822080844291184, + "learning_rate": 8.303452727795703e-07, + "loss": 0.7256, + "step": 7849 + }, + { + "epoch": 0.7079406592415566, + "grad_norm": 1.9381276621426602, + "learning_rate": 8.298714356868542e-07, + "loss": 0.8233, + "step": 7850 + }, + { + "epoch": 0.7080308427650268, + "grad_norm": 4.295956045199387, + "learning_rate": 8.293976984359605e-07, + "loss": 0.7566, + "step": 7851 + }, + { + "epoch": 0.708121026288497, + "grad_norm": 2.501252909462765, + "learning_rate": 8.289240610673092e-07, + "loss": 0.7476, + "step": 7852 + }, + { + "epoch": 0.7082112098119674, + "grad_norm": 1.5486482417165155, + "learning_rate": 8.284505236213144e-07, + "loss": 0.7796, + "step": 7853 + }, + { + "epoch": 0.7083013933354376, + "grad_norm": 1.8333207332293693, + "learning_rate": 8.279770861383806e-07, + "loss": 0.7384, + "step": 7854 + }, + { + "epoch": 0.7083915768589079, + "grad_norm": 1.5316114193759358, + "learning_rate": 8.275037486589042e-07, + "loss": 0.7425, + "step": 7855 + }, + { + "epoch": 0.7084817603823781, + "grad_norm": 0.6534671817936035, + "learning_rate": 8.270305112232739e-07, + "loss": 0.5594, + "step": 7856 + }, + { + "epoch": 0.7085719439058484, + "grad_norm": 1.63010272901749, + "learning_rate": 8.265573738718665e-07, + "loss": 0.7439, + "step": 7857 + }, + { + "epoch": 0.7086621274293187, + "grad_norm": 1.8918432929082354, + "learning_rate": 8.260843366450559e-07, + "loss": 0.705, + "step": 7858 + }, + { + "epoch": 0.7087523109527889, + "grad_norm": 1.5773438134838165, + "learning_rate": 8.256113995832017e-07, + "loss": 0.747, + "step": 7859 + }, + { + "epoch": 0.7088424944762591, + "grad_norm": 1.9575544864676497, + "learning_rate": 8.251385627266583e-07, + "loss": 0.7876, + "step": 7860 + }, + { + "epoch": 0.7089326779997295, + "grad_norm": 1.8425297982900222, + "learning_rate": 8.24665826115771e-07, + "loss": 0.757, + "step": 7861 + }, + { + "epoch": 0.7090228615231997, + "grad_norm": 3.0373126878793095, + "learning_rate": 8.241931897908763e-07, + "loss": 0.7649, + "step": 7862 + }, + { + "epoch": 0.7091130450466699, + "grad_norm": 1.6775354127514437, + "learning_rate": 8.237206537923016e-07, + "loss": 0.7727, + "step": 7863 + }, + { + "epoch": 0.7092032285701403, + "grad_norm": 2.243814655530626, + "learning_rate": 8.232482181603671e-07, + "loss": 0.739, + "step": 7864 + }, + { + "epoch": 0.7092934120936105, + "grad_norm": 2.555723674284467, + "learning_rate": 8.227758829353828e-07, + "loss": 0.7256, + "step": 7865 + }, + { + "epoch": 0.7093835956170808, + "grad_norm": 2.3033495106476316, + "learning_rate": 8.223036481576522e-07, + "loss": 0.7198, + "step": 7866 + }, + { + "epoch": 0.709473779140551, + "grad_norm": 1.9225285977145181, + "learning_rate": 8.218315138674672e-07, + "loss": 0.759, + "step": 7867 + }, + { + "epoch": 0.7095639626640213, + "grad_norm": 0.6219417373272137, + "learning_rate": 8.21359480105114e-07, + "loss": 0.5766, + "step": 7868 + }, + { + "epoch": 0.7096541461874916, + "grad_norm": 1.6740162052705532, + "learning_rate": 8.208875469108689e-07, + "loss": 0.7779, + "step": 7869 + }, + { + "epoch": 0.7097443297109618, + "grad_norm": 2.1355390063961264, + "learning_rate": 8.204157143249997e-07, + "loss": 0.7367, + "step": 7870 + }, + { + "epoch": 0.709834513234432, + "grad_norm": 1.4986881241421859, + "learning_rate": 8.199439823877668e-07, + "loss": 0.7162, + "step": 7871 + }, + { + "epoch": 0.7099246967579024, + "grad_norm": 1.5665207818604292, + "learning_rate": 8.194723511394186e-07, + "loss": 0.734, + "step": 7872 + }, + { + "epoch": 0.7100148802813726, + "grad_norm": 1.6725611481942455, + "learning_rate": 8.190008206202002e-07, + "loss": 0.6939, + "step": 7873 + }, + { + "epoch": 0.7101050638048428, + "grad_norm": 2.6101214774497907, + "learning_rate": 8.185293908703423e-07, + "loss": 0.7953, + "step": 7874 + }, + { + "epoch": 0.7101952473283131, + "grad_norm": 1.6241346402612855, + "learning_rate": 8.180580619300727e-07, + "loss": 0.7089, + "step": 7875 + }, + { + "epoch": 0.7102854308517834, + "grad_norm": 1.424644455975944, + "learning_rate": 8.175868338396057e-07, + "loss": 0.7025, + "step": 7876 + }, + { + "epoch": 0.7103756143752537, + "grad_norm": 2.6351103673763636, + "learning_rate": 8.171157066391499e-07, + "loss": 0.7135, + "step": 7877 + }, + { + "epoch": 0.7104657978987239, + "grad_norm": 2.014596484462519, + "learning_rate": 8.166446803689045e-07, + "loss": 0.7751, + "step": 7878 + }, + { + "epoch": 0.7105559814221941, + "grad_norm": 0.6679782978290121, + "learning_rate": 8.161737550690595e-07, + "loss": 0.5842, + "step": 7879 + }, + { + "epoch": 0.7106461649456645, + "grad_norm": 1.765533714972913, + "learning_rate": 8.157029307797976e-07, + "loss": 0.7747, + "step": 7880 + }, + { + "epoch": 0.7107363484691347, + "grad_norm": 1.6360365144483409, + "learning_rate": 8.152322075412925e-07, + "loss": 0.8022, + "step": 7881 + }, + { + "epoch": 0.7108265319926049, + "grad_norm": 1.6720885381071808, + "learning_rate": 8.147615853937073e-07, + "loss": 0.6934, + "step": 7882 + }, + { + "epoch": 0.7109167155160752, + "grad_norm": 1.7516339031437762, + "learning_rate": 8.142910643771992e-07, + "loss": 0.6661, + "step": 7883 + }, + { + "epoch": 0.7110068990395455, + "grad_norm": 7.591536587864974, + "learning_rate": 8.138206445319152e-07, + "loss": 0.7113, + "step": 7884 + }, + { + "epoch": 0.7110970825630157, + "grad_norm": 2.8955881760559357, + "learning_rate": 8.133503258979944e-07, + "loss": 0.649, + "step": 7885 + }, + { + "epoch": 0.711187266086486, + "grad_norm": 3.4334773715523323, + "learning_rate": 8.12880108515567e-07, + "loss": 0.6705, + "step": 7886 + }, + { + "epoch": 0.7112774496099563, + "grad_norm": 4.327838323580689, + "learning_rate": 8.124099924247543e-07, + "loss": 0.718, + "step": 7887 + }, + { + "epoch": 0.7113676331334265, + "grad_norm": 1.9298524979078142, + "learning_rate": 8.119399776656701e-07, + "loss": 0.65, + "step": 7888 + }, + { + "epoch": 0.7114578166568968, + "grad_norm": 1.5403925381787058, + "learning_rate": 8.114700642784167e-07, + "loss": 0.7444, + "step": 7889 + }, + { + "epoch": 0.711548000180367, + "grad_norm": 1.8735431733604857, + "learning_rate": 8.110002523030921e-07, + "loss": 0.8004, + "step": 7890 + }, + { + "epoch": 0.7116381837038374, + "grad_norm": 2.3613934503926646, + "learning_rate": 8.105305417797808e-07, + "loss": 0.767, + "step": 7891 + }, + { + "epoch": 0.7117283672273076, + "grad_norm": 3.1220167034320605, + "learning_rate": 8.100609327485635e-07, + "loss": 0.7852, + "step": 7892 + }, + { + "epoch": 0.7118185507507778, + "grad_norm": 1.9746892443161892, + "learning_rate": 8.095914252495082e-07, + "loss": 0.7982, + "step": 7893 + }, + { + "epoch": 0.7119087342742481, + "grad_norm": 2.146123687457565, + "learning_rate": 8.091220193226762e-07, + "loss": 0.7165, + "step": 7894 + }, + { + "epoch": 0.7119989177977184, + "grad_norm": 1.6858167377031577, + "learning_rate": 8.0865271500812e-07, + "loss": 0.7286, + "step": 7895 + }, + { + "epoch": 0.7120891013211886, + "grad_norm": 1.899704937861901, + "learning_rate": 8.081835123458831e-07, + "loss": 0.7847, + "step": 7896 + }, + { + "epoch": 0.7121792848446589, + "grad_norm": 1.575304059545376, + "learning_rate": 8.077144113760013e-07, + "loss": 0.7286, + "step": 7897 + }, + { + "epoch": 0.7122694683681291, + "grad_norm": 1.4216310061193145, + "learning_rate": 8.072454121384995e-07, + "loss": 0.7211, + "step": 7898 + }, + { + "epoch": 0.7123596518915994, + "grad_norm": 1.4311765826856016, + "learning_rate": 8.067765146733958e-07, + "loss": 0.7862, + "step": 7899 + }, + { + "epoch": 0.7124498354150697, + "grad_norm": 2.347204180716475, + "learning_rate": 8.063077190206993e-07, + "loss": 0.8255, + "step": 7900 + }, + { + "epoch": 0.7125400189385399, + "grad_norm": 0.6074617624218976, + "learning_rate": 8.058390252204101e-07, + "loss": 0.4937, + "step": 7901 + }, + { + "epoch": 0.7126302024620101, + "grad_norm": 2.1119404204013312, + "learning_rate": 8.0537043331252e-07, + "loss": 0.7378, + "step": 7902 + }, + { + "epoch": 0.7127203859854805, + "grad_norm": 2.804394039857119, + "learning_rate": 8.049019433370121e-07, + "loss": 0.6827, + "step": 7903 + }, + { + "epoch": 0.7128105695089507, + "grad_norm": 1.5077910742942184, + "learning_rate": 8.044335553338588e-07, + "loss": 0.7489, + "step": 7904 + }, + { + "epoch": 0.712900753032421, + "grad_norm": 1.5204589679102671, + "learning_rate": 8.039652693430281e-07, + "loss": 0.7533, + "step": 7905 + }, + { + "epoch": 0.7129909365558912, + "grad_norm": 2.9842615555377097, + "learning_rate": 8.034970854044742e-07, + "loss": 0.7189, + "step": 7906 + }, + { + "epoch": 0.7130811200793615, + "grad_norm": 1.752634313663238, + "learning_rate": 8.03029003558148e-07, + "loss": 0.8365, + "step": 7907 + }, + { + "epoch": 0.7131713036028318, + "grad_norm": 2.0623727239005123, + "learning_rate": 8.025610238439864e-07, + "loss": 0.8118, + "step": 7908 + }, + { + "epoch": 0.713261487126302, + "grad_norm": 2.0277657777033453, + "learning_rate": 8.020931463019207e-07, + "loss": 0.8253, + "step": 7909 + }, + { + "epoch": 0.7133516706497723, + "grad_norm": 1.7388021652985703, + "learning_rate": 8.016253709718732e-07, + "loss": 0.6882, + "step": 7910 + }, + { + "epoch": 0.7134418541732426, + "grad_norm": 2.1061582550986095, + "learning_rate": 8.011576978937567e-07, + "loss": 0.6518, + "step": 7911 + }, + { + "epoch": 0.7135320376967128, + "grad_norm": 1.617710121220937, + "learning_rate": 8.006901271074764e-07, + "loss": 0.7446, + "step": 7912 + }, + { + "epoch": 0.713622221220183, + "grad_norm": 1.7988575488737668, + "learning_rate": 8.002226586529261e-07, + "loss": 0.7876, + "step": 7913 + }, + { + "epoch": 0.7137124047436534, + "grad_norm": 4.998746333310499, + "learning_rate": 7.997552925699956e-07, + "loss": 0.8216, + "step": 7914 + }, + { + "epoch": 0.7138025882671236, + "grad_norm": 1.7752391385454067, + "learning_rate": 7.992880288985606e-07, + "loss": 0.8383, + "step": 7915 + }, + { + "epoch": 0.7138927717905938, + "grad_norm": 1.7705881360407911, + "learning_rate": 7.988208676784918e-07, + "loss": 0.7526, + "step": 7916 + }, + { + "epoch": 0.7139829553140641, + "grad_norm": 3.2367969235558816, + "learning_rate": 7.983538089496497e-07, + "loss": 0.7393, + "step": 7917 + }, + { + "epoch": 0.7140731388375344, + "grad_norm": 1.7696562127121742, + "learning_rate": 7.978868527518864e-07, + "loss": 0.7069, + "step": 7918 + }, + { + "epoch": 0.7141633223610047, + "grad_norm": 2.061313866064834, + "learning_rate": 7.974199991250455e-07, + "loss": 0.7277, + "step": 7919 + }, + { + "epoch": 0.7142535058844749, + "grad_norm": 1.8677021418687099, + "learning_rate": 7.969532481089616e-07, + "loss": 0.8454, + "step": 7920 + }, + { + "epoch": 0.7143436894079451, + "grad_norm": 1.9852278070865477, + "learning_rate": 7.964865997434589e-07, + "loss": 0.7564, + "step": 7921 + }, + { + "epoch": 0.7144338729314155, + "grad_norm": 1.6909022472332593, + "learning_rate": 7.96020054068357e-07, + "loss": 0.8185, + "step": 7922 + }, + { + "epoch": 0.7145240564548857, + "grad_norm": 1.228780987679653, + "learning_rate": 7.95553611123462e-07, + "loss": 0.7013, + "step": 7923 + }, + { + "epoch": 0.7146142399783559, + "grad_norm": 1.815314311300626, + "learning_rate": 7.950872709485741e-07, + "loss": 0.9321, + "step": 7924 + }, + { + "epoch": 0.7147044235018262, + "grad_norm": 2.2871003057102954, + "learning_rate": 7.946210335834842e-07, + "loss": 0.7975, + "step": 7925 + }, + { + "epoch": 0.7147946070252965, + "grad_norm": 1.5331430216871875, + "learning_rate": 7.94154899067974e-07, + "loss": 0.6691, + "step": 7926 + }, + { + "epoch": 0.7148847905487667, + "grad_norm": 1.7078299672454176, + "learning_rate": 7.936888674418177e-07, + "loss": 0.6723, + "step": 7927 + }, + { + "epoch": 0.714974974072237, + "grad_norm": 2.574484827301781, + "learning_rate": 7.932229387447771e-07, + "loss": 0.7504, + "step": 7928 + }, + { + "epoch": 0.7150651575957072, + "grad_norm": 2.5869490843231473, + "learning_rate": 7.927571130166109e-07, + "loss": 0.7392, + "step": 7929 + }, + { + "epoch": 0.7151553411191776, + "grad_norm": 0.8534483370237816, + "learning_rate": 7.922913902970632e-07, + "loss": 0.568, + "step": 7930 + }, + { + "epoch": 0.7152455246426478, + "grad_norm": 1.809932205237238, + "learning_rate": 7.918257706258744e-07, + "loss": 0.7414, + "step": 7931 + }, + { + "epoch": 0.715335708166118, + "grad_norm": 1.9997007622974268, + "learning_rate": 7.913602540427724e-07, + "loss": 0.7442, + "step": 7932 + }, + { + "epoch": 0.7154258916895884, + "grad_norm": 1.7808148705907811, + "learning_rate": 7.908948405874775e-07, + "loss": 0.7387, + "step": 7933 + }, + { + "epoch": 0.7155160752130586, + "grad_norm": 1.6622104995034244, + "learning_rate": 7.904295302997019e-07, + "loss": 0.6759, + "step": 7934 + }, + { + "epoch": 0.7156062587365288, + "grad_norm": 1.5176792970217385, + "learning_rate": 7.899643232191484e-07, + "loss": 0.7576, + "step": 7935 + }, + { + "epoch": 0.7156964422599991, + "grad_norm": 1.4821719234065487, + "learning_rate": 7.894992193855108e-07, + "loss": 0.7314, + "step": 7936 + }, + { + "epoch": 0.7157866257834694, + "grad_norm": 1.9387296650728276, + "learning_rate": 7.890342188384751e-07, + "loss": 0.6966, + "step": 7937 + }, + { + "epoch": 0.7158768093069396, + "grad_norm": 1.864915111440472, + "learning_rate": 7.885693216177165e-07, + "loss": 0.6865, + "step": 7938 + }, + { + "epoch": 0.7159669928304099, + "grad_norm": 1.5512569284092246, + "learning_rate": 7.88104527762903e-07, + "loss": 0.7407, + "step": 7939 + }, + { + "epoch": 0.7160571763538801, + "grad_norm": 4.016642994369038, + "learning_rate": 7.876398373136936e-07, + "loss": 0.7034, + "step": 7940 + }, + { + "epoch": 0.7161473598773505, + "grad_norm": 1.8957071017764606, + "learning_rate": 7.87175250309738e-07, + "loss": 0.7399, + "step": 7941 + }, + { + "epoch": 0.7162375434008207, + "grad_norm": 1.6745246497384825, + "learning_rate": 7.867107667906785e-07, + "loss": 0.8212, + "step": 7942 + }, + { + "epoch": 0.7163277269242909, + "grad_norm": 1.8070546268677803, + "learning_rate": 7.862463867961446e-07, + "loss": 0.7417, + "step": 7943 + }, + { + "epoch": 0.7164179104477612, + "grad_norm": 1.4052524419280468, + "learning_rate": 7.857821103657632e-07, + "loss": 0.8089, + "step": 7944 + }, + { + "epoch": 0.7165080939712315, + "grad_norm": 1.729492049645322, + "learning_rate": 7.853179375391459e-07, + "loss": 0.7731, + "step": 7945 + }, + { + "epoch": 0.7165982774947017, + "grad_norm": 1.6267341017555177, + "learning_rate": 7.848538683559012e-07, + "loss": 0.7307, + "step": 7946 + }, + { + "epoch": 0.716688461018172, + "grad_norm": 1.827962754051678, + "learning_rate": 7.843899028556238e-07, + "loss": 0.7844, + "step": 7947 + }, + { + "epoch": 0.7167786445416422, + "grad_norm": 1.934811480306273, + "learning_rate": 7.839260410779029e-07, + "loss": 0.7385, + "step": 7948 + }, + { + "epoch": 0.7168688280651125, + "grad_norm": 1.4721956620529602, + "learning_rate": 7.834622830623175e-07, + "loss": 0.7623, + "step": 7949 + }, + { + "epoch": 0.7169590115885828, + "grad_norm": 1.7057025548203386, + "learning_rate": 7.82998628848438e-07, + "loss": 0.7177, + "step": 7950 + }, + { + "epoch": 0.717049195112053, + "grad_norm": 2.1217775033116855, + "learning_rate": 7.825350784758261e-07, + "loss": 0.7346, + "step": 7951 + }, + { + "epoch": 0.7171393786355232, + "grad_norm": 1.4030899995036503, + "learning_rate": 7.820716319840342e-07, + "loss": 0.7068, + "step": 7952 + }, + { + "epoch": 0.7172295621589936, + "grad_norm": 2.021605026404016, + "learning_rate": 7.816082894126069e-07, + "loss": 0.8234, + "step": 7953 + }, + { + "epoch": 0.7173197456824638, + "grad_norm": 1.9673525376661773, + "learning_rate": 7.811450508010778e-07, + "loss": 0.718, + "step": 7954 + }, + { + "epoch": 0.717409929205934, + "grad_norm": 1.5710845112351122, + "learning_rate": 7.806819161889737e-07, + "loss": 0.7176, + "step": 7955 + }, + { + "epoch": 0.7175001127294043, + "grad_norm": 1.6191208197437312, + "learning_rate": 7.802188856158119e-07, + "loss": 0.7756, + "step": 7956 + }, + { + "epoch": 0.7175902962528746, + "grad_norm": 1.9619722106124373, + "learning_rate": 7.797559591211002e-07, + "loss": 0.7286, + "step": 7957 + }, + { + "epoch": 0.7176804797763449, + "grad_norm": 1.6150808572752973, + "learning_rate": 7.79293136744339e-07, + "loss": 0.7089, + "step": 7958 + }, + { + "epoch": 0.7177706632998151, + "grad_norm": 1.8182422833792298, + "learning_rate": 7.788304185250185e-07, + "loss": 0.7421, + "step": 7959 + }, + { + "epoch": 0.7178608468232854, + "grad_norm": 1.6695094262902639, + "learning_rate": 7.78367804502619e-07, + "loss": 0.7711, + "step": 7960 + }, + { + "epoch": 0.7179510303467557, + "grad_norm": 2.422867799125614, + "learning_rate": 7.779052947166156e-07, + "loss": 0.755, + "step": 7961 + }, + { + "epoch": 0.7180412138702259, + "grad_norm": 3.4816754065793933, + "learning_rate": 7.774428892064697e-07, + "loss": 0.7001, + "step": 7962 + }, + { + "epoch": 0.7181313973936961, + "grad_norm": 1.488469870321649, + "learning_rate": 7.769805880116391e-07, + "loss": 0.7371, + "step": 7963 + }, + { + "epoch": 0.7182215809171665, + "grad_norm": 2.3358773826594725, + "learning_rate": 7.765183911715678e-07, + "loss": 0.725, + "step": 7964 + }, + { + "epoch": 0.7183117644406367, + "grad_norm": 1.7273071892508285, + "learning_rate": 7.760562987256933e-07, + "loss": 0.6453, + "step": 7965 + }, + { + "epoch": 0.718401947964107, + "grad_norm": 2.120064220355458, + "learning_rate": 7.755943107134444e-07, + "loss": 0.7499, + "step": 7966 + }, + { + "epoch": 0.7184921314875772, + "grad_norm": 0.6369113279843696, + "learning_rate": 7.751324271742401e-07, + "loss": 0.5687, + "step": 7967 + }, + { + "epoch": 0.7185823150110475, + "grad_norm": 1.753916513119854, + "learning_rate": 7.746706481474916e-07, + "loss": 0.7591, + "step": 7968 + }, + { + "epoch": 0.7186724985345178, + "grad_norm": 0.5978248707378214, + "learning_rate": 7.742089736725992e-07, + "loss": 0.5371, + "step": 7969 + }, + { + "epoch": 0.718762682057988, + "grad_norm": 1.8729142986391547, + "learning_rate": 7.737474037889559e-07, + "loss": 0.6639, + "step": 7970 + }, + { + "epoch": 0.7188528655814582, + "grad_norm": 1.5219508032748994, + "learning_rate": 7.732859385359458e-07, + "loss": 0.712, + "step": 7971 + }, + { + "epoch": 0.7189430491049286, + "grad_norm": 1.673536170593656, + "learning_rate": 7.728245779529434e-07, + "loss": 0.6644, + "step": 7972 + }, + { + "epoch": 0.7190332326283988, + "grad_norm": 3.1694880180992957, + "learning_rate": 7.723633220793146e-07, + "loss": 0.7714, + "step": 7973 + }, + { + "epoch": 0.719123416151869, + "grad_norm": 1.8352034379617603, + "learning_rate": 7.719021709544162e-07, + "loss": 0.6345, + "step": 7974 + }, + { + "epoch": 0.7192135996753393, + "grad_norm": 1.8139421710603598, + "learning_rate": 7.714411246175964e-07, + "loss": 0.8337, + "step": 7975 + }, + { + "epoch": 0.7193037831988096, + "grad_norm": 1.649883413530792, + "learning_rate": 7.709801831081946e-07, + "loss": 0.664, + "step": 7976 + }, + { + "epoch": 0.7193939667222798, + "grad_norm": 3.3263580933173786, + "learning_rate": 7.705193464655391e-07, + "loss": 0.7041, + "step": 7977 + }, + { + "epoch": 0.7194841502457501, + "grad_norm": 1.3101741073019253, + "learning_rate": 7.700586147289534e-07, + "loss": 0.7177, + "step": 7978 + }, + { + "epoch": 0.7195743337692203, + "grad_norm": 2.345049891805322, + "learning_rate": 7.695979879377481e-07, + "loss": 0.6999, + "step": 7979 + }, + { + "epoch": 0.7196645172926907, + "grad_norm": 2.3482415458261414, + "learning_rate": 7.691374661312266e-07, + "loss": 0.7653, + "step": 7980 + }, + { + "epoch": 0.7197547008161609, + "grad_norm": 1.6709787693923612, + "learning_rate": 7.686770493486834e-07, + "loss": 0.7194, + "step": 7981 + }, + { + "epoch": 0.7198448843396311, + "grad_norm": 1.785784254065729, + "learning_rate": 7.68216737629404e-07, + "loss": 0.6901, + "step": 7982 + }, + { + "epoch": 0.7199350678631015, + "grad_norm": 1.3974271992810017, + "learning_rate": 7.67756531012665e-07, + "loss": 0.8091, + "step": 7983 + }, + { + "epoch": 0.7200252513865717, + "grad_norm": 1.4858988920344625, + "learning_rate": 7.67296429537732e-07, + "loss": 0.8007, + "step": 7984 + }, + { + "epoch": 0.7201154349100419, + "grad_norm": 1.7105371015417152, + "learning_rate": 7.668364332438661e-07, + "loss": 0.8097, + "step": 7985 + }, + { + "epoch": 0.7202056184335122, + "grad_norm": 1.665819914615712, + "learning_rate": 7.663765421703145e-07, + "loss": 0.7554, + "step": 7986 + }, + { + "epoch": 0.7202958019569825, + "grad_norm": 1.7960338448137074, + "learning_rate": 7.659167563563187e-07, + "loss": 0.679, + "step": 7987 + }, + { + "epoch": 0.7203859854804527, + "grad_norm": 1.939288882458956, + "learning_rate": 7.654570758411096e-07, + "loss": 0.7711, + "step": 7988 + }, + { + "epoch": 0.720476169003923, + "grad_norm": 2.2752917092233576, + "learning_rate": 7.649975006639103e-07, + "loss": 0.8012, + "step": 7989 + }, + { + "epoch": 0.7205663525273932, + "grad_norm": 0.6201013277684013, + "learning_rate": 7.645380308639337e-07, + "loss": 0.5516, + "step": 7990 + }, + { + "epoch": 0.7206565360508635, + "grad_norm": 1.9322170095767655, + "learning_rate": 7.640786664803853e-07, + "loss": 0.7876, + "step": 7991 + }, + { + "epoch": 0.7207467195743338, + "grad_norm": 1.6239036382828331, + "learning_rate": 7.636194075524587e-07, + "loss": 0.6627, + "step": 7992 + }, + { + "epoch": 0.720836903097804, + "grad_norm": 1.6663316469442786, + "learning_rate": 7.631602541193429e-07, + "loss": 0.6971, + "step": 7993 + }, + { + "epoch": 0.7209270866212742, + "grad_norm": 1.8880934988485838, + "learning_rate": 7.627012062202132e-07, + "loss": 0.7943, + "step": 7994 + }, + { + "epoch": 0.7210172701447446, + "grad_norm": 1.8947424702044884, + "learning_rate": 7.622422638942391e-07, + "loss": 0.742, + "step": 7995 + }, + { + "epoch": 0.7211074536682148, + "grad_norm": 22.239373916485725, + "learning_rate": 7.617834271805801e-07, + "loss": 0.8065, + "step": 7996 + }, + { + "epoch": 0.7211976371916851, + "grad_norm": 0.6480641956555491, + "learning_rate": 7.613246961183863e-07, + "loss": 0.5653, + "step": 7997 + }, + { + "epoch": 0.7212878207151553, + "grad_norm": 1.5150731448654888, + "learning_rate": 7.608660707468002e-07, + "loss": 0.7002, + "step": 7998 + }, + { + "epoch": 0.7213780042386256, + "grad_norm": 1.7404935667318513, + "learning_rate": 7.604075511049522e-07, + "loss": 0.6366, + "step": 7999 + }, + { + "epoch": 0.7214681877620959, + "grad_norm": 1.8831615203648457, + "learning_rate": 7.599491372319682e-07, + "loss": 0.6854, + "step": 8000 + }, + { + "epoch": 0.7215583712855661, + "grad_norm": 2.8247192277043305, + "learning_rate": 7.594908291669601e-07, + "loss": 0.7422, + "step": 8001 + }, + { + "epoch": 0.7216485548090363, + "grad_norm": 1.8009358834096076, + "learning_rate": 7.590326269490359e-07, + "loss": 0.7739, + "step": 8002 + }, + { + "epoch": 0.7217387383325067, + "grad_norm": 1.572323702100885, + "learning_rate": 7.585745306172899e-07, + "loss": 0.7002, + "step": 8003 + }, + { + "epoch": 0.7218289218559769, + "grad_norm": 2.2571559070085745, + "learning_rate": 7.5811654021081e-07, + "loss": 0.7231, + "step": 8004 + }, + { + "epoch": 0.7219191053794471, + "grad_norm": 1.3802016875868772, + "learning_rate": 7.576586557686748e-07, + "loss": 0.7714, + "step": 8005 + }, + { + "epoch": 0.7220092889029175, + "grad_norm": 3.7195554229713617, + "learning_rate": 7.572008773299531e-07, + "loss": 0.7098, + "step": 8006 + }, + { + "epoch": 0.7220994724263877, + "grad_norm": 1.9530916134841785, + "learning_rate": 7.567432049337055e-07, + "loss": 0.7789, + "step": 8007 + }, + { + "epoch": 0.722189655949858, + "grad_norm": 1.6688328811694169, + "learning_rate": 7.562856386189834e-07, + "loss": 0.6779, + "step": 8008 + }, + { + "epoch": 0.7222798394733282, + "grad_norm": 2.82759240968295, + "learning_rate": 7.558281784248275e-07, + "loss": 0.7591, + "step": 8009 + }, + { + "epoch": 0.7223700229967985, + "grad_norm": 1.7695278816106876, + "learning_rate": 7.553708243902721e-07, + "loss": 0.7345, + "step": 8010 + }, + { + "epoch": 0.7224602065202688, + "grad_norm": 2.485091483966999, + "learning_rate": 7.549135765543404e-07, + "loss": 0.6319, + "step": 8011 + }, + { + "epoch": 0.722550390043739, + "grad_norm": 2.4116502085424902, + "learning_rate": 7.544564349560481e-07, + "loss": 0.7644, + "step": 8012 + }, + { + "epoch": 0.7226405735672092, + "grad_norm": 1.6259620459656043, + "learning_rate": 7.539993996344009e-07, + "loss": 0.723, + "step": 8013 + }, + { + "epoch": 0.7227307570906796, + "grad_norm": 1.5695863629581845, + "learning_rate": 7.535424706283941e-07, + "loss": 0.7185, + "step": 8014 + }, + { + "epoch": 0.7228209406141498, + "grad_norm": 1.6075839845066981, + "learning_rate": 7.530856479770181e-07, + "loss": 0.7039, + "step": 8015 + }, + { + "epoch": 0.72291112413762, + "grad_norm": 1.6499433825633754, + "learning_rate": 7.526289317192484e-07, + "loss": 0.8223, + "step": 8016 + }, + { + "epoch": 0.7230013076610903, + "grad_norm": 2.475513416559811, + "learning_rate": 7.521723218940579e-07, + "loss": 0.74, + "step": 8017 + }, + { + "epoch": 0.7230914911845606, + "grad_norm": 1.8595909826674233, + "learning_rate": 7.517158185404038e-07, + "loss": 0.8045, + "step": 8018 + }, + { + "epoch": 0.7231816747080309, + "grad_norm": 2.280570987345124, + "learning_rate": 7.512594216972403e-07, + "loss": 0.768, + "step": 8019 + }, + { + "epoch": 0.7232718582315011, + "grad_norm": 11.085812590139318, + "learning_rate": 7.508031314035078e-07, + "loss": 0.778, + "step": 8020 + }, + { + "epoch": 0.7233620417549713, + "grad_norm": 2.0585815892593256, + "learning_rate": 7.503469476981401e-07, + "loss": 0.7647, + "step": 8021 + }, + { + "epoch": 0.7234522252784417, + "grad_norm": 1.7001620776386228, + "learning_rate": 7.498908706200613e-07, + "loss": 0.6162, + "step": 8022 + }, + { + "epoch": 0.7235424088019119, + "grad_norm": 1.3934462143530773, + "learning_rate": 7.494349002081866e-07, + "loss": 0.7866, + "step": 8023 + }, + { + "epoch": 0.7236325923253821, + "grad_norm": 1.6089789634737688, + "learning_rate": 7.489790365014224e-07, + "loss": 0.7746, + "step": 8024 + }, + { + "epoch": 0.7237227758488524, + "grad_norm": 1.6301057273416384, + "learning_rate": 7.485232795386642e-07, + "loss": 0.7579, + "step": 8025 + }, + { + "epoch": 0.7238129593723227, + "grad_norm": 1.482380621415241, + "learning_rate": 7.480676293588002e-07, + "loss": 0.7602, + "step": 8026 + }, + { + "epoch": 0.7239031428957929, + "grad_norm": 0.8170034241365282, + "learning_rate": 7.476120860007093e-07, + "loss": 0.5967, + "step": 8027 + }, + { + "epoch": 0.7239933264192632, + "grad_norm": 2.0227715416292678, + "learning_rate": 7.471566495032608e-07, + "loss": 0.7842, + "step": 8028 + }, + { + "epoch": 0.7240835099427335, + "grad_norm": 1.6466500955718262, + "learning_rate": 7.467013199053152e-07, + "loss": 0.6757, + "step": 8029 + }, + { + "epoch": 0.7241736934662037, + "grad_norm": 2.062958753024623, + "learning_rate": 7.46246097245724e-07, + "loss": 0.7434, + "step": 8030 + }, + { + "epoch": 0.724263876989674, + "grad_norm": 2.7950985344184476, + "learning_rate": 7.457909815633276e-07, + "loss": 0.7484, + "step": 8031 + }, + { + "epoch": 0.7243540605131442, + "grad_norm": 1.8972679271687958, + "learning_rate": 7.453359728969618e-07, + "loss": 0.7331, + "step": 8032 + }, + { + "epoch": 0.7244442440366146, + "grad_norm": 1.8324588366122765, + "learning_rate": 7.448810712854475e-07, + "loss": 0.6657, + "step": 8033 + }, + { + "epoch": 0.7245344275600848, + "grad_norm": 2.3121793369974544, + "learning_rate": 7.444262767676022e-07, + "loss": 0.7833, + "step": 8034 + }, + { + "epoch": 0.724624611083555, + "grad_norm": 1.6234928624660316, + "learning_rate": 7.439715893822296e-07, + "loss": 0.8159, + "step": 8035 + }, + { + "epoch": 0.7247147946070253, + "grad_norm": 1.9929916853054188, + "learning_rate": 7.435170091681264e-07, + "loss": 0.7669, + "step": 8036 + }, + { + "epoch": 0.7248049781304956, + "grad_norm": 3.730002425042149, + "learning_rate": 7.430625361640803e-07, + "loss": 0.7054, + "step": 8037 + }, + { + "epoch": 0.7248951616539658, + "grad_norm": 3.131679111095293, + "learning_rate": 7.426081704088694e-07, + "loss": 0.7833, + "step": 8038 + }, + { + "epoch": 0.7249853451774361, + "grad_norm": 3.15809350893954, + "learning_rate": 7.42153911941263e-07, + "loss": 0.8118, + "step": 8039 + }, + { + "epoch": 0.7250755287009063, + "grad_norm": 1.8232740288085627, + "learning_rate": 7.416997608000192e-07, + "loss": 0.7826, + "step": 8040 + }, + { + "epoch": 0.7251657122243766, + "grad_norm": 1.9204231298625785, + "learning_rate": 7.412457170238918e-07, + "loss": 0.7277, + "step": 8041 + }, + { + "epoch": 0.7252558957478469, + "grad_norm": 1.8414741953508185, + "learning_rate": 7.407917806516193e-07, + "loss": 0.6913, + "step": 8042 + }, + { + "epoch": 0.7253460792713171, + "grad_norm": 1.980844977342883, + "learning_rate": 7.403379517219354e-07, + "loss": 0.7348, + "step": 8043 + }, + { + "epoch": 0.7254362627947873, + "grad_norm": 1.5793862165891324, + "learning_rate": 7.398842302735636e-07, + "loss": 0.7742, + "step": 8044 + }, + { + "epoch": 0.7255264463182577, + "grad_norm": 0.6672174344282324, + "learning_rate": 7.394306163452171e-07, + "loss": 0.5462, + "step": 8045 + }, + { + "epoch": 0.7256166298417279, + "grad_norm": 1.367194170254376, + "learning_rate": 7.38977109975601e-07, + "loss": 0.7219, + "step": 8046 + }, + { + "epoch": 0.7257068133651982, + "grad_norm": 1.9781570335554524, + "learning_rate": 7.385237112034119e-07, + "loss": 0.658, + "step": 8047 + }, + { + "epoch": 0.7257969968886684, + "grad_norm": 1.604979057531894, + "learning_rate": 7.380704200673342e-07, + "loss": 0.8066, + "step": 8048 + }, + { + "epoch": 0.7258871804121387, + "grad_norm": 1.8266403821228212, + "learning_rate": 7.376172366060478e-07, + "loss": 0.767, + "step": 8049 + }, + { + "epoch": 0.725977363935609, + "grad_norm": 0.6420271346071036, + "learning_rate": 7.371641608582187e-07, + "loss": 0.5126, + "step": 8050 + }, + { + "epoch": 0.7260675474590792, + "grad_norm": 2.276859709661904, + "learning_rate": 7.367111928625067e-07, + "loss": 0.6158, + "step": 8051 + }, + { + "epoch": 0.7261577309825495, + "grad_norm": 1.8446262023346958, + "learning_rate": 7.362583326575613e-07, + "loss": 0.6576, + "step": 8052 + }, + { + "epoch": 0.7262479145060198, + "grad_norm": 1.8980981342925305, + "learning_rate": 7.358055802820234e-07, + "loss": 0.726, + "step": 8053 + }, + { + "epoch": 0.72633809802949, + "grad_norm": 6.280188019968997, + "learning_rate": 7.353529357745245e-07, + "loss": 0.7141, + "step": 8054 + }, + { + "epoch": 0.7264282815529602, + "grad_norm": 2.317106967588952, + "learning_rate": 7.349003991736851e-07, + "loss": 0.7845, + "step": 8055 + }, + { + "epoch": 0.7265184650764306, + "grad_norm": 1.8749190630928054, + "learning_rate": 7.344479705181206e-07, + "loss": 0.5949, + "step": 8056 + }, + { + "epoch": 0.7266086485999008, + "grad_norm": 1.5650459147533835, + "learning_rate": 7.339956498464322e-07, + "loss": 0.6931, + "step": 8057 + }, + { + "epoch": 0.726698832123371, + "grad_norm": 1.8390661325475555, + "learning_rate": 7.335434371972169e-07, + "loss": 0.79, + "step": 8058 + }, + { + "epoch": 0.7267890156468413, + "grad_norm": 1.5003120574600912, + "learning_rate": 7.33091332609058e-07, + "loss": 0.731, + "step": 8059 + }, + { + "epoch": 0.7268791991703116, + "grad_norm": 1.7089409871724475, + "learning_rate": 7.326393361205323e-07, + "loss": 0.7605, + "step": 8060 + }, + { + "epoch": 0.7269693826937819, + "grad_norm": 1.748450887865725, + "learning_rate": 7.321874477702068e-07, + "loss": 0.8216, + "step": 8061 + }, + { + "epoch": 0.7270595662172521, + "grad_norm": 0.749837619369319, + "learning_rate": 7.317356675966386e-07, + "loss": 0.5919, + "step": 8062 + }, + { + "epoch": 0.7271497497407223, + "grad_norm": 1.4330047312070764, + "learning_rate": 7.312839956383765e-07, + "loss": 0.7775, + "step": 8063 + }, + { + "epoch": 0.7272399332641927, + "grad_norm": 1.779095199662365, + "learning_rate": 7.308324319339603e-07, + "loss": 0.7199, + "step": 8064 + }, + { + "epoch": 0.7273301167876629, + "grad_norm": 2.0360633746777985, + "learning_rate": 7.303809765219182e-07, + "loss": 0.7434, + "step": 8065 + }, + { + "epoch": 0.7274203003111331, + "grad_norm": 2.2142868283704074, + "learning_rate": 7.299296294407719e-07, + "loss": 0.7876, + "step": 8066 + }, + { + "epoch": 0.7275104838346034, + "grad_norm": 1.6323540148809852, + "learning_rate": 7.294783907290327e-07, + "loss": 0.7854, + "step": 8067 + }, + { + "epoch": 0.7276006673580737, + "grad_norm": 2.0850393812087593, + "learning_rate": 7.290272604252028e-07, + "loss": 0.6114, + "step": 8068 + }, + { + "epoch": 0.727690850881544, + "grad_norm": 2.481935274827903, + "learning_rate": 7.285762385677758e-07, + "loss": 0.6491, + "step": 8069 + }, + { + "epoch": 0.7277810344050142, + "grad_norm": 1.7872450480006326, + "learning_rate": 7.281253251952335e-07, + "loss": 0.7798, + "step": 8070 + }, + { + "epoch": 0.7278712179284844, + "grad_norm": 1.5814126394712094, + "learning_rate": 7.276745203460526e-07, + "loss": 0.7197, + "step": 8071 + }, + { + "epoch": 0.7279614014519548, + "grad_norm": 0.7859698961986777, + "learning_rate": 7.272238240586959e-07, + "loss": 0.632, + "step": 8072 + }, + { + "epoch": 0.728051584975425, + "grad_norm": 1.2998249559663617, + "learning_rate": 7.267732363716219e-07, + "loss": 0.6771, + "step": 8073 + }, + { + "epoch": 0.7281417684988952, + "grad_norm": 2.195584317684755, + "learning_rate": 7.263227573232753e-07, + "loss": 0.6964, + "step": 8074 + }, + { + "epoch": 0.7282319520223655, + "grad_norm": 1.7591215557540931, + "learning_rate": 7.258723869520937e-07, + "loss": 0.8309, + "step": 8075 + }, + { + "epoch": 0.7283221355458358, + "grad_norm": 2.8245141176967263, + "learning_rate": 7.254221252965059e-07, + "loss": 0.7901, + "step": 8076 + }, + { + "epoch": 0.728412319069306, + "grad_norm": 2.1895172626993484, + "learning_rate": 7.249719723949301e-07, + "loss": 0.6581, + "step": 8077 + }, + { + "epoch": 0.7285025025927763, + "grad_norm": 1.4320460808335098, + "learning_rate": 7.245219282857761e-07, + "loss": 0.7648, + "step": 8078 + }, + { + "epoch": 0.7285926861162466, + "grad_norm": 1.5466715360981373, + "learning_rate": 7.240719930074442e-07, + "loss": 0.783, + "step": 8079 + }, + { + "epoch": 0.7286828696397168, + "grad_norm": 2.4094464757784335, + "learning_rate": 7.236221665983257e-07, + "loss": 0.7687, + "step": 8080 + }, + { + "epoch": 0.7287730531631871, + "grad_norm": 0.7308814367066658, + "learning_rate": 7.231724490968012e-07, + "loss": 0.5573, + "step": 8081 + }, + { + "epoch": 0.7288632366866573, + "grad_norm": 2.068616409828846, + "learning_rate": 7.227228405412438e-07, + "loss": 0.7123, + "step": 8082 + }, + { + "epoch": 0.7289534202101277, + "grad_norm": 2.2958143211542166, + "learning_rate": 7.222733409700165e-07, + "loss": 0.701, + "step": 8083 + }, + { + "epoch": 0.7290436037335979, + "grad_norm": 1.6140124716672761, + "learning_rate": 7.21823950421473e-07, + "loss": 0.7521, + "step": 8084 + }, + { + "epoch": 0.7291337872570681, + "grad_norm": 1.6317619338910545, + "learning_rate": 7.213746689339577e-07, + "loss": 0.7162, + "step": 8085 + }, + { + "epoch": 0.7292239707805384, + "grad_norm": 3.2583166962676096, + "learning_rate": 7.20925496545807e-07, + "loss": 0.7434, + "step": 8086 + }, + { + "epoch": 0.7293141543040087, + "grad_norm": 2.417941535463771, + "learning_rate": 7.20476433295344e-07, + "loss": 0.6529, + "step": 8087 + }, + { + "epoch": 0.7294043378274789, + "grad_norm": 1.4641697668231992, + "learning_rate": 7.200274792208882e-07, + "loss": 0.7826, + "step": 8088 + }, + { + "epoch": 0.7294945213509492, + "grad_norm": 1.5834821664560592, + "learning_rate": 7.195786343607444e-07, + "loss": 0.6315, + "step": 8089 + }, + { + "epoch": 0.7295847048744194, + "grad_norm": 2.1398680525058102, + "learning_rate": 7.191298987532131e-07, + "loss": 0.7379, + "step": 8090 + }, + { + "epoch": 0.7296748883978897, + "grad_norm": 2.178043087003279, + "learning_rate": 7.186812724365805e-07, + "loss": 0.7507, + "step": 8091 + }, + { + "epoch": 0.72976507192136, + "grad_norm": 1.8934698218637112, + "learning_rate": 7.182327554491272e-07, + "loss": 0.7547, + "step": 8092 + }, + { + "epoch": 0.7298552554448302, + "grad_norm": 0.6850069751175543, + "learning_rate": 7.177843478291225e-07, + "loss": 0.5871, + "step": 8093 + }, + { + "epoch": 0.7299454389683004, + "grad_norm": 1.9546044811545977, + "learning_rate": 7.173360496148276e-07, + "loss": 0.7751, + "step": 8094 + }, + { + "epoch": 0.7300356224917708, + "grad_norm": 2.551294062008263, + "learning_rate": 7.168878608444939e-07, + "loss": 0.7578, + "step": 8095 + }, + { + "epoch": 0.730125806015241, + "grad_norm": 2.3960628123193755, + "learning_rate": 7.164397815563623e-07, + "loss": 0.7435, + "step": 8096 + }, + { + "epoch": 0.7302159895387113, + "grad_norm": 1.88451754661887, + "learning_rate": 7.159918117886661e-07, + "loss": 0.7491, + "step": 8097 + }, + { + "epoch": 0.7303061730621815, + "grad_norm": 4.750083521058472, + "learning_rate": 7.155439515796284e-07, + "loss": 0.7612, + "step": 8098 + }, + { + "epoch": 0.7303963565856518, + "grad_norm": 1.874559668652807, + "learning_rate": 7.150962009674633e-07, + "loss": 0.7424, + "step": 8099 + }, + { + "epoch": 0.7304865401091221, + "grad_norm": 1.668344352652749, + "learning_rate": 7.146485599903751e-07, + "loss": 0.7894, + "step": 8100 + }, + { + "epoch": 0.7305767236325923, + "grad_norm": 1.5882386298965296, + "learning_rate": 7.142010286865592e-07, + "loss": 0.8164, + "step": 8101 + }, + { + "epoch": 0.7306669071560626, + "grad_norm": 0.5439870876522382, + "learning_rate": 7.137536070942012e-07, + "loss": 0.5479, + "step": 8102 + }, + { + "epoch": 0.7307570906795329, + "grad_norm": 1.671060595511198, + "learning_rate": 7.133062952514786e-07, + "loss": 0.7434, + "step": 8103 + }, + { + "epoch": 0.7308472742030031, + "grad_norm": 0.6431367295495856, + "learning_rate": 7.128590931965562e-07, + "loss": 0.604, + "step": 8104 + }, + { + "epoch": 0.7309374577264733, + "grad_norm": 1.4754389772377468, + "learning_rate": 7.124120009675945e-07, + "loss": 0.8057, + "step": 8105 + }, + { + "epoch": 0.7310276412499437, + "grad_norm": 2.3055143263269753, + "learning_rate": 7.119650186027399e-07, + "loss": 0.7701, + "step": 8106 + }, + { + "epoch": 0.7311178247734139, + "grad_norm": 5.552965200311374, + "learning_rate": 7.11518146140132e-07, + "loss": 0.8643, + "step": 8107 + }, + { + "epoch": 0.7312080082968841, + "grad_norm": 1.9938934800709351, + "learning_rate": 7.110713836179007e-07, + "loss": 0.7078, + "step": 8108 + }, + { + "epoch": 0.7312981918203544, + "grad_norm": 1.9032727740062798, + "learning_rate": 7.106247310741659e-07, + "loss": 0.6738, + "step": 8109 + }, + { + "epoch": 0.7313883753438247, + "grad_norm": 3.709784077532883, + "learning_rate": 7.101781885470393e-07, + "loss": 0.8127, + "step": 8110 + }, + { + "epoch": 0.731478558867295, + "grad_norm": 1.7863800728637598, + "learning_rate": 7.097317560746203e-07, + "loss": 0.7811, + "step": 8111 + }, + { + "epoch": 0.7315687423907652, + "grad_norm": 1.6255957905082192, + "learning_rate": 7.092854336950036e-07, + "loss": 0.8005, + "step": 8112 + }, + { + "epoch": 0.7316589259142354, + "grad_norm": 2.0985317001729706, + "learning_rate": 7.0883922144627e-07, + "loss": 0.7971, + "step": 8113 + }, + { + "epoch": 0.7317491094377058, + "grad_norm": 1.561229723515413, + "learning_rate": 7.083931193664934e-07, + "loss": 0.7786, + "step": 8114 + }, + { + "epoch": 0.731839292961176, + "grad_norm": 1.572377000737845, + "learning_rate": 7.079471274937378e-07, + "loss": 0.7908, + "step": 8115 + }, + { + "epoch": 0.7319294764846462, + "grad_norm": 2.548578733907832, + "learning_rate": 7.075012458660574e-07, + "loss": 0.7298, + "step": 8116 + }, + { + "epoch": 0.7320196600081165, + "grad_norm": 0.6619014807456756, + "learning_rate": 7.070554745214976e-07, + "loss": 0.5764, + "step": 8117 + }, + { + "epoch": 0.7321098435315868, + "grad_norm": 2.18739242289189, + "learning_rate": 7.066098134980947e-07, + "loss": 0.6963, + "step": 8118 + }, + { + "epoch": 0.732200027055057, + "grad_norm": 2.5182519311411813, + "learning_rate": 7.061642628338727e-07, + "loss": 0.7541, + "step": 8119 + }, + { + "epoch": 0.7322902105785273, + "grad_norm": 1.6259967974529188, + "learning_rate": 7.057188225668513e-07, + "loss": 0.7437, + "step": 8120 + }, + { + "epoch": 0.7323803941019975, + "grad_norm": 1.6448221198611408, + "learning_rate": 7.052734927350358e-07, + "loss": 0.8909, + "step": 8121 + }, + { + "epoch": 0.7324705776254679, + "grad_norm": 2.4146626044473893, + "learning_rate": 7.048282733764252e-07, + "loss": 0.77, + "step": 8122 + }, + { + "epoch": 0.7325607611489381, + "grad_norm": 1.8436857309293178, + "learning_rate": 7.043831645290077e-07, + "loss": 0.8237, + "step": 8123 + }, + { + "epoch": 0.7326509446724083, + "grad_norm": 1.66624057409392, + "learning_rate": 7.039381662307624e-07, + "loss": 0.7143, + "step": 8124 + }, + { + "epoch": 0.7327411281958787, + "grad_norm": 1.6623804608728696, + "learning_rate": 7.034932785196601e-07, + "loss": 0.7209, + "step": 8125 + }, + { + "epoch": 0.7328313117193489, + "grad_norm": 1.5217262242344471, + "learning_rate": 7.030485014336585e-07, + "loss": 0.7282, + "step": 8126 + }, + { + "epoch": 0.7329214952428191, + "grad_norm": 1.8868108059557305, + "learning_rate": 7.026038350107118e-07, + "loss": 0.7032, + "step": 8127 + }, + { + "epoch": 0.7330116787662894, + "grad_norm": 1.5761238478442252, + "learning_rate": 7.021592792887579e-07, + "loss": 0.8772, + "step": 8128 + }, + { + "epoch": 0.7331018622897597, + "grad_norm": 1.5385454732666586, + "learning_rate": 7.01714834305732e-07, + "loss": 0.7982, + "step": 8129 + }, + { + "epoch": 0.7331920458132299, + "grad_norm": 1.8148015635230812, + "learning_rate": 7.012705000995544e-07, + "loss": 0.706, + "step": 8130 + }, + { + "epoch": 0.7332822293367002, + "grad_norm": 1.6114706080469992, + "learning_rate": 7.008262767081392e-07, + "loss": 0.7384, + "step": 8131 + }, + { + "epoch": 0.7333724128601704, + "grad_norm": 22.10920315193815, + "learning_rate": 7.003821641693892e-07, + "loss": 0.7989, + "step": 8132 + }, + { + "epoch": 0.7334625963836408, + "grad_norm": 1.8031321507250975, + "learning_rate": 6.999381625211993e-07, + "loss": 0.7319, + "step": 8133 + }, + { + "epoch": 0.733552779907111, + "grad_norm": 1.7871083077182606, + "learning_rate": 6.994942718014536e-07, + "loss": 0.6677, + "step": 8134 + }, + { + "epoch": 0.7336429634305812, + "grad_norm": 2.723177669734599, + "learning_rate": 6.990504920480282e-07, + "loss": 0.6543, + "step": 8135 + }, + { + "epoch": 0.7337331469540515, + "grad_norm": 1.7817483171476742, + "learning_rate": 6.986068232987879e-07, + "loss": 0.7435, + "step": 8136 + }, + { + "epoch": 0.7338233304775218, + "grad_norm": 1.6112738952758883, + "learning_rate": 6.981632655915888e-07, + "loss": 0.6973, + "step": 8137 + }, + { + "epoch": 0.733913514000992, + "grad_norm": 5.411407480254874, + "learning_rate": 6.977198189642783e-07, + "loss": 0.8087, + "step": 8138 + }, + { + "epoch": 0.7340036975244623, + "grad_norm": 1.9616657743665318, + "learning_rate": 6.972764834546935e-07, + "loss": 0.7705, + "step": 8139 + }, + { + "epoch": 0.7340938810479325, + "grad_norm": 2.6309235358963585, + "learning_rate": 6.96833259100663e-07, + "loss": 0.7633, + "step": 8140 + }, + { + "epoch": 0.7341840645714028, + "grad_norm": 1.9864353087247244, + "learning_rate": 6.96390145940003e-07, + "loss": 0.7199, + "step": 8141 + }, + { + "epoch": 0.7342742480948731, + "grad_norm": 32.52130086213688, + "learning_rate": 6.959471440105253e-07, + "loss": 0.6989, + "step": 8142 + }, + { + "epoch": 0.7343644316183433, + "grad_norm": 1.7578167724557452, + "learning_rate": 6.955042533500261e-07, + "loss": 0.7789, + "step": 8143 + }, + { + "epoch": 0.7344546151418135, + "grad_norm": 2.490139493443663, + "learning_rate": 6.950614739962986e-07, + "loss": 0.7036, + "step": 8144 + }, + { + "epoch": 0.7345447986652839, + "grad_norm": 1.5640749051481468, + "learning_rate": 6.946188059871198e-07, + "loss": 0.6358, + "step": 8145 + }, + { + "epoch": 0.7346349821887541, + "grad_norm": 1.4949113042540918, + "learning_rate": 6.941762493602638e-07, + "loss": 0.8593, + "step": 8146 + }, + { + "epoch": 0.7347251657122243, + "grad_norm": 1.633511772269806, + "learning_rate": 6.937338041534899e-07, + "loss": 0.6837, + "step": 8147 + }, + { + "epoch": 0.7348153492356947, + "grad_norm": 1.923292986029553, + "learning_rate": 6.932914704045505e-07, + "loss": 0.7437, + "step": 8148 + }, + { + "epoch": 0.7349055327591649, + "grad_norm": 1.6665650654659843, + "learning_rate": 6.928492481511878e-07, + "loss": 0.773, + "step": 8149 + }, + { + "epoch": 0.7349957162826352, + "grad_norm": 1.9280264187779526, + "learning_rate": 6.924071374311349e-07, + "loss": 0.7005, + "step": 8150 + }, + { + "epoch": 0.7350858998061054, + "grad_norm": 0.5969104227070275, + "learning_rate": 6.919651382821157e-07, + "loss": 0.5584, + "step": 8151 + }, + { + "epoch": 0.7351760833295757, + "grad_norm": 1.602139252809907, + "learning_rate": 6.915232507418425e-07, + "loss": 0.7167, + "step": 8152 + }, + { + "epoch": 0.735266266853046, + "grad_norm": 1.960822964549126, + "learning_rate": 6.910814748480204e-07, + "loss": 0.636, + "step": 8153 + }, + { + "epoch": 0.7353564503765162, + "grad_norm": 2.4866011140823168, + "learning_rate": 6.906398106383445e-07, + "loss": 0.7658, + "step": 8154 + }, + { + "epoch": 0.7354466338999864, + "grad_norm": 1.474907002507664, + "learning_rate": 6.901982581504994e-07, + "loss": 0.7425, + "step": 8155 + }, + { + "epoch": 0.7355368174234568, + "grad_norm": 5.697223776950352, + "learning_rate": 6.897568174221611e-07, + "loss": 0.6195, + "step": 8156 + }, + { + "epoch": 0.735627000946927, + "grad_norm": 2.000777927740006, + "learning_rate": 6.893154884909966e-07, + "loss": 0.8127, + "step": 8157 + }, + { + "epoch": 0.7357171844703972, + "grad_norm": 0.605751217817656, + "learning_rate": 6.888742713946602e-07, + "loss": 0.5837, + "step": 8158 + }, + { + "epoch": 0.7358073679938675, + "grad_norm": 1.9491632182251282, + "learning_rate": 6.884331661708018e-07, + "loss": 0.7466, + "step": 8159 + }, + { + "epoch": 0.7358975515173378, + "grad_norm": 1.5161410367578159, + "learning_rate": 6.879921728570561e-07, + "loss": 0.7448, + "step": 8160 + }, + { + "epoch": 0.735987735040808, + "grad_norm": 1.7205226426833813, + "learning_rate": 6.875512914910539e-07, + "loss": 0.8202, + "step": 8161 + }, + { + "epoch": 0.7360779185642783, + "grad_norm": 1.648481467867426, + "learning_rate": 6.871105221104119e-07, + "loss": 0.694, + "step": 8162 + }, + { + "epoch": 0.7361681020877485, + "grad_norm": 4.583247212843541, + "learning_rate": 6.866698647527391e-07, + "loss": 0.7179, + "step": 8163 + }, + { + "epoch": 0.7362582856112189, + "grad_norm": 1.5706382906951408, + "learning_rate": 6.862293194556353e-07, + "loss": 0.7695, + "step": 8164 + }, + { + "epoch": 0.7363484691346891, + "grad_norm": 1.5817665161167682, + "learning_rate": 6.857888862566896e-07, + "loss": 0.6861, + "step": 8165 + }, + { + "epoch": 0.7364386526581593, + "grad_norm": 2.0800466042579453, + "learning_rate": 6.853485651934836e-07, + "loss": 0.7096, + "step": 8166 + }, + { + "epoch": 0.7365288361816296, + "grad_norm": 1.628630909873223, + "learning_rate": 6.849083563035855e-07, + "loss": 0.6036, + "step": 8167 + }, + { + "epoch": 0.7366190197050999, + "grad_norm": 0.713245752794608, + "learning_rate": 6.844682596245592e-07, + "loss": 0.6608, + "step": 8168 + }, + { + "epoch": 0.7367092032285701, + "grad_norm": 1.8601250938680567, + "learning_rate": 6.840282751939539e-07, + "loss": 0.8066, + "step": 8169 + }, + { + "epoch": 0.7367993867520404, + "grad_norm": 1.5207278681029928, + "learning_rate": 6.835884030493126e-07, + "loss": 0.7902, + "step": 8170 + }, + { + "epoch": 0.7368895702755107, + "grad_norm": 2.118516455637963, + "learning_rate": 6.831486432281672e-07, + "loss": 0.7669, + "step": 8171 + }, + { + "epoch": 0.736979753798981, + "grad_norm": 1.9552719523167306, + "learning_rate": 6.827089957680407e-07, + "loss": 0.76, + "step": 8172 + }, + { + "epoch": 0.7370699373224512, + "grad_norm": 1.5867730556343842, + "learning_rate": 6.822694607064461e-07, + "loss": 0.8613, + "step": 8173 + }, + { + "epoch": 0.7371601208459214, + "grad_norm": 2.0269184585226907, + "learning_rate": 6.818300380808877e-07, + "loss": 0.7662, + "step": 8174 + }, + { + "epoch": 0.7372503043693918, + "grad_norm": 1.5350038098465053, + "learning_rate": 6.813907279288574e-07, + "loss": 0.7366, + "step": 8175 + }, + { + "epoch": 0.737340487892862, + "grad_norm": 1.708162679134532, + "learning_rate": 6.809515302878422e-07, + "loss": 0.7723, + "step": 8176 + }, + { + "epoch": 0.7374306714163322, + "grad_norm": 3.668470906912148, + "learning_rate": 6.80512445195315e-07, + "loss": 0.7355, + "step": 8177 + }, + { + "epoch": 0.7375208549398025, + "grad_norm": 1.599993947136874, + "learning_rate": 6.800734726887416e-07, + "loss": 0.7647, + "step": 8178 + }, + { + "epoch": 0.7376110384632728, + "grad_norm": 2.673807170554533, + "learning_rate": 6.796346128055775e-07, + "loss": 0.7455, + "step": 8179 + }, + { + "epoch": 0.737701221986743, + "grad_norm": 1.6304121854847093, + "learning_rate": 6.791958655832684e-07, + "loss": 0.7167, + "step": 8180 + }, + { + "epoch": 0.7377914055102133, + "grad_norm": 2.0080178003700024, + "learning_rate": 6.787572310592518e-07, + "loss": 0.7108, + "step": 8181 + }, + { + "epoch": 0.7378815890336835, + "grad_norm": 2.0817387645606527, + "learning_rate": 6.783187092709521e-07, + "loss": 0.835, + "step": 8182 + }, + { + "epoch": 0.7379717725571538, + "grad_norm": 2.1522102660726317, + "learning_rate": 6.778803002557891e-07, + "loss": 0.7532, + "step": 8183 + }, + { + "epoch": 0.7380619560806241, + "grad_norm": 1.4316511999470878, + "learning_rate": 6.774420040511686e-07, + "loss": 0.7814, + "step": 8184 + }, + { + "epoch": 0.7381521396040943, + "grad_norm": 1.7027491583535737, + "learning_rate": 6.770038206944886e-07, + "loss": 0.7533, + "step": 8185 + }, + { + "epoch": 0.7382423231275645, + "grad_norm": 1.5839800852236532, + "learning_rate": 6.765657502231375e-07, + "loss": 0.7595, + "step": 8186 + }, + { + "epoch": 0.7383325066510349, + "grad_norm": 1.5813420807937977, + "learning_rate": 6.761277926744939e-07, + "loss": 0.812, + "step": 8187 + }, + { + "epoch": 0.7384226901745051, + "grad_norm": 1.4708644986115331, + "learning_rate": 6.756899480859268e-07, + "loss": 0.7373, + "step": 8188 + }, + { + "epoch": 0.7385128736979754, + "grad_norm": 1.8755853375083649, + "learning_rate": 6.752522164947956e-07, + "loss": 0.7931, + "step": 8189 + }, + { + "epoch": 0.7386030572214456, + "grad_norm": 1.4868103127440815, + "learning_rate": 6.748145979384498e-07, + "loss": 0.7987, + "step": 8190 + }, + { + "epoch": 0.7386932407449159, + "grad_norm": 2.178202680518834, + "learning_rate": 6.743770924542303e-07, + "loss": 0.69, + "step": 8191 + }, + { + "epoch": 0.7387834242683862, + "grad_norm": 0.6343628764755779, + "learning_rate": 6.739397000794658e-07, + "loss": 0.5704, + "step": 8192 + }, + { + "epoch": 0.7388736077918564, + "grad_norm": 1.9463946538247325, + "learning_rate": 6.735024208514782e-07, + "loss": 0.7236, + "step": 8193 + }, + { + "epoch": 0.7389637913153266, + "grad_norm": 1.5291358210114436, + "learning_rate": 6.73065254807578e-07, + "loss": 0.6608, + "step": 8194 + }, + { + "epoch": 0.739053974838797, + "grad_norm": 1.579097485935137, + "learning_rate": 6.726282019850669e-07, + "loss": 0.8008, + "step": 8195 + }, + { + "epoch": 0.7391441583622672, + "grad_norm": 1.688505791344396, + "learning_rate": 6.721912624212376e-07, + "loss": 0.8133, + "step": 8196 + }, + { + "epoch": 0.7392343418857374, + "grad_norm": 1.9607320731113331, + "learning_rate": 6.717544361533696e-07, + "loss": 0.6448, + "step": 8197 + }, + { + "epoch": 0.7393245254092078, + "grad_norm": 1.39929224244067, + "learning_rate": 6.713177232187386e-07, + "loss": 0.7295, + "step": 8198 + }, + { + "epoch": 0.739414708932678, + "grad_norm": 1.3837078665613727, + "learning_rate": 6.708811236546041e-07, + "loss": 0.7365, + "step": 8199 + }, + { + "epoch": 0.7395048924561483, + "grad_norm": 2.320145417950567, + "learning_rate": 6.704446374982224e-07, + "loss": 0.7233, + "step": 8200 + }, + { + "epoch": 0.7395950759796185, + "grad_norm": 1.9815809749601696, + "learning_rate": 6.700082647868346e-07, + "loss": 0.7836, + "step": 8201 + }, + { + "epoch": 0.7396852595030888, + "grad_norm": 1.4614154217547546, + "learning_rate": 6.695720055576751e-07, + "loss": 0.8015, + "step": 8202 + }, + { + "epoch": 0.7397754430265591, + "grad_norm": 2.481396407028651, + "learning_rate": 6.691358598479679e-07, + "loss": 0.7598, + "step": 8203 + }, + { + "epoch": 0.7398656265500293, + "grad_norm": 1.916872725259969, + "learning_rate": 6.686998276949276e-07, + "loss": 0.7384, + "step": 8204 + }, + { + "epoch": 0.7399558100734995, + "grad_norm": 1.8563485398772819, + "learning_rate": 6.682639091357587e-07, + "loss": 0.5858, + "step": 8205 + }, + { + "epoch": 0.7400459935969699, + "grad_norm": 1.680003291762624, + "learning_rate": 6.678281042076568e-07, + "loss": 0.7356, + "step": 8206 + }, + { + "epoch": 0.7401361771204401, + "grad_norm": 1.6838375189450165, + "learning_rate": 6.673924129478059e-07, + "loss": 0.7488, + "step": 8207 + }, + { + "epoch": 0.7402263606439103, + "grad_norm": 2.1540665559607355, + "learning_rate": 6.669568353933824e-07, + "loss": 0.6671, + "step": 8208 + }, + { + "epoch": 0.7403165441673806, + "grad_norm": 1.7435739927439895, + "learning_rate": 6.665213715815519e-07, + "loss": 0.6809, + "step": 8209 + }, + { + "epoch": 0.7404067276908509, + "grad_norm": 1.7559221063425565, + "learning_rate": 6.660860215494706e-07, + "loss": 0.7619, + "step": 8210 + }, + { + "epoch": 0.7404969112143212, + "grad_norm": 1.5941842927305623, + "learning_rate": 6.656507853342852e-07, + "loss": 0.701, + "step": 8211 + }, + { + "epoch": 0.7405870947377914, + "grad_norm": 1.4839228694387177, + "learning_rate": 6.652156629731323e-07, + "loss": 0.7562, + "step": 8212 + }, + { + "epoch": 0.7406772782612616, + "grad_norm": 2.346765536461574, + "learning_rate": 6.647806545031396e-07, + "loss": 0.8102, + "step": 8213 + }, + { + "epoch": 0.740767461784732, + "grad_norm": 1.5356664998677017, + "learning_rate": 6.643457599614224e-07, + "loss": 0.6725, + "step": 8214 + }, + { + "epoch": 0.7408576453082022, + "grad_norm": 1.6333969749326451, + "learning_rate": 6.63910979385091e-07, + "loss": 0.8017, + "step": 8215 + }, + { + "epoch": 0.7409478288316724, + "grad_norm": 2.6579067673244414, + "learning_rate": 6.634763128112409e-07, + "loss": 0.8359, + "step": 8216 + }, + { + "epoch": 0.7410380123551427, + "grad_norm": 2.377534618105597, + "learning_rate": 6.630417602769622e-07, + "loss": 0.7364, + "step": 8217 + }, + { + "epoch": 0.741128195878613, + "grad_norm": 1.6361335673335595, + "learning_rate": 6.62607321819332e-07, + "loss": 0.7532, + "step": 8218 + }, + { + "epoch": 0.7412183794020832, + "grad_norm": 1.3517455048780025, + "learning_rate": 6.621729974754196e-07, + "loss": 0.7468, + "step": 8219 + }, + { + "epoch": 0.7413085629255535, + "grad_norm": 1.7071789431021467, + "learning_rate": 6.617387872822835e-07, + "loss": 0.6972, + "step": 8220 + }, + { + "epoch": 0.7413987464490238, + "grad_norm": 1.847016302341458, + "learning_rate": 6.613046912769731e-07, + "loss": 0.719, + "step": 8221 + }, + { + "epoch": 0.741488929972494, + "grad_norm": 1.732687326627828, + "learning_rate": 6.608707094965289e-07, + "loss": 0.8135, + "step": 8222 + }, + { + "epoch": 0.7415791134959643, + "grad_norm": 1.9871006066511503, + "learning_rate": 6.604368419779787e-07, + "loss": 0.6764, + "step": 8223 + }, + { + "epoch": 0.7416692970194345, + "grad_norm": 1.635182244412722, + "learning_rate": 6.600030887583434e-07, + "loss": 0.7595, + "step": 8224 + }, + { + "epoch": 0.7417594805429049, + "grad_norm": 1.777378291538495, + "learning_rate": 6.595694498746336e-07, + "loss": 0.8562, + "step": 8225 + }, + { + "epoch": 0.7418496640663751, + "grad_norm": 2.3766553029256388, + "learning_rate": 6.591359253638491e-07, + "loss": 0.7226, + "step": 8226 + }, + { + "epoch": 0.7419398475898453, + "grad_norm": 3.2521823011869753, + "learning_rate": 6.587025152629808e-07, + "loss": 0.6376, + "step": 8227 + }, + { + "epoch": 0.7420300311133156, + "grad_norm": 2.315004590485143, + "learning_rate": 6.582692196090107e-07, + "loss": 0.7908, + "step": 8228 + }, + { + "epoch": 0.7421202146367859, + "grad_norm": 1.6013732798377067, + "learning_rate": 6.578360384389074e-07, + "loss": 0.7656, + "step": 8229 + }, + { + "epoch": 0.7422103981602561, + "grad_norm": 1.7080198868793126, + "learning_rate": 6.574029717896355e-07, + "loss": 0.7265, + "step": 8230 + }, + { + "epoch": 0.7423005816837264, + "grad_norm": 1.6119818453748527, + "learning_rate": 6.569700196981436e-07, + "loss": 0.7418, + "step": 8231 + }, + { + "epoch": 0.7423907652071966, + "grad_norm": 1.7908351728247345, + "learning_rate": 6.565371822013763e-07, + "loss": 0.6535, + "step": 8232 + }, + { + "epoch": 0.742480948730667, + "grad_norm": 12.156929055985547, + "learning_rate": 6.561044593362636e-07, + "loss": 0.8018, + "step": 8233 + }, + { + "epoch": 0.7425711322541372, + "grad_norm": 1.8779939272521353, + "learning_rate": 6.556718511397288e-07, + "loss": 0.7296, + "step": 8234 + }, + { + "epoch": 0.7426613157776074, + "grad_norm": 1.5798682225478349, + "learning_rate": 6.552393576486843e-07, + "loss": 0.8616, + "step": 8235 + }, + { + "epoch": 0.7427514993010776, + "grad_norm": 2.1564510085911093, + "learning_rate": 6.548069789000325e-07, + "loss": 0.7614, + "step": 8236 + }, + { + "epoch": 0.742841682824548, + "grad_norm": 1.883509273525464, + "learning_rate": 6.543747149306673e-07, + "loss": 0.7317, + "step": 8237 + }, + { + "epoch": 0.7429318663480182, + "grad_norm": 1.6967439716817125, + "learning_rate": 6.5394256577747e-07, + "loss": 0.7984, + "step": 8238 + }, + { + "epoch": 0.7430220498714885, + "grad_norm": 2.250817044578271, + "learning_rate": 6.535105314773161e-07, + "loss": 0.777, + "step": 8239 + }, + { + "epoch": 0.7431122333949587, + "grad_norm": 4.068454779771301, + "learning_rate": 6.530786120670677e-07, + "loss": 0.6782, + "step": 8240 + }, + { + "epoch": 0.743202416918429, + "grad_norm": 1.818143346227753, + "learning_rate": 6.526468075835787e-07, + "loss": 0.8058, + "step": 8241 + }, + { + "epoch": 0.7432926004418993, + "grad_norm": 2.4704503831017193, + "learning_rate": 6.522151180636937e-07, + "loss": 0.7933, + "step": 8242 + }, + { + "epoch": 0.7433827839653695, + "grad_norm": 3.071108649311591, + "learning_rate": 6.517835435442461e-07, + "loss": 0.7111, + "step": 8243 + }, + { + "epoch": 0.7434729674888398, + "grad_norm": 2.119661806769973, + "learning_rate": 6.513520840620606e-07, + "loss": 0.7681, + "step": 8244 + }, + { + "epoch": 0.7435631510123101, + "grad_norm": 1.7156398029070938, + "learning_rate": 6.509207396539525e-07, + "loss": 0.712, + "step": 8245 + }, + { + "epoch": 0.7436533345357803, + "grad_norm": 0.6221837489782868, + "learning_rate": 6.50489510356724e-07, + "loss": 0.5526, + "step": 8246 + }, + { + "epoch": 0.7437435180592505, + "grad_norm": 2.459084530730606, + "learning_rate": 6.500583962071732e-07, + "loss": 0.6795, + "step": 8247 + }, + { + "epoch": 0.7438337015827209, + "grad_norm": 2.227316852408509, + "learning_rate": 6.496273972420827e-07, + "loss": 0.7714, + "step": 8248 + }, + { + "epoch": 0.7439238851061911, + "grad_norm": 1.822407119016834, + "learning_rate": 6.491965134982287e-07, + "loss": 0.7993, + "step": 8249 + }, + { + "epoch": 0.7440140686296614, + "grad_norm": 1.9855572039995761, + "learning_rate": 6.487657450123765e-07, + "loss": 0.714, + "step": 8250 + }, + { + "epoch": 0.7441042521531316, + "grad_norm": 1.5944643102834186, + "learning_rate": 6.483350918212814e-07, + "loss": 0.7604, + "step": 8251 + }, + { + "epoch": 0.7441944356766019, + "grad_norm": 0.7090487280189506, + "learning_rate": 6.479045539616898e-07, + "loss": 0.5848, + "step": 8252 + }, + { + "epoch": 0.7442846192000722, + "grad_norm": 1.9646196180642352, + "learning_rate": 6.474741314703358e-07, + "loss": 0.7527, + "step": 8253 + }, + { + "epoch": 0.7443748027235424, + "grad_norm": 2.3720317160559, + "learning_rate": 6.47043824383948e-07, + "loss": 0.8075, + "step": 8254 + }, + { + "epoch": 0.7444649862470126, + "grad_norm": 1.8416253472205324, + "learning_rate": 6.466136327392399e-07, + "loss": 0.7839, + "step": 8255 + }, + { + "epoch": 0.744555169770483, + "grad_norm": 1.7568031634528092, + "learning_rate": 6.461835565729206e-07, + "loss": 0.783, + "step": 8256 + }, + { + "epoch": 0.7446453532939532, + "grad_norm": 2.2260064752558244, + "learning_rate": 6.457535959216844e-07, + "loss": 0.7766, + "step": 8257 + }, + { + "epoch": 0.7447355368174234, + "grad_norm": 1.8363454162653454, + "learning_rate": 6.453237508222186e-07, + "loss": 0.7083, + "step": 8258 + }, + { + "epoch": 0.7448257203408937, + "grad_norm": 1.5645942386225131, + "learning_rate": 6.448940213112e-07, + "loss": 0.7708, + "step": 8259 + }, + { + "epoch": 0.744915903864364, + "grad_norm": 1.4534262068118426, + "learning_rate": 6.444644074252954e-07, + "loss": 0.832, + "step": 8260 + }, + { + "epoch": 0.7450060873878342, + "grad_norm": 1.9339592482248833, + "learning_rate": 6.440349092011628e-07, + "loss": 0.8197, + "step": 8261 + }, + { + "epoch": 0.7450962709113045, + "grad_norm": 1.4766983696323626, + "learning_rate": 6.436055266754475e-07, + "loss": 0.7392, + "step": 8262 + }, + { + "epoch": 0.7451864544347747, + "grad_norm": 1.8503687104531419, + "learning_rate": 6.431762598847879e-07, + "loss": 0.7207, + "step": 8263 + }, + { + "epoch": 0.7452766379582451, + "grad_norm": 1.8111283604100152, + "learning_rate": 6.427471088658111e-07, + "loss": 0.73, + "step": 8264 + }, + { + "epoch": 0.7453668214817153, + "grad_norm": 2.1594953577972564, + "learning_rate": 6.42318073655135e-07, + "loss": 0.7483, + "step": 8265 + }, + { + "epoch": 0.7454570050051855, + "grad_norm": 3.4597557237424654, + "learning_rate": 6.41889154289367e-07, + "loss": 0.7161, + "step": 8266 + }, + { + "epoch": 0.7455471885286559, + "grad_norm": 1.780526733793348, + "learning_rate": 6.414603508051055e-07, + "loss": 0.8072, + "step": 8267 + }, + { + "epoch": 0.7456373720521261, + "grad_norm": 0.635181068038768, + "learning_rate": 6.410316632389365e-07, + "loss": 0.6046, + "step": 8268 + }, + { + "epoch": 0.7457275555755963, + "grad_norm": 2.4916668291208817, + "learning_rate": 6.406030916274406e-07, + "loss": 0.755, + "step": 8269 + }, + { + "epoch": 0.7458177390990666, + "grad_norm": 2.101936051874132, + "learning_rate": 6.401746360071831e-07, + "loss": 0.7088, + "step": 8270 + }, + { + "epoch": 0.7459079226225369, + "grad_norm": 1.7283917571259921, + "learning_rate": 6.397462964147251e-07, + "loss": 0.7114, + "step": 8271 + }, + { + "epoch": 0.7459981061460071, + "grad_norm": 2.353280792550355, + "learning_rate": 6.393180728866128e-07, + "loss": 0.7387, + "step": 8272 + }, + { + "epoch": 0.7460882896694774, + "grad_norm": 1.8774281991164201, + "learning_rate": 6.388899654593853e-07, + "loss": 0.7808, + "step": 8273 + }, + { + "epoch": 0.7461784731929476, + "grad_norm": 2.109600026878788, + "learning_rate": 6.384619741695709e-07, + "loss": 0.7744, + "step": 8274 + }, + { + "epoch": 0.746268656716418, + "grad_norm": 1.4604990092022443, + "learning_rate": 6.380340990536883e-07, + "loss": 0.7485, + "step": 8275 + }, + { + "epoch": 0.7463588402398882, + "grad_norm": 2.9339503491861034, + "learning_rate": 6.37606340148247e-07, + "loss": 0.7542, + "step": 8276 + }, + { + "epoch": 0.7464490237633584, + "grad_norm": 4.036587752041523, + "learning_rate": 6.371786974897433e-07, + "loss": 0.6554, + "step": 8277 + }, + { + "epoch": 0.7465392072868287, + "grad_norm": 1.413003345704218, + "learning_rate": 6.367511711146691e-07, + "loss": 0.7776, + "step": 8278 + }, + { + "epoch": 0.746629390810299, + "grad_norm": 2.389354193514777, + "learning_rate": 6.363237610595014e-07, + "loss": 0.6961, + "step": 8279 + }, + { + "epoch": 0.7467195743337692, + "grad_norm": 2.049090402287221, + "learning_rate": 6.358964673607094e-07, + "loss": 0.7334, + "step": 8280 + }, + { + "epoch": 0.7468097578572395, + "grad_norm": 1.6133558796392242, + "learning_rate": 6.354692900547525e-07, + "loss": 0.7804, + "step": 8281 + }, + { + "epoch": 0.7468999413807097, + "grad_norm": 1.7010280754198912, + "learning_rate": 6.350422291780797e-07, + "loss": 0.7244, + "step": 8282 + }, + { + "epoch": 0.74699012490418, + "grad_norm": 2.115803959770186, + "learning_rate": 6.346152847671302e-07, + "loss": 0.8476, + "step": 8283 + }, + { + "epoch": 0.7470803084276503, + "grad_norm": 2.996636889786775, + "learning_rate": 6.34188456858334e-07, + "loss": 0.7451, + "step": 8284 + }, + { + "epoch": 0.7471704919511205, + "grad_norm": 1.7681238501697738, + "learning_rate": 6.337617454881081e-07, + "loss": 0.7005, + "step": 8285 + }, + { + "epoch": 0.7472606754745907, + "grad_norm": 2.860956973848677, + "learning_rate": 6.333351506928651e-07, + "loss": 0.6637, + "step": 8286 + }, + { + "epoch": 0.7473508589980611, + "grad_norm": 1.6660511151700523, + "learning_rate": 6.329086725090018e-07, + "loss": 0.7185, + "step": 8287 + }, + { + "epoch": 0.7474410425215313, + "grad_norm": 1.8854076813822678, + "learning_rate": 6.324823109729087e-07, + "loss": 0.7118, + "step": 8288 + }, + { + "epoch": 0.7475312260450016, + "grad_norm": 3.7873449463785165, + "learning_rate": 6.320560661209653e-07, + "loss": 0.6806, + "step": 8289 + }, + { + "epoch": 0.7476214095684719, + "grad_norm": 2.481379304298425, + "learning_rate": 6.316299379895411e-07, + "loss": 0.6332, + "step": 8290 + }, + { + "epoch": 0.7477115930919421, + "grad_norm": 2.1807155577425767, + "learning_rate": 6.312039266149965e-07, + "loss": 0.7857, + "step": 8291 + }, + { + "epoch": 0.7478017766154124, + "grad_norm": 2.595875524072892, + "learning_rate": 6.307780320336789e-07, + "loss": 0.7832, + "step": 8292 + }, + { + "epoch": 0.7478919601388826, + "grad_norm": 1.8894167104969595, + "learning_rate": 6.303522542819306e-07, + "loss": 0.8082, + "step": 8293 + }, + { + "epoch": 0.7479821436623529, + "grad_norm": 2.127853609277064, + "learning_rate": 6.299265933960796e-07, + "loss": 0.7223, + "step": 8294 + }, + { + "epoch": 0.7480723271858232, + "grad_norm": 2.3311135996168972, + "learning_rate": 6.295010494124462e-07, + "loss": 0.6992, + "step": 8295 + }, + { + "epoch": 0.7481625107092934, + "grad_norm": 1.450307760791889, + "learning_rate": 6.290756223673399e-07, + "loss": 0.7244, + "step": 8296 + }, + { + "epoch": 0.7482526942327636, + "grad_norm": 1.3491882020355885, + "learning_rate": 6.28650312297061e-07, + "loss": 0.7803, + "step": 8297 + }, + { + "epoch": 0.748342877756234, + "grad_norm": 1.9534566368834088, + "learning_rate": 6.282251192378987e-07, + "loss": 0.6688, + "step": 8298 + }, + { + "epoch": 0.7484330612797042, + "grad_norm": 3.90024452910068, + "learning_rate": 6.278000432261334e-07, + "loss": 0.6775, + "step": 8299 + }, + { + "epoch": 0.7485232448031744, + "grad_norm": 2.512506957542693, + "learning_rate": 6.273750842980345e-07, + "loss": 0.7684, + "step": 8300 + }, + { + "epoch": 0.7486134283266447, + "grad_norm": 2.122797216986253, + "learning_rate": 6.269502424898625e-07, + "loss": 0.7752, + "step": 8301 + }, + { + "epoch": 0.748703611850115, + "grad_norm": 1.9709609801534569, + "learning_rate": 6.265255178378663e-07, + "loss": 0.7907, + "step": 8302 + }, + { + "epoch": 0.7487937953735853, + "grad_norm": 1.6170977139081313, + "learning_rate": 6.261009103782861e-07, + "loss": 0.6413, + "step": 8303 + }, + { + "epoch": 0.7488839788970555, + "grad_norm": 1.7940271781470245, + "learning_rate": 6.256764201473519e-07, + "loss": 0.7922, + "step": 8304 + }, + { + "epoch": 0.7489741624205257, + "grad_norm": 0.6226104595046502, + "learning_rate": 6.252520471812835e-07, + "loss": 0.5728, + "step": 8305 + }, + { + "epoch": 0.7490643459439961, + "grad_norm": 1.8396408714319477, + "learning_rate": 6.248277915162912e-07, + "loss": 0.7637, + "step": 8306 + }, + { + "epoch": 0.7491545294674663, + "grad_norm": 1.5255866907694418, + "learning_rate": 6.244036531885731e-07, + "loss": 0.6052, + "step": 8307 + }, + { + "epoch": 0.7492447129909365, + "grad_norm": 1.6938606760949237, + "learning_rate": 6.239796322343216e-07, + "loss": 0.6827, + "step": 8308 + }, + { + "epoch": 0.7493348965144068, + "grad_norm": 1.7216263025481489, + "learning_rate": 6.235557286897137e-07, + "loss": 0.6851, + "step": 8309 + }, + { + "epoch": 0.7494250800378771, + "grad_norm": 1.811530741962194, + "learning_rate": 6.231319425909223e-07, + "loss": 0.7432, + "step": 8310 + }, + { + "epoch": 0.7495152635613473, + "grad_norm": 1.6236629120317274, + "learning_rate": 6.227082739741045e-07, + "loss": 0.7812, + "step": 8311 + }, + { + "epoch": 0.7496054470848176, + "grad_norm": 1.530543183663073, + "learning_rate": 6.222847228754113e-07, + "loss": 0.772, + "step": 8312 + }, + { + "epoch": 0.7496956306082878, + "grad_norm": 1.8976404848225206, + "learning_rate": 6.218612893309823e-07, + "loss": 0.8451, + "step": 8313 + }, + { + "epoch": 0.7497858141317582, + "grad_norm": 1.537000795283512, + "learning_rate": 6.214379733769468e-07, + "loss": 0.6896, + "step": 8314 + }, + { + "epoch": 0.7498759976552284, + "grad_norm": 1.5899292269136505, + "learning_rate": 6.21014775049425e-07, + "loss": 0.7792, + "step": 8315 + }, + { + "epoch": 0.7499661811786986, + "grad_norm": 0.5726503244382285, + "learning_rate": 6.205916943845267e-07, + "loss": 0.564, + "step": 8316 + }, + { + "epoch": 0.750056364702169, + "grad_norm": 1.758203895227975, + "learning_rate": 6.201687314183504e-07, + "loss": 0.7928, + "step": 8317 + }, + { + "epoch": 0.7501465482256392, + "grad_norm": 0.6511719280899658, + "learning_rate": 6.197458861869862e-07, + "loss": 0.6116, + "step": 8318 + }, + { + "epoch": 0.7502367317491094, + "grad_norm": 1.8642151609390545, + "learning_rate": 6.193231587265138e-07, + "loss": 0.7923, + "step": 8319 + }, + { + "epoch": 0.7503269152725797, + "grad_norm": 2.566315206599537, + "learning_rate": 6.189005490730024e-07, + "loss": 0.7302, + "step": 8320 + }, + { + "epoch": 0.75041709879605, + "grad_norm": 3.1794009067806464, + "learning_rate": 6.184780572625115e-07, + "loss": 0.7695, + "step": 8321 + }, + { + "epoch": 0.7505072823195202, + "grad_norm": 1.7683521241110387, + "learning_rate": 6.180556833310902e-07, + "loss": 0.6485, + "step": 8322 + }, + { + "epoch": 0.7505974658429905, + "grad_norm": 1.5472825409374553, + "learning_rate": 6.176334273147788e-07, + "loss": 0.7851, + "step": 8323 + }, + { + "epoch": 0.7506876493664607, + "grad_norm": 2.0376667492741327, + "learning_rate": 6.172112892496042e-07, + "loss": 0.6713, + "step": 8324 + }, + { + "epoch": 0.750777832889931, + "grad_norm": 1.627425071526747, + "learning_rate": 6.167892691715883e-07, + "loss": 0.7906, + "step": 8325 + }, + { + "epoch": 0.7508680164134013, + "grad_norm": 1.6000064641106682, + "learning_rate": 6.163673671167378e-07, + "loss": 0.7325, + "step": 8326 + }, + { + "epoch": 0.7509581999368715, + "grad_norm": 1.5734741873380047, + "learning_rate": 6.15945583121054e-07, + "loss": 0.7896, + "step": 8327 + }, + { + "epoch": 0.7510483834603418, + "grad_norm": 2.167758910875172, + "learning_rate": 6.15523917220524e-07, + "loss": 0.6974, + "step": 8328 + }, + { + "epoch": 0.7511385669838121, + "grad_norm": 3.1875951416177735, + "learning_rate": 6.151023694511273e-07, + "loss": 0.8122, + "step": 8329 + }, + { + "epoch": 0.7512287505072823, + "grad_norm": 1.8190047515001782, + "learning_rate": 6.146809398488328e-07, + "loss": 0.6913, + "step": 8330 + }, + { + "epoch": 0.7513189340307526, + "grad_norm": 1.6372853757257932, + "learning_rate": 6.142596284495989e-07, + "loss": 0.7793, + "step": 8331 + }, + { + "epoch": 0.7514091175542228, + "grad_norm": 1.7077836881477526, + "learning_rate": 6.138384352893751e-07, + "loss": 0.7063, + "step": 8332 + }, + { + "epoch": 0.7514993010776931, + "grad_norm": 8.797097022065007, + "learning_rate": 6.134173604040987e-07, + "loss": 0.7413, + "step": 8333 + }, + { + "epoch": 0.7515894846011634, + "grad_norm": 1.958863944005836, + "learning_rate": 6.129964038296984e-07, + "loss": 0.7555, + "step": 8334 + }, + { + "epoch": 0.7516796681246336, + "grad_norm": 2.0657878447963265, + "learning_rate": 6.12575565602093e-07, + "loss": 0.8165, + "step": 8335 + }, + { + "epoch": 0.7517698516481038, + "grad_norm": 1.253025636990669, + "learning_rate": 6.121548457571905e-07, + "loss": 0.8183, + "step": 8336 + }, + { + "epoch": 0.7518600351715742, + "grad_norm": 4.825689695409036, + "learning_rate": 6.11734244330889e-07, + "loss": 0.6973, + "step": 8337 + }, + { + "epoch": 0.7519502186950444, + "grad_norm": 1.7793289667411576, + "learning_rate": 6.113137613590773e-07, + "loss": 0.7613, + "step": 8338 + }, + { + "epoch": 0.7520404022185146, + "grad_norm": 1.597454403389215, + "learning_rate": 6.108933968776313e-07, + "loss": 0.8063, + "step": 8339 + }, + { + "epoch": 0.752130585741985, + "grad_norm": 3.473085872477111, + "learning_rate": 6.104731509224212e-07, + "loss": 0.7456, + "step": 8340 + }, + { + "epoch": 0.7522207692654552, + "grad_norm": 2.6965715646497475, + "learning_rate": 6.100530235293027e-07, + "loss": 0.679, + "step": 8341 + }, + { + "epoch": 0.7523109527889255, + "grad_norm": 1.8342916700421807, + "learning_rate": 6.096330147341253e-07, + "loss": 0.7701, + "step": 8342 + }, + { + "epoch": 0.7524011363123957, + "grad_norm": 2.2777207129652117, + "learning_rate": 6.09213124572725e-07, + "loss": 0.8067, + "step": 8343 + }, + { + "epoch": 0.752491319835866, + "grad_norm": 1.3897843552453661, + "learning_rate": 6.087933530809297e-07, + "loss": 0.8231, + "step": 8344 + }, + { + "epoch": 0.7525815033593363, + "grad_norm": 1.532839514506945, + "learning_rate": 6.083737002945566e-07, + "loss": 0.7372, + "step": 8345 + }, + { + "epoch": 0.7526716868828065, + "grad_norm": 1.636599515882726, + "learning_rate": 6.079541662494126e-07, + "loss": 0.7043, + "step": 8346 + }, + { + "epoch": 0.7527618704062767, + "grad_norm": 2.874567911600935, + "learning_rate": 6.075347509812954e-07, + "loss": 0.7558, + "step": 8347 + }, + { + "epoch": 0.7528520539297471, + "grad_norm": 1.7763767168776146, + "learning_rate": 6.0711545452599e-07, + "loss": 0.7036, + "step": 8348 + }, + { + "epoch": 0.7529422374532173, + "grad_norm": 1.9700673857651587, + "learning_rate": 6.066962769192756e-07, + "loss": 0.6716, + "step": 8349 + }, + { + "epoch": 0.7530324209766875, + "grad_norm": 1.9313101799239503, + "learning_rate": 6.062772181969167e-07, + "loss": 0.6858, + "step": 8350 + }, + { + "epoch": 0.7531226045001578, + "grad_norm": 1.4454759659987535, + "learning_rate": 6.058582783946706e-07, + "loss": 0.8137, + "step": 8351 + }, + { + "epoch": 0.7532127880236281, + "grad_norm": 1.8370053696032165, + "learning_rate": 6.054394575482833e-07, + "loss": 0.7413, + "step": 8352 + }, + { + "epoch": 0.7533029715470984, + "grad_norm": 1.578518733394652, + "learning_rate": 6.05020755693491e-07, + "loss": 0.6647, + "step": 8353 + }, + { + "epoch": 0.7533931550705686, + "grad_norm": 2.223938476569147, + "learning_rate": 6.046021728660198e-07, + "loss": 0.6631, + "step": 8354 + }, + { + "epoch": 0.7534833385940388, + "grad_norm": 2.187087755867989, + "learning_rate": 6.041837091015858e-07, + "loss": 0.785, + "step": 8355 + }, + { + "epoch": 0.7535735221175092, + "grad_norm": 1.5051399224210598, + "learning_rate": 6.037653644358931e-07, + "loss": 0.8001, + "step": 8356 + }, + { + "epoch": 0.7536637056409794, + "grad_norm": 1.6509915493680138, + "learning_rate": 6.033471389046393e-07, + "loss": 0.7617, + "step": 8357 + }, + { + "epoch": 0.7537538891644496, + "grad_norm": 1.524898200467674, + "learning_rate": 6.029290325435084e-07, + "loss": 0.7041, + "step": 8358 + }, + { + "epoch": 0.7538440726879199, + "grad_norm": 1.9431762609042373, + "learning_rate": 6.025110453881756e-07, + "loss": 0.808, + "step": 8359 + }, + { + "epoch": 0.7539342562113902, + "grad_norm": 2.0020556614366343, + "learning_rate": 6.020931774743061e-07, + "loss": 0.6865, + "step": 8360 + }, + { + "epoch": 0.7540244397348604, + "grad_norm": 1.5789887736563435, + "learning_rate": 6.016754288375546e-07, + "loss": 0.6706, + "step": 8361 + }, + { + "epoch": 0.7541146232583307, + "grad_norm": 1.7388477559585542, + "learning_rate": 6.012577995135665e-07, + "loss": 0.7397, + "step": 8362 + }, + { + "epoch": 0.754204806781801, + "grad_norm": 2.646531581105738, + "learning_rate": 6.008402895379743e-07, + "loss": 0.6922, + "step": 8363 + }, + { + "epoch": 0.7542949903052713, + "grad_norm": 2.2536692899701785, + "learning_rate": 6.004228989464047e-07, + "loss": 0.8294, + "step": 8364 + }, + { + "epoch": 0.7543851738287415, + "grad_norm": 2.122756671460596, + "learning_rate": 6.000056277744692e-07, + "loss": 0.7703, + "step": 8365 + }, + { + "epoch": 0.7544753573522117, + "grad_norm": 2.0582163861162988, + "learning_rate": 5.995884760577745e-07, + "loss": 0.7213, + "step": 8366 + }, + { + "epoch": 0.7545655408756821, + "grad_norm": 0.6580971425926941, + "learning_rate": 5.99171443831912e-07, + "loss": 0.5645, + "step": 8367 + }, + { + "epoch": 0.7546557243991523, + "grad_norm": 1.4523683444316406, + "learning_rate": 5.98754531132466e-07, + "loss": 0.7456, + "step": 8368 + }, + { + "epoch": 0.7547459079226225, + "grad_norm": 2.8304406901986874, + "learning_rate": 5.983377379950099e-07, + "loss": 0.7526, + "step": 8369 + }, + { + "epoch": 0.7548360914460928, + "grad_norm": 1.632346784987412, + "learning_rate": 5.979210644551067e-07, + "loss": 0.687, + "step": 8370 + }, + { + "epoch": 0.7549262749695631, + "grad_norm": 2.175086521751655, + "learning_rate": 5.975045105483091e-07, + "loss": 0.7422, + "step": 8371 + }, + { + "epoch": 0.7550164584930333, + "grad_norm": 0.6019151322878091, + "learning_rate": 5.970880763101607e-07, + "loss": 0.588, + "step": 8372 + }, + { + "epoch": 0.7551066420165036, + "grad_norm": 2.0189609570896714, + "learning_rate": 5.966717617761925e-07, + "loss": 0.7668, + "step": 8373 + }, + { + "epoch": 0.7551968255399738, + "grad_norm": 1.817307837086083, + "learning_rate": 5.962555669819276e-07, + "loss": 0.728, + "step": 8374 + }, + { + "epoch": 0.7552870090634441, + "grad_norm": 1.5241528518798746, + "learning_rate": 5.958394919628777e-07, + "loss": 0.771, + "step": 8375 + }, + { + "epoch": 0.7553771925869144, + "grad_norm": 1.3467878114311365, + "learning_rate": 5.954235367545451e-07, + "loss": 0.6918, + "step": 8376 + }, + { + "epoch": 0.7554673761103846, + "grad_norm": 1.430313223567127, + "learning_rate": 5.950077013924213e-07, + "loss": 0.7429, + "step": 8377 + }, + { + "epoch": 0.7555575596338548, + "grad_norm": 1.722793123858325, + "learning_rate": 5.945919859119865e-07, + "loss": 0.8295, + "step": 8378 + }, + { + "epoch": 0.7556477431573252, + "grad_norm": 1.9414184078703192, + "learning_rate": 5.94176390348714e-07, + "loss": 0.6385, + "step": 8379 + }, + { + "epoch": 0.7557379266807954, + "grad_norm": 1.2882172096116824, + "learning_rate": 5.937609147380622e-07, + "loss": 0.8539, + "step": 8380 + }, + { + "epoch": 0.7558281102042657, + "grad_norm": 1.5560352866724299, + "learning_rate": 5.933455591154844e-07, + "loss": 0.6925, + "step": 8381 + }, + { + "epoch": 0.7559182937277359, + "grad_norm": 2.831514129781252, + "learning_rate": 5.929303235164191e-07, + "loss": 0.7167, + "step": 8382 + }, + { + "epoch": 0.7560084772512062, + "grad_norm": 5.54605053429291, + "learning_rate": 5.92515207976297e-07, + "loss": 0.7196, + "step": 8383 + }, + { + "epoch": 0.7560986607746765, + "grad_norm": 1.757100685956032, + "learning_rate": 5.921002125305383e-07, + "loss": 0.7383, + "step": 8384 + }, + { + "epoch": 0.7561888442981467, + "grad_norm": 1.9935297016182454, + "learning_rate": 5.916853372145525e-07, + "loss": 0.6707, + "step": 8385 + }, + { + "epoch": 0.756279027821617, + "grad_norm": 2.1036067507246536, + "learning_rate": 5.912705820637389e-07, + "loss": 0.6846, + "step": 8386 + }, + { + "epoch": 0.7563692113450873, + "grad_norm": 1.7648816189496253, + "learning_rate": 5.908559471134871e-07, + "loss": 0.8119, + "step": 8387 + }, + { + "epoch": 0.7564593948685575, + "grad_norm": 2.5136342197730777, + "learning_rate": 5.904414323991764e-07, + "loss": 0.6872, + "step": 8388 + }, + { + "epoch": 0.7565495783920277, + "grad_norm": 2.0853008770372243, + "learning_rate": 5.900270379561743e-07, + "loss": 0.7538, + "step": 8389 + }, + { + "epoch": 0.7566397619154981, + "grad_norm": 2.2341992502517405, + "learning_rate": 5.896127638198399e-07, + "loss": 0.7611, + "step": 8390 + }, + { + "epoch": 0.7567299454389683, + "grad_norm": 2.0525088054732943, + "learning_rate": 5.89198610025521e-07, + "loss": 0.75, + "step": 8391 + }, + { + "epoch": 0.7568201289624386, + "grad_norm": 2.140693941885651, + "learning_rate": 5.887845766085559e-07, + "loss": 0.815, + "step": 8392 + }, + { + "epoch": 0.7569103124859088, + "grad_norm": 1.8525885463253053, + "learning_rate": 5.883706636042722e-07, + "loss": 0.7058, + "step": 8393 + }, + { + "epoch": 0.7570004960093791, + "grad_norm": 3.659195715912653, + "learning_rate": 5.879568710479879e-07, + "loss": 0.6897, + "step": 8394 + }, + { + "epoch": 0.7570906795328494, + "grad_norm": 3.5881235943093643, + "learning_rate": 5.875431989750078e-07, + "loss": 0.8357, + "step": 8395 + }, + { + "epoch": 0.7571808630563196, + "grad_norm": 1.3869472691469218, + "learning_rate": 5.871296474206313e-07, + "loss": 0.6823, + "step": 8396 + }, + { + "epoch": 0.7572710465797898, + "grad_norm": 2.298636455210492, + "learning_rate": 5.867162164201427e-07, + "loss": 0.7526, + "step": 8397 + }, + { + "epoch": 0.7573612301032602, + "grad_norm": 1.8925932420061897, + "learning_rate": 5.863029060088205e-07, + "loss": 0.7738, + "step": 8398 + }, + { + "epoch": 0.7574514136267304, + "grad_norm": 1.7336430422706128, + "learning_rate": 5.858897162219289e-07, + "loss": 0.6454, + "step": 8399 + }, + { + "epoch": 0.7575415971502006, + "grad_norm": 1.9415517096263375, + "learning_rate": 5.854766470947238e-07, + "loss": 0.7963, + "step": 8400 + }, + { + "epoch": 0.7576317806736709, + "grad_norm": 2.0219135215898434, + "learning_rate": 5.850636986624511e-07, + "loss": 0.7358, + "step": 8401 + }, + { + "epoch": 0.7577219641971412, + "grad_norm": 1.858425122148145, + "learning_rate": 5.846508709603453e-07, + "loss": 0.7546, + "step": 8402 + }, + { + "epoch": 0.7578121477206115, + "grad_norm": 1.5266112305155994, + "learning_rate": 5.842381640236318e-07, + "loss": 0.7223, + "step": 8403 + }, + { + "epoch": 0.7579023312440817, + "grad_norm": 1.5809623969357722, + "learning_rate": 5.838255778875242e-07, + "loss": 0.7477, + "step": 8404 + }, + { + "epoch": 0.7579925147675519, + "grad_norm": 1.669399960701534, + "learning_rate": 5.83413112587227e-07, + "loss": 0.7684, + "step": 8405 + }, + { + "epoch": 0.7580826982910223, + "grad_norm": 1.6935854082265487, + "learning_rate": 5.830007681579338e-07, + "loss": 0.7351, + "step": 8406 + }, + { + "epoch": 0.7581728818144925, + "grad_norm": 2.26493288189135, + "learning_rate": 5.825885446348284e-07, + "loss": 0.7611, + "step": 8407 + }, + { + "epoch": 0.7582630653379627, + "grad_norm": 1.585626130004047, + "learning_rate": 5.821764420530842e-07, + "loss": 0.6876, + "step": 8408 + }, + { + "epoch": 0.7583532488614331, + "grad_norm": 1.7754441390664575, + "learning_rate": 5.817644604478633e-07, + "loss": 0.7662, + "step": 8409 + }, + { + "epoch": 0.7584434323849033, + "grad_norm": 2.6897640007045385, + "learning_rate": 5.81352599854319e-07, + "loss": 0.6939, + "step": 8410 + }, + { + "epoch": 0.7585336159083735, + "grad_norm": 1.872671143333722, + "learning_rate": 5.809408603075938e-07, + "loss": 0.781, + "step": 8411 + }, + { + "epoch": 0.7586237994318438, + "grad_norm": 2.2098569352461195, + "learning_rate": 5.805292418428176e-07, + "loss": 0.7276, + "step": 8412 + }, + { + "epoch": 0.7587139829553141, + "grad_norm": 2.986594608010953, + "learning_rate": 5.801177444951148e-07, + "loss": 0.737, + "step": 8413 + }, + { + "epoch": 0.7588041664787843, + "grad_norm": 1.2082549376652243, + "learning_rate": 5.797063682995944e-07, + "loss": 0.7776, + "step": 8414 + }, + { + "epoch": 0.7588943500022546, + "grad_norm": 2.20329305670664, + "learning_rate": 5.792951132913584e-07, + "loss": 0.8158, + "step": 8415 + }, + { + "epoch": 0.7589845335257248, + "grad_norm": 1.9177226046546254, + "learning_rate": 5.788839795054968e-07, + "loss": 0.8063, + "step": 8416 + }, + { + "epoch": 0.7590747170491952, + "grad_norm": 1.602928732593652, + "learning_rate": 5.784729669770898e-07, + "loss": 0.6636, + "step": 8417 + }, + { + "epoch": 0.7591649005726654, + "grad_norm": 1.6011520976408404, + "learning_rate": 5.780620757412084e-07, + "loss": 0.7478, + "step": 8418 + }, + { + "epoch": 0.7592550840961356, + "grad_norm": 1.9428452008044645, + "learning_rate": 5.776513058329098e-07, + "loss": 0.7672, + "step": 8419 + }, + { + "epoch": 0.7593452676196059, + "grad_norm": 2.414560309874399, + "learning_rate": 5.772406572872459e-07, + "loss": 0.7668, + "step": 8420 + }, + { + "epoch": 0.7594354511430762, + "grad_norm": 2.2628129015625724, + "learning_rate": 5.768301301392535e-07, + "loss": 0.7738, + "step": 8421 + }, + { + "epoch": 0.7595256346665464, + "grad_norm": 1.5909460496944359, + "learning_rate": 5.764197244239615e-07, + "loss": 0.7568, + "step": 8422 + }, + { + "epoch": 0.7596158181900167, + "grad_norm": 1.8387195043673645, + "learning_rate": 5.760094401763884e-07, + "loss": 0.7866, + "step": 8423 + }, + { + "epoch": 0.7597060017134869, + "grad_norm": 1.7832588111898275, + "learning_rate": 5.755992774315414e-07, + "loss": 0.736, + "step": 8424 + }, + { + "epoch": 0.7597961852369572, + "grad_norm": 1.837524637070099, + "learning_rate": 5.751892362244183e-07, + "loss": 0.7366, + "step": 8425 + }, + { + "epoch": 0.7598863687604275, + "grad_norm": 1.861970324373035, + "learning_rate": 5.747793165900065e-07, + "loss": 0.8038, + "step": 8426 + }, + { + "epoch": 0.7599765522838977, + "grad_norm": 1.5714266036999756, + "learning_rate": 5.743695185632806e-07, + "loss": 0.8038, + "step": 8427 + }, + { + "epoch": 0.7600667358073679, + "grad_norm": 1.9887960850370976, + "learning_rate": 5.739598421792098e-07, + "loss": 0.7889, + "step": 8428 + }, + { + "epoch": 0.7601569193308383, + "grad_norm": 2.5314209491789663, + "learning_rate": 5.735502874727474e-07, + "loss": 0.7267, + "step": 8429 + }, + { + "epoch": 0.7602471028543085, + "grad_norm": 1.6418482988188126, + "learning_rate": 5.731408544788398e-07, + "loss": 0.8757, + "step": 8430 + }, + { + "epoch": 0.7603372863777788, + "grad_norm": 3.410524818105288, + "learning_rate": 5.727315432324225e-07, + "loss": 0.6644, + "step": 8431 + }, + { + "epoch": 0.760427469901249, + "grad_norm": 1.9364489504002174, + "learning_rate": 5.723223537684196e-07, + "loss": 0.7356, + "step": 8432 + }, + { + "epoch": 0.7605176534247193, + "grad_norm": 2.3058893722968863, + "learning_rate": 5.719132861217462e-07, + "loss": 0.8225, + "step": 8433 + }, + { + "epoch": 0.7606078369481896, + "grad_norm": 1.6310260253551265, + "learning_rate": 5.715043403273044e-07, + "loss": 0.7131, + "step": 8434 + }, + { + "epoch": 0.7606980204716598, + "grad_norm": 2.3634457869827674, + "learning_rate": 5.710955164199902e-07, + "loss": 0.7105, + "step": 8435 + }, + { + "epoch": 0.7607882039951301, + "grad_norm": 1.6942083744982765, + "learning_rate": 5.706868144346841e-07, + "loss": 0.7666, + "step": 8436 + }, + { + "epoch": 0.7608783875186004, + "grad_norm": 1.654834358247031, + "learning_rate": 5.702782344062613e-07, + "loss": 0.747, + "step": 8437 + }, + { + "epoch": 0.7609685710420706, + "grad_norm": 2.180911250439527, + "learning_rate": 5.698697763695826e-07, + "loss": 0.7266, + "step": 8438 + }, + { + "epoch": 0.7610587545655408, + "grad_norm": 2.1550767443239187, + "learning_rate": 5.694614403595002e-07, + "loss": 0.7726, + "step": 8439 + }, + { + "epoch": 0.7611489380890112, + "grad_norm": 1.599152104899688, + "learning_rate": 5.690532264108554e-07, + "loss": 0.661, + "step": 8440 + }, + { + "epoch": 0.7612391216124814, + "grad_norm": 1.9049556277456463, + "learning_rate": 5.686451345584795e-07, + "loss": 0.7882, + "step": 8441 + }, + { + "epoch": 0.7613293051359517, + "grad_norm": 1.865769745852192, + "learning_rate": 5.682371648371933e-07, + "loss": 0.8434, + "step": 8442 + }, + { + "epoch": 0.7614194886594219, + "grad_norm": 1.5914735430580629, + "learning_rate": 5.678293172818074e-07, + "loss": 0.7099, + "step": 8443 + }, + { + "epoch": 0.7615096721828922, + "grad_norm": 3.950920364739794, + "learning_rate": 5.674215919271204e-07, + "loss": 0.7615, + "step": 8444 + }, + { + "epoch": 0.7615998557063625, + "grad_norm": 1.9823827880614024, + "learning_rate": 5.670139888079224e-07, + "loss": 0.6776, + "step": 8445 + }, + { + "epoch": 0.7616900392298327, + "grad_norm": 2.3504948521119307, + "learning_rate": 5.666065079589924e-07, + "loss": 0.8065, + "step": 8446 + }, + { + "epoch": 0.7617802227533029, + "grad_norm": 1.7372362746287877, + "learning_rate": 5.661991494150986e-07, + "loss": 0.6969, + "step": 8447 + }, + { + "epoch": 0.7618704062767733, + "grad_norm": 1.4973729493977153, + "learning_rate": 5.657919132109999e-07, + "loss": 0.7932, + "step": 8448 + }, + { + "epoch": 0.7619605898002435, + "grad_norm": 1.9103951334043079, + "learning_rate": 5.653847993814421e-07, + "loss": 0.7787, + "step": 8449 + }, + { + "epoch": 0.7620507733237137, + "grad_norm": 1.6961236235080048, + "learning_rate": 5.649778079611647e-07, + "loss": 0.7603, + "step": 8450 + }, + { + "epoch": 0.762140956847184, + "grad_norm": 1.9286043779898867, + "learning_rate": 5.645709389848923e-07, + "loss": 0.7294, + "step": 8451 + }, + { + "epoch": 0.7622311403706543, + "grad_norm": 1.7361250974303701, + "learning_rate": 5.641641924873435e-07, + "loss": 0.7056, + "step": 8452 + }, + { + "epoch": 0.7623213238941245, + "grad_norm": 4.976415706846999, + "learning_rate": 5.637575685032217e-07, + "loss": 0.7495, + "step": 8453 + }, + { + "epoch": 0.7624115074175948, + "grad_norm": 1.7050277859837248, + "learning_rate": 5.633510670672246e-07, + "loss": 0.828, + "step": 8454 + }, + { + "epoch": 0.762501690941065, + "grad_norm": 0.6225787469930434, + "learning_rate": 5.629446882140354e-07, + "loss": 0.605, + "step": 8455 + }, + { + "epoch": 0.7625918744645354, + "grad_norm": 1.4894309421651215, + "learning_rate": 5.625384319783295e-07, + "loss": 0.7592, + "step": 8456 + }, + { + "epoch": 0.7626820579880056, + "grad_norm": 2.1932570857367635, + "learning_rate": 5.621322983947705e-07, + "loss": 0.7016, + "step": 8457 + }, + { + "epoch": 0.7627722415114758, + "grad_norm": 1.8286759940013944, + "learning_rate": 5.617262874980122e-07, + "loss": 0.7844, + "step": 8458 + }, + { + "epoch": 0.7628624250349462, + "grad_norm": 1.665627974459227, + "learning_rate": 5.613203993226981e-07, + "loss": 0.793, + "step": 8459 + }, + { + "epoch": 0.7629526085584164, + "grad_norm": 4.809343367900767, + "learning_rate": 5.609146339034599e-07, + "loss": 0.7965, + "step": 8460 + }, + { + "epoch": 0.7630427920818866, + "grad_norm": 2.090678224898197, + "learning_rate": 5.605089912749199e-07, + "loss": 0.6914, + "step": 8461 + }, + { + "epoch": 0.7631329756053569, + "grad_norm": 2.4769167481262744, + "learning_rate": 5.601034714716901e-07, + "loss": 0.6907, + "step": 8462 + }, + { + "epoch": 0.7632231591288272, + "grad_norm": 1.3958410101532375, + "learning_rate": 5.59698074528372e-07, + "loss": 0.7204, + "step": 8463 + }, + { + "epoch": 0.7633133426522974, + "grad_norm": 1.5312037558287315, + "learning_rate": 5.592928004795555e-07, + "loss": 0.7206, + "step": 8464 + }, + { + "epoch": 0.7634035261757677, + "grad_norm": 4.96999933622728, + "learning_rate": 5.58887649359822e-07, + "loss": 0.756, + "step": 8465 + }, + { + "epoch": 0.7634937096992379, + "grad_norm": 1.659011841311513, + "learning_rate": 5.584826212037393e-07, + "loss": 0.6012, + "step": 8466 + }, + { + "epoch": 0.7635838932227083, + "grad_norm": 1.8445977024127342, + "learning_rate": 5.580777160458689e-07, + "loss": 0.6704, + "step": 8467 + }, + { + "epoch": 0.7636740767461785, + "grad_norm": 1.7283020922131183, + "learning_rate": 5.576729339207574e-07, + "loss": 0.7721, + "step": 8468 + }, + { + "epoch": 0.7637642602696487, + "grad_norm": 2.8520552536051436, + "learning_rate": 5.572682748629449e-07, + "loss": 0.8093, + "step": 8469 + }, + { + "epoch": 0.763854443793119, + "grad_norm": 1.571777116219112, + "learning_rate": 5.568637389069582e-07, + "loss": 0.7311, + "step": 8470 + }, + { + "epoch": 0.7639446273165893, + "grad_norm": 1.6290119969281578, + "learning_rate": 5.564593260873145e-07, + "loss": 0.7631, + "step": 8471 + }, + { + "epoch": 0.7640348108400595, + "grad_norm": 1.6900881240937582, + "learning_rate": 5.560550364385206e-07, + "loss": 0.7887, + "step": 8472 + }, + { + "epoch": 0.7641249943635298, + "grad_norm": 2.0048592664076414, + "learning_rate": 5.556508699950728e-07, + "loss": 0.8161, + "step": 8473 + }, + { + "epoch": 0.764215177887, + "grad_norm": 1.7452219721881375, + "learning_rate": 5.552468267914577e-07, + "loss": 0.6992, + "step": 8474 + }, + { + "epoch": 0.7643053614104703, + "grad_norm": 2.253604121593949, + "learning_rate": 5.548429068621481e-07, + "loss": 0.6645, + "step": 8475 + }, + { + "epoch": 0.7643955449339406, + "grad_norm": 0.6256901506842782, + "learning_rate": 5.544391102416115e-07, + "loss": 0.5749, + "step": 8476 + }, + { + "epoch": 0.7644857284574108, + "grad_norm": 1.8990817134727014, + "learning_rate": 5.540354369643003e-07, + "loss": 0.8047, + "step": 8477 + }, + { + "epoch": 0.764575911980881, + "grad_norm": 1.9904894604493726, + "learning_rate": 5.536318870646586e-07, + "loss": 0.6037, + "step": 8478 + }, + { + "epoch": 0.7646660955043514, + "grad_norm": 2.420395823965913, + "learning_rate": 5.532284605771194e-07, + "loss": 0.6558, + "step": 8479 + }, + { + "epoch": 0.7647562790278216, + "grad_norm": 1.5892032017114601, + "learning_rate": 5.528251575361052e-07, + "loss": 0.7123, + "step": 8480 + }, + { + "epoch": 0.7648464625512919, + "grad_norm": 1.4486235761405866, + "learning_rate": 5.524219779760284e-07, + "loss": 0.7432, + "step": 8481 + }, + { + "epoch": 0.7649366460747622, + "grad_norm": 1.8120968468955203, + "learning_rate": 5.520189219312907e-07, + "loss": 0.774, + "step": 8482 + }, + { + "epoch": 0.7650268295982324, + "grad_norm": 1.83683027921053, + "learning_rate": 5.516159894362817e-07, + "loss": 0.6303, + "step": 8483 + }, + { + "epoch": 0.7651170131217027, + "grad_norm": 4.8677940909125805, + "learning_rate": 5.512131805253839e-07, + "loss": 0.683, + "step": 8484 + }, + { + "epoch": 0.7652071966451729, + "grad_norm": 2.486639753951138, + "learning_rate": 5.508104952329653e-07, + "loss": 0.7581, + "step": 8485 + }, + { + "epoch": 0.7652973801686432, + "grad_norm": 2.1465275823463537, + "learning_rate": 5.504079335933862e-07, + "loss": 0.7758, + "step": 8486 + }, + { + "epoch": 0.7653875636921135, + "grad_norm": 1.9411908342982598, + "learning_rate": 5.500054956409952e-07, + "loss": 0.7741, + "step": 8487 + }, + { + "epoch": 0.7654777472155837, + "grad_norm": 2.71707777138476, + "learning_rate": 5.496031814101303e-07, + "loss": 0.7089, + "step": 8488 + }, + { + "epoch": 0.7655679307390539, + "grad_norm": 3.969061651533868, + "learning_rate": 5.492009909351203e-07, + "loss": 0.756, + "step": 8489 + }, + { + "epoch": 0.7656581142625243, + "grad_norm": 1.7562005684300284, + "learning_rate": 5.4879892425028e-07, + "loss": 0.6685, + "step": 8490 + }, + { + "epoch": 0.7657482977859945, + "grad_norm": 2.2383632678172187, + "learning_rate": 5.483969813899184e-07, + "loss": 0.882, + "step": 8491 + }, + { + "epoch": 0.7658384813094647, + "grad_norm": 1.4240759279389412, + "learning_rate": 5.479951623883299e-07, + "loss": 0.8282, + "step": 8492 + }, + { + "epoch": 0.765928664832935, + "grad_norm": 2.130701494530655, + "learning_rate": 5.475934672798004e-07, + "loss": 0.6828, + "step": 8493 + }, + { + "epoch": 0.7660188483564053, + "grad_norm": 1.5858423364209657, + "learning_rate": 5.471918960986047e-07, + "loss": 0.7118, + "step": 8494 + }, + { + "epoch": 0.7661090318798756, + "grad_norm": 1.6612530843621434, + "learning_rate": 5.467904488790071e-07, + "loss": 0.8428, + "step": 8495 + }, + { + "epoch": 0.7661992154033458, + "grad_norm": 2.070693992158517, + "learning_rate": 5.463891256552615e-07, + "loss": 0.7559, + "step": 8496 + }, + { + "epoch": 0.766289398926816, + "grad_norm": 1.664557441570534, + "learning_rate": 5.459879264616107e-07, + "loss": 0.6507, + "step": 8497 + }, + { + "epoch": 0.7663795824502864, + "grad_norm": 1.908865338371306, + "learning_rate": 5.455868513322874e-07, + "loss": 0.6307, + "step": 8498 + }, + { + "epoch": 0.7664697659737566, + "grad_norm": 1.78534375297425, + "learning_rate": 5.451859003015143e-07, + "loss": 0.6414, + "step": 8499 + }, + { + "epoch": 0.7665599494972268, + "grad_norm": 1.796414125466219, + "learning_rate": 5.447850734035009e-07, + "loss": 0.781, + "step": 8500 + }, + { + "epoch": 0.7666501330206971, + "grad_norm": 1.7245973476452778, + "learning_rate": 5.443843706724494e-07, + "loss": 0.726, + "step": 8501 + }, + { + "epoch": 0.7667403165441674, + "grad_norm": 1.6424812351821299, + "learning_rate": 5.439837921425494e-07, + "loss": 0.6743, + "step": 8502 + }, + { + "epoch": 0.7668305000676376, + "grad_norm": 5.401236971711661, + "learning_rate": 5.435833378479807e-07, + "loss": 0.7332, + "step": 8503 + }, + { + "epoch": 0.7669206835911079, + "grad_norm": 1.7052272456135398, + "learning_rate": 5.431830078229128e-07, + "loss": 0.607, + "step": 8504 + }, + { + "epoch": 0.7670108671145782, + "grad_norm": 1.7175777946300073, + "learning_rate": 5.427828021015022e-07, + "loss": 0.7628, + "step": 8505 + }, + { + "epoch": 0.7671010506380485, + "grad_norm": 2.0431253581077375, + "learning_rate": 5.42382720717899e-07, + "loss": 0.7764, + "step": 8506 + }, + { + "epoch": 0.7671912341615187, + "grad_norm": 3.3508023909517566, + "learning_rate": 5.419827637062384e-07, + "loss": 0.777, + "step": 8507 + }, + { + "epoch": 0.7672814176849889, + "grad_norm": 1.8063043394791722, + "learning_rate": 5.415829311006487e-07, + "loss": 0.8895, + "step": 8508 + }, + { + "epoch": 0.7673716012084593, + "grad_norm": 1.7759552950473483, + "learning_rate": 5.411832229352447e-07, + "loss": 0.8288, + "step": 8509 + }, + { + "epoch": 0.7674617847319295, + "grad_norm": 1.9806000254083567, + "learning_rate": 5.407836392441319e-07, + "loss": 0.8433, + "step": 8510 + }, + { + "epoch": 0.7675519682553997, + "grad_norm": 1.5047625593278275, + "learning_rate": 5.403841800614049e-07, + "loss": 0.7604, + "step": 8511 + }, + { + "epoch": 0.76764215177887, + "grad_norm": 1.3665050438160777, + "learning_rate": 5.39984845421148e-07, + "loss": 0.7028, + "step": 8512 + }, + { + "epoch": 0.7677323353023403, + "grad_norm": 2.651964388025685, + "learning_rate": 5.395856353574344e-07, + "loss": 0.73, + "step": 8513 + }, + { + "epoch": 0.7678225188258105, + "grad_norm": 0.591405934544105, + "learning_rate": 5.391865499043275e-07, + "loss": 0.5665, + "step": 8514 + }, + { + "epoch": 0.7679127023492808, + "grad_norm": 1.5989060118954737, + "learning_rate": 5.387875890958788e-07, + "loss": 0.7531, + "step": 8515 + }, + { + "epoch": 0.768002885872751, + "grad_norm": 1.4901682987787999, + "learning_rate": 5.383887529661298e-07, + "loss": 0.7709, + "step": 8516 + }, + { + "epoch": 0.7680930693962214, + "grad_norm": 2.0652000786603417, + "learning_rate": 5.379900415491116e-07, + "loss": 0.7043, + "step": 8517 + }, + { + "epoch": 0.7681832529196916, + "grad_norm": 1.9768896020717235, + "learning_rate": 5.375914548788447e-07, + "loss": 0.695, + "step": 8518 + }, + { + "epoch": 0.7682734364431618, + "grad_norm": 1.602967478726254, + "learning_rate": 5.371929929893384e-07, + "loss": 0.6252, + "step": 8519 + }, + { + "epoch": 0.768363619966632, + "grad_norm": 1.8688908552587091, + "learning_rate": 5.367946559145917e-07, + "loss": 0.7085, + "step": 8520 + }, + { + "epoch": 0.7684538034901024, + "grad_norm": 1.5000707291775945, + "learning_rate": 5.363964436885935e-07, + "loss": 0.7652, + "step": 8521 + }, + { + "epoch": 0.7685439870135726, + "grad_norm": 1.3917068388084315, + "learning_rate": 5.359983563453199e-07, + "loss": 0.8292, + "step": 8522 + }, + { + "epoch": 0.7686341705370429, + "grad_norm": 1.8090001220281338, + "learning_rate": 5.356003939187402e-07, + "loss": 0.8613, + "step": 8523 + }, + { + "epoch": 0.7687243540605131, + "grad_norm": 1.6907111098382663, + "learning_rate": 5.352025564428082e-07, + "loss": 0.7237, + "step": 8524 + }, + { + "epoch": 0.7688145375839834, + "grad_norm": 1.6933669075803846, + "learning_rate": 5.348048439514723e-07, + "loss": 0.788, + "step": 8525 + }, + { + "epoch": 0.7689047211074537, + "grad_norm": 1.860722414233792, + "learning_rate": 5.344072564786653e-07, + "loss": 0.6274, + "step": 8526 + }, + { + "epoch": 0.7689949046309239, + "grad_norm": 1.4575728568029778, + "learning_rate": 5.340097940583123e-07, + "loss": 0.7578, + "step": 8527 + }, + { + "epoch": 0.7690850881543942, + "grad_norm": 2.850492260900837, + "learning_rate": 5.336124567243275e-07, + "loss": 0.7063, + "step": 8528 + }, + { + "epoch": 0.7691752716778645, + "grad_norm": 1.8224873791167375, + "learning_rate": 5.33215244510613e-07, + "loss": 0.7518, + "step": 8529 + }, + { + "epoch": 0.7692654552013347, + "grad_norm": 1.768083194630837, + "learning_rate": 5.328181574510624e-07, + "loss": 0.7433, + "step": 8530 + }, + { + "epoch": 0.769355638724805, + "grad_norm": 1.7319450195210935, + "learning_rate": 5.324211955795559e-07, + "loss": 0.7949, + "step": 8531 + }, + { + "epoch": 0.7694458222482753, + "grad_norm": 2.69752981782825, + "learning_rate": 5.320243589299651e-07, + "loss": 0.6755, + "step": 8532 + }, + { + "epoch": 0.7695360057717455, + "grad_norm": 1.7231486129388112, + "learning_rate": 5.316276475361505e-07, + "loss": 0.7027, + "step": 8533 + }, + { + "epoch": 0.7696261892952158, + "grad_norm": 8.357350441167606, + "learning_rate": 5.312310614319613e-07, + "loss": 0.7886, + "step": 8534 + }, + { + "epoch": 0.769716372818686, + "grad_norm": 3.148421748952536, + "learning_rate": 5.308346006512367e-07, + "loss": 0.7203, + "step": 8535 + }, + { + "epoch": 0.7698065563421563, + "grad_norm": 1.6468324485652635, + "learning_rate": 5.30438265227805e-07, + "loss": 0.678, + "step": 8536 + }, + { + "epoch": 0.7698967398656266, + "grad_norm": 0.6169067661798605, + "learning_rate": 5.300420551954837e-07, + "loss": 0.5604, + "step": 8537 + }, + { + "epoch": 0.7699869233890968, + "grad_norm": 1.4348283072181347, + "learning_rate": 5.296459705880798e-07, + "loss": 0.6746, + "step": 8538 + }, + { + "epoch": 0.770077106912567, + "grad_norm": 1.4488356256072277, + "learning_rate": 5.292500114393881e-07, + "loss": 0.7995, + "step": 8539 + }, + { + "epoch": 0.7701672904360374, + "grad_norm": 2.1588268638758734, + "learning_rate": 5.288541777831963e-07, + "loss": 0.7712, + "step": 8540 + }, + { + "epoch": 0.7702574739595076, + "grad_norm": 1.5050520062253234, + "learning_rate": 5.284584696532772e-07, + "loss": 0.799, + "step": 8541 + }, + { + "epoch": 0.7703476574829778, + "grad_norm": 1.6770184269383481, + "learning_rate": 5.280628870833954e-07, + "loss": 0.7509, + "step": 8542 + }, + { + "epoch": 0.7704378410064481, + "grad_norm": 1.6465961603882335, + "learning_rate": 5.276674301073045e-07, + "loss": 0.7755, + "step": 8543 + }, + { + "epoch": 0.7705280245299184, + "grad_norm": 1.4643531893425907, + "learning_rate": 5.272720987587467e-07, + "loss": 0.7623, + "step": 8544 + }, + { + "epoch": 0.7706182080533887, + "grad_norm": 2.5539068960128906, + "learning_rate": 5.268768930714545e-07, + "loss": 0.7018, + "step": 8545 + }, + { + "epoch": 0.7707083915768589, + "grad_norm": 1.5091390358164647, + "learning_rate": 5.264818130791473e-07, + "loss": 0.7764, + "step": 8546 + }, + { + "epoch": 0.7707985751003291, + "grad_norm": 1.7641727923234494, + "learning_rate": 5.260868588155378e-07, + "loss": 0.7546, + "step": 8547 + }, + { + "epoch": 0.7708887586237995, + "grad_norm": 1.3783976884107418, + "learning_rate": 5.256920303143242e-07, + "loss": 0.7753, + "step": 8548 + }, + { + "epoch": 0.7709789421472697, + "grad_norm": 1.3802500976385583, + "learning_rate": 5.252973276091956e-07, + "loss": 0.8591, + "step": 8549 + }, + { + "epoch": 0.7710691256707399, + "grad_norm": 1.5187424522180952, + "learning_rate": 5.249027507338307e-07, + "loss": 0.8177, + "step": 8550 + }, + { + "epoch": 0.7711593091942102, + "grad_norm": 1.41440034487103, + "learning_rate": 5.245082997218966e-07, + "loss": 0.719, + "step": 8551 + }, + { + "epoch": 0.7712494927176805, + "grad_norm": 1.7651239545051975, + "learning_rate": 5.241139746070499e-07, + "loss": 0.7725, + "step": 8552 + }, + { + "epoch": 0.7713396762411507, + "grad_norm": 1.480217582714901, + "learning_rate": 5.237197754229376e-07, + "loss": 0.6284, + "step": 8553 + }, + { + "epoch": 0.771429859764621, + "grad_norm": 3.174106057542244, + "learning_rate": 5.233257022031931e-07, + "loss": 0.6536, + "step": 8554 + }, + { + "epoch": 0.7715200432880913, + "grad_norm": 2.3493323412505935, + "learning_rate": 5.229317549814432e-07, + "loss": 0.7592, + "step": 8555 + }, + { + "epoch": 0.7716102268115616, + "grad_norm": 2.3461141234182405, + "learning_rate": 5.225379337912998e-07, + "loss": 0.8271, + "step": 8556 + }, + { + "epoch": 0.7717004103350318, + "grad_norm": 1.6607931912335898, + "learning_rate": 5.221442386663663e-07, + "loss": 0.7602, + "step": 8557 + }, + { + "epoch": 0.771790593858502, + "grad_norm": 1.4710973284774587, + "learning_rate": 5.217506696402354e-07, + "loss": 0.7877, + "step": 8558 + }, + { + "epoch": 0.7718807773819724, + "grad_norm": 3.3906955975333197, + "learning_rate": 5.213572267464883e-07, + "loss": 0.7675, + "step": 8559 + }, + { + "epoch": 0.7719709609054426, + "grad_norm": 3.588741361527259, + "learning_rate": 5.209639100186965e-07, + "loss": 0.7241, + "step": 8560 + }, + { + "epoch": 0.7720611444289128, + "grad_norm": 0.6946997567904656, + "learning_rate": 5.205707194904179e-07, + "loss": 0.6412, + "step": 8561 + }, + { + "epoch": 0.7721513279523831, + "grad_norm": 1.9676174963411865, + "learning_rate": 5.201776551952042e-07, + "loss": 0.6729, + "step": 8562 + }, + { + "epoch": 0.7722415114758534, + "grad_norm": 2.3930957467818503, + "learning_rate": 5.197847171665914e-07, + "loss": 0.7125, + "step": 8563 + }, + { + "epoch": 0.7723316949993236, + "grad_norm": 1.9792608004486492, + "learning_rate": 5.193919054381095e-07, + "loss": 0.7361, + "step": 8564 + }, + { + "epoch": 0.7724218785227939, + "grad_norm": 1.6998041657240237, + "learning_rate": 5.189992200432738e-07, + "loss": 0.7123, + "step": 8565 + }, + { + "epoch": 0.7725120620462641, + "grad_norm": 2.5323424630907083, + "learning_rate": 5.186066610155906e-07, + "loss": 0.763, + "step": 8566 + }, + { + "epoch": 0.7726022455697344, + "grad_norm": 1.766831989770129, + "learning_rate": 5.182142283885555e-07, + "loss": 0.8017, + "step": 8567 + }, + { + "epoch": 0.7726924290932047, + "grad_norm": 0.6630467219266584, + "learning_rate": 5.178219221956528e-07, + "loss": 0.5313, + "step": 8568 + }, + { + "epoch": 0.7727826126166749, + "grad_norm": 1.6882377001058577, + "learning_rate": 5.174297424703565e-07, + "loss": 0.7676, + "step": 8569 + }, + { + "epoch": 0.7728727961401451, + "grad_norm": 1.6395852563532498, + "learning_rate": 5.170376892461299e-07, + "loss": 0.6777, + "step": 8570 + }, + { + "epoch": 0.7729629796636155, + "grad_norm": 0.7245286757781512, + "learning_rate": 5.16645762556424e-07, + "loss": 0.6309, + "step": 8571 + }, + { + "epoch": 0.7730531631870857, + "grad_norm": 1.511899794351193, + "learning_rate": 5.162539624346809e-07, + "loss": 0.7939, + "step": 8572 + }, + { + "epoch": 0.773143346710556, + "grad_norm": 2.5908255578831656, + "learning_rate": 5.158622889143309e-07, + "loss": 0.8087, + "step": 8573 + }, + { + "epoch": 0.7732335302340262, + "grad_norm": 1.8598775745422482, + "learning_rate": 5.154707420287939e-07, + "loss": 0.7067, + "step": 8574 + }, + { + "epoch": 0.7733237137574965, + "grad_norm": 2.574474825589482, + "learning_rate": 5.150793218114793e-07, + "loss": 0.6759, + "step": 8575 + }, + { + "epoch": 0.7734138972809668, + "grad_norm": 1.9355816727328914, + "learning_rate": 5.146880282957837e-07, + "loss": 0.7376, + "step": 8576 + }, + { + "epoch": 0.773504080804437, + "grad_norm": 1.624941091203303, + "learning_rate": 5.142968615150964e-07, + "loss": 0.7364, + "step": 8577 + }, + { + "epoch": 0.7735942643279073, + "grad_norm": 3.765608031681816, + "learning_rate": 5.139058215027921e-07, + "loss": 0.6778, + "step": 8578 + }, + { + "epoch": 0.7736844478513776, + "grad_norm": 1.3820018655209572, + "learning_rate": 5.135149082922383e-07, + "loss": 0.7338, + "step": 8579 + }, + { + "epoch": 0.7737746313748478, + "grad_norm": 5.983021874498908, + "learning_rate": 5.131241219167879e-07, + "loss": 0.6776, + "step": 8580 + }, + { + "epoch": 0.773864814898318, + "grad_norm": 0.8074418861334399, + "learning_rate": 5.127334624097869e-07, + "loss": 0.6229, + "step": 8581 + }, + { + "epoch": 0.7739549984217884, + "grad_norm": 2.4218302076566016, + "learning_rate": 5.123429298045672e-07, + "loss": 0.8022, + "step": 8582 + }, + { + "epoch": 0.7740451819452586, + "grad_norm": 1.5218867307363602, + "learning_rate": 5.119525241344515e-07, + "loss": 0.6365, + "step": 8583 + }, + { + "epoch": 0.7741353654687289, + "grad_norm": 1.584716101049618, + "learning_rate": 5.115622454327515e-07, + "loss": 0.7527, + "step": 8584 + }, + { + "epoch": 0.7742255489921991, + "grad_norm": 2.7347424069571282, + "learning_rate": 5.11172093732768e-07, + "loss": 0.7326, + "step": 8585 + }, + { + "epoch": 0.7743157325156694, + "grad_norm": 1.7522404497833086, + "learning_rate": 5.107820690677911e-07, + "loss": 0.7749, + "step": 8586 + }, + { + "epoch": 0.7744059160391397, + "grad_norm": 1.4978052613377726, + "learning_rate": 5.103921714710991e-07, + "loss": 0.6935, + "step": 8587 + }, + { + "epoch": 0.7744960995626099, + "grad_norm": 1.9411100165404012, + "learning_rate": 5.100024009759605e-07, + "loss": 0.7608, + "step": 8588 + }, + { + "epoch": 0.7745862830860801, + "grad_norm": 1.9400106591000665, + "learning_rate": 5.09612757615633e-07, + "loss": 0.7841, + "step": 8589 + }, + { + "epoch": 0.7746764666095505, + "grad_norm": 1.6598639735825098, + "learning_rate": 5.092232414233628e-07, + "loss": 0.7001, + "step": 8590 + }, + { + "epoch": 0.7747666501330207, + "grad_norm": 1.7819889610453814, + "learning_rate": 5.088338524323858e-07, + "loss": 0.7196, + "step": 8591 + }, + { + "epoch": 0.7748568336564909, + "grad_norm": 2.5295503349416024, + "learning_rate": 5.084445906759271e-07, + "loss": 0.8067, + "step": 8592 + }, + { + "epoch": 0.7749470171799612, + "grad_norm": 1.6715882670437816, + "learning_rate": 5.080554561871995e-07, + "loss": 0.809, + "step": 8593 + }, + { + "epoch": 0.7750372007034315, + "grad_norm": 1.4466587704268683, + "learning_rate": 5.076664489994078e-07, + "loss": 0.7778, + "step": 8594 + }, + { + "epoch": 0.7751273842269017, + "grad_norm": 1.850116019863832, + "learning_rate": 5.07277569145742e-07, + "loss": 0.7159, + "step": 8595 + }, + { + "epoch": 0.775217567750372, + "grad_norm": 3.431363402335748, + "learning_rate": 5.068888166593861e-07, + "loss": 0.6649, + "step": 8596 + }, + { + "epoch": 0.7753077512738422, + "grad_norm": 1.7826000250831002, + "learning_rate": 5.065001915735087e-07, + "loss": 0.7384, + "step": 8597 + }, + { + "epoch": 0.7753979347973126, + "grad_norm": 5.098998004991493, + "learning_rate": 5.061116939212702e-07, + "loss": 0.7114, + "step": 8598 + }, + { + "epoch": 0.7754881183207828, + "grad_norm": 1.7736318204515933, + "learning_rate": 5.05723323735819e-07, + "loss": 0.7436, + "step": 8599 + }, + { + "epoch": 0.775578301844253, + "grad_norm": 0.6251621751187475, + "learning_rate": 5.053350810502932e-07, + "loss": 0.5837, + "step": 8600 + }, + { + "epoch": 0.7756684853677234, + "grad_norm": 2.599471342920478, + "learning_rate": 5.049469658978202e-07, + "loss": 0.7412, + "step": 8601 + }, + { + "epoch": 0.7757586688911936, + "grad_norm": 0.6989244875456864, + "learning_rate": 5.045589783115147e-07, + "loss": 0.5317, + "step": 8602 + }, + { + "epoch": 0.7758488524146638, + "grad_norm": 10.290030489654894, + "learning_rate": 5.041711183244842e-07, + "loss": 0.7889, + "step": 8603 + }, + { + "epoch": 0.7759390359381341, + "grad_norm": 4.053137216424583, + "learning_rate": 5.037833859698211e-07, + "loss": 0.7631, + "step": 8604 + }, + { + "epoch": 0.7760292194616044, + "grad_norm": 1.7330734593032076, + "learning_rate": 5.033957812806096e-07, + "loss": 0.8131, + "step": 8605 + }, + { + "epoch": 0.7761194029850746, + "grad_norm": 2.3871991911720563, + "learning_rate": 5.030083042899223e-07, + "loss": 0.7956, + "step": 8606 + }, + { + "epoch": 0.7762095865085449, + "grad_norm": 1.727330654030806, + "learning_rate": 5.026209550308207e-07, + "loss": 0.7424, + "step": 8607 + }, + { + "epoch": 0.7762997700320151, + "grad_norm": 4.516270574865451, + "learning_rate": 5.022337335363558e-07, + "loss": 0.8075, + "step": 8608 + }, + { + "epoch": 0.7763899535554855, + "grad_norm": 1.5780577692497824, + "learning_rate": 5.018466398395677e-07, + "loss": 0.7719, + "step": 8609 + }, + { + "epoch": 0.7764801370789557, + "grad_norm": 4.780702048424025, + "learning_rate": 5.01459673973484e-07, + "loss": 0.6859, + "step": 8610 + }, + { + "epoch": 0.7765703206024259, + "grad_norm": 2.2360222353230594, + "learning_rate": 5.01072835971125e-07, + "loss": 0.8152, + "step": 8611 + }, + { + "epoch": 0.7766605041258962, + "grad_norm": 1.766170248727003, + "learning_rate": 5.006861258654959e-07, + "loss": 0.6522, + "step": 8612 + }, + { + "epoch": 0.7767506876493665, + "grad_norm": 2.1653206263194384, + "learning_rate": 5.002995436895938e-07, + "loss": 0.7767, + "step": 8613 + }, + { + "epoch": 0.7768408711728367, + "grad_norm": 1.9540514770372603, + "learning_rate": 4.999130894764039e-07, + "loss": 0.7055, + "step": 8614 + }, + { + "epoch": 0.776931054696307, + "grad_norm": 1.8119290866786089, + "learning_rate": 4.995267632589006e-07, + "loss": 0.6916, + "step": 8615 + }, + { + "epoch": 0.7770212382197772, + "grad_norm": 0.7177774234583203, + "learning_rate": 4.99140565070048e-07, + "loss": 0.581, + "step": 8616 + }, + { + "epoch": 0.7771114217432475, + "grad_norm": 1.4173968714584115, + "learning_rate": 4.987544949427969e-07, + "loss": 0.7691, + "step": 8617 + }, + { + "epoch": 0.7772016052667178, + "grad_norm": 0.6608049264248516, + "learning_rate": 4.98368552910091e-07, + "loss": 0.5881, + "step": 8618 + }, + { + "epoch": 0.777291788790188, + "grad_norm": 4.009528255729855, + "learning_rate": 4.979827390048596e-07, + "loss": 0.8631, + "step": 8619 + }, + { + "epoch": 0.7773819723136582, + "grad_norm": 1.5309564737303236, + "learning_rate": 4.975970532600231e-07, + "loss": 0.7139, + "step": 8620 + }, + { + "epoch": 0.7774721558371286, + "grad_norm": 2.9006857159964503, + "learning_rate": 4.972114957084901e-07, + "loss": 0.7701, + "step": 8621 + }, + { + "epoch": 0.7775623393605988, + "grad_norm": 2.495543226683589, + "learning_rate": 4.968260663831585e-07, + "loss": 0.7269, + "step": 8622 + }, + { + "epoch": 0.777652522884069, + "grad_norm": 1.3279929151489671, + "learning_rate": 4.964407653169154e-07, + "loss": 0.8581, + "step": 8623 + }, + { + "epoch": 0.7777427064075394, + "grad_norm": 1.4047838409279774, + "learning_rate": 4.960555925426366e-07, + "loss": 0.7149, + "step": 8624 + }, + { + "epoch": 0.7778328899310096, + "grad_norm": 2.06312089301957, + "learning_rate": 4.956705480931876e-07, + "loss": 0.8047, + "step": 8625 + }, + { + "epoch": 0.7779230734544799, + "grad_norm": 1.3619435040076078, + "learning_rate": 4.952856320014225e-07, + "loss": 0.7275, + "step": 8626 + }, + { + "epoch": 0.7780132569779501, + "grad_norm": 1.3467592211744535, + "learning_rate": 4.949008443001838e-07, + "loss": 0.6854, + "step": 8627 + }, + { + "epoch": 0.7781034405014204, + "grad_norm": 1.72517677728177, + "learning_rate": 4.945161850223041e-07, + "loss": 0.7599, + "step": 8628 + }, + { + "epoch": 0.7781936240248907, + "grad_norm": 2.1349208597349794, + "learning_rate": 4.941316542006044e-07, + "loss": 0.6763, + "step": 8629 + }, + { + "epoch": 0.7782838075483609, + "grad_norm": 1.3845918059379214, + "learning_rate": 4.937472518678956e-07, + "loss": 0.7708, + "step": 8630 + }, + { + "epoch": 0.7783739910718311, + "grad_norm": 1.994831081495667, + "learning_rate": 4.93362978056977e-07, + "loss": 0.746, + "step": 8631 + }, + { + "epoch": 0.7784641745953015, + "grad_norm": 1.7329799093104363, + "learning_rate": 4.929788328006355e-07, + "loss": 0.6989, + "step": 8632 + }, + { + "epoch": 0.7785543581187717, + "grad_norm": 2.0241081864238137, + "learning_rate": 4.925948161316506e-07, + "loss": 0.8192, + "step": 8633 + }, + { + "epoch": 0.778644541642242, + "grad_norm": 1.6553687144177933, + "learning_rate": 4.922109280827868e-07, + "loss": 0.736, + "step": 8634 + }, + { + "epoch": 0.7787347251657122, + "grad_norm": 2.0061855032467713, + "learning_rate": 4.918271686868016e-07, + "loss": 0.7418, + "step": 8635 + }, + { + "epoch": 0.7788249086891825, + "grad_norm": 1.9415453855213651, + "learning_rate": 4.914435379764379e-07, + "loss": 0.79, + "step": 8636 + }, + { + "epoch": 0.7789150922126528, + "grad_norm": 1.4463026624371127, + "learning_rate": 4.910600359844294e-07, + "loss": 0.7389, + "step": 8637 + }, + { + "epoch": 0.779005275736123, + "grad_norm": 2.0438704242282966, + "learning_rate": 4.90676662743499e-07, + "loss": 0.6554, + "step": 8638 + }, + { + "epoch": 0.7790954592595932, + "grad_norm": 1.7675057580792035, + "learning_rate": 4.902934182863581e-07, + "loss": 0.7415, + "step": 8639 + }, + { + "epoch": 0.7791856427830636, + "grad_norm": 2.0440190316077556, + "learning_rate": 4.899103026457069e-07, + "loss": 0.6234, + "step": 8640 + }, + { + "epoch": 0.7792758263065338, + "grad_norm": 0.592184185532124, + "learning_rate": 4.895273158542361e-07, + "loss": 0.5306, + "step": 8641 + }, + { + "epoch": 0.779366009830004, + "grad_norm": 1.6062110985227247, + "learning_rate": 4.891444579446227e-07, + "loss": 0.7614, + "step": 8642 + }, + { + "epoch": 0.7794561933534743, + "grad_norm": 2.1391039268210443, + "learning_rate": 4.887617289495349e-07, + "loss": 0.7037, + "step": 8643 + }, + { + "epoch": 0.7795463768769446, + "grad_norm": 0.5934040668022907, + "learning_rate": 4.883791289016292e-07, + "loss": 0.5118, + "step": 8644 + }, + { + "epoch": 0.7796365604004148, + "grad_norm": 2.4071274185655827, + "learning_rate": 4.879966578335514e-07, + "loss": 0.7549, + "step": 8645 + }, + { + "epoch": 0.7797267439238851, + "grad_norm": 1.9352849788917201, + "learning_rate": 4.876143157779358e-07, + "loss": 0.6435, + "step": 8646 + }, + { + "epoch": 0.7798169274473554, + "grad_norm": 1.9524767601007498, + "learning_rate": 4.872321027674058e-07, + "loss": 0.6371, + "step": 8647 + }, + { + "epoch": 0.7799071109708257, + "grad_norm": 1.810674208162321, + "learning_rate": 4.868500188345748e-07, + "loss": 0.8186, + "step": 8648 + }, + { + "epoch": 0.7799972944942959, + "grad_norm": 1.6645542188381195, + "learning_rate": 4.864680640120425e-07, + "loss": 0.8082, + "step": 8649 + }, + { + "epoch": 0.7800874780177661, + "grad_norm": 2.2073466657909706, + "learning_rate": 4.860862383324016e-07, + "loss": 0.7558, + "step": 8650 + }, + { + "epoch": 0.7801776615412365, + "grad_norm": 1.479242949690271, + "learning_rate": 4.857045418282295e-07, + "loss": 0.7398, + "step": 8651 + }, + { + "epoch": 0.7802678450647067, + "grad_norm": 2.1800821145367544, + "learning_rate": 4.853229745320966e-07, + "loss": 0.6959, + "step": 8652 + }, + { + "epoch": 0.7803580285881769, + "grad_norm": 1.712843438676244, + "learning_rate": 4.849415364765587e-07, + "loss": 0.7306, + "step": 8653 + }, + { + "epoch": 0.7804482121116472, + "grad_norm": 2.1750648181947105, + "learning_rate": 4.845602276941631e-07, + "loss": 0.8074, + "step": 8654 + }, + { + "epoch": 0.7805383956351175, + "grad_norm": 1.5946131594084407, + "learning_rate": 4.841790482174449e-07, + "loss": 0.7903, + "step": 8655 + }, + { + "epoch": 0.7806285791585877, + "grad_norm": 1.8204430136327803, + "learning_rate": 4.837979980789282e-07, + "loss": 0.7383, + "step": 8656 + }, + { + "epoch": 0.780718762682058, + "grad_norm": 1.6126053620448118, + "learning_rate": 4.834170773111273e-07, + "loss": 0.688, + "step": 8657 + }, + { + "epoch": 0.7808089462055282, + "grad_norm": 1.6347419983171692, + "learning_rate": 4.830362859465431e-07, + "loss": 0.7653, + "step": 8658 + }, + { + "epoch": 0.7808991297289986, + "grad_norm": 1.9614090675437696, + "learning_rate": 4.826556240176675e-07, + "loss": 0.8129, + "step": 8659 + }, + { + "epoch": 0.7809893132524688, + "grad_norm": 1.7983076272399368, + "learning_rate": 4.822750915569807e-07, + "loss": 0.7645, + "step": 8660 + }, + { + "epoch": 0.781079496775939, + "grad_norm": 3.437707235851929, + "learning_rate": 4.818946885969514e-07, + "loss": 0.821, + "step": 8661 + }, + { + "epoch": 0.7811696802994093, + "grad_norm": 3.0805936102603906, + "learning_rate": 4.815144151700383e-07, + "loss": 0.7907, + "step": 8662 + }, + { + "epoch": 0.7812598638228796, + "grad_norm": 2.218947925604702, + "learning_rate": 4.811342713086885e-07, + "loss": 0.7489, + "step": 8663 + }, + { + "epoch": 0.7813500473463498, + "grad_norm": 2.352574634401831, + "learning_rate": 4.807542570453367e-07, + "loss": 0.7869, + "step": 8664 + }, + { + "epoch": 0.7814402308698201, + "grad_norm": 1.7684012672445661, + "learning_rate": 4.803743724124098e-07, + "loss": 0.7135, + "step": 8665 + }, + { + "epoch": 0.7815304143932903, + "grad_norm": 1.9793183787239101, + "learning_rate": 4.799946174423192e-07, + "loss": 0.7548, + "step": 8666 + }, + { + "epoch": 0.7816205979167606, + "grad_norm": 0.5954394898423907, + "learning_rate": 4.796149921674706e-07, + "loss": 0.5326, + "step": 8667 + }, + { + "epoch": 0.7817107814402309, + "grad_norm": 1.5845743719759182, + "learning_rate": 4.792354966202534e-07, + "loss": 0.7046, + "step": 8668 + }, + { + "epoch": 0.7818009649637011, + "grad_norm": 2.6825385316497132, + "learning_rate": 4.788561308330489e-07, + "loss": 0.5763, + "step": 8669 + }, + { + "epoch": 0.7818911484871713, + "grad_norm": 2.2435253020173933, + "learning_rate": 4.784768948382272e-07, + "loss": 0.6961, + "step": 8670 + }, + { + "epoch": 0.7819813320106417, + "grad_norm": 1.5067292111983148, + "learning_rate": 4.780977886681461e-07, + "loss": 0.7539, + "step": 8671 + }, + { + "epoch": 0.7820715155341119, + "grad_norm": 1.8595111059814708, + "learning_rate": 4.777188123551541e-07, + "loss": 0.6964, + "step": 8672 + }, + { + "epoch": 0.7821616990575821, + "grad_norm": 2.222375835301608, + "learning_rate": 4.773399659315856e-07, + "loss": 0.7777, + "step": 8673 + }, + { + "epoch": 0.7822518825810525, + "grad_norm": 1.97360955799867, + "learning_rate": 4.769612494297681e-07, + "loss": 0.7285, + "step": 8674 + }, + { + "epoch": 0.7823420661045227, + "grad_norm": 0.6427443037266595, + "learning_rate": 4.765826628820142e-07, + "loss": 0.5549, + "step": 8675 + }, + { + "epoch": 0.782432249627993, + "grad_norm": 1.6707269880487237, + "learning_rate": 4.7620420632062775e-07, + "loss": 0.6595, + "step": 8676 + }, + { + "epoch": 0.7825224331514632, + "grad_norm": 2.1358201056745654, + "learning_rate": 4.758258797779002e-07, + "loss": 0.704, + "step": 8677 + }, + { + "epoch": 0.7826126166749335, + "grad_norm": 0.6167947026239253, + "learning_rate": 4.7544768328611317e-07, + "loss": 0.5651, + "step": 8678 + }, + { + "epoch": 0.7827028001984038, + "grad_norm": 2.013858819788224, + "learning_rate": 4.750696168775359e-07, + "loss": 0.7293, + "step": 8679 + }, + { + "epoch": 0.782792983721874, + "grad_norm": 0.6019759800735239, + "learning_rate": 4.746916805844279e-07, + "loss": 0.5525, + "step": 8680 + }, + { + "epoch": 0.7828831672453442, + "grad_norm": 1.470356488760643, + "learning_rate": 4.743138744390356e-07, + "loss": 0.7749, + "step": 8681 + }, + { + "epoch": 0.7829733507688146, + "grad_norm": 1.6662593184809382, + "learning_rate": 4.739361984735959e-07, + "loss": 0.8235, + "step": 8682 + }, + { + "epoch": 0.7830635342922848, + "grad_norm": 3.0502257091617095, + "learning_rate": 4.7355865272033455e-07, + "loss": 0.7023, + "step": 8683 + }, + { + "epoch": 0.783153717815755, + "grad_norm": 1.6083398100701145, + "learning_rate": 4.7318123721146563e-07, + "loss": 0.7425, + "step": 8684 + }, + { + "epoch": 0.7832439013392253, + "grad_norm": 1.4944542250299402, + "learning_rate": 4.728039519791924e-07, + "loss": 0.7365, + "step": 8685 + }, + { + "epoch": 0.7833340848626956, + "grad_norm": 0.6641565592853083, + "learning_rate": 4.72426797055707e-07, + "loss": 0.5989, + "step": 8686 + }, + { + "epoch": 0.7834242683861659, + "grad_norm": 1.9379615849131269, + "learning_rate": 4.720497724731904e-07, + "loss": 0.7091, + "step": 8687 + }, + { + "epoch": 0.7835144519096361, + "grad_norm": 1.7768988754702562, + "learning_rate": 4.7167287826381153e-07, + "loss": 0.7827, + "step": 8688 + }, + { + "epoch": 0.7836046354331063, + "grad_norm": 1.653217057835971, + "learning_rate": 4.712961144597307e-07, + "loss": 0.7505, + "step": 8689 + }, + { + "epoch": 0.7836948189565767, + "grad_norm": 2.206398770338771, + "learning_rate": 4.7091948109309343e-07, + "loss": 0.7893, + "step": 8690 + }, + { + "epoch": 0.7837850024800469, + "grad_norm": 1.8629370797384905, + "learning_rate": 4.705429781960384e-07, + "loss": 0.7549, + "step": 8691 + }, + { + "epoch": 0.7838751860035171, + "grad_norm": 1.7103713679694328, + "learning_rate": 4.7016660580068923e-07, + "loss": 0.7446, + "step": 8692 + }, + { + "epoch": 0.7839653695269874, + "grad_norm": 1.7416576041513632, + "learning_rate": 4.6979036393916093e-07, + "loss": 0.7762, + "step": 8693 + }, + { + "epoch": 0.7840555530504577, + "grad_norm": 1.6958197612672368, + "learning_rate": 4.6941425264355603e-07, + "loss": 0.8224, + "step": 8694 + }, + { + "epoch": 0.7841457365739279, + "grad_norm": 1.5705971531006546, + "learning_rate": 4.6903827194596666e-07, + "loss": 0.7618, + "step": 8695 + }, + { + "epoch": 0.7842359200973982, + "grad_norm": 2.0630364153963012, + "learning_rate": 4.686624218784743e-07, + "loss": 0.8034, + "step": 8696 + }, + { + "epoch": 0.7843261036208685, + "grad_norm": 2.3013449012462663, + "learning_rate": 4.6828670247314696e-07, + "loss": 0.744, + "step": 8697 + }, + { + "epoch": 0.7844162871443388, + "grad_norm": 2.251411630948507, + "learning_rate": 4.679111137620442e-07, + "loss": 0.761, + "step": 8698 + }, + { + "epoch": 0.784506470667809, + "grad_norm": 1.6418186749924473, + "learning_rate": 4.67535655777213e-07, + "loss": 0.7101, + "step": 8699 + }, + { + "epoch": 0.7845966541912792, + "grad_norm": 1.45953407809236, + "learning_rate": 4.6716032855068956e-07, + "loss": 0.7543, + "step": 8700 + }, + { + "epoch": 0.7846868377147496, + "grad_norm": 2.267701138257335, + "learning_rate": 4.6678513211449867e-07, + "loss": 0.7889, + "step": 8701 + }, + { + "epoch": 0.7847770212382198, + "grad_norm": 1.7508033543373973, + "learning_rate": 4.6641006650065516e-07, + "loss": 0.7184, + "step": 8702 + }, + { + "epoch": 0.78486720476169, + "grad_norm": 3.667131712350613, + "learning_rate": 4.6603513174115973e-07, + "loss": 0.7331, + "step": 8703 + }, + { + "epoch": 0.7849573882851603, + "grad_norm": 2.7590319825296223, + "learning_rate": 4.6566032786800625e-07, + "loss": 0.7628, + "step": 8704 + }, + { + "epoch": 0.7850475718086306, + "grad_norm": 1.607151532528184, + "learning_rate": 4.6528565491317274e-07, + "loss": 0.7294, + "step": 8705 + }, + { + "epoch": 0.7851377553321008, + "grad_norm": 2.0628178669444033, + "learning_rate": 4.649111129086305e-07, + "loss": 0.8147, + "step": 8706 + }, + { + "epoch": 0.7852279388555711, + "grad_norm": 1.6962523777607028, + "learning_rate": 4.6453670188633596e-07, + "loss": 0.7754, + "step": 8707 + }, + { + "epoch": 0.7853181223790413, + "grad_norm": 2.0568919376961383, + "learning_rate": 4.641624218782365e-07, + "loss": 0.7594, + "step": 8708 + }, + { + "epoch": 0.7854083059025116, + "grad_norm": 1.4479670710217922, + "learning_rate": 4.6378827291626765e-07, + "loss": 0.7167, + "step": 8709 + }, + { + "epoch": 0.7854984894259819, + "grad_norm": 1.3543231531349416, + "learning_rate": 4.634142550323541e-07, + "loss": 0.8013, + "step": 8710 + }, + { + "epoch": 0.7855886729494521, + "grad_norm": 3.334000536583924, + "learning_rate": 4.6304036825840943e-07, + "loss": 0.7974, + "step": 8711 + }, + { + "epoch": 0.7856788564729223, + "grad_norm": 1.4068523812223386, + "learning_rate": 4.626666126263341e-07, + "loss": 0.6527, + "step": 8712 + }, + { + "epoch": 0.7857690399963927, + "grad_norm": 1.6887386156174204, + "learning_rate": 4.622929881680213e-07, + "loss": 0.772, + "step": 8713 + }, + { + "epoch": 0.7858592235198629, + "grad_norm": 4.8401146813669795, + "learning_rate": 4.6191949491534887e-07, + "loss": 0.6652, + "step": 8714 + }, + { + "epoch": 0.7859494070433332, + "grad_norm": 1.8306839468423592, + "learning_rate": 4.6154613290018617e-07, + "loss": 0.7582, + "step": 8715 + }, + { + "epoch": 0.7860395905668034, + "grad_norm": 1.6520487204057017, + "learning_rate": 4.6117290215439043e-07, + "loss": 0.6977, + "step": 8716 + }, + { + "epoch": 0.7861297740902737, + "grad_norm": 2.4218325703503565, + "learning_rate": 4.6079980270980744e-07, + "loss": 0.6118, + "step": 8717 + }, + { + "epoch": 0.786219957613744, + "grad_norm": 3.374476922084468, + "learning_rate": 4.6042683459827245e-07, + "loss": 0.7238, + "step": 8718 + }, + { + "epoch": 0.7863101411372142, + "grad_norm": 0.7037134674983901, + "learning_rate": 4.600539978516098e-07, + "loss": 0.5635, + "step": 8719 + }, + { + "epoch": 0.7864003246606845, + "grad_norm": 1.7941858486503204, + "learning_rate": 4.5968129250163004e-07, + "loss": 0.7093, + "step": 8720 + }, + { + "epoch": 0.7864905081841548, + "grad_norm": 1.6791904867421625, + "learning_rate": 4.5930871858013653e-07, + "loss": 0.6815, + "step": 8721 + }, + { + "epoch": 0.786580691707625, + "grad_norm": 0.65441019016437, + "learning_rate": 4.589362761189182e-07, + "loss": 0.6212, + "step": 8722 + }, + { + "epoch": 0.7866708752310952, + "grad_norm": 1.6325432222261447, + "learning_rate": 4.585639651497539e-07, + "loss": 0.8151, + "step": 8723 + }, + { + "epoch": 0.7867610587545656, + "grad_norm": 1.5040256997763597, + "learning_rate": 4.581917857044115e-07, + "loss": 0.702, + "step": 8724 + }, + { + "epoch": 0.7868512422780358, + "grad_norm": 2.0651760658040215, + "learning_rate": 4.5781973781464734e-07, + "loss": 0.7831, + "step": 8725 + }, + { + "epoch": 0.7869414258015061, + "grad_norm": 2.2499690583538707, + "learning_rate": 4.574478215122073e-07, + "loss": 0.7871, + "step": 8726 + }, + { + "epoch": 0.7870316093249763, + "grad_norm": 2.553606463715014, + "learning_rate": 4.5707603682882357e-07, + "loss": 0.7284, + "step": 8727 + }, + { + "epoch": 0.7871217928484466, + "grad_norm": 3.885329967414412, + "learning_rate": 4.56704383796221e-07, + "loss": 0.6149, + "step": 8728 + }, + { + "epoch": 0.7872119763719169, + "grad_norm": 1.9967231728588146, + "learning_rate": 4.5633286244610956e-07, + "loss": 0.7868, + "step": 8729 + }, + { + "epoch": 0.7873021598953871, + "grad_norm": 1.4236882982289052, + "learning_rate": 4.5596147281018993e-07, + "loss": 0.813, + "step": 8730 + }, + { + "epoch": 0.7873923434188573, + "grad_norm": 1.8112112264649196, + "learning_rate": 4.5559021492015137e-07, + "loss": 0.8532, + "step": 8731 + }, + { + "epoch": 0.7874825269423277, + "grad_norm": 1.4191571764685604, + "learning_rate": 4.552190888076712e-07, + "loss": 0.6567, + "step": 8732 + }, + { + "epoch": 0.7875727104657979, + "grad_norm": 1.5848677461483567, + "learning_rate": 4.548480945044164e-07, + "loss": 0.7003, + "step": 8733 + }, + { + "epoch": 0.7876628939892681, + "grad_norm": 1.569265747399692, + "learning_rate": 4.54477232042042e-07, + "loss": 0.7448, + "step": 8734 + }, + { + "epoch": 0.7877530775127384, + "grad_norm": 1.5297133851710536, + "learning_rate": 4.541065014521921e-07, + "loss": 0.7457, + "step": 8735 + }, + { + "epoch": 0.7878432610362087, + "grad_norm": 3.6656992821024805, + "learning_rate": 4.5373590276649996e-07, + "loss": 0.6978, + "step": 8736 + }, + { + "epoch": 0.787933444559679, + "grad_norm": 1.390868647801113, + "learning_rate": 4.533654360165862e-07, + "loss": 0.7171, + "step": 8737 + }, + { + "epoch": 0.7880236280831492, + "grad_norm": 1.805025605736016, + "learning_rate": 4.5299510123406115e-07, + "loss": 0.8025, + "step": 8738 + }, + { + "epoch": 0.7881138116066194, + "grad_norm": 1.450169007547624, + "learning_rate": 4.5262489845052456e-07, + "loss": 0.7179, + "step": 8739 + }, + { + "epoch": 0.7882039951300898, + "grad_norm": 1.6259025121557638, + "learning_rate": 4.5225482769756353e-07, + "loss": 0.7136, + "step": 8740 + }, + { + "epoch": 0.78829417865356, + "grad_norm": 2.3519914305589227, + "learning_rate": 4.5188488900675545e-07, + "loss": 0.6775, + "step": 8741 + }, + { + "epoch": 0.7883843621770302, + "grad_norm": 1.9646056620629158, + "learning_rate": 4.5151508240966363e-07, + "loss": 0.7798, + "step": 8742 + }, + { + "epoch": 0.7884745457005006, + "grad_norm": 1.6325495019858742, + "learning_rate": 4.511454079378445e-07, + "loss": 0.7912, + "step": 8743 + }, + { + "epoch": 0.7885647292239708, + "grad_norm": 1.9879546431197057, + "learning_rate": 4.507758656228382e-07, + "loss": 0.6555, + "step": 8744 + }, + { + "epoch": 0.788654912747441, + "grad_norm": 2.7329110722679144, + "learning_rate": 4.5040645549617864e-07, + "loss": 0.8109, + "step": 8745 + }, + { + "epoch": 0.7887450962709113, + "grad_norm": 1.5338301427370629, + "learning_rate": 4.5003717758938384e-07, + "loss": 0.6638, + "step": 8746 + }, + { + "epoch": 0.7888352797943816, + "grad_norm": 1.6273657377445199, + "learning_rate": 4.4966803193396365e-07, + "loss": 0.6876, + "step": 8747 + }, + { + "epoch": 0.7889254633178518, + "grad_norm": 2.0706343490615207, + "learning_rate": 4.492990185614154e-07, + "loss": 0.7546, + "step": 8748 + }, + { + "epoch": 0.7890156468413221, + "grad_norm": 1.512680733341421, + "learning_rate": 4.489301375032255e-07, + "loss": 0.7387, + "step": 8749 + }, + { + "epoch": 0.7891058303647923, + "grad_norm": 2.4286331882157155, + "learning_rate": 4.4856138879086857e-07, + "loss": 0.7535, + "step": 8750 + }, + { + "epoch": 0.7891960138882627, + "grad_norm": 3.0359133999702856, + "learning_rate": 4.481927724558092e-07, + "loss": 0.6964, + "step": 8751 + }, + { + "epoch": 0.7892861974117329, + "grad_norm": 1.5672713861045338, + "learning_rate": 4.478242885294985e-07, + "loss": 0.7165, + "step": 8752 + }, + { + "epoch": 0.7893763809352031, + "grad_norm": 2.5092940187985784, + "learning_rate": 4.474559370433779e-07, + "loss": 0.7096, + "step": 8753 + }, + { + "epoch": 0.7894665644586734, + "grad_norm": 3.0898743060080474, + "learning_rate": 4.470877180288777e-07, + "loss": 0.6935, + "step": 8754 + }, + { + "epoch": 0.7895567479821437, + "grad_norm": 3.7923101920079305, + "learning_rate": 4.4671963151741574e-07, + "loss": 0.7164, + "step": 8755 + }, + { + "epoch": 0.7896469315056139, + "grad_norm": 1.420902464556149, + "learning_rate": 4.4635167754039973e-07, + "loss": 0.7231, + "step": 8756 + }, + { + "epoch": 0.7897371150290842, + "grad_norm": 1.3793659254500688, + "learning_rate": 4.459838561292253e-07, + "loss": 0.7699, + "step": 8757 + }, + { + "epoch": 0.7898272985525544, + "grad_norm": 2.8192193192509594, + "learning_rate": 4.456161673152774e-07, + "loss": 0.6668, + "step": 8758 + }, + { + "epoch": 0.7899174820760247, + "grad_norm": 1.6125394210136756, + "learning_rate": 4.4524861112992806e-07, + "loss": 0.7358, + "step": 8759 + }, + { + "epoch": 0.790007665599495, + "grad_norm": 1.789038445590031, + "learning_rate": 4.448811876045411e-07, + "loss": 0.7431, + "step": 8760 + }, + { + "epoch": 0.7900978491229652, + "grad_norm": 2.7182552720615805, + "learning_rate": 4.445138967704647e-07, + "loss": 0.8156, + "step": 8761 + }, + { + "epoch": 0.7901880326464354, + "grad_norm": 6.026146503438844, + "learning_rate": 4.4414673865904075e-07, + "loss": 0.7194, + "step": 8762 + }, + { + "epoch": 0.7902782161699058, + "grad_norm": 1.4684697147528656, + "learning_rate": 4.437797133015955e-07, + "loss": 0.7609, + "step": 8763 + }, + { + "epoch": 0.790368399693376, + "grad_norm": 1.7909199097328146, + "learning_rate": 4.4341282072944586e-07, + "loss": 0.6673, + "step": 8764 + }, + { + "epoch": 0.7904585832168463, + "grad_norm": 2.346739528113693, + "learning_rate": 4.430460609738973e-07, + "loss": 0.7675, + "step": 8765 + }, + { + "epoch": 0.7905487667403166, + "grad_norm": 0.7128509928516924, + "learning_rate": 4.4267943406624386e-07, + "loss": 0.6098, + "step": 8766 + }, + { + "epoch": 0.7906389502637868, + "grad_norm": 1.855316862615539, + "learning_rate": 4.4231294003776853e-07, + "loss": 0.8565, + "step": 8767 + }, + { + "epoch": 0.7907291337872571, + "grad_norm": 3.1390032020691057, + "learning_rate": 4.419465789197416e-07, + "loss": 0.7873, + "step": 8768 + }, + { + "epoch": 0.7908193173107273, + "grad_norm": 1.646104943120167, + "learning_rate": 4.415803507434237e-07, + "loss": 0.7377, + "step": 8769 + }, + { + "epoch": 0.7909095008341976, + "grad_norm": 1.7233681797117577, + "learning_rate": 4.4121425554006307e-07, + "loss": 0.7651, + "step": 8770 + }, + { + "epoch": 0.7909996843576679, + "grad_norm": 1.9149154319715487, + "learning_rate": 4.4084829334089744e-07, + "loss": 0.7306, + "step": 8771 + }, + { + "epoch": 0.7910898678811381, + "grad_norm": 1.5313839756490315, + "learning_rate": 4.404824641771525e-07, + "loss": 0.6858, + "step": 8772 + }, + { + "epoch": 0.7911800514046083, + "grad_norm": 1.5495771446559188, + "learning_rate": 4.4011676808004327e-07, + "loss": 0.8125, + "step": 8773 + }, + { + "epoch": 0.7912702349280787, + "grad_norm": 6.832809831368819, + "learning_rate": 4.3975120508077145e-07, + "loss": 0.7547, + "step": 8774 + }, + { + "epoch": 0.7913604184515489, + "grad_norm": 1.6595449242092695, + "learning_rate": 4.39385775210531e-07, + "loss": 0.7929, + "step": 8775 + }, + { + "epoch": 0.7914506019750192, + "grad_norm": 1.8408459568008189, + "learning_rate": 4.390204785005003e-07, + "loss": 0.7625, + "step": 8776 + }, + { + "epoch": 0.7915407854984894, + "grad_norm": 1.4297815510816596, + "learning_rate": 4.386553149818504e-07, + "loss": 0.7457, + "step": 8777 + }, + { + "epoch": 0.7916309690219597, + "grad_norm": 2.4428207816219896, + "learning_rate": 4.3829028468573793e-07, + "loss": 0.6125, + "step": 8778 + }, + { + "epoch": 0.79172115254543, + "grad_norm": 2.140610186671772, + "learning_rate": 4.3792538764330935e-07, + "loss": 0.6843, + "step": 8779 + }, + { + "epoch": 0.7918113360689002, + "grad_norm": 1.802626121965576, + "learning_rate": 4.3756062388569994e-07, + "loss": 0.7228, + "step": 8780 + }, + { + "epoch": 0.7919015195923704, + "grad_norm": 0.7314155684612795, + "learning_rate": 4.3719599344403346e-07, + "loss": 0.5802, + "step": 8781 + }, + { + "epoch": 0.7919917031158408, + "grad_norm": 1.4109256156682795, + "learning_rate": 4.3683149634942243e-07, + "loss": 0.7262, + "step": 8782 + }, + { + "epoch": 0.792081886639311, + "grad_norm": 6.376945049286617, + "learning_rate": 4.364671326329663e-07, + "loss": 0.6254, + "step": 8783 + }, + { + "epoch": 0.7921720701627812, + "grad_norm": 1.948857520503587, + "learning_rate": 4.3610290232575673e-07, + "loss": 0.7642, + "step": 8784 + }, + { + "epoch": 0.7922622536862515, + "grad_norm": 2.313529738994174, + "learning_rate": 4.357388054588702e-07, + "loss": 0.7498, + "step": 8785 + }, + { + "epoch": 0.7923524372097218, + "grad_norm": 1.646465912304737, + "learning_rate": 4.3537484206337405e-07, + "loss": 0.6891, + "step": 8786 + }, + { + "epoch": 0.792442620733192, + "grad_norm": 1.7617854002125166, + "learning_rate": 4.3501101217032366e-07, + "loss": 0.776, + "step": 8787 + }, + { + "epoch": 0.7925328042566623, + "grad_norm": 1.8964357222534372, + "learning_rate": 4.346473158107629e-07, + "loss": 0.6758, + "step": 8788 + }, + { + "epoch": 0.7926229877801325, + "grad_norm": 2.0499405363809347, + "learning_rate": 4.342837530157244e-07, + "loss": 0.7494, + "step": 8789 + }, + { + "epoch": 0.7927131713036029, + "grad_norm": 1.721467107240872, + "learning_rate": 4.3392032381622987e-07, + "loss": 0.6465, + "step": 8790 + }, + { + "epoch": 0.7928033548270731, + "grad_norm": 1.4297442815792683, + "learning_rate": 4.3355702824328765e-07, + "loss": 0.7058, + "step": 8791 + }, + { + "epoch": 0.7928935383505433, + "grad_norm": 1.4774469219555892, + "learning_rate": 4.3319386632789823e-07, + "loss": 0.7877, + "step": 8792 + }, + { + "epoch": 0.7929837218740137, + "grad_norm": 1.8220514333104492, + "learning_rate": 4.328308381010466e-07, + "loss": 0.74, + "step": 8793 + }, + { + "epoch": 0.7930739053974839, + "grad_norm": 1.5606955217553478, + "learning_rate": 4.3246794359370933e-07, + "loss": 0.7511, + "step": 8794 + }, + { + "epoch": 0.7931640889209541, + "grad_norm": 2.6015844029143205, + "learning_rate": 4.3210518283685025e-07, + "loss": 0.7231, + "step": 8795 + }, + { + "epoch": 0.7932542724444244, + "grad_norm": 6.2020665570144216, + "learning_rate": 4.317425558614225e-07, + "loss": 0.7406, + "step": 8796 + }, + { + "epoch": 0.7933444559678947, + "grad_norm": 1.3523699162341207, + "learning_rate": 4.3138006269836744e-07, + "loss": 0.7076, + "step": 8797 + }, + { + "epoch": 0.793434639491365, + "grad_norm": 1.9458135243073946, + "learning_rate": 4.3101770337861376e-07, + "loss": 0.7326, + "step": 8798 + }, + { + "epoch": 0.7935248230148352, + "grad_norm": 0.6723476122745871, + "learning_rate": 4.30655477933082e-07, + "loss": 0.5942, + "step": 8799 + }, + { + "epoch": 0.7936150065383054, + "grad_norm": 1.571481829211406, + "learning_rate": 4.30293386392677e-07, + "loss": 0.7695, + "step": 8800 + }, + { + "epoch": 0.7937051900617758, + "grad_norm": 1.52425014633621, + "learning_rate": 4.299314287882967e-07, + "loss": 0.7629, + "step": 8801 + }, + { + "epoch": 0.793795373585246, + "grad_norm": 2.0318252922555895, + "learning_rate": 4.2956960515082353e-07, + "loss": 0.7776, + "step": 8802 + }, + { + "epoch": 0.7938855571087162, + "grad_norm": 1.5292246132801872, + "learning_rate": 4.29207915511131e-07, + "loss": 0.7032, + "step": 8803 + }, + { + "epoch": 0.7939757406321865, + "grad_norm": 1.8953173078332917, + "learning_rate": 4.2884635990008024e-07, + "loss": 0.804, + "step": 8804 + }, + { + "epoch": 0.7940659241556568, + "grad_norm": 2.201612453828009, + "learning_rate": 4.284849383485214e-07, + "loss": 0.7113, + "step": 8805 + }, + { + "epoch": 0.794156107679127, + "grad_norm": 2.003572967952237, + "learning_rate": 4.2812365088729296e-07, + "loss": 0.7658, + "step": 8806 + }, + { + "epoch": 0.7942462912025973, + "grad_norm": 1.5508236050526982, + "learning_rate": 4.2776249754722227e-07, + "loss": 0.7633, + "step": 8807 + }, + { + "epoch": 0.7943364747260675, + "grad_norm": 2.1510445607853246, + "learning_rate": 4.27401478359124e-07, + "loss": 0.8139, + "step": 8808 + }, + { + "epoch": 0.7944266582495378, + "grad_norm": 3.11102680440236, + "learning_rate": 4.2704059335380283e-07, + "loss": 0.8463, + "step": 8809 + }, + { + "epoch": 0.7945168417730081, + "grad_norm": 1.470724604679009, + "learning_rate": 4.266798425620515e-07, + "loss": 0.8469, + "step": 8810 + }, + { + "epoch": 0.7946070252964783, + "grad_norm": 1.5296689649010327, + "learning_rate": 4.263192260146511e-07, + "loss": 0.7331, + "step": 8811 + }, + { + "epoch": 0.7946972088199485, + "grad_norm": 3.340599786504476, + "learning_rate": 4.2595874374237216e-07, + "loss": 0.7943, + "step": 8812 + }, + { + "epoch": 0.7947873923434189, + "grad_norm": 1.3986158204002022, + "learning_rate": 4.255983957759712e-07, + "loss": 0.7419, + "step": 8813 + }, + { + "epoch": 0.7948775758668891, + "grad_norm": 1.3619692372336376, + "learning_rate": 4.2523818214619745e-07, + "loss": 0.6836, + "step": 8814 + }, + { + "epoch": 0.7949677593903594, + "grad_norm": 1.6091140239183876, + "learning_rate": 4.24878102883784e-07, + "loss": 0.7927, + "step": 8815 + }, + { + "epoch": 0.7950579429138297, + "grad_norm": 1.5094521259955902, + "learning_rate": 4.24518158019457e-07, + "loss": 0.7513, + "step": 8816 + }, + { + "epoch": 0.7951481264372999, + "grad_norm": 1.6231627715559176, + "learning_rate": 4.241583475839274e-07, + "loss": 0.6863, + "step": 8817 + }, + { + "epoch": 0.7952383099607702, + "grad_norm": 1.9094543415573113, + "learning_rate": 4.237986716078965e-07, + "loss": 0.791, + "step": 8818 + }, + { + "epoch": 0.7953284934842404, + "grad_norm": 3.044374192797108, + "learning_rate": 4.2343913012205433e-07, + "loss": 0.7309, + "step": 8819 + }, + { + "epoch": 0.7954186770077107, + "grad_norm": 1.7435103385056465, + "learning_rate": 4.230797231570784e-07, + "loss": 0.721, + "step": 8820 + }, + { + "epoch": 0.795508860531181, + "grad_norm": 2.326648013232347, + "learning_rate": 4.227204507436357e-07, + "loss": 0.7504, + "step": 8821 + }, + { + "epoch": 0.7955990440546512, + "grad_norm": 1.41331757371741, + "learning_rate": 4.223613129123811e-07, + "loss": 0.6635, + "step": 8822 + }, + { + "epoch": 0.7956892275781214, + "grad_norm": 1.6788418082485952, + "learning_rate": 4.220023096939589e-07, + "loss": 0.6526, + "step": 8823 + }, + { + "epoch": 0.7957794111015918, + "grad_norm": 1.8070822675984888, + "learning_rate": 4.21643441119e-07, + "loss": 0.7392, + "step": 8824 + }, + { + "epoch": 0.795869594625062, + "grad_norm": 0.6269965231067438, + "learning_rate": 4.212847072181256e-07, + "loss": 0.6301, + "step": 8825 + }, + { + "epoch": 0.7959597781485322, + "grad_norm": 1.545250395943431, + "learning_rate": 4.2092610802194505e-07, + "loss": 0.7128, + "step": 8826 + }, + { + "epoch": 0.7960499616720025, + "grad_norm": 1.5020365400428517, + "learning_rate": 4.2056764356105587e-07, + "loss": 0.6588, + "step": 8827 + }, + { + "epoch": 0.7961401451954728, + "grad_norm": 1.71620371078892, + "learning_rate": 4.202093138660443e-07, + "loss": 0.7898, + "step": 8828 + }, + { + "epoch": 0.7962303287189431, + "grad_norm": 6.202699737961418, + "learning_rate": 4.198511189674854e-07, + "loss": 0.779, + "step": 8829 + }, + { + "epoch": 0.7963205122424133, + "grad_norm": 1.5860459614764932, + "learning_rate": 4.1949305889594066e-07, + "loss": 0.7595, + "step": 8830 + }, + { + "epoch": 0.7964106957658835, + "grad_norm": 1.5794253138327585, + "learning_rate": 4.191351336819642e-07, + "loss": 0.7366, + "step": 8831 + }, + { + "epoch": 0.7965008792893539, + "grad_norm": 1.4846010537936347, + "learning_rate": 4.187773433560939e-07, + "loss": 0.6853, + "step": 8832 + }, + { + "epoch": 0.7965910628128241, + "grad_norm": 2.5594585325887786, + "learning_rate": 4.184196879488604e-07, + "loss": 0.7445, + "step": 8833 + }, + { + "epoch": 0.7966812463362943, + "grad_norm": 2.553715698937961, + "learning_rate": 4.1806216749077936e-07, + "loss": 0.8077, + "step": 8834 + }, + { + "epoch": 0.7967714298597646, + "grad_norm": 1.7371747213278963, + "learning_rate": 4.177047820123569e-07, + "loss": 0.7879, + "step": 8835 + }, + { + "epoch": 0.7968616133832349, + "grad_norm": 1.6586656767764516, + "learning_rate": 4.1734753154408733e-07, + "loss": 0.6174, + "step": 8836 + }, + { + "epoch": 0.7969517969067051, + "grad_norm": 4.5169848866052575, + "learning_rate": 4.169904161164528e-07, + "loss": 0.7271, + "step": 8837 + }, + { + "epoch": 0.7970419804301754, + "grad_norm": 2.126151614326521, + "learning_rate": 4.1663343575992526e-07, + "loss": 0.8584, + "step": 8838 + }, + { + "epoch": 0.7971321639536457, + "grad_norm": 2.154701008246045, + "learning_rate": 4.1627659050496275e-07, + "loss": 0.7617, + "step": 8839 + }, + { + "epoch": 0.797222347477116, + "grad_norm": 1.5509615776472785, + "learning_rate": 4.1591988038201453e-07, + "loss": 0.6555, + "step": 8840 + }, + { + "epoch": 0.7973125310005862, + "grad_norm": 3.052562550146777, + "learning_rate": 4.155633054215164e-07, + "loss": 0.7902, + "step": 8841 + }, + { + "epoch": 0.7974027145240564, + "grad_norm": 2.157114491416056, + "learning_rate": 4.152068656538934e-07, + "loss": 0.6527, + "step": 8842 + }, + { + "epoch": 0.7974928980475268, + "grad_norm": 1.4670982000537913, + "learning_rate": 4.148505611095594e-07, + "loss": 0.8034, + "step": 8843 + }, + { + "epoch": 0.797583081570997, + "grad_norm": 1.940381030781506, + "learning_rate": 4.1449439181891563e-07, + "loss": 0.6251, + "step": 8844 + }, + { + "epoch": 0.7976732650944672, + "grad_norm": 1.618502022738131, + "learning_rate": 4.14138357812353e-07, + "loss": 0.7626, + "step": 8845 + }, + { + "epoch": 0.7977634486179375, + "grad_norm": 1.3760317486105487, + "learning_rate": 4.137824591202506e-07, + "loss": 0.7968, + "step": 8846 + }, + { + "epoch": 0.7978536321414078, + "grad_norm": 4.245337902410084, + "learning_rate": 4.134266957729737e-07, + "loss": 0.7311, + "step": 8847 + }, + { + "epoch": 0.797943815664878, + "grad_norm": 1.7015971058530814, + "learning_rate": 4.1307106780088065e-07, + "loss": 0.8345, + "step": 8848 + }, + { + "epoch": 0.7980339991883483, + "grad_norm": 1.4704381291131166, + "learning_rate": 4.1271557523431387e-07, + "loss": 0.7558, + "step": 8849 + }, + { + "epoch": 0.7981241827118185, + "grad_norm": 1.5884616102047024, + "learning_rate": 4.1236021810360634e-07, + "loss": 0.7314, + "step": 8850 + }, + { + "epoch": 0.7982143662352889, + "grad_norm": 2.4278619415084433, + "learning_rate": 4.120049964390793e-07, + "loss": 0.6903, + "step": 8851 + }, + { + "epoch": 0.7983045497587591, + "grad_norm": 1.7592939449631309, + "learning_rate": 4.116499102710418e-07, + "loss": 0.6937, + "step": 8852 + }, + { + "epoch": 0.7983947332822293, + "grad_norm": 1.9978111926080038, + "learning_rate": 4.112949596297928e-07, + "loss": 0.7174, + "step": 8853 + }, + { + "epoch": 0.7984849168056996, + "grad_norm": 1.7290178334478616, + "learning_rate": 4.1094014454561664e-07, + "loss": 0.6866, + "step": 8854 + }, + { + "epoch": 0.7985751003291699, + "grad_norm": 1.9191368880993538, + "learning_rate": 4.1058546504879057e-07, + "loss": 0.6485, + "step": 8855 + }, + { + "epoch": 0.7986652838526401, + "grad_norm": 1.983406070636315, + "learning_rate": 4.1023092116957583e-07, + "loss": 0.8076, + "step": 8856 + }, + { + "epoch": 0.7987554673761104, + "grad_norm": 1.509229083540741, + "learning_rate": 4.098765129382249e-07, + "loss": 0.7851, + "step": 8857 + }, + { + "epoch": 0.7988456508995806, + "grad_norm": 1.8579132360766195, + "learning_rate": 4.0952224038497764e-07, + "loss": 0.816, + "step": 8858 + }, + { + "epoch": 0.7989358344230509, + "grad_norm": 2.735131731225793, + "learning_rate": 4.091681035400627e-07, + "loss": 0.7417, + "step": 8859 + }, + { + "epoch": 0.7990260179465212, + "grad_norm": 1.750194402523547, + "learning_rate": 4.088141024336971e-07, + "loss": 0.7787, + "step": 8860 + }, + { + "epoch": 0.7991162014699914, + "grad_norm": 1.4078011329183506, + "learning_rate": 4.0846023709608636e-07, + "loss": 0.8218, + "step": 8861 + }, + { + "epoch": 0.7992063849934617, + "grad_norm": 2.374541288549047, + "learning_rate": 4.081065075574226e-07, + "loss": 0.8208, + "step": 8862 + }, + { + "epoch": 0.799296568516932, + "grad_norm": 2.1012672869296294, + "learning_rate": 4.077529138478906e-07, + "loss": 0.6755, + "step": 8863 + }, + { + "epoch": 0.7993867520404022, + "grad_norm": 0.6806728849564423, + "learning_rate": 4.073994559976588e-07, + "loss": 0.5417, + "step": 8864 + }, + { + "epoch": 0.7994769355638724, + "grad_norm": 2.2745706677111857, + "learning_rate": 4.0704613403688716e-07, + "loss": 0.708, + "step": 8865 + }, + { + "epoch": 0.7995671190873428, + "grad_norm": 2.2803544677452394, + "learning_rate": 4.0669294799572264e-07, + "loss": 0.8015, + "step": 8866 + }, + { + "epoch": 0.799657302610813, + "grad_norm": 5.023558619627523, + "learning_rate": 4.0633989790430113e-07, + "loss": 0.8472, + "step": 8867 + }, + { + "epoch": 0.7997474861342833, + "grad_norm": 2.035041791219654, + "learning_rate": 4.059869837927477e-07, + "loss": 0.7564, + "step": 8868 + }, + { + "epoch": 0.7998376696577535, + "grad_norm": 2.5705711907065476, + "learning_rate": 4.056342056911728e-07, + "loss": 0.7299, + "step": 8869 + }, + { + "epoch": 0.7999278531812238, + "grad_norm": 1.767597413352978, + "learning_rate": 4.052815636296798e-07, + "loss": 0.7624, + "step": 8870 + }, + { + "epoch": 0.8000180367046941, + "grad_norm": 1.4076123314407574, + "learning_rate": 4.0492905763835593e-07, + "loss": 0.7776, + "step": 8871 + }, + { + "epoch": 0.8001082202281643, + "grad_norm": 1.5564545217426604, + "learning_rate": 4.0457668774728115e-07, + "loss": 0.772, + "step": 8872 + }, + { + "epoch": 0.8001984037516345, + "grad_norm": 1.4495496346435635, + "learning_rate": 4.0422445398651985e-07, + "loss": 0.7606, + "step": 8873 + }, + { + "epoch": 0.8002885872751049, + "grad_norm": 1.640557278643229, + "learning_rate": 4.0387235638612706e-07, + "loss": 0.6695, + "step": 8874 + }, + { + "epoch": 0.8003787707985751, + "grad_norm": 2.000593931701115, + "learning_rate": 4.0352039497614586e-07, + "loss": 0.7503, + "step": 8875 + }, + { + "epoch": 0.8004689543220453, + "grad_norm": 2.563964611279389, + "learning_rate": 4.031685697866074e-07, + "loss": 0.7337, + "step": 8876 + }, + { + "epoch": 0.8005591378455156, + "grad_norm": 1.6230445246673717, + "learning_rate": 4.0281688084753165e-07, + "loss": 0.7837, + "step": 8877 + }, + { + "epoch": 0.8006493213689859, + "grad_norm": 1.682972521244237, + "learning_rate": 4.0246532818892675e-07, + "loss": 0.799, + "step": 8878 + }, + { + "epoch": 0.8007395048924562, + "grad_norm": 2.2613311066429103, + "learning_rate": 4.0211391184078814e-07, + "loss": 0.7845, + "step": 8879 + }, + { + "epoch": 0.8008296884159264, + "grad_norm": 1.7441134221090406, + "learning_rate": 4.0176263183310135e-07, + "loss": 0.7819, + "step": 8880 + }, + { + "epoch": 0.8009198719393966, + "grad_norm": 1.798081299930645, + "learning_rate": 4.0141148819583925e-07, + "loss": 0.8055, + "step": 8881 + }, + { + "epoch": 0.801010055462867, + "grad_norm": 0.5984480356429442, + "learning_rate": 4.010604809589637e-07, + "loss": 0.5702, + "step": 8882 + }, + { + "epoch": 0.8011002389863372, + "grad_norm": 1.6631776131468716, + "learning_rate": 4.0070961015242475e-07, + "loss": 0.7486, + "step": 8883 + }, + { + "epoch": 0.8011904225098074, + "grad_norm": 1.5527723775567672, + "learning_rate": 4.0035887580615933e-07, + "loss": 0.8174, + "step": 8884 + }, + { + "epoch": 0.8012806060332777, + "grad_norm": 2.021771662103939, + "learning_rate": 4.0000827795009594e-07, + "loss": 0.7979, + "step": 8885 + }, + { + "epoch": 0.801370789556748, + "grad_norm": 3.473187743837478, + "learning_rate": 3.996578166141475e-07, + "loss": 0.7111, + "step": 8886 + }, + { + "epoch": 0.8014609730802182, + "grad_norm": 10.55436040793593, + "learning_rate": 3.9930749182821955e-07, + "loss": 0.751, + "step": 8887 + }, + { + "epoch": 0.8015511566036885, + "grad_norm": 0.6039270276520986, + "learning_rate": 3.9895730362220116e-07, + "loss": 0.5948, + "step": 8888 + }, + { + "epoch": 0.8016413401271588, + "grad_norm": 2.2497100643272137, + "learning_rate": 3.986072520259749e-07, + "loss": 0.7849, + "step": 8889 + }, + { + "epoch": 0.801731523650629, + "grad_norm": 1.6158192324141953, + "learning_rate": 3.9825733706940736e-07, + "loss": 0.7385, + "step": 8890 + }, + { + "epoch": 0.8018217071740993, + "grad_norm": 1.4623749540945081, + "learning_rate": 3.979075587823557e-07, + "loss": 0.7658, + "step": 8891 + }, + { + "epoch": 0.8019118906975695, + "grad_norm": 2.4073903738000686, + "learning_rate": 3.9755791719466504e-07, + "loss": 0.6279, + "step": 8892 + }, + { + "epoch": 0.8020020742210399, + "grad_norm": 1.913463327475568, + "learning_rate": 3.9720841233616875e-07, + "loss": 0.7107, + "step": 8893 + }, + { + "epoch": 0.8020922577445101, + "grad_norm": 1.622035770863784, + "learning_rate": 3.968590442366888e-07, + "loss": 0.7181, + "step": 8894 + }, + { + "epoch": 0.8021824412679803, + "grad_norm": 1.5350992517348845, + "learning_rate": 3.9650981292603423e-07, + "loss": 0.7245, + "step": 8895 + }, + { + "epoch": 0.8022726247914506, + "grad_norm": 1.919203289056355, + "learning_rate": 3.961607184340041e-07, + "loss": 0.7283, + "step": 8896 + }, + { + "epoch": 0.8023628083149209, + "grad_norm": 1.3457471067995241, + "learning_rate": 3.9581176079038505e-07, + "loss": 0.6432, + "step": 8897 + }, + { + "epoch": 0.8024529918383911, + "grad_norm": 1.4481085871925845, + "learning_rate": 3.954629400249516e-07, + "loss": 0.6941, + "step": 8898 + }, + { + "epoch": 0.8025431753618614, + "grad_norm": 1.9065897279279196, + "learning_rate": 3.9511425616746787e-07, + "loss": 0.8131, + "step": 8899 + }, + { + "epoch": 0.8026333588853316, + "grad_norm": 2.4165453935207863, + "learning_rate": 3.947657092476853e-07, + "loss": 0.8099, + "step": 8900 + }, + { + "epoch": 0.802723542408802, + "grad_norm": 1.8668096950208144, + "learning_rate": 3.944172992953425e-07, + "loss": 0.759, + "step": 8901 + }, + { + "epoch": 0.8028137259322722, + "grad_norm": 1.601405140659299, + "learning_rate": 3.9406902634017e-07, + "loss": 0.7018, + "step": 8902 + }, + { + "epoch": 0.8029039094557424, + "grad_norm": 1.482193076026525, + "learning_rate": 3.9372089041188227e-07, + "loss": 0.7816, + "step": 8903 + }, + { + "epoch": 0.8029940929792126, + "grad_norm": 4.95356608932825, + "learning_rate": 3.9337289154018593e-07, + "loss": 0.6648, + "step": 8904 + }, + { + "epoch": 0.803084276502683, + "grad_norm": 1.5079854757282394, + "learning_rate": 3.930250297547728e-07, + "loss": 0.747, + "step": 8905 + }, + { + "epoch": 0.8031744600261532, + "grad_norm": 2.5062563812859553, + "learning_rate": 3.9267730508532513e-07, + "loss": 0.6297, + "step": 8906 + }, + { + "epoch": 0.8032646435496235, + "grad_norm": 1.6771029437638356, + "learning_rate": 3.923297175615121e-07, + "loss": 0.7357, + "step": 8907 + }, + { + "epoch": 0.8033548270730937, + "grad_norm": 1.597604242425022, + "learning_rate": 3.9198226721299243e-07, + "loss": 0.7897, + "step": 8908 + }, + { + "epoch": 0.803445010596564, + "grad_norm": 2.0812452024112424, + "learning_rate": 3.916349540694128e-07, + "loss": 0.7263, + "step": 8909 + }, + { + "epoch": 0.8035351941200343, + "grad_norm": 2.3494834453927442, + "learning_rate": 3.912877781604063e-07, + "loss": 0.7558, + "step": 8910 + }, + { + "epoch": 0.8036253776435045, + "grad_norm": 1.7072436726472848, + "learning_rate": 3.909407395155977e-07, + "loss": 0.7769, + "step": 8911 + }, + { + "epoch": 0.8037155611669748, + "grad_norm": 8.03460029752059, + "learning_rate": 3.9059383816459725e-07, + "loss": 0.694, + "step": 8912 + }, + { + "epoch": 0.8038057446904451, + "grad_norm": 5.920106810044392, + "learning_rate": 3.902470741370045e-07, + "loss": 0.811, + "step": 8913 + }, + { + "epoch": 0.8038959282139153, + "grad_norm": 1.4600352392777585, + "learning_rate": 3.8990044746240746e-07, + "loss": 0.7336, + "step": 8914 + }, + { + "epoch": 0.8039861117373855, + "grad_norm": 2.1813821528497113, + "learning_rate": 3.8955395817038237e-07, + "loss": 0.8449, + "step": 8915 + }, + { + "epoch": 0.8040762952608559, + "grad_norm": 1.6497728624928927, + "learning_rate": 3.892076062904934e-07, + "loss": 0.7161, + "step": 8916 + }, + { + "epoch": 0.8041664787843261, + "grad_norm": 2.0288073598258882, + "learning_rate": 3.8886139185229384e-07, + "loss": 0.7055, + "step": 8917 + }, + { + "epoch": 0.8042566623077964, + "grad_norm": 1.9452950963233793, + "learning_rate": 3.8851531488532284e-07, + "loss": 0.6608, + "step": 8918 + }, + { + "epoch": 0.8043468458312666, + "grad_norm": 1.8943814247900874, + "learning_rate": 3.88169375419112e-07, + "loss": 0.77, + "step": 8919 + }, + { + "epoch": 0.8044370293547369, + "grad_norm": 2.0823857504838914, + "learning_rate": 3.8782357348317717e-07, + "loss": 0.7197, + "step": 8920 + }, + { + "epoch": 0.8045272128782072, + "grad_norm": 2.3950922497083567, + "learning_rate": 3.8747790910702437e-07, + "loss": 0.7955, + "step": 8921 + }, + { + "epoch": 0.8046173964016774, + "grad_norm": 1.6103501559888111, + "learning_rate": 3.8713238232014776e-07, + "loss": 0.7487, + "step": 8922 + }, + { + "epoch": 0.8047075799251476, + "grad_norm": 1.8007396767672685, + "learning_rate": 3.867869931520296e-07, + "loss": 0.8069, + "step": 8923 + }, + { + "epoch": 0.804797763448618, + "grad_norm": 2.0385196591688466, + "learning_rate": 3.864417416321406e-07, + "loss": 0.7613, + "step": 8924 + }, + { + "epoch": 0.8048879469720882, + "grad_norm": 9.820804653414472, + "learning_rate": 3.8609662778993847e-07, + "loss": 0.7273, + "step": 8925 + }, + { + "epoch": 0.8049781304955584, + "grad_norm": 0.576914370299934, + "learning_rate": 3.85751651654872e-07, + "loss": 0.5579, + "step": 8926 + }, + { + "epoch": 0.8050683140190287, + "grad_norm": 1.724509697492834, + "learning_rate": 3.8540681325637505e-07, + "loss": 0.8591, + "step": 8927 + }, + { + "epoch": 0.805158497542499, + "grad_norm": 1.560952984768391, + "learning_rate": 3.8506211262387155e-07, + "loss": 0.7611, + "step": 8928 + }, + { + "epoch": 0.8052486810659693, + "grad_norm": 14.166656374460109, + "learning_rate": 3.847175497867732e-07, + "loss": 0.7627, + "step": 8929 + }, + { + "epoch": 0.8053388645894395, + "grad_norm": 1.7804042079688118, + "learning_rate": 3.843731247744801e-07, + "loss": 0.6948, + "step": 8930 + }, + { + "epoch": 0.8054290481129097, + "grad_norm": 1.6358350311804246, + "learning_rate": 3.8402883761638047e-07, + "loss": 0.7911, + "step": 8931 + }, + { + "epoch": 0.8055192316363801, + "grad_norm": 0.6699928045954001, + "learning_rate": 3.8368468834185076e-07, + "loss": 0.6138, + "step": 8932 + }, + { + "epoch": 0.8056094151598503, + "grad_norm": 3.5444603317393644, + "learning_rate": 3.8334067698025583e-07, + "loss": 0.7943, + "step": 8933 + }, + { + "epoch": 0.8056995986833205, + "grad_norm": 1.553905808116453, + "learning_rate": 3.8299680356094897e-07, + "loss": 0.6819, + "step": 8934 + }, + { + "epoch": 0.8057897822067909, + "grad_norm": 2.0362501371323907, + "learning_rate": 3.8265306811327024e-07, + "loss": 0.6447, + "step": 8935 + }, + { + "epoch": 0.8058799657302611, + "grad_norm": 3.907218873984875, + "learning_rate": 3.8230947066654994e-07, + "loss": 0.7184, + "step": 8936 + }, + { + "epoch": 0.8059701492537313, + "grad_norm": 3.1405408857478214, + "learning_rate": 3.819660112501053e-07, + "loss": 0.8117, + "step": 8937 + }, + { + "epoch": 0.8060603327772016, + "grad_norm": 0.6187502716526244, + "learning_rate": 3.816226898932422e-07, + "loss": 0.5976, + "step": 8938 + }, + { + "epoch": 0.8061505163006719, + "grad_norm": 2.1046471629422725, + "learning_rate": 3.812795066252557e-07, + "loss": 0.7398, + "step": 8939 + }, + { + "epoch": 0.8062406998241421, + "grad_norm": 2.0255973217517087, + "learning_rate": 3.8093646147542577e-07, + "loss": 0.7624, + "step": 8940 + }, + { + "epoch": 0.8063308833476124, + "grad_norm": 1.7914161359112444, + "learning_rate": 3.805935544730259e-07, + "loss": 0.7204, + "step": 8941 + }, + { + "epoch": 0.8064210668710826, + "grad_norm": 1.650505034865749, + "learning_rate": 3.802507856473118e-07, + "loss": 0.684, + "step": 8942 + }, + { + "epoch": 0.806511250394553, + "grad_norm": 2.368922236881593, + "learning_rate": 3.7990815502753317e-07, + "loss": 0.7537, + "step": 8943 + }, + { + "epoch": 0.8066014339180232, + "grad_norm": 1.7024215505397562, + "learning_rate": 3.795656626429231e-07, + "loss": 0.7539, + "step": 8944 + }, + { + "epoch": 0.8066916174414934, + "grad_norm": 2.3999497805745404, + "learning_rate": 3.792233085227059e-07, + "loss": 0.6745, + "step": 8945 + }, + { + "epoch": 0.8067818009649637, + "grad_norm": 2.0059669413785373, + "learning_rate": 3.788810926960928e-07, + "loss": 0.7612, + "step": 8946 + }, + { + "epoch": 0.806871984488434, + "grad_norm": 1.646620847718365, + "learning_rate": 3.785390151922836e-07, + "loss": 0.7357, + "step": 8947 + }, + { + "epoch": 0.8069621680119042, + "grad_norm": 2.2159681476824002, + "learning_rate": 3.781970760404665e-07, + "loss": 0.735, + "step": 8948 + }, + { + "epoch": 0.8070523515353745, + "grad_norm": 1.502019317716066, + "learning_rate": 3.778552752698176e-07, + "loss": 0.825, + "step": 8949 + }, + { + "epoch": 0.8071425350588447, + "grad_norm": 1.4848296422889304, + "learning_rate": 3.775136129095007e-07, + "loss": 0.6584, + "step": 8950 + }, + { + "epoch": 0.807232718582315, + "grad_norm": 3.3066251949304046, + "learning_rate": 3.771720889886685e-07, + "loss": 0.7198, + "step": 8951 + }, + { + "epoch": 0.8073229021057853, + "grad_norm": 0.695908794384892, + "learning_rate": 3.7683070353646194e-07, + "loss": 0.5963, + "step": 8952 + }, + { + "epoch": 0.8074130856292555, + "grad_norm": 1.7116088943364458, + "learning_rate": 3.7648945658200983e-07, + "loss": 0.712, + "step": 8953 + }, + { + "epoch": 0.8075032691527257, + "grad_norm": 2.780903569623316, + "learning_rate": 3.761483481544292e-07, + "loss": 0.6376, + "step": 8954 + }, + { + "epoch": 0.8075934526761961, + "grad_norm": 2.0484588552099448, + "learning_rate": 3.7580737828282525e-07, + "loss": 0.72, + "step": 8955 + }, + { + "epoch": 0.8076836361996663, + "grad_norm": 1.7649055973342684, + "learning_rate": 3.754665469962921e-07, + "loss": 0.728, + "step": 8956 + }, + { + "epoch": 0.8077738197231366, + "grad_norm": 1.7440271629003037, + "learning_rate": 3.7512585432390973e-07, + "loss": 0.7079, + "step": 8957 + }, + { + "epoch": 0.8078640032466069, + "grad_norm": 1.8907388700132173, + "learning_rate": 3.7478530029474987e-07, + "loss": 0.6881, + "step": 8958 + }, + { + "epoch": 0.8079541867700771, + "grad_norm": 0.6228735512278912, + "learning_rate": 3.7444488493786854e-07, + "loss": 0.628, + "step": 8959 + }, + { + "epoch": 0.8080443702935474, + "grad_norm": 2.072602240199876, + "learning_rate": 3.7410460828231405e-07, + "loss": 0.7982, + "step": 8960 + }, + { + "epoch": 0.8081345538170176, + "grad_norm": 1.4529514003956296, + "learning_rate": 3.737644703571188e-07, + "loss": 0.6839, + "step": 8961 + }, + { + "epoch": 0.8082247373404879, + "grad_norm": 1.7459924314319566, + "learning_rate": 3.734244711913059e-07, + "loss": 0.8246, + "step": 8962 + }, + { + "epoch": 0.8083149208639582, + "grad_norm": 1.5793995007032435, + "learning_rate": 3.7308461081388584e-07, + "loss": 0.7248, + "step": 8963 + }, + { + "epoch": 0.8084051043874284, + "grad_norm": 1.6671575141362371, + "learning_rate": 3.727448892538576e-07, + "loss": 0.6822, + "step": 8964 + }, + { + "epoch": 0.8084952879108986, + "grad_norm": 1.5156827306075835, + "learning_rate": 3.724053065402086e-07, + "loss": 0.8343, + "step": 8965 + }, + { + "epoch": 0.808585471434369, + "grad_norm": 1.5648432136949906, + "learning_rate": 3.7206586270191285e-07, + "loss": 0.7129, + "step": 8966 + }, + { + "epoch": 0.8086756549578392, + "grad_norm": 1.5623588498256928, + "learning_rate": 3.7172655776793385e-07, + "loss": 0.7967, + "step": 8967 + }, + { + "epoch": 0.8087658384813095, + "grad_norm": 1.6749547952274109, + "learning_rate": 3.7138739176722323e-07, + "loss": 0.643, + "step": 8968 + }, + { + "epoch": 0.8088560220047797, + "grad_norm": 1.8922417403746576, + "learning_rate": 3.710483647287206e-07, + "loss": 0.7133, + "step": 8969 + }, + { + "epoch": 0.80894620552825, + "grad_norm": 1.6267721343374693, + "learning_rate": 3.707094766813532e-07, + "loss": 0.6463, + "step": 8970 + }, + { + "epoch": 0.8090363890517203, + "grad_norm": 1.6754463056693736, + "learning_rate": 3.7037072765403754e-07, + "loss": 0.7136, + "step": 8971 + }, + { + "epoch": 0.8091265725751905, + "grad_norm": 2.022931363346707, + "learning_rate": 3.700321176756762e-07, + "loss": 0.7081, + "step": 8972 + }, + { + "epoch": 0.8092167560986607, + "grad_norm": 1.710414789208866, + "learning_rate": 3.69693646775163e-07, + "loss": 0.7943, + "step": 8973 + }, + { + "epoch": 0.8093069396221311, + "grad_norm": 2.0044577510774557, + "learning_rate": 3.693553149813764e-07, + "loss": 0.7358, + "step": 8974 + }, + { + "epoch": 0.8093971231456013, + "grad_norm": 0.5813860805737492, + "learning_rate": 3.690171223231866e-07, + "loss": 0.5515, + "step": 8975 + }, + { + "epoch": 0.8094873066690715, + "grad_norm": 1.3744143192344687, + "learning_rate": 3.6867906882944854e-07, + "loss": 0.8379, + "step": 8976 + }, + { + "epoch": 0.8095774901925418, + "grad_norm": 2.223372895300106, + "learning_rate": 3.6834115452900737e-07, + "loss": 0.7447, + "step": 8977 + }, + { + "epoch": 0.8096676737160121, + "grad_norm": 1.5264606703468795, + "learning_rate": 3.680033794506958e-07, + "loss": 0.6422, + "step": 8978 + }, + { + "epoch": 0.8097578572394823, + "grad_norm": 2.026351891885043, + "learning_rate": 3.676657436233346e-07, + "loss": 0.5768, + "step": 8979 + }, + { + "epoch": 0.8098480407629526, + "grad_norm": 1.5840443553927663, + "learning_rate": 3.6732824707573305e-07, + "loss": 0.7421, + "step": 8980 + }, + { + "epoch": 0.8099382242864229, + "grad_norm": 1.768808583613755, + "learning_rate": 3.6699088983668716e-07, + "loss": 0.7844, + "step": 8981 + }, + { + "epoch": 0.8100284078098932, + "grad_norm": 1.6802753218343496, + "learning_rate": 3.6665367193498376e-07, + "loss": 0.7008, + "step": 8982 + }, + { + "epoch": 0.8101185913333634, + "grad_norm": 2.44653562723239, + "learning_rate": 3.663165933993948e-07, + "loss": 0.7472, + "step": 8983 + }, + { + "epoch": 0.8102087748568336, + "grad_norm": 1.851577629961093, + "learning_rate": 3.659796542586822e-07, + "loss": 0.7677, + "step": 8984 + }, + { + "epoch": 0.810298958380304, + "grad_norm": 1.5317361021163174, + "learning_rate": 3.6564285454159526e-07, + "loss": 0.6995, + "step": 8985 + }, + { + "epoch": 0.8103891419037742, + "grad_norm": 1.5105303842114453, + "learning_rate": 3.653061942768718e-07, + "loss": 0.7912, + "step": 8986 + }, + { + "epoch": 0.8104793254272444, + "grad_norm": 1.7815911485007607, + "learning_rate": 3.649696734932375e-07, + "loss": 0.7026, + "step": 8987 + }, + { + "epoch": 0.8105695089507147, + "grad_norm": 3.68166134775335, + "learning_rate": 3.646332922194064e-07, + "loss": 0.6778, + "step": 8988 + }, + { + "epoch": 0.810659692474185, + "grad_norm": 1.7539423405729635, + "learning_rate": 3.6429705048407943e-07, + "loss": 0.7311, + "step": 8989 + }, + { + "epoch": 0.8107498759976552, + "grad_norm": 1.9367382644094893, + "learning_rate": 3.6396094831594804e-07, + "loss": 0.7433, + "step": 8990 + }, + { + "epoch": 0.8108400595211255, + "grad_norm": 1.6923705561172608, + "learning_rate": 3.6362498574368926e-07, + "loss": 0.6963, + "step": 8991 + }, + { + "epoch": 0.8109302430445957, + "grad_norm": 0.7048758853393342, + "learning_rate": 3.6328916279596935e-07, + "loss": 0.6246, + "step": 8992 + }, + { + "epoch": 0.811020426568066, + "grad_norm": 2.060904492626544, + "learning_rate": 3.6295347950144305e-07, + "loss": 0.7952, + "step": 8993 + }, + { + "epoch": 0.8111106100915363, + "grad_norm": 2.1802734899892497, + "learning_rate": 3.626179358887522e-07, + "loss": 0.5975, + "step": 8994 + }, + { + "epoch": 0.8112007936150065, + "grad_norm": 1.7705184506971492, + "learning_rate": 3.6228253198652816e-07, + "loss": 0.8036, + "step": 8995 + }, + { + "epoch": 0.8112909771384768, + "grad_norm": 2.3810303073573933, + "learning_rate": 3.6194726782338767e-07, + "loss": 0.6944, + "step": 8996 + }, + { + "epoch": 0.8113811606619471, + "grad_norm": 6.675191515057476, + "learning_rate": 3.6161214342793953e-07, + "loss": 0.7233, + "step": 8997 + }, + { + "epoch": 0.8114713441854173, + "grad_norm": 1.884904261416752, + "learning_rate": 3.612771588287764e-07, + "loss": 0.7741, + "step": 8998 + }, + { + "epoch": 0.8115615277088876, + "grad_norm": 2.096435305207895, + "learning_rate": 3.609423140544827e-07, + "loss": 0.8106, + "step": 8999 + }, + { + "epoch": 0.8116517112323578, + "grad_norm": 2.196565101004393, + "learning_rate": 3.6060760913362787e-07, + "loss": 0.7617, + "step": 9000 + }, + { + "epoch": 0.8117418947558281, + "grad_norm": 1.5441572416662612, + "learning_rate": 3.6027304409477146e-07, + "loss": 0.7501, + "step": 9001 + }, + { + "epoch": 0.8118320782792984, + "grad_norm": 1.8787373488679822, + "learning_rate": 3.599386189664604e-07, + "loss": 0.7159, + "step": 9002 + }, + { + "epoch": 0.8119222618027686, + "grad_norm": 1.9866242164625152, + "learning_rate": 3.5960433377722945e-07, + "loss": 0.7391, + "step": 9003 + }, + { + "epoch": 0.8120124453262388, + "grad_norm": 1.7391086623653393, + "learning_rate": 3.5927018855560174e-07, + "loss": 0.8046, + "step": 9004 + }, + { + "epoch": 0.8121026288497092, + "grad_norm": 1.571283448317736, + "learning_rate": 3.5893618333008904e-07, + "loss": 0.7339, + "step": 9005 + }, + { + "epoch": 0.8121928123731794, + "grad_norm": 1.687092201973984, + "learning_rate": 3.586023181291893e-07, + "loss": 0.8372, + "step": 9006 + }, + { + "epoch": 0.8122829958966497, + "grad_norm": 0.7575160754421141, + "learning_rate": 3.5826859298139044e-07, + "loss": 0.6114, + "step": 9007 + }, + { + "epoch": 0.81237317942012, + "grad_norm": 1.526948297141175, + "learning_rate": 3.5793500791516773e-07, + "loss": 0.7273, + "step": 9008 + }, + { + "epoch": 0.8124633629435902, + "grad_norm": 1.5398202957232827, + "learning_rate": 3.5760156295898415e-07, + "loss": 0.716, + "step": 9009 + }, + { + "epoch": 0.8125535464670605, + "grad_norm": 2.341379416372138, + "learning_rate": 3.5726825814129203e-07, + "loss": 0.7428, + "step": 9010 + }, + { + "epoch": 0.8126437299905307, + "grad_norm": 0.699564097274256, + "learning_rate": 3.5693509349052886e-07, + "loss": 0.6132, + "step": 9011 + }, + { + "epoch": 0.812733913514001, + "grad_norm": 1.6238879286614254, + "learning_rate": 3.5660206903512433e-07, + "loss": 0.8151, + "step": 9012 + }, + { + "epoch": 0.8128240970374713, + "grad_norm": 2.196043279794585, + "learning_rate": 3.56269184803492e-07, + "loss": 0.7969, + "step": 9013 + }, + { + "epoch": 0.8129142805609415, + "grad_norm": 1.5189801077127223, + "learning_rate": 3.5593644082403727e-07, + "loss": 0.7661, + "step": 9014 + }, + { + "epoch": 0.8130044640844117, + "grad_norm": 1.9531112670415636, + "learning_rate": 3.5560383712514994e-07, + "loss": 0.8424, + "step": 9015 + }, + { + "epoch": 0.8130946476078821, + "grad_norm": 1.6173505470531806, + "learning_rate": 3.5527137373521066e-07, + "loss": 0.7474, + "step": 9016 + }, + { + "epoch": 0.8131848311313523, + "grad_norm": 1.938098814934103, + "learning_rate": 3.5493905068258645e-07, + "loss": 0.7036, + "step": 9017 + }, + { + "epoch": 0.8132750146548225, + "grad_norm": 1.464149982195943, + "learning_rate": 3.546068679956333e-07, + "loss": 0.7603, + "step": 9018 + }, + { + "epoch": 0.8133651981782928, + "grad_norm": 1.6774919221459, + "learning_rate": 3.5427482570269487e-07, + "loss": 0.8079, + "step": 9019 + }, + { + "epoch": 0.8134553817017631, + "grad_norm": 2.956021619044059, + "learning_rate": 3.539429238321026e-07, + "loss": 0.5955, + "step": 9020 + }, + { + "epoch": 0.8135455652252334, + "grad_norm": 2.2517352300671774, + "learning_rate": 3.536111624121769e-07, + "loss": 0.5979, + "step": 9021 + }, + { + "epoch": 0.8136357487487036, + "grad_norm": 2.1054239073668546, + "learning_rate": 3.532795414712244e-07, + "loss": 0.7076, + "step": 9022 + }, + { + "epoch": 0.8137259322721738, + "grad_norm": 1.4273415584787499, + "learning_rate": 3.5294806103754124e-07, + "loss": 0.7642, + "step": 9023 + }, + { + "epoch": 0.8138161157956442, + "grad_norm": 1.6657549351151488, + "learning_rate": 3.526167211394115e-07, + "loss": 0.7597, + "step": 9024 + }, + { + "epoch": 0.8139062993191144, + "grad_norm": 4.095787414934989, + "learning_rate": 3.522855218051066e-07, + "loss": 0.6854, + "step": 9025 + }, + { + "epoch": 0.8139964828425846, + "grad_norm": 2.00451139418384, + "learning_rate": 3.5195446306288633e-07, + "loss": 0.7551, + "step": 9026 + }, + { + "epoch": 0.8140866663660549, + "grad_norm": 1.415816341715057, + "learning_rate": 3.51623544940999e-07, + "loss": 0.6894, + "step": 9027 + }, + { + "epoch": 0.8141768498895252, + "grad_norm": 1.460000293418123, + "learning_rate": 3.5129276746767886e-07, + "loss": 0.7317, + "step": 9028 + }, + { + "epoch": 0.8142670334129954, + "grad_norm": 1.9172547792600703, + "learning_rate": 3.5096213067115165e-07, + "loss": 0.7216, + "step": 9029 + }, + { + "epoch": 0.8143572169364657, + "grad_norm": 1.7061774521380255, + "learning_rate": 3.506316345796272e-07, + "loss": 0.7661, + "step": 9030 + }, + { + "epoch": 0.814447400459936, + "grad_norm": 0.596168936291154, + "learning_rate": 3.5030127922130714e-07, + "loss": 0.5357, + "step": 9031 + }, + { + "epoch": 0.8145375839834063, + "grad_norm": 1.5019224086840017, + "learning_rate": 3.4997106462437784e-07, + "loss": 0.6373, + "step": 9032 + }, + { + "epoch": 0.8146277675068765, + "grad_norm": 0.6558043919793448, + "learning_rate": 3.496409908170157e-07, + "loss": 0.562, + "step": 9033 + }, + { + "epoch": 0.8147179510303467, + "grad_norm": 2.925791680555506, + "learning_rate": 3.493110578273839e-07, + "loss": 0.7112, + "step": 9034 + }, + { + "epoch": 0.8148081345538171, + "grad_norm": 1.8813791477825998, + "learning_rate": 3.489812656836346e-07, + "loss": 0.8039, + "step": 9035 + }, + { + "epoch": 0.8148983180772873, + "grad_norm": 1.516059970379144, + "learning_rate": 3.486516144139078e-07, + "loss": 0.6593, + "step": 9036 + }, + { + "epoch": 0.8149885016007575, + "grad_norm": 1.5736450967525832, + "learning_rate": 3.4832210404632957e-07, + "loss": 0.7143, + "step": 9037 + }, + { + "epoch": 0.8150786851242278, + "grad_norm": 1.4341287472174808, + "learning_rate": 3.479927346090179e-07, + "loss": 0.7554, + "step": 9038 + }, + { + "epoch": 0.8151688686476981, + "grad_norm": 1.8439580509023379, + "learning_rate": 3.4766350613007455e-07, + "loss": 0.7337, + "step": 9039 + }, + { + "epoch": 0.8152590521711683, + "grad_norm": 1.7027525888349662, + "learning_rate": 3.4733441863759173e-07, + "loss": 0.7369, + "step": 9040 + }, + { + "epoch": 0.8153492356946386, + "grad_norm": 2.4131584573405815, + "learning_rate": 3.4700547215964916e-07, + "loss": 0.6397, + "step": 9041 + }, + { + "epoch": 0.8154394192181088, + "grad_norm": 1.4857212587208652, + "learning_rate": 3.46676666724314e-07, + "loss": 0.7237, + "step": 9042 + }, + { + "epoch": 0.8155296027415792, + "grad_norm": 2.862233218625403, + "learning_rate": 3.463480023596421e-07, + "loss": 0.74, + "step": 9043 + }, + { + "epoch": 0.8156197862650494, + "grad_norm": 2.207185506915646, + "learning_rate": 3.460194790936772e-07, + "loss": 0.8182, + "step": 9044 + }, + { + "epoch": 0.8157099697885196, + "grad_norm": 1.7461629081531003, + "learning_rate": 3.456910969544495e-07, + "loss": 0.7352, + "step": 9045 + }, + { + "epoch": 0.8158001533119899, + "grad_norm": 1.9660025908646146, + "learning_rate": 3.4536285596997994e-07, + "loss": 0.7755, + "step": 9046 + }, + { + "epoch": 0.8158903368354602, + "grad_norm": 1.9467227221065015, + "learning_rate": 3.450347561682747e-07, + "loss": 0.7038, + "step": 9047 + }, + { + "epoch": 0.8159805203589304, + "grad_norm": 5.353457372961534, + "learning_rate": 3.4470679757732945e-07, + "loss": 0.6873, + "step": 9048 + }, + { + "epoch": 0.8160707038824007, + "grad_norm": 2.1674721638829126, + "learning_rate": 3.4437898022512735e-07, + "loss": 0.8272, + "step": 9049 + }, + { + "epoch": 0.8161608874058709, + "grad_norm": 2.0505327927991015, + "learning_rate": 3.4405130413963977e-07, + "loss": 0.7476, + "step": 9050 + }, + { + "epoch": 0.8162510709293412, + "grad_norm": 2.3642690060337563, + "learning_rate": 3.437237693488262e-07, + "loss": 0.8141, + "step": 9051 + }, + { + "epoch": 0.8163412544528115, + "grad_norm": 1.3951212171987424, + "learning_rate": 3.433963758806322e-07, + "loss": 0.7596, + "step": 9052 + }, + { + "epoch": 0.8164314379762817, + "grad_norm": 1.654655480727923, + "learning_rate": 3.430691237629948e-07, + "loss": 0.6081, + "step": 9053 + }, + { + "epoch": 0.816521621499752, + "grad_norm": 2.919222520240734, + "learning_rate": 3.427420130238354e-07, + "loss": 0.6971, + "step": 9054 + }, + { + "epoch": 0.8166118050232223, + "grad_norm": 1.5404000448840582, + "learning_rate": 3.424150436910658e-07, + "loss": 0.8312, + "step": 9055 + }, + { + "epoch": 0.8167019885466925, + "grad_norm": 23.757982057106222, + "learning_rate": 3.420882157925842e-07, + "loss": 0.8024, + "step": 9056 + }, + { + "epoch": 0.8167921720701627, + "grad_norm": 2.557289594868263, + "learning_rate": 3.417615293562777e-07, + "loss": 0.7208, + "step": 9057 + }, + { + "epoch": 0.8168823555936331, + "grad_norm": 1.5946941197231952, + "learning_rate": 3.4143498441002105e-07, + "loss": 0.6537, + "step": 9058 + }, + { + "epoch": 0.8169725391171033, + "grad_norm": 1.5509834061699692, + "learning_rate": 3.411085809816767e-07, + "loss": 0.7521, + "step": 9059 + }, + { + "epoch": 0.8170627226405736, + "grad_norm": 1.8995152934343131, + "learning_rate": 3.407823190990953e-07, + "loss": 0.8651, + "step": 9060 + }, + { + "epoch": 0.8171529061640438, + "grad_norm": 1.9305294532139525, + "learning_rate": 3.4045619879011577e-07, + "loss": 0.8303, + "step": 9061 + }, + { + "epoch": 0.8172430896875141, + "grad_norm": 1.9987817868871522, + "learning_rate": 3.4013022008256334e-07, + "loss": 0.6733, + "step": 9062 + }, + { + "epoch": 0.8173332732109844, + "grad_norm": 1.8595972008529604, + "learning_rate": 3.398043830042532e-07, + "loss": 0.683, + "step": 9063 + }, + { + "epoch": 0.8174234567344546, + "grad_norm": 1.7666665440085507, + "learning_rate": 3.394786875829871e-07, + "loss": 0.7432, + "step": 9064 + }, + { + "epoch": 0.8175136402579248, + "grad_norm": 1.5500956444302765, + "learning_rate": 3.3915313384655564e-07, + "loss": 0.6271, + "step": 9065 + }, + { + "epoch": 0.8176038237813952, + "grad_norm": 1.3626100040790277, + "learning_rate": 3.388277218227369e-07, + "loss": 0.7593, + "step": 9066 + }, + { + "epoch": 0.8176940073048654, + "grad_norm": 2.1017732762708237, + "learning_rate": 3.3850245153929557e-07, + "loss": 0.7647, + "step": 9067 + }, + { + "epoch": 0.8177841908283356, + "grad_norm": 2.269135484970347, + "learning_rate": 3.381773230239875e-07, + "loss": 0.7046, + "step": 9068 + }, + { + "epoch": 0.8178743743518059, + "grad_norm": 1.4935165797495575, + "learning_rate": 3.3785233630455247e-07, + "loss": 0.7742, + "step": 9069 + }, + { + "epoch": 0.8179645578752762, + "grad_norm": 1.7686384024309398, + "learning_rate": 3.375274914087221e-07, + "loss": 0.7552, + "step": 9070 + }, + { + "epoch": 0.8180547413987465, + "grad_norm": 1.391066791779237, + "learning_rate": 3.3720278836421234e-07, + "loss": 0.7167, + "step": 9071 + }, + { + "epoch": 0.8181449249222167, + "grad_norm": 1.7626718450870327, + "learning_rate": 3.368782271987294e-07, + "loss": 0.7698, + "step": 9072 + }, + { + "epoch": 0.8182351084456869, + "grad_norm": 1.6233502596782785, + "learning_rate": 3.3655380793996636e-07, + "loss": 0.7514, + "step": 9073 + }, + { + "epoch": 0.8183252919691573, + "grad_norm": 1.5089275605173345, + "learning_rate": 3.362295306156047e-07, + "loss": 0.6623, + "step": 9074 + }, + { + "epoch": 0.8184154754926275, + "grad_norm": 1.6083046029379195, + "learning_rate": 3.3590539525331327e-07, + "loss": 0.7562, + "step": 9075 + }, + { + "epoch": 0.8185056590160977, + "grad_norm": 1.6315800884579617, + "learning_rate": 3.3558140188074967e-07, + "loss": 0.6935, + "step": 9076 + }, + { + "epoch": 0.8185958425395681, + "grad_norm": 4.340499869534803, + "learning_rate": 3.3525755052555817e-07, + "loss": 0.7744, + "step": 9077 + }, + { + "epoch": 0.8186860260630383, + "grad_norm": 0.7487076512685199, + "learning_rate": 3.3493384121537147e-07, + "loss": 0.6493, + "step": 9078 + }, + { + "epoch": 0.8187762095865085, + "grad_norm": 1.8901584144082282, + "learning_rate": 3.3461027397781075e-07, + "loss": 0.7436, + "step": 9079 + }, + { + "epoch": 0.8188663931099788, + "grad_norm": 4.076658723396637, + "learning_rate": 3.3428684884048397e-07, + "loss": 0.7274, + "step": 9080 + }, + { + "epoch": 0.8189565766334491, + "grad_norm": 1.520209938032089, + "learning_rate": 3.3396356583098826e-07, + "loss": 0.7459, + "step": 9081 + }, + { + "epoch": 0.8190467601569194, + "grad_norm": 1.7606503107745286, + "learning_rate": 3.3364042497690736e-07, + "loss": 0.7802, + "step": 9082 + }, + { + "epoch": 0.8191369436803896, + "grad_norm": 1.5050599268138067, + "learning_rate": 3.3331742630581405e-07, + "loss": 0.8001, + "step": 9083 + }, + { + "epoch": 0.8192271272038598, + "grad_norm": 1.418384753698564, + "learning_rate": 3.3299456984526717e-07, + "loss": 0.6907, + "step": 9084 + }, + { + "epoch": 0.8193173107273302, + "grad_norm": 1.4994439842952403, + "learning_rate": 3.3267185562281605e-07, + "loss": 0.7832, + "step": 9085 + }, + { + "epoch": 0.8194074942508004, + "grad_norm": 1.7293099340200386, + "learning_rate": 3.3234928366599514e-07, + "loss": 0.7381, + "step": 9086 + }, + { + "epoch": 0.8194976777742706, + "grad_norm": 1.7411700643469337, + "learning_rate": 3.3202685400232946e-07, + "loss": 0.7817, + "step": 9087 + }, + { + "epoch": 0.8195878612977409, + "grad_norm": 2.3663940106966725, + "learning_rate": 3.317045666593292e-07, + "loss": 0.7739, + "step": 9088 + }, + { + "epoch": 0.8196780448212112, + "grad_norm": 2.0965771163633353, + "learning_rate": 3.3138242166449426e-07, + "loss": 0.7515, + "step": 9089 + }, + { + "epoch": 0.8197682283446814, + "grad_norm": 1.5306632805769105, + "learning_rate": 3.310604190453117e-07, + "loss": 0.808, + "step": 9090 + }, + { + "epoch": 0.8198584118681517, + "grad_norm": 1.4677692953877433, + "learning_rate": 3.307385588292566e-07, + "loss": 0.694, + "step": 9091 + }, + { + "epoch": 0.8199485953916219, + "grad_norm": 1.6699635639898587, + "learning_rate": 3.304168410437924e-07, + "loss": 0.739, + "step": 9092 + }, + { + "epoch": 0.8200387789150922, + "grad_norm": 0.6317343062627954, + "learning_rate": 3.300952657163687e-07, + "loss": 0.6038, + "step": 9093 + }, + { + "epoch": 0.8201289624385625, + "grad_norm": 1.6275945637737548, + "learning_rate": 3.297738328744248e-07, + "loss": 0.726, + "step": 9094 + }, + { + "epoch": 0.8202191459620327, + "grad_norm": 1.813732714580507, + "learning_rate": 3.2945254254538714e-07, + "loss": 0.7367, + "step": 9095 + }, + { + "epoch": 0.820309329485503, + "grad_norm": 1.417278447539762, + "learning_rate": 3.2913139475666963e-07, + "loss": 0.71, + "step": 9096 + }, + { + "epoch": 0.8203995130089733, + "grad_norm": 2.1779722622663074, + "learning_rate": 3.288103895356749e-07, + "loss": 0.6935, + "step": 9097 + }, + { + "epoch": 0.8204896965324435, + "grad_norm": 1.9197287632169577, + "learning_rate": 3.284895269097927e-07, + "loss": 0.676, + "step": 9098 + }, + { + "epoch": 0.8205798800559138, + "grad_norm": 0.6697927189211097, + "learning_rate": 3.281688069063999e-07, + "loss": 0.5815, + "step": 9099 + }, + { + "epoch": 0.8206700635793841, + "grad_norm": 1.7671761240032275, + "learning_rate": 3.2784822955286396e-07, + "loss": 0.5691, + "step": 9100 + }, + { + "epoch": 0.8207602471028543, + "grad_norm": 1.4860937507700787, + "learning_rate": 3.275277948765365e-07, + "loss": 0.8154, + "step": 9101 + }, + { + "epoch": 0.8208504306263246, + "grad_norm": 2.9520807589724094, + "learning_rate": 3.2720750290475964e-07, + "loss": 0.784, + "step": 9102 + }, + { + "epoch": 0.8209406141497948, + "grad_norm": 2.009407330337447, + "learning_rate": 3.268873536648622e-07, + "loss": 0.7467, + "step": 9103 + }, + { + "epoch": 0.8210307976732651, + "grad_norm": 1.509424484421106, + "learning_rate": 3.265673471841612e-07, + "loss": 0.7174, + "step": 9104 + }, + { + "epoch": 0.8211209811967354, + "grad_norm": 0.6269200396497961, + "learning_rate": 3.262474834899616e-07, + "loss": 0.5404, + "step": 9105 + }, + { + "epoch": 0.8212111647202056, + "grad_norm": 0.6751323958643191, + "learning_rate": 3.2592776260955534e-07, + "loss": 0.6166, + "step": 9106 + }, + { + "epoch": 0.8213013482436758, + "grad_norm": 1.806311137078808, + "learning_rate": 3.256081845702239e-07, + "loss": 0.7864, + "step": 9107 + }, + { + "epoch": 0.8213915317671462, + "grad_norm": 4.445677194233971, + "learning_rate": 3.2528874939923335e-07, + "loss": 0.7924, + "step": 9108 + }, + { + "epoch": 0.8214817152906164, + "grad_norm": 1.9646306007607839, + "learning_rate": 3.2496945712384217e-07, + "loss": 0.702, + "step": 9109 + }, + { + "epoch": 0.8215718988140867, + "grad_norm": 1.8690520358243154, + "learning_rate": 3.246503077712923e-07, + "loss": 0.7256, + "step": 9110 + }, + { + "epoch": 0.8216620823375569, + "grad_norm": 1.640916852877455, + "learning_rate": 3.2433130136881625e-07, + "loss": 0.8359, + "step": 9111 + }, + { + "epoch": 0.8217522658610272, + "grad_norm": 2.50192024870452, + "learning_rate": 3.2401243794363287e-07, + "loss": 0.7424, + "step": 9112 + }, + { + "epoch": 0.8218424493844975, + "grad_norm": 1.7308810670699084, + "learning_rate": 3.236937175229495e-07, + "loss": 0.8035, + "step": 9113 + }, + { + "epoch": 0.8219326329079677, + "grad_norm": 1.8242764514888934, + "learning_rate": 3.233751401339615e-07, + "loss": 0.7188, + "step": 9114 + }, + { + "epoch": 0.8220228164314379, + "grad_norm": 2.0331311245206565, + "learning_rate": 3.2305670580385157e-07, + "loss": 0.8242, + "step": 9115 + }, + { + "epoch": 0.8221129999549083, + "grad_norm": 1.44970152183506, + "learning_rate": 3.227384145597898e-07, + "loss": 0.7447, + "step": 9116 + }, + { + "epoch": 0.8222031834783785, + "grad_norm": 2.0669331158362567, + "learning_rate": 3.224202664289346e-07, + "loss": 0.7336, + "step": 9117 + }, + { + "epoch": 0.8222933670018487, + "grad_norm": 3.0094836063930366, + "learning_rate": 3.2210226143843257e-07, + "loss": 0.6888, + "step": 9118 + }, + { + "epoch": 0.822383550525319, + "grad_norm": 0.6532242366471108, + "learning_rate": 3.217843996154173e-07, + "loss": 0.6134, + "step": 9119 + }, + { + "epoch": 0.8224737340487893, + "grad_norm": 0.6052895065296732, + "learning_rate": 3.2146668098701055e-07, + "loss": 0.529, + "step": 9120 + }, + { + "epoch": 0.8225639175722596, + "grad_norm": 2.317532090476195, + "learning_rate": 3.2114910558032215e-07, + "loss": 0.8727, + "step": 9121 + }, + { + "epoch": 0.8226541010957298, + "grad_norm": 3.6471693951423196, + "learning_rate": 3.2083167342244945e-07, + "loss": 0.7781, + "step": 9122 + }, + { + "epoch": 0.8227442846192, + "grad_norm": 1.8660273439176316, + "learning_rate": 3.205143845404763e-07, + "loss": 0.8424, + "step": 9123 + }, + { + "epoch": 0.8228344681426704, + "grad_norm": 1.573516689152635, + "learning_rate": 3.201972389614773e-07, + "loss": 0.7193, + "step": 9124 + }, + { + "epoch": 0.8229246516661406, + "grad_norm": 1.9430609852579617, + "learning_rate": 3.198802367125115e-07, + "loss": 0.7419, + "step": 9125 + }, + { + "epoch": 0.8230148351896108, + "grad_norm": 1.6915126000638974, + "learning_rate": 3.195633778206288e-07, + "loss": 0.754, + "step": 9126 + }, + { + "epoch": 0.8231050187130812, + "grad_norm": 2.3006599018627907, + "learning_rate": 3.19246662312864e-07, + "loss": 0.8215, + "step": 9127 + }, + { + "epoch": 0.8231952022365514, + "grad_norm": 0.6421700199096412, + "learning_rate": 3.189300902162417e-07, + "loss": 0.5791, + "step": 9128 + }, + { + "epoch": 0.8232853857600216, + "grad_norm": 2.484736674151663, + "learning_rate": 3.1861366155777327e-07, + "loss": 0.7603, + "step": 9129 + }, + { + "epoch": 0.8233755692834919, + "grad_norm": 1.4676921362577664, + "learning_rate": 3.182973763644583e-07, + "loss": 0.7542, + "step": 9130 + }, + { + "epoch": 0.8234657528069622, + "grad_norm": 2.6704537699714876, + "learning_rate": 3.1798123466328463e-07, + "loss": 0.6714, + "step": 9131 + }, + { + "epoch": 0.8235559363304324, + "grad_norm": 2.0459160341949105, + "learning_rate": 3.17665236481226e-07, + "loss": 0.7491, + "step": 9132 + }, + { + "epoch": 0.8236461198539027, + "grad_norm": 2.6125869626990017, + "learning_rate": 3.1734938184524576e-07, + "loss": 0.7084, + "step": 9133 + }, + { + "epoch": 0.8237363033773729, + "grad_norm": 1.8572843964056986, + "learning_rate": 3.1703367078229427e-07, + "loss": 0.7572, + "step": 9134 + }, + { + "epoch": 0.8238264869008433, + "grad_norm": 1.7625283286340878, + "learning_rate": 3.167181033193096e-07, + "loss": 0.7075, + "step": 9135 + }, + { + "epoch": 0.8239166704243135, + "grad_norm": 1.615517827801343, + "learning_rate": 3.16402679483218e-07, + "loss": 0.7591, + "step": 9136 + }, + { + "epoch": 0.8240068539477837, + "grad_norm": 2.277226911266312, + "learning_rate": 3.1608739930093366e-07, + "loss": 0.6789, + "step": 9137 + }, + { + "epoch": 0.824097037471254, + "grad_norm": 1.8507196624672708, + "learning_rate": 3.157722627993562e-07, + "loss": 0.6581, + "step": 9138 + }, + { + "epoch": 0.8241872209947243, + "grad_norm": 1.773587123928038, + "learning_rate": 3.1545727000537727e-07, + "loss": 0.7051, + "step": 9139 + }, + { + "epoch": 0.8242774045181945, + "grad_norm": 2.0967011790424497, + "learning_rate": 3.151424209458713e-07, + "loss": 0.7425, + "step": 9140 + }, + { + "epoch": 0.8243675880416648, + "grad_norm": 1.7427104626916379, + "learning_rate": 3.148277156477053e-07, + "loss": 0.7593, + "step": 9141 + }, + { + "epoch": 0.824457771565135, + "grad_norm": 2.215480598511111, + "learning_rate": 3.145131541377299e-07, + "loss": 0.7475, + "step": 9142 + }, + { + "epoch": 0.8245479550886053, + "grad_norm": 1.8117153837198203, + "learning_rate": 3.1419873644278606e-07, + "loss": 0.7124, + "step": 9143 + }, + { + "epoch": 0.8246381386120756, + "grad_norm": 7.987625565235901, + "learning_rate": 3.1388446258970147e-07, + "loss": 0.768, + "step": 9144 + }, + { + "epoch": 0.8247283221355458, + "grad_norm": 2.537282184591778, + "learning_rate": 3.1357033260529145e-07, + "loss": 0.7285, + "step": 9145 + }, + { + "epoch": 0.824818505659016, + "grad_norm": 1.5990210488742924, + "learning_rate": 3.1325634651636025e-07, + "loss": 0.8233, + "step": 9146 + }, + { + "epoch": 0.8249086891824864, + "grad_norm": 1.5067844344983488, + "learning_rate": 3.1294250434969694e-07, + "loss": 0.7883, + "step": 9147 + }, + { + "epoch": 0.8249988727059566, + "grad_norm": 2.898784112898185, + "learning_rate": 3.1262880613208274e-07, + "loss": 0.7873, + "step": 9148 + }, + { + "epoch": 0.8250890562294269, + "grad_norm": 2.396459105217729, + "learning_rate": 3.123152518902823e-07, + "loss": 0.7835, + "step": 9149 + }, + { + "epoch": 0.8251792397528972, + "grad_norm": 1.4525586737175338, + "learning_rate": 3.1200184165105017e-07, + "loss": 0.6976, + "step": 9150 + }, + { + "epoch": 0.8252694232763674, + "grad_norm": 1.765264491182887, + "learning_rate": 3.116885754411287e-07, + "loss": 0.7543, + "step": 9151 + }, + { + "epoch": 0.8253596067998377, + "grad_norm": 5.9664292542235255, + "learning_rate": 3.1137545328724703e-07, + "loss": 0.6791, + "step": 9152 + }, + { + "epoch": 0.8254497903233079, + "grad_norm": 2.204862410043516, + "learning_rate": 3.1106247521612285e-07, + "loss": 0.7489, + "step": 9153 + }, + { + "epoch": 0.8255399738467782, + "grad_norm": 2.3825531364992023, + "learning_rate": 3.107496412544612e-07, + "loss": 0.762, + "step": 9154 + }, + { + "epoch": 0.8256301573702485, + "grad_norm": 2.4262976841512915, + "learning_rate": 3.1043695142895397e-07, + "loss": 0.7707, + "step": 9155 + }, + { + "epoch": 0.8257203408937187, + "grad_norm": 1.6280733535980634, + "learning_rate": 3.101244057662828e-07, + "loss": 0.6669, + "step": 9156 + }, + { + "epoch": 0.8258105244171889, + "grad_norm": 1.7166212597492838, + "learning_rate": 3.098120042931152e-07, + "loss": 0.7339, + "step": 9157 + }, + { + "epoch": 0.8259007079406593, + "grad_norm": 1.6184003032469076, + "learning_rate": 3.0949974703610647e-07, + "loss": 0.7883, + "step": 9158 + }, + { + "epoch": 0.8259908914641295, + "grad_norm": 1.6975052535996862, + "learning_rate": 3.0918763402190107e-07, + "loss": 0.7356, + "step": 9159 + }, + { + "epoch": 0.8260810749875998, + "grad_norm": 1.8056865750565565, + "learning_rate": 3.088756652771296e-07, + "loss": 0.7015, + "step": 9160 + }, + { + "epoch": 0.82617125851107, + "grad_norm": 1.8063741621485783, + "learning_rate": 3.0856384082841147e-07, + "loss": 0.8425, + "step": 9161 + }, + { + "epoch": 0.8262614420345403, + "grad_norm": 1.6681120406698289, + "learning_rate": 3.0825216070235207e-07, + "loss": 0.6907, + "step": 9162 + }, + { + "epoch": 0.8263516255580106, + "grad_norm": 1.899282008584382, + "learning_rate": 3.0794062492554764e-07, + "loss": 0.8588, + "step": 9163 + }, + { + "epoch": 0.8264418090814808, + "grad_norm": 2.0313163159622345, + "learning_rate": 3.076292335245783e-07, + "loss": 0.7612, + "step": 9164 + }, + { + "epoch": 0.826531992604951, + "grad_norm": 2.3711683585229557, + "learning_rate": 3.073179865260145e-07, + "loss": 0.8181, + "step": 9165 + }, + { + "epoch": 0.8266221761284214, + "grad_norm": 1.5741878534886773, + "learning_rate": 3.070068839564135e-07, + "loss": 0.7528, + "step": 9166 + }, + { + "epoch": 0.8267123596518916, + "grad_norm": 1.9222950670126724, + "learning_rate": 3.0669592584232006e-07, + "loss": 0.7872, + "step": 9167 + }, + { + "epoch": 0.8268025431753618, + "grad_norm": 1.9267121515987822, + "learning_rate": 3.063851122102672e-07, + "loss": 0.6594, + "step": 9168 + }, + { + "epoch": 0.8268927266988321, + "grad_norm": 1.9696091487842478, + "learning_rate": 3.06074443086775e-07, + "loss": 0.6699, + "step": 9169 + }, + { + "epoch": 0.8269829102223024, + "grad_norm": 0.6155530800417682, + "learning_rate": 3.057639184983514e-07, + "loss": 0.531, + "step": 9170 + }, + { + "epoch": 0.8270730937457726, + "grad_norm": 1.3824553378653222, + "learning_rate": 3.054535384714927e-07, + "loss": 0.752, + "step": 9171 + }, + { + "epoch": 0.8271632772692429, + "grad_norm": 1.732069734482204, + "learning_rate": 3.0514330303268135e-07, + "loss": 0.9105, + "step": 9172 + }, + { + "epoch": 0.8272534607927132, + "grad_norm": 1.8389841327159022, + "learning_rate": 3.0483321220838876e-07, + "loss": 0.6989, + "step": 9173 + }, + { + "epoch": 0.8273436443161835, + "grad_norm": 2.1105027186582506, + "learning_rate": 3.045232660250734e-07, + "loss": 0.6029, + "step": 9174 + }, + { + "epoch": 0.8274338278396537, + "grad_norm": 1.609109282547883, + "learning_rate": 3.0421346450918185e-07, + "loss": 0.7768, + "step": 9175 + }, + { + "epoch": 0.8275240113631239, + "grad_norm": 1.8297184577492505, + "learning_rate": 3.039038076871485e-07, + "loss": 0.6936, + "step": 9176 + }, + { + "epoch": 0.8276141948865943, + "grad_norm": 1.6705011445046973, + "learning_rate": 3.035942955853934e-07, + "loss": 0.7638, + "step": 9177 + }, + { + "epoch": 0.8277043784100645, + "grad_norm": 1.7129839498744481, + "learning_rate": 3.0328492823032804e-07, + "loss": 0.7215, + "step": 9178 + }, + { + "epoch": 0.8277945619335347, + "grad_norm": 1.7294588263474935, + "learning_rate": 3.029757056483471e-07, + "loss": 0.684, + "step": 9179 + }, + { + "epoch": 0.827884745457005, + "grad_norm": 1.7514687913776255, + "learning_rate": 3.026666278658372e-07, + "loss": 0.6338, + "step": 9180 + }, + { + "epoch": 0.8279749289804753, + "grad_norm": 3.786888647311435, + "learning_rate": 3.023576949091691e-07, + "loss": 0.7445, + "step": 9181 + }, + { + "epoch": 0.8280651125039455, + "grad_norm": 2.081133163884304, + "learning_rate": 3.020489068047032e-07, + "loss": 0.6705, + "step": 9182 + }, + { + "epoch": 0.8281552960274158, + "grad_norm": 1.7102321058992502, + "learning_rate": 3.017402635787869e-07, + "loss": 0.7085, + "step": 9183 + }, + { + "epoch": 0.828245479550886, + "grad_norm": 1.6536037970378894, + "learning_rate": 3.0143176525775537e-07, + "loss": 0.7225, + "step": 9184 + }, + { + "epoch": 0.8283356630743564, + "grad_norm": 1.658716201035425, + "learning_rate": 3.0112341186793155e-07, + "loss": 0.6924, + "step": 9185 + }, + { + "epoch": 0.8284258465978266, + "grad_norm": 1.5037463610978463, + "learning_rate": 3.008152034356264e-07, + "loss": 0.6819, + "step": 9186 + }, + { + "epoch": 0.8285160301212968, + "grad_norm": 0.6314499396766116, + "learning_rate": 3.005071399871366e-07, + "loss": 0.5644, + "step": 9187 + }, + { + "epoch": 0.828606213644767, + "grad_norm": 1.934140338129567, + "learning_rate": 3.0019922154874853e-07, + "loss": 0.8073, + "step": 9188 + }, + { + "epoch": 0.8286963971682374, + "grad_norm": 1.8218328978186948, + "learning_rate": 2.998914481467356e-07, + "loss": 0.7077, + "step": 9189 + }, + { + "epoch": 0.8287865806917076, + "grad_norm": 1.822021271737434, + "learning_rate": 2.9958381980735837e-07, + "loss": 0.6956, + "step": 9190 + }, + { + "epoch": 0.8288767642151779, + "grad_norm": 2.3751287927339804, + "learning_rate": 2.992763365568658e-07, + "loss": 0.6725, + "step": 9191 + }, + { + "epoch": 0.8289669477386481, + "grad_norm": 1.9298633078361724, + "learning_rate": 2.98968998421494e-07, + "loss": 0.7461, + "step": 9192 + }, + { + "epoch": 0.8290571312621184, + "grad_norm": 1.6449049572605738, + "learning_rate": 2.98661805427467e-07, + "loss": 0.736, + "step": 9193 + }, + { + "epoch": 0.8291473147855887, + "grad_norm": 1.7212419651262274, + "learning_rate": 2.9835475760099483e-07, + "loss": 0.7595, + "step": 9194 + }, + { + "epoch": 0.8292374983090589, + "grad_norm": 1.5821860155297913, + "learning_rate": 2.9804785496827856e-07, + "loss": 0.7618, + "step": 9195 + }, + { + "epoch": 0.8293276818325293, + "grad_norm": 2.0995371535543907, + "learning_rate": 2.977410975555028e-07, + "loss": 0.7449, + "step": 9196 + }, + { + "epoch": 0.8294178653559995, + "grad_norm": 1.8106325987240637, + "learning_rate": 2.9743448538884376e-07, + "loss": 0.8047, + "step": 9197 + }, + { + "epoch": 0.8295080488794697, + "grad_norm": 1.8683724573799116, + "learning_rate": 2.9712801849446154e-07, + "loss": 0.7346, + "step": 9198 + }, + { + "epoch": 0.82959823240294, + "grad_norm": 1.8556463859253407, + "learning_rate": 2.9682169689850665e-07, + "loss": 0.693, + "step": 9199 + }, + { + "epoch": 0.8296884159264103, + "grad_norm": 0.6645600866917726, + "learning_rate": 2.9651552062711573e-07, + "loss": 0.5276, + "step": 9200 + }, + { + "epoch": 0.8297785994498805, + "grad_norm": 1.9548817782742793, + "learning_rate": 2.9620948970641333e-07, + "loss": 0.8197, + "step": 9201 + }, + { + "epoch": 0.8298687829733508, + "grad_norm": 4.681648162084567, + "learning_rate": 2.959036041625125e-07, + "loss": 0.761, + "step": 9202 + }, + { + "epoch": 0.829958966496821, + "grad_norm": 1.6874025811226612, + "learning_rate": 2.95597864021512e-07, + "loss": 0.6553, + "step": 9203 + }, + { + "epoch": 0.8300491500202913, + "grad_norm": 2.471201198869774, + "learning_rate": 2.9529226930949966e-07, + "loss": 0.7905, + "step": 9204 + }, + { + "epoch": 0.8301393335437616, + "grad_norm": 1.9208319089584973, + "learning_rate": 2.949868200525505e-07, + "loss": 0.7599, + "step": 9205 + }, + { + "epoch": 0.8302295170672318, + "grad_norm": 5.106752428342291, + "learning_rate": 2.9468151627672734e-07, + "loss": 0.7371, + "step": 9206 + }, + { + "epoch": 0.830319700590702, + "grad_norm": 1.7513744542556184, + "learning_rate": 2.9437635800808026e-07, + "loss": 0.7, + "step": 9207 + }, + { + "epoch": 0.8304098841141724, + "grad_norm": 2.06077815922809, + "learning_rate": 2.940713452726473e-07, + "loss": 0.7425, + "step": 9208 + }, + { + "epoch": 0.8305000676376426, + "grad_norm": 1.8237849175063068, + "learning_rate": 2.937664780964526e-07, + "loss": 0.7401, + "step": 9209 + }, + { + "epoch": 0.8305902511611128, + "grad_norm": 1.8907280885949282, + "learning_rate": 2.9346175650551133e-07, + "loss": 0.7549, + "step": 9210 + }, + { + "epoch": 0.8306804346845831, + "grad_norm": 1.552327727804834, + "learning_rate": 2.931571805258215e-07, + "loss": 0.7662, + "step": 9211 + }, + { + "epoch": 0.8307706182080534, + "grad_norm": 1.981646907773126, + "learning_rate": 2.9285275018337353e-07, + "loss": 0.7948, + "step": 9212 + }, + { + "epoch": 0.8308608017315237, + "grad_norm": 1.7204289645513589, + "learning_rate": 2.9254846550414146e-07, + "loss": 0.8434, + "step": 9213 + }, + { + "epoch": 0.8309509852549939, + "grad_norm": 2.5539317281764364, + "learning_rate": 2.922443265140893e-07, + "loss": 0.7981, + "step": 9214 + }, + { + "epoch": 0.8310411687784641, + "grad_norm": 1.703773515053434, + "learning_rate": 2.919403332391674e-07, + "loss": 0.7456, + "step": 9215 + }, + { + "epoch": 0.8311313523019345, + "grad_norm": 2.4414930648627093, + "learning_rate": 2.9163648570531464e-07, + "loss": 0.7662, + "step": 9216 + }, + { + "epoch": 0.8312215358254047, + "grad_norm": 2.5230136196450763, + "learning_rate": 2.9133278393845717e-07, + "loss": 0.6807, + "step": 9217 + }, + { + "epoch": 0.8313117193488749, + "grad_norm": 2.140719891971808, + "learning_rate": 2.9102922796450703e-07, + "loss": 0.8185, + "step": 9218 + }, + { + "epoch": 0.8314019028723453, + "grad_norm": 1.7295897858543057, + "learning_rate": 2.907258178093672e-07, + "loss": 0.7031, + "step": 9219 + }, + { + "epoch": 0.8314920863958155, + "grad_norm": 1.4420931040247225, + "learning_rate": 2.904225534989251e-07, + "loss": 0.6694, + "step": 9220 + }, + { + "epoch": 0.8315822699192857, + "grad_norm": 2.6403431939530604, + "learning_rate": 2.901194350590572e-07, + "loss": 0.6463, + "step": 9221 + }, + { + "epoch": 0.831672453442756, + "grad_norm": 2.0129768659586036, + "learning_rate": 2.898164625156274e-07, + "loss": 0.6269, + "step": 9222 + }, + { + "epoch": 0.8317626369662263, + "grad_norm": 2.2665448656267912, + "learning_rate": 2.8951363589448676e-07, + "loss": 0.7009, + "step": 9223 + }, + { + "epoch": 0.8318528204896966, + "grad_norm": 1.5298897287111877, + "learning_rate": 2.8921095522147434e-07, + "loss": 0.7406, + "step": 9224 + }, + { + "epoch": 0.8319430040131668, + "grad_norm": 1.6578206945678788, + "learning_rate": 2.8890842052241683e-07, + "loss": 0.7722, + "step": 9225 + }, + { + "epoch": 0.832033187536637, + "grad_norm": 0.5919065971762193, + "learning_rate": 2.886060318231267e-07, + "loss": 0.5269, + "step": 9226 + }, + { + "epoch": 0.8321233710601074, + "grad_norm": 5.257006556509195, + "learning_rate": 2.8830378914940755e-07, + "loss": 0.7713, + "step": 9227 + }, + { + "epoch": 0.8322135545835776, + "grad_norm": 7.561867679611637, + "learning_rate": 2.8800169252704675e-07, + "loss": 0.7303, + "step": 9228 + }, + { + "epoch": 0.8323037381070478, + "grad_norm": 1.6614617454310412, + "learning_rate": 2.8769974198182143e-07, + "loss": 0.7689, + "step": 9229 + }, + { + "epoch": 0.8323939216305181, + "grad_norm": 4.307080166416283, + "learning_rate": 2.873979375394955e-07, + "loss": 0.7448, + "step": 9230 + }, + { + "epoch": 0.8324841051539884, + "grad_norm": 1.8545400122190372, + "learning_rate": 2.870962792258209e-07, + "loss": 0.7781, + "step": 9231 + }, + { + "epoch": 0.8325742886774586, + "grad_norm": 3.22804887355029, + "learning_rate": 2.8679476706653716e-07, + "loss": 0.7534, + "step": 9232 + }, + { + "epoch": 0.8326644722009289, + "grad_norm": 1.9227089661924286, + "learning_rate": 2.864934010873692e-07, + "loss": 0.7683, + "step": 9233 + }, + { + "epoch": 0.8327546557243991, + "grad_norm": 1.463942675333289, + "learning_rate": 2.8619218131403357e-07, + "loss": 0.719, + "step": 9234 + }, + { + "epoch": 0.8328448392478695, + "grad_norm": 1.7477349200083103, + "learning_rate": 2.858911077722299e-07, + "loss": 0.7675, + "step": 9235 + }, + { + "epoch": 0.8329350227713397, + "grad_norm": 2.7008454517946747, + "learning_rate": 2.855901804876493e-07, + "loss": 0.6003, + "step": 9236 + }, + { + "epoch": 0.8330252062948099, + "grad_norm": 1.5499192155274593, + "learning_rate": 2.852893994859673e-07, + "loss": 0.7855, + "step": 9237 + }, + { + "epoch": 0.8331153898182801, + "grad_norm": 2.162003511974718, + "learning_rate": 2.849887647928484e-07, + "loss": 0.6875, + "step": 9238 + }, + { + "epoch": 0.8332055733417505, + "grad_norm": 2.3787672381194915, + "learning_rate": 2.8468827643394465e-07, + "loss": 0.648, + "step": 9239 + }, + { + "epoch": 0.8332957568652207, + "grad_norm": 1.6357816139138437, + "learning_rate": 2.843879344348954e-07, + "loss": 0.7531, + "step": 9240 + }, + { + "epoch": 0.833385940388691, + "grad_norm": 1.5003356558044354, + "learning_rate": 2.840877388213272e-07, + "loss": 0.7154, + "step": 9241 + }, + { + "epoch": 0.8334761239121612, + "grad_norm": 1.9869390427790805, + "learning_rate": 2.8378768961885515e-07, + "loss": 0.7481, + "step": 9242 + }, + { + "epoch": 0.8335663074356315, + "grad_norm": 3.411496942122094, + "learning_rate": 2.8348778685307983e-07, + "loss": 0.7757, + "step": 9243 + }, + { + "epoch": 0.8336564909591018, + "grad_norm": 1.769206777301514, + "learning_rate": 2.831880305495915e-07, + "loss": 0.8201, + "step": 9244 + }, + { + "epoch": 0.833746674482572, + "grad_norm": 0.6753146153619444, + "learning_rate": 2.828884207339668e-07, + "loss": 0.6119, + "step": 9245 + }, + { + "epoch": 0.8338368580060423, + "grad_norm": 1.9785943123928784, + "learning_rate": 2.8258895743177014e-07, + "loss": 0.8252, + "step": 9246 + }, + { + "epoch": 0.8339270415295126, + "grad_norm": 1.7044685034288278, + "learning_rate": 2.8228964066855356e-07, + "loss": 0.8191, + "step": 9247 + }, + { + "epoch": 0.8340172250529828, + "grad_norm": 1.9661059108158863, + "learning_rate": 2.819904704698555e-07, + "loss": 0.7236, + "step": 9248 + }, + { + "epoch": 0.834107408576453, + "grad_norm": 0.6393461255576148, + "learning_rate": 2.8169144686120437e-07, + "loss": 0.5279, + "step": 9249 + }, + { + "epoch": 0.8341975920999234, + "grad_norm": 6.020366751819695, + "learning_rate": 2.8139256986811254e-07, + "loss": 0.6576, + "step": 9250 + }, + { + "epoch": 0.8342877756233936, + "grad_norm": 1.7496935031063647, + "learning_rate": 2.8109383951608424e-07, + "loss": 0.7943, + "step": 9251 + }, + { + "epoch": 0.8343779591468639, + "grad_norm": 1.7436568560043828, + "learning_rate": 2.8079525583060683e-07, + "loss": 0.8077, + "step": 9252 + }, + { + "epoch": 0.8344681426703341, + "grad_norm": 1.38309036308044, + "learning_rate": 2.804968188371577e-07, + "loss": 0.7534, + "step": 9253 + }, + { + "epoch": 0.8345583261938044, + "grad_norm": 1.6589100188030783, + "learning_rate": 2.801985285612014e-07, + "loss": 0.7469, + "step": 9254 + }, + { + "epoch": 0.8346485097172747, + "grad_norm": 2.016136873291519, + "learning_rate": 2.7990038502818934e-07, + "loss": 0.716, + "step": 9255 + }, + { + "epoch": 0.8347386932407449, + "grad_norm": 1.4972276499187265, + "learning_rate": 2.796023882635612e-07, + "loss": 0.7223, + "step": 9256 + }, + { + "epoch": 0.8348288767642151, + "grad_norm": 1.541257116544185, + "learning_rate": 2.7930453829274323e-07, + "loss": 0.7941, + "step": 9257 + }, + { + "epoch": 0.8349190602876855, + "grad_norm": 4.297594044950653, + "learning_rate": 2.7900683514115054e-07, + "loss": 0.7894, + "step": 9258 + }, + { + "epoch": 0.8350092438111557, + "grad_norm": 1.925957785788453, + "learning_rate": 2.787092788341836e-07, + "loss": 0.6832, + "step": 9259 + }, + { + "epoch": 0.8350994273346259, + "grad_norm": 1.5920945339745671, + "learning_rate": 2.7841186939723195e-07, + "loss": 0.7445, + "step": 9260 + }, + { + "epoch": 0.8351896108580962, + "grad_norm": 1.7894156161650898, + "learning_rate": 2.7811460685567255e-07, + "loss": 0.7188, + "step": 9261 + }, + { + "epoch": 0.8352797943815665, + "grad_norm": 4.856020048335763, + "learning_rate": 2.778174912348692e-07, + "loss": 0.7414, + "step": 9262 + }, + { + "epoch": 0.8353699779050368, + "grad_norm": 2.298074940833154, + "learning_rate": 2.7752052256017354e-07, + "loss": 0.7093, + "step": 9263 + }, + { + "epoch": 0.835460161428507, + "grad_norm": 1.8023234815578149, + "learning_rate": 2.7722370085692493e-07, + "loss": 0.7159, + "step": 9264 + }, + { + "epoch": 0.8355503449519772, + "grad_norm": 2.0807657314202226, + "learning_rate": 2.769270261504486e-07, + "loss": 0.719, + "step": 9265 + }, + { + "epoch": 0.8356405284754476, + "grad_norm": 1.597059366889161, + "learning_rate": 2.7663049846606015e-07, + "loss": 0.7048, + "step": 9266 + }, + { + "epoch": 0.8357307119989178, + "grad_norm": 1.529461029713524, + "learning_rate": 2.763341178290592e-07, + "loss": 0.7611, + "step": 9267 + }, + { + "epoch": 0.835820895522388, + "grad_norm": 2.7385623803896455, + "learning_rate": 2.7603788426473663e-07, + "loss": 0.7713, + "step": 9268 + }, + { + "epoch": 0.8359110790458584, + "grad_norm": 2.128728176683796, + "learning_rate": 2.7574179779836695e-07, + "loss": 0.6791, + "step": 9269 + }, + { + "epoch": 0.8360012625693286, + "grad_norm": 1.6917395846932863, + "learning_rate": 2.754458584552146e-07, + "loss": 0.7053, + "step": 9270 + }, + { + "epoch": 0.8360914460927988, + "grad_norm": 1.9918004277001802, + "learning_rate": 2.751500662605308e-07, + "loss": 0.731, + "step": 9271 + }, + { + "epoch": 0.8361816296162691, + "grad_norm": 1.6842957911507148, + "learning_rate": 2.7485442123955383e-07, + "loss": 0.7039, + "step": 9272 + }, + { + "epoch": 0.8362718131397394, + "grad_norm": 1.7618941327612716, + "learning_rate": 2.7455892341751075e-07, + "loss": 0.6272, + "step": 9273 + }, + { + "epoch": 0.8363619966632097, + "grad_norm": 1.7653426898327067, + "learning_rate": 2.7426357281961365e-07, + "loss": 0.6177, + "step": 9274 + }, + { + "epoch": 0.8364521801866799, + "grad_norm": 1.891646054305426, + "learning_rate": 2.7396836947106416e-07, + "loss": 0.7467, + "step": 9275 + }, + { + "epoch": 0.8365423637101501, + "grad_norm": 1.7503199284890862, + "learning_rate": 2.736733133970506e-07, + "loss": 0.8399, + "step": 9276 + }, + { + "epoch": 0.8366325472336205, + "grad_norm": 1.4494988923872907, + "learning_rate": 2.7337840462274896e-07, + "loss": 0.7829, + "step": 9277 + }, + { + "epoch": 0.8367227307570907, + "grad_norm": 1.7701561773762977, + "learning_rate": 2.730836431733221e-07, + "loss": 0.7151, + "step": 9278 + }, + { + "epoch": 0.8368129142805609, + "grad_norm": 2.8234283903109043, + "learning_rate": 2.727890290739212e-07, + "loss": 0.7956, + "step": 9279 + }, + { + "epoch": 0.8369030978040312, + "grad_norm": 1.5043624384204664, + "learning_rate": 2.7249456234968395e-07, + "loss": 0.7798, + "step": 9280 + }, + { + "epoch": 0.8369932813275015, + "grad_norm": 1.5911926246002253, + "learning_rate": 2.722002430257364e-07, + "loss": 0.8133, + "step": 9281 + }, + { + "epoch": 0.8370834648509717, + "grad_norm": 1.9851218192437332, + "learning_rate": 2.7190607112719035e-07, + "loss": 0.7028, + "step": 9282 + }, + { + "epoch": 0.837173648374442, + "grad_norm": 2.7266795835523254, + "learning_rate": 2.716120466791476e-07, + "loss": 0.666, + "step": 9283 + }, + { + "epoch": 0.8372638318979122, + "grad_norm": 2.102182630065748, + "learning_rate": 2.7131816970669483e-07, + "loss": 0.7201, + "step": 9284 + }, + { + "epoch": 0.8373540154213825, + "grad_norm": 1.875152200243588, + "learning_rate": 2.7102444023490777e-07, + "loss": 0.7509, + "step": 9285 + }, + { + "epoch": 0.8374441989448528, + "grad_norm": 1.5712358028283342, + "learning_rate": 2.70730858288849e-07, + "loss": 0.8002, + "step": 9286 + }, + { + "epoch": 0.837534382468323, + "grad_norm": 1.7010280754198912, + "learning_rate": 2.704374238935685e-07, + "loss": 0.6813, + "step": 9287 + }, + { + "epoch": 0.8376245659917932, + "grad_norm": 0.673134200441465, + "learning_rate": 2.70144137074104e-07, + "loss": 0.575, + "step": 9288 + }, + { + "epoch": 0.8377147495152636, + "grad_norm": 1.9876980928752233, + "learning_rate": 2.6985099785547926e-07, + "loss": 0.7639, + "step": 9289 + }, + { + "epoch": 0.8378049330387338, + "grad_norm": 4.934612733744285, + "learning_rate": 2.695580062627083e-07, + "loss": 0.7723, + "step": 9290 + }, + { + "epoch": 0.8378951165622041, + "grad_norm": 3.4182625431770033, + "learning_rate": 2.692651623207891e-07, + "loss": 0.7395, + "step": 9291 + }, + { + "epoch": 0.8379853000856744, + "grad_norm": 1.4147588374725875, + "learning_rate": 2.689724660547097e-07, + "loss": 0.7315, + "step": 9292 + }, + { + "epoch": 0.8380754836091446, + "grad_norm": 0.6212014638718933, + "learning_rate": 2.686799174894441e-07, + "loss": 0.5717, + "step": 9293 + }, + { + "epoch": 0.8381656671326149, + "grad_norm": 0.777884175671443, + "learning_rate": 2.683875166499545e-07, + "loss": 0.5724, + "step": 9294 + }, + { + "epoch": 0.8382558506560851, + "grad_norm": 1.790301231419395, + "learning_rate": 2.680952635611899e-07, + "loss": 0.7514, + "step": 9295 + }, + { + "epoch": 0.8383460341795554, + "grad_norm": 1.7962399894844225, + "learning_rate": 2.678031582480875e-07, + "loss": 0.6713, + "step": 9296 + }, + { + "epoch": 0.8384362177030257, + "grad_norm": 1.6070659331425725, + "learning_rate": 2.6751120073557e-07, + "loss": 0.7855, + "step": 9297 + }, + { + "epoch": 0.8385264012264959, + "grad_norm": 1.6821495287424182, + "learning_rate": 2.672193910485505e-07, + "loss": 0.7118, + "step": 9298 + }, + { + "epoch": 0.8386165847499661, + "grad_norm": 1.7972898584865773, + "learning_rate": 2.669277292119265e-07, + "loss": 0.7433, + "step": 9299 + }, + { + "epoch": 0.8387067682734365, + "grad_norm": 2.080183687352206, + "learning_rate": 2.666362152505848e-07, + "loss": 0.7475, + "step": 9300 + }, + { + "epoch": 0.8387969517969067, + "grad_norm": 4.8248024776527645, + "learning_rate": 2.663448491893989e-07, + "loss": 0.6784, + "step": 9301 + }, + { + "epoch": 0.838887135320377, + "grad_norm": 1.7832227791479172, + "learning_rate": 2.6605363105322974e-07, + "loss": 0.7208, + "step": 9302 + }, + { + "epoch": 0.8389773188438472, + "grad_norm": 2.261175376963977, + "learning_rate": 2.657625608669263e-07, + "loss": 0.7438, + "step": 9303 + }, + { + "epoch": 0.8390675023673175, + "grad_norm": 1.8801100080851203, + "learning_rate": 2.654716386553224e-07, + "loss": 0.6587, + "step": 9304 + }, + { + "epoch": 0.8391576858907878, + "grad_norm": 1.8339921100034902, + "learning_rate": 2.651808644432436e-07, + "loss": 0.7356, + "step": 9305 + }, + { + "epoch": 0.839247869414258, + "grad_norm": 1.7313132539617162, + "learning_rate": 2.6489023825549807e-07, + "loss": 0.8783, + "step": 9306 + }, + { + "epoch": 0.8393380529377282, + "grad_norm": 1.97216319352526, + "learning_rate": 2.6459976011688547e-07, + "loss": 0.6715, + "step": 9307 + }, + { + "epoch": 0.8394282364611986, + "grad_norm": 1.9126619158251796, + "learning_rate": 2.6430943005219e-07, + "loss": 0.7382, + "step": 9308 + }, + { + "epoch": 0.8395184199846688, + "grad_norm": 1.401248217061769, + "learning_rate": 2.6401924808618447e-07, + "loss": 0.7706, + "step": 9309 + }, + { + "epoch": 0.839608603508139, + "grad_norm": 2.6324715322782914, + "learning_rate": 2.637292142436287e-07, + "loss": 0.7428, + "step": 9310 + }, + { + "epoch": 0.8396987870316093, + "grad_norm": 1.980481631144932, + "learning_rate": 2.6343932854927e-07, + "loss": 0.7137, + "step": 9311 + }, + { + "epoch": 0.8397889705550796, + "grad_norm": 0.6829212259284325, + "learning_rate": 2.6314959102784316e-07, + "loss": 0.5697, + "step": 9312 + }, + { + "epoch": 0.8398791540785498, + "grad_norm": 3.64536288177953, + "learning_rate": 2.6286000170407074e-07, + "loss": 0.744, + "step": 9313 + }, + { + "epoch": 0.8399693376020201, + "grad_norm": 2.1333110073033974, + "learning_rate": 2.625705606026607e-07, + "loss": 0.7191, + "step": 9314 + }, + { + "epoch": 0.8400595211254904, + "grad_norm": 2.5374479354252566, + "learning_rate": 2.622812677483106e-07, + "loss": 0.818, + "step": 9315 + }, + { + "epoch": 0.8401497046489607, + "grad_norm": 1.9130295437985705, + "learning_rate": 2.6199212316570453e-07, + "loss": 0.8007, + "step": 9316 + }, + { + "epoch": 0.8402398881724309, + "grad_norm": 1.643899683065852, + "learning_rate": 2.617031268795138e-07, + "loss": 0.7191, + "step": 9317 + }, + { + "epoch": 0.8403300716959011, + "grad_norm": 3.30464406807859, + "learning_rate": 2.614142789143976e-07, + "loss": 0.7708, + "step": 9318 + }, + { + "epoch": 0.8404202552193715, + "grad_norm": 1.482379173900191, + "learning_rate": 2.6112557929500047e-07, + "loss": 0.7223, + "step": 9319 + }, + { + "epoch": 0.8405104387428417, + "grad_norm": 1.935226583299986, + "learning_rate": 2.6083702804595817e-07, + "loss": 0.7676, + "step": 9320 + }, + { + "epoch": 0.8406006222663119, + "grad_norm": 1.4922639517647172, + "learning_rate": 2.6054862519188915e-07, + "loss": 0.7258, + "step": 9321 + }, + { + "epoch": 0.8406908057897822, + "grad_norm": 1.6628619237478983, + "learning_rate": 2.6026037075740357e-07, + "loss": 0.6832, + "step": 9322 + }, + { + "epoch": 0.8407809893132525, + "grad_norm": 1.863537209002425, + "learning_rate": 2.5997226476709524e-07, + "loss": 0.7058, + "step": 9323 + }, + { + "epoch": 0.8408711728367227, + "grad_norm": 1.9057116764583442, + "learning_rate": 2.5968430724554856e-07, + "loss": 0.7366, + "step": 9324 + }, + { + "epoch": 0.840961356360193, + "grad_norm": 1.9196328831410387, + "learning_rate": 2.5939649821733225e-07, + "loss": 0.6773, + "step": 9325 + }, + { + "epoch": 0.8410515398836632, + "grad_norm": 1.9055224421632748, + "learning_rate": 2.5910883770700433e-07, + "loss": 0.7115, + "step": 9326 + }, + { + "epoch": 0.8411417234071336, + "grad_norm": 1.3258109184366837, + "learning_rate": 2.5882132573910965e-07, + "loss": 0.583, + "step": 9327 + }, + { + "epoch": 0.8412319069306038, + "grad_norm": 1.2947322194494761, + "learning_rate": 2.585339623381801e-07, + "loss": 0.7787, + "step": 9328 + }, + { + "epoch": 0.841322090454074, + "grad_norm": 1.5025109890807287, + "learning_rate": 2.582467475287358e-07, + "loss": 0.7199, + "step": 9329 + }, + { + "epoch": 0.8414122739775443, + "grad_norm": 1.7643151615061532, + "learning_rate": 2.5795968133528224e-07, + "loss": 0.7523, + "step": 9330 + }, + { + "epoch": 0.8415024575010146, + "grad_norm": 1.6343045513075465, + "learning_rate": 2.576727637823144e-07, + "loss": 0.7439, + "step": 9331 + }, + { + "epoch": 0.8415926410244848, + "grad_norm": 1.8844369552321534, + "learning_rate": 2.5738599489431335e-07, + "loss": 0.7856, + "step": 9332 + }, + { + "epoch": 0.8416828245479551, + "grad_norm": 1.426006477020235, + "learning_rate": 2.5709937469574794e-07, + "loss": 0.8043, + "step": 9333 + }, + { + "epoch": 0.8417730080714253, + "grad_norm": 1.4045461292860044, + "learning_rate": 2.568129032110742e-07, + "loss": 0.7828, + "step": 9334 + }, + { + "epoch": 0.8418631915948956, + "grad_norm": 1.5630132976461764, + "learning_rate": 2.5652658046473565e-07, + "loss": 0.8218, + "step": 9335 + }, + { + "epoch": 0.8419533751183659, + "grad_norm": 1.7567322527074627, + "learning_rate": 2.5624040648116184e-07, + "loss": 0.7424, + "step": 9336 + }, + { + "epoch": 0.8420435586418361, + "grad_norm": 1.7255742623711927, + "learning_rate": 2.5595438128477245e-07, + "loss": 0.6871, + "step": 9337 + }, + { + "epoch": 0.8421337421653065, + "grad_norm": 3.18020753099315, + "learning_rate": 2.5566850489997096e-07, + "loss": 0.7515, + "step": 9338 + }, + { + "epoch": 0.8422239256887767, + "grad_norm": 2.241834446561792, + "learning_rate": 2.5538277735115166e-07, + "loss": 0.7705, + "step": 9339 + }, + { + "epoch": 0.8423141092122469, + "grad_norm": 2.6873967350482233, + "learning_rate": 2.5509719866269306e-07, + "loss": 0.7945, + "step": 9340 + }, + { + "epoch": 0.8424042927357172, + "grad_norm": 1.70875968109938, + "learning_rate": 2.548117688589628e-07, + "loss": 0.8048, + "step": 9341 + }, + { + "epoch": 0.8424944762591875, + "grad_norm": 2.138935842728024, + "learning_rate": 2.545264879643152e-07, + "loss": 0.7268, + "step": 9342 + }, + { + "epoch": 0.8425846597826577, + "grad_norm": 1.7807109083913413, + "learning_rate": 2.542413560030923e-07, + "loss": 0.7043, + "step": 9343 + }, + { + "epoch": 0.842674843306128, + "grad_norm": 4.291880531290806, + "learning_rate": 2.53956372999623e-07, + "loss": 0.6352, + "step": 9344 + }, + { + "epoch": 0.8427650268295982, + "grad_norm": 0.6494759895799677, + "learning_rate": 2.5367153897822293e-07, + "loss": 0.6094, + "step": 9345 + }, + { + "epoch": 0.8428552103530685, + "grad_norm": 1.5905184427179637, + "learning_rate": 2.5338685396319715e-07, + "loss": 0.7353, + "step": 9346 + }, + { + "epoch": 0.8429453938765388, + "grad_norm": 2.1814178926840158, + "learning_rate": 2.531023179788352e-07, + "loss": 0.6777, + "step": 9347 + }, + { + "epoch": 0.843035577400009, + "grad_norm": 1.5783790157330775, + "learning_rate": 2.528179310494158e-07, + "loss": 0.718, + "step": 9348 + }, + { + "epoch": 0.8431257609234792, + "grad_norm": 1.4129623011009815, + "learning_rate": 2.5253369319920436e-07, + "loss": 0.7628, + "step": 9349 + }, + { + "epoch": 0.8432159444469496, + "grad_norm": 2.717087511413229, + "learning_rate": 2.522496044524538e-07, + "loss": 0.7444, + "step": 9350 + }, + { + "epoch": 0.8433061279704198, + "grad_norm": 4.379911880651869, + "learning_rate": 2.5196566483340386e-07, + "loss": 0.8318, + "step": 9351 + }, + { + "epoch": 0.84339631149389, + "grad_norm": 1.6630255822045104, + "learning_rate": 2.516818743662825e-07, + "loss": 0.761, + "step": 9352 + }, + { + "epoch": 0.8434864950173603, + "grad_norm": 1.8903317184593815, + "learning_rate": 2.5139823307530285e-07, + "loss": 0.7409, + "step": 9353 + }, + { + "epoch": 0.8435766785408306, + "grad_norm": 9.285166931788135, + "learning_rate": 2.5111474098466836e-07, + "loss": 0.778, + "step": 9354 + }, + { + "epoch": 0.8436668620643009, + "grad_norm": 1.6293689976132788, + "learning_rate": 2.50831398118567e-07, + "loss": 0.7904, + "step": 9355 + }, + { + "epoch": 0.8437570455877711, + "grad_norm": 1.6423794039925537, + "learning_rate": 2.5054820450117576e-07, + "loss": 0.6875, + "step": 9356 + }, + { + "epoch": 0.8438472291112413, + "grad_norm": 1.684121458543797, + "learning_rate": 2.502651601566579e-07, + "loss": 0.7027, + "step": 9357 + }, + { + "epoch": 0.8439374126347117, + "grad_norm": 1.7573166889729972, + "learning_rate": 2.499822651091645e-07, + "loss": 0.6952, + "step": 9358 + }, + { + "epoch": 0.8440275961581819, + "grad_norm": 1.8819978463204168, + "learning_rate": 2.496995193828344e-07, + "loss": 0.7618, + "step": 9359 + }, + { + "epoch": 0.8441177796816521, + "grad_norm": 1.7494845312349914, + "learning_rate": 2.494169230017913e-07, + "loss": 0.7402, + "step": 9360 + }, + { + "epoch": 0.8442079632051224, + "grad_norm": 1.5103606202699835, + "learning_rate": 2.491344759901499e-07, + "loss": 0.7, + "step": 9361 + }, + { + "epoch": 0.8442981467285927, + "grad_norm": 5.780230458030996, + "learning_rate": 2.488521783720088e-07, + "loss": 0.7362, + "step": 9362 + }, + { + "epoch": 0.844388330252063, + "grad_norm": 2.4286358388004783, + "learning_rate": 2.4857003017145526e-07, + "loss": 0.7803, + "step": 9363 + }, + { + "epoch": 0.8444785137755332, + "grad_norm": 1.7182098927206522, + "learning_rate": 2.482880314125644e-07, + "loss": 0.7788, + "step": 9364 + }, + { + "epoch": 0.8445686972990035, + "grad_norm": 1.5982256110323616, + "learning_rate": 2.4800618211939726e-07, + "loss": 0.7331, + "step": 9365 + }, + { + "epoch": 0.8446588808224738, + "grad_norm": 2.253824692015794, + "learning_rate": 2.477244823160034e-07, + "loss": 0.7491, + "step": 9366 + }, + { + "epoch": 0.844749064345944, + "grad_norm": 1.6865227306653845, + "learning_rate": 2.474429320264184e-07, + "loss": 0.7663, + "step": 9367 + }, + { + "epoch": 0.8448392478694142, + "grad_norm": 1.6725622172964267, + "learning_rate": 2.47161531274666e-07, + "loss": 0.6999, + "step": 9368 + }, + { + "epoch": 0.8449294313928846, + "grad_norm": 1.6425875596581752, + "learning_rate": 2.4688028008475714e-07, + "loss": 0.6717, + "step": 9369 + }, + { + "epoch": 0.8450196149163548, + "grad_norm": 2.018567798057024, + "learning_rate": 2.465991784806891e-07, + "loss": 0.7369, + "step": 9370 + }, + { + "epoch": 0.845109798439825, + "grad_norm": 1.9134675015893716, + "learning_rate": 2.463182264864472e-07, + "loss": 0.8187, + "step": 9371 + }, + { + "epoch": 0.8451999819632953, + "grad_norm": 1.9523843810641746, + "learning_rate": 2.460374241260039e-07, + "loss": 0.6747, + "step": 9372 + }, + { + "epoch": 0.8452901654867656, + "grad_norm": 2.1627320975406135, + "learning_rate": 2.4575677142331884e-07, + "loss": 0.811, + "step": 9373 + }, + { + "epoch": 0.8453803490102358, + "grad_norm": 1.9115521843622367, + "learning_rate": 2.454762684023395e-07, + "loss": 0.6841, + "step": 9374 + }, + { + "epoch": 0.8454705325337061, + "grad_norm": 3.7412044372741105, + "learning_rate": 2.4519591508699823e-07, + "loss": 0.6903, + "step": 9375 + }, + { + "epoch": 0.8455607160571763, + "grad_norm": 1.839886642571609, + "learning_rate": 2.4491571150121815e-07, + "loss": 0.789, + "step": 9376 + }, + { + "epoch": 0.8456508995806467, + "grad_norm": 3.1888395000654604, + "learning_rate": 2.446356576689062e-07, + "loss": 0.7379, + "step": 9377 + }, + { + "epoch": 0.8457410831041169, + "grad_norm": 1.9126360502034148, + "learning_rate": 2.4435575361395976e-07, + "loss": 0.7542, + "step": 9378 + }, + { + "epoch": 0.8458312666275871, + "grad_norm": 1.741141171857562, + "learning_rate": 2.440759993602606e-07, + "loss": 0.7194, + "step": 9379 + }, + { + "epoch": 0.8459214501510574, + "grad_norm": 1.6927305327695943, + "learning_rate": 2.437963949316793e-07, + "loss": 0.6902, + "step": 9380 + }, + { + "epoch": 0.8460116336745277, + "grad_norm": 2.022668876623272, + "learning_rate": 2.435169403520729e-07, + "loss": 0.7315, + "step": 9381 + }, + { + "epoch": 0.8461018171979979, + "grad_norm": 1.8043115397631815, + "learning_rate": 2.4323763564528653e-07, + "loss": 0.7864, + "step": 9382 + }, + { + "epoch": 0.8461920007214682, + "grad_norm": 1.6393854454018986, + "learning_rate": 2.429584808351517e-07, + "loss": 0.7592, + "step": 9383 + }, + { + "epoch": 0.8462821842449384, + "grad_norm": 2.1482643751269195, + "learning_rate": 2.42679475945488e-07, + "loss": 0.8616, + "step": 9384 + }, + { + "epoch": 0.8463723677684087, + "grad_norm": 1.3767278392276534, + "learning_rate": 2.424006210001008e-07, + "loss": 0.74, + "step": 9385 + }, + { + "epoch": 0.846462551291879, + "grad_norm": 2.1353170480209367, + "learning_rate": 2.421219160227839e-07, + "loss": 0.769, + "step": 9386 + }, + { + "epoch": 0.8465527348153492, + "grad_norm": 1.5077942367791692, + "learning_rate": 2.4184336103731785e-07, + "loss": 0.7357, + "step": 9387 + }, + { + "epoch": 0.8466429183388196, + "grad_norm": 1.516306615717498, + "learning_rate": 2.4156495606747065e-07, + "loss": 0.7611, + "step": 9388 + }, + { + "epoch": 0.8467331018622898, + "grad_norm": 1.9195165662316436, + "learning_rate": 2.412867011369972e-07, + "loss": 0.8571, + "step": 9389 + }, + { + "epoch": 0.84682328538576, + "grad_norm": 3.043068097002294, + "learning_rate": 2.4100859626963997e-07, + "loss": 0.7524, + "step": 9390 + }, + { + "epoch": 0.8469134689092302, + "grad_norm": 1.384208190374749, + "learning_rate": 2.407306414891288e-07, + "loss": 0.7386, + "step": 9391 + }, + { + "epoch": 0.8470036524327006, + "grad_norm": 1.9430663841635987, + "learning_rate": 2.4045283681917893e-07, + "loss": 0.8436, + "step": 9392 + }, + { + "epoch": 0.8470938359561708, + "grad_norm": 1.7600718580011894, + "learning_rate": 2.4017518228349586e-07, + "loss": 0.8555, + "step": 9393 + }, + { + "epoch": 0.8471840194796411, + "grad_norm": 0.6109632189086028, + "learning_rate": 2.3989767790576887e-07, + "loss": 0.5921, + "step": 9394 + }, + { + "epoch": 0.8472742030031113, + "grad_norm": 1.9329453739513203, + "learning_rate": 2.396203237096781e-07, + "loss": 0.7107, + "step": 9395 + }, + { + "epoch": 0.8473643865265816, + "grad_norm": 1.8386194646727236, + "learning_rate": 2.393431197188873e-07, + "loss": 0.7346, + "step": 9396 + }, + { + "epoch": 0.8474545700500519, + "grad_norm": 2.626936697642642, + "learning_rate": 2.3906606595705004e-07, + "loss": 0.8105, + "step": 9397 + }, + { + "epoch": 0.8475447535735221, + "grad_norm": 2.2599478263233994, + "learning_rate": 2.387891624478056e-07, + "loss": 0.7935, + "step": 9398 + }, + { + "epoch": 0.8476349370969923, + "grad_norm": 1.5846739748357093, + "learning_rate": 2.3851240921478075e-07, + "loss": 0.7614, + "step": 9399 + }, + { + "epoch": 0.8477251206204627, + "grad_norm": 1.6295903002242873, + "learning_rate": 2.3823580628159057e-07, + "loss": 0.7248, + "step": 9400 + }, + { + "epoch": 0.8478153041439329, + "grad_norm": 2.089432663459882, + "learning_rate": 2.3795935367183517e-07, + "loss": 0.7257, + "step": 9401 + }, + { + "epoch": 0.8479054876674031, + "grad_norm": 1.539761457174851, + "learning_rate": 2.376830514091035e-07, + "loss": 0.7621, + "step": 9402 + }, + { + "epoch": 0.8479956711908734, + "grad_norm": 1.6138787075399306, + "learning_rate": 2.3740689951697135e-07, + "loss": 0.8238, + "step": 9403 + }, + { + "epoch": 0.8480858547143437, + "grad_norm": 2.0399090993487814, + "learning_rate": 2.371308980190012e-07, + "loss": 0.7412, + "step": 9404 + }, + { + "epoch": 0.848176038237814, + "grad_norm": 0.634334265054081, + "learning_rate": 2.3685504693874337e-07, + "loss": 0.598, + "step": 9405 + }, + { + "epoch": 0.8482662217612842, + "grad_norm": 3.9825161057007072, + "learning_rate": 2.3657934629973497e-07, + "loss": 0.7779, + "step": 9406 + }, + { + "epoch": 0.8483564052847544, + "grad_norm": 1.9266880832898556, + "learning_rate": 2.3630379612549944e-07, + "loss": 0.7472, + "step": 9407 + }, + { + "epoch": 0.8484465888082248, + "grad_norm": 0.7355881268667882, + "learning_rate": 2.3602839643954997e-07, + "loss": 0.6217, + "step": 9408 + }, + { + "epoch": 0.848536772331695, + "grad_norm": 1.3924968422900135, + "learning_rate": 2.3575314726538308e-07, + "loss": 0.7451, + "step": 9409 + }, + { + "epoch": 0.8486269558551652, + "grad_norm": 1.7149822900374838, + "learning_rate": 2.3547804862648645e-07, + "loss": 0.7814, + "step": 9410 + }, + { + "epoch": 0.8487171393786356, + "grad_norm": 1.5580492335266616, + "learning_rate": 2.3520310054633174e-07, + "loss": 0.7927, + "step": 9411 + }, + { + "epoch": 0.8488073229021058, + "grad_norm": 1.7802017875478127, + "learning_rate": 2.3492830304837973e-07, + "loss": 0.7063, + "step": 9412 + }, + { + "epoch": 0.848897506425576, + "grad_norm": 1.8297280350258134, + "learning_rate": 2.3465365615607723e-07, + "loss": 0.6889, + "step": 9413 + }, + { + "epoch": 0.8489876899490463, + "grad_norm": 3.3992318463714333, + "learning_rate": 2.3437915989285884e-07, + "loss": 0.7364, + "step": 9414 + }, + { + "epoch": 0.8490778734725166, + "grad_norm": 2.517032111289301, + "learning_rate": 2.3410481428214647e-07, + "loss": 0.8081, + "step": 9415 + }, + { + "epoch": 0.8491680569959869, + "grad_norm": 1.8413518406400344, + "learning_rate": 2.338306193473476e-07, + "loss": 0.771, + "step": 9416 + }, + { + "epoch": 0.8492582405194571, + "grad_norm": 2.2026481382885366, + "learning_rate": 2.3355657511185957e-07, + "loss": 0.7422, + "step": 9417 + }, + { + "epoch": 0.8493484240429273, + "grad_norm": 0.6590915423198875, + "learning_rate": 2.3328268159906428e-07, + "loss": 0.5627, + "step": 9418 + }, + { + "epoch": 0.8494386075663977, + "grad_norm": 2.207401318877147, + "learning_rate": 2.330089388323322e-07, + "loss": 0.793, + "step": 9419 + }, + { + "epoch": 0.8495287910898679, + "grad_norm": 1.8090994274214198, + "learning_rate": 2.327353468350204e-07, + "loss": 0.7791, + "step": 9420 + }, + { + "epoch": 0.8496189746133381, + "grad_norm": 1.6078225212415087, + "learning_rate": 2.3246190563047352e-07, + "loss": 0.7376, + "step": 9421 + }, + { + "epoch": 0.8497091581368084, + "grad_norm": 1.8986521783960848, + "learning_rate": 2.3218861524202293e-07, + "loss": 0.726, + "step": 9422 + }, + { + "epoch": 0.8497993416602787, + "grad_norm": 6.229653465956851, + "learning_rate": 2.3191547569298775e-07, + "loss": 0.7527, + "step": 9423 + }, + { + "epoch": 0.8498895251837489, + "grad_norm": 1.8221857476589913, + "learning_rate": 2.3164248700667245e-07, + "loss": 0.7351, + "step": 9424 + }, + { + "epoch": 0.8499797087072192, + "grad_norm": 2.6674855186977413, + "learning_rate": 2.313696492063717e-07, + "loss": 0.7773, + "step": 9425 + }, + { + "epoch": 0.8500698922306894, + "grad_norm": 1.953569529490869, + "learning_rate": 2.3109696231536401e-07, + "loss": 0.7309, + "step": 9426 + }, + { + "epoch": 0.8501600757541597, + "grad_norm": 1.6853868888170271, + "learning_rate": 2.3082442635691722e-07, + "loss": 0.7041, + "step": 9427 + }, + { + "epoch": 0.85025025927763, + "grad_norm": 1.8668234242740294, + "learning_rate": 2.305520413542854e-07, + "loss": 0.7538, + "step": 9428 + }, + { + "epoch": 0.8503404428011002, + "grad_norm": 1.2842897317245523, + "learning_rate": 2.3027980733071018e-07, + "loss": 0.6647, + "step": 9429 + }, + { + "epoch": 0.8504306263245704, + "grad_norm": 0.5984034635660331, + "learning_rate": 2.3000772430942027e-07, + "loss": 0.5448, + "step": 9430 + }, + { + "epoch": 0.8505208098480408, + "grad_norm": 1.632178662674417, + "learning_rate": 2.2973579231363028e-07, + "loss": 0.7589, + "step": 9431 + }, + { + "epoch": 0.850610993371511, + "grad_norm": 1.6543449385865774, + "learning_rate": 2.2946401136654446e-07, + "loss": 0.7683, + "step": 9432 + }, + { + "epoch": 0.8507011768949813, + "grad_norm": 1.5110072943298674, + "learning_rate": 2.2919238149135077e-07, + "loss": 0.7717, + "step": 9433 + }, + { + "epoch": 0.8507913604184516, + "grad_norm": 1.4625740880214104, + "learning_rate": 2.289209027112282e-07, + "loss": 0.7462, + "step": 9434 + }, + { + "epoch": 0.8508815439419218, + "grad_norm": 2.1816559247169427, + "learning_rate": 2.2864957504933934e-07, + "loss": 0.8046, + "step": 9435 + }, + { + "epoch": 0.8509717274653921, + "grad_norm": 2.765295752630778, + "learning_rate": 2.2837839852883589e-07, + "loss": 0.7251, + "step": 9436 + }, + { + "epoch": 0.8510619109888623, + "grad_norm": 1.8785823296951212, + "learning_rate": 2.2810737317285623e-07, + "loss": 0.7426, + "step": 9437 + }, + { + "epoch": 0.8511520945123326, + "grad_norm": 1.4158247615172286, + "learning_rate": 2.278364990045254e-07, + "loss": 0.684, + "step": 9438 + }, + { + "epoch": 0.8512422780358029, + "grad_norm": 2.940119772307362, + "learning_rate": 2.2756577604695625e-07, + "loss": 0.7357, + "step": 9439 + }, + { + "epoch": 0.8513324615592731, + "grad_norm": 2.064269838163654, + "learning_rate": 2.2729520432324855e-07, + "loss": 0.7638, + "step": 9440 + }, + { + "epoch": 0.8514226450827433, + "grad_norm": 1.5329329137139, + "learning_rate": 2.2702478385648826e-07, + "loss": 0.6625, + "step": 9441 + }, + { + "epoch": 0.8515128286062137, + "grad_norm": 1.7051107746507135, + "learning_rate": 2.2675451466974938e-07, + "loss": 0.7403, + "step": 9442 + }, + { + "epoch": 0.8516030121296839, + "grad_norm": 1.5982479874329565, + "learning_rate": 2.26484396786093e-07, + "loss": 0.6793, + "step": 9443 + }, + { + "epoch": 0.8516931956531542, + "grad_norm": 0.6925774907229166, + "learning_rate": 2.2621443022856667e-07, + "loss": 0.6032, + "step": 9444 + }, + { + "epoch": 0.8517833791766244, + "grad_norm": 2.0761647551765265, + "learning_rate": 2.2594461502020646e-07, + "loss": 0.7836, + "step": 9445 + }, + { + "epoch": 0.8518735627000947, + "grad_norm": 2.213467824863633, + "learning_rate": 2.2567495118403278e-07, + "loss": 0.6216, + "step": 9446 + }, + { + "epoch": 0.851963746223565, + "grad_norm": 2.709681767011193, + "learning_rate": 2.254054387430566e-07, + "loss": 0.7569, + "step": 9447 + }, + { + "epoch": 0.8520539297470352, + "grad_norm": 2.030321349091964, + "learning_rate": 2.2513607772027243e-07, + "loss": 0.7294, + "step": 9448 + }, + { + "epoch": 0.8521441132705054, + "grad_norm": 2.0394875971217568, + "learning_rate": 2.2486686813866562e-07, + "loss": 0.7079, + "step": 9449 + }, + { + "epoch": 0.8522342967939758, + "grad_norm": 2.1106929476449863, + "learning_rate": 2.2459781002120514e-07, + "loss": 0.799, + "step": 9450 + }, + { + "epoch": 0.852324480317446, + "grad_norm": 1.518860421267808, + "learning_rate": 2.243289033908491e-07, + "loss": 0.8103, + "step": 9451 + }, + { + "epoch": 0.8524146638409162, + "grad_norm": 2.322547632392116, + "learning_rate": 2.2406014827054176e-07, + "loss": 0.7597, + "step": 9452 + }, + { + "epoch": 0.8525048473643865, + "grad_norm": 4.586899820508282, + "learning_rate": 2.2379154468321525e-07, + "loss": 0.8079, + "step": 9453 + }, + { + "epoch": 0.8525950308878568, + "grad_norm": 1.4304846537199196, + "learning_rate": 2.2352309265178793e-07, + "loss": 0.7794, + "step": 9454 + }, + { + "epoch": 0.852685214411327, + "grad_norm": 2.466493565267772, + "learning_rate": 2.2325479219916565e-07, + "loss": 0.8106, + "step": 9455 + }, + { + "epoch": 0.8527753979347973, + "grad_norm": 1.622196640432757, + "learning_rate": 2.229866433482419e-07, + "loss": 0.7758, + "step": 9456 + }, + { + "epoch": 0.8528655814582676, + "grad_norm": 1.8522940691363614, + "learning_rate": 2.2271864612189552e-07, + "loss": 0.7958, + "step": 9457 + }, + { + "epoch": 0.8529557649817379, + "grad_norm": 1.7680368744944612, + "learning_rate": 2.2245080054299415e-07, + "loss": 0.7761, + "step": 9458 + }, + { + "epoch": 0.8530459485052081, + "grad_norm": 1.53198567059541, + "learning_rate": 2.2218310663439198e-07, + "loss": 0.6977, + "step": 9459 + }, + { + "epoch": 0.8531361320286783, + "grad_norm": 2.860537100023412, + "learning_rate": 2.2191556441892968e-07, + "loss": 0.7665, + "step": 9460 + }, + { + "epoch": 0.8532263155521487, + "grad_norm": 1.5179202603592463, + "learning_rate": 2.216481739194358e-07, + "loss": 0.7473, + "step": 9461 + }, + { + "epoch": 0.8533164990756189, + "grad_norm": 3.137099227537364, + "learning_rate": 2.2138093515872592e-07, + "loss": 0.6914, + "step": 9462 + }, + { + "epoch": 0.8534066825990891, + "grad_norm": 1.3629227743514396, + "learning_rate": 2.2111384815960087e-07, + "loss": 0.8273, + "step": 9463 + }, + { + "epoch": 0.8534968661225594, + "grad_norm": 1.5409272027121292, + "learning_rate": 2.208469129448518e-07, + "loss": 0.667, + "step": 9464 + }, + { + "epoch": 0.8535870496460297, + "grad_norm": 0.6303877351840946, + "learning_rate": 2.2058012953725357e-07, + "loss": 0.5688, + "step": 9465 + }, + { + "epoch": 0.8536772331695, + "grad_norm": 1.7400241986049956, + "learning_rate": 2.203134979595711e-07, + "loss": 0.7139, + "step": 9466 + }, + { + "epoch": 0.8537674166929702, + "grad_norm": 1.564594619581589, + "learning_rate": 2.2004701823455374e-07, + "loss": 0.6995, + "step": 9467 + }, + { + "epoch": 0.8538576002164404, + "grad_norm": 1.7550435955795067, + "learning_rate": 2.1978069038493906e-07, + "loss": 0.7725, + "step": 9468 + }, + { + "epoch": 0.8539477837399108, + "grad_norm": 1.9004108863056275, + "learning_rate": 2.1951451443345225e-07, + "loss": 0.7469, + "step": 9469 + }, + { + "epoch": 0.854037967263381, + "grad_norm": 1.6914319043053005, + "learning_rate": 2.1924849040280425e-07, + "loss": 0.7308, + "step": 9470 + }, + { + "epoch": 0.8541281507868512, + "grad_norm": 3.019489243623563, + "learning_rate": 2.1898261831569465e-07, + "loss": 0.8269, + "step": 9471 + }, + { + "epoch": 0.8542183343103215, + "grad_norm": 3.0843453851300358, + "learning_rate": 2.1871689819480798e-07, + "loss": 0.7501, + "step": 9472 + }, + { + "epoch": 0.8543085178337918, + "grad_norm": 1.3781889868173758, + "learning_rate": 2.1845133006281745e-07, + "loss": 0.7299, + "step": 9473 + }, + { + "epoch": 0.854398701357262, + "grad_norm": 1.546704292511109, + "learning_rate": 2.1818591394238294e-07, + "loss": 0.6021, + "step": 9474 + }, + { + "epoch": 0.8544888848807323, + "grad_norm": 1.6897191891063017, + "learning_rate": 2.1792064985615076e-07, + "loss": 0.7256, + "step": 9475 + }, + { + "epoch": 0.8545790684042025, + "grad_norm": 1.5773066296680398, + "learning_rate": 2.1765553782675528e-07, + "loss": 0.7146, + "step": 9476 + }, + { + "epoch": 0.8546692519276728, + "grad_norm": 1.4765675862542607, + "learning_rate": 2.1739057787681703e-07, + "loss": 0.8183, + "step": 9477 + }, + { + "epoch": 0.8547594354511431, + "grad_norm": 2.1032448313613403, + "learning_rate": 2.1712577002894372e-07, + "loss": 0.7709, + "step": 9478 + }, + { + "epoch": 0.8548496189746133, + "grad_norm": 1.5801676758992729, + "learning_rate": 2.1686111430573105e-07, + "loss": 0.7166, + "step": 9479 + }, + { + "epoch": 0.8549398024980835, + "grad_norm": 1.4922940680511907, + "learning_rate": 2.165966107297592e-07, + "loss": 0.7979, + "step": 9480 + }, + { + "epoch": 0.8550299860215539, + "grad_norm": 4.53787084487974, + "learning_rate": 2.16332259323599e-07, + "loss": 0.7719, + "step": 9481 + }, + { + "epoch": 0.8551201695450241, + "grad_norm": 1.8486174649802036, + "learning_rate": 2.1606806010980504e-07, + "loss": 0.6939, + "step": 9482 + }, + { + "epoch": 0.8552103530684944, + "grad_norm": 2.1335970934098207, + "learning_rate": 2.1580401311092067e-07, + "loss": 0.7381, + "step": 9483 + }, + { + "epoch": 0.8553005365919647, + "grad_norm": 1.496497834219036, + "learning_rate": 2.1554011834947604e-07, + "loss": 0.7336, + "step": 9484 + }, + { + "epoch": 0.8553907201154349, + "grad_norm": 1.8884883187851569, + "learning_rate": 2.1527637584798764e-07, + "loss": 0.7733, + "step": 9485 + }, + { + "epoch": 0.8554809036389052, + "grad_norm": 1.7885640878518023, + "learning_rate": 2.150127856289603e-07, + "loss": 0.8098, + "step": 9486 + }, + { + "epoch": 0.8555710871623754, + "grad_norm": 0.6207392178527381, + "learning_rate": 2.1474934771488363e-07, + "loss": 0.5572, + "step": 9487 + }, + { + "epoch": 0.8556612706858457, + "grad_norm": 1.5424276827733914, + "learning_rate": 2.1448606212823715e-07, + "loss": 0.7912, + "step": 9488 + }, + { + "epoch": 0.855751454209316, + "grad_norm": 1.672651378013361, + "learning_rate": 2.1422292889148452e-07, + "loss": 0.7287, + "step": 9489 + }, + { + "epoch": 0.8558416377327862, + "grad_norm": 1.594767002917467, + "learning_rate": 2.139599480270784e-07, + "loss": 0.7625, + "step": 9490 + }, + { + "epoch": 0.8559318212562564, + "grad_norm": 3.5225623285154755, + "learning_rate": 2.1369711955745773e-07, + "loss": 0.7344, + "step": 9491 + }, + { + "epoch": 0.8560220047797268, + "grad_norm": 2.4228510704793433, + "learning_rate": 2.1343444350504813e-07, + "loss": 0.7002, + "step": 9492 + }, + { + "epoch": 0.856112188303197, + "grad_norm": 1.5468876385413302, + "learning_rate": 2.1317191989226302e-07, + "loss": 0.7466, + "step": 9493 + }, + { + "epoch": 0.8562023718266673, + "grad_norm": 2.3289478051318198, + "learning_rate": 2.129095487415027e-07, + "loss": 0.73, + "step": 9494 + }, + { + "epoch": 0.8562925553501375, + "grad_norm": 2.0543029193169065, + "learning_rate": 2.1264733007515257e-07, + "loss": 0.7568, + "step": 9495 + }, + { + "epoch": 0.8563827388736078, + "grad_norm": 1.944975858308078, + "learning_rate": 2.1238526391558852e-07, + "loss": 0.7263, + "step": 9496 + }, + { + "epoch": 0.8564729223970781, + "grad_norm": 2.079810813956545, + "learning_rate": 2.1212335028517003e-07, + "loss": 0.882, + "step": 9497 + }, + { + "epoch": 0.8565631059205483, + "grad_norm": 1.9501171296961375, + "learning_rate": 2.1186158920624563e-07, + "loss": 0.7512, + "step": 9498 + }, + { + "epoch": 0.8566532894440185, + "grad_norm": 1.6482067941065273, + "learning_rate": 2.1159998070115015e-07, + "loss": 0.7452, + "step": 9499 + }, + { + "epoch": 0.8567434729674889, + "grad_norm": 1.778375615323674, + "learning_rate": 2.113385247922055e-07, + "loss": 0.727, + "step": 9500 + }, + { + "epoch": 0.8568336564909591, + "grad_norm": 2.147693852122851, + "learning_rate": 2.1107722150172068e-07, + "loss": 0.7518, + "step": 9501 + }, + { + "epoch": 0.8569238400144293, + "grad_norm": 4.973837880645031, + "learning_rate": 2.108160708519906e-07, + "loss": 0.7813, + "step": 9502 + }, + { + "epoch": 0.8570140235378996, + "grad_norm": 2.088633762780262, + "learning_rate": 2.1055507286529984e-07, + "loss": 0.7263, + "step": 9503 + }, + { + "epoch": 0.8571042070613699, + "grad_norm": 1.6757613316488162, + "learning_rate": 2.1029422756391612e-07, + "loss": 0.7653, + "step": 9504 + }, + { + "epoch": 0.8571943905848401, + "grad_norm": 1.5391759685379167, + "learning_rate": 2.1003353497009812e-07, + "loss": 0.6586, + "step": 9505 + }, + { + "epoch": 0.8572845741083104, + "grad_norm": 1.6006813416550414, + "learning_rate": 2.0977299510608825e-07, + "loss": 0.6686, + "step": 9506 + }, + { + "epoch": 0.8573747576317807, + "grad_norm": 1.9097676574234301, + "learning_rate": 2.0951260799411784e-07, + "loss": 0.7761, + "step": 9507 + }, + { + "epoch": 0.857464941155251, + "grad_norm": 1.9755452670909466, + "learning_rate": 2.0925237365640424e-07, + "loss": 0.737, + "step": 9508 + }, + { + "epoch": 0.8575551246787212, + "grad_norm": 0.6519577574751045, + "learning_rate": 2.0899229211515211e-07, + "loss": 0.5761, + "step": 9509 + }, + { + "epoch": 0.8576453082021914, + "grad_norm": 15.564333888087946, + "learning_rate": 2.0873236339255306e-07, + "loss": 0.6725, + "step": 9510 + }, + { + "epoch": 0.8577354917256618, + "grad_norm": 1.490516086014249, + "learning_rate": 2.0847258751078644e-07, + "loss": 0.6909, + "step": 9511 + }, + { + "epoch": 0.857825675249132, + "grad_norm": 4.01094916965762, + "learning_rate": 2.082129644920163e-07, + "loss": 0.7208, + "step": 9512 + }, + { + "epoch": 0.8579158587726022, + "grad_norm": 1.6207536823278537, + "learning_rate": 2.0795349435839605e-07, + "loss": 0.7546, + "step": 9513 + }, + { + "epoch": 0.8580060422960725, + "grad_norm": 2.3506815835014616, + "learning_rate": 2.0769417713206484e-07, + "loss": 0.7167, + "step": 9514 + }, + { + "epoch": 0.8580962258195428, + "grad_norm": 1.8406527091306177, + "learning_rate": 2.074350128351492e-07, + "loss": 0.806, + "step": 9515 + }, + { + "epoch": 0.858186409343013, + "grad_norm": 1.479248510254559, + "learning_rate": 2.0717600148976256e-07, + "loss": 0.6651, + "step": 9516 + }, + { + "epoch": 0.8582765928664833, + "grad_norm": 3.2389542601589567, + "learning_rate": 2.0691714311800436e-07, + "loss": 0.6828, + "step": 9517 + }, + { + "epoch": 0.8583667763899535, + "grad_norm": 2.9497158818900697, + "learning_rate": 2.066584377419631e-07, + "loss": 0.8022, + "step": 9518 + }, + { + "epoch": 0.8584569599134239, + "grad_norm": 1.6220745015420397, + "learning_rate": 2.0639988538371167e-07, + "loss": 0.7651, + "step": 9519 + }, + { + "epoch": 0.8585471434368941, + "grad_norm": 2.6408064740364803, + "learning_rate": 2.0614148606531258e-07, + "loss": 0.7117, + "step": 9520 + }, + { + "epoch": 0.8586373269603643, + "grad_norm": 2.466512221149301, + "learning_rate": 2.0588323980881285e-07, + "loss": 0.778, + "step": 9521 + }, + { + "epoch": 0.8587275104838346, + "grad_norm": 1.69444308862823, + "learning_rate": 2.0562514663624752e-07, + "loss": 0.7315, + "step": 9522 + }, + { + "epoch": 0.8588176940073049, + "grad_norm": 1.4594202169401544, + "learning_rate": 2.0536720656963902e-07, + "loss": 0.8104, + "step": 9523 + }, + { + "epoch": 0.8589078775307751, + "grad_norm": 2.050361168842937, + "learning_rate": 2.051094196309957e-07, + "loss": 0.7692, + "step": 9524 + }, + { + "epoch": 0.8589980610542454, + "grad_norm": 1.5379771941965767, + "learning_rate": 2.0485178584231378e-07, + "loss": 0.7041, + "step": 9525 + }, + { + "epoch": 0.8590882445777156, + "grad_norm": 0.6710979270583021, + "learning_rate": 2.0459430522557587e-07, + "loss": 0.5782, + "step": 9526 + }, + { + "epoch": 0.8591784281011859, + "grad_norm": 3.164995974135889, + "learning_rate": 2.0433697780275195e-07, + "loss": 0.7478, + "step": 9527 + }, + { + "epoch": 0.8592686116246562, + "grad_norm": 1.624872716173565, + "learning_rate": 2.040798035957978e-07, + "loss": 0.7626, + "step": 9528 + }, + { + "epoch": 0.8593587951481264, + "grad_norm": 0.7312720466615948, + "learning_rate": 2.038227826266574e-07, + "loss": 0.5936, + "step": 9529 + }, + { + "epoch": 0.8594489786715968, + "grad_norm": 4.0396775269933025, + "learning_rate": 2.0356591491726126e-07, + "loss": 0.752, + "step": 9530 + }, + { + "epoch": 0.859539162195067, + "grad_norm": 1.433735687131205, + "learning_rate": 2.033092004895267e-07, + "loss": 0.7846, + "step": 9531 + }, + { + "epoch": 0.8596293457185372, + "grad_norm": 1.689927368634814, + "learning_rate": 2.03052639365358e-07, + "loss": 0.7649, + "step": 9532 + }, + { + "epoch": 0.8597195292420075, + "grad_norm": 1.7168354815589484, + "learning_rate": 2.0279623156664694e-07, + "loss": 0.7648, + "step": 9533 + }, + { + "epoch": 0.8598097127654778, + "grad_norm": 1.6062650538826353, + "learning_rate": 2.0253997711527005e-07, + "loss": 0.8061, + "step": 9534 + }, + { + "epoch": 0.859899896288948, + "grad_norm": 1.9741729646482713, + "learning_rate": 2.0228387603309428e-07, + "loss": 0.7917, + "step": 9535 + }, + { + "epoch": 0.8599900798124183, + "grad_norm": 1.521451937485175, + "learning_rate": 2.0202792834197035e-07, + "loss": 0.7654, + "step": 9536 + }, + { + "epoch": 0.8600802633358885, + "grad_norm": 1.8987412700238924, + "learning_rate": 2.017721340637375e-07, + "loss": 0.6742, + "step": 9537 + }, + { + "epoch": 0.8601704468593588, + "grad_norm": 1.6969252970566777, + "learning_rate": 2.0151649322022134e-07, + "loss": 0.7332, + "step": 9538 + }, + { + "epoch": 0.8602606303828291, + "grad_norm": 1.9114402402875565, + "learning_rate": 2.012610058332349e-07, + "loss": 0.5579, + "step": 9539 + }, + { + "epoch": 0.8603508139062993, + "grad_norm": 0.6774845157416305, + "learning_rate": 2.010056719245774e-07, + "loss": 0.5741, + "step": 9540 + }, + { + "epoch": 0.8604409974297695, + "grad_norm": 1.57436027553286, + "learning_rate": 2.0075049151603563e-07, + "loss": 0.7671, + "step": 9541 + }, + { + "epoch": 0.8605311809532399, + "grad_norm": 1.5576528515215677, + "learning_rate": 2.0049546462938326e-07, + "loss": 0.7518, + "step": 9542 + }, + { + "epoch": 0.8606213644767101, + "grad_norm": 1.4836734921045487, + "learning_rate": 2.0024059128637961e-07, + "loss": 0.7652, + "step": 9543 + }, + { + "epoch": 0.8607115480001803, + "grad_norm": 1.472188860237337, + "learning_rate": 1.99985871508773e-07, + "loss": 0.6123, + "step": 9544 + }, + { + "epoch": 0.8608017315236506, + "grad_norm": 1.795163741227692, + "learning_rate": 1.9973130531829674e-07, + "loss": 0.7714, + "step": 9545 + }, + { + "epoch": 0.8608919150471209, + "grad_norm": 1.6970331978863564, + "learning_rate": 1.994768927366721e-07, + "loss": 0.6977, + "step": 9546 + }, + { + "epoch": 0.8609820985705912, + "grad_norm": 7.329449174072531, + "learning_rate": 1.992226337856069e-07, + "loss": 0.7147, + "step": 9547 + }, + { + "epoch": 0.8610722820940614, + "grad_norm": 1.5726354318104048, + "learning_rate": 1.9896852848679592e-07, + "loss": 0.7, + "step": 9548 + }, + { + "epoch": 0.8611624656175316, + "grad_norm": 2.2087526343199864, + "learning_rate": 1.9871457686192094e-07, + "loss": 0.7554, + "step": 9549 + }, + { + "epoch": 0.861252649141002, + "grad_norm": 1.8128015990735886, + "learning_rate": 1.984607789326509e-07, + "loss": 0.7418, + "step": 9550 + }, + { + "epoch": 0.8613428326644722, + "grad_norm": 3.5602002668307025, + "learning_rate": 1.982071347206402e-07, + "loss": 0.6941, + "step": 9551 + }, + { + "epoch": 0.8614330161879424, + "grad_norm": 1.4572267784667612, + "learning_rate": 1.9795364424753202e-07, + "loss": 0.7194, + "step": 9552 + }, + { + "epoch": 0.8615231997114128, + "grad_norm": 1.9157282218068592, + "learning_rate": 1.9770030753495505e-07, + "loss": 0.7106, + "step": 9553 + }, + { + "epoch": 0.861613383234883, + "grad_norm": 1.918100138901195, + "learning_rate": 1.9744712460452573e-07, + "loss": 0.7034, + "step": 9554 + }, + { + "epoch": 0.8617035667583532, + "grad_norm": 1.8361988287457145, + "learning_rate": 1.9719409547784703e-07, + "loss": 0.7287, + "step": 9555 + }, + { + "epoch": 0.8617937502818235, + "grad_norm": 2.285464952272335, + "learning_rate": 1.9694122017650837e-07, + "loss": 0.7682, + "step": 9556 + }, + { + "epoch": 0.8618839338052938, + "grad_norm": 1.5531488980885861, + "learning_rate": 1.9668849872208738e-07, + "loss": 0.7606, + "step": 9557 + }, + { + "epoch": 0.8619741173287641, + "grad_norm": 1.6241184190786933, + "learning_rate": 1.9643593113614632e-07, + "loss": 0.591, + "step": 9558 + }, + { + "epoch": 0.8620643008522343, + "grad_norm": 1.4560611553291543, + "learning_rate": 1.961835174402371e-07, + "loss": 0.7839, + "step": 9559 + }, + { + "epoch": 0.8621544843757045, + "grad_norm": 1.9716371226522416, + "learning_rate": 1.9593125765589535e-07, + "loss": 0.7149, + "step": 9560 + }, + { + "epoch": 0.8622446678991749, + "grad_norm": 1.5763198045706772, + "learning_rate": 1.9567915180464721e-07, + "loss": 0.6932, + "step": 9561 + }, + { + "epoch": 0.8623348514226451, + "grad_norm": 1.970969568696003, + "learning_rate": 1.9542719990800217e-07, + "loss": 0.7307, + "step": 9562 + }, + { + "epoch": 0.8624250349461153, + "grad_norm": 2.3400120746268414, + "learning_rate": 1.9517540198745896e-07, + "loss": 0.8137, + "step": 9563 + }, + { + "epoch": 0.8625152184695856, + "grad_norm": 8.874250353462962, + "learning_rate": 1.94923758064502e-07, + "loss": 0.7444, + "step": 9564 + }, + { + "epoch": 0.8626054019930559, + "grad_norm": 1.680035364369157, + "learning_rate": 1.9467226816060322e-07, + "loss": 0.7113, + "step": 9565 + }, + { + "epoch": 0.8626955855165261, + "grad_norm": 2.0699918678603093, + "learning_rate": 1.9442093229722122e-07, + "loss": 0.6841, + "step": 9566 + }, + { + "epoch": 0.8627857690399964, + "grad_norm": 2.308141467203276, + "learning_rate": 1.9416975049580085e-07, + "loss": 0.8247, + "step": 9567 + }, + { + "epoch": 0.8628759525634666, + "grad_norm": 7.375884180893242, + "learning_rate": 1.9391872277777456e-07, + "loss": 0.7311, + "step": 9568 + }, + { + "epoch": 0.862966136086937, + "grad_norm": 3.31536626066907, + "learning_rate": 1.9366784916456158e-07, + "loss": 0.7318, + "step": 9569 + }, + { + "epoch": 0.8630563196104072, + "grad_norm": 1.5631207568184642, + "learning_rate": 1.9341712967756774e-07, + "loss": 0.6978, + "step": 9570 + }, + { + "epoch": 0.8631465031338774, + "grad_norm": 2.053964814127087, + "learning_rate": 1.9316656433818566e-07, + "loss": 0.6543, + "step": 9571 + }, + { + "epoch": 0.8632366866573477, + "grad_norm": 1.6832919458771647, + "learning_rate": 1.929161531677954e-07, + "loss": 0.6446, + "step": 9572 + }, + { + "epoch": 0.863326870180818, + "grad_norm": 1.6490615458989017, + "learning_rate": 1.9266589618776251e-07, + "loss": 0.6989, + "step": 9573 + }, + { + "epoch": 0.8634170537042882, + "grad_norm": 1.9483487743008385, + "learning_rate": 1.924157934194417e-07, + "loss": 0.8004, + "step": 9574 + }, + { + "epoch": 0.8635072372277585, + "grad_norm": 2.2201180337337285, + "learning_rate": 1.9216584488417142e-07, + "loss": 0.6357, + "step": 9575 + }, + { + "epoch": 0.8635974207512288, + "grad_norm": 2.6053242564705377, + "learning_rate": 1.919160506032802e-07, + "loss": 0.7995, + "step": 9576 + }, + { + "epoch": 0.863687604274699, + "grad_norm": 1.5642207017093226, + "learning_rate": 1.916664105980812e-07, + "loss": 0.805, + "step": 9577 + }, + { + "epoch": 0.8637777877981693, + "grad_norm": 1.8201597747257754, + "learning_rate": 1.914169248898747e-07, + "loss": 0.7816, + "step": 9578 + }, + { + "epoch": 0.8638679713216395, + "grad_norm": 1.269560358740806, + "learning_rate": 1.9116759349994882e-07, + "loss": 0.6568, + "step": 9579 + }, + { + "epoch": 0.8639581548451098, + "grad_norm": 1.6039209631808486, + "learning_rate": 1.9091841644957763e-07, + "loss": 0.7837, + "step": 9580 + }, + { + "epoch": 0.8640483383685801, + "grad_norm": 1.4520751277886945, + "learning_rate": 1.9066939376002278e-07, + "loss": 0.8276, + "step": 9581 + }, + { + "epoch": 0.8641385218920503, + "grad_norm": 2.083647538968717, + "learning_rate": 1.9042052545253085e-07, + "loss": 0.6615, + "step": 9582 + }, + { + "epoch": 0.8642287054155205, + "grad_norm": 1.6985873046617967, + "learning_rate": 1.901718115483384e-07, + "loss": 0.7224, + "step": 9583 + }, + { + "epoch": 0.8643188889389909, + "grad_norm": 1.8287111304553059, + "learning_rate": 1.8992325206866598e-07, + "loss": 0.7274, + "step": 9584 + }, + { + "epoch": 0.8644090724624611, + "grad_norm": 1.3599967849917838, + "learning_rate": 1.8967484703472225e-07, + "loss": 0.7191, + "step": 9585 + }, + { + "epoch": 0.8644992559859314, + "grad_norm": 1.3796908818135445, + "learning_rate": 1.8942659646770288e-07, + "loss": 0.7904, + "step": 9586 + }, + { + "epoch": 0.8645894395094016, + "grad_norm": 1.877915722045706, + "learning_rate": 1.8917850038878936e-07, + "loss": 0.8086, + "step": 9587 + }, + { + "epoch": 0.8646796230328719, + "grad_norm": 1.5409984515143356, + "learning_rate": 1.8893055881915121e-07, + "loss": 0.6515, + "step": 9588 + }, + { + "epoch": 0.8647698065563422, + "grad_norm": 2.2127559449169123, + "learning_rate": 1.886827717799442e-07, + "loss": 0.6921, + "step": 9589 + }, + { + "epoch": 0.8648599900798124, + "grad_norm": 2.0001144376439752, + "learning_rate": 1.884351392923096e-07, + "loss": 0.7596, + "step": 9590 + }, + { + "epoch": 0.8649501736032826, + "grad_norm": 3.0910497113231767, + "learning_rate": 1.8818766137737896e-07, + "loss": 0.7805, + "step": 9591 + }, + { + "epoch": 0.865040357126753, + "grad_norm": 18.22003724145695, + "learning_rate": 1.8794033805626653e-07, + "loss": 0.7243, + "step": 9592 + }, + { + "epoch": 0.8651305406502232, + "grad_norm": 1.5745552403828484, + "learning_rate": 1.876931693500763e-07, + "loss": 0.7376, + "step": 9593 + }, + { + "epoch": 0.8652207241736934, + "grad_norm": 1.7747612268737873, + "learning_rate": 1.8744615527989783e-07, + "loss": 0.592, + "step": 9594 + }, + { + "epoch": 0.8653109076971637, + "grad_norm": 2.0017416046793595, + "learning_rate": 1.871992958668076e-07, + "loss": 0.8334, + "step": 9595 + }, + { + "epoch": 0.865401091220634, + "grad_norm": 2.5188383820506144, + "learning_rate": 1.8695259113186944e-07, + "loss": 0.789, + "step": 9596 + }, + { + "epoch": 0.8654912747441043, + "grad_norm": 2.0461225218496, + "learning_rate": 1.8670604109613252e-07, + "loss": 0.7074, + "step": 9597 + }, + { + "epoch": 0.8655814582675745, + "grad_norm": 1.5564019799228959, + "learning_rate": 1.8645964578063533e-07, + "loss": 0.7892, + "step": 9598 + }, + { + "epoch": 0.8656716417910447, + "grad_norm": 1.5296517419580886, + "learning_rate": 1.862134052064006e-07, + "loss": 0.7994, + "step": 9599 + }, + { + "epoch": 0.8657618253145151, + "grad_norm": 18.91748036162379, + "learning_rate": 1.8596731939443932e-07, + "loss": 0.7861, + "step": 9600 + }, + { + "epoch": 0.8658520088379853, + "grad_norm": 1.6493068048126664, + "learning_rate": 1.857213883657487e-07, + "loss": 0.7741, + "step": 9601 + }, + { + "epoch": 0.8659421923614555, + "grad_norm": 0.6426544839773066, + "learning_rate": 1.8547561214131303e-07, + "loss": 0.5507, + "step": 9602 + }, + { + "epoch": 0.8660323758849259, + "grad_norm": 121.27577045725168, + "learning_rate": 1.8522999074210355e-07, + "loss": 0.8205, + "step": 9603 + }, + { + "epoch": 0.8661225594083961, + "grad_norm": 2.4428632371232095, + "learning_rate": 1.849845241890775e-07, + "loss": 0.782, + "step": 9604 + }, + { + "epoch": 0.8662127429318663, + "grad_norm": 1.6560970451695787, + "learning_rate": 1.8473921250317992e-07, + "loss": 0.744, + "step": 9605 + }, + { + "epoch": 0.8663029264553366, + "grad_norm": 2.2276567948939987, + "learning_rate": 1.8449405570534225e-07, + "loss": 0.675, + "step": 9606 + }, + { + "epoch": 0.8663931099788069, + "grad_norm": 1.5434795837782116, + "learning_rate": 1.8424905381648204e-07, + "loss": 0.7891, + "step": 9607 + }, + { + "epoch": 0.8664832935022772, + "grad_norm": 2.1399099450459342, + "learning_rate": 1.8400420685750452e-07, + "loss": 0.7002, + "step": 9608 + }, + { + "epoch": 0.8665734770257474, + "grad_norm": 2.3775313088715166, + "learning_rate": 1.8375951484930142e-07, + "loss": 0.6811, + "step": 9609 + }, + { + "epoch": 0.8666636605492176, + "grad_norm": 4.952784577714055, + "learning_rate": 1.8351497781275094e-07, + "loss": 0.7471, + "step": 9610 + }, + { + "epoch": 0.866753844072688, + "grad_norm": 2.0286221451818878, + "learning_rate": 1.8327059576871907e-07, + "loss": 0.7277, + "step": 9611 + }, + { + "epoch": 0.8668440275961582, + "grad_norm": 1.4389819923106528, + "learning_rate": 1.8302636873805665e-07, + "loss": 0.7938, + "step": 9612 + }, + { + "epoch": 0.8669342111196284, + "grad_norm": 1.7905377298495473, + "learning_rate": 1.8278229674160373e-07, + "loss": 0.8214, + "step": 9613 + }, + { + "epoch": 0.8670243946430987, + "grad_norm": 1.332524541088512, + "learning_rate": 1.825383798001845e-07, + "loss": 0.6767, + "step": 9614 + }, + { + "epoch": 0.867114578166569, + "grad_norm": 1.6568840540606526, + "learning_rate": 1.8229461793461297e-07, + "loss": 0.689, + "step": 9615 + }, + { + "epoch": 0.8672047616900392, + "grad_norm": 1.625694639950157, + "learning_rate": 1.8205101116568698e-07, + "loss": 0.7586, + "step": 9616 + }, + { + "epoch": 0.8672949452135095, + "grad_norm": 1.9310857363473708, + "learning_rate": 1.818075595141928e-07, + "loss": 0.7943, + "step": 9617 + }, + { + "epoch": 0.8673851287369797, + "grad_norm": 2.398830915858477, + "learning_rate": 1.8156426300090288e-07, + "loss": 0.741, + "step": 9618 + }, + { + "epoch": 0.86747531226045, + "grad_norm": 1.4665032917119725, + "learning_rate": 1.8132112164657686e-07, + "loss": 0.7309, + "step": 9619 + }, + { + "epoch": 0.8675654957839203, + "grad_norm": 1.6716369744678294, + "learning_rate": 1.8107813547196106e-07, + "loss": 0.6381, + "step": 9620 + }, + { + "epoch": 0.8676556793073905, + "grad_norm": 1.9359033219812671, + "learning_rate": 1.8083530449778817e-07, + "loss": 0.7828, + "step": 9621 + }, + { + "epoch": 0.8677458628308607, + "grad_norm": 1.843809870378247, + "learning_rate": 1.8059262874477787e-07, + "loss": 0.6682, + "step": 9622 + }, + { + "epoch": 0.8678360463543311, + "grad_norm": 3.598133366164062, + "learning_rate": 1.8035010823363627e-07, + "loss": 0.7281, + "step": 9623 + }, + { + "epoch": 0.8679262298778013, + "grad_norm": 6.672613098626718, + "learning_rate": 1.8010774298505705e-07, + "loss": 0.782, + "step": 9624 + }, + { + "epoch": 0.8680164134012716, + "grad_norm": 2.6549014988100383, + "learning_rate": 1.7986553301972007e-07, + "loss": 0.835, + "step": 9625 + }, + { + "epoch": 0.8681065969247419, + "grad_norm": 1.8346115193774595, + "learning_rate": 1.7962347835829171e-07, + "loss": 0.7354, + "step": 9626 + }, + { + "epoch": 0.8681967804482121, + "grad_norm": 1.5058158343757992, + "learning_rate": 1.793815790214257e-07, + "loss": 0.7486, + "step": 9627 + }, + { + "epoch": 0.8682869639716824, + "grad_norm": 1.8445313947678363, + "learning_rate": 1.791398350297626e-07, + "loss": 0.7045, + "step": 9628 + }, + { + "epoch": 0.8683771474951526, + "grad_norm": 1.9296046559701485, + "learning_rate": 1.7889824640392813e-07, + "loss": 0.7496, + "step": 9629 + }, + { + "epoch": 0.868467331018623, + "grad_norm": 1.8461265764453787, + "learning_rate": 1.7865681316453741e-07, + "loss": 0.7241, + "step": 9630 + }, + { + "epoch": 0.8685575145420932, + "grad_norm": 1.7209060324151793, + "learning_rate": 1.7841553533218968e-07, + "loss": 0.7436, + "step": 9631 + }, + { + "epoch": 0.8686476980655634, + "grad_norm": 24.576709471316537, + "learning_rate": 1.7817441292747292e-07, + "loss": 0.8472, + "step": 9632 + }, + { + "epoch": 0.8687378815890336, + "grad_norm": 1.6920055006221884, + "learning_rate": 1.779334459709607e-07, + "loss": 0.7359, + "step": 9633 + }, + { + "epoch": 0.868828065112504, + "grad_norm": 2.0043141803131914, + "learning_rate": 1.7769263448321347e-07, + "loss": 0.7528, + "step": 9634 + }, + { + "epoch": 0.8689182486359742, + "grad_norm": 3.294935189103213, + "learning_rate": 1.7745197848477879e-07, + "loss": 0.736, + "step": 9635 + }, + { + "epoch": 0.8690084321594445, + "grad_norm": 2.0380077922848026, + "learning_rate": 1.7721147799619063e-07, + "loss": 0.8448, + "step": 9636 + }, + { + "epoch": 0.8690986156829147, + "grad_norm": 0.5848510333256652, + "learning_rate": 1.769711330379704e-07, + "loss": 0.5475, + "step": 9637 + }, + { + "epoch": 0.869188799206385, + "grad_norm": 1.9005310696714481, + "learning_rate": 1.767309436306248e-07, + "loss": 0.7923, + "step": 9638 + }, + { + "epoch": 0.8692789827298553, + "grad_norm": 1.9816514195186359, + "learning_rate": 1.764909097946483e-07, + "loss": 0.8082, + "step": 9639 + }, + { + "epoch": 0.8693691662533255, + "grad_norm": 1.9500200539560326, + "learning_rate": 1.7625103155052236e-07, + "loss": 0.8105, + "step": 9640 + }, + { + "epoch": 0.8694593497767957, + "grad_norm": 1.7428377588721822, + "learning_rate": 1.760113089187143e-07, + "loss": 0.7582, + "step": 9641 + }, + { + "epoch": 0.8695495333002661, + "grad_norm": 2.3408107383382637, + "learning_rate": 1.7577174191967868e-07, + "loss": 0.7699, + "step": 9642 + }, + { + "epoch": 0.8696397168237363, + "grad_norm": 1.6780147296045334, + "learning_rate": 1.755323305738574e-07, + "loss": 0.7556, + "step": 9643 + }, + { + "epoch": 0.8697299003472065, + "grad_norm": 1.4758195572218562, + "learning_rate": 1.7529307490167677e-07, + "loss": 0.7759, + "step": 9644 + }, + { + "epoch": 0.8698200838706768, + "grad_norm": 3.2416853550337112, + "learning_rate": 1.7505397492355288e-07, + "loss": 0.7281, + "step": 9645 + }, + { + "epoch": 0.8699102673941471, + "grad_norm": 1.8376163030654333, + "learning_rate": 1.7481503065988589e-07, + "loss": 0.776, + "step": 9646 + }, + { + "epoch": 0.8700004509176174, + "grad_norm": 1.374688850223531, + "learning_rate": 1.7457624213106526e-07, + "loss": 0.7107, + "step": 9647 + }, + { + "epoch": 0.8700906344410876, + "grad_norm": 1.750400565488853, + "learning_rate": 1.7433760935746465e-07, + "loss": 0.7469, + "step": 9648 + }, + { + "epoch": 0.8701808179645579, + "grad_norm": 1.9030024897789115, + "learning_rate": 1.740991323594456e-07, + "loss": 0.6901, + "step": 9649 + }, + { + "epoch": 0.8702710014880282, + "grad_norm": 1.585458241476097, + "learning_rate": 1.7386081115735651e-07, + "loss": 0.7232, + "step": 9650 + }, + { + "epoch": 0.8703611850114984, + "grad_norm": 2.285613498424103, + "learning_rate": 1.736226457715324e-07, + "loss": 0.7063, + "step": 9651 + }, + { + "epoch": 0.8704513685349686, + "grad_norm": 1.3814545769032274, + "learning_rate": 1.7338463622229505e-07, + "loss": 0.6974, + "step": 9652 + }, + { + "epoch": 0.870541552058439, + "grad_norm": 14.874818945031262, + "learning_rate": 1.7314678252995152e-07, + "loss": 0.689, + "step": 9653 + }, + { + "epoch": 0.8706317355819092, + "grad_norm": 1.6093788517285523, + "learning_rate": 1.7290908471479847e-07, + "loss": 0.6759, + "step": 9654 + }, + { + "epoch": 0.8707219191053794, + "grad_norm": 1.9872108681017617, + "learning_rate": 1.7267154279711637e-07, + "loss": 0.7412, + "step": 9655 + }, + { + "epoch": 0.8708121026288497, + "grad_norm": 1.406736374562119, + "learning_rate": 1.724341567971741e-07, + "loss": 0.791, + "step": 9656 + }, + { + "epoch": 0.87090228615232, + "grad_norm": 1.6558265864569648, + "learning_rate": 1.7219692673522657e-07, + "loss": 0.8435, + "step": 9657 + }, + { + "epoch": 0.8709924696757902, + "grad_norm": 1.6822053003334916, + "learning_rate": 1.7195985263151558e-07, + "loss": 0.7636, + "step": 9658 + }, + { + "epoch": 0.8710826531992605, + "grad_norm": 2.083305384130556, + "learning_rate": 1.7172293450626985e-07, + "loss": 0.7919, + "step": 9659 + }, + { + "epoch": 0.8711728367227307, + "grad_norm": 2.0553552971053124, + "learning_rate": 1.7148617237970475e-07, + "loss": 0.7449, + "step": 9660 + }, + { + "epoch": 0.8712630202462011, + "grad_norm": 1.748513135049652, + "learning_rate": 1.7124956627202102e-07, + "loss": 0.7087, + "step": 9661 + }, + { + "epoch": 0.8713532037696713, + "grad_norm": 3.790193809646818, + "learning_rate": 1.7101311620340852e-07, + "loss": 0.7604, + "step": 9662 + }, + { + "epoch": 0.8714433872931415, + "grad_norm": 1.7654347275017255, + "learning_rate": 1.7077682219404155e-07, + "loss": 0.692, + "step": 9663 + }, + { + "epoch": 0.8715335708166118, + "grad_norm": 7.017574863449005, + "learning_rate": 1.705406842640824e-07, + "loss": 0.8705, + "step": 9664 + }, + { + "epoch": 0.8716237543400821, + "grad_norm": 1.4697324835065533, + "learning_rate": 1.7030470243367946e-07, + "loss": 0.7343, + "step": 9665 + }, + { + "epoch": 0.8717139378635523, + "grad_norm": 1.3513886885704387, + "learning_rate": 1.7006887672296834e-07, + "loss": 0.6685, + "step": 9666 + }, + { + "epoch": 0.8718041213870226, + "grad_norm": 1.8014329901775352, + "learning_rate": 1.6983320715207094e-07, + "loss": 0.6879, + "step": 9667 + }, + { + "epoch": 0.8718943049104928, + "grad_norm": 1.4387890385055095, + "learning_rate": 1.6959769374109523e-07, + "loss": 0.7631, + "step": 9668 + }, + { + "epoch": 0.8719844884339631, + "grad_norm": 1.5468017098375197, + "learning_rate": 1.6936233651013754e-07, + "loss": 0.6812, + "step": 9669 + }, + { + "epoch": 0.8720746719574334, + "grad_norm": 1.5188123870720218, + "learning_rate": 1.691271354792787e-07, + "loss": 0.7744, + "step": 9670 + }, + { + "epoch": 0.8721648554809036, + "grad_norm": 1.462819972452116, + "learning_rate": 1.6889209066858866e-07, + "loss": 0.6238, + "step": 9671 + }, + { + "epoch": 0.872255039004374, + "grad_norm": 1.6413453473463075, + "learning_rate": 1.6865720209812185e-07, + "loss": 0.7892, + "step": 9672 + }, + { + "epoch": 0.8723452225278442, + "grad_norm": 2.0028006494879502, + "learning_rate": 1.684224697879204e-07, + "loss": 0.7107, + "step": 9673 + }, + { + "epoch": 0.8724354060513144, + "grad_norm": 3.102701972332143, + "learning_rate": 1.6818789375801302e-07, + "loss": 0.819, + "step": 9674 + }, + { + "epoch": 0.8725255895747847, + "grad_norm": 1.4437590627674834, + "learning_rate": 1.679534740284152e-07, + "loss": 0.6993, + "step": 9675 + }, + { + "epoch": 0.872615773098255, + "grad_norm": 1.660313282439961, + "learning_rate": 1.6771921061912853e-07, + "loss": 0.6855, + "step": 9676 + }, + { + "epoch": 0.8727059566217252, + "grad_norm": 1.8919997458387208, + "learning_rate": 1.6748510355014234e-07, + "loss": 0.6816, + "step": 9677 + }, + { + "epoch": 0.8727961401451955, + "grad_norm": 1.6745089166904443, + "learning_rate": 1.6725115284143132e-07, + "loss": 0.7552, + "step": 9678 + }, + { + "epoch": 0.8728863236686657, + "grad_norm": 2.6901600115808977, + "learning_rate": 1.670173585129575e-07, + "loss": 0.72, + "step": 9679 + }, + { + "epoch": 0.872976507192136, + "grad_norm": 1.624874256844332, + "learning_rate": 1.667837205846696e-07, + "loss": 0.802, + "step": 9680 + }, + { + "epoch": 0.8730666907156063, + "grad_norm": 2.1371016638729463, + "learning_rate": 1.6655023907650278e-07, + "loss": 0.7498, + "step": 9681 + }, + { + "epoch": 0.8731568742390765, + "grad_norm": 2.2272748911746243, + "learning_rate": 1.6631691400837954e-07, + "loss": 0.7816, + "step": 9682 + }, + { + "epoch": 0.8732470577625467, + "grad_norm": 1.5216085557123478, + "learning_rate": 1.6608374540020752e-07, + "loss": 0.7949, + "step": 9683 + }, + { + "epoch": 0.8733372412860171, + "grad_norm": 0.6888906978340736, + "learning_rate": 1.658507332718828e-07, + "loss": 0.5272, + "step": 9684 + }, + { + "epoch": 0.8734274248094873, + "grad_norm": 1.4758763409376898, + "learning_rate": 1.656178776432864e-07, + "loss": 0.6788, + "step": 9685 + }, + { + "epoch": 0.8735176083329576, + "grad_norm": 1.4873511744827956, + "learning_rate": 1.6538517853428814e-07, + "loss": 0.723, + "step": 9686 + }, + { + "epoch": 0.8736077918564278, + "grad_norm": 2.483014387945553, + "learning_rate": 1.6515263596474194e-07, + "loss": 0.724, + "step": 9687 + }, + { + "epoch": 0.8736979753798981, + "grad_norm": 2.996941119629134, + "learning_rate": 1.6492024995449017e-07, + "loss": 0.7305, + "step": 9688 + }, + { + "epoch": 0.8737881589033684, + "grad_norm": 1.671636546590342, + "learning_rate": 1.6468802052336116e-07, + "loss": 0.7075, + "step": 9689 + }, + { + "epoch": 0.8738783424268386, + "grad_norm": 1.6219685962990686, + "learning_rate": 1.6445594769116998e-07, + "loss": 0.7627, + "step": 9690 + }, + { + "epoch": 0.8739685259503088, + "grad_norm": 1.633362891779618, + "learning_rate": 1.6422403147771836e-07, + "loss": 0.8148, + "step": 9691 + }, + { + "epoch": 0.8740587094737792, + "grad_norm": 1.8995039342491573, + "learning_rate": 1.6399227190279485e-07, + "loss": 0.707, + "step": 9692 + }, + { + "epoch": 0.8741488929972494, + "grad_norm": 1.8762463559821194, + "learning_rate": 1.637606689861748e-07, + "loss": 0.6744, + "step": 9693 + }, + { + "epoch": 0.8742390765207196, + "grad_norm": 36.32381862916793, + "learning_rate": 1.6352922274761883e-07, + "loss": 0.7013, + "step": 9694 + }, + { + "epoch": 0.87432926004419, + "grad_norm": 1.8112651301476663, + "learning_rate": 1.6329793320687602e-07, + "loss": 0.6072, + "step": 9695 + }, + { + "epoch": 0.8744194435676602, + "grad_norm": 1.8617830471088603, + "learning_rate": 1.630668003836808e-07, + "loss": 0.781, + "step": 9696 + }, + { + "epoch": 0.8745096270911304, + "grad_norm": 5.424334493152457, + "learning_rate": 1.62835824297755e-07, + "loss": 0.7471, + "step": 9697 + }, + { + "epoch": 0.8745998106146007, + "grad_norm": 1.5054179730881718, + "learning_rate": 1.626050049688066e-07, + "loss": 0.7544, + "step": 9698 + }, + { + "epoch": 0.874689994138071, + "grad_norm": 11.110374523755059, + "learning_rate": 1.623743424165309e-07, + "loss": 0.7194, + "step": 9699 + }, + { + "epoch": 0.8747801776615413, + "grad_norm": 1.7610605029360067, + "learning_rate": 1.6214383666060826e-07, + "loss": 0.6953, + "step": 9700 + }, + { + "epoch": 0.8748703611850115, + "grad_norm": 2.576172615380864, + "learning_rate": 1.619134877207078e-07, + "loss": 0.7638, + "step": 9701 + }, + { + "epoch": 0.8749605447084817, + "grad_norm": 1.6181154399190116, + "learning_rate": 1.616832956164831e-07, + "loss": 0.7447, + "step": 9702 + }, + { + "epoch": 0.8750507282319521, + "grad_norm": 1.5755547469931679, + "learning_rate": 1.6145326036757667e-07, + "loss": 0.7135, + "step": 9703 + }, + { + "epoch": 0.8751409117554223, + "grad_norm": 9.796919778481199, + "learning_rate": 1.612233819936155e-07, + "loss": 0.7081, + "step": 9704 + }, + { + "epoch": 0.8752310952788925, + "grad_norm": 1.7368741973622523, + "learning_rate": 1.6099366051421414e-07, + "loss": 0.7408, + "step": 9705 + }, + { + "epoch": 0.8753212788023628, + "grad_norm": 1.5222628246161212, + "learning_rate": 1.6076409594897378e-07, + "loss": 0.803, + "step": 9706 + }, + { + "epoch": 0.8754114623258331, + "grad_norm": 0.662764481167924, + "learning_rate": 1.605346883174823e-07, + "loss": 0.5273, + "step": 9707 + }, + { + "epoch": 0.8755016458493033, + "grad_norm": 1.8881484275761018, + "learning_rate": 1.6030543763931427e-07, + "loss": 0.7169, + "step": 9708 + }, + { + "epoch": 0.8755918293727736, + "grad_norm": 1.6722401728182343, + "learning_rate": 1.600763439340298e-07, + "loss": 0.7618, + "step": 9709 + }, + { + "epoch": 0.8756820128962438, + "grad_norm": 1.6538780790226366, + "learning_rate": 1.5984740722117707e-07, + "loss": 0.7962, + "step": 9710 + }, + { + "epoch": 0.8757721964197142, + "grad_norm": 1.7133292981606154, + "learning_rate": 1.5961862752028998e-07, + "loss": 0.7566, + "step": 9711 + }, + { + "epoch": 0.8758623799431844, + "grad_norm": 1.386938201303147, + "learning_rate": 1.5939000485088937e-07, + "loss": 0.6526, + "step": 9712 + }, + { + "epoch": 0.8759525634666546, + "grad_norm": 1.9298917222162386, + "learning_rate": 1.5916153923248254e-07, + "loss": 0.8387, + "step": 9713 + }, + { + "epoch": 0.8760427469901249, + "grad_norm": 1.530222022209803, + "learning_rate": 1.5893323068456342e-07, + "loss": 0.7842, + "step": 9714 + }, + { + "epoch": 0.8761329305135952, + "grad_norm": 1.4892002582372563, + "learning_rate": 1.5870507922661248e-07, + "loss": 0.7901, + "step": 9715 + }, + { + "epoch": 0.8762231140370654, + "grad_norm": 0.6234892706540567, + "learning_rate": 1.5847708487809763e-07, + "loss": 0.5683, + "step": 9716 + }, + { + "epoch": 0.8763132975605357, + "grad_norm": 4.647726565579515, + "learning_rate": 1.5824924765847113e-07, + "loss": 0.8231, + "step": 9717 + }, + { + "epoch": 0.8764034810840059, + "grad_norm": 1.4324830453345332, + "learning_rate": 1.5802156758717478e-07, + "loss": 0.7221, + "step": 9718 + }, + { + "epoch": 0.8764936646074762, + "grad_norm": 1.7653248624660827, + "learning_rate": 1.5779404468363433e-07, + "loss": 0.7452, + "step": 9719 + }, + { + "epoch": 0.8765838481309465, + "grad_norm": 1.827634305389303, + "learning_rate": 1.5756667896726405e-07, + "loss": 0.797, + "step": 9720 + }, + { + "epoch": 0.8766740316544167, + "grad_norm": 8.762533312419961, + "learning_rate": 1.5733947045746377e-07, + "loss": 0.7622, + "step": 9721 + }, + { + "epoch": 0.876764215177887, + "grad_norm": 1.452594291003297, + "learning_rate": 1.5711241917362018e-07, + "loss": 0.809, + "step": 9722 + }, + { + "epoch": 0.8768543987013573, + "grad_norm": 2.512908226744956, + "learning_rate": 1.5688552513510688e-07, + "loss": 0.6875, + "step": 9723 + }, + { + "epoch": 0.8769445822248275, + "grad_norm": 2.874082170016599, + "learning_rate": 1.5665878836128266e-07, + "loss": 0.7663, + "step": 9724 + }, + { + "epoch": 0.8770347657482978, + "grad_norm": 1.4906593206841083, + "learning_rate": 1.5643220887149554e-07, + "loss": 0.7908, + "step": 9725 + }, + { + "epoch": 0.8771249492717681, + "grad_norm": 0.6630898027489976, + "learning_rate": 1.562057866850772e-07, + "loss": 0.5884, + "step": 9726 + }, + { + "epoch": 0.8772151327952383, + "grad_norm": 1.795268792134131, + "learning_rate": 1.5597952182134777e-07, + "loss": 0.6175, + "step": 9727 + }, + { + "epoch": 0.8773053163187086, + "grad_norm": 1.5146665393724075, + "learning_rate": 1.557534142996133e-07, + "loss": 0.7995, + "step": 9728 + }, + { + "epoch": 0.8773954998421788, + "grad_norm": 1.9785769002149214, + "learning_rate": 1.5552746413916662e-07, + "loss": 0.7775, + "step": 9729 + }, + { + "epoch": 0.8774856833656491, + "grad_norm": 1.7064712967782099, + "learning_rate": 1.5530167135928697e-07, + "loss": 0.7375, + "step": 9730 + }, + { + "epoch": 0.8775758668891194, + "grad_norm": 1.874984486833611, + "learning_rate": 1.5507603597924068e-07, + "loss": 0.6923, + "step": 9731 + }, + { + "epoch": 0.8776660504125896, + "grad_norm": 3.1019190047688214, + "learning_rate": 1.548505580182793e-07, + "loss": 0.7063, + "step": 9732 + }, + { + "epoch": 0.8777562339360598, + "grad_norm": 1.5553904722051666, + "learning_rate": 1.5462523749564271e-07, + "loss": 0.7453, + "step": 9733 + }, + { + "epoch": 0.8778464174595302, + "grad_norm": 1.532245857112423, + "learning_rate": 1.5440007443055602e-07, + "loss": 0.7397, + "step": 9734 + }, + { + "epoch": 0.8779366009830004, + "grad_norm": 4.2810417180317595, + "learning_rate": 1.541750688422314e-07, + "loss": 0.6796, + "step": 9735 + }, + { + "epoch": 0.8780267845064706, + "grad_norm": 1.975243471324827, + "learning_rate": 1.5395022074986797e-07, + "loss": 0.747, + "step": 9736 + }, + { + "epoch": 0.8781169680299409, + "grad_norm": 1.6615936877882418, + "learning_rate": 1.5372553017265033e-07, + "loss": 0.6873, + "step": 9737 + }, + { + "epoch": 0.8782071515534112, + "grad_norm": 2.2253046063000705, + "learning_rate": 1.5350099712975116e-07, + "loss": 0.8035, + "step": 9738 + }, + { + "epoch": 0.8782973350768815, + "grad_norm": 1.7992364376522167, + "learning_rate": 1.5327662164032785e-07, + "loss": 0.737, + "step": 9739 + }, + { + "epoch": 0.8783875186003517, + "grad_norm": 1.6033246278559525, + "learning_rate": 1.5305240372352656e-07, + "loss": 0.5937, + "step": 9740 + }, + { + "epoch": 0.8784777021238219, + "grad_norm": 1.9009613089800474, + "learning_rate": 1.5282834339847738e-07, + "loss": 0.7582, + "step": 9741 + }, + { + "epoch": 0.8785678856472923, + "grad_norm": 1.6957550174459504, + "learning_rate": 1.526044406842999e-07, + "loss": 0.7231, + "step": 9742 + }, + { + "epoch": 0.8786580691707625, + "grad_norm": 1.770480880069563, + "learning_rate": 1.523806956000977e-07, + "loss": 0.6735, + "step": 9743 + }, + { + "epoch": 0.8787482526942327, + "grad_norm": 1.8580417501236717, + "learning_rate": 1.5215710816496197e-07, + "loss": 0.7967, + "step": 9744 + }, + { + "epoch": 0.8788384362177031, + "grad_norm": 1.752031713655928, + "learning_rate": 1.5193367839797077e-07, + "loss": 0.6919, + "step": 9745 + }, + { + "epoch": 0.8789286197411733, + "grad_norm": 1.4559439112318848, + "learning_rate": 1.5171040631818842e-07, + "loss": 0.7431, + "step": 9746 + }, + { + "epoch": 0.8790188032646435, + "grad_norm": 1.7108261559892184, + "learning_rate": 1.5148729194466547e-07, + "loss": 0.6751, + "step": 9747 + }, + { + "epoch": 0.8791089867881138, + "grad_norm": 1.601741650493658, + "learning_rate": 1.5126433529643956e-07, + "loss": 0.8128, + "step": 9748 + }, + { + "epoch": 0.8791991703115841, + "grad_norm": 1.4143557586911946, + "learning_rate": 1.5104153639253436e-07, + "loss": 0.7416, + "step": 9749 + }, + { + "epoch": 0.8792893538350544, + "grad_norm": 0.6129256997608706, + "learning_rate": 1.5081889525196002e-07, + "loss": 0.6297, + "step": 9750 + }, + { + "epoch": 0.8793795373585246, + "grad_norm": 2.559025246147193, + "learning_rate": 1.5059641189371398e-07, + "loss": 0.798, + "step": 9751 + }, + { + "epoch": 0.8794697208819948, + "grad_norm": 1.5716538654802552, + "learning_rate": 1.503740863367795e-07, + "loss": 0.7599, + "step": 9752 + }, + { + "epoch": 0.8795599044054652, + "grad_norm": 1.830016828261582, + "learning_rate": 1.50151918600127e-07, + "loss": 0.7739, + "step": 9753 + }, + { + "epoch": 0.8796500879289354, + "grad_norm": 2.1958441582641544, + "learning_rate": 1.4992990870271217e-07, + "loss": 0.7909, + "step": 9754 + }, + { + "epoch": 0.8797402714524056, + "grad_norm": 1.992536686232875, + "learning_rate": 1.497080566634794e-07, + "loss": 0.7496, + "step": 9755 + }, + { + "epoch": 0.8798304549758759, + "grad_norm": 1.5232942171451695, + "learning_rate": 1.4948636250135693e-07, + "loss": 0.7552, + "step": 9756 + }, + { + "epoch": 0.8799206384993462, + "grad_norm": 1.7841202214963534, + "learning_rate": 1.4926482623526249e-07, + "loss": 0.8089, + "step": 9757 + }, + { + "epoch": 0.8800108220228164, + "grad_norm": 1.8257693249921456, + "learning_rate": 1.4904344788409694e-07, + "loss": 0.7461, + "step": 9758 + }, + { + "epoch": 0.8801010055462867, + "grad_norm": 1.572896018324304, + "learning_rate": 1.4882222746675143e-07, + "loss": 0.7095, + "step": 9759 + }, + { + "epoch": 0.8801911890697569, + "grad_norm": 1.6419847212578929, + "learning_rate": 1.4860116500210018e-07, + "loss": 0.7874, + "step": 9760 + }, + { + "epoch": 0.8802813725932273, + "grad_norm": 6.253419779740553, + "learning_rate": 1.4838026050900632e-07, + "loss": 0.815, + "step": 9761 + }, + { + "epoch": 0.8803715561166975, + "grad_norm": 6.160835843042087, + "learning_rate": 1.481595140063181e-07, + "loss": 0.792, + "step": 9762 + }, + { + "epoch": 0.8804617396401677, + "grad_norm": 1.8405275147404017, + "learning_rate": 1.4793892551287136e-07, + "loss": 0.8342, + "step": 9763 + }, + { + "epoch": 0.880551923163638, + "grad_norm": 1.7766249324460661, + "learning_rate": 1.4771849504748768e-07, + "loss": 0.6792, + "step": 9764 + }, + { + "epoch": 0.8806421066871083, + "grad_norm": 1.5015771045431072, + "learning_rate": 1.4749822262897517e-07, + "loss": 0.8217, + "step": 9765 + }, + { + "epoch": 0.8807322902105785, + "grad_norm": 1.889809361586898, + "learning_rate": 1.4727810827612895e-07, + "loss": 0.6904, + "step": 9766 + }, + { + "epoch": 0.8808224737340488, + "grad_norm": 2.2177858003769795, + "learning_rate": 1.470581520077303e-07, + "loss": 0.7534, + "step": 9767 + }, + { + "epoch": 0.8809126572575191, + "grad_norm": 2.0210639848202, + "learning_rate": 1.4683835384254705e-07, + "loss": 0.7433, + "step": 9768 + }, + { + "epoch": 0.8810028407809893, + "grad_norm": 1.6698541995607064, + "learning_rate": 1.4661871379933376e-07, + "loss": 0.8276, + "step": 9769 + }, + { + "epoch": 0.8810930243044596, + "grad_norm": 1.8135838392139223, + "learning_rate": 1.4639923189683169e-07, + "loss": 0.794, + "step": 9770 + }, + { + "epoch": 0.8811832078279298, + "grad_norm": 1.739061981776898, + "learning_rate": 1.461799081537669e-07, + "loss": 0.797, + "step": 9771 + }, + { + "epoch": 0.8812733913514001, + "grad_norm": 1.6202602219861733, + "learning_rate": 1.4596074258885514e-07, + "loss": 0.8117, + "step": 9772 + }, + { + "epoch": 0.8813635748748704, + "grad_norm": 1.7644067122676579, + "learning_rate": 1.4574173522079502e-07, + "loss": 0.7195, + "step": 9773 + }, + { + "epoch": 0.8814537583983406, + "grad_norm": 2.1493615954736245, + "learning_rate": 1.4552288606827513e-07, + "loss": 0.7229, + "step": 9774 + }, + { + "epoch": 0.8815439419218108, + "grad_norm": 1.651619682657408, + "learning_rate": 1.4530419514996761e-07, + "loss": 0.8232, + "step": 9775 + }, + { + "epoch": 0.8816341254452812, + "grad_norm": 0.691620120144549, + "learning_rate": 1.4508566248453291e-07, + "loss": 0.5889, + "step": 9776 + }, + { + "epoch": 0.8817243089687514, + "grad_norm": 1.922744492261669, + "learning_rate": 1.448672880906172e-07, + "loss": 0.7422, + "step": 9777 + }, + { + "epoch": 0.8818144924922217, + "grad_norm": 1.5518276422152006, + "learning_rate": 1.4464907198685382e-07, + "loss": 0.6982, + "step": 9778 + }, + { + "epoch": 0.8819046760156919, + "grad_norm": 1.736261115481428, + "learning_rate": 1.444310141918621e-07, + "loss": 0.7478, + "step": 9779 + }, + { + "epoch": 0.8819948595391622, + "grad_norm": 2.636172813128436, + "learning_rate": 1.4421311472424735e-07, + "loss": 0.7774, + "step": 9780 + }, + { + "epoch": 0.8820850430626325, + "grad_norm": 1.8886605768349052, + "learning_rate": 1.4399537360260273e-07, + "loss": 0.781, + "step": 9781 + }, + { + "epoch": 0.8821752265861027, + "grad_norm": 2.608211052153971, + "learning_rate": 1.4377779084550645e-07, + "loss": 0.77, + "step": 9782 + }, + { + "epoch": 0.8822654101095729, + "grad_norm": 1.722752298112487, + "learning_rate": 1.4356036647152413e-07, + "loss": 0.7377, + "step": 9783 + }, + { + "epoch": 0.8823555936330433, + "grad_norm": 1.4473833249889874, + "learning_rate": 1.4334310049920785e-07, + "loss": 0.7249, + "step": 9784 + }, + { + "epoch": 0.8824457771565135, + "grad_norm": 0.5985411282846871, + "learning_rate": 1.431259929470956e-07, + "loss": 0.5294, + "step": 9785 + }, + { + "epoch": 0.8825359606799837, + "grad_norm": 1.8155390156068119, + "learning_rate": 1.4290904383371237e-07, + "loss": 0.8244, + "step": 9786 + }, + { + "epoch": 0.882626144203454, + "grad_norm": 2.72197462002528, + "learning_rate": 1.4269225317756961e-07, + "loss": 0.7298, + "step": 9787 + }, + { + "epoch": 0.8827163277269243, + "grad_norm": 2.1109981367210504, + "learning_rate": 1.424756209971645e-07, + "loss": 0.7132, + "step": 9788 + }, + { + "epoch": 0.8828065112503946, + "grad_norm": 1.8710838747659875, + "learning_rate": 1.4225914731098199e-07, + "loss": 0.7441, + "step": 9789 + }, + { + "epoch": 0.8828966947738648, + "grad_norm": 1.7346564399633781, + "learning_rate": 1.4204283213749248e-07, + "loss": 0.6801, + "step": 9790 + }, + { + "epoch": 0.8829868782973351, + "grad_norm": 1.6829003412950747, + "learning_rate": 1.4182667549515315e-07, + "loss": 0.803, + "step": 9791 + }, + { + "epoch": 0.8830770618208054, + "grad_norm": 1.8180401416329468, + "learning_rate": 1.4161067740240752e-07, + "loss": 0.7263, + "step": 9792 + }, + { + "epoch": 0.8831672453442756, + "grad_norm": 1.858136317439663, + "learning_rate": 1.4139483787768614e-07, + "loss": 0.7389, + "step": 9793 + }, + { + "epoch": 0.8832574288677458, + "grad_norm": 2.4387995850074518, + "learning_rate": 1.4117915693940584e-07, + "loss": 0.6566, + "step": 9794 + }, + { + "epoch": 0.8833476123912162, + "grad_norm": 1.4107859043201307, + "learning_rate": 1.409636346059684e-07, + "loss": 0.7417, + "step": 9795 + }, + { + "epoch": 0.8834377959146864, + "grad_norm": 2.6960928409279745, + "learning_rate": 1.4074827089576501e-07, + "loss": 0.7803, + "step": 9796 + }, + { + "epoch": 0.8835279794381566, + "grad_norm": 1.83585464615239, + "learning_rate": 1.4053306582717085e-07, + "loss": 0.7594, + "step": 9797 + }, + { + "epoch": 0.8836181629616269, + "grad_norm": 1.4997422473705186, + "learning_rate": 1.4031801941854827e-07, + "loss": 0.7745, + "step": 9798 + }, + { + "epoch": 0.8837083464850972, + "grad_norm": 1.478917983019152, + "learning_rate": 1.401031316882466e-07, + "loss": 0.717, + "step": 9799 + }, + { + "epoch": 0.8837985300085675, + "grad_norm": 1.537817126807648, + "learning_rate": 1.39888402654601e-07, + "loss": 0.762, + "step": 9800 + }, + { + "epoch": 0.8838887135320377, + "grad_norm": 0.6106584677607396, + "learning_rate": 1.3967383233593344e-07, + "loss": 0.5876, + "step": 9801 + }, + { + "epoch": 0.8839788970555079, + "grad_norm": 3.1733437130328195, + "learning_rate": 1.3945942075055218e-07, + "loss": 0.7076, + "step": 9802 + }, + { + "epoch": 0.8840690805789783, + "grad_norm": 1.9322047321131726, + "learning_rate": 1.3924516791675212e-07, + "loss": 0.7667, + "step": 9803 + }, + { + "epoch": 0.8841592641024485, + "grad_norm": 1.4818502537830336, + "learning_rate": 1.3903107385281487e-07, + "loss": 0.6547, + "step": 9804 + }, + { + "epoch": 0.8842494476259187, + "grad_norm": 1.9360153909181792, + "learning_rate": 1.3881713857700717e-07, + "loss": 0.7585, + "step": 9805 + }, + { + "epoch": 0.884339631149389, + "grad_norm": 4.380274835263911, + "learning_rate": 1.3860336210758372e-07, + "loss": 0.7236, + "step": 9806 + }, + { + "epoch": 0.8844298146728593, + "grad_norm": 1.7619858093166574, + "learning_rate": 1.3838974446278506e-07, + "loss": 0.7083, + "step": 9807 + }, + { + "epoch": 0.8845199981963295, + "grad_norm": 1.543945544132454, + "learning_rate": 1.3817628566083817e-07, + "loss": 0.7224, + "step": 9808 + }, + { + "epoch": 0.8846101817197998, + "grad_norm": 1.8106252906385616, + "learning_rate": 1.3796298571995712e-07, + "loss": 0.7616, + "step": 9809 + }, + { + "epoch": 0.88470036524327, + "grad_norm": 4.771665653215772, + "learning_rate": 1.377498446583405e-07, + "loss": 0.647, + "step": 9810 + }, + { + "epoch": 0.8847905487667403, + "grad_norm": 2.154158863616722, + "learning_rate": 1.3753686249417596e-07, + "loss": 0.7212, + "step": 9811 + }, + { + "epoch": 0.8848807322902106, + "grad_norm": 48.10991545246936, + "learning_rate": 1.373240392456354e-07, + "loss": 0.7017, + "step": 9812 + }, + { + "epoch": 0.8849709158136808, + "grad_norm": 1.6888980725454192, + "learning_rate": 1.37111374930879e-07, + "loss": 0.6827, + "step": 9813 + }, + { + "epoch": 0.885061099337151, + "grad_norm": 1.5222381565825482, + "learning_rate": 1.3689886956805176e-07, + "loss": 0.6727, + "step": 9814 + }, + { + "epoch": 0.8851512828606214, + "grad_norm": 1.467738351002792, + "learning_rate": 1.3668652317528585e-07, + "loss": 0.7374, + "step": 9815 + }, + { + "epoch": 0.8852414663840916, + "grad_norm": 1.5774552841015317, + "learning_rate": 1.3647433577070012e-07, + "loss": 0.7653, + "step": 9816 + }, + { + "epoch": 0.8853316499075619, + "grad_norm": 0.7203149758252937, + "learning_rate": 1.3626230737239942e-07, + "loss": 0.5495, + "step": 9817 + }, + { + "epoch": 0.8854218334310322, + "grad_norm": 0.7035542767130476, + "learning_rate": 1.3605043799847527e-07, + "loss": 0.557, + "step": 9818 + }, + { + "epoch": 0.8855120169545024, + "grad_norm": 1.4712089084276707, + "learning_rate": 1.3583872766700567e-07, + "loss": 0.6825, + "step": 9819 + }, + { + "epoch": 0.8856022004779727, + "grad_norm": 1.5835429186724652, + "learning_rate": 1.3562717639605437e-07, + "loss": 0.7282, + "step": 9820 + }, + { + "epoch": 0.8856923840014429, + "grad_norm": 2.987954594329686, + "learning_rate": 1.3541578420367229e-07, + "loss": 0.7162, + "step": 9821 + }, + { + "epoch": 0.8857825675249132, + "grad_norm": 1.7401541574744213, + "learning_rate": 1.3520455110789697e-07, + "loss": 0.7624, + "step": 9822 + }, + { + "epoch": 0.8858727510483835, + "grad_norm": 1.4851112547534482, + "learning_rate": 1.3499347712675158e-07, + "loss": 0.7289, + "step": 9823 + }, + { + "epoch": 0.8859629345718537, + "grad_norm": 1.5313821073882035, + "learning_rate": 1.3478256227824635e-07, + "loss": 0.7082, + "step": 9824 + }, + { + "epoch": 0.8860531180953239, + "grad_norm": 1.845136993615249, + "learning_rate": 1.3457180658037759e-07, + "loss": 0.6553, + "step": 9825 + }, + { + "epoch": 0.8861433016187943, + "grad_norm": 2.2747750558836812, + "learning_rate": 1.3436121005112843e-07, + "loss": 0.7779, + "step": 9826 + }, + { + "epoch": 0.8862334851422645, + "grad_norm": 1.663428172247141, + "learning_rate": 1.3415077270846719e-07, + "loss": 0.7142, + "step": 9827 + }, + { + "epoch": 0.8863236686657348, + "grad_norm": 2.283289689902251, + "learning_rate": 1.3394049457035105e-07, + "loss": 0.8599, + "step": 9828 + }, + { + "epoch": 0.886413852189205, + "grad_norm": 1.5920854739817278, + "learning_rate": 1.3373037565472034e-07, + "loss": 0.7572, + "step": 9829 + }, + { + "epoch": 0.8865040357126753, + "grad_norm": 1.471779559821419, + "learning_rate": 1.3352041597950537e-07, + "loss": 0.7264, + "step": 9830 + }, + { + "epoch": 0.8865942192361456, + "grad_norm": 2.3937341754280705, + "learning_rate": 1.333106155626196e-07, + "loss": 0.7379, + "step": 9831 + }, + { + "epoch": 0.8866844027596158, + "grad_norm": 1.6694091723681728, + "learning_rate": 1.331009744219651e-07, + "loss": 0.6841, + "step": 9832 + }, + { + "epoch": 0.886774586283086, + "grad_norm": 0.672835639134027, + "learning_rate": 1.3289149257542943e-07, + "loss": 0.6125, + "step": 9833 + }, + { + "epoch": 0.8868647698065564, + "grad_norm": 1.6340265460571943, + "learning_rate": 1.3268217004088666e-07, + "loss": 0.6474, + "step": 9834 + }, + { + "epoch": 0.8869549533300266, + "grad_norm": 1.6358968997497116, + "learning_rate": 1.3247300683619788e-07, + "loss": 0.636, + "step": 9835 + }, + { + "epoch": 0.8870451368534968, + "grad_norm": 2.712699647227484, + "learning_rate": 1.3226400297920903e-07, + "loss": 0.7215, + "step": 9836 + }, + { + "epoch": 0.8871353203769671, + "grad_norm": 1.7267698275680208, + "learning_rate": 1.3205515848775428e-07, + "loss": 0.7287, + "step": 9837 + }, + { + "epoch": 0.8872255039004374, + "grad_norm": 1.680802370062659, + "learning_rate": 1.3184647337965316e-07, + "loss": 0.6843, + "step": 9838 + }, + { + "epoch": 0.8873156874239077, + "grad_norm": 1.4381644330510055, + "learning_rate": 1.3163794767271163e-07, + "loss": 0.6942, + "step": 9839 + }, + { + "epoch": 0.8874058709473779, + "grad_norm": 1.503884293036752, + "learning_rate": 1.314295813847226e-07, + "loss": 0.7154, + "step": 9840 + }, + { + "epoch": 0.8874960544708482, + "grad_norm": 1.8145945055588844, + "learning_rate": 1.3122137453346515e-07, + "loss": 0.8437, + "step": 9841 + }, + { + "epoch": 0.8875862379943185, + "grad_norm": 2.826412209162835, + "learning_rate": 1.3101332713670376e-07, + "loss": 0.8053, + "step": 9842 + }, + { + "epoch": 0.8876764215177887, + "grad_norm": 0.6626496937613404, + "learning_rate": 1.3080543921219133e-07, + "loss": 0.6065, + "step": 9843 + }, + { + "epoch": 0.8877666050412589, + "grad_norm": 1.69507607775973, + "learning_rate": 1.3059771077766478e-07, + "loss": 0.7379, + "step": 9844 + }, + { + "epoch": 0.8878567885647293, + "grad_norm": 2.6673285636404738, + "learning_rate": 1.3039014185085018e-07, + "loss": 0.7846, + "step": 9845 + }, + { + "epoch": 0.8879469720881995, + "grad_norm": 2.1825897692978637, + "learning_rate": 1.301827324494571e-07, + "loss": 0.7545, + "step": 9846 + }, + { + "epoch": 0.8880371556116697, + "grad_norm": 2.754255037337585, + "learning_rate": 1.2997548259118342e-07, + "loss": 0.7831, + "step": 9847 + }, + { + "epoch": 0.88812733913514, + "grad_norm": 1.8050525510751603, + "learning_rate": 1.2976839229371272e-07, + "loss": 0.7775, + "step": 9848 + }, + { + "epoch": 0.8882175226586103, + "grad_norm": 0.6247232062634187, + "learning_rate": 1.2956146157471515e-07, + "loss": 0.5957, + "step": 9849 + }, + { + "epoch": 0.8883077061820805, + "grad_norm": 1.577280555128022, + "learning_rate": 1.2935469045184745e-07, + "loss": 0.7633, + "step": 9850 + }, + { + "epoch": 0.8883978897055508, + "grad_norm": 2.5261337945137474, + "learning_rate": 1.291480789427517e-07, + "loss": 0.7209, + "step": 9851 + }, + { + "epoch": 0.888488073229021, + "grad_norm": 1.6150486757483877, + "learning_rate": 1.2894162706505807e-07, + "loss": 0.784, + "step": 9852 + }, + { + "epoch": 0.8885782567524914, + "grad_norm": 1.8199715356382493, + "learning_rate": 1.2873533483638155e-07, + "loss": 0.7397, + "step": 9853 + }, + { + "epoch": 0.8886684402759616, + "grad_norm": 1.7046509652283028, + "learning_rate": 1.285292022743243e-07, + "loss": 0.6918, + "step": 9854 + }, + { + "epoch": 0.8887586237994318, + "grad_norm": 1.5600860258982916, + "learning_rate": 1.2832322939647467e-07, + "loss": 0.6935, + "step": 9855 + }, + { + "epoch": 0.8888488073229021, + "grad_norm": 1.90782856852445, + "learning_rate": 1.281174162204075e-07, + "loss": 0.7284, + "step": 9856 + }, + { + "epoch": 0.8889389908463724, + "grad_norm": 1.9801798051053028, + "learning_rate": 1.2791176276368366e-07, + "loss": 0.6228, + "step": 9857 + }, + { + "epoch": 0.8890291743698426, + "grad_norm": 1.720775381902812, + "learning_rate": 1.2770626904385128e-07, + "loss": 0.7555, + "step": 9858 + }, + { + "epoch": 0.8891193578933129, + "grad_norm": 1.6964026958241774, + "learning_rate": 1.2750093507844306e-07, + "loss": 0.6061, + "step": 9859 + }, + { + "epoch": 0.8892095414167831, + "grad_norm": 1.4190966951357697, + "learning_rate": 1.272957608849805e-07, + "loss": 0.7811, + "step": 9860 + }, + { + "epoch": 0.8892997249402534, + "grad_norm": 1.4722086178053062, + "learning_rate": 1.270907464809694e-07, + "loss": 0.7187, + "step": 9861 + }, + { + "epoch": 0.8893899084637237, + "grad_norm": 1.3992353752507434, + "learning_rate": 1.2688589188390285e-07, + "loss": 0.7452, + "step": 9862 + }, + { + "epoch": 0.8894800919871939, + "grad_norm": 1.8540147726299612, + "learning_rate": 1.2668119711126023e-07, + "loss": 0.7563, + "step": 9863 + }, + { + "epoch": 0.8895702755106643, + "grad_norm": 4.7638754713180305, + "learning_rate": 1.2647666218050735e-07, + "loss": 0.7452, + "step": 9864 + }, + { + "epoch": 0.8896604590341345, + "grad_norm": 2.079273451494557, + "learning_rate": 1.2627228710909643e-07, + "loss": 0.7059, + "step": 9865 + }, + { + "epoch": 0.8897506425576047, + "grad_norm": 1.5217622593786502, + "learning_rate": 1.260680719144649e-07, + "loss": 0.7113, + "step": 9866 + }, + { + "epoch": 0.889840826081075, + "grad_norm": 3.571615903573436, + "learning_rate": 1.2586401661403877e-07, + "loss": 0.675, + "step": 9867 + }, + { + "epoch": 0.8899310096045453, + "grad_norm": 1.8392330379059119, + "learning_rate": 1.2566012122522817e-07, + "loss": 0.7218, + "step": 9868 + }, + { + "epoch": 0.8900211931280155, + "grad_norm": 1.872703226365218, + "learning_rate": 1.254563857654316e-07, + "loss": 0.7164, + "step": 9869 + }, + { + "epoch": 0.8901113766514858, + "grad_norm": 1.946761361566733, + "learning_rate": 1.2525281025203205e-07, + "loss": 0.6339, + "step": 9870 + }, + { + "epoch": 0.890201560174956, + "grad_norm": 4.017218010340308, + "learning_rate": 1.2504939470240006e-07, + "loss": 0.7306, + "step": 9871 + }, + { + "epoch": 0.8902917436984263, + "grad_norm": 1.684928630370117, + "learning_rate": 1.2484613913389196e-07, + "loss": 0.81, + "step": 9872 + }, + { + "epoch": 0.8903819272218966, + "grad_norm": 1.7403530161753518, + "learning_rate": 1.2464304356385057e-07, + "loss": 0.6587, + "step": 9873 + }, + { + "epoch": 0.8904721107453668, + "grad_norm": 2.4116931139118214, + "learning_rate": 1.2444010800960558e-07, + "loss": 0.8036, + "step": 9874 + }, + { + "epoch": 0.890562294268837, + "grad_norm": 1.565870769040371, + "learning_rate": 1.2423733248847267e-07, + "loss": 0.7296, + "step": 9875 + }, + { + "epoch": 0.8906524777923074, + "grad_norm": 2.592749161256407, + "learning_rate": 1.2403471701775293e-07, + "loss": 0.6867, + "step": 9876 + }, + { + "epoch": 0.8907426613157776, + "grad_norm": 1.7156256281394713, + "learning_rate": 1.2383226161473515e-07, + "loss": 0.7879, + "step": 9877 + }, + { + "epoch": 0.8908328448392479, + "grad_norm": 7.60203796969764, + "learning_rate": 1.2362996629669376e-07, + "loss": 0.8206, + "step": 9878 + }, + { + "epoch": 0.8909230283627181, + "grad_norm": 1.4700075001427966, + "learning_rate": 1.2342783108089007e-07, + "loss": 0.7544, + "step": 9879 + }, + { + "epoch": 0.8910132118861884, + "grad_norm": 1.8300344814116256, + "learning_rate": 1.2322585598457135e-07, + "loss": 0.8013, + "step": 9880 + }, + { + "epoch": 0.8911033954096587, + "grad_norm": 1.3368098662115617, + "learning_rate": 1.2302404102497034e-07, + "loss": 0.7188, + "step": 9881 + }, + { + "epoch": 0.8911935789331289, + "grad_norm": 5.576377787353985, + "learning_rate": 1.228223862193083e-07, + "loss": 0.6749, + "step": 9882 + }, + { + "epoch": 0.8912837624565991, + "grad_norm": 2.340826525498899, + "learning_rate": 1.2262089158479038e-07, + "loss": 0.7545, + "step": 9883 + }, + { + "epoch": 0.8913739459800695, + "grad_norm": 1.8184840633857207, + "learning_rate": 1.2241955713861042e-07, + "loss": 0.7905, + "step": 9884 + }, + { + "epoch": 0.8914641295035397, + "grad_norm": 2.2048387287702593, + "learning_rate": 1.222183828979464e-07, + "loss": 0.6879, + "step": 9885 + }, + { + "epoch": 0.8915543130270099, + "grad_norm": 1.4477015365211938, + "learning_rate": 1.2201736887996372e-07, + "loss": 0.741, + "step": 9886 + }, + { + "epoch": 0.8916444965504803, + "grad_norm": 1.819033128127406, + "learning_rate": 1.2181651510181444e-07, + "loss": 0.8218, + "step": 9887 + }, + { + "epoch": 0.8917346800739505, + "grad_norm": 1.6691827934325334, + "learning_rate": 1.2161582158063622e-07, + "loss": 0.7745, + "step": 9888 + }, + { + "epoch": 0.8918248635974207, + "grad_norm": 2.064982682864383, + "learning_rate": 1.214152883335533e-07, + "loss": 0.6486, + "step": 9889 + }, + { + "epoch": 0.891915047120891, + "grad_norm": 1.8783714500876536, + "learning_rate": 1.2121491537767648e-07, + "loss": 0.6922, + "step": 9890 + }, + { + "epoch": 0.8920052306443613, + "grad_norm": 1.8737356373425418, + "learning_rate": 1.2101470273010294e-07, + "loss": 0.7755, + "step": 9891 + }, + { + "epoch": 0.8920954141678316, + "grad_norm": 1.781550264983963, + "learning_rate": 1.2081465040791528e-07, + "loss": 0.7222, + "step": 9892 + }, + { + "epoch": 0.8921855976913018, + "grad_norm": 1.6157014072435458, + "learning_rate": 1.2061475842818335e-07, + "loss": 0.7098, + "step": 9893 + }, + { + "epoch": 0.892275781214772, + "grad_norm": 2.1117253525348825, + "learning_rate": 1.2041502680796313e-07, + "loss": 0.7216, + "step": 9894 + }, + { + "epoch": 0.8923659647382424, + "grad_norm": 1.5456963875184853, + "learning_rate": 1.2021545556429648e-07, + "loss": 0.7484, + "step": 9895 + }, + { + "epoch": 0.8924561482617126, + "grad_norm": 1.690977539987142, + "learning_rate": 1.2001604471421245e-07, + "loss": 0.7437, + "step": 9896 + }, + { + "epoch": 0.8925463317851828, + "grad_norm": 1.4310365576029729, + "learning_rate": 1.1981679427472567e-07, + "loss": 0.7947, + "step": 9897 + }, + { + "epoch": 0.8926365153086531, + "grad_norm": 1.822018981792843, + "learning_rate": 1.196177042628368e-07, + "loss": 0.7972, + "step": 9898 + }, + { + "epoch": 0.8927266988321234, + "grad_norm": 2.1254329240400662, + "learning_rate": 1.194187746955344e-07, + "loss": 0.7454, + "step": 9899 + }, + { + "epoch": 0.8928168823555936, + "grad_norm": 3.503640869705694, + "learning_rate": 1.1922000558979094e-07, + "loss": 0.7201, + "step": 9900 + }, + { + "epoch": 0.8929070658790639, + "grad_norm": 1.7198040591124042, + "learning_rate": 1.1902139696256752e-07, + "loss": 0.7763, + "step": 9901 + }, + { + "epoch": 0.8929972494025341, + "grad_norm": 1.9341646218717066, + "learning_rate": 1.188229488308099e-07, + "loss": 0.6788, + "step": 9902 + }, + { + "epoch": 0.8930874329260045, + "grad_norm": 0.5679382127111091, + "learning_rate": 1.1862466121145098e-07, + "loss": 0.5573, + "step": 9903 + }, + { + "epoch": 0.8931776164494747, + "grad_norm": 2.0262896493071776, + "learning_rate": 1.184265341214099e-07, + "loss": 0.6917, + "step": 9904 + }, + { + "epoch": 0.8932677999729449, + "grad_norm": 1.616067323325708, + "learning_rate": 1.182285675775918e-07, + "loss": 0.7673, + "step": 9905 + }, + { + "epoch": 0.8933579834964152, + "grad_norm": 1.9262397019236646, + "learning_rate": 1.1803076159688851e-07, + "loss": 0.7307, + "step": 9906 + }, + { + "epoch": 0.8934481670198855, + "grad_norm": 1.6032861134137768, + "learning_rate": 1.1783311619617741e-07, + "loss": 0.7785, + "step": 9907 + }, + { + "epoch": 0.8935383505433557, + "grad_norm": 1.6700069651201472, + "learning_rate": 1.1763563139232302e-07, + "loss": 0.7153, + "step": 9908 + }, + { + "epoch": 0.893628534066826, + "grad_norm": 2.6884124448913034, + "learning_rate": 1.1743830720217562e-07, + "loss": 0.6856, + "step": 9909 + }, + { + "epoch": 0.8937187175902963, + "grad_norm": 1.9776535942913946, + "learning_rate": 1.1724114364257243e-07, + "loss": 0.7256, + "step": 9910 + }, + { + "epoch": 0.8938089011137665, + "grad_norm": 2.5834990007336933, + "learning_rate": 1.1704414073033619e-07, + "loss": 0.7529, + "step": 9911 + }, + { + "epoch": 0.8938990846372368, + "grad_norm": 1.8933179608020942, + "learning_rate": 1.1684729848227636e-07, + "loss": 0.7856, + "step": 9912 + }, + { + "epoch": 0.893989268160707, + "grad_norm": 1.9084496236403257, + "learning_rate": 1.1665061691518884e-07, + "loss": 0.8034, + "step": 9913 + }, + { + "epoch": 0.8940794516841774, + "grad_norm": 1.684708440894968, + "learning_rate": 1.1645409604585532e-07, + "loss": 0.743, + "step": 9914 + }, + { + "epoch": 0.8941696352076476, + "grad_norm": 1.620282514804178, + "learning_rate": 1.162577358910437e-07, + "loss": 0.754, + "step": 9915 + }, + { + "epoch": 0.8942598187311178, + "grad_norm": 2.594002033535786, + "learning_rate": 1.160615364675095e-07, + "loss": 0.7221, + "step": 9916 + }, + { + "epoch": 0.894350002254588, + "grad_norm": 0.611250875546505, + "learning_rate": 1.1586549779199262e-07, + "loss": 0.5596, + "step": 9917 + }, + { + "epoch": 0.8944401857780584, + "grad_norm": 1.9325299043303266, + "learning_rate": 1.1566961988122037e-07, + "loss": 0.8039, + "step": 9918 + }, + { + "epoch": 0.8945303693015286, + "grad_norm": 2.076803260319565, + "learning_rate": 1.1547390275190627e-07, + "loss": 0.7133, + "step": 9919 + }, + { + "epoch": 0.8946205528249989, + "grad_norm": 1.7855284321662241, + "learning_rate": 1.1527834642075007e-07, + "loss": 0.6857, + "step": 9920 + }, + { + "epoch": 0.8947107363484691, + "grad_norm": 5.7395538310857965, + "learning_rate": 1.1508295090443797e-07, + "loss": 0.7299, + "step": 9921 + }, + { + "epoch": 0.8948009198719394, + "grad_norm": 1.432185174855645, + "learning_rate": 1.148877162196411e-07, + "loss": 0.6966, + "step": 9922 + }, + { + "epoch": 0.8948911033954097, + "grad_norm": 2.3939004039076783, + "learning_rate": 1.1469264238301924e-07, + "loss": 0.6785, + "step": 9923 + }, + { + "epoch": 0.8949812869188799, + "grad_norm": 1.7632296357314134, + "learning_rate": 1.1449772941121638e-07, + "loss": 0.7689, + "step": 9924 + }, + { + "epoch": 0.8950714704423501, + "grad_norm": 1.8639319861000194, + "learning_rate": 1.1430297732086369e-07, + "loss": 0.717, + "step": 9925 + }, + { + "epoch": 0.8951616539658205, + "grad_norm": 1.802025884105804, + "learning_rate": 1.1410838612857876e-07, + "loss": 0.7359, + "step": 9926 + }, + { + "epoch": 0.8952518374892907, + "grad_norm": 1.8508241364263303, + "learning_rate": 1.1391395585096497e-07, + "loss": 0.8417, + "step": 9927 + }, + { + "epoch": 0.895342021012761, + "grad_norm": 1.8986451463207492, + "learning_rate": 1.1371968650461216e-07, + "loss": 0.6751, + "step": 9928 + }, + { + "epoch": 0.8954322045362312, + "grad_norm": 0.7220418018989085, + "learning_rate": 1.1352557810609687e-07, + "loss": 0.5721, + "step": 9929 + }, + { + "epoch": 0.8955223880597015, + "grad_norm": 17.610727578100743, + "learning_rate": 1.1333163067198048e-07, + "loss": 0.7304, + "step": 9930 + }, + { + "epoch": 0.8956125715831718, + "grad_norm": 1.4886595077621543, + "learning_rate": 1.1313784421881311e-07, + "loss": 0.8081, + "step": 9931 + }, + { + "epoch": 0.895702755106642, + "grad_norm": 2.3725692959183142, + "learning_rate": 1.1294421876312865e-07, + "loss": 0.7772, + "step": 9932 + }, + { + "epoch": 0.8957929386301122, + "grad_norm": 1.8549463440160383, + "learning_rate": 1.1275075432144831e-07, + "loss": 0.7162, + "step": 9933 + }, + { + "epoch": 0.8958831221535826, + "grad_norm": 2.4487545662604187, + "learning_rate": 1.1255745091028002e-07, + "loss": 0.7597, + "step": 9934 + }, + { + "epoch": 0.8959733056770528, + "grad_norm": 1.9496339650056527, + "learning_rate": 1.1236430854611723e-07, + "loss": 0.7531, + "step": 9935 + }, + { + "epoch": 0.896063489200523, + "grad_norm": 27.04595991555682, + "learning_rate": 1.1217132724544032e-07, + "loss": 0.7073, + "step": 9936 + }, + { + "epoch": 0.8961536727239934, + "grad_norm": 3.7835335889968498, + "learning_rate": 1.1197850702471434e-07, + "loss": 0.7918, + "step": 9937 + }, + { + "epoch": 0.8962438562474636, + "grad_norm": 1.8028678803278533, + "learning_rate": 1.1178584790039348e-07, + "loss": 0.8098, + "step": 9938 + }, + { + "epoch": 0.8963340397709338, + "grad_norm": 1.8014619082897765, + "learning_rate": 1.1159334988891478e-07, + "loss": 0.6877, + "step": 9939 + }, + { + "epoch": 0.8964242232944041, + "grad_norm": 1.9431669973875967, + "learning_rate": 1.1140101300670446e-07, + "loss": 0.7213, + "step": 9940 + }, + { + "epoch": 0.8965144068178744, + "grad_norm": 1.5681113383069398, + "learning_rate": 1.1120883727017338e-07, + "loss": 0.6048, + "step": 9941 + }, + { + "epoch": 0.8966045903413447, + "grad_norm": 1.6141507297268083, + "learning_rate": 1.1101682269571889e-07, + "loss": 0.7968, + "step": 9942 + }, + { + "epoch": 0.8966947738648149, + "grad_norm": 2.126449875718704, + "learning_rate": 1.1082496929972473e-07, + "loss": 0.7226, + "step": 9943 + }, + { + "epoch": 0.8967849573882851, + "grad_norm": 2.03151126061712, + "learning_rate": 1.1063327709856096e-07, + "loss": 0.7176, + "step": 9944 + }, + { + "epoch": 0.8968751409117555, + "grad_norm": 1.8867163490295928, + "learning_rate": 1.1044174610858403e-07, + "loss": 0.6283, + "step": 9945 + }, + { + "epoch": 0.8969653244352257, + "grad_norm": 1.8198867103403746, + "learning_rate": 1.1025037634613643e-07, + "loss": 0.699, + "step": 9946 + }, + { + "epoch": 0.8970555079586959, + "grad_norm": 3.4500276647370773, + "learning_rate": 1.1005916782754643e-07, + "loss": 0.7676, + "step": 9947 + }, + { + "epoch": 0.8971456914821662, + "grad_norm": 7.585731071389967, + "learning_rate": 1.0986812056912919e-07, + "loss": 0.7291, + "step": 9948 + }, + { + "epoch": 0.8972358750056365, + "grad_norm": 1.8741163078701397, + "learning_rate": 1.0967723458718613e-07, + "loss": 0.6245, + "step": 9949 + }, + { + "epoch": 0.8973260585291067, + "grad_norm": 3.0133231277067942, + "learning_rate": 1.0948650989800445e-07, + "loss": 0.7108, + "step": 9950 + }, + { + "epoch": 0.897416242052577, + "grad_norm": 3.011042300005411, + "learning_rate": 1.0929594651785823e-07, + "loss": 0.7267, + "step": 9951 + }, + { + "epoch": 0.8975064255760472, + "grad_norm": 1.5638617113250255, + "learning_rate": 1.0910554446300646e-07, + "loss": 0.7934, + "step": 9952 + }, + { + "epoch": 0.8975966090995176, + "grad_norm": 1.7386769048051967, + "learning_rate": 1.089153037496966e-07, + "loss": 0.642, + "step": 9953 + }, + { + "epoch": 0.8976867926229878, + "grad_norm": 2.935306785445538, + "learning_rate": 1.0872522439415966e-07, + "loss": 0.7655, + "step": 9954 + }, + { + "epoch": 0.897776976146458, + "grad_norm": 1.9409930827643016, + "learning_rate": 1.0853530641261554e-07, + "loss": 0.6957, + "step": 9955 + }, + { + "epoch": 0.8978671596699282, + "grad_norm": 1.5062523711253137, + "learning_rate": 1.083455498212682e-07, + "loss": 0.7565, + "step": 9956 + }, + { + "epoch": 0.8979573431933986, + "grad_norm": 1.7100825154880195, + "learning_rate": 1.0815595463630911e-07, + "loss": 0.7787, + "step": 9957 + }, + { + "epoch": 0.8980475267168688, + "grad_norm": 2.0788886022553394, + "learning_rate": 1.0796652087391556e-07, + "loss": 0.6139, + "step": 9958 + }, + { + "epoch": 0.8981377102403391, + "grad_norm": 1.5259280460837434, + "learning_rate": 1.0777724855025083e-07, + "loss": 0.84, + "step": 9959 + }, + { + "epoch": 0.8982278937638094, + "grad_norm": 1.9060359271807379, + "learning_rate": 1.075881376814649e-07, + "loss": 0.7449, + "step": 9960 + }, + { + "epoch": 0.8983180772872796, + "grad_norm": 1.802989144058255, + "learning_rate": 1.0739918828369377e-07, + "loss": 0.7975, + "step": 9961 + }, + { + "epoch": 0.8984082608107499, + "grad_norm": 1.6083602669129216, + "learning_rate": 1.0721040037305983e-07, + "loss": 0.7501, + "step": 9962 + }, + { + "epoch": 0.8984984443342201, + "grad_norm": 1.8286092393980697, + "learning_rate": 1.0702177396567114e-07, + "loss": 0.6924, + "step": 9963 + }, + { + "epoch": 0.8985886278576904, + "grad_norm": 0.663706897442434, + "learning_rate": 1.0683330907762233e-07, + "loss": 0.5632, + "step": 9964 + }, + { + "epoch": 0.8986788113811607, + "grad_norm": 1.4315657654637381, + "learning_rate": 1.0664500572499435e-07, + "loss": 0.6738, + "step": 9965 + }, + { + "epoch": 0.8987689949046309, + "grad_norm": 1.813518435354795, + "learning_rate": 1.0645686392385455e-07, + "loss": 0.6843, + "step": 9966 + }, + { + "epoch": 0.8988591784281011, + "grad_norm": 1.4687139628433028, + "learning_rate": 1.0626888369025588e-07, + "loss": 0.6668, + "step": 9967 + }, + { + "epoch": 0.8989493619515715, + "grad_norm": 1.6484168896584401, + "learning_rate": 1.0608106504023817e-07, + "loss": 0.7675, + "step": 9968 + }, + { + "epoch": 0.8990395454750417, + "grad_norm": 1.4149199357313578, + "learning_rate": 1.0589340798982637e-07, + "loss": 0.7334, + "step": 9969 + }, + { + "epoch": 0.899129728998512, + "grad_norm": 2.5004429424803627, + "learning_rate": 1.057059125550337e-07, + "loss": 0.6841, + "step": 9970 + }, + { + "epoch": 0.8992199125219822, + "grad_norm": 1.4252784155860303, + "learning_rate": 1.0551857875185732e-07, + "loss": 0.7513, + "step": 9971 + }, + { + "epoch": 0.8993100960454525, + "grad_norm": 1.7472617661220102, + "learning_rate": 1.0533140659628181e-07, + "loss": 0.6255, + "step": 9972 + }, + { + "epoch": 0.8994002795689228, + "grad_norm": 1.8957676579205014, + "learning_rate": 1.0514439610427772e-07, + "loss": 0.8005, + "step": 9973 + }, + { + "epoch": 0.899490463092393, + "grad_norm": 4.016823675060114, + "learning_rate": 1.0495754729180206e-07, + "loss": 0.7687, + "step": 9974 + }, + { + "epoch": 0.8995806466158632, + "grad_norm": 1.536868783039241, + "learning_rate": 1.0477086017479741e-07, + "loss": 0.8419, + "step": 9975 + }, + { + "epoch": 0.8996708301393336, + "grad_norm": 2.086547419499394, + "learning_rate": 1.0458433476919327e-07, + "loss": 0.6991, + "step": 9976 + }, + { + "epoch": 0.8997610136628038, + "grad_norm": 1.58090674855082, + "learning_rate": 1.0439797109090509e-07, + "loss": 0.7696, + "step": 9977 + }, + { + "epoch": 0.899851197186274, + "grad_norm": 2.287154929393946, + "learning_rate": 1.0421176915583396e-07, + "loss": 0.7165, + "step": 9978 + }, + { + "epoch": 0.8999413807097443, + "grad_norm": 1.7102736487245918, + "learning_rate": 1.0402572897986828e-07, + "loss": 0.774, + "step": 9979 + }, + { + "epoch": 0.9000315642332146, + "grad_norm": 0.5891380888911254, + "learning_rate": 1.0383985057888134e-07, + "loss": 0.5682, + "step": 9980 + }, + { + "epoch": 0.9001217477566849, + "grad_norm": 2.828961833417277, + "learning_rate": 1.036541339687338e-07, + "loss": 0.647, + "step": 9981 + }, + { + "epoch": 0.9002119312801551, + "grad_norm": 0.651134375011307, + "learning_rate": 1.0346857916527186e-07, + "loss": 0.5384, + "step": 9982 + }, + { + "epoch": 0.9003021148036254, + "grad_norm": 1.9460178919417586, + "learning_rate": 1.0328318618432819e-07, + "loss": 0.7407, + "step": 9983 + }, + { + "epoch": 0.9003922983270957, + "grad_norm": 1.5820173098103032, + "learning_rate": 1.0309795504172148e-07, + "loss": 0.804, + "step": 9984 + }, + { + "epoch": 0.9004824818505659, + "grad_norm": 2.014003722618322, + "learning_rate": 1.0291288575325685e-07, + "loss": 0.7192, + "step": 9985 + }, + { + "epoch": 0.9005726653740361, + "grad_norm": 1.6812852933788285, + "learning_rate": 1.0272797833472502e-07, + "loss": 0.7064, + "step": 9986 + }, + { + "epoch": 0.9006628488975065, + "grad_norm": 0.6703743254775353, + "learning_rate": 1.0254323280190335e-07, + "loss": 0.548, + "step": 9987 + }, + { + "epoch": 0.9007530324209767, + "grad_norm": 2.1997268117000486, + "learning_rate": 1.023586491705557e-07, + "loss": 0.7206, + "step": 9988 + }, + { + "epoch": 0.9008432159444469, + "grad_norm": 2.722860539350877, + "learning_rate": 1.0217422745643145e-07, + "loss": 0.6826, + "step": 9989 + }, + { + "epoch": 0.9009333994679172, + "grad_norm": 1.9743937177456121, + "learning_rate": 1.0198996767526691e-07, + "loss": 0.6701, + "step": 9990 + }, + { + "epoch": 0.9010235829913875, + "grad_norm": 2.0428173818729793, + "learning_rate": 1.018058698427835e-07, + "loss": 0.7209, + "step": 9991 + }, + { + "epoch": 0.9011137665148578, + "grad_norm": 1.66965579762992, + "learning_rate": 1.0162193397469021e-07, + "loss": 0.7883, + "step": 9992 + }, + { + "epoch": 0.901203950038328, + "grad_norm": 1.724014888269877, + "learning_rate": 1.0143816008668049e-07, + "loss": 0.6773, + "step": 9993 + }, + { + "epoch": 0.9012941335617982, + "grad_norm": 0.6407301514089334, + "learning_rate": 1.0125454819443624e-07, + "loss": 0.5866, + "step": 9994 + }, + { + "epoch": 0.9013843170852686, + "grad_norm": 2.367221139599571, + "learning_rate": 1.0107109831362315e-07, + "loss": 0.6518, + "step": 9995 + }, + { + "epoch": 0.9014745006087388, + "grad_norm": 1.5064915222020583, + "learning_rate": 1.0088781045989447e-07, + "loss": 0.7473, + "step": 9996 + }, + { + "epoch": 0.901564684132209, + "grad_norm": 5.905207789267354, + "learning_rate": 1.0070468464888926e-07, + "loss": 0.7937, + "step": 9997 + }, + { + "epoch": 0.9016548676556793, + "grad_norm": 2.765027083942011, + "learning_rate": 1.0052172089623324e-07, + "loss": 0.7611, + "step": 9998 + }, + { + "epoch": 0.9017450511791496, + "grad_norm": 1.843657022491779, + "learning_rate": 1.0033891921753746e-07, + "loss": 0.7776, + "step": 9999 + }, + { + "epoch": 0.9018352347026198, + "grad_norm": 3.6849747109030635, + "learning_rate": 1.0015627962839968e-07, + "loss": 0.6554, + "step": 10000 + }, + { + "epoch": 0.9019254182260901, + "grad_norm": 3.747951456657351, + "learning_rate": 9.99738021444041e-08, + "loss": 0.7146, + "step": 10001 + }, + { + "epoch": 0.9020156017495603, + "grad_norm": 1.7298476781654912, + "learning_rate": 9.979148678112003e-08, + "loss": 0.6728, + "step": 10002 + }, + { + "epoch": 0.9021057852730306, + "grad_norm": 1.8457460105798114, + "learning_rate": 9.960933355410417e-08, + "loss": 0.8424, + "step": 10003 + }, + { + "epoch": 0.9021959687965009, + "grad_norm": 1.5615499278309242, + "learning_rate": 9.942734247889828e-08, + "loss": 0.6561, + "step": 10004 + }, + { + "epoch": 0.9022861523199711, + "grad_norm": 2.9696212293264783, + "learning_rate": 9.92455135710315e-08, + "loss": 0.748, + "step": 10005 + }, + { + "epoch": 0.9023763358434415, + "grad_norm": 2.908293467105427, + "learning_rate": 9.906384684601787e-08, + "loss": 0.6741, + "step": 10006 + }, + { + "epoch": 0.9024665193669117, + "grad_norm": 1.6428339302010562, + "learning_rate": 9.8882342319359e-08, + "loss": 0.7827, + "step": 10007 + }, + { + "epoch": 0.9025567028903819, + "grad_norm": 1.3607239880911421, + "learning_rate": 9.870100000654048e-08, + "loss": 0.6894, + "step": 10008 + }, + { + "epoch": 0.9026468864138522, + "grad_norm": 1.388098103255375, + "learning_rate": 9.851981992303704e-08, + "loss": 0.7136, + "step": 10009 + }, + { + "epoch": 0.9027370699373225, + "grad_norm": 1.5689119384589558, + "learning_rate": 9.833880208430678e-08, + "loss": 0.7439, + "step": 10010 + }, + { + "epoch": 0.9028272534607927, + "grad_norm": 1.7784403677169691, + "learning_rate": 9.815794650579601e-08, + "loss": 0.7354, + "step": 10011 + }, + { + "epoch": 0.902917436984263, + "grad_norm": 1.725229637783135, + "learning_rate": 9.797725320293548e-08, + "loss": 0.7152, + "step": 10012 + }, + { + "epoch": 0.9030076205077332, + "grad_norm": 3.6838698541555397, + "learning_rate": 9.779672219114354e-08, + "loss": 0.8146, + "step": 10013 + }, + { + "epoch": 0.9030978040312035, + "grad_norm": 2.0806320097722977, + "learning_rate": 9.761635348582386e-08, + "loss": 0.7212, + "step": 10014 + }, + { + "epoch": 0.9031879875546738, + "grad_norm": 1.6960240290442046, + "learning_rate": 9.743614710236658e-08, + "loss": 0.7722, + "step": 10015 + }, + { + "epoch": 0.903278171078144, + "grad_norm": 1.7033031265481824, + "learning_rate": 9.725610305614806e-08, + "loss": 0.6999, + "step": 10016 + }, + { + "epoch": 0.9033683546016142, + "grad_norm": 1.5940760110754768, + "learning_rate": 9.707622136253002e-08, + "loss": 0.7329, + "step": 10017 + }, + { + "epoch": 0.9034585381250846, + "grad_norm": 1.6933525463701342, + "learning_rate": 9.689650203686173e-08, + "loss": 0.7639, + "step": 10018 + }, + { + "epoch": 0.9035487216485548, + "grad_norm": 2.014309831086462, + "learning_rate": 9.671694509447715e-08, + "loss": 0.7176, + "step": 10019 + }, + { + "epoch": 0.903638905172025, + "grad_norm": 2.347301995942629, + "learning_rate": 9.653755055069757e-08, + "loss": 0.7571, + "step": 10020 + }, + { + "epoch": 0.9037290886954953, + "grad_norm": 1.269824138014422, + "learning_rate": 9.635831842082987e-08, + "loss": 0.7153, + "step": 10021 + }, + { + "epoch": 0.9038192722189656, + "grad_norm": 1.8839217594568531, + "learning_rate": 9.617924872016691e-08, + "loss": 0.7513, + "step": 10022 + }, + { + "epoch": 0.9039094557424359, + "grad_norm": 1.793340573786663, + "learning_rate": 9.600034146398806e-08, + "loss": 0.7809, + "step": 10023 + }, + { + "epoch": 0.9039996392659061, + "grad_norm": 0.6539114656775874, + "learning_rate": 9.582159666755863e-08, + "loss": 0.5817, + "step": 10024 + }, + { + "epoch": 0.9040898227893763, + "grad_norm": 1.4187618171098928, + "learning_rate": 9.564301434612976e-08, + "loss": 0.712, + "step": 10025 + }, + { + "epoch": 0.9041800063128467, + "grad_norm": 1.7171486764255661, + "learning_rate": 9.546459451494015e-08, + "loss": 0.7831, + "step": 10026 + }, + { + "epoch": 0.9042701898363169, + "grad_norm": 0.6311818293785038, + "learning_rate": 9.528633718921231e-08, + "loss": 0.6054, + "step": 10027 + }, + { + "epoch": 0.9043603733597871, + "grad_norm": 1.48419782685446, + "learning_rate": 9.510824238415672e-08, + "loss": 0.7895, + "step": 10028 + }, + { + "epoch": 0.9044505568832575, + "grad_norm": 3.131194117612776, + "learning_rate": 9.493031011496944e-08, + "loss": 0.7706, + "step": 10029 + }, + { + "epoch": 0.9045407404067277, + "grad_norm": 1.700786139534198, + "learning_rate": 9.475254039683234e-08, + "loss": 0.5982, + "step": 10030 + }, + { + "epoch": 0.904630923930198, + "grad_norm": 1.5128351076765765, + "learning_rate": 9.45749332449144e-08, + "loss": 0.7059, + "step": 10031 + }, + { + "epoch": 0.9047211074536682, + "grad_norm": 0.6480231282468932, + "learning_rate": 9.439748867436903e-08, + "loss": 0.5896, + "step": 10032 + }, + { + "epoch": 0.9048112909771385, + "grad_norm": 2.0503414009335326, + "learning_rate": 9.42202067003377e-08, + "loss": 0.7047, + "step": 10033 + }, + { + "epoch": 0.9049014745006088, + "grad_norm": 1.7442484071584052, + "learning_rate": 9.404308733794652e-08, + "loss": 0.8742, + "step": 10034 + }, + { + "epoch": 0.904991658024079, + "grad_norm": 1.4454301940836258, + "learning_rate": 9.38661306023083e-08, + "loss": 0.7602, + "step": 10035 + }, + { + "epoch": 0.9050818415475492, + "grad_norm": 3.7778010757201606, + "learning_rate": 9.368933650852229e-08, + "loss": 0.731, + "step": 10036 + }, + { + "epoch": 0.9051720250710196, + "grad_norm": 0.7060057014999672, + "learning_rate": 9.351270507167352e-08, + "loss": 0.5941, + "step": 10037 + }, + { + "epoch": 0.9052622085944898, + "grad_norm": 2.442264302340218, + "learning_rate": 9.333623630683285e-08, + "loss": 0.6818, + "step": 10038 + }, + { + "epoch": 0.90535239211796, + "grad_norm": 1.729833344166451, + "learning_rate": 9.315993022905799e-08, + "loss": 0.5916, + "step": 10039 + }, + { + "epoch": 0.9054425756414303, + "grad_norm": 2.239117903732719, + "learning_rate": 9.298378685339158e-08, + "loss": 0.7193, + "step": 10040 + }, + { + "epoch": 0.9055327591649006, + "grad_norm": 2.352879760445539, + "learning_rate": 9.280780619486406e-08, + "loss": 0.8034, + "step": 10041 + }, + { + "epoch": 0.9056229426883708, + "grad_norm": 1.8520070117834182, + "learning_rate": 9.26319882684905e-08, + "loss": 0.8219, + "step": 10042 + }, + { + "epoch": 0.9057131262118411, + "grad_norm": 1.7230276109738059, + "learning_rate": 9.245633308927293e-08, + "loss": 0.7077, + "step": 10043 + }, + { + "epoch": 0.9058033097353113, + "grad_norm": 1.677708369132338, + "learning_rate": 9.228084067219888e-08, + "loss": 0.7216, + "step": 10044 + }, + { + "epoch": 0.9058934932587817, + "grad_norm": 1.5823975692309926, + "learning_rate": 9.210551103224284e-08, + "loss": 0.6817, + "step": 10045 + }, + { + "epoch": 0.9059836767822519, + "grad_norm": 1.529223599876882, + "learning_rate": 9.193034418436463e-08, + "loss": 0.7918, + "step": 10046 + }, + { + "epoch": 0.9060738603057221, + "grad_norm": 0.6578894300623235, + "learning_rate": 9.175534014351005e-08, + "loss": 0.6142, + "step": 10047 + }, + { + "epoch": 0.9061640438291924, + "grad_norm": 2.3137519128297948, + "learning_rate": 9.158049892461228e-08, + "loss": 0.6832, + "step": 10048 + }, + { + "epoch": 0.9062542273526627, + "grad_norm": 1.751353353626744, + "learning_rate": 9.140582054258871e-08, + "loss": 0.6774, + "step": 10049 + }, + { + "epoch": 0.9063444108761329, + "grad_norm": 1.8273782753750976, + "learning_rate": 9.123130501234499e-08, + "loss": 0.749, + "step": 10050 + }, + { + "epoch": 0.9064345943996032, + "grad_norm": 1.7451661652201276, + "learning_rate": 9.105695234877098e-08, + "loss": 0.7898, + "step": 10051 + }, + { + "epoch": 0.9065247779230734, + "grad_norm": 5.391452360930278, + "learning_rate": 9.088276256674344e-08, + "loss": 0.7636, + "step": 10052 + }, + { + "epoch": 0.9066149614465437, + "grad_norm": 1.4700652381797192, + "learning_rate": 9.070873568112536e-08, + "loss": 0.7247, + "step": 10053 + }, + { + "epoch": 0.906705144970014, + "grad_norm": 3.059100697331818, + "learning_rate": 9.053487170676577e-08, + "loss": 0.8001, + "step": 10054 + }, + { + "epoch": 0.9067953284934842, + "grad_norm": 4.097113239685195, + "learning_rate": 9.036117065849968e-08, + "loss": 0.7996, + "step": 10055 + }, + { + "epoch": 0.9068855120169546, + "grad_norm": 2.1400289332136273, + "learning_rate": 9.018763255114837e-08, + "loss": 0.7199, + "step": 10056 + }, + { + "epoch": 0.9069756955404248, + "grad_norm": 1.6123473486983548, + "learning_rate": 9.00142573995184e-08, + "loss": 0.712, + "step": 10057 + }, + { + "epoch": 0.907065879063895, + "grad_norm": 11.121367472263696, + "learning_rate": 8.984104521840375e-08, + "loss": 0.8176, + "step": 10058 + }, + { + "epoch": 0.9071560625873653, + "grad_norm": 3.2772146933860484, + "learning_rate": 8.966799602258346e-08, + "loss": 0.7602, + "step": 10059 + }, + { + "epoch": 0.9072462461108356, + "grad_norm": 1.5705735478320588, + "learning_rate": 8.949510982682329e-08, + "loss": 0.7862, + "step": 10060 + }, + { + "epoch": 0.9073364296343058, + "grad_norm": 2.4703290693872284, + "learning_rate": 8.932238664587499e-08, + "loss": 0.8082, + "step": 10061 + }, + { + "epoch": 0.9074266131577761, + "grad_norm": 2.4139036848548576, + "learning_rate": 8.914982649447567e-08, + "loss": 0.8326, + "step": 10062 + }, + { + "epoch": 0.9075167966812463, + "grad_norm": 2.968615558993311, + "learning_rate": 8.897742938734975e-08, + "loss": 0.8075, + "step": 10063 + }, + { + "epoch": 0.9076069802047166, + "grad_norm": 1.8464077689313165, + "learning_rate": 8.880519533920661e-08, + "loss": 0.7722, + "step": 10064 + }, + { + "epoch": 0.9076971637281869, + "grad_norm": 1.324739831897294, + "learning_rate": 8.863312436474268e-08, + "loss": 0.7436, + "step": 10065 + }, + { + "epoch": 0.9077873472516571, + "grad_norm": 2.110327950433327, + "learning_rate": 8.846121647863936e-08, + "loss": 0.7492, + "step": 10066 + }, + { + "epoch": 0.9078775307751273, + "grad_norm": 1.4801388174113883, + "learning_rate": 8.828947169556555e-08, + "loss": 0.7435, + "step": 10067 + }, + { + "epoch": 0.9079677142985977, + "grad_norm": 1.4477663396652023, + "learning_rate": 8.81178900301749e-08, + "loss": 0.7065, + "step": 10068 + }, + { + "epoch": 0.9080578978220679, + "grad_norm": 1.3873268191196717, + "learning_rate": 8.794647149710787e-08, + "loss": 0.654, + "step": 10069 + }, + { + "epoch": 0.9081480813455381, + "grad_norm": 1.7109597958022165, + "learning_rate": 8.777521611099081e-08, + "loss": 0.7379, + "step": 10070 + }, + { + "epoch": 0.9082382648690084, + "grad_norm": 1.6024858743696686, + "learning_rate": 8.760412388643624e-08, + "loss": 0.6801, + "step": 10071 + }, + { + "epoch": 0.9083284483924787, + "grad_norm": 1.624154310980288, + "learning_rate": 8.74331948380429e-08, + "loss": 0.7555, + "step": 10072 + }, + { + "epoch": 0.908418631915949, + "grad_norm": 1.825178265683335, + "learning_rate": 8.726242898039516e-08, + "loss": 0.7416, + "step": 10073 + }, + { + "epoch": 0.9085088154394192, + "grad_norm": 2.5436959113291713, + "learning_rate": 8.709182632806334e-08, + "loss": 0.7497, + "step": 10074 + }, + { + "epoch": 0.9085989989628894, + "grad_norm": 1.8626923660170218, + "learning_rate": 8.692138689560469e-08, + "loss": 0.7498, + "step": 10075 + }, + { + "epoch": 0.9086891824863598, + "grad_norm": 2.044582101891365, + "learning_rate": 8.675111069756203e-08, + "loss": 0.7261, + "step": 10076 + }, + { + "epoch": 0.90877936600983, + "grad_norm": 1.6143451720113522, + "learning_rate": 8.658099774846395e-08, + "loss": 0.7114, + "step": 10077 + }, + { + "epoch": 0.9088695495333002, + "grad_norm": 1.3916659370389994, + "learning_rate": 8.641104806282595e-08, + "loss": 0.682, + "step": 10078 + }, + { + "epoch": 0.9089597330567706, + "grad_norm": 1.9734709313249759, + "learning_rate": 8.624126165514845e-08, + "loss": 0.749, + "step": 10079 + }, + { + "epoch": 0.9090499165802408, + "grad_norm": 8.523295624248865, + "learning_rate": 8.607163853991917e-08, + "loss": 0.7073, + "step": 10080 + }, + { + "epoch": 0.909140100103711, + "grad_norm": 1.6535993995054545, + "learning_rate": 8.590217873161054e-08, + "loss": 0.5869, + "step": 10081 + }, + { + "epoch": 0.9092302836271813, + "grad_norm": 1.598226580682881, + "learning_rate": 8.573288224468255e-08, + "loss": 0.7547, + "step": 10082 + }, + { + "epoch": 0.9093204671506516, + "grad_norm": 2.157128307199626, + "learning_rate": 8.556374909358011e-08, + "loss": 0.6706, + "step": 10083 + }, + { + "epoch": 0.9094106506741219, + "grad_norm": 1.7471039514517575, + "learning_rate": 8.539477929273476e-08, + "loss": 0.7134, + "step": 10084 + }, + { + "epoch": 0.9095008341975921, + "grad_norm": 1.4007456394653968, + "learning_rate": 8.522597285656386e-08, + "loss": 0.7712, + "step": 10085 + }, + { + "epoch": 0.9095910177210623, + "grad_norm": 0.6590590529082633, + "learning_rate": 8.505732979947078e-08, + "loss": 0.5899, + "step": 10086 + }, + { + "epoch": 0.9096812012445327, + "grad_norm": 1.6553709468421782, + "learning_rate": 8.488885013584557e-08, + "loss": 0.7561, + "step": 10087 + }, + { + "epoch": 0.9097713847680029, + "grad_norm": 1.7692182318775873, + "learning_rate": 8.472053388006295e-08, + "loss": 0.6819, + "step": 10088 + }, + { + "epoch": 0.9098615682914731, + "grad_norm": 1.8732096071326423, + "learning_rate": 8.455238104648565e-08, + "loss": 0.8271, + "step": 10089 + }, + { + "epoch": 0.9099517518149434, + "grad_norm": 2.1893105779519217, + "learning_rate": 8.438439164946043e-08, + "loss": 0.6118, + "step": 10090 + }, + { + "epoch": 0.9100419353384137, + "grad_norm": 1.7090504312010661, + "learning_rate": 8.42165657033218e-08, + "loss": 0.7531, + "step": 10091 + }, + { + "epoch": 0.9101321188618839, + "grad_norm": 1.9390659157033587, + "learning_rate": 8.4048903222389e-08, + "loss": 0.778, + "step": 10092 + }, + { + "epoch": 0.9102223023853542, + "grad_norm": 1.4883305098423905, + "learning_rate": 8.388140422096856e-08, + "loss": 0.7928, + "step": 10093 + }, + { + "epoch": 0.9103124859088244, + "grad_norm": 11.822240940675123, + "learning_rate": 8.371406871335173e-08, + "loss": 0.7602, + "step": 10094 + }, + { + "epoch": 0.9104026694322948, + "grad_norm": 1.6725341353190293, + "learning_rate": 8.354689671381732e-08, + "loss": 0.7411, + "step": 10095 + }, + { + "epoch": 0.910492852955765, + "grad_norm": 1.8561667324103943, + "learning_rate": 8.337988823662834e-08, + "loss": 0.7359, + "step": 10096 + }, + { + "epoch": 0.9105830364792352, + "grad_norm": 2.083421997726714, + "learning_rate": 8.321304329603607e-08, + "loss": 0.698, + "step": 10097 + }, + { + "epoch": 0.9106732200027055, + "grad_norm": 0.6226802452757548, + "learning_rate": 8.304636190627557e-08, + "loss": 0.5865, + "step": 10098 + }, + { + "epoch": 0.9107634035261758, + "grad_norm": 2.3960933599481056, + "learning_rate": 8.287984408156945e-08, + "loss": 0.76, + "step": 10099 + }, + { + "epoch": 0.910853587049646, + "grad_norm": 1.6436578238863175, + "learning_rate": 8.271348983612591e-08, + "loss": 0.6505, + "step": 10100 + }, + { + "epoch": 0.9109437705731163, + "grad_norm": 1.5175174813936902, + "learning_rate": 8.254729918413938e-08, + "loss": 0.7041, + "step": 10101 + }, + { + "epoch": 0.9110339540965866, + "grad_norm": 1.4198801488399302, + "learning_rate": 8.238127213979006e-08, + "loss": 0.7126, + "step": 10102 + }, + { + "epoch": 0.9111241376200568, + "grad_norm": 1.947056429080023, + "learning_rate": 8.221540871724398e-08, + "loss": 0.809, + "step": 10103 + }, + { + "epoch": 0.9112143211435271, + "grad_norm": 1.6641824168665122, + "learning_rate": 8.2049708930654e-08, + "loss": 0.6301, + "step": 10104 + }, + { + "epoch": 0.9113045046669973, + "grad_norm": 1.47540059872812, + "learning_rate": 8.188417279415793e-08, + "loss": 0.7061, + "step": 10105 + }, + { + "epoch": 0.9113946881904677, + "grad_norm": 2.1829392982742215, + "learning_rate": 8.171880032188117e-08, + "loss": 0.8337, + "step": 10106 + }, + { + "epoch": 0.9114848717139379, + "grad_norm": 1.4915300288922788, + "learning_rate": 8.155359152793351e-08, + "loss": 0.7201, + "step": 10107 + }, + { + "epoch": 0.9115750552374081, + "grad_norm": 1.7745576925306974, + "learning_rate": 8.138854642641147e-08, + "loss": 0.7335, + "step": 10108 + }, + { + "epoch": 0.9116652387608783, + "grad_norm": 1.6131263640086493, + "learning_rate": 8.122366503139777e-08, + "loss": 0.6686, + "step": 10109 + }, + { + "epoch": 0.9117554222843487, + "grad_norm": 1.920657384706786, + "learning_rate": 8.105894735696117e-08, + "loss": 0.681, + "step": 10110 + }, + { + "epoch": 0.9118456058078189, + "grad_norm": 2.230331710640762, + "learning_rate": 8.089439341715576e-08, + "loss": 0.7217, + "step": 10111 + }, + { + "epoch": 0.9119357893312892, + "grad_norm": 1.7148974156074552, + "learning_rate": 8.073000322602319e-08, + "loss": 0.7713, + "step": 10112 + }, + { + "epoch": 0.9120259728547594, + "grad_norm": 1.6520941074528324, + "learning_rate": 8.056577679758891e-08, + "loss": 0.7572, + "step": 10113 + }, + { + "epoch": 0.9121161563782297, + "grad_norm": 2.032197701846365, + "learning_rate": 8.040171414586638e-08, + "loss": 0.6951, + "step": 10114 + }, + { + "epoch": 0.9122063399017, + "grad_norm": 1.4524770132796714, + "learning_rate": 8.023781528485419e-08, + "loss": 0.7048, + "step": 10115 + }, + { + "epoch": 0.9122965234251702, + "grad_norm": 2.074311047815539, + "learning_rate": 8.00740802285369e-08, + "loss": 0.7648, + "step": 10116 + }, + { + "epoch": 0.9123867069486404, + "grad_norm": 1.5385027801964615, + "learning_rate": 7.99105089908858e-08, + "loss": 0.6065, + "step": 10117 + }, + { + "epoch": 0.9124768904721108, + "grad_norm": 2.049715008348233, + "learning_rate": 7.974710158585685e-08, + "loss": 0.7156, + "step": 10118 + }, + { + "epoch": 0.912567073995581, + "grad_norm": 1.822641021233228, + "learning_rate": 7.958385802739375e-08, + "loss": 0.766, + "step": 10119 + }, + { + "epoch": 0.9126572575190512, + "grad_norm": 4.70257463751885, + "learning_rate": 7.942077832942452e-08, + "loss": 0.7983, + "step": 10120 + }, + { + "epoch": 0.9127474410425215, + "grad_norm": 0.8115873346065262, + "learning_rate": 7.925786250586508e-08, + "loss": 0.5817, + "step": 10121 + }, + { + "epoch": 0.9128376245659918, + "grad_norm": 0.6574136544233258, + "learning_rate": 7.909511057061524e-08, + "loss": 0.5805, + "step": 10122 + }, + { + "epoch": 0.9129278080894621, + "grad_norm": 2.374675527794205, + "learning_rate": 7.893252253756234e-08, + "loss": 0.7258, + "step": 10123 + }, + { + "epoch": 0.9130179916129323, + "grad_norm": 1.9448874135516434, + "learning_rate": 7.877009842057925e-08, + "loss": 0.6702, + "step": 10124 + }, + { + "epoch": 0.9131081751364026, + "grad_norm": 1.6209515244932775, + "learning_rate": 7.860783823352512e-08, + "loss": 0.7659, + "step": 10125 + }, + { + "epoch": 0.9131983586598729, + "grad_norm": 1.5933340689259723, + "learning_rate": 7.844574199024445e-08, + "loss": 0.75, + "step": 10126 + }, + { + "epoch": 0.9132885421833431, + "grad_norm": 1.6022764521936126, + "learning_rate": 7.82838097045686e-08, + "loss": 0.7854, + "step": 10127 + }, + { + "epoch": 0.9133787257068133, + "grad_norm": 1.7242159540510649, + "learning_rate": 7.812204139031454e-08, + "loss": 0.7336, + "step": 10128 + }, + { + "epoch": 0.9134689092302837, + "grad_norm": 1.9355886937925308, + "learning_rate": 7.796043706128474e-08, + "loss": 0.7532, + "step": 10129 + }, + { + "epoch": 0.9135590927537539, + "grad_norm": 1.94321313048976, + "learning_rate": 7.779899673126844e-08, + "loss": 0.7679, + "step": 10130 + }, + { + "epoch": 0.9136492762772241, + "grad_norm": 1.5455555538583063, + "learning_rate": 7.76377204140406e-08, + "loss": 0.7817, + "step": 10131 + }, + { + "epoch": 0.9137394598006944, + "grad_norm": 1.860039087520607, + "learning_rate": 7.74766081233622e-08, + "loss": 0.729, + "step": 10132 + }, + { + "epoch": 0.9138296433241647, + "grad_norm": 1.607697510754755, + "learning_rate": 7.73156598729805e-08, + "loss": 0.8075, + "step": 10133 + }, + { + "epoch": 0.913919826847635, + "grad_norm": 1.7741607669978898, + "learning_rate": 7.715487567662849e-08, + "loss": 0.7934, + "step": 10134 + }, + { + "epoch": 0.9140100103711052, + "grad_norm": 2.9069754402690005, + "learning_rate": 7.69942555480243e-08, + "loss": 0.7087, + "step": 10135 + }, + { + "epoch": 0.9141001938945754, + "grad_norm": 1.741320954771476, + "learning_rate": 7.68337995008741e-08, + "loss": 0.8454, + "step": 10136 + }, + { + "epoch": 0.9141903774180458, + "grad_norm": 1.4730091431043275, + "learning_rate": 7.667350754886803e-08, + "loss": 0.5965, + "step": 10137 + }, + { + "epoch": 0.914280560941516, + "grad_norm": 3.4128763682978094, + "learning_rate": 7.651337970568361e-08, + "loss": 0.7344, + "step": 10138 + }, + { + "epoch": 0.9143707444649862, + "grad_norm": 1.6330733387510128, + "learning_rate": 7.635341598498368e-08, + "loss": 0.7052, + "step": 10139 + }, + { + "epoch": 0.9144609279884565, + "grad_norm": 2.391449929317947, + "learning_rate": 7.61936164004171e-08, + "loss": 0.7376, + "step": 10140 + }, + { + "epoch": 0.9145511115119268, + "grad_norm": 1.5511886868590101, + "learning_rate": 7.603398096561875e-08, + "loss": 0.7878, + "step": 10141 + }, + { + "epoch": 0.914641295035397, + "grad_norm": 1.7728665142683304, + "learning_rate": 7.587450969420994e-08, + "loss": 0.7479, + "step": 10142 + }, + { + "epoch": 0.9147314785588673, + "grad_norm": 0.6780831266914432, + "learning_rate": 7.571520259979757e-08, + "loss": 0.5607, + "step": 10143 + }, + { + "epoch": 0.9148216620823375, + "grad_norm": 1.9668185735753778, + "learning_rate": 7.555605969597455e-08, + "loss": 0.5784, + "step": 10144 + }, + { + "epoch": 0.9149118456058078, + "grad_norm": 1.7150915570745953, + "learning_rate": 7.539708099631959e-08, + "loss": 0.8246, + "step": 10145 + }, + { + "epoch": 0.9150020291292781, + "grad_norm": 1.494242107460504, + "learning_rate": 7.52382665143978e-08, + "loss": 0.7469, + "step": 10146 + }, + { + "epoch": 0.9150922126527483, + "grad_norm": 1.7897678633894891, + "learning_rate": 7.507961626376014e-08, + "loss": 0.7967, + "step": 10147 + }, + { + "epoch": 0.9151823961762187, + "grad_norm": 1.714831237052155, + "learning_rate": 7.492113025794378e-08, + "loss": 0.7794, + "step": 10148 + }, + { + "epoch": 0.9152725796996889, + "grad_norm": 1.6389472739098394, + "learning_rate": 7.476280851047101e-08, + "loss": 0.6894, + "step": 10149 + }, + { + "epoch": 0.9153627632231591, + "grad_norm": 1.1950329721762147, + "learning_rate": 7.460465103485125e-08, + "loss": 0.7622, + "step": 10150 + }, + { + "epoch": 0.9154529467466294, + "grad_norm": 1.7998701260655239, + "learning_rate": 7.444665784457948e-08, + "loss": 0.7323, + "step": 10151 + }, + { + "epoch": 0.9155431302700997, + "grad_norm": 2.048968111243397, + "learning_rate": 7.42888289531356e-08, + "loss": 0.698, + "step": 10152 + }, + { + "epoch": 0.9156333137935699, + "grad_norm": 1.583525453589982, + "learning_rate": 7.41311643739877e-08, + "loss": 0.6621, + "step": 10153 + }, + { + "epoch": 0.9157234973170402, + "grad_norm": 1.5532338616579457, + "learning_rate": 7.39736641205877e-08, + "loss": 0.7812, + "step": 10154 + }, + { + "epoch": 0.9158136808405104, + "grad_norm": 1.7062411018111612, + "learning_rate": 7.381632820637462e-08, + "loss": 0.7649, + "step": 10155 + }, + { + "epoch": 0.9159038643639807, + "grad_norm": 1.8126211454748908, + "learning_rate": 7.365915664477352e-08, + "loss": 0.7325, + "step": 10156 + }, + { + "epoch": 0.915994047887451, + "grad_norm": 1.845447211714376, + "learning_rate": 7.350214944919474e-08, + "loss": 0.7455, + "step": 10157 + }, + { + "epoch": 0.9160842314109212, + "grad_norm": 1.8209946299792676, + "learning_rate": 7.334530663303539e-08, + "loss": 0.7739, + "step": 10158 + }, + { + "epoch": 0.9161744149343914, + "grad_norm": 1.9670905127943434, + "learning_rate": 7.318862820967742e-08, + "loss": 0.75, + "step": 10159 + }, + { + "epoch": 0.9162645984578618, + "grad_norm": 1.6085909813314718, + "learning_rate": 7.303211419249056e-08, + "loss": 0.7703, + "step": 10160 + }, + { + "epoch": 0.916354781981332, + "grad_norm": 1.615431417234369, + "learning_rate": 7.287576459482858e-08, + "loss": 0.7577, + "step": 10161 + }, + { + "epoch": 0.9164449655048023, + "grad_norm": 1.926823270278433, + "learning_rate": 7.271957943003259e-08, + "loss": 0.7793, + "step": 10162 + }, + { + "epoch": 0.9165351490282725, + "grad_norm": 1.92230877206739, + "learning_rate": 7.256355871142883e-08, + "loss": 0.7401, + "step": 10163 + }, + { + "epoch": 0.9166253325517428, + "grad_norm": 2.736212499927058, + "learning_rate": 7.240770245233019e-08, + "loss": 0.7935, + "step": 10164 + }, + { + "epoch": 0.9167155160752131, + "grad_norm": 1.5745429753446794, + "learning_rate": 7.225201066603492e-08, + "loss": 0.6541, + "step": 10165 + }, + { + "epoch": 0.9168056995986833, + "grad_norm": 1.7494678369374608, + "learning_rate": 7.209648336582774e-08, + "loss": 0.8492, + "step": 10166 + }, + { + "epoch": 0.9168958831221535, + "grad_norm": 2.121321467475305, + "learning_rate": 7.19411205649787e-08, + "loss": 0.726, + "step": 10167 + }, + { + "epoch": 0.9169860666456239, + "grad_norm": 1.861621045083243, + "learning_rate": 7.178592227674474e-08, + "loss": 0.7644, + "step": 10168 + }, + { + "epoch": 0.9170762501690941, + "grad_norm": 1.9528220590255834, + "learning_rate": 7.163088851436771e-08, + "loss": 0.6553, + "step": 10169 + }, + { + "epoch": 0.9171664336925643, + "grad_norm": 1.5117748161387343, + "learning_rate": 7.147601929107639e-08, + "loss": 0.7061, + "step": 10170 + }, + { + "epoch": 0.9172566172160346, + "grad_norm": 1.8043370422662492, + "learning_rate": 7.132131462008461e-08, + "loss": 0.7913, + "step": 10171 + }, + { + "epoch": 0.9173468007395049, + "grad_norm": 1.704832288710357, + "learning_rate": 7.116677451459297e-08, + "loss": 0.7468, + "step": 10172 + }, + { + "epoch": 0.9174369842629752, + "grad_norm": 1.3182735612985732, + "learning_rate": 7.101239898778799e-08, + "loss": 0.7645, + "step": 10173 + }, + { + "epoch": 0.9175271677864454, + "grad_norm": 2.261585712637907, + "learning_rate": 7.085818805284094e-08, + "loss": 0.7778, + "step": 10174 + }, + { + "epoch": 0.9176173513099157, + "grad_norm": 1.7345045574015094, + "learning_rate": 7.070414172291083e-08, + "loss": 0.8227, + "step": 10175 + }, + { + "epoch": 0.917707534833386, + "grad_norm": 1.3759941928402764, + "learning_rate": 7.055026001114095e-08, + "loss": 0.6866, + "step": 10176 + }, + { + "epoch": 0.9177977183568562, + "grad_norm": 0.6451019072794656, + "learning_rate": 7.039654293066211e-08, + "loss": 0.5658, + "step": 10177 + }, + { + "epoch": 0.9178879018803264, + "grad_norm": 1.72037540220314, + "learning_rate": 7.024299049459003e-08, + "loss": 0.6629, + "step": 10178 + }, + { + "epoch": 0.9179780854037968, + "grad_norm": 2.1460147027954535, + "learning_rate": 7.008960271602627e-08, + "loss": 0.6579, + "step": 10179 + }, + { + "epoch": 0.918068268927267, + "grad_norm": 1.8221648782272213, + "learning_rate": 6.993637960805921e-08, + "loss": 0.7064, + "step": 10180 + }, + { + "epoch": 0.9181584524507372, + "grad_norm": 1.8056433322143568, + "learning_rate": 6.97833211837624e-08, + "loss": 0.7664, + "step": 10181 + }, + { + "epoch": 0.9182486359742075, + "grad_norm": 2.227856176565968, + "learning_rate": 6.963042745619562e-08, + "loss": 0.6879, + "step": 10182 + }, + { + "epoch": 0.9183388194976778, + "grad_norm": 1.7226979776116476, + "learning_rate": 6.947769843840511e-08, + "loss": 0.7096, + "step": 10183 + }, + { + "epoch": 0.918429003021148, + "grad_norm": 1.7909425410644215, + "learning_rate": 6.9325134143422e-08, + "loss": 0.7731, + "step": 10184 + }, + { + "epoch": 0.9185191865446183, + "grad_norm": 1.8104622665767216, + "learning_rate": 6.917273458426387e-08, + "loss": 0.7228, + "step": 10185 + }, + { + "epoch": 0.9186093700680885, + "grad_norm": 0.6868450122496921, + "learning_rate": 6.902049977393476e-08, + "loss": 0.5541, + "step": 10186 + }, + { + "epoch": 0.9186995535915589, + "grad_norm": 2.838333672885423, + "learning_rate": 6.886842972542362e-08, + "loss": 0.7465, + "step": 10187 + }, + { + "epoch": 0.9187897371150291, + "grad_norm": 1.5145886210786987, + "learning_rate": 6.871652445170672e-08, + "loss": 0.6299, + "step": 10188 + }, + { + "epoch": 0.9188799206384993, + "grad_norm": 1.6008433146101866, + "learning_rate": 6.856478396574416e-08, + "loss": 0.8267, + "step": 10189 + }, + { + "epoch": 0.9189701041619696, + "grad_norm": 3.0932066469175106, + "learning_rate": 6.841320828048491e-08, + "loss": 0.8012, + "step": 10190 + }, + { + "epoch": 0.9190602876854399, + "grad_norm": 1.7535647461236372, + "learning_rate": 6.826179740886062e-08, + "loss": 0.738, + "step": 10191 + }, + { + "epoch": 0.9191504712089101, + "grad_norm": 1.6276627512242892, + "learning_rate": 6.811055136379184e-08, + "loss": 0.8058, + "step": 10192 + }, + { + "epoch": 0.9192406547323804, + "grad_norm": 3.3941655812557165, + "learning_rate": 6.79594701581827e-08, + "loss": 0.7182, + "step": 10193 + }, + { + "epoch": 0.9193308382558506, + "grad_norm": 1.7065387076350593, + "learning_rate": 6.780855380492511e-08, + "loss": 0.7478, + "step": 10194 + }, + { + "epoch": 0.919421021779321, + "grad_norm": 2.6771850467660405, + "learning_rate": 6.765780231689544e-08, + "loss": 0.7546, + "step": 10195 + }, + { + "epoch": 0.9195112053027912, + "grad_norm": 2.1390989112402194, + "learning_rate": 6.750721570695695e-08, + "loss": 0.7967, + "step": 10196 + }, + { + "epoch": 0.9196013888262614, + "grad_norm": 3.150010844620702, + "learning_rate": 6.735679398795868e-08, + "loss": 0.7249, + "step": 10197 + }, + { + "epoch": 0.9196915723497318, + "grad_norm": 3.209736768916171, + "learning_rate": 6.720653717273506e-08, + "loss": 0.6805, + "step": 10198 + }, + { + "epoch": 0.919781755873202, + "grad_norm": 1.5663483340241935, + "learning_rate": 6.705644527410714e-08, + "loss": 0.7422, + "step": 10199 + }, + { + "epoch": 0.9198719393966722, + "grad_norm": 1.966154961638492, + "learning_rate": 6.690651830488136e-08, + "loss": 0.7099, + "step": 10200 + }, + { + "epoch": 0.9199621229201425, + "grad_norm": 2.354851738379152, + "learning_rate": 6.675675627785037e-08, + "loss": 0.7401, + "step": 10201 + }, + { + "epoch": 0.9200523064436128, + "grad_norm": 4.750634201275768, + "learning_rate": 6.660715920579263e-08, + "loss": 0.7603, + "step": 10202 + }, + { + "epoch": 0.920142489967083, + "grad_norm": 0.5940027953621565, + "learning_rate": 6.645772710147279e-08, + "loss": 0.5773, + "step": 10203 + }, + { + "epoch": 0.9202326734905533, + "grad_norm": 1.2817653456964042, + "learning_rate": 6.630845997764112e-08, + "loss": 0.7156, + "step": 10204 + }, + { + "epoch": 0.9203228570140235, + "grad_norm": 1.8001776819482216, + "learning_rate": 6.615935784703409e-08, + "loss": 0.6833, + "step": 10205 + }, + { + "epoch": 0.9204130405374938, + "grad_norm": 1.664495707255701, + "learning_rate": 6.601042072237328e-08, + "loss": 0.6673, + "step": 10206 + }, + { + "epoch": 0.9205032240609641, + "grad_norm": 1.767338149498865, + "learning_rate": 6.586164861636767e-08, + "loss": 0.7692, + "step": 10207 + }, + { + "epoch": 0.9205934075844343, + "grad_norm": 2.131454650360685, + "learning_rate": 6.571304154171065e-08, + "loss": 0.6563, + "step": 10208 + }, + { + "epoch": 0.9206835911079045, + "grad_norm": 2.020406212517751, + "learning_rate": 6.556459951108273e-08, + "loss": 0.7623, + "step": 10209 + }, + { + "epoch": 0.9207737746313749, + "grad_norm": 2.054367098466691, + "learning_rate": 6.541632253714957e-08, + "loss": 0.7227, + "step": 10210 + }, + { + "epoch": 0.9208639581548451, + "grad_norm": 1.8904634517031378, + "learning_rate": 6.526821063256261e-08, + "loss": 0.6692, + "step": 10211 + }, + { + "epoch": 0.9209541416783154, + "grad_norm": 2.672480107546534, + "learning_rate": 6.512026380996016e-08, + "loss": 0.6523, + "step": 10212 + }, + { + "epoch": 0.9210443252017856, + "grad_norm": 1.6917028717651634, + "learning_rate": 6.49724820819657e-08, + "loss": 0.7153, + "step": 10213 + }, + { + "epoch": 0.9211345087252559, + "grad_norm": 1.7490367963441535, + "learning_rate": 6.48248654611887e-08, + "loss": 0.7742, + "step": 10214 + }, + { + "epoch": 0.9212246922487262, + "grad_norm": 2.020499906529059, + "learning_rate": 6.467741396022419e-08, + "loss": 0.7892, + "step": 10215 + }, + { + "epoch": 0.9213148757721964, + "grad_norm": 1.8611954184394817, + "learning_rate": 6.453012759165455e-08, + "loss": 0.7511, + "step": 10216 + }, + { + "epoch": 0.9214050592956666, + "grad_norm": 2.564224686220878, + "learning_rate": 6.438300636804639e-08, + "loss": 0.7065, + "step": 10217 + }, + { + "epoch": 0.921495242819137, + "grad_norm": 1.6772174505150654, + "learning_rate": 6.423605030195278e-08, + "loss": 0.7767, + "step": 10218 + }, + { + "epoch": 0.9215854263426072, + "grad_norm": 2.266480698623455, + "learning_rate": 6.408925940591304e-08, + "loss": 0.8174, + "step": 10219 + }, + { + "epoch": 0.9216756098660774, + "grad_norm": 1.6556753205212162, + "learning_rate": 6.394263369245222e-08, + "loss": 0.6729, + "step": 10220 + }, + { + "epoch": 0.9217657933895478, + "grad_norm": 1.7346565774076375, + "learning_rate": 6.379617317408126e-08, + "loss": 0.7463, + "step": 10221 + }, + { + "epoch": 0.921855976913018, + "grad_norm": 1.9201782924601571, + "learning_rate": 6.364987786329723e-08, + "loss": 0.731, + "step": 10222 + }, + { + "epoch": 0.9219461604364882, + "grad_norm": 1.6591627312074997, + "learning_rate": 6.350374777258193e-08, + "loss": 0.7364, + "step": 10223 + }, + { + "epoch": 0.9220363439599585, + "grad_norm": 1.6919067910811225, + "learning_rate": 6.335778291440519e-08, + "loss": 0.7206, + "step": 10224 + }, + { + "epoch": 0.9221265274834288, + "grad_norm": 1.9600491278193963, + "learning_rate": 6.321198330122057e-08, + "loss": 0.6492, + "step": 10225 + }, + { + "epoch": 0.9222167110068991, + "grad_norm": 1.7003643514968156, + "learning_rate": 6.306634894546902e-08, + "loss": 0.6758, + "step": 10226 + }, + { + "epoch": 0.9223068945303693, + "grad_norm": 1.5207539716245735, + "learning_rate": 6.292087985957661e-08, + "loss": 0.6808, + "step": 10227 + }, + { + "epoch": 0.9223970780538395, + "grad_norm": 2.2418081780145553, + "learning_rate": 6.277557605595585e-08, + "loss": 0.7821, + "step": 10228 + }, + { + "epoch": 0.9224872615773099, + "grad_norm": 1.3234320810470521, + "learning_rate": 6.263043754700481e-08, + "loss": 0.6669, + "step": 10229 + }, + { + "epoch": 0.9225774451007801, + "grad_norm": 1.7381701487493733, + "learning_rate": 6.248546434510671e-08, + "loss": 0.7403, + "step": 10230 + }, + { + "epoch": 0.9226676286242503, + "grad_norm": 1.8067570829230213, + "learning_rate": 6.234065646263298e-08, + "loss": 0.6738, + "step": 10231 + }, + { + "epoch": 0.9227578121477206, + "grad_norm": 1.8878232514925046, + "learning_rate": 6.219601391193796e-08, + "loss": 0.647, + "step": 10232 + }, + { + "epoch": 0.9228479956711909, + "grad_norm": 2.2614929402739006, + "learning_rate": 6.205153670536423e-08, + "loss": 0.6678, + "step": 10233 + }, + { + "epoch": 0.9229381791946611, + "grad_norm": 1.5892839125597005, + "learning_rate": 6.190722485523902e-08, + "loss": 0.7758, + "step": 10234 + }, + { + "epoch": 0.9230283627181314, + "grad_norm": 1.4959777308963829, + "learning_rate": 6.176307837387607e-08, + "loss": 0.7166, + "step": 10235 + }, + { + "epoch": 0.9231185462416016, + "grad_norm": 1.6951785408218987, + "learning_rate": 6.16190972735744e-08, + "loss": 0.6894, + "step": 10236 + }, + { + "epoch": 0.923208729765072, + "grad_norm": 1.6419411602420166, + "learning_rate": 6.147528156661974e-08, + "loss": 0.7679, + "step": 10237 + }, + { + "epoch": 0.9232989132885422, + "grad_norm": 1.9168401998937499, + "learning_rate": 6.133163126528273e-08, + "loss": 0.6613, + "step": 10238 + }, + { + "epoch": 0.9233890968120124, + "grad_norm": 1.8954786311187557, + "learning_rate": 6.11881463818209e-08, + "loss": 0.7461, + "step": 10239 + }, + { + "epoch": 0.9234792803354827, + "grad_norm": 1.980130198658082, + "learning_rate": 6.104482692847668e-08, + "loss": 0.7311, + "step": 10240 + }, + { + "epoch": 0.923569463858953, + "grad_norm": 1.5446890596493412, + "learning_rate": 6.090167291747917e-08, + "loss": 0.7072, + "step": 10241 + }, + { + "epoch": 0.9236596473824232, + "grad_norm": 12.802583457586854, + "learning_rate": 6.075868436104303e-08, + "loss": 0.6382, + "step": 10242 + }, + { + "epoch": 0.9237498309058935, + "grad_norm": 1.8267719445670485, + "learning_rate": 6.061586127136875e-08, + "loss": 0.7287, + "step": 10243 + }, + { + "epoch": 0.9238400144293638, + "grad_norm": 1.5313403823024903, + "learning_rate": 6.047320366064324e-08, + "loss": 0.7058, + "step": 10244 + }, + { + "epoch": 0.923930197952834, + "grad_norm": 1.671801342633983, + "learning_rate": 6.033071154103786e-08, + "loss": 0.7499, + "step": 10245 + }, + { + "epoch": 0.9240203814763043, + "grad_norm": 1.6145223134313218, + "learning_rate": 6.018838492471178e-08, + "loss": 0.7604, + "step": 10246 + }, + { + "epoch": 0.9241105649997745, + "grad_norm": 0.620269468337918, + "learning_rate": 6.00462238238082e-08, + "loss": 0.5799, + "step": 10247 + }, + { + "epoch": 0.9242007485232449, + "grad_norm": 2.407934218689972, + "learning_rate": 5.990422825045827e-08, + "loss": 0.7054, + "step": 10248 + }, + { + "epoch": 0.9242909320467151, + "grad_norm": 1.6405412925074074, + "learning_rate": 5.976239821677675e-08, + "loss": 0.7194, + "step": 10249 + }, + { + "epoch": 0.9243811155701853, + "grad_norm": 6.011365933441599, + "learning_rate": 5.962073373486598e-08, + "loss": 0.6826, + "step": 10250 + }, + { + "epoch": 0.9244712990936556, + "grad_norm": 1.4167248863524031, + "learning_rate": 5.947923481681316e-08, + "loss": 0.7027, + "step": 10251 + }, + { + "epoch": 0.9245614826171259, + "grad_norm": 1.5555101834598088, + "learning_rate": 5.933790147469198e-08, + "loss": 0.7114, + "step": 10252 + }, + { + "epoch": 0.9246516661405961, + "grad_norm": 2.9667210120320417, + "learning_rate": 5.9196733720561665e-08, + "loss": 0.6923, + "step": 10253 + }, + { + "epoch": 0.9247418496640664, + "grad_norm": 3.7566831164492105, + "learning_rate": 5.905573156646793e-08, + "loss": 0.8244, + "step": 10254 + }, + { + "epoch": 0.9248320331875366, + "grad_norm": 2.0725295377865085, + "learning_rate": 5.8914895024441134e-08, + "loss": 0.8157, + "step": 10255 + }, + { + "epoch": 0.9249222167110069, + "grad_norm": 4.894237025399466, + "learning_rate": 5.877422410649857e-08, + "loss": 0.7716, + "step": 10256 + }, + { + "epoch": 0.9250124002344772, + "grad_norm": 1.8016861567941345, + "learning_rate": 5.863371882464285e-08, + "loss": 0.7575, + "step": 10257 + }, + { + "epoch": 0.9251025837579474, + "grad_norm": 20.519386614876705, + "learning_rate": 5.849337919086283e-08, + "loss": 0.7196, + "step": 10258 + }, + { + "epoch": 0.9251927672814176, + "grad_norm": 2.691531152138524, + "learning_rate": 5.835320521713316e-08, + "loss": 0.7174, + "step": 10259 + }, + { + "epoch": 0.925282950804888, + "grad_norm": 1.5884076504901827, + "learning_rate": 5.8213196915414264e-08, + "loss": 0.7248, + "step": 10260 + }, + { + "epoch": 0.9253731343283582, + "grad_norm": 2.8987077859467147, + "learning_rate": 5.807335429765237e-08, + "loss": 0.795, + "step": 10261 + }, + { + "epoch": 0.9254633178518284, + "grad_norm": 1.8905258152694115, + "learning_rate": 5.7933677375779034e-08, + "loss": 0.7891, + "step": 10262 + }, + { + "epoch": 0.9255535013752987, + "grad_norm": 1.55723026456528, + "learning_rate": 5.77941661617134e-08, + "loss": 0.7918, + "step": 10263 + }, + { + "epoch": 0.925643684898769, + "grad_norm": 1.7537645992179836, + "learning_rate": 5.765482066735816e-08, + "loss": 0.6745, + "step": 10264 + }, + { + "epoch": 0.9257338684222393, + "grad_norm": 2.1344094358824717, + "learning_rate": 5.7515640904604256e-08, + "loss": 0.744, + "step": 10265 + }, + { + "epoch": 0.9258240519457095, + "grad_norm": 1.6543569722794824, + "learning_rate": 5.7376626885326187e-08, + "loss": 0.7265, + "step": 10266 + }, + { + "epoch": 0.9259142354691798, + "grad_norm": 1.6043476729226065, + "learning_rate": 5.723777862138601e-08, + "loss": 0.8096, + "step": 10267 + }, + { + "epoch": 0.9260044189926501, + "grad_norm": 2.0512895807867753, + "learning_rate": 5.7099096124630705e-08, + "loss": 0.8202, + "step": 10268 + }, + { + "epoch": 0.9260946025161203, + "grad_norm": 1.2720153449980316, + "learning_rate": 5.696057940689347e-08, + "loss": 0.7492, + "step": 10269 + }, + { + "epoch": 0.9261847860395905, + "grad_norm": 2.039637108296184, + "learning_rate": 5.6822228479993736e-08, + "loss": 0.7625, + "step": 10270 + }, + { + "epoch": 0.9262749695630609, + "grad_norm": 2.867943271740973, + "learning_rate": 5.668404335573584e-08, + "loss": 0.7818, + "step": 10271 + }, + { + "epoch": 0.9263651530865311, + "grad_norm": 1.612171521108567, + "learning_rate": 5.654602404591058e-08, + "loss": 0.6857, + "step": 10272 + }, + { + "epoch": 0.9264553366100013, + "grad_norm": 3.091004203159186, + "learning_rate": 5.640817056229474e-08, + "loss": 0.7986, + "step": 10273 + }, + { + "epoch": 0.9265455201334716, + "grad_norm": 1.4968462214560505, + "learning_rate": 5.6270482916650706e-08, + "loss": 0.7841, + "step": 10274 + }, + { + "epoch": 0.9266357036569419, + "grad_norm": 1.6384501256025201, + "learning_rate": 5.613296112072663e-08, + "loss": 0.8174, + "step": 10275 + }, + { + "epoch": 0.9267258871804122, + "grad_norm": 6.044554587993753, + "learning_rate": 5.59956051862569e-08, + "loss": 0.7479, + "step": 10276 + }, + { + "epoch": 0.9268160707038824, + "grad_norm": 1.6864831474541122, + "learning_rate": 5.585841512496081e-08, + "loss": 0.8059, + "step": 10277 + }, + { + "epoch": 0.9269062542273526, + "grad_norm": 2.8807693312199203, + "learning_rate": 5.5721390948545e-08, + "loss": 0.8498, + "step": 10278 + }, + { + "epoch": 0.926996437750823, + "grad_norm": 1.8198210089562852, + "learning_rate": 5.558453266870056e-08, + "loss": 0.7858, + "step": 10279 + }, + { + "epoch": 0.9270866212742932, + "grad_norm": 1.8804611148845636, + "learning_rate": 5.544784029710525e-08, + "loss": 0.6835, + "step": 10280 + }, + { + "epoch": 0.9271768047977634, + "grad_norm": 3.056153552946018, + "learning_rate": 5.531131384542242e-08, + "loss": 0.8152, + "step": 10281 + }, + { + "epoch": 0.9272669883212337, + "grad_norm": 1.442457523877538, + "learning_rate": 5.51749533253012e-08, + "loss": 0.6921, + "step": 10282 + }, + { + "epoch": 0.927357171844704, + "grad_norm": 1.892254718118945, + "learning_rate": 5.503875874837649e-08, + "loss": 0.6841, + "step": 10283 + }, + { + "epoch": 0.9274473553681742, + "grad_norm": 1.6368964560822319, + "learning_rate": 5.4902730126269225e-08, + "loss": 0.7077, + "step": 10284 + }, + { + "epoch": 0.9275375388916445, + "grad_norm": 1.9767986654087997, + "learning_rate": 5.476686747058656e-08, + "loss": 0.7782, + "step": 10285 + }, + { + "epoch": 0.9276277224151147, + "grad_norm": 2.1450254777582325, + "learning_rate": 5.4631170792920124e-08, + "loss": 0.7199, + "step": 10286 + }, + { + "epoch": 0.927717905938585, + "grad_norm": 1.792535534217862, + "learning_rate": 5.449564010484953e-08, + "loss": 0.7187, + "step": 10287 + }, + { + "epoch": 0.9278080894620553, + "grad_norm": 2.1646480938571186, + "learning_rate": 5.436027541793775e-08, + "loss": 0.767, + "step": 10288 + }, + { + "epoch": 0.9278982729855255, + "grad_norm": 1.671393423153366, + "learning_rate": 5.4225076743735554e-08, + "loss": 0.7193, + "step": 10289 + }, + { + "epoch": 0.9279884565089958, + "grad_norm": 1.5318293351363437, + "learning_rate": 5.409004409377882e-08, + "loss": 0.6649, + "step": 10290 + }, + { + "epoch": 0.9280786400324661, + "grad_norm": 1.5609700914556524, + "learning_rate": 5.3955177479589e-08, + "loss": 0.6672, + "step": 10291 + }, + { + "epoch": 0.9281688235559363, + "grad_norm": 11.256284865558857, + "learning_rate": 5.3820476912674e-08, + "loss": 0.748, + "step": 10292 + }, + { + "epoch": 0.9282590070794066, + "grad_norm": 1.494000676401335, + "learning_rate": 5.3685942404527063e-08, + "loss": 0.7697, + "step": 10293 + }, + { + "epoch": 0.9283491906028769, + "grad_norm": 1.8823219388361678, + "learning_rate": 5.355157396662702e-08, + "loss": 0.7226, + "step": 10294 + }, + { + "epoch": 0.9284393741263471, + "grad_norm": 1.5368275947597423, + "learning_rate": 5.34173716104398e-08, + "loss": 0.7257, + "step": 10295 + }, + { + "epoch": 0.9285295576498174, + "grad_norm": 1.8164323702339003, + "learning_rate": 5.328333534741536e-08, + "loss": 0.7517, + "step": 10296 + }, + { + "epoch": 0.9286197411732876, + "grad_norm": 3.236537634192432, + "learning_rate": 5.314946518899099e-08, + "loss": 0.7115, + "step": 10297 + }, + { + "epoch": 0.928709924696758, + "grad_norm": 1.7972771236099105, + "learning_rate": 5.301576114658912e-08, + "loss": 0.8106, + "step": 10298 + }, + { + "epoch": 0.9288001082202282, + "grad_norm": 2.284871297170808, + "learning_rate": 5.288222323161795e-08, + "loss": 0.77, + "step": 10299 + }, + { + "epoch": 0.9288902917436984, + "grad_norm": 2.124882414313367, + "learning_rate": 5.274885145547214e-08, + "loss": 0.7939, + "step": 10300 + }, + { + "epoch": 0.9289804752671686, + "grad_norm": 2.1237361740012517, + "learning_rate": 5.261564582953082e-08, + "loss": 0.7339, + "step": 10301 + }, + { + "epoch": 0.929070658790639, + "grad_norm": 2.141246844554921, + "learning_rate": 5.248260636516066e-08, + "loss": 0.7202, + "step": 10302 + }, + { + "epoch": 0.9291608423141092, + "grad_norm": 1.9415950568204998, + "learning_rate": 5.2349733073712824e-08, + "loss": 0.6897, + "step": 10303 + }, + { + "epoch": 0.9292510258375795, + "grad_norm": 1.5568977631781136, + "learning_rate": 5.221702596652533e-08, + "loss": 0.8305, + "step": 10304 + }, + { + "epoch": 0.9293412093610497, + "grad_norm": 2.523570810109452, + "learning_rate": 5.208448505492091e-08, + "loss": 0.6886, + "step": 10305 + }, + { + "epoch": 0.92943139288452, + "grad_norm": 1.6920204368962308, + "learning_rate": 5.1952110350208965e-08, + "loss": 0.7246, + "step": 10306 + }, + { + "epoch": 0.9295215764079903, + "grad_norm": 1.8001505311965, + "learning_rate": 5.181990186368446e-08, + "loss": 0.6975, + "step": 10307 + }, + { + "epoch": 0.9296117599314605, + "grad_norm": 1.7763141710151304, + "learning_rate": 5.1687859606627915e-08, + "loss": 0.8376, + "step": 10308 + }, + { + "epoch": 0.9297019434549307, + "grad_norm": 2.1570611063967884, + "learning_rate": 5.1555983590306327e-08, + "loss": 0.653, + "step": 10309 + }, + { + "epoch": 0.9297921269784011, + "grad_norm": 0.6560112882332002, + "learning_rate": 5.1424273825971806e-08, + "loss": 0.5785, + "step": 10310 + }, + { + "epoch": 0.9298823105018713, + "grad_norm": 1.7284477608499786, + "learning_rate": 5.1292730324862475e-08, + "loss": 0.7021, + "step": 10311 + }, + { + "epoch": 0.9299724940253415, + "grad_norm": 1.5548757314550716, + "learning_rate": 5.116135309820224e-08, + "loss": 0.7573, + "step": 10312 + }, + { + "epoch": 0.9300626775488118, + "grad_norm": 1.9448470818978423, + "learning_rate": 5.103014215720147e-08, + "loss": 0.7744, + "step": 10313 + }, + { + "epoch": 0.9301528610722821, + "grad_norm": 1.7589068587790848, + "learning_rate": 5.0899097513055214e-08, + "loss": 0.7986, + "step": 10314 + }, + { + "epoch": 0.9302430445957524, + "grad_norm": 1.6799912288981764, + "learning_rate": 5.076821917694563e-08, + "loss": 0.73, + "step": 10315 + }, + { + "epoch": 0.9303332281192226, + "grad_norm": 1.8659311321343466, + "learning_rate": 5.063750716003889e-08, + "loss": 0.7882, + "step": 10316 + }, + { + "epoch": 0.9304234116426929, + "grad_norm": 3.8007950503186, + "learning_rate": 5.050696147348921e-08, + "loss": 0.725, + "step": 10317 + }, + { + "epoch": 0.9305135951661632, + "grad_norm": 1.9819140455621636, + "learning_rate": 5.037658212843454e-08, + "loss": 0.7133, + "step": 10318 + }, + { + "epoch": 0.9306037786896334, + "grad_norm": 1.457697248247251, + "learning_rate": 5.0246369136000444e-08, + "loss": 0.6321, + "step": 10319 + }, + { + "epoch": 0.9306939622131036, + "grad_norm": 1.6127563805028902, + "learning_rate": 5.011632250729691e-08, + "loss": 0.8191, + "step": 10320 + }, + { + "epoch": 0.930784145736574, + "grad_norm": 2.164322159265606, + "learning_rate": 4.998644225342019e-08, + "loss": 0.8322, + "step": 10321 + }, + { + "epoch": 0.9308743292600442, + "grad_norm": 1.9301407625325406, + "learning_rate": 4.9856728385452296e-08, + "loss": 0.7444, + "step": 10322 + }, + { + "epoch": 0.9309645127835144, + "grad_norm": 1.8356358057054085, + "learning_rate": 4.9727180914461485e-08, + "loss": 0.8316, + "step": 10323 + }, + { + "epoch": 0.9310546963069847, + "grad_norm": 1.9371306774939663, + "learning_rate": 4.959779985150137e-08, + "loss": 0.7511, + "step": 10324 + }, + { + "epoch": 0.931144879830455, + "grad_norm": 1.9850403278587838, + "learning_rate": 4.9468585207611105e-08, + "loss": 0.6936, + "step": 10325 + }, + { + "epoch": 0.9312350633539253, + "grad_norm": 1.8668593114264578, + "learning_rate": 4.9339536993816764e-08, + "loss": 0.7205, + "step": 10326 + }, + { + "epoch": 0.9313252468773955, + "grad_norm": 1.7280021637620449, + "learning_rate": 4.921065522112844e-08, + "loss": 0.6937, + "step": 10327 + }, + { + "epoch": 0.9314154304008657, + "grad_norm": 1.5885567669593734, + "learning_rate": 4.908193990054377e-08, + "loss": 0.7865, + "step": 10328 + }, + { + "epoch": 0.9315056139243361, + "grad_norm": 1.6347573848685322, + "learning_rate": 4.89533910430453e-08, + "loss": 0.7579, + "step": 10329 + }, + { + "epoch": 0.9315957974478063, + "grad_norm": 1.4752191154532377, + "learning_rate": 4.8825008659601376e-08, + "loss": 0.7318, + "step": 10330 + }, + { + "epoch": 0.9316859809712765, + "grad_norm": 1.8904082749418518, + "learning_rate": 4.869679276116634e-08, + "loss": 0.7517, + "step": 10331 + }, + { + "epoch": 0.9317761644947468, + "grad_norm": 1.492482101358665, + "learning_rate": 4.856874335868055e-08, + "loss": 0.7558, + "step": 10332 + }, + { + "epoch": 0.9318663480182171, + "grad_norm": 2.6986985319458903, + "learning_rate": 4.844086046306928e-08, + "loss": 0.8481, + "step": 10333 + }, + { + "epoch": 0.9319565315416873, + "grad_norm": 1.6472966734665067, + "learning_rate": 4.8313144085244896e-08, + "loss": 0.74, + "step": 10334 + }, + { + "epoch": 0.9320467150651576, + "grad_norm": 0.6498171136400057, + "learning_rate": 4.818559423610424e-08, + "loss": 0.5746, + "step": 10335 + }, + { + "epoch": 0.9321368985886278, + "grad_norm": 1.6110906252526889, + "learning_rate": 4.8058210926531284e-08, + "loss": 0.6853, + "step": 10336 + }, + { + "epoch": 0.9322270821120981, + "grad_norm": 1.5233570566702126, + "learning_rate": 4.7930994167394435e-08, + "loss": 0.7766, + "step": 10337 + }, + { + "epoch": 0.9323172656355684, + "grad_norm": 1.504583746164301, + "learning_rate": 4.7803943969548786e-08, + "loss": 0.7859, + "step": 10338 + }, + { + "epoch": 0.9324074491590386, + "grad_norm": 1.9236056591313202, + "learning_rate": 4.7677060343834784e-08, + "loss": 0.6879, + "step": 10339 + }, + { + "epoch": 0.932497632682509, + "grad_norm": 1.7384241624012278, + "learning_rate": 4.75503433010791e-08, + "loss": 0.7361, + "step": 10340 + }, + { + "epoch": 0.9325878162059792, + "grad_norm": 2.335851457216808, + "learning_rate": 4.742379285209419e-08, + "loss": 0.7263, + "step": 10341 + }, + { + "epoch": 0.9326779997294494, + "grad_norm": 2.9161711317271903, + "learning_rate": 4.72974090076772e-08, + "loss": 0.8408, + "step": 10342 + }, + { + "epoch": 0.9327681832529197, + "grad_norm": 1.4658208817818266, + "learning_rate": 4.717119177861262e-08, + "loss": 0.7003, + "step": 10343 + }, + { + "epoch": 0.93285836677639, + "grad_norm": 1.779460677511776, + "learning_rate": 4.70451411756696e-08, + "loss": 0.7576, + "step": 10344 + }, + { + "epoch": 0.9329485502998602, + "grad_norm": 1.6395312341784403, + "learning_rate": 4.691925720960355e-08, + "loss": 0.7799, + "step": 10345 + }, + { + "epoch": 0.9330387338233305, + "grad_norm": 2.133945262088931, + "learning_rate": 4.6793539891155645e-08, + "loss": 0.7691, + "step": 10346 + }, + { + "epoch": 0.9331289173468007, + "grad_norm": 3.285434053447986, + "learning_rate": 4.6667989231052864e-08, + "loss": 0.7234, + "step": 10347 + }, + { + "epoch": 0.933219100870271, + "grad_norm": 1.7229867907840821, + "learning_rate": 4.654260524000797e-08, + "loss": 0.755, + "step": 10348 + }, + { + "epoch": 0.9333092843937413, + "grad_norm": 1.56817344618059, + "learning_rate": 4.6417387928719076e-08, + "loss": 0.7307, + "step": 10349 + }, + { + "epoch": 0.9333994679172115, + "grad_norm": 1.7758946138809042, + "learning_rate": 4.629233730787052e-08, + "loss": 0.6154, + "step": 10350 + }, + { + "epoch": 0.9334896514406817, + "grad_norm": 5.983697041089314, + "learning_rate": 4.616745338813266e-08, + "loss": 0.7049, + "step": 10351 + }, + { + "epoch": 0.9335798349641521, + "grad_norm": 2.073866877191085, + "learning_rate": 4.6042736180160744e-08, + "loss": 0.7592, + "step": 10352 + }, + { + "epoch": 0.9336700184876223, + "grad_norm": 1.923842996452673, + "learning_rate": 4.591818569459671e-08, + "loss": 0.8596, + "step": 10353 + }, + { + "epoch": 0.9337602020110926, + "grad_norm": 1.635778407364854, + "learning_rate": 4.5793801942067614e-08, + "loss": 0.7645, + "step": 10354 + }, + { + "epoch": 0.9338503855345628, + "grad_norm": 2.4120831816328323, + "learning_rate": 4.566958493318673e-08, + "loss": 0.6128, + "step": 10355 + }, + { + "epoch": 0.9339405690580331, + "grad_norm": 0.605535318960522, + "learning_rate": 4.554553467855316e-08, + "loss": 0.5991, + "step": 10356 + }, + { + "epoch": 0.9340307525815034, + "grad_norm": 6.319829028750439, + "learning_rate": 4.5421651188751074e-08, + "loss": 0.7609, + "step": 10357 + }, + { + "epoch": 0.9341209361049736, + "grad_norm": 2.84741695534144, + "learning_rate": 4.529793447435137e-08, + "loss": 0.7551, + "step": 10358 + }, + { + "epoch": 0.9342111196284438, + "grad_norm": 2.4635679688504633, + "learning_rate": 4.5174384545909824e-08, + "loss": 0.7453, + "step": 10359 + }, + { + "epoch": 0.9343013031519142, + "grad_norm": 2.384862399500395, + "learning_rate": 4.505100141396867e-08, + "loss": 0.7855, + "step": 10360 + }, + { + "epoch": 0.9343914866753844, + "grad_norm": 2.1198813554565166, + "learning_rate": 4.492778508905548e-08, + "loss": 0.7803, + "step": 10361 + }, + { + "epoch": 0.9344816701988546, + "grad_norm": 1.6827642605667097, + "learning_rate": 4.480473558168385e-08, + "loss": 0.6821, + "step": 10362 + }, + { + "epoch": 0.934571853722325, + "grad_norm": 5.149439897911911, + "learning_rate": 4.4681852902352936e-08, + "loss": 0.6461, + "step": 10363 + }, + { + "epoch": 0.9346620372457952, + "grad_norm": 1.8402471727749312, + "learning_rate": 4.455913706154812e-08, + "loss": 0.6639, + "step": 10364 + }, + { + "epoch": 0.9347522207692655, + "grad_norm": 1.8403376019748534, + "learning_rate": 4.443658806973949e-08, + "loss": 0.7672, + "step": 10365 + }, + { + "epoch": 0.9348424042927357, + "grad_norm": 0.6668038947211347, + "learning_rate": 4.431420593738444e-08, + "loss": 0.6012, + "step": 10366 + }, + { + "epoch": 0.934932587816206, + "grad_norm": 1.8609023032395196, + "learning_rate": 4.419199067492485e-08, + "loss": 0.6551, + "step": 10367 + }, + { + "epoch": 0.9350227713396763, + "grad_norm": 1.5949103675219736, + "learning_rate": 4.4069942292788596e-08, + "loss": 0.7489, + "step": 10368 + }, + { + "epoch": 0.9351129548631465, + "grad_norm": 1.784144609479167, + "learning_rate": 4.39480608013898e-08, + "loss": 0.7479, + "step": 10369 + }, + { + "epoch": 0.9352031383866167, + "grad_norm": 1.8320431576730223, + "learning_rate": 4.3826346211128126e-08, + "loss": 0.8332, + "step": 10370 + }, + { + "epoch": 0.9352933219100871, + "grad_norm": 1.683586740147276, + "learning_rate": 4.370479853238884e-08, + "loss": 0.7473, + "step": 10371 + }, + { + "epoch": 0.9353835054335573, + "grad_norm": 1.6185405433674784, + "learning_rate": 4.3583417775542756e-08, + "loss": 0.706, + "step": 10372 + }, + { + "epoch": 0.9354736889570275, + "grad_norm": 1.9927486452158596, + "learning_rate": 4.3462203950947575e-08, + "loss": 0.6935, + "step": 10373 + }, + { + "epoch": 0.9355638724804978, + "grad_norm": 1.8676488458025018, + "learning_rate": 4.3341157068944814e-08, + "loss": 0.7291, + "step": 10374 + }, + { + "epoch": 0.9356540560039681, + "grad_norm": 1.7322714481179327, + "learning_rate": 4.322027713986376e-08, + "loss": 0.7574, + "step": 10375 + }, + { + "epoch": 0.9357442395274383, + "grad_norm": 1.912077454406696, + "learning_rate": 4.309956417401816e-08, + "loss": 0.8135, + "step": 10376 + }, + { + "epoch": 0.9358344230509086, + "grad_norm": 1.831062825502749, + "learning_rate": 4.297901818170801e-08, + "loss": 0.7128, + "step": 10377 + }, + { + "epoch": 0.9359246065743788, + "grad_norm": 1.659780015387024, + "learning_rate": 4.285863917321886e-08, + "loss": 0.673, + "step": 10378 + }, + { + "epoch": 0.9360147900978492, + "grad_norm": 1.943958531322012, + "learning_rate": 4.2738427158822253e-08, + "loss": 0.6816, + "step": 10379 + }, + { + "epoch": 0.9361049736213194, + "grad_norm": 1.6434414621157776, + "learning_rate": 4.261838214877511e-08, + "loss": 0.7539, + "step": 10380 + }, + { + "epoch": 0.9361951571447896, + "grad_norm": 2.1056341838777923, + "learning_rate": 4.249850415332079e-08, + "loss": 0.7489, + "step": 10381 + }, + { + "epoch": 0.9362853406682599, + "grad_norm": 1.7858656928043597, + "learning_rate": 4.237879318268756e-08, + "loss": 0.8012, + "step": 10382 + }, + { + "epoch": 0.9363755241917302, + "grad_norm": 1.482108624599734, + "learning_rate": 4.225924924708968e-08, + "loss": 0.6722, + "step": 10383 + }, + { + "epoch": 0.9364657077152004, + "grad_norm": 2.9195967171065855, + "learning_rate": 4.2139872356727665e-08, + "loss": 0.7558, + "step": 10384 + }, + { + "epoch": 0.9365558912386707, + "grad_norm": 1.7444305349849674, + "learning_rate": 4.202066252178738e-08, + "loss": 0.7077, + "step": 10385 + }, + { + "epoch": 0.936646074762141, + "grad_norm": 1.745580542326402, + "learning_rate": 4.1901619752440445e-08, + "loss": 0.8365, + "step": 10386 + }, + { + "epoch": 0.9367362582856112, + "grad_norm": 1.5501137937797926, + "learning_rate": 4.178274405884363e-08, + "loss": 0.7416, + "step": 10387 + }, + { + "epoch": 0.9368264418090815, + "grad_norm": 2.2904487032006213, + "learning_rate": 4.166403545114105e-08, + "loss": 0.7511, + "step": 10388 + }, + { + "epoch": 0.9369166253325517, + "grad_norm": 1.7328483804852537, + "learning_rate": 4.154549393946083e-08, + "loss": 0.7976, + "step": 10389 + }, + { + "epoch": 0.937006808856022, + "grad_norm": 2.3828804381644226, + "learning_rate": 4.14271195339182e-08, + "loss": 0.7746, + "step": 10390 + }, + { + "epoch": 0.9370969923794923, + "grad_norm": 2.032664114882403, + "learning_rate": 4.1308912244613084e-08, + "loss": 0.7101, + "step": 10391 + }, + { + "epoch": 0.9371871759029625, + "grad_norm": 1.7418335696228664, + "learning_rate": 4.1190872081631636e-08, + "loss": 0.6996, + "step": 10392 + }, + { + "epoch": 0.9372773594264328, + "grad_norm": 1.7432764866582544, + "learning_rate": 4.107299905504558e-08, + "loss": 0.8099, + "step": 10393 + }, + { + "epoch": 0.9373675429499031, + "grad_norm": 1.5867137042988537, + "learning_rate": 4.095529317491286e-08, + "loss": 0.7387, + "step": 10394 + }, + { + "epoch": 0.9374577264733733, + "grad_norm": 1.8340470990879634, + "learning_rate": 4.0837754451276575e-08, + "loss": 0.7506, + "step": 10395 + }, + { + "epoch": 0.9375479099968436, + "grad_norm": 2.508420496172864, + "learning_rate": 4.072038289416557e-08, + "loss": 0.7307, + "step": 10396 + }, + { + "epoch": 0.9376380935203138, + "grad_norm": 1.526899190141929, + "learning_rate": 4.0603178513595185e-08, + "loss": 0.7434, + "step": 10397 + }, + { + "epoch": 0.9377282770437841, + "grad_norm": 2.9663342736156944, + "learning_rate": 4.0486141319565624e-08, + "loss": 0.7512, + "step": 10398 + }, + { + "epoch": 0.9378184605672544, + "grad_norm": 1.8408071016601553, + "learning_rate": 4.0369271322062916e-08, + "loss": 0.686, + "step": 10399 + }, + { + "epoch": 0.9379086440907246, + "grad_norm": 1.709892058060896, + "learning_rate": 4.0252568531059295e-08, + "loss": 0.7647, + "step": 10400 + }, + { + "epoch": 0.9379988276141948, + "grad_norm": 10.446913808282128, + "learning_rate": 4.013603295651235e-08, + "loss": 0.7437, + "step": 10401 + }, + { + "epoch": 0.9380890111376652, + "grad_norm": 1.59670492263124, + "learning_rate": 4.001966460836592e-08, + "loss": 0.7732, + "step": 10402 + }, + { + "epoch": 0.9381791946611354, + "grad_norm": 1.7046646018723768, + "learning_rate": 3.990346349654894e-08, + "loss": 0.8196, + "step": 10403 + }, + { + "epoch": 0.9382693781846057, + "grad_norm": 1.6221066906620238, + "learning_rate": 3.9787429630975924e-08, + "loss": 0.7178, + "step": 10404 + }, + { + "epoch": 0.9383595617080759, + "grad_norm": 2.767369985998312, + "learning_rate": 3.967156302154828e-08, + "loss": 0.7322, + "step": 10405 + }, + { + "epoch": 0.9384497452315462, + "grad_norm": 3.8504385710768365, + "learning_rate": 3.955586367815189e-08, + "loss": 0.7004, + "step": 10406 + }, + { + "epoch": 0.9385399287550165, + "grad_norm": 1.7259290404822996, + "learning_rate": 3.944033161065907e-08, + "loss": 0.8125, + "step": 10407 + }, + { + "epoch": 0.9386301122784867, + "grad_norm": 1.5786796577108104, + "learning_rate": 3.93249668289275e-08, + "loss": 0.6277, + "step": 10408 + }, + { + "epoch": 0.9387202958019569, + "grad_norm": 1.5362252354934383, + "learning_rate": 3.920976934280063e-08, + "loss": 0.6706, + "step": 10409 + }, + { + "epoch": 0.9388104793254273, + "grad_norm": 1.7435164237051552, + "learning_rate": 3.909473916210815e-08, + "loss": 0.8267, + "step": 10410 + }, + { + "epoch": 0.9389006628488975, + "grad_norm": 3.3303110885796103, + "learning_rate": 3.897987629666488e-08, + "loss": 0.7004, + "step": 10411 + }, + { + "epoch": 0.9389908463723677, + "grad_norm": 1.5055403909627236, + "learning_rate": 3.886518075627143e-08, + "loss": 0.8152, + "step": 10412 + }, + { + "epoch": 0.9390810298958381, + "grad_norm": 2.1890320998374264, + "learning_rate": 3.875065255071419e-08, + "loss": 0.7445, + "step": 10413 + }, + { + "epoch": 0.9391712134193083, + "grad_norm": 2.18147275820153, + "learning_rate": 3.863629168976579e-08, + "loss": 0.7165, + "step": 10414 + }, + { + "epoch": 0.9392613969427785, + "grad_norm": 1.7718893006839265, + "learning_rate": 3.852209818318375e-08, + "loss": 0.7989, + "step": 10415 + }, + { + "epoch": 0.9393515804662488, + "grad_norm": 12.323202835505063, + "learning_rate": 3.840807204071161e-08, + "loss": 0.6505, + "step": 10416 + }, + { + "epoch": 0.9394417639897191, + "grad_norm": 1.6350645751333361, + "learning_rate": 3.829421327207894e-08, + "loss": 0.7259, + "step": 10417 + }, + { + "epoch": 0.9395319475131894, + "grad_norm": 1.57013652062166, + "learning_rate": 3.8180521887000825e-08, + "loss": 0.6983, + "step": 10418 + }, + { + "epoch": 0.9396221310366596, + "grad_norm": 0.5915721554781802, + "learning_rate": 3.806699789517775e-08, + "loss": 0.61, + "step": 10419 + }, + { + "epoch": 0.9397123145601298, + "grad_norm": 1.5615393164843048, + "learning_rate": 3.7953641306296635e-08, + "loss": 0.7259, + "step": 10420 + }, + { + "epoch": 0.9398024980836002, + "grad_norm": 1.7261535022014212, + "learning_rate": 3.784045213002951e-08, + "loss": 0.5902, + "step": 10421 + }, + { + "epoch": 0.9398926816070704, + "grad_norm": 2.0025756463515556, + "learning_rate": 3.7727430376033986e-08, + "loss": 0.7571, + "step": 10422 + }, + { + "epoch": 0.9399828651305406, + "grad_norm": 1.682093117475776, + "learning_rate": 3.7614576053954126e-08, + "loss": 0.7475, + "step": 10423 + }, + { + "epoch": 0.9400730486540109, + "grad_norm": 1.6846318065089436, + "learning_rate": 3.75018891734189e-08, + "loss": 0.7397, + "step": 10424 + }, + { + "epoch": 0.9401632321774812, + "grad_norm": 1.7974707901110634, + "learning_rate": 3.738936974404372e-08, + "loss": 0.7905, + "step": 10425 + }, + { + "epoch": 0.9402534157009514, + "grad_norm": 1.693714003514733, + "learning_rate": 3.7277017775429354e-08, + "loss": 0.8071, + "step": 10426 + }, + { + "epoch": 0.9403435992244217, + "grad_norm": 1.5505987487563555, + "learning_rate": 3.7164833277162136e-08, + "loss": 0.7681, + "step": 10427 + }, + { + "epoch": 0.9404337827478919, + "grad_norm": 1.3566327113198904, + "learning_rate": 3.705281625881418e-08, + "loss": 0.7246, + "step": 10428 + }, + { + "epoch": 0.9405239662713623, + "grad_norm": 1.6081891181341386, + "learning_rate": 3.694096672994362e-08, + "loss": 0.6529, + "step": 10429 + }, + { + "epoch": 0.9406141497948325, + "grad_norm": 1.5016181482965159, + "learning_rate": 3.682928470009394e-08, + "loss": 0.7985, + "step": 10430 + }, + { + "epoch": 0.9407043333183027, + "grad_norm": 2.4337906612375377, + "learning_rate": 3.6717770178794406e-08, + "loss": 0.705, + "step": 10431 + }, + { + "epoch": 0.940794516841773, + "grad_norm": 1.4060176657439167, + "learning_rate": 3.6606423175560287e-08, + "loss": 0.8492, + "step": 10432 + }, + { + "epoch": 0.9408847003652433, + "grad_norm": 0.6257891679470035, + "learning_rate": 3.649524369989221e-08, + "loss": 0.5463, + "step": 10433 + }, + { + "epoch": 0.9409748838887135, + "grad_norm": 2.8590040956503215, + "learning_rate": 3.638423176127636e-08, + "loss": 0.7769, + "step": 10434 + }, + { + "epoch": 0.9410650674121838, + "grad_norm": 1.4823858485411543, + "learning_rate": 3.6273387369185396e-08, + "loss": 0.7801, + "step": 10435 + }, + { + "epoch": 0.9411552509356541, + "grad_norm": 1.41830734539915, + "learning_rate": 3.616271053307685e-08, + "loss": 0.6871, + "step": 10436 + }, + { + "epoch": 0.9412454344591243, + "grad_norm": 1.8111186189567383, + "learning_rate": 3.6052201262394275e-08, + "loss": 0.7644, + "step": 10437 + }, + { + "epoch": 0.9413356179825946, + "grad_norm": 1.6004801744915274, + "learning_rate": 3.5941859566566816e-08, + "loss": 0.7994, + "step": 10438 + }, + { + "epoch": 0.9414258015060648, + "grad_norm": 0.6852883180108273, + "learning_rate": 3.583168545500981e-08, + "loss": 0.5685, + "step": 10439 + }, + { + "epoch": 0.9415159850295352, + "grad_norm": 1.5791245873446995, + "learning_rate": 3.5721678937123746e-08, + "loss": 0.6898, + "step": 10440 + }, + { + "epoch": 0.9416061685530054, + "grad_norm": 1.8147809376153317, + "learning_rate": 3.561184002229467e-08, + "loss": 0.8261, + "step": 10441 + }, + { + "epoch": 0.9416963520764756, + "grad_norm": 1.461806502937177, + "learning_rate": 3.550216871989531e-08, + "loss": 0.6912, + "step": 10442 + }, + { + "epoch": 0.9417865355999459, + "grad_norm": 0.6391463777296976, + "learning_rate": 3.539266503928262e-08, + "loss": 0.6469, + "step": 10443 + }, + { + "epoch": 0.9418767191234162, + "grad_norm": 1.6647261608024921, + "learning_rate": 3.528332898980091e-08, + "loss": 0.7566, + "step": 10444 + }, + { + "epoch": 0.9419669026468864, + "grad_norm": 1.8013435859378992, + "learning_rate": 3.517416058077849e-08, + "loss": 0.7133, + "step": 10445 + }, + { + "epoch": 0.9420570861703567, + "grad_norm": 2.076942968137428, + "learning_rate": 3.506515982153102e-08, + "loss": 0.6504, + "step": 10446 + }, + { + "epoch": 0.9421472696938269, + "grad_norm": 2.0096265857821014, + "learning_rate": 3.495632672135862e-08, + "loss": 0.7492, + "step": 10447 + }, + { + "epoch": 0.9422374532172972, + "grad_norm": 1.4837032203402316, + "learning_rate": 3.4847661289547417e-08, + "loss": 0.7754, + "step": 10448 + }, + { + "epoch": 0.9423276367407675, + "grad_norm": 2.2388453014202225, + "learning_rate": 3.473916353536932e-08, + "loss": 0.6986, + "step": 10449 + }, + { + "epoch": 0.9424178202642377, + "grad_norm": 1.5973461941507532, + "learning_rate": 3.463083346808249e-08, + "loss": 0.7122, + "step": 10450 + }, + { + "epoch": 0.9425080037877079, + "grad_norm": 1.8656692397401065, + "learning_rate": 3.452267109692975e-08, + "loss": 0.6451, + "step": 10451 + }, + { + "epoch": 0.9425981873111783, + "grad_norm": 1.7660918167240025, + "learning_rate": 3.441467643114016e-08, + "loss": 0.7455, + "step": 10452 + }, + { + "epoch": 0.9426883708346485, + "grad_norm": 0.5907831101487887, + "learning_rate": 3.430684947992857e-08, + "loss": 0.5576, + "step": 10453 + }, + { + "epoch": 0.9427785543581187, + "grad_norm": 4.6740197199420574, + "learning_rate": 3.419919025249518e-08, + "loss": 0.7598, + "step": 10454 + }, + { + "epoch": 0.942868737881589, + "grad_norm": 1.6137888112265466, + "learning_rate": 3.40916987580262e-08, + "loss": 0.7636, + "step": 10455 + }, + { + "epoch": 0.9429589214050593, + "grad_norm": 2.3212905004334, + "learning_rate": 3.398437500569362e-08, + "loss": 0.7232, + "step": 10456 + }, + { + "epoch": 0.9430491049285296, + "grad_norm": 1.5563248489948218, + "learning_rate": 3.3877219004654347e-08, + "loss": 0.6242, + "step": 10457 + }, + { + "epoch": 0.9431392884519998, + "grad_norm": 1.865791788744699, + "learning_rate": 3.3770230764051946e-08, + "loss": 0.7764, + "step": 10458 + }, + { + "epoch": 0.9432294719754701, + "grad_norm": 1.8477356714191933, + "learning_rate": 3.366341029301534e-08, + "loss": 0.7667, + "step": 10459 + }, + { + "epoch": 0.9433196554989404, + "grad_norm": 1.9273354099125706, + "learning_rate": 3.355675760065857e-08, + "loss": 0.7548, + "step": 10460 + }, + { + "epoch": 0.9434098390224106, + "grad_norm": 0.5762930015357463, + "learning_rate": 3.345027269608236e-08, + "loss": 0.5251, + "step": 10461 + }, + { + "epoch": 0.9435000225458808, + "grad_norm": 0.7446820549679796, + "learning_rate": 3.334395558837211e-08, + "loss": 0.5926, + "step": 10462 + }, + { + "epoch": 0.9435902060693512, + "grad_norm": 1.6438499362285917, + "learning_rate": 3.3237806286599667e-08, + "loss": 0.6529, + "step": 10463 + }, + { + "epoch": 0.9436803895928214, + "grad_norm": 1.7726311552003944, + "learning_rate": 3.313182479982224e-08, + "loss": 0.7335, + "step": 10464 + }, + { + "epoch": 0.9437705731162916, + "grad_norm": 1.6469660695789947, + "learning_rate": 3.302601113708259e-08, + "loss": 0.8127, + "step": 10465 + }, + { + "epoch": 0.9438607566397619, + "grad_norm": 1.2854894420645722, + "learning_rate": 3.292036530740972e-08, + "loss": 0.7944, + "step": 10466 + }, + { + "epoch": 0.9439509401632322, + "grad_norm": 1.7167404911993935, + "learning_rate": 3.2814887319817294e-08, + "loss": 0.7345, + "step": 10467 + }, + { + "epoch": 0.9440411236867025, + "grad_norm": 1.737495776384009, + "learning_rate": 3.270957718330591e-08, + "loss": 0.7872, + "step": 10468 + }, + { + "epoch": 0.9441313072101727, + "grad_norm": 1.3483500588224127, + "learning_rate": 3.260443490686082e-08, + "loss": 0.7915, + "step": 10469 + }, + { + "epoch": 0.9442214907336429, + "grad_norm": 1.9241353210874232, + "learning_rate": 3.249946049945351e-08, + "loss": 0.7637, + "step": 10470 + }, + { + "epoch": 0.9443116742571133, + "grad_norm": 1.8883103001876895, + "learning_rate": 3.239465397004082e-08, + "loss": 0.7595, + "step": 10471 + }, + { + "epoch": 0.9444018577805835, + "grad_norm": 2.3252498615503963, + "learning_rate": 3.229001532756559e-08, + "loss": 0.7083, + "step": 10472 + }, + { + "epoch": 0.9444920413040537, + "grad_norm": 4.238211044925456, + "learning_rate": 3.218554458095602e-08, + "loss": 0.7422, + "step": 10473 + }, + { + "epoch": 0.944582224827524, + "grad_norm": 6.713508015960888, + "learning_rate": 3.20812417391263e-08, + "loss": 0.7478, + "step": 10474 + }, + { + "epoch": 0.9446724083509943, + "grad_norm": 1.2552812114573975, + "learning_rate": 3.1977106810975764e-08, + "loss": 0.7375, + "step": 10475 + }, + { + "epoch": 0.9447625918744645, + "grad_norm": 0.7820750648198288, + "learning_rate": 3.187313980539042e-08, + "loss": 0.6218, + "step": 10476 + }, + { + "epoch": 0.9448527753979348, + "grad_norm": 1.4874871004971801, + "learning_rate": 3.176934073124071e-08, + "loss": 0.8269, + "step": 10477 + }, + { + "epoch": 0.944942958921405, + "grad_norm": 1.7121517801919042, + "learning_rate": 3.166570959738357e-08, + "loss": 0.7491, + "step": 10478 + }, + { + "epoch": 0.9450331424448754, + "grad_norm": 1.6830512144033423, + "learning_rate": 3.1562246412661476e-08, + "loss": 0.6827, + "step": 10479 + }, + { + "epoch": 0.9451233259683456, + "grad_norm": 1.90855774566381, + "learning_rate": 3.145895118590225e-08, + "loss": 0.7812, + "step": 10480 + }, + { + "epoch": 0.9452135094918158, + "grad_norm": 1.7192473645608175, + "learning_rate": 3.135582392591996e-08, + "loss": 0.6559, + "step": 10481 + }, + { + "epoch": 0.9453036930152862, + "grad_norm": 0.6027120283215532, + "learning_rate": 3.125286464151333e-08, + "loss": 0.536, + "step": 10482 + }, + { + "epoch": 0.9453938765387564, + "grad_norm": 1.8087604355263884, + "learning_rate": 3.115007334146824e-08, + "loss": 0.7197, + "step": 10483 + }, + { + "epoch": 0.9454840600622266, + "grad_norm": 0.6309510391770325, + "learning_rate": 3.104745003455478e-08, + "loss": 0.5571, + "step": 10484 + }, + { + "epoch": 0.9455742435856969, + "grad_norm": 1.5176813392446569, + "learning_rate": 3.094499472952972e-08, + "loss": 0.7659, + "step": 10485 + }, + { + "epoch": 0.9456644271091672, + "grad_norm": 1.9007280134529712, + "learning_rate": 3.084270743513495e-08, + "loss": 0.717, + "step": 10486 + }, + { + "epoch": 0.9457546106326374, + "grad_norm": 1.9698326524827512, + "learning_rate": 3.074058816009817e-08, + "loss": 0.6375, + "step": 10487 + }, + { + "epoch": 0.9458447941561077, + "grad_norm": 1.8377417606464332, + "learning_rate": 3.063863691313284e-08, + "loss": 0.6977, + "step": 10488 + }, + { + "epoch": 0.9459349776795779, + "grad_norm": 1.748057990864622, + "learning_rate": 3.0536853702937794e-08, + "loss": 0.6827, + "step": 10489 + }, + { + "epoch": 0.9460251612030482, + "grad_norm": 2.338013405949882, + "learning_rate": 3.043523853819807e-08, + "loss": 0.7451, + "step": 10490 + }, + { + "epoch": 0.9461153447265185, + "grad_norm": 1.6670910136109407, + "learning_rate": 3.0333791427583855e-08, + "loss": 0.771, + "step": 10491 + }, + { + "epoch": 0.9462055282499887, + "grad_norm": 1.6647599598829679, + "learning_rate": 3.023251237975111e-08, + "loss": 0.866, + "step": 10492 + }, + { + "epoch": 0.946295711773459, + "grad_norm": 0.6561555340305126, + "learning_rate": 3.0131401403341584e-08, + "loss": 0.5941, + "step": 10493 + }, + { + "epoch": 0.9463858952969293, + "grad_norm": 2.681837815093837, + "learning_rate": 3.00304585069826e-08, + "loss": 0.7255, + "step": 10494 + }, + { + "epoch": 0.9464760788203995, + "grad_norm": 1.6653159708340168, + "learning_rate": 2.992968369928728e-08, + "loss": 0.7353, + "step": 10495 + }, + { + "epoch": 0.9465662623438698, + "grad_norm": 1.744516022577722, + "learning_rate": 2.982907698885429e-08, + "loss": 0.7635, + "step": 10496 + }, + { + "epoch": 0.94665644586734, + "grad_norm": 2.157718655854151, + "learning_rate": 2.9728638384267645e-08, + "loss": 0.7546, + "step": 10497 + }, + { + "epoch": 0.9467466293908103, + "grad_norm": 2.0688040366131197, + "learning_rate": 2.962836789409784e-08, + "loss": 0.7498, + "step": 10498 + }, + { + "epoch": 0.9468368129142806, + "grad_norm": 1.472375412934201, + "learning_rate": 2.95282655268998e-08, + "loss": 0.8396, + "step": 10499 + }, + { + "epoch": 0.9469269964377508, + "grad_norm": 1.9726088112140907, + "learning_rate": 2.942833129121558e-08, + "loss": 0.747, + "step": 10500 + }, + { + "epoch": 0.947017179961221, + "grad_norm": 1.937608100582644, + "learning_rate": 2.9328565195571475e-08, + "loss": 0.6732, + "step": 10501 + }, + { + "epoch": 0.9471073634846914, + "grad_norm": 1.4972153883860433, + "learning_rate": 2.9228967248480675e-08, + "loss": 0.8139, + "step": 10502 + }, + { + "epoch": 0.9471975470081616, + "grad_norm": 1.9009437500980895, + "learning_rate": 2.912953745844082e-08, + "loss": 0.7835, + "step": 10503 + }, + { + "epoch": 0.9472877305316318, + "grad_norm": 1.9079497841718762, + "learning_rate": 2.9030275833936247e-08, + "loss": 0.7368, + "step": 10504 + }, + { + "epoch": 0.9473779140551022, + "grad_norm": 2.014153705008608, + "learning_rate": 2.893118238343617e-08, + "loss": 0.6473, + "step": 10505 + }, + { + "epoch": 0.9474680975785724, + "grad_norm": 2.19247035338172, + "learning_rate": 2.8832257115396052e-08, + "loss": 0.6572, + "step": 10506 + }, + { + "epoch": 0.9475582811020427, + "grad_norm": 2.096350236656028, + "learning_rate": 2.873350003825692e-08, + "loss": 0.7263, + "step": 10507 + }, + { + "epoch": 0.9476484646255129, + "grad_norm": 0.5640374471326202, + "learning_rate": 2.8634911160444696e-08, + "loss": 0.4722, + "step": 10508 + }, + { + "epoch": 0.9477386481489832, + "grad_norm": 1.6810853331520286, + "learning_rate": 2.853649049037199e-08, + "loss": 0.8082, + "step": 10509 + }, + { + "epoch": 0.9478288316724535, + "grad_norm": 1.9509959098581429, + "learning_rate": 2.8438238036436525e-08, + "loss": 0.798, + "step": 10510 + }, + { + "epoch": 0.9479190151959237, + "grad_norm": 4.02244090357508, + "learning_rate": 2.834015380702137e-08, + "loss": 0.7411, + "step": 10511 + }, + { + "epoch": 0.9480091987193939, + "grad_norm": 0.5810850247709946, + "learning_rate": 2.824223781049606e-08, + "loss": 0.5129, + "step": 10512 + }, + { + "epoch": 0.9480993822428643, + "grad_norm": 1.9664096545802618, + "learning_rate": 2.8144490055215465e-08, + "loss": 0.8091, + "step": 10513 + }, + { + "epoch": 0.9481895657663345, + "grad_norm": 1.5789267137079992, + "learning_rate": 2.8046910549519355e-08, + "loss": 0.7795, + "step": 10514 + }, + { + "epoch": 0.9482797492898047, + "grad_norm": 1.6489220941946858, + "learning_rate": 2.794949930173418e-08, + "loss": 0.7498, + "step": 10515 + }, + { + "epoch": 0.948369932813275, + "grad_norm": 1.9333975027659942, + "learning_rate": 2.7852256320171296e-08, + "loss": 0.7777, + "step": 10516 + }, + { + "epoch": 0.9484601163367453, + "grad_norm": 1.7623007208529973, + "learning_rate": 2.775518161312851e-08, + "loss": 0.6101, + "step": 10517 + }, + { + "epoch": 0.9485502998602156, + "grad_norm": 1.5160294612497869, + "learning_rate": 2.76582751888883e-08, + "loss": 0.7725, + "step": 10518 + }, + { + "epoch": 0.9486404833836858, + "grad_norm": 1.9778263439092363, + "learning_rate": 2.756153705571962e-08, + "loss": 0.7621, + "step": 10519 + }, + { + "epoch": 0.948730666907156, + "grad_norm": 1.4449172020656393, + "learning_rate": 2.74649672218763e-08, + "loss": 0.7268, + "step": 10520 + }, + { + "epoch": 0.9488208504306264, + "grad_norm": 1.5953060108915909, + "learning_rate": 2.7368565695598424e-08, + "loss": 0.733, + "step": 10521 + }, + { + "epoch": 0.9489110339540966, + "grad_norm": 1.7679237322906243, + "learning_rate": 2.727233248511185e-08, + "loss": 0.6716, + "step": 10522 + }, + { + "epoch": 0.9490012174775668, + "grad_norm": 2.111525957718064, + "learning_rate": 2.71762675986269e-08, + "loss": 0.7013, + "step": 10523 + }, + { + "epoch": 0.9490914010010371, + "grad_norm": 1.4699285121324879, + "learning_rate": 2.7080371044341242e-08, + "loss": 0.7874, + "step": 10524 + }, + { + "epoch": 0.9491815845245074, + "grad_norm": 2.094799974128821, + "learning_rate": 2.6984642830436556e-08, + "loss": 0.7762, + "step": 10525 + }, + { + "epoch": 0.9492717680479776, + "grad_norm": 1.4774720957746459, + "learning_rate": 2.688908296508141e-08, + "loss": 0.73, + "step": 10526 + }, + { + "epoch": 0.9493619515714479, + "grad_norm": 2.573405066299106, + "learning_rate": 2.679369145642929e-08, + "loss": 0.7972, + "step": 10527 + }, + { + "epoch": 0.9494521350949181, + "grad_norm": 2.0484931897494087, + "learning_rate": 2.669846831261946e-08, + "loss": 0.7647, + "step": 10528 + }, + { + "epoch": 0.9495423186183884, + "grad_norm": 2.032559134360935, + "learning_rate": 2.6603413541776976e-08, + "loss": 0.7332, + "step": 10529 + }, + { + "epoch": 0.9496325021418587, + "grad_norm": 1.5531771430428978, + "learning_rate": 2.6508527152012683e-08, + "loss": 0.6973, + "step": 10530 + }, + { + "epoch": 0.9497226856653289, + "grad_norm": 2.090385238515631, + "learning_rate": 2.641380915142233e-08, + "loss": 0.6495, + "step": 10531 + }, + { + "epoch": 0.9498128691887993, + "grad_norm": 2.05711980818143, + "learning_rate": 2.6319259548088334e-08, + "loss": 0.7683, + "step": 10532 + }, + { + "epoch": 0.9499030527122695, + "grad_norm": 1.8578114707693678, + "learning_rate": 2.6224878350077585e-08, + "loss": 0.7446, + "step": 10533 + }, + { + "epoch": 0.9499932362357397, + "grad_norm": 1.700347174929081, + "learning_rate": 2.6130665565443633e-08, + "loss": 0.6619, + "step": 10534 + }, + { + "epoch": 0.95008341975921, + "grad_norm": 1.8610352228675802, + "learning_rate": 2.603662120222494e-08, + "loss": 0.7473, + "step": 10535 + }, + { + "epoch": 0.9501736032826803, + "grad_norm": 2.1912681821687876, + "learning_rate": 2.59427452684462e-08, + "loss": 0.5933, + "step": 10536 + }, + { + "epoch": 0.9502637868061505, + "grad_norm": 1.4266078277439451, + "learning_rate": 2.5849037772117443e-08, + "loss": 0.7273, + "step": 10537 + }, + { + "epoch": 0.9503539703296208, + "grad_norm": 2.31945322483048, + "learning_rate": 2.575549872123384e-08, + "loss": 0.7674, + "step": 10538 + }, + { + "epoch": 0.950444153853091, + "grad_norm": 1.3741631562230736, + "learning_rate": 2.5662128123776994e-08, + "loss": 0.7509, + "step": 10539 + }, + { + "epoch": 0.9505343373765613, + "grad_norm": 1.4835920176161965, + "learning_rate": 2.5568925987713875e-08, + "loss": 0.7426, + "step": 10540 + }, + { + "epoch": 0.9506245209000316, + "grad_norm": 2.6462204016758704, + "learning_rate": 2.5475892320996785e-08, + "loss": 0.7181, + "step": 10541 + }, + { + "epoch": 0.9507147044235018, + "grad_norm": 0.6745697363256169, + "learning_rate": 2.5383027131564038e-08, + "loss": 0.5615, + "step": 10542 + }, + { + "epoch": 0.950804887946972, + "grad_norm": 1.7559320861222418, + "learning_rate": 2.52903304273393e-08, + "loss": 0.7189, + "step": 10543 + }, + { + "epoch": 0.9508950714704424, + "grad_norm": 1.499544233381407, + "learning_rate": 2.519780221623202e-08, + "loss": 0.734, + "step": 10544 + }, + { + "epoch": 0.9509852549939126, + "grad_norm": 1.5818648639355721, + "learning_rate": 2.510544250613722e-08, + "loss": 0.7055, + "step": 10545 + }, + { + "epoch": 0.9510754385173829, + "grad_norm": 2.6235919082214387, + "learning_rate": 2.501325130493548e-08, + "loss": 0.8444, + "step": 10546 + }, + { + "epoch": 0.9511656220408531, + "grad_norm": 0.7146215041246727, + "learning_rate": 2.4921228620493395e-08, + "loss": 0.5784, + "step": 10547 + }, + { + "epoch": 0.9512558055643234, + "grad_norm": 0.7399201946785228, + "learning_rate": 2.4829374460662244e-08, + "loss": 0.6149, + "step": 10548 + }, + { + "epoch": 0.9513459890877937, + "grad_norm": 1.5452553332547967, + "learning_rate": 2.473768883327976e-08, + "loss": 0.7634, + "step": 10549 + }, + { + "epoch": 0.9514361726112639, + "grad_norm": 2.077544740821715, + "learning_rate": 2.464617174616923e-08, + "loss": 0.7093, + "step": 10550 + }, + { + "epoch": 0.9515263561347341, + "grad_norm": 1.631717588560704, + "learning_rate": 2.455482320713953e-08, + "loss": 0.7436, + "step": 10551 + }, + { + "epoch": 0.9516165396582045, + "grad_norm": 2.9959794760027783, + "learning_rate": 2.4463643223984643e-08, + "loss": 0.8071, + "step": 10552 + }, + { + "epoch": 0.9517067231816747, + "grad_norm": 3.8256728434457283, + "learning_rate": 2.4372631804484567e-08, + "loss": 0.817, + "step": 10553 + }, + { + "epoch": 0.9517969067051449, + "grad_norm": 3.190263784139195, + "learning_rate": 2.4281788956405313e-08, + "loss": 0.8232, + "step": 10554 + }, + { + "epoch": 0.9518870902286153, + "grad_norm": 9.745174289669485, + "learning_rate": 2.4191114687497572e-08, + "loss": 0.6774, + "step": 10555 + }, + { + "epoch": 0.9519772737520855, + "grad_norm": 2.152553984245304, + "learning_rate": 2.4100609005498706e-08, + "loss": 0.7068, + "step": 10556 + }, + { + "epoch": 0.9520674572755558, + "grad_norm": 1.5565845666074454, + "learning_rate": 2.4010271918130764e-08, + "loss": 0.6772, + "step": 10557 + }, + { + "epoch": 0.952157640799026, + "grad_norm": 2.0232119637065833, + "learning_rate": 2.39201034331018e-08, + "loss": 0.7064, + "step": 10558 + }, + { + "epoch": 0.9522478243224963, + "grad_norm": 1.9502979417748674, + "learning_rate": 2.3830103558105663e-08, + "loss": 0.7363, + "step": 10559 + }, + { + "epoch": 0.9523380078459666, + "grad_norm": 2.037623222697111, + "learning_rate": 2.374027230082154e-08, + "loss": 0.755, + "step": 10560 + }, + { + "epoch": 0.9524281913694368, + "grad_norm": 0.6790832815488199, + "learning_rate": 2.365060966891441e-08, + "loss": 0.5953, + "step": 10561 + }, + { + "epoch": 0.952518374892907, + "grad_norm": 1.6507939133742873, + "learning_rate": 2.3561115670034827e-08, + "loss": 0.741, + "step": 10562 + }, + { + "epoch": 0.9526085584163774, + "grad_norm": 1.4376525797930273, + "learning_rate": 2.3471790311818675e-08, + "loss": 0.7646, + "step": 10563 + }, + { + "epoch": 0.9526987419398476, + "grad_norm": 2.2150574165635293, + "learning_rate": 2.338263360188808e-08, + "loss": 0.7354, + "step": 10564 + }, + { + "epoch": 0.9527889254633178, + "grad_norm": 1.740652460124788, + "learning_rate": 2.329364554784985e-08, + "loss": 0.6826, + "step": 10565 + }, + { + "epoch": 0.9528791089867881, + "grad_norm": 2.7825849094099846, + "learning_rate": 2.3204826157297465e-08, + "loss": 0.8036, + "step": 10566 + }, + { + "epoch": 0.9529692925102584, + "grad_norm": 1.7117509693506816, + "learning_rate": 2.3116175437809082e-08, + "loss": 0.6817, + "step": 10567 + }, + { + "epoch": 0.9530594760337286, + "grad_norm": 0.6421505743395298, + "learning_rate": 2.30276933969491e-08, + "loss": 0.5957, + "step": 10568 + }, + { + "epoch": 0.9531496595571989, + "grad_norm": 1.4013681232303061, + "learning_rate": 2.2939380042267255e-08, + "loss": 0.801, + "step": 10569 + }, + { + "epoch": 0.9532398430806691, + "grad_norm": 1.8140077567896769, + "learning_rate": 2.2851235381298627e-08, + "loss": 0.651, + "step": 10570 + }, + { + "epoch": 0.9533300266041395, + "grad_norm": 1.8066240629519568, + "learning_rate": 2.2763259421564986e-08, + "loss": 0.675, + "step": 10571 + }, + { + "epoch": 0.9534202101276097, + "grad_norm": 1.9744809615491716, + "learning_rate": 2.2675452170571873e-08, + "loss": 0.7132, + "step": 10572 + }, + { + "epoch": 0.9535103936510799, + "grad_norm": 2.0304319641843804, + "learning_rate": 2.2587813635812414e-08, + "loss": 0.7418, + "step": 10573 + }, + { + "epoch": 0.9536005771745502, + "grad_norm": 2.097215659473023, + "learning_rate": 2.2500343824763958e-08, + "loss": 0.8328, + "step": 10574 + }, + { + "epoch": 0.9536907606980205, + "grad_norm": 2.0083098630470255, + "learning_rate": 2.2413042744890088e-08, + "loss": 0.6948, + "step": 10575 + }, + { + "epoch": 0.9537809442214907, + "grad_norm": 1.7483813429363053, + "learning_rate": 2.2325910403639514e-08, + "loss": 0.7316, + "step": 10576 + }, + { + "epoch": 0.953871127744961, + "grad_norm": 1.6512306017028633, + "learning_rate": 2.223894680844718e-08, + "loss": 0.7038, + "step": 10577 + }, + { + "epoch": 0.9539613112684313, + "grad_norm": 2.025922508625592, + "learning_rate": 2.2152151966733146e-08, + "loss": 0.5833, + "step": 10578 + }, + { + "epoch": 0.9540514947919015, + "grad_norm": 2.3377740587129168, + "learning_rate": 2.2065525885903267e-08, + "loss": 0.7817, + "step": 10579 + }, + { + "epoch": 0.9541416783153718, + "grad_norm": 1.5527902653070138, + "learning_rate": 2.1979068573348747e-08, + "loss": 0.7907, + "step": 10580 + }, + { + "epoch": 0.954231861838842, + "grad_norm": 2.066500396602952, + "learning_rate": 2.1892780036447013e-08, + "loss": 0.7119, + "step": 10581 + }, + { + "epoch": 0.9543220453623124, + "grad_norm": 1.7791106778819137, + "learning_rate": 2.1806660282560175e-08, + "loss": 0.8867, + "step": 10582 + }, + { + "epoch": 0.9544122288857826, + "grad_norm": 1.4764075626848279, + "learning_rate": 2.1720709319037024e-08, + "loss": 0.7679, + "step": 10583 + }, + { + "epoch": 0.9545024124092528, + "grad_norm": 2.431553850578963, + "learning_rate": 2.1634927153211023e-08, + "loss": 0.7113, + "step": 10584 + }, + { + "epoch": 0.954592595932723, + "grad_norm": 2.308624938650893, + "learning_rate": 2.1549313792401437e-08, + "loss": 0.8086, + "step": 10585 + }, + { + "epoch": 0.9546827794561934, + "grad_norm": 1.9586537957925236, + "learning_rate": 2.1463869243913746e-08, + "loss": 0.8374, + "step": 10586 + }, + { + "epoch": 0.9547729629796636, + "grad_norm": 2.3137566528554405, + "learning_rate": 2.1378593515037902e-08, + "loss": 0.7234, + "step": 10587 + }, + { + "epoch": 0.9548631465031339, + "grad_norm": 1.50381675553016, + "learning_rate": 2.129348661305075e-08, + "loss": 0.8026, + "step": 10588 + }, + { + "epoch": 0.9549533300266041, + "grad_norm": 2.431740044089583, + "learning_rate": 2.1208548545213813e-08, + "loss": 0.7803, + "step": 10589 + }, + { + "epoch": 0.9550435135500744, + "grad_norm": 1.9128373565330163, + "learning_rate": 2.1123779318774404e-08, + "loss": 0.702, + "step": 10590 + }, + { + "epoch": 0.9551336970735447, + "grad_norm": 6.906026845235209, + "learning_rate": 2.1039178940965408e-08, + "loss": 0.688, + "step": 10591 + }, + { + "epoch": 0.9552238805970149, + "grad_norm": 1.4249345329371235, + "learning_rate": 2.0954747419005712e-08, + "loss": 0.8095, + "step": 10592 + }, + { + "epoch": 0.9553140641204851, + "grad_norm": 2.4503762948802845, + "learning_rate": 2.087048476009934e-08, + "loss": 0.7249, + "step": 10593 + }, + { + "epoch": 0.9554042476439555, + "grad_norm": 1.5707065220786365, + "learning_rate": 2.0786390971435862e-08, + "loss": 0.72, + "step": 10594 + }, + { + "epoch": 0.9554944311674257, + "grad_norm": 2.669668435358319, + "learning_rate": 2.070246606019088e-08, + "loss": 0.7073, + "step": 10595 + }, + { + "epoch": 0.955584614690896, + "grad_norm": 2.438838004631651, + "learning_rate": 2.0618710033525112e-08, + "loss": 0.665, + "step": 10596 + }, + { + "epoch": 0.9556747982143662, + "grad_norm": 1.6710123706635946, + "learning_rate": 2.053512289858528e-08, + "loss": 0.7123, + "step": 10597 + }, + { + "epoch": 0.9557649817378365, + "grad_norm": 2.042731481016028, + "learning_rate": 2.0451704662503456e-08, + "loss": 0.7865, + "step": 10598 + }, + { + "epoch": 0.9558551652613068, + "grad_norm": 1.893237366236614, + "learning_rate": 2.0368455332397282e-08, + "loss": 0.8276, + "step": 10599 + }, + { + "epoch": 0.955945348784777, + "grad_norm": 1.6779712514221945, + "learning_rate": 2.0285374915369967e-08, + "loss": 0.7588, + "step": 10600 + }, + { + "epoch": 0.9560355323082473, + "grad_norm": 2.2615873993729365, + "learning_rate": 2.020246341851073e-08, + "loss": 0.7888, + "step": 10601 + }, + { + "epoch": 0.9561257158317176, + "grad_norm": 2.8924569588164935, + "learning_rate": 2.0119720848893463e-08, + "loss": 0.7863, + "step": 10602 + }, + { + "epoch": 0.9562158993551878, + "grad_norm": 1.7946692819283425, + "learning_rate": 2.0037147213578964e-08, + "loss": 0.6935, + "step": 10603 + }, + { + "epoch": 0.956306082878658, + "grad_norm": 1.6126222904171619, + "learning_rate": 1.9954742519612265e-08, + "loss": 0.7718, + "step": 10604 + }, + { + "epoch": 0.9563962664021284, + "grad_norm": 1.5888844438118694, + "learning_rate": 1.9872506774024633e-08, + "loss": 0.7209, + "step": 10605 + }, + { + "epoch": 0.9564864499255986, + "grad_norm": 2.010220045314621, + "learning_rate": 1.979043998383334e-08, + "loss": 0.7523, + "step": 10606 + }, + { + "epoch": 0.9565766334490688, + "grad_norm": 3.1462993318784545, + "learning_rate": 1.970854215604034e-08, + "loss": 0.6856, + "step": 10607 + }, + { + "epoch": 0.9566668169725391, + "grad_norm": 1.5979566221783135, + "learning_rate": 1.9626813297633826e-08, + "loss": 0.7158, + "step": 10608 + }, + { + "epoch": 0.9567570004960094, + "grad_norm": 2.2543661352479836, + "learning_rate": 1.954525341558688e-08, + "loss": 0.646, + "step": 10609 + }, + { + "epoch": 0.9568471840194797, + "grad_norm": 2.048227576724493, + "learning_rate": 1.9463862516859498e-08, + "loss": 0.7142, + "step": 10610 + }, + { + "epoch": 0.9569373675429499, + "grad_norm": 1.7010917776108092, + "learning_rate": 1.938264060839545e-08, + "loss": 0.8337, + "step": 10611 + }, + { + "epoch": 0.9570275510664201, + "grad_norm": 1.7336412544497921, + "learning_rate": 1.9301587697126086e-08, + "loss": 0.7192, + "step": 10612 + }, + { + "epoch": 0.9571177345898905, + "grad_norm": 1.9826144120248896, + "learning_rate": 1.9220703789966318e-08, + "loss": 0.6987, + "step": 10613 + }, + { + "epoch": 0.9572079181133607, + "grad_norm": 1.6090121554620278, + "learning_rate": 1.913998889381818e-08, + "loss": 0.7053, + "step": 10614 + }, + { + "epoch": 0.9572981016368309, + "grad_norm": 2.3261747448292542, + "learning_rate": 1.9059443015568387e-08, + "loss": 0.7614, + "step": 10615 + }, + { + "epoch": 0.9573882851603012, + "grad_norm": 1.9831911295178082, + "learning_rate": 1.8979066162089884e-08, + "loss": 0.7597, + "step": 10616 + }, + { + "epoch": 0.9574784686837715, + "grad_norm": 4.675847128752175, + "learning_rate": 1.889885834024052e-08, + "loss": 0.7492, + "step": 10617 + }, + { + "epoch": 0.9575686522072417, + "grad_norm": 2.2544331851956114, + "learning_rate": 1.8818819556864374e-08, + "loss": 0.6975, + "step": 10618 + }, + { + "epoch": 0.957658835730712, + "grad_norm": 2.2158499018449116, + "learning_rate": 1.873894981879065e-08, + "loss": 0.7315, + "step": 10619 + }, + { + "epoch": 0.9577490192541822, + "grad_norm": 0.6347621741567858, + "learning_rate": 1.8659249132834342e-08, + "loss": 0.5911, + "step": 10620 + }, + { + "epoch": 0.9578392027776526, + "grad_norm": 1.521912657724184, + "learning_rate": 1.857971750579579e-08, + "loss": 0.7237, + "step": 10621 + }, + { + "epoch": 0.9579293863011228, + "grad_norm": 1.4164812677277911, + "learning_rate": 1.8500354944461116e-08, + "loss": 0.6287, + "step": 10622 + }, + { + "epoch": 0.958019569824593, + "grad_norm": 1.4988144321924703, + "learning_rate": 1.8421161455602242e-08, + "loss": 0.755, + "step": 10623 + }, + { + "epoch": 0.9581097533480634, + "grad_norm": 1.5944925710051077, + "learning_rate": 1.834213704597598e-08, + "loss": 0.7695, + "step": 10624 + }, + { + "epoch": 0.9581999368715336, + "grad_norm": 1.5014048990828524, + "learning_rate": 1.8263281722325385e-08, + "loss": 0.7553, + "step": 10625 + }, + { + "epoch": 0.9582901203950038, + "grad_norm": 1.950328136614288, + "learning_rate": 1.818459549137885e-08, + "loss": 0.7563, + "step": 10626 + }, + { + "epoch": 0.9583803039184741, + "grad_norm": 1.6985086294543972, + "learning_rate": 1.8106078359850117e-08, + "loss": 0.7332, + "step": 10627 + }, + { + "epoch": 0.9584704874419444, + "grad_norm": 3.9597959660457764, + "learning_rate": 1.802773033443894e-08, + "loss": 0.7651, + "step": 10628 + }, + { + "epoch": 0.9585606709654146, + "grad_norm": 1.646049469224135, + "learning_rate": 1.7949551421830413e-08, + "loss": 0.7132, + "step": 10629 + }, + { + "epoch": 0.9586508544888849, + "grad_norm": 1.526466137002843, + "learning_rate": 1.7871541628694752e-08, + "loss": 0.7056, + "step": 10630 + }, + { + "epoch": 0.9587410380123551, + "grad_norm": 2.8257579698096706, + "learning_rate": 1.779370096168864e-08, + "loss": 0.6581, + "step": 10631 + }, + { + "epoch": 0.9588312215358255, + "grad_norm": 1.9124450700638052, + "learning_rate": 1.771602942745387e-08, + "loss": 0.7109, + "step": 10632 + }, + { + "epoch": 0.9589214050592957, + "grad_norm": 4.0607981711862235, + "learning_rate": 1.763852703261759e-08, + "loss": 0.7276, + "step": 10633 + }, + { + "epoch": 0.9590115885827659, + "grad_norm": 1.7406516382997281, + "learning_rate": 1.756119378379295e-08, + "loss": 0.8086, + "step": 10634 + }, + { + "epoch": 0.9591017721062362, + "grad_norm": 1.5786917395877442, + "learning_rate": 1.7484029687578005e-08, + "loss": 0.8344, + "step": 10635 + }, + { + "epoch": 0.9591919556297065, + "grad_norm": 1.495829187810709, + "learning_rate": 1.740703475055727e-08, + "loss": 0.7214, + "step": 10636 + }, + { + "epoch": 0.9592821391531767, + "grad_norm": 2.740965829184635, + "learning_rate": 1.7330208979300153e-08, + "loss": 0.7821, + "step": 10637 + }, + { + "epoch": 0.959372322676647, + "grad_norm": 2.501741660932082, + "learning_rate": 1.725355238036208e-08, + "loss": 0.7883, + "step": 10638 + }, + { + "epoch": 0.9594625062001172, + "grad_norm": 1.6341986362274645, + "learning_rate": 1.7177064960283594e-08, + "loss": 0.7862, + "step": 10639 + }, + { + "epoch": 0.9595526897235875, + "grad_norm": 11.387089938203799, + "learning_rate": 1.7100746725591253e-08, + "loss": 0.7415, + "step": 10640 + }, + { + "epoch": 0.9596428732470578, + "grad_norm": 2.1626133663031486, + "learning_rate": 1.7024597682796517e-08, + "loss": 0.7405, + "step": 10641 + }, + { + "epoch": 0.959733056770528, + "grad_norm": 1.618629365645727, + "learning_rate": 1.6948617838397293e-08, + "loss": 0.8166, + "step": 10642 + }, + { + "epoch": 0.9598232402939982, + "grad_norm": 1.446565404605978, + "learning_rate": 1.6872807198876404e-08, + "loss": 0.7747, + "step": 10643 + }, + { + "epoch": 0.9599134238174686, + "grad_norm": 1.8317786329282504, + "learning_rate": 1.679716577070245e-08, + "loss": 0.8462, + "step": 10644 + }, + { + "epoch": 0.9600036073409388, + "grad_norm": 1.7152434906340486, + "learning_rate": 1.6721693560329596e-08, + "loss": 0.8339, + "step": 10645 + }, + { + "epoch": 0.960093790864409, + "grad_norm": 1.8512463521052138, + "learning_rate": 1.6646390574197366e-08, + "loss": 0.6552, + "step": 10646 + }, + { + "epoch": 0.9601839743878793, + "grad_norm": 0.6906491460937533, + "learning_rate": 1.6571256818731504e-08, + "loss": 0.5966, + "step": 10647 + }, + { + "epoch": 0.9602741579113496, + "grad_norm": 1.8240021734308378, + "learning_rate": 1.6496292300342218e-08, + "loss": 0.6761, + "step": 10648 + }, + { + "epoch": 0.9603643414348199, + "grad_norm": 2.244657744080715, + "learning_rate": 1.642149702542639e-08, + "loss": 0.6867, + "step": 10649 + }, + { + "epoch": 0.9604545249582901, + "grad_norm": 1.614630257529653, + "learning_rate": 1.634687100036558e-08, + "loss": 0.7088, + "step": 10650 + }, + { + "epoch": 0.9605447084817604, + "grad_norm": 2.5580740131410904, + "learning_rate": 1.627241423152781e-08, + "loss": 0.6778, + "step": 10651 + }, + { + "epoch": 0.9606348920052307, + "grad_norm": 2.3130631663219368, + "learning_rate": 1.619812672526555e-08, + "loss": 0.7943, + "step": 10652 + }, + { + "epoch": 0.9607250755287009, + "grad_norm": 1.8690589878984933, + "learning_rate": 1.6124008487917727e-08, + "loss": 0.7296, + "step": 10653 + }, + { + "epoch": 0.9608152590521711, + "grad_norm": 0.6470098520778774, + "learning_rate": 1.6050059525808623e-08, + "loss": 0.5946, + "step": 10654 + }, + { + "epoch": 0.9609054425756415, + "grad_norm": 0.6367551641056378, + "learning_rate": 1.597627984524763e-08, + "loss": 0.5904, + "step": 10655 + }, + { + "epoch": 0.9609956260991117, + "grad_norm": 1.6083547821327702, + "learning_rate": 1.590266945253038e-08, + "loss": 0.7872, + "step": 10656 + }, + { + "epoch": 0.9610858096225819, + "grad_norm": 1.3904396105001082, + "learning_rate": 1.582922835393763e-08, + "loss": 0.7598, + "step": 10657 + }, + { + "epoch": 0.9611759931460522, + "grad_norm": 1.529464692984304, + "learning_rate": 1.5755956555735473e-08, + "loss": 0.7197, + "step": 10658 + }, + { + "epoch": 0.9612661766695225, + "grad_norm": 0.717739057690671, + "learning_rate": 1.5682854064176244e-08, + "loss": 0.6649, + "step": 10659 + }, + { + "epoch": 0.9613563601929928, + "grad_norm": 1.9214733913702517, + "learning_rate": 1.5609920885497395e-08, + "loss": 0.6793, + "step": 10660 + }, + { + "epoch": 0.961446543716463, + "grad_norm": 1.6805075129778286, + "learning_rate": 1.5537157025921732e-08, + "loss": 0.7651, + "step": 10661 + }, + { + "epoch": 0.9615367272399332, + "grad_norm": 1.8461781692736583, + "learning_rate": 1.5464562491658285e-08, + "loss": 0.7253, + "step": 10662 + }, + { + "epoch": 0.9616269107634036, + "grad_norm": 2.034154839622603, + "learning_rate": 1.5392137288900764e-08, + "loss": 0.7784, + "step": 10663 + }, + { + "epoch": 0.9617170942868738, + "grad_norm": 0.598452144075107, + "learning_rate": 1.531988142382934e-08, + "loss": 0.5797, + "step": 10664 + }, + { + "epoch": 0.961807277810344, + "grad_norm": 4.464694016449779, + "learning_rate": 1.5247794902608634e-08, + "loss": 0.6745, + "step": 10665 + }, + { + "epoch": 0.9618974613338143, + "grad_norm": 1.8527781041208742, + "learning_rate": 1.5175877731390398e-08, + "loss": 0.7043, + "step": 10666 + }, + { + "epoch": 0.9619876448572846, + "grad_norm": 1.7744990461255024, + "learning_rate": 1.510412991631016e-08, + "loss": 0.7745, + "step": 10667 + }, + { + "epoch": 0.9620778283807548, + "grad_norm": 1.553950916413565, + "learning_rate": 1.503255146349014e-08, + "loss": 0.7725, + "step": 10668 + }, + { + "epoch": 0.9621680119042251, + "grad_norm": 1.8108153571079084, + "learning_rate": 1.4961142379037893e-08, + "loss": 0.7125, + "step": 10669 + }, + { + "epoch": 0.9622581954276953, + "grad_norm": 2.428221428918979, + "learning_rate": 1.4889902669046327e-08, + "loss": 0.7836, + "step": 10670 + }, + { + "epoch": 0.9623483789511657, + "grad_norm": 1.9210321702740705, + "learning_rate": 1.4818832339594135e-08, + "loss": 0.746, + "step": 10671 + }, + { + "epoch": 0.9624385624746359, + "grad_norm": 1.4607099060008923, + "learning_rate": 1.474793139674535e-08, + "loss": 0.6977, + "step": 10672 + }, + { + "epoch": 0.9625287459981061, + "grad_norm": 1.6284532534595528, + "learning_rate": 1.4677199846549581e-08, + "loss": 0.744, + "step": 10673 + }, + { + "epoch": 0.9626189295215765, + "grad_norm": 1.9763577924426485, + "learning_rate": 1.4606637695042224e-08, + "loss": 0.7461, + "step": 10674 + }, + { + "epoch": 0.9627091130450467, + "grad_norm": 4.008762299126556, + "learning_rate": 1.4536244948243793e-08, + "loss": 0.8401, + "step": 10675 + }, + { + "epoch": 0.9627992965685169, + "grad_norm": 1.7764481184749126, + "learning_rate": 1.4466021612160595e-08, + "loss": 0.7112, + "step": 10676 + }, + { + "epoch": 0.9628894800919872, + "grad_norm": 1.7688511768963748, + "learning_rate": 1.4395967692784505e-08, + "loss": 0.7103, + "step": 10677 + }, + { + "epoch": 0.9629796636154575, + "grad_norm": 1.7159168821981294, + "learning_rate": 1.4326083196092963e-08, + "loss": 0.6724, + "step": 10678 + }, + { + "epoch": 0.9630698471389277, + "grad_norm": 1.4316536146878303, + "learning_rate": 1.42563681280492e-08, + "loss": 0.7533, + "step": 10679 + }, + { + "epoch": 0.963160030662398, + "grad_norm": 1.9399310211817737, + "learning_rate": 1.4186822494600902e-08, + "loss": 0.72, + "step": 10680 + }, + { + "epoch": 0.9632502141858682, + "grad_norm": 2.560176633463256, + "learning_rate": 1.4117446301682877e-08, + "loss": 0.7326, + "step": 10681 + }, + { + "epoch": 0.9633403977093385, + "grad_norm": 1.645088300798938, + "learning_rate": 1.4048239555214392e-08, + "loss": 0.7006, + "step": 10682 + }, + { + "epoch": 0.9634305812328088, + "grad_norm": 1.5294399072789415, + "learning_rate": 1.3979202261100497e-08, + "loss": 0.7776, + "step": 10683 + }, + { + "epoch": 0.963520764756279, + "grad_norm": 2.870730173354934, + "learning_rate": 1.3910334425231817e-08, + "loss": 0.8088, + "step": 10684 + }, + { + "epoch": 0.9636109482797492, + "grad_norm": 1.4973428556496713, + "learning_rate": 1.384163605348454e-08, + "loss": 0.739, + "step": 10685 + }, + { + "epoch": 0.9637011318032196, + "grad_norm": 2.041664889068868, + "learning_rate": 1.3773107151720642e-08, + "loss": 0.8022, + "step": 10686 + }, + { + "epoch": 0.9637913153266898, + "grad_norm": 1.9095914352792256, + "learning_rate": 1.3704747725787003e-08, + "loss": 0.8065, + "step": 10687 + }, + { + "epoch": 0.9638814988501601, + "grad_norm": 0.7331864092409571, + "learning_rate": 1.3636557781516512e-08, + "loss": 0.5873, + "step": 10688 + }, + { + "epoch": 0.9639716823736303, + "grad_norm": 2.1066932868281625, + "learning_rate": 1.3568537324727847e-08, + "loss": 0.747, + "step": 10689 + }, + { + "epoch": 0.9640618658971006, + "grad_norm": 1.5228754033640253, + "learning_rate": 1.3500686361224589e-08, + "loss": 0.7426, + "step": 10690 + }, + { + "epoch": 0.9641520494205709, + "grad_norm": 1.9533623513007323, + "learning_rate": 1.3433004896796108e-08, + "loss": 0.7416, + "step": 10691 + }, + { + "epoch": 0.9642422329440411, + "grad_norm": 3.6299562444230347, + "learning_rate": 1.336549293721756e-08, + "loss": 0.6326, + "step": 10692 + }, + { + "epoch": 0.9643324164675113, + "grad_norm": 0.6788187723462701, + "learning_rate": 1.3298150488249227e-08, + "loss": 0.5731, + "step": 10693 + }, + { + "epoch": 0.9644225999909817, + "grad_norm": 2.179236143133989, + "learning_rate": 1.3230977555637401e-08, + "loss": 0.817, + "step": 10694 + }, + { + "epoch": 0.9645127835144519, + "grad_norm": 1.4571262359339012, + "learning_rate": 1.3163974145113499e-08, + "loss": 0.7001, + "step": 10695 + }, + { + "epoch": 0.9646029670379221, + "grad_norm": 12.855658660478138, + "learning_rate": 1.3097140262394723e-08, + "loss": 0.643, + "step": 10696 + }, + { + "epoch": 0.9646931505613925, + "grad_norm": 2.5832543412303783, + "learning_rate": 1.303047591318318e-08, + "loss": 0.7784, + "step": 10697 + }, + { + "epoch": 0.9647833340848627, + "grad_norm": 1.6439125183757528, + "learning_rate": 1.2963981103167875e-08, + "loss": 0.6747, + "step": 10698 + }, + { + "epoch": 0.964873517608333, + "grad_norm": 1.5277248989934, + "learning_rate": 1.2897655838021825e-08, + "loss": 0.7268, + "step": 10699 + }, + { + "epoch": 0.9649637011318032, + "grad_norm": 1.9845379739954845, + "learning_rate": 1.2831500123404726e-08, + "loss": 0.7518, + "step": 10700 + }, + { + "epoch": 0.9650538846552735, + "grad_norm": 2.3822939449199887, + "learning_rate": 1.2765513964961172e-08, + "loss": 0.8944, + "step": 10701 + }, + { + "epoch": 0.9651440681787438, + "grad_norm": 1.7544302720406173, + "learning_rate": 1.2699697368321549e-08, + "loss": 0.7409, + "step": 10702 + }, + { + "epoch": 0.965234251702214, + "grad_norm": 1.615644668075702, + "learning_rate": 1.2634050339101366e-08, + "loss": 0.7273, + "step": 10703 + }, + { + "epoch": 0.9653244352256842, + "grad_norm": 2.761973842726258, + "learning_rate": 1.2568572882902361e-08, + "loss": 0.721, + "step": 10704 + }, + { + "epoch": 0.9654146187491546, + "grad_norm": 1.8485689066517836, + "learning_rate": 1.2503265005311402e-08, + "loss": 0.7168, + "step": 10705 + }, + { + "epoch": 0.9655048022726248, + "grad_norm": 1.67022416797447, + "learning_rate": 1.2438126711900698e-08, + "loss": 0.7084, + "step": 10706 + }, + { + "epoch": 0.965594985796095, + "grad_norm": 1.5215717334666936, + "learning_rate": 1.2373158008228247e-08, + "loss": 0.5909, + "step": 10707 + }, + { + "epoch": 0.9656851693195653, + "grad_norm": 1.3852357913618831, + "learning_rate": 1.2308358899837833e-08, + "loss": 0.7212, + "step": 10708 + }, + { + "epoch": 0.9657753528430356, + "grad_norm": 1.7907066286664466, + "learning_rate": 1.224372939225815e-08, + "loss": 0.7195, + "step": 10709 + }, + { + "epoch": 0.9658655363665059, + "grad_norm": 1.7178305507445044, + "learning_rate": 1.2179269491003674e-08, + "loss": 0.823, + "step": 10710 + }, + { + "epoch": 0.9659557198899761, + "grad_norm": 1.9530364359803234, + "learning_rate": 1.2114979201574894e-08, + "loss": 0.6143, + "step": 10711 + }, + { + "epoch": 0.9660459034134463, + "grad_norm": 1.6054079364480165, + "learning_rate": 1.2050858529456975e-08, + "loss": 0.7611, + "step": 10712 + }, + { + "epoch": 0.9661360869369167, + "grad_norm": 1.6747598447440573, + "learning_rate": 1.1986907480121545e-08, + "loss": 0.7456, + "step": 10713 + }, + { + "epoch": 0.9662262704603869, + "grad_norm": 1.781153559165384, + "learning_rate": 1.192312605902468e-08, + "loss": 0.7664, + "step": 10714 + }, + { + "epoch": 0.9663164539838571, + "grad_norm": 0.7918972800476787, + "learning_rate": 1.1859514271608917e-08, + "loss": 0.6028, + "step": 10715 + }, + { + "epoch": 0.9664066375073274, + "grad_norm": 2.5515846751383022, + "learning_rate": 1.1796072123301914e-08, + "loss": 0.7772, + "step": 10716 + }, + { + "epoch": 0.9664968210307977, + "grad_norm": 0.6404533388843415, + "learning_rate": 1.1732799619516897e-08, + "loss": 0.5821, + "step": 10717 + }, + { + "epoch": 0.9665870045542679, + "grad_norm": 2.9414513167820786, + "learning_rate": 1.1669696765652659e-08, + "loss": 0.7299, + "step": 10718 + }, + { + "epoch": 0.9666771880777382, + "grad_norm": 2.0485038973524707, + "learning_rate": 1.1606763567093336e-08, + "loss": 0.7723, + "step": 10719 + }, + { + "epoch": 0.9667673716012085, + "grad_norm": 0.6738306902408145, + "learning_rate": 1.1544000029208857e-08, + "loss": 0.5729, + "step": 10720 + }, + { + "epoch": 0.9668575551246787, + "grad_norm": 1.7577705208746066, + "learning_rate": 1.148140615735449e-08, + "loss": 0.6021, + "step": 10721 + }, + { + "epoch": 0.966947738648149, + "grad_norm": 1.4784462835663368, + "learning_rate": 1.1418981956871076e-08, + "loss": 0.7209, + "step": 10722 + }, + { + "epoch": 0.9670379221716192, + "grad_norm": 2.135013101792169, + "learning_rate": 1.1356727433085245e-08, + "loss": 0.7328, + "step": 10723 + }, + { + "epoch": 0.9671281056950896, + "grad_norm": 1.482641553119506, + "learning_rate": 1.1294642591308524e-08, + "loss": 0.7666, + "step": 10724 + }, + { + "epoch": 0.9672182892185598, + "grad_norm": 2.3651001039251836, + "learning_rate": 1.1232727436838452e-08, + "loss": 0.7333, + "step": 10725 + }, + { + "epoch": 0.96730847274203, + "grad_norm": 2.2081374915272005, + "learning_rate": 1.1170981974958138e-08, + "loss": 0.7884, + "step": 10726 + }, + { + "epoch": 0.9673986562655003, + "grad_norm": 1.5557747877507224, + "learning_rate": 1.1109406210936035e-08, + "loss": 0.8096, + "step": 10727 + }, + { + "epoch": 0.9674888397889706, + "grad_norm": 2.383465086578616, + "learning_rate": 1.1048000150025939e-08, + "loss": 0.8601, + "step": 10728 + }, + { + "epoch": 0.9675790233124408, + "grad_norm": 3.3507228413082957, + "learning_rate": 1.0986763797467213e-08, + "loss": 0.6412, + "step": 10729 + }, + { + "epoch": 0.9676692068359111, + "grad_norm": 0.5876172719746302, + "learning_rate": 1.0925697158485459e-08, + "loss": 0.5446, + "step": 10730 + }, + { + "epoch": 0.9677593903593813, + "grad_norm": 1.679767127700147, + "learning_rate": 1.0864800238290727e-08, + "loss": 0.7506, + "step": 10731 + }, + { + "epoch": 0.9678495738828516, + "grad_norm": 1.7244335184264588, + "learning_rate": 1.0804073042079309e-08, + "loss": 0.7021, + "step": 10732 + }, + { + "epoch": 0.9679397574063219, + "grad_norm": 0.6419482172852338, + "learning_rate": 1.0743515575032392e-08, + "loss": 0.5888, + "step": 10733 + }, + { + "epoch": 0.9680299409297921, + "grad_norm": 2.155078182528151, + "learning_rate": 1.0683127842317619e-08, + "loss": 0.7554, + "step": 10734 + }, + { + "epoch": 0.9681201244532623, + "grad_norm": 3.7654410036333408, + "learning_rate": 1.0622909849087314e-08, + "loss": 0.7727, + "step": 10735 + }, + { + "epoch": 0.9682103079767327, + "grad_norm": 1.5890161108036875, + "learning_rate": 1.0562861600479588e-08, + "loss": 0.8039, + "step": 10736 + }, + { + "epoch": 0.9683004915002029, + "grad_norm": 1.9585801503363982, + "learning_rate": 1.0502983101618345e-08, + "loss": 0.836, + "step": 10737 + }, + { + "epoch": 0.9683906750236732, + "grad_norm": 1.921360039919608, + "learning_rate": 1.0443274357612386e-08, + "loss": 0.624, + "step": 10738 + }, + { + "epoch": 0.9684808585471434, + "grad_norm": 1.7377243006988359, + "learning_rate": 1.0383735373556524e-08, + "loss": 0.7576, + "step": 10739 + }, + { + "epoch": 0.9685710420706137, + "grad_norm": 1.4371717326972153, + "learning_rate": 1.0324366154531139e-08, + "loss": 0.7735, + "step": 10740 + }, + { + "epoch": 0.968661225594084, + "grad_norm": 1.9124420780561056, + "learning_rate": 1.0265166705601735e-08, + "loss": 0.7096, + "step": 10741 + }, + { + "epoch": 0.9687514091175542, + "grad_norm": 0.704815929407543, + "learning_rate": 1.0206137031819606e-08, + "loss": 0.611, + "step": 10742 + }, + { + "epoch": 0.9688415926410244, + "grad_norm": 4.683187510074372, + "learning_rate": 1.0147277138221388e-08, + "loss": 0.7568, + "step": 10743 + }, + { + "epoch": 0.9689317761644948, + "grad_norm": 4.331253628639804, + "learning_rate": 1.0088587029829287e-08, + "loss": 0.7783, + "step": 10744 + }, + { + "epoch": 0.969021959687965, + "grad_norm": 1.5501619346690523, + "learning_rate": 1.003006671165152e-08, + "loss": 0.8472, + "step": 10745 + }, + { + "epoch": 0.9691121432114352, + "grad_norm": 1.8738210150333503, + "learning_rate": 9.971716188680978e-09, + "loss": 0.7377, + "step": 10746 + }, + { + "epoch": 0.9692023267349056, + "grad_norm": 2.0451516620760266, + "learning_rate": 9.91353546589635e-09, + "loss": 0.7097, + "step": 10747 + }, + { + "epoch": 0.9692925102583758, + "grad_norm": 1.8587483223533627, + "learning_rate": 9.855524548262106e-09, + "loss": 0.6804, + "step": 10748 + }, + { + "epoch": 0.969382693781846, + "grad_norm": 1.796376830994295, + "learning_rate": 9.797683440728288e-09, + "loss": 0.6909, + "step": 10749 + }, + { + "epoch": 0.9694728773053163, + "grad_norm": 1.6614969741329957, + "learning_rate": 9.740012148229836e-09, + "loss": 0.6564, + "step": 10750 + }, + { + "epoch": 0.9695630608287866, + "grad_norm": 1.519165858095714, + "learning_rate": 9.682510675687705e-09, + "loss": 0.6945, + "step": 10751 + }, + { + "epoch": 0.9696532443522569, + "grad_norm": 1.9564041076994525, + "learning_rate": 9.625179028008191e-09, + "loss": 0.6648, + "step": 10752 + }, + { + "epoch": 0.9697434278757271, + "grad_norm": 1.7239090219995925, + "learning_rate": 9.568017210083379e-09, + "loss": 0.5888, + "step": 10753 + }, + { + "epoch": 0.9698336113991973, + "grad_norm": 2.2980120692388524, + "learning_rate": 9.511025226790259e-09, + "loss": 0.722, + "step": 10754 + }, + { + "epoch": 0.9699237949226677, + "grad_norm": 1.643291595127542, + "learning_rate": 9.454203082992052e-09, + "loss": 0.7175, + "step": 10755 + }, + { + "epoch": 0.9700139784461379, + "grad_norm": 2.3026787748734483, + "learning_rate": 9.3975507835371e-09, + "loss": 0.6551, + "step": 10756 + }, + { + "epoch": 0.9701041619696081, + "grad_norm": 0.6311331470845365, + "learning_rate": 9.341068333259094e-09, + "loss": 0.5651, + "step": 10757 + }, + { + "epoch": 0.9701943454930784, + "grad_norm": 1.9300755407937187, + "learning_rate": 9.28475573697729e-09, + "loss": 0.7197, + "step": 10758 + }, + { + "epoch": 0.9702845290165487, + "grad_norm": 1.6047035491348796, + "learning_rate": 9.228612999497177e-09, + "loss": 0.7451, + "step": 10759 + }, + { + "epoch": 0.970374712540019, + "grad_norm": 2.8831709145314126, + "learning_rate": 9.172640125608478e-09, + "loss": 0.6822, + "step": 10760 + }, + { + "epoch": 0.9704648960634892, + "grad_norm": 1.5711863349634967, + "learning_rate": 9.116837120087817e-09, + "loss": 0.757, + "step": 10761 + }, + { + "epoch": 0.9705550795869594, + "grad_norm": 1.9869835596465166, + "learning_rate": 9.061203987695832e-09, + "loss": 0.8113, + "step": 10762 + }, + { + "epoch": 0.9706452631104298, + "grad_norm": 2.0541034056831253, + "learning_rate": 9.005740733180055e-09, + "loss": 0.69, + "step": 10763 + }, + { + "epoch": 0.9707354466339, + "grad_norm": 1.8491966487980405, + "learning_rate": 8.950447361272483e-09, + "loss": 0.8259, + "step": 10764 + }, + { + "epoch": 0.9708256301573702, + "grad_norm": 4.153950219874959, + "learning_rate": 8.895323876691784e-09, + "loss": 0.7487, + "step": 10765 + }, + { + "epoch": 0.9709158136808405, + "grad_norm": 1.9915024121705889, + "learning_rate": 8.840370284140419e-09, + "loss": 0.7972, + "step": 10766 + }, + { + "epoch": 0.9710059972043108, + "grad_norm": 2.0604160068409336, + "learning_rate": 8.78558658830797e-09, + "loss": 0.6615, + "step": 10767 + }, + { + "epoch": 0.971096180727781, + "grad_norm": 0.7035399590011113, + "learning_rate": 8.730972793868696e-09, + "loss": 0.5771, + "step": 10768 + }, + { + "epoch": 0.9711863642512513, + "grad_norm": 2.612171159700685, + "learning_rate": 8.67652890548265e-09, + "loss": 0.7062, + "step": 10769 + }, + { + "epoch": 0.9712765477747216, + "grad_norm": 1.5133861247418106, + "learning_rate": 8.622254927795004e-09, + "loss": 0.7244, + "step": 10770 + }, + { + "epoch": 0.9713667312981918, + "grad_norm": 2.1349578240403035, + "learning_rate": 8.568150865436941e-09, + "loss": 0.6935, + "step": 10771 + }, + { + "epoch": 0.9714569148216621, + "grad_norm": 1.5044145790158443, + "learning_rate": 8.514216723024991e-09, + "loss": 0.754, + "step": 10772 + }, + { + "epoch": 0.9715470983451323, + "grad_norm": 2.392743636654017, + "learning_rate": 8.460452505161031e-09, + "loss": 0.6631, + "step": 10773 + }, + { + "epoch": 0.9716372818686027, + "grad_norm": 2.6536261670423023, + "learning_rate": 8.4068582164325e-09, + "loss": 0.7657, + "step": 10774 + }, + { + "epoch": 0.9717274653920729, + "grad_norm": 1.6307471687868937, + "learning_rate": 8.353433861412406e-09, + "loss": 0.7236, + "step": 10775 + }, + { + "epoch": 0.9718176489155431, + "grad_norm": 1.530870118021958, + "learning_rate": 8.300179444658883e-09, + "loss": 0.7016, + "step": 10776 + }, + { + "epoch": 0.9719078324390134, + "grad_norm": 4.226432304703947, + "learning_rate": 8.247094970716296e-09, + "loss": 0.8376, + "step": 10777 + }, + { + "epoch": 0.9719980159624837, + "grad_norm": 1.5191077890179983, + "learning_rate": 8.19418044411413e-09, + "loss": 0.7777, + "step": 10778 + }, + { + "epoch": 0.9720881994859539, + "grad_norm": 1.380970951736089, + "learning_rate": 8.141435869367219e-09, + "loss": 0.7438, + "step": 10779 + }, + { + "epoch": 0.9721783830094242, + "grad_norm": 1.4741653279064275, + "learning_rate": 8.088861250975742e-09, + "loss": 0.7782, + "step": 10780 + }, + { + "epoch": 0.9722685665328944, + "grad_norm": 23.179376788446692, + "learning_rate": 8.036456593426111e-09, + "loss": 0.708, + "step": 10781 + }, + { + "epoch": 0.9723587500563647, + "grad_norm": 1.8353901148537886, + "learning_rate": 7.984221901189415e-09, + "loss": 0.7773, + "step": 10782 + }, + { + "epoch": 0.972448933579835, + "grad_norm": 9.556805679202599, + "learning_rate": 7.932157178722976e-09, + "loss": 0.6645, + "step": 10783 + }, + { + "epoch": 0.9725391171033052, + "grad_norm": 2.3147344361937763, + "learning_rate": 7.880262430468799e-09, + "loss": 0.7102, + "step": 10784 + }, + { + "epoch": 0.9726293006267754, + "grad_norm": 1.788115271468833, + "learning_rate": 7.828537660855339e-09, + "loss": 0.6754, + "step": 10785 + }, + { + "epoch": 0.9727194841502458, + "grad_norm": 1.9429803066423812, + "learning_rate": 7.776982874295512e-09, + "loss": 0.7573, + "step": 10786 + }, + { + "epoch": 0.972809667673716, + "grad_norm": 2.20410231969277, + "learning_rate": 7.725598075188688e-09, + "loss": 0.6501, + "step": 10787 + }, + { + "epoch": 0.9728998511971862, + "grad_norm": 1.4370804879431438, + "learning_rate": 7.674383267918916e-09, + "loss": 0.6924, + "step": 10788 + }, + { + "epoch": 0.9729900347206565, + "grad_norm": 1.7302217673236235, + "learning_rate": 7.623338456856476e-09, + "loss": 0.7301, + "step": 10789 + }, + { + "epoch": 0.9730802182441268, + "grad_norm": 2.1345691642859363, + "learning_rate": 7.572463646356554e-09, + "loss": 0.6343, + "step": 10790 + }, + { + "epoch": 0.9731704017675971, + "grad_norm": 1.8366180457218022, + "learning_rate": 7.521758840760339e-09, + "loss": 0.7127, + "step": 10791 + }, + { + "epoch": 0.9732605852910673, + "grad_norm": 2.271026171507846, + "learning_rate": 7.471224044393931e-09, + "loss": 0.7011, + "step": 10792 + }, + { + "epoch": 0.9733507688145376, + "grad_norm": 1.6238343386224934, + "learning_rate": 7.420859261569434e-09, + "loss": 0.7983, + "step": 10793 + }, + { + "epoch": 0.9734409523380079, + "grad_norm": 2.0063567468878687, + "learning_rate": 7.370664496584078e-09, + "loss": 0.8009, + "step": 10794 + }, + { + "epoch": 0.9735311358614781, + "grad_norm": 1.95546874756152, + "learning_rate": 7.3206397537211026e-09, + "loss": 0.6645, + "step": 10795 + }, + { + "epoch": 0.9736213193849483, + "grad_norm": 10.058003244195833, + "learning_rate": 7.270785037248428e-09, + "loss": 0.7126, + "step": 10796 + }, + { + "epoch": 0.9737115029084187, + "grad_norm": 3.3321450181614405, + "learning_rate": 7.221100351420428e-09, + "loss": 0.7041, + "step": 10797 + }, + { + "epoch": 0.9738016864318889, + "grad_norm": 1.6111982813451895, + "learning_rate": 7.171585700475935e-09, + "loss": 0.668, + "step": 10798 + }, + { + "epoch": 0.9738918699553591, + "grad_norm": 1.7437305545064468, + "learning_rate": 7.122241088640235e-09, + "loss": 0.8021, + "step": 10799 + }, + { + "epoch": 0.9739820534788294, + "grad_norm": 1.7736444562570324, + "learning_rate": 7.073066520123516e-09, + "loss": 0.8626, + "step": 10800 + }, + { + "epoch": 0.9740722370022997, + "grad_norm": 2.7758553295090453, + "learning_rate": 7.0240619991217555e-09, + "loss": 0.6355, + "step": 10801 + }, + { + "epoch": 0.97416242052577, + "grad_norm": 1.5957507122886248, + "learning_rate": 6.975227529816052e-09, + "loss": 0.8114, + "step": 10802 + }, + { + "epoch": 0.9742526040492402, + "grad_norm": 1.8109734124349437, + "learning_rate": 6.926563116373296e-09, + "loss": 0.7858, + "step": 10803 + }, + { + "epoch": 0.9743427875727104, + "grad_norm": 1.3717494168510609, + "learning_rate": 6.878068762945943e-09, + "loss": 0.6758, + "step": 10804 + }, + { + "epoch": 0.9744329710961808, + "grad_norm": 1.4784917590131257, + "learning_rate": 6.829744473671794e-09, + "loss": 0.7783, + "step": 10805 + }, + { + "epoch": 0.974523154619651, + "grad_norm": 0.5872639455767469, + "learning_rate": 6.781590252674219e-09, + "loss": 0.4789, + "step": 10806 + }, + { + "epoch": 0.9746133381431212, + "grad_norm": 5.263185926663888, + "learning_rate": 6.733606104061484e-09, + "loss": 0.7488, + "step": 10807 + }, + { + "epoch": 0.9747035216665915, + "grad_norm": 1.4017653268541188, + "learning_rate": 6.6857920319283165e-09, + "loss": 0.6851, + "step": 10808 + }, + { + "epoch": 0.9747937051900618, + "grad_norm": 2.09713676171848, + "learning_rate": 6.638148040354563e-09, + "loss": 0.7154, + "step": 10809 + }, + { + "epoch": 0.974883888713532, + "grad_norm": 1.75487425329194, + "learning_rate": 6.590674133405194e-09, + "loss": 0.7014, + "step": 10810 + }, + { + "epoch": 0.9749740722370023, + "grad_norm": 1.9228979350204367, + "learning_rate": 6.5433703151311914e-09, + "loss": 0.7505, + "step": 10811 + }, + { + "epoch": 0.9750642557604725, + "grad_norm": 1.7165583422041186, + "learning_rate": 6.49623658956866e-09, + "loss": 0.7411, + "step": 10812 + }, + { + "epoch": 0.9751544392839429, + "grad_norm": 2.46219208239659, + "learning_rate": 6.44927296073905e-09, + "loss": 0.6164, + "step": 10813 + }, + { + "epoch": 0.9752446228074131, + "grad_norm": 1.5607184362340722, + "learning_rate": 6.402479432649821e-09, + "loss": 0.8231, + "step": 10814 + }, + { + "epoch": 0.9753348063308833, + "grad_norm": 1.7195223373466346, + "learning_rate": 6.355856009293781e-09, + "loss": 0.7323, + "step": 10815 + }, + { + "epoch": 0.9754249898543537, + "grad_norm": 1.932512015420096, + "learning_rate": 6.3094026946488575e-09, + "loss": 0.7168, + "step": 10816 + }, + { + "epoch": 0.9755151733778239, + "grad_norm": 1.5082856063598078, + "learning_rate": 6.2631194926787704e-09, + "loss": 0.7419, + "step": 10817 + }, + { + "epoch": 0.9756053569012941, + "grad_norm": 1.619209021215413, + "learning_rate": 6.217006407332581e-09, + "loss": 0.6363, + "step": 10818 + }, + { + "epoch": 0.9756955404247644, + "grad_norm": 1.4322897986197571, + "learning_rate": 6.1710634425453654e-09, + "loss": 0.6858, + "step": 10819 + }, + { + "epoch": 0.9757857239482347, + "grad_norm": 1.7728754572966938, + "learning_rate": 6.1252906022366544e-09, + "loss": 0.7622, + "step": 10820 + }, + { + "epoch": 0.9758759074717049, + "grad_norm": 11.09447483327498, + "learning_rate": 6.079687890312213e-09, + "loss": 0.7173, + "step": 10821 + }, + { + "epoch": 0.9759660909951752, + "grad_norm": 1.958919870183641, + "learning_rate": 6.034255310663372e-09, + "loss": 0.7202, + "step": 10822 + }, + { + "epoch": 0.9760562745186454, + "grad_norm": 1.8062385400322751, + "learning_rate": 5.988992867166143e-09, + "loss": 0.8371, + "step": 10823 + }, + { + "epoch": 0.9761464580421157, + "grad_norm": 1.9324874641633667, + "learning_rate": 5.943900563682991e-09, + "loss": 0.7777, + "step": 10824 + }, + { + "epoch": 0.976236641565586, + "grad_norm": 0.7148394975379406, + "learning_rate": 5.898978404061506e-09, + "loss": 0.6563, + "step": 10825 + }, + { + "epoch": 0.9763268250890562, + "grad_norm": 2.2830329094084245, + "learning_rate": 5.85422639213462e-09, + "loss": 0.7316, + "step": 10826 + }, + { + "epoch": 0.9764170086125264, + "grad_norm": 1.4901769384512182, + "learning_rate": 5.809644531720614e-09, + "loss": 0.7511, + "step": 10827 + }, + { + "epoch": 0.9765071921359968, + "grad_norm": 1.7502234180068135, + "learning_rate": 5.765232826623556e-09, + "loss": 0.7682, + "step": 10828 + }, + { + "epoch": 0.976597375659467, + "grad_norm": 1.9641462722429481, + "learning_rate": 5.720991280633081e-09, + "loss": 0.7192, + "step": 10829 + }, + { + "epoch": 0.9766875591829373, + "grad_norm": 1.4464543136781844, + "learning_rate": 5.676919897523724e-09, + "loss": 0.7233, + "step": 10830 + }, + { + "epoch": 0.9767777427064075, + "grad_norm": 1.4986485114904133, + "learning_rate": 5.633018681056256e-09, + "loss": 0.7434, + "step": 10831 + }, + { + "epoch": 0.9768679262298778, + "grad_norm": 3.029304745416114, + "learning_rate": 5.589287634976569e-09, + "loss": 0.7406, + "step": 10832 + }, + { + "epoch": 0.9769581097533481, + "grad_norm": 1.6372246518027331, + "learning_rate": 5.5457267630159014e-09, + "loss": 0.7106, + "step": 10833 + }, + { + "epoch": 0.9770482932768183, + "grad_norm": 2.7176701823888347, + "learning_rate": 5.5023360688910555e-09, + "loss": 0.7772, + "step": 10834 + }, + { + "epoch": 0.9771384768002885, + "grad_norm": 1.6235441875923615, + "learning_rate": 5.459115556304183e-09, + "loss": 0.7883, + "step": 10835 + }, + { + "epoch": 0.9772286603237589, + "grad_norm": 1.9686595351006793, + "learning_rate": 5.416065228943889e-09, + "loss": 0.6708, + "step": 10836 + }, + { + "epoch": 0.9773188438472291, + "grad_norm": 1.7057386868820281, + "learning_rate": 5.373185090482568e-09, + "loss": 0.6933, + "step": 10837 + }, + { + "epoch": 0.9774090273706993, + "grad_norm": 1.6592449963465314, + "learning_rate": 5.330475144579516e-09, + "loss": 0.7191, + "step": 10838 + }, + { + "epoch": 0.9774992108941697, + "grad_norm": 1.4490107055384425, + "learning_rate": 5.2879353948787065e-09, + "loss": 0.8264, + "step": 10839 + }, + { + "epoch": 0.9775893944176399, + "grad_norm": 2.035310062153377, + "learning_rate": 5.245565845010125e-09, + "loss": 0.7988, + "step": 10840 + }, + { + "epoch": 0.9776795779411102, + "grad_norm": 1.4763523335724602, + "learning_rate": 5.2033664985886575e-09, + "loss": 0.6713, + "step": 10841 + }, + { + "epoch": 0.9777697614645804, + "grad_norm": 1.6927489134007974, + "learning_rate": 5.161337359215201e-09, + "loss": 0.8827, + "step": 10842 + }, + { + "epoch": 0.9778599449880507, + "grad_norm": 1.9069270823219988, + "learning_rate": 5.119478430475999e-09, + "loss": 0.7516, + "step": 10843 + }, + { + "epoch": 0.977950128511521, + "grad_norm": 1.9759608623397102, + "learning_rate": 5.077789715942416e-09, + "loss": 0.6881, + "step": 10844 + }, + { + "epoch": 0.9780403120349912, + "grad_norm": 2.0825024155766467, + "learning_rate": 5.036271219171606e-09, + "loss": 0.7641, + "step": 10845 + }, + { + "epoch": 0.9781304955584614, + "grad_norm": 4.777327979647417, + "learning_rate": 4.994922943706514e-09, + "loss": 0.7766, + "step": 10846 + }, + { + "epoch": 0.9782206790819318, + "grad_norm": 1.6546479160087648, + "learning_rate": 4.953744893074763e-09, + "loss": 0.7197, + "step": 10847 + }, + { + "epoch": 0.978310862605402, + "grad_norm": 2.379892630198769, + "learning_rate": 4.912737070789985e-09, + "loss": 0.6671, + "step": 10848 + }, + { + "epoch": 0.9784010461288722, + "grad_norm": 1.7104222466783965, + "learning_rate": 4.871899480351604e-09, + "loss": 0.8161, + "step": 10849 + }, + { + "epoch": 0.9784912296523425, + "grad_norm": 1.645380014046907, + "learning_rate": 4.831232125243501e-09, + "loss": 0.7555, + "step": 10850 + }, + { + "epoch": 0.9785814131758128, + "grad_norm": 2.6032788912379683, + "learning_rate": 4.7907350089360086e-09, + "loss": 0.8421, + "step": 10851 + }, + { + "epoch": 0.978671596699283, + "grad_norm": 1.676623721751425, + "learning_rate": 4.750408134884365e-09, + "loss": 0.6383, + "step": 10852 + }, + { + "epoch": 0.9787617802227533, + "grad_norm": 1.6367325884614115, + "learning_rate": 4.710251506529816e-09, + "loss": 0.6494, + "step": 10853 + }, + { + "epoch": 0.9788519637462235, + "grad_norm": 1.5424827099886773, + "learning_rate": 4.6702651272982894e-09, + "loss": 0.7559, + "step": 10854 + }, + { + "epoch": 0.9789421472696939, + "grad_norm": 1.6682745409783541, + "learning_rate": 4.630449000602166e-09, + "loss": 0.7148, + "step": 10855 + }, + { + "epoch": 0.9790323307931641, + "grad_norm": 2.7187278132793975, + "learning_rate": 4.590803129838283e-09, + "loss": 0.8061, + "step": 10856 + }, + { + "epoch": 0.9791225143166343, + "grad_norm": 1.562310550648651, + "learning_rate": 4.551327518389714e-09, + "loss": 0.7475, + "step": 10857 + }, + { + "epoch": 0.9792126978401046, + "grad_norm": 2.0062508651398785, + "learning_rate": 4.512022169624652e-09, + "loss": 0.7306, + "step": 10858 + }, + { + "epoch": 0.9793028813635749, + "grad_norm": 1.7148320712515186, + "learning_rate": 4.472887086896637e-09, + "loss": 0.7126, + "step": 10859 + }, + { + "epoch": 0.9793930648870451, + "grad_norm": 1.7837455400866948, + "learning_rate": 4.433922273545443e-09, + "loss": 0.6905, + "step": 10860 + }, + { + "epoch": 0.9794832484105154, + "grad_norm": 0.6064161702155302, + "learning_rate": 4.395127732895299e-09, + "loss": 0.5813, + "step": 10861 + }, + { + "epoch": 0.9795734319339856, + "grad_norm": 2.0932073316900937, + "learning_rate": 4.356503468256445e-09, + "loss": 0.8219, + "step": 10862 + }, + { + "epoch": 0.979663615457456, + "grad_norm": 2.079210499802806, + "learning_rate": 4.318049482924913e-09, + "loss": 0.7516, + "step": 10863 + }, + { + "epoch": 0.9797537989809262, + "grad_norm": 1.8250502461871532, + "learning_rate": 4.279765780181188e-09, + "loss": 0.7079, + "step": 10864 + }, + { + "epoch": 0.9798439825043964, + "grad_norm": 2.279250431323097, + "learning_rate": 4.241652363291992e-09, + "loss": 0.7569, + "step": 10865 + }, + { + "epoch": 0.9799341660278668, + "grad_norm": 9.945852164634347, + "learning_rate": 4.203709235509834e-09, + "loss": 0.7128, + "step": 10866 + }, + { + "epoch": 0.980024349551337, + "grad_norm": 2.3447646932746564, + "learning_rate": 4.165936400071679e-09, + "loss": 0.7587, + "step": 10867 + }, + { + "epoch": 0.9801145330748072, + "grad_norm": 1.6579230872975275, + "learning_rate": 4.12833386020095e-09, + "loss": 0.6896, + "step": 10868 + }, + { + "epoch": 0.9802047165982775, + "grad_norm": 2.5042510130054687, + "learning_rate": 4.090901619105746e-09, + "loss": 0.7807, + "step": 10869 + }, + { + "epoch": 0.9802949001217478, + "grad_norm": 1.5765347163584902, + "learning_rate": 4.053639679980181e-09, + "loss": 0.6323, + "step": 10870 + }, + { + "epoch": 0.980385083645218, + "grad_norm": 1.8081460155501283, + "learning_rate": 4.01654804600371e-09, + "loss": 0.7486, + "step": 10871 + }, + { + "epoch": 0.9804752671686883, + "grad_norm": 2.1295955556947805, + "learning_rate": 3.9796267203409114e-09, + "loss": 0.7497, + "step": 10872 + }, + { + "epoch": 0.9805654506921585, + "grad_norm": 1.4725089163906946, + "learning_rate": 3.942875706142379e-09, + "loss": 0.7177, + "step": 10873 + }, + { + "epoch": 0.9806556342156288, + "grad_norm": 2.130882535785851, + "learning_rate": 3.906295006543825e-09, + "loss": 0.801, + "step": 10874 + }, + { + "epoch": 0.9807458177390991, + "grad_norm": 1.739003783517487, + "learning_rate": 3.8698846246665305e-09, + "loss": 0.7197, + "step": 10875 + }, + { + "epoch": 0.9808360012625693, + "grad_norm": 1.8417205632926523, + "learning_rate": 3.833644563617344e-09, + "loss": 0.7434, + "step": 10876 + }, + { + "epoch": 0.9809261847860395, + "grad_norm": 3.2176163538099405, + "learning_rate": 3.797574826488237e-09, + "loss": 0.7541, + "step": 10877 + }, + { + "epoch": 0.9810163683095099, + "grad_norm": 3.6058192056958482, + "learning_rate": 3.761675416356969e-09, + "loss": 0.6681, + "step": 10878 + }, + { + "epoch": 0.9811065518329801, + "grad_norm": 1.5328290154892386, + "learning_rate": 3.725946336286867e-09, + "loss": 0.6482, + "step": 10879 + }, + { + "epoch": 0.9811967353564504, + "grad_norm": 2.0690698889188623, + "learning_rate": 3.6903875893261604e-09, + "loss": 0.7158, + "step": 10880 + }, + { + "epoch": 0.9812869188799206, + "grad_norm": 1.6577482104981414, + "learning_rate": 3.6549991785093105e-09, + "loss": 0.6726, + "step": 10881 + }, + { + "epoch": 0.9813771024033909, + "grad_norm": 2.2940394746398884, + "learning_rate": 3.6197811068554575e-09, + "loss": 0.7306, + "step": 10882 + }, + { + "epoch": 0.9814672859268612, + "grad_norm": 1.6061535045568327, + "learning_rate": 3.584733377369975e-09, + "loss": 0.8262, + "step": 10883 + }, + { + "epoch": 0.9815574694503314, + "grad_norm": 5.295078209145432, + "learning_rate": 3.549855993043138e-09, + "loss": 0.7227, + "step": 10884 + }, + { + "epoch": 0.9816476529738016, + "grad_norm": 1.7860938350039073, + "learning_rate": 3.5151489568507887e-09, + "loss": 0.7235, + "step": 10885 + }, + { + "epoch": 0.981737836497272, + "grad_norm": 2.060975291474668, + "learning_rate": 3.4806122717545572e-09, + "loss": 0.7572, + "step": 10886 + }, + { + "epoch": 0.9818280200207422, + "grad_norm": 1.7461632494994814, + "learning_rate": 3.446245940701198e-09, + "loss": 0.7034, + "step": 10887 + }, + { + "epoch": 0.9819182035442124, + "grad_norm": 1.4830490614993446, + "learning_rate": 3.41204996662281e-09, + "loss": 0.6799, + "step": 10888 + }, + { + "epoch": 0.9820083870676828, + "grad_norm": 1.352125061794074, + "learning_rate": 3.3780243524375028e-09, + "loss": 0.7367, + "step": 10889 + }, + { + "epoch": 0.982098570591153, + "grad_norm": 1.7363411009296146, + "learning_rate": 3.3441691010485107e-09, + "loss": 0.7674, + "step": 10890 + }, + { + "epoch": 0.9821887541146233, + "grad_norm": 1.7068797028964613, + "learning_rate": 3.3104842153444113e-09, + "loss": 0.8164, + "step": 10891 + }, + { + "epoch": 0.9822789376380935, + "grad_norm": 1.7088444418157152, + "learning_rate": 3.27696969819935e-09, + "loss": 0.736, + "step": 10892 + }, + { + "epoch": 0.9823691211615638, + "grad_norm": 1.3714084535780762, + "learning_rate": 3.2436255524732615e-09, + "loss": 0.6951, + "step": 10893 + }, + { + "epoch": 0.9824593046850341, + "grad_norm": 1.87713812671529, + "learning_rate": 3.210451781010759e-09, + "loss": 0.7212, + "step": 10894 + }, + { + "epoch": 0.9825494882085043, + "grad_norm": 4.085270843678658, + "learning_rate": 3.1774483866426895e-09, + "loss": 0.6031, + "step": 10895 + }, + { + "epoch": 0.9826396717319745, + "grad_norm": 1.6827452041360411, + "learning_rate": 3.144615372185244e-09, + "loss": 0.7063, + "step": 10896 + }, + { + "epoch": 0.9827298552554449, + "grad_norm": 1.5584950810227223, + "learning_rate": 3.1119527404399604e-09, + "loss": 0.83, + "step": 10897 + }, + { + "epoch": 0.9828200387789151, + "grad_norm": 2.798681943474024, + "learning_rate": 3.0794604941932754e-09, + "loss": 0.7427, + "step": 10898 + }, + { + "epoch": 0.9829102223023853, + "grad_norm": 1.6002770452253634, + "learning_rate": 3.0471386362180827e-09, + "loss": 0.6404, + "step": 10899 + }, + { + "epoch": 0.9830004058258556, + "grad_norm": 3.529951143656116, + "learning_rate": 3.0149871692719542e-09, + "loss": 0.7912, + "step": 10900 + }, + { + "epoch": 0.9830905893493259, + "grad_norm": 1.4290831075250665, + "learning_rate": 2.9830060960984728e-09, + "loss": 0.7255, + "step": 10901 + }, + { + "epoch": 0.9831807728727961, + "grad_norm": 0.7025350426862991, + "learning_rate": 2.9511954194263442e-09, + "loss": 0.5842, + "step": 10902 + }, + { + "epoch": 0.9832709563962664, + "grad_norm": 1.9453218145798585, + "learning_rate": 2.9195551419698426e-09, + "loss": 0.7411, + "step": 10903 + }, + { + "epoch": 0.9833611399197366, + "grad_norm": 9.30486583779032, + "learning_rate": 2.888085266428808e-09, + "loss": 0.7013, + "step": 10904 + }, + { + "epoch": 0.983451323443207, + "grad_norm": 2.1352836629493854, + "learning_rate": 2.8567857954882037e-09, + "loss": 0.697, + "step": 10905 + }, + { + "epoch": 0.9835415069666772, + "grad_norm": 1.724727363049713, + "learning_rate": 2.82565673181856e-09, + "loss": 0.7939, + "step": 10906 + }, + { + "epoch": 0.9836316904901474, + "grad_norm": 1.7111493680580014, + "learning_rate": 2.7946980780764184e-09, + "loss": 0.7604, + "step": 10907 + }, + { + "epoch": 0.9837218740136177, + "grad_norm": 1.9099656460453645, + "learning_rate": 2.763909836903e-09, + "loss": 0.7975, + "step": 10908 + }, + { + "epoch": 0.983812057537088, + "grad_norm": 1.63821270116652, + "learning_rate": 2.7332920109255364e-09, + "loss": 0.6547, + "step": 10909 + }, + { + "epoch": 0.9839022410605582, + "grad_norm": 1.5839570556994047, + "learning_rate": 2.702844602756382e-09, + "loss": 0.7188, + "step": 10910 + }, + { + "epoch": 0.9839924245840285, + "grad_norm": 2.319545220820906, + "learning_rate": 2.6725676149936814e-09, + "loss": 0.7201, + "step": 10911 + }, + { + "epoch": 0.9840826081074988, + "grad_norm": 1.8305952548615387, + "learning_rate": 2.642461050220479e-09, + "loss": 0.8057, + "step": 10912 + }, + { + "epoch": 0.984172791630969, + "grad_norm": 1.4865178752790595, + "learning_rate": 2.612524911005609e-09, + "loss": 0.7224, + "step": 10913 + }, + { + "epoch": 0.9842629751544393, + "grad_norm": 1.5606357897659338, + "learning_rate": 2.582759199903917e-09, + "loss": 0.6157, + "step": 10914 + }, + { + "epoch": 0.9843531586779095, + "grad_norm": 1.5910744742100886, + "learning_rate": 2.553163919454704e-09, + "loss": 0.7008, + "step": 10915 + }, + { + "epoch": 0.9844433422013799, + "grad_norm": 1.7751352527290891, + "learning_rate": 2.523739072183506e-09, + "loss": 0.7218, + "step": 10916 + }, + { + "epoch": 0.9845335257248501, + "grad_norm": 1.5937780489977778, + "learning_rate": 2.4944846606007597e-09, + "loss": 0.6983, + "step": 10917 + }, + { + "epoch": 0.9846237092483203, + "grad_norm": 0.6148667220946401, + "learning_rate": 2.46540068720269e-09, + "loss": 0.6153, + "step": 10918 + }, + { + "epoch": 0.9847138927717906, + "grad_norm": 3.172243895478004, + "learning_rate": 2.4364871544708674e-09, + "loss": 0.7792, + "step": 10919 + }, + { + "epoch": 0.9848040762952609, + "grad_norm": 2.225852451544658, + "learning_rate": 2.4077440648726523e-09, + "loss": 0.6886, + "step": 10920 + }, + { + "epoch": 0.9848942598187311, + "grad_norm": 1.7896540301047275, + "learning_rate": 2.379171420860082e-09, + "loss": 0.8055, + "step": 10921 + }, + { + "epoch": 0.9849844433422014, + "grad_norm": 2.2736821518181336, + "learning_rate": 2.3507692248714296e-09, + "loss": 0.6619, + "step": 10922 + }, + { + "epoch": 0.9850746268656716, + "grad_norm": 1.7295402291861177, + "learning_rate": 2.322537479330089e-09, + "loss": 0.7519, + "step": 10923 + }, + { + "epoch": 0.9851648103891419, + "grad_norm": 1.6034401656318058, + "learning_rate": 2.2944761866450223e-09, + "loss": 0.8156, + "step": 10924 + }, + { + "epoch": 0.9852549939126122, + "grad_norm": 2.711261740870576, + "learning_rate": 2.266585349210315e-09, + "loss": 0.7231, + "step": 10925 + }, + { + "epoch": 0.9853451774360824, + "grad_norm": 2.3548285530029753, + "learning_rate": 2.2388649694060623e-09, + "loss": 0.6987, + "step": 10926 + }, + { + "epoch": 0.9854353609595526, + "grad_norm": 1.759859105886264, + "learning_rate": 2.211315049597262e-09, + "loss": 0.8, + "step": 10927 + }, + { + "epoch": 0.985525544483023, + "grad_norm": 1.599356793221361, + "learning_rate": 2.1839355921349224e-09, + "loss": 0.7741, + "step": 10928 + }, + { + "epoch": 0.9856157280064932, + "grad_norm": 2.0545282918653602, + "learning_rate": 2.156726599354952e-09, + "loss": 0.7307, + "step": 10929 + }, + { + "epoch": 0.9857059115299635, + "grad_norm": 1.8234716678611795, + "learning_rate": 2.129688073578828e-09, + "loss": 0.7634, + "step": 10930 + }, + { + "epoch": 0.9857960950534337, + "grad_norm": 2.3925940688244274, + "learning_rate": 2.1028200171142597e-09, + "loss": 0.8723, + "step": 10931 + }, + { + "epoch": 0.985886278576904, + "grad_norm": 2.2961762072439336, + "learning_rate": 2.076122432253191e-09, + "loss": 0.7215, + "step": 10932 + }, + { + "epoch": 0.9859764621003743, + "grad_norm": 2.3707241920418887, + "learning_rate": 2.0495953212738005e-09, + "loss": 0.7746, + "step": 10933 + }, + { + "epoch": 0.9860666456238445, + "grad_norm": 1.399453758214474, + "learning_rate": 2.0232386864396102e-09, + "loss": 0.8193, + "step": 10934 + }, + { + "epoch": 0.9861568291473148, + "grad_norm": 2.1748357118844046, + "learning_rate": 1.9970525299992656e-09, + "loss": 0.7622, + "step": 10935 + }, + { + "epoch": 0.9862470126707851, + "grad_norm": 1.9646864234626482, + "learning_rate": 1.9710368541874245e-09, + "loss": 0.7008, + "step": 10936 + }, + { + "epoch": 0.9863371961942553, + "grad_norm": 0.5725794377665201, + "learning_rate": 1.945191661223644e-09, + "loss": 0.5866, + "step": 10937 + }, + { + "epoch": 0.9864273797177255, + "grad_norm": 1.4975333595699638, + "learning_rate": 1.9195169533132714e-09, + "loss": 0.788, + "step": 10938 + }, + { + "epoch": 0.9865175632411959, + "grad_norm": 0.5882927162704472, + "learning_rate": 1.894012732646999e-09, + "loss": 0.5937, + "step": 10939 + }, + { + "epoch": 0.9866077467646661, + "grad_norm": 2.180133842244296, + "learning_rate": 1.8686790014010854e-09, + "loss": 0.6916, + "step": 10940 + }, + { + "epoch": 0.9866979302881363, + "grad_norm": 0.6039181450268233, + "learning_rate": 1.8435157617369134e-09, + "loss": 0.5589, + "step": 10941 + }, + { + "epoch": 0.9867881138116066, + "grad_norm": 1.6782300436719133, + "learning_rate": 1.818523015801876e-09, + "loss": 0.8787, + "step": 10942 + }, + { + "epoch": 0.9868782973350769, + "grad_norm": 1.901443236975038, + "learning_rate": 1.7937007657282677e-09, + "loss": 0.7003, + "step": 10943 + }, + { + "epoch": 0.9869684808585472, + "grad_norm": 1.9950574360851892, + "learning_rate": 1.7690490136341718e-09, + "loss": 0.685, + "step": 10944 + }, + { + "epoch": 0.9870586643820174, + "grad_norm": 1.8250007341984356, + "learning_rate": 1.744567761622795e-09, + "loss": 0.7371, + "step": 10945 + }, + { + "epoch": 0.9871488479054876, + "grad_norm": 1.6770867373350993, + "learning_rate": 1.7202570117831327e-09, + "loss": 0.7271, + "step": 10946 + }, + { + "epoch": 0.987239031428958, + "grad_norm": 2.489400135998505, + "learning_rate": 1.696116766189526e-09, + "loss": 0.7189, + "step": 10947 + }, + { + "epoch": 0.9873292149524282, + "grad_norm": 8.794480534881975, + "learning_rate": 1.6721470269021042e-09, + "loss": 0.7815, + "step": 10948 + }, + { + "epoch": 0.9874193984758984, + "grad_norm": 2.0191886212767387, + "learning_rate": 1.6483477959654546e-09, + "loss": 0.7094, + "step": 10949 + }, + { + "epoch": 0.9875095819993687, + "grad_norm": 1.850999126757679, + "learning_rate": 1.6247190754106187e-09, + "loss": 0.7208, + "step": 10950 + }, + { + "epoch": 0.987599765522839, + "grad_norm": 1.4416916314023405, + "learning_rate": 1.6012608672537619e-09, + "loss": 0.7052, + "step": 10951 + }, + { + "epoch": 0.9876899490463092, + "grad_norm": 3.390206306728996, + "learning_rate": 1.5779731734963942e-09, + "loss": 0.7501, + "step": 10952 + }, + { + "epoch": 0.9877801325697795, + "grad_norm": 16.449375054915176, + "learning_rate": 1.5548559961253705e-09, + "loss": 0.8042, + "step": 10953 + }, + { + "epoch": 0.9878703160932497, + "grad_norm": 1.6388717729497286, + "learning_rate": 1.5319093371135573e-09, + "loss": 0.6655, + "step": 10954 + }, + { + "epoch": 0.9879604996167201, + "grad_norm": 1.6509729926651633, + "learning_rate": 1.5091331984184997e-09, + "loss": 0.8263, + "step": 10955 + }, + { + "epoch": 0.9880506831401903, + "grad_norm": 1.5346685901704915, + "learning_rate": 1.486527581983754e-09, + "loss": 0.735, + "step": 10956 + }, + { + "epoch": 0.9881408666636605, + "grad_norm": 2.990982012032381, + "learning_rate": 1.4640924897382223e-09, + "loss": 0.8026, + "step": 10957 + }, + { + "epoch": 0.9882310501871309, + "grad_norm": 1.5502667474406508, + "learning_rate": 1.4418279235961506e-09, + "loss": 0.78, + "step": 10958 + }, + { + "epoch": 0.9883212337106011, + "grad_norm": 1.7147745105433767, + "learning_rate": 1.4197338854573526e-09, + "loss": 0.6441, + "step": 10959 + }, + { + "epoch": 0.9884114172340713, + "grad_norm": 1.9126335571129551, + "learning_rate": 1.3978103772067651e-09, + "loss": 0.6974, + "step": 10960 + }, + { + "epoch": 0.9885016007575416, + "grad_norm": 1.864346245835838, + "learning_rate": 1.3760574007153358e-09, + "loss": 0.6603, + "step": 10961 + }, + { + "epoch": 0.9885917842810119, + "grad_norm": 2.390197516224327, + "learning_rate": 1.3544749578389137e-09, + "loss": 0.8192, + "step": 10962 + }, + { + "epoch": 0.9886819678044821, + "grad_norm": 1.6184299874375352, + "learning_rate": 1.3330630504189143e-09, + "loss": 0.7921, + "step": 10963 + }, + { + "epoch": 0.9887721513279524, + "grad_norm": 1.953329212956862, + "learning_rate": 1.3118216802827652e-09, + "loss": 0.7068, + "step": 10964 + }, + { + "epoch": 0.9888623348514226, + "grad_norm": 2.05256061199614, + "learning_rate": 1.2907508492425722e-09, + "loss": 0.7096, + "step": 10965 + }, + { + "epoch": 0.988952518374893, + "grad_norm": 1.7610216474356062, + "learning_rate": 1.2698505590962305e-09, + "loss": 0.7026, + "step": 10966 + }, + { + "epoch": 0.9890427018983632, + "grad_norm": 1.6598972971295007, + "learning_rate": 1.2491208116272022e-09, + "loss": 0.7372, + "step": 10967 + }, + { + "epoch": 0.9891328854218334, + "grad_norm": 1.7804664092908455, + "learning_rate": 1.2285616086040728e-09, + "loss": 0.7719, + "step": 10968 + }, + { + "epoch": 0.9892230689453037, + "grad_norm": 1.3741372609659843, + "learning_rate": 1.2081729517812167e-09, + "loss": 0.8028, + "step": 10969 + }, + { + "epoch": 0.989313252468774, + "grad_norm": 1.5744842229726828, + "learning_rate": 1.1879548428983533e-09, + "loss": 0.7382, + "step": 10970 + }, + { + "epoch": 0.9894034359922442, + "grad_norm": 1.6779851759300108, + "learning_rate": 1.167907283680547e-09, + "loss": 0.6901, + "step": 10971 + }, + { + "epoch": 0.9894936195157145, + "grad_norm": 1.5785808848989833, + "learning_rate": 1.1480302758382077e-09, + "loss": 0.6806, + "step": 10972 + }, + { + "epoch": 0.9895838030391847, + "grad_norm": 1.6010758567771433, + "learning_rate": 1.1283238210675338e-09, + "loss": 0.7297, + "step": 10973 + }, + { + "epoch": 0.989673986562655, + "grad_norm": 2.4648308851870593, + "learning_rate": 1.1087879210498474e-09, + "loss": 0.6691, + "step": 10974 + }, + { + "epoch": 0.9897641700861253, + "grad_norm": 1.6413242847979286, + "learning_rate": 1.0894225774522592e-09, + "loss": 0.7855, + "step": 10975 + }, + { + "epoch": 0.9898543536095955, + "grad_norm": 0.6960678712292174, + "learning_rate": 1.070227791927003e-09, + "loss": 0.5953, + "step": 10976 + }, + { + "epoch": 0.9899445371330657, + "grad_norm": 3.2196331108954657, + "learning_rate": 1.0512035661118802e-09, + "loss": 0.8404, + "step": 10977 + }, + { + "epoch": 0.9900347206565361, + "grad_norm": 2.067619328220716, + "learning_rate": 1.0323499016300364e-09, + "loss": 0.761, + "step": 10978 + }, + { + "epoch": 0.9901249041800063, + "grad_norm": 1.7667894152817591, + "learning_rate": 1.013666800090407e-09, + "loss": 0.6918, + "step": 10979 + }, + { + "epoch": 0.9902150877034765, + "grad_norm": 8.16153426014326, + "learning_rate": 9.951542630870502e-10, + "loss": 0.724, + "step": 10980 + }, + { + "epoch": 0.9903052712269468, + "grad_norm": 2.056028791315081, + "learning_rate": 9.768122921995915e-10, + "loss": 0.78, + "step": 10981 + }, + { + "epoch": 0.9903954547504171, + "grad_norm": 2.143559247752837, + "learning_rate": 9.58640888992779e-10, + "loss": 0.6807, + "step": 10982 + }, + { + "epoch": 0.9904856382738874, + "grad_norm": 1.7149895191150406, + "learning_rate": 9.40640055017594e-10, + "loss": 0.6549, + "step": 10983 + }, + { + "epoch": 0.9905758217973576, + "grad_norm": 1.622834817228605, + "learning_rate": 9.228097918094757e-10, + "loss": 0.7401, + "step": 10984 + }, + { + "epoch": 0.9906660053208279, + "grad_norm": 1.9520071874509657, + "learning_rate": 9.051501008900952e-10, + "loss": 0.7115, + "step": 10985 + }, + { + "epoch": 0.9907561888442982, + "grad_norm": 2.115787057008684, + "learning_rate": 8.876609837662475e-10, + "loss": 0.7544, + "step": 10986 + }, + { + "epoch": 0.9908463723677684, + "grad_norm": 1.4885380240578883, + "learning_rate": 8.70342441930294e-10, + "loss": 0.8266, + "step": 10987 + }, + { + "epoch": 0.9909365558912386, + "grad_norm": 1.7601122922187329, + "learning_rate": 8.531944768594979e-10, + "loss": 0.71, + "step": 10988 + }, + { + "epoch": 0.991026739414709, + "grad_norm": 3.1405049011648942, + "learning_rate": 8.362170900175769e-10, + "loss": 0.8199, + "step": 10989 + }, + { + "epoch": 0.9911169229381792, + "grad_norm": 7.27508074004474, + "learning_rate": 8.194102828527061e-10, + "loss": 0.697, + "step": 10990 + }, + { + "epoch": 0.9912071064616494, + "grad_norm": 1.7379634266594928, + "learning_rate": 8.027740567992936e-10, + "loss": 0.7826, + "step": 10991 + }, + { + "epoch": 0.9912972899851197, + "grad_norm": 1.5112236848595717, + "learning_rate": 7.863084132766484e-10, + "loss": 0.7373, + "step": 10992 + }, + { + "epoch": 0.99138747350859, + "grad_norm": 1.6801104345809221, + "learning_rate": 7.700133536896469e-10, + "loss": 0.7556, + "step": 10993 + }, + { + "epoch": 0.9914776570320603, + "grad_norm": 1.5012617367707306, + "learning_rate": 7.538888794287324e-10, + "loss": 0.7104, + "step": 10994 + }, + { + "epoch": 0.9915678405555305, + "grad_norm": 1.5444484134702354, + "learning_rate": 7.379349918696931e-10, + "loss": 0.7621, + "step": 10995 + }, + { + "epoch": 0.9916580240790007, + "grad_norm": 1.8276799630244926, + "learning_rate": 7.221516923738846e-10, + "loss": 0.715, + "step": 10996 + }, + { + "epoch": 0.9917482076024711, + "grad_norm": 2.075874544171306, + "learning_rate": 7.065389822880075e-10, + "loss": 0.8462, + "step": 10997 + }, + { + "epoch": 0.9918383911259413, + "grad_norm": 3.246765875061049, + "learning_rate": 6.910968629443292e-10, + "loss": 0.8174, + "step": 10998 + }, + { + "epoch": 0.9919285746494115, + "grad_norm": 1.8482880429542856, + "learning_rate": 6.758253356602406e-10, + "loss": 0.7359, + "step": 10999 + }, + { + "epoch": 0.9920187581728818, + "grad_norm": 1.5529229967244007, + "learning_rate": 6.607244017389213e-10, + "loss": 0.8321, + "step": 11000 + }, + { + "epoch": 0.9921089416963521, + "grad_norm": 2.5034358256659064, + "learning_rate": 6.457940624686742e-10, + "loss": 0.7616, + "step": 11001 + }, + { + "epoch": 0.9921991252198223, + "grad_norm": 1.9188651109393027, + "learning_rate": 6.310343191238132e-10, + "loss": 0.8215, + "step": 11002 + }, + { + "epoch": 0.9922893087432926, + "grad_norm": 2.1598319556954566, + "learning_rate": 6.164451729635534e-10, + "loss": 0.6369, + "step": 11003 + }, + { + "epoch": 0.9923794922667628, + "grad_norm": 2.431816125381106, + "learning_rate": 6.020266252324546e-10, + "loss": 0.6157, + "step": 11004 + }, + { + "epoch": 0.9924696757902332, + "grad_norm": 1.719567745918533, + "learning_rate": 5.877786771610882e-10, + "loss": 0.7122, + "step": 11005 + }, + { + "epoch": 0.9925598593137034, + "grad_norm": 0.7498214827114977, + "learning_rate": 5.737013299651483e-10, + "loss": 0.6468, + "step": 11006 + }, + { + "epoch": 0.9926500428371736, + "grad_norm": 1.718025956669432, + "learning_rate": 5.597945848458963e-10, + "loss": 0.7439, + "step": 11007 + }, + { + "epoch": 0.992740226360644, + "grad_norm": 1.6026969055511502, + "learning_rate": 5.460584429894944e-10, + "loss": 0.7118, + "step": 11008 + }, + { + "epoch": 0.9928304098841142, + "grad_norm": 3.4075961471505387, + "learning_rate": 5.32492905568338e-10, + "loss": 0.7428, + "step": 11009 + }, + { + "epoch": 0.9929205934075844, + "grad_norm": 1.768433961165373, + "learning_rate": 5.190979737399459e-10, + "loss": 0.7758, + "step": 11010 + }, + { + "epoch": 0.9930107769310547, + "grad_norm": 1.4809893092993727, + "learning_rate": 5.058736486469594e-10, + "loss": 0.6298, + "step": 11011 + }, + { + "epoch": 0.993100960454525, + "grad_norm": 1.810364747994854, + "learning_rate": 4.928199314180314e-10, + "loss": 0.771, + "step": 11012 + }, + { + "epoch": 0.9931911439779952, + "grad_norm": 2.46605303279905, + "learning_rate": 4.799368231669376e-10, + "loss": 0.6588, + "step": 11013 + }, + { + "epoch": 0.9932813275014655, + "grad_norm": 2.061229198714046, + "learning_rate": 4.672243249927988e-10, + "loss": 0.7297, + "step": 11014 + }, + { + "epoch": 0.9933715110249357, + "grad_norm": 1.5755100302417804, + "learning_rate": 4.546824379803027e-10, + "loss": 0.7684, + "step": 11015 + }, + { + "epoch": 0.993461694548406, + "grad_norm": 2.229662854834342, + "learning_rate": 4.4231116319970454e-10, + "loss": 0.7653, + "step": 11016 + }, + { + "epoch": 0.9935518780718763, + "grad_norm": 1.894435290236495, + "learning_rate": 4.3011050170660423e-10, + "loss": 0.7452, + "step": 11017 + }, + { + "epoch": 0.9936420615953465, + "grad_norm": 2.047780775630415, + "learning_rate": 4.18080454542169e-10, + "loss": 0.7346, + "step": 11018 + }, + { + "epoch": 0.9937322451188167, + "grad_norm": 0.6637858768328998, + "learning_rate": 4.0622102273246694e-10, + "loss": 0.5764, + "step": 11019 + }, + { + "epoch": 0.9938224286422871, + "grad_norm": 2.3251051807197696, + "learning_rate": 3.945322072897994e-10, + "loss": 0.7986, + "step": 11020 + }, + { + "epoch": 0.9939126121657573, + "grad_norm": 3.665058317878408, + "learning_rate": 3.830140092111467e-10, + "loss": 0.7969, + "step": 11021 + }, + { + "epoch": 0.9940027956892276, + "grad_norm": 2.0885763442917424, + "learning_rate": 3.7166642947972225e-10, + "loss": 0.7409, + "step": 11022 + }, + { + "epoch": 0.9940929792126978, + "grad_norm": 1.9752998390027356, + "learning_rate": 3.604894690634186e-10, + "loss": 0.7172, + "step": 11023 + }, + { + "epoch": 0.9941831627361681, + "grad_norm": 6.669215223822077, + "learning_rate": 3.494831289161393e-10, + "loss": 0.8261, + "step": 11024 + }, + { + "epoch": 0.9942733462596384, + "grad_norm": 1.6077599430669491, + "learning_rate": 3.3864740997668897e-10, + "loss": 0.7749, + "step": 11025 + }, + { + "epoch": 0.9943635297831086, + "grad_norm": 1.5590965588111132, + "learning_rate": 3.279823131701054e-10, + "loss": 0.7638, + "step": 11026 + }, + { + "epoch": 0.9944537133065788, + "grad_norm": 2.242018955820958, + "learning_rate": 3.1748783940610536e-10, + "loss": 0.6963, + "step": 11027 + }, + { + "epoch": 0.9945438968300492, + "grad_norm": 1.5928932579902233, + "learning_rate": 3.071639895801947e-10, + "loss": 0.703, + "step": 11028 + }, + { + "epoch": 0.9946340803535194, + "grad_norm": 1.8864102634629585, + "learning_rate": 2.9701076457322447e-10, + "loss": 0.806, + "step": 11029 + }, + { + "epoch": 0.9947242638769896, + "grad_norm": 2.057433871090668, + "learning_rate": 2.870281652513906e-10, + "loss": 0.781, + "step": 11030 + }, + { + "epoch": 0.99481444740046, + "grad_norm": 1.76422955217873, + "learning_rate": 2.772161924669003e-10, + "loss": 0.7278, + "step": 11031 + }, + { + "epoch": 0.9949046309239302, + "grad_norm": 1.8716929199107668, + "learning_rate": 2.6757484705641765e-10, + "loss": 0.7257, + "step": 11032 + }, + { + "epoch": 0.9949948144474005, + "grad_norm": 1.872316729307563, + "learning_rate": 2.58104129843062e-10, + "loss": 0.7434, + "step": 11033 + }, + { + "epoch": 0.9950849979708707, + "grad_norm": 1.96596445084852, + "learning_rate": 2.4880404163463154e-10, + "loss": 0.6951, + "step": 11034 + }, + { + "epoch": 0.995175181494341, + "grad_norm": 1.9286616142822997, + "learning_rate": 2.3967458322471377e-10, + "loss": 0.7297, + "step": 11035 + }, + { + "epoch": 0.9952653650178113, + "grad_norm": 1.590921847180835, + "learning_rate": 2.307157553922412e-10, + "loss": 0.7539, + "step": 11036 + }, + { + "epoch": 0.9953555485412815, + "grad_norm": 1.4965913507815454, + "learning_rate": 2.2192755890193538e-10, + "loss": 0.7818, + "step": 11037 + }, + { + "epoch": 0.9954457320647517, + "grad_norm": 0.6218019682206751, + "learning_rate": 2.133099945034189e-10, + "loss": 0.546, + "step": 11038 + }, + { + "epoch": 0.9955359155882221, + "grad_norm": 2.8135717257367863, + "learning_rate": 2.048630629318815e-10, + "loss": 0.7628, + "step": 11039 + }, + { + "epoch": 0.9956260991116923, + "grad_norm": 1.4040218503099466, + "learning_rate": 1.965867649080799e-10, + "loss": 0.7359, + "step": 11040 + }, + { + "epoch": 0.9957162826351625, + "grad_norm": 0.6129166558051956, + "learning_rate": 1.8848110113856008e-10, + "loss": 0.5637, + "step": 11041 + }, + { + "epoch": 0.9958064661586328, + "grad_norm": 1.6625131190649993, + "learning_rate": 1.8054607231454687e-10, + "loss": 0.78, + "step": 11042 + }, + { + "epoch": 0.9958966496821031, + "grad_norm": 1.5638271798802144, + "learning_rate": 1.7278167911327635e-10, + "loss": 0.6959, + "step": 11043 + }, + { + "epoch": 0.9959868332055734, + "grad_norm": 1.6053699175261067, + "learning_rate": 1.6518792219710753e-10, + "loss": 0.7748, + "step": 11044 + }, + { + "epoch": 0.9960770167290436, + "grad_norm": 1.6930098827454405, + "learning_rate": 1.5776480221418865e-10, + "loss": 0.7465, + "step": 11045 + }, + { + "epoch": 0.9961672002525138, + "grad_norm": 3.9131547861347147, + "learning_rate": 1.505123197977909e-10, + "loss": 0.7448, + "step": 11046 + }, + { + "epoch": 0.9962573837759842, + "grad_norm": 1.724372061173526, + "learning_rate": 1.4343047556675258e-10, + "loss": 0.678, + "step": 11047 + }, + { + "epoch": 0.9963475672994544, + "grad_norm": 2.2647573980296283, + "learning_rate": 1.3651927012503506e-10, + "loss": 0.7437, + "step": 11048 + }, + { + "epoch": 0.9964377508229246, + "grad_norm": 1.8777323840872533, + "learning_rate": 1.297787040630549e-10, + "loss": 0.7462, + "step": 11049 + }, + { + "epoch": 0.9965279343463949, + "grad_norm": 2.683422810581046, + "learning_rate": 1.2320877795524153e-10, + "loss": 0.7676, + "step": 11050 + }, + { + "epoch": 0.9966181178698652, + "grad_norm": 2.016188668384235, + "learning_rate": 1.1680949236247962e-10, + "loss": 0.7245, + "step": 11051 + }, + { + "epoch": 0.9967083013933354, + "grad_norm": 1.57868600070773, + "learning_rate": 1.1058084783099886e-10, + "loss": 0.7477, + "step": 11052 + }, + { + "epoch": 0.9967984849168057, + "grad_norm": 2.211507339641303, + "learning_rate": 1.0452284489170793e-10, + "loss": 0.7533, + "step": 11053 + }, + { + "epoch": 0.996888668440276, + "grad_norm": 2.184767515091988, + "learning_rate": 9.86354840621928e-11, + "loss": 0.7262, + "step": 11054 + }, + { + "epoch": 0.9969788519637462, + "grad_norm": 3.4887947823723167, + "learning_rate": 9.291876584427427e-11, + "loss": 0.6192, + "step": 11055 + }, + { + "epoch": 0.9970690354872165, + "grad_norm": 2.131693340290717, + "learning_rate": 8.737269072578435e-11, + "loss": 0.7135, + "step": 11056 + }, + { + "epoch": 0.9971592190106867, + "grad_norm": 1.4994367495658116, + "learning_rate": 8.199725918012212e-11, + "loss": 0.7389, + "step": 11057 + }, + { + "epoch": 0.9972494025341571, + "grad_norm": 1.6607929758979938, + "learning_rate": 7.679247166603175e-11, + "loss": 0.7444, + "step": 11058 + }, + { + "epoch": 0.9973395860576273, + "grad_norm": 1.844617477915555, + "learning_rate": 7.17583286273804e-11, + "loss": 0.7445, + "step": 11059 + }, + { + "epoch": 0.9974297695810975, + "grad_norm": 1.809757854956306, + "learning_rate": 6.689483049360233e-11, + "loss": 0.7657, + "step": 11060 + }, + { + "epoch": 0.9975199531045678, + "grad_norm": 2.3715028614121025, + "learning_rate": 6.220197768014302e-11, + "loss": 0.715, + "step": 11061 + }, + { + "epoch": 0.9976101366280381, + "grad_norm": 1.999793876516644, + "learning_rate": 5.7679770587126806e-11, + "loss": 0.7602, + "step": 11062 + }, + { + "epoch": 0.9977003201515083, + "grad_norm": 0.6164481164520272, + "learning_rate": 5.33282096002452e-11, + "loss": 0.6658, + "step": 11063 + }, + { + "epoch": 0.9977905036749786, + "grad_norm": 1.9915265950489418, + "learning_rate": 4.914729509120086e-11, + "loss": 0.7134, + "step": 11064 + }, + { + "epoch": 0.9978806871984488, + "grad_norm": 1.4740987740008686, + "learning_rate": 4.513702741637537e-11, + "loss": 0.6669, + "step": 11065 + }, + { + "epoch": 0.9979708707219191, + "grad_norm": 3.2393231711638113, + "learning_rate": 4.129740691816153e-11, + "loss": 0.8042, + "step": 11066 + }, + { + "epoch": 0.9980610542453894, + "grad_norm": 1.3799733039789281, + "learning_rate": 3.762843392429715e-11, + "loss": 0.7601, + "step": 11067 + }, + { + "epoch": 0.9981512377688596, + "grad_norm": 1.8508771440776053, + "learning_rate": 3.413010874742106e-11, + "loss": 0.7199, + "step": 11068 + }, + { + "epoch": 0.9982414212923298, + "grad_norm": 2.778526344101661, + "learning_rate": 3.080243168618324e-11, + "loss": 0.7348, + "step": 11069 + }, + { + "epoch": 0.9983316048158002, + "grad_norm": 1.915173079933426, + "learning_rate": 2.7645403024800783e-11, + "loss": 0.6885, + "step": 11070 + }, + { + "epoch": 0.9984217883392704, + "grad_norm": 1.357126239445417, + "learning_rate": 2.4659023032391756e-11, + "loss": 0.8092, + "step": 11071 + }, + { + "epoch": 0.9985119718627407, + "grad_norm": 1.6403973376035073, + "learning_rate": 2.1843291963863364e-11, + "loss": 0.7861, + "step": 11072 + }, + { + "epoch": 0.9986021553862109, + "grad_norm": 1.6125888770398173, + "learning_rate": 1.9198210059245822e-11, + "loss": 0.7571, + "step": 11073 + }, + { + "epoch": 0.9986923389096812, + "grad_norm": 1.9919347864568018, + "learning_rate": 1.672377754458054e-11, + "loss": 0.7436, + "step": 11074 + }, + { + "epoch": 0.9987825224331515, + "grad_norm": 20.00877188006394, + "learning_rate": 1.4419994630809895e-11, + "loss": 0.7323, + "step": 11075 + }, + { + "epoch": 0.9988727059566217, + "grad_norm": 2.1621274594151094, + "learning_rate": 1.2286861514443358e-11, + "loss": 0.6932, + "step": 11076 + }, + { + "epoch": 0.998962889480092, + "grad_norm": 1.3248851078532171, + "learning_rate": 1.0324378377779553e-11, + "loss": 0.7789, + "step": 11077 + }, + { + "epoch": 0.9990530730035623, + "grad_norm": 2.974754284881345, + "learning_rate": 8.532545388018064e-12, + "loss": 0.7394, + "step": 11078 + }, + { + "epoch": 0.9991432565270325, + "grad_norm": 1.8211917963722504, + "learning_rate": 6.911362697925582e-12, + "loss": 0.5767, + "step": 11079 + }, + { + "epoch": 0.9992334400505027, + "grad_norm": 2.0852892150447815, + "learning_rate": 5.46083044605794e-12, + "loss": 0.717, + "step": 11080 + }, + { + "epoch": 0.9993236235739731, + "grad_norm": 1.6458990429987688, + "learning_rate": 4.1809487563160276e-12, + "loss": 0.7922, + "step": 11081 + }, + { + "epoch": 0.9994138070974433, + "grad_norm": 2.6489534578767273, + "learning_rate": 3.0717177375017e-12, + "loss": 0.7442, + "step": 11082 + }, + { + "epoch": 0.9995039906209136, + "grad_norm": 1.8609619420878256, + "learning_rate": 2.1331374846500495e-12, + "loss": 0.7363, + "step": 11083 + }, + { + "epoch": 0.9995941741443838, + "grad_norm": 2.0582749990415974, + "learning_rate": 1.3652080774750885e-12, + "loss": 0.838, + "step": 11084 + }, + { + "epoch": 0.9996843576678541, + "grad_norm": 1.3818844628322762, + "learning_rate": 7.679295817020204e-13, + "loss": 0.6564, + "step": 11085 + }, + { + "epoch": 0.9997745411913244, + "grad_norm": 1.7941582085669492, + "learning_rate": 3.413020484011042e-13, + "loss": 0.7195, + "step": 11086 + }, + { + "epoch": 0.9998647247147946, + "grad_norm": 1.6475317758323569, + "learning_rate": 8.53255139876552e-14, + "loss": 0.7831, + "step": 11087 + }, + { + "epoch": 0.9999549082382648, + "grad_norm": 3.1586735551382463, + "learning_rate": 0.0, + "loss": 0.6846, + "step": 11088 + }, + { + "epoch": 0.9999549082382648, + "step": 11088, + "total_flos": 1.347768590492762e+18, + "train_loss": 0.5537439968522966, + "train_runtime": 116542.4248, + "train_samples_per_second": 5.709, + "train_steps_per_second": 0.095 + } + ], + "logging_steps": 1.0, + "max_steps": 11088, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 100, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.347768590492762e+18, + "train_batch_size": 5, + "trial_name": null, + "trial_params": null +} diff --git a/sft_full/smoe_sigmoidgating/training_args.bin b/sft_full/smoe_sigmoidgating/training_args.bin new file mode 100644 index 0000000000000000000000000000000000000000..127c1aa11d905a2961efa77b5c382b114609011b --- /dev/null +++ b/sft_full/smoe_sigmoidgating/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2caf1feafb197632e80e375c16b7f23835bf99a2b5d929eaf7cb8771e1cb2d4e +size 8184